Eur. J. Biochem. 206, 587-593 (1992) FEBS 1992

Monoclonal antibodies against Pseudomonas aeruginosa elastase : A neutralizing antibody which recognizes a conformational epitope related to an active site of elastase Shin-ichi YOKOTA, Hiroshi OHTSUKA and Hiroshi NOGUCHI Biotechnology Laboratory, Takarazuka Research Center, Sumitomo Chemical Co. Ltd, Takarazuka, Hyogo, Japan Received January 23/February 28, 1992

-

EJB 92 0087

We have established seven murine hybridoma cell lines which produce monoclonal antibodies (mAbs) against Pseudomonas aeruginosa elastase. The seven mAbs recognized at least six different epitopes on the elastase molecule. All mAbs inhibited both enzymatic activities of elastase and protease, in which elastin fluorescein and hide powder azure were used as substrates, respectively. One of them, mAb E-4D3, strongly neutralized enzymatic activities of peptidase in which furylacryloyl-glycyl-leucinamidewas used as a substrate, as well as of elastase and protease. In contrast, the other six mAbs did not neutralize peptidase activity at all. The Ki value for furylacryloylglycyl-leucinamide of E-4D3, as well as its Fab fragment, was comparable to those for metalloprotease inhibitors such as phosphoramidon and Zincov inhibitor. The binding of mAb E-4D3 was inhibited by phosphoramidon and Zincov inhibitor, but not by metal chelators such as EDTA and ophenanthroline. A line of evidence suggests that mAb E-4D3 directly interacts with active site and highly neutralizes enzymatic activity of P. aeruginosa elastase. Data of Western blotting and ELISA suggest that mAb E-4D3 is likely to recognize an elastase molecule in a conformation-dependent manner as an epitope. In contrast, the neutralizing activity of the other mAbs against elastase and protease seems to be caused by a low accessibility of an enzyme to insoluble and high-molecular-mass substrates through the binding and steric hindrance of the mAbs to an enzyme.

Pseudomonas aeruginosa is one of causative factors for opportunistic infections. P. aeruginosa produces many extracellular pathogenic factors such as exotoxin A, exoenzyme s, elastase, alkaline protease and phospholipase C [ l , 21. Among them, elastase degrades many kinds of biologically important proteins such as elastin, immunoglobulin, q-protease inhibitor and complement components [2,3], and generates bradykinin [4]. Although the lethality of elastase is not so high in comparison with that of exotoxin A, elastase has been shown to enhance the virulence of P. aeruginosa strains [3, 51. Thus its pathogenicity is now considered to be an ‘aggresin’ activity [3]. Homma et al. [6, 71 have developed a multicomponent vaccine consisting of a common protective antigen (OEP) and toxoids of exotoxin A, elastase and alkaline protease for P. aeruginosa infections. Neutralizing antibodies to P.aeruginosa elastase have been expected to be one of candidates for therapeutics against P. aeruginosa infections. Extensive enzymological and chemical studies on P. aeruginosa elastase have been carried out [3, 8, 91. Elastase is

a member of the group of zinc-metalloproteases which have a zinc ion as an essential component for enzymatic activity [lo]. Recently, the primary structure was deduced from a full nucleotide sequence I l l , 121, and the three-dimensional structure was solved at 15-nm resolution [13]. However, immunochemical studies, especially on the neutralizing mechanism and the epitope of anti-elastase antibodies, have rarely been carried out. Such immunochemical studies should provide useful information on the efficacy of elastase vaccine and anti-elastase antibodies in immunotherapy against P. aeruginosa infections. Only one report [I 41 has been presented so far on the preparation of murine anti(P.aeruginosa elastase) mAbs but no information on elastase neutralizing activity of the mAbs was given in the report. In the present study, we have prepared seven P. aeruginosa elastase-specific murine mAbs and characterized them.

