The Zika virus NS5 protein binds HSP90 to suppress EGF-induced Akt signaling and trophoblast cell migration

Abstract

Zika virus (ZIKV) infection during pregnancy can cause congenital Zika virus syndrome (CZV), including fetal growth restriction and death. In the developing placenta, trophoblast cells respond to epidermal growth factor (EGF) to migrate into the decidua to facilitate implantation and fetal development. EGF activates the Akt protein kinase, a master regulator of trophoblast cell migration. Akt signaling and stability are dependent on heat shock protein 90 (HSP90), which mediates the maturation of proteins necessary for EGF/Akt signaling. Here we show that ZIKV infection inhibits EGF-mediated Akt activation and downstream signaling to suppress trophoblast migration. The ZIKV non-structural protein 5 (NS5) is sufficient to inhibit trophoblast migration through its binding interaction with HSP90, leading to suppression of Akt phosphorylation and inhibition of EGF-induced trophoblast migration. Thus, ZIKV NS5/HSP90 interactions play a key role in disruption of trophoblast function, revealing an underlying cause of improper placental development and fetal disease.

1. Introduction

Zika virus (ZIKV) is an emerging flavivirus that was first isolated in 1947 in Zika Forest, Uganda (1), and has since emerged in Oceania and the Americas (2,3). ZIKV is related to several important emerging viral pathogens, including West Nile virus, Dengue virus, and Yellow Fever virus (4). Flaviviruses are enveloped single-stranded RNA viruses. The RNA genome encodes three structural proteins (capsid, pre-membrane, and envelope) and seven non-structural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, NS5) (5,6). In the sub-tropical and tropical regions of the world, ZIKV is primarily transmitted by the Aedes aegypti mosquito (7,8). Importantly, ZIKV can also be spread through direct sexual contact, blood transfusion, and vertically during pregnancy (2,9). While ZIKV infection typically causes mild symptoms in adults, infection in pregnant women can lead to congenital Zika virus syndrome (CZS), which includes severe outcomes such as microcephaly, fetal growth restriction, and fetal demise (10–13). In 2015, ZIKV caused the largest epidemic in Brazil infecting between 440,000–1,300,000 people, resulting in a surge of neonates born with microcephaly, a birth defect associated with altered myelination in the central nervous system (CNS) (14–16). In 2016, the World Health Organization (WHO) declared ZIKV a Public Health Emergency of Concern (3,17). ZIKV is now endemic in many countries in the Americas (18,19) and remains a public health problem today, with Brazil alone reporting 1,065 cases of CZS in 2023 (20).

With ZIKV causing a broad range of adverse effects on the developing fetus, it remains paramount to define mechanisms leading to CZS. ZIKV productively infects the placenta where it targets placental macrophages, trophoblast cells, fibroblasts, and the maternal decidua basalis (14,21,22). Women infected during the first two trimesters of pregnancy are at the highest risk of congenital ZIKV infection leading to CZS and fetal demise (23–25). ZIKV infection in pregnant mothers and non-human primates has been associated with impaired placental development, including immature placental villi and a decreased number of syncytial knots, which contribute to placental insufficiency (24,26–28). Problematically, dysregulation in placentation can lead to poor pregnancy outcomes, such as miscarriages and intrauterine growth restriction (29,30). Trophoblast cells are essential for successful implantation and proper placentation, and any disruption in placental development has significant outcomes such as preeclampsia, fetal growth restriction, miscarriage, and stillbirth (30–33). Even minor periods of malnutrition could permanently impact fetal organ development (34,35). Trophoblast migration in early pregnancy is vital for the initial blastocyst implantation and primary villi formation. As pregnancy progresses, further migration of trophoblast cells facilitates the development of anchoring villi, which are crucial for attaching the placenta to the uterine wall. These anchoring villi give rise to extra villous trophoblast cells, which play a key role in remodeling maternal spiral arteries to ensure an adequate blood supply to the fetus (30,31,36).

Trophoblast migration is precisely controlled by maternal and fetal growth factors and cytokines, including insulin-like growth factors 1 and 2, placental growth factor, fibroblast growth factor, and epidermal growth factor (EGF) (37,38). EGF is essential for trophoblast migration and is produced throughout pregnancy (37,39). Dysregulation in EGF signaling during pregnancy leads to fetal growth restriction and pregnancy complications (37,38,40). EGF directs the activation of the Ak strain transforming (Akt) kinase pathway. In addition to migration, Akt plays key roles in multiple cellular processes, such as proliferation and protein synthesis (41–43). Activation of Akt controls cell migration in part by reorganizing the cytoskeleton via glycogen synthase kinase-3 beta (GSK3β), which affects microtubule architecture and cell polarity (42,44). Akt function is heavily reliant on Heat Shock Protein 90 (HSP90), a molecular chaperone that facilitates protein kinase signaling (45,46). HSP90 tethers Akt signaling proteins together and precisely controls the activation and dynamic behavior of the Akt signaling pathway (47). HSP90 dysregulation in the placenta has been associated with intrauterine fetal growth restriction in humans, and mutations in HSP90 in mice lead to failure of placental labyrinth formation (48,49).

