Purification of a novel, nucleoplasmin-like protein from somatic nuclei.

Abstract

We have purified a nucleoplasmin-like protein from the nuclei of somatic Xenopus laevis cells. This protein possesses a number of the distinctive features of nucleoplasmin isolated from oocytes or unfertilized eggs. The protein is recognized by both monoclonal and polyclonal antisera raised against egg nucleoplasmin. The protein has an oligomeric structure, which must be heated in SDS to completely dissociate, is acidic, phosphorylated and efficiently promotes the in vitro formation of chromatin. We have partially characterized this novel protein and because of its resemblance to nucleoplasmin isolated from oocytes or unfertilized eggs we have named this protein nucleoplasmin S.

Full text

Page 1

The EMBO Journal vol.6 no.13 pp.3945-3954, 1987
Purification of a novel, nucleoplasmin-like protein from somatic
nuclei
Matt Cotten' and Roger ChaLkley
Department of Molecular Physiology and Biophysics, Vanderbilt University,
Nashville, TN 37232, USA
'Present address: Institute for Molecular Pathology, Himmelpfortgasse 1/3,
A-1010 Vienna, Austria
Communicated by W.Franke
We have purified a nucleoplasmin-like protein from the nuclei
of somatic Xenopus laevis cells. This protein possesses a
number of the distinctive features of nucleoplasmin isolated
from oocytes or unfertilized eggs. The protein is recognized
by both monoclonal and polyclonal antisera raised against egg
nucleoplasmin. The protein has an oligomeric structure,
which must be heated in SDS to completely dissociate, is
acidic, phosphorylated and efficiently promotes the in vitro
formation ofchromatin. We have partially characterized this
novel protein and because of its resemblance to nucleoplanin
isolated from oocytes or unfertilized eggs we have named this
protein nucleoplasmin S.
Key
words:
chromatin/histone/nucleoplasmin/phosphoryla-
tion/somatic
Introduction
Nucleoplasmin and N1/N2 are acidic, phosphorylated proteins
which accumulate in association with histones in amphibian
oocytes (Kleinschmidt and Franke, 1982; Kleinschmidt et al.,
1985). These proteins promote the formation of soluble chromatin
structures in vitro, at physiological ionic strengths, from purified
histones and DNA, the property for which nucleoplasmin was
originally isolated (Laskey et al., 1978; Mills et al., 1980; Earn-
shaw et al., 1980). The notion that these proteins may play a
role in the transport and/or deposition of histones is supported
by the observation that shortly before fertilization, which trig-
gers a period of very rapid chromatin assembly, the oocyte form
of nucleoplasmin becomes massively phosphorylated (Sealy et
al., 1986; Cotten et al., 1986; Burglin et al., 1987). This
modification profoundly enhances the chromatin assembly ac-
tivity of this protein (Sealy et al., 1986; Cotten et al., 1986).
Both nucleoplasmin and N1/N2 are karyophilic proteins (De
Robertis et al., 1978; Dabauvalle and Franke, 1982) and the ac-
tive process of transport into the nucleus has been studied in detail
with nucleoplasmin (Dingwall et al., 1982; Feldherr et al., 1984;
Newmeyer et al., 1986a,b). In addition, there is evidence that
nucleoplasmin may be involved in RNA metabolism. The pro-
tein has been found associated with actively transcribed regions
of lampbrush chromosomes (Moreau et al., 1986). Immuno-
logical evidence suggests that both nucleoplasmin (Krohneand
Franke, 1980a,b) and N1/N2 (Krohne, 1985) are present in the
nuclei of a range of vertebrate cell types and are not a peculiar
feature of amphibian oogenesis. However, there are no publish-
ed reports ofthe purificationofnucleoplasminfrom a non-gamete
cell type.
We are studying the process of chromatin assembly and gene
IRL Press Limited, Oxford, England
activation in eukaryotic cells and are interested in further defin-
ing the role that nucleoplasmin might play in these processes in
somatic cell types. We set out to determine ifa nucleoplasmin-like
molecule could be isolated from the nucleus of a somatic cell.
Consistent with previous immunological observations (Krohne
and Franke, 1980a,b), we have identified a small amount of a
protein in a Xenopus kidney cell line which resembles the oocyte
form of nucleoplasmin. This protein is either present in the
cytoplasm or, more likely, it rapidly leaks out ofthe nucleus dur-
ing isolation procedures. We have also identified and purified
to homogeneity, a much more abundant, nucleoplasmin-like pro-
tein that is present in the nucleus of somatic cells. This second
protein has a slightly greater mol. wt than oocyte nucleoplasmin
but
like features. Like oocyte nucleoplasmin, this protein exists as
an oligomer (most likely either a pentamer or a hexamer) in solu-
tion, and requires boiling in SDS for complete dissociation to
the monomer. The protein is recognized on Western blots by
both an affinity-purified polyclonal and a monoclonal antibody
raised against egg nucleoplasmin. The protein binds histones in
vitro and possesses potent in vitro chromatin assembly activity.
The protein has a similar amino acid composition to oocyte
nucleoplasmin, with a slightly lower glutamic acid content, and
the protein is phosphorylated.
This second form of nucleoplasmin is present in all tissues of
Xenopus that we have examined. A similar protein is present in
chicken embryos, calf thymus, rat liver and HTC cells, a rat
hepatoma cell line. We suggest that this protein is common to
all somatic cell types and because of the protein's strong
resemblance to oocyte nucleoplasmin, we tentatively name this
protein nucleoplasmin S (for nucleoplasmin simulacrum, should
the protein be positively identified as a nucleoplasmin homologue,
the S can then signify somatic). We have yet to determine the
function of this protein in vivo. However, because the protein
is as abundant in non-replicating cells as it is in replicating cells,
the protein's function is probably not limited to replication.
it possesses a number of distinctive nucleoplasmin-
Results
Identification ofa nucleoplasmin-like molecule in somatic cells
We initiated our search for a somatic form of nucleoplasmin in
XLA cells, a cell line derived from Xenopus kidney. A 500 mM
whole cell extract was prepared and partially fractionated byam-
monium sulfate precipitation and elution from phenyl-Sepharose.
This material was analyzed by Western blotting, usinganaffinity-
purified, polyclonal, anti-nucleoplasmin antibody. Whenoocyte
or egg nucleoplasmin are analyzed with this antibody prepara-
tion an immunoreactive species of
sample is not boiled in SDS priortoelectrophoresis (Figure IA,
lane 2). However, if the sample is boiled in SDS in order to
disrupt the oligomericstructure oftheprotein,then theresulting
signal of the monomer is at least 20-fold diminished inintensity
(Figure lA, lane 1). The antibodyisapparently recognizingsome
feature of the oligomeric nucleoplasmin structure, relativelylittle
of which is preserved in the conformation of the dissociated
-150 kd is observed if the
3945

