Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Paralog knockout profiling identifies DUSP4 and DUSP6 as a digenic dependence in MAPK pathway-driven cancers

Abstract

Although single-gene perturbation screens have revealed a number of new targets, vulnerabilities specific to frequently altered drivers have not been uncovered. An important question is whether the compensatory relationship between functionally redundant genes masks potential therapeutic targets in single-gene perturbation studies. To identify digenic dependencies, we developed a CRISPR paralog targeting library to investigate the viability effects of disrupting 3,284 genes, 5,065 paralog pairs and 815 paralog families. We identified that dual inactivation of DUSP4 and DUSP6 selectively impairs growth in NRAS and BRAF mutant cells through the hyperactivation of MAPK signaling. Furthermore, cells resistant to MAPK pathway therapeutics become cross-sensitized to DUSP4 and DUSP6 perturbations such that the mechanisms of resistance to the inhibitors reinforce this mechanism of vulnerability. Together, multigene perturbation technologies unveil previously unrecognized digenic vulnerabilities that may be leveraged as new therapeutic targets in cancer.

This is a preview of subscription content, access via your institution

Access options

Rent or buy this article

Prices vary by article type

from$1.95

to$39.95

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Development and performance of a Digenic Paralog CRISPR library.
Fig. 2: Synergistic paralog dependencies and small-molecule inhibitor profiles.
Fig. 3: DUSP4/6 are synergistic dependencies in NRAS mutant cells through the hyperactivation of ERK.
Fig. 4: ERK2 DRS is required for DUSP4/6 binding.
Fig. 5: BRAF mutant melanomas resistant to MAPKi are hypersensitized to dual DUSP4/6 knockout.

Similar content being viewed by others

Data availability

The read counts for all screening data and subsequent analyses are provided as Supplementary Data and are available at the Sequence Read Archive under accession no. PRJNA745952. Paralog identification was obtained from ENSEMBL release 91. All genomic data from the CCLE are available at https://portals.broadinstitute.org/ccle/data. DepMap 20Q1 was used for all analyses except for Fig. 5c,d and Extended Data Fig. 6c where DepMap 21Q2 was used. Source data are provided with this paper.

Code availability

All custom code used for analysis is available on GitHub (https://github.com/sellerslab/ParalogV1_DUSP46).

References

  1. McDonald, E. R. 3rd et al. Project DRIVE: a compendium of cancer dependencies and synthetic lethal relationships uncovered by large-scale, deep RNAi screening. Cell 170, 577–592.e10 (2017).

    Article  CAS  PubMed  Google Scholar 

  2. Behan, F. M. et al. Prioritization of cancer therapeutic targets using CRISPR–Cas9 screens. Nature 568, 511–516 (2019).

    Article  CAS  PubMed  Google Scholar 

  3. Tsherniak, A. et al. Defining a cancer dependency map. Cell 170, 564–576.e16 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Chan, E. M. et al. WRN helicase is a synthetic lethal target in microsatellite unstable cancers. Nature 568, 551–556 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Mavrakis, K. J. et al. Disordered methionine metabolism in MTAP/CDKN2A-deleted cancers leads to dependence on PRMT5. Science 351, 1208–1213 (2016).

    Article  CAS  PubMed  Google Scholar 

  6. Lieb, S. et al. Werner syndrome helicase is a selective vulnerability of microsatellite instability-high tumor cells. eLife 8, e43333 (2019).

    Google Scholar 

  7. Kategaya, L., Perumal, S. K., Hager, J. H. & Belmont, L. D. Werner syndrome helicase is required for the survival of cancer cells with microsatellite instability. iScience 13, 488–497 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Kryukov, G. V. et al. MTAP deletion confers enhanced dependency on the PRMT5 arginine methyltransferase in cancer cells. Science 351, 1214–1218 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. De Kegel, B. & Ryan, C. J. Paralog buffering contributes to the variable essentiality of genes in cancer cell lines. PLoS Genet. 15, e1008466 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Kelly, M. R. et al. Combined proteomic and genetic interaction mapping reveals new RAS effector pathways and susceptibilities. Cancer Discov. 10, 1950–1967 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Kuzmin, E. et al. Systematic analysis of complex genetic interactions. Science 360, eaao1729 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  12. Tong, A. H. Y. et al. Global mapping of the yeast genetic interaction network. Science 303, 808–813 (2004).

