Two p90 Ribosomal S6 Kinase Isoforms Are Involved in the Regulation of Mitotic and Meiotic Arrest in Artemia

Background: Ribosomal S6 kinase (RSK) plays important roles in meiosis and mitosis.

Results: Ar-Rsk1 promoted cdc2 phosphorylation and thereby meiotic arrest, whereas Ar-Rsk2 knockdown resulted in mitotic arrest.

Conclusion: Ar-Rsk1, which lacks an ERK-docking motif, controls meiotic arrest, whereas Ar-Rsk2, which has an ERK-docking motif, controls mitotic arrest.

Significance: The data provide insight into the functions of RSK isoforms.

Abstract

There are multiple isoforms of p90 ribosomal S6 kinase (RSK), which regulate diverse cellular functions such as cell growth, proliferation, maturation, and motility. However, the relationship between the structures and functions of RSK isoforms remains undetermined. Artemia is a useful model in which to study cell cycle arrest because these animals undergo prolonged diapauses, a state of obligate dormancy. A novel RSK isoform was identified in Artemia, which was termed Ar-Rsk2. This isoform was compared with an RSK isoform that we previously identified in Artemia, termed Ar-Rsk1. Ar-Rsk2 has an ERK-docking motif, whereas Ar-Rsk1 does not. Western blot analysis revealed that Ar-Rsk1 was activated by phosphorylation, which blocked meiosis in oocytes. Knockdown of Ar-Rsk1 reduced the level of phosphorylated cdc2 and thereby suppressed cytostatic factor activity. This indicates that Ar-Rsk1 regulates the cytostatic factor in meiosis. Expression of Ar-Rsk2 was down-regulated in Artemia cysts in which mitosis was arrested. Knockdown of Ar-Rsk2 resulted in decreased levels of cyclin D3 and phosphorylated histone H3, and the production of pseudo-diapause cysts. This indicates that Ar-Rsk2 regulates mitotic arrest. PLK and ERK RNAi showed that Ar-Rsk2, but not Ar-Rsk1, could be activated by PLK-ERK in Artemia. This is the first study to report that RSK isoforms with and without an ERK-docking motif regulate mitosis and meiosis, respectively. This study provides insight into the relationship between the structures and functions of RSK isoforms.

Introduction

The ribosomal S6 kinase (RSK)² family comprises a group of ERK/MAPK effectors that regulate diverse cellular functions by phosphorylating nuclear and cytoplasmic targets. Hyperactive RSK signaling regulates cell division in several types of cancer, including breast, prostate, and lung (1–3). The RSK inhibitor SL-0101 arrests MCF-7 cells (a breast cancer cell line) in the G₁ phase, and thereby inhibits their proliferation (1). RSK is overexpressed in more than 50% of primary malignant lung lesions, whereas it is undetectable in normal epithelium (3). This suggests that RSK promotes the proliferation of tumor cells.

In most metazoa, unfertilized oocytes are arrested at meiotic metaphase because of a cytoplasmic activity called cytostatic factor (CSF) (4, 5). This ensures that oocytes are released into the ovisac (uterus) in an optimal physiological state, which reduces the risk of insemination failing, such as occurs in polyspermy (6). The E3 ubiquitin ligase anaphase-promoting complex/cyclosome can degrade the components of CSF, after which meiosis resumes (4). The MAPK-RSK pathway is required to maintain CSF-dependent meiotic arrest. In this pathway, RSK phosphorylates its substrate Erp1 (also called Emi2) on Ser³³⁵–Thr³³⁶, which leads to activation of Erp1. Activated Erp1 inhibits anaphase-promoting complex/cyclosome in vivo and in vitro (7, 8).

RSK isoforms contain an N-terminal kinase domain (NTKD) and a C-terminal kinase domain (CTKD), which are connected by a regulatory linker. All RSK isoforms, including the orthologues of RSK in Caenorhabditis elegans and Drosophila melanogaster, contain four essential phosphorylation sites (Ser²²¹, Ser³⁶³, Ser³⁸⁰, and Thr⁵⁷³ in human Rsk1) that are responsive to mitogenic stimuli (9). ERK1/2 phosphorylates Thr⁵⁷³ (located in the CTKD) and Ser³⁶³ by docking to the ERK-docking motif. The CTKD then autophosphorylates Ser-380, which creates a docking site for PDK1, which, in turn, phosphorylates Ser²²¹ in the NTKD. Phosphorylation of Ser²²¹ stabilizes NTKD in an active conformation, resulting in full activation of RSK (9–11). This hierarchical phosphorylation cascade is essential to allow active RSK isoforms to perform their various functions.

The ERK-docking motif (⁷³⁹LAQRRVRKLPSTTL⁷⁵²) is located near the C terminus of RSK and proves an important motif for RSK activation. In quiescent HEK293 cells, endogenous ERK1/2 associates with immunoprecipitated wild-type Rsk1. However, a C terminally truncated Rsk1, lacking the last 11 amino acids that encode the ERK-docking motif, does not associate with ERK1/2 (12). Roux et al. (12) transfected HEK293 cells with Rsk1 point mutants and found that the Rsk1 L739A, R742A, and R743A mutants were unable to bind ERK1/2. These mutations also prevented Rsk1 activation, as shown by kinase assays using substrates for the N-terminal kinase activity of Rsk1 (12). Therefore, the ERK-docking motif is required for ERK1/2 binding to RSK and is shown to be essential for ERK1/2-mediated activation.

