| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Laboratory of Phytopathology, Wageningen University, and Graduate School of Experimental Plant Sciences, 6709 PD Wageningen, The Netherlands
4 Corresponding author: Laboratory of Phytopathology, Wageningen University, Binnenhaven 5, 6709 PD Wageningen, The Netherlands.
E-mail: francine.govers{at}wur.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Generally, the sexual inheritance of molecular markers in P. infestans and related oomycetes appears to be Mendelian. So far, molecular markers have been exploited to construct genetic linkage maps from three oomycete species, P. infestans (VAN DER LEE et al. 1997), Phytophthora sojae (MAY et al. 2002), and Bremia lactucae (SICARD et al. 2003). In P. infestans, however, deviations occasionally were found due to the presence of one or three alleles of a locus in the progeny, suggesting the occurrence of translocations, aneuploidy, and hemizygous regions (JUDELSON 1996a; CARTER et al. 1999; JUDELSON and ROBERTS 1999; VAN DER LEE et al. 2001b), and in Phytophthora cinnamomi similar abnormalities were reported (DOBROWOLSKI et al. 2002). These studies were all based on multi-locus markers that did not span complete linkage groups.
To gain better insight into the genetics of P. infestans, we performed linkage analysis of markers in two F1 progenies, integrated the two maps, and scanned the individual progenies for abnormalities. Previously, we constructed an amplified fragment length polymorphism (AFLP) linkage map based on an F1 progeny obtained by crossing two Dutch field isolates and positioned the mating-type locus and six race-specific avirulence genes on this map (VAN DER LEE et al. 1997, 2001a). We extended the linkage analysis using the same progeny and analyzed segregation in a second F1 progeny obtained from two Mexican field isolates that are genetically unrelated to the Dutch field isolates. The maps are comparable; however, several cases were found where markers appear to be in a different order or not linked at all. This is particularly evident on linkage group III that contains the mating-type locus. The high-density maps also revealed trisomic progeny that originate from abnormalities during meiosis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Marker generation:
AFLP markers were generated as described by VOS et al. (1995), using the restriction enzyme combination of EcoRI-MseI with two selective bases on each side. DNA isolation and template preparation were described before (VAN DER LEE et al. 1997). Three types of markers were scored: A markers (genetic model Aa x aa), B markers (genetic model aa x Aa), and H markers (genetic model Aa x Aa). The same person scored all markers visually. Markers were named by the type of marker (A, B, or H) followed by the primer combination used to generate the marker and the position in the gel, either as a size estimate (indicated by the letter "s") or as a fragment number (indicated by an "f"). Markers that segregate in both crosses are called common markers and have an additional prefix C. Markers segregating in cross 71 are shown in uppercase (e.g., CAE + AA/M + CAs201.9), while cross 68 markers are shown in lowercase (e.g., cae + aa/m + cas 206.5). Fingerprinting of the two parental lines of cross 71 and 12 of its progeny was performed twice. Scoring accuracy was evaluated by the percentage of dissimilarity in the scorings between replicated individuals over the total number of scored markers using Microsoft Excel. For comparative studies precise length estimates of the AFLP markers are needed. Upon request the authors will provide more detailed information, including the visualization of the markers on gel. Apart from AFLP markers, some RFLP, RAPD, and allozyme markers were scored. These markers are also named by the probe, primer, or protein used to generate the marker followed by a fragment number or size. Some cross 68 markers were generated by A. Dyer and W. E. Fry from Cornell University and P. W. Tooley from the U.S. Department of Agriculture (Frederick, MD). For the RFLP, RAPD, and allozyme data, no quality assessment was performed. Allelic markers originating from the same RFLP or allozyme were treated as separate markers. The mating type was determined by checking for presence of oospores when co-cultured with tester strains of known mating type.
