Sorting out the role of the RSK tandem kinases in skeletal muscle myogenesis (and perhaps spermiogenesis (mouse t-complex))
This
project focuses on understanding the individual roles of the three 90 kD
S/T kinases Rsk
isoforms expressed in skeletal muscle cells. The isoforms are
encoded on distinct chromosomal loci in mice and man, one X-linked
(RSK2; patients with Coffin-Lowry Syndrome (CLS) are men
whose
1 X chromosome encodes null alleles of RSK2) and the two others (RSK1 and
RSK3) autosomal. Two additional
tandem kinases (MSK1 and MSK2) have just recently
been published
by Dario
Alessi's group in Dundee
(Scotland), one identical to a cDNA dubbed RSK-B by
Lesslauer's group in Basel, Switzerland (GenBank HSA010119), published in
the November 6 issue of JBC..
According to Alessi, MSK1 and MSK2 are also very abundant
in muscle, like the RSK1-2-3 kinases. The MSK or RSK-B kinases seem to lie
downstream
of p38 MAPK, not the ERK kinases like the true RSK1-2-3 isoforms. The
three MAP kinase pathways (see inset below) are canalized or shielded
from crosstalk from parallel MAP kinase pathways. The mechanism for
this canalization seems to involve scaffolding proteins such as JIP1
and MP1
as well as stable non-covalent associations of signaling molecules
into "signal transduction particles", such as the association between
ERK1 and RSKs. RSK3
is especially
interesting, because it features an additional nuclear targetting signal
near its N-terminus, and can phosphorylate a variety of transcription
factors (CREB, TCF/Lef, Myc, the estrogen receptor, and the Fos subunit
of AP-1), in addition to the NF-kappaB regulator I-kappa-B-alpha
(IKBA). The RSK
kinases can also phosphorylate histones H1 and H2B,
perhaps signaling a role in chromatin or chromatin remodelling; in
collaboration with Paolo Sassone-Corsi (Strasbourg), Dave Allis
(University of Virginia) has just demonstrated that the Coffin-Lowry
tandem kinase (RSK2) is an EGF-stimulated histone H3 kinase (Science). A
new
paper in EMBO
Journal (Deak et al., EMBO J 17, 4426 (1998) suggests that
the phosphorylation of CREB by RSK may be a red herring, and the true
CREB kinase may be another RSK-related kinase called MSK in mouse. We
are exploring the function of the RSK isoforms and their association with
MAP kinases in the mouse myoblast line C2C12, shown at left after 72 hr
in differentiation medium. Notice that the myotubes harbor >25 nuclei,
often aligned in chains. In this photograph, the cytoplasm is stained
by an antibody to skeletal muscle myosin heavy chain (Texas Red) and
the nuclei are stained with DAPI (epifluorescence facility of John
Tomkiel).
The RSK3
gene is located a 6q27
in humans; the Sanger
Centre in England has cloned two 150
kb segments and
has completely sequenced both
of them. The two files overlap; I have spliced them together and
created a single 243 kb sequence file, which
show that the human RSK3 gene (from cap site at +1 to the polyadenylatlon
signal) is 217,693 bp in length, featuring 21 exons, and one enormous
intron (between exons 1 and 2) of 88.1 kb. . Two expressed sequence tag
(EST) cloning projects
have picked up human RSK3 cDNAs, including the sequence identified as
N79949 (Hillier et al.) and the sequence identified as HSCOWE011
(Auffray
et al.). Inadvertently, Kim et al. (BBRC 254, 720-727 (1999)) cloned
the rat homolog of RSK3 (accession D83013), which they think is RSK1; they
are wrong,
because it is certainly RSK3. Bernhard Herrmann's group at the
Max-Planck Institute for Immunobiology at Freiburg has just published an
article in Mammalian
Genome about the Rsk3 gene in mouse, which maps
smack-dab in the middle of the region of chromosome 17 where John
Schimenti mapped Tcr (the genetic entity posited by Mary Lyon to be
the element mediating the response of t-complex alleles to a series
of nearby genes regulating spermiogenesis!). There are two wonderful
WWW resources on structure and function of kinases available at http://www.sdsc.edu/kinases
and a new one called Phosphobase maintainted by Andres Kreegipuu &
Nikolaj Blom in Denmark, found at this site.
