SARS-CoV-2 Receptor ACE2 Is an Interferon-Stimulated Gene in Human Airway Epithelial Cells and Is Detected in Specific Cell Subsets across Tissues

Abstract

There is pressing urgency to understand the pathogenesis of the severe acute respiratory syndrome coronavirus clade 2 (SARS-CoV-2), which causes the disease COVID-19. SARS-CoV-2 spike (S) protein binds angiotensin-converting enzyme 2 (ACE2), and in concert with host proteases, principally transmembrane serine protease 2 (TMPRSS2), promotes cellular entry. The cell subsets targeted by SARS-CoV-2 in host tissues and the factors that regulate ACE2 expression remain unknown. Here, we leverage human, non-human primate, and mouse single-cell RNA-sequencing (scRNA-seq) datasets across health and disease to uncover putative targets of SARS-CoV-2 among tissue-resident cell subsets. We identify ACE2 and TMPRSS2 co-expressing cells within lung type II pneumocytes, ileal absorptive enterocytes, and nasal goblet secretory cells. Strikingly, we discovered that ACE2 is a human interferon-stimulated gene (ISG) in vitro using airway epithelial cells and extend our findings to in vivo viral infections. Our data suggest that SARS-CoV-2 could exploit species-specific interferon-driven upregulation of ACE2, a tissue-protective mediator during lung injury, to enhance infection.

Keywords: ACE2; COVID-19; ISG; SARS-CoV-2; human; influenza; interferon; mouse; non-human primate; scRNA-seq.

Conflict of interest statement

Declaration of Interests A.R. is an SAB member of ThermoFisher Scientific, Neogene Therapeutics, Asimov, and Syros Pharmaceuticals; a co-founder of and equity holder in Celsius Therapeutics; and an equity holder in Immunitas Therapeutics. A.K.S. reports compensation for consulting and/or SAB membership from Merck, Honeycomb Biotechnologies, Cellarity, Cogen Therapeutics, Orche Bio, and Dahlia Biosciences. L.S.K. is on the SAB for HiFiBio; she reports research funding from Kymab Limited, Bristol Meyers Squibb, Magenta Therapeutics, BlueBird Bio, and Regeneron Pharmaceuticals and consulting fees from Equillium, FortySeven, Inc, Novartis, Inc, EMD Serono, Gilead Sciences, and Takeda Pharmaceuticals. A.S. is an employee of Johnson and Johnson. N.K. is an inventor on a patent using thyroid hormone mimetics in acute lung injury that is now being considered for intervention in COVID-19 patients. J.L. is a scientific consultant for 10X Genomics, Inc. O.R.R, is a co-inventor on patent applications filed by the Broad Institute to inventions relating to single-cell genomics applications, such as in PCT/US2018/060860 and US Provisional Application No. 62/745,259. S.T. in the last three years was a consultant at Genentech, Biogen, and Roche and is a member of the SAB of Foresite Labs. M.H.W. is now an employee of Pfizer. F.J.T. reports receiving consulting fees from Roche Diagnostics GmbH and ownership interest in Cellarity, Inc. P.H. is a co-inventor on a patent using artificial intelligence and high-resolution microscopy for COVID-19 infection testing based on serology.

Copyright © 2020 The Authors. Published by Elsevier Inc. All rights reserved.

Figures

Graphical abstract

Figure 1

Expression of ACE2 in Type II Pneumocytes in Healthy Lungs of Non-human Primates (A) Schematic of protocol for isolation of lung tissue at necropsy from healthy non-human primates (M. mulatta, n = 3), creation of scRNA-seq libraries by using Seq-Well v1, and computational analysis to identify cell types by using unbiased methods. UMAP projection of 3,793 single cells, points colored by cell identity (see STAR Methods). (B) Uniform manifold approximation and projection (UMAP) as in (A), points colored by detection of ACE2 (coronavirus receptor, top) or TMPRSS2 (coronavirus S protein priming for entry, bottom). Color coding is as follows: black, RNA positive; blue, RNA negative. (C) Dot plot of 2 defining genes for each cell type (Table S1) (Bonferroni-adjusted p ACE2 and TMPRSS2. Dot size represents fraction of cells within that type expressing a given gene, and color intensity represents binned count-based expression amount (log(scaled UMI+1)) among expressing cells. ACE2 is enriched in type II pneumocytes (6.7% expressing, Bonferroni-adjusted p = 8.62E−33), as is TMPRSS2 (29.5% expressing, Bonferroni-adjusted p = 8.73E−153). Of all type II pneumocytes, 3.8% co-express ACE2 and TMPRSS2 (Table S9). Red arrow indicates cell type with largest proportion of ACE2+TMPRSS2+ cells. (D) Genes differentially expressed among ACE2+ and ACE2− type II pneumocytes. (SCDE package, FDR-adjusted p < 0.05 for IFNGR2, NT5DC1, ARL6IP1, and TRIM27; full results can be found in Table S1). See also Table S1.

