Comparative Biophysical Studies of Hepatitis B Antigen, Subtypes adw and ayw

Transcription

1 JOURNAL OF VIROLOGY, Jan. 1975, p Copyright American Society for Microbiology Vol. 15, No. 1 Printed in USA. Comparative Biophysical Studies of Hepatitis B Antigen, Subtypes adw and ayw R. CHAIREZ,1 F. B. HOLLINGER, J. P. BRUNSCHWIG, AND G. R. DREESMAN* Department of Virology and Epidemiology, Baylor College of Medicine, Houston, Texas Received for publication 5 August 1974 Comparative biophysical and biochemical analyses were performed on purified preparations of hepatitis B antigen (HBS Ag) subtypes adw and ayw, including isoelectric ph evaluations, analysis of the different morphological forms, molecular weight determinations, and analysis of the polypeptides by polyacrylamide gel electrophoresis. Both HB8 Ag-positive plasma and purified HB8 Ag were analyzed by electrofocusing in a sucrose ampholyte gradient. Four distinct populations of HB. Ag with a ph range of to for unfractionated plasma samples and to for purified samples were detected in both adw and ayw preparations. Electron microscopic studies of each population of purified HB. Ag revealed 19- to 27-nm spheres in each fraction. Purified material labeled with "'I by the chloramine-t method behaved as one major vpopulation with an isoelectric ph value of Purified adw preparations revealed a major population with a molecular weight of 3.7 x 106 and a second one of 4.6 x 106. Purified preparations of ayw contained one population with a molecular weight of 4.6 x 106. Polyacrylamide gel electrophoretic analysis of purified HB6 Ag revealed nine polypeptides for ayw and seven for adw particles. These studies indicate that purified preparations of HB6 Ag are heterogeneous and that distinct differences can be detected between the two subtypes. The complexity in the physicochemical properties of the particles associated with hepatitis B antigen (HB. Ag) is becoming increasingly more apparent (16). Proteins, lipids, and carbohydrates have been identified as constituents of the purified particles (1, 3, 7, 14, 19, 20). Recently, Howard and Zuckerman (13) observed two distinct populations of HB. Ag with different charges. Our laboratory has demonstrated the heterogeneity of size, molecular weight, and overall charge properties in purified preparations of HB8 Ag, subtype adw (HBS Ag/adw) (6). The present study compares the morphology, isoelectric ph (pl), molecular weight, and polypeptide composition of purified preparations of HB. Ag, subtypes adw and ayw. MATERIALS AND METHODS HB8 Ag purification. HB. Ag preparations of the adw or ayw subtypes were purified as described previously (5). Essentially, the method involved ultracentrifugation of HB8 Ag-positive plasma, acidification of the resulting pellet at ph 2.4 for 1 h, and repelleting, followed by two sequential isopycnic density centrifugations in CsCl. Final purification was accomplished by rate zonal centrifugation in a preformed linear sucrose gradient, followed by removal ' Present address: Department of Biology, George Mason University, Fairfax, Va of the sucrose by dialysis against several changes of 0.15 M NaCl for 24 h. Immunochemical techniques. The presence of HB8 Ag antigenic activity was monitored by a microtiter complement fixation test (15) or by the doubleantibody radioimmunoassay (RIA-DA) technique (10, 11). Antibody to HB8 Ag (anti-hb.) used in these assays was prepared by injecting guinea pigs with purified HB. Ag as described elsewhere (5, 12). The subtype specificity of the HB. Ag-positive plasma was determined with the agar gel diffusion method by using specific guinea pig anti-hb/adw and anti-hbj/ ayw sera and rabbit anti-hbjadw and anti-hbj/adr sera. Protein determination. Total protein content of purified HB. Ag was determined by the Lowry method, with crystalline bovine albumin as a standard (17). lodination. Purified HB. Ag was labeled with '2II or 1"1I by the chloramine-t technique (9) with modifications described elsewhere (10). The specific activity of the labeled preparations ranged between 5 and 15 ACi/jLg. Isoelectric point determination. Isoelectric points were determined in an ampholyte ph gradient of 3.0 to 6.0 in an LKB electrofocusing column as described previously (6). All samples were dialyzed against four changes of 1% glycine for 48 h before electrofocusing. Electron microscopy. Purified HB. Ag samples were prepared for electron microscopy by the droplet- 182

