1. Cell Biology
  2. Neuroscience

Hypoxia-inducible factor cell non-autonomously regulates C. elegans stress responses and behavior via a nuclear receptor

  1. Corinne L Pender
  2. H Robert Horvitz  Is a corresponding author
  1. Massachusetts Institute of Technology, United States
https://doi.org/10.7554/eLife.36828
Share this article
Cite this article
  1. Corinne L Pender
  2. H Robert Horvitz
(2018)
Hypoxia-inducible factor cell non-autonomously regulates C. elegans stress responses and behavior via a nuclear receptor
eLife 7:e36828.
https://doi.org/10.7554/eLife.36828
  2. Download BibTeX
  3. Download .RIS
6 figures, 1 table and 3 additional files

Figures

Figure 1 with 2 supplements
Download asset Open asset
The cytochrome P450 gene cyp-36A1 is an effector of the hypoxia-response pathway that regulates behavior.

(A) Schematic of the screen design. Stages of embryonic development adapted from Ringstad and Horvitz (2008) and Paquin et al. (2016). (B–F) Distribution of stages of eggs newly laid by adult hermaphrodites, used as a proxy for egg retention time in utero, of animals of the indicated genotypes. (B) Stages of eggs laid by wild-type animals. (C) egl-9 loss-of-function (lf) mutants laid later stage eggs than the wild type (p<0.001, Chi-square test with Holm-Bonferroni correction). (D) cyp-36A1(n5666) suppressed the egg-laying defect of egl-9(lf) mutants (p<0.001). (E) The cyp-36A1(+) transgene, which contains wild-type cyp-36A1, rescued the suppression of the egg-laying defect observed in cyp-36A1(n5666); egl-9(lf) mutants (p<0.001). (F) cyp-36A1(n5666) mutants displayed wild-type egg laying (p>0.05). (G) egl-9(lf) mutants were defective in locomotion rate, and this defect was suppressed by hif-1(lf) and cyp-36A1(lf) mutations. Mean ± SD of n = 3 biological replicates, *p<0.05, **p<0.01 considered significant (Student’s t-test with Holm-Bonferroni correction). (H) egl-9(lf) mutants were defective in defecation rate, and this defect was suppressed by hif-1(lf) and cyp-36A1(lf). Mean ± SD of n ≥ 6 animals, ***p<0.001 considered significant (Student’s t-test with Holm-Bonferroni correction). (I) Relative expression of cyp-36A1 mRNA in the wild type, egl-9(lf), egl-9(lf) hif-1(lf), and hif-1(lf) mutants, measured by qRT-PCR and normalized to the expression of the large ribosomal subunit rpl-32. Mean ± SD of n = 3 biological replicates, ***p<0.001 considered significant. ns, not significant (Student’s t-test with Holm-Bonferroni correction). Alleles used for (B–F) were cyp-36A1(n5666), egl-9(n586), and nIs674 (nIs [cyp-36A1(+)]), and all strains used in (B–F) contained the nIs470 (Pcysl-2::gfp) transgene. Alleles used for (G–I) were egl-9(sa307), hif-1(ia4), and cyp-36A1(gk824636).

https://doi.org/10.7554/eLife.36828.002
Figure 1—figure supplement 1
Download asset Open asset
A screen for suppressors of the egl-9(lf) egg-laying defect.

(A) Summary of three cloned alleles from a suppressor screen of the egl-9(lf) egg-defect. (B) Gene diagram of hif-1; isoform F38A6.3b is shown. The ia4 deletion allele was generated by deletion screening (Jiang et al., 2001). Scale bar, 100 bases. (C) Gene diagram of cyp-36A1. The gk824636 allele is from the Million Mutation Project (Thompson et al., 2013). Scale bar, 100 bases.

https://doi.org/10.7554/eLife.36828.003
Figure 1—figure supplement 2
Download asset Open asset
cyp-36A1(lf) suppresses the egg-laying defect of hypoxia-exposed animals.

