METAGENOMIC ANALYSIS OF ANTIBIOTIC RESISTANT GENES IN A CONVENTIONAL AND MEMBRANE BIOREACTOR WASTEWATER TREATMENT PLANT  By  Camille McCall                      A THESIS  Submitted to  Michigan State University In partial fulfillment of the requirements for the degree of   Environmental Engineering  Master of Science  2016   ABSTRACT  METAGENOMIC ANALYSIS OF ANTIBIOTIC RESISTANT GENES IN A CONVENTIONAL AND MEMBRANE BIOREACTOR WASTEWATER TREATMENT PLANT  By  Camille McCall  Wastewater treatment plants (WWTPs) are known environments for the presence and transfer of antibiotic resistant genes (ARGs), an evolving environmental pollutant. This study aimed to explore the prevalence of ARGs and resistant bacteria in a conventional, and a membrane bioreactor (MBR) WWTP in Michigan (USA). A sequence-based metagenomic approach was implemented to detect the profile of ARGs in the activated sludge (AS), before disinfection (BD), and after disinfection (AD) treatment stages in each WWTP. Metagenomic alignment detected genes resistant to sulfonamide, tetracycline, macrolide, elfamycin, aminoglycoside, and -lactam classes of antibiotics to be prevalent ARGs in both WWTPs. Effluent samples yielded the highest presence of ARGs in each plant compared to AS and BD samples. Quantitative analysis found that 56.25% and 53.33% of a total of 23  among all samples, were not detected until after disinfection samples for the conventional and MBR WWTPs, respectively. To our knowledge, this study is the first to report the prevalence of elfamycin resistant genes in WWTPs.  In addition to this, Acinetobacter baumannii, Pseudomonas aeruginosa, and Pasteurella multocida were found to be predominant resistant bacteria in AD samples from each WWTP. The occurrence of ARGs increased in both WWTPs as treatment progressed further suggesting that increased wastewater treatment selects for antimicrobial resistance. iii TABLE OF CONTENTS   LIST OF TABLES ......................................................................................................................... iv  LIST OF FIGURES ....................................................................................................................... vi  1. INTRODUCTION .....................................................................................................................1  2. MATERIALS AND METHODS ...............................................................................................4 2.1. Sample Collection ..........................................................................................................4 2.2. Sample Processing and Filtration ...................................................................................4 2.3. Nucleic Acid Extraction .................................................................................................5 2.4. High-throughput Sequencing and Preprocessing ...........................................................5 2.5. Sequence-based Metagenomic Alignment and Calculations .........................................6 2.6. ARG Reference Databases .............................................................................................7 2.7. Statistical Assessment of Sequence Alignment ..............................................................7  3. RESULTS ..................................................................................................................................8 3.1. Preprocessing Quality Analysis ......................................................................................8 3.2. Bowtie2 and BWA-MEM alignment on Illumina high-throughput sequences ..............8 3.3. ARGs Profiles in Wastewater Treatment Plants ..........................................................10 3.4. ARG Nucleotide Reference Databases .........................................................................11 3.5. Occurrence of ARGs in Wastewater Treatment Plants ................................................12 3.6. Statistical Assessment using MAPQ Scores generated from Bowtie2 and BWA-MEM Alignments ............................................................................................................................17  4. DISCUSSION ..........................................................................................................................19 4.1. Occurrence of ARGs in WWTPs .................................................................................19 4.2. Occurrence of ARGs in Various Wastewater Treatment Stages ..................................22 4.3. Composition of ARBs in Chlorine and UV Treated Wastewater ................................24 4.4. Statistical Analysis of Sequence Alignment using MAPQ Scores...............................26  5. CONCLUSION ........................................................................................................................28  APPENDIX ....................................................................................................................................29  REFERENCES ..............................................................................................................................37          iv LIST OF TABLES   Table 2.1. Process characteristics for East Lansing and Traverse City wastewater treatment plants............................................................................................................................................... 4  Table 3.1. Quality control analysis results on raw sequence reads using FastQC8  Table 3.2. ARGs and with either Bowtie2 or BWA-MEM in AS, BD, and AD samples from CAS and MBR WWTPs. Contains genes detected in CARD and ARG-ANNOT reference databases. All annotations are generated from their respective reference database< 90% similarity (-).....................................................................................15   Table A.1. Bowtie2 alignment statistics ...........30  Table A.2. Bowtie2 alignment statistics per sample30  Table A.3. BWA-MEM alignment statis...........31   Table A.4. BWA-MEM alignment statis31  Table A.5. coverage in activated sludge (AS) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one data set32    Table A.6. Average depth and gene similarity with BWA-90% coverage in activated sludge (AS) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one da32   Table A.7. Average depth and gene similarity with Bowtie2 alignment for ARGs coverage in before disinfection (BD) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one 33   Table A.8. Average depth and gene similarity with BWA-90% coverage in before disinfection (BD) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one data 34  Table A.9. Average depth and gene similacoverage in after disinfection (AD) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one data set35  v Table A.10. Average depth and gene similarity with BWA-90% coverage in after disinfection (AD) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one 36                                           vi LIST OF FIGURES   Figure 1. Number of mapped bases in CARD and ARG-ANNOT reference databases per sample. The number of mapped bases was pulled from the standard CIGAR string from SAMtools statistical computation. Error bars indicate the number of mismatches computed given the corresponding BAM file   Figure 2. Comparison of average sequencing depth per sample when aligned with CARD and ARG-ANNOT nucleotide reference genomes for Bowtie2 and BWA-10  Figure 3or BWA-MEM in each WWTP. ARGs considered consist of genes from both reference databases12  Figure 4. Relative abundance of antibiotic classes per sample belonging to resistant genes -MEM. Genes were obtained from both CARD and ARG-ANNOT databases13                            1 1. INTRODUCTION   The development and overuse of antibiotics has led to the selection of antibiotic resistant genes (ARGs) in bacteria resulting in reduced susceptibility to a wide range of antimicrobial treatments. Antibiotics are largely used in animal production and are among the largest therapeutic treatment for bacterial infections in humans (Manaia et al., 2015; Schmieder and Edwards, 2012). It is said that within the first few years of introducing a new antibiotic, pathogens commonly known to persist in hospital settings, develop resistance (Schmieder and Edwards, 2012; Zhang et al., 2015b). To date, nosocomial (hospital) infections are one of the leading causes of death (Xia et al., 2016). Resistant pathogens of this nature escape controlled settings through, most commonly, hospital waste streams and are released into the environment. Antibiotic resistant genes have become an evolving environmental pollutant known to occur in ecosystems such as, soils, surface waters, and WWTPs (Manaia et al., 2015). WWTPs are known to select for antimicrobial resistance given the various treatment processes and the harboring of an abundance of diverse microbial communities (Mao et al., 2015; Murray et al., 1984).   