Receptor-Like Kinase ACR4 Restricts Formative Cell Divisions in the Arabidopsis Root

See allHide authors and affiliations

Science  24 Oct 2008:
Vol. 322, Issue 5901, pp. 594-597
DOI: 10.1126/science.1160158


During the development of multicellular organisms, organogenesis and pattern formation depend on formative divisions to specify and maintain pools of stem cells. In higher plants, these activities are essential to shape the final root architecture because the functioning of root apical meristems and the de novo formation of lateral roots entirely rely on it. We used transcript profiling on sorted pericycle cells undergoing lateral root initiation to identify the receptor-like kinase ACR4 of Arabidopsis as a key factor both in promoting formative cell divisions in the pericycle and in constraining the number of these divisions once organogenesis has been started. In the root tip meristem, ACR4 shows a similar action by controlling cell proliferation activity in the columella cell lineage. Thus, ACR4 function reveals a common mechanism of formative cell division control in the main root tip meristem and during lateral root initiation.

Unlike animals, plants produce new tissues and organs primarily postembryonically from pluripotent stem cells in the root and shoot meristems. In Arabidopsis root tips, the embryonic stem cell niche, a single layer of initial cells surrounding the quiescent center, is well characterized (1). However, the branching process in roots depends on the formation of new meristems starting from a limited number of pericycle lateral root founder cells (2). The mechanisms underpinning the restriction of formative cell division to a few pericycle cells and the specification of stem cell identity in this process remain unresolved.

To gain insight into this process, we performed live imaging on longitudinal pericycle cell files during lateral root initiation in Arabidopsis. Time-lapse recordings revealed a repeated cell division pattern composed of two successive rounds of asymmetric cell divisions, generating a central core of four small cells and two larger flanking cells (Fig. 1, A to D, fig. S1, and movie S1). To achieve this, the original pericycle lateral root founder cells undergo an initial asymmetric division to generate a smaller daughter cell and a larger flanking cell. The latter will undergo another asymmetric division, resulting in a central core of small cells. Hereafter, the process of anticlinal asymmetric cell divisions stops, and the two central cells change their axis of division by 90° and divide periclinally (Fig. 1E and fig. S1). The flanking and the adjacent undivided pericycle cells undergo few or no anticlinal divisions and will only contribute modestly to the flanks of the primordium (2).

Fig. 1.

Identification of asymmetric cell division genes during lateral root initiation. (A to E) In vivo analysis of consecutive divisions in one file of xylem pole pericycle cells visualized with the plasma membrane marker p35S::EGFP:LTI6a (black). Arrowheads mark the oblique cell wall between two adjacent pericycle cells (red), consecutive anticlinal asymmetric cell divisions (blue), and first periclinal divisions of the central cells (green). Color code used for the overlay corresponds to fig. S1. (F) Diagram depicting the different filters to which the microarray data were subjected to reveal a subset of 15 candidate regulators and the identification of ACR4.

Exogenous application of the plant hormone auxin induces cell division and results in lateral root initiation in the entire pericycle at the xylem poles in Arabidopsis (3). Auxin can thus be used to synchronously induce asymmetric cell divisions of cells, facilitating analysis of molecular mechanisms of lateral root formation. To identify molecular components controlling this essential cell division pattern, we characterized the transcript profile of auxin-activated pericycle cells precisely located at the xylem pole using fluorescence-activated cell sorting in combination with a highly synchronized time course (4, 5). We identified 1920 significantly differentially expressed genes (table S1). K-means clustering of the expression patterns leads to the delineation of 10 profiles that show temporal changes in expression levels (fig. S2). The reproducibility of the expression profiles, the resolution of our experimental approach, and the potential involvement in lateral root development were evaluated using quantitative polymerase chain reaction, promoter–beta-glucuronidase (GUS) reporters, and mutants, respectively (figs. S3 to S5 and table S2). This combination of synchronized lateral root induction and cell sorting yielded a transcript profile data set reflecting lateral root initiation with high spatial and temporal resolution.

To identify candidate factors involved in regulating the asymmetric cell division pattern, we subjected our data set to several filters (5) and identified 15 potential key regulatory genes for the process of asymmetric cell division and cell fate specification during lateral root initiation (Fig. 1F and table S3). One promising candidate, the only gene identified by all of the filter criteria (Fig. 1F and table S3), encodes the membrane-localized receptor-like kinase ARABIDOPSIS CRINKLY4 (ACR4, AT3G59420) (6). ACR4 is transcribed specifically in the small daughter cells after the first asymmetric pericycle cell division. Subsequently, the expression expands to the adjacent small daughter cells from the second asymmetric cell division, resulting in a central core-specific expression pattern (fig. S6, and movie S2).

Because several receptor-like kinases have been demonstrated to regulate patterning, establish cell identities, and specify cell fate (710), we investigated whether the asymmetric expression of ACR4 not only reflects organized pericycle division but also is causal for this process. We determined that acr4 exhibited a significant increase (19%) in the total number of lateral root meristems (LRMs)/cm, compared with wild-type (Fig. 2A and table S4). Because ACR4 is a member of the CRINKLY4 gene family in Arabidopsis (11), redundancy could be expected. Notwithstanding that the other family members were not identified in the transcript profiling, double- and triple-mutant combinations of acr4 with mutations in the other four family members (CRR1, CRR2, CRR3, and CRR4) exhibited even higher LRM densities (Fig. 2A and table S4). In wild type, LRMs exhibit a left-right alternation with regular spacing of lateral roots (12) and are never initiated opposite each other (Fig. 2, B and D, and table S5). In the mutants of the ACR4 gene family, LRMs were initiated close to one another (fig. S7) and often in the normally excluded opposite positions (Fig. 2, B and J, table S5, and fig. S7). The mutants also had stretches of a two-layered pericycle or fused primordia, both abnormal features (Fig. 2, B and E, and table S5). During lateral root initiation and primordium development, mutant pericycle cells frequently exhibited unusual mitotic activity and auxin response, as visualized by pCYCB1;1::GUS (13) and pDR5::GUS (14), respectively (Fig. 2, F to J). At a later stage, boundaries of the wild-type lateral root primordium showed expression of pLBD5::GUS (Fig. 2K), but mutant lateral root primordia lacked such clear identification of their borders (Fig. 2L). However, most of the primordia at aberrant positions arrest during development (fig. S7) and do not contribute to the overall root architecture. Hence, when only emerged lateral roots are taken into account, instead of the total number of LRMs, acr4 has a significantly lower density than wild type (Fig. 2C and table S4). To get better insight in the role of ACR4 during lateral root initiation, we overexpressed ACR4 tissue-specifically in the xylem pole pericycle and could show that this line had a higher density of emerged lateral roots (fig. S8). These observations suggest that ACR4 is required to coordinate pericycle cell divisions during lateral root initiation, where it functions mainly to prevent surrounding cells from dividing but also seems to be involved in the initiation event itself.

Fig. 2.

ACR4 represses pericycle cell divisions in flanking and adjacent cells during lateral root initiation. (A to C) Lateral root phenotypes of single, double, and triple mutants for five related receptor-like kinases, including ACR4. (A) Lateral root meristem (LRM) densities (including all primordial stages; mean ± SEM). (B) Percentage of aberrant lateral root positioning (mean ± SEM). (C) Lateral root densities, only including emerged lateral roots (mean ± SEM). *, statistically significant differences for values compared with wild type as determined by Student's t-test (P < 0.05). (D to L) Division pattern and marker analysis in pericycle cells during lateral root initiation in wild type [(D), (F), (H), and (K)] and representative examples of single or double mutants for five related receptor-like kinases, including ACR4 [(E), (G), (I), (J), and (L)]. Pericycle shows increased cell divisions (black arrowheads); thus, the primordium boundary and central core are not clearly defined in acr4crr3 (E) compared with wild-type primordium (D). Markers pCYCB1;1::GUS [(F) and (G)], pDR5:: GUS [(H) to (J)], and pLBD5::GUS [(K) and (L)] show altered expression patterns in mutants [(G), (I), (J), and (L)] compared with wild type [(F), (H), and (K)]. White, green, and pink arrowheads indicate pericycle cell file [(F) and (G)], opposing primordia (J) and lateral root organ boundaries [(K) and (L)], respectively; broken red line separates two layers of pericycle (E).

In mutants gnomR5 (15) and slr-1 (16), which lack asymmetric pericycle cell divisions and lateral roots, ACR4 expression was also absent from the pericycle but was normal in the mutant root tips (Fig. 3, A to C). In slr-1xCYCD3;1OE, zones of short pericycle cells can be observed (Fig. 3D, inset), but still no lateral roots are formed (16). No ACR4 expression could be detected in the zones of short pericycle cells, whereas root tip expression remained again unchanged (Fig. 3D). In gnomR5, auxin induced unstructured proliferation of the pericycle (Fig. 3E, inset) but no organized lateral roots (15). ACR4 was not expressed in the resulting multilayered pericycle of gnomR5 (Fig. 3E). As in auxin-treated gnomR5, wild-type roots, exposed to the vesicle trafficking inhibitor brefeldin A (BFA) and to auxin, showed homogeneous proliferation of the pericycle (15) (Fig. 3F) and no ACR4 expression (Fig. 3G). Thus, ACR4 is correlated with formative divisions and organogenesis, and functions in lateral root formation by suppressing proliferative cell divisions in nearby pericycle cells.

Fig. 3.

ACR4 expression correlates with formative divisions and organogenesis. (A) pACR4::H2B:YFP expression (green) in wild-type roots in the central core of the lateral root initiation site. (B to G) pACR4::H2B:YFP expression in seedling roots without [(B) to (D)] and with [(E) to (G)] auxin treatment in gnomR5 [(B) and (E)], slr-1 (C), slr-1xCYCD3;1OE at zone of short pericycle cells (D), and Col-0 treated with BFA [(F) and (G)]. The inset shows the unaltered expression in the main root tip [(B) and (C)], the zone of short pericycle cells (D), or the multilayered pericycle (E) (the last two are recalcitrant to propidium iodide staining). Black arrowheads, uncoordinated cell divisions [(D) to (G)]; white arrowheads, pericycle [(B) to (E)]. Broken red line marks the multilayered pericycle [(F) and (G)]. Propidium iodide (red) was used to visualize cell boundaries [(A) to (E)], and differential interference contrast optics to visualize pericycle cell size. Scale bars, 50 μm.

ACR4 is also expressed in the root apical meristem in columella and epidermis/lateral root cap initials and their respective derivatives (6) (fig. S9). Therefore, we investigated whether ACR4 has a similar role in mediating formative divisions in the root apex. Differentiation of columella cells can be monitored by lugol staining of the starch grains that are present in columella cells but not in columella initials (17). In contrast to wild-type roots, which generally exhibited organized tiers of columella root cap cells (Fig. 4A), the root cap of the acr4 mutant is frequently distorted (Fig. 4, B and C, and table S6). Wild-type roots display usually one, occasionally two, layer(s) of unstained cells below the quiescent center, resulting from a recent synchronous division of all columella initial cells (Fig. 4, A and D). The acr4 root tips, however, often displayed additional divisions in cells of the columella cell lineage (Fig. 4, E and F, and table S6), ultimately giving rise to a distorted root cap. In view of the redundancy between members of the ACR4 family, these aberrant divisions occur more often in the triple-mutant combinations (table S6). Therefore, ACR4 function is also required to restrict division activity in daughter cells of columella stem cells in the root apex.

Fig. 4.

ACR4 represses irregular divisions in columella stem cell daughter cells in the root apex. (A to F) Root apex at 2 days after germination shows regular columella cell tiers in wild type with one (A) or two layers of undifferentiated cells after a recent division of the columella initial (D) and disorganized columella in acr4 (B) with up to three undifferentiated cells in one tier (E). Inset in (E) shows an extra cell in the uppermost layer of differentiated columella in acr4 (black arrowhead). Violet, starch grains; red asterisks, undifferentiated and dividing cells; red arrowhead, quiescent center. [(C) and (F)] Percentage of root tips with disorganized columella (C) and additional columella cell divisions measured by the appearance of additional non-starch staining initials (F).

Plants and animals require asymmetric cell divisions, coinciding with the acquisition of the correct cell fate, for growth and reproduction (18, 19). For instance, the asymmetric cell division mechanism is central to the activity of stem cells (20), and recent studies have shown a correlation with cancer-like states of the cell when asymmetric cell divisions do not take place (21). We have shown that the receptor-like kinase ACR4 represses supernumerary formative divisions of root cells, both in pericycle cells during lateral root initiation and in the columella in the root apex (fig. S10). Our data suggest that ACR4 signaling is a critical homeostatic mechanism in mediating formative divisions in pluripotent root tissue during organogenesis and might act both cell autonomously and non–cell autonomously. Cell autonomously, ACR4 might be required for correct specification of lateral root primordia cells, as can be deduced from enhanced lateral root formation in the gain-of-function plants and from its strict transcriptional correlation with formative divisions in the pericycle. This is in agreement with ACR4-dependent cell specification that has been proposed for the L1 layer (6). Non–cell autonomously, ACR4-signaling might prevent neighboring pericycle cells from becoming triggered for lateral root initiation. Interestingly, a similar dual role for the maize homolog CRINKLY4 was proposed in the L1 layer based on the results of a genetic mosaic analysis (22). ACR4 function reveals the common mechanisms for both root apical meristem and developing lateral root primordia and is likely to be mechanistically related to its role in the L1 layer.

In the future, the ACR4 pathway regulating root apical and lateral meristem function may be worth comparing to the CLAVATA receptor-like kinase pathway that functions in the shoot apical meristem to maintain the size of the stem cell pool.

Supporting Online Material

Materials and Methods

Figs. S1 to S13

Tables S1 to S11

Movies S1 and S2


References and Notes

Stay Connected to Science

Navigate This Article