A temporal sequence of heterochronic gene activities promotes stage-specific developmental events in Caenorhabditis elegans

Abstract The heterochronic genes of the nematode Caenorhabditis elegans control the succession of postembryonic developmental events. The 4 core heterochronic genes lin-14, lin-28, hbl-1, and lin-41 act in a sequence to specify cell fates specific to each of the 4 larval stages. It was previously shown that lin-14 has 2 activities separated in time that promote L1 and L2 developmental events, respectively. Using the auxin-inducible degron system, we find that lin-28 and hbl-1 each have 2 activities that control L2 and L3 events which are also separated in time. Relative to events they control, both lin-28 and hbl-1 appear to act just prior to or concurrently with events of the L2. Relative to each other, lin-28 and hbl-1 appear to act simultaneously. By contrast, the lin-14 activity controlling L2 events precedes those of lin-28 and hbl-1 controlling the same events, suggesting that lin-14's regulation of lin-28 is responsible for the delay. Likewise, the activities of lin-28 and hbl-1 controlling L3 fates act well in advance of those fates, suggesting a similar regulatory gap. lin-41 acts early in the L3 to affect fates of the L4, although it was not possible to determine whether it too has 2 temporally separated activities. We also uncovered a feedback phenomenon that prevents the reactivation of heterochronic gene activity late in development after it has been downregulated. This study places the heterochronic gene activities into a timeline of postembryonic development relative to one another and to the developmental events whose timing they control.


Introduction
Postembryonic development of the nematode Caenorhabditis elegans occurs in 4 larval stages (L1-L4) where stage-specific developmental events-including cell divisions, differentiation, and morphogenesis-complete the formation of tissues and organs of the reproductive adult.The timing and sequence of these stagespecific events are under the control of the heterochronic genes, which act in a hierarchy that controls both the stage-specific events and the activities of later-acting heterochronic regulators (Rougvie and Moss 2013).Four heterochronic genes, lin-14, lin-28, hbl-1, and lin-41, are the core factors specifying stage-specific events.Although their phenotypic effects and genetic relationships have been well-characterized, it is not known precisely when they act relative to each other and to the events they control.Outstanding questions are: Do genes that control the same cell fate events act at the same time, and do the genes act during the execution of the event they control, or prior to that event?Ambros and Horvitz (1987) used temperature-sensitive alleles of lin-14 to determine its times of action.They shifted animals from the permissive to the restrictive temperature, and vice versa, at specific times during postembryonic development.These experiments showed that lin-14 has 2 activities that occur at different times: an early activity, named lin-14a, that occurs early in the L1 stage that promotes L1-specific developmental events, and a second activity, lin-14b, that occurs hours later, in the middle of the L1 stage that promotes L2-specific events.The animal's requirement for lin-14a occurs very close to the time the L1 events occur, whereas lin-14b occurs a few hours in advance of the L2 events that it controls (Ambros and Horvitz 1987).
Similar studies have not been carried out for the other core heterochronic genes, lin-28, hbl-1, and lin-41, because temperaturesensitive alleles of these have not been identified.It is known that lin-28 controls both L2-and L3-specific events, suggesting that it too has 2 separate activities (Ambros and Horvitz 1984;Vadla et al. 2012).Like lin-28 mutants, hypomorphic alleles of hbl-1 also cause skipping of L2 events and precocious differentiation, but it has not yet been suggested that it has 2 activities (Abrahante et al. 2003;Abbott et al. 2005).lin-41 controls events after lin-28 and hbl-1, but it is unclear what stage lin-41 acts on, whether L3 or L4 (Slack et al. 2000;Vadla et al. 2012).
To induce rapid degradation of these proteins without temperature-sensitive alleles, we used the auxin-inducible degron (AID) system adapted from Arabidopsis thaliana (Zhang et al. 2015;Hills-Muckey et al. 2022).We constructed alleles of each of https://doi.org/10.1093/g3journal/jkae130Advance Access Publication Date: 12 June 2024 Investigation these genes with the degron coding region fused to either the 5′-or 3′-end of a gene's open reading frame using CRISPR/Cas9.All of these genes were wild-type in phenotype in the absence of auxin analog when we used a modified TIR (TIR1(F79G); Hills-Muckey et al. 2022).This approach allowed us to approximate the times of action of these heterochronic genes and their temporal relationships to each other and the events they control.The precision of this method is limited by the fact that the kinetics of the degron system may vary for different proteins (Zhang et al. 2015;Hills-Muckey et al. 2022).Nevertheless, we have been able to roughly place the heterochronic gene activities on the timeline of postembryonic development as an essential part of explaining the developmental timing mechanism of C. elegans.

Strains and culture conditions
Nematodes were grown at 20°C on standard NGM plates seeded with Escherichia coli AMA1004 unless otherwise indicated.Males for analysis of male tail seam cells were produced by growing strains on bacteria containing the RNAi plasmid pLT651 (Timmons et al. 2014).

