Studies to Determine the Cause of Citrus Somaclonal Variation

Variation occurring from the in vitro cul- ture of citrus, namely somaclonal variation, is of importance not only to the genetic con- servation of germplasm in vitro, but also to genetic improvement. Theoretically, somaclonal variation is nearly the same thing as the somatic variation (namely bud- sport mutations), in the fi eld. The differ- ence is that the former is referred to the in vitro system, and the latter to the in vivo system. With the in vitro system, it is much easier to examine the variation at the cyto- logical and molecular levels than it is with what occurs in the fi eld, due to the much higher frequency of variation. For

Harvest date

C2-46 Ratio 16.96 19.41 11.29

SS 7.32 6.29 6.01

Colour score 34.6 33.7 32.8

C1-45 Ratio 17.54 17.38 11.21

SS 6.65 6.29 5.76

Colour score 34.6 33.4 33.2

N14-10 Ratio 21.14 23.7 12.59

SS 7.00 6.45 5.41

Colour score 33.6 32.5 33.2

Control Ratio 18.94 14.92 9.92

SS 6.97 4.71 4.18

Colour score 33.8 33.5 31.0

SS = pounds of soluble solids/90 lb box.

germplasm conservation, unwanted somaclonal variation is a key issue that affects the system’s effi ciency for maintain- ing the genetic stability of the conserved material; on the other hand, for genetic improvement as previously discussed, somaclonal variation is an important source of novel variation, particularly for species that are diffi cult to improve by conven- tional breeding such as sweet orange and grapefruit.

As described earlier, the reported vari- ations in sweet orange include maturity date, morphology, ploidy level and others. Changes in maturity date and seed content may have the most practical value. Auto- tetraploid sweet orange plants have been obtained via organogenesis from non- embryogenic callus and nucellar seedling internodes, and from embryogenic callus. Generally, somaclonal variation can be classifi ed into two categories, one is stable genetic variation that can be transmitted to the next generation, which includes ploidy variation, single gene mutation, movement of retroelements/transposons, etc. (Grandbastien, 1998; Kubis et al., 2003). The other is the epigenetic variation, such as DNA methylation (Fig. 9.1; Finnegan et al., 1996; Hao and Deng, 2002b), which

leads to variable gene transcription and translation. Epigenetic variation in general cannot be transmitted to the next genera- tion.

For the origin of stable somaclonal variation, there are two possible primary pathways; one is that the variation comes from pre-existing variation in the explant. Deng et al. (1985) examined the chromo- some number of the meristems of citrus cul- tivars, and verified the existence of variation in ploidy level in 2–7% of cells examined. Such pre-existing variation that occurs in somatic cells is rarely expressed or recovered, but may be recovered if such cells have an opportunity to regenerate in vitro, and as a result variable plants will be found among the regenerates. Another pathway is that the variation originated during in vitro culture. Potential induction factors include the medium, temperature and the growth regulators used during the culture. However, how these factors may cause variation is poorly understood. Previous experiments have verifi ed that 2, 4-dichlorophenoxyacetic acid (2, 4-D) enhanced ploidy variation in citrus non- embryogenic callus (Deng et al., 1985), but kinetin did not. Generally, the longer the time in culture (culture age), the higher the

Fig. 9.1. An embryogenic callus incapable (1) and capable (2) of embryogenesis of Newhall navel orange did not show any differences in their random amplifi ed polymorphic DNA (RAPD) profi le (3) with primer OPV-12. However, the methylation state of DNA was different; some of the sites were demethylated. (4) Methylation-sensitive amplifi cation polymorphism (MSAP) analysis with the primer pair HM-TCAA/E-ATG indicated the difference. Lanes M and H were calli incapable of embryoid regeneration, while ‘M’ and H’ were calli capable of regeneration. Arrowheads point to the demethylated sites.

rate of associated somaclonal variation. The most easily observed phenomenon is that the regeneration capacity of citrus embryo- genic callus decreases with culture age,

probably due to the accumulation of muta- tions. This phenomenon has been exploited to enhance the selection of somatic hybrids in protoplast fusion experiments by reduc-

Table 9.7. Chromosomal ploidy of different genotypes of citrus calluses.

