Control of clathrin-mediated endocytosis by NIMA family kinases

Endocytosis, the process by which cells internalize plasma membrane and associated cargo, is regulated extensively by posttranslational modifications. Previous studies suggested the potential involvement of scores of protein kinases in endocytic control, of which only a few have been validated in vivo. Here we show that the conserved NIMA-related kinases NEKL-2/NEK8/9 and NEKL-3/NEK6/7 (the NEKLs) control clathrin-mediated endocytosis in C. elegans. Loss of NEKL-2 or NEKL-3 activities leads to penetrant larval molting defects and to the abnormal localization of trafficking markers in arrested larvae. Using an auxin-based degron system, we also find that depletion of NEKLs in adult-stage C. elegans leads to gross clathrin mislocalization and to a dramatic reduction in clathrin mobility at the apical membrane. Using a non-biased genetic screen to identify suppressors of nekl molting defects, we identified several components and regulators of AP2, the major clathrin adapter complex acting at the plasma membrane. Strikingly, reduced AP2 activity rescues both nekl mutant molting defects as well as associated trafficking phenotypes, whereas increased levels of active AP2 exacerbate nekl defects. Moreover, in a unique example of mutual suppression, NEKL inhibition alleviates defects associated with reduced AP2 activity, attesting to the tight link between NEKL and AP2 functions. We also show that NEKLs are required for the clustering and internalization of membrane cargo required for molting. Notably, we find that human NEKs can rescue molting and trafficking defects in nekl mutant worms, suggesting that the control of intracellular trafficking is an evolutionarily conserved function of NEK family kinases.


Introduction
The cuticle of C. elegans is a flexible apical extracellular matrix consisting of cross-linked collagens, non-collagenous proteins, linked carbohydrates, and lipids [1,2]. The cuticle is essential for providing a protective barrier from the environment, for maintaining the proper shape and integrity of the organism, and for facilitating muscle-based locomotion by functioning as an exoskeleton [3]. Remodeling of the cuticle occurs at the end of each of four larval stages (L1-L4) through a process called molting. During molting, a new cuticle is synthesized under the old cuticle, which is then shed [3][4][5]. Molting enables organismal growth and allows for changes in the composition and organization of the cuticle at different life stages. Molting defects can occur when synthesis of the new cuticle is compromised or when shedding of the old cuticle is incomplete. A sizeable number of factors have been implicated in C. elegans molting including proteins involved in cuticle structure, protein modification, protein degradation, cell signaling, transcription, and intracellular trafficking [4,6].
During each molt, an accumulation of ribosomes, Golgi bodies and RNA is observed within the epidermal cells that produce the new cuticle, consistent with increased protein synthesis [1,2,5,7]. The secretion of essential structural proteins, along with the enzymatic activities required for cuticle replacement, is accomplished through exocytosis. Consistent with this, inhibition of sec-23, which encodes a component of COPII-coated vesicles required for the transport of proteins from the endoplasmic reticulum (ER) to the Golgi, leads to molting defects [8]. At the same time, endocytosis is required to balance exocytosis and thus maintain a relatively constant volume/area of apical plasma membrane. In addition, the recycling of old cuticle components may be enabled through the process of endocytosis.
We previously showed that NEKL-MLTs are expressed in punctate patterns in the major syncytial epidermis of the worm, hyp7, suggesting that NEKL-MLTs localize to one or more trafficking compartments [20,21]. In addition, we found that molting-defective nekl-mlt larvae exhibit abnormal morphology and/or localization of multiple trafficking markers in the apical region of hyp7, including a multi-copy clathrin heavy chain reporter [20,21]. However, it was unclear as to whether the apparent defects in clathrin-mediated endocytosis were a primary cause of the observed molting defects in nekl-mlt mutants or a secondary consequence of physiological effects caused by the presence of a double cuticle and the inability of the worms to feed.
Here we demonstrate that NEKL-MLTs regulate clathrin-mediated endocytosis within the context of an intact developing organism. We show that the function of NEKL-MLTs in trafficking is highly sensitive to the balance between the open and closed AP2 conformations and that AP2-associated phenotypes are also responsive to NEKL activity. In addition, we demonstrate that loss of NEKL functions leads to defects in LRP-1/LRP2 endocytosis, a cargo that is physiologically relevant to molting. Our combined findings indicate that defects in endocytosis are likely to be a major underlying basis for the observed molting defects in nekl mutants. Finally, we show that mammalian NEK6 and NEK7 can partially rescue endocytosis and molting defects in nekl-3 mutant worms.
Among nekl-2; nekl-3 animals containing the fd155 suppressor mutation, 98% progressed to adulthood versus only 1-2% for the parental nekl-2; nekl-3 strain (Fig 1B and 1F). We showed the fd155 causal mutation to be a T-to-A transversion in the 11 th exon of dpy-23, which leads to  3(gk894345) worms. The rescuing extrachromosomal array (fdEx286) expresses wild-type nekl-3 and the SUR-5::GFP reporter. White arrow indicates a growth-arrested nekl-2; nekl-3 larva that failed to inherit the array and exhibits a "corset" morphology, characteristic of nekl-mlt molting defects (inset in A). The bracket marks the constricted mid-body region, which contains a double cuticle. Bar size in A = 100 μm (for A and B); in inset, 20 μm. (B) DIC image of nekl-2; nekl-3 adult worm containing the suppressor mutation dpy-23(fd155). (C) Graphic representation of AP2 tetramer complex containing four subunits. (D) Gene structure diagram of dpy-23/μ including the locations of mutations. Point mutations (fd155, fd277, fd279) are indicated by arrows; indel (fd261) by the line ending in a small bracket. (E) Protein domain diagram of DPY-23/μ with corresponding allelic changes; fd261 is missing~30 bp of the proximal 5'UTR including a predicted SL1 transplice site and the start codon. The amino acid (aa) locations of the two domains are indicated. (F-I) Bar plots showing suppression of molting defects in nekl-mlt mutants by reduction in AP2 activity. Assays F-H were carried out in nekl-2(fd81); nekl-3(gk894345) double mutants. (G,I) RNAi was carried out in the indicated backgrounds a S434T substitution in a serine that is highly conserved (Fig 1D and 1E). A second independent mutation (fd277) was identified as a T to G transversion in the 8 th exon of dpy-23, leading to a Y275D substitution in a highly conserved tyrosine residue (Fig 1D and 1E). Both mutations, as well as a CRISPR/Cas9-generated predicted null allele of dpy-23 (fd279) (S1 Table), led to a high percentage of viable adults in the nekl-2; nekl-3 background (Fig 1D-1F), and partial suppression was observed in nekl-2; nekl-3 strains treated with dpy-23(RNAi) (Fig 1G).
Failure to observe suppression of nekl-mlt null alleles using RNAi to knockdown AP2 may be due to incomplete inactivation of the targeted subunit. We therefore determined if null alleles in apa-2/α or dpy-23/μ could suppress larval arrest in null nekl-mlt backgrounds. In the case of apa-2/α, we failed to observe any suppression in null nekl-2, nekl-3, and mlt-3 backgrounds (n > 1000 for each strain), indicating that complete loss of the α/σ hemicomplex is not sufficient to overcome the requirement for NEKL-2 and NEKL-3. Attempts to score suppression of nekl-mlt null alleles with dpy-23(e840), however, were not successful because the generated compound mutants were very sick and slow growing. using injection methods; control indicates non-injected siblings. (F-I) Error bars indicate 95% confidence intervals; pvalues were determined using Fischer's exact test where proportions were compared to the wild-type allele (F,H) or to the RNAi control (G,I): ���� p < 0.0001, ��� p < 0.001. (J) Guide to nekl-mlt alleles used in this study. Raw data are available in S1 File. https://doi.org/10.1371/journal.pgen.1008633.g001
FCHO-1 is a member of the muniscin protein family, members of which have important roles in clathrin-mediated endocytosis and are recruited to nascent pits early during endocytic vesicle formation [27,32,52,53]. Orthologs of FCHO-1 contain three characterized functional domains: an N-terminal F-BAR domain, a C-terminal Mu-homology domain, and an internal AP2 Activator domain ( Fig 1D). The F-BAR domain binds to lipid bilayers and aids in membrane curvature, whereas the Mu-homology domain binds to cargo and other endocytic adapter proteins. The AP2 Activator domain has recently been shown to facilitate the allosteric opening/activation of AP2 (Fig 2B), and loss of fcho-1 activity in C. elegans leads to a decrease in the levels of open/active AP2 [39][40][41].
The above findings imply that decreasing the amount of open/active AP2, either by reducing gross AP2 levels or by inhibiting its allosteric activation, can bypass the requirement for NEKL-MLT activity. As a further test of this model, we used CRISPR to generate an in-frame fcho-1 deletion (fd262), which is predicted to specifically remove the AP2 activator domain of FCHO-1 without affecting the F-BAR or Mu-Homology domains (Fig 2C and 2D) [40,41]. Importantly, fcho-1(fd262) strongly suppressed molting defects in nekl-2; nekl-3 mutants ( Fig  2E), consistent with suppression by fcho-1 occurring through a reduction in AP2 activity.

