We and others previously observed that a subset of intestinal polyps in the Apc^Min/+ mouse model still arise, and are larger, in the context of a hypomorphic Egfr allele (Egfr^wa2) or pharmacological EGFR inhibition [8, 9], suggesting the existence of an EGFR-independent mechanism by which polyps can arise. To directly test whether polyps can arise independently of EGFR, we used a conditional knockout allele of Egfr (Egfr^tm1Dwt also called Egfr^f) [19] in combination with Cre recombinase driven by the villin promoter that is expressed throughout the intestinal epithelia (Tg(Vil1-Cre)). At 100 days of age, mice with tissue-specific deletion of Egfr (Apc^Min/+, Egfr^f/f, Tg(Vil1-Cre)) displayed an 80% reduction in total polyp number compared to littermates with normal levels of EGFR (Apc^Min/+, Egfr^f/f; p < 0.0001) (Fig. 1a), similar to what was originally observed with genetic or pharmacologic reduction of EGFR activity [8]. No significant difference in the number of colon polyps was observed in the absence of Egfr (Supplementary Fig. 1), although few colonic polyps arose in the Apc^Min/+ model. Also, similar to the original study with reduced EGFR activity [8], the polyps that developed in the absence of EGFR were significantly larger (p < 0.05) than those developing under normal levels of EGFR (Fig. 1b).

Considering the limitations of the Apc^Min/+ mouse model to recapitulate human colon tumors, we also used a mouse model of inducible CRC to test whether colon polyps can also arise independently of EGFR. Colon polyps developing under normal levels of EGFR (Egfr^+/+, Apc^f/f, Kras^LSL/+) were considered EGFR-dependent, while polyps arising in the absence of EGFR were considered EGFR-independent (Egfr^f/f, Apc^f/f, Kras^LSL/+). We also analyzed mice with a wildtype Kras allele carrying Apc^f/f (Egfr^+/+, Apc^f/f, Kras^+/+) compared to animals with a combination of Apc^f/f and Egfr^f/f (Egfr^f/f, Apc^f/f, Kras^+/+). Following focal delivery of a CRE recombinase-expressing adenovirus (AdCre) to the distal colon, colonic polyps arose in the absence of EGFR irrespective of the presence or absence of oncogenic KRAS (Fig. 1c). No polyps developed by inactivating Apc alone in the presence of EGFR. Interestingly, mice lacking Egfr showed increased polyp penetrance compared to those with Egfr, irrespective of Kras status. Additionally, despite the focal induction of polyps with AdCre, polyp multiplicity was higher for EGFR-independent polyps than EGFR-dependent polyps (p < 0.001) (Fig. 1d).

We assessed the rate of polyp growth by serial colonoscopy every two weeks by comparing the ratio of the polyp and lumen cross-sectional areas (Fig. 1e). The EGFR-independent group showed a significantly higher luminal occlusion until week 15 (p < 0.05) (Fig. 1f), reaching occlusion much faster than the EGFR-dependent group. Taken together, these results definitely establish the existence of a subset of colon polyps that arise independent of EGFR and that in the absence of EGFR, polyps are more likely to arise and have an accelerated growth rate irrespective of Kras status.

To determine how EGFR-independent polyps can arise in the presence of EGFR, we analyzed RNA from intestinal polyps from Apc^Min/+, Egfr^f/f and Apc^Min/+, Egfr^f/f, Tg(Vil1-Cre) mice. Hierarchical clustering (Supplementary Fig. 2a) and PCA analysis (Supplementary Fig. 2b) from the transcriptome data demonstrated a clear distinction between Apc^Min/+, Egfr^f/f and Apc^Min/+, Egfr^f/f, Tg(Vil1-Cre) intestinal polyps. A subset of intestinal polyps from Apc^Min/+, Egfr^f/f mice, with normal levels of EGFR, clustered with Apc^Min/+, Egfr^f/f, Tg(Vil1-Cre) polyps. The presence of Apc^Min/+, Egfr^f/f that clustered with Apc^Min/+, Egfr^f/f, Tg(Vil1-Cre) polyps confirmed the existence of an EGFR-independent mechanism of CRC progression even in mice with normal EGFR present. Together, these results suggest that the EGFR-independent pathway is not dependent upon loss of EGFR, but occurs in a subset of polyps upon initiation.