The effects elicited by RSK isoforms depend on the cell type and isoform. Cell proliferation is promoted by some RSK isoforms, whereas it is inhibited by others (13–18). Rsk1 and Rsk2 positively regulate the proliferation of tumor cells (1, 13). Consistent with this, the level of Rsk1 is often higher in tumors than in healthy tissues (14). By contrast, levels of Rsk3 and Rsk4 are usually lower in tumors than in healthy tissues (15). In addition, Rsk4 plays an inhibitory role in embryogenesis (15), and participates in p53-dependent cell growth arrest (16) and oncogene-induced cell senescence (17). Rsk2 is highly expressed in Xenopus oocytes and helps control critical stages of the meiotic cell cycle (18). For example, Rsk2 controls G₂/M progression in meiosis I by phosphorylating and thereby inhibiting the kinase Myt1 (19). Rsk1 is thought to promote metaphase II arrest by phosphorylating and thereby activating the kinase Bub1, which is a mediator of anaphase-promoting complex inhibition (20).

Salt lakes on plateaus are the most hostile environments on earth, and hardly any animals can survive in these environments. One notable exception is Artemia, a small crustacean that can survive in extreme conditions such as high salinity, low levels of oxygen, and large change in temperature (21). To cope with such harsh habitats and widely different environmental conditions, Artemia have two independent reproductive pathways (22). Under unfavorable environmental conditions, mature Artemia produce and release encysted gastrula embryos that follow the diapause-destined developmental pathway and thereby enter diapause (an obligate dormancy) (23, 24). When environmental conditions are favorable, embryonic development precedes uninterrupted, and mature Artemia release swimming nauplius larvae.

Diapause-destined development can be divided into four stages, namely, pre-diapause, diapause, post-diapause, and nauplius. In the pre-diapause stage, fertilized embryos develop into late gastrulae in the uterus (21). Diapause embryos are composed of 4000 cells arrested at G₁/S phase. In these embryos, metabolic activity is greatly reduced and RNA/protein synthesis does not occur (25–28). Post-diapause embryos, in which the cell cycle is arrested at the G₂/M phase (29), are obtained by storing diapause embryos at −20 °C for at least 3 months (30). In favorable conditions, post-diapause embryos hatch, the diapause is terminated, and cell division and development resume, resulting in the generation of nauplii (30–32). The mechanisms underlying this development remain unclear. A complex enzymatic system, including the RSK regulatory pathway, is suggested to be involved in diapause formation and termination (30, 33, 34).

In this study, we identified a novel RSK isoform in Artemia, which we termed Ar-Rsk2. We compared this function of this kinase with that of another RSK isoform that we previously identified called Ar-Rsk1. Ar-Rsk1, which lacks an ERK-docking motif, played an important role in regulating CSF and meiotic arrest. By contrast, Ar-Rsk2, which has an ERK-docking motif, functioned in the regulation of mitotic arrest. This is the first study to report that RSK isoforms with and without an ERK-docking motif regulate mitosis and meiosis, respectively. Thus, this study provides an insight into the relationship between the structures and functions of RSK isoforms.

EXPERIMENTAL PROCEDURES

Animals

Specimens of amphigenic Artemia franciscana from salterns in San Francisco Bay were purchased from San Francisco Bay Brand, Hayward, CA, as dried (activated) encysted embryos. The daily light regime (h of light:h of dark) and the percentage of artificial seawater determine the reproductive mode of these animals (36). Diapause-destined Artemia were raised in 8% (w/v) artificial seawater at 25 °C in a 4-h light:20-h dark cycle, whereas Artemia that developed directly were raised in 4% (w/v) artificial seawater in a 16-h light:8-h dark cycle. Chlorella powder was supplied as brine shrimp food.

Cloning of Ar-Rsk1 and Ar-Rsk2

Ar-Rsk1 cDNA was isolated as described by Dai et al. (25). Ar-Rsk2 cDNA was obtained by performing two rounds of 3′ rapid amplification using the FirstChoice^TM RLM-RACE Kit and 5′ PCR with gene-specific primers (Table 1). The sequence of Ar-Rsk2 was imported into SeqMan Lasergene software. The sequences of the complete cDNA and the deduced peptide were determined using Blastn and Blastx, respectively, on the NCBI website. The nucleotide sequence of Ar-Rsk2 cDNA was submitted to GenBank^TM under accession number KJ679441.

TABLE 1.

Primers used for PCR

The sequences that are recognized by restriction endonucleases are underlined. TubulinF and TubulinR were designed based on the cDNA sequence stored under GenBank^TM accession number AF427598 (R, reverse; F, forward).