(A) Schematic depicting experimental design. (B–E) Distribution of stages of eggs laid by adult hermaphrodites. (B) Stages of eggs laid by wild-type animals grown in normoxia (21% O2). (C) Animals exposed to hypoxia (1% O2) for 24 hr as adults laid later stage eggs than those in normoxia (p<0.001, Chi-square test with Holm-Bonferroni correction). (D) cyp-36A1(gk824636) mutants grown in normoxia displayed wild-type egg laying (p>0.05). (E) cyp-36A1(gk824636) mutants exposed to 1% O2 laid eggs at earlier stages than wild-type animals in the same condition (p<0.001).

https://doi.org/10.7554/eLife.36828.004
Figure 2 with 1 supplement
Download asset Open asset
CYP-36A1 acts downstream of HIF-1 to regulate gene expression changes and stress responses.

(A) Blue circle: Genes that were at least twofold downregulated in egl-9(lf) mutants. Purple and orange circles: Subset of egl-9(lf)-downregulated genes that were significantly upregulated in egl-9(lf) hif-1(lf) double mutants (purple) or cyp-36A1(lf); egl-9(lf) double mutants (orange) vs. egl-9(lf) single mutants. (B) Blue circle: Genes that were at least twofold upregulated in egl-9(lf) mutants. Purple and orange circles: Subset of egl-9(lf)-upregulated genes that were significantly downregulated in egl-9(lf) hif-1(lf) double mutants (purple) or cyp-36A1(lf); egl-9(lf) double mutants (orange) vs. egl-9(lf) single mutants. Significance for all comparisons in (A) and (B) was based on two biological replicates and determined by the Benjamini-Hochberg procedure with a false-discovery rate of 0.05. (C) Survival of animals grown from the L4 larval stage on the pathogen Pseudomonas aeruginosa. Wild type (n = 86 animals) vs. egl-9(lf) (n = 154) p<0.001; egl-9(lf) vs. cyp-36A1(lf); egl-9(lf) (n = 83), p<0.001; cyp-36A1(lf); egl-9(lf) vs. cyp-36A1(lf); egl-9(lf); nIs [cyp-36A1(+)] (n = 97), p<0.001; wild type vs. cyp-36A1(lf) (n = 61) p<0.05, as determined by the log-rank (Mantel-Cox) test, correcting for multiple comparisons with the Holm-Bonferroni method. See figure supplement for replicate data. (D) Survival of animals to the L4 larval stage or later after growth for three days from the L1 larval stage on plates containing 5 μg/ml tunicamycin. Mean ± SD of n = 3 biological replicates. **p<0.01, ***p<0.001 considered significant. ns (p>0.05), not significant (Student’s t-test with Holm-Bonferroni correction). (E) Survival of animals exposed to 7.5 mM tert-butyl hydroperoxide for 10 hr as young adults. Mean ± SD of n = 5 biological replicates. *p<0.05 considered significant. ns (p>0.05), not significant (Student’s t-test with Holm-Bonferroni correction). Alleles used for (A) and (B) were egl-9(sa307), hif-1(ia4), and cyp-36A1(gk824636). Alleles used for (C–E) were egl-9(n586), cyp-36A1(n5666), and nIs674 (nIs [cyp-36A1(+)]), and all strains used in (C–E) contained the nIs470 (Pcysl-2::gfp) transgene.

https://doi.org/10.7554/eLife.36828.005
Figure 2—source data 1

RNA-seq GO enrichment analysis.

Gene ontology enrichment analysis terms for genes more than twofold upregulated in egl-9 mutants and significantly suppressed by hif-1(lf) and cyp-36A1(lf).

https://doi.org/10.7554/eLife.36828.007
Figure 2—figure supplement 1
Download asset Open asset
Replicate data for survival on Pseudomonas aeruginosa.

Replicate data for Figure 2C. Wild type (n = 150) vs. egl-9(lf) (n = 50), p<0.001; egl-9(lf) vs. cyp-36A1(lf); egl-9(lf) (n = 163), p<0.001; cyp-36A1(lf); egl-9(lf) vs. cyp-36A1(lf); egl-9(lf); nIs [cyp-36A1(+)] (n = 51), p<0.001; wild type vs. cyp-36A1(lf) (n = 35), p>0.05, as determined by the log-rank (Mantel-Cox) test, correcting for multiple comparisons with the Holm-Bonferroni method.

https://doi.org/10.7554/eLife.36828.006
Figure 3 with 4 supplements
Download asset Open asset
HIF-1 and CYP-36A1 cell non-autonomously regulate expression of a stress-responsive gene.

