Tracheid formation in cultures of pine (pinus sylvestris)

Cell cultures of angiosperm species can be induced to differentiate by subculture into media that contain various growth factors in definite concentration ratios. Changes in various enzyme activities have been associated with the changes in cell organization (Haddon & Northcote, 1976) and in the case of phenylalanine ammonia lyase (EC 4.3.1.5) a correlation has been established in several cultured species between a sharp rise in activity and the onset of differentiation (Haddon & Northcote, 1975). Phenylalanine ammonia lyase has been identified as a key enzyme in the synthesis of phenyl propanoid compounds, among which are precursors of lignin, an important constituent of the secondary cell wall present in differentiated cells (Grisebach, 1977). Enzymes for cell wall polysaccharide synthesis have also been shown to undergo changes in activity during the differentiation of cultures and this is consistent with the differences in composition between primary and secondary cell walls (Dalessandro & Northcote, 1977). The gymnosperm secondary cell wall has a β-(1→4) glucomannan as a major constituent but this polysaccharide is surprisingly present in only small amounts in the walls of cultured cells, which probably resemble cambial cells. Xyloglucans are known to be important hemicelluloses in the primary wall of angiosperms but have not previously been identified in Pinus sylvestris, in which the major xylose-containing polysaccharide is an arabinoxylan (Garegg & Lindberg, 1960).

We have developed a system for inducing differentiation in cultured cells of/-’, sylvestris L., in which the examination of polysaccharide synthesis is more amenable to the experimental control of conditions during development. We report here the identification of conditions favouring tracheid formation, which is correlated with changes in phenylalanine ammonia Ivase activity and an altered sugar composition of the cell wall.

The cell culture was derived from callus grown out of stem explants from 4-week-old seedlings of P. sylvestris grown under sterile conditions. Cultures were maintained in the PRL-4 medium of Gamborg (1965) supplemented with 1-naphthalaneacetic acid (NAA; 10·7μM, 2mgl⁻¹), 2,4-dichlorophenoxyacetic acid (9·0 μM, 2mg l⁻¹) and sucrose (2%) (maintenance medium). Additions to the induction medium are specified in Results. For callus growth the medium was solidified by addition of 1% agar. The cultures were grown at 25 °C in the dark and suspension cultures were shaken at 50 revs min⁻¹.

All chemicals were of analytical grade. All water used was double glass-distilled. Agar was obtained from Difco Laboratories, Detroit, USA. Lignin Pink was supplied by BDH, Poole, UK.

The method used was that adopted for pine tissue by Ronald & Söderhäll (1985). Duplicate samples were estimated for all the experiments. Protein was measured by the method of Read & Northcote (1981). Cells were homogenized with sand in a mortar, together with an equal vol of 0·1 M-Tris HC1, pH8·8, containing 0·05% mercaptoethanol at 4°C.

Cells were harvested for analysis by filtration, washed with water to remove medium, freeze-dried and stored at — 20°C. Before use the cells were ground to a fine powder in a pestle and mortar. Cell wall polysaccharides were analysed after delignification by the acid/chlorite method (Wise et al. 1946) followed by Soxhlet extraction for 30 min with water and 15 min with acetone. Samples (10 mg) were then hydrolysed in 3% (w/w) H₂SO₄ at 120°C, 103 kPa for 1 h. The hydrolysates were neutralized with BaCO₃ and reduced by addition of 15 mg NaBH4 for 2 h at 25°C. The reaction was stopped by addition of glacial acetic acid, and boric acid was removed by repeated addition of methanol and evaporation to dryness. The alditols produced were acetylated with 1 ml acetic anhydride and 0·l ml H₂SO₄ for 1h at 6O°C and then extracted with dichloromethane, dried by rotary evaporation and analysed on a Pye Unicam series 104 gas-liquid chromatograph (g.l.c.) fitted with a l·5 m column of 3% SP2340 on Supelcoport 100–200 mesh, run at 230–240°C (Turnham & Northcote, 1982).

