Angiotensin‐converting enzyme II expression and its implication in the association between COVID‐19 and allergic rhinitis

To the Editor,

Coronavirus disease 2019 (COVID‐19), caused by severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2), affects more than 17 million of people and results in more than 666 000 deaths all over the world. Although allergic diseases are highly prevalent globally, their risks for the development of COVID‐19 remain poorly understood.

Severe acute respiratory syndrome coronavirus 2 entry host cells via angiotensin‐converting enzyme II (ACE2). ¹ Upregulated ACE2 expression has been associated with increased risk of COVID‐19 in patients with chronic obstructive pulmonary disease, diabetes, and hypertension. ² ACE2 gene expression is enriched in nasal epithelial cells, highlighting the importance of nose as a portal for initial SARS‐CoV‐2 infection and transmission.

Allergic rhinitis (AR) is the most common disorder of nose and affects 10%‐40% of the population. ³ Previous studies reported low incidences of AR in COVID‐19 patients, ranging from 0% to 1.8% in China. ⁴ , ⁵ However, those results were generated solely based on the medical records, and AR comorbidity might not be well considered under actual emergency situation. ⁴ , ⁵ Moreover, the association between AR comorbidity and the disease severity of COVID‐19 and the role of ACE2 in this association are largely unknown.

Here, we retrospectively analyzed 1172 etiologically confirmed COVID‐19 patients discharged from Tongji Hospital, Wuhan, China from January 27, 2020 to March 10, 2020. Hospital electronic medical records were extracted and comorbidities were reevaluated by the telephone follow‐up. Both multivariate logistic regression and propensity score matching (PSM) analysis were performed to exclude the influence of potential confounding variables. In addition, repository inferior turbinate tissues and cells from 29 control subjects without AR and 29 patients with AR undergoing nasal septoplasty and collected before COVID‐19 era were used for the study of ACE2 expression. Human nasal epithelial cells (HNECs) were collected by epithelial scrapings of the inferior turbinates from subjects without AR. Primary HNECs were cultured with the air‐liquid interface method. The cell culture was also performed before COVID‐19 era, and the ACE2 expression was analyzed in this study. The RNA‐seq data of nasal and bronchial brushings from 7 patients with concomitant AR and asthma and 9 healthy controls were acquired from the Gene Expression Omnibus database (GSE101720). More information regarding subjects and methods is provided in this article's Online Supplement including Tables S1‐S2.

In our cohort, 115 (9.8%) patients reported physician‐diagnosed AR. COVID‐19 patients without AR were older than those with AR (61 [49‐69] vs 54 [40 −65]; P < .01). Patients with AR had a higher incidence of concomitant chronic liver disease (4.4% vs 1.1%; P = .02), and tended to have a higher incidence of asthma comorbidity (5.2% vs 2.2%; P = .06) and a lower incidence of hypertension (24.4% vs 32.5%; P = .07) than those without AR (Table S3). After adjusting for or propensity score matching for confounding factors, including age, gender, smoking status, and comorbidities (Table S3), no significant difference in frequencies of symptoms or laboratory results was found between COVID‐19 patients with and without AR (Table 1 and Table S4). Importantly, no difference in the frequencies of severe cases on admission, receiving mechanical ventilation and other treatments, or complications including severe acute respiratory syndrome was revealed for patients with and without AR either (Table 1). In PSM analysis, we were able to match 109 patients without AR to 109 patients with AR at a ratio of 1:1 (Table S3).

TABLE 1.

Clinical characteristics, complications, and treatments of COVID‐19 patients without and with AR before and after propensity score matching