Synchronization at hatching
To generate developmentally synchronized populations of animals, adults filled with eggs were washed from crowded plates into 15 ml tubes, spun down, and the excess liquid then was removed leaving a worm pellet.A total of 500-1000 µl of household bleach solution was added to the tube and vortexed every 2 minutes until cuticles of animals were dissolved enough to release eggs.Tubes then were filled with sterilized distilled water to slow the bleach activity and centrifuged at 400 g for 3 minutes, then the supernatant was discarded and eggs were washed twice with sterilized distilled water and then twice with M9.Then eggs were transferred into M9 and left on a shaker for 20-48 hours at room temperature.Newly hatched larvae will not initiate L1 development without a food source.Then the liquid with larvae was centrifuged to concentrate larvae and they were transferred to plates with food, either with or without 5-Ph-IAA.The 0 time point (0 hours of postembryonic development) was when the animals were placed on food.
We inserted minimal degron sequence (Morawska and Ulrich 2013;Zhang et al. 2015) at the 3′ end of the open reading frame in lin-14, lin-28, and hbl-1 and at the 5′ end of the open reading frame in lin-41 (Supplementary Table 2).
To make insertions, a hybrid dsDNA repair template was used as described in Dokshin et al. (2018).Repair templates were melted and cooled before injections (Ghanta and Mello 2020).The repair template then was added to the CRISPR mix to the final concentration of 100-500 ng/µl of DNA.
Progeny that had roller phenotypes was isolated and then, plates were screened for AID insertion.AID alleles were outcrossed 3×.

AID system
We used a modified auxin-inducible system with TIR1(F79G), using 5-Ph-IAA as the auxin analog (Zhang et al. 2015;Hills-Muckey et al. 2022).To make 5-Ph-IAA-containing plates, 50 µmol of 5-Ph-IAA that was spread on standard NGM plates to approximately 0.005 µM concentration in the agar prior to seeding.
Synchronized animals were placed either onto NGM plates without 5-Ph-IAA, then transferred to 5-Ph-IAA at the indicated time point, or onto plates with 5-Ph-IAA, then transferred to NGM plates without 5-Ph-IAA at the indicated time point.
Because it was not practical to cover all time points in a single experiment, experiments were performed for a subset of all the time points required to define auxin-sensitive periods.Supplementary Table 1 lists each experiment and the strains and time points covered in each, which are also indicated in the figures.

The times of actions for lin-14 determined by AID match those determined by temperature-sensitive alleles
To test the usefulness of the AID system for determining the times of action of the heterochronic genes, we analyzed the phenotypes of lin-14::AID animals transferred to and from the auxin analog 5-Ph-IAA during larval development.Shifting animals onto 5-Ph-IAA should be equivalent to shifting temperature-sensitive (ts) alleles to the nonpermissive temperature.
We first studied lin-14's control of L2 events (the lin-14b activity) by assessing seam cell number at adulthood of animals shifted at different times.Normally, the total number of seam cells increases during the L2 due to a stage-specific symmetric division in many of these cells (Sulston and Horvitz 1977).If the L2-specific seam cell lineage patterns are skipped, and L3 patterns occur precocious in their place due to the lack of lin-14b activity, adult seam cell numbers will be reduced (Ambros and Horvitz 1987).
We observed that the number of seam cells at adulthood was the lowest when lin-14::AID animals were shifted to 5-Ph-IAA at 4-8 hours after hatching (Fig. 1a).Larvae transferred to 5-Ph-IAA later, at 10 and 12 hours, had an intermediate number of seam cells, and most larvae executed normal seam cell division patterns when they were transferred to 5-Ph-IAA at or after 14 hours of postembryonic development (Fig. 1a).By 12 hours, most animals transferred to 5-Ph-IAA (70%) had fewer than 30% of seam cells expressing normal cell division patterns.In their work, Ambros and Horvitz (1987) had considered animals with 30% or fewer seam cells expressing L3 patterns equivalent to those grown at the permissive temperature.Therefore, we consider the end of lin-14's auxin-sensitive period to be around 12 hours of postembryonic development, which closely corresponds to that defined by Ambros and Horvitz using temperature-sensitive alleles.
Next, we examined the formation of adult alae.Adult alae are cuticle structures that form when seam cells differentiate at adulthood.Defective lin-14b, like other heterochronic mutants that result in precocious development, cause adult alae to form at least 1 stage early (Ambros and Horvitz 1984;Ambros and Horvitz 1987;Ambros 1989).It is not believed that lin-14 directly controls alae formation, rather that the time of alae formation is a consequence of lin-14's activity earlier, specifically on L2 events.lin-14::AID larvae transferred to 5-Ph-IAA at 6 hours had full precocious alae, but when transferred between 10 and 12 hours in development, they displayed partial or no precocious alae, and alae developed at the normal time when they were transferred to 5-Ph-IAA at 14 hours in development or later (Fig. 1b).
For most lin-14::AID animals grown with or without 5-Ph-IAA, L1 seam cell divisions occurred at 4-7 hours, and animals entered lethargus at the end of L1 between 16 and 18 hours of larval development.Thus, the end of the time period requiring lin-14b activity as determined by the AID system (its auxin-sensitive period) was the mid-to late-L1 stage, which matches the temperaturesensitive period of lin-14b defined using temperature-sensitive alleles (Ambros and Horvitz 1984).
lin-14a activity specifies developmental events of the L1, and lack of lin-14a causes L2 events to be executed in the L1 (Ambros and Horvitz 1987).Because lin-14(0) animals lack both lin-14a and lin-14b activities, they execute the L2-specific symmetric division of seam cells only once, precociously in the L1, and thereby end up with a normal adult seam cell count, despite other problems (Ambros and Horvitz 1987).It is important to note that we found that lin-14::AID animals grown continuously on 5-Ph-IAA did not have a normal count but rather had a reduced number of seam cells, indicating that lin-14a activity was not completely reduced to null levels (Fig. 1a).The degron, therefore, did not appear to completely inactivate lin-14, but rather resulted in a partial loss-of-function instead of a null phenotype.Nevertheless, we could still use this condition to characterize lin-14a timing.