2 VOL. 15, 1975 HEPATITIS B ANTIGEN, SUBTYPES adw AND ayw pseudoreplication method described previously by McCombs et al. (18). Briefly, a drop of HB. Ag suspension was diffused into a small block of 2% agar followed by addition of a drop of parlodion. The resulting film was then stripped from the agar block by floating on a 0.2% phosphotungstate acid solution (ph 7.0) containing 0.05% sucrose and subsequently placed on a copper grid (300 mesh), dried, and examined in a Hitachi llb electron microscope. Analytical ultracentrifugation. Molecular weight studies of purified HB8 Ag, subtypes adw and ayw, were done in a Spinco model E analytical ultracentrifuge by using Rayleigh interference optics. The molecular weight values were determined by the use of the high-speed equilibrium method of Yphantis (21). All runs were made in a double-sector interference cell in the An-J rotor at speeds of 3,000, 4,000, 5,200, and 6,400 rpm. Procedures for establishing equilibrium, photographing interference patterns, and measuring interference fringes were as described previously (6). Polyacrylamide gel electrophoresis. Disc polyacrylamide gel electrophoresis of the purified HB. Ag preparations was done in 10% gels. Eighty micrograms of purified HB. Ag/adw or HB. Ag/ayw was solubilized in 1% sodium dodecyl sulfate, 1% beta-mercaptoethanol, and 6 M urea by heating at 60 C for 10 min as described by Gerin et al. (8). Sodium dodecyl sulfate-polyacrylamide gels and buffer were prepared as previously described (3), and electrophoresis was performed essentially as described by Davis (4). Molecular weight markers used were bovine serum albumin, ovalbumin, chymotrypsinogen, and cytochrome c (Schwartz/Mann). Electrophoresis was performed at 4.0 ma per gel until the bromophenol blue tracking dye migrated at least 8 cm. The polypeptides were detected by staining with Coomassie brilliant blue (0.05% in 7% acetic acid) for 1 h and destaining in 7% acetic acid and 5% methanol. RESULTS pi of HB. Ag proteins. Iodinated purified HB8 Ag, unlabeled purified HB8 Ag, and HB8 Ag-positive plasma of subtypes adw or ayw were dialyzed against 1% glycine buffer and isoelectrofocused to equilibrium. Electrofocusing of iodinated HB8 Ag of the adw subtype showed the existence of two different patterns. Two adw sources designated adw (A) had a single pl at 3.9 ± 0.1, whereas two other adw sources designated adw (B) showed the isoelectric populations at ph 4.0 i 0.1 and , analogous to those previously reported by this laboratory (6) (Fig. 1A and B). Similar electrofocusing of eight different ayw sources revealed a single radiolabeled population at ph 3.9 i 0.1 (Fig. 1C). It should be pointed out that since.these studies were completed, similar iodinated preparations have been analyzed in a ph gradient range of 2.5 to 4.0. A somewhat lower pi value of 3.8 ± 0.1 was noted under these conditions. Electrofocusing of unlabeled purified preparations of HB8 Ag of the same origin as those a - z0c IL FRACTION NUMBER 12mI/fr.) FIG. 1. Electrofocusing of 125I-labeled, purified HB, Ag, subtypes adw, source (A); adw, source (B); and ayw, source (C), with an ampholyte ph range of 3.0 to 6.0 stabilized in sucrose. The material was electrophoresed for 44 h with a final voltage of 700 V. These patterns were reproduced for eight different iodinated preparations. studied above also resulted in two different patterns (Fig. 2A and B). Electrofocusing of two adw (A) preparations revealed a major antigenic population (pi = ) and three smaller populations with pi values of 3.9 ± 0.05, 4.0 ± 0.1, and , respectively, as monitored by RIA-DA. In contrast, the majority of the antigenic activity of another purified preparation of adw (B) was detected in two peaks with pi values of 3.9 ± 0.05 and 4.9, with some activity associated with two other populations (pl = 4.1 i 0.05 and 4.4 ± 0.1). The pattem observed by electrofocusing ayw preparations also showed heterogeneity in the overall charge properties; six populations observed in the majority of the ayw preparations had similar pi values of 3.9 ± 0.05, , 4.2 ± 0.1, 4.4 i 0.1, 4.9, and (Fig. 2C). The pattern of reactivity of the ayw preparations was most similar to the B source of adw. Since a shift to a more acid ph was observed by iodination of purified HB8 Ag, it seemed v

