Scand. J. clin. Lab. Invest. 37, 749-755, 1977.
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
Isolation and partial characterization of a glycineand serine-rich polypeptide from human serum high-density lipoproteins (HDL) S . - 0 . O L O FSSO N , G . F A G E R & A . G U S T A F S O N Departments of Medical Biochemistry and Medicine I, Sahlgren’s Hospital, University of Goteborg, Goteborg, Sweden
Olofsson, S.-O., Fager, G. & Gustafson, A. Isolation and partial characterization of a glycine- and serine-rich polypeptide from human serum high-density lipoprotein (HDL). Scand. J. clin. Lab. Invest. 37, 749-755, 1977. Lately, several minor polypeptides characterized by a high content of glycine and serine have been described in human serum lipoproteins. By column chromatography we have isolated a glycine- and serine-rich polypeptide from totally delipidized high-density lipoproteins, apo-HDL. The partial characterization of this polypeptide by total amino acid composition and by immunology utilizing monospecific antisera to polypeptides of human serum lipoproteins, would indicate a unique polypeptide. This low molecular weight (4900 Daltons) glycine- and serine-rich polypeptide was characterized by a mobility on polyacrylamide gel electrophoresis, similar to that of polypeptide A-I, a blocked NHz-terminal amino acid and a terminal COOH-fragment consisting of R-SerAla-Gly-Gly. There are similarities between this polypeptide and the protein moiety of a cholesterol ester-rich lipoprotein fraction, HDLSUP,obtained by in vitro incubation of HDL subfractions and described in earlier publications by our group. The identity between these twopolypeptides, however, cannot beconfirmed by the present study. Key-words: glycine; lipoprotein; high density; polypeptide; serine Sven-Olof Olofsson, M.D., Department of Medical Biochemistry, University of Goteborg, S-400 33 Goteborg, Sweden
In addition to five well-characterized apolipoproteins, and their constitutive polypeptides,*
* Apolipoprotein A, apoA (composed of the nonidentical polypeptides A-I and A-11) comprises the bulk of the protein moiety of high-density lipoproteins (HDL). ApoB (polypeptide composition unknown) dominates in very-low-density lipoproteins (VLDL) and low-density lipoproteins (LDL) and apoC (with nonidentical polypeptides c-1,C-11 and c-111) is common to VLDL and HDL. More recently, apoD [lo, 12,161 also referred to as A-I11 [8], and apoE (‘arginine-rich polypeptide’, [3, 21, 23, 261) have been included as minor polypeptide components.
data in the literature [15, 18, 22, 231 suggest that polypeptide chains, characterized by a relatively high content of serine and glycine are present - . within the lipoprotein system. We have presented evidence for a glycine- and serine rich polypeptide as the apolipoprotein of a cholesterol ester-rich lipoprotein released from highdensity lipoproteins (HDL) [15, 181 by in vitro incubation. In order to study the possible role of the glycine- and serine-rich polypeptides in serum lipoprotein structure and metabolism, we 749
750
S . - 0 . Olofsson, G. Fager & A . Gustafson
have isolated and characterized one of them. The present publication describes the preparation of this polypeptide from HDL, its further characterization by amino acid composition, immunology, COOH- and NH2terminal amino acids and molecular weight determination.
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
MATERIAL AND METHODS Blood was obtained from healthy normolipidaemic male volunteers (age 25-40 years), who had fasted overnight. Blood was allowed to clot for 60 min at +4"C and serum was recovered by low-speed centrifugation. Serum was immediately placed at an ambient temperature of tO"C and kept between 0 and +4"C throughout the experiments [17]. Isolation and purification of H D L . HDL was isolated by a two-step preparative ultracentrifugation procedure at solvent densities (D) 1.19 g/ml and 1.063 g/ml [17]. The further purification of HDL from contaminating lipoprotein B (LP-B) was carried out by column (K 9/30) chromatography on Concanavalin ASepharose 4B (Pharmacia Fine Chemicals, Sweden) [l I] using 0.2 mol/l sodium acetate, 0.1 mol/l sodium chloride and 0,001 mol/l of manganese chloride, magnesium chloride and calcium chloride (pH 6.8).Before this chromatography, the buffer composition of HDL was changed to the elution buffer by gel filtration on a column (K 15/30) of Sephadex G25 (Pharmacia Fine Chemicals, Sweden). The purified HDL was finally desalted on Sephadex G25 employing 0.05 mol/l phosphate buffer. The HDL preparation obtained showed a typical pattern on basic polyacrylamide gel electrophoresis (Fig. 2, Pattern A) and did not react against antisera to albumin or LP-B. All buffers used for the isolation and purification procedures were made up fresh in distilled, sterilized water containing 0.2% sodium azide. All reagents used were of analytical grade, except guanidine-HC1 which was purified on charcoal according to Kelly & Alaupovic [7]. Delipidization of HDL. Purified and lyophylized HDL preparations were totally delipidized by four extractions with a mixture of chloroform
and methanol (2:1, v/v) with approx. 12 ml of solvent per 10 mg of apolipoproteins. The delipidization procedure was completed with three washings with diethylether, after which apo-HDL was recovered by low-speed centrifugation and dried in uacuo. Fractionation of apo-HDL. Dried apo-HDL was immediately dissolved in 0.01 mol/l Tris-HCI buffer (pH 7.2) containing 0.05% of sodium EDTA and 6 mol/l of purified guanidine-HCI [7] and chromatographed on a column (K 26/100) of Sepharose 6B (Pharmacia Fine Chemicals, Sweden) equilibrated with the same buffer [21]. The total volume of the column was 438 ml and the void volume 134 ml. The flow rate was 6-8 ml/h and the effluent collected in 3.5 ml fractions. Fractions were read at 280 nm in a spectrophotometer (Zeiss). Before further characterization, the protein moieties were submitted to extensive dialysis against water, against 0.1 mol/l KCI and again water (with four daily changes with 100 times the volume of the dialysate) and finally lyophilized. Characterization and quantification of the protein moieties. Lipoproteins and apolipoprotein fractions were studied by double diffusion [20] in 1 % agar (Behringwerke, Marburg an der Lahn, W. Germany) employing barbital buffer (pH 8.6), ionic strength 0.05. Intact HDL was studied with rabbit sera containing antibodies to human albumin, a-llipoproteins and /3-lipoproteins (Behringwerke) and apo-HDL fractions also by monospecific antisera to apoA-I, apoA-11, apoC-I, apoC-1112, apoD and apoE. The latter antisera were made available by the courtesy of Dr P. Alaupovic, Oklahoma City, Oklahoma, USA. Apo-HDL fractions were dissolved in 0.1 % sodium dodesyl sulphate (SDS), 0.1 mol/l ammonium carbonate or 3% (by weight) Triton X-405 (KEBO AB, Stockholm, Sweden) in barbital buffer (pH 8.6). The concentration of fraction V applied to double diffusion varied in different experiments from 0.5 to 5 mglml. Basic polyacrylamide gel electrophoresis (PAGE) was performed in a Canalco Model 1200 unit, utilizing 10% acrylamide monomer (pH 8.6) and 8 mol/l urea [4, 191. Gels were fixed in 10% trichloracetic acid and stained with Coomassie Brilliant Blue [2]. The
