Figure 1.
Selection and identification of host factors conferring protection against YFV-induced cell death.
A schematic of the iterative selection process is shown.
Figure 2.
DNAJC14 confers resistance to YFV-induced cell death.
(A) Photographs 7 d after YFV challenge (moi = 1) of SW13 cells transduced with Round 3 of the selected lentiviral cDNA constructs compared to cells transduced with V1-GFP vector control. (B) The cells transduced with the Round 3 lentivirus pool and surviving YFV infection (Rd 3) were expanded and reinfected with YFV at the indicated moi. Crystal violet staining was performed 3 d later. Cells transduced with vector alone serve as a control (V1-GFP). (C) DNA was isolated from naïve SW13 or Round 3 (Rd 3) cells, and the lentiviral insert amplified by PCR. The major band was identified as encoding a truncated hamster DNAJC14. Sizes of the DNA markers (kb) are indicated to the left. (D) A schematic of human DNAJC14 is shown, with the putative transmembrane (TM) domains (gray), J domain (red) with conserved HPD sequence, zinc finger motifs (blue) and Jiv90 domain (orange) indicated. A schematic of the isolated hamster clone, showing homology to amino acids 305 to 702 of human DNAJC14, is shown below.
Figure 3.
DNAJC14 inhibits YFV replication.
(A). YFV replication is inhibited in Round 3 (Rd 3) cells. Rd 3 or naïve cells were infected with YFV (moi = 1) and at the indicated days after infection the amount of virus released into the medium was determined by plaque assay. A single well was used for each cell type, with replacement of the medium at each timepoint. (B). Hamster DNAJC14 inhibits YFV protein expression. SW13 cells were transduced with control lentivirus (V1-GFP) or lentivirus containing the truncated hamster DNAJC14 insert (clone 1-2, 1-3, 1-4 and 1-5) and were infected 2 d later with YFV (moi = 0.5). Cells were analyzed by Western blot for NS3 expression 3 d after infection. Naïve cells serve as a negative control and actin serves as a loading control. (C). Full-length and truncated human DNAJC14 inhibit YFV replication. SW13 cells were mock transduced (Mock) or transduced with the indicated lentiviral vectors and 2 d later were infected with YFV (moi = 0.5) as indicated. Cells were harvested and subjected to Western blot analysis 3 d after infection using antibodies as indicated to the right. V1-FL expresses myc epitope tagged full-length human DNAJC14 while V1-NT1 expresses myc tagged truncated human DNAJC14 (aa 305–702). Naïve cells were left untransduced and uninfected and serve as a negative control. For B and C migration of size standards (in kDa) is indicated.
Figure 4.
DNAJC14 inhibits multiple members of the Flaviviridae family.
SW13 (A, B) or Huh7.5 (C–E) cells were seeded in equal numbers into plates and transduced with V1 vector expressing GFP (GFP, filled circles), V1 vector expressing full-length human DNAJC14 (FL, open circles) or V1 vector expressing truncated (aa 305–702) human DNAJC14 (NT1, triangles). After 2 days, the cells were challenged (moi = 5) with YFV 17D (A), YFV Asibi (B), Kunjin virus (C), Langat virus (D) or were challenged (moi = 0.1) with HCV Jc1FLAG2/p7-nsGluc2A. The medium (A–D) or cells (E) were harvested at the indicated times for quantification of virus replication. For A and B, a single separate well was utilized for each time point, and virion production was enumerated by plaque assay. In both cases, there were fewer than 100 plaque-forming units at 12 h post infection. The dashed line indicates the sensitivity of the plaque assay. Pfu, plaque forming units. For C and D, duplicate wells were infected and the medium was harvested and replaced at each timepoint. Virion production since the prior time point was enumerated by focus forming assay as described in Materials and Methods. Data points represent the mean titer; error bars indicate the range. Similar results were obtained in an independent experiment for both Kunjin and Langat viruses. FFU, focus forming units. For E, cells were harvested at the indicated times after infection for measurement of luciferase activity as described in Materials and Methods. Data points represent mean values obtained from triplicate wells; error bars indicate the standard deviation. RLU, relative light units.