While ARGs in WWTPs have been explored in many facets, the effects of treatment on these genes are still unclear. Several studies report discrepancies in the effects of treatment on the behavior of ARGs throughout various wastewater treatment stages (Diehl and LaPara, 2010; Murray et al., 1984). The differences in how ARGs behave throughout waste treatment cycles can be caused by several factors that are at play. For example, biological treatment types, hydraulic detention time, presence of heavy metals, and temperature all play a role in the prevalence and removal of ARGs in wastewater treatment systems (Bondarczuk et al., 2016; Diehl and LaPara, 2010; Gao et al., 2012b; Mao et al., 2015; Novo and Manaia, 2010; Zhang et al., 2015c; Zhang et al., 2015b). Common ARGs found in waste streams such as tetracycline and sulfonamide resistant  2 genes have been studied at great lengths, generally by means of culturing and polymerase chain reaction (PCR) techniques (Mao et al., 2015, Gao et al., 2012a; Zhang et al., 2011; Gao et al., 2012b; Diehl and LaPara, 2010; Novo and Manaia, 2010; Munir et al., 2011). However, studies capturing a wide spectrum of ARGs in various types of wastewater treatment systems are limited. Considering the role of WWTPs in the selection or reduction of ARGs, the need to investigate the prevalence and diversity of these genes on a metagenomic scale has become increasingly important over the years.  Though, culture-based techniques have made it possible to detect ARGs in environmental samples, a reported 99% of environmental bacteria cannot be cultured (Varela and Manaia, 2013). This makes culture-base metagenomics insufficient in identifying novel, or a broad spectrum of antibiotic resistant bacteria within microbial communities (Schmieder and Edwards, 2012). The introduction of sequence-based metagenomics allows for the detection of a wide range of ARGs through genomic mapping and gene affiliation to national reference databases (Schmieder and Edwards, 2012). Sequence-based techniques involve sequencing a random sample of genetic material recovered directly from the environment without the use of culturing (Penders et al., 2013). Sequencing platforms such as Illumina, produce vast libraries of genetic information that are later sequenced and can then be post-processed for further analysis.   In this study, we used a sequenced-based metagenomic approach to assess the occurrence of ARGs in a conventional activated sludge (CAS) with chlorine (Cl) disinfection, and a membrane bioreactor (MBR) with ultraviolet (UV) disinfection WWTP in Michigan. Samples were extracted from the activated sludge, before disinfection, and after disinfection stages of each WWTP. Our  3 major goal is to identify predominant antibiotic resistant bacteria (ARB) in AD samples from each WWTP.                       4 2. MATERIALS AND METHODS 2.1. Sample Collection   Sewage samples were collected from the East Lansing and Traverse City WWTPs in Michigan (U.S.A.) in 2013. The characteristics of these WWTPs are shown in Table 2.1. Samples were taken from three different locations throughout the treatment process: activated sludge (AS), before disinfection (BD) and after disinfection (AD). Grab effluent samples were collected for bacterial isolation from each location. All samples were mixed and stored on ice, then transported to the laboratory for further processing.  Table 2.1. Process characteristics for East Lansing and Traverse City wastewater treatment plants.  East Lansing WWTP Traverse City WWTP Wastewater treatment process (biological treatment) Conventional Activated Sludge (CAS) Membrane Biological Reactor (MBR) Sludge Retention Time (SRT) 14 days 7.58 days Capacity 18.8 MGD 17.0 MGD Average Flow 13.4 MGD 8.5 MGD Discharge Rate 14.1 MGD 4.0 MGD Disinfection Chlorine (Cl) Ultraviolet (UV)   2.2. Sample Processing and Filtration   (Millipore, Billerica, MA). The volume of AD and BD samples filtered was 1 L for each of the four samples. The filters were collected in sterile 50 mL polypropylene tubes and 50 mL Phosphate Buffer (1X PBS) was added in each tube containing a filter. The tubes were vortexed for five minutes to allow the biomass layer on the filters to mix with water. 50 mL of the AS samples were also collected in sterile centrifuge tubes. All tubes were centrifuged for 20 min at 4500 rpm to concentrate the sample down to 2 mL. Supernatant was discarded and the concentrates were stored DNA extraction was performed.  5 2.3. Nucleic Acid Extraction  Bacterial DNA was extracted using a MagNA Pure Compact DNA extractor (Roche Applied MagNA Pure Compact utilizes a magnetic-bead technology for the isolation process. Sample amount of stored in a freezer at -20°C. DNA concentration was determined using the NanoDrop Spectrophotometer (NanoDrop® ND-1000, Wilmington, DE). 2.4. High-throughput Sequencing and Preprocessing  All six bacterial of DNA (per sample) was sent to the Research Technology Support Facility (RTSF) at Michigan State University. The NuGEN Ovation Ultralow Library System, with an input requirement of 1-100 ng of DNA, was used for all samples to accommodate for any sample containing low genetic material. After preparation, all libraries were sequenced on an Illumina platform (Illumina HiSeq2500, Roche Technologies) generating 150 bp paired-end reads.   The reads were returned as FASTQ.GZ files. All FASTQ.GZ files were processed using a Unix/Linux system offered through the MSU High Performance Computing Center (HPCC). All raw sequence reads were analyzed for quality using FastQC, a quality control tool for sequencing data (Andrews, 2010). Based on the quality control check, Illumina adapters and low quality bases were removed from all raw reads using Trimmomatic (Bolger et al., 2014). Finally, FastQC was performed once more to ensure the integrity of the sequence reads and the accuracy of the latter genetic alignment processes.    6 2.5. Sequence-based Metagenomic Alignment and Calculations  All six metagenomes were analyzed using Bowtie2 and Burrow-Wheeler Aligner-Maximal Exact Match (BWA-MEM), tools for aligning sequence reads to reference genomes. Bowtie2 utilizes the full-text minute (FM) index, based on the Burrows-Wheeler transform (BWT), which allows for gapped-read alignment (Langmead and Salzberg, 2012). Bowtie2 was run using default settings (end-to-end alignment, and a minimum threshold alignment score of -90) for each metagenome (Langmead and Salzberg, 2012). BWA-MEM, similar to Bowtie2, uses the FM index and allows for long gapped-read alignments. BWA-MEM was run using default settings (local alignment) (Li, 2013ity when mapped with either Bowtie2 or BWA-MEM was established in order for resistant genes to be considered in this study (Kristiansson et al., 2011; Shi et al., 2013; Zhang et al., 2015c). similarity were then clustered together for further analysis.    The breadth of coverage (gene similarity), and average depth and standard deviation (Std) were extracted for all ARGs in the reference genomes. Gene similarity, and average depth and standard deviation of all genes considered in this study can be found in the appendix section (Tables A.5  A.10) The gene similarity for each gene was determined by normalizing the number of unique aligned bases in each position to the length of the reference gene (Molnar and Ilie, 2014). The depth, also known as the redundancy of coverage, for each gene was calculated by normalizing the total number of bases aligned in the reference gene against its total length. The redundancy of coverage for each metagenome was calculated using the Lander/Waterman equation, where L is the read length, N is the number of aligned reads and G is the haploid genome size (Sims et al., 2014).   7 2.6. ARG Reference Databases  Reference genomes from the Comprehensive Antibiotic Resistance Database (CARD) and the Antibiotic Resistance Gene-ANNOTation (ARG-ANNOT) database were downloaded and used for alignment. Both databases are composed of antibiotic resistant gene nucleotide sequences in FASTA format and consist of various antibiotic classes. The CARD and ARG-ANNOT references are 2,848,122 bp and 1,463,892 bp, respectively and consist of genetic data imported from NCBI GenBank and peer-reviewed publications (Gupta et al., 2015; McArthur et al., 2013). CARD and ARG-ANNOT sequences are classified based on the CARD NCBI taxonomy ontology and PubMed publications (McArthur et al., 2013), and a unified nomenclature (Gupta et al., 2015), respectively. All sequences were curated prior to being introduced into the databases. Statistical data (mapped reads, error rate, mismatches, insertions, and overall mapping quality) was retrieved  output file using SAMtools.  