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
GIycine- and serine-rich polypeptide of H D L
concentration of samples applied to PAGE was 0.25-2 mg/ml. Analyses of total amino acids were performed on a JLC-5 AH Automatic Amino Acid Analyser equipped with an integrator (Jeol Ltd, Tokyo, Japan). Hydolysis of protein moieties was carried out in 6 mol/l HC1 in uacuo for 22 h [13]. Cystine and cysteine were determined as cysteic acid following performic acid oxidation [14]. Tryptophan was determined after hydrolysis with p-toluenesulphonic acid
751
VI-tVO
4
PI. NH2-terminal amino acid analyses were performed by a method adopted from Wood & Wang I271 and by a method modified by Shelburne & Quarfordt [21]. For comparison bovine serum albumin, porcine insulin and human apoA-I were used. COOH-terminal amino acid analysis was performed by incubation with carboxypeptidase A at 37°C [I]. Protein material was dissolved in ethylmorpholine buffer containing 0.1% SDS utilizing an enzyme substrate ratio of 1:40 and samples drawn at 5, 10, 15, 30, 60 and 180 min. The protein content was determined by a micro-Kjeldahl method [25] assuming 16% nitrogen. Molecular weight estimation. Gel filtration was carried out on Sephadex G75 (Pharmacia Fine Chemicals, Sweden) utilizing a column K26/100. The column was equilibrated and eluated with 0.01 mol/l Tris-HC1 buffer (pH 7.2) containing 0.05 % of sodium EDTA and 6 mol/l guanidineHCl [5]. Calibration was performed with purified samples of human apoA-I (rnol. wt. 28,000), human apoA-I1 (rnol. wt. 17,400) apoA-I1 monomer obtained by reduction and carboxy methylation of apoA-I1 [6] (rnol. wt. 8700), cytochrome c (rnol. wt. 13,400), ACTH (mol. wt. 4500) and glucagon (mol. wt. 3520). The void volume (V,) was determined by chromatography of Blue Dextran 2000. VI (the internal volume) Vo was estimated by chromatography of 14Calanine. The retention coefficient ( K d ) [5] was calculated for standards and samples.
+
RESULTS Isolation of glycine- and serine-rich polypeptide (fraction V ) from apo-HDL. Gel chromatography of apo-HDL gave five fractions (Fig. 1).
0.25
c
P
100
200
300
400
500
Elution volume ( m l )
FIG. I . Chromatographic separation of the lipid-free protein moiety of human serum high-density lipoproteins (apo-HDL) on Sepharose 6B, utilizing 0.01 mol/l Tris-HCI buffer (pH 7.2) containing 6 mol/l quanidine-HCI and 0.05% Na-EDTA (7). Vo = void volume and VI = internal volume.
Fraction I (eluted closely after the void volume) contained apo-A-I as judged from PAGE and immunological properties. Fraction I1 contained mainly apoE (the arginine-rich polypeptide) contaminated by apoA-I. Fraction 111 showed apoA-I and apoA-I1 and Fraction IV the C-I1 and C-I11 polypeptides. Fraction V, finally, was eluted with a buffer volume of 424 ml slightly ahead of V I + V o (438 ml), corresponding to a K , of 0.95. After dialysis, the protein yield of fraction V corresponded to 0.25 mg/100 ml of serum by the present technique. Partial characterization of fraction V . The characterization by double immune diffusion of fraction V dissolved in SDS or in 3% Triton against specific antisera to apoA-I, apoA-11, apoC-I, apoC-111, apoD, apoB and apoE failed to reveal any irnmunoprecipitation lines. Although no unequivocal interpretation can be made from a negative result on immuno-doublediffusion, these data indicate the absence of those polypeptides as contaminations. On basic polyacrylamide gel electrophoresis (Fig. 2) fraction V, both when dissolved in 8 mol/l urea and in 0.1 % SDS revealed only one single band taking protein stain in the region of
752
S.-0. Olofsson, G . Fager & A . Gustafson
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
A
B
C
FIG. 2. Patterns o n basic polyacrylamide gel electrophoresis (PAGE). A = protein moiety of HDL (apeHDL). B = fraction V dissolved in 0.1Y SDS. and c = fraction v dissolved in urea containiig stacking gel.
apoA-I. This finding suggests fraction V to contain a single polypeptide. Total amino acid composition of fraction V, was characterized by a high content of serine and glycine, but also by lack of tryptophan (Table 1). It appears as if methionine also is
missing, since after HC1-hydrolysis and performic acid oxidation neither methionine nor methionine sulphoxide or sulphone was detectable. The amino acid composition of fraction V differed from that of polypeptide A-1 by its content of isoleucine, cystine and/or cysteine. Furthermore, the amino acid composition differed from that of apoE (arginine-rich polypeptide) by its low arginine content. The described procedures [21, 271 for the analyses of NH2-terminal amino acid failed to reveal any DNS-amino acids indicating a blocked NH2-terminalamino acid of the protein moiety of fraction V. In agreement with earlier data, the present technique, however, showed the expected NH2-terminal amino acids for apoA-I, insulin and albumin, respectively. lncubation with carboxy-peptidase A (Fig. 3) yielded only glycine during the first 10 min of incubation indicating glycine to be COOHterminal amino acid. The release of glycine reached maximum after 15 min, to a molar ratio of 2:1 amino acid/protein, assuming mol. wt. 4900. This suggests the last two amino acids to be glycine. Our data also showed that the next two amino acids to be released were alanine and serine. This would give the tentative COOH-terminal fragment of R-Ser-Ala-Gly-Gly.
