Introduction

Zika virus (ZIKV) is an arthropod borne (arbo) virus and belongs to the family of Flaviviridae. Since its discovery in Uganda in 1947, ZIKV has invaded different territories around the world. In 2015, ZIKV was reported in Brazil resulting in an epidemic of unprecedented scale in the Americas and the Caribbean1,2. Severe complications due to ZIKV infection are observed during pregnancy, including congenital abnormalities such as microcephaly and subcortical calcifications3. In adults, acute symptomatic ZIKV infection is typically characterized by a self-limiting disease with fever, rash, conjunctivitis, myalgia and arthralgia/arthritis4. Arthralgia is reported in over 70% of symptomatic ZIKV cases including persistent or recurrent arthralgia for more than 30 days5,6.Virus-induced arthralgia has previously been reported following infection with the alphaviruses Ross River virus (RRV) and chikungunya virus (CHIKV)7,8. Infection with RRV can perturb osteoblast function and trigger pathologic bone loss due to induction of interleukin-6 (IL6), and contributes to virus-induced arthritis7. Osteoblasts are responsible for the deposition of bone matrix and mineralization of the bone matrix. However, osteoblast function may be altered due to infection as reported previously for other viruses such as hepatitis C virus (HCV), measles virus (MV) and human immunodeficiency virus (HIV)9.

ZIKV has been detected in synovial fluid of a patient with arthralgia10, and a recent study has shown the susceptibility of an osteoblast-like cell line for ZIKV infection11. Osteoblasts originate from mesenchymal stem cells (MSCs) and in a recent case of miscarriage associated with ZIKV infection, ZIKV displayed tropism for fetal MSC in the perichondrium12. However, it is not known if MSC- derived osteoblasts are susceptible to ZIKV infection, and whether infection of osteoblasts affects their function thereby contributing to ZIKV pathogenesis and ZIKV-associated osteoarticular complications. Therefore, in the current study, we determined the susceptibility of primary human MSC-derived osteoblasts to infection with ZIKV and its effect on differentiation, maturation and function of these cells.

Results

ZIKV infects osteoblasts with high titers

In order to determine whether the osteoblasts are susceptible to infection with ZIKV, we infected osteoblast cultures with ZIKV at an moi of 5. The infected cultures did not show evidence for a cytopathic effect (CPE) due to ZIKV infection throughout the culture period (Supplementary Fig. 1). CHIKV infection resulted in extensive CPE within 4 days post-infection and as such samples were not available for subsequent analyses. Similar to ZIKV, DENV infection did not result in CPE (supplementary Fig. 4). ZIKV infection of osteoblasts was confirmed by IFA at day 4, 7, 11 and 17 post-infection (Fig. 1 and Supplementary Fig. 2). Interestingly, despite using a high moi, not all cells were infected with ZIKV and the level of ZIKV infection remained similar throughout the experiment. However, ZIKV infected differentiating osteoblasts and produced high infectious titers of up to 107 TCID50/ml within 2 days post infection. No significant evidence of DENV infection was observed by IFA or virus titration (Supplementary Figs 34). Virus growth kinetics and IFA during the differentiation period showed that osteoblast cultures were persistently infected with ZIKV and shedding of infectious virions was observed over the period of three weeks post-infection for both donors (Fig. 1a–c).

Figure 1
figure 1

Replication of ZIKV in primary osteoblasts. Culture supernatant was collected at different time points after infection of primary osteoblasts by ZIKV (moi = 5). (a) Growth curve kinetics of ZIKV infection in osteoblasts from Donor 4266 (closed circles) and Donor 3520 (open squares) during differentiation over the period of 3 weeks. Error bars represent the standard error of mean (S.E.M). (b) Representative immunofluorescent images of ZIKV-infected cells stained for ZIKV antigen (green) and nuclei (blue), and (c) uninfected controls at day 4 post-infection. Magnification 200x.

