Human induced pluripotent stem cells are resistant to human cytomegalovirus infection primarily at the attachment level due to the reduced expression of cell-surface heparan sulfate

ABSTRACT

Cytomegalovirus (CMV), a type of herpes virus, is the predominant cause of congenital anomalies due to intrauterine infections in humans. Adverse outcomes related to intrauterine infections with human cytomegalovirus (HCMV) vary widely, depending on factors such as fetal infection timing, infection route, and viral virulence. The precise mechanism underlying HCMV susceptibility remains unclear. In this study, we compared the susceptibility of neonatal human dermal fibroblast cells (NHDFCs) and human induced pluripotent stem cells (hiPSCs) derived from NHDFCs, which are genetically identical to HCMV, using immunostaining, microarray, in situ hybridization, quantitative PCR, and scanning electron microscopy. These cells were previously used to compare CMV susceptibility, but the underlying mechanisms were not fully elucidated. HCMV susceptibility of hiPSCs was significantly lower in the earliest phase. No shared gene ontologies were observed immediately post-infection between the two cell types using microarray analysis. Early-stage expression of HCMV antigens and the HCMV genome was minimal in immunostaining and in in situ hybridization in hiPSCs. This strongly suggests that HCMV does not readily bind to hiPSC surfaces. Scanning electron microscopy performed using the NanoSuit method confirmed the scarcity of HCMV particles on hiPSC surfaces. The zeta potential and charge mapping of the charged surface in NHDFCs and hiPSCs exhibited minimal differences when assessed using zeta potential analyzer and scanning ion conductance microscopy; however, the expression of heparan sulfate (HS) was significantly lower in hiPSCs compared with that in NHDFCs. Thus, HS expression could be a primary determinant of HCMV resistance in hiPSCs at the attachment level.

IMPORTANCE

Numerous factors such as attachment, virus particle entry, transcription, and virus particle egress can affect viral susceptibility. Since 1984, pluripotent cells are known to be CMV resistant; however, the exact mechanism underlying this resistance remains elusive. Some researchers suggest inhibition in the initial phase of HCMV binding, while others have suggested the possibility of a sufficient amount of HCMV entering the cells to establish latency. This study demonstrates that HCMV particles rarely attach to the surfaces of hiPSCs. This is not due to limitations in the electrostatic interactions between the surface of hiPSCs and HCMV particles, but due to HS expression. Therefore, HS expression should be recognized as a key factor in determining the susceptibility of HCMV in congenital infection in vitro and in vivo. In the future, drugs targeting HS may become crucial for the treatment of congenital CMV infections. Thus, further research in this area is warranted.

INTRODUCTION

Cytomegalovirus (CMV), a member of the herpes virus family, is a major cause of intrauterine infections leading to congenital anomalies in humans. Depending on the timing, route, and virulence of the fetal infection, intrauterine infection with human cytomegalovirus (HCMV) may cause various abnormalities (1). However, mouse embryos injected with murine cytomegalovirus (MCMV)-infected blastocysts do not express viral genes, indicating a lack of susceptibility to MCMV (2).

Mouse embryonic stem (ES) cells appear to be non-permissive to MCMV infection, with the MCMV immediate-early (IE) promoter showing no activation in the ES cells from transgenic mice (3). Human NTera2/D1 embryonic carcinoma (NT2) cells serve as an effective model for studying the regulatory mechanisms underlying major immediate-early (MIE) enhancer/promoter silencing during quiescent HCMV infection (4–6). This is due to the blockage of HCMV replication in embryonic NT2 cells due to the inhibition of viral MIE gene expression, unlike in differentiated cells (4, 5, 7). In a previous study, we showed that the susceptibility of mouse ES cell to MCMV is inhibited in a multistep manner. In addition, mouse induced pluripotent stem (iPS) cells, similar to mouse ES cells, exhibit resistance to MCMV. These novel attributes of pluripotent stem cells (ES or iPS) related to CMV infection have been described previously (8, 9). Other researchers have confirmed that undifferentiated ES or human induced pluripotent stem cells (hiPSCs) do not permit HCMV infection (10–14). However, the key mechanisms underlying the resistance to HCMV in human pluripotent stem cells remain unknown. Berger et al. (12) suggested that this resistance arises from a block at the initial stage of viral binding, whereas another group, using laboratory (AD169) and clinical HCMV strains, proposed that HCMV enters ESCs and establishes latency (14).

The infectious behavior of CMV in human ES or pluripotent stem cells offers a valuable model for elucidating the mechanisms underlying CMV resistance during early embryogenesis. Here, by comparing the HCMV susceptibility of neonatal human dermal fibroblast cells (NHDFCs) and hiPSCs with the same genetic background, we aimed to elucidate the primary inhibitory mechanism of HCMV.

RESULTS

Establishment of hiPSCs from NHDFCs

NHDFCs (Fig. 1A) were reprogrammed using the CytoTune-iPS v.2.0 Sendai Reprogramming Kit. After several weeks of culture, clearly recognizable tightly packed colonies with morphologies similar to those of human embryonic stem cells (hESCs) appeared on the feeder cells. These colonies were transferred to and cultured in a feederless dish (Fig. 1A). The pluripotent markers (alkaline phosphatase, Nanog, Oct3/4, and SSEA-4) were expressed in the hiPSCs (Fig. 1B).

Fig 1.

Generation of hiPSCs from NHDFCs. (A) Tightly packed colony with morphology similar to that of a hES colony after culturing for several weeks post-transduction. NHDFCs were transduced overnight with Sendai viruses containing human Oct3/4, Sox2, Klf4, and c-Myc. Scale bar: 280 µm. (B) Expression of pluripotent markers [ALP, Nanog, Oct3/4, and SSEA-4 (red)] in hiPSCs with corresponding nuclear markers (DAPI, blue). Scale bar: 70 µm. DAPI, 4′,6-diamidino-2-phenylindole; hES, human embryonic stem; hiPSC, human induced pluripotent stem cell; NHDFC, neonatal human dermal fibroblast cell.