Primer	Length	Positions of nucleotides	Direction	Sequence (5′ – 3′)
	bp
3′F1	20	1772–1791	F	TGCGACTTTGGTTTCGCGAA
3′F2	20	1820–1839	F	ACGCCTTGCTACACTGCCAA
5′F	24	1–24	F	ACATTCTATGCAAATAAGCCTTCA
5′R	20	1787–1806	R	TTTTCTCTTAATTGTTTAGC
Ar-Rsk1QF	23	2212–2234	F	TATATATTTAGTACCAGGGACC
Ar-Rsk1QR	23	2428–2450	R	CATGAGAAGAATTTTCGGGGTGG
Ar-Rsk2QF	23	2208–2230	F	TGCGTGCGTTTACCAATCCTCAT
Ar-Rsk2QR	23	2358–2380	R	CAGTATCTGGCAACTTCACTCTG
TubulinF	20	446–465	F	GCAGTGGTCTACAAGGTTTC
TubulinR	22	774–795	R	ATCAAAACGAAGGCTGGCGGTG
Ar-Rsk1iF	27	2217–2235	F	GCTCTAGAATTTAGTACCAGGGACCGA
Ar-Rsk1iR	26	2447–2465	R	GGAATTCCTGCTCAATTTTACACATG
Ar-Rsk2iF	28	2210–2229	F	GCTCTAGACATGCGTTTACCAATCCTCA
Ar-Rsk2iR	30	2358–2380	R	GGAATTCCAGTATCTGGCAACTTCACTCTGCA
GFPiF	30	122–144	F	GCTCTAGAAACTTACCCTTAATTTTATTTGC
GFPiR	28	461–480	R	CATGCCATGGGCCATTCTTTGGTTTGTCTC
Ar-Rsk1EF	25	1788–1812	F	CTAAACAATTAAGAGAAAACGGACT
Ar-Rsk1ER	26	2215–2240	R	AAGTTCGGTCTTCTGGTACTAAATAT
Ar-Rsk2EF	20	1886–1905	F	TGTGACATTTGGAGCCTTGG
Ar-Rsk2ER	25	2274–2298	R	GTTTTAGCACGTGACTGTCTTCTTT

Semi-quantitative RT-PCR

Total RNA was extracted using TRIzol reagent according to the manufacturer's instructions. First-strand cDNA was synthesized from 1 μg of total RNA using M-MLV Reverse Transcriptase in a 10-μl reaction. Ar-Rsk1, Ar-rsk2, and α-tubulin cDNAs (GenBank accession number AF427598) were amplified separately in 25-μl reactions using 1 μl of the cDNA as a template. The primers used are shown in Table 1.

dsRNA Synthesis and Microinjection

Reconstructed plasmids that contained two inverted T7 polymerase sites flanking the cloning region were obtained (named pET-T7), as described previously (25). The 3′ terminus differs between Ar-Rsk1 and Ar-Rsk2. A 249-bp fragment of Ar-Rsk1 and a 171-bp fragment of Ar-Rsk2, including the 3′-UTR and parts of the 3′ coding region, were PCR amplified using the Ar-Rsk1iF/Ar-Rsk1iR and Ar-Rsk2iF/Ar-Rsk2iR primer sets, respectively (Table 1). These amplified fragments were digested with XbaI and EcoRI and subcloned into pET-T7. For the negative control, a 359-bp fragment of GFP was amplified from the pcDNA3.1/CT-GFP-TOPO® plasmid (using primers GFPiF and GFPiR; Table 1) and subcloned into pET-T7 between the XbaI and NcoI restriction sites. These recombinant plasmids were used to express dsRNA of GFP, Ar-Rsk1, and Ar-Rsk2. After transformation into Escherichia coli strain DH5α, the sequences of the inserted fragments in the plasmids were confirmed by DNA sequencing. The recombinant plasmids were transformed into the E. coli strain HT115, and dsRNAs were purified as described by Yodmuang et al. (37). dsRNA was microinjected into Artemia using the Ultra-MicroPump II equipped with a Micro4^TM MicroSyringe pump controller. One μg of dsRNA was injected per adult, prior to the formation of oocytes (34).

Southern Blot Analysis

Genomic DNA was isolated using Wizard Genomic DNA Purification Kit (Promega, Madison, WI). 25 μg of genomic DNA was digested with each of BamHI, EcoRI, KpnI or XbaI alone. Heat-inactivated samples were run on a 1% agarose/TAE gel. Blotting onto Immobilon-Ny⁺ (Millipore, catalog number INYC00010) was achieved using capillary action (35). DIG Easy Hyb Granules (catalog number 11796895001, Roche Applied Science) were reconstituted by adding water for both pre-hybridization and hybridization buffers. A hybridization probe, covering the common fragments of Ar-Rsk1 and Ar-Rsk2 (1696–1935 bp of Ar-Rsk1), was synthesized. Finally, the blot was exposed to x-ray films and developed using CSPD (catalog number 11755633001, Roche Applied Science).

Western Blotting Analysis

Proteins were extracted using TRIzol reagent (Invitrogen). From each sample, 25 mg of protein was separated by SDS-PAGE and transferred to a PVDF membrane (Millipore, Bedford, MA). Membranes were incubated with a primary antibody overnight at 4 °C and signals were detected using a BM Chemiluminescence Western blotting kit (Roche Applied Science). C-terminal fragments of Ar-Rsk1 (amino acids (aa) 701–714) and Ar-Rsk2 (aa 722–734) were chemically synthesized. Rabbits were immunized with these peptides to produce antibodies. Anti-cdc2 (ab47779), anti-phospho-cdc2 (Tyr¹⁵) (ab47594), and anti-phospho-RSK (Ser²²¹) (ab10695) antibodies were purchased from Abcam (Cambridge, UK). Anti-tubulin (T6199) (Sigma), anti-histone H3 (1326-1), anti-phospho-histone H3 (Ser¹⁰) (1173-1), anti-cyclin B1 (1495-1), anti-cyclin D3 (1846-1), and anti-phospho-RSK (Ser³⁸⁰) (1468-s) antibodies (Burlingame, CA) were purchased.

