Abstract Background Since the breakout of COVID-19, the mutated forms of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have shown enhanced rates of transmission and adaptation to humans. The variants of concern (VOC), designated Alpha, Beta, Gamma, Delta, and Omicron emerged independent of one another, and in turn rapidly became dominant. The success of each VOC, as well as the virus fitness, were enabled by altered intrinsic functional properties and, reasonably, to virus antigenicity changes, conferring the ability to evade a primed immune response. Methods We analysed the gene expression profiles of monocyte-derived macrophages (MDM) isolated from whole blood of healthy participants exposed to the 5 different SARS-CoV-2 VOC: D614G, Alpha (B.1.1.7), Gamma (P1), Delta (B.1.617.2), and Omicron BA.1 (B.1.1.529), and to the HCoV-OC43 strain, a coronavirus already present in the population before the SARS-CoV-2 pandemic. Whole transcriptome RNA-Seq, for both coding and non-coding RNAs, was then made. Results After exposure to the 5 VOC of MDM, we initially assessed the presence of the viral SARS-CoV-2 transcripts to confirm viral entry. We then analysed the RNA-Seq data and observed a significant deregulation of both coding and non-coding RNAs. In particular, our RNA-Seq analysis showed a significant up-regulation of several genes involved in different immunological processes, such as PARP9/PARP14 axes, in macrophages exposed to D614G, Alpha, and Gamma variants. Surprisingly, our data showed that macrophages exposed to the Delta variant exhibited a transcriptional profile more similar to the naïve control group, while macrophages exposed to the Omicron variant showed intermediate differentially expressed genes (DEGs) between the two groups. By checking the canonical markers for M1/M2 differentiation states, we did not observe any expression in macrophages exposed to the Delta variant, suggesting an M0 status, comparable to the naïve control group. Finally, we observed a significant deregulation of 3 main types of non-coding RNAs (ncRNAs): long non-coding RNAs (lncRNAs), microRNAs (miRNAs), and small nucleolar RNAs (snoRNAs), some of which are common to coronaviruses, and some specific to SARS-CoV-2. Conclusion The SARS-CoV-2-dependent alteration of the transcriptome of monocyte-derived macrophage (MDM)-infected cells can be linked to the chronological order of the variants’ appearance in the human population. Our data suggest an evolution of VOC in modulating the host immune response, with a strong change in pace beginning with the advent of the Delta variant. MDMs exposed to Delta showed a failure in the activation of the adaptive immune response, and this correlates with the more severe symptoms developed by people affected with this SARS-CoV-2 variant. Supplementary Information The online version contains supplementary material available at 10.1186/s12967-025-06158-2. Introduction The global pandemic of coronavirus disease 2019 (COVID-19), declared in March 2020, and caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has been drawing the attention of scientific and political communities worldwide. The reasons reside in the illness caused, and in the subsequent devastating effects on the world’s population, resulting in more than 7 million deaths ([48]https://www.who.int/emergencies/diseases/novel-coronavirus-2019). In particular, older adults aged ≥ 60 with medical comorbidities such as obesity, cardiovascular disease, and diabetes, and immunocompromised transplanted patients with solid organ or hematopoietic stem cells were considered at the greatest risk of becoming very ill [[49]1]. Coronaviruses are enveloped, positive-sense single-stranded RNA viruses widespread among birds and mammals, but have only been associated with human diseases capable of affecting the respiratory system [[50]2]. SARS-CoV-2 has been shown to be prone to genetic evolution, developing mutations over time and adapting to the human hosts. In particular, genetic variations in the receptor-binding domain (RBD) of the spike (S) glycoprotein result in facilitating viral entry into the host cell, and have implications for its pathogenesis. Among multiple variants of SARS-CoV-2, only a few are considered variants of concern (VOC), because they are associated with enhanced transmissibility or virulence, the ability to evade detection or decrease the effectiveness of therapeutics or vaccination, thus having a higher impact on public health. The virus entry into the host human cells is mediated by the binding of the RBD to a cell surface receptor angiotensin-converting enzyme-2 (ACE2) followed by the fusion with the host cell membrane through a proteolytic cleavage of the S protein, operated by transmembrane serine protease 2 (TMPRSS2) or by cathepsin B or L (CTS-B or–L) [[51]3]. Ciliated nasal secretory epithelial cells of the upper respiratory tract and pneumocytes of the lower respiratory tract, together with bronchial and endothelial cells of the pulmonary vasculature, co-express ACE2 and TMPRSS2 proteins [[52]4]. The initial role of resident macrophages followed by the recruitment of peripheral macrophages and other immune cells are crucial in severe lung inflammation due to SARS-CoV-2 infection [[53]5]. Macrophages are myeloid antigen-presenting cells that differentiate from circulating monocytes, and play a crucial role in activation of the host innate immune defence. In response to different signals, macrophages have the capacity to differentiate in classically-activated macrophages (M1) when stimulated by TLR ligands and IFN-γ or, alternatively, activated macrophages (M2) when stimulated by IL-4 and IL-13. The M1 phenotype is characterized by the expression of high levels of pro-inflammatory cytokines, elevated production of reactive nitrogen and oxygen species, and promotion of Th1 response with a robust microbicide activity. On the other hand, M2 macrophages are implicated mostly in parasite infections and in tissue remodelling and tumor progression. Macrophages have a variety of specialized functions mediated by the production of cytokines (IL-6, IL-1β, TNF-α, IL-12) and chemokines (CXCL8), aiding the recruitment of various inflammatory cells, such as neutrophils, monocytes, and natural killer cells to build a powerful innate immune response [[54]6, [55]7]. Thus, they play a crucial role in the development of a solid immune response, though it has also been reported that these cells can have a preservative role, and aid the dissemination of different pathogens. Some examples described in the literature concern the persistence of Mycobacterium tuberculosis in alveolar macrophages, as well as macrophages allowing the replication of the human cytomegalovirus and the MERS coronavirus within them. This evidence has prompted the scientific community to formulate the hypothesis that macrophages can act as “Trojan