Report

Genetic control of distal stem cell fate within root and embryonic meristems

See allHide authors and affiliations

Science  06 Feb 2015:
Vol. 347, Issue 6222, pp. 655-659
DOI: 10.1126/science.aaa0196

Genetic control of stem cell fate in plant roots

Without roots, most plants cannot thrive. Crawford et al. have now unearthed the robust control systems that build roots. Signaling by the plant hormone auxin triggers three genes that control the development of stem cells forming the root. With this trio of genes, any one of which can do the job, root development is backed up with fail-safe controls. The team could use the same system of controls to sprout roots in the wrong places, making roots instead of shoots.

Science, this issue p. 655

Abstract

The root meristem consists of populations of distal and proximal stem cells and an organizing center known as the quiescent center. During embryogenesis, initiation of the root meristem occurs when an asymmetric cell division of the hypophysis forms the distal stem cells and quiescent center. We have identified NO TRANSMITTING TRACT (NTT) and two closely related paralogs as being required for the initiation of the root meristem. All three genes are expressed in the hypophysis, and their expression is dependent on the auxin-signaling pathway. Expression of these genes is necessary for distal stem cell fate within the root meristem, whereas misexpression is sufficient to transform other stem cell populations to a distal stem cell fate in both the embryo and mature roots.

Development of plant roots depends on regulation of stem cell function in the root apical meristem, where the quiescent center separates two populations of stem cells into proximal and distal domains (Fig. 1A) (1, 2). The cells in the quiescent center rarely divide themselves but signal to surrounding stem cells to remain undifferentiated. The quiescent center is formed during embryo development when the uppermost cell of the suspensor, known as the hypophysis, divides asymmetrically to initiate the root meristem. Here we analyze the signals that first establish the quiescent center and distal stem cell populations.

Fig. 1 Phenotype of the nww mutant and expression of NWW genes in embryos.

(A) Stages in root meristem development, from an early globular-stage embryo to a mature root meristem with different regions of the meristem labeled and colored. QC, quiescent center. (B and C) Seedling phenotype of the wild type (B) and nww mutant (C). The arrow in (C) indicates the absence of the root. (D to G) Arrows indicate root meristem initiation in globular (glob) and heart-stage embryos of the wild type [(D) and (F)] and nww mutant [(E) and (G)]. (H to K) Recombineered NTT [(H) and (I)], WIP4 (J), and WIP5 (K) reporters in a 32-cell or early globular stage before root meristem initiation [arrows in (H), (J), and (K)] and a late globular stage (I). Arrows point to hypophysis (hypo) cell in (H), (J), and (K). Scale bars, 20 μm in (D) to (K).

The NO TRANSMITTING TRACT (NTT) gene encodes a putative zinc finger transcription factor that is required in developing carpels for transmitting tract formation (3, 4). NTT is also expressed in root meristems, although ntt homozygous mutant roots are normal. To determine if NTT acts redundantly with other genes, we combined mutations in NTT with mutations in WIP DOMAIN PROTEIN 4 (WIP4) and WIP5, the two genes most closely related to NTT (5) (see supplementary materials and methods). Although plants lacking one or two of these genes had normal roots, plants lacking all three genes—ntt, wip4, and wip5 (hereafter referred to as nww)—had no roots (Fig. 1, B and C, and fig. S1). Thus, the NWW function is required for root formation, and these three genes can each individually fulfill that function.

Development of wild-type (WT) and nww mutant embryos diverges at the 32-cell (early globular) stage, when the hypophysis normally divides asymmetrically to form the lens-shaped cell that goes on to generate the quiescent center (Fig. 1A). In nww mutants, this asymmetric cell division fails to occur, and the lens-shaped cell does not form (Fig. 1 and fig. S1). Because activity of the root meristem is dependent on the quiescent center, loss of the lens-shaped cell likely explains the absence of roots in the nww mutant.

As WT embryos continue to develop, the root meristem becomes more distinct. In nww mutant embryos, this region is highly disorganized, lacking a recognizable root meristem (Fig. 1 and fig. S1). In nww mutant embryos, cell divisions above the hypophysis appear normal at the globular stage, but after failure of that key asymmetric cell division, they become progressively more abnormal. These cell division defects may be a result of the loss of the quiescent center, which normally sends a signal to the proximal stem cells, and these defects are reflected postembryonically by the lack of vasculature tissue in the hypocotyl of the nww mutant (fig. S2).