Figure 5.
DNAJC14 inhibits a post entry step.
(A) and (B) V1-GFP- (GFP) or V1-hDNAJC14-FL (FL) transduced SW13 cells were electroporated 2 d later with in vitro transcribed YF-17D RNA to bypass the entry step. Cells and media were harvested at the indicated times after electroporation. (A) Western blot analysis was performed on equal volumes of the cell extracts using the antibodies indicated to the right; actin serves as a loading control and antibodies to myc detect the tagged DNAJC14 protein. Migration of size standards (in kDa) is indicated to the left. (B) Virus present in the medium from V1-GFP- (GFP, black circles) and V1-hDNAJC14-FL (FL, red triangles) transduced cells was enumerated by plaque assay. A single separate well was utilized for each time point. Pfu, plaque forming units. (C) Schematic of the YFV replicon construct, which expresses Renilla luciferase (RLuc) in place of the structural proteins. The locations of the YFV nonstructural proteins are indicated in the polyprotein, which is targeted to the ER by a signal sequence (red bar). UTR, untranslated region; 2A (black bar), the foot and mouth disease virus 2A autoproteolytic peptide. (D) V1-GFP- (GFP, circles) and V1-hDNAJC14-FL- (FL, triangles) transduced SW13 cells were electroporated with the wild type YF replicon RNA (filled symbols, solid lines) or with replication incompetent RNA containing a mutation in the RNA-dependent RNA polymerase (ΔDD, open symbols, dashed lines). At various times after electroporation, the cells were harvested and luciferase activity was determined. Data represents the mean luciferase value of triplicate samples; error bars indicate the standard deviation and are sometimes obscured by the symbol. RLU, relative light units. Similar results were obtained in an independent experiment utilizing Huh7.5 cells.
Figure 6.
(A) A schematic of DNAJC14 is shown with the truncation mutants indicated below. Arrows above the schematic indicate the location of point mutants engineered in the NT5 truncation mutant backbone. Each mutant contained a C-terminal myc epitope tag, indicated in green. (B and C) SW13 cells were transduced with lentivirus expressing the indicated mutants and 2 d later were challenged with YFV (moi = 5). Cells and media were harvested 1 d later. V, control V1 lentivirus; FL, lentivirus expressing full-length hDNAJC14; WT, the NT5 truncation mutant without any point mutations. (B) Western blot analysis was performed on equal volumes of the cell extracts using anti-myc antibody to detect the tagged DNAJC14 or mutant protein. Migration of size standards (in kDa) is indicated to the left. (C) YFV present in the medium was enumerated by plaque assay. Data represent mean values obtained from triplicate wells; error bars indicate the standard deviation. Pfu, plaque forming units.
Figure 7.
The C-terminus of DNAJC14 mediates self-interaction.
(A) Self-interaction of DNAJC14. HEK293T cells were cotransfected with pTrip-EGFP-hDNAJC14-NT5 (GFP-NT5) and pV1-hDNAJC14-FL (FL-myc) or the NT5 mutant (NT5-myc) as indicated. Cells were harvested 2 d later and myc-tagged DNAJC14 was immunoprecipitated using anti-myc antibody. Western blots were performed using anti-GFP and anti-myc antibodies as indicated. (B). Self-interaction is mediated by the C-terminus. HEK293T cells were cotransfected as indicated with pTrip-EGFP-hDNAJC14-NT5 (GFP-NT5) and pV1-hDNAJC14-NT5 (NT5-myc, left panels) or the NT5CT1 mutant (NT5CT1-myc, right panels) lacking the C terminal 77 amino acids. Cells were harvested 2 d later and DNAJC14 was immunoprecipitated using anti-GFP or control IgG as indicated. Western blots were performed using anti-GFP and anti-myc antibodies as indicated. For both A and B, arrows indicate the DNAJC14 proteins; migration of size standards (in kDa) is indicated to the left. The asterisk indicates immunoglobulin heavy chain.
