Pristionchus pacificus - Springer Link

Mol Gen Genomics (2003) 269: 715–722 DOI 10.1007/s00438-003-0881-8

O R I GI N A L P A P E R

J. Srinivasan Æ W. Sinz Æ T. Jesse L. Wiggers-Perebolte Æ K. Jansen Æ J. Buntjer M. van der Meulen Æ R. J. Sommer

An integrated physical and genetic map of the nematode Pristionchus pacificus Received: 15 May 2003 / Accepted: 6 June 2003 / Published online: 17 July 2003 Springer-Verlag 2003

Abstract The free-living nematode Pristionchus pacificus is one of several species that have recently been developed as a satellite system for comparative functional studies in evolutionary developmental biology. Comparisons of developmental processes between P. pacificus and the well established model organism Caenorhabditis elegans at the cellular and genetic levels provide detailed insight into the molecular changes that shape evolutionary transitions. To facilitate genetic analysis and cloning of mutations in P. pacificus, we previously generated a BAC-based genetic linkage map for this organism. Here, we describe the construction of a physical map of the P. pacificus genome based on AFLP fingerprint analysis of 7747 BAC clones. Most of the SSCP markers used to generate the genetic linkage map were derived from BAC ends, so that the physical genome map and the genetic map can be integrated. The contigs that make up the physical map are evenly distributed over the genetic linkage map and no clustering is observed, indicating that the physical map provides a valid representation of the P. pacificus genome. The integrated genome map thus provides a framework for positional cloning and the study of genome evolution in nematodes. Keywords Pristionchus pacificus Æ Caenorhabditis elegans Æ Genomics Æ Physical map Æ Amplified fragment length polymorphism (AFLP)

Communicated by G. Ju¨rgens J. Srinivasan Æ W. Sinz Æ R. J. Sommer (&) Abteilung fu¨r Evolutionsbiologie, Max-Planck Institut fu¨r Entwicklungsbiologie, Spemannstrasse 37, 72076 Tu¨bingen, Germany E-mail: [email protected] Tel.: +49-7071-601371 Fax: +49-7071-601498 T. Jesse Æ L. Wiggers-Perebolte Æ K. Jansen Æ J. Buntjer M. van der Meulen Keygene N.V., Agro Business Park 90, 6708 AE Wageningen, The Netherlands

Introduction Evolutionary developmental biology tries to elucidate how developmental and morphological diversity is generated over evolutionary time scales by comparing related species that differ in specific aspects of their development (Simpson 2002; Pires-DaSilva and Sommer 2003). Often, well-known model organisms, such as Drosophila or Caenorhabditis elegans, are compared to other insects or nematodes. Over the years, a small number of other species, so-called satellite organisms, have been selected for detailed analysis and comparison of selected developmental processes (Simpson 2002). In recent years, Pristionchus pacificus has become established as one such satellite organism that can contribute to the elucidation of diverse issues in evolutionary developmental biology (Eizinger et al. 1999; Felix et al. 2000; Sommer 2000). P. pacificus is a free-living nematode of the Diplogastridae family and, like the model organism Caenorhabditis elegans, it can be easily cultured in the laboratory. It has a 4-day life cycle and can be propagated by self-fertilizing hermaphrodites or by outcrossing with males that arise spontaneously (Sommer and Sternberg 1996). In addition, many cellular, genetic and molecular techniques that have been successfully used in C. elegans are also applicable to P. pacificus. Comparative studies of vulva development between C. elegans and P. pacificus have indicated that, although homologous precursor cells are involved in vulva formation, the cell-cell interactions required to form a normal vulva vary greatly between the two species (for review, see Sommer 2001). In addition, comparison of mutations in homeotic genes in both species has revealed substantial differences in the function of these genes during vulva formation (Eizinger and Sommer 1997; Jungblut and Sommer 1998; Sommer et al. 1998; Jungblut et al. 2001). We have previously described the construction of a BAC-based genetic linkage map for P. pacificus, which

