SARS-CoV-2 superinfection in CD14⁺ monocytes with latent human cytomegalovirus (HCMV) promotes inflammatory cascade

Highlights

• •SARS-CoV-2 Wuhan WA.1, Delta AY.4, and Omicron BA.1 variants can successfully infect CD14⁺monocytes.
• •Superinfection of CD14⁺monocytes found in cells latently infected with HCMV.
• •Transcriptomic analysis shows inflammatory pathway cascades following coinfection of SARS-CoV-2 and HCMV.

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the etiologic agent of coronavirus disease 2019 (COVID-19), has posed significant challenges to global health. While much attention has been directed towards understanding the primary mechanisms of SARS-CoV-2 infection, emerging evidence suggests co-infections or superinfections with other viruses may contribute to increased morbidity and mortality, particularly in severe cases of COVID-19. Among viruses that have been reported in patients with SARS-CoV-2, seropositivity for Human cytomegalovirus (HCMV) is associated with increased COVID-19 risk and hospitalization. HCMV is a ubiquitous beta-herpesvirus with a seroprevalence of 60–90 % worldwide and one of the leading causes of mortality in immunocompromised individuals. The primary sites of latency for HCMV include CD14⁺ monocytes and CD34⁺ hematopoietic cells. In this study, we sought to investigate SARS-CoV-2 infection of CD14⁺ monocytes latently infected with HCMV. We demonstrate that CD14⁺ cells are susceptible and permissive to SARS-CoV-2 infection and detect subgenomic transcripts indicative of replication. To further investigate the molecular changes triggered by SARS-CoV-2 infection in HCMV-latent CD14⁺ monocytes, we conducted RNA sequencing coupled with bioinformatic differential gene analysis. The results revealed significant differences in cytokine-cytokine receptor interactions and inflammatory pathways in cells superinfected with replication-competent SARS-CoV-2 compared to the heat-inactivated and mock controls. Notably, there was a significant upregulation in transcripts associated with pro-inflammatory response factors and a decrease in anti-inflammatory factors. Taken together, these findings provide a basis for the heightened inflammatory response, offering potential avenues for targeted therapeutic interventions among HCMV-infected severe cases of COVID-19.

Summary

COVID-19 patients infected with secondary viruses have been associated with a higher prevalence of severe symptoms. Individuals seropositive for human cytomegalovirus (HCMV) infection are at an increased risk for severe COVID-19 disease and hospitalization. HCMV reactivation has been reported in severe COVID-19 cases with respiratory failure and could be the result of co-infection with SARS-CoV-2 and HCMV. In a cell culture model of superinfection, HCMV has previously been shown to increase infection of SARS-CoV-2 of epithelial cells by upregulating the human angiotensin-converting enzyme-2 (ACE2) receptor. In this study, we utilize CD14⁺ monocytes, a major cell type that harbors latent HCMV, to investigate co-infection of SARS-CoV-2 and HCMV. This study is a first step toward understanding the mechanism that may facilitate increased COVID-19 disease severity in patients infected with SARS-CoV-2 and HCMV.

1. Introduction

Human coronaviruses are members of the Coronaviridae family and Coronavirinae subfamily. The subfamily is further split into four genera: Alphacoronavirus, Betacoronavirus, Gammacoronavirus, and Deltacoronavirus (Cui et al., 2019; Forni et al., 2017). Coronaviruses possess a relatively large positive-sense single-stranded genome of up to 32 kb with a genomic replication cycle consisting of only RNA (Hulswit et al., 2016; Domingo, 2016; Domingo et al., 2015). The genetic material of coronaviruses encodes four key structural proteins: the spike (S) protein, nucleocapsid (N) protein, membrane (M) protein, and envelope (E) protein. Each of these is essential for the formation of a functional infectious viral particle (Garoff et al., 1998; Mortola and Roy, 2004; Masters, 2006). The viral envelope contains the spike (S) and membrane (M) virus-specific glycoproteins (Schoeman and Fielding, 2019). It is proposed that the genetic variability of RNA viruses allows for higher cross-species transmission, specifically because of the broad tropism within the viral spike (S) entry protein and rate of point mutations (Hulswit et al., 2016; Domingo, 2016; Domingo et al., 2015). In the last 20 years, three medically significant betacoronaviruses have transferred to humans through zoonotic events, SARS-CoV (severe acute respiratory syndrome), MERS-CoV (Middle East respiratory syndrome), and SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2). Additionally, the betacoronaviruses HCoV-OC43 and HCoV-HKU1 circulate in the human population and are attributed to less severe cold-like symptoms to moderately severe influenza-like illness in humans (Zhang et al., 2018; Friedman et al., 2018).

Infection by SARS-CoV-2 in humans leads to the respiratory illness coronavirus disease 2019 (COVID-19) (Zheng, 2020). The wide range of severity of COVID-19 has been speculated to be multifactorial, including differences in viral variants, patient genetics, and comorbidities. Recent studies have attempted to understand further the role of co-infection or superinfection in COVID-19 disease severity (Perera et al., 2023). Co-infection of SARS-CoV-2 and influenza A virus promotes increased SARS-CoV-2 infectivity, and patients requiring hospitalization with oxygen support exhibited a higher prevalence of human cytomegalovirus (HCMV) seropositivity (Perera et al., 2023; Bai et al., 2021).

HCMV is a ubiquitous betaherpesvirus with a seroprevalence of 60–90 % worldwide (Staras et al., 2006). HCMV is one of the leading causes of mortality in immunocompromised individuals especially following solid-organ or bone marrow transplantation. Additionally, HCMV infection is a leading cause of congenital birth defects and can cause a wide range of syndromes ranging from microcephaly to hearing loss and cognitive disabilities. Additionally, individuals with compromised cell-mediated immunity in response to the IE antigen or pp65 proteins often exhibit active HCMV infection (Kumar et al., 2019). For healthy, immune-competent individuals, HCMV infection causes very few signs and symptoms during initial infection or episodes of reactivation (Griffiths and Reeves, 2021).

After initial infection, HCMV enters latency, where the viral genome is maintained in the absence of lytic gene expression, and there is no production of infectious virus. The main sites of latency for HCMV are CD14⁺ monocytes and CD34⁺ hematopoietic stem cells (Sissons et al., 2002). Reactivation of the HCMV occurs concurrently with differentiation and maturation of peripheral blood monocytes into macrophages, following their egress from the circulatory system and ingress into the various tissue compartments. This process is guided by cytokine and chemokine signaling. There are periodic episodes of HCMV reactivation, but for healthy individuals, this is recognized and controlled by the immune system and rarely leads to disease. Uncontrolled lytic reactivation can lead to HCMV-related diseases such as pneumonia, colitis, retinitis, and hepatitis (Griffiths and Reeves, 2021).