Previous studies in human neural stem cells have shown that ZIKV dysregulates Akt signaling to suppress neurogenesis, but the mechanisms by which ZIKV impacts Akt signaling are not defined (50). Moreover, the impact of ZIKV on Akt signaling in the placenta is not well understood. Flavivirus non-structural protein 5 (NS5), including ZIKV NS5, can bind to HSP90 and dysregulate its interaction with client kinases such as Janus (JAK) kinases and Akt (51,52). NS5 is an essential protein for viral replication, serving as an RNA-dependent RNA polymerase (RdRp), which is essential for RNA synthesis (53). Additionally, NS5 functions as a methyltransferase (MTase) that adds a 5’ RNA cap to facilitate polyprotein translation and help evade innate immunity (53). ZIKV NS5 binds to HSP90 to inhibit its chaperone function, leading to misfolding and proteasomal degradation of HSP90-client kinases (51). The NS5-HSP90 interaction facilitates viral evasion of types I and III interferon (IFN) signaling via degradation of JAKs, thereby supporting virus replication and spread (51,54,55). However, the interaction between ZIKV and HSP90 in the placenta and its impact on EGF-mediated Akt signaling is unknown. In this study, we examined the impact of ZIKV infection of trophoblast cells on EGF-mediated Akt signaling and subsequent EGF-induced cell migration. We demonstrated that ZIKV dysregulates EGF-dependent phosphorylation of Akt in trophoblast cells, which subsequently disrupts trophoblast migration. We further show that ZIKV NS5 binds to HSP90 in trophoblast cells, resulting in disruption of Akt phosphorylation and suppression of EGF-dependent trophoblast cell migration. We conclude that disruption of the EGF-Akt signaling axis in trophoblast cells is an important determinant of ZIKV-induced pregnancy complications underlying CZS.

2. Materials and methods

2.1. Cell lines

JAR (ATCC^® HTB-144^™) were cultured in RPMI-1640 (Fisher Scientific) supplemented with 10% fetal bovine serum (FBS; HyClone, Logan, UT, USA), 10mM HEPES (Fisher Scientific) and 1X Antibiotics/Antimycotic (Fisher Scientific). HEK293T (ATCC) and Vero (WHO, Geneva, Switzerland) cells were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS; HyClone, Logan, UT, USA), 10mM HEPES (Fisher Scientific) and 1X Antibiotics/Antimycotic (Fisher Scientific). All cell lines were negative for mycoplasma contamination. The cells used in our experiments were at passage number 3. To ensure consistency across all experiments, cell numbers were counted, and treatments and infections were applied when the cells reached approximately 80% confluency. All cells were maintained in humidified incubators set at 37°C with 5% CO₂.

For acute EGF treatment, cells were treated with 100ng/ml of EGF (Thermofisher) for 30 minutes.

2.2. ZIKV strains and working virus stocks preparation

ZIKV/French Polynesia/H/PF/2013 was provided by Ralph Baric (University of North Carolina), ZIKV/Honduras/R103451/VR-1848 was provided by Lucia Vojtech (University of Washington), and ZIKV Malaysia (ZIKV/Malaysia/1966/P6740) was provided by Robert Tesh (University of Texas Medical Branch, Galveston, TX, USA).

All infections for generation of working virus stocks were initiated using Vero cells seeded in a hyperflask and infected with ZIKV at an MOI of 0.02 in a final volume of 200mL (DMEM with 1% FBS). Infected cells were incubated in a humidified incubator set at 37 °C with 5% CO₂ for 2 h with gentle rocking. 550 mL of fresh media (2% DMEM) was subsequently added to the inoculum. 72 hours (h) post-virus adsorption, the media was removed, saved and replaced with fresh complete media and the infection let run for another 72h. The two supernatants were combined and concentrated with Amicon Ultra centrifugal filter unit (Millipore Sigma UFC901024) according to manufacturer’s protocol. The titers of working stocks were determined using a plaque assay performed on Vero cells. Six-well plates were seeded with Vero cells to reach approximately 80% confluency on the day of infection. Each well was inoculated with 300 μL of diluted ZIKV and incubated in a humidified chamber at 37 °C with 5% CO₂ for 1 hour with gentle rocking. After incubation wells were rinsed with 1xDPBS and 2 mL of an agarose overlay was added to each well. Plaques became visible after 4–5 days and were stained with 2 mL of a Neutral Red (Sigma N2889) agarose overlay for visualization and counting.

2.3. ZIKV infection

Cells were mock-infected or infected with ZIKV variants at an of MOI 10, prepared in complete cell culture media and incubated at 37°C for 2 h rocking. Following adsorption, the inoculum was removed and replaced with complete media at the appropriate timepoints (described in figure legends). All cells were maintained in humidified incubators set at 37°C with 5% CO₂.

2.4. Protein lysate quantification

Cells were lysed on ice with radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate) containing freshly added protease and phosphatase inhibitors cocktail (1X; Thermo Scientific no. 78440). Cell lysates were collected in 200μl of RIPA and stored at −80°C. Lysates were then centrifuged at 14,000 rpm for 15 min at 4°C to pellet the cellular debris. Clarified lysates were collected and quantified using a Bio-Rad protein assay kit (Bio-Rad).

