Genetic compensation induced by deleterious mutations but not gene knockdowns

Article metrics



Cells sense their environment and adapt to it by fine-tuning their transcriptome. Wired into this network of gene expression control are mechanisms to compensate for gene dosage. The increasing use of reverse genetics in zebrafish, and other model systems, has revealed profound differences between the phenotypes caused by genetic mutations and those caused by gene knockdowns at many loci1,2,3, an observation previously reported in mouse and Arabidopsis4,5,6,7. To identify the reasons underlying the phenotypic differences between mutants and knockdowns, we generated mutations in zebrafish egfl7, an endothelial extracellular matrix gene of therapeutic interest, as well as in vegfaa. Here we show that egfl7 mutants do not show any obvious phenotypes while animals injected with egfl7 morpholino (morphants) exhibit severe vascular defects. We further observe that egfl7 mutants are less sensitive than their wild-type siblings to Egfl7 knockdown, arguing against residual protein function in the mutants or significant off-target effects of the morpholinos when used at a moderate dose. Comparing egfl7 mutant and morphant proteomes and transcriptomes, we identify a set of proteins and genes that are upregulated in mutants but not in morphants. Among them are extracellular matrix genes that can rescue egfl7 morphants, indicating that they could be compensating for the loss of Egfl7 function in the phenotypically wild-type egfl7 mutants. Moreover, egfl7 CRISPR interference, which obstructs transcript elongation and causes severe vascular defects, does not cause the upregulation of these genes. Similarly, vegfaa mutants but not morphants show an upregulation of vegfab. Taken together, these data reveal the activation of a compensatory network to buffer against deleterious mutations, which was not observed after translational or transcriptional knockdown.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Generation of zebrafish egfl7 mutant alleles and sporadic brain haemorrhage in mutant larvae.
Figure 2: Zebrafish egfl7 mutant embryos are less sensitive to egfl7 morpholino injections.
Figure 3: Emilin3a is upregulated in mutant but not in morphant or CRISPRi embryos.
Figure 4: Emilin2 and Emilin3 can rescue egfl7 morphants.


  1. 1

    Kok, F. O. et al. Reverse genetic screening reveals poor correlation between morpholino-induced and mutant phenotypes in zebrafish. Dev. Cell 32, 97–108 (2015)

  2. 2

    Stainier, D. Y., Kontarakis, Z. & Rossi, A. Making sense of anti-sense data. Dev. Cell 32, 7–8 (2015)

  3. 3

    Law, S. H. & Sargent, T. D. The serine-threonine protein kinase PAK4 is dispensable in zebrafish: identification of a morpholino-generated pseudophenotype. PLoS ONE 9, e100268 (2014)

  4. 4

    Daude, N. et al. Knockout of the prion protein (PrP)-like Sprn gene does not produce embryonic lethality in combination with PrP(C)-deficiency. Proc. Natl Acad. Sci. USA 109, 9035–9040 (2012)

  5. 5

    De Souza, A. T. et al. Transcriptional and phenotypic comparisons of Ppara knockout and siRNA knockdown mice. Nucleic Acids Res. 34, 4486–4494 (2006)

  6. 6

    Smart, N. & Riley, P. R. Thymosin β4 in vascular development, response to research commentary. Circ. Res. 112, 29–30 (2013)

  7. 7

    Gao, Y. et al. Auxin binding protein 1 (ABP1) is not required for either auxin signaling or Arabidopsis development. Proc. Natl Acad. Sci. USA 112, 2275–2280 (2015)

  8. 8

    Anon. RNA interference on target. Nature Methods 3, 659 (2006)

  9. 9

    Robu, M. E. et al. p53 activation by knockdown technologies. PLoS Genet. 3, e78 (2007)

  10. 10

    Schmidt, M. et al. EGFL7 regulates the collective migration of endothelial cells by restricting their spatial distribution. Development 134, 2913–2923 (2007)

  11. 11

    Kuhnert, F. et al. Attribution of vascular phenotypes of the murine Egfl7 locus to the microRNA miR-126 . Development 135, 3989–3993 (2008)

  12. 12

    Parker, L. S. et al. The endothelial-cell-derived secreted factor Egfl7 regulates vascular tube formation. Nature 428, 754–758 (2004)

  13. 13

    Charpentier, M. S. et al. CASZ1 promotes vascular assembly and morphogenesis through the direct regulation of an EGFL7/RhoA-mediated pathway. Dev. Cell 25, 132–143 (2013)

  14. 14

    Huang, C. et al. VE-statin/Egfl7 siRNA inhibits angiogenesis in malignant glioma in vitro . Int. J. Clin. Exp. Pathol. 7, 1077–1084 (2014)

