Characterization of Structure and Expression of the Dzip1 Gene in the Rat and Mousei

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Genomics 87 (2006) 275 – 285 www.elsevier.com/locate/ygeno

Characterization of structure and expression of the Dzip1 gene in the rat and mousei

Benjamin J. Curry a, Janet E. Holt a, Eileen A. McLaughlin a,b, R. John Aitken a,b,*

a Reproductive Science Group, Discipline of Biological Science, School of Environmental and Life Sciences, University of Newcastle, Callaghan, NSW 2308, Australia b ARC Centre of Excellence for Biotechnology and Development, Discipline of Biological Science, School of Environmental and Life Sciences, University of Newcastle, Callaghan, NSW 2308, Australia

Received 14 March 2005; accepted 21 October 2005 Available online 20 December 2005

Abstract

A transcript encoding a rat homologue of DZIP1 (DAZ-interacting protein) was isolated from testis RNA. Like human DZIP1, it contains a C2H2 zinc finger domain. A predicted mouse homologue of DZIP1 was found in the GenBank database. Genome analysis indicated that while DZIP1 and mouse Dzip1 contain 22 and 20 exons, respectively, the rat sequence was intronless, confirmed by PCR on genomic DNA. This rat Dzip1 sequence is homologous to mouse Dzip1 exons 1–6 and DZIP1 exons 5–9. As this rat sequence was shorter than DZIP1 it was designated rat Dzip1S. The rat genome also contained a further predicted homologue of DZIP1 displaying conserved linkage homology with mouse Dzip1 and DZIP1. This sequence, if expressed, is the true rat homologue of DZIP1, designated rat Dzip1. Rat Dzip1S mRNA was present in all tissues examined by qualitative RT-RCR, and in situ hybridization of rat testis confirmed that expression of rat Dzip1S mRNA was confined to the spermatogenic lineage, specifically premeiotic spermatogonia. D 2005 Elsevier Inc. All rights reserved.

Keywords: DZIP1; DAZ1; DAZL; Testes; Male reproductive tract

In mammals, the Y chromosome is required for the There are two other members of the DAZ1 gene family. There formation of the testis and is necessary for spermatogenesis. is an autosomal homologue of DAZ1 named DAZL (DAZ-like), Evidence has accumulated over recent years that suggests the which maps to human chromosome 3 and is expressed only in DAZ1 (Deleted in Azoospermia) genes located on the Y male and female germ cells [7–10]. DAZL encodes an RNA- chromosome function early in germ-line stem cells and, in the binding protein with a single RNA-binding motif and a single human, are essential for spermatogenesis. Thus, men with DAZ repeat [10]. In mice, disruption of DAZL interferes with deleted DAZ1 genes have defects in spermatogenesis, with a male and female germ cell development [11,12]. The third characteristic phenotype involving absence of all germ cells, member of the DAZ1 family is BOLL,ahomologueof including spermatogonial stem cells [1–4]. There is an array of Drosophila Boule that also maps to human chromosome 2 four DAZ1 genes in two clusters located on the Y chromosome [13]. BOLL is a closer homologue to Drosophila Boule than to [5], and they encode RNA-binding proteins with a common either human DAZ1 or human DAZL, but like DAZL it contains a RNA-recognition motif and a series of 8–18 DAZ repeats [1]. single RNA-binding domain and a single DAZ repeat [13]. These genes are expressed in fetal gonocytes and spermatogo- BOLL is not expressed in primordial germ cells but is first nial stem cells [6]. detectable in spermatocytes and persists through meiosis to round spermatids [13]. Such results are consistent with BOLL as i Sequence data from this article have been deposited with the GenBank Data a meiotic regulator, as first proposed for the Drosophila gene Library under Accession No. AY847187. [14]. Through phylogenetic analysis, it is proposed that Boule * Corresponding author. Reproductive Science Group, Discipline of Biolog- ical Science, School of Environmental and Life Sciences, University of was the original DAZ1 family member, DAZL arose through Newcastle, Callaghan, NSW 2308, Australia. Fax: +61 2 49216308. gene duplication after vertebrate evolution, while a further E-mail address: [email protected] (R.J. Aitken). duplication resulted in Y-linked DAZ1 [13].

To ascertain the function of DAZ gene family members, Accession No. AB023213) [25], with an identity of 61%. This studies have been performed to identify binding partners for has subsequently been shown to be DZIP1 (GenBank these genes. Drosophila Boule regulates the expression of Accession No. NM_014934) [19,24]. The full-length sequence twine, a cdc25 phosphatase required for spermatocytes to in the current study was determined by 3V- and 5V-rapid complete meiosis [15], and recent evidence points to a similar amplification of cDNA ends (RACE) on rat testis total RNA. role for BOLL [16,17]. DAZ1, DAZL, and BOLL have been For 3V-RACE, nested PCR was necessary to produce a specific found to interact with PUM2 [18,19], a human homologue of band of 1.3 kb (Fig. 1A). For 5V-RACE, a primary reaction Drosophila pumilio encoding an RNA-binding protein that gave a specific band of 850 bp (Fig. 1B). These bands were interacts with other cofactors at the 3VUTR (untranslated sequenced and collated to produce a sequence of 2057 bp with region) of specific transcripts to regulate their translation an open reading frame of 332 amino acids. This was given [20–22]. DAZ1 and DAZL also interact through the DAZ GenBank Accession No. AY847187 and the gene was repeats with human DAZAP1 (DAZ-associated protein-1) and designated rat Dzip1S. The coding region was determined by DAZAP2 [23]. DAZAP1 encodes an RNA-binding protein the first methionine after an upstream in-frame stop codon. expressed most abundantly in the testis [23]. DAZ1 and DAZL A BLASTP search of the GenBank database was conducted also associate with human DZIP1 (DAZ-interacting protein), against the predicted amino acid sequence of rat Dzip1S. It was which encodes three different protein isoforms that contain a found to be 99% identical to a predicted rat protein C2H2 zinc finger domain [24]. The DZIPt1 isoform has two ‘‘hypothetical protein XP_579325’’ (GenBank Accession No. alternative exons, which insert 46 amino acids at the N- XP_344225). The predicted amino sequence was also found to terminus and result in premature truncation of the protein just be 65% identical (over 326 amino acids) to a further predicted carboxy-terminal to the C2H2 zinc finger domain [24]. The rat protein ‘‘similar to DAZ interacting protein 1’’ (GenBank DZIPt2 isoform has an alternative exon that adds 19 amino Accession No. XP_344461). The full-length sequence was acids beyond the C2H2 zinc finger domain [24]. In the present confirmed by PCR using primers designed to amplify the study, we report on the characterization and expression of the coding region of rat Dzip1S, which gave the correct predicted rat and mouse homologues of DZIP1 and, in addition, size of 1050 bp (Fig. 1C). Thus the rat sequence encodes a rat characterize the genomic organization of rat, mouse, and homologue of DZIP1 and has been designated rat Dzip1S. The human DZIP1. reasoning behind this nomenclature will be discussed below.

