Antigens from Blastomycin Purified Derivative

INFECTION AND IMMUNITY, Mar. 1976, p. 758-762 Copyright C 1976 American Society for Microbiology Vol. 13, No. 3 Printed in U.S.A. Preparative Isotachophoretic Separation of Skin Test Antigens from Blastomycin Purified Derivative MICHAEL V. LANCASTER' AND RONALD F. SPROUSE2 * School of Medicine and College of Veterinary Medicine, University of Missouri-Columbia, Columbia, Missouri 65201 Received for publication 5 November 1975 This investigation examined the theoretical and practical parameters requisite to preparatory isotachophoretic separation of blastomycin purified derivative. It resulted in a relatively simple, two-step procedure for preparation of blastomycin antigens in milligram quantities that exhibited sensitivity and specificity in experimentally infected guinea pigs. Analysis of the nine isotachophoretic fractions for skin test sensitivity and specificity provided some insight into the generally accepted unreliability of blastomycin when used for immunological evaluation. Blastomycin, the crude mycelial growth filtrate of Blastomyces dermatitidis, generally has been of little value in serodiagnosis, differential diagnosis, or epidemiological surveillance. This probably is the result of a composite of causes, such as extremely localized exposure of the host population to the etiological agent, considerable variation in host susceptibility to infection, and/or sensitization upon exposure to the etiological agent, and certainly the quality of the antigen used. This investigation was concerned with the latter point. A recently developed technique that combines the theoretical aspects of disc electrophoresis and isoelectric focusing has allowed the isolation of components based upon their relative electrophoretic mobilities. The technique, termed isotachophoresis, has been developed both theoretically and practically by several investigators (5). Consequently a series of requirements has been formulated which, when satisfied, offers a system for separation of multiple species with the characteristics of constant zone sharpening, concentration of individual zones, and equal migration velocities for all species. Isotachophoresis can provide a separation technique with better resolution than either disc electrophoresis or isoelectric focusing. Thus, the objective of this study was largescale isolation of homogeneous components from a partially purified derivative of blastomycin (BPD) by preparative isotachophoresis and subsequent evaluation of those components for skin test reactivity. ' Present address: Department of Dermatology, Army Institute of Research, The Presidio of San Francisco, San Francisco, Calif. 94129. 2 Temporary address: Department of Experimental Pathology and Anatomy, Mayo Clinic and Medical School, Rochester, Minn. 55901. MATERIALS AND METHODS Antigens. Blastomycin KCB-26 was obtained from the United States Public Health Service. BPD was obtained from blastomycin KCB-26 by gel filtration chromatography (R. F. Sprouse and M. V. Lancaster, submitted for publication) and used for subsequent steps in this study. Determination of pi. BPD was subjected to analytical polyacrylamide isoelectric focusing to determine specific isoelectric points (pl) for the major protein components and, more importantly, to ascertain the range of pl values for all the components. Such information, when combined with the principles of isotachophoresis, resulted in calculated optimal separation parameters, which were applied to the technique. Analytical polyacrylamide gel isoelectric focusing was carried out according to the method of Catsimpoolas (1). The gels were stained by the method of Griffith and Catsimpoolas (5) with mercuric chloride-bromophenol blue and densitometric tracings obtained using a model 542 Densicord recording electrophoresis densitometer (Colab Laboratories, Inc., Chicago Heights, Ill.) Each isoelectric focusing run consisted of gels with protein samples and gels containing only carrier ampholytes. After isoelectric focusing, the reference gels were sliced in 1-mm sections, respective sections were eluted in 1 ml of distilled water, and the ph of each aliquot was determined by using a conventional ph meter with a microelectrode. All procedures were performed at 20 C. Corresponding pi values of the stained gels then were estimated by superimposing ph gradients determined for the reference gels. Isotachophoretic system. Preparative isotachophoretic separation of BPD was accomplished by using an LKB 7900 Uniphor electrophoresis system (LKB, Bromma, Sweden). The method of Svendsen and Rose (11) with a modification of the elution chamber membrane suggested by Svendsen (10) was used for these studies. The system used a 3.75% polyacrylamide gel, prepared by photopolymerization of a mixture of stock solutions. The stock acryl- 758