TABLE I. Amino acid composition (moles per 1000 mol of amino acid residue) of fraction V (n = 5). Data are given as mean+ SEM. For comparison are given data on protein moiety of cholesterol ester-rich lipoprotein fraction, apo-HDLSUP,obtained after in virro incubation [18] and on glycine- and serine-rich polypeptide of Shore & Shore (22).
Amino acid
Fraction V Mean+ SEM
apo-HDLSUP ~ 7 1
Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Isoleucine Leucine Thyrosine Phenylalanine Lysine Histidine Arginine Tryptophan Methionine Cysteine/cystine
82+4 44+ 4 188+22 151+ 13 42+ 3 176+22 82+ 5 38+ 4 24+ 3 58+ 14 19+ 3 22+ 3 44+ 7 23+4 25+7 0 0 14
87 50 140 156 O* 165 59 50 30 90 25 28 45 19 41 n.d. trace 9
Polypeptide described by Shore & Shore [221 72 40 204 160 20 194 71 30 21 35 7 15 76 30 14 0 7 7
* Elution of hydrolysed amino acids not monitored at 440 nm in this series.
753
Glycine- and serine-rich polypeptide of HDL
LC
;2.0
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
E
0
Time iminl
FIG.3. COOH-terminal amino acid analysis. Action of carboxypeptidase A at 37°C on fraction V obtained, from human serum apo-HDL. Release of glycine ( C ) , alanine ( 0 )and serine ( x ) are given in relation to time of incubation.
4L 3
0.5
,.I
Relative r e t e r t i o n ( K d )
The homogeneity of fraction V was emphasized by its chromatographic behaviour on Sephadex G75 (Fig. 4).The fraction was eluted as a narrow, homogenous peak with a Kd of 0.51. The PAGE pattern of this peak was the same as that of fraction v. The Kd of fraction V on Sephadex GI5 correponded to a monomeric molecular weight of 4900 daltons (Fig. 5 ) . When estimated from data on relative amino amino acid composition (Table I), the molecular weight of 4920 daltons was obtained for the first multiple corresponding to forty-nine amino acid residues.
t
O O‘O
t
I
100
li
I \ I
200 300 400 Elution volume(rn1)
500
FIG. 4. Chromatographic behavior of fraction V, obtained from human serum apo-HDL by gel filtration on Sepharose 6B, when re-chromatographed on Sephadex G75 in 0.01 mol/l Tris-HC1 buffer containing 6 mol/l guanidine-HCI and 0.05% Na-EDTA. Inserted is pattern on basic PAGE of this fraction.
FIG.5 . Molecular size estimation by relative retention on Sephadex G75 (column K 26/100) utilizing 0.01 mol/l Tris-HC1 buffer containing 6 mol/l guanidine-HCI and 0.05% Na-EDTA. ApoA-I (28,000), apoA-I1 (1 7,400), Cytochrome c (13,400), apoA-I1 monomer (8700), ACTH (4540) and glucagon (3520) were used as references. Data plotted in semilogarithmic fashion. Relative retention of fraction V indicated by broken line.
DISCUSSION The present publication deals with the preparation of a glycine- and serine-rich polypeptide in high-density lipoproteins (HDL). Our data indicate that the present procedure gives a homogeneous preparation without contamination of any of the known apolipoproteins, apoA-I, apoA-11, apoC-I, apoC-11, apoC-111, apoB, apoD or apoE. The amino acid composition is characterized by a high content of glycine and serine. These amino acids, however, are common contaminants in protein preparations. In the present study, the content of glycine and serine was closely similar in repeated preparations (n = 5 ) . Furthermore, the good agreement between molecular weight estimation from amino acid composition and by gel filtration, would indicate contamination of free amino acids to be negligible. As to total amino acid composition, the lack of tryptophan and presumably of methionine is of interest. This glycine- and serine-rich polypeptide isolated from apo-HDL is further characterized by a blocked NH2-terminal amino acid and by glycine as COOH-terminal amino acid. Furthermore, gel filtration and calculations from total amino acid composition indicates a low mole-
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
754
S.-0. Olofsson, G . Fager & A . Gustafson
cular weight, 4900 daltons. In the present study, the yield of this glycine- and serine-rich polypeptide is low (0.25 mg/100 ml of serum). The isolated polypeptide reveals a low water solubility, but can be solubilized in sodium dodecyl sulphate (SDS), 8 mol/l urea, 3 % Triton X-405 and in 6 mol/l guanidine-HC1. When solubilized in these agents, the polypeptide fails to react with antisera to all known apolipoproteins. Recent data [18, 22, 231 would suggest the presence of several polypeptides rich in glycine and serine within human serum lipoprotein spectrum. A closer comparison between those is not possible because only incomplete data is available. We were able in a recent report [18] to isolate and partially characterize the apolipoprotein of a cholesterol ester-rich lipoprotein fraction obtained by in vitro incubation of HDL subfractions. This apolipoprotein was also characterized by a high content of glycine and serine (Table I). Furthermore, the finding of a blocked NH2-terminal amino acid was in common with the present polypeptide, as was its behaviour on PAGE. The comparison of total amino acid composition (Table I) would indicate certain differences as to relative amounts of amino acids. The apolipoprotein of the cholesterol ester-rich lipoprotein fraction, apoHDLSup,revealed a higher content of arginine and was not completely characterized as to proline and tryptophan content. It is possible that these differences were due to impurities in apo-HDL"P, rather than in the present preparation, fraction V. However, a final proof for identity between the protein moieties of HDLsup and fraction V, has to wait for immunology utilizing monospecific antibodies prepared against their protein moieties. The constantly low yield of these small polypeptides, has in our hands so far prevented the preparation of antibodies. Shore & Shore [22, 231 have described the presence. of glycine- and serine-rich apolipoprotein within human serum [22] as well as in very-low-density lipoproteins (VLDL) [23]. Those glycine- and serine-rich polypeptides were either not visualized on PAGE [22] or migrated as a fast moving band at the position of C-I11 [23]. The difference in relative amino acid composition other than glycine and serine seems furthermore to exclude identity to the present polypeptide.
Our earlier studies on the cholesterol esterrich lipoprotein fraction in HDL, HDLSuP, indicated the possibility of a physiological function of this lipoprotein fraction, as a mediator in cholesterol ester transport among lipoproteins. However, apo-HDLgupappears to be different from apoE (the arginine-rich polypeptide) [24] as well as from the protein moiety described by Zilversmit [28], both linked specifically to cholesterol ester transport.
