RNA-Seq time-course analysis of neural precursor cell transcriptome in response to herpes simplex Virus-1 infection

Our previous work described the impairment in proliferation, migration, and differentiation of NPCs infected with HSV-1 using a 3D model of NPCs (neurospheres) (Zheng et al. 2020a, 2022). To investigate the changes in transcription in human NPCs following HSV-1 infection, we conducted a time-course RNA-Seq study using NPCs-derived neurospheres. The method for generating NPCs from human iPSCs is detailed in the methods section. NPCs were infected with a genetically engineered HSV-1 KOS strain that expresses enhanced green fluorescent protein (EGFP) and red fluorescent protein (RFP) under the control of immediate early and late gene promoters, respectively (Zheng et al. 2020a). Cells were infected at the multiplicity of infections (MOIs) of 0.001 and 0.0001 in the presence or absence of antivirals 5BVdU + IFN-α. After one hour, the inocula were removed, cells were washed and dissociated, and were transferred into low attachment 6-well plate for the generation of neurospheres. Uninfected and infected neurospheres were harvested at days 3, 5, and 7 post-infection (p.i.) (Fig. 1). The choice of the MOIs relies on the evidence that antivirals 5BVdU + IFN-α (which demonstrate greater efficacy when compared to acyclovir) cannot effectively inhibit viral replication at MOIs higher than 0.001 across 7 days following infection (Zheng et al. 2020a). The expression of the fluorescent reporter genes at the different time points confirmed the infection of NPCs (data not shown).

RNA-Seq mapping results revealed a range of nucleotide coverage, from 24.6 million reads (day 7, MOI 0.0001 condition) to 51.2 million reads (day 7, uninfected condition in the presence of antivirals), as shown in Fig. 2a. Mapped reads ranged from 12.6 million reads to 50.2 million reads (mean 34.9, sd 8.4). Following human transcript mapping, the remaining unmapped reads were collected and mapped to the Human Herpesvirus strain KOS genome (GenBank: JQ780693.1) in a similar manner. Our analysis revealed a strong impact of antivirals (Avs) 5BVdU + IFN-α on HSV-1 infection during days 3 to 7 post-infection, with the most significant results observed on day 7, where infected NPCs exposed to the antiviral had a minimal number of viral transcripts (Figs. 2b and 3). An increase in the viral genes was detected on day 5 post-infection, suggesting a temporary surge in virus production. A total of 12,484 human genes were expressed at mean transcripts per million (TPM) > = 5 in uninfected neurospheres untreated with Avs in at least 1 condition, and 12,610 in uninfected neurospheres treated with Avs in at least 1 condition. Details of the RNA-Seq experiments can be found in the methods section.

In infected neurospheres, the number of differentially expressed host genes ranged from 37 − 3,647 at the three time points in the different conditions when compared to uninfected cells, defined as FDR corrected significance at p < = 0.05 based on edgeR analysis and max group mean of > = 5 TPM in expression (Fig. 4). At one or more of the three time points, a total of 4,547 genes demonstrated differential expression. Comparisons within days 5 and 7 in the absence of Avs were not considered reliable because several conditions on those days exhibited unacceptable RNA degradation based on QC (Fig. 2c).

The differentially expressed genes (DEGs) were categorized into up-regulated and down-regulated groups, with 2,652 unique genes being exclusively down-regulated across conditions, 1,601 genes being exclusively up-regulated, and 294 exhibiting a mixed profile (Fig. 4). Several genes were only differentially expressed at one of the three sampling times. Out of these, a greater number of genes (933) were differentially expressed at 3 dpi compared to 5 or 7 dpi in the presence of antivirals.

We employed Ingenuity Pathway Analysis (IPA) to determine the top ten hierarchically clustered, significantly dysregulated canonical pathways that were significantly altered in response to HSV-1 infection in the presence or absence of antivirals (Fig. 5). Statistical significance was calculated using right-tailed Fisher Exact Probability Tests; biological pathways showing p-value < 0.05 were considered statistically significant. The fifteen most significantly altered pathways in neurospheres infected in the presence or absence of antiviral are shown in Fig. 5a. According to IPA’s canonical pathway analysis of these commonly DEGs, oxidative phosphorylation, EIF2, apoptosis, assembly of RNA polymerase, and RHOGDI were generally found to be upregulated (Fig. 5). Conversely, CREB signaling, IL-15 production, WNT/Ca + pathway, neuroinflammation signaling, heparan sulfate, dermatan sulfate, and cholesterol biosynthesis were generally downregulated (Fig. 5).

