Reevaluation of the role of LIP-1 as an ERK/MPK-1 dual specificity phosphatase in the C. elegans germline

The fidelity of a signaling pathway depends on its tight regulation in space and time. Extracellular signal-regulated kinase (ERK) controls wide-ranging cellular processes to promote organismal development and tissue homeostasis. ERK activation depends on a reversible dual phosphorylation on the TEY motif in its active site by ERK kinase (MEK) and dephosphorylation by DUSPs (dual specificity phosphatases). LIP-1, a DUSP6/7 homolog, was proposed to function as an ERK (MPK-1) DUSP in the Caenorhabditis elegans germline primarily because of its phenotype, which morphologically mimics that of a RAS/let-60 gain-of-function mutant (i.e., small oocyte phenotype). Our investigations, however, reveal that loss of lip-1 does not lead to an increase in MPK-1 activity in vivo. Instead, we show that loss of lip-1 leads to 1) a decrease in MPK-1 phosphorylation, 2) lower MPK-1 substrate phosphorylation, 3) phenocopy of mpk-1 reduction-of-function (rather than gain-of-function) allele, and 4) a failure to rescue mpk-1–dependent germline or fertility defects. Moreover, using diverse genetic mutants, we show that the small oocyte phenotype does not correlate with increased ectopic MPK-1 activity and that ectopic increase in MPK-1 phosphorylation does not necessarily result in a small oocyte phenotype. Together, these data demonstrate that LIP-1 does not function as an MPK-1 DUSP in the C. elegans germline. Our results caution against overinterpretation of the mechanistic underpinnings of orthologous phenotypes, since they may be a result of independent mechanisms, and provide a framework for characterizing the distinct molecular targets through which LIP-1 may mediate its several germline functions.

Extracellular signal-regulated kinases (ERKs) are a group of serine/threonine protein kinases and classical members of mitogen activated protein kinases (MAPKs). The ERK MAPKs are terminal enzymes of a highly conserved three-tiered kinase signaling cascade, the RAS–ERK pathway (1, 2). Extracellular stimuli, including growth factors and insulin signaling induce the sequential activation of RAS–ERK pathway that orchestrates a wide range of cellular processes such as gene expression, proliferation, differentiation, and apoptosis to regulate tissue and organismal homeostasis (Fig. 1A) (1–3). Because the ERK MAPK signaling pathway regulates a myriad of developmental processes for controlled and ordered execution of the pathway, ERK activity is tightly monitored in space and time (4). MEK (also known as MAPK/ERK kinase) phosphorylates ERK at threonine and tyrosine residues (TEY motif), thus activating its function (1). Active ERK is then inactivated by dual specificity MAPK phosphatases (MKPs or DUSPs) that remove the phosphate residues. Together, MEK and DUSPs shape the magnitude, duration, and spatiotemporal profile of ERK activity (1, 4–6).Open in a separate windowFig. 1.lip-1 mutants are defective in pachytene exit and oocyte formation. (A) Schematic view of the conserved LET-60 (RAS)–MPK-1 (ERK) pathway showing that the regulation of ERK/MPK-1 activation depends on upstream kinase cascade and dephosphorylation depends on DUSPs. (B) Schematic view of a hermaphroditic C. elegans germline displaying the spatiotemporal nature of MPK-1 activation. The germline is oriented in a distal (*) to proximal direction from left to right. Proliferative PZ cells are in the distal region, capped by the distal tip cell (DTC). Germ cells enter meiotic prophase at the transition zone (TZ), followed by progression through different stages of meiotic prophase. The “loop region” is the anatomic bend in the U-shaped gonad. The −1 marks the oldest oocyte at the proximal end. Active MPK-1 is visualized by a specific dpMPK-1 antibody in two distinct regions of the germline: midpachytene, termed as zone 1, and proximal few oocytes, termed as zone 2. The intensity of the color (red) correlates with strength of MPK-1 di-phosphorylation. (C) Predicted activation of MPK-1 in the absence of DUSP: either distal to zone 1, called “precocious” activation, or in the late-pachytene/early-diplotene region (anatomically in the loop region), called “ectopic” activation. (D–I) Differential interference contrast microscopy images of germlines from indicated genotypes, age, and temperature to visualize germline