Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Received 22 May 1995 Heredity 76 (1996) 335—345 Genetic nature of a nontumour mutant isolated from tumour-prone amphidiploid Nicotiana glauca-Iangsdorffii (G G LL): a critical assessment* M. R. AHUJA BFH, Institute of Forest Genetics, Sieker Landstrasse 2, 22927 Grosshansdorf, Germany The genetic nature of a nontumour mutant, isolated by Izard (1957) following irradiation of germinating seeds of the tumour-prone amphidiploid Nicotiana glauca-langsdorffii (GGLL-T), is re-examined in the light of previous and current studies on tumour formation in Nicotiana hybrids. As there are lingering questions regarding the dominant nature of the nontumour trait, I shall present experimental evidence and theoretical arguments that suggest that the nontumour condition does not necessarily involve a single dominant gene mutation, but rather mutations affecting tumour expression. In the framework of a genetic model of tumour formation in Nicotiana (Ahuja, 1968), the mutations are probably in those ee loci of N glauca affecting tumour expression, and not in the I element of N langsdorffii, as previously suggested by Smith (1972) and Sekine et a!. (1993). As tumour-prone tissues are highly sensitive to environmental influences, including irradiation, it would be very difficult to obtain a mutation from tumour to a completely nontumour state, as in all likelihood, the irradiation treatments would accelerate tumour formation, or induce variation in the tumour expression. Instead, I propose a different approach for isolating a nontumour mutant, which involves irradiation of haploid cells (pollen) from N langsdorffii (I species) and then pollinating N glauca (ee species) with the irradiated pollen. As the cross N glauca x N langsdorffii normally yields a 100 per cent tumorous progeny, any deviation from this norm, i.e. a nontumour mutant, would be easily detected in the proposed cross. The reciprocal cross involving irradiated pollen from the ee parent N glauca may also be attempted to check the validity of the hypothesis. Keywords: genetic control, genetic tumours, irradiation effects, Nicotiana glauca-langsdorffii, nontumour mutant, tumour expression. ving two genomes each of N glauca (GG) and N langsdorffii (LL) in the amphidiploid (GGLL), and Introduction Since the discovery by Kostoff (1930) of nine interspecific tumorous hybrids in the genus Nicotiana, those with different combinations ranging from one N glauca genome with two of N langsdorffii (GLL) to three of N glauca and one of N langsdorffii almost 20 additional hybrid combinations have been reported in which the entire F1 is tumorous (Kehr & (GGGL). However, the tumour expression varies depending on the genomic combination. For Smith, 1954; Ahuja, 1965; Smith, 1972). One of these tumorous interspecific hybrid combinations involving Nicotiana glauca (2n = 24) and N langs- example, the tumours are relatively smaller in the triploid GLL compared with the diploid hybrid (GL) or the amphidiploid (GGLL) combinations (Ahuja & Hagen, 1967). dorffii (2n = 18) has been most extensively investi- gated from the standpoint of genetic nature and physiology of tumour formation (see Ahuja & During the course of his investigations with amphidiploid N glauca-langsdorfjii (GGLL), Izard (1957) irradiated germinating seeds of GGLL with a Hagen, 1967; Bayer, 1982; Smith, 1988). Tumours also occur in hybrids at different ploidy levels invol- dose of 1200 r, and isolated a plant that did not produce any tumours. Although this plant had a *Dedicated to George L. Hagen on his 70th birthday. 1996 The Genetical Society of Great Britain. 335 336 M. R. AHUJA chromosome complement of 42, as in the tumorous GGLL, it showed considerable meiotic abnormal- ities, including univalents and bridges. By the process of selfing, Izard isolated two homozygous lines, which were nontumorous, and showed somewhat normal meiosis. One of these lines (354) was subsequently employed by Smith & Stevenson (1961) for determining the genetic nature of the nontumour mutant (GGLL-NT), and subsequently by a number of other investigators for characterization of this mutant. As there have been conflict- ing reports on the behaviour of this nontumour mutant and its progeny, obtained by crosses with the tumorous amphidiploid GGLL, under greenhouse and field conditions in terms of tumour formation, and with regard to other stress factors such as irradiation and chemical treatments, I think it might be worthwhile to re-examine this mutant in the light of earlier genetic and physiological studies, as well as recent investigations involving Agrobacterium and progeny from the cross GGLL (T x NT) developed tumours at the radiation dosage level of 200 r and above per day. The progeny from the same cross (T x NT) remained nontumorous under the normal (field conditions), indicating that the environmental factors influence tumour expression. However, another cross involving the nontumour mutant (GGLL-NT) and N langsdorffii (LL) which yields the triploid hybrid combination GL(NT)L remains nontumorous under both normal field and irradiation environments. On the other hand, the progeny from the triploid hybrid combination GLL derived from a cross between tumorous GGLL-T and N iangsdorffii (LL) remains tumorous under the two sets of environmental conditions. To find biochemical genetical differences between the nontumour mutant (GGLL-NT) and the tumour-prone amphidiploid (GGLL-T), Bhatia et al. (1967) employed acrylamide gel electrophoresis. Their results revealed that the nontumour and molecular biology. In this report I shall present experimental evidence and theoretical arguments that suggest that the so-called nontumour mutant, obtained following irradiation of seeds from the tumour-prone leaf tissues showed electrophoretic variation in proteins and enzymes. Of the three enzyme systems examined, the nontumour and tumour-prone genotypes differed at least for two tumorous amphidiploid GGLL, does not necessarily involve a single dominant gene mutation, but rather mutations affecting tumour expression. esterase loci, and for one of the peroxidases; these were lacking in the nontumour mutant. The protein and enzyme patterns of the tumour tissue (normally tumours develop on the stem and not leaves) were strikingly different from those in the leaf tissues of the nontuniour mutant and the tumour-prone