Page 2

M.Cotten and R.Chalkley
which diminishes with boiling in SDS. The immunoreactivity of
these XLA proteins appears to be authentic. No signal is observed
when the Western is probed with normal rabbit serum nor is there
a signal when the antibody is omitted and the Western is incubated
only with labeled protein A (results not shown). Furthermore,
a similar pattern of reactivity is observed if the Western isprob-
ed with a monoclonal antibody prepared against egg nucleo-
plasmin (results not shown).
Thus, consistent with previous analyses (Krohne and Franke,
*1980a,b) protein immunoreactive to antibodiesprepared against
oocyte nucleoplasmin is present in somatic cells. However, we
have observed at least two types of signal: an abundant signal
with a greater mol. wt than oocyte nucleoplasmin, which is not
celhihIin r,,-rr-hInrit atiA
acid-soluble species with a mol. wt that is similar to oocyte
nucleoplasmin. To clarify the relation between these proteins we
proceeded to purify the more abundant immunoreactive species.
_
-.
*
ani-a IAfn_alA lpcec! nhvinAantrli_
Fig. 1. Panel A: nucleoplasmin S in whole cell extracts. XLA cells were
homogenized in 500 mM NaCl, 20 mM Tris, pH 8, 7 mM
mercaptoethanol, 5 mM EDTA, 1 mM benzamidine, 0.1 mM PMSF, 0.5%
Triton X-100, and centrifuged for 30 min
HB-4 rotor. The supernatant was adjusted to 35% saturation in ammonium
sulfate, centrifuged again and this supernatant was applied directly to a
phenyl-Sepharosecolumn equilibrated with 1.5 M ammonium sulfate,
20 mM Tris, pH 8. The column was washed with three column volumes of
this buffer and the bound protein was eluted with 20 mM Tris, pH 8. The
elutedprotein was analyzed by Western blotting. Bound antibody was
localized by 1251-labeled protein A incubation followed by autoradiography at
-70°C. Lanes 1 and 2, 0.1itgegg nucleoplasmin (purified as described by
Sealy et al., 1986); lanes 3 and 4, XLA whole cell extract; lanes 5 and 6,
5% perchloric-acid-soluble protein from the XLA whole cell extract after
dialysisto remove the acid; lanes 7 and 8, the material described for lanes
5 and 6precipitated overnight at -20°C with 4 vol ethanol and dissolved in
1/26 the initial volume. The samples in odd-numbered lanes were heated in
aboiling water bath for 10 min in SDS application buffer (b); the samples
in even-numbered lanes were applied directly to the gel in SDS application
buffer, without heating (n). Panel B: screening somatic cell extracts for
nucleoplasmin. XLA cells were fractionated into cytosolic and 100, 300 and
530 mM nuclear extracts as described in Materials and methods. The
extracts were dialyzed into np storage buffer and aliquots were resolved by
SDS-acrylamide electrophoresis, transferred to nitrocellulose and probed
with affinity-purified anti-nucleoplasmin antibody. Bound antibody was
localized by 1251-labeled protein incubation followed by autoradiography at
-70°C. Lanes 1 and 2, cytoplasm; lanes 3 and 4, 100 mM nuclear
extract; lanes 5 and 6, 300 mM nuclear extract; lanes 7 and 8, 530 mM
nuclear extract; lanes 8 and 9, material remaining with the nucleus
following 530 mM extraction; lane 11,
lane 12 0.1 ug egg nucleoplasmin. Samples in odd-numbered lanes were
heated in a boiling water bath for 10 min in SDS application buffer before
electrophoresis (b), samples in even-numbered lanes were applied in
application buffer without boiling (n).
(3-
at 10 000 r.p.m. in a Sorvall
1jigpurified egg nucleoplasmin;
monomer. In XLA cell extracts we find an immunoreactive,
>200-kd protein (Figure IA, lane 4). Similar to egg nucleo-
plasmin, the reactivity of this XLA protein is diminished when
thesampleis heated in SDS (Figure IA, lane 3). However, unlike
