Gene disruption in Cryptococcus neoformans and Cryptococcus gattii by in vitro transposition
Cryptococcus neoformans and Cryptococcus gattii are basidiomycetous fungi that infect immunocompromised and immunocompetent people. We developed an insertional mutagenesis strategy for these species based on in vitro transposition and we tested the method by disrupting the URA5 gene in a strain of C. neoformans and the CAP10 gene in three strains of C. gattii. We targeted plasmid DNA containing the URA5 gene or plasmid DNA containing the CAP10 gene from genomic libraries from the shotgun sequencing project for the C. gatti strain WM276. In the latter case, the availability of the end sequences of the clones from the assembled genomic sequence allows rapid selection of target genes for disruption. Modified transposons containing the nourseothricin (NAT) or neomycin (Neo) resistance cassettes were randomly inserted into the target DNA by in vitro transposition. The disrupted genes were used for biolistic transformation and homologous integration was subsequently confirmed by PCR and Southern blot analysis. These results demonstrate that the emerging genomic resources, combined with in vitro transposition into plasmid DNAs from shotgun sequencing libraries or cloned PCR products, will facilitate high-throughput genetic analysis in Cryptococcus species.
KeywordsGene disruption Transposon Basidiomycetes Virulence factors
The basidiomycete fungal pathogens Cryptococcus neoformans and Cryptococus gattii infect both immunocompetent and immunocompromised individuals to cause cryptococcosis. Cryptococcus strains infect the central nervous system and cause meningioencephalitis that can be fatal if not aggressively treated. Five different serotypes (A, B, C, D and AD) have been classified based on antigenic differences in the capsule of the fungus. Strains of the B and C serotypes have recently been recognized as a separate species called C. gattii, while serotypes A and D are considered different varieties called C. neoformans var. grubii and C. neoformans var. neoformans, respectively (Kwon-Chung et al. 2002). Serotype A is the most clinically prevalent in cryptococcosis patients in North America, followed by serotype D. Serotype B isolates (C. gattii) were considered to be limited to tropical and subtropical environments (Casadevall and Perfect 1998) but this idea has been called into question by the emergence of infections caused by C. gattii in immunocompetent individuals on Vancouver Island (Hoang et al. 2004).
The fungus produces a polysaccharide capsule that is a well-characterized virulence factor for the pathogen. Other virulence factors include the ability to grow at 37°C, melanin synthesis, and production of urease and phospholipase (Casadevall and Perfect 1998; Cox et al. 2000, 2001; Kwon-Chung and Rhodes 1986). There is considerable interest in determining the genetic and biochemical basis of capsule formation in C. neoformans and C. gattii because the capsule is antiphagocytic and exerts a negative influence on the host immune response (Buchanan and Murphy 1998). By isolating and complementing acapsular mutants, Kwon-Chung and colleagues identified four genes (CAP10, CAP59, CAP60 and CAP64) involved in capsule biosynthesis (Chang et al. 1996; Chang and Kwon-Chung 1994, 1998, 1999). These CAP genes are essential for capsule formation and for virulence in the murine model. Further studies demonstrated that the CAP10 gene product was located in the cytoplasm while Cap60 was located at the nuclear membrane (Chang and Kwon-Chung 1998, 1999). CAP10 is expressed at high levels in late-stationary-phase cells, and the expression level is modulated by the transcriptional factor STE12alpha (Chang and Kwon-Chung 1999). Other genes, such as the CAS genes, also play a role in capsule synthesis (Moyrand et al. 2002).
C. neoformans is an excellent model to study fungal pathogenesis. The fungus has a well-defined sexual cycle, it is predominantly haploid, and congenic alpha- and a mating-type strains are available (Hull and Heitman 2002; Lodge et al. 1994; Sia et al. 2000). The fungus is amenable to a variety of genetic and molecular techniques. For example, gene disruption is readily accomplished by biolistic transformation (Davidson et al. 2000), electroporation (Edman and Kwon-Chung 1990) or Agrobacterium-mediated transformation (Idnurm et al. 2004) to introduce deletion constructs. The reported rates of homologous recombination range from 0.008% to 50% depending on the parameters involved in the experiment (Fox et al. 2001; Salas et al. 1996; Nelson et al. 2003). Selectable markers, such as ADE2 and URA5, have been developed for transformation of auxotrophic recipient strains. Dominant antibiotic resistance markers for hygromycin, noursethricin, and G418 (neomycin) are also available (Cox et al. 1996; Hua et al. 2000). In addition, several animal models to study virulence have been developed (Hull and Heitman 2002; Perfect 2005). These approaches and tools have only recently been applied to the analysis of strains of C. gattii.