genotypes. From these results on the protein profiles and three enzyme systems, it is rather difficult to draw Genetic and phenotypic analysis of the nontumour mutant Smith & Stevenson (1961) carried out detailed genetic analysis to determine the nature of the GGLL-NT mutant. They made a series of crosses with the tumorous amphidiploid GGLL-T, to study the inheritance of the nontumour trait. Crosses between T x NT and NT x T yielded a nontumorous progeny, indicating that the nontumour condition is dominant. Backcrosses of the NT-F1 to GGLL-NT gave all nontumorous progeny, and to GGLL-T yielded a segregation ratio of 1:1 for the nontumour and tumour traits, further supporting the dominant mode of inheritance of the nontumour condition. However, selfing of the NT-F1 yielded an F2 popula- tion comprising of more than 90 per cent nontumour plants by the end of the growing season. This result would suggest that the validity of a single dominant gene interpretation is questionable. Futhermore, Smith & Stevenson (1961) investi- gated the effects of chronic gamma irradiation on the tumour and nontumour genotypes. The nontumour mutant line (GGLL-NT) did not develop tumours at any of the radiation dosage levels up to 450 r per 20 h day (for 108 days), whereas the any conclusions regarding the role of specific proteins/enzymes in tumour formation and/or suppression. A detailed cytogenetic analysis by Ahuja & Hagen (1967) of the hybrid derivatives obtained by selfing the partially fertile triploid (GLL) or backcrossing it to N langsdorffii (LL), showed that the tumour expression was enhanced by a threshold of certain N glauca chromosomes on the N langsdorffii background. In other words, the lower the N glauca chromosome number in the hybrid derivatives, the smaller the size of the tumours. In a hybrid derivative with a full genome of N langsdorffii along with two glauca chromosomes, very small tumours developed, compared with those in the triploid (GLL). These observations were interpreted to mean that a number of genes on the N glauca chromosomes enhance tumour expression in the N glauca-langsdorffii hybrid combination (Ahuja & Hagen, 1967). Smith (1988) also showed, following a detailed analysis in the N glauca-langsdorffii hybrid, that genes on the N glauca chromosomes enhance The Genetical Society of Great Britain, Heredity, 76, 335—345. NATURE OF A NONTUMOUR MUTANT IN N/COT/A NA 337 tumour expression, and those on N langsdorffii greenhouse. Only five of ten were raised to maturity. chromosomes have an opposite effect on the tumour Even from this small sample of five plants, two expression. Smith obtained a tumorous segregant developed tumours of variable sizes following the flowering phase. Tumours measured approximately 10 mm in one plant, and 5 mm in the second plant. These observations also raise questions regarding with a full genome of N langsdorffii plus a single N glauca chromosome. However, the tumour size was reduced in the segregant compared with that in the diploid (GL) or even the triploid (GLL), consistent with the observations of Ahuja & Hagen (1967). We had also received seeds of the tumorous amphidiploid (GGLL-T) and the nontumour mutant (GGLL-NT) line No. 354 from Dr Harold Smith, Brookhaven National Laboratory, Upton, New York. Fifty plants of the NT line were grown under field conditions in Philadelphia during the summer months (temperatures in the high 30°C). Tumours failed to develop on any of the 50 plants. Several plants showed slight bulgings on the main stem, perhaps reminiscent of the incipient tumour condition. Towards the end of the growing season four entire plants with bulgings were brought into the greenhouse, generally maintained at 23°C. All four plants developed tumours (Fig. 1) measuring approximately 5—10 mm in diameter and approach- ing in size the tumours in triploid GLL (M. R. Ahuja, unpublished data). In another experiment, ten plants from the nontumour line (GGLL-NT) were grown in the the behaviour of the nontumour mutant under greenhouse and the field conditions (M. R. Ahuja, unpublished data). Further studies on the nature of the nontumour mutant (GGLL-NT) were carried out by Durante et al. (1982). This Italian group also received the seed stocks for the tumour and nontumour genotypes (GGLL-T, GGLL-NT, and their hybrid T x NT) from Dr Smith. Small tumours developed on a number of wounded plants in the nontumourous F1 (NT x T) plants and apparently on nontumorous plants derived from backcrossing the F1 to the tumorous parent (T) in the greenhouse, thus, also questioning the dominant nature of the nontumour mutant. They irradiated the seeds of the tumorous GGLL-T and the nontumourous hybrid T x NT with X-rays. Tumours were induced at a high frequency on the seedlings derived from irradiated seeds of the nontumour T x NT hybrid after 50 kr of irradiation, whereas the tumour-prone GGLL-T developed 100 per cent tumours on the seedlings at 20 kr irradiation. Durante et a!. (1982) also obtained similar results following treatment with 6-azauradil in the tumour (GGLL-T) and the nontumour hybrid T x NT. In the tumour-prone GGLL-T almost 100 per cent tumours were induced, whereas in the F1 hybrid (T x NT) only about 50 per cent of the seedlings developed tumours following treatment with 6-azauracil. Furthermore, Durante et a!. (1989) showed that the neoplastic phenotype in the nontumour GGLL-NT can be restored by treatment of cells with 5-azacytidine, which is known to block methylation of the cytosines in DNA. They recovered habituated colonies in cell culture (but with no shooty differentiation) and the repetitive DNA sequences of treated cells showed a significantly lower level of methylation. The habituated clones obtained by treatment with azacytidine showed reduced chromosome number, close to the amphidiploid level (2n = 42), but also showed the presence of dicentric bridges and a breakage-fusion-bridge cycle. These authors attributed differences between their results and Smith & Stevenson's (1961) results regarding the response of the F1 T x NT to irraFig. 1 Tumours on stems of greenhouse-grown triploid Nicotiana glauca-/angsdoiffii (GLL), amphidiploid (GGLL-T), and the nontumour mutant (GGLL-NT). The tumours are from plants of comparable age. The Genetical Society of Great Britain, Heredity, 76, 335—345. diation or chemical treatment to the fact that Smith & Stevenson grew their plants under field conditions, whereas Durante et al. grew their tumour and nontumour genotypes under greenhouse conditions. 