Fat Study

bioRxiv preprint doi: https://doi.org/10.1101/2021.10.24.465626; this version posted October 25, 2021.
The copyright holder for this preprint

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
1 Title: SARS-CoV-2 infects human adipose tissue and elicits an inflammatory response
2 consistent with severe COVID-19
4 Authors:
5 Giovanny J. Martínez-Colón1†, Kalani Ratnasiri2†, Heping Chen1, Sizun Jiang5,9, Elizabeth
6 Zanley1, Arjun Rustagi1, Renu Verma1, Han Chen5, Jason R. Andrews1, Kirsten D. Mertz10,
7 Alexandar Tzankov8, Dan Azagury3, Jack Boyd4, Garry P. Nolan5, Christian M. Schürch7,
8 Matthias S. Matter8, Catherine A. Blish1,2,6†*, Tracey L. McLaughlin1†*
10 Affiliations:
1
11 Department of Medicine, Stanford University School of Medicine, Stanford, CA, USA
2
12 Program in Immunology, Stanford University School of Medicine, Stanford, CA, USA
3
13 Department of Surgery, Stanford University School of Medicine, Stanford, CA, USA
4
14 Department of Cardiothoracic Surgery, Stanford University School of Medicine, Stanford, CA,
15 USA
5
16 Department of Pathology, Stanford University School of Medicine, Stanford, CA, USA
6
17 Chan Zuckerberg Biohub, San Francisco, CA USA
7
18 Department of Pathology and Neuropathology, University Hospital and Comprehensive Cancer
19 Center Tübingen, Tübingen, Germany

