THE EXTERNAL SURFACE OF THE HUMAN BODY AS A HABITAT FOR NONFERMENTING NONPIGMENTED YEASTS'

Transcription

1 THE EXTERNAL SURFACE OF THE HUMAN BODY AS A HABITAT FOR NONFERMENTING NONPIGMENTED YEASTS' G. H. CONNELL2 AND C. E. SKINNER Department of Bacteriology and Public Health, State College of Washington, Pullman, Washington Received for publication June 1, 1953 In recent years the occurrence of yeasts, accurately diagnosed as to genus and species, has been established for various habitats. Much of this information has been reviewed by Mrak and Phaff (1948). Although a considerable amount of work has been done on the distribution of yeasts on the external surface of the human body, little knowledge of the actual specific identity of these yeasts has been established. The most nearly complete studies are those of Marwin (1949) and Ravitts (1949), both of whom have reviewed the previous literature. Repeated isolation of certain species from the same source, especially when these species are rare or unknown elsewhere, suggests that this source may well be a true habitat for these species. It was decided that a study of various external skin surfaces of the human body should be the subject for a more detailed study in order to determine whether or not yeasts, or certain species of yeasts, occur with any regularity on the human external body surface. Such knowledge of what species are to be expected as normal must necessarily precede knowledge of what species are abnormal or adventitious. EXPERIMENTAL METHODS A series of six tubes of sterile glucose-beefpeptone broth plus yeast extract, 0.1 per cent, ph 4.5 (Marwin, 1949), with sterile individual cotton tipped swabs was distributed to 250 experimental subjects, none of whom showed or 1 This work was supported in part by the Office of Naval Research as project NR: , contract Nonr-232(00) and in part by funds supplied as Project 17, under the State of Washington Initiative Measure number 171 for Medical and Biological Research. 2 A resumd of a thesis presented by the senior author (present address: Camp Detrick, Frederick, Maryland) as a part of the requirements for the Doctor of Philosophy degree, June, 1952, State College of Washington, Pullman, Washington. knew of any pathological skin conditions in himself. Each person was given detailed typed instructions for swabbing his own skin (axilla, lumbar region, inframammary, umbilicus, between fingers, between toes). These instructions called for the sterile swab to be moistened with the broth and then to be swabbed over the body surface with considerable pressure. The swab was then to be put directly into the broth tube, which was to be replugged immediately. Verbal instructions and demonstrations as well as the typed instructions were given. Most of the subjects were young adults, 17 to 30 years old. At first the subjects were instructed to refrain from bathing for a period of three days before taking cultures. After half of the isolates were secured, it was found that cultures could be obtained as readily a few hours after bathing as several days after. Hence, instructions to refrain from bathing were dispensed with in the latter half of the isolation program. All cultures were incubated at room temperature (about 24 C) for up to two weeks. As soon as growth was apparent and in any case by the end of the two week period, streaks were made on glucose-tartaric acid-beef-peptone agar poured plates at ph 4.8, and these were incubated at room temperature. Isolations were made from these plates, and in those cases where apparently different colonial types showed up, all types were isolated. Usually only a single type was observed on a plate. These apparently pure cultures were replated two or three times and were examined microscopically to verify the fact that only a single morphological vegetative type of cell was present. After purification, all isolates were tested for their ability to ferment glucose. The term "fermentation" is used here in the usual sense in the zymological literature, that is to mean gas production under anaerobic conditions, the alcoholic fermentation. Durham tubes with 5 per cent glucose, 1 per cent peptone, 0.1 per cent yeast extract at ph 5.8 to 6.0 were inoculated and incubated at room temperature. If pellicles 627

