Enzymatically Active Papain Preferentially Inducesan Allergic Response in Mice

Louise Chambers,* Alan Brown,† David I. Pritchard,† Suneal Sreedharan,‡Keith Brocklehurst,‡ and Noor A. Kalsheker*,1

*Division of Clinical Chemistry, School of Clinical Laboratory Sciences, University Hospital, Nottingham NG7 2UH;†School of Biological Sciences, University of Nottingham, Nottingham NG7 2RD; and ‡Laboratory of Structuraland Mechanistic Enzymology, Division of Biomedical Sciences, Queen Mary and Westfield College,University of London, Mile End Road, London E1 4NS, England

Received November 6, 1998

Human exposure to papain, a cysteine proteinase, isassociated with hypersensitivity reactions. We demon-strate in mice that enzymatically active papain pref-erentially induces an IgG1 response and results inmast cell degranulation, both features typical of anallergic reaction. Inactive papain, blocked with E-64,appears to desensitize mice to subsequent challengeby active enzyme. These results suggest that the enzy-matic activity of papain is critical in inducing an al-lergic response and that the use of inactive allergensmay be a possible strategy for desensitizing allergicindividuals. © 1998 Academic Press

The allergic response is characterized by an increasein circulating immunoglobulin E (IgE) to allergens andis supported mainly by cytokines and leukocytes asso-ciated with Type-2 (Th2) responses (1). Cross-linking ofreceptor-bound IgE by antigen results in the release ofpharmacologically active agents from mast cells thatgive rise to the symptoms associated with immediateand delayed hypersensitivity reactions.

There is growing body of evidence to suggest thatenzymatic activity is vital to the establishment of theTh2 milieu. This could be achieved by a number ofcomplementary enzymatic effects on the cells of theimmune system; for example by (i) directly inducingTh2 cytokine production by cells of the mast and baso-phil series (1) and (ii) cleavage of cell surface regula-tory molecules such as CD23 (2, 3). Many allergens areenzymes with proteolytic activity (4). As an indicationof the allergenicity of proteinases, workers exposed topapain, a cysteine proteinase, are at significant risk ofdeveloping allergic hypersensitivity (5) and there is asuggestion that this risk may spread to the consumer

(6). Papain is widely used in the food industry as ameat tenderizer, a beer clarification agent and in manypharmaceutical products (7). However, it has yet to beestablished whether the exposure of the immune sys-tem to inactivated allergenic enzymes undermines thisprocess, effectively leading to desensitization.

A previous report suggested that immunization ofmice with active papain may be associated with a Th2response (8) but no information was provided withregard to the characterization of the enzyme, the vali-dation of its purity, or the differential levels of antibodyresponse seen. In the present report, inactivated pa-pain seemed to reverse the immunological profile to-wards a Th1 rather than a Th2 response. A major housedust mite allergen, Der p 1, another cysteine protein-ase, is closely related to papain, and shares criticalfeatures, including conservation of the active site andsimilar three-dimensional structures (9, 10). It seemsplausible therefore, that both enzymes may share com-mon mechanisms capable of inducing an allergic re-sponse.

To test the possibility that vaccination with an inac-tivated model cysteinyl proteinase and allergen repre-sents a valid desensitization process, we have resortedto investigating the role of proteinase activity in vivoby comparing the effects of immunizing mice withenzymatically active papain that has been purified ex-tensively, and totally inactive papain formed by irre-versibly binding the cysteine proteinase inhibitortrans-epoxysuccinyl-L-leucylamido (4-guanidino)-butane(E-64) to the active site. Thus the two forms of papainare structurally similar but differ only in enzyme ac-tivity. In addition, the immunizations were performedin an adjuvant free environment, in order to adminis-ter the enzyme as close to its native state as possible.