Linkage analysis:
All individuals generated in the two crosses were analyzed for genetic distance using the Dice index within the NTSYS software package (ROHLF 1993). Isolates with a genetic similarity of 95% or more were considered genetically identical and only one of each set was used in the linkage analysis carried out in JoinMap 3.0 (http://www.kyazma.nl/join/jm_intro.htm). In the AFLP linkage analysis, progeny for which a large number of markers could not be scored reliably, as well as markers that could not be scored reliably on a large set of individuals, were set aside from the data set. Markers were assigned to linkage groups using LOD values ranging from 3.5 to 8.5. Only linkage groups with at least three markers were used for map calculation. To obtain a reliable grouping of the markers, we tested whether (i) the segregation ratio of the markers was in accordance with other markers in the group, (ii) the linkage phase of the markers could be determined relative to the other markers, (iii) the markers were linked to at least two distinct other loci given a LOD value of 1, and (iv) the markers showed clear polarity in the recombination events with other markers. Markers that did not meet these criteria were rejected. Finally, we analyzed whether the A and B markers separately had enough linkage data to support a single group, given a LOD threshold value of 3. To determine the order of markers in the linkage groups, we first calculated the order of markers originating from either the A1 parent (Aa x aa), resulting in an A map, or the A2 parent (aa x Aa), resulting in a B map. For the larger linkage groups with many A or B markers, A and B maps for each linkage phase (chromosome) also were calculated. The marker order was calculated using all linkage data with a LOD score of 1 or higher and a maximum distance of 40 cM with the Kosambi mapping function. A and B maps were used as a fixed framework for the integrated map containing H markers (Aa x Aa) as well as A and B markers. JoinMap 3.0 provides the opportunity to set aside markers that cause friction in the map, which is detected by a jump in the 
2 value for the map upon addition of each marker. This 2 was set at the default value 5. When all markers were analyzed like this, JoinMap performed a second round attempting to position the markers that initially caused friction and were set aside, thereby using the map calculated in the first round as fixed order and applying the same 2 restriction. Maps were drawn with MapChart (VOORRIPS 2002).
Alignment of the maps of cross 71 and cross 68:
The maps of cross 71 and cross 68 were aligned using markers that segregate in both crosses, the so-called common markers indicated with a C as prefix. Similar to the integration of the A and the B maps, which was done by markers present in both parents that segregated in the progeny, the alignment was based on markers that had an identical fragment size and an identical intensity and that segregated in both crosses. However, in contrast to markers generated within one cross, markers common in two crosses do not necessarily fit in the same genetic model in each of the crosses. For instance, an A marker in cross 71 (Aa x aa) may be an A (Aa x aa), B (aa x Aa), or H (Aa x Aa) marker in cross 68 and vice versa. The information for the comparison is more reliable when markers of the same type can be compared but more difficult if the comparison involves combinations of marker types. The alignment was done graphically and the position of the common markers was carefully examined using the calculations generated by JoinMap.
Graphical genotyping of the progeny:
For the identification of trisomic or monosomic progeny, we made a graphical display of the markers of each linkage group for each of the progeny. Trisomic progeny were detected by the presence of all A markers or all B markers of a particular linkage group regardless of the linkage phase (chromosome) from which the markers originate (see Figure 3 as an example). Only linkage groups with five or more markers originating from the same parent with at least two changes of the linkage phase were assayed. Graphical genotyping of the progeny was performed in Microsoft Excel and in GGT (VAN BERLOO 1999).
|
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
2 jump (7 and 4, respectively), indicating that there is not much friction in the data for marker ordering. Also, the comparison of linkage maps calculated for markers of a single linkage phase performed on linkage groups I, III, IV, V, and VIII did not reveal differences. The calculation of the integrated map (A, B, H) using the fixed order from the A and B map appeared to be slightly more difficult as judged by the 2 values and the compression of the genetic distance observed, for instance, at the top of LG IV (as shown in Figure 2C). Nevertheless, in most cases the relative position of the H markers in the map with the A and H markers is similar to the map with the B and H markers. An exception was LG III, which contains the mating-type locus. In this case the integration of the A and B map using the H markers caused great difficulties. Apart from the problems encountered in LG III (addressed below), most of the markers could be positioned on the integrated map—480 after the first and 508 after the second round. The 26 markers that could not be placed with this 2 restriction were not positioned on the map and are listed in the box in Figure 1. Most of the rejected markers fit best in dense regions where a single scoring error can lead to high 2 jumps. The added value of these markers is low.
|
|
|
|
The integrated linkage map of cross 71 comprises 23 linkage groups (Figure 1) of which 13 contain markers from both parents (A, B, and H markers). The latter are referred to as major linkage groups and are labeled with roman numbers. The remaining 10 minor linkage groups are composed of markers from a single parent and are named by their parental origin, A1 or A2, followed by a letter. Except for LG A1-a (81.6 cM) and A2-a (52.8 cM), the minor linkage groups are relatively small (3–32 cM). LG III-a, XII, A1-d, and A1-e showed strong aberrations in the segregation ratios (listed in Table 1). LG A1-d and LG A1-e contain only A markers of a single linkage phase; LG VIII and LG IX also contain a large number of markers of only a single linkage phase. The major linkage groups have a high marker density with an average interval of 2.1 cM. However, some linkage groups, such as LG XII, or regions, such as the top of LG XI or the lower region of LG IX, are low in marker density, particularly for markers derived from one of the parents or from a specific linkage phase. Some linkage groups contain clusters of markers in coupling phase of a single parent, originating from a single chromosome. The length of the linkage groups ranges from 3 to 101 cM. Particularly short for the number of markers it contains is LG VI with 22 markers on 10.7 cM; in this case it is even unclear if the map length in centimorgans extends beyond the experimental noise generated by the integration of the A and B map.