Excess open AP2 enhances nekl-mlt molting defects
Given that decreasing the amount of open AP2 led to the suppression of nekl-mlt defects, we next tested if increasing the levels of open AP2 could enhance molting defects in weak nekl-mlt loss-of-function backgrounds. ncap-1 encodes the C. elegans ortholog of the mammalian adap-tiN Ear-binding Coat-Associated Proteins (NECAP1 and NECAP2), which function in several different aspects of clathrin-mediated endocytosis [42,[54][55][56][57]. Recently, C. elegans NCAP-1 was shown to allosterically regulate AP2 to promote the closed/inactive conformation, and loss of ncap-1 suppresses the slow-growth phenotype of fcho-1 strains [42]. Thus, NCAP-1 acts in opposition to FCHO-1 to regulate AP2 conformation and activity ( Fig 3A) [39].
To test for enhancement of molting defects, we carried out RNAi of nekl-mlts in wild-type and ncap-1(mew39) deletion backgrounds using "weak" RNAi feeding methods, which cause only a partial reduction in nekl-mlt activities [20,21]. Whereas RNAi feeding of nekl-mlts in wild type resulted in little or no defective molting, significantly elevated levels of molting arrest were detected in the ncap-1(mew39) background ( Fig 3B). These findings imply that increased levels of open AP2 exacerbate nekl-mlt loss-of-function phenotypes. Correspondingly, RNAi (E) Assays were carried out in nekl-2(fd81); nekl-3(gk894345) double mutants using the indicated fcho-1 alleles. (F) RNAi was carried out in the indicated backgrounds using injection methods; control indicates non-injected siblings. (G) Assays were carried out in strong/null nekl-mlt backgrounds using the null fcho-1(ox477) allele. (E-G) Error bars indicate 95% confidence intervals. p-Values were determined using Fischer's exact test where proportions were compared to the wild-type allele (E,G) or to the RNAi control (F); ���� p < 0.0001, ��� p < 0.001. Raw data are available in S1 File. of ncap-1 mitigated suppression conferred by loss of fcho-1 in nekl-2; nekl-3 mutants (Fig 3C). These results are consistent with the reported opposing functions for NCAP-1 and FCHO-1 and support the model that suppression of nekl-mlts by fcho-1 and AP2 mutations is due to reduced levels of open AP2.
Because mammalian NECAPs have been suggested to have several distinct functions during endocytosis [54][55][56]58], we carried out additional tests to determine if increased open AP2 could enhance nekl-mlt defects. For these studies, we made use of several missense mutations in dpy-23/μ that shift the balance of AP2 toward the open state ("open mutants") [40]. As with ncap-1, we observed strong enhancement of molting defects after nekl-mlt RNAi feeding in dpy-23/μ strains containing E306K or T160A substitutions (Fig 3D-3E). The somewhat stronger findings observed for the E306K mutation correlates with biochemical assays showing that E306K leads to higher levels of open AP2 than the T160A substitution [40]. Collectively, our findings demonstrate that reducing the level of open AP2 suppresses nekl-mlt molting defects whereas increasing the level of open AP2 exacerbates defects.

Loss of NEKL-MLT activity suppresses AP2-associated defects
Loss of function of non-essential AP2 subunits (μ, α, and σ) and fcho-1 leads to reduced growth rates and the accumulation of fluid between the cuticle and epidermis [40,49]. This latter defect visibly manifests in adult-stage worms as bilateral bulges located near the junction of the hyp6 and hyp7 epidermal syncytia in the region of the neck (the "Jowls" phenotype; Fig 4A). Strikingly, we observed that loss of function in NEKL-MLT activity significantly suppressed the Jowls phenotype of fcho-1 and AP2 mutants (Fig 4A-4F). Specifically, partial suppression of Jowls was observed in the fcho-1(ox477) null mutant background, as well as in the presence of null alleles of apa-2/α and dpy-23/μ, and in aps-2(RNAi) animals ( Fig 4C-4F). Moreover, suppression was observed in strains with reduced or abolished function in either the NEKL-2 or NEKL-3 pathways (Fig 4C and 4F). Our observation that simultaneous loss of AP2 and NEKL-MLT activities leads to the mutual suppression of both molting-defective and Jowls phenotypes underscores the tight functional connection between NEKL-MLT and AP2 activities in vivo.

