OBJECTIVE: the introduction of molecular based tagging of genes for use in a breeding program at San Carlos University.
Luis Mejia, Amilcar Sanchez, and Carolina Zea, San Carlos University; D. P. Maxwell, M. Salus, and C. T. Martin, University of Wisconsin; F. Vidavski, Hebrew University, and Valerie Williamson, University of California-Davis
(Also, see Two step protocol for detection of Mi gene in problem germplasm)
Several options for molecular gene tagging were considered originally. The Volcani Institute group had limited success with the PCR markers for Verticillium wilt resistance and Hebrew University was using the REX markers for the nematode resistance gene (Mi). It was decided that this marker would be use to introduce these methods into the biotechnology curriculum at San Carlos University. The technique involved PCR and restriction enzyme digestions, which could be performed in Dr. Mejia’s laboratory at San Carlos University. Also, it would be very useful to develop breeding lines for Guatemala that had both resistance to begomoviruses and root knot nematode. This has not been possible with the Ty-1 gene, since this gene for TYLCV resistance is very closely linked to the Mi gene.
Primers REXF1/REXR2 (Chromosome 6) – REX locus
The REX primers amplify a molecular marker linked to Mi gene (root knot nematode resistance) in Chromosome 6 (Williamson et al. 1994. TAG 87:757-763). This primer pair amplifies a single band of around 750 bp from tomato breeding lines and hybrids. The REX fragment is then digested with TaqI restriction enzyme, and if a TaqI site (TCGA) is present, the REX fragment is linked to the Mi gene (ie an introgression from S. peruvianum). This gives a digestion of two fragments, 570 and 160 bp. The esculentum sequence does not have the sequence associated with the TaqI restriction site. A heterozygous plant would give three bands after digestion.
An M. Sc. student, Carolina Zea, at San Carlos University did her thesis research on the application of this marker for evaluation of germplasm in the breeding lines being developed in this CDR project. A cross was made between Marina, a commercial cultivar with the N gene (Mi gene) and the hybrid H5 (G1h x G10p) giving population H44. This H44 population was evaluated in the field in March 03 and plant 29 was resistant to begomoviruses. This plant was identified by the REX marker in Guatemala as being heterozygous for the Mi gene. One potential problem with digestions is the possibility that incomplete digestions could produce erratic results. In order to verify digestion results, several seedling progenies (F3 plants) were produced in Guatemala and the DNA from these plants was brought to Wisconsin and amplified with the REX primers. The amplified fragments were sent for direct sequencing, using the REXR2 primer. Sequence analysis of the REX PCR fragment clearly distinguished among the three classes of plants, homozygous for the TaqI site, heterozygous, and homozygous for no TaqI site. The presence of the TaqI site corresponds to the resistance gene marker and is associated with a SNP for A/C, T for Mi gene and G for Mi+ gene or L. esculentum phenotype (susceptible). A commercial hybrid, Better Boy, reported to have the Mi gene (N) was heterozygous (A/C), Gh1 and Gh2 were homozygous for the Mi gene marker (A/A), and Gh3, HC7880, M82 and L. esculentum var. cerasiforme were homozygous for the Mi+ gene (susceptible allele) marker (C/C).
At the Tomato Round Table Meeting in Guatemala in 2002, Dr. Judith Milo from Hebrew University of Jerusalem reported that the REX locus could be used to tag the Ty1 gene (chilense) for resistance to TYLCV. So we wanted to compare the REX locus sequence for a line with Ty1 gene (TY52), a line with S. habrochaites (L. hirsutum) (902), a control cultivar that is homozygous for the resistance Mi gene (Mi/Mi, Motelle) and the susceptible lines (Mi+/Mi+, Moneymaker and M82). Dr. Valerie Williamson supplied seeds of Motelle and Moneymaker, Dr. Favi Vidavski seeds of the 902 breeding line and M82, and Dr. H. Czosnek seeds for the TY52 and TY50 from D. Zamir. The TaqI site (CAPS marker) was evident in Motelle (TaqI site present) and restriction site was lacking in the REX fragment for the susceptible line (M82) (Fig. 5). The susceptible line had a SNP at this position and the sequence was C instead of A, as was present in Motelle (Fig.1). This result is consistent with the expected results for the TaqI digestions (Williamson et al. TAG 87:757). The three presumably susceptible lines, M82, TY50 and L. esculentum var. cerasiforme had the identical sequences for this 600 nt region. Four lines, Motelle, TY52, G2h, and 902, had the A nt for this TaqI site SNP. This was unexpected as TY52 and G2h and 902 were thought to be susceptible to root knot nematode. These three lines were subjected to a standard bioassay for root knot resistance. TY52 and 902 were found to be susceptible in the bioassay conducted at Hebrew University of Jerusalem (F. Vidavski, pers. com.) and Gh2 was resistant to root knot nematode at UC-Davis (V. Williamson, pers. com.). Thus, in two cases the presence of the TaqI SNP (A) gave false positive results. In this 600-nt region of the REX fragment, there are nine SNPs between the three susceptible lines (TY50, M82 and cerasiforme) and the three lines Motelle, Ih902, and Gh2. However, only seven of these SNPs were shared by TY52 and Motelle. There were four SNPs that could distinguish TY52 from Motelle. It was concluded that three types of sequences could be recognized by the unique SNPs for the REX locus, ie, esculentum type (M82), peruvianum type (Motelle), and chilense type (TY52). Dr. Valerie Williamson noted that the chilense-type sequence had two TaqI sites, and thus would give three fragments when digested with TaqI restriction enzyme. SNPs associated with the peruvianum-type sequence were easily detected in commercial hybrids (heterozygous for the REX locus) such as Better Boy and Marina.
