Protoplast regeneration from normal and bromodeoxyuridine-resistant sycamore callus

Tissue cultures of plant cells of many species can be induced under appropriate conditions to divide in an organized manner and to differentiate to whole plants (Steward, Mapes & Ammirato, 1969; Rao, Handro & Harada, 1973). It is now becoming experimentally possible to modify the genetic content of cultured cells and grow the tissues back to a plant. This could enable genetically useful traits, such as genes for disease resistance, to be introduced into a population of important plants where it could be propagated by conventional breeding techniques. This modification could range from the introduction of small amounts of foreign DNA to the introduction of organelles or the fusion of whole cells from genetically unrelated species. Plant cell hybridization would also be a very useful approach to studying problems of cellular organization and control. In one special case a hybrid plant has been produced from fused protoplasts and shown to be identical to the hybrid produced by conventional breeding techniques (Carlson, Smith & Dearing, 1972). We have been investigating: (1) conditions for the production of healthy protoplasts from a tissue culture of sycamore; (2) the effects of various reagents on the fusion of tobacco leaf, maize root and sycamore callus protoplasts; (3) conditions for the regeneration of walls by sycamore callus and for the resumption of cell division leading to callus formation; and (4) the production of mutant lines of tissue culture, which would be used in a system, developed for animal tissue culture cells, for selecting hybrid cells or protoplasts from a mixed population.

The HAT selective system (Littlefield, 1964, 1966; Szybalski, Szybalska & Ragni, 1962) utilizes animal cells lacking in either thymidine kinase (EC. 2.7.1.21) or hypoxanthine guanosine phosphoribosyl transferase (EC. 2.4.2.8) for the selection of hybrid cells. These mutant cells are selected for by their resistance to the base analogues bromodeoxyuridine or azaguanine (Littlefield, 1963; Kit, Dubbs, Piekarski & Hsu, 1963). Enzymic mutants of this sort would be extremely useful in plant protoplast fusion work as they would enable the selective growth of rare heterokaryons from a large number of homokaryons and unfused protoplasts. We have isolated a bromodeoxyuridine-resistant callus of sycamore and have also prepared protoplasts from it and studied their regeneration.

Materials were Analytical Reagent grade or its equivalent. Aminopterin was obtained from Serva Ltd (Microbio Laboratories), thymidine phosphates from Sigma UK Ltd. [6-³H]-thymidine (2 mCi/ml) was obtained from the Radiochemical Centre (Amersham). Calcofluor white was a gift of the American Cyanamid Co.

The tissue culture (S₄) used in this study was derived from a hypocotyl explant of sycamore (Acer pseudoplatanus). The isolation and conditions for growth and differentiation of the S₄ callus have been described (Wright & Northcote, 1972). The cells were grown in the PRL4 medium of Gamborg (1966) with the addition of 1 mg/1. of 1-naphthyl-acetic acid and 2 % sucrose (growth medium). Agar (1 %) was used for the growth of solid calluses. Liquid suspensions of cells were grown in 500-ml conical flasks containing 100 ml of medium and an inoculum of 10–25 ml of cells. The cells were shaken at 25 °C in the dark, and subcultured every 7–10 days. All cell cultures were maintained in a sterile condition. Media and glassware were autoclaved at 120 °C, at 103 kN m⁻² (15 lb in.⁻²) for 30 min. The constituents of HAT medium (hypoxanthine 10⁻⁴ M, aminopterin 10⁻⁵ M, thymidine 10⁻⁴ M) and bromodeoxyuridine were added to the solutions prior to autoclaving. Solutions containing high sorbitol concentrations or enzymes, were sterile-filtered by passage through a Millipore disk (0·45 μm). All operations were performed on a sterile air bench.

Packed cell volume was measured by centrifugation of liquid suspensions (10 ml) from replicate cultures at 50 g for 10 min in a sterile graduated centrifuge tube. The contents were then resuspended and replaced in the flask. Growth of solid tissues was measured by placing 3–5 replicate pieces of callus tissue of 0·1±0·05g on to 20 ml agar medium. The initial and later weights were measured and the mean percentage increase and standard error calculated.

