CALLUS FORMATION IN ANTHER CULTURE OF TEA CLONES,

T h ~ s tudy was carried out to regenerate haploids from cultured anthers of t e a clones. Morphological and histological studies on t h e an ther callus dcvclopmcnt revealed t h a t nuclei of' numerous microsporcs hegan to divide unerjually, forming multicellular s t~ucturcs during the first week of culture and anther lohe.; swelled gradually until hursting. The rate of callus induction was rapid during 6-10 wccks and compact greenish calli were formed from anthers . Calli llecamc more heterogeneous wit11 time in culture. The determination of ploidy levels in anther callus showed that two levels ofploidy were present in callus. In the callus, the pei~entage of haploid cells was more (68%) than that of diploid (6%). The study on cornpailson of callus growth in antllel-s of different clones indicated that the sulvival of anthers of three clones TRI 2043, TRI 2023 and TRI 2025 was high (Iigllcst was 98%, the lowest 78%) and calli were produced in anthers of all clones used in this tiial. TRI 2043 exhibited relatively more callus formation (76.2 mg) from an ther cultured in half Murashige and Skoog (MS) medium with 2,4 D and 13AP grown in light, followed by TRI 2023, TRI 2024, TRI 2025 and lTI777. In the dark, significant callus growth was ohsewed in four clones (TRI 2025, TRI 2024, TRI 2023 and TRI 777) . Calli that formed in light turned dark green, merislcmoidlike stiuctures after transferto the same medium without 2,4 D. However, plantlets could not he regenerated.


INTRODUCTION
Tea, Cantellia sinensis (L.) is cultivated commercially for its tender leaf which is used as a beverage. In Sri Lanka, it is grown under various types of soil and climatic conditions with different pests and disease problems. Therefore, it. is necessary to provide good planting material to growers in order to ensure highest possible yield of tea with high quality. As the demand for genetically uniform and agronomically superior planting material is increasing the creation of new and commercially desirable tea cultivars become essential and this can be achieved only by a proper programme of breeding. For commercial planting, high yielding clones with good leaf quality have been obtained from hybridization. However, selfincompatibility, low seed setting ability of high yielding clones1, rapid loss of viability of seeds and long durations of each sexual cycle have been barriers in the tea breeding programme.
Anther culture technique can play an important role in genetics and plant breeding programme of tea. Even though anther culture technique has been widely used in crop improvement in non-woody plants not much work has been done in woody plants. Anther culture in tea was first described in 196g2. But very few attempts have been made on anther culture of tea. Differentiation of anther callus into a radicated structure in which vascular bundles were found was reported in 19703. Further work on the differentiation rate of the root derived from tea anther cultures was describedin 1981" Therefore, an attempt was made to produce haploids from cultured anthers of Sri Lankan tea clones.

METHODS AND MATERIALS
Plant material : Unopened floral buds (anthers containing microspores at mid uninucleate stage) of clones TRI 2025, TRI 2023, TRI 2024, TRI 2043 and TRI 777 were harvested separately in petri dishes from the seed garden in St.Coombs estate, Talawakelle and stored in a refrigerator a t 5" C for 3 days. They were then sterilised with 70% ethanol for 1 min and 5% Clorox (sodium hypochlorite 5.25 7%) for 10 min. Then they were rinsed three times in sterile water and the anthers were excised from each bud.
Histological and morphological studies : In this experiment, excised 20 anthers of clone TRI 2025 were placed in each petri dish containing half MS medium (Murashige and SkoogS) supplemented with 2,4 D (2.0 mg/L) and BAP (1.0 mg/L) with 10 replicates. They were then incubated in white fluorescence light (16 h, 2000 lux). For cytological examination anthers were taken daily upto 7 days of inoculation and then weekly. They were examined as in wheat anther culture5The weight of anther was recorded every week upto the first month of culture initiation. At monthly intervals, calli were transferred to the same medium but without 2,4 D and kept in white fluorescence light (16 h, 2000 lux). Subculturing was done once a month. Colour and type of calli were observed and callus at 3"' month was used to determine the ploidy levels.
Analysis ofploidy Levels in callu,s : This experiment was done to confirm the haploid status of callus. Seventy five grams of callus was taken in a sterile petri dish and 1 ml of high resolution DNA k i t solution (commercial PARTEC buffer) was added to isolate the nuclei from cells. The sample was chopped gently with a razor blade in the petri dish containing the solution and the suspension was filtered through 50 pm nylon mesh into a small tube to separate nuclei from the debris. Up to 2 ml of DNA staining solution (commercial PARTEC solution) was then added.
The samples were maintained a t 0 "C following isolation of nuclei in order to decrease nuclease activity7. After the time required for staining, DNAof the sample was analysed in a Flow Cytometer (PARTEC, Munster, Germany). Integrals of each individual peak in the histogram were obtained using built in software8.
Callus formatiou : This experiment was done to study the response of anthers of different clones on callus formation. Ten excised anthers (microspores a t mid  Table 1.  Each treatment was replicated 6 times. Bottles (4 mm diameter) containing medium 1 were kept in white fluorescence light (16 h, 2000 lux) and the other two media were incubated in the dark. Subculturing was done once a month. In each treatment, the number of anthers that produced callus and the weight of callus were recorded a t the 3"' month. The data on the weight of callus was analysed using analysis of variance. Means were compared using Duncan's multiple range test!'. At the 3'" month calli were transferred to the same media but without 2 , 4 D and kept in white fluorescence light (16 h, 3000 lux) for the regeneration of shoots.