oocyte nucleoplasmin, a lower mol. wt, presumably monomeric,
signalispresent and its immunoreactivity is not diminished by
boiling in SDS (Figure IA, lanes 3 and 4). That a direct
precursor-product relationship exists between these two mol.
wt forms is demonstrated below.
One distinctive feature of oocyte nucleoplasmin is its solubili-
tyin 5% perchloric acid (Kleinschmidt et al., 1985), a property
shared with histone HI and the HMG proteins. To determine
if the XLA signal possesses this characteristic, the XLA extract
was adjusted to 5% PCA and the supernatant was analyzed for
nucleoplasmin content by Western blotting. We find no signal
in the unconcentrated sample (Figure la, lanes 5 and 6). If this
extract is concentrated 26-fold we still find none of the higher
mol. wt signal, but an oocyte nucleoplasmin-sized signal can now
be observed (Figures IA, lanes 7 and 8). Like the oocyte pro-
tein, this material has an immunoreactive high mol. wt signal
Screening cell fractions for nucleoplasmin S
XLA cells were divided into a series of subcellular fractions to
determine the location of nucleoplasmin S. Cells were homo-
genized in a Triton-containing buffer to prepare nuclei, these
nuclei were washed once with the Triton-containing buffer. The
initial cytoplasmic material was pooled with this Triton wash of
nuclei. It is quite possible, of course, that easily extracted nuclear
proteins may be present in this fraction. The nuclei were then
extracted sequentially with buffers containing 100, 300 and
530 mM NaCl. Finally, the remaining nuclear proteins were
released by sonicating the nuclear residue in the presence of SDS.
The proteins of these extracts were separated by SDS-acryl-
amide electrophoresis, transferred to nitrocellulose and probed
with an affinity-purified, polyclonal anti-nucleoplasmin antibody.
We find that the bulk of the major immunoreactive protein in
XLA cells, nucleoplasmin S, is released from the nucleus at
300 mM and higher ionic strength (Figure IB, lanes 5-8).
Although it is not shown here, if we extract with PCA and con-
centrate the samples as described above, the oocyte nucleo-
plasmin-sized signal can be found only in the cytoplasmic
fraction, consistent with reports that this protein is nuclear but
rapidly leaks out of the nucleus during fractionation (Krohne and
Franke, 1980a,b). It appears that only a portion of the higher
mol. wt signal is extracted from the nucleus with a single 300 mM
salt wash. We have found that repeated extractions with buffers
containing 300 mM NaCl can remove substantially all ofthe im-
munoreactive species while repeated washings with 100 mM
NaCl, even in the presence of0.5% Triton X-100, do not remove
the protein (unpublished data). The amount of material remain-
ing with the nucleus after 530 mM extraction appears to be a
significant fraction of the total signal in this experiment (Figure
iB, lanes 9 and 10) because the nuclei were extracted with
530 mM NaCl only once. Nearly quantitative removal of this
material can be obtained by repeating the 530 mM extraction.
Furthermore, ifchromatin is prepared by lysing nuclei in 10 mM
EDTA, the bulk ofthe immunoreactive protein remains associated
with the chromatin unless the ionic strength is raised above
300 mM NaCl (unpublished data).
Purification ofnucleoplasmin S
The purification of nucleoplasmin S is initiated by preparing
nuclei as described above and extracting them twice with 530 mM
NaCl (without the lower ionic strength extractions). This nuclear
extract is adjusted to 35% saturation in ammonium sulfate, the
supernatant of this preparation is applied directly to a phenyl-
3946
B
. ,:::"
A
1
aSW
40

