The expression profiles of CD209 mRNA and protein in different tissues, immune cell types, and cancer types were obtained from the online database, namely, The Human Protein Atlas website (https://www.proteinatlas.org). The expression data for CD209 mRNA and protein in pan-cancer can be downloaded from the UALCAN analysis website (http://ualcan.path.uab.edu/analysis.html). We also analyzed the single-cell RNA-seq data to search for CD209 expression data in the nasopharynx immune cells of COVID-19 patients, which is available on the UCSC cell browser portal (http://www.cells.ucsc.edu/).

The Kaplan–Meier plotter (Győrffy, 2021) was used to obtain the overall survival (OS) significance map data of CD209 in all cancer types (https://kmplot.com/analysis/), and the “log-rank test” module was used to test the hypothesis.

We used the default “Immune module” available in the TIMER2 database (http://timer.cistrome.org/) to analyze the relationship of CD209 expression with the abundances of NK cells, CD4⁺ T cells, CD8⁺ T cells, B cells, neutrophils, macrophages, cancer-associated fibroblasts, and dendritic cells by their corresponding gene modules. In a chosen cancer type, the correlation module could generate user-defined gene expression scatter plots, as well as their relative statistical significance by Spearman’s correlation and corresponding estimation.

To study the CD209 expression in SARS-CoV-2- or SARS-CoV-infected cells/organs, we reanalyzed the public available transcriptomic data (GSE52920, GSE171110, and GSE161881) from the Gene Expression Omnibus database (GEO), together with a proteome dataset (IPX0002393000) from the iProX database (http://111.198.139.98/page/home.html) (Nie et al., 2021). In particular, GSE52920 contained six lung tissues of mice, including three infected SARS-CoV and three controls. Vero E6 cells were mock-infected or infected with SARS-CoV-2 at an moi of 2 in the GSE161881 dataset. GSE171110 contained the whole-blood gene expression profiles of 44 COVID-19 patients and 10 healthy donors. RNA-seq data from tests between COVID-19 patients and their healthy controls were obtained from the IPX0002393000 dataset. A graphical visualization of CD209 data was prepared using Graph Prism 7.0 software (GraphPad Prism Software, United States).

Three datasets including GSE10072, GSE68465, and GSE50081 were employed to investigate the expression of CD209 in LUAD or LUSC patients with different smoking histories. All patients with smoking histories were divided into three groups: never smoked, former smoker, and current smoker. Graphs were prepared using Graph Prism 7.0 software.

We evaluated the gene correlations by Spearman’s correlation. Their correlation strength was determined using the absolute value guidelines as follows: 0.80–1.0 as “very strong,” 0.60–0.79 as “strong,” 0.40–0.59 as “medium,” 0.20–0.39 as “weak,” and 0.00–0.19 as “very weak.” Statistical significance was assessed by p values < 0.05.

The CD209 gene was most abundantly expressed in the lymph nodes, adipose tissue, and muscle tissues, while the CD209 protein was distributed mostly in the lymph nodes and lungs in male and female tissues (Figure 1A). Ultimately, we analyzed the CD209 gene expression in four public datasets (HPA, Consensus, GTEx, and FANTOM5). The expression of CD209 was detected at the highest levels within the adipose tissue and lymph nodes and at relatively lower levels in the small intestine, placenta, heart muscle, and urinary bladder. Intriguingly, adipose tissue demonstrated the abundant expression of CD209 in the GTEx and HPA public datasets, which was in line with the current concept that obese patients are more susceptible to infection by SARS-CoV-2 (Figure 1B). Recent studies have shown that SARS-COV-2 can directly infect primary DCs and macrophages, which may lead to the destruction of human LNs and spleens and subsequent lymphocytopenia (Feng et al., 2020). However, ACE2 was barely expressed in lymphoid cells. Our data showed that CD209 was a highly expressed gene in macrophages and DCs. Moreover, we investigated the CD209 gene expression in a variety of tissues through the HCCDB database. We found that adipose tissue and the small intestine had the highest CD209 gene expression (Figure 1C), which was consistent with the expression of ACE2. In summary, CD209 was expressed abundantly in various healthy human tissues, especially in adipose tissue, small intestine, and lymph nodes, which had higher susceptibility to infection by SARS-CoV-2.