2 628 G. H. CONNELL AND C. E. SKINNER [VOL. 66 were formed, the culture was shaken to cause this surface growth to settle to the bottom of the tube. Tubes were not considered negative unless no gas was visible in the inserted tube of the Durham apparatus at all times up to 30 days. Cultures showing no fermentation were rechecked periodically. At the same time the purified cultures were examined for ascospore formation on various culture media, including Gorodkova's (Skinner et al., 1947) and a vegetable juice (Wickerham, 1951) agar. Various temperatures were used. All the isolates which produced ascospores were fermenters except three isolates which were diagnosed as Lipomyces starkeyi. (Connell, Skinner, and Hurd, 1953). At the same time that yeasts from the body were being isolated, yeasts from the air also were isolated by exposing petri plates of acidified glucose agar to the air in laboratories, offices, or homes of the subjects. Another method of sampling the flora of the air involved the use of bubblers and impingers through which the room air was drawn. Organisms trapped in the sterile impinger fluid were isolated by plating and subjected to the same identification procedures as were the organisms isolated from the skin. The fermenting species were aasigned for further study and identification to Mr. Robert Hurd who had cooperated in the work to this point, and the writers retained the nonfermenters for the same purpose. Even though most nonfermenting yeasts are considered to be anascosporogenous, one final attempt to demonstrate spores was made on all the nonfermenting nonpigmented yeasts, which also were given a final check for fermentative ability. None was found to be fermentative, and only the three isolates of L. starkeyi were found to produce ascospores by Starkey's (1946) technique. The Dalmau technique as modified by Wickerham (1951) was used to detect the presence of pseudomycelium or of true mycelium, and this procedure was checked by the use of scratch plates of unsugared corn (maize) meal agar (Conant et al., 1944). In general, Wickerham's (1951) methods for the study of morphology were followed. RESULTS A total of 1,121 organisms was isolated from the body and from the air. After preliminary morphological study, some of these turned out to be organisms other than yeasts, and a small number of this group did not survive culturing on glucose-peptone agar. Most of these, previous to dying out, had been tentatively diagnosed either as smut sporidia or as yeast-like phases of certain dematiaceous molds, the so-called "black yeasts" (Skinner et al., 1947). When "black yeasts" are first isolated, they often resemble genuine yeasts both microscopically and in macroscopic colony appearance. Therefore, it was decided to keep them and to include them in general studies to be made later. Carri6n (1950) has recorded their occurrence on the skin of human beings and has indicated that some "black yeasts" are of considerable medical importance. Their importance in the dairy industry has been stressed from time to time. Including the bulk of the black yeasts, which did survive culture, a total of 784 isolates from the body and 235 from the air was maintained on laboratory media. The ability of certain carbohydrates or other organic substances to serve as sole sources of carbon and of nitrates to serve as sole sources of nitrogen was tested by Wickerham's techniques. Along with organisms isolated in this investigation many cultures of named species of yeasts were obtained from Wickerham, who in turn had obtained, these authenic strains from Lodder. Production of "starch-like" substances was demonstrated by incubation for up to 30 days in Wickerham's (1951) nitrogen base synthetic media plus glucose. In all cases which were recorded as positive, the results were definite although the reactions varied in intensity from strong to weak. Faintly positive starch reactions were sometimes intensified by growing cultures under constant agitation. Production of carotinoid pigment was determined by the method of Mrak et al. (1949). Other physiological methods were essentially those of Wickerham (1951). For purposes of comparison, isolates were divided into: (1) Fermenting, nonpigmented, imperfect, and ascogenous yeasts, the genera Saccharomyces Meyen emend. Rees, Torulopsis Berlese, and Candida Berkhout (in part). (2) Nonpigmented, nonfermenting, mostly anascosporogenous yeasts, the genera Cryptococcus Kuitzing emend. Vuillemin, Candida Berkhout (in part), and three isolates of Lipomycee starkeyi Lodder & van Rij (Connell, Skinner, and Hurd, 1953).

