Int J Colorectal Dis (2015) 30:1247–1254 DOI 10.1007/s00384-015-2263-1
ORIGINAL ARTICLE
Surgical diverticulitis is not associated with defects in the expression of wound healing genes Tara M. Connelly 1 & Arthur S. Berg 2 & Leonard R. Harris III 1 & Rafel Tappouni 3 & Dave Brinton 1 & Sue Deiling 1 & Walter A. Koltun 1
Accepted: 14 May 2015 / Published online: 24 May 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose The development of diverticuli may represent defects in collagen vascular tissue integrity possibly from a genetic predisposition. We evaluated the tissue expression of wound healing genes in sigmoid tissue from youthful patients undergoing surgery for diverticulitis and thus would more likely suffer from a genetic predisposition (SD mean age 39 ±0.9) versus controls in the form of patients over the age of 50 (mean age 52.9±10.5 years) without evidence of diverticular disease. Methods The mRNA expression of 84 genes associated with the extracellular matrix, cellular adhesion, growth factors, inflammatory cytokines, and signal transduction was evaluated in 16 SD and 15 control tissues using a Qiagen™ Wound Healing Array. Vitronectin, the gene protein with the highest potential significance on raw analysis, was further investigated using a Taqman assay with an additional 11 SD (total n= 27) and four control (total n=19) samples. Statistics were by Student’s t and Mann-Whitney tests with Bonferroni correction.
Poster presentation at the 2014 ASCRS meeting, May 17-21, 2014, Hollywood, FL * Walter A. Koltun [email protected] 1
Department of Surgery, Division of Colon and Rectal Surgery, College of Medicine, The Pennsylvania State University, Hershey, PA 17033-0850, USA
2
Department of Public Health Sciences, College of Medicine, The Pennsylvania State University, Hershey, PA, USA
3
Department of Radiology, Wake Forest Baptist Medical Center, Winston-Salem, NC, USA
Results No significant differences in mRNA expression between the SD and control tissue in the 84 measured genes were demonstrated after correction. Vitronectin mRNA expression was downregulated 2.7-fold in SD tissue vs. tissue from non-neoplastic control patients (p = 0.001 raw/0.08 corrected). However, on vitronectin TaqMan analysis, no difference in expression was seen in SD vs. all controls or in all subset comparisons. Conclusions The lack of significant alteration in mRNA expression of traditionally associated wound healing genes/ proteins in young SD patients suggests that such genes play a minor role in the genetic predisposition to youthful diverticulitis. Keywords Diverticulitis . Surgical genetics . Wound healing . Vitronectin . Diverticular disease
Introduction Diverticulitis has been conventionally thought of as a disease of old age that is due to both the loss of colonic compliance and increased intraluminal pressure secondary to constipation. Thus, lifestyle factors including diet and obesity have traditionally been the most commonly studied predictors of disease development [1–7]. The small studies performed to date have attempted to provide a better understanding of the pathophysiology of the disease by studying colonic wall physiology and the matrix metalloproteinases (MMPs) within the wall [8–13]. Our group has previously provided evidence for a possible tissue integrity defect in diverticulitis patients by demonstrating an increased risk of incisional hernia after surgery in diverticulitis patients when compared to cancer patients undergoing sigmoid colectomies [14].
1248
The observations of multiple affected family members, often with early onset of diverticular disease (DD) and the concomitant presence of both diverticular disease and genetic connective tissue disorders such as Ehlers-Danlos and polycystic kidney disease, suggest a genetic predisposition for DD [15–21]. We hypothesize that the development of diverticulosis is due to an environmental trigger (such as constipation) superimposed on a background of host genetic susceptibility, possibly a defect in genes associated with collagen integrity, the extracellular matrix, and/or tissue repair. Some, but not all, patients with diverticuli develop diverticulitis. We also hypothesize that the progression from diverticulosis to then the infectious process of diverticulitis (and then again to diverticulitis needing surgical intervention) requires additional environmental factors (e.g., bacteria) as well as additional host genetic defects, possibly in the immune system. We have previously demonstrated an association between diverticulitis requiring surgery (SD) with a single nucleotide polymorphism (SNP) in the immunoregulatory gene TNFSF15 [22]. It is very likely that other genes are involved in what is probably a complex genetic disorder involving several biological pathways. Patients diagnosed with DD at a young age typically have a difficult disease course. A common factor in such patients is a family history of DD which further suggests a genetic component to the disease. These youthful patients are understudied but may represent a study cohort that would have the greatest probability of discovery of genes affiliated with DD. The discovery of a genetic correlate for DD would focus further disease investigation into specific biologic pathways and may also assist in medical and/or surgical decisionmaking. The focus of the present study was to compare mRNA expression of known wound healing-associated genes in full-thickness sigmoid tissue from youthful patients undergoing surgical resection for diverticulitis (where a genetic defect would be most likely) and control sigmoid tissue from patients without DD, to further understand the likely underlying connective tissue aspect of the development of diverticulosis. Control patients were ones without evidence of any DD (diverticulosis or diverticulitis). Since such disease usually presents at an elderly age, to ensure the greatest likelihood of a group of patients without a genetic predisposition to DD, older patients were chosen as controls, since youthful or agematched controls could very possibly contain patients who later would develop DD.
Materials and methods A Qiagen™ Human Wound Healing RT2 Profiler™ Array (SABiosciences, Valencia, CA catalog number PAHS-121Z) was used according to the manufacturer’s recommended
Int J Colorectal Dis (2015) 30:1247–1254
protocol. The array measures the mRNA from 84 genes associated with extracellular matrix/cellular adhesion, growth factors, inflammatory cytokines, and signal transduction (Table 1) [23]. For the Wound Healing Array analysis, full-thickness sigmoid surgical specimens from 16 unrelated youthful SD patients (mean age 37.6±0.9 years, 11 plegmaceous or recurrent disease, 5 perforating or abscessing disease) and 15 control patients undergoing a sigmoid resection (mean age 52.9± 10.5 years, comprised of 2 dysmotility, 1 volvulus, 1 endometriosis, 3 familial adenomatous polyposis (FAP), 1 hereditary nonpolyposis colorectal cancer, and 7 colorectal cancer (CRC)) were selected from the Hershey Medical Center Division of Colon and Rectal Surgery’s Biobank. The absence of diverticular disease in all control patients was confirmed on preoperative imaging (CT scan or barium enema) by a specialist gastrointestinal radiologist and again on operative specimen pathology, at the time of surgery. The mRNA identified as potentially most significantly altered on the Wound Healing Array (vitronectin (VTN), see BResults^) was further specifically studied using a TaqMan gene expression assay (Applied Biosystems, Hs00940758_g1) using 100 ng of cDNA. For the TaqMan assay, an additional 11 youthful SD patients (mean age 39.5 ±6.1 years, 5 male) and four new control (mean age 64.4±7.9, 4 male, 1 unresectable polyp, 1 dysmotility, and 2 rectal cancers) patients were recruited. These samples were added to the original 31 samples (16 SD, 15 controls) used in the first Wound Healing Array experiment to yield a total of 27 SD and 19 control tissues for the subsequent VTN TaqMan assay study. Patient demographics are shown in Table 2. Patient comparison groups are show in Table 3. For both the Wound Healing Array and the VTN TaqMan assay, full-thickness sigmoid tissue was harvested at the time of surgical resection with the assistance of the operating surgeon. In the cases of cancer and/or polyps, normal-appearing tissue adjacent to the lesion was taken. The specimens were placed immediately in RNAlater (Ambion, Cat No. AM7021) and flash frozen in liquid nitrogen. Subsequent tissue pulverization was performed using a Biospecs MultiSample BioPulverizer (Cat. No. 59012MS). A TRIzol (Ambion, Cat. No. 15596018)/RNeasy Mini Kit (Qiagen, Cat. No. 74104) hybrid RNA extraction protocol was used to isolate RNA from the pulverized samples. Total isolated RNA quality was assured using an Agilent Bioanalyzer 2100. All samples included in the study had an RNA integrity number (RIN) value ≥7. The SuperScript III First-Strand kit (Invitrogen, Cat. No. 18080-051) using 400 ng of total RNA was used for cDNA conversion according to the manufacturer’s recommended protocol. All plates were read on an Applied Biosystems 7900HT Real-Time PCR System. RQManager (Applied Biosystems) software was
Int J Colorectal Dis (2015) 30:1247–1254 Table 1
Qiagen™ Human Wound Healing RT2 Profiler™ Array pathways and genes
Extracellular Remodeling enzymes matrix components COL14A1 COL1A1 COL1A2 COL3A1 COL4A1 COL4A3 COL5A1 COL5A2 COL5A3 VTN
1249
CTSG CTSK CTSV F13A1 F3 (tissue factor) FGA (fibrinogen) MMP1 MMP2 MMP7 MMP9 PLAT (tPA) PLAU (uPA) PLAUR (uPAR) PLG SERPINE1 (PAI-1) TIMP1
Cellular adhesion
Cytoskeleton
Inflammatory cytokines Growth factors and chemokines
CDH1 (E-cadherin) ITGA1 ITGA2 ITGA3 ITGA4 ITGA5 ITGA6 ITGAV ITGB1 ITGB3 ITGB5 ITGB6
ACTA2 (a-SMA) ACTC1 RAC1 RHOA TAGLN
CCL2 (MCP-1) CCL7 (MCP-3) CD40LG (TNFSF5) CXCL1 CXCL11 (ITACIP-9) CXCL2 CXCL5 (ENA-78/LIX) IFNG IL10 IL1B IL2 IL4 IL6
used for TaqMan result interpretation and RT2 Profiler™ PCR array data analysis software (Qiagen, Valencia, CA).
