- Email: [email protected]
Comparative analysis of pathogenicity of Cryptococcus neoformans serotypes A, D and AD in murine cryptococcosis
Journal of Infection (2005) 51, 10–16
www.elsevierhealth.com/journals/jinf
Comparative analysis of pathogenicity of Cryptococcus neoformans serotypes A, D and AD in murine cryptococcosis F. Barchiesia,*, M. Cogliatib, M.C. Espostob, E. Spreghinia, A.M. Schimizzia, B.L. Wickesc, G. Scalisea, M.A. Vivianib a
` Politecnica delle Marche, Azienda Istituto di Malattie Infettive e Medicina Pubblica, Universita Ospedaliera Umberto I8, Via Conca 60020 Torrette di Ancona, Ancona, Italy b ` degli Studi di Milano, IRCCS Ospedale Maggiore, Istituto di Igiene e Medicina Preventiva, Universita Milano, Italy c Department of Microbiology, University of Texas Health Science Center at San Antonio, San Antonio, TX 78229-3900, USA Accepted 21 July 2004 Available online 16 September 2004
KEYWORDS Cryptococcus neoformans; Serotype; Diploid; Haploid; Hybrid
Abstract Objectives. To characterize the pathogenicity of 15 strains of Cryptococcus neoformans belonging to several serotype/mating type allele patterns (Da, Da, Aa, Aa, Aa/Da and Da/Aa) in experimental models of murine cryptococcosis. Methods. CD1-infected mice were examined for survival and fungal loads in either brain or lung during the course of infection. Results. All strains, with the exception of one Da strain, produced melanin in vitro. Similarly, all strains were encapsulated and produced phospholipase. When CD1 mice were challenged intravenously (i.v.) with 5!105 CFU/mouse and observed for 60 days post-infection, a significant variation of mortality rate was observed among mice infected with different strains. Aa and Aa/Da strains all produced 100% mortality within the study period with mean survivals significantly shorter than those of mice infected with strains belonging to any other allele type (P!0.0001). A wide range of pathogenicity was shown by haploid and diploid strains presenting Da allele. This finding was confirmed by an intranasal model of challenge. To investigate the progression of infection, the mice were challenged i.v. with 5!104 CFU/mouse and tissue burden experiments (brain and lung) were performed on days 6 and 12 postinfection. Only the mice infected with Aa and Aa/Da strains showed a O1 log10 increase of CFU/g in both tissues throughout the study period. Conclusions. Our results suggest that the presence of the Aa mating type allele in either haploid or diploid strains is correlated with virulence, while the presence of the Aa or Da allele in haploid strains is associated with moderate or no virulence.
* Corresponding author. Tel.: C39-071-5963467; fax: C39-071-5963468. E-mail address: [email protected] (F. Barchiesi). 0163-4453/$30.00 Q 2004 The British Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jinf.2004.07.013
Pathogenicity of Cryptococcus neoformans ser. A, D and AD
11
Finally, either haploid or diploid strains presenting Da allele vary in virulence. Q 2004 The British Infection Society. Published by Elsevier Ltd. All rights reserved.
Introduction
Analysis of virulence factors
The yeast Cryptococcus neoformans is an encapsulated fungus that can cause infection in both immunocompetent and immunocompromised individuals. Its incidence has dramatically increased during the last two decades because of the AIDS epidemic, although impaired cellmediated immunity in general represents the main predisposing factor for development of cryptococcosis. Meningoencephalitis is the most frequent clinical presentation, although any organ can be affected. Cryptococcus neoformans has two mating types, MATa and MATa, and two varieties, var. neoformans and var. gatti, which can be further sub divided into four serotypes (A, B, C, D).1 An additional class, serotype, AD, has been shown to be diploid and results from serotype A and D hybridization.2,3 A relationship between mating type and virulence, with MATa found to be more virulent than MATa strains, was reported by Kwon Chung et al.4 Recent data also document a different degree of virulence among serotypes and mating types.5–7 In an attempt to further elucidate the relationship between mating type, serotype, and virulence, representative haploid and hybrid strains of each C. neoformans serotype were compared in terms of virulence factors and pathogenicity in an experimental murine model of cryptococcosis.
