Abstract
Genetic and environmental factors have been shown to work together to alter cancer risk. In this study we evaluate previously identified gene and lifestyle interactions in a candidate pathway that were associated with colon cancer risk to see if these interactions altered gene expression. We analyzed non-tumor RNA-seq data from 144 colon cancer patients who had genotype, recent cigarette smoking, diet, body mass index (BMI), and recent aspirin/non-steroidal anti-inflammatory use data. Using a false discovery rate of 0.1, we evaluated differential gene expression between high and low levels of lifestyle exposure and genotypes using DESeq2. Thirteen pathway genes and 17 SNPs within those genes were associated with altered expression of other genes in the pathway. BMI, NSAIDs use and dietary components of the oxidative balance score (OBS) also were associated with altered gene expression. SNPs previously identified as interacting with these lifestyle factors, altered expression of pathway genes. NSAIDs interacted with 10 genes (15 SNPs) within those genes to alter expression of 28 pathway genes; recent cigarette smoking interacted with seven genes (nine SNPs) to alter expression of 27 genes. BMI interacted with FLT1, KDR, SEPN1, TERT, TXNRD2, and VEGFA to alter expression of eight genes. Three genes (five SNPs) interacted with OBS to alter expression of 12 genes. These data provide support for previously identified lifestyle and gene interactions associated with colon cancer in that they altered expression of key pathway genes. The need to consider lifestyle factors in conjunction with genetic factors is illustrated.
Keywords: Gene expression, inflammation, BMI, NSAIDs, diet, colon cancer
Introduction
Interpreting the functionality between single-nucleotide polymorphisms (SNPs) based on associations with outcomes, such as cancer, is often difficult. Associations with tagSNPs can result from being in LD with other disease-causing SNPs or can be the result of chance findings. To help guard against erroneous associations various methods taking into account multiple comparisons are employed. Likewise, in silico programs are available to help predict functionality based on their involvement in splicing, transcription, translation, and post-translation [1,2], although studies have found that the prediction made by these programs do not correspond with associations observed in analytical studies [3]. We and others have utilized gene expression techniques to help determine potential functionality. However, gene expression associated with specific genotypes can help provide an indication of functionality if it exists, but lack of an association does not rule out functionality [4-6]. In our previous work, we shown that candidate genes and their associated SNPs influence expression of pathway genes to a greater extent than they do their own expression [4].
The examination of the interaction between genetic variants and lifestyle-related factors has been a hallmark of our ability to comprehend the complicated factors contributing to disease. It has long been hypothesized that genes and environment work together in creating disease risk and that consideration of one without the other leaves a void in our understanding of the carcinogenic process. Our examination of The Convergence of Hormones, Inflammation, and Energy-Related Factors (CHIEF) Pathway has illustrated both the risk associated with genetic variation [7] and that lifestyle factors further modify the risk associated with these genes [8,9]. Important lifestyle factors that interact with genes in the pathway are dietary components such as antioxidants that interact with MAPK [9,10], angiogenesis [8], and TLR [11] genes; aspirin and/or ibuprofen use with angiogenesis [8], cytokines [12,13], MAPK [9], JAK/STAT [14], TGFβ-signaling pathway [15], TLR [11], and selenoprotein genes [16]; cigarette smoking with angiogenesis [8], cytokines [13,17], MAPK [9], JAK/STAT [14], TGFB1 [15], and selenoprotein 16 pathway-related genes; and BMI with angiogenesis [8], estrogen-related genes [18], VDR [19], and cytokines [13,17].
In this paper, we build on our previous work with the CHIEF pathway, which is composed of genes associated with inflammation, angiogenesis, energy-related factors, and hormones [20]. Previously we examined associations between pathway genes and expression of select genes within the pathway [4]. In this study, we evaluate expression of all pathway-related genes with genotypes of pathway genes. We test our previously identified genes and SNPs that were associated with colon cancer using ARTP method to evaluate significance over the pathway [7]. We further evaluated the functionality of the pathway, by assessing how genotypes combined with lifestyle factors impact gene expression. We focus our assessment of lifestyle factors on variables associated with oxidative stress (i.e. cigarette smoking, recent use of aspirin and/or ibuprofen referred to as NSAIDs, body mass index (BMI), and our oxidative balance score (OBS) and its related dietary components). Based on our previous work, we hypothesize that genetic variation in the pathway influences expression of other genes in the pathway; the altered expression is influenced by the combination of genetic and lifestyle-related factors.
Methods
Total RNA was available from colonic non-tumor tissue for 175 colon cancer cases who were part of the Diet, Activity, and Lifestyle study, an incident, population-based, case-control study of colon cancer from Utah and the Kaiser Permanente Medical Research Program (KPMRP). Cases had tumor registry verification of a first primary adenocarcinoma of the colon and were diagnosed between October 1991 and September 1994. Tumor tissue blocks were obtained for 97% of all Utah cases and for 85% of all KPMRP cases [21] and included those who signed informed consent and those retrieved by local tumor registries and sent to study investigators without personal identifiers. Individuals with known adenomatous polyposis coli (APC), Crohn’s disease, or inflammatory bowel disease were not eligible for the study. The study was approved by the Institutional Review Board of the University of Utah and at KPMRP and all study participants signed informed consent.
Diet and lifestyle data
Data were collected by trained and certified interviewers using laptop computers. All interviews were audio-taped as previously described and reviewed for quality control purposes [22]. The referent period for the study was two years prior to diagnosis for cases and selection for controls. Dietary information was obtained for the referent year from an extensive diet history questionnaire adapted from the validated CARDIA diet history [23]. As part of the study questionnaire, information was collected on regular use and current use of aspirin and non-steroidal anti-inflammatory drugs and cigarette smoking history including start and stop dates for smoking. Measured height at the time of interview and self-reported weight from two years prior to diagnosis were used to calculate body mass index (kg/m2).
TagSNPs and genetic assessment
TagSNPs were selected using the following parameters: r2=0.8 defined LD blocks using a Caucasian LD map, minor allele frequency (MAF) >0.1, range =-1500 bps from the initiation codon to +1500 bps from the termination codon, and 1 SNP/LD bin. All markers were genotyped using a custom multiplexed bead array assay format based on Golden Gate chemistry (Illumina, San Diego, California). A genotyping call rate of 99.85% was attained. Blinded internal replicates represented 4.4% of the sample set. The duplicate concordance rate was 100.00%. Previous analysis using Adaptive Rank Truncation Product (ARTP) identified several genes as significant within the pathway [7]. We focus our genotype analysis with pathway gene expression on results of those 155 genes and their 1246 SNPs and the previously identified inheritance model (Supplemental Table 1 includes pathway genes evaluated along with their location and full name). In our previous analysis, genes were considered significant if the Adaptive Rank Truncation Product had a gene p value of 0.05 or less and a SNP p value of 0.05 or less. When evaluating genotype and lifestyle interactions with pathway gene expression we focused on those SNPs previously identified as interacting with lifestyle factors to modify colon cancer risk (Supplemental Table 2 contains a list of SNP and lifestyle interactions tested).
RNA processing
RNA was extracted from formalin-fixed paraffin embedded tissues. We assessed slides containing thin section of colon tissue and blocks that were prepared over the duration of the study prior to the time of RNA isolation to determine their suitability. Older slides produced comparable RNA quality as more recent slides and were not correlated with time lapse between slide preparation and RNA preparation. The study pathologist reviewed slides to delineate tumor and non-tumor tissue. Cells from non-tumor colon tissue were dissected from 1-4 sequential sections on aniline blue stained slides using an H&E slide for reference. Total RNA was extracted, isolated, and purified using the RecoverAll Total Nucleic Acid isolation kit (Ambion), RNA yields were determined using a NanoDrop spectrophotometer.