Molting-defective nekl mutants exhibit changes in epidermal clathrin that are suppressed by inhibition of fcho-1 and AP2
Given the strong genetic data linking the NEKL-MLTs to AP2, we next examined the role of NEKL-MLTs in clathrin-mediated endocytosis. We previously showed that molting-defective nekl-mlt mutants exhibit abnormal localization of a multi-copy clathrin heavy chain reporter [20,21,59]. To visualize clathrin at physiological levels, we used CRISPR/Cas9 to generate a clathrin heavy chain reporter construct with an N-terminal GFP tag (GFP::CHC-1), which showed the expected punctate localization pattern in the hyp7 epidermal syncytium ( Fig 5A). Consistent with our previous findings, GFP::CHC-1 localization was altered in molting-defective nekl-2; nekl-3 larvae (Fig 5A and 5B). Specifically, the average mean intensity of GFP:: CHC-1 was increased by 2.2-fold in the apical region of hyp7 in nekl-2; nekl-3 larvae arrested at the L2/L3 molt relative to wild-type late-stage L2 larvae ( Fig 5B). In addition, the percentage of GFP-positive pixels (above a uniformly applied threshold) was increased by 1.3-fold in nekl-2; nekl-3 molting-defective larvae relative to wild type, which may reflect an increase in the density and/or size of apical GFP::CHC-1 puncta (Fig 5A-5C).

Loss of NEKLs in adults leads to changes in clathrin that are independent of molting defects
Although our above findings suggest that the NEKLs regulate epidermal clathrin, the analysis of trafficking in nekl-mlt mutants is problematic because of potential secondary effects caused by the double cuticle and larval arrest phenotype. To circumvent this obstacle, we engineered regulatable NEKL kinases using the auxin-induced degradation system [60,61]. Importantly, by depleting NEKLs after the final molt, we eliminated the possibility that changes in clathrin larvae expressing CRISPR-tagged GFP::CHC-1 within the apical region of the hyp7 epidermal syncytium. GFP::CHC-1 localization is shown for wild-type, nekl-2(fd81); nekl-3(gk894345), nekl-2(fd81); fcho-1(fd131); nekl-3(gk894345), and nekl-2(fd81); dpy-23(fd155) nekl-3(gk894345) strains. Inverted fluorescence images are shown to aid clarity. Bar size in A = 5 μm (for A-D). Background subtraction was performed using the same parameters for all images; minimum and maximum pixel values were kept consistent for all images. (B,C) For individual larvae of the indicated genotypes, the mean GFP::CHC-1 intensity (B) and the percentage of GFP-positive pixels above threshold (C) were determined in the apical region of hyp7. Both the group mean and 95% confidence interval (error bars) are shown. p-Values for compared means were determined using two-tailed Mann-Whitney tests; ���� p < 0.0001, �� p < 0.01. Raw data are available in S1 File.
https://doi.org/10.1371/journal.pgen.1008633.g005 localization could be an indirect consequence of defective molting. Proteins tagged with an auxin-inducible degron (AID) are responsive to auxin, which binds to the AID motif leading to ubiquitination by the TIR1-SCF E3-ligase complex and degradation by the proteosome ( Fig  6A).
Exposure of CRISPR-tagged NEKL-2::AID and NEKL-3::AID day-1 adults to auxin led to the complete loss of both full-length tagged proteins within 20 h (Fig 6B). Consistent with NEKL loss of function, 100% of NEKL-2::AID and NEKL-3::AID L1 larvae exposed to auxin from hatching arrested with molting defects (Fig 6C) and displayed abnormal accumulation of apical GFP::CHC-1 (S1 Fig). We note that whereas auxin treatment resulted in the complete disappearance of any detectable NEKL-3::AID protein, we often observed a more rapidly migrating band in samples from auxin-treated NEKL-2::AID worms, suggesting that a partial fragment of NEKL-2::AID may be resistant to further degradation.
To determine the effects of auxin-induced NEKL::AID depletion on clathrin localization in adults, we exposed day-1 adults to auxin for 20 h prior to GFP::CHC-1 localization analysis. In control experiments with wild-type adults, auxin treatment alone did not significantly affect apical hyp7 GFP::CHC-1 localization (Fig 6J and 6K, S2 Fig; S1 and S2 Movies). Strikingly, we observed 2.5-fold and 3.6-fold increases in the average mean intensities of apical GFP::CHC-1 in auxin-treated NEKL-2::AID and NEKL-3::AID adults, respectively, relative to auxin-treated wild-type animals ( Fig 6D, 6G and 6J; S3 and S4 Movies). Likewise, auxin-treated NEKL-2:: AID and NEKL-3::AID adults displayed 1.4-fold and 1.7-fold increases, respectively, in the percentage of pixels above threshold relative to auxin-treated wild-type animals (Fig 6D, 6G and 6K). These data demonstrate that loss of NEKLs leads to abnormal apical clathrin localization through a mechanism that is independent of the molting cycle. The stronger effects observed for NEKL-3 could be due to an incomplete loss of NEKL-2 activity after auxin treatment ( Fig 6B) or to different requirements for NEKL-2 and NEKL-3 in clathrin-mediated endocytosis.
While conducting these studies, we also observed a clear effect of the AID tag on the activities of NEKL-2 and NEKL-3 even in the absence of auxin. Both mean GFP::CHC-1 levels and the percentage of pixels above threshold were increased in the apical hyp7 region of untreated NEKL-2::AID and NEKL-3::AID adults relative to untreated wild-type controls (Fig 6J and 6K; S5 and S6 Movies). The effect of the AID tag was strongest in the NEKL-3::AID strain, consistent with our observation that~20% of untreated NEKL-3::AID worms displayed molting defects (Fig 6C, 6J and 6K). Higher baseline levels in the untreated NEKL::AID strains resulted in somewhat less-dramatic fold changes in comparisons of untreated versus auxin-treated NEKL::AID adults. For example, the average intensity of GFP::CHC-1 increased by 1.6-fold and 1.8-fold in NEKL-2::AID and NEKL-3::AID auxin-treated strains, respectively, compared with age-matched untreated NEKL::AID controls (Fig 6E and 6H). Correspondingly, the percentage of pixels above threshold was increased by 1.2-fold and 1.3-fold in auxin-treated NEKL-2::AID and NEKL-3::AID adults, respectively, relative to untreated controls (Fig 6, 6F and 6I). The observed effects in untreated NEKL::AID strains could be due to the 45-aa AID tag sterically interfering with activities of the NEKL proteins. Alternatively, some auxinindependent activity of the TIR1-E3 ubiquitin-ligase complex could lead to reduced levels of NEKL::AID proteins relative to wild type.
As an additional test, we simultaneously examined clathrin (GFP::CHC-1) and a marker for AP2, P hyp7 ::mScarlet::DPY-23/μ2, in auxin-treated wild-type and NEKL-3::AID adults. Notably, the mean intensity and percentage of positive pixels increased for both clathrin and DPY-23/μ2 in NEKL-3::AID worms relative to wild type (Fig 7A-7I). The greater increase in mean intensity observed for clathrin versus DPY-23/μ2 may reflect the redistribution of clathrin to the apical surface from other compartments within the cell (Fig 7H). In contrast, the large majority of DPY-23/μ2 is expected to reside close the apical surface in wild type. We also note that the Pearson's r coefficient, which is a measure of CHC-1-DPY-23/μ2 colocalization, was significantly greater in auxin-treated NEKL-3::AID animals versus wild type (Fig 7G), also consistent with an increase in apical clathrin. Conversely, we observed a slight decrease in the extent of co-localization between clathrin and a marker for AP1, P hpy7 ::mScarlet::APM-1/μ1, a subunit of the clathrin adapter complex acting in the trans-Golgi compartment (S3 Fig). Altogether our findings indicate that loss of NEKLs leads to increased accumulation of apical clathrin, reflecting increased recruitment or retention of clathrin at or near the plasma membrane.