3 184 CHAIREZ ET AL. J. VIROL. 7 b I I z 0 I 0 (- 0n or 6 _ 6 2 cl 5 4 Z 6 ol FRACTION NUMBER (2milfr.i FIG. 2. Electrofocusing of unlabeled purified HB, Ag, subtypes adw, source (A); adw, source (B); and ayw, source (C). Conditions of electrophoresis are described in Fig. 1. These patterns were reproduced for three different purified preparations of each subtype. plausible that the purification procedures might also affect the overall charge of HB. Ag. Therefore, each of the unfractionated plasmas from which the above preparations had been derived was fractionated by electrofocusing with subsequent testing of the antigenic activity by RIA- DA. Similar patterns were observed with each of these three plasmas (Fig. 3). However, there were considerable differences in the relative proportion of each isoelectric population among the various sources. All adw and ayw sources had a major peak at pi = In addition, the adw (B) sources had a peak at pl = 5.4 ± 0.1, which possibly corresponded to the pi = 5.2 found in the adw (A) and ayw sources. It should be noted that all of the antigenic activity was associated with populations with pi values in a considerably higher ph range than those detected in the purified preparations. The antigenic activity associated with the more basic ph range (pl > 6) observed in FRACTION NUMBER (2mI/fr.) FIG. 3. Electrofocusing of unfractionated HB, Agpositive plasma of subtypes adw, source (A); adw, source (B); and ayw, source (C). Conditions of electrophoresis are described in Fig. 1. These pattems were reproduced for three different plasmas of each subtype. several preparations of purified HB8 Ag and with whole plasma has not been investigated further, but it appears to be an artifact of the fractionation procedure. Electron microscopy of HB8 Ag isoelectric populations. The morphology of particles containing HB. Ag in unfractionated plasma, each charged population of plasma, and each charged population of the purified HB8 Ag for the adw (A and B sources) and ayw subtypes was examined by electron microscopy. The populations isolated by isoelectric focusing of HB6 Ag-positive plasma and purified HB. Ag of each subtype were quite reproducible from one experiment to another. Unfractionated plasma of both subtypes contained 19- to 27-nm spheres, 20 x 50- to 20 x 100-nm tubules, and 42-nm Dane particles. The number of tubules and Dane particles per milliliter in the plasma before electrofocusing varied considerably and depended on the source and not the subtype

4 VOL. 15, 1975 HEPATITIS B ANTIGEN, SUBTYPES adw AND ayw 185 being studied. The various pi populations of electrofocused plasma for each subtype contained 19- to 27-nm spheres, tubules, or Dane particles (Table 1; Fig. 4). The relative size of corresponding charged populations in the different subtypes adw (A and B) and ayw (C) in plasma did not appear to reflect any difference in content of tubules or Dane particles. Rather, the size of the peaks in both plasma and purified HB. Ag appeared to reflect the relative concentration of the differently charged 19- to 27-nm spheres. Except for the occasional short tubular forms found in some pi populations of electrofocused, purified HB. Ag of both subtypes, pi populations tested contained mainly the 19- to 27-nm spheres (Table 1; Fig. 5). However, the proportion of spheres in the upper end of the 19- to 27-nm range was greater in the pi = population of the adw (A) source. Measurements were made of particle size of the pi populations of the electrofocused plasma and purified material of each HB. Ag source studied. These measurements showed that only the pl = population of purified adw (A) contained spheres averaging 25 nm, whereas the spheres in the other pl populations averaged 