Times of actions of heterochronic genes | 3
Shifting lin-14::AID animals away from 5-Ph-IAA is the equivalent of shifting ts-alleles to the permissive temperature.Most lin-14::AID larvae removed from 5-Ph-IAA in the first hour in development had a higher than normal number of seam cells at adulthood but no precocious alae (Fig. 2).The same phenotype was observed by Ambros and Horvitz (1984) as a result of the L2 fates being executed twice, in both the L1 and L2, and interpreted as a loss of lin-14a activity with normal lin-14b activity (lin-14a -b + ).Shifting animals at 2 hours resulted in a nearly wild-type level of seam cells and some precocious alae.These data indicate that the auxin-sensitive period for the lin-14a activity begins at 1 hour or before and that the auxin-sensitive period for the lin-14b begins around 2 hours.
Although we were not able to observe a time frame for the restoration of lin-14a activity, lin-14b activity could be fully restored when worms were removed from 5-Ph-IAA plates at 1 hour after start, and at 2 hours in some worms, whereas Ambros and Horvitz (1987) observed restoration following shifts up until 6-7 hours in development.Thus, the recovery from 5-Ph-IAA is not as rapid as initiating degradation of lin-14 with the auxin analog.But overall, these results showed that the AID system could be used to determine the times of actions for heterochronic genes and that times obtained when animals are moved to the auxin analog are relatively precise.
Intestinal nuclei divisions occur during the L1 stage, and these are also under the control of lin-14.C. elegans L1 animals hatch with 20 intestinal nuclei, but some of these undergo mitosis during the L1 molt to yield 30-34 intestinal nuclei by adulthood.Our own counts of the lin-14::AID strain showed that early L1 larvae had 20 ± 0.5 (n = 14) intestinal nuclei, and animals after L2 seam cell divisions had 30 ± 2.3 (n = 12) intestinal nuclei.Consistent with prior observations of lin-14(0) mutants, we observed that lin-14::AID animals grown continuously on 5-Ph-IAA had only 21.8 ± 2.4 (n = 15) intestinal nuclei at adulthood in contrast to normal 30-34 (Sulston and Horvitz 1977;Ambros and Horvitz 1984).
We determined the auxin-sensitive period for lin-14's control of intestinal nuclear divisions.Intestinal nuclei counts were reduced in lin-14::AID animals transferred to 5-Ph-IAA before 16 hours of postembryonic development.L1 lethargus was observed at 16-18 hours, and the animals in lethargus (without pharyngeal pumping) that were transferred to 5-Ph-IAA at 18 hours developed a normal number of intestinal nuclei (Fig. 3).The number of intestinal nuclei increased in animals transferred to 5-Ph-IAA at 13-14 hours in development as compared to those always grown on 5-Ph-IAA, yet was still less than the wild-type number, indicating that lin-14 activity promoting intestinal nuclei divisions was not completed at that time.Thus, this activity of lin-14 in the intestine has a slightly different time frame compared to its activities promoting seam cell fates.As a result, although the control of intestinal nuclei divisions and the lin-14a activity both occur in the L1 stage, these activities have different time frames and might be independent.

lin-28 acts later than lin-14 to promote L2 developmental events
Having established for lin-14 that the auxin-sensitive periods match the established temperature-sensitive periods, we employed this system to study lin-28 and hbl-1.lin-28 mutants are like lin-14b mutants in that they skip L2-specific events, including the L2-specific symmetric seam cell division that increases the number of seam cells in late stages.(Ambros andHorvitz 1984, 1987;Seggerson et al. 2002).When  lin-28:AID animals are grown continuously on 5-Ph-IAA, seam cell counts are reduced compared to wild-type (Fig. 4a).This reduced count is true for animals shifted onto 5-Ph-IAA from 8 to 14 hours.However, the counts increased near the wild-type level in animals transferred to 5-Ph-IAA from 16 to 20 hours of development (Fig. 4a).In general, wild-type seam cell counts were observed when animals were shifted at 22 hours or later; however, some lin-28::AID animals transferred to 5-Ph-IAA at 22 hours and later had 17 seam cells, the reason is unclear (Figs.4a and 5a).Thus, lin-28's time of action for specifying L2 cell fates as determined by AID was at the end of the L1 stage (Figs.4a and 5a).Note that our experiments directly compared the auxin-sensitive periods of lin-14 and lin-28 by shifting both strains in parallel at the same time points (Supplementary Table 1).We found that this period for lin-28 was significantly later than what we determined for lin-14b activity, the mid-L1 stage, so that these genes do appear to be acting separately, not simultaneously, to specify L2 fates.

