Nucleic Acids Research
Volume 6 Number 1 January 1979
Cell specific antiserum to chromosome scaffold proteins
Ailsa M.Campbell*, R.C.Briggs, R.E.Bird and L.S.Hnilica
Departments of Biochemistry and Molecular Biology, Vanderbilt University, Nashville, TN 37232, USA Received 25 August 1978
ABSTRACT Antiserum has been raised to a chromosomal protein fraction specific for Hela cells. The immunoactivity is located in the transcriptionally inactive regions of log phase chromatin. Digestion of metaphase chromosomes results in the purification of the immunoactivity in the scaffold region of the chromosomes. Extensive nuclease digestion of the scaffolds results in loss of activity. The data suggest that some of the proteins in the scaffold area are both tight binding and cell specific and may therefore play a sophisticated role in gene expression. INTRODUCTION
Immunochemical techniques have been employed in the analysis of chromosomal proteins in many laboratories.
Evidence has been presented to show that groups of non-histone proteins can be injected into animals to raise
antisera which are tissue specific (1-5), species specific (6-8), and tumor
specific (3,5,10,11).
The nature of the primary antigen has varied widely
between laboratories.
Some have immunised with whole chromatin (3,6,7),
some with non-histone proteins which have been freed of other nuclear
components (9,10) and some with dehistonised chromatin in which the histones and the bulk of the non-histone proteins have been removed by treatment of the chromatin with solutions of high ionic strength containing urea (1,2,4, 5,8). In the latter case, a variety of responses is obtained when dehistonised chromatin is injected into rabbits. A few respond by making antibodies to the DNA component of the antigen and some by making antibodies to that fraction of the non-histone proteins which is highly conserved among tissues. A substantial number of animals, however, produce antiserum which can be demonstrated by both microcomplement fixation and immunocytochemical localisation to be specific for the tissue of origin of the chromatin. The serum from these animals can then be used to attempt to identify the
C) Information Retrieval Limited 1 Falconberg Court London Wl V 5FG England
205
Nucleic Acids Research chromosomal localisation of the cell specific antigens. The localisation of any specific protein or set of proteins on the superstructure of the chromosome or chromatin requires a clear view of the Thus, overall architecture or packing of DNA inside the cell nucleus. while specific nuclear non-histone proteins can be readily assigned to identifiable organelles such as the nucleolus (10), the chromocentre (12), individual chromosomes in a cross species cell hybrid (8), or bands in a polytene chromosome (9), those proteins which cannot be readily identified with a morphologically distinct nuclear entity are less readily assigned A functional distinction which has to a particular chromosomal location. recently become available is that of the localisation of non-histone chromosomal proteins among the "active" as opposed to "inactive" cistronic regions (13-17). Selective nuclease digestion can produce an enrichment of various non-histone proteins in the supernatant of the treated chromatin while the undigested material can be demonstrated to have been depleted of the DNA which is known to be actively transcribed in any particular cell type. Such experiments can lead to the association of these readily released nonhistone proteins with the propagation of transcription, though the view has been challenged (29).
In other cases, non-histone proteins have been
identified with the mononucleosomal component of the chromatin indicating strong affinities for the nucleosomes themselves (18,19), or with the
oligonucleosomal component, indicating affinity for the internucleosomal DNA (20).
It is therefore possible not only to localise any specific non-
histone protein on the "inactive" or active nucleosomes but also to further
pinpoint the attachment site of the protein on or among the nucleosomal particles. The experiments described here show the localisation of an antigen or
group of antigens which are specific for Hela cells in comparison to other human tissues or cell culture lines.
The antigen can be shown to be a DNA
binding protein, located in the "inactive" fraction of the chromatin and not present on either the nucleosomes or the bulk of the internucleosomal DNA. It is, however, greatly enriched in the "scaffold" region of metaphase
chromosomes (21) suggesting that such proteins may play a more sophisticated role in gene expression than their name implies.
METHODS Hela cells were grown in modified eagles medium supplemented with 5% calf serum.
206
Metaphase chromosomes were prepared according to the-method of
Nucleic Acids Research Stubblefield and Wray (22).
Dehistonised chromatin was prepared as
described previously (23) and used to immunise New Zealand white rabbits as Scaffold material was obtained described by Chytil and Spelsberg (1). Adolph et al. (21) except that the of essentially according to the method nuclease digest was fractionated on a 2.5 x 100 cm Biogel A 50 M column. Dextran sulphate and heparin were used to dehistonise the chromosomes rather than 2.0 NaCl (21). Chromatin fractionation by partial DNase II digestion followed by
precipitation with MgCl2 was by the method of Gottesfeld (13). The immunological reactivity of the chromatin was determined by the method of quantitative microcomplement fixation as described by Wasserman Washed sheep red blood cells (GIBCO Diagnostics) were and Levine (24). activated with anti-sheep red blood cell serum (Capell Laboratories). Guinea pig serum complement (Capell) was titrated to give 100% cell lysis of the activated sheep red blood cells after 30 min of incubation at 370C.