Strains were grown on YPD agar (1% yeast extract, 2% peptone, 2% dextrose, 2% agar) for 24 h at 30 8C before each assay. Phospholipase activity was tested by inoculating 1!108 yeast/ml on egg yolk agar.10 Cultures were observed for formation of a precipitate around the colony for up to 14 days at 30 8C. Melanin production was tested by dropping 1 ml of yeast suspension (1!108 mlK1) onto L-DOPA agar using a Steer inoculator.11 The development of a dark brown colony within 24–48 h at 30 8C indicated a positive reaction. Capsule formation was tested on low iron medium.11 Plates were inoculated by dropping 1 ml of yeast suspension (1! 108 mlK1) onto plates using a Steer inoculator and incubating for 7 days at 25 8C. The capsule was
Materials and methods Sources of isolates and characterization A total of 15 strains of C. neoformans were selected in order to have representative strains of each mating type allelic pattern (Table 1). Eleven strains were isolated from humans, two from the environment and two others were laboratory strains. All the isolates were previously serotyped by slide agglutination with the Crypto Check kit (Iatron Labs., Tokyo, Japan) and genotyped by PCR fingerprinting.8 Mating type was determined by PCR of MFa, MFa and STE20 alleles.3 DNA content was measured by flow cytometry.9
Table 1 studied
List of the Cryptococcus neoformans strains
Straina
Mating type allelic patternb
Sourcec
H99d UA1993 IUM96-2828e IUM99-3617f JEC20g IUM99-5509 JEC21g UA477 UA491 UA4223 CBS132h UA486 UA1094 UA2341 UA2715
Aa Aa Aa Aa Da Da Da Da Da Da Aa/Da Aa/Da Aa/Da Aa/Da Aa/Da
CSF (HIVK) CSF (HIVK) Soil CSF (HIVK) Laboratory CSF (HIVC) Laboratory Blood (HIVC) Blood (HIVC) CSF (HIVK) Peach juice CSF (HIVC) CSF (HIVC) CSF (HIVC) CSF (HIVC)
a
IUM, Istituto di Igiene e Medicina Preventiva, Universita ` degli Studi, Milano, Italy; NIH, National Institutes of Health, Bethesda, MD, USA; CBS, Centraalbureau voor Schimmelcultures, Baarn, The Netherlands; UA, Universita ` degli Studi, Ancona, Italy. b Mating type was determined by PCR of MFa, MFa, STE20a, and STE20a genes specific for serotype A and D. c CSF, cerebrospinal fluid. d Strain kindly supplied by J. Perfect, Duke University, Durham, NC, USA. e Viviani et al., Med Mycol 2001; 39, 383–6. f Viviani et al. Emerg Infect Dis 2003; 9, 1179–80. g Kwon Chung et al. Infect Immun 1992; 60, 602–5. h Sanfelice Cryptococcus neoformans type strain.
12
F. Barchiesi et al.
visualized under the microscope in a suspension of India ink.
Animal studies A murine model of systemic cryptococcosis was established in CD1 male mice (weight, 25 g; Charles River Laboratories, Calco, Italy) by intravenous injection of viable yeast cells grown overnight in brain-heart infusion broth. In survival studies, the mice were challenged with approximately 5! 105 CFU/mouse, using 10 animals per each group. The mice were observed throughout day 60, and deaths were recorded daily. In tissue burden studies, the mice were challenged with approximately 5!104 CFU/mouse. After 6 and 12 days post-infection, five mice in each group were euthanized by CO2-induced asphyxia, and the number of viable CFU per gram of brain and lungs of each animal was determined by quantitative plating of organ homogenates on Sabouraud dextrose agar (SDA) plates. A selected group of isolates were also tested in mice by using a model of pulmonary cryptococcosis. Briefly, mice were anesthetized with ketamine. A 6 ml droplet containing approximately 1!107 CFU/mouse was placed on the nares while the diaphragm was compressed. The diaphragm was released and the mouse inhaled the droplet. Either in the intravenous or pulmonary model of infection, the CFU/mouse was confirmed by plating the final inoculum onto SDA plates. Animal experiments were conducted with the approval of the University of Ancona ethics committee.
Statistical analysis Survival data from the mouse experiments were analyzed by using a Kruskal–Wallis test. Student’s t test was used in evaluating the yeast colony counts from mice. A P!0.05 was considered significant.
Results All strains, with the exception of JEC20 (serotype D MATa), produced melanin (Fig. 1A). Similarly, all strains were encapsulated (data not shown) and produced phospholipase (Fig. 1B). Results of survival studies are presented in Table 2. The Aa strains, UA1993 and H99, showed 100% mortality between day 26 and 29 post-infection, respectively (PZ0.5). There was a marked difference in the survival of mice infected with the 4 serotype D MATa isolates. Only UA4223 yielded 100% mortality within the
Figure 1 Analysis of virulence factors. (A) Melanin production on L-DOPA agar. (B) Phospholipase production on egg yolk agar. Negative controls, Candida lusitaniae and Candida krusei, are the last two samples in the right bottom of the plates both in figures A and B.
study period. Mortality among the other 3 isolates ranged from 10 to 70%. A clear rank order of pathogenicity within this group of strains was found with UA4223OJEC21OUA477yUA491. None of the mice infected with JEC20 (serotype D MATa) succumbed within the study period. Although 40% of mice infected with the other serotype D MATa (IUM99-5509) died, the difference in percentage of mortality caused by the two strains did not reach a statistical significance (PZ0.055). A mortality of 40 and 60% resulted from infection with the two serotype A MATa isolates, IUM96-2828 and IUM993617, respectively (PZ0.2). For the hybrid strains, the three Aa/Da types caused 100% mortality within 60 days (range 25–45 days). The survival of mice infected with UA 486, but not with CBS 132, was significantly longer than that observed in mice infected with UA 1094 (PZ 0.03). Strains presenting the Aa/Da mating type pattern caused 30% (UA2715) and 90% (UA2341) mortality within the study period (PZ0.003). Fig. 2 shows cumulative survivals of mice infected with all strains grouped by mating type. Only Aa and Aa/Da types yielded 100% of mortality within the study period. Mortality of the other types ranged from 20% (Da types) and 60% (Aa/Da). No difference was found in the survival rate between the groups of mice infected with Aa and Aa/Da types (PZ0.1), while both types were found to be significantly more pathogenic than the other four mating types (P!0.0001). Similarly, no differences were found in the survival rates among the groups of mice infected with Aa/Da, Da and Aa types (P ranging from 0.1 to 0.5). While both Aa/Da and Da types were found to be significantly more pathogenic than Da types (PZ0.01 and 0.03, respectively), the survival rate of mice infected with Aa types did not differ with that of mice infected with Da types (PZ0.165). Since we observed a wide virulence variation among Da types, these isolates were also tested in a
Pathogenicity of Cryptococcus neoformans ser. A, D and AD Table 2
13
Survival of mice infected intravenously with equal number of yeast cells
Strains
Mating type allelic pattern
Mortality on day 60 post-infection (%)
Mean survival (daysGSD)
UA1993 H99 IUM96-2828 IUM99-3617 UA491 UA477 JEC21 UA4223 JEC20 IUM99-5509 UA486 CBS132 UA1094 UA2715 UA2341
Aa Aa Aa Aa Da Da Da Da Da Da Aa/Da Aa/Da Aa/Da Aa/Da Aa/Da
100 100 40 60 10 30 70 100 0 40 100 100 100 30 90
19G4 21G5 53G11 45G17 58G11 56G11 43G16a 30G8a,b O60 50G17 26G10 26G12 18G5c 51G17 34G15d
a b c d
vs. UA 491 and UA 477. vs. JEC21. vs. UA 486. vs. UA 2715.