Results Comparison of rat Dzip1S to mouse and human DZIP1

Cloning of cDNA encoding rat Dzip1S The deduced amino acid sequence of rat Dzip1S is shown in Fig. 2A. A motif search on the rat Dzip1S peptide sequence During a search for genes involved in the regulation of was performed (http://myhits.isb-sib.ch/cgi-bin/motif_scan) spermatogenesis in the rat, we isolated a partial cDNA sequence and produced one significant alignment, which was the of 812 bp from a rat testis cDNA library with an uninterrupted C2H2 zinc finger (pfamPF00096). This motif consists of open reading frame. This sequence aligned with 269 amino acids CX2–4CX12HX2–6H, where C represents cysteine, H histidine, of an uncharacterized human gene, KIAA0996 (GenBank and X any amino acid. The BLASTP analysis described above

Fig. 1. RACE PCR and cloning of rat Dzip1S. (A) 3V-RACE of rat Dzip1S. Forward primary and nested gene-specific primers were designed as described under Materials and methods, and RACE PCR was performed with a universal primer provided in the RACE kit (Clontech). (B) 5V-RACE of rat Dzip1S. Reverse gene- specific primers were designed as described under Materials and methods, and RACE PCR was performed with a universal primer provided in the RACE kit (Clontech). (C) Full-length PCR of the coding region of rat Dzip1S, using primers designed according to the rat Dzip1S sequence from RACE PCR. M, GeneRuler 100 bp DNA ladder (Fermentas). B.J. Curry et al. / Genomics 87 (2006) 275–285 277

Fig. 2. Deduced amino acid sequence of rat Dzip1S and alignments with mouse Dzip1 and DZIP1 proteins. (A) Amino acid sequence of the rat Dzip1S gene (GenBank Accession No. AY847187). The amino acid sequence is shown in single-letter code and starts from the first methionine (M) downstream of a stop codon. (B) Multiple sequence alignment of rat Dzip1S (GenBank Accession No. AAW31758), the predicted mouse Dzip1 (GenBank Accession No. NP_080219), and DZIP1 (GenBank Accession No. Q86YF9). Identical residues are boxed in black and chemically similar ones in gray. All sequences were aligned using the

CLUSTALW program. The conserved C2H2 zinc finger (pfamPF00096) motif is underlined. The basic regions flanking the zinc finger are double underlined. Rat Dzip1S, the short rat homologue of DZIP1; mouse Dzip1, predicted mo1use homologue of DZIP1; DZIP1, the human gene DAZ interacting protein 1. 278 B.J. Curry et al. / Genomics 87 (2006) 275–285 demonstrated that rat Dzip1S has a 60% identity with DZIP1 The mouse Dzip1 exon structure was determined by search- over 341 amino acids (GenBank Accession No. Q86YF9). The ing the mouse genome database with GenBank Accession No. corresponding nucleotide sequence for human DZIP1 is NM_025943. There was one positive match on chromosome GenBank Accession No. NM_014934. The BLASTP analysis 14, ‘‘Mus musculus chromosome 14 genomic contig, strain showed a 66% identity over 327 amino acids with a predicted C57BL/6J’’ (GenBank Accession No. NT_039609), which mouse DZIP1 homologue (GenBank Accession No. consisted of 20 individual hits corresponding to 20 exons (Fig. AAH98211). The corresponding nucleotide sequence for 3A). The DZIP1 exon structure was determined by searching mouse Dzip1 is GenBank Accession No. NM_025943. The the human genome database with GenBank Accession No. Mouse Genome Informatics designation for mouse Dzip1 is NM_014934, resulting in a hit on chromosome 13, ‘‘Homo MGI:1914311. A CLUSTALW analysis was performed on rat sapiens chromosome 13 genomic contig’’ (GenBank Accession Dzip1, mouse Dzip1, and DZIP1 (Fig. 2B). The C2H2 zinc No. NT_009952), which consisted of 22 individual hits, finger motif is quite well conserved between the three species, corresponding to 22 exons (Fig. 4A). In a fashion similar to with only small differences seen in the alignment (Fig. 2B). All that of the rat, DZIP1 is encoded on the complementary strand. sequences have a stretch of basic amino acid residues N- The intron/exon boundaries for mouse Dzip1 and DZIP1 were terminal to the C2H2 zinc finger, which is RRRKK in the rat determined based on the GT/AG splice rule. The 20 exons and 19 and mouse and KRRKK in the human (Fig. 2B). The rat also introns of mDZIP1 and the 22 exons and 21 introns of DZIP1 are has weakly basic residues C-terminal to the zinc finger domain, marked by the presence of highly conserved GT (at the 5V end of the sequence being KKAMAR (Fig. 2). These basic amino acid the intron) and AG (at the 3V end of the intron) sequences. This stretches are thought to strengthen the DNA binding ability of shows that although DZIP1 and mouse Dzip1 amino acid the C2H2 zinc finger domain. sequences share significant similarity (Fig. 2B), the exon structures of these genes do not align at all (Figs. 3A and 4A). Determination of rat Dzip1S genomic structure and location and comparison with mouse and human DZIP1 genomic structure Comparison of sequence homology of DZIP1 between rat, mouse, and human To obtain the exon structure of the rat Dzip1S gene, the cDNA sequence (GenBank Accession No. AY847187) was used as a When mouse Dzip1 (GenBank Accession No. NM_025943) template in a BLASTN search against the rat genome database, was searched against the rat genome there was one hit on resulting in one positive match on chromosome 14, ‘‘Rattus chromosome 14 from 358 to 781 bp on the mouse cDNA norvegicus chromosome 14 WGS supercontig’’ (GenBank sequence. This genomic sequence corresponds to 355–1322 bp Accession No. NW_047424). There was one hit corresponding on the rat Dzip1S cDNA sequence, and the alignment ends just to one exon, which in the rat genome covered the first 1600 bp before the end of the coding region of rat Dzip1S (Fig. 3B). of the cDNA sequence. The rat Dzip1S sequence is in the When a BLAST2 search was performed between rat Dzip1S reverse orientation in the genome on chromosome 14p21, and mouse Dzip1, there was 75% identity between the two which indicates that it is encoded on the complementary strand. sequences up to the point where the alignment ends at 1204 bp This region of the rat genome is incomplete; however, as the of mouse Dzip1 (Fig. 3B). Mouse Dzip1 consists of 20 exons, coding region ends at 1365 bp, it is confirmed that the coding and the exons and introns are of various sizes (Fig. 3A). The region of rat Dzip1S is confined to one exon. alignment with rat Dzip1S starts at the end of mouse exon 1,