ACKNOWLEDGMENTS The technical assistance of Miss Anita Jacobsson and Mrs Marie-Louise MBnsson and the secretarial aid of Miss Annika Hofde is gratefully acknowledged. The amino acid analyses were performed by Ulf Svanberg, B.Sc., and Mrs Monica Olsson. The authors also wish to express their appreciation to Drs P. Alaupovic and W. J. McConathy (Oklahoma City, Oklahoma, USA) for generous supply of monospecific antisera. This study was supported by grants from the Swedish Medical Research Council (19X-2100), the Swedish Oleo-Margarine Foundation for Nutritional Research, the Swedish National Association against Heart and Lung Disease, 'Goteborgs Lakaresallskap', and the Medical Faculty of Goteborg University.
REFERENCES 1 Ambler, R.P. Enzymatic hydrolysis with carboxypeptidases. Meth. Enzymol. XI, 155, 1967. 2 Chrambach, A., Reisfeld, R.A., Wyckoff, M. & Zaccari, J.A. Procedure for rapid and sensitive staining of protein fractionated by polyacrylamide gel electrophoresis. Analyt. Biochem. 20, 150, 1967. 3 Curry, M.D., McConathy, W.J., Alaupovic, P., Ledford, J.H. & Popovic, M. Determination of human apolipoprotein E by electroimmunoassay. Biochim. biophys. Acta, 439, 413, 1976. 4 Davis, B.J. Disc electrophoresis. 11. Methods and applications to human serum proteins. Ann. N . Y. Acad. Sci. 121, 404, 1964. 5 Fish, W.W., Mann, K.G. &Tanford, C. The estimation of polypeDtide chain molecular weights by gel filtration in 6 M guanidine hydrochloride. J. biol. Chem. 224,4989, 1969. 6 Jackson, L'R., Morrisett, J.D., Pownall, H.J. & Gotto, Jr., A.M. Human high-density lipoprotein, apolipoprotein glutamine 11. J. biol. Chem. 248, 5218, 1973. 7 Kelly, J.L. & Alaupovic, P. Lipid transport in the avian species. Part 1. Isolation and characterization
of apolipoproteins and major lipoprotein density
Glycine- and serine-rich polypeptide of HDL
8
9 10
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
11
12
13
classes of male Turkey serum. Atherosclerosis, 24, 155, 1976. Kostner, G. Studies on the composition and structure of human serum lipoproteins. Isolation and partial characterization of apolipoprotein A 111. Biochim. biophys. A d a , 336, 383, 1974. Liu, T.Y. & Chang, Y.H. Hydrolysis of proteins with p-toluene sulfonic acid. J. biol. Chem. 746,2842, 1971. McConathy, W.J. & Alaupovic, P. Isolation and partial characterization of apolipoprotein D : a new protein moiety of the human plasma lipoprotein system. FEBS Letters, 37, 178, 1973. McConathy, W.J. & Alaupovic, P. Studies on the interaction of Concanavalin A with major density classes of human plasma lipoproteins. Evidence for the specific binding of lipoprotein B in its associated and free forms. FEBS Letters, 41, 174, 1974. McConathy, W.J. & Alaupovic, P. Studies of the isolation and partial characterization of apolipoprotein D and lipoprotein D of human plasma. Biochemistry, 15, 515, 1976. Moore, S. & Stein, W.H. Merh. Enzymol. VI, 819, 1960. .. Moore, S. On the determination of cystine as cysteic acid. J. biol. Clrem. 238, 235, 1963. Olofsson, S.-O., Fager, G. & Gustafson, A. Studies on human serum high-density lipoproteins V1. Studies on a cholesterol ester releasing reaction in virro. Scand. J. elin. Lab. Invest. 36, 481, 1976. Olofsson, S.-0. & Gustafson, A. Degradation of high-density lipoproteins (HDL) in vitro. Scund. J. clin. Lab. Invest. 33, Suppl. 137, 57, 1974. Olofsson, S.-0. & Gustafson, A. Studies on human serum high-density lipoproteins. 11. Isolation of subfractions in the cold. Scand. J. clin. Lab. Invest. 34, 257, 1974. Olofsson, S . - 0 . & Gustafson, A. Studies on human serum high-density lipoproteins. V. Isolation and
19 20 21 22 23
24
25
~
14 I5
16 17
18
26
27 28
755
characterization of a cholesterol ester-rich lipoprotein after in vitro incubation. Scand. J. clin. Lab. Invest. 36, 67, 1976. Ornstein, L. Disc electrophoresis. I. Background and theory. Ann. N.A. Acad. Sci. 121, 321, 1964. Ouchterlonv. 0. Antieen-antibodv reactions in eels: types of reactions incoordinated systems of diffusion. Acta pathol. microbiol. scand. 32, 321, 1953. Shelburne, F.A. & Quarfordt, S.H. A new apoprotein of human plasma very low density lipoproteins. J. biol. Chem. 249, 1428, 1974. Shore, B. & Shore, V. Isolation and characterization of polypeptides of human serum lipoproteins. Biochemistry, 8,4510, 1969. Shore, V. & Shore, B. Heterogeneity of human plasma very low density lipoproteins. Separation of species differing in protein components. Biochemistry, 12, 502, 1973. Shore, B., Shore, V., Salel, A., Mason, D. & Zelis, R. An apolipoprotein preferentially enriched in cholesteryl ester-rich very low density lipoproteins. Biochem. biophys. Res. Commun. 58, 1, 1974. Strid, L. Phosphopeptides from a tryptic hydrolysate of human casein. Acta chem. scand. 15, 1423, 1961. Utermann, G., Isolation and partial characterization of an arginine-rich apolipoprotein from human plasma very-low-density lipoproteins: apolipoprotein E. Hoppe-Seylersphysiol. chem. 356, 1 1 13, 1975. Wood, R.K. & Wang, K.T. Separation of dansylamino acids by polyamide layer chromatography. Biochim. biophys. Acta, 133, 369, 1967. Zilversmit, D.B., Hughes, L.B. & Balmer, J. Stimulation of cholesterol ester exchange by lipoprotein-free rabbit plasma. Biochim. biophys. Acin, 409, 393, 1975.