Current treatments for HCMV target lytic replicating virus and are primarily used following reactivation rather than prophylaxis. In situations where HCMV-positive individuals become immunosuppressed, the likelihood of HCMV reactivation increases (Cook and Trgovcich, 2011). With the use of antiviral treatment, transplant patients have an incidence of roughly 10 % of experiencing HCMV disease within the first-year post-transplant (Stern et al., 2019). In their systematic review, Belsky et al. (2021) found higher levels of inflammatory markers, rates of intensive care, and hospital mortality in cancer and solid organ transplant patients experiencing COVID-19 disease. Infection with SARS-CoV-2 leads to viral suppression of natural killer (NK) cells, increased inflammation, and depletion of CD8⁺ T cells, creating ideal conditions for HCMV reactivation (Sissons et al., 2002; Belsky et al., 2021; Forrest et al., 2023; Choi et al., 2017; Clari et al., 2013; Castón et al., 2016). In a recent case study, patients experienced HCMV reactivation while hospitalized in the ICU for COVID-19 (Moniz et al., 2021). These patients had a negative viral load of HCMV prior to movement into the ICU. Additionally, HCMV seropositivity alone is sufficient to increase risks associated with COVID-19, specifically acute respiratory disease syndrome, or ARDS (Alanio et al., 2022). While SARS-CoV-2 infection may promote many of the ideal conditions for HCMV reactivation, HCMV has been reported to increase SARS-CoV-2 infection into epithelial cells by upregulating cell surface receptors (Perera et al., 2023).

Angiotensin-converting enzyme 2 (ACE2) is the functional receptor for the α-coronavirus, HCoV-NL63, and β-coronaviruses, SARS-CoV and SARS-CoV-2³. The ACE2 receptor was identified as being required for a successful in vivo SARS-CoV infection (Imai et al., 2008). In epithelial cells, HCMV upregulates ACE2, promoting a SARS-CoV-2 superinfection (Perera et al., 2023). The viral SARS-CoV-2 spike (S) protein is composed of separate subunits, namely S1 and S2, which play distinct roles in the viral life cycle. The S1 subunit primarily facilitates the binding process and the S2 subunit facilitates membrane fusion. The viral spike protein interacts with the human ACE2 receptor located on the host cell membrane through the S1 subunit within the receptor-binding domain (RBD). The RBD of SARS-CoV-2 exhibits a stronger binding interaction with soluble ACE2 when contrasted with the interaction of SARS-CoV RBD with the same receptor (Wrapp et al., 2020; Beyerstedt et al., 2021). Likewise, many studies have demonstrated the significance of Transmembrane Serine Protease 2 (TMPRSS2) in influencing the susceptibility of host cells to SARS-CoV-2 (Hoffmann et al., 2020). The cleavage of the viral spike protein, a crucial step for viral entry, is orchestrated through the enzymatic activity of TMPRSS2 present within the host cell (Abbasi et al., 2021).

As previous research has demonstrated an association between HCMV infection and the severity of COVID-19, our objective was to explore the infection of CD14⁺ monocytes with latent HCMV by SARS-CoV-2. We aimed to ascertain whether these cells could facilitate co-infection and if any alterations in the transcriptome could explain the severity of the disease. In this study, we establish the Omicron BA.1 variant of SARS-CoV-2 successfully infects CD14⁺ monocytes latently infected with HCMV strain TB40E. We also show that HCMV upregulates the ACE2 receptors on CD14⁺ monocytes, enhancing cell permissiveness of SARS-CoV-2. Our transcriptome analysis results suggest there is interplay between SARS-CoV-2 and HCMV that influences the cellular landscape to promote an increase in pro-inflammatory factors and decrease anti-inflammatory factors which may influence COVID-19 disease severity.

2. Materials and methods

2.1. Cell lines

Human Calu-3 lung adenocarcinoma cells (ATCC, cat. # HTB-55) were grown in antibiotic-free Dulbecco's modified Eagle medium (DMEM) (Corning, cat. # 45,000–304) supplemented with 10 % fetal bovine serum (FBS) (Corning CAT# 35–010-CV). The following reagent was obtained through BEI Resources, NIAID, NIH: Cercopithecus aethiops Kidney Epithelial Cells Expressing Transmembrane Protease, Serine 2 and Human Angiotensin-Converting Enzyme 2 (Vero E6-TMPRSS2-T2A-ACE2), NR-54,970. Vero E6 recombinant TMPRSS2-ACE2 were maintained in antibiotic-free Dulbecco's modified Eagle medium (DMEM) (Corning, cat. # 45,000–304) supplemented with 10 % fetal bovine serum (Corning, Ref. # 35–010-CV). Human PB CD14⁺ monocytes (CAT# 70,035.1) were obtained from STEMCELL Technologies. Cells were certified HCMV negative before shipment. Human CD14⁺ cells were maintained in Iscove DMEM (Hyclone) supplemented with 20 % heat-inactivated FBS, 50 ng/mL M-CSF, 50 ng/mL stem cell factor (SCF), 50 ng/mL G-CSF, 50 ng/mL GM-CSF, 50 ng/mL IL-3 (R&D Systems) at a density of 1 × 10⁶ cells/mL on low cell-binding plates (Nunc Hydrocell or Sphera). The medium was made fresh and replaced every 3 days. Cells were incubated with TB40E virus for 1 hr. then washed twice with Hanks Balanced Salt solution (HBSS). After 12–14 days post infection cells were transferred to 12-well plates. Calu-3 cells were seeded at a density of 75,000 cells into 12-well transwell compartments 4 h before SARS-CoV-2 infection. Following a stringent wash and replacement of media, Calu-3 cells in the top transwell were added to the 12-well plate containing latent or mock CD14⁺ cells. All cells were grown at 37 °C in a humidified incubator supplemented with 5 % CO₂.

2.2. Viruses

All HCMV viruses used in this study were BAC-derived TB40E GFP strains. For HCMV infection studies, cells were infected at a multiplicity of infections (MOI = 5) based on titration on human fibroblast (HF) cells. For SARS-CoV-2, cells were infected at a SARS-CoV-2 (MOI = 1) as based on titration on Vero E6 TMPRSS2-ACE2 cells. The following reagent was obtained through BEI Resources, NIAID, NIH: SARS-Related Coronavirus 2, Isolate hCoV-19/USA/GA-EHC-2811C/2021 (Lineage B.1.1.529; Omicron Variant), NR-56,481, contributed by Mehul Suthar. SARS-Related Coronavirus 2 – Isolate hCoV-19/USA/GA-EHC-2811C/2021 (Lineage B.1.1.529; Omicron Variant).

2.3. Immunofluorescence assay (IFA)

To perform immunofluorescence detection of HCMV and SARS-CoV-2 proteins, CD14⁺ monocytes with and without HCMV were seeded in 12-well dishes with coverslips at a density of 60,000 cells per well. The cells were directly infected with SARS-CoV-2 (MOI=1) or left as mock and incubated for 48 h. The supernatant was removed and samples were washed with PBS before being fixed with 4 % paraformaldehyde in PBS for 30 min. Cells were then washed three times before being permeabilized with 0.5 % Triton for 10 min. Samples were then blocked with 3 % goat serum for 30 min at room temperature. Primary antibody in BSA (2.5 mg/ml N1-Rabbit, CAT# PA5–119,601; 200 mL/well) and/or HCMV IE1 primary antibody (gift of Bill Britt, University of Alabama), 1:500, 200uL/well was incubated for 2 h at room temperature. The primary antibody was removed, and samples were washed three times with 1 % BSA. The secondary antibody (1:500, anti-Rabbit 594; 200 mL/well) was incubated for 30 min at room temperature, rocking. The secondary antibody was removed, and samples were washed three times with 1 % BSA. Wells were rinsed three times with PBS and coverslips were mounted on cell culture dishes with ProLong™ Gold Antifade Mountant with DAPI (Cat#: P36935).