Microarray analysis results revealed the upregulation of 88 stem cell genes (by more than twofold) in hiPSCs when compared to that in NHDFCs. Representative upregulated genes included ESRG (2,033.6-fold increase), POU5.1 (OCT3/4) (1,714.4-fold increase), NANOG (452.8-fold increase), and SOX2 (307.3-fold increase) (Data S1). Pluripotency was confirmed by the successful differentiation into neurons (data not shown) and the beating of cardiac cells (Movie S1).

Comparison of HCMV infection in hiPSCs and NHDFCs

NHDFCs (Fig. 2A) and hiPSCs cultured on feeder cells (Fig. 2D) were infected with HCMV (AD169) at a multiplicity of infection (MOI) of 10. Cytopathic changes were observed in NHDFCs at 3 (Fig. 2B) and 5 days post-infection (dpi) (Fig. 2C). In contrast, hiPSCs displayed no cytopathic changes and maintained their colony shape with clear edges at 3 (Fig. 2E) and 5 dpi (Fig. 2F).

Fig 2.

Comparison of HCMV (AD169) infection between NHDFCs and hiPSCs. (A–C) NHDFCs. (A) Control (uninfected). (B) 3 days post-HCMV infection at an MOI of 10. (C) Seven days post-HCMV infection at an MOI of 10. Arrows indicate cytopathic changes. (D–F) hiPSCs. (C) Control (uninfected). (D) Three days post-HCMV infection at an MOI of 10. (E) Seven days post-HCMV infection at an MOI of 10. Black lines represent the border of the hiPSC colonies. Scale bar: 100 µm. (G–K) HCMV pp65 expression in NHDFCs and hiPSCs 12 h post-infection. (G and I) HCMV pp65 antigen-positive (green) NHDFCs (G) and hiPSCs (I). (H and J) DAPI staining (blue) of NHDFCs (H) and hiPSCs (J). Scale bar: 50 µm. (K) Comparison of the HCMV pp65-positive cell ratio between NHDFCs and hiPSCs. ★★P < 0.01. (L–P) HCMV IE1/2 + pUL44 expression in NHDFCs and iPSCs 12 h post-infection. (L and N) HCMV IE1/2 + pUL44 antigen-positive (green) NHDFCs (L) and hiPSCs (N). (M and O) DAPI staining (blue) of NHDFCs (M) and hiPSCs (O). Scale bar: 40 µm. (K) Comparison of the HCMV IE1/2 + pUL44-positive cell ratio between NHDFCs and hiPSCs. ★★P < 0.01. (Q–S) NHDFCs after HCMV infection at an MOI of 10, 12 h post-infection (high magnification). Phase contrast image (Q), Immunofluorescence images of pp65 (R) and DAPI (S). (T–V) hiPSCs after HCMV infection at an MOI of 10, 12 h post-infection (high magnification). Phase contrast image (T), Immunofluorescence images of pp65 (U) and DAPI (V). Scale bar: 20 µm. HCMV, human cytomegalovirus; hiPSC, human induced pluripotent stem cell; IE1/2, immediate-early 1/2; MOI, multiplicity of infection; NHDFC, neonatal human dermal fibroblast cell.

Immunocytochemical analysis was performed to examine the differences in the infection phases between NHDFCs and feeder-free hiPSCs at an HCMV MOI of 10. Multiple phosphoprotein 65 (pp65)-positive cells were observed in NHDFCs (Fig. 2G); the cell locations were identified using 4′,6-diamidino-2-phenylindole (DAPI) staining (Fig. 2H) (infection ratio: 85.2% ± 2.1%). In contrast, pp65-positive cells were scarcely observed in hiPSCs (Fig. 2I and J) 12 h post-infection (hpi, infection ratio: 0.2% ± 0.3%), demonstrating a significant difference (Fig. 2K, P < 0.01).

Multiple cells positive for HCMV with immediate-early 1/2 (IE1/2) and early (E) (pUL44) antigens were observed in NHDFCs (Fig. 2L and M) (infection ratio: 93.6% ± 3.2%) but were hardly observed in hiPSCs (Fig. 2N and O) at 12 hpi (infection ratio: 0.6% ± 0.4%), demonstrating a significant difference (Fig. 2P, P < 0.01). In hiPSCs, pp65, a component of the virus particles, was examined using a high-resolution camera with high magnification. pp65 was highly expressed in NHDFCs (Fig. 2Q through S); however, no expression was observed in hiPSCs under the same conditions (Fig. 2T through V).

Comparison of the host gene expression between NHDFCs and hiPSCs at the very early phase of HCMV infection

NHDFCs and hiPSCs were infected at an MOI of 10, and mRNA was harvested from each cell type, 30 min after infection. The scatter plots obtained using raw data showed that 759 genes were upregulated (>2.0) and 180 were downregulated (<0.5) in NHDFCs (Fig. 3A). In contrast, 55 genes were upregulated (>1.8) and 61 were downregulated (< 0.56) in hiPSCs (Fig. 3A; Data S2). Cluster analysis indicated that the pattern of expression changes due to HCMV infection differed considerably between the two cell types (Fig. 3B). The Venn diagram (Fig. 3C) indicates that the common genes upregulated or downregulated by HCMV infection in both hiPSCs and NHDFCs were limited to seven and three genes, respectively. Results of a gene ontology biological process term analysis using DAVID Bioinformatics Resources revealed significant upregulation and downregulation of 43 and 16 genes, respectively, during HCMV infection in NHDFCs. The representative gene ontology categories included “regulation of transcription from RNA polymerase II promoter,” “protein autophosphorylation,” and “cell adhesion.” No significant gene ontology biological processes were identified in hiPSCs following HCMV infection. The present data indicate that no significant intracellular gene activation occurred in HCMV-infected hiPSCs at the very early phase of infection (Fig. 3D).

Fig 3.

