Planta (1990)180:285-292
P l a n t a 9 Springer-Verlag 1990
Patterns of expression of the JIM 4 arabinogalactan-protein epitope in cell cultures and during somatic embryogenesis in Daucus carota L. Nicola J. Stacey, Keith Roberts, and J. Paul Knox* Department of Cell Biology,John Innes Institute, Colney Lane, Norwich, NR4 7UH, UK
Abstract. Spatiotemp0ral patterns of expression of the cell-surface arabinogalactan-protein epitope defined by monoclonal antibody JIM4 (J.P. Knox et al., 1989, Development 106, 47-56) have been characterized by indirect immunofluorescence during the process of somatic embryogenesis in Daucus carota L. The JIM4 epitope (J4e) occurred on cells established in culture from hypocotyl explants which appeared to derive, at least in part, from the epidermal cells of the hypocotyl. Cultures maintained in the presence of 2,4-dichlorophenoxyacetic acid developed proembryogenic masses of which only infrequent cells at the surface expressed J4e. Sub-culture at a low cell density and withdrawl of the synthetic auxin resulted in an increase in J4e expression in most surface cells and most abundantly in surface layers of cells at the future shoot end of developing embryos. The transition to heart-shaped embryos occurred concurrently with the expression of J4e by groups of cells beneath the developing cotyledons, at the junction of the future root and shoot. At this stage, J4e was also expressed by a single well-defined layer of cells at the surface of the embryos. Advancement to the mature torpedo stage was accompanied by the expression of the epitope on cells forming two regions of the future stele and of cells associated with the cotyledonary provascular tissue characteristic of the carrot seedling. At this stage there was substantially less expression of the marker antigen by epidermal cells, although infrequent expression by isolated cells of the epidermis was maintained. The correlation of J4e expression with the development and distinction of plant tissue patterns during somatic embryogenesis indicates a role for plasma-membrane arabinogalactan proteins in these processes. * To whom correspondence should be addressed Abbreviations: AGP=arabinogalactan protein; 2,4-D=2,4-dichlorophenoxyacetic acid; J4e = JIM 4 epitope; PEM = proembryogenic mass
Key words: Arabinogalactan protein - Cell surface Daucus (somatic embryogenesis) - Embryogenesis (somatic) - Epitope (JIM4)
Introduction The recent derivation of a monoclonal antibody, JIM4, that recognizes an antigen showing position-specific expression during the early stages of the development of the carrot root apex provides a molecular marker of plant developmental processes (Knox et al. 1989). JIM4 recognizes an epitope (J4e), occurring on a set of plasmamembrane-associated arabinogalactan proteins (AGPs), that in the carrot root apex is expressed by two sections of the developing pericycle, other stele cells related by position, and epidermal cells (Knox et al. 1989). The function of these cell-surface glycoproteins is as yet unknown. To gain further insight into the possible role of these molecules during the acquisition of plant form and early developmental processes we have studied the expression of J4e during carrot somatic embryogenesis. The in-vitro formation of embryos has been studied extensively since the initial observations by Steward et al. (1958) and Reinert (1959). As yet, the molecular processes whereby cells are capable of entering culture as single cells and subsequently enter pathways of organization leading to viable embryos are unknown. No cell-surface molecules marking spatiotemporal patterns that reflect the acquisition of plant form during somatic embryogenesis are known (Sung et al. 1984). In this report we describe the expression of J4e during the formation of continuous cultures of callus cells from hypocotyl explants and also the spatiotemporal patterns of expression, reflecting the establishment of tissue patterns during the subsequent, hormonally controlled, formation of somatic embryos from the proembryogenic masses (PEMs) in these cultures.
286
Material and methods Plant mater&l. Carrot (Daucus carota L. cv. Early Nantes) seeds (W.W. Johnson & Son, Boston, Lincs., UK) were sterilized and germinated on moist filter paper. After 7-10 d, hypocotyls were excised, cut into 1-mm sections and incubated in Gamborg's B5 medium (Flow Laboratories, Rickmansworth, Herts., UK) supplemented with 2 ~tM 2,4-dichlorophenoxyacetic acid (2,4-D) and placed upon a rotary shaker (80 rpm). The methods of De Vries et al. (1988) were followed to maximize embryo formation. Hypocotyl explants were removed 19 d after culture. Cultures of suspended cells, in which PEMs developed, were maintained in medium containing 2,4-D and subcultured by a fivefold dilution every 10-14 d. Embryos were formed when samples, containing PEMs, were withdrawn (at least one week after the most recent subculture), washed extensively with and diluted into medium lacking 2,4-D. A low cell density of cells was produced by addition of 5 Ixl (settled cell volume) of cells into 5 ml of medium in a Petri dish (50 mm diameter). All cultures of carrot seedlings and cells were maintained at 25~ C in the dark. Protoplasts were prepared from the suspended single cells by overnight incubation with 0.5% (w/v) Driselase in 0.5 M mannitol and their reactivity with JIM4 determined by indirect immunofluorescence. Indirect immunofluorescence with the monoclonal antibody JIM4. Plant material from all stages of culture was removed from culture medium and immediately fixed in 4% (w/v) formaldehyde in a buffer containing 50mM 1,4-piperazinediethanesulfonic acid (Pipes), 5 mM MgSO4 and 5 mM ethylene glycol bis (/%aminoethylether)-N,N,N',N'-tetraacetic acid (EGTA), pH 6.9. Tissue was fixed for at least 1 h, washed briefly in water and mounted in OCT compound (Agar Aids, London, UK) and sectioned at - 1 0 ~ with a Bright 5030 Cryomicrotome (Bright Instrument Co., Huntington, Cambs., UK) Sections (7-10 ~tm) were collected on poly-L-lysine-coatedmulti-well slides and stored dry until use. The derivation and characterization of the anti-AGP monoclonal antibody have been described elsewhere (Knox et al. 1989). The sections were incubated for 5 min with 5% (v/v) calf serum in phosphate-buffered saline (CS/PBS) prior to incubation with a fivefold dilution of JIM4 hybridoma culture supernatant in CS/ PBS for 1 h. Wells were washed three times with PBS and then incubated with anti-rat-IgG,A,M-fluorescein isothiocyanate (ICN Biomedicals, Irvine, Cal., USA) at a 150-fold dilution in CS/PBS for a further I h. A final, extensive washing in PBS preceded a 30-s incubation with 4'6-diamidino-2-phenylindole(DAPI) at 1 ~tg. ml-1 in PBS, mounting in Citifluor antifade (Citifluor Ltd., City University, London, UK) and examination with a Zeiss (Oberkochen, FRG) Photomicroscope III. The DAPI is reactive with DNA and its fluorescence was observed at 45~490 nm to locate all cells of the treated sections. Fluorescein isothiocyanate fluorescencewas observed at 520-560 nm and photographed using Kodak (Hemel Hempsted, Herts., UK) TMAX400 film. Controls of no antibody or alternative monoclonal antibody culture supernatant (MAC 207, Pennell et al. 1989) were run and always observed to confirm that the distinct patterns of immunofluorescence shown and described were due to JIM4. Results Formation o f cell cultures f r o m carrot seedling hypocotyls. Expression of J4e was studied during the establishment, from carrot hypocotyl explants, of cultures with embryogenie potential and also during the development of somatic embryos as described by De Vries et al. (1988). The expression of J4e by cells of the hypocotyl at the time of excision of explants occurs in the epidermal cells