Cells and protoplasts were viewed on a Zeiss Ultraphot II equipped with Nomarski optics and u.v. lamp. Petri dishes were viewed on a Union inverted microscope. Cell walls and remnants of cell wall material were detected by staining the protoplasts for 1 min with 0·1 % Calcofluor in growth medium and viewing under u.v. light.

Protoplasts were isolated from liquid suspension cultures of callus using a mixture of macerozyme (a pectinase) and Cellulase Onozuka SS (Unwin Ltd, Welwyn Garden City). The Onozuka was purified by passage through a Sephadex G-25 column in glass-distilled water (25 g/80 ml on 800 ml bed volume). The front running dark brown peak of material excluded from the gel was collected and freeze-dried (Hanke & Northcote, 1974).

Cell suspension (20 ml) was mixed with an equal volume of 09 M sorbitol in growth medium. The cells were allowed to settle and the supernatant was drawn off down to 4 ml. To this was added 4 ml of 5 % cellulase and pectinase in 0·45 M sorbitol in growth medium.

The pH of this mixture was in the range 4·8–5·4 and no adjustment was made. The cells were shaken gently on a reciprocal shaker for 4–12 h during which time protoplasts were released into the medium. The protoplasts were separated from the remaining cell clumps by a 43-μm stainless steel mesh (R. Cadisch and Sons, Finchley), and then washed on a glass-fibre filter paper cone by dispensing 100 ml medium with 0·45 M sorbitol over them slowly with a Pasteur pipette. The last 10 ml of protoplast suspension in the filter funnel were poured off and then centrifuged at 45 g for 10 min to concentrate the protoplasts in 1–2 ml of medium. A sample was removed for counting in a haemocytometer and the rest of the protoplasts were embedded in agar medium by the method of Nagata & Takebe (1971). The final concentrations of components in the agar regeneration medium were agar 0·6 %, sorbitol 0·45 M, calcium 5 mM, glucose and ribose 0·25 g/1. (Kao et al. 1973), zeatin 12·5 μg/1., all in growth medium made up in Analar water.

Sterile maize (Zea mays var. Caldera 535) roots (15) each 1–3 cm long were taken from 4-day-old seedlings and chopped into 2-mm lengths before incubation in 8 ml purified cellulase (2·5 %) and pectinase (2·5 %) in 0·45 M sorbitol in growth medium. The mixture was gently shaken for 19 h or more and the protoplasts collected by passage through 63-/MT1 steel mesh and centrifugation.

Mesophyll protoplasts were prepared by a modification of the method of Power & Frearson (1973)-Mature leaves of Nicotiana Java or xanthi were surface-sterilized briefly in ethanol and then Milton solution (Richardson-Merrell Ltd., London) and washed in sterile water. The leaves were then allowed to wilt for 4 h before the lower epidermis was stripped from the leaf and the exposed mesophyll cells plasmolysed on 0·5 M sorbitol. The leaf pieces (0·5 g) were then laid on to 8 ml of purified cellulase (2·5 %) and pectinase (2·5 %). The flask was evacuated for 5 min to infiltrate the air spaces with enzyme solution. The leaf pieces were incubated for 12 h in enzyme solution before being strained through 43-μm steel mesh to separate the protoplasts.

Fusion experiments were performed on freshly isolated protoplasts usually taken directly from the enzyme solution. Protoplast suspensions were spun at 11–50g to pellet them and were then resuspended in the test solution and spun again to bring about protoplast contacts. The pellet was then gently resuspended with a Pasteur pipette and a sample viewed for protoplast fusion and viability.

