published_or_final_version / Pathology / Master / Master of Medical Sciences

The endocannabinoid system (ECS) is known to be involved in key aspects of cell maintenance within the human colon, as well as being dysregulated in pathophysiological conditions, including colon inflammation and cancer. However, the contribution of the ECS within each of these conditions has not been fully elucidated. This indicates that the current identification of key targets within the ECS that are involved in gut pathology could be used as potential novel therapeutics. Two experimental approaches were designed and optimised to give an insight into ECS signal regulation within the human colon and to screen ECS therapeutics, tetrahydrocannabinol (THC) and cannabidiol (CBD); a human colon ex vivo explant culture model and an innovative multiplexed quantitative gene expression technology, the GenomeLab GeXP system (Beckman Coulter). Gene targets were identified that are known markers of regulation and function in cells of healthy tissue. An assay, the hCellMarkerPlex was designed that incorporated twenty-three of these gene targets, epithelial (EZR, KRT18, SLC9A2), proliferation (PCNA, CCND1, MS4A12), differentiation (B4GANLT2, CDX1, CDX2), apoptotic (CASP3, NOX1, NTN1), fibroblast (FSP1, COL1A1), structural (ACTG2, CNN1, DES), gene transcription (HDAC1), stem cell (LGR5), endothelial (VWF) and mucin production (MUC2). The hCellMarkerPlex identified gene signatures which distinguished between normal, adenoma and carcinoma tissue, identifying cellular processes showing abnormal activity associated with pathological status. The resulting biomarker profiles were used to establish a human colon explant culture system. The human colon explant culture presents a novel model to study modulation of the ECS and screen ECS therapeutics. Combined with the GenomeLab GeXP System multiple components of the ECS were assessed at the gene regulatory level. A custom designed GeXP assay, the hECSplex, was developed. hECSplex gene expression signatures of EC receptors (CNR1, CNR2, GPR55 and TRPV1), ECS enzymes (NAPE-PLD, GDE1, DAGLA, DAGLB, FAAH, FAAH2 and PTGS2), inflammatory (IL1B, IL10, IL6, LEP, TNF and SOCS3), signalling pathway (ID1, BCL2, CFL1, BIRC5, TP53, MYC and KRAS), lipid production (SREBF1, ACACA), and plasma-membrane (OCLN) markers revealed altered expression of ECS components in carcinogenesis compared to normal tissue. Abstract vi . The hECSplex gene expression signature of colon explants showed that ECS was not altered during culture, emphasising the explant models capability as a pharmaceutical tool to test current and novel therapeutics. Applications of both THC and CBD to normal colon explants at different concentrations do not lead to any significant changes. Indicating the current pharmacological use of phytocannabinoids is causing no adverse effects in surrounding healthy colon tissue. The GenomeLab System presents new opportunities to interrogate multiple components of the endocannabinoid signalling system in small colon explant tissue samples, and in response to ECS therapeutics.

I-Abstract I.1-Objectives- To grossly evaluate colonic abnormalities within the Manitoba population by determining prevalence according to; age, gender, body mass index, body weight, body length, colon length, and fecal weight. I.2-Methods- A population study of 67 medico-legal autopsies from two major teaching hospitals was performed, examining the colon for abnormalities. The colon was resected just proximal to the ileocecal valve and just distal to the recto-sigmoid junction. Once the specimen was detached it was weighed with the autopsy scale opened and the feces removed. The specimen was patted dry and re-weighed. The specimen was laid on a blue specimen photograph board where the length was measured from the cecum to terminal sigmoid/rectum and photographed digitally. The specimen was then evaluated for any abnormalities. The specimen was then placed back into the body. I.3-Results Of the 67 colons assessed, 66% had an abnormality, of which 37.3% had diverticular disease and 24% had polyps, the two most common diseases. Age was the only significant factor (p=0.004) in this study affecting prevalence. The prevalence of multiple polyps was 63% with colon length being the only significant factor (p=0.0265) in this study affecting prevalence. I.4-Conclusions A progressive risk of increased abnormality formation is noted with age. Diverticular disease and polyps have similar prediction factors and disease prevalence. Many factors are suggested in the literature to influence the prediction of abnormalities, however only age was determined to be significant in this study. Multiplicity of polyps in the colon is significantly related to colon length.