ANGPT1 CSF2 (GM-CSF) CSF3 (GCSF) CTGF EGF FGF10 FGF2 FGF7 HBEGF (DTR) HGF IGF1 MIF PDGFA TGFA TGFB1 TNF VEGFA
Signal transduction
TGFβ: TGFB1 TGFBR3 STAT3 WNT: CTNNB1 WISP1 WNT5A Phosphorylation: MAPK1 (ERK2) MAPK3 (ERK1) PTEN Receptors: EGFR IL6ST (GP130) Other: PTGS2
Results Wound Healing Array
Statistical analysis Multiple group comparisons were made in both the Wound Healing Array (n=16 SD vs. 15 controls) and the subsequent focused VTN TaqMan gene expression assay (n=27 SD vs. 19 controls) cohorts (see Table 3): The Student’s t test was used for analysis of the Wound Healing Array using RT2 Profiler™ PCR array data analysis software (Qiagen, Valencia, CA). The Mann-Whitney test was used for subsequent statistical analysis of VTN using R software. A Bonferroni correction was applied to account for multiple comparisons of 84 genes in the Wound Healing Array.
None of the 84 genes on the array were upregulated greater than 1.5 times in the 16 SD patients’ tissue when compared to all control tissues (n= 15). However, several genes were downregulated in the SD tissue vs. all control tissue (Fig. 1). The most marked downregulation was seen in colonystimulating factor 3 (CSF3) and prostaglandin-endoperoxide synthase 2 (PTGS2). However, this dysregulation was not statistically significant. Only chemokine (CXC motif) ligand 2 (CXCL2) and CSF3 were found to be significantly downregulated on raw analysis. However, both comparisons lost significance after Bonferroni correction. The most potentially significant change found in any of the comparisons performed was seen in the SD vs. the four non-
Table 2 Patient demographics Diverticulitis array cohort n=16
Additional diverticulitis patients for gene candidate (VTN) study
Array controls n=15
n=11
Additional controls for gene candidate (VTN) study n=4
Age at diagnosis Years, SD Range (years) Male Smoking status Current Former Never Unknown
37.6±4.9 29.8–48.3 15 4 5 6 1
39.5±6.1 27.4–48.6 5 3 4 3 1
52.9±10.5 40.3–73.6 7 3 5 6 1
64.4±7.9 50.3–70.4 4 0 2 2 0
1250 Table 3 Patient subset comparisons
Int J Colorectal Dis (2015) 30:1247–1254
Cohort comparisons
Primary analysis using 84-gene array
Targeted gene (VTN) study
Number of patients
Number of patients
SD vs. all non diverticulitis controls* SD vs. cancer only
16 vs. 15 16 vs. 7
27 vs. 19 27 vs. 9
SD vs. non-neoplastic controls** Perforating vs. non-perforating SD
16 vs. 4 11 vs. 5
27 vs. 8 19 vs. 8
*Non-diverticular controls: cancer+HNPCC+FAP+dysmotility+volvulus+endometriosis **Non-neoplastic controls: dysmotility+volvulus+endometriosis
neoplastic control (constipation, dysmotility, endometriosis) tissue comparisons (Fig. 2). The VTN gene was found to be downregulated 2.7× relative to the non-neoplastic tissue (p= 0.001, prior to Bonferroni correction, p=0.08 after correction). Although several other genes were downregulated at least 2 fold in the SD tissue, no significant difference was demonstrated after Bonferroni correction. In the SD vs. cancer tissue (n=7) comparison, CSF3 expression level was significantly downregulated (3.4-fold) in the SD tissue (p=0.01) but significance was lost after Bonferroni correction (Fig. 3). The greatest difference in expression was seen when comparing the perforating SD (n=5) tissue vs. the non-perforating SD tissue (n=11) (Fig. 4). Matrix metalloproteinase 7 (MMP7) was downregulated 3.9× in the perforating tissue (p=0.037 before correction, NSD after correction).
upregulated in the non-perforating tissue when compared to controls. However, none of these comparisons reached statistical significance.
Discussion
With the addition of 15 patients as described above (11 SD and 4 controls), candidate gene expression study for the VTN gene using a TaqMan array demonstrated no difference in mRNA expression in SD vs. controls. In fact, expression was nearly identical in the two groups (Table 4). On subgroup analysis, VTN expression was downregulated in tissue with perforating diverticular disease when compared to both control and non-perforating diverticular tissue. Conversely, it was
This study utilized an array to study the mRNA expression of 84 genes found in pathways that may intuitively be associated with DD. We chose to study patients with youthful disease onset. Our control population had to be older patients to be sure that they would not have a genetic predisposition to diverticuli which typically occur later in life and may simply be a phenotypic variant of the same genetic defect. It would not be appropriate to compare a group of 30-year-olds with diverticulitis with a group of 30-year-olds without diverticulitis since approximately half of such Bcontrols^ would eventually develop DD by the age of 60 and thus could possibly contain the same underlying genetic defect as in the studied group. The study design used such an array to cast a Bwide net^ for the identification of potential gene targets for further study. Similarly, several comparisons between subset groups were evaluated. No obvious differences were seen, except for in the case of VTN, which suggested itself as a potential gene candidate. However, further evaluation with additional patients similarly showed no statistical difference. First, the gene
Fig. 1 Genes with downregulated mRNA expression on Qiagen™ Human Wound Healing RT 2 Profiler ™ Array in all surgical diverticulitis (n=16) vs. all control (n=15) tissues. CXCL2 chemokine (C-X-C motif) ligand 2, FGF10 fibroblast growth factor 10, WISP
WNT1-inducible-signaling pathway protein 1, IL6 interleukin 6, ACTA2 alpha-actin-2, ACTC1 alpha cardiac muscle 1, TAGLN transgelin, CSF3 colony-stimulating factor 3 (granulocyte), PTGS2 prostaglandinendoperoxide synthase 2, VTN vitronectin
VTN study
Int J Colorectal Dis (2015) 30:1247–1254
1251
Fig. 2 Genes with differential regulation of mRNA expression on Qiagen™ Human Wound Healing RT2 Profiler™ Array in all surgical diverticulitis patient (n = 16) vs. healthy control (n = 4) tissues (2 dysmotility, 1 volvulus, 1 endometriosis). FGF2 fibroblast growth factor 2, ANGPT1 angiopoietin 1, FGF7 fibroblast growth factor 7,
ITGB3 integrin beta-3, EGF epidermal growth factor, ACTC1 alpha cardiac muscle 1, ITGA integrin alpha 1, WISP WNT1-induciblesignaling pathway protein, ACTA2 alpha-actin-2, VTN vitronectin, PTGS2 prostaglandin-endoperoxide synthase 2, TAGLN transgelin
expression of all non-diverticular disease controls was compared to the diverticulitis patients. As this was a heterogeneous population, we also examined cancer patients separately. A potential overlap between genes expressed in the cancer patients and diverticulitis patients led us to exclude these patients in the additional analysis and study the expression from the four noncancer patients vs. diverticulitis patients. Lastly, we hypothesized that perforating and non-perforating DD
may have a differing genetic basis. Thus, we compared gene expression between these multiple groups; several genes on this array were found to have differential expression in the diverticulitis patients vs. control tissue on raw statistical analysis. Although statistical significance was lost in all comparisons after Bonferroni correction, changes associated with the VTN gene appeared to have the most significant raw p value. Thus, this gene was selected for further specific study.