N.J. Stacey et al. : Expression of arabinogalactan epitope and two regions of the vascular tissue (Fig. | A; Knox et al. 1989). The anatomy of the hypocotyl is characteristic of the root, with the transition to a vasculature characteristic of the shoot occurring only in the upper hypocotyl (Esau 1940). Six days after incubation of hypocoty[ explants in B 5 medium containing 2 laM 2,4-D, the epidermal layers were seen to be dissociated from the cortical regions and floating freely as sheets of cells (Fig. 1 B). Cells at the periphery of these sheets appeared to contribute to the culture as single large vacuolate cells. The preparation of protoplasts from these cells and reaction with J I M 4 indicated that they had maintained their expression of J4e during their transformation from epidermal into cultured cells (data not shown). At the early stages of culture all the suspended cells were of this large vacuolate morphology. After approximately 10 d of culture, callus, consisting of small highly cytoplasmic cells, was observed at the cut ends of the explants (Fig. 1 D). Cells of this callus, derived from subepidermal cells, were unreactive, or only very weakly reactive, with J I M 4 (Fig. 1 C). However, at the time of removal of the hypocotyl explants (day 19) the suspended cells, maintained in 2,4-D, were predominantly, if not exclusively, o f the large vacuolate type. Protoplasts prepared from cultures at this stage were all highly reactive with JIM 4. Twelve days after removal of hypocotyl explants from the culture of cells (day 31 of culture), small clusters of cells were noted in the medium and these displayed a dramatic change in J4e expression. Indirect immunofluorescence due to J I M 4 was only observable in isolated and infrequent cells at the surface of the clusters of small, tightly adhered cells. Representative examples are shown in Fig. 2A, C. These clumps are the proembryogenic masses (PEMs) of Halperin (1966), cells on the surface of which are capable of developing into embryos. At this stage, and all later stages, the single larger vacuolated cells, also present in the culture, displayed abundant (as determined by intensity of fluorescence) expression of J4e (not shown). The cultures were maintained for several months in this state by the presence of 2,4-D, with an increasing predominance of PEMs over the single vacuolate cells. The size of the PEMs varied between sets of established cultures but in all cases the JIM 4 epitope was expressed by only occasional cells at the surface of a PEM. A reduction to a basal level of infrequent J4e expression in the PEMs, represented by Fig. 2C, was observed to occur over time.
Formation o f somatic embryos. Withdrawl of the synthetic auxin and culture at low cell density resulted in the development of torpedo-stage embryos within 10-14 d. These conditions resulted in a dramatic increase of expression of J4e by cells of the PEMs within 24 h. All or most of the cells at the surface of the PEMs were strongly fluorescent due to J I M 4 binding 24 h after the commencement of the conditions allowing embryogenesis (Fig. 2B, D). Induced expression in larger, ramiform PEMs was commonly restricted to cells on the distal
N.J. Stacey et al. : Expression of arabinogalactan epitope
287
Fig. 1 A-D. Carrot hypocotyl explants cultured in Gamborg's B 5 medium supplemented with 2 I.tM 2,4-D. A Transverse section of the hypocotyl of a carrot seedling at the beginning of culture showing immunofluorescence due to JIM4 on epidermal cells (e) and two sectors of the vascular cylinder (v, the arrowheads indicate the centre of the sectors and the xylem axis). JIM4 did not react with cells of the cortex (c). B White-light image of a sheet of JIM 4reactive epidermal cells dissociated from hypocotyl explants 6 d after the beginning of culture. Arrowheads indicate the region con-
tributing large vacuolate cells to the culture of suspended cells. C Indirect immunofluorescence image showing the weak reactivity of JIM4 with a section of callus (arrowheads) derived from the culture of hypocotyl explants. Small double arrowheads indicate highly JIM 4-reactive cells of hypocotyl explant debris. D A DAPIstained image of nuclei of the section of callus shown in C to indicate its composition of small tightly adhered cells. Bars= 100 ~tm
surface o f the branching structures as shown in Fig. 2 D. In a study o f the timing of this increase in AGP-epitope expression the first increase occurred between 12 and 24 h after removal of the 2,4-D (data not shown). It is well established that embryos can develop from cells at the surface of the P E M s (Halperin 1966; De Vries et al. 1988). The earliest stages in the development of globular embryos were difficult to locate, but most surface cells and small groups of cells on the surface of P E M s displayed abundant J4e expression at this time. It has often been observed that several embryos can develop concurrently, linked by the future root poles through a central cluster of cells of the former PEM. This indicates that the establishment of the root-shoot polarity is a very early event and occurs in spatial rela-
tion to the PEM. At the globular stage the established root-shoot polarity is reflected by the J4e expression in several surface layers centred upon the developing shoot end of the embryo (Fig. 3A). The subsequent development of embryos was associated with distinct spatial patterns of expression of J4e. As embryos progressed to the heart-shaped stage the expression was restricted more clearly to one, well-defined, surface layer of cells and the first internal region of J4e expression was seen, in longitudinal section, as two regions o f expression reflecting the position of the ridges of the emerging cotyledons (Fig. 3 C, D). The internal expression of J4e appeared to arise as two distinct groups of cells, representing the cotyledonary node and the future root-shoot junction. These cells did not ap-
288
Fig. 2A-F. JIM4-reactive cells in proembryogenic masses derived from the culture of carrot hypocotyl explants. A Immunofluorescent image of J I M 4 reaction with cells in a section of a PEM produced in the presence of 2,4-D from a culture of suspended cells derived from hypocotyl explants. This PEM was isolated 10 d after the removal of the explant debris. Reactive cells were on the surface of the PEM (arrowhead) and the central cells were unreactive (u). B After 24 h of culture at low cell density in the absence of 2,4-D, a PEM equivalent to that shown in A displayed more abundant fluorescence due to J I M 4 in surface cells but cells in the centre remained unreactive (u). C J I M 4 reaction with PEMs, maintained in medium with 2,4-D, approx, six weeks after the
N.J. Stacey et al. : Expression of arabinogalactan epitope
removal of the explant debris. Infrequent isolated surface cells expressing J4e are indicated by arrowheads. D Section through a PEM equivalent to that of B, indicating that cells expressing J4e 24 h after removal of 2,4-D and culture at low cell density were restricted to distal surfaces of a ramiform PEM. Cells at the centre of composite clumps (u) and of the branching PEM (m) remained unreactive. E A negative control section, equivalent to D, from which J I M 4 treatment was omitted. F Another control section treated with a different antibody, M A C 207, in the place of JIM 4, showed reaction with all cells. B a r s = 100 lam
N.J. Stacey et al. : Expression of arabinogalactan epitope
289
