|
Figure 1. Fractionation of factors interfering with actin polymerization
present in 100,000 g supernatant from whole Xenopus ovaries
by gel filtration on a Sephadex G 150 column; 15 ml postribosomal
supernatant, prepared as described in Materials and Methods, were
loaded on the column and the protein eluted was measured by absorption
at 280 nm (solid line). Individual fractions were assayed
for their influence on the polymerization of rabbit skeletal muscle
actin as measured by low shear viscosimetry. The polymerization
of the G-actin solution was started by addition of KC1 and ATP to
final concentrations of 100 mM and 1 raM, respectively (continuous
line with dots) or 2 mM MgC12 instead of KCI (finely dotted
line). Active fractions, eluting in peak I (at a position corresponding
to Mr 180,000) were pooled and used for further purification. A
second activity peak (peak II, eluting at the position of ovalbumin),
vanished completely when actin polymerization was initiated by
2 mM MgCI2 instead of 100 mM KCI.
|
|
Figure 2. Monitoring of the actin-modulating protein during
purification. (a) SDS-PAGE (18%; for buffer system see references
2, 30) analysis of fractions obtained during the purification of the
major actin-modulating activity from Xenopus ovaries (staining
with Coomassie Blue). Lane 1, reference polypeptides (from top to
bottom): 13-galactosidase (Mr 116,000), phosphorylase a (Mr
93,000), BSA (Mr 68,000), actin (Mr 42,000); lane 2, total 100,000
g supernatant; lane 3, pooled peak I fractions obtained by gel filtration
of the 100,000 g supernatant on Sephadex G 150 (cf. Fig~ 1);
lane 4, 3 M guanidinium-HCl (Gu.HCI) eluate obtained after chromatography
of the pooled peak I fractions on DNAse I agarose. The
polypeptide of Mr '~ 105,000 (in this gel system) coelutes with actin
(open triangles); lane 5, peak activity fraction obtained after
chromatography of the dialyzed Gu-HC1 eluate (lane 4) on a
MonoQ anion exchange column; lane 6, the purified protein, obtained
by re.chromatography of the MonoQ fractions (shown in lane
5), on a MonoQ column in the presence of 6 M urea. The purified
protein showed a Mr of ,'~93,000 in the SDS-PAGE system used
here and its ratio to actin was somewhat higher after MonoQ separation
(lane 5) than before (lane 4). We do not know whether this
is a preparative artifact or reflects the enrichment of equimolar
complexes formed in the oocyte. (b) Two-dimensional gel electrophoreses
of the purified major actin-modulating protein of Xenopus
laevis ovary (first dimension: isoelectric focussing; second
dimension: SDS-PAGE). (B) BSA and rabbit muscle ct-actin (A)
were used as markers in coelectrophoresis. The brackets denote two
polypeptides each with an isoelectric variant.
|
|
Figure 3. Characteristics of purified major actin-modulating protein from Xenopus oocytes. (Left) Effects on the polymerization of rabbit
skeletal muscle actin as measured in an Ostwald-type viscometer. The purified protein was added to an actin solution (final concentration:
0.3 mg/ml) in various molar ratios (indicated at the right hand end of each curve) and the kinetics of actin polymerization were compared
with those of the control (C, only buffer was added to the actin solution). Polymerization was induced by addition of MgC12 and ATP
to concentrations of 2 and 1 mM, respectively. Molar ratios of actin to the Mr ,x, 94,000 protein (gelsolin is indicated). Note the earlier
initiation of actin polymerization and also the marked reduction of end viscosity in the presence of the actin-modulating protein. (Right)
Influence of the major actin-modulating protein from Xenopus laevis oocytes on preformed actin filaments. Skeletal muscle actin (0.3 mg/ml)
was polymerized at 25°C by addition of 2 mM MgCI2 and 1 mM ATP. After reaching a steady state, the purified actin-binding protein
was added in various molar ratios (indicated). The time of addition is marked by an arrow. The change in viscosity was measured using
an Ostwald-type viscosimeter. Addition of the purified protein in a molar ratio of 1:8 reduced the steady state viscosity to about 50%.
