Oncogenic targets Mmp7, S100a9, Nppb and Aldh1a3 from transcriptome profiling of FAP and Pirc adenomas are downregulated in response to tumor suppression by Clotam

Abstract

Intervention strategies in familial adenomatous polyposis (FAP) patients and other high-risk colorectal cancer (CRC) populations have highlighted a critical need for endoscopy combined with safe and effective preventive agents. We performed transcriptome profiling of colorectal adenomas from FAP patients and the polyposis in rat colon (Pirc) preclinical model, and prioritized molecular targets for prevention studies in vivo. At clinically relevant doses in the Pirc model, the drug Clotam (tolfenamic acid, TA) was highly effective at suppressing tumorigenesis both in the colon and in the small intestine, when administered alone or in combination with Sulindac. Cell proliferation in the colonic crypts was reduced significantly by TA, coincident with increased cleaved caspase-3 and decreased Survivin, β-catenin, cyclin D1 and matrix metalloproteinase 7. From the list of differentially expressed genes prioritized by transcriptome profiling, Mmp7, S100a9, Nppb and Aldh1a3 were defined as key oncogene candidates downregulated in colon tumors after TA treatment. Monthly colonoscopies revealed the rapid onset of tumor suppression by TA in the Pirc model, and the temporal changes in Mmp7, S100a9, Nppb and Aldh1a3, highlighting their value as potential early biomarkers for prevention in the clinical setting. We conclude that TA, an “old drug” repurposed from migraine, offers an exciting new therapeutic avenue in FAP and other high-risk CRC patient populations.

Colorectal cancer (CRC) is one of the leading causes of cancer-related death worldwide.¹ Mortality rates vary according to geographic region, gender, age, ethnicity and diet/lifestyle factors.^1,2 Genetic influences also contribute to the etiology of CRC, for both sporadic cases and hereditary syndromes, such as familial adenomatous polyposis (FAP).^2–4 Risk-reducing surgical intervention in FAP patients often is coupled to prevention strategies using nonsteroidal anti-inflammatory drugs (NSAIDs).^5–7 However, cardiovascular risks associated with some NSAIDs and selective cyclooxygenase-2 (COX-2) inhibitors prompted the testing of other preventive agents, such as statins, difluoromethylornithine, ursodeoxycholic acid, eicosapentaenoic acid and curcumin.^5–7 None of these agents is completely effective against adenoma growth in the gastrointestinal (GI) tract, and although much discussion surrounds the use of aspirin, Sulindac continues to be viewed by many as the standard of care in FAP patients.⁵

Sulindac is highly effective in preclinical models that mimic FAP, including the Apc-driven mutant polyposis in rat colon (Pirc).^8–10 In addition to the presence of a significant colon tumor burden, the Pirc model has a lifespan suitable for both prevention and promotion studies.^8–11 Using a new endoscopic tracking methodology,¹² we evaluated agents that might substitute for Sulindac, and identified Clotam (tolfenamic acid, TA) as a promising candidate in the Pirc model. As an “old drug” repurposed from migraine,¹³ TA is considered to be an NSAID that acts on noncanonical, COX-independent targets in colon cancer cells,^14,15 including β-catenin and specificity protein (Sp) transcription factors.^16–18 The current investigation reports, for the first time, the potent tumor suppression by TA at clinically relevant doses in the Pirc model. Among the mechanistic targets evaluated were several oncogene candidates identified by transcriptome profiling of human FAP and Pirc colonic adenomas.

Material and Methods

RNA-seq analysis

RNA-seq was performed on three colorectal adenomas and three adjacent mucosa samples from patients diagnosed with FAP at the University of Texas MD Anderson Cancer Center (Houston, TX). Informed consent was obtained from all individuals, and the study was approved by the Institutional Review Board (MD Anderson protocol number PA12–0327). Pathological characterization, tissue preservation, and sample preparation for molecular analyses were as reported.³ RNA-seq analysis also was performed on three large colon adenomas (grade 5 lesions, average volume 141 ± 74.9 mm³) and three adjacent mucosa samples from Pirc males aged 6–8 months, housed in AAALAC certified facilities at the Institute of Biosciences & Technology, Texas A&M University (Houston, TX). Rats were obtained under a licensing agreement from Taconic Biosciences (Albany, NY), and were bred in-house and genotyped as reported.⁸ All studies were approved by the Institutional Animal Care & Use Committee under protocol # 2014–0217-IBT. Frozen colon tumor samples and their matched controls were thawed and prepared for molecular analysis as reported, including RNA quality verification and cDNA library generation.^19–22 Sequencing of human samples was performed on an Illumina HiSeq 2000 sequencer at the MD Anderson Sequencing Core Facility, whereas rat samples were run on an Illumina HiSeq 2500 instrument in the Texas A&M Genomics and Bioinformatics Services Core.