Results of microarray analysis between NHDFCs and hiPSCs. (A) Scatter plots using raw data showing that 759 genes were upregulated (>2.0-fold) and 180 genes were downregulated (<0.5-fold) in NHDFCs (black dots). Fifty genes were upregulated (>1.8-fold) and 61 genes were downregulated (<0.56-fold) in hiPSCs (black dots). (B) Cluster analysis comparison between NHDFCs and hiPSCs. (C) Venn diagrams showing the commonly regulated genes between hiPSCs and NHDFCs due to HCMV infection: seven upregulated and three downregulated. (D) Gene ontology biological process terms in HCMV infection in NHDFCs. HCMV, human cytomegalovirus; hiPSC, human induced pluripotent stem cell; NHDFC, neonatal human dermal fibroblast cell.

Comparative analysis of the HCMV genome using in situ hybridization

NHDFCs and hiPSCs were infected with HCMV at MOIs of 1 and 10. After the cells were fixed, cell pellets were collected and paraffin sections were prepared for in situ hybridization to visualize the HCMV genome. HCMV DNA was visualized as green fluorescent protein (GFP) dot signals (indicative of HCMV genomes) at 1 hpi. Uninfected control samples showed no GFP dot signals in NHDFCs (Fig. 4A). When the infection load increased, the number of GFP dot signals representing the HCMV genome inside the cells increased in NHDFCs (Fig. 4B and C), and the same phenomenon was observed in hiPSCs (Fig. 4D through F). However, the number of signals in hiPSCs was significantly lower than that in NHDFCs. These results strongly suggest that HCMV particles have difficulty entering or attaching to hiPSCs.

Fig 4.

Visualization of the HCMV genome through in situ hybridization in NHDFCs and hiPSCs. (A–C) NHDFCs. (A) Control (uninfected). (B) HCMV (AD169) infection at an MOI of 1, 1 h post-infection. (C) HCMV (AD169) infection at an MOI of 10. The HCMV genome was visualized with GFP (green). (D–F) hiPSCs. (D) Control (uninfected). (E) HCMV (AD169) infection at an MOI of 1, 1 h post-infection. (F) HCMV (AD169) infection at an MOI of 10. The HCMV genome was visualized with GFP (green). Scale bar: 25 µm. GFP, green fluorescent protein; HCMV, human cytomegalovirus; hiPSC, human induced pluripotent stem cell; MOI, multiplicity of infection; NHDFC, neonatal human dermal fibroblast cell.

Observation and quantification of the HCMV particles on the cell surface

The HCMV particles were allowed to attach to the cell surface at 4°C for 1 h, at an MOI of 10, followed by three rinses with phosphate-buffered saline (PBS). The NanoSuit method, which provides a simple and quick way of observing wet biological specimens without traditional coating methods (15–19), was used to visualize the HCMV particles using field-emission scanning electron microscopy (FE-SEM). Numerous HCMV particles were observed on the surface of NHDFCs (Fig. 5A and B). Conversely, HCMV particles were rarely observed on the surface of hiPSCs (Fig. 5C and D).

Fig 5.

Observation and quantification of HCMV particles on the cell surface. (A–D) Observation of HCMV particles using FE-SEM and the NanoSuit method. (A) HCMV particles on the surface of NHDFCs. Scale bar: 10 µm. (B) Magnified area (square) of panel A. White arrows represent HCMV particles. Scale bar: 5 µm. (C) HCMV particles on the surface of hiPSCs. Scale bar: 10 µm. (D) Magnified area (square) of (C). White arrows represent HCMV particles. Scale bar: 5 µm. (E) Comparison of HCMV genome copies at the attachment level between NHDFCs and hiPSCs at an MOI of 10. ★★P < 0.01. FE-SEM, field-emission scanning electron microscopy; HCMV, human cytomegalovirus; hiPSC, human induced pluripotent stem cell; NHDFC, neonatal human dermal fibroblast cell.

The comparative cycle threshold (Ct) method (ΔΔCt method) was employed to compare the level of HCMV genome copies attached to NHDFCs and hiPSCs at an MOI of 10. The relative copy number of the HCMV IE1 gene was standardized using the RNase P gene copy number. Each sample type was tested five times. The average Ct for HCMV IE1 in NHDFCs was 27.26 ± 0.40 (standard deviation), while that for the RNase P gene was 30.43 ± 0.44. The ΔCt (HCMV IE1-RNase P) for NHDFCs was −3.17 ± 0.59. The average Ct for HCMV IE1 in hiPSCs was 33.00 ± 0.14, while that for the RNase P gene was 29.68 ± 0.17. Thus, the ΔCt (HCMV IE1-RNase P) for hiPSCs was 3.32 ± 0.22. The ΔΔCt was –6.49 ± 0.59, indicating a 89.9-fold (ranging from 59.7 to 135.3) difference in the HCMV genome abundance at the attachment phase (Fig. 5E; Table 1). These results suggested that hiPSCs possess mechanisms to resist HCMV infection during the attachment stage.

TABLE 1.

Calculation of fold change in HCMV IE1 gene expression between NHDFCs and hiPSCs using the comparative ΔΔCt method^a

Sample	HCMV IE1 average ct	RNase P average ct	Δct: HCMV IE1-RNase P	ΔΔct	Fold difference in HCMV IE1 gene of NHDFCs to HCMV IE1 gene of hiPSCs = 2-ΔΔCt
NHDFCs	27.26 ± 0.40	30.43 ± 0.44	−3.17 ± 0.59	ΔCt NHDFCs − ΔCt hiPSCs: −6.49 ± 0.59	89.9 (59.7–135.3)
hiPSCs	33.00 ± 0.14	29.68 ± 0.17	3.32 ± 0.22	ΔCt hiPSCs − ΔCt hiPSCs: 0 ± 0.22	1 (0.86–1.17)

^a hiPSC, human induced pluripotent stem cell; NHDFC, neonatal human dermal fibroblast cell.

