Isolation and characterisation of somatic hybrids between Lycopersicon esculentum and Lycopersicon peruvianum

English. Several desirable traits, such as disease resistances, have been introduced from more or less related wild Lycopersicon species, into the cultivated tomato, L.esculentum , by classical breeding techniques. Somatic hybridisation by protoplast fusion can enhance the germplasm pool available for tomato breeding, because this procedure allows to by-pass sexual crossing barriers. Especially the technique of asymmetric hybridisation, by which untreated recipient protoplasts are fused with donor protoplasts of which the major part of the genome has been eliminated, for instance by irradiation, might be very useful. The resulting asymmetric hybrids will contain only a small fraction of the donor genome, and, therefore, the number of backcrosses required to eliminate undesired donor traits might be relatively small; it might also be possible to breed at the (near) diploid level with such hybrids. Asymmetric somatic hybrids might also be useful for chromosome mapping. In this thesis several aspects of somatic hybridisation that are important for the application of this technique to the improvement of the tomato, were analysed in detail.The efficient isolation of symmetric somatic hybrids between L.esculentum and the wild species L.peruvianum by means of a double selection strategy is described in Chapter 2. The symmetric hybrids were selected on the basis of kanamycin resistance of L.esculentum and the superior regeneration capacity of L.peruvianum . These hybrids were very vigorous plants; most of them had a tetraploid chromosome number of 48 (2n = 4x). A minority of the hybrids was at the hexaploid level with chromosome numbers from 64 to 72. The hybrid nature of all regenerated plants was confirmed by analysis of isozyme markers, by their intermediate morphology and, in some cases, by the analysis of restriction fragment length polymorphisms (RFLPs). According to RFLP analysis 6 hexaploid hybrids that were tested all contained one diploid genome of L.esculentum and two diploid genomes of L.peruvianum . One of these hexaploids had genomes of two different L.peruvianum genotypes and was therefore considered to be derived from a triple protoplast fusion. The hexaploid plants resembled L.peruvianum more than the tetraploids did. The hybrids did flower and set fruits.The fertility of the tetraploid and hexaploid somatic hybrids was analysed after selfing and after backcrossing to each of both fusion parents (Chapter 3). Most of the somatic hybrids, especially the tetraploids, were fertile upon selfing and yielded many seeds, of which 79% germinated. These progeny plants were vigorous and often fertile after selfing. Backcrossing of the somatic hybrids with the L.esculentum parent did not yield any viable seeds; backcrossing with L. peruvianum yielded a few germinating seeds, but only if L. peruvianum was used as staminate parent. The plants derived from the backcross hexaploid hybrid x tetraploid L. peruvianum had a pentaploid chromosome number (2n - 5x - 60) and were vigorous, whereas the plants derived from the backcross tetraploid hybrid x diploid L. peruvianum grew retarded. To obtain information about the behaviour of chromosomes in the hybrids, the progenies of the selfings and backcrosses were analysed for the segregation of several traits, namely kanamycin resistance, isozyme patterns for acid phosphatase, locus Aps-1 , and some morphological characteristics. Most of the progenies segregated for kanamycin resistance as expected on the basis of the number of loci present in the kanamycin resistant fusion parent, whereas in some progenies the fraction of kanamycin resistant plants was smaller than expected. The segregation of the Aps-1 isozyme patterns and the variation for some of the morphological characteristics among the progeny plants indicated a tetrasomic inheritance of at least part of the genes in the hybrids.Asymmetric somatic hybrids between L. esculentum and L. peruvianum were selected on the basis of the regeneration capacity of L. peruvianum (Chapter 4). Selection against growth of L. peruvianum was effected by lethal doses (50, 300 or 1000 Gray) of gamma-irradiation applied to this species before protoplast fusion. The asymmetric hybrids needed a longer regeneration time than the symmetric hybrids, and showed a large variation in callus morphology, plant morphology and chromosome numbers. All asymmetric hybrid plants were aneuploid. After a low dose (50 Gy) most hybrids had a near-triploid chromosome number, whereas after a high dose (300 or 1000 Gy) most of the hybrids were near pentaploid. In spite of the aneuploid chromosome numbers, many