    Article  CAS  PubMed  Google Scholar 

  13. Costanzo, M. et al. A global genetic interaction network maps a wiring diagram of cellular function. Science 353, aaf1420 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  14. Fischer, B. et al. A map of directional genetic interactions in a metazoan cell. eLife 4, e05464 (2015).

    Google Scholar 

  15. Han, K. et al. Synergistic drug combinations for cancer identified in a CRISPR screen for pairwise genetic interactions. Nat. Biotechnol. 35, 463–474 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Norman, T. M. et al. Exploring genetic interaction manifolds constructed from rich single-cell phenotypes. Science 365, 786–793 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Najm, F. J. et al. Orthologous CRISPR–Cas9 enzymes for combinatorial genetic screens. Nat. Biotechnol. 36, 179–189 (2018).

    Article  CAS  PubMed  Google Scholar 

  18. Horlbeck, M. A. et al. Mapping the genetic landscape of human cells. Cell 174, 953–967.e22 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Shen, J. P. et al. Combinatorial CRISPR–Cas9 screens for de novo mapping of genetic interactions. Nat. Methods 14, 573–576 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Zhao, D. et al. Combinatorial CRISPR–Cas9 metabolic screens reveal critical redox control points dependent on the KEAP1-NRF2 regulatory axis. Mol. Cell 69, 699–708.e7 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Hart, T. & Moffat, J. BAGEL: a computational framework for identifying essential genes from pooled library screens. BMC Bioinformatics 17, 164 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  22. Helming, K. C. et al. ARID1B is a specific vulnerability in ARID1A-mutant cancers. Nat. Med. 20, 251–254 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Zamanighomi, M. et al. GEMINI: a variational Bayesian approach to identify genetic interactions from combinatorial CRISPR screens. Genome Biol. 20, 137 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

  24. Chen, X. et al. Combined PKC and MEK inhibition in uveal melanoma with GNAQ and GNA11 mutations. Oncogene 33, 4724–4734 (2014).

    Article  CAS  PubMed  Google Scholar 

  25. O’Leary, B., Finn, R. S. & Turner, N. C. Treating cancer with selective CDK4/6 inhibitors. Nat. Rev. Clin. Oncol. 13, 417–430 (2016).

    Article  PubMed  Google Scholar 

  26. Park, E. et al. Architecture of autoinhibited and active BRAF–MEK1–14-3-3 complexes. Nature 575, 545–550 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Camps, M., Nichols, A. & Arkinstall, S. Dual specificity phosphatases: a gene family for control of MAP kinase function. FASEB J. 14, 6–16 (2000).

    Article  CAS  PubMed  Google Scholar 

  28. Kidger, A. M. & Keyse, S. M. The regulation of oncogenic Ras/ERK signalling by dual-specificity mitogen activated protein kinase phosphatases (MKPs). Semin. Cell Dev. Biol. 50, 125–132 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Unni, A. M., Lockwood, W. W., Zejnullahu, K., Lee-Lin, S.-Q. & Varmus, H. Evidence that synthetic lethality underlies the mutual exclusivity of oncogenic KRAS and EGFR mutations in lung adenocarcinoma. eLife 4, e06907 (2015).

    Google Scholar 

  30. Leung, G. P. et al. Hyperactivation of MAPK signaling is deleterious to RAS/RAF-mutant melanoma. Mol. Cancer Res. 17, 199–211 (2019).

    Article  CAS  PubMed  Google Scholar 

  31. Brenan, L. et al. Phenotypic characterization of a comprehensive set of MAPK1/ERK2 missense mutants. Cell Rep. 17, 1171–1183 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Goetz, E. M., Ghandi, M., Treacy, D. J., Wagle, N. & Garraway, L. A. ERK mutations confer resistance to mitogen-activated protein kinase pathway inhibitors. Cancer Res. 74, 7079–7089 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Sanchez-Vega, F. et al. Oncogenic signaling pathways in The Cancer Genome Atlas. Cell 173, 321–337.e10 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Unni, A. M. et al. Hyperactivation of ERK by multiple mechanisms is toxic to RTK-RAS mutation-driven lung adenocarcinoma cells. eLife 7, e33718 (2018).

    Google Scholar 

  35. Liu, S., Sun, J.-P., Zhou, B. & Zhang, Z.-Y. Structural basis of docking interactions between ERK2 and MAP kinase phosphatase 3. Proc. Natl Acad. Sci. USA 103, 5326–5331 (2006).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Peng, D.-J., Zhou, J.-Y. & Wu, G. S. Post-translational regulation of mitogen-activated protein kinase phosphatase-2 (MKP-2) by ERK. Cell Cycle 9, 4650–4655 (2010).