The integration of the A and B map using the H markers caused difficulties in LG III, the LG containing the mating-type locus. Closer examination revealed a cluster of six H markers (underlined in Figure 1) strongly linked to the B markers in LG III but not linked at all to any of the A markers in this linkage group. This is remarkable since in the corresponding segment A markers are present and should have been linked if the H markers on the III-a map were in the same position. This indicates that the A and B map were dissimilar for LG III. Consequently, we did not construct an integrated map of LG III but instead constructed separate A and B maps called III-a and III-b. In the two maps the position of the H markers was calculated by adding each and every H marker separately to avoid interference by other H markers. We analyzed whether the six H markers from III-b, which had no linkage to the A markers in III-a, were linked to A markers in other linkage groups but we could not identify even weak linkage (LOD <1). Three other H markers had low LOD values as compared to the A markers. The H marker on top of III-b (CHE + CC/M + CTf16) had a LOD value of 1 with some A markers of III-a, and two H markers at the bottom of III-b (HE + GG/M + CCf13A and HE + CA/M + CCf19) were linked to three A markers that also linked together (AE + AA/M + CTs136.3, CAE + AA/M + CTs192.8, AE + AG/M + AAf19A). Although the LOD values were low, the data supported the fusion of these three A markers with LG III-a. Segregation ratios were similar and the direct distance between the A markers as well as between the H markers matched with the calculated distances and the distance of these markers on the III-b map. Therefore we added these markers to LG III-a despite the fact that the significance for linkage is lower (as indicated by the dashed line in LG III-a in Figure 1).
Map construction in cross 68:
From cross 68, 62 F1 progeny were available. In these progeny 465 AFLP, 17 RFLP, 31 RAPD, and two allozyme markers were scored as well as the mating type. The majority of the markers were AFLP markers and the reliability of this data set, as judged by the scoring of replicated samples, ranged from 99.8 to 99.9%. Fifteen progeny had information on <443 markers and these progeny were excluded from further analysis. Also DNA markers that were scored on <32 progeny were excluded from further analysis, as were markers that segregated with an aberrant segregation ratio (2 > 5). Linkage analysis was performed with the remaining 425 markers on 47 progeny. Using variable LOD values ranging from 3.5 to 8.5, 392 markers could be grouped into 24 linkage groups (Figure S1 and Table S1 at http://www.genetics.org/supplemental/): 8 major linkage groups containing A, B, and H markers and 16 minor linkage groups, 8 with markers originating from the A1 parent and 8 with markers from the A2 parent. The order of the markers was calculated but in some of the linkage groups the relatively low number of H markers and the limited number of progeny made integration of the A and B maps more difficult. For some linkage groups, integration was supported by multi-allelic markers which, when treated as independent markers, mapped within 4–10 cM in the same linkage group. Overall, the map generated from the linkage analysis in cross 68 is more fragmented than the map of cross 71.
Alignment of the maps of cross 71 and cross 68:
For alignment of the maps we used markers segregating in both crosses. These common markers were identified by fingerprinting the four parental lines and a number of their progeny side by side, followed by careful inspection of the fingerprints for bands with identical length and intensity. We started the alignment of the two maps by grouping the common markers per linkage group. Groups of two or more common markers that were mapped on one linkage group in cross 71 also mapped on one linkage group in cross 68 (Table 2) but there were clear exceptions, for instance, in LG I and LG III. We retested the grouping of these markers, and all groupings appeared reliable. The numbering of the linkage groups in cross 68 follows the numbering of the linkage groups in cross 71 as much as possible. As a consequence, linkage groups in cross 68 indicated by roman numbers are not necessarily major linkage groups and the A1 and A2 linkage groups are not necessarily minor linkage groups. Subsequently, we tested if linkage groups could be merged on the basis of linkage information obtained from the other cross. For example, the exchange of common markers from LG XIII and A2-a suggested that these two groups compose one linkage group. Markers on LG XIII71 and LG A2-a71, however, were not linked and could not be merged in one linkage group despite the fact that enough relevant marker data were available.
