IP6K1 is essential for chromatoid body formation and temporal regulation of Tnp2 and Prm2 expression in mouse spermatids

Inositol hexakisphosphate kinases (IP6Ks) are evolutionarily conserved inositol phosphate kinases that generate the small molecule messenger, 5-diphosphoinositol pentakisphosphate (5PP-IP₅ or 5-IP₇) from inositol hexakisphosphate (IP₆) (Saiardi et al., 1999; Thomas and Potter, 2014). There are three mammalian IP6K isoforms, IP6K1, IP6K2 and IP6K3, which have a highly conserved C-terminus and a more diverse N-terminus (Wang et al., 2014). IP6K1 and IP6K2 are expressed ubiquitously with varying levels of expression in different tissues (Saiardi et al., 1999), whereas IP6K3 is highly expressed in the cerebellum and skeletal muscle (Saiardi et al., 2001; Fu et al., 2015; Moritoh et al., 2016). Genomic deletion of Ip6k1 in mice leads to a dramatic loss of male fertility (Bhandari et al., 2008), but Ip6k2- and Ip6k3-knockout mice breed normally (Morrison et al., 2009; Fu et al., 2015; Moritoh et al., 2016). The specific occurrence of male sterility in Ip6k1-knockout mice, but not in mice devoid of the other IP6K isoforms, suggests a distinct role for IP6K1 in spermatogenesis.

Spermatogenesis is a complex and dynamic process during which germ cells develop in the seminiferous tubules of the testis, with Sertoli cells providing them with structural and metabolic support. The process of spermatogenesis involves three phases – mitotic, meiotic and haploid. During the mitotic phase, spermatogonial stem cells self-renew and produce spermatogonia that proliferate and differentiate, eventually generating primary spermatocytes. Primary spermatocytes undergo two successive meiotic divisions to produce haploid round spermatids. Round spermatids undergo a lengthy phase of spermatid differentiation called spermiogenesis. During this process, round spermatids go through complex biochemical and morphological changes resulting in formation of the acrosome and sperm tail, removal of excess cytoplasm and extensive nuclear condensation (Russell et al., 1990). Mature spermatids are released from the testis and are stored in the epididymis prior to ejaculation.

In mice, the cycle of the seminiferous epithelium is divided into 12 stages (stages I–XII). Each stage represents a particular arrangement of different types of developmentally synchronised germ cells as seen in a cross-section of a seminiferous tubule (Russell et al., 1990; Ahmed and de Rooij, 2009). Spermiogenesis is subdivided into 16 steps (steps 1–16) based on changes in acrosome structure and nuclear compaction of maturing spermatids (Russell et al., 1990; Ahmed and de Rooij, 2009). Spermiogenesis is accompanied by extensive chromatin reorganisation during which most of the nucleosomal histones are initially replaced by transition proteins (TNP1 and TNP2) and subsequently by protamines (PRM1 and PRM2) (Rousseaux et al., 2008). Differentiation of round spermatids into compact elongated spermatids makes the latter transcriptionally inactive, leading to global gene silencing due to extensive chromatin condensation (Braun, 1998). Genes that are required during the final steps of spermiogenesis are transcribed in round spermatids and their mRNAs are stored in a translationally repressed state until their timely translation is initiated (Braun, 1998). Premature expression of TNP2, PRM1 or PRM2 has been associated with defective spermatid morphology and infertility in mice (Lee et al., 1995; Tseden et al., 2007; Fukuda et al., 2013).

While previous histologic examination of Ip6k1^−/− mice revealed few advanced spermatids in the testes and no sperm in the epididymis, the functional basis for spermiogenesis failure in these mice was not explored (Bhandari et al., 2008). In this study, we show that IP6K1 is highly expressed in round spermatids. Loss of IP6K1 leads to failure in spermatid differentiation due to defects in elongation and condensation of sperm nuclei. We found that in round spermatids, IP6K1 is enriched in a perinuclear ribonucleoprotein complex called the chromatoid body, and noted that the chromatoid body is absent or fragmented in Ip6k1^−/− spermatids. We also observed premature synthesis of TNP2 and PRM2 in round spermatids from juvenile Ip6k1^−/− mice, suggesting that IP6K1 is an important component of the post-transcriptional regulatory machinery of testicular germ cells, and therefore an indispensable regulator of male germ cell differentiation.