For CD14⁺ IFA of ACE2, approximately 60,000 cells were added to coverslips following a 7-day HCMV infection in culture. Mock samples did not contain HCMV. Cells were allowed to settle and adhere to the coverslips for 30 min at room temperature. HCMV IE1 primary antibody (gift of Bill Britt, University of Alabama) and ACE2 primary antibody (CAT#: SMO754) were added at a concentration of 1:500, 200uL/well and incubated for 2 h at room temperature. The remaining steps were followed as described above.

2.4. ELISA

The supernatant levels of IL-1β and TNF-α were measured using the DuoSet ELISA and ancillary reagent kits (R&D systems). The protocol was followed according to the manufacturer's instructions. The supernatant was obtained from a co-culture transwell system, CD14⁺ latently infected with HCMV for 12 days were on the bottom, and Calu-3 cells following a 2-hour SARS-CoV-2 absorption were cultured in the top portion. After 72-hour incubation, a total of 1 ml of supernatant was collected and stored at −80 °C until use. Samples were spun at 1000 x G for 10 min to remove cellular debris and transferred to clean tubes. For ELISA, 100 ml of supernatant was added to each of the 3 wells for technical replicates. Each sample set was performed 3 times for a total of 3 biological replicates.

2.5. Plaque assay

For plaque assays, Vero E6 ACE2 cells (ATCC® CRL‐1586™) (100,000 cells/well) in 1 mL medium were seeded in a 12-well culture plate and incubated overnight. For infection, 3 serial dilutions of virus stock (SARS-CoV-2, BA.1.1.7) were prepared by combining 10 µL of virus stock and 90 µL of DMEM/10 %FBS and mixed thoroughly by slowly pipetting up and down. The 12-well plate with 90–95 % confluent cell monolayer was used for infection. Cell culture medium was removed, and cells were washed with sterile 1X DPBS (no calcium, no magnesium), followed by the addition of 500 µL of each dilution, Live BA.1 or Heat Inactivated. The infected cell plate was incubated at 37 °C with 5 % CO₂ for an hour with gentle rocking every 15 min to spread the virus inoculum evenly and prevent the monolayer from drying. After 1 h, the unabsorbed virus was removed by discarding the virus supernatant, washing the cells with sterile 1X DPBS (1 mL), and replenishing with CMC 1 mL overlay medium. The plate was then incubated at 37 °C with 5 % CO2 for 72 h. After 3 days of incubation, CMC was removed, and cells were washed with DPBS twice to remove all CMC from wells. Then, a 4 % paraformaldehyde solution (2 mL) (Thermo Scientific, Waltham, MA, USA, Cat # J19943-K2) was added, and the plate was incubated at room temperature for an hour. After removing formaldehyde, the cells were washed with 1X DPBS. The infected cells were stained with 0.2 % crystal violet solution (Sigma-Aldrich, St. Louis, MO, USA; Cat #C0775 in 20 % ethanol (2 mL) at room temperature for 30 min with a gentle rocking every 10 min. The staining solution was removed, and cells were washed with sterile distilled water (three times). The plate was inverted and left to dry on an absorbent pad for 1–2 h (or overnight). For each well/sample/dilution, the number of plaques for that dilution was counted, and the plaque forming unit (PFU per mL) was calculated using the equation-

$$\begin{array}{c}\text{PFU}/\text{mL}=\text{average}\text{number}\text{of}\text{plaques}\hfill \\ \left(\text{Dilution}\text{factor}\text{of}\text{the}\text{well}\mathrm{x}\text{volume}\text{of}\text{inoculum}\text{per}\text{plate}\right]\hfill \end{array}$$

The experiment was performed in triplicates (technical replicates) and in at least 2 sets of samples (biological replicates). The average number of plaques per sample was counted, and the titer of the virus in the stock/sample is represented as PFU/mL.

2.6. RNA-Extraction and qPCR

For total RNA extraction, samples were processed using spin-column purification with Direct-zol™ RNA Miniprep Plus kit (Cat # R2072, Zymo). Briefly, Trizol-LS was added directly to the cell culture dish to allow for cell lysis. Following a 5-minute incubation, an equal volume of ethanol was added to the lysed sample, mixed thoroughly, and transferred into a zymo-spin IICR column in a collection tube. Samples were further processed following the manufacturer's instructions. TaqPath™ 1-Step RT-qPCR Master Mix (CAT#: A15299), along with IDT DNA primers and probes, were used according to the manufacturer's instructions for RT-qPCR analysis. Intracellular SARS-CoV-2 genome copies were determined using SARS-CoV-2 RUO qPCR Primer & Probe Kit (IDT, Inc.) on Quantstudio 5 with TaqPath™ 1-Step RT-qPCR Master Mix (ThermoFisher Scientific). Predesigned primers and probes from IDT and primers from ThermoFisher as follows: IL1-β (Hs.PT.58.1518186), TNFα (Hs.PT.58.45380900), GAPDH (Hs.PT.39a.22214836), 7SK (Hs.PT.58.38798776.g).

2.7. Library preparation and RNA-sequencing

RNA was submitted to the Nevada Genomics Core, where libraries were prepared for sequencing by paired-end RNA-seq using Illumina Stranded mRNA prep and NextSeq 2000 (Illumina, San Diego, CA, USA).

2.8. Differential expression analysis

Sequence data (fastq files) were processed by the Nevada Bioinformatics Core. Raw_seq data was downloaded from Illumina Basespace and data was trimmed using fastp v0.20. After trimming, both overrepresented sequences, Per Seq GC content and Per Base N content, were corrected. Trimmed data was aligned to Grch38 Ensembl human reference using HISAT2 v2.2.1. HISAT2 alignment was used for downstream DEG analysis. The differential gene expression patterns were assessed using DESeq2 pipeline version 1.36.0 (Love et al., 2014). Following DESeq2, the Data were analyzed using Advaita Bio's iPathwayGuide (Draghici, 2007, Donato, 2013).

2.9. Statistical analysis

Unless indicated, statistical analyses were completed by one-way, two-way ANOVA, or unpaired student t-test with a p < 0.05. All statistical analyses were completed using GraphPad Prism software except for RNA sequencing analysis.