Correspondence to S. Yokota, Biotechnology Laboratory, Takarazuka Research Center, Sumitomo Chemical Co., Ltd., 4-2-1, Takatsukasa, Takarazuka, Hyogo, Japan 665 AhhreviatianJ. Fur-Acr-Gly-Leu-NH2, furylacryloyl-glycyl-leucinamide; HOLeu, N-hydroxy-L-leucine; NaCl/Pi, phosphatebuffered saline Enzymes. Pseudomonas aeruginosa elastase, P. aeruginosa alkaline protease, Vibrio cholerae hemagglutinin/protease, thermolysin, Serratia protease (EC 3.4.24.4); pancreatic elastase (EC 3.4.21.36); leukocyte elastase (EC 3.4.21.37).

Enzyme and its toxoid

~

_

_

MATERIALS AND METHODS Elastase prepared from P. aeruginosa was purchased from Nagase Biochemical Inc. (Osaka, Japan). Two kinds of elastase toxoids were used in this study. Toxoid-I, which was prepared by treatment with formalin [6], was kindly donated from Dr J. Y. Homma (Kitasato Institute, Tokyo, Japan). Toxoid-I1 was prepared according to the method of Morihara and Homma [15] using an irreversible inhibitor, C1CH2COHOLeu-Ala-Gly-NH, [16] (Peptide Institute, Osaka, Japan).

588 P. urruginosa alkaline protease was purchased from Nagase Biochemical Inc. Vibrio cholerae hemagglutinin/protease [I71was kindly donated by Dr R.A. Finkelstein (University of Missouri-Columbia, School of Medicine, Columbia, MS). Srrratia protease, thermolysin and human leukocyte elastase were purchased from Sigma (St. Louis, MO). Porcine pancreatic elastase type I and type I1 were purchased from Advance Biofactures Corp. (Lynbrook, NY).

Peptidase activity was determined by using furylacryloylglycyl-leucinamide (Fur-Acr-Gly-Leu-NHz, purchased from Peptide Institute Inc.) with a slight modification of the procedure of Feder and Schuck [21]. Briefly, 400 p11.25 mM FurAcr-Gly-Leu-NHz in Tris/maleate pH 7.0 containing l mM CaC12, 100 p1 mAb in NaC1/Pi, and 5 p1 elastase (0.1 mg/ ml) were incubated at 37°C for 1 h. After the reaction was terminated by adding EDTA, the decrease in absorbance at 345 nm, which was caused by hydrolysis of the peptide bond, was monitored.

Preparation of mAbs

Each Balb/c mouse (4-week-old female) was subcutaneously injected with 100 pg elastase emulsified in complete Freund's adjuvant, for the first immunization. For booster immunization (four times), 100 pg elastase in incomplete Freund's adjuvant was subcutaneously injected at intervals of 2 weeks. For the last immunization, 100 pg elastase in NaC1/ P, was intravenously injected a month after the last booster immunization. Three days after the last immunization, spleen was removed for cell fusion. The resulting spleen cells were fused with mouse myeloma cells P3X63-Ag8-653 (ATCC CRL 1580) using poly(ethy1ene glycol) according to the standard method [18, 191. Elastase-specific mAbs were screened by two methods, namely ELISA using toxoid-I1 as a coated antigen and neutralizing assay for elastase activity. Neutralizing assay was performed using elastin fluorescein (pass 400 mesh; Elastin Products Co., Pacific, MO) as a substrate; 100 pl hybridoma culture supernatant, 100 p1 elastin fluorescein (10 mg/ml) in 0.1 M Tris/maleate pH 7.0 containing 2 mM CaC12, and 0.5 pg elastase were mixed and incubated at 37°C for 3 h in 96-well inicroplate. After reaction, the plate was centrifuged at 3000 rpm for 10 min, then solubilized fluorescein in the supernatant was measured at an excitation wavelength of 490nm and an emission wavelength of 520nm. The hybridoma cells producing elastase-specific mAbs were cloned by a limiting-dilution method. mAbs were purified from mouse ascitic fluids by proteinA - Cellulofine (Seikagaku Kogyo Corp., Tokyo, Japan) column chromatography using MAPS-I1 protein A binding buffer (pH 9.0) and MAPS-I1 protein A elution buffer (pH 3.0) (Bio-Rad, Richmond, CA). Each mAb was dialyzed against NaCl/P,, and used as a mAb preparation. Assay of enzyme activity