Sequencing library preparation
Library construction was performed using the Illumina TruSeq Stranded Total RNA Sample Preparation Kit with Ribo-Zero. Briefly, Ribosomal RNA was removed from 100 ng total RNA using biotinylated Ribo-Zero oligos attached to magnetic beads that are complimentary to cytoplasmic rRNA. Following purification, the rRNA-depleted sample is fragmented with divalent cations under elevated temperatures and primed with random hexamers in preparation for cDNA synthesis. First strand reverse transcription is accomplished using Superscript II Reverse Transcriptase (Invitrogen). Second strand cDNA synthesis is accomplished using DNA polymerase I and Rnase H under conditions in which dUTP is substituted for dTTP, yielding blunt-ended cDNA fragments in which the second strand contains dUTP. An A-base is added to the blunt ends as a means to prepare the cDNA fragments for adapter ligation and block concatamer formation during the ligation step. Adapters containing a T-base overhang were ligated to the A-tailed DNA fragments. Ligated fragments were PCR-amplified (13 cycles) under conditions in which the PCR reaction enables amplification of the first strand cDNA product, whereas attempted amplification of the second strand product stalls at dUTP bases and therefore is not represented in the amplified library. The PCR-amplified library was purified using Agencourt AMPure XP beads (Beckman Coulter Genomics). The concentration of the amplified library was measured with a NanoDrop spectrophotometer and an aliquot of the library is resolved on an Agilent 2200 Tape Station to define the size distribution of the sequencing library.
Sequencing and data processing
Sequencing libraries (18 pM) were chemically denatured and applied to an Illumina TruSeq v3 single read flow cell using an Illumina cBot. Hybridized molecules were clonally amplified and annealed to sequencing primers with reagents from an Illumina TruSeq SR Cluster Kit v3-cBot-HS. Following transfer of the flowcell to an Illumina HiSeq instrument, a 50 cycle single-read sequence run was performed using TruSeq SBS v3 sequencing reagents. The single-end 50-base reads from the Illumina HiSeq2500 were aligned to a sequence database containing the human genome (build GRCh37/hg19, February 2009, from genome.ucsc.edu) plus all splice junctions generated using the USeq MakeTranscriptome application (version 8.8.1, available here: http://useq.sourceforge.net/). Alignment was performed using novoalign version 2.08.01 available from novocraft.com, which also trimmed any adapter sequence. Following alignment, genome alignments to splice junctions were translated back to genomic coordinates using the US eq Sam Transcriptome Parser application. The resulting alignments were sorted and indexed using the Picard Sort Sam application (version 1.100, available here: http://broadinstitute.github.io/picard/). Aligned read counts for each gene were calculated using pysam (https://code.google.com/p/pysam/) and samtools (http://samtools.sourceforge.net/). A python script using the pysam library was given a list of the genome coordinates for each gene, and counts to the exons and UTRs of those genes were calculated. Gene coordinates were downloaded from http://genome.ucsc.edu. We focused our analysis on differential gene expression of 236 genes related to the CHIEF pathway.
Statistical methods
Of the 197 initial tumor/non-tumor tissue pairs, 22 subjects failed quality control (QC) based on low number of sequence leaving 175 subjects with high quality expression data. Of these, 144 had questionnaire data for diet and lifestyle data, and 138 had genotype data for inclusion in the analysis. For each genetic, dietary, and lifestyle factor, our analysis centered on contrasting gene expression levels of individuals with lower intake or exposure levels to those of individuals with higher intake or exposure levels. For genotype associated we focused on the previously identified inheritance model (i.e. dominant, recessive, or additive) that was associated with colon cancer to define categories for contrast. To summarize risk associated with multiple exposures, we utilized an oxidative balance score (OBS) that consisted of 13 diet and lifestyle factors that were pro-oxidants (dietary iron and polyunsaturated fat and cigarette smoking) and anti-oxidants (vitamin C, vitamin E, selenium, beta carotene, lycopene, lutein/zeaxanthin, vitamin D, calcium, and folic acid and NSAID use) [8]. This score has been shown be associated with colon cancer [8]. To create the OBS, these diet and lifestyle factors were assigned values of 2 for low levels of exposure for each pro-oxidants or high exposure to anti-oxidants (low-risk), one for intermediate levels of exposure, and zero for high levels of exposure to pro-oxidants and low exposure to anti-oxidants (high-risk). The individual scores for the 13 variables were then combined to obtain the OBS. Higher summary score corresponded to greater oxidative balance; individual’s OBSs were categorized as low, intermediate, or high based on tertiles associated with the empirical distribution of the OBSs. Additionally, we evaluated specific components of the OBS. Dietary data were evaluated using nutrients per 1000 calories and then categorized using quartiles of intake based on sex-specific distributions, combining the second and third quartiles to form the intermediate group. Cigarette smoking was categorized as never, former, or current smoker. Use of NSAIDs (which included aspirin and/or non-steroidal anti-inflammatory drugs) was categorized as either being a recent user (i.e. using NSAIDs during the referent period) or a non-user. We evaluated BMI categorized as <25 kg/m2, 25-30 kg/m2, or >30 kg/m2.
For each variable of interest (genotype, specific dietary factors, recent NSAIDS use, recent cigarette smoking, BMI, and OBS), we assessed which genes displayed statistically significant differential expression between low and high categories of gene, diet, and lifestyle factor using the Bioconductor package DESeq2 written for the R statistical programming environment. DESeq2 assumes the RNA-seq counts are distributed according to negative binomial distributions. It utilizes generalized linear modeling to test individual null hypotheses of zero log2 fold changes between high and low categories (i.e. no differential expression); for each gene it employs both an independent-filtering method and the Benjamini and Hochberg (BH) [24] procedure to improve power and control the false discovery rate (FDR). The default DESeq2 options were applied, including the replacement of outliers, as defined by Cook’s distance, and the use of the Wald test. For further details regarding DESeq2, see Love et al. [25]. In identifying genes with differential expression, an FDR of 0.10 was used. To help describe the data we report the average DESeq2-adjusted gene expression levels (size factor adjusted counts) among individuals in the high and low categories of pathway genes. The fold change calculations associated with these genes was determined by DESeq2 and represents the log2 change in expression level (i.e. counts) for the high category compared to the low category.
Evaluation of gene and environment interactions on pathway gene expression also was done in DESeq2 as described above, with the exception of using the likelihood ratio test (LRT) in place of the Wald test in order to test all levels of the interaction term. This analysis focused on previously reported statistically significant interactions between genetic variants and lifestyle factors with expression levels of 236 pathway-related genes. We report mean normalized count values along with unadjusted and BH-adjusted p values where the FDR was set at <0.1.
Results
The mean age of the study population was 64.5 (Table 1). There were more men than women, more people who never smoked than were recent smokers, and more people who did not recently use NSAIDS (aspirin and/or ibuprofen) than used NSADIs. Almost 35% had a BMI <25, 36.8% were considered overweight (BMI 25-30), and 28.4% were considered obese (BMI >30).
Table 1.
