Download Cytoskeleton: What Does GTP Do for Septins? Dispatch

Current Biology, Vol. 12, R788–R790, November 19, 2002, ©2002 Elsevier Science Ltd. All rights reserved. PII S0960-9822(02)01295-2
Cytoskeleton: What Does GTP Do for
Septins?
Timothy J. Mitchison and Christine M. Field
The recent observation of GTP-promoted polymerization of a single septin polypeptide suggests
that this protein has tubulin-like biochemical properties. This model cannot, however, explain the GTPbiochemistry of heteromeric septin complexes from
cytosol.
The septins are a conserved family of GTP-binding
proteins. They were discovered in the budding yeast
Saccharomyces cerevisiae as a set of genes — CDC3,
CDC10, CDC11 and CDC12 — required for normal bud
morphology. Subsequent work suggested that septins
are the building blocks of the neck filaments which
organize the thin neck of cytoplasm separating the
mother and bud. [1]. Septins are also present in metazoan cells, where they are required for cytokinesis in
some systems, and implicated in a variety of other
processes involving organization of the cell cortex and
exocytosis [2–4]. Septins immunopurified from cytosol
exist as heteromeric complexes containing three or
four different septin polypeptides in a defined
stochiometry. These complexes can polymerize into
long filaments in vitro.
All septins contain sequence elements distantly
related to conserved motifs of the Ras/EF-Tu family of
small GTPases [2,5], and analysis of isolated septin
complexes from cells and Drosophila embryos showed
that each septin subunit indeed binds a molecule of
guanine nucleotide [6,7]. For all GTPases, understanding their biochemistry has hinged on understanding the
role of GTP binding and hydrolysis, and it seems likely
that the same will hold for septins. It is thus exciting that,
as reported recently in Current Biology, Mendoza,
Hyman and Glotzer [8] have observed rapid GTP hydrolysis by, and GTP-driven polymerization of, a single
septin polypeptide expressed in bacteria (Figure 1).
In previous work with septin complexes isolated
from cytosol, it was difficult to characterize the biochemical role of GTP for two reasons. First, the complexes contain three or four different septins, each with
its own GTP-binding site (Figure 1). Second, GTP
exchange was very slow or absent in isolated complexes, and their polymerization was not affected by
GTP [6,7,9]. Mendoza et al. [8] avoided the complexity
of multiple GTP-binding sites by expressing a single
septin polypeptide in bacteria. Earlier studies of bacterially produced septins [10,11] also reported rapid GTP
binding and hydrolysis, but Mendoza et al.’s [8] observation of GTP-promoted polymerization is the first time
that addition of nucleotide has been observed to influence septin behavior in vitro. As such, it may presage
Department of Cell Biology, Harvard Medical School, 240
Longwood Avenue, Boston, Massachusetts 02115-5731, USA.
Dispatch
a breakthrough in the septin field. Figure 1 summarizes
the reported biochemical differences between the
expressed single septin polypeptide and complexes
purified from cytosol.
What are the likely implications of these new observations for septin biochemistry? Mendoza et al.’s [8]
discussion of the physiological significance of their
results is very much influenced by the biochemistry of
β tubulin and the related bacterial protein FtsZ
(Figure 2A). They propose that GTP binding promotes
septin polymerization in cells. Their data are also consistent with the possibility that, as with β tubulin, GTP
hydrolysis promotes septin depolymerization.
To reconcile with the fact that their experiments
were performed with a single polypeptide, rather than
a heteromeric complex, Mendoza et al. [8] suggest
that single septin polypeptides may polymerize in the
cell under some circumstances. They cite genetic and
cytological data as suggesting that not all the septins
in a heteromeric complex may be required for function
[10,12]. Alternatively, they suggest that the whole
heteromeric complex behaves in the same way they
observe for a single septin. In our opinion, the published genetic and cytological data do not support the
view that single septins can function on their own,
though this question requires more research. And
while similarity to β tubulin is a reasonable hypothesis
for the role of GTP in septin biochemistry, other
models are also worth considering.
For all GTPases, GTP binding and hydrolysis allow
the protein to act as a molecular switch, but the
function of this switch differs markedly between
GTPase sub-families. In Figure 2 we summarize a
subset of the known roles of GTP in GTPase subfamilies, focusing on those of greatest potential relevance
to septins. The polymerization model (Figure 2A) is
based on the established β tubulin/FtsZ mechanism.