Received 10 June 1977 Accepted 26 August 1977
-
I
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
Isolation and partial characterization of a glycineand serine-rich polypeptide from human serum high-density lipoproteins (HDL) S . - 0 . O L O FSSO N , G . F A G E R & A . G U S T A F S O N Departments of Medical Biochemistry and Medicine I, Sahlgren’s Hospital, University of Goteborg, Goteborg, Sweden
Olofsson, S.-O., Fager, G. & Gustafson, A. Isolation and partial characterization of a glycine- and serine-rich polypeptide from human serum high-density lipoprotein (HDL). Scand. J. clin. Lab. Invest. 37, 749-755, 1977. Lately, several minor polypeptides characterized by a high content of glycine and serine have been described in human serum lipoproteins. By column chromatography we have isolated a glycine- and serine-rich polypeptide from totally delipidized high-density lipoproteins, apo-HDL. The partial characterization of this polypeptide by total amino acid composition and by immunology utilizing monospecific antisera to polypeptides of human serum lipoproteins, would indicate a unique polypeptide. This low molecular weight (4900 Daltons) glycine- and serine-rich polypeptide was characterized by a mobility on polyacrylamide gel electrophoresis, similar to that of polypeptide A-I, a blocked NHz-terminal amino acid and a terminal COOH-fragment consisting of R-SerAla-Gly-Gly. There are similarities between this polypeptide and the protein moiety of a cholesterol ester-rich lipoprotein fraction, HDLSUP,obtained by in vitro incubation of HDL subfractions and described in earlier publications by our group. The identity between these twopolypeptides, however, cannot beconfirmed by the present study. Key-words: glycine; lipoprotein; high density; polypeptide; serine Sven-Olof Olofsson, M.D., Department of Medical Biochemistry, University of Goteborg, S-400 33 Goteborg, Sweden
In addition to five well-characterized apolipoproteins, and their constitutive polypeptides,*
* Apolipoprotein A, apoA (composed of the nonidentical polypeptides A-I and A-11) comprises the bulk of the protein moiety of high-density lipoproteins (HDL). ApoB (polypeptide composition unknown) dominates in very-low-density lipoproteins (VLDL) and low-density lipoproteins (LDL) and apoC (with nonidentical polypeptides c-1,C-11 and c-111) is common to VLDL and HDL. More recently, apoD [lo, 12,161 also referred to as A-I11 [8], and apoE (‘arginine-rich polypeptide’, [3, 21, 23, 261) have been included as minor polypeptide components.
data in the literature [15, 18, 22, 231 suggest that polypeptide chains, characterized by a relatively high content of serine and glycine are present - . within the lipoprotein system. We have presented evidence for a glycine- and serine rich polypeptide as the apolipoprotein of a cholesterol ester-rich lipoprotein released from highdensity lipoproteins (HDL) [15, 181 by in vitro incubation. In order to study the possible role of the glycine- and serine-rich polypeptides in serum lipoprotein structure and metabolism, we 749
750
S . - 0 . Olofsson, G. Fager & A . Gustafson
have isolated and characterized one of them. The present publication describes the preparation of this polypeptide from HDL, its further characterization by amino acid composition, immunology, COOH- and NH2terminal amino acids and molecular weight determination.
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
MATERIAL AND METHODS Blood was obtained from healthy normolipidaemic male volunteers (age 25-40 years), who had fasted overnight. Blood was allowed to clot for 60 min at +4"C and serum was recovered by low-speed centrifugation. Serum was immediately placed at an ambient temperature of tO"C and kept between 0 and +4"C throughout the experiments [17]. Isolation and purification of H D L . HDL was isolated by a two-step preparative ultracentrifugation procedure at solvent densities (D) 1.19 g/ml and 1.063 g/ml [17]. The further purification of HDL from contaminating lipoprotein B (LP-B) was carried out by column (K 9/30) chromatography on Concanavalin ASepharose 4B (Pharmacia Fine Chemicals, Sweden) [l I] using 0.2 mol/l sodium acetate, 0.1 mol/l sodium chloride and 0,001 mol/l of manganese chloride, magnesium chloride and calcium chloride (pH 6.8).Before this chromatography, the buffer composition of HDL was changed to the elution buffer by gel filtration on a column (K 15/30) of Sephadex G25 (Pharmacia Fine Chemicals, Sweden). The purified HDL was finally desalted on Sephadex G25 employing 0.05 mol/l phosphate buffer. The HDL preparation obtained showed a typical pattern on basic polyacrylamide gel electrophoresis (Fig. 2, Pattern A) and did not react against antisera to albumin or LP-B. All buffers used for the isolation and purification procedures were made up fresh in distilled, sterilized water containing 0.2% sodium azide. All reagents used were of analytical grade, except guanidine-HC1 which was purified on charcoal according to Kelly & Alaupovic [7]. Delipidization of HDL. Purified and lyophylized HDL preparations were totally delipidized by four extractions with a mixture of chloroform
and methanol (2:1, v/v) with approx. 12 ml of solvent per 10 mg of apolipoproteins. The delipidization procedure was completed with three washings with diethylether, after which apo-HDL was recovered by low-speed centrifugation and dried in uacuo. Fractionation of apo-HDL. Dried apo-HDL was immediately dissolved in 0.01 mol/l Tris-HCI buffer (pH 7.2) containing 0.05% of sodium EDTA and 6 mol/l of purified guanidine-HCI [7] and chromatographed on a column (K 26/100) of Sepharose 6B (Pharmacia Fine Chemicals, Sweden) equilibrated with the same buffer [21]. The total volume of the column was 438 ml and the void volume 134 ml. The flow rate was 6-8 ml/h and the effluent collected in 3.5 ml fractions. Fractions were read at 280 nm in a spectrophotometer (Zeiss). Before further characterization, the protein moieties were submitted to extensive dialysis against water, against 0.1 mol/l KCI and again water (with four daily changes with 100 times the volume of the dialysate) and finally lyophilized. Characterization and quantification of the protein moieties. Lipoproteins and apolipoprotein fractions were studied by double diffusion [20] in 1 % agar (Behringwerke, Marburg an der Lahn, W. Germany) employing barbital buffer (pH 8.6), ionic strength 0.05. Intact HDL was studied with rabbit sera containing antibodies to human albumin, a-llipoproteins and /3-lipoproteins (Behringwerke) and apo-HDL fractions also by monospecific antisera to apoA-I, apoA-11, apoC-I, apoC-1112, apoD and apoE. The latter antisera were made available by the courtesy of Dr P. Alaupovic, Oklahoma City, Oklahoma, USA. Apo-HDL fractions were dissolved in 0.1 % sodium dodesyl sulphate (SDS), 0.1 mol/l ammonium carbonate or 3% (by weight) Triton X-405 (KEBO AB, Stockholm, Sweden) in barbital buffer (pH 8.6). The concentration of fraction V applied to double diffusion varied in different experiments from 0.5 to 5 mglml. Basic polyacrylamide gel electrophoresis (PAGE) was performed in a Canalco Model 1200 unit, utilizing 10% acrylamide monomer (pH 8.6) and 8 mol/l urea [4, 191. Gels were fixed in 10% trichloracetic acid and stained with Coomassie Brilliant Blue [2]. The