Co-immunoprecipitation

Samples were stirred in FA lysis buffer (50 mm HEPES, pH 7.5, 140 mm NaCl, 1 mm EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, protease inhibitors, and 40 units/ml of RNasin) on ice for 15 min. Lysates were incubated with Protein A Dynabeads for 1–2 h, and then clarified by centrifugation. Supernatants were incubated overnight at 4 °C with anti-Ar-Rsk1 or anti-Ar-Rsk2 antibodies or with chitin-binding protein antibody as a control. Beads were washed with extraction buffer twice and then mixed with 2× loading buffer (0.125 m Tris, pH 6.8, 4% SDS, 5% β-mercaptoethanol, 20% glycerol, and 0.02% bromphenol blue). Samples were boiled, separated by SDS-PAGE, and Western blotted.

Phosphorylation Assay of Ar-Rsk1 and Ar-Rsk2

Immunoprecipitation was performed using pan-antibodies that recognized both non-phosphorylated and phosphorylated forms of Ar-Rsk1 and Ar-Rsk2. Western blotting of these immunoprecipitates was performed using these pan-antibodies or with antibodies that specifically recognized phosphorylated forms of Ar-Rsk1 and Ar-Rsk2 (Ser²¹⁸ or Ser³⁸⁰). When a similar amount of total Ar-Rsk1 or Ar-Rsk2 was immunoprecipitated, the level of phosphorylated protein reflected the activity of the given isoform.

BrdU Incorporation Assay

Artemia were incubated in seawater containing 1 mm BrdU for 24 h, fixed with 4% (w/v) paraformaldehyde, and paraffin-embedded. Ten-μm thick tissue sections were incubated with a mouse monoclonal anti-BrdU antibody at 4 °C overnight, and then with alkaline phosphatase-conjugated anti-mouse IgG antibody. Staining was performed using nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate solution in the dark. The staining reaction was stopped by the addition of 10 mm TE buffer (10 mm Tris-Cl, pH 8.0, and 1 mm EDTA).

RESULTS

Characterization of Meiotic and Mitotic Arrest during Artemia Development

The oviparous and ovoviviparous life cycles of Artemia were controlled by modulating the different percentage of seawater and the daily light regime of the animals (36). In both pathways (Fig. 1A), oocytes appear in the ovary, gradually enlarge, and arrest in metaphase I in oviducts until fertilization occurs (Fig. 1A, lanes 1 and 2). Once fertilized, oocytes enter the ovisac, meiosis resumes, and embryo development proceeds (Fig. 1A, lane 3). In the oviparous pathway, fertilized embryos develop to the late gastrula stage in the uterus. Encysted gastrula embryos are released, which are in diapause and in which RNA/protein synthesis does not occur. When conditions are favorable, diapause embryos are activated and develop into post-diapause embryos, which can hatch and develop into nauplii and metanauplii (Fig. 1A, lanes 4–9). By contrast, in the oviviparous pathway, embryonic development is not interrupted and swimming nauplii are released directly (Fig. 1A, lanes 4′ and 5′).

FIGURE 1.

Development of Artemia and expression of meiotic and mitotic proteins at various stages. A, the two reproductive pathways of Artemia, the diapause-destined oviparous pathway (top) and the directly developing ovoviviparous pathway (bottom). Lanes 1, early oogenesis; 2, late oogenesis; 3, early embryos; 4, pre-diapause embryos; 5, diapause embryos; 6, post-diapause embryos; 7, 8 h-incubated embyos; 8, nauplii; 9, metanauplii; 4′, late embryos; 5′, nauplii. B, expression of the CSF components cdc2 and cyclin B1 at various stages of development. C, expression of mitosis markers at various stages of development. In B and C, the lane numbers correspond to the developmental stage shown in A. H3, histone H3; p-cdc2 (Y15), phospho-cdc2 (Tyr¹⁵); p-H3(S10), phospho-histone H3 (Ser¹⁰).

Meiotic arrest was examined by performing Western blot analysis of the two components of CSF, namely, cdc2 and cyclin B1 (Fig. 1B). Levels of total cdc2 and phospho-cdc2 (Tyr¹⁵) were high in late oogenesis, indicating that oocytes were arrested in meiosis. Following fertilization, levels of these were proteins decreased in early embryonic cells, indicating the resumption and completion of meiosis (Fig. 1B). Cyclin B1 was expressed during late oogenesis and in early embryonic cells.

Cell cycle progression during diapause formation was characterized by performing Western blot analysis of the mitosis markers cyclin D3 and phospho-histone H3 (Ser¹⁰) (Fig. 1C). The results indicated that mitosis occurred in pre-diapause embryos, but did not occur in diapause or post-diapause embryos. Mitosis eventually resumed following diapause termination, and embryos developed into nauplii and metanauplii. Taken together, these results show that meiosis was arrested in late oogenesis, whereas mitosis was arrested in diapause and post-diapause embryos.

