This is a correction to: Janine Kövilein, Adam Sorbie, Sevana Khaloian, Vanessa Küntzel, Miriam von Stern, Mohamed Ahmed, Sebastian Jarosch, Marianne Remke, Amira Metwaly, Elena M Reuss, Dirk H Busch, Matthieu Allez, Katja Steiger, Barbara Schraml, Olivia I Coleman, Dirk Haller, Susceptibility to inflammatory bowel diseases promotes invasive carcinomas in a murine model of ATF6-driven colon cancer, Journal of Crohn’s and Colitis, Volume 19, Issue 7, July 2025, jjaf102, https://doi.org/10.1093/ecco-jcc/jjaf102
There were errors in the details of Figures 1-4 in the original publication of the paper. These have been corrected in the article as follow.
Figure 1.
Interleukin-10 deficiency increases tumor susceptibility in nATF6IEC mice and drives formation of colonic invasive carcinomas. A) Survival curves of nATF6IEC;Il10-/- mice housed under specific pathogen-free (SPF) conditions. B) Colonic tumor incidence of tg/wt;Il10-/- mice categorized into responder (R = mice that developed tumors) and non-responder (NR = mice that did not show any tumors) at week 5, 12 and 15+. C) Colonic tumor numbers of nATF6IEC;Il10-/- mice at week 5, 12, and 15+. D) Dysplasia score resulting from histological scoring of susceptible (proximal) colon tissue of nATF6IEC;Il10-/- mice for mucosal architecture and atypia (0-3) at week 5, 12, and 15+. E) Representative images of H&E staining of susceptible colon tissue at week 12 (scale bars: 200 μm) and corresponding higher magnifications (rectangles; scale bars: 100 μm). F) Tumor incidence of 0.5 × 0.5 cm colon tissue sites identifying a tumor susceptible (colon tissue site 1 to 9) and non-susceptible (colon tissue site from 10 onwards) region with representative macroscopic images of tg/wt;Il10-/- mice; asterisks indicate tumors. G) Ki67 and PAS-AB staining of proximal or susceptible colon tissue of fl/fl;Il10-/- and NR and R tg/wt;Il10-/- mice (scale bars: 100 μm) and corresponding higher magnifications (rectangles; scale bars: 100 μm). Respective quantifications of each staining were analyzed in tumor susceptible colon tissue. H) H-score of nuclear (NUC) and cytoplasmic (CYT) ATF6 expression in colon biopsies of IBD patients stained immunohistochemically. Representative images of ATF6 expression in active and inactive disease as well as ATF6-high and ATF6-low expression in colon tissue (scale bars: 100 μm) and corresponding higher magnifications. wk = week; NT = non-tumor; NR = non-responder; R = responder; NUC = nuclear; CYT = cytoplasmic.
Figure 1, errors were emended in the genotypes within the legend of pane A; In panel B, the the mouse number (14) should stand on top of fl/fl and the label “15+wk” was misaligned; in panel E “3 (invasice cancer)” is corrected to read: “3 (invasive cancer)”; labelling of x-axis is realigned for panel F. Figure 1 should read:
Figure 2, in panels B and F, “Il10” is corrected to “IL10”; in panel E “CD4+ T cells” are corrected to: “CD+ cells”. Figure 2 should read:
Figure 1.
Interleukin-10 deficiency increases tumor susceptibility in nATF6IEC mice and drives formation of colonic invasive carcinomas. A) Survival curves of nATF6IEC;Il10-/- mice housed under specific pathogen-free (SPF) conditions. B) Colonic tumor incidence of tg/wt;Il10-/- mice categorized into responder (R = mice that developed tumors) and non-responder (NR = mice that did not show any tumors) at week 5, 12 and 15+. C) Colonic tumor numbers of nATF6IEC;Il10-/- mice at week 5, 12, and 15+. D) Dysplasia score resulting from histological scoring of susceptible (proximal) colon tissue of nATF6IEC;Il10-/- mice for mucosal architecture and atypia (0-3) at week 5, 12, and 15+. E) Representative images of H&E staining of susceptible colon tissue at week 12 (scale bars: 200 μm) and corresponding higher magnifications (rectangles; scale bars: 100 μm). F) Tumor incidence of 0.5 × 0.5 cm colon tissue sites identifying a tumor susceptible (colon tissue site 1 to 9) and non-susceptible (colon tissue site from 10 onwards) region with representative macroscopic images of tg/wt;Il10-/- mice; asterisks indicate tumors. G) Ki67 and PAS-AB staining of proximal or susceptible colon tissue of fl/fl;Il10-/- and NR and R tg/wt;Il10-/- mice (scale bars: 100 μm) and corresponding higher magnifications (rectangles; scale bars: 100 μm). Respective quantifications of each staining were analyzed in tumor susceptible colon tissue. H) H-score of nuclear (NUC) and cytoplasmic (CYT) ATF6 expression in colon biopsies of IBD patients stained immunohistochemically. Representative images of ATF6 expression in active and inactive disease as well as ATF6-high and ATF6-low expression in colon tissue (scale bars: 100 μm) and corresponding higher magnifications. wk = week; NT = non-tumor; NR = non-responder; R = responder; NUC = nuclear; CYT = cytoplasmic.