ZIKV perturbs osteoblast differentiation and mineralization

The effect of ZIKV infection on the differentiation and maturation of osteoblasts was determined in infected cultures by quantifying the ALP activity and mineral content at different time points (Day 7, 11, 14, 18 and 21) post infection. These time points were selected as optimal time points for quantification of differentiation and maturation of MSC-derived osteoblasts, based on previous studies13,14. In ZIKV-infected osteoblasts derived from 2 different MSC donors, ALP activity was significantly reduced at day 11 post-infection compared to uninfected controls (Fig. 2a,b). Additionally, ZIKV infection significantly reduced osteoblast maturation in terms of mineral content in ZIKV-infected osteoblasts compared to uninfected controls at day 18 and 21 post-infection (Fig. 2c,d).

Figure 2
figure 2

Effect of ZIKV infection on osteoblast differentiation and maturation. Effect of ZIKV infection on osteoblast differentiation is measured by alkaline phosphatase (ALP) activity in cultures from (a) Donor 4266 and (b) Donor 3520, and effect on mineralization is measured as a concentration of calcium present in the cultures from c) Donor 4266 and (d) Donor 3520. Results are compared between ZIKV -infected (white bars) and uninfected controls (black bars) and ALP levels are normalized against total protein. Error bars represent the standard error of mean. *p < 0.05.

ZIKV affects osteoblast marker genes

To quantify the impact of ZIKV infection on the expression levels of key factors during ZIKV infection, which play an instrumental role in determining the osteoblasts phenotype, we quantified gene expression of ALP, Runt-related transcription factor 2 (RUNX2, a key transcription factor for osteoblast differentiation) and a classical inflammatory mediator interleukin 6 (IL6). A significant reduction of ALP (Fig. 3a,b) and RUNX2 expression (Fig. 3c,d) was observed in ZIKV-infected differentiating osteoblasts compared to uninfected controls (p value < 0.05) at day 7 post-infection. Interestingly, the levels of IL6 were significantly increased in infected osteoblasts (Fig. 3e,f).

Figure 3
figure 3

Gene expression levels of analysed genes after ZIKV infection. Gene expression of key transcription factors from ZIKV infected osteoblasts (white bars) versus uninfected controls (black bars) in (Left panel) Donor 4266 and (Right panel) Donor 3520. Gene expression was corrected for house keeping gene, GAPDH. Error bars represent the standard error of mean. *p < 0.05.

Discussion

The development of arthralgia/arthritis has been a hallmark clinical symptom in several arbovirus infections, including ZIKV, CHIKV and dengue viruses9,15,16. While generally not considered as severe as in CHIKV infection, arthralgia/arthritis has been reported in over 70% of ZIKV cases6. In this study we investigated the susceptibility of MSC-derived osteoblasts to ZIKV infection and the effects of this infection on the phenotype of osteoblasts.

ZIKV readily infects differentiating primary human osteoblasts and grows to high titers within 2 days post infection. These findings are in line with a recent report which confirms the susceptibility of a human osteoblast-like cell line (HOBIT) to ZIKV11, as well as observations from other arthritogenic arboviruses such as CHIKV17. However, in contrast to previous studies we did not observe CPE in ZIKV-infected primary osteoblasts. This disparity may occur in part due to the differences in in vitro models, as primary osteoblasts derived from MSCs were used in the current study compared to an osteoblast-like cell line in the previous ZIKV study and bone fragment-derived osteoblasts used in the CHIKV study. Additionally, flavivirus replication has been shown to incompletely inhibit host cell macromolecular synthesis, which may result in non-cytopathic persistent infections18,19,20. In the absence of CPE, ZIKV infection resulted in a persistent infection of osteoblasts for up to 3 weeks post infection, which suggests there may be viral persistence similar to CHIKV17.