The different groupings already suggested that an overall integration of the two maps on the basis of common markers was not possible. We then compared the order of the markers within the five linkage groups for which four or more common markers were available (i.e., LG I, LG III, LG IV, LG VIII, and LG XI). In LG VIII and LG XI, alignment of the maps from cross 68 and cross 71 showed minor differences in the order of the markers that remained within the mapping resolution and also the distances between the markers were similar (Figure 2, D and E). In the comparison of LG IV (Figure 2C), the order and the distance between the markers was similar for four marker pairs but not for marker pair CAE + AA/M + CAs201.9-cae + aa/m + cas206.5. In cross 71, marker CAE + AA/M + CAs201.9 was mapped distal from markers HE + CA/M + CCf16 and AE + GG/M + CAf10 (Figure 1) and also the position of marker cae + aa/m + cas206.5 in cross 68 was clear from the relative position to markers cae + ac/m + cts228.3 and cae + ct/m + tgf17 (Figure S1 at http://www.genetics.org/supplemental/). In these cases all markers originated from the same parent and therefore provided maximum mapping resolution. Also, the fact that marker CAE + AA/M + CAs201.9 was quite distant from the other common markers in cross 71 whereas cae + aa/m + cas206.5 was close to the common markers in cross 68 made an identical position unlikely.
In cross 71 the A and B map of LG III could not be integrated (Figure 1) but there was no problem with the integration of LG III in cross 68. Of the six markers that distinguished III-a71 from III-b71 (HE + GT/M + GCf4, HE + CC/M + CCf10, HE + AC/M + CTs196.9, CHE + AC/M + CTs194.6, CHE + AA/M + CAs152.5, and HE + AC/M + TTf6), two (CHE + AA/M + CAs152.5 and CHE + AC/M + CTs194.6) had a corresponding marker in cross 68 (che + aa/m + cas161.8 and cbe + ac/m + cts182). Both these markers were positioned on LG III68 (Figure 2B). Furthermore, marker che11m15s161.8 was linked to both A and B markers in LG III68, demonstrating that the two parents of cross 68 are not dissimilar in this region. Further comparison revealed that the integrated map of LG III68 is dissimilar to both the III-a71 and the III-b71 map. The corresponding markers for markers CBE + AC/M + CCs508.7 and CBE + AG/M + CAs307.6 (cbe + ac/m + ccs588.6 and cae + ag/m + cas284.9, respectively) were not positioned on LG III68. Instead, cae + ag/m + cas284.9 was linked (LOD value of 5.9) to marker PEPI in cross 68 on a LG with no other common markers and marker cbe + ac/m + ccs588.6 was linked (LOD value of 6.5) to RG57/1H on LG I(part)68 and, conversely, marker CBE + GA/M + TGf3 corresponding to marker che + ga/m + tgf11 positioned on LG III68 was not mapped on LG III in cross 71 but on LG VII71. Nevertheless, seven other markers and the mating-type locus were linked and their relative positions were similar.
In LG I, the differences in the order and the distance of four of the six common markers was within the mapping resolution (Figure 2A). In cross 68, marker cbe + ct/m + tgf13a, corresponding to marker CAE + CT/M + TGf11 in cross 71, was not linked to any of the markers from LG I68, but this can be due to the absence of B and H markers in the corresponding part of the map in cross 68. However, marker cae + ct/m + ccf6b was located on LG I68 but the corresponding marker in cross 71 (CHE + CT/M + CCF17) was mapped on LG X71.
Part of the progeny is trisomic for one or more chromosomes:
While constructing a linkage map, the inheritance of markers in each individual progeny can be visualized in the order of the map by graphical genotyping. This procedure is generally used to identify errors in marker scoring. Such errors often result in apparent double crossing-over events, but these are unlikely to occur. In this study we used graphical genotyping to identify aberrant progeny. In both cross 68 and cross 71, part of the progeny contained all markers from a particular chromosomal pair from one of the parents. One example of trisomy of LG IV is shown in Figure 3. One of the progeny of cross 71, D12-17, contained all LG IV markers from the A1 parent regardless of whether the markers were in coupling phase (same chromosome) or in repulsion phase (homologous chromosome). In addition, D12-17 received a third chromosome represented by LG IV markers derived from the A2 parent. D12-17 thus received both homologous chromosomes of a specific pair from one parent and one recombinant chromosome from the other parent.