716

accelerates positional cloning in this organism (Srinivasan et al. 2002). The assembly of this map was aided by the high degree of sequence polymorphism that exists between the wild-type strain of P. pacificus from California and a strain from Port Angeles, Washington (Schlak 1997; Srinivasan et al. 2001). In addition, detailed sequence analysis of a 126-kb P. pacificus BAC clone indicated the existence of longrange synteny, but only limited microsynteny, between the genomes of P. pacificus and C. elegans (Lee et al. 2003). To further facilitate the cloning of mutations in P. pacificus, an integrated physical and genetic map is required. Integrated genetic and physical genome maps not only make map based cloning possible, but are also useful for genome sequencing projects. Physical genome maps have been constructed using a fingerprinting technique that recognizes restriction fragments common to different genomic clones, thus establishing overlaps between them. Overlapping genomic clones are then assembled into contigs (Marra et al. 1997). However, this fingerprinting technique is limited in its ability to order complex genomes, and it does not provide a link to genetic maps and sequence information a priori. Hence, additional PCR-based or hybridization-based methods must be used, in combination with fingerprinting, to construct high-quality genome Fig. 1 AFLP fingerprint pattern of individual BAC clones run on the MegaBace 1000 machine. AFLP fingerprinting was done with Hind III/MseI (+0/+0). Each lane represents an AFLP fingerprint of an individual BAC clone. The arrows at the top and bottom of the image indicate the region of the gel used for fingerprint analysis. The asterisks indicate the bands derived from the vector during AFLP analysis. The fingerprints obtained were then analyzed using BACXtractor (see Materials and methods)

maps of various organisms (Marra et al. 1997; Cao et al. 1999). Here, we describe the generation of a physical map of the P. pacificus genome based on AFLP analysis of 7747 BAC clones, and its integration with the genetic linkage map. We show that the contigs that comprise the physical map are equally distributed over the genetic linkage map and that no clustering is observed. Thus, the physical map provides a valid representation of the P. pacificus genome.

Materials and methods AFLP fingerprinting of BAC clones BAC DNA was isolated (Srinivasan et al. 2002), and Hind III/ MseI AFLP templates were prepared (Vos et al. 1995). A 5-ll aliquot of the restriction ligation mix was diluted tenfold in 10 mM TRIS-HCl pH 7.5, 0.1 mM EDTA buffer, and non selective amplification was carried out with fluorescently labeled Hind III primer (5¢-GACTGCGTACCAGCTT-3¢) and an unlabeled Mse I primer (5¢-GACGATGAGTCCTGAGTAA-3¢). PCR was performed in a total volume of 20 ll (Vos et al. 1995) for 35 cycles of DNA denaturation at 94C for 30 s, annealing step at 56C for 1 min, and extension at 72C for 1 min. The fluorescent primers were synthesized at MWG-Biotech (FAM and JOE) and Applied Biosystems (NED). All amplification reactions were performed in a PE-9700 thermocycler (Perkin Elmer). After the amplification step, the reaction mix was desalted using Sephadex

717 Fig. 2A–C Fingerprint analysis of a representative contig in the P. pacificus physical map. A Raw AFLP data for the contig 340 obtained using Hind II/ MseI(+0/+0). B FPC v6.2 fingerprint analysis window displaying representative fingerprint patterns of individual clones from ctg340. C FPC v6.2 contig window displaying a horizontal representation of ctg 340. The asterisks and signs to the right of some clones indicate canonical and approximate match clones, respectively (Soderlund et al. 1997)

718 Table 1 Automated contig assembly of the P. pacificus physical map at different Tolerance and Cutoff values Tolerance Cutoff Number of contigsa Maximum number of contigs Size distribution of contigs by BAC number

2 2 2 3 3 3

e-06 e-07 e-08 e-06 e-07 e-08

260 349 426 272 369 442

(370) (435) (495) (378) (440) (495)

4017 (491) 1827 (320) 549 (300) 3966 (340) 688 (314) 355 (259)

>25 BACs 25–10 BACs 3–9 BACs 2 BACs

Singletons

35 52 60 38 61 70

308 422 581 272 379 544

(76) (78) (75) (81) (82) (81)

52 66 86 57 77 90

(78) (88) (106) (76) (91) (104)

113 (148) 152 (179) 165 (188) 120 (151) 146 (172) 167(186)

60 (68) 79 (90) 115 (126) 57 (70) 85 (95) 115 (124)

a

In all cases, the values in parentheses indicate the number of contigs obtained after DQing the original contigs. Note that the number of singletons remains unchanged after Dqing

G50 filtration and subsequently electrokinetically injected into a MegaBACE 1000 capillary sequencer (Amersham Biosciences). In each capillary, an internal ET-ROX labeled size standard and three different BAC samples were amplified with FAM-, JOE- and NED-labeled Hind III primers, respectively, and were injected simultaneously. The ET-ROX size standard was purchased from Amersham Biosciences.

