A biochemical model for the initiation and maintenance of the quiescent center: implications for organization of root meristems

Apical meristems give rise to all tissue and organ systems of the postembryonic plant. Not only do they generate the cells from which the plant is constructed, but apical meristems also function as the organizing centers for postembryonic morphogenesis. The evidence for this conclusion has emerged gradually from many different studies addressing the initiation, organization, maintenance and function of apical meristems, especially root apical meristems (Steeves and Sussex, 1989).

The organization of root meristems has been studied through analyses of cell lineages, based on histological sections. Because cell lineages, or files, converge at the root pole, Hanstein (1868) postulated that a few cells, located at the pole could serve as the initials for the root, and through high mitotic activity, generate all the cells that make up the root. In an effort to define these cells precisely, Clowes (1953, 1954) performed surgical experiments on Zea mays root apices and showed that if these cells were damaged or removed, the remaining surrounding cells could directly reconstitute a complete apex. This led Clowes to conclude that the functional initials were actually located peripherally to the very central cells, in a region of the meristem later named the proximal meristem (Feldman and Torrey, 1975). Clowes’s (1956) use of thymidine labeling showed for the first time that the most central cells of the meristem actually divide infrequently, or not at all, and he named this population of cells the quiescent center (QC).

Since its discovery, much work has contributed to the characterization of the QC. While it is believed to be a feature of all angiosperm root apices, the most extensive analysis of the QC has been done on roots of maize, in which the QC can attain a size of 1000-1500 cells (Feldman and Torrey, 1976). Dolan et al. (1993) have shown that in Arabidopsis the QC comprises only four central cells derived from the hypophysis and is surrounded by cells that act as the initials for the files of cells that make up the root. Average cell cycle times within the QC of Zea are in the range of 170 hours. In contrast, the cell cycle time of the root cap initials, the most rapidly dividing cells in the root, is only 10-16 hours, and for the proximal meristem, 18-25 hours (Clowes, 1961).

Cells within the QC can be distinguished from surrounding meristem cells by their fainter histochemical staining (Goyal and Pillai, 1986), lower RNA content (Clowes, 1972), lower protein content (Jensen, 1958), and low RNA polymerase activity (Fisher, 1968). In situ hybridizations with a ³H-labeled histone gene probe and with [³H]polyuridylic acid (poly U) to root tissue sections have revealed populations of unlabeled cells which correspond precisely to the area defined as the developing QC in rice and Capsella for each probe respectively (Raghavan and Olmedilla, 1989; Raghavan, 1990).

Prior to discernible apical organization, cells at the presumptive root pole (the future root meristem) show uniformly high rates of mitosis, and stain densely with a variety of histochemical stains. Establishment and elaboration of a QC occurs gradually and precedes stable histological patterning in the meristem. From experiments using uptake of radiolabelled precursors, Clowes (1978) concluded that establishment of the QC preceded all histogenic events except establishment of a root cap meristem layer. More recently, using in situ hybridization of radiolabelled poly U to Capsella, Raghavan (1990) was able to show that very early during embryogenesis, before any root apical organization was evident, one cell and subsequently several were clearly distinguished by their relative lack of bound probe. He then demonstrated that these unlabelled cells were the origin of the QC. His results, in an elegant way, supported previous hypotheses that the organization of apical meristems in roots follows the establishment and elaboration of the QC. Other work concerning the establishment of histological patterning in developing lateral root primordia and in adventitious root primordia has also provided evidence for the establishment of a QC prior to histological organization of a root meristem (Rondet, 1961; Feldman, 1977).

Using microsurgical techniques, earlier workers had shown that it was possible to excise defined regions of the root apical meristem in maize, including the QC itself (Feldman, 1977). Results of these efforts showed that roots are able to regenerate a new apical meristem following surgical excisions, but that prior to the initiation of distinctive histological zonation, a QC reformed, appearing initially as a group of mitotically relatively inactive cells surrounded by the rapidly dividing cells remaining at the cut root stump (Feldman, 1977; Rost and Jones, 1988). From this work it was also concluded that the reformation of a QC precedes and is requisite for organization of root meristems.

