Soluble CD13 is a potential mediator of neutrophil-induced thrombogenic inflammation in SARS-CoV-2 infection

Abstract

The soluble variant of the ectopeptidase CD13 (sCD13), released from the cell surface by matrix metalloproteinase 14 (MMP14), is a potent pro-inflammatory mediator, displaying chemotactic, angiogenic, and arthritogenic properties through bradykinin receptor B1 (B1R). We revealed a link between sCD13 and amplified neutrophil-mediated inflammatory responses in SARS-CoV-2 infection. sCD13 was markedly elevated in patients with COVID-19 and correlated with disease severity and variants, ethnicity, inflammation markers, and neutrophil extracellular trap formation (NETosis). Neutrophils treated with sCD13 showed heightened NETosis and chemotaxis, which were inhibited by sCD13 receptor blockade. Meanwhile sCD13 did not induce platelet aggregation. Single-cell analysis of COVID-19 lungs revealed coexpression of CD13 and MMP14 by various cell types, and higher CD13 expression compared with controls. Neutrophils with high CD13 mRNA were enriched for genes associated with immaturity, though CD13 protein expression was lower. Histological examination of COVID-19 lungs revealed CD13-positive leukocytes trapped in vessels with fibrin thrombi. Flow cytometry verified the presence of B1R and a second sCD13 receptor, protease-activated receptor 4, on monocytes and neutrophils. These findings identify sCD13 as a potential instigator of COVID-19-associated NETosis, potentiating vascular stress and thromboembolic complications. The potent pro-inflammatory effects of sCD13 may contribute to severe COVID-19, suggesting that sCD13 and its receptors might be therapeutic targets.

Keywords: COVID-19; Endothelial cells; Immunology; Infectious disease; Neutrophils; Thrombosis.

Figure 1. Circulating sCD13 levels are elevated in patients with COVID-19 and correlate with inflammatory markers.

(A) Compared with healthy controls, significant elevation of sCD13 was observed in patients with COVID-19 (controls n = 48, COVID-19 n = 172). (B) Black patients showed significantly higher levels of sCD13 compared with White patients with COVID-19 (Asian n = 4, Black n = 77, White n = 73). (C) Significantly higher sCD13 levels were observed in patients with COVID-19 at each age group compared with healthy controls (controls n = 48, COVID-19 n = 172). (D) Patients requiring mechanical ventilation had significantly higher levels of sCD13 compared with patients breathing room air (RA) or with nasal cannula (NC) (controls n = 46, RA or NC n = 95, ventilation n = 72). (E) Significant positive correlation between sCD13 levels and clinical labs ferritin (n = 144) or lactate dehydrogenase (LDH, n = 118) was observed. D-dimer (n = 124) and C-reactive protein (CRP, n = 138) demonstrated positive slopes that were not statistically significant. (F) In a separate cohort, sCD13 was significantly elevated in patients with COVID-19 (n = 20) compared with healthy controls (n = 9), while levels did not significantly change 3 months after the diagnosis (n = 20). (G) Both symptomatic (n = 14) and asymptomatic patients (n = 25) showed significantly higher sCD13 levels compared with healthy controls (n = 21). (H) sCD13 levels were significantly higher in COVID-19 patients with the Alpha variant (n = 14) compared with healthy controls (n = 14) and patients with the Omicron variant (n = 21). (I) Significant elevation of sCD13 in patients with COVID-19 (n = 16) was also observed in a third cohort compared with healthy controls who were not vaccinated (n = 4). (J) Significant elevation of sCD13 was observed in COVID-19 patients with the Delta variant (n = 12) compared with patients with the Alpha (n = 8) or Omicron variants (n = 2). Results are expressed as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Significance was determined by Mann-Whitney test (A and C), Kruskal-Wallis test (B, D, F–H, and J), Spearman’s test (E), and unpaired, 2-tailed Student’s t test (I).

Figure 2. sCD13 correlates with NETosis-associated markers in patients with COVID-19 and activates neutrophils.

(A) sCD13 showed significant correlations with Cit-H3 (n = 172) and MPO-DNA (n = 172) in patients with COVID-19 while S100 A8/A9 (n = 172) showed a positive slope but not a statistically significant correlation. (B) Representative images of neutrophils isolated from peripheral blood and analyzed after stimulation with PBS or sCD13. Panels show merged images of NETs in which neutrophil elastase was stained green by immunofluorescence and DNA was stained blue by Hoechst 33342. n = 3 technical replicates. Scale bar: 100 μm. (C) sCD13 blocked the staining of PAR4-5F10 antibodies, which recognize only the inactivated/uncleaved form of PAR4 yet had no effect on PAR4-FITC antibodies, which recognize both the activated and inactivated forms. n = 3 technical replicates, repeated 2 times. Original magnification, ×1,000. (D) sCD13-induced NETosis was blocked by B1R inhibitor SSR-240612 and PAR4 inhibitor BMS-986120 (all n = 3). (E) PMA-induced NETosis was not impacted by the B1R or PAR4 inhibitors (all n = 3). Results are expressed as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Significance was determined by Spearman’s test (A) and 1-way ANOVA (D and E).

Figure 3. CD13 is highly expressed in lungs from patients with COVID-19.