The apoptosis signaling pathways exhibited upregulation across all time points at both multiplicities of infection (Fig. 5a). Autophagy exhibited upregulation at day 3 post-infection. However, a downregulation of autophagy was observed at days 5 and 7 post-infection (Fig. 5b). The PI3K/AKT signaling cascade exerts a suppressive effect on the lytic cycle of viral replication by maintaining a repressive chromatin state of the viral genome via the activation of eIF4E-binding protein (4E-BP). A key component of this pathway is the mammalian target of rapamycin (mTOR) (Kobayashi et al. 2012). Notably, in acutely infected neurospheres, it was observed that both the PI3K/AKT pathway and mTOR are downregulated (Fig. 5b).

Based on IPA analysis, we next aimed to identify the genes and/or pathways that impair neuroectodermal formation and the different aspects of NPC neurogenesis. The inhibition of TGF-b and BMP signaling pathways have been shown to induce neuroectodermal formation. In the context of HSV-1 infection of neurospheres, it has been observed that these signaling pathways are downregulated. Consequently, it could indicate that HSV-1 infection may, in principle, influence neuroectoderm formation. However, it is important to note that the inhibition of BMP signaling has been reported to initiate neural induction through the activation of fibroblast growth factor (FGF) signaling and ZIC genes (Marchal et al. 2009). Remarkably, our results show a downregulation of the FGF signaling in neurospheres infected at both MOIs. These findings suggest that HSV-1 could potentially influence neuroectodermal formation.

Our previous findings have shown the negative impact of HSV-1 on proliferation, self-renewal, and migration abilities of NPCs (Zheng et al. 2022). To gain insights into the underlying mechanisms, we analyzed the RNA sequencing data of infected neurospheres at the different time points and MOIs to identify potential candidate genes responsible for these impairments. Our analysis indicated that a number of crucial pathways are downregulated in infected cells. These include: the WNT/β-catenin pathway, which is known to be active in the ventricular zone during cortical development and plays a crucial role in regulating the proliferation of NPCs (Kuwahara et al. 2014) (Fig. 5b); the Notch signaling pathway, which is essential for the maintenance of neural stem cells (NSCs) by enhancing the NSC self-renewal and by inhibiting differentiation (Wang et al. 2009; Hitoshi et al. 2002; Lasky and Wu 2005); ephrin-B signaling, which seems to be important for regulating self-renewal (Qiu et al. 2008); the sonic hedgehog signaling pathway, which coordinates the proliferation and differentiation of neural stem/progenitor cells through its regulation of the cell cycle kinetics of radial glial cells and intermediate progenitor cells (Komada 2012). In addition, we compiled gene ontology (GO) gene sets related to NPC proliferation, migration and differentiation, and assessed dysregulation patterns in these genes. We observed that antivirals did not normalize upregulation of several genes involved in these processes. This suggests that the antiviral treatment had a more targeted or specific impact on the regulation of these groups of genes (Figs. 6 and 7).

A number of key genes involved in aspects of neurogenesis and maintenance were also downregulated: NR2F1 plays a role in controlling the long-term self-renewal of neural progenitor cells by regulating cell cycle genes and key master genes involved in cortical development, such as PAX6 (Fig. 6a-d) (Bertacchi et al. 2020); STAT3, which regulates the maintenance (Yoshimatsu et al. 2006), proliferation and differentiation of NPCs (Su et al. 2020); the high-mobility group DNA binding gene SOX2 (Fig. 6a and c-d), which is an important factor for the maintenance of neural stem cells in adult neurogenic areas, and its regulatory mutations cause neurodegeneration and affects adult neurogenesis (Ferri et al. 2004); epidermal growth factor (EGF) and fibroblast growth factor (FGF) signaling pathway genes, which are crucial in regulating the proliferation of neural stem cells (Fig. 5b) (Vescovi et al. 1993; Gritti et al. 1996); HES5, involved in the maintenance of neural stem cells (Fig. 6a and d) (Ohtsuka et al. 2001).