The exact number of trisomic progeny is difficult to give since some linkage groups contained only markers of a single linkage phase or the number of phase transitions was too low. As was mentioned above and illustrated in Figure 4, there was a strong tendency for tightly linked markers to occur in coupling phase. Still, by analyzing all linkage groups with two or more phase transitions we could obtain an estimate using these criteria: 8 cross 71 progeny were trisomic on the basis of the A map while 5 progeny were trisomic on the basis of the B map. When the information on the A and B map was combined, 11 progeny (16.1%) were trisomic: 8 for one linkage group and 1 for two, 1 for three, and 1 for six linkage groups (Table 3 and Table S2 at http://www.genetics.org/supplemental/). For some linkage groups no trisomy was found and consequently none of the progeny appeared to be triploid. Furthermore, in cross 71 we did not find progeny that were monosomic, as would be detected by the absence of all markers from one of the parents for a particular linkage group. Using the same criteria for cross 68, 10 progeny were trisomic on the basis of the B map while 1 was trisomic on the basis of the A map. In total, 11 progeny were trisomic and also one putative monosomic genotype was identified. If multi-allelic markers were available, the trisomic isolates always correlated with the identification of three alleles of RFLP or isoenzyme markers. However, the presence of a single or three RFLP or isoenzyme alleles frequently did not correlate to monosomy or trisomy of the linkage group.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
A high-density genetic linkage map for P. infestans using cross 71:
The markers analyzed consist largely of AFLP markers and reliability of this data set as judged by replicate samples of the same isolate was high (nearly 99%). All segregating markers, even those that showed strong deviation of the expected segregation ratios, were scored and analyzed. Over 90% of the high-quality markers are linked in 23 linkage groups and >85% of these markers could be positioned. Markers that could not be mapped are largely dominantly scored H markers, which intrinsically are less informative and therefore more difficult to group and position. The robustness of the grouping was tested by building the maps with the two independent marker types, which could be done for all linkage groups except for LG III. The integration of the A and the B maps resulted in a map with 508 AFLP markers distributed over 13 major and 10 minor linkage groups. The major linkage groups are dense in markers with an average marker spacing of 
2.0 cM. Nevertheless, some regions are low in marker density, particularly for markers originating from one of the parents or from one linkage phase, indicating that the map is far from saturated. Also, the number of linkage groups is remarkably high for the expected 8–10 chromosomes (SANSOME and BRASIER 1973). It seems that although >90% of the markers are linked, some parts of the genome remain uncovered, resulting in gaps in the linkage map. This problem may be caused by low polymorphism between the parental isolates for the homologous chromosomes in some regions. The alternative perspective would be that most markers originate from hypervariable regions and the tight clustering of markers in linkage phase seems to point in this direction. Hemizygous regions were identified before (JUDELSON 1996b; VAN DER LEE et al. 2001b) and may explain this observation.
Comparison with the previous maps of cross 71:
In general, the grouping in the new map fits well with that of the first genetic linkage map of P. infestans and is identical to the partial maps presented before (VAN DER LEE et al. 1997; 2001a). Compared to the previous map, LG I, IV, V, VIII, IX, A1-a, A2-a, A2-c, and A2-d remain within the same boundaries upon the addition of the new markers (Table 1). The higher marker density revealed dissimilarity between the III-a and III-b map that was not noted previously. Some minor linkage groups could be integrated in a major linkage group: LG A2-e was integrated into LG X, and A1-b, A2-b, and part of LG V were fused to the new LG XI. The most important differences from the previous map are found in LG II, VI, and VII, all of which are now split into two or more linkage groups (Table 1). In the case of LG II, markers with distorted segregation ratios erroneously merged three linkage groups and hence LG II is now split in LG II, LG XII, and LG A1-e. In the case of LG VI and LG VII, the number of bridging H markers in the previous map appeared to be too low and LG VI is now split in LG VI and LG A1-f, while LG VII is split in LG VII and LG XIII.
The current AFLP map was constructed with more markers using more stringent LOD values and stricter criteria. With the exception of three markers that were forced to the end of their linkage group by the JoinMap algorithm, the order and distance between the markers in the current map and the previous map is similar, at least within the mapping resolution. Occasionally markers from different parents "slide" over each other, which is quite understandable, given the mapping resolution of the H markers. The software program JoinMap3.0 is much more sophisticated than the JoinMap1.4 version used for the first map. JoinMap3.0 allows easy identification of markers with aberrant segregation ratios and markers that cause friction in the map are set aside temporarily. The Windows-based user-interface makes it easy to analyze effects of different settings and contributions of individual markers or individual progeny to the map. Nevertheless, construction of integrated maps based solely on dominant markers remains difficult and since there is no direct linkage information between A and B markers, integrated maps should be handled with caution. Reassessment of the present grouping will be needed when more markers are added. We anticipate that with more markers, groups will merge, generating a number of linkage groups that is closer to the predicted number of 8–10 chromosomes (SANSOME and BRASIER 1973).