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
GIycine- and serine-rich polypeptide of H D L
concentration of samples applied to PAGE was 0.25-2 mg/ml. Analyses of total amino acids were performed on a JLC-5 AH Automatic Amino Acid Analyser equipped with an integrator (Jeol Ltd, Tokyo, Japan). Hydolysis of protein moieties was carried out in 6 mol/l HC1 in uacuo for 22 h [13]. Cystine and cysteine were determined as cysteic acid following performic acid oxidation [14]. Tryptophan was determined after hydrolysis with p-toluenesulphonic acid
751
VI-tVO
4
PI. NH2-terminal amino acid analyses were performed by a method adopted from Wood & Wang I271 and by a method modified by Shelburne & Quarfordt [21]. For comparison bovine serum albumin, porcine insulin and human apoA-I were used. COOH-terminal amino acid analysis was performed by incubation with carboxypeptidase A at 37°C [I]. Protein material was dissolved in ethylmorpholine buffer containing 0.1% SDS utilizing an enzyme substrate ratio of 1:40 and samples drawn at 5, 10, 15, 30, 60 and 180 min. The protein content was determined by a micro-Kjeldahl method [25] assuming 16% nitrogen. Molecular weight estimation. Gel filtration was carried out on Sephadex G75 (Pharmacia Fine Chemicals, Sweden) utilizing a column K26/100. The column was equilibrated and eluated with 0.01 mol/l Tris-HC1 buffer (pH 7.2) containing 0.05 % of sodium EDTA and 6 mol/l guanidineHCl [5]. Calibration was performed with purified samples of human apoA-I (rnol. wt. 28,000), human apoA-I1 (rnol. wt. 17,400) apoA-I1 monomer obtained by reduction and carboxy methylation of apoA-I1 [6] (rnol. wt. 8700), cytochrome c (rnol. wt. 13,400), ACTH (mol. wt. 4500) and glucagon (mol. wt. 3520). The void volume (V,) was determined by chromatography of Blue Dextran 2000. VI (the internal volume) Vo was estimated by chromatography of 14Calanine. The retention coefficient ( K d ) [5] was calculated for standards and samples.
+
RESULTS Isolation of glycine- and serine-rich polypeptide (fraction V ) from apo-HDL. Gel chromatography of apo-HDL gave five fractions (Fig. 1).
0.25
c
P
100
200
300
400
500
Elution volume ( m l )
FIG. I . Chromatographic separation of the lipid-free protein moiety of human serum high-density lipoproteins (apo-HDL) on Sepharose 6B, utilizing 0.01 mol/l Tris-HCI buffer (pH 7.2) containing 6 mol/l quanidine-HCI and 0.05% Na-EDTA (7). Vo = void volume and VI = internal volume.
Fraction I (eluted closely after the void volume) contained apo-A-I as judged from PAGE and immunological properties. Fraction I1 contained mainly apoE (the arginine-rich polypeptide) contaminated by apoA-I. Fraction 111 showed apoA-I and apoA-I1 and Fraction IV the C-I1 and C-I11 polypeptides. Fraction V, finally, was eluted with a buffer volume of 424 ml slightly ahead of V I + V o (438 ml), corresponding to a K , of 0.95. After dialysis, the protein yield of fraction V corresponded to 0.25 mg/100 ml of serum by the present technique. Partial characterization of fraction V . The characterization by double immune diffusion of fraction V dissolved in SDS or in 3% Triton against specific antisera to apoA-I, apoA-11, apoC-I, apoC-111, apoD, apoB and apoE failed to reveal any irnmunoprecipitation lines. Although no unequivocal interpretation can be made from a negative result on immuno-doublediffusion, these data indicate the absence of those polypeptides as contaminations. On basic polyacrylamide gel electrophoresis (Fig. 2) fraction V, both when dissolved in 8 mol/l urea and in 0.1 % SDS revealed only one single band taking protein stain in the region of
752
S.-0. Olofsson, G . Fager & A . Gustafson
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
A
B
C
FIG. 2. Patterns o n basic polyacrylamide gel electrophoresis (PAGE). A = protein moiety of HDL (apeHDL). B = fraction V dissolved in 0.1Y SDS. and c = fraction v dissolved in urea containiig stacking gel.
apoA-I. This finding suggests fraction V to contain a single polypeptide. Total amino acid composition of fraction V, was characterized by a high content of serine and glycine, but also by lack of tryptophan (Table 1). It appears as if methionine also is
missing, since after HC1-hydrolysis and performic acid oxidation neither methionine nor methionine sulphoxide or sulphone was detectable. The amino acid composition of fraction V differed from that of polypeptide A-1 by its content of isoleucine, cystine and/or cysteine. Furthermore, the amino acid composition differed from that of apoE (arginine-rich polypeptide) by its low arginine content. The described procedures [21, 271 for the analyses of NH2-terminal amino acid failed to reveal any DNS-amino acids indicating a blocked NH2-terminalamino acid of the protein moiety of fraction V. In agreement with earlier data, the present technique, however, showed the expected NH2-terminal amino acids for apoA-I, insulin and albumin, respectively. lncubation with carboxy-peptidase A (Fig. 3) yielded only glycine during the first 10 min of incubation indicating glycine to be COOHterminal amino acid. The release of glycine reached maximum after 15 min, to a molar ratio of 2:1 amino acid/protein, assuming mol. wt. 4900. This suggests the last two amino acids to be glycine. Our data also showed that the next two amino acids to be released were alanine and serine. This would give the tentative COOH-terminal fragment of R-Ser-Ala-Gly-Gly.