To identify the CD209 expression in SARS-CoV-2-infected versus infected organs/cell lines, we reanalyzed the public available transcriptomic datasets (GSE52920, GSE161881, GSE171110, and iProX IPX0002393000) that have been uploaded recently to GEO and iProX. Compared to the healthy controls, the mRNA abundance of CD209 demonstrated significant decreases in the lungs of mice (Figure 1D, Supplementary Table S1), Vero E6 cells (Figure 1E, Supplementary Table S2) and peri-blood (Figure 1F, Supplementary Table S3), and testes of patients (Figure 1G, Supplementary Table S4) infected with SARS-CoV/SARS-CoV-2. Together with previous discoveries, these data strongly suggested that CD209 might play a critical role in mediating SARS-CoV-2 infection in patients.

Cancer patients suffer a poor prognosis after being infected with SARS-CoV-2 due to their impaired immunity (Li et al., 2020). Thus, we estimated the landscape profile of CD209, a potential SARS-CoV-2 receptor, in pan-carcinoma patients. In the TCGA dataset, we found that CD209 was expressed in various cancer types, while its mRNA abundance was highest in sarcoma (SARC) (Figure 2A), whereas the protein level of CD209 was most enriched in head and neck cancer (Figure 2B). Recent studies demonstrated that elderly patients with sarcoma may be at an increased risk of developing severe complications following SARS-CoV-2 infection (Wagner et al., 2021a; Wagner et al., 2021b). We further showed that the CD209 expression was positively associated with advanced age in SARC patients, especially in those elderly patients over 60 years old (Figure 2C).

Previous studies showed that ACE2 and CD209 were overexpressed in the healthy lungs of smokers, especially former smokers (Cai et al., 2020a; Cai et al., 2020b), but the relevance of this phenomenon in lung cancer remains unclear. Thus, we first examined the expression profiles of former smokers, nonsmokers, and current smokers in LUSC (lung squamous cell carcinoma) and LUAD (lung adenocarcinoma) patients. We found that the CD209 expression was significantly upregulated in former smokers with LUAD (Figures 2D,E, Supplementary Table S5, S6) and current smokers with LUSC (Figure 2F, Supplementary Table S7). These findings might help to interpret why lung cancer patients are more liable to be infected with COVID-19 (Liang et al., 2020).

The efficiency of CD209 in the survival of pan-cancer patients was further explored with the Kaplan–Meier plotter (Győrffy, 2021). We found that the reduced CD209 expression was associated with a poor overall survival rate in LUAD (p = 0.012, HR = 0.69), READ (rectum adenocarcinoma) (p = 0.052, HR = 0.46), and (cervical squamous cell carcinoma) (p = 0.034, HR = 0.52). Conversely, the high expression of CD209 predicted poor overall survival in KIRP (kidney renal papillary cell carcinoma) (p = 0.039, HR = 1.87), LUSC (p = 0.0061, HR = 1.48), KIRC (kidney renal clear cell carcinoma) (p = 0.031, HR = 1.42), ESCC (esophageal squamous cell carcinoma) (p = 0.058, HR = 2.42), LHC (liver hepatocellular carcinoma) (p = 0.064, HR = 1.41), and THYM (thymoma) (p = 0.039, HR = 340295023.97) (Figure 3). In summary, these data suggested that the particular distribution of CD209, an alternative receptor of SARS-CoV-2 (Amraei et al., 2021), in different cell types might lead to different prognoses in specific carcinoma patients, indicating that during the COVID-19 epidemic, intensive health care is needed for these cancer patients.

Patients severely infected by SARS-CoV-2 usually develop a “cytokine storm” (Mahallawi et al., 2018), which may be due to inflammation-induced lung injury and viral sepsis (Prompetchara et al., 2020). First, we investigated whether the liberation of these abnormal cytokines would be connected to the expression of CD209 across different types of immune cells in three different sub-datasets. Monocytes exhibited the highest CD209 expression, and DCs also expressed appreciable levels of CD209 (Figures 4A–C). Single-cell RNA-seq analysis also demonstrated similar results in monocyte-like inflammatory DCs (moDCs) and nonresident macrophages (nrMa) in the nasopharynx of COVID-19-infected patients (Figure 4D). Transcriptomic changes in COVID-19-related genes within virus-positive cells were highly upregulated in S100A8/A9, mainly in monocytes and megakaryocytes in peripheral blood, which might result in intensive “cytokine storms” that could be observed in severe patients (Ren et al., 2021). Myeloid dendritic cells, which have played a crucial role in adaptive immunity generation, are known for being able to present antigens toward T-cell surfaces and could sensitize naive T cells (Howard et al., 2004). These data illustrated that CD209 is widely expressed in various adaptive/innate immune cells, primarily in DC cells and monocytes, suggesting that CD209 is pivotal in the immune system and may account for the “cytokine storm” that occurs in some severe cases of SARS-CoV-2 infection.