22 nm. It should be noted that the adw (A) plasma contained abundant Dane particles and tubules both before and after electrofocusing. The Dane particles were not abundant in the ayw sources studied, and they were scarce in the adw (B) source. This difference in initial concentration of Dane particles explains the absence of Dane particles in the pi peaks of adw (B) plasma. Any disruption of the outer membranes of the Dane particles releases the 25- to 27-nm inner core. The fact that the adw (A) plasma contained TABLE 1. these particles in abundance may explain why only the adw (A) preparations of purified HB. Ag contained a major pi population (pi = 4.1 i 0.1) with spheres averaging 25 nm in diameter. Analytical centrifugation molecular weight determination. The molecular weights of purified HB. Ag of the adw (A) and ayw sources were determined by the Yphantis technique (21). The average molecular weight of the adw (A) source in two different experiments was 3.7 x 10', with an additional minor population of 4.6 x 106 observed at the lower rotor speeds (Table 2). Studies of three different ayw sources in two different experiments revealed an average molecular weight of 4.6 x 106. Highermolecular-weight populations were also observed at low rotor speeds for ayw sources, but the material appeared to be heterogeneous and a straight line plot was not obtained. Disc polyacrylamide gel electrophoresis of HB. Ag subtypes. Purified HB. Ag was dialyzed overnight against electrophoresis buffer and solubilized, and the resulting polypeptides were separated in 10% gels by disc polyacrylamide gel electrophoresis (Fig. 6). Coomassie blue staining revealed seven polypeptides for the adw subtype with estimated molecular weights of 19,000, 24,000, 27,000, 35,000, 40,000, 55,000, and approximately 120,000 (Table 3). Subtype ayw also had seven polypeptides of these molecular weights plus two additional weakly staining polypeptides of 69,000 and 105,000. DISCUSSION This study presents further evidence of heterogeneity in both the adw and ayw subtypes of Morphology of hepatitis B antigen (HB, Ag) populations purified by isoelectrofocusing Isoelectric ph populations of starting material HB. Ag Particle Purified HB. Ag Whole HB. Ag positive plasma subtype description0a 3.9± 4.1± 4.5± ± 5.1± 5.3± >. > >6.0 adw D (A) SP Jr + +b ± + + T ST ±j ±- + adw D (B) SP ± + +± ++ + T ± ± ST ±- ± ayw D + + (C) SP ++± ±±- T ±= + ±f + ST 4- ±= ad, Dane particle; SP, spherical particle; T, tubular form; ST, short tubular form. +, Relative abundance of each form, based on a scale of ± to

5 ..).f ;h.04>j-*.. "..... i 'P.:, 11, 4--%U I- I -e f"ar t. * S.s A*% 4>' :,2 ;; %, * ,,.. s%. s...,.... t X.4!.4 R.. '. '.'. v; g.4 2ilj,;5.. >l,. *. f7t - wi, # :-+e *-.. e:s g v \. S Q -e * +i b$da, *_'@' ^:.. X4 n w 4 SS o.7ap f e J; i :.$ *>r % ^ ' ; r f v,:!e lff ". :'. *, ^>s. L _,* 7>.e.*... I...,,.W,.. i' ; i.-... ori,w.j' 19 r < r ^' f 's.i *s *At.5 3 *_. >,. *:.t' 0L X': S:- s.. OF b., QrClSf.. ^s. s. f* g......,e,, ; ;e PV 4,,- :ṙ C^.. s is tb *- X_i; _ =.,,._z-: W;_ >;,.e.,. ws w.'4 r > SF&U w; ;seii Aiz *S<,,> s#.:e O.,., b. s b :S.. ^ a P: v '.. pl = 4*5 to ti. B pl=s*o * C wb s L...,t,., X;. teg9_x5,:w } pl=5<2-5@4.".;..%a I.. D..t. It -.N:..:.4 t i Cf r, -:ix 186 pi >6.O

6 VOL. 15, 1975 HEPATITIS B ANTIGEN, SUBTYPES adw AND ayw HB. Ag. In addition to the biophysical evidence for heterogeneity of pi populations and molecular weight values, electron microscopic examination of fractions positive for HB8 Ag also shows morphological heterogeneity. The initial pattern observed by electrofocusing of iodinated (by the chloramine-t method) purified HB. Ag of adw subtype, showing two populations with pl values of 4.0 and 4.4, has now been found to be characteristic of approximately 10 to 20% of adw sources. All the other sources show only a single radiolabeled pl at ph 3.9. With a lower gradient (2.5 to 4.0), a value of was observed. The significance of these charged populations can be inferred from the fact that primarily tyrosine and histidine residues are labeled with 125I. Thus, polypeptides rich in these two residues would be preferentially labeled. This