pulmonary model of murine cryptococcosis. The results are shown on Fig. 3. Again, C. neoformans 4223 was the most pathogenic strain showing 60% of mortality on day 60, while the percent of mortality for the other three strains ranged from 10 to 50% (P!0.05, 4223 vs. 491 and JEC 21). To investigate the progression of fungal infections caused by strains belonging to different mating types, six strains were selected for tissue burden experiments (Table 3). The mice infected with H99 (Aa) showed the highest increase of CFU/g in the brain tissue between day 6 and 12 postinfection (P ranging from !0.0001 to 0.038 vs. all
Figure 2
strains) followed by the mice infected with UA1094 (Aa/Da). Similar to that observed in the brain, the mice infected with H99 and UA1094 showed the highest increase of CFU/g of lung between day 6 and 12 post-infection (P ranging from !0.0001 to 0.021 vs. IUM 96-2828, IUM 99-5509 and UA 1094).
Discussion To our knowledge this is the first study in which the representative haploid and hybrid strains of each C. neoformans var. neoformans serotype were
Cumulative survivals of mice infected intravenously with all the strains grouped by mating type.
14
F. Barchiesi et al.
Figure 3
Survivals of mice infected intranasally with the four strains belonging to Da types.
compared in terms of virulence factors and pathogenicity in an experimental murine model of systemic cryptococcosis. The majority of these strains were clinical isolates. Our results showed that haploid and hybrid strains presenting the Aa mating type locus are highly virulent in the murine tail vein injection model. These results are consistent with data reported by Chaturvedi et al. who found that eight Aa/Da hybrid strains tested in an intravenous animal infection model were all as virulent as the highly pathogenic strain H99.6 Similarly, although in a different murine model of infection (inhalation), Lengeler et al. tested four hybrid strains (two strains each of Aa/Da and Aa/ Da) and found that one of the two Aa/Da strains was virulent, but less than H99.3 The remaining three strains were moderately virulent. The second Aa/ Da strain was CBS132, which in our model gave 100% mortality as did H99. The different route of inoculation may be the cause of the divergence in the results for this strain. When yeasts are inoculated by inhalation, several factors play an important role in the host infection: the yeast size (quite variable in C. neoformans and crucial to reach the alveoli), the capacity of adhesion to the pulmonary epithelial cells and the ability to invade the host. In the tail vein injection model there likely are other factors distinct from a pulmonary infection route that lead to a different response to infection. Though our sample size is admittedly small, the results also showed that haploid Aa strains are more virulent than haploid Aa and Da strains. Nonetheless, these results are consistent with data from two other studies which showed that haploid Aa
strains, 125.91 and IUM96-2828, were markedly less virulent than H99.12,13 In a recent report, Nielsen et al., showed the absence of relationship between virulence and mating type in serotype A isolates.7 However, the strains used in this study were heavily transformed in order to improve their fertility and to obtain congenic strains. In addition, such congenic strains have only the H99 background, the most virulent of the two parent strains. With regard to the Da haploid strains, they displayed a wide and heterogeneous level of pathogenicity as did the Aa/Da hybrid strains. This finding was confirmed by the additional model of Table 3 Yeast cells count increase between day 6 and 12 post-infection in tissues of mice infected with six strains of Cryptococcus neoformans Strain
Mating type allelic pattern
Mean CFUs increase (log10GSD) in the following tissues Brain
H99 IUM962828 UA4223 IUM995509 UA1094 UA2715 a b c d e
Lung
Aa Aa
1.7G0.3 0.9G0.4
1.1G0.2 b,c,d 0.1G0.0
PDF Da Da
0.7G0.2 0.7G0.8
0.8G0.3d 0.7G0.0d
Aa/Da Aa/Da
1.3G0.2b,e 0.7G0.1
1.3G0.4b,c,d 0.7G0.2d
vs. Any strain. vs. UA2715. vs. IUM99-5509. vs. IUM96-2828. vs. UA4223.