Fig. 3A-I. Patterns of J4e expression by carrot embryos induced in cultures of PEMs subcultured at low cell density in the absence of 2,4-D. A Longitudinal section of a late globular-stage embryo, r = root pole, s= shoot pole. Reactive cells in the surface layers centre on the shoot end. B A longitudinal section of an embryo at the early heart-shaped stage indicated two groups of internal cells reactive with JIM4, reflecting the position of emerging cotyledonary ridges (arrowhead). C Longitudinal section through a late heartshaped embryo. The J4e was expressed by a single defined layer of epidermal cells at the shoot apical end of the embryo. D, E At the early torpedo stage the cell groups beneath the emerging cotyledons were welt established and provascular tissues (of the future stele) also expressed J4e. F A longitudinal section through a mature torpedo-stage somatic embryo in which J4e is expressed in a manner corresponding to the carrot seedling. Position and angle of arrows indicate sections from an equivalent embryo shown in G-I. G Transverse section showing two sectors of provascular tissue that expressed J4e (large arrowhead). Isolated epidermal cells also expressed J4e (small arrowhead). H Transverse section showing a continuous circle of expression at the cotyledonary node. I Transverse section showing commencement of the ramification of provascular tissue into the cotyledons, a= cells of the future shoot apex which express J4e. Bars= 100 ~tm pear to be anatomically distinct at this stage. In the mature embryo the cells of this region expressing J4e form a complete circle (see below and Fig. 3 H). As the embryo developed further, J4e expression was observed in the provascular tissue of the future stele. This was in two regions reflecting the two sectors of the future stele (Fig. 3 D, E, G), related to the position of the diarch vascular structure in the mature root, and then as the cotyledons developed J4e expression occurred on cells associated with cotyledonary provascular traces (Fig. 3 F, I). At this stage the circle of expression at the root-shoot junction can be clearly discerned (Fig. 3 H), and also its correlation with ramifying provascular traces in the cotyledons (Fig. 3 F, I). A region of expression in surface layers at the site of the future shoot meristem was often observed (Fig. 3D, E, I). In
the mature embryo the epidermal layer o f cells did not uniformly express J4e, but infrequent groups of isolated cells continued to do so.
Developmental arrest and J4e expression. In certain cases, cultures maintained in the presence of 2,4-D at a lower density of cells can produce globular embryo-like structures, generally larger than the globular embryos developed from PEMs after 2,4-D withdrawl (Bokird et al. 1986). These are arrested embryos and subsequent withdrawl of 2,4-D results in organization directly into shoot and root meristems in a highly irregular way, resulting in monstrous embryos. The frequency of such anomalies has previously been noted (Halperin 1966; Ammirato 1987). Such arrested globular embryos maintained in 2,4-D have very few cells displaying J4e expression, often
290
Fig. 4. A Arrested globular stages of carrot embryos observed in cultures maintained for 10 d in medium containing 2,4-D. JIM4 was reactive with only a few cells. However, a potential root (r) and shoot (s) polarity does occur in relation to the centre of the PEM (m). B A section through equivalent tissue 24 h after culture at low cell density in medium containing no 2,4-D. J4e expression had increased and was most abundant towards the shoot pole of future embryos. Bars= 100 lam
none in a section or only several isolated weakly labelled cells (Fig. 4A). Removal of 2,4-D and subculture at low cell density resulted in expression of J4e by cells in the surface layers with decreasing abundance of expression by more proximal cells (Fig. 4B). As these structures developed further into larger and often irregular embryos, patterns of J4e expression correlating with the morphology-expression relationships seen iri normal embryos were observed (not shown).
Discussion We have no information on the function of the antigens bearing J4e and only limited knowledge of their biochemical nature. They are plasma-membrane-associated AGPs, and J4e appears to define a subset, or represent a modification of, a larger set of membrane-associated AGPs (Knox et al. 1989; Pennell et al. 1989). Arabinogalactan proteins have received considerable attention, but the function of this class of molecules has not yet emerged (Fincher et al. 1983). However, it is an important observation that AGPs associated with the plasma membrane of plant cells can contain information, in terms of epitope restriction to distinct tissues, or groups of cells, during early stages of plant development. We have used somatic embryogenesis, which is accessible to manipulation and hormonal control, to further study the expression of J4e. The patterns of expression of J4e during the develop-
N.J. Stacey et al. : Expression of arabinogalactan epitope
mental processes described in this report can be put into the two categories described below.
Establishment of the pattern of expression characteristic of the carrot seedling. In this first category of J4e expression, the establishment of the pattern of expression characteristic of the mature embryo and seedling (Knox et al. 1989), the epitope is expressed by two distinct regions of the future pericycle and associated cells, cells associated with future vascular traces in the cotyledon, isolated groups of epidermal cells and a group of cells at the cotyledonary node. Anatomical studies have been carried out upon developing somatic embryos of carrot (Halperin and Wetherall 1964; Halperin 1966; McWilliam et al. 1974; Schiavone and Cooke 1985; Ammirato 1987). The protoderm is the first tissue to be clearly defined, at the late globular stage, and the formation of the cotyledonary node is immediately subsequent to that of the protoderm. The provascular tissue of the future hypocotyl first appears during the transition to the heart-shaped stage and then differentiates acropetally. The somatic embryo reaches the maturity of its torpedo stage with the appearance of provascular tissue in the cotyledons. The expression of J4e seems to be correlated with this sequence by its appearance in the outer surface layers of cells at the globular stage, its restriction to a defined single layer at the heart-shaped stage, and its expression by cells associated with the future root-shoot junction (the cotyledonary node). Finally, J4e expression was associated with provascular tissue in the future hypocotyl and then in the cotyledons. The transition from radial to bilateral symmetry of the embryo (at the heart-shaped stage) is reflected in J4e expression by two groups of internal cells beneath the emerging cotyledons and locates the developing cotyledonary node. At the torpedo stage the bilateral symmetry is reflected in the bilateral expression of the provascular tissues. These correlations of J4e expression with the emerging anatomical patterns have been made by observation of the position of cell collec-
N.J. Stacey et al. : Expression of arabinogalactan epitope tives expressing J4e rather than by the precise observation of distinctive structural features. The timing o f the expression o f J4e by these groups of cells, in relation to known sequences of embryo development, indicate, as in the root meristem (Knox et al. 1989), that expression of this A G P epitope occurred at the time of the formation o f these distinct tissues. The decrease in J4e expression by surface layers o f the developing embryo during its transition from the globular to the torpedo stage is striking and may reflect an important role for surface cell layers during the early events of organization (see below). The reduction in expression may be caused by the shedding of cells or by switching off expression. The shedding of surface cells has been noted for all stages of embryos (Halperin 1966) and in certain cases highly JIM4-reactive cells have been noted to be loosely associated with the surface of embryos. Establishment o f cultures and proembryogenic masses.