Measurement of zeta potential on the surface of NHDFCs and hiPSCs

The structure of a cell for a flat-plate sample is shown in Fig. 6A. Specifically, a sheet-like or flat-plate sample was adhered to the bottom surface of a box-shaped quartz cell, wherein a monitor particle solution was injected. Electrophoretic measurements of the monitored particles were performed at each depth level in the cell (Fig. 6B). The electrophoretic velocity of the solid surface was calculated from the obtained electric osmosis profile and Mori and Okamoto’s formula. We compared the zeta potential of cell surfaces between living NHDFCs and confluent hiPSCs cultured on the glass using the surface zeta potential measurement method. Figure 6C shows representative electrophoretic measurements of the monitored particle data for both NHDFCs and hiPSCs. The average zeta potential of NHDFCs was −6.59 ± 2.02 mV, and that of hiPSCs was −8.58 ± 3.77 mV (n = 9). The zeta potential of both cell surfaces was slightly negative, and there were no substantial differences between the two cell types (Fig. 6D).

Fig 6.

Measurement of the zeta potential on the surface of NHDFCs and hiPSCs. (A) Schematic representation of the flat-surface zeta cell unit of a zeta potential particle size analyzer ELSZ-1000. (B) Representation of the principle of zeta potential measurement of flat material. Upper green rectangle: flat plate; lower yellow rectangle: quartz cell; left blue rectangle: cathode electrode; right orange rectangle: anode electrode. (C) Representative zeta potential measurement data of NHDFCs and hiPSCs. (D) Comparison of the zeta potential on the cell surfaces of NHDFCs and hiPSCs (n = 9). (E–G) Topographical imaging and mapping of charged surfaces using scanning ion conductance microscopy. (E) Optical microscopic image. Scale bar: 30 µm. (F) Surface topographical image. Scale bar: 6 µm. (G) Charge distribution image. The red arrow indicates the no-charge point. Scale bar: 6 µm. a.u., arbitrary unit; hiPSC, human induced pluripotent stem cell; NHDFC, neonatal human dermal fibroblast cell; n.s., not significant; SICM, scanning ion conductance microscopy.

Scanning ion conductance microscopy (SICM) was used to visualize a charge map of the cell surface. NHDFCs and hiPSCs were fixed in this SICM experiment; however, owing to the measurement circumstances, the fixed cells retained a surface charge on the cell membrane owing to the presence of structural proteins and lipids on it (20). Therefore, the SICM probe was placed at the border between the NHDFCs and hiPSCs (Fig. 6E), and a topographical image of the surface structure was obtained from the area bordered by the square in Fig. 6E (Fig. 6F). The charge distribution map shows that most of the cell surface carried a negative charge. This charge distribution did not differ between NHDFCs and hiPSCs (Fig. 6G).

Comparison of heparan sulfate expression between NHDFCs and hiPSCs

The number of cells positive for HCMV IE1/2 and pUL44 proteins was reduced by heparin in a dose-dependent manner, as shown in Fig. S1A and B. Figure S1C demonstrates a significant reduction in infection, as indicated by a dose-response curve represented by the Hill equation. Notably, a higher MOI required higher concentrations of heparin to achieve the IC₅₀ value. Specifically, at a higher MOI of 10, the IC₅₀ value was 5.26 µg/mL, while at a lower MOI of 2.5, the IC₅₀ value was 0.88 µg/mL. In addition, heparan sulfate (HS)-deficient cells (pgsD-677) (confirmed in Fig. S2A and B) showed significant reduction in MCMV and HCMV infection compared with that in Chinese hamster ovary (CHO) cells (CHO-K1) at the attachment level (Fig. S2C through E). Microarray analysis showed that the expression of HS proteoglycan 2 (transcript variants 1 and 2) was 4.43 and 3.07 times lower in hiPSCs compared to NHDFCs, respectively (Data S1). To analyze the differences in HS expression at the protein level, hiPSCs were co-cultured with NHDFCs and fixed with 4% paraformaldehyde (Fig. 7A and D). Subsequently, HS expression was examined using immunocytochemistry. HS expression was significantly lower in hiPSCs than that in NHDFCs (Fig. 7B, C, E, and F). The HS expression in hiPSCs was opposite to that of the pluripotency marker, OCT3/4 (Fig. 7G and H). We evaluated the expression level of HS on the surface of NHDFCs and hiPSCs using flow cytometry. We compared the mean fluorescence intensity (MFI) between the two cell types using the formula [(MFI of HS) − (MFI of IgM isotype)] (Fig. 7I). An average reduction of 49.4-fold difference in HS protein expression was observed on hiPSCs compared to that on NHDFCs. This reduction was significant, as confirmed in four independent experiments (P < 0.01) (Fig. 7J).

Fig 7.

Expression of heparan sulfate in NHDFCs and hiPSCs. (A–H) Co-culture of NHDFCs and hiPSCs. (A and D) Optical microscopic images. (B and E) Images showing the expression of heparan sulfate (red). (C and F) DAPI staining (blue) images. (D–F) Magnified area (square) of panels A–C. Scale bars: 100 µm (A–C) and 39 µm (D–F). (G) Image showing the expression of OCT3/4 (green). (H) Image showing the expression of heparan sulfate (red). Scale bar: 75 µm. (I) Comparison of the flow cytometry analysis results between isotype and HS antibody of NHDFCs and hiPSCs. (J) Comparison of mean fluorescent intensity difference from isotype control between NHDCs and hiPSCs. ★★P < 0.01 (n = 4). hiPSC, human induced pluripotent stem cell; NHDFC, neonatal human dermal fibroblast cell.

DISCUSSION

The susceptibility of hESCs to CMV infection has been explored in previous studies (12, 14). The advent of reprogramming of human somatic cells into hiPSCs has marked a significant milestone in stem cell research because hiPSCs can differentiate into any cell type and are ethically less controversial than hESCs (21). D’Aiuto et al. (10) observed that hiPSCs and hiPS-derived neurons do not allow HCMV infection. Previous studies and our present findings confirm that pluripotent stem cells, including hESCs and hiPSCs, exhibit resistance to CMV. However, the mechanisms underlying the resistance remain unclear.

