Gtsf1l and Gtsf2 Are Specifically Expressed in Gonocytes and Spermatids but Are Not Essential for Spermatogenesis

The unknown protein family 0224 (UPF0224) includes three members that are expressed in germ-line cells in mice: Gtsf1, Gtsf1l, and BC048502 (Gtsf2). These genes produce proteins with two repeats of the CHHC Zn-finger domain, a predicted RNA-binding motif, in the N terminus. We previously reported that Gtsf1 is essential for spermatogenesis and retrotransposon suppression. In this study, we investigated the expression patterns and functions of Gtsf1l and Gtsf2. Interestingly, Gtsf1l and Gtsf2 were found to be sequentially but not simultaneously expressed in gonocytes and spermatids. Pull-down experiments showed that both GTSF1L and GTSF2 can interact with PIWI-protein complexes. Nevertheless, knocking out Gtsf1, Gtsf2, or both did not cause defects in spermatogenesis or retrotransposon suppression in mice.


Introduction
In mammals, male reproduction is maintained by successive generations of sperm in the testes. Spermatogenesis involves mitosis, meiosis, and spermiogenesis [1]. Mitosis self-renews the spermatogonia, whose progeny differentiate into primary spermatocytes [2]. Meiosis generates haploid round spermatids and halves the number of chromosomes in the primary spermatocytes. Spermiogenesis, by which the haploid spermatids are processed to mature sperm, occurs in a progressive, organized manner within the seminiferous tubule, and tubule cross-sections reveal spermatogenic cells at 12 distinct stages (I-XII).
We previously used in silico screening to identify mouse genes that are specifically expressed in germ-line cells [11], and reported that one of these genes, Gtsf1/Cue110, is involved in spermatogenesis and retrotransposon suppression in murine testes [12]. We recently confirmed that Gtsf1 is also involved in piRNA biogenesis (unpublished). Here, we focused on Gtsf1l and on BC048502, which we designated Gtsf2. We previously showed by RT-PCR analysis that these genes, which belong to the mouse unknown protein family 0224 (UPF0224), are expressed in the testis [11]. However, the expression and function of Gtsf1l and Gtsf2 in spermatogenesis is unclear. In this study, we examined the Gtsf1l and Gtsf2 expression patterns in detail and described the phenotypes of mice lacking Gtsf1l, Gtsf2, or both.

Materials and Methods Animals
All experiments involving animals were carried out in accordance with institutional guidelines under the protocols (No. 21-089 and No. 26-066), which were approved by the Animal Care and Use Committee of the Osaka University Graduate School of Medicine. Mice were euthanized with an intraperitoneal injection of pentobarbital sodium at 180 mg/kg body weight.

RT-PCR
Total RNA was extracted from embryonic and postnatal testes at various ages using the acid guanidinium thiocyanate-phenol-chloroform method. Total RNA was pretreated with RNasefree DNase I, and then 1 μg of RNA was reverse-transcribed for 30 min at 42°C in a total volume of 40 μl, using an Oligo(dT) primer and ReverTraAce reverse transcriptase (Toyobo, Osaka, Japan), according to the manufacturer's instructions. The resulting cDNA was PCRamplified with gene-specific primers using 20-30 cycles of 98°C for 15 sec, 55-65°C for 30 sec, and 72°C for 60 sec, followed by an extension for 10 min at 72°C. The amplified products were separated by electrophoresis on a 2% agarose gel and detected by ethidium-bromide staining. Gtsf1l cDNA was amplified as a 297-bp product using the primer pair 5'-ATGAGAGGAGGG GACCCAGGAGAAAC-3' and 5'-AAGTGTCCTGCTGCCCAAAGTGTACG-3'. Gtsf2 cDNA was amplified as a 320-bp product using the primer pair 5'-CCAACTGTATCAACAGGACTG CAGT-3' and 5'-GGCAATGTCTCCATCAGTTTTTCTGC-3'. As an internal control, Gapdh cDNA was amplified as a 983-bp product using the primer pair 5'-CATGTAGGCCATGAGG TCCACCAC-3' and 5'-TGAAGGTCGGTGTGAACGGATTTGGC-3'.