2.5. Immunoblot analysis

Protein lysates were incubated with Laemmli buffer, adjusted to a final concentration of 1X (Bio-Rad, no. 1610747, 1610737) containing 10% β-mercaptoethanol (Thermo Fisher-Life Technologies, no. AC125472500) at 95°C for 5 min. Approximately 10 mg of total protein was loaded per lane onto 4 to 20% Criterion TGX gradient gels (Bio-Rad, no. 5671094) and electrophoresed at ~105V in 1X sodium dodecyl sulfate (SDS) running buffer (25 mM Tris, 192 mM glycine, 0.1% SDS). Proteins were then transferred onto nitrocellulose membranes (Fisher Scientific) at 90 V for 1 h at 4°C in 1X Towbin transfer buffer (25 mM Tris, 192 mM glycine, 0.01% SDS, 20% methanol). Membranes were blocked for 1 h at room temperature in Tris-buffered saline (TBS)-based Intercept blocking buffer (LI-COR) and stained overnight at 4°C with the primary antibodies diluted in blocking buffer: Phospho-Akt (Ser473) (Cat# 4060S; 1:1000; Cell Signaling Technology [CST]), Akt (Cat # 9272;1:1000; CST), Phospho-GSK3β (Ser9) (Cat# 5558S;1:1000; CST), GSK3β (Cat# 12456; 1:1000; CST), ZIKV NS5 (Cat # GTX133327; 1:2,500; GeneTex, Irvine, CA, USA), Actin (Cat# MAB1501; 1:1000; MilliporeSigma), HSP90 (Cat# ab240366,1:10000; Abcam). The following day, membranes were washed with TBS containing Tween 20 (TBST) and probed with secondary antibodies (either horseradish peroxidase (HRP) or Alexa Fluor 680-conjugated or Alexa Fluor 790-conjugated secondary antibodies diluted 1:10,000 in TBS for 1h at RT. Membranes were then washed with TBST followed by 1X TBS and imaged on an Odyssey CLx imager (LI-COR) or ChemiDoc XRS+ system (Bio-Rad). HRP bands were developed using chemiluminescence (ECL) prime Western blotting reagents (Fisher Scientific). Protein abundance was quantified in Image Studio.

2.6. Migration Assay

The protocol was modified from the method described by Angelova et al. (2013) (60). JAR cells were serum starved for 24h in cell media containing 1% FBS media. The next day, they were lifted with TryplE express enzyme (Thermofisher) and plated at 1×10⁵ cells in duplicates in FluoroBlok transwells (Fisher Scientific, no. 351152) or Translucent wells (Fisher Scientific, 087712) in a 24-well companion plate (Fisher Scientific, no. 08–771-25). Cells were subsequently treated with mock/EGF at 1ng/ml in serum starvation media and incubated for 24 hours for migration to happen. Cells were then stained with a viability dye, Calcein-AM (Biolegend, no. 425201, 1:1000), in HBSS (Thermofisher, no. 14025092). Confocal immunofluorescence images of live cells from six fields of view at different focal planes along the z-axis were then acquired on a Nikon Eclipse Ti microscope. The images were analyzed via Image J software.

The migration index was determined and used to quantify the proportion of migrated cells relative to the total cell population (migrated and unmigrated). Fluorescent z-plane images from FluoroBlok wells (migrated cells) and Translucent wells (migrated and unmigrated cells) were z-stack projected. A threshold value for detection, determined from mock wells, was uniformly applied across all samples, converting the images to binary format. Cell density was quantified in pixels using the ImageJ’s Density Intensity function. The Migration Index was then calculated using the formula:

Migration Index = (Density Intensity of Migrated Cells) / (Density Intensity of Total Cells).

2.7. P-Akt inhibition

JAR cells were serum starved in Opti-MEM and treated with 1μM API-1 (Cat# 3897, Tocris Biosciences) for 24h and treated with +/−EGF (100ng/ml) for 30 min before subsequently lysed in RIPA buffer and analyzed by immunoblotting.

For the migration assay, JAR cells were serum starved in 1% FBS for 24h and seeded in transwells containing 0.5μM API-1 or DMSO control for 4h. Next, the transwells were transferred to new companion wells containing 0.5μM of API-1 or DMSO +/− EGF.

2.8. HSP90 inhibition

JAR cells were treated with geldanamycin (GA) at 0.5μM for 24h before treatment with EGF (100ng/ml) for 30 min and lysed in RIPA for immunoblotting analysis. Mock cells were treated with DMSO. For migration assays, JAR cells were treated with 0.5μM of GA for 4h or DMSO control before transferring to transwells that had 0.5μM GA or DMSO +/−EGF.