horses,” internalizing pathogens within them, allowing their survival and spreading to other tissues [[56]8–[57]10]. Starting from this hypothesis, we decided to study their behaviour towards SARS-CoV-2 and its variants. We found that macrophages can internalize SARS-CoV-2, acting as passive carriers for the virus. Even if no complete replication is sustained in these cells, the virus is kept viable, and infection of permissive cells can occur upon co-culturing [[58]11]. Our previous results suggest that macrophages may aid in SARS-CoV-2 persistence and dissemination throughout the body, though the mechanisms behind this behaviour have yet to be elucidated. Thus, we decided to analyse the transcriptional landscape, both coding and non-coding RNAs, of macrophages after exposure to SARS-CoV-2 major VOC to identify the pathways altered by exposure to these VOC, and better understand macrophage behaviour in SARS-CoV-2 pathogenesis. Several gene expression studies have been done to shed light on the mechanisms by which SARS-CoV-2 alter normal host cell metabolism and signalling for its own fitness [[59]12, [60]13]. Little is known about ncRNAs, both long and short, but recent evidence has described how lncRNAs play a crucial role in the SARS-CoV-2 life cycle, as well as in the host immune response [[61]14–[62]16]. Materials and methods SARS-CoV-2 variants Experienced personnel carried out all procedures of SARS-COV-2 isolation, propagation, titration, and infection in a Biosafety Level 3 (BSL3) laboratory. All SARS-CoV-2 strains, Italian Reference (D614G, B.1.1), Alpha variant (B.1.1.7) Gamma (P1), Delta (B.1.617.2), and Omicron BA.1 (B.1.1.529) were isolated from nasopharyngeal swabs, and propagated on the VERO C1008 (Vero 76, clone E6, Vero E6; ATCC CRL-1586™) cell line. Complete genome sequencing [[63]17] was done in order to confirm the presence of variant-defining mutations, and sequences were submitted to GISAID [[64]15]. Virus variants were titrated on VERO E6 to prepare cell-free virus for monocyte-derived macrophages (MDM) infection. The titer of each variant was measured as a 50% tissue culture infectious dose (TCID50) in six replicas in a 96-well flat-bottom tissue-culture microtiter plate. Logarithmic dilution of stock virus, starting with 1:10 to 1:10 − 8 in the presence of 3 × 10^4 VERO E6 cells were incubated for 72 h at 33 °C in 5% CO2. The cells were observed for CPE and stained with Gram’s crystal violet solution (Merck, Darmstadt, Germany) plus 5% v/v formaldehyde 40% m/v (Carlo Erba SpA, Milan, Italy). The titer was calculated as 50% of the tissue culture infectious dose (TCID50) per mL using the Reed–Muench method, and expressed as its logarithmic value [[65]18]. Cell cultures and macrophage differentiation VERO E6 cells were cultured in fresh Eagle’s Minimum Essential Medium (EMEM, Lonza Group Ltd, Basel, Switzerland) supplemented with 1% v/v penicillin, streptomycin and glutamine (Euroclone SpA, Milan, Italy), and 10% or 2% v/v Fetal Bovine Serum (FBS; Euroclone SpA) for maintenance. Immune cells were cultured in RPMI 1640 (Lonza Group Ltd) supplemented with 1% v/v penicillin, streptomycin glutamine, 10% v/v FBS, and 800U/mL Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF; Thermo Fisher, Waltham, MA, USA). Peripheral blood mononuclear cells (PBMC) were isolated from eight buffy-coat units from healthy blood donors, given by the Blood Bank Processing and Validation Centre DMTE/SIMT Fondazione IRCCS Policlinico San Matteo according to the Italian law decree of the Ministry of Health, November 2, 2015 “Provisions relating to the quality and safety requirements of blood and blood components,” using Ficoll-Paque Plus™ (GE Healthcare, Chicago, IL, USA). PBMC were plated at a concentration of 5 × 10^6 cells/well of 24-well microplates (COSTAR, Corning Incorporated, NY, USA) for 3 hours at 37 C° 5% CO[2] to let the cells adhere in RPMI 1640 w/o FBS. Non-adherent cells were removed, and macrophages were differentiated by stimulation with GM-CSF 800 U/mL in RPMI 1640 w/ 10% v/v FBS for 7 days (a schematic representation is shown in Supplementary Fig. [66]1A). Macrophages were also grown on round glass slides for cell staining after virus infection. Time course experiments Macrophages were grown on both glass slides and in a 24-well microplate. After 7 days of differentiation, macrophages were then exposed to 10000TCID[50]/mL SARS-CoV-2 for 2 h at 33 °C, in a 5% CO[2] atmosphere. After removal of virus inoculum, cells were washed 3 times in PBS 1X, and maintained in RPMI 1640 w/o FBS with the addition of 1% penicillin, streptomycin, glutamine and GM-CSF 800U/mL. At each time point (24, 48, 72, and 96 h), supernatants were titrated on VERO E6 cells to detect complete virus replication with production of infectious virus. As a control for viral propagation and persistence, VERO E6 were infected in the same conditions, and the same experiments were done. Immunofluorescence assay Macrophages grown on glass slides were exposed to viable SARS-CoV-2, and at 24 and 96 h fixed in methanol/acetone 1:2 for 5 min, before staining with SARS-CoV-2 nucleoprotein/NP antibody, rabbit mAb (Sino Biological, Beijing, China) and anti-spike antibody, mouse mAb (Sino Biological), followed by donkey anti-rabbit Alexa Fluor 488 (Thermo Fisher) and a goat antimouse rhodamine (Chemicon, Tokyo, Japan) secondary antibodies, respectively. Nuclei were stained with DAPI (Sigma-Aldrich, MO, USA), and slides were examined at 20X and 40X with a Zeiss Axioplan 2 Imaging Microscope (ZEISS, Oberkochen, Germany) for the expression of nucleoprotein and spike protein. The antibodies for anti-nucleoprotein and anti-spike protein were reactive to all VOC, with the exception of Gamma. To assess the presence of SARS-CoV-2 viral particles after exposition to Gamma, human anti-SARS-CoV-2 serum and fluorescent anti-human γ-immunoglobulin (anti-hIgG) as secondary antibody were used. This serum was previously described as highly reactive, with a neutralizing titer against Gamma and anti-SARS-CoV-2 total γ-immunoglobulin (hIgG) levels over the limit of detection (> 1:640 and > 2080 BAU/mL, respectively) [[67]19]. After coating with the serum and washing of the glass slides, anti-hIgG FITC conjugated antibody (CAPPEL, Thomas Scientific, PA, USA) was used to stain the cells, and DAPI was used to stain the nuclei. Macrophage exposition to viable SARS-CoV2 variants Macrophages were exposed to 200uL of 10000TCID50/mL viable SARS-CoV2, for 2 h at 33 °C in a 5% CO[2] atmosphere. After exposure, the inoculum was removed, and fresh RPMI 1640 w/o FBS was added. After 24 h of incubation, the supernatant was discarded, and cells were treated with 700uL TRIzol (Thermo Fisher) and immediately stored at -80 °C until use for RNA extraction. At the same time, SARS-CoV-2 not exposed macrophages were treated at the same conditions, minus