Tumor-infiltrating immune cells (TIICs) play a crucial role in tumorigenesis and progression in the tumor microenvironment (TME). Therefore, we first investigated whether CD209 expression was correlated with TIIC levels in various cancer types from TIMER. The results showed that the CD209 expression was evidently correlated with the infiltration of CD8⁺ T cells (Figure 5A), CD4⁺ T cells (Figure 5B), Treg cells (Figure 5C), B cells (Figure 5D), monocytes (Figure 5E), NK cells (Figure 5F), myeloid dendritic cells (Figure 5G), neutrophils (Figure 5H), macrophages (Figure 5I), and cancer-associated fibroblasts (Figure 5J) in pan-carcinoma patients.

The higher expression of CD209, an alternative receptor of SARS-CoV-2, was associated with lower OS in patients with LUSC (Figure 3). Therefore, we further explored the correlation between CD209 and immune storms in LUSC patients. The CD209 expression showed significant positive correlations with tumor purity and infiltrating levels of immune cells in LUSC patients (Figure 6). In particular, the CD209 expression was found to be positively correlated with infiltrating levels of B cells (r = 0.229, p = 4.46e-06), macrophages (r = 0.645, p = 2.25e-57), CD4⁺ T cells (p = 3.28e-22, r = 0.424), neutrophils (r = 0.383, p = 4.29e-18), Treg cells (p = 9.21e-25, r = 0.447), CD8⁺ T cells (p = 4.25e-20, r = 0.403), and dendritic cells (p = 9.83e-45, r = 0.568) (Figure 6A). These findings strongly suggested that CD209 might play a specific role in tumor immune infiltration, especially the recruitment of macrophages and dendritic cells in LUSC.

To further study the correlation between CD209 and a variety of infiltrating immune cells, we explored the relationship between CD209 and various immune markers in LUSC, including M1/M2 macrophages (Figures 6B,C), Treg cells (Figure 6D), and memory T-cell markers (Figure 6E). Interestingly, we found that the CD209 expression level was positively associated with the immune markers of tumor-associated macrophages (TAMs) and M2 macrophages but not M1 macrophages (Figure 6B). There were significant correlations between CD209 and M2 marker genes, such as CD163 (r = 0.766, p = 2.81e-93), MS4A4A (r = 0.723, p = 2.61e-78), VSIG4 (r = 0.709, p = 3.78e-74), MRC1 (p = 2.11e-94, r = 0.768), CLEC10A (P = 2e-26, r = 0.461), and PDCD1LG2 (r = 0.573, p = 2.26e-43) (Figure 6C). Moreover, positive correlations were observed between CD209 and the infiltration levels of Treg cells with TGFB1 (p = 5.9e-08, r = 0.245) and FOXP3 (r = 0.498, p = 3.02e-31) (Figure 6D). In addition, we also found significant correlations between CD209 and marker genes of T-cell exhaustion, such as CTLA4 (p = 4.43e-23, r = 0.432), PD-1 (r = 0.397, p = 2.02e-19), TIGIT (p = 3.03e-22, r = 0.424), HAVCR2 (p = 1.56e-59, r = 0.654), and LAG3 (p = 7.01e-19, r = 0.391) (Figure 6F). We further found that CD209 expression significantly correlated with the levels of a panel of inflammatory cytokines and chemokines, including IL-1 (r = 0.259, p = 1.01e-08), IL-6 (r = 0.269, p = 2.48e-09), IL-7 (r = 0.253, p = 2.25e-08), IL-33 (r = 0.128, p = 9.72e-10), CCL2 (r = 0.446, p = 9.54e-25), and CXCL12 (r = 0.538, p = 3.27e-37) (Figure 6G). A previous study revealed that ACE2 showed a weak correlation with immune infiltration compared to CD209 (Shu et al., 2021). We further found that CD209 had a weak correlation with ACE2 in LUSC (p = 4.29e-1, cor = −0.036) (Figure 6H). In summary, these results revealed that CD209 was specifically correlated with immune infiltrating cells in specific tumors, which might ultimately contribute to the immune storm in SARS-CoV-2-infected cancer patients.