INFECTION AND IMMUNrrY, Nov. 1992, p. 4973-4975
Vol. 60, No. 11
Copyright X 1992, American Society for Microbiology
Human Neutrophil Azurocidin Synergizes with Leukocyte Elastase and Cathepsin G in the Killing of
Capnocytophaga sputigena KENNETH T. MIYASAKI* AND AMY L. BODEAU Section of Oral Biology and Dental Research Institute, UCLA School ofDentistry, Los Angeles, California 90024-1668
Received 19 May 1992/Accepted 31 August 1992
Azurocidin was purified in the presence of phenylmethylsulfonyl fluoride. Electrophoresis revealed at least seven species which exhibited N-terminal sequences consistent with azurocidin. Azurocidin exhibited no bactericidal activity against Capnocytophaga sputigena or other oral bacteria but synergized the bactericidal activity of enzymaticaily active elastase. Azurocidin also interacted synergistically with cathepsin G. The neutrophil neutral serine proteases (NSP) are a family of antimicrobial glycoproteins. Four members of the NSP family have been identified in neutrophils: cathepsin G, leukocyte elastase, p29b, and azurocidin and the related or identical molecule CAP37 (azurocidin/CAP37) (3, 8, 9, 12, 16). The natural substrates for the enzymatically active NSP are unknown (15). The bactericidal function of one NSP, cathepsin G, probably resides within discrete peptide domains (1). Reminiscent of most granzymes from cytotoxic T cells (4), azurocidin is enzymatically inert, and its catalytic serine is replaced by glycine (16). CAP37, a molecule which exhibits amino-terminal-sequence homology with azurocidin (first 20 amino acids), is also enzymatically inert and possesses a serine substitution for the catalytic histidine (3, 9). Unlike leukocyte elastase, cathepsin G, and p29b, neither azurocidin nor CAP37 binds diisopropyl fluorophosphate (3, 9). Both cathepsin G and elastase kill Capnocytophaga spp. (5). The killing of Capnocytophaga spp. by cathepsin G is dependent upon an intact enzyme active site, whereas the killing of most nonoral bacteria is not (5, 6). Recently, it has been observed that the killing of Pseudomonas aeruginosa is also dependent upon an intact enzyme active site, indicating that the enzyme-dependent mechanism is more widespread than previously appreciated (14). Both enzyme-dependent and enzyme-independent bactericidal activities are inhibited by plasma antiproteases (6). Azurocidin/CAP37 exhibits potent antimicrobial activities against nonoral bacteria (3, 12, 16), and we wanted to determine whether it exerted bactericidal effects against periodontal bacteria. The preparation of human neutrophils, granules, and granule extracts, the fractionation of granule extracts by gel filtration on Sephadex G-100 (Pharmacia-LKB Biotechnology, Piscataway, N.J.), the subfractionation of elastasecontaining Sephadex G-100 fraction C (which was microbicidal) into subfractions CO through C5 by cation-exchange chromatography on a Mono-S HR 5/5 column (PharmaciaLKB) (pH 4.7), and cationic polyacrylamide gel electrophoresis (CAT-PAGE) were performed as previously described (7). Subfraction C5 exhibited microbicidal activity and contained elastase, azurocidin, and unidentified components (7). Subfraction C5 (total protein, 1,925 ,ug, as assessed by the *