Despite these many studies we do not have definitive information about the factors that initiate and maintain quiescence in these cells nor do we have a convincing functional role for these cells. Past workers have suggested that the QC and the ‘ultimate initials’ that it contains are equivalent to ‘stem’ cells in animals (Barlow, 1978). Torrey (1972), and later Feldman (1975, 1979), proposed that the QC may be a site of hormone biosynthesis in the root and as a consequence of these localized, enhanced metabolic processes these cells were inhibited with regard to many other physiological activities. Other possible explanations for the quiescent state have focused either on the supposed nutritional status of the QC or on the possibility that physical constraints prevent cells of the QC from dividing (Clowes, 1972).

In this paper we present a new hypothesis for the formation of the quiescent center and provide data that address the cause and maintenance of the QC. We have used the perspective of examining the QC with regard to its position in a whole plant context. The QC is located at the most distal part of the root proper, the most terminal location on the path of polar transport from the shoot. Of the many substances polarly transported in the plant, auxin is one of the best studied and has been shown to affect root meristem organization. Here we provide a detailed mechanism linking polar auxin transport with the establishment and maintenance of the QC (Fig. 1). In this mechanism polar auxin is directly linked to quiescence through the action of ascorbate oxidase and ascorbic acid. Briefly, we report that auxin and ascorbate oxidase levels are high in the QC relative to surrounding cells and that the QC cells have low or undetectable levels of ascorbic acid, a compound known to be necessary for the transition from G₁ to S in the cell cycle. Having discussed the mechanism imposing quiescence we discuss the implications that this mechanism has for the establishment of pattern at the root apex.

Corn caryopses (var. Merit, Asgrow Seed Co., Kalamazoo, MI) were imbibed and germinated in the dark at 25 °C for 2 days. Tissue was collected by surgical removal of the cap and excision of the QC, (Feldman and Torrey, 1976). In this cultivar, the QC is separated from the proximal meristem by a weak, thin-walled junction making possible routine, clean dissections of isolated QCs. QCs were collected in a moist environment, quick frozen on dry ice and stored at −80 °C for extractions as described below. Approximately 2 mm of the remaining root stump, the proximal meristem region, was also collected and stored in this manner. Where mature root tissue was used, 1 cm segments located approximately 1 cm behind the tip were collected. For root tissue cultured with or without exogenous auxin, mature root segments (1 cm) were cultured on Murashige and Skoog (MS) medium, 3% sucrose, 0.8% agar in the presence of 0 or 1.0 mg/l of 2,4-D. Material was incubated in the dark at 25 °C and was harvested and collected as above after 2 and 4 days (Esaka et al., 1992).

Caryopses were germinated as above and grown for 3 days. Roots were severed at the base, below the scutellar node and the cut end was placed in a 0.5 ml microfuge tube containing half-strength MS medium, 10⁻⁸ M nonradioactive indoleacetic acid (IAA) and [¹⁴C]IAA (1 μCi/ml, specific activity = 4.8 Ci/mM) in 0.8% low melting temperature agarose (Feldman, 1981). After 24 hours of transport at room temperature in the dark, roots were removed from tubes and squashed between plastic wrap and Whatman blotting paper. A glass plate was placed over the plastic wrap and pressure was applied squashing the root firmly and uniformly onto the paper. Squashed roots were exposed to X-ray film for approximately 2 weeks.

