1. Cancer Biology
Download icon

Inhibiting IRE1α-endonuclease activity decreases tumor burden in a mouse model for hepatocellular carcinoma

  1. Nataša Pavlović
  2. Carlemi Calitz
  3. Kess Thanapirom
  4. Guiseppe Mazza
  5. Krista Rombouts
  6. Pär Gerwins
  7. Femke Heindryckx  Is a corresponding author
  1. Department of Medical Cell Biology, Uppsala University, Sweden
  2. Regenerative Medicine & Fibrosis Group, Institute for Liver and Digestive Health, University College London, United Kingdom
  3. Department of Radiology, Uppsala University Hospital, Sweden
Research Article
Cite this article as: eLife 2020;9:e55865 doi: 10.7554/eLife.55865
14 figures, 3 tables, 2 data sets and 3 additional files

Figures

Inhibiting IRE1α reduces tumor burden in vivo.

(A) Representative images of liver slides stained with hematoxylin and eosin (H and E), Sirius red and αSMA-antibodies. (B) tumor burden of mice with DEN-induced HCC treated with 4μ8C or vehicle-treated controls. (C) Quantification of percentage of collagen and (D) αSMA on liver slides. (E) mRNA expression of Pcna in liver tissue from mice with HCC treated with 4μ8C (F). Heatmap showing protein expression levels in healthy liver, DEN-induced HCC and DEN-induced HCC treated with 4μ8C from three biological replicates per group. p-Values were calculated via the Student's T-test, scale bars = 120 μm.

Figure 2 with 2 supplements
Increased expression of ER-stress markers in mice with HCC.

(A) mRNA expression of ER-stress markers Edem1, Ero1b, Grp94, Herp, Atf4, Eif2ak3, Ddit3, and Hspa5 in liver tissue from healthy mice; and tumor tissue and surrounding non-tumoral tissue from mice with DEN-induced HCC. (B) Hspa5-mRNA and (C) protein expression of BIP in murine liver tissue. (D) Ratio of spliced to unspliced XBP1 in liver tissue from healthy mice; and tumor tissue and surrounding non-tumoral tissue from mice with DEN-induced HCC, treated with 4μ8C. (E) Representative western blot image of spliced and unspliced XBP1 protein and vinculin in healthy liver, DEN-induced HCC and DEN-induced HCC treated with 4μ8C. (F) quantification of spliced and unspliced XBP1, normalized to total vinculin levels. (G) Ratio of spliced to unspliced XBP1 protein levels. (H) Representative images and (I) quantification of liver tissue sections stained with antibodies against spliced XBP1. p-Values were calculated via the Student's T-test with five biological replicates per group. Scale bars = 120 μm.

Figure 2—figure supplement 1
Activation of the unfolded protein response is mainly located in the stroma of mice with HCC.

Liver tissue from mice with DEN-induced HCC, stained with αSMA-antibodies and co-stained with antibodies against (A) spliced XBP1, (B) total XBP1, (C) IRE1α (D) phopho-IRE1α, and (E) BIP. Scale bars = 50 μm.

Figure 2—figure supplement 2
Expression of ER-stress markers is localized in close vicinity to αSMA.

Immunofluorescent images from tissue from mice with DEN-induced HCC, stained with αSMA-antibodies and co-stained with antibodies against (A) spliced XBP1, (B) total XBP1, (C) IRE1α, (D) phopho-IRE1α, and (E) BIP. (F) Immunofluorescent image from DEN-induced HCC stained with antibodies against spliced XBP1.

Activation of the unfolded protein response pathway is increased in patients with fibrotic HCC.

(A) Heat map showing gene-set enrichment analysis results from samples from fibrous HCC versus non-fibrous HCC. (C) Immunohistochemically stained liver biopsies from HCC-patients obtained from the human protein atlas, using antibodies against IRE1α-mediated actors of the unfolded protein response: WIPI1, SHC1, PPP2R5B, and BIP. (D) Kaplan-Meier survival curves of HCC-patients with high or low expression of WIPI1, SHC1, PPP2R5B, and BIP. p-Values were calculated via a Log-Rank test.