exposition to SARS-CoV-2, and used as the naïve control group for all experiments. RNA extraction and next-generation sequencing Total RNA was extracted using the RNeasy Mini kit (QIAGEN, Hilden, Germany). RNA integrity was evaluated with the Agilent 4200 TapeStation System (Agilent Technologies Ltd, Santa Clara, CA, USA). RNA-Seq libraries were prepared from 1 µg of RNA using Illumina TruSeq Stranded Total RNA (Illumina, San Diego, CA, USA) and TruSeq RNA UD Indexes 24 Indexes-96 Samples (Illumina). Briefly, total RNA input was deprived of ribosomal RNA, and fragmented before the synthesis of single and double strand cDNA. Then, double strand cDNA 3’ends were adenylated and Illumina indexing adapters were ligated. Library size distribution and quality was evaluated on the Agilent 4200 TapeStation System (Agilent Technologies Ltd). Finally, libraries were pooled in equimolar concentration based on Qubit (Invitrogen, Waltham, MA, USA) quantification, and sequencing was performed on a NextSeq™ 550 (Illumina) with 2 × 76 cycles, following the manufacturer’s instructions. Transcript assembly, sequence alignment, and quantification Quality control of the raw sequencing data was done using FastQC (v0.11.9, Babraham Institute, Babraham, UK). Subsequently, Trimmomatic (v0.39) [[68]20] was employed to eliminate low-quality reads and conduct adapter trimming. The extracted RNA sequences were aligned with the human reference genome (hg19) using STAR (v2.7.0) ([69]https://github.com/alexdobin/STAR/releases) [[70]21]. Transcript abundances were quantified with RSEM (v1.3.3) ([71]https://deweylab.github.io/RSEM/) [[72]22], and gene expression levels were normalized to transcripts per kilobase million (TPM) to account for differences in transcript lengths and sequencing depth across samples. Transcripts that did not reach a threshold of 10 normalized read counts after averaging across all the biological replicates were not considered for the analysis. Screening DEGs For differential expression analysis, gene expression values were transformed by applying the base-2 logarithm to the normalized TPM values, with 1 added to avoid taking the logarithm of zero, i.e., log2(TPM + 1). To identify differentially-expressed genes (DEGs), a Student’s t-test was done, with a significance threshold set at a p-value of < 0.05. The test was applied for each pair of conditions only when at least 3 out of 4 samples exhibited log2(TPM + 1) > 1, ensuring that genes with low expression across most samples were excluded from the analysis in order to minimize the likelihood of false discoveries. Data analysis was conducted in R (v4.3.2) and Python (v3.8.10). An average linkage algorithm was employed for hierarchical clustering of the filtered DEGs after applying a z-score transformation, using Euclidean distance as the distance metric; the results were visualized through a heatmap. Additionally, principal component analysis (PCA) was done on the filtered dataset. The transformed expression levels were then visualized using boxplots, such that the central line represents the mean value, the hinges correspond to the standard deviation (mean ± standard deviation), and the notches indicate the minimum and maximum values for each condition. Venn diagrams were used to analyze the overlap of differentially expressed transcripts among comparisons. Volcano plots were used to visualize down- and up-regulated DEGs for each variant. Functional enrichment analysis For functional gene annotation, the obtained sequences were annotated by searching various databases, including Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG). These deregulated genes were subjected to GO and KEGG significant enrichment analyses to identify the biological functions and related metabolic pathways in which these genes are involved. Results Analysis of macrophages exposed to different SARS-CoV-2 variants The aim of our study was to evaluate the effects of the internalization of different SARS-CoV-2 strains on monocyte-derived macrophages differentiated from PBMC in immunocompetent participants. To do so, we differentiated macrophages from isolated monocytes, exposed them to different SARS-CoV-2 strains (Fig. [73]1), and did a comprehensive analysis of the transcriptome profiles after viral exposure. We focused our attention on the comparison between the Italian Reference strain (D614G) and different variants that have been described in recent years as variants of concern (VOC) given their impact on human health (World Health Organization [74]https://www.who.int/en/activities/tracking-SARS-CoV-2-variants). The variants taken into consideration have been associated with a higher transmission rate and pathogenicity, a lower response to vaccination and natural immunity, and an increased severity of the disease. In particular, B.1.1.7 (Alpha), B.1.617.2 (Delta), P.1 (Gamma) and BA.1 (Omicron) VOC were considered. The HCoV-OC43 strain, instead, was added to our study as a coronavirus present in the population before the SARS-CoV-2 pandemic. In this study, from here on, we will refer to the particular strain of each SARS-CoV-2 variant as follows: Italian Reference strain as D614G; B.1.1.7 as Alpha; B.1.617.2 as Delta; P.1 as Gamma, and BA.1 as Omicron. Fig. 1. [75]Fig. 1 [76]Open in a new tab General workflow used in this study to characterize macrophages’ response to SARS-CoV2 exposure. Monocytes were isolated from PBMC collected from a group of immunocompetent human participants and were differentiated into macrophages. From macrophages exposed to SARS-CoV2 VOC, total RNA was extracted and used for preparation of libraries and whole transcriptome RNA-Seq analyses. This image was created using BioRender ([77]https://www.biorender.com/) After exposure of macrophages to 10000TCID50/mL cell-free viable SARS-CoV-2 VOC, the ability of producing viable viruses was assessed through titration onto VERO E6 cells and RNA quantification in both the supernatants and intracellular extracts. Each assay was repeated using VERO E6 cells as a positive control of viral replication. Titration curves of VERO E6 cells and macrophages are reported in Fig. [78]2. Fig. 2. [79]Fig. 2 [80]Open in a new tab Titration curves of VERO E6 cells and macrophages exposed to SARS-CoV-2 VOC. Tissue culture infective dose (TCID50) was determined for each variant in macrophages and VERO E6 cells used as positive control. Data show that all VOC were able to replicate in VERO E6 cells, while in macrophages viral proliferation was not detected. The TCID50 was calculated using the Reed–Muench method in 6 replicates, and expressed as the logarithm of virus titer Similarly to what has previously been described in macrophages exposed to D614G [[81]11], all SARS-CoV-2 VOC cultured on VERO E6 used in this study produced visible cytopathic effects (CPE) as shown in Supplementary Fig. [82]1B. The different VOC showed