3. Results

3.1. SARS-CoV-2 infects human peripheral blood CD14⁺ monocytes

Human CD14⁺ monocytes isolated from peripheral blood mononuclear cells were purchased from a commercial source (STEMCELL Technologies) and grown according to previously published reports to maintain them in an undifferentiated state (Rossetto et al., 2013; Hargett and Shenk, 2010). To determine if SARS-CoV-2 could successfully infect and replicate within CD14⁺ monocytes, cells were infected with SARS-CoV-2 for 48-hours before RNA was harvested. RT-qPCR was performed with primers and probes specific for SARS-CoV-2 nucleocapsid 1 (N1) and the cellular control (GAPDH). The nucleocapsid (N1) transcript copy number was calculated for the original Wuhan SARS-CoV-2 strain (WA.1) along with Delta (AY.4), and Omicron (BA.1) (Fig. 1A). A heat-inactivated control of each variant was used as a control for the possibility of remaining viral particles attached to the cell surface or viral particles entering the cell but not replicating. Both Delta (AY.4) and Omicron (BA.1) displayed higher N1 copy numbers compared to Wuhan (WA.1). Omicron BA.1 variant has unique mutations within the viral RBD that leads to better immune evasion and increased binding affinity to ACE2 (Xia et al., 2022; Zhang et al., 2021). Given that the Omicron BA.1 variant had the highest copy number in our CD14⁺ cells and was the predominant circulating variant during the study period, it was selected for all subsequent experiments.

Fig. 1.

The CD14⁺ monocytes are susceptible to SARS-CoV-2 infection with variants WA.1, AY.4, and BA.1 and are permissive to viral replication.

A. SARS-CoV-2 nucleocapsid (N1) copy number in CD14⁺ monocytes 48 h post-infection with heat-inactivated (HI) or replication-competent (live) variants WA.1, AY.4, and BA.1 (MOI=1.0). Total RNA was extracted and quantified by RT-qPCR using primers/probes for N1 and cellular 7SK to calculate the copy number. B. Plaque assay performed on Vero cells infected with HI and live BA.1 virus to determine the titer. C. Schematic diagram of co-culture system. Calu-3 lung cells were grown in the top transwell compartment, with latent HCMV CD14⁺ cells grown in the bottom compartment. Isolated transfer of SARS-CoV-2 from the top transwell into the bottom well for infection of CD14⁺ cells. Co-infected cells were harvested for RNA. D. 72 h post-co-culture, RT-qPCR of IE1/2 from HCMV CD14+ was performed. RNA was harvested and RT-qPCR was performed to measure the amount of IE1/2 normalized to cellular GAPDH. The fold change for the CD14+ cells cultured with either heat-inactivated or Live SARS-CoV-2 was calculated relative to mock-infected. E and F. ORF3a and N1 copy number in CD14⁺ monocytes following co-culture for 72 h with Calu-3 cells infected with HI or live BA.1 (MOI=1.0) virus, or mock controls. Total RNA was extracted and quantified by RT-qPCR using primers/probes for ORF3a, N1, and cellular 7SK to calculate the copy number. G. Fluorescent microscope images of CD14+ cells performed 72-hours post co-culture infection. Images taken at 40X with DAPI, IE1 (TB40E-GFP, Alexa Fluor 488), and SARS-CoV-2 N1 (Red, Alexa Fluor 594) in mock, HI, and live BA.1 SARS-CoV-2.

To calculate the concentration of SARS-CoV-2 to use for infection as well as to test the inactivation of the heat-inactivated controls, plaque assays were completed using 1 to 10 dilutions (Fig. 1B). Briefly, Vero-E6-TMPRSS2-ACE2 cell monolayer was infected with dilutions of SARS-CoV-2, starting with an unknown concentration. Following an absorption incubation of 2 h, a carboxymethylcellulose (CMC) overlay was applied to the infected monolayer to inhibit virus spread and limit virus growth to cell foci at the initial infection sites. Infected cells were incubated for 3 days, subsequently fixed using 4 % formaldehyde, and stained with crystal violet for plaque visualization. Plaques were then counted to determine the infectious virus titer used for future experiments. Infectious virus titers are measured in plaque‐forming units (PFU)/mL at 1.25 for the 1:1000 dilution. A multiplicity of infection (MOI) of 1 for all future experiments was used for the SARS-CoV-2 infection with the same volume added for heat-inactivated control.

After determining that SARS-CoV-2 can directly infect CD14⁺ monocytes, we proceeded to investigate SARS-CoV-2 infection of CD14⁺ monocytes harboring latent HCMV using a co-culture system with SARS-CoV-2 infected Calu-3 cells. The CD14⁺ monocytes were infected with HCMV (strain TB40E, MOI=5), and latency was established over a period of 12 days. Co-infection experiments were performed using a transwell plate system, where CD14⁺ cells were transferred to the bottom well, and Calu-3 cells were grown in the top transwell compartment. Before adding the Calu-3 to the transwell plate, they were infected with replication-competent (live) or heat-inactivated (HI) BA.1 SARS-CoV-2 for 2 h, followed by removing unattached particles. The Calu-3 transwells were then placed in the upper compartment and incubated with the CD14⁺ cells for 72 h. Following incubation, the CD14⁺ cells and supernatant were harvested for RT-qPCR, ELISA, RNA-sequencing, and transcriptome analysis. A schematic of the co-culture system is demonstrated in Fig. 1C. To examine whether the CD14+ cells containing latent HCMV underwent reactivation during the 72-hour co-infection with SARS-CoV-2, RT-qPCR of IE1/2 mRNA was performed in mock-HCMV, heat-inactivated, and live SARS-CoV-2-HCMV (Fig. 1D). The relative copies of IE1/2 transcripts was increased in the heat-inactivated and live SARS-CoV-2 compared to the mock-infected, suggesting HCMV remained latent in mock-treated cells and experienced some reactivation following co-infection.

The SARS-CoV-2 genome contains 9 accessory open reading frames (ORFs) 3a, 3b, 6, 7a, 7b, 8, 9b, 9c, and 10 (Naqvi et al., 2020). The structural and accessory proteins are produced through a series of nested subgenomic RNAs (sgRNAs), which consist of intermediate negative RNAs (Liu et al., 2014; Zhang et al., 2022). To confirm active replication of SARS-CoV-2 following infection of CD14⁺ cells we examined levels of subgenomic and genomic transcripts. It is noted that some recent studies debate the use of subgenomic RNAs as a marker of active replication (Alexandersen et al., 2020), while others suggest it can be reliably used as a marker of infectivity and replication (Santos Bravo et al., 2022; Rosenke et al., 2023; Grassi et al., 2022). We performed RT-qPCR measuring subgenomic replicative negative sense transcript ORF3a and the genomic nucleocapsid N1 (Fig. 1E-F). SARS-CoV-2 viral transcripts were analyzed from CD14⁺ cells with and without HCMV co-infection. Slightly higher, but not significantly higher levels of subgenomic ORF3a copies were found in sets with HCMV (Fig. 1E), and we found N1 copies to be comparable between sets (Fig. 1F).