To evaluate whether this unique type of CRC is potentially relevant in humans, we tested human genes orthologous to those in the 1,290 gene mouse EGFR-independent molecular signature using data in the Cancer Genome Atlas (TCGA) database. We found that human CRCs partition into three classes, two of which correspond to EGFR-dependent and independent polyps based on directionality of the gene expression levels (Supplementary Fig. 3a). Similar to the accelerated growth of EGFR-independent polyps in the Apc^Min/+ model, the majority of human are CRCs predicted to be EGFR-independent based on lower survival rates (Supplemental Fig. 3b), consistent with the observation that over 80% of human 3D CRC organoids do not respond to EGFR inhibition [20].

Comparative transcriptomics also revealed differential gene expression related to Egfr and Kras status in the inducible model of CRC (Supplementary Fig. 4a). When compared to their respective adjacent normal tissue, transcriptomic analysis revealed a group of genes that characterize the three polyps types (Supplementary Fig. 4b). Forty-four genes were found to be differentially expressed between EGFR-dependent (Egfr^+/+, Apc^f/f, Kras^LSL/+) and EGFR-independent (Egfr^f/f, Apc^f/f, Kras^LSL/+) colonic polyps (Supplementary Table 1). Up-regulated genes included Aadac, Sult1a1, Itpka Sult1c2, Bmp3 and Aldh1a, while downregulated genes included Ngf, Rnaset2b, Col5a3, and Erdr1. Ingenuity pathway analyis (IPA) suggested that there are several upstream regulators that determine the EGFR dependency of colon polyps (Supplementary Tables 2–5). Differentially expressed genes are enriched in canonical pathways involving melatonin degradation, serotonin degradation, and dopamine degradation (Supplementary Fig. 5a). The most significant potential upstream molecules associated with the differentially expressed genes were IL10RA (z-score = 3.148, p = 1.10×10⁻¹⁰), HNF1A (z-score = 2.392, p = 1.09×10⁻⁵), glucocorticoid (z-score = 1.982, p = 1.19×10⁻³), and beta-estradiol (z-score = 1.367, p = 1.69×10⁻²) (Supplementary Table 6). IPA analysis also suggested that the top three enriched signaling networks were dermatological diseases and conditions, organ morphology, small molecule biochemistry (Network 1), lipid metabolism, small molecule biochemistry, vitamin and mineral metabolism (Network 2), and cell to cell signaling interaction, drug metabolism, post-translational modification (Network 3) (Supplementary Fig. 5b). The core nodes in the signaling Network 1 include ERK1/2, MAPK and PI3K complex. For Network 2 and Network 3, the core nodes include IL1B and beta-estradiol, respectively. Analysis of the transcriptome data from Apc^Min/+, Egfr^f/f, Tg(Vil1-Cre) and Apc^Min/+, Egfr^f/f intestinal polyps, in the absence of Kras mutations, also indicated that IL10RA is a common upstream regulator of the differentially expressed genes (Supplementary Table 1).

We validated by quantitative-PCR (qPCR) five common genes (Aadac, Sult1a1, Aldh1a1, Maob, and Sult1c2) between the top five enriched canonical pathways and the upstream regulator IL10RA (Supplementary Fig. 6). We showed that the transcript levels of Aadac, Sult1a1, Aldh1a1 are up-regulated in polyps lacking EGFR, independent of Kras status. Maob and Sult1c2 are up-regulated only in polyps lacking EGFR with mutant Kras.