338 M. R. AHUJA Genetic tumours of diverse origins in Nicotiana exhibit different responses to environmental conditions. For example, the tumour-prone hybrid deriva- conditioned or those induced by the crown gall tives of N debneyi-tabacum x N longifiora, carrying a single longiflora chromosome or its fragment on the exogenous supply of essential phytohormones, such as cytokinin(s) and auxin(s) (Braun, 1958; Schaeffer & Smith, 1963; Ahuja & Hagen, 1966a). In contrast, debneyi-tabacum background (Ahuja, 1962), form bacterium Agrobacterium tumefaciens, are capable of supporting their growth in vitro without an tumours in the greenhouse or field conditions. cells from normal plants require these phyto- Progenies of another amphidiploid N debneyi-longi- hormones for their growth and differentiation. In flora, which forms tumours in the greenhouse, do not develop tumours under field conditions. The hybrid N glauca x N longifiora exhibit variable tumour formation in the greenhouse, but tumours fail to develop on plants from the same population grown under field conditions (Ahuja & Hagen, 1967). In a tomato hybrid combination carrying a dominantly inherited gene frosty spot (Frs) from Lycopersicon chilense on the genetic background of the N glauca-langsdorffii hybrid (GL) and amphidip- bid (GGLL) the tumour-prone tissues are autonomous with respect to kinetin and auxin, whereas the tissues from the nontumour mutant (GGLL-NT) had absolute requirements for kinetin as well as auxin for growth and differentiation of shoots like those of the parental species (Schaeffer et at., 1963). Cell lines established from tumour-prone genotypes exhibit a fully habituated state in culture (Meins, L. esculentum tumours developed on the ventral 1982, 1994). In a different tumorous hybrid surface of the leaves when grown in the greenhouse; however, tumours failed to develop under field combination involving N deneyi-tabacum x N longiflora the in vitro habituated tumour tissues grow and conditions (Doering & Ahuja, 1967). Therefore, it would appear that environmental conditions play a teratomas on the plants in vivo, and these teratoma- very important role in the expression of tumours. differentiate numerous aberrant shoots typical of tous growths do not generally differentiate roots with exogenously supplied auxins in vivtro or ex vitro Cytology Cytological studies by Smith & Stevenson (1961) revealed that the tumorous (GGLL-T) and the nontumorous (GGLL-NT) genotypes, and their hybrid T x NT have 42 chromosomes, as would be expected from the parental contributions of N glauca (2n 24) and N langsdorffli (2n = 18), thus confirming Izard's (1957) earlier chromosome counts on the nontumour mutant. The T and the NT produced 85—90 per cent stainable pollen, while the F1 T x NT showed 80 per cent stainable pollen. However, the meiosis was notably more irregular in the F1 hybrid compared with the parental amphidip- bids (T and NT). All three amphidiploids had lagging chromosomes or bridges at anaphase I of meiosis. Nevertheless, amphidiploids carrying complete genomes of two parents provide a buffering effect and adequate functional pollen and egg cells are produced, even if there are some meiotic irregularities. From the cytological observations it would appear that the nontumour mutant has not lost any chromosomes, but probably carries chromosome rearrangements, including deletion(s), or gene mutation(s). Genetic control or hormones in tumour and nontumour genotypes Generally plant tumour cells, which are genetically in a rooting bench (Ahuja & Hagen, 1966a; Ahuja, 1971), suggesting that the balance is heavily tipped in favour of kinetin in these tumour tissues. It was indicated by Schaeffer et at. (1963) and Feng et a!. (1990) that kinetin is an essential requirement for the differentiation of shooty morphology in the nontumour mutant GGLL-NT, resembling that of the tumorous GGLL-T. However, it is unlikely that kinetin alone will lead to the growth autonomy and habituated state in the nontumour mutant tissues, as auxin is also essential for the early growth phase. Our tissue culture studies showed that stem discs from plants with small tumours (ST) and nontumour plants in the same nontumour mutant (GGLL-NT) line did not grow on a modified basal Murashige & Skoog's (1962) medium (Ahuja & Hagen, 1966a), unsupplemented with hormones. The tissues from the ST genotype, however, showed slight enlarge- ment of the cells in the first week in culture. But subsequently the stem discs turned brown and there was no further growth. Perhaps this may result from the slightly higher amounts of the free auxin in the ST compared with NT genotypes, which may have provided the initial trigger. Subsequent cessation of growth of the ST tissues is dictated by its genotype, and is reflected in its inability to synthesize continuously the necessary growth-promoting substances. Both ST and NT tissues grew when the medium was enriched with 0.1 and 1 mg/L 1 indole-3-acetic acid (IAA). However, the differences in the growth rates The Genetical Society of Great Britain, Heredity, 76, 335—345. NATURE OF A NONTUMOUR MUTANT IN N/COT/A NA 339 of the ST and NT tissues were striking indeed. The growth of the NT was only slightly enhanced by the two concentrations of IAA, and further supplemen- prone tissues from diploid hybrid GL and tation of the medium with both phytohormones, higher in the GL compared with GGLL, and the IAA and kinetin, would be necessary for obtaining optimal growth, as was shown by Schaeaffer et al. (1963). By contrast, the growth of the ST tissue was GLL combination had approximately the same auxin levels as in N langsdorfii, but three times lower than greatly enhanced by both levels of IAA, and the growth of callus increased linearly with increasing concentrations of IAA (Fig. 2). At each concentration, comparatively the ST tissues grew ten times that of the NT tissues as determined by the dry weight increase (M. R. Ahuja, unpublished data). It may be pointed out that the aim of the investigation was to demonstrate differences in the auxin requirement between NT and ST segregants in the same population, and not to bring out the optima for their growth in vitro. It would appear that the plants with ST may be somewhat similar to the tumour-prone triploid GLL, which also requires auxin for its growth (Ahuja & Hagen, 1966b). Besides, tumourprone GGLL-T stem pith tissues may also require an initial treatment with auxin to trigger habituation in culture (Cheng & Smith, 1973), compared with constitutively habituated teratomatous tissues. The inherent growth requirements are also reflected in the endogenous auxin levels in the tumour•OmgiLJM UOlmgLIM. 