Figure 4 with 1 supplement
Tumor cells secrete factors that induce ER-stress in stellate cells, which contributes to their activation.

(A) mRNA-expression of ER-stress markers ATF6, ATF4, EIF2AK3, GADD34, EDEM1, DDIT3 and HSPA5, in stellate cells (LX2) co-cultured with cancer cells (HepG2 or Huh7) and treated with 4μ8C or control. (B) Detection of spliced (XBP1s) and unspliced XBP1 (XBP1u) via qPCR and (C) via digestion of the XBP1u-RT-qPCR product by Pst-I and subsequent visualization by separation of on agarose gel. (D) Quantified ratio of spliced and unspliced measured on agarose gel after digestion by Pst-I (E) mRNA expression of HSPA5 in stellate cells (LX2) co-cultured with cancer cells (HepG2 or Huh7) and treated with 4μ8C or control. (F) protein expression of p-IRE1α and vinculin in stellate cells (LX2) co-cultured with cancer cells (HepG2 or Huh7) in transwell assays and treated with 4μ8C or control. (G) mRNA-expression of stellate cell activation markers ACTA2 and (H) collagen in LX2-cells co-cultured with HepG2 or Huh7-cells and treated with or without 4μ8C. p-Values were calculated via ANOVA with 10 biological replicates per group.

Figure 4—figure supplement 1
Secretion of TGFβ by tumor cells activates stellate cells and induces ER-stress.

(A) Concentration of TGFβ in medium from tumor cells (HepG2 or Huh7) grown in mono-culture or co-cultured with LX2-stellate cells, treated with 4μ8C or control. (B) Concentration of TGFβ in medium from liver scaffolds engrafted with stellate cells (C) (LX2) and tumor cells (HepG2) treated with 4μ8C or control. mRNA-expression of the ER-stress markers DDIT3, (D) spliced XBP1, (E) unspliced XBP1 and (F) HSPA5 in hepatic stellate cells (LX2) grown as mono-culture or in co-cultures with the cancer cell lines HepG2 and Huh7 treated with the TGFβ receptor inhibitor SB-431541 or control. (G) mRNA-expression of stellate cell activation markers ACTA2 and (H) collagen in LX2-cells grown with HepG2 or Huh7-cells and treated with SB-431541 or control. p-Values were calculated via the Student's T-test from seven biological replicates per group.

Inhibiting IRE1α decreases stellate cell activation in human liver 3D scaffolds engrafted with stellate cells and tumor cells.

(A) Representative images of H and E and Sirius red stained slides of decellularized human liver scaffolds engrafted with LX2 stellate cells and HepG2-tumor cells treated with 4μ8C or control. (B) Quantification of collagen-stained area fraction of liver scaffolds engrafted with LX2 stellate cells and HepG2-tumor cells treated with 4μ8C or control. (C) mRNA-expression of the stellate cells activation marker collagen and ER-stress markers HSPA5, spliced XBP-1 (XBP1-S), and DDIT3 in liver scaffolds engrafted with stellate cells (LX2) and cancer cells (HepG2), treated with 4μ8C or control. p-Values were calculated via ANOVA from three biological replicates per group, scale bars = 100 μm.

Inhibition of IRE1α decreases tumor cell proliferation.

(A) PCNA mRNA-expression of HepG2 or Huh7-cells grown with LX2-cells in transwell inserts and treated with the IRE1α-inhibitor 4μ8C or control. (B) Relative cell number of LX2 and HepG2 or (C) LX2 and Huh7-cells treated with 4μ8C or control. (D) Representative images of tumor cells (HepG2 or Huh7) and LX2-stellate cells stained with antibodies against the HCC-marker EPCAM and the proliferation marker KI67. (E) Cell proliferation of HepG2 or HepG2+LX2 spheroids and (F) Huh7 or Huh7+LX2 spheroids treated with 4μ8C or control. p-Values were calculated via the Student's T-test from nine biological replicates per group, scale bars = 50 μm.