TABLE I. Amino acid composition (moles per 1000 mol of amino acid residue) of fraction V (n = 5). Data are given as mean+ SEM. For comparison are given data on protein moiety of cholesterol ester-rich lipoprotein fraction, apo-HDLSUP,obtained after in virro incubation [18] and on glycine- and serine-rich polypeptide of Shore & Shore (22).
Amino acid
Fraction V Mean+ SEM
apo-HDLSUP ~ 7 1
Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Isoleucine Leucine Thyrosine Phenylalanine Lysine Histidine Arginine Tryptophan Methionine Cysteine/cystine
82+4 44+ 4 188+22 151+ 13 42+ 3 176+22 82+ 5 38+ 4 24+ 3 58+ 14 19+ 3 22+ 3 44+ 7 23+4 25+7 0 0 14
87 50 140 156 O* 165 59 50 30 90 25 28 45 19 41 n.d. trace 9
Polypeptide described by Shore & Shore [221 72 40 204 160 20 194 71 30 21 35 7 15 76 30 14 0 7 7
* Elution of hydrolysed amino acids not monitored at 440 nm in this series.
753
Glycine- and serine-rich polypeptide of HDL
LC
;2.0
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
E
0
Time iminl
FIG.3. COOH-terminal amino acid analysis. Action of carboxypeptidase A at 37°C on fraction V obtained, from human serum apo-HDL. Release of glycine ( C ) , alanine ( 0 )and serine ( x ) are given in relation to time of incubation.
4L 3
0.5
,.I
Relative r e t e r t i o n ( K d )
The homogeneity of fraction V was emphasized by its chromatographic behaviour on Sephadex G75 (Fig. 4).The fraction was eluted as a narrow, homogenous peak with a Kd of 0.51. The PAGE pattern of this peak was the same as that of fraction v. The Kd of fraction V on Sephadex GI5 correponded to a monomeric molecular weight of 4900 daltons (Fig. 5 ) . When estimated from data on relative amino amino acid composition (Table I), the molecular weight of 4920 daltons was obtained for the first multiple corresponding to forty-nine amino acid residues.
t
O O‘O
t
I
100
li
I \ I
200 300 400 Elution volume(rn1)
500
FIG. 4. Chromatographic behavior of fraction V, obtained from human serum apo-HDL by gel filtration on Sepharose 6B, when re-chromatographed on Sephadex G75 in 0.01 mol/l Tris-HC1 buffer containing 6 mol/l guanidine-HCI and 0.05% Na-EDTA. Inserted is pattern on basic PAGE of this fraction.
FIG.5 . Molecular size estimation by relative retention on Sephadex G75 (column K 26/100) utilizing 0.01 mol/l Tris-HC1 buffer containing 6 mol/l guanidine-HCI and 0.05% Na-EDTA. ApoA-I (28,000), apoA-I1 (1 7,400), Cytochrome c (13,400), apoA-I1 monomer (8700), ACTH (4540) and glucagon (3520) were used as references. Data plotted in semilogarithmic fashion. Relative retention of fraction V indicated by broken line.
DISCUSSION The present publication deals with the preparation of a glycine- and serine-rich polypeptide in high-density lipoproteins (HDL). Our data indicate that the present procedure gives a homogeneous preparation without contamination of any of the known apolipoproteins, apoA-I, apoA-11, apoC-I, apoC-11, apoC-111, apoB, apoD or apoE. The amino acid composition is characterized by a high content of glycine and serine. These amino acids, however, are common contaminants in protein preparations. In the present study, the content of glycine and serine was closely similar in repeated preparations (n = 5 ) . Furthermore, the good agreement between molecular weight estimation from amino acid composition and by gel filtration, would indicate contamination of free amino acids to be negligible. As to total amino acid composition, the lack of tryptophan and presumably of methionine is of interest. This glycine- and serine-rich polypeptide isolated from apo-HDL is further characterized by a blocked NH2-terminal amino acid and by glycine as COOH-terminal amino acid. Furthermore, gel filtration and calculations from total amino acid composition indicates a low mole-
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
754
S.-0. Olofsson, G . Fager & A . Gustafson
cular weight, 4900 daltons. In the present study, the yield of this glycine- and serine-rich polypeptide is low (0.25 mg/100 ml of serum). The isolated polypeptide reveals a low water solubility, but can be solubilized in sodium dodecyl sulphate (SDS), 8 mol/l urea, 3 % Triton X-405 and in 6 mol/l guanidine-HC1. When solubilized in these agents, the polypeptide fails to react with antisera to all known apolipoproteins. Recent data [18, 22, 231 would suggest the presence of several polypeptides rich in glycine and serine within human serum lipoprotein spectrum. A closer comparison between those is not possible because only incomplete data is available. We were able in a recent report [18] to isolate and partially characterize the apolipoprotein of a cholesterol ester-rich lipoprotein fraction obtained by in vitro incubation of HDL subfractions. This apolipoprotein was also characterized by a high content of glycine and serine (Table I). Furthermore, the finding of a blocked NH2-terminal amino acid was in common with the present polypeptide, as was its behaviour on PAGE. The comparison of total amino acid composition (Table I) would indicate certain differences as to relative amounts of amino acids. The apolipoprotein of the cholesterol ester-rich lipoprotein fraction, apoHDLSup,revealed a higher content of arginine and was not completely characterized as to proline and tryptophan content. It is possible that these differences were due to impurities in apo-HDL"P, rather than in the present preparation, fraction V. However, a final proof for identity between the protein moieties of HDLsup and fraction V, has to wait for immunology utilizing monospecific antibodies prepared against their protein moieties. The constantly low yield of these small polypeptides, has in our hands so far prevented the preparation of antibodies. Shore & Shore [22, 231 have described the presence. of glycine- and serine-rich apolipoprotein within human serum [22] as well as in very-low-density lipoproteins (VLDL) [23]. Those glycine- and serine-rich polypeptides were either not visualized on PAGE [22] or migrated as a fast moving band at the position of C-I11 [23]. The difference in relative amino acid composition other than glycine and serine seems furthermore to exclude identity to the present polypeptide.
Our earlier studies on the cholesterol esterrich lipoprotein fraction in HDL, HDLSuP, indicated the possibility of a physiological function of this lipoprotein fraction, as a mediator in cholesterol ester transport among lipoproteins. However, apo-HDLgupappears to be different from apoE (the arginine-rich polypeptide) [24] as well as from the protein moiety described by Zilversmit [28], both linked specifically to cholesterol ester transport.