180 DIOm9IUAA 160 S _________ ______________ E uo C .5 120 __________________ ____________________________ 'C 100 C .2' 0 80- . _____ __________________ 60• 40 N glauca (Bayer, 1967). The amphidiploid GGLL showed auxin levels only slightly higher than in N glauca. In another hybrid combination involving N debneyi-tabacurn and N. longifiora and their tumour- prone hybrid derivatives, Bayer & Ahuja (1968) showed that the progeny from tumour-forming F1 hybrids and the early tumour segregants from the backcross generations had very similar auxin levels, but the auxin levels were much lower in the parents and the late tumour-forming segregants. The isozyme analysis of Ahuja & Gupta (1974) revealed that a peroxidase locus from N longifiora inherited in the tumour-prone hybrid derivatives of the latter backcross generation of the cross N debneyi-tabacurn x N longifiora was only activated in the tumorous tissues. These studies suggest that the endogenous auxin levels are determined by the genotype and the genomic relationships between parental species entering tumorous combinations, and these may be critical for the development of tumours in the Nicotiana hybrids. Studies by Ames and his associates (Ames & Mistretta, 1975; Ames, 1981) have shown that there may be a close correlation between 200 - 0. amphidiploid GGLL as determined by Bayer (1967). The extractable auxin (IAA) levels were several-fold ____________ ___________________ 20 0 I NT I I ST Fig. 2 Effect of indole-3-acetic acid (IAA) on in vitro growth of tissues from small tumour-prone (GGLL-ST) and nontumour-prone (GGLL-NT) segregants derived by selfing the nontumour mutant. Stem explants were grown on modified Murashige & Skoog (1962) medium, supplemented with 0.1 mg/L 1 IAA, and 1 mg/L 1 IAA for 5 weeks. Growth is represented by an increase over the initial dry weight. The Genetical Society of Great Britain, Heredity, 76, 335—345. decline in IAA levels and onset of tumour formation in hybrids grown in the greenhouse. Recently, SyOno and his group (Ichikawa et al., 1989; Fujita et a!., 1991; Ichikawa & SyOno, 1991) have provided support for this observation that tumour induction may be associated with a temporary decrease in endogenous levels of auxin in the plants, and after- wards there is a dramatic surge of some 30- to 80-fold in the auxin level within 5 days after tumour induction. These differences in the endogenous levels, observed by different investigators, before tumour formation may result from the techniques employed for auxin assays, for example bioassay of free or bound auxin, the genotype, in vitro or & vitro derived tissues, type of tissues employed (stem segments or apical tissues), and the time tissues were analysed before the flowering phase, which triggers tumour formation under normal conditions, that is, whether tissues were analysed at the elonga- tion phase or just before the flowering phase. However, auxin levels alone do not confer growth autonomy to tumour tissue in vivo or in vitro, but other biosynthetic systems, including cytokinins, must also accompany the characteristic teratomatous tumour growth. I tend to favour the hypothesis that 340 M.R.AHUJA the initial levels of auxins or cytokinins in the tumour-prone genotypes may be in a flux at the time of flowering, which triggers tumour formation under normal conditions. Accordingly, it is the inherent capacity of tumour-prone cells to turn on, among other things, auxin and cytokinin biosynthetic systems, sequentially or simultaneously at the flow- ering phase, or at the critical time in response to stresses before flowering, for the initiation and for the continuous and proliferative formation of the tumour tissue. Agrobacterium T-DNA genes and genetic only in the presence of the wild-type plasmid; ipt alone did not induce shoots in N langsdorffii. The wild-type and ipt + plasmids induced shoots at a low frequency (up to 6 per cent) in the nontumour mutant GGLL-NT, whereas in the tumour-prone GGLL-T, a high frequency of shoots (above 80 per cent) was induced by the wild-type and insertion mutant plasmids. Nacmias et a!. (1987) interpreted these results as implying that there are different genetic controls for initial cell proliferation and for continued autotrophic growth in genetic tumours. On the other hand, Feng et al. (1990) obtained the fully habituated phenotype in the nontumour mutant nontumour tissues have been further investigated by (GGLL-NT) following transformation with the Ti plasmid carrying the ipt gene. They showed that shooty morphology of the nontumour mutant was not only restored with kinetin, but also with the ipt employing T-DNA genes from the Ti plasmid of gene. The transformed NT tissues exhibited a shooty Agrobacterium tumefaciens. Genetic transformation by Agrobacterium in the genetic tumour system in phenotype, which was indistinguishable from the tu mou rs The growth requirements of the tumour and wild-type untransformed genetic tumours on GGLL- Nicotiana has revealed that certain wild-type or mutant genes from the T-DNA can be used for understanding growth autonomy in this system. Nacmias et al. (1987) transformed N glauca, N T. However, some questions remain unresolved regarding the transformation studies in the Nicotiana genetic tumour system with the Ti plasmids langsdorffii, tumorous amphidiploid GGLL-T, and the nontumour mutant GGLL-NT with wild-type by kinetin and the ipt gene are also not quite and mutant genes iaaH, iaaM (tms), and ipi from the Ti plasmid of Agrobacterium tumefaciens. The iaaH and iaaM loci are involved in enzymes which code for indole-3-acetic acid synthetase, and the ipt locus encodes for isopentyl transferase, which is a key enzyme in cytokinin biosynthesis. The results of Nacmias et al. (1987) showed that fully habituated calli (23 per cent) were obtained in N langsdorffii following transformation with only wild- type tms ipt + genes, but in N glauca the habituated calli (up to 90 per cent) were obtained either by the tms — ipt + gene or tms + ipt gene