We now demonstrate that a low dose of enzymati-cally active papain preferentially induces an IgG1 re-sponse in mice, an accepted surrogate isotype for the

1 To whom correspondence should be addressed. Fax: 01159709167.E-mail: noor.kalsheker@nottingham.ac.uk.

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS 253, 837–840 (1998)ARTICLE NO. RC989862

837 0006-291X/98 $25.00Copyright © 1998 by Academic PressAll rights of reproduction in any form reserved.

Th2 response, and results in increased mast cell de-granulation on re-exposure to the allergen. Further-more, mice immunized with inactive papain, and sub-sequently challenged with active enzyme, demonstratea reduction in the number of degranulated mast cells,suggesting that the use of inactive allergens may be apossible general strategy for desensitizing allergic in-dividuals.

MATERIALS AND METHODS

Materials. The fluorogenic substrate Cbz-Phe-Arg-AMC.HCl waspurchased from Calbiochem-Novabiochem Ltd., Beeston, Nottingham.Two-times crystallized papain, trans-epoxysuccinyl-L-leucylamido(4-guanidino)-butane (E64), N-benzoyl-L-arginine p-nitroanilide (L-BAPNA), 2,29-dipyridyl disulfide (2PDS) and n-t-glycine-p-Nitrophenylester (Boc-gly-pNPE) were purchased from Sigma–Aldrich Co. Ltd.,Poole, Dorset. Prepacked PD10 columns and SP-Sepharose was pur-chased from Pharmacia Biotech, St. Alban’s, Herts. Antibodies werepurchased from Seralab, Crawley Down, Sussex, UK, diluted 1:1000 inblocking agent).

Papain purification and enzyme activity. Papain was purifiedfrom papaya latex by FPLC using cation exchange chromatographyon SP-Sepharose. Elution was performed using linear salt gradientfrom 0.1 to 1 M Na1 in 0.1 M sodium acetate buffer pH 5.0, contain-ing 1 mM EDTA. The fractions of papain were pooled and reversiblyblocked with excess 2PDS and stored at 220°C until further use. Thepurity of the papain was by assessed using tricine–sodium dodecylsulfate–polyacrylamide gel electrophoresis (SDS–PAGE) as de-scribed (11) and stained with Coomassie brilliant blue R250. Theregeneration of active papain was achieved by mixing with 20 mMcysteine at pH 8.0, at room temperature for 20 min. The mixture wasthen applied to a 9-ml PD10 desalting column preequilibrated with0.1 M KCl and 1 mM EDTA to remove cysteine and 2PDS. Proteinfractions of 1 ml were collected and the absorbance peak at 280 nmcontaining activated papain. The concentration of papain was deter-mined using the extinction coefficient (e 5 56,000 M21 cm21) and themolecular weight of papain (Mr 23,350).

Proteolytic activity was assessed by monitoring the release of7-amino-4-methylcoumarin (AMC) from the synthetic peptide Cbz-Phe-Arg-AMC.HCl as previously described (12). The release of AMCresulting from the hydrolysis of 180 mM Cbz-Phe-Arg-AMC in 10 mMcysteine, 0.5 M Tris–HCl, pH 8.0, 2% DMSO was monitored at roomtemperature using a Hoeffer TK 100 mini-fluorometer (excitation365 nm, emission 465 nm) over a 2 minute period. All reactions werecarried out in 2 ml volumes and initiated by the addition of sub-strate. The irreversible cysteine proteinase inhibitor E64 was addedfrom a stock of 10 mM dissolved in distilled water and added to theactive papain until proteolytic hydrolysis was undetectable. An equalvolume of distilled water was added to the active papain to ensureequal dilution of the protein content in both samples.