Elastase activity was determined by using elastin fluorescein (pass 400 mesh); 100 p1 elastin fluorescein (10 mg/ml) suspension in 0.1 M Tris/maleate pH 7.0 containing 2 mM CaCl,, 100 pl of various concentrations of mAb solution in NaCl/P,, and 5 p1 elastase (0.1 mg/ml) were incubated at 37°C for 2 h with rapid shaking. The reaction was terminated by adding 10 p1 0.2 M EDTA and the mixture was centrifuged at 10 000 rpm for 10 min. The solubilized fluorescein in the supernatant was measured as described above. Protease activity was determined by a dye-releasing method using hide powder azure (Sigma), basically according to the method of Rinderknecht et al. [20]. Briefly, 400 p1 hide powder azure (10 mg/ml) suspension in 0.1 M sodium phosphate pH 7.0, 100 p1 mAb (1 mgiml), and 2 p1 elastase (0.1 mg/ml) were incubated at 37 "C for 1 h with rapid shaking. After the reaction was terminated by adding EDTA, the mixture was centrifuged at 10 000 rpm for 5 min. The solubilized dye in supernatant was measured by its absorbance at 595 nm.

ELISA

ELISA was performed as described previously [19]. Elastase or elastase toxoid was used as a coated antigen at a concentration of 10 pg/ml in 0.05 M sodium carbonate pH 9.6. Alkaline-phosphatase-conjugated anti-(mouse IgG) antibodies (Kirkegaad & Perry Laboratories Inc., Gaithersburg, MD) and sodium p-nitrophenylphosphate were used as secondary antibody and substrate, respectively. An unrelated mouse mAb A-1C9 (IgG1), recognizing human IgM, was used as a control. SDS/PAGE and Western blotting

SDSjPAGE was performed on a 10 - 20% polyacrylamide gradient gel (Daiichi Chemicals, Tokyo, Japan) according to the method of Laemmli [22]. Western blotting was performed according to the method of Towbin et al. [23] using a polyvinylidene difluoride filter (Millipore, Bedford, MA) as a transfer membrane. Specific binding was detected by using alkaline phosphatase-conjugated anti-(mouse IgG) antibodies and sodium 5-bromo-4-chloro-indoylphosphate/p-nitrotetrazolium blue chloride as second antibody and substrate, respectively. Analysis of antigen-antibody complex by gel filtration

The mode in binding of mAb to elastase was analyzed by identifying antigen-antibody complex formation using Superose 6 gel chromatography. Reaction mixture containing antibody and elastase was incubated at room temperature for 5min and then the mixture was applied to a Superose 6 column (1 x 30 cm; Pharmacia, Uppsala, Sweden) in NaC1/Pi at the flow rate of 0.6 ml/min performed with an FPLC system (Pharmacia). The elution profile was monitored by absorbance at 280 nm. Other materials and methods

Phosphoramidon [24] and ZincovTM inhibitor, 2-(Nhydroxycarboxamido) - 4- methylpentanoyl- L - alanyl- glycinamide [16], were purchased from Peptide Institute and Calbiochem (La Jolla, CA), respectively. The isotype of mAb was determined by using a mouse monoclonal antibody isotyping kit (Amersham Corp., Arlington Heights, IL). The Fab fragment of mAb was prepared by digestion with immobilized papain (Pierce, Rockford, IL), basically according to the standard method [25], and subsequently purified by protein-A Cellulofine and Superose 6 column chromatography. Phosphate-buffered saline (NaCl/Pi, pH 7.2) consists of 8 g NaCl, 0.2 g KCl, 2.9 g NaZHPO4. 12 HzO and 0.2 g KHZPO4/1.