2.7. Statistical Assessment of Sequence Alignment   Mapping quality (MAPQ) scores per sequence were extracted from the Sequence Alignment/Mapping (SAM) files generated from Bowtie2 and BWA-MEM alignments. The AD samples from the CAS and MBR WWTPs were considered in the statistical analysis to generate a base-level assessment were sequence reads that were aligned during primary alignment, paired, and mapped in its proper pair (i.e. among present SAM flags, these were: 83, 99, 147, and 163). The probability of incorrect sequence alignment was calculated using the Phred scale (Ruffalo et al., 2012).    8 3. RESULTS  3.1. Preprocessing Quality Analysis    Quality analysis results on raw reads yielded an average guanine-cytosine (GC) content and mean quality of 56.2% and 36 for all reads, respectively. The total number of sequences per paired-end read for all samples range between 8.9-11.7 Mbp with an average of 10.4 Mbp (Table 3.1).  Table 3.1. Quality control analysis results on raw sequence reads using FastQC Sample Sample Abbreviation Sequences Per Pair (bp) Average GC Content (%) Conventional Activated Sludge _ Activated Sludge CAS_AS 9994682 56 Conventional Activated Sludge _ Before Disinfection CAS_BD 10129270 55 Conventional Activated Sludge _  After Disinfection CAS_AD 11667426 53 Membrane Bioreactor  _ Activated Sludge MBR_AS 11042222 57 Membrane Bioreactor  _ Before Disinfection MBR_BD 8879956 58 Membrane Bioreactor  _ After Disinfection MBR_AD 10670819 58  3.2. Bowtie2 and BWA-MEM alignment on Illumina high-throughput sequences  The average error rate, and GC content of mapped reads was 2.49% and 2.42%, and 53.18% and 47.8%, for CARD and ARG-ANNOT reference databases, respectively when aligned with Bowtie2. BWA-MEM generated an average error rate of 5.44% and 1.80%, and average GC content of 51.11% and 52.67% for CARD and ARG-ANNOT databases, respectively (See Tables A.1-A.4 for alignment statistics per sample). The average and maximum read length for BWA-MEM local alignment was 149 bp and 150 bp, respectively for all samples, making it analogous to Bowtie2 end-to-end read alignment of 150 bp. The number of reads mapped with BWA-MEM was 5.98 and 5.12 times greater than Bowtie2 alignment against the CARD and ARG-ANNOT reference genomes, respectively.    9    Figure 1. Number of mapped bases in CARD and ARG-ANNOT reference databases per sample. The number of mapped bases was pulled from the standard CIGAR string from SAMtools statistical computation. Error bars indicate the number of mismatches computed given the corresponding BAM file.   The number of bases mapped per sample for each reference genome reported the lowest alignment in AS samples for each treatment utility with AD samples containing the greatest  10 number of mapped reads (Figure 1). BWA-MEM covered 49.5% and 26.29% of ARGs for CARD and ARG-ANNOT reference genomes, respectively. This was about 14.8 and 10.2 times greater than Bowtie2. However, a significant portion of detected ARGs revealed less than 10% gene similarity, 70.61% and 85.57% for CARD and ARG-ANNOT databases, respectively. The average depth per sample for each database during BWA-MEM alignment closely corresponded to the average depth obtained during Bowtie2 alignment (Figure 2).    Figure 2. Comparison of average sequencing depth per sample when aligned with CARD and ARG-ANNOT nucleotide reference genomes for Bowtie2 and BWA-MEM.   3.3. ARGs Profiles in Wastewater Treatment Plants   Bowtie2 alignment of CAS samples against the CARD database yielded greater alignment per base compared to the MBR WWTP by 1.5 and 1.7-fold for BD, and AD samples, respectively. 0.000.200.400.600.801.001.201.401.60CAS_ASCAS_BDCAS_ADMBR_ASMBR_BDMBR_ADAverage Depth (x)Sample SitesBWA_ARG-ANNOTBWA_CARDBowtie_ARG-ANNOTBowtie_CARD 11 A 1.1-fold difference was found between AS samples in CAS and MBR WWTPs. The ARG-ANNOT database found a greater number of aligned bases in AS and BD samples for the MBR plant by 1.7, and 1.1-fold, respectively. Similar to the CARD database, CAS samples yielded greater alignment in the AD sample by 1.4-fold compared to the MBR when aligned against the ARG-ANNOT database (Figure 1). BWA-MEM alignment detected a greater number of mapped bases in the MBR samples relative to the CAS samples for AS and AD sites during alignment against the CARD database. (Figure 1). Overall, results reveal a systematic trend of increasing ARGs as treatment progresses AD>BD>AS (Figure 1).  3.4. ARG Nucleotide Reference Databases  Alignment against the ARG-ANNOT reference database mapped a significantly lower number of sequence reads relative to CARD, approximately, 6.4 and 9.5 times less for Bowtie2 and BWA alignments, respectively. However, the lower number of mapped reads in the ARG-ANNOT reference database is expected considering its lesser genome size of 1463892 bp containing 1691 genes, compared to 2848122 bp with 3008 unique ARGs. ARG-ANNOT database detected only a small portion of ARGs with a gene id% in all samples given both aligners. Thus, the majority of prevalent ARGs found in all samples were identified when aligned to the CARD database. Genes resistant to Sulfonamide (sulI) (Genbank accession no. AF071413), detected in AS, BD, and AD samples, macrolides (msr(E), mph(E)) (Genbank accession no. JF769133), detected in AS, BD, and AD samples, and streptomycin (strA) (Genbank accession no. AB366441), detected in BD samples, with greater than 90% coverage were detected in ARG-ANNOT reference database. Tetracycline resistant gene (tet(39)) (Genbank accession no. AY743590), detected in BD and AD samples, was detected in both databases with identical gene  12 similarities. The remaining ARGs considered in latter discussions were derived from the CARD reference database.  3.5. Occurrence of ARGs in Wastewater Treatment Plants  A total of 23 different  90% gene similarity to its reference gene and therefore, were clustered together and considered in this study. The profiles of prevalent antibiotic resistant genes in each sample resemble the profiles of the total number of mapped bases presented in Figure 1, see Figure 3.   Figure 3.  when aligned with either Bowtie2 or BWA-MEM in each WWTP. ARGs considered consist of genes from both reference databases.  Genes with resistance to macrolides, sulfonamide-lactam, and aminoglycoside classes of antibiotics were predominant, with AD samples containing the highest occurrences of these ARGs (Figures 3 and 4).   29166915024681012141618ASBDADNumber of Detected ARGsTreatment StageCASMBR 13  Figure 4. Relative abundance of antibiotic classes per sample belonging to resistant genes  90% gene similarity when aligned with either Bowtie2 of BWA-MEM. Genes were obtained from both CARD and ARG-ANNOT databases.   Metagenomic alignment found sulfonamides, macrolides-lactam, and elfamycin ARGs to be predominant in AS samples (Table 3.2). SulI (GenBank accession no. AF071413) was detected with Bowtie2 alignment at a 98% and 96% similarity to the reference gene for CAS and MBR WWTPs, respectively. Elfamycin resistant (tuf) (Genbank accession no. CP002695) genes revealed 99% and 100% gene similarity when aligned with the MBR samples according to BWA-MEM alignment. -lactamase (blaRTG-5), and erythromycin (ermB) resistant genes conferred a 100% gene similarity with the MBR sequence reads for AS when aligned with BWA-MEM (GenBank accession nos. JQ364968 and X64695, respectively).   Tet39, sul1, msr(E), mph(E), streptomycin (strA, rpsL), and tuf resistant genes were prevalent in BD samples (Table 3.2). Macrolides and sulI resistant genes coverage in both WWTPs (Tables A.7-A.8). Tet(39) was detected in the MBR WWTP with a gene identity of 94.6% and 100% from Bowtie2 and BWA alignments, respectively. Tuf genes obtained 0102030405060MacrolideSulfonamideTetracyclineBeta-lactamaseElfamycinAminoglycosideMultidrug EffluxRelative Abundance (%)Antibiotic ClassTC_ADEL_ADTC_BDEL_BDTC_ASEL_AS 14 a gene homology as high as 100% when aligned with BWA-MEM for both CAS and MBR BD samples (Tables A.7-A.8).    