Translocations and other aberrations at LG III:
In cross 71, the A and the B map of LG III are dissimilar in the region close to the mating-type locus. A group of six H markers, in the middle of LG III-b, are not linked to markers positioned on LG III-a. This does not involve a deletion, since by nature the H markers (Aa x Aa) are present in both parents. This absence of linkage therefore is reminiscent of a translocation. However, we could not identify the repositioning of this region to any of the other linkage groups. The translocation does not seem to correlate to the mating type as such, since the parental isolates of cross 68 do not appear to differ in this region. JUDELSON et al. (1995) found only two of the four expected combinations of gametes in the progeny of some crosses, suggesting balanced lethals and possibly generated by balanced translocations. In cross 71 the translocation does not seem to be balanced: all four allelic combinations were found, but then again segregation ratios for the A1 and A2 mating-type locus and markers from the A1 parent on the same chromosome are distorted (VAN DER LEE et al. 1997). The progeny mapped in this study were generated from oospores obtained by infecting potato leaves with the parental strains and were recovered from sporulating lesions formed on leaves floating on water mixed with soil containing the oospores (in vivo; DRENTH et al. 1995). Remarkably, when progeny of the same parental strains were generated from oospores obtained by co-cultivation on rye medium and germination of these oospores on water agar (in vitro), the progeny showed no distorted segregation ratios for the mating-type locus or for markers linked to the mating type (T. VAN DER LEE and F. GOVERS, unpublished results). The fact that the distorted segregation ratios were found only in the in vivo progeny suggests that progeny with the A1 mating type have an advantage during in vivo development, survival, and/or growth. In this respect it may be significant that the A1 mating type was distributed all over the world while the A2 mating type was restricted to some areas (FRY et al. 1992). Even now, in populations where sexual reproduction occurs, the ratio of A1 to A2 is biased for the A1 mating type, particularly after prolonged periods that favor vegetative growth (ZWANKHUIZEN et al. 1998, 2000). Another remarkable feature is that in all studies reported, including this study, the A1 mating type is dominant, whereas A2 isolates of P. infestans can be self-fertile (SMART et al. 2000). Obviously, the mating-type locus is one of the most challenging and intriguing areas for genetic studies in P. infestans.
Alignment of two maps reveals more translocations:
The map of cross 68 with 363 markers positioned on 24 linkage groups is less dense and more fragmented than the map of cross 71. Following the successful integration of A and B maps on the basis of common H markers, we used comigrating AFLP fragments as common markers for integration of the maps of cross 71 and cross 68. We anticipated that the map of cross 71 could serve as a backbone for the map of cross 68 and that areas with low marker density, caused by low polymorphism between parental isolates of one cross, would benefit from markers obtained in the same region in the other cross. In general the grouping and order of the common markers in the two crosses matched, but we also found several dissimilarities. This points toward differences in grouping and/or ordering of markers in the parental isolates of cross 68 and cross 71 and is reminiscent of translocations (Table 2, Figure 2). It is true that comigrating AFLP fragments do not necessarily represent the same fragment, but since H markers could readily be used to integrate A and B maps there is no reason why C markers could not be used for the same purpose. These findings suggest that translocations are not rare in Phytophthora and other studies support this. In P. sojae rearrangements were found within a linkage group (MACGREGOR et al. 2002), whereas in P. infestans a marker linked to the mating-type locus appeared to have translocated to another linkage group (RANDALL et al. 2003). Translocations create the possibility of obtaining one, two, or three copies of a locus or region, which may result in balanced lethals (JUDELSON 1996a) or in high frequencies of nonviable oospores, as observed in many crosses (KNAPOVA et al. 2002). On the other hand, the flexibility to have one, two, or three copies of a genomic region can be advantageous for adaptation.
Trisomic progeny:
Previously, the occurrence of trisomic progeny in two Phytophthora species, P. infestans and P. cinnamoni, was postulated on the basis of the presence of three alleles of multi-locus markers (CARTER et al. 1999; DOBROWOLSKI et al. 2002). In those studies the number of markers was limited and the presence of three alleles could not be assessed for a complete linkage group. In this study we used a different approach to identify trisomic progeny and demonstrated that trisomy extends over the whole chromosome. For each linkage group with at least 5 markers from one of the parents and two linkage-phase transitions, we tested whether we could detect the presence of two copies of a chromosomal set of one of the parents using the linkage phase of the markers. If all markers from a particular parent are present in the progeny regardless of the linkage phase, then both parental chromosomes are transmitted to its offspring. As summarized in Table 3, we found significant numbers of trisomic progeny for specific chromosomes. Obviously, with this method trisomy of linkage groups with a few markers or only one or no phase transition remains undetected whereas more markers and more phase transitions result in more solid data. For LG IV, e.g., with 21 markers originating from parent 80029 and six phase transitions we could demonstrate that trisomy extends over at least 50 cM of a total of 99 cM (Figure 3). We also identified progeny that had only a part of a linkage group in triplicate (Table S2, at http://www.genetics.org/supplemental/). In cross 68, some of the multi-allelic markers suggested the presence of one or three copies of a chromosome pair but this was not confirmed by other markers on the same linkage group, whereas other linkage groups were clearly trisomic. This indicates that, apart from trisomy, other aberrations, possibly translocations or postfusion instability, do occur. Remarkably, in three progeny of cross 71 the intensity of a group of markers linked to the avirulence gene cluster Avr3-Avr10-Avr11 was significantly reduced to 50% (data not shown), also pointing at postfusion deletions or mitotic gene conversions (as described for P. sojae by CHAMNANPUNT et al. 2001). Because we can detect trisomy only in linkage groups with sufficient markers and phase transitions, assessment of the precise percentage of trisomic progeny is impossible but definitely 10–16% of the progeny from both crosses have one or more trisomic linkage groups. In contrast, we encountered only a single individual that was possibly monosomic for one linkage group and no progeny that were triploid.