The second category of JIM4-epitope expression concerns earlier events of embryogenesis, before the development of tissue boundaries, and during the establishment o f plant cells in culture. These patterns of expression indicate a role for plasma-membrane AGPs at the earliest stages of organization and of the distinction of cells within cell clusters. When the large vacuolate cells derived from JIM4reactive epidermal cells, and thought to be non-embryogenic, were cultured their reactivity with J I M 4 was maintained. Such cells are distinct in morphology from the PEMs and also in terms of J4e expression. Only infrequent isolated cells at the surface of the PEMs expressed J4e, and often only weakly in terms of intensity of fluorescence. The origin of PEMs, formed in increasing abundance as a result of continued culture in medium containing 2,4-D, is unclear. It is not known whether the PEMs are produced from of the large vacuolate cells or from the small highly cytoplasmic cells which are derived from callus proliferated at the cut ends of the hypocotyl explants and exist at a low frequency at the time of removal of the hypocotyl explants (Nomura and Komamine 1985; De Vries etal. 1988). The callus formed at the cut edge of explants, and presumably adding cells to the medium, is more characteristic of PEMs in terms of cell morphology and J4e expression. Embryos are thought to arise from single cells or groups o f cells at the surface of PEMs, but the relationship between them and J4e expression has not been determined. The mechanisms and processes whereby cells are coordinated to initiate somatic embryos are uncertain (Williams and Maheswaran 1986). The dramatic increase in J4e expression upon withdrawl of 2,4-D from the medium and culture at low cell density further indicates a developmental role for AGPs. The expression of J4e by surface cells of a cell mass identifies these cells in terms of position. The presence of 2,4-D and a high cell density appears to inhibit the full expression of this positional identity (as well as the development of embryos). The nature of this
291 switch, whether by existing cells or new daughter cells, has not been clarified. This distinction of surface cells may be required for the future development of organized structures at the surface of a PEM and also during the early globular and heart-shaped stages of embryo formation when cells o f the surface layers abundantly express J4e. In this latter case, surface cells o f developing embryos are clearly defined by anticlinical divisions and are distinct from the callus-like cells expressing J4e at the surface of a PEM. However, in both cases J4e is a marker of surface position and may reflect common developmental processes. The case of developmentally arrested (by 2,4-D) globular-stage embryos and the increase in expression of J4e prior to their continued development indicate that existing cells can be induced to express J4e. These observations also point to a functional, possibly regulatory, role for cell-surface AGPs. The observed gradient of expression, decreasing internally and distally upon release from arrest, further indicates that the antigens recognized by J I M 4 are positionally expressed in relation to organizing clusters of plant cells. In summary, the expression of J4e, as characterized so far, provides a marker for all stages of the acquisition of the major tissue patterns during carrot somatic embryogenesis. The ubiquity of differential expression of J4e by cells at all stages from PEMs to the mature embryo indicates that plasma-membrane AGPs are functionally concerned with the position of cells in relation to emerging plant form, since they are expressed in terms of surface position in cell clusters and in later developmental events are associated with the distinction of tissue blocks. This class of antigens does not appear to be tissue-specific, concerned with specific differentiation pathways, but seems to be a class of cell-surface glycoproteins concerned with the establishment of positional distinctions between cells and of tissue patterns within which cytodifferentiation processes can be contained. We thank Andrew Davis for photographic assistance and Roger Pennell for useful discussions. References Ammirato, P.V. (1987) Organizational events during somatic embryogenesis. In: Plant tissue and cell culture, pp. 57-81, Green, C.E., Somers, D.A., Hackett, W.P., Biesboer, D.D., eds. Alan R. Liss, New York Bokird, C., Choi, J.H., Sung, Z.R. (1986) Effect of 2,4-dichlorophenoxyacetic acid on the expression of embryogenic program in carrot. Plant Physiol. 81, 1143-1146 De Vries, S.C., Booij, H., Meyerink, P., Huisman, G., Wilde, H.D., Thomas, T.L., van Kammen, A. (1988) Acquisition of embryogenic potential in carrot cell-suspension cultures. Planta 176, 196-204 Esau, K. (1940) Developmental anatomy of the fleshy storage organ of Daucus carota. Hilgardia 13, 175-226 Fincher, G.B., Stone, B.A., Clarke, A.E. (1983) Arabinogalactan proteins: structure, biosynthesis and function. Annu. Rev. Plant Physiol. 34, 47-70 Halperin, W. (1966) Alternative morphogenetic events in cell suspensions. Am. J. Bot. 53, 443453
292 Halperin, W., Wetherall, D.F. (1964) Adventive embryony in tissue cultures of the wild carrot, Daucus carota. Am. J. Bot. 51, 274-283 Knox, J.P., Day, S., Roberts, K. (1989) A set of cell surface glycoproteins forms an early marker of cell position, and not cell type, in the root apical meristem of Daucus carota L. Development 106, 47-56 McWilliam, A.A., Smith, S.M., Street, H.E. (1974) The origin and development of embryoids in suspension cultures of carrot (Daucus carota L.). Ann. Bot. 38, 243-250 Nomura, K., Komamine, A. (1985) Identification and isolation of single cells that produce somatic embryos at a high frequency in a carrot suspension culture. Plant Physiol. 79, 988-991 Pennell, R.I., Knox, J.P., Scofield, G.N., Selvendran, R.R., Roberts, K. (1989) A family of abundant plasma membrane-associated glycoproteins related to the arabinogalactan proteins is unique to flowering plants. J. Cell Biol. 108, 1967-1977
N.J. Stacey et al. : Expression of arabinogalactan epitope Reinert, J. (1958) Morphogenese und ihre Kontrolle an Gewebekulturen aus Carotten. Naturwissenschaften 14, 344-345 Schiavone, F.M., Cooke, T.J. (1988) A geometric analysis of somatic embryo formation in carrot cell cultures. Can. J. Bot. 63, 1573-1578 Steward, F.C., Mapes, M.O., Mears, K. (1958) Growth and organized development of cultured ceils. II. Organization in cultures grown from freely suspended cells. Am. J. Bot. 45, 705 708 Sung, Z.R., Feinberg, A., Chorneau, R., Borkird, C., Furner, I., Smith, J., Terzi, M., LoSchiavo, F., Giuliano, G., Pitto, L., Nuti-Ronchi, V. (1984) Developmental biology of embryogenesis from carrot culture. Plant Mol. Biol. Rep. 2, 3-14 Williams, E.G., Maheswaran, G. (1986)Somatic embryogenesis: factors influencing coordinated behaviour of cells as an embryogenic group. Ann. Bot. 57, 443-462 Received 14 July; accepted 21 September 1989
P l a n t a 9 Springer-Verlag 1990
Patterns of expression of the JIM 4 arabinogalactan-protein epitope in cell cultures and during somatic embryogenesis in Daucus carota L. Nicola J. Stacey, Keith Roberts, and J. Paul Knox* Department of Cell Biology,John Innes Institute, Colney Lane, Norwich, NR4 7UH, UK
