IAI Accepts, published online ahead of print on 1 June 2009 Infect. Immun. doi:10.1128/IAI.00296-09 Copyright © 2009, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.

1 2

Most Environmental Isolates of Cryptococcus neoformans var. grubii (Serotype A)

3

are Not Lethal for Mice

4 5 6

Anastasia P. Litvintseva and Thomas G. Mitchell

7

Department of Molecular Genetics and Microbiology,

8

Duke University Medical Center, Durham, North Carolina, USA

9 10 11 12 13 14

Running title: Murine virulence of C. neoformans serotype A

15 16

Key words: Cryptococcus neoformans, virulence, environmental isolates, retrotransposon,

17

genotype.

18

Thomas G. Mitchell, corresponding author, to whom reprint requests should be addressed:

19

T.G. Mitchell , Box 3803, Department of Molecular Genetics and Microbiology, Duke University

20

Medical Center, Durham, NC 27710, USA. Tel, +919-684-5792; Fax, +919-684-2790; email,

21

[email protected]

22 23 24 25

1

26

ABSTRACT

27

Most cases of cryptococcosis are caused by Cryptococcus neoformans var. grubii (serotype

28

A), which is widespread in the environment, where it is primarily associated with pigeon

29

excreta. A number of molecular epidemiological studies indicate that many environmental and

30

clinical isolates of serotype A are indistinguishable. However, the murine virulence of

31

environmental strains of C. neoformans has not been thoroughly evaluated. We used the

32

murine intranasal model of cryptococcosis to compare the lethality of clinical and

33

environmental strains of serotype A that possessed identical genotypes as determined by

34

Amplified Fragment Length Polymorphisms (AFLP) and Multilocus Sequence Typing (MLST).

35

Eleven environmental strains were tested and only one caused disease within 60 days

36

postinfection, at which time the experiments were terminated. Conversely, seven of ten clinical

37

isolates were lethal for mice at median times of 19 to 40 days. Passing environmental isolates

38

in mice (up to three times) did not significantly increase their lethality. In follow-up studies, we

39

developed a new genotyping technique based on hybridization with TCN2 and TCN4

40

retrotransposon-specific probes. Although the retrotransposon banding patterns were unstable

41

after prolonged incubation in the laboratory, this method was able to differentiate clinical and

42

environmental strains that had the same AFLP/MLST genotypes.

43 44 45

INTRODUCTION

46

Cryptococcus neoformans var. grubii (serotype A) is a ubiquitous saprobic yeast. In

47

nature, it can be routinely isolated from avian, especially pigeon, habitats, soil contaminated with

48

avian excreta, and decayed wood (10, 18, 20, 25, 35, 36). Mammalian infection with C.

49

neoformans is acquired by the inhalation of airborne yeasts or spores. The ensuing pulmonary

50

infection is often asymptomatic. Alternatively, this exposure may lead to subacute, invasive

2

51

disease. The yeast cells may disseminate to any part of the body, but C. neoformans is

52

neurotropic, and the typical, life-threatening clinical manifestation of cryptococcosis is

53

meningoencephalitis. Cryptococcal disease occurs predominantly in people with impaired

54

immunity, but individuals with apparently normal host defenses may also develop grave

55

infections (5). With rare exceptions, cryptococcosis is not transmissible among humans or other

56

animals (5).

57

Although cryptococcosis can be caused by other species of Cryptococcus or any of the

58

three serotypes of C. neoformans (A, D or AD), most clinical and veterinary cases worldwide are

59

caused by isolates of serotype A, which is also the most prevalent serotype among

60

environmental samples. Genotypic analyses of isolates of serotype A have identified three

61

genetically isolated subpopulations, designated VNI, VNII and VNB (26-28). Representative

62

isolates of all three populations have been cultured from patients and the environment.

63

However, VNI strains of serotype A are globally dominant in patients, veterinary cases, and the

64

environment. For example, using Amplified Fragment Length Polymorphisms (AFLP) and

65

Multilocus Sequence Typing (MLST), we genotyped numerous environmental VNI isolates from

66

North America and found identical genotypes in clinical and environmental samples (25, 28). In

67

contradistinction, strains of VNII are much less common in patients, and they are exceptionally

68

rare in the environment (3, 28, 31). Regarding the VNB population, both clinical and

69

environmental isolates appear to be confined to southern Africa (28).

70

Since identical genotypes of VNI strains can be isolated from clinical specimens and

71

environmental samples, it is widely accepted that fully virulent strains are present in the

72

environment (5, 6, 21). However, few studies have experimentally tested the pathogenicity of

73

environmental isolates of C. neoformans, and none have compared the virulence of clinical and

74

environmental isolates with the same genotypes. In 2006, Silva et al. compared the virulence of

75

62 strains of serotype A that were isolated from pigeon excreta in Brazil (39). Using an

3

76

intravenous mouse model of experimental cryptococcosis, they observed that the environmental

77

strains varied in their ability to cause disease; some mice developed symptoms of

78

cryptococcosis, but others remained asymptomatic (39). In 1989, Fromtling et al. evaluated the

79

murine virulence of environmental and clinical isolates of C. neoformans from Puerto Rico (14).

80

They also used an intravenous model of infection, and they determined that the mean lethal

81

dose resulting in the death of 50% of the infected mice (LD50) was significantly higher for

82

environmental compared to clinical strains (14). In the next most recent report of the virulence of

83

environmental strains, published in 1963, the authors compared 21 clinical isolates of C.

84

neoformans with 47 isolates from soil contaminated with pigeon feces (19). The isolates were

85

tested for virulence in a murine intracerebral model of cryptococcosis, in which suspensions of

86

the yeast cells were injected directly into the cerebra of the mice. They concluded that “the

87

strains isolated from [human] cases of cryptococcosis were more virulent than the soil strains;

88

however, almost half [