observation is relevant to the RIA detection of HB8 Ag, which is based on the inhibition by an unknown amount of HB8 Ag with a standard amount of iodinated HB, Ag for antibody binding sites. Since the several charged HB. Ag populations do not iodinate equally well, their detection may be quite variable by current RIA testing procedures. The original plasma of preparations of adw (A), adw (B), and ayw were each electrofocused to further determine the charge heterogeneity of the three morphological forms of HB8 Ag, i.e., spheres, tubules, and Dane particles. The relative similarity and reproducibility of the pi populations between various sources of adw (A), adw (B), and ayw HB. Ag preparations indicate that these pi peaks are not experimental artifacts and that they represent similarly charged populations occurring in all subtypes studied. In a recent study (2), the morphological forms (15- to 22-nm spheres, 20 x 300- to 20 x 700-nm tubules, 42-nm Dane particles and 24-nm inner cores) of an HB. Ag-positive plasma were purified, iodinated, and studied biophysically. Each labeled morphological form showed a distinct pl ranging from 3.65 to 4.7; however, as demonstrated in this study, the process of iodination alters the overall charge of the particles or preferentially labels only the most basic population. The similarity in electrofocusing pattern between purified HB8 Ag of the adw (A), adw (B), and ayw sources also indicated that the patterns depict charged populations with some common 187 antigenic determinants. The only major difference between the subtypes is in the relative amount of each charged population, but this could reflect not only the presence of d or y but also the relative amount of a in each fraction as detected by the RIA-DA assay. Different amounts of each determinant may thus contribute to the charge differences. Perhaps the most significant difference between the electrofocusing patterns of plasma and of purified HB8 Ag is the appearance of a 3.9 population in the purified material that is not present in plasma and the concomitant disappearance or reduction in the populations with pl values of 5.2 to 5.4 that are present in plasma. The reduction of the 5.2 to 5.4 populations may have been due to disruption of antigen-antibody complexes or to the release of normal serum components from the HB8 Ag particles during the acidification step performed during purification. Electron microscopy did not show any significant morphological differences between the pi = 5.2 to 5.4 peak and any of the others. Similar morphological forms (20 to 25 nm) were also found by immune electron microscopy for the isoelectric ph populations reported by Howard and Zuckerman (13). Using partially purified material, they reported only two pi values for ad (3.65 and 4.33) and for ay (3.95 and 4.9). Their preparations with pi values of 3.95 and 4.9 corresponded exactly with two of our charged populations for purified HBE Ag. Electron microscopic examination of the purified isoelectric peaks in this study showed that the 25-nm spheres present in purified HB8 Ag preparations were observed with the greatest frequency at ph 4.1. This agrees with our previous report in that a population of particles containing spheres of this size, probably representing hepatitis B core antigen (HBC Ag), had two iodinated populations with pi values of 4.15 and 4.7. The present data would suggest that HBc Ag derived from the Dane particle may be associated with the antigenic activity detected at ph 4.1. Final evaluation of this observation must await the development of a sensitive method specific for the detection of HBc Ag. The previously published molecular weight determinations of purified HB8 Ag were 3.56 x 106 and 4.57 x 106 for the two different-sized particles in adw (B) purified preparations as FIG. 4. Electron micrographs of isoelectric point populations showing particle types representative of those found in electrofocused HB, Ag-positive plasma of subtypes adw and ayw. Panels A, B, C, and D show isoelectric points for pi 4.5, 5.0, 5.2 to 5.4, and >6.0, respectively. Preparations were stained with potassium phosphotungstate. x 110,000.