Inhibition of IRE1α decreases cell proliferation and improves liver function in human liver scaffolds engrafted with stellate cells and tumor cells.

(A) PCNA and (B) HNF4A expression of human liver scaffolds engrafted with HepG2-tumor cells and LX2-stellate cells, treated with 4μ8C or control. (C) Representative images of tumor cells (HepG2) and LX2-stellate cells stained with antibodies against the HCC-marker EPCAM and the proliferation marker KI67. p-Values were calculated via ANOVA on three biological replicates per group, scale bars = 100 μm.

Figure 8 with 1 supplement
Inhibition of IRE1α decreases cell migration.

(A) mRNA-expression of pro-metastatic markers MMP9 and (B) MMP1 in HepG2 and Huh7-cells co-cultured with LX2-cells and treated with 4μ8C or control. (C) Scratch wound on HepG2-cells and LX2-cells treated with 4μ8C or control. (D) Images of Cell Tracker stained HepG2-cells (Green) and LX2-cells (Red) invading the scratch area. (E) Quantification of wound size in HepG2-cells and LX2-cells treated with 4μ8C or control. (F) Number of HepG2-cells and LX2-cells invading the scratch wound after 24 hr in co-cultures and (G) mono-cultures. p-Values were calculated via the Student's T-test from 10 biological replicates per group (panel A and B) or six biological replicates per group (panel E-G), scale bars = 120 μm.

Figure 8—figure supplement 1
Inhibiting IRE1α decreases chemotaxis.

(A) Migration plots of LX2-cells co-cultured with HepG2-cells exposed to an FBS-gradient (increasing towards the right) and treated with control or (B) 4μ8C (C) Quantification of total migration and (D) directional migration of LX2-cells (co-cultured with HepG2-cells) toward an FBS-gradient with or without 4μ8C. (E) Migration plots of HepG2-cells co-cultured with LX2-stellate cells and exposed to an FBS-gradient and treated with control or (F) 4μ8C. (G) Quantification of total migration and (H) directional migration of HepG2-cells (co-cultured with LX2-cells) towards an FBS-gradient with or without 4μ8C. p-Values were calculated via the Student's T-test from three biological replicates per group. Red lines indicate migration toward the gradient, while black lines indicate migration away from the gradient.

Figure 9 with 1 supplement
Silencing IRE1α in LX2-cells mimics 4μ8C.

(A) ERN1-mRNA-expression of LX2-cells transfected with IRE1α-siRNA (si-IRE1α) or mock-transfected (Scr) (B) PCNA-mRNA-expression of HepG2-cells co-cultured with IRE1α-silenced LX2-cells or controls (C). Relative cell numbers in co-cultures of HepG2-cells and IRE1α-silenced LX2-cells or controls. (D) ERN1-mRNA-expression of HepG2- and Huh7-cells transfected with IRE1α-siRNA (si-IRE1α) or mock-transfected (Scr). (E) Relative cell numbers in co-cultures LX2-cells or and si-RNE. Transfected HepG2 or Huh7 cells or mock-transfected controls (Scr). p-Values were calculated via the Student's T-test from three biological replicates per group (panel A, B and D) or six biological replicates (panel C and E).

Figure 9—figure supplement 1
Proliferation and migration after silencing IRE1α in LX2-cells.

(A) Proliferation of spheroids of HepG2-cells and IRE1α-silenced LX2-cells or controls (B) Images and (C) quantification of αSMA-stained spheroids with HepG2-cells and IRE1α-silenced LX2-cells or controls. (D) Images and (E) quantification of scratch wound of HepG2-cells co-cultured with IRE1α-silenced LX2-cells or controls. p-Values were calculated via the Student's T-test from three biological replicates per group, scale bars = 50 μm (E) or 120 μm (G).