8
20 Institute of Medical Genetics and Pathology, University Hospital of Basel, University of Basel,
21 Basel, Switzerland
9
22 Center for Virology and Vaccine Research, Beth Israel Deaconess Medical Center, Boston, MA,
23 USA
1
bioRxiv preprint doi: https://doi.org/10.1101/2021.10.24.465626; this version posted October 25, 2021. The copyright holder for this preprint
10
24 Institute of Pathology, Cantonal Hospital Baselland, Liestal, Switzerland
25
26 †Equal contributions
27
28 *Corresponding authors
29 Tracey L. McLaughlin
30 Department of Medicine, Division of Endocrinology, Gerontology, and Metabolism
31 300 Pasteur Dr,
32 Stanford, CA 94305
33 tmclaugh@stanford.edu
34
35 Catherine A. Blish
36 Department of Medicine, Division of Infectious Disease and Geographic Medicine
37 300 Pasteur Dr., Lane Building, L134
38 Stanford, CA 94305
39 cblish@stanford.edu
40
41
42
43
44
45
2
46 One sentence summary: Our work provides the first in vivo evidence of SARS-CoV-2 infection
47 in human adipose tissue and describes the associated inflammation.
48
49 Abstract: The COVID-19 pandemic, caused by the viral pathogen SARS-CoV-2, has taken the
50 lives of millions of individuals around the world. Obesity is associated with adverse COVID-19
51 outcomes, but the underlying mechanism is unknown. In this report, we demonstrate that human
52 adipose tissue from multiple depots is permissive to SARS-CoV-2 infection and that infection
53 elicits an inflammatory response, including the secretion of known inflammatory mediators of
54 severe COVID-19. We identify two cellular targets of SARS-CoV-2 infection in adipose tissue:
55 mature adipocytes and adipose tissue macrophages. Adipose tissue macrophage infection is
56 largely restricted to a highly inflammatory subpopulation of macrophages, present at baseline,
57 that is further activated in response to SARS-CoV-2 infection. Preadipocytes, while not infected,
58 adopt a proinflammatory phenotype. We further demonstrate that SARS-CoV-2 RNA is
59 detectable in adipocytes in COVID-19 autopsy cases and is associated with an inflammatory
60 infiltrate. Collectively, our findings indicate that adipose tissue supports SARS-CoV-2 infection
61 and pathogenic inflammation and may explain the link between obesity and severe COVID-19.
62
63 Main Text:
64 INTRODUCTION
65 The COVID-19 pandemic, caused by the viral pathogen SARS-CoV-2, has infected over 250
66 million people and taken the lives of over 4.5 million individuals globally as of October 2021.
67 The clinical manifestations of COVID-19 range from asymptomatic infection, mild cold-like
68 symptoms, to severe pulmonary and extrapulmonary manifestations characterized by extreme
3
69 inflammation and cytokine storm (1). Obesity has emerged as an independent risk factor for
70 infection, severe disease, and mortality (2–6). While obesity is associated with comorbid
71 conditions also related to severe COVID-19, the independent relative risk of obesity is higher
72 than that of hypertension and type 2 diabetes (2, 5, 7). Further, obesity is a risk factor even in
73 young adults and children who do not have other comorbid conditions (8). Several distinct
74 mechanisms could underlie this association. Impaired respiratory mechanics may result from a
75 heavy chest wall, airway resistance, and/or presence of obstructive sleep apnea (9). The
76 metabolic milieu in obesity, particularly among individuals with insulin resistance, is
77 characterized by systemic inflammation and hypercoagulability (10–12) and could thus stimulate
78 a more robust inflammatory response to SARS-CoV-2. Impaired immune responses to viral
79 infection are another possibility, as obese individuals exhibit altered immune cell profiles at
80 baseline (13) and in response to influenza infection (7, 14).
81
82 Furthermore, a recent report demonstrated that SARS-CoV-2 can infect adipocytes in vitro (15);
83 it is not known whether SARS-CoV-2 infects other adipose tissue-resident cells and/or drives an
84 inflammatory response in adipose tissue. Other viruses have been shown to infect several cell
85 types within adipose tissue, including adipocytes (influenza A virus), adipose-stromal cells
86 (adenovirus 36, human cytomegalovirus), macrophages (simian immunodeficiency virus (SIV)),
87 and T cells (human immunodeficiency virus (HIV))(16–19). Complex interactions between
88 various cell types and adipocytes can drive significant inflammation, with reports of adipocyte-
89 derived chemoattractants such as monocyte chemoattractant protein-1 (MCP-1) leading to
90 macrophage infiltration (20), tumor necrosis factor alpha (TNF-ɑ) activating nuclear factor
91 kappa B (21, 22), and free fatty acids driving toll-like receptor 4-mediated inflammation and
4
92 insulin resistance (23). Thus, pronounced and/or prolonged SARS-CoV-2 viral replication and
93 inflammation might occur in those with obesity and contribute to severe disease. Of particular
94 concern is the possibility that viral infection of peri-organ fat could contribute to organ damage
95 via inflammation and downstream processes such as extracellular matrix deposition/fibrosis,
96 edema, impaired cellular function, endothelial dysfunction, and hypercoagulability (11, 13). To
97 date, COVID-19 profiling studies have generally excluded adipose tissue from analyses; in other
98 cases adipose tissue may have been lumped in with analyses of adjoining tissues (24, 25) These
99 studies have shown that SARS-CoV-2 RNA and proteins are detected across numerous tissues,
100 including the lung, brain, intestine, and pancreas (24, 26, 27). Thus, we undertook a study to test
101 the hypothesis that SARS-CoV-2 infects cells within human adipose tissue and incites an
102 inflammatory response. We harvested adipose tissue from multiple depots in uninfected obese
103 humans for in vitro infection and obtained autopsy specimens of various adipose depots in
104 individuals who died from COVID-19. Our results clearly show SARS-CoV-2 infection in
105 macrophages and adipocytes from multiple adipose depots, with an attendant increase in
106 inflammatory profile.
107
108 RESULTS
109 Establishment of a system to examine SARS-CoV-2 infection in human adipose tissue
110 To explore the ability of human adipose tissue to support SARS-CoV-2 infection and drive an
111 inflammatory response, we recruited participants undergoing bariatric or cardiothoracic surgery.
112 Freshly harvested tissue from the subcutaneous (SAT), visceral (VAT), pericardial (PAT) and
113 epicardial (EAT) adipose tissue depots was subjected to collagenase digestion to separate
5
114 stromal-vascular cells (SVC) from mature adipocytes. Clinical characteristics are shown in table
115 S1.
116
117
6
118
119 Figure 1. Exposure of stromal vascular cells (SVC) from adipose tissue to SARS-CoV-2
120 supports infection of macrophages. (A) Sketch of workflow. SVC of human adipose tissue was
121 isolated by collagenase digestion prior to viral infection. SVC was infected or left untreated
7
122 (mock) with SARS-CoV-2 (USA-WA1/2020) at a multiplicity of infection (MOI) of 1. (B)
123 Relative gene expression of SARS-CoV-2 (N gene) obtained by 1-step RTqPCR.at 24 hpi or
124 mock infection (SAT, n=6; VAT, n=4). (C) Relative gene expression of SARS-CoV-2 (N gene)
125 obtained by 1-step RTqPCR in cultures were treated with vehicle (DMSO) or 10µM remdesivir
126 to inhibit viral replication and maintained for 24, 48, 72, and 96 hours before harvest. Relative
127 expression was analyzed by ΔΔCt method relative to a mock sample using 18S rRNA as a
128 housekeeping gene. Each data point is an average of 3 technical replicates. (D) Frequency of
129 SARS-CoV-2 infected cells based on SARS-CoV-2 N protein detection by flow cytometry of
130 SAT (left, n=2) and VAT (right, n=2). Gating is detailed in supplemental figure 3. Statistical
131 analysis: (A) paired, two-sided, student’s t-test. **P<0.01, ***P<0.001. (B) Statistical analysis
132 was performed with a two-way ANOVA, multiple comparisons using statistical hypothesis
133 Sidak. *P<0.05. Data is presented as ± mean s.e.m. (A) Sketch was created with BioRender.
134
135 Human adipose tissue SVC supports SARS-CoV-2 infection
136 Freshly isolated SVC from human adipose tissue were exposed to SARS-CoV-2 (WA-01) at a
137 multiplicity of infection (MOI) of 1 or mock-infected. Infection and replication were assessed by
138 reverse transcription quantitative PCR (RTqPCR) for SARS-CoV-2 nucleocapsid (N) gene and
139 by flow cytometry for SARS-CoV-2 N protein (Figure 1A). Both genomic (Fig. 1B) and
140 subgenomic (fig. S1) viral RNA were detected in SARS-CoV-2 infected SVC from subcutaneous
141 (n=6 Fig.1B; n=2 fig. S1), omental (n=4 Fig.1B; n=2 fig. S1), pericardial (n=1 fig. S1), and
142 epicardial (n=1 fig. S1) adipose tissue. To confirm that N-gene detection represented active viral
143 infection, we cultured infected cells in the presence of remdesivir, an inhibitor of the viral RNA-
144 dependent RNA polymerase. We first confirmed the ability of remdesivir to inhibit viral RNA
8
145 accumulation in A549-ACE2 cells (fig. S2). In SARS-CoV-2-infected SVC from both SAT and
146 VAT, we observed accumulation of N-gene copies over the course of 96 hours, indicating viral
147 RNA replication (Fig. 1B). Importantly, remdesivir significantly reduced viral RNA levels at
148 later time points in both tissue types (Fig. 1C). Together, these data indicate that SVC can
149 support SARS-CoV-2 infection.
150
151 Next, to determine which of the SVC are infected, we performed flow cytometry to assess viral
152 N protein expression in six major SVC types (dendritic cells, macrophages, B cells, T cells,
153 preadipocytes, and endothelial cells) (Fig. 1A and fig. S3). In both SAT and VAT, SARS-CoV-2
154 N protein was primarily restricted to macrophages (defined as large granular
155 CD45+CD14+CD11c- cells) (Fig.1D). We next assessed the level of angiotensin-converting
156 enzyme 2 (ACE-2), the cellular receptor for SARS-CoV-2 (28) in the SVC. ACE-2 protein
157 expression was very limited in SVC, with no detectable expression in the VAT and very low-
158 level expression on ~3% of SAT macrophages when compared to isotype control (fig. S4). In
159 addition, we did not detect ACE2 mRNA in any of the six samples collected from SAT, VAT,
160 PAT or EAT (table S2). Thus, these data indicate that SARS-CoV-2 primarily infects
161 macrophages in SVC and may enter through a non-canonical entry receptor.
162
163
9
164
165 Figure 2. A subset of macrophages is infected with SARS-CoV-2. (A) Schematic of
166 experiment. The stromal vascular cells (SVC) were isolated from the SAT and VAT depots of
167 three different participants and infected with either mock or SARS-CoV-2 (MOI of 1.0). Each
168 sample was collected for scRNA-seq at 24 hpi. (B-C) UMAP representation of the SVC from all
169 participants (n=3) across 198,759 cells, (B) colored by manually annotated cell type and (C)
170 colored by infection and depot. (D) UMAP representation of all cells colored by SARS-CoV-2
171 cpm (log10). (E) Violin plot reveals the SARS-CoV-2 cpm values of all cells across SARS-CoV-
10
172 2 infected samples, showing only the 8 cell clusters with the highest composition of SARS-CoV-
173 2+ cells. Percentages above each cell type denote the percentage of cells within each cluster that
174 have over 10 SARS-CoV-2 reads. (F) UMAP projections of all macrophages from the scRNA-
175 seq dataset, colored by SARS-CoV-2 cpm (left), colored by CD68 expression (middle) and
176 colored by ACE2 expression (right). (G) Dotplot of the proportion of cells (dot size) in the
177 macrophage clusters split by infection condition expressing genes relevant for SARS-CoV-2
178 entry and antiviral defense, as well as SARS-CoV-2 cpm and macrophage cluster markers and
179 colored by scaled average expression. (A) Figure was created with BioRender.
11
180
181 Figure 3. The infected macrophage cluster is marked by increased chemokine expression.
182 (A) Volcano plot of the differentially expressed genes between macrophage clusters 2 (C2) and
183 12 (C12) across mock-infected samples. (B) Venn Diagram comparison of the significantly
184 differentially expressed genes (DEGs) and their direction of change across C2 versus C12 in (1)
185 mock-infected, (2) all SAT and (3) all VAT, and (4) SARS-CoV-2-infected conditions. (C)
12
186 Heatmap of the most significant DEGs between SARS-CoV-2+ versus bystander macrophages
187 within C2. (D) Normalized enrichment scores of top Reactome pathways, using significant
188 DEGs between SARS-CoV-2+ versus bystander C2 macrophages.
189
190 SARS-CoV-2 infects a distinct subset of macrophages
191 To comprehensively characterize the cells infected with SARS-CoV-2 in human adipose tissue,
192 we performed single-cell RNA sequencing (scRNA-seq) of SARS-CoV-2 and mock-infected
193 SVC isolated from both SAT and VAT from three bariatric surgery patients (Fig. 2A, table S3).
194 We generated 198,759 single-cell expression profiles. After performing dimensionality reduction
195 and Harmony batch integration (29), unbiased graph-based clustering resulted in 23 cell types
196 (C0-C22) with marker genes that aligned with previous datasets and included three major cell
197 groups: preadipocytes, immune cells, and endothelial cells (30, 31) (Fig. 2B, table S3, fig. S5).
198 We saw the greatest heterogeneity across the preadipocyte population for which we identified 14
199 distinct clusters, each of which was labeled by its top two cluster defining genes. We similarly
200 labeled two distinct macrophage clusters.
201
202 We detected SARS-CoV-2 at levels ranging up to 1079 transcripts per cell in a total of 8.8% of
203 cells in the infected samples (fig. S6) and did not detect a single SARS-CoV-2 transcript in the
204 mock-infected samples. The uniform manifold approximation and projection (UMAP) of SARS-
205 Cov-2-exposed samples colored by viral counts-per-million (cpm) showed that macrophages
206 contained the highest concentration of SARS-CoV-2 transcripts (Fig. 2D), with >85% of SARS-
207 CoV-2 exposed C2-macrophages and >11% C12-macrophages containing more than ten viral
208 transcripts (Fig. 2E). ACE2 transcript was not detected in the SVC macrophages and therefore
13
209 could not explain the difference in infection levels between C2-macrophages and C12-
210 macrophages (Fig. 2F, 2G). However, C2-macrophages expressed higher levels of BSG, NRP1/2,
211 CTSB and CTSL compared to C12-macrophages (Fig. 2G), all previously noted to be important
212 to SARS-CoV-2 infection and viral processing (32–34). Our data suggest that these alternative
213 entry receptors could contribute to SARS-CoV-2 entry into C2-macrophages.
214
215 To further evaluate how these two macrophage populations differ in SARS-CoV-2 susceptibility,
216 we examined the most differentially expressed genes between the macrophage cluster that was
217 highly enriched for SARS-CoV-2 reads (C2) versus the relatively uninfected cluster (C12) at
218 baseline (mock-infected) and identified 1000 significant differentially expressed genes (Fig 3A,
219 table S4). Notably, even in the mock condition, there was a significant upregulation of
220 inflammation related transcripts within C2 compared to C12, and in particular, transcripts for
221 numerous chemokine ligands (Fig. 3A, fig. S7B, table S6-7). Increased expression of CCL8 and
222 CCL3, chemokines involved in monocyte chemotaxis, highlight the potential of these infected
223 cells to induce the infiltration of inflammatory monocytes and macrophages (35, 36). There was
224 significant downregulation of 14 HLA genes across HLA Class I and Class II proteins, including
225 HLA-DRB1, HLA-DRA, HLA-DQB1, and HLA-DPB1, in the C2-macrophages compared to C12-
226 macrophages (Fig. 3A, table S4). Additionally, the C2-macrophage cluster demonstrated
227 enrichment for transcripts associated with perivascular macrophages (PVM) including CD163,
228 SELENOP, MARCO and APOE (Fig. 3B and (31, 37, 38)). In contrast, the relatively uninfected
229 C12 demonstrated upregulation of genes relating to lipid metabolism and other markers of lipid-
230 associated macrophages, including FABP4, TGM2, GSN, TREM2, LPL and PGD, suggesting that
14
231 this macrophage population might be implicated in lipid accumulation and trafficking and
232 derived from metabolic activation (30, 31, 37, 39).
233
234 SARS-CoV-2 infection drives an inflammatory response in macrophages
235 As major differences between the macrophage clusters include markers of inflammation (i.e.
236 IL1B, CXCL8, CCLs) (Fig. 3A-B), this raises the possibility that in vitro SARS-CoV-2 infection
237 drives the formation of the C2-macrophage cluster. However, this cluster is present in mock-
238 infected samples, and both SARS-CoV-2-infected and mock-infected samples displayed a
239 similar array of differentially expressed genes between these clusters (Fig. 3B, fig. S7A),
240 indicating that this inflammatory C2-macrophage cluster is naturally present in the SVC of
241 adipose tissue and is more susceptible to infection in vitro.
242
243 To explore the cell-intrinsic effects of infection, we compared the transcriptional profiles of C2-
244 macrophages that did (infected) or did not (bystander) contain SARS-CoV-2 transcripts within
245 the infected cultures (fig. S7C, Fig. 3C, table S5). Though there were baseline differences in
246 gene expression between participants, several genes strongly distinguished SARS-CoV-2+ C2
247 cells from bystander C2 cells by hierarchical clustering (Fig. 3C); many of these genes also
248 distinguish C2-macrophages from C12-macrophages (fig. S7D). For instance, CCL4, CCL8, and
249 CCL3 genes were significantly enriched in the infected cells within C2, and LYZ and THBS1
250 were significantly downregulated within the infected cells. As these genes contributed to
251 defining the C2 population from C12 cells, this result suggests that SARS-CoV-2 infection may
252 drive C2 features to further extremes. Pathway analysis of significant DEGs between the SARS-
253 CoV-2+ cells versus bystanders highlights the enrichment of pathways associated with IL-10
15
254 signaling (Fig. 3D, table S8), which is consistent with upregulation of this pathway in monocytes
255 of severe COVID-19 patients (40). Chemokine and other innate immune system-associated
256 pathways are also highly enriched in infected cells (Fig. 3D), which has been previously
257 demonstrated (41). Interestingly, there is a reduced enrichment in pathways related to translation
258 machinery of the host, which is supported by SARS-CoV-2 studies suggesting that SARS-CoV-2
259 infection has the ability to reshape translation, splicing, protein homeostasis and nucleic acid
260 metabolism pathways (42, 43). Thus, the macrophages with detectable SARS-CoV-2 RNA
261 display a dramatic transcriptional response with upregulation of inflammatory pathways.
262
16
263
264 Figure 4. Preadipocytes respond to SARS-CoV-2 exposure. (A, B) UMAP embedding of all
265 preadipocytes (n=140,867) colored by (A) cluster and (B) sample and infection type. (C) Cell
266 fraction bar plot clustered by sample and infection type within each cluster. (D) Feature plots
267 depicting expression of selected markers associated with preadipocyte cell states, cell types and
268 antiviral genes. (E) Box plots of average cytokine (top) and ISG- (bottom) module scores across
269 the preadipocytes of each participant and depot in both mock and SARS-CoV-2 infection
270 conditions. (F) Reactome pathway analysis was performed on the significant DEGs by
17
271 participant and cluster within SAT. Pathways that were represented and significant in at least
272 four of the participant-cluster subsets were included. Pathways clustered by Euclidean distance
273 (tree not shown) and split by the two major subtrees.
274
275 Adipocyte progenitors demonstrate an inflammatory response to SARS-CoV-2 infection of
276 macrophages
277 Given that exposure of preadipocytes to inflammatory cytokines such as TNF-α or IL-6 can alter
278 their phenotype (31, 44), we next evaluated whether SARS-CoV-2 infection of SVC led to a
279 bystander activation of preadipocytes. We embedded only preadipocytes and used unsupervised
280 clustering to identify 17 unique clusters (P0-P16) (Fig. 4A-C, fig. S8A, table S9). Notably, SAT
281 and VAT preadipocytes were highly transcriptionally unique, with 11/17 clusters each composed
282 of over 75% of cells derived from only one depot. Overall, there was not a dramatic shift in
283 cluster composition of SARS-CoV-2-infected and mock-infected samples, except for cluster P15
284 which was highly enriched with preadipocytes from SARS-CoV-2-infected SAT (Fig. 4C). This
285 cluster was highly inflammatory, with dramatic upregulation of two antiviral interferon
286 stimulated genes (ISGs), IFIT1 and ISG15 (fig. S8A), suggesting a relatively robust response to
287 infection in the SAT. The remaining clusters were distinguished by several unique DEGs (fig.
288 S8A, table S9), including CXCL14 and APOD in cluster P9, indicating enrichment with
289 adipocyte stem cells (ASCs) (30, 31, 45); KRT18 and MSLN in clusters P1 and P4 indicating
290 enrichment with mesothelial cells in VAT only, as previously reported (30, 45, 46). Further,
291 VAT-predominant clusters demonstrated higher levels of IFI27 (a marker of SARS-CoV-2
292 infection in the blood) than other clusters while IL-6 expression was high across all
18
293 preadipocytes in mock-conditions, highlighting the highly inflammatory nature of preadipocytes
294 at baseline.
295