In addition, an immunofluorescence assay (IFA) was performed to visualize SARS-CoV-2 and HCMV in CD14⁺ monocytes. Following a 12-day infection of CD14+ monocytes with HCMV, monocytes were transferred to bottom transwell compartments containing coverslips, and SARS-CoV-2 was added to the top transwell compartment containing Calu-3 cells. Following a 3-day co-culture infection with SARS-CoV-2, for a total of 15 days of HCMV infection, top transwell compartments were removed, and IFA was performed on bottom well coverslips. Primary antibodies specific for either SARS-CoV-2 nucleocapsid protein (red) or HCMV immediate-early 1 (IE1) (green) were used to visualize the respective infected cells. ProLong gold antifade with DAPI was used to stain the nucleus and mount the coverslips (Fig. 1G). No green signal was detected for experiments without HCMV (Mock-Mock, Mock-HI, and Mock-SARS-CoV-2). The SARS-CoV-2 nucleocapsid protein (red) was found in all cells regardless of HCMV status. We note that the HCMV TB40E strain used in these studies does express GFP under the control of a constitutive promoter, but when the virus is latent, as in these experiments, very little, if any, GFP signal is detectable. We cannot rule out that the green signal from the HCMV latent cells is exclusively from the antibody staining for IE1 but could also be from the GFP in the virus; however, this is still a marker for HCMV infection only seen in the HCMV-positive cells. The immunofluorescence assay (IFA) visually confirmed the infection of CD14+ monocytes in our cultured cells by detection of SARS-CoV-2 proteins.

3.2. The infection with HCMV increases ACE2 expression in CD14⁺ monocytes, increasing susceptibility to SARS-CoV-2 infection

Considering that the ACE2 receptor is required by SARS-CoV-2 to enter the cell, we next investigated the levels of ACE2 in CD14⁺ monocytes pre- and post-HCMV infection. To help answer the question of whether the increase in ACE2 happens prior to the establishment of latency, the expression of ACE2 was assessed at 7 and 14 dpi. The 7 dpi represents a time past the initial events of infection during the process of establishing latency, but it is not fully latent. The experimental design schematic is shown in Fig. 2A. HCMV immediate early −1 and −2 (IE1/2) transcripts were measured at 7 and 10 dpi using RT-qPCR, normalized with cellular GAPDH (Fig. 2B). As HCMV establishes latency the amount of IE1/2 decreases as the major immediate-early promoter is silenced following an initial burst of transcription upon infection. The decrease in IE1/2 from days 7 to 10 in our study follows a similar pattern of decreased IE expression noted in previous studies of CD14⁺ HCMV latency models (Rossetto et al., 2013; Hargett and Shenk, 2010). Protein lysates harvested from CD14⁺ cells at 7 days post-infection (dpi) and 14 dpi, along with mock controls for SDS-PAGE western blot (Fig. 2C). Image J analysis of western blot bands showed a roughly 3-fold higher level of ACE2 in cells containing HCMV (Fig. 2D). This data is consistent with previous reports that HCMV upregulates the expression of ACE2 in epithelial cells to promote a superinfection with SARS-CoV-2¹³. To visualize the co-expression of HCMV IE1 protein with ACE2, immunofluorescence assays were performed comparing mock with 7 dpi HCMV infected CD14⁺ monocytes (Fig. 2E). HCMV infected cells expressing IE1 are visible by the green signal and cellular ACE2 by the red signal, and DAPI was used for nuclear visualization. There is a heterogeneous mixture of cells, and some show infection with HCMV along with the expression of ACE2. Results confirm levels of ACE2 receptor are increased following infection and latency of HCMV in CD14⁺monocytes. These data reveal a possible reason for higher infection of SARS-CoV-2 in monocytes.

Fig. 2.

HCMV infection in CD14⁺ monocytes increases ACE2 expression.

A. Schematic of experimental design. Mock and HCMV-infected (MOI=5) CD14⁺ monocytes were cultured in non-adherent dishes for 7- and 14 days. At 7 days post-infection (dpi), a fraction of cells was collected and divided for use to assess proteins by western blot and immunofluorescence, and at 14 days, cells were collected for western blot. B. To monitor the establishment of latency, the expression of IE1/2 was measured by RT-qPCR at 7 and 10 dpi. C. Western blot from protein lysates at 7 and 14 dpi in mock and HCMV infected CD14⁺. Proteins were visualized using Anti-ACE2 antibody and cellular anti-β-Actin antibody. D. Protein quantification was further assessed with ImageJ analysis of the band of ACE2 compared to b-Actin. E. To visualize the expression of ACE2, immunofluorescence performed at 7 dpi of mock and HCMV-infected CD14⁺ monocytes. Green signal is HCMV anti-IE1 primary antibody with anti-mouse-488 secondary, and GFP from HCMV TB40E. The red signal is Anti-ACE2 rabbit primary antibody with anti-rabbit-594 secondary. Cellular nuclei were stained and visualized with DAPI (Blue).

3.3.1. Transcriptomic analysis of HCMV and SARS-CoV-2 co-infected CD14⁺ monocytes revealed suppression of anti-inflammatory and activation of pro-inflammatory immune pathways

To assess changes in the transcriptome, we performed RNA-seq from the CD14⁺ harboring latent HCMV and co-infected with either heat-inactivated (HI) or live SARS-CoV-2 as described co-cultured system. We performed bulk RNA-seq on the samples, and we noted that this is a heterogenous mixture of cells that were infected with HCMV along with SARS-CoV-2, but not every cell may contain both viruses, and our further analysis is on the population as a whole. All significantly differentially expressed (DE) genes represented in terms of their measured expression change with a log fold change greater than 1 and less than 0.05 for significance were included in the circle pie charts. The SARS-CoV-2 BA.1 genes (log fold change) are shown in green, while the heat-inactivated control sample genes are purple. Cellular gene expression differences between heat-inactivated and live BA.1 versus Mock with LogFC (Fold Change) with p-value < 0.05 (Fig. 3A). The principal component analysis of all 9 samples, 3 technical replicates of mock (black), heat-inactivated BA.1 (purple), and live BA.1 (green) revealed distinct groups for experimental conditions, indicating each sample group has a unique transcriptome profile (Fig. 3B). This was further strengthened following differential expression analysis showing 747 DE downregulated genes in the live SARS-CoV-2 versus Mock and 143 DE downregulated genes in the heat-inactivated versus Mock, with 61 DE genes in common between the two. Additionally, 1074 upregulated DE genes were found in SARS-CoV-2 live versus Mock and 64 upregulated DE genes in the heat-inactivated, with 238 in common.

Fig. 3.

Cellular transcriptome analysis of HCMV and SARS-CoV-2 co-infected CD14⁺ monocytes reveal a suppression of anti-inflammatory and activation of pro-inflammatory immune pathways.

A. All significantly differentially expressed (DE) cellular genes represented in terms of their measured expression change with a log fold change greater than 1 and less than 0.05 for significance. The number of differentially expressed cellular genes are shown in green. The number of differentially expressed cellular genes following infection with the HI virus is shown in purple. B. Principal component analysis (PCA) of three replicates within each group (mock, heat-inactivated BA1.HI and BA1.Live). C. Heat map of significantly downregulated differentially expressed genes (DEGs) in SARS-CoV-2 (BA1.Live) versus Mock. D. Heat map of significantly upregulated differentially expressed genes (DEGs) in SARS-CoV-2 (BA1.Live) versus Mock.