lin-28 and hbl-1 act simultaneously to control L2 cell fates, just prior to L2 seam cell divisions
Loss-of-function alleles of hbl-1 cause a phenotype similar to that of lin-28 mutants (Abrahante et al. 2003).We directly compared the auxin-sensitive periods of these 2 genes (Supplementary Table 1).The time frame for the end of hbl-1 activity promoting L2 seam cell fates was almost identical to that of lin-28 (Fig. 5).The majority of hbl-1::AID animals developed a normal number of seam cells when they were moved to 5-Ph-IAA at 24 hours in development or later (Fig. 5b).
The times of action of lin-28 and hbl-1 determined by AID appeared to coincide with the L1 lethargus that precedes the L1 molt.The symmetric seam cell divisions normally occur soon after the L1 molt (Sulston and Horvitz 1977).Considering the time needed for the process of degradation to start working, we conclude that the activities of lin-28 and hbl-1 are required right before the L2 seam cell divisions.lin-28 and hbl-1 act during the L2 to promote later events lin-28's effect on the time of adult alae formation is understood to be a consequence of its control of L3 fates, which is a separate activity from its control of L2 fates (Vadla et al. 2012).For both lin-28 and hbl-1, the auxin-sensitive period for precocious adult alae development was slightly later than that of the L2-specific seam cell increase (Figs. 4-6).Some animals transferred to 5-Ph-IAA at 20, 22, and 24 hours had normal numbers of seam cells and complete or gapped precocious alae (Fig. 6), which indicates that they executed the L2 seam cell fates normally and then skipped a later stage.Thus, like lin-14, both lin-28 and hbl-1 appear to have 2 activities separated in time.
We noted a difference in the appearance of the precocious alae of lin-28::AID and hbl-1::AID strains (Supplementary Fig. 1).In   6. lin-28 and hbl-1 have activities that promote seam cell fates after the L2 stage.a) Precocious alae in lin-28::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.The percent was calculated as the fraction of seam cells that developed alae vs the total number of seam cells for each animal.Bars indicate averages and 95% CIs.b) Precocious alae in hbl-1::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.Alae were scored as "full" when they could be easily followed from head to tail, "weak" when alae did not have clear gaps but also could not be followed completely, and "partial" when gaps were observed.A gap in the graph separates different experiments.Different dot shapes indicate different groups of synchronized animals.Times of actions of heterochronic genes | 5 lin-28::AID animals, precocious alae were strong even when they had gaps, whereas the hbl-1::AID alae were difficult to see because they were less pronounced or thinner.In most cases, it was difficult to estimate an approximate percent expressivity of the precocious alae in these animals.This observation suggests some difference in how lin-28 and hbl-1 affect the differentiation of seam cells.

lin-28 is subject to a feedback loop
In experiments where we shifted lin-28:AID animals from 5-Ph-IAA to plates without the auxin analog, we found that lin-28 activity did not recover.Specifically, L1 larvae that were placed on 5-Ph-IAA after synchronization and removed in 1 hour developed a null-like phenotype with reduced numbers of seam cells and precocious alae (11 ± 0.5 seam cells, 87% with full precocious alae, n = 8).This was surprising given the fact that experiments shifting these animals onto 5-Ph-IAA showed that the auxin-sensitive period for lin-28 extended well past 1 hour of postembryonic development.This was not true for hbl-1, lin-28's partner in controlling the L2 cell fates (see below).This finding suggests a feedback loop: once lin-28 activity is reduced, consequences of its downregulation make it impossible either for lin-28 expression to be restored or for lin-28 to have an effect on downstream events.This phenomenon prevented us from defining a precise start time of lin-28's activity in the L1.

