TAXONOMIC AND FUNCTIONAL CHARACTERIZATION OF MICROBIAL COMMUNITIES LINKED TO CHLORINATED SOLVENT, 1,4-DIOXANE AND RDX BIODEGRADATION IN GROUNDWATER AND SOIL MICROCOSMS By Hongyu Dang A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Environmental Engineering—Doctor of Philosophy 2021 ABSTRACT TAXONOMIC AND FUNCTIONAL CHARACTERIZATION OF MICROBIAL COMMUNITIES LINKED TO CHLORINATED SOLVENT, 1,4-DIOXANE AND RDX BIODEGRADATION IN GROUNDWATER AND SOIL MICROCOSMS By Hongyu Dang Bioremediation is becoming increasing popular for the remediation of sites contaminated with a range of different contaminants. Molecular methods such as 16S rRNA gene amplicon sequencing, shotgun sequencing, and high throughput quantitative PCR offer much potential for examining the microorganisms and functional genes associated with contaminant biodegradation, which can provide critical additional lines of evidence for effective site remediation. In this work, the first project examined the taxonomic and functional biomarkers associated with chlorinated solvent and 1,4-dioxane biodegradation in groundwater from five contaminated sites. Each site had previously been bioaugmented with the commercially available dechlorinating mixed culture, SDC-9. The results highlighted the occurrence of numerous genera previously linked to chlorinated solvent biodegradation. The functional gene analysis indicated two reductive dehalogenase genes (vcrA and tceA) from Dehalococcoides mccartyi were abundant. Additionally, aerobic and anaerobic biomarkers for the biodegradation of various chlorinated compounds were observed across all sites. The approach used (shotgun sequencing) is advantageous over many other methods because an unlimited number of functional genes can be examined and a more complete picture of the functional abilities of microbial communities can be depicted. Another research project evaluated the functional genes and species associated with RDX biodegradation at a RDX contaminated Navy site where biostimulation had been adopted. For this, DNA samples extracted from groundwater samples pre- and post- biostimulation were subject to shotgun sequencing and high throughput qPCR. DNA sequences from thirty-one RDX biodegraders were detected, with the most abundant species being Variovorax sp. JS1663. Further, nine RDX biodegrading species significantly (p<0.05) increased in abundance following biostimulation. Both the sequencing data and qPCR indicated xenA and xenB exhibited the highest relative abundance among the six functional genes examined. Four genes, diaA, nsfI, xenA and pnrB, exhibited higher relative abundance values in some wells following biostimulation. The study provides a comprehensive approach for assessing biomarkers during RDX bioremediation and provides evidence that biostimulation generated a positive impact on a set of key species and genes. A third study examined laboratory microcosms to determine the phylotypes and functional genes associated with the biodegradation of cis-dichloroethene (cDCE) and 1,4- dioxane. The impact of amending with lactate on cDCE and 1,4-dioxane biodegradation was also investigated. Stable isotope probing (SIP) was then used to determine which phylotypes were actively involved in biodegradation. The most enriched phylotypes for 13C assimilation from 1,4- dioxane included Rhodopseudomonas and Rhodanobacter. The dominant enriched phylotypes for 13C assimilation from cDCE included Bacteriovorax, Pseudomonas and Sphingomonas. The functional genes associated with the degradation of these contaminants was predicted using PICRUSt2. The results suggest aerobic concurrent biodegradation of cDCE and 1,4-dioxane should be considered for chlorinated solvent site remediation. Overall, the data generated and approaches utilized in all three projects have the potential to be incorporated into diagnostic molecular tools for assessing biodegradation potential and for evaluating bioremediation performance at contaminated sites. Copyright by HONGYU DANG 2021 To my Mom and Dad, for encouraging me to go on every adventure, nurturing me to cherish the value of knowledge and everything else. v ACKNOWLEDGEMENTS I would like to express my deepest appreciation to Dr. Alison Cupples, who continually supported and guided me in the past four years. Without her help and exceptional scientific guidance, the work in this thesis would not have been possible. I would like to thank my committee members, Dr. Syed Hashsham, Dr. Irene Xagoraraki and Dr. Brian Teppen, for their valuable research advice. I would like to thank the current and past members in our laboratory and Dr. Hashsham’s laboratory, particularly, Yogendra Kanitkar, Robert Stedtfeld and Maggie Williams, for their willingness to teach me many methods during the early stages of my doctoral program. I am also grateful to the research efforts of Jenny Collier and Vidhya Ramalingam, as their work built foundations for my research. Also, thanks Dr. Benli Chai for introducing me to bioinformatic tools, which had lasting effects across this thesis. In addition, thanks to my friends, Jianzhou He and Xun Qian for enhancing my life and providing constructive advice in my research during my time at MSU. Lastly, I would like to thank my family for their endless love and support. vi TABLE OF CONTENTS LIST OF TABLES ....................................................................................................................................... ix LIST OF FIGURES ...................................................................................................................................... x KEY TO ABBREVIATIONS .................................................................................................................... xvi CHAPTER 1 INTRODUCTION ............................................................................................................... 1 1. Chlorinated Solvent, RDX and 1,4-Dioxane Contamination ............................................................... 1 2. Biodegradation of Chlorinated Solvents, 1,4-Dioxane and RDX ......................................................... 2 3. High Throughput Sequencing for Monitoring Biodegradation ............................................................ 4 4. Dissertation Outline and Objectives ..................................................................................................... 5 REFERENCES ............................................................................................................................................. 7 CHAPTER 2 ABUNDANCE OF CHLORINATED SOLVENT AND 1,4-DIOXANE DEGRADING MICROORGANISMS AT FIVE CHLORINATED SOLVENT CONTAMINATED SITES DETERMINED VIA SHOTGUN SEQUENCING .................................................................................... 13 Abstract ................................................................................................................................................... 13 1. Introduction ........................................................................................................................................ 14 2. Methods .............................................................................................................................................. 16 2.1 DNA Extraction from Groundwater and SDC-9 .......................................................................... 16 2.2 Sequencing and Taxonomic Analysis .......................................................................................... 16 2.3 Reference Sequences Collection, Functional Gene Analysis, qPCR ........................................... 17 3. Results ................................................................................................................................................ 17 3.1 Sequencing and Taxonomic Analysis .......................................................................................... 17 3.2 Occurrence of Chlorinated Solvent Degrading Microorganisms in SDC-9 and In Situ............... 18 3.3 Functional Gene Analysis ............................................................................................................ 21 4. Discussion ........................................................................................................................................... 30 Acknowledgements ................................................................................................................................ 35 APPENDIX ................................................................................................................................................. 36 REFERENCES ........................................................................................................................................... 84 CHAPTER 3 DIVERSITY AND ABUNDANCE OF THE FUNCTIONAL GENES AND BACTERIA ASSOCIATED WITH RDX DEGRADATION AT A CONTAMINATED SITE PRE- AND POST- BIOSTIMULATION .................................................................................................................................. 94 Abstract ................................................................................................................................................... 94 1. Introduction ........................................................................................................................................ 95 2. Materials and Methods ....................................................................................................................... 97 2.1 Sample Collection and DNA Extraction ...................................................................................... 97 2.2 High Throughput Sequencing ...................................................................................................... 97 2.3 Taxonomic Analysis ..................................................................................................................... 98 2.4 Functional Gene Analysis ............................................................................................................ 99 2.5 Functional Gene Phylogenetic Trees .......................................................................................... 100 2.6 Co-occurrence Network of Genera............................................................................................. 100 2.7 Analysis for Species Associated with RDX Biodegradation ..................................................... 101 2.8 Statistical Analysis ..................................................................................................................... 101 2.9 High Throughput Quantitative PCR ........................................................................................... 102 vii 3. Results .............................................................................................................................................. 103 3.1 RDX Concentrations .................................................................................................................. 103 3.2 Microbial Community Analysis based on MG-RAST ............................................................... 103 3.3 Functional Genes Associated with RDX Biodegradation .......................................................... 105 3.4 Co-occurrence of Genera Associated with RDX Biodegradation .............................................. 112 3.5 Presence of Known RDX Degraders .......................................................................................... 114 3.6 KEGG Pathways ........................................................................................................................ 116 3.7 High Throughput qPCR ............................................................................................................. 116 4. Discussion ......................................................................................................................................... 116 Acknowledgements .............................................................................................................................. 121 APPENDIX ............................................................................................................................................... 122 REFERENCES ......................................................................................................................................... 140 CHAPTER 4 IDENTIFICATION OF THE PHYLOTYPES AND PREDICTED FUNCTIONAL GENES INVOLVED IN CIS-DICHLOROETHENE AND 1,4-DIOXANE AEROBIC BIODEGRADATION IN SOIL MICROCOSMS .................................................................................... 148 Abstract ................................................................................................................................................. 148 1. Introduction ...................................................................................................................................... 149 2. Methods ............................................................................................................................................ 153 2.1 Chemicals and Soil Inocula ........................................................................................................ 153 2.2 Microcosms Setup ...................................................................................................................... 153 2.3 Analytic Methods ....................................................................................................................... 154 2.4 DNA extraction, Fractionation and Sequencing ......................................................................... 155 2.5 Analysis of Sequencing Data ..................................................................................................... 157 2.6 Function Prediction and Correlation .......................................................................................... 158 2.7 Species Associated with 1,4-dioxane and DCE Degradation ..................................................... 159 3. Results .............................................................................................................................................. 159 3.1 Degradation of 1,4-Dioxane and cDCE With or Without Lactate ............................................. 159 3.2 Microbial Community Analysis ................................................................................................. 164 3.3 Phylotypes Responsible for 13C Uptake from cDCE and 1,4-Dioxane ...................................... 167 3.4 Co-Occurrence Networks ........................................................................................................... 171 3.5 Predicted Functions and Correlations with OTUs ...................................................................... 172 4. Discussion ......................................................................................................................................... 173 Acknowledgements .............................................................................................................................. 179 APPENDIX ............................................................................................................................................... 180 REFERENCES ......................................................................................................................................... 194 CHAPTER 5 CONCLUSIONS AND FUTURE RESEARCH DIRECTIONS ..................................... 203 viii LIST OF TABLES Table 1. 1. Number of active superfund sites with above contaminants as of 04/06/2021 (2) ..................... 2 Supplementary Table 2. 1. Groundwater and sampling data. ..................................................................... 46 Supplementary Table 2. 2. Groundwater and SDC-9 MG-RAST sequence analysis data. ........................ 47 Supplementary Table 2. 3. Genomes used for collecting functional protein sequences. ............................ 48 Supplementary Table 2. 4. Number of collected genomes and dereplicated RDases. ................................ 49 Table 3. 1. Collection dates, RDX concentrations, locations, DNA concentrations and names of groundwater sampling wells. ...................................................................................................................... 98 Supplementary Table 3. 1. MG-RAST analysis data for datasets from DNA extracts of groundwater samples pre- and post-biostimulation. ...................................................................................................... 123 Supplementary Table 3. 2. FunGene filters for obtaining the reference sequences and the number of collected sequences before and after dereplication. .................................................................................. 124 Supplementary Table 3. 3. Identified RDX degraders with the lowest rank name and taxonomy ID from NCBI. ........................................................................................................................................................ 124 Supplementary Table 3. 4. STAMP analysis parameters for generating Supplementary Figures 3.4 and 3.5. Filters were applied to limit the number of genera or functions shown in each figure. All tests used Welch’s t-test (two sided) with the default CI option (Welch’s inverted) and default multiple test correction (no correction). For each test, two groups were compared (pre and post biostimulation samples). ................................................................................................................................................... 126 Supplementary Table 3. 5. Spearman’s rank correlation parameters between gene copy number of qPCR and relative abundance of genes associated with RDX degradation. ........................................................ 127 Supplementary Table 4. 1. Identified 1,4-dioxane and DCE degraders with the lowest rank name and taxonomy ID from NCBI. ......................................................................................................................... 181 Supplementary Table 4. 2. Enriched OTUs captured by the co-occurrence network. These OTUs were enriched in heavy fractions of 13C labeled chemicals amended samples determined by STAMP. ........... 182 ix LIST OF FIGURES Figure 2. 1. Relative abundance (%, as determined using MG-RAST) of methanotrophs and genera associated with chlorinated solvent biodegradation in groundwater from San Antonio (A), Tulsa (B), Quantico (C), Edison (D), Indian Head (E) and SDC-9 (F). The genus Dehalococcoides was present in all groundwater samples ranging from 0.1 – 3.5%. Note,"MW" in name refers to a monitoring well and "IW" in name refers to an injection well. The insert in F does not include Dehalococcoides or Desulfitobacterium to enable a y-axis with a different scale. ..................................................................... 19 Figure 2. 2. Normalized relative abundance (%, as determined by DIAMOND) of genes associated with reductive dechlorination in Dehalococcoides mccartyi (A), Dehalogenimonas spp. (B), Dehalobacter spp. (C) and Desulfitobacterium spp. (D) in SDC-9 (inserts) and in groundwater from the five chlorinated solvent sites. The highest abundance values are from tceA and vcrA from Dehalococcoides, followed by cerA from Dehalogenimonas. ..................................................................................................................... 23 Figure 2. 3. Normalized relative abundance (%, determined with DIAMOND) of genes (A) and relative abundance (%, determined with MG-RAST) of genera (B) previously associated with 1,4-dioxane degradation in all groundwater samples and in SDC-9. The relative abundance of Pseudonocardia was zero in all groundwater samples and in SDC-9. Methylosinus trichosporium OB3b mmoX was the dominant 1,4-dioxane degrading gene in the majority of the groundwater samples. ................................. 24 Figure 2. 4. Normalized relative abundance (%, determined with DIAMOND) of genes associated with the chlorinated solvent reductive dechlorination (A) and the aerobic degradation of the chlorinated solvents (B, C) in SDC-9 (insert for A) and in groundwater from the five chlorinated solvent sites. The aerobic genes occurred at lower levels compared to the anaerobic genes. Note, the analysis approach differed from the approach used to generated figure 2, in that all sequences from the databases were compared to each dataset. ........................................................................................................................... 26 Figure 2. 5. Principle component analyses of functional genes (A) and genera (B) associated with chlorinated solvent and 1,4-dioxane biodegradation in all groundwater samples. ..................................... 29 Supplementary Figure 2. 1. TCE plume maps for the Edison, NJ site. TCE contour maps for the site prior to addition of emulsified oil and dehalogenating culture SDC-9 in 2009 are provided for the shallow zone (A) and deep zone at the site (B). Well 303S is located in the shallow zone and well 114 is located in the deep zone. Post-treatment contour maps in 2010 for the shallow zone (C) and deep zone (D) are also provided. All values are in µg/L. The wells from which samples were collected and analyzed are indicated with arrows. ................................................................................................................................. 50 Supplementary Figure 2. 2. Cis-DCE Plume maps for the Edison, NJ site. Cis-DCE contour maps for the site prior to addition of emulsified oil and dehalogenating culture SDC-9 in 2009 are provided for the shallow zone (A) and deep zone at the site (B). Well 303S is located in the shallow zone and well 114 is located in the deep zone. Post-treatment contour maps in 2010 for the shallow zone (C) and deep zone (D) are also provided. All values are in µg/L. The wells from which samples were collected and analyzed are indicated with arrows. ........................................................................................................................... 54 Supplementary Figure 2. 3. Demonstration plot layout at the Quantico, VA site. The cathode and anode wells are indicated by red and green symbols, respectively. This system was used to supply H2 to support x reductive dechlorination of cis-DCE downgradient of a landfill. See data in Supplementary Figures 17- 19. ............................................................................................................................................................... 58 Supplementary Figure 2. 4. Concentration data for cis-DCE at the Quantico, VA site. The groundwater samples were collected on Day 243 from wells CW-2, PMW-2, CW-2, AW-1, MW-15R, and PMW-4. . 59 Supplementary Figure 2. 5. Concentration data for vinyl chloride at the Quantico, VA site. The groundwater samples were collected on Day 243 from wells CW-2, PMW-2, CW-2, AW-1, MW-15R, and PMW-4. ................................................................................................................................................ 60 Supplementary Figure 2. 6. Concentration data for ethene at the Quantico, VA site. The groundwater samples were collected on Day 243 from wells CW-2, PMW-2, CW-2, AW-1, MW-15R, and PMW-4. . 61 Supplementary Figure 2. 7. Demonstration plot layout at the Indian Head, Md site. Injection wells (IWs) were amended with lactate, diammonium phosphate, potassium bicarbonate (for pH adjustment) and dehalogenating culture SDC-9. Monitoring wells (MWs) were used to measure system performance. A low voltage was used to maintain system pH. Anodes for this system are shown in the figure. Wells that were sampled are indicated by arrows. See MW data in Supplementary Figures 21-22. No analytical data are available for the IWs. ............................................................................................................................ 62 Supplementary Figure 2. 8. Concentration data for cVOCs, ethene and ethane in well MW38 at the Indian Head, Md site. The groundwater samples were collected on 6/22/16......................................................... 63 Supplementary Figure 2. 9. Concentration data for cVOCs, ethene and ethane in well MW40 at the Indian Head, Md site. The groundwater samples were collected on 6/22/16......................................................... 64 Supplementary Figure 2. 10. Demonstration Plot layout at the Tulsa, Ok site. IWs are emulsified oil and dehalogenating culture SDC-9 injection wells and MWs are groundwater monitoring wells. See data in Supplementary Figures 24-26. .................................................................................................................... 65 Supplementary Figure 2. 11. Concentration data for TCE in injection wells (IWs) at the Tulsa, OK Site. The groundwater samples were collected on 6/09/15. ................................................................................ 66 Supplementary Figure 2. 12. Concentration data for TCE in monitoring wells (MWs) at the Tulsa, OK Site. The groundwater samples were collected on 6/09/15. ....................................................................... 67 Supplementary Figure 2. 13. Concentration data for 1,4-dioxane in injection wells (IWs) at the Tulsa, OK Site. The groundwater samples were collected on 6/09/15. ....................................................................... 68 Supplementary Figure 2. 14. Injection points and locations of monitoring wells SS050MW113 (113) and SS050MW514 (514) at the San Antonio, TX, Site. Analytical data are provided for each well. Groundwater samples were collected on 7/28/16. BZ = benzene. ............................................................. 69 Supplementary Figure 2. 15. Injection points and location of monitoring well SS050MW035 (35) at the San Antonio, TX, Site. Analytical data are provided. Groundwater samples were collected on 7/28/16... 70 Supplementary Figure 2. 16. Rarefaction curves for microbial communities in groundwater and in SDC-9. .................................................................................................................................................................... 71 Supplementary Figure 2. 17. Classification of microbial communities in two samples of SDC-9 (data analyzed with MG-RAST). ......................................................................................................................... 72 xi Supplementary Figure 2. 18. Classification of microbial communities in three monitoring well groundwater samples from San Antonio (data analyzed with MG-RAST). ............................................... 73 Supplementary Figure 2. 19. Classification of microbial communities in injection well (A and B) and monitoring well (C, D and E) groundwater samples from Tulsa (data analyzed with MG-RAST). .......... 74 Supplementary Figure 2. 20. Classification of microbial communities in groundwater injection well (A) and monitoring well (B, C, D) samples from Quantico (data analyzed with MG-RAST). ......................... 76 Supplementary Figure 2. 21. Classification of microbial communities in groundwater monitoring well samples from Edison (data analyzed with MG-RAST). ............................................................................. 77 Supplementary Figure 2. 22. Classification of microbial communities in groundwater injection (A, B) and monitoring well (C, D) samples from Indian Head (data analyzed with MG-RAST). ............................... 78 Supplementary Figure 2. 23. Normalized relative abundance (%) of fdhA in SDC-9 (insert) and in groundwater from the five chlorinated solvent sites (data analyzed with DIAMOND). ............................ 79 Supplementary Figure 2. 24. Normalized relative abundance (%) of Dehalococcoidies mccartyi hydrogenase genes hupLS (A), vhcAG (B), hymABCD (C) and echABCEF (D) in SDC-9 (inserts) and in groundwater from the five chlorinated solvent sites (data analyzed with DIAMOND). ............................ 80 Supplementary Figure 2. 25. Normalized relative abundance (%) of Dehalococcoidies mccartyi corrinoid metabolism genes btuFCD (A), cbiA, cbiB, cbiZ (B) and cobA, cobB, cobC, cobD, cobQ, cobS, cobT, cobU (C) in SDC-9 (inserts) and in groundwater from the five chlorinated solvent sites (data analyzed with DIAMOND). ....................................................................................................................................... 81 Supplementary Figure 2. 26. Comparison between normalized relative abundance of vcrA, tceA and sum of RDases to fdhA (data analyzed with DIAMOND).................................................................................. 82 Supplementary Figure 2. 27. Comparison between vcrA gene copies (per L) determined via qPCR and shotgun sequencing (normalized relative abundance, %, MG-RAST). The results from two shotgun sequencing quantification methods are shown (as discussed in the text). .................................................. 83 Figure 3. 1. Phylogenetic trees were built for the aligned reference sequences of functional genes (A-F), the reference sequences were colored by phylum or class. The bars on the right illustrated the relative abundance (%) of aligned reference sequences in different samples, light and dark red denoted pre- and post-biostimulation from MW32. light and dark orange denoted pre- and post-biostimulation from MW62, light and dark green denoted pre- and post-biostimulation from MW66, light and dark blue denoted pre- and post-biostimulation from MW67, purple denoted different wells with only post- biostimulated samples. .............................................................................................................................. 107 Figure 3. 2. Normalized relative abundance (%) of the total aerobic (nfsI, pnrB and xplA) (A) and anaerobic (diaA, xenA and xenB) (B) functional genes relevant to RDX biodegradation across all monitoring wells (MW) in replicate DNA extracts. The legend terms post and pre refer to the post- and pre-biostimulation samples, respectively. ................................................................................................. 111 Figure 3. 3. Co-occurrence network based on spearman correlation (rho > 0.85 and p-value < 0.01) of main genera found in all samples from post-biostimulated wells. Only genus with an average abundance > 0.1% and present in at least 50% of samples were considered. Node size indicates the relative abundance (0.1% ~ 5.46%). Nodes colored in red: identified genus associated with RDX degradation. Nodes colored xii in orange: potential genus to generate diaA. Nodes colored in pink: potential genus to generate pnrB. Nodes colored in blue: potential genus to generate xplA. Nodes colored in green: potential genus to generate xenA or xenB. No potential genus to generate nfsI was found.................................................... 113 Figure 3. 4. Phylogram constructed with reads assigned (identity ≥ 85% and query coverage ≥ 85%) to the species associated with RDX degradation across all monitoring wells (MW) in replicate DNA extracts. Each species was colored with phylum or class from Proteobacteria. The bars in the outside indicated the normalized counts assigned to the species, missing bars meant zero counts. ........................................... 115 Figure 3. 5. Species associated with RDX degradation showed significant differences before and after biostimulation across all wells. The extended error bar was created using Welch’s t-test (two sided) with the default CI option (Welch’s inverted), default multiple test correction (no correction) and default p value filter of 0.05. .................................................................................................................................... 115 Supplementary Figure 3. 1. Principle component analysis of all samples based on the genus results from MG-RAST. Clustered samples were marked in circles. ........................................................................... 127 Supplementary Figure 3. 2. The most abundant phylotypes in each sample at the class (A), order (B) family (C), and genus (D) levels. For each classification, phylotypes with an average relative abundance across all samples less than 1% were placed within "other". .................................................................... 128 Supplementary Figure 3. 3. Relative abundance (%) of the 20 most abundant genera in duplicated samples from each well. Pre and post refer to the pre- and post-biostimulation samples, respectively. ................ 129 Supplementary Figure 3. 4. A comparison of those significantly different between pre- and post- biostimulation wells from the abundant genera (relative abundance ≥ 1.5) (p < 0.05, Welch's two sided t- test). ........................................................................................................................................................... 130 Supplementary Figure 3. 5. Comparison of the most abundant genera pre- and post biostimulation in MW32 (A), MW62 (B), MW66 (C) and MW67 (D) with significant differences (p < 0.05, Welch's two sided t-test). ............................................................................................................................................... 131 Supplementary Figure 3. 6. Normalized relative abundance (%) of the functional genes relevant to RDX biodegradation across all monitoring wells (MW) in replicate DNA extracts. The legend terms post and pre refer to the post- and pre-biostimulation samples, respectively. ......................................................... 133 Supplementary Figure 3. 7. Taxonomy of microorganisms associated with aligned references sequences of functional genes: diaA (phylum level, A), nfsI (genus level, B), pnrB (genus level, C), xenA (order level, D), xenB (order level, E) and xplA (genus level, F) sequences across all soils. .............................. 134 Supplementary Figure 3. 8. Co-occurrence network based on spearman correlation (rho > 0.85 and p- value < 0.01) of main genera found in all samples from post-biostimulated wells. Only genus with an average abundance > 0.1% and present in at least 50% of samples were considered. Node size indicates the relative abundance (0.1% ~ 5.46%). The network was process with Modularity function of Gephi to group nodes colored into 7 different modules with default setting and a resolution of 0.85. ................... 136 Supplementary Figure 3. 9. Comparison of degradation pathway (A) and functions (B) in category of Xenobiotics biodegradation and metabolism between pre- and post biostimulation wells. For degradation pathway analysis, default options were used for the two groups comparison (p < 0.05, Welch's two sided t-test). For functions analysis, an extra filter was added as difference in mean proportions > 1%. .......... 137 xiii Supplementary Figure 3. 10. Heatmap of groundwater and Red Cedar River (RC) log10 gene copies per milliliter (A) and sediment log10 gene copies per gram (B). Grey cells indicate either no amplification or false positive amplification. In the sample name, post and pre refer to the post- and pre-biostimulation samples, J_ refer to the samples from previous work (24). ...................................................................... 138 Figure 4. 1. 1,4-Dioxane (A) and cDCE (B) concentrations in triplicate sample microcosms (purple [A] and blue [B]) and triplicate abiotic controls (red) inoculated with soil 1 and different amendments. The shaded areas indicate 95% confidence intervals along the linear regression model. ................................ 162 Figure 4. 2. 1,4-Dioxane (A) and cDCE (B) concentrations in triplicate sample microcosms (purple [A] and blue [B]) and triplicate abiotic controls (red) inoculated with soil 2 and different amendments. The shaded areas indicate 95% confidence intervals along the linear regression model. ................................ 163 Figure 4. 3. Non-metric Multi-dimensional Scaling (NMDS) plots (A, B) and alpha diversity measurements (C, D) for the cDCE (A, C) and 1,4-dioxane (B, D) SIP experiments with soil 1 and 2. . 165 Figure 4. 4. Classification to the phylum level for both replicates and soils amended with 1,4-dioxane (upper plot) or cDCE (lower plot) with a rarefied even depth of 95% of the minimum sum of OTU counts, each column represents cumulative values for three fractions (A). The classification (family level) of the top 30 OTUs (across all samples) within the most dominant phylum (Proteobacteria) (B) without rarefaction. ................................................................................................................................................ 166 Figure 4. 5. Boxplots with Wilcoxon Rank test results between phylotypes enriched (as determined by STAMP) in 13C amended heavy fractions (red dots) compared to the 12C amended heavy fractions (purple dots) by soil (upper is Soil 1 and lower is Soil 2) and by replicate of 1,4-dioxane amended samples. The graphs on the right have a different y-axis. P values of 0.0001, 0.001, 0.01, 0.05, >0.05 are presented by ****, ***, **, *, ns. ........................................................................................................... 169 Figure 4. 6. Boxplots with Wilcoxon Rank test results between phylotypes enriched (as determined by STAMP) in 13C amended heavy fractions (green dots) compared to the 12C amended heavy fractions (blue dots) by soil (upper is Soil 1 and lower is Soil 2) and by replicate of cDCE amended samples. The graphs for Soil 2 have a different y-axis. P values of 0.0001, 0.001, 0.01, 0.05, >0.05 are presented by ****, ***, **, *, ns. ................................................................................................................................. 170 Figure 4. 7. Correlation of KO functions associated with degradation and OTUs with average abundance higher than 0.05% from 1,4-dioxane (A) and cDCE in SIP tests. 18 OTUs had an absolute correlation coefficient high than 0.59 with at least 4 of function in 1,4-dioxane SIP. 20 OTUs had an absolute correlation coefficient high than 0.6 with at least 2 of function in 1,4-dioxane SIP. ................................ 173 Supplementary Figure 4. 1. Average concentration of labeled and unlabeled 1,4-dioxane (A, B), and cDCE (C, D) in triplicate sample microcosms and triplicate abiotic controls inoculated with soil 1 (A, C) and 2 (B, D)............................................................................................................................................... 183 Supplementary Figure 4. 2. Rarefaction curves for DNA extracts in heavy and light fractions from 1,4- dioxane (A) and cDCE (B) SIP experiments in microcosms amended with soil 1. and 2. ....................... 184 Supplementary Figure 4. 3. Non-metric Multi-dimensional Scaling (NMDS) plot and alpha diversity measurements for sequencing results of 1,4-dioxane and cDCE SIP tests by KBS soil 1 and 2. ............. 185 Supplementary Figure 4. 4. The most abundant (top 40) genera (by mean, with phylum) in all SIP fractions from 1,4-dioxane (A) and cDCE (B) amended microcosms inoculated with soil 1 or 2. .......... 186 xiv Supplementary Figure 4. 5. Species or strains previously associated with 1,4-dioxane (A) or cDCE (B) biodegradation present in KBS soil 1 (red) and soil 2 (blue) from shotgun sequencing data. .................. 187 Supplementary Figure 4. 6. Phylotypes statistically enriched (Welch's two sided t-test, p<0.05) in heavy fractions (1.730-1.747 g/mL) of 13C 1,4-dioxane amended samples compared to fractions of comparable buoyant density (1.730-1.748 g/mL) in 12C 1,4-dioxane amended samples in soil 1 (A) and 2 (B). Values and error bars represent averages and standard deviations from three fractions each (with each fraction being sequenced and quantified in triplicate). After removing the background phylotypes that were also enriched in light fractions, a total of 282 and 28 phylotypes were enriched 1,4-dioxane amended samples for soil 1 and 2, respectively. The figure only displayed phylotypes with a difference of average abundance from 13C 1,4-dioxane and amended 12C 1,4-dioxane samples higher than 0.15% (A) and 0.01% (B) for soil 1 and soil 2. The insert was in a smaller scale. ...................................................................... 188 Supplementary Figure 4. 