Genomic survey of the non-cultivatable opportunistic human pathogen, Enterocytozoon bieneusi

Abstract

Enterocytozoon bieneusi is the most common microsporidian associated with human disease, particularly in the immunocompromised population. In the setting of HIV infection, it is associated with diarrhea and wasting syndrome. Like all microsporidia, E. bieneusi is an obligate, intracellular parasite, but unlike others, it is in direct contact with the host cell cytoplasm. Studies of E. bieneusi have been greatly limited due to the absence of genomic data and lack of a robust cultivation system. Here, we present the first large-scale genomic dataset for E. bieneusi. Approximately 3.86 Mb of unique sequence was generated by paired end Sanger sequencing, representing about 64% of the estimated 6 Mb genome. A total of 3,804 genes were identified in E. bieneusi, of which 1,702 encode proteins with assigned functions. Of these, 653 are homologs of Encephalitozoon cuniculi proteins. Only one E. bieneusi protein with assigned function had no E. cuniculi homolog. The shared proteins were, in general, evenly distributed among the functional categories, with the exception of a dearth of genes encoding proteins associated with pathways for fatty acid and core carbon metabolism. Short intergenic regions, high gene density, and shortened protein-coding sequences were observed in the E. bieneusi genome, all traits consistent with genomic compaction. Our findings suggest that E. bieneusi is a likely model for extreme genome reduction and host dependence.

Figure 1. Estimation of the genome size of E. bieneusi based on pulsed field electrophoresis analysis.

The E. bieneusi chromosomes were separated by electrophoresis (see Methods) and stained with ethidium bromide. The sizes of the chromosomal bands (lane 2; 0.92, 1.0, 1.06 Mb) were estimated using the MCID 3.0 software (Imaging Research Inc., St. Catharines, Canada). Based on densitometry analysis (ImageQuant TL 1-D analysis software; GE Healthcare Bio-Sciences Corp., Piscataway, NJ), the ratios of the bands were estimated to be 1∶4∶1. Using these ratios, the genome was estimated to be ∼6 Mb. S. cerevisiae chromosomal size standards (Bio-Rad) were included (lane 1) and their sizes (Mb) are shown.

Figure 2. Comparison of the lengths of E. bieneusi proteins with their respective S. cerevisiae (Sc) and E. cuniculi (Ec) homologs.

Only E. bieneusi proteins with assigned functions that were located on the larger contigs were included in these comparisons. (A). Lengths of the E. bieneusi proteins (n = 566) relative to their S. cerevisiae homologs, expressed as a percentage. (B). Lengths of the E. bieneusi proteins (n = 580) relative to their E. cuniculi homologs, expressed as a percentage. E. bieneusi proteins that were shorter or larger than their respective homologs have percentages less than 100% or greater than 100%, respectively.

Figure 3. Grouped bar graph showing the distribution percentages of E. bieneusi (blue) and E. cuniculi (red) proteins among the functional categories.

E. bieneusi proteins (653 proteins) were assigned to one of eleven functional categories listed in Katinka et al. . For comparison, the distribution percentages of the E. cuniculi proteins (884 proteins) were included . The corresponding gene lists for both organisms are presented in Table S4.