To determine if the pattern of expression and protein accumulation for NTT correlated with the observed mutant phenotypes, we created a recombineered construct expressing 2XYPET fused to NTT under the control of the entire NTT locus, gNTT-n2YPET (6, 7). This construct was capable of complementing the ntt single- and nww triple-mutant phenotypes. The expression pattern was indistinguishable from in situ hybridization patterns (fig. S3). NTT was expressed in the hypophysis cell in globular-staged embryos (Fig. 1H) and was maintained as the hypophysis gave rise to the apical and basal daughter cells (Fig. 1I). WIP4 and WIP5 were similarly expressed in the hypophysis but, in contrast with NTT, not in the suspensor (Fig. 1, H to K). Thus, the NWW genes are all expressed in the hypophysis, where they act redundantly to promote root meristem initiation.

The rootless phenotype of the nww triple mutant resembles mutants in the auxin response factor MONOPTEROS (MP/ARF5) (8). We therefore examined accumulation of NTT in mp mutants to determine if NTT accumulation was dependent on MP. Whereas gNTT-n2YPET accumulation was observed in the hypophysis of WT embryos (Fig. 2A), little or no accumulation in the hypophysis was detected in mp mutants (Fig. 2B and fig. S4). This suggests that NTT acts downstream of MP in promoting root meristem initiation.

Fig. 2 Expression analysis in WT and mutant embryos.

(A and B) Arrows indicate gNTT-n2YPET accumulation in divided hypophysis of WT (A) and mp mutant (B) globular-stage embryos. (C and D) In situ hybridization of MP in globular-stage embryos. To ensure that we were not detecting background signal, we used a segregating mp mutant population along with an in situ probe that only detected WT MP transcript. Arrows show expression in outlined hypophysis (C) of early globular-stage embryo, but no expression in a phenotypically mp mutant (D) late globular-stage embryo. (E) Accumulation of recombineered gMP-n2YPET throughout an early globular embryo. The arrow indicates accumulation within the hypophysis. (F) Diagram of the genetic pathway to promote root initiation. (G and H) Normal initiation of pDR5-GFP expression is indicated by arrows in the wild type (G) and nww mutant (H). (I and J) Normal initiation of pWOX5-GFP expression indicated by arrows in the wild type (I) and nww mutant (J). Scale bars, 20 μm in (A) to (E) and (G) to (J).

To better understand the relationship between MP and NTT, we examined MP expression within the embryo. A recombineered MP 2XYPET reporter, gMP-n2YPET, accumulated broadly in the embryo, including all of the suspensor region that includes the hypophysis (Fig. 2E and fig. S5). Similarly, in situ data and the pMP-MP–green fluorescent protein (GFP) reporter showed expression in the hypophysis (fig. S5, D to G). We detected MP expression in WT embryos but not in the mp mutant embryos (Fig. 2, C and D). RNA-seq data confirm expression of MP in the suspensor region (http://seedgenenetwork.net) (fig. S5H). We conclude that MP is expressed in the hypophysis, where it promotes root meristem initiation by activating the NWW genes. To provide further support for the idea that NTT functions downstream of MP, we used a chromatin immunoprecipitation assay and found that MP binds in vivo to a conserved AuxRE binding site within the NTT intron (fig. S4F). Similar AuxRE sites are conserved in the introns of both WIP4 and WIP5, and these AuxRE sites are also found in orthologs of the NWW genes in other Brassicaceae species (fig. S4G). These data are consistent with NTT functioning downstream of MP and suggest that MP directly binds to NTT regulatory sequences to promote expression.

To provide further insights into the nature of the defects seen in the nww triple mutant, we analyzed the expression of several relevant markers. We first looked at the expression of the pDR5-GFP marker that contains several AuxRE sites and reflects auxin signaling (911). In WT embryos, pDR5-GFP expression was observed in the hypophysis, and the loss of this expression in mp mutants has been linked to the absence of root meristem initiation (12). We found that the initial expression of the pDR5-GFP marker was similar in both WT and nww mutant embryos at the globular stage (Fig. 2, G and H, and fig. S6, A and B). Thus, in contrast to MP, our data show that the NWW genes are not required for the initial activation of the pDR5-GFP marker, consistent with the NWW genes functioning downstream of MP. At later stages of embryo development, expression of pDR5-GFP was much reduced in the nww mutant as compared with the wild type, presumably a consequence of the failure of the root meristem to form (fig. S6).

