A genomic safe haven for mutant complementation in Cryptococcus neoformans

Abstract

Just as Koch's postulates formed the foundation of early infectious disease study, Stanley Falkow's molecular Koch's postulates define best practice in determining whether a specific gene contributes to virulence of a pathogen. Fundamentally, these molecular postulates state that if a gene is involved in virulence, its removal will compromise virulence. Likewise, its reintroduction should restore virulence to the mutant. These approaches are widely employed in Cryptococcus neoformans, where gene deletion via biolistic transformation is a well-established technique. However, the complementation of these mutants is less straightforward. Currently, one of three approaches will be taken: the gene is reintroduced at the original locus, the gene is reintroduced into a random site in the genome, or the mutant is not complemented at all. Depending on which approach is utilized, the mutant may be complemented but other genes are potentially disrupted in the process. To counter the drawbacks of the current approaches to complementation we have created a new tool to assist in this key step in the study of a gene's role in virulence. We have identified and characterized a small gene-free region in the C. neoformans genome dubbed the "safe haven", and constructed a plasmid vector that targets DNA constructs to this preselected site. The plasmid vector integrates with high frequency, effectively complementing a mutant strain without disrupting adjacent genes. qRT-PCR of the flanking genes on either side of the safe haven site following integration of the targeting vector revealed no changes in their expression, and no secondary phenotypes were observed in a range of phenotypic assays including an intranasal murine infection model. Combined, these data confirm that we have successfully created a much-needed molecular resource for the Cryptococcus community, enabling the reliable fulfillment of the molecular Koch's postulates.

Fig 1. Gene distribution and spacing in the Cryptococcal genome.

Distribution of intergenic distances in C. neoformans with the frequency of neighboring genes.

Fig 2. Targeted integration vector.

A. The vector created has several key features. These include an intact multicloning site and blue-white screening for cloning the genetic construct of interest, a dominant selectable marker (here NAT, but also constructed in NEO and HYG versions), bla, a bacterial selection marker, and 5’ and 3’ flanking regions that are homologous to the safe haven site in the genome. The polylinker sequence contains three rare cutting restriction enzyme recognition sites that are used to linearize the vector so that the 5’ and 3’ regions are subsequently flanking the construct. B. Representation of the selected safe haven site, and how the linearized vector inserts via homologous recombination.

Fig 3. Multiplex PCR for screening successful genomic integration events.

A. Depiction of the genome with the primers used in the multiplex PCR. B. Representative multiplex colony PCR results from transformants of H99 with the empty vector cut with BaeI. Transformants yielding two bands indicate the construct has integrated correctly, while incorrect transformants only have one band.

Fig 4. Non-homologous ends inhibit successful homologous integration.

Illustration of the polylinker sequence after being cut with each of three rare cutting restriction enzymes. The construct depiction indicates where the marker and gene as well as the two flanking regions would be located after linearization. The red sequence shows residual polylinker sequence after digestion. Percent integration indicates the proportion of antibiotic resistant colonies in which the targeting vector was correctly integrated at the safe haven site as determined by multiplex PCR after biolistic transformation.

Fig 5. Integration of the vector into the safe haven site does not affect expression of flanking genes.

Transcript abundance of CNAG_00777 and CNAG_00778 relative to ACT1 with (H99 + empty vector) and without (H99) integration of vector at the safe haven site. Values show mean, error bars show S.E.M.

Fig 6. An ade2Δ mutant cannot grow on YNB media lacking adenine.

All complemented strains created in this study have wild-type levels of growth on YNB media and are of normal color.

Fig 7. Expression of ADE2 and flanking genes in an ade2Δ strain complemented with targeted or random integration.

Expression of (A) CNAG_02295, (B) ADE2 and (C) CNAG_02293 relative to ACT1 in various strains. The black arrow indicates where the NAT selectable marker was inserted to complement in the strain “Genomic Location”. Values show mean, error bars show S.E.M. * = P<0.05; *** = P<0.01.

Fig 8. Virulence in a mouse model.

A. No significant difference was found between H99, H99 carrying the empty targeting plasmid, and the ade2Δ strain complemented with ADE2 using the safe haven targeting plasmid. While the ade2Δ mutant and the ade2Δ strain carrying the empty safe haven showed no significant difference from each other, a significant difference in survival was seen between each and H99 (P<0.0001). Significant differences were also observed between H99 and complemented strains Random #1, Random #2 and Genomic Location (P<0.0001, P<0.0001 & P<0.05 respectively).