Fingerprint analysis using BACXtractor Raw MegaBACE trace files (rsd) were processed using dedicated Keygene software (Xpose 2.0) for the removal of dye signal crosstalk and for size standard calling. All fingerprint traces were collected in the application BACXtractor (Keygene NV). Fingerprint peaks were scored under the following conditions: 20% signal rise, min 20% scale level. Traces were automatically filtered for quality based on the number of peaks found: too many or too few peaks suggest problems with PCR. Fragment size data were exported from BACXtractor to ‘‘bands’’ files for further analysis in FPC (see below). Since the larger AFLP fragments are less abundant than the smaller ones (as is to be expected on theoretical grounds), these fragments can be more reliably classified as identical. Besides, the chance of the occurrence of co-migrating, but non-homologous, fragments decreases with increasing fragment size, leading to more informative data for contig construction. Therefore, we decided to limit the exported data to a subset of fragments in the size range of 260–825 nt. Before export, all mobilities were multiplied by a factor of 10 to make the data compatible with FPC.

Contig assembly using FPC All contig construction was performed using the latest version of FPC (v6.2) downloaded from the Arizona Unversity Genomic Institute webpage (http://www.genome.arizona.edu/software/fpc/). To construct contigs, the FPC program parameters, e.g., tolerance and cutoff value, had to be determined empirically. The parameter ‘Tolerance’ refers to the maximum distance (measured in tenths of a millimeter) by which the mobilities of two bands from two different BAC clones may differ and still be considered as the same band. This was determined by viewing a set of clones that contained similar fingerprints. Tolerance values were varied and the effect on a BAC clone was visualized by comparing bands of that BAC clone with similar bands from other clones. A low tolerance value decreases the window size and increases the stringency of comparison. The fixed setting of the Tolerance parameter was applied. The cutoff score is the threshold value representing the maximum allowable probability of a chance match between any two BAC clones (also known as the ‘Sulston score’). The lower the Sulston score, the lower the probability that a match has arisen due to chance and the higher the probability that it represents a real

overlap. During contig assembly, to declare an overlap between two clones, approximately 50% of the bands need to be common to the two clones.

Q clones and the DQer function During assembly, a low stringency cutoff results in contigs with many questionable (Q) clones (i.e. many false positives); a high stringency cutoff results in too many contigs (i.e. many false negatives). The normal strategy employed, therefore, is an initial assembly of contigs using a moderate cutoff, followed by subsequent analysis at a lower cutoff. FPCv6.2 incorporates a function called DQer, which performs this function automatically. Contigs with Q clones above a certain user-defined cutoff are reassembled up to three times, and each time the cutoff is lowered by a factor of 10 (e.g. 1e-07, 1e-08, 1e-09). This assembly may result in one or more Contig Build (CB) maps. Sometimes during the DQing process, a contig with many Q clones is split into multiple contigs with reduced numbers of Q clones. Occasionally, a contig with many Qs may not change even when the cutoff is reduced by a factor of 3. This is indicative of a repetitive fingerprint.

Results and discussion AFLP-based BAC fingerprinting We have constructed a Hind III BAC library of P. pacificus var. California containing 13,440 clones, and generated 133 Single-Strand Conformation Polymorphism (SSCP) markers from BAC-end and EST sequences (Srinivasan et al. 2002). To generate a physical map of the genome of P. pacificus, we selected more than 9000 BAC clones from the Hind III library, including those BAC clones that had been used to generate the SSCP markers. All the BAC clones were AFLP fingerprinted using the primer combination Hind III/MseI (+0/+0; i.e., no selective bases in the AFLP reaction) (Fig. 1). The resulting fingerprints were digitized and scanned for quality control using the program BACxtractor (see Materials and methods for details), and scanned manually for repetitive elements as these produce excessive numbers of bands. In order to optimize contig assembly, a screening window of 260–825 bp was chosen to create the physical map from the AFLP data.