Depletion of NEKLs greatly reduces clathrin exchange on apical membranes of the epidermis
Our results above show overaccumulation of clathrin on apical hyp7 epidermal membranes after depletion of NEKL-2 or NEKL-3. Such an overaccumulation could be due to a higher rate of coated pit formation, a reduced rate of coated pit release from the plasma membrane, or a reduced rate of clathrin-coated vesicle uncoating. One way to differentiate among these possibilities is to measure clathrin exchange between the membrane-associated and cytoplasmic pools via fluorescence recovery after photobleaching (FRAP) assays using GFP::CHC-1. If coated pit assembly rates were increased in animals lacking NEKLs we would expect to observe increased recovery rates for GFP::CHC-1 after photobleaching. Conversely, if NEKL depletion leads to a decrease in the disassembly of coated vesicles, as occurs after auxilin depletion, we would expect to observe a decrease in recovery after photobleaching [62]. In addition, certain perturbations that block endocytosis in vivo (e.g., exposure to hypertonic sucrose media or intracellular potassium depletion) lead to the genesis of abnormal "clathrin microcages" just below the surface of the plasma membrane, which also block recovery from photobleaching [63]. Moreover, perturbation of the scission enzyme dynamin, which results in failed pinching off of clathrin-coated vesicles, does not affect GFP-clathrin recovery from photobleaching, even though coated pits over-accumulate on the plasma membrane [63]. Therefore, no change in recovery after photobleaching could indicate a role for the NEKLs in vesicle scission.
To determine if the NEKLs affect clathrin dynamics, we carried out FRAP in NEKL-2::AID and NEKL-3::AID strains~20 h after exposure to auxin. In the case of wild-type controls, the mobile fraction of GFP::CHC-1 in the apical hyp7 region appeared to be slightly higher in untreated (75%) versus auxin-treated (64%) animals, although this difference was at most marginally significant (p = 0.071; Fig 8A, S4 Fig, S7 and S8 Movies). Strikingly, the GFP::CHC-1 mobile fraction in NEKL-3::AID adults was reduced to only 20% in auxin-treated worms, a more than 3-fold reduction relative to auxin-treated wild-type animals (Fig 8E, 8F and 8H, S4  Fig, S9 Movie). Similarly, the mobile fraction in NEKL-2::AID strains was just 44% in auxin-treated animals, a 1.5-fold reduction relative to auxin-treated wild type (Fig 8C, 8F and 8G, S4  Fig, S10 Movie). Thus, loss of either NEKL-2 or NEKL-3 results in a strong reduction in clathrin exchange. We note that some bleached spots exhibited homogenous recovery within the bleached area whereas others exhibited nonhomogeneous recovery. This was true for both NEKL-2::AID and NEKl-3::AID auxin-treated worms, although nonhomogeneous recovery was observed more frequently in NEKL-2::AID (16/21) than in NEKL-3::AID (5/17) samples. Similar to what we observed for GFP::CHC-1 mean intensities and pixels above threshold (Fig 6), the AID tag appeared to cause a partial loss of NEKL-2 and NEKL-3 activities, even in the absence of auxin; the mobile fraction in untreated NEKL-2::AID and NEKL-3::AID worms was 61% and 49%, respectively (Fig 8B, 8D and 8F-8H, S4 Fig, S11 and 12 Movies). In addition, the more robust effects observed for NEKL-3::AID versus NEKL-2::AID adults are consistent with the stronger effects observed for NEKL-3::AID in the GFP::CHC-1 localization assays ( Fig  6). Overall, our FRAP data demonstrate that NEKLs strongly affect clathrin exchange in the apical epidermis. Coupled together, our findings indicate that loss of NEKLs leads to the increased stability of clathrin structures located at or near the apical hyp7 surface, a defect consistent with reduced rates of clathrin uncoating and/or the formation of clathrin microcages.