Helicase-like Transcription Factor (HLTF) is a DNA helicase protein which is homologous to SNF/SWI family. It has been demonstrated to be a functional homolog of yeast Rad5, required for the maintenance of genomic stability. Although the physiologic role of HLTF is largely unknown,inactivation of HLTF by promoter hypermethylation has been found in more than 40% human colon cancers. In this study, we have applied mouse transgenic approaches to determine whether loss of HLTF function could be important for colorectal carcinogenesis. HLTF knockout mice were generated by the deletion of first 5 exons of the HLTF gene. The complete loss of HLTF expression in HLTF -/- mice was confirmed by northern blot and real time RT-PCR assays. HLTF -/- mice did not show any developmental defects within a 2-year observation indicating that HLTF is dispensable for mouse development. Furthermore, HLTF -/- mice were free of intestinal or colorectal tumors or other types of tumors, suggesting that loss of HLTF function alone is not sufficient to drive oncogenic transformation in intestinal track and other tissues. To determine whether loss of HLTF function could cooperate with other tumor suppressors in the formation of colorectal cancers, we have bred HLTF knockout mice with the mutant mice for APC (adenomtous polyposis coli) and P53. In HLTF -/-APC Min/+ mice, a significantly increased formation of intestinal adenocarcinoma and colorectal cancers were observed. Although very few HLTF -/-P53 -/- mice developed colorectal cancers, these mice had increased incidence of the formation of metastatic lymphomas. Cytogenetic analysis of colorectal cancer cells derived from HLTF -/-APC Min/+ mice demonstrated a high incidence of gross chromosomal instabilities, including Robertsonian fusions, fragments and aneuploidy. All these genetic alterations were not observed in the intestinal tumor cells from APC Min/+, implicating that loss of HLTF function could induce genomic instability which contributes to intestinal carcinogenesis. To further investigate the role of HLTF in colorectal carcinogenesis, we have also applied a shRNA knockdown approach to down-regulate HLTF expression in human HCT-116 colon cancer cells. HCT-116 cells highly express HLTF and show less chromosomal instability, making these cells as a very useful model to investigate the loss of function of HLTF in human colorectal carcinogenesis. Using Western blot approach, we confirmed that HLTF knockdown HCT-116 cells had less than 5% of HLTF expression as compared to the scramble controls. By inoculating HLTF knockdown HCT-116 cells to Rag1 -/-IL2 -/- immunocompromised mice, we further demonstrated that HLTF knockdown promote tumor growth and invasion. Moreover, spectral karyotyping analysis revealed that HLTF knockdown human colon cancer cells had significantly increased chromosomal instability, including both aneuploidy and chromosomal translocation. Taken together, our work strongly indicates that loss of HLTF function can promote the malignant transformation of intestinal or colonic adenomas to carcinomas by inducing genomic instability. Given the high frequency of epigenetic inactivation by hypermethylation of HLTF in human colon cancers, our studies strongly suggest that this epigenetic alteration could be directly involved in the development of colorectal cancer rather than a consequence of this carcinogenesis.

This study investigated enteric nervous system (ENS) involvement in intestinal secretion induced by cyclic GMP-dependent secretagogues. The investigation was based upon the study of transepithelial ion transport in rat small and large intestine and in guinea pig caecum using voltage-clamped in vitro preparations mounted in Ussing chambers. ENS participation was established from the use of neural blocking agents (tetrodotoxin (TTX), bicuculline and capsaicin). The relative contribution of the myenteric plexus was assessed by selectively stripping tissues of the longitudinal muscle layer. All tissues, both unstripped and stripped, responded to <I>Escherichia coli </I>STa enterotoxin, guanylin and the nitric oxide (NO) donor sodium nitroprusside (SNP) with a dose-dependent increase in inward short circuit current (I_SC). Regarding STa/guanylin, TTX inhibited this I_SCin unstripped rat distal colon, ileum and guinea pig caecum, demonstrating that the ENS plays an important role in these tissues. In rat distal colon, TTX induced an abolition of the STa/guanylin response in both preparations, indicating submucous plexus involvement. In rat proximal colon there was no discernible TTX-sensitive component observed. The ileum displayed partial control from both the myenteric and submucous plexuses, whereas the caecum exhibited partial control from the myenteric plexus alone. Bicuculline inhibited STa action to a significant degree in the caecum while capsaicin inhibited secretion in the proximal colon. In rat small intestine, the SNP-induced I_SC was inhibited by TTX in both unstripped and stripped tissues. In contrast, inhibitory pathways were shown to exist in distal colon exposed to SNP, TTX revealing an enhancement of SNP-induced secretion in the stripped preparation. Thus, although there is clear involvement of the ENS in the actions of STa/guanylin and SNP, it is not possible to make a general statement regarding its contribution throughout the length of the alimentary canal due to the extent of inter segmental and inter species variations.