2.9. ReCLIP Analysis of Endogenous Proteins

The protocol was modified from the method described by Roby et.al (2020) (51). JAR cells were infected with ZIKV strains at an MOI of 10 for 24h or 48h before washing twice with PBS and crosslinked with 0.5mM dithiobis (succinimidyl propionate) (DSP) (Cat# 22586, ThermoFisher) in PBS. Crosslinked cells were quenched with TBS at RT for 10min and washed with 1X TBS on ice. Cells were subsequently lysed on ice in RIPA and sonicated twice with 20s pulses and 30s pauses in-between. Lysates were cleared via centrifugation at 14,000 rpm for 15 min at 4°C and quantified using BCA assay (ThermoFisher). For each IP, 600μg of lysate was incubated with 4μg of anti HSP90 (Cat# SC-13119, Santa Cruz Biotechnology) or IgG2A (Cat#: 401501, Biolegend) in 500uL final volume at 4C° overnight with rotation. The following day 50μl of Protein G Dynabeads (ThermoFisher) (washed once with RIPA prior to use) was added to lysates in 500μl final volume and incubated for 3h at 4°C with rotation. Samples were subsequently applied to a magnetic rack to precipitate beads and washed four times with 400μl of RIPA. Protein complexes were eluted from beads and DSP crosslinker was reduced by incubation with 65μl of loading buffer (three parts RIPA to one-part 4X Laemmli Sample Buffer (Cat# 1610747, Bio-Rad) + 10% β-mercaptoethanol) and incubated at 95°C for 5min. Samples were analyzed via immunoblotting as appropriate with 15μl of immunoprecipitate loaded per lane. Membranes were blocked and stained overnight at 4°C with primary antibodies in blocking buffer: NS5 (Cat# GTX133327, Genetex), HSP90 (Cat# ab203126, Abcam), NS3 (Cat# GTX634387, Genetex), Capsid (Cat# GTX133317, Genetex), GAPDH (Cat# 8884, Cell Signaling). The following day, membranes were washed and probed with secondary antibody: HRP conjugated anti-Rabbit (Cat# 111–035-003, Jackson ImmunoResearch Secondary Antibodies) and developed HRP bands using chemiluminescence (ECL) prime Western blotting reagents (Fisher Scientific) and imaged on a ChemiDoc XRS+ system (Bio-Rad).

2.10. DNA constructs

PcDNA3.1+(control) and pcDNA3.1 ZIKV/Malaysia NS5 constructs were generated via infusion cloning as previously described (54).

ZIKV/Malaysia NS5 sequences were amplified from ZIKV/Malaysia cDNA and were cloned into pSLIK doxycycline-inducible expression system using TaKaRa infusion cloning kit using the following primer sequence: Forward: TGATCACTAGCGTACATGGGAGGTGGAACGGG, Reverse: TCTTCCAATTCGTACTTACTTGTCATCGTCATCCTTGTAATCGATG. Doxycycline inducible plasmid vector (pSLIK tightTRE puro) was gifted by Oberst lab, University of Washington. Lentivirus particles were produced by co-transfecting with NS5 or mCherry pSLIK plasmids (2.5 μg), along with packaging plasmids (pMD2G (0.7 μg) and PsPAX2(1.8 μg)) using TransIT-LT1(Cat# MIR 2304, MirusBio) transfection reagent and Opti-MEM according to manufacturer’s protocol. Lentivirus supernatants were collected at 48h and 72h post-transfection and concentrated with TAKARA Lenti-X Concentrator (Cat# 631231) according to the manufacturer’s protocol. Transfection efficiency was measured by immunoblotting and flow cytometry analysis to quantify NS5 expression and mCherry florescent cells, respectfully.

2.11. Cell transduction and Doxycycline induction

JAR cells were transduced with NS5 or mCherry (control) lentivirus (at a 1:6 dilution) in the presence of Polybrene (Millipore Sigma) according to the manufacturer’s protocol. At 48h post-transduction, cells were selected with 4 μg /ml of puromycin and incubated until confluency was achieved.

Transduced cells were serum starved in starvation media (culture media with 1% FBS) for 24h before they were treated with 1 μg/ml of Doxycycline for 24h and either collected in RIPA for immunoblotting or transferred to transwells for the transwell migration assay.

2.12. Cell Transfection

HEK293T cells were seeded on poly-L-ornithine coated 6-well plates and 24h later washed twice with PBS and incubated in 2.5ml of Opti-MEM. Cells were subsequently transfected with 2.5μg of plasmid with TransIT-LT1 (Cat# MIR 2304, MirusBio) transfection reagent according to the manufacturer’s protocol. Cells were incubated for 48h before they were lysed in RIPA buffer for immunoblot analysis.

2.13. Statistical analysis

The results are reported as the mean ± SEM of 3 independent experiments (unless mentioned otherwise in figure legend). Statistical analysis was performed with a one-way analysis of variance (ANOVA) followed by Tukey’s multiple-comparison test using GraphPad Software (9.5.1). An alpha level (p-value) of <0.05 was considered significant.

2.14. Illustrations

Figures were generated using BioRender.

3. Results

3.1. EGF-mediated Akt signaling in trophoblast cells is dysregulated by ZIKV

To examine the effect of ZIKV infection on Akt signaling, we assessed acute responses to EGF treatment in JAR cells, a first trimester trophoblast cell line (56,57), following infection with two different epidemic ZIKV strains: ZIKV/Honduras/R103451/VR-1848 and ZIKV/French Polynesia/H/PF/2013 (58, 59) or mock-infection (control). Cells were cultured in media alone (control) or treated with EGF for 30 minutes followed by immunoblot analysis. Treatment of mock-infected cells with EGF resulted in the phosphorylation of Akt and its downstream signaling target protein, GSK3β (Figures 1A and B). In contrast, EGF treatment of ZIKV-infected cells did not induce phosphorylation of Akt and GSK3β (Figure 1B). Total Akt levels were also reduced following ZIKV infection with a more pronounced reduction observed with the ZIKV/Honduras strain. Densitometric analysis of the ratio of phosphorylated Akt (P-Akt) to total Akt and of phosphorylated GSK3β (P-GSK3β) to total GSK3β demonstrated a significant reduction in EGF-induced Akt and GSK3β phosphorylation linked to ZIKV infection similarly for both virus strains (Figures 1C, D, E and F). These observations indicate that ZIKV reduced Akt abundance in addition to impairing EGF-induced Akt phosphorylation and downstream GSK3β phosphorylation for both virus strains.