Abstract. Spatiotemp0ral patterns of expression of the cell-surface arabinogalactan-protein epitope defined by monoclonal antibody JIM4 (J.P. Knox et al., 1989, Development 106, 47-56) have been characterized by indirect immunofluorescence during the process of somatic embryogenesis in Daucus carota L. The JIM4 epitope (J4e) occurred on cells established in culture from hypocotyl explants which appeared to derive, at least in part, from the epidermal cells of the hypocotyl. Cultures maintained in the presence of 2,4-dichlorophenoxyacetic acid developed proembryogenic masses of which only infrequent cells at the surface expressed J4e. Sub-culture at a low cell density and withdrawl of the synthetic auxin resulted in an increase in J4e expression in most surface cells and most abundantly in surface layers of cells at the future shoot end of developing embryos. The transition to heart-shaped embryos occurred concurrently with the expression of J4e by groups of cells beneath the developing cotyledons, at the junction of the future root and shoot. At this stage, J4e was also expressed by a single well-defined layer of cells at the surface of the embryos. Advancement to the mature torpedo stage was accompanied by the expression of the epitope on cells forming two regions of the future stele and of cells associated with the cotyledonary provascular tissue characteristic of the carrot seedling. At this stage there was substantially less expression of the marker antigen by epidermal cells, although infrequent expression by isolated cells of the epidermis was maintained. The correlation of J4e expression with the development and distinction of plant tissue patterns during somatic embryogenesis indicates a role for plasma-membrane arabinogalactan proteins in these processes. * To whom correspondence should be addressed Abbreviations: AGP=arabinogalactan protein; 2,4-D=2,4-dichlorophenoxyacetic acid; J4e = JIM 4 epitope; PEM = proembryogenic mass
Key words: Arabinogalactan protein - Cell surface Daucus (somatic embryogenesis) - Embryogenesis (somatic) - Epitope (JIM4)
Introduction The recent derivation of a monoclonal antibody, JIM4, that recognizes an antigen showing position-specific expression during the early stages of the development of the carrot root apex provides a molecular marker of plant developmental processes (Knox et al. 1989). JIM4 recognizes an epitope (J4e), occurring on a set of plasmamembrane-associated arabinogalactan proteins (AGPs), that in the carrot root apex is expressed by two sections of the developing pericycle, other stele cells related by position, and epidermal cells (Knox et al. 1989). The function of these cell-surface glycoproteins is as yet unknown. To gain further insight into the possible role of these molecules during the acquisition of plant form and early developmental processes we have studied the expression of J4e during carrot somatic embryogenesis. The in-vitro formation of embryos has been studied extensively since the initial observations by Steward et al. (1958) and Reinert (1959). As yet, the molecular processes whereby cells are capable of entering culture as single cells and subsequently enter pathways of organization leading to viable embryos are unknown. No cell-surface molecules marking spatiotemporal patterns that reflect the acquisition of plant form during somatic embryogenesis are known (Sung et al. 1984). In this report we describe the expression of J4e during the formation of continuous cultures of callus cells from hypocotyl explants and also the spatiotemporal patterns of expression, reflecting the establishment of tissue patterns during the subsequent, hormonally controlled, formation of somatic embryos from the proembryogenic masses (PEMs) in these cultures.
286
Material and methods Plant mater&l. Carrot (Daucus carota L. cv. Early Nantes) seeds (W.W. Johnson & Son, Boston, Lincs., UK) were sterilized and germinated on moist filter paper. After 7-10 d, hypocotyls were excised, cut into 1-mm sections and incubated in Gamborg's B5 medium (Flow Laboratories, Rickmansworth, Herts., UK) supplemented with 2 ~tM 2,4-dichlorophenoxyacetic acid (2,4-D) and placed upon a rotary shaker (80 rpm). The methods of De Vries et al. (1988) were followed to maximize embryo formation. Hypocotyl explants were removed 19 d after culture. Cultures of suspended cells, in which PEMs developed, were maintained in medium containing 2,4-D and subcultured by a fivefold dilution every 10-14 d. Embryos were formed when samples, containing PEMs, were withdrawn (at least one week after the most recent subculture), washed extensively with and diluted into medium lacking 2,4-D. A low cell density of cells was produced by addition of 5 Ixl (settled cell volume) of cells into 5 ml of medium in a Petri dish (50 mm diameter). All cultures of carrot seedlings and cells were maintained at 25~ C in the dark. Protoplasts were prepared from the suspended single cells by overnight incubation with 0.5% (w/v) Driselase in 0.5 M mannitol and their reactivity with JIM4 determined by indirect immunofluorescence. Indirect immunofluorescence with the monoclonal antibody JIM4. Plant material from all stages of culture was removed from culture medium and immediately fixed in 4% (w/v) formaldehyde in a buffer containing 50mM 1,4-piperazinediethanesulfonic acid (Pipes), 5 mM MgSO4 and 5 mM ethylene glycol bis (/%aminoethylether)-N,N,N',N'-tetraacetic acid (EGTA), pH 6.9. Tissue was fixed for at least 1 h, washed briefly in water and mounted in OCT compound (Agar Aids, London, UK) and sectioned at - 1 0 ~ with a Bright 5030 Cryomicrotome (Bright Instrument Co., Huntington, Cambs., UK) Sections (7-10 ~tm) were collected on poly-L-lysine-coatedmulti-well slides and stored dry until use. The derivation and characterization of the anti-AGP monoclonal antibody have been described elsewhere (Knox et al. 1989). The sections were incubated for 5 min with 5% (v/v) calf serum in phosphate-buffered saline (CS/PBS) prior to incubation with a fivefold dilution of JIM4 hybridoma culture supernatant in CS/ PBS for 1 h. Wells were washed three times with PBS and then incubated with anti-rat-IgG,A,M-fluorescein isothiocyanate (ICN Biomedicals, Irvine, Cal., USA) at a 150-fold dilution in CS/PBS for a further I h. A final, extensive washing in PBS preceded a 30-s incubation with 4'6-diamidino-2-phenylindole(DAPI) at 1 ~tg. ml-1 in PBS, mounting in Citifluor antifade (Citifluor Ltd., City University, London, UK) and examination with a Zeiss (Oberkochen, FRG) Photomicroscope III. The DAPI is reactive with DNA and its fluorescence was observed at 45~490 nm to locate all cells of the treated sections. Fluorescein isothiocyanate fluorescencewas observed at 520-560 nm and photographed using Kodak (Hemel Hempsted, Herts., UK) TMAX400 film. Controls of no antibody or alternative monoclonal antibody culture supernatant (MAC 207, Pennell et al. 1989) were run and always observed to confirm that the distinct patterns of immunofluorescence shown and described were due to JIM4. Results Formation o f cell cultures f r o m carrot seedling hypocotyls. Expression of J4e was studied during the establishment, from carrot hypocotyl explants, of cultures with embryogenie potential and also during the development of somatic embryos as described by De Vries et al. (1988). The expression of J4e by cells of the hypocotyl at the time of excision of explants occurs in the epidermal cells
