Osório L et al. (2019), ISM1 regulates NODAL signaling and asymmetric o...

XB-ART-56012

J Cell Biol January 1, 2019; 218 (7): 2388-2402.

ISM1 regulates NODAL signaling and asymmetric organ morphogenesis during development.

Osório L , Wu X , Wang L , Jiang Z , Neideck C , Sheng G , Zhou Z .

Abstract
Isthmin1 (ISM1) was originally identified as a fibroblast group factor expressed in Xenopus laevis embryonic brain, but its biological functions remain unclear. The spatiotemporal distribution of ISM1, with high expression in the anterior primitive streak of the chick embryo and the anterior mesendoderm of the mouse embryo, suggested that ISM1 may regulate signaling by the NODAL subfamily of TGB-β cytokines that control embryo patterning. We report that ISM1 is an inhibitor of NODAL signaling. ISM1 has little effect on TGF-β1, ACTIVIN-A, or BMP4 signaling but specifically inhibits NODAL-induced phosphorylation of SMAD2. In line with this observation, ectopic ISM1 causes defective left-right asymmetry and abnormal heart positioning in chick embryos. Mechanistically, ISM1 interacts with NODAL ligand and type I receptor ACVR1B through its AMOP domain, which compromises the NODAL-ACVR1B interaction and down-regulates phosphorylation of SMAD2. Therefore, we identify ISM1 as an extracellular antagonist of NODAL and reveal a negative regulatory mechanism that provides greater plasticity for the fine-tuning of NODAL signaling.

PubMed ID: 31171630
PMC ID: PMC6605798
Article link: J Cell Biol

Species referenced: Xenopus laevis
Genes referenced: actb acvr1b acvr2b bmp4 cer1 ecd gdf1 ism1 lefty myc nodal pitx2 smad2 snai1 tdgf1.3 tsr1