Bioinformatics analyses were performed using published methodologies.^23–26 For human FAP samples, RNA-seq reads were mapped to human genome assembly hg19 using STAR. Gene expression was quantified by RSEM, and differentially expressed genes (DEGs) were detected by limma, based on expected counts from the RSEM output. RNA-seq reads from the Pirc preclinical model were mapped to the rat genome assembly rn6 using STAR (version 2.4.2a). The number of reads mapped to each gene was determined by htseq-count. Reads per kilobase of transcript per million mapped reads (RPKM) values were used to quantify gene expression, and heatmaps were generated using gene RPKM values. Detection of DEGs was performed by DESeq2 and Cuffdiff. Gene ontology and pathway enrichment analysis was performed using DAVID.

Preclinical studies

In the first study, Pirc males at 4 weeks of age were housed 2–3 per cage and assigned randomly to one of the following groups: (i) untreated controls; (ii) Sulindac at 320 parts per million (ppm) in the diet; (iii) TA at 50 mg/kg body weight, three time per week by oral gavage; or (iv) TA + Sulindac at the corresponding doses used in (ii) and (iii), above. Monthly endoscopy examinations were conducted, as reported,¹² until the study was terminated at 4 months. In a follow-up study lasting 5 months, TA was administered in the diet at 287 ppm, and in addition to monthly colonoscopy examinations, each rat was injected i.p. with bromodeoxyuridine (BrdU) 1 hr prior to sacrifice, as reported.^27,28 The entire GI tract and other tissues were inspected carefully for possible toxicity or other signs of abnormality, and a record was kept on the weekly body weight and food consumption for each animal. The number, size and position of each lesion was recorded by individuals blinded to the treatment. Tumor and adjacent normal-looking tissues were removed, and one portion was fixed in 10% buffered formalin while other portions were flash-frozen in liquid nitrogen and stored at −80°C for molecular analyses.

Immunoblotting and immunohistochemistry

Adenomas and normal tissues were examined using immunoblotting methodologies reported.^29–31 Primary antibodies were from sources described previously for cleaved caspase-3, Survivin, cyclin D1, Mmp7, β-catenin, Sp1 and β-actin,^{16–21,27–33} whereas Aldh1a3 was from Abcam (Cambridge, MA) and S100a9 was from ThermoFisher Scientific (Bartlesville, OK). Commercial Nppb antibodies proved unsuitable for use in rat colon (data not shown). Immunohistochemical detection and quantification of cleaved caspase 3- and BrdU-positive cells was as described previously.^27,28 At least three separate experiments were conducted for each of the proteins analyzed.

qRT-PCR analysis

As reported in detail elsewhere,^19–21 frozen samples of tumor and adjacent tissue were thawed and the mRNA was extracted (RNAeasy kit, Qiagen, Valencia, CA), followed by reverse-transcriptase using SuperScript III (ThermoFisher Scientific, Houston, TX). Targets of interest were measured by PCR and normalized to glyceraldehyde-3-phosphate dehydrogenase (Gapdh). Typically, 40 cycles of PCR were run on a LightCycler 480 II system (Roche, Annapolis), in a 20-μl reaction volume containing cDNAs, SYBR green dye (Roche) and gene-specific primers. Each mRNA was quantified by determining the point at which the fluorescence accumulation entered the exponential phase (C_t), and the C_t ratio of the target gene to Gapdh was calculated for each sample. At least three separate experiments were performed for each sample.

Statistical analyses

Unless indicated otherwise, data were plotted as mean ± SD and compared using Student’s t test for paired comparisons, or analysis of variance for group comparisons. In the figures, significant outcomes were shown as follows: *p < 0.05, **p < 0.01, ***p < 0.001.