Page 3

Nucleoplasmin-like protein from somatic nuclei
1I n
2
3
4
5
iI
oI
bn
l I
nbIDnbr1 b n I-
C
205
116
97
6a6
.._ lif
1
I
-F
~
'3;w;$
..O
*
*Ce
_,eumr
fl
4 t45
029
_w
0
21
-5
1-1
4
---r
5,
pool
')
Vol.
r
SC
Fig. 2. Phenyl-Sepharose fractionation of chromatin assembly activity. Fractions containing proteins eluted by reverse salt gradient from phenyl-Sepharose
chromatography were pooled and dialyzed into NP storage buffer. Panel A. Aliquots of 20
content by SDS-acrylamide electrophoresis followed by staining with Coomassie Blue. The positions of the nucleoplasmin S oligomer (npsO) monomer
(npsM) and of N1/N2 are indicated; m, mol. wt standards of carbonic anhydrase, egg albumin, bovine serum albumin, phosphorylase b, (3-galactosidase and
myosin with approximate mol. wts indicated in kilodaltons. Panel B. The phenyl-Sepharose pools assayed for chromatin assembly activity using 10, 30 or 90
jil of each pool (indicated by 1, 3 and 9 at the top of the figure). Assembly assays were performed as described in Materials and methods using 0.3
relaxed pBR322 per assay and a histone to DNA ratio of 1. The DNA from the assembly was resolved on a 2% agarose TPE gel and stained with ethidium
bromide. M, Supercoiled pBR322 marker; r, the relaxed pBR322 DNA upon which the assembly was performed.
ud (containing
5 jig total protein) were analyzed for protein
Mg
Sepharose column and eluted with a reverse salt gradient. The
proteins present in pooled fractions from the phenyl-Sepharose
column are shown in Figure 2A. Both nucleoplasmin S and a
pair of
oocyte N1/N2 proteins (see below) elute early in the gradient.
-100 000-kd proteins, which could be related to the
In contrast, histone HI (the only histone likely to have been ex-
tracted from nuclei under these ionic conditions) is not eluted
until late in the gradient and will not affect the assembly assays
to be described below. The monomeric/oligomeric behavior of
nucleoplasmin S is most apparent in pool 4 (Figure 2A).
3947
A
-
Pr.
.k *W
b0bwA
I..
_r~~~~~~~~~p
_;J
e
'r
n pDs
in-
B
i
I
%---411
r.-0
1 3 9
1 3 9
i 3
1
1
nIDs0
--.
t
4'.
.
N 1 N2
ftft.
I
I"
1