Identification of Two RSK Isoforms in Artemia

We studied the functions of RSK isoforms in the regulation of meiotic and mitotic arrest in Artemia. We previously reported that an Ar-RSK isoform, termed Ar-Rsk1, plays an important role in diapause termination in Artemia (29). This isoform lacks a typical ERK-docking motif. In the current study, we identified another Ar-RSK isoform in Artemia, termed Ar-Rsk2, which has an ERK-docking motif. Sequence analysis revealed that the cDNA of Ar-Rsk2 contains an open reading frame that encodes a protein of 734 aa, with a predicted mass of 83.7 kDa (Fig. 2A). The deduced aa sequence of this protein is similar to those of other RSK family members. Ar-Rsk2 has a 98% sequence identity with Ar-Rsk1 (Fig. 2B), 59% with human and mouse Rsk1/2, 60% with human and mouse Rsk3, and 58% with human Rsk4. Phylogenetic analysis suggested that Ar-Rsk2 can be grouped with the RSK expressed in D. melanogaster (56% sequence identity) (Fig. 2D).

FIGURE 2.

Molecular characterization of Ar-Rsk1 and Ar-Rsk2. A, nucleotide and aa sequences of Ar-Rsk2. Numbers on the left indicate the nucleotide and aa positions. B, sequence alignment of Ar-Rsk1 and Ar-Rsk2. Numbers on the right indicate the aa position; underlined sequences indicate the CTKDs and NTKDs; asterisks indicate typical phosphorylation sites; the box indicates a typical ERK-docking motif. Accession number of Ar-Rsk1: KJ679442; accession number of Ar-Rsk2: KJ679441. C, structures of Ar-Rsk1 and Ar-Rsk2. D, phylogenetic analysis of RSK isoforms from Artemia with those from other species. E, Southern blot analysis of Ar-Rsk1 and Ar-Rsk2. Lane 1, genomic DNA was digested with BamHI alone; lane 2, with EcoRI alone; lane 3, with KpnI alone; lane 4, with XbaI alone.

To address the genomic aspects of Ar-Rsk1 and Ar-Rsk2, we enzyme digested the genomic DNA with BamHI, EcoRI, KpnI, and XbaI and then incubated the enzyme-digested product with a probe covering the common fragments of Ar-Rsk1 and Ar-Rsk2. The results of Southern blot showed that only one band in each enzyme lane had been hybridized by the probe (Fig. 2E). This indicated that Ar-Rsk1 and Ar-Rsk2 are spliced from one unique single-copy gene.

The deduced aa sequence showed that Ar-Rsk2 has two distinct kinase domains, NTKD and CTKD, which are joined by a linker region (Fig. 2C). This is identical to the structure of Ar-Rsk1. Ar-Rsk2 contains five phosphorylation sites (Ser²¹⁸, Thr³⁵⁹, Ser³⁶³, Ser³⁸⁰, and Thr⁵⁷¹), which were identified according to RSK consensus phosphorylation sites (9). However, aa 720–724 of Ar-Rsk2 (Leu-Ala-Lys-Arg-Arg) constitute an ERK-docking site (14), which Ar-Rsk1 lacks (Fig. 2, B and C).

Expression Pattern of Ar-Rsk1 and Ar-Rsk2 during Diapause Formation and Termination

Semi-quantitative RT-PCR showed that Ar-Rsk1 was stably expressed throughout Artemia development (Fig. 3, A and B). The level of Ar-Rsk2 mRNA was high in early embryos, nauplii, and metanauplii (Fig. 3, A and B), in which cells are dividing. However, expression of Ar-Rsk2 was suppressed in diapause and post-diapause embryos (Fig. 3B), in which cells are in mitotic arrest.

FIGURE 3.

Expression and activation of Ar-Rsk1 and Ar-Rsk2 during Artemia development. A and B, semi-quantitative RT-PCR analysis of Ar-Rsk1 and Ar-Rsk2 expression. C, the specific protein expression of Ar-Rsk1 and Ar-Rsk2 each in cephalothorax, abdomen, or embryo during the development of Artemia. D and E, Western blot analysis of Ar-Rsk1 and Ar-Rsk2. F, the specific phosphorylation levels of Ar-Rsk1 each in cephalothorax, abdomen, or embryos. G and H, levels of phosphorylated Ar-Rsk1 during Artemia development. I, the specific phosphorylation level of Ar-Rsk2 each in cephalothorax, abdomen, or embryos. J and K, levels of phosphorylated Ar-Rsk2 during Artemia development. Lane 1, early oogenesis; lane 2, late oogenesis; lane 3, early embryos; lane 4, pre-diapause embryos; lane 5, diapause embryos; lane 6, post-diapause embryos; lane 7, 8-h incubated embyos; lane 8, nauplii; lane 9, metanauplii. P, positive control (protein sample from late oogenesis). ce, cephalothorax; ab, abdomen; em, embryos.