Results

RNA-seq analysis of FAP and Pirc adenomas defines novel molecular targets in common

Our initial approach was to compare, for the first time, RNA-seq datasets from three human FAP colorectal adenomas and three colon polyps from Pirc males aged 8–10 months. KEGG analysis revealed common pathways, such as Wnt signaling, chemokine signaling, cell adhesion molecules, extracellular matrix (ECM)-receptor signaling and drug metabolism (Fig. 1a). Among the differentially expressed genes in Pirc adenomas, 1,795 were downregulated and 1,666 were upregulated compared with their corresponding levels in adjacent normal-looking rat colonic mucosa (Fig. 1b). After applying an arbitrary cut-off of 8-fold in either direction, and accepting genes that were common to both rat and human, the search was narrowed to 71 tumor suppressor and 62 oncogene candidates in Pirc colon polyps. Of these genes, 20 were downregulated and 32 were upregulated >2-fold in human FAP patient samples (Fig. 1b). Subsequent hierarchical clustering clearly segregated normal colon from the corresponding adenomatous polyp in both FAP and Pirc (Fig. 2).

Figure 1.

Common pathways identified by RNA-seq analysis of FAP and Pirc colorectal adenomas. (a) KEGG pathway terms that were shared by human and rat (black bars) included Wnt signaling, cell adhesion molecules, chemokine signaling and ECM-receptor signaling. (b) Prioritization of molecular targets from RNA-seq. Differentially expressed genes (DEGs) in Pirc adenomas, compared with adjacent normal rat colonic mucosa, were filtered using the cutoffs shown in the figure. Twenty downregulated and 32 upregulated genes were carried forward for subsequent validation, see Supporting Information Figures S2 and S3.

Figure 2.

Heatmaps of differentially expressed genes in FAP and Pirc colorectal adenomas. Each polyp and its matched control (e.g., Polyp 1|Normal 1) clustered separately for both (A) FAP and (B) Pirc, following DAVID analysis on DEGs prioritized as described in Figure 1b. [Color figure can be viewed at wileyonlinelibrary.com]

TA suppresses tumorigenesis at clinically relevant doses in the Pirc model

Pirc males were assigned to groups as follows (Fig. 3a): (i) untreated controls; (ii) 320 ppm Sulindac in the diet, human equivalent dose 371 mg daily; (iii) TA at 50 mg/kg body weight, three time/week by oral gavage, human equivalent dose 490 mg daily; or (iv) TA + Sulindac, at the doses used in (ii) and (iii) above. Monthly endoscopies revealed tumor suppression by Sulindac, TA and TA + Sulindac at 2, 3 and 4 months of treatment (Fig. 3b). When the study was terminated at 4 months, the average number of colon tumors/rat recorded at necropsy (Fig. 3c) corresponded well with the colonoscopy data (4-month time-point, Fig. 3b). Sulindac, TA and TA + Sulindac inhibited the number of tumors significantly, both in the colon and small intestine (Figs. 3c and 3d). The total tumor burden, which takes account of the size and number of lesions at each site, also was reduced significantly by the test agents given alone or in combination (Figs. 3e and 3f).

Figure 3.

Tumor suppression in the Pirc model after treatment with TA, alone or in combination with Sulindac. (a) Pirc males at 4 weeks of age were assigned to the following groups: (i) untreated controls; (ii) Sulindac at 320 parts per million (ppm) in the diet, human equivalent dose 371 mg daily; (iii) TA at 50 mg/kg body weight, three time per week by oral gavage, human equivalent dose 490 mg daily; or (iv) TA + Sulindac, at the doses used in (ii) and (iii). Colonoscopies (arrows) were performed each month, as reported,¹² and the study was terminated at 4 months. (b) Endoscopic tracking of tumor suppression by the test agents. (c–f) Tumor outcomes in the colon and small intestine at 4 months, expressed in terms of number of polyps per animal, or total tumor burden (mm³). Each data-point designates an individual animal, with the n value as shown in panel (a). [Color figure can be viewed at wileyonlinelibrary.com]

A follow-up study administered TA in the diet at 287 ppm, or the human equivalent dose of 418 mg TA daily (Fig. 4a). In addition to monthly colonoscopies, each rat was injected with BrdU in order to examine changes in colonic cell proliferation, as reported.^27,28 Endoscopic examination revealed tumor suppression by dietary TA at 2, 3, 4 and 5 months of treatment (Fig. 4b), and this was corroborated at the 5-month necropsy, both in the colon and the small intestine (Figs. 4c–4f). No signs of GI toxicity were observed in any of the rats treated with TA, by either gavage or dietary routes of administration, and no deleterious effects were observed on animal body weights (Supporting Information Figure S1) or weekly food consumption (data not shown). Careful inspection of the gastric mucosa and other parts of the GI tract revealed no anomalies, and blood and urinary parameters were within the normal range.

Figure 4.