Genomic resources are rapidly accumulating for C. neoformans and C. gattii. Specifically, two genome sequences have been described for serotype D strains (Loftus et al. 2005), and sequences have been obtained for a serotype A strain and two serotype B strains (http://www.fungal.genome.duke.edu/; http://www.bcgsc.ca; http://www.broad.mit.edu/annotation/fgi/). In addition, a serial analysis gene expression (SAGE) program is in progress to characterize the transcriptomes of C. neoformans strains under different conditions related to virulence (Steen et al. 2002, 2003; Lian et al. 2004). Microarrays have also recently become available to the research community to allow rapid analysis of gene expression (http://www.genome.wustl.edu/activity/ma/cneoformans/).
Given the accumulating genomic sequence and expression data, it is clear that simple and efficient gene disruption approaches are needed to explore the functions of Cryptococcal genes. Gene disruption by homologous recombination after biolistic transformation has proven successful in C. neoformans and C. gattii. Davidson et al. (2002) developed an overlapping PCR strategy to make constructs with large regions of homology for gene disruption without cloning. Nelson et al. (2003) examined the minimum sequence requirements for efficient homologous recombination and found that symmetric or asymmetrical constructs with 300 bp or more of flanking sequence on each side of the construct were efficiently targeted for gene disruption in C. neoformans var. grubii. In the context of their findings, these authors also proposed that an in vitro transposition strategy coupled with PCR amplification of target genes would facilitate the preparation of deletion constructs. Here we describe the development and application of such a transposon-based strategy to generate disruption constructs without overlapping PCR and cloning in order to rapidly disrupt genes of interest, particularly in C. gattii. This approach was motivated by access to a large bank of plasmid clones with known end sequences from our shotgun sequencing project for the serotype B strain WM276 of C. gattii. Relatively few genes have been disrupted in this species and the combination of in vitro transposition with the genome project should allow high-throughput genetic analysis. To test this approach, plasmids carrying the CAP10 gene were identified from the libraries and the gene was disrupted in three different strains of C. gattii. We also disrupted the well-characterized URA5 gene in the serotype A strain H99 of C. neoformans to confirm the applicability of the protocol in this better characterized serotype.
Materials and methods
Fungal strains, plasmids and media
Serotype A strain H99 (C. neoformans var. grubii) and serotype B strains WM276, NIH444 and A1MR265 (C. gattii) were used in the study. WM276 belongs to subtype VGI of C. gattii, NIH444 and A1MR265 are both in subtype VGII by molecular typing (Kidd et al. 2004, 2005). All strains were maintained on YPD medium (1% yeast extract, 2% peptone, 2% dextrose, and 2% agar). Plasmid pGPS3 was purchased from New England Biolabs, USA Plasmids pJAF1 containing a neomycin resistance cassette, pCH233 containing a noursethricin resistance cassette, and pJMN97-3 carrying a URA5 gene fragment were generously provided by Dr. J. Heitman (Duke University). YPD plates containing neomycin (200 μg/ml) were used for selecting CAP10 disruption transformants and YPD plates containing noursethricin (100 μg/ml) were used to select URA5 disruption transformants.
Modification of the GPS3 transposon
Targets for gene disruption
Two targets chosen for disruption, CAP10 (in serotype B) and URA5 (in serotype A), are easily screened for a phenotype (by India ink staining and growth on 5-FOA, respectively). A stand-alone local BLAST was performed with the end sequence datasets of the WM276 shotgun sequencing libraries CN022 and CN23E to identify the clones containing inserts with the largest portion of CAP10. Both libraries had an average insert size of approximately 2 kb: CN022 contains 35,328 sequenced clones and CN23E contains 24,096 sequenced cloned. In the present study, the clone CN23E63cH10 that contains the full-length CAP10 gene was selected for gene disruption. The plasmid pJMN97-3 carrying the URA5 gene was used to generate the URA5 disruption construct for the serotype A strain H99.