Characteristics	Unmatched		P value	Adjusted P value a , *	Matched (1:1) a , *		P value
	Without AR	With AR			Without AR	With AR
Subject, N	1057	115	‐	‐	109	109	‐
Systemic signs and symptoms, N (%)	‐	‐	‐	‐	‐	‐	‐
Fever	823 (77.9)	98 (85.2)	.07	.09	85 (78.0)	95 (87.2)	.10
Cough	700 (66.2)	67 (58.3)	.09	.08	73 (67.0)	62 (56.9)	.16
Shortness of breath	455 (43.1)	40 (34.8)	.09	.07	52 (47.7)	38 (34.9)	.07
Fatigue	261 (24.7)	24 (20.9)	.36	.43	30 (27.5)	22 (20.2)	.27
Anorexia	253 (23.9)	21 (18.3)	.17	.09	26 (23.9)	19 (17.4)	.32
Diarrhea	182 (17.2)	24 (20.9)	.33	.42	16 (14.7)	21 (19.3)	.47
Myalgia	157 (14.9)	15 (13.0)	.60	.62	18 (16.5)	12 (11.0)	.33
Nausea and vomiting	78 (7.4)	7 (6.1)	.61	.55	5 (4.6)	5 (4.6)	.99
Headache	78 (7.4)	9 (7.8)	.86	.94	13 (11.9)	9 (8.3)	.50
Dizziness	47 (4.5)	5 (4.4)	.96	.99	3 (2.8)	5 (4.6)	.72
Severe COVID‐19 cases, N (%)	183 (17.3)	16 (13.9)	.36	.79	17 (15.6)	16 (14.7)	.99
Complications, N (%)	‐	‐	‐	‐
Acute liver injury	70 (6.6)	9 (7.8)	.56	.68	6 (5.5)	7 (6.4)	.99
ARDS	64 (6.1)	5 (4.3)	.44	.57	5 (4.6)	5 (4.6)	.99
Acute kidney injury	61 (5.8)	2 (1.7)	.08	.28	3 (2.8)	2 (1.8)	.99
Acute myocardial injury	30 (2.8)	1 (0.9)	.35	.29	3 (2.8)	1 (0.9)	.62
Shock	7 (0.7)	2 (1.7)	.21	.38	0 (0)	2 (1.8)	.50
In‐hospital treatments, N (%)	‐	‐	‐	‐
Oxygen supplementation	844 (79.9)	85 (73.9)	.14	.29	87 (79.8)	79 (72.5)	.27
Mechanical ventilation	41 (3.9)	3 (2.6)	.48	.66	4 (3.7)	3 (2.8)	.99
Antibiotic treatment	766 (72.5)	84 (73.0)	.90	.73	83 (76.1)	79 (72.5)	.64
Antiviral treatment	985 (93.2)	107 (93.0)	.95	.99	102 (93.6)	101 (92.7)	.99
Glucocorticoid therapy	273 (25.8)	22 (19.1)	.12	.06	32 (29.4)	20 (18.3)	.08
Intravenous immunoglobulin therapy	171 (16.2)	18 (15.7)	.88	.96	19 (17.4)	16 (14.7)	.71

Data are presented numbers with percentages for categorical variables.

For categorical variables, Chi‐square or Fisher's exact test was applied to compare the difference in proportions between groups when appropriate.

Abbreviations: AR, allergic rhinitis; ARDS, acute respiratory distress syndrome; COVID‐19, coronavirus disease 2019.

^a Age, gender, smoking status, and all comorbidities were selected for propensity score matching. Please see Table S3.

^*In adjusted logistic regression analysis, adjusted variables included age, gender, smoking, and comorbidities (hypertension, diabetes, cardiovascular disease cerebrovascular diseases, malignancy, chronic liver diseases, chronic kidney diseases, chronic rhinosinusitis, asthma, and chronic obstructive pulmonary disease).

We next studied ACE2 expression and found its mRNA and protein expression in nasal tissues was comparable between AR patients and control subjects (Figure 1A,B). ACE2 mRNA expression was downregulated by IL‐4 and IL‐13, whereas upregulated by IFN‐α, IFN‐γ, and TNF‐α in cultured HNECs (Figure 1C). We discovered that the mRNA expression of type 2 response genes, including ST6GAL1, POSTN, and CCL26, was increased in nasal tissues in AR patients compared with that in non‐AR controls, however, the mRNA expression of IFN response genes, including CXCL10 and CXCL11, was comparable between AR patients and non‐AR controls (Figure [Link], [Link]). ACE2 expression positively correlated with the expression of IFN response genes, but not type 2 response genes, in nasal tissues when analyzing AR and control subjects together or separately (Figure 1D‐E and Figure [Link], [Link]). Our findings of ACE2 expression in nasal tissues in AR patients and its relationship with the expression of IFN response genes were confirmed in nasal brushing cells from patients with concomitant AR and asthma by analyzing the public database (GSE101720) (Figures S3 and S4).

FIGURE 1.

Angiotensin‐converting enzyme II expression and its correlation with IFN and type 2 response in AR patients and healthy controls. A, The mRNA expression level of ACE2 gene in inferior turbinate tissues from AR patients and controls as detected by quantitative RT‐PCR. Control, n = 19; AR, n = 19. B, Immunohistochemistry study of ACE2 expression in inferior turbinate tissues from AR patients and controls. The representative photomicrographs are shown (original magnification ×400). The staining intensity in epithelium was quantified, and the results were presented as average optical density value per unit area. Control, n = 10; AR, n = 10. C, Human nasal epithelial cells (HNECs) scraped from inferior turbinate mucosa of control subjects were cultured with an air‐liquid interface method. After differentiation, HNECs were stimulated with various cytokines, dexamethasone (DEX), poly (I:C) (dsRNA), lipopolysaccharides (LPS), and staphylococcal enterotoxin B (SEB) for 6 h. After stimulation, cells were harvested for quantitative RT‐PCR assay (n = 6). D and E, ACE2 mRNA expression positively correlated with the mRNA expression of IFN response genes (CXCL10 and CXCL11) (D), but not type 2 response genes (ST6GAL1, POSTN, and CCL26 expression) (E) in tissues when analyzing AR patients and control subjects together (n = 38). Tissue data are expressed as medians and interquartile ranges and analyzed by Mann‐Whitney U 2‐tailed test. Cell culture data are expressed as medians ± SEM and analyzed by unpaired Student's t test. Spearman's correlation was used for correlation analysis. ACE2, angiotensin‐converting enzyme II; AR, allergic rhinitis; IL, interleukin; TNF‐α, tumor necrosis factor (TNF) α; IFN, interferon; TSLP, thymic stromal lymphopoietin; CCL, chemokine (C‐C motif) ligand; CXCL, chemokine (C‐X‐C motif) ligand; ST6GAL1, beta‐galactoside alpha‐2,6‐sialyltransferase 1; POSTN, periostin