Fig. 3. Comparative genomic organization of mouse Dzip1 and rat Dzip1S. (A) The exon structure of mouse Dzip1 (M1-M20) based on a predicted sequence (GenBank Accession No. NM_025943). Introns and exons are drawn to scale. The number of base pairs is given for each exon and for intron 4. Length of intron 4 is given in italics. The coding region is shown from 255 to 2183 bp. (B) A comparison of the genomic organizations of mouse Dzip1 and rat Dzip1S. A BLAST2 analysis between mouse Dzip1 and rat Dzip1S showed that the rat sequence aligns to the first six exons of the mouse Dzip1 sequence from base pair 355 to 1322. This alignment has an identity of 75% and contains the majority of the coding sequence of rat Dzip1S. B.J. Curry et al. / Genomics 87 (2006) 275–285 279

Fig. 4. Comparative genomic organization of DZIP1 and rat Dzip1S. (A) The exon structure of DZIP1 (H1–H22) based on the sequence of GenBank Accession No. NM_014934. Introns and exons are drawn to scale. The number of base pairs is given for each exon. The coding region is shown from 834 to 3380 bp. (B) A comparison of the genomic organizations of DZIP1 and rat Dzip1S. Exons H5–H9 of DZIP1 are shown, along with the whole of the rat Dzip1S sequence. A BLAST2 comparison between the DZIP1 and the rat Dzip1S sequence showed an alignment between nucleotide 77 in exon 5 through to nucleotide 35 of exon 9 of human DZIP1 and nucleotides 436–1330 of rat Dzip1S. includes the whole of exons 2–4, and ends 26 bp through exon (Figs. 2B and 5C), but does not show conserved linkage 5(Fig. 3B). homology. When the DZIP1 cDNA sequence (GenBank Accession No. NM_014934) was searched against the rat genome there was RT-PCR analysis of rat and mouse DZIP one hit on chromosome 14 from 948 to 1857 bp on the human DZIP1. This genomic sequence corresponds to 436–1330 bp RRT-PCR of rat Dzip1S and mouse Dzip1 was performed on the rat Dzip1S cDNA sequence and the alignment ends just on total RNA from multiple rat and mouse tissues (Figs. 6A– before the end of the coding region of rat Dzip1S (Fig. 4B). 6D), to obtain a qualitative expression of rat and mouse Dzip1 When a BLAST2 search was performed between rat Dzip1S mRNA. Rat Dzip1S was expressed in all tissues examined (Fig. and DZIP1, there was 75% identity found between the two 6A). Mouse DZIP1 was also present in all tissues examined sequences up until where the alignment ends at 1839 bp of except for liver (Fig. 6C). A h-actin PCR is shown for both DZIP1 (Fig. 4B). DZIP1 consists of 22 exons, and the exons species tissue panels as a control for cDNA integrity (Figs. 6B and introns are of various sizes (Fig. 4A). The alignment with and 6D). The 151-bp PCR product of mouse Dzip1 shown in rat Dzip1S starts partway through human exon 5, includes the Fig. 6C was sequenced and found to be identical to GenBank whole of exons 6–8, and ends 35 bp through exon 9 (Fig. 4B). Accession No. NM_025943, from 992 to 1142 bp. The above mouse DZIP1 cDNA search on the rat genome also gave one positive match on chromosome 15, ‘‘Rattus Confirmation of intronless status of rDZIP1S norvegicus chromosome 15 genomic contig’’ (GenBank Acces- sion No. NW_047455). There were 26 hits corresponding to 26 To confirm the intronless status of rat Dzip1S compared to exons. This resulted in the identification of a second predicted the intron-containing mouse Dzip1, PCR was performed on homologous rat Dzip1 sequence (GenBank Accession No. cDNA and genomic DNA (gDNA) for both mouse and rat XP_344461). This sequence and rDzip1S (GenBank Accession testis. For rat PCR, the Dzip1 primers were the same as used No. AAW31758) are homologous over 326 residues with an previously for Fig. 1C, covering the full coding region of rat identity of 65%. A CLUSTALW analysis shows that sequence Dzip1S. The same PCR amplicon of 1050 bp was produced XP_344461 is much larger than AAW31758, being longer at with both cDNA and gDNA (Fig. 7A), indicating that there the amino and carboxyl-terminals (Fig. 5A). A rat genome were no introns present in rat Dzip1S. Importantly, PCR with analysis shows that this larger sequence maps at chromosome h-actin confirmed the presence of gDNA as the h-actin primers 15 and shows conserved linkage homology with mouse Dzip1 were designed across large introns, so no discrete amplicon was on chromosome 14 (GenBank Accession No. NM_025943) seen with gDNA, but the predicted amplicon of 1128 bp was and DZIP on chromosome 13 (GenBank Accession No. seen with cDNA (Fig. 7B). NM_014934) (Fig. 5B). Genes that show linkage homology With the mouse the situation is entirely different. The cDNA are CLDN10, DNAJC3, and UGCGL2. This makes it likely amplicon of mouse Dzip1 had the expected size of 151 bp; that sequence XP_344461 is the rat homologue of DZIP1 and however, the amplicon produced from gDNA was 1088 bp (Fig. should be designated rat Dzip1. Sequence AAW31758, isolated 7C). This was expected as the PCR primers for mouse Dzip1 in the present study, we have designated rat Dzip1S, as it is the were designed in exons 4 and 5, and there is an intervening short homologue of rat Dzip1. As described earlier, rat Dzip1S intron of 937 bp (Fig. 3A). Like the rat, h-actin PCR with is also homologous to the mouse and human versions of DZIP1 mouse gDNA did not produce an amplicon (Fig. 7D). 280 B.J. Curry et al. / Genomics 87 (2006) 275–285