  15. 15

    Cermak, T. et al. Efficient design and assembly of custom TALEN and other TAL effector-based constructs for DNA targeting. Nucleic Acids Res. 39, e82 (2011)

  16. 16

    Nichol, D. & Stuhlmann, H. EGFL7: a unique angiogenic signaling factor in vascular development and disease. Blood 119, 1345–1352 (2012)

  17. 17

    Chi, N. C. et al. Foxn4 directly regulates tbx2b expression and atrioventricular canal formation. Genes Dev. 22, 734–739 (2008)

  18. 18

    Jin, S. W., Beis, D., Mitchell, T., Chen, J. N. & Stainier, D. Y. Cellular and molecular analyses of vascular tube and lumen formation in zebrafish. Development 132, 5199–5209 (2005)

  19. 19

    Larson, M. H. et al. CRISPR interference (CRISPRi) for sequence-specific control of gene expression. Nature Protocols 8, 2180–2196 (2013)

  20. 20

    Zanetti, M. et al. EMILIN-1 deficiency induces elastogenesis and vascular cell defects. Mol. Cell. Biol. 24, 638–650 (2004)

  21. 21

    Lelievre, E. et al. VE-statin/egfl7 regulates vascular elastogenesis by interacting with lysyl oxidases. EMBO J. 27, 1658–1670 (2008)

  22. 22

    Sulem, P. et al. Identification of a large set of rare complete human knockouts. Nature Genet. 47, 448–452 (2015)

  23. 23

    Doyle, E. L. et al. TAL effector specificity for base 0 of the DNA target is altered in a complex, effector- and assay-dependent manner by substitutions for the tryptophan in cryptic repeat -1. PLoS ONE 8, e82120 (2013)

  24. 24

    Bedell, V. M. et al. In vivo genome editing using a high-efficiency TALEN system. Nature 491, 114–118 (2012)

  25. 25

    Jao, L. E., Wente, S. R. & Chen, W. Efficient multiplex biallelic zebrafish genome editing using a CRISPR nuclease system. Proc. Natl Acad. Sci. USA 110, 13904–13909 (2013)

  26. 26

    Nolte, H. et al. Global protein expression profiling of zebrafish organs based on in vivo incorporation of stable isotopes. J. Proteome Res. 13, 2162–2174 (2014)

  27. 27

    Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nature Biotechnol. 26, 1367–1372 (2008)

  28. 28

    Cox, J. et al. Andromeda: a peptide search engine integrated into the MaxQuant environment. J. Proteome Res. 10, 1794–1805 (2011)

  29. 29

    Tusher, V. G., Tibshirani, R. & Chu, G. Significance analysis of microarrays applied to the ionizing radiation response. Proc. Natl Acad. Sci. USA 98, 5116–5121 (2001)

Download references


We thank H.-B. Kwon and other members of the laboratory, past and present, as well as K. Sampath, D. Wainstock, C. Moens and M. Grether, for discussions, comments on the manuscript and/or reagents, and the Max Planck Society, Packard foundation, and EMBO for funding.

Author information

All authors were involved in the experimental design, data analysis and writing. Experiments were performed by all except M.K. and D.Y.R.S., who also supervised the project.

Correspondence to Didier Y. R. Stainier.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 Generation and identification of zebrafish egfl7 mutant alleles.

a, TALENs were designed to target exon 3 of egfl7 which encodes part of the EMI domain. Sequence alignment of part of exon 3 from the WT, egfl7s980 and egfl7s981 alleles shows TALEN indels: Δ3/s980 (three nucleotide deletion) and Δ4/s981 (five nucleotide deletion), and one nucleotide insertion (yellow). b, Genotyping example of single embryos sampled from a population of egfl7 WT, egfl7 s981/+ and egfl7s981/s981 fish using high-resolution melt analysis. The green curve corresponds to the WT allele and the red one to the egfl7s981 allele. Heterozygous embryos have both alleles and thus the melting profile (in blue) is a composition of the WT and mutant curves.

Extended Data Figure 2 The egfl7s981 mutation leads to egfl7 mRNA degradation, reduced protein expression and impaired secretion.

a, The egfl7s981 mutation leads to egfl7 mRNA degradation: egfl7 mRNA expression in 24 hpf WT, egfl7s981/s981 and egfl7s980/s980 embryos. Expression normalized to gapdh. b, The egfl7s981 (p.Gln49Leufs*30) mutation leads to strongly reduced protein expression. Western blot analyses of Egfl7-Myc-tag expression in transfected HUVEC cells. Egfl7 WT and Egfl7s980 protein expression was strongly detected in the medium whereas the Egfl7s981 isoform was strongly reduced in the cells and very poorly secreted (right), or undetectable in both (left). Furthermore, Egfl7s981 shares high similarity to the truncated protein produced in the original Egfl7 mutant mouse in which the protein was not detectable using an Egfl7 antibody10.