Immunization protocol. Initially in a pilot experiment, 5- to6-week-old male Balb/c mice (Harlan Olac Ltd., Bicester, Oxon) weresplit into 3 groups (2 mice in each group). Prior to immunization allmice were bled from the tail vein. Individual groups were immunizedsubcutaneously with either 10 mg of active or inactive papain. Acontrol group was immunized with storage buffer (1 mM KCl, 100mM EDTA) alone. On days 14 and 28 the mice were bled from the tailvein and immunized again with the same quantities of papain. 14days after the final immunization (day 42) the mice were sacrificed,bled from the thorax and the spleen removed for analysis of cytokinesecretion. Splenocytes were harvested and cultured in vitro as de-scribed by (13) and then stimulated with Concanavalin A (5 mg/ml)for 48 h at 37°C in an atmosphere supplemented with 5% CO2.Concentrations of interleukin-5 (IL-5) and interferon-g (IFN-g) in

culture supernatants were determined in triplicate by a captureELISA using a previously described method (14).

In a more extensive confirmatory experiment, 48 male Balb/c micewere split into four groups containing 12 mice in each group. Prior toimmunization the mice were bled from the tail vein. Groups A and Bwere immunized with 1 mg of active papain while groups C and Dreceived 1 mg of inactive papain. On day 14, six mice from each groupwere bled from the tail vein. Following this, 1 mg of active papain(Groups A and C) or inactive papain (Groups B and D) was injectedinto the peritoneal cavity. After 15 min the mice were sacrificed and5 ml of Hanks buffered saline (HBSS) injected into the peritonealcavity. The peritoneum was massaged gently before being openedand approximately 3 ml of the HBSS (containing peritoneal mastcells) removed into a sterile universal container. The cells wereharvested by centrifugation (600g, 10 min.) and resuspended in 200ml RPMI (Sigma). 100 ml of cell suspension was applied to a micro-scope slide using a cytospin (970 rpm for 2 min). The slides werestained with Giemsa (1:20) and the percentage of degranulated mastcells was determined by microscopic analysis following the observa-tion of a minimum of 200 cells.

The remaining mice received a further immunization of 1 mg ofactive and inactive papain respectively. On day 28 the remainingmice were bled, and challenged intraperitoneally with active andinactive papain as described earlier.

Blood taken from the tail vein or thoracic cavity was allowed to clotand then centrifuged at 13,000g for 15 minutes. The plasma wascollected into fresh polypropylene tubes and stored at 220°C untilrequired.

ELISA for IgG1, IgG2a. Purified papain was diluted to 5 mg/ml in0.05 M sodium carbonate/bicarbonate buffer pH 9.6. Fifty microlitersof this solution was used to coat the wells of a 96-well polystyreneplate overnight at 4°C. The plate was then washed 3 times with Trisbuffered saline, 0.05 % Tween 20, pH 7.4 (TBS/Tween) and the wellsblocked with 200 ml of 5% Marvel milk powder/TBS (blocking agent)for 1 h at 37°C. The plate was washed as before and 50 ml of mouseserum (diluted 1:50 in blocking agent) was added to individual wellsand the plate incubated at 4°C overnight. After this period, the platewas washed again and 50 ml of sheep anti mouse IgG1 or IgG2a-alkaline phosphatase was added to individual wells and the plateincubated for a further 90 min at 37°C. The plates were washedagain and antibody binding visualized by the addition of 100 ml ofp-nitro-phenyl-phosphate substrate (1 mg/ml in 0.1 M glycine, 1 mMcalcium chloride, 1 mM magnesium chloride, pH 10.4). All assayswere carried out in triplicate. ELISA values are expressed as themean absorbance at 405 nm 6 SEM after the subtraction of anegative buffer control.

RESULTS AND DISCUSSION

The FPLC traces with corresponding substrate ac-tivities are shown in Fig. 1a. The single peak obtainedby FPLC was identified as a single band on SDS PAGE(Fig. 1b). The mean specific activity of the papain usedin the immunizations against the substrate CbZ-Phe-Arg-AMC.HCl was 50 nmol 6 5.85 (SEM) of AMCreleased mg21 s21 and was undetectable for inacti-vated papain. In the initial experiment immunizationwith 10 mg of active papain resulted in a differentialIgG1 response at 28 days (following two immuniza-tions) compared with inactive papain. This differentialresponse was almost lost by 42 days (results notshown). In contrast there was no significant IgG2aresponse. When the experiment was repeated with alower dose of papain (1 mg) a significant differential

Vol. 253, No. 3, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

838

IgG1 response was seen on day 28 between mice im-munized with active and inactive papain (Fig. 2).