(A) PT24B8.5::gfp expression of the indicated genotypes (n = 10 animals per image). Scale bars, 100 μm. (B) Expression of cyp-36A1(+) specifically in neurons, hypoderm, muscle, or intestine of cyp-36A1(lf); egl-9(lf) animals increased expression of PT24B8.5::gfp in intestine (n = 10 animals per image). Scale bars, 100 μm. (C) Quantification of fluorescence intensity for (A) and (B), measured as average intensity for a 300 μm section of the intestine in the midbody of each animal, as indicated. *p<0.05, **p<0.01, ***p<0.001 considered significant (Student’s t-test with Holm-Bonferroni correction). Mean ± SD of n = 10 animals. See figure supplement for replicate data. (D) Expression of hif-1(P621A), which encodes a stable variant of HIF-1 (Pocock and Hobert, 2008), specifically in neurons, hypoderm, muscle, or intestine increased expression of PT24B8.5::gfp in intestine; increased expression was suppressed by cyp-36A1(lf) (n = 10 animals per image). Scale bars, 100 μm. (E) Quantification of fluorescence intensity for (D), measured as average intensity for a 300 μm section of the intestine in the midbody of each animal, as indicated. *p<0.05, ***p<0.01, ***p<0.001 considered significant (Student’s t-test with Holm-Bonferroni correction). Mean ± SD of n = 10 animals. See figure supplement for replicate data. Alleles used were egl-9(sa307), hif-1(ia4), cyp-36A1(gk824636), nEx2699[Pneurons::hif-1(P621A)], otEx3156 [Phypoderm::hif-1(P621A)], otEx3165 [Pmuscle::hif-1(P621A)], nEx2860 [Pintestine:: hif-1(P621A)], nEx2853 [Pneurons::cyp(+)], nEx2856 [Phypoderm::cyp(+)], nEx2859 [Pmuscle::cyp(+)], and nEx2849 [Pintestine::cyp(+)]. All strains contained the agIs219 (PT24B8.5::gfp) transgene.

https://doi.org/10.7554/eLife.36828.008
Figure 3—figure supplement 1
Download asset Open asset
cyp-36A1 expression in multiple tissues rescues the egg-laying phenotype of cyp-36A1(lf); egl-9(lf).

(A–H) Distribution of stages of eggs laid by adult hermaphrodites of the indicated genotypes. All strains contained the agIs219 (PT24B8.5::gfp) transgene. (A) Stages of eggs laid by wild-type animals. (B) egl-9(sa307) animals laid later stage eggs than wild type (p<0.001, Chi-square test with Holm-Bonferroni correction). (C) cyp-36A1(gk824636) suppressed the egg-laying defect of egl-9(sa307) mutants (p<0.001). (D) Stages of eggs laid by cyp-36A1(gk824636) single mutants were not significantly different from those laid by wild-type animals (p>0.05). (E–H) Expression of cyp-36A1(+) specifically in neurons, hypoderm, muscle, or intestine of cyp-36A1(gk824636); egl-9(sa307) animals all rescued the suppression by cyp-36A1(lf) of the egg-laying defect of egl-9(lf) (p<0.001).

https://doi.org/10.7554/eLife.36828.009
Figure 3—figure supplement 2
Download asset Open asset
cyp-36A1 is expressed in many tissues.

(A–D) Paired fluorescent (left in each panel) and Nomarski (right in each panel) micrographs showing expression of a transcriptional Pcyp-36A1::gfp reporter (nIs682) in an adult worm. Expression was observed in the head (A), midbody (B and C), and tail (D), including in neurons, body-wall muscle, vulval muscle, intestine, and hypoderm, as indicated. Scale bars, 20 μm.

https://doi.org/10.7554/eLife.36828.010
Figure 3—figure supplement 3
Download asset Open asset
Replicate data for PT24B8.5::gfp reporter expression.

Replicate data for Figure 3C. Quantification of n = 10 animals. ***p<0.001 considered significant. ns (p>0.05), not significant (Student’s t-test with Holm-Bonferroni correction).

https://doi.org/10.7554/eLife.36828.011
Figure 3—figure supplement 4
Download asset Open asset
Replicate data for PT24B8.5::gfp reporter expression.