This project is a collaboration
with Dr. Minoru S. H.
Ko, M.D., Ph.D., George Grunberger, M.D., and John Schimenti, Ph.D. at the
Jackson Laboratory in Bar Harbor, ME.
This summer (1998), two
summer students are working in my lab on this project. One student (Mr
Christopher P Gayer, shown at his poster session; checked shirt) will be a
senior at Alma College this fall. He is constructing
plasmids designed to express antisense segments of mouse RSK-1, RSK-2,
and RSK-3 cDNAs, and expression constructs for GST-IKB fusion proteins.
He is also putting together a construct designed to produce a dominant
negative (D205N) RSK-1 cDNA under the control of a T7 promoter, for
microinjection experiments. Another summer student, Mr Sahil Sood,
will be a sophomore this fall at Johns
Hopkins University. Sahil put together a targeting
construct (pPNT-RSK2523) designed for knockout
of the mouse RSK-3 gene through
homologous recombination in embryonic stem (ES) cells. In
collaboration with Dr Ye-Shih Ho at the WSU Institute for Chemical
Toxicology, we now have 13 ES cell candidate clones which apparently have
undergone homologous recombination with our targeting vector to create a
null RSK-3 allele, defective in exon 6.
Phylogenetic analysis of the tandem kinase family which
includes RSK kinases from worms, flies, frogs, chicks, rat, mouse and man
and the recently-discovered MSK kinase family. This analysis suggests
that the frog, rat, and chick RSK kinases are probably RSK-1 orthlogs.
There is a partial sequence (not shown) of a RSK kinase from the
xebrafish Danio rerio which almost certainly is a RSK-2 ortholog. The
fly and worm RSK
kinases, only 1 per species, are not easily sorted
into RSK-1, RSK-2 or RSK-3, although this tree suggests that the single
RSK kinase in worms may actually be a RSK-1.
Our laboratory is using the C2C12 myogenic model to not only sort out the individual function of the 3 RSK isoforms, but to clarify the biochemistry and cell biology of RSK-ERK interactions, which may mediate evolutionarily-conserved biological phenomena at the level of the cell as diverse as oocyte maturation in frogs and myogenesis in mammalian skeletal muscle. One route we are taking to explore the structure and function of the ERK-RSK "particle" involves the use of aptamers in the yeast 2-hybrid system, described in a recent PNAS article by Misha Kolonin and Russ Finley. This project is being carried out by a first year graduate student in the lab, Mr Dawei Wang, who has now cloned an intact (2.2 kb) mouse RSK3 open reading frame from mouse skeletal muscle, to be shuttled into the LexA fusion vector pEG202.
Simpler eukaryotes may help sort out the logic of MAP kinase signaling pathways through genetic analysis of epistasis
The MAP kinase is probably about 1 billion years old, appearing in
unicellular eukaryotes such as Saccharomyces cerevisiae (yeast), which
surprisingly encodes five MAP
kinases. The genome of the free-living roundworm Caenorhabditis
elegans is now 98%
sequenced, and seems to encode three MAP kinases.
One of these (Sur-1, Suppressor of Ras-1) is well-studied, and is
involved in detemination of vulval cell fate (Wu
and Han, 1994; Lackner
et al., 1994) and in "exit from
pachytene" (Church
et al., 1995). There is no RSK or MSK gene in yeast, but there is
one tandem kinase in C elegans, and it seems to be a bona fide RSK
kinase. Does the RSK gene in C. elegans function in the mpk-1/sur-1
pathway? For example, what is the effect of a RSK null mutation or RSK
RNAi
on vulval development? In mammals, RSK can phosphorylate glycogen
synthase kinase (gsk)-3beta. Can RSK phosphorylate gsk-3 (sgg-1) in C.
elegans? Does RSK thus function in spindle re-positioning and the
centrosome cycle in C. elegans?
The Goustin laboratory is located on the Fifth Floor of the Biological Sciences Building, WSU Main Campus, across from the newly-renovated historic Old Main (below).
For more information about positions available, contact:
Anton
Scott Goustin, Ph.D. (313-993-7688)
asg@cmb.biosci.wayne.edu