a
Pathogenicity of Cryptococcus neoformans ser. A, D and AD pulmonary infection. Kwon-Chung and Hill reported similar variation during the first study on virulence and mating type.1 The reason for this variability was attributed to a difference in karyotype arrangement and thus in the genome background. This possibility was addressed by constructing congenic MATa and MATa serotype D strains and testing them for virulence using the mouse model. The results showed that the MATa strain was more virulent than the MATa strain.4 The limit of this study was that only one genomic background was tested in the congenic strains. In a more recent review, it was suggested that multiple congenic strains, constructed in different backgrounds, should be tested to evaluate the association of virulence with mating type.14 These strains have been successfully constructed and are presently being tested (unpublished data). The in vitro results of melanin, capsule and phospholipase production showed that all the tested strains expressed these virulence factors, although to different extents. This variability was not associated with mating type or virulence. Similarly, the variability in virulence of Da and Aa/Da strains was not correlated with different levels of production of these factors. Overall, these findings suggest that while mating type can influence virulence, factors such as ploidy and serotype also must be considered. This conclusion supports earlier studies of serotype and mating type specific genes. For example, disruption of the STE20a gene in serotype A strain H99 was found to cause a cytokinesis defect and low capsule production, which likely contributed to the strong reduction in virulence.15 When the same gene was disrupted in serotype D congenic strains there was little effect on virulence. Similar results were observed for the STE12 homologs. In this case, the serotype D ste12a disruptants were attenuated for virulence, whereas serotype A ste12a mutants were fully virulent.5,16 Lastly, it is clear that hybrid strains comprise an important component of C. neoformans isolates. They are readily recovered from clinical isolates and as our results show, are virulent in experimental models. The clinical significance of these isolates, both in frequency of recovery and virulence, suggests more detailed studies of these strains may be warranted in the future.
Acknowledgements M.A.V. is supported by a grant from MIUR (2001065221). B.L.W. is a Burroughs-Wellcome
15
New Investigator in Molecular Pathogenic Mycology and is supported by US Public Health Service Grant R29AI43522 and 1R01AI54946 from the National Institutes of Health. G.S. is supported by a grant from MIUR (2001065221) and by a grant from Istituto Superiore di Sanita ` (III AIDS project contract no. 50C.29), Rome, Italy.
References 1. Kwon-Chung KJ, Hill WB. Sexuality and pathogenicity of Filobasidiella neoformans (Cryptococcus neoformans). In: Vanbreuseghem R, De Vroy C, editors. Sexuality and pathogenicity of fungi. Paris: Page Masson; 1981. 2. Cogliati M, Esposto MC, Clarke DL, Wickes BL, Viviani MA. Origin of Cryptococcus neoformans var. neoformans diploid strains. J Clin Microbiol 2001;39:3889–94. 3. Lengeler KB, Cox GM, Heitman J. Serotype AD strains of Cryptococcus neoformans are diploid or aneuploid and are heterozygous at the mating-type locus. Infect Immun 2001; 69:115–22. 4. Kwon-Chung KJ, Edman JC, Wickes BL. Genetic association of mating types and virulence in Cryptococcus neoformans. Infect Immun 1992;60:602–5. 5. Chang YC, Wickes BL, Miller GF, Penoyer LA, Kwon-Chung KJ. Cryptococcus neoformans STE12 alpha regulates virulence but is not essential for mating. J Exp Med 2000;191:871–82. 6. Chaturvedi V, Fan J, Stein B, Behr MJ, Samsonoff WA, Wickes BL, Chaturvedi S. Molecular and genetic analyses of mating pheromones reveal intervariety mating or hybridization in Cryptococcus neoformans. Infect Immun 2002;70: 5225–35. 7. Nielsen K, Cox GM, Wang P, Toffaletti DL, Perfect JR, Heitman J. Sexual cycle of Cryptococcus neoformans var. grubii and virulence of congenic a and isolates. Infect Immun 2003;71:4831–41. 8. Viviani MA, Wen H, Roverselli A, Caldarelli-Stefano R, Cogliati M, Ferrante P, Tortorano AM. Identification by polymerase chain reaction fingerprinting of Cryptococcus neoformans serotype AD. J Med Vet Mycol 1997;35:355–60. 9. Tanaka R, Taguchi H, Takeo K, Miyaji M, Nishimura K. Determination of ploidy in Cryptococcus neoformans by flow cytometry. J Med Vet Mycol 1996;34:299–301. 10. Price MF, Wilkinson ID, Gentry LO. Plate method for detection of phospholipase activity in Candida albicans. Sabouraudia 1982;20:7–14. 11. Kwon-Chung KJ, Polacheck I, Popkin TJ. Melanin-lacking mutants of Cryptococcus neoformans and their virulence for mice. J Bacteriol 1982;150:1414–21. 12. Keller SM, Viviani MA, Esposto MC, Cogliati M, Wickes BL. Molecular and genetic characterization of a serotype A MATa Cryptococcus neoformans isolate. Microbiology 2003;149: 131–42. 13. Lengeler KB, Wang P, Cox GM, Perfect JR, Heitman J. Identification of the MATa mating-type locus of Cryptococcus neoformans reveals a serotype A MATa strain thought to have been extinct. Proc Natl Acad Sci USA 2000;97: 14455–60. 14. Wickes BL. The role of mating type and morphology in Cryptococcus neoformans pathogenesis. J Med Microbiol 2002;292:1–17. 15. Wang P, Nichols CB, Lengeler KB, Cardenas ME, Cox GM,
16 Perfect JR, Heitman J. Mating-type-specific and nonspecific PAK kinases play shared and divergent roles in Cryptococcus neoformans. Eukaryot Cell 2002;1:257–72. 16. Yue C, Cavallo LM, Alspaugh JA, Wang P, Cox GM, Perfect JR,
F. Barchiesi et al. Heitman J. The STE12 alpha homolog is required for haploid filamentation but largely dispensable for mating and virulence in Cryptococcus neoformans. Genetics 1999;153: 1601–15.