Figure 3.
Mucosa-associated microbiota of responder mice is enriched in CAC-relevant pathobionts. A) Luminal (cecal) and mucosal (colonic) α-diversity (richness, Shannon Effective and Faith’s phylogenetic diversity) in fl/fl;Il10-/- control mice and NR and R tg/wt;Il10-/- mice. B) Luminal and mucosal β-diversity between NR and R tg/wt;Il10-/- mice based on generalized UniFrac distance. Differences between groups were tested using PERMANOVA. C) ROC curves comparing true vs false positive rates of Random Forest, L1-penalised Lasso and Ridge regression models trained on luminal and mucosal data. Mucosal data was randomly subsampled to match luminal sample size (original n=57, subsampled=49). All models were trained using repeated 5-fold cross-validation with 5 repeats. D) Fluorescent in situ hybridization (FISH) targeting 16S rRNA (EUB338 probe) in combination with immunostaining of mucus (UEA1) in the proximal colon of fl/fl;Il10-/- and tg/wt;Il10-/- mice (scale bars: 100 μm) and corresponding higher magnifications (rectangles; scale bars: 100 μm). Nuclei are counterstained with DAPI. Quantification of the bacterial distance to the epithelial surface (o = outer layer, s = stratified layer, e = epithelia). E) Spearman-correlation of centered-log-ratio (CLR) transformed abundance of tumor enriched ASVs classified as Bacteroides, Desulfovibrio and Lachnospiraceae and tumor number in 12-week-old tg/wt;Il10-/- mice. F) Phylogenetic tree showing the similarities between microbiota profiles based on generalized UniFrac distances in mucosal biopsies derived from CD patients (n = 40) and UC patients (n = 24). Individual taxonomic composition at the phylum level is shown as stacked bar plots around the phylogram. Innermost ring shows stratification based on disease phenotype, CD patient samples (blue) and UC patient samples (grey), denoted by an asterisk (*); the second ring shows stratification based on disease activity, active disease (red), inactive disease (yellow), denoted by two asterisks (**); the third ring shows stratification based on ATF6 classification, high ATF6 (green), low ATF6 (grey), denoted by 3 asterisks (***). G) Correlations of H-Scores of nuclear (NUC) and cytoplasmic (CYT) ATF6 expression in colon biopsies of IBD patients with bacterial genera. NT = non-tumor; NR = non-responder; R = responder; AUC = area under the curve; o = outer layer, s = stratified layer, e = epithelia; UEA1 = ulex europaeus agglutinin 1; EUB = Eubacteria FISH probe; DAPI = 4’,6-diamidino-2-phenylindole; ASV = amplicon sequence variant; CD = Crohn’s disease; UC = ulcerative colitis; NUC = nuclear; CYT = cytoplasmic.
In Figure 3, panel B, the corrected p-value for Mucosal is corrected to read: “(p-value 0.015, corr. p-value 0.0225)”instead of: “(p-value 0.015, corr. p-value 0.225)”; “NMDS pot” is corrected to read: “NMDS plot”; in panel E, R-values are emended so that top data figures are positive and lower are negative. Figure 3 should read:
Figure 2.
nATF6IEC;Il10-/- mice show changes in innate immunity. A) Inflammation score for susceptible (proximal) and non-susceptible (distal) colon tissue. nATF6IEC;Il10-/- mice were histologically scored for immune cell infiltration and epithelial damage resulting in an inflammation score (0-9). B) Lipocalin-2 (LCN2) levels in fecal samples of nATF6IEC;Il10-/- mice. C) UMAPs depicting cell abundance and distribution of CD3+CD90+ lymphocytes in colon tissue of nATF6IEC;Il10-/- mice. Colon tissue was divided according to tumor susceptible region (proximal-mid = 2/3 colon length) and non-susceptible region (distal = 1/3 colon length). D) UMAPs depicting cell abundance and distribution of remaining CD3-CD90.2- cells of susceptible and non-susceptible colon tissue of nATF6IEC;Il10-/- mice. E) Quantification of total leukocytes, neutrophils, monocytes and CD4+ T cells given as cell counts per square (0.5 × 0.5 cm colon tissue) and % of leukocytes. F) Quantification of immune cell populations in the colonic lamina propria of R nATF6IEC;Il10-/- mice at 15 + weeks of age by ChipCytometry staining. Immune cells are stained as follows: CD3 (red), CD11b (green), Ly6G (blue). Nuclei are stained with Hoechst (grey). R = responder; NT = non-tumor; T = tumor.