Tumor suppression in the Pirc model after dietary administration of TA. (a) Rats were given basal AIN93 diet (AIN controls), or AIN diet containing TA at 287 ppm. In addition to monthly colonoscopies, each rat was injected i.p. with BrdU, 1 hr before sacrifice, as reported.^27,28 (b) Endoscopic tracking of tumor suppression by dietary TA. (c–f) Tumor outcomes in the colon and small intestine at 5 months, expressed in terms of number of polyps per animal, or total tumor burden (mm³). Each data-point designates an individual animal, with the n value as shown in panel “(a).” [Color figure can be viewed at wileyonlinelibrary.com]

Oncogene candidates identified by RNA-seq were downregulated by TA in colon tumors

Differentially expressed genes from the RNA-seq studies (Fig. 2) were validated, in their entirety, via comprehensive qRT-PCR analysis (Supporting Information Figures S2 and S3, respectively). Among the complete set of tumor suppressor and oncogene candidates, TA altered Mmp7, S100a9, Nppb and Aldh1a3 significantly in Pirc colon tumors, downregulating their expression levels to those seen in the adjacent normal-looking colonic mucosa (Figs. 5a–5d). Temporal studies revealed that Mmp7 levels were dysregulated early in colon tumor development, and remained overexpressed throughout the course of the investigation (Fig. 5e, orange line). Treatment with TA resulted in consistent downregulation of Mmp7, in contrast to S100a9, where the expression levels were more variable (Fig. 5f). Initially, elevated levels of Nppb were followed by a gradual decline (Fig. 5g, orange line), whereas Aldh1a3 levels increased over time (Fig. 5h, orange line). Like Mmp7, both Nppb and Aldh1a3 were consistently downregulated by dietary TA (compare the orange and blue lines in Figs. 5g and 5h). The findings suggested that Mmp7 and Nppb might serve as early prevention biomarkers, with Aldh1a3 appearing as a slightly later target for TA downregulation.

Figure 5.

Oncogenic targets prioritized by RNA-seq analysis and the temporal changes after TA treatment. (a–d) Among the 52 DEGs prioritized from RNA-seq analysis, and validated using qRT-PCR (Supporting Information Figures S2 and S3), dietary TA significantly downregulated Mmp7, S100a9, Nppb and Aldh1a3 in Pirc adenomatous polyps. (e–h) Temporal changes for Mmp7, S100a9, Nppb and Aldh1a3 in Pirc colon polyps. The qRT-PCR assays were repeated at least three times, using glyceraldehyde-3-phosphate dehydrogenase (Gapdh) for normalization. Data = mean ± SD. [Color figure can be viewed at wileyonlinelibrary.com]

TA inhibits colonic cell proliferation and induces apoptosis coincident with reduced β-catenin and β-catenin/T-cell factor (Tcf) downstream targets

Immunoblotting was performed on colonic adenomas from TA-treated rats and from controls given AIN93 basal diet (Fig. 6a). The preventive agent increased cleaved caspase-3 and decreased Survivin, β-catenin, cyclin D1, Mmp7, S100a9 and Aldh1a3 protein levels in Pirc colon tumors. The findings for Mmp7, S100A9 and Aldh1a3 proteins corroborate the qRT-PCR data showing downregulation of the corresponding target genes by TA (Fig. 5). Survivin, β-catenin, cyclin D1 and Mmp7 immunoblot data were consistent with previous studies that used TA in colorectal and other human cancer cell lines.^{16–18,32,33} No apparent effect was seen on Sp transcription factors, such as Sp1 (Fig. 6a), which were implicated in prior reports with TA.^{16–18,32,33} Moreover, TA did not alter COX-1 or COX-2 protein expression levels in Pirc colon tumors (data not shown), consistent with its reported COX-independent actions.^14,15 However, an assessment of COX enzyme activity and prostaglandin levels will be required to fully corroborate the COX-dependent versus COX-independent actions of TA in the Pirc model.

Figure 6.

Molecular changes associated with TA treatment in the Pirc model. (a) Immunoblotting of colon tumors from rats fed TA in the diet identified increased cleaved caspase-3 and decreased Survivin, coincident with downregulation of β-catenin, cyclin D1, Mmp7, S100a9 and Aldh1a3. No suitable commercial antibody was found for Nppb (see text). Each blot is representative of at least two independent experiments. (b) The colonic crypt labeling index was determined in the distal, middle and proximal regions of the colon, as reported.^27,28 BrdU-positive cells, arrows. Data = mean ± SD, n = 5–6 animals. (c) Corresponding crypt labeling data for cleaved caspase 3.

