Fibronectin observed in the nuclear matrix of HeLa tumour cells

Fibronectin observed in the nuclear matrix of HeLa tumour cells
GEROLD ZERLAUTH, J^ZEFA WESIERSKA-GADEK and GEORG SAUERMANN*
Institute of Tiniiorbiology-Cancer Research of the University of Vienna, A-1090 Vienna, Austria
• Author for correspondence
Summary
We have investigated the intracellular distribution of insoluble fibronectin in HeLa tumour
cells. By indirect immunofluorescence microscopy fibronectin •was detected in the nuclear
region, but not in the region of the cell surface.
Isolated nuclei and isolated nuclear matrices also
stained for fibronectin. By quantitative enzymelinked immunosorbent assay (ELISA), fibronectin was found almost exclusively in the subcellular fraction of isolated nuclear matrices. Using
immunoblotting techniques fibronectin was
detected in nuclear matrices isolated from cells
grown in standard and in fibronectin-depleted
medium. The data demonstrate that fibronectin in
HeLa tumour cells is preferentially associated
with the nuclear matrix.
Introduction
skeletal framework of the interphase nucleus (Berezney
& Coffey, 1977).
Fibronectins are large glycoproteins which are considered to play a major role in cell-substratum and
cell-cell interaction. Insoluble fibronectin has been
found on the surface of most untransformed cells and
in basement membranes. Alterations in the synthesis
and localization of fibronectin have, however, been
observed to result from the malignant transformation
of cells. Though capable of synthesizing the glycoprotein, different types of tumour cells have been
reported to be depleted of cell surface fibronectin (for
reviews, see Hynes & Yamada, 1982; Morgan &
Garrod, 1984; Vaheri & Mosher, 1978). There are also
indications of an alternative splicing of fibronectin
mRNA (Paul et al. 1986) and of structural differences
at the protein level in transformed cells (Borsi et al.
1985; Matsuura & Hakomori, 1985). Furthermore, the
presence of nuclear fibronectin in autopsy specimens of
human liver tumours has been reported (Jagirdarei al.
1985).
In the present study the intracellular distribution of
insoluble fibronectin in HeLa tumour cells has been
investigated by immunofluorescence microscopy,
immunosorbent assays and immunoblotting. The data
demonstrate that the fibronectin of the tumour cells is
found in association with the nuclear matrix, the
Journal of Cell Science 89, 415-421 (1988)
Printed in Great Britain © The Company of Biologists Limited 1988
Key words: fibronectin, nuclear matrix, HeLa tumour
cells.
Materials and methods
Cell culture
HeLa S3 cells were grown at 37°C in monolayer cultures in
Eagle's Minimal Essential Medium supplemented with 10%
foetal calf serum in an atmosphere of 5 % CO2 in air. In
some experiments fibronectin-depleted foetal calf serum was
used. The fibronectin. was removed by affinity binding to
gelatin-Sepharose.
Isolation of nuclei and nuclear matrices
The nuclei were isolated by variations of the methods of
Hodge et al. (1977) and Capco et al. (1982) and the nuclear
matrices were isolated following the method of Van Eekelen
& Van Venrooij (1981) as described previously (WesierskaGadek & Sauermann, 1985). Briefly, washed HeLa cells were
homogenized in hypotonic buffer in the presence of 0-5%
Nonidet P-40. After sucrose step centrifugation, the nuclei
were washed in 1 % Nonidet P-40, 0-5 % sodium deoxycholate in low-salt buffer, and subsequently digested with
5 0 0 / i g m r ' DNase I and 100/igml" 1 pancreatic RNase.
After a sucrose step centrifugation the nuclease-treated nuclei
were twice extracted with 0-4M-ammonium sulphate and
washed. All solutions except that used for nuclease digestion
contained the protease inhibitors iY-cr-tosyl-L-lysine chloromethyl ketone, A'-cr-tosyl-L-phenylalanine chloromethyl
415
ketone, and aprotonin, at concentrations of 1/igml ' each,
plus 1 mM-pepstatin and 1 mM-phenylmethylsulphonyl fluoride.