This is the model favored by Mendoza et al. [8]; it
explains their data with isolated septin polypeptide
well, but does not account for behavior of heteromeric
complexes (Figure 1).
The folding model, illustrated in Figure 2B, is
exemplified by α tubulin, and is worth considering for
septins. It is possible that the GTPase domain of the
septin expressed in Escherichia coli was unfolded,
that GTP binding triggered its folding, and that this in
turn promoted polymerization and hydrolysis. A
pessimistic extension is that the GTP-driven polymerization seen by Mendoza et al. [8] is not in fact a clue
to septin biochemistry, but an artifact of bacterial
expression. We do not favor this idea, in part because
addition of GDP did not promote polymerization. But
it would be worth checking the folding state of the E.
coli-expressed single septin.
The complex assembly model, illustrated in Figure
2C, describes the biochemistry of several important
GTPases, including Ras and EF-Tu, and may well be
relevant to septins. With only one septin present,
Current Biology
R789
Exchange and hydrolysis
Polymerization
Pi
Single septin
expressed
in E. coli
GTP
–
–
GTP
FAST
GDP + Pi
?
Heteromeric
complex from
cytosol
GTP
GTP
GDP GDP
GTP
GDP GDP
High salt
Lo w salt
Dialysis
SLOW
GDP + Pi
Current Biology
Figure 1. Comparison of an isolated septin complex with a recombinant single septin polypeptide.
A Xenopus septin expressed in E. coli [8] was purified without bound nucleotide. This septin exchanged and hydrolysed GTP rapidly.
Addition of GTP promoted polymerization into paired filaments containing bound GDP. Septins were purified as heteromeric complexes [3,7,9] from cytosol of Drosophila embryos, yeast or baculovirus-infected insect cells. These heteromeric complexes contained
tightly bound nucleotide (one molecule per septin polypeptide, GDP:GTP ratio about 2:1). Pure heteromeric complexes exchanged
GTP very slowly. The yeast complex polymerized into paired filaments, and the baculovirus-expressed complex into bundles and rings
when the salt was lowered by dialysis. Addition of GTP did not affect polymerization of heteromeric complexes.
Mendoza et al. [8] observed that GTP binding promoted
formation of a homodimer interface, leading to polymerization. But when a different septin is also present,
perhaps the same biochemical switch promotes formation of a heterodimer interface, and thus assembly of a
discrete complex, without polymerization. This model is
consistent with the known biochemistry of heteromeric
complexes (Figure 1) and not inconsistent with the data
of Mendoza et al. [8]. An interesting extension of this
model is that GTP hydrolysis might promote disassembly of the heteromeric complex, so that regulation of
complex assembly might be important for septin
biology. In the complex assembly model, it is switching
between isolated septin molecules and heteromeric
complexes that is key to the biology, rather than
switching between subunit and polymer as in the polymerization model.
What needs to be done next? At a biochemical
level, the most important direction seems to be to try
to bridge between the new observations with a single
septin and previous observations with isolated
heteromeric complexes. It will be interesting to mix
two (or three) different expressed and correctly folded
septins, for example, and ask whether GTP addition
promotes polymerization of a single septin or the
A
GTP
Polymerization
Figure 2. Models for the role of GTP in the
biochemistry of septins. (See text.)
GDP
GTP
GTP
GTP
GTP
GTP
GDP
B
assembly of discrete heteromeric complexes. The
only published attempt to date at reconstituting a
heteromeric complex from single expressed subunits
was not successful [13], but this does not preclude
the possibility.
Another important direction will be to try to find
conditions where isolated heteromeric complexes
exchange nucleotide more rapidly, and thus determine
whether guanine nucleotide regulates polymerization
or complex assembly — or some other, unexpected
function. This may require identification of a septin
GTP-exchange promoting protein. At a cellular level, a
key question in our minds is to what extent septin
function depends on polymerization, or even on
heteromeric complex formation. Analysis of yeast
strains deleted for one septin suggested that not all
septin function requires normal polymerization into
filaments [9]; whether all functions of septins depend
on heteromeric complex formation has not been
rigorously tested. A related question is whether
isolated septin molecules exist in the cell, and if so,
whether they have a function.
Beyond these issues of assembly state, the work of
Mendoza et al. [8] may help generate tools to probe
septin biology. For small GTPases, mutants deficient in
β-tubulin,
FtsZ
GTP
Folding
GTP
C
Complex
assembly
GTP
GDP
α-tubulin
GDP
GTP
EF-Tu,
Ras
Current Biology
Dispatch
R790
GTP hydrolysis have been key tools for elucidating cellular function, and the same will likely be true for septins.
