Intestine-Specific Mttp Deletion Increases the Severity of Experimental Colitis and Leads to Greater Tumor Burden in a Model of Colitis Associated Cancer

Background Gut derived lipid factors have been implicated in systemic injury and inflammation but the precise pathways involved are unknown. In addition, dietary fat intake and obesity are independent risk factors for the development of colorectal cancer. Here we studied the severity of experimental colitis and the development of colitis associated cancer (CAC) in mice with an inducible block in chylomicron secretion and fat malabsorption, following intestine-specific deletion of microsomal triglyceride transfer protein (Mttp-IKO). Methodology/Principal Findings Mttp-IKO mice exhibited more severe injury with ∼90% mortality following dextran sodium sulfate (DSS) induced colitis, compared to <20% in controls. Intestinal permeability was increased in Mttp-IKO mice compared to controls, both at baseline and after DSS administration, in association with increased circulating levels of TNFα. DSS treatment increased colonic mRNA expression of IL-1β and IL-17A as well as inflammasome expression in both genotypes, but the abundance of TNFα was selectively increased in DSS treated Mttp-IKO mice. There was a 2-fold increase in colonic tumor burden in Mttp-IKO mice following azoxymethane/DSS treatment, which was associated with increased colonic inflammation as well as alterations in cytokine expression. To examine the pathways by which alterations in fatty acid abundance might interact with cytokine signaling to regulate colonic epithelial growth, we used primary murine myofibroblasts to demonstrate that palmitate induced expression of amphiregulin and epiregulin and augmented the increase in both of these growth mediators when added to IL-1βor to TNFα. Conclusions These studies demonstrate that Mttp-IKO mice, despite absorbing virtually no dietary fat, exhibit augmented fatty acid dependent signaling that in turn exacerbates colonic injury and increases tumor formation.


Introduction
The continued epidemic of obesity and its associated comorbidities has fueled interest in understanding the pathways and mediators involved, particularly in relation to inflammation and cancer. Obesity and its related metabolic complications have been demonstrated to play a role in a number of cancers, including colorectal cancer (CRC) and consequently dietary modification and weight control are viewed as key modifiable risk factors [1]. However, the precise mechanisms and pathways by which obesity contributes to CRC risk are still incompletely understood and are likely to include alterations in systemic inflammation and cytokine signaling, insulin resistance as well as local and systemic effects of altered nutrient uptake including fatty acids, cholesterol and bile acids [2]. Because of its importance as an energy source, dietary fat intake has been a target for intervention as a strategy to mitigate the effects of obesity and inflammation on CRC risk [3].
In addition to the role for obesity and fat intake in modulating inflammation and CRC risk, there is emerging information suggesting that the small intestine plays an important role as a mediator in the inflammatory response to systemic sepsis [4,5]. In this scenario, it has been suggested that the small intestine secretes lipid products derived from luminal digestion, which are then transported in mesenteric lymphatics, where they induce organ failure at distant sites [6,7]. Other work has reinforced this gutlymph hypothesis by demonstrating that ligation of the mesenteric duct (thereby eliminating delivery of chylomicrons from the intestine into the systemic circulation) abrogates the systemic effects of shock, findings that point to the importance of lipid mediators arising from the intestine in the pathogenesis of systemic inflammation [8,9]. Other work has implicated lipoproteins as vehicles for transporting growth factors and morphogens, including mammalian Wnt3, suggesting a more fundamental role for lipid transport in the regulation of epithelial growth and proliferation [10].
In this study, we have examined the role of intestinal chylomicron secretion in order to understand the role of intestinal lipid transport in the pathogenesis of colonic inflammation and colitis associate cancer (CAC). We recently demonstrated that mice with conditional intestine-specific deletion of microsomal triglyceride transfer protein (Mttp-IKO) develop virtually complete block of intestinal absorption [11] and exhibit a survival advantage when challenged with Pseudomonas aeruginosa, the most common cause of gram negative pneumonia [12]. Those findings raised the possibility that blocking intestinal chylomicron secretion might also attenuate the effects of chemical induced experimental colitis and in turn abrogate colorectal cancer development following azoxymethane/dextran sodium sulfate (AOM/DSS) challenge [13]. Our findings, however, revealed that Mttp-IKO mice developed worse colonic injury and increased tumor burden. Furthermore, we found that altered fatty acid signaling may play a key role in promoting these phenotypes.