Based on the transcriptomic analysis showing that IL10RA is a common upstream regulator of EGFR-independent polyps, we examined the role of IL10 signaling in EGFR-independent polyp growth by initially measuring Il10ra and Il10 transcriptomic levels in Egfr^+/+, Apc^f/f, Kras^LSL/+ and Egfr^f/f, Apc^f/f, Kras^LSL/+ colon polyps. We did not observe any statistically significant differences in transcript levels of Il10ra (Fig. 2a), however we did see an increase in the transcript levels of Il10 (Fig. 2b) and a well-known downstream target of IL10RA, Socs3 (Fig. 2c) suggesting that IL10RA is activated in EGFR-independent polyps. Consistent with elevated Il10 expression in Egfr^f/f, Apc^f/f, Kras^LSL/+ polyps, we also observed a substantial increase in IL10 levels in serum from animals with Egfr^f/f, Apc^f/f, Kras^LSL/+ colon polyps compared with serum from mice with Egfr^+/+, Apc^f/f, Kras^LSL/+ polyps (Fig. 2d). Increased levels of IL10 in serum have been associated with poor prognosis and lower survival in CRC patients [21, 22].

Because IL10 is known to regulate the differentiation of M2 macrophages [23], we evaluated whether accumulation of tumor infiltrating lymphocytes (TILs) is increased in EGFR-independent polyps. Immunohistochemical evaluation showed that Egfr^f/f, Apc^f/f, Kras^LSL/+ colon polyps exhibited abundant clusters of CD45-positive leukocytes (Supplementary Fig. 7) and total macrophages positive for F4/80 (Fig. 2e–f). When the type of macrophages present in EGFR-independent polyps was analyzed, they were found to be CD163-positive M2 macrophages (Fig. 2g) with lower presence of iNOS-positive M1 macrophages (Fig. 2h). An accumulation of CD3-positive T-cells was observed in both Egfr^+/+, Apc^f/f, Kras^LSL/+ and Egfr^f/f, Apc^f/f, Kras^LSL/+ colon polyps (Supplementary Fig. 6). Based on these results, the combination of EGFR-independence and mutant Kras appears to be more important in the formation of the TME than EGFR dependency or Kras status alone.

The immunosuppressive cytokine IL10 has been associated with poor prognosis in colon cancer [21, 22, 24]. Although macrophages are involved in anti-tumor defenses, production of IL10 by tumor cells may allow malignant cells to escape cell-mediated immune defenses. In order to determine if IL10 induces an anergic state in T-cell infiltrating EGFR-independent colon polyps, we examined the expression of several anergy-associated genes in EGFR-independent colon polyps compared to adjacent control normal colon tissue and EGFR-dependent polyps (Supplementary Fig. 8). Anergy-inducing genes were activated in T-cell infiltrating EGFR-independent colon polyps, suggesting that in the absence of EGFR, polyps escape the immune system through anergy by increasing IL10 levels.

Consistent with the activation of IL10 signaling in EGFR-independent colon polyps, we showed that the inhibition of Rous sarcoma oncogene (Src), a downstream regulator of IL10 signaling, also reduced colon tumorigenesis in vivo (Supplementary Fig. 9). We used mice carrying the Src^tm1Sor targeted mutation to reduce SRC activity in Apc^Min/+ mice. At five-months of age, small intestinal polyp number in Apc^Min/+, Src^tm1Sor/+ mice was reduced 35% compared to Apc^Min/+, Src^+/+ littermates (p = 0.0167). In addition, small intestinal polyps were approximately 20% smaller in Apc^Min/+, Src^tm1Sor/+ mice compared to polyps in their Apc^Min/+, Src^+/+ littermates (p = 0.0133). To evaluate potential efficacy of EGFR and SRC inhibitor combinatorial therapy, Apc^Min/+, Src^+/+ and Apc^Min/+, Src^tm1Sor/+ mice were treated with the small molecule EGFR inhibitor, AG1478. AG1478 reduced colonic polyp number in Apc^Min/+, Src^tm1Sor/+ mice by 70%, with a cooperative effect when comparing Apc^Min/+, Src^+/+ with Apc^Min/+, Src^tm1Sor/+ mice on AG1478 (p = 0.0444). Together these data suggest that IL10 signaling, potentially through SRC activation, promotes an increase in cell proliferation in EGFR-independent colon polyps