constructs, — indicating that the two parental species have different requirements for attaining a habituated state. The nontumour mutant GGLL-NT could also be habituated by the same insertion mutants (tms ipt + or tms + ipt ) gene constructs as in N glauca, but at a significantly lower frequency (up to 20 per cent). On the other hand, high incidences (100 per cent) of habituated calli were obtained in the tumour GGLL-T with wild-type and mutant gene constructs. However, differentiation of the shooty phenotype was under a different genetic control. The N glauca tissues did not differentiate any shoots when transformed with the wild-type plasmid, but differentiated shoots (some 18 per cent) only in the presence of the ipt + gene, while the N langsdorffii tissues differentiated some shoots (4 per cent) from Agrnbacterium. The shooty phenotypes induced comparable. The shooty phenotype induced by Feng et al. (1990) in the nontumour mutant tissues in vitro was not induced by kinetin alone; the media also contained auxin for growth of the callus. It is also not clear in this study whether nontumour tissues were preincubated in an auxin-containing medium before transformation with the ipt gene. The levels of cytokinins in the tissues before and after transformation of the nontumour mutant are also not comparable to those in the untransformed wild-type GGLL tumour. Of the three cytokinins measured, only isopentyl adenosine (IPA) levels showed significant differences between the GGLL-T (85 units) and the untransformed GGLL-NT (6 units) genotypes. However, after transformation of NT with the ipt gene, the IPA levels increased three times to 18 units, but way below the IPA levels in tumour GGLL-T. The levels of two other cytokinins, transzeatin riboside (t-ZR) and dihydrozeatin riboside (DHZR), are also puzzling. There are no significant differences in the levels of t-ZR (T = 2.1; NT = 1.2) and DHZR (T 8.7; NT = 7.3) between the tumour GGLL-T and nontumour GULL-NT. However, after transformation with the ipt gene the levels of these cytokinins are more than 100 units in the NT genotype. The total levels of the three cytokinins are at least two and half times higher in the transformed NT compared with the T genotype. Based on their results, Feng et a!. (1990) concluded that genetic The Genetical Society of Great Britain, Heredity, 76, 335—345. NATURE OF A NONTUMOUR MUTANT IN NICOTIANA 341 tumours are caused, at least in part, by the elevated levels of cytokinin in the interspecific Nicotiana species hybrids. But that seems to be only part of Conclusions and theoretical considerations Taken together these observations question the the story. The role of Agrobacterium rhizogenes T-DNA from validity of a single dominant gene interpretation of the mutation from tumour to nontumour state in N the Ri plasmid has been extensively investigated in gtauca-langsdorffii. It would appear that by X- hairy root disease, growth and differentiation in irradiation of Nicotiana glauca-tangsdorffii germinating seeds, Izard (1957) may have produced genetic variability affecting growth and tumour expression, and these genetic variants may have included cryptic plants (see Constantino et a!., 1994). Recent studies have shown the presence of rol gene sequences designated Ng rot (Furner et a!., 1986) and more recently as cT-DNA (cellular T-DNA; Aoki et al., 1994) from the Ri plasmid in the untransformed N glauca, one of the parents involved in the tumourprone hybrids. Furner et a!. (1986) found more than 80 per cent homology between the endogenous N glauca Ng rol sequences (rolB, and rotC) and the TL region of the Ri plasmid. As the Ng rol sequences are not transcribed in N glauca, Furner et at. (1986) considered that they may be pseuodogenes. However, recent studies by Ichikawa et at. (1990) have shown that the Ng rolB and Ng rotC gene sequences are transcribed in the tumour tissue of the hybrid N gtauca x N langsdorffii; the tran scription of rolB increased, whereas rolC decreased, with an increase in the endogenous auxin level in the tissue. However, transcription of these cT-DNA genes was completely suppressed by exogenous application of auxin. In addition to rolB and rotC sequences, two other genes homologous to open reading frames 13 and 14 (ORFs 13 and 14) of Ri plasmid are also present in the cT-DNA of untransformed N glauca, which are also expressed in the genetic tumours on N glauca x N langsdorffii, but not in the leaf tissue of the tumour-prone hybrid (Aoki et a!., 1994). According to Aoki et al. (1994) ORFs 13 and 14 may be involved in the synthesis of growth substances similar to cytokinin. If untrans- formed N glauca has all the necessary cT-DNA genes for auxin and cytokinin autonomy, the question arises why does it not develop tumours when treated with chemicals or with irradiation? What is the contribution of N langsdorffii in the N glauca x N langsdorffii tumorous combination? Does N tangsdorffi contribute gene(s) for an additional! different cytokinin, or for deregulation of hormone synthesizing genes, or possibly demethylation of the Ng cT-DNA, at the time of initiation of tumors in a hybrid system? The main questions, however, still need to be addressed: what is the specific role of the Ng cT-DNA in the development of tumours and do other Nicotiana species involved in the tumorous combinations also carry oncogens that are derived from Agrobacterium? The Genetical Society of Great Britain, Heredity, 76, 335—345. genetic aberrations. Subsequently, he isolated two nontumour lines under field conditions. No gross structural chromosome abnormalities were observed by Izard (1957) and Smith & Stevenson (1961) in the nontumour line 354, which carried 42 chromosomes (the summation of two parental contributions) as does the tumorous amphidiploid GGLL. Subsequent studies with the nontumour mutant showed that it behaves differently under field and greenhouse conditions, and that small tumours develop on the nontumour mutant under greenhouse conditions, and in the F1 progenies derived following crosses between T x NT. Tumours also developed on nontumour derivatives following irradiation (Smith & Stevenson, 1961; Durante et al., 1982). However, one critical cross involving NT x NT, and its F2 progenies have not been analysed so far. As the first breeding results suggested that the nontumour condition was transmitted as a dominant trait, the NT x NT cross