pp65, a part of the viral tegument and the most abundant protein in virions, plays a major role in extracellular virions (22) and is expressed during the early/late infection phase of protein expression (23). pp65 is transported to the nucleus of permissively infected cells immediately after the fusion of viral and cellular membranes post-entry. It then relocates to the cytoplasm at the later stages of infection (24, 25). Penkert and Kalejta (14) reported that the tegument proteins, pp65 and pp71, entered the cytoplasm of ESCs after infection with a lab-adapted AD169 strain or with the clinical virus isolate FIX, suggesting a potential route for HCMV entry into ESCs to establish latency. HCMV latency was established through the delivery of the viral genome to the nucleus without initiating lytic-phase gene expression. Specifically, the products of the major IE locus, IE1 and IE2, were not expressed. This represents a significant difference from our current findings in the present study. Our previous mouse models using ES and iPSCs showed that the entry of MCMV into pluripotent cells was blocked (8, 9). The findings of Berger et al. (12) are in agreement with our previous and current data regarding the block at the initial stage of viral binding, as demonstrated by (i) the failure of the major viral tegument protein, pp65, to be transported into the nucleus or cytoplasm, reflecting an absence of viral internalization, and (ii) a lack of HCMV DNA accumulation during viral incubation with hESCs. They concluded that platelet-derived growth factor receptor alpha was critical for developmentally acquired HCMV susceptibility.

Microarray analysis results showed that changes in gene expression in hiPSCs following infection were significantly lower than those in NHDFCs; however, these changes were unrelated to the infection process. No gene ontology process terms were indicated in hiPSCs after infection. These results strongly suggest that the infection process, including the suppressive or enhancing mechanisms after infection, is not activated in hiPSCs. The in situ hybridization results showed that HCMV rarely entered hiPSCs when compared to NHDFCs at 1 hpi. This finding strongly supports the lack of pp65 expression in hiPSCs following infection. Some GFP dots were observed in hiPSCs; therefore, it is possible that a small number of HCMV particles entered the hiPSCs. The suppression of the HCMV major IE gene expression, as indicated in previous studies (3–7, 14), might be functioning below the detection level in hiPSCs.

The NanoSuit method and comparison of the relative HCMV genome copy number at the attachment level provide direct evidence that hiPSCs are inherently resistant to HCMV particle adhesion. Viruses rely on electrostatic interactions for optimal virion assembly and attachment (26). These non-specific electrostatic interactions allow viruses to concentrate on the cell surface and transfer to specific receptors (27). The potentials of resting and fixed cells are usually negative (20). Lipids and (glyco)proteins are the main constituents of biological membranes. The sugar moieties of glycoproteins, glycolipids, and attached glycocalyx components, such as hyaluronic acid, can harbor ionizable groups that confer a net negative charge on the outer surface of the plasma membrane. The aggregated surface charge of the outer membrane is indirectly estimated by measuring the zeta potential (the potential at the slipping plane) via electrophoresis (28). The surface charge of the cell differs with the cell type. Cancerous cells are more negatively charged than non-cancerous cells (29). In this study, the average zeta potential of the living cell surface was not significantly different between NHDFCs and hiPSCs. In addition, the charge distributions of the fixed NDHFCs and hiPSCs, measured using SICM, did not differ. Therefore, the surface charge of NHDFCs and hiPSCs did not play a significant role in their differential susceptibility to HCMV.

The interaction between the cell surface and virus entry involves downstream co-receptor interactions, ultimately leading to fusion between the virus envelope and the cell membrane. Several enveloped viruses, including herpesviruses, attach to host cells by initially interacting with cell-surface HS proteoglycans, followed by specific co-receptor engagement, which culminates in virus-host membrane fusion and virus entry (30). HCMV preferentially binds to uniquely sulfated and polymerized HS (a member of sulfated glycosaminoglycans). The attachment of HCMV to cell-surface HS moieties can be inhibited by heparin, a member of the sulfated glycosaminoglycans (31, 32). This was also confirmed in our infection experiments using heparin treatment and heparan sulfate-deficient cells. hiPSCs had a lower HS expression than NHDFCs, both semi-quantitively and quantitatively. Mitra et al. investigated the impact of specific types of sulfation and the degree of polymerization in terms of the number of monosaccharide units (dp) in HS chains on both human and mouse CMV infection and binding (32). Mouse ES cells express relatively low-sulfated HS molecules when compared to that in a variety of murine tissues (33, 34). However, a clear increase in the expression of HS biosynthetic enzymes, which results in increased HS chain sulfation, is observed during ES cell differentiation (34). HS biosynthesis involves stepwise transfer of different sugar residues, leading to the elongation of HS chains (35). The preliminary microarray results indicated the downregulation of exostosin glycosyltransferases (EXT1 and EXT2), N-deacetylase/N-sulfotransferases (NDST1 and NDST2), D-glucuronyl C5-epimerase, and heparan sulfate glucosamine 3-O-sulfotransferase 3B1 in hiPSCs when compared to that in NHDFCs. These factors could explain the low HS expression in hiPSCs. In addition, the chain lengths of HS could be shorter, and the sulfation patterns of HS in hiPSCs could be insufficient for CMV attachment and infectivity.

