|Home | About | Journals | Submit | Contact Us | Français|
C. elegans vulval development has been an important paradigm for investigating cell fate specification and elucidating core signaling pathways and modulators (reviewed in ). The Vulval Precursor Cells (VPCs) are spatially patterned during the L3 stage by the EGFR-Ras-MAPK-mediated inductive signal and the LIN-12/Notch-mediated lateral signal. The pattern is both precise and robust , due to multiple positive and negative feedback mechanisms underlying crosstalk between these pathways (reviewed in ). Signaling is also regulated temporally, as constitutive activation of the spatial patterning pathways does not alter the timing of VPC fate specification [4, 5]. The heterochronic genes, including the microRNA lin-4 and its target lin-14, constitute a temporal control mechanism used in different contexts (reviewed in [6, 7]; ). We find that lin-4 specifically controls the activity of LIN-12/Notch through lin-14, but not other known targets, and that persistent lin-14 blocks LIN-12 activity without interfering with the key events of LIN-12/Notch signal transduction. In the L2 stage, there is sufficient lin-14 activity to inhibit constitutive lin-12. Our results suggest that lin-4 and lin-14 contribute to spatial patterning through temporal gating of LIN-12. We propose that in the L2 stage lin-14 sets a high threshold for LIN-12 activation to help prevent premature activation of LIN-12 by ligands expressed in other cells in the vicinity, thereby contributing to the precision and robustness of VPC fate patterning.
The Vulval Precursor Cells (VPCs) are spatially patterned during the L3 stage. The LIN-3/EGF inductive signal produced by the anchor cell of the gonad causes the closest VPC, P6.p, to adopt the 1° vulval fate and to produce ligands for LIN-12/Notch. These ligands constitute a lateral signal that causes the neighboring VPCs, P5.p and P7.p, to adopt the 2° vulval fate.
Heterochronic genes were good candidates for regulating the timing of spatial patterning. In lin-4(e912) null mutants [lin-4(0)], seam cells born in any larval stage adopt the fate normally associated only with the L1 stage. If, by analogy with seam cells, VPCs retain an L1 temporal identity, they should not respond to the spatial patterning signals in the L3 stage.
To assess the ability of P6.p to respond to the inductive signal we examined 1° fate markers: egl-17, which is expressed in response to EGFR-Ras-MAPK activation in the L2 and L3 stages , and apx-1 and lag-2, the lateral signal genes, which are expressed only in the L3 stage (; X. Karp and I. G., unpublished observations). All three markers are expressed at their normal times in lin-4(0) mutants and, as in wild-type, 1° marker expression depends on the presence of the gonad (Fig. 1A). All VPCs are competent to respond to the inductive signal, since a 1° fate marker is also expressed in all VPCs in lin-15(n309), in which the inductive signal is ectopically expressed in the major hypodermal syncytium  (Fig. 1B).
In contrast, P5.p and P7.p do not appear to respond to the LIN-12/Notch-mediated lateral signal, as evident in the lack of expression of 2° fate markers: two direct LIN-12 transcriptional targets, lst-5 and mir-61 (Fig. 1C), as well as lin-11 (Fig.2A; ). Expression of the lateral signal genes (Fig. 1A) and normal LIN-12 downregulation (data not shown) suggest that the lateral signal is sent, and most VPCs divide at the normal time in lin-4(0) mutants (Fig. S1A,B; [12, 13]), indicating that this block to LIN-12 signaling does not result from failure to pass through S phase .
The ability of VPCs to respond to the inductive signal in lin-4(0) mutants suggests that they have an L3 “temporal identity”, and the loss of the 2° fate suggests that lin-4 is specifically required for lin-12/Notch activity at this stage.
lin-4 expression begins in the VPCs soon after they are born in the L1 stage ( and Fig. S1C,D) and appears to function autonomously to allow lateral signaling to occur (Fig. S1E,F). To test whether loss of lin-4 activity blocks signal sending or signal reception, we assessed whether constitutive LIN-12 activity can overcome the lateral signaling defect in lin-4(0).
After LIN-12/Notch is activated by ligand, proteolytic cleavages in the extracellular and transmembrane domains release the intracellular domain, which translocates to the nucleus and activates target gene transcription (reviewed in ). Constitutive lin-12 activity causes all six VPCs to adopt the 2° fate and results in a Multivulva phenotype (Fig. 2A). We used three different constitutively active forms of LIN-12 to probe the level at which lin-4(0) blocks LIN-12 signaling: LIN-12(n137), a missense mutant that requires both extracellular and transmembrane cleavages to release the intracellular domain; LIN-12(ΔEΔDTS)::GFP, which requires transmembrane cleavage; and LIN-12(intraΔP), which consists of the untethered intracellular domain (see Supplemental Experimental Procedures). lin-4(0) completely suppressed all three constitutively active forms (Fig. 2A and data not shown), consistent with a block downstream of the cleavage events.
