Modulation of the In Vitro Candidacidal Activity of Human Neutrophil
Defensins by Target Cell Metabolism and Divalent Cations
Robert 1. Lehrer,*" Tomas Ganz,*1 Dorothy Szklarek,* and Michael E. Selstedlt
Departments ofMedicine* and Pathology,"I and Will Rogers Institute Pulmonary Research Laboratory, UCLA School ofMedicine,
University ofCalifornia, Los Angeles, Los Angeles, California 90024; and tDepartment ofMedicine,
Veterans Administration Medical Center, West Los Angeles, California 90073
We tested the in vitro susceptibility of Candida albicans to
three defensins from human neutrophilic granulocytes
(HNP-1, 2, and 3), a homologous defensin from rabbit leuko-
cytes (NP-1), and four unrelated cationic peptides. Although
the primary amino acid sequences of HNP-1, 2, and 3 are
identical except for a single amino-terminal amino acid alter-
ation, HNP-1 and HNP-2 killed C. albicans but HNP-3 did
not. C. albicans blastoconidia were protected from HNP-I
when incubations were performed in the absence of oxygen or
in the presence of inhibitors that blocked both of its mitochon-
drial respiratory pathways. Neither anaerobiosis nor mito-
chondrial inhibitors substantially protected C. albicans ex-
posed to NP-1, poly-L-arginine, poly-L-lysine, or mellitin.
Human neutrophilic granulocyte defensin-mediated candidaci-
dal activity was inhibited by both Mg2+ and Can, and was
unaffected by Fe2". In contrast, Fe2" inhibited the candidacidal
activity ofNP-i and all ofthe model cationic peptides, whereas
Mg2+ inhibited none of them. These data demonstrate that
susceptibility of C. albicans to human defensins depends both
on the ionic environment and on the metabolic state of the
target cell. The latter finding suggests that leukocyte-mediated
microbicidal mechanisms may manifest oxygen dependence for
reasons unrelated to the production of reactive oxygen inter-
mediates by the leukocyte.
Defensins are small (mol wt 3,500-4,000), variably cationic,
cystine-rich antimicrobial peptides that are major constituents
of PMN azurophil granules. They are active in vitro against
gram-positive and gram-negative bacteria, certain fungi, and
enveloped viruses (1-7). By virtue oftheir net positive charges
at physiological pH, defensins may be' considered to be "cat-
Although the primary structures of the human defensins
are established (1), their antimicrobial mechanisms are not
known. To delineate the antifungal properties of human de-
fensins, we tested human PMN defensins against blastoconidia
of C. albicans. Certain metabolic inhibitors and divalent cat-
Address correspondence and reprint requests to Dr. Lehrer, Depart-
ment ofMedicine, UCLA School ofMedicine, Los Angeles, CA 90024.
Receivedfor publication 27 April 1987 and in revisedform 8 De-
ions showed dissimilar effects on the candidacidal activities of
the most potent human (HNP-1) and rabbit (NP- 1) defensins
(5, 6). By comparing both ofthese defensins to selected natural
or synthetic "model" cationic peptides, we identified several
properties that appear unique to the human PMN defensins.
This article describes our observations.
Proteins andpeptides. Normal human neutrophilic granulocytes were
obtained by leukopheresis and their defensins were purified as pre-
viously described (3). NP-I was purified from rabbit peritoneal granu-
locytes (5). Salmon sperm protamine (free base), mellitin, poly-L-lysine
hydrobromide (average mol wt 17,000) and poly-L-arginine hydro-
chloride (average mol wt 40,000) were obtained from Sigma Chemical
Co., St. Louis, MO and used without further purification. Certain
properties ofthese proteins and peptides are summarized in Table I.
Candidacidal testing. Susceptibility of C. albicans strain 820 to
defensins and other cationic peptides was tested as previously de-
scribed (6). In brief, 106CFU/ml ofwashed C. albicans blastoconidia
from an 18-h broth culture were incubated at 370C for up to,4 h with
specified concentrations ofthe peptides in 10mM Na phosphate buffer
(pH 7.4, ionic strength 1.36 mS). Colony counts were obtained, in
duplicate, from each of three serial, 10-fold-diluted samples. Experi-
mental results were confirmed on at least three separate occasions,
unless otherwise noted.