There are some limitations to this study. The laboratory and clinical strains of HCMV display different cell tropisms, which can be attributed to the presence or absence of genes UL128 through UL151, especially UL128, UL130, and UL131, and their corresponding host receptors (36, 37). We used only the laboratory strain, AD169, as we were unable to obtain the short passage clinical strain of HCMV or a repaired AD169 virus. Successful viral entry includes downstream co-receptor interactions that ultimately lead to the fusion of the viral envelope with the cell membrane (30). The interaction between cell-surface HS and the HCMV envelope represents the primary event in the intricate process of virus entry. The gM/gN complex in HCMV is instrumental in the adherence of the virus to the host cells, potentially by facilitating its interactions with HS proteoglycans present on the cell surface, as indicated by affinity chromatography (38). HCMV gB interacts with HS, leading to virus attachment (31). Applying a soluble version of gB hinders the entry of HCMV into cells (39). The binding and infectivity of HCMV are reduced by soluble forms of heparin and HS. This reduction is observed in cells processed with heparinases or those that cannot generate HS (40). Berger et al. (12), using both the laboratory strain (AD169) and the clinical HCMV strains, suggested that resistance arises due to a barrier at the initial stage of viral binding. The pentamer-negative AD169 virus, a clinical strain, and the pentamer-repaired AD169 virus have gM/gN and gB in common; these have the capacity to bind HS for the attachment. Therefore, employing only AD169 for elucidating the phenomena observed in this study is apparently sufficient. However, there are no reports suggesting that the absence of gM and gN in virions impacts entry or that their absence affects entry due to missing interactions with HS. This may be due to the essential nature of both gM and gN for replication of HCMV. Deletions in the C-terminal cytoplasmic tail of gM (41) or the carboxy-terminal domain of gN (42) result in replication-deficient viruses, which may complicate attachment and entry experiments using gM/gN mutant HCMV. Isaacson and Compton (43) reported that HCMV gB is not absolutely required for virus attachment or assembly and egress from infected cells using gB-null HCMV virions. However, heparin inhibits gB-null HCMV at the attachment level. This suggests that the gM/gN complex or another unknown protein may compensate for gB in binding to HS. Further investigation is required in this area. Additionally, inhibitory effects of heparin and heparinase not only on the laboratory strain AD169 (44) but also on the TB/40E strain of HCMV, which possesses a relatively intact genome (45, 46), have been reported. Although gM/gN or gB is a strong candidate for binding to HS for attachment, other targets may exist in the clinical strain of HCMV. Incorporating a clinical strain of HCMV in future experiments would further reinforce these findings. Until then, the assumption that strain AD169 possesses all mechanisms used by clinical strains of CMV to enter iPSCs remains speculative. While we do not claim to have elucidated every resistance mechanism of hiPSCs against HCMV, our data strongly indicate that reduced HS expression in hiPSCs contributes to primary resistance at the attachment level.

Furthermore, we used an in vitro experimental model to observe differences in HCMV susceptibility and believe that the level of HS expression could influence HCMV infectivity in vivo. Therefore, the relationship between them should be investigated in future studies using three-dimensional organelle culture or clinical congenital autopsies from CMV infection cases.

In conclusion, our results clearly demonstrated that HCMV barely attaches to the surface of hiPSCs due to low HS expression. Among various factors to consider, this reduced HS expression stands out as the primary determinant of HCMV resistance in hiPSCs. We reaffirmed the significance of the CMV infection mechanism involving HS during embryonic development. In the future, drugs targeting HS could become crucial for the treatment of congenital CMV infections. Therefore, further research in this area is warranted.

MATERIALS AND METHODS

Generation of hiPSCs and cell culture

NHDFCs were cultured in Dulbecco’s modified Eagle medium (DMEM), supplemented with 10% (vol/vol) fetal bovine serum (FBS). They were transduced overnight with Sendai viruses containing human Oct3/4, Sox2, Klf4, and c-Myc (CytoTune-iPS Sendai Reprogramming Kit; Thermo Fisher Scientific, Waltham, MA, USA) at an MOI of 15. The medium was changed daily for the next 2 days, and the cells were resuspended in expansion DMEM supplemented with 10% (vol/vol) FBS. On day 3, the cells were recovered via centrifugation (650 × g for 5 min) and plated on mitomycin-C-treated mouse embryonic fibroblasts (feeder cells) in an expansion medium for two additional days. On day 6, half of the medium was changed to human pluripotent stem cell medium [DMEM/F12 supplemented with 20% Knock Out Serum Replacer, 1-mM L-glutamine, 1% (vol/vol) penicillin/streptomycin, and 100-µM 2-mercaptoethanol (all from Life Technologies, Waltham, MA, USA) or 20-ng/mL basic fibroblast growth factor (ReproCELL Inc. Kanagawa, Japan)]. For feeder-free culture, iMatrix-511 (0.5 ng/mL) was coated on the plate at 37°C for 1 h. ReproFF2 (ReproCELL Inc.) culture with 20-ng/mL basic fibroblast growth factor was used for maintaining hiPSC proliferation.

Virus

The HCMV strain AD169, provided by the Department of Regenerative and Infectious pathology, Hamamatsu University School of Medicine, was cultivated in human embryonic lung fibroblasts using an MOI of 0.01. This process was continued until cytopathic effects were noticeable. To generate a viral preparation, the cells were subjected to sonication, and the solution was clarified via centrifugation at 2,000 × g. The concentration of the virus was then determined using a conventional plaque assay performed on human embryonic lung fibroblasts. Recombinant viruses, derived from the Smith strain of wild-type MCMV (gene accession number U68299) capable of expressing the EGFP (Clontech, Palo Alto, CA, USA), were used in this study. This recombinant virus was engineered to express an EGFP gene insert under the control of an EF-1α/HTLV composite promoter, combining the EF-1α core promoter with the 5′-untranslated region of the human T-cell leukemia virus (8, 47).

Microarray analysis

Total RNA was extracted from each sample of either two independently prepared batches of 50,000 control (NHDFCs and hiPSCs) and HCMV-infected (NHDFCs and hiPSCs) cells 30 min post-infection at an MOI of 10, using the RNeasy Mini Kit (QIAGEN, Tokyo, Japan) and diluted with RNase-free water. Additionally, RNA was amplified using the Ovation Pico WTA System V2 (NuGEN Technologies, San Carlos, CA, USA). Gene expression profiling was commissioned to Medical & Biological Laboratories Co. Ltd. (SurePrint G3 Human Gene Expression 8 × 60K v.3 Microarray kit; Agilent Technologies, Santa Clara, CA, USA). Distribution was conducted through an agency (MBL Co., Ltd., Tokyo, Japan). RNA-Seq data were analyzed using the Subio platform software (Subio Inc.). Global normalization was performed at the 75th percentile, with a low signal cut-off value of 30. If the normalizing signal was below 30, it was replaced with a log₂ ratio for each control sample. To define the set for analysis, probes were removed if 50% of the Agilent’s flag glsWellAbove BGs were 0 and the log₂ ratios of all samples were between −0.5 and 0.5. Thereafter, hierarchical clustering was performed on the set of genes. For hiPSCs, differentially expressed genes were extracted with criteria of ×1.8 up or down and Student’s t-test with a P value of <0.1. For NHDFCs, with criteria of ×2.0 up or down. The up- and downregulated genes were compared using a Venn diagram. Next, an enrichment analysis on gene ontology biological processes was performed using DAVID Bioinformatics Resources (https://david.ncifcrf.gov).

