The term “somaclonal variation” has been used to refer to the variation which has been observed in plants regenerated from cell and tissue cultures, but in general not from axillary bud or shoot tip cultures. As pointed out by Karp (1989), somaclonal variation is a phenomenon of broad taxonomic occurrence, reported for species of different ploidy levels, and for outcrossing and inbreeding, vegetatively and seed propagated, and cultivated and non-cultivated plants. Characters affected include both qualitative and quantitative traits.
Pre-existing intercellular variation may be a contributing factor in some cases, but many instances clearly involve variation induced during culture. At the molecular level, evidence has been presented for the involvement of many phenomena, including gross alterations in chromosome number and structure, point mutations, mitotic recombination, and the amplification, deamplification, deletion, transposition or methylation of DNA sequences in nuclear, mitochondrial or chloroplast genomes (Larkin 1987, Evans 1989, van den Bulk 1991, Brown 1991, Karp 1989). Repetitive sequences may have a general role in somaclonal variation. Anomalies which seem peculiar to somaclonal populations are the occasional occurrence of homozygous variants (perhaps due to mitotic crossing over), and instances of directed changes resulting in population shifts (Larkin 1987, Karp 1989). Examples of the latter include reports of whole population shifts towards resistance to diseases caused by agents such as Helminthosporium sacchari in sugarcane and Verticillium albo-atrum in alfalfa (van den Bulk 1991). In general though, the frequency of variants displaying a desired trait is very low. The chromosomal or molecular changes may result in stable alterations which are transmitted to sexual progeny, but variants frequently are not stable, particularly through meiosis. Some of the unstable changes may be due to the activation of transposable elements, while some may have a basis in DNA methylation (Karp 1989). Many instances of selection of variant cell lines concern epigenetic variation - adaptations to selection pressure which are reversed when the pressure is removed.
The incidence of somaclonal variation is influenced by genotype, by ploidy level (polyploids giving rise to greater variation), tissue source, culture time and procedure (Larkin 1987, Karp 1989). Contrary to some early suppositions though, protoplasts are probably not more predisposed to somaclonal variation than cells (Larkin 1987).
Traits with respect to which somaclonal variants have been detected include:
Variation induced with respect to resistance to disease, including fungal, bacterial and viral diseases, was well reviewed by Van den Bulk (1991). That author's listing of crops/diseases for which variants displaying increased resistance have been selected is reproduced below in Table 3.
Table 3. Somaclonal Variants for Disease Resistance
| Crop | Pathogen or disease |
| Rice | Helminthosporium oryzae, Xanthomonas oryzae |
| Wheat | Helminthosporium sativum, Pseudomonas syringae |
| Barley | Rhynchosporium secalis, Helminthosporium sativum |
| Oats | Helminthosporium victoriae |
| Maize | Helminthosporium maydis |
| Sugarcane | Helminthosporium sacchari, Sclerospora sacchari, Ustilago scitaminea, Puccinia melanocephala, Fiji disease |
| Tobacco | Phytophthora parasitica, Pseudomonas syringae, P. solanacearum, tobacco mosaic virus |
| Eggplant | little leaf disease |
| Potato | Alternaria solani, Phytophthora infestans, Streptomyces scabies, potato virus X, potato virus Y, potato leaf roll |
| Tomato | virus Fusarium oxysproum, Pseudomonas solanacearum, |
| HOp | tomato mosaic virus |
| Alfalfa | Verticillium albo-atrum Verticillium albo-atrum, Fusarium solani, F. oxysporum |
| Celery | Fusarium oxysporum Fusarium oxysporum, Septoria apii, Cercospora apii, |
| Lettuce | Pseudomonas cichorii |
| Rape | Bremia lactucae, lettuce mosaic virus |
| Banana | Phoma lingam, Alternaria brassicicola |
| Peach | Fusarium oxysporum |
| Poplar | Xanthomonas campestris, Pseudomonas syringae Septoria musiva, Melampsora medusae |
A forest tree species not included in the above review is Larix decidua, for which somaclones resistant to Gremmeniella abietina have been identified, although performance of regenerated plants was not reported (Skilling et al. 1983). The majority of the citations in Table 3 involve selection at the whole plant level, after regeneration from culture. Many also involve in vitro selection. For about half of the cases reported, resistance has been inherited in a stable manner. Mode of inheritance has been established in only a few cases, but Mendelian, quantitative and maternal inheritance have all been implicated (Van den Bulk 1991). Heritability of resistance has not been established in many cases. In some cases, it has been shown that plants selected for resistance were changed in other, unwanted traits as well. Many reports exist also of unsuccessful attempts to select regenerants with increased resistance.