morphology. The loop region is on the right in the photographs and oocytes on the ventral side. Oocytes are numbered from proximal to distal polarity (toward loop). The most proximal oocyte is labeled as −1. Arrowheads indicate oocytes, and arrows indicate pachytene-stage germ cells. (J) Quantification of germlines of the indicated genotypes, with pachytene-progression–defective phenotypes expressed as a percentage. (K–P) Dissected DAPI-stained germlines of the indicated genotypes (mid-L4 + 24 h at 25 °C) displaying germline morphology. Insets are magnified views of germ cell(s) at the proximal gonad (after loop region). The dissected germlines are oriented with the distal on the left (*) to proximal on the right of the photograph, according to the meiotic progression. Arrowheads indicate oocytes, and arrows indicate pachytene germ cells. The total number of germlines (n) analyzed per genotype is indicated in each panel (scale bars, 25 μm).The Caenorhabditis elegans oogenic germline, like most complex biological systems, displays a controlled spatiotemporal pattern of ERK (MPK-1 in C. elegans) activity (7–11). Active MPK-1, as assayed using an antibody that detects dual phosphorylated MPK-1 at threonine and tyrosine of the TEY motif (7, 12), is visualized in midpachytene (termed as zone 1 of activation). However, MPK-1 is dephosphorylated, and thus, its activity is very low in the late-pachytene and early-diplotene region of the germline, which corresponds to the anatomic “loop” of the C. elegans U-shaped gonad (Fig. 1B). MPK-1 phosphorylation is again visible in the proximal diakinesis oocytes (termed as zone 2) in a hermaphroditic germline (7–9). Zone 2 activation is mediated by a secreted sperm signal (major sperm protein, or MSP), which antagonizes the VAB-1 Ephrin receptor (13). Thus, zone 2 activation is absent in C. elegans females, which do not produce sperm (7). In a wild-type oogenic hermaphroditic germline, active MPK-1 has not been visualized in the distal germline, from the progenitor zone (PZ) to midpachytene, and is very low in the loop region of the germline. Because total MPK-1 protein is expressed throughout the germline (8), the striking spatiotemporal activation pattern of MPK-1 observed using the dual-phosphospecific antibody suggests localized activation and inactivation of MPK-1 through MEK and DUSPs.In the oogenic hermaphroditic germline, the phenotypic consequences of MPK-1 activation are complex. In genetic mutants of the mpk-1 pathway, changes to the MPK-1 activation pattern along the spatiotemporal axis, as well as alterations to signal strength, produce distinct phenotypes. For example, a complete loss of MPK-1 activity in a null allele causes the oogenic germ cells to arrest in early- to midpachytene (8, 14). In the absence of MPK-1 activity, the germ cells fail to launch the apoptotic program because they do not progress into midpachytene, the stage in which meiotic checkpoint activation culls errors (9, 15). Reduction of MPK-1 signal strength using temperature-sensitive (ts) alleles, however, produced different phenotypes depending on the time at which MPK-1 activation was reduced during oogenic development. These mpk-1(ts) germlines exhibit increased apoptosis (due to higher levels of meiotic asynapsis defects; Ref. 11), a pachytene-progression defect in which pachytene-stage cells linger and are observed in the loop region, and fewer oocytes with an increased size relative to wild-type animals (8). Conversely, in RAS/let-60 gain-of-function mutants, the spatiotemporal pattern of MPK-1 activation is different from the wild-type in two regions: 1) in midpachytene, the germline displays “precocious” activation of MPK-1, and 2) the loop region exhibits “ectopic” MPK-1 activation (Fig. 1C). These animals, unlike the wild-type, display multiple small oocytes (8). Because of the striking increase in oocyte number in the RAS/let-60 gain-of-function mutants, an increase in oocyte number has been considered as a readout for MPK-1 activation. Mutants displaying multiple small oocytes are thus interpreted to be a consequence of increased MPK-1 activity.The C. elegans genome has 29 predicted DUSPs, of which LIP-1 (lateral signal induced phosphatase-1) bears homology with mammalian DUSP6/7 (16, 17). Genetic evidence suggested that loss of lip-1 negatively regulates MPK-1 during somatic vulval development (17). In vitro, in mammalian