Page 4

M.Cotten and R.ChaLkley
'LQfit
bnbnAnnnn
lb
IjLN
rn
I
.w
b
in
n
0
Ini
I-)
ft
-\j
n
U
up
mm-
'S
1
2 3 4
5 6
7 8 9 1011 12 13
1
2 3 4 5 6 7 8 9 10 11 1223
Fig. 3. Hydroxyapatite purification of somatic nucleoplasmin. Theappropriate phenyl-Sepharose pool,afterdialysisintonp storagebuffer to remove residual
ammonium sulfate, was adjusted to 1 M NaCl, 10 mM sodiumphosphate, pH 7, and resolvedby hydroxyapatite chromatographyas described in Materials
and methods. Samples eluting in each step were resolved onduplicate 8%acrylamide-SDS gels.Onegelwas stained with Coomassie Blue to determine total
protein content (panel A); the contents of the second gel were transferred to nitrocellulose andprobedwithaffinity-purified anti-nucleoplasmin antibody to
determine nucleoplasmin content (panel B). Input, phenyl-Sepharose pool4(see Figure 2A); ft, hydroxyapatitecolumnflowthrough; 2M, the material eluted
with 2 M NaCl, 10 mM phosphate pH 7; 100 etc., the material eluted with the indicated concentration ofphosphate, pH7(in mM); m, mol. wt markers as
in Figure 2A.
I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
I
b
sl n
A
ID.IL
B
530
ggt1
Llln
--S
I
DD...1
n 1I
n
n
'1 1
e_
0.d
_
G 7a 910
...S*
.
1 2
3
58
/
9 10
Fig. 4. A summary of somatic nucleoplasmin purification. Aliquots of material at each stage of the somatic nucleoplasmin purification were resolved by
SDS-acrylamide electrophoresis on duplicate 8% acrylamide gels. One of the gels was stained with Coomassie Blue to determine total protein content (panel
A), the second gel was transferred to nitrocellulose and probed with affinity-purified anti-nucleoplasmin antibody to determine nucleoplasmin content (panel
B). Egg, Egg nucleoplasmin, 0.8 pg per lane in panel A, 0.8 pg (boiled) and 0.08 pg (not boiled) in panel B; 530 NE, 530 mM nuclear extract of XLA
cells; AS super, 35% saturated ammonium sulfate supernatant of a 530 mM nuclear extract of XLA cells; phen-Seph, the somatic nucleoplasmin-containing
pool from phenyl-Sepharose chromatography; HA, material eluted from hydroxyapatite by 400 mM phosphate; b, n, samples in SDS application buffer were
either heated for 10 min in a boiling water bath (b) before electrophoresis or applied to the gel in application buffer without the boiling water bath treatment
(n); m, mol. wt markers as in Figure 2A.
At this point in the purification, the in vitro chromatin assembly
activity ofthe partially purified fractions can be assayed (Figure
2B). We find that a certain amount of assembly activity is
associated with the phenyl-Sepharose pool containing primarily
the Nl/N2-like protein (pool 2). This is consistent with previous
3948
reports of the interaction of this protein with histones (Klein-
schmidt et al., 1985). However, a more potent assembly activi-
ty is found in those fractions which contain both the Nl/N2-like
protein and nucleoplasmin S (pool 3) and nucleoplasmin S without
the N1/N2-like proteins (pool 4). Fractions enriched in other pro-
3 3
b n
U
A
.,
m.
"a
4
I
.:
-...
el.f,,,W
-040i.
..
o..j ..,