Attachment assay and quantitative reverse transcription PCR

HCMV was allowed to attach at 4°C for 1 h, followed by three washes with 1-mL PBS. The cells were then lysed in PBS, and DNA was extracted and analyzed using quantitative PCR to detect viral genomes. The PCR primers used were from the HCMV IE1 gene: upstream, 5ʹ-GACTAGTGTGATGCTGGCCAAG-3ʹ, and downstream, 5ʹ- GCTACAATAGCCTCTTCCTCATCTG-3ʹ. A fluorogenic probe (5ʹ-carboxyfluorescein-AGCCTGAGGTTATCAGTGTAATGAAGCGCC-3ʹ) was inserted between the PCR primers. The RNaseP gene was chosen for normalization using the TaqMan RNase P Detection Reagents Kit (Thermo Fisher Scientific).

Scanning electron microscopy of virus particles on the cell surface using the NanoSuit method

A NanoSuit type III solution (Nisshin EM Co. Ltd., Tokyo, Japan) was applied to cultured cells on a cover glass, covering the entire culture and allowed to stand for 1 min. The sections were then spin-coated to remove excess solution (650 × g, 30 s) and air-dried. Next, the specimens were directly introduced into the scanning electron microscopy, where a NanoSuit was formed following irradiation with an electron beam. FE-SEM was performed using a Hitachi S-4800 instrument (Hitachi High-Technologies, Tokyo, Japan) operated at an acceleration voltage of 1.0 kV. The vacuum level of the observation chamber was 10⁻³ to 10⁻⁷ Pa. The secondary electrons were detected using the signal from the lower detectors. Other details were as follows: working distance = 8 mm; aperture size, ϕ = 100 µm; scan speed = 10–15 frames/s each beam.

Immunocytochemistry

NHDFCs and hiPSCs were fixed with 4% (vol/vol) paraformaldehyde (PFA) in PBS for 20 min at room temperature (20°C–25°C). The hiPSCs on feeder cells in the dishes were stained using the Human Pluripotent Stem Cell Marker Antibody Panel (anti-h/m/r Alkaline Phosphatase Purified Mouse Monoclonal IgG1 Clone B4-78; Anti-hNanog Affinity Purified Goat IgG; anti-hOct 3/4 Affinity Purified Goat IgG; anti-h/mSSEA-1 Purified Mouse Monoclonal IgM Clone MC-480; and anti-h/mSSEA-4 Purified Mouse Monoclonal IgG3 Clone MC-813–70; R&D Systems Inc., Minneapolis, MN, USA). Immunocytochemical staining was performed according to manufacturer’s instructions. The infected cells were fixed and stained with mouse monoclonal antibodies against HCMV IE1/2 + pUL44 antigens (Clones CCH2/DDG9, Agilent Technologies) and mouse monoclonal pp65 (Clones 2A6; Leica Biosystems, Deer Park, IL, USA). NHDFCs and hiPSCs were stained with mouse IgM antibody against HS (F58-10E4 clone; Amsbio, Abingdon, UK). The secondary antibody was Alexa Fluor 488 goat anti-rabbit, anti-rat, or anti-mouse IgG (H + L), Alexa Fluor 546 goat anti-mouse IgM (Molecular Probes, Invitrogen, CA, USA), incubated for 30 min at 20°C–25°C. DAPI (Dojindo, Kumamoto, Japan) was used to stain the nuclei.

In situ hybridization

The probe for DNA in situ hybridization was prepared from the HCMV DNA genome (DY-380) using a bacterial artificial chromosome system via nick translation, as previously described (8). In situ hybridization of HCMV DNA was performed, as previously described (8, 48) for the paraffin sections from fixed pellet cells. The cells were treated with RNase (100 µg/mL in PBS; Boehringer Ingelheim, Biberach, Germany) for the detection of viral DNA.

Zeta potential of the surface of cells

The zeta potential is the electric potential at the shear plane of a particle or cell surface. This potential can be indirectly measured on a flat surface or sheet-shaped, cell-coated glass using the Mori and Okamoto’s equation (https://www.otsukael.com/weblearn/chapter/learnid/3/page/2), which utilizes the electroosmotic flow in the liquid generated by the potential on the sample surface. hiPSCs and NHDFCs were grown on glass until they reached confluence. The surface charges of hiPSCs and NHDFCs were determined using a zeta potential and particle size analyzer, ELSZ-1000 (Otsuka Electronics Co., Ltd., Osaka, Japan). After soaking the specimens in 10-mM NaCl, into which the monitoring particles were injected, the zeta potential of the surface was measured using a flat-surface zeta cell unit (ELSZ, Otsuka Electronics Co., Ltd.). The electrophoretic velocity of the solid surface was calculated using the obtained electric osmosis profile and the Mori and Okamoto formula. The equation used was as follows:

$${\displaystyle U_{obs}(Z)=A{U}_0(Z/b{)}^2+\mathrm{\Delta }{U}_0(Z/b)+(1-A)U_0+U_p,}$$

where U_obs(Z) is the apparent velocity measured at position Z;

Z is the distance from the cell center;

A = 1 / [(2/3) − (0.420166 / k)];

K = a/b is the ratio of the side lengths a and b of the cell cross section (a > b);

U_p is the true particle mobility;

U₀ is the average solvent flow velocity on the upper and lower surfaces of the cell;

ΔU₀ is the difference between the flow velocities of the solvent on the upper and lower surfaces of the cell.

Topographical imaging and mapping of charged surfaces using SICM with a theta nanopipette

SICM was used to visualize the charge map of the cell surface. SICM is used for imaging samples in liquids, at the nanometer scale (49, 50). Imaging the sample surface is challenging due to the influence of the charge of the sample. However, Shirasawa et al. developed a unique SICM technology that enables the simultaneous measurement of a topographical image and cell-surface charge (51). The SICM unit was mounted onto the sample stage of an inverted microscope. The nanopipette was moved using the x-, y-, and z-axis stages driven by piezo motors for coarse positioning. An x- and y-axis flat piezoscanner and a z-axis piezoelectric element were used for fine positioning and scanning. The ion current was detected using a current to voltage converter, and the output signal was fed to a personal computer. To obtain a topographical image, two electrodes were inserted into each channel of the theta nanopipette. However, to obtain a charge map, the bath electrode was inserted in the same way into another nanopipette placed in the bath solution. By switching the setup, topographical imaging and charge mapping were performed.