Finally, the colonic crypt cell labeling index was determined for distal, middle and proximal regions of the rat colon, as reported.^27,28 Dietary TA treatment reduced cell proliferation in the distal and middle regions, but had no significant effect in the proximal colon (Fig. 6b). Upon close inspection of the basal, middle and luminal regions within each colonic crypt column, inhibition of cell proliferation by TA was largely restricted to the bottom half of the crypt, with little or no effect in the upper apical regions (data not shown). However, TA treatment increased the cleaved caspase 3 index, most notably in the middle and proximal regions, with positively labeled cells largely restricted to the apical cells of the colonic crypt (Fig. 6c, white arrows).

Discussion

This is the first report to describe the significant suppression by TA of both small intestine and colorectal adenomas in the Pirc preclinical model of human FAP. A reduction in the colonic crypt BrdU labeling index, coupled with loss of Survivin and increased cleaved caspase-3, suggested a net shift in the balance of apoptosis versus cell proliferation, consistent with the preventive actions of TA in vivo. TA was effective when given via oral gavage or by dietary administration, and exhibited equal or greater efficacy than Sulindac when tested under the conditions reported here.

Downregulation of Survivin in Pirc colon polyps recapitulated prior observations with TA in various human cancer cell lines and mouse tumor xenografts.^16–18,33 However, no effect was seen on Sp1, a reported molecular target of TA in some studies.^16–18,32 This might be due to the genetic background of the Pirc model,⁸ in which dysregulated Apc/β-catenin rather Sp transcription factor signaling serves as the dominant driver of tumorigenesis. Interestingly, in the Apc-mutant mouse kindred, Apc^Min/+,³⁴ short-term TA treatment also downregulated cyclin D1 independent of Sp1.³⁵ In this investigation, Pirc colon tumors had reduced expression of β-catenin and its downstream targets cyclin D1 and Mmp7,^36,37 extending prior observations on β-catenin degradation in TA-treated colon cancer cells.¹⁸ Thus, prevention by TA in the Pirc model might arise due to the preferential targeting of β-catenin and its upstream or downstream regulators.

In support of this idea, unbiased transcriptomic profiling of human FAP and Pirc colonic adenomas identified Wnt signaling as one of the top pathways affected, along with ECM-receptor and chemokine signaling, cell adhesion molecules, and drug metabolism. Validation of RNA-seq data by qRT-PCR analysis corroborated many of the most highly upregulated and downregulated genes that were common to both species. Among the list of fifty-two most differentially altered genes were four that responded with significant downregulation following TA treatment in the Pirc model, namely S100a9, Aldh1a3, Nppb and Mmp7. MMP7 is a well-known oncogenic factor in human CRC,³⁸ and is one of the most consistently overexpressed genes in adenomas from FAP patients.³⁹ As a downstream target of β-catenin/Tcf (37), Mmp7 downregulation would be in accordance with the preventive actions of TA in vivo. What, if anything, might connect β-catenin to the other oncogene candidates prioritized here, namely, Aldh1a3, Nppb and S100a9?

The S100 family includes S100A8 and S100A9 calcium-binding proteins linked to CRC progression, invasion, and metastasis.⁴⁰ Comparative proteomics analysis identified S100A9 as increased markedly in human CRC and the metastases.⁴¹ Human CRC cells treated with recombinant S100A9 or S100A8 proteins exhibited increased viability and migration due, in part, to upregulation of β-catenin and MMP7.⁴² Thus, downregulation of S100A8 and/or S100A9 would be predicted to attenuate β-catenin and MMP7 expression, consistent with the TA data reported here in Pirc colon tumors (Fig. 6a).

Among the nineteen human aldehyde dehydrogenase (ALDH) genes, ALDH1A1, ALDH1A2 and ALDH1A3 are markers of normal tissue stem cells and cancer stem cells, and are involved in differentiation and self-renewal.⁴³ In HT29 human colon cancer cells, miR-125a/b expression was found to regulate chemoresistance via upregulation of ALDH1A3 and MCL1.⁴⁴ Overexpression of ALDH1A3 was detected in chemoresistant Lovo-1 human colonic epithelial cells, and RNAi-mediated silencing of ALDH1A3 sensitized cells to focal adhesion kinase inhibition, resulting in cell cycle arrest and decreased cell viability.⁴⁵ Notably, several ALDH isoforms have been identified as direct transcriptional targets of β-catenin/Tcf.⁴⁶ Thus, further studies appear to be warranted in Pirc colon tumors on the link between β-catenin downregulation, reduced β-catenin/Tcf levels and attenuated Aldh1a3 expression following TA intake.

