Expression of HCMV-Encoded miRNA in Subjects Acutely Coinfected with HIV: Correlation with Inflammation and Immune Activation

Abstract

Human cytomegalovirus (HCMV) coinfection is associated with a faster HIV disease progression and adverse clinical outcomes. HCMV-encoded miRNA expression, in individuals acutely infected with HIV (AHI), compared to those with HCMV monoinfection, was investigated in relation to viral replication and inflammation/immune activation. Sixteen individuals with AHI coinfected with HCMV were analyzed at serodiagnosis (T0) and after 6 (T1) and 12 (T2) months of antiretroviral therapy initiated within one week from serodiagnosis. Fourteen HCMV monoinfected subjects were also studied. Plasma RNA was reverse-transcribed and amplified with a panel designed to detect 14 different HCMV-microRNAs (miRNAs). VEGF-A and IP-10 plasma levels were quantified using ELISA. Except for hcmv-miR-70-3p, detected in all subjects, hcmv-miR-UL112-3p, hcmv-miR-US25-1-5p, hcmv-miR-US25-2-3p, hcmv-miR-US4-5p, hcmv-miR-US5-1, hcmv-miR-US5-2-3p, hcmv-miR-UL36-3p, and hcmv-miR-UL36-5p were significantly more frequently detected when HCMV DNA was present (lytic infection). In latent HCMV infection, hcmv-miR-UL22A-5p and hcmv-miR-UL148D were more frequently observed in HIV/HCMV-coinfected individuals, compared to mono-HCMV infection. Hcmv-miR-UL22A-5p and hcmv-miR-US33-5p showed a direct correlation with HIV-1 RNA. Notable positive correlations between hcmv-miR-UL22A-5p and the interferon-gamma-inducible protein 10 (IP-10), as well as between hcmv-miR-UL148D and the vascular endothelial growth factor A (VEGF-A), were also observed. HCMV-miRNA expression varies between lytic and latent infection and differs in HIV coinfection. In HCMV/HIV coinfection, increased levels of hcmv-miR-UL148D, associated with VEGF-A production, seem to be less linked to HIV viremia with respect to hcmv-miR-UL22A-5p and hcmv-miR-US33-5p. A deeper understanding of HCMV-encoded miRNA biology may facilitate the comprehension of HCMV/HIV coinfection pathogenetic mechanisms.

Keywords: HCMV; HIV; VEGF; inflammation; miRNA.

Figure 1

HCMV miRNA detection rates in lytic and latent infections. The y axis displays the observed frequencies, and the x axis shows the fourteen HCMV-encoded miRNAs analyzed. Lytic infections are displayed in green, while latent infection is shown in purple. Significant differences in the miRNA frequencies between the two groups are reported with * p < 0.05 and ** p < 0.001.

Figure 2

HCMV miRNA detection rates in subjects with HCMV only and those with HIV coinfection. The y axis displays the percentages of the observed frequencies; the x axis shows the fourteen HCMV-encoded miRNAs analyzed. HCMV monoinfections are represented in blue, while HCMV/HIV coinfections are in orange. Significant differences in the miRNA frequencies between the two groups are reported with * p < 0.05.

Figure 3

The relationship between the expression of HCMV-encoded miRNAs, which are commonly observed during lytic infection, and VEGF-A plasma levels. The correlation between the hcmv-miR-US25-1-5p expression levels and VEGF-A plasma concentrations in HCMV lytic infection (Panel A). The correlation between the hcmv-miR-US4-5p expression levels and VEGF-A plasma concentrations in HCMV lytic infection (Panel B). The correlation between hcmv-miR-US5-1 and VEGF-A plasma concentrations in HCMV lytic infection (Panel C). In all panels, the solid line corresponds to the linear regression line fitted to the data, while the dotted lines represent the 95% confidence interval of the regression line. Moreover, the varying hcmv-miR expression levels are presented as ΔCt, normalized to the amplification of the endogenously expressed hsa-miR-6090.

Figure 4

Correlation between HCMV-encoded miRNAs and HIV viremia in HIV-coinfected subjects. Correlation between hcmv-miR-UL22A-5p and HIV-1 RNA (Panel A). Correlation between hcmv-miR-US33-5p and HIV-1 RNA (Panel B). In all panels, the solid line corresponds to the linear regression line fitted to the data, while the dotted lines represent the 95% confidence interval of the regression line. Moreover, varying hcmv-miR expression levels are presented as ΔCt, normalized to amplification of endogenously expressed hsa-miR-6090.

Figure 5

Correlation between HCMV-encoded miRNAs and inflammatory cytokine plasma levels in latently HCMV/HIV-coinfected subjects. Correlations between hcmv-miR-UL22A-5p and IP-10 plasma levels (Panel A). Correlations between hcmv-miR-UL148D and VEGF-A plasma levels (Panel B). In all panels, the solid line corresponds to the linear regression line fitted to the data, while the dotted lines represent the 95% confidence interval of the regression line. Moreover, varying hcmv-miR expression levels are presented as ΔCt, normalized to amplification of endogenously expressed hsa-miR-6090.

Figure 6

Diagram describing analytical workflow. Designed with Biorender (https://www.biorender.com/, last accessed on 6 June 2025).