Osteoblasts play an important role in bone remodeling, which is a tightly regulated process, requiring a balance in bone resorption and bone formation. Interestingly, persistent ZIKV infection had a direct effect on the differentiation, maturation and function of primary osteoblasts. Osteoblast differentiation comprises a highly coordinated sequence of events and is regulated by the activities of several key transcription factors. The key transcription regulator RUNX2 plays a central role by modulating the commitment of osteoprogenitors and expression of major bone matrix genes by activating different pathways21. Following osteoblast commitment, increased levels of ALP are considered as the marker of osteogenic differentiation as it is one of the first functional enzymes observed prior to the process of mineralization. Indeed, by reducing levels of the mineralization inhibitor pyrophosphate, ALP activity is critical for the mineralization process22. Thus, our findings showing significantly reduced expressions of RUNX2 and ALP in persistently infected osteoblasts followed by diminished mineral depositions are of particular interest in view of ZIKV-induced alterations in the osteoblast phenotype.

In addition to bone formation biomarkers, elevated levels of IL6 comparable to that described for other arthritogenic arboviruses were also observed following ZIKV infection. This key pro-inflammatory mediator plays a pivotal role in the pathophysiology of rheumatoid arthritis, where it has been associated with stimulating neutrophil migration, osteoclast maturation and pannus proliferation23. During alphavirus infection, IL6 has been shown to indirectly disturb the bone homeostasis by stimulating the induction of bone resorption mediators in osteoblast cultures, resulting in bone loss and joint inflammation7,17. Future studies will focus on identifying key biomarkers which will allow us to predict the osteoarticular complications and disease severity following ZIKV infection.

In conclusion, these data clearly demonstrate that osteoblasts are susceptible to infection by ZIKV and that infection results in reduced differentiation and maturation of osteoblasts. The impaired function of osteoblasts can subsequently trigger an imbalance in bone homeostasis and induce bone-related disorders. These data warrant further studies to delineate the molecular mechanisms behind ZIKV pathogenesis in bone development in vitro and in vivo.

Materials and Methods

Cell cultures

Human bone marrow-derived MSCs of two healthy donors (Donor 4266 and 3520, 33 and 20 year old males, respectively) were purchased from Lonza (PT-2501). MSCs were differentiated into osteoblasts as described previously24. Briefly, MSCs were cultured in αMEM medium (Gibco, Thermofisher) supplemented with 10% heat-inactivated fetal calf serum a (FCS, Sigma), 20 mM Hepes (Sigma), 100 U/mL penicillin (Lonza) and 100 µg/mL streptomycin (Thermofisher) and 1.8 mM CaCl2 (Sigma) at 37 °C and 5% CO2 in a humidified atmosphere. Medium on day 3 post-seeding was supplemented with 100 mM dexamethasone (dex) and 10 mM β-glycerophosphate (osteogenic medium) resulting in differentiation into mineralizing osteoblasts within 2–3 weeks14. The media were refreshed twice per week. Vero cells (African green monkey kidney epithelial cells, ATCC CCL-81) were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Lonza, the Netherlands) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Greiner Bio-One, Austria), 2 mM L-glutamin (Lonza), 1% sodium bicarbonate (Lonza), 1% Hepes (Lonza), 100 U/mL penicillin (Lonza) and 100 µg/mL streptomycin (Lonza) at 37 °C and 5% CO2 in a humidified atmosphere.

Virus

Zika virus Suriname ZIKVNL00013 (ZIKVAS-Sur16) was isolated from a patient in The Netherlands (EVAg no. 011V-01621)12. This strain was previously shown to have a similar phenotype as the prototypic Asian lineage of ZIKV25. The virus stock used in this study was grown in Vero cells and passage number 3 was used for the current study. Virus titers in the supernatants were determined by endpoint titrations on Vero cells as described previously25. In some experiments, dengue virus (DEN 2 16681 strain) and chikungunya virus (CHIKV/IND/NL10/152) were included as controls. Briefly, tenfold serial dilution were inoculated onto a monolayer of Vero cells in a 96-wells plate (2 × 104 cells/well). Cytopathic effect (CPE) was used as read out and determined at 5 days post-infection (dpi), and virus titers were calculated as the 50% tissue culture infective dose (TCID50) using the Spearman-Kärber method26. An initial 1:10 dilution of supernatant resulted in a detection limit of 101.5 TCID50/ ml.

