A Structural Model for Human Transfer Factor

A STRUCTURAL MODEL FOR HUMAN

TRANSFER FACTOR DISCUSSION PAPER Discussants: Denis R. Burger, Pamela A. Wampler, Arthur A. Vandenbark,

and David H. Regan Veterans Administration Medical Center and Northwest Cancer Research Foundation Portland, Oregon 97201

Transfer factor (TF) specifically denotes the ability of dialyzable leukocyte extracts (DLE) to transfer dermal reactivity in man. None of the other in vitro or in vivo phenomena attributed to DLE adequately detect or assess the dermal transfer component nor are these appropriate models for this reaction. Since this activity can still be evaluated only in a transfer system in man, the nature, specificity and mode of action of the responsible component (s) have remained controversial. The strategy that we employed to investigate the nature of the dermal transfer component was ( 1 ) to develop a reproducible system to assess passive transfer in man; (2) to fractionate DLE by exclusion chromatography, electrofocusing, and high pressure reverse phase (HPRP) chromatography in order to purify the active component for structural analysis; and (3) to probe the structure of TF before final purification by assessing TF sensitivities to enzymes with selective substrate activity. THEKL,H TRANSFER SYSTEM

Our approach was to use large batches of TF from leukapheresed donors who were immunized to a nonmicrobial antigen (keyhole limpet hemocyanin, KLH). Since TF activity can be assessed before and after immunization, we can relate the dermal reactivity in the recipient to the immunization of the donor.’, In this way our data argue for the concept of a specific TF raised through immunization:’ and support the specificity of TF reported in the literature.4- 6 A crucial feature of this transfer system is the naive status of the recipients before receiving TF from KLH-immunized donors. We presumed that this antigen would rarely be encountered by the recipient population. Considerable evidence has been obtained that indicates this is the case,G although the extent to which the recipients are naive to KLM in absolute terms is unknown. For this reason we can avoid subjecting recipients to a “priming” exposure in the form of a skin test, which would be required to select nonreactive individuals to microbial antigens. The blind evaluation of skin tests in recipients of TF from immune and control donors in the experiments reported here and previously 2 - ‘I has shown that dermal transfer can be detected sensitively. The importance of blind evaluation has been recently emphasized by Neidhart et ~ l . who , ~ were

236 0077-8923/79/0332-0236

$01.75/0 @ 1979, NYAS

Burger et al.: Human Transfer Factor

237

unable to assess systemic transfer accurately because of high frequency of background reactivity in their recipient pool. Additional controls that we employed included a phytohemagglutinin (PHA) skin test on all recipients to assure their ability to mount a dermal response (a positive control) and a coccidioidin skin test as a negative control. All recipients responded to the PHA skin test. The frequency of positive reactors to coccidioidin in TF recipients was on the same order of magnitude as the frequency observed in non-TF controls. This low frequency of reactivity to coccidioidin was expected since our recipient and control pools have had limited exposure to the Sonoran life zone where Coccidioides immitis is endemic. Taken together, these data are consistent with the selective recruitment of a KLHresponsive clone in recipients of TF from KLH-immune donors. SUMMARY OF FRACTIONATION OF EXTRACTS CONTAINING TF We have previously shown by Sephadex G25 exclusion chromatography that dermal transfer activity found in DLE was confined to a single fraction. This fraction (IIIA) eluted at 2.1 V,, and had an U V 254/280 ratio of about 5 . Purification of the IIIa fraction by electrofocusing indicated that the active component had a PI of 1.6. Since formation of gradients below pH 2.5 with available ampholytes was not possible, we turned to HPRP chromatography for further fractionation. Preparative HPRP chromatography was used to isolate the multiple components in fraction IIIa in quantities suitable for biological testing. We were surprised to find activity in two regions that were chromatoThese data are summarized in TABLE 1. graphically distinct:', Alkaline phosphatase treatment converted inosine monophosphate, which cochromatographed with TF activity, to inosine. This effected a considerable purification since less than 1% of the UV absorbing material remained in the active fraction. Fluorogenic monitoring of column effluents for primary amines showed that most peptides also eluted in this fraction.E Subsequent purification will involve alterations in the matrix-solvent system to achieve retention of the