The chromatin concentration ranges were tested by incubating various chromatin dilutions 18 hr at 40C, in the presence of titrated complement, Activated sheep red blood cells with 0.1 ml of antiserum diluted 1:200. were then added and after a 30 min incubation at 370C the extent of red All the anticell lysis was determined spectrophotometrically at 413 nm. sera used also showed positive reactivity on immuno localisation with per-
oxidase anti horse radish peroxidase reagent (25) when assayed on logarithIt was not possible to achieve immunochemical staining of metaphase chromosomes or scaffold using this technique as the extensive washing during the immunolocalisation procedure
mic cells grown on glass slides (Fig. 2).
removed all the chromosomes from the slide, and all common fixatives desHowever, the metaphase material could be troyed immunological activity.
readily identified by conventional staining. Trypsin digestion was carried out with 1000 units/ml of (Sigma) trypsin Trypsin inhibitor (1 mg) was then per mg chromatin DNA for 30 min at 37 The added to prevent trypsin inhibition of the complement fixation assay. Ribonuclease digestion buffer was 10 mMtris, 40 mM NaCl 1 mM EDTA pH 7.5.
(200 units/ml/mg chromatin) was carried out under the same conditions. Gel electrophoresis of the phenol extracted DNA was carried out in 6% polyacrylamide gels with Hae III fragments of Col El DNA as markers (30,31). Protein electrophoresis was carried out by the method of Weber and Osborn (32) modified to contain urea and a 4% stacking gel to allow for protein separation in the presence of DNA (33).
207
Nucleic Acids Research RESULTS
Figure 1 shows the complement fixation activity of the Hela antiserum. It is clear that there is little or no reactivity with other human tissues or tumor cell lines, or indeed with rat tumour cells.
This has been
The antiserum rereadily confirmed by immunolocalisation (Fig. 2) (26). acted, however, with equal efficiency with chromatin from Hela S3 cell lines. The cell cycle appearance of the dependence of this antibody has been
extensively analysed (26) and it is known to be present in similar amounts
throughout the cell cycle, unlike antiserum raised to human LNSV cell lines The material is clearly a by very similar immunisation procedures (8).
100
w
x
z
z
w 2Xw 50 -j a.
I
0
5 1.25 25 uG DNA IN CHROMATIN FIGURE 1:
10
Complement fixation of Hela cell chromatin with antiserum * o o A A
Hela chromatin Wl 38 human lung fibroblast chromatin Human placental chromatin Human placental DNA, double stranded M.W.>la7 Normal human breast, human breast tumor, lung tumor and colon tumor and Novikoff Hepatoma chromatin.
The antiserum dilution was 1/200. Two animals were tested for the full range of specificity and another two for a limited range (human breast, endometrium and DNA). Three animals showed antibody at low titre with no specificity and were not further used. 208
Nucleic Acids Research strongly antigenic component of the Hela dehistonised chromatin as all the animals which were immunised produced antiserum with the same range of specificity. The antigen is absolutely dependent on the presence of DNA for its immunoactivity (Fig. 2) (26). The results of fractionation into "active" and "inactive" chromatin are shown in Figure 3. The magnesium precipitable material shows a slight increase in activity over the untreated chromatin suggesting that the antigenic protein(s)
are not located on actively transcribing regions of the Essentially similar results are obtained when DNase I susceptibility of the chromatin is tested. Such data indicate that the activity does not lie on or among those nucleosomes which are on the DNA undergoing trans-
genome.
cription but it remains possible that the activity is located on or among the bulk of the rest of the nucleosomes. If this were the case, it should
ts,,; #E.
f:i::djA
A
FIGURE
2:
Immunocytochemical
localisation
A
Hela
cells
with
immune
serum
B
Hela
cells
with
immune
serum
2000
units
micrococcal
cells
with
C
Hela
D
The
by
sane phase
group
cells
Hela
specific
pretreated
proteins
with
nuclease
preimmune of
of
X
as
serum in
C
v?isualised
contrast
209
Nucleic Acids Research l00 r
Precipitote
0
x Lu.
aI 0.
2
0.