Western blotting
Testes were homogenized in TNE buffer [10 mM Tris-HCl, pH 7.5, 1 mM EDTA, 150 mM NaCl, 0.1% NP-40, and Protease Inhibitor Cocktail III (Sigma-Aldrich, St. Louis, MO)]. BMT-10 cells transfected with pGTSF2 were lysed directly using TNE buffer. Each lysate was mixed with a half-volume of 3x SDS Sample Buffer (New England Biolabs, Beverly, MA) and 1/ 30-volume of 1 M DTT, and heated at 99°C for 5 min. A total of 10 μg protein was loaded per lane, separated on a 15% SDS-polyacrylamide gel, and then transferred onto an Immobilon-P membrane (Millipore, New Bedford, MA). The membranes were blocked with 3% skim milk in TTBS (10 mM Tris-HCl, pH 7.6, 137 mM NaCl, and 0.1% Tween 20) for 1 h at room temperature, incubated overnight at 4°C with anti-GTSF1L (1:2500) or anti-GTSF2 (1:1000) in Alignment of the amino-acid residues of the mouse UPF0224-family proteins using CLUSTAL-X. The first and second CHHC Zn-finger domains are indicated in blue and red, respectively. Asterisks indicate amino-acid residues that are conserved among GTSF1, GTSF1L, and GTSF2. Pluses indicate amino-acid residues that are conserved between GTSF1L and GTSF2.
blocking solution, washed three times in TTBS, and incubated with horseradish peroxidaseconjugated goat anti-rabbit Ig (1:2000; Dako, Kyoto, Japan) in blocking solution for 1 h at room temperature. The membranes were again washed three times in TTBS, and signals were detected with the ECL Western Blotting Detection Kit (GE Healthcare).

Gtsf1l-and Gtsf2-null mice
To generate a Gtsf1l-null mouse, we used the CRISPR/Cas9 system and a double-nicking strategy [15,16]. Two plasmids containing the expression cassettes for hCas9 (D10A) and sgRNA were constructed by ligating two annealed single-stranded oligo DNAs including the CRISPR target sequence (S1A Fig) into the BbsI site of the pX335 vector (www.addgene.org/CRISPR). Cryopreserved pronuclear-stage BDF1 eggs (CLEA Japan, Tokyo, Japan) were thawed and cultured for 10 min in EmbryoMax KSOM Medium with 1/2 Amino Acids (Cat. #MR-106-D; Merck-Millipore, Darmstadt, Germany). The two pX335 plasmids (5 ng/μl each) were microinjected into the pronuclei of each egg, the eggs were cultured in the same medium overnight, and two-celled embryos were transferred into the oviducts of pseudopregnant female ICR mice. To genotype founder mice, genomic fragments containing the CRISPR target sequence were PCR-amplified using PrimeSTAR MAX (Takara, Shiga, Japan) and sequenced.

Histological analysis
For histology, testes were fixed overnight with 4% paraformaldehyde in PBS at 4°C, washed in PBS, embedded in paraffin, and sectioned at 5 μm. The sections were prepared and stained with hematoxylin and eosin.

Bisulfite methylation analysis
Genomic DNA isolated from the testes of 8-week-old mice was treated with bisulfate using the Epitect Bisulfite Kit (Qiagen, Germantown, MD). A long terminal repeat (LTR) region of IAP on chromosome 3qD was arbitrarily selected for analysis by nested PCR. The 5' region of the L1-type Gf was amplified by specific primers [5]. The products were cloned with the TOPO TA Cloning Kit (Invitrogen), and the resulting single colonies were picked and grown. Plasmid DNAs were extracted from the colonies, and the insert DNA of each plasmid was sequenced.

Results and Discussion
The two N-terminal CHHC-domain repeats of GTSF1L and GTSF2 are highly homologous GTSF1L and GTSF2 are UPF0224 proteins, which contain two N-terminal copies of a CHHCtype Zn-finger domain (Fig 1A and 1B); this domain is predicted to recognize and bind RNA [19]. Amino-acid sequence alignment of mouse GTSF1, GTSF1L, and GTSF2 proteins ( Fig  1B) showed that the N-terminal regions containing these Zn-finger domains were highly homologous in these three UPF0224-family genes, whereas the homology of the C-terminal regions, whose function is unknown, was low. [11]. Here, we analyzed the Gtsf1l and Gtsf2 mRNA expression in embryonic and postnatal testes by RT-PCR (Fig 2A). Gtsf1l transcripts were detected earlier than embryonic day (E)15.5, and from postnatal day (P)22 onward. Gtsf2 transcripts were detected from E16.5 to E18.5, and from P20 onward. Thus, the expression of both genes decreased temporarily around the time of birth. In the embryonic testes, Gtsf1l and Gtsf2 were expressed sequentially and alternately. In the postnatal period, these genes began to be expressed at different time points: Gtsf2 at P20 and Gtsf1l at P22. Because the first wave of spermatogenesis reaches the stage of round spermatids at P20 [20], these results suggest that the Gtsf1l and Gtsf2 mRNA expression is initiated at the stage of round spermatids or later. To determine the GTSF1L and GTSF2 expression patterns during spermatogenesis, we analyzed adult mouse testes by immunofluorescence using antibodies against the C-terminal portions of GTSF1L and GTSF2 (Fig 1A). We found distinct GTSF1L and GTSF2 signals in the inner space of the seminiferous tubules (Fig 2B-2K). Spermatogonia, spermatocytes, or round spermatids were not stained for GTSF1L (Fig 2B-2F). The cytoplasm of early elongating spermatids was strongly stained for GTSF1L at stage VII-X, but the staining decreased in the late elongating spermatids at stage XI-XII. On the other hand, the GTSF2 staining was not detected in spermatogonia, spermatocytes, early round spermatids, or late elongating spermatids, but was clearly seen in the cytoplasm of late round spermatids at stage IV-VI (Fig 2G-2K). GTSF2 staining decreased in the early elongating spermatids at stage VII-VIII. These results indicated that the GTSF2 and GTSF1L proteins are expressed sequentially and transiently at specific stages of spermatogenesis in the postnatal testes. Since proteins that are exclusively expressed in spermiogenic cells are rare, the expression patterns of GTSF2 and GTSF1L make them useful specific markers: GTSF2 for late round spermatids and GTSF1L for early elongating spermatids.