589 Table 1. Binding activity of mAb to P. aevuginosa elastase and its toxoids, determined by ELISA, Western blotting and gel filtration methods. The binding activity of each mAb was assessed by several different methods. Toxoid-I and toxoid-I1 of elastase were prepared by treatment with formalin and C1CH2CO-HOLeu-Ala- Gly-NH2,respectively. Polyclonal refers to rabbit anti-(elastase-toxoid-I) antibodies prepared from antiserum. In ELISA, elastase or its toxoid was used as a solid antigen and binding activity was expressed as absorbance at 405 nm. The results of Western blotting are also shown in Fig. 1 . Binding activity was expressed as follows: strong; + +, medium ; +, weak; -,no binding. In gel filtration analysis, the mixture of antibody and antigen, in a molar ratio 1:1, was applied to a Superose 6 column. Binding activity was expressed as percentage binding calculated from the equation: 100 x { 1-[(free elastase)/(total elastase applied)]}; nd: not determined.

+ + +,

mAb

ELISA native

Western blotting

Gel filtratiuon

toxoid-I1

native

0 0.04 0.22 0.20 0.28 0.26 0.08 nd

toxoid-I1

97 > 91 > 97 > 97 > 97 > 97

%

A405

E-4D3 E.2B1 E-2C1 E-2C4 E-6A5 E-lB2 E-7B5 Polyclonal

toxoid-I

0 1.12 0.66 0.46 1.26 1.30 0.38 nd

> 97 > 97 > 97 > 97 > 97 > 97 > 97 > 97

++ ++t

++ + ++ ++

1 2 3 4 5 6 7 8 9 1 0

RESULTS mAbs 94

Seven hybridoma cell lines producing murine anti-elastase mAbs were established and selected on the basis of both binding to elastase toxoids-I1 and neutralization of elastase activity. Each mAb neutralized elastase activity when elastin fluorescein was used as a substrate. And all mAbs except E4D3 showed positive binding in ELISA with either elastase or toxoid-I1 as a coated antigen. All mAbs were determined to be IgGl ( K ) . Binding of mAb to elastase and its toxoids

Binding activity of the mAbs was tested by three methods, namely ELISA, Western blotting and gel filtration (Table 1 and Fig. 1). While the binding activity of each mAb determined by ELISA and Western blotting was different, the specific binding to elastase was observed for all mAbs in gel filtration. Among these mAbs, the binding of E-4D3 to native elastase was not shown to be positive by either ELISA or Western blotting. Furthermore, E-4D3 did not bind to toxoidI1 determined by gel filtration. In contrast, the mAbs other than E-4D3 showed binding to toxoid-11. Binding activity of all mAbs and polyclonal antibody to toxoid-I, which was treated with formalin and showed reduced antigenicity [15], varied between antibodies. Especially, the binding of E-4D3 and E-7B2 to toxoid-I was completely lost in contrast to their full binding to native elastase in gel filtration. Figure 1 shows the results of Western blotting in the presence of 2-mercaptoethanol. Similar results were obtained in the absence of 2-mercaptoethanol (data not shown). In Western blotting, E-4D3 and E-2C4 gave no positive binding to elastase, whereas the other mAbs showed the specific binding to a molecule with a molecular mass of about 33 kDa, corresponding to elastase. In the blotting with some mAbs as well as polyclonal antibody, some minor bands were observed in the molecular mass region lower than 30 kDa. These bands seemed to be degraded products derived from elastase.