Figure 3.
Mucosa-associated microbiota of responder mice is enriched in CAC-relevant pathobionts. A) Luminal (cecal) and mucosal (colonic) α-diversity (richness, Shannon Effective and Faith’s phylogenetic diversity) in fl/fl;Il10-/- control mice and NR and R tg/wt;Il10-/- mice. B) Luminal and mucosal β-diversity between NR and R tg/wt;Il10-/- mice based on generalized UniFrac distance. Differences between groups were tested using PERMANOVA. C) ROC curves comparing true vs false positive rates of Random Forest, L1-penalised Lasso and Ridge regression models trained on luminal and mucosal data. Mucosal data was randomly subsampled to match luminal sample size (original n=57, subsampled=49). All models were trained using repeated 5-fold cross-validation with 5 repeats. D) Fluorescent in situ hybridization (FISH) targeting 16S rRNA (EUB338 probe) in combination with immunostaining of mucus (UEA1) in the proximal colon of fl/fl;Il10-/- and tg/wt;Il10-/- mice (scale bars: 100 μm) and corresponding higher magnifications (rectangles; scale bars: 100 μm). Nuclei are counterstained with DAPI. Quantification of the bacterial distance to the epithelial surface (o = outer layer, s = stratified layer, e = epithelia). E) Spearman-correlation of centered-log-ratio (CLR) transformed abundance of tumor enriched ASVs classified as Bacteroides, Desulfovibrio and Lachnospiraceae and tumor number in 12-week-old tg/wt;Il10-/- mice. F) Phylogenetic tree showing the similarities between microbiota profiles based on generalized UniFrac distances in mucosal biopsies derived from CD patients (n = 40) and UC patients (n = 24). Individual taxonomic composition at the phylum level is shown as stacked bar plots around the phylogram. Innermost ring shows stratification based on disease phenotype, CD patient samples (blue) and UC patient samples (grey), denoted by an asterisk (*); the second ring shows stratification based on disease activity, active disease (red), inactive disease (yellow), denoted by two asterisks (**); the third ring shows stratification based on ATF6 classification, high ATF6 (green), low ATF6 (grey), denoted by 3 asterisks (***). G) Correlations of H-Scores of nuclear (NUC) and cytoplasmic (CYT) ATF6 expression in colon biopsies of IBD patients with bacterial genera. NT = non-tumor; NR = non-responder; R = responder; AUC = area under the curve; o = outer layer, s = stratified layer, e = epithelia; UEA1 = ulex europaeus agglutinin 1; EUB = Eubacteria FISH probe; DAPI = 4’,6-diamidino-2-phenylindole; ASV = amplicon sequence variant; CD = Crohn’s disease; UC = ulcerative colitis; NUC = nuclear; CYT = cytoplasmic.
Figure 4.
Colonization with an IBD-relevant minimal consortium induces tumor formation in germ-free mice. A) Survival curve of germ-free (GF) and SIHUMI colonized nATF6IEC;Il10-/- mice. GF mice were euthanized at 25 weeks of age. For SIHUMI colonization, mice were gavaged at the age of 4 weeks and colonized for 12 weeks. B) Colonic tumor incidence of nATF6IEC;Il10-/- mice categorized into non-responder (NR), responder (R) or non-tumor (NT). C) Colonic tumor numbers and D) volume in nATF6IEC;Il10-/- mice at endpoint. For tumor volume, only the biggest tumor was considered. E) Dysplasia score of susceptible colon tissue, resulting from histologically scoring GF and SIHUMI colonized nATF6IEC;Il10-/- mice for mucosal architecture and atypia (0-3), and F) inflammation score (0-9), based on immune cell infiltration and epithelial damage. G) Lipocalin-2 (LCN2) levels in proximal colonic content of GF and SIHUMI colonized nATF6IEC;Il10-/- mice. Quantifications of H) proliferating cells (Ki67) per crypt and I) mucin-filled goblet cells (GC) per 100 μm2 in susceptible colon tissue of GF and SIHUMI colonized mice. J) Bacterial abundance of SIHUMI strains in mice at week 1, 6, and 12 after gavage. K) Bacterial abundance of SIHUMI strains colonized in tg/wt;Il10-/- separated according to R and NR after 8 weeks of gavage (the time point where most mice dropped out of the experiment). L) Survival curve of humanized nATF6IEC;Il10-/- mice (one CD patient, one UC patient). Mice were gavaged at the age of 4 weeks and colonized for 12 weeks. LCN2 levels in susceptible colonic content of M) CD and N) UC colonized mice. O) Colonic tumor incidence categorized into NR and R and P) tumor numbers of CD colonized mice. Q) Colonic tumor incidence categorized into NR and R and R) tumor numbers of UC colonized mice. GF = germ-free; SIHUMI = simplified human microbiota; NT = non-tumor; NR = non-responder; R = responder; wk = week; CD = Crohn’s disease; UC = ulcerative colitis.