3 1953] SURFACE OF HUMAN BODY AS HABITAT FOR YEASTS 629 (3) Carotinoid pigment containing, budding, anascosporogenous yeasts, not reproducing by ballistospores: the genus Rhodotorula Harrison. (4) "Black yeasts". (5) Carotinoid pigment containing anascosporogenous yeasts, reproducing by ballistospores (repetition) and by budding: the genus Sporobolomyces Kluyver & van Niel. Tables 1 and 2. Table 1 indicates that the yeastflora on the body differs considerably from that of the air around the body. Considering for the TABLE 1 Types of organisms isolated from the be true yeasts. The three categories are (1) carotinoid pigment containing yeasts: the genera Rhodotorula and Sporobolomyces, (2) nonpigmented, nonfermenters: Cryptococcus, Lipomyces, and Candida, (3) fermenters: Torulopsis, Candida, and Saccharomyces. Note how the carotinoid pigment containing yeasts predominate percentage-wise in the air and how the fermenters predominate on the body. Clearly there is a different ratio of one type of yeast to another on the skin and in the air, and statistical analysis body and from the air FROM BODY FROM AIR TYPE OP ORGANISM TOTAL PER CENT Fermenters-nonpigmented Nonfermenters-nonpigmented Rhodotorula species "Black yeasts"-dematiaceae Sporobolomyce speciesc Total , TABLE 2 Types of yeasts, exclusive of dematiaceous organisms, isolated from the body and from the air TYPX YEAST FROM BODY FROM AIR Carotinoid yeasts Colorless nonfermenters Fermenters Total moment only the first two categories of table 1, nonpigmented fermenters and nonfermenters, it is seen that 180 nonfermenters and 507 fermenters were isolated from the body; and from the air, 26 fermenters and 27 nonfermenters. A statistical analysis shows that the hypothesis that there is no significant difference between these two habitats as to the proportion of fermenters and nonfermenters is untenable (X2 = 14.9, n = 1). Likewise the ratio of "black yeasts" to fermenters is greater in the air than on the body; the body, however, does have a preponderance of fermenting yeasts. Table 3 shows the results leaving out consideration of the "black yeasts", which as mentioned previously are not considered to TOTAL proves this statement. Moreover, this statement is not weakened by the fact that the yeasts were isolated from the two sources by different techniques since when air samples were drawn through impingers and the liquid subsequently was plated out, the proportion of Rhodotorula and "black yeasts" to others was even larger than that when plates were exposed to the air. The difference, however, was not sufficient to indicate that the two methods were significantly different as to selectivity of any one type of yeast. Tables 3 and 4. Further evidence indicates that the skin harbors a different flora from that of the air around the body. Swabbings of floors,

4 630 G. H. CONNELL AND C. E. SKINNER [VOL. 66 walls, window panes, furniture, table tops, rugs, and other materials with which the body comes in contact were sampled exactly as was the skin, except that for each swab considerably more area was included. These swabs were handled exactly as the skin swabs were handled and inoculated into the same medium. Of 132 attempts only 3 isolates of "yeasts" were obtained, 2 "black yeasts" and 1 Rhodotorula. It is obvious that qualitatively and quantitatively the yeasts faces of the body. Note that when arranged in order of frequency of isolation, fermenters and nonfermenters are in almost the same sequence on all 5 parts of the body sampled, and moreover the percentage of occurrence is rather close, one to the other. The distribution of the nonfermenting, nonpigmented, anascosporogenous yeasts into species gave less difficulty than was anticipated. At first all of them were placed in 10 species and TABLE 3 Comparison of nonfermenting types of yeasts, exclusive of "black yeasts" FROM BODY FROM AIR TYPE YEAST TOTAL Rhodotorula species Sporobolomyces species Nonpigmented yeasts Total TABLE 4 Numbers of nonpigmented yeasts isolated from various body surfaces-genera Saccharomyces, Lipomyces, Cryptococcus, and Candida PERENTERS NONPERMZNTERS PART OF TRE BODY TOTAL PER CENT Between toes Between fingers Inframammary Umbilicus Lumbar region Axilla Total on the body are different from those