Analysis of splenocyte culture supernatants forIL-5 and IFN-g concentrations following Con A stim-ulation showed that the group of mice immunizedwith active papain released 115 units/ml of IL-5 and58 units/ml of IFN-g In contrast the mice immunizedwith inactive papain failed to produce IL-5 and only34 units/ml of IFN-g. The production of IL-5 a prin-cipal eosinophil-activating cytokine (15) suggeststhat active papain preferentially induced a Th2 type

response. In mice from group A, immunized andchallenged (on day 28) with active papain showed asignificantly greater number of degranulated mastcells when compared with group D mice (on day 28)that had been immunized and challenged with inac-tive enzyme. Furthermore, in group C prior immuni-zation with inactive enzyme, resulted in reducedmast cell degranulation following subsequent chal-lenge by active enzyme on day 28 (Fig. 3).

These data suggest that the cysteine proteinase pa-pain in mice, appears to be necessary for the inductionof a Th2 response. In the mouse, IgG1 is capable ofmediating antigen dependent airways obstruction, afeature of asthma by a mechanism which is believed tobe similar to induced by IgE in humans (16). Analysisof histamine release from challenged mast cells in con-junction with papain-specific IgE, in addition to a num-ber of variables such as strain of mouse, the protocolsfor antigen sensitization challenge and the routes forimmunization, should be considered when this model isdissected further.

The house dust mite proteinase, Der p 1 is closelyrelated to papain and it has a number of propertieswhich may make it allergenic. Der p 1 has been shownto increase the permeability of the bronchial mucosa,facilitating access to antigen presenting cells (17), cat-alytically inactivating a1-antitrypsin, the major anti-proteinase defense in the lungs (18). Cleavage of thelow affinity receptor for IgE (CD23) by Der p 1 maypotentially result in excess IgE stimulation (2) whilecleavage of the interleukin-2 receptor CD 25 whichmodulates immunoglobulin synthesis (IL-2 is an im-

FIG. 1. (a) Papain was purified from papaya latex by FPLC using cation exchange chromatography on SP-Sepharose and eluting witha linear salt gradient from 0.1 to 1.0 M Na1 ions in 0.1 M sodium acetate buffer, containing 1 mM EDTA. Papain (PAP) and caricain (CAR)are the first and last components of the latex to be eluted. The remaining protein peak situated between these two components consists ofthe chymopapains A, B, and M. The hydrolysis of L-BAPNA and Boc-gly-pNPE by the components are also shown. (b) SDS–PAGE profile ofpapain eluting from SP-Sepharose.

FIG. 2. Balb/c mice (12 mice per group) were immunized subcu-taneously with either 1 mg of active (F) or inactive (E) papain on days0 and 14. On days 0, 14, and 28 mice were bleed from the tail andpapain specific IgG1 levels were determined by ELISA. IgG1 levelsare expressed as the mean absorbances (405 nm) 6 SEM.

Vol. 253, No. 3, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

839

portant stimulus for activated T cells) results in a biastoward a Th2 response (19). Der p 1 has also beenreported to have a direct effect on rodent mast cellstransfected with the a-sub-unit of the human IgE highaffinity receptor, by increasing mediator release inde-pendently of IgE (20). In a guinea pig model proteolyticenzymes, present in detergents, instilled intratrache-ally, have been shown to enhance the allergic re-sponses of guinea pigs to nonproteolytic detergent en-zymes in a mixture (21), thus providing a generalmechanism for increased allergenicity.

In conclusion, we report that the immunization ofenzymatically inactivated allergen can underminethe Th2 process, possibly by desensitization and pre-venting the occurrence of an allergic response. Theimmunological mechanisms by which these areachieved are unknown at present, but it is temptingto speculate that inactivated allergen is prevented

from inducing the Th2 milieu, and instead stimulatescontra-regulatory Th1 pathways.