www.elsevierhealth.com/journals/jinf
Comparative analysis of pathogenicity of Cryptococcus neoformans serotypes A, D and AD in murine cryptococcosis F. Barchiesia,*, M. Cogliatib, M.C. Espostob, E. Spreghinia, A.M. Schimizzia, B.L. Wickesc, G. Scalisea, M.A. Vivianib a
` Politecnica delle Marche, Azienda Istituto di Malattie Infettive e Medicina Pubblica, Universita Ospedaliera Umberto I8, Via Conca 60020 Torrette di Ancona, Ancona, Italy b ` degli Studi di Milano, IRCCS Ospedale Maggiore, Istituto di Igiene e Medicina Preventiva, Universita Milano, Italy c Department of Microbiology, University of Texas Health Science Center at San Antonio, San Antonio, TX 78229-3900, USA Accepted 21 July 2004 Available online 16 September 2004
KEYWORDS Cryptococcus neoformans; Serotype; Diploid; Haploid; Hybrid
Abstract Objectives. To characterize the pathogenicity of 15 strains of Cryptococcus neoformans belonging to several serotype/mating type allele patterns (Da, Da, Aa, Aa, Aa/Da and Da/Aa) in experimental models of murine cryptococcosis. Methods. CD1-infected mice were examined for survival and fungal loads in either brain or lung during the course of infection. Results. All strains, with the exception of one Da strain, produced melanin in vitro. Similarly, all strains were encapsulated and produced phospholipase. When CD1 mice were challenged intravenously (i.v.) with 5!105 CFU/mouse and observed for 60 days post-infection, a significant variation of mortality rate was observed among mice infected with different strains. Aa and Aa/Da strains all produced 100% mortality within the study period with mean survivals significantly shorter than those of mice infected with strains belonging to any other allele type (P!0.0001). A wide range of pathogenicity was shown by haploid and diploid strains presenting Da allele. This finding was confirmed by an intranasal model of challenge. To investigate the progression of infection, the mice were challenged i.v. with 5!104 CFU/mouse and tissue burden experiments (brain and lung) were performed on days 6 and 12 postinfection. Only the mice infected with Aa and Aa/Da strains showed a O1 log10 increase of CFU/g in both tissues throughout the study period. Conclusions. Our results suggest that the presence of the Aa mating type allele in either haploid or diploid strains is correlated with virulence, while the presence of the Aa or Da allele in haploid strains is associated with moderate or no virulence.
* Corresponding author. Tel.: C39-071-5963467; fax: C39-071-5963468. E-mail address: [email protected] (F. Barchiesi). 0163-4453/$30.00 Q 2004 The British Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jinf.2004.07.013
Pathogenicity of Cryptococcus neoformans ser. A, D and AD
11
Finally, either haploid or diploid strains presenting Da allele vary in virulence. Q 2004 The British Infection Society. Published by Elsevier Ltd. All rights reserved.
Introduction
Analysis of virulence factors
The yeast Cryptococcus neoformans is an encapsulated fungus that can cause infection in both immunocompetent and immunocompromised individuals. Its incidence has dramatically increased during the last two decades because of the AIDS epidemic, although impaired cellmediated immunity in general represents the main predisposing factor for development of cryptococcosis. Meningoencephalitis is the most frequent clinical presentation, although any organ can be affected. Cryptococcus neoformans has two mating types, MATa and MATa, and two varieties, var. neoformans and var. gatti, which can be further sub divided into four serotypes (A, B, C, D).1 An additional class, serotype, AD, has been shown to be diploid and results from serotype A and D hybridization.2,3 A relationship between mating type and virulence, with MATa found to be more virulent than MATa strains, was reported by Kwon Chung et al.4 Recent data also document a different degree of virulence among serotypes and mating types.5–7 In an attempt to further elucidate the relationship between mating type, serotype, and virulence, representative haploid and hybrid strains of each C. neoformans serotype were compared in terms of virulence factors and pathogenicity in an experimental murine model of cryptococcosis.