Histological study
Nuclei of numerous microspores began to divide unequally, forming multicellular structures during the first week of culture. Microspores of those in which the division was observed increased in their original volume. The weight of anther significantly increased during the first month of culture initiation ( Figure  1). Anther lobes swelled gradually until bursting. Callus mass consisted of tracheid elements and parenchyma cells with large vacuoles. All the different steps of embryogenesis from bicellular to embryo stages were observed during the first 3-8 weeks of culture. Very few number of small embryos was noticed in callus formed from anther, but it was not visible to the naked eye.

Morphological study
Anther lobes started swelling in the lSt week and 68% of anthers swelled in the lSt month. Then, they split longitudinally and burst. Bursting of explants started in the 3'" week. The rate of callus induction was rapid during 6-10 weeks. The callus formation from anthers was not observed in the first month ( Figure  2). Compact greenish calli formed from anthers. When the callus aged, the development of the callus was slow and it became more heterogeneous with time in culture. Plant regeneration did not occur in any of the callus tissues during the 5 months of culture.

Ploidy levels in callus
In t l~j s study, first and second peaks were obtained a t channel 80 and 160 respectively. The ratio of the mean channel numbers of peaks in the DNA histogram was calculated7. The value 1:2 obtained suggests that two levels of ploidy were present in the callus. When considering peak index and position at channel, first and secorld peaks should be haploid and diploid respectively. In the anther callus, haploid cells were more (68%) than that of diploid (6%). The resulting histogram with two peaks had coefficients of variation less than 8%.

Callus formation
The survival of anthers of three clones TRI 2043, TRI 2023 and TRI 2025 was high (highest was 98%, the lowest 78%)) in all media tested (Figure 3). After G weeks of incubation, compact greenish callus developed from anthers grown in light and whitish callus in the dark. Further, it was observed t h a t calli were produced from anthers of all clones and the effect of different niedja on callus growth also varied. Callus induction was first observed in anthers of clone TRI 2023 and TRI 2043 cultured in light and also the frequency of their callus formation was about 50%) ( Figure 4).
Of the five clones tested, TRI 2043 a t the 3"' month of incubation exhibited relatively more callus formation (76.2 mg) from anther cultured in medium with 2,4 D and BAP grown in light, followed by TRI 2023 (43 mg) and TRI 2024 (30 mg) as given 111 Figure 5. Significant callus formation was observed in anthers of four clones (TRI 2025, TRI 2024, TRI 2043 and TRI 777) when anthers were cultured in the medium with 2,4 D and kinetin or in combination with IAA and grown in the dark.    Figure 4 : The percentage of callus formation in anthers of five clones.

Regeneration of shoot b u d from callus
At the third month, calli developed from anthers of the five clones were transferred to the same medium devoid of 2,4 D and grown in light. Calli that grew in light turned dark green. Meristemoid like structures were observed after 5 weeks of transfer of calli to the same medium without 2,4 D. When calli formed in the dark were transferred to light condition, most of them turned brown and died. In some instances, they turned greenish yellow in colour after 2 months of transfer.