LIN-12(ΔEΔDTS)::GFP also allows us to visualize the localization of the intracellular domain of LIN-12. Although the Multivulva phenotype was completely suppressed, lin-4(0) did not alter the apparent level or nuclear accumulation of GFP (Fig. 2B). These results strongly suggest that loss of lin-4 acts downstream of receptor cleavage and nuclear import to block LIN-12 activity.
Loss of lin-4 causes persistence of its two confirmed targets, LIN-14 and LIN-28, and may also cause persistent hbl-1 activity[17–21]. Although the activity of these genes must be down-regulated for correct temporal specification of seam cells, we found that only LIN-14 appears to be a relevant target for lateral signaling: loss or reduction in lin-14 activity, but not of lin-28 or hbl-1, restored expression of the 2° fate in a lin-4(0) background (Fig. 3A; Fig. S2A). Thus, LIN-14 is necessary to block lateral signaling in lin-4(0).
Persistent LIN-14 is also sufficient to block lateral signaling. In lin-14(n355) hermaphrodites, in which deletion of the lin-4 binding sites in the 3' UTR results in persistent LIN-14 accumulation , markers for the 1° fate (Fig. 3B; Fig. S2B), but not the 2° fate (Fig. 3B), were expressed. In contrast, a form of lin-28 engineered to lack the lin-4 binding site causes a retarded phenotype in lateral hypodermis but does not affect expression of a 2° fate marker (Fig. S2C), further supporting the inference that persistent LIN-28 does not cause a lateral signaling defect. Taken together, these results suggest that LIN-14 inhibits lin-12/Notch signaling and must be down-regulated in VPCs by lin-4 to permit 2° fate specification.
The temperature-sensitive period for LIN-12-mediated 2° fate specification is in the L3 stage prior to VPC division . If LIN-14 blocks lateral signaling per se, we would expect to be able to reverse this block by lowering lin-14 activity near this time. Alternatively, if LIN-14 prevents VPCs from maturing and attaining competence to respond to the lateral signal, we might expect that reversing this block might require that lin-14 activity be lowered through much of the L2 stage, or even, prior to the L2 stage, when lin-14 acts to prevent precocious vulval induction .
To determine the temperature-sensitive period for blocking lateral signaling, we first attempted to use lin-14(n355n679ts), an allele that encodes a heat-sensitive LIN-14 protein (due to the n679 missense mutation) that is no longer subject to lin-4 regulation (due to n355, which deletes the lin-4 binding sites) [22–24]. However, at 15°C, lateral signaling was not affected (Fig. S3), suggesting that the mutant LIN-14(n679) protein is less active or less stable and is not sufficiently active to block LIN-12. By adding lin-4(0) to stimulate potential positive feedback from lin-28 [19, 25], we increased lin-14(n355n679ts) activity sufficiently to inhibit lateral signaling at 15° (Fig. S3), but the activity was still heat-sensitive such that at 25°C, most lin-4(0); lin-14(n355n679ts) hermaphrodites have normal lateral signaling (Fig. 4A).
When lin-4(0); lin-14(n355n679ts) hermaphrodites were shifted from 15° to 25° at different times, we found that a shift to 25° prior to the L2 molt permitted lateral signaling, but a shift at or after the L2 molt did not (Fig. 4A). Thus, reducing lin-14 activity before lateral signaling rescues the defect caused by loss of lin-4, but maintaining high lin-14 activity near the time of lateral signaling blocks lin-12 activity. These observations suggest that LIN-14 inhibits lateral signaling per se rather than the competence of the VPCs to adopt the 2° fate.
Although LIN-12 is present in the L2 stage , constitutively active LIN-12 does not specify the 2° fate until the L3 stage , indicating that there is a mechanism to prevent premature LIN-12 signaling. Since persistent LIN-14 blocks lin-12 activity in heterochronic mutant contexts, lin-14 might be part of this mechanism. Although functional LIN-14::GFP fades away rapidly after the VPCs are born and cannot be detected in the VPCs in the L2 stage ([24, 27]; data not shown), a low level of LIN-14 may still be present and detectable in a functional assay.