Chromosome ploidy levels (%)

ing or eliminating the regeneration of diploid plants from the embryogenic parent (Grosser et al., 2000). To get rid of the inter- ruption of pre-existing variation, single pro- toplast sibling lines of Newhall sweet orange were fi rst established by two rounds of low-density protoplast culture in agar- solidified medium. These diploid lines originated from a single protoplast and have the same genetic background. During the regeneration process, chromosome counts were conducted, and the results showed that polyploid cells were found as early as 40 days after culture initiation (Hao and Deng, 2002a).

Chromosome number variation in citrus embryogenic callus is a common phe- nomenon. Embryogenic callus from eight

citrus cultivars was examined, and the per- centage of cells with variable chromosome numbers varied from 4.9 to 9.1% of the total cells (Table 9.7; Fig. 9.2). Most of the cells with variable chromosome numbers were tetraploid. Ten microcalli, all of single pro- toplast descent, were examined for ploidy level. Data presented in Table 9.8 showed that eight out of the ten microcalli had clear chromosomal images and four out of the eight calli had cells with variable chromo- some numbers. Out of the total observed cells of the eight calli, 91.7% of the cells were diploid, and 3.7% were variable. Interestingly, the percentage of cells with variable chromosome numbers did not increase over time (based on examining cells from subsequent subcultures). From

Table 9.8. Chromosomal variations of citrus calluses derived from protoplasts 40 days after culture initiation.

Fig. 9.2. Different ploidy cells found in the embryogenic callus of Valencia sweet orange: (1) Haploid 2 n =

x = 9; (2) diploid 2 n = 2 x = 18; (3) tetraploid 2 n = 4 x = 36; (4) hexaploid 2 n = 6 x = 54.

Table 9.7, data indicate that the subculture year was not positively correlated with the variation.

Sibling lines of Newhall sweet orange were examined for ploidy stability. Data from Table 9.9 showed that the ploidy level was relatively stable after one year in cul- ture. Ploidy level changes apparently reach an equilibrium, in contrast to other types of variation that appear to increase over time in culture. In this case, the Newhall embryo- genic callus has been subcultured for more than six years, much longer than an annual crop, yet the callus still has capacity for somatic embryogenesis. The chromosomal stability is probably one of the reasons. Another reason for this stability may be due to the culture medium used, which contains no variation-inducing growth regulators.

To understand better the mechanism of ploidy level stability in citrus callus, we

tested the mitotic index of cells with differ- ent ploidy levels (Table 9.10). The results showed that tetraploid cells have a lower mitotic index, and a higher percentage of apoptosis (Fig. 9.3; Hao et al., 2002), which can explain the ploidy stability phenome- non, i.e. diploid cell cultures continuously generate a small percentage of cells with variable chromosome numbers, but these cells with variable chromosome numbers generally have a lower competitive ability and, therefore, remain at a low level in the mixed population.

We also examined the chromosome numbers of callus, embryoids and plantlets derived from the same callus line of ‘Anliucheng’ sweet orange. We observed more cells with variable chromosome num- bers in the callus than in the embryoids, which had more variable cells than the resulting plantlets. This result indicates that

Before subculture After subculturing for 1 year

the regeneration pathway screens for normal diploid cells. Normal cells seem to have a higher capacity for plant regeneration.

Another point worthy of further study is the hypothesis that the in vitro system can generate more variation than in vivo culture over the same length of time. If such a comparison is based on the timing of cell division, maybe both systems would have the same variation rate. However, cell divi- sion in the in vitro system is continuous, whereas over the same period of time, cell division in the in vivo system is inter- rupted. Therefore, the in vitro system offers more opportunities for variation to occur.