|Home | About | Journals | Submit | Contact Us | Français|
For eukaryotic pathogens that are depauperate in genetic variation, multilocus microsatellite typing (MLMT) offers an accurate and reproducible method of characterizing genetic diversity; herein we describe the development of an MLMT system for the emerging pathogenic fungus Penicillium marneffei based on 23 microsatellite loci. Screening isolates held within the Centraalbureau voor Schimmelcultures culture collection demonstrate high levels of genetic diversity and 100% reproducibility. This MLMT system provides a powerful epidemiological tool to analyse the underlying parameters that are responsible for the emergence of P. marneffei in human HIV-positive populations.
Penicillium marneffei is an asexual pathogenic fungus of the family Trichocomaceae that has emerged since 1990 as a significant human mycosis. Emergence of P. marneffei has occurred in concert with the epidemic of AIDS in Southeast Asia where it is classified as an AIDS-related indicator disease (Li et al. 1992). The endemic range of the pathogen is confined to Southeast Asia where autochthonous isolations are known from northeast India, Thailand, the Guangxi region of China, Hong Kong, Taiwan, Viet Nam and Indonesia (Ajello et al. 1995). In these regions, P. marneffei is a naturally occurring sylvatic infection within a high proportion of bamboo rat species (Ajello et al. 1995). However, it is not known whether bamboo rats are (i) an obligate stage in P. marneffei’s life cycle and/or are (ii) a zoonotic foci for human infection. In order to address these questions, there is an urgent need for well-characterized neutral genetic markers with which to analyse the population structure of P. marneffei.
Although multilocus sequence typing (MLST) (Maiden et al. 1998) is becoming the established method of discriminating isolates for many pathogens, insufficient genetic variation in eukaryotic species can pose a problem ( Taylor & Fisher 2003). In such cases, loci with increased variability are required. Studies on the ascomycete fungus Coccidioides immitis and C. posadasii have shown that multilocus microsatellite typing (MLMT) systems work well in discriminating individuals, populations and species (Fisher et al. 2002). Here, we describe the development of an MLMT system for P. marneffei based on a sample of 24 isolates (Table 1).
The NCBI blastn tool was used to search the 1.7 mega-bases of P. marneffei sequences deposited in GenBank for loci that contained microsatellite repeats. All possible permutations of the dinucleotide motifs (AC)7, (AG)7, (AT)7 and (CG)7 were used to screen P. marneffei entries in GenBank. This process was repeated for tri- and tetra-nucleotide motifs, with a minimum specified repeat length of six (for trinucleotide repeats) and five (for tetranucleotide repeats). This search resulted in 30 dinucleotide, 14 tri-nucleotide and five tetranucleotide repeats being discovered. From these 49 sequences, a set of 24 that contained the longest microsatellite repeats were chosen and polymerase chain reaction (PCR) primers were designed (Table 2). To facilitate multiplex PCR amplification, four groups of six loci were designed to amplify with nonoverlapping sizes. The forward primer for each locus was fluorescently labelled with either FAM, HEX or NED in order that the loci could be genotyped using an automated sequencer.
A panel of 24 isolates were assembled comprising 17 clinical isolates from the collection held at the Centraalbureau voor Schimmelcultures (CBS) and the type specimen, isolated from a bamboo rat from Viet Nam in 1959 (Capponi et al. 1956). In addition, clinical isolates were chosen from the collection of Dr Nongnuch Vanittanakom (Table 1). The Chiang Mai isolates had been isolated from three patients on two independent hospital visits, resulting in two isolations for each patient. Assuming that mixed infections are rare for cases of disseminated penicilliosis, then these duplicated isolates should have an identical multilocus genotype, and therefore act as positive control for the reproducibility of the MLMT system. Isolates were grown on Sabourad’s agar at room temperature and DNA was extracted from 7-day-old cultures as described previously (Vanittanakom et al. 1996). All isolates were typed ‘blind’ to control for user bias.
Primers for each of the chosen microsatellite loci were aliquoted into four multiplex pools to a concentration of 2 μm. Subsequently, 1 μL of a 1/10 dilution of the DNA from each isolate was amplified using the Qiagen Multiplex PCR Kit (Qiagen), with a working primer concentration of 0.2 μm. Cycling conditions were as follows: 95 °C for 15 min, followed by 35 cycles of 94 °C for 30 s; 57 °C for 90 s; 72 °C for 60 s, and a final extension of 60 °C for 30 min. Each reaction was then diluted 1:6 with dH2O and 1 μL of the dilutant was heat denatured with 8·75 μL of formamide loading buffer at 95 °C for 2 min. Subsequently, the PCR products were electrophoresed through a 96-capillary array using POP-6 and a Rox-500 internal size standard (Applied Biosystems). Alleles were scored using genotyper software (Applied Biosystems) and unique genotypes were then assigned a specific microsatellite type (MT) identifier. Of the 24 loci, 23 amplified and typeability was high, with 550/552 alleles being successfully characterized (99.6%). Only a single allele was found within each P. marneffei strain, confirming that the fungus is haploid. Of the 23 typeable loci, 21 were polymorphic with between two and 14 alleles within our sample. When the ‘blinds’ had been removed from our panel of DNA, it was found that isolates taken from the same patient on different occasions (NV16/NV19, NV12/NV30, NV56/NV59) had identical multilocus genotypes, demonstrating 100% reproducibility of the MLMT system at all 23 loci. This shows the that the loci are not mutating at a high enough rate to generate intra-isolate polymorphisms over the timescales observed here. Within this dataset, the index of association (IA) was highly significant (IA = 3.414, P < 0.01), showing that there is extensive multilocus linkage disequilibria. This disequilibria probably results from the pathogens asexual mode of reproduction, although there may be a component that is attributable to phylogeographical structure.
In common with MLST, MLMT is an exact method by which different laboratories can compare their results to those accomplished by other laboratories. The increasing availability of automated sequencers means that MLMT genotypes can be rapidly generated and added to online databases of multilocus genotypes. To aid this process, an SQL server relational database of P. marneffei MLMT genotypes is held at http://pmarneffei.multilocus.net/. Adoption of our MLMT system for P. marneffei will allow an accrual of genetic information from a number of collaborating laboratories. Once sufficient coverage is generated, then sophisticated spatio-temporal epidemiological surveillance is possible, as well as allowing questions on the evolution and adaptation of P. marneffei to be addressed.
This work was supported by a Wellcome Trust Biodiversity Fellowship to MC Fisher.