The assay is modified from Schwarz & Fites (1970). Suspension cells (4 days after sub-culture) were collected on 2 layers of muslin, and 5 g fresh weight were added to 5 ml of cold 0 4 M phosphate buffer (NaH₂PO₄. 2H₂O, 24-3 g/1.; K₂HPO₄, 42·2 g/1.) pH 7·0. This was sonicated twice at 4 °C for 2·5 min to break the cells. The sonicate was centrifuged at 30000g for 30 min and the supernatant used as the enzyme preparation. The assay mixture consisted of enzyme extract 50 μl; ATP 50 mM, MgCl₂, 4 mM, 50 μl; [6-³H]thymidine 0·19 mM, 325 μCi/ml, 5 μl; all in 0·2 M phosphate buffer pH 7·0. The assay mixture was incubated at 30 °C for 30 min and the reaction terminated by the addition of 0·3 ml cold methanol. The precipitate was removed by centrifugation and 150 μl of the supernatant applied to Whatman No. 1 paper for electrophoresis at 4 kV, 30 min in acetic acid:pyridine: water (4:40:756) pH 3·5 which separated the products, thymidine phosphates, from the unchanged thymidine which remained at the origin. Marker spots of thymidine and thymidine phosphates were detected by u.v. contact photography. The electrophoresis papers were cut into 4 × 1 cm strips and counted in Triton:toluene (1.2) scintillant containing 10% water. The vials were left 12 h to elute the radioactivity from the strips before counting on a Philips model A liquid-scintillation analyser. Protein concentration was determined by the absorbance at 260 and 280 nm (Warburg & Christian, 1941). Enzyme activity was calculated from percentage of counts in thymidine monophosphate over the total counts in that track of the electrophoretogram, as pmol thymidine monophosphate/min/mg protein. The enzyme activity was shown to be proportional to both enzyme amount and time over the assay period, and a zero value for the activity without enzyme present was subtracted.

Two replicate cultures of cell suspensions (25 ml) of S₄ and the mutant, SOB4 were incubated, with shaking, at 24±2°C. [6-³H]thymidine (10⁶ dpm/ml and 88 μM final concentration) was added to the cells in the medium in which they were growing. Samples (4 ml) were taken at 0, 1·5, 10, 30, 60 and 120 min and the medium removed by placing the cells on 4 layers of muslin under reduced pressure in a Millipore castle apparatus (Rubery & Sheldrake, 1973) for 30 s. The cells were then scraped into preweighed scintillation bottles and weighed. Triton:toluene scintillation fluid (10 ml) was added to each bottle and the cells were soaked overnight before counting. The cells were not washed, so the value at zero time represents the label bound to the cell wall. A calibration curve of dry weight against wet weight was obtained for both cell lines and the results expressed as cpm/mg dry weight of cells. Approximately 5 % of the total radioactivity was taken up during the course of the experiment.

From Fig. 1 it can be seen that the S₄ (normal) callus is inhibited by bromodeoxy-uridine concentrations greater than 3 μg/ml. Liquid suspension cultures of S₄ were therefore grown in 1 μg/ml and later 2 μg/ml bromodeoxyuridine and samples taken out and subcultured at intervals in 10 or 30 μg/ml bromodeoxyuridine. After 4 months’ growth a line of cells SOB4 was obtained which could grow in the presence of both 10 and 30μg/ml bromodeoxyuridine (Fig. 2). The slower growth of the S₄ callus in normal growth medium is due to its fairly recent origin from solid callus.

The SOB4 cells grown for 6 weeks on normal growth medium increased 30-fold in fresh weight and still retained their resistance to bromodeoxyuridine when transferred back to a medium containing it. Fifty calluses regenerated from protoplasts of SOB4 in the absence of bromodeoxyuridine all retained their resistance to it at 10 μg/ml. Thus the altered phenotype is heritable.

