—....-,_.-. .. "70-" . . . . , .. . . .. ‘ . . . . v . . . .. . . . .. . A . V. ._. .. .. y. . . .. . . . .. ._ . ‘ .4 . .. .. . . .. . . . . V: . .. .. . ‘. . . . i T ._ , . . , . , . . _ . ‘y _ , ‘ . ,. .. .. . , . l A. . _ . . _ . v . . . . . pails. $4.91 MICHIGAN S 7/ NIVERSITY LIBRARI I, 1111 11111 111111 1211111111 3 129 005 $6865 LIBRARY Michigan State J-'__"lMiVflflnli{ay This is to certify that the dissertation entitled Characterization of Resistance to Tomato Mosaic Virus in Tomato Somaclones presented by Sandra Lynn Schiller Smith has been accepted towards fulfillment of the requirements for Ph.D. Botany & Plant Pathology degree in Major professor Date June 9, 1989 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Is An Affirmative Action/Equal Opportunity Institution CHARACTERIZATION OF RESISTANCE TO TOMATO MOSAIC VIRUS IN TOMATO SOMACLONES By Sandra Lynn Schiller Smith A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1989 C704 3/24~ ABSTRACT CHARACTERIZATION OF RESISTANCE TO TOMATO MOSAIC VIRUS IN TOMATO SOMACLONES BY Sandra Lynn Schiller Smith Six tomato somaclones from a tomato line that was fully susceptible to tomato mosaic virus (ToMV), were found to be resistant to ToMV multiplication. The purpose of the research of this dissertation was to characterize this resistance and determine its mode of inheritance. Screening for ToMV resistance was accomplished by inoculation of seedlings, incubation, and enzyme-linked immunosorbant assay (ELISA) for the detection of ToMV coat protein. The resistance was shown to be stably inherited, with nearly complete penetrance through several generations. Based on evidence from crosses and backcrosses to the parental line, the trait appeared to involve nuclear genes(s) that were incompletely dominant and showed a gene dosage dependence. Screening of progeny from reciprocal crosses indicated there was also a maternal effect in the inheritance of the trait. Because of this, several cytoplasmic components were investigated, but chloroplast DNA did not differ between susceptible and somaclonal resistant plant lines according to limited restriction endonuclease analyses, and no endogenous double-stranded RNA was found. In crosses with isogenic tomato lines expressing known resistance genes, it was shown that the somaclonal resistance was additive with Tm-l but not with Tm-Z. The type of resistance seen in the somaclonal lines was similar to that of the gene Tm-l in that it suppressed Sandra Lynn Schiller Smith symptom formation, limited virus multiplication, was not temperature sensitive and had similar virus strain responses. There was no resistance to tobacco etch virus or cucumber mosaic virus, thus a specificity to tobamoviruses was evident. Studies of virus movement in susceptible versus resistant plants indicated that transport of virus from the inoculated leaf was inhibited in the resistant plants. In grafting experiments, it was shown that virus multiplication was delayed in resistant scions grafted to susceptible, inoculated rootstocks. In the reciprocal graft, there was also a delay in virus multiplication in susceptible scions grafted onto resistant, inoculated rootstocks. In examining resistance at the cellular level, immunofluorescent staining of protoplasts released from inoculated resistant plants indicated restriction of virus titer and location. ACKNOWLEDGMENTS I would like to thank Dr. Harry Murakishi for his guidance and the kindness which he extended to me. I would like to express my appreciation for the work which was done by Kristen Barden Rorrer in producing the resistant tomato somaclones, and for the undergraduate students who helped in the assay of many tomato samples: Cathi Brandi, Susan Ruhl, Michael Collins, Todd Bridges, Rodney Cox, Kevin Fiddes and Rebecca Daly. I am grateful for the contributions of time and ideas from Randy Niedz (in counting chromosomes) and from Marcia Murry (in developing the techniques for organelle molecular genetics). I would also like to thank the members of my graduate committee, Rebecca Grumet, Joseph Saunders, Shauna Somerville and especially Barbara Sears, for their counsel and the time they devoted to thinking about these somaclones. I am grateful for the support and encouragement given by my family and friends, who made it possible to accomplish my goals. ii TABLE OF CONTENTS LIST OF TABLES ..................................................... vii LIST OF FIGURES ..................................................... xi LIST OF ABBREVIATIONS .............................................. xiv GENERAL INTRODUCTION ................................................. l ORIGIN OF SOMACLONES ............................................ 3 GENETICS ........................................................ 4 Evaluation of the plant material ............................. 4 VIROLOGY ........................................................ 6 Evaluation of the plant-virus interaction .................... 6 CHAPTER 1: GENETICS CHARACTERISTICS OF THE ToMV-RESISTANT SOMACLONES ................................................ 7 INTRODUCTION ......................................................... 7 LITERATURE REVIEW .................................................... 7 MATERIALS AND METHODS ............................................... 12 PLANT AND VIRUS MATERIAL ....................................... 12 Plants ...................................................... 12 Viruses ..................................................... 12 ARE SOMACLONES NORMAL PLANTS? .................................. 13 Cross morphology ............................................ 13 Determination of chromosome number .......................... 13 Seed germination rate ....................................... 14 Pollen viability ............................................ 14 SCREENING FOR RESISTANCE TO TMV ................................ 15 Screening of tomato somaclones .............................. 15 Purification of virus ....................................... 16 Enzyme-linked immunosorbant assay ........................... 17 Local lesion assay .......................................... 20 Detection of viral inclusions ............................... 21 iii Analysis of ELISA values .................................... 22 Cross-pollination of tomatoes ............................... 22 Seed collection ............................................. 23 GENETIC ANALYSIS ............................................... 24 Rating of resistance in tomato somaclones ................... 24 Crosses between somaclonal lines and susceptible tomato lines ................................................ 27 Seed transmission of virus .................................. 28 Crosses with known resistance genes ......................... 29 MOLECULAR GENETICS ............................................. 29 Purification and restriction analysis of chloroplast DNA (chNA) ................................................. 29 Gel electrophoresis ......................................... 31 Extraction and analysis of double-stranded RNA (dsRNA) ...... 32 Photography ................................................. 34 RESULTS AND DISCUSSION .............................................. 35 ARE SOMACLONES NORMAL PLANTS? .................................. 35 Cross morphology ............................................ 35 Chromosome number ........................................... 35 Seed germination rate ....................................... 35 Pollen viability ............................................ 35 SCREENING FOR RESISTANCE TO TMV ................................ 38 Suitability of resistance ratings ........................... 38 Heritability of resistance .................................. 41 Causes of incomplete penetrance ............................. 44 Influence of virus on phenotype ............................. 45 Chimerism ................................................... 49 Comparison of somaclonal lines .............................. 49 RESULTS OF SCREENING CROSSES ................................... 51 Results of reciprocal crosses to parental line .............. 52 Detection of reciprocal differences .................. 52 Significance of reciprocal differences for each somaclonal line ................................. 52 Are there differences among the somaclonal 1ines?....54 Estimation of resistance due to the maternal effect ...................................... 56 Were the original somaclones homozygous for the resistance trait? ................................ 59 Estimation of environmental effects .................. 61 Results of screening the segregating generation ............. 63 Detection of reciprocal differences .................. 63 Was there unintentional selection of resistant F2 plants ............................................ 65 Significance of reciprocal differences ............... 67 iv Are there differences among the somaclonal lines?....67 Estimation of resistance due to the maternal effect ............................................... 68 Were the original somaclones homozygous for the resistance trait? ................................ 68 Estimation of environmental effects .................. 74 Reciprocal crosses with Bonny Best .......................... 76 Backcrosses ................................................. 8O Detection of differences among backcrosses ........... 83 Are there differences among the somaclonal lines?....86 Distribution of ELISA values for backcrosses ......... 86 Summary of results of screening somaclonal line 247 for resistance to ToMV ...................................... 86 Summary of results of screening somaclonal line 12 for resistance to ToMV ...................................... 90 Seed transmission of ToMV ................................... 93 Results of crosses with known resistance genes .............. 94 MOLECULAR GENETICS ............................................. 98 Chloroplast DNA ............................................. 98 Double-stranded RNA ......................................... 99 GENETIC ANALYSIS .............................................. 102 Genetic models ............................................. 102 Model 1: Two additive genes plus a cytoplasmic effect ...... 103 Model 2: Incompletely dominant gene(s) with an additive cytoplasmic interaction .................................... 111 SUMMARY ............................................................ 118 CHAPTER 2: CHARACTERIZATION OF PLANT-VIRUS INTERACTIONS OF THE ToMV RESISTANT SOMACLONES ............................... 122 INTRODUCTION ....................................................... 122 LITERATURE REVIEW .................................................. 122 Tomato mosaic virus life cycle ............................. 122 TMV resistance genes in tomato ............................. 126 Characteristics of resistance genes ........................ 126 Temperature sensitivity .................................... 128 Protoplast infection ....................................... 128 RNA polymerase activity .................................... 129 Grafting experiments ....................................... 129 Resistance mechanisms ...................................... 131 MATERIALS AND METHODS .............................................. 132 RESISTANCE RESPONSE TO DIFFERENT CHALLENGES ................... 132 Response of somaclones to different virus strains .......... 132 Response of somaclones to inoculation of different V viruses .................................................... 132 Response of somaclones to inoculation and incubation at different temperatures ..................................... 133 VIRUS MOVEMENT ................................................ 133 Virus movement in tomato ................................... 133 Virus movement in grafted tomatoes ......................... 135 PROTOPLAST ISOLATION AND SCREENING FOR INFECTION .............. 137 Protoplast isolation ....................................... 137 Protoplast viability ....................................... 139 Preparation and staining of infected protoplasts ........... 139 Detection of virus infection by fluorescence microscopy....l40 Photography ................................................ 140 Preparation of FITC-conjugate .............................. 140 Cross-absorption of FITC-conjugate with acetone powder of healthy tobacco leaves .................................. 141 RESULTS AND DISCUSSION ............................................ 143 RESPONSE TO VIRUS CHALLENGE ................................... 143 Response to different strains of ToMV ...................... 143 Response of somaclonal lines to other viruses .............. 143 Response of somaclonal resistance to high temperature ...... 145 MOVEMENT OF VIRUS ............................................. 146 Systemic movement of virus in susceptible and resistant plants ..................................................... 146 Movement of virus in grafted plants ........................ 148 PROTOPLASTS ................................................... 152 Isolation of protoplasts from inoculated plants ............ 152 SUMMARY AND GENERAL CONCLUSIONS .................................... 158 LIST OF REFERENCES ................................................. 163 APPENDIX A: Regeneration and screening of tomato somaclones for resistance to tobacco mosaic virus ................. 171 APPENDIX B: ELISA virus detection procedures ....................... 177 APPENDIX C: Pedigrees and ratings of reciprocal crosses ............ 182 vi LIST OF TABLES Table 1. Resistance of original somaclones (R) and progeny (R1) to TMV-Flavum ........................................................ 5 Table 2. Comparisons of seed germination rates. Rates are compared among each generation of selfed tomato somaclones, the parental line, GCRI-26, and the isogenic line, GCRI-237, which is homozygous for the Tm-l gene ..................................... 36 Table 3. Comparison of pollen germination rates among pollen samples from somaclonal tomato lines and the parental line, GCRI-26 ............................................................. 37 Table 4. Comparison of resistance rating systems, for susceptible and resistant control lines. S:I:R rating systems are S-high virus titer at 10 d (pi), I-high virus titer at 20 d and ano or low virus titer at 10 and 20 d. S:R ratings are S-high virus titer at 10 d and R-no or low virus titer at 10 d. HzMzL ratings are H-high virus titer (1.3-2.0 AU), M-medium virus titer (0.5-1.3 AU) and L-low virus titer (0.0-0.5 AU) all at 10 d. Mean ELISA value (AU) was calculated from standard- ized values ......................................................... 38 Table 5. Resistance ratings of selfed generations of somaclonal lines. Data compiled for up to five generations for each somaclonal line using the S:I:R rating system which classified plants in three groups according to the length of time the plants remained free of detectable virus. 8 (susceptible) were plants with virus detected at 10 d pi, I (intermediate) were plants with virus detected at 20 d pi and R (resistant) were plants with no detectable virus at 20 d pi. * indicates groups which were screened with TMV-Fl instead of ToMV-0 ................... 43 Table 6. Summary of responses for resistant somaclones in response to inoculation by TMV-0 using A) the S:R rating at 10 d pi (S-virus titer greater than 0.50 AU and R-virus titer less than 0.50 AU) or B) the mean ELISA value (with results of LSD range test between somaclonal lines of each generation). F-values from the analysis of variance tables are listed below for each generation ................................................. 50 Table 7. Testing for significant reciprocal differences for each somaclonal line. ELISA means for F1 progeny of reciprocal crosses were used to calculate the analysis of variance ............. 53 vii Table 8. Summary of results after screening F tomato plants for resistance to ToMV (at 10 d pi). A) S:I:R ratings. B) Mean ELISA values. Lower case letters in parentheses following each ELISA mean indicate the results of LSD multiple range testing among families, where means with the same letter are not significantly different from each other (p-0.05). C) Results of analysis of variance among somaclone families ....................... 55 Table 9. Summary of S:R resistance ratings and mean ELISA values (from 10 d pi) of F1 progeny of reciprocal crosses between each somaclonal line and the paternal line, GCRI-26. The letters in parentheses denote the groups of plants that were screened in the same environmental group. Brackets and (+) indicate that the crosses were made using the same R1 parent plants. Within each environmental group, F1 plants were siblings. Small letters to the right indicate groups determined by LSD multiple range tests (p-0.05) ................................ 60 Table 10. Results of analysis of variance and LSD multiple range tests (p-0.05) of ELISA means of F1 progeny of pooled somaclone families. A) Among environmental groups. Groups were identified by upper and lower case letters which were assigned in alphabetical order at the time of screening. B) Among somaclone F families within environmental groups. C) Among somaclone F1 families ...................................... 62 Table 11. Summary of results after screening F1 and F2 tomato plants for resistance to ToMV (at 10 d pi). A) % S:I:R ratings. B) Mean ELISA values. Lower case letters in parentheses following each ELISA mean indicate results of LSD multiple range testing among families. C) Results of analysis of variance among somaclone families. D) Comparison of reciprocal F1 and F2 families ............................................................ 66 Table 12. Summary of resistance ratings at 10 d pi of F2 progeny of reciprocal crosses between somaclonal lines and the susceptible line, GCRI-26. (+) means groups are siblings. The letters in parentheses denote the environmental group. Plants within each group were siblings ................................................. 71 Table 13. Results of analysis of variance and LSD multiple range tests (p-0.05) of ELISA means of F2 progeny of pooled somaclone families. A) Among environmental groups. B) Among somaclone F families within environmental groups. C) Among somaclone F2 families ............................................... 75 Table 14. Results of screening of cross-pollinated plants from reciprocal crosses of tomato somaclone R2 with the susceptible cultivar, Bonny Best. A) S:I:R ratings and B) Mean ELISA values, with LSD multiple range tests (p-0.05) to compare ELISA means between crosses within each somaclonal line ......................... 79 Table 15. Mean ELISA values for each backcross shown in Figure 12. The results of analysis of variance and LSD multiple range tests (p-0.05) were included ........................................ 84 viii Table 16. Orthogonal comparisons between backcrosses using ELISA values at 10 d pi from the pooled somaclone families .......... 85 Table 17. S:R ratings (10 d pi) of backcrosses which were screened with ToMV. Results of each environmental group were shown for each somaclone family ..................................... 87 Table 18. Results of screening of backcrosses with ToMV. A) Expressed as percentages (S:R rating at 10 d pi). B) Mean ELISA values at 10 d pi ............................................. 88 Table 19. Summary of results from screening somaclonal line 247, using the percent S:R at 10 d pi, the ELISA mean and the H:M:L ratings (high - 1.33-2.0 AU: medium - 0.51-l.32 AU: low - 0.00-0.50 AU) for virus titer ....................................... 91 Table 20. Summary of results from screening somaclonal line 12, using the percent S:R at 10 d pi, the ELISA mean and the HzMzL ratings (high - 1.33-2.0 AU: medium - 0.51-1.32 AU: low - 0.00- 0.50 AU) for virus titer ............................................ 92 Table 21. Results of testing for the presence of ToMV using ELISA in different parts of infected or healthy plants .............. 94 Table 22. Results of screening reciprocal crosses made between somaclonal line 12 or 247 and tomatoes bearing the resis ance genes Tm-l/Tm-l (GCRI-237), Tm-Z/Tm-Z (GCRI-236) or Tm-Z /Tm-22 (CORT-267) .......................................................... 96 Table 23. Model 1: Description of proposed genetic model of inheritance of somaclonally derived resistance to ToMV ............. 104 Table 24. Model 1: List of predicted genotypes resulting from all possible crosses between the hypothesized genotypes of somaclonal lines and GCRI-26 of Table 23 ........................... 106 Table 25. Model 1: Summary of genotypes and phenotypes for resistance to ToMV in tomatoes with an additive resistance plus a cytoplasmic resistance factor .................................... 107 Table 26. Model 1: List of combinations of F1 and F2 predicted results from crosses of all possible genotypes of somaclonal lines and GCRI-26....................................... ........... 109 Table 27. Model 1: Comparison of predicted and observed results of an additive duplicate gene model using data from somaclonal line 247 ................................................ 110 Table 28. Model 2: Descriptions of a proposed genetic model for the inheritance of ToMV resistance seen in tomato somaclone lines. The model is based on an incompletely dominant nuclear gene with a maternal effect. A) Definition of terms. B) Discription of genotypes and additive rank resulting from gene dosage ........................................................ 113 ix Table 29. Model 2: Additive nuclear-cytoplasmic interaction. A) Assignment of additive ranking values for different genotypes. B) Proposed genotypes and additive gene dosage values ............. 115 Table 30. Model 2: Additive nuclear-cytoplasmic interaction. Comparison of predicted resistance rank with mean ELISA values (from 10 d pi) for pooled somaclonal families .............. 117 Table 31. List of TMV resistance genes in tomato and some of their characteristics .............................................. 127 Table 32. Response of somaclonal lines to inoculation by three strains of TMV ..................................................... 143 Table 33. Ratings (S:I:R) of plants inoculated with ToMV-O and incubated at either 25° or 30° C for 30 d. ( ) indicates number of plants tested ................................................... 145 Table 34. Correlation of ELISA values to protoplast yield and to percentage of infected protoplasts isolated from susceptible and resistant tomato lines. Also included are ELISA values of protoplast samples and of CPW buffer used in rinsing the protoplasts during preparation ..................................... 155 Table I. Range and mean of ELISA values for healthy and TMV-Flavum infected GCRI-26 tomato plants .......................... 173 Table II. Resistance of original somaclones (R) to TMV-Flavum. Back inoculations of extracts of somaclones to N. glutinosa were also negative ................................................. 173 Table III. Resistance of R progeny of somaclones to TMV-Flavum. ( ) - Number 0% plants in assay ....................... 173 Table Cl. Pedigrees and ratings of reciprocal crosses. Listed on the left is genotype (female parent x pollen parent) followed by the S:I:R rating of the plants. Results of screening of F2 are presented on the same line as the F (or the next line if there is more than one set). The letters in parentheses denote the groups of plants that were screened at the same time. The last column represents the designation of the F1 plant which produced the seed for the F screening. *indicates a plant which was screened by TMV-F avum rather than ToMV-0. "un" indicates that F seed for screening were collected from uninoculated and unrated plants .................... 182 LIST OF FIGURES Figure 1. Inclusion bodies of ToMV-0 in tomato leaf hair cells appear crystalline (x500) ........................................... 21 Figure 2. Distribution of ELISA values at 10 d pi for the following genotypes: A) susceptible +/+ (GCRI-26) B) resistant Tm-l/Tm-l (GCRI-237) and C) resistant (somaclonal R1) ............... 40 Figure 3. Pedigrees and resistance ratings of all generations of tomato somaclones screened. The (S:I:R) ratings classified plants in three groups according to the length of time the plants remained free of detectable virus. S (susceptible) were plants with virus detected at 10 d pi, I (intermediate) were plants with virus detected at 20 d pi and R (resistant) were plants with no detectable virus at 20 d. # indicates groups that were selected from seeds of parents with an S or I rating. * indicates groups which were inoculated with TMV-Flavum (all other plants were inoculated with ToMV-0). Upper- and lower-case letters indicate the groups of plants which were screened together. After the R generation, each subsequent generation was labelled with a number indicating the plant which parented the generation. The numbers designating the plant lines are underlined. The S:I:R ratios (representing individuals) are given for each group screened, except in the cases marked with "un" which means the group was uninoculated ....................................................... 46 Figure 4. Comparison of virus resistance in F of reciprocal crosses between somaclonal lines (pooled) and the susceptible parental line, GCRI-26. Scored at 3 stages in their development: 10, 20, and 30 d pi ................................................ 53 Figure 5. The distribution of ELISA values for virus detected in plants at 10 d pi, from the crosses A) between GCRI-26 (+/+) and GCRI-237 (Tm-l/Tm-l), B) GCRI-26 x somaclone lines 215, 219, 247, 322 and 330, and C) the same somaclone lines x GCRI-26 ........ 57 Figure 6. Resistance to ToMV in progeny of reciprocal crosses of tomato somaclonal lines with the parent susceptible line, GCRI-26, for both the F1 and F2. A) Bars represent the percentage of virus-free plants at 10, 20 and 30 d. B) Testing for significant differences for each somaclonal line ................... 64 xi Figure 7. The distribution of ELISA values for virus detected in F2 progeny at 10 d pi from crosses between A) GCRI-26 and somaclonal lines and B) the reciprocal cross, of somaclonal lines and GCRI-26 ................................................... 69 Figure 8. Pedigree of reciprocal crosses of somaclonal line 247. Different R1 plants were used for the different numbered crosses shown. For each family of progeny screened, the environmental group letter (in parentheses) was given, followed by the S:R rating at 10 d pi and the mean ELISA value. Results of LSD tests were included for comparison of means of each particular cross .................................................... 73 Figure 9. A) Graph of resistance to ToMV in reciprocal crosses of tomato somaclones with the susceptible line, Bonny Best for both F and F progeny, scored at 3 stages in their development: 10, 20 and 30 d pi B) mean ELISA values ............................. 77 Figure 10. The distribution of ELISA values for virus detected in plants from A) resistant somaclonal lines and susceptible Bonny Best, B) the F1 progeny of the cross Bonny Best x Sc and C) the F1 progeny of the reciprocal cross, Sc x Bonny Best .......... 81 Figure 11. The distribution of ELISA values for virus detected in F2 progeny plants from the crosses A) Bonny Best x Sc and B) Sc x Bonny Best ..................................................... 82 Figure 12. Results of screening backcrosses for resistance to ToMV. Bars represent the ELISA mean for each backcross, with lines indicating standard error of the mean (using pooled data from all somaclonal lines) ..................................... 84 Figure 13. The distribution of ELISA values for virus detected in plants from the backcrosses made between susceptible GCRI-26, resistant somaclonal lines and the F1 progeny of these. Upper graph) Backcrosses A-D used the somaclonal lines once in the pedigree and Lower graph) backcrosses E-H used the somaclonal lines twice in the pedigree ......................................... 89 Figure 14. Restriction patterns of chNA from parental tomato line GCRI-26 compared to somaclone line 247 (A) and 12 (B). Lanes marked 0 are undigested chNA. Lambda DNA fragments (restricted by Hind III) were used as markers ...................... 100 Figure 15. Extraction of dsRNA species from tomato somaclones 12 and 247 and from parental line 26. Controls show bands after extraction from ToMV-infected tomato (faint) at 6.15 kb and from pepper at 14 kb ........................................... 101 Figure 16. Graph of resistance ranking versus mean ELISA values for each cross and backcross between GCRI-26 and the pooled somaclonal lines. (Line was determined by linear regression analysis ................................................ 117 xii Figure 17. Tomato mosaic virus life cycle. a) virus enters cell through broken membrane and b) is uncoated while being translated from the 5' end c) the RNA is replicated (found in the double-stranded replicative form) and many subgenomic RNA's are made (found as the replicative intermediate with single- stranded RNA tails). e) viral proteins are synthesized from the subgenomic RNA's on the plant ribosomes. Virus is transported from f) cell to cell and then g) throughout the plant. h) typical symptoms are mosaic leaves, and viral inclusions can be detected in the leaf hair cells .................................... 123 Figure 18. Diagram of sampling of inoculated tomato plants for detection of virus spread over time and space. The triangle indicates the leaflet which was inoculated. Numbering and lettering identified each leaflet for later reference ............. Figure 19. Diagram of whip graft (A) and whip and tongue graft (B) ......................................................... Figure 20. Response of resistant tomato somaclonal lines to virus challenge with ToMV, CMV or TEV ............................. Figure 21. The movement of ToMV-O in tomatoes. Susceptible GCRI-26 tomatoes (A) allow the movement of virus as illustrated by the shaded leaves. At four days, the virus was spreading in the inoculated leaf, then moved quickly to the upper part of the plant. After 20 d the entire plant contained virus. In the resistant somaclonal line 247 (B), the virus remained in the inoculated leaf. An open triangle marks the inoculated leaflet... Figure 22. Results of experiments in which grafted plants were inoculated in the outermost lower leaf of the rootstock, and the movement of the virus was followed using ELISA to measure virus titer in young leaves of the rootstock and scion. Grafts were made between A) two susceptible plants as a control. B) and C) reciprocal grafts between resistant (somaclonal line 247) and susceptible (GCRI-26) plants. Illustration represents grafts .134 .136 .144 .147 at 10 d ............................................................ 149 Figure 23. Results of grafting experiments using resistant (somaclonal line 12) and susceptible (GCRI-26) plant material after inoculation with ToMV-0 and screening young leaves from rootstock and scion using ELISA. Illustration represents the grafts at 10 d ..................................................... 151 Figure 24. A) Infected cells isolated from GCRI-26. B) One infected cell was present among thousands of healthy cells in plant 12-1. C) Cells from resistant plant 247-4 show low levels of virus in most cells. (x600) .............................. . ...... 154 xiii LIST OF ABBREVIATIONS AU ................... Absorbance units. BSA .................. Bovine serum albumin. CMV .................. Cucumber mosaic virus. CP ................... Coat protein. chNA ................ Chloroplast deoxyribonucleic acid. CPW .................. Cell protoplast wash. DE-22 ................ DEAE-cellulose ion exchange material. DNA .................. Deoxyribonucleic acid dsRNA ................ Double-stranded ribonucleic acid. EDTA ................. Ethylenediamine tetraacetic acid. ELISA ................ Enzyme-linked immunosorbant assay. EtOH ................. Ethanol. F1 ................... First filial generation (after cross-pollination). F2 ................... Second filial generation (after cross-pollination). FITC ................. Fluoroisothiocyanate. GCRI ................. Glasshouse Crops Research Institute. H:M:L ................ ToMV resistance rating of high, medium or low for plants dependent on the amount of virus detected by ELISA in the plant after inoculation. LSD .................. Fisher's least significant difference multiple range test. MS ................... Murashige-Skoog plant cell growth medium. PBS .................. Phosphate-buffered saline. PBS-Tween ............ Phosphate-buffered saline + Tween 20 (a detergent) xiv PEG .................. Polyethylene glycol. pi ................... post-inoculation. PVX .................. Potato virus X. R0 ................... Regenerated plant or somaclone. R1, R2, etc .......... Self-pollination generations of somaclone with subscript numbers designating the number of the generation. RF ................... Replicative form of virus. RFLP ................. Restriction fragment length polymorphism. RI ................... Replicative intermediate of virus. RNase A .............. Ribonuclease A. Sc ................... Somaclone. SDS .................. Sodium dodecyl sulphate. S:I:R (or S:R) ....... ToMV resistance rating of susceptible, intermediate or resistant depending on the amount of virus detected in the plant by ELISA. SNK .................. Student-Newman-Keuls multiple range test. STE .................. Buffer composed of sodium chloride, Tris and EDTA. TAE .................. Buffer composed of Tris, acetate and EDTA. TEV .................. Tobacco etch virus. Tm-l, Tm-2, Tm-ZZ....T0MV-resistance genes described in tomato. Tm-3 ................. Tentative name for somaclonal ToMV-resistance gene. TMV .................. Tobacco mosaic virus. TMV-F1 ............... Flavum strain of tobacco mosaic virus. ToMV-0 ............... Common strain of tomato mosaic virus. XV INTRODUCTORY SUMMARY The goal of this dissertation research was to characterize tomato plants that had been selected for resistance to tobacco mosaic virus (TMV). These plants were recovered after regeneration from tissue culture of a fully susceptible tomato line. A new type of resistance was obtained, which appears to be encoded by a nuclear gene, with the addition of a maternal effect. The characterization of this resistance extends into several areas, including classical and molecular genetics, and virology, which will be treated in separate chapters. Tomatoes were specifically chosen for this work because much is known about the genetics of the tomato. Two recent linkage maps are available (Mutschler et a1., 1987). The classical linkage map contains 327 loci defined by isozymes, morphological markers, resistance genes and physiological mutants, and the molecular linkage map which consists of 150 restriction fragment length polymorphisms (RFLPs) created by using random single copy DNA probes as well as clones of known genes. In addition, the chloroplast and mitochondrial genomes of tomato are characterized (Schiller et a1., 1982; Piechulla et a1., 1985; McClean and Hanson, 1986; Palmer and Zamir, 1982; Hause et a1., 1986). Tomato mosaic virus (ToMV) is a strain of tobacco mosaic virus and one of the best characterized of plant viruses (CMI/AAB Description of Plant Viruses, 1977). These viruses, which infect a broad host range, 2 are closely related members of the tobamovirus group. They are composed of a single strand of RNA which is coated with protein subunits to make a straight rod ca.300 x 18 nm. Tomato mosaic virus is widespread and often epidemic in tomato, and is economically damaging in both glasshouse and outdoor tomato crops, with fruit yield reduced by 15 to 25% (Broadbent, 1964). Virus diseases, in contrast to those caused by bacteria and fungi, cannot be controlled directly by chemicals or antibiotics. Resistant varieties have offered the most effective means of control (Gibbs and Harrison, 1976). There are three well-studied genes for resistance to ToMV in tomato: Tm-l, which slows virus replication and suppresses symptom formation, and Tm-Z and Tm-22 (which are allelic) effectively prevent infection of ToMV by a hypersensitive response. Near-isogenic tomato lines with these resistance genes are available. Unfortunately, strains of ToMV have been isolated which overcome the resistance conferred by each of these alleles (Pelham, 1972; Rast, 1975). So far there are no strains which overcome the resistance of tomatoes which have both Tm-l and Tm-Z or Tm-ZZ. RNA genomes have a high rate of mutation (Reanney, 1982), thus it is only a matter of time before current resistant cultivars are overcome. Common ToMV field isolates, which may contain several virus strains, have been found to include strains able to overcome each type of resistance (Alexander, 1971; Rast, 1975). It has been demonstrated that a strain may change markedly on passage through some resistant tomato genotypes (litter and Murakishi, 1969; Pelham et a1., 1970), which seem to select for more virulent virus. 