In vitro transposition
Primers used in the study
Transformation, screening and phenotype studies
Biolistic transformation was performed using linear DNA by the method described by Davidson et al. (2000). Transformants were selected on YPD plates containing either nourseothricin (100 μg/ml for ura5 mutants) or neomycin (G418; 200 μg/ml for cap10 mutants). Colony PCR was performed with Extaq polymerase (Takara) to screen for the putative disruptants. The PCR reaction consisted of 30 cycles of 15 s at 94°C, 15 s at 55°C, and 3 min at 72°C with an initial denaturation of 12 min at 94°C and a final extension of 7 min at 72°C. For serotype B strains (WM276, NIH444 and A1MR265), colony PCR was first performed with primers, CAP10intR and CAP10intF (Table 1), and the presence of a PCR band of 240 bps indicated the wild-type genotype (“negative” screening). Those transformants that did not produce a band of 240 bps were subjected to a second round of PCR (“positive” screening) with primers GPS-N and a primer from the genomic sequence of CAP10 gene flanking region (CAP10-outR, Table 1). Putative CAP10 disruptants were verified by Southern blot analysis. Capsule production was induced by growth in low-iron medium (LIM) overnight at 30°C, and the capsule was examined by differential interference contrast (DIC) microscopy after India ink negative staining. For the URA5 disruption, transformants on YPD plates containing nourseothricin (100 μg/ml) were screened for growth on 5-FOA (1 mg/ml), and on YNB minimal medium without uracil (drop-out test), and confirmed by Southern blot analysis.
Southern blot analysis
Genomic DNA from each strain was isolated following the protocol as described by Perkin et al. (1988), and approximately 10 μg of genomic DNA was digested to completion. The fragments were separated by agarose gel electrophoresis and transferred to nylon filters using 10× SSC. Probes were produced using a Random Priming Kit (Amersham Pharmacia Biotech) by incorporation of 32P-dCTP according to the manufacturer’s protocol. The 1.7 kb CAP10 gene-specific probe DNA was produced by digestion of plasmid pCN23E63cH10 with EcoRV. The ~470 bp fragment of URA5 gene was produced by digestion of pJMN97-3 with EcoRV.
We demonstrated the applicability of the in vitro transposition-based strategy by disrupting the CAP10 gene in the serotype B strains WM276, A1MR265 and NIH444, and the URA5 gene in the serotype A strain H99. Two shotgun sequencing libraries, CN022 and CN23E, containing ~2 kb inserts in the BstX1 site of pBR194c and pBR194e, respectively, were available for serotype B strain WM276. These libraries were constructed for the whole genome sequencing project of the C. gattii strain WM276 and paired sequencing reads were obtained for approximately 60,000 clones. The assembled genome sequence including the paired end reads for each clone is available in the public domain (http://www.bcgsc.ca). By stand-alone local BLAST using the CAP10 gene from a serotype D strain B-4500 (Chang and Kwon-Chung 1999) as query sequence, four clones in the CN022 library (CN02275dA10, CN02262cG04, CN02225cD02 and CN02295bG12) and four clones in CN023 library (CN23E63cH10, CN23E31bE10, CN23E19aG11 and CN23E21dD08) were identified to contain the CAP10 gene. Clone CN23E63cH10 contained the full length CAP10 gene (2,586 bp insert) and was selected for further study.
An in vitro transposition reaction was performed by mixing the modified transposon vector pGPS3-Neo and the target DNA (pCN23E63cH10) in the presence of TnsABC transposase. Following E. coli transformation, the resulting colonies were screened by colony PCR; clones with putative insertions in the CAP10 gene produced a PCR band (<2.6 kb) in one of two PCR reactions, using either primers BR194-F and GPS-N, or primers BR194-F and GPS-S (data not shown). The presence of a PCR product also provided information on the orientation of the transposon insertion. Among 20 colonies screened, 12 colonies appeared to have a transposon insertion in the CAP10 gene while the remaining insertions occurred in the vector. The insert in pCN23E63cH10 is 2,586 bps and contained sequences from −557 to 2,029 nt of the CAP10 gene in WM276 (the open reading frame extends from 558 to 2,563 nt in the insert). To confirm the results from the PCR screen that indicated CAP10 disruption, plasmid DNA from four of the 12 colonies (pCAP10-Neo4, pCAP10-Neo9, pCAP10-Neo11, and pCAP10-Neo14) was used for sequence analysis of the insertion site. Transposon insertion sites 15 nt (pCAP10-Neo4), 175 nt (pCAP10-Neo9), 1,233 nt (pCAP10-Neo14), and 1,455 nt (pCAP10-Neo11) were present downstream from the putative start codon, respectively (Fig. 1b). pCAP10-Neo9 was chosen to transform three C. gattii strains.