VOL. 13, 1976 amide solution contained 30 g of acrylamide and 1 g of N,N'-methylenebisacrylamide, and was made up to 100 ml with water. The stock sucrose solution was 25% in water. The stock catalyst solution was 4 mg of riboflavin in 100 ml of water. The stock gel buffer solution contained 30 ml of glacial acetic acid, 20 g of tris(hydroxymethyl)aminomethane, and 3 ml of N,N,N',N'-tetramethylethylenediamine, made up to 1,000 ml with distilled water (ph 4.4). The leading electrolyte was the same as the stock gel buffer solution (acetate, ph 4.4). The terminating electrolyte was 6 g of tris(hydroxymethyl)aminomethane and 30 g of glycine in 2,000 ml of water (glycinate, ph 8.3). Gels were prepared by mixing acrylamide solution, sucrose solution, gel buffer solution, and catalyst solution with water in ratios of 1:1:1:1:4, respectively, and decanting into a glass column (25 by 400 mm). Water was layered over the gel solution to produce a flat surface upon photopolymerization. Column assembly. The column was assembled upon completion of the polymerization step, and leading electrolyte buffer was introduced into the lower electrode chamber (anode) and into the elution chamber. Terminating electrolyte buffer was introduced into the upper electrode chamber (cathode). Elution was accomplished by using the column modification described by Svendsen (10). The dialysis membrane normally used to form the elution chamber was replaced with a cellulose acetate filter. Eluant was pumped into the chamber at 5 ml/h and effluent was pumped out of the column at 20 ml/h. The result was forced flow of elution buffer upwards through the cellulose acetate membrane. Such forced flow provided more efficient washing of the elution chamber and thus sharper elution patterns. Isotachophoresis. Once elution was initiated, 3.25 ml of ampholyte-sample mixture was layered at the gel-buffer interface and current was applied. A maximum of 10 W was allowed for a separation performed at 4 C. Effluent was continuously monitored at 280 nm, absorbance was recorded, and fractions were collected in 5-ml aliquots. Desalting and characterization. Individual fractions subsequently were desalted and further purified by gel filtration on a column (0.9 by 40 cm) of Sephadex G-15 dextran. All fractions were lyophilized, weighed, and reconstituted to 100 gg (dry weight)/ml with nonpyrogenic distilled water. Protein was determined quantitatively by the modified Folin phenol method (7), using crystalline bovine serum albumin as standard. Fractions were diluted to a final concentration of approximately 50 gg/ml. Skin testing. Guinea pigs used for these studies were sensitized and/or infected with either B. dermatitidis or Histoplasma capsulatum as described by Goodman et al. (4). BPD fractions were assayed for skin test reactivity by intradermal inoculations (Mantoux) of approximately 5 jig of protein dissolved in 0.1 ml of sterile nonpyrogenic distilled water. Guinea pig sensitivity was evaluated concurrently by inoculation of respective animals with either BPD or histoplasmin HKC-5 (1:37.5). Each BPD fraction was tested on 25 B. dermatitidis- and ISOTACHOPHORESIS OF BPD 759 25 H. capsulatum-sensitized and/or -infected guinea pigs. Each antigen also was simultaneously tested on 12 nonsensitized control animals. Any inoculation site exhibiting questionable validity caused by inflammation due to site preparation, scratching, etc., was eliminated from the data base. Reactions were read at 24, 48, and 72 h after inoculation, and indurations were recorded in 5-mm increments, beginning with 0.5 mm. Negative reaction was recorded for 0-mm induration. Only 48-h readings were used for interpretation in these studies. RESULTS Figure 1 represents the results of analytical polyacrylamide gel isoelectric focusing of BPD. Significant concentrations of proteinaceous material appeared between values of ph 4 and 8 in eight distinct zones. These data indicated that a wide-range (pi 3 to 10) mixture of ampholyte spacer ion would be appropriate for isotachophoretic separation of BPD. Preparative isotachophoresis of BPD resulted in the separation of components shown in Fig. 2. Aliquots pooled to produce individual fractions are distinguished by arrows and designated with letters A through J. Three discrete areas of absorbancy, fractions F, H, and J, were obvious, and one less discrete area, fractions B, C, and D, indicated a triplet combination of components. Zones of similar amphoteric species migrated closely spaced in an isotachophoretic system, and consequently some zonal mixing occurred during elution. Gel filtration chromatography, used to separate buffer salts and low-molecular-weight spacer ions from BPD components, also proved capable of separating these individual BPD components mixed during the elution step. This composite preparative isotachophoretic gel filtration procedure provided milligram quantities of chromatographically pure, salt-free BPD components for subsequent skin test evaluation. Table 