3 The resistances which are available in tomato have been crossed into the cultivated tomato from wild tomato species. Because these resistance genes are often linked with undesirable characteristics such as temperature sensitivity, severe stunting and chlorosis (Russell, 1978), conventional plant breeding can be more difficult and time consuming. In the search for new resistance genes, an alternative approach seemed desirable. It was demonstrated that it is possible to obtain ToMV resistance by regenerating tomato plants from tissue culture (Barden et a1., 1986). In this work, the tissue culture- derived resistance was characterized in order to determine the mechanism of resistance, and how it was generated. ORIGIN OF SOMACLONES The tomato somaclones were derived through tissue culture by Kristen Barden (1985). The description and selection of the six somaclones studied in this dissertation are described in detail in Appendix A (Barden et a1., 1986). The fully susceptible (+/+) line, GCRI-26, derived from the variety ‘Craigella' (Glasshouse Crops Research Institute, Littlehampten, U.K) was chosen for the starting material because it is isogenic to a series of lines with the known ToMV resistance genes in tomato. This allowed direct comparison to the resistances conferred by these genes. To regenerate plants, sections removed from fully-expanded leaves of GCRI-26 plants were placed on regeneration medium. Direct adventi- tious shoots formed after four weeks, were rooted and moved into soilless peat. The tomato somaclones were inoculated with TMV-Flavum, a yellowing strain of TMV. Visual mosaic symptoms were detectable in 4 susceptible plants 7-10 days post-inoculation. Symptomless plants were screened using enzyme-linked immunosorbant assay (ELISA) to determine the presence of viral coat protein in the plants. Local lesion assays were also done to determine the presence of infective virus. Of 370 somaclones, 18 were virus-free after the initial screen- ing. Records were not kept of the origin of each of these somaclones, but most of them were regenerated from different leaf sections. Twelve of the virus-free somaclones became infected with virus after re- inoculation. The remaining six virus-free somaclones (Table 1) were self-pollinated for seed collection. Seed was collected from each plant for evaluation of the resistance of the next generation, designated R1. The R1 plants were inoculated with TMV-Flavum and screened for resistance. Most plants were shown to be virus-free at 30 days post-inoculation (pi). It was therefore shown that somaclones resistant to TMV could be obtained from a recognized fully susceptible source cultivar using ordinary tissue culture methods without the addition of a known mutagen. The type of resistance shown by the somaclones was a delay in symptom expression and virus multiplication. This dissertation includes the results of further studies to determine the nature of the resistance and its inheritance. GENETICS EEElEQLLQD 2i Eh; plant material. The somaclonal plants were found to resemble the parental plant material in every way except for the resistance trait. This determination included comparisons of gross morphology (to screen for undesirable traits), chromosome number, seed Table 1. Resistance of original somaclones (R) and progeny (R1) to TMV-Flavum. R ELISA Duration of R1 virus- Days Plant valuesa experiment (d) free/total post-inoc. (AU) Sc 12 0.00 51 15/15 30 0.00 11/11 41 Sc 215 0.03 161 13/13 30 0.04 Sc 219 0.03 173 8/9 30 0.02 Sc 247 0.03 101 13/14 30 0.04 Sc 322 0.04 137 12/16 30 0.04 Sc 330 0.04 95 14/18 28 0.04 GCRI-26 >0.50 20 0/27 30 Uninoc. 0.00- - 6/6 55 (range) 0.07 aELISA to detect TMV-Flavum coat protein. (Taken from Barden et a1., 1986.) 6 germination rate and pollen viability. The heritability of the viral resistance was then determined (to distinguish from epigenetic mech- anisms), and the mode of inheritance was studied. Rating systems were devised to identify resistant phenotypes and were used to compare self- pollinated generations, crosses with the parental line (F1) and their progeny (F2) as well as backcrossed plant lines. Because a complex genetic pattern was evident which included non-Mendelian inheritance, an effort was made to detect changes in the cytoplasm at the molecular level. To compare organelle genomes to the parental line, chloroplast DNA (chNA) was extracted and cut with four endonucleases for visualization of restriction fragment length polymor- phisms (RFLPs). The same goal for mitochondrial DNA was not achieved. Also, screening for the presence of double-stranded RNA (dsRNA) was conducted to determine whether the presence of such a molecule might play a role in the viral resistance. VIROLOGY Exglugtign of the plan§;xiru§,in§ezaggign. The somaclonally-derived resistance was compared to resistance conferred by other known TMV re- sistance genes. The plants were challenged by the common tomato virus strain (ToMV-0) under different conditions, by different ToMV strains and by other viruses. The systemic movement of the virus through sus- ceptible tomato plants was compared to the movement in resistant soma- clones. The movement of virus was also monitored in grafted resistant and susceptible plant material. Finally, the presence of virus was de- tected at the cellular level by releasing protoplasts from inoculated plants and immunofluorescent staining to detect virus coat protein. CHAPTER 1 CHAPTER 1 GEEEIIC CHARAQIERIZAIION 9: IEE Tony-RESISIANI SOMACLONES. INTRODUCTION In characterizing the genetics of resistance to ToMV in tomato somaclones, several questions were addressed. The somaclonal lines were compared to the parental line to determine if there were deleterious traits in the resistant somaclones, the heritability of the resistance, and the mode of inheritance. The type of resistance was also examined using virological assays and compared to other known types of TMV resistance in an effort to ascertain the origin of the somaclonal resistance. LITERATURE REVIEW Plant tissue culture is the 13 zigzg, aseptic cultivation of plants, using plant organs, meristems, callus culture, protoplasts or single cells (Thorpe, 1981). Plants derived from many forms of tissue culture are called somaclones, and the variation seen in plants derived from tissue culture has been termed somaclonal variation (Larkin and Scowcroft, 1981). The genetic variability generated during plant cell culture has been reviewed extensively (Larkin and Scowcroft, 1981; Evans and Sharp, 1986; Evans et a1., 1984; Brettel and Ingram, 1979; Meins, 1983 and D'Amato, 1977) and evaluated for use in the selection of plants resistant to disease (Daub, 1986). Plant tissue culture has been proposed as a source of novel genetic variabiltiy to be exploited in selection programs (Buiatti et a1., 1985) and a review of the 7 8 literature supports the notion that plant cell culture itself generates genetic variability. It is still debated whether somaclonal variation is a result of genetic differences pre-existing in somatic cells of the explant or is induced during the tissue culture phase by components of the tissue culture medium (Evans and Sharp, 1986). Some cases of genetic variability after plant tissue culture have been attributed to gross chromosomal changes such as altered chromosome number (D'Amato, 1978) and translocations (in oat, McCoy et a1., 1982; in celery, Orton, 1983). Many other cases of genetic variation have been characterized as single gene mutations which could be due to large or small gene rearrangements, gene amplification or deletion, transpos- able elements, sister chromatid exchange or cryptic virus elimination (Larkin and Scowcroft, 1981) as well as cytoplasmic changes in genomes of organelles (Gengenbach et a1., 1977). Single gene mutations have been detected in tomato (Evans and Sharp, 1983), maize (Edallo et a1., 1981), potato (Ramulu et a1., 1984), tobacco (Prat, 1983) and many more (Larkin and Scowcroft, 1981; Evans and Sharp, 1986). Multigenic inher- itance has been established for a tissue-culture derived resistance to Fusarium wilt in celery (Heath-Pagliuso et a1., 1988). Cytoplasmic changes have been detected and confirmed by RFLP mapping of mito- chondrial DNA in potato (Kemble and Shepard, 1984) and in maize (Gengenbach et a1., 1977). In other cases, plant regeneration has resulted in transient variation, termed epigenetic, as well as genetic variation. Epigenetic variation involves selective gene expression that occurs with regularity in response to certain inducers present in the tissue culture system, and is not transmitted meiotically (Meins, 1983). Cytokinin habituation in tobacco is an example of epigenetic variation (Meins and Lutz, 1980). Epigenetic inheritance must be tested by looking at the R1 generation to determine whether the trait is expressed after meiosis. Tomato has served as a useful model in studying somaclonal variation because the genome has been extensively mapped using morphological traits, isozymes and RFLPs. In one study of somaclonal variation in tomato, plants were regenerated and self-pollinated for screening of the R1 progeny for morphological changes (Evans and Sharp, 1983). Seven recessive and three dominant traits (all showing normal single gene inheritance) were characterized (Evans et a1., 1984). In another approach, Sibi et a1. (1984) studied the recombination between two linked genes on a chromosome in tomato plants regenerated from tissue culture, using two sets of nuclear gene markers. They found significant increases in the recombination rate in about half of the somaclones tested. Buiatti et a1. (1985) self-pollinated 88 tomato somaclones to screen for chlorophyll deficiencies and other morphological abnormalities. About 17% of the R1 progeny had deficiencies, all seemingly recessive. They also found chimeric variegation, suggesting that the regenerated plant originated from more than one cell in culture. In another case, regenerated tomato plants with phenotypic variations were isolated and analyzed genetically (Sibi, 1986). Self- fertilization revealed that 56% of plants were mutant for five characters noted. Another 17% of the plants presented new character- istics but did not give segregating progeny, thus appearing to be homozygous. Of these, several showed differences between progeny of reciprocal crosses to the control line, mostly paternally and some 10 maternally oriented. Because of the stability of the effects after selfing the crosses between variants, as well as their asymmetrical behavior in reciprocal crosses, it was concluded that this phenomenon was not directly caused by heterozygous coding DNA, nor by cytoplasmic effects alone. The term ‘epigenic modification' was used to describe this type of inheritance, and was attributed to modifications of heritable biological structures or dynamic systems, not involving expressed genes directly. In an approach to assess genome stability in cultured tomatoes, O'Connell et a1. (1986) analyzed isozyme patterns for eight enzymes and found no differences in patterns between regenerated plants and parental plant material. However, after callus culture (protoplast- derived callus grown for one month) or interspecific fusion with L. pennellii, they found variation in isozyme patterns. Similarly, Nagy et a1. (1983) compared mitochondrial DNA RFLPs of tobacco before and after regeneration from tissue culture and found no differences, but could detect differences after interspecific protoplast fusion. One of the limitations to utilizing somaclonal variation in a resistance breeding program is the selection and identification of desired mutant phenotypes. To recognize the desired resistance, an appropriate screening and selection scheme must be designed and utilized. For tomato, several approaches have been successful. In one example, tomato somaclones were selected for resistance to Fusarium wilt by including the toxin, fusaric acid, in the culture medium. After the successful selection of several lines and screening two generations of progeny, it was determined that a single gene for resistance was involved (Shahin and Spivey, 1986). In another approach, unselected somaclones of ‘Bulgaria 12' (esculentum x ll pimpinellifolium) were generated, self-pollinated and the progeny screened for novel resistance to Clavibacter michiganense subsp. michiganense (CMM) (R. DeVries, personal communication). Several resistant lines of tomatoes were recovered, but the mode of inheritance was not established. Another strategy to generate resistance in cultured cells was to include a toxic molecule in the tissue culture medium to select for resistant individuals. In tomato, success was achieved in gaining glyphosate tolerance (Smith et a1., 1986), aluminum resistance (Meredith, 1978) and tolerance to paraquat (Thomas and Pratt, 1982). The paraquat-tolerant mutants were carried to the next generation to determine that the trait was heritable. Tomatoes resistant to ToMV were recovered in this study. After determining the best system for rating resistance, the inheritance of this resistance was established by genetic analysis. Through self- pollinations, reciprocal crosses with the parental line, and F1 backcrosses, as well as reciprocal crosses with another susceptible line of tomato, it was possible to estimate penetrance, expressivity, and the amount of cytoplasmic effect. Models to explain the mode of inheritance were discussed. It was also possible to compare somaclones to one another to establish their similarities. To assess the relationship of the somaclonal resistance to other ToMV resistance genes, reciprocal crosses were made with the near-isogenic tomato lines for Tm-l, Tm-Z and Tm-22. This genetic characterization comprises the first part of this dissertation. 12 MATERIALS AND METHODS PLANT AND VIRUS MATERIAL Elanta. The original source material for the somaclones was GCRI-26, a fully TMV susceptible line (+/+) of ‘Craigella' (Glasshouse Crops Research Institute, Littlehampten, U.K.). Near-isogenic lines of ‘Craigella' with the TMV resistance genes, Tm-l (GCRI-237), Tm-Z (GCRI-236) and Tm-22 (GCRI-267) were obtained from the same source for use in comparisons. The seeds from GCRI lines and seeds collected from the tomato somaclone lines were sown in soilless peat (Baccto Professional Planting Mix, Michigan Peat 60., Houston, TX) that was moistened with distilled water and covered with plastic wrap. Germinating seeds were kept at room temperature (about 240 C) under cool white fluorescent lights supplying 60 uEm"2s'1 with a sixteen hour light cycle. Seeds usually germinated in five to seven days. Seedlings were transplanted into 1" x 2" pots at 9-12 d in preparation for screening. Plants were watered daily with distilled water containing fertilizer at one tablespoon of Peters 20-20-20 mix (Peters Fertilizer Products, Allentown, PA) per gallon. Eitaaaa. The common strain of Tomato mosaic virus (ToMV-O) was kindly provided by F. Motoyoshi of the National Institute of Agrobiological Research, Ibakari, Japan, and was increased in GCRI-26 tomato plants. The virus strain used in the early screening and selection of the original somaclones was TMV-Flavum (furnished by H. Jockusch of Max Planck Institut, Tfibingen, Federal Republic of Germany, and increased in GCRI-26 tomato plants), a strain of TMV that causes bright yellow mosaic 13 symptoms on susceptible tomatoes. This strain is a tobacco strain, and presumably less virulent in tomato than ToMV-0. Therefore, further screening of the progeny of the somaclones was done using ToMV-O. Inoculum consisted of virus from the leaves of infected tomatoes that were collected, ground in 0.01 M sodium-potassium (NaK)-phosphate buffer, pH 7.0, at a dilution of l g:10 ml, filtered through four layers of cheesecloth, divided into aliquots and frozen at -200 C. Both strains of virus were preserved in this manner and used throughout all of the screening of the somaclones. This insured that the same virus strains were applied to all of the groups of tomatoes. ARE SOMACLONES NORMAL PLANTS? £1255 matphalagy. Somaclonal R, R1, R2 and R3 plants were observed for features that were different from the original source material, GCRI-26. Gross morphology of seedlings and mature, fruiting plants was noted. Datatmtgattan at ahgamoaome naaaat. Root tip squashes were made from rapidly growing roots of transplanted tomatoes following procedures adapted from Krikorian et a1. (1983). Sections (5-10 mm) were cut from exposed root tips and incubated at room temperature in 0.05% colchicine. After 5 h, the colchicine was replaced with a fixative (a 3:1 mixture of ethanol and acetic acid) and the tips were incubated overnight at room temperature. The tips were treated with 1 N HCl for ten min to soften the tissue, transferred to microscope slides and then wetted with additional fixative. The root tips were stained by the addition of several drops of acetocarmine stain (10% w/v acetocarmine in 45% acetic acid). The tip was excised at about 0.5 mm and pressed with the flat end of a glass rod to macerate, with more dye added as needed to keep 14 the tip moist. A coverslip was carefully added and gentle pressure was applied to squash the cells. Cells were viewed at 600x magnification using phase contrast and photographed using 160 ASA tungsten film, with 5.5 volt light source and exposures of 1/5, 1/10 and 1/25 s. For each line of tomato, several plants were sampled and from each plant, chromosomes from ten or more cells were counted. Saga gatmiaatian tata. Records were kept of all seeds planted as described earlier and the number of seeds that germinated. Percent germination was determined at 10 to 14 days after planting, though in a few cases it was as early as 8 days or as late as 28 days. For comparison of rates, control plants (GCRI-26) were included in every group, which was usually comprised of 20-30 seeds of each somaclonal line. Each group of seeds planted served as a replication (10-34 replications for each line), and from this, the mean and standard error were calculated. The germination rates of each somaclonal line were then tested using standard analysis of variance statistics and compared using several multiple range tests to determine whether the somaclonal lines differed significantly from the parental line (using the Number Cruncher Statistical System, or NCSS, software package). Pallan xtatility. Pollen viability was determined using two different methods (Brewbaker, 1964). For each blossom examined, at least 500 pollen grains were scored. For each plant line, this was repeated several times using both methods. After collection of pollen from fully-opened flowers, pollen was released, placed on a microscope slide and stained with acetocarmine. Viable pollen grains excluded the dye, and looked intact and round, while those pollen grains which were stained by the dye often appeared shriveled. The second method was to 15 germinate the pollen in vitro. Again, pollen was collected from opened flowers but then released onto pollen-germination medium (20% w/v sucrose, 1 mg/l H3803 and 1.5% agar) in a petri plate. After 1-3 h, germinated and ungerminated pollen grains were counted. There was very good agreement in the results obtained by these two methods of determining pollen viability. Standard statistical methods (using NCSS software) were used to calculate means, standard errors, analysis of variance table and multiple range tests using the pollen germination data. SCREENING FOR RESISTANCE TO TMV Sttaaning 9f tamatg aamatlanaa. For screening seedlings for resistance to TMV, plants which had been transplanted to 1" x 2" pots at 9-12 days were allowed to recover for at least one day and then inoculated with virus. Since the first true leaves were Just beginning to expand and were not of uniform size, only the cotyledons were inoculated. For inoculation, frozen virus stock was thawed and diluted to 1:20 with 0.01 M NaK-phosphate buffer (pH 7). This was applied to carborundum-dusted cotyledons with a cotton swab. The concentration of ToMV was approximately 0.6 mg protein/ml, as determined by ELISA (using purified ToMV as a standard). Cotyledons were allowed to air dry and were then rinsed with distilled water to remove the excess virus and carborundum. The inoculated tomato plants were placed on shelves in the laboratory, under cool-white fluorescent lights supplying 60 uEm'zs’1 with a l6-hour light period at room temperature (24° C). Plants were watered daily with distilled water containing 1 tablespoon Peters 20-20- 16 20 fertilizer per gallon. In earlier screening of tomatoes, plants were transplanted to larger pots at 30 d and moved to a controlled environmental chamber or to the greenhouse. Later, plants were kept in small pots on the lab shelf for 70 days unless they were needed for the production of seeds, in which case they were transplanted into large pots and kept in the greenhouse. When it was time to assay the tomatoes for the presence of virus, small sections of leaf near the growing tip of the plant were removed. It was important to take the samples from the youngest, most rapidly dividing tissue because it had been established that the virus appears in that part of the plant very soon after infection, even if infection is initiated in other parts of the plant (Samuel, 1934). The leaf sections were weighed, and trimmed if necessary, to provide 20 mg fresh weight. Three types of virus assay were employed during different phases of the screening. Enzyme-linked immunosorbant assay (ELISA) for detection of ToMV coat protein was used in all screening of plants. Presence of viral coat protein has been shown to be closely correlated with infectivity, with the presence of ToMV-RNA in the GCRI tomato lines, and with ToMV (Fraser and Loughlin, 1980). ELISA data were supplemented with rub-inoculation to local lesion hosts or detection of viral inclusions using the light microscope. Eatifiitatian at xitaa. Virus was rub-inoculated onto leaves of 6-8 week-old Havana-38 tobacco plants and allowed to incubate for 2 to 3 weeks. Tobacco leaves were then removed and weighed. The leaf tissue was placed in the freezer and stored at -200 C until use. For the purification, the method of Gooding and Herbert (1967) was followed with slight modifications. The frozen tissue was thawed and homogenized 17 (l g tissue/2 ml 0.1 M NaK-phosphate buffer, pH 7.0) in a blender for three min. The homogenate was filtered through four layers of cheesecloth and the filtrate distributed into 250 ml centrifuge bottles. They were centrifuged at 10,000 rpm for 15 min at 40 C in a Sorvall GSA rotor (16,400 x g) to remove cellular debris. The supernatant was collected, adjusted to 8% (w/v) polyethylene glycol (PEG) MW 8000 and to 0.2 M NaCl. The final volume was noted and the preparation was stirred at room temperature for 30 min to precipitate the virus. The entire mixture was centrifuged at 10,000 rpm in a Sorvall SS-34 rotor (9400 x g) for 10 min at 4° C. The virus precipitate was then resuspended in 0.01 M NaK-phosphate buffer, pH 7.0, at 1/5 the total volume used in the PEG precipitation. Three repetitions of high (46,000 x g) and low (12,000 x g) speed centrifugations were performed to further purify the virus. The high speed ultracentrifugation of the virus solution was performed at 25,000 rpm for 2.5 h at 40 C in a Beckman T170 rotor to pellet the virus. The pellets were resuspended in 0.01 M NaK-phosphate buffer, pH 7.0 after overnight incubation at 4° C and then centrifuged for 10 min at 10,000 rpm (SS-34 rotor) to remove additional cell debris. Concentration of virus was estimated based on absorbance at 260 nm. At this wavelength, the extinction coefficient for purified TMV absorbance (1 mg/ml) is 3.0 (Zaitlin and Israel, 1975). Purity of the preparation was measured by comparing the ratio of absorbance at 260 to 280 nm (a ratio of 1.17 is best). The purified virus was filtered through a Morton ultra-fine fritted disc filter (0.9-1.4 u pore diameter) to remove microorganisms, and stored in 0.01 M phosphate 18 buffer. After filtering, the absorbance at 260 nm was measured and the virus was diluted to a concentration of 1 mg/ml and stored in 1 ml aliquots at 4° C. Enayma;liakag immunoaotbant aaaay. The procedures used for ELISA were based on those of Clark and Adams (1977) and are detailed in Appendix B. Antisera against ToMV-0 and TMV-Fl were elicited in rabbits following three weekly intramuscular injections of 1 mg/ml purified virus emulsified in Freund's adjuvant (Difco Co., Detroit, MI). The rabbits were bled at the ear 10-14 days after the final injection. After allowing the blood to clot in 50 ml centrifuge tubes, the antiserum was collected following centrifugation in a Sorvall SS34 rotor at 5000 rpm for 10 minutes. The antiserum was then diluted 1:10 with distilled water, and precipitated by the addition of 10 volumes of saturated ammonium sulfate solution. After one hour of precipitation at room temperature, the antiserum was centrifuged at 6000 rpm for five minutes in a Sorvall type SS-34 rotor to collect the precipitated proteins. The protein pellet was resuspended in two volumes of 1/2 strength phosphate buffered saline (PBS) at pH 7.4. Phosphate buffered saline consisted of 0.1 M sodium-potassium phosphate and 0.85% NaCl, pH 7.4. The antibody solution was dialyzed against three changes of 1/2 strength PBS including once overnight to desalt. This solution was washed through a DEAE-cellulose ion exchange column (Whatman DE 23) using 1/2 strength PBS; the fractions which were collected were monitored at 280 nm on a Beckman DB spectrOphotometer. The purified gamma-globulin was adjusted to have an absorbance reading of 1.4 units, which corresponds to a concentration of 1 mg/ml. Purified gamma-globulin was stored at -200 C in small aliquots. 19 Purified gamma-globulin of each virus was conjugated to the indicator enzyme, alkaline phosphatase. For each 1 mg of gamma- globulin, 2 mg of alkaline phosphatase type VII (Sigma P-5521) was used. These were mixed and dialyzed against three changes of normal strength PBS. Glutaraldehyde was added to the antibody-enzyme mixture to make a final concentration of 0.05% (v/v), to initiate crosslinking of the proteins. This reaction was allowed to proceed at room temperature for 4-5 hours. Glutaraldehyde was then removed by dialysis against three changes of PBS + 0.01% sodium azide. Bovine serum albumin was added to a concentration of 5 mg/ml to stabilize the conjugate. The conjugate was stored at 40 C for up to six months. The activity of the conjugated alkaline phosphatase diminished over time, but could be adjusted by increasing the concentration used in the assay. Purified gamma-globulin and gamma-globulin-enzyme conjugate was diluted for use in the enzyme-linked immunosorbant assay. Tests were made to determine the optimum concentration, so that the assay was complete between 30-60 min. The resulting dilution was usually best between 1:1000 and 1:500. Flat-bottomed microtiter plates (Dynatech Labs, Inc., Alexandria, Virginia) were coated with purified gamma-globulin for the appropriate virus. The gamma-globulin was diluted in a coating buffer comprised of 0.05 M sodium carbonate at pH 9.6. Each well was filled with 200 ul of buffer and plates were incubated for 2-4 hours at 37° C or overnight at 40 C. Plates were rinsed three times with PBS + 0.05% Tween-20 (PBS- Tween), then 200 ul of sample was loaded into each well. Samples were prepared at a 1:100 dilution by grinding 20 mg leaf slices in 2 m1 grinding buffer (PBS-Tween with the addition of 2% w/v polyvinyl- 20 pyrrolidone, MW 10,000, and 0.2% w/v ovalbumin). Duplicate wells were used for each sample, and each plate contained four controls (in duplicate): a) grinding buffer alone, b) virus-free tomato, c) infected tomato and d) a standard amount of purified virus (100 ug ToMV/ml). Samples were incubated 4-6 hours at 370 C or overnight at 40 C. Plates were again rinsed three times with PBS-Tween. The gamma-globulin conjugate was then diluted appropriately in grinding buffer, and 200 ul was applied to each well, and incubated 3-4 hours at 37° C or overnight at 4° C. After rinsing three times with PBS-Tween, 300 ul of a reaction mixture containing 1 mg p-nitrophenyl phosphate/m1 in 10% v/v diethanolamine, pH 9.8, was added. Absorbance readings were taken at 405 nm for each well using a Microelisa minireader II (Dynatech Labs, Alexandria, VA) and recorded on a printer (CIEX 80F/T, Components Express Incorporated, Santa Ana, CA 92705). Absorbance readings were taken between 30 minutes and 4 hours, determined by the time when the wells containing 100 ug/ml ToMV-O reached an absorbance between 1.00 and 2.55 absorbance units (AU) for the best range of values. Latal laaiaa aaaay. Inoculated tomatoes were also assayed by inoculation to Nicotiana glutinosa or to N. tabacum xanthi-nc, which are indicator plants for both TMV-Flavum and for ToMV-O. If the tomato plant was infected with these viruses, then the sap would cause local lesions to appear on the tobacco several days after rub-inoculation. This assay was done to determine the presence of infective virus particles, since ELISA detects only the ToMV coat protein. A sample of tomato leaf from the newest growth was taken and ground in 0.01 M phosphate buffer at a dilution of 1:20. Carborundum (600 mesh) was 21 dusted on the tobacco leaf and the sample was applied using a cotton swab. After several minutes, the tobacco leaves were rinsed with distilled water. Local lesions developed after 4-5 days in the greenhouse, and the number of lesions roughly correlated to the amount of virus present. Later it was determined that the plant samples in the ELISA grinding buffer were suitable to inoculate the indicator tobacco plants provided the inoculated leaves were rinsed immediately after inoculation. Dataattaa at zttal ingluaioaa. A third assay to determine the presence of ToMV in the tomatoes was used on several occasions. Viral inclusions in the hair cells of the leaves could be seen in the susceptible tomatoes at 5 days after inoculation. Following the procedure described by McWhorter (1965), a thin section of rapidly expanding leaf tissue was excised, mounted in distilled water under a coverslip and viewed by bright field microscopy at 100-200x magnification. The inclusion bodies looked like crystals, often hexagonal (Figure 1), in the cytoplasm of the hair cells. The presence of inclusion bodies correlated very well with ELISA values above 0.50 AU. Figure 1. Inclusion bodies of ToMV-O in tomato leaf hair cells appear crystalline. (x500) 22 Aaalxaia at ELLSA values. After processing of the samples and the readings taken at 405 nm, the duplicate absorbance values were averaged and recorded in a notebook and entered in a computer database. If duplicate values did not agree, the sample was checked again in another pair of duplicate wells. Initially, if values were between 0.00 and 0.10 AU, the plant was considered to be disease-free. Values between 0.10 and 0.50 AU indicated that some virus was present, but often the plants recovered from this low titer of virus and were found to be free of virus at the next time of assay. Plants were always reassayed after having ELISA values in this range, since some were in the process of producing a high virus titer and others were able to overcome the infection. Plants which had ELISA values from 0.50 to 2.55 AU were considered to be infected with virus and were usually discarded after having high virus titer over a period of two time points, unless the plants were needed for breeding. After the earliest screening of tomato somaclones, a regular schedule of ELISA was established so the results could be compared from group to group. Assays were taken at 10, 20, 30, 50, 70 and 150 days (plus or minus one day) until it was determined that the 150 day assay could be eliminated. The later groups (R2-R5 and crosses) were assayed until 70 days post-inoculation (pi) or 30 days pi. QIQEELDQIlIDEEIQD at tgaatoes. Plants were maintained in 8-10" pots in the greenhouse with cool-white fluorescent light provided in a 16 h light cycle. Plants were watered daily with tap water containing Peters 20-20-20 mix fertilizer at the rate of one tablespoon per gallon. Using procedures from T. Kamps (personal communication), pollen from individual plants was collected and stored in petri dishes held at 4° C 23 for two to three weeks. To prevent stray pollination, open blossoms were removed from plants to be pollinated. Flower buds that had not yet begun to open were used for pollination. First, the blossom was opened to expose the anthers, then using extra sharp forceps (Uni-fit size 3, Regine, Switzerland), the anthers were removed, leaving the stigma intact. At this time, the pollen was not yet mature and remained attached to the anthers. After one to two days, mature pollen from fully opened flowers of the desired plant was applied to the stigma. Pollen was applied by touching the stigma to freshly released pollen, three times in five days. Hands and tools were carefully washed in 95% ethanol after handling each kind of pollen to ensure that unintentional pollinations did not take place. Careful records of parents were kept in a notebook and by tags which were placed on each blossom. Fruit set could be observed about seven days after emasculation if the pollinations were successful. After fruit set was detected, fruit ripened in about 45 days. Reciprocal crosses were made using different R1 or R2 plants of each somaclonal line and the susceptible parent line (GCRI-26) until there were 4-5 fruits produced for each of the desired crosses. For screening of the F2 progeny, seed was collected from either uninoculated F1 plants (unselected) or from previously inoculated and screened plants (selected for resistance or susceptibility). Sets of backcrosses were made between F1 crosses and somaclone lines (plants of the R1 or R2 generation) or the susceptible GCRI-26. There was only one complete set of backcrosses, for somaclonal line 247. 24 §££Q agllecttoa. Seeds were collected by cutting open tomato fruits and removing the gel layer from the seeds in one of three ways. Sometimes seeds were placed on a paper towel and scraped back and forth with a razor blade to mechanically remove the gel from the seeds. Most often, seeds were placed in a large test tube along with the juice of the fruit. Concentrated hydrochloric acid was added to about 12% and allowed to sit at room temperature for one to five hours. Seeds were then transferred to a sieve and rinsed in tap water. This acid treatment was also effective in destroying virus on the seed coat (Taylor et a1., 1961). The last method for removal of the seed gel was to place seeds in a test tube and allow the tomato juice to ferment over several days at room temperature. This was only used once, since it was unpleasant to handle the fermented juice. Seeds were moved to a sieve and rinsed in tap water. After the gel was removed, seeds were air dried on a laboratory table for several hours or overnight before packaging. Seed packets were placed in a dessicator for several days and then stored in the refrigerator until ready for use. Seeds collected from all self-pollinated fruits of each plant were pooled, but seeds from each cross-pollinated fruit were kept in a separate envelope. GENETIC ANALYSIS Rating 2f zasiatanae ta tomato somacloaaa. For proper classification of resistant phenotypes, it was essential to devise a rating system which reflected the predicted phenotypes for a given model. The assays employed in the study gave an estimate of virus concentration for each plant over a period of time, therefore the rating systems could incorporate virus titer and/or time. Symptomatology was not useful in rating for resistance for several reasons; the lack of clear mosaic 25 symptoms in susceptible leaves using ToMV-O, the time necessary for other symptoms to appear (such as stunting), and the lack of correlation between symptom formation and virus titer. The last reason can be exemplified by the resistance seen in plants expressing the Tm-l gene. This gene is dominant for symptom suppression but incompletely dominant for limitation of virus multiplication (Fraser and Loughlin, 1980). The following rating systems were evaluated. At the time of screening, both virus titer and time were used in rating resistance because virus titer was measured in the plants at 10, 20, 30, 50 and 70 d pi. The earliest ratings of resistance in somaclones were very strict; plants were considered susceptible if there was detectable virus at any point in time. Later, the plants were grouped into three classes: S- susceptible (virus titer above 0.50 AU at 10 d pi), I- intermediate (virus titer above 0.50 AU at 20 d pi) and R9 resistant (virus titer less than 0.50 AU at 10 and 20 d pi). This non-standardized rating system was the least refined since it was developed during the early part of screening. Using this rating system, there was unexplained variation between replicate groups screened (possibly due to variable expressivity caused by seasonal and environ- mental effects or horticultural practices). Since the ratings were somewhat arbitrarily defined by the choide of assay time (pi), it was felt that perhaps some sensitivity in the measure of resistance was lost. Earlier investigators of TMV-resistance noted similar problems (Phillip et a1., 1965) and realized that the virus responded to resistance over time, with symptoms and virus titer dependent on age of 26 difficult. In another study, Holmes (1954) scored resistance five days after inoculation by examination of the inoculated leaf for symptoms caused by strain 1952D of TMV, which causes lesions on susceptible leaves. By limiting the resistance response to this early point in time, he was able to discern single gene inheritance of what later became known as the Tm-l gene for resistance. In our study, because there was unexplained variation over time, it was decided that 10 d pi was most valid for describing resistance. Susceptible control plants consistently showed a high virus titer even by 7 d pi, so susceptibility to infection would be detectable at 10 d pi. This simplified system used only two classes: S- susceptible (virus titer above 0.50 AU) and R- resistant (virus titer below 0.50 AU) at 10 d pi. However, there was variation in the virus titer in plants rated S, so further classification was possible. ELISA values from 10 d pi were standardized for each microtiter plate so that the range of absorbance values was from 0.00 to 2.00 AU (using purified ToMV-O or maximum control values to standardize). The distribution of these adjusted absorbance values was studied in the form of histograms, and cluster analysis (using NCSS software) was performed. Using data from the screening of each cross and backcross, a rating system was created to classify plants into three discrete groups. Plants with ELISA values between 0.0 and 0.5 AU were considered resistant (LP low), plants with values between 0.5 and 1.3 AU were intermediate (MP medium) and plants with values from 1.3 to 2.0 AU were considered susceptible (H- high). This classification accounted for all but 6.6% of the variation in the cluster analysis, and correlated visually with groupings from the histograms. Of the susceptible 27 controls, 89.2% were rated "H" and of the self-pollinated somaclonal lines, 97.4% of the plants were rated "L", thus giving confidence in this rating system. As another measure of resistance, standardized ELISA absorbance values were used to calculate the mean, variance and standard error for each line and each cross or backcross. These values were also used for analysis of variance and multiple range tests. While these means correlated well with overall levels of resistance, they did not describe the discrete phenotypes seen. In the genetic analysis of a complex trait like viral resistance, it was important to consider all of the rating systems to evaluate proposed genetic models. groaaea between somacloaal liaaa aaa aasgeptibla tamata linaa. To determine the mode of inheritance, resistant R1 plants were used to make reciprocal crosses to the original parent line, GCRI-26. Screening of the progeny (F1) would reveal either non-Mendelian inheritance or heterozygosity of resistance in the R1. If the reciprocal progenies were not identical in the level of resistance seen, a cytoplasmic factor would be indicated. It is possible that the somaclones (R) were heterozygous for resistance gene(s), which would be seen as phenotypic variation between F1 families in which different R1 parent plants were used to obtain the F1 and segregation within F1 families. The F1 progeny were then self-pollinated to produce the F2 generation. Because some F2 were from uninoculated plants and others were from previously screened plants, the possibility of inadvertant selection for resistance was examined. A comparison between unselected and selected F2 plants was made using analysis of variance and multiple range tests. 