We showed that Il10 transcripts in colon polyps and IL10 levels in serum are increased in mice with Egfr^f/f, Apc^f/f, Kras^LSL/+ colon polyps. EGFR-independent colon polyps also showed an increase in cell proliferation (Ki67+ cells) when compared with Egfr^+/+, Apc^f/f, Kras^LSL/+ polyps (Supplementary Fig. 7). To evaluate the role of IL10 in promoting polyp growth, we performed in vitro experiments using colon polyp organoids. Polyp organoids were treated with IL10 or an IL10 neutralizing antibody to evaluate the effect of IL10 signaling on colon polyp growth.

Similar to in vivo results, we observed that single-cell suspensions from established Egfr^f/f, Apc^f/f, Kras^LSL/+ organoids grew faster and showed higher cell proliferation than Egfr^+/+, Apc^f/f, Kras^LSL/+ organoids at 24 hours after seeding (Fig. 3a). We also found increased levels of IL10 in the conditioned media of Egfr^f/f, Apc^f/f, Kras^LSL/+ organoids (Fig. 3a), suggesting that epithelial cells were the source of elevated serum IL10 in vivo. To test the effect of IL10 on cell proliferation, cells were seeded and after serum starvation for 12 hours, organoids were treated with mouse recombinant IL10 protein, which significantly increased cell proliferation of Egfr^+/+, Apc^f/f, Kras^LSL/+ organoids in a dose-dependent manner (Fig. 3b). This effect was not observed in EGFR-independent Egfr^f/f, Apc^f/f, Kras^LSL/+ organoids that presumably are already producing sufficient levels of IL10 for maximal growth. Conversely, proliferation was attenuated in Egfr^f/f, Apc^f/f, Kras^LSL/+ organoids treated with an IL10 neutralizing antibody (Fig. 3c), while EGFR-dependent Egfr^+/+, Apc^f/f, Kras^LSL/+ organoids showed increased growth and IL10 levels (Supplementary Fig. 10) in the media after treatment with EGFR inhibitor AG1478 (Fig. 3d). Furthermore, a cooperative effect of EGFR inhibitor and IL10 neutralizing antibody on organoid size was observed irrespective of EGFR status (Fig. 3e), suggesting that targeting both pathways could overcome the activation of compensatory pathways and be therapeutically beneficial in both EGFR-dependent and independent CRC. Together, these data confirm that IL10 is required for EGFR-independent cell proliferation and its presence can increase growth of EGFR-dependent polyps.

To evaluate the potential therapeutic utility of anti-IL10 neutralizing antibody treatment on CRC in vivo, mice with Egfr^+/+, Apc^f/f, Kras^LSL/+ and Egfr^f/f, Apc^f/f, Kras^LSL/+ colon polyps were administrated either anti-IL10 neutralizing antibody or vehicle (PBS). Confirming the in vitro results, anti-IL10 antibody administration significantly reduced Egfr^f/f, Apc^f/f, Kras^LSL/+ colon polyp size, with one polyp disappearing completely upon anti-IL10 treatment, compared to control mice (Fig. 4). There was no effect of anti-IL10 treatment on polyps from Egfr^+/+, Apc^f/f, Kras^LSL/+ mice, indicating a potential therapeutic avenue for EGFR-independent polyps.

All animal studies were conducted under protocols approved by the Texas A&M University Institution Animal Care and Use Committee guidelines. C57BL/6J (B6)-Apc^Min/+ [37] and B6.129S7- Src^tm1Sor [38] were obtained from The Jackson Laboratory (Bar Harbor, ME). B6;D2-Tg(Vil-cre)20Syr [39] and B6.129-Kras^tm4Tyj [40] were obtained from NCI-Frederick. Apc^tm2Rak [41] mice were obtained from the International Mouse Strain Resources. We previously generated B6.129- Egfr^tm1dwt [19]. All strains were crossed for at least 10 generations to the B6 background. Mice were housed five per cage, fed Purina Mills Lab Diet 2919 and maintained at 22° under a 12-hr light cycle. Mice were euthanized by CO₂ asphyxiation for tissue collection.