Using polyclonal antibodies specific to Ar-Rsk1 or Ar-Rsk2, the bands of 81 and 83 kDa were identified in Artemia, respectively. The results of the Western blot indicated that Ar-Rsk1 was expressed in all tested tissues, with the exception of those of the cephalothorax and embryos during embryonic development. The expression of Ar-Rsk2 was specific to the oocyte-containing or embryo-containing abdomen of Artemia (Fig. 3C). The level of Ar-Rsk1 expression remained constant throughout development (Fig. 3, D and E), however, Ar-Rsk2 expression was reduced in pre-diapause embryos, and was suppressed in diapause and post-diapause embryos. Thereafter, Ar-Rsk2 expression gradually increased and was highest in nauplii, in which cells are dividing (Fig. 3, D and E).

The consensus phosphorylation sites of Ar-Rsk1 and Ar-Rsk2 were mapped (Fig. 2B). Immunoprecipitations were performed using polyclonal antibodies that specifically recognized Ar-Rsk1 or Ar-Rsk2. Western blotting of these immunoprecipitates was performed using antibodies that specifically recognized Ar-Rsk1 and Ar-Rsk2 phosphorylated at Ser²¹⁸ (Ser²²¹ in human Rsk1) and Ser³⁸⁰. The results indicated that Ar-Rsk1 was activated in oocyte-containing abdomen during oogenesis, in which meiosis was blocked by the activity of CSF (Fig. 3, F and G). Following fertilization, the level of phosphorylated Ar-Rsk1 decreased, indicating the resumption and completion of meiosis (Fig. 3G). Ar-Rsk1 was not activated during embryonic development (Fig. 3H). These results indicate that Ar-Rsk1 is involved in the initiation and maintenance of meiotic arrest.

Ar-Rsk2 was activated in the oocyte-containing and embryo-containing abdomen (Fig. 3I). The level of phosphorylated Ar-Rsk2 was higher in early embryos than in oocytes (Fig. 3J). The level of phosphorylated Ar-Rsk2 was suppressed in diapause and post-diapause embryos and resumed after diapause termination (Fig. 3K). These results indicate that Ar-Rsk2 is involved in mitotic arrest during early embryonic development.

Ar-Rsk1 Controls Meiotic Arrest during Oogenesis in Artemia

To further study the functions of Ar-RSKs in meiotic arrest, Ar-Rsk1 or Ar-Rsk2 was knocked down in immature Artemia using RNAi. Protein levels of Ar-Rsk1 and Ar-Rsk2 were reduced by more than 90% following injection of Ar-Rsk1 dsRNA or Ar-Rsk2 dsRNA, respectively (Fig. 4A). The signals of Ar-Rsk1 in the Ar-Rsk2 RNAi and Ar-Rsk2 in the Ar-Rsk1 RNAi were detected to demonstrate the knockdown specificity (Fig. 4B). The protein levels of cyclin B1 and total cdc2 did not change during late oogenesis following knockdown of Ar-Rsk1, whereas the level of phospho-cdc2 (Tyr¹⁵) was clearly decreased (Fig. 4C). Phosphorylation and activation of cdc2 are necessary to form the active cdc2-cyclin B1 complex. These results indicate that phosphorylated Ar-Rsk1 is involved in the formation and activation of this complex, which arrests oocytes in meiosis. By contrast, the levels of total cdc2, phospho-cdc2 (Tyr¹⁵), and cyclin B1 did not change during late oogenesis following knockdown of Ar-Rsk2 (Fig. 4D), indicating that Ar-Rsk2 is not involved in the initiation or maintenance of meiotic arrest.

FIGURE 4.

Effects of knockdown of Ar-Rsk1 and Ar-Rsk2 on reproduction in Artemia. A, knockdown efficiency following injection of GFP dsRNA (GFPi), Ar-Rsk1 dsRNA (Ar-Rsk1i), or Ar-Rsk2 dsRNA (Ar-Rsk2i). B, knockdown specificity. Detection of the signal of Ar-Rsk1 in the Ar-Rsk2 RNAi or Ar-Rsk2 in the Ar-Rsk1 RNAi. C, Western blot analysis of cdc2, phospho-cdc2 (Tyr¹⁵), and cyclin B1 following GFPi and Ar-Rsk1i treatment. D, Western blot analysis of cdc2, phospho-cdc2 (Tyr¹⁵), and cyclin B1 following GFPi or Ar-Rsk2i treatment. E, offspring produced via the directly developing (ovoviviparous) and diapause-destined (oviparous) pathways following GFPi, Ar-Rsk1i, or Ar-Rsk2i treatments. Artemia that directly developed following GFPi (a), Ar-Rsk1i (b), or Ar-Rsk2i (c) treatment are indicated. Diapause-destined Artemia embryos following GFPi (a′), Ar-Rsk1i (b′), or Ar-Rsk2i (c′) treatment. EO, early oogenesis; LO, late oogenesis; EE, early embryos. p-cdc2 (Y15), phospho-cdc2 (Tyr¹⁵). co, coxopodite; ba, basipodite; en, endopodite; ex, exopodite.