Figure 2

Select Lung Epithelial Cells from Control, HIV-1-Infected, and Mycobacterium-tuberculosis-Infected Human Donors Co-Express ACE2 and TMPRSS2 (A) Schematic of protocol for isolation of human lung tissue from surgical excess, creation of scRNA-seq libraries by using Seq-Well S3, and computational analysis to identify cell types by using unbiased methods. Shown on the right is a UMAP projection of 18,915 cells across 8 donors (n = 3 TB+HIV+; n = 3 TB+; n = 2 non-infected patients). Cells represented by points, colored according to cell type (see STAR Methods). (B) UMAP projection as in (A), points colored by detection of ACE2 (top) or TMPRSS2 (bottom). Color coding is as follows: black, RNA positive; blue, RNA negative. (C) Dot plot of 2 defining genes for each cell type (FDR-adjusted p ACE2 and TMPRSS2; dot size represents fraction of cells within cell type expressing a given gene, and color intensity represents binned count-based expression amount (log(scaled UMI+1)) among expressing cells. All cluster-defining genes are provided in Table S2. Red arrow indicates cell types with largest proportion of ACE2+TMPRSS2+ cells. (D) Volcano plot identifying significantly upregulated genes in ACE2+TMPRSS2+ pneumocytes compared with all remaining pneumocytes. Red points represent genes with a FDR-adjusted p < 0.05, and log2(fold change) >1.5. Text highlighting specific genes; the full list is available in Table S2. (E) Expression of ACE2 across human donors by HIV and TB status (p = 0.009 by likelihood-ratio test). See also Table S2.

Figure S1

NHP Tuberculosis Infected Lung and Granuloma, Related to Figures 1 and 2 (A). UMAP projection of epithelial cells (1,099 cells) colored by annotated cell type, tissue source, and gating as ACE2+TMPRSS2+ cells. ACE2+TMPRSS2+ cells comprise 11% of ciliated cells, 16% of club cells, 10% type I pneumocytes, and 22% type II pneumocytes. Data generated using Seq-Well S3 (Table S3). (B). Number of cells (left) and % (right) ACE2+TMPRSS2+ cells by tissue source (granuloma versus uninvolved lung) and cell type. Ciliated cells and club cells were omitted from this analysis as we detected too few cells (< 7 total cells) belonging to these clusters in the granulomas. Statistical significance assessed by Fisher Exact Test (Table S3). (C). Dot plot of top cluster defining genes for each epithelial cell type and ACE2 and TMPRSS2. Dot size represents fraction of cells expressing, and color intensity represents average log(normalized UMI + 1) among all cells in each group scaled between 0 and 1 by gene. ACE2 expression is enriched in club cells (Bimodal test, Bonferroni-corrected p < 0.001), ciliated cells (p < 0.005), and type I pneumocytes (p < 0.001). TMPRSS2 expression is enriched in type I pneumocytes (p < 0.001) and ciliated cells (p < 0.001) (Table S3). (D). Dot plot of genes differentially expressed between ACE2+TMPRSS2+ epithelial cells versus rest (Bimodal test, Bonferroni-corrected p < 0.01, log fold change > 0.5). (Table S3, c = number of cells, n = number of animals).

