JOURNAL OF BACTERIOLOGY, June 1969, p. 996-1004 Vol. 98, No. 3 Copyright © 1969 American Society for Microbiology Printed in U.S.A. Interaction of Candida albicans with Human Leukocytes and Serum' ROBERT I. LEHRER AND MARTIN J. CLINE Cancer Research Institute and the Department of Medicine, University of California Medical Center, San Francisco, California 94122 Received for publication 10 January 1969 A quantitative assay of candidacidal activity based on differential staining of non- viable Candida albicans by methylene blue was developed and applied to studies of leukocytes from normal individuals and patients with fungal and other infections. Serum factors were necessary for optimal phagocytosis of C. albicans but lacked direct candidacidal activity. Normal human neutrophils (38 studies) killed 29.0 a 7.4% of ingested C. albicans in 1 hr. Eosinophils and monocytes killed a smaller percentage. Neutrophil candidacidal activity did not require protein or ribonucleic acid synthesis by the leukocyte but was inhibited by anaerobic conditions, potas- sium cyanide, and colchicine. Leukocytes of a patient with hereditary myeloperoxi- dase deficiency and of three children with chronic granulomatous disease phago- cytized C. albicans normally, yet failed to kill them. Our data suggest that the neutrophil can play an important role in resistance to Candida infection and that the lysosomal enzyme myeloperoxidase and its oxidant substrate hydrogen peroxide are the major participants in neutrophil candidacidal activity. Although certain disorders and therapeutic pro- on Sabouraud agar slants at room temperature and cedures appear to alter man's susceptibility to were transferred every 2 months. fungal infection, the factors involved in either Test organisms were inoculated into 50 ml of Sa- bouraud dextrose broth (Difco) and cultured or altered responses to fungal challenge 2% normal at 33 C. Under these conditions, the Candida cells are Candida albicans is incompletely defined. rep- grew only in yeast phase. Only cultures with fewer than resentative of the group of opportunistic patho- 5% nonviable organisms, as determined by dye exclu- gens which cause disease in the impaired host (3, sion (vide infra), were employed in assays. This condi- 5, 13, 44). This study was undertaken to develop a tion was fulfilled by most cultures 3 to 14 days old. quantitative method capable of measuring the Candida cells were quantitated in a hemocytometer candidacidal activity of human leukocytes and and resuspended in Hanks balanced salt solution serum and to define the metabolic pathways of (BSS) for use in candidacidal assays. the leukocyte which participate in the candidaci- Leukocyte preparation. Peripheral venous blood 5 units of per ml dal process. The new method and results of its containing heparin (Lilly, benzyl alcohol preservative) was mixed with half its volume application to studies of leukocyte function and of 3% dextran in normal saline (39) and allowed to simplified in vitro models of Candida infection sediment at room temperature for 30 to 45 min. The are presented. leukocyte-rich supernatant fluid was removed, cen- trifuged at 150 X g for 10 min, and washed twice MATERIALS AND METHODS with BSS containing 10% fetal calf serum and 5 units of heparin per ml. Leukocyte and absolute polymor- C. albicans. A strain of C. albicans (UC 820) was phonuclear (PMN) cell concentrations were deter- obtained from the Department of Clinical Pathology, mined by hemocytometer count. Just before being University of California. It showed features typical of forma- added to the assay, the leukocytes were centrifuged at C. albicans (2, 27), including chlamydospore 150 X g for 8 min and resuspended in BSS at a final tion, rapid filamentation in the presence of serum (38), concentration of 107 PMN/ml. Erythrocyte contami- on EMB characteristic growth pattern agar, fermenta- nation of the preparations was variable, usually ap- with acid and gas pro- tion of dextrose and maltose one red cell per white cell. Platelet con- and to proximating duction, lack of lactose fermentation, failure tamination also varied, but usually was less than one gas in sucrose. Stock cultures were maintained produce platelet per five leukocytes. Serum. Fresh serum was obtained from several I This work was presented in part at the 1968 Meeting of the American Federation for Clinical Research, Atlantic City, N.J. normal group AB donors, stored individually in 4-ml (Clin. Res. 16:331, 1968). volumes on dry ice, and used within 1 month. In some 996 VOL. 98, 1969 INTERACTION OF C. ALBICANS WITH HUMAN LEUKOCYTES 997 experiments leukocytes were studied in the donor again decanted and replaced with 1 ml of BSS con- patient's serum. taining 2 X 106 Candida cells and 25% normal AB Neutrophil candidacidal assay. Equal volumes (0.25 serum. After a 30-min incubation period, the