The greatest number of prevalent ARGs during Bowtie2 and BWA-MEM alignment occurred in AD samples (Figure 3). Significant portions of ARGs were not detected until AD samples in either Bowtie2 or BWA alignment. A total of 16 prevalent ARGs were detected in CAS samples, of which 56.25% of them arose after disinfection. MBR samples detected 15 ARGs, of which 53.33% of them were detected after disinfection (Figure 3 and Table 3.2). The portion of genes that persisted in AD samples from the activated sludge process was 12.50% and 33.33% for CAS and MBR treatment plants, respectively. BlaRTG-5, tet39, sul1, aminoglycoside (aadA6, msr(E), mph(E), tuf, rpsL and multidrug resistant efflux transporter (mexF) ARGs were found to be prominent in both AD samples (Table 3.2).              15 Table 3.2. -MEM in AS, BD, and AD samples from CAS and MBR WWTPs. Contains genes detected in CARD and ARG-ANNOT reference databases. All annotations are generated from their respective reference database. < 90% similarity (-).  Database Accession No. Best Hit NCBI Genbank Accession No. Gene ARG Class ARB CAS AS MBRAS CAS BD MBRBD CAS AD MBRAD AM087411.1.gene3 AM087411 aadA6 Aminoglycoside Pseudomonas aeruginosa - - - - + + JQ364968.1.gene6 JQ364968 blaRTG-5 Beta-lactamase Acinetobacter baumannii - + - + + + X58272.1.gene1 X58272 blaOXA-5 Beta-lactamase Pseudomonas aeruginosa - - - - - + X64695.1.gene9 X64695 ermB Macrolides Streptococcus pyogenes - + - - - - NC_002516.2.882884 NC_002516 mexF Multidrug Efflux Pseudomonas aeruginosa - - - - + + (MLS)MphE:JF769133:8777-9661:885 JF769133 mph(E) Macrolides Pasteurella multocida - + + + + + (MLS)MsrE:JF769133:7246-8721:1476 JF769133 msr(E)_1 Macrolides Pasteurella multocida - - + + + + EF102240.1.gene3 EF102240 msr(E)_2 Macrolides Acinetobacter baumannii - - - - + - NC_007682.3.4246769 NC_007682 msr(E)_3 Macrolides Escherichia coli - - - - - + NC_010481.6155782 NC_010481 msr(E)_4 Macrolides Acinetobacter baumannii - - - - + + CP003022.1.gene337 CP003022 msr(E)_5 Macrolides Pasteurella multocida - - - - - + AL123456.3.gene715 AL123456 rpsL_1 Aminoglycoside Mycobacterium tuberculosis - - - + - - CP003248.2.gene711 CP003248 rpsL_2 Aminoglycoside Mycobacterium tuberculosis - - - + - + (AGly)StrA:AB366441:22458-23261:804 AB366441 strA Aminoglycoside Salmonella enterica - - + - - - (Sul)SulI:AF071413:6700-7539:840 AF071413 sulI_1 Sulfonamide Escherichia coli + + + + + + AJ223604.1.gene9 AJ223604 sulI_2 Sulfonamide Pseudomonas aeruginosa - - + - - - AY743590.gene AY743590 tet(39) Tetracycline Acinetobacter LUH5605 - - - + + - CP002695.1.gene3614 CP002695 tuf_1 Elfamycins Bordetella pertussis + + + + + + NC_011595.7072242 NC_011595 tuf_2 Elfamycins Acinetobacter baumannii - - + + + +   16 Table 3.2.  CP002695.1.gene10 CP002695 tuf_3 Elfamycins Bordetella pertussis - + + + + + CP000647.1.gene3761 CP000647 tuf_4 Elfamycins Klebsiella pneumoniae - - - - + - NC_002516.2.881718 NC_002516 tufA Elfamycins Pseudomonas aeruginosa - - - - + + NC_002516.2.881697 NC_002516 tufB Elfamycins Pseudomonas aeruginosa - - + - + + 17 E.coli, Acinetobacter baumannii, Pasteurella multocida, and Pseudomonas aeruginosa composed the majority of antibiotic resistant bacteria (ARB) in AD samples (Table 3.2). Resistant gene msr(E)_2 belonging to Acinetobacter baumannii possessed a 97% gene identify in the AD sample for CAS treatment plant during Bowtie2 alignment, but was not detected in the MBR sample. Sulfonamide resistant gene, sul1, belonging to E.coli conferred a 98% (97%) and 98% (98%) gene identity for CAS and MBR after disinfection samples, respectively (Table A.9  A.10).      Beta-lactamase (RTG-5) belonging to Acinetobacter baumannii conferred a 99% gene identity in AD samples for the CAS treatment plant with an average depth of 1.27 and 1.36x in Bowtie2 and BWA-MEM alignment processes, respectively. While blaRTG-5 resistant gene went undetected in the MBR AD sample during Bowtie2 alignment, a 94% similarity was found during alignment with BWA-MEM. Pasteurella multocida resistant gene, mph(E), possessed a 98% (93%) and 91% (98%) identity to samples from the CAS and MBR AD samples, respectively, during alignment with Bowtie2 (BWA-MEM). After alignment with BWA-MEM, AD samples from CAS and MBR conferred a 100% gene identity with an average depth of 3.51 and 2.56, respectively to aadA6 resistant gene belonging to Pseudomonas aeruginosa (Table A.9  A.10). 3.6. Statistical Assessment using MAPQ Scores generated from Bowtie2 and BWA-MEM Alignments  MAPQ scores were extracted from Bowtie2 and BWA alignment from their respective SAM file. BWA-MEM produced MAPQ scores ranging from 0 to 60 (p<0.001). Average MAPQ scores generated during alignment with the CARD database were 18.65 (p=0.01) and 15.76 (p=0.03) for CAS and MBR samples, respectively. Alignment with the ARG-ANNOT databases revealed an average MAPQ score of 55.89 (p<0.001) and 59 (p<0.001) for CAS and MBR samples, respectively.  18 The lowest score reported for Bowtie2 alignment was 0 (p=1) with the highest being 255 (MAPQ score not available). The MAPQ score of selected sequences in Bowtie2 was reported as 255 for the majority of unique reads (reads with only one sequence alignment reported). Multi-reads (reads with multiple alignments) reported in Bowtie2 generally contained a score of either 0 or 1 (p>0.5) depending on the alignment score of the best alignment, and the difference between the best and second best alignment scores. Due to the absence of usable MAPQ scores for unique reads for Bowtie2, its base-line sequence mapping quality could not be fully assessed in this study.                   19 4. DISCUSSION   4.1. Occurrence of ARGs in WWTPs  Here, we use a sequence-based metagenomic approach to investigate the prevalence of a broad spectrum of ARGs in a conventional activated sludge WWTP with chlorine disinfection, and membrane bioreactor with UV disinfection.  In this study, genes resistant to tetracycline, sulfonamide, -lactam, aminoglycoside, elfamycin, and macrolide classes of antibiotics along with a multidrug efflux transporter (mexF) were detected in both CAS and MBR WWTPs.    Tetracycline and sulfonamide resistant genes are among the most common ARGs found in wastewater treatment systems, and other aquatic environments due to their widespread use (Gao et al., 2012a; Mao et al., 2015; Zhang et al., 2009a; Zhang et al., 2015a). Tet(39), and sulI were detected in BD and AD samples. Tetracycline and sulfonamide resistant genes are known to persist during wastewater treatment (Mao et al., 2015; Szczepanowski et al., 2009; Vaz-Moreira et al., 2015). The ability of these genes to withstand various treatment processes greatly increases the risks of these genes contaminating natural water bodies. Mao et al., (2015) evaluated the profile of 30 different ARGs in WWTPs in China and revealed that tet and sul1 genes were enriched in the final effluent of both selected WWTPs. Another study, conducted between hospital effluent, and the influent and effluent of an urban WWTP, revealed higher concentrations of sulfonamide resistant genes in the treated wastewater and hospital effluent compared to the raw sewage entering the WWTP (Vaz-Moreira et al., 2015). In contrast, Munir et al., 2011 observed a reduction in tetO, tetW, and sul1 resistant genes in the effluent of five different WWTPs in Michigan containing various treatment types. Another study, conducted on tetracycline resistant genes, tetA and tetC, in different water environments, found that the concentration of both tetracycline resistant genes were reduced in the final effluent of a sewage treatment plant in China (Zhang et al., 2009a).  20 Theses discrepancies indicate that antibiotic resistance selection is dependent on factors beyond that of solely wastewater treatment.  Class A, RTG-5 (CARB-14), beta-lactamase, conferring resistance to carbenicillin (Bonnin et al., 2012; Couture et al., 1992), was detected in all MBR samples and in the AD sample of the CAS WWTP. Class D, OXA-5, beta-lactamase, conferring resistance to oxacillin (Couture et al., 1992), was detected in the MBR AD sample. A primary mechanism behind -lactamase resistance, generally for gram-negative bacteria, is -lactamase enzymes (Couture et al., 1992; Xia et al., 2016). Beta-lactams are a widely used antibiotic with wide occurrences of resistance within the microbial community due to its extensive use and low toxicity (Zhang et al., 2009b). A comprehensive study on bacterial plasmids isolated from the activated sludge and final effluent of a WWTP in Germany also -lactamase hydrolyzing enzyme, OXA-5 (Genbank accession no. JQ364968) present in AS and final effluent samples as well as resistance to several other class A--lactamases (Szczepanowski et al., 2009).   