296 To more globally investigate the effects of SARS-CoV-2 infection on SAT and VAT
297 preadipocytes, we looked at the total number of DEGs upon SARS-CoV-2 infection across each
298 participant’s SAT and VAT and noticed that for each participant, the SAT had a higher number
299 of DEGs than the paired VAT sample, suggesting that the SAT had a stronger transcriptional
300 response than VAT (fig. S8B). We then used known ISGs and cytokine genes to generate ISG
301 and cytokine gene scores for each participant, infection condition and depot (Fig. 4E, table S14).
302 Infection induced both ISGs and cytokines, but SAT demonstrated a greater increase in ISG and
303 cytokine responses than did VAT. These data also revealed that in these three participants, the
304 VAT depot has a higher baseline cytokine expression that is further increased upon SARS-CoV-
305 2 exposure (Fig. 4D, fig. S9A).
306
307 Next, we wanted to see whether particular preadipocyte clusters within the SAT and VAT
308 showed a stronger response to SARS-CoV-2 exposure. We employed perturbation analysis
309 previously described (47, 48) to identify which cluster within each depot and participant
310 demonstrated the strongest changes across infection conditions (fig. S8C, S8D). Interestingly,
311 across both depots, the ASC-like cluster, P9, along with nearby clusters, P16 and P12, showed
312 the lowest perturbation relative to other clusters. While donors varied in their most highly
313 perturbed clusters, this analysis demonstrated that SAT showed consistently stronger
314 perturbation than VAT upon SARS-CoV-2 infection. Therefore, we then focused on the SAT to
315 understand the particular pathways by cluster that contribute to the overall inflammatory
19
316 response in the depot. To do this, we identified the DEGs between SARS-CoV-2 exposed versus
317 mock preadipocytes by cluster, depot, and participant and employed gene set enrichment analysis
318 using the Reactome database to identify relevant pathways to understand the response within
319 each participant and each cluster (Fig. 4F, table S10-13). Clusters P5 and P2 consistently showed
320 transcriptional enrichment for genes relevant to an immune response, more specifically related to
321 cytokine signaling. Additionally, pathway analysis by cluster and participant shows that in
322 participant 1 and 2, which seemed to have a more dramatic response to SARS-CoV-2 infection,
323 we see most of the adipose clusters showing enrichment in responses relevant to the immune
324 system. Participant 3 only had 2/17 clusters revealing enrichment for immune-related pathways.
325 Negatively enriched pathways relate to viral transcription and translation pathways which
326 suggests that preadipocytes are also expressing genes to suppress viral production. A similar
327 analysis was also performed on the VAT depot which demonstrated a similar enrichment for
328 immune response pathways across preadipocyte clusters, with participant 1 eliciting more
329 dramatic changes across their preadipocyte compartment (fig. S8D). Overall, these data indicate
330 that SARS-CoV-2 infection within the SVC macrophages drives inflammatory responses in the
331 neighboring preadipocyte cells.
332
333
334
335
336
337
338
20
Table 1: 80-plex Luminex analysis of Stromal Vascular Compartment from human adipose tissue
Results
SAT (n=8) VAT (n=6)
Analyte Category Mean p value/ q Mean p value/ q Known role in COVID-19 References
Fold value Fold value
change change
IP-10 Chemokine 19.3 0.01/0.06 17.04 0.03/0.18 Elevated in acute cases and critically (49–57)
ill patients; associated with disease
severity and ICU admission;
predictor of disease progression;
biomarker of impaired T cell
response in acute infection
MCP2 0.74 0.02/0.1 1.12 0.03/0.18 Elevated in critically ill patients (56)
MCP3 1.32 0.04/0.2 0.97 0.84/1 Elevated in serum and blood; (50, 53–
associated with disease severity and 56)
predictor of disease progression
CXCL13 1.19 0.02/0.1 0.96 0.63/0.98 Elevated in serum (58)
CCL27 0.99 0.74/0.97 1.12 0.03/0.18 Secreted by HeLa cells in response (59)
to SARS-CoV-2’s ORF7a
PDGFAA Growth 5.41 0.01/0.06 4.06 0.03/0.18 Associated with severe disease and (51)
Factor ICU admission
PDGFAB 1.12 0.01/0.06 1.11 0.09/0.36 Elevated in serum and associated (51, 55)
/BB with ICU admission
VEGF 3.81 0.01/0.06 2.52 0.03/0.18 Elevated in serum; role in COVID- (55, 60)
19 brain inflammation
21
MCSF 2.57 0.01/0.06 2.07 0.03/0.18 Elevated in serum (52)
GM-CSF 1.81 0.08/0.31 1.49 0.03/0.18 Elevated in serum, including in (52–55)
mechanically ventilated patients
LIF 0.57 0.01/0.06 1.44 0.03/0.18 Proposed to treat COVID-19 patients (61)
by decreasing IL-6
TGF-α 2.01 0.01/0.06 1.81 0.03/0.18 Not known role
IL-4 Th2 1.96 0.01/0.06 2.02 0.03/0.18 Elevated in serum (51, 55)
Cytokines
IL-13 1.13 0.08/0.31 1.22 0.03/0.18 Elevated in serum (51)
MIF Inflammat 12.16 0.01/0.06 10.81 0.03/0.18 Elevated in critically ill patients (62)
ory
IL-21 1.08 0.01/0.06 0.92 0.25/0.69 Elevated in serum compared to non- (63)
cytokines
COVID-19 pneumonia patients
TNF-β 1.17 0.01/0.06 1.23 0.03/0.18 Not Known role
sICAM1 Cell 0.87 0.04/0.2 0.94 1/1 Elevated in non-survivor compared (64)
adhesion to survivor
SCF HRC 0.9 0.02/0.13 1.08 0.03/0.18 Not known role
Abbreviations: Interleukin (IL), Interferon gamma inducible protein 10 (IP-10), Macrophage colony stimulation factor
(MCSF), Platelet-derived growth factors (PDGF), Transforming growth factor alpha (TGFα), Tumor necrosis factor beta
(TNF-β), Vascular endothelial growth factor (VEGF), Macrophage migration inhibitory factor (MIF), Granulocyte colony
stimulation factor (GM-CSF), Leukemia inhibitory factor (LIF), Monocyte chemoattractant protein 2 (MCP2), C-X-C motif
ligand (CXCL), Stem cell factor (SCF), C-C motif chemokine ligand (CCL), Soluble intercellular adhesion molecule-1
(sICAM1), Hematopoiesis regulatory cytokine (HRC)
Statistical analysis: Wilcoxon rank test corrected for multiple testing by the Benjamini-Hochberg method. False discovery rate
of 10%. Numbers were rounded to their significant digits.
339
22
340 Table 1. 80-plex Luminex analysis of stromal vascular cells from human adipose tissue
341 exposed to SARS-CoV-2. Table summarizing results of 80-plex Luminex assay performed in
342 supernatants of SVC (SAT, n=8; VAT, n=6) cultures that were infected with SARS-CoV-2 or
343 mock-infected for 24 hours. First column shows analytes of interest. Second column categorizes
344 analytes in either chemokine, growth factor, Th2 cytokine, inflammatory cytokine, cell adhesion
345 molecule, or hematopoiesis regulatory cytokine. Columns 3 to 6 summarizes statistical results
346 splitted by mean fold change and p and q (Adjusted p value) value obtained from Wilcoxon
347 signed rank test. Columns 7 and 8 summarizes the known role of each analyte in COVID-19.
348 Bottom of the table contains abbreviations and descriptions of statistics.
349
350 SARS-CoV-2 infection of SVC induces inflammation
351 To better understand the impact of SARS-CoV-2-mediated infection on inflammatory responses
352 within the SVC, we measured the relative gene expression of IL-6 by RTqPCR (fig. S11A) in
353 SARS-CoV-2 infected versus mock-infected SVC. We selected IL-6 due to its high presence in
354 severe cases of COVID-19 (65), its association with low-grade inflammation in obese
355 individuals (66), and its likely role in COVID-19 immunopathogenesis (67). SARS-CoV-2-
356 infection drove significantly increased IL-6 expression in SAT (n=6) (fig. S11A). IL-6 transcripts
357 were also increased in VAT (n=6), but the change was not statistically significant (fig. S11A).
358 Additionally, we measured gene expression of five interferon-related genes (IFNA1, IFNB1,
359 ISG15, IFI27, and IER3) in SAT and VAT infected with SARS-CoV-2 over a period of 96 hpi
360 (fig. S10). We identified increased expression of IFNA1, IFNB1, and ISG15 as viral RNA
361 accumulated in both SAT and VAT demonstrating induction of antiviral response (fig. S10).
362
23
363 To better understand the impact of SARS-CoV-2 infection on inflammatory responses within the
364 SVC, we tested supernatants from infected versus mock-infected cultures for protein expression
365 of various secreted inflammatory markers by Luminex. As compared to mock infection, SARS-
366 CoV-2 infection of SAT and VAT resulted in upregulation of several chemokines, cytokines,
367 growth factors, and other inflammatory mediators (Table 1). In both SAT and VAT, the elevated
368 production of IP-10 is particularly noteworthy as this cytokine has been reported to be elevated
369 in the serum of critically ill COVID-19 patients (49). Additionally, we found the platelet-derived
370 growth factor (PDGF)-AA, and AB/BB (PDGFAB/BB) together with the type 2 (TH2) immune
371 factor interleukin-4 (IL-4) to be elevated in SARS-CoV-2 infected SAT. Similarly, PDGF-AA,
372 IL-4, and IL-13 (another TH2 factor) are elevated in the SARS-CoV-2 infected VAT. Together,
373 these data suggest that adipose tissue could contribute to the secretion of the cytokines and
374 vascular factors in the setting of COVID-19 infection.
375
376
377
24
378
379
380
381 Figure 5. Mature and in vitro differentiated adipocytes support SARS-CoV-2 infection.
382 Mature adipocytes (MA) of human adipose tissue were isolated by collagenase digestion prior to
383 viral infection. MA were infected or left untreated (mock) for 24 hours with SARS-COV-2
384 (USA-WA1/2020) at a MOI:1. (A) Measurements of genomic (gRNA) and subgenomic
385 (sgRNA) SARS-CoV-2 genome copy numbers infected MA from subcutaneous (left; n=3) and
386 visceral (right; n=2 omentum; n=1 epicardial; n=1 pericardial) adipose tissue obtained by
387 absolute gene quantification using 1-step RTqPCR TaqManTM and reported as fold change of
388 infected to mock). (B) Relative gene expression of N gene in adipocytes differentiated in vitro
25
389 from pericardial preadipocytes using adipocyte differentiation media for 0, 3, 6, and 13 days
390 before left untreated or infected with SARS-CoV-2 at a MOI:1. Results were obtained by 1-step
391 RTqPCR and analyzed by ΔΔCt method using 18s as a housekeeping gene. (C) Adipocytes
392 differentiated in vitro from preadipocytes obtained from pericardial adipose tissue were infected
393 or mock infected with SARS-COV-2 (USA-WA1/2020) at a MOI of 1 for 1 hour, followed by
394 washing and removing the virus and replacing with media treated with vehicle (DMSO) or 10µM
395 remdesivir to inhibit viral replication. Cultures were maintained for 24, 48, 72, and 96 hpi, after
396 which gene expression was obtained by 1-step RTqPCR for the N gene. Relative expression was
397 analyzed by ΔΔCt method relative to the mock sample. (C) Each data point is an average of 3
398 technical replicates and are presented as ± mean s.e.m.
399
400 Mature and in vitro differentiated adipocytes support SARS-CoV-2 infection
401 We next explored whether mature adipocytes could support SARS-CoV-2 infection. Mature
402 adipocytes were infected with SARS-CoV-2 or mock infected. We detected both genomic and
403 subgenomic SARS-CoV-2 (N gene) in infected mature adipocytes from both SAT (n=3) and
404 VAT (n=2 omentum; n=1 EAT; n=1 PAT) by absolute gene quantification 24 hpi (Fig. 5A). We
405 also measured secreted inflammatory mediators by Luminex in these infected mature fat samples
406 (fig. S9A, S9B). There was a trend for increased secretion of VEGF, PDGFAA, PDGFAB/BB
407 and IP-10 upon infection, but no significant differences in cytokines levels (fig. S9A-B).
408
409 We next evaluated infection in differentiated preadipocytes (68). As a control of successful
410 differentiation, we confirmed an increase in gene expression of fatty acid binding protein 4
411 (FABP4) (69) on differentiated adipocytes (table S2B). ACE2 expression was not detected in
26
412 undifferentiated preadipocytes (day 0) but was detected by day 3 in SAT, VAT, and PAT (table
413 S2B). Adipocytes differentiated from PAT were infected at days 3, 6, and 13 of differentiation.
414 We detected SARS-CoV-2 (N gene) at 24 hpi in all differentiated adipocytes (Fig. 5B).
415 Adipocytes differentiated from SAT had detectable genomic and subgenomic SARS-CoV-2
416 RNA detected 24hpi on adipocytes at day 0 and day 3 of differentiation (fig. S9C-D). To
417 determine whether SARS-CoV-2 can replicate in differentiated adipocytes, we infected day 13
418 differentiated adipocytes, and cultured them with either vehicle or remdesivir. We observed
419 increased viral reads in differentiated adipocytes by 72 hpi; this viral RNA replication was
420 partially inhibited by remdesivir (Fig. 5C). These results show that both freshly isolated mature
421 adipocytes and adipocytes differentiated in culture can be infected by SARS-CoV-2 and
422 represent an additional replication site.
423
424
425
27
426
427 Figure 6. SARS-CoV-2 RNA and immune infiltration are present in adipose tissue of
428 autopsy samples from COVID-19 patients. RNA in situ hybridization (ISH) on epicardial fat
429 from heart autopsies from patients (Top, autopsy #9; Bottom, #10) who succumbed to COVID-
430 19. Assays were performed using probes against SARS-CoV-2 Spike mRNA. Red arrowheads
28
431 show ISH positive signals and blue arrows show inflammatory cells. (A and C) Overview of the
432 heart tissue section (2mm), and (B and D) magnified view (20um) of the represented region. (E)
433 Interface of epicardial fat and myocardium (50um). Note the inflammatory infiltration (blue
434 arrows) only in the epicardial fat. Image has been rotated 90°.
435
436 SARS-CoV-2 is detected in the adipose tissue from deceased COVID-19 patients.
437 We next sought to find evidence for SARS-CoV-2 infection of adipose tissue in vivo. We
438 performed RTqPCR for SARS-CoV-2 genes in lung, fat, heart, and kidney tissue samples in a
439 cohort of eight COVID-19 autopsy cases (Table 2; autopsy # 1-8). Studies assessing SARS-CoV-
440 2 levels in the lungs and heart of several subjects here have been previously published (70–72).
441 In addition to detecting SARS-CoV-2 in the lung and heart, we also detected SARS-CoV-2 in
442 epicardial, visceral, and subcutaneous adipose tissue and kidney (Table 2). As expected, SARS-
443 CoV-2 detection was highest in the lung (mean Ct=25). Interestingly, it was followed by adipose
444 tissue (mean Ct=33), heart (mean Ct=34) and kidney (mean Ct=34). In addition, we performed
445 RNA in situ hybridization (ISH) on epicardial fat (Fig. 6). We detected SARS-CoV-2 in the
446 cytoplasm of adipocytes in the epicardium and not in immune cells or myocardium (Fig. 6A-D).
447 Mononuclear cell infiltration, mainly consisting of lymphocytes, is observed in areas showing
448 positive SARS-CoV-2 signal in epicardium (Fig. 6A-E). Our results provide evidence of SARS-
449 CoV-2 detection in the adipose tissue of COVID-19 autopsies, indicating that adipose tissue
450 tissue can harbor SARS-CoV-2 infection and contribute to pathogenic inflammation.
451
452
453
29
Table 2: Evidence of SARS-CoV-2 detection in adipose tissue from deceased individual
RTqPCR for SARS-CoV-2 (Ct
values)
Autop Sex Age Ethnic BMI DHBD Comorbidities Lung Heart Kidney Adipose Type of
sy # (years) ity (kg/m2) (Days) adipose
analyzed
1 F 84 C 24 12 CAD, HTN, UD UD UD UD Epicardia
metastatic lung CA, l
T2D, AKI,
dementia
2 M 85 C 26 5 HTN, dyslipidemia, 23 33 33 33 Visceral
T2D, prostate CA
in remission, OSA,
cirrhosis, leukemia
in remission
3 F 81 C 26 4 CAD, HTN, PAD, 22 34 36 36 Visceral
AKI, former
smoker
4 M 71 C 24 NH PAD, CAD, 27 33 UD 33 Visceral
valvular heart
disease, AKI
5 M 65 C 26 8 Dyslipidemia, 23 34 34 33 Epicardia
CAD, COPD, l
HTN, T2D, OSA,
AKI
6 M 89 C N/A 5 CAD, 21 31 N/A 28 Epicardia
cardiomyopathy, l
dyslipidemia, T2D,
30
OSA, former
smoker, COPD
7 M 85 N/A 30 5 Former smoker, 26 35 UD 34 Subcutan
HTN, eous
cardiomyopathy,
COPD, AKI,
dementia
8 M 58 C 47 7 Former smoker, 35 37 33 36 Subcutan
HTN, obesity, eous
CAD, CHF, AKI
9* F 78 C 35.2 NH obesity, AV-block N/A N/A N/A 34 Epicardia
III with pacemaker, l
HTN, cerebellar
infarcts
10* F 87 C N/A 1 Dementia, obesity, N/A N/A N/A 32 Epicardia
CAD, HTN, T2D l
Note: All Ct values presented are an average of ORF1ab, S, and N genes. All heart RNA samples were collected from a
mixture of myocardium and epicardium. Any Ct value below 37 was considered a true SARS-CoV-2 signal. Autopsies labeled
with (*) means cases were from Tübingen and not part of the Basel cohort.
Abbreviations: not available (N/A), undetected (UD), threshold cycles (Ct), open reading frame 1ab (ORF1ab), spike (S),
nucleocapsid (N), days of hospitalization before death (DHBD), not hospitalized (NH), not available (N/A), Caucasian (C),
male (M), female (F), body mass index (BMI), coronary artery disease (CAD), hypertension (HTN), type 2 diabetes mellitus
(T2D), cancer (CA), acute renal insufficiency (AKI), obstructive sleep apnea (OSA), peripheral artery disease (PAD), chronic
obstructive pulmonary disease (COPD), congestive heart failure (CHF).
454
455 Table 2. Evidence of SARS-CoV-2 detection in adipose tissue from deceased individuals.
456 Table summarizing SARS-CoV-2 PCR signals from 8 autopsy cases, seven of them from
457 patients deceased from COVID-19. All threshold cycles (Ct) presented are an average of three
31
458 SARS-CoV-2 gene sequences for open reading frame 1ab (ORF1ab), spike (S), and nucleocapsid
459 (N). RNA was isolated from either epicardial, visceral, or subcutaneous fat. Heart RNA samples
460 were collected from a mixture of myocardium and epicardium. Samples not presented in this
461 study were marked as not available “N/A”. When the PCR signal was undetected, the sample
462 was marked as “UD”.
463
464 DISCUSSION
465 Here, we report that SARS-CoV-2 infects human adipose tissue in COVID-19 patients and in
466 vitro. Significantly extending a prior report that human adipocytes can be infected in vitro (15),
467 we identify both tissue-resident macrophages and mature adipocytes as target cells for SARS-
468 CoV-2 infection (16, 17, 19). Additionally, this study reveals that in vitro infection leads to
469 activation of inflammatory pathways in macrophage and preadipocytes and the secretion of
470 inflammatory factors associated with severe COVID-19. Finally, we provide the first evidence
471 that this may be relevant to disease pathogenesis in humans in vivo, as we detected SARS-CoV-2
472 RNA in VAT, EAT, and SAT of COVID-19 autopsies. We also demonstrate histologic evidence
473 of inflammation adjacent to viral signals in adipose tissue in an autopsy sample. Together, this
474 in-depth analysis of adipose susceptibility and inflammatory response to SARS-CoV-2 infection
475 suggests that adipose tissue may serve as a potential reservoir for SARS-CoV-2 and potentiator
476 of systemic and regional inflammation, possibly contributing to severe clinical disease in obese
477 individuals infected with SARS-CoV-2.
478
479 One of our most intriguing findings was that SARS-CoV-2 infection of adipose SVC was
480 primarily restricted to just one of two macrophage clusters. The two macrophage populations
32
481 were defined as: C2-macrophages, an inflammatory cluster with IL1B and CCL3 as its most
482 distinguishing transcripts, and C12-macrophages, characterized by significant enrichment of
483 HLA class II transcripts and LYZ and alarmin S100A8 expression. C2-macrophages were
484 predominately infected, with over 85% of C2 cells within infected SVC carrying SARS-CoV-2
485 transcripts. These two clusters do not fall into a classical M1 and M2 classification: instead,
486 macrophages activated in response to adipose tissue signals may exhibit a unique phenotype (39,
487 73). The C2 macrophages express transcripts characteristic of PVMs, which have been
488 previously described in adipose tissue (31) and are highly phagocytic (74). PVMs can become
489 virally infected: in the brain, PVMs have been shown to become infected by HIV and SIV (38,
490 75). Interestingly, SARS-CoV-2 infection further drives the C2/C12 dichotomy as chemokine
491 markers are further upregulated while LYZ and HLA expression is further downregulated in
492 infected SARS-CoV-2 transcript-containing C2 cells versus their bystander counterparts. LYZ is
493 of particular interest as prior studies have shown that lysozyme can play a role in antiviral
494 activity (76–78), which suggests that low lysozyme levels (as found in the susceptible C2
495 cluster) may be advantageous for the virus. For example, influenza infection reduces lysozyme
496 secretion in neutrophils (79) and the combination of lysozyme and lactoferrin therapy reduces
497 Bovine viral diarrhoea virus titers in vitro (76, 78). This data suggests that lysozyme therapy
498 could reduce SARS-CoV-2 levels in the adipose tissue. We also observed significant reduction in
499 HLA Class I and II genes relevant to antigen presentation in the infected C2-macrophages
500 cluster, consistent with findings in blood monocytes in severe COVID-19 disease (80, 81) which
501 reduces immune surveillance and targeting of infected host cells, another infection-induced
502 change which could be advantageous for virus propagation. Together, these data highlight the
503 pivotal role macrophages may play in disease pathogenesis and the need for a better
33
504 understanding of how tissue macrophages at different sites are infected by and control SARS-
505 CoV-2 infection.
506
507 The role of macrophages in supporting SARS-CoV-2 replication has been a subject of significant
508 interest as there is no consensus about whether macrophages support viral replication. SARS-