7. Phylotypes statistically enriched (Welch's two sided t-test, p<0.05) in heavy fractions (1.733-1.744 g/mL) of 13C cDCE amended samples compared to fractions of comparable buoyant density (1.733-1.745 g/mL) in 12C DCE amended samples in soil 1 (A) and 2 (B). Values and error bars represent averages and standard deviations from three fractions each (with each fraction being sequenced and quantified in triplicate). After removing the background phylotypes that were also enriched in light fractions, a total of 30 and 25 phylotypes were enriched in DCE amended samples for soil 1 and soil 2, respectively . The figure only displayed phylotypes with a difference of average abundance from 13C DCE and amended 12C DCE samples higher than 0.1% (A) and 0.05% (B) for soil 1 and soil 2. The insert was in a smaller scale. ............................................................................................ 189 Supplementary Figure 4. 8. Co-occurrence networked based on spearman correlation (rho > 0.70 and p- value < 0.01) for the main OTUs from microbial communities for 1,4-dioxane (A) and cDCE (B) degradation. Connected nodes with lines had a rho > 0.7. Filters for main OTUs: present in at least 50% of samples, average abundance > 0.06% (A) and > 0.1% (B). There were 166 (A) and 172 (B) nodes met the filters. The networks were colored with OTUs show significant difference (p<0.05) of samples from heavy fraction of soil 1 and 2 amended with 13C labeled 1,4-dioxane or cDCE. Number of nodes belonging to that group was in the parentheses. ....................................................................................... 190 Supplementary Figure 4. 9. Co-occurrence networked based on spearman correlation (rho > 0.70 and p- value < 0.01) for the main OTUs from microbial communities for 1,4-dioxane degradation. Connected nodes with lines had a rho > 0.7. Filters for main OTUs: present in at least 50% of samples, average abundance > 0.06%. There were 166 nodes met the filters. A: OTUs were colored by phylum, B: OTUs were colored by if its abundance is significantly higher in DNA with C13 isotope high BD value fractions from soil 1 or soil 2. Number of nodes belonging to that group was in the parentheses. ......................... 191 Supplementary Figure 4. 10. Co-occurrence networked based on spearman correlation (rho > 0.70 and p- value < 0.01) for the main OTUs from microbial communities for cDCE degradation. Connected nodes with lines had a rho > 0.7. Filters for main OTUs: present in at least 50% of samples, average abundance > 0.1%. There were 172 nodes met the filters. A: OTUs were colored by phylum, B: OTUs were colored by if its abundance is significantly higher in DNA with C13 isotope high BD value fractions from soil 1 or soil 2. Number of nodes belonging to that group was in the parentheses. ......................... 192 Supplementary Figure 4. 11. KO functions associated with 1,4-dioxane and cDCE in SIP fractions obtained from microcosm replicates (R1 and R2) in both soil 1 (T1) and 2 (T2). ................................... 193 xv KEY TO ABBREVIATIONS PCE TCE Tetrachloroethene Trichloroethene DCE Dichloroethene VC Vinyl Chloride RDX Hexahydro-1,3,5-trinitro-1,3,5-triazine qPCR Quantitative Polymerase Chain Reaction RDase Reductive Dehalogenase SIP Stable Isotope Probing MG-RAST Metagenomics Rapid Annotation using Subsystem Technology DIAMOND Double Index Alignment of Next-Generation Sequencing Data KBS LTER Kellogg Biological Station Long-Term Ecological Research STAMP Statistical Analysis of Taxonomic and Functional Profiles KEGG Kyoto Encyclopedia of Genes and Genomes OTU Operational Taxonomic Unit xvi CHAPTER 1 Introduction 1. Chlorinated Solvent, RDX and 1,4-Dioxane Contamination Chlorinated Solvent Contamination. The chlorinated solvents tetrachloroethene (PCE) and trichloroethene (TCE) and their metabolites, cis-dichloroethene (cDCE) and vinyl chloride (VC), are persistent groundwater contaminants, requiring remediation because of their risks to human health (1). From the list of co-contaminants found in soil and groundwater, the chlorinated solvents and their metabolites are particularly prevalent (found at > 3000 active superfund sites (2), Table 1) and problematic due to their tendency to form large, dissolved-phase plumes and their recalcitrant nature (3). RDX Contamination. RDX (Hexahydro-1,3,5-trinitro-1,3,5-triazine), also known as Royal Demolition Explosive, is a synthetic product and commonly used explosive (4). It is classified as a possible human carcinogen and its exposure can lead to irritation, nausea, kidney damage and other health impacts (4). RDX has caused soil, groundwater and sediment contamination because of the denotation of ordnance, firing of munitions on military training ranges, and the manufacturing and transport of munitions. Currently, there are 39 active sites with RDX on the National Priority List (2) (Table 1.1). 1,4-Dioxane Contamination. 1,4-Dioxane, a probable human carcinogen and common chlorinated solvent stabilizer, has been found at numerous contaminated sites across the U.S. (4, 5). In an examination from 49 remediation installations at U.S. Air Force sites, 1,4-dioxane was detected in 781 groundwater wells, with the percentage of 1,4-dioxane with TCE in all 1,4- dioxane detection-positive wells being 64.4% (6). In an evaluation of >2000 sites in California, 1 the chlorinated solvents were found in 94% of the sites with detections of 1,4-dioxane (7). There are 69 active super fund sites contaminated with 1,4-dioxane (2) (Table 1). Table 1. 1. Number of active superfund sites with above contaminants as of 04/06/2021 (2) Contaminants Chloroethene RDX 1,4-dioxane PCE TCE DCE (cis- and trans-DCE) VC Number of active sites 898 1007 815 677 39 68 2. Biodegradation of Chlorinated Solvents, 1,4-Dioxane and RDX In the past decade, biostimulation (e.g. the addition of electron donors) has become increasingly popular for many organic contaminants. For the chlorinated solvents, in many cases, bioaugmentation is practiced, involving the addition of both electron donor and mixed microbial communities. Bioaugmentation is starting to become more recognized as a potential remediation method for 1,4-dioxane and RDX, but the applications are currently more limited compared to the chlorinated solvents. Biodegradation of Chlorinated Solvents. Dehalococcoides mccartyi, is a key microorganism for the complete transformation of PCE to the non-hazardous end product, ethene (8, 9). D. mccartyi strains reduce chlorinated compounds obtaining energy from the reduction process (10-12). Examples of commercially available mixed cultures containing D. mccartyi for chlorinated solvent remediation include SDC-9 (from APTIM) and KB-1 (from SiREM) (13). It was estimated that several hundred sites in the U.S. have been subject to bioaugmentation with cultures containing D. mccartyi (14). Following bioaugmentation, remediation professionals commonly monitor D. mccartyi populations, typically targeting reductive dehalogenase (RDase) genes such as pceA, tceA, vcrA and bvcA (15-17) using quantitative PCR (qPCR) on nucleic acids extracted from groundwater (18-20). Other genera expressing RDases include 2 Dehalogenimonas (21), Desulfitobacterium (22), Dehalobacter (23), Geobacter (24), Sulfurospirillum (25) and Anaeromyxobacter (26). In addition to anaerobic dehalogenation, chloroethenes can also be biodegraded aerobically. The genes encoding for the enzymes associated with aerobic VC degradation include etnC (alkene monooxygenase) and etnE (epoxyalkane: CoM transferase) (27-31). Also, the gene encoding for cytochrome P450 from Polaromonas sp. strain JS666 initializes the biodegradation of cis-1,2-dichoroethene (32). Other genes associated with co-metabolism of chlorinated ethenes include the α subunits of soluble and particulate methane monooxygenases (mmoX and pmoA) (33-35). Biodegradation of RDX. RDX biodegradation is initiated by a number of bacteria and their associated enzymes under aerobic or anaerobic conditions. Under aerobic conditions, nfsI (present in both Morganella morganii strain B2 and Enterobacter cloacae strain 96-3) encodes a type I nitroreductase, which is responsible for the nitroreduction of RDX (36). Another RDX degrading functional gene, pnrB, was associated with Pseudomonas sp. and Stenotrophomonas maltophilia (37). Microorganisms within the genera Rhodococcus, Gordonia, Williamsia and Microbacterium have the well-studied xplA gene, which has been linked with nitro group removal and ring cleavage (38-41). Under anaerobic conditions, RDX transformation through nitro group denitration was initiated by the enzyme encoded by diaA from Clostridium kluyveri (42, 43). Finally, xenA and xenB, from the genus Pseudomonas, encode enzymes for the transformation of RDX to methylenedinitramine (44). These genes have previously been monitored at contaminated sites as evidence for RDX degradation, often using qPCR (45-47). Biodegradation of 1,4-Dioxane. Many bacteria and enyzymes have been associated with the metabolic or co-metabolic biodegradation of 1,4-dioxane under aerobic conditions (48, 49), 3 with limited information available for anaerobic biodegradation. Pseudonocardia dioxanivorans strain CB1190 is a well studied 1,4-dioxane degrader able to use 1,4-dioxane as the carbon source via dioxane monooxygenase (50, 51). Other enyzmes involved in 1,4-dioxane biodegradation included toluene monooxygenase, propane monooxygenase, tetrahydrofuran monooxygenase and methane monooxygenase (49). At contaminated sites, 1,4-dioxane is often present with the chlorinated solvents, which can impact aerobic 1,4-dioxane biodegradation. For example, for P. dioxanivorans CB1190, 1,1- DCE and cDCE had a greater effect on 1,4-dioxane biodegradation compared to TCE, while the effect of 1,1,1-trichloroethane (1,1-TCA) was negligible (52). 3. High Throughput Sequencing for Monitoring Biodegradation Current approaches to detect biodegraders (targeting 16S rRNA genes or functional genes) for many contaminants in groundwater have typically focused on quantitative polymerase chain reaction or qPCR (16, 19, 45, 47, 53-56). Although qPCR has a high level of sensitivity, it can only target a limited number of genes (unless high throughput qPCR is used). For example, during the natural attenuation of chlorinated solvents (57-59), following biostimulation (60) and batch 1,4-dioxane biodegradation (61), 16S rRNA gene amplicon sequencing has been used to monitor the dynamics of microbial communities. However, functions from the microbial communities can only be predicted (62, 63) rather than directly detected. High throughput sequencing offers an additional valuable tool for monitoring biomarkers environmental samples because a limitless number of biomarkers can be investigated. In contrast to 16S rRNA gene amplicon sequencing, shotgun sequencing captures random pieces of DNA, thus can sequence both the taxonomic and the potential functional characteristics of microbial communities (64). 4 4. Dissertation Outline and Objectives The following summarizes the key objectives of each of the following chapters. Chapter 2 has been published (Dang, H., Kanitkar, Y. H., Stedtfeld, R. D., Hatzinger, P. B., Hashsham, S. A. and A. M. Cupples. 2018. Abundance of chlorinated solvent and 1,4-dioxane degrading microorganisms at five chlorinated solvent contaminated sites determined via shotgun sequencing. Environmental Science and Technology. 52 (23): 13914–13924), whereas Chapters 3 and 4 are currently being prepared for submission to peer reviewed journals. Chapter 2. The project examined nucleic acids extracted from SDC-9 and groundwater from five chloroethene contaminated sites, previously bioaugmented with SDC-9. The overall objective was to develop the methodology for both taxonomic and functional analysis for chlorinated solvent contaminated sites. The specific objectives were to 1) determine the relative abundance of genera associated with chloroethene biodegradation; 2) investigate the relative abundance of reductive dehalogenases and other functional biomarkers involved in the biodegradation of chlorinated contaminants and 1,4-dioxane and 3) correlate the abundance of all biomarkers across individual wells. Chapter 3. The project examined nucleic acids extracted from groundwater at an RDX contaminated military site using shotgun sequencing and high throughput qPCR. The specific objectives were to 1) determine the relative abundance of each functional gene, 2) ascertain the taxonomy of the microorganisms associated with each functional gene, 3) investigate changes in gene abundance following biostimulation and 4) ascertain if previously identified RDX degraders were present at the site and if their abundance changed following biostimulation. Chapter 4. The study examined the concurrent biodegradation of cDCE and 1,4-dioxane in laboratory soil microcosms. The specific objectives were to 1) examine removal rates of the 5 co-contaminants cDCE and 1,4-dioxane by two soil microcosms, with and without lactate addition, 2) identify the microorganisms responsible for the uptake of 13C from cDCE and 1,4- dioxane and 3) determine the functional genes present and correlate their presence to specific bacteria. Chapter 5. This chapter generalizes the key findings and provides suggestions for future research. 6 REFERENCES 7 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES Mattes TE, Alexander AK, Coleman NV. 2010. Aerobic biodegradation of the chloroethenes: pathways, enzymes, ecology, and evolution. FEMS Microbiol Rev 34:445-75. EPA. 2021. Search Superfund Site Information. https://cumulis.epa.gov/supercpad/cursites/srchsites.cfm. Yang Y, McCarty PL. 2000. Biologically Enhanced Dissolution of Tetrachloroethene DNAPL. Environmental Science & Technology 34:2979-2984. EPA. 2017. Technical Fact Sheet – Hexahydro-1,3,5-trinitro1,3,5-triazine (RDX). https://www.epa.gov/sites/production/files/2017-10/documents/ffrro_ecfactsheet_rdx_9-15- 17_508.pdf. Zenker MJ, Borden RC, Barlaz MA. 2003. Occurrence and Treatment of 1,4-Dioxane in Aqueous Environments. Environmental Engineering Science 20:423-432. Anderson RH, Anderson JK, Bower PA. 2012. Co-occurrence of 1,4-dioxane with trichloroethylene in chlorinated solvent groundwater plumes at US Air Force installations: Fact or fiction. Integrated Environmental Assessment and Management 8:731-737. Adamson DT, Mahendra S, Walker KL, Rauch SR, Sengupta S, Newell CJ. 2014. A Multisite Survey To Identify the Scale of the 1,4-Dioxane Problem at Contaminated Groundwater Sites. Environmental Science & Technology Letters 1:254-258. Muller JA, Rosner BM, von Abendroth G, Meshulam-Simon G, McCarty PL, Spormann AM. 2004. Molecular identification of the catabolic vinyl chloride reductase from Dehalococcoides sp strain VS and its environmental distribution. Applied and Environmental Microbiology 70:4880- 4888. He JZ, Ritalahti KM, Aiello MR, Loffler FE. 2003. Complete detoxification of vinyl chloride by an anaerobic enrichment culture and identification of the reductively dechlorinating population as a Dehalococcoides species. Applied and Environmental Microbiology 69:996-1003. 10. Maymo-Gatell X, Anguish T, Zinder SH. 1999. Reductive dechlorination of chlorinated ethenes and 1,2-dichloroethane by "Dehalococcoides ethenogenes" 195. Applied and Environmental Microbiology 65:3108-3113. 11. 12. He J, Sung Y, Krajmalnik-Brown R, Ritalahti KM, Loffler FE. 2005. Isolation and characterization of Dehalococcoides sp strain FL2, a trichloroethene (TCE)- and 1,2- dichloroethene-respiring anaerobe. Environmental Microbiology 7:1442-1450. Krajmalnik-Brown R, Holscher T, Thomson IN, Saunders FM, Ritalahti KM, Loffler FE. 2004. Genetic identification of a putative vinyl chloride reductase in Dehalococcoides sp strain BAV1. Applied and Environmental Microbiology 70:6347-6351. 8 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. Steffan RJ, Vainberg S. 2013. Production and handling of Dehalococcoides bioaugmentation cultures, p 89-115. In Stroo HF, Leeson A, Ward CH (ed), Bioaugmentation for Groundwater Remediation. Springer, New York. Lyon DY, Vogel TM. 2013. Bioaugmentation for groundwater remediation: an overview, p 1-38. In Stroo HF, Leeson A, Ward CH (ed), Bioaugmentation for groundwater remediation. Springer, New York. Kanitkar YH, Stedtfeld RD, Hatzinger PB, Hashsham SA, Cupples AM. 2017. Development and application of a rapid, user-friendly, and inexpensive method to detect Dehalococcoides sp reductive dehalogenase genes from groundwater. Applied Microbiology and Biotechnology 101:4827-4835. Perez-de-Mora A, Zila A, McMaster ML, Edwards EA. 2014. Bioremediation of chlorinated ethenes in fractured bedrock and associated changes in dechlorinating and nondechlorinating microbial populations. Environmental Science & Technology 48:5770-5779. Petrovskis EA, Amber WR, Walker CB. 2013. Microbial monitoring during bioaugmentation with Dehalococcoides, p 171-197. In Stroo HF, Leeson A, Ward CH (ed), Bioaugmentation for groundwater remediation. Springer, New York. Hatt JK, Loffler FE. 2012. Quantitative real-time PCR (qPCR) detection chemistries affect enumeration of the Dehalococcoides 16S rRNA gene in groundwater. Journal of Microbiological Methods 88:263-270. Lee PKH, Macbeth TW, Sorenson KS, Deeb RA, Alvarez-Cohen L. 2008. Quantifying genes and transcripts to assess the in situ physiology of "Dehalococcoides" spp. in a trichloroethene- contaminated groundwater site. Applied and Environmental Microbiology 74:2728-2739. Liang Y, Liu XK, Singletary MA, Wang K, Mattes TE. 2017. Relationships between the abundance and expression of functional genes from vinyl chloride (VC)-degrading bacteria and geochemical parameters at VC-contaminated sites. Environmental Science & Technology 51:12164-12174. Yang Y, Higgins SA, Yan J, Simsir B, Chourey K, Iyer R, Hettich RL, Baldwin B, Ogles DM, Loffler FE. 2017. Grape pomace compost harbors organohalide-respiring Dehalogenimonas species with novel reductive dehalogenase genes. ISME J 11:2767-2780. Futagami T, Fukaki Y, Fujihara H, Takegawa K, Goto M, Furukawa K. 2013. Evaluation of the inhibitory effects of chloroform on ortho-chlorophenol- and chloroethene-dechlorinating Desulfitobacterium strains. AMB Express 3:30. 23. Maillard J, Schumacher W, Vazquez F, Regeard C, Hagen WR, Holliger C. 2003. Characterization of the corrinoid iron-sulfur protein tetrachloroethene reductive dehalogenase of Dehalobacter restrictus. Appl Environ Microbiol 69:4628-38. 24. Sung Y, Fletcher KE, Ritalahti KM, Apkarian RP, Ramos-Hernandez N, Sanford RA, Mesbah NM, Loffler FE. 2006. Geobacter lovleyi sp. nov. strain SZ, a novel metal-reducing and tetrachloroethene-dechlorinating bacterium. Appl Environ Microbiol 72:2775-82. 9 25. 26. 27. 28. 29. Buttet GF, Holliger C, Maillard J. 2013. Functional genotyping of Sulfurospirillum spp. in mixed cultures allowed the identification of a new tetrachloroethene reductive dehalogenase. Appl Environ Microbiol 79:6941-7. He Q, Sanford RA. 2002. Induction characteristics of reductive dehalogenation in the ortho- halophenol-respiring bacterium, Anaeromyxobacter dehalogenans. Biodegradation 13:307-16. Hartmans S, Debont JAM. 1992. Aerobic vinyl chloride metabolism in Mycobacterium aurum L1. Applied and Environmental Microbiology 58:1220-1226. Danko AS, Saski CA, TomkinS JP, Freedman DL. 2006. Involvement of coenzyme M during aerobic biodegradation of vinyl chloride and ethene by Pseudomonas putida strain AJ and Ochrobactrum sp strain TD. Applied and Environmental Microbiology 72:3756-3758. Coleman NV, Spain JC. 2003. Epoxyalkane: Coenzyme M transferase in the ethene and vinyl chloride biodegradation pathways of Mycobacterium strain JS60. Journal of Bacteriology 185:5536-5545. 30. Mattes TE, Coleman NV, Spain JC, Gossett JM. 2005. Physiological and molecular genetic analyses of vinyl chloride and ethene biodegradation in Nocardioides sp strain JS614. Archives of Microbiology 183:95-106. 31. 32. 33. 34. 35. 36. 37. Coleman NV, Spain JC. 2003. Distribution of the coenzyme m pathway of epoxide metabolism among ethene- and vinyl chloride-degrading Mycobacterium strains. Applied and Environmental Microbiology 69:6041-6046. Nishino SF, Shin KA, Gossett JM, Spain JC. 2013. Cytochrome P450 Initiates Degradation of cis-Dichloroethene by Polaromonas sp Strain JS666. Applied and Environmental Microbiology 79:2263-2272. Lee SW, Keeney DR, Lim DH, Dispirito AA, Semrau JD. 2006. Mixed pollutant degradation by Methylosinus trichosporium OB3b expressing either soluble or particulate methane monooxygenase: Can the tortoise beat the hare? Applied and Environmental Microbiology 72:7503-7509. Yoon S, Im J, Bandow N, DiSpirito AA, Semrau JD. 2011. Constitutive expression of pMMO by Methylocystis strain SB2 when grown on multi-carbon substrates: implications for biodegradation of chlorinated ethenes. Environmental Microbiology Reports 3:182-188. Chang HL, Alvarez-Cohen L. 1996. Biodegradation of individual and multiple chlorinated aliphatic hydrocarbons by methane-oxidizing cultures. Applied and Environmental Microbiology 62:3371-3377. Kitts CL, Green CE, Otley RA, Alvarez MA, Unkefer PJ. 2000. Type I nitroreductases in soil Enterobacteria reduce TNT (2,4,6,-trinitrotoluene) and RDX (hexahydro-1,3,5-trinitro-1,3,5- triazine). Can J Microbiol 46:278-82. Caballero A, Lazaro JJ, Ramos JL, Esteve-Nunez A. 2005. PnrA, a new nitroreductase-family enzyme in the TNT-degrading strain Pseudomonas putida JLR11. Environ Microbiol 7:1211-9. 10 38. 39. 40. 41. 42. 43. 44. 45. Indest KJ, Crocker FH, Athow R. 2007. A TaqMan polymerase chain reaction method for monitoring RDX-degrading bacteria based on the xplA functional gene. J Microbiol Methods 68:267-74. Bernstein A, Adar E, Nejidat A, Ronen Z. 2011. Isolation and characterization of RDX-degrading Rhodococcus species from a contaminated aquifer. Biodegradation 22:997-1005. Seth-Smith HM, Rosser SJ, Basran A, Travis ER, Dabbs ER, Nicklin S, Bruce NC. 2002. Cloning, sequencing, and characterization of the hexahydro-1,3,5-Trinitro-1,3,5-triazine degradation gene cluster from Rhodococcus rhodochrous. Appl Environ Microbiol 68:4764-71. Andeer PF, Stahl DA, Bruce NC, Strand SE. 2009. Lateral transfer of genes for hexahydro-1,3,5- trinitro-1,3,5-triazine (RDX) degradation. Appl Environ Microbiol 75:3258-62. Bhushan B, Halasz A, Spain JC, Hawari J. 2002. Diaphorase catalyzed biotransformation of RDX via N-denitration mechanism. Biochem Biophys Res Commun 296:779-84. Chakraborty S, Sakka M, Kimura T, Sakka K. 2008. Cloning and expression of a Clostridium kluyveri gene responsible for diaphorase activity. Biosci Biotechnol Biochem 72:735-41. Fuller ME, McClay K, Hawari J, Paquet L, Malone TE, Fox BG, Steffan RJ. 2009. Transformation of RDX and other energetic compounds by xenobiotic reductases XenA and XenB. Appl Microbiol Biotechnol 84:535-44. Fuller ME, McClay K, Higham M, Hatzinger PB, Steffan RJ. 2010. Hexahydro-1,3,5-trinitro- 1,3,5-triazine (RDX) Bioremediation in Groundwater: Are Known RDX-Degrading Bacteria the Dominant Players? Bioremediation Journal 14:121-134. 46. Michalsen MM, King AS, Rule RA, Fuller ME, Hatzinger PB, Condee CW, Crocker FH, Indest KJ, Jung CM, Istok JD. 2016. Evaluation of Biostimulation and Bioaugmentation To Stimulate Hexahydro-1,3,5-trinitro-1,3,5,-triazine Degradation in an Aerobic Groundwater Aquifer. Environ Sci Technol 50:7625-32. 47. Michalsen MM, King AS, Istok JD, Crocker FH, Fuller ME, Kucharzyk KH, Gander MJ. 2020. Spatially-distinct redox conditions and degradation rates following field-scale bioaugmentation for RDX-contaminated groundwater remediation. Journal of Hazardous Materials 387:121529. 48. Mahendra S, Alvarez-Cohen L. 2006. Kinetics of 1,4-dioxane biodegradation by monooxygenase-expressing bacteria. Environ Sci Technol 40:5435-42. 49. 50. 51. He Y, Mathieu J, Yang Y, Yu PF, da Silva MLB, Alvarez PJJ. 2017. 1,4-Dioxane biodegradation by Mycobacterium dioxanotrophicus PH-06 is associated with a group-6 soluble di-iron monooxygenase. Environmental Science & Technology Letters 4:494-499. Grostern A, Sales CM, Zhuang W-Q, Erbilgin O, Alvarez-Cohen L. 2012. Glyoxylate Metabolism Is a Key Feature of the Metabolic Degradation of 1,4-Dioxane by Pseudonocardia dioxanivorans Strain CB1190. Applied and Environmental Microbiology 78:3298. Gedalanga PB, Pornwongthong P, Mora R, Chiang SY, Baldwin B, Ogles D, Mahendra S. 2014. Identification of biomarker genes to predict biodegradation of 1,4-dioxane. Appl Environ Microbiol 80:3209-18. 11 Zhang S, Gedalanga PB, Mahendra S. 2016. Biodegradation Kinetics of 1,4-Dioxane in 52. Chlorinated Solvent Mixtures. Environ Sci Technol 50:9599-607. 53. Crocker FH, Indest KJ, Jung CM, Hancock DE, Fuller ME, Hatzinger PB, Vainberg S, Istok JD, Wilson E, Michalsen MM. 2015. Evaluation of microbial transport during aerobic bioaugmentation of an RDX-contaminated aquifer. Biodegradation 26:443-51. 54. 55. 56. 57. 58. 59. 60. Li M, Mathieu J, Liu Y, Van Orden ET, Yang Y, Fiorenza S, Alvarez PJJ. 2014. The Abundance of Tetrahydrofuran/Dioxane Monooxygenase Genes (thmA/dxmA) and 1,4-Dioxane Degradation Activity Are Significantly Correlated at Various Impacted Aquifers. Environmental Science & Technology Letters 1:122-127. He Y, Mathieu J, da Silva MLB, Li M, Alvarez PJJ. 2018. 1,4-Dioxane-degrading consortia can be enriched from uncontaminated soils: prevalence of Mycobacterium and soluble di-iron monooxygenase genes. Microbial Biotechnology 11:189-198. Ritalahti KM, Amos BK, Sung Y, Wu Q, Koenigsberg SS, Loffler FE. 2006. Quantitative PCR targeting 16S rRNA and reductive dehalogenase genes simultaneously monitors multiple Dehalococcoides strains. Appl Environ Microbiol 72:2765-74. Kotik M, Davidova A, Voriskova J, Baldrian P. 2013. Bacterial communities in tetrachloroethene-polluted groundwaters: A case study. Science of the Total Environment 454:517-527. Nemecek J, Dolinova I, Machackova J, Spanek R, Sevcu A, Lederer T, Cernik M. 2017. Stratification of chlorinated ethenes natural attenuation in an alluvial aquifer assessed by hydrochemical and biomolecular tools. Chemosphere 184:1157-1167. Simsir B, Yan J, Im J, Graves D, Loffler FE. 2017. Natural attenuation in streambed sediment receiving chlorinated solvents from underlying fracture networks. Environmental Science & Technology 51:4821-4830. Atashgahi S, Lu Y, Zheng Y, Saccenti E, Suarez-Diez M, Ramiro-Garcia J, Eisenmann H, Elsner M, Stams AJM, Springael D, Dejonghe W, Smidt H. 2017. Geochemical and microbial community determinants of reductive dechlorination at a site biostimulated with glycerol. Environmental Microbiology 19:968-981. 61. Miao Y, Johnson NW, Gedalanga PB, Adamson D, Newell C, Mahendra S. 2019. Response and recovery of microbial communities subjected to oxidative and biological treatments of 1,4- dioxane and co-contaminants. Water Research 149:74-85. 62. 63. 64. Douglas GM, Maffei VJ, Zaneveld JR, Yurgel SN, Brown JR, Taylor CM, Huttenhower C, Langille MGI. 2020. PICRUSt2 for prediction of metagenome functions. Nat Biotechnol 38:685- 688. Aßhauer KP, Wemheuer B, Daniel R, Meinicke P. 2015. Tax4Fun: predicting functional profiles from metagenomic 16S rRNA data. Bioinformatics 31:2882-2884. Quince C, Walker AW, Simpson JT, Loman NJ, Segata N. 2017. Shotgun metagenomics, from sampling to analysis. Nat Biotechnol 35:833-844. 12 CHAPTER 2 Abundance of Chlorinated Solvent and 1,4-Dioxane Degrading Microorganisms at Five Chlorinated Solvent Contaminated Sites Determined via Shotgun Sequencing This chapter is a modified version of a published work in Environmental Science & Technology: Hongyu Dang, Yogendra H. Kanitkar, Robert D. Stedtfeld, Paul B. Hatzinger, Syed A. Hashsham, and Alison M. Cupple Abundance of Chlorinated Solvent and 1,4-Dioxane Degrading Microorganisms at Five Chlorinated Solvent Contaminated Sites Determined via Shotgun Sequencing Environmental Science & Technology 2018 52 (23), 13914-13924. Abstract Shotgun sequencing was used for the quantification of taxonomic and functional biomarkers associated with chlorinated solvent bioremediation in twenty groundwater samples (five sites), following bioaugmentation with SDC-9. The analysis determined the abundance of 1) genera associated with chlorinated solvent degradation, 2) reductive dehalogenase (RDases) genes, 3) genes associated with 1,4-dioxane removal, 4) genes associated with aerobic chlorinated solvent degradation and 5) D. mccartyi genes associated with hydrogen and corrinoid metabolism. The taxonomic analysis revealed numerous genera previously linked to chlorinated solvent degradation, including Dehalococcoides, Desulfitobacterium and Dehalogenimonas. The functional gene analysis indicated vcrA and tceA from D. mccartyi were the RDases with the highest relative abundance. Reads aligning with both aerobic and anaerobic biomarkers were observed across all sites. Aerobic solvent degradation genes, etnC or etnE, were detected in at least one sample from each site, as were pmoA and mmoX. The most abundant 1,4-dioxane biomarker detected was Methylosinus trichosporium OB3b mmoX. Reads aligning to thmA or Pseudonocardia were not found. The work illustrates the importance of shotgun sequencing to provide a more complete picture of the functional abilities of microbial communities. The 13 approach is advantageous over current methods because an unlimited number of functional genes can be quantified. 1. Introduction The chlorinated solvents tetrachloroethene (PCE) and trichloroethene (TCE) and their metabolites, dichloroethene (DCE) and vinyl chloride (VC), are persistent groundwater contaminants, requiring remediation because of their risks to human health. Remediation efforts have involved biostimulation, through the addition of carbon sources, or bioaugmentation, which involves the injection of mixed microbial cultures containing Dehalococcoides mccartyi (1). D. mccartyi is a key microorganism for the complete transformation of these chemicals to the non- hazardous end product, ethene (2, 3). D. mccartyi strains reduce chlorinated compounds obtaining energy from the reduction process (4-6). Examples of commercially available mixed cultures containing D. mccartyi for chlorinated solvent remediation include SDC-9 (from APTIM, formerly CB&I, also marketed under several different names) and KB-1 (from SiREM) (1). It was estimated that several hundred sites in the US have been subject to bioaugmentation with cultures containing D. mccartyi (7). With the expansion of this remedial practice over the last decade, the number of sites in the US now numbers well over 2,300, and bioaugmentation has been performed in at least 11 other countries (P Hatzinger, pers comm). Following bioaugmentation, remediation professionals commonly monitor D. mccartyi populations, typically targeting reductive dehalogenase (RDase) genes such as vcrA (8-10) using quantitative PCR (qPCR) on nucleic acids extracted from groundwater (11-13). While qPCR has been successful for documenting the occurrence and dechlorinating activity of D. mccartyi (9, 12, 14, 15) most laboratories only have the instrumentation (bench-top real-time thermal cycler) to target a small number of functional genes. Next generation 14 sequencing (NGS) is now becoming the tool of choice for environmental samples. For example, 16S rRNA gene amplicon NGS (16S rRNA-NGS) has been used to monitor microbial communities during chlorinated solvent natural attenuation (16-18), following biostimulation (9, 19) (20-22), during zerovalent iron-based (22, 23) and thermal-based (24, 25) chlorinated solvent remediation. In contrast to 16S rRNA-NGS, shotgun (or whole genome) sequencing offers the opportunity to investigate both the taxonomic and the potential functional characteristics of microbial communities. However, only a limited number of researchers have adopted this approach for describing chlorinated solvent groundwater microbial communities. Notably, these studies have primarily focused on taxonomic data, without specifically addressing RDases or other functional genes related to chlorinated solvent degradation (26, 27). Others have examined dehalogenating genes in forest soils using shotgun sequencing (28). To our knowledge, the current work represents the first study to target contaminant degrading functional genes in groundwater from chlorinated solvent contaminated sites using shotgun sequencing. The samples included groundwater (from twenty injection or monitoring wells, post bioaugmentation with SDC-9) from five contaminated sites as well as the bioaugmentation culture, SDC-9. Although other researchers have used NGS to study D. mccartyi containing enrichment cultures (e.g. KB-1, D2, ANAS) (29, 30), limited data is available on SDC-9. The overall objective was to develop the methodology to quantify chlorinated solvent and 1,4- dioxane degrading microorganisms in contaminated site groundwater using both taxonomic and functional analyses. We propose that this approach (or a derivative) will ultimately be the method of choice for predicting biodegradation potential at contaminated sites. 15 2. Methods 2.1 DNA Extraction from Groundwater and SDC-9 Groundwater samples from injection and monitoring wells were collected at five different chlorinated solvent sites (San Antonio TX, Tulsa OK, Edison NJ, Quantico VA, and Indian Head MD) through traditional low-flow sampling (31). Only one of the five locations (Tulsa, OK) was known to be contaminated with 1,4-dioxane. The water was pumped into sterile amber bottles (1L), which were placed on ice and then shipped overnight to Michigan State University. All sites were previously bioaugmented with the commercially available reductive dechlorinating culture, SDC-9 (32, 33). Details concerning groundwater sampling times and site characteristics have been summarized (Supplementary Table 2.1). Additional site information, when available, has also been provided (e.g. plume maps, plot layouts, concentration data over time) for each site (Supplementary Figures 2.1-15). DNA was extracted (collection on a filter, bead-beating and chemical lysis) from groundwater and mixed culture (SDC-9) samples using the PowerWater DNA isolation kit (Mo Bio Laboratories, a Qiagen Company) and previously described methods (8, 34). 2.2 Sequencing and Taxonomic Analysis DNA extracts from twenty groundwater samples and the culture SDC-9 were submitted for library generation and sequencing to the Research Technology Support Facility Genomics Core at Michigan State University (MSU). Details on the preparation of libraries, the sequencing platform (Illumina HiSeq 4000) and the taxonomic analysis (Meta Genome Rapid Annotation using Subsystem Technology (MG-RAST) (35) are provided in the Supplementary Section (Supplementary Table 2.2). 16 2.3 Reference Sequences Collection, Functional Gene Analysis, qPCR Two approaches were employed to analyze the functional gene data. First, protein sequences associated with RDases for published genomes were collected from the National Center for Biotechnology Information (NCBI). The microorganisms and genome information utilized in this analysis has been summarized (Supplementary Tables 2.3 and 2.4). Secondly, to enable a wider number of sequences to be examined, protein sequences were collected from additional sources e.g. Functional Gene Pipeline and Repository (FunGene) (36), NCBI BLAST. DIAMOND (double index alignment of next-generation sequencing data ) (37) was used as the alignment tool for all functional genes. A stringent screening criteria approach (minimum sequence identity of 90% and alignment length of 49 amino acids) was adopted because of the similarity in many of the D. mccartyi genes (e.g. hydrogenases and corrinoid metabolism genes) between different strains. Detailed information on the collection of these sequences and the DIAMOND analysis has been provided (Supplementary Section). Quantitative PCR was performed to enumerate vcrA gene copies in each DNA extract using methods previously developed (34, 38) (see Supplementary Section). 3. Results 3.1 Sequencing and Taxonomic Analysis From the twenty groundwater DNA extracts, the majority (seventeen) generated between ~4 and ~6 million sequences each, post quality control. Three samples (PMW2, MWAW1, IW7) produced lower sequence counts (157,000, 471,513 and 1,547,247). The average sequence length varied from 226 to 241 bp (standard deviations from 34 to 41 bp) (Supplementary Table 2.2). The rarefaction curves plateaued indicating the analysis had captured the majority of the diversity within the samples (Supplementary Figure 2.16). 17 Sequencing analysis of SDC-9 indicated the genera Dehalococcoides (31% of all sequences) and Methanocorpusculum (10%) were major components of the culture (Supplementary Figure 2.17). Other important microorganisms included those within the phyla Bacteroidetes (23%, primarily the genera Parabacteroides and Bacteroides), Firmicutes (19%, primarily Desulfitobacterium, Desulfotomaculum, Clostridium and Bacillus) and Proteobacteria (9%). For the groundwater, between two and five samples were studied for each of the five sites, with Proteobacteria and Archaea being dominant in many samples (Supplementary Figures 2.18-22) 3.2 Occurrence of Chlorinated Solvent Degrading Microorganisms in SDC-9 and In Situ The sequencing data for each site was examined to determine the relative abundance of genera previously associated with chlorinated solvent degradation (Figure 2.1). It is important to note that this analysis is only at the genus level and therefore, except for Dehalococcoides, may overestimate the abundance of possible degrading microorganisms. Dehalobacter and Desulfomonile were not detected in any of the culture or groundwater samples by MG-RAST and are not included in Figure 2.1. 18 A C E 5 4 3 2 1 0 20 15 10 5 0 10 8 6 4 2 0 MW35 MW113 MW514 IWCW2 MWAW1 MW15R PMW2 PMW4 IW5 MW38 IW7 MW40 20 15 10 5 0 5 4 3 2 1 0 F 35 30 25 20 15 10 5 0 B MW2 IW3 MW3 IW4 MW4 IW6 D MW114 MW303S 0.5 0.4 0.3 0.2 0.1 0 SDC9-1 SDC9-2 Figure 2. 1. Relative abundance (%, as determined using MG-RAST) of methanotrophs and genera associated with chlorinated solvent biodegradation in groundwater from San Antonio (A), Tulsa (B), Quantico (C), Edison (D), Indian Head (E) and SDC-9 (F). The genus Dehalococcoides was present in all groundwater samples ranging from 0.1 – 3.5%. Note,"MW" in name refers to a monitoring well and "IW" in name refers to an injection well. The insert in F does not include Dehalococcoides or Desulfitobacterium to enable a y-axis with a different scale. 19 The relative abundance of methanotrophs in the groundwater samples was also investigated (Figure 2.1). Methanotrophs are important because of their ability to use particulate and soluble methane monoxygenases (pMMO and sMMO) to cometabolically oxidize several chlorinated solvents (39-41). Dehalococcoides, the key dechlorinating genera in SDC-9 (31% in SDC-9), was detected in every sample at every site (averages for each site ranging from 0.2 to 1.4%). The sites had been bioaugmented with SDC-9 from ~ 6.5 months (Quantico) to more than 6 years (Edison) prior to groundwater sample collection (Supplementary, Table 1). The abundance of Dehalococcoides was greater in the injection wells (IW3, IW4, IW5, IW, CW2) compared to the monitoring wells (Figure 2.1B, C). Dehalococcoides relative abundance levels (0.14-0.26%) were lowest at the Edison site (Figure 2.1D) which had the longest time between bioaugmentation and sample collection (76 months). The lower Dehalococcoides levels at the Quantico site (0.15-0.19%, Figure 2.1C) are puzzling, since it had the shortest time between bioaugmentation and sampling (6.5 months), and may be related to the electron donor utilized (hydrogen compared to a fermentable substrate). At the Tulsa site, Dehalococcoides relative abundance levels were on the higher side (monitoring wells, 0.44 -0.96%, Figure 2.1B), perhaps as a result of higher TCE concentrations at the time of sampling (Supplementary Figure 2.12). Dehalococcoides abundance levels were also higher at the Indian Head site (0.40-0.75%), possibly related to a shorter time between bioaugmentation and sampling (9 months). Desulfitobacterium was detected at all five sites, although the relative abundance (average ranging from 0.1 to 0.4%) was typically less than that of Dehalococcoides. Except for Dehalococcoides, Desulfitobacterium was present at a higher relative abundance in SDC-9 (2.7%) compared to other dechlorinating microorganisms (<0.4%). At three sites, Geobacter was 20 the most abundant genus in this group (Figure 2.1A, B and C) and at two sites, it was either the second or third most abundant (Figure 2.1D and E). The five methanotrophs examined were present only at low levels in SDC-9 (averages ranging from 0.006-0.035%). In the groundwater samples, Methylococcus or Methylosinus were typically the most abundant, followed by Methylobacterium and Methylocella. 