Determination of the Potential Tumor-Suppressive Effects of Gsdme in a Chemically Induced and in a Genetically Modified Intestinal Cancer Mouse Model

Gasdermin E (GSDME), also known as deafness autosomal dominant 5 (DFNA5) and previously identified to be an inducer of regulated cell death, is frequently epigenetically inactivated in different cancer types, suggesting that GSDME is a tumor suppressor gene. In this study, we aimed to evaluate the tumor-suppressive effects of GSDME in two intestinal cancer mouse models. To mimic the silencing of GSDME by methylation as observed in human cancers, a Gsdme knockout (KO) mouse was developed. The effect of GSDME on tumorigenesis was studied both in a chemically induced and in a genetic intestinal cancer mouse model, as strong evidence shows that GSDME plays a role in human colorectal cancer and representative mouse models for intestinal cancer are available. Azoxymethane (AOM) was used to induce colorectal tumors in the chemically induced intestinal cancer model (n = 100). For the genetic intestinal cancer model, Apc1638N/+ mice were used (n = 37). In both experiments, the number of mice bearing microscopic proliferative lesions, the number and type of lesions per mouse and the histopathological features of the adenocarcinomas were compared between Gsdme KO and wild type (WT) mice. Unfortunately, we found no major differences between Gsdme KO and WT mice, neither for the number of affected mice nor for the multiplicity of proliferative lesions in the mice. However, recent breakthroughs on gasdermin function indicate that GSDME is an executioner of necrotic cell death. Therefore, it is possible that GSDME may be important for creating an inflammatory microenvironment around the tumor. This is in line with the trend towards more severe inflammation in WT compared to Gsdme KO mice, that we observed in our study. We conclude that the effect of GSDME in tumor biology is probably more subtle than previously thought.

. Reference genes used for normalization of Gsdme expression in brain and colon tissue. The geNorm expression stability value of the reference gene (M) and the coefficient of variation of the normalized reference gene relative quantities (CV) are shown for every reference gene. Good reference genes have a M < 0.5 and a CV < 0.2. All primers, except the ones for Rpl13a (forward primer: 5'-CACTCTGGAGGAGAAACGGAAGG-3' and reverse primer: 5'-GCAGGCATGAGGCAAACAGTC-3'), were primer mixes from the Mouse geNormPLUS kit (Primerdesign, Chandler's Ford, United Kingdom). Table S2. Overview of all lesions scored in the large intestine of AOM-treated mice, sacrificed at respectively 20, 22 or 24 weeks of age.  The number of mice with at least one specific lesion throughout the whole large intestine at respectively 20, 22 or 24 weeks of age, or in one specific part of the large intestine (proximal, mid 1, mid 2, distal), in the Gsdme KO and in the WT group is indicated. Proliferative change includes: typical hyperplasia, atypical hyperplasia, adenoma and/or adenocarcinoma.   figure 1). In literature, these hepatocellular inclusions after AOM treatment have been reported also in rats and would seem to be cytoplasmic invaginations [1]. These 'intranuclear' -intracytoplasmic inclusion bodies are probably proliferations of smooth endoplasmic reticulum (induction of cytochrome P450) induced by AOM. Mice of the C57BL/6 strain show high inducibility of cytochromes in the liver [2].

Reference Genes -Brain
Dilated sinusoids are associated with hypoxia and can be associated with liver inflammation; or can be seen around hepatic nodules, due to hemodynamic changes during sacrifice. Foci of hyperplasia were sometimes demarcated by fibrous tissue that accentuated the nodules (macroscopically visible; Supplementary figure 1). Hepatic cells in these nodules were usually of normal size or larger or smaller than normal hepatocytes. These lesions were also reported in literature in rats [1]. Cirrhosis develops as a chronic stage after acute liver injury (necrosis) and is seen after administration of several xenobiotics [3]. The cirrhosis encountered here is most probably related to the AOM treatment.
In one WT mouse a focal nodular lesion was seen. In a focal area of biliary duct proliferation there was deposition of crystalloid and strongly basophilic substances, most probably calcium, in the lumina of the ducts associated with degeneration of the biliary walls. In one Gsdme KO mouse a nodular adenoma was found in the liver. In addition, there were some other lesions seen: congestion (N = 40/54 WT, N = 40/46 KO), which is probably related to circulation disorders during sacrifice; hepatocellular macro-and microvesicular vacuolation (N = 24/54 WT, N = 24/46 KO) which is a result of fat accumulation and is related to the nutritional status of the animal, but can also be related to hepatic dysfunction (fatty change). Moreover, microgranulomes (N = 41/54 WT, N = 39/46 KO), small aggregates of inflammatory cells grouped around small zones of degenerate hepatocytes, were found. Finally, portal inflammation of mononuclear cells (N = 6/54 WT, N = 11/46 KO) was observed, which can be associated with biliary tract lesions. Lastly, in one Gsdme KO mouse extramedullar hematopoiesis was seen. Hematopoiesis can occur in the liver when normal hematopoiesis in the bone marrow and/or spleen is disturbed, for example when the bone marrow and/or spleen are invaded with tumor cells (lymphoma). For none of those hepatic lesions we were able to find a statistical significant difference between WT and Gsdme KO mice.

Supplementary Information LUNGS -CHEMICAL EXPERIMENT
Lungs from 22 Gsdme KO and 16 WT mice were analyzed. These lungs revealed red patches at necropsy. Histological examination revealed congestion (N = 12/16 WT, N = 12/22 KO), edema (N = 1/16 WT, N = 2/22 KO) and/or areas of atelectasis (N = 4/16 WT, N = 6/22 KO). These lesions are probably related to hemodynamic changes during sacrifice. Two WT and two Gsdme KO mice showed hemorrhage of the lungs. One Gsdme KO mouse showed focal, minimal, interstitial mononuclear cell infiltration. In one WT mouse pneumonia was seen. This animal presented at necropsy with purple-colored patches on the lung surfaces. Microscopically, a marked inflammatory infiltrate of neutrophilic granulocytes and mononuclear cells was multifocally present in the lung alveoli and septa. This inflammation can have different causes. In aged mice it has been associated with aspiration of foreign material in relation to general bad health.
In the lungs, hyperemia (N = 12/24 WT, N = 8/13 KO) and hemorrhages (N = 3/24 WT, N = 3/13 KO) were found, which were probably related to circulation disorders during sacrifice. In one Apc 1638N/+ Gsdme WT and one Apc 1638N/+ Gsdme KO mouse lymphocytic interstitial infiltration was found which indicated (focal) inflammation in the lungs. In one Apc 1638N/+ Gsdme WT mouse areas of atelectasis and emphysema were observed, which were probably related to respiration disorders during sacrifice. None of these lesions were significantly different between Apc 1638N/+ Gsdme WT and Apc 1638N/+ Gsdme KO mice.