Multidrug efflux transporter, mexF, was detected in AD samples of both CAS and MBR treatment plants. The multidrug resistance pump, mexF, is a part of the resistance-nodulation-cell division (RND) family, encoding resistance in Pseudomonas aeruginosa to antibiotics such as, chloramphenicol, trimethoprim, and fluoroquinolones. The efflux pump is said to consist of three proteins, which collaborate to expel antibiotics from the cytoplasm or periplasm of the microorganism (Aires et al., 2002). A number of studies have reported RND resistance pumps in aquatic environments (Gomez-Alvarez et al., 2012; Szczepanowski et al., 2009).  Aminoglycosides, aadA6 and rpsL, conferring resistance to streptomycin and spectinomycin (Finken et al., 1993; Papadovasilaki et al., 2015) were also present in this study. Aminoglycosides  21 consist of three groups, phosphotransferases (APHs), acetyltransferases (AACs), and the adenyltransferases (ANTs) (Wright, 1999). They are used against a wide range of aerobic bacteria, including Mycobacterium tuberculosis (Fonseca et al., 2015). Streptomycin was the first antibiotic used to treat TB (Honore et al., 1994). M.tuberculosis bacteria resistant to this drug are often classified as extensively drug-resistant strains (Fonseca et al., 2015). Mutations in the ribosomal protein S12 encoding gene, rpsL, is said to cause resistance to streptomycin in M.tuberculosis (Finken et al., 1993). Aminoglycoside resistant genes have become increasingly common in water environments (Shi et al., 2013; Szczepanowski et al., 2009) and have been reported in the effluent of treated wastewater in several instances (Szczepanowski et al., 2009; Vaz-Moreira et al., 2015).   Macrolide and elfamycin resistant genes were the most abundant genes detected in this study with resistant genes occurring in all six samples. After disinfection samples produced the greatest abundance of these genes compared to AS and BD samples. Genes resistant to macrolide classes of antibiotics composed 30.43% (7 out of 23) of all prevalent ARGs detected in this study. Macrolide resistant genes msr(E) and mph(E), detected in this study, encode a macrolide efflux pump, and macrolide-inactivating phosphotransferase, respectively (Rose et al., 2012). Rose et al., (2012) revealed that msr(E) and mph(E) were consistently expressed in tandem when detected in Pasteurella multocida, which is revealed in one instance in our study. Erythromycin resistant gene ermB, belonging to macrolide resistant classes of antibiotics, was detected in the AS sample from the MBR WWTP. A recent study, conducted on a bench-scale activated sludge treatment process, observed a survival rate of 100% among bacteria that were highly resistance to erythromycin in the effluent of the activated sludge process. It also revealed that the proportions of ermB remained constant in the activated sludge process (Guo et al., 2015). In addition to this, erythromycin  22 resistant genes are known to persist in the final effluent of WWTPs (Berglund et al., 2015; Sidrach-Cardona et al., 2014).   Similar to macrolide ARGs, elfamycin resistant genes were the second most prevalent genes detected in this study composing 26.09% (6 out of 23) of ARGs detected. Elfamycins are a class of naturally occurring antibiotics, consisting of antibiotics such as Kirromycin, Aurodox, and Efrotomycin (Hall et al., 1989; Sottani et al., 1993; Miele et al., 1994). Elfamycins inhibit bacterial growth by binding to the elongation factor Tu polypeptide, a component responsible for bacterial protein synthesis (Sottani et al., 1993). To our knowledge, no recent studies on the prevalence of elfamycin resistant genes in sewage treatment plants have been documented. Sequence-based metagenomics has revealed the occurrence of these genes in all treatment stages with reports of 100% gene similarity to the reference gene. 4.2. Occurrence of ARGs in Various Wastewater Treatment Stages   Activated sludge samples from both WWTPs attained the lowest occurrence of ARGs compared to before disinfection and after disinfection samples with no notable difference between the two biological treatment processes. It has been reported that biological treatment types, loading rates, sludge retention times and the production of biosolids, to name a few, impact the presence of resistant genes (Kim et al., 2010; Munir et al., 2011; Novo and Manaia, 2010) in biological treatment processes. Kim et al. (2010) reported the occurrence of tetracycline resistant genes in three WWTPs located in New York and Connecticut, with varying biological treatment functions. The study revealed that the AS processes did not contribute significantly to the enrichment of tetracycline resistant genes, and found that the fraction of these genes were lower in AS processes compared to the before disinfection processes. This was consistent with our study, which revealed a systematic trend of increasing occurrence of ARGs as treatment progressed.   23 The low profile of ARGs in the AS process could be the result of several factors such as, nutrient abundance, and adsorption to sludge biosolids. Munir et al. (2011) reported an increase in the concentration of tet and sul ARGs studied in the biosolids of five different WWTPs by several orders of magnitude relative to before and after disinfection samples. These ARGs can potentially be exposed to the natural environment via land applications causing further exposure and spread of antibiotic resistance (Schmieder and Edwards 2012). A study conducted by Novo and Manaia (2010) documented the prevalence of ARGs in three different WWTPs with varying biological treatment processes. The study revealed that ARGs resistant to tetracycline persisted in the treated effluent of the conventional activated sludge WWTP compared to the trickling filter, and submerged aeration filter facilities. The mechanisms behind the prevalence of ARGs after biological treatment systems are still unclear. Further studies are needed to assess various operational, functional, and environmental factors that contribute to the composition of resistant genes in wastewater microbial communities.  Furthermore, a systematic increase was observed in the profile of ARGs from AS to BD samples for both WWTPs. There was an increase in genes conferring resistance to elfamycins, and the occurrence of tet(39), strA, and rpsL resistant genes in BD samples for each WWTP. Similar to our results, other studies have reported an increase in antibiotic resistant genes like tetracycline and macrolides in before disinfection samples (Guo et al., 2015; Kim et al. 2010) suggesting that resistant genes arose either in the effluent of the biological treatment processes, or during secondary clarification.   After disinfection samples revealed the largest occurrence of ARGs in both CAS and MBR WWTPs. A significant number of the 23 total ARGs 24 similarity, were not detected until after disinfection in both CAS and MBR treatment plants, 56.25% and 53.33%, respectively.  Consistent with recent studies, a number of disinfection processes fail to be effective against reducing ARGs, and in some cases, an increase in the occurrence of these genes was observed in the treated effluent (Huang et al., 2011; Murray et al., 1984; Shi et al., 2013; Yuan et al., 2015). A previous study revealed a significant increase in ARGs in water sampled from a chlorine disinfection tank as opposed to the raw source water of a water treatment plant in Nanjing, China. The results also revealed a significant increase in mobile genetic elements in the chlorinated water compared to the raw water sources (Shi et al., 2013). Other studies reported that chlorine disinfection had little effect on reducing the presence of ARGs in the effluent of WWTPs (Yuan et al., 2015; Munir et al., 2011). On the other hand, several studies have reported a decrease in the presence of ARGs after chlorination (Huang et al., 2011; Mao et al., 2015; Murray et al., 1984).   Similar to chlorine disinfection, recent studies detected an increase in antibiotic resistance after UV disinfection (Kim et al., 2010). In our study, there was an increase in detected ARGs after UV radiation treatment, but there was no significant difference between the occurrences of ARGs following chlorination compared to UV.  4.3. Composition of ARBs in Chlorine and UV Treated Wastewater    Acinetobacter baumannii, Pseudomonas aeruginosa, and Pasteurella multocida composed the majority of antibiotic resistant bacteria found in AD samples in both WWTPs. According to the CARD and ARG-ANNOT resistance gene ontologies (ARO), a number of prevalent ARGs found in this study: including macrolides, elfamyins, multidrug, aminoglycoside -lactamase resistant genes, belong generally to the abovementioned gram-negative bacteria. A. baumannii is an opportunist pathogen responsible for a number of nosocomial infections and reported outbreaks  25 within recent years (Zarrilli et al., 2008). Since the rise of its antibiotic resistant characteristics in the 1960s, A. baumannii has since been characterized as a multidrug resist bacteria known to develop resistance within the first year of introducing a new antibiotic for treatment (Gonzalez-Villoria and Valverde-Garduno, 2016).  