28 Outcrosses of the somaclonal lines were made to another susceptible tomato line, Bonny Best, to determine levels of expression in a different genetic background. Both F1 and F2 lines were screened and rated for resistance. Reciprocal crosses were made of the F1 plants back to both parents (somaclones and GCRI-26) to generate a set of backcrosses. There are eight possible backcross combinations, which were completed where possible and progeny were screened for viral resistance. Saga ttanamtaatan at xttaa. Because seed transmission of the virus could alter the results of the genetic screening, the amount of seed transmission of ToMV-O was determined in GCRI-26 control plants and in some somaclonal crossing lines. Taylor et a1. (1961) showed that seed transmission of ToMV in tomato occurred mainly by handling the seedlings during transplanting, with inadvertant contamination by virus on the seed coat. Only a small percentage of seeds were found with virus in the endosperm and none were detected with virus in the embryo. They found that the virus present on the seed coat was effectively removed by acid treatment of the seed (25 m1 HCl per 5 1b fruit for 3 h, estimated to be about 1% v/v HCL). To test virus transmission in susceptible tomatoes, 60 seeds of GCRI-26 were soaked overnight in ToMV inoculum (infected sap diluted 1:20 with 0.01 M phosphate buffer, pH 7.0). Seeds were planted (one per square) and allowed to germinate undisturbed. At 18 d the plants were sampled and ELISA values determined. As a control to detect other sources of virus contamination, 30 uninoculated seeds were similarly tested. 29 In another experiment, ELISA values were determined for virus- infected fruit-bearing plants, fruit juice and seed from the fruit after acid washing. These were compared with healthy plants, fruit and seed. Seedlings germinated from the seed of infected fruit were tested for virus using ELISA. There was also indirect evidence from experiments designed to study virus movement in susceptible tomatoes. Seeds from infected plants were germinated and transplanted to 3" pots. After 25-30 d representative plants were sampled for ELISA to ensure that plants were virus-free before inoculation. Czaaaaa 21th knagn xaaiatanaa ganaa. To test for allelism between the somaclones and other characterized resistance genes, reciprocal crosses were completed and the F1 progeny were screened and rated for resistance. If high levels of resistance were maintained, with the genes being additive for resistance, this would serve as preliminary evidence for the genes being alleles of the same locus. If less resistance was observed, then the genes would not be considered additive or allelic. The plants used for these reciprocal crosses were near-isogenic lines of ‘Craigella' which were homozygous for resistance genes: GCRI-237 (Tm-l/Tm-l), GCRI-236 (Tm-Z/Tm-Z) and GCRI- 267 (Tm-ZZ/Tm-ZZ). As additional controls, reciprocal crosses were made between these homozygous lines to GCRI-26 (+/+). To test for allelism between the different somaclones, crosses were made between them. Again, additivity of the genes to achieve a high level of resistance would indicate allelism, however, these plant lines have not been screened to date. 30 MOLECULAR GENETICS Patttiaatiaa aaa testtictioa analysis at chlgtoplast DNA (cppNA). Procedures were adapted from Kut and Flick (1986), for in-organello restriction digestion of DNA in purified chloroplasts. Leaves of plants grown under cool-white fluorescent light supplying 60 uEm'zs"1 with a 16 hour photoperiod were picked, weighed and refrigerated overnight in the dark. Five to ten g of young leaves were homogenized in a cold mini- blender in 40 ml isolation buffer which consisted of 0.35 M sorbitol, 50 mM Tris-HC1 pH 8.0, 5 mM EDTA, 0.1% bovine serum albumin (BSA) (w/v), 15 mM 2-mercaptoethanol, 1 mM spermine and 1 mM spermidine (both from Sigma). Four bursts of 1-5 sec in the mini-blender were adequate to homogenize the leaf tissue. The homogenate was pressed through a double layer of cheesecloth and filtered through two layers of Miracloth into a 50 ml centrifuge tube on ice. All of the following steps were performed at 4° C. To pellet nuclei and whole cells, the homogenate was centrifuged by bringing the speed up to 3000 rpm in a Sorvall type SS-34 rotor (1000 x g) and immediately allowing the rotor to slow down. The supernatant fluid was transferred to 50 m1 tubes and centrifuged at 2500 rpm in 88-34 rotor (750 x g) for 10 min to pellet the chloroplasts. After discarding the supernatant, the pellet was gently resuspended with an artist's paintbrush in 2 ml isolation buffer. The Chloroplasts were further purified on a step gradient, using 12 ml 30% w/v sucrose layered onto 20 ml 60% w/v sucrose (both in isolation buffer). Each sample was divided into two fractions which were loaded onto two gradients. The gradients were centrifuged for 30 min at 26,000 rpm at 4° C in a Sorvall AH-629 swinging bucket rotor (90,000 x g). The chloroplasts which banded at the 30%-60% sucrose interphase were removed 31 with a pasteur pipette and the two fractions of each sample were combined. Chloroplasts were washed by the addition of 15 m1 isolation buffer followed by centrifugation at 2500 rpm in an 88-34 rotor (750 x g) for 10 min to collect the purified chloroplasts. The pellet was resuspended in 2.5 ml isolation buffer and divided between 5 Eppendorf tubes (0.5 ml each). These fractions were centrifuged at full speed on a Beckman Microfuge E (15,850 x g) for one min and the resulting supernatant was discarded. The pellet was saved. The chloroplast envelopes were made more permeable by the addition of 100 ul swelling buffer (0.2 M NaCl, 20 mM MgC12, 10 mM Tris-HCl pH 7.8 and 0.01% nuclease-free BSA) per 5-10 g starting material and incubated at 37° C for 15 min. After swelling, 100 units (30-50 U/g fresh weight recommended) of the desired restriction enzyme was added to each tube and incubated for 1-2 hours at 370 C. The digestion was stopped by the addition of 20 ul 10% sodium dodecyl sulfate (SDS). To each tube, 0.18 g of CsCl was added and gently mixed, and warmed to 500 C for several min. The tubes were then centrifuged at full speed (15,850 x g) for 90 s at room temperature. The clear layer below the green pellicle was removed using a pasteur pipette and transferred to a microfuge tube containing 0.4 ml sterile water. This was centrifuged at full speed for 5 min at 40 C to remove insoluble material, and the supernatant was divided into two fractions and moved into new microfuge tubes. To each tube, 0.5 ml ice-cold 95% ethanol was added and the precipitation of the DNA was allowed to proceed at -200 C overnight. The samples were centrifuged at full speed for 10 min at 40 C and pellets were washed with 0.5 ml 80% ice-cold ethanol and centrifuged 32 again for 5 min at 40 C. Pellets were dried in a vacuum dessicator and resuspended in 10 ul TE (10 mM Tris, 1 mM EDTA). Duplicate samples were combined after resuspension. Gal eleattapboresis. Samples were prepared for electrophoresis by the addition of gel loading buffer (0.25% bromophenol blue, 0.25% xylene cyanol and 25% Ficoll, type 400 in H20) containing 1 ul RNase A (Sigma) (Maniatis et a1., 1982). Digestion of RNA was allowed to proceed for 10 min at room temperature. The samples were loaded in wells of a 0.6% agarose gel containing 0.5 ug/ml ethidium bromide in TAE buffer (0.04 M Tris-acetate pH 7.8 and 1 mM EDTA), and electrophoresed for 14 h at 30 mA constant current for a 25 on gel, or 1-3 h at 50 mA for a 10 cm gel. Size markers consisting of lambda DNA restricted with Hind III were included for each gel, with fragments that were 23.13 kb, 9.42, 6.7, 4.35, 2.32, 2.02 and 0.564 kb. Exttaatlan ana analyaia at aouble-sttaaaea RNA LQEEHAI- Procedures for the extraction of dsRNA were adapted from the protocol described by Fulbright et a1. (1983) and modified for use in higher plants by Haufler and Fulbright (personal communication). Fully expanded leaves were picked from plants grown in the laboratory under cool-white fluorescent lights supplying 60 uEm'zs'1 with a 16 h photoperiod. For positive controls which served as size markers, samples from tomato plants infected with ToMV-O were extracted as well as samples from healthy pepper plants (Qapaiaam aaaaam cv. ‘California Wonder') that have endogenous dsRNA (l4 kd) as determined by A. Dodds (personal communication). The leaves were either used immediately or were wrapped in aluminum foil and stored at -200 C until use. For each sample, 5-10 g of leaf tissue was placed in a chilled 33 mortar and was pulverized with a pestle in the presence of liquid nitrogen. The powder was transferred to a chilled centrifuge bottle or tube and four volumes extraction buffer (2 mM Tris-HCl pH 6.8, 20 mM NaCl, 0.2 mM EDTA, 3% SDS, 0.5% 2-mercaptoethanol and 0.5 mg/ml fractionated bentonite) was added. To this was added four volumes of STE-saturated phenol and two volumes of chloroform. (STE buffer consisted of 10 mM Tris-HCl pH 6.8, 100 mM NaCl and 1 mM EDTA.) This mixture was incubated in a shaking ice bath for 30 min. The mixture was then centrifuged at 7000 rpm in an 88-34 rotor (6000 x g) for 15 min. The upper, aqueous layer was taken by pipet and was filtered through two layers of cheesecloth. To adjust the ethanol content to 15-18%, 0.2 volumes of 95% ethanol was added. Double-stranded RNA was selectively bound to a column consisting of Whatman CF-ll cellulose (Anspec Co., Ann Arbor, MI) in the presence of 15-18% ethanol. To prepare this column for samples of up to 20 g starting material, 2.5 g of CF-ll cellulose was mixed with 25 ml 85% STE:15% EtOH. The slurry was poured into a 50 ml syringe which was plugged with a disk of Miracloth. Samples (up to 160 ml each) were slowly added to the column down a glass rod, and were rinsed with a total of 120 ml 85% STE:15% EtOH added in 20 m1 aliquots. Each aliquot was allowed to drain before adding the next. The dsRNA fraction was then eluted with 1x STE by adding 1 ml, 1 ml, 6 ml and 6 m1 portions. The eluate was collected in a 50 ml centrifuge tube, and 28 ml 95% EtOH and 0.7 m1 M 2.5 sodium acetate (pH 5.2) were added to precipitate the dsRNA. The extracted dsRNA was placed at -200 C overnight to precipitate it. The extract was centrifuged at 9000 rpm in an SS-34 34 rotor (9,800 x g) for 30 minutes. The pellets were resuspended in 0.5 ml 1x STE and transferred to microfuge tubes. The nucleic acids were precipitated again by the addition of 0.5 ml 95% EtOH and 50 ul 2.5 M sodium acetate and cooled to -20° C overnight. The supernatant was removed after centrifuging for 15 min at 4° C at full speed (15,850 x g) in a Beckman microfuge. The pellet was dried and resuspended in 0.1 m1 loading buffer (1x TAE, 0.25% bromophenol blue, 30% glycerol) and was electrophoresed on a 0.6% agarose gel (using 0.5 ug/ml ethidium bromide in 1x TAE) as described previously. Photography. Photography of DNA or dsRNA gels was done using a Polaroid MP-4 Land camera (equipped with filters 2B and 23A) set up with a Foto UV 300 DNA Transilluminator (Fotodyne, Inc., New Berlin, WI). For routine records of gels, Polaroid type 57 sheet film was used with f5.6 and 1/2 8 exposure. For higher quality photographs plus negatives, type 55 Polaroid sheet film was used, with a typical exposure of 2 min and an aperture setting of f5.6. 35 RESULTS AND DISCUSSION ARE SOMACLONES NORMAL PLANTS? fitaaa matphalagy. There were no notable differences in gross morphology between the somaclonal lines and the parental line, except after the plants had been inoculated with virus. After inoculation, the parental line became stunted and showed symptoms of virus infection, including mosaic and strapped leaves, but the resistant somaclonal lines did not have the same diseased appearance. However, dwarf plants were apparent among both the uninoculated parental line and the uninoculated somaclonal lines at the approximate occurrence of one in 16 plants. QBIQEQEQEE namaat. There was no difference in chromosome number between the somaclones and the parental line. All lines that were examined had the normal (Zn-24) number of chromosomes (Rick, 1974). No determina- tions were done for somaclone lines 322 and 330. Stag gatmiaatian xata. Statistical comparison of seed germination rates in Table 2 showed that there was no significant difference in germina- tion between somaclonal lines and the parental line, GCRI-26, when tested using the Student-Newman-Keuls (SNK) multiple range test (p-0.05). {align ziahtltty. The results of pollen germination studies (Table 3) indicated that there was no significant difference between the somaclones and the parental line, GCRI-26, using the SNK multiple range test (p-0.05). Though pollen viability was normal, somaclone line 12 did not produce as many seeds as the other lines of tomatoes (quantitative data not collected). 36 Table 2. Comparisons of seed germination rates. Rates are compared among each generation of selfed tomato somaclones, the parental line, GCRI-26, and the isogenic line, GCRI-237, which is homozygous for the Tm- 1 gene. Replications Seed % (Environmental Generations Line germination S.E. groups) Rl-R3 Sc 12 82.6 4.52 n-l9 Rl’R4 Sc 215 73.8 5.61 n-23 Rl-R3 Sc 219 70.8 7.15 n-17 Rl'RS Sc 247 82.5 3.14 n-34 Rl-R3 Sc 322 81.3 4.62 n-15 Rl-R3 Sc 330 78.7 5.92 n~17 -- GCRI-26 88.48 2.60 n-29 -- GCRI-237 84.3 7.68 n-10 anot significantly different from the other values (p-0.05) using the Student-Newman-Keuls multiple range test. any of the lines. No differences were seen among 37 Table 3. Comparison of pollen germination rates among pollen samples from somaclonal tomato lines and the parental line, GCRI-26. Pollen % # plants Line germination S.E. examined Sc 12 74.2 2.66 n-3 Sc 215 66.5 7.43 n-9 Sc 219 77.8 5.68 n-5 Sc 247 68.3 6.33 n-8 Sc 322 88.6 6.18 n-3 Sc 330 58.5 6.58 n-8 GCRI-26 78.4a 3 47 n-16 {.15.}; """"""" éiéé'éfigi """"""""""""""""""""""" £1not significantly different from the other values (p-0.05) using the Student-Newman-Keuls multiple range test. No differences were seen among any of the lines. 38 SCREENING FOR RESISTANCE TO TMV fiaitaatltty at tesistance ratings. Because of the complexity of the plant-virus interaction in studies of resistance, it is important to be accurate in the description of susceptible and resistant plant phenotypes. The three approaches to rating plants for resistance are compared in Table 4, for susceptible controls (+/+), resistant controls (Tm-I/Tm-l) and resistant somaclonal R1 plants. For the susceptible control line, GCRI-26, the S:I:R and S:R ratings indicated almost complete susceptibility of individuals, and Table 4. Comparison of resistance rating systems, for susceptible and resistant control lines. S:I:R ratings are S-high virus titer at 10 d (pi), I-high virus titer at 20 d and R-no or low virus titer at 10 and 20 d. S:R ratings are S-high virus titer at 10 d and ano or low virus titer at 10 d. H:M:L ratings are H-high virus titer (1.3-2.0 AU), M-medium virus titer (0.5-1.3 AU) and Lnlow virus titer (0.0-0.5 AU) all at 10 d. Mean ELISA value (AU) was calculated from standardized values. Pooled Plant line GCRI-26 GCRI-237 Somaclone R1 Genotype +/+ Tm-l/Tm-l unknown S:I:R (%) 99:1:0 7:17:76 3:12:85 S:R (%) 99:1 7:93 3:97 H:M:L (%) 89:10:1 4:2:94 0:4:96 ELISA mean (10 d) 1.755 0.165 0.043 Standard error 0.015 0.050 0.011 Variance 0.097 0.178 0.026 Sample number 425 70 224 39 the H:M:L rating focused on the variation seen in levels of virus detected among the plants. With each rating system, 99% of the GCRI-26 plants had allowed virus to accumulate to levels above 0.50 AU by 10 d pi, indicating nearly complete penetrance of the susceptible trait. Figure 2A details the distribution of ELISA values of these susceptible plants. Most plants had a virus titer above 1.50 AU, but 11% of the plants accumulated less virus, indicating there was some variation in the expressivity of susceptibility, most probably reflecting subtle differences in the genetic backgrounds of the inbred GCRI-26 plants. The mean ELISA value is quite high (on a scale of 0.0 to 2.0 AU) with a fairly low standard error and variance. The resistance trait seen in GCRI-237, the homozygous tomato line expressing the Tm-l gene, was not expressed with complete penetrance. The S:I:R and S:R ratings indicated 93% of the plants were resistant at 10 d pi, and that 17% more of the plants accumulated virus by 20 d pi. This is in agreement with the description of this resistance as causing a delay in virus multiplication and having a tolerance to low levels of virus (Fraser and Loughlin, 1980). The H:M:L rating also supports this, as does the ELISA mean. The ELISA mean indicates that there is a very low level of virus present in the population, with a larger variance and standard deviation seen than in the susceptible controls, reflecting the wider variation in expressivity of Tm-l. Figure 28 details the distribution of ELISA values seen among plants of this genotype. Resistance in the somaclones was more consistent. Most of the R1 somaclones screened (97%) were free of virus at 10 days pi and were often still free of virus at 150 d. In each rating system, there was 96-97% penetrance of the resistance trait seen in the somaclonal 4O A 100%, = 1 com-23 (-/-I .2 cm 4.3 a '3 a «mat o I °‘ / - 1 1 0 40st // a ' l a / * o 20%» 1 1 u 3. _ m ,; . _ Q, . _...,+ 0‘ 1 a 1 a . 1 ‘ ”L '. ”a; ' o .10 .33 .so 37 ..‘a {a 1.10 133 m 1.07 Hm 2.0 ELISA values (AU at 405 nm) B l00$ l 1 com-237 (Tm-1/‘I’m-1) l S .05} OF. H l a ' I "" l 5 00% t 1 8‘ ’ l G. . c... 40‘ "' \ o - . u '4 l 5 20% “I ‘ l‘ § ' \ Q: 0‘ m 1 J 1 I A n 1 o .13 .33 .so .57 .u 1.0 no us use 157 min ELISA values (AU at 405 ml Somaclone R1 1 Percent of Population I.” .33 .50 .07 .83 1.0 1.“ 1.33 1.60 1G7 1.;3 2.0 ELISA value (AU at 405 nm) Figure 2. Distribution of ELISA values at 10 d pi for the following genotypes: A) susceptible +/+ (GCRI-26) B) resistant Tm-I/Tm-l (GCRI- 237) and C) resistant (somaclonal R1). 41 progeny. The mean ELISA value was very low on the scale of 0.0 to 2.0 AU, with a low standard error and low variance. Examination of the distribution of ELISA values in Figure 2C allows one to conclude that there was little variation in expression of this resistance, and that the plants which were rated as being susceptible did not accumulate as much virus as did more than 90% of those which were susceptible controls. With this in mind, it can be stated that there was almost complete penetrance of this resistance trait and only 4% of the plants showed variation in the expressivity. When compared to the resistance of the Tm-l gene, the somaclonal resistance was stronger in every rating system. When standard analysis of variance was tabulated for each somaclonal R1 and GCRI-237, it was determined that there was a significant difference (5% level) between the mean ELISA values (F-3.83 with 6 and 287 df). Multiple range tests showed that GCRI-237 was significantly different from each of somaclonal R1 lines (except Sc 12 which had fewer samples). Excluding Sc 12, there were no differences among the somaclonal R1 lines. Each of the rating systems accurately reflected the genotypes of the control lines. The S:I:R rating system, which incorporates a longer period of time, is useful as a more rigorous test of resistance. The H:M:L rating system is more fine-tuned, and very descriptive when paired with the mean of ELISA values. Later, its usefulness in describing intermediate phenotypes will be apparent. Uglitahilitx 2f taatataaaa. When the progeny generations of somaclones were inoculated and screened for the presence of ToMV-O, it was clear that the resistance was heritable. The results of screening for all somaclone generations are summarized and presented in Table 5 using the S:I:R (individuals) rating system. For each somaclonal line, at least 42 Table 5. Resistance ratings of selfed generations of somaclonal lines. Data were compiled for up to five generations for each somaclonal line using the S:I:R rating system which classified plants in three groups according to the length of time the plants remained free of detectable virus. 8 (susceptible) were plants with virus detected at 10 d pi, I (intermediate) were plants with virus detected at 20 d pi and R (resistant) were plants with no detectable virus at 20 d pi. Each entry represents plants which were screened in separate groups, which were totalled at the bottom. * indicates groups which were inoculated with TMV-Fl instead of ToMV-0. 43 Table 5 . TOTAL 5:17:183 *0:0:ll 0:0:12 0:2:10 Sc 12 0:1:15 0:0:12 0:2:10 0:0:12 0:0:16 :0:9 *0:0:5 *0:0:16 :7 0:7:12 0 3:2 0:2:10 o . 1:1 :0'1 :44 :0 0:8:68 3:9:71 Sc 215 19:16:204 1:1:10 0:0:12 0: 0:2:10 *0:0:13 1 0:0:18 :4 *2:1:5 2:2:20 0:0:12 0 1 0:0:12 :4 :2 1:1:10 :6 0:0:12 12:0:0 1:10:54 033:9 0:0:12 0.0.5 9.9.12 0.0.12 3:1:56 Sc 219 13:3:74 127 3:16: 0:2 0:0:18 *1:0:6 0'2'13 :9 0:1:11 0:4:20 *0:4:2 0:0:14 :2 0:0:12 0:0'12 0:1:64 12 3:11:43 So 247 4:17:446 0:1:11 0:0:12 0:0:11 0:0:12 0:0:12 0:0:12 *0:0:6 1:0:11 '0:0:14 0:1:11 0 0:1:2 0:0:12 0:0:14 .0. 0:0:12 0:0:12 0:0:12 0:0:12 0:2:36 0:0:12 :7 :0 :3 0:0:16 0:0:12 0:1:11 0 :0 0:0:12 0:0:18 0:0:71 :0:8 12 0:2:10 :7 1:4:99 1° 00 0:0:12 0:1:11 12 2:8:152 1:25:116 1:3 0:0:12 *0:4:12 Sc 322 :0 0:0:12 0:0:12 0:0:12 :3: :1 7 *0:5:5 010262 0:22:46 13:24:131 1: 0:0:14 Sc 330 *0:0:18 1:1:10 0:0:12 5: :5 :3 *0:5:3 5:2 18 0:0:10 4:6: :3:9 5:6:59 4:12:54 44 three generations were screened for resistance. Examination of the data from each somaclonal line indicated that the virus resistance was consistently expressed in the progeny, with the conclusion that the resistance was genetically stable. Since the trait was heritable, this eliminated the possibility that the resistance trait of the somaclones ‘was epigenetic. The consistency of the resistance through three or more generations, averaging 97% penetrance, supports the idea that the resistance trait is true breeding. If the resistance was caused by a single gene, it would be homozygous. Segregation between resistant and susceptible phenotypes would be evident in the progeny if there was a single gene in the heterozygous state. If the resistance was inherited multigenically, it could be possible that the original somaclones were heterozygous for resistance genes, but only if each combination of genes was always expressed as a resistant phenotype in the subsequent generations. Further examination of this issue will be done as the crossing data are discussed. Qaaaaa,a£ inaamplata paaattanaa. As noted earlier, only 97% of the somaclonal R1 lines expressed full resistance. Highly resistant plants were selected to bear fruit in most cases, but when less resistant plants (in the I category) were used for seed production, their progeny were not less resistant, indicating that the I phenotype did not reflect the genotype in every case. In these less resistant plants, resistance was lost at an early time (20 to 30 d pi), similar to loss of resistance conferred by the Tm-l gene. High virus titer was detected by ELISA and the virus was infective, causing local lesions when inoculated on N. glutinosa. To determine if the genotype was less resistant, seed were collected from 45 the plants and resulting progeny were screened by virus inoculation and ELISA. The progeny in each case were resistant. These results are presented in a detailed pedigree for each somaclonal line in Figure 3, using the S:I:R (individuals) rating system. Cases in which the progeny of infected plants were resistant are marked with #. The resistance trait was predominantly true-breeding. The most striking exception can be seen when examining the pedigree for Sc 215 (Figure 3), which had one group of seedlings from one fruit that were all infected with virus at 10 d pi, yet siblings screened in another group were resistant (marked by #). When progeny of an infected plant from the first group were inoculated with ToMV and screened, they were virus-free and rated as resistant (marked by S). This phenotypic variability with seemingly constant genotype (incomplete penetrance) suggested that perhaps there were other conditions influencing phenotype expression. This could have been due to a less resistant genetic background, or to mutation of the virus to a strain which could overcome the resistance. ,Inflaanaa at Etta; an no e. Evidence to support the theory that the virus had mutated to overcome the resistance was strong. Leaves from putatively resistant tomato plants that contained high virus titer were collected and examined. The symptoms were noticably different from the common strain, with a more yellow mosaic pattern. When the leaves were ground in phosphate buffer, and the sap was used to inoculate a group of resistant somaclonal seedlings, it was found that none of the resistant somaclones escaped infection at 10 d p.i. This was an indication that the virus had become more virulent with passage through the resistant somaclones. Zitter and Murakishi (1969) reported a 46 Figure 3. Pedigrees and resistance ratings of all generations of tomato somaclones screened. The S:I:R ratings classified plants in three groups according to the length of time the plants remained free of detectable virus. S (susceptible) were plants with virus detected at 10 d pi, I (intermediate) were plants with virus detected at 20 d pi and R (resistant) were plants with no detectable virus at 20 d. # indicates groups that were selected from seeds of parents with an S or I rating. * indicates groups which were inoculated with TMV-Flavum (all other plants were inoculated with ToMV-O). Upper- and lower-case letters indicate the groups of plants which were screened together. After the R1 generation, each subsequent generation was labelled with a number indicating the plant which parented the generation. The numbers designating the plant lines are underlined. The S:I:R ratios (representing individuals) are given for each group screened, except in the cases marked with "un" which means the group was uninoculated. 47 SOMACLONE 12 J I I I I I 1'2 A 3.2 :A 2 0:0:11 0:0:9 0 0 5 0:0:16 2:0:3 l L I I l l I I 12-7 12-9 2-8 * 2- '12-}; * - 4 0:0:12 3'2z7 0:7:12 0:5:5 0:3:6 0:4:8 0:0:12 0:0:12 0:0:16 12'7-2 12-7-8 lz-B-Q 12-8‘10 0:2:10 0:0:12 0:2:10 0:2:10 SOMACLONE 215 J l I l I I l I I 2.3 2 E 12 .Q E. E Q 0:0:13 (un) 0:0:4 2:1:5 1:0:6 0:0:12 0:0:5 0:0:12 119;: 3.1:;- zis_-§. s-b 2.1223 2121: 0:2:10 0'0'18 0:0:12 1 1:10 12:0:0 0:0:12 ' I 0:0:12 l 215-2'10 215°5-5 215-5-12 gzg-z-gg - -4 0:0:12 0:1:4 0:0:12 1:2:0 0:3:0 L 0:4:8 I I 215-5-5-3t 2 - ° -4 1:1:10 1:1:10 SOMACLONE 219 l I I I I l I I I I! E 2 1.1.3. 2 E SE Q 1:0:6 1:5:2 (un) 0:4:2 1:2:3 0:0:12 0:0:12 0:0:6 19'2 21_9'_l 9'9 11.9.2}. 1E1 0:0:18 0:0:9 0:1:11 0:0:14 0:0:12 Ml 0:2:7 0:2:13 Figure 3. N 48 SOMACLONE 247 l I I I I I I I I I :A E 2 2 A Q 2 Si x. 0:0:14 0:0:12 (un) 0-0:14 0:0:7 1:3:4 0:0:16 0:0:12 0:0:7 1 1 H———. rL——i r 1 1 2247-17 247-1 247-1 347-2 347-7 247-6 247-12 211:1 331;; 0:0:8 0:0:12 0:0:12 0:0:12 0:0:7 0:0:9 0:1:11 0:0:12 0:0:12 0:1:2 2:1:9 0:0:18 0:0:7 0:0:12 0:1:4 0:4:5 I r__1 l I I I I 247-17-4 247-1'3 247'1'9 247-7‘5 247'6'1 Z47'§-§ z47r§'4 Z47-§'§ 247-4'5 0:0:12 0:0:12 0:0:12 1:0:11 0:0:12 0:1:11 0:0:8 0:2:10 0:1:11 | I r I I 247-17-4-3 347-1-3-8 247-1-9-1 247-6-4'3 247-6-4-7 247-4'5-4 0:0:12 0:0:11 0:0:12 0:0:12 0:0:12 0:0:12 | I I 247-17-4-3-5 247- -4- -9 241-6-4-z-lz 0:1:11 0:1:11 0:0:16 SOMACLONE 322 l - I I I I I I I 2 fl 2 .C. 3.2 E 9 (un) 0:4:12 0:3:7 0:7:1 0:5:5 0:3:9 0:0:12 112-21 32 - L221 122;: m 12.239. 0:0:12 0:0:2 0:0:12 0:0:12 0:0:12 0:0:12 | 322-3'11 1:3:8 SOMACLONE 330 l I I I I I I I 13. 2 E 1.12. E. 2 E 0:0:18 (un) 4:3:5 0:5:3 0:3:9 0:1:7 0:0:12 330-4 119;; 330-5 33 - 332;; 339:12 0:0:14 0:0:12 5 5 2 0 0 10 0:1'9 0:0:12 I 1:1:2 1:1:10 2:4:6 Figure 3. (continued) 49 similar occurance in resistant tomatoes. It would therefore be wise to consider virus mutation as a cause of some loss of resistance. A second type of inconsistency in resistant plants was seen in occasional plants that would have a high virus titer at one time of assay (ie. 20 d), but not at the time of the next assay (ie. 30 d). Since only the growing tip of the plant was sampled at each time point, it appeared that the virus which was present at 20 d did not move into the newest leaf at 30 d. There were examples of this phenomenon among all the somaclonal lines, and especially among the F1 lines which were cross-pollinated with GCRI-26. Because of this observation, a series of experiments was initiated to study the movement of the virus in the plants (included in the next chapter of the dissertation). thmatiam. The possibility that the somaclones were chimeric plants was examined by looking at the R1 progeny of each somaclone. For each somaclone, different replications to test R1 plants were taken from different seed lots corresponding to different fruits. Seed lots were kept in separate envelopes. Accidental fire has since destroyed the envelopes and their contents so that identification of fruits and seed lots cannot be done. The possibility that the somaclones were chimeric cannot be completely eliminated, but it is doubtful that chimerism would play a role in the characterization of the somaclonal resistance, since there was no clear evidence of differences between replications of R1 progeny. Campattaan at aamaalaaal liaaa. The S:R ratings at 10 d pi from each somaclonal line were summarized in Table 6 and converted into percent- ages for comparison between somaclones. The mean ELISA value for each somaclonal line and generation was also included, with results of analysis of variance and Fisher's least significant difference (LSD) 50 Table 6. Summary of responses of resistant somaclones in response to inoculation by TMV-O using A) the S:R rating at 10 d pi (S-virus titer greater than 0.50 AU and R-virus titer less than 0.50 AU) or B) the mean ELISA value (with results of LSD range test among somaclonal lines of each generation). F-values from the analysis of variance tables are listed below each generation. A. Percent S:R at 10 days. Line R1 R2 R3 R4 R5 MEAN i5 """" 2'56 '''''' L26; """ 6'i66 """""""""""""""" 3'53"" 215 5 95 14:86 2 98 8 92 8 92 219 5 95 0:100 0:100 2 98 247 1:99 1:99 1:99 0:100 0:100 1:99 322 0:100 0:100 8 92 l 99 330 6 94 7:93 14 86 8 92 ’ """"" 3'59 """" L'éi """" é'éé """" £255 """ 6'5; """" S'éi"" B. Mean ELISA value and results of LSD (p-0.05) between each generation and between somaclones. Means with common letters were not significantly different from each other. Line R1 R2 R3 R4 R5 MEAN* i; """" 6'i5; """ 6’i6L’2Afi3'6'6L; """""""""""""" 6’65L’Eééi 215 0.038 0.267 (c) 0.089 0.114 0.131 (c) 219 0 025 0 021 (a) 0 055 0 034 (a) 247 0.036 0.027 (a) 0.040 0.017 0.038 0.035 (a) 322 0 027 0 009 (a) 0 090 0 042 (ab) 330 0 038 0 245 (be) 0 090 0 124 (c) i';;i;;"iii§'2;;3'"éfé%;; """ iiii'2;;§ """"""""""" £Ti3;;"" (RI'R3) *Total SE-0.006, variance-0.051 **significantly different from tabulated F value (p-0.01) 51 multiple comparison tests indicated. From these data it was possible to ask if the somaclones were similar to each other. In the R1 generation, the F value from analysis of variance was low, indicating that there were not significant differences. In the R2 generation, Sc 215 and Sc 330 differed significantly from the other somaclonal lines, but relatively few plants caused this difference. There were no significant differences seen among somaclonal lines in the R3 generation. Because there were not consistent differences seen between somaclones in the first three generations, there was not conclusive proof that the resistance of the somaclones was the same or that it differed. Each line seemed to breed true for the resistance and showed a high level of penetrance, also indicating that there were not clear differences between the somaclonal lines. This evidence lends support to the bulking of data for selfed somaclonal lines. Further examination of this question was done in the analyses of the crosses. RESULTS OF SCREENING CROSSES The first step in the genetic analysis was to examine the resistance of the F1 progeny of reciprocal crosses between the susceptible parent line, GCRI-26, and each of the somaclone lines. Since each of the somaclonal lines breeds true, that the resistance gene is homozygous is a reasonable assumption to make, although the trait might also be multigenic. When making reciprocal crosses with plants assumed to be homozygous for a trait, the nuclear backgrounds of the progeny will be identical. If differences were seen between progeny of the reciprocal crosses, it would indicate that the trait might be cytoplasmically encoded. 52 Other information that could be learned from the screening of F1 progeny is the degree of dominance of a nuclear genetic trait. If all progeny displayed full resistance, the resistance gene would be dominant, and if none of the progeny displayed resistance, the resistance gene would be recessive. If intermediate levels of resistance were found in the progeny, then incomplete dominance of the gene could be possible, as well as more complex explanations. Raanlta at tacinrocal arossaa ta parental ling. Datattian at I2212£2££l.21§f2£2322§- After inoculation and screening of the F1 progeny, examination of early results indicated that there were differences among the F1 of the reciprocal crosses for the somaclonal lines (Figure 4). When the somaclonal parent lines were used as the female parent (Sc x 26), the offspring showed a high level of resistance, although their resistance to ToMV infection was not as strong as that displayed by the offspring of self-crosses (Sc). In contrast, when the somaclonal plants were used as the male parent (26 x Sc), only a very low level of resistance was evident. But again, susceptibility to the virus was not as complete as that displayed by the susceptible control plants (26). When the reciprocal crosses were compared to each other using analysis of variance of the ELISA means, the F value was very high (33.73, with l and 279 df) confirming the reciprocal differences (using pooled data from all somaclone families). mawmmwwm. It was evident after comparison of ELISA means of reciprocal crosses for each somaclonal line (Table 7) that there were maternally biased differences between the reciprocal crosses in most cases. Results of analysis of variance for reciprocal crosses within each somaclone family 53 10 DAYS [:3 20 DAYS [:1 so DAYS 8025* """"""""""""""" I 60%. 40%>t Percent of Population 20% 4/1)’ . 1'“ ‘1 "171531 :':=5:':'I 80 Line , , . . .. ‘.‘.' .nA, 4 n4-.‘. I:‘( I ~ ‘1 80x26 0%» A ‘ Figure 4. Comparison of virus resistance in F1 of reciprocal crosses between somaclonal lines (pooled) and the susceptible parental line, GCRI-26, scored at 3 stages in their development: 10, 20, and 30 d pi. Table 7. Testing for significant reciprocal differences for each somaclonal line. ELISA means for F1 progeny of reciprocal crosses were used to calculate the analysis of variance. S 6 26 a St # ELISA # ELISA F Line plants mean plants mean value Sc 12 32 0.868 80 0.778 0.47 n s Sc 215 40 0.362 36 0.761 9.86** Sc 219 33 0.538 14 0.866 3.66 n s Sc 247 35 0.246 41 0.804 15.44** Sc 322 10 0.059 24 0.812 20.41** Sc 330 18 0.162 24 0.662 10.38** Pooled 168 0 385 219 0 791 33.73** GCRI-237 12 0.821 12 0.832 0.00 n.s. LIE‘I/+) **Highly significantly different (p-0.01). 54 indicated that the differences were highly significant (p-0.01), except for reciprocal crosses of somaclones 12 and 219. It is also helpful to evaluate the same plants using the S:I:R ratings (Table 8A). Notably, the distribution of plants in the S:I:R ratings showed a considerable maternal bias toward higher resistance among each somaclone family. When ELISA values were used for statistical analysis, the reciprocal differences were not as consistent. Reciprocal crosses made between GCRI-26 and GCRI-237 did not show differences using either rating. Ata thata differances anang £h2 sgmaalonal ltnaa? It was possible to examine the similarity of inheritance of resistance among the somaclonal lines by comparing the S:I:R ratings and ELISA values of F1 progeny, and by calculating the F value from the analysis of variance of ELISA values (Table 8). Evaluation of S:I:R ratings for differences among somaclone families was difficult, so initial comparisons were made using ELISA values. Of the maternal somaclone crosses (Sc x 26), the F value was 5.57 (with 6 and 156 df), indicating very significant differences among the somaclone families (p-0.01). Results of the LSD multiple range test showed that this was due to F1 progeny of somaclonal line 12, which differed significantly from the other somaclonal lines. There was variation between F1 progeny of somaclonal line 219 and lines 322 and 330, and progeny of lines 215 and 247 were intermediate. Overall, these differences were not as great as that of line 12. Line GCRI-237 (Tm-1 gene) was included in the comparison and was shown to be significantly different from all somaclonal lines except line 12. However, when the S:I:R ratings were compared between 12 and 237, different distributions were apparent, thus weakening the significance of this result. 