67

+

-

43-

30 20.1

14.4

-c

-

Fig. 1. Western blotting analysis of P. aevuginosa elastase with antielastase mAbs. Lanes 1 and 2 show the molecular mass markers (Pharmacia, values on left in kDa) and the elastase preparation (5 pg) stained by Coomassie brilliant blue, respectively. Lanes 3- 10 show immunoblotting of elastase (1 pg/lane) with: lane 3, E-4D3; lane 4, E-2B1; lane 5, E-2C1; lane 6, E-2C4; lane 7, E-6A5; lane 8, E-7B2; lane 9, E-7B5; lane 10, rabbit anti-(elastase toxoid-I) antiserum.

Epitope analysis

Differences in epitopes recognized by mAbs (epitope mapping) on an elastase molecule were determined by investigating the differential formation of complexes between the mAb and antigen in gel filtration [27]. Elastase and two mAbs were mixed at a molar ratio of 2 : l : l and then applied to a column of Superose 6. Fig. 2 C and D shows the chromatogram of the mixture of two mAbs and elastase. When elastase was mixed with mAbs E-2B1 and E-7B2 (Fig. 2 C), a complex with a molecular mass larger than that of the complex between the homongeneous mAb (E-2B1) and elastase (Fig. 2 B) was

590 E-2B1 E-ijA5

E-705

E-7B2

E-4D3

nnnn

uu E-2C1

Fig. 3. Relative spatial distribution of epitopes (epitope mapping) recognized by anti-elastase mAbs deduced from complex formation between two mAbs via an elastase. The scheme was deduced from the data of Table 2. Overlapping lines denote the proximity of each of the epitopes from the data that the two mAbs cannot bind to an elastase molecule simultaneously. This scheme only shows the configurational relationship between two epitopes but does not indicate the sequential position of each epitope.

P a?

I

I4

I

I

II

I

E-2C4

Retention time (min)

Fig. 2. Detection of complex formation between two mAbs via elastase on Superose 6 column chromatography. (A) mAb E-2B1, (B) mixture of E-2B1 and elastase at a molar ratio of 1 : 1 ; (C) mixture of E-2B1, E-7B2 and elastase at a molar ratio of 1 :I :2; (D) mixture of E-2B1, E-2C1 and elastase at a molar ratio of 1 :1:2. Arrows V,, 1 , 2 and 3 indicate the elution positions of blue dextrdn, the complex of mAb E2B1 and elastase at a molar ratio of 1:1, mAb E-2B1, and elastase, respectively.

The results of the other pairs of mAbs are summarized in Table 2. If the epitopes of a pair of mAbs were independent, the major peak of protein, in which a larger complex would form between two mAbs via elastase, would be eluted at about 22.7 min or earlier. On the other hand, if the two epitopes overlapped or were mutually exclusive because of their proximity, the major peak would be eluted near the position of a complex between an antibody and elastase, namely between 24.6 - 25.2 min. The mutual relationship scheme of difference of the epitopes deduced from the results of Table 2 is shown in Fig. 3. The epitopes for mAbs were estimated to be different from each other except E-2B1 and E-6A5, but one epitope was close to some of the other mAbs except E-4D3. Neutralization of enzyme activity by mAbs

formed. Formation of the larger complex was attributed to the simultaneous binding of two heterogeneous mAbs to an elastase molecule. The result indicates that E-2B1 and E7B2 recognize quite different (independent) epitopes on the elastase. In the case of the pair of mAbs E-2B1 and E-2C1 (Fig. 2 D), however, the elution position of complex was near that of the homogeneous E-2B1 -elastase complex. The epitopes of E-2B1 and E-2C1 were mutually exclusive and not distinguished from each other in this experiment because the two mAbs were not able to bind simultaneously to an elaslase molecule, probably due to steric hindrance. The result indicates proximity or overlapping of the epitopes of E-2B1 and E-2C1.