Fig. 3 Genes with downregulated mRNA expression on Qiagen™ Human Wound Healing RT 2 Profiler ™ Array in all surgical diverticulitis (n=16) vs. cancer (n=7) tissues. FGF10 fibroblast growth factor 10, ITGB3 integrin beta-3, COL5A type V collagen, IL6 interleukin
6, ACTA2 alpha-actin-2, ACTC1 alpha cardiac muscle 1, TAGLN transgelin, PTGS2 prostaglandin-endoperoxide synthase 2, CSF3 colony-stimulating factor 3 (granulocyte)
1252
Int J Colorectal Dis (2015) 30:1247–1254
Fig. 4 Genes with differential mRNA expression on Qiagen™ Human Wound Healing RT2 Profiler ™ Array in all perforating surgical diverticulitis (n=5) vs. non-perforating surgical diverticulitis (n=11) tissue. MMP7 matrix metalloproteinase 7, CSF3 colony-stimulating
factor 3 (granulocyte), MMP1 matrix metalloproteinase 1, CTSV cathepsin V, SERPINE1 serpine pepsidase inhibitor/plasminogen activator type 1, F3 coagulation factor III, VTN vitronectin, PTGS2 prostaglandinendoperoxide synthase 2, PLG plasminogen
When designing the second half of the experiment looking specifically at VTN using the TaqMan array, we chose to include the tissue that was already studied on the wound healing array but added newly recruited patients for additional power. Although not a traditional discovery+confirmatory cohort design, this was done due to relatively low overall patient numbers, and with this larger cohort, statistical significance was not achieved. The majority of previous functional studies on diverticulitis have looked at the collagen, elastin, and proteoglycancontaining extracellular matrix (ECM) which provides the integrity and flexibility of the colonic wall [11, 24]. VTN is a glycoprotein found in the ECM and has a role in cell adhesion as well as the prevention of lysis by the complement pathway and hemostasis [25]. It may also assist in promoting inflammation as a host defense against infection [26]. The matrix metalloproteinase (MMP) family of zinc-containing endopeptidases degrades the ECM, while tissue inhibitors of MMPs (TIMPs) inhibit them and thus prevent degradation. The MMPs and TIMPs are the most studied physiologic components of the colon wall in diverticulitis patients. Using rtPCR, Mimura et al. demonstrated no difference in MMP1, 2, and 9 levels between tissues from the diverticulum of six cancer specimens with asymptomatic diverticular disease, tissue adjacent to an abscess or stricture in 5 Bcomplicated
diverticulosis,^ and 11 colorectal cancer control tissues, similar to our study. However, TIMP1 and 2 mRNA transcripts were higher in tissue affected by diverticular disease vs. control tissue with a significant difference seen in the muscularis propria [12]. Lakatos et al. specifically examined MMP9 in tissue from patients with another inflammatory colonic pathology, ulcerative colitis (UC), and found expression levels were more than 4× higher in UC patients than in DD patients [27]. Similarly, Altadill et al. showed a downregulation of MMP2, MMP9, and MMP13 expression in inflamed diverticulitis vs. Crohn’s disease (CD) mucosa in over 50 patients using rtPCR, immunohistochemistry, and western blots. However, higher expression of MMP1, TIMP1, and TIMP3 was seen in both inflamed and noninflamed diverticulitis mucosa vs. CD [9]. In the present study, the MMP genes demonstrated significant downregulation on raw analysis in the penetrating vs. nonpenetrating comparisons. However, in all other comparisons, fold change was less than 1.3 (SD vs. all controls, SD vs. healthy controls, SD vs. cancer tissue). The one TIMP included on our array, TIMP1, did not show significant differential regulation on raw analysis on any of our comparisons, conflicting with these other studies [9, 12]. Signaling and inflammatory pathways in diverticulitis have been less studied. One study evaluated histamine receptor expression in over 100 diverticulitis patients (57 complicated, 44 non complicated) and sigmoid cancer controls. A significant association with complicated diverticulitis and Ban allergic predisposition^ (pollen, food, pets, etc.) was noted, and a difference in H1R and H2R expression between healthy and diverticulitis patient tissues was demonstrated [28]. Costedio et al. performed immunohistochemistry, rtPCR, and enzyme immunoassay kit for 5HT in stimulated and unstimulated tissues to study the different genes in the serotonin signaling pathway in sigmoid tissue from 22 healthy controls, 16
Table 4
VTN TaqMan mRNA expression comparisons
Comparison
Fold change
p value
All SD vs. controls Perforating SD vs. controls Non-perforating SD vs. controls Perforating vs. non-perforating
1.03 −0.52 1.4 −3.1
0.50 0.23 0.94 0.12
Int J Colorectal Dis (2015) 30:1247–1254
diverticulosis patients, and 13 diverticulitis patients with matched normal adjacent tissue and found decreased serotonin-selective reuptake transporter SERT (5HT transporter) in the sigmoid mucosa of patients with a history of diverticulitis when compared with controls [29]. How this relates to pathophysiology of diverticulitis is unclear and, like other observations, may be Bepiphenomenon.^ Limitations The major limitation to the present study is the low number of healthy control tissues. In order to study the colonic wall in its full thickness as was chosen for this project, full surgical resection specimens are required not just colonoscopic biopsies since a defect in the muscular wall of the colon (not the mucosa) lead to eventual diverticuli. Full-thickness colonic samples from healthy controls are difficult to obtain, especially from youthful patients. Future studies utilizing colonoscopy biopsy specimens may be warranted to increase patient numbers but may not capture dysregulated gene expression in the entire colonic wall. Another limitation is the slightly varied distance from our specimens to the abscess/perforation or even inflamed diverticuli as expression may vary in these different locations. However, all specimens were taken as close to the foci of disease as possible to minimize this effect. Another criticism is the heterogeneity of our control population. However, the main criteria for inclusion into this category was the absence of DD and an older age at resection was necessary, to ensure that diverticulosis would not develop later in life, as may occur and confound the results, if a younger control population was chosen. Results could be confounded due to the inclusion of the five patients with perforating disease. Perforating disease is an acute infective process with macrophage infiltration and acute inflammation which may alter results. This was noted in de Vries’ study where dectin, a pattern recognition receptor expressed on myeloid cells, showed increased expression in tissue with severe vs. mild diverticulitis. The authors noted that this increased expression may likely be due to macrophage expression, on which dectin is present [30]. To address this, in the present study, the expression between the perforating and non-perforating diverticulitis tissues was compared. No difference was seen between mRNA expressions in nonperforating vs. perforating DD in any of the 84 genes after Bonferroni correction. However, the number of samples with perforating disease overall was low.