Crosses were set up in four cages using trios of two females Egfr^f/+, Apc^Min/+ and one male Egfr^f/+, Apc^Min/+. Over six months, littermates were genotyped and, 37 male and 38 female Egfr^f/f, Apc^Min/+ mice and 50 male and 50 female Egfr^+/+, Apc^Min/+ mice were euthanized at 100 days of age to evaluate the effect of EGFR in intestinal tumorigenesis. The number of animals was based on the prediction that 10–20% of polyps in animals with normal levels of Egfr would be transcriptionally similar to Egfr deficient polyps.

Crosses were set up in four cages using trios of two females Egfr^f/+, Apc^f/f, Kras^LSL/+ and one male Egfr^f/+, Apc^f/f, Kras^+/+, littermates were genotyped and 90 mice allocated to the following groups. The number of animals per group was estimated using JMP based on α-value of 0.05 and a desired power of 0.8: 5 females and 5 males Egfr^+/+, Apc^f/f, Kras^+/+; 15 females and 15 males Egfr^f/f, Apc^f/f, Kras^+/+; 10 females and 10 males Egfr^+/+, Apc^f/f, Kras^LSL/+; and 15 females and 15 males Egfr^f/f, Apc^f/f, Kras^LSL/+. Eight cohorts of 12 (3-month-old) mice were randomized in each cohort, with equal representation of each genotype used to locally induce colon polyps using the non-surgical exposure of AdCre described below. Animals that developed anal prolapse and met euthanasia criteria were excluded from transcriptomic analysis.

Src^tm1Sor/+ females were crossed with Apc^Min/+ males to generate Apc^Min/+ Src^+/+ and Apc^Min/+ Src^tm1Sor/+. Mice were randomly assigned to controls or the AG1478 EGFR inhibitor (LC Labs) treatment group. The number of mice were equally split by sex and included: 19 Apc^Min/+ Src^+/+; 24 Apc^Min/+ Src^tm1Sor/+; 19 Apc^Min/+ Src^+/+ treated with 144mg/kg AG1478; and 24 Apc^Min/+ Src^tm1Sor/+ treated with 144mg/kg AG1478.

Mice were genotyped for the Apc^Min allele as previously described [8]. Mice were genotyped for the Egfr^tm1Dmt allele as previously described [19]. Mice were genotyped using PCR with the primers described below:

Vil-Cre

Cre-S1, 5’ - GTGATGAGGTTCGCAAGAAC

Cre-AS1, 5’ - AGCATTGCTGTCACTTGGTC

Polyethylene tubing (I.D. 1.4 mm, O.D. 1.90 mm; Becton Dickinson, Sparks, MD) was cut to sizes appropriate for mice (10 cm). A 1 cm window was notched into the tubing and the end of the tubing was closed with edges being rounded to avoid perforation of the bowel. Marks corresponding to 1 cm intervals were made on the tubing. A longitudinal stripe was also applied corresponding to the orientation of the window. Mice were anesthetized using 2% isoflurane. The colon was irrigated with phosphate buffered saline (PBS). A narrow ribbon of GelFoam (Pharmacia and Upjohn, NY, NY) was inserted into the window cut into the polyethylene tube. 200 μL of 0.05% trypsin (Hyclone, Logan, VT) was injected into the tubing and inserted into the mouse colon at the desired depth and radial orientation. After 10 minutes, the slotted tube was removed and a soft flocked brush (Puritan Purflock Ultra Flocked 25–3316-U) was introduced at the same intraluminal location. The brush was then used to abrade the epithelium for up to 3 minutes. After PBS irrigation, a slotted tubing containing GelFoam was then filled with 200μL PBS containing 10⁹ PFU of Cre recombinase-expressing adenovirus (AdCre) (Ad5CMVCre and Ad5CMVEmpty, University of Iowa Gene Transfer Vector Core, IA). After 30 minutes of incubation, the tubing was removed. Owing to the anatomical limitations of the mouse, only the most distal half of the colon (~4 cm) could be inoculated in this way. Mice were given Banamine for analgesia and recovered quickly after the procedure. Induction with AdCre was performed in nine different cohorts. Each cohort had a group of mice representing each of the genotypes in the study.