Animals
Mttp flox/flox villin-Cre-ER T2 (Mttp-IKO) mice (along with littermate controls) were used for these studies, in a background of ,75% C57BL/6 and ,25% 129/SvJ. Cre recombinase expression in villus epithelial cells was induced by five daily intraperitoneal injections of 1 mg tamoxifen (Sigma), as described previously [11]. Experiments were performed on mice consuming a regular low fat rodent chow. Experimental colitis was induced by administration of 2.5% dextran sodium sulfate (DSS) (MW 40,000-50,000, Cat#9011-18-1, Affymetrix, Inc, Cleveland, Ohio.), for the indicated times in the figure legends and weighed daily. CAC was induced by injection of 8 week old mice with 10 mg/kg body weight azoxymethane (AOM, Sigma) followed by three cycles (5 days) of DSS starting at 5 th (2%), 26 th (2.5%) and 46 th (2.5%) day after AOM injection, minor modifications of the protocol used by others [14]. Mice were sacrificed at 9 weeks after AOM injection and tissues collected for analysis. All animal studies were approved by the Animal Studies Committees of Washington University School of Medicine (#20100146) and were conducted in strict accordance with the National Institutes of Health guidelines for the use of laboratory animals.

Histomorphological analyses
Samples of small intestine and colon were fixed and embedded for sectioning and hematoxylin and eosin (H&E) staining. Where indicated, intestinal sections were stained for lipid droplets using osmium tetroxide as described [12] Histologic scoring of DSS injury was undertaken and quantitated using the parameters described [15]. Intestinal proliferation was examined in mice injected with 5-Bromo-29deoxyuridine (BrdU) (200 ml volume, 5 mg/ml diluted in normal saline; Sigma), 2 hours prior to sacrifice. Samples were processed for BrdU staining as described [12]. Analysis of colon tumor burden was undertaken using samples fixed in 10% formalin (Sigma) and pinned for scoring by an investigator blinded to genotype. Each sample was photographed using a Nikon SMZ800 dissecting microscope and a Photometrics CooLSNAPcf camera (Imaging Processing Services). The area of each section was determined and the size of tumors quantitated using Metavue software (Molecular Devices). H&E sections were also reviewed by a pathologist (IN), blinded to genotype.

Intestinal permeability and cytokine assays
Intestinal permeability was determined following injection of FITC labeled dextran (FD-4, MW 4000, Sigma). Control and Mttp-IKO mice were gavaged with FD-4 (400 mg/Kg body weight), prior to and 7 days after 2.5% DSS treatment and sera were collected 4 hour later. Serum FITC was measured on a fluorimeter (Synergy HT, BioTekH) at excitation 485/20, emission 528/20. For LPS determinations, sera were diluted 1:10 and heated at 70uC for 15 minutes to inactive inhibitors. LPS was then measured using the LAL chromogenic endotoxin quantitation kit according the manufacturer's instructions (ThermoScientific, Rockford, IL). Serum TNFa and IL-1b were quantitated by ELISA using kits purchased from BD Biosciences (San Jose, CA) or R&D systems (Minneapolis, MN) following the manufacturers' instructions.

Real time quantitative polymerase chain reaction
Total RNA was isolated from intestinal tissues using Trizol either under standard protocols or for samples treated with DSS, using the modification described by Kerr and colleagues [16]. RNAs were quantitated using primer pairs (sequences available upon request) designed by Primer express software (Applied Biosystems). Relative mRNA abundance is expressed relative to control mice without DSS treatment, normalized to GAPDH.

Colonic lipid content, Fecal lipid content and protein expression
Total lipids were extracted from feces as well as from proximal colon, as previously described [11]. Triglyceride (TG), total cholesterol (TC) and free fatty acid (FFA) were measured enzymatically using Wako kits (Wako Chemicals, Richmond, VA). Western blotting was conducted in standard fashion using antibodies whose source is indicated in the relevant figure legend. For colonic TNFa and IL1b content samples of descending colon were homogenized in buffer containing 20 mM Tris (pH 7.4), 1 mM Sodium orthovanadate, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton, 50 mM Sodium fluoride, 50 mM bglycerophosphate and 16 Complete mini protease inhibitor cocktail (Roche, Mannhein, Germany). TNFa and IL1b levels were measured in clarified supernatants using the ELISA kits described above and normalized to protein concentration.