We next examined the expression of WUSCHEL-RELATED HOMEOBOX5 (WOX5), a gene that is expressed in the quiescent center and is required for columella stem cell maintenance (13). During root meristem initiation, pWOX5-GFP is first detected in the hypophysis and is absent in mp mutants (13), which suggests that, similar to pDR5-GFP, pWOX5-GFP is also dependent on MP expression. In contrast, we found that the initial expression of pWOX5-GFP was normal in the nww mutant (Fig. 2, I and J), although later expression was diffuse and mislocalized apically within the nww mutant embryo (fig. S6). Thus, NWW genes are not required for the initial activation of WOX5 but are required for maintenance of the WOX5 expression pattern within the root meristem. Thus, even though the MP and NWW genes are normally required for root initiation, other MP-dependent targets are still activated independently of NWW genes (Fig. 2F).

Expression of the three NWW genes persists in the root meristem during postembryonic root development. We detected gNTT-n2YPET expression within the mature root meristem in the quiescent center, columella stem cells (also referred to as columella initials), and columella cells (Fig. 3A). We similarly found that gWIP4-n2YPET and gWIP5-n2YPET were expressed in the mature root meristem, although their expression was restricted to the quiescent center and columella initial cells (Fig. 3B and fig. S3J). Thus, the NWW genes may be needed to maintain meristem integrity in the developing root.

Fig. 3 Control of distal meristem fate within the root.

(A and B) Accumulation of NTT (A) and WIP4 (B) in the quiescent center and columella initials (arrow) of the root meristem. (C and D) Rescued root phenotype of the wild type (C) and nww mutant (D) after auxin treatment and 2 days after transfer to MS media. (E to L) Phenotype of the wild type [(E), (G), (I), and (K)] and nww mutant [(F), (H), (J), and (L)]. The QC25 marker is indicated by an arrow in the wild type (E) and is absent in nww mutant (F). pPIN7-PIN7-GFP marker, with distal expression indicated by an arrow in the wild type (G) and absent in the nww mutant (H). pWOX5-GFP [(I) and (J)] and pDR5-GFP [(K) and (L)] markers in the wild type [(I) and (K)] and nww mutant [(J) and (L)]. Scale bars, 20 μm in (A), (B), and (E) to (L); 1 mm in (C) and (D).

The plant hormone auxin can promote root initiation in tissue culture. Despite the requirement for MP for root initiation in the embryo, phenotypically WT roots can be induced if mp mutants are treated with auxin (14). We found that exogenous application of auxin could similarly rescue root formation in nww mutants (Fig. 3, C and D, and fig. S7). However, in contrast to mp mutants, the rescued roots of nww mutants are highly abnormal. The nww-rescued roots were larger than WT roots and failed to form amyloplasts, a marker for distal stem cell fate (Fig. 3, E and F). As a result, these rescued roots lack a gravitropic response. Our data indicate that the NWW genes are required to pattern distal stem cells within the root meristem. To further test this idea, we analyzed the PIN-FORMED 7 marker (pPIN7-PIN7-GFP), which is normally expressed in cells derived from both the proximal and distal regions of the root meristem (Fig. 3, G and H). Consistent with our hypothesis, we found that PIN7 expression was not detected in the distal region of the nww mutant root meristem yet was still present in the proximal region.

Because the stereotypical pattern of cell divisions within the root meristem is dependent on signals from the quiescent center (15), the increased size and disorganization of the nww mutant roots suggest that the quiescent center may be abnormal. We examined the expression of two quiescent center markers, QC25 and WOX5. We found that the QC25 marker was not detected in the rescued roots of the nww mutant (Fig. 3, E and F), consistent with the idea that the quiescent center is not normal. As different quiescent center markers depend on different stem cell regulators (13), we also analyzed the expression of WOX5. In WT roots, pWOX5-GFP expression was detected only within the four cells of the quiescent center, whereas expression in the nww mutant expanded to include a larger number of cells (Fig. 3, I and J).