When Ip6k1^−/− female mice were mated with Ip6k1^−/− male mice, no pregnancies were observed, but when the same females were mated with heterozygous male mice, pups were born according to Mendelian ratios (Bhandari et al., 2008), indicating that IP6K1 is necessary for gametogenesis in males but not in females. On pairing adult Ip6k1^−/− males with wild-type females, we found that these mice could successfully mount the females and produce coital plugs, suggesting that defects in their male sexual behaviour are not responsible for their infertility. We dissected and compared the testes and epididymides of adult (2–3 months old) Ip6k1^+/+ and Ip6k1^−/− male mice. Testes and epididymides are smaller in Ip6k1^−/− mice compared to those in their Ip6k1^+/+ littermates, but do not show any gross morphological defects (Fig. 1A). To determine whether the reduction in epididymis and testis size can be attributed to the lower body weight of Ip6k1^−/− mice (Bhandari et al., 2008), we normalised the average epididymis or testis weight of each mouse to its body weight, and noted that this ratio was reduced by ∼30% for both tissues in Ip6k1^−/− compared to that in Ip6k1^+/+ mice (Fig. 1B,C). A reduction in serum testosterone levels has been shown to correlate with reduced weight of the testis and epididymis (Xu et al., 2007; Luria et al., 2009), but we observed no significant difference between serum testosterone levels in Ip6k1^+/+ and Ip6k1^−/− mice (Fig. 1D). Histological examination of epididymides revealed densely packed mature spermatozoa in Ip6k1^+/+ mice, but abnormal round cells in the epididymidal lumen of Ip6k1^−/− mice (Fig. 1E). Germ cells isolated from the epididymal lumen were examined more closely through DAPI staining to reveal mature spermatozoa in Ip6k1^+/+ mice, and primarily round cells with a few abnormal elongated spermatozoa in Ip6k1^−/− mice (Fig. 1F). This was reflected as a dramatic decrease in the sperm count in Ip6k1^−/− mice (Fig. 1G). TUNEL staining of epididymal sections from Ip6k1^−/− mice revealed that nearly all the germ cells in the epididymal lumen carry fragmented DNA, suggesting that they are undergoing degeneration (Fig. 1H). These results clearly demonstrate that male infertility associated with the loss of IP6K1 is primarily due to azoospermia, that is, the absence of mature spermatozoa in the epididymis.

Ip6k1 is known to be actively transcribed in a variety of mouse tissues but its transcript levels are specifically enriched in testis and brain (Saiardi et al., 1999). To gain further insight into the role of IP6K1 in male fertility, we examined its expression profile during the course of sperm development in mouse testes. Testes collected from 14-, 18-, 28-, 35- and 60-day-old mice were subjected to immunoblotting to detect IP6K1 levels (Fig. 2A). 14 day postpartum (dpp) testes contain Sertoli cells, spermatogonia, leptotene, zygotene and early pachytene spermatocytes; 18 dpp testes possess spermatogonia, early and late pachytene spermatocytes; 28 ddp testes contain round spermatids and early elongating spermatids in addition to these germ cells, and 35 dpp and adult (60 dpp) testes contain all germ cells including elongated spermatids (Bellvé et al., 1977; Kluin et al., 1982). IP6K1 was detected at low levels in 14 dpp testes; its expression increased significantly in 18 dpp testes and continued to rise during later stages of sperm development, suggesting that this protein is more abundant in late meiotic and post-meiotic germ cells (Fig. 2A,B).

To identify the precise cell types expressing Ip6k1, we performed immunofluorescence to detect IP6K1 in testis sections from Ip6k1^+/+ mice, with a Ip6k1^−/− section serving as the negative control (Fig. 2C,D). IP6K1 was localised to the cytoplasm of late pachytene (Fig. 2Ci) and diplotene spermatocytes (Fig. 2Cii). In post-meiotic round spermatids, IP6K1 was expressed at high levels in the cytoplasm and appeared concentrated in cytoplasmic granules (Fig. 2Ciii). We did not detect intense IP6K1 staining in Sertoli cells, spermatogonia, leptotene, zygotene, or early pachytene spermatocytes (Fig. 2C). However, as 14 dpp tubules contain only these cell types (Bellvé et al., 1977), it is likely that some of these express low levels of IP6K1. This possibility is supported by the observation that the regions of the Ip6k1^+/+ sections containing these cells (Fig. 2C) show a higher immunofluorescence signal than the background observed in Ip6k1^−/− sections (Fig. 2D). The increase in IP6K1 levels in 60 dpp testes (Fig. 2A,B) may be due to its higher expression in round spermatids which constitute the major population of testicular cells in adult testis. The high expression level of IP6K1 in late spermatocytes and round spermatids (Fig. 2E) is commensurate with a role for this protein in spermiogenesis.

Histological examination of stage X/XI, and stage I tubules from 2-month-old Ip6k1^+/+ and Ip6k1^−/− mice revealed no obvious differences in the composition and arrangement of cells (Fig. S1A). Although elongating spermatids are present in the earlier developmental stages, stage VIII tubules, which normally contain mature spermatids close to the lumen, showed no condensed spermatids in Ip6k1^−/− testes (Fig. S1A), suggesting that elongating spermatids are lost during development in these mice. Germ cell layers towards the basement membrane of seminiferous tubule cross-sections were largely unaltered between Ip6k1^+/+ and Ip6k1^−/− mice, with no detectable changes in spermatogonia and spermatocytes (Figs S1A and S2). This observation was corroborated by flow cytometry analysis of DNA content in isolated testicular cells. Ip6k1^+/+ testicular cells revealed four different peaks corresponding to fully condensed elongated spermatids (HC), round spermatids (1C), spermatogonia and somatic cells (2C), and primary spermatocytes (4C) (Krishnamurthy et al., 2000) (Fig. S1B). In Ip6k1^−/− testicular cells, the percentage of elongated spermatids (HC) was drastically reduced (Fig. S1B,C) and there was a proportional increase in the 4C and 2C cell populations. Interestingly, the Ip6k1^−/− HC peak could not be clearly resolved from 1C, which is suggestive of reduced DNA compaction in these spermatids (Fig. S1B).