Following knockdown of Ar-Rsk1, typically malformed nauplii were produced with various uromere deformities and shortened swimming setae (Fig. 4E, b). Nauplii produced by GFP dsRNA-injected Artemia did not exhibit these malformations (Fig. 4E, a). These malformed nauplius could not swim normally and most died within 1 week. Individuals that had been injected with Ar-Rsk2 dsRNA released pseudo-diapause cysts (Fig. 4E, c). These did not develop into nauplii. Taken together, these data indicate that Ar-Rsk1 is important for oogenesis and the development of embryos after fertilization, whereas Ar-Rsk2 is involved in the development of embryos.

Ar-Rsk2 Controls Mitotic Arrest during Embryo Development in Artemia

To investigate the functions of Ar-Rsk1 and Ar-Rsk2 in the regulation of mitotic arrest during the development of Artemia, Ar-Rsk1 dsRNA or Ar-Rsk2 dsRNA was injected during the early stages of embryo development, after meiosis was completed. Ar-Rsk2 was weakly expressed in diapause and post-diapause embryos; therefore, we performed RNAi in Artemia that were following the ovoviviparous (directly developing) pathway.

Western blot analysis showed that the protein levels of Ar-Rsk1 were reduced by more than 80% and those of Ar-Rsk2 were reduced by more than 90% (Fig. 5A) following injection of Ar-Rsk1 or Ar-Rsk2 dsRNA, respectively. The expression of neither Ar-Rsk1 in the Ar-Rsk2 RNAi nor Ar-Rsk2 in the Ar-Rsk1 RNAi was affected following knockdown (Fig. 5B). Up to 90% of individuals injected with Ar-Rsk2 dsRNA released pseudo-diapause cysts, whereas individuals injected with GFP or Ar-Rsk1 dsRNA released nauplii (Fig. 5C). Furthermore, the BrdU assay showed that following Ar-Rsk2 knockdown, cells did not divide and embryos did not develop (Fig. 5D). By contrast, following injection of GFP dsRNA or Ar-Rsk1 dsRNA, cell division occurred normally and embryos developed to the blastula stage (Fig. 5D).

FIGURE 5.

Effect of knockdown of Ar-Rsk1 and Ar-Rsk2 on embryonic development of Artemia. A, knockdown efficiency following injection of GFP dsRNA (GFPi), Ar-Rsk1 dsRNA (Ar-Rsk1i), or Ar-Rsk2 dsRNA (Ar-Rsk2i). B, knockdown specificity. C, offspring produced following GFPi (a), Ar-Rsk1i (b), or Ar-Rsk2i (c) treatments. D, BrdU assay of knockdown of Ar-Rsk1 and Ar-Rsk2 on embryonic development of Artemia. BrdU(+) indicates labeling with an anti-BrdU antibody. BrdU(−) indicates labeling in which the anti-BrdU antibody was omitted. E, Western blot analysis of cyclin D3 and phospho-histone H3 (Ser¹⁰) following GFPi or Ar-Rsk1i treatment. F, Western blot analysis of cyclin D3 and phospho-histone H3 (Ser¹⁰) following GFPi or Ar-Rsk2i treatment. LO, late oogenesis. EE, early embryos; LE, late embryos. H3, histone H3; p-H3 (S10), phospho-histone H3 (Ser¹⁰).

Western blot analysis showed that levels of cyclin D3 and phospho-histone H3 (Ser-10) were decreased following knockdown of Ar-Rsk2 (Fig. 5F), whereas they were not changed following knockdown of Ar-Rsk1 (Fig. 5E). These results indicate that Ar-Rsk2 is involved in regulation of mitosis and plays important roles in diapause formation in Artemia. In summary, Ar-Rsk2 functions in diapause formation by regulating mitotic arrest, whereas Ar-Rsk1 functions in oocyte development by regulating meiotic arrest.

DISCUSSION

Artemia is a useful model in which to study cell cycle arrest because cells arrest in meiosis and mitosis during oogenesis and embryonic development, respectively. In the present study, we found that Ar-Rsk1 and Ar-Rsk2 have different expression patterns and levels of phosphorylation during development, and play markedly different roles in meiotic and mitotic arrest.

RSK isoforms are emerging as multifunctional effecters that have distinct functions. Ectopic expression of Rsk2 increases the proliferation rate and anchorage-independent transformation of mouse embryonic fibroblasts (38). Rsk1 and Rsk2 may also promote G₁ phase progression by controlling the activity of p27^kip1, a cyclin-dependent kinase 2 inhibitor (39). However, Rsk3 was recently shown to act as a tumor suppressor in ovarian cancer (40), and several lines of evidence suggest that Rsk4 negatively regulates cell proliferation (41). The current study shows that RSK isoforms also have distinct functions in Artemia; Ar-Rsk1 regulates meiosis, whereas Ar-Rsk2 regulates mitosis.

RSK isoforms interact with ERK through their C-terminal docking domains, which have a ERK-docking motif consensus sequence (42). This motif corresponds to a hydrophobic residue closely followed by two lysine or arginine residues, which are positively charged (Leu-Xaa²-Lys/Arg-Lys/Arg-Xaa⁵-Leu). Although this motif is not conserved in RSK orthologues from Drosophila or C. elegans, it was previously thought to be the domain that is required first during the hierarchical phosphorylation of RSK by ERK (43).