Figure 3

NHP and Human Ileal Absorptive Enterocytes Co-Express ACE2 and TMPRSS2 (A) Expression ACE2 across diverse tissues in healthy NHPs (n = 3 animals; 52,858 cells). (B) Schematic of protocol for isolation of NHP ileum (n = 5) at necropsy for scRNA-seq using Seq-Well v1, and computational pipeline to identify cell types by using unbiased methods. Shown on the right is a UMAP projection of 4,515 cells colored by cell type. (C) Dot plot of 2 defining genes for each cell type, with ACE2 and TMPRSS2. Dot size represents fraction of cells within cell type expressing a given gene, and color intensity represents binned count-based expression amounts (log(scaled UMI+1)) among expressing cells. All cluster defining genes are provided in Table S4. Red arrow indicates cell type with largest proportion of ACE2+TMPRSS2+ cells. (D) Schematic of protocol for isolation of human ileal cells from endoscopic pinch biopsies in non-inflamed regions (n = 13). Shown on the right is a tSNE plot of 13,689 epithelial cells selected from original dataset generated by 10x 3′ v2 (see Figure S2), colored by cellular subsets. (E). Dot plot of 2 defining genes for each cell type, with ACE2 and TMPRSS2. Dot size represents fraction of cells within cell type expressing a given gene, and color intensity represents binned count-based expression amounts (log(scaled UMI+1)) among expressing cells. All cluster defining genes are provided in Table S5. Red arrow indicates cell type with largest proportion of ACE2+TMPRSS2+ cells. (F). Expression of ACE2 (left) and TMPRSS2 (right) among all epithelial subsets from human donors. See also Figure S2 and Tables S4 and S5.

Figure S2

Human and NHP Ileum, Related to Figure 3 (A). Top: tSNE projection of all cells from healthy pediatric human ileum within a previously-unpublished 10x 3′ v2 dataset (115,569 cells). Black: higher expression of ACE2 (left), TMPRSS2 (right). Bottom: Corresponding violin plots of expression values for ACE2 (left) and TMPRSS2 (right). Solid line: epithelial cells. (B). Co-expression of ACE2 and TMPRSS2 by epithelial cell subset. Number indicates % of ACE2+TMPRSS2+ cells by cell subset. (C). tSNE projection of 13,689 cells as in Figure 3D, cells colored by co-expression of ACE2 and TMPRSS2 (black). (D). Expression of ACE2 and canonical interferon-responsive genes among absorptive enterocytes from Healthy (n = 2) and SHIV-infected, anti-retroviral treated animals (n = 3). Bonferroni-adjusted p-values by Wilcoxon test (healthy: 510 cells, SHIV-infected: 636 cells).

Figure 4

Healthy and Allergic Inflamed Human Nasal Mucosa Co-Express ACE2 and TMPRSS2 in a Subset of Goblet Secretory Cells (A) Schematic for sampling of n = 12 ethmoid sinus surgical samples and n = 9 inferior turbinate nasal scrapings to generate scRNA-seq libraries by using Seq-Well v1. See Ordovas-Montanes et al., (2018). (B) Dot plot of all cell types from ethmoid-sinus-derived cells (n = 6 non-polyp CRS samples, n = 6 polyp CRS samples). Two defining genes for each cell type, in addition to CDHR3 (rhinovirus receptor), ACE2, TMPRSS2, and JAK1. Dot size represents fraction of cells within that type expressing a given gene, and color intensity represents binned count-based expression amounts (log(scaled UMI+1)) among expressing cells (see Table S6 for statistics by subset). Red arrow indicates cell types with largest proportion of ACE2+TMPRSS2+ cells. (C) Dot plot for 2 defining genes for each cell type identified from granular clustering of epithelial cells (18,325 single cells) derived from both ethmoid sinus and inferior turbinate sampling (healthy inferior turbinate [3,681 cells; n = 3 samples], polyp-bearing patient inferior turbinate [1,370 cells; n = 4 samples], non-polyp ethmoid sinus surgical samples [5,928 cells; n = 6 samples], and polyp surgical and scraping samples directly from polyp in ethmoid sinus [7,346 cells; n = 8 samples]). Red arrow indicates cell type with largest proportion of ACE2+TMPRSS2+ cells. (D) tSNE of 18,325 single epithelial cells from inferior turbinate and ethmoid sinus (omitting immune cells). Colored by cell types 3,152 basal, 3,089 differentiating, 8,840 secretory, 1,105 ciliated, and 2,139 glandular cells. (E) tSNE as in (D), identifying epithelial cells co-expressing ACE2 and TMPRSS2 (30 cells, black points). (F) tSNE as in (D), colored by detailed cell types with higher granularity, as in (C). (G) Individual differentially expressed genes between ACE2+TMPRSS2+ cells and all other secretory epithelial cells (see Table S6 for full gene list with statistics). Bonferroni-adjusted likelihood-ratio test p < 0.02 for all genes displayed. (H) Stacked bar plot of each subset of epithelial cells among all epithelial cells by donor (each bar) and sampling location (noted below graph) (unpaired t test p ACE2+TMPRSS2+ cells. See also Figure S3 and Table S6.