TABLE 1 PROPERTIES OF HUMAN TRANSFER FACTOR Dermal Response Mean f S.D. (N) (1) Nonimmune donors do not transfer

(2) No TF activity before immunization but evident after immunization (3) TF elutes at 2-3 V, on Sephadex G25

compared to other fractions (4) TF has a PI ca. 1.6 (5) TF activity in two HPRP

regions (6) TF is not destroyed by alkaline phosphatase (7) TF activity in one HPRP region after alkaline phosphatase treatment .~

1 f l 2 2 1 14 k 3 15 2 3 2 f 1 12 2 4 10.5 4 13.5 f 5 12.4 f 2

*

12

f2

238

Annals New York Academy of Sciences

active component. These experiments and others designed to determine if dermal transfer activity for antigens other than KLH also resides in this region are in progress. A STRUCTURAL MODELFOR TF The resistance of TF to pancreatic deoxyribonuclease, pancreatic ribonuclease, and trypsin reported in the early l i t e r a t ~ r e , ~suggested . that the active moiety was not a conventional nucleic acid or protein. A polypeptide composition for TF has been an attractive hypothesis since it is feasible to generate the multiple combinations required to account for specificity with a dictionary of amino acids. A diversity of greater than 1Olo combinations can be generated from a polypeptide chain with 8 residues available for substitution where from 1 to 7 residue substitutions are required to derive a new specificity.1° Previous to this report, the evidence for a polypeptide moiety in TF has come from animal transfer systems 11*l 3 or has been inconclusive.'* Our approach to determine if a polypeptide component was essential for TF was to select two nonspecific peptidases (pronase and proteinase K ) for degradation experiments. Both enzymes destroyed dermal transfer (from greater than 10 to less than 5 mm skin tests) when employed in high concentrations (TABLE2). To examine further the polypeptide structure of TF, we used terminalspecific peptidases in additional experiments. Carboxypeptidase A, which is specific for the carboxy-terminus, destroyed TF activity. Leucine aminopeptidase, which is specific for the amino-terminus, did not destroy activity (TABLE2). TABLE2 ENZYMATIC TREATMENT OF TF-CONTAINING FRACTIONS DLE Fraction * DLE IIIa IIIa IIIa IIIa IIIa IIIa IIIa IIIa IIIa IIIa IIIa, APT IIIa, APT

Enzymatic Treatment t None None Sham, no enzyme Pronase, 2 mg/TFU Pronase, 20 mg/TPU Pronase, 30 mg/TFU+traysylol Proteinase K, 1 mg/TFU Proteinase K, 25 mg/TFU Leucine amino peptidase Carboxy peptidase A Alkaline phosphatase Phosphodiesterase I Phosphodiesterase I1

Dermal Response t x f S.D. (N) 13.9 f 5.3 13.2 rt 2.6 11.6 2 2.0 17, 20 3.8 f 0.5 11, 18 10.0 f 1.0 3.0 2 1.0 14, 15 0, 2 13.4 rt 1.9 4.8 f 0.5 11.3 rt 1.0

(8) (8) (5) (2) (4) (2) (4) (3) (2) (2) (10)

(4) (5)

* DLE=vacuum dialyzed extract; IIIa =Sephadex Fraction IIIa; IIIa, APT= alkaline phosphatase treated Sephadex Fraction IIIa. t Enzyme treatments carried out according to reference 6. $ Skin tests with 100 pg KLH were administered 3-5 days after systemic TF injection and read for diameter of induration at 24 hr in a blinded protocol.

Burger et al.: Human Transfer Factor

239

Since previous work suggested that endonucleases did not destroy TF act i ~ i t y , ~we , selected two exonucleases for further investigation. Snake venom phosphodiesterase I is a 3’ exonuclease that requires a free 3‘ hydroxyl and yields mononucleoside-5’-phosphates. Bovine spleen phosphodiesterase I1 is a 5’ exonuclease that requires a free 5‘ hydroxyl and yields mononucleoside-3’-phosphates. Phosphodiesterase I destroyed activity, whereas phosphodiesterase I1 did not. Contamination of phosphodiesterase I with 5’ nucleotidases was less than 0.001% as judged by cleavage of AMP to adenosine. Experiments are in progress to identify the mononucleoside 5’-phosphates generated during the snake venom destruction. These experiments suggest that TF possesses a peptide component, and a phosphodiester linkage to a moiety with a free 3’ hydroxyl. Retention characteristics on HPRP chromatography and behavior to alkaline phosphatase treatment are consistent with an additional phosphate residue in TF. One model consistent with these characteristics and data from other laboratories is presented in FIGURE1. The proposed structure represents a working model which allows design of additional definitive experiments.