Because of the simplicity of adding herbicides to culture media, herbicide resistance was an early target of somaclonal research. The review by Hughes (1983) demonstrates that cell lines and/or plants have been selected for resistance to most types of herbicide:
Plants regenerated from cultures selected at the cell stage may or may not be resistant. In greenhouse trials, somaclonal variants of poplar hybrids have demonstrated tolerance to glyphosate and sulfometuron methyl (Michler and Haissig 1988), in the latter case at ten times the concentration required to kill the original clones (Michler 1988). Sexual transmission of the herbicide resistance has been demonstrated for picloram (tobacco), paraquat (tobacco and tomato), amitrole (tobacco) and chlorsulfuron and sulfometuron methyl (tobacco) (Chaleff 1986b). Where investigated, herbicide tolerance often has been shown to be a genetically simple trait - dominant and semidominant nuclear allelles at one to three loci for resistance to paraquat in tomato and to picloram, chlorsulfuron and sulfometuron methyl in tobacco (Chaleff 1986b). Atrazine resistance, evident in plants regenerated from resistant cell lines of soybean, is not passed on to progeny as a simple nuclear dominant trait, but may be cytoplasmically inherited (Wrather & Freytag 1991). Plants regenerated from Nicotiana debneyi calli tolerant to amitrole displayed some tolerance, but transmission of the resistance trait to F1 progeny did not seem to display either simple Mendelian patterns or maternal inheritance (Swartzberg et al. 1985).
Once again, technical simplicity has meant that identification of salt tolerant variants has been a popular subject for research. In some cases, somaclones displaying tolerance stable through subsequent sexual generations have been identified, e.g. in Brassica juncea (Jain et al. 1991) and alfalfa (Winicov 1990), as a dominant mutation in the latter case (Winicov 1991). In several reports, salt tolerant plants have been regenerated from selected cell lines, although stability through meiosis is yet to be demonstrated, e.g. for Prunus (Ochatt & Power 1989) and Citrus (Spiegel-Roy & Ben-Hayyim 1985). Many reports though concern selection for tolerance which is not stable or the stability of which has not been demonstrated.
Aluminium tolerance was expressed in Nicotiana plumbaginifolia plants regenerated from variant cell lines selected in culture (Meredith et al. 1988). Distribution of tolerance among sexual progeny of the first generation was consistent with that expected for control by a single dominant gene. In subsequent generations though, this clear segregation pattern was not evident. Rice cell lines selected for aluminium tolerance yielded tolerant plants (Van Sint Jan & Bouharmont 1992). For potato, 30% of tolerant cell lines yielded tolerant plants, but only 5% of these clones displayed tolerance after four micropropagation cycles (Wersuhn et al. 1988). These interesting examples illustrate the complexity of somaclonal variation, and highlight the importance of studies, in particular genetic studies, to determine the stability and value of variants identified. In most cases, somaclonal variants displaying tolerance to stresses have not been subjected to such studies. Tolerance to higher levels of cadmium in tobacco cell lines was stable in the absence of cadmium, but not tested in whole plants (Huang & Goldsbrough 1988). These tolerant cell lines also showed higher tolerance to heat shock and cold treatments than the unselected cells. Cold tolerant wheat plants were regenerated from cell lines surviving immersion in liquid nitrogen (Kendall 1991). Wheat cell lines selected for resistance to hydroxyproline, resistance which was stable in the absence of the agent, showed increased frost tolerance (Tantau & Dorffling 1991). Leaf explants from plants regenerated from cotton cell lines resistant to high temperatures also displayed increased tolerance (Trolinder & Shang 1991). Tissue culture derived variation for improved tolerance to acid soils, drought and insects has been reported for sorghum (Miller et al. 1991).