Strains were grown on YPD agar (1% yeast extract, 2% peptone, 2% dextrose, 2% agar) for 24 h at 30 8C before each assay. Phospholipase activity was tested by inoculating 1!108 yeast/ml on egg yolk agar.10 Cultures were observed for formation of a precipitate around the colony for up to 14 days at 30 8C. Melanin production was tested by dropping 1 ml of yeast suspension (1!108 mlK1) onto L-DOPA agar using a Steer inoculator.11 The development of a dark brown colony within 24–48 h at 30 8C indicated a positive reaction. Capsule formation was tested on low iron medium.11 Plates were inoculated by dropping 1 ml of yeast suspension (1! 108 mlK1) onto plates using a Steer inoculator and incubating for 7 days at 25 8C. The capsule was
Materials and methods Sources of isolates and characterization A total of 15 strains of C. neoformans were selected in order to have representative strains of each mating type allelic pattern (Table 1). Eleven strains were isolated from humans, two from the environment and two others were laboratory strains. All the isolates were previously serotyped by slide agglutination with the Crypto Check kit (Iatron Labs., Tokyo, Japan) and genotyped by PCR fingerprinting.8 Mating type was determined by PCR of MFa, MFa and STE20 alleles.3 DNA content was measured by flow cytometry.9
Table 1 studied
List of the Cryptococcus neoformans strains
Straina
Mating type allelic patternb
Sourcec
H99d UA1993 IUM96-2828e IUM99-3617f JEC20g IUM99-5509 JEC21g UA477 UA491 UA4223 CBS132h UA486 UA1094 UA2341 UA2715
Aa Aa Aa Aa Da Da Da Da Da Da Aa/Da Aa/Da Aa/Da Aa/Da Aa/Da
CSF (HIVK) CSF (HIVK) Soil CSF (HIVK) Laboratory CSF (HIVC) Laboratory Blood (HIVC) Blood (HIVC) CSF (HIVK) Peach juice CSF (HIVC) CSF (HIVC) CSF (HIVC) CSF (HIVC)
a
IUM, Istituto di Igiene e Medicina Preventiva, Universita ` degli Studi, Milano, Italy; NIH, National Institutes of Health, Bethesda, MD, USA; CBS, Centraalbureau voor Schimmelcultures, Baarn, The Netherlands; UA, Universita ` degli Studi, Ancona, Italy. b Mating type was determined by PCR of MFa, MFa, STE20a, and STE20a genes specific for serotype A and D. c CSF, cerebrospinal fluid. d Strain kindly supplied by J. Perfect, Duke University, Durham, NC, USA. e Viviani et al., Med Mycol 2001; 39, 383–6. f Viviani et al. Emerg Infect Dis 2003; 9, 1179–80. g Kwon Chung et al. Infect Immun 1992; 60, 602–5. h Sanfelice Cryptococcus neoformans type strain.
12
F. Barchiesi et al.
visualized under the microscope in a suspension of India ink.
Animal studies A murine model of systemic cryptococcosis was established in CD1 male mice (weight, 25 g; Charles River Laboratories, Calco, Italy) by intravenous injection of viable yeast cells grown overnight in brain-heart infusion broth. In survival studies, the mice were challenged with approximately 5! 105 CFU/mouse, using 10 animals per each group. The mice were observed throughout day 60, and deaths were recorded daily. In tissue burden studies, the mice were challenged with approximately 5!104 CFU/mouse. After 6 and 12 days post-infection, five mice in each group were euthanized by CO2-induced asphyxia, and the number of viable CFU per gram of brain and lungs of each animal was determined by quantitative plating of organ homogenates on Sabouraud dextrose agar (SDA) plates. A selected group of isolates were also tested in mice by using a model of pulmonary cryptococcosis. Briefly, mice were anesthetized with ketamine. A 6 ml droplet containing approximately 1!107 CFU/mouse was placed on the nares while the diaphragm was compressed. The diaphragm was released and the mouse inhaled the droplet. Either in the intravenous or pulmonary model of infection, the CFU/mouse was confirmed by plating the final inoculum onto SDA plates. Animal experiments were conducted with the approval of the University of Ancona ethics committee.
Statistical analysis Survival data from the mouse experiments were analyzed by using a Kruskal–Wallis test. Student’s t test was used in evaluating the yeast colony counts from mice. A P!0.05 was considered significant.
Results All strains, with the exception of JEC20 (serotype D MATa), produced melanin (Fig. 1A). Similarly, all strains were encapsulated (data not shown) and produced phospholipase (Fig. 1B). Results of survival studies are presented in Table 2. The Aa strains, UA1993 and H99, showed 100% mortality between day 26 and 29 post-infection, respectively (PZ0.5). There was a marked difference in the survival of mice infected with the 4 serotype D MATa isolates. Only UA4223 yielded 100% mortality within the
Figure 1 Analysis of virulence factors. (A) Melanin production on L-DOPA agar. (B) Phospholipase production on egg yolk agar. Negative controls, Candida lusitaniae and Candida krusei, are the last two samples in the right bottom of the plates both in figures A and B.
study period. Mortality among the other 3 isolates ranged from 10 to 70%. A clear rank order of pathogenicity within this group of strains was found with UA4223OJEC21OUA477yUA491. None of the mice infected with JEC20 (serotype D MATa) succumbed within the study period. Although 40% of mice infected with the other serotype D MATa (IUM99-5509) died, the difference in percentage of mortality caused by the two strains did not reach a statistical significance (PZ0.055). A mortality of 40 and 60% resulted from infection with the two serotype A MATa isolates, IUM96-2828 and IUM993617, respectively (PZ0.2). For the hybrid strains, the three Aa/Da types caused 100% mortality within 60 days (range 25–45 days). The survival of mice infected with UA 486, but not with CBS 132, was significantly longer than that observed in mice infected with UA 1094 (PZ 0.03). Strains presenting the Aa/Da mating type pattern caused 30% (UA2715) and 90% (UA2341) mortality within the study period (PZ0.003). Fig. 2 shows cumulative survivals of mice infected with all strains grouped by mating type. Only Aa and Aa/Da types yielded 100% of mortality within the study period. Mortality of the other types ranged from 20% (Da types) and 60% (Aa/Da). No difference was found in the survival rate between the groups of mice infected with Aa and Aa/Da types (PZ0.1), while both types were found to be significantly more pathogenic than the other four mating types (P!0.0001). Similarly, no differences were found in the survival rates among the groups of mice infected with Aa/Da, Da and Aa types (P ranging from 0.1 to 0.5). While both Aa/Da and Da types were found to be significantly more pathogenic than Da types (PZ0.01 and 0.03, respectively), the survival rate of mice infected with Aa types did not differ with that of mice infected with Da types (PZ0.165). Since we observed a wide virulence variation among Da types, these isolates were also tested in a
Pathogenicity of Cryptococcus neoformans ser. A, D and AD Table 2
13
Survival of mice infected intravenously with equal number of yeast cells
Strains
Mating type allelic pattern
Mortality on day 60 post-infection (%)
Mean survival (daysGSD)
UA1993 H99 IUM96-2828 IUM99-3617 UA491 UA477 JEC21 UA4223 JEC20 IUM99-5509 UA486 CBS132 UA1094 UA2715 UA2341
Aa Aa Aa Aa Da Da Da Da Da Da Aa/Da Aa/Da Aa/Da Aa/Da Aa/Da
100 100 40 60 10 30 70 100 0 40 100 100 100 30 90
19G4 21G5 53G11 45G17 58G11 56G11 43G16a 30G8a,b O60 50G17 26G10 26G12 18G5c 51G17 34G15d
a b c d
vs. UA 491 and UA 477. vs. JEC21. vs. UA 486. vs. UA 2715.