mono- ml) of AB serum, leukocyte suspension, and BSS cytes were washed twice with warm BSS and serum to were added to sterile plastic tubes (12 by 75 mm). All remove the nonphagocytized yeast cells; they were tubes were prepared in duplicate; a third tube con- then incubated for an additional 60 min in fresh BSS taining all components except leukocytes served as a containing AB serum. In the tubes without cover slips, control. Sterile technique was used throughout. the incubation was terminated by the addition of The tubes were incubated for 10 min at 37 C. A 0.25 ml of 2.5% sodium deoxycholate (pH 8.7) to lyse 0.25-mi volume of C. albicans at a concentration of 107 the monocytes. Methylene blue was added, and the yeast cells per ml was then added, and the tubes were suspensions were handled and counted as in the neu- rotated (30 rev/min) at 37 C for 60 min. After 15 trophil assay. For assessment of phagocytosis, cover min, a drop was taken for direct examination and slips from the other Leighton tubes were stained and preparation of stained smears to confirm that all examined. added organisms had been ingested. At 60 min, 0.25 ml of 2.5% sodium deoxycholate (pH 8.7) was added RESULTS to each tube. At this concentration, deoxycholate causes immediate lysis of the blood cells without Differential staining of viable and nonviable damage to the Candida cells. Methylene blue, 0.01% Candida. We established the ability of methylene in distilled water, was then added to achieve a final blue to specifically stain nonviable C. albicans in volume of 4 to 5 ml and a final dye concentration of several ways. Usually, fewer than 5% of the cells about 0.0075% or 2 X 10-4 M. The Candida cell in the Candida broth cultures were stained by suspensions were centrifuged at 1,100 X g for 15 min 2 X 10-4 M methylene blue. If the yeast cells were at 4 C and resuspended in about 0.5 ml of the residual killed by boiling for 10 min, all were intensely supernatant fluid. Thereafter, the tubes were kept in stained. Staining occurred promptly and was in- an ice-water bath until they could be examined mi- dependent of pH in the of 5 to 8. A croscopically. At least 300 Candida cells from each range repre- tube were examined to determine the percentage sentative mixture of heat-killed and viable Can- stained. To derive the candidacidal activity due to the dida cells after exposure to 2 X 10-4 M methylene action of phagocytes, the percentage of stained yeast blue is shown in Fig. 1. cells in the control tubes, usually 0.5 to 2.5%, was The specificity of the methylene blue staining of subtracted from that in the experimental tubes. All phagocytized C. albicans cells was studied as tubes were counted in a "blind" manner to avoid follows. After an inoculum of Candida cells was subjective bias. ingested by normal leukocytes, the leukocytes Viable Candida cells, which were unstained, clearly were lysed with deoxycholate, and methylene blue differed from the nonviable organisms, which took a was added. A drop of the lysed uniform, intense blue cytoplasmic stain. If Candida suspension was cells were left at 0 C in the presence of methylene blue placed on a slide containing a thin strip of for more than 3 hr, some of the organisms acquired a Sabouraud agar, sealed with a cover slip, and diffuse, faintly blue tinge; therefore, observations were observed microscopically at intervals for 8 hr. concluded within this time period. During several such experiments, unstained Can- Candidacidal activity was independent of the Can- dida cells were seen to bud and form micro- dida-neutrophil ratio over a range of 0.5 to 1.5 Candida colonies, whereas the blue-stained Candida cells cells per neutrophil and of serum concentrations in the always failed to bud. range of 6 to 50%. When cells of C. albicans cultured The possibility that methylene blue was itself for 72 hr or more were used, no multiplication oc- responsible for killing the phagocytized Candida curred during the 60-min incubation period of the assay, ensuring that the percentage killed was directly cells that were stainable was excluded by the proportional to the absolute number killed. Initial ex- following experiment. Ingested Candida cells were periments revealed that leukocytes killed by repeated again recovered from normal leukocytes by de- freezing and thawing lacked candidacidal activity in oxycholate lysis and divided into duplicate por- this system. tions. One portion was diluted with Sabouraud Monocyte candidacidal assay. Monocytes were iso- broth, and pour plates were made in Sabouraud lated from heparinized peripheral blood of normal agar for colony counts. The second portion was donors by previously described methods (8) and diluted with 0.01 % methylene blue, and the con- suspended at 106 cells per ml in medium TC 199 con- centration of Candida cells was determined in a taining 30% fetal calf serum.