Indirect immunofluorescence microscopy
Cells grown on glass coverslips were fixed in 3-7% formaldehyde at room temperature for 20min, then permeabihzed
with 0-2% Triton X-100 in phosphate-buffered saline for
1 min (Beyth & Culp, 1984). Isolated nuclei and nuclear
matrices were applied directly to glass coverslips and fixed.
The washed samples were incubated with rabbit anti-human
fibronectin antibody (Cappel Lab., Cochranville, PA) in
1:40 dilution at 37°C for 30min, then washed and incubated
with anti-rabbit immunoglobulin G (IgG) coupled to rhodamine (Dakopatts, Glostrup, Denmark). In control experiments with human fibroblasts and HeLa tumour cells the
anti-human fibronectin antibody was either omitted or preabsorbed with fibronectin. These specimens did not stain.
SDS—polyacrylamide gel electrophoresis and
immunodetection of fibronectin
Proteins were separated by electrophoresis in 7 5 % and 12 %
polyacrylamide gels as described by Laemmli (1970), except
that 7 M-urea was included in the gel buffer. Routinely, 5 % /3mercaptoethanol was included in the sample buffer. The
separated proteins were electrophoretically transferred to
nitrocellulose sheets as described by Towbin et al. (1979)
with the addition of 0-01 % SDS in the blotting buffer. A
blotting time of 40 h ensured quantitative transfer of high
molecular weight proteins (Lee el al. 1984). To detect
fibronectin, the blots were incubated with horseradish peroxidase-conjugated anti-human fibronectin antibody (Dakopatts) in 1:200 dilution (Figs 3-5). Colour formation by
horseradish peroxidase was induced by incubation with
chloronaphthol and hydrogen peroxide (Ruoslahti et al.
1982). Alternatively, the blots were incubated with a monoclonal anti-human fibronectin antibody (IST-1) specified as
nonreacting with bovine fibronectin (Sera-Lab), and then
exposed to biotinylated anti-mouse IgG and, finally, to
[ l25 I]streptavidin (Fig. 6). The immune complexes were
detected by autoradiography.
Enzyme-lmked immunosoi'beiit assay (ELISA) for
fibroneclin
Cells and subcellular fractions were extracted with 6 M-urea,
1 % Triton X-100 and 1 mM-phenylmethylsulphonyl fluoride
in phosphate-buffered saline (Mosher & Vaheri, 1978). The
fibronectin concentration was determined by an ELISA
method described previously (Zerlauth & Wolf, 1984). In
brief, samples were placed in microtitre plates coated with
gelatin (Type I; Sigma Chemical Co., St Louis, MO). The
gelatin-bound fibronectin was quantified by the use of
horseradish peroxidase-conjugated anti-fibronectin IgG
(Dakopatts) as antibody and of o-phenylenediamine as enzyme substrate. In each assay a standard curve of human
plasma fibronectin was made. The curve was linear between 4
and 40 ng of fibronectin and values obtained for identical
samples assayed in parallel varied less than 5 %. The protein
content was measured according to Bradford et al. (1976).
416
G. Zerlauth et al.
Results
Immunofluorescence micmscopy for fibronectin
Indirect immunofluorescence studies were performed
to localize fibronectin in cultured HeLa tumour cells.
Cells grown on glass coverslips were fixed in situ and
the glycoprotein identified by the use of anti-human
fibronectin antibody. Bound antibody was then stained
with rhodamine-conjugated second antibody.
The unpermeabilized cells did not stain in the
cellular or pericellular region (not shown), indicating
the absence of extracellular fibronectin. To permit the
passage of the antibodies through the plasma membrane, cells were permeabilized with Triton X-100.
The permeabilized cells stained intensely and
uniformly for fibronectin in the nuclear region
(Fig. 1A). The experiments, thus, indicated the existence of fibronectin in the tumour cell nucleus.
Fig. IB shows that isolated nuclei also stained for
fibronectin. Apparently, fibronectin is tightly associated with the nucleus, since the nuclei had been
extensively washed with ionic and non-ionic detergents
during the isolation procedure.
