ACE2: Entry Receptor for SARS-CoV-2

In order to better understand the complex relationships between host and virus genetic dependencies, the team used a broad range of analytical and experimental methods to validate their results. This integrative approach included genome editing, single-cell sequencing, confocal imaging and computational analyses of gene expression and proteomic datasets.

After intensive research, the scientists and doctors claim they have found 30 genes that block the virus from infecting human cells including RAB7A, a gene that seems to regulate the ACE-2 receptor that the virus binds to and uses to enter the cell. The spike protein’s first contact with a human cell is through ACE-2 receptor.

“Our findings confirmed what scientists believe to be true about ACE-2 receptor’s role in infection; it holds the key to unlocking the virus,” said Dr. tenOever. “It also revealed the virus needs a toolbox of components to infect human cells. Everything must be in alignment for the virus to enter human cells.”



The team discovered that the top-ranked genes — those whose loss reduces viral infection substantially — clustered into a handful of protein complexes, including vacuolar ATPases, Retromer, Commander, Arp2/3, and PI3K. Many of these protein complexes are involved in trafficking proteins to and from the cell membrane.

“We were very pleased to see multiple genes within the same family as top-ranked hits in our genome-wide screen. This gave us a high degree of confidence that these protein families were crucial to the virus lifecycle, either for getting into human cells or successful viral replication,” said Dr. Zharko Daniloski, a postdoctoral fellow in the Sanjana Lab and co-first author of the study.

Using proteomic data, they found that several of the top-ranked host genes directly interact with the virus’s own proteins, highlighting their central role in the viral lifecycle. The team also analyzed common host genes required for other viral pathogens, such as Zika or H1N1 pandemic influenza.



Cholesterol and the virus
The research team also identified drugs that are currently on the market for different diseases that they claim block the entry of Covid-19 into human cells by increasing cellular cholesterol. In particular, they found three drugs currently on the market were more than 100-fold more effective in stopping viral entry in human lung cells:

Amlodipine, brand name Norvasc, by Pfizer, to treat high blood pressure and angina.
Tamoxifen, brand name Soltamox by Fortovia Therapeutics, an estrogen modulator, to treat breast cancer.
Ilomastat, brand name Galardin, it’s a matrix metalloprotease inhibitor, that now being manufactured by many companies; a chemotherapy agent, with applications for skincare and anti-aging products.
The other five drugs that were tested — called PIK-111, Compound 19, SAR 405, Autophinib, ALLN -- are used in research but are not yet branded and used in clinical trials for existing diseases.


Their findings offer insight into novel therapies that may be effective in treating Covid-19 and reveal the underlying molecular targets of those therapies.

The bioengineers in New York were working on other projects with gene-editing technology from CRISPR but quickly pivoted to studying the coronavirus when it swept through the metropolitan area last March. “Seeing the tragic impact of Covid-19 here in New York and across the world, we felt that we could use the high-throughput CRISPR gene editing tools that we have applied to other diseases to understand what are the key human genes required by the SARS-CoV-2 virus,” said Dr. Sanjana.

Dr. Neville Sanjana and his team at the New York Genome Center used CRISPR to identify the genes that can protect human cells against Covid-19.

As he explained, “current treatments for SARS-CoV-2 infection currently go after the virus itself, but this study offers a better understanding of how host genes influence viral entry and will enable new avenues for therapeutic discovery.”

Previously, Dr. Sanjana has applied genome-wide CRISPR screens to identify the genetic drivers of diverse diseases, including drug resistance in melanoma, immunotherapy failure, lung cancer metastasis, innate immunity, inborn metabolic disorders and muscular dystrophy.

“The hope is that the data from this study— which pinpoints required genes for SARS-CoV-2 infection — could in the future work be combined with human genome sequencing data to identify individuals that might be either more susceptible or more resistant to Covid-19,” Dr. Sanjana said.

The New York team is not the first to use CRISPR gene editing techniques to fight Covid-19. Other bioengineering groups at MIT and Stanford have been using CRISPR to develop ways to fight the SARS-CoV-2 and develop diagnostic tools for Covid-19.

Results
A High-Throughput Screen to Identify Genes Required for SARS-CoV-2 Infection
To identify key genes required for SARS-CoV-2 infection, we performed a genome-scale loss-of-function screen targeting 19,050 genes in the human genome using the GeCKOv2 CRISPR-Cas9 library (Sanjana et al., 2014). The GeCKOv2 library contains 122,411 CRISPR single-guide RNAs (sgRNAs) (6 guide RNAs per gene) and has previously been used in CRISPR screens for drug resistance, immunotherapy, synthetic lethality, mitochondrial disease, and therapeutic discovery for muscular dystrophy (Erb et al., 2017; Jain et al., 2016; Lek et al., 2020; Patel et al., 2017; Shalem et al., 2014). First, we transduced a human alveolar basal epithelial carcinoma cell line (A549) that constitutively expresses ACE2 (referred to as A549ACE2) with an all-in-one lentiviral vector containing Cas9, guide RNAs from the GeCKOv2 human library, and a puromycin resistance gene. The transduction was performed at a low multiplicity of infection (MOI ∼0.2) to ensure that most cells would receive only one guide RNA construct (Figure 1A). We then selected with puromycin so that only library-transduced cells remained. We also measured the survival rate after puromycin selection was complete (3 days) to ensure high coverage of the 122,411 guide RNAs (∼1,000 cells per guide RNA). After puromycin selection was complete, we cultured the cells for 9 days to ensure protein depletion after CRISPR gene targeting.