hbl-1 has 2 activities separated in time
In contrast to lin-28, hbl-1's activity was restored when shifted away from the auxin analog 5-Ph-IAA.When synchronized hbl-1::AID animals were grown on 5-Ph-IAA from the start of postembryonic development and then transferred away at 4 hour intervals, both seam cell numbers and adult alae formation were mostly wild-type up until 20 hours, indicating recovery from the degradation over that timeframe (Supplementary Table 1; Fig. 7).It was not until the hbl-1::AID animals were on 5-Ph-IAA for 24 hours or longer did precocious phenotypes occur (Fig. 7).
To define the gene's auxin-sensitive period for determining L2 seam cell fates, hbl-1::AID animals were either transferred to 5-Ph-IAA or removed from it in 2-hour intervals, after which both seam cell counts and precocious alae formation were assessed (Supplementary Table 1; Fig. 8).Experiments where animals were shifted to 5-Ph-IAA showed that the auxin-sensitive period is over by about 22 hours of postembryonic development, as seam cell counts were mostly wild-type from that point on (Fig. 8a).In experiments where hbl-1::AID animals were moved off the auxin analog, seam cells remained normal in animals removed from 5-Ph-IAA before 18 hours of larval development and showed a strong loss-of-function phenotype in those removed at 24 hours or later (Fig. 8b).These results define hbl-1's auxinsensitive period to between 18 and 22 hours of postembryonic development.We observed L2 seam cell divisions to occur between 21 and 24 hours in hbl-1::AID animals grown either without or continuously on 5-Ph-IAA.Therefore, hbl-1's time of action appears to occur just before or during the L2 seam cell divisions.
Because lin-28's effect on the L2-specific division pattern and seam cell differentiation at adulthood reflect 2 different activities of the same gene, we wished to assess whether the same is true for hbl-1 (Vadla, et al. 2012).We observed that full precocious alae appeared in hbl-1::AID animals that had been transferred to 5-Ph-IAA before 24 hours in development.The development of precocious alae became less frequent in later transfers, until they disappeared altogether in animals transferred at 30 hours in development (Fig. 9a).Similarly, in animals removed from 5-Ph-IAA at 24-26 hours in development, precocious alae sometimes appeared (Fig. 9b).Since most animals did not develop precocious alae when they were moved to 5-Ph-IAA at 28 hours or later, and most animals formed precocious alae if they were removed from 5-Ph-IAA at the same time, this must be shortly after the time that hbl-1 acts to control later events.Thus, hbl-1's activity promoting seam cell fates after the L2 (24-26 hours) is separated in time from its earlier activity controlling seam cell divisions (18-22 hours).This second activity ceases in 4-6 hours after the L2 divisions are completed, which is approximately 4-6 hours before the L2 molt.

hbl-1 deficiency causes skipping of L3 developmental events
Hermaphrodite seam cell lineages show the same division patterns in L3 and L4 stages, so it is not possible to distinguish skipping of L3 vs L4 cell fates in these lineages.However, certain male seam cell lineages, those that give rise to the male tail rays, do have pattern differences between these stages (Sulston and Horvitz 1977).Like the seam cells in the L2, 5 male seam cells undergo additional symmetric division in the L3 that will ultimately produce 9 ray precursor cells that will form the rays of the male tail (Sulston and Horvitz 1977).We predicted that if the L3-specific fates were skipped there would be a reduced number of ray precursor cells; if the L4 fates were skipped, the number of ray precursor cells would not be reduced.
First, we grew hbl-1::AID animals continuously on 5-Ph-IAA and examined male tails between 32 and 50 hours when ray cell divisions occur.We observed that the number of ray precursor cells was significantly reduced and ray differentiation occurred precociously: at 32 hours, ray cells already completed L4-specific divisions, while in animals grown on plates without auxin, they were still undergoing L3-specific divisions (n = 15, Fig. 10).Note that in addition to occurring in the L2, symmetric seam cell divisions also occur in the L3 in ray-producing seam cells V6 and V7 (Sulston and Horvitz 1977).So, the reduction of ray precursor cells in hbl-1:: AID animals grown continuously on the auxin analog could be due to loss of these proliferative divisions in both L2 and L3.
Next, hbl-1::AID animals were moved to auxin between 18 and 22 hours of postembryonic development (Fig. 11).We examined both the non-ray-producing seam cells, V1-V4, and the ray-producing seam cells V5, V6, and T. For each time point, 7 animals were observed, all animals had patterns similar to those represented on micrographs in Fig. 11.Animals moved to auxin at either 18 or 20 hours showed precocious differentiation of precursor cells and a reduced number of rays, but had a reduced overall number of seam cells (12-13), meaning that the reduction in rays may have been due, at least in part, to skipping the symmetric division of the L2.In contrast, animals moved to auxin at the slightly later time point of 22 hours of development showed reduced number of rays but overall the wild-type number of seam cells (16) (Fig. 11), meaning that the L2-specific symmetric division was not skipped.We interpret these observations to mean that the reduction in the second of Precocious alae in hbl-1::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.b) Precocious alae in hbl-1::AID animals removed from 5-Ph-IAA at the indicated times after synchronization.Alae were scored in mixed populations of molting L3 to late L4 animals due to variations in the developmental speed between individuals after synchronization.Alae were scored as "full" if they could be easily followed from head to tail, "weak" if they did not have clear gaps but could not be easily followed, and "partial" when gaps were observed.Gaps in graphs separate different experiments.Different dot shapes indicate different groups of synchronized animals.Fig. 10.Reduction of hbl-1 activity causes a skipping of L3 stage and premature cell fates in male tail.a) Fluorescence micrograph of a male tail at 32 hours in a hbl-1::AID animal grown without 5-Ph-IAA.The number of ray precursor cells appears reduced; however, those cells proceeded to undergo further divisions characteristic of these cells.b) Fluorescent micrograph of male tail at 32 hours in a hbl-1::AID animal grown continuously on 5-Ph-IAA.Seam cell nuclei and cell junctions visualized with SCM::GFP and ajm-1::GFP, respectively.Animals are oriented anterior to the left, dorsal side up.