Replicate data for Figure 3E. Quantification of n = 10 animals. **p<0.01, ***p<0.001 considered significant (Student’s t-test with Holm-Bonferroni correction).

https://doi.org/10.7554/eLife.36828.012
Figure 4 with 2 supplements
Download asset Open asset
The nuclear hormone receptor NHR-46 acts downstream of CYP-36A1.

(A) nhr-46 gene diagram; isoform C45E5.6b is shown. n6125 and n6126 are in the NHR-46 ligand-binding domain. Scale bar, 100 bases. (B) PT24B8.5::gfp fluorescence of the indicated genotypes (n = 10 animals per image). Scale bars, 100 μm. (C) Quantification of fluorescence intensity for (B), measured as average intensity for a 300 μm section of the intestine in the midbody of each animal, as indicated. *p<0.05, ***p<0.001 considered significant (Student’s t-test with Holm-Bonferroni correction). Mean ±SD of n = 10 animals. See figure supplement for replicate data. (D) Distribution of stages of eggs newly laid by adult hermaphrodites of the indicated genotypes. ***p<0.001 considered significant. ns (p>0.05), not significant (Chi-square test with Holm-Bonferroni correction). (E) Survival of animals grown from the L4 larval stage on the pathogen Pseudomonas aeruginosa. Wild type (n = 129) vs. egl-9(lf) (n = 67) p<0.001; egl-9(lf) vs. cyp-36A1(lf); egl-9(lf) (n = 129), p<0.001; cyp-36A1(lf); egl-9(lf) vs. cyp-36A1(lf); nhr-46(lf); egl-9(lf) (n = 94), p<0.001; wild type vs. cyp-36A1(lf) (n = 106), p<0.05; wild type vs. nhr-46(lf) (n = 121), p<0.001, as determined by the log-rank (Mantel-Cox) test, correcting for multiple comparisons with the Holm-Bonferroni method. egl-9(lf) allele was egl-9(n586). See figure supplement for replicate data. (F) Survival of animals to the L4 larval stage or later after growth for three days from the L1 larval stage on plates containing 5 μg/ml tunicamycin. Mean ± SD of n = 3 biological replicates. ***p<0.001 considered significant. ns (p>0.05), not significant (Student’s t-test with Holm-Bonferroni correction). (G) Survival of animals exposed to 7.5 mM tert-butyl hydroperoxide for 10 hr as young adults. Mean ± SD of n = 4 biological replicates. *p<0.05 considered significant. ns, not significant (Student’s t-test with Holm-Bonferroni correction). Alleles used for (B–G) were egl-9(sa307), cyp-36A1(gk824636), nhr-46(n6126), and nEx2586 (nEx [nhr-46(+)]) except where otherwise noted. All strains in (B), (C), (D), (F), and (G) contained the agIs219 (PT24B8.5::gfp) transgene.

https://doi.org/10.7554/eLife.36828.013
Figure 4—figure supplement 1
Download asset Open asset
Replicate data for PT24B8.5::gfp reporter expression.

Replicate data for Figure 4C. Quantification of n = 10 animals. ***p<0.001 considered significant (Student’s t-test with Holm-Bonferroni correction).

https://doi.org/10.7554/eLife.36828.014
Figure 4—figure supplement 2
Download asset Open asset
Replicate data for survival on Pseudomonas aeruginosa.

Replicate data for Figure 4E. Wild type (n = 34) vs. egl-9(n586) (n = 61), p<0.001; egl-9(lf) vs. cyp-36A1(lf); egl-9(lf) (n = 52), p<0.05; cyp-36A1(lf); egl-9(lf) vs. cyp-36A1(lf); nhr-46(lf); egl-9(lf) (n = 39), p<0.001; wild type vs. cyp-36A1(lf) (n = 91), p>0.05; wild type vs. nhr-46(lf) (n = 54), p<0.05, as determined by the log-rank (Mantel-Cox) test, correcting for multiple comparisons with the Holm-Bonferroni method.

https://doi.org/10.7554/eLife.36828.015
Figure 5 with 1 supplement
Download asset Open asset
NHR-46 acts in different tissues to regulate T24B8.5 expression and egg laying.