3.3 Functional Gene Analysis The groundwater sequencing data were aligned to characterized RDases from D. mccartyi and three other genera (Dehalogenimonas, Dehalobacter and Desulfitobacterium) (Figure 2.2). Not surprisingly, RDases from D. mccartyi were the most abundant (Figure 2.2A). Samples from Tulsa illustrated some of the highest values for tceA and vcrA, again a pattern perhaps caused by the higher chlorinated ethene concentrations at this site (Supplementary Table 2.1, Supplementary Figure 2.12). Following Tulsa, the wells at Indian Head contained the second most abundant reads aligning to RDases from D. mccartyi. These results agree with the MG- RAST analysis, which illustrated the highest relative abundance of Dehalococcoides at Indian Head and Tulsa (Figure 2.1B and E). The abundance of RDases from Dehalogenimonas, Dehalobacter and Desulfitobacterium were found in lower numbers and the results varied between sites (Figure 2.2B, C, D). The majority of reads aligning with cerA and tdrA from Dehalogenimonas were from Tulsa (MW2, MW3, MW4, IW3, IW4, IW6), followed by Indian Head (IW5, IW7, MW38, MW40) and Edison (MW303S) (Figure 2.2B). The average relative abundance values for Dehalogenimonas from the MG-RAST analysis indicated the highest values for San Antonio, Edison and Indian Head (Figure 2.1A, D, E). Reads aligning to RDases from Dehalobacter and Desulfitobacterium were less abundant but were found in at least one well from three of the five sites (except San 21 Antonio and Edison) (Figure 2.2C, D). Although Desulfitobacterium was detected with the MG- RAST analysis, Dehalobacter was not. Additional differences between the MG-RAST and the functional gene data sets included the presence of the genera Anaeromyxobacter and Sulfurospirillum with MG-RAST, but the absence of functional genes (associated with the removal of chlorinated chemicals) from these microorganisms. Also, Geobacter and Polaromonas were present at all sites, however, reads aligning to pceA of Geobacter lovleyi and cytochrome P450 of Polaromonas JS666 were observed from only one sample each (MW40 and MW4, respectively, data not shown). These findings emphasize the importance of functional gene analysis to clearly define in situ potential biodegradation capabilities. The majority of the RDases found in SDC-9 were from D. mccartyi, with tceA and vcrA being the most abundant (~two orders of magnitude higher than the RDases from other species) (inserts in Figure 2.2). RDases from Dehalogenimonas, Dehalobacter, Desulfitobacterium were also present in SDC-9. Reads aligning to the genes associated with the aerobic degradation of 1,4-dioxane (42) were also investigated (Figure 2.3). From the twelve genes examined, only six were identified in the groundwater samples (Figure 2.3A). These genes were detected in at least one sample from all five sites, despite the fact that only one of the sites (Tulsa) was known to be contaminated with 1,4-dioxane. Surprisingly, no genes associated with Pseudonocardia were detected. The MG-RAST taxonomic data were examined for the presence of the genera associated with these genes 22 ) % ( e c n a d n u b A e v i t a l e R d e z i l a m r o N 1E-5 0E+0 B cerA tdrA 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I D 1E-3 8E-4 4E-4 0E+0 1E-4 8E-5 6E-5 4E-5 2E-5 0E+0 1E-2 5E-3 0E+0 1 - 9 C D S 2 - 9 C D S 4 W I 6 W I 2 W C W I 2 W M P S 3 0 3 W M probable chlorophenol RDase 1 W A W M R 5 1 W M 4 1 1 W M 4 W M P 5 W I 7 W I 8 3 W M 0 4 W M dcaA cprK/cprC like protein pceC/pceT pceA 4E-3 2E-3 0E+0 1 - 9 C D S 2 - 9 C D S 6E-3 A bvcA tceA vcrA pceA 4E-3 2E-3 0E+0 0.08 0.06 0.04 0.02 0.00 1 - 9 C D S 2 - 9 C D S 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 2E-5 C cprA pceA 4.E-4 2.E-4 0.E+0 1 - 9 C D S 2 - 9 C D S 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M Figure 2. 2. Normalized relative abundance (%, as determined by DIAMOND) of genes associated with reductive dechlorination in Dehalococcoides mccartyi (A), Dehalogenimonas spp. (B), Dehalobacter spp. (C) and Desulfitobacterium spp. (D) in SDC-9 (inserts) and in groundwater from the five chlorinated solvent sites. The highest abundance values are from tceA and vcrA from Dehalococcoides, followed by cerA from Dehalogenimonas. 23 ) % ( e c n a d n u b A e v i t a l e R d e z i l a m r o N 4E-3 3E-3 2E-3 1E-3 0E+0 A Methylosinus trichosporium OB3b mmoX Pseudomonas mendocina KR1 tmoA Pseudomonas pickettii PKO1 tbuA1 Burkholderia cepacia G4 tomA3 Rhodococcus jostii RHA1 prmA Rhodococcus sp. RR1 prmA Pseudonocardia tetrahydrofuranoxydans thmA Pseudonocardia sp. ENV478 thmA Pseudonocardia dioxanivorans CB1190 thmA Rhodococcus sp. YYL thmA Mycobacterium sp. ENV421 prmA Mycobacterium dioxanotrophicus PH-06 prmA B ) % ( e c n a d n u b A e v i t a l e R 40 30 20 10 0 Pseudomonas Mycobacterium Rhodococcus Burkholderia Methylosinus Pseudonocardia 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W C 1 W A 2 W M P R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 1 - 9 C D S 2 - 9 C D S 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W C 1 W A 2 W M P R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 1 - 9 C D S 2 - 9 C D S Figure 2. 3. Normalized relative abundance (%, determined with DIAMOND) of genes (A) and relative abundance (%, determined with MG- RAST) of genera (B) previously associated with 1,4-dioxane degradation in all groundwater samples and in SDC-9. The relative abundance of Pseudonocardia was zero in all groundwater samples and in SDC-9. Methylosinus trichosporium OB3b mmoX was the dominant 1,4-dioxane degrading gene in the majority of the groundwater samples. 24 (Figure 2.3B). From this group, Pseudomonas was the most dominant genus, followed by Burkholderia, Mycobacterium, Methylosinus and Rhodococcus. Similar to the functional gene data, the genus Pseudonocardia was not detected in any groundwater sample. The shotgun data sets were also queried against reference databases that contained both RDases from complete genomes as well as those from uncultured microorganisms (Figure 2.4A). The results were consistent with those found using sequences from complete genomes only (Figure 2.2A). Reads aligning with the genes associated with the aerobic degradation of the chlorinated ethenes (pmoA, mmoX, etnC, etnE) (40, 41, 43) were detected in the groundwater samples from a number of samples from Edison and Indian Head (Figure 2.4B, C). Additionally, etnC and etnE were also found at high levels in the monitoring wells from the Tulsa site, again perhaps as a result of higher cVOC concentrations at the time of sampling. Notably, the highest normalized relative abundance values for etnC and etnE were two orders of magnitude lower than vcrA or tceA. 25 6E-3 4E-3 2E-3 0E+0 8E-05 6E-05 4E-05 2E-05 0E+00 ) % ( e c n a d n u b A e v i t a l e R d e z i l a m r o N A bvcA vcrA tceA pceA 0.06 0.04 0.02 0.00 1 - 9 C D S 2 - 9 C D S 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M C etnE etnC 4E-04 B pmoA mmoX 3E-04 2E-04 1E-04 0E+00 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 1 - 9 C D S 2 - 9 C D S 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 1 - 9 C D S 2 - 9 C D S Figure 2. 4. Normalized relative abundance (%, determined with DIAMOND) of genes associated with the chlorinated solvent reductive dechlorination (A) and the aerobic degradation of the chlorinated solvents (B, C) in SDC-9 (insert for A) and in groundwater from the five chlorinated solvent sites. The aerobic genes occurred at lower levels compared to the anaerobic genes. Note, the analysis approach differed from the approach used to generated Figure 2.2, in that all sequences from the databases were compared to each dataset. 26 The DIAMOND analysis included alignments to a gene encoding for a formate dehydrogenase-like protein (fdhA), hydrogenase genes (hup, vhc, hym and ech) and corrinoid metabolism genes (btu, cbi and cob) from D. mccartyi. In previous research, the formate dehydrogenase-like protein was found to be highly expressed and ubiquitous in D. mccartyi, representing a specific indictor for activity (44). Hydrogenases are thought to oxidize H2, the electron donor for D. mccartyi (45). Corrinoid metabolism genes are relevant for up-taking and transforming of cobamides and cobinamide, which are critical for D. mccartyi RDases (45). Samples containing the most abundant reads of fdhA were from Tulsa following by samples from Indian Head (Supplementary Figure 2.23). The abundance patterns for the hydrogenase and corrinoid metabolism genes across samples were similar to those for vcrA, tceA and fdhA (Supplementary Figures 2.24 and 2.25). The fdhA abundance patterns across samples were similar to those observed for tceA and vcrA (Spearman’s rank correlation coefficients 0.939 and 0.89 for fdhA vs. vcrA and fdhA vs. tceA, respectively, p values both < 0.0001, Supplementary Figure 2.26), indicating this gene acts as an effective biomarker for D. mccartyi. To investigate the accuracy of the shotgun sequencing data quantification method, the relative abundance of vcrA determined via shotgun sequencing was compared to vcrA gene copies determined via qPCR (Supplementary Figure 2.27). In general, the abundance of vcrA determined using shotgun sequencing correlated well (Spearman’s rank correlation coefficient 0.808, p value < 0.0001) with the qPCR data (3.9 X 104 to 7.0 X 109 vcr gene copies per L). Principal component analyses were completed for the functional genes (Figure 2.5A) and genera (Figure 2.5B) associated with chlorinated solvent and 1,4-dioxane biodegradation. The genes tdrA, vcrA and tceA were positively correlated to fhdA as well as the hydrogenase and corrinoid metabolism genes, consistent with their similar abundance distribution in the wells. 27 These genes correlated with injective wells from the Tulsa site, which would be expected considering the high relative abundance of Dehalococcoides in these samples. Genes relevant to aerobic chlorinated ethene degradation correlated with mmoX (from M. trichosporium OB3b) suggesting the genetic potential for degradation of these co-contaminants occurs at the same site. In this case, the genes correlated with MW114 from the Edison site. The remaining genes associated with 1,4-dioxane degradation correlated together (bottom left quadrant) perhaps indicating multiple functional genes will contribute to 1,4-dioxane degradation at the same site. RDases (pceA) from Desulfitobacterium and Dehalobacter also correlated together, along with MW3 (from Tulsa) which was previously found to contain RDases from these genera. For the taxonomic principal component analysis (Figure 2.5B), the anaerobic genera Dehalococcoides, Desulfitobacterium and Desulfuromonas correlated together along with the injection wells from the Tulsa site. For the methanotrophs, Methylococcus and Methylobacterium illustrated a positive correlation to each other and to the wells from several sites e.g. MW114, MW303, PMW4. The genera PCA is less meaningful because it is unknown if the majority of these microorganisms are truly associated with contaminant degradation. 28 A 10 mmoX (OB3b) mmoX pmoA etnE Active variables Active observations ) % 0 0 . 4 1 ( 2 F 5 0 -5 -10 ) % 5 8 . 5 1 ( 2 F 6 5 4 3 2 1 0 -1 -2 -3 -4 MW114 etnC bvcA MW303S tdrA MW38 MW514 MW4 MWAW1 PMW4 MW113 PMW2 cprK/cprC like IW7 MW35 protein MW15R MW40 IW5 pceA MW2 cprA IW4 IW3 IW6 hymABCD tceA vcrA cobABCDQSTU hupLS fdhA btuFCD cbiABZ echABCEF vhcAG tmoA tbuA1 IWCW2 tomA3 probable MW3 chlorophenol pceC/pceT RDase RR1 prmA RHA1 prmA Dehalobacter pceA Desulfitobacterium -6 -4 -2 0 pceA 2 4 6 8 10 12 F1 (37.77 %) Active variables Active observations Methylacidiphilum MW38 B Anaeromyxobacter Dehalogenimonas Desulfuromonas MW35 Desulfitobacterium MW113 IW4 Dehalococcoides IW3 IW6 MW514 Methylosinus Methylocella Methylobacterium Methylococcus MW303S PMW4 MW114 MW40 Nocardioides Geobacter Sulfurospirillum MW3 IWCW2 IW7 Pseudomonas MW2 MW4 Burkholderia PMW2 Mycobacterium Polaromonas Propionibacterium Rhodococcus MWAW1 IW5 Mycobacterium -6 -4 -2 0 2 4 6 F1 (29.82 %) MW15R Figure 2. 5. Principle component analyses of functional genes (A) and genera (B) associated with chlorinated solvent and 1,4-dioxane biodegradation in all groundwater samples. 29 4. Discussion Biostimulation and bioaugmentation are becoming increasingly popular approaches for the remediation of groundwater contaminated with PCE, TCE and their daughter products. However, limited research has focused on groundwater microbial communities post bioaugmentation. This work is important because of the requirement for Dehalococcoides to co-exist with other “supporting” microorganisms and to survive over time. Further, it is also valuable to determine if other chlorinated solvent degrading microorganisms are present, and the extent to which these organisms persist following bioaugmentation with exogenous strains. Not surprisingly, the genus Dehalococcoides was a major component of SDC-9. This was also reported for another common bioaugmentation culture, KB-1 (29). Of additional interest is the presence (4% relative abundance) of Desulfitobacterium in SDC- 9, as this genus has also been associated with dechlorination (46-50). Similarly, others have reported Desulfitobacterium type RDase genes in Dehalococcoides enriched cultures (29). Other genera linked to chlorinated solvent degradation were also detected in SDC-9 (as discussed above); however, their relative abundance in the culture was low compared to Dehalococcoides or Desulfitobacterium. As in other Dehalococcoides enrichment cultures, SDC-9 contained methanogens (Methanocorpusculum), acetogens (Clostridiaceae) and Geobacter (9, 29). Geobacter has previously been associated with interspecies corrinoid transfer with Dehalococcoides (51). In addition, Geobacter has also been associated with dechlorination (52, 53). The genera Thermosinus and Selenomonas within the family Veillonellaceae were detected in SDC-9 at low levels (3% and 0.4%, respectively). Veillonellaceae were previously found to be important corrinoid supplying microorganisms to Dehalococcoides in another 30 enrichment culture (54). SDC-9 contained Desulfovibrio (2.5%), which, in previous research, was linked to more robust dechlorination rates and growth when grown in co-culture with Dehalococcoides (55). It was reported that Desulfovibrio can support Dehalococcoides by providing acetate, hydrogen and corrinoid cofactors (55). Following Dehalococcoides and Methanocorpusculum, the third and fourth most abundant genera in SDC-9 were Bacteroides (5.4%) and Parabacteroides (10%) (within Bacteroidetes). Members of the Bacteroidetes phylum have also been reported as important bacteria in other dechlorinating mixed cultures (29) and in contaminated groundwater (9). For the groundwater samples, Geobacter was more abundant at all sites compared to the SDC-9 culture and may therefore be important in playing a supportive role for Dehalococcoides at contaminated sites. In contrast, Thermosinus and Selenomonas were not detected in any groundwater samples. Other potentially supportive microorganisms, including Desulfovibrio, Bacteroides and Parabacteroides, were detected in the groundwater at all sites (ranging from 0.1- 4.2%) and therefore may also play a supportive role for Dehalococcoides in situ. Similar to many previous studies examining microbial communities at chlorinated solvent sites (undergoing some kind of bioremediation), the genera Dehalococcoides, Dehalogenimonas and Geobacter were found in groundwater from all five sites (9, 19-27). The current study identified Desulfitobacterium and Anaeromyxobacter in the majority of samples and these genera have also been frequently detected at contaminated sites (17, 20-22, 24). In contrast, fewer previous studies have reported the presence of Polaromonas and Nocardioides (16, 17, 25). Previous researchers have also detected methanotrophs in situ (16, 18, 24). It was surprising that Dehalobacter was absent in the MG-RAST data, as this genus has been commonly reported in groundwater from chlorinated solvent contaminated sites (17, 19, 21, 22, 24). However, cprA 31 and pceA from Dehalobacter were found in the functional gene analysis, suggesting this genus could be present, but at levels undetectable by the MG-RAST analysis. Although taxonomic data is important for characterizing microbial communities in situ, it is well recognized that certain limitations are associated with such data. A key limitation concerns an inability to classify to the species level when short sequences are analyzed. This issue is particularly relevant to bioremediation applications, as it impacts an identification of a known degrader, e.g. Polaromonas JS666 (56) or Geobacter lovleyi (57), over others in the same genus that are not capable of contaminant degradation. In the current study, relying on taxonomic data alone would have been misleading, because although the genera Polaromonas and Geobacter were present, the functional genes were largely absent (P450 from Polaromonas JS666 and pceA from Geobacter were detected only once). Another related limitation concerns the inability of taxonomic data to provide in-depth information on function. This concern is important when considering D. mccartyi, as strains with similar 16S rRNA gene sequences may contain different RDases. Clearly, to generate a full picture of the functional abilities of microorganisms to degrade contaminants in situ, both taxonomic and functional analyses are needed. The taxonomic and functional analysis detected both aerobic and anaerobic biomarkers across the five sites. For example, both vcrA and etnC were found in MW2, MW3, MW4 from the Tulsa site (although the values for vcrA were higher). This trend has previously been noted for groundwater from other chlorinated solvent sites (13, 16, 58). The genes etnC or etnE were detected in at least one groundwater sample from each site, with the normalized relative abundance values covering a wide range. Similarly, pmoA and mmoX were detected in at least one groundwater sample from each site and 32 were particularly abundant at the Edison site. Given the occurrence of these genes in the current study, future research directions should include a consideration of both aerobic and anaerobic genes when accounting for chlorinated solvent removal rates. To our knowledge, this study represents the first analysis of the genes associated with 1,4-dioxane degradation in groundwater using shotgun sequencing. Here, from the twelve sequences investigated, the most abundant number of reads (collectively, in all groundwater samples) aligned to Methylosinus trichosporium OB3b mmoX, followed by Burkholderia cepacia G4 tomA3 and Pseudomonas pickettii PKO1 tbuA1. Notably, although mmoX from M. trichosporium OB3b has been associated with 1,4-dioxane degradation at high concentrations (59), at low, environmental relevant concentrations, no removal was observed (60). Three others (Pseudomonas mendocina KR1 tmoA, Rhodococcus jostii RHA1 prmA, Rhodococcus sp. RR1 prmA) were detected at lower levels in at least one well from each site. In some cases, remarkably, the normalized relative abundance values were in the same range as those for vcrA and tceA, even though 1,4-dioxane was not previously reported at 4 of the 5 sites, and reducing conditions (i.e., negative oxidation-reduction potential; nORP) generally prevailed. Previously, others have observed thmA in samples from 1,4-dioxane contaminated sites using qPCR primers designed to thmA from Pseudonocardia (61-63). However, reads aligning to thmA were not found in the current study. The taxonomic data confirmed this finding, as the genus Pseudonocardia was absent from the MG-RAST results. Reads aligning to Mycobacterium 1,4- dioxane degrading gene sequences (prmA) were also not detected in the current study, even though the taxonomic MG-RAST data indicated this genus was present. This discrepancy again illustrates the importance of functional gene data to corroborate taxonomic data and assumptions about function. Further, the current work illustrates the importance of shotgun sequencing to 33 provide a more complete picture of the potential of in situ microbial communities to degrade 1,4-dioxane compared to qPCR, which typically only targets a small number of genes. Previous research indicated that transcripts of the proteins Fdh and Hup may be better indicators of cell respiration compared to RDases (64, 65). In fact, it was concluded that HupL transcripts were the most robust activity biomarker across multiple D. mccartyi strains (66). Given importance of Hup, the relative abundance of fdhA and other genes encoding for hydrogenases from D. mccartyi were investigated in the groundwater samples. Building on the approach developed in the current study, future research could include shotgun sequencing of transcripts to obtain an improved indicator of D. mccartyi cell respiration. These gene targets, as well as those involved in corrinoid metabolism, could be used as additional biomarkers for D. mccartyi. To examine the quantitative robustness of the data generated, the normalized relative abundance values for vcrA were compared to those obtained via TaqMan qPCR. The correlation indicated the methods produced similar values across a range of concentrations for the five sites. Two important future research directions for using shotgun sequencing for bioremediation applications will be 1) to determine detection limits and 2) to generate more in depth comparisons to values determined with qPCR. In summary, methods were developed to determine the abundance of genes associated with chlorinated solvent and 1,4-dioxane biodegradation in groundwater samples from multiple samples from multiple contaminated sites. The use of shotgun sequencing enabled a larger selection of genes to be targeted compared to traditional qPCR. In fact, the number of functional genes that can be analyzed is limitless. The 34 method also does not require primer design or primer assay verification for each target (as is the case for qPCR). The most labor-intensive part of the approach involved the collection of reference fasta files for the DIAMOND alignment (following this, all remaining steps were not time consuming). The sequencing price is perhaps the largest limitation to the method. In the current study, for 22 samples, the cost was approximately $210 per sample. However, it is likely that sequencing costs will drop as the technology evolves, making the approach more attractive. The data indicated the presence of both aerobic and anaerobic biomarkers for chlorinated solvent degradation. Not surprisingly, the taxonomic data alone was insufficient to determine the functional abilities of these communities. The relative abundance of hydrogenases and corrinoid metabolism genes suggest these may be appropriate additional biomarkers for D. mccartyi. The approach developed will enable researchers to investigate the abundance of any contaminant degrading gene in any sample, greatly expanding the analytical toolbox for natural attenuation, biostimulation or bioaugmentation. Acknowledgements Thanks to Simon Vainberg, Sheryl Streger, Robert E. Mayer, Michael Martinez and David Lippincott from APTIM Federal Services for providing groundwater samples. Thanks to James Cole and Benli Chai from RDP (MSU). Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Views, opinions, and/or findings contained in this report are those of the authors and should not be construed as an official Department of Defense position or decision unless designated by other official documentation. 35 APPENDIX 36 APPENDIX Text A1. Background on Functional Genes The abundance of the genes associated with reductive dechlorination were examined from the genera Dehalococcoides, Dehalogenimonas, Desulfitobacterium, Dehalobacter, Geobacter, Sulfurospirillum and Anaeromyxobacter. The genes encoding for enzymes associated with aerobic VC degradation were also targeted (etnC/alkene monooxygenase and etnE/epoxyalkane:CoM transferase) (67-71). Also, the gene encoding for cytochrome P450 from Polaromonas JS666 was investigated, as this initializes the degradation of cis-1,2-dichoroethene (72). The genes encoding for the α subunits of soluble and particulate methane monooxygenases (mmoX and pmoA) were examined due to their role in chlorinated ethene degradation (41, 73, 74). The genes encoding the enzymes associated with 1,4-dioxane biodegradation (as summarized in (42)) were also investigated. This chemical is a probable human carcinogen (75) and is frequently detected at sites where the chlorinated solvents are present (76-78). Finally, the genes encoding for enzymes associated with hydrogen metabolism (fdhA, hupSL, vhcAG, hymABCD and echABCEF) and corrinoid metabolism (btuFCD, cobA, cobB, cobC, cobD, cobQ, cobS, cobT, cobU, cbiA, cbiB and cbiZ) in D. mccartyi were also quantified (45). Methods Text A2. Collection of Sequences for Functional Genes A primary source for RDases was the Functional Gene Pipeline and Repository (FunGene) (36) website using the link ‘vcrA_ver2’. For this, the contents (e.g. score, protein and nucleotide accession numbers, microorganism name, length of the protein) of the topmost 3000 sequences from ‘vcrA_ver2’ were exported into excel. Then, accession number lists of RDases for each 37 species were created by setting a filter in Excel. Each accession number was checked for accuracy using complete genome information (NCBI) (Supplementary Table 2.3). The protein sequence fasta files were also downloaded from FunGene (again by selecting the same 3000 sequences of ‘vcrA_ver2’). The accession number RDase lists were used to collect RDases reference protein sequences from the protein sequence fasta files (from FunGene) using a tool (Readseq.jar) developed by Ribosomal Database Project (https://github.com/rdpstaff/RDPTools). The overall process produced individual RDase protein sequence files for each microorganism. A number of microorganisms (including Dehalococcoides mccartyi JNA, SG1, Sulfurospirillum strains, Anaeromyxobacter dehalogenans 2CP-C, Geobacter lovleyi SZ, Dehalobacter E1 and FTH1, Desulfitobacterium sp. PCE1, Desulfitobacterium hafniense TCP-A and PCP-1 and Polaromonas sp. JS666) did not have functional gene data on FunGene, therefore their reference sequences were collected manually by downloading the fasta files from NCBI complete genomes (Supplementary Table 2.3). The vcrA reference sequences (39 vcrA sequences, protein fasta files) were collected from the link ‘vcrA’ in FunGene by selecting those sequences with a score higher than 900 (Hidden Markov Model score alignment by FunGene). The tceA and bvcA reference sequences were collected using both NCBI and FunGene. Protein sequences (AAN85588, AAT64888) previously used for designing tceA and bvcA primers (38) were first used to collect sequences from the NCBI database using BLAST (79). Sequences with a maximum score higher than 900 from the BLAST search were collected (31 tceA sequences, 13 bvcA sequences). In FunGene, using the ‘Probe Match Search’ function, primers for tceA (TceA1270F, TceA1336R and TceA1294Probe) and bvcA (Bvc925F, Bvc1017R and Bvc977Probe) (38) were used to search the first 3000 sequences of ‘vcrA_ver2’, producing 38 tceA sequences and 11 bvcA sequences. 38 The tceA and bvcA sequences from the two sources were compared. The tceA sequences from NCBI (except ADV18463) were all present in the tceA sequences obtained from FunGene. Therefore, the final tceA reference list consisted of the sequences from FunGene along with ADV18463. A similar approach was used for generating the bvcA reference list. Sequences for pceA (5 sequences) and fdrA (30 sequences) from Dehalococcoides mccartyi were collected by downloading the fasta files from the NCBI complete genomes. The sequence for the putative VC RDase (cerA) from Dehalogenimonas (80) was kindly provided by Dr. Frank Loeffler (Locus Tag JP09_004725, Protein ID PPD58423.1). Reference sequences for etnC and etnE (31 etnC sequences, 95 etnE sequences) were collected from FunGene using scores higher than 700 and 500, respectively. Additionally, primers for etnC (RTC_F (etnC) and RTC_R (etnC)) and etnE (RTC_F (etnE) and RTC_R (etnE)) (43) were used with the ‘Probe Match Search’ function in FunGene to search for sequences in all pages of ‘etnC’ and ‘etnE’, resulting in 9 etnC sequences, 31 etnE sequences. Reference sequences for etnC and etnE (20 etnC sequences, 53 etnE sequences) were also collected from UniProt. The final etnC and etnE reference sequences were generated by combining all data sets discussed above. mmoX and pmoA reference sequences (21 mmoX sequences, 30 pmoA sequences) were first collected using ‘mmoX’ and ‘pmoA’ links in FunGene (sequences with a score higher than 980 and 500, respectively). Additionally, all other sequences annotated as ‘mmoX’ or ‘soluble methane monooxygenase’ in all pages of mmoX in FunGene were also collected. For this, information, such as score, protein and nucleotide accession number, name of the microorganism, length of the protein, was imported to excel. Then, a filter in excel was set for the name of the gene to create an accession number list for mmoX. The accession number list was 39 used to collect reference protein sequences from protein sequences downloaded from FunGene using Readseq.jar (generating 580 sequences). Sequences annotated as ‘pmoA’ or ‘particulate methane monooxygenase’ were also collected using methods similar to those described from mmoX (generating 8327 sequences). A list of functional genes (12 sequences) associated with 1,4-dioxane metabolism or cometabolism was obtained from a recent publication (42). The protein sequences of these genes were then collected from NCBI. The functional genes associated with hydrogen and corrinoid metabolism in D. mccartyi were also examined. Reference sequences for all hydrogenase (hupLS, vhcAG, echABCEF and hymABCD) and corrinoid (btuFCD, cbiABZ and cobABCDQSTU) metabolism genes were collected by using NCBI BLAST search. Additional information on the collection of sequences associated with hydrogen and corrinoid metabolism is provided in the supplementary section. Sequences ACZ61293.1 and ACZ61294.1 from D. mccartyi VS were used for starting the BLAST search for hupL and hupS, separately. Then hupL and hupS reference sequences (13 hupL sequences, 6 hupS sequences) were collected with an identity > 95% and >94%, respectively. All identity values were selected because of the large identity decrease after the last selected reference sequences. vhcA and vhcG reference sequences (9 vhcA sequences, 8 vhcG sequences) were collected with an identity > 90%. The sequences used for the BLAST search were ACZ61705.1 and ACZ61704.1 from D. mccartyi VS. hymA1 and hymA2 reference sequences (4 hymA1 sequences, 5 hymA2 sequences) were collected with an identity > 98% and >96%, respectively. The sequences used for starting the BLAST search were ACZ61326.1 and ACZ61777.1 from D. mccartyi VS. hymB1 and hymB2 (3 hymB1 sequences, 19 hymB2 sequences) were collected with an identity > 98% and > 97%, respectively. The sequences used 40 for starting the BLAST search were ACZ61327.1 and ACZ61778.1 from D. mccartyi VS. hymC1 and hymC2 (15 hymC1 sequences, 15 hymC2 sequences) were collected with an identity >96% and >87%, respectively. The sequences used to start the BLAST search were ACZ61328.1 and ACZ61779.1 from D. mccartyi VS. hymD1 (11 hymD1 sequences) was collected with an identity > 89%. The sequence used to start the BLAST search was ACZ61329.1 from D. mccartyi VS. Additional hymABC genes were found in D. mccartyi 195 and following the similar nomenclature for the genes, they were named hymA3, A4, B3 and C3. hymA3 and hymA4 (9 hymA3 sequences, 13 hymA4 sequences) were collected with identities > 98% and > 93%, respectively. The sequences for the BLAST search were AAW39863.1 and AAW40249.1. hymB3 (18 sequences) was collected with an identity > 94%. The sequence used for starting the BLAST search was AAW39862.1 hymC3 (13 sequences) was collected with an identity > 90%. The sequence used for starting the BLAST search was AAW39861.1 The sequences used for starting BLAST search for echABCEF were from D. mccartyi CBDB1 with accession number of CAI82985.1, CAI82986.1, CAI82987.1, CAI82992.1 and CAI82993.1. echABCEF reference sequences (23 echA sequences, 16 echB sequences, 7 echC sequences, 10 echE sequences, 11 echF sequences) were collected with an identity > 92%, 94%, 94%, 96% and 84%, respectively. The sequences used for starting BLAST search for btuFCD were from D. mccartyi DCMB5 with accession number of AGG06280.1, AGG06281.1 and AGG06282.1. btuFCD reference sequences (17 btuF sequences, 14 btuC sequences, 6 btuD sequences) were collected with an identity > 89%, 93% and 93%, respectively. The sequences used for starting BLAST search for cbi were from D. mccartyi VS. cbiA and cbiB reference sequences (16 cbiA sequences, 13 cbiB sequences) were collected both with 41 an identity > 90%. The sequences used for starting the BLAST search were ACZ61308.1 and ACZ61741.1. There were four cbiZ sequences from D. mccartyi VS (hereafter named cbiZ1234). The accession number of the sequences of cbiZ1234 for the BLAST search were ACZ61242.1, ACZ61249.1, ACZ61740.1 and ACZ62455.1. cbiZ1234 reference sequences (16 cbiZ1 sequences, 11 cbiZ2 sequences, 10 cbiZ3 sequences, 46 cbiZ4 sequences) were collected with an identity > 72%, 97%, 92% and 87%, respectively. The majority of the sequences used for starting BLAST search for cob were from D. mccartyi VS, with one from cobC from D. mccartyi CBDB1. The accession number of sequences used for starting BLAST search for cobA123 were with of AAW40449.1, AAW39561.1 and AAW39547.1. cobA123 reference sequences (8 cobA1 sequences, 14 cobA2 sequences, 14 cobA3 sequences) were collected with an identity > 91%, 92% and 94%, respectively. The accession number of sequences used for starting BLAST search for cobBCQ were with of AAW40541.1, CAI82815.1 and AAW39791.1. cobBCQ reference sequences (16 cobB sequences, 10 cobC sequences, 20 cobQ sequences) were collected with an identity > 89%, 90% and 92%, respectively. cobD1 and cobD4 reference sequences (11 cobD1 sequences, 17 cobD4 sequences) were collected with an identity > 84% and 80%, respectively. The sequences used for starting the BLAST search were AAW40448.1 and AAW39562.1. cobS1, cobT1 and cobU1 reference sequences (14 cobS1 sequences, 12 cobT1 sequences, 14 cobU1 sequences) were collected with an identity > 90%, 94% and 90%, respectively. The sequences used for starting the BLAST search were AAW40093.1, AAW40094.1 and AAW40091.1. 42 The BLAST search of cobD2 and cobD3 generated the same results as cbiB. Also, the BLAST results of cobS2, cobT2 and cobU2 were the same as those of cobS1, cobT1 and cobU1, respectively. Therefore, the results of cobD2, cobD3 cobS2, cobT2 and cobU2 were not included in the analysis. Text A3. Library Preparation, Sequencing, MG-RAST and DIAMOND analysis The Takara ThruPLEX low input DNA library preparation kit was used to generate libraries based on manufacturer’s recommendations. Completed libraries were subject to quality control and quantification using a combination of Qubit dsDNA HS and Caliper LabChipGX HS DNA assays. All libraries were pooled in equimolar amounts to a maximum usable volume based on quantification obtained using the Kapa Biosystems Illumina Library Quantification qPCR kit. This pool was loaded on one lane of an Illumina HiSeq 4000 flow cell and sequenced in a 2x150 bp paired end format. Base calling was performed by Illumina Real Time Analysis (RTA) v2.7.6 and output of the RTA was demultiplexed and converted to FastQ format with Illumina Bcl2fastq v2.18.0. The Meta Genome Rapid Annotation using Subsystem Technology (MG-RAST) (35) version 4.0.2. was used for the taxonomic analysis of the metagenomes. The processing pipeline included merging paired end reads, SolexaQA (81) to trim low-quality regions from FASTQ data and dereplication to remove artificial duplicate reads. Gene calling was completed using FragGeneScan (82). For taxonomic profiles, the best hit classification at a maximum e-value of 1e−5, a minimum identity of 60% and a minimum alignment length of 15 against the ReqSeq database (83) were used. The MG-RAST plugin Krona was used to illustrate the taxonomic composition of each sample (84). MG-RAST was used to generate rarefaction curves. MG-RAST ID numbers and pre- and post- QC sequencing data have been summarized (Supplementary Table 43 2.2) and the datasets are publicly available on MG-RAST. The following chlorinated solvent degrading genera were investigated in the MG-RAST data: Anaeromyxobacter (85), Dehalococcoides (2, 4-6, 86, 87), Polaromonas (56, 72), Nocardioides (70, 88), Desulfitobacterium (47-50), Geobacter (52), Sulfurospirillum (89-91), Dehalobacter (92-94), Desulfomonile (95, 96), Desulfuromonas (97, 98), Propionibacterium (99), Mycobacterium (67, 100), Dehalobacter (93, 101), Desulfomonile, (102) and Dehalogenimonas (103-106). DIAMOND (double index alignment of next-generation sequencing data ) (37) was used as the alignment tool for all functional genes. The collected protein sequences from same species or with the same function were aligned to themselves for dereplication (removing sequences with 100% identity) and one representative sequence was left as the reference for that group. Then, low quality sequences and Illumina adapters sequences were removed using Trimmomatic (107). The shotgun reads were then aligned to the dereplicated references for each groundwater sample and SDC-9 using DIAMOND. Only reads that exhibited an identity of ≥ 90% and an alignment length ≥ 49 amino acids to the reference sequences were counted as aligned reads to each sequence. For each, relative abundance values were calculated using the number of aligned reads divided by the total number of sequences for each sample. The relative abundance values were then normalized by (divided by) the number of dereplicated reference sequences for each gene to produce normalized relative abundance values. Details concerning qPCR targeted towards vcrA are provided in the Supplementary Section. Text A4. vcrA qPCR The PCR tubes (20 µL reactions) contained 10 µL iTaq Universal Probe Supermix (Bio-Rad), 1.2 µL TaqMan probe (38, 108) and balance water to 18 µL. PCR amplifications were performed in three stages: 1) 95 °C for 15 min, 2) 40 cycle of 95 °C for 15 s, 58 °C for 1 min, 3) a slow 44 ramp of 1% to 95 °C for 15 s and 58 °C for 15 s. DNA templates and plasmid standards (containing a partial vcrA gene in a GenScript plasmid) were added to each reaction as 2 µL aliquots. All qPCR experiments were performed in a bench top thermal cycler (C1000 Touch Thermal Cycler, Bio-Rad). 45 Supplementary Table 2. 1. Groundwater and sampling data. Site Date Bioaugmented Date Sampled Quantico, VA 12/2/2015 5/17/2016 Months between Inoculation & Sampling 6.5 Tulsa, OK 8/2013 6/9/2015 22 Indian Head, MD 9/23/15 6/22/2016 9 San Antonio, TX Edison, NJ 113 & 514: 10/7 & 10/8/2014 35: 10/17 & 10/30/2014 7/8/2009 7/28/2016 20 11/10/2015 76 Basic Geochemistry pH 5-8 ORP: -100 to 0 Dis Fe: 0-140 mg/L pH 6.1-7.1 ORP: -64 to 248 Dis Fe: 0-110 mg/L pH 6.2-8.9 ORP: 37 to -326 pH 5.6-6.8 ORP: -32 to 237 Overall: pH 5-6.5 ORP: -100 to 3 Dis Fe: 0.7 – 13 mg/L MW-114: pH 6.0 ORP: -70 Dis Fe: 5.19 mg/L MW-303S: pH 6.1 ORP: -80 Dis Fe: 2.05 mg/L 46 Carbon Source cVOCs at time of sampling (µg/L) None, H2 generated via proton reduction EOS lactate EVO Lactate, yeast extract, potassium bicarbonate VC – 0-60 Cis-DCE - 0-120 VC – 60-830 Cis-DCE – 110-1500 TCE – 200-9000 Trans-DCE – 3-17 1,1-DCE – 5-1400 1,2-DCA – 1-34 1,1-DCA – Trace levels (<5) 1,4-Dioxane – 78-220 VC – 0-29 Cis-DCE – 0-178 TCE – 0-40 Carbon Disulfide – 1-8 MEK – 0-8 VC – 2.5-8.4 Cis-DCE – 2.5-7.4 TCE – 0-2 Overall: VC – 0-1170 Cis-DCE – 0-1190 TCE – 0-8 1,2,4-Trimethylbenzene – 0-8. Trace levels (<5) of benzene, MTBE, ethylbenzene, xylenes, isopropyl benzene, 1,3,5- trimethylbenzene, sec-butylbenzene, 1,1-DCE, tDCE MW-114: VC – 83 Cis-DCE – 70 TCE – 8 MW-303S: VC – 2J Cis-DCE – 4J Supplementary Table 2. 2. Groundwater and SDC-9 MG-RAST sequence analysis data. Location Monitoring Well MG-RAST ID # Pre QC Sequence Count Post QC Sequence Count Post QC Mean Length SDC-9 San Antonio, TX Tulsa, OK Quantico, VA Edison, NJ Indian Head, MD Culture-1 Culture-2 MW35 MW113 MW514 MW2 MW3 MW4 IW3 IW4 IW6 MWCW2 PMW2 MWAW1 MW 15R PMW4 MW114 MW303S IW5 IW7 MW38 MW40 mgm4795922.3 mgm4795924.3 mgm4795328.3 mgm4795332.3 mgm4795329.3 mgm4795334.3 mgm4795333.3 mgm4795342.3 mgm4795340.3 mgm4795341.3 mgm4795673.3 mgm4795675.3 mgm4795339.3 mgm4795335.3 mgm4795679.3 mgm4795678.3 mgm4795676.3 mgm4795677.3 mgm4795927.3 mgm4795847.3 mgm4795845.3 mgm4795846.3 6,845,624 6,181,247 5,301,996 6,185,927 5,934,109 5,691,547 6,872,780 6,200,534 5,889,710 6,938,129 7,800,767 5,171,923 662,422 233,177 6,710,609 5,429,417 6,174,464 5,824,346 5,674,151 2,036,212 5,832,647 5,265,951 5,090,799 4,478,198 4,513,530 5,404,716 4,847,401 4,714,786 5,425,995 5,327,773 4,891,993 5,228,938 6,112,693 3,998,002 471,513 157,539 5,018,326 4,433,348 5,242,089 5,008,287 4,837,429 1,547,247 5,071,187 4,480,623 236 ± 37 bp 229 ± 39 bp 239 ± 35 bp 239 ± 35 bp 240 ± 35 bp 239 ± 35 bp 227 ± 38 bp 238 ± 35 bp 241 ± 34 bp 229 ± 38 bp 230 ± 38 bp 241 ± 35 bp 231 ± 39 bp 226 ± 38 bp 241 ± 35 bp 241 ± 35 bp 240 ± 35 bp 239 ± 35 bp 241 ± 34 bp 219 ± 41 bp 240 ± 35 bp 241 ± 35 bp 47 Supplementary Table 2. 