Genes regulating the migration of early-born neurons, such as ROBO1 (Gonda et al. 2013) ROBO2 (Guerrero-Cazares et al. 2017), SLIT1 (Deboux et al. 2020), and ASTN1 (Wilson et al. 2010) were downregulated (Fig. 6d). Slit-Robo signaling has also been found to play a role in the regulation of the NPCs proliferation as well as cell branching and elongation (Borrell et al. 2012; Andrews et al. 2008). Conversely, Reelin, which regulates NPCs migration (Courtès et al. 2011), was found upregulated by HSV-1 infection. GATA2, which activates genes specific to the GABAergic neuron subtype in the midbrain (Lahti et al. 2013), and PAX2, which is involved in the specification of interneurons in the dorsal horn (Batista and Lewis 2008), are upregulated at both MOIs, but are both downregulated by day 7 (Fig. 6d). RASD1, proposed to exert an antiviral activity (Wyler et al. 2019) was found significantly upregulated on day 3 p.i. at both MOIs.

MSX1, whose overexpression in mouse neurospheres promotes oligodendrogenesis (Roybon et al. 2008), is overexpressed at day 3 post-infection (p.i.) at both MOIs (Fig. 6d). HES5, which inhibits both astrocyte and oligodendrocyte differentiation (Wu et al. 2003), is downregulated at both MOIs but upregulated at day 7 p.i. (Figure 6a and d). MYT1, which is involved in proliferation and differentiation of oligodendrocytes precursor cells (OPCs) (Wu et al. 2003), is upregulated at day 3 (Fig. 6d). OLIG2, which regulates key aspects of the oligodendrocytes development (Zhang et al. 2022), is upregulated at MOI 0.001 and downregulated at MOI 0.0001 (Fig. 6d). ID4, which inhibits OPC differentiation, is downregulated at both MOIs (Fig. 6d). Taken together, these results indicate that HSV-1 infection may favor oligodendrocyte differentiation from NPCs. However, ID2, which inhibits OPC differentiation (Emery and Lu 2015), is upregulated at day 3 at MOI 0.0001. Furthermore, DLX2, a modulator of neurons versus oligodendrocytes development in the ventral embryonic forebrain (Petryniak et al. 2007), is upregulated at day 3 at both MOIs. Overall, the net effect of the opposing gene expressions on oligodendrocyte differentiation of HSV-1-infected NPCs is uncertain. In HSV-1-infected NPCs, various genes that regulate neuronal differentiation and maturation were found to be dysregulated (Fig. 6). These include NLGN1, which induces neurite outgrowth (Fig. 6d) (Gjørlund et al. 2012), PAX6 (Fig. 6a-d) (Sansom et al. 2009), and TIMP2 (Fig. 6d) (Pérez-Martínez and Jaworski 2005), which promotes neuronal differentiation. HSV-1 affects glutamate ionotropic receptors AMPA type GRIA1 and GRIA3, glutamate ionotropic receptors delta type GRID1 and GRID2, and glutamate ionotropic receptors kainate type GRIK1, GRIK2, GRIK3, and GRIK4 (Fig. 8). NTRK2 (Zagrebelsky et al. 2020), KIF1A (Fan and Lai 2022), and PDLIM5 (Herrick et al. 2010) genes regulate dendritic spine formation, and were downregulated in all the conditions (Fig. 6).

Consistent with previous findings, cholesterol biosynthesis was significantly downregulated in HSV-1-infected neurospheres (Blanc et al. 2011; Wang et al. 2020; Sviridov and Bukrinsky 2014; Prabhu et al. 2016; Dai et al. 2022; Chen et al. 2023; Takano et al. 2011; Ma et al. 2022; Huang et al. 2023; Wudiri and Nicola 2017; Cagno et al. 2017). In fact, on day 3 p.i. at MOI 0.001 in the absence of antivirals IPA analysis showed that the superpathway of cholesterol biosynthesis (p = 1.90E-0; z-score= -3.606), as well as the cholesterol biosynthesis I, II, and III pathways (p = 3.33E-08; z-score =-3), were most significantly downregulated. Similarly, at the same time point at MOI 0.0001, IPA analyses also indicated a significant downregulation of cholesterol biosynthesis I, II, and III pathways (p = 3.41E-08; z-score =-3) (Fig. 9). This downregulation of the sterol metabolic network was efficiently inhibited when infected cells are treated with 5BVdU + IFN-α. These results uncover a dependency role for the suppression of the sterol metabolic network in HSV-1-infected neural progenitor cells. Our findings indicate that HSV-1 infection leads to a significant reduction in the expression of genes involved in the sterol metabolic pathway, including those involved in cholesterol synthesis and transport. These findings provide new insights into the complex interactions between viruses and host cell metabolism in human CNS and suggests that targeting this metabolic pathway may represent a potential strategy for the development of antiviral therapies against HSV-1 infection.