Coomassie dye-binding method of Bradford ) was refractionated by cation-exchange chromatography by using Mono-S equilibrated with 50 mM HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid) (pH 8.2) and eluted with a concave 0 to 2 M NaCl gradient (0.5 ml/min). For this purification, a portion of Sephadex G-100 fraction C and Mono-S subfraction C5 were inactivated with phenylmethylsulfonyl fluoride (Sigma Chemical Co., St. Louis, Mo.) (20 pg/ml). Three asymmetric peak areas were obtained, and they were designated CSA through C5E as shown in Fig. 1. Pooled C5A and C5B contained less than 2% (38 ,g) of the total protein. Pooled CSC, the major peak obtained byA220, contained 37% (718 ,ug) of the total protein and was identified as elastase by both CAT-PAGE and N-terminal-sequence criteria. The remaining 61% of protein in subfraction C5 was contained within C5D and CSE, which accounted for 389 and 781 p,g of protein, respectively. C5D and C5E were analyzed by CAT-PAGE (Fig. 2). Seven Coomassie-stainable bands were revealed, four in C5E and six in CSD (three bands were shared by both pooled fractions). Amino acid composition analysis of pooled fractions CSD and CSE revealed negligible tyrosine residues, consistent with their low A2w values relative to that of the pooled fraction C5C. The four major bands (bands 1 through 4) and a pool of the three minor bands (bands 5 to 7) were electroeluted by presoaking the bands for 3 h in sodium dodecyl sulfate (SDS)-2.1% PAGE-20 mM ammonium bicarbonate and applying a current (10 mA) for 3 h in 0.1% SDS-20 mM ammonium bicarbonate by using a model 1750 electroelution tank (ISCO Inc., Lincoln, Nebr.). N-terminal amino acid analysis was performed by automated Edman degradation (Porton Instruments, Inc., Tarzana, Calif.). All bands exhibited N-terminal sequences distinct from other neutrophil NSP (elastase, cathepsin G, and p29b) and identical to the sequences reported for azurocidin (Table 1). Pooled fraction CSE exhibited a molecular mass range of 25 to 29 kDa, and CSD exhibited a higher range of molecular mass (28 to 36 kDa), as assessed by SDS-PAGE (data not shown). Azurocidin is known to consist of at least three glycoisomers (16). The size heterogeneity we observed would explain the reported disparity between the molecular weights of azurocidin and CAP37 (3, 9, 16). Also, it may explain why CAP37 is reportedly a fairly minor azurophil granule protein but azurocidin is a major azurophil granule protein (3, 11). CAP37 is probably a purified high-molecular-
Corresponding author. 4973
TABLE 1. Amino-terminal sequences of pooled fractions C5D and C5E compared with those of elastase, p29b, and cathepsin GI
C5D 0.0 0
1. Purification of azurocidin from subfraction CS with a
Mono-S HR 5/5
column equilibrated with
50 mM HEPES, pH 8.2.
Azurocidin eluted as a complex peak which was more cationic than elastase.
Because of the relatively
azurocidin, the 280-nm peak heights are deceptively low compared with those of elastase.
Leukocyte elastase AGP7 (p29b) Cathepsin G Azurocidin
1 2 3 4 5-7
IVGGBR IVGGR IVGGR IVGGBR IVGGR
KARPR QFPFL ABIQN KAR KAR KARPR QFPFL KAR
IVGGR IVGGH IIGGR IVGGR
HARPH EAQPH ESRPH KARPR
AWPFM BRPYM SBRPYM QFPFL
VSLQL ASLEM AYLQI ABIQN
a Leukocyte elastase, p29b, and cathepsin G sequences were determined by Salvesen et al. (10), Sinha et al. (13), and Wilde et al. (16). bAmino acids: A, alanine; E, glutamic acid; F, phenylalanine; G, glycine; H, histidine; I, isoleucine; K, lysine; L, leucine; M, methionine; N, asparagine; P, proline; Q, glutamine; R, arginine; S, serine; V, valine; W, tryptophan; Y, tyrosine.