I-Abstract I.1-Objectives- To grossly evaluate colonic abnormalities within the Manitoba population by determining prevalence according to; age, gender, body mass index, body weight, body length, colon length, and fecal weight. I.2-Methods- A population study of 67 medico-legal autopsies from two major teaching hospitals was performed, examining the colon for abnormalities. The colon was resected just proximal to the ileocecal valve and just distal to the recto-sigmoid junction. Once the specimen was detached it was weighed with the autopsy scale opened and the feces removed. The specimen was patted dry and re-weighed. The specimen was laid on a blue specimen photograph board where the length was measured from the cecum to terminal sigmoid/rectum and photographed digitally. The specimen was then evaluated for any abnormalities. The specimen was then placed back into the body. I.3-Results Of the 67 colons assessed, 66% had an abnormality, of which 37.3% had diverticular disease and 24% had polyps, the two most common diseases. Age was the only significant factor (p=0.004) in this study affecting prevalence. The prevalence of multiple polyps was 63% with colon length being the only significant factor (p=0.0265) in this study affecting prevalence. I.4-Conclusions A progressive risk of increased abnormality formation is noted with age. Diverticular disease and polyps have similar prediction factors and disease prevalence. Many factors are suggested in the literature to influence the prediction of abnormalities, however only age was determined to be significant in this study. Multiplicity of polyps in the colon is significantly related to colon length.

Helicase-like Transcription Factor (HLTF) is a DNA helicase protein which is homologous to SNF/SWI family. It has been demonstrated to be a functional homolog of yeast Rad5, required for the maintenance of genomic stability. Although the physiologic role of HLTF is largely unknown,inactivation of HLTF by promoter hypermethylation has been found in more than 40% human colon cancers. In this study, we have applied mouse transgenic approaches to determine whether loss of HLTF function could be important for colorectal carcinogenesis. HLTF knockout mice were generated by the deletion of first 5 exons of the HLTF gene. The complete loss of HLTF expression in HLTF -/- mice was confirmed by northern blot and real time RT-PCR assays. HLTF -/- mice did not show any developmental defects within a 2-year observation indicating that HLTF is dispensable for mouse development. Furthermore, HLTF -/- mice were free of intestinal or colorectal tumors or other types of tumors, suggesting that loss of HLTF function alone is not sufficient to drive oncogenic transformation in intestinal track and other tissues. To determine whether loss of HLTF function could cooperate with other tumor suppressors in the formation of colorectal cancers, we have bred HLTF knockout mice with the mutant mice for APC (adenomtous polyposis coli) and P53. In HLTF -/-APC Min/+ mice, a significantly increased formation of intestinal adenocarcinoma and colorectal cancers were observed. Although very few HLTF -/-P53 -/- mice developed colorectal cancers, these mice had increased incidence of the formation of metastatic lymphomas. Cytogenetic analysis of colorectal cancer cells derived from HLTF -/-APC Min/+ mice demonstrated a high incidence of gross chromosomal instabilities, including Robertsonian fusions, fragments and aneuploidy. All these genetic alterations were not observed in the intestinal tumor cells from APC Min/+, implicating that loss of HLTF function could induce genomic instability which contributes to intestinal carcinogenesis. To further investigate the role of HLTF in colorectal carcinogenesis, we have also applied a shRNA knockdown approach to down-regulate HLTF expression in human HCT-116 colon cancer cells. HCT-116 cells highly express HLTF and show less chromosomal instability, making these cells as a very useful model to investigate the loss of function of HLTF in human colorectal carcinogenesis. Using Western blot approach, we confirmed that HLTF knockdown HCT-116 cells had less than 5% of HLTF expression as compared to the scramble controls. By inoculating HLTF knockdown HCT-116 cells to Rag1 -/-IL2 -/- immunocompromised mice, we further demonstrated that HLTF knockdown promote tumor growth and invasion. Moreover, spectral karyotyping analysis revealed that HLTF knockdown human colon cancer cells had significantly increased chromosomal instability, including both aneuploidy and chromosomal translocation. Taken together, our work strongly indicates that loss of HLTF function can promote the malignant transformation of intestinal or colonic adenomas to carcinomas by inducing genomic instability. Given the high frequency of epigenetic inactivation by hypermethylation of HLTF in human colon cancers, our studies strongly suggest that this epigenetic alteration could be directly involved in the development of colorectal cancer rather than a consequence of this carcinogenesis.