Samples of cells previously delignified were then extracted with 0·5M-sodium hexametaphosphate (60°C, 6h), dimethylsulphoxide (DMSO) (22°C, 16 h) and 0·5M-NaBH₄ (pH 9·1, 22°C, 4 h) to remove pectic wall material. A further extraction with 3% NaOH at 22°C for 6 h was used to obtain a hemicellulose fraction. The sugar composition of all the fractions was analysed and 80% of the total xylose residues was present in the sodium hydroxide extract. After neutralization with glacial acetic acid and dialysis against tap water for 12 h the polysaccharides were precipitated by addition of 4vol. of ethanol. These polysaccharides were chromatographed on cellulose columns (Whatman, CF1, 20mm × 70 mm) eluted consecutively with water, 8M-urea and 0·5M-NaOH for the purification of xyloglucan (Joseleau & Cham-bat, 1984). Fractions (1 ml) were collected and assayed for total carbohydrate by the phenol/sulphuric acid method (Dubois et al. 1956).

The polysaccharides were methylated by the procedure of Haworth & Percival (1932). The partially methylated products were dialysed for 12 h against running tap water and then re-methylated by the procedure of Hakomori (1964). These procedures were repeated alternatively for a total of six methylations. The methylated polysaccharide was hydro-lysed with 3% (w/w) H₂SO₄ for 1 h at 120°C, 103 kPa and neutralized with BaCO₃. The methylated sugars were identified by g.l.c. on a 3 m column packed with 3% SP234O on 100/120 Supelcoport run at 190°C. Material was collected at the exit from the column in a 0°C trap and this was used to measure the radioactivity in fractions corresponding to the peaks of methylated sugars (Heiniger & Delmer, 1977). The radioactivity was measured after addition of 2 ml of scintillant containing 4g PPO and 50 mg POPOP in 11 of toluene.

The spectra were acquired using a Bruker AM 400 WB nuclear magnetic resonance spectrometer operating at 100·6MHz for ¹³C. Nominal pulses were repeated at 4-s intervals with a sweep of 29·4kHz and a sweep size of 32 000. A 10 mm carbon-13 probe was used with a deuterium lock and gated broad band proton decoupling (Gorin, 1981). Samples were dissolved in ²H₂O.

Freeze-dried and triturated cells were extracted for 1 h with acetone and toluene in a Soxhlet apparatus. The wall lignin content was estimated by the method of Jeffs & Northcote (1966) using a guaiacol standard.

Cultures were examined using a Zeiss Ultraphot Mkll equipped for Nomarski differential interference contrast. For identification of lignified tissue preparations were stained with Lignin Pink and counterstained with Chlorazol Black (Cannon, 1941). The Lignin Pink fluorescence was detected using epifluorescence optics with an excitation filter Zeiss BP546/9 (green) and barrier filter LP59O (red). Cells were separated by sonication with a Dawes Soniprobe and counted in a Sedgewick Rafter plankton counting chamber.

Growth of the cultures was monitored by determination of the packed cell volume of cells subcultured into maintenance and induction media (Fig. 1). Cultures were routinely subcultured every 21 days and they retained their ability to produce tracheids after transfer to induction medium over many maintenance subcultures, even after 40 subcultures (approx. 2 years). Transfer of cells to induction medium resulted in increased growth and prolonged viability of the cells that did not differentiate to tracheids, probably due to the increased availability of carbon in the medium.

In order to identify a combination of growth factors capable of inducing cellular differentiation, cells were subcultured into one of a series of media (PRL4) containing 1, 2 or 5 mg l⁻¹ of NAA together with 0·2, 0·4 or l·0 mg l⁻¹ of kinetin. The time-course of phenylalanine ammonia lyase activity was followed and the majority of media used gave an induction peak of phenylalanine ammonia lyase (PAL) activity. This was not observed when the maintenance medium was used (Fig. 2). Controls containing only kinetin showed very poor growth and only limited induction of phenylalanine ammonia lyase. Controls containing only auxin showed no induction. The highest activity was seen in the induction medium with the highest levels of both auxin and cytokinin. This series of conditions was repeated with the addition of gibberellic acid (30 UM) to all the media, but no appreciable influence on phenylalanine ammonia lyase induction was observed. To define further the conditions for phenylalanine am-monia lyase induction, cells were grown in a series of media in which the concentrations of the growth factors, or sucrose, were varied (Figs 3, 4, 5). The activity of phenylalanine ammonia lyase was assayed only at the time of expected appearance of the induction peak, i.e. 17 days after subculture. The most favourable growth factor levels indicated were 10 mg 1⁻¹ NAA, 2mg l⁻¹ kinetin and 8% sucrose. The number of tracheids formed was also measured (Figs 3, 4, 5) and the same combination of growth factors that favoured phenylalanine ammonia lyase induction also gave the highest percentage of tracheids, except when sucrose concentration was varied. Subsequently a lower level of sucrose (6%) was adopted for the standard induction medium to avoid excessive accumulation of starch granules, which had an adverse effect on the formation of tracheids. The effect of the growth factors on phenylalanine ammonia lyase induction was not due to an alteration in the timing of the phenylalanine ammonia lyase peak, since the activity measured at a second time point, 21 days after subculture, was invariably lower than at 17 days. The maximum induction of phenylalanine ammonia lyase was, however, found at the same concentration of growth factors as when the activity was measured after 17 days.