Figure 1. ZIKV dysregulates EGF dependent Akt signaling.

A) Model diagram showing that EGF treatment induces phosphorylation of Akt and GSK3β, activating downstream signaling pathways. B) JAR cells were infected with ZIKV Honduras (48h) or ZIKV French Polynesia (48h) at MOI of 10 and treated with 100ng/ml of EGF or mock for 30min. Cells were analyzed via immunoblotting. Densitometry of phospho-protein to total protein normalized to mock was calculated in C) and D) for ZIKV French Polynesia and E) and F) for ZIKV Honduras. Results are representative of three independent experiments except E and F (n=2). For all figures error bars represent the standard error of the mean (SEM). All statistical analyses were performed with one-way analysis of variance (ANOVA) followed by Tukey’s multiple-comparison test. (*, P<0.05; **, P<0.005; ***, P<0.0002)

3.2. In trophoblasts, EGF-dependent migration is inhibited by ZIKV

To measure EGF-mediated cell migration, we optimized an in vitro migration assay to precisely measure cell migration under different experimental conditions (60) (Figure 2A). To first determine whether Akt signaling function is critical for EGF-mediated trophoblast migration in our JAR model, we utilized the Akt inhibitor, API-1 (61) to block Akt phosphorylation. API-1 treatment of cells inhibited EGF-mediated phosphorylation of Akt (Figure 2B). We next measured cell migration via the transwell migration assay under Akt inhibition by API-1 (Figures 2C and 2D). EGF treatment induced migration in DMSO control cells however API-1 treated cultures exhibited reduced levels of cell migration.

Figure 2. Akt phosphorylation is required for trophoblast migration.

A) A model of transwell migration assay. B) Immunoblot of JAR cells treated with API-1 for 24h followed by EGF treatment for 30min.C) Transwell migration assay images in the presence or absence of API-1 and or EGF (1ng/ml). Representative stitched images from 4 fields of view are shown for each well. D) Migration index. Error bars represent the standard error of the mean (SEM). The results are representative of three independent experiments. Statistical analysis was performed with a one-way analysis of variance (ANOVA) followed by Tukey’s multiple-comparison test. (*, P<0.05)

To assess the impact of ZIKV infection on cell migration, cultures of ZIKV-infected cells were subjected to the transwell migration assay. JAR cells were infected with ZIKV/Honduras (Figures 3A and B) or ZIKV/French Polynesia (Figures 3C and D) and or mock-infected. EGF treatment induced the migration of mock-infected controls compared to cells treated with media only. In contrast, ZIKV-infected cells were unresponsive to EGF stimulation and exhibited lower levels of migration (Figure 3). To account for the cytopathic effect of ZIKV infection that could possibly reduce cell migration output, a migration index was defined (see Methods section 2.6). The migration index of both ZIKV/Honduras-infected and ZIKV/French Polynesia-infected cells indicated a similar significant reduction in EGF-induced cell migration (Figures 3B and D). Thus, ZIKV inhibition of Akt signaling imparts a blockade to EGF-mediated trophoblast cell migration.

Figure 3. ZIKV impairs EGF dependent migration in trophoblasts.

Migration assay of ZIKV infected cells. JAR cells were infected with ZIKV Honduras (A and B) or ZIKV French Polynesia (C and D) at an MOI of 10 for 24h. Cells were then transferred to FluoroBlok or translucent transwell inserts and incubated for 24 hours in the presence of 1 ng/ml EGF or a mock treatment in the outer well. Representative stitched images from 4 fields of view are shown for each well. One-way ANOVA test was used to compare means of migration indexes. Error bars represent the standard error of the mean (SEM). The results are representative of three independent experiments. Statistical analyses were performed with a one-way analysis of variance (ANOVA) followed by Tukey’s multiple-comparison test. (***, P<0.0002).

3.3. EGF-mediated signaling and migration require the HSP90 chaperone function

HSP90 is an essential chaperone of Akt that mediates protein maturation and stability to facilitate signaling competency (47, 48). Treatment of JAR cells with geldanamycin (GA), an HSP90 inhibitor that blocks chaperone function (62–64), significantly reduced EGF-induced Akt and GSK3β phosphorylation levels and led to reduced Akt abundance (Figure 4A). Densitometric analysis confirmed a significant decrease in the ratio of phosphorylated Akt to Akt in GA-treated cells, indicating that inhibition of HSP90 blocks Akt phosphorylation and reduces total Akt levels, likely due to protein misfolding and degradation (48, 53) (Figure 4B). To assess the impact of HSP90 inhibition on trophoblast function, we conducted transwell migration assays on GA-treated cells. GA treatment resulted in a significant decrease in trophoblast migration compared to mock-treated cells (Figures 4C and D). These findings demonstrate that HSP90 function is critical for proper EGF-mediated signaling and migration of trophoblast cells.

Figure 4. HSP90 is critical for EGF signaling and migration.