509 CoV-2 and the related coronaviruses SARS-CoV-1 and MERS can enter macrophages (82–84),
510 but the downstream effects remain unclear. One study shows that neither MERS-CoV nor SARS-
511 CoV-1 can replicate in human macrophages in vitro (82), another found that MERS-CoV but not
512 SARS-CoV-1 can replicate within human monocyte-derived macrophages (83), and yet another
513 study demonstrated abortive SARS-CoV-1 infection of human macrophages (84). For SARS-
514 CoV-2, a recent study reported that the virus fails to replicate or induce an inflammatory
515 response in human monocyte-derived macrophages in vitro (85), while another study showed
516 that the SARS-CoV-2 spike protein can induce macrophage activation in murine cells (86).
517 Importantly, these studies focused on blood-derived macrophages, and tissue-resident
518 macrophages may have different properties. SARS-CoV-2 has been detected in human alveolar
519 macrophages COVID-19 patients (87). While it is possible that infection in adipose tissue
520 macrophages is abortive, two lines of evidence suggest that this infection is productive. First, we
521 detected both genomic RNA and positive sense sgRNA, an intermediate RNA made during viral
522 replication (88); however, these data should be interpreted with caution as sgRNAs can be quite
523 stable and therefore may not fully reflect active replication (89, 90). Second, we observed
524 increased viral RNA accumulation over time in SVC infections, suggestive of viral replication,
525 which was decreased by remdesivir, an inhibitor of the viral RNA-dependent RNA polymerase.
526 Interestingly, we also observed enrichment in inflammatory pathways and a reduction in host
34
527 translation pathways specifically in infected macrophages, indicative of viral hijacking of host
528 translation (91). Such suppression of host translation has been also noted in SARS-CoV-1
529 infection: the NSP1 protein of SARS-CoV-1 suppresses host translation and gene expression,
530 including those of type I interferons (43). Together, these data highlight the dramatic effects of
531 SARS-CoV-2 infection on macrophages.
532
533 We also found that SARS-CoV-2 can infect human adipocytes both in vitro and in vivo.
534 Previously, researchers demonstrated that SARS-CoV-2 is detectable in adipose tissue of SARS-
535 CoV-2-infected hamsters, and that human adipocytes derived from breast tissue can support
536 infection in vitro (15). Here we demonstrate infection in multiple different adipose depots,
537 including critical peri-organ depots. We show this through (1) detection of both genomic and
538 subgenomic SARS-CoV-2 N gene expression in infected mature adipocytes and (2) infected
539 differentiated preadipocytes, (3) demonstration of a reduction in SARS-CoV-2 viral load in
540 differentiated preadipocytes post-infection in response to remdesivir treatment, and (4) histology
541 showing SARS-CoV-2 in adipocytes in a COVID-19 autopsy. This data adds to a growing body
542 of evidence that adipose tissue can serve as a reservoir for RNA viruses including influenza A
543 virus and HIV (16, 17, 19). Collectively, this data suggests that adipose tissue may serve as a
544 reservoir for SARS-CoV-2.
545
546 Our data show that preadipocytes adopted proinflammatory phenotypes following SARS-CoV-2
547 infection of SVC, despite not being infected themselves. We noted 17 different preadipocyte
548 clusters that varied in their representation among VAT and SAT depots. Interestingly, we found
549 that adipocyte stem cell-like clusters demonstrated more subdued transcriptional changes in
35
550 response to infection compared to other clusters across depots and participants. Pathways
551 upregulated across various preadipocyte clusters in response to infection included interferon and
552 interleukin signaling, both of which are important to controlling viral infection. Previous reports
553 have demonstrated that IL-6 and TNF-ɑ cytokine exposure can impair preadipocyte
554 differentiation and lipid accumulation, instead promoting the transition of preadipocytes towards
555 an inflammatory state (44, 92). Thus, the SARS-CoV-2 infection of adipose tissue could drive a
556 proinflammatory cascade promoted by preadipocyte activation. Interestingly, SAT preadipocytes
557 had a more dramatic inflammatory response than VAT preadipocytes. This was surprising
558 because several studies have pointed to the VAT being a better predictor of COVID-19 severity
559 than the SAT (93–95). For example, a recent single-center cohort study demonstrated that
560 increased VAT thickness and lower SAT thickness increased the risk of COVID-19 ICU
561 admission, independent of BMI (95). While bulk transcriptomic data collected from in vivo
562 SARS-CoV-2 infection of hamsters concluded that VAT had a stronger antiviral response than
563 SAT, this study included only mature adipocytes and studied the response 48 hpi (15). While our
564 human data similarly demonstrates higher cytokine expression of VAT preadipocytes at baseline,
565 it also suggests that the SAT preadipocytes might play an underappreciated role early upon
566 infection by driving an increased inflammatory response.
567
568 A significant finding of our study is the dramatic inflammatory response following SARS-CoV-2
569 infection of adipose tissue, particularly within the SVC. Across multiple experiments, at both the
570 RNA and protein level, we saw increased expression of cytokines, ISGs and other molecules
571 related to inflammation and antiviral pathways 24 hours after SVC infection. Many of the
572 molecules we found upregulated upon SARS-CoV-2 infection are associated with COVID-19
36
573 severity: IP-10, PDGFAA, PDGFAB/BB, IL-4, and MCSF all have been reported to be elevated
574 in the serum of critically ill COVID-19 patients (49, 51, 52), MIF has been described as a
575 predictor of poor outcome on mechanically ventilated COVID-19 patients (62), and VEGF may
576 play a crucial role in COVID-19 related brain inflammation (60). Further, we detected increased
577 type I interferon transcripts and ISG15 in SVC of both SAT and VAT 72 and 96 hpi, indicative
578 of a persistent antiviral response during viral replication. While significantly less dramatic, we
579 also saw trends for increased inflammatory cytokine and chemokines following infection of
580 mature adipocytes. Notably, we observed both induction of transcripts (CCL8 and CCL3) and
581 demonstrated secretion of chemokines (MCP2 and MCP3) in infected SVC cultures that can
582 attract macrophages. This is entirely consistent with our autopsy finding of an inflammatory
583 infiltrate associated with SARS-CoV-2+ cells and suggests that the antiviral response may be
584 dragging more susceptible macrophages to the site of infection. These data suggest that targeted
585 inhibition of inflammation in the adipose tissue could improve outcomes in COVID-19 subjects.
586 In fact, therapeutics targeting inflammation of the adipose tissue during obesity-induced
587 inflammation have been proposed as a treatment for metabolic disease (96). For instance,
588 salicylate, a cyclooxygenase inhibitor, reduces inflammation of the adipose tissue in obese
589 individuals and has been proposed as a therapeutic strategy in COVID-19 patients due to its anti-
590 inflammatory properties and antiviral activity against both DNA and RNA viruses (97–103).
591 Therefore, drugs reducing inflammation of the adipose tissue in obese individuals should be
592 further explored in COVID-19 subjects.
593
594 The mechanism(s) of viral entry into adipocytes and SVC macrophages remain unclear but may
595 not be through the canonical ACE2 receptor. While ACE2 mRNA expression has been reported
37
596 in adipose tissue, (113) and its level is affected by diet and obesity (114), we detected no ACE2
597 protein in SVC from VAT, and almost no expression in SAT. We could not detect ACE2 gene
598 expression in SVC from either SAT or VAT, either by RTqPCR with primers validated on
599 kidney tissue or by scRNA-seq. Similarly, we did not detect ACE2 in mature adipocytes, though
600 we did identify upregulation of ACE2 transcripts during in vitro differentiation of preadipocytes.
601 Another recent study also detected higher levels of NRP1 and FURIN (two other proposed
602 SARS-CoV-2 viral entry factors) compared to ACE2 and TMPRSS2 in murine and human
603 adipocytes (15). In both macrophages and adipocytes, entry could occur via use of an alternate
604 entry receptor. Genome-scale CRISPR loss-of-function screens in susceptible human cells have
605 identified multiple SARS-CoV-2 entry factors in addition to ACE2 (115). Alternative entry
606 receptors reported include CD147, Neuropilin-1, Dipeptidyl peptidase 4 (DPP4), alanyl
607 aminopeptidase (ANPEP), glutamyl aminopeptidase (ENPEP), and angiotensin II receptor type 2
608 (AGTR2) (104, 105). In macrophages, entry could also occur via phagocytosis of viral particles,
609 phagocytosis of infected cells, or, if antibodies were present, via antibody-dependent entry.
610 Elucidating the entry mechanisms will be an important area of future investigation given its
611 implications for disease pathogenesis. For example, Neuropilin-1 has may mediate entry into
612 neural cells and play a role in neurologic manifestations of COVID-19 (106), and yet antibody
613 therapies that only block interactions with ACE2 may not block such entry if it is relevant in
614 vivo.
615
616 Our study has several limitations. Our numbers of replicates were limited for some assays, such
617 as the evaluation of cytokine secretion following infection of mature adipocytes. Nonetheless, we
618 observed significant indications of inflammatory responses. It is possible that we were unable to
38
619 fully wash input virus following infection of mature adipocytes due to the high lipid content,
620 size, and fragility of the cells, falsely increasing the viral signal. However, our detection of
621 sgRNA, the time-dependent increase in viral RNA accumulation that was inhibited by
622 remdesivir, and the detection of SARS-CoV-2 RNA in autopsy samples all provide orthogonal
623 support for true infection of adipocytes. Our autopsy studies were limited in number, and we
624 were only able to perform confirmatory ISH on epicardium, not in the subcutaneous, omental, or
625 pericardial fat due to limited autopsy tissue availability. All experiments were performed with
626 the WA-01 strain of SARS-CoV-2 and no experiments with its variants were performed, and
627 plaque assays to confirm viral production were not performed. As tissue donors were obese, an
628 important area of future investigation will be to study the effects of SARS-CoV-2 infection on
629 lean adipose tissue, as well as to study the adipose tissue of those with ‘long COVID’.
630
631 Overall, here we provide evidence that two cell types within human adipose tissue are permissive
632 to SARS-CoV-2 infection. This adds to data showing susceptibility of other tissues including
633 heart, kidneys, pharynx, liver, brain, and pancreas (1, 27, 107). SARS-CoV-2 RNA was detected
634 in autopsy samples at a higher viral load than in adjacent organs, the kidney or heart.
635 Importantly, infection of adipose tissue drives an inflammatory response in infected
636 macrophages and preadipocytes. Thus, multiple cells within adipose tissue likely participate in
637 both viral replication and inflammation. Importantly, we demonstrated infection and
638 inflammation in adipose tissue adjacent to critical organs such as the heart and intestine, thus
639 pointing to the potential for adipose tissue potentiation of organ damage in severe COVID-19.
640 Furthermore, if adipose cells constitute a reservoir for viral infection, obesity may contribute not
641 only to severe acute disease, but also to long-COVID syndrome. Collectively, our data implies
39
642 that infection in adipose tissue may partially explain the link between obesity and severe
643 COVID-19. More efforts to understand the complexity and contributions of this tissue to
644 COVID-19 pathogenesis are warranted.
645
646 MATERIALS AND METHODS
647 Study design: The aim of this study was to determine if human adipose tissue is permissive to
648 SARS-CoV-2 infection. We obtained adipose tissue from consented subjects and exposed these
649 samples to SARS-CoV-2 and measured viral entry, replication, and inflammatory pathways by
650 flow cytometry, RTqPCR, scRNA-seq, and Luminex. All analyses were performed in an
651 unbiased fashion. Formalin-fixed and paraffin embedded lung, kidney, adipose and heart tissue
652 from COVID-19 autopsy samples were evaluated for viral RNA. The n value was not controlled
653 and was dependent on availability of samples.
654
655 Subjects: Study participants were recruited from the Adult Bariatric Surgery and Cardiothoracic
656 Surgery clinics at Stanford University Medical Center during the preoperative visit. Eligibility
657 requirements include >25 yrs of age, and exclusion criteria included chronic inflammatory
658 conditions, pregnancy/lactation, use of weight loss medications, and current/prior diagnosis of
659 COVID-19. The protocol was approved by the Stanford Institutional Review Board and all
660 subjects gave written, informed consent. Tissue samples from COVID-19 deceased patients were
661 obtained from University Hospital Tübingen, Tübingen, Germany, or from University Hospital
662 Basel or Cantonal Hospital Baselland, Switzerland. SARS-CoV-2 infected tissue was obtained
663 during autopsy and processed as previously described (71). The use of SARS-CoV-2 infected
664 tissue was approved by the ethics commission of Northern Switzerland (EKNZ; study ID: 2020-
40
665 00969). All COVID-19 patients or their relatives consented to the use of tissue for research
666 purposes.
667
668 Preparation, isolation, and differentiation of adipose tissue: On the day of bariatric surgery,
669 approximately 2-3g each of subcutaneous abdominal (SAT) and omental visceral adipose (VAT)
670 tissue was harvested intraoperatively and immediately processed. For cardiothoracic surgery
671 patients, 2g of PAT, 1g of EAT, and 1-2g of SAT chest wall was obtained intraoperatively and
672 immediately processed. Tissue was subject to collagenase digestion for separation of mature
673 adipocytes (MA) and SVC, and for differentiation of preadipocytes as previously described (108)
674 with details in supplemental methods.
675
676 Virus and cell lines: The USA WA1/2020 strain of SARS-CoV-2 was obtained from BEI
677 Resources, passaged in VeroE6 cells, and tittered by Avicel (FMC Biopolymer) overlay plaque
678 assay on VeroE6 cells. Passage 3 virus was used for all infections. VeroE6 cells were obtained
679 from ATCC and were mycoplasma free. A549-ACE2 was a gift from Ralf Bartenschlager and
680 was mycoplasma free.
681
682 SARS-CoV-2 infections of differentiated adipocytes, SVC, MA, and A549-ACE2: Cells were
683 infected with SARS-CoV-2 (2019-nCOV/USA-WA1/2020) at a MOI of 1 for 1 hour before
684 washing input virus and replacing media for incubation. Remdesivir was used at 10 µM.
685
686 RNA isolation and quantification: At the time of collection, cells were washed with PBS
687 followed by incubation with TRIzol LS (Thermo) reagent for 15-20mins for cell lysis and virus
41
688 inactivation and RNA extraction. RNA was resuspended in water and quantified by absorbance
689 (260/280) using a NanoDropTM spectrophotometer (Thermo Scientific).
690
691 Genomic and subgenomic absolute gene quantification by RTqPCR: 5µl of the total RNA
692 was used for 1-step RTqPCR. Genomic N-gene quantification was done with the use of CDC
693 qualified primers and probes amplifying the N1 region, n2019-nCoV (Biosearch technologies,
694 KIT-NCOV-PP1-1000). For subgenomic N-gene quantification, E-gene sgRNA forward primer
695 for SARS-CoV-2 leader sequence was combined with CDC N1 gene reverse primer and probe to
696 detect N-gene sgRNA as previously shown (89, 109). RNA and primers were mixed with the
697 TaqPath 1-step RTqPCR master mix (Applied Biosystems, A15299). A standard curve for Ct
698 values and genome copy numbers was obtained using pET21b+ plasmid with the N-gene inserts.
699 All the samples were analyzed in technical duplicates. The Ct cutoff for positive samples was
700 <38 with amplification observed in both duplicates. The samples were analyzed on a
701 StepOnePlusTM real time PCR machine (Applied Biosystems) using the following parameters:
702 (stage 1) 10 minutes at 50ºC for reverse transcription (RT), followed by (stage 2) 3 minutes at
703 95ºC for initial denaturation and (stage 3) 40 cycles of 10 seconds at 95ºC, 15 seconds at 56ºC,
704 and 5 seconds at 72ºC.
705
706 RTqPCR for relative gene quantification: RNA was isolated as described above. Either the 1-
707 step method using TaqPath (Applied Biosystems, A15299) or the 2-step process using the
708 Superscript III first-strand synthesis system was performed (Applied biosystems, A25742) (for
709 details see the Supplemental Materials and Material).
710
42
711 SARS-CoV-2 detection in autopsies: Autopsy samples were prepared as previously described
712 (110). Briefly, RNA was isolated from formalin-fixed and paraffin embedded tissue with the use
713 of Maxwell RSC RNA FFPE (Promega) according to manufacturer recommendations. RTqPCR
714 was performed using the TaqManTM 2019-nCoV control kit v1 (Thermo Scientific, A47533) to
715 target the three viral genes: ORF1ab, S, and N genes, and the human RPPH1 gene. According to
716 the manufacturer's recommendation, a Ct value below 37 in at least two out of three viral
717 genomic regions was considered positive. A case was considered negative if Ct values were
718 above 40. Values between 37 and 40 were considered undetermined and the assay was repeated.
719 Samples were always run in duplicates. RNA-ISH was used to detect SARS-CoV-2 Spike
720 mRNA in tissue samples (for details see the Supplemental Materials and Material).
721
722 Flow cytometry: A single cell suspension was stained with Fc block, a viability stain, and
723 surface stained for CD45, CD3, CD14, CD34, CD11c, and CD31 before fixing and
724 permeabilizing to stain with anti-SARS-CoV-2 N protein as described in supplemental methods.
725
726 Sample preparation for Luminex: Supernatants from mock or infected SVC were collected
727 after 24 hpi into low-binding protein collection tubes. Supernatants from mature adipocytes were
728 collected by first removing floating mature adipocytes with a cut wide pipette tip and pipetting
729 remaining media into low-protein binding collection tubes. Supernatants were stored at -80oC.
730 To remove samples from BSL3 containment, supernatants were thawed and mixed with 10%
731 TritonX-100 (Sigma-Aldrich, T9284) for a final concentration of 1% TritonX-100 and incubated
732 for 20 minutes at room temperature for viral inactivation. Supernatants were then removed from
43
733 BSL3 and transferred to a low-binding protein 96 well plate for 80plex Luminex by the Human
734 Immune Monitoring Center (HIMC) at Stanford University.
735
736 Single cell RNA sequencing (scRNA-seq): The gel beads-in-emulsion (GEM) single cell 3’ kit
737 v3.1 dual index kit (10X Genomics) was used following manufacturer’s recommendations with
738 slight modifications. Briefly, SVC from SAT and VAT was left untreated or infected with
739 SARS-CoV-2 as described above. After 24hpi, cells were collected, washed, and diluted at a
740 density of 1x103 cells per µl in cold DMEM (Life technologies; 11885-092) media supplemented
741 with 10% FBS (Corning, MT35016CV). 10,000 cells per lane were loaded onto a Chromium
742 Controller in the BSC per manufacturer’s instructions. Following GEM creation, samples were
743 transferred into PCR tube strips prior to transferring into a PTC-200 thermocycler (MJ Research)
744 for RT. The RT parameters were the following: 45 minutes at 53oC, followed by 5 minutes at
745 85oC, then 15 minutes at 60oC and finally samples were kept at 4oC. Barcoded cDNA was
746 removed to BSL2, and sequencing libraries were prepared per manufacturer's recommendation,
747 with a TapeStation 4200 (Agilent) used for quality control. Libraries were pooled for a final
748 concentration of 10nM and sequenced on a NovaSeq S2 (Ilumina) at the Chan Zuckerberg
749 Biohub (San Francisco).
750
751 Alignment and preprocessing of scRNA-seq data: The quality of the raw FASTQ data was
752 examined with FASTQC and then aligned (cell ranger count) to a custom genome including
753 human genome (hg38) and the complete genome sequence of SARS-CoV-2 (2019-nCOV/USA-
754 WA1/2020) (GenBank: MN985325.1) using the “Cell Ranger” software package v6.0.0 (10x
755 Genomics). R (4.0.4) was used for all downstream analyses. Resulting filtered feature-barcode-
44
756 matrices were processed using the R package Seurat (v4.0.0). Briefly, count matrices were
757 merged and loaded into Seurat with SARS-CoV-2 counts removed and appended to the
758 metadata. All genes represented in < 10 cells were excluded from downstream analysis. Cells
759 were filtered based on the following criteria: less than 200 distinct genes, less than 100 unique
760 molecular identifiers (UMIs), and greater than 15% UMIs from mitochondrial genes. Each batch
761 was then individually normalized using the “SCTransform” function that included regression for
762 percent mitochondria. Integration features were then calculated using the
763 “SelectIntegrationFeatures” function and passed into “VariableFeatures” of the merged object to