Article Images: [+] show captions

	Figure 1. Mouse ISM1 is a secreted soluble protein that is glycosylated at asparagine residues 39 and 282. (A) Western blot of ISM1 in WCL and CM samples of HEK293T cells transiently transfected with mouse ISM1 and digested with PNGase F or Endo H as indicated. ISM1 protein is detected as a band of ∼70 kD in both WCL and CM without digestion. (B) HEK293T cells were transiently transfected with ISM1 or its N39Q, N282Q, and N39Q/N282Q point mutated forms. WCL samples were digested with or without PNGase F and subjected to Western blotting to detect ISM1. The size of intracellular ISM1 shifted from 70 kD to ∼65 and 60 kD following single or combined N point mutations, respectively. (C) Western blot of ISM1 in WCL and CM samples of HEK293T cells transiently transfected with wild-type ISM1 or its N39Q, N282Q, and N39Q/N282Q point mutated forms, in the presence or absence of 1 ng/ml tunicamycin. The protein molecular weight markers in kilodaltons are indicated by numbers on the left side of the Western blots.
	Figure 2. ISM1 modulates R-SMAD activation by members of the TGF-β superfamily in a ligand-dependent manner. (A) Western blot of pSMAD2 in WCL of serum-starved HEK293T cells treated with 40 ng/ml TGF-β1 or 40 ng/ml ACTIVIN-A in the presence of mock or ISM1 CM. For each panel, two biological replicates were performed. (B) Western blot of pSMAD2 in WCL of serum-starved HEK293T-CRIPTO cells treated with 100 ng/ml NODAL or 100 ng/ml GDF1 in the presence or absence of ISM1 CM as indicated. For each panel, two biological replicates were performed. (C) Western blot of pSMAD1/5/8 in WCL of serum-starved HEK293T cells treated with 50 ng/ml BMP4 in the presence or absence of ISM1 CM. For each panel, two biological replicates were performed. (D) Western blot of pSMAD2 in WCL of serum-starved HEK293T-CRIPTO cells treated with 100 ng/ml NODAL in the presence of 0, 100, or 200 ng/ml ISM1 protein. (E) Quantification of the intensity of pSMAD2 relative to β-actin in three biological replicates. The protein molecular weights in kilodaltons are indicated by numbers on the left side of the Western blots. Data represent mean ± SEM. ACTB, β-actin.
	Figure 3. Effect of ISM1 on SMAD2/FOXH1 transcriptional activity and gene expression of NODAL downstream targets. (A) Fold changes of SMAD2/FOXH1 transcriptional activity represented by relative luciferase units (RLU) in serum-starved HEK293T-CRIPTO cells treated with 100 ng/ml NODAL in the presence of 0, 100, or 200 ng/ml of rISM1 protein. Data from three independent experiments of dual luciferase reporter assay. (B) NODAL-induced transcriptional activity of SMAD2/FOXH1 in serum-starved HEK293T-CRIPTO cells in the presence of mock or increasing doses of ISM1 CM. The dosage of CM used is represented by the percentage of CM in the culture medium. Data from three independent experiments. Relative ISM1 protein levels in each treatment are shown by Western blot (WB). (C and D) Fold change of transcription of A3-luc reporter represented by RLU in serum-starved HEK293T cells treated with 40 ng/ml TGF-β1 (C) or 40 ng/ml ACTIVIN-A (D) in the presence of mock or increasing doses of ISM1 CM. Data from three independent experiments. Relative ISM1 protein levels in each treatment are shown by Western blot. Data represent mean ± SEM; statistical analyses were performed using unpaired, two-tailed t test, and statistically significant P values are indicated. (E) Schematic diagram of the experimental approach to test the inhibitory effect of ISM1 in NODAL-induced NODAL gene expression in chick embryos. When implanted on the right side of embryos at stages HH4–6, NODAL-soaked beads induce ectopic NODAL expression. The ability of NODAL to induce its own expression was tested in the presence of purified mouse rISM1 protein in embryos cultured ex ovo. (F) NODAL whole-mount in situ hybridization in embryos implanted with rNODAL + PBS– or rNODAL + rISM1–soaked beads. Arrowheads point to the ectopic NODAL expression in the right LPM, while asterisks indicate the endogenous NODAL expression on the left LPM. Embryos are shown in ventral view, and orientation along the AP and LR axes are indicated. Scale bar, 25 µm. (G) Percentage of embryos showing ectopic NODAL expression on the right LPM following implantation with beads soaked with rNODAL + PBS or rNODAL + rISM1.