Description of study population
| Mean | STD | ||
|---|---|---|---|
| Age | 64.5 | 9.8 | |
| Oxidative balance score | |||
| Low | 9.7 | 2 | |
| High | 17.7 | 1.7 | |
| N | % | ||
| Sex | |||
| Male | 80 | 55.60% | |
| Female | 64 | 44.40% | |
| Smoking status | |||
| Never | 60 | 41.70% | |
| Former | 65 | 45.10% | |
| Current | 19 | 13.20% | |
| Recent NSAID use | |||
| No | 89 | 62.70% | |
| Yes | 53 | 37.30% | |
| BMI (kg/m2) | |||
| <25 | 50 | 34.7% | |
| 25-30 | 53 | 36.8% | |
| >30 | 41 | 28.4% | |
No pathway genes significantly altered their own gene expression (data not shown in table). However, several pathway genes altered expression of other genes in the pathway (Table 2 shows all de-regulated genes with FDR of <0.1). Thirteen pathway genes and 17 SNPs within those genes were associated with altered expression of other genes in the pathway. The most significant associations (adjusted p values less than 0.05) were for BMPR1B rs17616243 and rs1863652 altering expression of ALDH1A1 and FGF1 respectively, MAP3K9 rs11628333 altering expression of TYK2 and AKT2, SMAD3 rs1498506 altering expression of NFκB1, TERT rs2853668 altering expression of TYK2 and IRF3, TGFB1 rs4803455 altering expression of IRS2, and TNF rs1800630 altering expression of TEK.
Table 2.
Pathway gene expression altered by other pathway genes
| SNP rs# (Inheritance model) | Gene | Mean expression | Log2 fold change | Adjusted p value | |||
|---|---|---|---|---|---|---|---|
|
|
|||||||
| Gene | Altered | Genotype 0 | Genotype 1 | p value | |||
| BMPR1A | rs7088641 (Dominant) | IL6R | 57.87 | 42.91 | -0.35 | 7.50E-04 | 9.75E-02 |
| BMPR1B | rs17616243 (Additive) | ALDH1A1 | 20.66 | 2.55 | -1.4 | 1.34E-04 | 2.68E-02 |
| NFAM1 | 10.19 | 0.44 | -1.29 | 7.14E-04 | 7.14E-02 | ||
| DDIT4 | 16.21 | 3.76 | -1.16 | 1.42E-03 | 9.49E-02 | ||
| rs1863652 (Recessive) | FGF1 | 1.46 | 6.08 | 0.59 | 6.05E-05 | 1.28E-02 | |
| EIF4EBP3 | 7.74 | 19.65 | 0.63 | 9.05E-04 | 9.59E-02 | ||
| IL6R | rs4845623 (Recessive) | AR | 39.12 | 87.89 | 0.47 | 7.16E-04 | 9.36E-02 |
| TCF7L2 | 111.08 | 80.34 | -0.35 | 1.22E-03 | 9.36E-02 | ||
| rs7549250 (Dominant) | STAT3 | 127.92 | 109.64 | -0.18 | 7.56E-03 | 9.18E-02 | |
| TCF7L2 | 89.02 | 111.63 | 0.25 | 7.65E-03 | 9.18E-02 | ||
| IL8 | rs4073 (Additive) | VEGFA | 125.18 | 166.75 | 0.33 | 6.80E-03 | 8.16E-02 |
| IRF3 | rs2304204 (Recessive) | NADSYN1 | 124.94 | 99.08 | -0.33 | 1.96E-03 | 9.40E-02 |
| MAP3K3 | rs11658329 (Dominant) | AR | 30.76 | 64.29 | 0.4 | 1.35E-03 | 9.58E-02 |
| MAP3K9 | rs11628333 (Recessive) | TYK2 | 70.2 | 106.66 | 0.53 | 1.11E-04 | 7.90E-03 |
| AKT2 | 62.48 | 84.71 | 0.37 | 1.01E-03 | 3.57E-02 | ||
| TSC2 | 73.43 | 103.79 | 0.41 | 4.02E-03 | 9.50E-02 | ||
| rs17176971 (Recessive) | PIK3CB | 43.12 | 21.03 | -0.99 | 4.04E-04 | 6.96E-02 | |
| AKT2 | 64.09 | 116.65 | 0.74 | 8.69E-04 | 6.96E-02 | ||
| NFAT5 | 200.65 | 109.9 | -0.73 | 1.22E-03 | 6.96E-02 | ||
| TYK2 | 73.26 | 144.3 | 0.87 | 1.73E-03 | 6.96E-02 | ||
| SMAD2 | 104.07 | 57.18 | -0.71 | 1.75E-03 | 6.96E-02 | ||
| MAP3K11 | 64.72 | 115.11 | 0.76 | 3.08E-03 | 9.97E-02 | ||
| LEPR | 7.23 | 23.15 | 1.04 | 4.60E-03 | 9.97E-02 | ||
| DUSP7 | 6.93 | 20.9 | 1.01 | 4.69E-03 | 9.97E-02 | ||
| SMAD4 | 65.76 | 39.2 | -0.74 | 4.77E-03 | 9.97E-02 | ||
| PRRX1 | 6.29 | 23.93 | 1.06 | 5.01E-03 | 9.97E-02 | ||
| SMAD3 | rs1498506 (Additive) | NFKB1 | 34.45 | 50.44 | 0.43 | 1.12E-04 | 2.65E-02 |
| SMAD7 | rs12953717 (Additive) | RAF1 | 71.17 | 91.1 | 0.3 | 9.88E-04 | 9.38E-02 |
| STAT5B | rs6503691 (Additive) | CALM2 | 146.84 | 74.06 | -0.7 | 9.09E-04 | 9.59E-02 |
| MAPK1 | 88.42 | 43.75 | -0.59 | 2.02E-03 | 9.59E-02 | ||
| AR | 40.4 | 130.94 | 0.83 | 3.59E-03 | 9.59E-02 | ||
| MAPK3 | 42.49 | 86.92 | 0.69 | 3.88E-03 | 9.59E-02 | ||
| PTEN | 81.58 | 42.12 | -0.64 | 4.06E-03 | 9.59E-02 | ||
| TERT | rs2736100 (Recessive) | ERBB2 | 110.38 | 150.32 | 0.37 | 4.00E-03 | 9.58E-02 |
| IRF1 | 85.86 | 62.54 | -0.38 | 6.90E-03 | 9.58E-02 | ||
| STAT1 | 135.39 | 93.11 | -0.41 | 7.98E-03 | 9.58E-02 | ||
| rs2853668 (Additive) | TYK2 | 69.98 | 135.4 | 0.73 | 1.26E-04 | 1.95E-02 | |
| IRF3 | 26.39 | 53.36 | 0.76 | 1.74E-04 | 1.95E-02 | ||
| PPARG | 43.67 | 25.14 | -0.72 | 9.25E-04 | 6.91E-02 | ||
| TSC2 | 73.63 | 126.51 | 0.62 | 1.50E-03 | 7.93E-02 | ||
| MAP3K11 | 61.11 | 97 | 0.56 | 2.04E-03 | 7.93E-02 | ||
| MMP2 | 58.24 | 116.28 | 0.74 | 2.12E-03 | 7.93E-02 | ||
| TRAF2 | 14.29 | 25.32 | 0.68 | 3.74E-03 | 9.98E-02 | ||
| STAT5A | 16 | 31.23 | 0.63 | 4.13E-03 | 9.98E-02 | ||
| TLR1 | 7.4 | 22.42 | 0.73 | 4.36E-03 | 9.98E-02 | ||
| DUSP4 | 7.47 | 15.92 | 0.72 | 4.46E-03 | 9.98E-02 | ||
| TGFB1 | rs4803455 (Additive) | IRS2 | 51.95 | 84.95 | 0.5 | 3.07E-04 | 2.55E-02 |
| JUNB | 62.6 | 90.06 | 0.39 | 2.11E-03 | 8.78E-02 | ||
| TNF | rs1800630 (Dominant) | TEK | 6.09 | 12.27 | 0.47 | 1.99E-04 | 4.70E-02 |
Genotype 0 refers to the homozygote common for the additive or dominant model and homozygote common/heterozygote for the recessive model. Genotype 1 refers to the homozygote variant for the additive or recessive model and heterozygote/homozygote variant for the dominant model. Log2 fold change = difference between coefficients (i.e. genotype1/2 coeff.-genotype 0 coeff.), on log2 scale.