Because Ih902 is the main source of resistance to begomoviruses being used in Guatemala and also the TOMATO MERC countries (Palestine, Lebanon, Jordan, Egypt, Morocco, and Tunisia), it was decided that a new PCR-based method for detection of the Mi gene should be developed. Dr. S. Tanksley, Cornell University, had developed a PCR-based method (CAPS) for detection of the Mi gene and it is protected by intellectual property by the Cornell University Foundation. Dr. Maxwell negotiated with the Cornell University Foundation to evaluate the Tanksley PCR-based method (Cornell method). He was given permission to evaluate the marker for research only and that they could not be used in a breeding program. The Cornell method was evaluated with the same germplasm as had been tested for the REX primers. As with the REX locus CAPS marker, the Cornell method gave false positives for TY52 and 902. The expected results for linkage to the Mi gene were obtained with Motelle and G2h. The PCR fragments from the Cornell primer pair were sequenced and as before the TY52 could be distinguished from Motelle and M82 by several SNPs, thus the Cornell PCR fragment could distinguish between chilense-type, peruvianum-type and esculentum-type sequences. The surprise was that Ih902 had the chilense-type sequence and not the peruvianum type sequence. Unfortunately, this information can not be used to develop a specific PCR-based system, which would distinguish the three types of sequences, because of the intellectual property agreement with Cornell University Foundation.
Dr. Valerie Williamson’s group cloned the Mi gene region and sequenced a BAC clone (U81378) and the sequence for two genomic regions for the Mi1.2 and Mi1.1 gene (Plant Cell 10:1,307-1,319). Several primers specific to the promoter sequence of the functional gene, Mi.12, were designed and tested. From this effort, primer Mi12F1 (5’ gca att cta gat cta gct att tgt tgt tc 3’) and Mi12R2 (5’ cct gct cgt tta cca tta ctt ttc caa cc 3’) gave the most encouraging results. This primer pair gave a 720 bp and 620 bp fragment with Motelle (resistant) and Moneymarker (susceptible), respectively. For the heterozygous cultivars, Better Boy and Marina, two fragments (620 and 720 bp) were obtained. These data are consistent with the results from the REX locus and the Cornell locus. Six other lines that are thought to be susceptible to room knot nematode all gave fragments of 620 bp. Gh2, which is resistant, gave a 720 bp fragment. Line Ty52, which gives false positives with the REX locus and Cornell locus gave a 1,000-bp fragment. Also, Gc9, which as the chilense-type sequence for REX locus and Cornell locus, had a 1000-bp fragment. These PCR primers should be useful to follow the TY1 gene as it gives a distinct fragment size from the esculentum- and peruvianum-type fragments. The fragment from S. habrochaites (L. hirsutum) was 800 bp. Line Ih902 gave the peruvianum-type fragment, 720 bp, which would be a false positive. These fragments were sequenced and compared to the published sequence for the BAC clone in GenBank. Surprisingly, the sequences from the 620-bp or the 720-bp fragments did not match sequences of this BAC clone, which means that a different region of the tomato genome had been amplified. These two sequences were sent to Dr. Valerie Williamson (UC-Davis) and her group matched the sequence obtained from Motelle (peruvianum type) with Mi1.4 in the 2p group of Mi genes in this same region and the sequence from Moneymaker (esculentum type) with the Mi1F in the 1e region. Thus, the sequences did match those in the region of the Mi gene clusters of Chromosome 6, but they did not match the original region (promoter of Mi1.2) used to design the primers. In June 2004, Seah et al. (2004. TAG 108:1635-1642) reported that the nematode-resistant region has the Mi-1 gene and 6 other homologs, which are grouped into two clusters (four in one and three in the other) and are separated by 300 kb. The REX locus is located in this 300-kb region. The organization of this Mi gene region in susceptible tomato was also report by Seah et al. (2004), and they suggested that there has an inversion of the 300-kb region between the two clusters. Also, they found two additional Mi-1 homologs that mapped to chromosome 5.
Lines that were presumably susceptible gave either the 620-bp fragment (cerasiforme, M82, Gh3, Gc16, TY50) or the 1000-bp fragment (TY52 and Gc9). The 620-bp fragments were sequenced except for M82 and they all were identical to the esculentum-type sequence for Moneymaker. The 720-bp fragment, characteristic of the peruvianum-type sequence from Motelle, was for Gh1 gave the peruvianum-like sequence.
In conclusions, both the CAPS markers for the REX locus and the Cornell locus give false positives with Ih902 and TY52. When the fragments are sequenced, the peruvianum- and chilense-type sequences can be identified by unique SNP patterns. The Mi12F1/Mi12R2 primers give four different size fragments for each type, esculentum- (620 bp), peruvianum- (720 bp), hirsutum- (800 bp), and chilense-type (1,000 bp). Line Ih902 appears to be chimeric in this region with sequences both from chilense (Cornell locus) and peruvianum (REX and Mi12F1/Mi12R2 loci) and this may explain the difficulty in developing a PCR-based method to distinguish 902 from Motelle. As indicated above this is important since most of the resistance to begomoviruses is being derived from 902.
One last point, Gh2, which was selected from FAVI 12, has both resistance to root knot nematode and begomoviruses and has the excepted sequence and PCR fragments for the Mi-1 gene. This is an excellent line to use to incorporate resistance for both begomoviruses and root knot nematode into the various programs in the TOMATO MERC project.