In situ hybridization of rat Dzip1S to rat testis sections testicular cells (Fig. 8B), indicating the specificity of the antisense cRNA hybridization. rat Dzip1S message was The expression of rat Dzip1S in the male germ line was expressed specifically in premeiotic spermatogonia in the confirmed by in situ hybridization using a DIG-labeled rat testis. Other differentiating germ cells, including the fully probe. Hybridization with an antisense cRNA probe showed formed spermatozoa within the tubule lumen, were com- specific staining to the spermatogenic cell lineage (Figs. 8A pletely unlabeled (Figs. 8A and 8C), as were the Sertoli cells and 8C), compared to that of a hybridization using a sense that extend from the basement membrane to the lumen. Thus cRNA probe, which did not result in any staining of rat rat Dzip1S was expressed only in cells of the male germ cell

Fig. 5. Schematic alignment of the two rat homologues of rat Dzip1 with mouse and human homologues. (A) Multiple sequence alignment of the predicted rat Dzip1 (GenBank Accession No. XP_344461) and rat Dzip1S (GenBank Accession No. AAW31758) peptide sequences. Identical residues are boxed in black and chemically similar ones in gray. All sequences were aligned using the CLUSTALW program. The conserved C2H2 zinc finger (pfamPF00096) motif is underlined. The basic regions flanking the zinc finger are double underlined. (B) Schematic alignment of the peptide sequence of rat Dzip1 found on chromosome 15 with that of the mouse Dzip1 (GenBank Accession No. NP_080219) and DZIP1 (GenBank Accession No. Q86YF9) peptide sequences. Numbers shown are peptide residue numbers. Dotted lines show where alignments start and end. (C) Schematic alignment of the peptide sequence of rat Dzip1S found on chromosome 14 with that of the mouse Dzip1 and DZIP1 peptide sequences. Numbers shown are peptide residue numbers. Dotted lines show where alignments start and end. B.J. Curry et al. / Genomics 87 (2006) 275–285 281

Fig. 5 (continued). lineage and was not expressed in the terminally differentiated encodes at least three different protein isoforms that contain a spermatozoon. C2H2 zinc finger domain [24]. We have identified a rat homologue of DZIP1 using RACE-PCR, called rat Dzip1S, Discussion and extended the study to investigate the existence and expression of the mouse homologue of DZIP1, and to compare The human DAZ genes encompass a family that is deemed the genomic organization of rat Dzip1S, mouse Dzip1, and to play a role in early male germ cell development [2,3,12,14]. DZIP1. We also examined the expression of rat Dzip1S in rat The genes DAZ1, DAZL, and BOLL encode RNA-binding testis sections by in situ hybridization. proteins, and their binding partners such as PUM2 [19,22], RACE-PCR identified a 2057-bp cDNA sequence in the rat DAZAP1,andDAZAP2 [23] also encode RNA-binding with an open reading frame of 332 amino acids that was proteins. DAZ1 and DAZL also associate with DZIP1, which confirmed by RT-PCR to consist of a single transcript. Sequencing of this amplicon confirmed the cDNA sequence obtained through RACE. A motif search on the putative protein sequence of rat Dzip1S produced one significant alignment,

Fig. 7. Confirmation of intronless status of rat Dzip1S. PCR was performed on rat Fig. 6. RT-PCR analysis of rat Dzip1S and mouse Dzip1 from rat and mouse tissue and mouse testis cDNA and genomic DNA (gDNA). (A and B) PCR of rat testis panels. (A and B) RT-PCR analysis of rat tissue panel with primers specific to (A) cDNA and gDNA with primers specific to (A) rat Dzip1S and (B) b-actin. rat Dzip1S and (B) b-actin. (C and D) RT-PCR analysis of mouse tissue panel with Predicted size for rat Dzip1S amplicons for cDNA and gDNA is 1150 bp. (C and primers specific to (C) mouse Dzip1 and (D) b-actin. Lanes: T, testis; E, D) PCR of mouse testis cDNA and gDNA with primers specific to (C) mouse epididymis; SM, skeletal muscle; Lu, lung; Li, liver; SI, small intestine; B, brain; Dzip1 and (D) b-actin. Predicted size for mouse Dzip1 amplicons for cDNA is P, prostate; K, kidney; H, heart; M, GeneRuler 100 bp DNA ladder (Fermentas). 151 bp and for gDNA is 1088 bp. M, GeneRuler 100 bp DNA ladder (Fermentas). 282 B.J. Curry et al. / Genomics 87 (2006) 275–285

Fig. 8. Expression of rat Dzip1S mRNA in rat testis by in situ hybridization. (A) Antisense cRNA control for the rat Dzip1S DIG probe indicating specificity of the antisense cRNA interaction. Original magnification: Â100. (B) Sense cRNA for the rat Dzip1S DIG probe. Original magnification: Â100. (C) A Â200 magnification of the antisense cRNA of the rat Dzip1S DIG probe. The signal for rat Dzip1S was specific to cells of the male germ cell lineage, restricted to premeiotic spermatogonia (spg). L, Lumen of seminiferous tubule. S, Sertoli cells.