55 Table 8. Summary of results after screening F tomato plants for resistance to ToMV (at 10 d pi). A) S:I:R rat ngs. B) Mean ELISA values. Lower case letters in parentheses following each ELISA mean indicate the results of LSD multiple range testing among families, where means with the same letter are not significantly different from each other (p-0.05). C) Result of analysis of variance among somaclone families. A) S:I:R ratings at 10 d pi. Sc_x_26 26_x_Sc # #- Line plants (8 IzR) plants (S I R) Sc 12 32 53:38:9 80 61:3l:8 Sc 215 40 23:45:32 36 50:50:0 Sc 219 33 30:45:25 14 57:29:14 Sc 247 35 17:40:43 41 68:29:3 Sc 322 10 0:60:40 24 63:33:4 Sc 330 18 11:61:28 24 63:25:12 GCRI-237 12 75:25:0 12 75:25:0 (Tm-l/+) B) Mean ELISA values at 10 d pi. Line ELISA mean ELISA mean Sc 12 0 868 (c) 0.778 Sc 215 0 362 (ab) 0.761 Sc 219 0 538 (b) 0 866 Sc 247 0 246 (ab) 0.804 Sc 322 0 059 (a) 0.812 Sc 330 0 162 (a) 0.662 GCRI-237 (Tm-1) 0.821 (c) 0.832 C) Results of analysis of variance among somaclone families. F value 5.57** 0.32 (n.s.) **Highly significant difference (p-0.01). 56 Of the crosses in which somaclonal lines were the pollen parent (26 x Sc), no significant differences were found among somaclonal lines and GCRI-237, with a very low F value (0.32, with 6 and 211 df). That there was not a significant difference among somaclonal lines and GCRI-237 when these were used as the pollen parent was of interest, perhaps indicating that the mode of inheritance of somaclonal and Tm-l resistance genes is similar. Analysis of ELISA values indicated that, unlike the other somaclonal families, somaclone 12 did not show a maternal effect. But because this was not substantiated by the S:I:R ratings, evidence provided by analysis of further screening is necessary before resistance of somaclone line 12 can be proven to be different from the other lines. In conclusion, there was no clear evidence in the F1 cross data to suggest that the somaclonal lines differ in their inheritance of resistance. However, as a precaution, F1 data from crosses with somaclonal line 12 were not bulked in further analysis. Wefmflstmedusmm mataraal 23.2%- To estimate the maternal effect, the distributions of ELISA values were compared for reciprocal F1 generations (Figure 5). As an example of resistance without cytoplasmic inheritance, progeny of crosses between GCRI-26 and GCRI-237 (Tm-l/Tm-l) were screened for resistance and compared to the somaclonal lines. Plants with the hybrid genotype Tm-l/+ showed intermediate virus titer (Figure 5A), with some variation in expressivity that resulted in a normal distribution of ELISA values, slightly favoring plants with lower titers of virus. Results of the two reciprocal crosses were combined for GCRI-237, since there was no difference observed between them. However, the distribution of ELISA 57 > Tin-II- .o‘L' nuan-ILI17 Percent of Population ............. ..................... I, ‘.’ ”.1. . . l . .. . '. L .' .l.-u:‘.‘.'.‘ l‘ "(J-13...} i... .'.';"":"I .'.'.'.'..‘.‘1' .1' ..".'.. n l 0 ”Al-.IS‘.O0 .07 .08 10 III 183 LOO 107 1.8 1L0 ELISAvaIuoHAUathnIII 100‘ 26xSomacIonoIF1I lllln"0J77 Percent of Population o .. .3. u ,7 . “mt.m .“o ..m m to ELISA value. (AU at 406 Ml) i000 O Somaclone): 28(F1I llll|41381 Percent of Population o J” ”unumt. m to ELISA VIII“ (AU at 406 no) Figure 5. The distribution of ELISA values for virus detected in plants at 10 d pi, from the crosses A) between GCRI-26 (+/+) and GCRI-237 (Tm- I/Tm-l), B) GCRI-26 x somaclone lines 215, 219, 247, 322 and 330, and C) the same somaclone lines x GCRI-26. 58 values from progeny of reciprocal crosses using the somaclonal lines (excluding Sc 12) differed significantly. The distribution of ELISA values from F1 progeny of crosses using the somaclonal lines as the pollen parent (26 x Sc)(Figure SB) resembled that of Tm-1/+ (Figure 5A), with the majority of plants having intermediate ELISA values, but again favoring lower virus titers. In the progeny of the reciprocal cross (Sc x 26) much less virus multiplied, with the distribution skewed toward less virus (Figure 5C). This was also reflected in the ELISA means, of 0.777 AU for 26 x Sc compared to 0.321 AU for Sc x 26. In the more resistant progeny of Sc x 26, 71.2% of the plants displayed a high level of resistance, with ELISA titers of less than 0.50 AU, compared to only 34.1% of the progeny of 26 x Sc, a distribution shift, with 37.1% of the plants with more resistance to virus multiplication. These differences in progeny phenotypes from reciprocal crosses suggested either maternal inheritance or a maternal effect. Maternal inheritance usually means that a trait is encoded by organelle DNA. In tomato the transmission of cytoplasmic genes is usually uniparental maternal (Kirk and Tilney-Bassett, 1978), however, there may be incompletely dominant or multigenic nuclear genes, or other types of cytoplasmic factors responsible for the intermediate levels of resistance seen in the somaclonal lines. For example, an extra- chromosomal plasmid or a cryptic virus could be involved. In this regard, a number of plants have been shown to contain double-stranded RNAs (dsRNA) of no known function or phenotype, including beans (Wakarchuk and Hamilton, 1985) and some varieties of solanaceous plants; pepper, tomato and tobacco (A. Dodds, personal communication; Ikegami and Fraenkel-Conrat, 1979). Because a maternal bias in inheritance was observed in most of the somaclonal lines, it was logical to examine them 59 for the presence of any such unusual nucleic acids which could be correlated to the resistance trait. There was less resistance to virus replication seen in the F1 progeny compared to the somaclonal parent line. Therefore, if there was also a nuclear gene for the resistance trait, resistance was not a dominant allele, rather it was incompletely dominant or additive. If there had been complete dominance, there would not be the notable reduction in resistance in the F1 plants relative to the resistant somaclone lines, unless it was heterozygous. Kata tha aziginal somaclonea homozygoua £21 tna taatatanaa trait? If the original somaclones were homozygous for the resistance trait, then variation among groups of F1 progeny should mostly be due to environmental differences. Because there seemed to be variation among groups of siblings screened at different times (termed environmental groups), the variation attributed to environment needed to be estimated and compared to the variation among somaclonal families. Detailed results of screening the F1 seedlings from cross- pollination can be studied by referring to Table 9. Close examination of these data allowed identification of variation among environmental groups with the same parents, which are marked by a (+) to the left of the rating. These replicates were not necessarily from seed of the same fruit. Results of screening are presented using the 10 d pi S:R rating system and the ELISA means. Using the ELISA values, LSD range tests were employed to identify ELISA means of environmental groups which were significantly different from each other within somaclonal families (p-0.05), and are identified by the lower case letters to the right of the ELISA means. ELISA means with letters in common were not significantly different. It was clear that there was much variation due 60 Table 9. Summary of S:R resistance ratings and mean ELISA values (from 10 d pi) of F1 progeny of reciprocal crosses between each somaclonal line and the parental line, GCRI-26. The letters in parentheses denote the groups of plants that were screened in the same environmental group. Brackets and (+) indicate that the crosses were made using the same R1 parent plants. Within each environmental group, F1 plants were siblings. Small letters to the right indicate groups determined by LSD multiple range tests (p-0.05). Sc x 26 26 x Sc S:R (10 d) ELISA S:R (10 d) ELISA Line rating mean rating mean Somaclone 12 3:3(I) 0.908 a [+10:2(I) 0.659 a 6:8(1) 0.738 a + 7:5(2) 0.659 a §ta(BB) 0 764 a +10:2(N) 1.069 b 17:15 0.868 I.» 9:3(1) 1.284 b (53:47%) + 3:5(Q) 0.500 a 5:7(H) 0.641 a §t1(BB) 9,414 a 49:31 0.778 n.s (61:39%) **sig Somaclone 215 0:12(I) 0.000 a + 5:7(H) 0.570 a [+3:3(Q) 0.507 b I+ 9:3(I) 1.156 b +2:4(Q) 0.493 b [+ 1:5(Q) 0.439 a 232(2) b + 331(Q) ab 9:31 0.362 18:18 0.761 (23:77%) **sig (50:50%) *sig. Somaclone 219 3:3(N) 0.923 b [+ 8:4(N) 1.006 a 2:4(Q) 0.343 ab + Q;2(dl) thzg b +1:5(Q) 0.286 a 8:6 0.516 [+1:7(d1) 0.308 a (57:43%) +§1&(d2) 0.853 ab 10:23 0.543 (30:70%) n.s. **sig Somaclone 247 0:12(Q) 0.000 a [+ 4:5(H) 0.328 a 0:5(T) 0.000 ab + 3:5(I) 0.500 a 4:2(2) 0.749 c 11:1(N) 0.963 b 2419(33) 9..th b 121033) L222 b 6:29 0.246 28:13 0.804 (l7:83%) **sig (68:32%) **sig Somaclone 322 0:3(Q) 0.193 a [+ 5:1(N) 0.911 b Qt1(Q) 0,092 a + 4:8(2) 0.490 a 0:10 0.059 §;Q(N) ltfifia b (0:100%) 15:9 0.812 n.s (63:37%) **sig Somaclone 330 +0:9(S) 0.043 a 11:1(N) 0.977 a [+2:4(2) 0.427 a mm 2.312 b Qt3(S) QtQQQ a 15:9 0.662 2:ll:5 0.162 (63:37%) (11:89%) n.s. **sig 61 to environmental group, but results were inconclusive when assessing variation among somaclonal families. There was no clear evidence to determine whether or not the original somaclones were heterozygous. If the somaclones were heterozygous, the R1 progeny would segregate into different genotypes (ie: RR, Rr and rr), thus giving F1 progeny that would show differing segregation patterns when homozygous or heterozygous R1 parents were used in crosses. Families from one parent should not show differing segregation patterns, but progeny from differing parents might. In Table 9 there were as many differences among groups of siblings as there were among groups that were not siblings, regardless of environmental group. fiatinatian at anvironnental attacta. The results of LSD multiple range testing (p-0.05) are presented in Table 10A for comparisons between environmental groups screened. Results of multiple range tests of groups (designated by lower case letters in parentheses) show that there was significant variation among ELISA means. This environmental variation is probably refected in the broad distribution of ELISA values seen in Figure 5. Because the experimental design gave an unbalanced case, analysis of variance for interactions between environmental group and somaclone families could not be calculated. Instead, the data were analyzed for variation of pooled somaclone families within each environmental group (Table 10B) and for variation of environmental groups within each somaclone family (Table 10C). The results of these analyses of variance suggested that of the large variation between environmental groups seen in Table 10A was responsible for much of the total variation. For the Sc x 26 families, the significant variation was due to Sc 215 in Group I, and Sc 247 in 62 Table 10. Results of analysis of variance and LSD multiple range tests (p-0.05) of ELISA means of F1 progeny of pooled somaclone families. A) Among environmental groups. Groups were identified by upper and lower case letters which were assigned in alphabetical order at the time of screening. B) Among somaclone F families within environmental groups. C) Among somaclone F1 fami ies. A. Variation in pooled somaclone families among environmental groups. c 5 22 EnvironmenEaI # ELISA nv ronmen a ELISA Group plants mean (LSD) Group plants mean (LSD) T 0.000 a d 0.026 a S 12 0.030 a Z 36 .507 a Q 46 0.226 a H 33 0.530 a BB 12 0.344 a Q 20 0.534 a I 32 0.493 bc 1 44 0.936 b d 14 0.564 bc N 59 1.031 b 2 24 0.586 bc BB 12 1.204 b N 6 0.923 c) P value between 3.71* 9.54** gnVi I gnmental groups B. Variation among pooled somaclone F families within environmental groups (in groups where comparison can be made). §a_x_2_6 Sc 263559. c Group lines plants F value Group lines plants F value "i """ i2 """" 26 """"""" i """ i2 """" ii """""" 215 12 11-2°** 215 12 2 68 247 8 (n s ) N 219 6 '--- N 12 12 219 12 247 12 0 41 322 12 (n s ) 330 12 """" """" Q 2. 2 .0 n s 247 12 3 13* 322 10 "i ...... 2i§ ...... i2 """"""""" 2"- 12 12 247 6 (0-93 322 12 (i :5) 329 6 “-S°) 3:9 12. ° ° C. Variation among somaclone F1 families among different environmental groups. #22. 2233552 Line plants groups F value plants groups F value Sc 12 '-"52 ' 2 0.18 80 6 4.11** Sc 215 40 9.01** 36 3 2.99* Sc 219 33 3 2.52 14 2 12.52** Sc 247 35 4 6.15** 41 4 6.81** Sc 332 10 2 3.86 24 2 10.94** Sc 330 18 2 2.28 24 2 18.32** *"Significant differences (p-U.U§). **Highly significant differences (p-0.01). 63 Group Q. In each case, only one set of progeny differed from the others, possibly displaying differences due to heterozygous somaclones. However, this was not seen consistently in all somaclone families (possibly because of the relatively low sample numbers). Among the 26 x Sc families, there was not much variation among the families within environmental groups (Table 10B), rather the variation seemed to be among F1 families (Table 10C). That there were not differences among the somaclone families, but there were differences within the somaclonal lines indicates one of two things. One possibility is that each somaclone was heterozygous for resistance and produced different F1 genotypes, but this seems unlikely considering that the somaclonal lines breed true. Another possibility is that most of the variation was due to environmental effects, and the somaclonal lines can each be considered homozygous and similar to one another. Analysis of more crossing data may help to discern between these possibilities. mammwm. ngggggign of reciprocal diffgrgnggg. The next part of the genetic analysis was to screen the F2, or segregating generation of the recipro- cal crosses. These plants were obtained by self-pollinating several F1 plants from each somaclone family to obtain seeds which were germinated, inoculated with ToMV and screened for virus using ELISA. A general presentation of these results is made in Figure 6A, comparing the percentage of virus-free F1 and F2 progeny from the reciprocal crosses. When the less resistant F1 (26 x So) were self-pollinated, the F2 progeny showed more resistance than the F1. In the reciprocal cross (Sc x 26), the resistance of the F2 progeny was about the same as in the F1. For further analysis, the mean ELISA values were listed (Figure 6B). 64 A. Resistance response of reciprocal crosses. E] 10 DAYS - 20 DAYS 30 mm 100%~ "-1 ‘ 13' 80%. 1 0-1 p k 3 F'— O. O ‘L 6011* r_. ‘H O “E 40% ~ 0) U L. a) ’(""”’ 0‘ 20% L [—‘l ”*1/21...I‘i:1.-:: 0% * I if"! ................ 26 26x8c Scx26 26x80 8ox28 8c Line B. ELISA values (from 10 d pi) for F2 progeny of reciprocal crosses. # ELISA # ELISA F Line plants mean plants mean value So 12 36 0 356 58 0.626 3.95 n.s. So 215 44 0 386 28 0.588 1.82 n.s. So 219 29 0 491 34 0.774 2 80 n.s. Sc 247 107 0 400 74 0 441 0.20 n.s. Sc 322 32 0 623 44 0.658 0.06 n.s. Sc 330 37 0 665 56 0.782 0.52 n.s. Pooled 285 0.461 294 0.628 9.26** **High1y significant difference (p-0.01). Figure 6. Resistance to ToMV in progeny of reciprocal crosses of tomato somaclonal lines with the parent susceptible line, GCRI-26, for both the F and F . A) Bars represent the percentage of virus-free plants at 10, 20 and 3 d. B) Testing for significant reciprocal differences for each somaclonal line. 65 Examination of the mean ELISA values of the F2 progeny resulted in the recognition of a slight maternal bias. Though progeny plants in each somaclone family had less virus accumulation for the Sc x 26 cross than for the the reciprocal cross, none of the differences were significant unless the data from somaclone families were pooled. Evaluation of S:I:R ratings was needed to supplement these observations. The detailed pedigrees of all the F1 and F2 plants which were screened are presented in Appendix C. From these pedigrees, 10 d pi S:I:R ratings of the F2 progeny were compiled (Table 11) and supplemented with the ELISA means for analysis of variance and LSD multiple range testing. After determining whether inadvertent selection for resistance occurred in the choice of F2 progeny, further analysis could be done. mmmmmmmfiwrzm? To ensure that the results obtained were not skewed by possible selection of F2 progeny of plants that were more resistant, a comparison was made between F2 progeny that were taken from fruit of unscreened plants (unselected) and from plants that had been screened for resistance and were rated either susceptible (S) or resistant (R). The ELISA values for each plant were used for analysis of variance between selected and unselected F2 progeny, and demonstrated that there were not significant differences. P values were 3.18 (with 1 and 292 df) for selected versus unselected 26 x Sc F2 and 4.94 (with 1 and 283 df) for selected versus unselected Sc x 26 F2, both were below the value of 5.02 for significance (p-0.05). Because there were not significant differences between selected and unselected F2 plants, all F2 data were used in further analysis. 66 Table 11. Summary of results after screening F1 and F2 tomato plants for resistance to ToMV (at 10 d pi). A) % S:I:R ratings. B) mean ELISA values. Lower case letters in parentheses following each ELISA mean indicate results of LSD multiple range testing among families. C) Result of analysis of variance among somaclone families. D) Comparison of reciprocal F1 and F2 families. A. % S:I:R ratings at 10 d pi. Sc x 26 26 x Sc Line F1 F2 F1 F2 12 (%) 40:46:14 16:24:60 61:31:8 42:24:34 215 (%) 23:45:32 20:18:62 50:50:0 43:21:36 219 (%) 30:45:25 37:19:44 57:29:14 50:13:37 247 (%) 17:40:43 26:9:65 68:29:3 28:15:57 322 (%) 0:60:40 41:18:41 63:33:4 39:27:34 330 (%) 11:61:28 35:23:42 63:25:12 45:17:38 B. Mean ELISA values at 10 d pi (AU at 405 nm). l2 0.868(c) 0.356 0.778 0.626 215 0.362(ab) 0.386 0.761 0.588 219 0.538(b) 0.491 0.866 0.774 247 0.246(ab) 0.400 0.804 0.442 322 0.059(a) 0.623 0.812 0.658 330 0.162(a) 0.665 0.662 0.782 C. Results of analysis of variance among somaclone families. F value 5.57** 1.80 0.32 2.00 (df) (6,156) (5,279) (6,211) (5,288) D. Comparison of reciprocal F1 and F2 families (pooled). Family ELISA mean (LSD) F value (df) F1 SC X 26 0.385 (a) F1 26 x Sc 0.791 (c) 16.92** (3 and 932) F2 Sc x 26 0.461 (a) F2 26 x Sc 0.628 (b) **Highly significant difference (p-0.01). 67 Significance g; reciprocal differences. It was evident after comparison of mean ELISA values for F2 progeny (Figure 6B) that there was a slight maternal bias, which was not significant within crosses from each somaclonal line. However, when the data were bulked, the reciprocal differences were highly significant. Examination of the S:I:R ratings for the same plants (Table 11A) indicated that there were consistently small differences between the F2 progeny of reciprocal crosses. In comparison to the considerable differences between the F1 progeny, it became clear that the maternal bias was greatly reduced. This unusual transmission pattern for resistance will be important in consideration of a model to explain the inheritance of the resistance. A1; thgxg differences among the somaglgnél lines? It was possible to examine the similarity of the somaclonal lines to each other by comparing the results of screening the F1 and F2 generations. The S:I:R ratings and mean ELISA values at 10 d pi (Table 11) were available for direct comparisons. The differences among the ratings were difficult to interpret, so ELISA values were employed to calculate the analysis of variance among somaclonal lines for each of the reciprocal crosses for each generation. Analysis of variance indicated that there were not significant differences among the somaclonal lines in the F2 generation (including Sc 12). The F values were low for F2 progeny of both of the reciprocal crosses, with 1.80 for Sc x 26 F2 and 2.00 for 26 x Sc F2. There was no clear evidence for differences among the somaclonal lines, which again helps to justify the pooling of data from the lines. When testing for differences between F1 and F2 progeny of the reciprocal crosses (Table 11D), the F value was 16.92, indicating that there were very significant differences among the four ELISA means 68 (p-0.01). Results of LSD multiple range testing (p-0.05) indicated that there was not a significant difference between the F1 and F2 of the cross Sc x 26, but there were differences between these and the F1 and F2 of the cross 26 x Sc. Also, there was a considerable difference between the F1 and F2 generations of the cross 26 x Sc. These comparisons substantiate the initial observations seen in Figure 6, which showed differences in F1 progeny of reciprocal crosses biased for maternal inheritance of the resistance, with the F2 generations showing similarly high levels of resistance, with an unusual increase of resistance seen from the F1 to the F2 generation of the 26 x Sc cross. Estimation mi resistange gm; £9 gng maternal effiggg. The distribution of ELISA values for all F2 plants screened was examined in Figure 7. The distributions were similar for the F2 progeny of reciprocal crosses, except that a greater portion of plants fell into the range of 0.00-0.16 AU for F2 progeny of the Sc x 26 cross. There were only 42.5% of plants in this range for the 26 x Sc cross (Figure 7A) compared to 58.2% of the Sc x 26 cross (Figure 7B), a difference of 15.7% between the reciprocal crosses. This was only about half the difference seen between reciprocal crosses in the F1 generation. There were more 26 x Sc F2 plants in the very high ELISA range (between 1.33-2.0 AU) than there were Sc x 26 F2, an increase of only 5.6% of the plants in this range. It was notable that there were not discrete groups seen in the segregating F2 generation, possibly due to environmental effects, thus making it more difficult to make genetic models to explain the inheritance of the resistance. mmmwwmmmfimmm If original somaclones were homozygous for the resistance trait, the R1 plants used in cross-pollination would all be homozygous, and the F1 69 was A 26 x Someciono (F2) : w‘ 6 man - 6.626 2 OJ 3 ems) 5 O. O B: ..., 40%\ O u . . 5 o 20‘ ' 3 h I ~. Q} . N ;-‘ 1y 1 . 4. A. 114 x ..x . . 11.41. ...fl .1 i o .16 .33 .66 .67 .66 1.00 116 1.63 160 161 166 2.0 ELISA value. (AU It 406 nm) 1006 B Somaclone x 26 (F2) 6011 b nun - 0.476 C 0 fl 3 ath F! 3 8 406 ~ “I f O t 2 o 666 1— ‘ In 0 ...... m ' . .‘ .... . ...... .V.'..,.... 6 .16 1.66 .66 61.661166116166166 1.67 166 2.0 ELISA values (AU at 405 nm) Figure 7. The distribution of ELISA values for virus detected in F2 progeny at 10 d pi from crosses between A) GCRI-26 and somaclonal lines and B) the reciprocal cross, of somaclonal lines and GCRI-26. 56 {EC Wide 70 progeny of crosses using different R1 parents from the same somaclonal line would not show segregation. The F2 progeny would segregate into different resistance phenotypes, but there would not be variation in segregation patterns among sets of F2 progeny from different F1 parents. Therefore, there should not be more variation among distantly related replications than among sibling replications. In replications among cousins, significant variation in segregation could occur only if the original somaclone was heterozygous for resistance. To examine the data for evidence of heterozygous parentage, both the 10 d pi S:R ratings and ELISA means were used for each replication of each somaclonal line (listed in Table 12). In this table, siblings screened in different environmental groups are marked with (+) and other data sets are from more distantly related plants within the somaclonal family. The results of LSD multiple range tests for differences in replications of each reciprocal cross for each somaclone are included to the right of each ELISA mean. It was evident that somaclone lines 12, 215 and 219 did not show significant differences among the data sets from environmental groups, so that these lines did not provide evidence to support the theory that the original plants were heterozygous. It must be noted that the theory was not disproven; perhaps more replications of these lines might allow detection of differences. Similarly, in somaclonal lines 322 and 330, more replications would have been useful to clarify the ambiguous results. In each of these lines, there were differences between data sets for the F2 progeny of one cross (Sc x 26), but not in the reciprocal cross (26 x Sc). Enough data sets of somaclone family 247 were screened to provide evidence to support the possibility that the original somaclone was 71 Table 12. Summary of resistance ratings at 10 d pi of F2 progeny of reciprocal crosses between somaclonal lines and the susceptible line, GCRI-26. (+) means groups are siblings. The letters in parentheses denote the environmental group. Plants within each group were siblings. Sc x 26 26 x Sc S:R (10 d) ELISA S:R (10 d) ELISA Plant line rating mean rating mean Sc 12 2:10(T) 0.279 a 5:11(R) 0.368 a 2:12(T) 0.445 a 6:8(2) 0.856 b 2:9(n) 0.318 a 3:4(z) 1.005 ab 6:31 0.356 4:8(W) 0.462 ab (16:84%) 9;:(3) 0,294 ab 26:36 0.626 n.s. (42:58%) n.s Sc 215 2:14(R) 0.143 a 9 7(R) 0.779 a + 5:11(i) 0.685 b 3:9(n) 0,334 a + 2;i1(n) 9,309 ab 12:16 0.588 9:35 0.386 (43:57%) (20:80%) n.s. n.s So 219 8:8(W) 0.151 a + 4:9(W) 0.482 a + 1:5(d1) 0.247 a + 3:2(d1) 0.840 a + 3:7(d2) 0.271 a + 8 4(d2) 1 028 a ° 0 4 8 a a 12 20 0.491 15 15 0.774 (38 62%) n.s. (50 50%) n.s. Sc 247 + 0:16(c) 0.053 a 11:5(R) 0.740 c + 1:15(u) 0.236 ab + 4.12(u) 0.484 bc 0:16(s) 0.032 a + 6:10(x) 0.725 c 0:l4(x) 0.102 a + 0 13(3) 0.091 ab 12:4(i) 0.927 d + 14 99924 a + 11:5(i) 0.888 cd 21:54 0.442 + 4;2(s) 9,550 bc (28:72%) 27:79 0.400 (25:75%) **sig **sig So 322 11:5(c) 0.909 a 10:6(i) 0.945 b 2;14(n) 9,936 b + 4:8(W) 0.448 a 13:19 0.623 + 9;;9Lnl ab (41:59%) 17:27 0.658 **sig (39:61%) n.s. Sc 330 11:5(c) 1.011 b + 5:8(W) 0.472 a 4:12(y) 0.565 ab +10:6(y) 1.186 b 2;;31n) 99999 a 5:11(x) 0.718 ab 17:31 0.665 gillil 99119 ab (45:55%) 26:32 0.782 *sig (46:54%) n.s. 72 heterozygous for resistance. The data from both the F1 and F2 generations were compiled to make the pedigree presented in Figure 8. 0f the 247 x 26 crosses, the first example was consistently highly resistant in the F1 and each of the four F2 families, but not as high as the self-pollinated somaclonal line. In the second example, the F1 plants were very resistant but their F2 progeny (taken from a single fruit) were not, indicating genetic and/or environmental variation between generations. The third and fourth examples each displayed intermediate levels of virus accumulation in the F1 and F2 generations. When the different crosses were compared to each other, it seemed clear that there were consistent differences between them, thus supporting the idea that there were different R1 genotypes used for the crosses. These results are confounded because data sets were from different environmental groups. In the reciprocal (26 x 247) crosses, differences were also observed between crosses. The first example showed an intermediate level of virus accumulation in the F1 progeny and more virus accumu- lation in the F2 generation. The second and third examples were similar in that the F1 allowed somewhat high levels of virus to accumulate. The F2 progeny of the second cross gave strong evidence for genetic differences in the progeny of different F1 plants. Two sets of data (from different environmental groups) of F2 progeny were screened from two different F1 plants, resulting in convincing differences. The two data sets of F2 screened from one F1 plant showed an intermediate level of virus present, but the data sets of F2 progeny from another F1 plant showed a consistent resistance to virus accumulation. These differences were evident in both the S:R rating system and the 73 Somaclone Line 247 x GCRI-26 F1 progeny 247x26 247x26 247x26 247x26 L. 2.. 3... A. S:R mean S:R mean S:R mean S:R mean (0)0:12 0.000 a (T)0:5 0.000 ab (2)4:2 0.749 c (BB)2:10 0.334 b F2 S:R mean S:R mean S:R mean _(c)0:16 0.053a —(1)12:a 0.927d L<1>11z5 0.888 cd (0)1:15 0.236 ab (5) 4:9 0.550 bc r (s)0:l6 0.032 a L-v(x)0:l4 0.102 a GCRI-26 x Somaclone Line 247 F1 progeny 26x247 26x247 26x247 J... 2.. 3... S:R mean S:R mean S:R mean (H)4:5 0.328 a (N)1l:1 0.963 b (BB)10:2 1.204 b (I)3:5 0.500 a F2 S:R mean S:R mean [_ (R)ll:5 0.740 c F_(u)4:l2 0.484 bc (x)6110 0.725 c _-(s)0:13 0.091 ab (u)0:l4 0.024 a Figure 8. Pedigree of reciprocal crosses of somaclonal line 247. Different R1 plants were used for the different numbered crosses shown. For each family of progeny screened, the environmental group letter (in parentheses) is given, followed by the S:R rating at 10 d pi and the mean ELISA value. Results of LSD tests were included for comparison of means of each generation of each particular cross. 74 comparisons of ELISA means, and differences between environmental groups were not large, indicating that there may be genetic differences. Overall, these data from both generations of the reciprocal crosses support the hypothesis that the original somaclone 247 might have been heterozygous for resistance, however, the mode of inheritance for the resistance would need to be complex to explain why the self- pollinated generations did not display segregation into different phenotypes. A genetic model which incorporated heterozygosity was considered and will be discussed in a later section. However, a closer examination of the effect of environment was important. Estimation 9f environmental effec;§. Just as was seen in the screening of the F1 progeny, there was notable variation among environmental groups of the F2 progeny. The results of testing for significant differences among groups are presented in Table 13A. The F values were very significant (p-0.01) for F2 progeny of both reciprocal crosses, and multiple range testing showed where the differences occurred among environmental groups. These differences among groups represent a considerable environmental influence on screening. Again, because experimental design gave an unbalanced case, variance due to interactions among environmental group and genotype could not be calculated. Data were analyzed for variation among somaclone families within each environmental group (Table 13B) and for variation of environmental groups within each somaclone family (Table 130). Variation among progeny of each somaclonal line was not significant, except in the case of Group c (due to progeny of Sc 247 x 26, which is also reflected in 75 Table 13. Results of analysis of variance and LSD multiple range tests (p-0.05) of ELISA means of F2 progeny of pooled somaclone families. A) Among environmental groups. B) Among somaclone F2 families within environmental groups. C) Among somaclone F2 families. A. Variation among F2 environmental groups. x 6 29 n SQ nv ronmen a # nv ronmen a Group plants mean (LSD) Group plants mean (LSD) R 16 0.143 a; s 13 0.091 a . a; 812.2 a. a 23 32.2% . u . a . a . 11 2? a 13 32:: ab 5 . a . c n 25 0.331 ab) x 48 0.657 bc T 25 0.372 abc; 1 29 0.852 cd a 14 0.565 bcd z 18 0.873 cd 16 0.657 cd) d 17 0.962 d; c 48 0.658 d; y 14 1.186 d i 48 0.833 d F value between 4.81** 4.57** snxiranmanfal 622226 B. Variation among pooled somaclone F families within environmental groups (Egggrigps where comparisong can be made 6 __K Sc "# c Group lines plants F value Group lines plants F value R 215 16 ---- R 12 16 215 16 l 83 247 16 (n s ) W 219 16 ---- W 12 12 219 13 0.01 322 12 (n s.) 330 13 i 215 16 1.16 1 322 16 0.72 247 32 (n.s.) 330 13 (n.s ) n 12 11 0 16 n 215 12 3.28 215 13 ( ' 219 4 (n.s.) 219 2 “-5) "g """ i215 """ if, """""""" g """" II """ II """"" III" 322 16 15.97** 339 16 C. Variation among somaclone F2 families among environmental groups. $3126 ___£__A§JL§&____ # # # Line plants groups F value plants groups F value Sc 12 0.29 1.42 Sc 215 3.83 3.47 Sc 219 1.32 1.31 Sc 247 9.74** 5.07** So 322 8.89** 2.72 Sc 330 5.06* 2.09 *‘Significant differences (p-U.U5). **Highly significant differences (p-0.01). 76 Table 13C). Otherwise notable differences were not seen among somaclone families when examined within the environmental groups. Consistent variation among data sets within each somaclonal family was not observed (Table 13C). Only among the progeny families of somaclone 247 was there significant variation, but it was not possible to determine the source. In general, because most testing did not attribute variation to differences among somaclone families or within somaclone families, it is likely that most of the variation is due to environmental factors. Beginxggal g12§§§§ nigh finnny EQSE- To test the level of somaclonal resistance when moved into another susceptible cultivar, several somaclonal lines were outcrossed to Bonny Best tomatoes. This step was taken in order to determine the general level of resistance with another genetic background. It was also important to confirm that there was a difference in progeny phenotypes of reciprocal crosses, thus eliminating the possibility that the line GCRI-26 was the cause of the previously detected differences. Results of screening the F1 and F2 of reciprocal crosses with Bonny Best are presented graphically in Figure 9. In the F1 progeny, there appeared to be some maternal effect in the inheritance of the resistance, and in the F2 generation, there was increased resistance (greater number of resistant plants), and the loss of the maternal effect. Overall, there was less resistance expressed in progeny of these crosses than progeny of crosses with GCRI-26. This type of change in expressivity was expected, since it was reported previously in studies with the Tm-l gene (Fraser, 1985). A more detailed inspection of the resistance ratings and mean ELISA values was required for further analysis. 77 A. [:1 10 DAYS - 20 DAYS 30 DAYS 100%b 8 :3 60% F 2 .... 3 a 60% F 1 $4 0 ....) 40% "' c m o .............. 8 3; °' 20* 0% 1 f_.Ll . ...; , ;:-.;;‘ BB BBxSc 83x80 86x88 86 Line B. ELISA mean 1.730 1.209 0.732 0.825 0.875 0.043 (AU at 10 d pi) S:R rating 24:0 35:1 24:12 15:14 12:12 0:24 (10 d pi) Figure 9. A) graph of resistance to ToMV in reciprocal crosses of tomato somaclones with the susceptible line, Bonny Best for both F and F progeny, scored at 3 stages in their development: 10, 20 and 30 d pi. B mean ELISA values. 78 The pedigrees of the crosses are listed in Table 14, with both S:I:R ratings and mean ELISA values (10 d pi). The crosses for soma- clone lines 215 and 330 were done reciprocally with individual plants, thus eliminating the possibility that there would be complications due to differing nuclear backgrounds of the somaclonal parents when the progeny were examined for any differences between the crosses. The F1 plants were screened in one set and F2 plants were screened in another set, eliminating as much variation due to environment as possible. The results of screening the respective F1 and F2 progeny for each somaclone were similar, using both S:I:R ratings (Table 14A) and mean ELISA values (Table 14B). Analysis of variance confirmed that there were not significant differences among somaclonal families. Pooling of the data then allowed greater precision in comparisons between the F1 and F2 progeny by providing a larger sample number. The differences seen in ELISA means between reciprocal crosses were statistically significant using the pooled data. The results of LCD multiple range testing are presented in Table 14B. These differences were consistent in the F1 progeny for each somaclonal line, though were not always significantly different when an analysis of variance was done, again due to small sample number. Differences between the ELISA means of the F2 progeny were not evident in individual somaclonal lines, confirming the same result as when the somaclonal lines were pooled. A11 F1 and F2 progeny were significantly different from the control Bonny Best plants, showing less virus accumulation when the somaclonal resistance genes were present. Again to estimate the amount of resistance due to cytoplasmic inheritance, it was useful to look at the distribution of ELISA values 79 Table 14. Results of screening of cross-pollinated plants from reciprocal crosses of tomato somaclone R with the susceptible cultivar, Bonny Best. A) S:I:R ratings and B) Mean ELISA values, with LSD multiple range tests (p-0.05) to compare ELISA means between crosses within each somaclonal line. A. S:I:R ratings at 10 d pi. Sc 3 BB BB n 52 12-2 8:1:3 5 1 6 - - 12-9 - - 12:0 0 9 5 2 215-6 7 5 0 7:2:3 12 0 0 6 2 4 330-5 9:2 1 - 11 1 0 - Pooled 24:8:4 12:3:9 35:1:0 15:7:6 Control Bonny Best 24:0:0 B. Mean ELISA values (AU at 405 nm) at 10 d pi. Sc R2 59 3 BB 99 n Sc F Parent F1 F2 F1 F2 value 12-2 0.682(a) 0.637(a) - - 12-9 - - 1.334(b) 0.942(ab) 3.03* 215-6 0.706(a) 1.111(ab) 1.207(b) 0.707(ab) 2 20 330-5 0.803(a) - 1.088(a) - 3.38 Pooled 0.732(a) 0.875(a) 1.209(b) 0.825(a) l4.40** Control Bonnie Best 1.730(c) *Significant differences (p-0.05). **Highly significant differences (p-0.01). 