Figure S3

Nasal and Sinus Mucosa, Related to Figures 4 and 5 (A). Expression of ACE2 and TMPRSS2 across donors. (B). Enhanced capture of ACE2 mRNA with second strand synthesis protocol employed in Seq-Well S3. Dot size represents fraction of cells expressing. (C). Cultured human primary basal epithelial cells at confluence were treated with increasing doses (0.1 to 10ng/mL) of IFNα2, IFNγ, IL-4, IL-13, IL-17A, and IL-1B for 12 h and bulk RNA-seq analysis was performed (Replicate experiment using Human Donor 1 as in Figure 5) (D). ACE2 expression by stimulation condition. Wilcoxon test between each cytokine (combined doses) versus rest: IFNα Bonferroni-adjusted p = 4.1E-07; IFNγ Bonferroni-adjusted p = 9.3E-03; all else n.s. ∗∗∗ p < 0.001. (E). ACE2 expression by IFNα2 dose. Bonferroni-corrected t-test compared to 0 ng/mL condition: ∗∗∗ p < 0.001, ∗ p < 0.05. (F). ACE2 expression by IFNγ dose. Bonferroni-corrected t-test compared to 0 ng/mL condition: ∗∗∗ p < 0.001, ∗ p < 0.05. (G). IFITM1 expression by IFNα2 dose. Bonferroni-corrected t-test compared to 0 ng/mL condition: ∗∗∗ p < 0.001. (H). IFITM1 expression by IFNγ dose. Bonferroni-corrected t-test compared to 0 ng/mL condition: ∗∗∗ p < 0.001. (I). GBP5 expression among cultured human primary basal epithelial cells. Wilcoxon test: IFNα versus IFNγ Bonferroni-adjusted p = 2.94E-07; IFNγ Bonferroni-adjusted p = 9.3E-03. TP10K: transcripts per 10,000 reads. ∗∗∗ p < 0.001. (J). GBP5 expression by IFNα2 dose. Bonferroni-corrected t-test compared to 0 ng/mL condition: ∗∗∗ p < 0.001. (K). GBP5 expression by IFNγ dose. Bonferroni-corrected t-test compared to 0 ng/mL condition: ∗∗∗ p < 0.001.

Figure 5

ACE2 is an Interferon-Stimulated Gene in Primary Human Barrier Tissue Epithelial Cells (A–D) Basal epithelial cells from distinct sources were cultured to confluence and treated with increasing doses (0.1–10 ng/mL) of IFN-α2, IFN-γ, IL-4, IL-17A, and/or IFN-β for 12 h and bulk RNA-seq analysis was performed. Expression of ACE2 (human) or Ace2 (mouse) by cell type and stimulation condition. (A) Primary mouse basal cells from tracheal epithelium are shown. (B) BEAS-2B human bronchial cell line is shown. (C) Primary human basal cells from nasal scraping, Donor 1, is shown. (D) Primary human basal cells from nasal scraping, Donor 2. Abbreviation is as follows: TP10K, transcripts per 10,000 reads. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, Bonferroni-corrected t test compared with untreated condition. (E–H) Co-expression of STAT1/Stat1 and ACE2/Ace2 by cell type. (E) Primary mouse basal cells from tracheal epithelium are shown. (F) BEAS-2B human bronchial cell line is shown. (G) Primary human basal cells from nasal scraping, Donor 1, are shown. (H) Primary human basal cells from nasal scraping, Donor 2 are shown. Abbreviation is as follows: TP10K, transcripts per 10,000 reads. Statistical significance assessed by Spearman’s rank correlation. (I–L) Expression of ACE2 in primary human basal cells from nasal scrapings across a range of concentrations of IFN-γ or IFN-α2. (I) IFN-α2 dose response in Donor 1 (p < 0.001 by one-way ANOVA) is shown. (J) IFN-γ dose response in Donor 1 (p < 0.01 by one-way ANOVA) is shown. (K) IFN-α2 dose response in Donor 2 (p < 0.001 by one-way ANOVA) is shown. (L) IFN-γ dose response in Donor 2 (p < 0.001 by one-way ANOVA). Abbreviation is as follows: TP10K, transcripts per 10,000 reads. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, Bonferroni-corrected post hoc testing compared with 0 ng/mL condition. See also Figures S3 and S4 and Table S7.