1253
This suggests that extracellular matrix and growth factorassociated genes probably play a minor role in the possible genetic defects predisposing youthful individuals to diverticulitis. Acknowledgments Division of Colon and Rectal Surgery is the recipient of the Carlino Fund for IBD Research. Conflict of interest The authors declare that they have no competing interests.
References 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
Conclusion The current project demonstrated a lack of significant mRNA dysregulation of over 80 genes traditionally associated with wound healing and tissue integrity in young patients with SD.
13.
Strate LL, Liu YL, Aldoori WH, Syngal S, Giovannucci EL (2009) Obesity increases the risks of diverticulitis and diverticular bleeding. Gastroenterology 136:115–122. doi:10.1053/j.gastro.2008.09. 025 Zaidi E, Daly B (2006) CT and clinical features of acute diverticulitis in an urban U.S. population: rising frequency in young, obese adults. AJR Am J Roentgenol 187:689–694 Rosemar A, Angeras U, Rosengren A (2008) Body mass index and diverticular disease: a 28-year follow-up study in men. Dis Colon Rectum 51:450–455 Manousos O, Day NE, Tzonou A, Papadimitriou C, Kapetanakis A, Polychronopoulou-Trichopoulou A, Trichopoulos D (1985) Diet and other factors in the aetiology of diverticulosis: an epidemiological study in Greece. Gut 26:544–549 Aldoori WH, Giovannucci EL, Rimm EB, Wing AL, Trichopoulos DV, Willett WC (1994) A prospective study of diet and the risk of symptomatic diverticular disease in men. Am J Clin Nutr 60:757– 764 Aldoori WH, Giovannucci EL, Rimm EB, Ascherio A, Stampfer MJ, Colditz GA, Wing AL, Trichopoulos DV, Willett WC (1995) Prospective study of physical activity and the risk of symptomatic diverticular disease in men. Gut 36:276–282 Crowe FL, Appleby PN, Allen NE, Key TJ (2011) Diet and risk of diverticular disease in Oxford cohort of European Prospective Investigation into Cancer and Nutrition (EPIC): prospective study of British vegetarians and non-vegetarians. BMJ 343:d4131. doi: 10.1136/bmj.d4131 Ulmer TF, Rosch R, Mossdorf A, Alizai H, Binnebosel M, Neumann U (2014) Colonic wall changes in patients with diverticular disease—is there a predisposition for a complicated course? Int J Surg. doi:10.1016/j.ijsu.2014.03.014 Altadill A, Eiro N, Gonzalez LO, Junquera S, González-Quintana JM, Sánchez MR, Andicoechea A, Saro C, Rodrigo L, Vizoso FJ (2012) Comparative analysis of the expression of metalloproteases and their inhibitors in resected Crohn’s disease and complicated diverticular disease. Inflamm Bowel Dis 18:120–130. doi:10. 1002/ibd.21682 Wess L, Eastwood MA, Wess TJ, Busuttil A, Miller A (1995) Cross linking of collagen is increased in colonic diverticulosis. Gut 37: 91–94 Stumpf M, Krones CJ, Klinge U, Rosch R, Junge K, Schumpelick V (2006) Collagen in colon disease. Hernia 10:498–501 Mimura T, Bateman AC, Lee RL, Johnson PA, McDonald PJ, Talbot IC, Kamm MA, MacDonald TT, Pender SL (2004) Upregulation of collagen and tissue inhibitors of matrix metalloproteinase in colonic diverticular disease. Dis Colon Rectum 47:371– 378, discussion 378-379 Stumpf M, Cao W, Klinge U, Klosterhalfen B, Kasperk R, Schumpelick V (2001) Increased distribution of collagen type III
1254 and reduced expression of matrix metalloproteinase 1 in patients with diverticular disease. Int J Color Dis 16:271–275 14. Pogacnik J, Messaris E, Deiling SM, Connelly TM, Berg AS, Stewart DB, McKenna KJ, Poritz LS, Koltun WA (2014) Increased risk of incisional hernia after sigmoid colectomy for diverticulitis compared with colon cancer. J Am Coll Surg 218:920– 928. doi:10.1016/j.jamcollsurg.2014.01.050 15. Claassen AT, Mourad-Baars PE, Mearin ML, Hilhorst-Hofstee Y, Gerritsen van der Hoop A (2006) Two siblings below the age of 20 years with diverticular disease. Int J Color Dis 21:190–191 16. Schlotthauer HL (1946) Familial diverticulosis of the colon: report of seven cases in one family of nine persons. Ann Surg 124:497– 502 17. Omojola MF, Mangete E (1988) Diverticula of the colon in three Nigerian siblings. Trop Geogr Med 40:54–57 18. Santin BJ, Prasad V, Caniano DA (2009) Colonic diverticulitis in adolescents: an index case and associated syndromes. Pediatr Surg Int 25:901–905. doi:10.1007/s00383-009-2472-1 19. Granlund J, Svensson T, Olen O, Hjern F, Pedersen NL, Magnusson PK, Schmidt PT (2012) The genetic influence on diverticular disease—a twin study. Aliment Pharmacol Ther. doi:10.1111/j.13652036.2012.05069.x 20. Manwaring M (2011) Diverticular disease: genetic, geographic and environmental aspects. Semin Colon Rectal Surg 22:148–153 21. Scheff RT, Zuckerman G, Harter H, Delmez J, Koehler R (1980) Diverticular disease in patients with chronic renal failure due to polycystic kidney disease. Ann Intern Med 92:202–204 22. Connelly TM, Berg AS, Hegarty JP, Deiling S, Brinton D, Poritz LS, Koltun WA (2014) The TNFSF15 gene single nucleotide polymorphism Rs7848647 is associated with surgical diverticulitis. Ann Surg 259:1132–1137. doi:10.1097/SLA.0000000000000232
Int J Colorectal Dis (2015) 30:1247–1254 23.
Qiagen. Human Wound Healing Array (2014) http://www. sabiosciences.com/rt_pcr_product/HTML/PAHS-121Z.html. Accessed April 24, 2014 24. Melendez-Zajgla J, Del Pozo L, Ceballos G, Maldonado V (2008) Tissue inhibitor of metalloproteinases-4. The road less traveled. Mol Cancer 7:85. doi:10.1186/1476-4598-7-85 25. Preissner KT, Seiffert D (1998) Role of vitronectin and its receptors in haemostasis and vascular remodeling. Thromb Res 89:1–21 26. Singh B, Su YC, Riesbeck K (2010) Vitronectin in bacterial pathogenesis: a host protein used in complement escape and cellular invasion. Mol Microbiol 78:545–560. doi:10.1111/j.1365-2958. 2010.07373.x 27. Lakatos G, Sipos F, Miheller P, Hritz I, Varga MZ, Juhász M, Molnár B, Tulassay Z, Herszényi L (2012) The behavior of matrix metalloproteinase-9 in lymphocytic colitis, collagenous colitis and ulcerative colitis. Pathol Oncol Res 18:85–91. doi:10.1007/s12253011-9420-9 28. von Rahden BH, Jurowich C, Kircher S, Lazariotou M, Jung M, Germer CT, Grimm M (2012) Allergic predisposition, histamine and histamine receptor expression (H1R, H2R) are associated with complicated courses of sigmoid diverticulitis. J Gastrointest Surg 16:173–182. doi:10.1007/s11605-011-1702-8, discussion 182 29. Costedio MM, Coates MD, Danielson AB, Buttolph TR 3rd, Blaszyk HJ, Mawe GM, Hyman NH (2008) Serotonin signaling in diverticular disease. J Gastrointest Surg 12:1439–1445. doi:10. 1007/s11605-008-0536-5 30. de Vries HS, Plantinga TS, van Krieken JH, Stienstra R, van Bodegraven AA, Festen EA, Weersma RK, Crusius JB, Linskens RK, Joosten LA, Netea MG, de Jong DJ (2009) Genetic association analysis of the functional c.714T>G polymorphism and mucosal expression of dectin-1 in inflammatory bowel disease. PLoS One 4, e7818. doi:10.1371/journal.pone.0007818
ORIGINAL ARTICLE
Surgical diverticulitis is not associated with defects in the expression of wound healing genes Tara M. Connelly 1 & Arthur S. Berg 2 & Leonard R. Harris III 1 & Rafel Tappouni 3 & Dave Brinton 1 & Sue Deiling 1 & Walter A. Koltun 1
Accepted: 14 May 2015 / Published online: 24 May 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose The development of diverticuli may represent defects in collagen vascular tissue integrity possibly from a genetic predisposition. We evaluated the tissue expression of wound healing genes in sigmoid tissue from youthful patients undergoing surgery for diverticulitis and thus would more likely suffer from a genetic predisposition (SD mean age 39 ±0.9) versus controls in the form of patients over the age of 50 (mean age 52.9±10.5 years) without evidence of diverticular disease. Methods The mRNA expression of 84 genes associated with the extracellular matrix, cellular adhesion, growth factors, inflammatory cytokines, and signal transduction was evaluated in 16 SD and 15 control tissues using a Qiagen™ Wound Healing Array. Vitronectin, the gene protein with the highest potential significance on raw analysis, was further investigated using a Taqman assay with an additional 11 SD (total n= 27) and four control (total n=19) samples. Statistics were by Student’s t and Mann-Whitney tests with Bonferroni correction.