To closely examine the development of Ip6k1^−/− spermatids, we identified the 16 developmental steps of spermiogenesis based on the shape of the nucleus and acrosome (Russell et al., 1990), by co-staining testis sections with DAPI and peanut agglutinin (PNA), a lectin that binds to glycoconjugates on the outer acrosomal membrane (Aviles et al., 1997) (Fig. S2). Step 1 to 12 spermatids are observed in stages I through XII, and step 13 to 16 spermatids are seen in stages I through VIII of the seminiferous epithelium (Fig. 2E). Analysis of adult stage XI seminiferous tubules revealed that, in Ip6k1^−/− mice, the round spermatids advance to step 10–11 elongating spermatids, but their nuclear morphology is abnormal (Fig. 3A). By stage XII (step 12) Ip6k1^+/+ spermatids attain their condensed comma shape, whereas Ip6k1^−/− tubules show irregularly shaped decondensed spermatid nuclei similar to those seen in stage XI (Fig. 3B). By step 14 (stage II–III), Ip6k1^+/+ spermatid nuclei were further condensed but Ip6k1^−/− spermatids remained decondensed (Fig. 3C). Finally, as observed in the histological sections (Fig. S1A), stage VIII (step 16) fully condensed ready-to-release spermatids were seen in Ip6k1^+/+ tubules, but were entirely absent in Ip6k1^−/− mice (Fig. 3D). Stage VIII tubules also show step 8 round spermatids with a fully developed acrosome, which appear identical in Ip6k1^+/+ and Ip6k1^−/− mice. This was observed more clearly in isolated round spermatids stained with PNA (Fig. S3A).

For a closer examination of spermatid morphology, we isolated elongated spermatids corresponding to steps 13 to 16 of spermatid differentiation by undertaking transillumination-assisted microdissection of seminiferous tubules, and stained them with DAPI to detect their nuclei. Ip6k1^−/− spermatids displayed irregularly shaped heads and a bent or blunt apex, lacking the typical hook-shaped appearance of Ip6k1^+/+ spermatids (Fig. 3E). Consistent with this, transmission electron microscopy of elongating/elongated spermatids revealed less condensed and loosely packed deformed nuclei with uneven density in Ip6k1^−/− spermatids (Fig. 3F,G) in comparison to tightly packed and homogeneously condensed Ip6k1^+/+ spermatids (Fig. 3H,I). Misshapen sperm heads have also been shown to arise due to anomalies in the formation of the manchette, a transient tubulin-rich structure formed around the nucleus in step 10–12 elongating spermatids (Kierszenbaum et al., 2007). We stained testis sections to detect α-tubulin, which marks the manchette, and noted normal formation of this structure despite the abnormal shape of Ip6k1^−/− spermatid nuclei (Fig. S3B). Taken together, these data suggest that improper nuclear condensation of Ip6k1^−/− elongating spermatids is not due to defects in acrosome or manchette formation, and may arise due to deficiencies in sperm DNA condensation.

To follow the process of DNA condensation during spermiogenesis, we tracked the presence of histone H4 in Ip6k1^+/+ and Ip6k1^−/− elongating spermatids (Fig. S4A–C). As expected, histone H4 was visible in both Ip6k1^+/+ and Ip6k1^−/− step 10–11 (stage X/XI) early elongating spermatids (Fig. 4A). As Ip6k1^+/+ spermatids advanced to step 12 (stage XII), histone H4 was no longer visible, but Ip6k1^−/− spermatids in stage XII and II–IV tubules contained histone H4, suggesting that these spermatids do not progress beyond step 11 (Fig. 4B,C). Elongating spermatids in stage II–III Ip6k1^−/− tubules also stained positive for in situ TUNEL labelling, and showed the presence of cleaved caspase 3, suggesting that they undergo apoptosis and are eventually lost (Fig. 4D,E).

A close look at the pattern of expression of IP6K1 in round spermatids reveals that it is enriched in granules within the cytoplasm (Fig. 2Ciii). These granules are reminiscent of chromatoid bodies, which are perinuclear ribonucleoprotein complexes found in round spermatids at steps 1 through 8, and are involved in mRNA translational control, mRNA decay and small RNA-mediated gene regulation (Parvinen, 2005; Kotaja and Sassone-Corsi, 2007; Meikar et al., 2011). Co-staining of squash preparations of Ip6k1^+/+ round spermatids to detect IP6K1 and the chromatoid body marker protein MIWI (also known as PIWIL1) revealed that IP6K1 is indeed localised to the chromatoid body in round spermatids in addition to being present in the cytoplasm (Fig. 5A). We therefore examined the architectural integrity of the chromatoid body in Ip6k1^+/+ and Ip6k1^−/− spermatids. On staining testis sections to detect MVH (also known as DDX4), an RNA helicase enriched in the chromatoid body, a conspicuous perinuclear spot was observed in Ip6k1^+/+ round spermatids (Fig. 5B). However, most Ip6k1^−/− spermatids did not show any chromatoid body staining, with some spermatids showing tiny fragmented spots (Fig. 5B). MVH and MIWI are the two most abundant proteins in the chromatoid body, constituting ∼40% (mole percentage) of the chromatoid body proteome (Meikar et al., 2014). We noted that the protein level of MIWI is unaltered and that of MVH is marginally increased in Ip6k1^−/− testes, suggesting that the disruption of the chromatoid body in Ip6k1^−/− mice is not due to reduced levels of these proteins (Fig. 5C,D). The ultrastructure of the chromatoid body was further examined by transmission electron microscopy. In contrast to massive sponge-like perinuclear aggregates observed in Ip6k1^+/+ spermatids, Ip6k1^−/− spermatids showed altered chromatoid body morphologies (Fig. 5E). While a few Ip6k1^−/− spermatids contained an intact chromatoid body, in most of these spermatids the chromatoid body was either fragmented or absent (Fig. 5F). These data demonstrate that IP6K1 has a critical function in the assembly of the chromatoid body, which in turn has been shown to be essential for normal spermatid development (Vasileva et al., 2009; Tanaka et al., 2011).