Figure S4

Published Studies of Epithelial Cells Following Interferon Treatment Related to Figure 5 (A). Fold change of ACE2 expression among human or mouse datasets following Type I or Type II interferon treatment compared to untreated control. Generated from publicly available microarray data curated at interferome.org. Includes all studies with abs(fold-change) > 1. (B). Location of transcription factors binding regions spanning −1500 bp to +500 bp from the transcription start site of ACE2 (human, top) or Ace2 (mouse, bottom). Generated from TRANSFAC data using the interferome.org database (Matys et al., 2003, Rusinova et al., 2013).

Figure S5

Mouse Nasal Epithelium Following Interferon-α Exposure Related to Figure 6 (A). Schematic: mice were exposed to 10,000 units of IFN-α or saline by intranasal application (n = 2 per group). After 12 h, animals were sacrificed and nasal epithelium was dissected and dissociated for scRNA-seq using Seq-Well S3. (B). Dot plot of 2 defining genes for each cell type, with Ace2, Tmprss2, and Cdhr3. Dot size represents fraction of cells within cell type expressing, and color intensity binned count-based expression level (log(scaled UMI+1)) among expressing cells. All cluster defining genes are provided in Table S8. Red arrows: cell types with largest proportion of Ace2+ cells. Dendrogram (left) by person correlation over differentially expressed genes with Ward clustering. (C). UMAP of Basal Epithelial Cells (380 cells) across 4 mice. Black: Saline-treated mouse; red: IFN-α treated. (D). UMAP of Basal Epithelial Cells as in C, points colored by detection of Ace2. Black: RNA positive, blue: RNA negative (6.6% Ace2+, Bonferroni-adjusted p = 1.1E-10 for Basal Epithelial Cell expression versus all other cells). (E). Schematic: wildtype (WT) and IFNγ-receptor knockout (IFNγR−/−) mice were infected intranasally with murine gamma-herpesvirus-68 (MHV68). Cells from whole lung were digested for scRNA-seq using Drop-seq (yielding 5,558 Epcam+ cells). (F). Expression of Ace2 by epithelial cell type, wild type (WT) mice. Statistical significance by Wilcoxon rank sum test with Bonferroni correction. (G). Expression of Ace2 among type II pneumocytes binned by infection status in WT mice. All pairwise comparisons non-significant (p > 0.05) by Wilcoxon rank sum test. (H). Percent of Ace2+ cells by infection condition (uninfected, bystander cells in MHV68-infected mouse, MHV68 RNA+ cells) and mouse genotype (WT, IFNγR −/−). Black bars: Ace2+ positive cells; white bars: Ace2- cells. (I). Schematic of RNA-Seq data from (Matos et al., 2019) of human lung explants (n = 5 donors) exposed to influenza A virus (IAV, H3N2) at 24 h post infection. (J). Expression of SFTPC (surfactant protein C, a marker of type II pneumocytes) versus ACE2 among mock-infected lung explants. Statistical significance assessed by Pearson’s correlation, r = 0.93, p = 0.021. TPM: transcripts per million. (K). SFTPC expression among matched donors following mock or IAV infection for 24 h. Statistical significance assessed by ratio paired t test, p = 0.86. (L). ACE2 expression among matched donors following mock or IAV infection for 24 h. Statistical significance assessed by ratio paired t test, p = 0.0054. (M). Western blot of fully-differentiated air-liquid interface cultures from bronchial cells derived from 4 human donors with asthma. Cells from each donor were treated with 10 ng/mL IFNγ for 24 h, and compared to a matched untreated condition. ACE2 protein: AF933 (R&D). Fold changes quantified for IFNγ treated versus untreated for each patient donor following normalization to GAPDH.