C. neoformans and C. gattii are important pathogens of humans and experimentally tractable model systems to study fungal pathogenesis. The genomes of five representative strains from different serotypes have been sequenced, including JEC21 and B3501a (D), H99 (A), A1MR265 and WM276 (B). Such advances will enable extensive comparative genomics studies among C. neoformans, C. gattii and the other fungal species. In addition, gene expression studies by SAGE, microarrays, and subtractive hybridization, as well as proteomic studies, are identifying many interesting genes for subsequent genetic investigation. Rapid and high-throughput protocols to test gene function are therefore essential for the efficient use of the completed genomes and the expression data. In C. neoformans and C. gattii, cloning and/or PCR strategies have been previously used to make the disruption/deletion constructs of targeted alleles. However, such strategies can be time-consuming, expensive and technically challenging. The in vitro transposition-based strategy described here is relatively rapid and straightforward in its implementation, and it is cost effective. It does not require multiple cloning and/or PCR steps to generate the constructs for targeted alleles, although cloned PCR products can serve as templates. In addition, the primers used in the colony PCR steps to identify disruption constructs (either GPS-S and BR194-F, or GPS-N and BR194-F) were designed from the transposon and vector sequences, and therefore can be universally used for screening insertions in all targeted genes presented in the WM276 shotgun sequencing libraries. The transposon strategy also has the advantage of potentially identifying clones with different insertion sites within a gene to generate alleles with potentially novel phenotypic effects. Finally, the procedure is useful for introducing selectable markers into cloned DNA fragments to facilitate complementation of mutations during genetic reconstitution studies and into large insert DNA clones (e.g., BAC clones) for transformation into Cryptococcus cells. The latter approach may be useful for studying the functions of larger genomic elements such as centromeres, or for the complementation of large deletion mutations.
The Tn7-based transposition system is known to randomly generate single, simple insertions into target DNA (Biery et al. 2001), which makes in vitro transposition useful to initiate genome-wide mutagenesis studies in filamentous fungi (Hamer et al. 2001). The GPS3-based protocol has been used in other fungal species, such as Aspergillus fumigatus (Jadoun et al. 2004), the plant pathogenic fungi Mycosphaerella graminicola (Zwiers and DeWaard 2001) and Stagonospora nodorum (Solomon et al. 2004), and the human pathogen, Candida glabrata (Castano et al. 2003). Similarly, TAGKO (transposon-arrayed gene knockout), a technique developed based on the random insertion of transposons, was successfully used to disrupt targeted genes in Magnaporthe grisea (Hamer et al. 2001) and M. graminicola (Adachi et al. 2002). Unfortunately, such studies have been limited to ascomyceteous fungi. A previous investigation in M. graminicola suggested that an increase in gene-targeting frequency might be achieved if the transposon-based strategy is combined with the Agrobacterium-mediated transformation (Adachi et al. 2002). Agrobacterium-mediated transformation has been recently applied to C. neoformans (Idnurm et al. 2004) and it will be interesting to see if the gene-targeting frequency could be improved by combination of Agrobacterium-mediated transformation and disruption construct generation by the GPS3 transposon-based strategy.
The CAP10 disruption construct derived from the DNA of strain WM276 was successfully used to disrupt the gene in the different subtype strains A1MR265 and NIH444. The cap10 mutants from three serotype B strains (WM276, A1MR265 and NIH444) demonstrated an acapsular phenotype that was similar to that described for a serotype D strain (Chang and Kwon-Chung 1999). The efficiencies of homologous recombination in all serotype B strains are generally high (20–60%), although both NIH444 and A1MR265 were lower than WM276. This was likely due to sequence divergence among the subtype strains, which is known to affect gene targeting in C. neoformans. Global comparison of genome content revealed that the similarity between the two recent sequenced genomes, WM276 and A1MR265, is only around 95% and alignment of the CAP10 sequences showed that the similarity between WM276 and A1MR265 is 96%. There is limited genome information available for NIH444. The present study reveals that the VGI disruption construct can disrupt genes in the different subtype VGII strain, even though the similarity between the subtypes is only around 95–96%, but not in a different serotype (A or D) with a similarity of 92–96%. It is interesting to note that previous studies revealed that approximately 4–8% divergence of intragenic regions from serotype A to D may hinder gene targeting, thus preventing homologous recombination (Davidson et al. 2000). It was suggested that the same strain background should be used for the gene disruption construct and the recipient (Davidson et al. 2000). In general, the successful disruption of the CAP10 allele in A1MR265 by a disruption construct from a cross-subtype allele of WM276 indicates that it will be possible to use the easily accessible WM276 shotgun sequence libraries as a resource to knock out genes in other serotype B strains (including A1MR265). The finding that the disruption construct for the CAP10 gene of WM276 is applicable to generate knockout strains for the cross-subtype alleles in strains A1MR265 and NIH444 is intriguing given that gene disruption alleles for serotype A did not prove to be applicable to serotype D, and vice versa (e.g., Davidson et al. 2000). For the serotype B strains, it is not known whether the region where CAP10 is located is a “hot spot” for recombination, or whether the DNA recombination machinery (e.g., the DNA mismatch repair system) is different between the serotype A, B, and D strains. In Saccharomyces cerevisiae, E. coli, and Salmonella sp., it is known that the DNA mismatch repair system contributes to the requirement of identical sequences for high-efficiency homologous recombination (Earley and Crouse 1998; Rayssiguier et al. 1989).