Loss of NEKLs leads to defects in the trafficking of a physiologically relevant membrane cargo
The observed effects on clathrin localization and dynamics following NEKL::AID depletion led us to investigate the role of NEKLs in regulating plasma membrane cargo. LRP-1 is orthologous to human LDL receptor related protein 2 (LRP2), is expressed in hyp7, and is required for normal molting [15]. LRP-1 is thought to bind to and transport low-density lipoproteins into the epidermis from the extracellular space between the epidermis and cuticle. Once internalized, the release and breakdown of LDLs may provide a key source of sterols, which are required for generating hormonal cues necessary for the molting process [4,13].
Consistent with previous reports, LRP-1::GFP was expressed in wild-type in a punctate pattern in the apical epidermis of larvae and adults (Fig 10A and 10C) [15,64]. Moreover, these  puncta correspond to clathrin-coated pits or vesicles based on co-localization with a marker for AP2 (S6 Fig). Notably, depletion of NEKL-3::AID with auxin led to a dramatic change in the localization pattern of LRP-1::GFP in the apical hyp7 region of adults (Fig 10A and 10B; S13 and S14 Movies). Specifically, LRP-1::GFP localization became highly diffusive in the plane of the apical membrane after NEKL-3::AID depletion. This effect was reflected by a 3-fold increase in the percentage of GFP-positive pixels above threshold in auxin-treated wild-type versus NEKL-3::AID adults (Fig 10A and 10B). Furthermore, no overlap in values between auxin-treated wild type and NEKL-3::AID worms was observed (Fig 10B), indicating that apical LRP-1::GFP localization provides a particularly strong readout for NEKL defects. Consistent with previous data, a modest increase in the percentage of positive pixels was observed in untreated NEKL-3::AID adults as compared with wild type (Fig 10B). In addition, we observed a modest increase in the percentage of LRP-1::GFP-positive pixels in medial planes of hyp7 in both treated and untreated NEK-3::AID adults relative to wild type (S6 Fig;  S13 and S14 Movies). Together, our data suggest that NEKL-3 is required for the apical  -1(mew39), and dpy-23(mew74) (C) adults. The group mean and 95% confidence interval (error bars) are shown. p-Values were determined using two-tailed Mann-Whitney tests; ���� p < 0.0001, �� p < 0.01. Raw data are available in S1 File.

NEKLs affect trafficking through a mechanism that is distinct from that of NCAP-1
Several pieces of genetic data suggest that the NEKLs could function through a molecular mechanism that is similar to that of NCAP-1. For example, mutations in ncap-1 and the nekls all suppress the phenotype of AP2 loss-of-function mutants (Fig 4) [42], as do mutations in dpy-23/μ that promote the open AP2 conformation [40]. Furthermore, loss of ncap-1 or dpy-23/μ open mutants strongly enhanced molting defects in partial loss-of-function nekl backgrounds (Fig 3), which can occur when proteins act in a common pathway or process.
To determine if NCAP-1 and the NEKLs might act through a similar mechanism, we first examined CHC-1::GFP in ncap-1(mew39) null mutants. Although mean levels of apical GFP:: CHC-1 were increased by 1.2-fold in ncap-1(mew39) mutants (Fig 11A and 11B), these changes were far less pronounced than the~2-or 3-fold increases observed between auxintreated wild-type and NEKL::AID adults (Fig 6). Moreover, no statistical difference was detected in the percentage of GFP-positive pixels between wild-type and ncap-1(mew39) worms ( Fig 11C). Consistent with a lack of strong effects observed for ncap-1 mutants, we detected no differences in GFP::CHC-1 mean intensity and pixels above threshold between wild type and dpy-23(mew74) open mutants (Fig 11A-11C). Likewise, we failed to detect statistical differences in clathrin localization in ncap-1(mew39); dpy-23(mew74) double mutants, although there appeared to be a trend towards a reduced number of the larger clathrin puncta and an increased number of smaller puncta in this strain (Fig 11A-11C).
As an additional test, we examined GFP::CHC-1 recovery after photobleaching in ncap-1 (mew39) and dpy-23(mew74) mutants. Notably, whereas NEKL::AID auxin-treated worms showed a pronounced reduction in the mobile GFP fraction relative to wild type (Fig 8, S4 Fig), we observed no significant differences between wild-type, ncap-1(mew39), and dpy-23(mew74) worms (Fig 10D-10F). Thus, an increase in the level of open/active AP2 does not appear to result in a marked change in the mobility of clathrin at the apical plasma membrane in hyp7.
We next examined LRP-1::GFP localization in ncap-1(mew39) and dpy-23(mew74) adults. Although we observed a modest increase in the percentage of pixels above threshold (1.2-to 1.5-fold) in these strains relative to wild type, the effects were again much weaker than what we observed in auxin-treated NEKL-3::AID worms (Figs 10C, 10D and 6). These findings strongly suggest that the NEKLs act through a mechanism that is fundamentally different than NCAP-1 and are therefore unlikely to regulate the conformation of AP2 (e.g., to promote the closed state). Nevertheless, our collective data demonstrate that the NEKLs affect a trafficking process that is highly sensitive to the balance between open and closed AP2 conformations.

The mammalian NEKL-3 orthologs, NEK6 and NEK7, rescue molting and clathrin defects associated with nekl-3 loss
The human NIMA kinase (NEK) family of proteins includes two closely related orthologs of NEKL-3, NEK6 and NEK7, which are~70% identical and~85% similar to NEKL-3 and~80% identical and 90% similar to each other [20]. Given their high degree of conservation, we wanted to determine if the human homologs could rescue molting and clathrin-associated defects in nekl-3 loss-of-function backgrounds.
We generated transgenes expressing human NEK6 and NEK7 cDNAs under the control of the wild-type nekl-3 promoter. Our constructs included C-terminal fusions of the NEKs to an intron-containing GFP cassette as well as cDNA-only constructs that lack the GFP tag. Notably, NIMA kinases regulate endocytosis both P nekl-3 ::NEK6::GFP and P nekl-3 ::NEK7::GFP were able to rescue molting defects in null nekl-3(gk506) and hypomorphic nekl-3(sv3) mutants, although their level of rescue was weaker than what was obtained using a~5-kb region derived from the nekl-3 genomic locus (Fig 12A;  S7 Fig) [20]. In addition to full rescue to the adult stage, which was observed in~50% of nekl-3 (gk506) worms expressing human NEK::GFP transgenes (Fig 12A), many transgene-positive worms exhibited at least partial rescue based on their size and stage of arrest (S7 Fig). A NEK6 cDNA-only construct (P nekl-3 ::NEK6) also provided significant rescue in both backgrounds, albeit to a lesser degree than the NEK6::GFP fusion (Fig 12A; S7 Fig). In contrast, a NEK7 cDNA-only construct (P nekl-3 ::NEK7) failed to rescue molting defects in either background (Fig 12A; S7 Fig). The reduced rescuing activity of both the NEK6 and NEK7 cDNAonly constructs is likely due to weak expression, as intron-containing markers boost the expression of cDNAs in C. elegans and other systems [65,66]. Consistent with our rescue data for nekl-3 mutants, P nekl-3 ::NEK6, but not P nekl-3 ::NEK7, was able to partially rescue molting defects following auxin treatment of NEKL-3::AID strains, although rescue was again less robust than for wild-type nekl-3 (Fig 12B).
We next tested if wild-type nekl-3 and P nekl-3 ::NEK6 could rescue GFP::CHC-1 localization defects in NEKL-3::AID strains. As shown in Fig 6, auxin-induced NEKL-3::AID depletion led to a 1.8-fold increase in mean apical GFP::CHC-1 intensity relative to untreated NEKL-3::AID worms. Furthermore, as expected, expression of the wild-type NEKL-3 protein fully rescued clathrin defects in auxin-treated NEKL-3::AID worms (Fig 12C, 12E and 12F). Specifically, no increase was observed in the mean intensity of GFP::CHC-1 or in the percentage of pixels above threshold in treated versus untreated NEKL-3::AID-nekl-3 + worms. Notably, NEKL-3:: AID worms expressing NEK6 displayed only a 1.2-fold increase in the mean intensity of GFP:: CHC-1 relative to untreated NEKL-3::AID-NEK6 + controls ( Fig 12D) and this fold change was significantly lower than NEKL-3::AID worms containing no transgene (Fig 12E). NEK6 expression did not, however, rescue the percentage of positive pixels above threshold relative to the no-transgene control (Fig 12D and 12F). We note that direct comparisons of GFP:: CHC-1 localization data in Figs 6 and 12 were not possible because of differences in marker composition and thresholding procedures. Nevertheless, our findings demonstrate that both human NEK6 and NEK7 can rescue molting defects in nekl-3 mutants and that NEK6 can partially rescue clathrin defects in strains with reduced NEKL-3 function. These findings strongly suggest that the trafficking functions demonstrated for NEKL-3 are conserved across species.