Times of actions of heterochronic genes | 7
hbl-1's 2 activities can cause the skipping of L3-specific divisions to lead to precocious L4 and adult cell fates.

lin-41 acts during the L3 stage
lin-41 has multiple roles in the animal including the control of developmental timing (Slack et al. 2000;Spike et al. 2014).lin-41 null alleles are sterile, whereas loss-of-function alleles cause a precocious phenotype where seam cells differentiate precociously, producing adult alae in the L4, but seam cell counts are normal.Again, it has not been specifically addressed whether lin-41 loss-of-function mutants skip L3 or L4 stage events.
When lin-41::AID animals were grown continuously on 5-Ph-IAA, short precocious alae patches (less than 50% expressivity) were observed in 38% of the animals (Fig. 12a).Animals transferred to 5-Ph-IAA at 2-hour intervals from 16 to 36 hours displayed precocious alae patches (Fig. 12a).No precocious alae was observed when animals were transferred at 38 hours or later, indicating that lin-41's auxin-sensitive period ends between 36 and 38 hours of postembryonic development (Fig. 12a).L3 seam cell divisions were observed at 36 hours in development and the L3 molt around 40 hours, so, auxin sensitivity for lin-41 ended during the L3 stage.
Interestingly, the fraction of animals with precocious alae patches was significantly higher in animals transferred at 16 and 18 hours than later (P < 0.05 comparing the 16/18 hours group with the 20-40 hours group, by Fisher's exact test).We have no explanation for this effect, but it suggests a somewhat increased sensitivity of the heterochronic pathway to lin-41 reduction during early postembryonic stages, possibly similar to lin-28's feedback loop mentioned above.
lin-41::AID animals moved away from 5-Ph-IAA from 32 hours and later showed percentages of precocious alae patches comparable to animals grown continuously on the auxin analog (Fig. 12b).Surprisingly, removing lin-41::AID animals from the auxin analog at 12, 20, and 28 hours resulted in a small amount of precocious alae, again indicating a lasting effect of lin-41 reduction during the L1 and L2 stages.Otherwise, these data suggest that lin-41 acts primarily during the L3 stage.

lin-41 activity deficiency likely causes skipping of L4 developmental events
We observed that male lin-41::AID animals grown on auxin had a precocious tail tip retraction and disrupted morphogenesis of rays and fan, as has been previously described for loss-of-function mutants (Del Rio-Albrechtsen et al. 2006).However, it appears that most ray precursor cells were produced, indicating that the proliferative cell divisions that produce them occurred normally (n = 10, Fig. 13, Sulston and Horvitz 1977).So, in contrast with hbl-1, reduction of lin-41 using the AID system did not cause skipping of L3 events in most male tail cells.We therefore interpret the precocious differentiation of the hypodermal cells in both male and hermaphrodite as reflecting the skipping of L4-specific cell fates when lin-41 activity is reduced.