Additional advantages of the transposon-based strategy are that it works in a variety of strain backgrounds, as indicated by URA5 disruption in the serotype A strain H99, and it can be modified for other fungi. Shotgun sequencing libraries, such as those constructed for the genome sequencing projects for the different strains of Cryptococcus, provide a valuable resource for the high-throughout gene-function tests. Future sequencing projects for fungi could similarly couple the in vitro transposition strategy with sequencing libraries to facilitate subsequent genetic analysis. For C.neoformans, homologous recombination requires a large region of homology, and the minimal flanking region of similarity was proposed to be larger than 300 bps (Nelson et al. 2002). Plasmid libraries included in shotgun sequencing programs normally contain inserts that are 2 kb and larger, which means that suitable disruption constructs can easily be selected after the in vitro transposition reaction. In addition, the sequences of all clones in shotgun sequencing libraries are easily accessible by bioinoformatics analysis. Thus, the high coverage of libraries should allow selection of any gene(s) of interest for disruption. Of course, access to the clone libraries is not essential for the procedure because cloned PCR fragments carrying genes of interest also serve as substrates for in vitro transposition.
Funding was provided by the National Institute of Allergy and Infectious Disease (5RO1AI053721) and the Canadian Institutes of Health Research (CIHR) to JWK. WM276 shotgun sequencing was completed at the Michael Smith Genome Science Centre (GSC) with funding from Genome Canada to JWK. We thank the sequencing group at the GSC for construction of the libraries.
- Adachi K, Nelson GH, Peoples KA, Frank SA, Montenegro-Chamorro MV, DeZwaan TM, Ramamurthy L, Shuster JR, Hamer L, Tanzer MM (2002) Efficient gene identification and targeted gene disruption in the wheat blotch fungus Mycosphaerella graminicola using TAGKO. Curr Genet 42:123–127PubMedCrossRefGoogle Scholar
- Casadevall A, Perfect JR (1998) Cryptococcus neoformans. ASM Press, WashingtonGoogle Scholar
- Hamer L, Adachi K, Montenegro-Chamorro MV, Tanzer MM, Mahanty SK, Lo C, Tarpey RW, Skalchunes AR, Heiniger RW, Frank SA, Darveaux BA, Lampe DJ, Slater TM, Ramamurthy L, DeZwaan TM, Nelson GH, Shuster JR, Woessner J, Hamer JE (2001) Gene discovery and gene function assignment in filamentous fungi. Proc Natl Acad Sci USA 98:5110–5115PubMedCrossRefGoogle Scholar
- Kidd SE, Hagen F, Tscharke RL, Huynh M, Bartlett KH, Fyfe M, Macdougall L, Boekhout T, Kwon-Chung KJ, Meyer W (2004) A rare genotype of Cryptococcus gattii caused the cryptococcosis outbreak on Vancouver Island (British Columbia, Canada). Proc Natl Acad Sci USA 101:17258–17263PubMedCrossRefGoogle Scholar
- Loftus BJ, Fung E, Roncaglia P, Rowley D, Amedeo P, Bruno D, Vamathevan J, Miranda M, Anderson IJ, Fraser JA, Allen JE, Bosdet IE, Brent MR, Chiu R, Doering TL, Donlin MJ, D’Souza CA, Fox DS, Grinberg V, Fu J, Fukushima M, Haas BJ, Huang JC, Janbon G, Jones SJ, Koo HL, Krzywinski MI, Kwon-Chung JK, Lengeler KB, Maiti R, Marra MA, Marra RE, Mathewson CA, Mitchell TG, Pertea M, Riggs FR, Salzberg SL, Schein JE, Shvartsbeyn A, Shin H, Shumway M, Specht CA, Suh BB, Tenney A, Utterback TR, Wickes BL, Wortman JR, Wye NH, Kronstad JW, Lodge JK, Heitman J, Davis RW, Fraser CM, Hyman RW (2005) The genome of the basidiomycetous yeast and human pathogen Cryptococcus neoformans Science 307:1321–1324PubMedCrossRefGoogle Scholar