The temporal regulation of sperm DNA condensation is dependent on the timely expression of transition proteins and protamines. The mRNAs encoding these proteins are transcribed in round spermatids, but are stored in a translationally repressed state to be translated during spermatid elongation (Kistler et al., 1996; Braun, 1998; Kleene, 2013). Translational control of mRNAs in round spermatids involves the interaction of cis-elements in the mRNA with RNA-binding proteins and microRNAs, and differs for individual mRNAs (Kleene, 2013). The role of the chromatoid body in translational repression of spermatid mRNAs has been debated (Parvinen, 2005; Kotaja and Sassone-Corsi, 2007; Kleene and Cullinane, 2011). mRNAs encoding TNP2 and PRM2 have been shown to localise to the chromatoid body (Saunders et al., 1992; Fukuda et al., 2013), and disruption of the chromatoid body has been shown to cause male infertility and can also lead to aberrant cytoplasmic aggregation of TNP2 and PRM2 (Reuter et al., 2011; Tanaka et al., 2011). It has previously been shown that Prm2 mRNA is expressed in 28 dpp juvenile mouse testes, but significant levels of PRM2 protein can be detected only at 32 dpp (Fukuda et al., 2013). Therefore, to investigate whether the loss of the chromatoid body in Ip6k1^−/− spermatids leads to changes in TNP2 and PRM2 expression, we examined the levels of these proteins in 28- and 35-day-old juvenile testes. At 28 dpp, neither protein was detected in Ip6k1^+/+ testes whereas Ip6k1^−/− testes showed premature expression of both these proteins (Fig. 6A,B). TNP2 and PRM2 expression was detected in both Ip6k1^+/+ and Ip6k1^−/− 35 dpp testes, but protein levels were higher in knockout mice. In contrast to protein levels, there was either a decrease or no change in the levels of Tnp2 and Prm2 transcripts in 28 and 35 dpp Ip6k1^−/− compared with that seen in Ip6k1^+/+ testes (Fig. 6C). This suggests that the premature expression of these proteins is not attributable to an increase in mRNA abundance, but instead is a result of derepression of their translational silencing. Testicular sections of 28 dpp mice revealed that TNP2 is highly expressed in the nuclei of a few abnormally condensed Ip6k1^−/− spermatids (Fig. 6D), and PRM2 is prematurely expressed in the nuclei of most of the Ip6k1^−/− round spermatids (Fig. 6E). In adult mice, PRM2 showed uniform nuclear staining in Ip6k1^+/+ spermatids, but weak nuclear staining in some Ip6k1^−/− elongating spermatids and aberrant cytoplasmic localisation in the form of large aggregates around spermatid nuclei (Fig. 6F). The seminiferous tubules of 6-month-old Ip6k1^−/− mice harboured prematurely condensed round spermatids that were positive for expression of TNP2 and PRM2 (Fig. 6G). Similar patterns of aberrant TNP2 and PRM2 expression and localisation have been observed in Miwi- and Tdrd6-knockout mice, which exhibit defects in chromatoid body assembly (Reuter et al., 2011; Tanaka et al., 2011). Taken together, these data suggest that IP6K1 is essential for the post-transcriptional regulation of Tnp2 and Prm2 mRNA expression in round spermatids.

This study investigates the underlying basis of spermatogenic arrest in mice lacking the inositol phosphate kinase IP6K1. Our data show that IP6K1 is essential for histone removal and sperm head elongation during spermatid differentiation. We show that IP6K1 is a novel and essential component of chromatoid bodies, which are either disintegrated or absent in round spermatids lacking IP6K1. We also observe premature translation of TNP2 and PRM2 in round spermatids of juvenile mice, and aberrant localisation of these proteins in adult mice, leading to abnormal DNA condensation in elongating spermatids (Fig. 7). These deformed elongating spermatids ultimately undergo apoptosis, which is responsible for azoospermia and male infertility in Ip6k1^−/− mice.

To date, IP6K1 is the only inositol polyphosphate metabolic enzyme shown to participate in mammalian gametogenesis. Another Ip6k1 mutant mouse strain generated by gene trapping technology introduced a retroviral insertion between exons 2 and 3 (coding exons 1 and 2), resulting in the loss of Ip6k1 transcript (http://www.informatics.jax.org/allele/key/36801). Preliminary histological analysis of these male mice revealed bilateral epididymal azoospermia and testicular degeneration, suggesting that they would display male sterility, similar to what we observe upon deletion of the terminal exon 6. Deletion of the other IP6K isoforms, IP6K2 and IP6K3, has no effect on spermatogenesis (Morrison et al., 2009; Fu et al., 2015; Moritoh et al., 2016), indicating that the different IP6K isoforms have non-overlapping functions in mammalian reproductive physiology. Treatment of mice with the pan-IP6K inhibitor TNP was shown to have no effect on male fertility, which was attributed to the inability of this drug to penetrate the blood–testis barrier (Ghoshal et al., 2016).