Supernatont 2
4
6
8
10
.ugDNA in chromatin
FIGURE 3:
DNase II fractionation of chromatin from logarithmically growing Hela cells
be possible to locate the antigenic material among the oligonucleosomes produced by microccocal nuclease digestion as this is known to have little Figure 4 shown the selective action among the two types of chromatin. Little or no results of such an experiment using Biogel fractionation. the fractionated of area in any be detected activity could at any stage Such data cleardigest except in the very high molecular weight material. ly indicate that the activity can be enriched in the high molecular weight
product of a microccocal nuclease digest but do not necessarily indicate Metaphase chromosomes were therethat the material is scaffold protein. fore purified, digested and dehistonised according to the method of Adolph The results of such fractionation (Figure 5) show a clear et al. (21). but small high molecular weight peak which has only 1% of the DNA but all The protein/DNA ratio of the of the immunological activity (Figure 6).
material is 3.3/1, compared to that of the whole chromosomes of 2.9/1, clearly indicating that the increase in activity in the purified scaffolds is not dependent on whether DNA or protein is used as an assay standard. Microccocal nuclease digestion results in extensive precipitation As a control experiment to test of histones and non-histone proteins. whether the digestion produced some type of aggregate which was enriched in the antigen(s) and not readily disassociated by heparin and dextran Sonication sulphate, the metaphase chromosomes were extensively sonicated.
210
Nucleic Acids Research
i-
OD 2600.2 -
+
0-
02Lt
1 .1
FIGURE 4:
0
100
0.46
0 4
j,2
6 21 26 31 36 41 46 51 56 TUBE NO.
Biogel fractionation of log phase Hela chromatin digested with microccocal nuclease. Approximately 10 OD260 units were applied to each column.
The arrows indicate the percentage complement fixed per 2.5 ug chromatin DNA by a 1/200 solution of antiserum.
readily releases the antigen(s) into a non sedimentable fraction of the When this soluble material was digested with microccocal chromatin. nuclease and analysed on the Biogel column, no front peak was obtained (Fig. 5). The appearance of the antigenic proteins among the scaffolds does not, therefore, reflect an artifact of digestion. A second type of control experiment of this nature is shown in Figure 7. It could be argued that an antigen which was readily dislodged from its DNA binding site might be located among the nucleosomes in the cell but be displaced during digestion. Many of these non histone proteins precipitate when they are not bound to DNA and so the antigenic material could attach
itself to the residual scaffold DNA. However, the protein(s) are clearly very tightly bound to DNA as levels of microccocal nuclease of greater than 400 units are required to abolish immunoactivity when the Similar material is digested in the presence of protease inhibitors. results are obtained on the digestion of whole chromatin (25). Nucleosomes, on the other hand are susceptible to digestion with ten fold
211
Nucleic Acids Research
OD260 C
OD20
FRACTION NUMBER
FIGURE 5:
Metaphase chromosome fractionation according to the method of Adolph et al. (21).
(a) Normal metaphase chromosomes (b) Sonicated metaphase chromosomes
smaller amounts of enzyme.
It is, therefore, unlikely that the antigen
would be dislodged from oligonucleosomes in preference to the nucleosomes Pretreatment of the chromatin with trypsin also resulted themselves. in the loss of antigenicity but RNase did not affect the antigen. Figure 8
shows the proteins contained in the scaffold fraction.
There are six to eight major bands and several minor ones.
All preparatiors
consistently appeared to show a small amount of histone, even when further This small molecular weight dehistonised and recycled through the column. material may, however, be non histone protein which binds tightly to The data cannot be directly heterochromatic DNA of Hela Cells (34).
compared to other published work (21) as this entailed fluorographic analysis of methionine labelled protein, a procedure which must be expected to give There are no banding patterns dependent on the amino acid composition. Figure 9 also shows that detectable nucleosomal DNA fragments (Fig. 9). 212
Nucleic Acids Research
0
w
x
z
w LJI
w
-J a-
0 C-
FIGURE 6:
Complement fixation of metaphase chromosomes and scaffolds
A Scaffolds O Small molecular weight material (Tubes 31-51 on Figure 5(a)) * Metaphase chromosomes
the scaffold DNA can be largely localised into a diffuse band in the size range 80 to 140 base pairs, but that at least one larger discrete DNA component in the size range of 4000 nucleotides is also present.
The antigen is absolutely dependent on the presence of homologous DNA for its immuno-
activity but it is not yet clear which DNA size class contains the binding
site.
DISCUSSION
The production of antiserum which has specificity for Hela chromatin in contrast to chromatin from other human tissues and cell lines indicates some distinct features not observed in the Novikoff hepatoma system (11). In the latter case the antigen can also be shown to be present in other rat tumor cells such as the Walker carcinosarcoma and also in embryonic liver. 213
Nucleic Acids Research
Xl 0
w x
U-
50-
w
-j a.