pulmonary model of murine cryptococcosis. The results are shown on Fig. 3. Again, C. neoformans 4223 was the most pathogenic strain showing 60% of mortality on day 60, while the percent of mortality for the other three strains ranged from 10 to 50% (P!0.05, 4223 vs. 491 and JEC 21). To investigate the progression of fungal infections caused by strains belonging to different mating types, six strains were selected for tissue burden experiments (Table 3). The mice infected with H99 (Aa) showed the highest increase of CFU/g in the brain tissue between day 6 and 12 postinfection (P ranging from !0.0001 to 0.038 vs. all
Figure 2
strains) followed by the mice infected with UA1094 (Aa/Da). Similar to that observed in the brain, the mice infected with H99 and UA1094 showed the highest increase of CFU/g of lung between day 6 and 12 post-infection (P ranging from !0.0001 to 0.021 vs. IUM 96-2828, IUM 99-5509 and UA 1094).
Discussion To our knowledge this is the first study in which the representative haploid and hybrid strains of each C. neoformans var. neoformans serotype were
Cumulative survivals of mice infected intravenously with all the strains grouped by mating type.
14
F. Barchiesi et al.
Figure 3
Survivals of mice infected intranasally with the four strains belonging to Da types.
compared in terms of virulence factors and pathogenicity in an experimental murine model of systemic cryptococcosis. The majority of these strains were clinical isolates. Our results showed that haploid and hybrid strains presenting the Aa mating type locus are highly virulent in the murine tail vein injection model. These results are consistent with data reported by Chaturvedi et al. who found that eight Aa/Da hybrid strains tested in an intravenous animal infection model were all as virulent as the highly pathogenic strain H99.6 Similarly, although in a different murine model of infection (inhalation), Lengeler et al. tested four hybrid strains (two strains each of Aa/Da and Aa/ Da) and found that one of the two Aa/Da strains was virulent, but less than H99.3 The remaining three strains were moderately virulent. The second Aa/ Da strain was CBS132, which in our model gave 100% mortality as did H99. The different route of inoculation may be the cause of the divergence in the results for this strain. When yeasts are inoculated by inhalation, several factors play an important role in the host infection: the yeast size (quite variable in C. neoformans and crucial to reach the alveoli), the capacity of adhesion to the pulmonary epithelial cells and the ability to invade the host. In the tail vein injection model there likely are other factors distinct from a pulmonary infection route that lead to a different response to infection. Though our sample size is admittedly small, the results also showed that haploid Aa strains are more virulent than haploid Aa and Da strains. Nonetheless, these results are consistent with data from two other studies which showed that haploid Aa
strains, 125.91 and IUM96-2828, were markedly less virulent than H99.12,13 In a recent report, Nielsen et al., showed the absence of relationship between virulence and mating type in serotype A isolates.7 However, the strains used in this study were heavily transformed in order to improve their fertility and to obtain congenic strains. In addition, such congenic strains have only the H99 background, the most virulent of the two parent strains. With regard to the Da haploid strains, they displayed a wide and heterogeneous level of pathogenicity as did the Aa/Da hybrid strains. This finding was confirmed by the additional model of Table 3 Yeast cells count increase between day 6 and 12 post-infection in tissues of mice infected with six strains of Cryptococcus neoformans Strain
Mating type allelic pattern
Mean CFUs increase (log10GSD) in the following tissues Brain
H99 IUM962828 UA4223 IUM995509 UA1094 UA2715 a b c d e
Lung
Aa Aa
1.7G0.3 0.9G0.4
1.1G0.2 b,c,d 0.1G0.0
PDF Da Da
0.7G0.2 0.7G0.8
0.8G0.3d 0.7G0.0d
Aa/Da Aa/Da
1.3G0.2b,e 0.7G0.1
1.3G0.4b,c,d 0.7G0.2d
vs. Any strain. vs. UA2715. vs. IUM99-5509. vs. IUM96-2828. vs. UA4223.