7 -eq, A I.- pl =3.9 9'-- 4-> Downloaded from fnre,4,4i '4 pi~i on July 11, 2018 by guest pl= ! -f I:r...,..50. jk-" "L 188 p6. 0 pi >6.0

8 VOL. 15, 1975 HEPATITIS B ANTIGEN, SUBTYPES adw AND ayw TABLE 2. Analytical ultracentrifugation measurement of molecular weights of purified HB, Ag, subtypes adw and ayw Rotor speed Average" apparent mol wt (x 10') adw subtype ayw subtype' 3, , , , Avg mol wt 'Molecular weight values given for subtype adw were from a different source than used by Dreesman et al. (6); values for subtype ayw were averages of two experiments using three different sources of ayw. bthe ayw subtype also appeared to have an additional higher-molecular-weight component at low rotor speeds, but it was not possible to obtain a reliable value for this second component. TABLE 3. Polypeptidesa of HB. Ag, subtypes adw and ayw Protein HB. Ag/adw5 HB. Aglayw' 1 120, , , , ,000 55, ,000 40, ,000 35, ,000 27, ,000 24, ,000 19,000 adisc polyacrylamide gel electrophoresis was performed on 10% gels; 80,gg of HB. Ag of each subtype was solubilized in 1% sodium dodecyl sulfate, 1% beta-mercaptoethanol, and 6 M urea and heated at 60 C for 10 min. I Relative concentrations of the major polypeptides were estimated on the basis of staining intensity with Coomassie blue; major proteins are designated by italics. reported by us (6). Further analytical centrifugation molecular weight determinations of purified HB. Ag of the adw (A) and ayw subtypes showed that the ayw had a higher molecular weight than the main component of the adw (A) and adw (B) sources. The main molecular weight component in ayw sources is 4.6 x 106, which is 900,000 larger than the main molecular weight component of the adw subtype. E 5644 UN 7~~~~~~~~~~~ 0e RELATIVE MIGRATION FIG. 6. Solubilized samples of purified preparations of (A) HB, Ag/adw and (B) HB, Aglayw separated by polyacrylamide gel electrophoresis in 10% sodium dodecyl sulfate-urea (ph 8.8) gels. The gels were fixed, stained with Coomassie brilliant blue, destained, and scanned at 550 nm with a Gilford 240 linear transport apparatus. We have detected two minor polypeptides in association with HB. Ag/ayw that were not observed in the stained gels containing equal amounts of solubilized adw subtype. In this regard, Gerin (7) reported the presence of seven and six polypeptides for HB8 Ag/ayw and HB8 Ag/adw, respectively. A previous study from this laboratory (2) has shown an identical glycoprotein composition for the two subtypes. It might be added that purified HB8 Ag/adw iodinated to high specific activity has small amounts of the two polypeptides observed only in HB, Aglayw preparations in the present study: the 69,000- and 105,000-molecularweight species (3). The significance of these differently charged particles and of the larger molecular weight for the ayw subtype remains to be determined. However, a more detailed biochemical analysis of the different morphological forms present in the two subtypes should help elucidate the biophysical differences reported here. ACKNOWLEDGMENTS This study was supported in part by research contract DADA 17-73C-3074 from the U.S. Army Medical Research and Development Command and by Public Health Serice research grant HE from the National Heart and Lung Institute. We wish to thank Margaret Huebner, Melinda Freeman, and Karen Landry for excellent technical assistance. FIG. 5. Electron micrographs of isoelectric point populations showing 19 to 27-nm spheres representative of those found in electrofocused purified HB. Ag of subtypes adw and ayw. Panels A, B, C, D, E, and F show isoelectric points for pi 3.9, 4.1, 4.5, 4.9, 5.3, and >6.0, respectively. Preparations were stained with potassium phosphotungstate. x110,000.