The parental isolates of cross 71 are both field isolates and their genetic diversity is in line with the diversity found in the Dutch field population. No crossing barriers appear to exist between the parental isolates, as the progeny were no more heterozygous than the parental isolates (data not shown). The parental isolates were also used for mixed inoculations in field trails and numerous oospores were found (DRENTH et al. 1995). The cross 71 progeny, including the trisomic progeny, are pathogenic on potato. We therefore think that trisomy in P. infestans is not an artifact and that it occurs in a natural environment. Oomycetes have a multinucleate coenocytic mycelium and variable chromosome numbers may be less problematic. We hypothesize that trisomy and subsequent instability of chromosomal regions, or loss of complete chromosomes, contribute significantly to the notorious genetic flexibility of P. infestans.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
|
FOOTNOTES |
|---|
2 Present address: Di.Bi.Ag.A., Polytechnic University of Marche, 60131 Ancona, Italy.
3 Present address: RSBS, Australian National University, Canberra, ACT 2601 Australia.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
CARTER, D. A., K. W. BUCK, S. A. ARCHER, T. VAN DER LEE, R. C. SHATTOCK et al., 1999 The detection of nonhybrid, trisomic, and triploid offspring in sexual progeny of a mating of Phytophthora infestans. Fungal Genet. Biol. 26: 198–208.[CrossRef][Medline]
CHAMNANPUNT, J., W. SHAN and B. TYLER, 2001 High frequency mitotic gene conversion in genetic hybrids of the oomycete Phytophthora sojae. Proc. Natl. Acad. Sci. USA 98: 14530–14535.
DOBROWOLSKI, M. P., I. C. TOMMERUP, H. D. BLAKEMAN and P. A. O'BRIEN, 2002 Non-Mendelian inheritance revealed in a genetic analysis of sexual progeny of Phytophthora cinnamomi with microsatellite markers. Fungal Genet. Biol. 35: 197–212.[CrossRef][Medline]
DRENTH, A., E. M. JANSEN and F. GOVERS, 1995 Formation and survival of oospores of Phytophthora infestans under natural conditions. Plant Pathol. 44: 86–94.
ERWIN, D. C., and O. K. RIBEIRO, 1996 Phytophthora Diseases Worldwide. APS Press, St. Paul.
FRY, W. E., S. B. GOODWIN, J. M. MATUSZAK, L. J. SPIELMAN, M. G. MILGROOM et al., 1992 Population genetics and intercontinental migrations of Phytophthora infestans. Annu. Rev. Phytopathol. 30: 107–130.
GARELIK, G., 2002 Agriculture. Taking the bite out of potato blight. Science 298: 1702–1704.
GOODWIN, S. B., A. DRENTH and W. E. FRY, 1992 Cloning and genetic analyses of two highly polymorphic, moderately repetitive nuclear DNAs from Phytophthora infestans. Curr. Genet. 22: 107–115.[CrossRef][Medline]
JUDELSON, H. S., 1996a Genetic and physical variability at the mating type locus of the oomycete, Phytophthora infestans. Genetics 144: 1005–1013.[Abstract]
JUDELSON, H. S., 1996b Chromosomal heteromorphism linked to the mating type locus of the oomycete Phytophthora infestans. Mol. Gen. Genet. 252: 155–161.[CrossRef][Medline]
JUDELSON, H. S., and S. ROBERTS, 1999 Multiple loci determining insensitivity to phenylamide fungicides in Phytophthora infestans. Phytopathology 89: 754–760.