Discussion
Previously, time-of-action data existed for only 1 heterochronic gene, lin-14, which was obtained using temperature-sensitive alleles (Ambros and Horvitz 1987).The times of action of other core heterochronic genes were only suggested by indirect evidence: the earliest developmental events they affected and their expression patterns.lin-28, for example, is expressed from the start of postembryonic development through the L2 and L3, but the first developmental event it controls is at the beginning of the L2, about the time it begins to be downregulated (Ambros and Horvitz 1984;Moss et al. 1997).In order to better understand the mechanistic relationships among the heterochronic genes, it is important to know more precisely when their activities are needed to specify stage-specific cell fates.
We found that 2 other heterochronic genes, lin-28 and hbl-1, each have 2 activities that, like lin-14's, are separated in time (Fig. 14).Relative to events they control, both lin-28 and hbl-1 appear to be required just prior to or concurrent with the L2-specific symmetric seam cell division which they control, and relative to each other, lin-28 and hbl-1 act essentially simultaneously.By contrast, lin-14's second activity (lin-14b) ends well before those of lin-28 and hbl-1 despite the fact that they control the same developmental events.Developmental timing regulators may act directly on the events they control, or they may act in advance of those events, perhaps by setting up a regulatory condition or acting through a series of molecular events that causes a delay.We know that lin-14 controls L2 cell fates by transcriptionally repressing the expression of microRNA genes that repress lin-28 and hbl-1 (Tsialikas et al. 2017).Possibly, this regulation is responsible for the separation in the genes' times of action: given that downregulation of lin-14 would cause transcriptional activation of microRNA genes, it may take time for that transcriptional activation to lead to microRNA-mediated repression of lin-28 and hbl-1 in time to have an effect on L2 fates.
Relative to the L3 events they control, lin-28 and hbl-1 are more like lin-14b in they act well in advance of those events: from the early-to mid-L2 to control events that happen in the early L3 (Fig. 14).Relative to each other, these genes appear staggered in time, with lin-28 acting just prior to hbl-1.Genetic analysis suggests indeed that lin-28 acts indirectly through hbl-1 to control cell fates (Vadla et al 2012;Ilbay and Ambros 2019;Ilbay et al. 2021).Thus, whereas the first activities of lin-14, lin-28, and hbl-1  Times of actions of heterochronic genes | 9 act just prior to or even concurrent with the events, the second of each gene's 2 activities that acts well in advance of the events it controls (Fig. 14; Ambros and Horvitz 1987).
We identified a single activity of lin-41 that acts in the early-to mid-L3 to control events of the L4 (Fig. 14).It is possible that yet unidentified series of molecular events intervenes between lin-41's time of action and the L4 events it controls.Because null alleles of lin-41 are sterile, we know that our lin-41::AID allele did not eliminate lin-41 activity on 5-Ph-IAA because these animals were fertile when grown on the auxin analog (data not shown).Because we did not reduce lin-41 activity, we cannot not be confident we have detected all of its role in the heterochronic pathway.Specifically it is theoretically possible that lin-41, like the other genes, has 2 activities, one of which regulates the L3, as previously was speculated (Vadla et al. 2012).Therefore, it may be due to the inability of our lin-41::AID allele which did not fully reduce lin-41 activity that we may not have detected any activity regulating L3 events that lin-41 might have.
There are other limitations of the AID system for precise timing of these genes.First, we know that the degron system behaves differently with different proteins (Hills-Muckey et al. 2022).Differences in auxin-sensitive periods may be true differences or due to differences in the degradation and activity kinetics of these factors (For instance, the HBL-1 protein may take longer to degrade than LIN-28).So, a truly precise comparison between the times of activities of the heterochronic genes is not possible.Another limitation, which is well known, is that the degradation and reactivation occur at best on the order of an hour or 2, even when they occur robustly (Zhang et al. 2015;Ashley et al. 2021).This means that estimations of start and end times of the heterochronic genes' periods of action cannot be precise within hours.
We unexpectedly uncovered a phenomenon that may reflect a feature of heterochronic gene hierarchy.Although restoration of activity usually occurs when auxin is removed, we found that reduction of lin-28 activity was irreversible, and consequently, were unable to define the start of its auxin-sensitive period.Remarkably, a null phenotype occurred even when removal from auxin was 18 hours or more from the auxin-sensitive period, theoretically enough time for the protein to recover (data not shown; Zhang et al. 2015).We know that lin-28 transcription is continuous through larval development, even into late stages, and that its temporal regulation is due to microRNAs (Seggerson et al. 2002).The lin-28 gene has the interesting property of being both a negative regulator and a target of let-7-family microRNAs (Vadla et al. 2012;Tsialikas et al. 2017).It is possible that the auxin-induced downregulation of lin-28 leads to an upregulation of let-7, which may contribute to the irreversible repression of lin-28 expression.Nevertheless, we currently lack a complete explanation for this phenomenon, suggesting that there is more to learn about lin-28's regulation.
Interestingly, we also observed a similar phenomenon for lin-41.We found that lin-41 activity was required mostly between 32 and 38 hours of development to control when adult alae appeared.But the precocious alae phenotype is more penetrant in animals grown continuously on 5-Ph-IAA or moved to the auxin analog before 20 hours, suggesting that lin-41 might have some activity during the L1 and L2 stages.It is possible that lin-41, like lin-28, inhibits a negative regulator of itself, which, when derepressed prior to lin-41's auxin-sensitive period, can lead to further reduction in lin-41 activity.
Our study has attempted to place the heterochronic gene activities into a timeline of postembryonic development relative to one another and to the developmental events whose timing they control.Further studies would be needed to confirm the timeline of heterochronic gene activities, and further mechanistic understanding will be needed to reveal how these genes control these specific events and how they influence the times of action of each other in the heterochronic gene hierarchy.

Fig. 1 .
Fig. 1. lin-14 acts during the mid-L1 stage to promote L2 seam cell fates.a) Numbers of seam cells in lin-14::AID animals moved to 5-Ph-IAA at the indicated times after L1 synchronization.Seam cells were scored in mixed populations of late L3 and early L4 animals due to variations in the developmental speed between individuals after synchronization.Different datapoint shapes indicate different groups of synchronized animals.Horizontal dotted line indicates the wild-type number of seam cells, 16. "C" indicates animals grown continuously on 5-Ph-IAA.Bars indicate averages and 95% CI. b) Precocious adult alae of early L4 stage animals from (a).

Fig. 2 .
Fig. 2. lin-14 activity controlling L1 seam cell fates occurs prior to 2 hours of postembryonic development.a) Numbers of seam cells in lin-14::AID animals removed from 5-Ph-IAA at the indicated times after synchronization.Seam cells were scored in mixed populations of late L3 to late L4 animals due to variations in the developmental speed between individuals after synchronization.Horizontal dotted line indicates the exact wild-type number of seam cells.Bars indicate averages and 95% CIs.b) Precocious adult alae of L4 stage animals from (a).
Fig. 3. lin-14 activity promotes intestinal nuclei divisions prior to 16 hours of postembryonic development.Intestinal nuclei numbers in lin-14::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.The horizontal dotted line indicates the average number of intestinal nuclei in wild-type animals.A(−) indicates animals grown continuously without 5-Ph-IAA.A(+) indicates animals grown continuously on 5-Ph-IAA.Bars indicate averages and 95% CIs.Different datapoint shapes indicate different groups of synchronized animals needed to cover all the time points.