BMI, NSAIDs use, and dietary components of the OBS were associated with altered gene expression (Table 3), although neither cigarette smoking nor the summary OBS altered expression of any pathway genes with a FDR of 0.1. Most factors altered only one gene, although folic acid, vitamin D, and lutein/zeaxanthin each altered expression of two pathway genes. SOD2 expression was altered by both folic acid and vitamin C, while RUNX2 expression was altered by both folic acid and beta carotene.
Table 3.
Associations between lifestyle factors and gene expression
| Gene | Average adjusted count | Log2 fold change1 | Adjusted p value | |||
|---|---|---|---|---|---|---|
|
|
||||||
| Variable | Altered | Low tertile | High tertile | P value | ||
| BMI | MAP3K7 | 42.15 | 28.15 | -0.39 | 2.50E-04 | 5.90E-02 |
| NSAIDs | IL6R | 56.85 | 42.20 | -0.34 | 7.90E-05 | 1.87E-02 |
| Folic acid | SOD2 | 201.42 | 154.47 | -0.32 | 1.45E-03 | 9.43E-02 |
| RUNX2 | 26.16 | 38.40 | 0.44 | 1.75E-04 | 2.27E-02 | |
| Vitamin D | STAT1 | 96.41 | 133.40 | 0.37 | 2.93E-03 | 8.59E-02 |
| IRF1 | 67.82 | 86.66 | 0.32 | 4.77E-03 | 8.59E-02 | |
| Beta carotene | RUNX2 | 23.00 | 33.68 | 0.41 | 8.05E-05 | 1.90E-02 |
| Lutein | MAPK14 | 72.92 | 88.79 | 0.26 | 4.10E-03 | 9.84E-02 |
| JUNB | 97.22 | 77.05 | -0.31 | 3.92E-03 | 9.84E-02 | |
| Vitamin C | SOD2 | 185.23 | 149.01 | -0.27 | 5.29E-03 | 6.35E-02 |
Log2 fold Change p value from DESeq 2 is the difference between coefficients (i.e. high tertile coeff.-low tertile coeff.) on log2 scale.
Focusing on genes and their associated SNPs that have previously been shown to interact with NSAIDs, cigarette smoking, BMI, and OBS to alter colon cancer risk, we observed that several of these interactions were associated with significant differential expression in pathway genes (Tables 4 and 5). NSAIDs interacted with 10 genes and 15 SNPs within those genes to alter expression of 28 pathway genes (Table 4). Three SNPs in KDR altered gene expression, while two SNPs in IRF2, TXNRD1, and TNXRD2 altered expression of multiple genes. Other groups of gene families that altered expression were interferon regulatory factors (IRF2, IRF5, and IRF6) altering SMAD6, EPX, NPY2R, MAP3K10, SOD2, PCK1, TEP1, and MYO15B. KDR, a VEGFA receptor, altered expression of CXCR2, NFAT5, AKT1, PTEN, CALM3, VDR, and DUSP4. Recent cigarette smoking interacted with seven genes and nine SNPs to alter expression of 27 genes. JAK2 rs10815160 altered expression of 13 of these genes, included altered expression of MTOR, BMP1 and 4, SEPN1, MAP3K3, TLR2, PDGF, SOD2, and PTGIS. Expression of MAPK1 was altered by two SNPs in NOS2A and VEGFA interacting with cigarette smoking. FLT1, a receptor for VEGFA, interacted with smoking to alter expression of VEGFA.
Table 4.
Interactions between recent aspirin/NSAID use and smoking with pathway genes on expression of genes in pathway
| Means of normalized counts | |||||||||
|
|
|||||||||
| No recent aspirin/NSAID use | Recent aspirin/NSAID use | ||||||||
|
|
|||||||||
| Gene SNP | Expression gene | Homozygote common | Heterozygote | Homozygote variant | Homozygote common | Heterozygote | Homozygote variant | P value | Adjusted p value |
|
| |||||||||
| EPX_rs10853004 | BMPR1A | 10.17 | 7.84 | 0.00 | 6.27 | 6.39 | 6.91 | 5.43E-06 | 1.28E-03 |
| PTGS1 | 18.68 | 13.60 | 3.67 | 15.56 | 12.93 | 21.62 | 1.83E-04 | 2.15E-02 | |
| MMP2 | 70.90 | 54.11 | 13.08 | 70.80 | 81.80 | 83.79 | 7.55E-04 | 5.94E-02 | |
| IL15_rs17461269_Rec | IRF2 | 21.13 | 12.71 | 17.44 | 33.25 | 1.10E-04 | 2.59E-02 | ||
| IRF2_rs3756093_Dom | SMAD6 | 7.62 | 14.58 | 12.11 | 5.65 | 4.58E-04 | 9.72E-02 | ||
| IRF2_rs9684244 | EPX | 1.01 | 0.17 | 0.61 | 0.25 | 2.44 | 0.00 | 2.74E-04 | 3.89E-02 |
| NPY2R | 0.95 | 0.09 | 1.09 | 0.01 | 0.84 | 0.00 | 3.30E-04 | 3.89E-02 | |
| IRF5_rs1874328 | MAP3K10 | 11.26 | 6.26 | 3.98 | 4.89 | 14.39 | 8.66 | 3.25E-04 | 7.67E-02 |
| IRF6_rs2013196_Dom | SOD2 | 160.71 | 131.18 | 138.05 | 179.08 | 9.76E-04 | 1.17E-02 | ||
| PCK1 | 137.18 | 112.33 | 99.75 | 182.62 | 9.37E-03 | 4.39E-02 | |||
