Dicarbonyl Electrophiles Mediate Inflammation-Induced Gastrointestinal Carcinogenesis

Abstract

Background & aims: Inflammation in the gastrointestinal tract may lead to the development of cancer. Dicarbonyl electrophiles, such as isolevuglandins (isoLGs), are generated from lipid peroxidation during the inflammatory response and form covalent adducts with amine-containing macromolecules. Thus, we sought to determine the role of dicarbonyl electrophiles in inflammation-associated carcinogenesis.

Methods: The formation of isoLG adducts was analyzed in the gastric tissues of patients infected with Helicobacter pylori from gastritis to precancerous intestinal metaplasia, in human gastric organoids, and in patients with colitis and colitis-associated carcinoma (CAC). The effect on cancer development of a potent scavenger of dicarbonyl electrophiles, 5-ethyl-2-hydroxybenzylamine (EtHOBA), was determined in transgenic FVB/N insulin-gastrin (INS-GAS) mice and Mongolian gerbils as models of H pylori-induced carcinogenesis and in C57BL/6 mice treated with azoxymethane-dextran sulfate sodium as a model of CAC. The effect of EtHOBA on mutations in gastric epithelial cells of H pylori-infected INS-GAS mice was assessed by whole-exome sequencing.

Results: We show increased isoLG adducts in gastric epithelial cell nuclei in patients with gastritis and intestinal metaplasia and in human gastric organoids infected with H pylori. EtHOBA inhibited gastric carcinoma in infected INS-GAS mice and gerbils and attenuated isoLG adducts, DNA damage, and somatic mutation frequency. Additionally, isoLG adducts were elevated in tissues from patients with colitis, colitis-associated dysplasia, and CAC as well as in dysplastic tumors of C57BL/6 mice treated with azoxymethane-dextran sulfate sodium. In this model, EtHOBA significantly reduced adduct formation, tumorigenesis, and dysplasia severity.

Conclusions: Dicarbonyl electrophiles represent a link between inflammation and somatic genomic alterations and are thus key targets for cancer chemoprevention.

Keywords: Chemoprevention; Colorectal cancer; Electrophiles; Stomach Cancer.

Figure 1.

Levels of isoLG-lysyl adducts in GECs in response to H. pylori. A, Representative photomicrographs of human gastric tissues from uninfected individuals (n=4) and patients with NAG (n=13), MAG (n=9), or IM (n=7) immunostained with the D11 antibody. Scale bars, 50 μm. B, Quantification of nuclear D11 staining shown in A. C-D, Human antral gastroids from three different subjects (C, n=2 independent experiments for each) and AGS cells (D, n=3 independent experiments) were infected or not with H. pylori for 24 h in the presence or absence of 100 μM EtHOBA. The intracellular levels of isoLG-lysyl adducts were measured by LC/ESI/MS/MS.

Figure 2.

Effect of EtHOBA on H. pylori-induced carcinogenesis in INS-GAS mice. FVB/N INS-GAS mice were infected or not with H. pylori. EtHOBA (7.5 mg/ml) was given in the drinking water. After 56 days, animals were euthanized, and the stomach was removed (A). The gastric concentration of EtHOBA (B) and isoLG-lysyl adducts (C) were determined by LC/ESI/MS/MS; each dot represents a mouse. D, Colonization of the stomach by H. pylori was assessed by serial dilution and culture. E, Histologic gastritis was determined by scoring H&E staining (H). In D and E, n=39 and n=30 mice, infected with H. pylori and infected mice + EtHOBA, respectively. F-H, The frequency of LGD and IMC in infected mice (F) and the extent of dysplasia and cancer (G) were determined from the H&E staining (H). In F, the number of animals is indicated in each bar, and the same number of mice was included in G. In F, ND, no dysplasia, and in H, the scale bars represent 50 μm. I-J, PET/CT scanning of INS-GAS mice ± H. pylori ± EtHOBA injected with ¹⁸F-NaF (I); the stomach is surrounded by a yellow circle. The quantification of ¹⁸F-NaF uptake in the stomach is shown in J.

Figure 3.

Effect of EtHOBA on H. pylori pathogenesis in Mongolian gerbils. Gerbils were infected or not with H. pylori; n=21 gerbils infected with H. pylori and n=20 infected gerbils + EtHOBA. EtHOBA (7.5 mg/ml) was given in the drinking water. After 56 days, the colonization of the stomach by H. pylori was assessed by serial dilution and culture (A) and gastritis was determined by histopathological scoring (B); all the uninfected gerbils had normal histology (n=10 per group). The frequency of cancer was determined in infected gerbils (C) from the H&E staining (D); scale bars, 50 μm. For all the panels, n=21 gerbils infected with H. pylori and n=20 gerbils infected with H. pylori + EtHOBA.

Figure 4.

Inhibition of H. pylori-induced DNA damage by EtHOBA. A-B, The level of pH2AX was assessed by IHC in the gastric tissues of FVB/N INS-GAS mice ± H. pylori ± EtHOBA (A) and was then quantified in a blinded manner by a GI pathologist (B); the images shown in A are representative of the number of animals shown in B. C-E, WES was performed on a longitudinal strip of the stomach (n=10 INS-GAS, n=12 INS-GAS + EtHOBA, n=12 INS-GAS + H. pylori, and n=8 INS-GAS + H. pylori + EtHOBA). The burden of non-synonymous SNPs (C), the identification of the most frequently mutated genes (D), and VAF distributions (E) are shown. In the oncoplot (D): green, missense mutations; orange, splice-site mutations; red, nonsense mutations; purple, frame shift insertions; blue, frame shift deletions; black, multihit mutations.

Figure 5.

Increased electrophile adduct formation with colitis and CAC. A human TMA containing cores from uninvolved tissues (U) with normal histology (n=6 UC and n=2 CD), patients with colitis (n=23 UC and n=7 CD), LGD (n=18 UC and n=6 CD), HGD (n=4 UC and n=2 CD), and CAC (n=11 UC and n=3 CD) was immunostained with the D11 antibody (A) and the nuclear staining was quantified (B). Black dots depict patients with UC and red dots represents patients with CD. Scale bars, 50 μm.

Figure 6.

Impact of the electrophile scavenger on colon carcinogenesis. C57BL/6 mice were treated with AOM-DSS and were given 1.5 mg/ml EtHOBA according to the experimental protocol depicted in A. After 56 days, colons were removed and tumor number (B), average tumor size per mouse (C), and tumor burden (D) were determined macroscopically. Then H&E staining (E) was used to assess the number of adenomas (F), the frequency of LGD and HGD (G) and histologic colitis in the non-tumor areas (H). In E, each arrow depicts a tumor and the scale bars represent 50 μm.

Figure 7.

Generation of isoLG-lysyl adducts in mice with CAC. Immunostaining with D11 antibody was performed in the colon of mice treated with AOM-DSS ± EtHOBA, according to the experimental procedure shown in Figure 6A; these photomicrographs are representative data of the staining performed on 4 mice per group. Scale bars, 50 μm.