764 maintain the repeatedly variable features across each dataset. Within each batch and condition
765 (tissue and infection status), cells identified as doublets by both “DoubletFinder” (v3) (using pN
766 = 0.25, pK = 0.09, PCs = 1:50, anticipated collision rate = 10%) and scds (top 10% of cells
767 ranked by hybrid scores) were removed (n = 3,165 cells) from the analysis. After applying these
768 filtering steps, the dataset contained 198,759 high-quality cells. Principal component (PC)
769 analysis (PCA) was performed on the data. The resulting data were corrected for batch effects
770 using the “Harmony” package (29) with the top 50 PCs. Uniform Manifold Approximation and
771 Projection for Dimension Reduction (UMAP) coordinate generation and clustering were
772 performed using the “RunUMAP”, “FindNeighbors”, and “FindClusters” (resolution = 0.6)
773 functions in Seurat with PCs 1-50. Manual annotation of each cellular cluster was performed by
774 finding the differentially expressed genes using Seurat's “FindAllMarkers” function with default
775 Wilcoxon rank-sum test and comparing those markers with known cell type-specific genes from
776 previous datasets (30, 31). “FindMarkers” function using the MAST algorithm (latent.vars =
777 ‘participant’) based on a Bonferroni-adjusted P < 0.05 and a log2 fold change > 0.25 was used
778 for targeted differential gene expression analysis.
45
779
780 scRNA-seq analyses: Gene ontology and KEGG pathway analysis was performed using R
781 package stringdb (111). Reactome gene set enrichment analysis was performed using the R
782 package fgsea (112), considering ranked gene lists. The Seurat function AddModuleScore() was
783 used to score individual cells by expression of either a list of genes relating to ISGs or cytokines.
784 This function generates an average module score by calculating the mean expression of each
785 gene in the module corrected for expression of a random set of similarly expressing genes. Gene
786 lists used to define each module are defined in table S15. Heatmaps were generated using
787 ComplexHeatmap, Seurat and pheatmap packages, Violin plots were generated using ggplot2,
788 and dotplots were generated using the Seurat packages in R. A Github repository for all original
789 code used for analysis and visualization will be made public upon publication.
790
791 Images: Pictures were taken with the use of a EVOS XL core cell imaging system (Thermo
792 Fisher Scientific) with an objective of 10x.
793
794 Statistical analysis: GraphPad Prism version 9.1.0 (216) and R (4.0.4) were used for statistical
795 analysis. When comparing mock and infected groups a paired, two-sided, student’s t-test was
796 applied. When comparing more than two groups a two-way ANOVA, multiple comparisons
797 using statistical hypothesis Sidak was performed. In Luminex analysis a paired Wilcoxon signed
798 rank test with desired false discovery rate of 10% was performed. Data bars are always presented
799 as ± mean s.e.m.
800
801 Supplementary Materials
46
802 Materials and Methods
803 Fig. S1. Genomic and Subgenomic measurement of SARS-CoV-2 on in vitro infected SVC.
804 Fig. S2. Validation of Remdesivir treatments in A549-ACE2 cells.
805 Fig. S3. Gating strategy of SVC in human adipose tissue.
806 Fig. S4. Limited ACE2 protein detection in SVC from SAT and VAT.
807 Fig. S5. Infection, depot and participant breakdown by cluster annotation.
808 Fig. S6. Distribution of SARS-CoV-2 transcripts across all cells.
809 Fig. S7. C2- and C12-macrophages are distinctly different upon mock and SARS-CoV-2
810 infection.
811 Fig. S8. Characterization of preadipocytes across VAT and SAT.
812 Fig. S9. Mature and in vitro differentiated adipocytes harbor SARS-CoV-2 RNA and mount
813 inflammatory responses after exposure to SARS-CoV-2 in vitro.
814 Fig. S10. Increased interferon related genes in adipose tissue post-SARS-CoV-2 infection.
815 Fig. S11. Elevated IL-6 gene expression and secretion of inflammatory mediators in SVC post in
816 vitro SARS-CoV-2 infection.
817 Table S1. Adipose tissue participant’s demographic, medical information, and sample use.
818 Table S2. No expression of ACE2 in adipose tissue but increased ACE2 expression in in vitro
819 differentiated adipocytes at 3 days of differentiation.
820 Table S3. Marker genes for each cluster defined in combined Seurat analysis.
47
821 Table S4. Differentially expressed transcripts within macrophages only, across different subsets
822 of infection, mock and depot.
823 Table S5. Differentially expressed transcripts between infected (SARS-CoV-2+) versus
824 bystander populations within the C2 and C12 infected macrophage populations.
825 Table S6. GO and KEGG term enrichment for markers of C12 macrophages.
826 Table S7. GO and KEGG term enrichment for markers of C2 macrophages.
827 Table S8. Reactome pathways for differentially expressed genes between infected (SARS-CoV-
828 2+) versus bystander C2 macrophages.
829 Table S9. Marker genes for each cluster defined from the Seurat analysis of Preadipocytes only.
830 Table S10. Differentially expressed genes between SARS-CoV-2 exposed versus mock exposed
831 preadipocytes, by cluster and participant, within the SAT
832 Table S11. Differentially expressed genes between SARS-CoV-2 exposed versus mock exposed
833 preadipocytes, by cluster and participant, within the VAT.
834 Table S12. Reactome pathways for differentially expressed genes between SARS-CoV-2
835 exposed versus mock exposed preadipocytes, by cluster and participant, within the SAT.
836 Table S13. Reactome pathways for differentially expressed genes between SARS-CoV-2
837 exposed versus mock exposed preadipocytes, by cluster and participant, within the VAT.
838 Table S14. ISG and cytokine gene modules.
839
840 References and Notes
841
48
842 1. A. Gupta, M. V. Madhavan, K. Sehgal, N. Nair, S. Mahajan, T. S. Sehrawat, B. Bikdeli, N.
843 Ahluwalia, J. C. Ausiello, E. Y. Wan, D. E. Freedberg, A. J. Kirtane, S. A. Parikh, M. S. Maurer,
844 A. S. Nordvig, D. Accili, J. M. Bathon, S. Mohan, K. A. Bauer, M. B. Leon, H. M. Krumholz, N.
845 Uriel, M. R. Mehra, M. S. V. Elkind, G. W. Stone, A. Schwartz, D. D. Ho, J. P. Bilezikian, D.
846 W. Landry, Extrapulmonary manifestations of COVID-19, Nat Med 26, 1017–1032 (2020).
847 2. C. M. Petrilli, S. A. Jones, J. Yang, H. Rajagopalan, L. O’Donnell, Y. Chernyak, K. A. Tobin,
848 R. J. Cerfolio, F. Francois, L. I. Horwitz, Factors associated with hospital admission and critical
849 illness among 5279 people with coronavirus disease 2019 in New York City: prospective cohort
850 study, Bmj 369, m1966 (2020).
851 3. A. Simonnet, M. Chetboun, J. Poissy, V. Raverdy, J. Noulette, A. Duhamel, J. Labreuche, D.
852 Mathieu, F. Pattou, M. Jourdain, R. Caizzo, M. Caplan, N. Cousin, T. Duburcq, A. Durand, A. E.
853 kalioubie, R. Favory, B. Garcia, P. Girardie, J. Goutay, M. Houard, E. Jaillette, N. Kostuj, G.
854 Ledoux, D. Mathieu, A. S. Moreau, C. Niles, S. Nseir, T. Onimus, E. Parmentier, S. Préau, L.
855 Robriquet, A. Rouze, S. Six, H. Verkindt, High Prevalence of Obesity in Severe Acute
856 Respiratory Syndrome Coronavirus‐2 (SARS‐CoV‐2) Requiring Invasive Mechanical
857 Ventilation, Obesity 28, 1195–1199 (2020).
858 4. Q. Ruan, K. Yang, W. Wang, L. Jiang, J. Song, Clinical predictors of mortality due to
859 COVID-19 based on an analysis of data of 150 patients from Wuhan, China, Intens Care Med
860 46, 846–848 (2020).
861 5. S. Y. Tartof, L. Qian, V. Hong, R. Wei, R. F. Nadjafi, H. Fischer, Z. Li, S. F. Shaw, S. L.
862 Caparosa, C. L. Nau, T. Saxena, G. K. Rieg, B. K. Ackerson, A. L. Sharp, J. Skarbinski, T. K.
863 Naik, S. B. Murali, Obesity and Mortality Among Patients Diagnosed With COVID-19: Results
864 From an Integrated Health Care Organization, Ann Intern Med 173, 773–781 (2020).
49
865 6. B. M. Popkin, S. Du, W. D. Green, M. A. Beck, T. Algaith, C. H. Herbst, R. F. Alsukait, M.
866 Alluhidan, N. Alazemi, M. Shekar, Individuals with obesity and COVID‐19: A global
867 perspective on the epidemiology and biological relationships, Obes Rev 21, e13128 (2020).
868 7. D. Moriconi, S. Masi, E. Rebelos, A. Virdis, M. L. Manca, S. D. Marco, S. Taddei, M.
869 Nannipieri, Obesity prolongs the hospital stay in patients affected by COVID-19, and may
870 impact on SARS-COV-2 shedding, Obes Res Clin Pract 14, 205–209 (2020).
871 8. M. Deng, Y. Qi, L. Deng, H. Wang, Y. Xu, Z. Li, Z. Meng, J. Tang, Z. Dai, Obesity as a
872 Potential Predictor of Disease Severity in Young COVID‐19 Patients: A Retrospective Study,
873 Obesity 28, 1815–1825 (2020).
874 9. C. Zammit, H. Liddicoat, I. Moonsie, H. Makker, Obesity and respiratory diseases, Int J Gen
875 Medicine 3, 335–343 (2010).
876 10. G. Reaven, F. Abbasi, T. McLaughlin, Obesity, Insulin Resistance, and Cardiovascular
877 Disease, Recent Prog Horm Res 59, 207–223 (2004).
878 11. T. McLaughlin, G. Allison, F. Abbasi, C. Lamendola, G. Reaven, Prevalence of insulin
879 resistance and associated cardiovascular disease risk factors among normal weight, overweight,
880 and obese individuals, Metabolis 53, 495–499 (2004).
881 12. F. Abbasi, T. McLaughlin, C. Lamendola, I. Lipinska, G. Tofler, G. M. Reaven, Comparison
882 of Plasminogen Activator Inhibitor-1 Concentration in Insulin-Resistant Versus Insulin-Sensitive
883 Healthy Women, Arterioscler Thromb Vasc Biol 19, 2818–21 (1999).
884 13. T. McLaughlin, S. E. Ackerman, L. Shen, E. Engleman, Role of innate and adaptive
885 immunity in obesity-associated metabolic disease, J Clin Invest 127, 5–13 (2017).
886 14. H. E. Maier, R. Lopez, N. Sanchez, S. Ng, L. Gresh, S. Ojeda, R. Burger-Calderon, G. Kuan,
887 E. Harris, A. Balmaseda, A. Gordon, Obesity Increases the Duration of Influenza A Virus
50
888 Shedding in Adults, J Infect Dis 218, 1378–1382 (2018).
889 15. M. Reiterer, M. Rajan, N. Gómez-Banoy, J. D. Lau, L. G. Gomez-Escobar, L. Ma, A. Gilani,
890 S. Alvarez-Mulett, E. T. Sholle, V. Chandar, Y. Bram, K. Hoffman, P. Bhardwaj, P. Piloco, A.
891 Rubio-Navarro, S. Uhl, L. Carrau, S. Houhgton, D. Redmond, A. P. Shukla, P. Goyal, K. A.
892 Brown, B. R. tenOever, L. C. Alonso, R. E. Schwartz, E. J. Schenck, M. M. Safford, J. C. Lo,
893 Hyperglycemia in Acute COVID-19 is Characterized by Insulin Resistance and Adipose Tissue
894 Infectivity by SARS-CoV-2, Cell Metab (2021), doi:10.1016/j.cmet.2021.09.009.
895 16. J. Couturier, J. W. Suliburk, J. M. Brown, D. J. Luke, N. Agarwal, X. Yu, C. Nguyen, D.
896 Iyer, C. A. Kozinetz, P. A. Overbeek, M. L. Metzker, A. Balasubramanyam, D. E. Lewis, Human
897 adipose tissue as a reservoir for memory CD4+ T cells and HIV, Aids 29, 667–674 (2015).
898 17. H. Nishimura, S. Itamura, T. Iwasaki, T. Kurata, M. Tashiro, Characterization of human
899 influenza A (H5N1) virus infection in mice: neuro-, pneumo- and adipotropic infection, J Gen
900 Virol 81, 2503–2510 (2000).
901 18. K. J. Zwezdaryk, M. B. Ferris, A. L. Strong, C. A. Morris, B. A. Bunnell, N. V. Dhurandhar,
902 J. M. Gimble, D. E. Sullivan, Human cytomegalovirus infection of human adipose-derived
903 stromal/stem cells restricts differentiation along the adipogenic lineage, Adipocyte 5, 53–64
904 (2015).
905 19. A. Damouche, T. Lazure, V. Avettand-Fènoël, N. Huot, N. Dejucq-Rainsford, A.-P. Satie, A.
906 Mélard, L. David, C. Gommet, J. Ghosn, N. Noel, G. Pourcher, V. Martinez, S. Benoist, V.
907 Béréziat, A. Cosma, B. Favier, B. Vaslin, C. Rouzioux, J. Capeau, M. Müller-Trutwin, N.
908 Dereuddre-Bosquet, R. L. Grand, O. Lambotte, C. Bourgeois, Adipose Tissue Is a Neglected
909 Viral Reservoir and an Inflammatory Site during Chronic HIV and SIV Infection, Plos Pathog
910 11, e1005153 (2015).
51
911 20. H. Kanda, S. Tateya, Y. Tamori, K. Kotani, K. Hiasa, R. Kitazawa, S. Kitazawa, H. Miyachi,
912 S. Maeda, K. Egashira, M. Kasuga, MCP-1 contributes to macrophage infiltration into adipose
913 tissue, insulin resistance, and hepatic steatosis in obesity, J Clin Invest 116, 1494–1505 (2006).
914 21. B. E. Wisse, The Inflammatory Syndrome&colon; The Role of Adipose Tissue Cytokines in
915 Metabolic Disorders Linked to Obesity, J Am Soc Nephrol 15, 2792–2800 (2004).
916 22. G. S. Hotamisligil, N. S. Shargill, B. M. Spiegelman, Adipose Expression of Tumor Necrosis
917 Factor-α: Direct Role in Obesity-Linked Insulin Resistance, Science 259, 87–91 (1993).
918 23. H. Shi, M. V. Kokoeva, K. Inouye, I. Tzameli, H. Yin, J. S. Flier, TLR4 links innate
919 immunity and fatty acid–induced insulin resistance, J Clin Invest 116, 3015–3025 (2006).
920 24. T. M. Delorey, C. G. K. Ziegler, G. Heimberg, R. Normand, Y. Yang, Å. Segerstolpe, D.
921 Abbondanza, S. J. Fleming, A. Subramanian, D. T. Montoro, K. A. Jagadeesh, K. K. Dey, P.
922 Sen, M. Slyper, Y. H. Pita-Juárez, D. Phillips, J. Biermann, Z. Bloom-Ackermann, N. Barkas, A.
923 Ganna, J. Gomez, J. C. Melms, I. Katsyv, E. Normandin, P. Naderi, Y. V. Popov, S. S. Raju, S.
924 Niezen, L. T.-Y. Tsai, K. J. Siddle, M. Sud, V. M. Tran, S. K. Vellarikkal, Y. Wang, L. Amir-
925 Zilberstein, D. S. Atri, J. Beechem, O. R. Brook, J. Chen, P. Divakar, P. Dorceus, J. M. Engreitz,
926 A. Essene, D. M. Fitzgerald, R. Fropf, S. Gazal, J. Gould, J. Grzyb, T. Harvey, J. Hecht, T.
927 Hether, J. Jané-Valbuena, M. Leney-Greene, H. Ma, C. McCabe, D. E. McLoughlin, E. M.
928 Miller, C. Muus, M. Niemi, R. Padera, L. Pan, D. Pant, C. Pe’er, J. Pfiffner-Borges, C. J. Pinto,
929 J. Plaisted, J. Reeves, M. Ross, M. Rudy, E. H. Rueckert, M. Siciliano, A. Sturm, E. Todres, A.
930 Waghray, S. Warren, S. Zhang, D. R. Zollinger, L. Cosimi, R. M. Gupta, N. Hacohen, H.
931 Hibshoosh, W. Hide, A. L. Price, J. Rajagopal, P. R. Tata, S. Riedel, G. Szabo, T. L. Tickle, P.
932 T. Ellinor, D. Hung, P. C. Sabeti, R. Novak, R. Rogers, D. E. Ingber, Z. G. Jiang, D. Juric, M.
933 Babadi, S. L. Farhi, B. Izar, J. R. Stone, I. S. Vlachos, I. H. Solomon, O. Ashenberg, C. B. M.
52
934 Porter, B. Li, A. K. Shalek, A.-C. Villani, O. Rozenblatt-Rosen, A. Regev, COVID-19 tissue
935 atlases reveal SARS-CoV-2 pathology and cellular targets, Nature 595, 107–113 (2021).
936 25. T. T. S. Consortium, S. R. Quake, The Tabula Sapiens: a single cell transcriptomic atlas of
937 multiple organs from individual human donors, Biorxiv , 2021.07.19.452956 (2021).
938 26. E. Song, C. Zhang, B. Israelow, A. Lu-Culligan, A. V. Prado, S. Skriabine, P. Lu, O.-E.
939 Weizman, F. Liu, Y. Dai, K. Szigeti-Buck, Y. Yasumoto, G. Wang, C. Castaldi, J. Heltke, E. Ng,
940 J. Wheeler, M. M. Alfajaro, E. Levavasseur, B. Fontes, N. G. Ravindra, D. V. Dijk, S. Mane, M.
941 Gunel, A. Ring, S. A. J. Kazmi, K. Zhang, C. B. Wilen, T. L. Horvath, I. Plu, S. Haik, J.-L.
942 Thomas, A. Louvi, S. F. Farhadian, A. Huttner, D. Seilhean, N. Renier, K. Bilguvar, A. Iwasaki,
943 Neuroinvasion of SARS-CoV-2 in human and mouse brain, J Exp Med 218, e20202135 (2021).
944 27. J. A. Müller, R. Groß, C. Conzelmann, J. Krüger, U. Merle, J. Steinhart, T. Weil, L. Koepke,
945 C. P. Bozzo, C. Read, G. Fois, T. Eiseler, J. Gehrmann, J. van Vuuren, I. M. Wessbecher, M.
946 Frick, I. G. Costa, M. Breunig, B. Grüner, L. Peters, M. Schuster, S. Liebau, T. Seufferlein, S.
947 Stenger, A. Stenzinger, P. E. MacDonald, F. Kirchhoff, K. M. J. Sparrer, P. Walther, H. Lickert,
948 T. F. E. Barth, M. Wagner, J. Münch, S. Heller, A. Kleger, SARS-CoV-2 infects and replicates
949 in cells of the human endocrine and exocrine pancreas, Nat Metabolism 3, 149–165 (2021).
950 28. M. Hoffmann, H. Kleine-Weber, S. Schroeder, N. Krüger, T. Herrler, S. Erichsen, T. S.
951 Schiergens, G. Herrler, N.-H. Wu, A. Nitsche, M. A. Müller, C. Drosten, S. Pöhlmann, SARS-
952 CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven
953 Protease Inhibitor, Cell 181, 271-280.e8 (2020).
954 29. I. Korsunsky, N. Millard, J. Fan, K. Slowikowski, F. Zhang, K. Wei, Y. Baglaenko, M.
955 Brenner, P. Loh, S. Raychaudhuri, Fast, sensitive and accurate integration of single-cell data
956 with Harmony, Nat Methods 16, 1289–1296 (2019).
53
957 30. J. Vijay, M.-F. Gauthier, R. L. Biswell, D. A. Louiselle, J. J. Johnston, W. A. Cheung, B.
958 Belden, A. Pramatarova, L. Biertho, M. Gibson, M.-M. Simon, H. Djambazian, A. Staffa, G.
959 Bourque, A. Laitinen, J. Nystedt, M.-C. Vohl, J. D. Fraser, T. Pastinen, A. Tchernof, E.
960 Grundberg, Single-cell analysis of human adipose tissue identifies depot- and disease-specific
961 cell types, Nat Metabolism 2, 97–109 (2020).
962 31. A. D. Hildreth, F. Ma, Y. Y. Wong, R. Sun, M. Pellegrini, T. E. O’Sullivan, Single-cell
963 sequencing of human white adipose tissue identifies new cell states in health and obesity, Nat
964 Immunol , 1–15 (2021).
965 32. Q. Wang, Y. Zhang, L. Wu, S. Niu, C. Song, Z. Zhang, G. Lu, C. Qiao, Y. Hu, K.-Y. Yuen,
966 Q. Wang, H. Zhou, J. Yan, J. Qi, Structural and Functional Basis of SARS-CoV-2 Entry by
967 Using Human ACE2, Cell 181, 894-904.e9 (2020).
968 33. M.-M. Zhao, W.-L. Yang, F.-Y. Yang, L. Zhang, W.-J. Huang, W. Hou, C.-F. Fan, R.-H. Jin,
969 Y.-M. Feng, Y.-C. Wang, J.-K. Yang, Cathepsin L plays a key role in SARS-CoV-2 infection in
970 humans and humanized mice and is a promising target for new drug development, Signal
971 Transduct Target Ther 6, 134 (2021).
972 34. L. Cantuti-Castelvetri, R. Ojha, L. D. Pedro, M. Djannatian, J. Franz, S. Kuivanen, F. van der
973 Meer, K. Kallio, T. Kaya, M. Anastasina, T. Smura, L. Levanov, L. Szirovicza, A. Tobi, H.
974 Kallio-Kokko, P. Österlund, M. Joensuu, F. A. Meunier, S. J. Butcher, M. S. Winkler, B.
975 Mollenhauer, A. Helenius, O. Gokce, T. Teesalu, J. Hepojoki, O. Vapalahti, C. Stadelmann, G.
976 B. and M. Simons, Neuropilin-1 facilitates SARS-CoV-2 cell entry and infectivity, (2020).
977 35. J. Huber, F. W. Kiefer, M. Zeyda, B. Ludvik, G. R. Silberhumer, G. Prager, G. J. Zlabinger,
978 T. M. Stulnig, CC Chemokine and CC Chemokine Receptor Profiles in Visceral and
979 Subcutaneous Adipose Tissue Are Altered in Human Obesity, J Clin Endocrinol Metabolism 93,
54
980 3215–3221 (2008).
981 36. K. Asano, N. Takahashi, M. Ushiki, M. Monya, F. Aihara, E. Kuboki, S. Moriyama, M. Iida,
982 H. Kitamura, C.-H. Qiu, T. Watanabe, M. Tanaka, Intestinal CD169+ macrophages initiate
983 mucosal inflammation by secreting CCL8 that recruits inflammatory monocytes, Nat Commun 6,
984 7802 (2015).
985 37. S. Chakarov, H. Y. Lim, L. Tan, S. Y. Lim, P. See, J. Lum, X.-M. Zhang, S. Foo, S.
986 Nakamizo, K. Duan, W. T. Kong, R. Gentek, A. Balachander, D. Carbajo, C. Bleriot, B.
987 Malleret, J. K. C. Tam, S. Baig, M. Shabeer, S.-A. E. S. Toh, A. Schlitzer, A. Larbi, T. Marichal,
988 B. Malissen, J. Chen, M. Poidinger, K. Kabashima, M. Bajenoff, L. G. Ng, V. Angeli, F.
989 Ginhoux, Two distinct interstitial macrophage populations coexist across tissues in specific
990 subtissular niches, Science 363, eaau0964 (2019).
991 38. W.-K. Kim, X. Alvarez, J. Fisher, B. Bronfin, S. Westmoreland, J. McLaurin, K. Williams,
992 CD163 Identifies Perivascular Macrophages in Normal and Viral Encephalitic Brains and
993 Potential Precursors to Perivascular Macrophages in Blood, Am J Pathology 168, 822–834
994 (2006).
995 39. M. Kratz, B. R. Coats, K. B. Hisert, D. Hagman, V. Mutskov, E. Peris, K. Q. Schoenfelt, J.
996 N. Kuzma, I. Larson, P. S. Billing, R. W. Landerholm, M. Crouthamel, D. Gozal, S. Hwang, P.
997 K. Singh, L. Becker, Metabolic Dysfunction Drives a Mechanistically Distinct Proinflammatory
998 Phenotype in Adipose Tissue Macrophages, Cell Metab 20, 614–625 (2014).
999 40. L. Lu, H. Zhang, D. J. Dauphars, and Y.-W. He, A Potential Role of Interleukin 10 in
1000 COVID-19 Pathogenesis, (2021).
1001 41. E. Mick, J. Kamm, A. O. Pisco, K. Ratnasiri, J. M. Babik, G. Castañeda, J. L. DeRisi, A. M.
1002 Detweiler, S. L. Hao, K. N. Kangelaris, G. R. Kumar, L. M. Li, S. A. Mann, N. Neff, P. A.
55
1003 Prasad, P. H. Serpa, S. J. Shah, N. Spottiswoode, M. Tan, C. S. Calfee, S. A. Christenson, A.
1004 Kistler, C. Langelier, Upper airway gene expression reveals suppressed immune responses to
1005 SARS-CoV-2 compared with other respiratory viruses, Nat Commun 11, 5854 (2020).
1006 42. D. Bojkova, K. Klann, B. Koch, M. Widera, D. Krause, S. Ciesek, J. Cinatl, C. Münch,
1007 Proteomics of SARS-CoV-2-infected host cells reveals therapy targets, Nature 583, 469–472
1008 (2020).
1009 43. W. Kamitani, C. Huang, K. Narayanan, K. G. Lokugamage, S. Makino, A two-pronged
1010 strategy to suppress host protein synthesis by SARS coronavirus Nsp1 protein, Nat Struct Mol
1011 Biol 16, 1134–1140 (2009).
1012 44. B. Gustafson, S. Gogg, S. Hedjazifar, L. Jenndahl, A. Hammarstedt, U. Smith, Inflammation
1013 and impaired adipogenesis in hypertrophic obesity in man, Am J Physiol-endoc M 297, E999–
1014 E1003 (2009).
1015 45. C. Hepler, B. Shan, Q. Zhang, G. H. Henry, M. Shao, L. Vishvanath, A. L. Ghaben, A. B.
1016 Mobley, D. Strand, G. C. Hon, R. K. Gupta, Identification of functionally distinct fibro-
1017 inflammatory and adipogenic stromal subpopulations in visceral adipose tissue of adult mice,
1018 Elife 7, e39636 (2018).
1019 46. Y.-Y. Chau, R. Bandiera, A. Serrels, O. M. Martínez-Estrada, W. Qing, M. Lee, J. Slight, A.
1020 Thornburn, R. Berry, S. McHaffie, R. H. Stimson, B. R. Walker, R. M. Chapuli, A. Schedl, N.
1021 Hastie, Visceral and subcutaneous fat have different origins and evidence supports a mesothelial
1022 source, Nat Cell Biol 16, 367–375 (2014).
1023 47. A. J. Wilk, M. J. Lee, B. Wei, B. Parks, R. Pi, G. J. Martínez-Colón, T. Ranganath, N. Q.
1024 Zhao, S. Taylor, W. Becker, S. C.-19 Biobank, T. Ranganath, N. Q. Zhao, A. J. Wilk, R.
1025 Vergara, J. L. McKechnie, L. de la Parte, K. W. Dantzler, M. Ty, N. Kathale, G. J. Martinez-
56
1026 Colon, A. Rustagi, G. Ivison, R. Pi, M. J. Lee, R. Brewer, T. Hollis, A. Baird, M. Ugur, M. Tal,