plants grew vigorously, flowered and, in some cases, set fruits. In general the morphology of the asymmetric hybrids was intermediate between that of L. esculentum and symmetric somatic hybrids of both species, and approached the morphology of L. esculentum more after a high dose of irradiation. The high dose hybrids also grew wore slowly, flowered and set fruits less than the low dose hybrids. No viable seeds could be obtained from any asymmetric hybrid.The asymmetric hybrids were also analysed for the retention of specific genes and alleles of the donor L. peruvianum (Chapter 5). About 50% of the asymmetric 30H-hybrids (L. peruvianum parent irradiated with 300 Gy) had retained the kanamycin resistance trait from donor plants with one hemizygous resistance locus. L. peruvianum specific alleles of the isozyme markers Aps-1 and glutamate oxaloacetate transaminase were present in at least 70% of the hybrids; the retention of donor alleles was lower in 30H-than in 5H-hybrids (donor irradiated with 50 Gy). On the average, 74% of the L. peruvianum specific alleles (one or both) of 18 morphological markers, which were located on 10 of the 12 tomato chromosomes, were detected in the 30H-hybrids. It was estimated that each allele of a given marker was, on the average, present in half of the 30H-hybrids. In 36% of the 30H-hybrids, only one of both recessive morphological markers that were located on a same L. esculentum chromosome was complemented by the corresponding L. peruvianum allele. This is an indication for frequent breakage of the L. peruvianum chromosomes. Several hybrid calli regenerated genotypically different plants, which suggested that some segregation occurred in these calli before shoot regeneration.Fifteen asymmetric hybrids, eight 5H- and seven 30H-hybrids, were analysed with 29 RFLP markers (Chapter 6); each tomato chromosome was represented by at least two such markers. Retention of alleles from the irradiated L. peruvianum donor genome in the asymmetric hybrids ranged from 31% to 83%. On the average, 67% of the 58 alleles of the 29 RFLP loci was present in the 5H- hybrids and 53% in the 30H-hybrids. The 5H-hybrids contained more complete L. peruvianum chromosomes, as determined by the retention of RFLP alleles of the loci on one chromosome of the donor, than the 30H-hybrids, whereas the 30H-hybrids contained more incomplete chromosomes; this indicated a more frequent breakage of L. peruvianum chromosomes in the 30H-hybrids. In most hybrids some L. esculentum alleles were lost. Three L.peruvianum loci, located on chromosome 2, 4 and 7, respectively, were present in each asymmetric hybrid, which may suggest linkage to the regeneration capacity trait which was used in selecting the asymmetric hybrids. The asymmetric hybrids were also analysed with a probe for ribosomal DNA (rDNA). The amount of rDNA from L. peruvianum retained in the hybrids varied strongly: in one hybrid amplification had occurred, whereas in others L. peruvianum rDNA was either absent or (in)completely present. The amount of L. esculentum specific rDNA was decreased in most asymmetric hybrids.The analyses of the asymmetric somatic hybrids showed that irradiation before fusion eliminated the L. peruvianum genome only to a limited extent. In addition, the hybrids were sterile despite their ability to flower and set fruits. It can be concluded therefore that application of these hybrids in breeding programs is very restricted. However, fertile progeny plants derived from selfed symmetric hybrids might be used for this purpose, because segregation for L. esculentum specific traits was observed. The use of the asymmetric hybrids for somatic mapping is not favourable, because of the limited elimination of the donor genome, the large number of incomplete donor chromosomes and the probable presence of rearranged chromosomes. It is possible that highly asymmetric hybrids can be obtained by the use of other selectable markers or other procedures to induce elimination of the donor genome.

Dutch; Flemish. In dit proefschrift zijn experimenten beschreven, waarin wordt nagegaan wat de mogelijkheden en moeilijkheden zijn van de overdracht van eigenschappen uit een wilde soort naar de cultuurtomaat m.b.v. protoplastenfusie. De volgende aspecten, die van belang zijn voor de toepassing van de protoplastenfusie bij de veredeling van de tomaat, zijn nader onderzocht: het ontwikkelen van een doeltreffende selectiemethode voor de isolatie van somatische hybriden; het isoleren van asymmetrische somatische hybriden, waarbij de wilde soort bestraald wordt voor de fusie; en het karakteriseren van de hybriden om te bepalen wat de bijdrage is van de (bestraalde) wilde soort aan het genoom van de hybriden