Oxygen consumption. The effect of various metabolic inhibitors
was tested in a model K-IC Oxygraph (Gilson Medical Electronics,
Inc., Middleton, WI) fitted with dual Clark-YSI electrodes (Yellow
Springs Instrument Co., Yellow Springs, OH) in thermostatted, stirred
chambers (vol 1.2 ml) that were maintained at 37'C.
Anaerobic experiments. Anaerobic experiments were performed in
a gloveless anaerobic chamber (Anaerobe Systems, Santa Clara, CA)
generously made available to us by Prof. S. Finegold and Dr. H.
Wexler (Veterans Administration Medical Center, West Los Angeles,
CA). C. albicans was grown aerobically for 18 h in Sabouraud's-2%
dextrose broth, as in ouraerobicexperiments,washed twiceaerobically
in 10 mM sodium phosphate buffer, pH 7.4, counted in ahemocyto-
meter, and again centrifuged. The wet pellet, containing a known
number of CFU, was transferred via a special exchange passageinto
the anaerobic chamber, wherein it was taken upin 5 nil ofanaerobic
buffer, centrifuged, and resuspended in anaerobic buffer. All media
within the chamber had been equilibratedwith an atmosphereof80%
nitrogen, 10% C02, and 10% H2 for 24-40 h, and all procedures
(primary incubation, diluting, plating, secondary incubation)were
performed anaerobically through tightly sealed access ports.Because
C. albicans colonies developed quite slowly under anaerobic condi-
tions, in most instances the petridishes were transferred after 42 h of
anaerobic incubation to an aerobic incubator where theywere addi-
tionally incubated for 24-48 h before CFU were counted.
Activity of defensins against C. albicans. Despite possessing
similar primary structures, the human neutrophil's defensins
differed in relative potency against C. albicans. Fig. 1 depicts
Candidacidal Mechanisms ofHuman Defensins
J. Clin. Invest.
©The American Society for Clinical Investigation, Inc.
Volume 81, June 1988, 1829-1835
29. Cech, P., and R. I. Lehrer. 1984. Heterogeneity of human
neutrophil phagolysosomes: functional consequences for candidacidal
activity. Blood. 64:147-15 1.
30. Pollack, C., S. C. Straley, and M. S. Klempner. 1986. Probing
the phagolysosomal environment of human macrophages with a
Ca2+-responsive operon fusion in Yersinia pestis. Nature (Lond.).
31. Kamaya, T. 1970. Lytic action of lysozyme on Candida albi-
cans. Mycopathol. Mycol. Appl. 42:197-207.
32. Marquis, G., S. Montplaisir, S. Garzon, H. Strykowski, and P.
Auger. 1982. Fungitoxicity of muramidase: ultrastructural damage to
Candida albicans. Lab. Invest. 46:627-636.
33. Davies, M. B. 1980. Peptide uptake in Candida albicans. J.
Gen. Microbiol. 121:181-186.
34. Naider, F., J. M. Becker, and E. Katzir-Katchalski. 1974. Utili-
zation of methionine-containing peptides and their derivatives by a
methionine-requiring auxotroph ofSaccharomyces cerevisiae. J. Bio.
35. Scheit, K. H., and P. M. Bhargava. 1985. Effect ofseminalplas-
min, an antimicrobial protein from bull semen, on growth and macro-
molecular synthesis in Candida albicans. Ind. J. Biochem. Biophys.
36. Cope, J. E. 1980. The porosity of the cell wall of Candida
albicans. J. Gen. Microbiol. 119:253-255.
37. Scherrer, R., L. Louden, and P. Gerhardt. 1974. Porosity ofthe
yeast cell wall and membrane. J. Bacteriol. 118:534-540.
38. Haberman, E., and J. Jentsch. 1967. Sequenz analys sed Melli-
tins aus den tryptischen und peptischen Spaltstucken. Hoppe-Seyler's
Z. Physiol. Chem. 348:37-50.
39. Dayhoff, M. 0. 1972. Atlas ofProtein Sequence and Structure.
Volume 5: D-278. National Biomedical Research Foundation,
Georgetown University Medical Center, Washington, D.C.
Candidacidal Mechanisms ofHuman Defensins