	Figure 4. The inhibitory effects of ISM1, LEFTY1, and CER1 on NODAL signaling. (A) Serum-starved HEK293T-CRIPTO cells treated with 100 ng/ml rNODAL in the presence of mock, LEFTY1, ISM1, or CER1 CM at indicated dilutions. WCL samples were subjected to Western blotting analyses for pSMAD2. Total SMAD2/3 and β-actin were used as internal controls. The protein levels of LEFTY1, ISM1, and CER1 representing their relative concentrations in the CMs were analyzed by Western blotting using Flag antibodies. (B) Quantification of the intensity of pSMAD2 relative to SMAD2/3 in three biological replicates. Data represent mean ± SEM. The unpaired, two-tailed t test was used for statistical analyses. Statistically significant P values are indicated. (C) Dual luciferase reporter assay to determine the SMAD2/FOXH1 transcriptional activity in HEK293T-CRIPTO cells treated by 100 ng/ml rNODAL for 24 h, together with CM of mock, FLAG-LEFTY1, FLAG-ISM1, or FLAG-CER1 at the indicated dilutions. Data represent mean ± SEM of four independent experiments. P values of statistical analyses using unpaired, two-tailed t test are indicated in each individual experimental group.
	Figure 5. Interaction of ISM1 with NODAL ligand and its type I receptor ACVR1B. (A) HEK293T-CRIPTO cells were transiently transfected with ISM1 and FLAG-NODAL as indicated. CM samples were immunoprecipitated (IP) with anti-FLAG or anti-ISM1 antibodies and subjected to Western blot using antibodies against ISM1 and FLAG, respectively. Both precursor and mature forms of NODAL were pulled down from CM. (B) In situ PLA to detect ACVR1B-ISM1 complexes (red dots) in HeLa cells transiently expressing ACVR1B-HA plasmid at low dosage, in the presence or absence of ISM1 CM. Cells were counterstained with DAPI to visualize nuclei. The number of PLA signals per cell (lower left) and the percentage of cells that have PLA signals (lower right) were counted in a total of 200 cells. Scale bars, 20 µm. Data represent mean ± SEM. The unpaired, two-tailed t test was used for statistical analysis. (C) Schematic representation of the soluble forms of the ECD of ACVR1B and ACVR2B. ACVR1BECD was fused to either the HA epitope or to human IgG Fc fragment. ACVR2BECD was fused to MYC/His epitope. (D) HEK293T were transiently transfected with ISM1 and ACVR1BECD-Fc as indicated. CM samples were immunoprecipitated with protein G agarose beads and immunoblotted with anti-ISM1 and anti-IgG antibodies. (E) HEK293T were transiently transfected with ISM1 and ACVR2BECD-MYC as indicated. CM samples were immunoprecipitated with anti-ISM1 antibodies and subjected to Western blot with anti-MYC and anti-ISM1 antibodies. (F) WCL of HEK293T-CRIPTO cells transiently transfected with mock or ISM1 plasmids were precipitated with anti-ISM1 antibodies and analyzed by Western blot with antibodies against FLAG and ISM1. Arrowheads in E and F indicate the heavy chain of the IgG (50 kD). Protein molecular weight in kilodaltons is indicated by numbers on the left side of Western blots.
	Figure 6. AMOP domain is required for the function of ISM1 as an antagonist of NODAL signaling. (A) Diagram of domain-deleted ISM1 constructs lacking either TSR1 or AMOP (ΔTSR1 and ΔAMOP, respectively). The N-glycosylation sites at positions N39 and N282 are indicated. N39 is located in the N terminus, 10 amino acids downstream from the signal peptide (amino acids 1–29), and N282 is located in the region between TSR1 and AMOP domains (amino acids 260–285). (B) Western blot of ISM1 in WCL and CM samples of HEK293T cells transiently transfected with wild-type ISM1, ΔTSR1-ISM1, or ΔAMOP-ISM1. Samples were digested with PNGase F or Endo H before being subjected to Western blotting. (C) Mapping of NODAL-interacting domain of ISM1. HEK293T cells were transiently transfected with FLAG-NODAL together with either wild-type ISM1 or domain-deleted ISM1 mutants. CM samples were immunoprecipitated (IP) with anti-FLAG or anti-ISM1 antibodies and subjected to Western blotting with anti-ISM1 and anti-FLAG antibodies. (D) Mapping of ACVR1BECD-interacting domain of ISM1. HEK293T cells transiently expressing ACVR1BECD were cotransfected with wild-type ISM1 or domain-deleted ISM1 mutants. ACVR2BECD were also cotransfected with ACVR1BECD to potentially improve the interaction between ISM1 and ACVR1BECD. CM samples were immunoprecipitated with anti-ISM1 antibodies and immunoblotted with anti-HA antibodies. ISM1 interaction with ACVR1BECD is no longer observed when AMOP domain is absent. (E) Western blot of pSMAD2 in WCL of P19C6 cells treated with 100 ng/ml NODAL in the presence of mock, HIS-CER1, FLAG-LEFTY1, wild-type ISM1, ΔTSR1-ISM1, or ΔAMOP-ISM1 CM. Western blots for HIS (CER1), FLAG (LEFTY1), and ISM1 in CM samples is shown.