80 for the F1 and F2 plants screened. In Figure 10A, the control parental lines were graphed to show that there was a sharp difference between the two genotypes. The somaclonal lines were always in the range less than 0.50 AU and the susceptible Bonny Best plants were always much above 0.50 AU. It appeared that there was variation in the expressivity of susceptiblity in Bonny Best that almost looked like two discrete groupings. In both of the F1 progenies from the reciprocal crosses, there were intermediate levels of virus accumulation seen (Figure 10B and C), but there were 22% more plants in the resistant range of 0.00-0.50 AU when the somaclonal lines were used as the seed parent in the cross than when used as the pollen parent. As in crosses of somaclones with GCRI- 26, the F1 resistance levels are intermediate relative to the parents, thus looking much like an incompletely dominant trait. When the ELISA distributions of the F2 progeny from the reciprocal crosses were examined (Figure 11), there seemed to be clear segregation into classes having high or low virus titers, but lacking plants with mid-point values. This distribution is similar to that of crosses to GCRI-26 (Figure 7), but because of the elimination of environmental effects, the peaks are more discrete in the crosses with Bonny Best. When the F2 progeny of the reciprocal crosses were compared for differences indicating maternal inheritance of the resistance, there were no differences seen. The F2 progeny of So x Bonny Best had only 3% more plants in the range of 0.0-0.5 AU than the reciprocal Bonny Best x Sc F2. This supported the other evidence for the absence of the maternal inheritance of resistance in the F2 generation. finnhnxnnggg. Completion of backcrosses of the F1 plants to the somaclone R1 or R2 and susceptible GCRI-26 parents was accomplished for 81 3661616111 80 Bonny But noun-0.03 nun-1.78 40§§fi ...... o‘fgfi3, , . 0 .16 .66 .60 .67 .63 10 1.16 1.60 160 1.67 1.66 8.0 Percent of Population % l I; 1006 Bonnyeutxmeolonflfl) g m. moon-1.200 «-1 4.) 1‘. :1 0011* o. o 6. ca 40%* o .1.) 8 ... o P" f... o a. 0 .3 .88 ‘0 .87 .33 1.0 I.” 1.381.” 1.07 1.” 8.0 (: «ms SouunfiuurxlaonnyaoctF1 g m.- nun-0.788 on .6.) 6 —1 s oosr 8' a. GH 1- o 406 +3 : 0 O h o o. 'uflnummm m m m u ELISA values (AU .1 405 M’ Figure 10. The distribution of ELISA values for virus detected in plants from A) resistant somaclonal lines and susceptible Bonny Best, B) the F1 progeny of the cross Bonny Best x Sc and C) the F1 progeny of the reciprocal cross, Sc x Bonny Best. 82 1006 = Bonnyaut x Somaclone F2 .2 ’0‘ .— noun 0 0.025 a.) as III-i 5 o. BO‘F o a. M 0 406+ a.) c: o \. o I- 1. 206 ‘\ 0 ~ v ' 9- . ,. ; 0‘ - A v!- 5 r V l " A”. .A . 1.1.51" 1 . 1.1...A. ' ..l 0 .16 .66 .60 .67 .66 10 116 166 160 167 166 6.0 ELISA values (AU 61 405 am) 1006 a Somaclone x BomyBoet F2 2 sun - 0.076 +3 606 a 1—6 a a. ‘64 ° 406- «6.) = \ w ‘ ‘1 2 20% r- \_ / 1 0- {..,. mga f ___ 5 ‘J.,, 0 166660 .0-7“ .66 1.0 1.10.6133160107103 6.0 ELISA values (AU I! 405 nm) Figure 11. The distribution of ELISA values for virus detected in F progeny plants from the crosses A) Bonny Best x Sc and B Sc x Bonny Best. 83 many of the possible crosses, but only somaclonal line 247 had a complete set of the eight backcrosses. In initial comparisons, ELISA means from the resistance screenings were combined for all somaclonal lines and are presented graphically in Figure 12. The corresponding ELISA means and standard error of the mean are listed in Table 15. Detection of differences among backcrosses. When the susceptible line, GCRI-26, was used twice in the crossing scheme, high levels of virus were allowed to accumulate in the progeny. However, when the somaclone lines contributed twice to the pedigree, less virus accumulated in the progeny plants, to show that a higher level of resistance was expressed, though not as high as in the self-crossed somaclones. This provided evidence for an additive effect that could be due to incomplete dominance of the resistance gene(s). More detailed analysis was necessary to establish the significance of the differences seen. The results of analysis of variance and LSD multiple range tests (p-0.05) in Table 15 confirmed that the differences in ELISA means were very significant among the backcrosses and controls. The very large P value of 657.9 (with 13 and 1970 df) provided strong evidence for differences, which were then identified using the range test. Both resistant and susceptible controls were significantly different from the backcrosses, and among the backcrosses there were differences. Backcrosses A-D were clearly different from backcrosses E-H, which confirmed the gene dosage effect. For direct comparisons of selected backcrosses, orthogonal contrasts were done (using NCSS Software). When backcrosses ABCD were contrasted to EFGH for the combined somaclones (Table 16), the T value was 9.71 (395 df), confirming highly significant differences (p-0.01) due to gene dosage. 84 i A 26xl26xSc) E 126xSclec 1'; ; B 26xlScx26] F [Scx26lec 1. i 3 J; C 126xSclx26 G ScxlzexScl 1: i..' ; 1 3: . DISchGlxze HScxlScx26l f? ‘1 z 3 11 1 <1 :: 33'.1_.I T: W 3 . 1% 3 1 1 . .3 3113 33151115 2 1: 111::111 é ': : 3 1 5 9 1 11 9 :0 :1 1 f 1 1 31 1 i i . . : ; .. z z 3 1 I 3 i . i 2 I 1 3J '1 1 : 1 1 ' 1 1! 1 1 1 i 2 + 14 .gy4: 5 1 g: 1 1 j 1 1 1 1 1 i 1 "'1: 31:11:11 L111! 11,6 111'1111 11:.11_L. :3 131:;111 11,1131: - '11 13s'1111 11 11:11 3 ; {1131:11 12:1113: . «3.: 33313.11: 1f11111371l+rrj -, ~ 5 : C' E C J "I : Figure 12. Results of screening backcrosses for resistance to ToMV. Bars represent the ELISA mean for each backcross, with lines indicating standard error of the mean (using pooled data from all somaclonal lines). Table 15. Mean ELISA values for each backcross shown in Figure 12. The results of analysis of variance and LSD multiple range tests (p-0.05) were included. # ELISA Backcross plants mean (AU) S.E. LSD Control 26 425 1.755 0.0151 a A. 26x[26xSc] 83 1.384 0.0624 b B. 26x[Scx26] 70 1.448 0.0634 b C. [26xSc]x26 28 1.133 0.1387 c D. [Scx26]x26 66 1.104 0.0841 c E. [26xSc]xSc 78 0.413 0 0710 e F. [Scx26]xSc 10 0.758 0.2965 d G. Scx[26xSc] 36 0.636 0.1148 d H. Scx[Scx26] 46 0.332 0 0740 e Control Sc 224 0.0436 0.0108 f F value --- ---- 6S7.9** **Highly significant differences (p-0.01). 85 Table 16. Orthogonal comparisons between backcrosses using ELISA values at 10 d pi from the pooled somaclone families. # ELISA Backcross plants mean ------- ORTHOGONAL COMPARISONS -------- A. 26x[26xSc] 83 1.384 -1 -l B. 26x[Scx26] 70 1.448 -1 -1 C [26xSc]x26 28 l 133 -l l D [Scx26]x26 66 1 104 -1 1 E. [26xSc]xSc 78 0.413 1 -1 F. [Scx26]xSc 10 0.758 1 -1 G. Scx[26xSc] 36 0 636 l 1 H. Scx[Scx26] 46 0.332 1 l T value: 9.71** 2.27* 2.69** 0.56 0.71 (395 df) (n.s.) (n.s.) *Significant difference (p-0.05). **Highly significant difference (p-0.01). When looking for maternally biased reciprocal differences, backcross B was contrasted to D, and backcross E was contrasted to G. In both cases, differences were seen, but only in backcross B versus D was there significantly less virus accumulation in progeny with the somaclonal cytoplasm; in backcross E versus G, the differences in virus accumulation were not significant. Therefore, the maternal bias seen between F1 reciprocal crosses did not consistently appear in the progeny of backcrosses. When the other reciprocal pairs were contrasted, backcrosses A and C showed significant differences even with the same cytoplasmic backgrounds. Backcrosses F and H were not significantly different. Analysis of ELISA data from backcrosses of somaclone line 247 were nearly identical to the pooled results. A larger sample would help in discerning the true differences among the backcrosses. 86 Are EDSEQ differences emegg Ehé somaelonal linee? To look at variation among somaclone families, the S:R ratings (at 10 d pi) were listed for data sets from each environmental group (Table 17). There was consistency among the somaclone families in which similar crosses were duplicated. For a closer look, the results were converted to percentage S:R in Table 18A, with the corresponding ELISA means found in Table 188. When differences among somaclone families were tested for each backcross using analysis of variance of ELISA values, these differences were not significant for each case. Thus, the pooling of the data from different somaclonal lines was valid. Disgzibugion e: gelgé xeleee fie; baekerosges. In Figure 13, it was evident that backcrosses A-D showed fewer plants with low virus titer (upper graph) compared to backcrosses E-H (lower graph). The biggest differences seen in the distribution of ELISA values seemed to be between these two groups, which were divided into those backcrosses which used the somaclonal lines once (A-D) or twice (E-H). Looking at the reciprocal backcrosses B and D, it was noted that backcross D, which had the cytoplasm of the somaclonal line, had more progeny in the low ELISA value range. Of the other reciprocal pair which had differing cytoplasmic backgrounds (E and G), there was not a striking difference between the distributions of ELISA values. As an added note, there were not discrete phenotype groups evident in the distribution of ELISA values, which suggests variation due to environmental effects. mummgmmwmmmwn 1252. The next step in the genetic analysis was to look at all of the data from one somaclonal line. Somaclonal line 247 was chosen for this analysis for two reasons: it displayed both a high level of resistance 87 Table 17. S:R ratings (10 d pi) of backcrosses which were screened with ToMV. Results of each environmental group were shown for each somaclone family. Backcross 12 215 219 247 322 330 SUM A. 26x[26xSc] 10:0 11:1 - 12:0 - - 12:0 leg 10:0 81 2 10;1 18:1 22 0 ------ 32:1 (%) 98 2 B 26x[Scx26] 10:0 12 0 - 7:9 - 12 0 19.2 9.31 60 10 22 0 16:10 ------ (%) 86 14 C [26xSc]x26 - - - 9:3 - 4:2 L2 21 7 17:5 ------ (%) 75 25 D [Scx26]x26 - - 8 2 9:1 9:3 10 2 7:5 51:15 m ------ 24:8 (%) 77 23 E [26xSc]xSc 5:7 0:5 - 0 6 - 4 5 2:8 51; 5:7 4 9 20 58 03.11 ...... 12:39 (%) 26 24 F [Scx26]xSc - - - 4 6 - - 4 6 (%) 40 60 G Scx[26xSc] 6:6 - - 3:7 - - 4:4 15 21 LA ------ 12:14 (%) 42 58 H. Scx[Scx26] - 3:9 3:9 0:10 - 3:9 9:37 88 Table 18. Results of screening of backcrosses with ToMV. A) Expressed as percentages (S:R rating at 10 d pi). B) Mean ELISA values at 10 d pi. A. Percent S:R rating for each somaclonal line. Beckcroee 12 215 219 247 322 330 SUM A. 26xl26x8c] 97:3 95:5 -- 100:0 -- -- 98:2 3. 26xlScx26] 100:0 100:0 -- 62:38 -- 100:0 86:14 C. [2618c1126 -- -- -- 77:23 -- 67:33 75:25 D. [Scx261126 -- -- 80:20 75:25 75:25 84:16 77:23 "QT'IQ;L;LIL;"QQI;7"33;; """ I """ Q12;;""'ZZ """ 22;; """ 5.32"" P. [SchBJXSc -- -- -- 40:60 -- -- 40:60 6. Scx[26xSc] 46:54 -- -- 30:70 -- -- 42:58 E. ScxlSchS] -- 25:75 25:75 0:100 -- 25:75 20:80 B. ELISA means for eech eomeclonel line. Beckcroee 12 215 218 247 322 330 SUM A. 258(258801 1.427 0.871 -- 1.673 -- -- 1.384 B. 263(8c126] 1.568 1.354 -- 1.447 -- 1.520 1.448 C. [26xSc1326 -- -- -- 1.180 -- 0.963 1.133 D. [Sc2261126 -- -- 1.304 1.018 1.136 1.138 1.105 ”QT'IQQLQLJL;"IQLWLTQQI'": """ ELI; """ 33"" ...... ...... F. [SchBIXSc -- -- -- 0.758 -- -- 0.758 G. Scx[26xSc] 0.682 -- -- 0.516 -- -- 0.636 B. SCXISCXZBJ " 0.401 0.488 0.008 -- 0.376 0.332 89 / // I: A) 23 x 128x801 a) 20 x [83x28] 1 / a) [28x83] x 20 - 0) [some] 1: 23 g 1 00% '1 E 80% . 3 8 607.3 “a 40% 4 :7 1 1'" yz/b ’.1.f.§.§.;.§-§E.f 3:¢*:?:3:3:3:3:?: g 20* '1 . ........ .. Q/ :i-:-:-:-;-:-;-; :1'1L3‘15'323 :-:-:~:-:-:~' g 0* j I l l T V I T T l r I T .18 .83 .60 .67 .83 1.0 1.2 1.3 1.5 1.7 1.8 2.0 ELISA values (AU at 405 nm) / / [:1 E) [28x80] 1: 83 F) [80x26] 1: 80 // a) 83 3: [28x80] - H) 80 x [80x26] E 100$1K +3 2. 30951 5 Q a 80% 1 “5 40151 a =7 2 a: a 8 20% "1 iij- C: E E 5/5 E E a ;i:l'-:-.-:f:1'fj 55:33:53}; g 0* 1 2:: a a a .18 .33 .50 .87 .33 1.0 12 1.3 1.5 1.7 1.8 2.0 ELISA values (AU at 405 nm) Figure 13. The distribution of ELISA values for virus detected in plants from the backcrosses made between susceptible GCRI-26, resistant somaclonal lines and the F progeny of these. Upper graph) Back- crosses A-D used the somaclonal lines once in the pedigree and Lower graph) backcrosses E-H used the somaclonal lines twice in the pedigree. 90 and a clear maternal effect. Also, extra plants of line 247 were screened at each level to provide larger numbers for the statistical work. A summary of results from screening somaclone line 247 (using three rating systems) is listed in Table 19. Somaclone line 247 was found to breed true for resistance after self-pollination. There were clear differences in reciprocal crosses in the F1 favoring a maternal effect in the inheritance of resistance, but not in the F2. As noted in Figure 8, there was weak evidence for heterozygosity in the original somaclone, which must be considered when genetic models for the inheritance of resistance are conceived. For the backcrosses, more resistance was seen in the crosses which used somaclone line 247 twice compared to the crosses in which GCRI-26 was used twice, indicating the additive resistance trait is dependent on gene dosage. mumgmwmummm 19mg. Somaclonal line 12 was examined separately since a maternal bias was not detected between F1 progeny of reciprocal crosses to GCRI-26. Data from this line also incorporates results of crosses with Bonny Best. A summary of results for this somaclonal line is presented in Table 20, using three rating systems. In the R1 generation only 14 plants were assayed. After self-pollination, resistance ratings improved in each generation, and selection for resistance may have been an important factor in this notable improvement. Because so few R1 plants were assayed, the possibility that the original somaclone was heterozygous for resistance is impossible to explore and cannot be eliminated. However, there was no evidence of segregation of self- pollinated progeny into different phenotypes, so that it was likely that the resistance was true breeding. 91 Table 19. Summary of results from screening somaclonal line 247, using the percent S:R at 10 d pi, the ELISA mean and the H:M:L ratings (high - 1.33-2.0 AU:medium - 0.51-1.32 AUzlow - 0.00-0.50 AU) for virus titer. Sample % ELISA % Cross size S:R mean S.E. H:M:L 247 R1 50 1:99 0.04 0.023 0 4:96 247 R2 122 1:99 0.03 0.006 0 1:99 247 R3 132 1'99 0.04 0.006 0 0:100 247 R4 71 0:100 0.02 0.004 0 0:100 247 R5 40 0:100 0.04 0.009 0 0:100 RI'RS 415 1 99 0.03 0 004 0 1 99 26 x 247 F1 41 68:32 0.80 0.092 19:44:37 247 x 26 F1 35 17:83 0.25 0.077 9:9:82 (26 x 247) F2 74 28:72 0.44 0.069 13:13:74 (247 x 26) F2 107 26:74 0.40 0.057 15:11:74 Backcrosses A) 26x[26x247] 22 100:0 1.67 0 074 82 18 0 B) 26x[247x26] 21 62:38 1.45 0.152 71:10.19 C) [26x247]x26 22 77:23 1.18 0.162 54:23:23 D) [247x26]x26 32 75:25 1.02 0 120 31:44 25 E) [26x247]x247 13 31:69 0.55 0.232 31:0:69 F) [247x26]x247 10 40:60 0.76 0.286 40:0:60 G) 247x[26x247] 10 30:70 0.52 0.262 20:10:70 H) 247x[247x26] 10 0:100 0.01 0.003 0:0:100 92 Table 20. Summary of results from screening somaclonal line 12, using the percent S:R at 10 d pi, the ELISA mean and the H:M:L ratings (high - l.33-2.0 AU: medium - O.51-l.32 AU: low - 0.00-0.50 AU) for virus titer. Sample % ELISA % Cross size S:R mean S.E. H:M:L 12 R1 14 4:96 0.134 0 090 0 14:86 12 R2 83 4 96 0.104 0 054 5 0:95 12 R3 76 0 100 0.044 0 010 0 0:100 Rl-R3 156 3 97 0.094 0 022 2 1 97 26 x 12 F1 80 61:39 0.778 0.064 18:46:36 12 x 26 F1 32 53:47 0.868 0.133 37:19:44 (26 x 12) F2 63 42:58 0.626 0.090 21:19:60 (12 x 26) F2 36 16:84 0.356 0.093 14:11:75 BonBest x 12 F1 12 100:0 1.334 0.087 50:50:0 12 x BonBest F1 12 75:25 0.682 0.126 8:59:33 (BonBest x 12) F2 16 56:44 0.942 0.213 44:12:44 (12 x BonBest) F2 12 42:58 0.637 0.203 25:17:58 Backcrosses (incomplete set) A) 26x[26x12] 34 97:3 1.427 0.096 62:32:6 B) 26x[12x26] 10 100:0 1.568 0.111 60:40:0 C) [26x12]x26 0 - - - - D) [12x26]x26 0 - - - - E) [26x12]x12 50 23:77 0.377 0.078 10:20:70 F) [12x26]x12 0 - - - - G) 12x[26x12] 26 46:54 0.682 0.126 23:27:50 H) 12x[12x26] 0 - - - - 93 The F1 progeny of reciprocal crosses made with somaclonal line 12 and GCRI-26 did not differ significantly from each other. Though the ELISA means of the F2 generation did not differ significantly when tested statistically, the S:R and H:M:L ratings were not similar between the reciprocal crosses, indicating that the reciprocal differences might be real. The results of screening outcrosses made to Bonny Best were not consistent with the reciprocal crosses with GCRI-26, because reciprocal differences were seen in the inheritance of resistance between the crosses with Bonny Best. The differences between F2 progeny were not as strong. The ratings of backcrosses for somaclonal line 12 were consistent with the pattern of gene dosage dependence seen with the other somaclone families. When the somaclonal line was used only once in the pedigree, there was less resistance than when the somaclonal line was used twice. When reciprocal backcrosses were compared (E versus G), the progeny with the somaclonal cytoplasm allowed more virus accumulation, so that again there was no maternally biased transmission of the resistance. Seed grenemieeien 9f may. Some plant-virus combinations allow seed transmission of the virus (Gibbs and Harrison, 1976), which can resemble cytoplasmic inheritance of infection. To determine the role of endo- genous ToMV in the differences in virus titer seen between reciprocal crosses, the seed-transmissibility of ToMV was assessed. The results provided evidence to confirm that ToMV was not seed transmitted. Of 60 GCRI-26 seeds which were soaked in virus inoculum and then allowed to germinate undisturbed, no virus was detected by ELISA at 18 d pi (Murakishi, personal communication). In the same experiment, of the 30 GCRI-26 seeds which were not treated with the virus before 94 germination, none were found to be infected by virus at 18 d. This evidence indirectly supported that of the previous reports that ToMV was not transmitted through the seed (Taylor et a1., 1961). A stronger line of evidence was provided with the results of screening uninoculated seedlings of GCRI-26 which were produced on infected parent plants. The absence of ToMV after 25-30 d, in these susceptible GCRI-26 plants (which were used in studies of the movement of virus) was consistent for all plants screened. This was an important result in eliminating seed transmission of virus as part of the ‘resistance' mechanism.. In another experiment, ELISA values were determined for virus- infected fruit-bearing plants, fruit juice and seed from the fruit after acid washing (Table 21). From these data, it was clear that there was virus present at high levels in the fruit of infected plants, but that the acid washing treatment commonly used in this work effectively destroyed the virus (coat protein) on the seed coat. Table 21. Results of testing for the presence of ToMV using ELISA in different parts of infected or healthy plants. Plant Fruit Seed ELISA ELISA ELISA ean mean emeen Infected plants 1.48 AU 1.61 AU 0.00 AU Healthy plants 0.03 AU n.d. 0 00 AU Eeeelee 2f exeeeee with knogg :esistagce genee. To test for allelism between the somaclones and the other characterized resistance genes, fresh reciprocal crosses were completed and the F1 progeny were 95 inoculated with ToMV and screened using ELISA. Somaclone lines 12 and 247 were used in crosses with the isogenic lines GCRI-237 (Tm-1 gene), GCRI-236 (Tm-2 gene) and GCRI-267 (Tm-22 gene). The Tm-l gene is incompletely dominant for limitation of virus multiplication (Fraser and Loughlin, 1980), and the genes of the Tm-Z locus are incompletely dominant for the hypersensitive response to ToMV (Hall, 1980). When the Tm-Z genes are heterozygous, temperature dependent systemic necrosis occurs, which is considered to be incomplete dominance of the hypersensitive response. The Tm-Z gene is less effective than the Tm-Z2 gene against most strains of ToMV. For incompletely dominant genes, if the F1 progeny were to show high levels of resistance, then the genes being tested could be considered to be additive in nature, and might be allelic to each other. The same results might also indicate that there are genes at two different loci which have a incompletely dominant effect. Less resistance in the F1 progeny would serve as evidence that the genes tested were not additive. Resistance screening of the segregating F2 generation of the crosses would be necessary to establish allelism of genes, but this screening was not completed. Table 22 gives the results of screening the F1 progeny of the reciprocal crosses at 10 d pi. Resistance was intermediate when somaclonal lines or GCRI-237 were crossed with susceptible GCRI-26, indicating incomplete dominance of resistance for these genes. It is notable that by classification in S:R or H:M:L ratings there appears to be segregation in the F1 progeny. This is likely to be due to variable expressivity of these genes, as was also noted in the distribution of ELISA values in Figure 5. It is also possible that at the the time of 96 Table 22. Results of screening reciprocal crosses made between somaclonal line 12 or 247 and tomatoes bearing the resisgance genes Tm-l/Tm-l (GCRI-237), Tm-Z/Tm-Z (GCRI-236) or Tm-ZZ/Tm-Z (GCRI-267). 10 d ELISA value S:R H:M:L Genotype Cross (mean) (S.E.) rating rating Sc/+ 26 x 247 0.946 0.408 10:2 7:4:1 247 x 26 0.246 0.306 2:10 1:4:7 26 x 12 0.414 0.479 5:7 2:4:6 12 x 26 0 764 0.484 8 4 6 4 2 Tm-l/+ 26 x 237 0.671 O 345 9 3 3 6 3 237 x 26 0 659 0 336 9 3 3 6 3 Sc/Tm-l 247 x 237 0.001 0.004 0:12 0:0:12 237 x 247 0.003 0.005 0:6 0:0:6 12 x 237 0.062 0.042 0:12 0:0:12 237 x 12 0.027 0.010 0:12 0:0:12 Sc/Tm-Z 247 x 236 0 603 0 348 8 4 3 6 3 236 x 247 0 760 O 452 9 3 4 5 3 Sc/Tm-Zz 247 x 267 0.012 0 006 o 12 o o 12 267 x 247 0 001 0 003 0 12 0 O 12 Controls Sc/Sc 247 x 247 0.025 0 044 0 12 0 0 12 12 x 12 O 003 0.006 0 5 O 0 5 Tm-l/Tm-l 237 x 237 0.149 0.256 1:11 1:0:11 +/+ 26 x 26 1.513 0.051 12:0 12:0:0 97 assay, partial resistance had already been overcome. The results listed here support a genetic model for incompletely dominant resistance in the somaclonal lines, since the Tm-l gene has a similar intermediate level of resistance to virus multiplication seen in the somaclone lines. Significantly, crosses made to GCRI-237 using either somaclonal line produced progeny which were fully resistant to ToMV, indicating additivity of the somaclonal resistance genes with Tm-l. Screening of the F2 plants would allow differentiation between allelism and additivity of resistance genes. Because the crosses with GCRI-237 using somaclone lines 12 or 247 each resulted in complete resistance, this supported earlier results indicating that there were not differences between the lines of somaclones. Screening of crosses between the somaclonal lines themselves would provide further evidence of allelism of the resistance genes recovered from the somaclones. When somaclonal line 247 was crossed with GCRI-236 (Tm-2 gene), the progeny had intermediate levels of virus after 10 d pi. This evidence indicated that somaclonal resistance and the Tm-2 gene were not additive. In fact, because the Tm-2 gene did not display dominance in the F1 progeny, there may have been interference in the resistance normally seen in plants with the genotype Tm-2/+ by the presence of the somaclonal resistance gene. Control plants of the Tm-2/+ genotype would be important for comparison, but none were assayed. When the somaclonal lines were crossed with plants having Tm-Zz, the progeny plants did not allow virus to replicate, thus showing possible additivity of resistance. Alternatively, the Tm-22 gene may act almost completely dominant or epistatic for resistance and may 98 not actually be additive with the somaclonal resistance. Further screening of the F2 progeny would be necessary to discern between these possibilities. Again, controls with the Tm-22/+ genotype would have been useful for comparison, but were not screened. MOLECULAR GENETICS Because there appeared to be a cytoplasmic factor in the inheritance of TMV resistance in the tomato somaclones, the possibility that there were detectable changes in cytoplasmic nucleic acids was explored. Somaclonal lines were examined for changes in the chloroplast genome and for the presence of unusual nucleic acids (dsRNA) in the cytoplasm which might be associated with the resistance trait. One can imagine that the dsRNA might act as a protective element against viral replication, and could have been activated by the "genome shock“ of the tissue culture conditions. Gene activation resulting in the expression of a repressor of TMV replication or excision out of the genome within which it has integrated could confer the resistance seen in the somaclonal lines. Changes have been detected in mitochondrial DNA of potato (Kemble and Shepard, 1984) and in maize (Gengenbach et a1., 1977) after tissue culture operations. Because chloroplasts or mitochondria might be involved in the life cycle of ToMV, changes in the chloroplast genome were studied. Similar attempts to purify and analyze mitochondrial DNA were not successful. thezeeleeg DEA. Restriction fragments of chNA from somaclones 12 and 247 were compared to chNA of GCRI-26, the parental line. Using four restriction enzymes, no RFLPs could be seen on the gel (Figure 14). This does not mean that there are no changes in either somaclonal 99 plastid genome, only that if changes occurred, they were not detected by these enzymes. Banding patterns were similar to those published for tomato (Piechulla, et.al., 1985; Palmer and Zamir, 1982). Therefore, if there was a change in the chNA which altered virus replication, the change would not be a major one. on e- t a ded REA. When procedures were used to extract dsRNA from plant material, no dsRNA was detected in GCRI-26 or the somaclones (Figure 15). A band of dsRNA corresponding to the replicative form of ToMV (6.15 kb) was faintly visible after extraction from a mildly infected GCRI-26 plant. As a positive control, dsRNA was extracted from pepper (14 kd). Because no dsRNA was found in the uninfected tomato lines, this provided evidence that dsRNA (expressed constitutively) does not play a role in the resistance seen in the tomato somaclones. 100 fl Pvu II 0 Hind III p Pst I EcoR I 26 26ABAA26BAB B26AA26B Figure 14. Restriction patterns of chNA from parental tomato line GCRI-26 compared to somaclone line 247 (A) and 12 (B). Lanes marked 6 are undigested chNA. Lambda DNA fragments (restricted by Hind III) were used as markers. 101 A A h3 247 26 12 26 TMV pepper h3 Figure 15. Extraction of dsRNA species from tomato somaclones 12 and 247 and from parental line 26. Controls show bands after extraction from ToMV-infected tomato (faint) at 6.15 kb and from pepper at 14 kb. 102 GENETIC ANALYSIS genetic models. With the completion of resistance screening for the somaclonal lines and crosses, it was possible to test models of inheritance of the resistance. It became clear that there were certain data that eliminated very simple genetic models from consideration. There were three lines of evidence which need to be considered in the proposals for a model of inheritance of the somaclonal resistance. First, nearly all R1 plants screened were resistant to ToMV accumulation. Because there was no segregation of self-pollinated progeny into different phenotypes, a single gene for resistance would have had to be homozygous in the original somaclone. If the original somaclone were not homozygous, then multiple genes would be needed in the model. The second line of evidence involved the maternally biased inheritance. In tomatoes, cytoplasmic inheritance has been reported to be uniparental maternal (Kirk and Tilney-Bassett, 1978), so that a trait inherited from the seed-bearing parent would be passed to all of the progeny, and the trait would not be seen in any progeny if available only from the pollen parent. Yet in the somaclonal resistance, the results of screening progeny of reciprocal crosses indicated intermediate levels of resistance. The maternal bias in the inheritance of resistance nearly disappeared in the F2 generation and in the backcross progeny. There was no model involving a cytoplasmic gene that would provide an explanation for these results was impossible, so other explanations for the maternal bias in transmission of inheritance were needed. Lastly, the intermediate levels of virus accumulation seen in F1 progeny from crosses and backcrosses suggest the possibility that the 103 somaclonal resistance was due to completely dominant nuclear gene(s) was unlikely. There was more evidence to support models that were based on incompletely dominant or additive genes. Another more general complication was the high level of variation due to environmental effects that was seen among groups of plants screened at different times. This reduced the ability to resolve differences in resistance among the lines, and in retrospect, more attention was needed in the experimental design of the progeny evaluation. The assay for evaluation of resistance may not have been optimal either, as demonstrated by the large variation seen among plants with the genotype, Tm-1/+ (Figure 5 and Table 22). To make up for this, care was taken to keep all possible models in mind for evaluation. Two models are presented and evaluated. M2Q21.11 122 EQQIEIEQ sense Elflfi fl extsnlaamis Efffisfi. In this model, a scheme was proposed for resistance based on two gene pairs plus a cytoplasmic factor. The two genes are additive and unlinked, with alleles for susceptibility (Ts) and alleles for resistance (Tr) plus the cytoplasmic resistance factor A or the susceptibility cytOplasm a. For this model, the assumption was made that the original somaclone had a heterozygous genotype, but that this heterozygosity was masked in the self-crossed progeny because of the additive effect of two nuclear genes plus a cytoplasmically encoded gene which gave similar phenotypes. The description of this model can be found in Table 23, with the list of assumptions which were made. The descriptions of phenotypes were dependent on gene dosage, and are listed as susceptible (S) versus resistant (R) or rated on a scale of zero to two (roughly corresponding to the scale of ELISA values). Also listed in Table 23 are the phenotype descriptions of the selfed 104 Table 23. Model 1: Description of proposed genetic model of inheritance of somaclonally derived resistance to ToMV. Assumptions: 1. There was a cytoplasmic mutational event that caused a change from a to A cytoplasm which enhances the effect of nuclear resistance genes. 2. There was a masked heterozygous gene in the susceptible parent line (genotype TsTsTsTr). 3. Mutation or deletion occurred at two unlinked nuclear genes which are duplicate and additive. QCRI-Zé Somaclone Event: a TrTsTsTs --------- > A TrTrTsTr (mutation or deletion) Definitions A is a cytoplasmically encoded resistance gene a is the lack of this gene in the cytoplasm Tr is the additive resistant allele Ts is the susceptible allele Phenotype Ranking We a A a A TsTsTsTs S S 2.0 1.5 TsTsTsTr S S 1.5 1.0 TsTsTrTr S R 1.0 0.5 TsTrTrTr R R 0.5 0.0 TrTrTrTr R R 0.0 0.0 Selfed GCRI-26 Rank TrTsTsTs ----------- 9 1 a TsTsTsTs S 2.0 2 a TrTsTsTs S 1.5 l a TrTrTsTs § leg all S l 5 mean Selfed somaclone gagk TrTrTsTr ------------ 9.1 A TrTrTsTs R 0.5 2 A TrTrTrTs R 0.0 l A TrTrTrTr R 9‘9 all R O 17 mean 105 GCRI-26 and the selfed somaclonal lines. It was noted that the predicted phenotypes were similar to those observed, and when looking at the distributions of ELISA values for these plant lines (Figure 2) there was close agreement to the predicted distributions. Table 24 summarizes the predicted results for F1 progeny of crosses between the somaclonal lines and GCRI-26. The complexity of these predictions increases when all nine genotype combinations are included, but results in only four genotype classes, which are listed. The predicted results are consistent with the observed results for the reciprocal F1 families, and perhaps might explain some of the variation that was observed between F1 progeny from different somaclonal R1 parents. The pooled predicted and observed S:R ratings were the same for the F1, with 3:1 for the a cytoplasm and 1:3 for the A cytoplasm. The numerical rankings for predicted and observed crosses were also similar, in that the mean value for the A cytoplasm was approximately half of the value for the a cytoplasm in both cases. The next step is to examine the results of self-pollination to produce the F2 progeny, presented in Table 25. When each of the four F1 genotypes were selfed to produce F2 families, different ratios resulted, as shown by both S:R rating and the numerical ranking (with the same scale as ELISA values). In each case, the resistance rating was higher in the presence of A, the resistant somaclonal cytoplasm. When all of the resulting F2 phenotypes were pooled for each of the different cytoplasmic backgrounds, the predicted ratings could be compared to the observed ratings for the pooled somaclonal lines. The predicted S:R ratio of 3:1 for the F2 progeny with the susceptible cytoplasm (a) was the inverse of the observed S:R ratio, which was 1:3. For the resistant cytoplasm (A), the 106 Table 24. Model 1: List of predicted genotypes resulting from all possible crosses between the hypothesized genotypes of somaclonal lines and GCRI-26 of Table 23. -Cross- Resulting e A Somaclone GCRI-26 Genotypes Class S:R Rank S:R Rank 1)TrTrTrTr x TsTsTsTs TrTsTrTs 2 S 1.0 R 0.5 2)TrTrTrTr x TsTrTsTs TrTrTrTs 4 R 0.5 R 0.0 TrTsTrTs 2 S 1,0 R 915 1.25 0.25 3)TrTrTrTr x TrTrTsTs TrTrTrTs 4 R 0.5 R 0.0 4)TrTrTsTr x TsTsTsTs TrTsTrts 2 S 1.0 R 0.5 TrTsTsTs 1 S 1‘; S 1,0 1 75 1.25 5)TrTrTsTr x TsTrTsTs TrTsTsTs 1 S 1.5 S 1.0 TrTsTrTs 2 S 1.0 R 0.5 TrTrTsTs 3 S 1.0 R 0.5 TrTrTsTr 4 R 0,5 R 9‘9 1.0 0.5 6)TrTrTsTr x TrTrTsTs TrTrTrTs 4 R 0.5 R 0.0 TrTrTsTs 3 S .119 R 9‘5 1 25 0.25 7)TrTrTsTs x TsTsTsTs TrTsTsTs 1 S 1.5 S 1.0 8)TrTrTsTs x TsTrTsTs TrTsTsTs l S 1.5 S 1.0 TrTrTsTs 3 S leg R 9‘5 1 25 0.75 9)TrTrTsTs x TrTrTsTs TrTrTsTs 3 S 1.0 R 0.5 107 Table 25. Model 1: Summary of genotypes and phenotypes for resistance to ToMV in tomatoes with an additive resistance plus a cytoplasmic resistance factor. Cross Genotypes Cytoplasm a Cytoplasm A F 26 x Sc Sc x 26 2% x Sc Rank Bank Sc x 26 l. TrTsTsTs S 1.5 S 1.0 2. TrTsTrTs S 1.0 R 0.5 3. TrTrTsTs S 1.0 R 0.5 4. TrTrTsTr g 9‘; R ,QeQ 3S:lR 1.0 18:3R 0.5 Expected F2: (four types of self-crosses) 1. TrTsTsTs ------ e l TrTrTsTs S 1.0 R 0.5 2 TrTsTsTs S 1.5 S 1.0 1 TsTsTsTs S 2‘9 5 1‘5 all 8 1 5 38:1R 1.0 2. TrTsTrTs ------ 9 1 TrTrTrTr R 0.0 R 0.0 2 TrTrTsTr R 0.5 R 0.0 l TrTrTsTs S 1.0 R 0.5 2 TrTsTrTr R 0.5 R 0.0 4 TrTsTrTs S 1.0 R 0.5 2 TrTsTsTs S 1.5 S 1.0 1 TsTsTrTr S 1.0 R 0.5 2 TsTsTrTs S 1.5 S 1.0 1 TsTsTsTs 5 2‘9 5 1‘5 115 SR 1.0 SS:1lR 0.53 3. TrTrTsTs ----—-> all TrTrTsTs all S 1 0 all R 0 5 4. TrTrTsTr ------ e 1 TrTrTrTr R 0.0 R 0.0 2 TrTrTsTr R 0.5 R 0.0 1 TrTrTsTs S 1‘9 3 9*; 1S BR 0.5 all R 0.13 Pooled total for F2 3S:1R 1.0 18:3R 0.54 108 predicted and observed resistance ratios were very close to the same. The predicted numerical rankings between these reciprocal F2 progeny were similar to those of the F1 progeny, with the resistant cytoplasm (A) reducing the value by about half. However, the observed ELISA means of the reciprocal F2 progeny were not significantly different from each other, so that there was a clear discrepancy between the predicted and the observed values. Though this weakened the support for this model, it did not eliminate the model, since it was possible that the observed resistance rating were skewed by inadvertant selection of more resistant F1 plants for self-pollination to get the next generation. To examine the data for this possibility, it was necessary to predict the possible results from each crossing combination. From the previous two tables, the predicted values were listed to show the ranking of F2 progeny from each of the nine F1 crosses. These are presented in Table 26. From this chart, it was then possible to look at the crossing pedigree of individual somaclones to see if the model would fit. The comparison of predicted and observed ratings was made using dat from somaclonal line 247, which had the most plants sampled (Table 27). In Table 27, the observed results of resistance screening of line 247 are listed on the left side of the table, and the best fit for the predicted genotypes is listed on the right of side of the table. It seemed that some of the data did fit the predicted model, but with notable exceptions, especially when both S:R ratings and ELISA means were compared. For instance, only the progeny of the first cross of 247 x 26 gave a close fit of observed and predicted values. In the other crosses, the ELISA values were fairly similar between observed values and predicted rankings, but the S:R ratios were very different. 109 Table 26. Model 1: List of combinations of F1 and F2 predicted results from crosses of all possible genotypes of somaclonal lines and GCRI-26. A) 26 x Sc (cytoplasm a) F S“ ' k Cr0ss ‘———FIE* Iégln§§-— FI“‘LQfl'—_F_" - ................................................. Z---- 1) all S 11 5 0 1 0 2) 1:1 11:5 0.75 1.0 1:3 0.5 3) all R 1:3 0 5 0 5 4) all S all S 1 25 1.5 11:5 1.0 5) 3 1 all 8 1 0 1.5 11:5 1.0 all S 1.0 1:3 0.5 6) 1 1 all 8 0 75 1.0 1: 0.5 7) all S all S 1 5 1 5 8) all S all S 1 25 1.5 all S 1.0 9) all S all 8 1 0 1 0 1) all R 5:11 o 5 o 5 2) all R 5 11 o 25 0.5 a 1 R 0.13 3) all R all R o o 13 4) 1:1 1.3 0.25 1.0 5-11 0.53 5) 1:3 1:3 0.5 1.0 5:11 0.53 all R 0.5 all R 0.13 ...... 65"' ' all R all R 0.25 0.5 all R 0.13 ' 7) 611 s 1 3 o 1 o ' " é)' 1 1 1:3 0 75 1.0 all R 0.5 \O V m H H 56 n1 ...: ...: W O U1 C U'| 110 Table 27. Model 1: Comparison of predicted and observed results of an additive duplicate gene model using data from somaclonal line 247. £1, £2 predicted predicted ELISA ELISA Cross E} F; Croae S R mean 8 R mean type* S-R rank 8 R rank 26 x 247 1. 4:5 (0.33) 3:5 (0.50) ----11:5 (0.74) 3 all R (0.5) -- 1:3 (0.5) 2. 11:1 (0.96) ----4:12 (0.48) 5 1:1 (0.75)-- 1:3 (0.5) 6:10 (0.73) or 11:5 (1.0) or all S (1.0) or all S (1.5) ----0 13 (0 09) 0 14 (0 02) 3. 10:2 (1.20) ----none 4,8 all 5 (1.25) ' ' 247 x 26 1. 0:12 (0.00) ----0:16 (0.05) 3 all R (0.0) --e11 R (0.13) 1:15 (0.24) "°'0:16 (0.03) 2. 0:5 (0.00) ----12:4 (0.93) 4 1:1 (0.25) -- 1:3 (1.0) or 5:11 (0.5) 3. 4:2 (0.80) ----11:5 (0.89) 8 1:1 (0.75) -' 1:3 (1.0) 4:9 (0.55) or all R (0.5) 4. 2:10 (0.33) ----none 2.6 all R (0.25) ' ’ *In reference to croaa types listed in Table 26. 