N.J. Stacey et al. : Expression of arabinogalactan epitope and two regions of the vascular tissue (Fig. | A; Knox et al. 1989). The anatomy of the hypocotyl is characteristic of the root, with the transition to a vasculature characteristic of the shoot occurring only in the upper hypocotyl (Esau 1940). Six days after incubation of hypocoty[ explants in B 5 medium containing 2 laM 2,4-D, the epidermal layers were seen to be dissociated from the cortical regions and floating freely as sheets of cells (Fig. 1 B). Cells at the periphery of these sheets appeared to contribute to the culture as single large vacuolate cells. The preparation of protoplasts from these cells and reaction with J I M 4 indicated that they had maintained their expression of J4e during their transformation from epidermal into cultured cells (data not shown). At the early stages of culture all the suspended cells were of this large vacuolate morphology. After approximately 10 d of culture, callus, consisting of small highly cytoplasmic cells, was observed at the cut ends of the explants (Fig. 1 D). Cells of this callus, derived from subepidermal cells, were unreactive, or only very weakly reactive, with J I M 4 (Fig. 1 C). However, at the time of removal of the hypocotyl explants (day 19) the suspended cells, maintained in 2,4-D, were predominantly, if not exclusively, o f the large vacuolate type. Protoplasts prepared from cultures at this stage were all highly reactive with JIM 4. Twelve days after removal of hypocotyl explants from the culture of cells (day 31 of culture), small clusters of cells were noted in the medium and these displayed a dramatic change in J4e expression. Indirect immunofluorescence due to J I M 4 was only observable in isolated and infrequent cells at the surface of the clusters of small, tightly adhered cells. Representative examples are shown in Fig. 2A, C. These clumps are the proembryogenic masses (PEMs) of Halperin (1966), cells on the surface of which are capable of developing into embryos. At this stage, and all later stages, the single larger vacuolated cells, also present in the culture, displayed abundant (as determined by intensity of fluorescence) expression of J4e (not shown). The cultures were maintained for several months in this state by the presence of 2,4-D, with an increasing predominance of PEMs over the single vacuolate cells. The size of the PEMs varied between sets of established cultures but in all cases the JIM 4 epitope was expressed by only occasional cells at the surface of a PEM. A reduction to a basal level of infrequent J4e expression in the PEMs, represented by Fig. 2C, was observed to occur over time.
Formation o f somatic embryos. Withdrawl of the synthetic auxin and culture at low cell density resulted in the development of torpedo-stage embryos within 10-14 d. These conditions resulted in a dramatic increase of expression of J4e by cells of the PEMs within 24 h. All or most of the cells at the surface of the PEMs were strongly fluorescent due to J I M 4 binding 24 h after the commencement of the conditions allowing embryogenesis (Fig. 2B, D). Induced expression in larger, ramiform PEMs was commonly restricted to cells on the distal
N.J. Stacey et al. : Expression of arabinogalactan epitope
287
Fig. 1 A-D. Carrot hypocotyl explants cultured in Gamborg's B 5 medium supplemented with 2 I.tM 2,4-D. A Transverse section of the hypocotyl of a carrot seedling at the beginning of culture showing immunofluorescence due to JIM4 on epidermal cells (e) and two sectors of the vascular cylinder (v, the arrowheads indicate the centre of the sectors and the xylem axis). JIM4 did not react with cells of the cortex (c). B White-light image of a sheet of JIM 4reactive epidermal cells dissociated from hypocotyl explants 6 d after the beginning of culture. Arrowheads indicate the region con-
tributing large vacuolate cells to the culture of suspended cells. C Indirect immunofluorescence image showing the weak reactivity of JIM4 with a section of callus (arrowheads) derived from the culture of hypocotyl explants. Small double arrowheads indicate highly JIM 4-reactive cells of hypocotyl explant debris. D A DAPIstained image of nuclei of the section of callus shown in C to indicate its composition of small tightly adhered cells. Bars= 100 ~tm
surface o f the branching structures as shown in Fig. 2 D. In a study o f the timing of this increase in AGP-epitope expression the first increase occurred between 12 and 24 h after removal of the 2,4-D (data not shown). It is well established that embryos can develop from cells at the surface of the P E M s (Halperin 1966; De Vries et al. 1988). The earliest stages in the development of globular embryos were difficult to locate, but most surface cells and small groups of cells on the surface of P E M s displayed abundant J4e expression at this time. It has often been observed that several embryos can develop concurrently, linked by the future root poles through a central cluster of cells of the former PEM. This indicates that the establishment of the root-shoot polarity is a very early event and occurs in spatial rela-
tion to the PEM. At the globular stage the established root-shoot polarity is reflected by the J4e expression in several surface layers centred upon the developing shoot end of the embryo (Fig. 3A). The subsequent development of embryos was associated with distinct spatial patterns of expression of J4e. As embryos progressed to the heart-shaped stage the expression was restricted more clearly to one, well-defined, surface layer of cells and the first internal region of J4e expression was seen, in longitudinal section, as two regions o f expression reflecting the position of the ridges of the emerging cotyledons (Fig. 3 C, D). The internal expression of J4e appeared to arise as two distinct groups of cells, representing the cotyledonary node and the future root-shoot junction. These cells did not ap-
288
Fig. 2A-F. JIM4-reactive cells in proembryogenic masses derived from the culture of carrot hypocotyl explants. A Immunofluorescent image of J I M 4 reaction with cells in a section of a PEM produced in the presence of 2,4-D from a culture of suspended cells derived from hypocotyl explants. This PEM was isolated 10 d after the removal of the explant debris. Reactive cells were on the surface of the PEM (arrowhead) and the central cells were unreactive (u). B After 24 h of culture at low cell density in the absence of 2,4-D, a PEM equivalent to that shown in A displayed more abundant fluorescence due to J I M 4 in surface cells but cells in the centre remained unreactive (u). C J I M 4 reaction with PEMs, maintained in medium with 2,4-D, approx, six weeks after the
N.J. Stacey et al. : Expression of arabinogalactan epitope
removal of the explant debris. Infrequent isolated surface cells expressing J4e are indicated by arrowheads. D Section through a PEM equivalent to that of B, indicating that cells expressing J4e 24 h after removal of 2,4-D and culture at low cell density were restricted to distal surfaces of a ramiform PEM. Cells at the centre of composite clumps (u) and of the branching PEM (m) remained unreactive. E A negative control section, equivalent to D, from which J I M 4 treatment was omitted. F Another control section treated with a different antibody, M A C 207, in the place of JIM 4, showed reaction with all cells. B a r s = 100 lam
N.J. Stacey et al. : Expression of arabinogalactan epitope
289