Cos-1 cultured cells, Myc-tagged LIP-1 protein was shown to dephosphorylate mammalian ERK1/2 (16). Coupled with the homology to mammalian DUSPs, the authors concluded that LIP-1 functions as an MPK-1 DUSP in vivo. In the C. elegans germline, immunofluorescence staining using an anti-LIP-1 antibody showed that the total protein is expressed from the proximal one-third of the PZ region and throughout the pachytene as membrane-associated bright puncta (18). LIP-1 was proposed to function as an MPK-1 DUSP, in the germline, from two lines of evidence (18), which we reevaluated based on the reasoning outlined below. First, in the prior report, the authors showed that in a feminized germline, which does not produce any sperm signal, loss of lip-1 led to an increase in phosphorylated MPK-1 in zone 2 (Fig. 1B). However, in the absence of the sperm signal, MPK-1 cannot be phosphorylated in zone 2 to begin with (7, 13) (Fig. 1A). In this context, inactivation or absence of a DUSP (LIP-1, in this case) should not lead to an increase in the level of phosphorylated MPK-1 since it was never phosphorylated. Second, the authors observed that loss of lip-1 led to ectopic (loop region) MPK-1 activation in hermaphrodites coupled with an increase in oocyte numbers. The authors interpreted this phenotype to be similar to that of RAS/let-60 gain-of-function mutant germlines (18). However, recent work has revealed that the increased oocyte production in RAS/let-60 gain-of-function animals is due to the “precocious” activation of MPK-1 in the early-pachytene, rather than the “ectopic” MPK-1 activation in the loop region (Fig. 1C) (11). Together, these two lines of reasoning led us to reinvestigate the role of LIP-1 as an MPK-1 DUSP in the C. elegans germline and to determine where in the germline spatiotemporal axis LIP-1 might function to regulate oocyte formation, using cytology, genetics, and phenotypic analyses.Contrary to what was previously published (18), our results show that 1) precocious or ectopic MPK-1 activity is not detected in the absence of lip-1—in fact, we found that loss of lip-1 led to lower MPK-1 activation; 2) loss of lip-1 fails to rescue the pachytene-progression and fertility defects observed upon reducing mpk-1 function; and 3) germlines with loss of lip-1 displayed an mpk-1 loss-of-function–like oocyte phenotype, rather than a gain-of-function–like oocyte phenotype, and 4) led to lower MPK-1 substrate phosphorylation. Moreover, we show that mutants in other genes, such as ooc-5 (human ortholog of torsinA AAA+ ATPase), also exhibit multiple small oocytes (19, 20) but do not present with ectopic MPK-1 activity, suggesting that increased oocyte number is not invariably equivalent to, or due to, increased MPK-1 phosphorylation. In support of this, we observed that loss of rskn-1 (human ortholog of RPS6KA, ribosomal protein S6 kinase A), which results in increased ectopic activation of MPK-1 in the loop region of the germline, does not exhibit increased oocyte numbers. This finding demonstrates that ectopic MPK-1 activation does not necessarily cause oocyte numbers to increase. Finally, in wild-type C. elegans diplotene oocytes, the synaptonemal complex (SC) central proteins are removed from the long arm of the chromosome axis to allow for accurate chromosome segregation (21). However, RAS/let-60(ga89ts) gain-of-function mutants have been shown to retain the SC central proteins on the long arm (10). Nadarajan et al. (10) reported that loss of lip-1 also leads to retention of the SC central protein to the long chromosomal arm and proposed that this was because of an increase in MPK-1 activation. We found that while the SC central element proteins are retained on the long arm of the chromosome in diplotene oocytes in both RAS/let-60(ga89ts) gain-of-function and lip-1 mutant oocytes, they are not retained in the rskn-1 mutant germlines, which display increased MPK-1 activation in oocytes. Thus, the retention of the SC central proteins in lip-1 mutant germlines likely occurs through MPK-1–independent mechanisms, suggesting that multiple regulatory processes, both independent of and dependent on ectopic MPK-1 phosphorylation, control SC disassembly. Together, these data demonstrate that LIP-1 does not function as an MPK-1 DUSP in the context of the C. elegans germline and may have multiple other targets through which it mediates its several germline functions.