Repositioning septins within the core particle

Abstract

Abbreviations (Define non-standard or uncommon abbreviations). AbstractSeptins are GTP binding proteins considered to be a novel component of the cytoskeleton. They polymerize into filaments based on hetero-oligomeric core particles which, in humans, are either hexamers or octamers composed of two copies each of either three or four different septins from the 13 available. Not all combinations are possible as it is believed that these must obey substitution rules which determine that different septins must be derived from four distinct and well-established sub-groups. Here, we have purified and characterized one such combinations, SEPT5-SEPT6-SEPT7, and used TEM to derive the first structural information concerning its assembly. The full complex was purified using an affinity tag attached to only one of its components (SEPT7) and was able to bind to and perturb lipid bilayers. Although the complex assembled into elongated hexameric particles, the position of SEPT5 was incompatible with that predicted by the reported structure of SEPT2-SEPT6-SEPT7 based on the substitution rules. MBP-fusion constructs for SEPT5 and SEPT2 and immuno-staining clearly show that these septins occupy the terminal positions of the SEPT5-SEPT6-SEPT7 and SEPT2-SEPT6-SEPT7 hexamers, respectively. In so doing they expose a so-called NC interface which we show to be more susceptible to perturbation at high salt concentrations. Our results show that the true structure of the hexamer is inverted with respect to that described previously and, as such, is more compatible with that reported for yeast. Taken together, our results suggest that the mechanisms involved in spontaneous self-assembly of septin core particles and their filaments deserve further reflection.

Turning it inside out: The organization of human septin heterooligomers

“…Hints from the literature notwithstanding, it is fair to say that until very recently the septin field accepted the arrangement of the septins within human heterohexamers as proposed by Sirajuddin et al () and within heterooctamers as proposed by Kim et al (). Everything changed with the deposition on the same day in March 2019 of two preprints in the bioRxiv (Mendonça et al, ; Soroor et al, ). One study, a revised version of which is published in this issue of Cytoskeleton , uses antibody decoration and MBP‐tagged subunits of two different representatives of the SEPT2 group, SEPT2 itself and SEPT5, to provide compelling support for the conclusion that the order of subunits in these human septin heterohexamers is SEPT2 (or SEPT5)–SEPT6–SEPT7–SEPT7–SEPT6–SEPT2 (or SEPT5) (Figure d), an order that is the exact inverse of that deduced by Sirajuddin et al ().…”

“…Mendonça et al () set out to purify a complex containing human SEPT5, SEPT6, and SEPT7 following heterologous co‐expression in E . coli .…”

“…Intrigued by this possibility, we noted in the Methods section of the Sirajuddin et al () study that they found it necessary to stabilize their hexamers via prefixation with glutaraldehyde before they examined their complexes by EM. Such a treatment was not needed in any other single‐particle EM study of other septin complexes, including those of Mendonça et al (). What could have destabilized the hexamers of Sirajuddin et al?…”

“…Hints from the literature notwithstanding, it is fair to say that until very recently the septin field accepted the arrangement of the septins within human heterohexamers as proposed by Sirajuddin et al (2007) and within heterooctamers as proposed by Kim et al (2011). Everything changed with the deposition on the same day in March 2019 of two preprints in the bioRxiv (Mendonça et al, 2019;Soroor et al, 2019). One study, a revised version of which is published in this issue…”

“…The other study is complementary because it uses primarily in vivo evidence to arrive independently at the same conclusion and, further, provides evidence that a SEPT9-SEPT9 NC homodimer occupies the central location in human septin heterooctamers. Mendonça et al (2019) set out to purify a complex containing human SEPT5, SEPT6, and SEPT7 following heterologous coexpression in E. coli. SEPT5 is closely related to SEPT2 (Pan et al, 2007); hence, the authors expected to purify hexamers containing a 1:1:1 ratio of these three proteins and, indeed, they did.…”