NEKLs regulate clathrin-mediated endocytosis
In this study, we have demonstrated the NEKLs to be novel regulators of clathrin-mediated endocytosis. More specifically, depletion of NEKL-2 or NEKL-3 at the adult stage led to a strong increase in the levels of clathrin and to a dramatic decrease in the mobility of clathrin at the apical surface of hyp7. These findings demonstrate that endocytic defects following NEKL depletion occur independently of molting defects and are thus not merely a secondary consequence of defective molting. The physiological relevance of our findings is further bolstered by our approach to studying NEKLs within their native context and within an intact developing organism. Moreover, given the requirement for endocytic trafficking factors in the molting process [4], our findings indicate that molting defects in nekl-mlt mutants are likely to be a consequence of abnormal trafficking.
https://doi.org/10.1371/journal.pgen.1008633.g012 loss-of-function mutations of individual AP2 subunits as well as mutations in an allosteric activator of AP2, FCHO-1. Likewise, mutations that increase the levels of open/active AP2 strongly enhanced nekl-mlt defects, including dpy-23/μ open mutants and loss of function of NCAP-1, an allosteric inhibitor of AP2. Additionally, defects in clathrin localization and mobility after NEKL-2 and NEKL-3 depletion in adults were rescued by mutations that decrease the levels of open/active AP2. Finally, depletion of NEKL-3 in adults led to dramatic defects in LRP-1/LRP2 endocytosis, which is required for normal molting and is internalized in multiple systems via clathrin-mediated endocytosis [64,67]. Taken together, our findings indicate that loss of NEKLs leads to defects in the internalization of cargo that is required for normal molting.
Given our specific findings for AP2 and FCHO-1, a question arises as to whether mutations in other genes connected to early steps of endocytosis can suppress defects in nekl-mlts. Notably, loss of function in itsn-1/intersectin and ehs-1/EPS15, which have been shown to physically and functionally interact with FCHO-AP2 [29,48,53,[68][69][70], failed to suppress molting defects in nekl-2; nekl-3 mutants (S8 Fig). In fact, mutations in itsn-1 and ehs-1 were found to enhance the penetrance of molting defects following partial loss of nekl-mlt function by RNAi ( S8 Fig). Likewise, similar enhancement effects were observed in strains containing mutations in other core clathrin-pathway components including unc-11/PICALM, unc-57/endophilin A2, sel-5/AAK1, and lst-4/SNX9/18/33 (S8 Fig). These results indicate that nekl-mlts are suppressed specifically by a reduction in AP2 activity and not by the generic impairment of the endocytic pathway. In addition, these findings are consistent with our model that depletion of NEKL activity reduces the uptake of plasma membrane cargo (e.g., LRP-1), a defect that can be further aggravated by inhibiting other components of the endocytic pathway.