The cells of SOB4 and S₄ calluses were grown on HAT medium to test whether this medium would support the growth of normal callus and the changed callus. Aminopterin at 10⁻⁵M was chosen for the HAT medium after growing S₄ callus on varying amounts in normal growth medium (Fig. 3). The S₄ callus grows much better over a prolonged period in HAT medium than the SOB4 callus (Fig. 4). The difference is obvious after both 3 and 6 weeks’ growth. This confirms that there is an alteration in the ability to utilize thymidine in the SOB4 callus.

Two possible causes of the inability to utilize the thymidine in the medium are first that the enzyme thymidine kinase may be lacking in the mutant cells, and secondly that the mutant cells may be unable to take up thymidine from the medium. Both of these hypotheses were tested. Table 1 shows that the levels of the thymidine kinase are the same in both calluses. Fig. 5 demonstrates that both calluses are capable of taking up thymidine at very similar rates.

Protoplasts on isolation from both calluses were round, vacuolate, and exhibited cytoplasmic streaming (Figs. 6–8). Staining with Calcofluor white showed that there were often a few specks of fluorescent material left on the surface of the protoplasts, but no live whole cells. A few crumpled cell fragments passed through the 43-μm mesh and were brightly fluorescent. It was observed that preparing the protoplasts in 0·45 M sorbitol produced fewer multinucleate protoplasts than when they were prepared in 0·34 M sorbitol.

At concentrations of protoplasts greater than 5 × 10⁴ per ml wall regeneration and cell division were observed in the regeneration medium. This differs from normal medium in that it contains zeatin and higher levels of calcium. Walls were synthesized over the first 5 days by most protoplasts, the cells became ovoid and budding sometimes occurred; this latter phenomenon often led to cell death (Fig. 9). The first divisions occurred after 5 days (Fig. 10), giving rise in some cases (0·2 %) to sustained divisions and morphologically normal callus cells (Figs. 11–13). However a large proportion of cells failed to regenerate in this manner, and either died or became large and irregular in shape (Figs. 13, 14). Small clumps of cells about 1 mm in diameter were plated on to normal agar medium with pieces of nurse callus, and they then grew to morphologically normal calluses (Fig. 14) and were subcultured in the usual way.

Sodium nitrate was the first agent suggested for protoplast fusion (Power, Cummins & Cocking, 1970). It did not cause any detectable rate of fusion in sycamore protoplasts. However, maize root protoplasts and tobacco mesophyll protoplasts were, very rarely, seen to undergo fusion (Figs. 16A–F, 17A–E) in 0·30 M and 0·35 M NaNO₃ respectively. The time course of fusion varied from 10 min to 6 h. Protoplasts tightly stuck together with a double layer of chloroplasts (Fig. 18) were also seen, but they did not progress any further towards fusion.

Other agents which have been reported to cause cell fusion in animal cells are lysolecithin and glycerol mono-oleate (Croce, Sawicki, Kritchevsky & Koprowski, 1971; Lucy et al. 1971). Lysolecithin was extremely toxic to protoplasts causing rapid bursting in aqueous solutions at 200 μg/ml. This toxicity could be reduced by the presence of bovine serum albumin or lipid droplets formed from glycerol trioleate and lecithin. Glycerol mono-oleate in emulsion was also tried as a possible fusion agent for sycamore protoplasts but was not successful.

It is a prerequisite of hybridization experiments that the preparation of protoplasts and the regeneration of the population of protoplasts to a callus should be more or less a matter of routine. The results presented here go some way towards this aim for the sycamore callus system. Protoplasts from this callus could also be important experimentally in studies of cell wall composition and synthesis, as a very detailed model of the sycamore callus cell wall has been put forward (Talmadge, Keegstra, Bauer & Albersheim, 1973; Bauer, Talmadge, Keegstra & Albersheim, 1973; Keegstra, Talmadge, Bauer & Albersheim, 1973). The effect of zeatin in increasing the division frequency of cultured protoplasts parallels the effect noted by Mackenzie, Konar & Street (1972) on the initiation of division in low-density lag phase cultures of sycamore cells in liquid suspension. It probably indicates that at these densities of protoplasts they do not synthesize enough endogenous cytokinin to initiate cell divisions in the regenerated protoplasts. The higher calcium levels also improve protoplast viability as has been noted by Kao et al. (1973). The low yield of normal actively dividing cells in these experiments is in accord with the results with other protoplast systems (Potrykus & Durand, 1972), although with tobacco mesophyll protoplasts a 60% plating efficiency has been observed (Nagata & Takebe, 1971). The low yield coupled with the various forms of abnormality described (Figs. 9, 13, 14) illustrate that the process of regeneration is a complex process which can go wrong at many stages, and that the regeneration medium could be improved.