    Article  CAS  PubMed  Google Scholar 

  37. Nazarian, R. et al. Melanomas acquire resistance to B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature 468, 973–977 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Villanueva, J. et al. Acquired resistance to BRAF inhibitors mediated by a RAF kinase switch in melanoma can be overcome by cotargeting MEK and IGF-1R/PI3K. Cancer Cell 18, 683–695 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Corcoran, R. B. et al. EGFR-mediated re-activation of MAPK signaling contributes to insensitivity of BRAF-mutant colorectal cancers to RAF inhibition with vemurafenib. Cancer Discov. 2, 227–235 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Hatzivassiliou, G. et al. RAF inhibitors prime wild-type RAF to activate the MAPK pathway and enhance growth. Nature 464, 431–435 (2010).

    Article  CAS  PubMed  Google Scholar 

  41. Das Thakur, M. et al. Modelling vemurafenib resistance in melanoma reveals a strategy to forestall drug resistance. Nature 494, 251–255 (2013).

    Article  CAS  PubMed  Google Scholar 

  42. Johannessen, C. M. et al. COT drives resistance to RAF inhibition through MAP kinase pathway reactivation. Nature 468, 968–972 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Wagle, N. et al. Dissecting therapeutic resistance to RAF inhibition in melanoma by tumor genomic profiling. J. Clin. Oncol. 29, 3085–3096 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Konieczkowski, D. J., Johannessen, C. M. & Garraway, L. A. A convergence-based framework for cancer drug resistance. Cancer Cell 33, 801–815 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Viswanathan, S. R. et al. Genome-scale analysis identifies paralog lethality as a vulnerability of chromosome 1p loss in cancer. Nat. Genet. 50, 937–943 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. Series B Stat. Methodol. 57, 289–300 (1995).

    Google Scholar 

  47. Doench, J. G. et al. Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR–Cas9. Nat. Biotechnol. 34, 184–191 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank W. Kaelin (Dana-Farber Cancer Institute), M. Meyerson (Dana-Farber Cancer Institute) and L. Lum (Loxo Oncology) for helpful discussions and the Broad Genomics Platform and Genetic Perturbation Platform for their contribution. This work was supported by grants from Ludwig Cancer Research at Harvard (no. 500506) and Broad/IBM Cancer Resistance Research Project (G.G. and L.P.). T.I. is supported by the Department of Defense Peer Reviewed Cancer Research Program Horizon Award (no. W81XWH-19-1-0271). J.G.D. is supported by the Next Generation Fund at the Broad Institute of MIT and Harvard.

Author information

Authors and Affiliations

Authors

Contributions

T.I. and W.R.S. conceived the studies. J.G.D. designed the strategy for library cloning and generated the pRDA_026 vector. M.Z., T.I., R.L., A.W., S.J. and J.M.K. performed the bioinformatic and statistical analysis. L.C. analyzed the TCGA expression dataset. J.G.D. and D.E.R. provided management support for library cloning. T.I., S.D., M.J.Y. and D.P. performed the CRISPR screens. C.T.L. performed the structural analyses. C.M.J., M.V.R. and B.R.P. generated the resistant cell lines. G.G., L.P. and F.V. provided management support for the Avana DepMap CRISPR screens on resistant cells. T.I., D.M., D.J.R. and M.J.Y. performed the biological validation and analysis. T.I., R.L. and M.Z. prepared the figures and tables for the main text and extended data. T.I. and W.R.S. wrote the manuscript with critical reading and feedback from the other coauthors.

Corresponding author

Correspondence to William R. Sellers.

Ethics declarations

Competing interests

During the conduct of this research, W.R.S. was or is a Board or Scientific Advisory Board member and equity holder in Peloton Therapeutics, IDEAYA Biosciences, Civetta Therapeutics, Scorpion Therapeutics and Bluebird and has consulted for Array, Astex, Dynamo Therapeutics, Ipsen, PearlRiver Bio, Sanofi and Servier and receives research funding from Pfizer Pharmaceuticals, Merck, IDEAYA Biosciences and Deerfield Management. J.G.D. consults for Foghorn Therapeutics, Maze Therapeutics, Merck, Agios and Pfizer; he also consults for and has equity in Tango Therapeutics. D.E.R. receives research funding from Functional Genomics Consortium members AbbVie, Janssen, Merck and Vir. G.G. received research funding from IBM and Pharmacyclics and is a founder, consultant and holds private equity in Scorpion Therapeutics. T.I. is a current employee and equity holder of Scorpion Therapeutics. F.V. receives funding from Novo Ventures. C.M.J. is a current Novartis employee and stockholder. A.W. is a current employee of Boehringer Ingelheim. The other authors declare no competing interests.