1027 D. Bogusch, G. Nahass, K. Haider, K. Q. T. Tran, L. Simpson, H. Din, J. Roque, R. Mann, I.
1028 Chang, E. Do, A. Fernandes, S.-C. Lyu, W. Zhang, M. Manohar, J. Krempski, A. Visweswaran,
1029 E. J. Zudock, K. Jee, K. Kumar, J. A. Newberry, J. V. Quinn, D. Schreiber, E. A. Ashley, C. A.
1030 Blish, A. L. Blomkalns, K. C. Nadeau, R. O’Hara, A. J. Rogers, S. Yang, D. Jimenez-Morales,
1031 A. L. Blomkalns, R. O’Hara, E. A. Ashley, K. C. Nadeau, S. Yang, S. Holmes, M. Rabinovitch,
1032 A. J. Rogers, W. J. Greenleaf, C. A. Blish, Multi-omic profiling reveals widespread
1033 dysregulation of innate immunity and hematopoiesis in COVID-19, J Exp Med 218, e20210582
1034 (2021).
1035 48. E. Papalexi, E. P. Mimitou, A. W. Butler, S. Foster, B. Bracken, W. M. Mauck, H.-H.
1036 Wessels, Y. Hao, B. Z. Yeung, P. Smibert, R. Satija, Characterizing the molecular regulation of
1037 inhibitory immune checkpoints with multimodal single-cell screens, Nat Genet 53, 322–331
1038 (2021).
1039 49. Y. Chen, J. Wang, C. Liu, L. Su, D. Zhang, J. Fan, Y. Yang, M. Xiao, J. Xie, Y. Xu, Y. Li, S.
1040 Zhang, IP-10 and MCP-1 as biomarkers associated with disease severity of COVID-19, Mol Med
1041 26, 97 (2020).
1042 50. Y. Yang, C. Shen, J. Li, J. Yuan, J. Wei, F. Huang, F. Wang, G. Li, Y. Li, L. Xing, L. Peng,
1043 M. Yang, M. Cao, H. Zheng, W. Wu, R. Zou, D. Li, Z. Xu, H. Wang, M. Zhang, Z. Zhang, G. F.
1044 Gao, C. Jiang, L. Liu, Y. Liu, Plasma IP-10 and MCP-3 levels are highly associated with disease
1045 severity and predict the progression of COVID-19, J Allergy Clin Immun 146, 119-127.e4
1046 (2020).
1047 51. A. C. Petrey, F. Qeadan, E. A. Middleton, I. V. Pinchuk, R. A. Campbell, E. J. Beswick,
1048 Cytokine release syndrome in COVID‐19: Innate immune, vascular, and platelet pathogenic
57
1049 factors differ in severity of disease and sex, J Leukocyte Biol 109, 55–66 (2021).
1050 52. L. Zhou, K. Huntington, S. Zhang, L. Carlsen, E.-Y. So, C. Parker, I. Sahin, H. Safran, S.
1051 Kamle, C.-M. Lee, C. G. Lee, J. A. Elias, K. S. Campbell, M. T. Naik, W. J. Atwood, E.
1052 Youssef, J. A. Pachter, A. Navaraj, A. A. Seyhan, O. Liang, W. S. El-Deiry, Natural Killer cell
1053 activation, reduced ACE2, TMPRSS2, cytokines G-CSF, M-CSF and SARS-CoV-2-S
1054 pseudovirus infectivity by MEK inhibitor treatment of human cells, Biorxiv , 2020.08.02.230839
1055 (2020).
1056 53. M. Blot, M. Jacquier, L.-S. A. Glele, G. Beltramo, M. Nguyen, P. Bonniaud, S. Prin, P.
1057 Andreu, B. Bouhemad, J.-B. Bour, C. Binquet, L. Piroth, J.-P. P. de Barros, D. Masson, J.-P.
1058 Quenot, P.-E. Charles, P. study group, F. Aptel, A. Dargent, M. Georges, M. Labruyère, L.
1059 Lagrost, A. Large, S. Monier, J.-B. Roudaut, C. Thomas, CXCL10 could drive longer duration of
1060 mechanical ventilation during COVID-19 ARDS, Crit Care 24, 632 (2020).
1061 54. M. Blot, J.-B. Bour, J. P. Quenot, A. Bourredjem, M. Nguyen, J. Guy, S. Monier, M.
1062 Georges, A. Large, A. Dargent, A. Guilhem, S. Mouries-Martin, J. Barben, B. Bouhemad, P.-E.
1063 Charles, P. Chavanet, C. Binquet, L. Piroth, P. Andreu, F. Aptel, M. Labruyère, S. Prin, G.
1064 Beltramo, P. Bonniaud, P. Bielefeld, H. Devilliers, B. Bonnotte, M. Buisson, A. Putot, The
1065 dysregulated innate immune response in severe COVID-19 pneumonia that could drive poorer
1066 outcome, J Transl Med 18, 457 (2020).
1067 55. C. Huang, Y. Wang, X. Li, L. Ren, J. Zhao, Y. Hu, L. Zhang, G. Fan, J. Xu, X. Gu, Z.
1068 Cheng, T. Yu, J. Xia, Y. Wei, W. Wu, X. Xie, W. Yin, H. Li, M. Liu, Y. Xiao, H. Gao, L. Guo,
1069 J. Xie, G. Wang, R. Jiang, Z. Gao, Q. Jin, J. Wang, B. Cao, Clinical features of patients infected
1070 with 2019 novel coronavirus in Wuhan, China, Lancet 395, 497–506 (2020).
1071 56. J. T. Sims, V. Krishnan, C.-Y. Chang, S. M. Engle, G. Casalini, G. H. Rodgers, N. Bivi, B. J.
58
1072 Nickoloff, R. J. Konrad, S. de Bono, R. E. Higgs, R. J. Benschop, S. Ottaviani, A. Cardoso, A.
1073 Nirula, M. Corbellino, J. Stebbing, Characterization of the cytokine storm reflects
1074 hyperinflammatory endothelial dysfunction in COVID-19, J Allergy Clin Immun 147, 107–111
1075 (2021).
1076 57. C. R. Moderbacher, S. I. Ramirez, J. M. Dan, A. Grifoni, K. M. Hastie, D. Weiskopf, S.
1077 Belanger, R. K. Abbott, C. Kim, J. Choi, Y. Kato, E. G. Crotty, C. Kim, S. A. Rawlings, J.
1078 Mateus, L. P. V. Tse, A. Frazier, R. Baric, B. Peters, J. Greenbaum, E. O. Saphire, D. M. Smith,
1079 A. Sette, S. Crotty, Antigen-Specific Adaptive Immunity to SARS-CoV-2 in Acute COVID-19
1080 and Associations with Age and Disease Severity, Cell 183, 996-1012.e19 (2020).
1081 58. A. M. Horspool, T. Kieffer, B. P. Russ, M. A. DeJong, M. A. Wolf, J. M. Karakiozis, B. J.
1082 Hickey, P. Fagone, D. H. Tacker, J. R. Bevere, I. Martinez, M. Barbier, P. L. Perrotta, F. H.
1083 Damron, Interplay of Antibody and Cytokine Production Reveals CXCL13 as a Potential Novel
1084 Biomarker of Lethal SARS-CoV-2 Infection, Msphere 6 (2021), doi:10.1128/msphere.01324-20.
1085 59. C.-M. Su, L. Wang, D. Yoo, Activation of NF-κB and induction of proinflammatory cytokine
1086 expressions mediated by ORF7a protein of SARS-CoV-2, Sci Rep-uk 11, 13464 (2021).
1087 60. X.-X. Yin, X.-R. Zheng, W. Peng, M.-L. Wu, X.-Y. Mao, Vascular Endothelial Growth
1088 Factor (VEGF) as a Vital Target for Brain Inflammation during the COVID-19 Outbreak, Acs
1089 Chem Neurosci 11, 1704–1705 (2020).
1090 61. S. M. Metcalfe, LIF and the lung’s stem cell niche: is failure to use LIF to protect against
1091 COVID-19 a grave omission in managing the pandemic?, Future Virol 15, 659–662 (2020).
1092 62. C. Bleilevens, J. Soppert, A. Hoffmann, T. Breuer, J. Bernhagen, L. Martin, L. Stiehler, G.
1093 Marx, M. Dreher, C. Stoppe, T.-P. Simon, Macrophage Migration Inhibitory Factor (MIF)
1094 Plasma Concentration in Critically Ill COVID-19 Patients: A Prospective Observational Study,
59
1095 Diagnostics 11, 332 (2021).
1096 63. N. A. A. Öztürk, A. Ursavaş, A. G. Dilektaşli, E. Demirdöğen, N. F. Coşkun, D. Ediger, A.
1097 E. Uzaslan, D. Y. Ermiş, M. Karaca, O. E. Terzi, M. Bayram, D. Ö. Topçu, B. Yiğitliler, A.
1098 Yurttaş, S. Maharramov, G. Yazici, H. B. Oral, M. Karadağ, Interleukin-21: A Potential
1099 Biomarker For Diagnosis and Predicting Prognosis in COVID-19 Patients, Turk J Med Sci
1100 (2021), doi:10.3906/sag-2102-24.
1101 64. A. G. Vassiliou, C. Keskinidou, E. Jahaj, P. Gallos, I. Dimopoulou, A. Kotanidou, S. E.
1102 Orfanos, ICU Admission Levels of Endothelial Biomarkers as Predictors of Mortality in
1103 Critically Ill COVID-19 Patients, Cells 10, 186 (2021).
1104 65. P. Sabaka, A. Koščálová, I. Straka, J. Hodosy, R. Lipták, B. Kmotorková, M. Kachlíková, A.
1105 Kušnírová, Role of interleukin 6 as a predictive factor for a severe course of Covid-19:
1106 retrospective data analysis of patients from a long-term care facility during Covid-19 outbreak,
1107 Bmc Infect Dis 21, 308 (2021).
1108 66. K. Timper, J. L. Denson, S. M. Steculorum, C. Heilinger, L. Engström-Ruud, C. M.
1109 Wunderlich, S. Rose-John, F. T. Wunderlich, J. C. Brüning, IL-6 Improves Energy and Glucose
1110 Homeostasis in Obesity via Enhanced Central IL-6 trans-Signaling, Cell Reports 19, 267–280
1111 (2017).
1112 67. S. A. Jones, C. A. Hunter, Is IL-6 a key cytokine target for therapy in COVID-19?, Nat Rev
1113 Immunol 21, 337–339 (2021).
1114 68. N. A. Baker, L. A. Muir, C. N. Lumeng, R. W. O’Rourke, Thermogenic Fat, Methods and
1115 Protocols, Methods Mol Biology 1566, 61–76 (2017).
1116 69. T. Garin-Shkolnik, A. Rudich, G. S. Hotamisligil, M. Rubinstein, FABP4 Attenuates PPAR
1117 and Adipogenesis and Is Inversely Correlated With PPAR in Adipose Tissues, Diabetes 63,
60
1118 900–911 (2013).
1119 70. R. Nienhold, Y. Ciani, V. H. Koelzer, A. Tzankov, J. D. Haslbauer, T. Menter, N. Schwab,
1120 M. Henkel, A. Frank, V. Zsikla, N. Willi, W. Kempf, T. Hoyler, M. Barbareschi, H. Moch, M.
1121 Tolnay, G. Cathomas, F. Demichelis, T. Junt, K. D. Mertz, Two distinct immunopathological
1122 profiles in autopsy lungs of COVID-19, Nat Commun 11, 5086 (2020).
1123 71. T. Menter, J. D. Haslbauer, R. Nienhold, S. Savic, H. Hopfer, N. Deigendesch, S. Frank, D.
1124 Turek, N. Willi, H. Pargger, S. Bassetti, J. D. Leuppi, G. Cathomas, M. Tolnay, K. D. Mertz, A.
1125 Tzankov, Postmortem examination of COVID‐19 patients reveals diffuse alveolar damage with
1126 severe capillary congestion and variegated findings in lungs and other organs suggesting vascular
1127 dysfunction, Histopathology 77, 198–209 (2020).
1128 72. J. D. Haslbauer, A. Tzankov, K. D. Mertz, N. Schwab, R. Nienhold, R. Twerenbold, G.
1129 Leibundgut, A. K. Stalder, M. Matter, K. Glatz, Characterisation of cardiac pathology in 23
1130 autopsies of lethal COVID‐19, J Pathology Clin Res 7, 326–337 (2021).
1131 73. D. L. Morris, K. Singer, and C. N. Lumenga, Adipose tissue macrophages: phenotypic
1132 plasticity and diversity in lean and obese states, (2011).
1133 74. A. Lapenna, M. D. Palma, C. E. Lewis, Perivascular macrophages in health and disease, Nat
1134 Rev Immunol 18, 689–702 (2018).
1135 75. H. Koppensteiner, R. Brack-Werner, M. Schindler, Macrophages and their relevance in
1136 Human Immunodeficiency Virus Type I infection, Retrovirology 9, 82 (2012).
1137 76. J. K. Mann, T. Ndung’u, The potential of lactoferrin, ovotransferrin and lysozyme as
1138 antiviral and immune-modulating agents in COVID-19, Future Virol 15, 609–624 (2020).
1139 77. S. Lee-Huang, V. Maiorov, P. L. Huang, A. Ng, H. C. Lee, Y.-T. Chang, N. Kallenbach, P.
1140 L. Huang, H.-C. Chen, Structural and Functional Modeling of Human Lysozyme Reveals a
61
1141 Unique Nonapeptide, HL9, with Anti-HIV Activity †, Biochemistry-us 44, 4648–4655 (2005).
1142 78. J. Małaczewska, E. Kaczorek-Łukowska, R. Wójcik, A. K. Siwicki, Antiviral effects of nisin,
1143 lysozyme, lactoferrin and their mixtures against bovine viral diarrhoea virus, Bmc Vet Res 15,
1144 318 (2019).
1145 79. G. PANG, R. CLANCY, M. CONG, M. ORTEGA, R. ZHIGANG, G. REEVES, Influenza
1146 Virus Inhibits Lysozyme Secretion by Sputum Neutrophils in Subjects with Chronic Bronchial
1147 Sepsis, 161, 718–722 (2000).
1148 80. P. S. Arunachalam, F. Wimmers, C. K. P. Mok, R. A. P. M. Perera, M. Scott, T. Hagan, N.
1149 Sigal, Y. Feng, L. Bristow, O. T.-Y. Tsang, D. Wagh, J. Coller, K. L. Pellegrini, D. Kazmin, G.
1150 Alaaeddine, W. S. Leung, J. M. C. Chan, T. S. H. Chik, C. Y. C. Choi, C. Huerta, M. P.
1151 McCullough, H. Lv, E. Anderson, S. Edupuganti, A. A. Upadhyay, S. E. Bosinger, H. T.
1152 Maecker, P. Khatri, N. Rouphael, M. Peiris, B. Pulendran, Systems biological assessment of
1153 immunity to mild versus severe COVID-19 infection in humans, Science 369, 1210–1220
1154 (2020).
1155 81. A. J. Wilk, A. Rustagi, N. Q. Zhao, J. Roque, G. J. Martínez-Colón, J. L. McKechnie, G. T.
1156 Ivison, T. Ranganath, R. Vergara, T. Hollis, L. J. Simpson, P. Grant, A. Subramanian, A. J.
1157 Rogers, C. A. Blish, A single-cell atlas of the peripheral immune response in patients with severe
1158 COVID-19, Nat Med 26, 1070–1076 (2020).
1159 82. J. Tynell, V. Westenius, E. Rönkkö, V. J. Munster, K. Melén, P. Österlund, I. Julkunen,
1160 Middle East respiratory syndrome coronavirus shows poor replication but significant induction
1161 of antiviral responses in human monocyte-derived macrophages and dendritic cells, J Gen Virol
1162 97, 344–355 (2016).
1163 83. J. Zhou, H. Chu, C. Li, B. H.-Y. Wong, Z.-S. Cheng, V. K.-M. Poon, T. Sun, C. C.-Y. Lau,
62
1164 K. K.-Y. Wong, J. Y.-W. Chan, J. F.-W. Chan, K. K.-W. To, K.-H. Chan, B.-J. Zheng, K.-Y.
1165 Yuen, Active Replication of Middle East Respiratory Syndrome Coronavirus and Aberrant
1166 Induction of Inflammatory Cytokines and Chemokines in Human Macrophages: Implications for
1167 Pathogenesis, J Infect Dis 209, 1331–1342 (2014).
1168 84. C. Y. Cheung, L. L. M. Poon, I. H. Y. Ng, W. Luk, S.-F. Sia, M. H. S. Wu, K.-H. Chan, K.-
1169 Y. Yuen, S. Gordon, Y. Guan, J. S. M. Peiris, Cytokine Responses in Severe Acute Respiratory
1170 Syndrome Coronavirus-Infected Macrophages In Vitro: Possible Relevance to Pathogenesis, J
1171 Virol 79, 7819–7826 (2005).
1172 85. K. P. Y. Hui, M.-C. Cheung, R. A. P. M. Perera, K.-C. Ng, C. H. T. Bui, J. C. W. Ho, M. M.
1173 T. Ng, D. I. T. Kuok, K. C. Shih, S.-W. Tsao, L. L. M. Poon, M. Peiris, J. M. Nicholls, M. C. W.
1174 Chan, Tropism, replication competence, and innate immune responses of the coronavirus SARS-
1175 CoV-2 in human respiratory tract and conjunctiva: an analysis in ex-vivo and in-vitro cultures,
1176 Lancet Respir Medicine 8, 687–695 (2020).
1177 86. X. Cao, Y. Tian, V. Nguyen, Y. Zhang, C. Gao, R. Yin, W. Carver, D. Fan, H. Albrecht, T.
1178 Cui, W. Tan, Spike Protein of SARS-CoV-2 Activates Macrophages and Contributes to
1179 Induction of Acute Lung Inflammations in Mice, Biorxiv , 2020.12.07.414706 (2020).
1180 87. R. A. Grant, L. Morales-Nebreda, N. S. Markov, S. Swaminathan, M. Querrey, E. R.
1181 Guzman, D. A. Abbott, H. K. Donnelly, A. Donayre, I. A. Goldberg, Z. M. Klug, N. Borkowski,
1182 Z. Lu, H. Kihshen, Y. Politanska, L. Sichizya, M. Kang, A. Shilatifard, C. Qi, J. W. Lomasney,
1183 A. C. Argento, J. M. Kruser, E. S. Malsin, C. O. Pickens, S. B. Smith, J. M. Walter, A. E.
1184 Pawlowski, D. Schneider, P. Nannapaneni, H. Abdala-Valencia, A. Bharat, C. J. Gottardi, G. R.
1185 S. Budinger, A. V. Misharin, B. D. Singer, R. G. Wunderink, R. A. Grant, L. Morales-Nebreda,
1186 N. S. Markov, S. Swaminathan, M. Querrey, E. R. Guzman, D. A. Abbott, H. K. Donnelly, A.
63
1187 Donayre, I. A. Goldberg, Z. M. Klug, N. Borkowski, Z. Lu, H. Kihshen, Y. Politanska, L.
1188 Sichizya, M. Kang, A. Shilatifard, C. Qi, J. W. Lomasney, A. C. Argento, J. M. Kruser, E. S.
1189 Malsin, C. O. Pickens, S. B. Smith, J. M. Walter, A. E. Pawlowski, D. Schneider, P.
1190 Nannapaneni, H. Abdala-Valencia, A. Bharat, C. J. Gottardi, G. R. S. Budinger, A. V. Misharin,
1191 B. D. Singer, R. G. Wunderink, A. A. Wagh, A. R. Hauser, A. R. Wolfe, A. Thakrar, A. V.
1192 Yeldandi, A. A. Wang, A. R. Levenson, A. M. Joudi, B. Tran, C. A. Gao, C. Kurihara, C. J.
1193 Schroedl, C. M. Horvath, D. Meza, D. D. Odell, D. W. Kamp, D. R. Winter, E. A. Ozer, E. D.
1194 Shanes, E. T. Bartom, E. J. Rendleman, E. M. Leibenguth, F. Wehbe, G. Y. Liu, G. T. Gadhvi,
1195 H. T. Navarro, J. I. Sznajder, J. E. Dematte, J. Le, J. M. Arnold, J. C. Du, J. Coleman, J. I.
1196 Bailey, J. S. Deters, J. A. Fiala, J. Starren, K. M. Ridge, K. Secunda, K. Aren, K. L. Gates, K.
1197 Todd, L. D. Gradone, L. N. Textor, L. F. Wolfe, L. L. Pesce, L. A. N. Amaral, M. L.
1198 Rosenbaum, M. Kandpal, M. Jain, M. A. Sala, M. Saine, M. Carns, M. J. Alexander, M. J.
1199 Cuttica, M. H. Prickett, N. H. Khan, N. S. Chandel, N. D. Soulakis, O. R. Rivas, P. C. Seed, P.
1200 A. Reyfman, P. D. Go, P. H. S. Sporn, P. R. Cooper, R. Tomic, R. Patel, R. Garza-Castillon, R.
1201 Kalhan, R. I. Morimoto, R. J. Mylvaganam, S. S. Kim, S. W. M. Gatesy, S. Thakkar, S. B.
1202 Maamar, S. Han, S. R. Rosenberg, S. Nozick, S. J. Green, S. R. Russell, T. A. Poor, T. J. Zak, T.
1203 A. Lombardo, T. Stoeger, T. Shamaly, Z. Ren, Circuits between infected macrophages and T
1204 cells in SARS-CoV-2 pneumonia, Nature 590, 635–641 (2021).
1205 88. D. Kim, J.-Y. Lee, J.-S. Yang, J. W. Kim, V. N. Kim, H. Chang, The Architecture of SARS-
1206 CoV-2 Transcriptome, Cell 181, 914-921.e10 (2020).
1207 89. R. Verma, E. Kim, G. J. Martinez-Colón, P. Jagannathan, A. Rustagi, J. Parsonnet, H.
1208 Bonilla, C. Khosla, M. Holubar, A. Subramanian, U. Singh, Y. Maldonado, C. A. Blish, J. R.
1209 Andrews, SARS-CoV-2 subgenomic RNA kinetics in longitudinal clinical samples, Open Forum
64
1210 Infect Dis 8, ofab310- (2021).
1211 90. D. E. Dimcheff, A. L. Valesano, K. E. Rumfelt, W. J. Fitzsimmons, C. Blair, C. Mirabelli, J.
1212 G. Petrie, E. T. Martin, C. Bhambhani, M. Tewari, A. S. Lauring, SARS-CoV-2 Total and
1213 Subgenomic RNA Viral Load in Hospitalized Patients, J Infect Dis , jiab215- (2021).
1214 91. M. Bushell, P. Sarnow, Hijacking the translation apparatus by RNA viruses, J Cell Biology
1215 158, 395–399 (2002).
1216 92. B. Gustafson, U. Smith, Cytokines Promote Wnt Signaling and Inflammation and Impair the
1217 Normal Differentiation and Lipid Accumulation in 3T3-L1 Preadipocytes*, J Biol Chem 281,
1218 9507–9516 (2006).
1219 93. M. Watanabe, D. Caruso, D. Tuccinardi, R. Risi, M. Zerunian, M. Polici, F. Pucciarelli, M.
1220 Tarallo, L. Strigari, S. Manfrini, S. Mariani, S. Basciani, C. Lubrano, A. Laghi, L. Gnessi,
1221 Visceral fat shows the strongest association with the need of intensive care in patients with
1222 COVID-19, Metabolis 111, 154319 (2020).
1223 94. H. Chandarana, B. Dane, A. Mikheev, M. T. Taffel, Y. Feng, H. Rusinek, Visceral adipose
1224 tissue in patients with COVID-19: risk stratification for severity, Abdom Radiol 46, 818–825
1225 (2021).
1226 95. S. Battisti, C. Pedone, N. Napoli, E. Russo, V. Agnoletti, S. G. Nigra, C. Dengo, M.
1227 Mughetti, C. Conte, P. Pozzilli, E. Giampalma, R. Strollo, Computed Tomography Highlights
1228 Increased Visceral Adiposity Associated With Critical Illness in COVID-19, Diabetes Care 43,
1229 e129–e130 (2020).
1230 96. S. M. Reilly, A. R. Saltiel, Adapting to obesity with adipose tissue inflammation, Nat Rev
1231 Endocrinol 13, 633–643 (2017).
1232 97. B. Glatthaar‐Saalmüller, K. H. Mair, A. Saalmüller, Antiviral activity of aspirin against RNA
65
1233 viruses of the respiratory tract—an in vitro study, Influenza Other Resp 11, 85–92 (2017).
1234 98. A. B. Goldfine, V. Fonseca, K. A. Jablonski, Y.-D. I. Chen, L. Tipton, M. A. Staten, S. E.
1235 Shoelson, T. I. U. S. in T. 2 D. S. Team, Salicylate (Salsalate) in Patients With Type 2 Diabetes:
1236 A Randomized Trial, Ann Intern Med 159, 1 (2013).
1237 99. V. Bianconi, F. Violi, F. Fallarino, P. Pignatelli, A. Sahebkar, M. Pirro, Is Acetylsalicylic
1238 Acid a Safe and Potentially Useful Choice for Adult Patients with COVID-19 ?, Drugs 80, 1383–
1239 1396 (2020).
1240 100. S. E, Y. ZX, F. VJ, H. ES, E. SE., Aspirin attenuates cytomegalovirus infectivity and gene
1241 expression mediated by cyclooxygenase-2 in coronary artery smooth muscle cells., Circ Res.
1242 (1998), doi:10.1161/01.res.83.2.210.
1243 101. A. Sánchez-García, C. P. Ríos-Ibarra, A. R. Rincón-Sánchez, R. Ortiz-López, A. Garza-
1244 Juárez, J. Morlett-Chávez, H. Martínez-Rodríguez, A. M. Rivas-Estilla, Use of proteomic
1245 analysis tools to identify HCV-proteins down-regulated by acetylsalicylic acid, Ann Hepatol 12,
1246 725–732 (2013).
1247 102. C. Müller, M. Hardt, D. Schwudke, B. W. Neuman, S. Pleschka, J. Ziebuhr, Inhibition of
1248 Cytosolic Phospholipase A2α Impairs an Early Step of Coronavirus Replication in Cell Culture,
1249 J Virol 92, e01463-17 (2018).
1250 103. P. V, B. S, D. B. G, B. M, In vitro activity of acetylsalicylic acid on replication of varicella-
1251 zoster virus, New Microbiol. (n.d.).
1252 104. S. F. Masre, N. F. Jufri, F. W. Ibrahim, S. H. A. Raub, Classical and alternative receptors
1253 for SARS‐CoV‐2 therapeutic strategy, Rev Med Virol 31, 1–9 (2021).
1254 105. M. Puray-Chavez, K. M. LaPak, T. P. Schrank, J. L. Elliott, D. P. Bhatt, M. J. Agajanian, R.
1255 Jasuja, D. Q. Lawson, K. Davis, P. W. Rothlauf, Z. Liu, H. Jo, N. Lee, K. Tenneti, J. E.
66
1256 Eschbach, C. S. Mugisha, E. M. Cousins, E. W. Cloer, H. R. Vuong, L. A. VanBlargan, A. L.
1257 Bailey, P. Gilchuk, J. E. Crowe, M. S. Diamond, D. N. Hayes, S. P. J. Whelan, A. Horani, S. L.
1258 Brody, D. Goldfarb, M. B. Major, S. B. Kutluay, Systematic analysis of SARS-CoV-2 infection
1259 of an ACE2-negative human airway cell, Cell Reports 36, 109364 (2021).
1260 106. J. Davies, H. S. Randeva, K. Chatha, M. Hall, D. A. Spandidos, E. Karteris, I. Kyrou,
1261 Neuropilin-1 as a new potential SARS-CoV-2 infection mediator implicated in the neurologic
1262 features and central nervous system involvement of COVID-19, Mol Med Rep 22, 4221–4226
1263 (2020).
1264 107. P. VG, L. M, L. MT, S. JP, W. MN, A. L, C. S, H. A, W. N, L. S, B. F, L. S, P. S, S. AS, E.
1265 C, G. O, G. M, W. D, W. T, K. S, P. K, A. M, H. TB., Multiorgan and Renal Tropism of SARS-
1266 CoV-2, N Engl J Med (2020), doi:10.1056/nejmc2011400.
1267 108. L. J. Liew, H. T. Ong, R. J. Dilley, Isolation and Culture of Adipose-Derived Stromal Cells
1268 from Subcutaneous Fat, Methods Mol Biology 1627, 193–203 (2017).
1269 109. R. Wölfel, V. M. Corman, W. Guggemos, M. Seilmaier, S. Zange, M. A. Müller, D.
1270 Niemeyer, T. C. Jones, P. Vollmar, C. Rothe, M. Hoelscher, T. Bleicker, S. Brünink, J.
1271 Schneider, R. Ehmann, K. Zwirglmaier, C. Drosten, C. Wendtner, Virological assessment of
1272 hospitalized patients with COVID-2019, Nature 581, 465–469 (2020).
1273 110. H. Bösmüller, S. Traxler, M. Bitzer, H. Häberle, W. Raiser, D. Nann, L. Frauenfeld, A.
1274 Vogelsberg, K. Klingel, F. Fend, The evolution of pulmonary pathology in fatal COVID-19
1275 disease: an autopsy study with clinical correlation, Virchows Arch 477, 349–357 (2020).
1276 111. D. Szklarczyk, A. L. Gable, D. Lyon, A. Junge, S. Wyder, J. Huerta-Cepas, M. Simonovic,
1277 N. T. Doncheva, J. H. Morris, P. Bork, L. J. Jensen, C. von Mering, STRING v11: protein–
1278 protein association networks with increased coverage, supporting functional discovery in
67
1279 genome-wide experimental datasets, Nucleic Acids Res 47, gky1131 (2018).
1280 112. G. Korotkevich, V. Sukhov, N. Budin, B. Shpak, M. N. Artyomov, A. Sergushichev, Fast
1281 gene set enrichment analysis, Biorxiv, 060012 (2021).
1282 113. S. Al-Benna, Association of high level gene expression of ACE2 in adipose tissue with