	Figure 7. ISM1 compromises formation of NODAL–ACVR1B complexes. (A) A mix of FLAG-NODAL CM and HIS-sCRIPTO (soluble form of CRIPTO containing its extracellular domain) CM was incubated with increasing concentrations of ISM1 CM. The mixed CM samples were immunoprecipitated (IP) with anti-FLAG antibodies and analyzed by Western blot with antibodies against HIS, FLAG, and ISM1. (B) HIS-sCRIPTO CM and ACRV1BECD-Fc CM were mixed with increasing concentrations of ISM1 CM. Protein G agarose beads were used to precipitate the ACRV1BECD-Fc complexes, followed by Western blotting with the antibodies against HIS, IgG, and ISM1. (C) FLAG-NODAL CM and ACRV1BECD-Fc CM were incubated together in the presence of increasing concentrations of ISM1 CM before immunoprecipitation with protein G agarose beads. The precipitates were immunoblotted with antibodies against FLAG, IgG, and ISM1. (D) In situ PLA to detect the NODAL–ACVR1B complexes (red dots) in HeLa cells transiently expressing a low dose of ACVR1B-HA plasmid and treated with FLAG-NODAL CM in the presence of either mock or ISM1 CM. Cells were counterstained with DAPI to visualize nuclei. The number of PLA signals per cell (lower left) and the percentage of cells that have PLA signals (lower right) were counted in a total of 200 cells. Scale bars, 20 µm. Data represent mean ± SEM. The unpaired, two-tailed t test was used for statistical analyses.
	Figure 8. Inhibition of NODAL signaling by ISM1 leads to defects in LR patterning in the chick embryo. (A) NODAL whole-mount in situ hybridization in ex ovo cultured chick embryos treated with mock or ISM1 CM. Arrowheads indicate NODAL expression in the left LPM. (B) Percentage of embryos with aberrant (reduced or absent) NODAL expression treated with either mock or ISM1 CM. (C) Relative NODAL transcription in the left half of embryos treated with mock or ISM1 CM determined by qPCR. (D) CER1 whole-mount in situ hybridization in ex ovo cultured chick embryos treated with mock or ISM1 CM. Arrowheads indicate CER1 expression in the left LPM. (E) Percentage of embryos showing reduced CER1 expression in the presence or absence of ISM1 CM. (F and G) Relative levels of CER1 (F) and PITX2 (G) transcripts in the left half of embryos in the presence or absence of ISM1. (H) Whole-mount in situ hybridization of LEFTY1 in ex ovo cultured chick embryos treated with mock or ISM1 CM. No obvious difference in LEFTY1 expression along the embryonic midline was observed. (I) Percentage of embryos with aberrant LEFTY1 expression after culture with mock or ISM1 CM. (J) Percentage of embryos showing a right-sided heart position after mock or ISM1 CM treatment. (K) SNAI whole-mount in situ hybridization in ex ovo cultured chick embryos treated with mock or ISM1 CM. Arrowheads indicate SNAI expression in the right LPM. (L) Percentage of embryos showing abnormal SNAI1 expression in the presence or absence of ISM1. For qPCR analyses, 10–12 embryos were pooled together for RNA purification. Two independent biological replicates were included. Three replicates of qPCR reactions were performed for each independent sample. Data represent mean ± SEM. The unpaired, two-tailed t test was used for statistical analysis. Statistically significant P values are indicated. All the embryos are shown in ventral view, and orientation along the AP and LR axes is indicated. Scale bars, 25 µm.
	Figure 9. Working model for the inhibitory effect of ISM1 on NODAL signaling. ISM1, a newly identified extracellular antagonist of NODAL signaling, interacts with both NODAL ligand and receptor complex ACVR2A–ACVR1B. ISM1 does not interact with CRIPTO but compromises the formation of NODAL–ACVR1B complex, therefore negatively regulating among DAN/CER, LEFTY1, and ISM1 on NODAL signaling as presented in the diagram. The table summarizes the binding profile of three NODAL antagonists with NODAL signaling components and their relative inhibitory strength.

References [+] :

Adams, The thrombospondins. 2012, Pubmed