similar trends, with viral peaks at 72 h, except for the Gamma variant, which reached plateau 48 h after inoculum. Conversely, no CPE was observed after inoculation of supernatants from macrophages exposed to SARS-CoV-2 VOCs at each time-point. To evaluate viral entry in macrophages, the presence of both nucleoprotein and spike protein was assessed by immunofluorescence using either specific anti-SARS-CoV-2 protein antibodies or by indirect staining with reactive human serum (Fig. [83]3). Fig. 3. [84]Fig. 3 [85]Open in a new tab Nucleoprotein and Spike protein expression in macrophages exposed to SARS-CoV2 VOC. Representative immunofluorescence images of cells stained with the anti-nucleoprotein (green) and anti-spike protein (red). Nuclei were stained with DAPI (blue). Images were taken at 20X magnification for the nucleoprotein (left) and at 40X magnification for the spike protein (right). The nucleoprotein was detectable starting at 24 h (A), while the spike protein was detectable only after 96 h at higher magnification (B). The scale bar is reported in D614G exposed macrophages. (C) Fixed Gamma-exposed macrophages were coated with SARS-CoV-2 highly reactive human serum, and then stained with FITC conjugated (green) anti-hIgG antibody; nuclei were stained with DAPI (blue). Magnification 20X. The presence of viral proteins is identified by the green fluorescence. The scale bar is reported in the 24 h post-exposition panel Results of the immunofluorescence assay show a kinetics of viral internalization similar for all the VOC analyzed. Starting at 24 h post-infection, the nucleoprotein is identifiable, and its presence increases in a time-dependent manner throughout the entire time-course assay (Fig. [86]3A). For the spike protein, we observed a lagging in the identification, with the first detectable appearance 96 h post-infection and at low density, needing higher magnification to be detected (Fig. [87]3B). As for the Gamma variant, even if indirectly determined, we can observe a similar time-dependent trend of viral particle identification. The γ-immunoglobulins bound to Gamma-exposed macrophages increase throughout the time course analysis, similar to what was observed for the other variants (Fig. [88]3C). To assess whether the SARS-CoV-2 VOCs internalized in the macrophages were still viable, and thus capable of infecting permissive cells, macrophages were co-cultured overnight with VERO E6 infected with each VOC, and were then cleared from infected VERO E6 by means of migration through a 5 μm filter, as previously described [[89]11] (Supplementary Fig. [90]1C). CPE was detected at 72 h post co-cultivation (PCC) on VERO E6, and increased to complete cell detachment at 120 h PCC, with a delayed gap of 24 h with respect to what has been previously reported for D614G (Supplementary Fig. [91]2). Considering that no viable virus was released in the supernatant of SARS-CoV-2-exposed macrophages, the presence of infection in co-cultivation with permissive cells demonstrates, for all the VOC used, the direct transfer of viral particles from the macrophages. This is consonant with what has previously been described with D614G [[92]8], and further consolidates the hypothesis of macrophages as “Trojan horses” for SARS-CoV-2. We then assessed the internalization efficiency of all the VOC used by quantifying the viral SARS-CoV-2 transcripts in intracellular RNAs extracted from exposed macrophages at 24 h. We checked the expression of the most conserved open reading frames (ORFs) of SARS-CoV-2, and in particular ORF1a and ORF1b located at the 5′ end of the SARS-CoV-2 genome; ORF3a, ORF6, ORF7a, ORF7b, ORF8, and ORF10, considered species-specific since they are present only in some strains [[93]23]; S (spike protein); E (envelope); M (membrane) and N (nucleocapsid). Looking at the viral counts, Delta and Omicron, in their active form, showed up to a 100-yield of viral particles higher compared to D614G, Alpha, and Gamma active variants (Table [94]1), thus confirming the presence of the viral RNA genome. Table 1. SARS-CoV-2 VOC transcripts counts VOC CTRL D614G Alpha Delta Gamma Omicron ORF1ab 0 7894 299 307,053 4940 17,995 ORF1ab 0 7 14 4 5 6.5 S 0 1419 52 49,558 888 31,108 N 0 698 28 15,510 468 14,193 ORF3a 2 242 9.75 7879 159 5391 M 0 256 10.7 7872 168 5380 ORF8 0 113 4 2947 63.5 1987 ORF7a 0 88 7.75 2090 71 1727 E 0 30 2 1034 17 577. ORF6 0 22 0.5 519.5 5.5 320 ORF7b 0 2 0 39.25 3 44 ORF10 0 6.5 0.25 52.25 4.25 93 [95]Open in a new tab Transcript counts of the twelve ORFs for each sample of the five variants, D614G, Alpha, Delta, Gamma, and Omicron Differential expression analysis of transcriptome profiles driven by SARS-CoV-2 VOC Once the efficiency of the viral internalization was confirmed, we wanted to characterize the transcriptomic profiles of the monocyte-derived macrophages exposed to all SARS-CoV-2 VOC, and to the OC43 coronavirus strain. All of them were then compared with the naïve control group. Transcripts that did not reach a threshold of 10 normalized read counts after averaging across all the biological replicates were not considered for the analysis. Our data showed that only 16’353 transcripts of the whole dataset reached the threshold of 10 normalized read counts out of the 57’820 total transcripts (Supplementary Tables [96]1 and [97]2). Heat map analysis of the differentially expressed genes clearly showed that the transcriptome profile of the host was markedly different in participants’ macrophages exposed to D614G, Alpha, and Gamma variants compared to the naïve control group, while Omicron and Delta variants clustered with the control group (Fig. [98]4A). Notably, the OC43 coronavirus strain showed a transcriptomic pattern similar to SARS-CoV-2 D614G, Alpha, and Gamma variants. Fig. 4. [99]Fig. 4 [100]Open in a new tab Heat map, PCA, Venn diagrams, volcano plots and bubble charts of the identified genes in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron SARS-COV2 variants. (A-B) Heat map and PCA analyses of the 16’353 genes identified in the macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants and in control samples. C-D) Venn diagrams showing the total number of significantly up-regulated (C) and down-regulated (D) genes (p-value ≤ 0.05) across macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. Venn diagrams were generated using the web-tool designed and maintained by A. Saurin ([101]https://www.biotools.fr/misc/venny). E) Volcano plots showing the differentially expressed genes (DEGs, p < 0.05) in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. Red dots represent up-regulated genes, and blue dots the down-regulated genes. F) Bubble chart for KEGG pathway enrichment analysis showing 10 immunological up-regulated datasets across the different conditions. Combined scores are shown by the circle area, while the circle color represents the range of the