1283 mortality of COVID-19 infection in obese patients, Obes Medicine 19, 100283 (2020).
1284 114. S. Gómez-Zorita, I. Milton-Laskibar, L. García-Arellano, M. González, M. P. Portillo, An
1285 Overview of Adipose Tissue ACE2 Modulation by Diet and Obesity. Potential Implications in
1286 COVID-19 Infection and Severity, Int J Mol Sci 22, 7975 (2021).
1287 115. Z. Daniloski, T. X. Jordan, H.-H. Wessels, D. A. Hoagland, S. Kasela, M. Legut, S.
1288 Maniatis, E. P. Mimitou, L. Lu, E. Geller, O. Danziger, B. R. Rosenberg, H. Phatnani, P.
1289 Smibert, T. Lappalainen, B. R. tenOever, N. E. Sanjana, Identification of Required Host Factors
1290 for SARS-CoV-2 Infection in Human Cells, Cell 184, 92-105.e16 (2021).
1291
1292 Acknowledgements: We thank the patients and their families for their consent to use their
1293 tissues. We would like to thank the Stanford Bariatric Surgery and Cardiothoracic Surgery clinic
1294 staff for assisting with participant recruitment and tissue harvesting. We thank Dr. Karin Klingel,
1295 Dr. Selina Traxler, Karen Greif, Dr. Massimo Granai and Dr. Hans Bösmüller (Department of
1296 Pathology, University Hospital Tübingen, Germany) for providing samples and logistic support.
1297 We would like to thank Anna Stalder and Jan Schneeberger (Institute of Medical Genetics and
1298 Pathology, University Hospital of Basel, Switzerland) for data analysis and technical assistance.
1299 We are thankful to Ralf Bartenschlager (University of Heidelberg, Germany) for providing
1300 A549-ACE2. The following reagent was deposited by the Centers for Disease Control and
1301 Prevention and obtained through BEI Resources, NIAID, NIH: SARS-Related Coronavirus 2,
68
1302 Isolate USA-WA1/2020, NR-52281. We thank Yael Rosenberg-Hasson, technical director of
1303 immunological assays at Stanford University’s HIMC, for Luminex assays. We thank Jaishree
1304 Garhyan, director of Stanford’s BSL3 service center, for BSL3 surveillance and training. We are
1305 grateful to Angela Detweiler and the Chan Zuckerberg Biohub foundation for sequencing.
1306
1307 Funding:
1308 National Institutes of Health grant R21AI159024 (TLM)
1309 American Diabetes Association grant 7-20-COVID-213 (TLM)
1310 Stanford University Innovative Medicines Accelerator COVID-19 Response grant (CAB, TLM)
1311 Botnar Research Centre for Child Health grant Emergency Response to COVID-19 grant (SJ,
1312 CMS, KDM, AT, MSM, GPN)
1313 Swiss National Science Foundation grant 320030_189275 (MSM)
1314 Chan Zuckerberg Biohub Investigator Program (CAB)
1315 National Institutes of Health grant 5T32 AI007502 (AR)
1316 National Science Foundation Graduate Research Fellowship 2019282939 (KR)
1317 Bill and Melinda Gates Foundation OPP1113682 (JRA)
1318
1319 Author contributions:
1320 Conceptualization: GMC, HeC, CAB, TLM
1321 Methodology: GMC, KR, HeC, SJ, HaC, CAB, TLM
1322 Investigation: GMC, KR, HeC, SZ, SJ, CMS, MSM, HaC, RV, AR
1323 Resources: EZ, DA, JB, AT, KDM, MSM, CMS, CAB, TLM
1324 Data Curation: GMC, KR, CAB, TLM
69
1325 Visualization: GMC, KR, SZ, SJ, CMS, MSM
1326 Funding acquisition: CAB, TLM, SJ, CMS, AT, KDM, MSM, GPN
1327 Supervision: JRA, GPN, CAB, TLM
1328 Writing – original draft: GMC, KR, HeC, CAB, TLM
1329 Writing – review & editing: GMC, KR, HeC, SJ, EZ, AR, RV, HaC, JRA, DA, JB, GPN, CMS,
1330 MSM, CAB, TLM
1331
1332 Competing interests: CAB is on the Scientific Advisory Boards of Catamaran Bio and
1333 DeepCell. CMS is on the Scientific Advisory Board of and has received research funding from
1334 Enable Medicine, Inc., both outside the current work. MSM has served as a consultant for
1335 Novartis and Glaxo Smith Kline and received speaker's honoraria from ThermoFisher and
1336 Merck, all outside the current work.
1337
1338 Data and materials availability: Data from scRNA-seq will be deposited with the Gene
1339 Expression Omnibus. A Github repository for all original code used for analysis and
1340 visualization will be made public upon publication.
1341 Figures
1342 Fig. 1. Exposure of stromal vascular cells (SVC) from adipose tissue to SARS-CoV-2
1343 supports infection of macrophages. (A) Sketch of workflow. SVC of human adipose tissue was
1344 isolated by collagenase digestion prior to viral infection. SVC was infected or left untreated
1345 (mock) with SARS-CoV-2 (USA-WA1/2020) at a multiplicity of infection (MOI) of 1. (B)
1346 Relative gene expression of SARS-CoV-2 (N gene) obtained by 1-step RTqPCR.at 24 hpi or
1347 mock infection (SAT, n=6; VAT, n=4). (C) Relative gene expression of SARS-CoV-2 (N gene)
70
1348 obtained by 1-step RTqPCR in cultures were treated with vehicle (DMSO) or 10µM remdesivir
1349 to inhibit viral replication and maintained for 24, 48, 72, and 96 hours before harvest. Relative
1350 expression was analyzed by ΔΔCt method relative to a mock sample using 18S rRNA as a
1351 housekeeping gene. Each data point is an average of 3 technical replicates. (D) Frequency of
1352 SARS-CoV-2 infected cells based on SARS-CoV-2 N protein detection by flow cytometry of
1353 SAT (left, n=2) and VAT (right, n=2). Gating is detailed in supplemental figure 3. Statistical
1354 analysis: (A) paired, two-sided, student’s t-test. **P<0.01, ***P<0.001. (B) Statistical analysis
1355 was performed with a two-way ANOVA, multiple comparisons using statistical hypothesis
1356 Sidak. *P<0.05. Data is presented as ± mean s.e.m. (A) Sketch was created with BioRender.
1357 Fig. 2. A subset of macrophages is infected with SARS-CoV-2. (A) Schematic of experiment.
1358 The stromal vascular cells (SVC) were isolated from the SAT and VAT depots of three different
1359 participants and infected with either mock or SARS-CoV-2 (MOI of 1.0). Each sample was
1360 collected for scRNA-seq at 24 hpi. (B-C) UMAP representation of the SVC from all participants
1361 (n=3) across 198,759 cells, (B) colored by manually annotated cell type and (C) colored by
1362 infection and depot. (D) UMAP representation of all cells colored by SARS-CoV-2 cpm (log10).
1363 (E) Violin plot reveals the SARS-CoV-2 cpm values of all cells across SARS-CoV-2 infected
1364 samples, showing only the 8 cell clusters with the highest composition of SARS-CoV-2+ cells.
1365 Percentages above each cell type denote the percentage of cells within each cluster that have
1366 over 10 SARS-CoV-2 reads. (F) UMAP projections of all macrophages from the scRNA-seq
1367 dataset, colored by SARS-CoV-2 cpm (left), colored by CD68 expression (middle) and colored
1368 by ACE2 expression (right). (G) Dotplot of the proportion of cells (dot size) in the macrophage
1369 clusters split by infection condition expressing genes relevant for SARS-CoV-2 entry and
1370 antiviral defense, as well as SARS-CoV-2 cpm and macrophage cluster markers and colored by
71
1371 scaled average expression. (A) Sketch was created with BioRender.
1372 Fig. 3. The infected macrophage cluster is marked by increased chemokine expression. (A)
1373 Volcano plot of the differentially expressed genes between macrophage clusters 2 (C2) and 12
1374 (C12) across mock-infected samples. (B) Venn Diagram comparison of the significantly
1375 differentially expressed genes (DEGs) and their direction of change across C2 versus C12 in (1)
1376 mock-infected, (2) all SAT and (3) all VAT, and (4) SARS-CoV-2-infected conditions. (C)
1377 Heatmap of the most significant DEGs between SARS-CoV-2+ versus bystander macrophages
1378 within C2. (D) Normalized enrichment scores of top Reactome pathways, using significant
1379 DEGs between SARS-CoV-2+ versus bystander C2 macrophages.
1380 Fig. 4. Preadipocytes respond to SARS-CoV-2 exposure. (A,B) UMAP embedding of all
1381 preadipocytes (n=140,867) colored by (A) cluster and (B) sample and infection type. (C) Cell
1382 fraction bar plot clustered by sample and infection type within each cluster. (D) Feature plots
1383 depicting expression of selected markers associated with preadipocyte cell states, cell types and
1384 antiviral genes. (E) Box plots of average cytokine (top) and ISG- (bottom) module scores across
1385 the preadipocytes of each participant and depot in both mock and SARS-CoV-2 infection
1386 conditions. (F) Reactome pathway analysis was performed on the significant DEGs by
1387 participant and cluster within SAT. Pathways that were represented and significant in at least
1388 four of the participant-cluster subsets were included. Pathways clustered by Euclidean distance
1389 (tree not shown) and split by the two major subtrees.
1390 Fig. 5. Mature and in vitro differentiated adipocytes support SARS-CoV-2 infection.
1391 Mature adipocytes (MA) of human adipose tissue were isolated by collagenase digestion prior to
1392 viral infection. MA were infected or left untreated (mock) for 24 hours with SARS-COV-2
72
1393 (USA-WA1/2020) at a MOI:1. (A) Measurements of genomic (gRNA) and subgenomic
1394 (sgRNA) SARS-CoV-2 genome copy numbers in infected MA from subcutaneous (left; n=3)
1395 and visceral (right; n=2 omentum; n=1 epicardial; n=1 pericardial) adipose tissue obtained by
1396 absolute gene quantification using 1-step RTqPCR TaqManTM and reported as fold change of
1397 infected to mock. (B) Relative gene expression of N gene in adipocytes differentiated in vitro
1398 from pericardial preadipocytes using adipocyte differentiation media for 0, 3, 6, and 13 days
1399 before left untreated or infected with SARS-CoV-2 at a MOI:1. Results were obtained by 1-step
1400 RTqPCR and analyzed by ΔΔCt method using 18s as a housekeeping gene. (C) Adipocytes
1401 differentiated in vitro from preadipocytes obtained from pericardial adipose tissue were infected
1402 or mock infected with SARS-COV-2 (USA-WA1/2020) at a MOI of 1 for 1 hour, followed by
1403 washing and removing the virus and replacing with media treated with vehicle (DMSO) or 10µM
1404 remdesivir to inhibit viral replication. Cultures were maintained for 24, 48, 72, and 96 hpi, after
1405 which gene expression was obtained by 1-step RTqPCR for the N gene. Relative expression was
1406 analyzed by ΔΔCt method relative to the mock sample. (C) Each data point is an average of 3
1407 technical replicates and are presented as ± mean s.e.m.
1408 Fig. 6. SARS-CoV-2 RNA and immune infiltration are present in adipose tissue of autopsy
1409 samples from COVID-19 patients. RNA in situ hybridization (ISH) on epicardial fat from heart
1410 autopsies from patients (Top, autopsy #9; Bottom, autopsy #10) who succumbed to COVID-19.
1411 Assays were performed using probes against SARS-CoV-2 Spike mRNA. Red arrowheads show
1412 ISH positive signals and blue arrows show inflammatory cells. (A and C) Overview of the heart
1413 tissue section (2mm), and (B and D) magnified view (20µm) of the represented region. (E)
1414 Interface of epicardial fat and myocardium (50µm). Note the inflammatory infiltration (blue
1415 arrows) only in the epicardial fat. Image has been rotated 90°.
73
1416 Table 1. 80-plex Luminex analysis of stromal vascular cells from human adipose tissue exposed
1417 to SARS-CoV-2. Table summarizing results of 80-plex Luminex assay performed in
1418 supernatants of SVC (SAT, n=8; VAT, n=6) cultures that were infected with SARS-CoV-2 or
1419 mock-infected for 24 hours. First column shows analytes of interest. Second column categorizes
1420 analytes in either chemokine, growth factor, Th2 cytokine, inflammatory cytokine, cell adhesion
1421 molecule, or hematopoiesis regulatory cytokine. Columns 3 to 6 summarizes statistical results
1422 split by mean fold change and p and q (Adjusted p value) value obtained from Wilcoxon signed
1423 rank test. Columns 7 and 8 summarizes the known role of each analyte in COVID-19. Bottom of
1424 the table contains abbreviations and descriptions of statistics.
1425 Table 2. Evidence of SARS-CoV-2 detection in adipose tissue from deceased individuals. Table
1426 summarizing SARS-CoV-2 PCR signals from 8 autopsy cases, seven of them from patients
1427 deceased from COVID-19. All threshold cycles (Ct) presented are an average of three SARS-
1428 CoV-2 gene sequences for open reading frame 1ab (ORF1ab), spike (S), and nucleocapsid (N).
1429 RNA was isolated from either epicardial, visceral, or subcutaneous fat. Heart RNA samples were
1430 collected from a mixture of myocardium and epicardium. Samples not presented in this study
1431 were marked as not available “N/A”. When the PCR signal was undetected, the sample was
1432 marked as “UD”.
1433
1434 SUPPLEMENTARY MATERIAL
1435 Materials and Methods
1436 Adipose Tissue Processing: Adipose tissue samples were minced and then digested by
1437 collagenase I (1 mg/mL) (Worthington Biochem. Corp, USA) at 37ºC for 60 minutes, in KRBH
74
1438 buffer containing BSA (2%), adenosine (250 uM), and P/S (1x), then filtered through a 500-µm
1439 nylon mesh, followed by separation of MA and SVC. MA were utilized immediately; SVC was
1440 collected by centrifugation at 500 xg for 5 minutes at room temperature. The SVC pellet was
1441 incubated in erythrocyte lysis buffer (Invitrogen) for 10 minutes at 37ºC, followed by
1442 centrifugation as above. The cell pellet was resuspended in HBSS (with 2% BSA) and
1443 centrifuged for another 5 minutes at 500 xg at RT. The SVC pellets were resuspended in growth
1444 medium (DMEM/F12, 10% FBS, and 1% P/S), filtered through a 75-µm cell strainer, and
1445 cultured for expansion or immediate use.
1446
1447 Differentiation of human preadipocytes: As previously described (107), isolated SVC was
1448 expanded in DMEM/F12 containing FBS (10%) and P/S (1%), split to expand, and cultured until
1449 confluence. When cells reached 100% confluence, differentiation was induced using
1450 differentiation media, DM-2 (Zenbio, Inc) consisting of insulin, dexamethasone and
1451 isobutylmethylxanthine (IBMX) and pioglitazone supplemented with 10% FBS. After 3 days, the
1452 media was changed to adipocyte maintenance media, AM-1 (Zenbio, Inc) containing only
1453 insulin, dexamethasone in DMEM/F12 supplemented with 10% FBS (Fetal bovine serum). The
1454 cells were then left to differentiate for 3, 6, 12, and 14 days, with culture medium (AM-1)
1455 changed every 3 days. Day 0 preadipocytes were collected at time of confluence, just before
1456 adipogenesis induction. During differentiation, cells were collected at all the other time points
1457 and utilized immediately for experiments as described. Adipogenesis was confirmed by oil
1458 droplet formation and by increased expression of fatty acid-binding protein 4 gene, Fabp4.
1459
75
1460 SARS-CoV-2 infections of differentiated adipocytes, SVC, and A549-ACE2: Cells were
1461 seeded a day before infection by culturing 4x105-1x106 cells per well in a 6-well plate (Corning).
1462 On the day of infection, SVC was centrifuged at 500g for 5mins, and washed with infection
1463 media (DMEM, 2%FBS, and 1% Pen-strep). Adherent differentiated adipocytes and A549-
1464 ACE2 were washed with infection media. A549-ACE2 cells were cultured under the presence of
1465 623µg/ml of Geneticin (Thermo Fisher Scientific; 10121035) for selection of ACE2 expressing
1466 cells. Viral infection was performed with SARS-CoV-2 (2019-nCOV/USA-WA1/2020) at a
1467 MOI of 1 for 1 hour while gently rocking before cells were washed and culture in culture media
1468 (DMEM, 10%FBS, and 1% Pen-strep) at 37oC with 5%CO2 under BSL3 containment. When
1469 remdesivir was used, cells were subjected to multiple washes with PBS 1 hour after infection
1470 before culturing under the presence of vehicle, dimethylsulfoxide (DMSO) (Sigma; D2650), or
1471 10µM Remdesivir (Gilead) in DMSO and cultured for longer periods (24, 48, 72, and 96hpi).
1472
1473 SARS-CoV-2 infections of MA: MA media was changed into infection media by penetrating
1474 the fat tissue with a 22-gauge polytetrafluoroethylene (PTFE) (Grainger) blunt needle attached to
1475 a 3cc syringe (Grainger), or by gently transferring the fat with a wide manually cut p100 tip into
1476 a 5ml conical collection tube containing 2mls of warm media. Viral infection was performed
1477 with the WA/01 strain of SARS-COV-2 (2019-nCOV/USA-WA1/2020) at a multiplicity of
1478 infection (MOI) of 1 by gently adding virus to the top of the floating cells under BSL3
1479 containment. MA was incubated for 1 hour while gently rocking at 37oC with 5%CO2. Media
1480 was then removed using as mentioned above prior to adding culture media (DMEM, 10%FBS,
1481 and 1% Pen-strep) and incubating at 37oC with 5%CO2.
1482
76
1483 Flow Cytometry: Cells were collected using Trypsin-EDTA (Thermo Scientific, 25200072) for
1484 5-10mins at 37oC and 5% CO2. Cells were then centrifuged, 500g for 5 minutes, and washed
1485 with PBS (Thermo Scientific, 10010023) and transferred into a 96-well plate, then washed twice
1486 more with FACS buffer (PBS, 2%FBS (Corning, MT35016CV), and 5mM EDTA (Hoefer,
1487 GR123)). Cells were incubated in Zombie aqua cell viability dye (BioLegend, 423102) at room
1488 temperature for 20 minutes, washed twice, incubated at room temperature with Fc block
1489 (BioLegend, 422302) for 5 minutes, and surface stained for 30 minutes at 4oC. Surface staining
1490 contained antibodies for CD45 (BioLegend, 368503), CD3 (BioLegend, 317343), CD14
1491 (BioLegend,301835), CD19 (BioLegend, 302261), CD11c (BioLegend, 371503), CD34
1492 (BioLegend, 343615), CD31 (BioLegend, 303133). In fig. S4, surface staining also contained
1493 antibodies against ACE2 (R&D, fab9332r) or isotype control (R&D, ICOO3R). After surface
1494 staining, cells were washed twice with FACS buffer before fixation with 4% Paraformaldehyde
1495 (PFA) (Electron Microscopy Sciences, 15710) for 1 hour at 4oC, cells were then permeabilized
1496 (eBioscience, 00-8333-56) for 10 minutes at room temperature before intracellular staining for
1497 45 minutes at 4oC. Intracellular staining contained an anti-SARS-CoV-2 N protein antibody
1498 (Sino Biological, 40143-T62). After intracellular staining, cells were washed with PBS, and
1499 diluted in 1%PFA and PBS before analyzing in a CytekTM Aurora. FCS files were collected and
1500 analyzed using FlowJoTM. Gating was done as shown in Fig. 1A.
1501
1502 1-step RTqPCR for relative gene quantification: RNA was then mixed with TaqPath 1-step
1503 RTqPCR master mix (Applied Biosystems, A15299) and primers for SARS-CoV-2 N-gene
1504 (Biosearch technologies; n2019-nCoV KIT-NCOV-PP1-1000), IL-6 (Thermo scientific,
1505 Hs00174131_m1), IFNA1 (Thermo scientific, Hs04189288_g1), IFNB1 (Thermo scientific,
77
1506 Hs01077958_s1), ISG15 (Thermo scientific, Hs01921425_s1), IFI27 (Thermo Scientific,
1507 Hs01086373_g1),and IER3 (Thermo scientific, Hs00174674_m1) for relative gene
1508 quantification. Relative expression was calculated by obtaining ∆∆CT, using the endogenous
1509 control eukaryotic 18S rRNA (Thermo Scientific, 4352930E) and mock samples as the calibrator
1510 control. The samples were analyzed on a QuantStudioTM 3 (Applied Biosystems) with the
1511 following parameters: 2 minutes at 25oC for uracil-DNA glycosylase (UNG) incubation, 15
1512 minutes at 50oC for RT, 2 minutes at 95oC for polymerase activation, and 40 cycles of 15
1513 seconds at 95oC and 30 seconds of 60oC for amplification.
1514
1515 2-step RTqPCR for ACE2, Fabp4, and 18s: RT on isolated RNA was performed with
1516 Superscript III first-strand synthesis system (Thermo scientific, 18080051). For qPCR, <100ng
1517 of cDNA were mixed with designed reverse and forward primers and PowerUp SYBR green
1518 master mix (Applied biosystems, A25742). The samples were analyzed on a QuantStudioTM 3
1519 (Applied Biosystems) with the following parameters: 10 minutes at 95oC for polymerase
1520 activation, and 40 cycles of 15 seconds at 95oC and 60 seconds of 60oC for amplification. The
1521 designed primers were the following: ACE2 forward (TAACCACGAAGCCGAAGACC) and
1522 reverse (TTGGGCAAGTGTGGACTGTT), Fabp4 forward (TGGGCCAGGAATTTGACGAA),
1523 and reverse (CACATGTACCAGGACACCCC); 18s forward
1524 (GGCCCTGTAATTGGAATGAGTC) and reverse (CCAAGATCCAACTACGAGCTT). All
1525 primers were designed with mRNA sequence using National Center for Biotechnology
1526 Information primer designing tool, Primer-BLAST. All sequences are Sequence 5’→3’.
1527
78
1528 RNA-ISH: For deparaffinization, slides were baked at 70 °C for 1 hour in a temperature-
1529 controlled oven, then immersed in fresh xylene twice for 5 minutes each. Rehydration was
1530 performed using a Leica ST4020 Linear Stainer (Leica Biosystems) programmed to three dips
1531 per wash for 180 seconds each with the following buffers: xylene x 3, 100% ethanol x 2, 95%
1532 ethanol x 2, 80% ethanol x 1, 70% ethanol x 1, and ddH2O x 3. Heat induced epitope retrieval
1533 was subsequently performed in a Lab Vision PT module (Thermo Fisher) using the Dako Target
1534 Retrieval Solution, pH 9 (DAKO Agilent, S236784-2) at 97 °C for 10 minutes, followed by
1535 controlled cooling down to 65 °C. Slides were removed from the PT module and further cooled
1536 to room temperature for 30 minutes before rinsing briefly in sterile, nuclease-free water (ddH2O)
1537 twice. A 15 minutes hydrogen peroxide block was subsequently performed at 40 °C (Bio-
1538 Techne, 322335). Slides were then washed twice for 2 minutes each in ddH2O before an
1539 overnight hybridization at 40 °C with probes against the SARS-CoV-2 Spike mRNA (Bio-
1540 Techne, 848561). Amplification of the ISH probes was performed the next day according to the
1541 manufacturer's protocol (Bio-Techne, 322350). Slides were then dried, mounted, and
1542 subsequently scanned via an Aperio slide scanner at the Stanford School of Medicine Histology
1543 Core.
1544
1545
79

