Male infertility is a significant problem in all mammalian species [1]. In humans, it is estimated that 15% of couples are infertile and in one third of these cases infertility can be attributed solely to the male partner [2]. In productive livestock the economic losses due to reproductive inefficiency in males can be substantial, particularly when infertility affects a genetically superior individual [3]. Although some instances of male factor infertility can be explained by infections, environmental causes, immunological or hormonal deficiencies, many are caused by genetic factors. Problems with the production and maturation of spermatozoa are the most common causes of male infertility resulting in a low sperm count, morphologically abnormal spermatozoa or reduced sperm motility [4–6]. Despite efforts to reveal the genes involved in spermatogenesis and their functions, little is known about the underlying causes of male infertility. Therefore, the localization and identification of mutations affecting specifically the spermatogenesis provide invaluable information for research on the causes of male infertility.

Mammalian spermatogenesis is a complex process, where diploid spermatogonia develop into haploid, highly specialized spermatozoa. Spermatogenesis occurs in the seminiferous tubules of the testis and can be divided in proliferative phase, meiotic phase and differentiation (spermiogenesis) [7]. During the proliferative phase spermatogonia go through several mitotic divisions. The final mitotic division of differentiated spermatogonia gives rise to the primary spermatocytes. Meiosis of primary spermatocytes leads to the production of secondary spermatocytes after the first meiotic division, while haploid round spermatids are formed following the second meiotic division. After meiosis, spermatids are connected with cytoplasmic bridges sharing transcripts and proteins [8]. In spermiogenesis the nucleus of the germ cell is remodeled and the acrosome and sperm tail are formed. Finally, mature spermatozoa are released into the lumen of seminiferous epithelium and transported to the epididymis for further maturation. During the whole process, the expression and the interactions between various genes and their protein products are required and regulated in an ordered manner. Some of these genes or their alternative transcripts are specifically expressed in the testis. Identification of these genes and their roles is important in understanding the mechanism of spermatogenesis.

Spermatogenic arrest at various stages of spermatogenesis is known to cause infertility in several mammalian species [9–11], but only recently genes and interaction networks behind these defects have been under investigation. Different mouse models have revealed several genes associated with male infertility caused by spermatogenic arrest [12–15], but the full understanding of events leading to spermatogenic arrest at specific phases requires further studies. Within the Finnish Yorkshire population several boars with azoospermia were identified during years 1995-2008. In this study we report an infertility defect in the Finnish Yorkshire boars caused by spermatogenic arrest (SA) in early meiotic cells. The aim of the study was to identify the mutation causing the SA defect and reveal the role of the affected gene in spermatogenesis. The disease associated region was localized to porcine chromosome 12 within a 1.9 Mbp region and a promising candidate gene, Testis expressed 14 (Tex14), was identified. Sequencing of the Tex14 mRNA showed a SA specific insertion in exon 27, which resulted in a premature translation stop codon and decreased expression of the TEX14 mRNA and protein products specifically in the testis.

The SA affected boars show almost no late meiotic and postmeiotic germ cells and reduced amount of spermatogonia. This phenotype is very similar to Tex14 knockout mice phenotype and therefore the search for causal mutation was focused on Tex14 gene. In Tex14 knockout mice, initially markedly reduced numbers of late meiotic and post meiotic cells are detected and in older animals more severe phenotype with decrease in all germ cell numbers is present [15]. In addition, the vacuolization is increased in older animals in the mouse and was evident also in the pig. In mice, as spermatogonia divide and differentiate, TEX14 together with other midbody proteins form intercellular bridges, a stable structure formed by mitotic and meiotic divisions and maintained until formation of spermatozoa. TEX14 interaction with CEP55 has been shown to be critical for subverting abscission toward a stable intercellular bridge [19].

In Tex14 knockout male mice, germ cells continue to form midbodies during telophase of cytokinesis, but the midbody is transient, as in somatic cells, and no intercellular bridges are formed [17]. Thus, Tex14 knockout mice completely lack intercellular bridges, and spermatogenesis fails before the first meiotic division resulting in male infertility [15]. In Finnish Yorkshire boars the spermatogenic arrest at first meiotic division appears also to be caused by a mutation in the Tex14 gene. Although several polymorphisms were detected within the Tex14 coding frame, a 51 bp insertion in exon 27 appears to be the obvious cause for the defect. This insertion results in differential splicing of exon 27, which creates a premature translation stop codon after position 3858 nt in the SA affected Tex14 transcript. The expression of all examined Tex14 fragments were markedly down regulated in the SA affected boar and the protein product of 150 kDa was absent. The probable cause for the down regulation of the Tex14 mRNA is the nonsense mediated degradation of the incorrectly spliced mRNA, but we have not entirely excluded the possibility of down regulation of mRNA transcription or changes in cytoplasmic mRNA stability. The sequencing of the promoter region (152 bp upstream of the transcription initiation site) did not show any SA associated mutations indicating that the transcription is not affected by mutations in the promoter region, but regulatory regions further upstream have not been investigated. However, the protein product of TEX14 appears to be absent in the SA affected testis highlighting the role of this gene in development of the SA defect.