which was the C2H2 zinc finger (pfamPF00096). Zinc fingers The rat Dzip1S gene described in the present study has only are usually associated with the regulation of transcription and one exon and is thus considered intronless. This compares with consist of a projecting ‘‘finger’’ structure that is stabilized by a DZIP1, which has 22 exons and mouse Dzip1 with 20 exons zinc ion bound by cysteine and histidine residues. In typical (Figs. 3 and 4). The intronless nature of rat Dzip1S was C2H2 zinc fingers the number of tandemly repeating finger confirmed in the current study with amplicons of different sizes domains varies from 2 to 37 with a minimum of 2 units produced when Dzip1 PCR was performed on rat and mouse required for specific high-affinity DNA binding (reviewed in gDNA (Fig. 7). This indicated the presence of introns in the [26]). However, proteins that require only a single zinc finger gDNA that encodes mouse Dzip1 but not in the gDNA that of the C2H2 family for DNA binding have been recently encodes rat Dzip1S. There are many examples of intronless identified. These include the Drosophila GAGA transcription genes expressed specifically in the male germ line [30–34]. factor [27,28] and the Arabidopsis SUPERMAN protein [29]. These are commonly processed genes that are believed to be As well as the C2H2 zinc finger motif, these proteins have two retrotransposons generated by making a reverse transcriptase basic flanking regions, which stabilize DNA binding [27,29]. copy of an mRNA and inserting the DNA copy into genomic The latest addition to this list is DZIP1 [24]. The rat homologue DNA in a germ-line cell [35,36]. Typically such processed described in the present study, rat Dzip1S, has a stretch of basic sequences form nonfunctional pseudogenes with no promoter amino acid residues flanking the C2H2 zinc finger (Fig. 2). Rat sequences, lack of transcription, and sequence degradation. As Dzip1S has a strong N-terminal basic sequence of RRRKK and such these genes are not expressed in all cells; however, genes a weaker C-terminal basic sequence of KKAMAR (Fig. 2A). such as those described above have obtained a promoter active in The mouse Dzip1 protein has similar basic sequences flanking testicular germ cells. Thus rat Dzip1S either is a processed gene the C2H2 zinc finger motif, while the human protein has only that has evolved from an intron-containing progenitor, such as an N-terminal flanking basic sequence (Fig. 2B). The human PGK-2 [30], or has eliminated introns directly by gene transcripts that DZIP1 regulate with the zinc finger domain conversion. As rat Dzip1S is expressed in a variety of tissues it is are yet to be determined. more likely that is has arisen through gene conversion. B.J. Curry et al. / Genomics 87 (2006) 275–285 283