0
0.1
0.2
03
pG DNA IN SCAFFOLD FIGURE 7:
Microccocal nuclease digestion of scaffold material. Peaks were pooled and concentrated in dialysis tubing in Sephadex G200. Microccocal nuclease was then added as shown to samples containing 0.5 ml chromosome DNA at 6 pg/ml * No nuclease * 150 units nuclease 0 450 units nuclease
The Hela antigen, on the other hand, does not appear to react with other human tumors and human placental trophoblasts.
al tissue was not obtainable for assays.
Unfortunately, human cervic-
It is clear, however, that the
antiserum is either specific to the human cervix and the Hela tumor line
derived from it, or to the Hela cells alone. Cell lines which have been in culture for more than two decades may have undergone many changes and the possibility that the specific nature of the antiserum has recent evolutionary origins cannot be ruled out, though there is no precedent amont the non It is, however, unlikely that a protein histone proteins of animal tumors. should undergo such extreme evolutionary divergence in so short a time and
indeed any protein which were to do so becomes an absorbing scientific problem in its own right. It, therefore, remains most likely that the antito both the tissue and the tumor or to the tumor alone. is either gen specific
214
Nucleic Acids Research
FIGURE 8:
Gel electrophoresis of the proteins in the antigen enriched fraction.
The localisation of such a specific protein on the chromosome is, therefore,
of major importance in the analysis of gene expression in differentiated
cells. The absolute dependence of the antigen on the presence of DNA for its activity indicates that it is the same type of antigen as has been shown to be present in Novikoff hepatoma (11). The results of the digestion experiments show that both protein and DNA must be present and neither component
alone will react.
Presumably either the antibody recognises sites on the
protein which are dependent on binding DNA for their key conformational
features, or it recognises features on the DNA only present in combination A relevant point relating to the nature of the antigen with the protein. The requirement for DNA is its extreme insolubility in the absence of DNA. may therefore reflect a simple physical requirement to be effectively solubilised so that the antigenic sites can be exposed and this seems to be the most rational explanation of the observation.
Nuclease digestion of chromatin frequently leads to the formation of
215
Nucleic Acids Research
FIGURE 9:
Gel electrophoresis of the DNA in the antigen enriched fraction. The marker sizes from the smallest upwards are 43,69,84,179 and 247 base pairs. The high molecular weight markers correspond to groups of bands of a similar size (30,31).
residual high molecular weight heterochromatic material which is selectively
digested.
It is for this reason that analysis of the morphologically
distinct chromosomes is of exceptional value in localisation experiments of this type. Digestion of log phase nuclei can readily lead to the enrichment of the residual chromatin in nucleolar material or material containing
the satellite DNA component.
The Hela antiserum is not specific to the
nucleoli as it shows no specific staining on immunolocalisation and reacts
216
Nucleic Acids Research readily with sonicated chromatin in which the nucleoli have been removed For similar reasons we have no evidence which suggests by centrifugation, that the protein is located among the satellite proteins but the possibility cannot be finally excluded as immunocytochemical staining of metaphase
chromosomes was not possible (see Methods section).
Indeed the digestion
of individual metaphase chromosomes in other laboratories has been claimed to give complete digestion with only kinetochores present as morphologically It is, however, doubtdistinct entities at the end of the reaction (12). ful as to whether the techniques employed in such a study would bave detect-
ed the scaffold material. The concept of scaffold proteins on the chromosome implies a strictly
Metaphase chromosomes structural role analogous to that of the histones. are now well known to be organised into discrete loops (27) each of which The fact that interphase cells are has about 100 kilobase pairs of DNA. also believed to exist in domains of similar dimensions (28) clearly indicates that the proteins at the base of the loops are not present for only a minor metaphase packagirg role and evidence is accumulating which suggests It that some form of scaffold is present throughout the cell cycle (21). is not possible to state categorically whether each loop represents a gene, or set of linked genes as the actual size of the gene coding for any
particular protein is known in only a very few cases but it presents an The initiation of transcription of any extremely plausible hypothesis. particular loop may then be at least in part due to interaction of the regulatory proteins with those scafoold components which control the events Thus, regulatory proteins which are occurring at the base of each loop. specific for different cell types may be expected to copurify with the chromosome scaffolds. We conclude that the specific antiserum produced to dehistonised Hela chromatin reacts preferentially with those proteins which are not readily
digested free of their bound DNA and are located physically on, or close to, the scaffold region of metaphase chromosomes. ACKNOWLEDGEMENTS We should like to thank Drs. G. & J. Stein for providing the Hela R.E.B. thanks the University Research cells for earlier immunisations. Council and the Natural Science Council of Vanderbilt University for support. This investigation was supported by a grant from the National Institutes of Health (CA-18389).