Septin sequences diverge from Ras at key residues
implicated in catalyzing GTP hydrolysis, and so far,
GTPase-deficient mutants have not been reported.
Expression of septins with mutations that prevent GTP
binding cause interesting biological effects [10,14], but
as the GTPase domain in these mutants may not be correctly folded, these data are hard to interpret at a molecular level. With a system where GTP binding and
GTPase activity can be easily measured, it may be possible to systematically test septin mutants for decreased
(or increased) GTPase rate.
Septins have been slow to yield their biochemical
secrets, but the first observation of a functional effect
of GTP should have a tonic effect on the field. While
the implications of the new observation are yet clear,
we are confident they will stimulate new interest and
experimental approaches, and are thus a major step
forward. (And as a final note, aficionados should bear
in mind that a new nomenclature has recently been
adopted for mammalian septins [15].)
References
1. Gladfelter, A.S., Pringle, J.R., and Lew, D.J. (2001). The septin
cortex at the yeast mother-bud neck. Curr. Opin. Microbiol. 4,
681–689.
2. Field, C.M., and Kellogg, D. (1999). Septins: cytoskeletal polymers
or signalling GTPases. Trends Cell Biol. 9, 387–394.
3. Trimble, W.S. (1999). Septins: A Highly conserved family of membrane-associated GTPases with functions in cell division and
beyond. J. Membr. Biol. 169, 75–81.
4. Kartmann, B., and Roth, D. (2001). Novel roles for mammalizn
septins: from vesicle trafficking to oncogenesis. J. Cell Sci. 114,
839–844.
5. Bourne, H.R., Sanders, D.A., and McCormick, F. (1991). The GTPase
superfamily: conserved structure and molecular mechanism. Nature
349, 117–127.
6. Field, C.M., al-Awar, O., Rosenblatt, J., Wong, M.L., Alberts, B., and
Mitchison, T.J. (1996). A purified septin complex forms filaments
and exhibits GTPase activity. J. Cell Biol. 133, 605–616.
7. Kinoshita, M., Field, C.M., Coughlin, M.L., Straight, M.F., and Mitchison, T.J. (2002). Self- and actin-templated assembly of mammalian
septins. Dev. Cell, in press.
8. Mendoza, M., Hyman, A.A., and Glotzer, M. (2002). GTP binding
induces filament assembly of a reconstituted septin. Curr. Biol.,
29th October issue.
9. Frazier, J.A., Wong, M.L., Longtine, M.S., Pringle, J.R., Mann, M.,
Mitchison, T.J., and Field, C.M. (1998). Polymerization of purified
yeast septins: evidence that organized filament arrays may not be
required for septin function. J. Cell Biol. 143, 737–749.
10. Kinoshita, M., Kumar, S., Mizoguchi, A., Ide, C., Kinoshita, A.,
Haraguchi, T., Hiraoka, Y., and Noda, M. (1997). Nedd5, a mammalian septin, is a novel cytoskeletal component interacting with
actin-based structures. Genes Dev. 11, 1535–1547.
11. Zhang, J., Kong, C., Xie, H., McPherson, P., Grinstein, S., and
Trimble, W.S. (1999). Phosphatidylinositol polyphosphate binding to
the mammalian septin H5 is modulated by GTP. Curr. Biol. 9,
1458–1467.
12. Adam, J.C., Pringle, J.R., and Peifer, M. (2000). Evidence for functional differentiation among Drosophila septins in cytokinesis and
cellularization. Mol. Biol. Cell 11, 3123–3125.
13. Joberty, J., Perlungher, R.R., Sheffield, P.J., Kinoshita, M., Noda,
M., Haystead, T., and Macara, I.G. (2001). Borg proteins control
septin organization and are negatively regulated by Cdc42. Nat. Cell
Biol. 3, 861–866.
14. Beites, C.L., Xie, H., Bowser, R., and Trimble, W. (1999). The septin
CDCrel-1 binds syntaxin and inhibits exocytosis. Nat. Neurosci. 2,
434–439.
15. Macara, I.G., Baldarelli, R., Field, C.M., Glotzer, M., Hayashi, Y.,
Hsu, S., Kennedy, M. B., Kinoshita, M., Longtine, M., Low, C., et al.
(2002). Mammalian septins nomenclature. Mol. Biol. Cell, in press.