Protein SYCP2 Is an Ancient Component of the Metazoan Synaptonemal Complex

Meiosis is a type of cell division that leads to the formation of haploid cells. It is divided into 2 sequent divisions: meiosis I and meiosis II. During the first meiotic division, the chromosome set is reduced to half by the separation of the homologous chromosomes. The sister chromatids are not separated until the second division, which is very similar to mitosis, finally providing 4 haploid germ cells from 1 diploid precursor.

Three meiosis-specific features need to interplay during the first division to achieve a successful chromosome reduction: (a) the cohesion of chromatids, which is protected at the centromeres during metaphase I to prevent a premature separation of the sister chromatids; (b) the recombination of homologous chromosomes that provides interconnection (chiasmata) of the homologous chromosomes, and (c) a stable synapsis of the homologous chromosomes, which is mediated by the synaptonemal complex (SC) [Page and Hawley, 2004].

In the electron microscope, a typical SC appears as a ladder-like structure interconnecting the homologs over their entire length [Fraune et al., 2012a]. Its assembly starts during leptonema with the formation of the axial elements. As meiosis progresses, during zygonema, axial elements (now called lateral elements, LEs) recruit numerous transverse filaments (TFs), which in a zipper-like manner overlap and interact with TFs from the opposing LEs, hence initiating the formation of the central element (CE), which stabilizes the interactions of the TFs. SCs are fully assembled in pachynema and become disassembled during diplonema of the first meiotic prophase [Zickler and Kleckner, 1999; Page and Hawley, 2003; Fraune et al., 2012a].

The tripartite structure of the SC (LEs, TFs, CE) is highly conserved in sexually reproducing organisms [Gillies, 1975; von Wettstein et al., 1984]. Very recently, we demonstrated that mammalian SC proteins derived from an ancient structure that was present in the ancestor of metazoans [Fraune et al., 2012b, 2013].Noteworthy, the ancient origin of SC proteins was unexpected at that time, as the mammalian SC proteins do not share any detectable sequence homologies with the characterized SC proteins of the other metazoan meiosis model organisms, i.e. Drosophila melanogaster and Caenorhabditis elegans [Fraune et al., 2012b, 2013].An in-depth phylogenetic analysis, however, finally disclosed the conservation of most characterized mouse SC proteins from basal branching metazoans to mammals [Fraune et al., 2012b, 2013]. Our previous studies demonstrated that the major structural SC proteins SYCP1, the protein of the TFs [Meuwissen et al., 1992], and SYCP3, a component of the LEs [Lammers et al., 1994], as well as the 2 CE proteins SYCE2 and TEX12 [Costa et al., 2005; Hamer et al., 2006] are ancient in Metazoa. Orthologs of all 4 proteins were found in several animal lineages, even down to the basal species Hydra, and were experimentally identified as protein components of the cnidarian SC. The deduced evolutionary sequence of SC formation further suggested that the mammalian components SYCE1 and SYCE3 - both specific components of the CE [Costa et al., 2005; Schramm et al., 2011] - newly emerged in the branches leading to Bilateria (SYCE1) and vertebrates (SYCE3), respectively. It appears likely that during the divergence of the ecdysozoan species strong diversification of the existing SC proteins might have occurred, which could explain the lack of sequence homology between the mammalian SC proteins and the SC components of D. melanogaster and C. elegans[Fraune et al., 2012b, 2013].

With SYCP2, a second major protein constituent of the LEs was identified in mammals, besides SYCP3 [Offenberg et al., 1998]. SYCP2 is the largest SC protein yet described and in the mouse consists of 1,500 amino acids. It appears to be involved in LE assembly as well as linking LEs and TFs [Offenberg et al., 1998; Yang et al., 2006; Winkel et al., 2009]. In contrast to the other mammalian SC proteins, the origin of mammalian SYCP2 has not been considered in previous analyses. Here, we investigated this aspect taking advantage of the vastly increasing number of genomic data referring to numerous different clades and species. Complementing bioinformatic results, we also performed expression analysis at the RNA and protein level.

Sequence analysis and assembly of the dataset were performed as described previously [Fraune et al., 2013]. Homologs of SYCP2 were retrieved from public databases at NCBI (http://blast.ncbi.nlm.nih.gov) and Ensembl (http://www.ensembl.org/index.html) with BLASTp and tBLASTn [Altschul et al., 1997] using the mouse SYCP2 (RefSeq: NP_796165.2) as query. Newly detected homologs were used as seeds for additional BLAST search attempts to detect more distantly related sequences as well. The identified sequences were used to build a specific HMM profile employing the HMMER 3.0 webserver (http://hmmer.janelia.org). The resulting profile was used to query the non-redundant database (hmmsearch option). The survey of public sequence databases allowed us to retrieve SYCP2 homologs. The corresponding protein sequences were aligned using MAFFT version 7 [Katoh and Standley, 2013] with the linsi option.