JAR cells were treated with geldanamycin (GA) to inhibit HSP90. A) Cells were treated with 0.5μM of GA for 24h and subsequently activated with 100ng/ml of EGF for 30 min and analyzed via immunoblotting. B) Densitometry of phospho-protein to total protein normalized to mock. C) Migration assay in the presence of GA. JAR cells were treated with 0.5uM GA for 24h then transferred to FluoroBlok or translucent transwell inserts and incubated for 24 hours in the presence of 1 ng/ml EGF or a mock treatment in the outer well. Representative stitched images from 4 fields of view are shown for each well. D) Migration Index. Error bars represent the standard error of the mean (SEM). The results are representative of three independent experiments. Statistical analysis was performed with one-way analysis of variance (ANOVA) followed by Tukey’s multiple-comparison test. (***, P<0.0002).

3.4. ZIKV NS5 interacts with HSP90 in trophoblast cells

Flavivirus proteins, including the NS5 protein, interact with HSP90 (53, 54). To determine if endogenous HSP90 in trophoblast cells interacts with ZIKV NS5, we conducted a reversible cross-linked immunoprecipitation of HSP90 (ReCLIP) and immunoblot assay. We evaluated HSP90/NS5 interaction during infection of JAR cells with three different ZIKV strains: ZIKV/Honduras, ZIKV/French Polynesia, and ZIKV/Malaysia (1966/P6740), the prototypical Asian lineage ZIKV strain (65). ReCLIP/immunoblot analysis revealed that NS5 from each ZIKV strain interacts with endogenous HSP90, as NS5 specifically co-precipitated with anti-HSP90 antibody but not with isotype control antibody (Figure 5A). NS5/HSP90 complex formation was apparent with all strains and there was no significant difference between relative abundance of NS5 recovered from co-immunoprecipitation with HSP90 between Zika virus strains (Figure 5B). We also identified the ZIKV NS3 protein but not the viral capsid protein within the HSP90 complex (Figure 5A). ZIKV NS3 but not the capsid protein is a known binding partner of NS5 (66), thereby confirming the specificity of our ReCLIP approach. These results demonstrate that the HSP90/NS5 interaction occurs during ZIKV infection in trophoblast cells.

Figure 5. ZIKV NS5 interacts with HSP90 in trophoblast cells.

A) Co-immunoprecipitation of ZIKV NS5 and HSP90 was performed using ReCLIP. JAR cells were infected with ZIKV/French Polynesia, ZIKV/Honduras and ZIKV/Malaysia at MOI 10. At 24hpi or 48hpi cells were crosslinked with DSP and quenched with TBS. Samples were then lysed and subjected to immunoprecipitation using an antibody specific to HSP90 or an isotype control (IgG2a). Immunocomplexes were captured with Dynabeads protein G, washed, and analyzed by SDS-PAGE followed by immunoblotting. The blot was probed with antibodies against (HSP90, NS5, NS3, capsid and GAPDH) to detect co-precipitated proteins. Immunoprecipitated samples (IP) and control samples (IgG2a) were included for comparison. B) Densitometry of relative abundance of ZIKV NS5 levels to HSP90 levels. The result is a representative of three independent experiments. Statistical analysis was performed with one-way analysis of variance (ANOVA) followed by Tukey’s multiple-comparison test.

3.5. ZIKV NS5 impairs Akt phosphorylation and EGF-mediated migration

ZIKV NS5 binds to HSP90 during infection, inhibiting HSP90 function and negatively impacting client kinase activity and stability to ultimately suppress protein kinase signaling (54). To determine if the ZIKV NS5 protein alone was sufficient to suppress Akt activity and EGF-induced trophoblast migration, we expressed ZIKV NS5 or empty vector control in 293T cells and measured endogenous Akt and p-Akt levels. Cells expressing NS5 had reduced levels of p-Akt and p-Akt to Akt ratio (Figure 6A and 6B). To define the impact of ZIKV NS5 on trophoblast cells, we generated doxycycline-inducible ZIKV NS5 or mCherry (control)-expressing JAR cell lines. Doxycycline treatment effectively induced the expression of ZIKV NS5 and mCherry (Figures 6C and 6E). To evaluate the effect of ZIKV NS5 on EGF mediated activation of Akt, we induced NS5 or mCherry (control) in JAR cells and treated them with media alone or EGF for 30 minutes followed by immunoblot analysis (Figure 6C). Treatment of mCherry expressing cells with EGF resulted in the phosphorylation of Akt. In contrast, EGF treatment of NS5 expressing cells had lower levels of phosphorylated Akt (Figures 6C and D). Densitometric analysis of the ratio of phosphorylated Akt (P-Akt) to total Akt of NS5 expressing cells demonstrated a significant reduction in EGF-induced Akt similar to that of ZIKV infection (Figure 6D). To elucidate the effect of ZIKV NS5 on EGF mediated trophoblast migration, we conducted transwell migration assays using the doxycycline-inducible NS5 and mCherry JAR trophoblast cells treated with media alone or with EGF. MCherry-expressing cells migrated across the transwell in response to EGF treatment. In contrast, ZIKV NS5-expressing cells failed to migrate in response to EGF treatment (Figures 6E and F). These findings demonstrate that ZIKV NS5 alone is sufficient to suppress Akt phosphorylation even in the presence of EGF and inhibit EGF-mediated trophoblast migration.

Figure 6. ZIKV NS5 blocks EGF-dependent trophoblast migration.