In mammals, the synthesis of 5-IP₇ relies on a cascade of metabolic enzymes, starting with the release of I(1,4,5)P₃ from phosphatidylinositol 4,5-bisphosphate [PI(4,5)P₂] (Fig. S4D). I(1,4,5)P₃ is converted into I(1,3,4,5)P₄ by inositol 1,4,5-trisphosphate 3-kinases (IP3Ks) of which there are three isoforms in mammals (Leyman et al., 2007). I(1,3,4,5)P₄ is subsequently converted into I(1,3,4,5,6)P₅ by inositol polyphosphate multikinase (IPMK) (Fujii and York, 2005), and I(1,3,4,5,6)P₅ is converted into IP₆ by inositol-pentakisphosphate 2-kinase (IPPK) (Verbsky et al., 2005a), both of which have only single isoforms in mammals. In an alternative pathway, I(1,3,4,5)P₄ is first dephosphorylated to I(1,3,4)P₃ and subsequently phosphorylated by ITPK1 to yield I(1,3,4,6)P₄, which is then phosphorylated to I(1,3,4,5,6)P₅ by IPMK (Verbsky et al., 2005b). Targeted mutant mice have been generated for all of the enzymes responsible for the synthesis of inositol phosphate precursors to 5-IP₇. Loss of any of the three IP3K isoforms did not alter viability or fertility in mice (Jun et al., 1998; Pouillon et al., 2003). However, mouse embryonic fibroblasts lacking all three IP3K isoforms have been shown to synthesise IP₄, IP₅ and IP₆, making use of a redundancy in the metabolic pathway that allows IPMK to take over the function of IP3Ks (Leyman et al., 2007) (Fig. S4D). A hypomorphic mutant for ITPK1 is viable and fertile (Wilson et al., 2009). Although the levels of inositol polyphosphates were not measured in cells derived from these mice, it is known that IPMK can catalyse the synthesis of I(1,3,4,5,6)P₅ from I(1,3,4,5)P₄ in mammalian cells, suggesting that ITPK1 is not essential for the synthesis of IP₆ (Fujii and York, 2005). Mice lacking IPMK die at around embryonic day 9.5, and cells derived from these embryos have negligible levels of IP₅ and IP₇, and an ∼90% reduction in IP₆ (Frederick et al., 2005; Maag et al., 2011). Homozygous deletion of IPPK, the only enzyme responsible for IP₆ synthesis, led to early embryonic lethality and no embryos could be derived for inositol polyphosphate analysis (Verbsky et al., 2005a). IP₅ and IP₆ may have specific functions in spermatogenesis that are independent of their role as 5-IP₇ precursors. Male germ-cell-specific knockout mice for IPMK and IPPK will be required to address this possibility. Since no knockout mice have been generated for the two mammalian PPIP5K isoforms that are responsible for the synthesis of IP₈ [1,5(PP)₂-IP₄] from 5-IP₇ (Fig. S4D), the importance of this inositol pyrophosphate in male gametogenesis is unknown.

Our most interesting finding is the identification of IP6K1 as a novel protein component of the chromatoid body. The expression pattern of IP6K1 in male germ cells resembles that of other integral chromatoid body components such as MIWI and the tudor domain-containing proteins TDRD6 and TDRD7. These proteins, like IP6K1, are highly expressed in the cytoplasm of late pachytene cells and continue to be expressed at high levels in the cytoplasm and chromatoid body of round spermatids (Deng and Lin, 2002; Vasileva et al., 2009; Tanaka et al., 2011). The loss of TDRD6 or TDRD7, which regulate ribonucleoprotein remodelling in chromatoid bodies, leads to male-specific sterility due to abnormalities in chromatoid body formation (Vasileva et al., 2009; Tanaka et al., 2011). Loss of the MIWI RNase activity has been shown to result in a similar pattern of PRM2 aggregation around elongating spermatids to that we observed in Ip6k1^−/− mice (Reuter et al., 2011).

Precise timing and levels of transition protein and protamine expression are important for the progression of spermiogenesis, and premature translation of these proteins can lead to spermiogenic defects similar to those observed in Ip6k1^−/− mice (Lee et al., 1995; Tseden et al., 2007; Fukuda et al., 2013). Transgenic mice carrying Tnp2 with a modified 3′ untranslated region (UTR) that causes its premature expression in round spermatids display abnormal sperm head morphology and male infertility (Tseden et al., 2007). Mice lacking CBF-A (also known as HNRNPAB), an RNA-binding protein involved in temporal regulation of Prm2 translation, display precocious translation of PRM2 leading to premature chromatin condensation in a subset of elongating spermatids (Fukuda et al., 2013). Transcripts encoding TNP2 and PRM2 have been shown to localise to the chromatoid body in step 7–8 round spermatids (Saunders et al., 1992; Fukuda et al., 2013), but, to our knowledge, there are no studies examining whether disruption of the chromatoid body alters the post-transcriptional regulation of these transcripts. Our data reveals that chromatoid body disruption associated with loss of IP6K1 correlates with premature translation of Tnp2 and Prm2, suggesting that the chromatoid body localisation of these transcripts may be directly relevant to their translational repression. IP6K1 may also indirectly regulate translation of these transcripts independently of its involvement in chromatoid body formation.