Figure 1 A Genome-Scale CRISPR Loss-of-Function Screen to Identify Genes that Prevent SARS-CoV-2 Infection of Human Alveolar Epithelial Carcinoma Cells

Next, we infected the GeCKOv2 pool of A549ACE2 cells with SARS-CoV-2 virus (Isolate USA-WA1/2020 NR-52281) at either a high (0.3) or a low (0.01) MOI. We verified that SARS-CoV-2 infects A549ACE2 cells by staining for the nucleocapsid (N) protein at 24 h post-infection (Figure 1B), and at day 6 post-infection, we measured cell survival for both the high and low MOI conditions (Figure 1C). As expected, the higher MOI infection resulted in fewer surviving cells at day 6 post-infection. Next, we extracted genomic DNA, and via amplicon sequencing, we quantified guide abundance in each biological condition (Figure 1A). To confirm that library representation was properly maintained, we computed the correlation between the guide representation in the plasmid library and after puromycin selection (r = 0.84) (Figure S1A). In contrast, after SARS-CoV-2 infection, there was a much greater degree of guide dropout, as expected given that SARS-CoV-2 rapidly kills A549ACE2 cells without CRISPR perturbations (Figures 1C and 1D).



Figure S1 Genome-wide Loss-of-Function CRISPR Screen Enriched Gene Identification, Related to Figure 1
Show full caption

Using robust-rank aggregation (RRA) on the guide relative enrichments, we computed gene-level scores to identify genes where loss-of-function mutations led to enrichment within the pool (Figure 1E; Kolde et al., 2012). We identified ∼1,000 genes with significant RRA enrichment (p < 0.05) (Figure S1B). We also used two other previously published methods to compute gene enrichments (RIGER weighted-sum and second-best rank) and found a high degree of overlap between enriched genes identified by all three methods (Figure S1C; Chen et al., 2015; Luo et al., 2008). We also found a high degree of shared genes across both the low and high SARS-CoV-2 MOI conditions: when examining the top 50 most enriched genes, we found that 27 of them were shared between the low and high MOI conditions (Figure 1F; Table S1), suggesting that several host genes involved in viral pathogenesis function independently of viral dose. An independent genome-scale CRISPR screen for SARS-CoV-2 infection also performed in A549 that overexpress ACE2 but with a different CRISPR library identified similar top-ranked genes (Zhu et al., 2020), highlighting the robustness of our results (Figure S1D).

Enriched Genes Are Involved in Multiple Aspects of the Viral Life Cycle and Are Broadly Expressed
Upon close examination of the most enriched genes, we found genes involved in key aspects of viral entry and replication (Figure 2; Du et al., 2009). For example, the well-established entry receptor angiotensin-converting enzyme 2 (ACE2) receptor was ranked as the 8th most-enriched gene in the low MOI screen and 12th in the high MOI screen (Table S1; Hoffmann et al., 2020a; Zhou et al., 2020). Among the top 50 enriched genes, we identified several sets of related genes that function together in complexes, giving us further confidence in the genome-scale screen (Figures 2 and 3A ). We found genes essential for initial attachment and endocytosis (ACE2, RAB7A, and 4 members of the ARP2/3 complex: ACTR2, ACTR3, ARPC3, and ARPC4), Spike protein cleavage and viral membrane fusion (CTSL and 13 members of the vacuolar-ATPase proton pump: ATP6AP1, ATP6AP2, ATP6V0B ATP6V0C, ATP6V0D1, ATP6V1A, ATP6V1B2, ATP6V1C1, ATP6V1E1, ATP6V1G1, ATP6V1H, TMEM199, and TOR1AIP1), endosome recycling (4 members of the endosomal protein sorting Retromer complex: VPS26A, VPS29, VPS35, and SNX27; 4 members of the endosomal trafficking Commander complex: COMMD2, COMMD3, COMMD3-BMI1, and COMMD4; and 3 members of the PI3K pathway: PIK3C3/VPS34, WDR81, and ACP5), ER-Golgi trafficking (DPM3, ERMP1, PPID, and CHST14), and transcriptional modulators (SLTM and SPEN). A consistent theme among the enriched complexes is endosome function and regulation (V-ATPase proton pump, Retromer, Commander, class 3 PI3Ks) (Banerjee and Kane, 2020; Mallam and Marcotte, 2017; McNally and Cullen, 2018). Gene set enrichment analysis on the full ranked list of genes identified significantly enriched Gene Ontology (GO) categories for endosome processing, transport, and acidification and categories related to cytokinesis and virion attachment (false discovery rate [FDR] q < 0.1) (Figures 3B and S2A–S2D ; Table S2; Subramanian et al., 2005).