in the air or those on surfaces with which the body comes in contact. It seems likely that the skin surface may well be a true habitat for yeasts. Croft and Black (1938) also concluded that yeasts lived saprophytically on normal skins. However, the data consisted simply in their own and other published records of the fact that yeasts (mostly, except for a few species of the genus "Monilia", unclassified) had been isolated from human skin surfaces. Table 4 indicates that there is no significant difference between fermenters and nonfermenters as to their relative distribution on various surone variety of Torulopsis according to Lodder's (1934) monograph, and 3 species of Candida according to Diddens and Lodder (1942). Only 4 isolates of Candida seemed to belong to other, and as far as could be determined, undescribed species. The authors are inclined to conservative taxonomy, and 20 or more species of Cryptococcus, most of them new, might have been recognized if the writers had been of the "splitting" school. When the text of Lodder and Kreger-van Rij (1952) was first consulted, it was found that few changes had to be made to bring these results into their new concepts. In general, their taxonomy and their classification

5 19531 SURFACE OF HUMAN BODY AS hiabitat FOR YEASTS 631 were followed with some exceptions. Below is a discussion of where the taxonomy of table 5 differs from that of Lodder and van Rij: (1) All strains which were put into Cryptococcus minor produce "starch". Therefore, they cannot be called Torulopsis famata as Lodder and van Rij named their strains which had previously been called Torulopsis minor. Until a further study is made, the authors do not wish to make a new combination. Since two strains of T. famata obtained from Lodder failed to form starch but did ferment glucose slightly, the writers are inclined to think that the isolates of this study are different from T. famata. (2) Cryptococcus aerius is being maintained since the authentic strain which came originally from Lodder produced starch when tested by Wickerham's (1951) method. Comparison of the new isolates with this culture leads the writers to identify their isolates with it. (3) Cryptococcus rotundatus is retained. The new isolates and the authentic culture from Lodder had not undergone the change in culture which Lodder and van Rij (1952) recorded, and the writers, therefore, see no reason for not recognizing the species. Ten of the eleven isolates from air of Candida zeylanoides were first diagnosed as Torulopsis uvae. These 10 isolates as well as the authentic culture from Lodder regularly failed to produce starch and on all media failed to produce a pseudomycelium. Occasionally a small bubble of gas was produced from glucose broth after prolonged incubation. Lodder and van Rij (1952) have found that their cultures have developed, since 1934, an ability to form pseudomycelium and on this basis have combined Torulopsis uvae with Candida zeylanoides. The authors do not agree entirely with this taxonomy (Connell and Skinner, 1951), but they accept the nomenclature for the present. Table 5. Table 5 shows that species of yeasts which were found on the skin and those found in the air are proportionally different in occurrence. Note especially the absence of some of the species in air. Note also the very large number of Candida zeylanoides from the air as compared to the small number from the skin. The difference in distribution of species of nonfermenting colorless yeasts as well as the difference previously noted between fermenting and nonfermenting pigmented yeasts, together indicate that either some of the yeasts on the skin are multiplying there, or the skin is selectively killing off some yeasts more rapidly than others. The latter hypothesis is difficult to accept after one has had the experience of isolating yeasts readily from skin surfaces and of having to really search rather hard for yeasts in the air and on surfaces TABLE 5 Comparison of frequency of isolation of species of nonfermenting colorless yeasts from the body and from the air spzczs FROM FROM BODY AIR Cryptococcus aerius (Saito) Nannizzi Cryptococcus albidus (Saito) Skinner Cryptococcus diftluens (Zach) Lodder & van Rij Cryptococcus laurentii (Kufferath) Skinner