Determining the role of NEKLs in clathrin-mediated endocytosis
One explanation to account for a number of our genetic and cell biological observations is that the NEKLs could promote the closed state of AP2. However, several observations argue against this model. (1) In contrast to clathrin localization in nekl mutants and NEKL-depleted adults, clathrin localization in ncap-1 and dpy-23-open mutants was largely unaffected. (2) Unlike clathrin mobility in NEKL-depleted adults, clathrin mobility in ncap-1 and dpy-23-open mutants was indistinguishable from that of wild type. (3) Defects in LRP-1 localization in ncap-1 and dpy-23-open mutants were much less severe than those observed after depletion of NEKL-3. (4) We failed to observe molting defects in ncap-1(mew39), dpy-23(mew74) and ncap-1 (mew39); dpy-23(mew74) double mutants, indicating that increased levels of open/active AP2 alone are not sufficient to induce molting defects. These observations indicate that the NEKLs act through a mechanism that is distinct from that of NCAP-1. Although it remains possible that the NEKLs may carry out multiple functions in endocytosis, including a role in AP2 regulation, we favor a model whereby the NEKLs control a trafficking step that is highly sensitive to the balance between open and closed AP2 conformations.
Based on our FRAP and localization data of clathrin, we suggest that the NEKLs may promote the uncoating of clathrin from internalized coated vesicles. Disassembly of the clathrin coat following membrane scission is carried out by the conserved uncoating ATPase Hsc70, together with its co-chaperone(s), auxilin/GAK [29,71]. Auxilin bound to clathrin recruits Hsc70 to clathrin-coated vesicles, and Hsc70 then interacts with the clathrin heavy chain to alter the conformation of the triskelia, leading to coat disassembly. Notably, inhibition of DNJ-25, the C. elegans ortholog of auxilin, leads to increased clathrin accumulation within the hyp7 epidermis, decreased clathrin mobility (in coelomocytes), and molting defects [62], all of which are observed in nekl-mlt mutants. These observations are consistent with the NEKLs acting either in parallel or upstream of Hsc70-auxilin to promote uncoating, models that will be tested in future studies.
Importantly, we observed that altering the balance between open and closed AP2 conformations led to strong genetic modulation of the molting and clathrin-associated phenotypes in nekl-mlt mutants. How might altering the balance of open/closed AP2 impact the efficiency of clathrin uncoating? It is well established that the conformational opening of AP2 promotes clathrin assembly at membranes and the subsequent formation of coated pits [26,27,29,30]. As such, one appealing hypothesis is that the conformational closing of AP2 may in turn facilitate the uncoating of clathrin from internalized vesicles. Although this kind of reciprocal function for AP2 in both the coating and uncoating of clathrin has not, to our knowledge, been directly demonstrated, it is consistent with much of our genetic and cell biological data.
Notably, auxilin can bind to both clathrin and AP2 [72], providing a possible mechanism by which uncoating could be coupled to the conformation of AP2. In addition, it has been suggested that clathrin uncoating may require the disruption of contacts between AP complexes and clathrin [73] and that clathrin and AP2 uncoating are linked [74]. Furthermore, the levels of PIP2 in membranes affect clathrin uncoating [75][76][77], despite the absence of direct binding of clathrin to membranes. In contrast, AP2 contains a PIP2-binding site that is available for membrane interactions only in the open conformation [27,38,39], suggesting that the conformation of AP2 could affect clathrin uncoating.
A number of early studies using cell culture systems, however, demonstrated that clathrin and AP2 uncoating/exchange can take place largely independently of each other. For example, clathrin release can occur in the absence of AP2 release from membranes [78][79][80][81]. In addition, cytosolic AP2 can be exchanged with membrane-bound AP2 in vesicles containing a stabilized clathrin coat [74,82,83]. Moreover, we failed to observe reduced recovery of clathrin after photobleaching in ncap-1 and dpy-23-open mutants, suggesting that, in an otherwise wild-type background, shifting the AP2 balance toward the open/active conformation does not detectably alter the kinetics of clathrin exchange. Additional studies will be necessary to elucidate the precise mechanism behind AP2 suppression of nekl-mlt phenotypes.
We also note that that the reduced mobility of clathrin observed in NEKL-depleted strains could be due to the formation of structures termed "clathrin microcages". Microcages are extremely small, sharply curved, unusual polymers of clathrin that lack any membrane [84]. Microcages have been observed in mammalian cells that overexpress an inactive form of Hsc70 [85], when potassium or ATP has been depleted from cells, or when cells are exposed to hypertonic sucrose [63,84]. Similar to what we observed in NEKL-depleted worms, the fluorescence recovery of clathrin after photobleaching was strongly impaired in cells with abundant microcages [63,74]. These microcages may in fact sequester cytosolic clathrin, leading to reduced clathrin availability and lower levels of endocytosis. Consistent with this, conditions that induce microcages also lead to the dispersal of LDL receptors on the membrane surface [84], a phenotype we observed for the LDL receptor, LRP-1, in NEKL-depleted strains. Future EM studies will determine the ultrastructure of the exchange-resistant clathrin foci in NEKLdepleted worms.

Trafficking functions of NIMA kinase family members are likely conserved across species
Our rescue of nekl-3 molting defects with constructs expressing human NEK6 and NEK7 strongly indicates that these proteins carry our similar functions in diverse organisms. Moreover, we observed rescue of clathrin mislocalization defects in nekl-3 mutants with human NEK6. Although the large majority of papers on NEK6 and NEK7 have focused on functions associated with cell division, a high-throughput analysis of the human kinome by Zerial and colleagues indicated that clathrin-mediated endocytosis is strongly decreased in cells when NEK6 or NEK7 is targeted by siRNAs [86]. Likewise this study also identified moderate defects in endocytosis following siRNA treatment of NEK8 and NEK9, the closest human homologs of nekl-2. A second large-scale study by the Zerial group also reported abnormal trafficking after knockdown of NEK6 and NEK8 based on multiple parameters [87]. In addition, there is some evidence to suggest that NIMA family members in A. nidulans and S. cerevisiae carry out functions connected to endocytosis [88]. Lastly, protein association studies of NEK6 and NEK7 have identified factors connected to both trafficking (e.g., AP2A1 and AP2B1) and to cytoskeletal factors that impact endocytic functions (e.g., CDC42 and actin) [89][90][91], although the functional significance of these interactions was not explored.
Given these collective observations, we propose that the control of intracellular trafficking may be an ancient and conserved function for members of the NIMA kinase family. This would represent a role for mammalian NEK kinases that has been largely overlooked but could be relevant to their involvement in diseases including cancer and ciliopathies [23,[92][93][94][95][96][97][98][99][100][101][102][103][104][105][106]. Future studies in both C. elegans and mammalian systems will be important to establish the precise functions of NIMA-related kinases in cellular trafficking and human disease processes.