Inhibiting IRE1α alters generation of ROS.

(A) intracellular ROS-levels in LX2, HepG2, and Huh7 cells treated with 50 μM 4μ8C, 100 μM 4μ8C or controls. (B) intracellular ROS-levels in LX2, HepG2 and Huh7 cells transfected with IRE1α-siRNA (si-IRE1α) or mock-transfected (Scr). p-Values were calculated via the Student's T-test from three biological replicates per group.

Author response image 1
Effect of ER-stress on ROS production.

(A) the ER-stress inducer increases ROS levels in HepG2 and Huh7 cells. (B) The ROS-scavenging effect of 4u8C increases over time.

Author response image 2
Effect of silencing IRE1a in different co-culture conditions.

(A) LX2-cells transfected with si-RNA targeting IRE1a or mock-transfected (Scr) in mono-culture or co-culture with tumor cells (HepG2 and Huh7). (B) HepG2 or Huh7-cells transfected with si-RNA targeting IRE1a or mock-transfected (Scr) in mono-culture or co-culture with LX2-stellate cells.

Author response image 3
Visible bands in lane 2 and 4, which corresponds to LX2-monocultures and LX2+Hep2 co-cultures treated with 4u8C.
Author response image 4
Protein levels of IRE1a in mock transfected and si-IRE1a transfected Huh7 cells.