Replication kinetics of ZIKV

MSCs were seeded three days prior to infection (day = −3). Three days post-seeding (day = 0), MSCs were stimulated into osteoblasts by adding osteogenic medium. After 6 hours of stimulation, osteoblast cultures were infected at a multiplicity of infection (moi) of 5 with ZIKVfor 1 hour at 37 °C in 5% CO2. The moi was based on titers determined on Vero cells. After incubation, the supernatant was removed and cells were washed three times with αMEM medium containing 10% heat-inactivated FCS followed by osteogenic medium and cells were cultured twice weekly up to three weeks depending on the experiment. Uninfected controls were cultured in parallel. To determine the ZIKV infectious titers produced, cell supernatants were collected at different time points post infection. Supernatant was stored at −80 °C until further use. Experiments were performed in triplicate with two different MSC donors.

Immunofluorescence assay

Infected cells from the replication growth kinetics assay were fixed with 4% PFA at days 4, 7, 11 and 17 post infection, permeabilized with 70% ethanol and stained using an immunofluorescence assay (IFA) as described previously25. Briefly, cells were incubated with mouse monoclonal antibody anti-flavivirus group antigen (MAB10216) clone D1–4G2-4-15 (Millipore, Germany) followed by staining with goat anti-mouse IgG conjugated with Alexa Fluor 488 (Life technologies, the Netherlands). After incubation, cells were mounted in ProLong® Diamond Antifade Mountant with DAPI (Life technologies, USA). Uninfected cells and ZIKV-infected cells stained with mouse isotype IgG2a antibody (Dako, Denmark) were used as negative controls. ZIKV-infected cells were identified by using a Zeiss LSM 700 confocal laser scanning microscope fitted on an Axio observer Z1 inverted microscope (Zeiss, Breda, the Netherlands). All images were processed using Zen 2010 software (Zeiss).

Alkaline phosphatase, mineralization, and protein assays

Alkaline phosphatase (ALP) and calcium measurements were performed at different time points based on gene expression data as described previously13,14. Briefly, ALP activity was determined by an enzymatic reaction, where the ALP-mediated conversion of para-nitrophenylphosphate (pNPP) (Sigma) to paranitrophenol (PNP) during 10 min at 37 °C was measured at 405 nm. ALP results were adjusted for protein content of the cell lysates. For protein measurement, 200 µL of working reagent was added to 25 µL of cell lysate. The mixture was incubated for 30 min at 37 °C, cooled down to room temperature (RT) for 10 min, and absorbance was measured at 595 nm. For calcium measurements, cell lysates were incubated overnight with 0.24 M HCl at 4 °C. Calcium content was determined colorimetrically using a calcium assay reagent prepared by combining 1 M ethanolamine buffer (pH 10.6) with 0.35 mM O-cresolphthalein complex one in a ratio of 1:1. All measurements were performed using a Victor2 plate reader.

Quantification of mRNA expression

RNA was extracted from ZIKV-infected and uninfected cells at different time points during osteoblast differentiation using TRIzol reagent (Thermo Fisher Scientific). RNA isolation, cDNA synthesis, and PCR reactions were performed as described previously24. Oligonucleotide primer pairs were designed to be either on exon boundaries or spanning at least one intron (Table 1). Gene expression was corrected for the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH).

Table 1 Primer sequences of the analyzed genes.

Statistical analysis

The statistical analyses were performed using GraphPad Prism 5.01 software. All results are expressed as means with standard error of the mean (S.E.M.). Mann Whitney U test was used for the comparison between two groups (infected versus un-infected). P value ≤ 0.05 was considered significant.