weight isomer of azurocidin. C5E was used throughout the
remaining experiments as azurocidin. Bactericidal scribed
(5) by using ionic conditions which we believed
would reflect those of the initial, alkalinized phagolysosome and the late-stage, mildly acidified phagolysosome (11). In
preliminary 0.0 studies,
bactericidal activity of the azurocidin- and elastase-contain-
at pH 6.0, 7.0, and 8.0. Optimal killing
was observed at pH 8.0, and for this reason, we selected pH 8.0 for initial screening of purified azurocidin. Screenings at pH 8.0 revealed no antimicrobial activity of azurocidin (100
ATCC 29523, NCTC 9709, or FDC-Y4; Haemophilus aph-
of13252; azurocidin~~~~~~. ... C5 corrodens a h sdtruhu rophilusisomer ATCC Eikenella ATCC 23834;
exermnt':;zuoidn remaining~~~~~~~~ wer Bactercidalassay Eschernchia pefre coili
Capnocytophaga sp. strain ATCC 33123, ATCC 33124, or ATCC 27872; or
ML-35. Because it was
FIG. CAT-PiaGEooaalysoisdof proole reported that azurocidin and CAP37 kill fractionso optimallyC5D at aandtC5E mildly
acidic pH (3,
12), we tested the antimicrobial
andss afterelab ielle
stdes, we preliminary~~~~~~~ c,t usedose aftracktgagent
Scenig .at 2923 D. ATCC~~~~~~~~~~~~~~~. 9709,'orF..p'... Haeophlu ATC Eienll corrodens?:....:.:4 rophilus~~~~~~~~~~~~~~~~~~~~*. fo iniial screnin 8~~~~~~~~~~~~~~~~~~~~~~....
of purified auoin. C50
acidic~~~~~~ pH 3, 12,
we tete :. *'E5
FIG. 2. CAT-PAGE analysis of pooled fractions CSD) and CSE (20 pg per lane). Bands are labelled 1 through 7. CYT C, cytochrome c, used as a tracking agent.
azurocidin against E. coli ML-35, Capnocytophaga sputigena ATCC 33123, and A. actinomycetemcomitans at pH 5.5. Again, no microbicidal effects were observed, although the growth of E. coli was impeded to some extent. Thus, our oral bacteria appear to be somewhat resistant to azurocidin, which is not unusual since Proteus mirabilis, Proteus vulgaris, Serratia marcescens, and various bacilli, staphylococci, and streptococci also are reported to be resistant to CAP37 (12). The above results were somewhat perplexing because our previous studies had demonstrated that subfraction C5 was bactericidal against C. sputigena ATCC 33123 (7). We estimated that subfraction C5 was about two-thirds azurocidin and one-third elastase, and therefore, we suspected that azurocidin may work in concert with elastase (or other NSP) to exert an antimicrobial effect against oral bacteria. Accordingly, we tested the interaction of azurocidin with commercially purified elastase and cathepsin G (Biodesign International, Kennebunkport, Maine). Killing was observed when sublethal concentrations of elastase (10 jig/ml) were combined with azurocidin at a concentration of 100 p,g/ml (Fig. 3A). The interactive antimicrobial effect of elastase and azurocidin was reproducible (observed in six of six other experiments) and could be blocked by a2-macroglobulin (Sigma Chemical Co.) and the elastase-specific inhibitor
N-methoxysuccinyl-alanylalanylprolylvalyl chloromethyl ketone (Sigma) at 10 p,g/ml, a concentration which inhibits at least 98% of enzyme activity after 15 min (data not shown). The synergistic antimicrobial activity of elastase and azurocidin was also blocked by concentrations of NaCl greater than 20 mM. A similar synergistic interaction was observed when cathepsin G (10 p,g/ml) and azurocidin (100 p,g/ml) were combined (Fig. 3B). These studies reveal an enzyme-dependent interaction among the neutrophil NSP. Although speculative, several possible explanations exist. We hypothesize that the NSP serve as their own substrates (rather than operating enzymatically upon vital bacterial targets). We favor this explanation because autoreactivity is pervasive in immune and hematologic systems, and this is particularly true for the serine proteases. For example, serum complement is a host defense system in which seine proteases interact with other host molecules (including other serine proteases) in order to exert an antimicrobial effect. Enzymatic activities of the
VOL. 60, 1992 4.0-
100 igmNl azurocikin 10 pg/mi azurocidin I ;Lgirrd azwocidin 0 glr azurocicin
< 1.5 S
C. sputigena ATCC 33123
[CATHEPSIN GJ. ILglnil
FIG. 3. Synergistic antimicrobial effects of azurocidin with elastase (A) and cathepsin G (B) (pH 8.0; 37'C; 2-h incubation) against C. sputigena ATCC 33123. Values are means of quadruplicate assays. Bars, standard deviations.
neutrophil NSP may expose sequestered antimicrobial domains which are sequestered because they are toxic to host tissues. We are presently trying to determine the nature of (i) the enzyme-dependent microbicidal activities of the NSP and (ii) the interaction between azurocidin and other neutrophil NSP. This study was supported by a US-PHS grant from NIH-NIDR (DE08161). K.T.M. is a recipient of a Research Career Development award DE00282 from the NIDR. We acknowledge the skillful amino-terminal-sequence determinations performed by Audree V. Fowler and Richard Stevens of the UCLA Protein Microsequencing Service, aided by NIH BRS Shared Instrument grant SlORR05554.