adjusted p-value. G) Bubble chart for Reactome pathway enrichment analysis showing 10 immunological up-regulated datasets across the different conditions. Combined scores are shown by the circle area, while the circle color represents the range of the adjusted p-value. This image was created using BioRender ([102]https://www.biorender.com/) We also evaluated the similarity among the samples through principal-component analysis (PCA), modelling the overall transcriptome expression. The variance explained by the first two components of PCA (42.01 and 12.74%) confirmed significant differences in the distribution across the VOC and the control group, and clearly highlighted the overlap in gene expression profiles between Delta and the control group (Fig. [103]4B). Venn analysis revealed that 176 DEGs were common to all examined conditions, while no common genes were down-regulated among the different conditions (Fig. [104]4C and D). The largest commonality in up-regulated DEGs was seen when comparing the D614G-strain-exposed macrophages with the SARS-CoV-2 Alpha-variant-exposed macrophages, while the SARS-CoV-2 Omicron-strain-exposed macrophages showed the less similar DEG pattern, which correlates with the observation that only few genes were differentially expressed compared to the control, and were indeed perturbed by the presence of the virus variant. We next determined the differences in the transcriptomic profiles between each variant and the control group. As shown in Fig. [105]4E, the vast majority of the transcripts were generally up-regulated, rather than down-regulated. In the OC43 coronavirus-strain-exposed macrophages, 2885 genes were significantly (p < 0.05) up-regulated and 65 down-regulated; SARS-CoV-2 D614G-strain-exposed macrophages showed 2004 significantly up-regulated genes, and 44 down-regulated genes; SARS-CoV-2 Alpha-variant-exposed macrophages 2288 and 46 significantly deregulated genes, up-regulated and down-regulated, respectively; SARS-CoV-2 Gamma-strain-exposed macrophages showed 2530 significantly up-regulated genes and 44 down-regulated genes. The different transcriptional trends of the Delta and Omicron variants observed with both the hierarchical clustering and the PCA analysis, were also confirmed with a direct comparison to the control group. In fact, we found only 8 significantly up-regulated and 6 down-regulated genes in the SARS-CoV-2 Delta-strain-exposed macrophages and 558 significantly up-regulated and 18 down-regulated genes in the SARS-CoV-2 Omicron-strain-exposed macrophages. Biological processes and pathways regulated by DEGs (both up-regulated and down-regulated genes) were identified based on KEGG and Reactome databases for all samples (Supplementary Figs. [106]3 and [107]4). We then compared different key immunological processes showing top counts and high adjusted p-values across the samples (Fig. [108]4F and G). Identification of deregulated genes for the different SARS-CoV-2 VOC Intrigued by the striking different transcriptional profile of the Delta and Omicron variants, compared to the other VOC and the naïve control, we wanted to go deeper in our analysis by checking the expression of some genes that play a key role in the immune response. Interestingly, some of the most differentially expressed genes in our samples are involved in a wide range of immunological events, at the transcriptional and cellular levels. Among them, we found highly differentially expressed PARP9 and PARP14, members of the PARP superfamily. Both PARP9 and PARP14 mRNA levels were strongly and significantly increased in macrophages exposed to the OC43 coronavirus strain, SARS-CoV-2 D614G, Alpha and Gamma variants, compared to the control group. Delta showed very low mRNA levels of PARP9, its ligand DTX3L, and the downstream transcription factor STAT1, while the SARS-CoV-2 Omicron variant showed lower mRNA levels of these factors compared to the other VOC (Fig. [109]5). Fig. 5. [110]Fig. 5 [111]Open in a new tab Relative gene expression of transcripts related to PARP9 and PARP14 pathways in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. A) Box plots showing the statistically differentially expressed genes of the PARP9 axis in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. B) Box plots showing the statistically differentially expressed genes of the PARP14 axis in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. Significance was calculated using Student t-tests. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. This image was created using BioRender ([112]https://www.biorender.com/) PARP14, STAT6, and interferon regulatory factor 3 (IRF3) mRNAs showed a similar expression profile among the different VOC and compared to the control, similar to the PARP9/DTXL3/STAT1 axis. Since it is crucial for an immunological response that macrophages became polarized toward the M1 or M2 subgroups, we wanted to assess the maturation of the macrophages exposed to each variant and the control group, analyzing the canonical markers for M1/M2. In particular, we analyzed the expression of CD80, CD86, FCGR1A (CD11), and FCGR2A (CD32) as M1 markers, while ARG1, VEGF, TGF-β, and CD206 were checked as M2 markers. In Fig. [113]6, we reported the statistically significant M1 and M2 markers up-regulated in SARS-CoV-2 VOCs, OC43 and SARS-CoV-2 VOCs D614G, Alpha, and Gamma. Fig. 6. [114]Fig. 6 [115]Open in a new tab Relative gene expression of M1 and M2 representative markers in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. In the upper panel are shown box plots of the statistically significant M1 representative markers: FCGR1A, ITGAM, FCGR2A and CD86 genes in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. In the lower panel are shown box plots of the statistically significant M2 representative markers: CD163, MRC1, TGFB1 and VEGFA in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. Significance was calculated using Student t-tests. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. This image was created using BioRender ([116]https://www.biorender.com/) Conversely, the expression of both M1 and M2 macrophage markers was absent in macrophages infected with SARS-CoV-2 Delta, thus suggesting an M0 status, comparable to the naïve control group. Similar to Delta exposure, macrophages exposed to the SARS-CoV-2 Omicron variant showed a mild expression for some of the M1/M2 markers, such as FCGR1A, ITGAM, and MRC1. Other M1/M2 markers were, instead, up-regulated by Omicron to the same extent observed in response to the other VOC. Non-coding RNAs in response to infection of SARS-CoV-2 variants Similar to protein-coding RNA transcripts, several non-coding RNAs (ncRNAs) are differentially regulated upon infection by the different SARS-CoV-2 variants. We observed a significant deregulation of 3 main types of ncRNAs: long non-coding RNAs (lncRNAs), microRNAs (miRNAs), and small nucleolar RNA (snoRNAs). Compared to naïve control group, we observed the deregulated expression of several lncRNAs, such as MALAT1, NEAT1, ITGB2-AS1, SETD5-AS1, and SH3BP5-AS1, as previously reported [[117]16]. These transcripts were generically up-regulated upon coronavirus infection, as evidenced by their increased expression also in response to OC43 infection (Fig. [118]7). Fig. 7. [119]Fig. 7 [120]Open in a new tab Relative gene expression of lncRNAs in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. Box plots showing the statistically differentially expression of the lncRNAs, MALAT1, NEAT1, ITGB2_AS1, SETD5_AS1 and SH3BP5_AS1 in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. Significance was calculated using Student t-tests. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. This image was created using BioRender ([121]https://www.biorender.com/) Other lncRNAs were, instead, more up-regulated with specific SARS-CoV-2 variants, such as HLA-AS1 for Gamma, and TTTY15 for D614G, Gamma and Alpha. Curiously, Delta infection did not promote any transcription of lncRNAs (Supplementary Table [122]3). [123]AL132780.1 was the only lncRNA found significantly down-regulated in response to OC43, Omicron, Alpha, and Delta (Supplementary Table [124]3). MiRNAs were also deregulated upon SARS-CoV-2 infection, as previously reported [[125]14, [126]16, [127]24]. Delta and Omicron exposure down-regulated the expression of miR-3648. Conversely, miR-142 and miR-3916 were significantly up-regulated by all strains in the study, except for Delta and Omicron (Fig. [128]8A). MiR-941-3, on the other hand, was significantly up-regulated in response after Omicron exposure (Fig. [129]8A). Fig. 8. [130]Fig. 8 [131]Open in a new tab Relative gene expression of miRNAs and snoRNA in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants A) Relative gene expression of miRNAs in macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. B) Venn diagrams showing the total number of significantly deregulated snoRNAs (p-value ≤ 0.05) across macrophages exposed to OC43, D614G, Alpha, Delta, Gamma, and Omicron variants. The Venn diagram was generated using the web-tool designed and maintained by A. Saurin ([132]https://www.biotools.fr/misc/venny). This image was created using BioRender ([133]https://www.biorender.com/) We found that all SARS-CoV-2 variants generally modulated three classes of snoRNAs: H/ACA box small nucleolar RNA A (snoRA), C/D box small nucleolar RNA (snoRD) and small Cajal body-specific RNAs (scaRNAs) (Fig. [134]8B). In our dataset, we observed a significant up-regulation of 7 snoRNAs (snoRA2A, snoRA3, snoRA48, snoRA62, snoRA63, snoRA80 and snoRD89) in response to macrophage exposure to all SARS-CoV-2 variants but Delta and Omicron. Among them, 4 snoRNAs (snoRA2A, snoRA3, snoRA63, and snoRD89) were also significantly up-regulated by OC43 exposure, thus suggesting a common mechanism in response to coronaviruses. Only snorD67 was found significantly up-regulated by any coronavirus exposure, except for Omicron (Supplementary Table [135]3). Alpha was the variant leading to the highest number of upregulated snoRNAs (14 snoRAs and 12 snoRD), while it down-regulated only snoRA70 and snoRA72. Conversely, Omicron significantly up-regulated only snoRA16, while it led to a significant down-regulation of 4 snoRNAs (snoRD3A, snoRD94, snoRA54, and scaRNA21) (Supplementary Table [136]3). Delta led to the significant up-regulation of 4 snoRNAs (snoRA11B, snoRA71C, snoRD3C, and snoRD67), while no snoRNAs were found down-regulated (Supplementary Table [137]3). Several snoRNAs were exclusively modulated by a specific variant, such as snoRD3C for Delta; snoRA7B, snoRA60, snoRA70, snoRA72, and snoRD118 for Alpha; snora12, snoRA47, snoRA58, snoRA77, and scaRNA9 for Gamma, and snoRA6, snoRA71A, snoRA73, snoRD3A, and scaRNA21 for D614G. Other snoRNAs were commonly shared between at least 2 SARS-CoV-2 variants, such as snoRA45 after D614G and Alpha exposure; snoRA11, after D614G and Gamma exposure; snoRA48, snoRA62, and snoRA80 in response to Alpha, Gamma, and D614G (Fig. [138]8B). Discussion Since its appearance in Wuhan in December 2019, SARS-CoV-2 has threatened the global population, causing more than 7 million deaths. In the first months of 2020, SARS-CoV-2 showed itself to be a highly dangerous human pathogen, but with limited phenotypic change and subsequent adaptation, compared to its later evolution [[139]25]. Otherwise, in late 2020, novel SARS-CoV-2 variants arose, typically characterized by point mutations on the spike protein. New variants exhibited enhanced rates of transmission and major capability in adaptation to humans [[140]26]. Among them, all the variants that were associated with a worse prognosis, regionally or globally, were named by the World Health Organization as “variants of concern” (VOC), and in chronological order the major concerns were Alpha, beta, Gamma, Delta, and Omicron. Missense mutations occurred in the spike protein that characterize the different VOC, resulting in an altered transmissibility and antigenicity that has been shown to confer the ability to evade a primed immune response [[141]27]. Innate immune cells, which include macrophages together with monocytes, neutrophils, and dendritic cells (DCs) represent the first players in the host defence process in response to microbial infections, regulating cytokine release and antigen presentation, and leading to the inflammatory response cascade [[142]28]. Macrophages, through their secretory function, play a key role in the crosstalk with the other immune cells, coordinating the immune response, and developing an effective protection against infective agents [[143]29], both by processing corpuscular antigens to soluble peptides in order to be recognized by B cells, and by presenting antigens to T lymphocytes [[144]30]. These functions clearly explain how macrophages have a key role in the maintenance of the immune homeostasis, and how alterations in these processes can lead to severe immune-impairing mechanisms [[145]31]. It has been previously demonstrated how SARS-CoV-2 is able to use multiple mechanisms to enter host cells [[146]31]. Indeed, despite the fact that viral replication occurs primarily in the respiratory epithelium, it has been shown that CD169 + lung macrophages are also able to transfer SARS-CoV-2 viral particles to adjacent lung regions, strengthening the hypothesis of macrophages as a “Trojan horse” in COVID-19 [[147]32, [148]33]. Due to the central role played by the immune response, and by macrophages in particular, in this study we aimed to compare and deeply analyse the response of monocyte-derived macrophages isolated from healthy participants to five different SARS-CoV-2 VOC. In particular, we focused on the role of the D614G variant, the first notable spike substitution that arose early in