was perhaps considered unnecessary. We might ask the question: if the nontumour mutant derived from the tumorous state does not involve a single dominant gene mutation, then what is the nature of this mutation? There are at least three other possibilities which present themselves: (i) that the nontumour mutation is deficient in transmission; (ii) that the nontumour condition results from an incomplete dominant mutation; and (iii) that the nontumour state involves more than one gene mutation, which control tumour expression. The first possibility does not appear to be tenable, as the entire F1-NT x T remains nontumorous under field conditions, but tumours can be induced by wounding in the greenhouse, and under the irradiation environment. The second possibility that the nontumour state is controlled in inheritance by a partially dominant gene is also questionable, as the F1 is nontumorous, and the F2 segregation ratios are skewed towards the nontumour condition (about 90 per cent of the progeny is nontumorous in the F2). This leaves us with the third alternative, which states that more than one mutation is involved in the 342 M.R.AHUJA origin of the nontumorous state with variable expression of tumours under different sets of environments. The question arises whether the N. longiflora contributes a relatively simply inherited gene(s) on the N glauca or N langsdorffli genomes have mutated to account for the nontumour state? The following is an attempt to answer this question. factors (more than one locus) from N debneyi- It has been shown by earlier studies that some tumour genotypes in Nicotiana and Lycopersicon respond differently to two sets of environments, that is, field and greenhouse conditions (Ahuja & Hagen, 1967; Doering & Ahuja, 1967). As some of the plants from the nontumour mutant line form small tumours in situ, but do not grow on the hormone- free medium in vitro, it would appear as though genes affecting tumour growth and development may have been impaired. The differential growth rates of the tissues from the tumour-forming and nontumorous plants from the same population in vitro also strengthen the view concerning genetic variability in the nontumour mutant line. In many respects the tumorous plants in the nontumour mutant line resemble the triploid GLL (with one genome of N glauca and two genomes of N langsdorffii): they both form small tumours and require exogenous supply of phytohormones for their growth and shooty differentiation (Ahuja & Hagen, 1966b). Based on detailed cytogenetic studies, Ahuja & Hagen (1967) suggested that a number of genes on certain N glauca chromosomes enhance tumour expression in the N glauca-langsdorffii hybrid genetic component designated I, essential for tumour initiation, which interacts with the ee genetic tabacum to form tumours. Earlier Näf (1958) had proposed that Nicotiana species entering tumorous combinations may be divided into two groups: one arbitrarily designated the 'plus' group consists largely of the species of the section Alatae with nine or ten chromosome pairs; whereas the 'minus' group consists of a variety of species coming from diverse sections of the genus, with 12 chromosome pairs or a multiple thereof. Barring a few exceptions, most species of Nicotiana entering tumorous combinations can be classified according Naf's scheme. According to Ahuja's hypothesis, the 'plus' group species such as N longiflora and N langsdorffii are the I carriers, whereas the 'minus' group species such as N debney, N tabacum and N glauca are the ee contributors involved in enhancement of tumour expression. In his review on plant genetic tumours, Smith (1972) wrote: 'If the hypothesis of Ahuja (1968) is correct, the mutation for the nontumorous condition may be expected to involve an I factor in the chromosome of N langsdorffii'. A similar statement was recently made by Sekine et a!. (1993). I would have agreed with Dr Smith if the nontumour condition had been transmitted as a single dominant trait in the selfed and backcross progenies, and tumours did not develop on the NT mutant or its nontumor- combination. It is tempting to suggest that by irradiation of tumorous amphidiploid N glauca-langs- ous progeny under different conditions. But that dorffii seeds, the genes on the N glauca genome may have been impaired. In another interspecific tumorous hybrid combination involving N debneyi-tabacum and N longifiora a nontumour condition does not seem to be transmitted as a dominant trait in the crosses, and the detailed cytogenetic study, accompanied by a repeated backcross programme, revealed that a specific N longifiora chromosome or its fragment in an otherwise N debneyi-tabacum background was adequate for the development of tumours (Ahuja, 1962). The tumour condition was transmitted as a dominant trait in this hybrid combination and, there was no diminishing of tumour expression in the backcross derivatives, compared with the tumorous F1. These results suggested that gene(s) on a single N longifiora chromosome are sufficient, whereas from N deneyi-tabacum a number of genes distributed over several chromosomes are required for tumour formation. Based on these observations, along with cytogenetic and tissue culture studies involving hormone requirements in the N debneyitabacum-longifiora and N glauca-langsdorffii hybrid complexes, Ahuja (1968) proposed a hypothesis that does not seem to be the case. Because the tumour expression is greatly modified by the environmental factors and irradiation, it is tempting to suggest that ee genes on the N glauca parent in the amphidiploid N glauca-langsdorffii may have been impaired by irradiation. That I factors on the N langsdorffii genome have not mutated comes from the cross involving the nontumour mutant and N langsdorfjii (GGLL-NT x LL) from Smith & Steven- son (1961), which gave a nontumour triploid (GLL-NT), indicating that addition of a normal genome of N langsdorffii did not initiate tumours, and that the radiation-induced mutations are probably in the ee factors of N glauca. It would be interesting to investigate if the mutation/deletions are in the cT-DNA, involving any specific rol or ORF genes, in the genome of N glauca. Stresses of diverse nature, such as wounding, chemical treatments, irradiation, etc. are known to initiate and enhance tumour formation at any stage of tumour-prone Nicotiana hybrids or its tumourThe Genetical