| TEP1 | 151.71 | 162.52 | 166.69 | 120.52 | 1.10E-02 | 4.39E-02 | |||
| CALM2 | 143.26 | 123.86 | 146.08 | 183.63 | 2.26E-02 | 5.97E-02 | |||
| MYO15B | 224.71 | 250.99 | 316.52 | 227.45 | 2.49E-02 | 5.97E-02 | |||
| KDR_rs12502008 | CXCR2 | 1.28 | 0.38 | 1.33 | 0.01 | 3.28 | 0.62 | 3.72E-04 | 8.77E-02 |
| KDR_rs12505758 | NFAT5 | 217.15 | 163.37 | 277.80 | 174.75 | 225.63 | 224.75 | 1.23E-04 | 7.27E-03 |
| AKT1 | 63.07 | 83.02 | 50.67 | 80.30 | 62.72 | 169.60 | 3.63E-04 | 1.07E-02 | |
| PTEN | 85.89 | 71.04 | 59.42 | 66.62 | 86.68 | 110.31 | 6.87E-03 | 8.41E-02 | |
| CALM3 | 56.09 | 65.32 | 15.69 | 68.04 | 53.95 | 66.18 | 6.94E-03 | 8.41E-02 | |
| VDR | 65.99 | 66.13 | 31.90 | 67.61 | 67.89 | 108.93 | 7.13E-03 | 8.41E-02 | |
| KDR_rs2071559 | DUSP4 | 6.37 | 11.62 | 8.42 | 42.94 | 7.94 | 11.78 | 2.61E-05 | 6.16E-03 |
| SMAD3_rs7173811 | MAP3K11 | 78.03 | 58.13 | 61.42 | 51.75 | 71.07 | 64.89 | 1.36E-03 | 9.62E-02 |
| TERT_rs2853668_Rec | TEP1 | 155.26 | 137.59 | 144.04 | 276.95 | 3.57E-03 | 9.24E-02 | ||
| RAF1 | 76.37 | 100.04 | 73.28 | 50.24 | 6.52E-03 | 9.24E-02 | |||
| CALM2 | 144.25 | 135.69 | 146.86 | 48.94 | 7.35E-03 | 9.24E-02 | |||
| HIF1A | 70.84 | 86.23 | 68.33 | 29.58 | 7.70E-03 | 9.24E-02 | |||
| TXNRD1_rs4523760_Dom | TYK2 | 68.29 | 63.29 | 69.72 | 118.63 | 3.94E-04 | 9.30E-02 | ||
| TXNRD1_rs4964778_Dom | MAP3K10 | 7.91 | 6.72 | 8.37 | 21.96 | 8.74E-04 | 9.80E-02 | ||
| RNF146 | 14.77 | 20.44 | 15.95 | 8.39 | 1.11E-03 | 9.80E-02 | |||
| TXNRD2_rs3788314 | DUSP4 | 7.98 | 9.42 | 9.76 | 40.06 | 8.09 | 8.39 | 6.12E-04 | 9.36E-02 |
| TXNRD2_rs756661 | GC | 0.15 | 2.74 | 0.00 | 0.08 | 0.07 | 2.43 | 3.11E-04 | 7.34E-02 |
|
| |||||||||
| Non smoker/non recent smoker | Recent smoker | ||||||||
|
|
|
||||||||
| FLT1_rs17086609 | KLF6 | 229.24 | 239.11 | 221.55 | 308.38 | 177.89 | 131.96 | 9.97E-03 | 7.59E-02 |
| SOD2 | 148.05 | 147.92 | 138.79 | 210.38 | 136.58 | 251.93 | 1.99E-02 | 7.59E-02 | |
| MYO15B | 275.77 | 227.96 | 219.43 | 215.22 | 345.29 | 227.94 | 2.26E-02 | 7.59E-02 | |
| VEGFA | 134.10 | 138.03 | 120.01 | 211.52 | 123.18 | 149.96 | 2.53E-02 | 7.59E-02 | |
| JAK2_rs10815160_Rec | MTOR | 88.98 | 191.21 | 102.96 | 66.33 | 6.57E-05 | 1.31E-02 | ||
| BMP4 | 10.15 | 22.21 | 11.79 | 1.47 | 3.01E-04 | 3.01E-02 | |||
| SEPN1 | 45.62 | 62.67 | 46.34 | 19.27 | 9.51E-04 | 3.68E-02 | |||
| ACVR1 | 13.75 | 22.36 | 14.09 | 4.53 | 9.86E-04 | 3.68E-02 | |||
| TEK | 8.62 | 10.10 | 5.36 | 0.00 | 1.04E-03 | 3.68E-02 | |||
| MAP3K3 | 24.18 | 30.74 | 28.39 | 8.78 | 1.10E-03 | 3.68E-02 | |||
| TLR4 | 18.33 | 10.04 | 21.24 | 3.28 | 2.69E-03 | 7.69E-02 | |||
| PDGFB | 5.95 | 8.17 | 6.80 | 0.80 | 3.87E-03 | 8.41E-02 | |||
| PTGIS | 4.54 | 9.98 | 3.89 | 0.00 | 3.92E-03 | 8.41E-02 | |||
| SOD2 | 149.71 | 92.69 | 168.85 | 224.31 | 4.21E-03 | 8.41E-02 | |||
| MMP2 | 63.81 | 109.97 | 86.44 | 32.06 | 5.75E-03 | 9.87E-02 | |||
| NADSYN1 | 121.43 | 172.70 | 118.70 | 100.38 | 6.07E-03 | 9.87E-02 | |||
| BMP1 | 27.57 | 59.13 | 33.62 | 21.48 | 6.41E-03 | 9.87E-02 | |||
| MAP3K11_rs7116712_Rec | TLR3 | 6.66 | 9.41 | 7.17 | 0.00 | 2.35E-04 | 5.55E-02 | ||
| NFAM1_rs13055337_Dom | IL6R | 44.87 | 59.81 | 65.26 | 38.78 | 1.06E-03 | 8.84E-02 | ||
| NOS2A_rs2274894 | MYO15B | 252.75 | 249.64 | 251.49 | 424.95 | 234.31 | 126.57 | 2.21E-03 | 7.94E-02 |
| NOS2A_rs3729508_Rec | DUSP6 | 37.98 | 20.44 | 31.95 | 58.25 | 3.01E-05 | 7.07E-03 | ||
| MAPK1 | 86.36 | 62.87 | 81.36 | 123.68 | 1.53E-04 | 1.80E-02 | |||
| NOS2A_rs944725 | MAPK1 | 79.50 | 87.45 | 82.33 | 117.18 | 74.90 | 51.83 | 4.58E-04 | 7.00E-02 |
| RPS6KA2 | 28.15 | 25.33 | 28.98 | 20.39 | 42.41 | 18.08 | 1.80E-03 | 9.71E-02 | |
| PTGS1 | 17.50 | 13.84 | 14.55 | 8.80 | 28.80 | 35.35 | 1.90E-03 | 9.71E-02 | |
| TLR2_rs3804099_Rec | DUSP1 | 71.70 | 48.59 | 48.11 | 99.72 | 1.94E-03 | 9.30E-02 | ||
| VEGFA_rs3025033 | CALM2 | 145.27 | 146.44 | 0.00 | 160.76 | 131.48 | 143.56 | 2.97E-04 | 1.75E-02 |
| MAPK1 | 86.43 | 80.47 | 0.00 | 95.45 | 93.55 | 96.92 | 1.55E-03 | 4.56E-02 | |
| DUSP1 | 69.86 | 62.91 | 0.00 | 42.41 | 80.78 | 50.29 | 4.29E-03 | 8.43E-02 | |
Table 5.
