**3. Results**

### *3.1. ERAP2 Allelic Variants Analyses*

Analysis of ERAP2 SNP prevalence was aligned with European population distribution reported in the U.S. National Library of Medicine Database [https://www.ncbi.nlm.nih.gov/snp/rs2549782? fbclid=IwAR1ZdwC747PDWvtAzt6hZBV5j7oFZiPkLjY-JdSee1Plzvym7fhJVQc1Aks (data not shown). Among the 20 genotyped HC: 6 were HomoA, 8 heterozygous, 6 HomoB. Subsequent analyses were performed only on PBMCs and MDMs isolated from HomoB and heteroAB donors to exclude confounding results. Indeed, HomoA individuals express negligible levels of ERAP2/Iso2 and Iso3.

Calu3 cell lines were heterozygous for rs2248374 ERAP2 genotype.

### *3.2. mRNA Expression of ERAPs in PBMCs from Subjects Carrying Di*ff*erent ERAP2 Genotypes Following Microbial Stimulation*

To verify whether the expression of ERAP2/Iso3 isoform is exclusively flu-specific or it may be triggered by other stimuli, we analyzed its expression on PBMCs isolated from HomoB and HeteroAB HC following HIV-AT-2, i-SARS-CoV-2, CMV, LPS, flu, IFN<sup>α</sup>, and IL-1β. As expected, the expression of the newly identified ERAP2/Iso3 was significantly augmented in PBMCs from all the subjects included in the study in response to flu (*p* < 0.002). However, even following CMV (*p* < 0.004), LPS (*p* < 0.002), HIV-AT-2 (*p* < 0.003), i-SARS-CoV-2 (*p* < 0.002) and IFNα (*p* < 0.003) stimulation, we observed a significant increase of its expression (Figure 2A). Conversely, IL-1β addition to cell culture did not result in ERAP2/Iso3 induction (Figure 2A). ERAP2/Iso1 expression was observed only in PBMCs from HeteroAB subjects and was induced following all the microbial-stimuli employed plus IFNα but not in response to IL-1β (Figure 2B). However, statistical significance was observed exclusively following HIV-AT-2 (*p* < 0.025), i-SARS-CoV-2 (*p* < 0.03) and CMV (*p* < 0.02). Likewise, ERAP1 expression was induced by all the stimuli employed excepting IL-1β and reached statistical significance following flu (*p* < 0.002), HIV-AT-2 (*p* < 0.03) CMV (*p* < 0.002), i-SARS-CoV-2 (*p* < 0.008) LPS (*p* < 0.05). We also observed a reduction following IL-1β (*p* < 0,001) stimulations (Figure 2C).

**Figure 2.** ERAP1, ERAP2/Iso1 and ERAP2/Iso3 mrna expression is increased following microbial stimulation. pbmcs isolated from 8 heteroab and 6 homob individuals were in vitro stimulated with microbial antigens (flu, CMV) inactivated viruses (i-SARS-CoV-2, HIV-AT-2) bacterial by-products (LPS) or inflammatory stimuli (IFN<sup>α</sup>, IL-1β) for ten h. mRNA expression for ERAP2/Iso3 (**A**), ERAP2/Iso1 (**B**) and ERAP1 (**C**) were assessed by RT-Real-Time PCR. Results are shown as the media of the relative expression units to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and β-actin reference genes calculated by the 2−ΔΔCt equation. (**D**) The microbial-dependent genetic control of ERAP2/Iso3 expression is underlined by the observation that its abundances are nearly doubled in HapB homozygotes compared to heterozygotes. Results are expressed as mean ± ES. \* = *p* < 0.05; \*\* = *p* < 0.01.

Notably, the microbial-dependent genetic control of ERAP2 isoform usage is sustained by further evidence. Indeed, there was a significant correlation between whole microbial transcript quantity and ERAP2/Iso3 transcript abundances in heterozygotes compared to HapB homozygotes (Figure 2D).