3. Genomes used for collecting functional protein sequences. Replicons Size (Mb) GC% Organism/Name Strain Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalococcoides mccartyi Dehalogenimonas lykanthroporepellens Dehalogenimonas formicexedens Dehalogenimonas alkenigignens Dehalogenimonas Geobacter lovleyi Sulfurospirillum Sulfurospirillum 195 CG5 CBDB1 BAV1 VS GT DCMB5 BTF08 GY50 CG4 CG1 IBARAKI 11a5 CG3 KBTCE2 KBDCA1 KBDCA2 KBTCE3 KBDCA3 KBVC2 KBVC1 KBTCE1 UCH-ATV1 MB 11a JNA SG1 WBC-2 EV-VC EV-TCE BL-DC-9 NSZ-14 IP3-3 WBC-2 1.46972 1.36215 1.3955 1.34189 1.41346 1.36015 1.4319 1.45233 1.40742 1.38231 1.48668 1.45106 1.46791 1.52129 1.3292 1.42846 1.39432 1.2716 1.33749 1.33773 1.3599 1.38891 1.38778 1.57151 1.32452 1.46251 1.42874 1.37458 1.4716 1.3573 1.68651 2.092789 1.85 1.72573 48.9 47.2 47 47.2 47.3 47.3 47.1 47.3 47 48.7 46.9 47 46.87 46.9 49.1 47.4 47.5 49.3 47.6 47.2 47.3 47.3 48.8 48.3 47.2 47.1 47.1 47.4 46.8 48.5 55.5 54 55.9 49.2 chromosome:NC_002936.3/CP000027.1 chromosome:NZ_CP006951.1/CP006951.1 chromosome:NC_007356.1/AJ965256.1 chromosome:NC_009455.1/CP000688.1 chromosome:NC_013552.1/CP001827.1 chromosome:NC_013890.1/CP001924.1 chromosome:NC_020386.1/CP004079.1 chromosome:NC_020387.1/CP004080.1 chromosome:NC_022964.1/CP006730.1 chromosome:NZ_CP006950.1/CP006950.1 chromosome:NZ_CP006949.1/CP006949.1 chromosome Unknown:NZ_AP014563.1/AP014563.1 chromosome:NZ_CP011127.1/CP011127.1 plasmid pDhc6:NZ_CP011128.1/CP011128.1 chromosome:NZ_CP013074.1/CP013074.1 chromosome:NZ_CP019865.1/CP019865.1 chromosome:NZ_CP019867.1/CP019867.1 chromosome:NZ_CP019868.1/CP019868.1 chromosome:NZ_CP019866.1/CP019866.1 chromosome:NZ_CP019946.1/CP019946.1 chromosome:NZ_CP019969.1/CP019969.1 chromosome:NZ_CP019968.1/CP019968.1 chromosome:NZ_CP019999.1/CP019999.1 chromosome:NZ_AP017649.1/AP017649.1 - - - - chromosome:CP017572.1 - - chromosome:NC_014314.1/CP002084.1 chromosome:NZ_CP018258.1/CP018258.1 - chromosome:CP011392.1 WGS - - - - - - - - - - - - - - - - - - - - - - - JGYD01 JGVX01 JSWM01 JPRE01 - LZFK01 LZFJ01 - - LFDV01 - Gene Protein # of RDase 1582 1459 1479 1444 1505 1468 1524 1556 1499 1470 1600 1556 1587 1657 1431 1563 1523 1361 1441 1440 1456 1502 1489 1711 1415 1582 1535 1466 1535 1451 1732 2176 1940 1800 3640 2943 2882 1497 1395 1412 1374 1432 1399 1461 1485 1427 1401 1527 1471 1521 1589 1364 1483 1443 1295 1372 1378 1393 1441 1408 1614 1339 1515 1467 1386 1475 1369 1650 2091 1856 1721 3552 2701 2793 19 25 32 10 37 20 23 20 26 13 32 28 30 20 4 6 6 4 7 16 21 16 15 27 8 26 28 15 28 11 20 25 29 22 2 2 2 Release Date 10/3/2001 8/4/2014 8/19/2005 5/7/2007 12/3/2009 2/17/2010 2/22/2013 2/22/2013 11/26/2013 8/4/2014 8/4/2014 9/9/2015 4/5/2016 12/6/2016 2/28/2017 2/28/2017 2/28/2017 2/28/2017 3/6/2017 3/9/2017 3/9/2017 3/15/2017 7/8/2017 12/4/2015 12/4/2015 1/15/2016 8/18/2014 10/12/2016 7/1/2016 7/1/2016 1/28/2014 10/17/2017 6/23/2016 5/8/2015 6/26/2016 5/31/2017 9/18/2017 SZ 0.077113 52.97 chromosome:NC_010815.1/CP001090 - plasmid pGLOV01: NC_010815.1/CP001090.1 SL2-1 JPD-1 2.87654 2.81409 38.7 38.8 chromosome:NZ_CP021416.1/CP021416.1 chromosome:NZ_CP023275.1/CP023275.1 - - 48 Supplementary Table 2. 3. (continued) SL2-2 Sulfurospirillum Sulfurospirillum halorespirans Sulfurospirillum multivorans Anaeromyxobacter dehalogenans Dehalobacter restrictus Dehalobacter Dehalobacter Dehalobacter Dehalobacter Dehalobacter Dehalobacter Desulfitobacterium hafniense Desulfitobacterium hafniense Desulfitobacterium hafniense Desulfitobacterium hafniense Desulfitobacterium hafniense Desulfitobacterium Desulfitobacterium chlororespirans Desulfitobacterium dichloroeliminans Desulfitobacterium dehalogenans Polaromonas JS666 DSM 13726 DSM 12446 2CP-C DSM 9455 DCA CF E1 FTH1 UNSWDHB TeCB1 DCB-2 Y51 TCP-A PCP-1 PCE-S PCE1 DSM 11544 LMG P-21439 ATCC 51507 2.87661 3.03 3.18 5.01348 2.94 3.06995 3.09205 2.95026 6.32936 3.20156 3.13322 5.27913 5.72753 4.96723 5.56321 5.6667 4.22 5.61 3.62 4.32 38.7 41.3 40.9 74.9 44.6 44.6 44.3 43.8 58.9 44.9 44 47.5 47.4 47.3 47.5 47.3 45 47.3 44.2 45 JS666 5.89868 62.00 chromosome:NZ_CP021979.1/CP021979.1 chromosome:NZ_CP017111.1/CP017111.1 chromosome:NZ_CP007201.1/CP007201.1 chromosome:NC_007760.1/CP000251.1 chromosome:NZ_CP007033.1/CP007033.1 chromosome:NC_018866.1/CP003869.1 chromosome:NC_018867.1/CP003870.1 - - - - chromosome:NC_011830.1/CP001336.1 chromosome:NC_007907.1/AP008230.1 - - - - - chromosome:NC_019903.1/CP003344.1 chromosome:NC_018017.1/CP003348.1 chromosome:NC_007948.1/CP000316.1 plasmid 1:NC_007949.1/CP000317.1 plasmid 2:NC_007950.1/CP000318.1 - - - - - - - CANE01 AQYY01 AUUR01 MCHF01 - - AQZD01 ARAZ01 - AQZF01 FRDN01 - - - 2924 3035 3288 4522 2848 2974 2985 2866 5934 3105 3106 5038 5484 4839 5327 5490 4070 5367 3463 4212 5660 2699 2967 3186 4416 2647 2848 2882 2719 5727 2944 2961 4821 5227 4556 5095 5417 3873 5282 3300 3974 5485 2 2 2 2 23 18 18 7 33 19 24 7 2 5 7 6 6 2 1 7 1a 6/22/2017 9/8/2017 4/29/2015 1/27/2006 5/14/2014 10/16/2012 10/16/2012 9/19/2012 4/19/2013 8/9/2013 8/18/2016 1/5/2009 3/10/2006 4/22/2013 4/22/2013 - 9/16/2013 12/2/3016 6/17/2013 9/10/2015 2006/04/10 a: Cytochrome P450 (ABE47160.1) from Polaromonas JS666 is not a RDase but catalyzes the initial step of cDCE degradation. Supplementary Table 2. 4. Number of collected genomes and dereplicated RDases. Microorganism Dehalococcoides mccartyi Dehalogenimonas Anaeromyxobacter Dehalobacter Geobacter Sulfurospirillum Desulfitobacterium Polaromonas Number of collected genome Dereplicated RDase number 30 4 1 7 1 5 9 1 317 91 2 103 2 6 36 1 (not an RDase) 49 Supplementary Figure 2. 1. TCE plume maps for the Edison, NJ site. TCE contour maps for the site prior to addition of emulsified oil and dehalogenating culture SDC-9 in 2009 are provided for the shallow zone (A) and deep zone at the site (B). Well 303S is located in the shallow zone and well 114 is located in the deep zone. Post-treatment contour maps in 2010 for the shallow zone (C) and deep zone (D) are also provided. All values are in µg/L. The wells from which samples were collected and analyzed are indicated with arrows. A: TCE UPPER ZONE 50 Supplementary Figure 2. 1. (continued) B: TCE LOWER ZONE 51 Supplementary Figure 2. 1. (continued) C: TCE UPPER ZONE 52 Supplementary Figure 2. 1. (continued) D: TCE LOWER ZONE 53 Supplementary Figure 2. 2. Cis-DCE Plume maps for the Edison, NJ site. Cis-DCE contour maps for the site prior to addition of emulsified oil and dehalogenating culture SDC-9 in 2009 are provided for the shallow zone (A) and deep zone at the site (B). Well 303S is located in the shallow zone and well 114 is located in the deep zone. Post-treatment contour maps in 2010 for the shallow zone (C) and deep zone (D) are also provided. All values are in µg/L. The wells from which samples were collected and analyzed are indicated with arrows. A: DCE UPPER ZONE 54 Supplementary Figure 2. 2. (continued) B: DCE LOWER ZONE 55 Supplementary Figure 2. 2. (continued) C: DCE UPPER ZONE 56 Supplementary Figure 2. 2. (continued) D: DCE LOWER ZONE 57 Supplementary Figure 2. 3. Demonstration plot layout at the Quantico, VA site. The cathode and anode wells are indicated by red and green symbols, respectively. This system was used to supply H2 to support reductive dechlorination of cis-DCE downgradient of a landfill. See data in Supplementary Figures 17-19. 58 Supplementary Figure 2. 4. Concentration data for cis-DCE at the Quantico, VA site. The groundwater samples were collected on Day 243 from wells CW- 2, PMW-2, CW-2, AW-1, MW-15R, and PMW-4. 59 Supplementary Figure 2. 5. Concentration data for vinyl chloride at the Quantico, VA site. The groundwater samples were collected on Day 243 from wells CW-2, PMW-2, CW-2, AW-1, MW-15R, and PMW-4. 60 Supplementary Figure 2. 6. Concentration data for ethene at the Quantico, VA site. The groundwater samples were collected on Day 243 from wells CW-2, PMW-2, CW-2, AW-1, MW-15R, and PMW-4. 61 Supplementary Figure 2. 7. Demonstration plot layout at the Indian Head, Md site. Injection wells (IWs) were amended with lactate, diammonium phosphate, potassium bicarbonate (for pH adjustment) and dehalogenating culture SDC-9. Monitoring wells (MWs) were used to measure system performance. A low voltage was used to maintain system pH. Anodes for this system are shown in the figure. Wells that were sampled are indicated by arrows. See MW data in Supplementary Figures 21-22. No analytical data are available for the IWs. 62 Supplementary Figure 2. 8. Concentration data for cVOCs, ethene and ethane in well MW38 at the Indian Head, Md site. The groundwater samples were collected on 6/22/16. 63 Supplementary Figure 2. 9. Concentration data for cVOCs, ethene and ethane in well MW40 at the Indian Head, Md site. The groundwater samples were collected on 6/22/16. 64 Supplementary Figure 2. 10. Demonstration Plot layout at the Tulsa, Ok site. IWs are emulsified oil and dehalogenating culture SDC-9 injection wells and MWs are groundwater monitoring wells. See data in Supplementary Figures 24-26. Groundwater flow 65 Supplementary Figure 2. 11. Concentration data for TCE in injection wells (IWs) at the Tulsa, OK Site. The groundwater samples were collected on 6/09/15. 66 Supplementary Figure 2. 12. Concentration data for TCE in monitoring wells (MWs) at the Tulsa, OK Site. The groundwater samples were collected on 6/09/15. 67 Supplementary Figure 2. 13. Concentration data for 1,4-dioxane in injection wells (IWs) at the Tulsa, OK Site. The groundwater samples were collected on 6/09/15. 68 Supplementary Figure 2. 14. Injection points and locations of monitoring wells SS050MW113 (113) and SS050MW514 (514) at the San Antonio, TX, Site. Analytical data are provided for each well. Groundwater samples were collected on 7/28/16. BZ = benzene. 69 Supplementary Figure 2. 15. Injection points and location of monitoring well SS050MW035 (35) at the San Antonio, TX, Site. Analytical data are provided. Groundwater samples were collected on 7/28/16. 70 Supplementary Figure 2. 16. Rarefaction curves for microbial communities in groundwater and in SDC-9. 71 A. SDC-9, replicate 1 B. SDC-9, replicate 2 Supplementary Figure 2. 17. Classification of microbial communities in two samples of SDC-9 (data analyzed with MG-RAST). 72 A. San Antonio, Monitoring Well 35 B. San Antonio, Monitoring Well 113 C. San Antonio, Monitoring Well 514 Supplementary Figure 2. 18. Classification of microbial communities in three monitoring well groundwater samples from San Antonio (data analyzed with MG-RAST). 73 A. Tulsa, Injection Well 4 B. Tulsa, Injection Well 6 Supplementary Figure 2. 19. Classification of microbial communities in injection well (A and B) and monitoring well (C, D and E) groundwater samples from Tulsa (data analyzed with MG-RAST). 74 Supplementary Figure 2. 19. (continued) C. Tulsa, Monitoring Well 2 D. Tulsa, Monitoring Well 3 E. Tulsa, Monitoring Well 4 75 A. Quantico, Cathode Injection Well CW2 B. Quantico Monitoring Well PMW2 (first from barrier) C. Quantico Monitoring Well MW15 (second from barrier) D. Quantico Monitoring Well PMW4 (third from barrier) Supplementary Figure 2. 20. Classification of microbial communities in groundwater injection well (A) and monitoring well (B, C, D) samples from Quantico (data analyzed with MG-RAST). 76 A. Edison, Monitoring Well 114 B. Edison, Monitoring Well 303 Supplementary Figure 2. 21. Classification of microbial communities in groundwater monitoring well samples from Edison (data analyzed with MG- RAST). 77 A. Indian Head, Injection Well 5 B. Indian Head, Injection Well 7 C. Indian Head, Monitoring Well 38 D. Indian Head, Monitoring Well 40 Supplementary Figure 2. 22. Classification of microbial communities in groundwater injection (A, B) and monitoring well (C, D) samples from Indian Head (data analyzed with MG-RAST). 78 0.08 0.06 0.04 0.02 0.00 1 - 9 C D S 2 - 9 C D S fdhA ) % ( e c n a d n u b A e v i t a l e R d e z i l a m r o N 6E-3 5E-3 4E-3 3E-3 2E-3 1E-3 0E+0 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W C W I 2 W M P 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M Supplementary Figure 2. 23. Normalized relative abundance (%) of fdhA in SDC-9 (insert) and in groundwater from the five chlorinated solvent sites (data analyzed with DIAMOND). 79 A hupS hupL 5E-3 4E-3 3E-3 ) 2E-3 0.06 0.04 0.02 0.00 1 - 9 C D S 2 - 9 C D S 1E-3 0E+0 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 2E-2 C hymD1 hymB3 hymA3 hymC3 hymB2 hymA2 hymC2 hymB1 hymA1 hymC1 hymA4 % ( e c n a d n u b A e v i t a l e R d e z i l a m r o N 2E-2 1E-2 5E-3 0E+0 0.04 0.03 0.02 0.01 0.00 1 - 9 C D S 2 - 9 C D S 4E-3 B vhcG vhcA 3E-3 2E-3 1E-3 0E+0 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 2E-2 echF echE echC echB echA D 1E-2 8E-3 4E-3 0E+0 0.15 0.10 0.05 0.00 1 - 9 C D S 2 - 9 C D S 0.20 0.15 0.10 0.05 0.00 1 - 9 C D S 2 - 9 C D S 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I 2 W M P 2 W C W I 1 W A W M R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M Supplementary Figure 2. 24. Normalized relative abundance (%) of Dehalococcoidies mccartyi hydrogenase genes hupLS (A), vhcAG (B), hymABCD (C) and echABCEF (D) in SDC-9 (inserts) and in groundwater from the five chlorinated solvent sites (data analyzed with DIAMOND). 80 B cbiZ4 cbiZ3 cbiZ2 cbiZ1 cbiB 0.15 0.10 0.05 0.00 1 - 9 C D S 2 - 9 C D S 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I … A W M 2 W M P 2 W C W I R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M 1E-2 8E-3 4E-3 0E+0 5E-3 4E-3 3E-3 2E-3 1E-3 0E+0 2E-2 1E-2 8E-3 4E-3 0E+0 ) % ( e c n a d n u b A e v i t a l e R d e z i l a m r o N A btuF btuD btuC 0.04 0.03 0.02 0.01 0.00 1 - 9 C D S 2 - 9 C D S 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I … A W M 2 W M P 2 W C W I R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M cobU1 cobS1 cobD4 cobC cobA3 cobA1 cobT1 cobQ cobD1 cobB cobA2 C 0.20 0.10 0.00 1 - 9 C D S 2 - 9 C D S 5 3 W M 3 1 1 W M 4 1 5 W M 2 W M 3 W M 4 W M 3 W I 4 W I 6 W I … A W M 2 W M P 2 W C W I R 5 1 W M 4 W M P 4 1 1 W M S 3 0 3 W M 5 W I 7 W I 8 3 W M 0 4 W M Supplementary Figure 2. 25. Normalized relative abundance (%) of Dehalococcoidies mccartyi corrinoid metabolism genes btuFCD (A), cbiA, cbiB, cbiZ (B) and cobA, cobB, cobC, cobD, cobQ, cobS, cobT, cobU (C) in SDC-9 (inserts) and in groundwater from the five chlorinated solvent sites (data analyzed with DIAMOND). 81 i y t r a c c m . D m o r f s e s a D R f o e c n a d n u b a e v i t a l e r d e z i l a m r o N 0.007 0.006 0.005 0.004 0.003 0.002 0.001 0 0 tceA vcrA total 0.001 0.002 0.003 0.004 0.005 0.006 Normalized relative abundance of fdhA Supplementary Figure 2. 26. Comparison between normalized relative abundance of vcrA, tceA and sum of RDases to fdhA (data analyzed with DIAMOND). 82 0.003 ) % ( e c n a d n u b A e v i t a l e R d e z i l a m r o N 0.002 0.001 0.000 0 vcrA from genomes only All vcrA sequences 5E+09 1E+10 Gene copies per L (qPCR) Supplementary Figure 2. 27. Comparison between vcrA gene copies (per L) determined via qPCR and shotgun sequencing (normalized relative abundance, %, MG-RAST). The results from two shotgun sequencing quantification methods are shown (as discussed in the text). 83 REFERENCES 84 REFERENCES Steffan RJ, Vainberg S. 2013. Production and handling of Dehalococcoides bioaugmentation Muller JA, Rosner BM, von Abendroth G, Meshulam-Simon G, McCarty PL, Spormann AM. Maymo-Gatell X, Anguish T, Zinder SH. 1999. Reductive dechlorination of chlorinated ethenes He JZ, Ritalahti KM, Aiello MR, Loffler FE. 2003. Complete detoxification of vinyl chloride by 1. cultures, p 89-115. In Stroo HF, Leeson A, Ward CH (ed), Bioaugmentation for Groundwater Remediation. Springer, New York. 2. 2004. Molecular identification of the catabolic vinyl chloride reductase from Dehalococcoides sp strain VS and its environmental distribution. Applied and Environmental Microbiology 70:4880-4888. 3. an anaerobic enrichment culture and identification of the reductively dechlorinating population as a Dehalococcoides species. Applied and Environmental Microbiology 69:996-1003. 4. and 1,2-dichloroethane by "Dehalococcoides ethenogenes" 195. Applied and Environmental Microbiology 65:3108-3113. 5. characterization of Dehalococcoides sp strain FL2, a trichloroethene (TCE)- and 1,2-dichloroethene- respiring anaerobe. Environmental Microbiology 7:1442-1450. 6. Genetic identification of a putative vinyl chloride reductase in Dehalococcoides sp strain BAV1. Applied and Environmental Microbiology 70:6347-6351. 7. In Stroo HF, Leeson A, Ward CH (ed), Bioaugmentation for groundwater remediation. Springer, New York. 8. application of a rapid, user-friendly, and inexpensive method to detect Dehalococcoides sp reductive dehalogenase genes from groundwater. Applied Microbiology and Biotechnology 101:4827-4835. 9. ethenes in fractured bedrock and associated changes in dechlorinating and nondechlorinating microbial populations. Environmental Science & Technology 48:5770-5779. 10. with Dehalococcoides, p 171-197. In Stroo HF, Leeson A, Ward CH (ed), Bioaugmentation for groundwater remediation. Springer, New York. 11. enumeration of the Dehalococcoides 16S rRNA gene in groundwater. Journal of Microbiological Methods 88:263-270. Kanitkar YH, Stedtfeld RD, Hatzinger PB, Hashsham SA, Cupples AM. 2017. Development and Perez-de-Mora A, Zila A, McMaster ML, Edwards EA. 2014. Bioremediation of chlorinated Petrovskis EA, Amber WR, Walker CB. 2013. Microbial monitoring during bioaugmentation Hatt JK, Loffler FE. 2012. Quantitative real-time PCR (qPCR) detection chemistries affect He J, Sung Y, Krajmalnik-Brown R, Ritalahti KM, Loffler FE. 2005. Isolation and Krajmalnik-Brown R, Holscher T, Thomson IN, Saunders FM, Ritalahti KM, Loffler FE. 2004. Lyon DY, Vogel TM. 2013. Bioaugmentation for groundwater remediation: an overview, p 1-38. 85 Lee PKH, Macbeth TW, Sorenson KS, Deeb RA, Alvarez-Cohen L. 2008. Quantifying genes and Liang Y, Liu XK, Singletary MA, Wang K, Mattes TE. 2017. Relationships between the van der Zaan B, Hannes F, Hoekstra N, Rijnaarts H, de Vos WM, Smidt H, Gerritse J. 2010. Kotik M, Davidova A, Voriskova J, Baldrian P. 2013. Bacterial communities in Nemecek J, Dolinova I, Machackova J, Spanek R, Sevcu A, Lederer T, Cernik M. 2017. 12. transcripts to assess the in situ physiology of "Dehalococcoides" spp. in a trichloroethene-contaminated groundwater site. Applied and Environmental Microbiology 74:2728-2739. 13. abundance and expression of functional genes from vinyl chloride (VC)-degrading bacteria and geochemical parameters at VC-contaminated sites. Environmental Science & Technology 51:12164- 12174. 14. Correlation of Dehalococcoides 16S rRNA and chloroethene-reductive dehalogenase genes with geochemical conditions in chloroethene-contaminated groundwater. Applied and Environmental Microbiology 76:843-850. 15. Munro JE, Kimyon O, Rich DJ, Koenig J, Tang SH, Low A, Lee M, Manefield M, Coleman NV. 2017. Co-occurrence of genes for aerobic and anaerobic biodegradation of dichloroethane in organochlorine-contaminated groundwater. Fems Microbiology Ecology 93. 16. tetrachloroethene-polluted groundwaters: A case study. Science of the Total Environment 454:517-527. 17. Stratification of chlorinated ethenes natural attenuation in an alluvial aquifer assessed by hydrochemical and biomolecular tools. Chemosphere 184:1157-1167. 18. receiving chlorinated solvents from underlying fracture networks. Environmental Science & Technology 51:4821-4830. 19. M, Stams AJM, Springael D, Dejonghe W, Smidt H. 2017. Geochemical and microbial community determinants of reductive dechlorination at a site biostimulated with glycerol. Environmental Microbiology 19:968-981. 20. microbial community from chlorinated solvent-contaminated groundwater after biostimulation using the metagenome analysis. Journal of Hazardous Materials 302:144-150. 21. Curvers C, Boucher D, Vogel TM, Peyretaillade E, Peyret P. 2012. In situ TCE degradation mediated by complex dehalorespiring communities during biostimulation processes. Microbial Biotechnology 5:642- 653. 22. Filip J, Cajthaml T. 2016. Combined nano-biotechnology for in-situ remediation of mixed contamination of groundwater by hexavalent chromium and chlorinated solvents. Science of the Total Environment 563:822-834. 23. EA, O'Carroll DM. 2016. Long-term field study of microbial community and dechlorinating activity Atashgahi S, Lu Y, Zheng Y, Saccenti E, Suarez-Diez M, Ramiro-Garcia J, Eisenmann H, Elsner Simsir B, Yan J, Im J, Graves D, Loffler FE. 2017. Natural attenuation in streambed sediment Kao CM, Liao HY, Chien CC, Tseng YK, Tang P, Lin CE, Chen SC. 2016. The change of Dugat-Bony E, Biderre-Petit C, Jaziri F, David MM, Denonfoux J, Lyon DY, Richard JY, Nemecek J, Pokorny P, Lhotsky O, Knytl V, Najmanova P, Steinova J, Cernik M, Filipova A, Kocur CMD, Lomheim L, Molenda O, Weber KP, Austrins LM, Sleep BE, Boparai HK, Edwards 86 Badin A, Broholm MM, Jacobsen CS, Palau J, Dennis P, Hunkeler D. 2016. Identification of Reiss RA, Guerra P, Makhnin O. 2016. Metagenome phylogenetic profiling of microbial Adetutu EM, Gundry TD, Patil SS, Golneshin A, Adigun J, Bhaskarla V, Aleer S, Shahsavari E, Nemecek J, Steinova J, Spanek R, Pluhar T, Pokorny P, Najmanova P, Knytl V, Cernik M. 2018. following carboxymethyl cellulose-stabilized nanoscale zero-valent iron injection. Environmental Science & Technology 50:7658-7670. 24. Thermally enhanced in situ bioremediation of groundwater contaminated with chlorinated solvents - A field test. Science of the Total Environment 622:743-755. 25. abiotic and biotic reductive dechlorination in a chlorinated ethene plume after thermal source remediation by means of isotopic and molecular biology tools. Journal of Contaminant Hydrology 192:1-19. 26. community evolution in a tetrachloroethen-contaminated aquifer responding to enhanced reductive dechlorination protocols. Standards in Genomic Sciences 11. 27. Ross E, Ball AS. 2015. Exploiting the intrinsic microbial degradative potential for field-based in situ dechlorination of trichloroethene contaminated groundwater. Journal of Hazardous Materials 300:48-57. 28. Weigold P, El-Hadidi M, Ruecker A, Huson DH, Scholten T, Jochmann M, Kappler A, Behrens S. 2016. A metagenomic-based survey of microbial (de)halogenation potential in a German forest soil. Scientific Reports 6. 29. Hug LA, Beiko RG, Rowe AR, Richardson RE, Edwards EA. 2012. Comparative metagenomics of three Dehalococcoides-containing enrichment cultures: the role of the non-dechlorinating community. Bmc Genomics 13. 30. analysis of a stable trichloroethene-degrading microbial community. Isme Journal 6:1702-1714. 31. procedures, EPA/540/S-95/504. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington, DC. 32. for treating chlorinated ethenes. Ground Water Monitoring and Remediation 30:113-124. 33. ethenes using Dehalococcoides sp.: Comparison between batch and column experiments. Chemosphere 75:141-148. 34. isothermal amplification (LAMP) for rapid detection and quantification of Dehalococcoides biomarker genes in commercial reductive dechlorinating cultures KB-1 and SDC-9. Applied and Environmental Microbiology 82:1799-1806. 35. Meyer F, Paarmann D, D'Souza M, Olson R, Glass EM, Kubal M, Paczian T, Rodriguez A, Stevens R, Wilke A, Wilkening J, Edwards RA. 2008. The metagenomics RAST server - a public resource for the automatic phylogenetic and functional analysis of metagenomes. Bmc Bioinformatics 9. Schaefer CE, Lippincott DR, Steffan RJ. 2010. Field-scale evaluation of bioaugmentation dosage Brisson VL, West KA, Lee PKH, Tringe SG, Brodie EL, Alvarez-Cohen L. 2012. Metagenomic Puls RW, Barcelona MJ. 1996. Low-flow (minimal drawdown) ground-water sampling Schaefer CE, Condee CW, Vainberg S, Steffan RJ. 2009. Bioaugmentation for chlorinated Kanitkar YH, Stedtfeld RD, Steffan RJ, Hashsham SA, Cupples AM. 2016. Loop-mediated 87 DiSpirito A, J. , Gulledge JC, Murrell AK, Shiemke ME, Lidstrom, Krema CL. 1992. Fish JA, Chai BL, Wang Q, Sun YN, Brown CT, Tiedje JM, Cole JR. 2013. FunGene: the Buchfink B, Xie C, Huson DH. 2015. Fast and sensitive protein alignment using DIAMOND. Ritalahti KM, Amos BK, Sung Y, Wu QZ, Koenigsberg SS, Loffler FE. 2006. Quantitative PCR 36. functional gene pipeline and repository. Frontiers in Microbiology 4. 37. Nature Methods 12:59-60. 38. targeting 16S rRNA and reductive dehalogenase genes simultaneously monitors multiple Dehalococcoides strains. Applied and Environmental Microbiology 72:2765-2774. 39. Trichloroethylene oxidation by the membrane-associated methane monooxygenase in type I, type II, and type x methanotrophs. Biodegradation 2:151-164. 40. Wymore RA, Lee MH, Keener WK, Miller AR, Colwell FS, Watwood ME, Sorenson KS. 2007. Field evidence for intrinsic aerobic chlorinated ethene cometabolism by methanotrophs expressing soluble methane monooxygenase. Biodegradation 11:125-139. 41. Lee SW, Keeney DR, Lim DH, Dispirito AA, Semrau JD. 2006. Mixed pollutant degradation by Methylosinus trichosporium OB3b expressing either soluble or particulate methane monooxygenase: Can the tortoise beat the hare? Applied and Environmental Microbiology 72:7503-7509. 42. He Y, Mathieu J, Yang Y, Yu PF, da Silva MLB, Alvarez PJJ. 2017. 1,4-Dioxane biodegradation by Mycobacterium dioxanotrophicus PH-06 is associated with a group-6 soluble di-iron monooxygenase. Environmental Science & Technology Letters 4:494-499. 43. Jin YO, Mattes TE. 2010. A quantitative PCR assay for aerobic, vinyl chloride- and ethene- assimilating microorganisms in groundwater. Environmental Science & Technology 44:9036-9041. 44. Morris RM, Fung JM, Rahm BG, Zhang S, Freedman DL, Zinder SH, Richardson RE. 2007. Comparative proteomics of Dehalococcoides spp. reveals strain-specific peptides associated with activity. Appl Environ Microbiol 73:320-6. 45. overview of genomic, transcriptomic and proteomic analyses in organohalide respiration research. Fems Microbiology Ecology 94. 46. putative reductive dehalogenases in Desulfitobacterium hafniense/frappieri and Dehalococcoides ethenogenes. Canadian Journal of Microbiology 48:697-706. 47. Desulfitobacterium and identification of its functional reductase gene. Plos One 10. 48. dehalogenans gen-nov, sp-nov, an anaerobic bacterium which reductively dechlorinates chlorophenolic compounds. International Journal of Systematic Bacteriology 44:612-619. 49. characterization of a tetrachloroethene dechlorinating Desulfitobacterium sp strain Y51: a review. Journal of Industrial Microbiology & Biotechnology 32:534-541. Turkowsky D, Jehmlich N, Diekert G, Adrian L, von Bergen M, Goris T. 2018. An integrative Zhao SY, Ding C, He JZ. 2015. Detoxification of 1,1,2-trichloroethane to ethene by Utkin I, Woese C, Wiegel J. 1994. Isolation and characterization of Desulfitobacterium Villemur R, Saucier M, Gauthier A, Beaudet R. 2002. Occurrence of several genes encoding Furukawa K, Suyama A, Tsuboi Y, Futagami T, Goto M. 2005. Biochemical and molecular 88 Doong RA, Lee CC, Lien CM. 2014. Enhanced dechlorination of carbon tetrachloride by Sung Y, Fletcher KF, Ritalaliti KM, Apkarian RP, Ramos-Hernandez N, Sanford RA, Mesbah Gerritse J, Drzyzga O, Kloetstra G, Keijmel M, Wiersum LP, Hutson R, Collins MD, Gottschal 50. JC. 1999. Influence of different electron donors and accepters on dehalorespiration of tetrachloroethene by Desulfitobacterium frappieri TCE1. Applied and Environmental Microbiology 65:5212-5221. Yan J, Ritalahti KM, Wagner DD, Loffler FE. 2012. Unexpected specificity of interspecies 51. cobamide transfer from Geobacter spp. to organohalide-respiring Dehalococcoides mccartyi strains. Applied and Environmental Microbiology 78:6630-6636. 52. NM, Loffler FE. 2006. Geobacter lovleyi sp nov strain SZ, a novel metal-reducing and tetrachloroethene- dechlorinating bacterium. Applied and Environmental Microbiology 72:2775-2782. 53. Geobacter sulfurreducens in the presence of naturally occurring quinones and ferrihydrite. Chemosphere 97:54-63. 54. Men YJ, Yu K, Baelum J, Gao Y, Tremblay J, Prestat E, Stenuit B, Tringe SG, Jansson J, Zhang T, Alvarez-Cohen L. 2017. Metagenomic and metatranscriptomic analyses reveal the structure and dynamics of a dechlorinating community containing Dehalococcoides mccartyi and corrinoid-providing microorganisms under cobalamin-limited conditions. Applied and Environmental Microbiology 83. 55. Men YJ, Feil H, VerBerkmoes NC, Shah MB, Johnson DR, Lee PKH, West KA, Zinder SH, Andersen GL, Alvarez-Cohen L. 2012. Sustainable syntrophic growth of Dehalococcoides ethenogenes strain 195 with Desulfovibrio vulgaris Hildenborough and Methanobacterium congolense: global transcriptomic and proteomic analyses. Isme Journal 6:410-421. 56. Mattes TE, Alexander AK, Richardson PM, Munk AC, Han CS, Stothard P, Coleman NV. 2008. The genome of Polaromonas sp strain JS666: Insights into the evolution of a hydrocarbon- and xenobiotic-degrading bacterium, and features of relevance to biotechnology. Applied and Environmental Microbiology 74:6405-6416. 57. Wagner DD, Hug LA, Hatt JK, Spitzmiller MR, Padilla-Crespo E, Ritalahti KM, Edwards EA, Konstantinidis KT, Loffler FE. 2012. Genomic determinants of organohalide-respiration in Geobacter lovleyi, an unusual member of the Geobacteraceae. Bmc Genomics 13. 58. Liang Y, Cook LJ, Mattes TE. 2017. Temporal abundance and activity trends of vinyl chloride (VC)-degrading bacteria in a dilute VC plume at Naval Air Station Oceana. Environmental Science and Pollution Research 24:13760-13774. 59. Mahendra S, Alvarez-Cohen L. 2006. Kinetics of 1,4-dioxane biodegradation by monooxygenase-expressing bacteria. Environmental Science & Technology 40:5435-5442. 60. Hatzinger PB, Banerjee R, Rezes R, Streger SH, McClay K, Schaefer CE. 2017. Potential for cometabolic biodegradation of 1,4-dioxane in aquifers with methane or ethane as primary substrates. Biodegradation 28:453-468. 61. abundance of tetrahydrofuran/dioxane monooxygenase genes (thmA/dxmA) and 1,4-dioxane degradation activity are significantly correlated at various impacted aquifers. Environmental Science & Technology Letters 1:122-127. Li MY, Mathieu J, Liu YY, Van Orden ET, Yang Y, Fiorenza S, Alvarez PJJ. 2014. The 89 da Silva MLB, Woroszylo C, Castillo NF, Adamson DT, Alvarez PJJ. 2018. Associating potential Rahm BG, Richardson RE. 2008. Dehalococcoides' gene transcripts as quantitative bioindicators Coleman NV, Spain JC. 2003. Epoxyalkane: Coenzyme M transferase in the ethene and vinyl Rahm BG, Richardson RE. 2008. Correlation of respiratory gene expression levels and pseudo- Heavner GLW, Mansfeldt CB, Debs GE, Hellerstedt ST, Rowe AR, Richardson RE. 2018. Hartmans S, Debont JAM. 1992. Aerobic vinyl chloride metabolism in Mycobacterium aurum Danko AS, Saski CA, TomkinS JP, Freedman DL. 2006. Involvement of coenzyme M during Gedalanga P, Madison A, Miao Y, Richards T, Hatton J, DiGuiseppi WH, Wilson J, Mahendra S. 62. 2016. A multiple lines of evidence framework to evaluate intrinsic biodegradation of 1,4-dioxane. Remediation-the Journal of Environmental Cleanup Costs Technologies & Techniques 27:93-114. 63. 1,4-dioxane biodegradation activity with groundwater geochemical parameters at four different contaminated sites. Journal of Environmental Management 206:60-64. 64. of tetrachloroethene, trichloroethene, and cis-1,2-dichloroethene dehalorespiration rates. Environmental Science & Technology 42:5099-5105. 65. steady-state PCE respiration rates in Dehalococcoides ethenogenes. Environmental Science & Technology 42:416-421. 66. Biomarkers' responses to reductive dechlorination rates and oxygen stress in bioaugmentation culture KB- 1 (TM). Microorganisms 6. 67. L1. Applied and Environmental Microbiology 58:1220-1226. 68. aerobic biodegradation of vinyl chloride and ethene by Pseudomonas putida strain AJ and Ochrobactrum sp strain TD. Applied and Environmental Microbiology 72:3756-3758. 69. chloride biodegradation pathways of Mycobacterium strain JS60. Journal of Bacteriology 185:5536-5545. 70. Mattes TE, Coleman NV, Spain JC, Gossett JM. 2005. Physiological and molecular genetic analyses of vinyl chloride and ethene biodegradation in Nocardioides sp strain JS614. Archives of Microbiology 183:95-106. 71. among ethene- and vinyl chloride-degrading Mycobacterium strains. Applied and Environmental Microbiology 69:6041-6046. 72. dichloroethene by Polaromonas sp strain JS666. Applied and Environmental Microbiology 79:2263- 2272. 73. Methylocystis strain SB2 when grown on multi-carbon substrates: implications for biodegradation of chlorinated ethenes. Environmental Microbiology Reports 3:182-188. 74. aliphatic hydrocarbons by methane-oxidizing cultures. Applied and Environmental Microbiology 62:3371-3377. 75. Toxicology and Industrial Health 12:1-43. Yoon S, Im J, Bandow N, DiSpirito AA, Semrau JD. 2011. Constitutive expression of pMMO by Chang HL, Alvarez-Cohen L. 1996. Biodegradation of individual and multiple chlorinated DeRosa CT, Wilbur S, Holler J, Richter P, Stevens YW. 1996. Health evaluation of 1,4-dioxane. Coleman NV, Spain JC. 2003. Distribution of the coenzyme m pathway of epoxide metabolism Nishino SF, Shin KA, Gossett JM, Spain JC. 2013. Cytochrome P450 initiates degradation of cis- 90 Adamson DT, Mahendra S, Walker KL, Rauch SR, Sengupta S, Newell CJ. 2014. A multisite Yang Y, Higgins SA, Yan J, Simsir B, Chourey K, Iyer R, Hettich RL, Baldwin B, Ogles DM, Cox MP, Peterson DA, Biggs PJ. 2010. SolexaQA: At-a-glance quality assessment of Illumina Rho MN, Tang HX, Ye YZ. 2010. FragGeneScan: predicting genes in short and error-prone Pruitt KD, Tatusova T, Maglott DR. 2005. NCBI Reference Sequence (RefSeq): a curated non- Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. Adamson DT, Anderson RH, Mahendra S, Newell CJ. 2015. Evidence of 1,4-dioxane attenuation 76. Mohr T, Stickney J, DiGuiseppi W. 2010. Environmental Investigation and Remediation: 1,4- Dioxane and Other Solvent Stabilizers. CRC Press, Boca Raton, ML. 77. at groundwater sites contaminated with chlorinated solvents and 1,4-dioxane. Environmental Science & Technology 49:6510-6518. 78. survey to identify the scale of the 1,4-dioxane problem atcontaminated groundwater sites. Environmental Science & Technology Letters 1:254-258. 79. Journal of Molecular Biology 215:403-410. 80. Loffler FE. 2017. Grape pomace compost harbors organohalide-respiring Dehalogenimonas species with novel reductive dehalogenase genes. Isme Journal 11:2767-2780. 81. second-generation sequencing data. Bmc Bioinformatics 11:485:1-6. 82. reads. Nucleic Acids Research 38. 83. redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Research 33:D501- D504. 84. browser. Bmc Bioinformatics 12. 85. dehalogenans gen. nov., sp nov., an aryl-halorespiring facultative anaerobic myxobacterium. Applied and Environmental Microbiology 68:893-900. 86. chlorinated compound respiring bacterium Dehalococcoides species strain CBDB1. Nature Biotechnology 23:1269-1273. 87. H, Zinder SH, Spormann AM. 2015. Dehalococcoides mccartyi gen. nov., sp. nov., obligately organohalide-respiring anaerobic bacteria relevant to halogen cycling and bioremediation, belong to a novel bacterial class, Dehalococcoidia classis nov., order Dehalococcoidales ord. nov. and family Dehalococcoidaceae fam. nov., within the phylum Chloroflexi (vol 63, pg 625, 2013). International Journal of Systematic and Evolutionary Microbiology 65:2015-2015. 88. Rio TG, Goodwin LA, Hammon NM, Han SS, Hauser LJ, Israni S, Kim E, Kyrpides N, Land ML, Lapidus A, Larimer FW, Lucas S, Pitluck S, Richardson P, Schmutz J, Tapia R, Thompson S, Tice HN, Spain JC, Gossett JG, Mattes TE. 2011. Genome sequence of the ethene- and vinyl chloride-oxidizing Actinomycete Nocardioides sp strain JS614. Journal of Bacteriology 193:3399-3400. Ondov BD, Bergman NH, Phillippy AM. 2011. Interactive metagenomic visualization in a Web Coleman NV, Wilson NL, Barry K, Brettin TS, Bruce DC, Copeland A, Dalin E, Detter JC, del Sanford RA, Cole JR, Tiedje JM. 2002. Characterization and description of Anaeromyxobacter Kube M, Beck A, Zinder SH, Kuhl H, Reinhardt R, Adrian L. 2005. Genome sequence of the Loffler FE, Yan J, Ritalahti KM, Adrian L, Edwards EA, Konstantinidis KT, Muller JA, Fullerton 91 Goris T, Schubert T, Gadkari J, Wubet T, Tarkka M, Buscot F, Adrian L, Diekert G. 2014. Zhang S, Wondrousch D, Cooper M, Zinder SH, Schuurmann G, Adrian L. 2017. Anaerobic Grostern A, Edwards EA. 2006. Growth of Dehalobacter and Dehalococcoides spp. during Deweerd KA, Mandelco L, Tanner RS, Woese CR, Suflita JM. 1990. Desulfomonile tiedjei gen- Luijten MLGC, de Weert J, Smidt H, Boschker HTS, de Vos WM, Schraa G, Stams AJM. 2003. Goris T, Schiffmann CL, Gadkari J, Schubert T, Seifert J, Jehmlich N, Von Bergen M, Diekert G. 89. Description of Sulfurospirillum halorespirans sp nov., an anaerobic, tetrachloroethene-respiring bacterium, and transfer of Dehalospirillum multivorans to the genus Sulfurospirillum as Sulfurospirillum multivorans comb. nov. International Journal of Systematic and Evolutionary Microbiology 53:787-793. 90. 2015. Proteomics of the organohalide-respiring Epsilonproteobacterium Sulfurospirillum multivorans adapted to tetrachloroethene and other energy substrates. Scientific Reports 5. 91. Insights into organohalide respiration and the versatile catabolism of Sulfurospirillum multivorans gained from comparative genomics and physiological studies. Environmental Microbiology 16:3562-3580. 92. dehalogenation of chloroanilines by Dehalococcoides mccartyi strain CBDB1 and Dehalobacter strain 14DCB1 via different pathways as related to molecular electronic structure. Environmental Science & Technology 51:3714-3724. 93. Holliger C, Hahn D, Harmsen H, Ludwig W, Schumacher W, Tindall B, Vazquez F, Weiss N, Zehnder AJB. 1998. Dehalobacter restrictus gen. nov. and sp. nov., a strictly anaerobic bacterium that reductively dechlorinates tetra- and trichloroethene in an anaerobic respiration. Archives of Microbiology 169:313-321. 94. degradation of chlorinated ethanes. Applied and Environmental Microbiology 72:428-436. 95. nov and sp-nov, a novel anaerobic, dehalogenating, sulfate-reducing bacterium. Archives of Microbiology 154:23-30. 96. dehalogenating bacterium from marine sediments. International Journal of Systematic and Evolutionary Microbiology 51:365-371. 97. trichloroethylene as electron acceptors. International Journal of Systematic Bacteriology 47:1262-1263. 98. Sung Y, Ritalahti KM, Sanford RA, Urbance JW, Flynn SJ, Tiedje JM, Loffler FE. 2003. Characterization of two tetrachloroethene-reducing, acetate-oxidizing anaerobic bacteria and their description as Desulfuromonas michiganensis sp nov. Applied and Environmental Microbiology 69:2964- 2974. 99. tetrachloroethylene- and cis-1,2-dichloroethylene-dechlorinating propionibacteria. Journal of Industrial Microbiology & Biotechnology 38:1667-1677. 100. Coleman NV, Mattes TE, Gossett JM, Spain JC. 2002. Phylogenetic and kinetic diversity of aerobic vinyl chloride-assimilating bacteria from contaminated sites. Applied and Environmental Microbiology 68:6162-6171. 101. Wild A, Hermann R, Leisinger T. 1997. Isolation of an anaerobic bacterium which reductively dechlorinates tetrachloroethene and trichloroethene. Biodegradation 7:507-511. Sun BL, Cole JR, Tiedje JM. 2001. Desulfomonile limimaris sp nov., an anaerobic Krumholz LR. 1997. Desulfuromonas chloroethenica sp. nov. uses tetrachloroethylene and Chang YC, Ikeutsu K, Toyama T, Choi D, Kikuchi S. 2011. Isolation and characterization of 92 Fathepure BZ, Tiedje JM. 1994. Reductive dechlorination of tetrachloroethylene by a 102. chlorobenzoate-enriched biofilm reactor. Environmental Science & Technology 28:746-752. 103. Mortan SH, Martin-Gonzalez L, Vicenta T, Caminal G, Nijenhuis I, Adrian L, Marco-Urrea E. 2017. Detoxification of 1,1,2-trichloroethane to ethene in a bioreactor co-culture of Dehalogenimonas and Dehalococcoides mccartyi strains. Journal of Hazardous Materials 331:218-225. 104. Moe WM, Yan J, Nobre MF, da Costa MS, Rainey FA. 2009. Dehalogenimonas lykanthroporepellens gen. nov., sp nov., a reductively dehalogenating bacterium isolated from chlorinated solvent-contaminated groundwater. International Journal of Systematic and Evolutionary Microbiology 59:2692-2697. 105. Bowman KS, Nobre MF, da Costa MS, Rainey FA, Moe WM. 2013. Dehalogenimonas alkenigignens sp nov., a chlorinated-alkane-dehalogenating bacterium isolated from groundwater. International Journal of Systematic and Evolutionary Microbiology 63:1492-1498. 106. Molenda O, Quaile AT, Edwards EA. 2016. Dehalogenimonas sp strain WBC-2 genome and identification of its trans-dichloroethene reductive dehalogenase, TdrA. Applied and Environmental Microbiology 82:40-50. 107. Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30:2114-20. 108. Kanitkar YH, Stedtfeld RD, Steffan RJ, Hashsham SA, Cupples AM. 2016. Loop-Mediated Isothermal Amplification (LAMP) for Rapid Detection and Quantification of Dehalococcoides Biomarker Genes in Commercial Reductive Dechlorinating Cultures KB-1 and SDC-9. Appl Environ Microbiol 82:1799-1806. 