To investigate whether the nuclear fibronectin was
associated with the chromatin or with the nuclear
matrix, the nuclei were digested with DNase I and the
chromatin constituents extracted with high-salt buffer.
To ensure the removal of ribonucleoproteins from the
nuclear matrix, the nuclei were also treated with
ribonuclease A. SDS-gel electrophoresis revealed that
the chromatin constitutents were effectively separated
from the nuclear matrix (Fig. 2). Comparison of the
10-20 K (K = 103Mr) zones of lanes D and E demonstrates that the bulk of the core histones was released by
high-salt treatment of the nuclease-treated nuclei. As
regards the DNA, previously published experiments
have shown that less than O'l % of the nuclear DNA
remained in the isolated nuclear matrix prepared as
above (Wesierska-Gadek & Sauermann, 1985). The
isolated nuclear matrix shown to be free of chromatin
constitutents also stained for fibronectin in immunofluorescence microscopy (Fig. 1C). This suggests that
fibronectin in HeLa tumour cells occurs in association
with the nuclear matrix.
To quantify the fibronectin in the different subcellular fractions an ELISA was used. As the assay used is
based on two selection criteria for fibronectin, i.e.
gelatin binding plus interaction with a specific antibody, it is a reliable method for the analysis of small
amounts of fibronectin in a complex mixture of proteins (Zerlauth & Wolf, 1984). Table 1 shows that
fibronectin was present in the isolated nuclei, while it
was not detectable in the postnuclear supernatant of the
cell homogenate. Furthermore, fibronectin could not
be detected in the combined supernatants resulting
from nuclease and high-salt treatment of isolated
Fig. 1. Immunofluorescence
detection of fibronectin.
A, permeabilized cell; B, isolated
nuclei; C, isolated nuclear matrix.
Staining with rhodamine-conjugated
anti-human fibronectin as in
Materials and methods. D, isolated
nuclei; E, isolated nuclear matrices
in phase contrast. X700.
nuclei. Apparently, the fibronectin content of the
chromatin-derived fraction is very low. Significant
amounts of fibronectin were, on the other hand,
observed in the isolated nuclear matrix. Approximately
90 % of the nuclear fibronectin was present in the
nuclear matrix.
Characterization offibivnectin by imtnunoblotting
To characterize further the fibronectin detected in the
isolated nuclear matrix, matrix samples were electrophoresed in SDS-polyacrylamide gels. The separated
proteins were electrophoretically transferred to nitrocellulose membranes and the blotted fibronectin was
identified by reaction with anti-human fibronectin
antibody. Coomassie Blue staining revealed a number
of protein bands in the nuclear matrix (Fig. 3, lane 1).
The anti-human fibronectin antibody recognized a
double band characteristic of the fibronectin subunit at
approximately 220K (Fig. 3, lane 2). This was observed after electrophoresis under reducing conditions.
When, however, the routinely employed /3-mercaptoethanol was omitted from the sample buffer, the
antibody detected the 440 K fibronectin dimer as an
additional band (Fig. 4B lanes 1, 3). These results lend
further support to the identification of fibronectin in
the isolated nuclear matrix.
To ensure that the fibronectin originated from the
HeLa tumour cells and not from the calf serum present
in the cell culture medium, control experiments were
performed. First, fibronectin was removed from calf
serum by affinity binding. Nuclear matrices were
prepared from HeLa tumour cells grown in fibronectin-depleted medium. Fig. 5 shows a blot of the
Fibivnectin in the nuclear matrix
417
gelatin-bound fraction of the nuclear matrix sample. It
can be seen that the polyclonal antibody still recognized
the characteristic fibronectin double band at 220K.
Second, experiments were performed with anti-human
fibronectin monoclonal antibody, IST-1, which does
not cross-react with bovine fibronectin (Zardi et al.