Figure 6

In Vivo Administration of Interferons in Mice Does Not Induce Ace2, and ACE2 Is Induced in Goblet Secretory Cells during Human Influenza Infection (A) UMAP of 11,358 single cells from mouse nasal epithelium (n = 4). (B) UMAP projection as in (A), points colored by detection of Ace2 (SARS-CoV-2 receptor homolog). Color coding is as follows: black, RNA positive; blue, RNA negative. (C) Percent of Ace2+ cells by treatment condition (n = 4 arrays per condition; n = 2 arrays per mouse). Black bars indicate Ace2+ cells; white bars indicate Ace2− cells. p = 0.4 by Student’s t test. (D) Heatmap of cell-type-defining genes (Trp63 and Krt17), interferon-induced genes (Irf7, Stat1, Irf9, and Oasl2), and Ace2 among basal epithelial cells, separated by cells derived from saline-treated mice (left) and IFN-α-treated mice (right). Statistical significance by likelihood-ratio test with Bonferroni correction is shown. A full list of differentially expressed genes can be found in Table S8. (E) Schematic for sampling cells derived from nasal washes of n = 18 human donors with and without current influenza A or B infection for Seq-Well v1 (35,840 single cells). See Cao et al., (2020). (F and G) ACE2 expression among goblet cells (F) and squamous cells (G) by infection status. Shown are Healthy Donor cells from influenza-negative donors (white); Bystander Cells from influenza A (IAV)- or influenza B (IBV)-infected donors, no intracellular viral RNA detected (black); Flu Viral RNA+ Cells with detectable intracellular influenza A or B viral RNA (red). Statistical significance by Wilcoxon test with Bonferroni correction, n.s. for Bystander versus Flu Viral RNA+. See also Figure S5 and Tables S6 and S8.

Figure S6

Power Calculations and Statistical Modeling of ACE2 Capture and Dropout Related to STAR Methods (A). Probability of capturing and transcribing at least 1 ACE2 cDNA molecule, as a function of the capture/reverse transcription efficiency for a single molecule and the number of ACE2 molecules expressed in an individual cell. Note that Drop-Seq provides a capture/transcription efficiency of approximately 11-13%, setting a floor on this parameter, and the experimental platforms used in this study are either equivalent or superior (Macosko et al., 2015). (B). Distribution of ACE2 fractional abundance within individual cells’ cDNA libraries (i.e., ACE2 UMIs / total number of reads), across non-human primate lung and ileum cell populations (see Figures 1 and 3). Mean fractional abundance among ACE2+ lung cells = 5.0E-5; mean fractional abundance among ACE2+ ileum cells = 2.7E-4. (C). Distribution of the number of reads within non-human primate lung and ileum cell populations (see Figures 1 and 3). Mean ± SEM reads among all lung cells = 28,512 ± 344; ACE2+ lung cells = 28,553 ± 2,988; all ileum cells = 14,864 ± 288; ACE2+ ileum cells = 10,591 ± 441. (D). Probability of observing at least one transcript for a gene of interest (e.g., ACE2) within an individual cell, as a function of sequencing depth and the gene’s fractional abundance (i.e., ACE2 reads / all reads) within the cell’s cDNA library. Fractional abundance provides the probability that a single read corresponds to the gene of interest, and presented heatmap indicates the probability that at least one read in the total number of reads allocated to the cell (i.e., from 103 to 106) originates from the gene of interest. Mean read depths and ACE2 fractional abundances for each tissue produce a 93.7% probability of detecting at least 1 ACE2 read in ileum cells, and a 76.0% chance for lung cells. Outlined rectangles highlight the regimes where cells from lung (turquoise) and ileum (pink) samples typically lie. (E). Number of ACE2+ cells within each cluster, as a function of average read depth for all cells in that cluster. Number of cells detected as ACE2+ is not correlated with read depth, even across relatively wide ranges of average read depths (Pearson’s r = −0.31, n.s.). (F). Probability of observing a particular number of cells positive for a gene of interest within a cluster, as a function of number of cells in the cluster. Probabilities were calculated under a negative binomial distribution with parameter p = 0.063 (the proportion of ACE2+ cells among type II pneumocytes presented in Figure 1; STAR Methods). The horizontal gray line indicates the arbitrary cut-off value of p = 0.05. (G). Given a population of cells with a known proportion that are positive for a gene of interest, probability of observing no positive cells (i.e., false negative identification of the cluster; solid lines) and probability of observing at least one positive cell as a function of cluster size.