Previous studies have shown that constitutively spliced in a pattern identical to that of DAZ1 and DAZL proteins [7].In introns are essential for optimal gene expression [37,38].Ithas human adult testis, DZIP1 was found to be localized to the also been shown that in higher eukaryotes intron-containing spermatogonia [24], in a fashion similar to that of DAZ1 and and intronless versions of otherwise identical genes can exhibit DAZL [7]. In mouse testis, previous work has shown DZIP1 to dramatically different expression profiles, the presence of be localized to spermatogonia and primary and secondary introns usually enhancing expression [39,40]. In the present spermatocytes using a polyclonal antibody raised to human study, the expression of rat Dzip1S does not seem to have been DZIP1 peptides [24]. In the current study we used a DIG-labeled affected by a lack of introns, as it was found to be expressed in rat Dzip1S probe to investigate the localization of rat Dzip1S all tissues of a rat tissue panel by RT-PCR (Fig. 6). mRNA in rat testis sections. We found that expression of rat There was a second rat Dzip1 homologue identified when Dzip1S mRNA was restricted to cells of the spermatogenic cell mouse Dzip1 was searched against the rat genome. This is a lineage, specifically premeiotic spermatogonia (Fig. 8). This predicted sequence (GenBank Accession No. XP_344461). confirmed previous immunohistochemical studies [24] and puts Unlike rat Dzip1S it is an intron-containing gene, with 26 rat Dzip1S in an ideal location to interact with DAZ1 and DAZL exons. This rat Dzip1 predicted protein is much larger than rat and influence male germ cell development. Dzip1S, being composed of 1219 amino acids, but they are This study has identified by RACE PCR rat Dzip1S, a rat homologous over 326 residues with a 65% identity. However, homologue of DZIP1 that is highly expressed in the adult testis this predicted rat Dzip1 homologue, which maps to chromo- and has an intact C2H2 zinc finger domain, suggesting a role in some 15, shows conserved linkage homology with mouse regulation of transcription. Its genomic structure was compared Dzip1 on mouse chromosome 14 and with DZIP1 on human to that of mouse Dzip1 and DZIP1. This analysis indicated that chromosome 15. Other than DZIP1, genes that show conserved rat Dzip1S is intronless. Both rat Dzip1S and mouse Dzip1 were linkage homology are Claudin 10 (GenBank Accession No. found to be expressed in the testis and rat Dzip1S expression was NM_006984), DnaJ (Hsp40) homologue (GenBank Accession found to be restricted to cells of the spermatogenic lineage by in No. NM_020121), and UDP-glucose ceramide glucosyltrans- situ hybridization of rat testis. Further work on rat Dzip1S will ferase-like 2 (GenBank Accession No. NM_006260). With this involve confirmation of its RNA binding capacity and identi- evidence it is proposed that this predicted rat sequence is the fication of binding partners such as DAZ1 or DAZL. true rat homologue of DZIP1 and is therefore named rat Dzip1. As the rat homologue described in this study is homologous to Materials and methods rat Dzip1 but is shorter, it is proposed that it be called rat Dzip1S. Chemicals Rat Dzip1S described in the present study has a predicted protein that is truncated compared to DZIP and mouse Dzip1. Unless otherwise specified, all chemicals were purchased from Sigma– The question arises as to the function of this protein. Moore et Aldrich (Castle Hill, NSW, Australia). al. [24] found that DZIPt1 interacts with DAZ1 at the same Isolation of total RNA from mouse and rat solid tissue region of DAZ1 protein required for the formation of homo- and heterodimers with itself and the DAZL protein [23,24,41]. Total RNA was extracted from solid tissues as previously described [42]. DZIPt1, like rDzip1S, is truncated compared to full-length DZIP1, and the minimum region of DZIPt1 that is required for Molecular cloning of the rat homologue of DZIP1 DAZ interaction was the C2H2 zinc finger region. As this region is present in rat Dzip1S, it is probable that rat Dzip1S In conducting a search for novel NAD(P)H oxidoreductases in a rat testis cDNA library (BD Biosciences Clontech, Palo Alto, CA, USA), we employed a will also bind to DAZ1 and be part of a protein cluster with degenerate PCR strategy, designing primers based on conserved regions of DAZ1, DAZL, and possibly other DAZ-interacting proteins known ferric reductases. These were FRE1 (GenBank Accession No. M86908) such as PUM2, DAZAP1, and DAZAP2, to regulate transcrip- [43] and FRE2 (GenBank Accession No. Z28220) [44], and the degenerate tion of specific genes. primers were forward, 5V-G/TTITTC/TA/CGIA/CA/GIA/TA/GIG/TG/TITAC/ By Northern blot analysis, Moore et al. [24] found DZIP1 TGA-3V, and reverse, 5V-TCITTIAG/A/CA/GTC/TIGGICG/TICC-3V. Inosine was used to decrease the amount of degeneracy of the primers. The full-length expression to be predominantly in the testis, with much lower gene of rat Dzip1S was determined by 5V- and 3V-RACE, using the SMART levels in the skeletal muscle and ovary. They also used RACE Amplification Kit from Clontech. RACE-ready cDNA was prepared quantitative PCR to detect expression of DZIP1 in fetal brain, from 1 Ag of rat testis total RNA using Powerscript reverse transcriptase ovary, and testis and also in adult oocytes and testis. The (Clontech), according to the RACE kit. Primary and nested 5V- and 3V-RACE, current study used RT-PCR to confirm that rat Dzip1S and respectively, were required to determine the full-length sequence. The gene- specific primers for 5V- and 3V-RACE were designed from the initial 812-bp mouse DZIP1 are expressed in brain and testis and in a range cDNA sequence. The 5V-RACE gene-specific primer was 5V-CCACTCAGGCT- of other adult tissues. It is probable that we have detected CAGGCGTAACCG-3V. The 3V-RACE gene-specific primers were primary, mouse and rat Dzip1 homologues in tissues other than those 5V-CGGCCATTACCAGTGTGGTTTTTGTG-3V, and nested, 5V-AGAGT- previously shown for DZIP1, as we used the more sensitive GAGGTCAACATGGGCAGGGAG-3V. The primers were at a final concen- technique of PCR, rather than Northern blotting [24]. tration of 0.2 AM. Touchdown PCR was performed with the following reaction conditions: 5 cycles of 94-C for 30 s, 72-C for 3 min; 5 cycles of 94-C for 30 s, There is evidence that DZIP1 is expressed in male germ cells. 70-C for 30 s, 72-C for 3 min; 25 cycles of 94-C for 30 s, 68-C for 30 s, 72-C In human fetal testis sections, Moore et al. [24] found that DZIP1 for 3 min. The PCR products were fractionated on 1% agarose gels prestained localized to the nucleus and cytoplasm of primordial germ cells with ethidium bromide. PCR products were gel purified and cloned into pGEM 284 B.J. Curry et al. / Genomics 87 (2006) 275–285

Easy-T (Promega Corp., Annandale, NSW, Australia) and sequenced with an NSW, Australia) was used to produce DIG-labeled probes with T7 RNA ABI 377 Prism DNA sequencer (Applied Biosystems, Scoresby, VIC, polymerase producing antisense probes and SP6 RNA polymerase producing Australia). The full-length rat Dzip1S sequence was collated from the sense probes. Rat testes were treated, probed with rat Dzip1S, washed, and overlapping sequences obtained from the rounds of 5V- and 3V-RACE. imaged as described previously [45]. For the cloning of the full-length coding region of rat Dzip1S, the Advantage 2 Taq mix from Clontech was used. The forward primer was 5V- TGGGTGTAGAGTCTTGGGAC-3V and the reverse primer was 5V-CTCAAG- Acknowledgment GATGTGCAGCGTCT-3V. The expected amplicon size was 1150 bp. The primers were at a final concentration of 0.2 AM. The PCR conditions were 1 This work was supported by the ARC Centre of Excellence - - - - cycle of 94 C for 5 min; 35 cycles of 95 C for 45 s, 57 C for 45 s, 72 C for 2 in Biotechnology and Development. min; 1 cycle of 72-C for 10 min. The PCR products were fractionated on 0.8% agarose gels prestained with ethidium bromide. PCR products were gel purified, cloned into pGEM Easy-T, and sequenced as described above. References