Figure 8.
DNAJC14 does not inhibit NS2/3 cleavage of YFV and HCV.
YFV replication is inhibited in cells inducibly expressing the NT5 mutant form of DNAJC14. (A) T-REx-293-NT5 cells were left uninduced or induced by treatment with doxycycline (Dox) for the indicated h and lysates were analyzed by Western blot using anti-myc and anti-actin (loading control) antibodies. Migration of size standards (in kDa) is indicated to the left. (B) T-REx-293-NT5 (NT5) or T-REx-293-LacZ (LacZ) cells were induced to express DNAJC14-NT5 or β-galactosidase, respectively, by 24 h treatment with doxycycline. The cells were then infected with YFV (moi = 5) and the medium was harvested and replaced at each timepoint. Virion production since the prior time point was enumerated by plaque assay. Pfu, plaque forming units. (C) YFV and HCV NS2/3 cleavage in T-REx-293-NT5 cells. T-REx-293-NT5 cells were left untreated or were induced to express DNAJC14-NT5 by 24 h treatment with doxycycline (Dox) as indicated. The cells were then cotransfected with pEGFP (loading control) and either pFlag-HCV-NS2/3(181) or pFlag-YFV-NS2/3(181). As controls, NS2-3 proteins containing active site mutations in the HCV NS2 (H143A) or YFV NS3 (S138A) proteases and incapable of cleavage activity were also expressed as indicated. Cells were harvested 1 day later and analyzed by Western blot using anti-Flag antibody (top panel) or anti-myc and anti-GFP antibodies (bottom panel). Arrows indicate the migration of the relevant proteins and migration of size standards (in kDa) is indicated on the left.
Figure 9.
DNAJC14 is recruited to YFV replication complexes.
(A) SW13 cells were left untransduced (top row) or were transduced (lower 2 rows) with the noninhibitory V1-hDNAJC14 mutants FL-H471Q or CT1 as indicated. Two d later the cells were mock treated (left panels) or were challenged with YFV (moi = 5, right 3 panels). After an additional 2 d, the cells were fixed and immunostained with rabbit anti-YFV NS3 polyclonal antibodies (NS3), and mouse anti-calnexin antibody (calnexin) or mouse anti-myc monoclonal antibody (myc) as indicated. AF488-conjugated anti-mouse IgG and AF594-conjugated anti-rabbit IgG antibodies were used as secondary antibodies. The cells were analyzed by confocal microscopy and representative images are shown. Calnexin or DNAJC14 mutants are shown in green, YFV NS3 is shown in red, and the merged images are shown on the right. (B) SW13 cells were left untransduced or were transduced with the V1-hDNAJC14-CT1 mutant (CT1-myc) as indicated and were infected 2 d later with YFV (moi = 1). After 2 d of infection, myc-tagged DNAJC14-CT1 was immunoprecipitated using anti-myc antibody. Western blots were performed using antibodies against NS3, calnexin and the myc epitope tag as indicated. (C) SW13 cells were left uninfected or were infected with YFV (moi = 1) as indicated. The cells were fixed 1 d later and analyzed by confocal microscopy for endogenous DNAJC14 (red) and double stranded RNA (dsRNA, green). The merged image is shown on the right. Arrows indicate several areas of colocalized DNAJC14 and dsRNA.
Figure 10.
Modulation of DNAJC14 levels by siRNA alters YFV replication.