Society of Great Britain, Heredity, 76, 335—345. NATURE OF A NONTUMOUR MUTANT IN NICOT/ANA 343 prone hybrid derivatives. With this kind of a germ- Parent line-conditioned tumour-prone system, which is highly sensitive to environmental stresses, it would be rather difficult to induce a mutation from tumour to nontumour state by irradiation in a tumour-prone amphidiploid, or a diploid, because this will in all Nicotiana langsdorfihi (LL) Nicotiana glauca (GG) 7 Irradiation Gametes likelihood enhance or alter tumour expression. Pollen Experimental data are consistent with this hypothesis. Many investigators (Sparrow et a!., 1956; Hagen et al., 1961; Smith & Stevenson, 1961; Ahuja & Cameron, 1963; Conklin & Smith, 1969; Durante et a!., 1982) have irradiated tumour-prone Nicotiana hybrid seeds, or plants at various stages of development, but so far no one has been able to induce a mutation in the nontumour direction in this system. Izard also attempted it, but in my opinion may have Fl GL Nontumorous, if I mutated succeeded in introducing genetic variability for tumour expression, possibily impairing the ee genes on the N glauca chromosomes. It would appear, therefore, that in the germlineconditioned tumour-prone hybrid system in which the whole plant is a 'potential tumour', the somatic tissues or seed are not suitable material for the induction of a nontumour mutation, in particular in the complex amphidiploid. Radiation may influence plant tumours in at least two different ways, but the distinction between these may be operationally somewhat difficult. The first type of effect is by the disturbance of metabolism, in particular the hormone biosynthetic systems, which would then trigger tumorous growth. This mechanism appears to be more relevant to tumour-prone Nicotiana hybrids which maintain a precarious balance between normal growth, before flowering, and neoplastic development, after flowering. The balance can be easily disturbed at any stage of vegetative growth before flowering by a variety of diverse stresses, including irradiation. The second type of radiation effect may involve mutations in the nuclear genes (including chromosome aberrations or gene mutations). The mutations may affect tumour expression, ranging from new genotypes with nontumour state, to those with variable tumour incidence, as in the case of the nontumour mutant in the tumour-prone amphidiploid Nicotiana glauca-langsdorffi produced by Izard (1957). Theoretically, it should be possible to induce a mutation in the I element in a diploid, but it would be more likely to induce mutations in the ee elements, as more of them are available as targets. It is a numbers game involving probability events. Instead, I propose a slightly different approach to the problem of mutability in the Nicotiana tumour system. I suggest irradiation of haploid cells (pollen) The Genetical Society of Great Britain, Heredity, 76, 335—345. Chromosome doubling GGL'L' Nontuinorous amphidiploid Fig. 3 Diagram showing a crossing scheme involving irradiation of pollen from Nicotiana langsdorffii (I species), and then pollinating N glauca (ee species) with the irradiated pollen to produce the F1. As the cross N glauca x N langsdorffii normally yields a 100 per cent tumorous progeny, any deviation from this norm (a nontumour mutant) would be easily detected, in which case an I element has been impaired by irradiation. The chromosome number of the nontumour mutant could then be doubled to produce a homozygous and fertile nontumorous amphidiploid (GGL'L'). The reciprocal cross utilizing the irradiated pollen from the ee parent N glauca should also be attempted to test the validity of the hypothesis. from one of the normal Nicotiana species, which enters into tumorous combination, and then producing specific hybrids utilizing the irradiated pollen. In the test systems, I suggest irradiation of pollen from an I carrier species, N langsdorffii, with appropriate doses, and then pollination of the ee species, N glauca, with the irradiated pollen (Fig. 3). As the cross N glauca x N langsdorffii normally yields a 100 per cent tumorous progeny, any deviation from this norm, that is, a nontumour mutant, would be easily detected in the proposed cross, in which case an I element has been impaired by irradiation. To test the alternative hypothesis, the reciprocal cross utiliz- ing the irradiated pollen from the ee parent N glauca should also be attempted. But it is anticipated 344 M.R.AHUJA on the basis of my hypothesis, that the latter cross may produce variation in the tumour expression in the hybrids. Both these possibilities are testable. The F1 (GL) progeny showing a nontumour mutant (GL'-NT), if produced, could be maintained in tissue culture, or its chromosome number doubled by colchicine to produce a nontumour amphidiploid (GGL'L'-NT), which would be homozygous for the nontumour trait, and at the same time fertile. BAYER, M. H. 1982. Genetic tumors: physiological aspects of tumor formation in interspecific hybrids. In: Kahl, G. and Schell, J. (eds) Molecular Biology of Plant Tumors, pp. 33—67. Academic Press, New York. BAYER, M. H. AND A}IUJA, M. R. 1968. Tumor formation in Nicotiana: auxin levels and auxin inhibitors in normal and tumor-prone genotypes. Planta, 79, 292—298. BHATIA, C. R., BUIATI1, M. AND SMITH, H. H. 1967. Electro- phoretic variation in proteins and enzymes of the tumor-forming hybrid Nicotiana glauca x N langsdorffii and its parents. Am. .1. Bot., 54, 1237—1241. BRAUN, A. C. 1958. A physiological basis for autonomous growth of the crown-gall tumor cells. Proc. Nazi. Acad. References Sci. US.A., 44, 344—349. M. R. 1962. A cytogenetic study of heritable tumors in Nicotiana species hybrids. Genetics, 47, 865—880. AHuJA, M. R. 1965. Genetic control of tumor formation in higher plants. Q. Rev. Biol., 40, 329—340. AHUJA, M. R. 1968. An hypothesis and evidence concerning AHUJA, the genetic components controlling tumor formation in Nicotiana. Mo!. Gen. Genet., 103, 176—184. AHUJA, M. R. 1971. Genetic control of phytohormones in tumor and non-tumor genotypes in Nicotiana. Indian J. Exp. Biol., 9, 60—68. AHUJA, M. R. AND CAMERON, D. R. 1963. The effects of X-irradiation on seedling tumor production in Nicotiana species and hybrids. Rad. Bot., 3, 55—57. AHUJA, M. R. AND GUPTA, V. K. 1974. Control of tumor- associated peroxidase in a genetic tumor system in Nicotiana. Experientia, 30, 1007—1008. A1-IUJA, M. R. AND HAGEN, G. L. 1966a. Morphogenesis in Nicotiana debneyi-tabacum, N longiflora and their tumor-forming hybrid derivatives in vitro. Devl. Biol., 13, 408—423. CHENG, T.