the pandemic, and which conferred approximately a 20% growth advantage [[149]34]; Alpha (PANGO lineage10 B.1.1.7); Gamma (P.1); Delta (B.1.617.2), and Omicron (BA.1). Moreover, in our study we also added HCoV-OC43, an ancestral coronavirus strain present in the population, to better define the differences in the pathways activated by OC43 and all the SARS-CoV-2 VOC that arose subsequently. We first assessed the possibility of viral replication inside macrophages by viral titration in a time course assay up to 96 h post-exposure. We observed that when macrophages are exposed to the SARS-CoV-2 VOCs they do not release in the supernatant viable SARS-CoV-2 particles, suggesting that no active replication occurs in these cells, in line with our previous results [[150]8]. We also observed that the nucleoprotein is detectable in macrophages starting at 24 h, and increases over time, while the kinetic of the spike protein is slower, becoming detectable at 96 h after exposure. As we demonstrated that all SARS-CoV-2 VOCs are effectively internalized in the macrophages, there is an increase of viral particles in the macrophages over time, even without production of viable viral particles detectable in the supernatants. Thus, we wanted to assess whether direct transfer of viable virus was an option. We co-cultivated macrophages exposed to the different VOC with uninfected VERO E6 cells and, as soon as 72 h post co-cultivation, a cytopathic effect was detectable. This means that direct transfer of viral particles had occurred, in accordance both with our previous results [[151]8] and also with other instances previously demonstrated. For example, it has been reported that the human cytomegalovirus can be internalized by polymorphonuclear cells and transported through the placenta to the amniotic sac where the virus can spread to the fetus, with severe sequelae [[152]35]. Moreover, Panuska et al. found that alveolar macrophages support both transcription and protein expression of the respiratory syncytial virus (RSV), and produce infectious RSV for prolonged periods (over 2 weeks), suggesting that macrophages act as a reservoir for RSV, and aid in viral extrapulmonary dissemination and prolonged virus shedding [[153]36]. Even if the hypothesis of “Trojan horses” for macrophages is already known in virology, the implications in SARS-CoV-2 dissemination and persistence have yet to be evaluated. Understanding the behaviour of macrophages upon encountering SARS-CoV-2, particularly for the ever-changing VOC, is critical in better managing the disease. For this reason, we decided to have an in-depth look at the transcriptional profiles of macrophages exposed to SARS-CoV-2 VOC. Once established that the macrophages exposed to the viruses contained viral particles, we proceeded to perform the RNA-Seq in order to assess a possible alteration of the MDM transcriptomic profiles. First, we quantified the level of viral internalization by checking the expression of some of the most conserved ORFs of the SARS-CoV-2 genome. In particular, located at the 5′ end of the genome, there are ORF1a and ORF1b, two large overlapping ORFs encoding for the replicase pp1a, encoded by ORF1a, and pp1ab, and generated when a -1 programmed ribosomal frameshift favors a stop codon-skipping process, leading to the translation of the polyprotein pp1ab instead of pp1a. These two proteins are conserved in coronaviruses, and were found to be the most transcribed and present across the variants. Moreover, we checked the expression of some of the other most conserved ORFs. Going from the 5′ to the 3′ of the genome, we checked S (spike protein), E (envelope), M (membrane) and N (nucleocapsid). Finally, we also checked for ORF3a, ORF6, ORF7a, ORF7b, ORF8, and ORF10, considered species-specific since they are present only in some strains [[154]23]. Heat map analysis and principal component analysis (PCA) modelling of the differentially expressed genes clearly showed that the transcriptome profile of the host is markedly different in participants’ macrophages exposed to D614G, Alpha, and Gamma variants compared to the naïve control group. Surprisingly, the Delta variant strongly clustered with the naïve control group rather than with the chronologically previous variants, showing a change of pace. The Omicron variant showed an intermediate transcriptomic profile between these two main groups. Notably, OC43 showed a transcriptomic pattern similar to SARS-CoV-2 D614G, Alpha, and Gamma variants. Analysing the overall deregulated genes of the exposed macrophages to virus variants compared to the naïve control group, the majority of the transcripts were generally up-regulated, rather than down-regulated, with the exception of Delta. In particular, in OC43, SARS-CoV-2 D614G, Alpha and Gamma exposed macrophages, 2885, 2004, 2288, 2530 genes were significantly up-regulated, respectively, and 65, 44, 46, 44 genes significantly down-regulated. In the SARS-CoV-2 Delta-strain-exposed macrophages we found only 8 significantly up-regulated and 6 down-regulated genes, while there were 558 significantly up-regulated and 18 down-regulated genes in the SARS-CoV-2 Omicron-strain-exposed macrophages, confirming the different transcriptional trends of the Delta and Omicron variants observed with both the hierarchical clustering and the PCA analysis. From the KEGG and Reactome dataset analysis of the DEGs, it was clear that macrophages exposed to OC43 and to SARS-CoV-2 variants D614G, Alpha, and Gamma strongly activate immunological key pathways such as the C-type lectin signaling pathway, NOD-like receptor signaling pathway, chemokine signaling pathway, neutrophil degranulation, neutrophil extracellular trap formation, and FC Gamma R-mediated phagocytosis. Moreover, among them, we also found pathways that have been associated with the worsening of symptoms in SARS-CoV-2 patients, such as platelet activation, lipid and atherosclerosis, and VEGF signaling. These pathways were not regulated by SARS-CoV-2 Delta and Omicron variants, confirming the previously assessed evidence obtained with the PCA analysis. Looking at some of the most differentially expressed genes in our samples, we observed that some are involved in a wide range of immunological events, at transcriptional and cellular levels. Among them, we found highly differentially expressed PARP9 and PARP14, members of the PARP superfamily. PARP14, together with STAT6, has been shown to regulate the expression of macrophage tissue factor, a mediator of thrombosis and inflammation, at a post-transcriptional level [[155]37]. PARP9 is considered a transcriptional co-activator of type I IFNR signal transduction; indeed, in B lymphocytes, by binding to deltex E3 ubiquitin ligase 3 L (DTX3L), it acts as STAT1-associated component. Both PARP9 and PARP14 mRNA levels are strongly and significantly increased in macrophages