3.3.2. Transcriptomic analysis of HCMV and SARS-CoV-2 co-infected CD14⁺ monocytes revealed suppression of anti-inflammatory and activation of pro-inflammatory immune pathways

To further elucidate the changes occurring within the cellular transcriptome during co-infection, we used iPathwayGuide to identify the top upstream genes activated and inhibited for live SARS-CoV-2 versus the mock samples containing HCMV (Fig. 3C, D). In the activated upregulated genes, the most significant was the cytokine interleukin-1β (IL-1β) (Fig. 4). IL-1β is a multifunctional cytokine affecting nearly all cell types and is considered a potent pro-inflammatory molecule (Dinarello, 1996). It has been demonstrated in monocytes that activation induces cleavage of proIL-1β, and mature IL-1β is secreted (Dinarello, 1996). In severe COVID-19 disease, IL-1β was found to be significantly increased within lung areas of viral production (Han et al., 2020).

Fig. 4.

Cellular transcriptome analysis of HCMV and SARS-CoV-2 co-infected CD14⁺ monocytes reveal upstream genes activated and inhibited.

Summary of DEGs at adjusted p 〈 0.05. BA.1 vs Mock, LFC 〉 0 (up): 3764, 22 %, LFC < 0 (down): 3682, 21 %. HI vs Mock, LFC > 0 (up): 2496, 15 %, LFC < 0 (down): 2218, 13 %. A. and B. Global analysis of cellular gene expression differences between HI and live BA.1 versus Mock with LogFC (Fold Change) with p value < 0.05. The graph depicts the topmost significant DEGs. On the right side of the panel are interaction string diagrams between the top DEGs that are activated (A) or inhibited (B).

Within the top upstream activated genes, we found high differentiated levels of Toll-like receptors (TLRs). Toll-like receptors recognize pathogen-associated molecular patterns and recognize a wide range of ligands from bacteria and viruses (Janssens and Beyaert, 2003). The top increased TLRs from our sample set included TLR2 and TLR4 (Fig. 4A). The plasma membrane receptor, TLR4, is involved in the recognition of viruses by binding to viral envelope proteins. In previous SARS and H5N1 influenza virus studies, TLR4 was associated with increased acute lung injury (Janssens and Beyaert, 2003; Kijpittayarit et al., 2007; Takeuchi and Akira, 2010). In silico studies have identified a notably robust interaction between the surface TLR4 and the spike protein of SARS-CoV-2 (Aboudounya and Heads, 2021). The signaling through TLR4 is also associated with the activation of transcription factors NF-kB and IRF3^52,53. Studies have suggested polymorphisms in innate immune receptors, such as TLR2 and TLR4, are associated with an increased risk of primary HCMV infection and disease severity (Cervera et al., 2007). In patients receiving liver transplants, the presence of single nucleotide polymorphism in TLR2 carries an increased risk of HCMV disease and viral load (Kijpittayarit et al., 2007). These types of studies also highlight the difficulty in assigning the disease severity of viral infections to a particular factor, given diverse host genetics.

One of the top upregulated genes with upstream genes activated included nuclear factor kappa B (NF-κB). NF-κB is a family of multifunctional transcription factors that regulate gene expression for innate and adaptive immunity, inflammation, and cell differentiation and survival (Mizgerd, 2006; Lawrence, 2009). In HCMV infection, NF-κB is required for efficient production of IE mRNA and proteins (DeMeritt et al., 2004). In severe COVID-19 cases, elevated levels of proinflammatory markers have been reported (Hariharan et al., 2021). In studies done with SARS-CoV nucleocapsid protein (N protein), NF-κB was significantly increased in a dose-dependent manner in Vero cells (Liao et al., 2005). In previous SARS-CoV studies, it was observed that peripheral blood mononuclear cells (PBMCs) treated with the S protein exhibited an increase in NF-κB (Dosch et al., 2009). The activation of NF-κB induced by β-coronaviruses involves a pathway mediated by the myeloid differentiation primary response 88 (MyD88) pathway through pattern-recognition receptors (PPRs) (Birra et al., 2020; Piras and Selvarajoo, 2014). This activation results in the production of various cytokines, such as interleukin-6 (IL-6), TNF-α, and chemokines, matching the increase seen in our transcriptomics analysis.

IL-6 is a versatile cytokine with pleiotropic effects, influencing a wide range of physiological processes like cell proliferation, differentiation of lymphocytes, cell survival, and apoptosis (Kamimura et al., 2003). IL-6 affects the immune, hematopoietic, inflammatory, nervous, and endocrine systems and is necessary for host defense (Kamimura et al., 2003; Akira et al., 1993). During infection of HCMV, IL-6 secretion prevents apoptosis by increasing the antiapoptotic factor survivin (Botto et al., 2011). During HCMV pneumonia, IL-6 levels significantly increase and, to a lesser extent, increase during allograft rejection (Humbert et al., 1993). Clinical studies report that IL-6 can be used to predict the severity of COVID-19 disease, with high plasma IL-6 levels being attributed to cytokine storm (Liu et al., 2020). Additionally, high levels of IL-1β, IL-6, IL-8, and TNF-α were reported in plasma of hospitalized patients with SARS-CoV and MERS-CoV (Wang et al., 2022). Altogether, these differentially expressed activated genes show significant changes in the inflammatory and cytokine pathways.

Further analysis of transcriptomics revealed suppression of anti-inflammatory modulators. The top genes with upstream genes predicted as inhibited included G-protein signaling modulators (GPSM1 and GPSM2) and Interleukin-10 (IL-10) (Fig. 4B). GPSM1, also known as an activator of G protein signaling 3 (AGS3), serves as a receptor-independent activator for G protein signaling (Cismowski et al., 1999). GPSM1 and GPSM2 share 59 % sequence homology and have the same domain structure (Cismowski et al., 1999; Takesono et al., 1999). These proteins have been implicated in G-protein signaling pathways and aiding in cell devision (Blumer et al., 2002). A deficiency in GPSM1 prevents NF-κB inflammatory signaling induced by TLR4 in macrophages, mitigating the pro-inflammatory response (Yan et al., 2022). As shown in Fig. 3F, GPSM genes had an overall negative log fold change. Decreased expression of GPSM1/AGS3 has been associated with reduced levels of phosphorylated cyclic AMP response element-binding protein (p-CREB), resulting in less apoptosis (Shao et al., 2014).

The largest fold change for upstream genes inhibited was Interleukin-10 (IL-10) (Fig. 4B). Initially recognized for its capacity to impede the activation and effector function of T cells, monocytes, and macrophages, IL-10 is a multifunctional cytokine that exerts diverse effects on a wide range of hemopoietic cell types (Moore et al., 2001). IL-10 is an anti-inflammatory marker that inhibits cytokine production and mononuclear cell function. A suppression of IL-10 levels assists cells in the clearance of viral infections (Brooks et al., 2006). IL-10 acts as a potent anti-inflammatory inhibitor of IL-1β, TNF-α, and IL-6 and enhances the expression of CD14⁷². Cytokine expression in monocytes is subdued by the HCMV-encoded viral IL-10 (cmvIL-10) through the inhibition of NF-κB activity (Nachtwey and Spencer, 2008). Elevated levels of TNF-α, IFN-γ, IL-2, IL-4, IL-6, and IL-10 were observed in severe cases of COVID-19 and correlated with poor prognosis (Han et al., 2020). Interestingly, critical cases showed significantly higher serum levels of IL-6 and IL-10⁴⁶.