Poster presentation at the 2014 ASCRS meeting, May 17-21, 2014, Hollywood, FL * Walter A. Koltun [email protected] 1
Department of Surgery, Division of Colon and Rectal Surgery, College of Medicine, The Pennsylvania State University, Hershey, PA 17033-0850, USA
2
Department of Public Health Sciences, College of Medicine, The Pennsylvania State University, Hershey, PA, USA
3
Department of Radiology, Wake Forest Baptist Medical Center, Winston-Salem, NC, USA
Results No significant differences in mRNA expression between the SD and control tissue in the 84 measured genes were demonstrated after correction. Vitronectin mRNA expression was downregulated 2.7-fold in SD tissue vs. tissue from non-neoplastic control patients (p = 0.001 raw/0.08 corrected). However, on vitronectin TaqMan analysis, no difference in expression was seen in SD vs. all controls or in all subset comparisons. Conclusions The lack of significant alteration in mRNA expression of traditionally associated wound healing genes/ proteins in young SD patients suggests that such genes play a minor role in the genetic predisposition to youthful diverticulitis. Keywords Diverticulitis . Surgical genetics . Wound healing . Vitronectin . Diverticular disease
Introduction Diverticulitis has been conventionally thought of as a disease of old age that is due to both the loss of colonic compliance and increased intraluminal pressure secondary to constipation. Thus, lifestyle factors including diet and obesity have traditionally been the most commonly studied predictors of disease development [1–7]. The small studies performed to date have attempted to provide a better understanding of the pathophysiology of the disease by studying colonic wall physiology and the matrix metalloproteinases (MMPs) within the wall [8–13]. Our group has previously provided evidence for a possible tissue integrity defect in diverticulitis patients by demonstrating an increased risk of incisional hernia after surgery in diverticulitis patients when compared to cancer patients undergoing sigmoid colectomies [14].
1248
The observations of multiple affected family members, often with early onset of diverticular disease (DD) and the concomitant presence of both diverticular disease and genetic connective tissue disorders such as Ehlers-Danlos and polycystic kidney disease, suggest a genetic predisposition for DD [15–21]. We hypothesize that the development of diverticulosis is due to an environmental trigger (such as constipation) superimposed on a background of host genetic susceptibility, possibly a defect in genes associated with collagen integrity, the extracellular matrix, and/or tissue repair. Some, but not all, patients with diverticuli develop diverticulitis. We also hypothesize that the progression from diverticulosis to then the infectious process of diverticulitis (and then again to diverticulitis needing surgical intervention) requires additional environmental factors (e.g., bacteria) as well as additional host genetic defects, possibly in the immune system. We have previously demonstrated an association between diverticulitis requiring surgery (SD) with a single nucleotide polymorphism (SNP) in the immunoregulatory gene TNFSF15 [22]. It is very likely that other genes are involved in what is probably a complex genetic disorder involving several biological pathways. Patients diagnosed with DD at a young age typically have a difficult disease course. A common factor in such patients is a family history of DD which further suggests a genetic component to the disease. These youthful patients are understudied but may represent a study cohort that would have the greatest probability of discovery of genes affiliated with DD. The discovery of a genetic correlate for DD would focus further disease investigation into specific biologic pathways and may also assist in medical and/or surgical decisionmaking. The focus of the present study was to compare mRNA expression of known wound healing-associated genes in full-thickness sigmoid tissue from youthful patients undergoing surgical resection for diverticulitis (where a genetic defect would be most likely) and control sigmoid tissue from patients without DD, to further understand the likely underlying connective tissue aspect of the development of diverticulosis. Control patients were ones without evidence of any DD (diverticulosis or diverticulitis). Since such disease usually presents at an elderly age, to ensure the greatest likelihood of a group of patients without a genetic predisposition to DD, older patients were chosen as controls, since youthful or agematched controls could very possibly contain patients who later would develop DD.
Materials and methods A Qiagen™ Human Wound Healing RT2 Profiler™ Array (SABiosciences, Valencia, CA catalog number PAHS-121Z) was used according to the manufacturer’s recommended
Int J Colorectal Dis (2015) 30:1247–1254
protocol. The array measures the mRNA from 84 genes associated with extracellular matrix/cellular adhesion, growth factors, inflammatory cytokines, and signal transduction (Table 1) [23]. For the Wound Healing Array analysis, full-thickness sigmoid surgical specimens from 16 unrelated youthful SD patients (mean age 37.6±0.9 years, 11 plegmaceous or recurrent disease, 5 perforating or abscessing disease) and 15 control patients undergoing a sigmoid resection (mean age 52.9± 10.5 years, comprised of 2 dysmotility, 1 volvulus, 1 endometriosis, 3 familial adenomatous polyposis (FAP), 1 hereditary nonpolyposis colorectal cancer, and 7 colorectal cancer (CRC)) were selected from the Hershey Medical Center Division of Colon and Rectal Surgery’s Biobank. The absence of diverticular disease in all control patients was confirmed on preoperative imaging (CT scan or barium enema) by a specialist gastrointestinal radiologist and again on operative specimen pathology, at the time of surgery. The mRNA identified as potentially most significantly altered on the Wound Healing Array (vitronectin (VTN), see BResults^) was further specifically studied using a TaqMan gene expression assay (Applied Biosystems, Hs00940758_g1) using 100 ng of cDNA. For the TaqMan assay, an additional 11 youthful SD patients (mean age 39.5 ±6.1 years, 5 male) and four new control (mean age 64.4±7.9, 4 male, 1 unresectable polyp, 1 dysmotility, and 2 rectal cancers) patients were recruited. These samples were added to the original 31 samples (16 SD, 15 controls) used in the first Wound Healing Array experiment to yield a total of 27 SD and 19 control tissues for the subsequent VTN TaqMan assay study. Patient demographics are shown in Table 2. Patient comparison groups are show in Table 3. For both the Wound Healing Array and the VTN TaqMan assay, full-thickness sigmoid tissue was harvested at the time of surgical resection with the assistance of the operating surgeon. In the cases of cancer and/or polyps, normal-appearing tissue adjacent to the lesion was taken. The specimens were placed immediately in RNAlater (Ambion, Cat No. AM7021) and flash frozen in liquid nitrogen. Subsequent tissue pulverization was performed using a Biospecs MultiSample BioPulverizer (Cat. No. 59012MS). A TRIzol (Ambion, Cat. No. 15596018)/RNeasy Mini Kit (Qiagen, Cat. No. 74104) hybrid RNA extraction protocol was used to isolate RNA from the pulverized samples. Total isolated RNA quality was assured using an Agilent Bioanalyzer 2100. All samples included in the study had an RNA integrity number (RIN) value ≥7. The SuperScript III First-Strand kit (Invitrogen, Cat. No. 18080-051) using 400 ng of total RNA was used for cDNA conversion according to the manufacturer’s recommended protocol. All plates were read on an Applied Biosystems 7900HT Real-Time PCR System. RQManager (Applied Biosystems) software was
Int J Colorectal Dis (2015) 30:1247–1254 Table 1
Qiagen™ Human Wound Healing RT2 Profiler™ Array pathways and genes
Extracellular Remodeling enzymes matrix components COL14A1 COL1A1 COL1A2 COL3A1 COL4A1 COL4A3 COL5A1 COL5A2 COL5A3 VTN
1249
CTSG CTSK CTSV F13A1 F3 (tissue factor) FGA (fibrinogen) MMP1 MMP2 MMP7 MMP9 PLAT (tPA) PLAU (uPA) PLAUR (uPAR) PLG SERPINE1 (PAI-1) TIMP1
Cellular adhesion
Cytoskeleton
Inflammatory cytokines Growth factors and chemokines
CDH1 (E-cadherin) ITGA1 ITGA2 ITGA3 ITGA4 ITGA5 ITGA6 ITGAV ITGB1 ITGB3 ITGB5 ITGB6
ACTA2 (a-SMA) ACTC1 RAC1 RHOA TAGLN
CCL2 (MCP-1) CCL7 (MCP-3) CD40LG (TNFSF5) CXCL1 CXCL11 (ITACIP-9) CXCL2 CXCL5 (ENA-78/LIX) IFNG IL10 IL1B IL2 IL4 IL6
used for TaqMan result interpretation and RT2 Profiler™ PCR array data analysis software (Qiagen, Valencia, CA).