Additional information

Peer review information Nature Genetics thanks Peter Jackson and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Single-gene perturbation screens potentially miss dependencies of functionally paralogous genes.

(a) Stacked bar graph showing the percentage of non-essential or pan-essential human genes with or without paralogs. (b) Box and whisker plots for trametinib (BRD:BRD-K12343256-001-08-9) sensitivity from the Cancer Therapeutics Response Portal (CTRPv2.0_2015 dataset; https://portals.broadinstitute.org/ctrp/). Trametinib was dosed at 16 concentrations in duplicate. Percent-viability curves were fit and the area-under-concentration-response curve (AUC) was calculated. The AUCs of NRASMUT (n=31) and NRASWT (n=485) cells are shown (left). NRAS, MAP2K1, and MAP2K2 CERES dependence scores in NRASMUT (n=47) and NRASWT (n=692) cells from DepMap screen (right). The centerline, lower hinge, and upper hinge correspond to the 50th, 25th, and 75th percentiles, respectively. The upper and lower whiskers extend from the upper and lower hinges to the largest and smallest values no further than 1.5 * IQR (interquartile range). All observations beyond the whiskers are shown in black dots. Two-sided Wilcox P-values are shown. (c) Schematic of the dual sgRNA cloning strategy. (d) Library representation for pDNA from Digenic Paralog, Big Papi, CDKO and early time point gDNA from Shen-Mali (combined from 293T, A549, and HeLa) and Zhao-Mali (combined from A549 and Hela). (e) Mismatch reads were calculated as the percentage of reads with unintended pairs of sgRNAs from the pDNA and gDNA (at 21 days post library transduction).

Extended Data Fig. 2 Robustness and reproducibility of Digenic Paralog CRISPR screens.

(a) Pearson correlation heatmap of raw LFCs between all pairwise combinations of the Digenic Paralog CRISPR screens. (b) Scatterplot of the raw LFC for each gene in the Digenic Paralog CRISPR library against the raw LFC from Avana DepMap CRISPR dataset (20Q1 public dataset) across ten cell lines with available DepMap data. Each dot is annotated by color based on the essentiality profile. Pearson correlation between the two screens was calculated using non-essential and pan-essential genes. (c) Pearson correlation between the average LFC of target guides in position A-B versus position B-A across 11 cell lines.

Extended Data Fig. 3 Characterization of synergistic paralog genes and selectivity of the PRKC paralog family in GNAQ-mutant uveal melanoma.

(a) Percent of synergistic pairs in the Digenic Paralog CRISPR library with FDR below 5% across 11 cell lines. (b) Distribution of genes that exhibited synergy. (c) Distribution of paralog pairs that exhibited synergy across protein class (top) and size of paralog family (bottom). (d) Frequency of synergistic pairs observed across a varying numbers of cell lines. (e) The distribution of non-synergistic pairs and synergistic pairs by whether the genes were expressed across Cancer Cell Line Encyclopedia (CCLE) RNA-seq samples (expressed gene defined as log2(TPM+1) > 2 in at least 50% of 1270 cell lines). (f) Heatmap of GEMINI synergy FDRs for gene pairs within the PRKC paralog family. (g) Heatmap of the LFC for gene pairs within the PRKC paralog family. Results are shown for individual cell lines.

Extended Data Fig. 4 DUSP4 and DUSP6 as dependencies in NRAS-mutant cells.