A) 239T cells were transfected with ZIKV NS5 or vector only for 48h and subjected to immunoblotting. B) Densitometry of (A). Statistical analysis was performed with a two-tailed unpaired t-test. (****, P<0.0001). C) Immunoblot analysis of JAR cells transduced with ZIKV NS5 plasmid or mCherry doxycycline-inducible plasmid (control) and treated with doxycycline for 24h and subsequently activated with 100ng/ml of EGF for 30 min and analyzed via immunoblotting. D) Densitometry of (C). E) Migration assay of transduced JAR cells treated with doxycycline for 24 hours, and then transferred to FluoroBlok or translucent transwell inserts. Cells were incubated for 24 hours with 1 ng/ml EGF or a mock treatment in the outer well. Representative stitched images from 4 fields of view are shown for each well. F) Migration index. Error bars represent the standard error of the mean (SEM). The results are representative of three independent experiments. Statistical analysis was performed with a one-way analysis of variance (ANOVA) followed by Tukey’s multiple-comparison test. (***, P<0.0002).

4. Discussion

Previous studies have shown that ZIKV can infect the placenta and disrupt placental development, leading to placental villi immaturity and insufficiency. As such, it is critical to understand the impact of ZIKV infection on trophoblast cell function (23–26,29). A study on trophoblast cells from congenital Zika syndrome (CZS) discordant twins revealed that trophoblasts from CZS-affected twins exhibited significantly decreased gene expression in pathways associated with migration and extracellular matrix organization compared to the uninfected twin, potentially suggesting that cell migration could be compromised as un underlying feature of CZS (80). However, it remains unclear whether trophoblast migration is impaired through a functional lens and how mechanistically Zika virus disrupts this process. Our study now demonstrates that ZIKV infection dysregulates Akt signaling in trophoblast cells, leading to inhibition of EGF-induced Akt-dependent trophoblast migration. We further show that this dysregulation is concomitant with the interaction between ZIKV NS5 and HSP90, and that the NS5 protein itself is sufficient to inhibit Akt phosphorylation and disrupt Akt-dependent trophoblast cell migration. Akt plays a key role in regulating cell migration through several interconnected mechanisms. It promotes glucose uptake and glycolysis, providing the energy necessary for cellular movement (67, 68) and it facilitates cytoskeletal reorganization by influencing microtubule architecture and cell polarity, partly through the phosphorylation and inactivation of GSK3β (42,44). Akt also enhances protein synthesis, including matrix-degrading proteases such as MMP9, and regulates adhesion molecules and integrins (42,43). Given these actions of Akt, we propose that the dysregulation of Akt observed in our system, through reduced phosphorylation and total abundance, could negatively impact trophoblast migration by disrupting multiple downstream processes.

In vitro trophoblast migration studies of viral infection can be challenging to analyze, especially considering cell viability during infection with cytopathic viruses such as ZIKV. To address this challenge, we used a trophoblast migration transwell system with FluoroBlok wells, which allowed for precise identification of migrated cells by obscuring the visibility of unmigrated cells. We also stained cells with Calcein AM viability dye that allows visualization of only live cells with intact membranes (69). This enabled us to specifically quantify live/migrated cells with minimal handling. In addition to controlling cell migration, Akt also promotes cell proliferation and inhibits apoptosis (70). In light of our findings that ZIKV dysregulates Akt signaling in trophoblast cells, we needed to account for potential virus-induced cell death and inhibition of proliferation during migration. Using Z-stacking in our image analyses, we captured both migrated and unmigrated cells from translucent wells for each condition. We also calculated a migration index by normalizing the Calcein AM fluorescence of migrated cells to total cells. This ensured that cell viability was considered in all output calculations for unbiased comparisons. Our observations suggest that NS5-mediated suppression of trophoblast cell migration could be a factor in viral disruption of placental development during ZIKV infection. Furthermore, our data provide valuable groundwork for future animal studies, particularly emphasizing the importance of examining these pathway dysregulation phenotypes in non-human primate models to gain a more comprehensive understanding of the ZIKV pregnancy outcomes downstream of Akt signaling pathway.

The critical roles of Akt kinase in cell biology are not limited to trophoblast cells, as Akt also plays a pivotal role in brain development and neurogenesis. Dysregulation of Akt signaling is linked to brain developmental diseases and microcephaly (71, 72). Additionally, the Akt pathway is essential for the migration of neural progenitor cells for proper CNS development (73). ZIKV has been demonstrated to dysregulate Akt signaling in the CNS during neurodevelopment (74) by inhibiting Akt-dependent neurogenesis (50) and ocular cell development (75), though the mechanisms underlying this dysregulation have not yet been defined. Here, we demonstrate for the first time that phosphorylation of the Akt protein kinase is blocked and the total protein abundance of Akt is reduced during ZIKV infection concomitant with NS5-mediated inhibition of HSP90. Our findings provide insights that can guide future studies by evaluating Akt function in other ZIKV-targeted tissues and assessing the downstream effects of viral disruption of Akt signaling. In addition to trophoblast cells, ZIKV NS5 protein has been shown to inhibit HSP90 in epithelial and neural stem cells to impair innate immune responses (51,54). Inhibition of HSP90 results in the destabilization and degradation of JAK and other client proteins via the proteasome (46,51,76). Other studies have shown that multiple viral proteins of other flaviviruses, such as Dengue virus, interact with HSP90 (52). While it is not yet clear whether other ZIKV proteins besides NS5 interact with HSP90, our data suggest that the ZIKV capsid protein does not engage with HSP90. This outcome indicates that interactions between ZIKV proteins and HSP90 in trophoblast cells are specific to certain viral proteins. Notably, we observed that ZIKV NS3 forms a complex with HSP90, which is consistent with its interaction with NS5 in the ZIKV replication complex (66). However, the possibility of a direct NS3-HSP90 interaction cannot be excluded. Additionally, while ZIKV NS4A and NS4B proteins have been shown to inhibit Akt signaling in human fetal neural stem cells, their interaction with HSP90 has not been studied (50). During infection with Dengue virus, however, NS3 and NS4B were shown to interact with HSP90 (52). Our findings establish that ZIKV NS5 forms a complex with HSP90 and is sufficient to inhibit Akt phosphorylation and trophoblast migration. We postulate that during ZIKV infection, the overall effect of multiple viral proteins on HSP90 interaction and regulation may further contribute to NS5 regulation of Akt function. Future research to elucidate the exact amino acid partners in the NS5/HSP90 interaction could provide valuable insights and guide the development of more effective therapeutics.

