Figure 1.
Long-term co-incubation of C. glabrata with RAW 264.7 macrophages yields a C. glabrata strain with pseudohyphae-like morphology.
A. Daily passage of C. glabrata with RAW 264.7 macrophages led to the formation of a pseudohyphae-like growth phenotype. The left panel shows the C. glabrata Evo strain interacting with RAW 264.7 cells. In contrast to the parental strain (WT; middle picture) the evolved strain (Evo) formed clumps (right picture) in liquid media. B. The parental wild type (WT) formed smooth colonies on YPD agar; colonies of the evolved strain (Evo) grew with a strongly wrinkled morphology. Both strains appeared purple colored on CHROMagar plates (right picture), characteristic for C. glabrata. C. Transmission electron micrographs indicate a cell separation defect of the evolved (Evo) strain in comparison to the WT, and show enlarged septa between mother and daughter cells, as well as an increased thickness of the outer cell wall layer.
Figure 2.
Increased fitness of the evolved strain in macrophages in direct competition.
Macrophages were infected with WT and Evo cells at a ratio of 100∶1 at day 0 and the relative proportion between the two strains monitored daily. The ratio reversed after a few days of coincubation, demonstrating an advantage for the Evo strain during interaction with macrophages. Mean values and standard deviations of three independent experiments are shown.
Figure 3.
Microevolutionary adaptation results in altered host-pathogen interactions.
A. The uptake by RAW 246.7 macrophages after 15 min to 6 hours of co-incubation was analyzed by differential staining (see Material and Methods). Both strains were internalized to a similar extent a all time points. (n≥3). B. Following 24 h co-incubation, the evolved strain (Evo) damaged RAW 246.7 macrophages, but not epithelial TR 146 cells, to a higher extent than the parental strain (WT) as measured by lactate dehydrogenase (LDH) release. (n≥3).
Figure 4.
Enhanced virulence and altered organ tropism of the evolved strain.
A. Mice were intravenously infected with 5×107 C. glabrata cells on day 0. Body weight of animals was monitored daily. During the first four days, infection with the evolved strain (Evo) led to severe loss in body weight, in contrast to mice infected with the parental strain (WT). B. Fungal burden was determined by culture from tissue homogenates of five animals per treatment group and time point. The distribution of fungal burden differed significantly between the wild type (WT) and the evolved strain (Evo) at the early time point. In the brain, the burden on day 2 p.i. was more than 100× higher for the evolved strain than for the wild type (median 1.1×107 cfu/ml vs. 4.9×104 cfu/ml). This difference was not present at later time points (days 7 and 21 p.i.), and both strains persisted at comparable levels in the organ. C. Representative histological images of brain tissue of infected mice. Brains infected with the evolved strain (Evo) showed large and numerous microcolonies (upper right picture, microcolonies are indicated by black arrows). Evo cells formed larger clumps of cells; the wild type (WT) formed only few small microcolonies in the brain (upper and lower left picture, a microcolony is indicated by a black arrow). Please note that the PAS stain used does not allow reliable differentiation between neuronal and endothelial cells.
Figure 5.
Differences in the host cytokine response to wild type and Evo strains in vivo and in vitro.
A. Selected cytokines and MPO levels were measured in murine organs at day 2 after infection with wild type or Evo strain. Significantly higher cytokine levels were found specifically in the brain of mice infected with the Evo strain, reflecting the transient high fungal burden in this organ. B. Release of the same cytokines by murine RAW246.7 macrophages was tested for the different strains. Only strains bearing the point mutation in the CHS2 gene (Evo and CHSEvo) elicited an increased release of TNFα.
Figure 6.
No large-scale genomic changes, but single SNP differences can be detected between WT and Evo strains.
Following alignment of Solexa/Illumina reads for the genomes of strains ATCC2001 and Evo on the C. glabrata reference genome, an average coverage score was calculated for each 1 kb region and normalized to the coverage obtained across the whole genome. These coverage ratio are shown in log2 scale. C. glabrata chromosomes A to M are shown in alternating black and grey colors. The location of SNPs identified in both strains relative to the reference genome is shown with green diamonds. The location of SNPs that distinguish the two strains is shown with red diamonds, the SNP on chromosome I responsible for the phenotype of strain Evo being shown in larger size. Note that several 1 kb regions harbored more than one SNP and are nevertheless represented using a single diamond.
Table 1.
Single Nucleotide Polymorphisms distinguishing the wild type (WT) and the evolved strain (Evo).
Figure 7.
A single nucleotide exchange is sufficient to produce the evolved phenotype.
A. Sanger sequencing confirmed the sequence alteration in CHS2 of the Evo strain (identified by whole genome sequencing), which led to an Asn→Lys (WT→Evo) amino acid exchange in the protein. Following a counter-selection experiment, the gene reverted to its original sequence (Rev), concomitant with the reversal to the original yeast growth form. B. This single nucleotide exchange observed in of the Evo strain was introduced into the WT strain by PCR amplification of CHS2 from the Evo strain, and cotransformation of this fragment with an PCR-amplified HIS3 marker including an overlapping (U1) region. C. The resulting strain was called CHSEvo, and correct integration was tested by sequencing. Similarly, CHSWT was created by amplifying the WT CHS2 gene and following the same cloning strategy. D. Morphologies of the WT, Evo, CHSWT and CHSEvo strain. The introduction of the Evo CHS2 gene into CHSEvo resulted in a growth form indistinguishable from the original Evo strain. The reintroduction of the WT gene did not change morphology (CHSWT).
Figure 8.
Introduction of a single nucleotide exchange into CHS2 results in increased macrophage damage.
Following 24 h co-incubation with macrophages, the CHS2Evo strain, containing the Evo allele of the CHS2 gene, elicited the same increased LDH release from macrophages as the Evo strain. Reintroduction of the wild type CHS2 gene (CHSWT) into the WT strain did not lead to a significant change in its damage potential. (n≥3).