### *3.3. Gene Expression of Immune Selected E*ff*ectors in PBMCs Following Microbial Stimulation*

To verify if the increased expression of ERAPs in response to microbial stimulation could be extended to other factors involved in the orchestration of the immune response, mRNA expression of 70 selected effectors was investigated by the innovative Quantigene Plex Gene expression technology. The genes whose mRNA expression was upregulated are implicated in almost all phases of immune response, including: chemokines, cytokines and cytokine receptors, pathogen recognition receptor, inflammasome, cholesterol metabolism, interferon stimulated genes, adhesion molecules, activation/inhibition markers, antigen presentation factors (Figure 3). Notably, the gene expression pattern was partially shared by all the stimuli employed and partially pathogen-specific as summarized in Figure 3. In particular, following i-SARS-CoV-2-stimulation a significantly higher transcription rate was observed for: CCL2 (*p* < 0.05); CCL5 (*p* < 0.035), HMGCS1 (*p* < 0.002), PYCARD (*p* < 0.002), CASP1 (*p* < 0.001), CD44 (*p* < 0.016), CD274 (*p* > 0.008), IL-8 (*p* < 0.003); IL-1β (*p* < 0.006), ABCA1 (*p* < 0.02), IL-6R (*p* < 0.004), CCL3 (*p* < 0.01), IFNγ (*p* < 0.05); TAP1 (*p* < 0.05).

**Figure 3.** mRNA expression of genes involved in the anti-microbial immune response was modulated in response to different pathogens. Quantigene Plex Gene expression technology was applied to quantify gene expression on PBMCs isolated from 8 HeteroAB and 6 HomoB individuals and stimulated with microbial antigens (flu, CMV) inactivated viruses (i-SARS-CoV-2, HIV-AT-2) bacterial by-products (LPS) or inflammatory stimuli (IFN<sup>α</sup>, IL-1β). Gene expression (mean values) is shown as a color scale from white to blue (Heatmap). Only statistically significant p values from T-test comparison between unstimulated and stimulated PBMCs are shown in table.

Unlike ERAP2 expression, no differences were observed in mRNA expression levels of all the analyzed genes in PBMCs from HomoB and HeteroAB subjects in response to all the stimuli engaged (data not shown).

### *3.4. mRNA Expression of ERAPs in In Vitro SARS-CoV-2 Infected Calu3 Cell Lines and HIV-Infected PBMCs*

To verify whether ERAP2/Iso3 expression varies in response to growing viral concentrations we adopted two in vitro model of infection. Thus, Calu3 cell lines were infected with different SARS-CoV-2 viral input and after 48 hours' viral replication as well as ERAP mRNA expression were assessed. As expected, SARS-CoV-2 replication increased according to the rising viral input as assessed analyzing both N1 (MOI 0.5 vs. 5: *p* < 0.01; MOI 0.5 vs. 1000: *p* < 0.001) and N2 (MOI 0.5 vs. 5: *p* < 0.01; MOI 0.5 vs. 1000: *p* < 0.001) (Figure 4A). Images of cellular cytopathic effect on SARS-CoV-2 infected cells at 48 h showed that despite robust SARS-CoV-2 replication in Calu3 cells, substantial cell death was detected only in cells infected with 1000 MOI (Figure 4B). Notably, this increase was coupled with

a progressive rise of ERAP2/Iso3 expression compared to the uninfected condition (MOI 0.5: *p* < 0.04; MOI 5: *p* < 0.01; MOI 1000: *p* < 0.01). Likewise, ERAP2/Iso1 and ERAP1 expression was induced in a viral-dose dependent manner, although statistical significance was observed only for ERAP2/Iso1 (MOI 0.5: *p* < 0.05) (Figure 4C).

**Figure 4.** In vitro SARS-CoV-2 infection assay on Calu3 cells. (**A**) SARS-CoV-2 replication was assessed on Calu3 cells infected at 0.5, 5, and 1000 MOI 48 h post-infection. Viral copy number quantification was performed by generating a standard curve from the quantified 2019-nCoV\_N positive plasmid control for Nucleocapsid (N) region 1 and 2, showing a significant increase in response to the increased viral input. (**B**) SARS-CoV-2-induced cytopathic effects were assessed in Calu3 infected cells. Representative images of SARS-CoV-2 infected-cells at 0.5 and 1000 MOI are reported. At 48 h post-infection, typical cytopathic effects, including cell rounding, detachment, degeneration, and syncytium formation were seen only in cells infected at 1000 MOI. Cells were imaged by optical microscope observation (ZOE™ Fluorescent Cell Imager, Bio-Rad, Hercules, CA, USA). (**C**) ERAP2/Iso3, Iso1, and ERAP1 mRNA expression by in vitro SARS-CoV-2 infected Calu3 cells increased according to the rising viral input. Mean values ± ES are reported. \* = *p* < 0.05; \*\* = *p* < 0.01; \*\*\* = *p* < 0.001. MOI = multiplicity ofinfection.