Fig. 3A-I. Patterns of J4e expression by carrot embryos induced in cultures of PEMs subcultured at low cell density in the absence of 2,4-D. A Longitudinal section of a late globular-stage embryo, r = root pole, s= shoot pole. Reactive cells in the surface layers centre on the shoot end. B A longitudinal section of an embryo at the early heart-shaped stage indicated two groups of internal cells reactive with JIM4, reflecting the position of emerging cotyledonary ridges (arrowhead). C Longitudinal section through a late heartshaped embryo. The J4e was expressed by a single defined layer of epidermal cells at the shoot apical end of the embryo. D, E At the early torpedo stage the cell groups beneath the emerging cotyledons were welt established and provascular tissues (of the future stele) also expressed J4e. F A longitudinal section through a mature torpedo-stage somatic embryo in which J4e is expressed in a manner corresponding to the carrot seedling. Position and angle of arrows indicate sections from an equivalent embryo shown in G-I. G Transverse section showing two sectors of provascular tissue that expressed J4e (large arrowhead). Isolated epidermal cells also expressed J4e (small arrowhead). H Transverse section showing a continuous circle of expression at the cotyledonary node. I Transverse section showing commencement of the ramification of provascular tissue into the cotyledons, a= cells of the future shoot apex which express J4e. Bars= 100 ~tm pear to be anatomically distinct at this stage. In the mature embryo the cells of this region expressing J4e form a complete circle (see below and Fig. 3 H). As the embryo developed further, J4e expression was observed in the provascular tissue of the future stele. This was in two regions reflecting the two sectors of the future stele (Fig. 3 D, E, G), related to the position of the diarch vascular structure in the mature root, and then as the cotyledons developed J4e expression occurred on cells associated with cotyledonary provascular traces (Fig. 3 F, I). At this stage the circle of expression at the root-shoot junction can be clearly discerned (Fig. 3 H), and also its correlation with ramifying provascular traces in the cotyledons (Fig. 3 F, I). A region of expression in surface layers at the site of the future shoot meristem was often observed (Fig. 3D, E, I). In
the mature embryo the epidermal layer o f cells did not uniformly express J4e, but infrequent groups of isolated cells continued to do so.
Developmental arrest and J4e expression. In certain cases, cultures maintained in the presence of 2,4-D at a lower density of cells can produce globular embryo-like structures, generally larger than the globular embryos developed from PEMs after 2,4-D withdrawl (Bokird et al. 1986). These are arrested embryos and subsequent withdrawl of 2,4-D results in organization directly into shoot and root meristems in a highly irregular way, resulting in monstrous embryos. The frequency of such anomalies has previously been noted (Halperin 1966; Ammirato 1987). Such arrested globular embryos maintained in 2,4-D have very few cells displaying J4e expression, often
290
Fig. 4. A Arrested globular stages of carrot embryos observed in cultures maintained for 10 d in medium containing 2,4-D. JIM4 was reactive with only a few cells. However, a potential root (r) and shoot (s) polarity does occur in relation to the centre of the PEM (m). B A section through equivalent tissue 24 h after culture at low cell density in medium containing no 2,4-D. J4e expression had increased and was most abundant towards the shoot pole of future embryos. Bars= 100 lam
none in a section or only several isolated weakly labelled cells (Fig. 4A). Removal of 2,4-D and subculture at low cell density resulted in expression of J4e by cells in the surface layers with decreasing abundance of expression by more proximal cells (Fig. 4B). As these structures developed further into larger and often irregular embryos, patterns of J4e expression correlating with the morphology-expression relationships seen iri normal embryos were observed (not shown).
Discussion We have no information on the function of the antigens bearing J4e and only limited knowledge of their biochemical nature. They are plasma-membrane-associated AGPs, and J4e appears to define a subset, or represent a modification of, a larger set of membrane-associated AGPs (Knox et al. 1989; Pennell et al. 1989). Arabinogalactan proteins have received considerable attention, but the function of this class of molecules has not yet emerged (Fincher et al. 1983). However, it is an important observation that AGPs associated with the plasma membrane of plant cells can contain information, in terms of epitope restriction to distinct tissues, or groups of cells, during early stages of plant development. We have used somatic embryogenesis, which is accessible to manipulation and hormonal control, to further study the expression of J4e. The patterns of expression of J4e during the develop-
N.J. Stacey et al. : Expression of arabinogalactan epitope
mental processes described in this report can be put into the two categories described below.
Establishment of the pattern of expression characteristic of the carrot seedling. In this first category of J4e expression, the establishment of the pattern of expression characteristic of the mature embryo and seedling (Knox et al. 1989), the epitope is expressed by two distinct regions of the future pericycle and associated cells, cells associated with future vascular traces in the cotyledon, isolated groups of epidermal cells and a group of cells at the cotyledonary node. Anatomical studies have been carried out upon developing somatic embryos of carrot (Halperin and Wetherall 1964; Halperin 1966; McWilliam et al. 1974; Schiavone and Cooke 1985; Ammirato 1987). The protoderm is the first tissue to be clearly defined, at the late globular stage, and the formation of the cotyledonary node is immediately subsequent to that of the protoderm. The provascular tissue of the future hypocotyl first appears during the transition to the heart-shaped stage and then differentiates acropetally. The somatic embryo reaches the maturity of its torpedo stage with the appearance of provascular tissue in the cotyledons. The expression of J4e seems to be correlated with this sequence by its appearance in the outer surface layers of cells at the globular stage, its restriction to a defined single layer at the heart-shaped stage, and its expression by cells associated with the future root-shoot junction (the cotyledonary node). Finally, J4e expression was associated with provascular tissue in the future hypocotyl and then in the cotyledons. The transition from radial to bilateral symmetry of the embryo (at the heart-shaped stage) is reflected in J4e expression by two groups of internal cells beneath the emerging cotyledons and locates the developing cotyledonary node. At the torpedo stage the bilateral symmetry is reflected in the bilateral expression of the provascular tissues. These correlations of J4e expression with the emerging anatomical patterns have been made by observation of the position of cell collec-
N.J. Stacey et al. : Expression of arabinogalactan epitope tives expressing J4e rather than by the precise observation of distinctive structural features. The timing o f the expression o f J4e by these groups of cells, in relation to known sequences of embryo development, indicate, as in the root meristem (Knox et al. 1989), that expression of this A G P epitope occurred at the time of the formation o f these distinct tissues. The decrease in J4e expression by surface layers o f the developing embryo during its transition from the globular to the torpedo stage is striking and may reflect an important role for surface cell layers during the early events of organization (see below). The reduction in expression may be caused by the shedding of cells or by switching off expression. The shedding of surface cells has been noted for all stages of embryos (Halperin 1966) and in certain cases highly JIM4-reactive cells have been noted to be loosely associated with the surface of embryos. Establishment o f cultures and proembryogenic masses.