SW13 cells transduced with V1-GFP (Vector) or V1-hDNAJC14-FL (DNAJC14) were treated with irrelevant control siRNA or siRNA targeting DNAJC14 as indicated and were infected with YFV (moi = 5). After 24 h the medium from triplicate samples was collected for virus titration, while 2 samples were pooled for RNA isolation and cells in the remaining sample were harvested for Western analysis. (A) DNAJC14 RNA levels were determined by quantitative RT-PCR. For each sample, DNAJC14 RNA levels were normalized to levels of GAPDH RNA and the ratio present in the vector control cells treated with the irrelevant control siRNA was set to 1. Bars represent mean relative levels obtained from triplicate RT reactions; error bars indicate the standard deviation. Asterisks indicate a significant difference from the cells transduced with control vector and receiving the control siRNA (students t test; *p<0.05, **p<0.01, ***p<0.001). (B) Virus present in the medium was titered by plaque assay. Bars represent mean titers from triplicate samples; error bars indicate the standard deviation. Pfu, plaque forming units. Asterisks indicate a significant difference in virus production compared to cells transduced with control vector and receiving the control siRNA (students t test; ***p<0.001). Virus titers obtained after silencing DNAJC14 in the vector control versus DNAJC14 transduced cells were not statistically different (ns). (C) Western blot analysis was performed on the silenced samples using the indicated antibodies. Whether the cells were transduced with V1-hDNAJC14-FL (DNAJC14-myc+) or control vector (−) is indicated. Migration of size markers (in kDa) is indicated to the left. The asterisk indicates a non-specific band.
Figure 11.
DNAJC14 inhibits YFV in a temporal and dose-dependent manner.
(A) SW13 cells were mock transduced (Mock) or transduced with Trip-RFP-hNZAP (NZAP), Trip-RFP-hDNAJC14-FL (FL) or Trip-RFP-hDNAJC14-NT1 (NT1) and infected 2 d later with YFV-Venus (moi = 5). Cells were fixed at 1.5 d post infection and analyzed by flow cytometry. The Venus (y-axis) and RFP (x-axis) fluorescence intensities of the cells are shown; gates to indicate expression of the transduced protein or productive infection were set on Mock transduced, uninfected cells (not shown). (B) SW13 cells were transduced with Trip-RFP-hDNAJC14-NT1 and infected 2 d later with YFV-Venus (moi = 5) and analyzed as in (A) at the indicated days after infection. (C) SW13 cells were left untransduced (closed circles) or were transduced with Trip-RFP-hDNAJC14-NT1 (RFP-NT1, open circles) and infected 2 d later with YFV-Venus (moi = 5). At the indicated times, the medium was removed and YFV present in the medium was quantified by plaque assay. Each data point represents the mean titer obtained from duplicate wells; error bars indicating the range are obscured by the symbols. The dotted line indicates the sensitivity of the plaque assay.
Figure 12.
Proposed model of DNAJC14 function.
(A) Incoming viral RNA is translated to produce the inactive form(s) of a viral protein(s) required for RNA replication complex formation (A and B, yellow and brown stars). When the stoichiometry of the substrate protein(s) and the host chaperone machinery, consisting of DNAJC14 (DNAJ, blue) and Hsp70 (green), is appropriate, proper folding (AB) allows the formation of replication complexes which generate new progeny viral RNA. This RNA is further translated to produce more substrate, which after undergoing the chaperone process results in the formation of additional replication complexes and amplification of the RNA replication process. Overexpression of DNAJC14 mutants (red) lacking the C-terminal self-interaction domain (CT mutants), or with mutations in the J domain (H471Q, L466P) has no effect on viral replication, since the mutants lack features necessary for stable interaction and the chaperone complex is not disrupted. (B) Overexpression of full-length DNAJC14 (FL) or N-terminal truncation mutants (NT mutants) results in their incorporation into the chaperone complexes due to the presence of the C-terminal interaction domain and an intact J domain. Disruption of the normal stoichiometry of the substrate/chaperone complex results in a failure to properly fold the viral protein and a failure to generate replication complexes. With time, however, continued translation of the incoming genome results in increased concentrations of the viral protein. The appropriate stoichiometry is restored allowing the chaperone process to proceed and viral replication complexes to be generated.