C, buffer control.
|
|
Figure 4. Detection of the major actin-modulating protein in different
subcellular fractions ofXenopus laevis oocytes (a) as well as in
eggs and early embryos (b) by SDS-PAGE and immunoblotting
(only the autoradiograms are shown). (a) The amount of the protein
of two manually isolated ooplasms (lane 1) as compared with that
detected in 20 nuclei manually isolated from TCA-fixed oocytes
(lane 2). Note the much higher concentration of the protein in the
cytoplasm. In lanes 3-6 the protein present in soluble and pelletable
fractions was compared with 0.25 lag purified protein (lane 7): lane
3. two total oocytes; lane 4, low speed pellet (10 min, 15,000 g, 4°C)
of three homogenized oocytes; lane 5, supernatant fraction corresponding
to lane 4; lane 6, high speed (1 h x 100,000 g, 4°C) supernatant
corresponding to lanes 4 and 5. The proteins of the different
fractions were separated by SDS-PAGE, transferred to nitrocellulose
and probed using guinea pig antibodies raised against the purified major actin-modulating protein for Xenopus laevis ovary
(see Fig. 2 a, lane 6). (b) Comparison of the amount of reactive
protein present in Xenopus laevis ovaries (lane 1:1 h x 100,000 g
supernatant) with that present in fertilized eggs (lane 2, equivalent
to lane 1) and in embryos of the morula stage (stage 6.5; lane 3).
|
|
Figure 5. Comparison of the major actin-modulating protein of Mr
"~ 93,000 from Xenopus laevis oocytes with villin by two-dimensional
gel electrophoresis (as in Fig. 2 b) and immunoblotting. Proteins
of postribosomal extracts from total Xenopus ovaries (a) and
whole intestinal mucosa of Xenopus (b) were separated by twodimensional
gel electrophoresis and stained by Coomassie Blue (a
and b) or processed, in parallel, for immunoblotting (autoradiograms
are shown in a', a", b', b"). After transfer of the proteins
to nitrocellulose, polypeptides were reacted with rabbit antisera
against porcine villin (a', b', immunoblots) and guinea pig antibodies
against the purified Mr ~ 93,000 actin-modulating protein from
Xenopus ovaries (a", b"). In (a', a", b' and b") the position of actin
(A) as determined after staining with Ponceau S is indicated. Note
that the villin antiserum reacted with a polypeptide much more basic
(arrowheads) than the Mr '~ 93,000 polypeptide (arrows) detected
with the specific antibodies. The villin antisera showed only
a very weak reaction with a polypeptide at a similar position in
Xenopus laevis ovary (a'; arrowhead) which, however, gave a peptide
map completely different from that of the intestinal villin spot
shown in b. The antibodies against the actin-modulating protein of
Xenopus laevis oocytes reacted only with the polypeptide which was
nearly isoelectric with actin (a" and b"; arrow), as expected from
Fig. 2 b, and did not cross react with villin or the villin-related
"spot" in ovaries. We have no explanation for the reactivity of villin
antisera with the actin-modulating component of ovaries (a'), but
not with that present in intestinal tissue (b').
|
|
Figure 6. Peptide maps of polypeptides reacive with antibodies against the actin-modulating Mr ~ 93,000 protein from Xenopus laevis
oocytes, identified by immunoblotting as shown in Fig. 8. Polypeptides were excised from gels, radioiodinated, digested with TPCK-treated
trypsin, and the resulting peptides were separated by electrophoresis (E) and chromatography (C) on cellulose plates. (a) Mr * 93,000
polypeptide present in oocyte; (b) corresponding polypeptide of culture A6 cells; (c) corresponding polypeptide of heart muscle tissue.
Peptide maps of the actin-modulating protein from A6 cells and heart muscle were very similar to each other, whereas the oocyte protein
showed, besides similarities (brackets) a number of different peptides (for discussion, see text).
|
|
Figure 7. Immunofluorescence microscopy showing the distribution
of the actin-modulating protein of Mr '~ 93,000 in cryostat sections
of Xenopus laevis oocytes (a and b) and intestine (c). The sections were fixed with 10% formaldehyde and treated with antibodies
(antiserum diluted 1:10) against the actin-modulating protein (a).
General staining of the oocyte cortex (demarcated by brackets) and
ooplasm (b, phase-contrast image coresponding to a). O, ooplasm.
In the intestine (c) epithelial cells (E) and smooth muscle (M) were
stained intensely, whereas cells of the lamina propria (LP) were
weakly, if at all, reactive. Bars, 50 Ixm.
|
|
Figure 8. Nucleotide sequence
of cDNA (clone Lgel X) selected
from ~,gtll expression
library of Xenopus laevis ovary
(for details, see reference 31)
and the amino acid sequence
derived therefrom (one-letter
code). The stop signal is denoted
by an asterisk, the polyadenylation
signal is underlined.
|
|
Figure 9. Comparison of the
amino acid sequence determined
from human plasma
gelsolin cDNA (upper line;
taken from reference 32) with
portions of the amino acid sequence
of the actin-modulating
protein of Mr ,,o 93,000
from Xenopus laevis ovaries
(lower line), specifically the
amino-terminal sequence of
the native protein (sequence
underlined; X, amino acid
not clearly identified) and the
eDNA-derived amino acid sequence
shown in Fig. 8. Identical
residues are denoted by
asterisks. The "villin homology
box" of residues 414-444
(upper sequence) is discussed
in the text.
|