Extended Data Figure 3 Vessel integrity and permeability do not appear to be affected in egfl7s981/s981 larvae.

A fluorescent molecule (2000 kDa FITC-dextran) was injected directly into the circulation of 72 hpf Tg(kdrl:HRAS:mCherry) larvae that previously showed haemorrhage, which was mostly localized around the hindbrain ventricle. Confocal micrographs of 72 hpf Tg(kdrl:HRAS:mCherry) expression, FITC-dextran and MERGE of WT and egfl7s981/s981 larvae in (a) lateral and (b) dorsal views. The FITC-dextran did not accumulate to the sites of haemorrhage, suggesting that these sites had clotted and vascular integrity had been restored after the initial blood leakage.

Extended Data Figure 4 In vivo genome editing: Myc-tag introduction in the egfl7 endogenous locus.

a, TALENs targeting the egfl7 stop codon created double-stranded breaks in the chromosomal DNA. Homology-directed repair precisely incorporated the Myc tag exogenous sequence (ssDNA) at the cut site. b, Western blot analysis of Egfl7-Myc-tag expression in 24 hpf control and morphant embryos. Egfl7 Myc-tag signal was reduced by around 80% in morphants (1 ng egfl7 MO) compared with uninjected. Expression normalized to tubulin (P ≤ 0.05). Error bars, s.e.m. (n = 3).

Extended Data Figure 5 The egfl7 morpholino does not significantly affect p53 mRNA expression at 1 ng per embryo but it does so at higher doses.

mRNA expression of p53 in 24 hpf WT, egfl7 Δ3 (egfl7s980 ) and egfl7 Δ4 (egfl7s981 ) mutant, and morphant (1, 2 and 4 ng injected) embryos. Expression normalized to gapdh. Error bars, s.e.m. of technical triplicates.

Extended Data Figure 6 The egfl7 transcript elongation inhibition causes a phenotype similar to the one seen in morphants.

a, gRNAs of egfl7 targeting the template (T) strand in exon 2 and non-template (NT) strand in the 5′ UTR and exon 2. b, Expression of egfl7 in non-template (NT) gRNA and template (T) gRNA-injected embryos relative to uninjected (CT) siblings at 20 hpf. qPCR data, pools of ten embryos each, expression normalized to gapdh (P ≤ 0.05). Error bars, s.e.m. (n = 3). c, Lateral view confocal micrographs of 48 hpf Tg(kdrl:GFP) embryos injected with egfl7 template and non-template CRISPRi. Template CRISPRi (top) embryos are indistinguishable from non-injected siblings, while non-template CRISPRi embryos exhibit different degrees of vascular defects (middle: mild; bottom: severe).

Extended Data Figure 7 Single-shot proteomics to assess changes between WT and egfl7s981 mutant embryos.

a, Schematic visualization of proteomic workflow. Embryos were lysed in urea buffer, and proteins were digested in-solution using trypsin and measured on a QExactive bench top instrument. Acquired spectra were analysed against the Uniprot zebrafish database (2014) using MaxQuant. b, Scatter plot matrix shows high correlation between biological replicates. Reproducibility was determined by a Pearson correlation coefficient.

Extended Data Figure 8 Emilin3a expression is upregulated in mutant but not morphant embryos.

a, Volcano plot showing significantly dysregulated proteins between egfl7 morphant and WT embryos at 24 hpf using label-free quantification. Emilin3a (blue) levels were not significantly different between morphant and WT embryos. Emilin3b is also highlighted in blue. b, Bar plot showing upregulation of emilin family members in 24 hpf egfl7 mutants compared with WT and morphants, as assessed from RNA-seq data (WT expression set at 1 for each gene).

Extended Data Figure 9 Expression of vegfab is upregulated in vegfaa mutant embryos but not in morphants, or vegfaa dominant negative-injected embryos; qPCR data, pools of ten embryos each, expression normalized to gapdh (P ≤ 0.05).

Error bars, s.e.m. (n = 5). a, mRNA expression of vegfab in 24 hpf vegfaa WT, mutant and morphant embryos. b, mRNA expression of vegfab in 24 hpf vegfaa WT and vegfaa dominant negative-injected embryos (two different dominant negatives were injected).

Supplementary information

Supplementary Data

This file contains Supplementary Table 1. (XLSX 20 kb)

Supplementary Data

This file contains Supplementary Table 2. (XLSX 6556 kb)

Supplementary Data

This file contains Supplementary Table 3. (XLSX 3762 kb)

PowerPoint slides

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Rossi, A., Kontarakis, Z., Gerri, C. et al. Genetic compensation induced by deleterious mutations but not gene knockdowns. Nature 524, 230–233 (2015) doi:10.1038/nature14580

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.