ANGPT1 CSF2 (GM-CSF) CSF3 (GCSF) CTGF EGF FGF10 FGF2 FGF7 HBEGF (DTR) HGF IGF1 MIF PDGFA TGFA TGFB1 TNF VEGFA
Signal transduction
TGFβ: TGFB1 TGFBR3 STAT3 WNT: CTNNB1 WISP1 WNT5A Phosphorylation: MAPK1 (ERK2) MAPK3 (ERK1) PTEN Receptors: EGFR IL6ST (GP130) Other: PTGS2
Results Wound Healing Array
Statistical analysis Multiple group comparisons were made in both the Wound Healing Array (n=16 SD vs. 15 controls) and the subsequent focused VTN TaqMan gene expression assay (n=27 SD vs. 19 controls) cohorts (see Table 3): The Student’s t test was used for analysis of the Wound Healing Array using RT2 Profiler™ PCR array data analysis software (Qiagen, Valencia, CA). The Mann-Whitney test was used for subsequent statistical analysis of VTN using R software. A Bonferroni correction was applied to account for multiple comparisons of 84 genes in the Wound Healing Array.
None of the 84 genes on the array were upregulated greater than 1.5 times in the 16 SD patients’ tissue when compared to all control tissues (n= 15). However, several genes were downregulated in the SD tissue vs. all control tissue (Fig. 1). The most marked downregulation was seen in colonystimulating factor 3 (CSF3) and prostaglandin-endoperoxide synthase 2 (PTGS2). However, this dysregulation was not statistically significant. Only chemokine (CXC motif) ligand 2 (CXCL2) and CSF3 were found to be significantly downregulated on raw analysis. However, both comparisons lost significance after Bonferroni correction. The most potentially significant change found in any of the comparisons performed was seen in the SD vs. the four non-
Table 2 Patient demographics Diverticulitis array cohort n=16
Additional diverticulitis patients for gene candidate (VTN) study
Array controls n=15
n=11
Additional controls for gene candidate (VTN) study n=4
Age at diagnosis Years, SD Range (years) Male Smoking status Current Former Never Unknown
37.6±4.9 29.8–48.3 15 4 5 6 1
39.5±6.1 27.4–48.6 5 3 4 3 1
52.9±10.5 40.3–73.6 7 3 5 6 1
64.4±7.9 50.3–70.4 4 0 2 2 0
1250 Table 3 Patient subset comparisons
Int J Colorectal Dis (2015) 30:1247–1254
Cohort comparisons
Primary analysis using 84-gene array
Targeted gene (VTN) study
Number of patients
Number of patients
SD vs. all non diverticulitis controls* SD vs. cancer only
16 vs. 15 16 vs. 7
27 vs. 19 27 vs. 9
SD vs. non-neoplastic controls** Perforating vs. non-perforating SD
16 vs. 4 11 vs. 5
27 vs. 8 19 vs. 8
*Non-diverticular controls: cancer+HNPCC+FAP+dysmotility+volvulus+endometriosis **Non-neoplastic controls: dysmotility+volvulus+endometriosis
neoplastic control (constipation, dysmotility, endometriosis) tissue comparisons (Fig. 2). The VTN gene was found to be downregulated 2.7× relative to the non-neoplastic tissue (p= 0.001, prior to Bonferroni correction, p=0.08 after correction). Although several other genes were downregulated at least 2 fold in the SD tissue, no significant difference was demonstrated after Bonferroni correction. In the SD vs. cancer tissue (n=7) comparison, CSF3 expression level was significantly downregulated (3.4-fold) in the SD tissue (p=0.01) but significance was lost after Bonferroni correction (Fig. 3). The greatest difference in expression was seen when comparing the perforating SD (n=5) tissue vs. the non-perforating SD tissue (n=11) (Fig. 4). Matrix metalloproteinase 7 (MMP7) was downregulated 3.9× in the perforating tissue (p=0.037 before correction, NSD after correction).
upregulated in the non-perforating tissue when compared to controls. However, none of these comparisons reached statistical significance.
Discussion
With the addition of 15 patients as described above (11 SD and 4 controls), candidate gene expression study for the VTN gene using a TaqMan array demonstrated no difference in mRNA expression in SD vs. controls. In fact, expression was nearly identical in the two groups (Table 4). On subgroup analysis, VTN expression was downregulated in tissue with perforating diverticular disease when compared to both control and non-perforating diverticular tissue. Conversely, it was
This study utilized an array to study the mRNA expression of 84 genes found in pathways that may intuitively be associated with DD. We chose to study patients with youthful disease onset. Our control population had to be older patients to be sure that they would not have a genetic predisposition to diverticuli which typically occur later in life and may simply be a phenotypic variant of the same genetic defect. It would not be appropriate to compare a group of 30-year-olds with diverticulitis with a group of 30-year-olds without diverticulitis since approximately half of such Bcontrols^ would eventually develop DD by the age of 60 and thus could possibly contain the same underlying genetic defect as in the studied group. The study design used such an array to cast a Bwide net^ for the identification of potential gene targets for further study. Similarly, several comparisons between subset groups were evaluated. No obvious differences were seen, except for in the case of VTN, which suggested itself as a potential gene candidate. However, further evaluation with additional patients similarly showed no statistical difference. First, the gene
Fig. 1 Genes with downregulated mRNA expression on Qiagen™ Human Wound Healing RT 2 Profiler ™ Array in all surgical diverticulitis (n=16) vs. all control (n=15) tissues. CXCL2 chemokine (C-X-C motif) ligand 2, FGF10 fibroblast growth factor 10, WISP
WNT1-inducible-signaling pathway protein 1, IL6 interleukin 6, ACTA2 alpha-actin-2, ACTC1 alpha cardiac muscle 1, TAGLN transgelin, CSF3 colony-stimulating factor 3 (granulocyte), PTGS2 prostaglandinendoperoxide synthase 2, VTN vitronectin
VTN study
Int J Colorectal Dis (2015) 30:1247–1254
1251
Fig. 2 Genes with differential regulation of mRNA expression on Qiagen™ Human Wound Healing RT2 Profiler™ Array in all surgical diverticulitis patient (n = 16) vs. healthy control (n = 4) tissues (2 dysmotility, 1 volvulus, 1 endometriosis). FGF2 fibroblast growth factor 2, ANGPT1 angiopoietin 1, FGF7 fibroblast growth factor 7,
ITGB3 integrin beta-3, EGF epidermal growth factor, ACTC1 alpha cardiac muscle 1, ITGA integrin alpha 1, WISP WNT1-induciblesignaling pathway protein, ACTA2 alpha-actin-2, VTN vitronectin, PTGS2 prostaglandin-endoperoxide synthase 2, TAGLN transgelin
expression of all non-diverticular disease controls was compared to the diverticulitis patients. As this was a heterogeneous population, we also examined cancer patients separately. A potential overlap between genes expressed in the cancer patients and diverticulitis patients led us to exclude these patients in the additional analysis and study the expression from the four noncancer patients vs. diverticulitis patients. Lastly, we hypothesized that perforating and non-perforating DD
may have a differing genetic basis. Thus, we compared gene expression between these multiple groups; several genes on this array were found to have differential expression in the diverticulitis patients vs. control tissue on raw statistical analysis. Although statistical significance was lost in all comparisons after Bonferroni correction, changes associated with the VTN gene appeared to have the most significant raw p value. Thus, this gene was selected for further specific study.