Preparation of genomic DNA from rat and mouse testis [1] R. Reijo, et al., Diverse spermatogenic defects in humans caused by Y chromosome deletions encompassing a novel RNA-binding protein gene, Freshly isolated mouse and rat testis was homogenized in 5 ml of STE buffer Nat. Genet. 10 (1995) 383–393. (500 mM NaCl, 100 mM Tris–Cl, pH 8.0, 10 mM EDTA) on ice. To the [2] R. Reijo, R.K. Alagappan, P. Pattrizio, D.C. Page, Severe oligozoospermia homogenate was added 250 Al of 20% SDS, 50 Alofh-mercaptoethanol, and 500 resulting from deletions of azoospermia factor gene on Y chromosome, Al of 20mg/ml proteinase K. The solution was mixed and incubated at 55-C Lancet 347 (1996) 1290–1293. overnight. An equal volume of phenol, pH 8.0, was added to the mixture, shaken [3] C. Foresta, et al., Y-chromosome deletions in idiopathic severe testiculo- vigorously for 60 s, and centrifuged at 3000g for 15 min. The aqueous layer was pathies, J. Clin. Endocrinol. Metab. 82 (1997) 1075–1080. removed to a fresh tube and an equal volume of phenol:chloroform:isoamyl [4] J.L. Pryor, et al., Microdeletions in the Y chromosome of infertile men, alcohol (25:24:1) was added, shaken vigorously, and centrifuged at 3000g for 15 N. Engl. J. Med. 336 (1997) 534–539. min. The aqueous layer was removed to a fresh tube and an equal volume of [5] R. Saxena, et al., Four DAZ genes in two clusters found in the AZFc chloroform was added, shaken vigorously for 15 min, and centrifuged at 3000g region of the human Y chromosome, Genomics 67 (2000) 256–267. for 15 min. The aqueous layer was removed to a fresh tube and the DNA was [6] R.A. Reijo, et al., DAZ family proteins exist throughout male germ cell precipitated with 1/10 volume 3 M sodium acetate and 2 volumes of 100% development and transit from nucleus to cytoplasm at meiosis in humans ethanol. and mice, Biol. Reprod. 63 (2000) 1490–1496. [7] H.J. Cooke, M. Lee, S. Kerr, M. Ruggiu, A murine homolog of the human DAZ gene is autosomal and expressed only in male and female gonads, RT-PCR of rat Dzip1S and mouse Dzip1 and PCR on genomic DNA Hum. Mol. Genet. 5 (1996) 513–516. [8] Z. Shan, A SPGY copy homologous to the mouse gene Dazla and the Five micrograms of total RNA from a panel of rat and mouse tissues was Drosophila gene boule is autosomal and expressed only in the human male reverse transcribed with oligo(dT)15 primer (Promega) and M-MLV reverse À gonad, Hum. Mol. Genet. 5 (1996) 2005–2111. transcriptase RNase H (Promega). Polymerase chain reaction for rat Dzip1S [9] R. Saxena, et al., The DAZ gene cluster on the human Y chromosome was performed with Taq polymerase from Fermentas (Hanover, MD, USA). arose from an autosomal gene that was transposed, repeatedly amplified Primers were designed for rat Dzip1S and mouse Dzip1. For the rat Dzip1S and pruned, Nat. Genet. 14 (1996) 292–299. V tissue panel RT-PCR, the primer sequences were forward, 5 -AAAAGG- [10] E. Seboun, et al., Gene sequence, localization and evolutionary conser- V V CAATGGCAAGGAAC-3 , and reverse, 5 -TCCTCCTCCCAACTAACCAC- vation of DAZLA, a candidate male sterility gene, Genomics 41 (1997) 3V. These primers generated a band of 187 bp. The PCR conditions were as 227–235. - - - follows: 1 cycle of 94 C for 5 min; 35 cycles of 95 C for 45 s, 60 C for 45 s, [11] M. Ruggiu, et al., The mouse Dazla gene encodes a cytoplasmic protein - - 72 C for 2 min; 1 cycle of 72 C for 10 min. essential for gametogenesis, Nature 389 (1997) 73–77. _ For mouse Dzip1 (GenBank Accession No. NM 025943), the primer [12] R. Reijo, et al., Mouse autosomal homolog of DAZ, a candidate male sequences were forward, 5V-CTCTCATGCAGTTCGCTTC-3V, and reverse, 5V- sterility gene in humans, is expressed in male germ cells before and after V AAATTCCCTCATAAACATCCCC-3 . These primers generated a band of 151 puberty, Genomics 35 (1996) 346–352. - - bp. The PCR conditions were 1 cycle of 94 C for 5 min; 35 cycles of 95 C for [13] E.Y. Xu, F.L. Moore, R.A. Reijo Pera, A gene family required for human - - - 45 s, 60 C for 45 s, 72 C for 2 min; 1 cycle of 72 C for 10 min. The primer germ cell development evolved from an ancient meiotic gene conserved in sequences for h-actin PCR were forward, 5V-ATGGATGATGA- metazoans, Proc. Natl. Acad. Sci. USA 98 (2001) 7414–7419. V V TATCGCCGCG-3 , and reverse, 5 -CTAGAAGCATTTGCGGTGGACGATG- [14] C.G. Eberhart, J.Z. Maines, S.A. Wasserman, Meiotic cell cycle V - GAGGGGCC-3 . The PCR conditions were as follows: 1 cycle of 94 C for 5 requirement for a fly homologue of human Deleted in Azoospermia, - - - - min; 35 cycles of 95 C for 45 s, 60 C for 45 s, 72 C for 2 min; 1 cycle of 72 C Nature 381 (1997) 783–785. for 10 min. These primers generated a band of 1128 bp. The PCR products [15] J.Z. Maines, S.A. Wasserman, Post-transcriptional regulation of the were run on 1% agarose gels prestained with ethidium bromide. meiotic Cdc25 protein Twine by the Dazl orthologue Boule, Nat. Cell PCR was also performed on 1 ng of rat and mouse gDNA to compare the Biol. 1 (1999) 171–174. products obtained with cDNA. The primers used for the rat gDNA were the [16] E.Y. Xu, et al., Human BOULE gene rescues meiotic defects in infertile same as those used for the full-length rat Dzip1S determination as described flies, Hum. Mol. Genet. 12 (2003) 169–175. above with the same PCR conditions. The primers used for the mouse gDNA [17] C.M. Luetjens, Association of meiotic arrest with lack of BOULE were the same as those used for the RT-PCR for the tissue panel as described protein expression in infertile men, J. Clin. Endocrinol. Metab. 89 above with the same PCR conditions. (2004) 1926–1933. [18] F.L. Moore, et al., Human pumilio-2 is expressed in embryonic stem cells In situ hybridization of rat Dzip1S to rat testis sections and germ cells and interacts with DAZ (Deleted in Azoospermia) and DAZ-like proteins, Proc. Natl. Acad. Sci. USA 100 (2003) 538–543. For in situ hybridization, the 187-bp amplicon produced by RT-PCR of rat [19] J. Urano, M.S. Fox, R.A. Reijo Pera, Interaction of the conserved meiotic Dzip1S RNA was ligated into pGEM Easy T (Promega) and linearized by regulators, BOULE (BOL) and PUMILIO-2 (PUM2), Mol. Reprod. Dev. restriction enzyme digestion. The peptide sequence of rat Dzip1S covered by 71 (2005) 290–298. the DIG probe was 232-LKEQLQVTRTQLESVQHTHRVRPSTDYERQKI- [20] J. Sonoda, R.P. Wharton, Drosophila brain tumour is a translational NEEDFQKLVVSWEEE-278. A DIG RNA labeling kit (Roche, Castle Hill, repressor, Genes Dev. 15 (2001) 762–773. B.J. Curry et al. / Genomics 87 (2006) 275–285 285