Primary murine intestinal myofibroblasts
Myofibroblasts were isolated from C57BL/6 wild type mice as described [17] and 1610 5 cells per well were seeded in 6-well plates and cultured overnight. The phenotype of these cells was verified by positive staining for smooth muscle actin and vimentin and negative staining for desmin and cytokeratin [13,17]. These myofibroblasts cells were then cultured for 12 hours in DMEM containing 0.25% FBA and 0.6% Fatty acid free bovine serum albumin (BSA, A8806, Sigma), supplemented with either buffer alone (BSA, Control), 200 mM Palmitate-BSA (PAL), TNFa at 100 ng/ml (TNFa) (R&D), IL1b at 10 ng/ml (IL1b) (R&D) or combinations of 200 mM Palmitate-BSA plus TNFa 100 ng/ml (TNFa+ PAL) or 200 mM Palmitate -BSA and IL1b 10 ng/ml (IL1b+ PAL), as indicated in the legend. These studies were all performed on myofibroblasts at passage 6-8. Sodium palmitate was conjugated with fatty acid free BSA at ratio of 2:1 (Palmitate: BSA), to make a 3 mM palmitate-BSA stock solution. Total RNA was extracted from individual samples and mRNAs quantitated as described above.

Statistical Analysis
Multiple group comparisons of continuous data sets were performed using one-way analysis of variance followed by post-hoc t-test using Prism 4.0 (GraphPad Software, San Diego, CA) and Microsoft Excel. Data are reported as means 6 SEM. Survival studies were analyzed via the log-rank test. A p value ,0.05 was considered to be statistically significant.

Increased severity of DSS colitis in Mttp-IKO mice
We administered 2.5% DSS in drinking water for 12 days in order to evaluate the impact on overall survival by genotype. Our findings revealed that Mttp-IKO mice exhibit accelerated death and markedly reduced survival (1/10) compared to control mice (9/11) Figure 1A. The augmented injury phenotype in Mttp-IKO mice was associated with greater weight loss in shorter (7 days) experiments ( Figure 1C) and a delayed return to baseline weight ( Figure 1B), along with reduced colon length ( Figure 1D, E). There was also more severe histologic injury as evidenced by increased lamina propria inflammation and crypt drop-out in the descending colon in Mttp-IKO mice at both 5 days (

DSS administration does not induce macroscopic small intestinal injury in Mttp-IKO mice and is not associated with colonic lipid accumulation
Our previous studies demonstrated that blocking chylomicron secretion from the intestine of Mttp-IKO mice results in massive lipid engorgement with villus distortion [11], raising the possibility that the high mortality encountered following DSS administration might reflect gross damage in the small intestine of these mice. However, this was not the case. As seen in Figure 3 A-D, we observed the expected lipid accumulation in small intestinal villi in Mttp-IKO mice but there was no gross or histological evidence of ulceration in any region of the small intestine following 7 days of DSS treatment. In addition, since Mttp is expressed in the colon of mice (albeit at low levels [18,19]) we explored the possibility that colonic lipid accumulation might contribute to the phenotype observed in Mttp-IKO mice. However, this again was not the case. As seen in Figure 3E-H, while we detected abundant lipid droplet accumulation in the small intestine of Mttp-IKO mice (as previously noted [11]) there were only rare, scattered osmium staining lipid droplets seen in the colon of Mttp-IKO mice and there was no accumulation of triglyceride, cholesterol or free fatty acid within colonic mucosa as detectable by enzymatic assay ( Figure 3I).

