Figure 1.
Interference with RA transport into the nucleus is sufficient to induce transcriptional repression of genes downstream of RARα.
(A) WB analysis showing that HME1 cells express RARα (left). Immunoprecipitation (IP) followed by WB showing that HME1 express CRABP2. IP of Hela cells transiently transfected with CRABP2-V5 served as positive control (right). (B) Transcriptional activation of two RA-responsive genes, RARβ2 and CRBP1, in response to RA (72 h) demonstrates a functional RA-RARα signaling in HME1 cells. (C) Transient HME1 transfection with wild type CRABP2-V5, followed by immunocytochemistry with anti-V5 (red) and DAPI nuclear staining (blue), shows that exogenous CRABP2-V5 can translocate from the cytoplasm into the nucleus after treatment with 0.1 μM RA for 30 min. (left). In contrast, exogenous CRABP2-KRK-V5 mutant carrying a mutated nuclear localization signal (NLS) is not able to enter the nucleus under the same conditions (right). (D) WB analysis showing the expression of the CRABP2-KRK-V5 protein in the HME1-derived clone KRK-15, but not in the HME1 control clone EV7. In vitro transcribed and translated CRABP2-KRK-V5 was used as a positive control (left). Both KRK-15 and EV-7 cells express RARα (right). (E) Both RARβ2 and CRBP1 transcription are significantly less inducible by RA (72 h) in KRK-15 cells relative to the control EV7 cells.
Figure 2.
Evidence of chromatin repression at RA-responsive genes downstream of RARα consequent to CRABP2 knock-down.
(A) The HME1-derived stable clones Si-CRABP2-A6 and Si-CRABP2-C6, carrying two distinct CRABP2-targeting shRNA sequences (CARBP2-A and CRABP2-C, respectively), display a significant decrease in CRABP2 transcript relative to the control clone Mock-13 (left). The level of RARα expression is similar in Si-CRABP2-A6, Si-CRABP2-C6 and the control clone Mock13 (right). (B) Both RARβ2 (left) and CRBP1 (right) are significantly less inducible by RA (72 h) in Si-CRABP2-A6 and Si-CRABP2-C6 clones relative to the control Mock13 clone. (C) qChIP analysis with anti-acetyl histone H4 (Ac-H4) showing that RARβ2 (top) and CRBP1 (bottom) chromatin of both Si-CRABP2-A6 and Si-CRABP2-C6 is marked by H4 hypoacetylation at the RARE-containing regulatory regions relative to the control clone Mock-13. (D) RA-induced RARβ2 (top) and CRBP1 (bottom) transcription can be restored in Si-CRABP2-A6 cells by treatment with the HDAC inhibitor TSA. (E) qChIP with anti-Polymerase II (Pol II) showing decreased occupancy of Pol II at the RARE-containing regions of both RARβ2 (top) and CRBP1 (bottom) in Si-CRABP2-A6 and Si-CRABP2-C6.
Figure 3.
Hampering CRABP2 function in HME1 cells leads to biological phenotypes that reflect homozygosis for epigenetically silent RARβ2 and CRBP1 alleles.
(A) HME1 cells knocked down for RARβ2 (Si-RARβ2) develop resistance to RA growth-inhibitory action (left). The HME1 clones KRK-15, Si-CRABP2-A6 and Si-CRABP2-C6, with impaired CRABP2 function, show a significantly higher fraction of RA-resistant cells than the cognate control clones EV7 and Mock-13 (right). (B) Scheme showing that RA-resistance is expected only in cells homozygous for RARβ2 alleles non-permissive for transcription (np/np), but not in cells either homozygous for permissive RARβ2 alleles (p/p), or heterozygous for permissive and non-permissive RARβ2 alleles (p/np). (C) HME1 cells knocked down for CRBP1 (Si-CRBP1, top left) are unable to form hollow, polarized acini in three-dimensional (3D) culture, as shown by confocal fluorescence microscopy (nuclei are visualized in blue, integrin in green, and the Golgi apparatus in red). HME1 clones with an impaired CRABP2 function (KRK15, Si-CRABP2-A6 and Si-CRABP2-C6) are also unable of proper acinar morphogenesis. (D) Scheme showing that impaired acinar morphogenesis is expected only in cells that have developed homozygosis for non-permissive CRBP1 alleles.
Figure 4.
Evidence of CpG hypermethylation corroborates the occurrence of epigenetic silencing at both RARβ2 and CRBP1 consequent to impaired CRABP2 function.
(A) Treatment with the demethylating agent 5-Aza can restore RA-induced transcription from repressed RARβ2 and CRBP1 chromatin in Si-CRABP2-A6 cells (left and right, respectively). (B) Quantitative MSP detecting methylated (M) alleles shows hypermethylation of RARβ2 (left) and CRBP1 (right) CpG-rich regulatory regions in Si-CRABP2-A6 and Si-CRABP2-C6 clones.
Figure 5.
Derangement of CRABP2 function exerts a chromatin repression effect branching downstream of RARβ2.
(A) RA-induced transcription of CYP26A1 is significantly downregulated in both HME1 cells knocked down for CRABP2 (Si-CRABP2-A6 and Si-CRABP2-C6 clones, left), and HME1 cells carrying the CRABP2-KRK mutant (KRK-15 clone, right) relative to control cells (Mock 13 and EV7 clones, respectively). (B) qChIP analysis with anti-acetyl histone H4 showing that CYP26A1 chromatin in Si-CRABP2-A6 and Si-CRABP2-C6 clones is marked by a significant H4 hypoacetylation of a region encompassing the CYP26A1 proximal RARE (left). RA-induced CYP16A1 transcription can be restored in Si-CRABP2-A6 by treatment with the HDAC inhibitor TSA for 72 h (right). (C) Bisulfite sequencing showing that HME1, like the CYP26A1-positive cell line T47D, is unmethylated in the proximal RARE-containing CpG island, while the CYP26A1-negative cell line MDA-MB-231 is fully methylated (left). Quantitative MSP with primers recognizing only methylated (M) alleles can detect methylation in MDA-MB-231, but not in T47D or HME1 cells (right). (D) Quantitative MSP analysis showing hypermethylation of CYP2A1 proximal CpG island in Si-CRABP2-A6 and Si-CRABP2-C6 clones (left). RA-induced CYP26A1 transcription can be efficiently restored in Si-CRABP2-A6 cells by treatment with the demethylating agent 5-Aza (right). (E) CYP26A1 transcription can still be induced by RA in HME1 cells knock down for CRBP1 (Si-CRBP1). Thus, CYP26A1 epigenetic downregulation is not consequent to CRBP1 epigenetic silencing.
Figure 6.
Epigenetic repression of a pleiotropic gene network as a consequence of a defective RA transport onto RARα.
(A) RA transport onto RARα by CRABP2 enables the transcriptional activation of a RA-responsive gene network involved in retinol (ROH)-RA metabolism, control of cell growth, and morphogenesis. (B) Interference with CRABP2-mediated RA transport onto RARα leads to epigenetic repression of this gene network, with pleiotropic biological outcome.