Number of BACs

491 233 182 172 167 152 131 125 121 118 112 108 107 106 102 93 92 91 90 87 83 80 75 62 61 61 59 59 58 55 52 48 47 47 46 45 44 44 43 42 42 41 40 40 40 39 39 38

Contig

Ctg350 Ctg278 Ctg370 Ctg299 Ctg362 Ctg337 Ctg343 Ctg365 Ctg319 Ctg297 Ctg314 Ctg336 Ctg369 Ctg2 Ctg298 Ctg5 Ctg309 Ctg367 Ctg8 Ctg308 Ctg338 Ctg262 Ctg361 Ctg315 Ctg3 Ctg28 Ctg33 Ctg268 Ctg303 Ctg323 Ctg1 Ctg296 Ctg18 Ctg272 Ctg305 Ctg35 Ctg274 Ctg366 Ctg322 Ctg263 Ctg316 Ctg26 Ctg66 Ctg36 Ctg14 Ctg292 Ctg273 Ctg261

888 680 560 816 540 648 580 596 340 556 420 544 368 548 416 576 464 432 224 592 364 464 360 308 388 288 408 396 336 192 412 272 240 188 308 332 264 232 268 324 216 264 316 228 192 444 340 404

Total length (kb) Ctg19 Ctg57 Ctg105 Ctg37 Ctg12 Ctg328 Ctg17 Ctg47 Ctg96 Ctg300 Ctg6 Ctg41 Ctg321 Ctg92 Ctg127 Ctg275 Ctg295 Ctg43 Ctg265 Ctg31 Ctg11 Ctg291 Ctg24 Ctg310 Ctg313 Ctg88 Ctg330 Ctg72 Ctg87 Ctg56 Ctg70 Ctg9 Ctg144 Ctg73 Ctg30 Ctg306 Ctg15 Ctg22 Ctg266 Ctg293 Ctg304 Ctg352 Ctg348 Ctg4 Ctg360 Ctg312 Ctg102 Ctg54

Contig 32 32 31 31 30 29 29 29 29 29 28 28 27 27 27 27 26 26 26 25 24 24 24 23 23 23 22 22 22 22 22 22 22 21 21 21 21 20 20 20 20 20 19 18 18 18 17 17

Number of BACs 204 132 264 136 220 272 228 196 192 124 188 188 232 224 132 132 324 180 124 160 244 148 124 208 204 188 284 196 184 168 148 144 136 284 204 188 172 240 228 168 164 160 116 148 140 132 188 172


Contig 16 16 16 15 15 15 15 15 15 15 14 14 14 14 14 14 13 13 13 12 12 12 12 12 12 12 12 12 11 11 11 11 11 11 11 11 11 11 11 11 10 10 10 10 10 10 10 10

Number of BACs 108 104 96 216 200 152 148 140 128 92 156 132 124 124 108 88 180 132 96 172 156 144 136 132 128 100 76 68 184 156 136 132 124 120 108 108 104 100 80 80 184 120 120 108 108 104 100 96


Contig 9 9 9 9 9 9 9 9 9 9 9 9 9 8 8 8 8 8 8 8 8 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 6 6 6 6 6 6 6 6 6 6 6 6

Number of BACs

Table 2 Automated contig assembly of the 7747 clones of the physical map at a Tolerance of 2 and a Cutoff value of 10-6

124 120 112 112 108 108 104 84 80 80 80 76 72 128 116 112 108 100 96 80 76 140 132 124 112 100 96 92 92 76 76 68 64 52 48 44 128 124 96 88 88 84 84 84 80 80 80 72


Contig 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 3 3 3 3

Number of BACs 112 96 92 88 88 76 76 72 68 56 52 52 52 48 40 116 92 84 84 84 84 80 80 76 76 76 72 72 68 68 68 64 60 56 52 52 48 44 44 44 44 36 36 36 88 84 84 76

Total length (kb)

719

76 76 72 72 72 64 64 60 60 3 3 3 3 3 3 3 3 3 Ctg130 Ctg136 Ctg159 Ctg187 Ctg335 Ctg142 Ctg170 Ctg183 Ctg206 64 64 52 52 48 48 120 120 112 6 6 6 6 6 6 5 5 5 Ctg116 Ctg357 Ctg147 Ctg161 Ctg190 Ctg289 Ctg80 Ctg200 Ctg133 92 92 84 76 64 52 52 132 124 10 10 10 10 10 10 10 9 9 Ctg58 Ctg122 Ctg83 Ctg329 Ctg355 Ctg99 Ctg320 Ctg89 Ctg95 17 17 17 17 16 16 16 16 16 Ctg27 Ctg38 Ctg82 Ctg327 Ctg94 Ctg85 Ctg106 Ctg52 Ctg71