93 CHAPTER 3 Diversity and Abundance of the Functional Genes and Bacteria Associated with RDX Degradation at a Contaminated Site Pre- and Post- Biostimulation This chapter is being prepared for submission to a peer reviewed journal: Hongyu Dang and Alison M. Cupples. Diversity and Abundance of the Functional Genes and Bacteria Associated with RDX Degradation at a Contaminated Site Pre- and Post- Biostimulation. Abstract Bioremediation is becoming an increasingly popular approach for the remediation of sites contaminated with the explosive hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX). Multiple lines of evidence are often needed to assess the success of such approaches, with molecular studies frequently providing important information on the abundance of key biodegrading species. Towards this goal, the current study utilized shotgun sequencing to determine the abundance and diversity of functional genes (xenA, xenB, xplA, diaA, pnrB, nfsI) and species previously associated with RDX biodegradation in groundwater before and after biostimulation at an RDX contaminated Navy Site. For this, DNA was extracted from four and seven groundwater wells pre- and post- biostimulation, respectively. From a set of 65 previously identified RDX degraders, 31 were found within the groundwater samples, with the most abundant species being Variovorax sp. JS1663, Pseudomonas fluorescens, Pseudomonas putida and Stenotrophomonas maltophilia. Further, 9 RDX degrading species significantly (p<0.05) increased in abundance following biostimulation. Both the sequencing data and qPCR indicated xenA and xenB exhibited the highest relative abundance among the six genes. Several genes (diaA, nsfI, xenA and pnrB) exhibited higher relative abundance values in some wells following biostimulation. The study provides a comprehensive approach for assessing biomarkers during RDX bioremediation and 94 provides evidence that biostimulation generated a positive impact on a set of key species and genes. 1. Introduction Hexahydro-1,3,5-trinitro-1,3,5-triazine, also known as Royal Demolition Explosive or RDX, is a synthetic product and commonly used explosive (1). This chemical is classified as a possible human carcinogen and has different impacts on human health (1). Due to the denotation of ordnance, firing of munitions on military training ranges, and the manufacturing and transport of munitions, RDX has caused soil, groundwater and sediment contamination, with 39 identified sites currently on the National Priority List (2). Although a range of approaches (incineration, activated carbon columns and hydrogen peroxide) have been used to remediate RDX contaminated sites (1, 3, 4), more recently, interest has turned to bioremediation because of the potential cost savings. Successful pilot- and full-scale explosives bioremediation demonstrations have occurred at multiple DoD facilities including the Nebraska Ordnance Plant, Cornhusker Army Ammunition Plant, and Milan Army Ammunition Plant (5-10). In a study involving push- pull tests to measure in situ RDX degradation rates following the addition of corn syrup, lactose, emulsified oil, and ethanol, biostimulation with corn syrup was the most successful, illustrating 82% RDX removal (11). Bioaugmentation with Gordonia sp. strain KTR9 has also been shown to be an effective RDX bioremediation approach (8-10). A number of bacteria and functional genes have been associated with the biodegradation of RDX under aerobic or anaerobic conditions. A type I nitroreductase (encoded by nfsI) present in both Morganella morganii strain B2 and Enterobacter cloacae strain 96-3 was responsible for the nitroreduction of RDX (12). Type I, or oxygen insensitive, nitroreductases, use a two- electron reduction mechanism to reduce nitro groups under aerobic conditions (13, 14). Another 95 functional gene, pnrB, from a Pseudomonas sp. and Stenotrophomonas maltophilia, was also associated with RDX degradation (15). The enzyme encoded by the well-studied xplA gene has been linked with nitro group removal and ring cleavage by the genera Rhodococcus, Gordonia, Williamsia and Microbacterium (16-19). Under anaerobic conditions, the enzyme encoded by diaA from Clostridium kluyveri initiated RDX transformation through nitro group denitration (20, 21). Finally xenA and xenB, associated with the genus Pseudomonas, encode enzymes that primarily transform RDX to methylenedinitramine (22). Monitoring of these genes has previously been deployed at contaminated sites as evidence for RDX degradation (8, 9, 23). Current approaches to detect RDX degraders in groundwater have typically focused on PCR (9, 10). Although this approach has a high level of sensitivity, it is often limited by the number of genes that can be targeted. More recently, our group used high throughput qPCR to quantify key RDX degrading genes in groundwater at a contaminated site before and after biostimulation (24). However, the specificity of the qPCR primers was not evaluated and the possibility of false positives could not be ruled out. Building on our previous work, the overall aim of the current study was to quantify and explore the diversity of the genes and bacteria previously associated with RDX at an RDX contaminated site both before and after biostimulation with fructose. Although others have used shotgun sequencing to investigate these functional genes in ovine rumen, the approach, to our knowledge, has yet to be used on groundwater from an RDX contaminated site (25). The specific objectives were 1) to determine the relative abundance of each functional gene, 2) to ascertain the taxonomy of the microorganisms associated with each functional gene, 3) to investigate changes in gene abundance following biostimulation and 4) to ascertain if previously identified RDX degraders were present at the site and if their abundance changed following biostimulation. The approach 96 has the potential to provide a greater depth of knowledge compared to commonly used methods and represents a promising tool for evaluating biodegradation potential at RDX contaminated sites. 2. Materials and Methods 2.1 Sample Collection and DNA Extraction Groundwater was collected between 2017 and 2019 from an RDX contaminated site, both before and after biostimulation. Site and remediation details as well as the collection of pre- biostimulation samples were described in a previous study (24). Samples collected in 1-L amber glass bottles were shipped to the laboratory overnight on ice and were stored in the dark at 4 °C prior to DNA extraction. Samples were collected both before biostimulation (injection of fructose amended groundwater) and post biostimulation (~1.5 yr. later). All post biostimulation samples used for the shotgun sequencing analysis were newly collected since the previous study. Approximately 1L of groundwater was flowed through a 47-mm-diameter 0.22-µm filter (GSWG047S6, Millipore) using a vacuum pump and then the filter was put into the PowerBead tube from DNeasy PowerWater kit (Qiagen, Germany). The rest of the DNA extraction followed the manufacturer’s protocol. Extractions were performed in triplicate, were eluted in 50 µL and stored at -20 °C for further use. 2.2 High Throughput Sequencing DNA extracts (Table 3.1) were submitted for library generation and sequencing to the Research Technology Support Facility (RTSF) Genomics Core at Michigan State University (MSU). The libraries were prepared using the Takara SMARTer ThruPLEX DNA-Seq Kit and SMARTer DNA HT Dual Index Kit following manufacturer's recommendations. Completed libraries were quantified using a combination of Qubit dsDNA HS and Agilent 4200 TapeStation HS 97 DNA1000 assays. The libraries were pooled in equimolar proportions and the pool was quantified using the Kapa Biosystems Illumina Library Quantification qPCR kit. This pool was loaded onto one lane of an Illumina HiSeq 4000 flow cell and sequencing was performed in a 2x150bp paired end format using a HiSeq 4000 300 cycle SBS reagent kit. Base calling was performed by Illumina Real Time Analysis (RTA) v2.7.7 and output of RTA was demultiplexed and converted to FastQ format with Illumina Bcl2fastq v2.19.1. Table 3. 1. Collection dates, RDX concentrations, locations, DNA concentrations and names of groundwater sampling wells. Source Perched aquifer Shallow aquifer Perched aquifer Perched aquifer Shallow aquifer Perched aquifer Perched aquifer Well name MW22 MW32 MW48 MW60R MW62 MW66 MW67 Biostimulation statusa Date of collection - Post Pre Post - Post - Post Pre Post Pre Post Pre Post Not collected April 2019 Nov. 2017 April 2019 Not collected April 2019 Not collected April 2019 Nov. 2017 April 2019 Jan. 2018 April 2019 Jan. 2018 April 2019 RDX Concentration concentration (ng/µL) (µg/L) - <1 (4/30/18) 8 (9/11/17) <1 (4/30/18) 59 (1/27/18) 3 (4/28/18) 133 (1/27/18) 99 (4/28/18) 23 (9/11/17) <1 (4/30/18) 103 (1/27/18) 3 (4/28/18) 24 (1/27/18) 4 (4/28/18) Rep 1 Rep 2 - 35.1 11.8 48.0 - 46.8 - 18.7 101.0 55.0 19.8 44.5 4.9 40.7 - 44.9 15.2 48.3 - 46.2 - 29.3 133.0 54.0 18.1 48.4 5.6 33.5 aAll pre-bioaugmentation samples were collected in a previous study(24) 2.3 Taxonomic Analysis Taxonomic analysis of the metagenomes was achieved using the Meta Genome Rapid Annotation using Subsystem Technology (MG-RAST) (26) (Version 4.0.3.). The processing pipeline involved merging paired end reads, trimming low-quality regions with SolexaQA (27) and removing the artificial duplicate reads with dereplication. Gene calling was performed using FragGeneScan (28). Default setting (best hit classification, 10-5 e-value, 60% identity, and a minimal alignment length of 15 amino acids) with the databases ReqSeq (29) and KEGG Orthologs (KO) (30) were used for taxonomic and functional gene profiling. MG-RAST ID numbers and sequencing data have been summarized (Supplementary Table 3.1) and the datasets 98 are publicly available on MG-RAST. The MG-RAST data files were downloaded and analyzed in Microsoft Excel 2016 to generate the most abundant phylotypes in each sample. 2.4 Functional Gene Analysis Reference sequences for the functional genes relevant to RDX degradation (diaA (21), nfsI (12), pnrB (31), xenA (22), xenB (22) and xplA (16, 19)) were collected from FunGene (32) using a minimum HMM coverage of 70% (Supplementary Table 3.2). FunGene filters (Supplementary Table 3.2) were set for collecting reference sequences with no less than 60% identity to the consensus sequence of that gene. Unaligned protein sequences were downloaded and dereplicated by the function Clustering.jar derep developed by Ribosomal Database Project (https://github.com/rdpstaff/RDPTools). Dereplicated reference sequences were used to create the database in DIAMOND (double index alignment of next-generation sequencing data) (33), which was the alignment tool for all of functional genes. Before alignment, low quality sequences and Illumina adapters were removed using Trimmomatic (34) (Version 0.36) with the Paired End Mode settings, as described in the Trimmomatic manual (http://www.usadellab.org/cms/uploads/supplementary/Trimmomatic/TrimmomaticManual_V0. 32.pdf). As stated above, the processed datasets were then aligned to the dereplicated references using DIAMOND. Only reads that exhibited an identity of ≥ 60 % and an alignment length ≥ 49 amino acids to the reference sequences were retained as aligned reads to each sequence. For each, relative abundance values were calculated using the number of aligned reads divided by the total number of sequences for each sample. The relative abundance values were then normalized by (divided by) the number of dereplicated reference sequences for each gene to produce normalized relative abundance values. 99 2.5 Functional Gene Phylogenetic Trees Trees were generated using the 20 most abundant sequences for each target gene (averaged across all samples). For this, text files with the appropriate accession numbers were uploaded to COBALT: constraint-based alignment tool for multiple protein sequences (35). The downloaded alignments (fasta plus gaps) from COBALT were submitted for MAFFT (multiple alignment using fast Fourier transform) alignment using an online server (36) (Version 7). The website was also used to generate trees (by the Neighbor-Joining method) and the trees were exported in Newick format. The downloaded tree files were then uploaded to Interactive Tree of Life (37) (Version 5.5.1). Sequences were colored by phylum (and in some classes, by class) and relative abundance values were added using the Datasets function called simple bar chart. Additionally, the total relative abundance values of all aligned reference sequences for each gene were summed across all samples to generate a pie chart at the most appropriate classification level for each dataset (phylum, order, genus). The taxonomic information for all aligned reference sequences was obtained using the R package taxonomizr (38), RStudio (39) (Version 0.9.24) and R (40) (Version 4.0.2). 2.6 Co-occurrence Network of Genera Genera found by MG-RAST with at least 0.1% average relative abundance and 50% occurrence in post biostimulated wells were selected for building the correlation network. The correlation of the filtered genera was calculated with R packages Hmisc (41) (Version 4.4-1) and Matrix (42) (Version 1.2-18) using Spearman ranking correlation. Strong correlations (correlation coefficient ≥ 0.85) and Benjamini-Hochberg method adjusted p value (p < 0.01) were set to filter the correlation results. The filtered correlation results were used to build occurrence network with the R package igraph (43) (Version 1.2.6) and this was then visualized in Gephi (44). Both the 100 identified genera associated with RDX degradation and the genera previously associated with the functional genes generation were marked with different colors. 2.7 Analysis for Species Associated with RDX Biodegradation Shotgun sequences processed by Trimmomatic were merged with fastq-join (45) (Version 1.3.1). Previously identified RDX degraders were searched for in the National Center for Biotechnology Information (NCBI) taxonomy browser to find their lowest ranks (primarily rank of species) and taxonomy IDs (Supplementary Table 3.3). The merged reads were then aligned to the NCBI protein database (nr) with the taxids option in DIAMOND (33) (version 2.0.6). The searched results were restricted to maximum of 10 sequences with identity ≥ 85 % and query coverage ≥ 85 % and the resulting files were then imported into Megan (46) (community edition Version 6.19.7) for species taxonomy assignment. The phylogenetic tree of those species and normalized counts of reads were then visualized in Interactive Tree of Life (37) (Version 5.5.1). 2.8 Statistical Analysis The software, Statistical Analysis of Taxonomic and Functional Profiles, (STAMP) (47) (Version 2.1.3) was used to analyze both the metagenomic data (from MG-RAST) and the output of the RDX degrading species comparison across all wells from Megan. Specifically, STAMP was used to detect differences in the relative proportions of the taxonomic and functional profiles between various samples. STAMP analysis included Welch’s two-sided t-test for two groups (samples and live controls) (p < 0.05) to generate extended error bar figures. The parameters for the generation of these figures are listed (Supplementary Table 3.4). It was also used to generate heatmaps for the most abundant genera (relative abundance > 1.5%) and functional genes. Principle component analysis (PCA) based on genus was also completed in STAMP. 101 2.9 High Throughput Quantitative PCR The SmartChip Real Time PCR system was used to quantify the functional genes associated with RDX degradation in the DNA extracts using 12 primer sets developed previously by our group (24) in a 12 assay X 384 sample configuration. A subset of 12 primers were selected based our previous research on the assays’ combined theoretical coverage, their performance on the SmartChip and the need for only one plasmid per gene (24). Standards involved 10-fold serial plasmid dilutions (101-107 copies/reaction) with plasmids described in the same previous work. Gene copy numbers for the plasmids were calculated following the work of Ritalahti et al (48). Primers and plasmids were manufactured by Integrated DNA Technologies (IDT, Coralville, IA) and GenScript Biotech (Piscataway, NJ). Samples and assays were dispensed into a 72 × 72 nanowell chip with the Multisample Nanodispenser. On the chip, the final 100 nL individual reactions consisted of 1× LightCycler 480 SYBR Green I Master (Roche Applied Sciences, Indianapolis, IN), 0.5 μM each of the forward and reverse primers, DNA, and balance PCR grade water. Plasmid dilution series for each gene were in triplicate to generate the standard curves. DNA extracts were run on the chip primarily in triplicates or duplicates and negative controls contained water. Reactions that did not amplify or were identified as false positives were considered as missing data for all analyses. Reactions with a Ct value higher than 28 were assigned as false positives, as recommended for the SmartChip System (49). The gene copy number per milliliter or per gram of starting material was transformed from the gene copy number per reaction calculated based on the standard curve for the corresponding primer set and plasmid. Heatmaps were generated with log10 gene copy number per milliliter or gram using R package ggplot2 (50) (Version 3.3.2). 102 3. Results 3.1 RDX Concentrations RDX concentrations before and after biostimulation indicated the approach was successful for reducing contaminant concentrations. For example, a reduction from 103 µg/L to 3 µg/L was observed for MW66. The majority of the other wells (except MW60R) illustrated RDX reductions to 4 µg/L or below. 3.2 Microbial Community Analysis based on MG-RAST The overall similarity of microbial profiling of samples were tested with PCA based on the genus level results from MG-RAST (Supplementary Figure 3.1). The samples from different wells with similar microbial profiling were clustered together. Wells MW32 and 62 post clustered together (red circle). MW32 pre, MW66 pre and MW67 post clustered (green circle) as did MW62 pre, MW67 pre, MW22 post and MW60R (blue circle). The pre and post well samples clustered separately, indicating the microbial communities changed following biostimulation. The main phylotypes from the microbial profiles were determined for all samples at the class, order, family and genus levels (Supplementary Figure 3.2). The communities were primarily composed of Beta-, Alpha-, Gamma-, Delta- and Epsilon Proteobacteria (total average abundance > 38%, 11%, 9% and 5%) (Supplementary Figure 3.2A). However, in MW48, the class Clostridia (abundance >14 %) was also important. At order level, the majority of sequences classified as Burkholderiales (average abundance ~ 31.7%) (Beta Proteobacteria), Rhizobiales (average abundance ~ 5.5%) (Alpha Proteobacteria), Desulfuromonadales (average abundance ~ 4.6%) (Delta Proteobacteria), Bacteroidales (average abundance ~ 3.7%) (Bacteroidia), Pseudomonadales (average abundance ~ 3.7%) (Gamma Proteobacteria), Actinomycetales (average abundance ~ 3.7%) (Actinobacteria), Rhodocyclales (average abundance ~ 3.5%) 103 (Alpha Proteobacteria) and Clostridiales (average abundance ~ 3.0%) (Clostridia) (Supplementary Figure 3.2B). Campylobacterales (abundance >22%) (Epsilon Proteobacteria) in MW32 and MW62 and Sphingomonadales (average abundance > 5%) (Alpha Proteobacteria) in MW62 and MW67 were also particularly abundant. At the family level, phylotypes primarily classified within the Comamonadaceae (average abundance ~ 18.3%) (Burkholderiales), Burkholderiaceae (average abundance ~ 6.5%) (Burkholderiales), Oxalobacteraceae (average abundance ~ 3.8%) (Burkholderiales), Rhodocyclaceae (average abundance ~ 3.5%) (Rhodocyclales), Geobacteraceae (average abundance ~ 3.5%) (Desulfuromonadales) and Pseudomonadaceae (average abundance ~ 3.3%) (Pseudomonadales), unclassified Burkholderiales (average abundance ~ 2.2%) and Bradyrhizobiaceae (average abundance ~ 2.1%) (Rhizobiales) (Supplementary Figure 3.2C). At the genus level, the dominant genera were Polaromonas (average abundance ~ 5.0%) (Comamonadaceae), Acidovorax (average abundance ~ 5.0%) (Comamonadaceae), Albidiferax (average abundance ~ 3.9%) (Comamonadaceae), Geobacter (average abundance ~ 3.5%) (Geobacteraceae), Arcobacter (average abundance ~ 3.5%) (Campylobacteraceae), Pseudomonas (average abundance ~ 3.0%) (Pseudomonadaceae) and Burkholderia (average abundance ~ 2.8%) (Burkholderiaceae) (Supplementary Figure 3.2D). The 20 most abundant genera were characterized for each well. Eleven genera were present in at least 6 wells, including: Polaromonas, Acidovorax, Albidiferax, Geobacter, Pseudomonas, Burkholderia, Bacteroides, Cupriavidus, Dechloromonas, Variovorax and Leptothrix (Supplementary Figure 3.3). Six genera were more abundant following biostimulation (Acrobacter, Geobacter, Bacteroides, Clostridium, Paludibacter and Pelobacter) and 6 were less abundant (Leptothrix, Variovorax, Methylibium, Cupriavidus, Verminephrobacter, 104 Bradyrhizobium and Caulobacter) (t-test, p <0.05, targeting >1.5% abundance threshold) (Supplementary Figure 3.4). The most abundant genera from pre- and post- biostimulation of individual wells were also determined to investigate the changes in relative abundance for individual wells (Supplementary Figure 3.5). In MW32, Albidiferax, Bacteroides, Sulfuricurvum, Sulfurimonas, Paludibacter, Parabacteroides, Clostridium, Dechloromonas and unclassified Campylobacterales became abundant. In MW62, genera that increased after biostimulation were Geobacter, Pelobacter Bacteroides, Paludibacter and Desulfovibrio. Only two genera, Dechloromonas and Clostridium showed significant increases after biostimulation in MW66. Among those wells, Acidovorax, Poloromonas and Variovorax significantly decreased in post biostimulated samples. 3.3 Functional Genes Associated with RDX Biodegradation All six functional genes previously linked to RDX biodegradation were detected in the groundwater samples from this site (Supplementary Figure 3.6). Normalized relative abundance values were highest for xenB, followed by xenA, and both were present in all samples analyzed. Following this, diaA, nsfI and pnrB all illustrated similar normalized relative abundance values. The normalized relative abundance of xplA was low or absent in the majority of samples except for a pre-biostimulation sample (MW32). The taxonomic classifications, along with relative abundance values (pre- and post- biostimulation), of the most aligned reference sequences (top 20) for each gene are shown with phylogenetic trees (Figure 3.1). The most diverse set of sequences was obtained from diaA and the relative abundance of each varied across wells. Sequences classifying with Clostridium kluyveri (RDX degrader known to contain diaA) were present in the majority of wells (Figure 105 3.1A). The most abundant (again, top 20) nsfI (nitroreductase) sequences were almost exclusively classified with the Gamma Proteobacteria (Figure 3.1B). The RDX degrading nitroreductases from Morganella morganii strain B2 and Enterobacter cloacae strain 96-3 (12) were not found. Three functional genes (pnrB, xenA and xenB) all classified within the Alpha, Beta, Delta, or Gamma Proteobacteria (Figure 3.1C, D &E). Previous studies have associated the RDX degrading pnrB gene with Pseudomonas and Stenotrophomonas (31, 51) and although Stenotrophomonas was not detected here, the majority of pnrB sequences were classified as Pseudomonas (Figure 3.1C). The most abundant pnrB sequences classified as Azotobacter. The relative abundance of pnrB from Azotobacter was particularly high in post biostimulation wells (MW66 post, MW67 post, 106 Relative abundance in soils MW62 pre-(in light) and post-(in dark) biostimulation MW67 pre-(in light) and post-(in dark) biostimulation MW32 pre-(in light) and post-(in dark) biostimulation MW66 pre-(in light) and post-(in dark) biostimulation MW22, 48, 60R post-biostimulation from light to dark A: diaA B: nfsI nfsI in black were from Gammaproteobacteria Figure 3. 1. Phylogenetic trees were built for the aligned reference sequences of functional genes (A-F), the reference sequences were colored by phylum or class. The bars on the right illustrated the relative abundance (%) of aligned reference sequences in different samples, light and dark red denoted pre- and post-biostimulation from MW32. light and dark orange denoted pre- and post-biostimulation from MW62, light and dark green denoted pre- and post-biostimulation from MW66, light and dark blue denoted pre- and post-biostimulation from MW67, purple denoted different wells with only post-biostimulated samples. 107 Figure 3. 1. (continued) C: pnrB pnrB in black were from Gammaproteobacteria D: xenA xenA in black were from Betaproteobacteria 108 Figure 3. 1. (continued) E: xenB xenB in black were from Betaproteobacteria F: xplA All xplA were from Actinobacteria 109 MW48 and MW60R). For xenA (previously associated with RDX degradation by Pseudomonas spp.(22)), the most abundance sequences were classified as Sorangium cellulosum (Delta Proteobacteria) (Figure 3.1D) and two sequences were classified within the genus Pseudomonas. Other classifications for xenA included Acidovorax, Adenella, Methylobacillus, Methylovorus, Burkholderiales (Beta Proteobacteria) and Phyllobacterium (Alpha Proteobacteria). For xenB, the most abundant sequences were classified within the Beta Proteobacteria with no sequences classifying as Pseudomonas (previously linked to RDX degradation by xenB (22)) (Figure 3.1E). The well-studied xplA gene has been associated with the genera Rhodococcus, Gordonia, Williamsia (suborder Corynebacterineae, phylum Actinobacteria) and Microbacterium (suborder Micrococcineae, phylum Actinobacteria) (16-19, 52-56). In the present study, sequences classifying with the genera Gordonia, Rhodococcus, Williamsia and Microbacterium were some of the most abundant xplA sequences observed (Figure 3.1E). The relative abundance of the aerobic (nfsI, pnrB and xplA) and anaerobic (diaA, xenA and xenB) functional genes pre- and post- biostimulation was also investigated (Figure 3.2). The anaerobic genes illustrated greater levels of relative abundance (Figure 3.2B) compared to the aerobic genes (Figure 3.2A). In the four wells with both pre and post biostimulation data, two wells illustrated a dominance in aerobic genes pre biostimulation (MW32 and MW62) and two were more abundant post biostimulation (MW66 and MW67) (Figure 3.2A). For the anaerobic genes, three of the four wells illustrated a greater abundance post biostimulation (Figure 3.2B). 110 e v i t a l e r d e z i l a m r o N ) % ( e c n a d n u b a 9E-05 6E-05 3E-05 0E+00 pre1 pre2 post1 post2 A sum of nfsI, pnrB and xplA pre1 pre2 post1 post2 B sum of diaA, xenA and xenB 4E-04 3E-04 2E-04 1E-04 0E+00 Figure 3. 2. Normalized relative abundance (%) of the total aerobic (nfsI, pnrB and xplA) (A) and anaerobic (diaA, xenA and xenB) (B) functional genes relevant to RDX biodegradation across all monitoring wells (MW) in replicate DNA extracts. The legend terms post and pre refer to the post- and pre-biostimulation samples, respectively. To better illustrate the diversity of the taxonomic classifications associated with each gene, the taxonomic information of all functional gene reference sequences observed, across all wells, was determined (Supplementary Figure 3.7). The pie-charts generated were classified to the phylum, order or genus level, depending on the data for that gene (to allow a readable number of labels). Similar to the most abundant data for diaA (as discussed above) when all sequences were included for diaA, a large number of phyla were noted, with the majority classifying as Proteobacteria, Firmicutes and Chloroflexi. The majority of the nsfI sequences were classified within the genera Vibrio, Klebsiella, Cedecea and Enterobacter. For pnrB, the majority were classified within the genera Pseudomonas, Azobacter, Pantoea, Massilia and Burkholderia. The patterns for xenA and xenB were similar (classified to the order level), with Burkholderiales, Pseudomonadales, Enterobacterales and Rhizobiales being commonly found. The least amount of diversity was noted for xplA with sequences classifying within the genera Rhodococcus, Williamsia, Gordonia, Mycobacterium, Mycolicibacterium and Microbacterium. The genera associated with those functional genes that had not been identified for RDX degradation were denoted as potential degraders, which were used for following co-occurrence analysis. 111 3.4 Co-occurrence of Genera Associated with RDX Biodegradation A co-occurrence network, with strong correlation between each node, was built for better illustrating the relationship of the main genera found with MG-RAST (Figure 3.3). A total of 16 identified genera (colored in red) associated with RDX degradation were found. Based on classifications of those functional genes, potential genera for RDX degradation were also denoted on the network. The majority of the potential genera (28, colored in green) were found for xenA or xenB, followed by diaA (3, colored in orange). Only one genus was shown for pnrB and xplA. The co-occurrence network was also processed to group those nodes into 7 different modules (Supplementary Figure 3.8). The modules colored in dark green (bottom right corner) and blue grouped Clostridium which was identified for generating diaA, all three potential genera for diaA and several other identified genera. Rhodococcus and the potential genera for xplA were grouped in purple. The modules colored in light green and orange grouped the rest of identified genera and most potential genera for xenA or xenB. 112 Identified genera Potential genera for xenA and xenB Potential genera for pnrB Potential genera for diaA Potential genera for xplA Others Figure 3. 3. Co-occurrence network based on spearman correlation (rho > 0.85 and p-value < 0.01) of main genera found in all samples from post-biostimulated wells. Only genus with an average abundance > 0.1% and present in at least 50% of samples were considered. Node size indicates the relative abundance (0.1% ~ 5.46%). Nodes colored in red: identified genus associated with RDX degradation. Nodes colored in orange: potential genus to generate diaA. Nodes colored in pink: potential genus to generate pnrB. Nodes colored in blue: potential genus to generate xplA. Nodes colored in green: potential genus to generate xenA or xenB. No potential genus to generate nfsI was found. 113 3.5 Presence of Known RDX Degraders Based on the result of MG-RAST, a number of genera associated with RDX degradation were found. The classification to those species within those genera were performed with cutoffs of identity ≥ 85 % and query coverage ≥ 85 %. A phylogram tree for species previously identified as RDX degraders was generated (Figure 3.4). The analysis indicated the presence of 31 RDX degrading bacterial species across all samples. Two fungal species previously associated with RDX degradation were also detected (data not shown). From the 31 bacterial species identified, the genus Variovorax showed the highest number of alignments. Others with higher alignments included the genera Pseudomonas, Stenotrophomonas, Geobacter and Agrobacterium. The statistical analysis indicated 9 RDX degrading species demonstrated a significant increase in abundance (p<0.05) in the post-biostimulation samples compared to the pre-biostimulation samples (Figure 3.5). 114 Normalized counts in soils MW32 pre-(in light) and post-(in dark) biostimulation MW62 pre-(in light) and post-(in dark) biostimulation MW66 pre-(in light) and post-(in dark) biostimulation MW67 pre-(in light) and post-(in dark) biostimulation MW22, 48, 60R post-biostimulation from light to dark Species classification Bacteroidetes Actinobacteria Firmicutes Alphaproteobacteria Betaproteobacteria Deltaproteobacteria Gammaproteobacteria Figure 3. 4. Phylogram constructed with reads assigned (identity ≥ 85% and query coverage ≥ 85%) to the species associated with RDX degradation across all monitoring wells (MW) in replicate DNA extracts. Each species was colored with phylum or class from Proteobacteria. The bars in the outside indicated the normalized counts assigned to the species, missing bars meant zero counts. Figure 3. 5. Species associated with RDX degradation showed significant differences before and after biostimulation across all wells. The extended error bar was created using Welch’s t-test (two sided) with the default CI option (Welch’s inverted), default multiple test correction (no correction) and default p value filter of 0.05. 115 3.6 KEGG Pathways The changes in the relative abundance of genes associated with xenobiotics biodegradation and metabolism were investigated using MG-RAST and STAMP (Supplementary Figure 3.9). After biostimulation, there was a significant increase for the genes involved in nitrotoluene degradation (Supplementary Figure 3.9 A). At function level, hydrogenase, carboxylesterase, N- ethylmaleimide reductase, 4-carboxymuconolactone decarboxylase and catechol 2,3-dioxygenase significantly were significantly more abundant after biostimulation (Supplementary Figure 3.9 B). 3.7 High Throughput qPCR All samples were amplified with 16S rRNA specific primers as well as functional gene primers (Supplementary Figure 3.10 A and B). Samples from this study and a previous study (24) were included in the analysis. Three genes (xenA, xenA and nsfI) were commonly found in both the groundwater and soil samples, indicating their possible widespread occurrence in the environment. In both the groundwater and sediment DNA extracts, diaA was not detected. The gene pnrB (primer: pnrB_PS5) and xplA were only detected in a limited number of groundwater DNA extracts. The maximum copy number was correlated with the relative abundance of that gene by the Spearman’s rank correlation test (Supplementary Table 3.5). Two genes (xenA and xplA) illustrated a significant correlation (p<0.05) between the two methods (shotgun sequencing analysis and qPCR). 4. Discussion RDX concentration changes indicated the biostimulation approach was successful at the majority of groundwater wells. To our knowledge, this is the first investigation of the genes and phylotypes involved in RDX biodegradation using shotgun sequencing in groundwater from an 116 RDX contaminated site. Although several genes (diaA, nsfI, xenA and pnrB) exhibited higher relative abundance values in some wells following biostimulation, the differences of relative abundance between the pre- and post- biostimulation were not significant or consistent. The lack of a statistical difference in the current work may be related to the number of samples studied. When these genes were grouped by aerobic or anaerobic conditions, in the shallow aquifer (MW32 and MW62) there was a trend to transfer from aerobic to anaerobic functional genes. While in perched aquifer, although anaerobic genes were dominant, overall aerobic genes were still enriched (MW66), which may share similarity with a slow aerobic RDX degradation under microaerophilic (dissolved oxygen < 0.04 mg/L) condition (57). In a previous study, using environmental samples from two Navy sites, our group found that both xplA and xenA significantly increased during RDX biodegradation compared to the controls in both groundwater and sediment microcosms (58). Further, in a limited number of microcosms in the previous study, xenB gene copy numbers increased. Comparing the current results to those previously obtained, must be performed with caution, as previous studies have used different detection methods and/or examined different sites. For example, one of the first studies on these functional genes used conventional PCR on groundwater from two sites (Picatinny Arsenal and Pueblo Chemical Depot) where RDX bioremediation (through the addition of organic substrates) was being examined (23). In that case, the researchers did not detect the targeted genes (xplA, xenA xenB, onr and hydA) in any of the groundwater samples. They suggested the lack of detection may have resulted from i) the absence of the genes, ii) low gene copy numbers or iii) limitations associated with the primers used. Several studies have targeted a subset of these genes during the evaluation of bioaugmentation for RDX remediation. In 2015, xenB and xplA were targeted at Umatilla 117 Chemical Depot (UMCD) in Umatilla, OR, as part of two forced-gradient bacterial transport tests of a mixed culture (strains in the genera Gordonia, Rhodococcus, Psuedomonas) or a single culture of Gordonia sp. strain KTR9. Through qPCR of xplA, xenB and a marker gene (kanamycin resistance gene), the researchers found that the three RDX-degrading strains were effectively introduced and transported within the aquifer (10). Another study at UMCD compared RDX removal rates under bioaugmentation with Gordonia sp. strain KTR9 to rates with biostimulation (low or high fructose) and also targeted xplA (8). They found that bioaugmentation achieved RDX concentration reductions comparable to those obtained by high carbon biostimulation while requiring substantially less fructose and thus resulting in cost benefits and less secondary water quality impacts. More recently, the genes associated with RDX biodegradation were investigated at the same site as the current study (Naval Base Kitsap, Bangor Site F near Silverdale, WA) (9, 24). One such project investigated xplA and xenB during bioaugmentation with Gordonia sp. KTR9 and Pseudomonas fluorescens strain I-C cells and found that these strains were transported 13 m downgradient over 1 month. The research also demonstrated that bioaugmentation was a viable technology for accelerating RDX cleanup. The other study (by our group), designed new primers for high throughput quantitative PCR to target all six genes (24). The final 49 newly designed primer sets improved upon the theoretical coverage of published primer sets, and this improvement corresponded to more detections in the environmental samples. All genes, except diaA, were detected in the site samples, with xenA and xenB being the most common, agreeing with the results presented here. Here, a key finding was the detection of a wide range of RDX degraders from numerous genera. Further, the identified genera appeared in the 20 most abundant genera (Supplementary 118 Figure 3.3) including: Geobacter, Pseudomonas, Burkholderia, Clostridium, Variovorax, Desulfovibrio, Bacillus, Desulfitobacterium, Prevotella, Rhodococcus. The potential genera for RDX degradation were dominant in at least 10 samples including: Acidovorax, Cupriavidus, Janthinobacterium, Dechloromonas, Ralstonia, Bradyrhizobium (all associated with xenA or xenB). Potential genera for generation pnrB, including: Azotobacter and Bordetella were only found in one sample for each as was Mycobacterium (associated with xplA). With the presence of the functional genes, the species associated with RDX degradation were explored based on relatively strict thresholds (identity ≥ 85 % and query coverage ≥ 85 %) so that the reads aligned very specifically to a taxon, even if the reads aligned to a less specific gene (46). From the analysis of KEGG pathways in category of xenobiotics biodegradation and metabolism, nitrotoluene degradation was significantly increased after biostimulation. The nitroreductase (KEGG ID K10679) from nitrotoluene degradation was found to be more abundant after biostimulation while it was not statistically significant. Nitroreductase was identified from Enterobacter cloacae isolated from a munitions facility because of its ability to metabolize trinitrotoluene (59, 60). This was consistent with the result of relative abundance of nfsI (Supplementary Figure 3.6). To date, previous studies have used other detection methods to explore microbial diversity at RDX contaminated sites. For example, the diversity of RDX degraders was examined at Picatinny Arsenal and Pueblo Chemical Depot using 16S rRNA gene amplicon sequencing (23). In that study, Rhizobiales and Geobacter were detected in nonbiostimulated samples and Bacteroidetes were detected in biostimulated samples. Pseudomonas and Clostridium were identified in both types of samples. In another project (using clone libraries), microbial communities were examined before and after the addition of acetate or lactate at 119 Picatinny Arsenal (61). In that project, Beta Proteobacteria accounted for more than half of the phylotypes after the addition of substrates. Alpha Proteobacteria and Gamma Proteobacteria decreased, while Delta Proteobacteria increased. Actinobacteria and Firmicutes were also detected, and Clostridia were enriched in samples following lactate addition. A similar pattern of Beta Proteobacteria dominance was also observed in groundwater samples at least two months after the addition of acetate at Iowa Army Ammunition Plant (62). Another study examined microbial community changes before and after biostimulation at Los Alamos National Laboratory. At that site, Rhodococcus (more than 28%) and Pseudomonas (about 6%) were abundant in the indigenous microbial community (63). The abundance of Pseudomonas increased to ~ 50% with the addition of safflower oil while that of Rhodococcus decreased to less than 5% with the addition of either acetate or safflower oil. In another experiment, waste glycerol (WG) was added to enhance in situ RDX biodegradation (64). Geobacter, Clostridium, Klebsiella and Bacteroidales, and Sulfuricurvum became enriched in WG impacted monitoring wells. Previous cost-effective analysis indicated a cost of $79-254 achieved an average RDX transformation rate of 1.20/day in bioaugmentation while a cost of $4 achieved an average rate of 0.49/day in bioaugmentation (8). This low cost for biostimulation (20-63 times lower) will take only 2.5 times longer than bioaugmentation to remove the same amount of RDX. With both diverse functional genes and degraders detected in the indigenous microbial community, it has been suggested that biostimulation is a reasonable and effective alternative to bioaugmentation when cost is a major concern. In summary, the functional genes and species associated with RDX were both detected in pre- and post-biostimulated samples. However, although sequences aligning with known RDX 120 degraders were present, it is unclear if these microorganisms were involved in RDX biodegradation at this site. The approach highlighted the importance of xenA and xenB and demonstrated that a large number of identified and potential RDX degraders were present both pre- and post-bioaugmentation. Further, a subset of these functional gene and degraders was significantly enriched following biostimulation, providing an additional line of evidence for assess the biodegradation potential and evaluating the success of the remediation approach. As the cost of the shotgun sequencing is likely to decrease, in the future, this approach has the potential to be deployed at a larger number of contaminated sites. Acknowledgements Thanks to M. M. Michalsen (U.S. Army Engineer Research Development Center) for providing the groundwater. Thanks to Malcolm Gander (Naval Facilities Engineering Command) for partially funding this project and thanks to Craig Tobias (University of Connecticut) for facilitating project funding. 