Tables

Table 1
A proteomics array using the Olink Mouse Exploratory assay – source data Figure 1F.
CTLDenDEN+4 u8cStatistical significance
Protein nameBiological processMeanSt. DevAverageSt. DevAverageSt. DevDEN vs CtrlDEN vs 4 u8CCtrl vs 4 u8c
ClmpNot prognostic in HCC1.680.142.971.002.480.64*
Yes1HCC promotor7.110.297.510.207.440.19*
Foxo1Tumor suppressor4.150.064.120.733.870.49
Pla2g4aHCC promotor3.420.385.701.365.040.80**
Prdx5HCC promotor7.370.497.230.266.670.34*
TgfaTumor growth factor5.360.526.810.646.930.88**
EpoUnfavorable prognotic marker3.200.343.710.353.370.33
Axin1HCC promotor4.240.384.800.374.390.35
FstHCC promotor5.870.318.040.737.500.71**
NadkNot prognostic in HCC10.100.1310.140.1810.300.27
Snap29Not prognostic in HCC7.700.327.870.327.620.30
S100a4HCC promotor2.730.747.010.626.850.97**
KitlgMetastasis2.480.423.740.623.310.98*
Gfra1HCC promotor4.400.355.070.404.920.39*
Ppp1r2Not prognostic in HCC4.370.164.860.464.470.43
Cyr61HCC promotor2.400.534.141.643.131.22*
AhrNot prognostic in HCC6.950.467.680.747.380.64
Ccl2HCC promotor4.590.589.692.048.93.**
QdprNot prognostic in HCC7.710.117.720.147.540.15
FasHCC promotor8.660.188.830.188.700.18
Riox2HCC promotor7.100.157.710.387.590.14**
EpcamHCC promotor1.560.333.161.143.270.89*
Ccl3Prognostic marker1.490.394.421.863.731.07**
Crim1HCC promotor2.460.283.711.093.210.56**
HgfTumor growth factor6.690.357.941.017.410.71*
Sez6l2HCC promotor−0.290.150.610.530.190.29*
Il1aInflammation and fibrosis6.650.518.350.657.620.54**
Ddah1HCC promotor8.040.228.180.057.840.18*
Acvrl1Not prognostic in HCC2.090.183.441.312.810.47
Cxcl9Inflammation and fibrosis3.680.867.711.686.651.58**
Map2k6Not prognostic in HCC7.750.157.980.417.880.28
Casp3Tumor surrpressor9.220.199.740.359.430.26
PdgfbTumor growth factor3.520.314.961.273.970.40*
Igsf3Unfavorable prognotic marker3.120.284.190.823.640.72
Cxcl1HCC promotor3.770.405.740.785.060.51**
Pak4HCC promotor3.470.424.390.683.930.54
LplNot prognostic in HCC1.660.402.440.452.020.60
Dctn2Unfavorable prognotic marker5.481.315.670.704.980.55
Ntf3Not prognostic in HCC2.160.272.800.712.270.40
Tnfsf12HCC promotor5.280.356.000.765.590.62
Ccl20Unfavorable prognotic marker5.200.345.920.815.530.66
Fli1HCC promotor1.910.223.731.382.980.83
Tpp1Unfavorable prognotic marker3.670.384.240.643.730.50
Parp1Unfavorable prognotic marker10.300.7210.930.4910.510.62
Table 2
Genes the contributed to the core-enrichment of the GSEA.
ProbeDescriptionRank Gene listRank Metric scoreCore enrichmentUPR branch
ASNSAsparagine synthetase (glutamine-hydrolyzing) [Source:HGNC Symbol;Acc:HGNC:753]2070.940YesPerk
PPP2R5BProtein phosphatase two regulatory subunit B'beta [Source:HGNC Symbol;Acc:HGNC:9310]4230.821YesIre1a
CCL2C-C motif chemokine ligand 2 [Source:HGNC Symbol;Acc:HGNC:10618]8470.689YesIre1a and Perk
EXOSC9Exosome component 9 [Source:HGNC Symbol;Acc:HGNC:9137]10040.654YesIre1a and Perk
WIPI1WD repeat domain, phosphoinositide interacting 1 [Source:HGNC Symbol;Acc:HGNC:25471]10220.649YesIre1a
KDELR3KDEL endoplasmic reticulum protein retention receptor 3 [Source:HGNC Symbol;Acc:HGNC:6306]11060.635YesIre1a
SHC1SHC adaptor protein 1 [Source:HGNC Symbol;Acc:HGNC:10840]26910.432YesIre1a
TPP1Tripeptidyl peptidase 1 [Source:HGNC Symbol;Acc:HGNC:2073]28840.414YesIre1a
HDGFHeparin binding growth factor [Source:HGNC Symbol;Acc:HGNC:4856]32350.386YesIre1a
TLN1Talin 1 [Source:HGNC Symbol;Acc:HGNC:11845]32640.384YesIre1a
EXTL3Exostosin like glycosyltransferase 3 [Source:HGNC Symbol;Acc:HGNC:3518]34880.365YesIre1a
TSPYL2TSPY like 2 [Source:HGNC Symbol;Acc:HGNC:24358]36800.350YesIre1a
MBTPS1Membrane-bound transcription factor peptidase, site 1 [Source:HGNC Symbol;Acc:HGNC:15456]39960.327YesAtf6