It has been reported that Ip6k1 transcripts rank second amongst the top five most abundant transcripts in the chromatoid body, alongside transition proteins and protamines (Meikar et al., 2014). In the same study, functional classification of 88 chromatoid body-associated proteins revealed that they are predominantly involved in RNA binding and other RNA regulatory functions. IP6K1, on the other hand, is an inositol phosphate kinase, and possesses no obvious RNA-interacting regions. It is possible that 5-IP₇ synthesised by IP6K1 may be essential for the function of one or more chromatoid body proteins. Alternatively, IP6K1 may act independently of its catalytic activity as a scaffold protein to maintain architectural integrity of the chromatoid body.

All animal experiments were performed in compliance with guidelines provided by the Committee for the Purpose of Control and Supervision of Experiments on Animals, Ministry of Environment, Forest, and Climate Change, Government of India. All animal experiments were approved by the Institutional Animal Ethics Committee (Protocol number PCD/CDFD/02 – version 2). Mice (Mus musculus, strain C57BL/6) used for this study were housed in the Centre for DNA Fingerprinting and Diagnostics animal facility located within the premises of Vimta Labs, Hyderabad. The Ip6k1 gene knockout mouse was generated as previously described (Bhandari et al., 2008). This mouse carries a deletion spanning the splice site, coding region and a portion of the 3′ UTR of the terminal exon (exon 6) of Ip6k1. Ip6k1^+/+ and Ip6k1^−/− littermates were produced by breeding Ip6k1^+/− mice. Age-matched juvenile and adult 8- to 16-week-old male mice were used for the collection of testes. Images shown in the figures are representative of single experiments, but each experiment was repeated with n=3–5 Ip6k1^+/+ and Ip6k1^−/− mice, and similar results were obtained. The number of animals used to obtain quantitative data (n) is indicated in the figure legends. Data were graphed and statistical analyses were performed as indicated in the figure legends, using GraphPad Prism 5. Statistical significance was defined as P<0.05.

All chemicals were purchased from Sigma-Aldrich, unless otherwise specified. Primary antibodies and their respective dilutions used for immunofluorescence (IF) and immunoblotting (IB) were: anti-TNP2 (Santa Cruz Biotechnology, sc21106; 1:200 IF, 1:500 IB); anti-PRM2 (Santa Cruz Biotechnology, sc-23104; 1:200 IF, 1:500 IB); anti-IP6K1 (Sigma-Aldrich, HPA040825; 1:300 IF), anti-IP6K1 (Santa Cruz Biotechnology, sc-10419; 1:500 IB); anti-cleaved caspase 3 (Cell Signaling Technology, 9661; 1:400 IF); anti-histone H4 (Abcam, ab10158; 1:300 IF); anti-MIWI (Santa Cruz Biotechnology, sc-22685; 1:200 IF, 1:500 IB); anti-MVH (Abcam, ab13840; 1:300 IF, 1:1000 IB); anti-α-tubulin (Sigma-Aldrich, T9025; 1:500 IF, 1:5000 IB); and anti-GAPDH (Sigma-Aldrich, G8795, 1:5000 IB) antibodies. Alexa Fluor 594-conjugated PNA lectin (Molecular Probes, L32459) was used at a dilution of 1:500.

Epididymides were dissected and collected in Dulbecco's modified Eagle's medium, cut into pieces and flushed out to release spermatozoa. These were filtered through a 40 µm mesh to remove debris. The resuspended spermatozoa were then counted with a haemocytometer under a light microscope according to standard protocols.

Blood collected from the retro-orbital sinus of Ip6k1^+/+ and Ip6k1 ^−/− male mice was allowed to clot for 1 h at room temperature. The clots were centrifuged at 1100 g for 30 min at 4°C. Serum samples were collected and stored in sealed tubes at −80°C until assayed. Serum testosterone concentration was measured using an enzyme-linked immunosorbent assay (Enzo Life Sciences; ADI-900-065), following the manufacturer's guidelines. Serum samples diluted (1:10) in assay buffer were assayed in duplicate. The cross-reactivity of this assay is 14.6% for 19-hydroxytestosterone and 7.2% for androstenedione, but less than 1% for any other hormone or metabolite.

Ip6k1^+/+ and Ip6k1^−/− male mice of different age groups were euthanised by carbon dioxide inhalation, and the testes and epididymides were dissected. After isolation, testes and epididymides were briefly washed with phosphate-buffered saline (PBS) and fixed in Bouin's solution for 48 h at 4°C. To remove the last traces of fixative, samples were washed in 70% ethanol, dehydrated, and embedded in paraffin wax. 4 µm sections were prepared on glass slides, cleared in xylene and dehydrated in a graded series of ethanol. The sections were then stained with haematoxylin and eosin. Images were acquired with a bright-field light microscope (Nikon ECLIPSE Ni-U, NIS Elements acquisition software, 20×0.5 N.A. or 40×0.75 N.A. objectives). Stages of the seminiferous epithelia were identified according to established methods (Russell et al., 1990; Ahmed and de Rooij, 2009). Ip6k1^−/− sections were staged based on cells other than elongating or elongated spermatids.