We observed that ZIKV/Honduras led to a greater reduction of total Akt levels compared to infection by ZIKV/French Polynesia, yet both virus strains similarly reduced the relative abundance of phosphorylated Akt. Moreover, both viruses similarly inhibited trophoblast migration, suggesting that phosphorylation of Akt is a critical determinant of trophoblast migration. This conclusion is further supported by the observation that chemical inhibition of Akt phosphorylation was sufficient to prevent trophoblast migration. Additionally, exogenous NS5 expression in trophoblast cells inhibited Akt phosphorylation without reducing Akt levels, yet effectively blocked trophoblast migration, suggesting that inhibition of Akt phosphorylation is the main factor preventing EGF-mediated migration. As such, the reduction of total Akt abundance during ZIKV infection further exacerbates the disruption of Akt signaling by reducing the pool of Akt available for phosphorylation/activation.

Our findings offer valuable insights into the molecular mechanisms by which ZIKV impairs placental development, specifically through its interaction with the EGF signaling pathway. ZIKV remains a significant public health concern, with 24,000 confirmed cases reported in the Americas in 2024 alone (18). Children born with CZS continue to face lifelong challenges, including severe motor skill impairment, impairment of linguistic development, and vision impairment among many other developmental challenges (75). By further exploring the mechanisms behind the impact of ZIKV infection on pregnancy, we can develop interventions to mitigate these outcomes and improve the quality of life for affected families. Additionally, understanding the ZIKV/host interface of NS5, HSP90, and EGF/Akt regulation provides broader insights into other congenital infections. Notably, between July and August 2024, the WHO and CDC released an alert and travel advisory for pregnant women regarding the Oropouche virus as infection with this virus was linked to over 8,000 confirmed cases in the Americas and potential vertical transmission in Brazil that resulted in stillbirth, spontaneous abortion, and microcephaly (76,77). These emerging virus threats reinforce the need to understand how viruses impact pregnancy and fetal health and underscore the need for continued research into viral infections during pregnancy.

5. Conclusions

Zika virus (ZIKV) infects trophoblast cells and impairs their migration, even in the presence of the potent migration activator EGF. This outcome is directed through ZIKV NS5 protein-mediated interaction and inhibition of HSP90, which destabilizes Akt and prevents its activation by EGF. Consequently, Akt signaling is disrupted, ultimately leading to Akt degradation (Figure 7). Notably, chemical inhibition of HSP90 replicates the effects of ZIKV, reducing Akt phosphorylation and stability. Akt is a critical kinase in cells involved in cytoskeleton rearrangement, glucose uptake, glycolysis, protein synthesis, and proliferation—all of which are necessary for cell migration regulation (42–44,65–66). The inhibition of Akt signaling by ZIKV compromises trophoblast EGF-mediated migration, a process critical for implantation, proper placentation, and remodeling of maternal spiral arteries (30,36,39). This disruption can contribute to placental immaturity and insufficiency (24,26–28), which may result in fetal growth restriction or, in severe cases, fetal demise (10–13). These findings highlight a potential mechanism by which ZIKV impacts pregnancy outcomes.

Figure 7. Diagram of ZIKV inhibition of Akt signaling in trophoblast cells.

Zika virus (ZIKV) infection in trophoblast cells leads to the inhibition of HSP90, primarily mediated by the NS5 protein, which interferes with the proper stabilization of Akt and other HSP90-dependent kinases. This inhibition prevents Akt from being activated by maternal growth factors such as EGF. Consequently, Akt signaling is disrupted, contributing to its degradation. A decrease in p-Akt levels leads to reduced cytoskeletal rearrangement, glucose uptake, protein synthesis, and proliferation, all of which are essential for effective trophoblast migration (42–44,65–66). The impaired Akt pathway hampers trophoblast migration into the maternal decidua. This results in the formation of immature placental villi incapable of supporting a developing fetus (24,26). These effects can contribute to placental immaturity, potentially leading to fetal growth restriction or, in severe cases, fetal demise (10–13).

• The Akt signaling pathway in the placenta is dysregulated by ZIKV

• ZIKV disrupts EGF-mediated trophoblast migration

• HSP90 is required for trophoblast EGF-mediated Akt signaling and migration

• ZIKV NS5 interacts with HSP90

• ZIKV NS5 disrupts EGF-mediated Akt signaling and trophoblast migration