Together, these findings shed light on how HSV-1 infection interferes with the complex regulation of neural progenitor cells that may cause some of the extensive neuropathogenesis in HSV encephalitis patients.

A KOS-based recombinant virus in which enhanced green fluorescent protein (EGFP) and monomeric red fluorescent protein (RFP) are reporters whose expression is driven by the viral promoters ICP0 and Glycoprotein C, respectively (HSV-1 DualFP) (Zheng et al. 2020a) was employed in this study. The virus stock was prepared in the D’Aiuto laboratory at the University of Pittsburgh. 80–90% confluent monolayers of Vero cells were infected at a multiplicity of infection (MOI of 3 in DMEM medium supplemented with 2% FBS). After 2 h the inoculum was removed, cells were washed and cultured for 2–3 days, until the appearance of full cytopathic effect (CPE). The cells were scraped and transferred along with the culture supernatant into 15 ml conical tubes. Cells were centrifuged at 1000 rpm for 5 min. The culture supernatant was removed, leaving behind 1.5 ml, and the cell pellet was resuspended using a vortex for 1–2 min. Cells were freeze-thawed three times. Debris was then removed by centrifuging at 3000 rpm for 5 min and the top culture supernatant containing cell-free viral particles was stored at − 80 °C until use. Virus titers were determined by the standard plaque assay as described below.

Human-induced pluripotent stem cells (hiPSCs) were cultured in mTesR™ plus on Matrigel-coated tissue culture-treated plates (STEMCELL Technologies). The hiPSCs were established at the National Institute of Mental Health (NIMH) Center for Collaborative Studies of Mental Disorders-funded Rutgers University Cell and DNA Repository (RUCDR) (http://​www.​rucdr.​org/​mental-health). The control steps included the analysis of pluripotency markers NANOG, Oct4, TRA60, TRA811, OSX2 and SSEA4. We subsequently conducted karyotyping, array comparative genomic hybridization (aCGH) assays and short tandem repeat (STR) profiling and compared them with donor genomic DNA to evaluate structural changes in genomic DNA during the generation of hiPSC lines.

Human NPCs were derived from hiPSC line 73–56010-02 as previously described (Zheng et al. 2022). Briefly, hiPSCs were cultured in mTeSR1-plus medium supplemented with dual SMAD inhibitors SB 431542 and LDN 193189 to promote neural induction. After 8–10 days, neural rosettes were manually isolated, transferred into Matrigel coated plates and cultured in StemDiff Neural Progenitor Medium (STEMCELL Technologies) for the expansion of NPCs. The expression of the NPCs markers SOX1 and PAX6 was analyzed (Fig. 1). All cells were cultured in standard conditions (37 °C, 5% CO2, and 100% humidity).