Page 5

Nucleoplasmin-like protein from somatic nuclei
rn.S
__,>. ~
2Xr
r
gg
12~~~~~~~~~~~~~::
C_
-
Fig.5.Chromatinassemblywith
purified
nucleoplasminSandN./N2.
Chromatin assembly was performed as described in Materials and methods
using 0.3 utg relaxed pBR322 per assay. Each reaction was centrifuged for 3
mmn at 15 000 g to identify precipitated material; p, the material present in
the pellet from this centrifugation;
M, Supercoiled pBR322 marker; R, the relaxed pBR322 used as the
s, the material present in the supernatant.
substrate for the assembly reaction. In
all panels, DNA from the assembly
assays was recovered by ethanol precipitation and resolved in 2% agarose
TPE gels. The resulting DNA pattern
ethidium bromide. Panel A: assembly with no added nucleoplasm-in S (lanes
1 and 2) or increasing quantities of nucleoplasmin S (lanes 3-14) expressed
was visualized by staining with
as nucleoplasmin S to DNA ratios above the figure. The histone to DNA
1. Panel B: assembly with purified NI/N2-like proteins at the
indicated N1/N2 to DNA ratios. The histone to DNA ratio was
C: assembly with either nucleoplasmin S (NPS)
histone concentration. The indicated histone to DNA ratios were used with
nucleoplasmin S to DNA ratio of 1.25 or an NI/N2 to DNA ratio of 2.7.
ratio was
1. Panel
or NI/N2 as a function of
a
teins showed no appreciable assembly activity (pools
To determine if the chromatinassembly activity in these frac-
of nucleoplasmin
NI1/N2-like proteins, a furtherpurificationof these proteins was
obtainedusing hydroxyapatite chromatography. We have observ-
ed that both oocyte and eggnucleoplasmin bind hydroxyapatite
efficiently andrequire high phosphateconcentrations for elution,
with the more highly phosphorylated eggnucleoplasmin requir-
ing higher (upto 500 mM) phosphateconcentrations for elution
than the less phosphorylated oocyte protein (eluted by 400 mM
phosphate, unpublished observation). Both nucleoplasmin S and
1 and 5).
tionsisindeed
a
function
S
and
of the
Table I. Amino acid composition of XLA nucleoplasmin S and N1/N2
compared with that of X.laevis egg nucleoplasmin (Eamshaw et al., 1980)
and Xlaevis oocyte N1/N2 (derived from the amino acid sequence publish-
ed by Kleinschmidt et al., 1986)
Amino acid
Egg NP
NPS
Oo N1/N2
XLA 'N1/N2'
Asn
Gln
Ser
Gly
His
Arg
Thr
Ala
Pro
Tyr
Val
Met
Cys
Ile
Leu
Phe
Lys
9.2
20.0
6.7
8.2
2.2
2.2
5.4
7.2
7.3
2.1
6.2
1.0
8.2
12.5
8.1
7.2
1.3
3.7
6.4
6.6
7.2
1.9
6.2
1.4
1.4
4.3
8.9
2.5
12.1
11.4
21.9
9.0
4.4
1.5
1.9
5.9
9.0
3.4
1.4
4.2
2.4
0.8
3.1
6.8
1.0
10.7
12.1
16.8
6.6
11.2
0.3
4.1
4.6
8.0
4.9
1.3
4.1
1.2
2.0
6.0
4.3
3.7
12.0
-
3.2
6.2
2.3
11.3
Samples of eletroeluted nucleoplasmin S (Figure 6A, lanes 3 and 4) or
hydroxyapatite-purified N1/N2 (Figure 6A, lane 5) were analyzed on a
Waters Pico-tag system. Values are expressed as moles percent.
the NI/N2-like proteins bind to hydroxyapatite in 10 mM phos-
phate in the presence ofup to 2 M NaCl. The protein-bound resin
is then washed with a step gradient of increasing phosphate con-
centration. Both nucleoplasmin S and the Nl/N2-like proteins,
but very few other proteins, remain bound in 300 mM phosphate
and are eluted with 400 mM phosphate. The hydroxyapatite elu-
tion profile ofphenyl-Sepharose pool 4 (substantially free ofthe
NI/N2-proteins; see Figure 2A) is shown in Figure 3, demon-
strating the purification of nucleoplasmin S. When the nucleo-
plasmin-S-containing or Nl/N2-containing phenyl-Sepharose
pools are chromatographed on hydroxyapatite, essentially pure
(>95% by protein staining) nucleoplasmin S or N1/N2-like pro-
teins can be obtained in the 400 mM phosphate eluent. Nucleo-
plasmin S purified in this manner can be seen in Figure 3A.
Western analysis of the hydroxyapatite purification of nucleo-
plasmin S (Figure 3B) demonstrates that the immunoreactive
species is indeed the protein observed by Coomassie Blue
staining.
A summary ofthe purification ofnucleoplasmin S is displayed
in Figure 4. A substantial purification is obtained with the am-
monium sulfate fractionation. The separation of the Nl/N2-like
proteins from nucleoplasmin S is achieved by judiciously pool-
ing phenyl-Sepharose fractions (see Figure 2A; only the poolcon-
taining nucleoplasmin S alone is shown in Figures 3 and 4) and
a substantial purification ofboth proteins is obtained by hydroxy-
apatite chromatography. From the results of a number ofprepara-
tions we estimate that both nucleoplasmin S and the Nl/N2-like
proteins are present in XLA cells at 0.1-0.5% of the histone
(or DNA) mass.
Chromatin assembly activity
The chromatin assembly activity of these purified nuclear pro-
teins can now be tested (Figure 5). We have previously found
that a major problem interferingwith in vitro chromatinassembly
at physiological ionic strengths is the formation ofhighmol. wt
protein-DNA aggregates (CottenandChaLldey, 1985).The for-
mation of theseaggregatescan be monitoredby centrifuging
assemblyreactions for a briefperiodat the end of the reaction
3949