Sequence information of the Hydra vulgaris AEP transcriptome was obtained from the public database on the Compagen server [Hemmrich et al., 2012]. Annotated sequence alignments were designed using CHROMA version 1.0 [Goodstadt and Ponting, 2001]. The identity threshold for grouping of the residues was set to 60%. Seven groups were created, depending on different features of the amino acids: identical (*), charged (:), Ser/Thr (:), aliphatic (:), aromatic (:), polar (:), and hydrophobic (:).

H. vulgaris from the strain AEP [Martin et al., 1997] were cultured at 18°C following standard procedures [Lenhoff and Brown, 1970]. Testes formation was induced by feeding the animals daily for at least 1 week, then starving them for up to 5 days [Wittlieb et al., 2006].

RNA isolation was carried out as described previously [Fraune et al., 2013]. Synthesis of a HySycp2 cDNA fragment was performed by PCR using Phusion® DNA polymerase (Thermo Scientific, St. Leon-Rot, Germany) and the following primer pair: HySycp2_part_5′ ATGGTTTCAATTGATGACCG and HySycp2_shortpart_3′ CATAGAATCATCTTGGACATAAGTTG (60°C annealing temperature). The amplified fragment was cloned into the pSC-B-amp/kan PCR cloning vector (Agilent Technologies, Böblingen, Germany) and sequenced by GATC (Konstanz, Germany).

In situ hybridization was performed on whole animals as reported in earlier studies [Grens et al., 1996; Fraune et al., 2012b]. The antisense RNA probe of 297 bp used in the current study was synthesized from the cloned cDNA of HySycp2.

The cDNA sequence coding for a HySYCP2 polypeptide with 99 amino acids (see above) was cloned into pET21a vector (Novagen, Darmstadt, Germany) in frame to the sequence coding for a C-terminal His-tag. Following bacterial expression, the His-tagged polypeptide was purified using a nickel-nitrilotriacetic acid agarose matrix (Qiagen, Hilden, Germany). The purified peptide was used to generate anti-HySYCP2 antibodies in rabbit and guinea pig, respectively. Immunization was conducted by Seqlab (Göttingen, Germany). The final bleedings were affinity purified with the HiTrap system according to the manufacturer's protocol (GE Healthcare, Munich, Germany). Rabbit anti-HySYCP1 and anti-HySYCP3 antibodies were described elsewhere [Fraune et al., 2012b]. Anti-actin antibody (A4700) was purchased from Sigma (Steinheim, Germany).

For immunoblot analysis, protein probes of different fractions of Hydra(head, mid-piece, foot, and testis) were separated on an 8% (v/v) acrylamide gel and transferred to a nitrocellulose membrane at 25 V for 16 h at 4°C using a wet/tank blotting system (Bio-Rad Laboratories, Munich). Detection of the HySYCP2 protein was done according to previous protocols [Fraune et al., 2012b] with minor changes. The nitrocellulose membrane was blocked overnight in PBS containing 0.2% Tween-20 (PBST) and 5% milk at 4°C. Afterwards, the membrane was incubated with rabbit anti-HySYCP2 antibodies (1:20,000) or mouse anti-actin antibody (1:10,000) diluted in PBST containing milk for 16 and 2 h, respectively. Secondary antibodies (goat anti-rabbit or anti-mouse peroxidase conjugate; Dianova, Hamburg, Germany) were used at a dilution of 1:10,000 in PBST with 5% milk for 1 h at room temperature. All washing steps were performed using PBST.

Immunofluorescence microscopy on cryosections of Hydra testis and on chromosome spreads [de Boer et al., 2009] of Hydraspermatocytes was carried out following the protocols in Fraune et al. [2012b]. The affinity-purified guinea pig anti-HySYCP2 antibodies were used at a dilution of 1:200.

Confocal images were taken using a Leica TCS-SP2 confocal laser-scanning microscope (Leica, Wetzlar) equipped with a 63×/1.40 HCX PL APO lbd.BL oil immersion objective. Immunofluorescence images in the current study represent 2D projections from series of ∼20 optical sections per cell, generated by the maximum projection algorithm (Leica). The images were pseudocolored using the Leica TCS-SP2 software. Final processing of the digital images was done with Adobe Photoshop CS5 (Adobe Systems). Imaging of whole-mount in situ hybridization was performed using the Binocular SZ 61 (Olympus, Hamburg), equipped with the EC3 camera (Leica).