RNA was isolated from frozen tissues collected as described above. RNA was isolated from material using a modification of the method of Puissant and Houdebine (1990). Approximately 150 individual frozen QCs were ground in a small grinding tube in 25 μl guanidinium buffer on ice. 2.5 μl 2 M sodium acetate, pH 4.0, and 25 μl water saturated phenol/chloroform was added. The solution was transferred to a microfuge tube, vortexed and centrifuged at 12,000 g for 10 minutes at 4 °C. The upper phase was recovered and precipitated with isopropanol. The pellet was resuspended in 25 μl 4 M LiCl, then centrifuged at 3000 g for 10 minutes at 4 °C. The resulting pellet was redissolved in Dep-treated H₂0, 0.1% SDS, extracted with an equal volume of chloroform, and the aqueous phase was adjusted to 0.2 M sodium acetate, pH 5.0, and precipitated again with isopropanol. The resulting pellet was resuspended in Dep-treated H₂O, and RNA was quantitated and used for northern blot analysis. RNA was extracted from proximal meristems and other root sources by the same method except the volumes were scaled up to process the tissues which were more easily collected in larger amounts.

RNA electrophoresis, blotting and hybridization were performed essentially as described previously (Maniatis et al., 1989). 5 or 10 μg of total RNA was electrophoresed in a 1% agarose gel containing formaldehyde and blotted to Nytran (Schleicher & Schuell). Equal loading of RNA was confirmed by ethidium bromide staining of the gel before transfer to the membrane. RNAs were probed with the near full length cDNA clone for cucumber AAO, pASO11 (Ohkawa et al., 1989). The probe was radiolabeled using random hexamer priming with the Prime-a-Gene method (Promega).

In situ hybridizations were done according to the method of Jackson (1991). Maize root tips were hybridized with ³⁵S-labeled sense and antisense riboprobes (synthesized with Ribo Probe kit, Stratagene) coding for elongation factor-α cloned from radish (Kerk, 1990). Slides were hybridized overnight, washed in 2× SSC, 50% formamide at 50 °C, dried and exposed to Kodak NTB-2 emulsion.

Ascorbic acid oxidase (AAO) activity was assayed by following the decrease in the spectrophotometric absorbance of ascorbic acid over time in the presence of proteins extracted from QCs, proximal meristems, and other root tissues using the method of Oberbacher and Vines (1963). Tissue homogenates were prepared from freshly collected quiescent centers, proximal meristems and cultured root segments. Tissues were ground in 5 parts (w/v) 0.1 M potassium phosphate buffer, pH 7.0, on ice and centrifuged at 10,000 g for 15 minutes at 4°C. 10 μl samples of these homogenates were added to the reaction mixture containing 0.05 M potassium phosphate buffer (pH 7.0), 0.5 mM EDTA, and 0.15 mM L-ascorbic acid in a volume of 1.0 ml (Esaka et al., 1988). One unit of activity was defined as the amount of enzyme which oxidizes 1.0 μmol of L-ascorbic acid per minute and was converted from the change in A₂₆₅ at 25 °C over time (Oberbacher and Vines 1963). Protein concentration was determined using the Bio-Rad protein assay and units of activity were normalized to this amount.

This procedure uses the unique ability of ascorbic acid to reduce silver nitrate to silver in acidic conditions (Chayen, 1953). Root tips were quick-frozen in isopentane, cooled in a dry ice bath and dehydrated in several changes of absolute ethanol at −30 °C. The ethanol was replaced with toluene and the tissue infiltrated with paraffin and sectioned at 10 μm (Jensen, 1962). Ascorbic acid was localized according to the methods of Jensen and Kavaljian (1956). Black deposits of metallic silver indicate regions where ascorbic acid is present. The control for this procedure is to expose the sectioned material to a copper sulfate solution that oxidizes all the ascorbic acid to dehydroascorbic acid, which does not react with the silver nitrate.

Tissue for immunolocalization was fixed in freshly prepared 2% aqueous ethyl-3-(-3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) (Sigma) on ice for 30 minutes under vacuum (Shi et al., 1993), followed by postfixation in 2.5% paraformaldehyde/0.25% glutaraldehyde in 0.1 M phosphate buffer, pH 7.0, at 4 °C overnight. Binding of antibody (Ab) was carried out essentially as described by Chichiricco et al. (1989) using alkaline phosphatase for detection. The monoclonal antibody (mAb) to auxin has been shown to be specific to free auxin in Zea root tips (Shi et al., 1993). Several controls were carried out to confirm specificity. The controls included: (1) omitting prefixation with EDC, (2) omitting incubation with the primary mAb and (3) incubation with the mAb previously exposed to an excess of auxin in incubation solution. The AAO antibody was a polyclonal Ab (Esaka et al., 1988) and was used for localizations in tissue sections as above, with the omission of the prefixation in EDC.

Corn caryopses were imbibed and germinated for 3 days as described above. Caryopses were then placed upon parafilm covered deep Petri dishes such that the root extended through holes punched in the film into solution contained in the dish. Solutions were either water or 0.1 mM ascorbic acid both maintained at pH 5.9. Roots were incubated at 25 °C in the dark for 24 hours with gentle agitation. Bromodeoxyuridine (BrdU) was added to 10 μM and roots were incubated for a further 24 hours. Root tips were then excised, fixed and sectioned as described above.

A modification of the procedure recommended by Boehringer Mannheim for use with their Anti-BrdU antibody was used. Sections were hydrolyzed in 1 N HCl for 1.5 hours at 37 °C, then neutralized by immersion in 0.1 M borate buffer, pH 8.5, washed with PBS, and incubated with anti-BrdU antibody (Developmental Studies Hybridoma Bank, NICHD) for 2 hours. Slides were washed in PBS and incubated with rabbit anti-mouse IgG-FITC (Southern Biotechnology Associates, Inc.) overnight in the dark at room temperature in a humidified chamber. Slides were washed and covered with a drop of 50% glycerol, 0.15% N-propyl gallate in PBS and a coverslip. The stained material was observed with a Zeiss Axiophot microscope equipped with a Zeiss ZVS-47DEC video camera. Video frames from the ZVS-47DEC were digitized and displayed on a Macintosh Power Mac 8100/80AV and arranged using Adobe Photoshop v. 3.0 software.

We established that auxin accumulates in the region of the QC by using the following two methods: (1) by demonstrating regions of accumulation of polarly transported [¹⁴C]IAA using autoradiography and (2) immunolocalization of IAA to root tip tissue sections. The first approach allowed visualization of the path of polar IAA transport in 3-day old maize seedling roots using [¹⁴C]IAA. Roots were severed from the hypocotyl and exposed to [¹⁴C]IAA at the cut surface. Transport was allowed to proceed for 24 hours and the roots were processed and exposed to X-ray film. Fig. 2 shows a representative root. Radioactivity was localized in the vascular tissue and in the root tip. A characteristic feature was the low level of signal in the region behind the tip, the region of cell elongation. Signal can also be seen in vascular traces leading to developing lateral root primordia.

Higher resolution of auxin localization in the root tip was obtained using a monoclonal antibody to auxin. This antibody has previously been shown to have high specificity for free auxin (Shi et al., 1993). Fig. 3A shows the alkaline phosphatase staining pattern indicating antibody binding to auxin. There was distinctive dark staining in the region corresponding to the quiescent center. The root cap also showed high levels of auxin, as did the outer cortical cells of the root proper. The vascular tissue showed significant staining as well. The root cap meristem region between the QC and the cap had much less auxin as did the inner cortex and epidermis. Two controls are also shown. First was the pattern seen when tissue sections were treated as above but without the primary antibody in the dilution buffer (Fig. 3B). There was little detectable staining in the QC. The other control shows the result of incubating the antibody with an excess of auxin in solution prior to exposure of the antibody to the tissue section (Fig. 3C). With this control, sites antigenic for auxin should become saturated prior to exposure to auxin in the tissue sections, and hence should not be able to bind auxin during immunolocalization. This pattern showed some generalized background staining but the marked differential distribution seen with the antibody alone was not detectable. The staining in the quiescent center, root cap, outer cortical cells and central cylinder was very reduced when the antibody was pretreated with auxin.

Auxin has an effect on AAO levels in the root. AAO activity increases in tissue cultured with auxin. Roots cultured with auxin for 4 days showed a ten-fold increase in activity compared to control roots (Fig. 4A). Northern blots of RNA prepared from portions of the same root tissue as that used for the activity assays show message levels of AAO peak at day 2 and decrease by day 4 (Fig. 4B). Esaka et al. (1992) also observed this same mRNA profile in pumpkin fruit tissue cultured with auxin. These data provide evidence that culture with auxin results in an increased level of AAO mRNA and enzyme activity in root tissue, and that a continued increase in the mRNA level does not underlie the increasing levels of enzyme activity, since AAO activity levels continue to increase even though mRNA levels decrease before day 4.

The following experiments examined AAO mRNA and protein distribution, and AAO activity in the different regions of the root. The cultivar of maize used for this work has a slightly weakened cell wall zone between the QC region and the proximal meristem, making it possible to remove the root cap and collect individual QCs. Northern blots of total RNA probed with a full-length cDNA for AAO show high levels of mRNA in the QC and lower levels in the proximal meristem and mature root region (an ethidium bromide stained gel is shown as a control for equal loading of the blotted gel, Fig. 5A,B). The same filter was reprobed with a cDNA for a cell cycle gene, p34cdc2, (Colasanti, J. et al., 1988) to demonstrate the different levels of these two messages in the total mRNA extracted from these tissues (Fig. 5C). No detectable mRNA was present in the QC, while high levels were detected in the proximal meristem where a high rate of cell division occurs. There is lower signal in the mature root as would be expected in tissues with fewer dividing cells.

The distribution of AAO protein in these regions of the tip is similar to the pattern of mRNA distribution. AAO is localized very distinctly in the QC but occurs at much lower levels in the proximal meristem (Fig. 6A). The root cap meristem region has relatively low levels of AAO protein. When this pattern is compared to the in situ hybridization pattern of elongation factor-α, a translation factor that has been shown to be a marker for cells undergoing active protein synthesis, the two patterns can be seen as the inverse of each other (Fig. 6B). While elongation factor-α is at very low levels in the QC, root cap, central vascular cylinder and outer cortical cells, AAO protein is relatively high in these tissues. The AAO pattern is also very similar to the pattern of auxin distribution previously shown with the auxin antibody (compare with Fig. 3A). Spectrophotometric assays of AAO activity in protein extracts made from QCs and proximal meristems show that this increased level of AAO protein in the QC corresponds to much higher activity levels of AAO in the QC (Fig. 7). Thus we have shown that the QC has high levels of AAO mRNA, AAO protein and AAO activity, compared to the immediately surrounding cells of the proximal and root cap meristems.

The primary substrate for AAO is ascorbic acid, which is utilized in several metabolic processes, and has been reported to be necessary for the transition from G₁ to S in the cell cycle of several plants (Chinoy, 1984). Others have also suggested that ascorbic acid may have a regulatory role in cell proliferation (Innocenti et al., 1989; Arrigoni et al., 1989; Liso et al., 1984; Citterio et al., 1994). Moreover, it is known that cells in the quiescent center have extended G₁ phases and divide rarely (Clowes, 1975). Localizations of ascorbic acid show it to be absent or present at undetectable levels in quiescent center cells (Fig. 8). At high magnification, the proximal meristem region shows orthogonal cell files with a clearly visible mitotic figure. The silver deposits that indicate the presence of ascorbic acid in these cells appear as black dots. Cells in the QC region are much less regularly shaped, show no mitotic figures and contain no silver deposits.

Roots incubated for 48 hours in 0.1 mM ascorbic acid showed increased levels of cell division activity compared to roots incubated in water. Most striking was the activation of all cells in the quiescent center as viewed in median sections through the root apex (Fig. 9). The immunofluorescent signal over the nuclei indicates cells that had incorporated BrdU during nuclear DNA replication and thus marked those that had passed from G₁ through S of the cell cycle during the period of the experimental treatment. The control roots in water showed a prominent quiescent center and low levels of signal in the area of the stele. The root cap meristem was evident and appeared to be composed of two clearly labeled cell tiers. In contrast, the ascorbic acid-treated roots had an overall higher level of BrdU incorporation, which is reflected not only by the brighter fluorescence of the majority of nuclei, but also by enhanced cytoplasmic fluorescence, perhaps indicative of increased rates of organelle DNA replication (Fujie et al., 1993). No QC could be distinguished, but just distal to the root cap juncture, the root cap meristem showed at least 4 prominent tiers of dividing cells. Thus, applying ascorbic acid directly to roots activates cell division in the majority if not all quiescent center cells.

In summary, the QC has relatively higher levels of auxin, AAO mRNA, AAO protein, and AAO activity, and lower or undetectable levels of ascorbic acid compared to the more rapidly dividing cells surrounding it. In addition, the general pattern of auxin distribution in the root tip is coincident with the pattern of AAO protein localization; both of which appear to be the inverse of the pattern of elongation factor-α, which has been used as an indicator of high levels protein synthesis and as a negative delineator of the quiescent center (Kerk, 1990).

We have proposed a new model for the establishment and maintenance of the quiescent center that is derived from consideration of its position and possible function in a whole plant context. As a terminal region with regard to the transport of many substances, the QC is located in a potentially dynamic region of the plant root. Recent studies of phloem transport and unloading in Arabidopsis roots have shown that fluorescent dye tracers can accumulate in the quiescent center after phloem unloading in the elongation zone and symplastic transport to the tip (Oparka et al., 1994). Our studies of transport of [¹⁴C]IAA in maize seedling roots showed accumulation of radioactivity at the root tip and also in the region of cell elongation. This latter area corresponds anatomically to the region of phloem unloading in Arabidopsis roots.

Antibody localization of auxin revealed that the root cap and quiescent center contained relatively higher levels of IAA than the immediately surrounding proximal meristem and root cap meristem cells. The vascular tissue, pericycle and outer cortical cells also showed high levels of antibody binding and these tissues also correlate anatomically with the regions in which Oparka et al. (1994) observed phloem transport and unloading. Salamatova (1993) has also reported similar patterns of auxin distribution in Monstera roots, using color reagents.

The effects on growth and development of the balance of hormone levels in plant tissues is well established. Localized accumulation of a hormone in a small population of cells could cause them to have significantly different metabolic properties than the cells surrounding them. Indeed, other investigators of the quiescent center have hypothesized that the hormone balance in the quiescent center may be different from that in adjacent regions, but speculations as to the exact difference have varied widely.

Many auxin responsive genes have now been identified and promoters for some of these have been shown to drive transcription of reporter genes in response to auxin exposure. Here we report that exogenous auxin can increase the level of AAO mRNA in root segments and that the mRNA level of AAO is significantly higher in quiescent center cells than in proximal meristem cells or mature cells of the root. Auxin was previously shown to increase levels of ascorbate oxidase activity in tobacco pith tissue and cultured pumpkin fruit (Newcomb, 1951; Esaka et al., 1992) and the present results show the same effect in segments of maize root tissue. It would be of interest now to determine the mechanism for these higher levels of AAO in the quiescent center, and to determine if this gene has an auxin-inducible or auxin-sensitive promoter.

AAO is a plant-specific copper-binding protein that has been localized mainly to the cell wall but there are reports of localization to other cellular compartments in many different cell types. It is very effective at oxidizing ascorbic acid, a compound necessary for many metabolic reactions and for the transition from G₁ to S in the cell cycle. We have shown high levels of AAO protein and AAO activity in QC cells relative to proximal meristem cells, and we suggest this causes the lack of detectable ascorbic acid in the QC, and as a consequence, the reduced mitotic activity there. These results are consistent with other reports of the involvement of ascorbic acid in the progression of G₁ phases in the cell cycle as well as in other aspects of plant cell growth (Cordoba et al., 1994). Moreover, as a redox reaction, the oxidation and ratio of ascorbic acid to dehydroascorbic acid in regions of the root suggests that the ascorbate system could be part of a larger redox regulatory system. There is a growing body of data that show redox potential can influence the regulation of gene expression and that modulation of the redox potential in cells is indeed an important regulatory system for cellular functions (Allen, 1993; Crane, 1994). When root tips are cultured in the presence of ascorbic acid, cells in the QC are activated to divide, and when root tips are treated with an inhibitor of ascorbic acid biosynthesis, cell division is inhibited and cells arrest in G₁ (Liso et al., 1984, 1988). Hence we conclude that the localized depletion of a substance that is essential for many cellular processes, and especially for the completion of the cell cycle, should be viewed as an important factor in maintaining the quiescent state of these cells.

How can this biochemical model be used to answer questions of meristem organization? Hejnowicz and Hejnowicz (1991) have modeled the formation of root apical patterning using a biophysical perspective based on growth tensor analysis. One aspect of their modeling predicts the cellular pattern generated when a focus of slow growth is imposed in an otherwise actively growing uniform growth field. Using given rules for cell wall placement with respect to the principal directions of growth, the model quite strikingly predicts the formation of a closed meristem cell pattern as shown in maize root apices (Hejnowicz and Hejnowicz, 1991). When a different tensor is applied, such that the focus region becomes the region of maximal growth, the model predicts the pattern in a pteridophyte root with a single apical cell. This suggests that root apical patterning may be controlled by regulation of relative growth rates at this focal point in the developing root meristem.

Our model is essentially a mechanism for controlling the growth rate at this focal point. Therefore, experimental changes that alter the steady state would be expected to have an effect on apical patterning. For instance, inhibition of polar auxin transport severely disrupts organization of the root meristem. This has been shown for embryonic roots, lateral roots, intact primary roots, and regenerating roots in many families of plants (Schiavone and Cooke, 1987; Liu et al., 1993; Hinchee and Rost, 1992; Stange, 1988; Kerk and Feldman, 1994). Hence alterations of other components of the model that relate to the growth rate at the quiescent center should also disrupt normal patterning of the root apex.

We suggest that our proposed biochemical model complements this biophysical growth model by providing a mechanism for regulating the relative growth rate in this focal region of a developing root meristem. Our model (Fig. 1) accounts for the localized depletion of ascorbic acid in the QC, and hence for the rate of cell growth and division in this specific region of the root tip. In addition our work supports the models of Hejnowicz and Hejnowicz which predict that localized control of growth rates here is a determining aspect in generating the pattern of cell walls in a given root tip. These growth rates may be different in different species and at different times during development, and thus underlie the diversity of anatomical patterns that have been reported in root tips. A similar type of mechanism may even underlie the organization of other kinds of meristems.

We are very grateful to Dr Ian Sussex for his critical reading of the manuscript and helpful discussions and suggestions. We are also very grateful to Dr Steven Ruzin who was so generous with his help with microscopy and image processing in the NSF Center for Plant Development at Berkeley. Thanks also to Dr A. Shinmyo, Department of Fermentation Technology, Osaka University, Japan, for his generous gift of the ascorbate oxidase cDNA clone and antibodies that we used in this study. We thank Prof. E. W. Weiler, Ruhr-Universität Bochum, for the monoclonal antibody to auxin, and we also thank Dr Sundaresan, Cold Spring Harbor Labs for the p34cdc2 cDNA clone. This work was supported by a USDA post-doctoral fellowship to N. Kerk, and an NSF grant to L. Feldman.