1980). The antibody has been described to be specific
for the Hep-2 domain of human fibronectin (Sekiguchi
et al. 1985). Fig. 6 shows that the antibody recognized
Table 1. Distribution offibronectin in subcellular
fractions
Fraction
Postnuclear supernatant
of cell homogenate
Nuclei
DNane-, RNase-,
(NH4)2SO4-released
Nuclear matrix
Protein per
2X108 cells
( m g)
Fibronectir i
90-2
*
20-9
16-7
157
•
2-1
ISO
(ng)
Amount of
fibronectin
(ng)
per mg
protein
7-S
71
Isolation of subcellular fractions and assays as in Materials and
methods.
• Below the limit of detection.
M r x10 - 3
205-
116-30
94-
68-
•
f
43-
-12-5
Fig. 2. Protein patterns of the subnuclear fractions.
Isolation of subnuclear fractions and 12% polyacrylamide
gel electrophoresis as in Materials and methods. Coomassie
Blue staining. A. Nuclei; B, proteins released from nuclei
by DNase I digestion; C, proteins released by RNase A
digestion; D, proteins extracted with 0'4M-ammonium
sulphate; E, nuclear matrix.
418
G. Zerlauth et al.
I
mk
Fig. 3. Identification of fibronectin by immunoblotting.
Electrophoretic separation of nuclear matrix proteins in
7 5 % polyacrylamide gel and immunological identification
of blotted fibronectin as in Materials and methods. Lane 1,
Coomassie Blue staining; lane 2, immunostaining.
H
xio" 3
205-
116 —
94-
Fig. 4. Electrophoretic separation of
fibronectin under different reducing
conditions. Electrophoresis and
immunostaining as in Fig. 3.
A. Coomassie Blue staining;
B, immunostaining. Lanes 1, 2, human
plasma fibronectin; lanes 3, 4, nuclear
matrix proteins; lanes 1, 3, samples
dissolved in the absence of /Smercaptoethanol; lanes 2, 4, samples
dissolved in the presence of 5 % /Smercaptoethanol.
68-
B
Thus, in addition to immunofluorcscence microscopy and quantitative analysis by ELISA, electrophoretic analysis and immunoblotting with mono- and
polyclonal antibodies demonstrate the presence of
fibronectin in the nuclear matrix of HeLa tumour cells.
Mr
x 1(T3
205-
Fig. 5. Fibronectin in nuclear
matrices of cells grown in fibronectindepleted medium. Cell growth in
fibronectin-depleted medium as in
Materials and methods. The
fibronectin of isolated nuclear
matrices was bound to and eluted
from gelatin—Sepharose.
Immunostaining as in Fig. 3.
fibronectin in the nuclear matrix sample isolated from
cells grown in fibronectin-depleted medium. Using
[ l25 I]streptavidin for detection, the fibronectin double
band was not resolved in the autoradiography of both
nuclear matrix and control plasma fibronectin (Sekigu-
chxetal. 1985).
Discussion
The present data show that the intracellular insoluble
fibronectin in HeLa tumour cells is localized in the cell
nucleus. Interestingly, the major part of the nuclear
fibronectin was found in tight association with the
isolated nuclear matrix and not in association with the
chromatin. These conclusions result from concordant
data obtained by such different experimental approaches as immunofluorescence microscopy, immunoblotting and quantitative analysis of subcellular
fractions.
The nuclear matrix is operationally defined as the
insoluble structure remaining after DNase and highsalt treatment of isolated nuclei (Berezney & Coffey,
1977). It is considered to represent the main structural
Fibmnectin in the nuclear matrix
419
References
X1(T 3
BARRACK, E. R. & COFFEY, D. S. (1982). Biological
205Fig. 6. Detection of
fibronectin by a monoclonal
antibody. Electrophoresis in
7-5 % polyacrylamide gel
and immunoblotting with
the monoclonal antibody
IST-1 as in Materials and
methods. Lane 1, human
plasma fibronectin; lane 2,
nuclear matrix sample.
properties of the nuclear matrix: steroid hormone
binding. Recent Pmg. Homi. Res. 38, 133-195.
BEREZNEY, R. (1979). Dynamic properties of the nuclear
matrix. In The Cell Nucleus, vol. 7 (ed. H. Busch), pp.
413-456. New York: Academic Press.