ACKNOWLEDGMENTS The technical assistance of Miss Anita Jacobsson and Mrs Marie-Louise MBnsson and the secretarial aid of Miss Annika Hofde is gratefully acknowledged. The amino acid analyses were performed by Ulf Svanberg, B.Sc., and Mrs Monica Olsson. The authors also wish to express their appreciation to Drs P. Alaupovic and W. J. McConathy (Oklahoma City, Oklahoma, USA) for generous supply of monospecific antisera. This study was supported by grants from the Swedish Medical Research Council (19X-2100), the Swedish Oleo-Margarine Foundation for Nutritional Research, the Swedish National Association against Heart and Lung Disease, 'Goteborgs Lakaresallskap', and the Medical Faculty of Goteborg University.
REFERENCES 1 Ambler, R.P. Enzymatic hydrolysis with carboxypeptidases. Meth. Enzymol. XI, 155, 1967. 2 Chrambach, A., Reisfeld, R.A., Wyckoff, M. & Zaccari, J.A. Procedure for rapid and sensitive staining of protein fractionated by polyacrylamide gel electrophoresis. Analyt. Biochem. 20, 150, 1967. 3 Curry, M.D., McConathy, W.J., Alaupovic, P., Ledford, J.H. & Popovic, M. Determination of human apolipoprotein E by electroimmunoassay. Biochim. biophys. Acta, 439, 413, 1976. 4 Davis, B.J. Disc electrophoresis. 11. Methods and applications to human serum proteins. Ann. N . Y. Acad. Sci. 121, 404, 1964. 5 Fish, W.W., Mann, K.G. &Tanford, C. The estimation of polypeDtide chain molecular weights by gel filtration in 6 M guanidine hydrochloride. J. biol. Chem. 224,4989, 1969. 6 Jackson, L'R., Morrisett, J.D., Pownall, H.J. & Gotto, Jr., A.M. Human high-density lipoprotein, apolipoprotein glutamine 11. J. biol. Chem. 248, 5218, 1973. 7 Kelly, J.L. & Alaupovic, P. Lipid transport in the avian species. Part 1. Isolation and characterization
of apolipoproteins and major lipoprotein density
Glycine- and serine-rich polypeptide of HDL
8
9 10
Scand J Clin Lab Invest Downloaded from informahealthcare.com by Nyu Medical Center on 01/11/15 For personal use only.
11
12
13
classes of male Turkey serum. Atherosclerosis, 24, 155, 1976. Kostner, G. Studies on the composition and structure of human serum lipoproteins. Isolation and partial characterization of apolipoprotein A 111. Biochim. biophys. A d a , 336, 383, 1974. Liu, T.Y. & Chang, Y.H. Hydrolysis of proteins with p-toluene sulfonic acid. J. biol. Chem. 746,2842, 1971. McConathy, W.J. & Alaupovic, P. Isolation and partial characterization of apolipoprotein D : a new protein moiety of the human plasma lipoprotein system. FEBS Letters, 37, 178, 1973. McConathy, W.J. & Alaupovic, P. Studies on the interaction of Concanavalin A with major density classes of human plasma lipoproteins. Evidence for the specific binding of lipoprotein B in its associated and free forms. FEBS Letters, 41, 174, 1974. McConathy, W.J. & Alaupovic, P. Studies of the isolation and partial characterization of apolipoprotein D and lipoprotein D of human plasma. Biochemistry, 15, 515, 1976. Moore, S. & Stein, W.H. Merh. Enzymol. VI, 819, 1960. .. Moore, S. On the determination of cystine as cysteic acid. J. biol. Clrem. 238, 235, 1963. Olofsson, S.-O., Fager, G. & Gustafson, A. Studies on human serum high-density lipoproteins V1. Studies on a cholesterol ester releasing reaction in virro. Scand. J. elin. Lab. Invest. 36, 481, 1976. Olofsson, S.-0. & Gustafson, A. Degradation of high-density lipoproteins (HDL) in vitro. Scund. J. clin. Lab. Invest. 33, Suppl. 137, 57, 1974. Olofsson, S.-0. & Gustafson, A. Studies on human serum high-density lipoproteins. 11. Isolation of subfractions in the cold. Scand. J. clin. Lab. Invest. 34, 257, 1974. Olofsson, S . - 0 . & Gustafson, A. Studies on human serum high-density lipoproteins. V. Isolation and
19 20 21 22 23
24
25
~
14 I5
16 17
18
26
27 28
755
characterization of a cholesterol ester-rich lipoprotein after in vitro incubation. Scand. J. clin. Lab. Invest. 36, 67, 1976. Ornstein, L. Disc electrophoresis. I. Background and theory. Ann. N.A. Acad. Sci. 121, 321, 1964. Ouchterlonv. 0. Antieen-antibodv reactions in eels: types of reactions incoordinated systems of diffusion. Acta pathol. microbiol. scand. 32, 321, 1953. Shelburne, F.A. & Quarfordt, S.H. A new apoprotein of human plasma very low density lipoproteins. J. biol. Chem. 249, 1428, 1974. Shore, B. & Shore, V. Isolation and characterization of polypeptides of human serum lipoproteins. Biochemistry, 8,4510, 1969. Shore, V. & Shore, B. Heterogeneity of human plasma very low density lipoproteins. Separation of species differing in protein components. Biochemistry, 12, 502, 1973. Shore, B., Shore, V., Salel, A., Mason, D. & Zelis, R. An apolipoprotein preferentially enriched in cholesteryl ester-rich very low density lipoproteins. Biochem. biophys. Res. Commun. 58, 1, 1974. Strid, L. Phosphopeptides from a tryptic hydrolysate of human casein. Acta chem. scand. 15, 1423, 1961. Utermann, G., Isolation and partial characterization of an arginine-rich apolipoprotein from human plasma very-low-density lipoproteins: apolipoprotein E. Hoppe-Seylersphysiol. chem. 356, 1 1 13, 1975. Wood, R.K. & Wang, K.T. Separation of dansylamino acids by polyamide layer chromatography. Biochim. biophys. Acta, 133, 369, 1967. Zilversmit, D.B., Hughes, L.B. & Balmer, J. Stimulation of cholesterol ester exchange by lipoprotein-free rabbit plasma. Biochim. biophys. Acin, 409, 393, 1975.
Received 10 June 1977 Accepted 26 August 1977
-
I