Flow cytometry

To detect cell-surface expression of HS in NHDFCs and hiPSCs, wild-type (CHO-K1) and HS-deficient mutant CHO-K1 (pgsD-677) cells were washed with PBS(−) thrice and were dissociated with non-enzymatic cell dissociation solution (Sartorius, Tokyo, Japan) for 5 min at 37°C. The cells were dispersed with each culture medium, centrifuged (500 × g) for 5 min at 4°C and fixed with 4% PFA for 5 min at 20°C–25°C. The cells were washed with 2% fetal calf serum (FCS)/PBS thrice with centrifugation (500 × g) for 5 min at 4°C. Human Fc receptor blocker (20 µL) (Clear Back; MBL Co., Ltd.) was added to 10⁵ cells, and the cells were incubated for 5 min at 20°C–25°C. The cells were then incubated with mouse IgM (isotype control; MBL Co., Ltd.) (10 µg/mL) and mouse IgM antibody against HS (F58-10E4 clone, Amsbio) (10 µg/mL) for 30 min at 4°C, respectively. The cells were washed with 2% FCS/PBS once with centrifugation (500 × g) for 5 min at 4°C and then incubated with Alexa Fluor 546 goat anti-mouse IgM (Molecular Probes, Invitrogen) for 30 min at 4°C. The cells were washed with 2% FCS/PBS once with centrifugation (500 × g) for 5 min at 4°C. The cells were dispersed with 500-µL 2% FCS/PBS; 10⁴ cells were analyzed with GAllios Flow Cytometry (Beckman Coulter, Brea, CA, USA). The difference in the MFI of HS [(MFI of HS) – (MFI of IgM isotype)] was compared between NHDFCs and hiPSCs.

HCMV infection of NHDFCs with heparin treatment

NHDFCs were seeded in 48-well tissue culture plates and cultured until confluent. These cells were pre-treated for 1 h with varying concentrations of heparin (0, 0.5, 1.0, 2.5, 5.0, 10.0, 25.0, and 100.0 µg/mL; Sigma-Aldrich Co., MO, USA). Following this, NHDFCs were infected with HCMV (strain AD169) at an MOI of 2.5 or 10, mock-infected, either in the presence or absence of heparin. The cultures were maintained for an additional 12 hpi. Subsequently, the cells were fixed using 4% PFA and subjected to immunostaining with an antibody specific to HCMV IE1/2 + pUL44 antigens (Clones CCH2/DDG9, Agilent Technologies) for 1 h at 20°C–25°C. This was followed by incubation with Alexa Fluor 488-conjugated goat anti-mouse IgG (H + L) as the secondary antibody (Molecular Probes, Invitrogen), for 1 h at 20°C–25°C. Nuclei were stained with DAPI (Dojindo). Data were obtained from three independent experiments. A dose-response curve, modeled using the Hill equation, is depicted to illustrate the relative infection rate in comparison to HCMV infection without heparin treatment. We used Dr. Fit software (available at http://sourceforge.net/projects/drfit/) to fit the Hill equation to the dose-response curve, aiding in the accurate determination of the IC₅₀.

CMV infection on heparan sulfate-deficient cells

CHO-K1 and pgsD-677 were purchased from the American Type Culture Collection (Manassas, VA, USA). Cells were maintained in Ham’s F-12K medium supplemented with 10% FBS, 100 µg/mL of streptomycin, and 100 IU/mL of penicillin. CHO-K1 and pgsD-677 were seeded in 48-well tissue culture plates and cultured until confluent. HCMV or MCMV was allowed to attach at 4°C for 1 h, followed by two washes with 1-mL PBS. The cultures were then sustained for an additional 12 or 24 hpi. After this period, the cells were fixed with 4% PFA and analyzed for GFP expression in the case of MCMV, or subjected to immunostaining for HCMV using an antibody specific to the HCMV IE1/2 + pUL44 antigens (Clones CCH2/DDG9, Agilent Technologies). These results were replicated in four separate experiments.

Statistical analysis

Statistical analysis was performed using a two-sided unpaired Student’s t-test for comparisons. All data were compared using a paired t-test. P values less than 0.05 or 0.01 were considered statistically significant.

DATA AVAILABILITY

The data supporting the findings of this study are derived from microarray analyses. These analyses were performed on total RNA extracted from two independently prepared batches of 50,000 cells each, encompassing both control groups [neonatal human dermal fibroblast cells (NHDFCs) and human induced pluripotent stem cells (hiPSCs)] and human cytomegalovirus-infected groups (NHDFCs and hiPSCs), collected 30 min post-infection at an multiplicity of infection of 10. The resulting data sets, characterized by RNA sequencing, have been deposited in the Gene Expression Omnibus (GEO) repository. These data sets are publicly accessible and can be retrieved using the GEO accession number GSE241636.

ETHICS APPROVAL

Generation of human induced pluripotent stem cells from neonatal human dermal fibroblast cells (derived from fetal skin fibroblasts and purchased from Lonza, Basel, Switzerland) was approved by the ethics committee of Hamamatsu University School of Medicine (approval number 12–186).

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/jvi.01278-23.

Data S1. jvi.01278-23-s0001.xlsx.

Comparison of mRNA expression between hiPSCs and NHDFCs (both cell types are not HCMV infected).

Data S2. jvi.01278-23-s0002.xlsx.

Changes of mRNA expression in hiPSCs and NHDFCs after HCMV infection.

Supplemental figures. jvi.01278-23-s0003.docx.

Fig. S1 and S2.

Movie S1. jvi.01278-23-s0004.mp4.

Cardiac cell differentiation from hiPSCs.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.