111 Thus, this proposed model was not supported by the data from resistance screening of either pooled somaclonal lines nor by comparisons of individual plant lines. The complexity of the analysis of this model was also a serious drawback, especially with the high level of environmental error. Further analysis, including the backcross results, failed to provide support for the model. Therefore, due to its great complexity, the evaluation of backcross results is not presented. Hegel 2; Ineempletely dominant gene(s) EiLh en egdigive eyeeeleemie ineezeegien. One major problem in formulating models for testing was that the maternal bias in the inheritance of the resistance was not seen consistently through all generations of crosses. It seemed that there was not a feasible model which incorporated an organellar gene for resistance. The maternally biased inheritance that was so strong in the F1 progeny of the reciprocal crosses was not evident in the F2 progeny nor the backcrosses. The possiblity that seed transmission of ToMV was the reason for the greater virus titers in susceptible plants was eliminated in the experiments on seed transmission. If this maternal bias was not due to a resistance gene encoded in an organelle, nor due to transmission of virus through the maternal cytoplasm, nor due to dsRNA presence, few choices remain to explain this aspect of inheritance of the resistance. Perhaps a maternal effect, also called predetermination, or nuclear- cytoplasmic interaction may be involved, though a mechanistic model or explanation for these would be difficult to formulate. Maternal predetermination describes a transitional phenotypic effect that normally wears off in one or more generations (Levine, 1973). It usually involves the imposition of characteristics on the cytoplasm of the ova by an outside force. Strictly speaking, this is a 112 developmental process rather than a mode of inheritance, but it can mimic extrachromosomal inheritance. Maternal effect, essentially the same as predetermination, is the determination of the phenotype of the offspring by cytoplasmic elements coded for by the nuclear genotype of the mother (Elseth and Baumgardner, 1984). Maternal effects can be due to complex interactions, as seen in mice in which adult size is influenced by maternal uterine genotype (Cowley, et a1., 1989). This is in contrast to nuclear-cytoplasmic gene interactions, such as cytoplasmic male sterility and the nuclear restorer genes of maize, which are due to the interaction of products encoded chromosomally and cytoplasmically (Elseth and Baumgardner, 1984). This model, described in Table 28A, includes a single nuclear gene and incorporates a maternal effect. It is assumed that the gene was homozygous in the original somaclone, and was expressed in an additive way, so that heterozygous plants would show an intermediate level of resistance. A precedence for an incompletely dominant ToMV resistance allele is the tomato resistance gene, Tm-l (Fraser and Loughlin, 1980). Because of the additivity seen between Tm-l and the somaclonal resistance, a tentative term used for the somaclonal gene in this model was Tm-3. Final designation depends on further genetic analysis to determine allelism to Tm-l. Additionally, there was a favorable change in the offspring, determined by the presence of the resistance genotype in the female parent, which was also dependent upon dosage of the resistance gene. The gene dosage ranking for this model is presented in Table 28B. Values are assigned depending on the gene dosage of the plant, and an additional lesser value was assigned that was dependent upon the genotype of the female parent. 113 Table 28. Model 2: Description of a proposed genetic model for the inheritance of ToMV resistance seen in tomato somaclone lines. The model is based on an incompletely dominant nuclear gene with a maternal effect. A) Definition of terms. B) Description of genotypes and additive rank resulting from gene dosage. A. Definition of terms. Tm-3 is an incompletely-dominant nuclear-encoded resistance gene. + is the wild-type susceptible allele. AA is a cytoplasm type determined by the homozygous nuclear resistance gene of the female parent, Tm-3/Tm-3. A is a cytoplasm type determined by the heterozygous female parent, Tm-3/+. a is the normal type of cytoplasm determined by the wild-type female parent, +/+. B. Additive ranking (gene dosage) Genotype Assigned gene dosage +/+ 0 Tm-3/+ 2 Tm-3/Tm-3 4 Female genotype Cytoplasm Assigned gene dosage 114 This ranking was composed of an additive nuclear component (incompletely dominant) and an additive maternal effect factor which increased resistance by a factor which was dependent on gene dosage of the female parent. A gene dosage value was assigned to each category, which were then added together for ranking of the genotypes. Detailed description of genotypes and additive gene dosage ranking are presented in Table 29. The homozygous starting material (A Tm-3/Tm- 3) bred true for resistance through all generations. The F1 progeny of reciprocal crosses of somaclonal lines with GCRI-26 produced plants with intermediate resistance (Tm-3/+). Because of the maternal effect, there was an enhancement of the resistance of Tm-3, resulting in a lower level of virus accumulation seen among the progeny. Upon self-pollination of the F1 to get the segregating F2 population, there would be three nuclear genotypes with respect to virus resistance: Tm-3/Tm-3, Tm-3/+ and +/+. Each of these would be more resistant in the presence of (A or AA), from the maternal effect. There would be four types of backcrosses of F1 to the parent lines; those which involved the somaclonal line twice and had the somaclonal cytoplasm (AA), those which involved the somaclonal line twice but had the susceptible cytoplasmic background (a), those which involved the somaclonal line once and had the somaclonal cytoplasm (A), and those which had the somaclonal line involved once but had the susceptible cytoplasmic background (a). Table 29 lists the predicted genotypes and additive resistance ranking (gene dosage) derived from these crosses. To examine the data for evidence of nuclear-cytoplasmic interaction, this resistance-ranking of the cross-pollinated lines and the backcrosses was used for comparison of the model with observed crosses. ELISA values were employed. Since the S:R rating system was 115 Table 29. Model 2: Additive nuclear-cytoplasmic interaction. A) Assignment of additive ranking values for different genotypes. B) Proposed genotypes and additive gene dosage values. A. Assigned gene dosage values. Progeny genotype Value Female parent genotype Value +/+ 0 +/+ 0 Tm-3/+ 2 Tm-3/+ 1 Tm-J/Tm-B 4 Tm-3/Tm-3 2 B. Description of genotypes and additive rank resulting from crosses. Parents Resulting Genotypes Additive Rank Mean Selfed Tm-3/Tm-3 All AA Tm-3/Tm-3 2+4-6 6 Reciprocal crosses (F1) Tm-3/Tm-3 x +/+ All AA Tm-3/+ 2+2-4 4 (Sc x 26) +/+ x Tm-3/Tm-3 All a Tm-3/+ 0+2-2 2 (26 x Sc) Reciprocal crosses (F2) Selfed Tm-3/+ A Tm-3/Tm-3 1+4-5 (Sc x 26) 2 A Tm-3/+ 1+2-3 3 1 A +/+ 1+0-1 Selfed Tm-3/+ 1 A Tm-3/Tm-3 1+4-5 (26 x Sc) 2 A Tm-3/+ 1+2-3 3 1 A +/+ 1+0-1 W +/+ x Tm-3/+ 1 a Tm-3/+ 0+2-2 1 (A and B) 1 a +/+ 0+0-0 Tm-3/+ x +/+ 1 A Tm-3/+ 1+2-3 2 (C and D) 1 A +/+ 1+0-1 Tm-3/+ x Tm-3/Tm-3 1 A Tm-3/Tm-3 1+4-5 4 (E and F) 1 A Tm-3/+ 1+2-3 Tm-3/Tm-3 x Tm-3/+ 1 AA T 5 (G and H) 1 AA Tm-3/+ 2+2-4 116 qualitative, based on virus presence or absence, the intermediate levels of virus titer could not be identified using the S:R ratings. As a more direct measure of virus titer, the standardized ELISA values from each plant were used to calculate the mean. To test for significant differences among the ranks, analysis of variance and range tests were done on all of the data, and results of the LSD multiple range test were presented in Table 30. In some cases there were significant differences of the experimental results between crosses with the same rank, possibly due to the relatively small sample number. When the additive gene dosage values were plotted versus mean ELISA values (Figure 16), there was a linear relationship. Linear regression analysis gave a correlation coefficient (r2) of 0.844, which indicated there was a fairly good fit. Thus a model using an incompletely dominant nuclear gene which also gives an additive maternal effect fits the experimental results found after resistance screening of the somaclonal lines. The mechanism of a nuclear gene with a cytoplasmic effect for this resistance could take many forms, involving the interaction of gene products at any of several locations. These locations could include components of the membrane of either the plastids or the mitochondria, or perhaps even the thylakoid membranes. Because the method of evaluation of resistance in these studies involved inoculation of the cotyledons, it was possible that there was still an effect from an element of the maternal cytoplasm. Further speculation on a mechanism for resistance that incorporates nuclear-cytoplasmic interaction, requires more detailed knowledge of the replication and life cycle of ToMV. Table 30. Model 2: Additive nuclear-cytoplasmic interaction. Comparison of predicted resistance rank with mean ELISA values (from 117 10 d pi) for pooled somaclonal families. additive ELISA Parents gene dosage mean SE LSD 26 x 26 0 l 75 0.015 a A) 26 x [26xSc] 1 1.38 0.064 b B) 26 x [Scx26] 1 1.45 0.063 b C) [26xSc] x 26 2 1.13 0.139 c D) [Scx26] x 26 2 1.11 0.084 c 26 x Sc (F1) 2 0.79 0.038 d [26xSc]x[26xSc] (F2) 3 0.63 0.040 e [Scx26]x[Scx26] 3 0.46 0.037 f Sc x 26 (F1) 4 0.38 0.044 f E) [26xSc] x Sc 4 0.41 0.071 f F) [Scx26] x Sc 4 0.76 0.297 de G) Sc x [26xSc] 5 0.64 0.115 de H) Sc x [Scx26] 5 0.33 0.074 f Sc x Sc (R1) 6 0.03 0.011 g 2 1. l m i a. L i g 1.5 b g 3 I V ‘5 3 1‘ 5' L 3 C3 E g ()5 O «I» ' F ‘3 5 ..l I“ o J 1 1 1 [D O 2 3 4 6 Resistance rank Figure 16. Graph of resistance ranking versus mean ELISA values for each cross and backcross between GCRI-26 and the pooled somaclonal (Line was determined by linear regression analysis.) lines. 118 SUMMARY In characterizing the resistance to ToMV in tomato somaclones, it was shown that symptom formation was suppressed and virus multiplication was limited within the plants. The somaclonal lines appeared to be normal otherwise, with the gross morphology, pollen viability and chromosome number being similar to the susceptible parental line. The resistance to ToMV found in these six tomato somaclonal lines is stably inherited through several generations, with almost complete penetrance of the resistance trait, and very little variation in expressivity in the selfed generations, thus appearing to breed true. After progeny of reciprocal crosses were screened and rated for resistance, it was shown that there was a 37.1% difference in ratings favoring progeny of the crosses in which the somaclonal lines were used as the female parent compared to the reciprocal cross. This maternal bias in the inheritance of resistance was not seen in all generations of crosses and backcrosses, eliminating involement of classical cytoplasmic inheritance. Although a maternally transmitted cytoplasmic factor was concluded to contribute to the inheritance of TMV resistance, the major portion of the resistance appeared to be determined by an incompletely dominant nuclear gene. This pattern of inheritance was found in varying degrees for all six of the resistant somaclones regenerated (representing approximately 1.6% of the somaclones regenerated in the study). Though careful records were not kept regarding the source of each somaclone, at least two of the TMV resistant somaclones were isolated from separate calli (Barden, 1985). To determine the relationships of the resistance from different somaclonal lines, test crosses between lines would be 119 necessary. If crosses are made between different somaclonal lines, the resulting progeny can be screened for resistance to ToMV. If the genes are additive, full resistance would result in the progeny, and resistance screening of the segregating F2 generation would give the information to establish whether the genes are allelic or not. For further characterization and comparison to each other, these somaclonal genes could each be tested for linkage to different markers (including isozymes and/or RFLPs known in tomato). It was shown that the resistance of these six tomato somaclonal lines derived from tissue culture was a type similar to that of the Tm-l gene, an incompletely dominant gene originally found in L. chilense (Pelham, 1972). Because the somaclonal resistance gene was phenotypically additive with Tm-l in crosses, the somaclonal gene was tentatively called Tm-3. In addition to the incompletely dominant nuclear gene(s), the somaclonal plant lines had a maternally transmitted component that enhanced the resistance of Tm-3. Because the genetic evidence for maternal cytoplasmic inheritance was atypical, a maternal effect of the nuclear gene was deduced, which could be from a nuclear- cytoplasmic interaction in which a cytoplasmic component limited the multiplication of ToMV when the nuclear resistance gene was present in the female parent. This limitation of virus multiplication was dependent on the nuclear gene dosage itself, and the gene dosage of the female parent. A similar pattern of inheritance was described in TMV-resistant hybrid tomato plants produced by conventional breeding methods between L. esculentum and L. chilense (Phillip et a1., 1965). An intermediate level of resistance was seen in the reciprocal crosses with a slight maternal bias. Further genetic characterization was not completed, 120 though it was later speculated that this resistance gene was probably Tm-I, with variation of expression due to the genetic background of the plant cultivar. In the further genetic characterization of the somaclonal resistance genes, several steps would be useful for analysis. First, the completion of tests for allelism and for linkage between Tm-3 and Tm-l should be done. This would include screening the segregating generation of the crosses between GCRI-237 (Tm-l) and the somaclonal line 247. If all of the F2 progeny are resistant, then the genes are allelic (or closely linked). To test for linkage, progeny of the following cross, ([237x26] x [247x26]) x 26, would be screened, however, because of the additive nature of the resistance, the results might not be clear. It would also be possible to determine which chromosome carries the somaclonal resistance gene. Sets of seeds are available with markers for each chromosome of tomato. By crossing the somaclonal lines with these and then screening the F2 generation, the resistance gene could be mapped to a specific chromosome. It might not be as simple as stated, since the chromosome which carries the resistance gene Tm-l has not yet been identified, though early attempts reported mapping of resistance to chromosome 5 (Clayberg et a1., 1960). To gain further information for testing proposed models of inheritance of the somaclonal resistance genes, it would be useful to try to separate the nuclear gene from the cyt0plasmic gene, thus eliminating the added complexity. It might then be possible to determine whether the somaclonal resistance is encoded by a single, incompletely dominant gene or not. Inversely, the strength of the 121 cytoplasmic factor could be determined (if it could be separated from the nuclear portion of the resistance). As another step in characterization, screening of the next generation of the crosses and backcrosses would be particularly useful, especially if these progeny were selected from parent plants which were identified as susceptible versus resistant. This would allow for the confirmation of genotypes that were predicted in the genetic models. In all further screening, the experimental design should follow a randomized block design, with careful attention paid to environmental differences between replications. Though many precautions were taken in the characterization of the somaclonal resistance, additional steps may help in identifying the sources of the variation in the experimental results. The amount of genetic variation could then be determined and separated from the environmental variation. When these are separated, the differences between progeny of each cross would be more meaningful, and generation mean analysis could be done. Overall, the use of somaclonal variation to generate a beneficial trait was successful. Yet it was suprising that mutations occurred in as many as six somaclones, resulting in a resistance trait which is very similar to one found in wild tomato. This leads to speculation on mechanisms of mutation, which might involve areas of DNA that are more susceptible to mutation, possible involvement of deletions of highly repetitive gene families, or secondary structural changes which change the expression of genes. These can only be tested by gaining more information on the role of the resistance genes in inhibiting virus multiplication. The next part of this work includes characterization of the type of resistance observed in the tomato somaclones. CHAPTER 2 CHAPTER 2 WW“ INQEWWQEIHB ToMV RESISTAEI SOMACLONES INTRODUCTION In characterizing the type of resistance to ToMV found in tomato somaclones, several approaches were taken. Plants were challenged with different strains and viruses under different conditions, systemic virus movement was studied, and infected protoplasts were studied. This resistance, encoded by both nuclear and cytoplasmic genes, offered a unique system for study. The objectives were to study the mechanisms involved in the resistance and to use that knowledge to gain a better understanding of the plant-pathogen relationship. To gain an understanding of the mechanisms that may be employed in the resistance of the tomato somaclones, the current knowledge about the cycle of the virus and about other resistance genes was reviewed. LITERATURE REVIEW Iemetg mefieie xiree life eyele. ToMV, a member of the tobamovirus group, is composed of a single strand of RNA and is coated with protein subunits to make a straight rod of about 300 x 18 nm. The virus particles enter the epidermal cells through breaks in the cell membrane caused by injury (Figure 17a). It is believed that 1 x 106 virions per cell are required to initiate infection (Takebe, 1983). However, when plants are inoculated with more than one strain, the infection is established with only one of the strains (Otsuki and Takebe, 1978). 122 123 REPLICATION b UNCOATING C R F d I! I 29"“: D D“ - + \e + O TOMATO MOSAIC VIRUS ‘."' LIFE CYCLE é... TRANSUTION F— e 9 7 —* "'— 7"" , an m .15 <3 , \ 2"- ¢ ., CELL-TO-CELL / a.) ‘1 1 / I VIRUS MOVEMENT Figure 17. Tomato mosaic virus life cycle. a) virus enters cell through broken membrane and b) is uncoated while being translated from the 5' end c) the RNA is replicated (found in the double-stranded replicative form) and many subgenomic RNA's are made (found as the replicative intermediate with single-stranded RNA tails). e) viral proteins are synthesized from the subgenomic RNA's on the plant ribosomes. Virus is transported from f) cell to cell and then g) throughout the plant. h) typical symptoms are mosaic leaves, and viral inclusions can be detected in the leaf hair cells. 124 Once in the cell, the virus is uncoated as the 5' end is translated (Figure 17b). This process, called cotranslational disassembly (Wilson and Watkins, 1986), occurs on the host ribosomes. The first protein product is associated with replication of the viral RNA, probably a component of the viral replicase. Replication of the negative strand of RNA (opposing base pairs of the coding strand) begins early in the infection (Figure 17c). Once there are copies of the negative strand, double-stranded RNA (dsRNA) can be extracted from plant cells (Ralph and Wojcik, 1969). This double-stranded form has been called the replicative form (RF). From the negative strand of RNA, many sub- genomic copies are made of the RNAs which code for viral proteins (Figure 17d). This complex can be isolated as the replicative inter- mediate (RI), with many single-stranded tails. The subgenomic pieces of RNA are then translated on plant ribosomes (Figure 17e). Four viral proteins have been well characterized (Goelet et a1., 1982; Atebekov and Dorokov, 1984). A 126-kD protein is encoded in the first part of the genome (5' end) and a 183-kD protein has also been found which is a readthrough product of the 126-kD protein. Both of these proteins are implicated in replicase function, and have been detected in the cytoplasm of infected tobacco cells by immuno-gold labelling (Hills et a1., 1987). The viral proteins, including the 126- and l83-kD proteins, are thought to be associated with granular inclusion bodies found in virus-infected cells (Saito et a1., 1987; Hills et a1., 1987). A 30-kD protein is made from another reading frame downstream from the first two proteins. This protein is involved in cell-to-cell transport of the virus (Ohno et a1., 1983; Deom et a1., 1987). This protein has been localized by immuno-gold labelling to the plasmodes- 125 mata (Tomenius et a1., 1987). Virus strains with mutations in this coding region are unable to move from the initially infected cell (Leonard and Zaitlin, 1982). The fourth gene is the 18-kD coat protein gene, which is found near the 3' end of the genome. Coat protein accumulates in the thylakoid membranes of the chloroplasts (Reinero and Beachy, 1986) as well as throughout the cytoplasm, and assembles with the nascent positive-stranded RNA to form new virions. Virus transport, reviewed by Atabekov and Dorokov (1984), is active at two steps. Cell to cell transport, probably via plasmodesmata, is active (Figure 17f). The virus then enters the phloem, is rapidly carried to the roots and the top of the plant (Figure 17g), and then over a longer period (15 to 20 d) spreads to all parts of the plant. It has been suggested that movement both into the phloem and back out to the mesophyll cells are active steps in viral movement. After movement of the virus through the plant, symptoms of virus infection are apparent, including the typical mosaic mottling of the leaves (Figure 17h) and the appearance of viral inclusions in the leaf hair cells. The inclusions are made from millions of virus particles which form hexagonal crystals that can be detected with a light microscope (Figure 1). The life cycle of the virus is complex and not completely understood at this time, but what is known can be used as a basis for speculating about the mechanisms which plants use to interfere with the life cycle to cause resistance. 126 TM! resistance genes 13 tomato. There are three well-characterized ToMV resistance genes in tomato, which have each been incorporated into cultivated tomato from wild tomato species. There is also pathogen- derived resistance in transgenic plants expressing TMV coat protein (CP) (Nelson et a1., 1988). Table 31 lists the resistance genes and several characteristics of the resistance traits. Chexeegeriegiee 21 reeieeenee geeee. There are three well-studied genes for resistance to ToMV in tomato. The gene Tm-I is co-dominant for limitation of virus replication and dominant for suppression of symptom formation (Fraser and Loughlin, 1980). At a different locus, the allelic genes, Tm-Z (Hall, 1980) and 161-22 (Alexander, 1963) effectively prevent infection of ToMV by a hypersensitive response, usually evident as local lesions. At higher temperature, or in heterozygous plants, the hypersensitive resistance response may cause systemic necrosis. These three genes have been bred into many commercial cultivars as Tm-l/Tm-l, Tm-Z/Tm-Z2 to give complete resistance to date. Also, each of these alleles has been incorporated into isogenic lines of tomato, and have been tested with different virus strains (see Table 31). This work done by Pelham (1972) has been very useful in aiding studies of resistance in tomato, and serves as an ideal system for study of resistance mechanisms. Because the resistant somaclones in this study were derived from GCRI-26, one of Pelham's near-isogenic plant lines, direct comparison of resistance can be made between the somaclones and the other resistance alleles. The transgenic tomato lines that express TMV-GP provide resistance to all strains of ToMV tested, but are not resistant to TMV-RNA (no coat protein) (Nelson et a1., 1988). This mechanism of resistance, which 127 Table 31. List of TMV resistance genes in tomato and some of their characteristics. Gene Mode of Type of Overcoming Response to Resistance Gene Source Inheritance Resistance Strains Temperature Level Tm-l‘I Lycopersicon nuclear limited ToMV-1 sensitive] protoplast chilense (dominant/ multiplication. ToMV-1.2 resistant co-dominant) symptom ToMV-1.22 suppression PI! L. nuclear limited n.d. n.d. protoplast 235673b chilense plus multiplication (Tm-1) maternal Tm-2c L. chromosome hypersensitive ToMV-2 sensitive callus peruvianum 9 (local lesion) ToMV-1 2 co-dominant 2d 2 Tm-z L. chromosome hypersensitive ToMV-2 sensitive callus peruvianum 9 (local lesion) ToMV-1 2 co-dominant +CP pathogen- nuclear cross- TMV-RNA n.d. protoplast trans- derived protection senic (TMV) Sc: tissue nuclear, limited ToMV-1 resistant n.d. cultured maternal multiplication L.esculentum effect .Bolmes, 1954. bPhillip et a1., cflell, 1980. dAlexander, 1963. .Nelson et el., tBarden et a1., 1988. 1986. 1965. 128 resembles cross-protection, is thought to be due to inhibition of uncoating and initial translation of viral protein (Register and Beachy, 1988). Iempegetuge gensitivity. When resistant tomato plants are inoculated with virus and incubated at high temperature (28-300 C), there is sometimes a breakdown in the resistance. With Tm-l, the results have been ambivalent with reports of temperature sensitivity, especially when heterozygous (Pelham, 1972) and stability (Fraser and Loughlin, 1980). The genes Tm-Z and Tm-22 both lose resistance at higher temperature (Pelham, 1972; Stobbs and MacNeill, 1980; Cirulli and Alexander, 1969), allowing systemic virus multiplication and spread. When such plants are moved to normal temperatures, the plants become systemically necrotic. Protoplast iefeceion. When isolated protoplasts were inoculated with ToMV, the cells with Tm-l inhibited virus multiplication (Motoyoshi and Oshima, 1977), thus the resistance was expressed at the cellular level. However, protoplasts from lines with Tm-Z or Tm-22 were as susceptible to ToMV inoculation as were proto- plasts from the fully susceptible line, GCRI-26 (Motoyoshi and Oshima, 1975; Stobbs and MacNeill, 1980). The resistance was not expressed at the cellular level, but when callus cultures were inoculated with ToMV, the Tm-Z and Tm-22 lines showed resistance (Toyoda et a1., 1983). There was no resistance seen in GCRI-26 callus cultures. Protoplasts from transgenic tobacco plants (similar to the tomato transgenic resistance) were resistant to infection by TMV, but not by TMV-RNA or TMV which had been treated at pH 8.0 (Register and Beachy, 1988). This result provides evidence for the hypothesis that protection is due to prevention of TMV uncoating. 129 RNA polymerase activity. In ToMV-infected susceptible tomatoes, activity of RNA-dependent RNA polymerase increases two times the normal levels observed in healthy plants (Evans et a1., 1984). Infection with most RNA viruses leads to increased enzyme activity (May et a1., 1970; Romaine and Zaitlin, 1978), and some or all of this activity may be responsible for virus reproduction. In tomato plants with the Tm-l resistance gene, there is no increase in this RNA polymerase activity (Evans et a1., 1984). Furthermore, when RNA-dependent RNA polymerase was purified and ToMV-RNA added as a template, the replicative form (RF) and replicative intermediate (RI) of ToMV were detected only when enzyme of ToMV-infected susceptible plants, but not the enzyme from healthy susceptible plants or inoculated plants containing the Tm-l gene was used. Strains of TMV that overcome Tm-l resistance did stimulate RNA polymerase activity (Evans et a1., 1985). These results suggest that Tm-l resistance operates by preventing synthesis or modifying RNA- dependent RNA polymerase activity needed to form a functional viral replicase, though other explanations are possible (Fraser, 1985). Grafting experimen£§. Study of virus movement in reciprocally grafted tomato has been done for each resistance gene. Selman and Yahampath (1973) made reciprocal grafts between a cultivar with the Tm-l gene and a susceptible cultivar with a different genetic background. The scion was inoculated with ToMV and virus was assayed by inoculation on N. glutinosa. When the rootstock was susceptible, the resistant scion was less resistant, allowing the multiplication of virus. When the rootstock was resistant, the susceptible scion became infected with virus, indicating that there was no viral inhibitor being transported upward. In either case, the virus moved freely across the graft and 130 replicated throughout the plant, indicating that the Tm-l gene did not inhibit transport of virus or cause systemic resistance. Pilowsky (1971) studied reciprocal grafts between plants with Tm- 22 and infected susceptible plants, and found that in either position (rootstock or scion) the Tm-ZZ plants became systemically necrotic, even when the graft was severed after seven days. Stobbs and MacNeill (1980) studied virus movement when grafts were made between plants expressing Tm-Z and susceptible plants. Stem inserts of one genotype were made between rootstock and scion of the other genotype, and ToMV was inoculated on the rootstocks, after the graft unions were established. The resistant insert did not alter virus movement; when a leaf of the susceptible rootstock was inoculated, the virus moved into the scion, but when a leaf of the resistant rootstock was inoculated, the virus remained in the inoculated leaf. When grafted as rootstock or scion, the resistant plants developed systemic necrosis when the grafted partner was TMV-infected. In the same study, virus movement was found to occur through the phloem when examined using electron microscopy. It was suggested that the virus by-passes the normal genetic resistance mechanisms by entering the resistant mesophyll from the phloem. The virus was localized if the only inoculum was on a resistant leaf. Preliminary reports of grafting experiments with transgenic tobacco plant material are intriguing. When inserted stem sections of a transgenic plant with TMV-GP were grafted between the top and bottom of a systemic host of TMV (N. tabacum cv. Xanthi) followed by inoculation of a lower leaf with TMV, virus movement through the graft was not inhibited (Wisniewski and Beachy, 1988). However, when the stem section included a leaf, the virus did not move through the graft to infect the top of the plant. One explanation is that resistance might be due to 131 the export of viral coat protein from the transgenic leaf into other parts of the plant. It was suggested by Beachy (1988) that this resistance may be caused by a physiological boost from nutrients produced in the healthy transgenic leaf which caused greater expression of the CP in the phloem cells, thus stopping transport. Be§1§£egee mechanigms. Hypothetically, there are three kinds of general resistance mechanisms (Fraser, 1985). Positive resistance is that in which the host produces a factor which stops the virus. The hyper- sensitive response to virus might exemplify this kind of resistance mechanism. Negative resistance is due to a factor which the host lacks that is essential to viral replication, as might be the case with a protein product such as RNA-dependent RNA polymerase to which a specific virus may be highly adapted. The other category includes physical or chemical barriers, to either prevent access of the virus or to cause destabilization of the virus with changes in pH or ionic strength. There are also a variety of resistant phenotypes (Fraser, 1985), including plants which are entirely symptomless, plants which localize the virus into necrotic or chlorotic lesions, and plants which allow systemic spread of the virus but are not affected with symptoms. In any of these phenotypes, the increase of virus in time and space is limited, as shown by reduced levels of virus multiplication at the cellular, plant or crop level. The possibility that host genetic background may influence resistance expression emphasizes the importance of using near- isogenic lines when comparing resistance genes. It becomes more critical to test the resistance in different ways in order to establish a mechanism. The following section of experiments does this. 132 MATERIALS AND METHODS RESISTANCE RESPONSE TO DIFFERENT CHALLENGES Beepogse e: somaclones £9 differen; gigee egzains. Tomato seeds were planted and cultivated as described previously. Groups of tomatoes representing the R1 generation of all six somaclonal lines and the susceptible line, GCRI-26, were divided and inoculated with ToMV-O, TMV-Fl or a strain that overcomes the resistance of the Tm-l gene, TMV-l (described by Pelham, 1972 and kindly provided by F. Motoyoshi, University of Tokyo, Japan). Assay of the inoculated plants was done at 10 d pi, and continued for 20, 30, 50, and 70 d for ToMV-O and TMV-F1. Plants that became infected with TMV-1 were discarded immediately after assaying to prevent unintentional spread to other tomatoes. Assay of virus presence was done using ELISA as described except for TMV-1, which was detected by observation of symptoms and inclusion bodies in leaf hair cells. Comparisons of ToMV-O and TMV-F1 were repeated three times with six to 16 plants of each line. Comparison of TMV-l to ToMV-O was done once with twelve plants of each line. mammmfimm. Grouvs of tomatoes representing each of the R1 generation somaclonal lines (except 12) and GCRI-26 were divided and inoculated with the pepper strain of cucumber mosaic virus (CMV, CMI/AAB Description of Plant Viruses, 1979) or with the wild tobacco strain of tobacco etch virus (TEV, CMI/AAB Description of Plant Viruses, 1980), which were both provided by T. Zitter, Cornell University, New York. The inoculum was prepared from frozen, infected tobacco tissue which was ground at 1 g:20 m1 of 0.01 M NaK-phosphate buffer, pH 7.0. Detection of the virus was done using 133 ELISA, with gamma-globulin specific for each virus. Each gamma-globulin was purified from rabbit antiserum (from American Type Culture Collection, Rockville, MD) as described previously. Plants were assayed at 10 and 20 d pi. Twelve plants of each line were tested with each virus. Resnonse 9f §omaclones £2 inocnlagion ene incubatinn en niggerenn e u e . Groups of tomatoes representing R1 and R2 generations of somaclones, GCRI-26 (+/+), and the resistant tomato line GCRI-237 (Tm-l/Tm-l) were inoculated with ToMV-O and divided into sets which were incubated at 250 C or 300 C, with otherwise similar conditions. Plants were assayed for the presence of virus at 10, 20, and 30 d pi (and one group was assayed at 50, 70 and 150 d pi as well) using ELISA. The first experiment included six plants of each line for each temperature and was replicated using six to nine plants of each line for each temperature. VIRUS MOVEMENT Vigne movement in eenene. Plants were transplanted into 4" pots at 12 d. The groups represented R2 plants from somaclone lines 12, 215 and 247, and the susceptible line, GCRI-26. One group was inoculated with ToMV-O when the plants were 32 d old and the second group was inoculated at 25 d old. The outermost leaflet of the first true leaf was marked by punching a hole with a wide-bore needle, and after dusting with carborundum, it was inoculated with virus. The inoculum was not rinsed off. Ten to 12 plants of GCRI-26 and somaclone 247 were tested in each replication. At each time point (4, 7, 10 and 20 days pi), leaflets were removed from each leaf of the plant as shown in Figure 18. Two or three plants of each line were sampled each day. Once plants had been 134 Figure 18. Diagram of sampling of inoculated tomato plants for detection of virus spread over time and space. The triangle indicates the leaflet which was inoculated. Numbering and lettering identified each leaflet for later reference. 135 sampled, they were not used for further assay, to eliminate the liklihood that plants were inadvertantly inoculated during sampling. Leaf samples were assayed using ELISA, and the absorbance values for leaflets from each corresponding leaf were averaged. Plants from the second repetition were used to isolate protoplasts for detection of infection at the cellular level. As controls, susceptible and resistant plants were mock-inoculated and assayed. For an added control in the second group, the inoculated leaflet was detached at 48 h to determine the time required for the virus to leave the inoculated leaf. 21135 mnvemeng in g1e££ee geneneee. To study the movement of virus between reciprocal grafts of resistant and susceptible plant material, several types of grafts were established. Whip grafts (see Figure 19A) were done using young seedlings by wetting a clean razor blade and removing the top of l8-d-old plants with a V-shaped cut. Tops were removed from two plants and exchanged, resulting in a rootstock with cotyledons only, and a scion with three young leaves. The graft was held in place using 1/2" adhesive tape. Plants were sealed in polyethyene bags to increase the humidity, and were allowed to bond for several weeks. A cotyledon of the rootstock of the grafted plant was then inoculated. Of the ten whip grafts made, three grafted plants survived for the assays. The second set of plants was grafted, in this case using older plants (20 to 30 d). Again, V-shaped cuts were made near the tops of two plants and the tops were exchanged. After the whip graft was established (25-30 d), there were two true leaves remaining below the graft plus secondary shoots, and two to three young leaves on the scion. The outermost leaflet of the first true leaf of the rootstock was 136 Figure 19. Diagram of whip graft (A) and whip and tongue graft (B). 