BEREZNEY, R. & COFFEY, D. S. (1977). Nuclear matrix.
Isolation and characterization of a framework structure
from rat liver nuclei. J. Cell Biol. 73, 616-637.
BEYTH, R. J. & CULP, L. A. (1984). Complementary
adhesive responses of human skin fibroblasts to the cellbinding domain of fibronectin and the heparan sulfatebinding protein, platelet factor-4. Expl Cell Res. 155,
537-548.
BORSI, L., ALLEMANNI, G., CASTELLANI, P., ROSELLINI,
C , Di VINCI, A. & ZARDI, L. (1985). Structural
framework of the interphase nucleus (for review, see
Barrack & Coffey, 1982).
Our conclusions regarding the intranuclear localization of fibronectin are in contrast to those of Zardi et
al. (1979) who reported the existence of chromatinassociated fibronectin. However, this discrepancy can
be explained by differences in the procedures for
subcellular fractionation. From the experimental protocol of Zardi et al. (1979) it can be deduced that their
'chromatin' fraction contained, among other cellular
components, also the nuclear matrix. In our present
experiments the chromatin constituents were clearly
separated from the residual nuclear matrix by nuclease
digestion and high-salt treatment of isolated nuclei.
This is demonstrated by electrophoretic analysis of the
proteins in the individual subnuclear fractions (Fig. 2)
and by quantitative analysis of the DNA-derived material (Wesierska-Gadek & Sauermann, 1985).
The question is open as to whether the association of
fibronectin with the nuclear matrix in the HeLa
tumour cells is due to alterations of the fibronectin
molecule or to specific properties of the nuclear matrix.
As regards the properties of the nuclear matrix of
neoplastic cells, knowledge is still scarce (for review,
see Berezney, 1979). On the other hand, the formation
of fibronectin variants with differing peptide domains
(Borsi e/a/. 1985; Castellelanief al. 1986; Sekiguchie*
al. 1985) and differing binding behaviour (Borsi et al.
1985) has been observed in tumour cells. Apparently,
alternative splicing of the primary transcript of the
fibronectin gene occurs (Colombi et al. 1986; Paul et
al. 1986). Furthermore, transformation-dependent differences in the degree of branching, internal fucosylation, and terminal sialysation of fibronectin have been
reported (Nichols et al. 1986).
We are indebted to Mrs E. Kainzbauer and Mr D. Printz
for their excellent technical assistance.
420
G. Zerlauth et al.
differences in the cell binding region of human
fibronectin molecules isolated from cultured normal and
tumor-derived cells. FEBS Lett. 192, 71-74.
BRADFORD, M. M. (1976). A rapid and sensitive method
for the quantitation of microgram quantities of protein
utilizing the principle of protein-dye binding. Analvt.
Biochem. 72, 248-254.
CAPCO, D. G., WAN, K. M. & PENMAN, S. (1982). The
nuclear matrix: three-dimensional architecture and
protein composition. Cell 29, 847-858.
CASTELLANI, P., SIRJ, A., ROSELLINI, C , INFUSINI, E.,
BORSI, L. & ZARDI, L. (1986). Transformed human cells
release different fibronectin variants than do normal
cells. J . Cell Biol. 103, 1671-1677.
COLOMBI, M., BARLATI, S., KORNBLIHTT, A., BARALLE, F.
E. & VAHERI, A. (1986). A family of fibronectin mRNAs
in human normal and transformed cells. Biochim.
biophys. Ada 868, 207-214.
HODGE, L. D., MANCINI, P., DAVIS, F. M. & HEYWOOD,
P. (1977). Nuclear matrix of HeLa S 3 cells. Polypeptide
composition during adenovirus infection and in phases of
the cell cycle. J . Cell Biol. 72, 194-208.
HYNES, R. O. & YAMADA, K. M. (1982). Fibronectins:
Multifunctional modular glycoproteins. J. Cell Biol. 95,
369-377.
JAGIRDAR, J., ISHAK, K. G., COLOMBO, M., BRAMBILLA,
C. & PARONETTO, F. (1985). Fibronectin patterns in
hepatocellular carcinoma and its clinical significance.
