Which cell type starts off as a diploid spermatogonium, spends some time as a round, haploid spermatid, but ends up being elongate and haploid with a powerful flagellum? A spermatozoan, of course. What regulates the differentiation of spermatids into spermatazoa is less obvious, but Gliki et al. have uncovered a role for the junctional adhesion molecule (JAM)-C in the polarization of round spermatids, by virtue of its ability to recruit a cell-polarity complex.

To ascertain the function of JAM-C directly, the authors disrupted the Jam-C gene by homologous recombination in mice. Of the surviving ∼60% of the offspring, the males had testes that were half the size of those of their littermates, and they couldn't make mature sperm cells; there were no differentiated, elongated spermatids. Analysis of the testes of wild-type mice showed JAM-C expression throughout most of the round spermatogenic cells, although some JAM-C was concentrated at the front (the presumptive head). As spermiogenesis proceeded, JAM-C was restricted to specialized adhesion structures — junctional plaques — that anchor spermatids to the signal- and nutrient-providing Sertoli cells. Closer inspection showed that another isoform — JAM-B — was present on Sertoli cells, so that, in principle, trans-interactions could occur between these two immunoglobulin-domain-containing proteins.

Having excluded a role for JAM-C in the maintenance of the blood–testes barrier (defects in which cause infertility), the authors noticed that morphological signs of polarization, such as the formation of an acrosome (the specialized penetrating vesicular organelle), were absent in the round Jam-C−/− spermatids. However, although cytoskeletal defects occurred in Jam-C−/− spermatids, these, along with defects in acrosome formation, weren't thought to be responsible for impaired polarization in spermatids.

Gliki et al. therefore looked at known mediators of cell polarization — specifically, Par6, Cdc42 and protein kinase Cλ (PKCλ; also referred to as atypical PKC (aPKC)). In wild-type spermatids, these proteins colocalized with JAM-C in the head regions, whereas their distribution was not polarized in Jam-C−/− spermatids. Another protein, PATJ, through interactions with PALS1 and crumbs3 (CRB3), can also regulate polarity, and the Par6–Cdc42–aPKC and CRB–PALS1–PATJ complexes can directly interact. PATJ also showed some colocalization with JAM-C at the junctional plaques.

How, then, might JAM-C affect these polarity proteins? To study this, the authors mimicked the binding of spermatids to Sertoli cells by providing cultured primary spermatids with soluble versions of JAM-B or JAM-C fused to immunoglobulin (Ig) domains. Before the addition of these fusion proteins, some PATJ was polarized and Cdc42 and PKCλ were broadly distributed in both wild-type and Jam-C−/− cells. But only in wild-type spermatids did the addition of the JAM-B and JAM-C fusion proteins cause Cdc42, PKCλ and PATJ to redistribute to the fusion-protein-binding sites. The authors also coimmunoprecipitated complexes containing JAM-B and JAM-C from testes, as well as complexes of JAM-B, Par6, Cdc42, PKCλ and PATJ with JAM-C. The polarized distribution of Par6 has been shown to require an interaction with Cdc42, and the authors showed that adding JAM-C–Ig fusion proteins to JAM-C-overexpressing cells increased activated Cdc42 — a potential way to recruit Par6 and the rest of the polarity proteins.

So JAM-C has a key role in mouse spermiogenesis during the polarization of spermatids by recruiting a polarity complex to junctional plaques. Because the proteins that are involved also show a similar distribution and behaviour in human spermatids, further investigation of JAM-C could uncover important information for the treatment of human infertility.