1 indicates the results of skin testing sensitized guinea pigs with BPD, isotachophoretic fractions of BPD, and histoplasmin HKC-5. A maximum of 25 sites was read for fractions A and E and a minimum of 15 sites was evaluated for fraction H. All other antigens were tested on 16 to 22 different guinea pigs. Fractions B and F were reactive only on the B. dermatitidis-infected animals. Fraction F elicited an induration of 10 to 15 mm, whereas a similar quantity of fraction B induced only a 5- to 10-mm reaction. In addition, fraction B elicited positive reactions on only 55% of the animals tested, whereas fraction F induced reactions on 100%. Fraction G appeared to display specificity for H. capsulatum-infected animals. However, the individual reactions were highly variable in size, and only 4 of the 16 animals tested exhibited positive reactions. Fraction D, the trailing portion of the B-C-D triplet, was the only other

760 LANCASTER AND SPROUSE INFECT. IMMUN. 0-o -0 I.- SM o0 o0 I a- z 0' 0 i 2 3 4 5 6 7 8 9 10 I1 GEL LENGTH( cm' FIG. 1. Isoelectric focusing pattern for BPD. The solid line shows the densitometric analysis of mercuric chloride-bromophenol blue-stained BPD proteins in the gel length. The broken line indicates the ph of respective 1-cm sections of control gel without BPD. Cw 4,, 0 w t5i z 3 125 I E F H i i's * A T fs k* if is lii ix is ix TUBE NUMBER FIG. 2. Isotachophoretic pattern for BPD. Fractions A through J are represented between the arrows and represent respective pooled 5-ml aliquots. fraction eliciting greater than 5-mm induration. DISCUSSION Numerous reports over the years have indicated that crude blastomycin is an extremely heterogeneous mixture of antigenic components. Even more recent reports of as yet highly speculative purified antigens either imply only partial purification (Sprouse and Lancaster, submitted for publication) or fail to prove homogeneity (2). These data indicate that once several operating parameters have been determined either empirically or by experimentation, isotachophoresis is capable of largescale, high-resolution separation of slow-migrating and dilute components of BPD. Amphoteric species generally migrate in an isotachophoretic system in order of their respective isoelectric points. Consequently, carrier ampholytes, introduced as spacer ions in such a system, must have a range of isoelectric points encompassing the pl values of the proteinaceous components to be separated. Initially, individual skin test reactivity was not known for any of the BPD components; therefore an ampholyte range was chosen to cover the entire spectrum of pi values. That pl range was determined to be ph 4 to 8 by analytical isoelectric focusing. However, once skin test reactive com-

VOL. 13, 1976 TABLE 1. ISOTACHOPHORESIS OF BPD 761 Results of skin testing isotachophoretic fractions of KCB-26 BPD 48-h induration (mm) Antigen Protein (pg/test) Infected animals B. dermatitidis H. capsulatum Control animals BPD 5.0 15-20a NP Negc Histoplasmin HKC-5 1:75 dilution NT 10-15 Neg BPD fractions A 3.8 0-5 0-5 Neg B 4.9 5-10 Neg Neg C 4.6 Neg Neg Neg D 4.7 5-10 0-5 Neg E 2.9 Neg Neg Neg F 5.4 10-15 Neg Neg G 5.1 Neg 5-10 Neg H 4.9 0-5 0-5 Neg J 5.1 0-5 0-5 Neg Mean induration measured to nearest 5-mm interval for all animals tested. NT, Not tested. MNeg, Antigen was tested and failed to elicit a reaction. ponents were characterized, the ampholyte range could be reduced to cover a narrower range of pi values, encompassing only those components of interest. This not only improved resolution of the desired components, but also reduced elutional cross-contamination of components with similar electrophoretic mobilities and allowed application of larger samples to the system. Separation conditions must be such that amphoteric species, i.e., proteins and spacer ions, are stacked between a fast leading ion and a slow terminating ion. Since migrational velocities for the BPD proteins were not known, the velocities of leading and terminating ions were adjusted empirically to encompass a wide velocity range (-35 x 10-; to -0.5 x 10- cm2/v.s). Isoelectric focusing data also indicated the presence of BPD species with pi values as low as 4 to 4.5. Thus, a gel buffer ph of 4.4 was determined to meet the ph requirements for this system while providing the leading and terminating ions with appropriate electrophoretic mobilities. Maximal resolution is dependent on all ionic species reaching equilibrium, at which point all species have constant concentrations relative to one another as predicted by Kohlrausch's regulating function. Consequently, sample size-to-column cross-sectional area (52.5 ml of ampholyte/mm2 of gel surface) and gel length to minimum time required for equilibrium resolution of all the species involved must be experimentally determined. Total sample size was determined to be 2 to 5 mg of BPD in a total volume of 2 ml and 1.25 ml of 