Paraffin-embedded testis sections were deparaffinised in xylene, dehydrated in a graded ethanol series, boiled in 10 mM sodium citrate buffer (pH 6) in a microwave oven for 20 min to retrieve the antigen, washed in PBS, permeabilised in 0.5% Triton X-100 for 10 min, blocked in blocking buffer (5% BSA+0.1% Triton X-100 or 4% FBS+0.1% Triton X-100) for 1 h at room temperature (RT) and then incubated overnight at 4°C with primary antibodies diluted in blocking buffer. The slides were washed with PBS and stained with fluorophore-conjugated secondary antibodies, including Alexa Fluor 488-conjugated or Alexa Fluor 568-conjugated goat anti-rabbit-IgG (Molecular Probes), Alexa Fluor 488-conjugated goat anti-mouse-IgG (Molecular Probes), Alexa Fluor 488-conjugated donkey anti-goat-IgG (Santa Cruz Biotechnology) and Alexa Fluor 555-conjugated donkey anti-rabbit-IgG (Santa Cruz Biotechnology). All secondary antibodies were used at 1:500 dilution in blocking buffer and incubated at RT for 1 h. Finally, the slides were mounted in antifade mounting medium containing DAPI (H-1200, Vector Labs). Images were acquired with an LSM 510 (LSM acquisition software) or LSM 700 (Zen acquisition software) confocal microscope (Zeiss) equipped with 405, 488, 555/561 nm and two-photon lasers, and fitted with a 63×1.4 NA objective.

TUNEL staining was performed using an APO-BrdU™ TUNEL assay kit (Sigma-Aldrich, A35125) according to the manufacturer's instructions. Briefly, testis sections were deparaffinised as described above and then labelled with BrdU in the presence of terminal deoxynucleotidyl transferase (TdT) enzyme for 2 h at 37°C. After two washes, the sections were incubated with Alexa Fluor 488-conjugated anti-BrdU antibody supplied with the kit for 1 h at RT. The slides were mounted and imaged as described above.

Testes from Ip6k1^+/+ and Ip6k1^−/− male mice were dissected and briefly washed in chilled 1× PBS. Tunica albuginea were removed and seminiferous tubules were minced and collected in RIPA buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA, 10 mM NaF, 1 mM PMSF) containing freshly added protease inhibitor cocktail (Sigma-Aldrich). The tissues were subjected to homogenisation (D-1 Homogenizer-Disperser, Miccra) and sonication (Bioruptor, Diagenode). The extracts were centrifuged at 20,000 g for 15 min at 4˚C and supernatants were immediately mixed with 4× SDS sample buffer to a final dilution of 1×. The samples were heated at 95°C for 5 min, resolved by SDS-PAGE and transferred onto PVDF membranes (GE Healthcare). Western blotting was performed to detect the protein of interest and to assess equal gel loading by sequential probing of blots with experimental and control antibodies. Alternatively, when the difference in size between the experimental and control proteins was significant, a blot corresponding to the same gel was cut and probed in parallel. Membranes were blocked with 5% skimmed milk for 1 h at RT, followed by incubation with the primary antibody overnight at 4°C. After washing three times with TBST buffer (20 mM Tris-HCl pH 7.6, 150 mM NaCl and 0.1% Tween 20), membranes were incubated at RT for 1 h with HRP-conjugated secondary antibodies as follows: goat anti-rabbit IgG conjugated to HRP (Southern Biotech, 4010-05; 1:5000) for MVH and GAPDH; donkey anti-goat IgG conjugated to HRP (Santa Cruz, sc2020; 1:3000) for IP6K1, TNP2, PRM2 and MIWI; and goat anti-mouse IgG conjugated to HRP (Southern Biotech, 1031-05; 1:10,000) for α-tubulin. After washing the membrane three times with TBST, specific proteins bands were detected using a chemiluminescence reagent (ECL Prime, GE Healthcare). Chemiluminescence was detected using the FlourChem E (Protein Simple) documentation system. Where required, blots were stripped using Restore™ Western Blot Stripping Buffer (ThermoFisher Scientific), and re-probed with the control antibody. Densitometry analysis of bands was performed by using ImageJ software (Fiji) or the multiplex band analysis tool in AlphaView software (Protein Simple). The band intensity of the protein of interest was normalised to the loading control.

Transmission electron microscopy (TEM) was performed as previously reported (Chemes, 2013). All reagents used for TEM were obtained from Electron Microscopy Sciences unless specified otherwise. Testes from adult mice were excised and directly fixed in 4% glutaraldehyde in sodium cacodylate buffer (0.1 M, pH 7.4). After washing in a fresh solution of 4% glutaraldehyde, testes were cut into smaller pieces (0.5–1 mm) and then immersed in the same fixative for 3 h at RT. After washing with sodium cacodylate buffer, the tissues were again fixed in 2% osmium tetroxide for 3 h at RT. The testis pieces were then dehydrated through a graded series of ethanol (50, 70 and 96%) at 4°C. To achieve complete dehydration, tissues were immersed in 100% ethanol (three changes, 20 min each) followed by propylene oxide (three changes, 20 min each) at RT. After dehydration, tissues were initially infiltrated with a 1:1 mixture of propylene oxide and Embed–Araldite mix [1:1:3 mix of Embed 812:Araldite 502:dodecenyl succinic anhydride (DDSA)] containing 3% BDMP-30 for 2 h at RT and then treated with only Embed–Araldite mix at RT (two changes for 3 h each and one change overnight). Tissue blocks were finally polymerised in Embed–Araldite mix at 60°C for 72 h. Ultra-thin sections were cut on an ultra-microtome (Leica) and mounted on 200-mesh uncoated copper grids (Ted Pella Inc.). Sections were stained with 2% uranyl acetate replacement stain for 1 h and 0.5% lead citrate in distilled water for 30 min. The observations were performed with a Tecnai G2 Spirit BioTWIN transmission electron microscope (FEI).