(a) Heatmap of GEMINI synergy FDRs for gene pairs associated with MAPK signaling (KEGG_MAPK_SIGNALING_PATHWAY). (b) Paralog pairs showing differential dependence across indications (TP53 mutation, NRAS/KRAS mutation, MYC amplification, CDKN2A deletion). The mean difference of LFC and P-values (empirical Bayes moderated t-test) are plotted. Each point is annotated by the color and size of the mean difference of the GEMINI synergy score. (c) LFCs of individual sgRNAs targeting DUSP4, DUSP6, or both in indicated cell lines. Inferred LFC is denoted with a line where each dot represents a sgRNA combination. (d) Normalized pERK1/2 (Thr202/Try204) signal. MELJUSO cells were infected with sgRNA targeting chromosome 2 or DUSP6 in combination with doxycycline (DOX)-inducible shRNA targeting LacZ or DUSP4. Cells were treated with or without 2uM of DOX for 4 days and pERK1/2 was measured by AlphaLISA. (e) Immunoblot for DUSP4, DUSP6, phospho-ERK1/2 and ERK1/2 were performed 2 days following sgRNA transduction and antibiotic selection in MELJUSO cells ectopically expressing sgRNA-resistant wild-type or phosphatase-dead DUSP4 (C280S) or DUSP6 (C293S) cDNA and infected with lentivirus producing sgRNAs targeting DUSP4, DUSP6 or both. cDNA expressing LacZ was used as a control. (f) MAPK wild-type squamous cell carcinoma cell line (A431) ectopically expressing DOX-inducible GFP, wild-type NRAS, or NRAS(Q61R). Cells were treated with 2uM DOX for 2 days and immunoblotted for the indicated proteins. (g) Relative viability in cells infected with a dual promoter lentiviral vector expressing two S. pyogenes sgRNAs in presence of DMSO or trametinib (trametinib; 4nM). Two different sgRNAs pairs were used against DUSP4 or DUSP6 and sgRNAs targeting AAVS1 and chromosome 2 intergenic sites were used as controls. (h) LFC of individual sgRNAs for DUSP4, DUSP6 or both in DMSO (left) or SCH772984 (right) treatment. Inferred LFC is denoted with a line where each dot represents a sgRNA combination. The experiment in d was from two independent experiments and g from three independent experiments. Data are mean ± s.e.m. *P-value ≤ 0.01, **P-value≤ 0.001, unpaired two-sided t-test.

Source data

Extended Data Fig. 5 Association between DUSP4 dependency and DUSP4/6 levels.

(a) Top 5 genes correlated with BRAF CERES score (left - Pearson) or BRAF mutation (right - Point-Biserial) from the DepMap screens. (b) Scatterplot of CERES scores for BRAF and DUSP4. (c) Scatter plot of CERES score for DUSP4 compared to the expression of DUSP4 or DUSP6. (d) Scatter plot of CERES score for DUSP4 compared to the mass spectrometry-based proteomics levels of DUSP4 or DUSP6. All expression values are in log2(TPM +1). Proteomic levels are shown as normalized log2-transformed ratios to the bridge sample in each Tandem Mass Tags (TMT) 10-plex. (e) Relative viability of cells infected with sgRNAs pairs targeting DUSP4 or DUSP6. AAVS1 and chromosome 2 intergenic sgRNAs were used as controls. P-values were calculated based on linear regression analysis. Experiments in e were from three independent experiments. Data are mean ± s.e.m. *P-value ≤ 0.01, unpaired two-sided t-test for e. P-values for b-d were calculated based on linear regression analysis.

Extended Data Fig. 6 Cells resistant to MAPK inhibitors are cross-sensitized to DUSP4 knockout.

(a) Percent cell viability of parental (gray), TDr (blue), or TDSr (red) A375 cell lines treated with the varying concentrations of trametinib, dabrafenib, or SCH772985 measured after 4 days using CellTiter-Glo. Cell viability was normalized to DMSO control. Data are mean ± s.e.m. of biological duplicates. (b) Schematic of HT144 cell lines resistant to 25nM dabrafenib (Dr). (c) Rank-ordered depiction of the difference in CERES score between HT144 cells resistant to dabrafenib (Dr) and parental HT144 from genome-wide single-gene CRISPR screen. Delta CERES was calculated by CERES HT144Dr - CERES HT144Parental.

Supplementary information

Reporting Summary.

Supplementary Tables

Supplementary Tables 1–13.

Source data

Source Data Fig. 3

Unprocessed western blots for Fig. 3.

Source Data Fig. 4

Unprocessed western blots for Fig. 4.

Source Data Fig. 5

Unprocessed western blots for Fig. 5.

Source Data Extended Data Fig. 4

Unprocessed western blots for Extended Data Fig. 4.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ito, T., Young, M.J., Li, R. et al. Paralog knockout profiling identifies DUSP4 and DUSP6 as a digenic dependence in MAPK pathway-driven cancers. Nat Genet 53, 1664–1672 (2021). https://doi.org/10.1038/s41588-021-00967-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41588-021-00967-z

This article is cited by

Search

Quick links

Nature Briefing: Translational Research

Sign up for the Nature Briefing: Translational Research newsletter — top stories in biotechnology, drug discovery and pharma.

Get what matters in translational research, free to your inbox weekly. Sign up for Nature Briefing: Translational Research