40% spacer ampholyte solution. A gel length of approximately 15 cm was determined to be optimal for separation of the 2-ml sample of BPD. Finally, a modification of the elution system was required as described earlier (11), incorporating a cellulose acetate membrane in the system to minimize cross-contamination by the highly charged, low-molecular-weight spacer ions. Several inferences may be drawn from the results of skin testing. At least two components (fractions B and F) of BPD appear to be reactive only on guinea pigs sensitized to B. dermatitidis. Fraction F appears to be the most sensitive as well as the most specific since its mean induration was at least twice that of any other fraction. In addition, fraction F was the only component to elicit reactions on 100% of the B. dermatitidis-infected animals while producing negative results in all H. capsulatum-infected animals. The data also revealed a striking absence of sensitivity to any BPD component in any animal shown to be highly sensitive to histoplasmin HKC-5 (1:37.5), whereas several components elicited cross-reactive responses. Only one component (fraction F) elicited a response in B. dermatitidis-infected animals that was comparable to that of BPD. In summary, these data provide insight into two of the principle problems, variation in sensitivity and nonspecificity, encountered in using crude blastomycin as an immunological tool. Unfractionated blastomycin KCB-25 and KCB-26 both have been shown to possess at least workable levels of reactivity in human (3), canine (8), and guinea pig (9) testing. Both antigens also exhibit considerable nonspecific-

762 LANCASTER AND SPROUSE INFECT. T IMMUN. ity. The presence in high concentration of a single, antigenically specific component, either fraction B or F, would result in good reactivity for any crude antigen. However, variation in concentration of that component due to growth conditions, strain variation, or other factors would lead to variation of reactivity in each lot of blastomycin prepared. Such lot to lot variation then would reflect the acknowledged variation in sensitivity for various blastomycins. On the other hand, the susceptible host in naturally occurring sensitization and/or infection is exposed to a multiplicity of antigenic determinants from the hyphae, yeast, and various intermediate forms of B. dermatitidis. That same host in turn elicits a composite delayed response when skin tested with the heterologous mixture of components occurring in crude blastomycin. The data presented here would suggest that most of these components are poorly reactive, some are cross-reactive, and only a few exhibit any degree of sensitivity and specificity. Thus, it is not surprising that blastomycin is markedly unreliable when used for in vivo and in vitro testing. In conclusion, this study indicated a relatively simple, two-step purification procedure for preparation of blastomycin antigens in milligram quantities that exhibited sensitivity and specificity in experimentally infected guinea pigs. Whether or not these same antigens prove efficacious for human testing remains to be elucidated. However, at least one could presuppose some improvement of the present untenable use of crude blastomycin for human serological and skin test evaluation. LITERATURE CITED 1. Catsimpoolas, N. 1968. Micro isoelectric focusing in polyacrylamide gel columns. Anal. Biochem. 25:480-482. 2. Cox, R. A., and H. W. Larsh. 1974. Yeast- and mycelialphase antigens ofblastomyces dermratitidis: comparison using disc gel electrophoresis. Infect. Immun. 10:48-53. 3. Furculow, M. L., E. W. Chick, and J. Bussey. 1970. Prevelence and incidence of human and canine blastomycosis. Am. Rev. Respir. Dis. 102:60-67. 4. Goodman, N. L., R. Sprouse, and H. Larsh. 1968. Histoplasmin potency as affected by culture age. Sabouraudia 6:273-284. 5. Griffith, A., and N. Catsimpoolas. 1972. General aspects of analytical isotachophoresis of proteins in polyacrylamide gels. Anal. Biochem. 45:192-201. 6. Haglund, H. 1970. Isotachophoresis - a principle for analytical and preparatory separation of substances such as proteins, peptides, nucleotides, weak acids, and metals. Sci. Tools 17:2-12. 7. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. 8. Menges, R. W., M. L. Furcolow, L. A. Selby, H. R. Ellis, and R. F. Habermann. 1965. Clinical and epidemiologic studies on 79 canine blastomycosis cases in Arkansas. Am. J. Epidemiol. 81:164-179. 9. Sprouse, R. F. 1976. Mycoses, p. 153-161. J. E. Wagner and P. J. Manning (ed.), In Biology of the guinea pig. Academic Press Inc., New York. 10. Svendsen, P. J. 1972. On elution systems for column electrophoresis in gels -a universal elution system for column electrophoresis. Sci. Tools 19:21-22. 11. Svendsen, P. J., and C. Rose. 1970. Separation of proteins using Ampholene carrier ampholytes as buffer and spacer ions in an isotachophoresis system. Sci. Tools 17:13-17.