Similar to A. baumannii, Pseudomonas aeruginosa is an opportunistic hospital-acquired pathogen commonly found in moist, high nutrient environments such as WWTPs (Slekovec et al., 2012). P. aeruginosa accounts for 15% of hospital acquired infections to date, mainly attacking intensive care units. These pathogenic bacteria also play a role in community-acquired infections due to exposure to contaminated water sources (Slekovec et al., 2012). Slekovec et al., 2012 found the presence of P. aeruginosa to be most abundant in water extracted from the effluent of a hospital waste stream and the treated sludge from a WWTP, than the other environments sampled. Furthermore, the study revealed the appearance of multidrug resistant strains of P.aeruginosa in the hospital wastewater, treated wastewater, and receiving river water downstream the WWTP. The occurrence of these types of resistant opportunistic pathogens in the environment is of increasing importance concerning human health (Bouki et al., 2013; Varela and Manaia, 2013).   Pasteurella multocida was also reported in this study and persisted in each of the treatment stages in the MBR WWTP. P. multocida is a pathogenic bacterium occurring in the natural oral flora of several species of animals (Weber et al., 1984). P. multocida are commonly classified as zoonotic pathogens known to cause various infectious diseases in both animals and humans (Rose et al., 2012; Weber et al., 1984). Human infections usually occur as a result of an animal bite causing a number of diseases including meningitis and pneumonia (Weber et al., 1984). Reports  26 have classified several antibodies that are no longer effective against P. multocida due to its reduced susceptibility to the drug (Rose et al., 2012; Sellyei et al., 2009; Weber et al., 1984).   Other bacteria such as E.coli, Bordetella pertussis, Klebsiella pneumonia, and streptomycin resistant Mycobacterium tuberculosis were also reported in AD samples.  Several studies reveal significant concentrations of antimicrobial resistant pathogenic bacteria in treated wastewater despite reductions of these bacteria throughout the treatment process (Everage et al., 2014; Munir et al., 2011; Slekovec et al., 2012). These studies indicate that resistant bacteria are disseminating into natural water bodies despite treatment (Gao et al., 2012a).   4.4. Statistical Analysis of Sequence Alignment using MAPQ Scores  Genetic alignment of AS, BD, and AD samples from each treatment utility was performed using Bowtie2 and BWA-MEM, which have proven to be accurate sequence aligners (Langmead and Salzberg, 2012; Li, 2013). Sequence-based metagenomics is most accurate when identifying known genes (Schmieder and Edwards, 2012), therefore ARGs found in all samples using metagenomic analysis only highly suggest the presence of these ARGs in this study. ARGs with a gene similarity less than 100% should be interpreted with caution as they could belong to some variation of the known reference gene, or novel resistance determinant (Schmieder and Edwards, 2012). Despite vast enhancements in the efficiency and accuracy of genetic aligners, they still have their limitations. Genetic alignment is exposed to a number of performance errors. For example, insertion or deletions (indels) generated during sequencing runs, as well as duplicate regions in the reference genome can lead to incorrect mapping during the alignment process (Li and Homer, 2010; Olson et al., 2015). Most sequence aligners generate sequence mapping quality scores for users to gage the confidence level of aligned sequences (Li and Homer, 2010; Ruffalo et al., 2011).    27 This study provided a baseline assessment of the unaltered sequence alignment errors rates  MAPQ scores when operating under default conditions. It was found that BWA-MEM produced more useful MAPQ scores compared to Bowtie2 (Ruffalo et al., 2011) and the number of sequences passing threshold requirements was greatest in BWA-MEM when aligned with either reference database for AD samples.   Bowtie2 MAPQ scores for high quality, unique-read alignments were not accounted for when aligned against either reference database, while multi-read alignments generally received a score of either 0 or 1 depending on Bowtie2 l (Langmead and Salzberg, 2012). Thus, MAPQ scores for Bowtie2 could not be assessed without further analysis. The integrity of MAPQ scores is not only important in evaluating the accuracy of genetic alignment, but also for downstream applications such as variant analysis (Li and Homer, 2010; Olson et al., 2015). Therefore, it is a best practice to realign MAPQ scores for accuracy before proceeding to downstream applications (Olson et al., 2015).             28 5. CONCLUSION  The rise of antibiotic resistance genes in the environment stimulated the need to study the development of these genes in favorable environments such as WWTPs. Thus, our study aimed to explore the profile and occurrence of these genes in WWTPs using a sequenced-based metagenomic approach. This study allowed for the detection of a broad spectrum of ARGs in the activated sludge, before disinfection and after disinfection treatment stages of two WWTPs in Michigan. The results reveal a systematic increase in the number of ARGs as treatment progressed suggesting that wastewater treatment promotes the development of antimicrobial resistant genes. On the basis of conflicting studies, the selection for ARGs in WTTPs is still not fully understood. However, it is evident that a number of factors impact the selection or removal of ARGs in sewage treatment systems. Our study further contributes to the pool of information on the dissemination of ARGs in WWTPs given various operational conditions and treatment stages. Further studies, assessing a wide range of functional, operational, and environmental factors, are needed to fully understand the behavior of these genes in treatment plants and in the environment.          29             APPENDIX              30 Table A.1. Bowtie2 alignment statistics per sample    Table A.2. Bowtie2 alignment statistics per sample              31 Table A.3. BWA-MEM alignment statistics per sample    Table A.4. BWA-MEM alignment statistics per sample              32 Table A.5. Average depth and gene similarity with Bowtie2 alignment for ARG% coverage in activated sludge (AS) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one data set.   Database Accession No. Gene  Length GC% Avg Depth (x) Depth Std Gene Similarity (%) ELAS TCAS ELAS TCAS ELAS TCAS CP002695.1.gene10 tuf 1191 62.64 1.65 1.77 1.43 1.80 82.28 66.16 CP002695.1.gene3614 tuf 1191 62.64 0.52 2.09 0.50 2.31 52.98 88.58 (Sul)SulI:AF071413:6700-7539:840 sulI 855 60.58 7.91 6.57 3.00 4.14 98.25 96.14 (MLS)MphE:JF769133:8777-9661:885 mphE 900 35.78 0.40 3.65 0.49 1.47 40.78 96.33 JQ364968.1.gene6 blaRTG-5 378 43.12 - 1.15 - 0.85 ND 70.37 X64695.1.gene9 ermB 111 31.53 - - - - ND ND   Table A.6. Average depth and gene similarity with BWA-samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one data set.   Database Accession No. Gene Length GC% Avg Depth (x) Depth Std Gene Similarity (%) ELAS TCAS ELAS TCAS ELAS TCAS CP002695.1.gene10 tuf 1191 62.64 4.16 14.90 3.66 8.85 81.95 99.50 CP002695.1.gene3614 tuf 1191 62.64 5.58 15.91 3.65 7.81 96.31 100 (Sul)SulI:AF071413:6700-7539:840 sulI 855 60.58 0.04 0.87 0.21 0.76 4.44 74.62 (MLS)MphE:JF769133:8777-9661:885 mphE 900 35.78 - 1.28 - 0.77 ND 80.56 JQ364968.1.gene6 blaRTG-5 378 43.12 - 2.80 - 1.66 ND 100 X64695.1.gene9 ermB 111 31.53 - 1.01 - 0.09 ND 100   33 Table A.7. Average depth a90% coverage in before disinfection (BD) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one data set.   Database Accession No. Gene  Length GC% Avg Depth (x) Depth Std Gene Similarity (%) ELBD TCBD ELBD TCBD ELBD TCBD AY743590.gene tet39 1188 40.57 1.32 2.36 1.34 0.89 64.81 94.61 (AGly)StrA:AB366441:22458-23261:804 strA 818 55.26 2.10 0.37 1.23 0.48 97.68 36.67 (Sul)SulI:AF071413:6700-7539:840 sulI 855 60.58 10.94 5.89 5.21 2.63 97.43 93.68 (MLS)MphE:JF769133:8777-9661:885 mph(E) 900 35.78 2.87 6.33 1.73 3.43 95.78 95.89 (MLS)MsrE:JF769133:7246-8721:1476 msr(E) 1501 39.31 5.05 6.31 2.67 2.91 94.47 94.74 CP002695.1.gene10 tuf 1191 62.64 1.27 1.74 1.29 2.32 75.06 73.97 CP002695.1.gene3614 tuf 1191 62.64 0.95 2.31 1.14 2.31 45.59 64.74 NC_002516.2.881697 tuf 1194 59.63 1.48 0.75 1.52 0.95 58.71 45.06 AJ223604.1.gene9 sulI 348 50 - - - - ND ND NC_011595.7072242 tuf 705 45.11 0.9872 1.18 1.10 1.54 49.08 44.54 JQ364968.1.gene6 blaRTG-5 378 43.12 - 1.18 - 0.88 ND 68.25 AL123456.3.gene715 rpsL 375 62.67 - - - - ND ND           34 Table A.8. Average depth and gene similarity with BWA-MEM alignment for ARGs with 90% coverage in before disinfection (BD) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one data set.   Database Accession No. Gene  Length GC% Avg Depth (x) Depth Std Gene Similarity (%) ELBD TCBD ELBD TCBD ELBD TCBD AY743590.gene tet39 1188 40.57 1.16 2.26 1.21 0.92 62.79 100 (AGly)StrA:AB366441:22458-23261:804 strA 818 55.26 0.53 0.54 0.83 0.55 31.30 51.34 (Sul)SulI:AF071413:6700-7539:840 sulI 855 60.58 3.21 2.93 1.93 1.87 93.92 89.82 (MLS)MphE:JF769133:8777-9661:885 mph(E) 900 35.78 1.18 3.86 0.66 2.15 86.33 98.33 (MLS)MsrE:JF769133:7246-8721:1476 msr(E) 1501 39.31 0.78 1.97 0.99 1.40 46.50 81.81 CP002695.1.gene10 tuf 1191 62.64 23.59 28.30 13.96 16.49 100 100 CP002695.1.gene3614 tuf 1191 62.64 25.95 32.10 14.64 17.13 97.90 100 NC_002516.2.881697 tuf 1194 59.63 18.05 20.28 15.16 17.41 93.80 89.28 AJ223604.1.gene9 sulI 348 50 1.14 0.60 0.55 0.56 90.80 56.90 NC_011595.7072242 tuf 705 45.11 5.39 3.69 4.92 2.35 90.21 90.35 JQ364968.1.gene6 blaRTG-5 378 43.12 0.47 1.95 0.50 0.45 47.62 99.74 AL123456.3.gene715 rpsL 375 62.67 2.37 4.61 2.85 3.43 57.87 99.20           35 Table A.9. 90% coverage in after disinfection (AD) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one data set.   Database Accession No. Gene Length GC% Avg Depth (x) Depth Std Gene Similarity (%) ELAD TCAD ELAD TCAD ELAD TCAD (Sul)SulI:AF071413:6700-7539:840 sul1 855 60.58 12.05 5.40 6.23 2.06 97.89 98.25 AM087411.1.gene3 aadA6 846 53.66 2.59 0.68 1.60 0.68 81.44 56.50 AY743590.gene tet39 1188 40.57 1.23 1.85 0.59 1.17 93.60 83.08 CP002695.1.gene3614 tuf 1191 62.64 0.97 2.02 0.95 1.31 64.90 95.05 EF102240.1.gene3 msrE 1476 39.97 2.01 1.69 0.79 1.58 97.83 75.14 (MLS)MphE:JF769133:8777-9661:885 mphE 900 35.78 6.36 5.00 2.41 2.79 98.33 90.78 (MLS)MsrE:JF769133:7246-8721:1476 msrE 1501 39.31 9.53 9.06 4.26 3.57 98.33 97.14 JQ364968.1.gene6 blaRTG-5 378 43.12 1.27 0.53 0.46 0.50 99.21 53.44 NC_002516.2.881697 tufB 1194 59.63 4.54 1.49 4.16 2.29 87.02 43.30 NC_007682.3.4246769 msrE 1476 39.97 2.46 2.10 1.45 1.53 85.98 92.21 NC_010481.6155782 msrE 1476 39.97 1.57 2.59 1.24 1.65 71.34 91.80 NC_011595.7072242 tuf 705 45.11 0.25 1.00 0.43 1.14 24.82 51.49 CP002695.1.gene10 tuf 1191 62.64 0.74 1.24 1.22 1.04 43.07 75.82 NC_002516.2.882884 mexF 3189 65.57 0.52 0.21 1.06 0.47 27.34 22.42 CP000647.1.gene3761 tuf 1185 54.09 0.64 0.54 1.38 0.72 18.65 40.68 NC_002516.2.881718 tufA 1194 59.63 0.48 0.53 1.01 0.97 22.11 28.06 CP003248.2.gene711 rpsL 375 62.67 - - - - ND ND CP003022.1.gene337 msrE 1476 39.97 1.72 1.37 1.79 1.07 64.16 75.07 X58272.1.gene1 blaOXA-5 804 40.17 0.22 0.60 0.42 0.69 22.14 49.13      36 Table A.10. Average depth and gene similarity with BWA-0% coverage in after disinfection (AD) samples for East Lansing (CAS) and Traverse City (MBR) WWTPs. ARGs from both CARD and ARG-ANNOT reference database are integrated into one data set.   Database Accession No. Gene Length GC% Avg Depth (x) Depth Std Gene Similarity (%) ELAD TCAD ELAD TCAD ELAD TCAD (Sul)SulI:AF071413:6700-7539:840 sul1 855 60.58 2.66 2.25 1.32 0.87 97.19 98.33 AM087411.1.gene3 aadA6 846 53.66 3.51 2.56 4.25 1.19 100 100 AY743590.gene tet39 1188 40.57 1.12 0.82 0.86 0.83 82.15 55.96 CP002695.1.gene3614 tuf 1191 62.64 24.15 47.95 14.13 26.90 100 100 EF102240.1.gene3 msrE 1476 39.97 1.64 1.01 1.18 2.31 78.46 41.02 (MLS)MphE:JF769133:8777-9661:885 mphE 900 35.78 1.55 2.61 0.70 1.07 92.56 98.25 (MLS)MsrE:JF769133:7246-8721:1476 msrE 1501 39.31 1.34 3.05 1.20 1.31 71.55 96.67 JQ364968.1.gene6 blaRTG-5 378 43.12 1.36 1.89 0.50 1.17 99.21 93.90 NC_002516.2.881697 tufB 1194 59.63 24.97 34.68 14.62 24.84 96.40 92.55 NC_007682.3.4246769 msrE 1476 39.97 1.14 1.89 0.64 1.17 86.79 93.90 NC_010481.6155782 msrE 1476 39.97 1.40 0.82 0.75 0.83 91.26 55.96 NC_011595.7072242 tuf 705 45.11 10.41 4.54 10.43 4.52 100 90.78 CP002695.1.gene10 tuf 1191 62.64 21.20 45.58 13.25 26.59 98.74 100 NC_002516.2.882884 mexF 3189 65.57 10.59 13.75 11.47 13.03 94.45 91.66 CP000647.1.gene3761 tuf 1185 54.09 4.23 5.73 2.35 5.48 93.67 75.87 NC_002516.2.881718 tufA 1194 59.63 16.57 26.85 15.49 27.23 93.30 90.62 CP003248.2.gene711 rpsL 375 62.67 3.49 3.15 3.46 4.04 91.20 55.73 CP003022.1.gene337 msrE 1476 39.97 1.17 2.66 0.84 1.53 78.66 96.27 X58272.1.gene1 blaOXA-5 804 40.17 0.22 0.60 0.42 0.65 22.14 91.17     37  REFERENCES  38 REFERENCES   Aires, J. R., Pechere, J. C., Van Delden, C., & Köhler, T. (2002). Amino acid residues essential for function of the MexF efflux pump protein of Pseudomonas aeruginosa. Antimicrobial agents and chemotherapy, 46(7), 2169-2173.  Andrews, S. (2010). FastQC: A quality control tool for high throughput sequence data. Reference Source. Berglund, B., Fick, J., & Lindgren, P. E. (2015). Urban Wastewater Effluent Increases Antibiotic Resistance Gene Concentrations in a Receiving Northern European River. Environmental Toxicology and Chemistry, 34(1), 192-196. Bolger, A. M., Lohse, M., & Usadel, B. (2014). Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics, btu170. Bondarczuk, K., Markowicz, A., & Piotrowska-Seget, Z. (2016). The urgent need for risk assessment on the antibiotic resistance spread via sewage sludge land application. Environment International, 87, 49-55. Bonnin, R. A., Poirel, L., & Nordmann, P. (2012). A novel and hybrid composite transposon at the origin of acquisition of bla RTG-5 in Acinetobacter baumannii. International journal of antimicrobial agents, 40(3), 257-259. Bouki, C., Venieri, D., & Diamadopoulos, E. (2013). Detection and fate of antibiotic resistant bacteria in wastewater treatment plants: A review. Ecotoxicology and Environmental Safety, 91, 1-9. Couture, F., Lachapelle, J., & Levesque, R. C. (1992). Phylogeny of LCR1 and OXA5 with class A and class D lactamases. Molecular microbiology, 6(12), 1693-1705. Diehl, D. L., & LaPara, T. M. (2010). Effect of temperature on the fate of genes encoding tetracycline resistance and the integrase of class 1 integrons within anaerobic and aerobic digesters treating municipal wastewater solids. Environmental science & technology, 44(23), 9128-9133. Everage, T. J., Boopathy, R., Nathaniel, R., LaFleur, G., & Doucet, J. (2014). A survey of antibiotic-resistant bacteria in a sewage treatment plant in Thibodaux, Louisiana, USA. International Biodeterioration & Biodegradation, 95, 2-10. Fonseca, J. D., Knight, G. M., & McHugh, T. D. (2015). The complex evolution of antibiotic resistance in Mycobacterium tuberculosis. International Journal of Infectious Diseases, 32, 94-100.   39 Finken, M., Kirschner, P., Meier, A., Wrede, A., & Böttger, E. C. (1993). Molecular basis of streptomycin resistance in Mycobacterium tuberculosis: alterations of the ribosomal protein S12 gene and point mutations within a functional 16S ribosomal RNA pseudoknot. Molecular microbiology, 9(6), 1239-1246. Gao, P., Mao, D., Luo, Y., Wang, L., Xu, B., & Xu, L. (2012a). Occurrence of sulfonamide and tetracycline-resistant bacteria and resistance genes in aquaculture environment. Water Research, 46(7), 2355-2364. Gao, P., Munir, M., & Xagoraraki, I. (2012b). Correlation of tetracycline and sulfonamide antibiotics with corresponding resistance genes and resistant bacteria in a conventional municipal wastewater treatment plant. Science of the Total Environment, 421, 173-183. Gomez-Alvarez, V., Revetta, R. P., & Santo Domingo, J. W. (2012). Metagenomic analyses of drinking water receiving different disinfection treatments. Applied and environmental microbiology, 78(17), 6095-6102. Gonzalez-Villoria, A. M., & Valverde-Garduno, V. (2016). Antibiotic-Resistant Acinetobacter baumannii Increasing Success Remains a Challenge as a Nosocomial Pathogen. Journal  Guo, M. T., Yuan, Q. B., & Yang, J. (2015). Insights into the amplification of bacterial resistance to erythromycin in activated sludge. Chemosphere, 136, 79-85. Gupta, S. K., Padmanabhan, B. R., Diene, S. M., Lopez-Rojas, R., Kempf, M., Landraud, L., & Rolain, J. M. (2014). ARG-ANNOT, a new bioinformatic tool to discover antibiotic resistance genes in bacterial genomes. Antimicrobial agents and chemotherapy, 58(1), 212-220. Hall, C. C., Watkins, J. D., & Georgopapadakou, N. H. (1989). Effects of elfamycins on elongation factor Tu from Escherichia coli and Staphylococcus aureus. Antimicrobial agents and chemotherapy, 33(3), 322-325. Honore, N., & Cole, S. T. (1994). Streptomycin resistance in mycobacteria. Antimicrobial agents and chemotherapy, 38(2), 238-242. Huang, J. J., Hu, H. Y., Tang, F., Li, Y., Lu, S. Q., & Lu, Y. (2011). Inactivation and reactivation of antibiotic-resistant bacteria by chlorination in secondary effluents of a municipal wastewater treatment plant. Water Research, 45(9), 2775-2781. Kim, S., Park, H., & Chandran, K. (2010). Propensity of activated sludge to amplify or attenuate tetracycline resistance genes and tetracycline resistant bacteria: A mathematical modeling approach. Chemosphere, 78(9), 1071-1077. Kristiansson, E., Fick, J., Janzon, A., Grabic, R., Rutgersson, C., Weijdegård, B., ... & Larsson, D. J. (2011). Pyrosequencing of antibiotic-contaminated river sediments reveals high levels of resistance and gene transfer elements. PloS one, 6(2), e17038.   