The methods available for fusion of protoplasts which we have tried give very low yields of fused products. Other chemical methods have been described (Kameya, 1973; Ito, 1973; Potrykus & Hoffmann, 1973; Kao & Michayluk, 1974) but none is as effective or as general as that achieved in animal cell fusion by the use of Sendai virus (Harris, Watkins, Ford & Schoefl, 1966). This then highlights the importance of an effective selection system.

The SOB4 line of cells is a mutant in that it is a stable heritable phenotypic change. The SOB4 callus is not a mutant lacking in thymidine kinase. The levels of the enzyme in both calluses are comparable to those found in dividing Jerusalem artichoke explants (Harland, Jackson & Yeoman, 1973). The SOB4 callus is not deficient in the ability to take up thymidine at the concentration range in which thymidine is present in the HAT medium. However, the cells are sensitive to the HAT medium. It thus seems that the mutant cells have some other defect in the uptake and utilization of thymidine. Mutant animal cell lines have been described which are defective in thymidine kinase (Littlefield, 1964) and thymidine transport (Breslow & Goldsby, 1969) but the SOB4 line is neither of these. Whether the SOB4 cells could be useful in hybridization experiments would depend on how the fusion product behaved with respect to its growth on HAT medium.

The selective enrichment technique we have used allows very rare mutant cells to take over the cell population as they are at a selective advantage in low concentrations of bromodeoxyuridine. Also bromodeoxyuridine is a mutagen which causes mistakes during DNA replication, so it would enhance the genetic variability of the cell population.

The pattern of growth of S₄ on the HAT medium indicates that the callus takes some time to adjust to the effects of the new medium. The callus on HAT, after 3 weeks of quite low growth, starts to grow at approximately the same rate as callus on normal medium. Presumably the cellular metabolism has adjusted to the use of exogenously supplied nucleoside precursors. However the S₄ growing just on aminopterin medium also begins to grow, which would suggest that the high degree of inhibition of increase in fresh weight observed here and of protein synthesis measured by Lescure (1969) in sycamore callus is not irreversible. Lescure (1973) has reported an azaguanine-resistant callus which might also be useful in the HAT system. However the basis of the resistance was not established.

In current work on cell hybridization with plant tissues the heterokaryons are detected microscopically, which means that the actual fusion event must be observed, or the protoplasts used must be very distinctive. This can be either by the presence of chloroplasts (Kao & Michayluk, 1974; Giles, 1972; Power & Frearson, 1973) or by using protoplasts prepared from flower petals (Potrykus, 1971). The only chemical selection system which has been successfully employed was that in the hybridization of mesophyll protoplasts from Nicotiana glauca × Nicotiana langsdorfii (Carlson et ai 1972). In that case the sexual hybrid plant and tissue cultures from it were available and it was known that the hybrid cells were unusual in not requiring the presence of an exogenous supply of auxin. This was then utilized by growing the protoplasts after fusion in medium lacking auxin to grow only hybrid cells. This elegant experiment has shown the practicability of asexual plant cell hybridization and it now remains to make it more widely applicable by a general selection method such as the HAT system, and improved techniques of protoplast fusion and culture.

One of us (S.W.J.B.) wishes to thank the Science Research Council for a grant during the tenure of which this work was carried out. We would also like to thank Mr L. Jewitt for assistance with the preparation of the photographic plates.