Immunofluorescence and H&E staining were performed on lung tissues obtained from autopsies of patients with cause of death attributed to COVID-19, targeting CD13, Cit-H3, and nuclear staining DAPI. (A) A pulmonary blood vessel containing a fibrin thrombus (arrow) with numerous entrapped leukocytes with positivity for CD13 (green). These cells can also be seen adherent to the vessel wall. Rare cells with histone H3 labeling (magenta) are interspersed. (B) Small pulmonary blood vessel containing a blood clot with extensive CD13 labeling (red), particularly in areas adjacent to the endothelial lining of the vessel. (C) Alveolar spaces are lined by capillaries. Desquamated pneumocytes and alveolar macrophages can be seen within the spaces (arrows), along with cells positive for histone H3 (magenta). Histone H3–labeled cells can also be seen along the endothelial lining and within the adjacent interstitium. Occasional cells with CD13 positivity (green) can also be seen along the alveolar endothelial lining. Representative lung tissue images of 3 patients with COVID-19 are shown. Original magnification, ×200–×400. Scale bar: 20×: 50 μm; 40×: 20 μm.

Figure 4. Cells expressing CD13, MMP14, B1R, and PAR4 are present in lungs in patients with COVID-19, and cellular levels of CD13 mRNA correlate with disease severity and neutrophil maturity.

Single-cell RNA-Seq results of nasopharyngeal/pharyngeal swabs, bronchial brushings and bronchial lavages were generated from patients with COVID-19 (n = 19) and healthy controls (n = 5). Data are extracted from Chua et al. (29). (A) Both ANPEP (codes for CD13) and MMP14 are expressed on various epithelial cells and macrophages. ANPEP is also expressed on neutrophils while MMP14 is expressed on mast cells. BDKRB1 (codes for B1R) is expressed on some epithelial cells with the highest expression on secretory cells and ciliated cells. F2RL3 (codes for PAR4) was barely detected in this dataset but seemed to be expressed by epithelial cells. Cell abbreviations are defined in Supplemental Figure 8. (B) Cellular ANPEP expression in nasopharyngeal cells is elevated in patients with COVID-19 (moderate COVID-19 n = 8, severe COVID-19 n = 11) compared with healthy controls (n = 5). (C) Using the median expression levels of ANPEP in patients, patients with COVID-19 were divided into 2 populations: ANPEP-high and ANPEP-low. Differentially expressed genes in neutrophils from these 2 groups were shown in the volcano plot (P < 1 × 10^–20, |log₂(fold-change)| < 0.25). (D) Pathway analysis of the differentially expressed genes in ANPEP-high and ANPEP-low neutrophils showed pathways related to neutrophil degranulation, leukocyte activation, and inflammation. (E) The neutrophils from ANPEP-high patients with COVID-19 (n = 10) showed a gene signature of immature-like neutrophils characterized by the overexpression of genes coding for several granule-content proteins (healthy controls n = 5, ANPEP-low n = 9). (F) The score for neutrophil immaturity was higher in critically ill patients with COVID-19 (n = 11) compared with moderate patients (left, n = 8). The median neutrophil-immaturity score of neutrophils in ANPEP-high patients (n = 10) was lower than that in ANPEP-low patients (right, n = 9). (G) The neutrophil-immaturity signature was most prominent in ANPEP-high patients (n = 10) as developing neutrophils versus mature neutrophils from the peripheral blood were examined (healthy controls n = 5, ANPEP-low n = 9). Results are expressed as mean ± SD. ****P < 0.0001. Significance was determined by 1-way ANOVA (B and G) and Mann-Whitney test (F).

Figure 5. CD13, PAR4, and B1R are highly expressed on neutrophils and monocytes from healthy donors.

(A) Gating scheme of flow cytometry analysis on monocytes and neutrophils. (B) CD13, B1R, and PAR4 were expressed on monocytes and neutrophils, while MMP14 showed minimal expression on these cells (n = 3). (C) Histograms showing the changes in CD13, PAR4, and B1R expression in neutrophils and monocytes with or without IL-1β, TNF-α, or IL-6 stimulation. (D) Quantification of relative expression of CD13, PAR4, and B1R on neutrophils and monocytes after stimulation with IL-1β, TNF-α, or IL-6 compared with unstimulated control from 3 healthy donors. Results are expressed as mean ± SD. *P < 0.05, **P < 0.01, ****P < 0.0001. Significance was determined by 1-way ANOVA. FMO, fluorescence minus 1 (gating control); NT, not treated.

Figure 6. Divergent CD13 expression is observed in mature and immature neutrophils.

(A) Reanalysis of single-cell RNA-Seq results generated from circulating human neutrophils (32) revealed that immature neutrophils have higher ANPEP expression. Dot plot of ANPEP, IFN, and neutrophil maturity genes for each neutrophil cluster, showing the average expression level and the percentage of cells expressing the gene in each cluster. (B) Gating scheme of flow cytometry analysis on mature and immature neutrophils isolated from whole blood. CD10⁺CD16^hi defines mature neutrophils while CD10^–CD16^lo defines immature neutrophils. (C) Significantly higher expression of CD13 and lower expression of PAR4 were observed in mature neutrophils compared with immature neutrophils while B1R expression showed similar levels. Data generated from 5–10 healthy controls. Results are expressed as mean ± SD. **P < 0.01. Significance was determined by Wilcoxon’s test.