In contrast to nasal epithelial cells, ACE2 gene expression was decreased in bronchial epithelial cells in patients with concomitant AR and allergic asthma as compared to that in healthy controls by analyzing GSE101720 datasheet (Figure [Link], [Link]A). ACE2 gene expression negatively correlated with the expression of type 2 response genes (POSTN, CLCA1, and IL1RL1), but not IFN response genes (IFI6, CXCL10, and CXCL11) in bronchial epithelial cells when analyzing all the subjects or patients with asthma and AR alone (Figure [Link], [Link]B and Figure [Link], [Link]). We further found nasal epithelial cells had higher ACE2 and CXCL10 expression levels, but lower POSTN, CLCA1, CPA3, and IL1RL1 expression compared with those in bronchial epithelial cells (Figure [Link], [Link]A) in patients with concomitant AR and asthma. Consistently, the gene expression ratios of CXCL10/ST6GAL1, CXCL10/POSTN, CXCL10/CPA3, and CXCL10/IL1RL1 were higher in nasal epithelial cells than in bronchial epithelial cells (Figure [Link], [Link]B), suggesting a predominant IFN and type 2 response in upper and lower airways, respectively, under allergic condition.

In this study, we carefully confirmed and reevaluated AR comorbidity in discharged COVID‐19 patients by telephone follow‐up. This may be the reason that the AR prevalence in our cohort (9.8%) was higher than those previously reported in Chinese COVID‐19 patients (0%‐1.8%). ⁴ , ⁵ The prevalence of AR in our COVID‐19 cohort is comparable to that in general populations in Wuhan (9.7%). ⁶ In addition, we did not find any association between AR comorbidity and disease severity in COVID‐19 patients. Chhiba KD et al have recently reported that AR was not associated with an increased risk of COVID‐19‐related hospitalization. ⁷ Collectively, these results indicate that primary AR may not modify the risk for COVID‐19.

In line with recent reports, ⁸ , ⁹ we found that type 2 cytokines downregulated, whereas IFNs upregulated ACE2 gene expression in HNECs. However, ACE2 expression correlated with IFN response, but not type 2 response, in nasal tissues and epithelial cells, underscoring a predominant role of IFNs in regulating ACE2 expression in upper airways. ⁹

Allergic rhinitis and allergic asthma are frequently co‐occurred and may share common immune‐pathogenic mechanisms such as type 2 inflammation. However, we found that patients with concomitant AR and allergic asthma had reduced ACE2 expression in bronchial epithelial cells, which is likely regulated by type 2 response. The difference in ACE2 expression in allergic nasal and bronchial epithelial cells is possibly related to the different tension of IFN and type 2 response in upper and lower airways under allergic condition. Therefore, the counter effect of IFN and type 2 response may have an important role in regulating ACE2 expression in airways. The reduced ACE2 expression in allergic asthmatics may suggest a lower risk for COVID‐19. Low prevalence of asthma (0%‐0.9%) was observed in patients with COVID‐19 in several studies in China. ⁴ In a USA cohort, Chhiba et al ⁷ recorded a relatively high prevalence of asthma (14%); however, there was no significant difference in hospitalization rate or mortality between patients with and without asthma.

Our study has several limitations. First, self‐reported symptoms and comorbidities might lead to the potential misestimation of the prevalence and the strength of association with the clinical outcomes. Second, we did not include fatal cases since no subsequent follow‐up confirmation of comorbidities could be made for them. Third, we could not get the demographic information of subjects in the public gene dataset, and the number of subjects was limited. We therefore could not preclude the potential bias. However, the nasal epithelial cell results were consistent with our nasal tissue results derived from a relatively larger cohort.

In conclusion, for the first time, we provide the evidence that AR comorbidity may not have significant modifying effect on the development and expression of COVID‐19. ACE2 expression is not altered in AR patients. ACE2 gene expression in airways is regulated, at least in part, by the counter effect of type 2 and IFN inflammation.

CONFLICTS OF INTEREST

The authors declare that they have no conflicts of interest.

Supporting information

Fig S1

Fig S2

Fig S3

Fig S4

Fig S5

Fig S6

Fig S7

App S1