PDIA5Protein disulfide isomerase family A member 5 [Source:HGNC Symbol;Acc:HGNC:24811]45300.294YesIre1a
DCTN1Dynactin subunit 1 [Source:HGNC Symbol;Acc:HGNC:2711]46380.287YesIre1a
DNAJC3DnaJ heat-shock protein family (Hsp40) member C3 [Source:HGNC Symbol;Acc:HGNC:9439]47610.281YesIre1a
SULT1A4Sulfotransferase family 1A member 4 [Source:HGNC Symbol;Acc:HGNC:30004]49380.272YesIre1a
PARNPoly(A)-specific ribonuclease [Source:HGNC Symbol;Acc:HGNC:8609]50370.266YesPerk
ADD1Adducin 1 [Source:HGNC Symbol;Acc:HGNC:243]53750.250YesIre1a
ERN1Endoplasmic reticulum to nucleus signaling 1 [Source:HGNC Symbol;Acc:HGNC:3449]54110.248YesIre1a
Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Strain, strain background (Mus musculus)Sv129 miceTaconic129S6HCC mouse model, Heindryckx et al., 2010; Heindryckx et al., 2012
Cell line (Homo sapiens)HepG2ATCCHB-8065
Cell line (Homo sapiens)Huh7Gifted, Karolinska institute
Cell line (Homo sapiens)LX2Sigma-AldrichSCC064
Transfected construct (human)si-IRE1αThermoFishers2004320,1–1 µM
Transfected construct (human)Si-Ctrl; ScrThermoFisher43908430,1–1 µM
AntibodyKI67 (rat monoclonal)eBioscienceSolA151:100
AntibodyEPCAM (rabbit polyclonal)Abcamab719161:100
AntibodySpliced XBP1 (goat monoclonal)AbcamAb855461:50
AntibodyTotal XBP1 (Rabbit polyclonal)AbcamAb371525 µg/ml
AntibodyIRE1a (rabbit polyclonal)AbcamAb370731 µg/ml
Antibodyp-IRE1 (rabbit polyclonal)AbNovaPAB124351:100
AntibodyαSMA (Rabbit Polyclonal)ThermoFisher7104871:200
AntibodyαSMA (Goat monocolonal)AbcamAb210271–2 µg/ml
AntibodyBIP (goat polyclonal)AbcamAb210271 µg/ml
AntibodyVinculin (Mouse monoclonal)ThermoFisher14-9777-821–5 µg/ml
Peptide, recombinant proteinPst-IThermoFisherER0615
Commercial assay or kitPierce BCA-protein assay kitThermoFisher233225
Commercial assay or kitEZNA RNA isolation Kit IIVWRR6934-02
Commercial assay or kitRNeasy Universal Mini KitQiagen73404
Commercial assay or kitDiva Decloacker solutionBiocareDV2004
Commercial assay or kitDCFDA - Cellular ROS Detection Assay KitAbcamab113851
Chemical compound, drugN-Nitrosodiethylamine, DENSigma-Aldrich1002877809
Chemical compound, drug4μ8CSigma-AldrichSML0949-25MGHeindryckx et al., 2016
Chemical compound, drugSB-431541, TGF-ß receptor inhibitorTocris161410 μM
Chemical compound, drugResazurinSigma-AldrichR7017-1G1:80 dilution
Commercial assay or kitIngenio electroporation solutionMirus Bio LLCMIR50114Ice-cold
Commercial assay or kitCellTracker Red CMTPXThermoFisherC345521 μM
Commercial assay or kitCellTracker Green CMFDAThermoFisherC29251 μM
Other12-well CorningCostar TranswellplatesSigma-Aldrich3460Calitz et al., 2020
OtherCorning CostarUltra-Low attachment 96-well platesSigma-AldrichCLS3471Calitz et al., 2019
OtherCellDirectorGradienTech11-001-10Fuchs et al., 2020

Data availability

Proteomics data has been deposited in Dryad with the following DOI: https://doi.org/10.5061/dryad.6wwpzgmv2.

The following data sets were generated
    1. Heindryckx F
    (2020) Dryad Digital Repository
    Protein expression of hepatocellular carcinoma in a fibrotic liver in mice.
    https://doi.org/10.5061/dryad.6wwpzgmv2
The following previously published data sets were used
    1. Seok JY
    2. Na DC
    3. Woo HG
    4. Roncalli M
    5. Kwon SM
    6. Yoo JE
    7. Ahn EY
    8. Kim GI
    9. Choi J
    10. Kim YB
    11. Park YN
    (2020) NCBI Gene Expression Omnibus
    ID GSE31370. Fibrous stromal component in hepatocellular carcinoma reveals a cholangiocarcinoma-like gene expression trait and epithelial-mesenchymal transition.

Additional files

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)

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

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