JUDELSON, H. S., L. J. SPIELMAN and R. C. SHATTOCK, 1995 Genetic mapping and non-Mendelian segregation of mating type loci in the oomycete, Phytophthora infestans. Genetics 141: 503–512.[Abstract]
KNAPOVA, G., A. SCHLENZIG and U. GISI, 2002 Crosses between isolates of Phytophthora infestans from potato and tomato and characterization of F-1 and F-2 progeny for phenotypic and molecular markers. Plant Pathol. 51: 698–709.[CrossRef]
MACGREGOR, T., M. BHATTACHARYYA, B. TYLER, R. BHAT, A. F. SCHMITTHENNER et al., 2002 Genetic and physical mapping of Avrla in Phytophthora sojae. Genetics 160: 949–959.
MAY, K. J., S. C. WHISSON, R. S. ZWART, I. R. SEARLE, J. A. G. IRWIN et al., 2002 Inheritance and mapping of 11 avirulence genes in Phytophthora sojae. Fungal Genet. Biol. 37: 1–12.[CrossRef][Medline]
RANDALL, T. A., A. H. FONG and H. S. JUDELSON, 2003 Chromosomal heteromorphism and an apparent translocation detected using a BAC contig spanning the mating type locus of Phytophthora infestans. Fungal Genet. Biol. 389: 75–84.[CrossRef]
ROHLF, F. J., 1993 NTSYS-pc: Numerical Taxonomy and Multivariate Analysis System, Version 1.80. Applied Biostatistics, New York.
SANSOME, E., and C. M. BRASIER, 1973 Diploidy and chromosomal structural hybridity in Phytophthora infestans. Nature 241: 344–345.[CrossRef]
SICARD, D., E. LEGG, S. BROWN, N. K. BABU, O. OCHOA et al., 2003 A genetic map of the lettuce downy mildew pathogen, Bremia lactucae, constructed from molecular markers and avirulence genes. Fungal Genet. Biol. 39: 16–30.[CrossRef][Medline]
SMART, C. D., H. MAYTON, E. S. G. MIZUBUTI, M. R. WILLMANN and W. E. FRY, 2000 Environmental and genetic factors influencing self-fertility in Phytophthora infestans. Phytopathology 90: 987–994.
VAN BERLOO, R., 1999 GGT: software for the display of graphical genotypes. J. Hered. 90: 328.
VAN DER LEE, T., I. J. WITTE DE, A. DRENTH, C. ALFONSO and F. GOVERS, 1997 AFLP linkage map of the oomycete Phytophthora infestans. Fungal Genet. Biol. 21: 278–291.[CrossRef][Medline]
VAN DER LEE, T., A. ROBOLD, A. TESTA, J. W. VAN ‘T KLOOSTER and F. GOVERS, 2001a Mapping of avirulence genes in Phytophthora infestans with amplified fragment length polymorphism markers selected by bulked segregant analysis. Genetics 157: 949–956.
VAN DER LEE, T., A. TESTA, J. W. VAN ‘T KLOOSTER, G. VAN DEN BERG-VELTHUIS and F. GOVERS, 2001b Chromosomal deletion in isolates of Phytophthora infestans correlates with virulence on R3, R10, and R11 potato lines. Mol. Plant-Microbe Interact. 14: 1444–1452.[Medline]
VOORRIPS, R. E., 2002 MapChart: software for the graphical presentation of linkage maps and QTLs. J. Heredity 93: 77–78.
VOS, P., R. HOGERS, M. BLEEKER, M. REIJANS, T. VAN DER LEE et al., 1995 AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res. 23: 4407–4414.
ZWANKHUIZEN, M. J., F. GOVERS and J. C. ZADOKS, 1998 Development of potato late blight epidemics: disease foci, disease gradients, and infection sources. Phytopathology 88: 754–763.
ZWANKHUIZEN, M. J., F. GOVERS and J. C. ZADOKS, 2000 Inoculum sources and genotypic diversity of Phytophthora infestans in Southern Flevoland, the Netherlands. Eur. J. Plant Pathol. 106: 667–680.[CrossRef]
This article has been cited by other articles:
 |
|
 |
|
  S. R. Harutyunyan, Z. Zhao, T. d. Hartog, K. Bouwmeester, A. J. Minnaard, B. L. Feringa, and F. Govers Biologically active Phytophthora mating hormone prepared by catalytic asymmetric total synthesis PNAS, June 24, 2008; 105(25): 8507 - 8512. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. H.Y. Jiang, R. Weide, P. J.I. van de Vondervoort, and F. Govers Amplification generates modular diversity at an avirulence locus in the pathogen Phytophthora Genome Res., July 1, 2006; 16(7): 827 - 840. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Fraser, J. C. Huang, R. Pukkila-Worley, J. A. Alspaugh, T. G. Mitchell, and J. Heitman Chromosomal Translocation and Segmental Duplication in Cryptococcus neoformans Eukaryot. Cell, February 1, 2005; 4(2): 401 - 406. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