In summary, somaclonal variation has been a popular subject for research, particularly during the 1980s. Forest tree species have not been neglected in this research. Significant work has been conducted on resistance to both disease and herbicides in poplars, and to disease in Larix decidua. A few commercially useful cultivars have been produced for crop species. With respect to disease resistance, for example, new cultivars have been released for sugarcane (resistant to Fiji disease), tomato and celery (resistant to Fusarium diseases) (van den Bulk 1991). Tomes' (1990) documentation of somaclonal variants released or in field trials included chlorsulfuron and imidazolinone resistance in rape, imidazolinone resistance in corn, high solids and Fusarium race 2 resistance in tomato, a white flowered form of lucerne, potato virus Y resistance in tobacco, and fall armyworm resistance in sorghum. In general though, somaclonal variation has not had the agricultural impact that early results promised. A significant problem is that many different phenomena are involved in variation induced in culture. Some of these are useful, and some not. In general, somaclonal variation is still poorly understood, and certainly not controllable. Furthermore, practical application is dependent on the ability to regenerate from the cell lines. This is still a problem for many species, and somaclonal variants often show reduced regenerability even for normally tractable species. Few studies have incorporated sufficient genetic examination to evaluate the stability or usefulness of the variants selected. This is not a field where success with a new crop/trait could be predicted with any confidence. Such a research program would be very empirical with the odds against success.
Assuming stability of the characteristics and the ability to regenerate from the variant cell lines, the use of somaclonal variation is likely to be of most interest where:
Insufficient natural variation is available to provide the level sought for the trait; or
Existing variation is not easily used in breeding.
Many crop species in current use show greatly reduced genetic variability as a result of a long history of selection and breeding. Peach germplasm, for example, is narrow, and regeneration of plants from selected cell lines provided resistance to bacterial spot which was not available in commercial cultivars or a large number of introductions (Hammerschlag 1986). Commercial and breeding populations of most forest tree species generally have not been subjected to the same reduction of genetic variation. The poplars are a possible exception. Even for genetically broad populations, however, there still will be traits for which insufficient natural variation is available. A survey of populations of a number of pine species, for example, revealed no useful resistance to Gremmeniella abietina (Skilling et al. 1983). Similarly, tolerance to many herbicides is not known to exist in angiospermous tree species and would thus be difficult to achieve through traditional breeding (Michler & Haissig 1988). Cold tolerance in many Eucalyptus species is another example. It has been suggested (Larkin 1987) that plants have present in their genomes a far richer array of genetic information than that expressed, that the phenotype represents a limited subset of the genes present, and that somehow cell culture gives access to some of this genetic resource.
A proven technique being available, somaclonal variation may be of value also where natural variation is difficult to manipulate. This approach may in some cases be less disruptive to the commercial genotype (and therefore quicker) than a backcrossing program with a wild type. For forest tree species, breeding programmes for poplars are based on crossing desirable traits, in particular resistance, into commercial clones, and a successful somaclonal approach may offer some efficiencies, particularly if combined with in vitro selection.
In general though, somaclonal variation offers very little for the genetic improvement of most of the major industrial forest tree species - the quantitative traits of most interest e.g. vigour, stem form and wood properties, are not traits which have been usefully altered in vitro. Poplars, often capable of regeneration from cell lines, and for which disease resistance is an important criterion, are perhaps the only candidates worthy of consideration among these species.
No immediate applicability is evident to the other tropical hardwoods or to non-industrial species, for which genetic variation naturally available is generally poorly defined. To summarize the status with respect to developing countries, this technology has very limited application in the short or imtermediate term, and any research would be high risk (with a low probability of success). This research should not be a high priority. In the longer term, the promised benefits of somaclonal variation may be more reliably delivered by genetic engineering approaches.