Increased intestinal permeability in Mttp-IKO mice and the role of an adaptive stress response
While the findings above imply that there are no gross alterations in villus integrity in Mttp-IKO mice, we explored the possibility that more subtle alterations might be associated with barrier dysfunction. To address this possibility, we administered FITC labeled dextran by gavage to mice of both genotypes, either before or after 7 days of DSS administration. Our findings revealed that serum FITC levels were significantly higher in Mttp-IKO mice under both conditions ( Figure 4A). These findings suggest that there is altered barrier function at baseline in Mttp-IKO mice that becomes further impaired in the setting of DSS injury. We next considered the possibility that the increased intestinal permeability phenotype in Mttp-IKO mice might in part reflect alterations in the expression of integral membrane proteins involved in tight junction maintenance and whose altered expression has been linked to defects in human inflammatory bowel disease [20,21]. We found that expression of HNF4a, ZO-1 and Lamb1 was decreased at baseline in the small intestine of Mttp-IKO mice and was decreased following DSS treatment, particularly in control mice ( Figure 4B). There was a corresponding decrease in ZO-1 expression in the colon upon DSS treatment with similar responses in control and Mttp-IKO mice ( Figure 4C).
We also examined parameters of the unfolded protein response (UPR) and integrated stress response pathways in the small intestine and colon of both genotypes and the impact of DSS exposure. Those findings demonstrated increased baseline expression of Grp78 and Chop in the small intestine of Mttp-IKO mice ( Figure 4B), as previously noted [22]. However, DSS treatment failed to produce a further increase in any of these stress markers, but rather tended to reduce expression of Grp78 and Chop in the small intestine of Mttp-IKO mice ( Figure 4B). By contrast, there were no baseline changes in UPR mRNAs in the colon of Mttp-IKO mice, but DSS treatment resulted in increased abundance of Xbp-1s in both genotypes ( Figure 4C). These findings suggest that there is no baseline UPR response in the colon of Mttp-IKO mice. To examine the possibility that an adaptive ER stress response might influence the injury phenotype observed in Mttp-IKO mice, we treated mice with the chemical chaperone tauroursodeoxycholic acid (TUDCA) since previous studies have shown that this strategy attenuates the UPR in metabolic stress conditions in mice [22,23]. These findings reveal that TUDCA treatment failed to rescue the severe phenotype of DSS treated Mttp-IKO mice and if anything led to accelerated death in these animals ( Figure 4D). These findings together lead us to conclude that the features of ER stress and the UPR noted in the small intestine of Mttp-IKO mice are adaptive in nature and that abrogation of this response leads to worse injury.

Enhanced colonic inflammation in Mttp-IKO mice in response to DSS injury
We observed the expected increase in expression of several cytokine mRNAs in the colon of DSS treated mice, including IL-1b, IL-17A and Nlrp3, with comparable induction in both genotypes ( Figure 5A). However we observed a greater increase in TNFa expression in the colon of DSS treated Mttp-IKO mice compared to controls ( Figure 5A). We also examined serum levels of LPS, IL-1b and TNFa, which have been previously associated with intestinal injury in mice treated with DSS [14,24]. Those findings revealed that serum TNFa levels were consistently elevated in Mttp-IKO mice, both at baseline and following DSS treatment while serum levels of LPS and IL-1b were more variable ( Figure 5B-D).

Colonic injury and inflammation is associated with increased susceptibility to CAC in Mttp-IKO mice
A substantial body of data suggests that inflammation is a major driver of tumorigenesis and work suggests a causative link mediated via pathways involving inflammatory cytokine signaling [25,26]. Given the propensity of Mttp-IKO mice to manifest a more severe phenotype in response to DSS mediated injury, we asked whether this increased colonic inflammation was associated with an altered colonic tumor burden following administration of azoxymethane followed by three cycles of DSS ( Figure 6A). The findings reveal greater tumor multiplicity ( Figure 6B) and overall tumor burden ( Figure 6C, D) (verified by pathologic examination by a pathologist blinded to genotype) in Mttp-IKO mice in association with increased cellular proliferation as evidenced by BrdU incorporation (Figure 6E, F).