121 APPENDIX 122 Supplementary Table 3. 1. MG-RAST analysis data for datasets from DNA extracts of groundwater samples pre- and post-biostimulation. Monitoring Well MG-RAST ID Upload: bp Count Upload: Sequences Count Upload: Mean Sequence Length Upload: Mean GC percent Artificial Duplicate Reads: Sequence Post QC: bp Count Count Post QC: Sequences Count Post QC: Mean Sequence Length Post QC: Mean GC percent APPENDIX MW22_post1 mgm4886589.3 1,741,804,251 bp 7476242 233 ± 36 bp 53 ± 11 % 1292469 1,429,965,062 bp 6112829 234 ± 36 bp 53 ± 11 % MW22_post2 mgm4886608.3 2,099,539,542 bp 9109772 230 ± 36 bp 52 ± 10 % 2270230 1,576,080,242 bp 6769037 233 ± 36 bp 52 ± 10 % MW32_post1 mgm4886604.3 1,954,836,759 bp 8263883 237 ± 35 bp 46 ± 14 % 1328041 1,629,268,603 bp 6860711 237 ± 35 bp 47 ± 14 % MW32_post2 mgm4886595.3 1,985,444,228 bp 8565024 232 ± 36 bp 47 ± 13 % 1350194 1,661,970,978 bp 7139317 233 ± 36 bp 48 ± 13 % MW32_pre1 mgm4886601.3 1,661,003,748 bp 7047448 236 ± 35 bp 60 ± 11 % MW32_pre2 mgm4886594.3 1,527,083,893 bp 6406285 238 ± 34 bp 61 ± 10 % 759357 749912 1,459,056,388 bp 6194085 236 ± 35 bp 60 ± 11 % 1,327,512,629 bp 5572112 238 ± 34 bp 61 ± 10 % MW48_post1 mgm4886590.3 1,586,090,675 bp 6782319 234 ± 36 bp 43 ± 14 % 1322672 1,268,425,462 bp 5399369 235 ± 36 bp 44 ± 14 % MW48_post2 mgm4886587.3 1,778,065,022 bp 7522747 236 ± 35 bp 41 ± 12 % 1291252 1,465,710,838 bp 6169619 238 ± 35 bp 41 ± 12 % MW60R_post1 mgm4886592.3 1,899,735,011 bp 8131466 234 ± 36 bp 62 ± 10 % 1707435 1,496,289,863 bp 6337044 236 ± 35 bp 61 ± 10 % MW60R_post2 mgm4886588.3 1,653,372,637 bp 7091470 233 ± 36 bp 62 ± 10 % 1291110 1,341,117,043 bp 5707179 235 ± 36 bp 61 ± 10 % MW62_post1 mgm4886599.3 1,422,045,219 bp 5955796 239 ± 34 bp 49 ± 16 % MW62_post2 mgm4886607.3 1,534,652,060 bp 6486643 237 ± 35 bp 51 ± 16 % MW62_pre1 mgm4886598.3 1,189,280,134 bp 4946375 240 ± 34 bp 59 ± 9 % MW62_pre2 mgm4886605.3 1,101,957,157 bp 4558998 242 ± 33 bp 60 ± 8 % MW66_post1 mgm4886593.3 1,614,662,781 bp 6850124 236 ± 35 bp 54 ± 12 % xMW66_post2 mgm4886596.3 1,587,569,892 bp 6796355 234 ± 36 bp 52 ± 13 % MW66_pre1 mgm4886597.3 1,503,207,116 bp 6258351 240 ± 34 bp 58 ± 10 % MW66_pre2 mgm4886600.3 1,308,725,413 bp 5468213 239 ± 34 bp 58 ± 10 % MW67_post1 mgm4886591.3 1,322,169,825 bp 5542737 239 ± 34 bp 54 ± 13 % MW67_post2 mgm4886606.3 1,512,555,910 bp 6371805 237 ± 35 bp 55 ± 12 % MW67_pre1 mgm4886602.3 1,517,750,130 bp 6345506 239 ± 34 bp 60 ± 8 % MW67_pre2 mgm4886603.3 1,392,698,023 bp 5787985 241 ± 34 bp 61 ± 8 % 920874 991426 612288 546177 788212 747488 663943 583529 716539 835319 764462 696637 1,192,591,889 bp 4975023 240 ± 34 bp 49 ± 16 % 1,289,201,671 bp 5427940 238 ± 35 bp 51 ± 16 % 1,030,173,642 bp 4281608 241 ± 34 bp 59 ± 9 % 956,669,806 bp 3955392 242 ± 33 bp 60 ± 9 % 1,411,464,163 bp 5985586 236 ± 35 bp 54 ± 12 % 1,397,130,817 bp 5981263 234 ± 36 bp 52 ± 13 % 1,323,804,007 bp 5513662 240 ± 34 bp 58 ± 10 % 1,153,919,440 bp 4823230 239 ± 34 bp 58 ± 10 % 1,137,449,146 bp 4764623 239 ± 34 bp 54 ± 13 % 1,298,629,352 bp 5465345 238 ± 35 bp 55 ± 12 % 1,320,419,861 bp 5517949 239 ± 34 bp 60 ± 8 % 1,211,531,946 bp 5029835 241 ± 33 bp 60 ± 8 % 123 Supplementary Table 3. 2. FunGene filters for obtaining the reference sequences and the number of collected sequences before and after dereplication. Gene Webpage name Minimum HMM coverage Minimum score Number of sequences Number of sequences collected after dereplication diaA nfsI pnrB xenA xenB xplA diaA_new nfsI pnrB xenA xenB xplA 70% 70% 70% 70% 70% 70% 350 302 309 497 475 1000 116 15742 275 3085 6336 11 90 653 131 957 1371 7 Supplementary Table 3. 3. Identified RDX degraders with the lowest rank name and taxonomy ID from NCBI. reference Lowest rank name in NCBI NCBI taxonomy ID Number of subtrees D Acetobacterium malicum Acetobacterium paludosum NCBI Rank species species 52692 52693 Acremonium sp. HAW-OCF3 species 311340 Anaeromyxobacter dehalogenans species 161493 Anaerovibrio lipolyticus species 82374 Bacillus sp. HPB-2 species 259962 Bacillus sp. HPB-3 species 259965 Bullera unica Burkholderia sp. Citrobacter freundii species species species 57474 36773 546 Clostridium acetobutylicum ATCC 824 strain 272562 Paraclostridium bifermentans Paraclostridium bifermentans species species 1490 1490 Clostridium geopurificans species 558153 Clostridium kluyveri species 1534 Clostridium polysaccharolyticum species 29364 Clostridium sp. EDB2 species 261021 Clostridium sp. Clostridium sp. Clostridium sp. Clostridium sp. Clostridium sp. species species species species species 1506 1506 1506 1506 1506 Desulfitobacterium chlororespirans species 51616 Desulfovibrio desulfuricans Desulfovibrio sp. Desulfovibrio sp. Desulfovibrio desulfuricans Desulfovibrio desulfuricans Desulfovibrio gigas Desulfovibrio vulgaris Enterobacter cloacae species species species species species species species species 876 885 885 876 876 879 881 550 124 0 0 0 2 3 0 0 0 0 E 18 0 6 6 0 2 1 1 0 E 0 E 0 E 0 E 0 E 1 3 0 E 0 E 3 3 1 4 43 Strain or species for RDX or its metabolites degradtion Acetobacterium malicum Strain HAAP-1 Acetobacterium paludosum Acremonium sp. HAW-OCF3 Anaeromyxobacter dehalogenans Strain K (no ATCC number) Anaerovibrio lipolyticus Bacillus (HPB2) Bacillus (HPB3) Bullera unica strain HAW-OCF2 Burkholderia sp.BL Citrobacter freundii Clostridium acetobutylicum (ATCC 824) Clostridium bifermentans A Clostridium bifermentans strain HAW-1 Clostridium geopurificans MJ1T Clostridium kluyveri ATCC8527 Clostridium polysaccharolyticum Clostridium sp. EDB2 (65) (66) (67) (68) (69) (70) (70) (67) (71) (72) (73) (74) (75) (76) (21) (69) (77) Clostridium sp. HAW-E3 (75, 78) Clostridium sp. HAW-EB17 Clostridium sp. HAW-G3 Clostridium sp. HAW-G4 Clostridium sp. HAW-HC1 Desulfitobacterium chlororespirans Strain Co23 Desulfovibrio desulfuricans EFX-DES Desulfovibrio sp. HAW-EB18 Desulfovibrio sp. HAW-ES2 Desulfovibrio spp. desulfuricans A Desulfovibrio spp. desulfuricans B Desulfovibrio spp. gigas Desulfovibrio spp. vulgaris Enterobacter cloacae strain 96-3 (79) (78) (75, 78) (75, 78) (68) (80) (79) (78) (81) (81) (81) (81) (12) Supplementary Table 3. 3. (continued) Geobacter metallireducens Geobacter sulfurreducens Gordonia sp. KTR9 Gordonia sp. YY1 Halomonas (HAW-OC4) Klebsiella pneumoniae Strain SCZ-1 Marinobacter (HAW-OC1) Methylobacterium extorquens Methylobacterium organophilum Methylobacterium rhodesianum Methylobacterium sp. JS178 Methylobacterium sp. strain BJ001 Morganella morganii Morganella morganii strain B2 Penicillium sp. HAW-OCF5 Phanerochaete chrysosporium Prevotella ruminicola Providencia rettgeri Pseudoalteromonas (HAW-OC2) Pseudoalteromonas (HAW-OC5) Pseudomonas (HPB1) Pseudomonas fluorescens I-C Pseudomonas putida II-B Pseudomonas sp. HK-6 Rhizobium rhizogenes BL A Rhodococcus sp. Strain A Rhodococcus sp. strain DN22 Rhodococcus species isolate T7 Rhodococcus species isolate T9N Rhodococcus strain YH1 Rhodococcus rhodochrous strain 11Y Rhodotorula mucilaginosa strain HAW- OCF1 Serratia marcescens Shewanella halifaxensis sp. strain HAW- EB4T Shewanella oneidensis Strain MR1 Shewanella sediminis sp. strain HAW- EB3T Shewanella sp. HAW EB1 Shewanella sp. HAW EB2 Shewanella sp. HAW-EB5 Stenotrophomonas maltophilia OK-5 Stenotrophomonas maltophilia PB1 (68) (68) (56) (82) (83) (84) (83) (85) (85) (85) (86) (85) (72) (12) (67) (87) (69) (72) (83) (83) (70) (22) (22) (88) (71) (57) (89) (17) (17) (54) (18) (67) (90) (91) (68) (92) (79) (79) (79) (31) (93) Geobacter metallireducens Geobacter sulfurreducens species species 28232 35554 Gordonia sp. KTR9 species 337191 Gordonia sp. YY1 species 396712 Halomonas sp. species 1486246 Klebsiella pneumoniae species 573 Marinobacter sp. species 50741 Methylorubrum extorquens species Methylobacterium organophilum species 408 410 Methylorubrum rhodesianum species 29427 Methylobacterium sp. JS178 species 316459 Methylorubrum sp. species 2282524 Morganella morganii Morganella morganii species species 582 582 Penicillium sp. HAW-OCF5 species 311341 Phanerochaete chrysosporium species 5306 Prevotella ruminicola Providencia rettgeri Pseudoalteromonas sp. Pseudoalteromonas sp. Pseudomonas sp. Pseudomonas fluorescens Pseudomonas putida species species species species species species species 839 587 53249 53249 306 294 303 Pseudomonas sp. HK-6 species 342605 Agrobacterium rhizogenes Rhodococcus sp. species species 359 1831 Rhodococcus sp. DN22 species 357684 Rhodococcus sp. T7 species 627444 Rhodococcus sp. T9N species 627445 Rhodococcus sp. YH1 species 89066 Rhodococcus rhodochrous species 1829 Rhodotorula mucilaginosa species 5537 Serratia marcescens species 615 Shewanella halifaxensis HAW-EB4 strain 458817 Shewanella oneidensis MR-1 strain 211586 Shewanella sediminis HAW-EB3 strain 425104 Shewanella sp. species 50422 Shewanella sp. species 50422 Shewanella atlantica species 271099 Stenotrophomonas maltophilia species Stenotrophomonas maltophilia species 40324 40324 125 2 3 0 0 0 E 406 0 E 5 0 0 0 0 E 10 10 0 1 2 3 0 E 0 E 0 E 47 39 0 1 0 0 0 0 0 8 0 27 0 0 0 0 E 0 E 0 37 37 Supplementary Table 3. 3. (continued) Streptococcus bovis A Variovorax sp. Strain JS1663 Williamsia sp. KTR4 Desulfosporosinus B (69) (94) (56) (95) Streptococcus equinus species 1335 Variovorax sp. JS1663 species 1851577 Williamsia sp. KTR4 species 337192 Desulfosporosinus genus 79206 Fusobacteria isolate HAW-EB21 B (75, 79) Fusobacteria phylum 32066 Aspergillus niger C Cladosporium cladosporioides C (96) (71) Aspergillus niger species 5061 Cladosporium cladosporioides species 29917 7 0 0 NC NC 12 1 A: The names of those identified species in the paper were revised in NCBI to another name. B: Rank of the two were higher than species. C: The two species were fungus, due to too many clades, the species names could not be displayed when analyzed in Megan. D: The number means the identified microorganism within that species, NC means not checked. E: The identified strain name from the paper could not be searched in NCBI taxonomy browser, for example: Burkholderia sp.BL was assigned to Burkholderia sp. which belonged to unclassified Burkholderia. Supplementary Table 3. 4. STAMP analysis parameters for generating Supplementary Figures 3.4 and 3.5. Filters were applied to limit the number of genera or functions shown in each figure. All tests used Welch’s t-test (two sided) with the default CI option (Welch’s inverted) and default multiple test correction (no correction). For each test, two groups were compared (pre and post biostimulation samples). A Genus MW32 Genus MW62 Genus MW66 Genus MW67 Parent Level Profile Level Unclassified Filtering p-value filter > B Parent Level Profile Level Unclassified Filtering p-value filter > Effect size filter 1 Different between two proportions Effect size < Entire sample Entire sample Entire sample Entire sample Genus Retain Genus Retain Genus Retain Genus Retain unclassified reads unclassified reads unclassified reads unclassified reads 0.05 0.05 0.05 0.05 Function 45 most different Entire sample Function Retain unclassified reads 0.05 0.09 126 Supplementary Table 3. 5. Spearman’s rank correlation parameters between gene copy number of qPCR and relative abundance of genes associated with RDX degradation. Gene nfsI pnrB xenA xenB xplA S 1288.4 780.89 450 932.98 212.09 p value 0.1826 0.4402 0.02197 0.8836 0.00013 rho -0.3296 0.19412 0.535604 0.037171 0.781128 Supplementary Figure 3. 1. Principle component analysis of all samples based on the genus results from MG-RAST. Clustered samples were marked in circles. 127 Supplementary Figure 3. 2. The most abundant phylotypes in each sample at the class (A), order (B) family (C), and genus (D) levels. For each classification, phylotypes with an average relative abundance across all samples less than 1% were placed within "other". 128 Supplementary Figure 3. 3. Relative abundance (%) of the 20 most abundant genera in duplicated samples from each well. Pre and post refer to the pre- and post-biostimulation samples, respectively. 129 Supplementary Figure 3. 4. A comparison of those significantly different between pre- and post- biostimulation wells from the abundant genera (relative abundance ≥ 1.5) (p < 0.05, Welch's two sided t- test). 130 A B B Supplementary Figure 3. 5. Comparison of the most abundant genera pre- and post biostimulation in MW32 (A), MW62 (B), MW66 (C) and MW67 (D) with significant differences (p < 0.05, Welch's two sided t-test). 131 C B D B Supplementary Figure 3. 5. (continued) 132 pre1 pre2 post1 post2 diaA pre1 pre2 post1 post2 pnrB pre1 pre2 post1 post2 xenB 8E-05 6E-05 4E-05 2E-05 0E+00 6E-05 4E-05 2E-05 0E+00 ) % ( e c n a d n u b a e v i t a l e r d e z i l a m r o N 3E-04 2E-04 1E-04 0E+00 1.2E-05 9.0E-06 6.0E-06 3.0E-06 0.0E+00 1.2E-04 9.0E-05 6.0E-05 3.0E-05 0.0E+00 8E-05 6E-05 4E-05 2E-05 0E+00 pre1 pre2 post1 post2 nfsI pre1 pre2 post1 post2 xenA pre1 pre2 post1 post2 xplA Supplementary Figure 3. 6. Normalized relative abundance (%) of the functional genes relevant to RDX biodegradation across all monitoring wells (MW) in replicate DNA extracts. The legend terms post and pre refer to the post- and pre-biostimulation samples, respectively. 133 A. diaA Proteobacteria unclassified Bacteria unclassified marine sediment metagenome Firmicutes Other Euryarchaeota Elusimicrobia Chloroflexi Candidatus Atribacteria Candidatus Cloacimonetes B. nfsI Klebsiella CedeceaEnterobacter Citrobacter Edwardsiella Beauveria Escherichia Buttiauxella Salmonella Raoultella Other Vibrio C. pnrB Azotobacter Pantoea Massilia Burkholderia Paraburkholderia Other Pseudomonas Actinobacteria candidate division WOR-3 candidate division Zixibacteria Candidatus Desantisbacteria Candidatus Latescibacteria Planctomycetes Spirochaetes Thermotogae unclassified metagenome Trabulsiella Shigella Leclercia Kluyvera Pseudescherichia Pluralibacter Photobacterium Yokenella Lelliottia Vibrio Bordetella Phytobacter Supplementary Figure 3. 7. Taxonomy of microorganisms associated with aligned references sequences of functional genes: diaA (phylum level, A), nfsI (genus level, B), pnrB (genus level, C), xenA (order level, D), xenB (order level, E) and xplA (genus level, F) sequences across all soils. 134 D. xenA Enterobacterales RhizobialesRhodospirillales Alteromonadales Nitrosomonadales Supplementary Figure 3. 7. (continued) Myxococcales Pseudomonadales E. xenB Rhizobiales Pseudomonadales Enterobacterales Burkholderiales F. xplA Williamsia Cellvibrionales Thiotrichales Aeromonadales Hydrogenophilales Verrucomicrobiales Pseudonocardiales Vibrionales Immundisolibacterales Neisseriales Desulfovibrionales Flavobacteriales Pirellulales Bacteriovoracales Rhodocyclales Isosphaerales Nitrosomonadales Ferrovales Chthoniobacterales Aeromonadales Micropepsales Myxococcales Desulfuromonadales Holosporales Xanthomonadales Rhodocyclales Oceanospirillales Chromatiales Caulobacterales Sphingomonadales Other Burkholderiales Caulobacterales Rhodospirillales Oceanospirillales Thiotrichales Methylococcales Chromatiales Opitutales Bdellovibrionales Rhodobacterales Acidobacteriales Verrucomicrobiales Micrococcales Other Gordonia Microbacterium Mycobacterium Rhodococcus Mycolicibacterium 135 Supplementary Figure 3. 8. Co-occurrence network based on spearman correlation (rho > 0.85 and p- value < 0.01) of main genera found in all samples from post-biostimulated wells. Only genus with an average abundance > 0.1% and present in at least 50% of samples were considered. Node size indicates the relative abundance (0.1% ~ 5.46%). The network was process with Modularity function of Gephi to group nodes colored into 7 different modules with default setting and a resolution of 0.85. 136 A B B B Supplementary Figure 3. 9. Comparison of degradation pathway (A) and functions (B) in category of Xenobiotics biodegradation and metabolism between pre- and post biostimulation wells. For degradation pathway analysis, default options were used for the two groups comparison (p < 0.05, Welch's two sided t-test). For functions analysis, an extra filter was added as difference in mean proportions > 1%. 137 A Supplementary Figure 3. 10. Heatmap of groundwater and Red Cedar River (RC) log10 gene copies per milliliter (A) and sediment log10 gene copies per gram (B). Grey cells indicate either no amplification or false positive amplification. In the sample name, post and pre refer to the post- and pre-biostimulation samples, J_ refer to the samples from previous work (24). 138 Supplementary Figure 3. 10. (continued) B 139 REFERENCES 140 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES EPA. 2017. Technical Fact Sheet – Hexahydro-1,3,5-trinitro1,3,5-triazine (RDX). https://www.epa.gov/sites/production/files/2017-10/documents/ffrro_ecfactsheet_rdx_9-15- 17_508.pdf. EPA. 2020. Search Superfund Site Information. https://cumulis.epa.gov/supercpad/cursites/srchsites.cfm. Agency for Toxic Substances and Disease Registry (ATSDR). 2012. Toxicological Profile for RDX. Felt D, Johnson JL, Larson S, Hubbard B, Henry K, Nestler C, Ballard JH. 2013 Evaluation of Treatment Technologies for Wastewater from Insensitive Munitions Production The US Army Engineer Research and Development Center, Wani AH, O'Neal BR, Davis JL, Hansen LD. 2002. Treatability Study for Biologically Active Zone Enhancement (BAZE) for In Situ RDX Degradation in Groundwater. U.S. Army Engineer Research and Development Center, Vicksburg, Mississippi. EPA. 2015. Third Five-Year Review For Cornhusker Army Ammunition Plant Grand Island, Nebraska Environmental Protection Agency Region 7, Lenexa, Kansas. Arcadis. 2004. In-situ enhanced anaerobic explosive treatment field test report, Milan Army Ammunition Plant Milan, Tennessee. Arcadis, Knoxville, Tennessee. Michalsen MM, King AS, Rule RA, Fuller ME, Hatzinger PB, Condee CW, Crocker FH, Indest KJ, Jung CM, Istok JD. 2016. Evaluation of Biostimulation and Bioaugmentation To Stimulate Hexahydro-1,3,5-trinitro-1,3,5,-triazine Degradation in an Aerobic Groundwater Aquifer. Environ Sci Technol 50:7625-32. Michalsen MM, King AS, Istok JD, Crocker FH, Fuller ME, Kucharzyk KH, Gander MJ. 2020. Spatially-distinct redox conditions and degradation rates following field-scale bioaugmentation for RDX-contaminated groundwater remediation. Journal of Hazardous Materials 387:121529. Crocker FH, Indest KJ, Jung CM, Hancock DE, Fuller ME, Hatzinger PB, Vainberg S, Istok JD, Wilson E, Michalsen MM. 2015. Evaluation of microbial transport during aerobic bioaugmentation of an RDX-contaminated aquifer. Biodegradation 26:443-51. 11. Michalsen MM, Weiss R, King A, Gent D, Medina VF, Istok JD. 2013. Push-Pull Tests for Estimating RDX and TNT Degradation Rates in Groundwater. Groundwater Monitoring & Remediation 33:61-68. 12. 13. Kitts CL, Green CE, Otley RA, Alvarez MA, Unkefer PJ. 2000. Type I nitroreductases in soil Enterobacteria reduce TNT (2,4,6,-trinitrotoluene) and RDX (hexahydro-1,3,5-trinitro-1,3,5- triazine). Can J Microbiol 46:278-82. Peterson FJ, Mason RP, Hovsepian J, Holtzman JL. 1979. Oxygen-sensitive and -insensitive nitroreduction by Escherichia coli and rat hepatic microsomes. J Biol Chem 254:4009-14. 141 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Li WZ, Yin P, Yang YX. 1987. [Properties of TNT-degrading enzymes in intact cells of Citrobacter freundii]. Wei Sheng Wu Xue Bao 27:257-63. Caballero A, Lazaro JJ, Ramos JL, Esteve-Nunez A. 2005. PnrA, a new nitroreductase-family enzyme in the TNT-degrading strain Pseudomonas putida JLR11. Environ Microbiol 7:1211-9. Indest KJ, Crocker FH, Athow R. 2007. A TaqMan polymerase chain reaction method for monitoring RDX-degrading bacteria based on the xplA functional gene. J Microbiol Methods 68:267-74. Bernstein A, Adar E, Nejidat A, Ronen Z. 2011. Isolation and characterization of RDX-degrading Rhodococcus species from a contaminated aquifer. Biodegradation 22:997-1005. Seth-Smith HM, Rosser SJ, Basran A, Travis ER, Dabbs ER, Nicklin S, Bruce NC. 2002. Cloning, sequencing, and characterization of the hexahydro-1,3,5-Trinitro-1,3,5-triazine degradation gene cluster from Rhodococcus rhodochrous. Appl Environ Microbiol 68:4764-71. Andeer PF, Stahl DA, Bruce NC, Strand SE. 2009. Lateral transfer of genes for hexahydro-1,3,5- trinitro-1,3,5-triazine (RDX) degradation. Appl Environ Microbiol 75:3258-62. Bhushan B, Halasz A, Spain JC, Hawari J. 2002. Diaphorase catalyzed biotransformation of RDX via N-denitration mechanism. Biochem Biophys Res Commun 296:779-84. Chakraborty S, Sakka M, Kimura T, Sakka K. 2008. Cloning and expression of a Clostridium kluyveri gene responsible for diaphorase activity. Biosci Biotechnol Biochem 72:735-41. Fuller ME, McClay K, Hawari J, Paquet L, Malone TE, Fox BG, Steffan RJ. 2009. Transformation of RDX and other energetic compounds by xenobiotic reductases XenA and XenB. Appl Microbiol Biotechnol 84:535-44. Fuller ME, McClay K, Higham M, Hatzinger PB, Steffan RJ. 2010. Hexahydro-1,3,5-trinitro- 1,3,5-triazine (RDX) Bioremediation in Groundwater: Are Known RDX-Degrading Bacteria the Dominant Players? Bioremediation Journal 14:121-134. Collier JM, Chai B, Cole JR, Michalsen MM, Cupples AM. 2019. High throughput quantification of the functional genes associated with RDX biodegradation using the SmartChip real-time PCR system. Appl Microbiol Biotechnol 103:7161-7175. Li RW, Giarrizzo JG, Wu S, Li W, Duringer JM, Craig AM. 2014. Metagenomic Insights into the RDX-Degrading Potential of the Ovine Rumen Microbiome. PLOS ONE 9:e110505. 26. Meyer F, Paarmann D, D'Souza M, Olson R, Glass EM, Kubal M, Paczian T, Rodriguez A, Stevens R, Wilke A, Wilkening J, Edwards RA. 2008. The metagenomics RAST server - a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinformatics 9:386. 27. 28. Cox MP, Peterson DA, Biggs PJ. 2010. SolexaQA: At-a-glance quality assessment of Illumina second-generation sequencing data. BMC Bioinformatics 11:485. Rho M, Tang H, Ye Y. 2010. FragGeneScan: predicting genes in short and error-prone reads. Nucleic Acids Res 38:e191. 142 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. Pruitt KD, Tatusova T, Maglott DR. 2005. NCBI Reference Sequence (RefSeq): a curated non- redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res 33:D501-4. Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M. 2016. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res 44:D457-62. Lee BU, Choi MS, Kim DM, Oh KH. 2017. Genome Shuffling of Stenotrophomonas maltophilia OK-5 for Improving the Degradation of Explosive RDX (Hexahydro-1,3,5-trinitro-1,3,5-triazine). Curr Microbiol 74:268-276. Fish JA, Chai B, Wang Q, Sun Y, Brown CT, Tiedje JM, Cole JR. 2013. FunGene: the functional gene pipeline and repository. Front Microbiol 4:291. Buchfink B, Xie C, Huson DH. 2015. Fast and sensitive protein alignment using DIAMOND. Nature Methods 12:59-60. Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30:2114-20. Papadopoulos JS, Agarwala R. 2007. COBALT: constraint-based alignment tool for multiple protein sequences. Bioinformatics 23:1073-9. Katoh K, Rozewicki J, Yamada KD. 2019. MAFFT online service: multiple sequence alignment, interactive sequence choice and visualization. Brief Bioinform 20:1160-1166. Letunic I, Bork P. 2019. Interactive Tree Of Life (iTOL) v4: recent updates and new developments. Nucleic Acids Res 47:W256-W259. Sherrill-Mix S. 2019. taxonomizr: Functions to Work with NCBI Accessions and Taxonomy. , vR package version 0.5.3. . https://CRAN.R-project.org/package=taxonomizr. RStudio Team. 2020. RStudio: Integrated Development for R., RStudio, PBC. Boston, MA, http://www.rstudio.com/. R Core Team. 2018. R: A language and environment for statistical computing, R Foundation for Statistical Computing, Vienna, Austria. , https://www.R-project.org/. Harrell FE. 2020. Hmisc: Harrell Miscellaneous, https://cran.r- project.org/web/packages/Hmisc/index.html. Bates D, Maechler M, Davis TA, Oehlschlägel J, Riedy J. 2019. Matrix: Sparse and Dense Matrix Classes and Methods, https://cran.r-project.org/web/packages/Matrix/index.html. Csardi G, Nepusz T. 2006. The igraph software package for complex network research. InterJournal, Complex Systems 1695. Bastian M, Heymann S, Jacomy M. 2006. Gephi: an open source software for exploring and manipulating networks. International AAAI Conference on Weblogs and Social Media. Aronesty E. 2013. Comparison of Sequencing Utility Programs. The Open Bioinformatics Journal 7:1-8. 143 46. 47. 48. 49. Huson DH, Beier S, Flade I, Gorska A, El-Hadidi M, Mitra S, Ruscheweyh HJ, Tappu R. 2016. MEGAN Community Edition - Interactive Exploration and Analysis of Large-Scale Microbiome Sequencing Data. PLoS Comput Biol 12:e1004957. Parks DH, Tyson GW, Hugenholtz P, Beiko RG. 2014. STAMP: statistical analysis of taxonomic and functional profiles. Bioinformatics 30:3123-4. Ritalahti KM, Amos BK, Sung Y, Wu Q, Koenigsberg SS, Loffler FE. 2006. Quantitative PCR targeting 16S rRNA and reductive dehalogenase genes simultaneously monitors multiple Dehalococcoides strains. Appl Environ Microbiol 72:2765-74. Stedtfeld RD, Williams MR, Fakher U, Johnson TA, Stedtfeld TM, Wang F, Khalife WT, Hughes M, Etchebarne BE, Tiedje JM, Hashsham SA. 2016. Antimicrobial resistance dashboard application for mapping environmental occurrence and resistant pathogens. FEMS Microbiol Ecol 92. 50. Wickham H. 2016. ggplot2: Elegant Graphics for Data Analysis., https://ggplot2.tidyverse.org. 51. Kahng H-Y, Lee B-U, Cho Y-S, Oh K-H. 2007. Purification and characterization of the NAD(P)H-nitroreductase for the catabolism of 2,4,6-trinitrotoluene (TNT) in Pseudomonas sp. HK-6. Biotechnology and Bioprocess Engineering 12:433. 52. 53. 54. 55. 56. 57. Rylott EL, Jackson RG, Sabbadin F, Seth-Smith HMB, Edwards J, Chong CS, Strand SE, Grogan G, Bruce NC. 2011. The explosive-degrading cytochrome P450 XplA: Biochemistry, structural features and prospects for bioremediation. Biochimica Et Biophysica Acta-Proteins and Proteomics 1814:230-236. Coleman NV, Nelson DR, Duxbury T. 1998. Aerobic biodegradation of hexahydro-1,3,5-trinitro- 1,3,5-triazine (RDX) as a nitrogen source by a Rhodococcus sp., strain DN22. Soil Biology & Biochemistry 30:1159-1167. Nejidat A, Kafka L, Tekoah Y, Ronen Z. 2008. Effect of organic and inorganic nitrogenous compounds on RDX degradation and cytochrome P-450 expression in Rhodococcus strain YH1. Biodegradation 19:313-320. Seth-Smith HMB, Edwards J, Rosser SJ, Rathbone DA, Bruce NC. 2008. The explosive- degrading cytochrome P450 system is highly conserved among strains of Rhodococcus spp. Applied and Environmental Microbiology 74:4550-4552. Thompson KT, Crocker FH, Fredrickson HL. 2005. Mineralization of the cyclic nitramine explosive hexahydro-1,3,5-trinitro-1,3,5-triazine by Gordonia and Williamsia spp. Applied and Environmental Microbiology 71:8265-8272. Fuller ME, Perreault N, Hawari J. 2010. Microaerophilic degradation of hexahydro-1,3,5-trinitro- 1,3,5-triazine (RDX) by three Rhodococcus strains. Letters in Applied Microbiology 51:313-318. 58. Wilson FP, Cupples AM. 2016. Microbial community characterization and functional gene quantification in RDX-degrading microcosms derived from sediment and groundwater at two naval sites. Applied Microbiology and Biotechnology 100:7297-7309. 144 59. 60. 61. 62. Bryant C, DeLuca M. 1991. Purification and characterization of an oxygen-insensitive NAD(P)H nitroreductase from Enterobacter cloacae. J Biol Chem 266:4119-25. Bryant C, Hubbard L, McElroy WD. 1991. Cloning, nucleotide sequence, and expression of the nitroreductase gene from Enterobacter cloacae. J Biol Chem 266:4126-30. Kwon MJ, O'Loughlin EJ, Antonopoulos DA, Finneran KT. 2011. Geochemical and microbiological processes contributing to the transformation of hexahydro-1,3,5-trinitro-1,3,5- triazine (RDX) in contaminated aquifer material. Chemosphere 84:1223-30. Livermore JA, Jin YO, Arnseth RW, Lepuil M, Mattes TE. 2013. Microbial community dynamics during acetate biostimulation of RDX-contaminated groundwater. Environ Sci Technol 47:7672- 8. 63. Wang D, Boukhalfa H, Marina O, Ware DS, Goering TJ, Sun F, Daligault HE, Lo CC, Vuyisich M, Starkenburg SR. 2017. Biostimulation and microbial community profiling reveal insights on RDX transformation in groundwater. Microbiologyopen 6. 64. 65. 66. 67. 68. 69. 70. 71. 72. Jugnia L-B, Manno D, Dodard S, Greer CW, Hendry M. 2019. Manipulating redox conditions to enhance in situ bioremediation of RDX in groundwater at a contaminated site. Science of The Total Environment 676:368-377. Adrian NR, Arnett CM. 2004. Anaerobic biodegradation of hexahydro-1,3,5-trinitro-1,3,5- triazine (RDX) by Acetobacterium malicum strain HAAP-1 isolated from a methanogenic mixed culture. Curr Microbiol 48:332-40. Sherburne LA, Shrout JD, Alvarez PJ. 2005. Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) degradation by Acetobacterium paludosum. Biodegradation 16:539-47. Bhatt M, Zhao JS, Halasz A, Hawari J. 2006. Biodegradation of hexahydro-1,3,5-trinitro-1,3,5- triazine by novel fungi isolated from unexploded ordnance contaminated marine sediment. J Ind Microbiol Biotechnol 33:850-8. Kwon MJ, Finneran KT. 2008. Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) and Octahydro- 1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) Biodegradation Kinetics Amongst Several Fe(III)- Reducing Genera. Soil and Sediment Contamination: An International Journal 17:189-203. Eaton HL, Duringer JM, Murty LD, Craig AM. 2013. Anaerobic bioremediation of RDX by ovine whole rumen fluid and pure culture isolates. Appl Microbiol Biotechnol 97:3699-710. Singh R, Soni P, Kumar P, Purohit S, Singh A. 2009. Biodegradation of high explosive production effluent containing RDX and HMX by denitrifying bacteria. World Journal of Microbiology and Biotechnology 25:269-275. Lee S-Y, Brodman BW. 2004. Biodegradation of 1,3,5-Trinitro-1,3,5-triazine (RDX). Journal of Environmental Science and Health, Part A 39:61-75. Kitts CL, Cunningham DP, Unkefer PJ. 1994. Isolation of three hexahydro-1,3,5-trinitro-1,3,5- triazine-degrading species of the family Enterobacteriaceae from nitramine explosive- contaminated soil. Applied and Environmental Microbiology 60:4608. 145 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. Zhang C, Hughes JB. 2003. Biodegradation pathways of hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) by Clostridium acetobutylicum cell-free extract. Chemosphere 50:665-671. Regan KM, Crawford RL. 1994. Characterization of Clostridium bifermentans and its biotransformation of 2,4,6-trinitrotoluene (TNT) and 1,3,5-triaza-1,3,5-trinitrocyclohexane (RDX). Biotechnology Letters 16:1081-1086. Zhao J-S, Paquet L, Halasz A, Manno D, Hawari J. 2004. Metabolism of octahydro-1,3,5,7- tetranitro-1,3,5,7-tetrazocine by Clostridium bifermentans strain HAW-1 and several other H2- producing fermentative anaerobic bacteria. FEMS Microbiology Letters 237:65-72. Kwon MJ, Wei N, Millerick K, Popovic J, Finneran K. 2014. Clostridium geopurificans Strain MJ1 sp. nov., A Strictly Anaerobic Bacterium that Grows via Fermentation and Reduces the Cyclic Nitramine Explosive Hexahydro-1,3,5-Trinitro-1,3,5-Triazine (RDX). Current Microbiology 68:743-750. Bhushan B, Halasz A, Thiboutot S, Ampleman G, Hawari J. 2004. Chemotaxis-mediated biodegradation of cyclic nitramine explosives RDX, HMX, and CL-20 by Clostridium sp. EDB2. Biochemical and Biophysical Research Communications 316:816-821. Zhao J-S, Spain J, Hawari J. 2003. Phylogenetic and metabolic diversity of hexahydro-1,3,5- trinitro-1,3,5-triazine (RDX)-transforming bacteria in strictly anaerobic mixed cultures enriched on RDX as nitrogen source. FEMS Microbiology Ecology 46:189-196. Zhao J-S, Spain J, Thiboutot S, Ampleman G, Greer C, Hawari J. 2004. Phylogeny of cyclic nitramine-degrading psychrophilic bacteria in marine sediment and their potential role in the natural attenuation of explosives. FEMS Microbiology Ecology 49:349-357. Arnett CM, Adrian NR. 2009. Cosubstrate independent mineralization of hexahydro-1,3,5- trinitro-1,3,5-triazine (RDX) by a Desulfovibrio species under anaerobic conditions. Biodegradation 20:15-26. Boopathy R, Gurgas M, Ullian J, Manning JF. 1998. Metabolism of Explosive Compounds by Sulfate-Reducing Bacteria. Current Microbiology 37:127-131. Ronen Z, Yanovich Y, Goldin R, Adar E. 2008. Metabolism of the explosive hexahydro-1,3,5- trinitro-1,3,5-triazine (RDX) in a contaminated vadose zone. Chemosphere 73:1492-1498. Bhatt M, Zhao J-S, Monteil-Rivera F, Hawari J. 2005. Biodegradation of cyclic nitramines by tropical marine sediment bacteria. Journal of Industrial Microbiology and Biotechnology 32:261- 267. Zhao J-S, Halasz A, Paquet L, Beaulieu C, Hawari J. 2002. Biodegradation of Hexahydro-1,3,5- Trinitro-1,3,5-Triazine and Its Mononitroso Derivative Hexahydro-1-Nitroso-3,5-Dinitro-1,3,5- Triazine by Klebsiella pneumoniae Strain SCZ-1 Isolated from an Anaerobic Sludge. Applied and Environmental Microbiology 68:5336. Van Aken B, Yoon JM, Schnoor JL. 2004. Biodegradation of Nitro-Substituted Explosives 2,4,6- Trinitrotoluene, Hexahydro-1,3,5-Trinitro-1,3,5-Triazine, and Octahydro-1,3,5,7-Tetranitro-1,3,5- Tetrazocine by a Phytosymbiotic Methylobacterium sp. Associated with Poplar Tissues (Populus deltoides x nigra DN34). Applied and Environmental Microbiology 70:508. 146 86. 87. 88. 89. 90. 91. 92. 93. Fournier D, Trott S, Hawari J, Spain J. 2005. Metabolism of the Aliphatic Nitramine 4-Nitro-2,4- Diazabutanal by Methylobacterium sp. Strain JS178. Applied and Environmental Microbiology 71:4199. Fournier D, Halasz A, Spain J, Spanggord RJ, Bottaro JC, Hawari J. 2004. Biodegradation of the Hexahydro-1,3,5-Trinitro-1,3,5-Triazine Ring Cleavage Product 4-Nitro-2,4-Diazabutanal by Phanerochaete chrysosporium. Applied and Environmental Microbiology 70:1123. Chang HW, Kahng HY, Kim SI, Chun JW, Oh KH. 2004. Characterization of Pseudomonas sp. HK-6 cells responding to explosive RDX (hexahydro-1,3,5-trinitro-1,3,5-triazine). Applied Microbiology and Biotechnology 65:323-329. Coleman NV, Spain JC, Duxbury T. 2002. Evidence that RDX biodegradation by Rhodococcus strain DN22 is plasmid-borne and involves a cytochrome p-450. Journal of Applied Microbiology 93:463-472. Young DM, Unkefer PJ, Ogden KL. 1997. Biotransformation of hexahydro-1,3,5-trinitro-1,3,5- triazine (RDX) by a prospective consortium and its most effective isolate Serratia marcescens. Biotechnology and Bioengineering 53:515-522. Zhao J-S, Manno D, Leggiadro C, apos, Neil D, Hawari J. 2006. Shewanella halifaxensis sp. nov., a novel obligately respiratory and denitrifying psychrophile. International Journal of Systematic and Evolutionary Microbiology 56:205-212. Zhao J-S, Manno D, Beaulieu C, Paquet L, Hawari J. 2005. Shewanella sediminis sp. nov., a novel Na+-requiring and hexahydro-1,3,5-trinitro-1,3,5-triazine-degrading bacterium from marine sediment. International Journal of Systematic and Evolutionary Microbiology 55:1511-1520. Binks PR, Nicklin S, Bruce NC. 1995. Degradation of hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) by Stenotrophomonas maltophilia PB1. Applied and Environmental Microbiology 61:1318. 94. Mahan KM, Zheng H, Fida TT, Parry RJ, Graham DE, Spain JC. 2017. Iron-Dependent Enzyme Catalyzes the Initial Step in Biodegradation of Nitroglycine by Variovorax sp. Strain JS1663. Applied and Environmental Microbiology 83:e00457-17. Cho K-C, Lee DG, Fuller ME, Hatzinger PB, Condee CW, Chu K-H. 2015. Application of 13C and 15N stable isotope probing to characterize RDX degrading microbial communities under different electron-accepting conditions. Journal of Hazardous Materials 297:42-51. Bhushan B, Halasz A, Spain J, Thiboutot S, Ampleman G, Hawari J. 2002. Biotransformation of Hexahydro-1,3,5-trinitro-1,3,5-triazine Catalyzed by a NAD(P)H:  Nitrate Oxidoreductase from Aspergillus niger. Environmental Science & Technology 36:3104-3108. 95. 96. 147 CHAPTER 4 Identification of the Phylotypes and Predicted Functional Genes Involved in cis- Dichloroethene and 1,4-Dioxane Aerobic Biodegradation in Soil Microcosms This chapter is being prepared for submission to a peer reviewed journal: Hongyu Dang and Alison M. Cupples. Identification of the Phylotypes and Predicted Functional Genes Involved in cis-Dichloroethene and 1,4-Dioxane Aerobic Biodegradation in Soil Microcosms. Abstract Co-contamination with chlorinated compounds and 1,4-dioxane has been reported at many sites. Recently, there has been an increased interest in aerobic bioremediation because of the potential to degrade multiple contaminants concurrently. However, the likelihood of implementing a successful bioremediation approach is dependent on the microorganisms present. Towards improving bioremediation efficacy, the current study examined laboratory microcosms (inoculated separately with two soils) to determine the phylotypes and functional genes associated with the biodegradation of two common co-contaminants (cis-dichloroethene [cDCE] and 1,4-dioxane). The impact of amending microcosms with lactate on cDCE and 1,4-dioxane biodegradation was also investigated. In one soil, when all three substrates were present (1,4- dioxane, cDCE, lactate), 1,4-dioxane removal was slower compared to when only one additional substrate was added (cDCE or lactate). In contrast, in microcosms amended with another soil, 1,4-dioxane removal trends for all three treatments were similar, indicating the present of either lactate or cDCE or both did not impact 1,4-dioxane biodegradation. Lactate appeared to improve the biological removal of cDCE in microcosms inoculated with either soil. Stable isotope probing (SIP) was then used to determine which phylotypes were actively involved in carbon uptake from cDCE and 1,4-dioxane in both soil communities. The most enriched phylotypes for 13C assimilation from 1,4-dioxane included Rhodopseudomonas and Rhodanobacter. Propane 148 monooxygenase was predicted (by PICRUSt2) to be dominant in the 1,4-dioxane amended microbial communities and propane monoxygenase gene abundance values correlated with other enriched (but less abundant) phylotypes for 13C-1,4-dioxane assimilation. The dominant enriched phylotypes for 13C assimilation from cDCE included Bacteriovorax, Pseudomonas and Sphingomonas. In the cDCE amended soil microcosms, PICRUSt2 predicted the presence of DNA encoding glutathione S-transferase (a known cDCE upregulated enzyme). Overall, the work demonstrated concurrent removal of cDCE and 1,4-dioxane by indigenous soil microbial communities and the enhancement of cDCE removal by lactate. The data generated on the phylotypes responsible for carbon uptake (as determined by SIP) could be incorporated into diagnostic molecular methods for site characterization. The results suggest aerobic concurrent biodegradation of cDCE and 1,4-dioxane should be considered for chlorinated solvent site remediation. 