40 Langmead, B., & Salzberg, S. L. (2012). Fast gapped-read alignment with Bowtie 2. Nature methods, 9(4), 357-359. Li, H. (2013). Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. arXiv preprint arXiv:1303.3997. Li, H., & Homer, N. (2010). A survey of sequence alignment algorithms for next-generation sequencing. Briefings in bioinformatics, 11(5), 473-483. McArthur, A. G., Waglechner, N., Nizam, F., Yan, A., Azad, M. A., Baylay, A. J., ... & Kalan, L. (2013). The comprehensive antibiotic resistance database. Antimicrobial agents and chemotherapy, 57(7), 3348-3357. Manaia, C. M., Macedo, G., Fatta-Kassinos, D., & Nunes, O. C. (2015). Antibiotic resistance in urban aquatic environments: can it be controlled? Applied microbiology and biotechnology, 1-15. Mao, D., Yu, S., Rysz, M., Luo, Y., Yang, F., Li, F., ... & Alvarez, P. J. J. (2015). Prevalence and proliferation of antibiotic resistance genes in two municipal wastewater treatment plants. Water research, 85, 458-466. Miele, A., Goldstein, B. P., Bandera, M., Jarvis, C., Resconi, A., & Williams, R. J. (1994). Differential Susceptibilities of Enterococcal Species to Elfamycin Antibiotics. Journal of Clinical Microbiology, 32(8), 2016-2018. Molnar, M., & Ilie, L. (2014). Correcting illumina data. Briefings in bioinformatics, bbu029. Munir, M., Wong, K., & Xagoraraki, I. (2011). Release of antibiotic resistant bacteria and genes in the effluent and biosolids of five wastewater utilities in Michigan. Water research, 45(2), 681-693. Murray, G. E., Tobin, R. S., Junkins, B., & Kushner, D. J. (1984). Effect of chlorination on antibiotic resistance profiles of sewage-related bacteria. Applied and environmental microbiology, 48(1), 73-77. Novo, A., & Manaia, C. M. (2010). Factors influencing antibiotic resistance burden in municipal wastewater treatment plants. Applied Microbiology and Biotechnology, 87(3), 1157-1166. Olson, N. D., Lund, S. P., Colman, R. E., Foster, J. T., Sahl, J. W., Schupp, J. M., ... & Zook, J. M. (2015). Best practices for evaluating single nucleotide variant calling methods for microbial genomics. Frontiers in genetics, 6. Papadovasilaki, M., Oberthür, D., Gessmann, R., Sarrou, I., Betzel, C., Scoulica, E., & Petratos, K. (2015). Biophysical and enzymatic properties of aminoglycoside adenylyltransferase AadA6 from Pseudomonas aeruginosa. Biochemistry and Biophysics Reports, 4, 152-157.   41 Penders, J., Stobberingh, E. E., Savelkoul, P. H., & Wolffs, P. F. (2013). The human microbiome as a reservoir of antimicrobial resistance. Front Microbiol, 4(87), 542-549. Rose, S., Desmolaize, B., Jaju, P., Wilhelm, C., Warrass, R., & Douthwaite, S. (2012). Multiplex PCR to identify macrolide resistance determinants in Mannheimia haemolytica and Pasteurella multocida. Antimicrobial agents and chemotherapy, 56(7), 3664-3669. Ruffalo, M., LaFramboise, T., & Koyutürk, M. (2011). Comparative analysis of algorithms for next-generation sequencing read alignment. Bioinformatics, 27(20), 2790-2796. Schmieder, R., & Edwards, R. (2012). Insights into antibiotic resistance through metagenomic approaches. Future microbiology, 7(1), 73-89. Sellyei, B., Varga, Z., Szentesi-Samu, K., Kaszanyitzky, E., & Magyar, T. (2009). Antimicrobial Susceptibility of Pasteurella Multocida Isolated from Swine and Poultry. Acta Veterinaria Hungarica, 57(3), 357-367.  Shi, P., Jia, S., Zhang, X. X., Zhang, T., Cheng, S., & Li, A. (2013). Metagenomic insights into chlorination effects on microbial antibiotic resistance in drinking water. Water research, 47(1), 111-120. Sidrach-Cardona, R., Hijosa-Valsero, M., Marti, E., Balcazar, J. L., & Becares, E. (2014). Prevalence of antibiotic-resistant fecal bacteria in a river impacted by both an antibiotic production plant and urban treated discharges. Science of the Total Environment, 488, 220-227. Sims, D., Sudbery, I., Ilott, N. E., Heger, A., & Ponting, C. P. (2014). Sequencing depth and coverage: key considerations in genomic analyses. Nature Reviews Genetics, 15(2), 121-132. Slekovec, C., Plantin, J., Cholley, P., Thouverez, M., Talon, D., Bertrand, X., et al. (2012). Tracking Down Antibiotic-Resistant Pseudomonas aeruginosa Isolates in a Wastewater Network. Plos One, 7(12). Sottani, C., Islam, K., Soffientini, A., Zerilli, L. F., & Seraglia, R. (1993). Studies on the Interaction of Elfamycin Antibiotics with Elongation Factor-Tu by Mass Spectroscopic Techniques. Rapid Communications in Mass Spectrometry, 7(7), 680-683. Szczepanowski, R., Linke, B., Krahn, I., Gartemann, K. H., Guetzkow, T., Eichler, W., ... & Schlueter, A. (2009). Detection of 140 clinically relevant antibiotic-resistance genes in the plasmid metagenome of wastewater treatment plant bacteria showing reduced susceptibility to selected antibiotics. Microbiology, 155(7), 2306-2319. Varela, A. R., & Manaia, C. M. (2013). Human health implications of clinically relevant bacteria in wastewater habitats. Environmental Science and Pollution Research, 20(6), 3550-3569.   42 Vaz-Moreira, I., Varela, A. R., Pereira, T. V., Fochat, R. C., & Manaia, C. M. (2015). Multidrug Resistance in quinolone-resistant gram-negative bacteria isolated from hospital effluent and the municipal wastewater treatment plant. Microbial Drug Resistance. Weber, D. J., Wolfson, J. S., Swartz, M. N., & Hooper, D. C. (1984). Pasteurella-Multocida Infections - Report of 34 Cases and Review of the Literature. Medicine, 63(3), 133-154. Wright, G. D. (1999). Aminoglycoside-modifying enzymes. Current opinion in microbiology, 2(5), 499-503. Xia, J., Gao, J., & Tang, W. (2016). Nosocomial infection and its molecular mechanisms of antibiotic resistance. Bioscience trends, (0). Yuan, Q. B., Guo, M. T., & Yang, J. (2015). Fate of Antibiotic Resistant Bacteria and Genes during Wastewater Chlorination: Implication for Antibiotic Resistance Control. Plos One, 10(3). Zarrilli, R., Vitale, D., Di Popolo, A., Bagattini, M., Daoud, Z., Khan, A. U., ... & Triassi, M. (2008). A plasmid-borne blaOXA-58 gene confers imipenem resistance to Acinetobacter baumannii isolates from a Lebanese hospital. Antimicrobial agents and chemotherapy, 52(11), 4115-4120. Zhang, C. M., Du, C., Xu, H., Miao, Y. H., Cheng, Y. Y., Tang, H., et al. (2015a). Occurrence of tetracycline-resistant fecal coliforms and their resistance genes in an urban river impacted by municipal wastewater treatment plant discharges. Journal of Environmental Science and Health Part a-Toxic/Hazardous Substances & Environmental Engineering, 50(7), 744-749. Zhang, S. H., Han, B., Gu, J., Wang, C., Wang, P. F., Ma, Y. Y., et al. (2015b). Fate of antibiotic resistant cultivable heterotrophic bacteria and antibiotic resistance genes in wastewater treatment processes. Chemosphere, 135, 138-145. Zhang, T., Yang, Y., & Pruden, A. (2015c). Effect of temperature on removal of antibiotic resistance genes by anaerobic digestion of activated sludge revealed by metagenomic approach. Applied microbiology and biotechnology, 99(18), 7771-7779. Zhang, X., Wu, B., Zhang, Y., Zhang, T., Yang, L., Fang, H. H., ... & Cheng, S. (2009a). Class 1 integronase gene and tetracycline resistance genes tetA and tetC in different water environments of Jiangsu Province, China. Ecotoxicology, 18(6), 652-660. Zhang, X. X., & Zhang, T. (2011). Occurrence, abundance, and diversity of tetracycline resistance genes in 15 sewage treatment plants across China and other global locations. Environmental science & technology, 45(7), 2598-2604. Zhang, X. X., Zhang, T., & Fang, H. H. (2009b). Antibiotic resistance genes in water environment. Applied microbiology and biotechnology, 82(3), 397-414.