The second category of JIM4-epitope expression concerns earlier events of embryogenesis, before the development of tissue boundaries, and during the establishment o f plant cells in culture. These patterns of expression indicate a role for plasma-membrane AGPs at the earliest stages of organization and of the distinction of cells within cell clusters. When the large vacuolate cells derived from JIM4reactive epidermal cells, and thought to be non-embryogenic, were cultured their reactivity with J I M 4 was maintained. Such cells are distinct in morphology from the PEMs and also in terms of J4e expression. Only infrequent isolated cells at the surface of the PEMs expressed J4e, and often only weakly in terms of intensity of fluorescence. The origin of PEMs, formed in increasing abundance as a result of continued culture in medium containing 2,4-D, is unclear. It is not known whether the PEMs are produced from of the large vacuolate cells or from the small highly cytoplasmic cells which are derived from callus proliferated at the cut ends of the hypocotyl explants and exist at a low frequency at the time of removal of the hypocotyl explants (Nomura and Komamine 1985; De Vries etal. 1988). The callus formed at the cut edge of explants, and presumably adding cells to the medium, is more characteristic of PEMs in terms of cell morphology and J4e expression. Embryos are thought to arise from single cells or groups o f cells at the surface of PEMs, but the relationship between them and J4e expression has not been determined. The mechanisms and processes whereby cells are coordinated to initiate somatic embryos are uncertain (Williams and Maheswaran 1986). The dramatic increase in J4e expression upon withdrawl of 2,4-D from the medium and culture at low cell density further indicates a developmental role for AGPs. The expression of J4e by surface cells of a cell mass identifies these cells in terms of position. The presence of 2,4-D and a high cell density appears to inhibit the full expression of this positional identity (as well as the development of embryos). The nature of this
291 switch, whether by existing cells or new daughter cells, has not been clarified. This distinction of surface cells may be required for the future development of organized structures at the surface of a PEM and also during the early globular and heart-shaped stages of embryo formation when cells o f the surface layers abundantly express J4e. In this latter case, surface cells o f developing embryos are clearly defined by anticlinical divisions and are distinct from the callus-like cells expressing J4e at the surface of a PEM. However, in both cases J4e is a marker of surface position and may reflect common developmental processes. The case of developmentally arrested (by 2,4-D) globular-stage embryos and the increase in expression of J4e prior to their continued development indicate that existing cells can be induced to express J4e. These observations also point to a functional, possibly regulatory, role for cell-surface AGPs. The observed gradient of expression, decreasing internally and distally upon release from arrest, further indicates that the antigens recognized by J I M 4 are positionally expressed in relation to organizing clusters of plant cells. In summary, the expression of J4e, as characterized so far, provides a marker for all stages of the acquisition of the major tissue patterns during carrot somatic embryogenesis. The ubiquity of differential expression of J4e by cells at all stages from PEMs to the mature embryo indicates that plasma-membrane AGPs are functionally concerned with the position of cells in relation to emerging plant form, since they are expressed in terms of surface position in cell clusters and in later developmental events are associated with the distinction of tissue blocks. This class of antigens does not appear to be tissue-specific, concerned with specific differentiation pathways, but seems to be a class of cell-surface glycoproteins concerned with the establishment of positional distinctions between cells and of tissue patterns within which cytodifferentiation processes can be contained. We thank Andrew Davis for photographic assistance and Roger Pennell for useful discussions. References Ammirato, P.V. (1987) Organizational events during somatic embryogenesis. In: Plant tissue and cell culture, pp. 57-81, Green, C.E., Somers, D.A., Hackett, W.P., Biesboer, D.D., eds. Alan R. Liss, New York Bokird, C., Choi, J.H., Sung, Z.R. (1986) Effect of 2,4-dichlorophenoxyacetic acid on the expression of embryogenic program in carrot. Plant Physiol. 81, 1143-1146 De Vries, S.C., Booij, H., Meyerink, P., Huisman, G., Wilde, H.D., Thomas, T.L., van Kammen, A. (1988) Acquisition of embryogenic potential in carrot cell-suspension cultures. Planta 176, 196-204 Esau, K. (1940) Developmental anatomy of the fleshy storage organ of Daucus carota. Hilgardia 13, 175-226 Fincher, G.B., Stone, B.A., Clarke, A.E. (1983) Arabinogalactan proteins: structure, biosynthesis and function. Annu. Rev. Plant Physiol. 34, 47-70 Halperin, W. (1966) Alternative morphogenetic events in cell suspensions. Am. J. Bot. 53, 443453
292 Halperin, W., Wetherall, D.F. (1964) Adventive embryony in tissue cultures of the wild carrot, Daucus carota. Am. J. Bot. 51, 274-283 Knox, J.P., Day, S., Roberts, K. (1989) A set of cell surface glycoproteins forms an early marker of cell position, and not cell type, in the root apical meristem of Daucus carota L. Development 106, 47-56 McWilliam, A.A., Smith, S.M., Street, H.E. (1974) The origin and development of embryoids in suspension cultures of carrot (Daucus carota L.). Ann. Bot. 38, 243-250 Nomura, K., Komamine, A. (1985) Identification and isolation of single cells that produce somatic embryos at a high frequency in a carrot suspension culture. Plant Physiol. 79, 988-991 Pennell, R.I., Knox, J.P., Scofield, G.N., Selvendran, R.R., Roberts, K. (1989) A family of abundant plasma membrane-associated glycoproteins related to the arabinogalactan proteins is unique to flowering plants. J. Cell Biol. 108, 1967-1977
N.J. Stacey et al. : Expression of arabinogalactan epitope Reinert, J. (1958) Morphogenese und ihre Kontrolle an Gewebekulturen aus Carotten. Naturwissenschaften 14, 344-345 Schiavone, F.M., Cooke, T.J. (1988) A geometric analysis of somatic embryo formation in carrot cell cultures. Can. J. Bot. 63, 1573-1578 Steward, F.C., Mapes, M.O., Mears, K. (1958) Growth and organized development of cultured ceils. II. Organization in cultures grown from freely suspended cells. Am. J. Bot. 45, 705 708 Sung, Z.R., Feinberg, A., Chorneau, R., Borkird, C., Furner, I., Smith, J., Terzi, M., LoSchiavo, F., Giuliano, G., Pitto, L., Nuti-Ronchi, V. (1984) Developmental biology of embryogenesis from carrot culture. Plant Mol. Biol. Rep. 2, 3-14 Williams, E.G., Maheswaran, G. (1986)Somatic embryogenesis: factors influencing coordinated behaviour of cells as an embryogenic group. Ann. Bot. 57, 443-462 Received 14 July; accepted 21 September 1989