An intensive database survey identified potential homologs of the mammalian SYCP2. Significant hits were detected in major metazoan phyla including Placozoa, Cnidaria, Crustacea, Plathelminthes, Annelida, Mollusca, Echinodermata, and several vertebrate classes. Most prominent sequence similarities were found in regions corresponding to the N-terminal part of the mouse SYCP2 protein (GenBank Acc. No. NP_796165). Accurate sequence comparison further revealed a region of 108 amino acids (amino acids 211-318 of the mouse SYCP2), which is highly conserved even between distant metazoan clades as exemplified by the CHROMA alignment of a representative taxonomic sample of SYCP2 homologs (fig. 1). The broad taxonomic distribution of SYCP2 homologs in Metazoa suggested that this protein is ancient in Metazoa. Then, the question is whether this ancestral SYCP2 was part of a SC in the ancestor of Metazoa or if SYCP2 was recruited secondarily in the mammalian SC. To address this question, we investigated the function of the SYCP2 homolog found in an early diverging metazoan lineage, the cnidarian species H. vulgaris (GenBank Acc. No. XP_004210390).

To answer the question on the identity and possible function of the putative Hydra SYCP2 homolog, we performed a detailed expression analysis of the identified sequence in H. vulgarisAEP at the mRNA and protein level. Using the detected H. vulgaris protein sequence as seed, we searched for H. vulgarisAEP transcriptome data [Hemmrich and Bosch, 2008; Hemmrich et al., 2012] coding for SYCP2-related sequences. We identified a related contig (HAEP_T-CDS_v02_43970) coding for a 99- amino-acid-long peptide, which aligns to the H. vulgaris sequence with a sequence identity of 96%. Using sequence-specific primers, we amplified a 297-bp-long cDNA sequence corresponding to part of the coding sequence of the putative H. vulgaris AEP SYCP2 gene (for primer sequence, see Material and Methods). This cDNA fragment was used as a template to synthesize a specific antisense RNA probe for whole-mount in situ hybridization. The analysis yielded a strong signal in the basal layer of the testes, which mainly harbors meiotic cell types (fig. 2A) [Kuznetsov et al., 2001; Fraune et al., 2014]. No additional signals were detected in the Hydra body. A comparison of the obtained staining pattern with those of previous in situ hybridizations against mRNAs encoding already characterized Hydra SC proteins demonstrated a remarkable similarity [Fraune et al., 2012b, 2013]. The testis-specific expression was confirmed by RT-PCR. For this experiment, 4 different tissue fractions (i.e. head, mid-piece, foot, and testis) were prepared [Fraune et al., 2012b]. Under these experimental conditions, a signal was exclusively detected in the testis fraction, revealing a PCR product of ∼300 bp (fig. 2B). The size of the band was as expected since the same primer pair as for cDNA cloning (297 bp amplified) was used.

Next, the expression of the putative Hydra SYCP2 homolog was characterized at the protein level. To this end, we raised antibodies against the corresponding protein fragment of Hydra (99 amino acids encoded by the amplified cDNA described above). The Western blot analysis yielded a prominent signal, which exclusively appeared in the testis lane. Hydra head, mid-piece and foot protein fractions were negative. The testicular protein recognized by the antibody has a calculated mass of 150 kDa (fig. 2C). The size of this protein band is in good agreement with that of mouse SYCP2, which comprises 1,500 amino acids.

Finally, the antibodies were used to localize the protein by immunofluorescence microscopy. On cryosections from Hydra testes, the antibody recognized thread-like nuclear structures in spermatocytes that correspond to SCs (fig. 3, insets). The signal was specific for spermatocytes as no signal was observed within any other cell type. As in previous studies, double-label immunofluorescence microscopy on chromosome spreads at different meiotic stages was used to demonstrate SYCP2 localization (fig. 4) [Fraune et al., 2012b, 2013]. As already described in other species, HySYCP2 colocalizes with the LE protein HySYCP3 but not with the central region (CR)-protein HySYCP1. This is particularly evident during diplonema (fig. 4B, C) when LEs are still present but CR disassembly has already started. As expected for a LE-specific protein, HySYCP2 antibodies label LEs even at the sites lacking a CR (fig. 4C, insets).

The protein investigated here is a component of SC lateral elements of the basal metazoan Hydra. We identified it as a homolog of mouse SYCP2, which is why we named it HySYCP2. Therefore, we conclude that protein SYCP2 belongs to the group of ancient SC components, which together with SYCP1, SYCP3, SYCE2, and TEX12 [Fraune et al., 2013] were already present in the last common ancestor of Metazoa over 500 million years ago and gave rise to modern SCs in present-day animals.

We thank Thomas Bosch (University of Kiel) for generous supply of Hydra cultures. This study was supported by the DFG Priority Program (SPP1348): Mechanisms of genome haploidization and the French National Agency for Research (ANR-10-BINF- 01-01). C.B.-A. is a member of the Institut Universitaire de France.

J.F. and M.A. contributed equally to this work.