The purpose of this study was to determine the absorption of [13C]5-formyltetrahydrofolic acid across the human colon with an intact microflora. Bioavailability was determined after administration of a pH-dependent acrylic co-polymer coated barium sulphate caplet containing 855 nmol and measuring the plasma appearance of [13C5]5-methyltetrahydrofolic acid compared to an intravenous injection of the same compound (214 nmol). Blood samples serially collected after both test doses were analyzed by microbiological assay and tandem mass spectrometry to determine total folates and ratio of labeled to unlabeled folates, respectively. Caplet disintegration in the colon was quantitatively assessed by fluoroscopic imaging for six of nine subjects. The plasma [13C5]5-methyltetrahydrofolic acid appearance rate was 4 + 0.9 nmol/h (intravenous) and 0.35 + 0.02 nmol/h (caplet). Mean apparent absorption across the colon was 35%. These findings suggest that physiological doses of natural folate are absorbed across the human colon in the presence of an undisturbed microbiota.

The purpose of this study was to determine the absorption of [13C]5-formyltetrahydrofolic acid across the human colon with an intact microflora. Bioavailability was determined after administration of a pH-dependent acrylic co-polymer coated barium sulphate caplet containing 855 nmol and measuring the plasma appearance of [13C5]5-methyltetrahydrofolic acid compared to an intravenous injection of the same compound (214 nmol). Blood samples serially collected after both test doses were analyzed by microbiological assay and tandem mass spectrometry to determine total folates and ratio of labeled to unlabeled folates, respectively. Caplet disintegration in the colon was quantitatively assessed by fluoroscopic imaging for six of nine subjects. The plasma [13C5]5-methyltetrahydrofolic acid appearance rate was 4 + 0.9 nmol/h (intravenous) and 0.35 + 0.02 nmol/h (caplet). Mean apparent absorption across the colon was 35%. These findings suggest that physiological doses of natural folate are absorbed across the human colon in the presence of an undisturbed microbiota.

Many studies show that estradiol (E2) and consumption of soy and its primary phytoestrogen component genistein (GEN) can inhibit the formation of colon tumors. However, the effects of E2 and GEN at physiologically relevant levels in non-diseased colonocytes have yet to be investigated. We hypothesized that E2 and GEN could prove to be chemo-protective agents in the colon by moderately increasing apoptosis and decreasing proliferation in a healthy system. Thus, the presented studies focused on evaluating the effects of E2 and GEN in non-malignant colonocytes in vitro and in vivo to determine how the compounds influence the physiology of these cells. E2 (1 nM/L) and GEN treatments (1 and 10 microM/L) decreased cell growth, increased apoptosis, and increased p53 transcriptional activity in young adult mouse colonocytes, a non-malignant cell line. To study further the effects of E2 and GEN in healthy colonic epithelia, we evaluated physiologic changes in colonic crypts in ovariectomized mice given an E2 pellet, 1,000 ppm GEN diet, or a phytoestrogen free diet. As seen in vitro, E2 treated animals had significantly higher rates of apoptosis with GEN trending in the same fashion. These data demonstrate that E2 and GEN alter the physiology of non-malignant colonocytes. Collectively, with our previous data, this suggests that E2 and GEN influence colonocyte physiology and this state may partially explain how these compounds decrease risk of colon cancer.