a
Pathogenicity of Cryptococcus neoformans ser. A, D and AD pulmonary infection. Kwon-Chung and Hill reported similar variation during the first study on virulence and mating type.1 The reason for this variability was attributed to a difference in karyotype arrangement and thus in the genome background. This possibility was addressed by constructing congenic MATa and MATa serotype D strains and testing them for virulence using the mouse model. The results showed that the MATa strain was more virulent than the MATa strain.4 The limit of this study was that only one genomic background was tested in the congenic strains. In a more recent review, it was suggested that multiple congenic strains, constructed in different backgrounds, should be tested to evaluate the association of virulence with mating type.14 These strains have been successfully constructed and are presently being tested (unpublished data). The in vitro results of melanin, capsule and phospholipase production showed that all the tested strains expressed these virulence factors, although to different extents. This variability was not associated with mating type or virulence. Similarly, the variability in virulence of Da and Aa/Da strains was not correlated with different levels of production of these factors. Overall, these findings suggest that while mating type can influence virulence, factors such as ploidy and serotype also must be considered. This conclusion supports earlier studies of serotype and mating type specific genes. For example, disruption of the STE20a gene in serotype A strain H99 was found to cause a cytokinesis defect and low capsule production, which likely contributed to the strong reduction in virulence.15 When the same gene was disrupted in serotype D congenic strains there was little effect on virulence. Similar results were observed for the STE12 homologs. In this case, the serotype D ste12a disruptants were attenuated for virulence, whereas serotype A ste12a mutants were fully virulent.5,16 Lastly, it is clear that hybrid strains comprise an important component of C. neoformans isolates. They are readily recovered from clinical isolates and as our results show, are virulent in experimental models. The clinical significance of these isolates, both in frequency of recovery and virulence, suggests more detailed studies of these strains may be warranted in the future.
Acknowledgements M.A.V. is supported by a grant from MIUR (2001065221). B.L.W. is a Burroughs-Wellcome
15
New Investigator in Molecular Pathogenic Mycology and is supported by US Public Health Service Grant R29AI43522 and 1R01AI54946 from the National Institutes of Health. G.S. is supported by a grant from MIUR (2001065221) and by a grant from Istituto Superiore di Sanita ` (III AIDS project contract no. 50C.29), Rome, Italy.
References 1. Kwon-Chung KJ, Hill WB. Sexuality and pathogenicity of Filobasidiella neoformans (Cryptococcus neoformans). In: Vanbreuseghem R, De Vroy C, editors. Sexuality and pathogenicity of fungi. Paris: Page Masson; 1981. 2. Cogliati M, Esposto MC, Clarke DL, Wickes BL, Viviani MA. Origin of Cryptococcus neoformans var. neoformans diploid strains. J Clin Microbiol 2001;39:3889–94. 3. Lengeler KB, Cox GM, Heitman J. Serotype AD strains of Cryptococcus neoformans are diploid or aneuploid and are heterozygous at the mating-type locus. Infect Immun 2001; 69:115–22. 4. Kwon-Chung KJ, Edman JC, Wickes BL. Genetic association of mating types and virulence in Cryptococcus neoformans. Infect Immun 1992;60:602–5. 5. Chang YC, Wickes BL, Miller GF, Penoyer LA, Kwon-Chung KJ. Cryptococcus neoformans STE12 alpha regulates virulence but is not essential for mating. J Exp Med 2000;191:871–82. 6. Chaturvedi V, Fan J, Stein B, Behr MJ, Samsonoff WA, Wickes BL, Chaturvedi S. Molecular and genetic analyses of mating pheromones reveal intervariety mating or hybridization in Cryptococcus neoformans. Infect Immun 2002;70: 5225–35. 7. Nielsen K, Cox GM, Wang P, Toffaletti DL, Perfect JR, Heitman J. Sexual cycle of Cryptococcus neoformans var. grubii and virulence of congenic a and isolates. Infect Immun 2003;71:4831–41. 8. Viviani MA, Wen H, Roverselli A, Caldarelli-Stefano R, Cogliati M, Ferrante P, Tortorano AM. Identification by polymerase chain reaction fingerprinting of Cryptococcus neoformans serotype AD. J Med Vet Mycol 1997;35:355–60. 9. Tanaka R, Taguchi H, Takeo K, Miyaji M, Nishimura K. Determination of ploidy in Cryptococcus neoformans by flow cytometry. J Med Vet Mycol 1996;34:299–301. 10. Price MF, Wilkinson ID, Gentry LO. Plate method for detection of phospholipase activity in Candida albicans. Sabouraudia 1982;20:7–14. 11. Kwon-Chung KJ, Polacheck I, Popkin TJ. Melanin-lacking mutants of Cryptococcus neoformans and their virulence for mice. J Bacteriol 1982;150:1414–21. 12. Keller SM, Viviani MA, Esposto MC, Cogliati M, Wickes BL. Molecular and genetic characterization of a serotype A MATa Cryptococcus neoformans isolate. Microbiology 2003;149: 131–42. 13. Lengeler KB, Wang P, Cox GM, Perfect JR, Heitman J. Identification of the MATa mating-type locus of Cryptococcus neoformans reveals a serotype A MATa strain thought to have been extinct. Proc Natl Acad Sci USA 2000;97: 14455–60. 14. Wickes BL. The role of mating type and morphology in Cryptococcus neoformans pathogenesis. J Med Microbiol 2002;292:1–17. 15. Wang P, Nichols CB, Lengeler KB, Cardenas ME, Cox GM,
16 Perfect JR, Heitman J. Mating-type-specific and nonspecific PAK kinases play shared and divergent roles in Cryptococcus neoformans. Eukaryot Cell 2002;1:257–72. 16. Yue C, Cavallo LM, Alspaugh JA, Wang P, Cox GM, Perfect JR,
F. Barchiesi et al. Heitman J. The STE12 alpha homolog is required for haploid filamentation but largely dispensable for mating and virulence in Cryptococcus neoformans. Genetics 1999;153: 1601–15.