137 inoculated. Of the ten grafts made, again only three were successful and used for the assays. The third type of graft was established using an approach graft which incorporated a tongue (whip and tongue graft, Figure 19B). This method was used successfully on older plants (33-39 d) and was less stressful to the plants. The graft was made on the stem between the third and fourth true leaves in most cases, and was bound with adhesive tape. After two weeks, one stem was severed below the graft and the opposing top was removed. The outermost leaflet of the first true leaf was inoculated a week after that. Of the ten whip and tongue grafts made, five were used for assay. Two others were contaminated with virus before the date of inoculation and were eliminated. Assay of the grafted plants was done using ELISA at 0, 5, 7 and 10 d pi. Some plants were assayed further at 21, 34, 39 and 44 d pi. The first group of grafted plants were assayed from every leaf of both rootstock and scion at each time point. Later groups were assayed in one or two leaves from both parts. Three pairs of reciprocal grafts were established between somaclone 247 (R2) and GCRI-26, and one pair between between somaclone 12 (R3) and GCRI-26. Two control grafts were made between two plants of GCRI-26 to eliminate the possibility that delayed virus spread was due to the graft itself. PROTOPLAST ISOLATION AND SCREENING FOR INFECTION Egotenlast isoiation. Tomatoes were grown in the laboratory (22-250 C) under cool white fluorescent lights supplying 60 uEm'zs"1 on a 16 hour light cycle. The plants were fertilized daily when they were watered with one tablespoon of Peters 20-20-20 per gallon. At about four weeks 138 plants were used for isolation of protoplasts. The condition of the plants was optimal for about six weeks when new leaves were used, but the yield of protoplasts per gram of leaf declined as the plants aged. Light green, tender leaves were picked, weighed and used to isolate protoplasts. Protoplast isolation procedures were adapted from Kassanis and White (1974). Leaves were surface sterilized by washing for ten min in 15% bleach solution containing two drops of Tween-20 per 250 ml. In a laminar flow hood, the leaves were then rinsed three times with sterile distilled water. All following steps were performed using aseptic techniques. From leaves which were infected with TMV, leaves were stacked and sliced transversely in 1 mm thin sections. Leaf strips were placed in a petri dish containing 20 ml of an enzyme solution consisting of 1% cellulysin, 0.5% macerase, (Calbiochem, Behring Diagnostics, San Diego, CA), 0.4 M sorbitol in cell protoplast wash (CPW) salts (0.2 mM KH2P04, 1 mM KNO3, 1 mM MgSO4'7 H20, 1 uM KI, 0.1 uM CuSOa'S H20, and 10 mM CaC12'2 H20) at pH 7.0. This was incubated on a slow shaker in the dark at room temperature, until the solution was visibly green and many protoplasts could be seen when viewed using the inverted microscope (12 to 18 h). The enzyme solution containing protoplasts and leaf debris was filtered through a 60 um sieve into a petri plate, and remaining leaf debris was gently rinsed with CPW salts with 0.4 M sorbitol at pH 6.5. The protoplasts were transferred carefully to a 30 m1 round-bottomed centrifuge tube to pellet the protoplasts. The protoplasts were centrifuged at 35 x g or 400 rpm in an HNS-II IEC No. 2355 tabletop centrifuge for 10 min. The supernatant was removed by suction, the 139 pellet was gently resuspended by rolling in a small amount of washing buffer, and the protoplasts were moved to a 15 m1 tube. The protoplasts were washed in 15 ml of sorbitol-CPW buffer, centrifuged again at 35 x g and the supernatant was removed by suction. The protoplasts were rinsed a total of three times in this manner. The final step in purification of the protoplasts was to float the intact protoplasts on a solution made of Lymphoprep (Accurate Chemical and Scientific Corp., Westbury, NY) and 0.5 M sucrose-CPW salts at 1:2 for healthy protoplasts or at a 2:1 ratio for protoplasts which were isolated from virus-infected plants. To determine whether the exposure to Lymphoprep (9.6% w/v sodium metrizoate and 5.6% w/v Ficoll; 1.077 g/ml density and 300 mOsm osmolarity) inhibits virus infection or detection, some isolations of protoplasts were done omitting the flotation step. Yield of protoplasts was determined by counting them with a hemocytometer, and calculating the number of protoplasts released per gram of starting material. Concentration of protoplasts was then adjusted to the proper amount for each experiment. BIQLQELQEE viebiiity. The viability of protoplasts after isolation was determined in two ways. Protoplasts were examined using an inverted light microscope and rated for their general appearance. Living cells remained perfectly round with an intact membrane. Dead cells were misshapen or shriveled. The second method to determine viability was to observe cells after the addition of drops of Evan's blue dye (0.05% w/v Evan's blue in 0.7 M mannitol, pH 7). Live cells excluded the dye, but dead cells took on a blue color after several min. W129 and ital-slag 2!. images mm Following the procedure by Otsuki and Takebe (1969), the infected protoplasts were harvested and concentrated by centrifugation. A small drop of solution 140 containing protoplasts was placed onto a glass slide which was coated with Mayer's albumin (egg white + phenol), swirled to spread the cells evenly and allowed to dry. The dried protoplasts were fixed by immersion in 95% ethanol for 10 min at room temperature, followed by a rinse for 10 min in PBS. To stain the protoplasts, 20 ul of FITC- conjugate (antibody against viral coat protein labelled with fluorescein isothiocyanate) was placed on the slide, covering the cells. The slide was incubated for one h in a moist chamber at 370 C in a water bath. The chamber was lined with wet filter paper so as to maintain a high level of humidity. Excess FITC-conjugate was removed by a 10 min rinse in PBS, then the sample was prepared for viewing by carefully blotting the excess PBS away, adding a drop of 10% glycerol (in PBS) and placing a coverslip over the cells. DEEQQELQB e; yinne infieenien by iluorescence niexeeeeny. FITC-labelled cells were viewed with a Zeiss GFL epiluminescent microscope equipped with exciter filters KP-490 and LP-445, dichroic reflector 510 and barrier filter 520. Infected cells contained specks which were a bright yellow-green in color. Uninfected cells exhibited slight light green fluorescence, and dead cells appeared dull orange-brown in color. When chlorophyll was present, cells fluoresced red. EhQEQEEQEEX- Photographs of the fluorescing cells were taken with ASA 400 daylight film at 12 volts. Exposures were between 12 min (for heavily infected cells) and 28 min (for uninfected cells). Ezenerenien efi EllgeenningeEee The procedure used was adapted from that of Spendlove, 1967. Serum was collected from rabbits following three virus injections at 2-week intervals. The gamma globulin fraction was precipitated by the dropwise addition of 25 ml (NH4)280A (saturated 141 solution) to 50 ml of serum with stirring. This was mixed for 3 h at 00 C. The precipitate was collected by centrifugation at 5000 rpm in a SS-34 rotor for 15 min. The pellet was then dissolved in 40 m1 of PBS, and was precipitated a second time by the dropwise addition of 20 m1 saturated (NH4)2804 while stirring for one h at 00 C. After centrifugation at 6000 rpm for 15 min, the pellet was dissolved in 6 ml PBS. The gamma globulin was then dialyzed against five changes of PBS over a period of 24 h. The solution was then clarified by centrifugation at 14,000 x g for 30 min and the supernatant containing the gamma globulin was collected. The protein concentration was determined by measuring the absorption at 280 nm, using the extinction coefficient of 1.8 for rabbit gamma globulin (McGuigan and Eisen, 1968). To label the gamma globulin with fluorescein isothiocyanate (FITC) (ICN Immunobiologicals, Lisle, IL) 2 ml of the purified gamma globulin was stirred while adding 1.2 ml of FITC solution dropwise. For the FITC solution, 15 mg FITC was mixed with 1 g protein, made in freshly prepared 0.1 M NazHPOa. An additional 0.4 ml of freshly prepared Na2P04 was added, followed by the addition of 0.4 ml 0.04 M NaOH (pH 9.4-9.6). The solution was incubated for 30 min at 250 C without stirring. The solution was loaded onto a Sephadex G-25 column (2.5 x 15 cm) which was previosly equilibrated with PBS. The conjugate was then eluted with PBS and the first visible (greenish-yellow) band was collected in fractions. The eluate was centrifuged at 15,000 rpm for 30 min to clarify. The supernatant containing the conjugate was adjusted to 0.1% NaN3 and stored at 4° C. MWfimeamfi—etflm geeeeee leeyeen Acetone powder of tobacco leaves was prepared according to the procedure of Otsuki (1976). Ten g of healthy leaves were picked, 142 placed in 100 ml cold acetone (00 C) and homogenized at maximum speed in a Waring blender for 5 min. More acetone was added to make a volume of 300 ml. The homogenate was washed on a Buchner funnel lined with Whatman No.1 filter paper with cold acetone until the filtrate was nearly colorless. The homogenate was further pulverized by grinding in a dry mortar, resulting in an almost powder-like substance. This was dried under vacuum overnight and stored in a dessicator at room temperature. The acetone powder was prepared for cross-absorption by several washes with PBS, pH 7.0. To 40 ml PBS, 0.2 g of powder was added and stirred for 10 min. The suspension of powder was centrifuged at 3000 rpm in an SS-34 rotor for 7 min, and the pellet was resuspended in 40 ml PBS, and centrifuged again as before. The pellet containing the acetone powder was saved, resuspended in 35 m1 PBS and divided into two tubes for cross-absorption of the different FITC-conjugate bands collected from the Sephadex column. To determine the optimum amount of cross- absorption necessary, one tube contained approximately twice the amount of the pellet as the second tube. Each fraction was centrifuged again and 2 m1 of FITC-conjugate solution was added to each pellet. This was stirred with the aid of a glass rod and the mixture was incubated for 30 min at 370 C. After unwanted antibodies were absorbed to the tobacco acetone powder, a centrifugation at 3000 rpm for 7 min removed the solids, and the supernatant fraction containing the FITC-conjugate was saved. Tests for the optimum amount of acetone powder needed for cross- absorption were made by trial and error method using identical sets of infected protoplasts. The various fractions were labelled and stored at 4° C. 143 RESULTS AND DISCUSSION RESPONSE TO VIRUS CHALLENGE Reenenee £2 different stgains ef ToMV. The somaclonal lines displayed similar resistance to ToMV-O and TMV-F1, but were not resistant to TMV-1 (Table 32). This response was similar to that reported for plants with the Tm-l gene for resistance (Pelham, 1972). Table 32. Response* of somaclonal lines to inoculation by three strains of TMV. Line ToMV-O TMV-Flavum TMV-1 12 R R S 215 R R S 219 R R S 247 R R S 322 R R S 330 R R S GCRI-26 S S S *Ratings were based on the presence of virus, detected by ELISA, symptoms or inclusion bodies. R - resistant, S - susceptible. Remus: 2f asses-1510.31 lines 122 22112.: rite-tea. After inoculation with ToMV-O, the somaclones were virus-free for more than 30 d (Figure 20). There were several plants among somaclone lines 215, 219, 247 and 330 which showed a delay in infection after inoculation with CMV, but all plants were infected by 20 d pi. When inoculated with TEV, all plants had high levels of virus at 10 d pi. It was concluded that the soma- clones which had been selected for TMV resistance were not resistant to other viruses. This was in contrast to a type of resistance found in a 144 26 216 219 247 322330 ' 26 216 219 247 322330 ° 26 216 219 247 322330 0) :3 l E 1 > CMV TEV IL 0 1 L \ * } m 111 . 1 3:1 o s —- ' * u. ' ' "’3 \ z 006 p 4 < 1 .J 1 a _. u_ 0.4 1' '1 z '1 9 Oak * \\‘ r- - ,1 tr . . \ 2 \ 1 ‘1 .1. 0 . l k .1.- .l. .1... .1... .3... O I l 10 DAYS 20 DAYS Figure 20. Response of resistant tomato somaclonal lines to virus challenge with ToMV, CMV or TEV. 145 tobacco line, T.I. 245, which gave a non—specific resistance to virus infection (Thomas and Fulton, 1968) and which was correlated with a reduced number of ectodesmata. Response pf eomacional pesistance pp nigh tempepetupe. The somaclonal resistance to ToMV was not changed when inoculated plants were grown at high temperature (Table 33). The resistance ratings of the somaclone lines at 25 and 300 C are similar, but plants did not thrive in the warmer conditions and were stunted. Severe stunting Table 33. Ratings (S:I:R) of plants inoculated with ToMV-O and incubated at either 25 or 300 C for 30 d. ( ) indicates number of plants tested. Plant line 25° c 30° c Somaclone 12 0:0:6 (6) 0:0:6 (6) Somaclone 215 0:0:15 (15) 0:0:15*(15) Somaclone 219 0:0:15 (15) 0:0:15 (15) Somaclone 247 0:0:15 (15) 0:1:l4*(15) Somaclone 322 1:0:11 (12) 0:2:10 (12) Somaclone 330 0:2:10 (12) 1:0:11 (12) GCRI-26 12:020 (12) 12:0:0 (12) GCRI-237 (Tm-I/Tm-l) 0:0:9 (9) 2:0:7 (9) *group with plants which recovered from infection. was seen in infected GCRI-26 plants at the higher temperature. There were several cases of ‘recovery' in the somaclones (marked with asterisk), when a plant showed high virus titer at one sample time and did not show the presence of virus at the next sampling. Two plants of 146 line 237 (Tm-l/Tm—l) which had a medium virus titer at 10 d died after necrotic collapse before they could be sampled again at 20 d. It is possible that line 237 was less resistant at the higher temperature, however, due to the small sample size no conclusive statement can be made. The somaclonal lines were not affected by the higher temperature. MOVEMENT OF VIRUS WWW e QMELW tibeandmaistantm. The movement of ToMV-O was fairly rapid in susceptible plants (Figure 21). At four days, virus was detected in the inoculated leaflet and in neighboring leaflets. By the seventh day after inoculation, there was a high titer of virus in the upper leaves of the plant, which were expanding with rapid growth. The virus spread through the rest of the plant. By 20 d, the whole plant contained a high virus titer. This result substantiated the classical work done on virus movement in tomato by Samuel (1934) and the more recent report of movement of potato spindle tuber viroid in potato (Palukaitis, 1987). The long-distance, intraplant transport was consistent with the pattern of movement of photosynthetic products, moving in the phloem from fully expanded leaves, to developing upper leaves. When this experiment was repeated using somaclonal line 247, high levels of virus did not develop anywhere in the plant. The inoculated leaf had a detectable level of virus, and over time, a trace level of virus was measured in the upper parts of the plants, but at the lower limits of detection by ELISA (values in the range of 0.02 to 0.09 AU). 147 20 d 20 d Figure 21. The movement of ToMV-O in tomatoes. Susceptible GCRI-26 tomatoes (A) allow the movement of virus as illustrated by the shaded leaves. At four d, the virus was spreading in the inoculated leaf, then moved quickly to the upper part of the plant. After 20 d the entire plant contained virus. In the resistant somaclonal line 247 (B), the virus remained in the inoculated leaf. An open triangle marks the inoculated leaflet. 148 Intermediate levels of virus (ELISA values in the range of 0.05 to 0.32 AU) were detected in plants of line 12 at 15-20 d after inoculation. It was evident that the virus was allowed to multiply and move systemically in these plants, but the multiplication was limited, since the virus titer was much lower than expected for fully susceptible plants. Differences in the resistance gene from each line may explain the consistently greater virus titer in somaclone line 12 compared to line 247 . It would be of interest to test for complementation between these two lines. Perhaps both somaclonal lines limit replication, but line 247 may limit systemic transport of the virus as well, but at this point it is difficult to separate the two effects. npyeneng pf yipne in grafted pienne. In the control grafts that were made with GCRI-26 on GCRI-26, there was no delay in virus movement (Figure 22A). Virus appeared in the rootstock leaves and the scion leaves in equal amounts at five d (ELISA values from 0.50 to 0.89 AU) and at ten d (values from 1.11 to 1.20 AU). Because the virus multiplied and spread to all parts of the grafted plant quickly, the possibility that limitation of virus movement due to the graft union was eliminated. When the susceptible line GCRI-26 was used for the inoculated rootstock and the resistant line 247 as the scion, the virus replicated and moved throughout the leaves of the susceptible rootstock to attain a high titer (ELISA values from 1.70 to 1.75 AU) by 7 to 10 d (Figure 22B). In each case, accumulation of virus was delayed in both rootstock and scion compared to the control grafts. In one graft (Figure 22, B1) the virus was delayed in multiplying to a high titer in the resistant scion, even with the presence of the large virus concentration in the 149 MEAN ELISA VALUES (AU at 405 nm) M Rootstock 0 d 5 d 7 d 11 d 21 d 39 d A) Control 1. zg 0.01 0.59 1.14 1.68 -* - 26 0.00 0.50 1.11 1.68 - - 2. 2g 0.00 0.86 1.20 1.68 - - 26 0.00 0.89 1.18 1.68 - - B) 247 scion l. 231 - 0.16 0.04 0.02 1.35 1.03 26 - - 1.75 1.75 1.77 0.80 2. 231 - 0.02 0.00 0.00 1.36 1.00 26 - - 0.14 0.20 1.75 0.88 3. 251 0 00 0 14 1.55 - - - 26 0.00 0 26 1.64 - - - C) 247 rootstock 1. 2g - 0.08 0.04 0.09 0.01 0.33 247 - - 0.20 0.24 0.11 0.69 2. 2g 0.00 0.00 0.00 0.00 0.00 0.02 247 0.00 - 0.00 0.00 - - 3. 2e 0.00 0.02 0.00 0.09 - - 247 0.00 0.00 0.00 0.07 - - *not determined Figure 22. Results of experiments in which grafted plants were inoculated in the outermost lower leaf of the rootstock, and the movement of the virus was followed, using ELISA to measure virus titer in young leaves of the rootstock and scion. Grafts were made between A) two susceptible plants as a control. B) and C) reciprocal grafts between resistant (somaclonal line 247) and susceptible (GCRI-26) plants. Illustrations represent grafts at 10 d. 150 rootstock. In two other grafts of this type, there was a lower titer of virus present in the rootstock, and virus moved into the resistant scion and reached high titer in 7 to 21 d. The accumulation of virus in the somaclonal material could have been due to the selection of a virus strain which was able to overcome the somaclonal resistance. In each grafted plant, the delay seen in virus multiplication in the resistant F scion was significant when compared to the control graft of susceptible ( plants. In the reciprocal graft of GCRI-26 (scion) and somaclone line 247 (rootstock), the inoculated resistant rootstock did not show the presence of virus (except for the virus in the inoculated leaf, with ELISA values from 0.00 to 0.14 AU) (Figure 22 C). The virus did not appear in the susceptible scion for more than 21 days pi, and when it did appear (with ELISA values of over 1.00 AU), the rootstock became infected as well. This later appearance of virus could be due to selection of a more virulent virus strain or by unintentional inoculation of the susceptible scion during sampling. In one grafted plant, the virus did not multiply in the susceptible scion for more than 40 d, which indicated that the resistant rootstock was restricting the movement of the small amount of virus that was present in the rootstock. If infective virus had moved into the susceptible scion, there would have been a high virus titer in five d. Because of the long delay in accumulation of virus in the scion, it was concluded that the virus was not allowed to spread systemically to the upper part from the resistant rootstock. Perhaps the active step of virus movement into the phloem was stopped, or there was simply not enough virus to allow for movement. Other explanations are possible as well. Resistance due to a systemically active molecule could not be eliminated. 151 Somaclone l2 MEAN ELISA VALUES (AU) _,§aigs_ Rootstock 0 d 5 d 7 d 11 d 21 d 39 d A i2 -* 0.01 0.45 0 44 0 24 1 00 26 - - 1 70 1.76 l 03 0 93 B. _e 0 00 0.00 0 00 0 05 - - 12 - 0.68 - 0 O4 - - *not determined Figure 23. Results of grafting experiments using resistant (somaclonal line 12) and susceptible (GCRI-26) plant material after inoculation with ToMV-O and screening young leaves from rootstock and scion using ELISA. Illustrations represent the grafts at 10 d. 152 The reciprocal grafts between GCRI-26 and somaclonal line 12 were similar to the above grafts in their response (Figure 23), but the resistance did not seem as strong in general, which was consistant with the results from the study of virus movement in plants of line 12. When GCRI-26 was the rootstock, there were high levels of virus present in the lower leaves, but only intermediate levels of virus accumulated in the leaves of the resistant scion (ELISA values from 0.45 to 0.24 AU) for longer than 20 d (Figure 23A). This indicated that there was an active limitation of virus multiplication in the resistant scion, but not complete inhibition. When somaclonal line 12 served as the rootstock, a fairly high level of virus was detected in the inoculated leaf at 5 d (0.68 AU), but did not multiply in the susceptible scion for more than 10 d (Figure 23B). This delay was significant when compared to the control susceptible grafts, and gave support to the hypothesis that the somaclonal resistance includes restriction of systemic movement in the plant. Another explanation could be that the virus only moves out of the leaf when it has reached a high enough titer, but there was no evidence to discern between these two causes of limited virus spread. PROTOPLASTS Isslstien Rf RIQLQElééEfi frgm ingguleted plenpe. Infected protoplasts seemed very fragile, and protoplast yields from plants with high virus titer (ELISA values of more than 1.00) were lower than yields normally obtained from healthy plants. Protoplasts isolated from infected GCRI- 26 (ELISA value of 1.85 AU) were smaller than protoplasts from healthy plants, and were associated with more cellular debris (Figure 24A). After staining with FITC-anti-coat protein antibody conjugate, 51.4% of the protoplasts prepared from infected GCRI-26 were highly fluorescent, 153 Figure 24. A) Infected protoplasts isolated from GCRI-26. B) One infected protoplast was present among thousands of healthy protoplasts in plant 12-1. C) Protoplasts from resistant plant 247-4 show low levels of virus in most protoplasts. (x600) 154 where the entire protoplast fluoresced with bright yellow-green, indicating the presence of virus. Uninfected protoplasts appeared darker. Cells which had been in contact with Lymphoprep during the isolation appeared red. Presumably less chlorophyll was extracted during fixation because the Ficoll left on the protoplast interfered with the extraction. Protoplasts which were isolated from inoculated somaclones were variable in yield and in proportion of infected to healthy protoplasts. The mean ELISA value of the plants (averaged from leaves throughout the plant) did not correlate very well with the appearance of infected protoplasts after protoplast isolation, even though samples for each were collected the same day (Table 34). The breaking and loss of infected protoplasts during the purification may account for this dicrepancy. For several of the protoplast preparations, the sorbitol- CPW buffers from rinsing the protoplasts were saved for testing by ELISA. These ELISA values give an estimation of the amount of virus released during the protoplast isolation procedure. As shown in Table 34, one plant (12-1) which had an ELISA value of 0.79 AU, yielded only a few infected protoplasts out of thousands of healthy protoplasts (Figure 24B). This was repeated with several plants. Another suprising exception was with protoplasts isolated from plants with very low ELISA values (0.05 AU) that appeared to contain virus in many of the protoplasts at a low level (Figure 24C). In this example, protoplasts were isolated from a plant (247-4) which was symptomless and had an ELISA value of 0.05 AU at 15 d pi. Of these protoplasts, 65.6% were labelled with the FITC-antibody. However, the fluorescent labelling was not seen completely throughout each 155 Table 34. Correlation of ELISA values to protoplast yield and to percentage of infected protoplasts isolated from susceptible and resistant tomato lines. Also included are ELISA values of protoplast samples and samples of CPW buffer used in rinsing the protoplasts during preparation. mean protoplast mean ELISA plant yield/gm protoplast % infected of CPW plant ELISA (x 10 ) ELISA protoplasts buffer GCRI-26 Healthy 26-17 0.00 4.53 - - - 26-18 0.00 5.29 - - - 26-19 0.00 8.76 - - - 26-21 0.00 §,§4 - - - 6.27 (mean) Infected 26-22 1.58 3.55 0.96 - - 26.248 1.83 2.96 1.61 51.4 - 26-26 1.14 9‘11 0.96 56.9 1.53 2.43 (mean) Inoculated somaclone lines 247-4° 0.06 3.17 0.00 65.6* - 247-8 0.13 3.26 0.00 <0.01 - 247-5 0.04 2.16 0.01 <0.01 0.00 247-6 0.05 1.71 - 29.1* 0.10 12-1b 0.13 4.62 0.02 <0.01 - 12.1 0.15 3.15 0.00 <0.01* 0.02 215-1 0.16 5.10 0.00 <0.01 - *indicates plants from which partially infected protoplasts were detected. asee Figure 16A. bsee Figure 16B. csee Figure 16C. 156 protoplast, unlike protoplasts from infected GCRI-26 plants. Instead, the virus seemed to be localized within compartments in the cytoplasm. The same phenomenon was found in several plants sampled. The presence of virus coat protein has been detected in inclusion bodies in the cytoplasm (Hills et a1., 1987) and has also been shown to accumulate in thylakoid membranes (Reinero and Beachy, 1986). It would be interesting to determine the location of the viral coat protein, especially in consideration of the cytoplasmic inheritance displayed by the somaclones. The results seen in Figure 240 were puzzling, since it was felt that the ELISA was a very sensitive measure of virus (1 ng/ml was the lower limit of detection). Because the ELISA values did not correlate with the number of protoplasts which were stained for virus, an explanation was sought. Perhaps the immunofluorescence microscopy was more sensitive than the ELISA method for detecting low levels of virus. There has been a report of time-resolved fluoroimmunoassay, which uses double antibody sandwich technique with a fluorescent label, capable of detecting virus (PVX) as low as 100 pg/ml (Siitari and Kurppa, 1987). This was found to give a five- to 100-fold increase in sensitivity over conventional ELISA. With this precedent, it may be possible to explain the relationship between ELISA values and the degree of fluorescent labelling. In the case of a high plant ELISA value with no fluorescence- labelled protoplasts, most of the infected protoplasts could have been lost during protoplast purification, making it impossible to compare these measures of infection. Loss of virus during protoplast rinsing was not evident in the case of one plant examined (12-l) which had ELISA 157 values of 0.15 AU for the plant which was relatively low, and 0.02 AU for the CPW buffer, which was very low. More of these assays would be necessary for a conclusive statement to be made. That a low level (0.01-0.06 AU) of virus was present in protoplasts of the highly resistant somaclonal plant lines was somewhat suprising. However, this has also been noted in resistant potato plant lines after inoculation with PVX (Tavantzis, 1988). Further examination of this phenomenon using molecular biological techniques could verify virus presence. Detection of viral RNA by hybridization to labelled cDNA probes would give more sensitivity. One of several types of in situ hybridization could be used to localize the virus at the ultrastructural level, possibly giving more insight into the mechanism of resistance seen in the tomato somaclonal lines. It would be important to determine whether the ToMV coat protein is associated with the chloroplast membranes versus inclusion bodies. SUMMARY AND GENERAL CONCLUSIONS In characterizing the resistance to ToMV in tomato somaclones, it was shown that symptom formation was suppressed and virus multiplication was limited within the plants. After genetic analysis, this resistance appeared to be encoded by an incompletely dominant nuclear gene which interacted with the cytoplasm to cause a maternal effect. The resistance of the six somaclonal lines seemed to have a similar pattern of inheritance, with the possible exception of somaclonal line 12. Two somaclonal lines, 12 and 247, used in test crosses, were each found to be additive with the ToMV resistance gene, Tm-l but not to Tm-Z. Test crosses with the other known resistance gene, Tm-ZZ, produced resistant F1 progeny, though is have been due to the dominance of Tm-ZZ. The ToMV-resistant tomato somaclonal lines in this study did not respond differently from one another after inoculation with different virus strains and other viruses. Their resistance was limited to the common strains of tobamoviruses. The somaclonal lines were also similar in response to high temperature incubation after inoculation, remaining resistant at 300 C for more than 30 d. The somaclonal resistance was similar to the resistance given by the Tm-l gene; each line limited viral multiplication to varying degrees and suppressed symptoms, but was overcome by the tomato mosaic virus strain, TMV-l, and was not affected by high temperature incubation. Of the six somaclonal lines, the degree of resistance found in somaclonal line 247 was the highest and most consistent. For compar- 158 159 ison, somaclonal lines 12 and 247 were chosen for studies of virus movement. In these studies, after inoculation with ToMV, very low virus titers were present in inoculated leaves of somaclonal line 247, but the virus did not move into the plant systemically. This restriction of virus movement could have been due to deficient levels of virus, or to a deficiency in the transport function. However, somaclone 12 allowed a medium virus titer to accumulate throughout the plant, but did not allow a high virus titer. In contrast, susceptible control tomatoes allowed the virus to spread throughout the plant in high titer in 5 d. In grafts of resistant and susceptible tomato, inoculated susceptible rootstock allowed multiplication of virus, but the upper, resistant scion delayed multiplication of virus. Somaclonal line 247 completely delayed multiplication for more than 10 d, but somaclonal line 12 allowed medium virus titer to accumulate, which was maintained for more than 20 d. In the reciprocal grafts, inoculated resistant rootstock restricted the virus to the inoculated leaf. In susceptible control grafts, virus levels were high throughout the scion and rootstock at 5 d. When plants infected with ToMV were used as a source of protoplasts, it was found that the protoplast yield was lowered when high levels of virus were present, probably due to protoplast breakage. In a few of the resistant plants from somaclonal lines, there was a low level of virus found in many protoplasts which was detected by fluoresence microscopy. The virus (coat protein) appeared to be limited to discrete areas in the cytoplasm, which might possibly be chloroplasts. One might speculate that this restriction of-virus or 160 virus component to low levels could be related to the mechanism of resistance, and might be due to the genetic element which displays the maternal effect in the inheritance of resistance. The resistant plants seemed to limit the multiplication of the virus in more than one way. After looking at protoplasts released from inoculated, resistant plants, lower levels of virus were present in each cell, indicating that there was an inhibition of replication at the cellular level. There was also a restriction of virus movement, especially evident in the grafting experiments. How these two effects are encoded and expressed to give resistance is still not clear. The resistance trait in each somaclonal line may be inherited by a single, incompletely dominant nuclear gene, Tm-3, with a maternally transmitted factor as well. The role of each of these components can only be ascertained by further virological and genetical research. Completion of characterization and comparison of each somaclonal line will also provide the information necessary to examine mechanisms of resistance to determine whether they differ among somaclonal lines. Further study of virus movement in resistant plants is essential. It would be useful to compare susceptible and resistant plants for the cell-to-cell spread of virus in inoculated leaves, and to compare the systemic spread though the plant with in situ virus detection methods. Other variations of grafting experiments would also provide more information on the mechanism of virus transport and its inhibition in the resistant plants. To determine the ability of the somaclonal lines to resist ToMV multiplication at the cellular level, further attempts at protoplast 161 infection should be made, preferably using electroporation, which can give up to 90% infection levels. The infection of cells from the resistant somaclonal lines should be contrasted to susceptible cells (+/+) and cells with the genotype Tm-l/Tm-l. First, the infectivity of the common strain of ToMV should be determined, and then the infectivity of TMV-l, the strain which overcomes the resistance of the Tm-l gene, should be determined. These results will help characterize the level of resistance and help to clarify the relationship of the somaclonal resistance, Tm-3 to the resistance of Tm-l. Since RNA-dependent RNA polymerase activity has been implicated in the resistance of the Tm-l gene, it would be useful to examine this enzyme activity in the somaclonal lines. Using somaclonal lines versus the control lines GCRI-26 (+/+) and GCRI-237 (Tm-l/Tm-l), the level of RNA-dependent RNA polymerase activity should be measured in unchallenged plants versus plants inoculated with ToMV-O or TMV-1. If differences are present which are similar to those reported for Tm-l, further studies of this enzyme activity would be useful in understanding the mechanism of resistance to ToMV replication measured at the cellular level. More information is needed about the life cycle of ToMV and about resistance to virus multiplication in plants to formulate resistance mechanisms. Current rapid gains in knowledge about the order and site of replicative events of ToMV, about the viral genes and gene products, and about the role of plant gene products in the multiplication and transport of ToMV will facilitate testing of hypothesized resistance mechanisms in the next decade. Perhaps the somaclonal lines will provide a source of variation of resistance for testing. 162 Possible resistance mechanisms might involve ribosomal components, RNA-dependent RNA polymerase activities, components of the membranes (endoplasmic reticulum, mitochondrial or chloroplast) which bind the active RNA polymerase, components of the viral inclusion bodies which seem to play a role in replication and assembly of the virus particle, chloroplast thylakoid membranes which accumulate viral coat protein, or plants cell products which are needed to facilitate transport of the infective virus through the plant. These are only a few of the possible areas in which resistance to virus multiplication could occur, as either positive or negative resistance mechanisms. Another area of research which deserves further study is the mechanism of mutation in somaclonal variation. Among possible mechanisms to examine are DNA amplification or deletion, somatic rearrangement of genes (particularly those in multigene families), unequal somatic crossing over, altered nucleotide methylation patterns, perturbation of DNA replication by altered nucleotide pools, silencing or activation of genes by mutations in associated non-coding regions and transposable elements (Scowcroft, 1985). The increasing availability and use of molecular probes will provide some of the experimental tools necessary to understand the contribution of such mechanisms to somaclonal variation. It is of interest that the viral resistances generated through somaclonal variation resemble that of a viral resistance gene found in a wild tomato species, both genetically and in their general response to virus challenge. It is worthy of note as an example of genetic flux, and when better characterized, may give insight into the evolution of disease resistance traits. LIST OF REFERENCES LIST OF REFERENCES Alexander, L.J. 1963. Transfer of a dominant type of resistance to the 4 known Ohio pathogenic strains of tobacco mosaic virus (TMV) from Lycopersicon peruvianum to L. esculentum. Phytopathology 53:869. Alexander, L.J. 1971. Host-pathogen dynamics of tobacco mosaic virus in tomato. Phytopathology 61:611-626. Atabekov, J.G. and Y.L. Dorokov. 1984. Plant virus-specific transport function and resistance of plants to viruses. Adv. in Virus Research 29:313-364. Barden, K.A. 1985. Variation among tomato seedlings and somaclonal response to inoculation with tomato mosaic virus. M.S. Thesis, Michigan State University. Barden, K.A., S. Schiller Smith and H.H. Murakishi. 1986. Regeneration and screening of tomato somaclones for resistance to tobacco mosaic virus. Plant Sci. Lett. 45:209-213. Brettell, R.I.S. and D.S. Ingram. 1979. Tissue culture in the production of novel disease-resistant crop plants. Biol. Rev. 55:329- 345. Brewbaker, J.L. 1964. In Agricultural Genetics, p 105, Prentice-Hall, Edgewood Cliffs, N.J. Broadbent, L. 1964. The epidemiology of tomato mosaic VII. The effect of TMV on tomato fruit yield and quality under glass. Ann. Appl. Biol. 54:209-224. Buiatti, M., G. Marcheschi, F. Tognoni, M. Lipucci di Paola, F. Collina Grenci and G. Martini. 1985. Genetic variability induced by tissue culture in the tomato (Lycopersicon esculentum). Z. Pflanzenzuchtg. 94:162-165. Cirulli, M. and L.J. Alexander. 1969. Influence of temperature and strain of tobacco mosaic virus on resistant tomato breeding line derived from Lycopersicon peruvianum. Phytopathology 59:1287-1297. Clark, M.F. and W. Adams. 1977. Characteristics of the microplate method of enzyme-linked immunosorbant assay. J. Gen. Virol. 