Inflammatory signals coupled with altered fatty acid signaling drive growth mediator secretion in Mttp-IKO mice
In order to understand in more depth the mechanisms underlying the proliferative phenotype observed in the colon of Mttp-IKO mice in response to AOM/DSS, we again examined the expression patterns of IL-1b and TNFa. We observed increased mRNA abundance of IL-1b in Mttp-IKO mice ( Figure 7A) and a trend to increased IL-1b protein expression ( Figure 7B). The expression of TNFa mRNA was comparable by genotype but we observed a significant increase in TNFa protein levels in colonic tissue from Mttp-IKO mice ( Figure 7B). These findings again support the hypothesis that an altered inflammatory response likely contributes to the increased tumor burden observed in Mttp-IKO mice.
To explore the mechanisms at play in this response, we undertook an examination of possible mediators of this proproliferative phenotype. Transcript profiling revealed no significant changes in mRNA expression of several candidate mediators ( Figure 7A), but there was a trend to increased amphiregulin protein expression and an increase in phospho-Akt abundance ( Figure 7C, D). These findings suggested to us that the secretion of mediators from intestinal myofibroblasts might play a role in growth modulation via paracrine pathways. In support of this possibility there is work showing a role for stromal myofibroblast production of amphiregulin and epiregulin in promoting colon tumor growth in the setting of inflamed human colon [27]. Our approach was guided by earlier findings demonstrating that small intestinal epithelial cells from Mttp-IKO mice undergo lipotoxic injury following short term high fat feeding [22]. Based on those observations, we suspected that the sustained growth promoting effects of intestinal Mttp deletion were potentially mediated by myofibroblasts.
In order to explore this possibility in more detail, we turned to primary murine intestinal myofibroblasts in order to examine amphiregulin and epiregulin regulation. We first established that fecal lipid content was increased as a result of fat malabsorption in Mttp-IKO mice, as previously documented [11]. In particular, we demonstrated increased fecal free fatty acid (FFA) abundance as well as increased content of cholesterol and phospholipid in Mttp-IKO mice ( Figure 7E). We determined that the fecal FFA concentration in control mice (,60 mmol/g) was similar to that reported recently [28]. We focused our attention on the abundant saturated fatty acids, selecting palmitic acid as an initial target in which to examine the regulation of both amphiregulin and epiregulin [29]. Our findings demonstrate striking upregulation of amphiregulin and epiregulin mRNA in primary murine intestinal myofibroblasts by 200 mM palmitate ( Figure 7E). Furthermore, the known induction of amphiregulin and epiregulin by TNFa and IL-1b [30] was further enhanced by the inclusion of palmitate ( Figure 7E). These findings suggest that fatty acids play a role in modifying growth responses in conjunction with other known mediators. In line with this suggestion we demonstrated increased small intestinal expression of amphiregulin expression in Mttp-IKO mice following exposure to a high fat western diet ( Figure 7F, G).   Taken together, the findings strongly suggest that increased colonic fatty acid availability, produced by defective small intestinal chylomicron assembly and the consequent block in normal lipid absorption, exacerbates inflammatory and other adaptive pathways that increase colonic proliferation.