REFERENCES 1. Bangalore, N., J. Travis, V. C. Onunka, J. Pohl, and W. M. Shafer. 1990. Identification of the primary antimicrobial domains in human neutrophil cathepsin G. J. Biol. Chem. 265: 13584-13588. 2. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 7:248-254. 3. Campanelli, D., P. A. Detmers, C. F. Nathan, and J. E. Gabay. 1990. Azurocidin and a homologous serine protease from neutrophils: differential antimicrobial and proteolytic properties. J. Clin. Invest. 85:904-915. 4. Jenne, D. E., and J. Tschopp. 1988. Granzymes, a family of serine proteases released from the granules of cytolytic T lymphocytes upon T cell receptor stimulation. Immunol. Rev. 103:53-71. 5. Miyasaki, K. T., and A. L. Bodeau. 1991. In vitro killing of oral Capnocytophaga by granule fractions of human neutrophils is associated with cathepsin G activity. J. Clin. Invest. 87:1585-
1593. 6. Miyasaki, K. T., and A. L. Bodeau. 1991. In vitro killing of Actinobacillus actinomycetemcomitans and Capnocytophaga
spp. by human neutrophil cathepsin G and elastase. Infect. Immun. 59:3015-3020. Miyasaki, K T., A. L. Bodeau, and T. F. Flemmig. 1991. Differential killing of Actinobacillus actinomycetemcomitans and Capnocytophaga spp. by human neutrophil granule components. Infect. Immun. 59:3760-3767. Niles, J. L., R. T. McCluskey, M. F. Ahmad, and M. A. Arnaout. 1989. Wegener's granulomatosis autoantigen is a novel serine protease. Blood 74:1888-1893. Pereira, A. H., W. M. Shafer, J. Pohl, L. E. Martin, and J. K. Spitznagel. 1990. CAP 37, a human neutrophil-derived chemotactic factor with monocyte specific activity. J. Clin. Invest. 85:1468-1476. Salvesen, G., D. Farley, J. Shuman, A. Przybyla, C. Reily, and J. Travis. 1987. Molecular cloning of human cathepsin G: structural similarity to mast cell and cytotoxic T lymphocyte proteinases. Biochemistry 26:2289-2293. Segal, A. W., M. Geisow, R. Garcia, A. Harper, and R. Miller. 1981. The respiratory burst of phagocytic cells is associated with a rise in vacuolar pH. Nature (London) 290:406-409. Shafer, W. M., L. E. Martin, and J. K. Spitznagel. 1984. Cationic antimicrobial proteins isolated from human neutrophil granulocytes in the presence of diisopropyl fluorophosphate. Infect. Immun. 45:29-35. Sinha, S., W. Watorek, S. Karr, J. Giles, W. Bode, and J. Travis. 1987. Primary structure of human neutrophil elastase. Proc. Natl. Acad. Sci. USA 84:2228-2232. Wasiluk, K. R., K. M. Skubitz, and B. H. Gray. 1991. Comparison of granule proteins from human polymorphonuclear leukocytes which are bactericidal toward Pseudomonas aeruginosa. Infect. Immun. 59:4193-4200. Watorek, W., D. Farley, G. Salvesen, and J. Travis. 1988. Neutrophil elastase and cathepsin G: structure, function, and biological control. Adv. Exp. Med. Biol. 240:23-31. Wilde, C. G., J. L. Snable, J. E. Griffith, and R. W. Scott. 1990. Characterization of two azurophil granule proteases with activesite homology to neutrophil elastase. J. Biol. Chem. 265:20382041.