In Figure 4 panel L, “water after gavage” is corrected to read: “weeks after gavage”. Figure 4 should read:
Figure 2.
nATF6IEC;Il10nATF6IEC;Il10-/- mice show changes in innate immunity. A) Inflammation score for susceptible (proximal) and non-susceptible (distal) colon tissue. nATF6IEC;Il10-/- mice were histologically scored for immune cell infiltration and epithelial damage resulting in an inflammation score (0-9). B) Lipocalin-2 (LCN2) levels in fecal samples of nATF6IEC;Il10-/- mice. C) UMAPs depicting cell abundance and distribution of CD3+CD90+ lymphocytes in colon tissue of nATF6IEC;Il10-/- mice. Colon tissue was divided according to tumor susceptible region (proximal-mid = 2/3 colon length) and non-susceptible region (distal = 1/3 colon length). D) UMAPs depicting cell abundance and distribution of remaining CD3-CD90.2- cells of susceptible and non-susceptible colon tissue of nATF6IEC;Il10-/- mice. E) Quantification of total leukocytes, neutrophils, monocytes and CD4+ T cells given as cell counts per square (0.5 × 0.5 cm colon tissue) and % of leukocytes. F) Quantification of immune cell populations in the colonic lamina propria of R nATF6IEC;Il10-/- mice at 15 + weeks of age by ChipCytometry staining. Immune cells are stained as follows: CD3 (red), CD11b (green), Ly6G (blue). Nuclei are stained with Hoechst (grey). R = responder; NT = non-tumor; T = tumor.
Figure 4.
Colonization with an IBD-relevant minimal consortium induces tumor formation in germ-free mice. A) Survival curve of germ-free (GF) and SIHUMI colonized nATF6IEC;Il10-/- mice. GF mice were euthanized at 25 weeks of age. For SIHUMI colonization, mice were gavaged at the age of 4 weeks and colonized for 12 weeks. B) Colonic tumor incidence of nATF6IEC;Il10-/- mice categorized into non-responder (NR), responder (R) or non-tumor (NT). C) Colonic tumor numbers and D) volume in nATF6IEC;Il10-/- mice at endpoint. For tumor volume, only the biggest tumor was considered. E) Dysplasia score of susceptible colon tissue, resulting from histologically scoring GF and SIHUMI colonized nATF6IEC;Il10-/- mice for mucosal architecture and atypia (0-3), and F) inflammation score (0-9), based on immune cell infiltration and epithelial damage. G) Lipocalin-2 (LCN2) levels in proximal colonic content of GF and SIHUMI colonized nATF6IEC;Il10-/- mice. Quantifications of H) proliferating cells (Ki67) per crypt and I) mucin-filled goblet cells (GC) per 100 μm2 in susceptible colon tissue of GF and SIHUMI colonized mice. J) Bacterial abundance of SIHUMI strains in mice at week 1, 6, and 12 after gavage. K) Bacterial abundance of SIHUMI strains colonized in tg/wt;Il10-/- separated according to R and NR after 8 weeks of gavage (the time point where most mice dropped out of the experiment). L) Survival curve of humanized nATF6IEC;Il10-/- mice (one CD patient, one UC patient). Mice were gavaged at the age of 4 weeks and colonized for 12 weeks. LCN2 levels in susceptible colonic content of M) CD and N) UC colonized mice. O) Colonic tumor incidence categorized into NR and R and P) tumor numbers of CD colonized mice. Q) Colonic tumor incidence categorized into NR and R and R) tumor numbers of UC colonized mice. GF = germ-free; SIHUMI = simplified human microbiota; NT = non-tumor; NR = non-responder; R = responder; wk = week; CD = Crohn’s disease; UC = ulcerative colitis.