In the present study, Ar-Rsk2 was found to contain a typical ERK-docking motif and Ar-Rsk1 was not. To identify the signals triggering the activation of these RSKs, knockdown of PLK and ERK was performed (Fig. 6A). The results of ERK RNAi showed that the phosphorylation of Ar-Rsk2 at Ser²¹⁸ and Ser³⁸⁰ had decreased, whereas the phosphorylation of Ar-Rsk1 remained unaffected (Fig. 6B, a). PLK RNAi induced activation of Ar-Rsk2 by phosphorylation at Ser²¹⁸ and Ser³⁸⁰ but had no effect on Ar-Rsk1 activation (Fig. 6B, b). This result indicated that only Ar-Rsk2, but not Ar-Rsk1, could be activated by the ERK pathway in Artemia and this activation was dependent on the ERK-docking motif near the C terminus of Ar-Rsk2.

FIGURE 6.

The identification of the signals triggering Ar-Rsk1 and Ar-Rsk2 activation. A, the expression of Ar-Rsk1 and Ar-Rsk2 following injection of GFP dsRNA (GFPi), ERK dsRNA (ERKi), or PLK dsRNA (PLKi). B, the phosphorylation level of Ar-Rsk1 and Ar-Rsk2 following injection of GFP dsRNA (GFPi), ERK dsRNA (ERKi), or PLK dsRNA (PLKi). a, the phosphorylation level of Ar-Rsk1 and Ar-Rsk2 following ERKi. b, the phosphorylation level of Ar-Rsk1 and Ar-Rsk2 following PLKi.

Like the case of Ar-Rsk1 in Artemia, five human RSK δ-1 proteins (NP_036556, NP_001129610, NP_001274148, NP_001274149, and NP_001274150) and a Xenopus RSK δ-1 protein (XP_002934734) all belong to the RSK family, but lack a typical ERK-docking domain. Based on our results, these RSKs could not be activated by PLK and ERK pathways, however, the mechanism of activation of the RSKs without the ERK-docking motif remains large unclear. In the present study, the function of Ar-Rsk1 was shown to control meiotic arrest during oogenesis. Thus, the biological function of these RSKs from other species without the ERK-docking motif is considered involved in the regulation of meiosis during oogenesis.

RSKs are involved in the regulation of cell proliferation in various malignancies and genetic diseases by directly or indirectly modulating the cell cycle machinery (44–46). We report that Ar-Rsk2 promotes cell proliferation during early embryonic development in Artemia by regulating the cell cycle. These findings are consistent with reports that all RSK isoforms (except Rsk4) are ubiquitously expressed at the mRNA level in regions that contain rapidly proliferating cells in mice and humans (14).

Artemia that developed via the oviviparous pathway were extremely sensitive to knockdown of Ar-Rsk2 on mitosis at early stages of development. These effects were more pronounced than those reported by Smith et al. (1) in the human breast cancer cell line MCF-7; proliferation of these cells is suppressed following treatment with an inhibitor of RSK function or knockdown of RSK expression. We propose the novel hypothesis that RSK isoforms specifically regulate mitosis in early embryonic cells and in other distinct types of cells, such as tumor-initiating cells. This may explain why triple-negative breast cancers, which contain a higher proportion of tumor-initiating cells than other breast cancer subtypes, are particularly sensitive to RSK inhibition (47). However, this hypothesis needs to be investigated further.

Extensive study of meiotic arrest during oocyte maturation revealed that the MAPK-RSK pathway is essential for CSF activity (8, 48). Little is known about CSF-dependent meiotic arrest in invertebrate species (49, 50). In Artemia, oocytes are arrested in metaphase I until fertilization occurs (22). Ar-Rsk1 was phosphorylated and active when CSF activity was high, whereas it was dephosphorylated when meiosis resumed (Fig. 3). Knockdown of Ar-Rsk1 resulted in dephosphorylation of cdc2 in oocytes and termination of meiotic arrest (Fig. 5). Most of these oocytes completed meiosis, and embryo development proceeded normally to produce nauplii. However, some of these nauplii exhibited malformations, such as shortened swimming setae and a jagged uromere. These malformations reflect the biological importance of meiotic arrest in oocytes; this ensures that oocytes are released into the ovisac (uterus) in the optimal physiological state, which prevents insemination failing or occurring abnormally. Our findings show that the Ar-Rsk1 pathway participates in maintaining the CSF-dependent meiotic arrest in Artemia, which is required for oocyte maturation. These results highlight the complexity of meiotic regulation, which has evolved to ensure that reproduction is precisely controlled.

RSK isoforms control cell proliferation by regulating mediators of the cell cycle (51, 52). RSK with ERK-docking motif phosphorylates serum response factor and contributes to transcriptional activation of c-Fos. c-Fos protein is stabilized by ERK and RSK phosphorylation and functions as part of the c-Fos-JUN activator protein-1 complex to activate expression of cyclin D, which contributes to G₁/S phase progression (51). In meiosis, RSK without ERK-docking motif phosphorylates and inactivates Myt1, a membrane-associated kinase that phosphorylates cdc2 on Tyr and Thr residues (52). This inhibits cdc2, leading to cell cycle progression. This is the first study to report that RSK isoforms with and without an ERK-docking motif regulate mitosis and meiosis, respectively. This study provides insight into the relationship between the structures and functions of RSK isoforms.