34:475-483. Clayberg, C.D., L. Butler, C.M. Rick and P.A. Young. 1960. Second list of known genes in the tomato. J. Heredity 51:167-174. 163 164 CMI/AAB Description of Plant Viruses. #156 Tomato Mosaic Virus, Sept. 1976. Commonwealth Mycological Institute, Kew, Surrey, U.K. CMI/AAB Description of Plant Viruses. #177 Tobamovirus Group, Sept. 1977. Commonwealth Mycological Institute, Kew, Surrey, U.K. CMI/AAB Description of Plant Viruses. #213 Cucumber Mosaic Virus, July 1979. Commonwealth Mycological Institute, Kew, Surrey, U.K. CMI/AAB Description of Plant Viruses. #258 Tobacco Etch Virus, 1980. Commonwealth Mycological Institute, Kew, Surrey, U.K. Cowley, D.E., D. Pomp, W.R. Atchley, E.J. Eisen and D. Hawkins-Brown. 1989. The impact of maternal uterine genotype on postnatal growth and adult body size in mice. Genetics 122:193-203. D'Amato, F. 1978. Chromosome number variation in cultured cells and regenerated plants. In Frontiers of Plant Tissue Culture, pp. 287-295. (edited by T.A. Thorpe), IAPTC/University of Calgary, Calgary. Daub, M.E. 1986. Tissue culture and the selection of resistance to pathogens. Ann. Rev. Phytopath. 24:159-186. Deom, C.M., M.J. Oliver and R.N. Beachy. 1987. The 30-kilodalton gene product of tobacco mosaic virus potentiates virus movement. Science 237:389-394. Edallo, S., C. Zuccinalli, M. Perenzin and F. Salamini. 1981. Chromosomal variation and frequency of spontaneous mutation associated with in vitro culture and plant regeneration in maize. Maydica 26:29- 56. Elseth, G.D. and R.D. Baumgardener. 1984. Extensions of gene transmission analysis. In Genetics. pp. 313-325. Addison-Wesley Publishing Company, Reading, MA. Evans, D.A. and W.R. Sharp. 1983. Single gene mutations in tomato plants regenerated from tissue culture. Science 221:949-951. Evans, D.A. and W.R. Sharp. 1986. Applications of somaclonal variation. Bio/Technology 4:528-532. Evans, D.A., W.R. Sharp and H.P. Medina-Filho. 1984. Somaclonal and gametoclonal variation. Am. J. Bot. 71:759-774. Evans, D.M.A., J.A. Bryant and R.S.S. Fraser. 1984. Characterization of RNA-polymerases in healthy and TMV-infected tomato plants. Ann. of Botany 54:271-281. Evans, D.M.A., R.S.S. Fraser and J.A. Bryant. 1985. RNA-dependent RNA polymerase activities in tomato plants susceptible or resistant to TMV. Ann. of Botany 55:587-591. 165 Fraser, R.S.S. 1985. Genetics of host resistance to viruses and of virulence. In Mechanisms of Resistance to Plant Diseases. pp. 62-79, p. 180. (Edited by R.S.S. Fraser) Martinus Nijhoff/Dr W. Junk Publishers, Dordrect. Fraser, R.S.S. and S.A.R. Loughlin. 1980. Resistance to TMV in tomato: effects of the Tm-l gene on virus multiplication. J. Gen. Virol. 48:87- 96. Fulbright, D.W., W.H. Weidlich, K.Z. Haufler, C.S. Thomas and C.P. Paul. 1983. Chestnut blight and recovering American chestnut trees in Michigan. Can. J. Bot. 61:3164-3171. Gengenbach, B.G., J.A. Connelly, D.R. Pring and M.F. Conde. 1977. Mitochondrial DNA variation in maize plants regenerated from cell cultures. Theor. Appl. Genet. 59:161-167. Gibbs, A. and B. Harrison. 1976. Plant Virology: The Principles. pp. 183-184, p. 225. Halsted Press, New York. Goelet, P., C.P. Lomonossoff, P.J.G. Butler, M.F. Akam, M.J. Gait and J. Karn. 1982. Nucleotide sequence of tobacco mosaic virus RNA. Proc. Natl. Acad. Sci. USA. 79:5818-5822. Gooding, G.V. and T.T. Herbert. 1967. A simple technique for purification of tobacco mosaic virus in large quantities. Phytopathology 57:1285. Hall, T.J. 1980. Resistance at the Tm-Z locus in the tomato to ToMV. Euphytica 29:189-197. Hause, B., F. Baldauf, K. Stock, C. Wasternack and M. Metzlaff. 1986. Molecular analysis of mitochondrial DNA fron tomato cell suspensoin cultures. Curr. Genet. 10:785-790. Heath-Pagliuso, S., J. Pullman and L. Rappaport. 1988. Somaclonal variation in celeryzscreening for resistance to Fusarium oxysporum F.5p. apii. Theor. Appl. Genet. 75:446-451. Hills, G.J., K.A. Plaskitt, N.D. Young, D.D. Dunigan, J.W. Watts, T.M.A. Wilson and M. Zaitlin. 1987. Immunogold localization of the intracellular sites of structural and non-structural tobacco mosaic virus proteins. Virology 161:488-496. Holmes, F.O. 1954. Inheritance of resistance to infection by TMV in tomato. Phytopathology 44:640-642. Ikegami, M. and H. Fraenkel-Conrat. 1979. Characterization of double- stranded ribonucleic acid in tobacco leaves. Proc. Natl. Acad. Sci. USA. 76:3637-3640. Kassanis, B. and R.F. White. 1914. A.simplified method of obtaining tobacco protoplasts for infection with TMV. J. Gen. Virol. 24:447-452. 166 Kemble, R.J. and J.F. Shepard. 1984. Cytoplasmic DNA variation in a potato protoclonal population. Theor. Appl. Genet. 69:211-216. Kirk, J.T.O. and R.A.E. Tilney-Bassett. 1978. The Plastids. pp. 405- 413. Elsevier/North Holland Biomedical Press, Amsterdam. Krikorian, A.D., S.A. O'Conner and M.S. Fritter. 1983. Chromosome number variation and karyotype stability in cultures and culture-derived plants. pp 541-581. In Handbook of Plant Cell Culture. Vol. 1: Techniques for Propagation and Breeding. (Edited by D.A. Evans, W.R. Sharp, P.V. Ammirato and Y. Yamada.) MacMillan Publishing Co., New York. Rut, S.A. and C.E. Flick. 1986. A miniprep system for analysis of chNA restriction digest fragments in Nicotiana. Plant Mol. Biol. Reporter 4:48-55. Larkin, P.J. and W.R. Scowcroft. 1981. Somaclonal variation- a new source of variability from cell cultures for plant improvement. Theor. Appl. Genet. 60:197-214. Leonard, D.A. and M. Zaitlin. 1982. A temperature-sensitive strain of TMV defective in cell-to-cell movement generates an altered viral-coded protein. Virology 117:416-424. Levine, L. 1973. Extrachromosomal inheritance. In Biology of the Gene. pp. 180-181. G.V. Mosby Company, St. Louis. Maniatis, T., E.F. Fritsch and J. Sambrook. 1982. Molecular cloning (A laboratory manual). Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. May, J.T., J.M. Gilliland and R.H. Symons. 1970. Properties of a plant virus-induced RNA polymerase in particulate fractions of cucumbers infected with cucumber mosaic virus. Virology 41:653-664. McClean, P.E. and M.R. Hanson. 1986. MtDNA sequence divergence among Lycopersicon and related Solanum species. Genetics 112:619-667. McCoy, T., R.L. Phillips and H.W. Rines. 1982. Cytogenetic analysis of plants regenerated from oat (Avena sativa) tissue culture: high frequency of partial chromosome loss. Can. J. Genet. Cytology 24:37-50. McGuigan, J.E. and H.N. Eisen. 1968. Differences in spectral properties and tryptophan content among rabbit anti-2,4-dinitrophenyl antibodies of the gamma-G-immunoglobulin class. Biochemistry 7:1919- 1928. McWhorter, F.P. 1965. Plant virus inclusions. Ann. Rev. Phytopath. 3:287. Meins, F., Jr. 1983. Heritable variation in plant cell culture. Ann. Rev. Plant Physiol. 34:327-346. 167 Meins, F., Jr. and J. Lutz. 1980. The induction of cytokinin habituation in primary pith explants of tobacco. Planta 149:402-407. Meredith, C.P. 1978. Selection and characterization of aluminum- resistant variants from tomato cell cultures. Plant Sci. Lett. 12:25- 34. Motoyoshi, F. and N. Oshima. 1975. Infection with TMV of leaf mesophyll protoplasts from susceptible and resistant lines of tomatoes. J. Gen. Virol. 29:81-91. Motoyoshi, F. and N. Oshima. 1977. Expression of genetically controlled resistance to TMV infection in isolated tomato leaf mesophyll protoplasts. J. Gen. Virol. 34:499-506. Murakishi, H.H. 1969. Infection of tomato callus cells with TMV-RNA. Phytopathology 58:993-996. Murakishi, H.H. and P.S. Carlson. 1982. In vitro selection of Nicotiana sylvestris variants with limited resistance to TMV. Plant Cell Reports 1:94-97. Murakishi, H., M. Lesney and P. Carlson. 1984. Protoplasts and plant viruses. Adv. in Cell Culture 3:1-55. Murashige, T. and F. Skoog. 1962. A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant. 15:473-497. Mutschler, M.A., S.D. Tanksley and C.M. Rick. 1987 linkage maps of the tomato (Lycopersicon esculentum). Tomato Genetics Coop. Rep. 37:5-34. Nagy, F., G. Lazar, L. Menczel and P. Maliga. 1983. A heteroplasmic state induced by protoplast fusion is a necessary condition for detecting rearrangements in Nicotiana mtDNA. Theor. Appl. Genet. 66:203-207. Nelson, R.S., S.M. McCormick, X. DeLannay, P. Dube, J. Layton, E.J. Horsch, S.G. Rogers, R.T. Fraley and R.N. Beachy. 1988. Virus tolerance, plant growth, and field performance of transgenic tomato plants expressing coat protein from tobacco mosaic virus. Bio/Technology 6:403-409. O'Connell, M.A., L.P. Hosticka and M.R. Hanson. 1986. Examination of genome stability in cultured Lycopersicon. Plant Cell Reports 5:276- 279. Ohno, T., N. Takamatsu and T. Meshi. 1983. Single amino acid substitution in 30 k protein of TMV defective in virus transport function. Virology 131:255-258. 0rton,T.J. 1983. Spontaneous electrophoretic and chromosomal variability in callus cultures and regenerated plants of celery. Theor. Appl. Genet. 67:17-24. 168 Otsuki, Y. and I. Takebe. 1969. Fluorescent antibody staining of tobacco mosaic virus antigen in tobacco mesophyll protoplasts. Virology 38:497-499. Otsuki, Y. and I. Takebe. 1978. Production of mixedly coated particles in tobacco mesophyll protoplasts doubly infected by strains of tobacco mosaic virus. Virology 84:162-171. Palmer, J.D. and D. Zamir. 1982. Chloroplast DNA evolution and phylogenetic relationships in Lycopersicon. Proc. Natl. Acad. Sci. USA. 79:5006-5010. Palukaitis, P. 1987. Potato spindle tuber viroid: investigation of the long-distance, intra-plant transport route. Virology 158:239-241. Pelham, J. 1972. Strain-type interaction of TMV in tomato. Ann. Appl. Biol. 71:219-228. Pelham, J., J.T. Fletcher and J.H. Hawkins. 1970. The establishment of a new strain of tobacco mosaic virus resulting from the use of resistant varieties of tomato. Ann. Appl. Biol. 65:293-297. Phillip, M.J., S. Honma and H.H. Murakishi. 1965. Studies on the inheritance of resistance to tobacco mosaic virus in the tomato. Euphytica 14:231-236. Piechulla, B., K.R. Chonoles Imlay and W. Gruissem. 1985. Plastid gene expression during fruit ripening in tomato. Plant Mol. Biol. 5:373-384. Pilowsky, M. 1971. Grafting studies with Im;za stock. Tomato Genetics Coop. Rep. 21:36. Prat, D. 1983. Genetic variability induced in Nicotiana sylvestris by protoplast culture. Theor. Appl. Genet. 64:223-230. Ralph, R.K. and S.J. Wojcik. 1969. Double-stranded TMV-RNA. Virology 37:276-282. Ramulu, R.S., P. Dijkhaus and S. Roest. 1984. Genetic instability in protoclones of potato (Solanum tuberosum L. cv. ‘Bintje'): new types of variation after vegetative propagation. Theor. Appl. Genet. 68:515-519. Rast, A.T. 1975. Variability of tobacco mosaic virus in relation to control of tomato mosaic in glasshouse tomato crops by resistance breeding and cross protection. Verslagen Van Landbouradkundige Des Rizksland Agricultural Research Reports 834:1- 76. C.A.P.D., Wageningnen. Reanney, D.C. 1982. The evolution of RNA viruses. Ann. Rev. Microbiol. 36:47-73. Register, J.C. III and R.N. Beachy. 1988. Towards an understanding of the mechanism of protection against TMV infection in transgenic tobacco plants. J. Cell Biol. 12C:288. 169 Reinero, A. and R.N. Beachy. 1986. Association of TMV coat protein with chloroplast membranes in virus-infected leaves. Plant M01. Biol. 6:291-301. Rick, C.M. 1974. The Tomato. In Handbook of Genetics, Vol.2. Plants, Plant Viruses, and Protists. pp. 247-280. (Edited by R.C. Rock.) Plenum Press, New York. Romaine, C.P. and M. Zaitlin. 1978. RNA-dependent RNA polymerases in uninfected and TMV-infected tobacco leaves: viral induced stimulation of a host polymerase activity. Virology 86:241-253. Russell, G.E. 1978. Plant Breeding for Pest and Disease Resistance. p. 257. Buttersworth, London. Samuel, G. 1934. The movement of tobacco mosaic virus within the plant. Ann. Appl. Biol. 21:90-111. Saito, T., D. Hosokawa, T. Meshi and Y. Okada. 1987. Immunocytochemical localization of the 130K and 180K proteins (putative replicase components) of tobacco mosaic virus. Virology 160:477-481. Schiller, B., R.C. Herrman and G. Melchers. 1982. Restriction endonuclease analysis of plastid DNA from tomato, potato and some of their somatic hybrids. Molec. Gen. Genet. 186:453-459. Scowcroft, W.R. 1985. Somaclonal variation: The myth of clonal uniformity. In Genetic Flux in Plants. pp. 217-245. (Edited by Bu Hohn and E.S. Dennis), Springer-Verlag Wein, New York. Selman, 1.11. and A.C.I. Yahampath. 1973. Some physiological characteristics of two tomato cultivars, one tolerant and one susceptible to tobacco mosaic virus. Ann. of Botany 37:853-865. Shahin, E.A. and R. Spivey. 1986. A single dominant gene for Fusarium wilt resistance in protoplast derived tomato plants. Theor. Appl. Genet. 73:164-169. Sibi, M. 1986. Non-Mendelian heredity- Genetic analysis of variant plants regenerated from in vitro culture: epigenetics and epigenic. In Somaclonal Variations and Crop Improvement. (Edited by J. Semal.) Martinus Nijhoff Publishers, Dordrecht. Sibi, M., M. Biglary, and Y. Demarly. 1984. Increase in the rate of recombinants in tomato (Lycopersicon esculentum L.) after in vitro regeneration. Theor. Appl. Genet. 68:317-321. Siitari, H. and A. Kurppa. 1987. Time-resolved fluoroimmunoassay in the detection of plant viruses. J. Gen. Virol. 68:1423-1428. Smith, C.M., D. Pratt and G.A. Thompson. 1986. Increased 5- enolpyruvylshikimic acid 3-phosphate synthase activity in a glyphosate- tolerant variant strain of tomato cells. Plant Cell Reports 5:298-301. 170 Spendlove, R.S. 1967. Microscopic Techniques. In Methods in Virology, Vol. III. pp. 475-520. (Edited by K. Maramorosch and H. Koprowski), Academic Press, New York. Stobbs, L.W. and B.H. MacNeill. 1980. Increase of tobacco mosaic virus in graft-inoculated TMV-resistant tomatoes. Can. J. of Plant Pathol. 2:217-221. Takebe, I. 1981. Inoculation of protoplasts with plant viruses. in Cell Culture and Somatic Cell. pp. 492-502. Academic Press, New York. Takebe, I. 1983. Protoplasts in plant virus research. Intl. Rev. of Cytology 16:89-111. Taylor, R.H., R.G. Grogan and K.A. Kimble. 1961. Transmission of tobacco mosaic virus in tomato seed. Phytopathology 51:837-842. Thomas, P. and R.W. Fulton. 1968. Correlation of the ectodesmata number with nonspecific resistance to initial virus infection. Virology 34:459-469. Thomas, B.R. and D. Pratt. 1982. Isolation of paraquat-tolerant mutants from tomato cell cultures. Theor. Appl. Genet. 63:169-176. Thorpe, T.A. 1981. Plant Tissue Culture. pp. 21-22. Academic Press, New York. Tomenius, K., D. Clapham and T. Meshi. 1987. Localization by immunogold cytochemistry of the virus-coded 30K protein in plasmodesmata of leaves infected with tobacco mosaic virus. Virology 160:363-371. Toyoda, H., M. Yamamoto and T. Harai. 1983. Resistance mechanism of cultured plant cells to TMV. (II)Resistance of calli, protoplasts, cotyledons, leaf discs and intact leaves of tomato plants. Ann. Phytopath. Soc. Japan. 49:639-646. Wakarchuk, D.A. and R.I. Hamilton. 1985. Cellular double-stranded RNA in Phaseolus vulgaris. Plant Mol. Biol. 5:55-63. Wilson, T.M.A. and P.A.C. Watkins. 1986. Influence of exogenous viral coat protein on the cotranslational disassembly of tobacco mosaic virus (TMV) particles in vitro. Virology 149:132-135. Wisniewski, L.A. and R.N. Beachy. 1988. The use of grafting to study a mechanism of cross protection in transgenic plants containing the TMV coat protein gene. J. Cell Biol. 12:275. Zaitlin, M. and H.W. Israel. 1975. Tobacco mosaic virus (type strain). C.M.I./A.A.B. Description of Plant Viruses, No. 151. Zitter, T.A. and H.H. Murakishi. 1969. Nature of increased virulence in TMV after passage in resistant tomato plants. Phytopathology 59:1736-1739. APPENDICES APPENDIX A APPENDIX A Regeneration and screening of tomato somaclones for resistance to tobacco mosaic virus 171 Plant Science. 45 (1986) 209—213 209 Elsevier Scientific Publishers Ireland Ltd. REGENERATION AND SCREENING OF TOMATO SOMACLONES FOR RESISTANCE. TO TOBACCO MOSAIC VIRUS ILA. BARDEN. S. SCHILLER SMITH and H.H. MURAKISI-II‘ Department of Botany and Plant Pathology. Michigan State University. East Lansing. MI 48824-1312 (C.S.A. ) (Received February 21st. 1986) (Revision received April 23rd, 1986) (Accepted May 7th, 1986) Leaf discs of tomato (Lycopersicon esculentum. Mill.) from a fully tobacco mosaic virus (TMV Husceptible (+/+) isogenic line, GCRI-26 (Glasshouse Crops Research Institute, Littlehampten, D.R.), were used to regener- ate somaclones (Sc). Out of 370 somaclones inoculated with TMV-Flavum. six were eventually selected as virus-free and putatively resistant. R, progeny from self -pollination of the six somaclones showed TMV-resistance for varying periods ranging from 28 to 55 days (extent of test) while seedlings of the source plant usually became infected in 10—20 days. In resistant plants. no symptom were v'uible and TMV could not be detected either by enzyme- linked immunosorbant assay (ELISA) or by back inoculations to Nicotiana glutinous. Key words: tomato; Lycopersicon esculentum; somaclonal variation; mosaic virus resistance Introduction Somaclonal variation is increasingly being investigated as an alternative source of genetic variation for crop improvement [1] and for disease resistance [2]. Tissue culture-derived plants have provided novel sources of resist- ance to Fiji virus disease of sugar cane [3], early [4] and late [5] blight of potato, southern leaf blight of maize [6], tobacco mosaic virus of tobacco [7], Verticillium wilt of lucerne [8], toxins of Pseudomonas syringes and Alternaria alternate in tobacco [9] and to the toxin, victorin, in oats [10]. In our preliminary work [11], we found 18 of 370 somaclones of a fully susceptible tomato, GCRI-26 [12] were free of virus 1 month after inoculation with TMV-Flavum. We present here results of further studies on 'To whom correspondence should be sent. Abbreviations: BA, 6-benzyladenine: ELISA, enzyme- linked immunosorbant assay; IAA. indoleacetic acid; MS, Murashige-Sltoog; Sc. somaclone; TMV, tobacco mosaic virus: Tween 20, polyethylene sorbitan monolaurate. the resistance of the somaclones which were derived from GCRI-26. Materials and methods Regeneration of tomato plants Tomato (L. esculentum. Mill.) seed of a fully susceptible (+/+) isogenic line, GCRI- 26, based on the variety Craigella, were kindly furnished by Dr. T.J. Hall of the Glass- house Crops Research Institute. Leaf explants were taken from over 50 different 4-week-old plants grown in a soilless peat mixture (Sun- shine mix 1, Fisons Western Corp. Vancouver, BC, Canada) at 25 : 1°C with a 16-h photo- period (cool white fluorescent tubes supply- ing 60 all m" s“). Fully expanded leaves were surface sterilized for 15 min in a solu- tion containing 10% (v/v) Big Chief Bleach (Patterson Laboratories, Detroit, MI) plus 0.1% polythylene sorbitan monolaurate (Tween 20) (Sigma Chemical, St. Louis, MO 63178) followed by three rinses in sterile water. Leaf discs, 6 mm in diameter, were cut out with the aid of a cork borer and 5—8 were placed on 25 ml of shoot- 0168-9452/86/80350 0 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland 172 210 inducing medium which consisted of Mura- shige-Skoog (MS) medium [13] supplemented with 0.3 mg/l indoleacetic acid (IAA) (Sigma) and 3.0 mg/l 6-benzyladenine (BA) (Sigma) and solidified with 0.7% (w/v) agar. The explants were cultured in 100 x 15 mm plastic petri plates and maintained at 23 : 1°C with a 16-h photoperiod (cool-white fluorescent tubes supplying 30 ME m" S"). Various concentrations of BA and IAA were tested in the shoot-inducing medium to determine the optimum combination for shoot induction. Young shoots formed after 4 weeks; these were subcultured in 150 X 25 mm culture tubes containing 10 ml of root-inducing medium consisting of MS medium supplemented with Gamborg [14] B5 vitamins. The pH of the medium was adjusted to 5.8 with dilute HCl or NaOH and then solidified with 0.7% agar. The root system was well developed within 2—3 weeks at which time the plants were removed from the tubes and the agar was gently washed off the roots. The rooted shoots were then transplanted in soilless peat mix- ture in 7 .5-cm styrofoam pots. Plants were placed in a plastic box and covered with a polyethylene film for 3-4 days. The film was removed and the plants placed at 23 t 1°C with a 16-h photoperiod (cool-white fluorescent tubes supplying 65 uE m" s") for an additional 4—5 days. Inoculation of Sc A yellowing strain of TMV-Flavum was kindly provided by Dr. H. Jockusch, Max Planck Institut, Tubingen, West Germany. The virus was increased in GCRI-26 tomato plants and the sap extracted by grinding leaves in a mortar in 0.01 M potassium phosphate buffer (pl-I 7 .0) (20 ml of buffer/g of leaves). Sc were inoculated by rubbing Carborundum-dusted leaves with a cotton swab previously dipped in infective sap. TMV-Flavum was used because infected plants could be readily distinguished by the yellow-mosaic symptoms. After inoculation, plants were transferred to a growth chamber 173 at 25 s 1°C with a 16-h photoperiod (cool- white fluorescent tubes supplying 65 uE m" 5"), Symptoms usually appeared in 7-10 days in susceptible control GCRI-26 seedlings. ELISA The double antibody sandwich microplate ELISA procedure of Clark and Adams [15] was used to determine the presence or ab- sence of virus. Antiserum to TMV-Flavum was produced in a rabbit by a series of three intramuscular injections of 1 mg each of purified virus. The initial injection was made with virus emulsified in Freund’s complete adjuvant and the second and third with Freund’s incomplete adjuvant (Difco Co., Detroit, MI 48201). The gamma globulin fraction was obtained by ammonium sulfate precipitation of the antiserum followed by passage through a 1 x 10 cm column of DE- 22 (Whatman Ltd.) and was conjugated with alkaline phosphatase (Type VII S, Sigma) [15]. Optimum dilution of gamma globan for coating plates was 1:1000 (v/v) and for the conjugate was 1:800 (v/v). Leaf tissue samples were ground with a mortar and pestle in PBS-Tween buffer containing 0.2 g KHIPO“ 2.3 g Na,l-IPO., 8.0 g NaCl, 0.2 g KCI and 0.5 ml Tween 20/1; 5 ml of buffer was used per g of tissue. All ELISA buffers contained 0.02% sodium aside as a preserv- ative. Absorbance values at 405 mm were recorded 30 or 60 min following addition of P-nitrophenyl phosphate, the substrate for alkaline phosphatase. Back inoculations were done with the same samples used in the ELISA by applying the extract with a cotton swab to Carbo- rundum-dusted leaves of N. glutinosa. The leaves were immediately rinsed with water after inoculation. Four half leaves were inoculated for each sample. Local lesions were counted 4-6 days later. Selection of virus-free somaclones and 9'08“? After the initial inoculation with TMV- Flavum and subsequent ELISA test survivors were transplanted to 10-cm pots of peat mixture and maintained in a greenhouse at 22-28°C with a 16-h photoperiod (cool- white fluorescent lighting supplying 60—65 11E m‘: 5“). After a second inoculation, survivors were transplanted to 20-cm pots and grown to the flowering stage. The plants were self-pollinated and seeds were collected from each plant in order to evaluate the resistance of the next generation, designated as R. [16]. R( plants were inoculated with TMV-Flavum and evaluated under the same conditions as described above. Results Plant regeneration The largest number of shoots in this test, 0.7/leaf disc, was obtained with 13.3 11M BA, and 1.7 11M IAA. A total of 370 Sc were regenerated and transplanted to soilless peat mixture. Virus screening of somaclones and progeny After the 370 Sc were inoculated with TMV-Flavum, 18 were judged to be free of virus based on ELISA. If ELISA values exceeded 0.50, plants were judged to be diseased and values for 57 symptomatic, infected plants ranged from 0.36- to 2.55. Values for 52 uninoculated healthy plants ranged from 0.00 to 0.03 (Table 1). Plants with values less than 0.5 were saved and Table I. Range and mean of ELISA values for healthy and TMV-Flavum infected GCRI-26 tomato plants. ELISA values Sample Range of mean : S.D. No.° ELISA values Healthy 0.005 t 0.01 52 0.00—0.03 GCRI-26 TMV-Flavum 1.27 t 0.77 57 036-255 infected GCRI-26 'Samples taken from groups of plants planted at 2-3-week intervals over a period of 8 months and assayed in duplicate. 211 Table II. Resistance of original somaclones (R) to TMV-Flavum. Back inoculations of extracts of somaclones to N. glutinosa were also negative. ELISA° Duration of values experiment Sc 12 0.00 51 days 0.00 Sc 215 0.03 161 days 0.04 Sc 219 0.03 173 days 0.02 Sc 247 0.03 101 days 0.04 Sc 322 0.04 137 days 0.04 Sc 330 0.04 95 days 0.04 'ELISA values for GCRI-26 plants inoculated with TMV-Flavum were >0.50 by 20 days: healthy control plants had values ranging from 0.00 to 0.07 . Readings are for duplicate wells. rescreened. After the second inoculation and screening, surviving plants were con- sidered resistant if ELISA values did not exceed two times the mean of the uninocul- ated control. Following reinoculation of the 18 survivors, six remained free of TMV- Flavum (Table II). R. progeny of Sc 12 and Sc 247 were inoculated with TMV- Flavum. The plants were tested by ELISA over a 28-55-day period. Results shown in Table III indicate that all of the 15 R) Table III. Resistance of R, progeny of somaclones to TMV-Flavum. ( ) = Number of plants in assay. ELISA values ELISA values at 30 days‘ at 55 days‘ Sc 12 0.00 s 0.00 (15) 0.03 : 001(15)" Sc 247 0.01 :003 (13) 0.05 $0.02 (4)b 0.56 (1) 0.74 t 0.16 (10) GCRI-26 1.15 z 0.45 (27) - Uninoculated 0.00 r 0.00 (2) 0.006 t 0.009 (6) GCRI-26 ‘Mean 2 S.D. bBaclt inoculations of extracts of resistant soma- clones at 55 days to N. glutinosa were also negative. 174 ”flu“ 27 H I, . I. I. I.)- " pl] 'lfl "l _, )- n. urea 41 use use use are ls-ssretssvsatterlesesleusa M Fig. 1. Resistance of R, progeny of somaclones to TMV-Flavum. The mean t S.D. of 99 ELISA values of putatively resistant somaclones wu 0.03 t 0.07 and the mean : S.D. of 16 ELISA values of control plants of inoculated GCRI-26 was 0.68 t 0.19. Back inoculations using extract from surviving somaclones at the end of tests, to N. glutinosa were also negative. progeny of Sc 12 were free of TMV-Flavum. Four of 14 R, progeny of Se 247 were free of TMV-Flavum. Back inoculations to N. glu tinosa were also negative. R, progeny of Sc 215, Sc 219, Sc 322, Sc 330 and additional Sc 12 were tested later with a second screening. Figure 1 shows that R, progeny of Sc 12 and those of Sc 215 were free of TMV-Flavum for the duration of the test but Sc 219, Sc 322 and Sc 330 showed some loss of resistance. Progeny of Se 219 appeared more resistant than that of Se 322 while Sc 330 was intermediate. The GCRI-26 control plants became infected at a faster rate and all had symptoms or high ELISA values by 30 days. Discussion The results indicate that the type of resistance shown by these somaclones is a delay in symptom expression and virus multiplication lasting for 41-55 days which was the extent of the tests. The resistant somaclones were obtained from a recognized fully 'susceptible source cultivar using ordi- nary tissue culture methods without the addition of a known mutagen. A sample of 57 plants from this source showed complete susceptibility, strongly suggesting that the resistant somaclones were displaying a new form of resistance. A similar type of TMV resistance has previously been described in tobacco somaclones following gamma radiation [7]. Different genes for TMV- resistance in tomato have been incorporated into the cultivar Craigella by Pelham [17]. These include the Tm-l gene which prevents symptom development and multiplication of the common strain of tomato mosaic virus [17,18]. Tm-2 and I‘m-2‘ genes elicit a hypersensitive response to the common strain of virus [17]. Tm-Z’ confers a high level of resistance to most strains of TMV [19]. The delay in symptom expression and inhibition of virus multiplication shown by our somaclones seem similar to that expressed by the Tm-l gene. However, further virological and genetic studies are needed to determine the exact nature of the resistance and its inheritance shown by the somaclones. Acknowledgments Journal Article No. 11909 of the Michigan Agricultural Experiment Station. This work was supported by U.S.D.A. grant No. 84-CRCR-1-1400. A graduate assistantship to K.A.B. was kindly provided by the Upjohn Co., Kalamazoo, MI. Thanks are expressed to Dr. J. Reynolds of the Upjohn Co. and Dr. R. Smith of International Plant Research Institute, San Carlos, CA 94070 for providing information on tomato shoot and root-inducing media, respectively. Appreciation is expressed to Drs. D. Rams- dell, J. Saunders and K. Sink for helpful suggestions. References 1 P.J. Larkin and W.R. Scowcroft, Theor. Appl. Genet.. 60 (1981) 197. 2 R.I.S. Brettel and D.S. Ingram, Biol. Rev., 54 (I979)329. 175 10 11 M. Krishnamurthi and J. Tlaskal, Proc. Int. Soc. Sugarcane Technol.. 15 (1974) 130. U. Matern. G. Strobe! and J. Shepard, Proc. Natl. Acad. Sci. U.S.A., 75 (1978) 4935. M. Behnke, Theor. Appl. Genet.. 56 (1980) 151. R.G. Gengenbach, C.E. Green and C.M. Donovan, Proc. Natl. Acad. Sci. U.S.A., 74 (1977) 5113. H.H. Murakishi and P.S. Carlson. Plant Cell Rep.. 1 (1982) 94. AD. Latunde-Duda and J.A. Lucas, Plant Sci. Lett., 32 (1983) 205. P. Thanutong, I. Furusawa and M. Yamamoto, Theor. Appl. Genet.. 66 (1983) 209. H.W. Rines and H.H. Luke, Theor. Appl. Genet.. 71 (1985) 16. ILA. Barden and H.H. Murakishi, Phytopatho- logy. 75 (1985) 1359 (Abstract No. 644). 176 13 14 15 16 17 18 19 213 R.S.S. Fraser, S.A.R. Loughlin and J.C. Conner, J. Gen. Virol., 50 (1980) 221. T. Murashige and F. Skoog. Physiol. Plant, 15 (1962) 473. CL Gamborg, R.A. Miller and K. Ojima, Exp. Cell Res.. 50 (1968) 151. ' M.F. Clark and A.N. Adams. J. Gen. Virol.. 34 (1977) 475. RS. Chaleff, Genetics of higher plants; appli- cations of cell culture, Cambridge Univ. Pres, Cambridge, 1981, p. 94. J. Pelham, Ann. Appl. Biol., 71 (1972) 219. R.S.S. Fraser and S.A.R. Loughlin, J. Gen. Virol., 48 (1980) 87. T.J. Hall, Euphytica, 29 (1960) 189. APPENDIX B APPENDIX B ELISA Virus Detection Procedures CONTENTS 1. ELISA Protocol 2. Chemicals 3. Coating and conjugate globulin preparation 177 178 1) ELISA Prggocol. (See Section 2 for composition of buffers) Reference: Clark, M. F. and A. N. Adams. 1977. Characteristics of the microplate method of enzyme-linked immunosorbant assay for the detection of plant viruses. J. Gen. Virol. 34:475-483. Note: All glassware is siliconized to prevent protein absorption. 1. Add 200 ul purified gamma-globulin which has been appropriately diluted in coating buffer per well of the microtiter plate. For each incubation step cover the plate with an airtight cover such as Saran Wrap. Incubate 4 hours at 37° C. Can be stored longer times at 5° or - 200 C. (Some coating globulins can be reused: check your system.) 2. Wash by flooding wells with PBS-Tween. Leave at least 3 minutes. Repeat wash three times. Empty plate. Take care to aspirate out bubbles in plate. - 3. Add 200 ul aliquots of the test samples appropriately diluted in virus grinding buffer to duplicate wells. Leave at 50 C overnight or at 370 C for 4-6 hours. Include a buffer check, a healthy check and a known virus check in each plate. 4. Wash plate three times as in step 2. 5. Add 200 ul aliquots of enzyme-labelled gamma-globulin (conjugate) of appropriate dilution to each well in PBS-Tween PVP egg albumin buffer. Incubate at 370 for 3-6 hours. (Some conjugates can be rescued: check your system.) 6. Wash plate three times as in step 2. 7. Add 300 ul aliquots of freshly prepared p-nitrophenol phosphate (1 mg/ml) in substrate buffer to each well. Reserve a beaker just for this purpose. Incubate at room temperature for 1 hour of less, as necessary to observe reaction. Save extra substrate for spectrophotometer reference - it should be water-clear. 8. Slow down reaction by adding 50 ul 3 M NaOH to each well. Read final results within 1 hour. Best at 0D (A405) near 1; color is not stable. 9. Assess results by a) visual observation, b) measurement of absorbance at 405 nm in a spectrophotometer or ELISA reader when quantitative data are desired. We consider the mean A405 of the healthy wells of the plates -/- 2 standard deviations to be the threshold value for positive virus identification in that plate. 2) 179 Chemicals. All glassware used with ELISA antibodies should be pretreated with a silicon coating solution to prevent loss by absorption. Coating buffer (0.5 M sodium carbonate, pH 9.6) Na2C03 1.59 g Ncho3 2.93 g NaN3 0.2 g (0.02% w/v) In 1 liter H20 adjusted to pH 9.6 with HCl. (Can be stored at 5° C, but check pH. Stock solution okay.) PBS-Tween washing solution (0.1 M sodium-potassium phosphate, 0.85% NaCl, 0.05% Tween-20, pH 7.4) 1 liter 3 liters NaCl 8.0 g 24.0 g KH2P04 0.2 g 0.6 g Na2HP04 1.65 g 3.45 g 7 H20 KCl 0.2 g 0.6 g NaN3 0.2 g 0.6 g (0.02% w/v) Tween-20 0.5 ml 1.5 ml In 1 liter H20. Should have pH - 7.4. (Can be stored at 5° C. Stock solution okay.) Virus grinding buffer (also used as conjugate buffer) PBS-Tween (see 3 above) 1 liter Polyvinylpyrrolidone (MW 10,000) 20 g (2% w/v) Ovalbumin (crude egg albumin) 2 g (0.2% w/v) Store at 5° C and use within one month. Healthy sap can be added to conjugate buffer to cross-absorb undesirable antibodies in conjugated gamma-globulin. Incubate 1 hour at 25° 0 before filling wells. Substrate buffer (10% diethanolamine, pH 9.8) diethanolamine 97 ml Make up to 1 liter with H20. Adjust to pH 9.8 with cone. HCl. Recheck pH before use and readjust to pH 9.8. Use liquid diethanolamine. 180 3) Coating gag conjugate globulin prgpgggglgg. 1. Purification of coating gamma-globulin. (Other recognised procedures may be used instead.) Choose an antiserum with a high gel double diffusion (Ouchterlony) titer against the strain of virus being studied and a low anti-healthy titer. l. 2. 10. To 1.0 ml antiserum add 9 ml distilled water. Slowly and dropwise add 10 m1 saturated ammonium sulphate solution while stirring. . Leave 30-60 min at room temperature. . Centrifuge at 6000 rpm for 5 min to collect precipitate. . Dissolve precipitate in 2 m1 1/2-strength PBS (1:1 PBS:H20 without Tween-20). . Dialyze 3 times against 500 ml 1/2-PBS (including once overnight). . Filter through 3-5 cm DE 23 cellulose (diethylaminoethyl cellulose) pre-equilibrated in 1/2-PBS. Use a short chromatographic column or a broken 10 m1 pipet plugged with glass wool. See DE 23 manufacturer's directions and follow them carefully. Bio-Rad Affigel Blue has also been suggested for trial as a more rapid one-step globulin purifi- cation method. . Wash gamma-globulin through DE 23 column with l/2-PBS. . Monitor effluent at 280 nm and collect first protein fraction to elute. Collect in 2 m1 fractions. Best should be about 3 0D. Measure A2 0 and adjust strength of gamma-globulin to read aproximater 1.4 0D (about 1 mg/ml). 2. Conjugation of enzyme with gamma-globulin l. 2. Centrifuge 1 m1 (-2 mg) Sigma No. P-4502 alkaline phosphatase at 6000 rpm for 5 min. Discard supernatent liquid. (This step is omitted when using Sigma No. P-5521.) Dissolve precipitate directly in 1 ml (-1 mg) purified gamma- globulin. Be careful of alkaline phosphatase pellet, which is soft. Add globulin to centrifuge tube. 181 (Continued) 3. Dialyze 3 times against 500 m1 PBS (1 X conc, no Tween-20) (twice 1 hour apart + overnight at 5° C). 4. Add fresh glutaraldehyde solution to make a final glutaraldehyde concentration of 0.05% (v/v). Mix well. 5. Leave 4 hours at room temperature. A yellow-brown color should develop. (Color is faint and ”sandy" but may stay water-clear and still work.) 6. Dialyze 3 times against 500 ml PBS + 0.01% sodium azide as preservative to remove glutaraldehyde. 7. Add bovine serun albumin, 5 mg/ml and store at 4° C. (Sigma No. A-7638.) APPENDIX C APPENDIX C Pedigrees and ratings of reciprocal crosses. 182 183 Table Cl. Pedigrees and ratings of reciprocal crosses. Listed on the left is genotype (female parent x pollen parent) followed by the S:I:R rating of the plants. Results of screening of F2 are presented on the same line as the F1 (or the next line if there is more than one set). The letters in parentheses denote the groups of plants that were screened at the same time. The last column represents the designation of the F1 plant which produced the seed for the F screening. *indicates a plant which was screened by TMV-Flavum rather than ToMV-0. "un" indicates that F seed for screening were collected from uninoculated and unrated plants. Somaglong 12 F1 F2 no. A12-9 x 26 3:2:1(I) 2:2:8(T) 1 A12-12* x 26 6:6:2(I) 2:3:9(T) 6 2:4:5(n) 7 A12-9-9 x 26 8:4:0(BB) 26 x A12-8 10:2:0(I) 5:4:7(R) lun 7:4:1(Z) 6:5:3(z) 2 26 x A12-9* 10:1:1(N) 3:2:2(z) lun 9:3:0(I) 3:4:1(Q) 4:1:7(W) lun 26 x A12-10* 5:7:0(H) 8:3:2(R) lun 26 x A12-9-9 5:4:3(BB) Seamless 2.12 F1 F2 n0- D215-8* x 26 0:3:9(I) 2:1:13(R) lun G215-7 x 26 3:3:0(Q) 2:4:0(Q) G215-10 x 26 4:4:4(Z) 5:6:5(i) 5 2:1:9(n) 5 26 x B215-1* 5:7:0(H) 9:3:4(R) lun 9:3:0(I) 3:3:6(n) 10 26 x F215-2 1:5:0(Q) 3:3:0(Q) 184 Table C1. (continued) Somaclgne 212 F1 F2 no. D219-l x 26 3:3:0(N) 8:3:5(W) lun F219-4 x 26 2:4:0(Q) G219-5 x 26 1:3:2(Q) 1:2:3(d1) lun 3:1:6(d2) lun 1:2:5(dl) 3:3:l(d2) 26 x D219-1(0) 8:4:0(N) 4:2:7(W) un 3:0:2(d1) un 8:2:2(d2) un 0:0:2(dl) fiQEéQlQDE 231 F1 F2 no. G247-l x 26 0:3:9(Q) 0:2:14(c) lun 0:0:16(s) 11 1:1:14(u) lun 0:0:l4(x) 2un F247-1 x 26 0:1:4(T) 12:2:2(i) 3 F247-2 x 26 4:1:1(Z) ll:4:l(i) 2un 4:1:8(s) 2un 6247-4-5-4 x 26 2:9:1(BB) 26 x A247-15* 4:5:0(H) 3:4:1(I) 11:0:5(R) lun 26 x A247-7 11:1:0(N) 4:5:7(u) lun 6:2:8(x) lun 0:2:ll(s) 2un 0:2:12(u) 2un 26 x G247-4-5-4 10:2:0(BB) 185 Table Cl. (continued) Somaclone 322 F1 F2 no. D322-3 x 26 0:3:0(Q) D322-10 x 26 0:3:4(Q) ll:4:l(c) 9 2:2:12(x) 6 26 x D322-8 5:1:0(N) 4:7:1(Z) 10:3:3(i) 12 26 x D322-3 6:0:0(N) 4:5:3(W) lun 3:4:9(x) lun S ma one 33_ F1 F2 no. D330-5 x 26 O:6:3(S) 11:3:2(c) 9 2:3:1(Z) 4:5:7(y) 3 D330-7 x 26 0:2:1(S) 2:3:11(x) 11 26 x D330-3 11:1:0(N) 5:3:5(W) lun 10:2:4(y) lun 26 x D330-4 4:5:3(2) 5:1:10(x) 11 6:4:3(i) 12 MICHIGAN STATE UNIV. LIBRARIES llHllllll"llllllllllllllllll”lllllllllllllllllllll 31293005756865