REFERENCES

1. Sutton, B. J., and Gould, H. J. (1993) Nature 366, 421–428.2. Hewitt, C. R. A., Brown, A. P., Hart, B. J., and Pritchard, D. I.

(1995) J. Exp. Med. 182, 1537–1544.3. Schulz, O., Laing, P., Sewell, H., and Shakib, F. (1995) Eur.

J. Immunol. 25, 3191–3194.4. Robinson, C., Kalsheker, N. A., Srinivasan, N., King, C. M.,

Garrod, D. R., Thompson, P. J., and Stewart, G. A. (1997) Clin.Exp. Allergy 27, 10–21.

5. Baur, X., Konig, G., Bencze, K., and Fruhmann, K. (1982) Clin.Allergy 12, 9–16.

6. Mansfield, L. E., Ting, S., Haverly, R. W., and Yeo, T. J. (1985)Ann. Allergy 55, 541–543.

7. Brocklehurst, K., Baines, B. S., and Kierston, M. P. J. (1981)Topics Enzyme Ferment. Biol. 5, 262–335

8. Finkelman, F. D., and Urban, J. F., Jr. (1992) Parasitol. Today 8,311–314.

9. Topham, C. M., Srinivasan, N., Thorpe, C. J., Overington, J. P.,and Kalsheker, N. A. (1994) Prot. Eng. 7, 869–894.

10. Brocklehurst, K., Watts, A. B., Patel, M., Verma, C., andThomas, E. W. (1998) in Comprehensive Biological Catalysis(Sinnott, M. L., Ed.), Vol. I, pp. 81–423, Academic Press, London.

11. Scagger, H., and Von Jagow, G. (1987) Anal. Biochem. 166,368–379.

12. Barrett, A. J. (1980) Biochem. J. 187, 909–912.13. Robinson, K., Bellaby, T., and Wakelin, D. (1994) Immunology

82, 261–267.14. Else, K., and Grencis, R. K. (1991) Immunology 72, 508–513.15. Wardlaw, A. J., Moqbel, R., and Kay, A. B. (1993) Adv. Immunol.

54, 229–270.16. Miuajima, I., Dombrowicz, D., Martin, T. R., Ravetch, J. V.,

Kinet, J-P., and Galli, S. J. (1997) J. Clin. Invest. 99, 901–914.17. Herbert, C. A., King, C. M., Ring, P. C., Holgate, S. T., Stewart,

G. A., Thompson, R. J., and Robinson, C. (1995) Am. J. Respir.Cell Mol. Biol. 12, 369–378.

18. Kalsheker, N. A., Deam, S., Chambers, L., Sreedharan, S.,Brocklehurst, K., and Lomas, D. A. (1997) Biochem. Biophys.Res. Commun. 221, 59–61.

19. Schulz, O., Sewell, H. F., and Shakib, F. (1998) J. Exp. Med. 187,271–275.

20. Machiado, D. C., Horton, D., Harrop, R., Peachell, P. T., andHelm, B. A. (1996) Eur. J. Immunol. 26, 2972–2980.

21. Sarlo, K., Ritz, H. L., Fletcher, R. E., Schrotel, K. R., and Clark,E. D. (1997) J. Clin. Immunol. 100, 480–487.

FIG. 3. Forty-eight Balb/c mice were split into four groups (12mice per group). Groups A and B were immunized with 1 mg of activepapain. Groups C and D were immunized with 1 mg of inactivepapain. On days 14 and 28, six mice from each group were challengedinterperitoneally with 1 mg of active (Groups A and C) or inactivepapain (Groups B and D) and the number of degranulated mast cellsdetermined. Results are expressed as the percentage of degranulatedmast cells 6 SEM following the observation of a minimum of 200cells. * denotes significant reduction in mast cell degranulation (P #0.05) compared with mice immunized and challenged with activepapain.

Vol. 253, No. 3, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

840