Cryptococcus laurentii (Kufferath) Skinner var flavescens (Saito) Lodder & van Rij Cryptococcus luteolus (Saito) Skinner Cryptococcus minor Pollacci & Nannizzi Cryptococcus rotundatus (Redaelli) Nannizzi apud Pollacci Lipomyces starkeyi Lodder & van Rij Candida lipolytica (Harrison) Diddens & Lodder Candida mesenterica (Geiger) Diddens & Lodder Candida zeylanoides (Castellani) Langeron & Guerra Candida spp Total of the living room or the laboratory. There were no isolations of the organism obtained from skin that was carefully described by Ruiz (1943) as a new species, Torulopsis orbicularis, nor was Cryptococcus neoformars isolated. Very few of the isolates grew well at 37 C, and these few were clearly not C. neoformans. A special study was made with all the 207 isolates of table 5 as to their ability to utilize various chemicals as sources of carbon. Wickerham (1951) has found some of these compounds

6 632 G. H. CONNELL AND C. E. SKINNER [VOL. 66 very useful in classifying species of Hansenula. The following substances were used following the techniques and media described by Wickerham: glucose, galactose, maltose, sucrose, lactose, cellobiose, trehalose, raffinose, melibiose, melizitose, inulin, soluble starch, xylose, L-arabinose, rhamnose, ethyl alcohol, glycerol, i-erythritol, adonitol, dulcitol, mannitol, sorbitol, inositol, a-methylglucoside, potassium gluconate, calcium 2-ketogluconate, lactic acid, succinic acid, and citric acid. In addition, the ability of the cultures to grow in media devoid of vitamins, to grow with KNO3 as the sole nitrogen source, to liquefy gelatin, to grow at 37 C and at 5 C, and to grow in media of high osmotic concentration was tested. For the organisms studied, it was found that none of the carbon compound assimilation studies, and only one of the other biochemical tests, gave any promise of usefulness in classification for species of Cryptococcus that could not be obtained with the sugar utilization and other biochemical tests used by Lodder (1934). There was little evidence of any sort of agreement of Wickerham's (1951) cultural tests, one with another, and the writers were compelled to accept the Lodder-van Rij system almost in its entirety. The one new character that did appear to have value was synthesis of "starch-like" substances. The ability of some yeasts to produce from glucose polysaccharides which give a blue test with iodine was described by Aschner and associates (1945). The studies reported herein were nearly completed before the appearance of the Lodder and Kreger-van Rij (1952) book, and the writers had come independently to the conclusion that "starch" synthesis rather than capsule production as Almeida and associates (1941) had suggested, or fermentation as suggested by Skinner (1950), probably would be the best single character to separate Cryptococcus from Torulopsis. Personal communication from Lodder had given an indication that she was using this character, and the writers were ready to accept it before the appearance of her book. The results of this study are even more definitive than are those of Lodder and van Rij since there is not a single isolate of the whole collection of some 700 Torulopsis-Cryptococcus isolates in which the authors recorded either a failure to produce starch for a nonfermenter or starch production by a fermenter. The technique used, that of Wickerham (1951), is apparently more sensitive than that used by Lodder and van Rij. ACKNOWLEDGMENTS The authors wish to thank the officials of the Northern Regional Research Laboratories of the United States Department of Agriculture, Peoria, Illinois, for allowing the senior author to participate in an intensive training course in zymologic taxonomic techniques under the capable direction of Dr. L. J. Wickerham. Thanks also are due Dr. F. Stodola of the same institution for synthesizing three of the compounds used in the carbon assimilation studies. SUMMARY A study was made of 784 isolates of yeasts and yeast-like fungi obtained from various human skin surfaces and of 235 organisms of similar types isolated from the air. Yeasts and yeast-like fungi 'were isolated much more frequently from the skin surfaces than from the air, or from wall, floor, or glass surfaces of rooms. The percentage of "black