(A) Expression of nhr-46(+) in the intestine but not in neurons or muscle rescued the high PT24B8.5::gfp expression in the intestine of cyp-36A1(lf); nhr-46(lf); egl-9(lf) mutants (n = 10 animals per image). Scale bars, 100 μm. (B) Quantification of fluorescence intensity for (A), measured as average intensity for a 300 μm section of the intestine in the midbody of each animal, as indicated. **p<0.01, ***p<0.001 considered significant. ns (p>0.05), not significant (Student’s t-test with Holm-Bonferroni correction). Mean ±SD of n = 10 animals. See figure supplement for replicate data. (C) Distribution of stages of eggs laid by adult hermaphrodites. Expression of nhr-46(+) in the intestine partially rescued the egg-laying defect of cyp-36A1(lf); nhr-46(lf); egl-9(lf) mutants. Neuronal nhr-46(+) expression fully rescued the egg-laying defect. Expression of nhr-46(+) in muscle did not rescue the egg-laying defect of cyp-36A1(lf); nhr-46(lf); egl-9(lf) mutants. ***p<0.001 considered significant. ns (p>0.05), not significant (Chi-square test with Holm-Bonferroni correction). Alleles used were egl-9(sa307), cyp-36A1(gk824636), nhr-46(n6126), nEx2713 [Pneurons::nhr-46(+)], nEx2715 [Pmuscle::nhr-46(+)], and nEx2864 [Pintestine::nhr-46(+)]. All strains contained the agIs219 (PT24B8.5::gfp) transgene.

https://doi.org/10.7554/eLife.36828.016
Figure 5—figure supplement 1
Download asset Open asset
Replicate data for PT24B8.5::gfp reporter expression.

Replicate data for Figure 5B. Quantification of n = 10 animals. **p<0.01, ***p<0.001 considered significant. ns (p>0.05), not significant (Student’s t-test with Holm-Bonferroni correction).

https://doi.org/10.7554/eLife.36828.017
Figure 6
Download asset Open asset
Model for the regulation of physiology and behavior by NHR-46 and CYP-36A1.

(A) The genetic pathway in which egl-9 inhibits hif-1, which activates cyp-36A1, which in turn inhibits nhr-46. (B) Model for how CYP-36A1 and NHR-46 function downstream of HIF-1. We suggest that CYP-36A1, which is transcriptionally upregulated by HIF-1, generates a hormone that binds NHR-46, thereby promoting transcriptional and physiological changes. See text for details.