3.3.3. TNF signaling, inflammatory responses, and cytokine-cytokine receptor pathways were all significantly altered in the SARS-CoV-2 infected HCMV monocytes

To further investigate the transcriptomic changes within co-infected CD14⁺ monocytes, we analyzed the biological pathways using iPathwayGuide. A chord diagram was used to display the top 6 most significant biological pathways using Bonferroni at -log10(p) (Fig. 5A). Those top 6 pathways were: cytokine-cytokine receptor interaction, viral protein interaction with cytokine and cytokine receptor, chemokine signaling pathway, NF-κB signaling pathway, IL-17 signaling pathway, toll-like receptor signaling pathway. Among the viral protein interaction with cytokine and cytokine receptors, the highest fold change genes are shown in a string map diagram (Fig. 5B). Interestingly, both IL-6 and TNF are present. Two other significant pathways included inflammatory response and TNF signaling (Fig. 5C and D, respectively).

Fig. 5.

Modulation of immune pathways in CD14⁺ cells co-infected with HCMV and SARS-CoV-2.

A. Chord diagram generated in iPathwayGuide for top 6 pathways with most important genes. Cytokine-cytokine receptor interaction (light green), viral protein interaction with cytokine and cytokine receptor (red), chemokine signaling pathway (yellow), NF-kappa B signaling pathway (light teal), IL-17 signaling pathway (blue), Toll-like receptor signaling pathway (purple). The -log10(p) Bonferroni in green A-F. Log2(FC) in pink. B. Inflammatory response genes fold change with corresponding string map. C. Viral protein interaction with cytokine and cytokine receptor genes fold change with corresponding string map. D. TNF Signaling genes fold change with corresponding string map. String maps were generated using STRING (Szklarczyk et al., 2023).

Within these categories, the common genes include IL-1β, TNF-α, NF-κB, and IL-6. Notably, most of the genes and biological processes showing differential expression were linked to inflammatory response and the immune system. However, the exact mechanisms involved in HCMV and SARS-CoV-2 co-infection regulation of pro- and anti-inflammatory signaling pathways will require further investigation. Taken together, these studies suggest that latently infected monocytes can experience infection with SARS-CoV-2, and this event kicks off a cascade of pro-inflammatory cytokines and decreases suppressive inflammatory markers, resulting in an ideal state for acute respiratory disease.

3.3.4. Increased TNF-α and IL-1β levels in HCMV and SARS-CoV-2 co-infected cells

The signaling pathways of the innate immune response, specifically those involving IL-1 and TNF-α, are essential conserved innate antimicrobial responses (Medzhitov and Janeway, 1998). Infection of monocytes with HCMV causes a sustained expression of IL-1β (Yurochko and Huang, 1999). Based on our transcriptome analysis, we chose to further measure levels of TNF-α and IL-1β through ELISA assays and RT-qPCR. The experiments were performed as described above, where CD14⁺ cells were infected with HCMV and cultured for approximately 12 days to allow the virus to become latent. Then, the CD14⁺ cells were transferred to the bottom of a transwell plate, with Calu-3 cells newly infected with SARS-CoV-2 in the upper compartment. The cells were incubated for 72 h before harvesting. The supernatant was collected and used for the ELISA, while the cells were harvested for RNA extraction. First, we measured the mRNA of TNF-α and IL-1β by RT-qPCR; transcripts were measured using TaqPath and predesigned primers and probes for TNF-α and IL-1β. Transcripts were normalized to cellular 7SK, and the ∆∆CT method was used to calculate fold change (Fig. 6A and C). Heat-inactivated and live SARS-CoV-2 were compared to HCMV latent untreated monocytes.

Fig. 6.

TNF-α and IL-1β levels are increased in CD14⁺ co-infected with HCMV and SARS-CoV-2.

A. and C. Co-infected CD14⁺ RNA was harvested following infections with a transwell system wherein HCMV was incubated with CD14⁺monocytes for 12 days before being transferred to the bottom compartment. The top compartment contained Calu-3 cells subsequently infected with SARS-CoV-2 for 3 days after combining the top and bottom transwells. To measure changes in TNF-α and IL-1β transcripts RT-qPCR was performed using TaqPath and predesigned primer/probes for cellular 7SK, TNF-α, and IL-1β. The ∆∆CT analysis was used to compare HI and live BA.1 to mock-infected transcripts normalized to 7SK. B. and D. The concentration of TNF-α and IL-1β was measured in the co-culture bottom compartment only supernatant by ELISA. The concentrations of the samples were interpolated and calculated from a standard curve of the indicated cytokine.

The supernatant was used to quantify IL-1β and TNF-α by ELISA. The samples were run in triplicate using the appropriate standards. The standard curve was generated from TNF-α (pg/mL) and IL-1β (pg/mL) following the manufacturer's instructions. ELISA interpolated summary data of TNF-α and IL-1β following 72-hour SARS-CoV-2 infection in latent HCMV cells (Fig. 6B and D, respectively). Taken together, the RT-qPCR and ELISA assays demonstrate an increase in cytokine transcript and cytokine secretion following SARS-CoV-2 co-infection of CD14⁺ cells.

3.4. Schematic for SARS-CoV-2 and HCMV co-infection model with RNA-sequencing defined transcriptomic changes

The goal of this study was to assess the transcriptomic changes in CD14⁺ monocytes during a co-infection with SARS-CoV-2 and HCMV (Fig. 7). We propose two potential ways that CD14⁺ cells with latent HCMV can become infected with SARS-CoV-2, first is from the hematogenous spread of SARS-CoV-2 from the lung environment to contact CD14⁺ cells in the circulatory system or ingress of HCMV infected monocytes into the lung environment results in co-infection. The second is infected monocytes entering the lung environment from the circulatory system where it encounters SARS-CoV-2 infected lung tissues. The significant cellular transcriptomic changes described in the prior results section are denoted in the diagram.

Fig. 7.

Schematic diagram of HCMV and SARS-CoV-2 interactions within host lung microenvironment.

Scenario 1. SARS-CoV-2 egresses from lung cells into the circulatory system where latent HCMV monocytes reside. SARS-CoV-2 enters monocytes containing HCMV and causes lytic reactivation of HCMV within the circulatory system. Scenario 2. Monocytes containing HCMV migrate to the lung environment from the circulatory system. SARS-CoV-2 infects monocytes, resulting in a co-infected cell. Differentiation of monocytes into macrophages induces HCMV lytic reactivation in the lung. Table of downregulated genes and inhibited upstream regulators (Red) and upregulated genes and activated upstream regulators (Yellow) of live SARS-CoV-2 BA.1 versus the mock control.