DISCUSSION
Microspores grow and develop into multicellular masses, embryos or pollen calli under favorable conditions1". The formation of these structures within the anther lobes could be the reason for the swelling of explants in culture. The two different processes t h a t exist simultaneously are the formation of somatic calli and the initiation of microspore development into small embryos or pollen callild. A first symmetric r~litotic division gives rise t o proembryos while asymmetric divisions originate callusn. Even though connective tissue of the anthers proliferated in priority during the first 2-3 weeks of culture this somatic callus then mostly degenerated while callus and embryos develop vigorously from microspores'". Sucrose is generally used a t a concentration of 2-4% in anther culture1*. Effect of a high sucrose level in the inductio~l medium is directly related to the regulation of osmotic pressure during growth induction and it prevents callus formation from somatic anther tissue1:'. In responsive anthers, the wall tissues gradually turn brown and after 3-8 weeks burst open due to the pressure exerted by the growing pollerl callus or pollell plantn. The ploidy levels of the callus were determined by measuring the nuclear DNA content of the cells. The DNA content was measured automatically and rapidly using Ploidy Analyzer. All cells belonging to one peak have the same quality of measured cell substances (e.g. DNA). In the case of a DNA measurement of plant cells, a single peak represents one plojdy level and the peaks of higher ploidy levels appear on the right side of the histograms. As polyploid cells contain more than two chromosome sets, calculation of t h e 2C (C=DNA contents of hap1.oid set of chromosomes) DNA content requires a knowledge of the ploidy level of the material'. In this experiment two levels of ploidy were present in the callus. With the knowledge of the material, these two levels of ploidy should be haploid and diploid. The resulting histogram with two peaks had coef'ficjents. of variation less than S %. Coeffjcient of variation (cv=standard deviation divided by the mean) z1sna1l.y range fron~ 1 to 1 0 5% ' . Low cv wlues nlalre it possjble to estimate small changes in nuclear DNA content. The higher cv values in the measurenlents of chopped plant tissues seem t,o be artefacts of mechanical destruction during choppjngl! Its s~pplication is attractive for ploid y screening in a large nunlber of plants ".
Compact greenish calli developed from anthers cultured in medium with 2,4 D and BAP grown in light arid whitish calli in niediunl with 2,4 D and lrjnetin or in conlbination with IAA grown in t h e dark. Auxin and cytoltiniri together are necessary for good callus formation'" Higher. concentrat~ions of non-phenoxy auxins (IAA, NAA or IBA) alone often pronlot,e root initiation rather than shoots whereas phenoxy-auxin (2,4 D ) pro~notes callus growth and ernbry~genesis'~. Callus growth in anthers of each clone varied according to the culture media and co~ldjtions. Tliis variation in cal!us growth may be due t,o the genotype of the donor plant. Similar findings were reported by ot11er.s~"~". According to these authors, when anthers of rice varieties are cultured in vitro they do not respond equally in producing callus and in regenerating plants. This could be due to genetic or environmental. characteristics of different rice varieties. Calli t,hat grew in l.ig11t t~~r n e d dark green after transfer of calli to the same medium without 2, 4 D. But plant regeneration did not occur in any of' callus tissues during the five months of' inc~lbation. The use of 2,4 D in combina.tion with BAP proved to be better than that of' I M in conlbiliation with either lrjnetin or zeatin for the callus formation fiom sour cherry ant11er.s"'. Regeneration of shoot,s was observedfour nlont,l~s after transfer of callus tissues to the same medium wit;l~out 2,4 D.
Anther lobes swelled gradually due to the f'ormatiorl ofn~ult,~cellular structures. Then they burst and produced the call1 and embryos Two levels of ploidy (haplo~d and diplo~d) were present In the ant,l~er callus. The percentage of haplo~d cells in the callus was more (GSV)) than that of diploid (6%). Therefore, it would be possible to regenerate haploids from the anther callus. The survival of anthers of three clones such a s TRI 2043, TRI 2023 and TRI 2025 was high (78-98%) aniong the clones tested. TRI 2043 was the most effective 111 producing the calli from anthers grown in light. I11 dark, significant callus growtl~ was observed in anthers of four clones (TRI 2025, TRI 2024, TRI 2043 and TRI 777). Though the callus formation from anthers was successfhl, plantlets could not be regenerated yet. Therefore, continued effort and screening of'calh with varlous conlbinatlons of hormones is required to produce the haploids.