Interactions between BMI and OBS and pathway genes on expression of genes in pathway
| Means of normalized counts | ||||||||||||
|
|
||||||||||||
| Normal (BMI <25) | Overweight (BMI 25 to <30) | Obese (BMI ≥30) | ||||||||||
|
|
||||||||||||
| Expressed | Homozygote | Homozygote | Homozygote | Homozygote | ||||||||
|
|
|
|
|
|||||||||
| Gene SNP | Gene | Common | Heterozygote | Variant | Common | Heterozygote | Variant | Common | Heterozygote | Variant | P value | Adjusted p value |
|
| ||||||||||||
| FLT1_rs3936415 | NRG2 | 0.27 | 5.91 | 0.71 | 0.92 | 0.96 | 2.69 | 1.25 | 0.28 | 1.81 | 1.30E-04 | 3.07E-02 |
| MMP9 | 2.77 | 6.69 | 2.20 | 2.32 | 4.12 | 6.72 | 5.46 | 1.54 | 17.18 | 4.62E-04 | 5.45E-02 | |
| KDR_rs2219471_Dom | AR | 21.70 | 47.64 | 50.87 | 27.88 | 93.02 | 31.83 | 8.89E-04 | 9.42E-02 | |||
| SEPN1_rs11247735 | AR | 45.74 | 35.82 | 25.08 | 17.82 | 38.66 | 67.67 | 9.33 | 84.12 | 119.79 | 1.43E-04 | 3.37E-02 |
| TERT | 8.72 | 0.86 | 2.50 | 1.08 | 1.78 | 2.03 | 0.14 | 1.03 | 0.95 | 5.27E-04 | 6.22E-02 | |
| TERT_rs10069690 | PCK1 | 163.78 | 88.71 | 125.30 | 112.00 | 136.08 | 136.84 | 102.53 | 171.90 | 74.87 | 3.97E-03 | 9.52E-02 |
| TERT_rs2242652_Dom | MMP9 | 6.20 | 1.64 | 4.06 | 4.20 | 3.18 | 7.71 | 4.53E-04 | 9.60E-02 | |||
| TXNRD2_rs1044732_Dom | SOD2 | 145.68 | 197.59 | 152.60 | 153.28 | 154.07 | 138.55 | 6.91E-03 | 8.30E-02 | |||
| VEGFA_rs25648_Rec | IL6R | 52.55 | 112.12 | 48.57 | 40.46 | 51.10 | 11.01 | 7.18E-04 | 9.33E-02 | |||
|
| ||||||||||||
| Low oxidative balance score | Intermediate oxidative balance score | High oxidative balance score | ||||||||||
|
|
|
|
||||||||||
| FLT1_rs12429309_Dom | AR | 63.42 | 23.92 | 71.81 | 26.58 | 27.77 | 73.42 | 9.33E-04 | 9.89E-02 | |||
| KDR_rs1531290 | SDR16C5 | 2.16 | 1.91 | 17.58 | 0.75 | 1.22 | 2.48 | 3.62 | 1.82 | 0.15 | 8.56E-05 | 2.02E-02 |
| KDR_rs2305948_Dom | FGF1 | 0.78 | 3.92 | 1.11 | 2.08 | 3.46 | 0.12 | 1.24E-04 | 1.15E-02 | |||
| PIK3CG | 12.40 | 30.55 | 23.17 | 17.26 | 19.50 | 8.93 | 1.42E-04 | 1.15E-02 | ||||
| MAP3K1 | 36.61 | 63.92 | 43.50 | 45.74 | 47.40 | 30.29 | 1.46E-04 | 1.15E-02 | ||||
| TYK2 | 72.83 | 116.78 | 78.62 | 62.03 | 76.74 | 45.71 | 2.06E-04 | 1.22E-02 | ||||
| TGFB2 | 1.35 | 8.86 | 2.26 | 1.64 | 2.68 | 4.20 | 3.91E-04 | 1.85E-02 | ||||
| TLR3 | 5.69 | 15.19 | 8.06 | 5.35 | 7.36 | 2.39 | 1.83E-03 | 7.19E-02 | ||||
| MMP2 | 56.10 | 108.50 | 65.09 | 74.35 | 73.00 | 46.60 | 2.24E-03 | 7.54E-02 | ||||
| DHCR7 | 5.54 | 12.53 | 6.29 | 7.92 | 9.03 | 3.93 | 3.00E-03 | 8.85E-02 | ||||
| NOS2A_rs4795067 | KLF6 | 231.23 | 204.66 | 303.46 | 188.84 | 236.28 | 211.24 | 276.26 | 251.45 | 128.71 | 8.10E-03 | 9.72E-02 |
| NOS2A_rs8072199_Rec | MMP2 | 59.82 | 109.87 | 73.42 | 32.12 | 74.12 | 33.76 | 5.97E-04 | 9.85E-02 | |||
There were fewer interactions between BMI and OBS and genes to alter gene expression than observed for either NSAIDs or cigarette smoking (Table 5). BMI interacted with FLT1, KDR, SEPN1, TERT, TXNRD2, and VEGFA to alter expression of eight genes. Interestingly, both KDR and SEPN1 altered expression of AR, while TXNRD2, a selenoprotein like SEPN1, altered expression of SOD2. VEGFA altered expression of IL6R. Three genes and five SNPs interacted with OBS to alter expression of 12 genes. Two of these genes, FLT1 and KDR, are VEGFA receptors, while the other gene, inducible nitric oxide synthase (NOS2A), is induced by cytokines and has been reported to play a role in oxidative stress-induced inflammation. FLT1 also interacted with OBS to alter AR expression.
Discussion
Our assessment of the CHIEF pathway genes and diet and lifestyle factors related to oxidative stress and inflammation supported our hypothesis that genes in the pathway influence genes expression of pathway genes and interact with lifestyle factors to alter gene expression of pathway genes. These findings add to the body of literature that recognize the importance of including lifestyle factors to more fully understand how genetic factors contribute to risk. The major changes in gene expression came from the combination of genetic and lifestyle factors rather than either genetic or lifestyle factor independently.
The genetic variants that we identified as altering gene expression did not alter their own gene expression, but altered expression of other genes in the pathway. The TGFβ-signaling pathway has repeatedly been associated with colon cancer [26,27]. In our data of the 13 genes influencing expression of other genes, five were in the TGFβ-signaling pathway, BMPR1A, BMPR1B, SMAD3, SMAD7, and TGFB1. Several studies suggest the importance of the BMP receptors, given that BMPs signal through their type I and II receptors 28. BMPR1A and BMPR1B are the two best characterized type I receptors. Substrates for these receptors include Smad proteins that play a central role in BMP signaling. Smad7 also is involved in inflammation-related pathways and has been shown to modulate TGF-β and wnt-signaling [29]. Genetic variation in the Smad7 gene on 8q21 has been identified through numerous genome-wide association studies (GWAS) as being associated with colorectal cancer [30]. SMAD3 rs1498506 previously associated with colon cancer after multiple comparison adjustment, was associated with NFκB1 expression; SMAD7 rs12953717, which we have previously identified as influencing colon cancer and has been identified through numerous genome-wide association studies (GWAS) as being associated with colorectal cancer (CRC) [30], altered expression of RAF1, a member of the MAP3K family. We previously reported that TGFB1 rs4803455 altered colon cancer risk by interacting with SMAD2 [15]. Two MAP3K family members, MAP3K3 and MAP3K9, altered expression of several genes, with MAP3K9 altering expression of 13 genes. TERT also altered expression of 13 genes, including ERBB2, STAT1, PPARG, TSC2, STAT5A, TLR1, and MAP3K11. One of the TERT SNPs that we found to influence expression levels of 10 genes, was rs2853668, which associated with colon cancer risk in our data and was identified in a genome-wide association study (GWAS) to be significantly associated with colorectal cancer risk [31]. These findings provide support for previously identified associations.
There have been few studies that have evaluated diet and lifestyle factors with gene expression [32-34]. Challenges of evaluating diet and other lifestyle factors with gene expression stem from the fact that the gene expression profile is time dependent and therefore relevant to current exposure. Exposure from current diet and lifestyle factors may not be recent enough to maintain an altered gene expression even if they do have an effect. Thus, although we assessed current smokers, current NSAID users, and diet close to the time of diagnosis, when the tissue would have been biopsied, the exposure may have been too far removed to impact expression. This could explain why few genes were de-regulated for NSAIDs specifically or for dietary factors and the OBS which was comprised mainly of dietary factors. BMI, which is a more constant exposure, was only associated with MAP3K7 expression. Likewise, nutrients evaluated were only associated with one or two genes when applying a FDR of <0.1.