Fig. 3 Genes with downregulated mRNA expression on Qiagen™ Human Wound Healing RT 2 Profiler ™ Array in all surgical diverticulitis (n=16) vs. cancer (n=7) tissues. FGF10 fibroblast growth factor 10, ITGB3 integrin beta-3, COL5A type V collagen, IL6 interleukin
6, ACTA2 alpha-actin-2, ACTC1 alpha cardiac muscle 1, TAGLN transgelin, PTGS2 prostaglandin-endoperoxide synthase 2, CSF3 colony-stimulating factor 3 (granulocyte)
1252
Int J Colorectal Dis (2015) 30:1247–1254
Fig. 4 Genes with differential mRNA expression on Qiagen™ Human Wound Healing RT2 Profiler ™ Array in all perforating surgical diverticulitis (n=5) vs. non-perforating surgical diverticulitis (n=11) tissue. MMP7 matrix metalloproteinase 7, CSF3 colony-stimulating
factor 3 (granulocyte), MMP1 matrix metalloproteinase 1, CTSV cathepsin V, SERPINE1 serpine pepsidase inhibitor/plasminogen activator type 1, F3 coagulation factor III, VTN vitronectin, PTGS2 prostaglandinendoperoxide synthase 2, PLG plasminogen
When designing the second half of the experiment looking specifically at VTN using the TaqMan array, we chose to include the tissue that was already studied on the wound healing array but added newly recruited patients for additional power. Although not a traditional discovery+confirmatory cohort design, this was done due to relatively low overall patient numbers, and with this larger cohort, statistical significance was not achieved. The majority of previous functional studies on diverticulitis have looked at the collagen, elastin, and proteoglycancontaining extracellular matrix (ECM) which provides the integrity and flexibility of the colonic wall [11, 24]. VTN is a glycoprotein found in the ECM and has a role in cell adhesion as well as the prevention of lysis by the complement pathway and hemostasis [25]. It may also assist in promoting inflammation as a host defense against infection [26]. The matrix metalloproteinase (MMP) family of zinc-containing endopeptidases degrades the ECM, while tissue inhibitors of MMPs (TIMPs) inhibit them and thus prevent degradation. The MMPs and TIMPs are the most studied physiologic components of the colon wall in diverticulitis patients. Using rtPCR, Mimura et al. demonstrated no difference in MMP1, 2, and 9 levels between tissues from the diverticulum of six cancer specimens with asymptomatic diverticular disease, tissue adjacent to an abscess or stricture in 5 Bcomplicated
diverticulosis,^ and 11 colorectal cancer control tissues, similar to our study. However, TIMP1 and 2 mRNA transcripts were higher in tissue affected by diverticular disease vs. control tissue with a significant difference seen in the muscularis propria [12]. Lakatos et al. specifically examined MMP9 in tissue from patients with another inflammatory colonic pathology, ulcerative colitis (UC), and found expression levels were more than 4× higher in UC patients than in DD patients [27]. Similarly, Altadill et al. showed a downregulation of MMP2, MMP9, and MMP13 expression in inflamed diverticulitis vs. Crohn’s disease (CD) mucosa in over 50 patients using rtPCR, immunohistochemistry, and western blots. However, higher expression of MMP1, TIMP1, and TIMP3 was seen in both inflamed and noninflamed diverticulitis mucosa vs. CD [9]. In the present study, the MMP genes demonstrated significant downregulation on raw analysis in the penetrating vs. nonpenetrating comparisons. However, in all other comparisons, fold change was less than 1.3 (SD vs. all controls, SD vs. healthy controls, SD vs. cancer tissue). The one TIMP included on our array, TIMP1, did not show significant differential regulation on raw analysis on any of our comparisons, conflicting with these other studies [9, 12]. Signaling and inflammatory pathways in diverticulitis have been less studied. One study evaluated histamine receptor expression in over 100 diverticulitis patients (57 complicated, 44 non complicated) and sigmoid cancer controls. A significant association with complicated diverticulitis and Ban allergic predisposition^ (pollen, food, pets, etc.) was noted, and a difference in H1R and H2R expression between healthy and diverticulitis patient tissues was demonstrated [28]. Costedio et al. performed immunohistochemistry, rtPCR, and enzyme immunoassay kit for 5HT in stimulated and unstimulated tissues to study the different genes in the serotonin signaling pathway in sigmoid tissue from 22 healthy controls, 16
Table 4
VTN TaqMan mRNA expression comparisons
Comparison
Fold change
p value
All SD vs. controls Perforating SD vs. controls Non-perforating SD vs. controls Perforating vs. non-perforating
1.03 −0.52 1.4 −3.1
0.50 0.23 0.94 0.12
Int J Colorectal Dis (2015) 30:1247–1254
diverticulosis patients, and 13 diverticulitis patients with matched normal adjacent tissue and found decreased serotonin-selective reuptake transporter SERT (5HT transporter) in the sigmoid mucosa of patients with a history of diverticulitis when compared with controls [29]. How this relates to pathophysiology of diverticulitis is unclear and, like other observations, may be Bepiphenomenon.^ Limitations The major limitation to the present study is the low number of healthy control tissues. In order to study the colonic wall in its full thickness as was chosen for this project, full surgical resection specimens are required not just colonoscopic biopsies since a defect in the muscular wall of the colon (not the mucosa) lead to eventual diverticuli. Full-thickness colonic samples from healthy controls are difficult to obtain, especially from youthful patients. Future studies utilizing colonoscopy biopsy specimens may be warranted to increase patient numbers but may not capture dysregulated gene expression in the entire colonic wall. Another limitation is the slightly varied distance from our specimens to the abscess/perforation or even inflamed diverticuli as expression may vary in these different locations. However, all specimens were taken as close to the foci of disease as possible to minimize this effect. Another criticism is the heterogeneity of our control population. However, the main criteria for inclusion into this category was the absence of DD and an older age at resection was necessary, to ensure that diverticulosis would not develop later in life, as may occur and confound the results, if a younger control population was chosen. Results could be confounded due to the inclusion of the five patients with perforating disease. Perforating disease is an acute infective process with macrophage infiltration and acute inflammation which may alter results. This was noted in de Vries’ study where dectin, a pattern recognition receptor expressed on myeloid cells, showed increased expression in tissue with severe vs. mild diverticulitis. The authors noted that this increased expression may likely be due to macrophage expression, on which dectin is present [30]. To address this, in the present study, the expression between the perforating and non-perforating diverticulitis tissues was compared. No difference was seen between mRNA expressions in nonperforating vs. perforating DD in any of the 84 genes after Bonferroni correction. However, the number of samples with perforating disease overall was low.