Little has been reported linking human CRC and NPPB, the gene coding for Natriuretic Peptide B (Nppb). Nppb, also known as brain natriuretic peptide, upregulates aquaporin 3 in HT29 cells via protein kinase-A and -G dependent pathways, linking these actions to water transport and barrier functions in the GI tract.⁴⁷ Anticancer effects of natriuretic peptides in colonic and other cancer cell lines were reported to involve changes in metabolic functions and Ras signaling.⁴⁸ More pertinent, Wnt inhibitory factor 1-mediated β-catenin downregulation was linked to loss of Nppb,⁴⁹ implicating Nppb as a β-catenin-regulated target.

In this investigation, TA and Sulindac were highly effective as individual test agents, but no firm conclusion could be drawn as to whether the combination might provide for additive or synergistic tumor suppression. This question will await follow-up studies using different doses of the respective test agents. It was apparent, however, that TA + Sulindac offered cooperative rather than antagonistic outcomes when used as a combination strategy in the Pirc model. This is pertinent, given the counteractive effects that sometimes arise in the clinical setting, as in FAP patients treated with celecoxib plus ursodeoxycholic acid.⁵⁰

Importantly, no gastric ulceration or other deleterious GI effects were observed in rats given the individual or combined test agents. Experiments with TA + Sulindac were not designed with a specific clinical application in mind, but sought to compare a “classical” COX-1/COX-2 inhibitor and a second NSAID postulated to act via other mechanisms relevant to colon cancer prevention.^14–18 More interesting might be the concomitant use of an agent that has a good safety margin, and that blocks the cascade to cancer within the adenoma cells that survive the primary blockade by NSAID treatment. Fenamates typically exhibit an improved safety margin over other NSAIDs in the clinical setting, and TA has a low reported frequency of GI toxicity and adverse events on renal variables.⁵¹ A double-blind, randomized trial of TA as a rapid release formulation in the treatment of migraine found that the drug was well tolerated, with occasional adverse events that were mild to moderate (e.g., flushing, fatigue, headache).⁵²

In this investigation, due to the relatively short duration of the preclinical study protocols, none of the adenomatous polyps was observed to progress to adenocarcinoma. A general shift from high-grade dysplasia to low-grade dysplasia was apparent in response to TA treatment (data not shown), but we did not specifically determine preneoplastic lesions, such as colonic aberrant crypt foci or multiple mucin depleted foci.^53–55 It is apparent that TA blocks some important oncogenic factors in the early stages of adenoma formation, but further studies are warranted on the four major targets from the transcriptomic profiling and how they might regulate the proliferation and apoptosis end-points examined here. This could provide additional insights into the changes (or lack thereof) in Survivin, cyclin D1, and Sp1, in the context of prior reports implicating these as key targets of TA.^{16–18,32,33} It will be of interest to ascertain whether the same gene signature involving Mmp7, S100a9, Aldh1a3 and Nppb in colon polyps applies to small intestine tumors, where TA and Sulindac also offered significant protection in the Pirc model. With the more widespread adoption of surgical procedures such as colectomy with ileorectal anastomosis, the pattern of mortality has shifted in FAP patients, with duodenal cancer now becoming the primary cause of death.⁵ Thus, alternative treatment strategies are urgently sought in high-risk cancer syndromes, such as FAP and hereditary nonpolyposis colorectal cancer, combining endoscopy with improved preventive agents.^5,56 We conclude that TA appears to offer an exciting new therapeutic avenue in such at-risk patient populations.

Supplementary Material

Additional Supporting Information may be found in the online version of this article.

What’s new?

Sulindac is viewed in some clinical settings as the standard of care for familial adenomatous polyposis (FAP) patients, but a need exists for safe and effective alternatives. Using the polyposis in rat colon (Pirc) model, here the authors observed significant tumor suppression by Clotam/tolfenamic acid. Rapid onset of tumor suppression coincided with downregulation of Mmp7, S100a9, Nppb and Aldh1a3, highlighting these oncogenic targets from FAP and Pirc transcriptome profiling as potential early biomarkers for prevention in the clinical setting. Clotam, an “old drug” repurposed from migraine, offers a new therapeutic avenue in FAP and other high-risk colorectal cancer patient populations.