Strains and maintenance
C. elegans strains were maintained according to standard protocols [107] and were propagated at 22˚C. Strains used in this study are listed in S2 Table. CRISPR/Cas9 CRISPR/Cas9 ribonucleoproteins in combination with the dpy-10 co-CRISPR method [108][109][110] was used to generate genomic lesions except for the pw27(nekl-2::aid) and pw29(nekl-3:: aid) alleles, which were created using the self-excising cassette method [111]. pw27(nekl-2::aid) and pw29(nekl-3::aid) guide RNA and repair template plasmids were created from existing plasmids pDD268 and pDD268, respectively [21], using Gibson assembly to directly replace mNeonGreen sequences with AID sequences. Cas9 enzyme was purchased from University of California, Berkeley, whereas, crRNA, tracrRNA, and the repair templates were brought from GE Healthcare Dharmacon, Inc. Briefly, 7.8 μl Berkeley Cas9 (6.4 μg/μl), 0.75 μl 3 M KCl, 0.75 μl 200 mM HEPES (pH 7.4), 5 μl 0.17 mM tracrRNA, 0.8 μl 0.3 mM dpy-10 crRNA, 2 μl 0.3 mM target crRNA, and 0.75 μl of distilled water were mixed and incubated at 3˚C for 15 minutes. After incubation, 16 μl of dpy-10 repair template and 1.6 μl of 10 μM target repair template was added, and the mixture was injected into the gonads of adult worms of the desired strains. See S1 Text for information on the sequences of oligos used for all CRISPR/Cas9 studies; sequence information on the obtained CRISPR/Cas9 lesions is indicated in S1 Table. Western blot analysis L4 animals were picked to NGM plates with or without 1mM auxin and harvested after 20 hrs at 20˚C. For each lane 100 young adult animals of each genotype/condition were handpicked into 30 μl of lysis buffer (100 mM Tris pH 6.8, 8% SDS, 20 mM β-mercaptoethanol) and incubated at 37˚C for 60 min in a thermocycler. Then 5 μl of 6X Laemmli sample buffer was added, and the samples were further incubated at 100˚C for 5 min. Extracted proteins were separated by 4-20% gradient ExpressPlus PAGE gel (Genscript) and blotted to nitrocellulose using Genscript MOPS buffer. After blocking in 5% non-fat milk for 60 min, the blot was rinsed and probed with mouse Monoclonal ANTI-FLAG M2 (Sigma, 1:1000 dilution) followed by HRPconjugated Goat anti-mouse antibody (1:10,000). Signals were detected using Supersignal West Pico chemiluminescent substrate (Thermo Scientific) and exposure to film. The blot was then stripped and probed with rabbit anti-actin antibody (Sigma A2066, 1:1000) followed by HRP-conjugated Goat anti-rabbit antibody (1:10,000).
RNAi dsRNAs corresponding to apa-2, dpy-23, aps-2, ncap-1, and fcho-1 were generated using standard methods [116], followed by injection at 0.8-1.0 μg/μl into worm gonads (see S1 Text for oligo sequences). For molting enhancement studies, RNAi feeding was performed using bacterial strains from Geneservice following standard protocols [117]. Worm strains were first grown for one (Fig 3B and 3D) or two (Fig 3E) generations on lin-35(RNAi) plates to increase RNAi susceptibility [118]. Gravid adults were then transferred to experimental RNAi plates and allowed to lay eggs for~24 h, and F1 progeny were scored for molting defects after an additional~72 h. Because mlt-3(RNAi) induced a high percentage of molting defects in wild type, mlt-3(RNAi) bacteria were diluted 1:10 with the control gfp(RNAi) strain.
For FRAP assays, anesthetized worms were analyzed immediately after being placed on slides (<10 min). The apical region of hyp7 was brought into focus, and a circular spot (7.272 μm in diameter) was photobleached using an iLas2 system (BioVision Technologies) with a 56-ms pulse of a 405-nm laser set at 50% power. After photobleaching, GFP::CHC-1 FRAP was detected by imaging every 5 s for a total of 180 s. Data acquired for Figs 8 and 11 were carried out at the same time using the same microscope and acquisition procedures. For this reason, some identical control data is present in these figures. Supplementary movies 7-12 are 200 s long and have been condensed 25× (8 s).

Image analysis
To quantify the mean intensity and the percentage of fluorescence-positive pixels above threshold (Figs 5,6 and 8-11), background fluorescence was subtracted from z-stack images using Fiji software (NIH) available at https://imagej.net/Fiji/Downloads). For a given z-plane of interest, the polygon selection tool was used to demarcate the region of hyp7 followed by mean intensity measurement. The percentage of fluorescence-positive pixels for the region of interest was determined after thresholding, and the same thresholding algorithm was used for strain comparisons.
FRAP time-lapse images were aligned to correct for any movement of the worms by using the "Rigid Body" Transformation method in the StackReg plugin (available at https:// imagej.net/StackReg). Next, the time-lapse images were background subtracted and analyzed using the custom-written Stowers plugin (available at http://research.stowers.org/ imagejplugins), which can be used with the Fiji software. The photobleached region was selected and mean intensities were quantified before photobleaching and after photobleaching for each time point. Fitting of FRAP curves was performed using batch FRAP fit in the Stowers plugin. Fitted curves were then normalized to prebleach mean intensities and then averaged to obtain final recovery curves. Mobile fractions were quantified using the values obtained from the fitted curves. The mobile fraction is determined by the following equation: amplitude/(prefrap-baseline) (see https://research.stowers.org/imagejplugins/ImageJ_ tutorial2.html).

Auxin treatment experiments
Auxin was purchased as indole-3-acetic acid from Alfa Aesar. In these experiments, L4-stage worms were transferred to plates and left to develop into adults (~20 h). A 0.4 M (100×) stock auxin solution was made by dissolving 0.7 g of auxin in 10 mL 100% ethanol. Each plate containing day-1 adults was treated with a mixture of 25 μl of stock auxin solution and 225 μl of distilled water.

Statistics
Statistical tests were performed as indicated using software from Prism GraphPad. Statistical tests comparing fold change ratios (Fig 12; S5 Fig) were carried out as described by Fay and Gerow [119]. Background subtraction was performed using the same parameters for all images; minimum and maximum pixel values were kept consistent for all images. Bar in A = 5 μm (for all panels). Mean GFP::CHC-1 intensities (B,E,H,K) and the percentage of GFP-positive pixels above threshold (C,F,I,L) were determined for day-2 adults. Panels A-C show data for the apical region of hyp7; panels D-L show data for a medial region of hyp7. (B,C,E,F,H,I,K,L) Both the group mean and 95% confidence interval (error bars) are shown. Note that wild-type showed small but statically significant increases in medial GFP::CHC-1 intensity and pixels above threshold after exposure to auxin, suggesting that auxin itself could exert a weak effect on GFP::CHC-1 localization. p-Values were determined using two-tailed Mann-Whitney tests; �� p < 0.01, � p < 0.05. Raw data are available in S1 File. :GFP and a marker for AP2, P dpy-7 ::mScarlet::DPY-23. mScarlet is represented as magenta, overlap is white. (B) Representative confocal images of LRP-1::GFP in auxin treated wild-type and NEKL-3::AID adults. Images show the medial plane of hyp7 (also see S13 and S14 Movies). Inverted fluorescence was used to aid clarity. Bar size in A,B = 5 μm. (C) Percentage of GFP-positive pixels above threshold were determined for individual adults. p-Values were determined using two-tailed Mann-Whitney tests; ���� p < 0.0001. Raw data are available in S1 File. (TIFF) S7 Fig. Supplemental NEK rescue data. (A,B) Bar plot showing rescue of molting defects in nekl-3(sv3) strains with the indicated transgenes. NEK6::GFP and NEK7::GFP refer to P nekl-3 :: NEK6::GFP and P nekl-3 ::NEK7::GFP, respectively. NEK6 and NEK7 refer to P nekl-3 ::NEK6 and P nekl-3 ::NEK7, respectively. p-Values were determined using Fischer's exact test; ���� p < 0.0001. (B) Bar plot showing the percentage of L1/L2 versus L2-L4 arrested larvae in transgene-positive nekl-3(gk506) mutants. Note that~40-60% of transgene-positive arrested larvae bypass the L1/L2 arrest point, whereas the large majority of transgene-minus worms arrest at L1/L2. Given~50% rescue to adulthood by the NEK6::GFP and NEK7::GFP transgenes (Fig 12A), partial-to-full rescue occurs at a frequency of~75% in transgene-positive nekl-3(gk506) mutants.