1253
This suggests that extracellular matrix and growth factorassociated genes probably play a minor role in the possible genetic defects predisposing youthful individuals to diverticulitis. Acknowledgments Division of Colon and Rectal Surgery is the recipient of the Carlino Fund for IBD Research. Conflict of interest The authors declare that they have no competing interests.
References 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
Conclusion The current project demonstrated a lack of significant mRNA dysregulation of over 80 genes traditionally associated with wound healing and tissue integrity in young patients with SD.
13.
Strate LL, Liu YL, Aldoori WH, Syngal S, Giovannucci EL (2009) Obesity increases the risks of diverticulitis and diverticular bleeding. Gastroenterology 136:115–122. doi:10.1053/j.gastro.2008.09. 025 Zaidi E, Daly B (2006) CT and clinical features of acute diverticulitis in an urban U.S. population: rising frequency in young, obese adults. AJR Am J Roentgenol 187:689–694 Rosemar A, Angeras U, Rosengren A (2008) Body mass index and diverticular disease: a 28-year follow-up study in men. Dis Colon Rectum 51:450–455 Manousos O, Day NE, Tzonou A, Papadimitriou C, Kapetanakis A, Polychronopoulou-Trichopoulou A, Trichopoulos D (1985) Diet and other factors in the aetiology of diverticulosis: an epidemiological study in Greece. Gut 26:544–549 Aldoori WH, Giovannucci EL, Rimm EB, Wing AL, Trichopoulos DV, Willett WC (1994) A prospective study of diet and the risk of symptomatic diverticular disease in men. Am J Clin Nutr 60:757– 764 Aldoori WH, Giovannucci EL, Rimm EB, Ascherio A, Stampfer MJ, Colditz GA, Wing AL, Trichopoulos DV, Willett WC (1995) Prospective study of physical activity and the risk of symptomatic diverticular disease in men. Gut 36:276–282 Crowe FL, Appleby PN, Allen NE, Key TJ (2011) Diet and risk of diverticular disease in Oxford cohort of European Prospective Investigation into Cancer and Nutrition (EPIC): prospective study of British vegetarians and non-vegetarians. BMJ 343:d4131. doi: 10.1136/bmj.d4131 Ulmer TF, Rosch R, Mossdorf A, Alizai H, Binnebosel M, Neumann U (2014) Colonic wall changes in patients with diverticular disease—is there a predisposition for a complicated course? Int J Surg. doi:10.1016/j.ijsu.2014.03.014 Altadill A, Eiro N, Gonzalez LO, Junquera S, González-Quintana JM, Sánchez MR, Andicoechea A, Saro C, Rodrigo L, Vizoso FJ (2012) Comparative analysis of the expression of metalloproteases and their inhibitors in resected Crohn’s disease and complicated diverticular disease. Inflamm Bowel Dis 18:120–130. doi:10. 1002/ibd.21682 Wess L, Eastwood MA, Wess TJ, Busuttil A, Miller A (1995) Cross linking of collagen is increased in colonic diverticulosis. Gut 37: 91–94 Stumpf M, Krones CJ, Klinge U, Rosch R, Junge K, Schumpelick V (2006) Collagen in colon disease. Hernia 10:498–501 Mimura T, Bateman AC, Lee RL, Johnson PA, McDonald PJ, Talbot IC, Kamm MA, MacDonald TT, Pender SL (2004) Upregulation of collagen and tissue inhibitors of matrix metalloproteinase in colonic diverticular disease. Dis Colon Rectum 47:371– 378, discussion 378-379 Stumpf M, Cao W, Klinge U, Klosterhalfen B, Kasperk R, Schumpelick V (2001) Increased distribution of collagen type III
1254 and reduced expression of matrix metalloproteinase 1 in patients with diverticular disease. Int J Color Dis 16:271–275 14. Pogacnik J, Messaris E, Deiling SM, Connelly TM, Berg AS, Stewart DB, McKenna KJ, Poritz LS, Koltun WA (2014) Increased risk of incisional hernia after sigmoid colectomy for diverticulitis compared with colon cancer. J Am Coll Surg 218:920– 928. doi:10.1016/j.jamcollsurg.2014.01.050 15. Claassen AT, Mourad-Baars PE, Mearin ML, Hilhorst-Hofstee Y, Gerritsen van der Hoop A (2006) Two siblings below the age of 20 years with diverticular disease. Int J Color Dis 21:190–191 16. Schlotthauer HL (1946) Familial diverticulosis of the colon: report of seven cases in one family of nine persons. Ann Surg 124:497– 502 17. Omojola MF, Mangete E (1988) Diverticula of the colon in three Nigerian siblings. Trop Geogr Med 40:54–57 18. Santin BJ, Prasad V, Caniano DA (2009) Colonic diverticulitis in adolescents: an index case and associated syndromes. Pediatr Surg Int 25:901–905. doi:10.1007/s00383-009-2472-1 19. Granlund J, Svensson T, Olen O, Hjern F, Pedersen NL, Magnusson PK, Schmidt PT (2012) The genetic influence on diverticular disease—a twin study. Aliment Pharmacol Ther. doi:10.1111/j.13652036.2012.05069.x 20. Manwaring M (2011) Diverticular disease: genetic, geographic and environmental aspects. Semin Colon Rectal Surg 22:148–153 21. Scheff RT, Zuckerman G, Harter H, Delmez J, Koehler R (1980) Diverticular disease in patients with chronic renal failure due to polycystic kidney disease. Ann Intern Med 92:202–204 22. Connelly TM, Berg AS, Hegarty JP, Deiling S, Brinton D, Poritz LS, Koltun WA (2014) The TNFSF15 gene single nucleotide polymorphism Rs7848647 is associated with surgical diverticulitis. Ann Surg 259:1132–1137. doi:10.1097/SLA.0000000000000232
Int J Colorectal Dis (2015) 30:1247–1254 23.
Qiagen. Human Wound Healing Array (2014) http://www. sabiosciences.com/rt_pcr_product/HTML/PAHS-121Z.html. Accessed April 24, 2014 24. Melendez-Zajgla J, Del Pozo L, Ceballos G, Maldonado V (2008) Tissue inhibitor of metalloproteinases-4. The road less traveled. Mol Cancer 7:85. doi:10.1186/1476-4598-7-85 25. Preissner KT, Seiffert D (1998) Role of vitronectin and its receptors in haemostasis and vascular remodeling. Thromb Res 89:1–21 26. Singh B, Su YC, Riesbeck K (2010) Vitronectin in bacterial pathogenesis: a host protein used in complement escape and cellular invasion. Mol Microbiol 78:545–560. doi:10.1111/j.1365-2958. 2010.07373.x 27. Lakatos G, Sipos F, Miheller P, Hritz I, Varga MZ, Juhász M, Molnár B, Tulassay Z, Herszényi L (2012) The behavior of matrix metalloproteinase-9 in lymphocytic colitis, collagenous colitis and ulcerative colitis. Pathol Oncol Res 18:85–91. doi:10.1007/s12253011-9420-9 28. von Rahden BH, Jurowich C, Kircher S, Lazariotou M, Jung M, Germer CT, Grimm M (2012) Allergic predisposition, histamine and histamine receptor expression (H1R, H2R) are associated with complicated courses of sigmoid diverticulitis. J Gastrointest Surg 16:173–182. doi:10.1007/s11605-011-1702-8, discussion 182 29. Costedio MM, Coates MD, Danielson AB, Buttolph TR 3rd, Blaszyk HJ, Mawe GM, Hyman NH (2008) Serotonin signaling in diverticular disease. J Gastrointest Surg 12:1439–1445. doi:10. 1007/s11605-008-0536-5 30. de Vries HS, Plantinga TS, van Krieken JH, Stienstra R, van Bodegraven AA, Festen EA, Weersma RK, Crusius JB, Linskens RK, Joosten LA, Netea MG, de Jong DJ (2009) Genetic association analysis of the functional c.714T>G polymorphism and mucosal expression of dectin-1 in inflammatory bowel disease. PLoS One 4, e7818. doi:10.1371/journal.pone.0007818