Fat Study

Uploaded by

Document Informationclick to expand document informationcovid

Document Informationclick to expand document information

Copyright:

Available Formats

Fat Study

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fat Study

Uploaded by

Copyright:

Available Formats

bioRxiv preprint doi: https://doi.org/10.1101/2021.10.24.465626; this version posted October 25, 2021.

The copyright holder for this preprint

2 consistent with severe COVID-19

5 Giovanny J. Martínez-Colón1†, Kalani Ratnasiri2†, Heping Chen1, Sizun Jiang5,9, Elizabeth

8 Matthias S. Matter8, Catherine A. Blish1,2,6†*, Tracey L. McLaughlin1†*

19 Center Tübingen, Tübingen, Germany

30 Department of Medicine, Division of Endocrinology, Gerontology, and Metabolism

31 300 Pasteur Dr,

36 Department of Medicine, Division of Infectious Disease and Geographic Medicine

37 300 Pasteur Dr., Lane Building, L134

47 in human adipose tissue and describes the associated inflammation.

53 elicits an inflammatory response, including the secretion of known inflammatory mediators of

56 largely restricted to a highly inflammatory subpopulation of macrophages, present at baseline,

58 adopt a proinflammatory phenotype. We further demonstrate that SARS-CoV-2 RNA is

59 detectable in adipocytes in COVID-19 autopsy cases and is associated with an inflammatory

68 symptoms, to severe pulmonary and extrapulmonary manifestations characterized by extreme

76 metabolic milieu in obesity, particularly among individuals with insulin resistance, is

78 a more robust inflammatory response to SARS-CoV-2. Impaired immune responses to viral

80 baseline (13) and in response to influenza infection (7, 14).

86 (adenovirus 36, human cytomegalovirus), macrophages (simian immunodeficiency virus (SIV)),

87 and T cells (human immunodeficiency virus (HIV))(16–19). Complex interactions between

89 derived chemoattractants such as monocyte chemoattractant protein-1 (MCP-1) leading to

95 via inflammation and downstream processes such as extracellular matrix deposition/fibrosis,

106 inflammatory profile.

109 Establishment of a system to examine SARS-CoV-2 infection in human adipose tissue

111 inflammatory response, we recruited participants undergoing bariatric or cardiothoracic surgery.

122 (mock) with SARS-CoV-2 (USA-WA1/2020) at a multiplicity of infection (MOI) of 1. (B)

135 Human adipose tissue SVC supports SARS-CoV-2 infection

149 support SARS-CoV-2 infection.

154 N protein was primarily restricted to macrophages (defined as large granular

155 CD45+CD14+CD11c- cells) (Fig.1D). We next assessed the level of angiotensin-converting

165 Figure 2. A subset of macrophages is infected with SARS-CoV-2. (A) Schematic of

188 DEGs between SARS-CoV-2+ versus bystander C2 macrophages.

190 SARS-CoV-2 infects a distinct subset of macrophages

192 we performed single-cell RNA sequencing (scRNA-seq) of SARS-CoV-2 and mock-infected

200 labeled two distinct macrophage clusters.

213 entry receptors could contribute to SARS-CoV-2 entry into C2-macrophages.

232 derived from metabolic activation (30, 31, 37, 39).

234 SARS-CoV-2 infection drives an inflammatory response in macrophages

241 adipose tissue and is more susceptible to infection in vitro.

261 display a dramatic transcriptional response with upregulation of inflammatory pathways.

275 Adipocyte progenitors demonstrate an inflammatory response to SARS-CoV-2 infection of

291 VAT-predominant clusters demonstrated higher levels of IFI27 (a marker of SARS-CoV-2

293 preadipocytes in mock-conditions, highlighting the highly inflammatory nature of preadipocytes

305 2 exposure (Fig. 4D, fig. S9A).

308 showed a stronger response to SARS-CoV-2 exposure. We employed perturbation analysis

331 neighboring preadipocyte cells.

SAT (n=8) VAT (n=6)

Fold value Fold value

ill patients; associated with disease

severity and ICU admission;

predictor of disease progression;

biomarker of impaired T cell

response in acute infection

associated with disease severity and 56)

predictor of disease progression

CXCL13 1.19 0.02/0.1 0.96 0.63/0.98 Elevated in serum (58)

8 Matthias S. Matter8, Catherine A. Blish1,2,6†, Tracey L. McLaughlin1†