152 124 116 112 192 160 144 128 112

Number of BACs

38 38 38 37 35 34 33 32 32 Ctg20 Ctg25 Ctg13 Ctg340 Ctg16 Ctg281 Ctg276 Ctg7 Ctg267

236 228 184 396 216 180 136 256 220

Number of BACs

Total length (kb)

The first step in contig assembly involved the collection of DNA fingerprints for more than 9000 BAC clones. The overall efficiency of the AFLP fingerprinting procedure was 75%. Of the 25% failures, 12% were due to poor-quality DNA, 9% to rDNA and empty clones, 3% to MegaBACE errors, and 1% to failures in the AFLP reaction. On average, 25 AFLP bands per BAC clone were used for fingerprint analysis. Following image analysis, 7747 BAC clones were assembled using FPC v6.2 (Soderlund et al. 1997, 2000; Chen et al. 2002; Cone et al. 2002) to generate a physical map of the P. pacificus genome (Fig. 2 and Table 1). To optimize contig assembly, tolerance and cutoff values for map generation had to be determined empirically (see Materials and methods for details). Tolerance values were varied from 1 to 4 and the effect of the change was visualized by comparing bands of one clone with similar bands of other clones. From this analysis, a fixed tolerance of 2 was selected for the construction of the P. pacificus genome map. To determine the ‘Sulston score’, the cutoff value was varied and the effect on a contig that had been confirmed by hybridization on a BAC filter was examined. Changes in cutoff value beyond a certain threshold result in the contig breaking up into smaller contigs. Based on this criterion, a cutoff value of 10)6 was chosen. The results of the automated contig assembly at various tolerance and cutoff values are shown in Table 1. For further analysis, we selected a tolerance of 2 and a cutoff value of 10)6, and Table 2 lists the contigs obtained under these conditions and their sizes. The automated contig assembly generates a number of contigs containing questionable (Q) clones. In FPCv6.2, the questionable clones are automatically reanalyzed by a new function DQer (see Materials and methods for details). Table 1 shows the results of the contig assembly procedure before and after DQing for different cutoff values. In total, 7747 clones were assembled into 260 contigs before DQing, and 370 contigs with an average of five BAC clones per contig emerged after DQing. The latter assembly provided the data shown in Table 2. Verification of physical map data by comparison with colony hybridization data

Contig

Table 2 (Contd.)

Assembly of the BAC fingerprints

Contig

Total length (kb)

Contig

Number of BACs

Total length (kb)

Contig

Number of BACs

Total length (kb)

Contig

Number of BACs

Total length (kb)

720

The contigs generated by FPC v6.2 were verified by checking the hybridization data acquired for different BAC-end fragments and SSCP markers. In nearly all cases, hybridization results confirmed the fingerprintbased contig assembly. Specifically, we hybridized a total of 104 ESTs, gene or BAC-end fragments to the BAC library. In approximately 80% of these experiments, we observed additional clones in the hybridization experiments besides the clones predicted from

721

Fig. 3 An integrated genetic and physical map of P. pacificus. SSCP markers are spaced more or less evenly throughout the genome. Note that the six chromosomes of P. pacificus have different genetic lengths and that they are drawn at different scales. Scales for each chromosome are highlighted at the top of each chromosome

the physical map. This is due to the fact that only 58% of the BAC library has been successfully fingerprinted. Hybridization experiments should also allow fusion of certain contigs. In a few cases, BAC walking by hybridization linked together individual contigs that could not be assembled by FPC due to limited overlap between the terminal BAC clones (H. Xiao, M. Zheng and R. J. Sommer, unpublished observation). In a small set of hybridization experiments, we found multiple signals due to the fact that some hybridization probes contained repetitive DNA. Most of these BAC signals were from the larger contigs containing more than 100 individual BAC clones. Finally, in seven cases, no BAC clones were detected after hybridization with gene fragments indicating the presence of some gaps in the BAC library. To overcome this problem, we are currently constructing a second BAC library using a different restriction enzyme. Taken together, these data indicate that the AFLP fingerprint-based physical map provides adequate coverage of the P. pacificus genome. Several experiments have suggested that the P. pacificus genome is approximately 100 Mb long (Sommer et al. 1996). On average, the Hind III BAC library should provide ten-fold coverage of the genome (Srinivasan et al. 2002), so that the fingerprinting of 58% of the BAC clones results in a 5–6· coverage.