In the Finnish Yorkshire pigs no effect on female reproduction has been reported and no other symptoms except infertility in boars were detected. Thus, the insertion within Tex14 appears to only affect spermatogenesis. This has also been seen in mice. In Tex14 knockout females no significant difference in the number of germ cells at prenatal day 11.5 was detected and litter size was only slightly affected, although the intercellular bridges were missing [18]. In this study we have also analyzed the expression of Tex14 in the testis, ovary, oviduct and uterus of control pigs. Tex14 appears to be highly expressed in the testis and substantially lower expression or no expression was detected in the female reproductive organs indicating a more prominent role of TEX14 in the testis. This result is consistent with studies in the mouse showing that TEX14 is not essential for postnatal oocyte growth and ovarian folliculogenesis [18].

The ancestral SA insertion sequence locates in the pig genome within Tex14 intron 27 360 bp downstream from the SA insertion site. All affected boars also have the original sequence in intron 27, thus the disease causing fragment appears to be copied in the same orientation to exon 27. The SA insertion region was not identified as an interspersed repeat or a low complexity DNA sequence, thus it appears to be a unique duplication event in the porcine genome. In the human, indels cause multiple genetic diseases [20] and are a source of natural intra- and interspecific genetic variation [21–24]. Short unique sequences are being actively duplicated in mammalian genomes and are often separated by some distance [25]. It has been postulated that distantly located duplicates are generated by direct tandem duplications, and separated apart by later insertions [26]. However, the SA insertion into an existing target sequence was created with minimal collateral damage to the target. This separated tandem duplication may have arisen by recombination after duplication event, since duplications undergo high frequencies of recombination and are consequently unstable [25]. Duplications have been shown to attribute from replication slippage errors [27, 28], but also the significance of recombination to the formation of small indels, in particular insertions, has been demonstrated [29, 30]. However, these mechanisms require an existing duplicate in the DNA sequence [28]. In the pig, within the region containing the SA insertion sequence and original sequence within Tex14 intron 27 no duplication sequences were present in the wild type animals. One possible mechanism for the SA duplication event could also be the imperfect repair of DNA double-strand breaks by classical nonhomologous end joining (NHEJ) [31–33]. NHEJ requires only short microhomologies of 1-4 bp and can even ligate overhangs without homologies [28, 34].

The identification of the insertion within Tex14 exon 27 enables the gene assisted selection against the defect. Although the SA carrier status is fairly easy to determine on an agarose gel, we further examined the possibility to use SNPs on Illumina PorcineSNP60 beadchip for identification of the SA carriers and affected animals. Two SNPs (ALGA0066210 and ASGA0054360) were strongly associated with the SA defect and a haplotype of these SNPs appears to be a usable marker for marker assisted selection against the defect at least within the Finnish Yorkshire population.

Tex14 gene appears to be crucial for succesful spermatogenesis in the mouse and pig. The functional studies revealing the role of TEX14 in the intercellular bridges between developing male germ cells indicate the importance of this gene in sperm production across species. A statistically significant difference in Tex14 gene expression has also been identified in azoospermic men [35]. However, no statistically significant associations between Tex14 SNPs and azoospermia have been identified in this study or in other genome wide association analyses of azoospermic men [36]. The lack of statistically significant associations may be partly due to the heterogenity of azoospermic cases and low frequency of Tex14 mutations. Genome wide association analyses of heterogenic samples require high number of cases and controls. Thus, the low amount of samples hinders the analyses in these studies and larger genome-wide association studies are required in order to identify novel SNPs associated with azoospermia. However, Tex14 is a good candidate gene for causal mutations in azoospermic men.

In this study we report a specific infertility defect in Yorkshire pigs. The infertility is caused by spermatogenic arrest in early meiotic sperm cells resulting in absence of spermatocytes and spermatids. The genome wide scan located the disease associated area within 2 Mbp region in porcine chromosome 12. Sequencing of a candidate gene, Tex14, revealed several polymorphisms within the coding sequence. The most likely causal mutation was identified within exon 27 resulting in premature stop codon in the mRNA sequence. In addition, the expression of TEX14 mRNA and protein products was markedly down regulated in the testis of a SA affected boar compared to control animals. The genomic mutation appeared to be a 51 bp insertion, which originates from Tex14 intron 27. Thus, the SA insertion is probably an original duplication event separated by recombination.