[21] M. Wickens, D.S. Bernstein, J. Kimble, R. Parker, A PUF family portrait: [33] M. Hisano, S. Yamada, H. Tanaka, Y. Nishimune, M. Nozaki, Genomic 3VUTR regulation as a way of life, Trends Genet. 18 (2002) 150–157. structure and promoter activity of the testis haploid germ cell-specific [22] M. Fox, J. Urano, R.A. Reijo Pera, Identification and characterization of intronless genes, Tact1 and Tact2, Mol. Reprod. Dev. 65 (2003) 148–156. RNA sequences to which human PUMILO-2 (PUM2) and deleted in [34] Z. Hao, et al., Expression analysis of the human testis-specific azoospermia-like (DAZL) bind, Genomics 85 (2005) 92–105. serine/threonine kinase (TSSK) homologues: a TSSK member is present [23] S. Tsui, et al., Identification of two novel proteins that interact with germ- in the equatorial segment of human sperm, Mol. Hum. Reprod. 10 (2004) cell-specific RNA-binding proteins DAZ and DAZL1, Genomics 65 433–444. (2000) 266–273. [35] C. Esnault, J. Maestre, T. Heidman, Human LINE retrotransposons [24] F.L. Moore, J. Jaruzelska, D.M. Dorfman, R.A. Reijo-Pera, Identification generate processed pseudogenes, Nat. Genetics 24 (2000) 363–367. of a novel gene, DZIP (DAZ-interacting protein), that encodes a protein [36] A.J. Mighell, N.R. Smith, P.A. Robinson, A.F. Markham, Vertebrate that interacts with DAZ (deleted in azoospermia) and is expressed in pseudogenes, FEBS Lett. 468 (2000) 109–114. embryonic stem cells and germ cells, Genomics 83 (2004) 834–843. [37] A.R. Buchman, P. Berg, Comparison of intron-dependent and intron- [25] T. Nagase, et al., Prediction of the coding sequences of unidentified human independent gene expression, Mol. Cell. Biol. 8 (1988) 4395–4405. genes: XIII. The complete sequences of 100 new cDNA clones from brain [38] R.D. Palmiter, E.P. Sandgren, M.R. Avarbock, D.D. Allen, R.L. Brinster, which code for large proteins in vitro, DNA Res. 6 (1999) 65–73. Heterologous introns can enhance expression of transgenes in mice, Proc. [26] A. Klug, J.W.R. Schwabe, Zinc fingers, FASEB J. 9 (1995) 597–604. Natl. Acad. Sci. USA 88 (1991) 478–482. [27] P.V. Pedone, et al., The single Cys2–His2 zinc finger domain of the [39] S. Lu, B.R. Cullen, Analysis of the stimulatory effect of splicing on GAGA protein flanked by basic residues is sufficient for high-affinity mRNA production and utilization in mammalian cells, RNA 9 (2003) specific DNA binding, Proc. Natl. Acad. Sci. USA 93 (1996) 2822–2826. 618–630. [28] J.G. Omichinski, P.V. Pedone, G. Felsenfeld, A.M. Gronenborn, G.M. [40] A. Nott, S.H. Meislen, M.J. Moore, A quantitative analysis of intron Clore, The solution structure of a specific GAGA factor–DNA complex effects on mammalian gene expression, RNA 9 (2003) 607–617. reveals a modular binding node, Nat. Struct. Biol. 4 (1997) 122–130. [41] M. Ruggiu, H. Cooke, In vivo and in vitro analysis of homodimerisation [29] N. Dathan, et al., The Arabidopsis SUPERMAN protein is able to activity of the mouse Dazl protein, Gene 252 (2000) 119–126. specifically bind DNA through its single Cys2–His2 zinc finger motif, [42] B.J. Curry, et al., Identification of a novel splice variant of mouse and Nucleic Acids Res. 30 (2002) 4945–4951. rat cytochrome b5/cytochrome b5 reductase, Genomics 83 (2004) [30] J.R. McCarrey, K. Thomas, Human testis-specific PGK gene lacks introns 425–438. and possesses characteristics of a processed gene, Nature 326 (1987) [43] A. Dancis, D.G. Roman, G.J. Anderson, A.G. Hinnebusch, R.D. Klausner, 501–505. Ferric reductase of Saccharomyces cerevisiae: molecular characterisation, [31] H.H. Dahl, R.M. Brown, W.M. Hutchinson, C. Maragos, G.K. Brown, A role in iron uptake, and transcriptional control by iron, Proc. Natl. Acad. testis specific form of the human pyruvate dehydrogenase E1a subunit is Sci. USA 89 (1992) 3869–3873. coded for by an intronless gene on chromosome 4, Genomics 8 (1990) [44] E. Georgatsou, D. Alexandraki, Two distinctly regulated genes are 225–232. required for ferric reduction, the first step of iron uptake in Saccharomyces [32] N. Reinton, et al., The gene encoding the Cg catalytic subunit of cAMP- cerevisiae, Mol. Cell Biol. 14 (1994) 3065–3073. dependent protein kinase is a transcribed retroposon, Genomics 49 (1998) [45] B. Curry, et al., Identification of RARhoGAP, a novel putative RhoGAP 290–297. gene expressed in male germ cells, Genomics 84 (2004) 406–418.