Neural progenitor cells (NPCs) were seeded onto 12-well matrigel-coated plates and cultured in STEMdiff™ Neural Progenitor (NP) medium until they were 80% confluent. On the day of infection, cells were infected with HSV-1 DualFP at MOI 0.001 and 0.0001 with or without the presence of antivirals (E)-5-(2-bromovinyl)-2′-deoxyuridine (5BVdU; 30 µM) and alpha interferon (IFN-α; 125 U/ml) (N = 3). The media for uninfected treated control wells were switched to StemDiff™ NP medium supplemented with 5BVdU + IFNalpha at the same time. One hour later, the infectious inocula were removed and culture wells were gently rinsed once with PBS. Corresponding media were added afterwards, and cells were manually dissociated and transferred into low-attachment 6-well plates. For each well one million cells were seeded. The conditions were: (i) MOI 0.001 treated with 5BVdU + IFN-α; (ii) MOI 0.0001 treated with 5BVdU + IFN-α; (iii) uninfected but treated with 5BVdU + IFN-α; (iv) MOI 0.001 untreated; (v) MOI 0.0001 untreated; (vi) uninfected and untreated (N = 3). Low-attachment plates were left on an orbital shaker in the incubator to form homogenous neurospheres. In total there were three sets containing all the conditions described above, they were harvested on Day 3 post infection, Day 5 post infection and Day 7 post infection, respectively. For each replicate well of each condition, we collected all spheres (or degenerating pieces for those infected but untreated on Day 7) along with all the media from the culture well. They were centrifuged at 10,000 rpm for 2 min and the supernatants were transferred, and the pellets kept at -80˚C. Pellets were dissociated and lysed with 200µL Buffer RLT plus provided in Qiagen RNeasy plus mini kit. Samples were kept at -80˚C until further RNA extraction based on the manufacturer’s instructions.

Total RNA libraries were generated using the Illumina TruSeq Stranded Total RNA Sample Preparation Guide, Revision E. The first step involved the removal of ribosomal and mitochondrial RNA using biotinylated, target-specific oligomers combined with Ribo-Zero rRNA removal beads. Following purification, remaining RNA was fragmented using divalent cations under elevated temperature, which were then copied into first strand cDNA using reverse transcriptase and random primers, followed by second strand cDNA synthesis using DNA Polymerase I and RNase H. Subsequently, a single adenosine base was added to each of the cDNA fragments, followed by ligation of an adapter. The products were purified and enriched with PCR to create the final cDNA library. A total of 12 cDNAs (two MOIs, two treatments ×three sample times) were generated. The cDNA libraries were validated using KAPA Biosystems primer premix kit with Illumina-compatible DNA primers and Qubit 2.0 fluorimeter. Quality was examined using an Agilent Bioanalyzer Tapestation 2200. The cDNA libraries were pooled at a final concentration of 1.8pM. Cluster generation and 100 bp paired-read dual-indexed sequencing was performed on Illumina NExtseq 500 (Children’s hospital of Pittsburgh, University of Pittsburgh). Sequencing read quality was assessed using fastQC v0.11.4 and CLCbio v11.0.1 software. The average number of reads per sample was 39.5 million (SD = 4.8 million reads) (Fig. 2).

Sequences were trimmed based on quality score using the modified-Mott trimming algorithm as implemented in CLC bio software, using a trim cutoff error probability of 0.05. Ambiguous bases were trimmed using a post trim maximal ambiguous base cutoff of 2. The trimmed reads were then mapped to the human genome GRCh38/hg38, using sequence and annotation provided by Ensembl (release 82). Approximately 92% of reads were mapped in pairs (SD = 1.14) across all samples, and 97.7% of reads were mapped in total (SD = 0.45).

Following human mapping, the remaining unmapped reads were collected and mapped to the Human Herpesvirus strain KOS genome (GenBank: JQ780693.1) in a similar manner.

Functional analysis of differentially expressed genes (DEG) was performed using Qiagen’s Ingenuity Pathway Analysis (IPA, Qiagen Bioinformatics, https://​www.​qiagenbioinforma​tics.​com/​products/​ingenuity-pathway-analysis/​). IPA provides tools to interpret DEG datasets in the context of biological pathways⁴¹. Canonical pathway analysis identified biological pathways from the IPA library of canonical pathways that were most significant in relation to the R430-treated DEG data set. The significance of the association was measured by (1) a ratio (the number of genes from the data set mapped to the pathway divided by the total number of genes present in the pathway-map) and (2) a p-value, calculated by Fisher’s exact test. Pathways Activity Analysis, a function of IPA, enables prediction of the overall activation/inhibition states of the canonical pathways based on a z-score algorithm.

Genes involved in aspects of neurogenesis were compiled from the Gene Ontology (GO) Consortium resources (http://​geneontology.​org/​). GO terms involved in neuronal differentiation, migration and proliferation were compiled, and these functional groupings of genes were assessed with regard to their dysregulation in HSV-1 infected cells compared to uninfected cells, both in the presence of and absence of antivirals.