Discussion
The findings in this report indicate that blocking chylomicron secretion from the small intestine is associated with increased severity of chemical (DSS) colitis and leads to increased tumor burden in a model of CAC. The mechanisms and pathways mediating these phenotypes include alterations in intestinal barrier function with increased permeability and increased serum levels of TNFa, coupled with activation of myofibroblast secretion of growth factors mediated at least in part through altered fatty acid signaling. Together the findings link aspects of intestinal lipid metabolism with intestinal inflammation and also to pathways that promote development of colorectal cancer. Some of these core observations merit more detailed discussion.
Previous studies have implicated lipid-derived factors secreted from the small intestine as key components in mediating distal organ failure in systemic sepsis [8,9] although the precise lipid components and lipoprotein particles involved (chylomicrons, low or high density lipoprotein) are yet to be determined. Other work has demonstrated that bacterial LPS absorption by the small intestine involves chylomicron formation as evidenced by diminished serum LPS levels in animals treated with Pluronic L-81, which functions within enterocytes to block chylomicron formation [31]. Here we observed that serum LPS levels tended to increase following DSS treatment but we found no differences between control and Mttp-IKO mice. By contrast, we found elevated serum levels of TNFa in Mttp-IKO mice, both at baseline and following DSS treatment, suggesting that blocking chylomicron formation was associated with increased production or altered turnover of this proinflammatory cytokine in the setting of acute chemical colitis. Further work will be required to resolve the cellular origins of TNFa and to address the important questions of production and clearance rates of this cytokine in DSS mediated colonic injury. Nevertheless, our findings are consistent with other work that has implicated lipid-dependent alterations in intestinal delivery of dietary antigens that modulate systemic responses, including the release of inflammatory mediators from adipose tissue [32,33].
Our recent findings demonstrated that Mttp-IKO mice were protected against pneumonia-associated sepsis induced intestinal injury and exhibited decreased mortality [12], which led us to speculate that blocking intestinal chylomicron secretion might actually be protective against other models associated with systemic inflammations. However, our findings demonstrated the opposite conclusion. Here we show that Mttp-IKO mice have a far worse outcome when challenged in a model of chemical (DSS) colitis. The apparent discordance in these outcomes lead us to suggest that blocking chylomicron assembly and secretion may protect animals from the secondary manifestations of intestinal injury only when the source of sepsis involves extraintestinal tissues (such as the lung). In the setting of pneumonia-associated sepsis, for example, Mttp-IKO mice exhibit a compensatory increase in hepatic HDL production [12]. By contrast, the current findings strongly imply that blocking chylomicron formation and decreasing dietary lipid absorption exacerbates the systemic effects of inflammatory stimuli associated with colonic injury.
Another consideration in attempting to understand the pathways by which blocking chylomicron secretion exaggerates the inflammation associated with colonic injury is that the genetic pathways involved in Mttp-dependent lipid delivery into the endoplasmic reticulum [34] overlap with Mttp-dependent pathways involved in lipid antigen processing. In this alternative setting, Mttp has been shown to play a key role in lipidating CD1 antigen presenting family members [35]. This is relevant because lipid antigen loading onto CD1d molecules is critical for presentation to natural killer T (NKT) cells [36]. In relation to the current findings, it is worth noting that human subjects with abetalipoproteinemia (ABL) (in whom the MTTP gene is defective) have been shown to exhibit defective CD1 function as a result of impaired lipid antigen loading [37]. It is also worth noting that while the incidence of colorectal cancer in ABL subjects is unknown, there is a case report of ileal adenocarcinoma in a patient with mild ABL [38]. Thus a confluence of findings suggest that Mttp may function in pathways beyond dietary lipid absorption and might conceivably play a role in the pathogenesis of intestinal inflammatory and protective processes mediated by CD1 dependent pathways [39]. In keeping with this suggestion, studies demonstrated that Mttp deletion was protective against experimental oxazalone-induced colitis, which is dependent on the presumptively haptenated antigen being presented via invariant NKT cells [40]. Those studies however did not address the specific role of intestinal epithelial Mttp deletion, but rather used an Mx-1 (interferon-inducible) Cre [40]. The question of whether intestinespecific Mttp deletion (as in the current study) exhibits the same protective phenotype as that observed using the interferon inducible system employed in those earlier studies will await clarification.
The current findings also raise the question of how dietary fat and intestinal lipid metabolism might influence inflammation and pathways relevant to the pathogenesis of colorectal cancer. It has been shown for example that mice chronically fed high fat diets exhibit increased mortality in some models of sepsis [41] while other studies demonstrated that high fat diets induce low grade endotoxemia in both experimental animals and in studies using intestinal cell culture [42]. Recent studies in otherwise healthy human subjects has further shown that feeding a western-style high fat diet produces features of systemic inflammation, including endotoxemia [43]. In keeping with these findings, other studies in obese subjects demonstrated that diet-induced weight loss led to reductions in serum cytokines associated with systemic inflammation (including TNFa, IL-1b), [44]. Taken together, the findings from in vitro studies together with studies using experimental animal models as well as human studies reinforce the concept that strategies including reducing dietary fat intake may attenuate the pro-inflammatory state associated with obesity and in turn reduce the risk of colorectal cancer. It bears emphasis that the current studies were undertaken in two genotypes of mice fed low fat rodent chow, suggesting that the increased fatty acid flux in Mttp-IKO mice resulted from the unabsorbed lipid rather than from increased dietary fat intake or from obesity (since Mttp-IKO mice are lean), or from increased colonic accumulation of neutral lipid. In this regard, we validated fecal FFA content in control, chow-fed mice (,60 mmol/g, Figure 7E) was similar to that recently reported [28]. In addition, those authors demonstrated that palmitic acid represented the dominant saturated FA species found in feces [28]. Extrapolating from their findings, we estimate that palmitic acid is present in control, chow-fed mouse feces at a concentration of ,20 millimolar [28]. These are important considerations in evaluating the effects of palmitic acid as used in the current studies. We selected a concentration (200 mM) previously shown in cell models to stimulate lipotoxic adaptive responses [29]. Our findings demonstrate a significant induction of epiregulin and amphiregulin expression in intestinal myofibroblasts, even at FFA concentrations log orders lower than found within the colonic lumen. Based upon these findings, it would be of interest to examine the effects if any of short-term exposure to high fat diets on colorectal cancer development or progression in Mttp-IKO mice.
Finally, our findings raise the possibility that there may be other adverse effects of Mttp deletion beyond lipid malabsorption. Pharmacologic compounds that block Mttp activity are currently in clinical trials, including some that are by design intestinespecific [45][46][47][48][49]. Our findings raise the concern that human subjects consuming these agents might be more susceptible to inflammatory conditions affecting the intestine and potentially also at greater risk of developing colorectal tumors.