As the cultures aged after transfer to fresh medium, their colour darkened from the original light brown and there was an accumulation of pigment in the cells. This change could be observed with both maintenance and induction media, but took place more rapidly in the latter. Cells in high sucrose media also accumulated starch granules. In appearance the cells on maintenance medium were normal tissue-cultured cells and were generally large and vacuolate with thin walls throughout the life of the culture. Cells with differentiated walls were never observed. Differentiated walls were observed in cells in the induction medium that gave the maximum induction peak of phenylalanine ammonia lyase after approximately 25 days (Fig. 6). At this time tracheids became visible, with thickened walls penetrated by numerous pits, some of which resembled the bordered pits of mature pine tracheids. Some cells exhibited rows of pits more regularly arranged over the cell surface, while other cells showed only a random arrangement. Infrequently cells were observed with thickened walls but no pits. The cultures usually formed small aggregates but differentiation was found to occur equally in the centre and at the periphery of clumps and no tendency towards organized growth was observed. The pitted cells exhibited birefringence under polarized light and showed appreciable fluorescence when stained with Lignin Pink. Cultures stained with Lignin Pink and counter-stained with Chlorazol Black showed uptake of the stain mainly in the pitted cells.

Cells were taken from six different cultures, 30 days after subculture, for analysis of the cell wall sugar composition. The cultures had been grown in pairs using either induction medium or maintenance medium. The results are given in Fig. 7. The most appreciable difference in wall composition is the twofold increase in mannose content observed in cells from induction medium compared with cells from maintenance medium. The neutral sugar composition of xylem from a 20-year-old pine tree was examined for comparison with the cultures. Fig. 7 shows that the cultured pine cells contain higher levels of arabinose and galactose, even after induction, than is the case with pine xylem isolated from pine stems. This can be attributed to the pectins of the primary walls of undifferentiated cells remaining in the culture.

The walls of cells cultured in induction medium were compared with those of cells from maintenance medium and an increase in lignin content was observed. The values obtained from three cultures in three separate experiments, harvested after 30 days growth, were generally slightly higher (4·6% as a percentage of total dry tissue weight) in induction medium than in maintenance medium (3·5%).

Polysaccharides obtained by extraction of suspension-cultured cells with 3% NaOH were separated on a cellulose column. The first fraction eluted with water had a high percentage of arabinose and contained mainly residual pectin and some water-soluble xylam Later fractions obtained by further elution with water and elution with urea and NaOH were preponderately composed of xylose and glucose, and were combined for further examination. Methylation of the polysaccharide yielded only the 2,3,6-tri-O-methyl glucose and a similar amount of dimethyl glucose. Of the methyl xylose recovered 80% was present as trimethyl xylose and 20% as dimethyl xylose. The methyl glucose derivatives formed are consistent with the occurrence of a β-(1→4)-linked glucan chain substituted with single xylose units. The presence of dimethyl xylose may be due to further substitution of the xylose side groups or the presence of a homoxylan.