Integration of the genetic and physical maps The genetic linkage map of P. pacificus was generated using SSCP markers from BAC-end and EST sequences (Srinivasan et al. 2002), and the physical map of the P. pacificus genome was assembled using AFLP fingerprints of individual BAC clones. Thus, many of the SSCP markers on the genetic map can be anchored to the physical map and vice versa. Incorporating the SSCP markers into the P. pacificus physical map resulted in the identification of 53 contigs containing one or more SSCP markers (Fig. 3). Specifically, 48 of these 53 contigs have one marker, whereas the remaining five have more than one marker (Fig. 3). In those cases in which two or more SSCP markers are in the same physical contig, the positions of these markers on the genetic linkage map are identical (data not shown). Most of the SSCP markers that were successfully linked to individual contigs are from BAC ends rather than from ESTs. Our data suggest that the BAC-endderived markers are more or less evenly distributed on the genetic map. The physical map assembly does not identify large clusters of BACs in any one region. In addition, the size distribution of the fingerprinted contigs, i.e., the size of the region covered and the number of constituent clones, suggested that the physical map coverage is quite even. These results are consistent with an even distribution of the physical map contigs throughout the genome of P. pacificus (Fig. 3). This is an important finding, and not a trivial observation. In C. elegans, the genome shows several interesting features that are common to the five autosomes at least; i.e., there are higher rates of recombination and higher levels of heterochromatin and repetitive elements in the distal

722

arms of the chromosomes than in the central regions (Barnes et al. 1995). In principle, such a scenario could bias the coverage of specific regions in genomic libraries, and hence also in a BAC-based physical map. Thus, in P. pacificus, the absence of central clustering of SSCPtagged contigs argues in favor of an equal representation of the BAC library and the physical map. However, it should also be noted that the even distribution of the SSCP markers might result from the fact that most of these markers are in randomly chosen BAC ends rather than in ESTs, microsatellites or any other potentially uneven distributed DNA fragments. Genomics in non-model organisms The generation of integrated genetic and physical maps provides a framework for large-scale genome analysis. Currently, most biological research is focused on studies in model organisms, such as Arabidopsis thaliana, C. elegans , Drosophila or mouse. In all these research communities, the generation of integrated genetic and physical maps provided the necessary framework for map based cloning and eventually, for genome sequencing (The C. elegans Sequencing Consortium 1998; Marra et al. 1999; Waterston et al. 2002). More recently, genomic approaches have also been applied in other areas of biology, such as evolutionary biology. For example, the generation of a genetic linkage map in stickleback has opened new lines of research towards the analysis of morphological evolution in this group of fishes (Peichel et al. 2001). Similarly, the generation of a genetic linkage map for the nematode species P. pacificus has facilitated mapping of mutants in this organism (Srinivasan et al. 2002). However, to provide a complete genomic framework for molecular studies in this organism, the integration of genetic and physical maps is a necessary prerequisite. This study provides the first detailed genome map for a non-model organism in the field of evolutionary developmental biology. Acknowledgements We thank G. Otto, A. Pires-daSilva and K.-Z. Lee for comments on the manuscript. The work described in this manuscript was financially supported by the Max-Planck Society. R. J. S. is a Max-Planck Society Investigator.