To test this hypothesis, we used lin-14(n179ts), a heat-sensitive allele that remains subject to regulation by lin-4: at 15° lin-14(n179ts) has relatively normal lin-14 activity, but at 25°, activity is reduced [12, 23]. When we maintained lin-12(n137); lin-14(n179ts) hermaphrodites at 15° until the L1 molt (to prevent precocious vulval induction; ) and shifted them to 25° to reduce lin-14 activity, we found that expression of the lin-12 target gene mir-61 is enhanced in the L2 stage (Fig. 4B). The degree of reporter expression seen at the L2 stage in this background is comparable to that seen at the L3 stage in a lin-12(n137) background (Fig. 4B), suggesting that in wild type, lin-14 negatively regulates lin-12 activity at the L2 stage.
In the lateral hypodermis, mutations in heterochronic genes cause seam cells born in one larval stage to adopt fates that are characteristic of other larval stages . In particular, lin-4(0) causes seam cells to display a "retarded" phenotype, such that seam cells born in subsequent larval stages adopt the fate normally found only in the L1 stage. This retarded phenotype results from the persistence of the lin-4 targets LIN-14 and LIN-28[19, 29]. In contrast, we found that the VPCs of lin-4(0) mutants do not have an overall retarded phenotype as they appear to respond to the inductive signal. Instead, they have a specific and dramatic failure to respond to the LIN-12/Notch-mediated lateral signal. This failure results from persistence of LIN-14, and does not involve LIN-28 or HBL-1. We also found that LIN-14 blocks lin-12 constitutive activity without affecting processing or nuclear import of the LIN-12 intracellular domain, implying that signal transduction per se is unaffected. As LIN-14 can function as a transcriptional repressor , a simple model is that LIN-14 acts to antagonize expression of a key target, or targets, of the lateral signaling pathway.
We further observed that the normal level of lin-14 activity present in the L2 stage is sufficient to inhibit constitutive LIN-12 activity. This observation suggests a novel and unexpected role for lin-14 in contributing to the precision and robustness of VPC fate patterning by inhibiting premature signaling by LIN-12 if inappropriately activated by ligands produced by other cells in the local environment (Fig. 4C,D).
LIN-12/Notch signaling is used in multiple cellular contexts and at multiple times during animal development, requiring mechanisms to ensure that desired signaling occurs at the correct time and place, and conversely, that undesired signaling does not occur at an inappropriate time or place. The importance of preventing inappropriate activation of Notch has been demonstrated in the developing C. elegans gonad, where certain somatic gonadal cells effectively form a physical barrier that prevents other somatic gonadal cells that express the LIN-12 ligand LAG-2 from inappropriately activating Notch in the proximal germline .
When it is not anatomically possible to isolate cells, there may be analogous genetic barriers. During the L2 stage, the VPCs constitute the ventral hypodermis of the mid-body, extending over a large area and in close proximity to many cells known to express lag-2 (; X. Zhang and I. G., unpublished observations; see Fig. 4C), and potentially other ligands . This anatomy would preclude isolating ligand-expressing cells from the VPCs by a physical barrier as in the somatic gonad. A genetic barrier such as that afforded by LIN-14 would help prevent VPCs from adopting the 2° fate outside of the spatial patterning context of the L3 stage. When lin-14 activity is further lowered in the L3 stage, potentially by expression of other microRNAs [15, 33–37], ectopic activation is less likely because lag-2 is no longer expressed in ventral cord neurons (X. Zhang and I. G., unpublished observations) and the inductive signal has led to LIN-12 downregulation in P6.p .
Here, we have shown that the microRNA lin-4 is essential for lin-12/Notch signaling and that it functions downstream of signal transduction by inhibiting a new negative regulator of this pathway, LIN-14. In contrast, in Drosophila, a different logical circuit that incorporates microRNAs that act downstream of Notch signal transduction has been described: three families of microRNAs inhibit two different sets of Notch transcriptional targets, presumably opposing the effects of Notch signal transduction . These observations illustrate different ways that microRNAs can be used to modulate Notch signaling to ensure correct cell fate specification during development.
We thank Oliver Hobert, Xantha Karp, Daniel Shaye, Claire de la Cova, and Marcus Vargas for helpful discussions and critical reading of the manuscript, and Victor Ambros, Xinyong Zhang and Min-sung Choi for other helpful advice and information during the course of this work. We are grateful to Richard Ruiz and Xinlan Zhou for excellent technical assistance. Some of the strains used in this study were provided by the Caenorhabditis Genetics Center (CGC), which is supported by the National Institutes of Health-National Center for Research Resources. This work was supported by a grant to I.G. from the National Institutes of Health (R01CA095389). I.G. is an Investigator with the Howard Hughes Medical Institute.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.