1. Introduction The clean-up of sites with mixed contamination poses a significant challenge to the remediation community. Developing synergistic approaches that could reduce the concentrations of multiple contaminants has the potential to result in considerable cost savings. From the list of co- contaminants found in soil and groundwater, the chlorinated solvents and their metabolites (tetrachloroethene [PCE], trichloroethene [TCE], cis-dichloroethene [cDCE], vinyl chloride [VC]) are particularly prevalent (found at > 3,000 Department of Defense sites) and problematic due to their tendency to form large, dissolved-phase plumes, their recalcitrant nature and the subsequent risk to human health. Remediation efforts have frequently involved biostimulation, through the addition of carbon sources, or bioaugmentation, which involves in the injection of mixed microbial cultures containing Dehalococcoides mccartyi (1). With the expansion of this 149 remedial practice over the last decade, the number of sites in the US now numbers well over 2,300, and bioaugmentation has been performed in at least 11 other countries (P Hatzinger, personal communication). Although clearly a very successful approach, bioremediation with D. mccartyi involves several significant limitations and thus may not be appropriate for all chlorinated solvent contaminated sites. Specifically, it is unlikely to be employed at large oxic sites because of the requirement for highly reducing conditions for D. mccartyi and the associated cost of driving such large sites anaerobic. Secondly, the approach will be less desirable at sites with multiple contaminants if those co-contaminants can be degraded more easily under aerobic conditions (e.g. benzene, toluene, 1,4-dioxane). Further, the accumulation of the known human carcinogen, VC, from the dechlorination process represents a significant risk if complete dechlorination does not occur. Additionally, driving sites anaerobic can result in long-term secondary groundwater impacts such as hydrogen sulfide formation, acidification, mobilization of reduced metals and methane accumulation. In contrast, aerobic approaches have the advantage that the geochemistry of the site is not significantly impacted. cDCE is a major degradation product of TCE by both abiotic and biotic degradation (2). For example, at a TCE contaminated site (Dover Air Force Base, DE) dechlorination by the indigenous microbial community only transformed TCE to cDCE (3). In laboratory batch and column tests for enhanced biological dissolution of PCE, cDCE was the main product of PCE dehalogenation and accumulated when PCE and TCE were present at high concentrations (4, 5). In fact, “cDCE stall” is a well-recognized term in the remediation community for the 150 accumulation of cDCE at chlorinated solvent sites. Given the common occurrence of cDCE, identifying the potential for cDCE transformation remains an important issue. 1,4-Dioxane, a probable human carcinogen and common chlorinated solvent stabilizer, has been found at numerous contaminated sites across the U.S. (6, 7). In an examination from 49 remediation installations at U.S. Air Force sites, 1,4-dioxane was detected in 781 groundwater wells, and 64% of wells that contained 1,4-D also contained TCE (8). In an evaluation of >2000 sites in California, the chlorinated solvents were found in 94% of the sites with detections of 1,4- dioxane (9). Many bacteria have been identified to metabolically or co-metabolically degrade 1,4- dioxane under aerobic conditions (10, 11). However, aerobic 1,4-dioxane degradation can be impacted by the presence of chlorinated compounds. For Pseudonocardia dioxanivorans strain CB1190, researchers reported 1,1,1-trichloroethane (1,1,1-TCA) and 1,1-dichloroethene (1,1- DCE) illustrated similar inhibitory effects on 1,4-dioxane degradation (12). In the same study, 1,1-DCE was a slightly more potent inhibitor for 1,4-dioxane degradation than 1,1,1-TCA for Pseudomonas mendocina strain KR1, while 1,1,1-TCA was a much more potent inhibitor for 1,4-dioxane degradation than 1,1-DCE for a Escherichia coli recombinant strain expressing toluene-4-monooxygenase from strain KR1 (12). A later study with P. dioxanivorans CB1190 indicated chlorinated compounds inhibited 1,4-dioxane biodegradation in the following order: 1,1-DCE > cDCE > TCE > 1,1,1-TCA (13). To address the problem of co-contamination, efforts have focused specifically on the removal of both chlorinated compounds and 1,4-dioxane. To degrade TCE, DCEs with 1,4- dioxane, P. dioxanivorans CB1190 was combined with hydrogen peroxide and tungstated zirconia, which partially removed those contaminants with the remainder being degraded by P. 151 dioxanivorans CB1190 (14, 15). In another study, P. dioxanivorans CB1190 was combined with the anaerobic bioaugmentation culture KB-1 (a chloroethene degrading consortium) resulting in TCE transformation to cDCE, as well as cDCE and 1,4-dioxane degradation by P. dioxanivorans CB1190 (16). Another strain Azoarcus sp. DD4, was found to degrade 1,4-dioxane with 1,1-DCE using propane as the main substrate (17). More recently, Azoarcus sp. DD4 was sequentially used with SDC-9 (another chloroethene degrading consortium) to achieve transformation of TCE to cDCE and VC by SDC-9 and co-metabolic removal of VC, cDCE and 1,4-dioxane by Azoarcus sp. DD4 with the addition of propane (18). These studies suggest mixed microbial communities will likely be needed to facilitate co-contamination remediation. An interesting question arising from these trends concerns the biodegrading abilities of indigenous mixed communities and their potential contribution to site remediation. Towards understanding the potential of natural mixed communities, the current work builds on previous research documenting aerobic 1,4-dioxane biodegradation in soil microcosms (19). In the current work, stable isotope probing (SIP) is utilized to identify which microorganisms are involved in carbon uptake from cDCE and 1,4-dioxane. SIP is a cultivation independent method to link identity with function (20) such as contaminant biodegradation (21- 25). As aerobic contaminant biodegradation often relies on co-metabolism, the impact of an additional substrate (lactate) was also investigated. Lactate was selected because it is commonly used in biostimulation (to drive sites anaerobic) (3, 26) and would therefore already be acceptable to many regulatory agencies. The objectives were to 1) examine removal rates of the co-contaminants cDCE and 1,4-dioxane, with and without lactate addition, with indigenous mixed microbial communities 2) identify the microorganisms responsible for the uptake of 13C from cDCE as well as from 1,4-dioxane during biodegradation and 3) predict the functional 152 genes present and correlate their presence to specific phylotypes. The overall rationale behind the current project is to provide knowledge to enhance the aerobic remediation of two important groundwater contaminants (cDCE, 1,4-dioxane) for oxic sites. 2. Methods 2.1 Chemicals and Soil Inocula Unlabeled 1,4-dioxane (99.8%) and cDCE were purchased from Sigma-Aldrich (MO, USA). Labeled 1,4-dioxane [(13C)4H8O2] was purchased from Santa Cruz Biotechnology (TX, USA) with 99.2% isotopic purity and 98% purity, and labeled cDCE [13C2H2Cl2] was purchased from Sigma-Aldrich (MO, USA) with 99% isotopic purity and 97% purity. Two soils were collected from 5 sampling stations in 6 replicate plots within Treatments 1 and 2 at the Michigan State University (MSU) Main Cropping System Experiment at Kellogg Biological Station Long-Term Ecological Research (KBS LTER) (42°24′N, 85°23′W). Both soils received conventional levels of chemical inputs, however, Treatment 1 is chisel plowed and Treatment 2 is under no-till management. For additional information see https://lter-kbs-msu- edu.proxy1.cl.msu.edu/research/site-description-and-maps/. All samples for each treatment were mixed, then stored at 4 °C in the dark. These soils were selected because the analysis of shotgun sequencing data generated from a previous study (27) indicated the presence of numerous microorganisms previously associated with 1,4-dioxane and cDCE biodegradation (as discussed in the results section). 2.2 Microcosms Setup For each set of amendments, microcosms were established in 160 mL serum bottles (wrapped with aluminum foil) with 10 g of soil and 20 mL of media. For each soil, triplicate microcosms were amended with one of the following four sets of amendments: 1) cDCE, 1,4-dioxane and 153 lactate, 2) cDCE and 1,4-dioxane, without lactate, 3) 1,4-dioxane with lactate and 4) cDCE with lactate. All microcosms were closed with a rubber seal and aluminum crimp. For each soil, all four treatments included triplicate abiotic autoclaved controls. All bottles were incubated at room temperature (~21 °C) on a shaker (110 rpm). The media was based on that used to enrich the 1,4- dioxane degrader Pseudonocardia dioxanivorans CB1190 (except nitrilotriacetic acid was removed) (28). One liter of the media contained 100 mL of a buffer stock [K2HPO4 (32.4 g/L), KH2PO4 (10 g/L), NH4Cl (20 g/L)] and 100 mL of a trace metal stock [MgSO4.7H2O (2 g/L), FeSO4.7H2O (0.12 g/L), MnSO4.H2O (0.03 g/L), ZnSO4.7H2O (0.03 g/L), and CoCl2.6H2O (0.01 g/L)]. The initial liquid concentrations of 1,4-dioxane and cDCE were ~ 6 mg/L and ~ 4 mg/L. The liquid concentration of cDCE was calculated based on Henry’s law (29). The treatments with sodium lactate were amended at 0.56 g/L (~5 mM). Additional microcosms were established for each soil (160 mL bottles, 10 g soil 1 or 2, same media) for the SIP experiments. For each soil, triplicate abiotic control microcosms (sterilized by autoclaving) and six microcosms were amended with unlabeled 1,4-dioxane or cDCE (similar concentrations as above). Another six microcosms were amended with 13C labeled 1,4-dioxane or 13C labeled cDCE. As the above experiments indicated cDCE was improved by the addition of lactate, the cDCE bottles were also amended with 5 mM of lactate (and closed with a rubber seal and aluminum crimp). The 1,4-dioxane amended bottles were not amended with lactate and were opened for 6 hours every three days for aeration. 2.3 Analytic Methods Liquid samples (0.1 mL) were withdrawn (with sterilized disposable needles and a 1 mL syringe), then filtered (with a 0.22 µm, 4 mm nylon syringe filter, Thomas Scientific, NJ) for 1,4-dioxane analysis. The filtered samples were injected into a GC-FID (Hewlett Packard 5890) 154 equipped with a column (Restek, Stabilwax-DB, 30m, 0.53 mmID, 1µm) using a similar method to that previously described (30). The injector temperature was maintained at 220 °C and the detector temperature was set at 250 °C. The oven temperature was programmed to initiate at 80 °C for 1 min, then increased to 140 °C with a ramp of 20 °C /min. The gas phase concentration of cDCE was determined (1 mL gastight syringe, 0.2 mL of the gaseous sample) with a GC-FID (Hewlett Packard 5890) equipped with a capillary column (Alltech, AT-624, 30m × 0.53mm ID × 3.0µm) using a similar method described in a previous study (31). The injector temperature was maintained at 180 °C and the detector at 240 °C. The oven temperature was programmed to initiate at 45 °C for 4 min, then increased to 165 °C with a ramp of 20 °C /min, held at 165 °C for 1 min. 2.4 DNA extraction, Fractionation and Sequencing Duplicate soil 1 and soil 2 inoculated microcosms amended with either labeled or unlabeled chemicals (16 bottles, 2 chemicals, 2 with unlabeled amendment X 2 with labeled amendment X 2 soils) were sacrificed for DNA extraction at ~50% cDCE or 1,4-dioxane removal using QIAGEN PowerSoil DNA extraction kit as per manufacturer’s protocol. DNA extracts (approx. 10 μg) were loaded into Quick-Seal polyallomer tubes (13 by 51 mm, 5.1 mL; Beckman Coulter (Brea, CA) along with a Tris-EDTA (10 mM Tris, 1 mM EDTA, pH 8)-CsCl solution for ultracentrifugation. The density of the mixture inside the tube was determined with a model AR200 digital refractometer (Leica Microsystems Inc., Buffalo Grove, IL), and it was adjusted to a final value of 1.730 g/mL by adding small volumes of CsCl solution or TE buffer until the tube could be sealed. The sealed tubes were then ultracentrifuged at 178,000×g (20 °C) for 46 h in a StepSaver 70 V6 vertical titanium rotor (8 by 5.1 mL capacity) within a Sorvall WX 80 Ultra Series centrifuge (Thermo Scientific, Waltham, MA). Following ultracentrifugation, the 155 tubes were placed onto a fraction system (Beckman Coulter) and fractions (∼26, 200 μL) were collected. The buoyant density of each fraction was measured, and CsCl was removed by glycogen-assisted ethanol precipitation. The DNA concentration in each fraction was quantified using the Quant-iT™ dsDNA High-Sensitivity Assay Kit. For each chemical (labeled and unlabeled) and soil, triplicate DNA extracts of three fractions with higher buoyant density (1.73-1.75 g/mL) and one fraction with lighter buoyant density (~1.70 g/mL) were submitted for 16S rRNA gene amplicon sequencing at Research Technology Support Facility (RTSF) at MSU. The heavy buoyant density fractions were selected based on the minimum concentration of DNA required for sequencing. In total, two 96 well plates were submitted for sequencing (2 chemicals, 2 soils, 4 fractions, 3 replicate fractions, 2 microcosm replicates, 2 isotopes). The V4 hypervariable region of the 16S rRNA gene was amplified using dual indexed Illumina compatible primers 515f/806r as described by James Kozich (32). PCR products were batch normalized using Invitrogen SequalPrep DNA Normalization plates and the products recovered from the plates pooled. The pool was cleaned up and concentrated using AmpureXP magnetic beads; it was QC’d and quantified using a combination of Qubit dsDNA HS, Agilent 4200 TapeStation HS DNA1000 and Kapa Illumina Library Quantification qPCR assays. The pool was loaded onto an Illumina MiSeq v2 standard flow cell and sequencing was performed in a 2x250 bp paired end format using a MiSeq v2 500 cycle reagent cartridge. Custom sequencing and index primers were added to appropriate wells of the reagent cartridge. Base calling was performed by Illumina Real Time Analysis (RTA) v1.18.54 and output of RTA was demultiplexed and converted to FastQ format with Illumina Bcl2fastq v2.19.1. The sequencing data for 1,4-dioxane and cDCE SIP was submitted to NCBI 156 under Bioproject PRJNA719874 (accession numbers SAMN18623434 to SAMN18623529) and PRJNA719920 (accession numbers SAMN18624005 to SAMN18624100), respectively. 2.5 Analysis of Sequencing Data The amplicon sequencing data in the fastq file format was analyzed with Mothur (version 1.44.2) (33) using the MiSeq Standard Operating Procedure (32). The procedure included trimming the raw sequences and quality control. The database used for alignment was SILVA bacteria database (Release 138) for the V4 region (34). Chimeras, mitochondrial and chloroplast lineage sequences were removed, then the sequences were classified into OTUs. The downstream analysis was conducted using microbiome (35) (version 1.10.0), phyloseq (36) (version 1.32.0), ampvis2 (37) (version 2.6.5), ggplot2 (38) (version 3.3.2), Hmisc (39) (version 4.4-1), Matrix (40) (version 1.2-18), igraph (41) (version 1.2.6), ggpubr (42) (version 0.4.0) in R (43) (version 4.0.4) with R studio (44) (version 1.1.456). Additionally, the software Statistical Analysis of Taxonomic and Functional Profiles (STAMP) (45) (version 2.1.3) was utilized to statistically analyze the results. The package microbiome was used to combine the OTUs shared file, taxonomy and metadata and it was also used to transform the counts of reads for OTUs into relative abundance. Phyloseq and ggplot2 were used for the Non-metric Multi-dimensional Scaling (NMDS) plots, alpha diversity measurements plots, the bar plot for the classification of the microbial community at phylum level for different soils and treatments and for exporting a subset of OTUs information based on the variables in metadata. Ampvis2 was used to generate the rarefaction curves and heatmaps of average abundance at the genus level of the sample replicates. OTUs with at least 0.06% average relative abundance and 50% occurrence were selected for building the correlation network. The packages Hmisc and Matrix were used to calculate the correlation of OTUs with 157 Spearman correlation. Strong correlations (correlation coefficient ≥ 0.7) and Benjamini- Hochberg method adjusted p value (p < 0.01) were set to filter the correlation result. The filtered correlation result were used to build occurrence network with the package igraph and these were visualized in Gephi (46). The OTUs enriched in the heavy fractions of the 13C labeled cDCE or 1,4-dioxane amended microcosms were determined using STAMP. Specifically, OTU relative abundance was compared between the heavy fractions of the microcosms amended with the labeled chemical and the heavy fractions of the microcosms amended with the unlabeled chemical (Welch’s two-sided t-test, p < 0.05, with default settings). STAMP was also used to investigate which OTUs were enriched in the light fractions of the labeled amended microcosms compared to the light fractions of the unlabeled amended microcosms to eliminate false positives in the heavy fraction analysis. In addition, the enriched OTUs were subject to further statistical analysis in RStudio (Wilcoxon Rank Test, ggplot2 and ggpubr). 2.6 Function Prediction and Correlation The microbial functions from KEGG orthologs (KO) (47) were predicted from the sequencing data using the PICRUSt2 pipeline (48). The functions related to 1,4-dioxane and cDCE degradation identified in previous research were manually picked (10, 11, 49-53) to generate the heatmaps of relative abundance across all sample replicates with the R package ComplexHeatmap (54) (version 2.4.3). In addition, OTUs from 1,4-dioxane and cDCE SIP experiments with an average relative abundance ≥ 0.05% were collected and pooled together with functions associated with 1,4-dioxane and cDCE degradation for running Spearman correlations. OTUs correlated with at least 4 and 2 functions for 1,4-dioxane and cDCE 158 degradation, respectively, with absolute values of correlation coefficients higher than 0.6 were chosen for plotting the heatmaps with the same R package. 2.7 Species Associated with 1,4-dioxane and DCE Degradation Previously, our group submitted DNA extracts from the same soils (Treatments 1 and 2 from KBS) for shotgun sequencing (27). In the current work, the shotgun sequences (processed by Trimmomatic (55)) were assembled with Megahit (56) (version 1.2.4) using the pair end plus single end option. Minimum and maximum kmer sizes were 27 and 127 with the kmer size step of 10. Previously identified 1,4-dioxane and cDCE degraders were searched for in the National Center for Biotechnology Information (NCBI) taxonomy browser to find their lowest ranks (primarily rank of species) and taxonomy IDs (Supplementary Table 4.1). The assembled reads were then aligned to the NCBI nucleotide database (nt) with the taxids option in BLASTN (57) (version 2.10.0-Linux_x86_64). The results were restricted to evalue ≤ 1× 10-5 (with output format 6 ) and identity ≥ 60 % and the resulting files were then imported into Megan (58) (community edition version 6.19.7). Each BLASTN output was processed to map the Megan genomic DNA accession database for generating the phylogenetic trees of the species associated with 1,4-dioxane or DCE degradation. 3. Results 3.1 Degradation of 1,4-Dioxane and cDCE With or Without Lactate 1,4-Dioxane and cDCE concentrations were determined in microcosms with inocula from two soils, with or without the additional amendment(s) of lactate/1,4-dioxane/cDCE (Figures 1 and 2). For soil 1 inoculated microcosms, when all three substrates were added together, the differentiation between removal trends for 1,4-dioxane between the samples and abiotic controls was not strong, with 95% confidence intervals (CIs) for the regression lines overlapping the 159 entire course of the incubation (Figure 4.1A, part A). However, when only lactate was added, the 1,4-dioxane regression lines 95% CIs between the abiotic controls and samples separated before day 20 (Figure 4.1A, part B). Similarly, when only cDCE was added, the 1,4-dioxane regression lines 95% CIs between the samples and abiotic controls separated at ~ day 20 (Figure 4.1, part C). Based on these trends, for this particular microbial community, one hypothesis is that when all three substrates are present (Figure 4.1A, part A), 1,4-dioxane removal is slower compared to when only one additional substrate is added (Figure 4.1A, parts B and C). Compared to soil 1 microcosms, the impact of additional chemicals on 1,4-dioxane removal was different for soil 2 microcosms. That is, 1,4-dioxane biodegradation was similar for all three treatments, indicating the present of either lactate or cDCE or both did not impact 1,4-dioxane removal in this soil community (Figure 4.2A, parts A-C). The most notable trend for cDCE biodegradation in both soil 1 and soil 2 microcosms concerns the addition of lactate. In soil 1 microcosms, without the addition of lactate, cDCE removal was slower, as indicated by the slope of the regression line (0.015) (Figure 4.1B, part C) compared to regression line slopes (0.022 and 0.021) from the treatments with lactate (Figure 4.1B, parts A and B). In soil 2 microcosms, although the cDCE regression line slope (0.026) was greater when lactate was not added (Figure 4.2B, part C), the overlap between the regression lines 95% CIs remained until ~ day 40, compared to ~ day 10 for both of the lactate amended treatments (Figure 4.2C, parts A and B). The trends for both soils support the hypothesis that the presence of lactate improves the biological removal of cDCE. Based on these results, lactate was added to the cDCE SIP experiments but not to the 1,4-dioxane SIP experiments. For soil 1 microcosms, the cDCE regression line slopes were similar (0.022 vs. 0.021) when either lactate was added with 1,4-dioxane (Figure 4.1B, part A), or when only lactate was 160 added (Figure 4.1B, part B), suggesting 1,4-dioxane does not impact cDCE removal in this soil community. For soil 2 microcosms, the cDCE regression line slopes differed (0.022 vs. 0.017) when either lactate was added with 1,4-dioxane (Figure 4.2B, part A), or when lactate was added by itself (Figure 4.2B, part B), again suggesting 1,4-dioxane likely does not impact cDCE removal (when lactate is present). 161 A. 1,4-Dioxane B. cDCE Figure 4. 1. 1,4-Dioxane (A) and cDCE (B) concentrations in triplicate sample microcosms (purple [A] and blue [B]) and triplicate abiotic controls (red) inoculated with soil 1 and different amendments. The shaded areas indicate 95% confidence intervals along the linear regression model. 162 A. 1,4-Dioxane B. cDCE Figure 4. 2. 1,4-Dioxane (A) and cDCE (B) concentrations in triplicate sample microcosms (purple [A] and blue [B]) and triplicate abiotic controls (red) inoculated with soil 2 and different amendments. The shaded areas indicate 95% confidence intervals along the linear regression model. In microcosms amended with all three substrates, decreases in cDCE concentrations occurred earlier than decreases for 1,4-dioxane in both soils 1 and 2 (as shown by an earlier 163 separation of the regression line 95% CIs between the samples and controls, Figure 4.1A, B, part A, Figure 4.2A, B part A). The pattern suggests there is a sequential removal for cDCE and 1,4- dioxane with the addition of lactate. In comparison, when the microcosms were amended with only 1,4-dioxane and cDCE (Figure 4.1A, B, part C, Figure 4.2A, B part C) the trend was less clear. In soil 2 microcosms, 1,4-dioxane removal started before cDCE removal while in soil 1 microcosms, the removal for 1,4-dioxane and cDCE started at a similar time. The SIP experiments involved the addition of cDCE and 1,4-dioxane separately to microcosms inoculated with each soil (Supplementary Figure 4.1). Triplicate samples for each were sacrificed at 44 days for DNA extraction (~50% removal of 1,4-dioxane or cDCE). The concentration of cDCE in abiotic controls declined towards the end of the study, likely a result of gas phase leakage through the previously punctured rubber septa (Supplementary Figure 4.1 C and D). 3.2 Microbial Community Analysis The rarefaction curves of the SIP fractions plateaued, indicating the majority of the species were sequenced (Supplementary Figure 4.2). A larger number of species were found in microcosms amended with soil 2 from all fractions (Supplementary Figure 4.2). The NMDS analysis suggested the community composition was different between the light and heavy fractions in both the cDCE (Figure 4.3A) and 1,4-dioxane (Figure 4.3B) amended microcosms, indicating a successful fractionation process. While a clear separation between the two soils was visible in the fractions originating from the cDCE amendments (Figure 4.3A), the separation was less pronounced in the 1,4-dioxane fractions (Figure 4.3B), suggesting a greater similarity in the latter samples. Alpha diversity and richness indices (Figure 4.3C and D) indicated a greater distinction between the light and heavy fractions of the cDCE amended samples compared to 164 those amended with 1,4-dioxane. NMDS analysis also provided clear distinctions between the communities based on the amended substrate (cDCE or 1,4-dioxane) (Supplementary Figure 4.3). The diversity and richness indices were higher in the samples amended with cDCE compared to those amended with 1,4-dioxane (Supplementary Figure 4.3). A C B D Figure 4. 3. Non-metric Multi-dimensional Scaling (NMDS) plots (A, B) and alpha diversity measurements (C, D) for the cDCE (A, C) and 1,4-dioxane (B, D) SIP experiments with soil 1 and 2. Using the rarefied even depth of 95% of the minimum sum of OTU counts, 32 phyla were identified (Figure 4.4A). Major phyla included Proteobacteria and Actinobacteria, and these were more abundant in all heavy fractions compared to the light fractions, while Bacteroidetes was dominant in the light fractions. Gemmatimonadetes was more abundant in the heavy fractions only in the cDCE amended samples. Other major phyla included Chloroflexi, 165 Firmicutes and Verrucomicrobia. In many cases, a clear distinction is visible between the 12C and 13C amended fractions, indicating differences between communities, as discussed below. A B Figure 4. 4. Classification to the phylum level for both replicates and soils amended with 1,4-dioxane (upper plot) or cDCE (lower plot) with a rarefied even depth of 95% of the minimum sum of OTU counts, each column represents cumulative values for three fractions (A). The classification (family level) of the top 30 OTUs (across all samples) within the most dominant phylum (Proteobacteria) (B) without rarefaction. 166 As Proteobacteria represented the phylum with the greatest number of sequences, the most abundant families (top 30 OTUs) within this phylum were determined (Figure 4.4B). Those that illustrated a higher abundance in the cDCE amended samples included: Bacteriovoracaceae, Bdellovibrionaceae, CCD24_fa, Chromobacteriaceae, Deltaproteobacteria_unclassified, Pseudomonadaceae, Sphingomonadaceae and Steroidobacteraceae. In the 1,4-dioxane amended samples, Rhodanobacteraceae was more abundant and several samples illustrated a high abundance of Burkholderiaceae and Xanthobacteraceae. At the genus level for all phyla, the most abundant genera in the 1,4-dioxane amended samples classified as Rhodanobacter, Chujaibacter (both Proteobacteria) and an uncharacterized genus within Bacteroidetes (Supplementary Figure 4.4). While in the cDCE amended samples, the most abundant genera included unclassified Bacteria, Pseudomonas (Proteobacteria) and Gp6 (Actinobacteria) (Supplementary Figure 4.4). The current work also involved the analysis of shotgun sequencing data from the same two soils from a previous study (27). Here, multiple species previously associated with 1,4- dioxane and cDCE degradation were identified, including Pseudonocardia dioxanivorans (1,4- dioxane degrader) and Polaromonas sp. JS666 (cDCE degrader) (Supplementary Figure 4.5). 3.3 Phylotypes Responsible for 13C Uptake from cDCE and 1,4-Dioxane Phylotypes enriched in the heavy fractions of the 13C cDCE or 13C 1,4-dioxane amended samples compared to the controls (heavy fractions from 12C cDCE or 12C 1,4-dioxane amended samples) were determined using Welch's two sided t-test (within STAMP, p < 0.05) (Supplementary Figure 4.6 and 4.7). The dominant enriched genera for 13C uptake from 1,4-dioxane included Rhodopseudomonas and Rhodanobacter (Supplementary Figure 4.6). Whereas the dominant enriched genera for 13C uptake from cDCE included Bacteriovorax, Pseudomonas and 167 Sphingomonas (Supplementary Figure 4.7). An additional statistical test (Wilcoxon Rank, p < 0.05) confirmed the enrichment of Rhodopseudomonas and Rhodanobacter in one or both replicates of both soils (Figure 4.5). For cDCE, multiple genera were enriched in replicates of soil 1 and 2 (Figure 4.6). Enriched genera in soil 1 included: Bacteriovorax, Bradyrhizobium and two unclassified genera from Blastocatellaceae. Enriched genera in soil 2 included: Bradyrhizobium, Caulobacter, an uncultured genus within Vicinamibacterales, Pseudomonas and Sphingomonas. The greater diversity of dominant enriched OTUs in cDCE microcosms between soils, compared to 1,4-dioxane microcosms between soils, is consistent with the NMDS analysis indicating clear distinctions between cDCE communities between soils compared to 1,4- dioxane communities between soils (Figure 4.3). 168 Soil 1 Soil 2 Figure 4. 5. Boxplots with Wilcoxon Rank test results between phylotypes enriched (as determined by STAMP) in 13C amended heavy fractions (red dots) compared to the 12C amended heavy fractions (purple dots) by soil (upper is Soil 1 and lower is Soil 2) and by replicate of 1,4-dioxane amended samples. The graphs on the right have a different y-axis. P values of 0.0001, 0.001, 0.01, 0.05, >0.05 are presented by ****, ***, **, *, ns. 169 Soil 1 Soil 2 Figure 4. 6. Boxplots with Wilcoxon Rank test results between phylotypes enriched (as determined by STAMP) in 13C amended heavy fractions (green dots) compared to the 12C amended heavy fractions (blue dots) by soil (upper is Soil 1 and lower is Soil 2) and by replicate of cDCE amended samples. The graphs for Soil 2 have a different y-axis. P values of 0.0001, 0.001, 0.01, 0.05, >0.05 are presented by ****, ***, **, *, ns. 170 3.4 Co-Occurrence Networks Co-occurrence networks were generated to illustrate the differences between soil 1 and soil 2 microbial communities involved in 1,4-dioxane and cDCE degradation (those enriched in 13C heavy fractions, STAMP analysis) (Supplementary Figure 4.8). The OTUs with a correlation coefficient > 0.7 were connected with lines. The main genera were represented by 166 and 172 nodes (analyzed as OTUs present in at least 50% of the samples and with the abundance ≥ 0.06%) for the degradation of 1,4-dioxane and cDCE, respectively. For the microbial communities associated with 1,4-dioxane biodegradation, 67 OTUs showed a significant difference between soil 1 and 2 (26 and 41 were more abundant in soil 1 and soil 2, respectively) (Supplementary Figure 4.8 A). In contrast, more OTUs (99) illustrated a significant difference between the soil 1 and soil 2 microbial communities for those involved in cDCE biodegradation (44 and 45 OTUs were more abundant in soil 1 and 2, and these OTUs were clearly separated) (Supplementary Figure 4.8 B). The network also illustrated the relationships between the enriched and other abundant OTUs. The enriched OTUs (by STAMP analysis) displayed on the networks are summarized (Supplementary Table 4.2). In the microbial communities associated with 1,4-dioxane biodegradation, the majority of OTUs classified as Actinobacteria, Proteobacteria and Acidobacteria (Supplementary Figure 4.9A). Actinobacteria and Proteobacteria were related to each other. A total of 58 and 8 enriched OTUs were displayed in soil 1 and 2, respectively. Most of the enriched OTUs were Actinobacteria and Proteobacteria (Supplementary Figure 4.9B). In the microbial communities associated with cDCE biodegradation, the majority of the OTUs were Proteobacteria, Bacteroidetes, Acidobacteria and Gemmatimonadetes (Supplementary Figure 4.10A). Proteobacteria connected more with Gemmatimonadetes. A total 171 of 26 and 16 enriched OTUs were displayed in soil 1 and 2, respectively. The enriched OTUs were Bacteroidetes, Acidobacteria in soil 1 while the enriched OTUs were Proteobacteria in soil 2 (Supplementary Figure 4.10B). 3.5 Predicted Functions and Correlations with OTUs In the current work, PICRUSt2 predicted KO functions formerly associated with 1,4-dioxane biodegradation in the 1,4-dioxane amended microcosms included toluene monooxygenase, propane monooxygenase (most abundant) and methane monooxygenase (10, 11) (Supplementary Figure 4.11). In the cDCE amended microcosms, the abundant function associated with cDCE included glutathione S-transferase (51, 59). For 1,4-dioxane, correlations between gene and phylotype abundance indicated propane monooxygenase positively correlated with Rokubacteriales, KD4-96 (Chloroflexi), Gitt-GS-36 (Chloroflexi) and uncultured genera from Vicinamibacterales and Gemmatimonadaceae (Figure 4.7A). For cDCE, glutathione S- transferase was positively correlated with BIrii4 and an unclassified genus from Xanthomonadales (Figure 4.7B). 172 A B Figure 4. 7. Correlation of KO functions associated with degradation and OTUs with average abundance higher than 0.05% from 1,4-dioxane (A) and cDCE in SIP tests. 18 OTUs had an absolute correlation coefficient high than 0.59 with at least 4 of function in 1,4-dioxane SIP. 20 OTUs had an absolute correlation coefficient high than 0.6 with at least 2 of function in 1,4-dioxane SIP. 4. Discussion There have been many reports of the common occurrence of TCE and 1,4-dioxane at contaminated sites (8, 9, 18). As TCE is reduced to cDCE by both indigenous microbial communities or dechlorinating cultures (3, 16, 18, 26) and “cDCE-stall” is common at contaminated sites the removal of this metabolite is also of concern. Previously, we reported 1,4- dioxane biodegradation in the two soils examined here (19). Here, we build on that research by 173 investigating the potential for the concurrent biodegradation of both cDCE and 1,4-dioxane. Further, the microorganisms involved in the uptake of carbon from each chemical were identified using DNA-based SIP. Additionally, the functional genes involved in the degradation of cDCE and 1,4-dioxane were predicted using PICRUSt2 (48) and their abundance was correlated to OTUs present. The impact of additional treatments on 1,4-dioxane biodegradation differed between the two soils. For soil 1, when all three substrates were present (lactate, cDCE and 1,4-dioxane), 1,4- dioxane removal was slower. However, in soil 2, 1,4-dioxane biodegradation was similar for all three treatments, indicating the presence of either lactate or cDCE or both did not impact 1,4- dioxane removal in this soil community. Inhibition of 1,4-dioxane biodegradation by additional substrates has been noted by others. For example, when propane was added to Azoarcus sp. DD4, co-metabolism of 1,4-dioxane was delayed and followed the co-metabolism of chloroethenes (1,1-dichlorothene, VC and cDCE) (17, 18). Further, research with Pseudonocardia dioxanivorans CB1190 indicated from four common co-contaminants (1,1- DCE, 1,1,1-TCA, cDCE, TCE), cDCE was the second most inhibitory chemical to aerobic 1,4- dioxane degradation (13). The most notable trend for cDCE removal was the positive impact of lactate. Also, when lactate was present, 1,4-dioxane did not impact cDCE removal and decreases in cDCE concentrations occurred earlier than 1,4-dioxane decreases in both soil microcosms. In comparison, when the microcosms were amended with only 1,4-dioxane and cDCE the trend was less clear. In soil 2 microcosms, 1,4-dioxane removal started before cDCE removal while in soil 1 microcosms, the removal for 1,4-dioxane and cDCE started at a similar time. 174 Unlike previous studies which have primarily involved the biodegradation of co-contaminants by isolates or commercially available mixed communities (16-18), this work examined contaminant biodegradation by indigenous microbial communities. The NMDS analysis indicated cDCE produced a clear difference between the microbial communities of the two soils. The difference between the two microbial communities was less distinct for the soil microcosms amended with 1,4-dioxane. These trends could suggest cDCE is more important for impacting microbial community structure, perhaps through inhibition or as a beneficial substrate. Interestingly, the microbial richness and diversity levels were also higher in the cDCE amended samples. Also, the enriched phylotypes illustrated greater differences between soils in the cDCE amended microcosms, compared to the 1,4-dioxane amended microcosms. To date, many aerobic 1,4-dioxane and cDCE degraders have been identified (Supplementary Table 4.1). In the current study, multiple 1,4-dioxane and cDCE degraders were detected in shotgun sequencing data from samples inoculated with both soils. To determine if these species were actively involved in biodegradation, SIP was employed to determine which microorganisms were responsible for carbon uptake from each chemical. From the twelve 1,4- dioxane degrading phylotypes identified by shotgun sequencing, only two were associated with carbon uptake from 1,4-dioxane. Specifically, the previously reported 1,4-dioxane degraders Rhodanobacter sp. and Xanthobacter flavus were detected via shotgun sequencing and OTUs classifying as Rhodanobacter and the family Xanthobacteraceae were detected via SIP. From the nineteen cDCE degrading phylotypes detected via shotgun sequencing only one genus (Pseudomonas) was detected via SIP. These results provide support for the importance of SIP, over sequencing alone, for connecting identity with function. 