In vitro HIV-1 infection of PBMCs isolated from HapB HC produced similar results. Indeed, as the viral input raised, viral replication quantified through p24 concentration analyses at 5 days post infection increased (0.1 vs. 1 ng p24 HIV-1Bal/1 × 10<sup>6</sup> PBMCs: *p* < 0.02; 0.1 vs. 10 ng p24 HIV-1Bal/1 × 10<sup>6</sup> PBMCs: *p* < 0.001) (Figure 5A). Likewise, ERAP2/Iso3 gene expression increased compared to the uninfected condition (0.1 ng p24 HIV-1Bal/1 × 10<sup>6</sup> PBMCs: *p* < 0.003; 0.1 ng p24 HIV-1Bal/1 × 10<sup>6</sup> PBMCs: *p* < 0.04; 10 ng p24 HIV-1Bal/1 × 10<sup>6</sup> PBMCs: *p* < 0.03) (Figure 5B). ERAP2/Iso1 (0.1 ng p24 HIV-1Bal/1 × 10<sup>6</sup> PBMCs: *p* < 0.05) and ERAP1 (0.1 ng p24 HIV-1Bal/1 × 10<sup>6</sup> PBMCs: *p* < 0.05) mRNA levels showed a similar trend (Figure 5B).

**Figure 5.** In vitro HIV-1 infection assay on PBMCs. (**A**) HIV-1 replication was assessed by p24 quantification on PBMCs isolated from 8 HeteroAB and 6 HomoB individuals infected with 0.1, 1, and 10 ng p24 HIV-1Bal/1 × 10<sup>6</sup> PBMCs 5 days post infection. Results showed a significant increase in viral replication according to the increased viral input. (**B**) ERAP2/Iso3, Iso1, and ERAP1 mRNA expression by in vitro HIV-1-infected PBMCs increased according to the rising viral input. Mean values ± ES are reported. \* = *p* < 0.05; \*\* = *p* < 0.01; \*\*\* = *p* < 0.001.

### *3.5. ERAP2*/*Iso3 Protein Production by MDMs from HeteroAB Subjects Following Microbial Specific Stimulation*

To verify if the short ERAP2/Iso3 isoform would function as an RNA or is translated into a protein product we performed a western blot assay on MDMs from 3 HeteroAB subjects triggered with different microbial antigens. Remarkably, the antibody which recognizes the full-length ERAP2 (Iso1) was able to detect also one short protein isoform (∼50 kDa) in CMV, flu, HIV-AT-2, i-SARS-CoV-2, LPS, IFNα stimulated cells from HeteroAB subjects, suggesting the translation of the short microbial-specific ERAP2/Iso3 (Figure 6). Conversely, following IL-1β stimulation, only ERAP2/Iso1 isoform was detected. The production of ERAP1 protein was observed in all the stimulated conditions except IL-1β (Figure 6). As the proteins extracted from the 3 HeteroAB subjects were sub-pooled for WB analyses, statistical evaluation of the results was not possible.

**Figure 6.** ERAP1, ERAP2/Iso1, and ERAP2/Iso3 production in pathogen stimulated monocyte-derived macrophages (MDMs). MDMs differentiated from 3 HeteroAB participants stimulated for 36 h with microbial antigens (flu, CMV) inactivated viruses (i-SARS-CoV-2, HIV-AT-2) bacterial by-products (LPS) or inflammatory stimuli (IFN<sup>α</sup>, IL-1β) were tested for protein using primary antibodies specific to a ERAP1 (goat polyclonal), ERAP2 and β-actin. Proteins extracted from the 3 HeteroAB subjects were sub-pooled for WB analyses. Histograms representing ERAP1, ERAP2/Iso1, and ERAP2/Iso3 densitometric quantification. Quantification was performed by Quantity One 4.6.6 software (Bio-Rad) and normalization was permed on GAPDH.