The ¹³C n.m.r. spectra was recorded from a sample of the polysaccharide (Fig. 8). The major C-l peak at 102·5p.p.m. (parts per million) indicated that an appreciable proportion of the xylose present was μ-(1→4)-linked, but the higher field peak at 99·04 p.p.m. corresponded to the C-l of xylose α-(l→6)-linked to glucose. The glucose C-l signal appeared at 103·22p.p.m. in accordance with β-(l→4) linkage of a xyloglucan backbone chain (Mori et al. 1980). That substitution took place at the C-6 of glucose can be confirmed by examination of the high field signals for the 6-position carbons. The peak at 60–88 p.p.m. was assigned to the C-6 of unsubstituted glucose residues and the peak at 63·79 p.p.m. was due to downfield shifting of substituted glucose C-6 atoms (Mori et al. 1980) and suggests that a major contribution to the spectrum was due to α-xylopyranosyl residues (1→6)-linked to a β-(l→4) glucopyranosyl chain. Comparison of the areas for the two xylose C-l peaks indicated that 28% of the xylose present participated in a α-(l→6) links to glucose.

It has been demonstrated that alteration in the growth regulator balance of the medium can induce phenylalanine ammonia lyase activity. Cells grown in a medium selected on the basis of its ability to cause the induction of phenylalanine ammonia lyase were observed to form tracheids with thickened walls, and thus confirmed that differentiation had been induced. Previous studies on tracheid formation using conifer tissue cultures have identified effects of temperature and light (Durzan et al. 1973) and boric acid (Webb, 1982), but no relationship with growth regulator levels has been demonstrated. The importance of cytokinins for the induction of differentiation has been observed in other conifer cultures and this is supported by observations on intact plants, where there is a correlation of cambial activity with seasonal fluctuations in cytokinin levels in P. sylvestris (Kubowicz, 1979). However, there has been a lack of similar emphasis on the role of auxins in differentiation of cultures, although auxins have also been shown to vary seasonally in P. sylvestris (Savidge & Wareing, 1984). Possibly this was because many of the studies have been concerned with differentiation from relatively large masses of preformed tissue so that there could be effects bought about by the action of endogenously synthesized regulators. In the suspension culture we have examined, the presence of auxin was essential for the maintenance of growth, and without auxin the induction of phenylalanine ammonia lyase was very poor. The significance of the high sucrose levels that was found to favour differentiation is less clear, since sucrose levels are known to decrease in spring when cambial activity is highest (Holl, 1985) and sucrose can also inhibit the rooting of P. sylvestris callus (Toribio & Pardos, 1982). In angiosperms it has been suggested that xylem and phloem differentiation occur in response to an auxin/sucrose balance (Minocha & Halperin, 1974), but Aloni (1980) has demonstrated that the differentiation of sieve and tracheary elements is dependent upon auxin alone. This was not true for the suspension cultures used in this work, since tracheid formation required the presence of both sucrose and auxin at levels that were previously stated to give rise to phloem.

Cells cultured in maintenance medium, which did not differentiate, had very little mannose present in their walls. During differentiation and secondary thickening the glucomannan content of the wall increased. This glucomannan has been identified by its isolation and n.m.r. spectroscopy (data not shown) and was similar to that isolated from the xylem of pine trees.

A xyloglucan was identified in the primary walls of the suspension cultured cells of pine. Xyloglucans have been extensively identified in angiosperms but there have been few investigations of their presence in gymnosperms. The suggested degree of substitution of 0·5 is similar to, but slightly lower than, that of xyloglucans from angiosperms, in which values between 0·8 and 0·5 are common (McNeil et al. 1985). The attachment of further sugars to xylose side-chains has been reported to involve arabinose, galactose and fucose. Arabinose and galactose were both present in the polysaccharide preparation and the C-l signals for these sugars were detected. However, the arabinose signals were not indicative of monsubstitution to xylose residues but rather, corresponded to signals characteristic of (1→5)-linked arabinan (Joseleau et al. 1977), which could have been present as a contaminant in the preparation. It is possible that growth in culture favoured the deposition of xyloglucan but it is improbable that the synthesis of a polysaccharide not previously found in the wall was initiated. Thus it can be suggested that the hemicellulose of the pine primary wall contains a xyloglucan, which is replaced by glucomannan in the secondary wall.

We are grateful to Robin Davey for skilled assistance with the gas-liquid chromatography, and to Peter Morris of the Cambridge Biochemical NMR group for advice on NMR spectroscopy. One of us (L.R.) acknowledges financial assistance from a SERC research studentship.