175 In the current work, many genera were enriched during the SIP experiments, suggesting a wide range of microorganisms were assimilating carbon from the biodegradation of 1,4-dioxane or cDCE. Significantly enriched genera from the biodegradation of 13C-1,4-dioxane included Rhodopseudomonas and Rhodanobacter. Consistent with the current study, Rhodopseudomonas was previously associated with the incorporation of 13C from 1,4-dioxane in aerobic experiments with activated sludge, and its abundance increased with the degradation of 1,4-dioxane in both batch tests and a full-scale treatment system (22). Rhodanobacter was also reported as a metabolizer for 1,4-dioxane, with the addition of tetrahydrofuran accelerating 1,4-dioxane degradation (60). Combined with the results from the current work, these studies indicate the importance of both genera for 1,4-dioxane biodegradation and future work should examine their occurrence and activity at 1,4-dioxane contaminated sites. The other SIP identified genera during 1,4-dioxane degradation illustrated lower relative abundance levels compared to Rhodopseudomonas and Rhodanobacter. An OTU classifying as Afipia was enriched in microcosms inoculated with soil 2. Similarly, others have linked this genus (Afipia sp. D1) to the assimilation of carbon from 1,4-dioxane (61). Afipia was also abundant in uncontaminated soil microcosms during 1,4-dioxane degradation (62, 63). Here, an unclassified genus from Xanthobacteraceae was associated with carbon uptake from 1,4-dioxane in soil 1 microcosms. This family includes the 1,4-dioxane degraders Xanthobacter sp. YN2 (64) and Xanthobacter flavus DT8 (65). Xanthobacteraceae significantly increased in 1,4-dioxane degradation tests with domestic wastewater activated sludge, and a novel 1,4-dioxane- hydroxylating monooxygenase was identified from Xanthobacter strains (66). Another enriched OTU from the 1,4-dioxane SIP study classified within the family Xanthomonadaceae. This family was previously linked to 1,4-dioxane degradation in activated sludge from a full-scale 176 bioreactor for landfill leachate treatment (67, 68). In other 1,4-dioxane degradation studies, dominant or enriched genera included Chryseobacterium, Dokdonella, Pseudonocardia, Bradyrhizobium, Mycobacterium, Nocardioides, and Kribbella (19, 69), however these genera were not identified via SIP in the current work. Dominant genera significantly enriched in the biodegradation of 13C-cDCE in either or both soil microcosms and replicates, included Bacteriovorax, Pseudomonas and Sphingomonas. The dominance of Pseudomonas is consistent with previous studies associating this genus with cDCE degradation (Supplementary Table 4.1). Although isolates from the genus Bacteriovorax have not been previously linked with cDCE biodegradation, this genus has previously been associated with hydrocarbon biodegradation (70, 71). In the current work, two enriched genera (Sphingomonas and Bradyrhizobium) during 13C-cDCE degradation were abundant or enriched in previous 1,4-dioxane degradation studies (14, 15, 68). Sphingomonas was dominant during 1,4-dioxane degradation by P. dioxanivorans CB1190 when residuals of chlorinated volatile organic compounds (including cDCE) were present (14, 15). However, in the current study, these genera were not associated with 13C uptake from 1,4-dioxane. Interestingly, there were multiple novel genera, with no previous links to cDCE or 1,4-dioxane, identified in the current study as carbon assimilators. Multiple functions for 1,4-dioxane and cDCE biodegradation were predicted in the soil microcosms using PICRUSt2 (48). The most abundant function for 1,4-dioxane biodegradation was propane monooxygenase. Many OTUs positively correlated with propane monooxygenase, including for example KD4-96 (Chloroflexi), an uncultured genus from the class of Vicinamibacterales and from the family of Gemmatimonadaceae. The high abundance of propane monooxygenase is consistent with previous work describing the dominance of propane 177 monooxygenase from Rhodococcus jostii RHA1 and Rhodococcus sp. RR1 in 1,4-dioxane degrading microcosms inoculated with these soils (19). Several other previously identified 1,4- dioxane degrading enzymes were predicted to be present in the soil microcosms, however, additional research is needed to confirm if these enzymes are active. The functional genes associated with aerobic cDCE degradation include cytochrome P450 monooxygenase and glutathione S-transferase Polaromonas sp. strain JS666 (50, 51, 59). In the current study, both biomarkers correlated with a number of OTUs, but not Polaromonas strain JS666. However, these OTUs were not enriched in the SIP experiments, suggesting other enzymes may be involved or other methods (beyond the predictions provided by PICRUSt) are needed to obtain such information. In summary, this work demonstrated the concurrent removal of cDCE and 1,4-dioxane by indigenous soil microbial communities and the enhancement of cDCE removal by lactate. Through the use of SIP, multiple genera, both previously identified and not previously identified degraders, were enriched and benefited from the degradation of 1,4-dioxane and cDCE. In addition, a wide range of genes involved in the degradation were predicted to be associated with contaminant removal. These genera and genes were more diverse than previously reported. The extraction of DNA at only one time point during biodegradation is a potential limitation of the current study. Further, it is unknown if the enriched genera participated in carbon uptake from 1,4-dioxane and cDCE, or from their metabolites. Combining the current research with more quantitative approaches (e.g. qPCR, RNA-Seq) would enhance the information gained from the functional gene analysis. The data generated in the current study has the potential to be incorporated into diagnostic tests for assessing biodegradation potential at contaminated sites, for example, quantification of Rhodopseudomonas and Rhodanobacter at 1,4-dioxane contaminated 178 sites. Although the results suggest aerobic concurrent biodegradation of cDCE and 1,4-dioxane should be considered for chlorinated solvent site remediation, additional research is needed to determine if appropriately low contaminant concentrations can be reached. Acknowledgements Thanks to Stacey VanderWulp from MSU for providing the soil samples from KBS LTER. This work was partially supported by NSF grant number 1902250. 179 APPENDIX 180 APPENDIX Supplementary Table 4. 1. Identified 1,4-dioxane and DCE degraders with the lowest rank name and taxonomy ID from NCBI. Strains/species for 1,4-dioxane degradation Species/strain name in NCBI NCBI rank Pseudonocardia dioxanivorans CB1190 Pseudonocardia dioxanivorans CB1190 Rhodococcus ruber 219 Pseudonocardia benzennivorans B5 Mycobacterium sp. PH-06 Afipia sp. D1 Mycobacterium sp. D6 Mycobacterium sp. D11 Pseudonocardia sp. D17 Acinetobacter baumannii DD1 Rhodanobacter AYS5 Xanthobacter flavus DT8 Rhodococcus ruber Pseudonocardia benzenivorans Mycobacterium dioxanotrophicus Afipia sp. D1 Mycobacterium sp. D6 Mycobacterium sp. D11 Pseudonocardia sp. D17 Acinetobacter baumannii Rhodanobacter sp. Xanthobacter flavus Rhodococcus aetherivorans JCM 14343 Rhodococcus aetherivorans Pseudonocardia tetrahydrofuranoxydans sp. K1 Pseudonocardia tetrahydrofuranoxydans Pseudonocardia sp. ENV478 Pseudonocardia sp. ENV478 Rhodococcus ruber T1 Rhodococcus ruber T5 Rhodococcus ruber ENV 425 Rhodococcus RR1 Flavobacterium sp. Mycobacterium vaccae Mycobacterium sp. ENV 421 Pseudomonas mendocina KR1 Ralstonia pickettii PKO1 Burkholderia cepacia G4 Rhodococcus ruber Rhodococcus ruber Rhodococcus ruber Rhodococcus sp. RR1 Flavobacterium sp. Mycolicibacterium vaccae Mycobacterium sp. ENV421 Pseudomonas mendocina Ralstonia pickettii Burkholderia cepacia Methylosinus trichosporium OB3b Methylosinus trichosporium OB3b Pseudonocardia acacia JCM Pseudonocardia acaciae Pseudonocardia asaccharolytica JCM Pseudonocardia asaccharolytica Pseudomonas pickettii PKO1 Rhodococcus sp. YYL Rhodococcus josti RHA1 Pseudonocardia K1 Mycobacterium vaccae JOB5 Ralstonia pickettii Rhodococcus sp. YYL Rhodococcus jostii RHA1 Pseudonocardia sp. Mycolicibacterium vaccae Rhodococcus rhodochrous ATCC 21198 Rhodococcus rhodochrous ATCC 21198 strain species species species species species species species species species species species species species species species species species species species species species species species strain species species species species strain species species strain NCBI taxID 675635 1830 228005 482462 882658 882659 882660 882661 470 1883446 281 191292 102884 377619 1830 1830 1830 402393 239 1810 1213407 300 329 292 595536 551276 54010 329 423618 101510 60912 1810 1429046 Number of subtrees 0 4 0 0 0 0 0 0 1003 0 0 1 1 0 4 4 4 0 0 A 3 0 7 5 5 0 1 1 5 0 0 0 A 3 0 Reference (10, 11, 28) (72) (10, 73) (11, 74) (61) (61) (61) (61) (75) (60) (65) (76) (11, 77) (11, 78) (61) (61) (79) (10, 11, 80) (81) (82) (11, 83) (10, 11, 84) (10, 85) (10, 11, 86) (10, 11, 87) (76) (76) (85) (11, 88) (11, 89) (10, 90) (10, 91) (92) Strains/species for aerobic DCE degradation Species/strain name in NCBI NCBI rank NCBI taxID Number of subtrees Reference Methylosinus trichosporium OB3b Methylosinus trichosporium OB3b Rhodococcus rhodochrous ATCC 21198 Rhodococcus rhodochrous ATCC 21198 Xanthobacter autotrophicus Ralstonia pickettii PKO1 Cupriavidus necator JMP134 Burkholderia vietnamenisis G4 Polaromonas chloroethenica JS666 Xanthobacter autotrophicus Ralstonia pickettii Cupriavidus pinatubonensis JMP134 Burkholderia vietnamiensis G4 Polaromonas sp. JS666 Methylococcus capsulatus Bath Methylococcus capsulatus str. Bath Pseduomonas stutzeri OX1 Nocardioides CF8 Rhodococcus globerulus AD45 Gordonia rubripertinca B-276 Pseudomonas stutzeri Nocardioides sp. CF8 Rhodococcus globerulus Gordonia rubripertincta Mycobacterium chubuense NBB4 Mycolicibacterium chubuense NBB4 Rhodococcus sp. Ralstonia sp. Variovorax sp. Comamonas testosteroni RF2 Bacillus sp. Pseudomonas sp. OX1 Ralstonia sp. TRW-1 Pseudomonas sp. YKD221 Rhodococcus sp. Strain AD45 Pseudomonas plecoglossicida Methylocystis sp. strain M Mycobacterium sp. strain TRW-2 Mycobacterium vaccae strain JOB5 Pseudomonas sp. strain JR1 Pseudomonas butanavora Pseudomonas putida strain F1 Rhodococcus sp. strain PB1 Rhodococcus sp. Ralstonia sp. Variovorax sp. Comamonas testosteroni Bacillus sp. Pseudomonas sp. OX1 Ralstonia sp. Pseudomonas sp. Rhodococcus sp. AD45 Pseudomonas plecoglossicida Methylocystis sp. M Mycobacterium sp. Mycolicibacterium vaccae Pseudomonas sp. JR1 Thauera butanivorans Pseudomonas putida F1 Rhodococcus sp. 181 strain strain species species strain strain species strain species species species species strain species species species species species species species species species species species species species species species strain species 595536 1429046 280 329 264198 269482 296591 243233 316 110319 33008 36822 710421 1831 54061 1871043 285 1409 320855 54061 306 103808 70775 51782 1785 1810 47160 86174 351746 1831 0 0 1 5 0 0 0 0 19 0 1 1 0 0 A 0 A 0 A 12 0 A 0 0 A 0 A 0 2 0 0 3 0 1 0 0 A (93) (92) (94) (94) (94) (94) (94) (94) (94) (94) (94) (94) (94) (95) (95) (95) (96) (97) (98) (99) (100) (101) (102) (103) (103) (103) (103) (103) (103) (103) Supplementary Table 4. 1.(continued) Rhodococcus erythropolis strain BD1 Rhodococcus erythropolis Xanthobacter sp. strain Py2 Xanthobacter autotrophicus Py2 species strain 1833 78245 10 0 (103) (103) A: The identified strain name from the paper could not be searched in NCBI taxonomy browser, for example: Pseudonocardia K1 was assigned to Pseudonocardia sp. which belonged to unclassified Pseudonocardia. Supplementary Table 4. 2. Enriched OTUs captured by the co-occurrence network. These OTUs were enriched in heavy fractions of 13C labeled chemicals amended samples determined by STAMP. 1,4-dioxane degradation soil 1 Otu00004_Rhodopseudomonas Otu00010_Rhodanobacter Otu00014_uncultured_ge99 Otu00016_Sphingomonas Otu00019_Xanthobacteraceae_unclassified7 cDCE degradation soil 1 Otu00001_Bacteriovorax Otu00012_RB4 Otu00013_Blastocatellaceae_unclassified Otu00027_uncultured Otu00031_Flavobacterium 1,4-dioxane degradation soil 2 Otu00001_Rhodanobacter Otu00004_Rhodopseudomonas Otu00003_Chujaibacter Otu00006_Chujaibacter Otu00008_Acidipila Otu00013_Ktedonobacteraceae_unclassifi ed Otu00026_Ensifer92 Otu00126_Chujaibacter Otu00041_37-3_ge Otu00045_KD4-96_ge Otu00057_Bradyrhizobium9 Otu00060_Blastocatellaceae_unclassified6 Otu00061_Blastocatellaceae_unclassified59 Otu00074_uncultured Otu00075_Chitinophagaceae_unclassified7 9 Otu00076_Vicinamibacteraceae_ge Otu00096_uncultured_ge Otu00116_uncultured95 Otu00122_Subgroup__ge Otu00140_uncultured_ge Otu00142_NS-2_marine_group_ge Otu00143_Blastocatellaceae_unclassified8 Otu00148_Pseudomonas Otu00160_MND Otu00161_Chthonomonadales_ge Otu00174_Chitinophaga Otu00179_Nitrosomonadaceae_unclassified 98 Otu00186_Sphingomonas98 Otu00191_Gemmatimonas Otu00197_Ellin57 Otu00233_uncultured99 cDCE degradation soil 1 Otu00002_Pseudomonas Otu00005_Sphingomonas Otu00014_Ellin667 Otu00020_Vicinamibacteraceae_ge Otu00022_Luteimonas98 Otu00057_Bradyrhizobium9 Otu00070_Adhaeribacter Otu00081_SC-I-84_ge Otu00095_Crenobacter Otu00114_Fimbriimonadaceae_ge Otu00136_Dechloromonas86 Otu00137_Polaromonas63 Otu00149_Caulobacter Otu00169_uncultured Otu00185_Novosphingobium97 Otu00199_Altererythrobacter Otu00200_Sphingobacteriaceae_unclassifi ed Otu00022_Nitrosospira Otu00023_KD4-96_ge Otu00025_Candidatus_Udaeobacter Otu00031_KD4-96_ge Otu00035_Vicinamibacteraceae_ge Otu00036_Pedobacter Otu00039_Bacteria_unclassified Otu00041_uncultured Otu00043_Holophaga Otu00044_Blastocatellaceae_unclassified Otu00045_uncultured Otu00046_uncultured_ge Otu00047_Arenimonas Otu00049_KD4-96_ge Otu00050_uncultured Otu00052_Subgroup_7_ge Otu00055_uncultured Otu00058_Nakamurella Otu00059_uncultured Otu00061_KD4-96_ge Otu00063_Gemmatimonas Otu00068_uncultured Otu00070_Alicyclobacillus Otu00075_Ellin667 Otu00082_Pseudolabrys96 Otu00088_Microbacteriaceae_unclassified9 Otu00089_Luteimonas97 Otu00095_Gammaproteobacteria_unclassifi ed Otu00097_uncultured Otu00099_uncultured Otu00115_Ellin65595 Otu00116_MND Otu00119_Rokubacteriales_ge Otu00122_uncultured Otu00124_Haliangium Otu00125_Luedemannella6 Otu00129_uncultured_ge Otu00130_Candidatus_Solibacter Otu00132_Nitrospira Otu00134_AD3_ge Otu00138_Phenylobacterium Otu00140_Subgroup_7_ge Otu00141_Acidibacter Otu00142_uncultured_ge99 Otu00148_uncultured_ge Otu00149_uncultured_ge54 Otu00150_WPS-2_ge Otu00151_MB-A2-8_ge Otu00152_uncultured Otu00153_Xanthomonadaceae_unclassified Otu00154_Elsterales_unclassified Otu00164_uncultured_ge Otu00169_Alcaligenaceae_unclassified Otu00171_Gemmatimonas Otu00173_MB-A2-8_ge 182 A B 1,4-Dioxane C13-1,4-Dioxane control 8 6 4 2 1,4-Dioxane c13-1,4-Dioxane control 0 20 40 60 80 100 0 20 40 60 80 100 C cDCE C13-cDCE controls 8 6 4 2 0 D cDCE C13-cDCE controls L / g m , C L / g m , C 8 6 4 2 8 6 4 2 0 10 20 30 40 50 0 10 time, day 20 30 time, day 40 50 Supplementary Figure 4. 1. Average concentration of labeled and unlabeled 1,4-dioxane (A, B), and cDCE (C, D) in triplicate sample microcosms and triplicate abiotic controls inoculated with soil 1 (A, C) and 2 (B, D). 183 A B Supplementary Figure 4. 2. Rarefaction curves for DNA extracts in heavy and light fractions from 1,4- dioxane (A) and cDCE (B) SIP experiments in microcosms amended with soil 1. and 2. 184 Supplementary Figure 4. 3. Non-metric Multi-dimensional Scaling (NMDS) plot and alpha diversity measurements for sequencing results of 1,4-dioxane and cDCE SIP tests by KBS soil 1 and 2. 185 A B Supplementary Figure 4. 4. The most abundant (top 40) genera (by mean, with phylum) in all SIP fractions from 1,4-dioxane (A) and cDCE (B) amended microcosms inoculated with soil 1 or 2. 186 A B Supplementary Figure 4. 5. Species or strains previously associated with 1,4-dioxane (A) or cDCE (B) biodegradation present in KBS soil 1 (red) and soil 2 (blue) from shotgun sequencing data. 187 A B 24 20 16 12 8 4 0 ) % ( e c n a d n u b a e v i t a l e R 60 50 40 30 20 10 0 ) % ( e c n a d n u b a e v i t a l e R T1-14D-R1 T1-14D-R2 T1-c13-14D-R1 T1-c13-14D-R2 2 1.5 1 0.5 0 ) % ( e c n a d n u b a e v i t a l e R T2-14D-R1 T2-14D-R2 T2-c13-14D-R1 T2-c13-14D-R2 Supplementary Figure 4. 6. Phylotypes statistically enriched (Welch's two sided t-test, p<0.05) in heavy fractions (1.730-1.747 g/mL) of 13C 1,4-dioxane amended samples compared to fractions of comparable buoyant density (1.730-1.748 g/mL) in 12C 1,4-dioxane amended samples in soil 1 (A) and 2 (B). Values and error bars represent averages and standard deviations from three fractions each (with each fraction being sequenced and quantified in triplicate). After removing the background phylotypes that were also enriched in light fractions, a total of 282 and 28 phylotypes were enriched 1,4-dioxane amended samples for soil 1 and 2, respectively. The figure only displayed phylotypes with a difference of average abundance from 13C 1,4-dioxane and amended 12C 1,4-dioxane samples higher than 0.15% (A) and 0.01% (B) for soil 1 and soil 2. The insert was in a smaller scale. 188 T1-cDCE-R1 T1-cDCE-R2 T1-c13-cDCE-R1 T1-c13-cDCE-R2 1.5 1 0.5 0 T2-cDCE-R1 T2-cDCE-R2 T2-c13-cDCE-R1 T2-c13-cDCE-R2 A 2 1.6 1.2 0.8 0.4 0 ) % ( e c n a d n u b a e v i t a l e R 10 B 8 6 4 2 0 ) % ( e c n a d n u b a e v i t a l e R Supplementary Figure 4. 7. Phylotypes statistically enriched (Welch's two sided t-test, p<0.05) in heavy fractions (1.733-1.744 g/mL) of 13C cDCE amended samples compared to fractions of comparable buoyant density (1.733-1.745 g/mL) in 12C DCE amended samples in soil 1 (A) and 2 (B). Values and error bars represent averages and standard deviations from three fractions each (with each fraction being sequenced and quantified in triplicate). After removing the background phylotypes that were also enriched in light fractions, a total of 30 and 25 phylotypes were enriched in DCE amended samples for soil 1 and soil 2, respectively . The figure only displayed phylotypes with a difference of average abundance from 13C DCE and amended 12C DCE samples higher than 0.1% (A) and 0.05% (B) for soil 1 and soil 2. The insert was in a smaller scale. 189 A. 1,4-dioxane B. cDCE A, 1,4-dioxane More in soil 1 (26) More in soil 2 (41) No significant difference (99) B, cDCE More in soil 1 (44) More in soil 2 (55) No significant difference (73) Supplementary Figure 4. 8. Co-occurrence networked based on spearman correlation (rho > 0.70 and p-value < 0.01) for the main OTUs from microbial communities for 1,4-dioxane (A) and cDCE (B) degradation. Connected nodes with lines had a rho > 0.7. Filters for main OTUs: present in at least 50% of samples, average abundance > 0.06% (A) and > 0.1% (B). There were 166 (A) and 172 (B) nodes met the filters. The networks were colored with OTUs show significant difference (p<0.05) of samples from heavy fraction of soil 1 and 2 amended with 13C labeled 1,4- dioxane or cDCE. Number of nodes belonging to that group was in the parentheses. 190 A. By phylum B. By enriched A, by phylum Actinobacteria (44) Proteobacteria (43) Acidobacteria (21) Bacteroidetes (13) Gemmatimonadetes (11) Others (34) B, by enriched Enriched in soil 1 (57) Enriched in soil 2 (7) Enriched in both (1) Not enriched (101) Supplementary Figure 4. 9. Co-occurrence networked based on spearman correlation (rho > 0.70 and p-value < 0.01) for the main OTUs from microbial communities for 1,4-dioxane degradation. Connected nodes with lines had a rho > 0.7. Filters for main OTUs: present in at least 50% of samples, average abundance > 0.06%. There were 166 nodes met the filters. A: OTUs were colored by phylum, B: OTUs were colored by if its abundance is significantly higher in DNA with C13 isotope high BD value fractions from soil 1 or soil 2. Number of nodes belonging to that group was in the parentheses. 191 A. By phylum B. By enriched A, by Phylum Proteobacteria (52) Bacteroidetes (31) Acidobacteria (28) Gemmatimonadetes (25) Verrucomicrobia (9) Others (27) B, by enriched Enriched in soil 1 (25) Enriched in soil 2 (15) Enriched in both (1) Not enriched (131) Supplementary Figure 4. 10. Co-occurrence networked based on spearman correlation (rho > 0.70 and p-value < 0.01) for the main OTUs from microbial communities for cDCE degradation. Connected nodes with lines had a rho > 0.7. Filters for main OTUs: present in at least 50% of samples, average abundance > 0.1%. There were 172 nodes met the filters. A: OTUs were colored by phylum, B: OTUs were colored by if its abundance is significantly higher in DNA with C13 isotope high BD value fractions from soil 1 or soil 2. Number of nodes belonging to that group was in the parentheses. 192 Supplementary Figure 4. 11. KO functions associated with 1,4-dioxane and cDCE in SIP fractions obtained from microcosm replicates (R1 and R2) in both soil 1 (T1) and 2 (T2). 193 REFERENCES 194 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES Steffan RJ, Vainberg S. 2013. Production and handling of Dehalococcoides bioaugmentation cultures, p 89-115. In Stroo HF, Leeson A, Ward CH (ed), Bioaugmentation for Groundwater Remediation. Springer, New York. Brown RA, Mueller JG, Seech AG, Henderson JK, Wilson JT. 2009. Interactions between biological and abiotic pathways in the reduction of chlorinated solvents. Remediation Journal 20:9-20. Ellis DE, Lutz EJ, Odom JM, Buchanan RJ, Bartlett CL, Lee MD, Harkness MR, DeWeerd KA. 2000. Bioaugmentation for Accelerated In Situ Anaerobic Bioremediation. Environmental Science & Technology 34:2254-2260. Yang Y, McCarty PL. 2002. Comparison between donor substrates for biologically enhanced tetrachloroethene DNAPL dissolution. Environ Sci Technol 36:3400-4. Yang Y, McCarty PL. 2000. Biologically Enhanced Dissolution of Tetrachloroethene DNAPL. Environmental Science & Technology 34:2979-2984. EPA. 2017. Technical Fact Sheet - 1,4-Dioxane. https://www.epa.gov/sites/production/files/2014- 03/documents/ffrro_factsheet_contaminant_14-dioxane_january2014_final.pdf. Zenker MJ, Borden RC, Barlaz MA. 2003. Occurrence and Treatment of 1,4-Dioxane in Aqueous Environments. Environmental Engineering Science 20:423-432. Anderson RH, Anderson JK, Bower PA. 2012. Co-occurrence of 1,4-dioxane with trichloroethylene in chlorinated solvent groundwater plumes at US Air Force installations: Fact or fiction. Integrated Environmental Assessment and Management 8:731-737. Adamson DT, Mahendra S, Walker KL, Rauch SR, Sengupta S, Newell CJ. 2014. A Multisite Survey To Identify the Scale of the 1,4-Dioxane Problem at Contaminated Groundwater Sites. Environmental Science & Technology Letters 1:254-258. 10. Mahendra S, Alvarez-Cohen L. 2006. Kinetics of 1,4-dioxane biodegradation by monooxygenase-expressing bacteria. Environ Sci Technol 40:5435-42. 11. He Y, Mathieu J, Yang Y, Yu P, da Silva MLB, Alvarez PJJ. 2017. 1,4-Dioxane Biodegradation by Mycobacterium dioxanotrophicus PH-06 Is Associated with a Group-6 Soluble Di-Iron Monooxygenase. Environmental Science & Technology Letters 4:494-499. 12. Mahendra S, Grostern A, Alvarez-Cohen L. 2013. The impact of chlorinated solvent co- contaminants on the biodegradation kinetics of 1,4-dioxane. Chemosphere 91:88-92. 13. Zhang S, Gedalanga PB, Mahendra S. 2016. Biodegradation Kinetics of 1,4-Dioxane in Chlorinated Solvent Mixtures. Environ Sci Technol 50:9599-607. 195 14. Miao Y, Johnson NW, Gedalanga PB, Adamson D, Newell C, Mahendra S. 2019. Response and recovery of microbial communities subjected to oxidative and biological treatments of 1,4- dioxane and co-contaminants. Water Research 149:74-85. 15. Miao Y, Johnson NW, Phan T, Heck K, Gedalanga PB, Zheng X, Adamson D, Newell C, Wong MS, Mahendra S. 2020. Monitoring, assessment, and prediction of microbial shifts in coupled catalysis and biodegradation of 1,4-dioxane and co-contaminants. Water Research 173:115540. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Polasko AL, Zulli A, Gedalanga PB, Pornwongthong P, Mahendra S. 2019. A Mixed Microbial Community for the Biodegradation of Chlorinated Ethenes and 1,4-Dioxane. Environmental Science & Technology Letters 6:49-54. Deng D, Li F, Wu C, Li M. 2018. Synchronic Biotransformation of 1,4-Dioxane and 1,1- Dichloroethylene by a Gram-Negative Propanotroph Azoarcus sp. DD4. Environmental Science & Technology Letters 5:526-532. Li F, Deng D, Zeng L, Abrams S, Li M. 2021. Sequential anaerobic and aerobic bioaugmentation for commingled groundwater contamination of trichloroethene and 1,4-dioxane. Sci Total Environ 774:145118. Ramalingam V, Cupples AM. 2020. Enrichment of novel Actinomycetales and the detection of monooxygenases during aerobic 1,4-dioxane biodegradation with uncontaminated and contaminated inocula. Appl Microbiol Biotechnol 104:2255-2269. Radajewski S, Ineson P, Parekh NR, Murrell JC. 2000. Stable-isotope probing as a tool in microbial ecology. Nature 403:646-649. Paes F, Liu X, Mattes TE, Cupples AM. 2015. Elucidating carbon uptake from vinyl chloride using stable isotope probing and Illumina sequencing. Applied Microbiology and Biotechnology 99:7735-7743. Aoyagi T, Morishita F, Sugiyama Y, Ichikawa D, Mayumi D, Kikuchi Y, Ogata A, Muraoka K, Habe H, Hori T. 2018. Identification of active and taxonomically diverse 1,4-dioxane degraders in a full-scale activated sludge system by high-sensitivity stable isotope probing. ISME J 12:2376-2388. Cho K-C, Lee DG, Roh H, Fuller ME, Hatzinger PB, Chu K-H. 2013. Application of 13C-stable isotope probing to identify RDX-degrading microorganisms in groundwater. Environmental Pollution 178:350-360. Jayamani I, Cupples AM. 2015. Stable isotope probing reveals the importance of Comamonas and Pseudomonadaceae in RDX degradation in samples from a Navy detonation site. Environ Sci Pollut Res Int 22:10340-50. Sun W, Sun X, Cupples AM. 2012. Anaerobic methyl tert-butyl ether-degrading microorganisms identified in wastewater treatment plant samples by stable isotope probing. Appl Environ Microbiol 78:2973-80. Fennell DE, Carroll AB, Gossett JM, Zinder SH. 2001. Assessment of Indigenous Reductive Dechlorinating Potential at a TCE-Contaminated Site Using Microcosms, Polymerase Chain Reaction Analysis, and Site Data. Environmental Science & Technology 35:1830-1839. 196 27. 28. 29. Thelusmond JR, Strathmann TJ, Cupples AM. 2019. Carbamazepine, triclocarban and triclosan biodegradation and the phylotypes and functional genes associated with xenobiotic degradation in four agricultural soils. Sci Total Environ 657:1138-1149. Parales RE, Adamus JE, White N, May HD. 1994. Degradation of 1,4-dioxane by an Actinomycete in pure culture. Appl Environ Microbiol 60:4527-30. Gossett JM. 1987. Measurement of Henry's law constants for C1 and C2 chlorinated hydrocarbons. Environmental Science & Technology 21:202-208. 30. Myers MA, Johnson NW, Marin EZ, Pornwongthong P, Liu Y, Gedalanga PB, Mahendra S. 2018. Abiotic and bioaugmented granular activated carbon for the treatment of 1,4-dioxane- contaminated water. Environ Pollut 240:916-924. 31. 32. 33. 34. 35. Freedman DL, Gossett JM. 1989. Biological reductive dechlorination of tetrachloroethylene and trichloroethylene to ethylene under methanogenic conditions. Appl Environ Microbiol 55:2144- 51. Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. 2013. Development of a dual- index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq Illumina sequencing platform. Appl Environ Microbiol 79:5112-20. Schloss PD. 2009. A high-throughput DNA sequence aligner for microbial ecology studies. PLoS One 4:e8230. Pruesse E, Quast C, Knittel K, Fuchs BM, Ludwig W, Peplies J, Glockner FO. 2007. SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res 35:7188-96. Lahti L, Shetty S. 2017. Tools for microbiome analysis in R, http://microbiome.github.com/microbiome. 36. McMurdie PJ, Holmes S. 2013. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One 8:e61217. 37. Andersen KS, Kirkegaard RH, Karst SM, Albertsen M. 2018. ampvis2: an R package to analyse and visualise 16S rRNA amplicon data. bioRxiv doi:10.1101/299537:299537. 38. Wickham H. 2016. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. 39. Harrell FE. 2020. Hmisc: Harrell Miscellaneous, https://cran.r- project.org/web/packages/Hmisc/index.html. 40. 41. 42. Bates D, Maechler M, Davis TA, Oehlschlägel J, Riedy J. 2019. Matrix: Sparse and Dense Matrix Classes and Methods, https://cran.r-project.org/web/packages/Matrix/index.html. Csardi G, Nepusz T. 2006. The igraph software package for complex network research. InterJournal, Complex Systems 1695. Kassambara A. 2020. ggpubr: 'ggplot2' Based Publication Ready Plots., https://CRAN.R- project.org/package=ggpubr. 197 R Core Team. 2018. R: A language and environment for statistical computing, R Foundation for Statistical Computing, Vienna, Austria. , https://www.R-project.org/. RStudio Team. 2020. RStudio: Integrated Development for R., RStudio, PBC. Boston, MA, http://www.rstudio.com/. Parks DH, Tyson GW, Hugenholtz P, Beiko RG. 2014. STAMP: statistical analysis of taxonomic and functional profiles. Bioinformatics 30:3123-4. Bastian M, Heymann S, Jacomy M. 2006. Gephi: an open source software for exploring and manipulating networks. International AAAI Conference on Weblogs and Social Media. Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M. 2016. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res 44:D457-62. Douglas GM, Maffei VJ, Zaneveld JR, Yurgel SN, Brown JR, Taylor CM, Huttenhower C, Langille MGI. 2020. PICRUSt2 for prediction of metagenome functions. Nat Biotechnol 38:685- 688. Grostern A, Sales CM, Zhuang W-Q, Erbilgin O, Alvarez-Cohen L. 2012. Glyoxylate Metabolism Is a Key Feature of the Metabolic Degradation of 1,4-Dioxane by Pseudonocardia dioxanivorans Strain CB1190. Applied and Environmental Microbiology 78:3298. Nishino SF, Shin KA, Gossett JM, Spain JC. 2013. Cytochrome P450 Initiates Degradation of cis-Dichloroethene by Polaromonas sp Strain JS666. Applied and Environmental Microbiology 79:2263-2272. Giddings CGS, Jennings LK, Gossett JM. 2010. Microcosm Assessment of a DNA Probe Applied to Aerobic Degradation of cis-1,2-Dichloroethene by Polaromonas sp. Strain JS666. Groundwater Monitoring & Remediation 30:97-105. Van Hylckama VJ, De Koning W, Janssen DB. 1997. Effect of Chlorinated Ethene Conversion on Viability and Activity of Methylosinus trichosporium OB3b. Appl Environ Microbiol 63:4961-4. Semprini L. 1997. Strategies for the aerobic co-metabolism of chlorinated solvents. Curr Opin Biotechnol 8:296-308. Gu Z, Eils R, Schlesner M. 2016. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32:2847-9. Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30:2114-20. Li D, Luo R, Liu CM, Leung CM, Ting HF, Sadakane K, Yamashita H, Lam TW. 2016. MEGAHIT v1.0: A fast and scalable metagenome assembler driven by advanced methodologies and community practices. Methods 102:3-11. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. J Mol Biol 215:403-10. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 198 58. 59. 60. 61. 62. 63. Huson DH, Beier S, Flade I, Gorska A, El-Hadidi M, Mitra S, Ruscheweyh HJ, Tappu R. 2016. MEGAN Community Edition - Interactive Exploration and Analysis of Large-Scale Microbiome Sequencing Data. PLoS Comput Biol 12:e1004957. Jennings LK, Chartrand MM, Lacrampe-Couloume G, Lollar BS, Spain JC, Gossett JM. 2009. Proteomic and transcriptomic analyses reveal genes upregulated by cis-dichloroethene in Polaromonas sp. strain JS666. Appl Environ Microbiol 75:3733-44. Pugazhendi A. 2015. Biodegradation of 1,4-dioxane by Rhodanobacter AYS5 and the role of additional substrates. Annals of Microbiology 65. Sei K, Miyagaki K, Kakinoki T, Fukugasako K, Inoue D, Ike M. 2013. Isolation and characterization of bacterial strains that have high ability to degrade 1,4-dioxane as a sole carbon and energy source. Biodegradation 24:665-674. He Y, Mathieu J, da Silva MLB, Li M, Alvarez PJJ. 2018. 1,4-Dioxane-degrading consortia can be enriched from uncontaminated soils: prevalence of Mycobacterium and soluble di-iron monooxygenase genes. Microbial Biotechnology 11:189-198. Nam JH, Ventura JS, Yeom IT, Lee Y, Jahng D. 2016. Structural and Kinetic Characteristics of 1,4-Dioxane-Degrading Bacterial Consortia Containing the Phylum TM7. J Microbiol Biotechnol 26:1951-1964. 64. Ma F, Wang Y, Yang J, Guo H, Su D, Yu L. 2021. Degradation of 1,4-Dioxane by Xanthobacter sp. YN2. Curr Microbiol 78:992-1005. 65. 66. 67. 68. 69. 70. Chen D-Z, Jin X-J, Chen J, Ye J-X, Jiang N-X, Chen J-M. 2016. Intermediates and substrate interaction of 1,4-dioxane degradation by the effective metabolizer Xanthobacter flavus DT8. International Biodeterioration & Biodegradation 106:133-140. Chen R, Miao Y, Liu Y, Zhang L, Zhong M, Adams JM, Dong Y, Mahendra S. 2021. Identification of novel 1,4-dioxane degraders and related genes from activated sludge by taxonomic and functional gene sequence analysis. J Hazard Mater 412:125157. Xiong Y, Mason OU, Lowe A, Zhou C, Chen G, Tang Y. 2019. Microbial Community Analysis Provides Insights into the Effects of Tetrahydrofuran on 1,4-Dioxane Biodegradation. Applied and Environmental Microbiology 85:e00244-19. Xiong Y, Mason OU, Lowe A, Zhang Z, Zhou C, Chen G, Villalonga MJ, Tang Y. 2020. Investigating promising substrates for promoting 1,4-dioxane biodegradation: effects of ethane and tetrahydrofuran on microbial consortia. Biodegradation 31:171-182. Tusher TR, Shimizu T, Inoue C, Chien MF. 2019. Enrichment and Analysis of Stable 1,4- dioxane-Degrading Microbial Consortia Consisting of Novel Dioxane-Degraders. Microorganisms 8. Hu P, Dubinsky EA, Probst AJ, Wang J, Sieber CMK, Tom LM, Gardinali PR, Banfield JF, Atlas RM, Andersen GL. 2017. Simulation of Deepwater Horizon oil plume reveals substrate specialization within a complex community of hydrocarbon degraders. Proceedings of the National Academy of Sciences 114:7432-7437. 199 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. Bacosa HP, Erdner DL, Rosenheim BE, Shetty P, Seitz KW, Baker BJ, Liu Z. 2018. Hydrocarbon degradation and response of seafloor sediment bacterial community in the northern Gulf of Mexico to light Louisiana sweet crude oil. ISME J 12:2532-2543. Bock C, Kroppenstedt RM, Diekmann H. 1996. Degradation and bioconversion of aliphatic and aromatic hydrocarbons by Rhodococcus ruber 219. Applied Microbiology and Biotechnology 45:408-410. Kampfer P, Kroppenstedt RM. 2004. Pseudonocardia benzenivorans sp. nov. Int J Syst Evol Microbiol 54:749-751. Kim YM, Jeon JR, Murugesan K, Kim EJ, Chang YS. 2009. Biodegradation of 1,4-dioxane and transformation of related cyclic compounds by a newly isolated Mycobacterium sp. PH-06. Biodegradation 20:511-9. Huang H, Shen D, Li N, Shan D, Shentu J, Zhou Y. 2014. Biodegradation of 1,4-Dioxane by a Novel Strain and Its Biodegradation Pathway. Water, Air, & Soil Pollution 225:2135. Inoue D, Tsunoda T, Sawada K, Yamamoto N, Saito Y, Sei K, Ike M. 2016. 1,4-Dioxane degradation potential of members of the genera Pseudonocardia and Rhodococcus. Biodegradation 27:277-286. Kohlweyer U, Thiemer B, Schrader T, Andreesen JR. 2000. Tetrahydrofuran degradation by a newly isolated culture of Pseudonocardia sp. strain K1. FEMS Microbiol Lett 186:301-6. Vainberg S, McClay K, Masuda H, Root D, Condee C, Zylstra GJ, Steffan RJ. 2006. Biodegradation of ether pollutants by Pseudonocardia sp. strain ENV478. Appl Environ Microbiol 72:5218-24. Steffan RJ, McClay K, Vainberg S, Condee CW, Zhang D. 1997. Biodegradation of the gasoline oxygenates methyl tert-butyl ether, ethyl tert-butyl ether, and tert-amyl methyl ether by propane- oxidizing bacteria. Applied and Environmental Microbiology 63:4216. Stringfellow WT, Alvarez-Cohen L. 1999. Evaluating the relationship between the sorption of PAHs to bacterial biomass and biodegradation. Water Research 33:2535-2544. Sun B, Ko K, Ramsay JA. 2011. Biodegradation of 1,4-dioxane by a Flavobacterium. Biodegradation 22:651-9. Burback BL, Perry JJ. 1993. Biodegradation and biotransformation of groundwater pollutant mixtures by Mycobacterium vaccae. Appl Environ Microbiol 59:1025-9. 83. Masuda H, McClay K, Steffan RJ, Zylstra GJ. 2012. Characterization of three propane-inducible oxygenases in Mycobacterium sp. strain ENV421. Lett Appl Microbiol 55:175-81. 84. Whited GM, Gibson DT. 1991. Separation and partial characterization of the enzymes of the toluene-4-monooxygenase catabolic pathway in Pseudomonas mendocina KR1. J Bacteriol 173:3017-20. 200 85. 86. Kukor JJ, Olsen RH. 1990. Molecular cloning, characterization, and regulation of a Pseudomonas pickettii PKO1 gene encoding phenol hydroxylase and expression of the gene in Pseudomonas aeruginosa PAO1c. J Bacteriol 172:4624-30. Nelson MJ, Montgomery SO, O'Neill E J, Pritchard PH. 1986. Aerobic metabolism of trichloroethylene by a bacterial isolate. Appl Environ Microbiol 52:383-4. 87. Whittenbury R, Phillips KC, Wilkinson JF. 1970. Enrichment, Isolation and Some Properties of Methane-utilizing Bacteria. Microbiology 61:205-218. 88. 89. 90. 91. 92. 93. Yao Y, Lv Z, Min H, Lv Z, Jiao H. 2009. Isolation, identification and characterization of a novel Rhodococcus sp. strain in biodegradation of tetrahydrofuran and its medium optimization using sequential statistics-based experimental designs. Bioresource Technology 100:2762-2769. Sharp JO, Sales CM, LeBlanc JC, Liu J, Wood TK, Eltis LD, Mohn WW, Alvarez-Cohen L. 2007. An Inducible Propane Monooxygenase Is Responsible for Nitrosodimethylamine Degradation by Rhodococcus sp. Strain RHA1. Applied and Environmental Microbiology 73:6930. Thiemer B, Andreesen JR, Schrader T. 2003. Cloning and characterization of a gene cluster involved in tetrahydrofuran degradation in Pseudonocardia sp. strain K1. Arch Microbiol 179:266-77. Smith CA, O'Reilly KT, Hyman MR. 2003. Characterization of the initial reactions during the cometabolic oxidation of methyl tert-butyl ether by propane-grown Mycobacterium vaccae JOB5. Appl Environ Microbiol 69:796-804. Rasmussen MT, Saito AM, Hyman MR, Semprini L. 2020. Co-encapsulation of slow release compounds and Rhodococcus rhodochrous ATCC 21198 in gellan gum beads to promote the long-term aerobic cometabolic transformation of 1,1,1-trichloroethane, cis-1,2-dichloroethene and 1,4-dioxane. Environ Sci Process Impacts 22:771-791. Lee SW, Keeney DR, Lim DH, Dispirito AA, Semrau JD. 2006. Mixed pollutant degradation by Methylosinus trichosporium OB3b expressing either soluble or particulate methane monooxygenase: can the tortoise beat the hare? Appl Environ Microbiol 72:7503-9. 94. Mattes TE, Alexander AK, Coleman NV. 2010. Aerobic biodegradation of the chloroethenes: pathways, enzymes, ecology, and evolution. FEMS Microbiol Rev 34:445-75. 95. 96. 97. Elango V, Kurtz HD, Freedman DL. 2011. Aerobic cometabolism of trichloroethene and cis- dichloroethene with benzene and chlorinated benzenes as growth substrates. Chemosphere 84:247-253. Zalesak M, Ruzicka J, Vicha R, Dvorackova M. 2017. Cometabolic degradation of dichloroethenes by Comamonas testosteroni RF2. Chemosphere 186:919-927. Aulenta F, Verdini R, Zeppilli M, Zanaroli G, Fava F, Rossetti S, Majone M. 2013. Electrochemical stimulation of microbial cis-dichloroethene (cis-DCE) oxidation by an ethene- assimilating culture. N Biotechnol 30:749-55. 201 98. 99. Chauhan S, Barbieri P, Wood TK. 1998. Oxidation of trichloroethylene, 1,1-dichloroethylene, and chloroform by toluene/o-xylene monooxygenase from Pseudomonas stutzeri OX1. Appl Environ Microbiol 64:3023-4. Elango VK, Liggenstoffer AS, Fathepure BZ. 2006. Biodegradation of vinyl chloride and cis- dichloroethene by a Ralstonia sp. strain TRW-1. Appl Microbiol Biotechnol 72:1270-5. 100. Yonezuka K, Araki N, Shimodaira J, Ohji S, Hosoyama A, Numata M, Yamazoe A, Kasai D, Masai E, Fujita N, Ezaki T, Fukuda M. 2016. Isolation and characterization of a bacterial strain that degrades cis-dichloroethenein the absence of aromatic inducers. The Journal of General and Applied Microbiology 62:118-125. 101. 102. van Hylckama Vlieg JET, Kingma J, van den Wijngaard AJ, Janssen DB. 1998. A Glutathione- Transferase with Activity towards cis-1,2-Dichloroepoxyethane Is Involved in Isoprene Utilization by Rhodococcus sp. Strain AD45. Applied and Environmental Microbiology 64:2800. Li J, de Toledo RA, Chung J, Shim H. 2014. Removal of mixture of cis-1,2- dichloroethylene/trichloroethylene/benzene, toluene, ethylbenzene, and xylenes from contaminated soil by Pseudomonas plecoglossicida. Journal of Chemical Technology & Biotechnology 89:1934-1940. 103. Dolinova I, Strojsova M, Cernik M, Nemecek J, Machackova J, Sevcu A. 2017. Microbial degradation of chloroethenes: a review. Environ Sci Pollut Res Int 24:13262-13283. 202 CHAPTER 5 Conclusions and Future Research Directions The key findings from each chapter are summarized below: Chapter 2. Analysis via shotgun sequencing of groundwater at five SDC-9 bioaugmented chlorinated solvent contaminated sites indicated the presence of DNA from numerous biodegraders, including Dehalococcoides, Desulfitobacterium and Dehalogenimonas. Further, DNA sequences from both anaerobic (pceA, tceA, vcrA and bvcA) and aerobic (etnE, etnC, mmoX and pmoA) functional genes were also detected. Additionally, DNA sequences from hydrogenases and functional genes associated with corrinoid metabolism and 1,4-dioxane degradation were also observed. Chapter 3. The analysis of groundwater from a biostimulated RDX contaminated site indicated DNA from more than thirty RDX degrading species were present in the pre- and post- biostimulated groundwater samples, with Variovorax sp. JS1663, Pseudomonas fluorescens, Pseudomonas putida and Stenotrophomonas maltophilia being the most abundant. From these, nine RDX degrading species significantly (p<0.05) increased in abundance following biostimulation. Both shotgun sequencing and qPCR indicated the most abundant RDX degrading functional genes were xenA and xenB. The relative abundance percentages of three RDX degrading genes (diaA, nsfI and pnrB) were similar and xplA was low or absent in most of the samples. Chapter 4. This study identified phylotypes associated with 1,4-dioxane and cDCE biodegradation using 16S rRNA gene amplicon sequencing coupled with SIP. In the 1,4-dioxane degrading microcosms two genera (Rhodopseudomonas and Rhodanobacter) were associated with the majority of 13C assimilation from 1,4-dioxane. In the cDCE degrading microcosms, the 203 dominant genera for 13C assimilation included Bacteriovorax, Pseudomonas and Sphingomonas. The predicted functions associated with 1,4-dioxane and cDCE biodegradation were also determined. Overall, the work demonstrated concurrent removal of cDCE and 1,4-dioxane by indigenous soil microbial communities and the enhancement of cDCE removal by lactate. The data generated on the phylotypes responsible for carbon uptake (as determined by SIP) could be incorporated into diagnostic molecular methods for site characterization. The results suggest aerobic concurrent biodegradation of cDCE and 1,4-dioxane should be considered for chlorinated solvent site remediation. Future research could include manipulating the DNA concentration submitted for sequencing, so that the comparison across samples would be based on the changes of absolute values rather than the relative abundance of a taxon or gene. Further, as additional sequencing data becomes available, data mining activities could improve our understanding of biodegradation potential across sites. Also, future research should include the correlation of geochemical data with molecular data to determine which factors are beneficial or impact the functional genera. In addition, future research could involve RNA-seq (RNA is reverse transcribed to cDNA, and submitted for high throughput sequencing) to reveal the active functions during contaminant biodegradation. 204