Cancer 56, 1643-1648.
LAEMMLI, U. K. (1970). Cleavage of structural proteins
during the assembly of the head of bactenophage T4.
Nature, Land. 227, 680-685.
LEE, G., HYNES, R. O. & KIRSCHNER, M. (1984).
Temporal and spatial regulation of fibronectin in early
Xenopus development. Cell 36, 729-740.
MATSUURA, H. & HAKOMORI, S. (1985). The oncofetal
domain of fibronectin defined by monoclonal antibody
FDC-6: its presence in fibronectins from fetal and tumor
tissues and its absence in those from normal adult tissues
and plasma. Proc. natn. Acad. Sci. U.SA. 82,
6517-6521.
MORGAN, J. & GARROD, D. (1984). HeLa cells form focal
contacts that are not fibronectin dependent. .7. Cell Sci.
66, 133-145.
MOSHER, D. F. & VAHERI, A. (1978). Thrombin stimulates
the production and release of a major surface-associated
glycoprotein (fibronectin) in cultures of human
fibroblasts. Expl Cell Res. 112, 323-334.
gels to nitrocellulose sheets: procedure and some
applications. Proc. natn. Acad. Sci. U.SA. 76,
4350-4354.
VAHERI, A. & MOSHER, D. F. (1978). High molecular
weight, cell surface-associated glycoprotein (fibronectin)
lost in malignant transformation. Biochim. biophys. Ada
516, 1-25.
NICHOLS, E. J., FENDERSON, B. A., CARTER, W. G. &
VAN EEKELEN, C. A. G. & VAN VENROOIJ, W. J. (1981).
HAKAMORI, S. (1986). Domain-specific distribution of
carbohydrates in human fibronectins and the
transformation-dependent translocation of branched type
2 chain defined by monoclonal antibody C6. J. biol.
Chem. 261, 11295-11301.
PAUL, J. I., SCHWARZBAUER, J. E., TAMKUN, J. W. &
HYNES, R. O. (1986). Cell-type-specific fibronectin
subunits generated by alternative splicing. J. biol. Chem.
261, 12258-12265.
RUOSLAHTI, E., HAYMAN, E. G., PIERSCHBACHER, M. &
ENGVALL, E. (1982). Fibronectin: purification,
immunochemical properties, and biological activities.
Meth. Enzym. 82, 803-831.
SEKIGUCHI, K., SIRI, A., ZARDI, L. & HAKOMORI, S.
(1985). Differences in domain structure between human
fibronectins isolated from plasma and from culture
supernatants of normal and transformed fibroblasts.
Studies with domain-specific antibodies. Jf. biol. Chem.
260, 5105-5114.
TOWBIN, H., STAEHELIN, T. & GORDON, J. (1979).
Electrophoretic transfer of proteins from polyacrylamide
HnRNA and its attachment to a nuclear protein matrix.
J. Cell Biol. 88, 554-563.
WESIERSKA-GADEK, J. & SAUERMANN, G. (1985).
Modification of nuclear matrix proteins by ADPribosylation. Association of nuclear ADPribosyltransferase with the nuclear matrix. Eur.J.
Biochem. 153, 421-428.
ZARDI, L., CARNEMOLLA, B., SIRI, A., SANTI, L. &
ACCOLLA, R. S. (1980). Somatic cell hybrids producing
antibodies specific to human fibronectin. Int.Jf. Cancer
25, 325-329.
ZARDI, L., SIRI, A., CARNEMOLLA, B., SANTI, L.,
GARDNER, W. D. & HOCH, S. O. (1979). Fibronectin: a
chromatin-associated protein? Cell 18, 649-657.
ZERLAUTH, G. & WOLF, G. (1984). Kinetics of fibronectin
release from fibroblasts in response to 12-0tetradecanoyl-phorbol-13-acetate and retinoic acid.
Carcinogenesis 5, 863-868.
(Received 10 August 1987 -Accepted, in revisedfonn,
27 November 1987)
Fibronectin in the nuclear matrix
421