Fig. 4 .
Fig. 4. The activity of lin-28 that promotes L2 seam cell fates occurs later than the similar activity of lin-14.a) Numbers of seam cells in lin-28::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.Seam cells were scored in mixed populations of late L3 to late L4 animals due to variations in the developmental speed between individuals after synchronization.The horizontal dotted line indicates the wild-type number of seam cells."C" indicates animals grown continuously on 5-Ph-IAA.Bars indicate averages and 95% CIs.b) Precocious adult alae of L4 stage animals from (a).Different dot shapes indicate different groups of synchronized animals.

Fig. 5 .
Fig. 5.The activities of lin-28 and hbl-1 that promote L2 seam cell fates coincide in time.a) Numbers of seam cells in lin-28::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.Seam cells were scored in mixed populations of late L4 and young adult animals due to variations in the developmental speed between individuals after synchronization.b) Numbers of seam cells in hbl-1::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.Seam cells were scored in mixed populations of late L3 to late L4 animals.Horizontal dotted lines indicate the wild-type number of seam cells."C" indicates animals grown continuously on 5-Ph-IAA.Bars indicate averages and 95% CIs.Different dot shapes indicate different groups of synchronized animals.
Fig.6.lin-28 and hbl-1 have activities that promote seam cell fates after the L2 stage.a) Precocious alae in lin-28::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.The percent was calculated as the fraction of seam cells that developed alae vs the total number of seam cells for each animal.Bars indicate averages and 95% CIs.b) Precocious alae in hbl-1::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.Alae were scored as "full" when they could be easily followed from head to tail, "weak" when alae did not have clear gaps but also could not be followed completely, and "partial" when gaps were observed.A gap in the graph separates different experiments.Different dot shapes indicate different groups of synchronized animals.

Fig. 7 .Fig. 8 .
Fig. 7. hbl-1 activity is restored after removal from 5-Ph-IAA.a) Numbers of seam cells in hbl-1::AID animals removed from 5-Ph-IAA at the indicated times after synchronization.Seam cells were scored in mixed populations of late L3 to late L4 animals due to variations in the developmental speed between individuals after synchronization.Horizontal dotted lines indicate the wild-type number of seam cells.Bars indicate averages and 95% CIs.b) Precocious adult alae of L4 stage animals from (a).Different dot shapes indicate different groups of synchronized animals.

Fig. 9 .
Fig. 9.The time of the second hbl-1 activity is distinct from the first.a)Precocious alae in hbl-1::AID animals moved to 5-Ph-IAA at the indicated times after synchronization.b) Precocious alae in hbl-1::AID animals removed from 5-Ph-IAA at the indicated times after synchronization.Alae were scored in mixed populations of molting L3 to late L4 animals due to variations in the developmental speed between individuals after synchronization.Alae were scored as "full" if they could be easily followed from head to tail, "weak" if they did not have clear gaps but could not be easily followed, and "partial" when gaps were observed.Gaps in graphs separate different experiments.Different dot shapes indicate different groups of synchronized animals.

Fig. 11 .
Fig.11.HBL-1 degradation at certain times in development causes a skipping of L3 stage and premature cell fates in male tail.a) hbl-1::AID grown without 5-Ph-IAA.b) Tail of a male moved to auxin at 18 hours, this individual also had 12 total seam cells.c) Tail of a male moved to auxin at 20 hours, this individual also had 13 total seam cells.d) Tail of a male moved to auxin at 22 hours, this individual also had 16 total seam cells.Seam cell nuclei and cell junctions visualized with SCM::GFP and ajm-1:: GFP, respectively.Arrowheads indicate ray precursor lineages.Animals are oriented anterior to the left, dorsal side up.

Fig. 13 .
Fig. 13.lin-41 activity reduction causes premature tail tip retraction but no change in the number of ray precursor cells.a) DIC micrograph of lin-41::AID male grown on 5-Ph-IAA observed at 40 hours; b) fluorescent microscopy of the same tail showing seam cell nuclei and cell junctions visualized with SCM::GFP and ajm-1::GFP, respectively.The number of ray precursor cells appears normal(Sulston and Horvitz 1977).The animal is oriented anterior to the left, dorsal side up.

Fig. 14 .
Fig. 14.Times of actions for heterochronic genes in C. elegans development.The lethargus periods preceding and including the molts are indicated by gray gradients.Vertical boxes indicate approximate periods of stage-specific seam cell divisions, and bars indicate the genes' auxin-sensitive periods affecting stage-specific cell fates: lin-14's upper bars: L1 and L2 fates, lower bar: intestinal nuclear divisions, lin-28's and hbl-1's bars: L2 and L3 fates, lin-41's bar: L4 fates.The bars indicate the periods when most animals showed a strong loss-of-function phenotype.The gradient in the bars is to indicate that we did not determine precisely the start time of the auxin-sensitive period.