The procedure followed was similar to that described earlier (Kotaja et al., 2004). 8-week-old male mice were euthanised as described above, and testes were removed and decapsulated and seminiferous tubules transferred to a Petri dish containing PBS. When viewed under a transilluminating dissection microscope, seminiferous tubules produce different patterns of light absorption giving four different regions; weak spot (stages XII–I), strong spot (stages II–VI), dark zone (stages VII–VIII) and pale zone (stages IX–XI). The dark zone containing elongated spermatids was cut into 1 mm-long pieces and placed in 50 µl of 100 mM sucrose. Elongated spermatids were obtained in suspension by gentle pipetting of the minced tissue for 1 min. Cells were dried on slides pre-dipped in 1% PFA containing 0.15% Triton X-100. Since dark zones were not clearly visible in Ip6k1^−/− tubules, cells in the region between the strong spot and pale zone were isolated and analysed. The dried slides were briefly washed in PBS followed by cell fixation in 4% PFA for 10 min, and mounted in antifade mounting medium containing DAPI (H-1200, Vector Labs). Images were acquired using an LSM 700 confocal microscope (Zeiss; Zen acquisition software; 405 nm laser) using a 63×1.4 NA objective.

Quantification of mRNA levels was carried out by reverse transcription real-time quantitative PCR (RT-qPCR). Briefly, total cellular RNA was isolated from 28 dpp and 35 dpp Ip6k1^+/+ and Ip6k1^−/− testes by using TRIzol reagent (Invitrogen) followed by a RNeasy Mini Kit (Qiagen) according to the manufacturers' instructions. 2 μg RNA was used for first-strand cDNA synthesis by reverse transcription with SuperScript Reverse Transcriptase III (Invitrogen) and oligonucleotide dT primers. qPCR was performed using gene-specific primers on the ABI 7500 real-time PCR system (Applied Biosystems) with MESA GREEN qPCR MasterMix Plus for SYBR® Assay Low ROX (Eurogentec) for detection. All samples were run in two or three technical replicates. The sequences of primers used were: Tnp2 forward, 5′-GCTCAGGGCGAAGATACAAGTG-3′ and reverse, 5′-TGTGACATCATCCCAACAGTCC-3′; Prm2 forward, 5′-GTAGGAGGCACCATCACTAAGC-3′ and reverse, 5′-AGACATCGACATGGAATGGTG-3′; and Gapdh forward, 5′-GGAGAAGGCCGGGGCCCACTTGAA-3′ and reverse: 5′-GCATGGACTGTGGTCATGAGCCCTTCCAC-3′. The difference in transcript levels was calculated using the fold change (ΔΔC_t method) (Schmittgen and Livak, 2008). ΔC_t is the C_t value for the gene of interest normalised to the C_t value of the respective Gapdh control in both Ip6k1^+/+ and Ip6k1^−/− testes. ΔΔC_t values were calculated as a relative change in ΔC_t of the target gene in Ip6k1^−/− with respect to Ip6k1^+/+ testes. Fold changes were expressed as 2^−ΔΔCt.

Flow cytometry analysis of testicular cells was performed as described previously (Krishnamurthy et al., 2000). Testes were dissected, the tunica albuginea was removed, and the seminiferous tubules were minced in Ca²⁺- and Mg²⁺-free PBS (Gibco). Cells were dispersed by gentle aspiration, filtered using a 40 µm nylon filter, and washed in PBS by centrifuging at 800 g for 5 min. The cells were re-suspended in PBS, fixed in 70% chilled ethanol, and stored at 4°C for 24 h or at −20°C for up to 1 week until further analysis. 1×10⁶–2×10⁶ ethanol-fixed testicular cells were washed with PBS and treated with 0.25% pepsin solution for 10 min at 37°C. Finally, cells were stained with propidium iodide (PI) staining solution (25 µg/ml PI, 40 mg/ml RNase and 0.03% Nonidet P-40 in PBS) at RT for 20 min. The PI-stained cells were analysed on an Accuri C6 flow cytometer (BD Biosciences; excitation 488 nm; emission 585/40 nm).

We sincerely thank Dr Solomon Snyder at the Johns Hopkins University School of Medicine for the gift of Ip6k1^+/− mice. We acknowledge the infrastructure and staff of the CDFD experimental animal facility, especially Suman Komjeti, Jayant P. Hole and Sridhar Kavela for help with breeding and maintenance of mice. Renu Pasricha and H. Krishnamurthy at the Centre for Cellular and Molecular Platforms (C-CAMP), Bangalore, are thanked for assistance in TEM imaging and discussions on flow cytometry, respectively. We acknowledge the assistance of Joy P. Vallentyne and A. L. Annapurna with confocal imaging. We are very grateful to Manasa Chanduri and Sitalakshmi Thampatty for manuscript proofreading and helpful suggestions. We thank all members of the Laboratory of Cellular Signalling for helpful comments.