Mice were anesthetized using 2% isoflurane and the colons were flushed with PBS. The Coloview System was used to monitor polyp formation and growth in the distal half of the colon every two weeks as previously described (Karl Storz, Goleta, CA) [42]. ImageJ analysis was utilized to measure the percent lumen occlusion as previously described [43].

For inhibiting IL10 in vivo, five mice per genotype (Egfr^f/f, Apc^f/f, Kras^LSL/+, and Egfr^+/+, Apc^f/f, Kras^LSL/+) with established AdCre-induced colonic polyps were monitored for four months. Intraperitoneally (IP) injection with the anti-IL10 was administered (200μg/mouse at the beginning of the treatment; 100μg/mouse and 20μg/mouse injections were IP administrated once/week with two days between injections for four weeks) and PBS as control. After treatment animas were sacrificed, and polyp measurements were taken to quantify treatment effects.

The polyp number and diameter were measured for the entire length of the small intestine and colon using a dissecting microscope and in-scope micrometer at 5x magnification. The smallest polyps that can be counted are approximately 0.3 mm in diameter. Polyp scoring was performed without knowledge of genotype by the investigator. Changes in polyp growth rate were recorded grossly as polyp size. In addition to polyp size, polyps were carefully scored based on number and location along the gastro-intestinal (GI) tract.

Intestinal tissues or colon samples were collected and fixed in 10% neutral buffered formalin. The processed tissues were embedded in paraffin and sectioned (7 μm). Every 50 μm, sections were taken and stained with H&E.

Colon polyps were deparaffinized in xylene followed by rehydration in 100%, 90%, and 70% ethanol and distilled water. The slides were then incubated in fresh hematoxylin (Merck, Darmstadt, Germany) for 5 min and washed in distilled water, followed by incubation in acidified eosin solution (Sigma, Deisenhofen, Germany) for 1 min and washing. Finally, the slides were dehydrated in 90% and 100% ethanol, air dried, and mounted. The H&E stained colon sections were assessed by a pathologist (A.P.) for differentiation of polyp tissue from non-polyp tissue. Sections from 5 different animals per genotype were analyzed to generate SEM.

Immunohistochemical procedures were performed as described [44]. Antigen-retrieval was performed by boiling for 20 min in citrate buffer, pH 6.0. Sections were treated with 0.3% hydrogen peroxide in PBS for 30 min, washed in PBS, blocked in PBS plus 3% specific serum and 0.1% Triton X- 100, and then incubated with primary antibodies and HRP-conjugated specific anti-rabbit secondary antibody (Vector Laboratories, Inc). Antigen-antibody complexes were detected with DAB peroxidase substrate kit (Vector Laboratories, Burlingame, CA) according to the manufacturer’s protocol. Suppliers for primary antibodies were Abcam, (CD163-ab182422, iNOS-ab15323, Ki67-ab15580, Tunel Assay Kit-ab206386); Santa Cruz Biotechnology, (CD3-sc-1127); Biolegend, Inc (F4/80-BM8), eBioscience, (CD45-14-0451). The total number of cells and all the positively stained cells in polyp area of the core was counted using Fiji and expressed as a percentage of the total number of cells in the polyp core. Sections from five different animals per genotype were analyzed to generate SEM.