Neutralization of enzymatic activity of elastase by mAbs was performed using several substrates (Table 3). All mAbs neutralized both elastase and protease activity assayed using elastin fluorescein and hide powder azure as substrates, respectively, whereas unrelated mAb A-1C9 used as a control did not show any neutralizing activity. Larger amounts of mAb were required for neutralizing enzyme activity in the experiment using hide powder azure than that using elastin fluorescein. E-4D3 showed the most potent neutralizing activity against elastase and protease. In the peptidase assay using Fur-Acr-Gly-Leu-NH2 as a substrate, only E-4D3 showed neutralizing activity. The other mAbs had no influence on peptidase activity, including the control mAb A-1C9.

Table 2. Gel filtration analysis of the complex formation between elastase and pairs of anti-elastase mAbs. Two different mAbs and elastase were mixed in a molar ratio 1 : 1 :2, incubated, and then applied to a column of Superose 6 as described in Materials and Methods. Values represent mean retention time of the main peak eluted from the Superose 6 column in triplicate experiments. E-4D3 and complexes of E-4D3 and elastase (in a molar ratio 1 : 1) were eluted at 25.7 min and 24.8 min, respectively. An asterisk (*) indicates that a larger complex, which was attributable to two mAbs binding to an elastase molecule simultaneously, was formed. The complex formation indicates that the two mAbs recognize independent epitopes each other. mAb

Retention time of main peak eluted in the presence of mAb E-4D3

E-2B1

E-2C1

E-2C4

E-6A5

E-7B2

E-7B5

21.9* 24.6

19.7* 24.1 24.1

22.2* 22.4* 22.6* 24.8

22.1* 25.0 24.7 22.6* 24.6

22.3* 22.3* 22.4* 25.0 22.4* 24.6

18.9* 22.1* 25.2 24.6 22.4* 22.5* 24.1

min b4D3 E-2B1 E-2C1 E-2C4 E-6A5 E-1B2 ~.-7n5

24.6

591 A

12

3

I1

I

D

ll

I

mAb (pglml)

Fig. 4. Neutralization of enzymatic activity of P. aeruginosa elastase ( O ) ,thermolysin (0) and V . cholevae hemagglutinin/protease ( A ) by mAb E-4D3. Neutralizing activity of mAb E-4D3 for peptidase activity using Fur-Acr-Gly-Leu-NH2 as a substrate was tested as described in Materials and Methods. Briefly, a 500-pl reaction mixture containing 1 mM Fur-Acr-Gly-Leu-NH,, various concentrations of mAb E-4D3 and enzyme (0.5 pg elastase, 0.2 pg thermolysin or 0.7 pg hemagglutinin/protease) was incubated at 37 "C for 1 h. Neutralizing activity was expressed as residual activity in the presence of mAb E4D3. Addition of an unrelated mAb did not affect the peptidase activity of any protease.

Table 3. Neutralizing activity of mAb against P. aeruginosa elastase. Neutralizing activity of each mAb was determined by using elastin fluorescein, hide powder azure and Fur-Acr-Gly-Leu-NH2 as a substrate for elastase, protease and peptidase activity, respectively. Neutralizing activity was expressed as a concentration required for 50% inhibition of enzyme activity (ID5o).In the case of hide powder azure, neutralizing activity was expressed as a percentage of the value when 200 pg/ml mAb was added in the reaction mixture. Detailed assay conditions are described in Materials and Methods. Polyclonal refers to rabbit anti-(elastase-toxoid-I) antibodies. A-1C9 is an unrelated mouse mAb (IgC,); nd, not determined. mAb

mAb concn giving 50% inhibition of elastase

Neutralizing activity of protease

peptidase

Fg/ml

%

~

E-4D3 E-4D3 Fab E-2B1 E-2C1 E-2C4 E-2C4 Fab E-6A5 E-7B2 E-7B5 Polyclonal A-1C9

3.2 2.6 7.1 16.3 4.9 6.8 10.8 14.2 7.4 68 > 1000

9.6 6.6 > 2000 > 2000 > 2000 > 2000 > 2000 > 2000 > 2000 230 > 2000

68 nd 45 21 36 nd 27 26 31 nd