GTSF1L and GTSF2 interact with the MIWI, MILI, and TDRD1 complexes
Recently, Drosophila Gtsf1 was reported to associate with PIWI proteins via its central region [21,22]. Furthermore, we found that mouse GTSF1 associates with the MIWI, MILI, and MIWI2 complexes via its central region (unpublished). The timing of expression of the MIWI, MILI, TDRD1, and TDRD6 complexes partially overlaps that of GTSF1L and GTSF2. Therefore, to determine whether GTSF1L and GTSF2 interact with the MIWI, MILI, TDRD1, or TDRD6 complex, we performed pull-down experiments from mouse testis lysates using recombinant GTSF1L and GTSF2 proteins fused to GST. We found that full-length GTSF1L and GTSF2 pulled down MIWI, MILI, or TDRD1, but not TDRD6 (Fig 3B). Further analysis showed that GTSF1L and GTSF2 associated with these complexes via an N-terminal Zn-finger domain, rather than through a central region like GTSF1. This difference in the interacting domain may suggest functional differences between GTSF1 and GTSF1L or GTSF2. To clarify whether this interaction depended on RNA, we performed pull-down assays using RNase A. We found that GTSF1L's pull-down of the MIWI, MILI, and TDRD1 complexes depended on RNA (Fig 3B, left panel), but GTSF2's pull-down was independent of RNA ( Fig 3B, right  panel).
Next, to reveal whether Gtsf1l and/or Gtsf2 are involved in suppressing retrotransposons, we examined the retrotransposon expression and CpG DNA methylation in the retrotransposon promoter regions. First, we examined the methylation by bisulfate sequencing analysis (S4A Fig). Next, we examined the retrotransposon expression by RT-qPCR and immunofluorescence (S4B-S4F Fig). Our results indicated that the loss of Gtsf1l and/or Gtsf2 does not affect retrotransposon suppression.  This study found that although GTSF1L and GTSF2 associate with the PIWI and Tudor complexes, Gtsf1l and Gtsf2 are dispensable for spermatogenesis and retrotransposon suppression. Considering the specific expression patterns of Gtsf1l and Gtsf2 and the structure of their products, these genes may play some roles in spermatogenesis, but may be required only in specific circumstances or genetic backgrounds. The homology between GTSF1L and GTSF2 is high (approximately 78%) in the N-terminal regions (ZnF; Fig 3A), but low (approximately 30%) in the C-terminal regions (dZnF; Fig 3A). These C-terminal regions may contribute to the difference of GTSF1L and GTSF2 in the affinity to other molecules and/or the functions. Examining the effects of Gtsf1l and/or Gtsf2 overexpression may further clarify the roles of GTSF1L and GTSF2 in spermatogenesis in mice. ) and the 5.4-kb IΔ1-type IAP in chromosome 3qD were arbitrarily analyzed using previously described PCR primers [5]. Filled and open circles represent methylated and unmethylated CpGs, respectively; the percentages of methylated CpGs are shown. (B) Quantitative RT-PCR analysis of Line-1 and IAP expression in the testes from 8-week-old Gtsf1l +/-/Gtsf2 +/-, Gtsf1l -/-/Gtsf2 +/-, Gtsf1l +/-/ Gtsf2 -/-, and Gtsf1l -/-/Gtsf2 -/mice; data were normalized to the Gapdh expression. Error bars indicate standard deviation. (C-F) Frozen sections of the adult testes from 8-week-old Gtsf1l +/+ /Gtsf2 +/+ (C, E) and Gtsf1l -/-/Gtsf2 -/-(D, F) mice were immunostained using anti-L1ORF1p (C, D) and anti-IAP GAG (E, F) antibodies (green). Nuclei were stained with DAPI (blue). Scale bar: 500 μm. (TIF)