-Y. AND SMITH, H. H. 1973. The influence of genomes on the autonomous growth of pith cultures of Nicotiana glauca-langsdorffii hybrids. Planta, 113, 29—34. CONKLIN, M. E. AND SMITH, H. H. 1969. Endogeneous beta irradiation and tumor production in an amphidiploid Nicotiana hybrid. Am. .1. Bat., 55, 473—476. CONSTANTINO, P., CAPONE, I., CARDARELLI, M., DE PAOLIS, A., MAURO, M. L. AND TROVATO, M. 1994. Bacterial plant oncogenes: the rol gene saga. Genetics, 94, 203—211. DOERING, G. R. AND AHUJA, M. R. 1967. Morphogenetic studies of a genetically controlled tumor-like condition in Lycopersicon hybrids. Planta, 75, 85—93. DURANTE, M., GERI, C., BUIATFI, M., BARONCELLI, S., PARENTI, R., RONCHI, V. N., MARTINI, G., GRENCI, F. C., GRISVARD, J. AND GUILLE, a. 1982. DNA heterogeneity and genetic control of tumorigenesis in Nicotiana tumorous and nontumorous genotypes. Devl. Genet., 3, 25—39. DURANTE, M., CECCHINI, E., NATALI, L., CITTI, L., GERI, C., PARENTI, R. AND RONCHI, V. N. 1989. 5-Azacytidine- AHUJA, M. R. AND HAGEN, G. L. 1966b. Chromosomes and induced tumorous transformation and DNA hyper- nutritional requirements of a tumor-forming Nicotiana methylation in Nicotiana tissue cultures. DevI. Genet., hybrid and its derivatives. Am. J. Bot., 53, 609. AHUJA, M. R. AND HAGEN, 0. L. 1967. Cytogenetics of 10, 298—303. FENG, X.-H., DUBE, S. K., BOTTINO, P. J. AND KUNG, S.-D. tumor-bearing interspecific triploid Nicotiana glaucalangsdorffii and its hybrid derivatives. J. Hered., 58, 103— 108. AMES, I. H. 1981. Genetic tumors in plants with special emphasis on Nicotiana hybrids. In: Kaiser, H. E. (ed.) Neoplasms — Comparative Pathology and Growth in Animals, Plants and Man, pp. 483—491. Williams and Wilkins, Baltimore. AMES, I. H. AND MISTRETFA, p w. 1975. Auxin: its role in genetic tumor induction. P1. Physiol., 56, 744—746. AOKI, S., KAWAOKA, A., SEKINE, M., ICHIKAWA, T., FUJITA, T., SHINMYO, A. AND SYONO, K. 1994. Sequence of the cellu- lar T-DNA in the untransformed genome of Nicotiana glauca that is homologous to ORFs 13 and 14 of the Ri plasmid and analysis of its expression in genetic tumors of N glauca x N langsdorffii. Mol. Gen. Genet., 243, 706—710. BAYER, M. H. 1967. Thin-layer chromatography of auxin and inhibitors in Nicotiana glauca, N langsdorffii and 1990. Restoration of shooty morphology of a nontumorous mutant of Nicotiana glauca x N langsdorffii by cyto- kinin and the isopentyltransferase gene. Plant Mol. Biol., 15, 407—420. FUJITA, T., ICHIKAWA, T. AND SYONO, K. 1991. Changes in morphology, levels of endogenous IAA and protein composition in relation to the development of tobacco genetic tumor induced in the dark. Plant Cell Physiol., 32, 169—177. FURNER, I. J., HUFFMAN, 0. A., AMASINO, R. M., GARFINKLE, D. J., GORDON, M. P. AND NESTER, E. w. 1986. An Agro- bacterium transformation in the evolution of the genus Nicotiana. Nature, 319, 422—427. HAGEN, G. L., GUNCKLE, J. E. AND SPARROW, A. H. 1961. Morphology and histology of tumor types induced by X, gamma, and beta irradiation of a tobacco hybrid. Am. J. Bot., 48, 691—699. ICHIKAWA, T., KOBAYASHI, M., NAKAGAWA, S., SAKURAI, A. AND SYONO, K. 1989. Morphological observations and three of their tumor-forming hybrids. Planta, 72, qualitative and quantitative studies of auxins after 329—337. induction of tobacco genetic tumors. Plant Cell Physiol., The Genetical Society of Great Britain, Heredity, 76, 335—345. NATURE OF A NONTUMOUR MUTANT !N NICOTIANA 345 30, 57—63. ICHIKAWA, T., OZEKI, Y. AND SYONO, K. 1990. Evidence for the expression of the rol genes of Nicotiana glauca in genetic tumors of N glauca x N langsdorfJii. Mo!. Gen. Genet., 220, 177—180. ICHIKAWA, T. AND SYONO, K. 1991. Tobacco genetic tumors. Plant Cell Physiol., 32, 1123—1128. IZARD, c. 1957. Obtention et fixation de lignees tumorales et non tumorales a partir de mutations experimentales de l'hybride N glauca x N langsdorffii. C. R. Acad. Agric. (France), 43, 325—327. KEHR, A. E. AND SMITH, H. H. 1954. Genetic tumors in Nicotiana hybrids. Brookhaven Symp. Biol., 6, 55—78. KOSTOFF, D. 1930. Tumors and other malformations on certain Nicotiana hybrids. Zentr Bakteriol. Parasitenk Infekt. Jena, 81, 244—260. MEINS, F. Jr. 1982. Habituation of cultured plant cells. In: KahI, G. and Schell, J. (eds) Molecular Biology of Plant Tumors, pp. 3—31. Academic Press, New York. MEINS, F. Jr. 1994. Habituation of cultured cells for cyto- kinins. In: Mok, D. W. S. and Mok, M. C. (eds) Cytokinins — Chemistry, Activity, and Function, pp. 269—287. CRC Press, Boca Raton, FL. MURASHIGE, T. AND SKOOG, F. 1962. A revised medium for tiana sp. and hybrids induced by Agrobacterium tumefaciens T-DNA mutants. Devi. Genet., 8, 6 1—71. NAF, u. 1958. Studies on tumor formation in Nicotiana hybrids. I. The classification of the parents into two etiologically significant groups. Growth, 22, 167—180. SCHAEFFER, G. W. AND SMITH, H. H. 1963. Auxin—kinetin interactions in tissue culture of Nicotiana species and tumor-conditioned hybrids. P1. Physiol., 38, 291—297. SCHAEFFER, G. W., SMITH, H. H. AND PERKUS, M. P. 1963. Growth factor interactions in the tissue culture of tumorous and nontumorous Nicotiana glauca-langsdorffii. Am. .1. Bot,, 50, 766—771. SEKINE, M., ICHIKAWA, T. AND SYONO, K. 1993. A tumoriza- tion-redifferentiation system involving tobacco genetic tumors in tissue culture. J. Plant Res. (Special Issue), 3, 83—95. sMITH, H. H. 1972. Plant genetic tumors. Progr exp. Tumor Res., 15, 138—164. sMITH, H. H. 1988. The inheritance of genetic tumors in Nicotiana hybrids. J. Hered., 79, 277—283. SMITh, H. H. AND STEVENSON, H. Q. 1961. Genetic control and radiation effects in Nicotiana tumors. Z. Vererbungsl., 92, 100—118. SPARROW, A. H., GUNCKLE, J. E., SCHAIRER, L. A. AND rapid growth and bioassays with Nicotiana tissue HAGEN, G. L. 1956. Tumor formation and other morpho- cultures. Physiol. Plant., 15, 473—497. genetic responses in an amphidiploid tobacco hybrid exposed to chronic gamma irradiation. Am. J. Bot., 43, NACMIAS, B., UGOLINI, S., RICCI, M. D., PELLEGRINI, M. G., BOGANI, P., BETFINI, P., INZE, D. AND BUIATI1, M. 1987. Tumor formation and morphogenesis on different Nico- The Genetical Society of Great Britain, Heredity, 76, 335—345. 377—388.