217
Nucleic Acids Research *Present address: Department of Biochemistry, University of Glasgow, Glasgow GI 2 8QQ, UK REFERENCES
1. 2. 3. 4.
5.
6. 7. 8.
9. 10.
11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
21. 22. 23.
Chytil, F. & Spelsberg, T.C. (1971) Nature New Biology 233 215-218 Wakabayashi, K. & Hnilica, L.S. (1973) Nature New Biology 242 153-155 Zardi, L., Lin, J.C. & Baserga, R. (1973) Nature New Biology 245 211-213 Briggs, R.C., Chiu, J.F. Hnilica, L.S., Chytil, F. Rogers, L.W. & Page, D.L. (Cell Differentiation - in press) Campbell, A.M., Briggs, R.C., Zimmer, M.S.. Krejewska, W.M., Chiu, J.F. $ Hnilica, L.S. (1978) (Biological Markers in Neoplasia, Elsevier p. 369-384. Ed. Ruddon, R.). Okita, K. & Zardi, L. (1974) Exp. Cell Res. 86 59-68 Zardi, L. (1975) Eur. J. Biochem. 55 231 Tsutsui, V., Chang, H.L., & Baserga, R. (1977) Cell Biology Int. Rep. 1 301-308 Silver, L.M. & Elgin, S.C.R. (1976) Proc. Natl. Acad. Sci. U.S. 73, 423-427 Davis, F.M., Busch, R.K., Yeoman, L.C. & Busch, H. (1978) Cancer Research 38 1906-1915 Fujitani, H., Chiu, J.F. & Hnilica, L.S. (1978) Proc. Natl. Acad. Sci. U.S. 75 1943-1946 Rattner, J.B., Krystal, G. & Hankalo, B.A. (1978) Chromosoma 66 259-268 Gottesfeld, J.M. & Partington, G.A. (1977) Cell 12 953-962 Weintraub, H. & Groudine, M. (1976) Science 193 848-856 Garel, A. & Axel, R. (1976) Proc. Natl. Acad. Sci. U.S. 73 3966-3970 Panet, A. & Cedar, H. (1977) Cell 11 933-940 Vidali, G., Boffa, L.C. & Allfrey, V.G. (1977) Cell 12 409-415 Goldknopf, I.L., French, M.F., Musso, R. & Busch, H. (1977) Proc. Natl. Acad. Sci. U.S. 74 5492-5495 Goodwin, G. H., Woodhead, L. & Johns, E.W. (1977) FEBS Letters 73 85-88 Neumann, J., Whittaker, R., Blenherd, B. & Ingram, V. (1978) Nucleic Acids Research 5 1675-1687 Adolph, K., Chang, S.M., Paulson, J.R. & Laemnli, U.K. (1977) Proc. Natl. Acad. Sci. U.S. 74 4937-4941 Wray, W. & Stubblefield, E. (1970) Exp. Cell Res. 59 469-478 Spelsberg, T.C., Hnilica, L.S. & Ansevin, A.T. (1971) Biochim. Biophys. Acta. 228
24. 25.
550-561
Wasserman, E. & Levine, L. (1961) J. Immunol. 87 290-291 Sternberger, L.A. (1974) Irmnunocytochemistry (Prentics Hall, Englewood
Cliffs) 26. 27. 28. 29.
30. 31. 32. 33. 34.
218
Briggs, R.C., Stein, G., Stein, J., Campbell, A.M., Chiu, J.F. & Hnilica, L.S. in preparation Adolf, K.W., Cheng, S.M. & Laemmli, V.K. (1977) Cell 12 805-816 Campbell, A.M. (1978) Trends in Biochemical Sciences 3 107-108 Goodwin, G.H. & Johns, E.W. (1978) Biochim. Biophys. Acta. 519 279-284 Tomizawa, J., Ohmori, H. & Bird, R.E. (1977) Proc. Natl. Acad. Sci. 74 1865-1869 Oka, A. & Takanami, M. (1976) Nature 264 193-196 Weber, K. & Osborn, M. (1969) J. Biol. Chem. 244 4406-4411 Smith, M.C. & Chae, C.B. (1973) Biochim. Biophys. Acta 317 10-15 Pederson, T. & Bhorjee, J.S. (1975) Biochemistry 14, 3238-3242