Our previous analyses have shown that multiple lifestyle factors interact with pathway genes to influence colon cancer risk. Important lifestyle factors previously identified as interacting with genes in the pathway are diet and MAPK [9,10], angiogenesis [8], and TLR [11] genes; NSAIDs use with angiogenesis [8], estrogen-related genes [18], cytokines [12,13], MAPK [9], JAK/STAT [14], TGFβ-signaling pathway [15], TLR [11], TERT [35], and selenoproteins [16]; cigarette smoking with angiogenesis [8], cytokines [13,17], MAPK [9], JAK/STAT [14], TGFβ [15], and selenoproteins [16] sub-pathways; BMI with angiogenesis [8], estrogen-related genes [18], VDR [19], TERT [35], and cytokines [13,17]. In these analyses, we have shown that many of the previously identified interactions alter gene expression in the CHIEF pathway. We demonstrated that many of the interactions between specific SNPs and lifestyle factors altered gene expression of important pathway genes, lending support for the previously identified associations. The SNPs within these genes that we previously reported significant interactions with, are also the SNPs previously identified as interacting with lifestyle factors. We identified altered gene expression associated with 50% of the previously identified interactions for NSAIDS and OBS, 29% of the previously identified interactions with smoking, and 35% of those identified with BMI. For instance, with NSAIDs we detected significant associations for 2 of the 4 previously identified IRF2 SNPs that altered colon cancer risk; we observed significant differential gene expression with the only significant interaction between NSAIDs and IRF5, and for one of the two IRF6 SNPs. For FLT1, KDR, and VEGF, we observed significant interaction for all previously identified SNP interacting with NSAID use, cigarette smoking, and BMI. Likewise for TERT we saw differential gene expression associated with the previously identified interaction between rs2853668 and NSAIDs use and TERT rs10069690 and rs2242652 with BMI. For TXNRD1 and TXNRD2 we were able to replicate significant findings with NSAIDs for four of the previously identified six SNPs that interacted with NSAIDs to alter colon cancer risk.
While we have provided data that support a biological basis for previously identified interactions, our study has limitations. As we previously stated, gene expression data represent expression at a given point in time, in this case at the time of surgery to remove the colon tumor. Various variables are impacted by this limitation. Genotypes do not change over time and therefore should not vary by the timing of the biopsy, while variables such as NSAID use and diet would have variability and likely inhibit our ability to detect associations that might exist. Likewise, the associations which we did detect could have a larger change if the measurement had been closer to the time of tissue extraction. Additionally, we have utilized colonic non-tumor tissue, so genes would have to be expressed in colon tissue for detection. These diet and lifestyle factors could influence other genes in other tissue sources. Utilizing colonic tissue that is located close to a tumor could alter gene expression, however, it would alter it is a manner that was not dependent on diet or lifestyle exposure. While we did not show altered gene expression with all previously identified interactions, it should be kept in mind that altered gene expression is one aspect of functionality and failure to see an association does not mean that genes couldn’t alter protein levels or other aspects of functionality we were unable to evaluate.
Our data illustrate the importance of evaluating broader pathways to determine functionality of genes and SNPs. Genes within pathways were more likely to have expression altered than were the genes themselves. The importance of evaluating the interaction between genes and environment in terms of functionality also has been demonstrated. In many instances, the combination of factors altered expression of pathway genes that neither gene nor lifestyle factor altered alone. This is one of the first, if not the first study, to illustrate this point. However, since these findings are from only one small study, replication is needed in other similar studies.
Acknowledgements
This study was funded by NCI grants CA48998. The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official view of the National Cancer Institute. We would like to acknowledge the contributions of Dr. Bette Caan, Judy Morse and Donna Schaffer and the Kaiser Permanente Medical Research Program, and Sandra Edwards for data collection and organization, Jennifer Herrick for data management, Erica Wolff and Michael Hoffman for RNA extraction, Wade Samowitz for slide review, and Brett Milash at the Bioinformatics Core Facility at the University of Utah.
Disclosure of conflict of interest
None.
Supporting Information
References
- 1.Lee PH, Shatkay H. F-SNP: computationally predicted functional SNPs for disease association studies. Nucleic Acids Res. 2008;36:D820–4. doi: 10.1093/nar/gkm904. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Yuan HY, Chiou JJ, Tseng WH, Liu CH, Liu CK, Lin YJ, Wang HH, Yao A, Chen YT, Hsu CN. FASTSNP: an always up-to-date and extendable service for SNP function analysis and prioritization. Nucleic Acids Res. 2006;34:W635–41. doi: 10.1093/nar/gkl236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Penney RB, Lundgreen A, Yao-Borengasser A, Koroth-Edavana V, Williams S, Wolff R, Slattery ML, Kadlubar S. Lack of correlation between in silico projection of function and quantitative real-time PCR-determined gene expression levels in colon tissue. Pharmgenomics Pers Med. 2013;6:99–103. doi: 10.2147/PGPM.S49199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Slattery ML, Lundgreen A, Mullany LE, Penney RB, Wolff RK. Influence of CHIEF pathway genes on gene expression: a pathway approach to functionality. Int J Mol Epidemiol Genet. 2014;5:100–11. [PMC free article] [PubMed] [Google Scholar]
- 5.Gonzalez-Moreno O, Calvo A, Joshi BH, Abasolo I, Leland P, Wang Z, Montuenga L, Puri RK, Green JE. Gene expression profiling identifies IL-13 receptor alpha 2 chain as a therapeutic target in prostate tumor cells overexpressing adrenomedullin. Int J Cancer. 2005;114:870–8. doi: 10.1002/ijc.20789. [DOI] [PubMed] [Google Scholar]
- 6.Hoffstedt J, Eriksson P, Mottagui-Tabar S, Arner P. A polymorphism in the leptin promoter region (-2548 G/A) influences gene expression and adipose tissue secretion of leptin. Horm Metab Res. 2002;34:355–9. doi: 10.1055/s-2002-33466. [DOI] [PubMed] [Google Scholar]
- 7.Slattery ML, Wolff RK, Lundgreen A. A pathway approach to evaluating the association between the CHIEF pathway and risk of colorectal cancer. Carcinogenesis. 2015;36:49–59. doi: 10.1093/carcin/bgu213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Slattery ML, Lundgreen A, Welbourn B, Wolff RK, Corcoran C. Oxidative balance and colon and rectal cancer: Interaction of lifestyle factors and genes. Mutat Res. 2012;734:30–40. doi: 10.1016/j.mrfmmm.2012.04.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Slattery ML, Lundgreen A, Wolff RK. MAP kinase genes and colon and rectal cancer. Carcinogenesis. 2012;33:2398–408. doi: 10.1093/carcin/bgs305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Slattery ML, Lundgreen A, Wolff RK. Dietary Influence on MAPK-Signaling Pathways and Risk of Colon and Rectal Cancer. Nutr Cancer. 2013;65:729–38. doi: 10.1080/01635581.2013.795599. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Slattery ML, Herrick JS, Bondurant KL, Wolff RK. Toll-like receptor genes and their association with colon and rectal cancer development and prognosis. Int J Cancer. 2012;130:2974–80. doi: 10.1002/ijc.26314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Slattery ML, Lundgreen A, Bondurant KL, Wolff RK. Tumor necrosis factor-related genes and colon and rectal cancer. Int J Mol Epidemiol Genet. 2011;2:328–38. [PMC free article] [PubMed] [Google Scholar]