9 190 CHAIREZ ET AL. J. VIROL. LITERATURE CITED 1. Burrell, C. J., E. Proudfoot, G. A. Keen, and B. P. Marmion Carbohydrates in hepatitis B antigen. Nature N. Biol. 243: Chairez, R., F. B. Hollinger, J. L. Melnick, and G. R. Dreesman Biophysical properties of purified morphologic forms of hepatitis B antigen. Intervirology 3: Chairez, R., S. Steiner, J. L. Melnick, and G. R. Dreesman Glycoproteins associated with hepatitis B antigen. Intervirology 1: Davis, B. J Disc-electrophoresis. II. Method and application to human serum proteins. Ann. N.Y. Acad. Sci. 121: Dreesman, G. R., F. B. Hollinger, R. M. McCombs, and J. L. Melnick Production of potent anti-australia antigen sera of high specificity and sensitivity in goats. Infect. Immunity 5: Dreesman, G. R., F. B. Hollinger, J. R. Suriano, R. S. Fujioka, J. P. Brunschwig, and J. L. Melnick Biophysical and biochemical heterogeneity of purified hepatitis B antigen. J. Virol. 10: Gerin, J. L Isolation and physicochemical characteristics of HB Ag, p In G. N. Vyas, H. A. Perkins, and R. Schmid (ed.), Hepatitis and blood transfusion. Grune and Stratton, New York. 8. Gerin, J. L., P. V. Holland, and R. H. Purcell Australia antigen: large-scale purification from human serum and biochemical studies of its proteins. J. Virol. 7: Greenwood, F. C., W. M. Hunter, and J. S. Glover The preparation of "3'I-labeled human growth hormone of high specific radioactivity. Biochem. J. 89: Hollinger, F. B Hepatitis, p In Manual of clinical microbiology. American Society for Microbiology, Washington, D.C. 11. Hollinger, F. B., V. Vorndam, and G. R. Dreesman Assay of Australia antigen and antibody employing double-antibody and solid-phase radioimmunoassay techniques and comparison with the passive hemagglutination methods. J. Immunol. 107: Hollinger, F. B., C. Wasi, G. R. Dreesman, and J. L. Melnick Subtyping of hepatitis B antigen by use of monospecific antibody-coated cells. J. Infect. Dis. 128: Howard, C. P., and A. J. Zuckerman Electrofocusing of hepatitis B antigen. J. Gen. Virol. 20: Kim, C. Y., and D. M. Bissell Stability of the lipid and protein of hepatitis-associated (Australia) antigen. J. Infect. Dis. 123: Laboratory Branch Task Force, and H. L. Casey Standardized diagnostic complement fixation method and adaptation to micro test. In Public Health Service Monogr. no. 74, Washington, D.C. 16. LeBouvier, G. L., and R. W. McCollum Australia (hepatitis-associated) antigen: physicochemical and immunological characteristics, p In K. M. Smith, M. A. Lauffer, and F. B. Bang (ed.), Advances in virus research, vol. 16. Academic Press Inc., New York. 17. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193: McCombs, R. M., M. Benyesh-Melnick, and J. P. Brunschwig Biophysical studies of vesicular stomatitis virus. J. Bacteriol. 91: Millman, I., L. A. Loeb, M. E. Bayer, and B. S. Blumberg Australia antigen (a hepatitisassociated antigen). Purification and physical properties. J. Exp. Med. 131: Steiner, S., M. T. Huebner, and G. R. Dreesman Major polar lipids of hepatitis B antigen preparations: evidence for the presence of a glycosphingolipid. J. Virol. 14: Yphantis, D. A Equilibrium ultracentrifugation of dilute solutions. Biochemistry 3: