329
Biochimica et Biophysica Acta, 623 (1980) 329--338 © Elsevier/North-Holland Biomedical Press
BBA 38414
THE ISOLATION OF THREE NEW HIGH MOBILITY GROUP NUCLEAR PROTEINS
GRAHAM H. GOODWIN, ELIZABETH BROWN, JOHN M. WALKER and ERNEST W. JOHNS
Chester Beatty Research Institute, Institute of Cancer Research, Royal Cancer Hospital, Fulham Road, London SW3 6JB (U.K.) (Received November 22nd, 1979)
Key words: High mobility group protein; Non-histone protein; Chromatin; Amino acid sequence; (Calf thymus)
Summary In addition to the four high mobility group non-histone chromosomal proteins HMG 1, 2, 14 and 17 and histone H1, perchloric acid extracts of nuclei contain a number of other smaller low molecular weight proteins. Three of these proteins (HMG 18, 19A, and 19B) have been purified and characterized. Protein HMG 18 has high lysine and alanine contents, resembling histone HI. Proteins HMG 19A and 19B have high contents of basic and acidic amino acids and resemble HMG proteins 1, 2, 14 and 17. N-terminal sequence analyses of the proteins show that they are not degradation products of histones or the other HMG proteins. However, there are sequence similarities between HMG 18 and histone H5, and between HMG 19B and HMG 17, supporting the view that the HMG proteins and the lysine-rich histones are functionally related. Introduction During the course of our investigations on the high mobility group (HMG) proteins in calf thymus and liver nuclei we have noted the presence of a number of other low molecular weight proteins (I-IMG 18, 19 and 20), which co-extracted with 5% perchloric acid [1,2]. One of these proteins (HMG 20) has been identified as ubiquitin [1]. In this paper we report the isolation of three other proteins, one of which is similar in amino acid composition to histone H1, whilst the other two are typical HMG proteins, having high contents of acidic and basic amino acids. These proteins are present in nuclei but may not necessarily be DNA-associated. Abbreviations: HMG, high mobility group; HPLC, high pressure liquid chromatography.
330 Methods
Preparation o f calf thymus nuclei Fresh calf thymus tissue was blended in 2.2 M sucrose/3 mM MgC12. Phenylmethyl sulphonyl fluoride (0.5 mM) was added and the mixture filtered through gauze. The filtrate was layered over 2.2 M sucrose/3 mM MgC12/0.5 mM phenylmethysulphonylfluoride and centrifuged at 30 000 X g for 1 h. The nuclear pellet was resuspended in the 2.2 M sucrose solution and sedimented again. Examination of the final nuclear pellet by electron microscopy showed that the nuclei were very clean with little cytoplasmic contamination. The HMG protein fraction was extracted from these nuclei with 5% perchloric acid and precipitated with acetone [2]. Isolation o f HMG pro teins 18, 19A and 19B from perchloric acid ex tracts o f calf thymus tissue Thymus (1 kg) was extracted three times with 5% perchloric acid. The total extract (2.8 1) was filtered through Whatman glass fiber paper and made 0.3 M HC1. Acetone (3.5 vols.) was added to precipitate histone protein H1 which was removed b y centrifugation. The proteins in the supernatant (HMG proteins, some H1 and ubiquitin) were precipitated by the addition of a further 2.5 vols. of acetone. The proteins from two such preparations were combined and fractionated b y ethanol-HC1 precipitation as described before for fractionating HMG proteins [3] (see Scheme I). Briefly, the proteins were redissolved in 0.1 N HC1 at a concentration of 50 mg/ml and 5 vols. of ethanol-HC1 (99 : 1 v/v, ethanol-conc. HC1) were added. Any precipitate was removed b y centrifugation (2000 X g for 10 rain). A further 7.5 vols. of ethanol-HC1 were added to the supernatant to precipitate some of the HMG 1 and 2, and all of the HMG 14, 17 and 18. After sedimenting this precipitate (2000 × g for 10 rain) to give fraction 1, the supernatant was filtered and the volume of the filtrate measured. Acetone (6 vols. of the measured volume) was added to precipitate the remaining HMG 1 and 2, ub.iquitin and HMG proteins 19A and 19B (fraction 2). HMG 18 was isolated from the fraction 1 proteins as follows. Fraction 1 protein (500 rag) was redissolved in 10 mM sodium borate buffer, pH 8.8, titrated to pH 9, and NaC1 was added to make the total Na ÷ concentration 0.1 M. The solution was clarified b y centrifugation and loaded onto a 2.5 × 58 cm CMSephadex C-25 column equilibrated with 10 mM borate buffer, pH 8 . 8 . 1 5 0 ml of 0.1 M NaC1 in borate buffer was pumped through the column (1 ml/min) followed b y a 4 1 salt gradient (0.1--1.2 M NaC1). Protein HMG 18, eluting at 0.65 M NaC1 (Fig. 2), was recovered b y dialysing the pooled fractions against 50% ethanol containing 10 mM HC1 to concentrate the protein which was then precipitated b y making the solution 0.1 M HC1 and adding 6 vols. of acetone (yield 2--3 rag). Proteins HMG 19A and 19B were isolated from fraction 2 proteins as follows. Fraction 2 proteins (400 mg) were fractionated on a CM-Sephadex column as described above except the protein was loaded in 0.2 M NaC1 and the salt gradient (4 l) was 0.2--1.2 M NaC1. Proteins HMG 19A and 19B, eluting at a b o u t 0.35 M NaC1 (Fig. 4) were collected as described above. To
331 obtain a better separation of 19A and 19B the proteins were combined and redissolved in 10 mM sodium acetate, pH 5.5. The solution was titrated to pH, 5.5 and loaded onto a CM-cellulose (Whatman) column (1.5 X 8 cm). A 200 ml salt gradient (0.1--1.0 M NaC1) was pumped through the column (1 ml/min). Proteins HMG 19A and 19B (Fig. 6) were collected as described above. (Yields: 3.5 mg and 1.0 mg for 19A and 19B, respectively). N-terminal amino acid sequence determination Protein HMG 18 (2.5 mg) and protein HMG 19B (3.0 mg) were analysed on a Beckman protein sequenator using a 0.1 M quadrol buffer programmer essentially as described by Hunkapiller and Hood [4]. Released thiazolinone derivatives were converted to the phenylthiohydantoin amino acids by incubation in 20% (v/v) trifluoroacetic acid at 80°C for 10 min. Phenylthiohydantoin amino acids were identified both by high pressure liquid chromatography (HPLC) and by back hydrolysis to the free amino acid. HPLC was carried out on a DuPont 830 liquid chromatogram using a Partisil PX5 5/25 ODS column (Whatman) at 65°C. Phenylthiohydantoin amino acids were eluted using a linear gradient from 15 to 48% (v/v) acetonitrile/0.01 M sodium acetate (pH 4.5) for 6.5 min, then holding at 48% for a further 10 min. Back hydrolysis was carried out by heating in hydriodic acid (64%) at l l 0 ° C for 24 h. Amino acids thus produced were analysed on a Rank Hilger Chromospek amino acid ana!yser. Analytical techniques Amino acid analyses were carried out as described previously [1 ]. Proteins were analysed electrophoretically on 15 or 20% polyacrylamide-sodium dodecyl sulphate (SDS) gels as described previously [5]. Results and Discussion
Fig. 1 shows a sodium dodecyl sulphate (SDS)-gel electrophoretic analysis of the perchloric acid-extracted proteins (minus most of the histone protein HI) from purified calf thymus nuclei. In addition to the four main HMG proteins, HMG 1, 2, 14 and 17, two minor bands, HMG 18 and 19, are seen running ahead of HMG 17. As we show below, HMG 19 is, in fact, composed of two components, 19A and 19B. HMG 19B is present in smaller quantities and runs between HMG 18 and HMG 19A and is not always completely resolved on these gels. (A fourth low molecular weight protein, ubiquitin, migrates even faster than these three [ 1 ], but is not always present in perchloric acid extracts of highly purified nuclei.) When this perchloric acid protein extract from nuclei is fractionated on a CM-Sephadex column at pH 8.8, as for the fractionation of HMG proteins [2], HMG 18 is the last protein to be eluted at about 0.65 M NaC1, and HMG 19A and 19B elute almost together with HMG 17 at about 0.35 M HC1 (data not shown). In order to obtain these three minor proteins in a pure form in sufficient quantities for further characterisation it was necessary to use perchloric extracts of the total tissue (rather than purified nuclei) since larger quantities of material can be obtained in this way and also proteolytic degradation is minimised [2]. The fractionation scheme of Scheme I was
332 THYMUS PCA e x t r a c t i o n and fractional a c e t o n e precipitation
HMG PROTEINS AND UBIQUITIN (+H1) Redissolve A d d 5 vols. ethanol-HCl. Centrifuge A d d 7.5 vols. ethanol-HC1 to s u p e m a t a n t Centrifuge
PRECIPITATE ( F R A C T I O N I)
S U P E R N A T A N T ( F R A C T I O N 2) CM-Sephadex
CM-Sephadex
HMG1
HMG2
HMG14
HMG17
H1 H M G 1 8
Ubiquitin HMG1 HMG2 HMG 19A, B . CM-cellulose HMG 19A
HMG 19B
S c h e m e I. F r a c t i o n a t i o n o f Perchloric acid (PCA)-extracted proteins.
carried out on this material, the ethanol-HC1 precipitation being required to separate HMG 17 from HMG 19A and 19B. The ethanol-HCl-insoluble fraction 1, containing HMG 18, was fractionated by CM-Sephadex chromatography (Fig. 2). HMG 18 elutes at about 0.65 M NaC1. The SDS gel electrophoretic analysis (Fig. 3) shows that the protein is slightly contaminated with a faster moving component. The ethanol-HCl-soluble fraction 2, containing proteins HMG 19A and 19B, was fractionated in the same way (Fig. 4) by CM-Sephadex chromatography. The major c o m p o n e n t eluting at 0.35 M NaC1 is protein HMG 19A (Fig. 5c) but is contaminated slightly with a minor component, HMG 19B, eluting on the trailing edge of the peak (Fig. 5b). These two proteins could be better separated and obtained in a purer form by combining the A and B fractions from this column and subjecting then to chromatography on CM-ceUulose at pH 5.5 (Fig. 6). Fig. 7 shows the SDS-polyacrylamide gel electrophoresis analysis of fractions A and B from this column. The third peak is a heterogenous mixture of contaminants which have separated away on this column (not shown). The amino acid analyses of the three proteins HMG 18, 19A and 19B are given in Table I. Protein HMG 18 is highly basic and has high contents of lysine and alanine and thus resembles histone H1. Proteins HMG 19A and HMG 19B resemble one another in having similar amino acid compositions. The major differences between the two proteins are in the aspartic acid and arginine contents. These differences have been consistently observed in three preparations of these two proteins. Proteins HMG 19A and B are characterised by their high contents of basic and acidic amino acids and in this respect t h e y resemble the previously isolated HMG proteins. The N-terminal amino acid sequences of HMG 18 and HMG 19B are shown in Scheme II. HMG 19Ahas a blocked amino acid terminal and was therefore not amenable to sequence analysis. A single
50
100
150
FractionNo.
Fig. i (Top left). sodium dodecyl sulpbate (SDS)-polyacwlemide extiacted proteins from purified calf thymes rulei.
gel electrophoreeie of perc~oric acid-
Fig. 2 (Bottom left). CMSephadex chromatograpkw of fraction 1 proteins. 20 ml fractions were collected. HMG 18 (hatched peak) was recovered by acetone precipitation. Fig. 8. SDS-polyaerylamide gel electrophoresis of (a) total thymus perchloric acid-extract, (b) protein HMG 18 isoleted by CM-Sephadex chromatography (Fig. 2).
sequence was obtained for both proteins HMG 18 and 19B, confirming the purity of these proteins. Amino acid analysis suggests that HMG 18 is approx. 85 residues long, or a multiple thereof. However, the electrophoretic mobility of HMG 18 on SDS gel
334
HMG 18
10 Lys-Val- His- Gly-Ser- Leu-Ala-Arg-Ala-Gly20 -Lys-Val- Arg-Gly-Gln-Thr-Pro-Lys-Val-Ala-Lys-Gln-Glu-Lys-Lys-
10 HMG 19B Thr-Arg-Gly- Asn-Gln-Arg-Glu-Leu-Ala- Gly20 -Gln-Lys-Asn-Met-Lys-Lys-Gln-Lys-Asp-Pro-Val- Lys-GlyS c h e m e II. N - t e r m i n a l a m i n o a c i d s e q u e n c e s o f p r o t e i n s H M G l S a n d H M G 1 9 B .
electrophoresis suggests that the protein cannot be any longer than approx. 85 residues. Analysis of the N-terminal amino acid sequence and the amino acid composition of the remaining portion of the molecule shows that there can be no extensive acidic regions in the molecule. In this respect HMG 18 resembles the histones, and differs from the four HMG proteins 1, 2, 14 and 17 which all have regions of negative charge. 36% of the first 25 residues are basic residues, but since this is the same figure for the overall basic composition of the molecule there is at present no evidence for the clustering of the basic residues in the molecule. The amino acid analysis of HMG 19A again supports a size of approx. 85 residues. The exact size of HMG 19B is difficult to determine from the amino acid analysis, but since the protein runs in the same region asHMG 18 and 19A on SDS gel electrophoresis a size in the region of 100 residues can be assumed. The N-terminal amino acid sequence of HMG 19B shows no evidence of clustering of any particular groups of amino acids. Comparison of the N-terminal
3.0
HMG1
2.0
"1.0 .0.8 HMG19
! 20
40
60
80
100
.o. 6
120
Fraction No. F i g . 4. CM-Sephadex c h r o m a t o g r a p h y o f fraction 2 p r o t e i n s . 10 m l f r a c t i o n s w e r e c o l l e c t e d . The t w o HMG 19 (A a n d B) f r a c t i o n s w e r e recovered b y a c e t o n e precipitation.
335 i i ¸¸ !i !ili!i
Fig. 5. SDS-polyacrylamide gel electrophoresis of (a) total t h y m u s perchloric acid-extract, (b) fraction B from the CM-Sephadex chromatography (Fig. 4), (c) fraction A from the CM-Sephadex chromatography (Fig. 4).
1.0.
CC B~
1.0
II
0.4
A
~ ,,¢
0.8 0.6
0.5
T
0.2 r/ i
20
40 Fraction1 No.
|
i
60
80
F i g . 6. C M - c e l l u l o s e c h r o m a t o g r a p h y of combined fractions A and B from the preceding CM-Sephadex chromatography (Fig. 4). 2 ml fractions were collected. HMG 19A and 19B (hatched peaks) were recovered by acetone precipitation.
336
Fig. 7. S D S - p o l y a c r y l a m i d e gel e l e c t r o p h o r e s i s of (a) p r o t e i n H M G 1 9 A , (b) p r o t e i n H M G 1 9 B o b t a i n e d b y t h e CM-ce]lulose c h r o m a t o g r a p h y (Fig. 6). TABLE I AMINO ACID COMPOSITIONS (mol%) OF P R O T E I N S HMG 18, 19A AND 19B
Asp Thr Ser Glu Pro GIy Ala Cys Vai Met ~le Leu Tyr Phe His Lys Arg
HMG 18
HMG 19A
H M G 19B
5.0 4.8 3.8 5.2 6.1 8.3 15.6 -7.9 0.9 -1.2 1.3 2.6 1.2 27.6 8.5
4.3 2.2 3.8 18.0 5.4 10.5 10.6 -1,1 1.3 1,3 5,1 0,8 2.0 0.4 29.8 3.4
9.4 2.5 6.0 18.1 4.2 7.2 10.2 -2.0 1.8 1.5 3.9 0.9 1.7 0.6 22.4 7.8
337 10 20 HMG 18 -Arg-Ala-Gly-Lys-Val-Arg-Gly-Gln-Thr-Pro-Lys-Val- Ala-Lys-Gln-Glu-Lys40 50 -Arg-Ala-Glu-Lys-Ser- Arg-Gly-Gly-Ser- Set- Arg-Gln-Se r- Ile- Gln LysH5
HMG 19B -Lys-Lys HMG 17
•
20 Gln-Lys-.Asp_-Pro-Val-Lys-Gly
50 -Lys-Lys-Gl:~-Glu-Lys-Val- P r o .
Lys-Gly-
S c h e m e Ill.Comparison of sequences from H M G IS, H 5 [12], H M G 1 9 B and H M G amino acids can be derived from o n e another by singlebase substitutions).
17 [9]. (Underlined
sequences of HMG 18 and 19B with the k n o w n sequences of the histones and HMG proteins 1, 2, 14 and 17 shows no extensive sequence homologies to suggest that they are degradation products of histones or other HMG proteins. The fact that these proteins are present in perchloric acid extracts of tissue obtained from calves immediately after slaughter and transported to the laboratory in liquid nitrogen also suggests that they are n o t degradation products [2]. The nuclear localization o f these proteins is not yet known, b u t it seems highly likely that these proteins are b o u n d to nucleic acid since they have high contents of basic amino acids like the DNA-bound histones and HMG proteins 1, 2, 14 and 17. In support of this view, we would like to point o u t that there is a section in the N-terminal sequence of HMG 18 that has some resemblance to a region in histone H5 (Scheme III); similarly, HMG 19B has a section of sequence resembling a region in the HMG 17 molecule (Scheme III). Because HMG 18 and 19 are present in nuclei in such small quantities (01.% b y weight o f the DNA) they would be difficult to detect in nucleos0mes, especially since they migrate in approximately the histone H4 position on SDS gels. However, such bands can sometimes be seen in nucleosomes (Fig. 5 of Ref. 6) b u t it is n o t possible to say with certainty whether or n o t these are proteins HMG 18 and 19. Also we have found that (like HMG 1 and 2) some o f the proteins can be washed o u t of the nucleus with low ionic strength buffers and hence t h e y may n o t be b o u n d to DNA b u t rather to RNA. Examination of calf t h y m u s and liver cytoplasmic fractions (crude 'mitochondrial', 'microsomal' and 'ribosomal' fractions) reveal the presence of a band running close to (but n o t exactly coincidental with) the HMG 18 band in all three fractions (not shown). This might mean that HMG 18 is, for example, a ribosomal protein or other R N A - b o u n d protein and is only found in isolated nuclei as a result of the transient presence of newly synthesised ribosomes or ribonucleoprotein particles in the nucleus. Contamination of the nuclei with ribosomes attached to the outer nuclear membrane is also possible. Conversely, since in the electron micrographs the nuclear membrane of a not inconsiderable portion o f nuclei appears to be partly ruptured, the presence of HMG 18 in the cytoplasm m a y be due to release from the nucleus of structures containing HMG 18. Thus, there is still some d o u b t as to the true subcellular location of
338 HMG 18, and further work is required to determine this. However, it should be pointed out that HMG 18 does not correspond to any of the so far purified ribosomal proteins (see Ref. 7 for references therein) or protein bound to heterogenous nuclear RNA [8]. Proteins HMG 19A and 19B appear to be located predominantly in the nucleus but again it should be stressed that further studies are required to establish conclusively where the proteins are bound and until then these proteins can only be tentatively classed with the previously isolated high mobility group proteins (HMG 1, 2, 14 and 17) on the basis of their similar amino acid composition and solubility properties. From our previous sequence studies of the HMG proteins [9,10] and from our recent demonstration that HMG 14 and HMG 17 (and maybe HMG 1 and 2) are bound to the linker DNA immediately contiguous to the core particle of the nucleosome [11] it would appear that these HMG proteins are closely related to histone HI and H5. The results of this paper, showing that there are some similarities between the N-terminal sequence of HMG 18 and histone H5 and between HMG 19B and HMG 17, is consistent with the view that the seven HMG proteins isolated to date (HMG 1, 2, 14, 17, 18, 19A, 19B) and the two lysine-rich histones HI and H5 form a family of related proteins, which, by binding to specific regions of the genome, after the structure of the chromatin thereby controlling genetic expression [ 11 ]. Acknowledgements We would like to thank Carol Wright for excellent technical assistance, Dr. J.R.B. Hastings for amino acid analyses, and M. Birbeck and D. Robertson for the electron microscopy. This work was supported by grants from the Medical Research Council. References 1 2 3 4 5 6 7 8 9 10 11 12
Walker, J.M., Goodwin, G.H. and Johns, E.W. (1978) FEBS Lett. 9 0 , 2 3 7 - - 3 3 0 Goodwin, G.H., Walker, J.M. and Johns, E.W. (1978( Biochim. Biophys. Acta 5 1 9 , 2 3 3 - - 2 4 2 Goodwin, G.H., Rabbani, A., Nicolas, R.H. and Johns, E.W. (1977) FEBS Lett. 80, 413--416 Hunkapillar, M. an d Hood, L. (1978) Biochemistry 17, 2124--2133 Goodwin, G.H. and Johns, E.W. (1978) Biochim. Biophys. Acta 5 1 9 , 2 7 9 - - 2 8 4 Mathew, C.G.P., Goodwin, G.H. and Johns, E.W. (1979) Nucleic Acid Res. 6, 167--179 McConkey, E.H., Bielka, H., Gordon, J., Lastick, S.M., Lin, A., Ogata, K.. Reboud, J.-P., Traugh, J.A., Traut, R.R., Warner, J.R., Welfle, H. and Wool, I.G. (1979) Mol. Gen. Genet. 169, 1--6 Karn, J., Vidali, G., Boffa, L.C. and Allfrey, V.G. (1977) J. Biol. Chem. 252, 7307--7321 Walker, J.M., Hastings, J.R.B. and Johns. E.W. (1977) Eur. J. Biochem. 76, 461--467 Walker, J.M., Goodwin, G.H. and Johns, E.W. (1979) FEBS Lett. 100, 394--398 Goodwin, G.H., Mathew, C.G.P., Wright, C.A., Venkov, C.D. and Johns, E.W. (1979) Nucleic Acid Res., in t h e p r e s s Sautiere. P., Briand, G., Kmiecik, D., Loy, O. and Bisenke, G. (1976) FEBS Lett. 63, 164--169