Fig 1.
CBP1 alleles show a range of lytic abilities.
BMDMs were treated with 2.5 μg/mL tunicamycin (Tm), infected with indicated Hc strains at an MOI of 5, or mock infected (uninf). Macrophage lysis was measured by lactate dehydrogenase (LDH) release into culture supernatants and is presented as the percentage of total LDH in the supernatant and lysate of uninfected macrophages at 3 hpi. The lysis at each time point is an average of duplicate measurements of wells infected in triplicate, resulting in six total measurements, ± standard deviation. Macrophages treated with Tm or infected with wildtype Hc, cbp1+CBP1, cbp1+Pb_CBP1, or cbp1+cbp1D10A showed significantly more lysis compared to uninfected BMDMs at the indicated time points (**p<0.01, two-tailed Mann-Whitney test).
Fig 2.
Hc does not cause ER stress in infected BMDMs.
BMDMs were infected with the indicated Hc strains at an MOI of 5, mock infected (uninf), or treated with 2.5 μg/mL tunicamycin (Tm). (A) Unspliced (Xbp1u) and spliced (Xbp1s) isoforms of Xbp1 were detected by non-quantitative RT-PCR. (B-D) Relative abundances of (B) ERdj4, (C) SEL1L, and (D) BiP were measured by RT-qPCR. (E) Phosphorylated PERK (Thr 980) was detected by Western blot at 12 hpi. (F) BMDMs were treated with the PERK-specific inhibitor GSK2606414 (3 μM) or vehicle (DMSO) for 12 hpi, and caspase-3/7 activity was assessed and normalized to vehicle-treated uninfected cells. Each value is an average of triplicate wells ± standard deviation. **p<0.01, two-tailed t test.
Fig 3.
Lytic CBP1 alleles activate the integrated stress response in infected macrophages.
BMDMs were treated with 2.5 μg/mL tunicamycin (Tm), infected with indicated Hc strains at an MOI of 5, or mock infected (uninf). Phosphorylated eIF2α (Ser51) (A) and ATF4 (B) were assessed by Western blot at 12 hpi. Representative blots are shown. The average signal intensity of phospho-eIF2α or ATF4 relative to the corresponding α-tubulin loading control from three replicates is shown in the bar graphs.*p<0.05, **p<0.01, compared to wildtype, two-tailed t-test. (C) Relative abundances of CHOP and TRIB3 transcripts were assessed by RT-qPCR at the indicated time points and normalized to uninfected BMDMs at 3 hpi.
Fig 4.
Lytic CBP1 alleles cause a decrease in phospho-Akt levels in infected macrophages.
BMDMs were infected with the indicated Hc strains, and phospho-Akt (Thr308) and total Akt were assessed by Western blot at (A) 12 hpi and (B) 48 hpi. Representative blots are shown. The average signal intensity of phospho-Akt relative to total Akt from three replicates is shown in the bar graphs.*p<0.05, **p<0.01, compared to wildtype, two-tailed t-test.
Fig 5.
Lytic CBP1 alleles cause an increase in caspase-8 activity in infected macrophages.
BMDMs were treated with 2.5μg/mL tunicamycin (Tm), infected with indicated Hc strains at an MOI of 5, or mock infected (uninf). Caspase-8 activity was measured at 24 and 48 hpi and normalized to the activity in uninfected cells at 24 hpi. Each value is an average of triplicate wells ± standard deviation. **p<0.005, compared to wildtype, two-tailed t-test.
Fig 6.
The host genes CHOP and TRIB3 are necessary for robust caspase-3/7 activity and macrophage death during Hc infection.
(A) Wildtype, CHOP-/-, and TRIB3-/- macrophages were mock infected or infected with wildtype Hc at an MOI of 5 or 10. Caspase-3/7 activity was measured 24 hpi and normalized to the activity in uninfected macrophages for each genotype. Each value is an average of triplicate wells ± standard deviation. *p<0.05, compared to wildtype, two-tailed t-test. (B) Wildtype, CHOP-/-, and TRIB3-/- BMDMs (MΦ) were infected with wildtype Hc at an MOI of 1 or mock infected (uninf), and macrophage lysis was measured by LDH release. BMDM lysis is presented as the percentage of total LDH in the supernatant and lysate of uninfected macrophages at 3 hpi. The lysis at each time point is an average of duplicate measurements of wells infected in triplicate, resulting in six total measurements, ± standard deviation.
Fig 7.
CHOP is required for optimal spread of Hc during macrophage infection.
(A) Uninfected wildtype BMDMs were seeded in 6-well plates. Above them, transwells with wildtype or CHOP-/- BMDMs, either uninfected or infected with wildtype Hc, were placed with the macrophages on the underside of the transwell. After the onset of lysis of the infected macrophages, the transwells were removed. The fungal burdens immediately after transwell removal (B) and lysis kinetics (C) of the bottom macrophages were assessed by CFU enumeration and LDH release assay, respectively. For CFUs, ***p<0.005 by two-tailed t-test; for macrophage lysis, *p<0.05 by two-tailed Mann-Whitney.
Fig 8.
CHOP-/- mice are resistant to Hc infection.
(A) C57BL/6 and CHOP-/- mice were mock infected (uninf) or infected with 1x106 Hc yeast and the percentage of apoptotic (viability dye+ and caspase-3/7+) alveolar macrophages (CD45+ SiglecF+ CD11bmid CD11c+ F4/80+) was measured at 3 dpi. *p<0.05, as determined by ANOVA. (B) CFUs from lungs and spleens of C57BL/6 and CHOP-/- mice infected with 3 x 105 Hc yeast. *p<0.05, **p<0.005, as determined by independent 1-way ANOVA analyses of log transformed CFUs at each time point. (C) Kaplan-Meier survival curves of C57BL/6 and CHOP-/- mice mock infected (uninf; n = 2) or infected with 1 x 106 Hc yeast (n = 11). *p<0.05, logrank test.
Fig 9.
Model of Cbp1-mediated host-cell death during Hc infection.
During infection, Hc yeast produce a large amount of the secreted protein Cbp1 (red triangle). Cbp1 induces phosphorylation of the mammalian protein eIF2α through an unknown mechanism. The increase in phospho-eIF2 leads to the preferential translation of the transcription factor ATF4, which leads to the expression of CHOP and TRIB3. The pseudokinase Tribbles 3 inhibits Akt phosphorylation, thereby promoting Bax/Bak oligomerization, ultimately resulting in caspase-3/7 activation, which is enhanced by caspase-8 activation. The activation of the executioner caspases ultimately results in macrophage death, allowing for the release of live fungal cells.