-0

-

-0

-

-

Phorphodirrtrrarr I

Phorphddiestrrorr II

-0

-

Pronatr and Protrinarr K

‘-O-P4 0

Alkalinr Phorphataro

PO,

6 Destruction of

T F Activity Altrration in T F Bohovior on Chromatography T F Rrristont to Trratrnrnt

FIGURE 1. Structural model for transfer factor.

240

Annals New York Academy of Sciences REFERENCES

1. VANDENBARK, A. A., D. R. BURGER,D. L. DREYER,G. D. DAVES& R. M. VETTO. 1977. Human Transfer Factor: Fractionation by electrofocusing and high pressure, reverse phase chromatography. J. Immunol. 118: 636. 2. BURGER, D. R., A. A. VANDENBARK, P. FINKE & R. M. VETTO. 1977. De novo appearance of KLH transfer factor following immunization. Cell. Immunol. 29: 410. H. S. 1974. The Harvey Lectures, Series 68. Academic Press. New 3. LAWRENCE, York. 4. BURGER,D. R., A. A. VANDENBARK, W. DUNNICK,W. G. KRAYBILL & R. M. VETTO. 1978. Properties of human transfer factor from KLH-immunized donors: Dissociation of dermal transfer and proliferation augmenting activities. J. Reticuloendothel. SOC.24: 359. 5. KIRKPATRICK, C. H. Transfer of delayed cutaneous hypersensitivity with transfer factor. Cell Immunol. 41: 62. 6. BURGER,D. R.,A. A. VANDENBARK, W.DUNNICK,W. KRAYBILL, G. D. DAVES& R. M. VETTO. 1979. Human Transfer Factor: Structural properties suggested by HPRP chromatography and enzymatic sensitivities. J. Immunol. 122: 1091. G. D. DAVES,W. A. ANDERSON, JR., R. M. 7. BURGER,D. R., A. A. VANDENBARK, VEITO & P. FINKE.1976. Human transfer factor: Fractionation and biologic activity. J. Immunol. 117: 789. J. A., N. CHRISTMS,E. N. METZ,S. P. BALCERAK & A. F. Lo BUGLIO. 8. NEIDHART, 1978. Skin test conversion following transfer factor; a double-blind study of normal individuals. J. Allergy Clin. Immunol. 61: 115. H. S. 1969. Transfer Factor. In Advances in Immunology. Vol. 11. 3. LAWRENCE, F. S. Dixon, Jr. and H. G. Kunkel, Eds. : 239-350. Academic Press. New York. C. H. & T. K. SMITH.1977. In Regulatory mechanisms in lympho10. KIRKPATRICK, cyte activation. D. 0.Lucas, Ed. : 174-188. Academic Press. New York. 1976. Human transfer factor 11. WILSON,G. B., T. M. WELCH& H. H. FUDENBERO. in guinea pigs: Partial purification of the active component. In Transfer Factor: Basic Properties and Clinical Applications. M. S. Ascher, C. H. Kirkpatrick, and A. A. Gottlieb, Eds. : 409-423. Academic Press. New York. W. & F. H. BACH.1976. Guinea Pig “Transfer Factor” in vitro, Physio 12. DUNNICK, co-chemical properties and partial purification. I n Transfer Factor: Basic Properties and Clinical Applications. M. S. Ascher, C. H. Kirkpatrick, and A. A. Gottlieb, Eds. : 185-195. Academic Press. New York. & M. DINOWITZ. 1977. Transfer of 13. RIFKIND,D., J. A. FREY, A. PETERSEN delayed hypersensitivity in mice to microbial antigens with dialyzable transfer factor. Infect. Immunol. 16:258. 1973. Studies on 14. SPITLER,L. E., D. WEBB,C. VON MULLER& H. H. FUDENBERG. the characterization of transfer factor. J. Clin. Invest. 52: 80a.