Because WOX5 has been shown to be downstream of auxin signaling, we also analyzed pDR5-GFP expression to determine whether auxin signaling is altered in the nww mutant root (Fig. 3, K and L, and fig. S7). In WT roots, pDR5-GFP was observed in only the quiescent center and columella cells. In the rescued roots of the nww mutant, pDR5-GFP was misexpressed throughout the presumptive root cap. Although auxin signaling still occurs in the nww mutant, the altered pDR5-GFP expression pattern suggests that the roots are unable to respond to auxin in an appropriate manner. Thus, the NWW genes are required for appropriate development of the quiescent center and the distal root meristem.

Misexpression of NTT might be sufficient to transform other stem cells within the root meristem into a distal fate. To test this prediction, we created an inducible line by fusing NTT to the glucocorticoid receptor (p35S-NTT-GR) (16). In the wild type, columella cells are restricted to the distal region within the root meristem. In contrast, induction of NTT activity within mature root meristems caused ectopic production of columella cells (Fig. 4, A and B). The QC25 marker, normally expressed only in the quiescent center, was ectopically expressed in the proximal region when NTT activity was induced. Thus, NTT is both necessary and sufficient to pattern distal stem cell identity within the root meristem. The p35S-NTT-GR line can also mimic the root-inducing effects of exogenous auxin application. When WT seedlings were transferred to normal media after germinating on 10-μm 1-naphthaleneacetic acid (NAA), extra roots were produced. Similarly, extra roots formed after p35S-NTT-GR seeds were germinated in the presence of the dexamethasone inducer and then transferred to normal media (fig. S8). This is consistent with a role of the NWW genes in mediating an auxin signal for root initiation.

Fig. 4 Analyses of NTT misexpression.

(A and B) Root meristem of p35S-NTT-GR uninduced (A) and induced (B) roots stained for amyloplasts (brackets show WT staining) and with the QC25 marker with blue staining. (C and D) Seedlings of pAS1-LHG4 driver line control (C) and pAS1>>NTT (D). (E and F) Torpedo stage of WT (E) and pML1>>NTT (F) embryos with pDR5-GFP marker. Scale bars, 20 μm in (A), (B), (E), and (F).

NTT misexpression can also change stem cell fate within the embryo. In WT embryos, the apical region gives rise to the shoot apical meristem and two groups of primordial cells known as the cotyledon initials. The AS1 gene is strongly expressed in the cotyledon initials in transition-stage embryos (17). Misexpression of NTT under the control of the AS1 promoter caused roots to form instead of cotyledons in the resultant seedlings (Fig. 4, C and D, and fig. S8, E to G). This suggests that NTT expression is sufficient to transform cotyledon primordia to a root meristem fate within the apical region of the embryo.

More widespread NTT misexpression in the protodermal layer of the apical cells of early globular-stage embryos using the AtML1 promoter (pML1>>NTT) (18) resulted in embryos with asymmetrical structure, losing both the cotyledons and the shoot apical meristem (Fig. 4, E and F, and fig S8). Taken together, these studies support a model in which NTT misexpression is sufficient to pattern basal stem cell identity within the embryo and distal stem cell identity in the root meristem (fig. S8K).

There is tremendous interest in identifying the major pathways that specify stem cells in both animal and plant systems. Identification of the NWW genes will help to explain the formation of stem cells and may ultimately allow for the manipulation of the root to enhance agricultural yield. Additionally, although many regulators have been found to pattern plant meristems, it is likely that additional intrinsic factors remain undiscovered due to genetic redundancy, as is the case with the NWW genes.

Correction (10 February 2015): During the production process, an incorrect legend was substituted for the Fig. 4 legend in the print version and online PDF. This has now been corrected.

Supplementary Materials

www.sciencemag.org/content/347/6222/655/suppl/DC1

Materials and Methods

Figs. S1 to S8

Table S1

References (1929)

References and Notes

  1. Acknowledgments: We thank A. Gallavotti and B. Bargmann for their critical comments and M. Estelle and Y. Zhang for use and help with the confocal microscope. This work was funded by the NSF (grants IOS-0817544 and IOS-1121055 to M.F.Y. and IOS-0822411 to J.A.L.) and the NIH (NIH/NIGMS grant 5 R01 GM072764 to J.A.L.). The supplementary materials contain additional data.
View Abstract

Navigate This Article