References Barnes TM, Kohara Y, Coulson A, Hekimi S (1995) Meiotic recombination, noncoding DNA and genomic organization in Caenorhabditis elegans. Genetics 141:159–179 Cao Y, et al (1999) A 12-Mb complete coverage BAC contig map in human chromosome 16p13.1–p11.2. Genome Res 9:763–774 Chen M, et al (2002) An integrated physical and genetic map of the rice genome. Plant Cell 14:537–545 Cone KC, et al (2002) Genetic, physical, and informatics resources for maize. On the road to an integrated map. Plant Physiol 130:1598–1605 Eizinger A, Sommer RJ (1997) The homeotic gene lin-39 and the evolution of nematode epidermal cell fates. Science 278:452–455 Eizinger A, Jungblut B, Sommer RJ (1999) Evolutionary change in the functional specificity of genes. Trends Genet 15:197–202

Felix MA, De Ley P, Sommer RJ, Frisse L, Nadler SA, Thomas WK, Vanfleteren J, Sternberg PW (2000) Evolution of vulva development in the Cephalobina (Nematoda). Dev Biol 221:68–86 Jungblut BA, Sommer RJ (1998) The Pristionchus pacificus mab-5 gene is involved in the regulation of ventral epidermal cell fates. Curr Biol 8:775–778 Jungblut B, Pires-daSilva A, Sommer RJ (2001) Formation of the egg-laying system in Pristionchus pacificus requires complex interactions between gonadal, mesodermal and epidermal tissues and does not rely on single cell inductions. Development 128:3395–3404 Lee KZ, Eizinger A, Nandakumar R, Schuster SC, Sommer RJ (2003) Limited microsynteny between the genomes of Pristionchus pacificus and Caenorhabditis elegans. Nucleic Acids Res 31:2553–2560 Marra M, et al (1999) A map for sequence analysis of the Arabidopsis thaliana genome. Nat Genet 22:265–270 Marra MA, Kucaba TA, Dietrich NL, Green ED, Brownstein B, Wilson RK, McDonald KM, Hillier LW, McPherson JD, Waterston RH (1997) High throughput fingerprint analysis of large-insert clones. Genome Res 7:1072–1084 Peichel CL, Nereng KS, Ohgi KA, Cole BL, Colosimo PF, Buerkle CA, Schluter D, Kingsley DM (2001) The genetic architecture of divergence between threespine stickleback species. Nature 414:901–905 Pires-DaSilva A, Sommer RJ (2003) The evolution of signalling pathways in animal development. Nat Rev Genet 4:39–49. Schlak I, Eizinger A, Sommer RJ (1997) High rate of restriction fragment length polymorphisms between two populations of the nematode Pristionchus pacificus (Diplogastridae). J Zool Syst Evol 35:137–142 Simpson P (2002) Evolution of development in closely related species of flies and worms. Nat Rev Genet 3:907 Soderlund C, Longden I, Mott R (1997) FPC: a system for building contigs from restriction fingerprinted clones. Comput Appl Biosci 13:523–535 Soderlund C, Humphray S, Dunham A, French L (2000) Contigs built with fingerprints, markers, and FPC V4.7. Genome Res 10:1772–1787 Sommer RJ (2000) Evolution of nematode development. Curr Opin Genet Dev 10:443–448 Sommer RJ (2001) As good as they get: cells in nematode vulva development and evolution. Curr Opin Cell Biol 13:715–720 Sommer RJ, Sternberg PW (1996) Evolution of nematode vulval fate patterning. Dev Biol 173:396–407 Sommer RJ, Carta LK, Seong-youn Kim A, Sternberg PW (1996) Morphological, genetic and molecular description of Pristionchus pacificus sp. n. (Nematoda: Neodiplogastridae). Fund Appl Nematol 19:511–521 Sommer RJ, Eizinger A, Lee KZ, Jungblut B, Bubeck A, Schlak I (1998) The Pristionchus HOX gene Ppa-lin-39 inhibits programmed cell death to specify the vulva equivalence group and is not required during vulval induction. Development 125:3865–3873 Srinivasan J, Pires-daSilva A, Gutierrez A, Zheng M, Jungblut B, Witte H, Schlak I, Sommer RJ (2001) Microevolutionary analysis of the nematode genus Pristionchus suggests a recent evolution of redundant developmental mechanisms during vulva formation. Evol Dev 3:229–240 Srinivasan J, et al (2002) A Bacterial Artificial Chromosome-based genetic linkage map of the nematode Pristionchus pacificus. Genetics 162:129–134 The C. elegans Sequencing Consortium (1998) Genome sequence of the nematode C. elegans: a platform for investigating biology. Science 282:2012–2018 Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Hornes M, Frijters A, Pot J, Peleman J, Kuiper M, Zabeau M (1995) AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res 23:4407–4414 Waterston RH, et al (2002) Initial sequencing and comparative analysis of the mouse genome. Nature 420:520–562