Discussion
This study describes the identification of TMEM203 as a novel regulator of ER calcium levels. TMEM203 was identified here by its ability upon overexpression to mediate calcium dependent nuclear relocalization of the CRCT1 and NFAT transcription factors. TMEM203 is an evolutionary conserved resident ER protein that bears little amino acid sequence similarity to other proteins. TMEM203 was required to maintain normal ER calcium levels in a variety of cell types. Mice lacking Tmem203 expression were viable though male knockout mice were infertile and exhibited a severe block in spermiogenesis and spermiation which were also accompanied by altered accumulation of cellular calcium. Thus Tmem203 represents an evolutionarily conserved regulator of intracellular calcium homeostasis and provides a causal link between intracellular calcium regulation and spermiogenesis.
The ability of TMEM203 to induce nuclear translocation of NFAT and CRTC1 required activation of calcineurin likely following severe loss of ER calcium stores and activation of store operated calcium entry (SOCE). The mechanism by which TMEM203 overexpression releases ER calcium stores is not clear. Exogenously expressed TMEM203 was localized to the ER (Fig 2A) and was found in a complex with a number of calcium regulatory proteins, including STIM1, SERCA2 and IP3R (Fig 2B). The activity of TMEM203 could be explained by regulation of one or more complexed proteins, for example by either inhibition of calcium refilling via SERCA2 or activation of the IP3R. Alternatively, overexpressed TMEM203 could form a pore, increasing the leak of calcium stores as has been suggested for other overexpressed proteins such as BCL-2, the Bax inhibitory protein-1 or presinillins [43,44].
TMEM203 depletion, either via siRNA or through gene disruption also lowered ER calcium stores (Fig 3A–3D). Thus, at least in some cell types, TMEM203 appeared to play an essential role in maintaining ER calcium. It is not clear why overexpression and loss of TMEM203 expression both resulted in reduction of ER calcium stores. One explanation is that overexpression of TMEM203 may act in a dominant negative fashion, perhaps titrating out either endogenous TMEM203 or TMEM203 interacting proteins. However, it is worth noting that the effects of TMEM203 inhibition and overexpression are distinct in some ways. For instance, while overexpression of TMEM203 appears to elevate the basal cytosolic calcium levels and activate NFAT and CRTC1 (Fig 1A–1D), loss of Tmem203 has an inhibitory effect on NFAT dependent gene expression in mouse fibroblasts (S5 Fig). These observations suggest that overexpression and reduction of TMEM203 have distinct physiologic affects and that TMEM203 likely regulates ER calcium via multiple mechanisms.
In this regard it is worth noting that other proteins, notably of the BCL-2 family, have been reported to have bidirectional/opposing effects on ER calcium stores. For example, overexpression of BCL-2 lowers ER calcium stores by several reported mechanisms including acting as a leak channel, inhibiting refilling by SERCA and by enhancing activation and calcium release by the IP3R [45]. Conversely BCL-2 has also been suggested to act as an inhibitor of the IP3R, preventing calcium release during stress. Similarly, BAX overexpression has been reported to reduce ER calcium stores [46], whereas inactivation of BAX and BAK also reduces ER calcium levels through enhanced calcium leak and IP3R activation [47,48]. The precise mechanism by which TMEM203 affects ER calcium stores will await further studies including the biochemical purification and characterization of TMEM203 and associated protein complexes.
Tmem203 deficient male mice were infertile with a failure to complete spermatogenesis and spermiogenesis. At present, morphologic analysis does not pinpoint a single aspect of spermatogenesis affected by TMEM203 deficiency. Haploid spermatids appear to be reduced in numbers and spermiogenesis fails after nuclear condensation, resulting in a large reduction in elongated spermatids, and accumulation of small numbers of degenerative malformed sperm heads which fail to undergo spermiation (Fig 5A–5E).While the exact cause of spermiogenic failure is not clear, calcium homeostasis in testicular cells was clearly defective; suggesting that defective spermiogenesis may be due to altered handling of calcium stores (Fig 7A–7D). Expression profiling experiments demonstrated significantly altered expression of calcium pumps and channels in Tmem203 deficient mice (Fig 6C). The increased expression of the main calcium extrusion pump, Pmca1 and decreased expression of multiple import channels seems likely to contribute to modestly reduced basal calcium levels and highly reduced retention of calcium after SOCE seen in vitro with primary spermatocytes. The altered expression of calcium pumps and channels might be due to either an attempt by cells to compensate for loss of Tmem203 or indirectly due to developmental defects. It is worth noting that loss of Bcl2 also appears to result in a dramatically increased rate of calcium extrusion in pancreatic acinar cells due to increased function of the Pmca calcium pump [49].
A clear role for calcium stores in spermiogenesis has not been directly described although a number of genes regulated by calcium have been shown to be essential for sperm development. C elegans lacking expression of mammalian homologs of calreticulin (Crt-1)—the calcium binding chaperone protein or calcineurin (Cna or Cnb)—the calcium dependent phosphatase are sterile due to defects in sperm development in addition to oocyte development and fertilization [50,51]
A similar spermiogenesis defect as in Tmem203 null mice was described with Calcium and integrin binding 1 (CIB1)/calmyrin null mice [52]. CIB1 is a homolog of Calmodulin and Calcineurin B and regulates calcium dependent signaling events [53]. It will be of interest to determine if CIB1 and Tmem203 act in a common pathway involved in calcium/calcineurin signaling and spermatogenesis in mammals. Finally, Bax is also required for spermiogenesis in mice [54] and as mentioned above, also modulates levels of ER calcium similar to Tmem203. The similarity in phenotypes seen in Tmem203 and Bcl2 family protein deficient mice [37] suggests that these proteins may converge on common mechanisms that govern ER calcium stores and cellular calcium handling.
The defective calcium handling in testicular cell preparations, which were predominantly haploid spermatocytes (Fig 7A–7D), suggests that Tmem203 deficiency directly affects development or survival of spermatocytes. However, it is also possible that defects Sertoli and Lydig cells could also play some or a major role in the infertility in Tmem203 deficient mice. Spermiation, the process of releasing mature sperm from the Sertoli cells into the seminiferous tubule was completely absent in Tmem203 deficient mice, while it is possible that Tmem203 is directly involved in spermiation, it seems likely this defect is due to the severe loss of maturing spermatids, miss-shappen post meiotic spermatids and the likely clearance of remaining malformed spermatids by Sertoli cells before spermiation. The simplest explanation for the phenotypes observed is that the defective calcium handling within spermatocytes results in developmental arrest or cell death of spermatocytes shortly after the stage of nuclear condensation.
Tmem203 is a highly conserved gene and is absolutely requirement for fertility. Tmem203’s role in maintaining steady state ER calcium levels suggests it may modulate a variety of biological processes. Indeed, preliminary analysis suggests that TMEM203 deficiency also results in other phenotypes. TMEM203 deficient mice are about 10% smaller than WT siblings, have mild immunologic defects at least in vitro, and exhibit evidence for impaired neurologic function, though these phenotypes have not been sufficiently characterized for publication yet. The absolute requirement for TME203 in fertility would seem to be sufficient to explain its high level of evolutionary conservation. Tmem203 null mice may serve as an excellent tool to understand the regulation of intracellular calcium levels and its interdependence on carefully orchestrated modulation of ER calcium stores. Tmem203’s role in spermatogenesis adds to the complexity of this fascinating process and strongly suggests a role of intracellular calcium homeostasis during mammalian spermatogenesis.