4. Discussion and conclusions

The severity of SARS-CoV-2 infections can differ significantly between individuals, and the underlying reasons for this variation are not fully understood. Our findings demonstrate that HCMV can enhance SARS-CoV-2 permissiveness in monocytes by increasing the expression of cellular ACE2. Individuals admitted to hospitals with COVID-19 exhibited a notable prevalence of secondary infections. Among the risk factors associated with secondary infections are early ICU necessity, respiratory failure, and severe lymphopenia (Ripa et al., 2021). HCMV is a highly prevalent virus, persisting throughout a person's lifetime. Advanced age has been recognized as a significant risk factor for the severity of COVID-19 disease, with mortality rates rising in each successive decade of life. Elderly individuals have the highest prevalence of HCMV, advanced age coupled with HCMV can contribute to immune senesce leading to a lack of antibody production following immunization, and decreased responses to vaccines (Wikby et al., 2005; Saurwein-Teissl et al., 2002; Trzonkowski et al., 2003). These risks combined with our data indicating immune system and inflammatory response disruption suggest that HCMV latent infection can adversely affect COVID-19 disease severity.

This study aimed to address what changes occur when CD14⁺ monocytes harboring latent HCMV are infected with SARS-CoV-2. We first tested the ability of SARS-CoV-2 variants to infect CD14⁺monocytes without the presence of HCMV. We found the Omicron BA.1 variant to have the highest copy number of SARS-CoV-2 nucleocapsid over the original Wuhan (WA.1) strain and Delta (AY.4) variant. Immunofluorescence assays in CD14⁺ cells co-infected with HCMV and SARS-CoV-2 showed the presence of HCMV IE1 and SARS-CoV-2 nucleocapsid proteins expression in the same cells, indicating a successful co-infection. Examining the levels of ACE2 in CD14⁺ monocytes showed an increase in ACE2 expression in cells infected with HCMV compared to mock-infected cells. This suggests the HCMV could increase cell permissiveness of SARS-CoV-2 through the increase in ACE2 expression. We then sought to compare the cellular transcriptome between HCMV-positive monocytes co-infected with heat-inactivated versus live SARS-CoV-2 virus. Using bulk RNA sequencing, we found that co-infection of CD14⁺ monocytes resulted in an increase in pro-inflammatory and a decrease in anti-inflammatory markers. Together, these responses indicate a possible cause for differences in cytokine storm and other severe COVID-19 pathologies.

The HCMV secretion profile is notably rich in IL-6, which plays a crucial role in promoting prolonged cell survival. The mechanism employed by HCMV to increase cell survival involves IL-6 receptor-mediated signaling, leading to survivin upregulation and suppression of apoptosis (Botto et al., 2011). Interestingly, high levels of IL-6 are associated with severe COVID-19 diseases, including cytokine storm and multi-organ failure (Majidpoor and Mortezaee, 2022). Our data shows significantly higher levels of IL-6 in samples with HCMV and SARS-CoV-2 present than in samples with HCMV alone or heat-inactivated SARS-CoV-2.

While we focused our analysis on immune factors, there were other interesting genes were regulated, including Spastic Paraplegia 7 (SPG7) which was downregulated. SPG7 encodes a mitochondrial metalloprotease protein involved in membrane trafficking, intracellular motility, organelle biogenesis, protein folding, and proteolysis. Interestingly, a new study discovered multiple SARS-CoV-2 viral proteins, including M protein, ORF3a, ORF9b and 9C, ORF10, and NSP6, bind to mitochondrial permeability transition pore complex proteins, including SPG7 and CCDC58 (Ramachandran et al., 2022).

One limitation of the study is that we performed RNA-sequencing and transcriptome analysis only in HCMV infected CD14⁺monocytes. In the future, we would like to include sequencing profiles of other relevant cell types that harbor latent HCMV, such as CD34⁺cells. Our current study used bulk RNA-seq for the transcriptome analysis, and we did not determine what percentage of CD14+ cells were co-infected, left uninfected, or had a single infection with either HCMV or SARS-CoV-2. Another limitation includes not being able to track infection or replication accurately. Our current system of a 48- or 72-hour SARS-CoV-2 infection allows for re-entry and additional replication. To address this in future studies, we plan to use a serum-blocking system. This system would allow for an initial one-hour infection of SARS-CoV-2, followed by blocking re-entry by serum-treating cells. This system could help accurately determine if one viral replication cycle has occurred.

Overall, our data demonstrate successful co-infection of SARS-CoV-2 and HCMV within monocytes and significant transcriptome profile differences between the mock-HCMV, heat-inactivated SARS-CoV-2 with HCMV, and live SARS-CoV-2 with HCMV within CD14⁺ cells. The data presented here points to potential causes of immune system dysregulation, which could lead to the increase in disease severity associated with HCMV status.

In summary, the severity of COVID-19 is influenced by various factors, and the presence of an additional latent pathogen, HCMV, could have a significant impact. Reactivation of HCMV during COVID-19 is a risk factor for higher mortality. In one study done by Pérez-Granda et al. (2023), they found 29.7 % of patients with a median age of 64 and 11.42 % HCMV seropositivity was admitted to the ICU. COVID-19 patients who were HCMV positive experience longer hospital stays and are more frequently admitted to the ICU and placed under mechanical ventilation (Pérez-Granda et al., 2023). These clinical cases and ongoing studies suggest a necessity for further research into the HCMV status of COVID-19 patients. Additional studies will help determine whether antiviral drugs targeting HCMV could alleviate the severity of COVID-19 in individuals with HCMV seropositivity. In HCMV-induced disease, TNF-α is produced by HCMV-specific T cells (Clement and Humphreys, 2019; Bolovan-Fritts et al., 2004). Interestingly, TNF-α and IL-1β induce immediate early (IE) gene transcription to promote reactivation of latent HCMV (Clement and Humphreys, 2019; Simon et al., 2005; Forte et al., 2018).

Author statement

We declare that we have no financial or personal relationships with others or organizations that could inappropriately influence our work. While preparing this work, the authors did not use any generative AI or AI-assisted technologies in the writing process. All content was written, reviewed, and edited manually by the authors, who take full responsibility for the content of the publication. We confirm that the work described in this manuscript has not been published previously and is not under consideration for publication elsewhere. All authors have approved the submission of this manuscript, and if accepted, it will not be published elsewhere in the same form, in English, or any other language, including electronically, without the written consent of the copyright holder.

Funding

This work was supported by departmental funds and the Nevada IDeA Network of Biomedical Research Excellence (NV-INBRE) pilot grant from the National Institute of General Medical Sciences (GM103440 and GM104944).

CRediT authorship contribution statement

Shannon Harger Payen: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation. Kabita Adhikari: Investigation, Data curation. Juli Petereit: Visualization, Software. Timsy Uppal: Supervision. Cyprian C. Rossetto: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition. Subhash C. Verma: Project administration, Data curation, Investigation, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no conflicts of interest with the contents of this article.

Data availability

• All sequences for the transcriptomic analysis generated for this study have been deposited into NCBI's Sequence Read Archive (SRA) and are accessible through the following accession number: PRJNA10765588.