yeasts", of Rhodotorula, and of Sporobolomyces was significantly lower on the skin than in the air, and a significantly higher percentage of fermenters was found on the skin than in the air. When the 180 nonfermenting, nonpigmented yeasts isolated from the body and the 27 of similar types isolated from the air were studied in detail, it was found that in some cases the percentages of the same species from each habitat varied decidedly. The following species are recorded as having been isolated both from the human skin and from the air, and for many of them this is the first recorded instance of their having been isolated from the human body: Cryptococcus albidus, C. diffluens, C. laurentii, C. laurentii var flavescens, C. luteolus, Candida lipolytica, and Candida zeylanoides. The following organisms were isolated from the body but not from the air: Cryptococcus aerius, C. minor, C. rotundatus, Lipomyces starkeyi, and Candida nmsenterica. Most of these are new habitat records. It is believed that various human skin surfaces actually serve as a habitat for certain yeasts. A study of utilization of 28 organic compounds as sole carbon sources and of other biochemical tests was made of all the nonfermenting, nonpigmented yeasts isolated in this study. With one exception none of these biochemical tests,

7 19531 SURFACE OF HUMAN BODY AS HABITAT FOR YEASTS 633 other than those already used by Lodder in 1934, was found to be of value in classifying this particular group of yeasts. The production of a substance which turned blue with iodine and commonly called "starch" in recent zymological literature was found to be an excellent and reliable taxonomic character and was found to correlate with ability to ferment glucose. None of the fermenting yeasts isolated produced "starch", and none of the nonfermenting colorless yeasts without pseudomycelium failed to produce it. It is concluded that the introduction by Lodder and van Rij of "starch" production as a taxonomic criterion to separate the genera Torulopsis and Cryptococcus was justified. REFERENCES ALMEIDA, F. P. DE, LACAZ, C. DA S., AND BARROS, O. DE 1941 Orientagao pratica para a identificag&o das leveduras. Rev. Inst. Adolfo Lutz (Brasil), 1, ASCHNER, M., MAGER, J., AND LIEBOWITZ, J Production of extra-cellular starch in cultures of capsulated yeasts. Nature, 156, 295. CARRI6N, A. T Yeast-like dematiaceous fungi infecting the human skin. Arch. Dermatol. and Syphilol., 61, CONANT, N. F., MARTIN, D. S., SMITH, D. T., BAKER, R. D., AND CALLOWAY, J. L Manual of clinical mycology. W. B. Saunders Company, Philadelphia and London. CONNELL, G. H., AND SKINNER, C. E The origin of some species of the genus Candida in non-pseudomycelium-producing anascosporogenous yeasts. Mycopathologia, 6, CONNELL, G. H., SKINNER, C. E., AND HURD, R. C Lipomyces starkeyi on the skin surface of the human body. Mycologia, in press. CROFT, C. C., AND BLACK, L. A Yeastlike fungi isolated from normal skins. J. Lab. Clin. Med., 23, DIDDENS, H. A., AND LODDER, J Die anaskosporogenen Hefen, Halfte II. North Holland Publishing Company, Amsterdam. LODDER, J Die anaskosporogenen Hefen, Hialfte I. Verhand. Akad. Wet. Amsterdam, Afdeel Natuurk. (2 ser.), 32, LODDER, J., AND KREGER-VAN RIJ, N. J. W The yeasts, a taxonomic study. North Holland Publishing Company, Amsterdam. MARWIN, R. M Relative incidence of Candida albicans on the skins of persons with and without skin diseases. J. Invest. Dermatol., 12, MRAK, E. M., AND PHAFF, H. J Yeasts. Ann. Rev. Microbiol., 2, MRAK, E. M., PHAFF, H. J., AND MACKINNEY, G A simple test for carotinoid pigments in yeasts. J. Bact., 57, RAVITTS, H. G Cutaneous cryptococcosis. A survey of Cryptococcus on normal and pathologic skin. J. Invest. Dermatol., 12, RuIz, 0. M Estudio de una nueva especie de levaduras del g6nero Torulopsis Berlese, aislado de las escamas de la piel humana. Anales inst. biol. (Univ. nac. M6x.), 14, SKINNER, C. E Generic name for imperfect yeasts, Cryptococcus or Torulopsis? Am. Midland Naturalist, 43, SKINNER, C. E., EMMONS, C. W., AND TSuCmYA, H. M Henrici's molds, yeasts, and actinomycetes. 2nd ed. John Wiley and Sons, Inc., New York. STARKEY, R. L Lipid production by a soil yeast. J. Bact., 51, WICKERHAM, L. J Taxonomy of yeasts. U. S. Dept. Agric. Tech. Bull