https://doi.org/10.7554/eLife.36828.018

Tables

Key resources table
Reagent type
(species) or resource		DesignationSource or referenceIdentifiersAdditional
information
Strain, strain background
(Caenorhabditis elegans)		AU78Dennis KimagIs9 III
Strain, strain background
(C. elegans)		CB6088Jonathan Hodgkinegl-9(sa307) V hif-1(ia4) V
Strain, strain background
(C. elegans)		JT307Creg Darbyegl-9(sa307) V
Strain, strain background
(C. elegans)		MT20483Dengke Ma/Bob
Horvitz		nIs470 IV
Strain, strain background
(C. elegans)		MT22836this papercyp-36A1(gk824636) I ; egl-9(sa307) V
Strain, strain background
(C. elegans)		MT23218this papernIs682 X
Strain, strain background
(C. elegans)		MT24164this papercyp-36A1(gk824636) I ; agIs9 III
Strain, strain background
(C. elegans)		MT24165this papercyp-36A1(gk824636) I ; agIs9 III ;
egl-9(sa307) V
Strain, strain background
(C. elegans)		MT24166this paperagIs9 III ; egl-9(sa307) V hif-1(ia4) V
Strain, strain background
(C. elegans)		MT24167this paperagIs9 III ; egl-9(sa307) V
Strain, strain background
(C. elegans)		MT24177this paperagIs9 III ; hif-1(ia4) V
Strain, strain background
(C. elegans)		MT24179this papercyp-36A1(n5666) I ; (nIs470) IV
Strain, strain background
(C. elegans)		MT24520this papercyp-36A1(n5666) I ; (nIs470) IV ;
egl-9(n586ts) V ; nIs674
Strain, strain background
(C. elegans)		MT24622this papercyp-36A1(n5666) I ; (nIs470) IV ;
egl-9(n586ts) V
Strain, strain background
(C. elegans)		MT24684this papercyp-36A1(gk824636) I ; agIs9 III ;
nhr-46(n6126) IV ; egl-9(sa307) V ;
nEx2586
Strain, strain background
(C. elegans)		MT24690this paperagIs9 III ; otEx3165
Strain, strain background
(C. elegans)		MT24692this paperagIs9 III ; otEx3156
Strain, strain background
(C. elegans)		MT24911this papercyp-36A1(gk824636) I ; agIs9 III ;
nhr-46(n6126) IV ; egl-9(sa307) V
Strain, strain background
(C. elegans)		MT25047this paperagIs9 III ; nEx2699
Strain, strain background
(C. elegans)		MT25048this papercyp-36A1(gk824636) I ; agIs9 III ;
nEx2699
Strain, strain background
(C. elegans)		MT25050this papercyp-36A1(gk824636) I ; agIs9 III ;
otEx3156
Strain, strain background
(C. elegans)		MT25051this papercyp-36A1(gk824636) I ; agIs9 III ;
otEx3165
Strain, strain background
(C. elegans)		MT25104this papercyp-36A1(gk824636) I ; agIs9 III ;
nhr-46(n6126) IV ; egl-9(sa307) V ;
nEx2713
Strain, strain background
(C. elegans)		MT25107this papercyp-36A1(gk824636) I ; agIs9 III ;
nhr-46(n6126) IV ; egl-9(sa307) V ;
nEx2715
Strain, strain background
(C. elegans)		MT25196this paperagIs9 III ; nhr-46(n6125) IV
Strain, strain background
(C. elegans)		MT25197this papercyp-36A1(gk824636) I ; agIs9 III ;
nhr-46(n6125) IV ; egl-9(sa307) V
Strain, strain background
(C. elegans)		MT25211this papercyp-36A1(gk824636) I
Strain, strain background
(C. elegans)		MT25212this papercyp-36A1(gk824636) I ; egl-9(n586ts) V
Strain, strain background
(C. elegans)		MT25213this papercyp-36A1(gk824636) I ;
nhr-46(n6126) IV ; egl-9(n586ts) V
Strain, strain background
(C. elegans)		MT25214this papernhr-46(n6126) IV
Strain, strain background
(C. elegans)		MT25215this paperegl-9(n586ts) V
Strain, strain background
(C. elegans)		MT25596this papercyp-36A1(gk824636) I ; agIs9 III ;
egl-9(sa307) V ; nEx2853
Strain, strain background
(C. elegans)		MT25599this papercyp-36A1(gk824636) I ; agIs9 III ;
egl-9(sa307) V ; nEx2856
Strain, strain background
(C. elegans)		MT25601this papercyp-36A1(gk824636) I ; agIs9 III ;
egl-9(sa307) V ; nEx2849
Strain, strain background
(C. elegans)		MT25605this papercyp-36A1(gk824636) I ; agIs9 III ;
egl-9(sa307) V ; nEx2859
Strain, strain background
(C. elegans)		MT25606this paperagIs9 III ; nEx2860
Strain, strain background
(C. elegans)		MT25610this papercyp-36A1(gk824636) I ; agIs9 III ;
nhr-46(n6126) IV ; egl-9(sa307) V ; nEx2864
Strain, strain background
(C. elegans)		MT25611this papercyp-36A1(gk824636) I ; agIs9 III ; nEx2860
Strain, strain background
(C. elegans)		MT25627this paperagIs9 III ; nhr-46(n6126) IV
Strain, strain background
(C. elegans)		ZG31Huaqi Jiang/Jo Anne Powell-Coffmanhif-1(ia4) V

Additional files

Supplementary file 1

Genes that were significantly differentially expressed between wild type and egl-9(lf) and at least twofold downregulated in egl-9(lf) vs. wild type.

https://doi.org/10.7554/eLife.36828.019
Supplementary file 2

Genes that were significantly differentially expressed between wild type and egl-9(lf) and at least twofold upregulated in egl-9(lf) vs. wild type.

https://doi.org/10.7554/eLife.36828.020
Transparent reporting form
https://doi.org/10.7554/eLife.36828.021

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Corinne L Pender
  2. H Robert Horvitz
(2018)
Hypoxia-inducible factor cell non-autonomously regulates C. elegans stress responses and behavior via a nuclear receptor
eLife 7:e36828.
https://doi.org/10.7554/eLife.36828