Epithelial-only intestinal minigut cultures from freshly isolated Egfr^+/+, Apc^f/f, Kras^LSL/+ and Egfr^f/f, Apc^f/f, Kras^LSL/+ mouse colonic polyps were established essentially as described previously [45]. Crypts were isolated by calcium chelation and mechanical agitation, then embedded in Matrigel (Corning^™ Matrigel^™ GFR Membrane Matrix, ThermoFisher CB40230C) droplets with N2 and B27 supplements (ThermoFisher 17502001 and 17504001, respectively). Once established, cultures were broken up by adding 0.25% of trypsin and incubating for 10 minutes. Single cell suspensions of 2000 cells/well were allowed to grow for 24 hours to test differences in cell proliferation. Because Egfr^f/f, Apc^f/f, Kras^LSL/+ organoids grow faster and show higher cell proliferation at 24h after seeding than Egfr^+/+, Apc^f/f, Kras^LSL/+, cells were allowed to grow under serum starvation conditions for 12h to ensure similar number of cells at time of treatment. To evaluate cell proliferation of the organoids, treatment with mouse IL10 recombinant protein, eBioscience^™ (ThermoFisher 14-8101-62), IL10 neutralizing antibody (Clone JES052A5 from R&D MAB417), or EGFR inhibitor AG1478 (LC Labs) was left for an additional 72 hours. A range of four concentrations of each compound was exposed to 3D organoids: mouse recombinant IL10 (5, 10, 20 and 30 ng/ml), IL10 neutralizing antibody (0.5, 1, 2 and 3 ug/ml), EGFR inhibitor AG1478 (0.02, 0.1, 0.5, and 1 mM). On the same plate a medium control and a DMSO control were included. Cell proliferation was measured using CyQUANT^™ Cell Proliferation Assay Kit for cells in culture (ThermoFisher C7026). The fluorescent signal was quantified using a microplate reader (Cytation4 BioteK) with excitation at 485 nm and emission detection at 530 nm. Every condition was repeated in three wells per experiment and three times in independent tests. RNA from organoids was isolated using PicoPure^™ RNA Isolation Kit (ThermoFisher KIT0214).

IL10 concentration in mouse serum was measured using IL10 Mouse Instant ELISA^™, (ThermoFisher BMS614INDT). The measurement was performed according to the manufacturer’s instructions.

A total of three sequencing runs were performed to sequence 56 samples on NextSeq 500 sequencing instrument at Texas A&M Institute for Genome Sciences and Society Molecular Genomics Core using high output kit v2. A total of 1.5 billion 75 bp single-end reads were checked for adapter sequences and low-quality bases using Trimmomatic [46], resulting in approximately 1.4 billion filtered reads (96%). RNA-Seq reads were aligned to mouse assembly mm10 using HISAT2 version 2.0.5 [47] with an overall mapping rate of approximately 97%. Raw gene counts were generated with feature Counts package [48], while discarding ambiguous read mappings. Normalized read counts and gene expression tests were performed using DESeq2 [49] following recommended guidelines by the authors. Ingenuity Pathway Analysis (IPA, Qiagen) was used to analyzed differentially expressed genes between the different groups. RNAseq data was deposited in NIH: BioProject ID PRJNA635118.

Data in the Cancer Genome Atlas (TCGA) database was obtained through proper permissions. RNAseq data from 512 patients designated as TCGA-COAD (Colorectal Adenocarcinoma) was downloaded using the TCGAbiolinks R package [50]. Counts were catalogued at TCGA as generated by the HTSeq package [51]. Filtering was performed to consider patient data where each patient record was designated with “Primary_Tumor” metadata column resulting in a total raw count set for all genes in the orthologous list. The genes were used to perform differential expression analysis using DESeq2 [49].

Genes with significant changes in expression between EGFR-dependent and independent polyps, based on ANOVA analysis, were confirmed by qRT-PCR. cDNA was synthesized from total RNA from each polyp using the QuantiTect Reverse Transcription Kit (Qiagen 205314). PCR reactions were set up in 96-well plates, all samples were run in triplicate. Analysis was performed on a LightCycler 96 Thermocycler (Roche) using LightCycler 480 Sybr Green I Master reaction mix. Specific primers were designed to amplify a fragment from the genes (Supplementary Table 7).

The nonparametric Mann–Whitney U two-sided test was used to analyze polyp data. To compare the statistical difference between 2 groups, unpaired student’s t test was used. P-values less than 0.05 was considered as significantly different.