exposed to OC43, SARS-CoV-2 D614G, Alpha, and Gamma variants compared to the naïve control group. Surprisingly, Delta and Omicron variants highly differed in the expression levels of these factors compared to the other SARS-CoV-2 variants, showing instead a high similarity with the naïve control group. In fact, Delta showed very low mRNA levels of PARP9, its ligand DTX3L, and the downstream transcription factor STAT1. Intriguingly, the SARS-CoV-2 Omicron variant also showed lower mRNA levels of these factors compared to the other VOC. Interestingly, PARP14, STAT6, and interferon regulatory factor 3 (IRF3) mRNAs also showed a similar expression profile among the different VOC, and compared to the naïve control group, similar to the PARP9/DTXL3/STAT1 axis. Notably, the expression of these factors in the SARS-CoV-2 Omicron variant showed a lower expression. Since both PARP9 and PARP14 are described as responsible for macrophage activation [[156]38], and recent findings have found their leading role in the hyper inflammatory state through a molecular mimicking phenomenon [[157]39], it is interesting to observe that SARS-CoV-2 Delta and Omicron variants showed a similar expression profile compared to the naïve control group, while they strongly differed from the previous VOC, thus strengthening our hypothesis of a lack of macrophage activation probably driven by PARP9/PARP14 mediators. The different mRNA expression profile of macrophages exposed to SARS-CoV-2 Delta and Omicron, compared to the other VOC included in our study, led us to consider possible changes in macrophage phenotype regulation among the different variants. M0, M1, and M2 macrophage states are highly regulated by specific signaling pathways and genetic programs. Usually, the full characterization of the M1 and M2 macrophages is done by considering more traits, since the combination of specific expression markers is used together with cytokines, specific pathways and receptors [[158]40]. Generally speaking, M1 macrophages express pro-inflammatory and cytotoxic genes, such as CD80, CD86, FCGR1A (CD11), and FCGR2A (CD32), and play a role in killing pathogens, and in tumor onset. M2 macrophages, instead, show anti-inflammatory traits, and are more involved in repairing functions. Enzymes and cytokines typically produced from M2 are ARG1, VEGF, and TGF-β, and CD206 [[159]41]. Surprisingly, looking at the expression of both M1 and M2 macrophages markers, it is clearly shown how macrophages infected with SARS-CoV-2 Delta did not express the canonical markers of M1/M2 differentiation states, suggesting an M0 status, comparable to the naïve control group. Furthermore, also macrophages exposed to the SARS-CoV-2 Omicron variant showed a lower expression of most of the M1/M2 markers, suggesting a similar behavior compared to the Delta variant. In order to have a comprehensive view of the macrophages’ transcriptional machinery we wanted to analyze the non-coding RNA counterpart. In line with the literature, we observed an alteration in the expression of the host ncRNAs [[160]13]. In particular, we found a deregulated expression of lncRNAs, miRNAs, and snoRNAs. LncRNAs and miRNAs have been described as playing a crucial role in the SARS-CoV-2 life cycle, and in the host immune response [[161]11–[162]13, [163]37]. We found a general up-regulation of lncRNAs and miRNAs with all SARS-CoV-2 VOC and OC43, but not Delta, thus suggesting that the regulation of these transcripts generally occurs upon encounter with the coronavirus. The fact that Delta did not trigger any up-regulation of lncRNAs indicates that these transcripts are likely involved mainly in the host innate immune response, which is not triggered by exposure to the Delta variant. Interestingly, we found that miR3648 was down-regulated by both Delta and Omicron, while its levels were unchanged by the other SARS-CoV-2 VOC. MiR3648 expression is triggered in response to endoplasmic reticulum (ER) stress, as observed during the Japanese encephalitis virus infection [[164]42, [165]43]. Therefore, it is tempting to speculate that the down-regulation of miR3648 expression might be necessary for an unperturbed viral replication of Delta and Omicron variants. However, additional studies are needed to support this hypothesis. Less is known about snoRNAs and COVID-19. In fact, some studies have implicated snoRNAs in numerous physiological and pathological viral processes, where they are known to be exploited by RNA viruses to favor their life-cycle, including SARS-CoV-2 [[166]44–[167]46]. The fact that we found snoRNAs differentially regulated upon SARS-CoV-2 VOCs infection raises the question of the putative functional roles played by these ncRNAs during coronavirus infections. SnoRNAs are known to modify cellular ribosomal RNA (rRNA) through the post-transcriptional modifications pseudouridine and 2’-O-methylation, which regulate the structure and stability of RNAs [[168]47]. Interestingly, these post-transcriptional modifications have been found in the SARS-CoV-2 genomic and RNA transcripts [[169]48–[170]50]. Thus, it is plausible to think that SARS-CoV-2 may trigger the transcription of specific host snoRNAs to post-transcriptionally modify and stabilize viral RNAs during its life cycle. However, future experimental validation is necessary to confirm this hypothesis. In conclusion, in our study we confirmed that macrophages can act as Trojan horses for SARS-CoV-2. Interestingly, by varying the VOC the effect in vitro did not change. Macrophages internalize SARS-CoV-2, which does not replicate inside them, but can be transmitted to permissive cells in co-cultivation. Here we have presented an exhaustive analysis of the SARS-CoV-2 VOC-exposed macrophage transcriptomes, revealing significant alterations in gene expression, in both coding and non-coding RNAs. Surprisingly, exposing macrophages to Delta did not have a significant impact on the transcriptional profile compared to the non-exposed macrophages, but produced significant differences when compared to the transcriptional profiles of other VOC. Altogether, our data strongly indicate that the SARS-CoV-2 Delta variant leads to a failure in the activation of the adaptive immune response in human macrophages isolated from immunocompetent humans, and this could explain an immune evasion mechanism driven by Delta, and leading to the more severe symptoms developed by people affected with this variant. Our study paves the way for future research on the molecular basis that regulates the immune response to viral infections, key players of host-pathogen crosstalk. Electronic supplementary material Below is the link to the electronic supplementary material. [171]Supplementary Material 1^ (15.3MB, xlsx) [172]Supplementary Material 2^ (43.9MB, xlsx) [173]Supplementary Material 3^ (2.5MB, xlsx) [174]Supplementary Material 4^ (2.3MB, docx) Acknowledgements