- 13.Bondurant KL, Lundgreen A, Herrick JS, Kadlubar S, Wolff RK, Slattery ML. Interleukin genes and associations with colon and rectal cancer risk and overall survival. Int J Cancer. 2013;132:905–15. doi: 10.1002/ijc.27660. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Slattery ML, Lundgreen A, Kadlubar SA, Bondurant KL, Wolff RK. JAK/STAT/SOCSsignaling pathway and colon and rectal cancer. Mol Carcinog. 2013;52:155–66. doi: 10.1002/mc.21841. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Slattery ML, Lundgreen A, Wolff RK, Herrick JS, Caan BJ. Genetic variation in the transforming growth factor-beta-signaling pathway, lifestyle factors, and risk of colon or rectal cancer. Dis Colon Rectum. 2012;55:532–40. doi: 10.1097/DCR.0b013e31824b5feb. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Slattery ML, Lundgreen A, Welbourn B, Corcoran C, Wolff RK. Genetic variation in selenoprotein genes, lifestyle, and risk of colon and rectal cancer. PLoS One. 2012;7:e37312. doi: 10.1371/journal.pone.0037312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Slattery ML, Lundgreen A, Bondurant KL, Wolff RK. Interferon-signaling pathway: associations with colon and rectal cancer risk and subsequent survival. Carcinogenesis. 2011;32:1660–7. doi: 10.1093/carcin/bgr189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Slattery ML, Lundgreen A, Herrick JS, Kadlubar S, Caan BJ, Potter JD, Wolff RK. Variation in the CYP19A1 gene and risk of colon and rectal cancer. Cancer Causes Control. 2011;22:955–63. doi: 10.1007/s10552-011-9768-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Slattery ML, Murtaugh M, Caan B, Ma KN, Wolff R, Samowitz W. Associations between BMI, energy intake, energy expenditure, VDR genotype and colon and rectal cancers (United States) Cancer Causes Control. 2004;15:863–72. doi: 10.1007/s10552-004-1048-6. [DOI] [PubMed] [Google Scholar]
- 20.Slattery ML, Fitzpatrick FA. Convergence of hormones, inflammation, and energy-related factors: a novel pathway of cancer etiology. Cancer Prev Res (Phila) 2009;2:922–30. doi: 10.1158/1940-6207.CAPR-08-0191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Slattery ML, Edwards SL, Palmer L, Curtin K, Morse J, Anderson K, Samowitz W. Use of archival tissue in epidemiologic studies: collection procedures and assessment of potential sources of bias. Mutat Res. 2000;432:7–14. doi: 10.1016/s1383-5726(99)00010-2. [DOI] [PubMed] [Google Scholar]
- 22.Edwards S, Slattery ML, Mori M, Berry TD, Caan BJ, Palmer P, Potter JD. Objective system for interviewer performance evaluation for use in epidemiologic studies. Am J Epidemiol. 1994;140:1020–8. doi: 10.1093/oxfordjournals.aje.a117192. [DOI] [PubMed] [Google Scholar]
- 23.Liu K, Slattery M, Jacobs D Jr, Cutter G, McDonald A, Van Horn L, Hilner JE, Caan B, Bragg C, Dyer A, et al. A study of the reliability and comparative validity of the cardia dietary history. Ethn Dis. 1994;4:15–27. [PubMed] [Google Scholar]
- 24.Benjamini YaH Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society. 1995;57:289–300. [Google Scholar]
- 25.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNAseq data with DESeq2. Genome Biol. 2014;15:550. doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Slattery ML, Herrick J, Lundgreen A, Wolff RK. Genetic variation in the TGF{beta}-signaling pathway and colon and rectal cancer risk. Cancer Epidemiol Biomarkers Prev. 2011;20:57–69. doi: 10.1158/1055-9965.EPI-10-0843. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Slattery ML, Lundgreen A, Herrick JS, Kadlubar S, Caan BJ, Potter JD, Wolff RK. Genetic variation in bone morphogenetic protein and colon and rectal cancer. International journal of cancer. J Int Du Cancer. 2012;130:653–64. doi: 10.1002/ijc.26047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Chen D, Zhao M, Mundy GR. Bone morphogenetic proteins. Growth Factors. 2004;22:233–41. doi: 10.1080/08977190412331279890. [DOI] [PubMed] [Google Scholar]
- 29.ten Dijke P, Hill CS. New insights into TGF-beta-Smad signalling. Trends in biochemical sciences. 2004;29:265–73. doi: 10.1016/j.tibs.2004.03.008. [DOI] [PubMed] [Google Scholar]
- 30.Broderick P, Carvajal-Carmona L, Pittman AM, Webb E, Howarth K, Rowan A, Lubbe S, Spain S, Sullivan K, Fielding S, Jaeger E, Vijayakrishnan J, Kemp Z, Gorman M, Chandler I, Papaemmanuil E, Penegar S, Wood W, Sellick G, Qureshi M, Teixeira A, Domingo E, Barclay E, Martin L, Sieber O CORGI Consortium. Kerr D, Gray R, Peto J, Cazier JB, Tomlinson I, Houlston RS. A genome-wide association study shows that common alleles of SMAD7 influence colorectal cancer risk. Nat Genet. 2007;39:1315–7. doi: 10.1038/ng.2007.18. [DOI] [PubMed] [Google Scholar]
- 31.Peters U, Hutter CM, Hsu L, Schumacher FR, Conti DV, Carlson CS, Edlund CK, Haile RW, Gallinger S, Zanke BW, Lemire M, Rangrej J, Vijayaraghavan R, Chan AT, Hazra A, Hunter DJ, Ma J, Fuchs CS, Giovannucci EL, Kraft P, Liu Y, Chen L, Jiao S, Makar KW, Taverna D, Gruber SB, Rennert G, Moreno V, Ulrich CM, Woods MO, Green RC, Parfrey PS, Prentice RL, Kooperberg C, Jackson RD, Lacroix AZ, Caan BJ, Hayes RB, Berndt SI, Chanock SJ, Schoen RE, Chang-Claude J, Hoffmeister M, Brenner H, Frank B, Bézieau S, Küry S, Slattery ML, Hopper JL, Jenkins MA, Le Marchand L, Lindor NM, Newcomb PA, Seminara D, Hudson TJ, Duggan DJ, Potter JD, Casey G. Meta-analysis of new genome-wide association studies of colorectal cancer risk. Hum Genet. 2011;131:217–34. doi: 10.1007/s00439-011-1055-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Hu R, Kong AN. Activation of MAP kinases, apoptosis and nutrigenomics of gene expression elicited by dietary cancer-prevention compounds. Nutrition. 2004;20:83–8. doi: 10.1016/j.nut.2003.09.015. [DOI] [PubMed] [Google Scholar]
- 33.Ornish D, Magbanua MJ, Weidner G, Weinberg V, Kemp C, Green C, Mattie MD, Marlin R, Simko J, Shinohara K, Haqq CM, Carroll PR. Changes in prostate gene expression in men undergoing an intensive nutrition and lifestyle intervention. Proc Natl Acad Sci U S A. 2008;105:8369–74. doi: 10.1073/pnas.0803080105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Landi MT, Dracheva T, Rotunno M, Figueroa JD, Liu H, Dasgupta A, Mann FE, Fukuoka J, Hames M, Bergen AW, Murphy SE, Yang P, Pesatori AC, Consonni D, Bertazzi PA, Wacholder S, Shih JH, Caporaso NE, Jen J. Gene expression signature of cigarette smoking and its role in lung adenocarcinoma development and survival. PLoS One. 2008;3:e1651. doi: 10.1371/journal.pone.0001651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Pellatt AJ, Wolff RK, Herrick J, Lundgreen A, Slattery ML. TERT’s role in colorectal carcinogenesis. Mol Carcinog. 2013;52:507–13. doi: 10.1002/mc.21885. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.