Gastrointestinal Symptom Cluster is Associated With Epigenetic Regulation of Lymphotoxin Beta in Oncology Patients Receiving Chemotherapy

Patients and Settings

This analysis is part of a larger study that evaluated symptom clusters in oncology outpatients receiving chemotherapy (Harris, Kober, Cooper, et al., 2022). Eligible patients were ≥18 years of age; had a diagnosis of breast, lung, gastrointestinal, or gynecologic cancer; had received chemotherapy within the preceding 4 weeks; were scheduled to receive at least two additional cycles of chemotherapy; were able to read, write, and understand English; and gave written informed consent. Patients were recruited from two Comprehensive Cancer Centers, one Veteran’s Affairs hospital, and four community-based oncology programs.

Study Procedures

The study was approved by the Institutional Review Board at each of the study sites. Of the 2234 patients approached, 1343 consented to participate (60.1% response rate). The major reason for refusal was being overwhelmed with their cancer treatment. Eligible patients were approached in the infusion unit during their first or second cycle of chemotherapy by a member of the research team to discuss study participation and obtain written informed consent. Data from the enrollment assessment (i.e., symptoms in the week prior to the patient’s second or third cycle of chemotherapy) were used in this analysis. At enrollment, a total of 1071 patients provided a blood sample for the DNA methylation analyses. Medical records were reviewed for disease and treatment information.

Instruments

Patients completed a demographic questionnaire, Karnofsky Performance Status (KPS) scale (Karnofsky, 1977), and Self-Administered Comorbidity Questionnaire (Sangha et al., 2003). Toxicity of each patient’s chemotherapy regimen was rated using the MAX2 index (Extermann et al., 2004).

A modified version of the 32-item Memorial Symptom Assessment Scale (MSAS) was used to evaluate the occurrence, severity, and distress of 38 common symptoms associated with cancer and its treatment (Portenoy et al., 1994). Six additional symptoms were added: hot flashes, chest tightness, difficulty breathing, abdominal cramps, increased appetite, and weight gain. Using the MSAS, patients were asked to indicate whether they had experienced each symptom in the past week (i.e., symptom occurrence). The patients’ responses to the occurrence items were used to create the symptom clusters. The validity and reliability of the MSAS are well-established (Portenoy et al., 1994).

Data Analyses

Descriptive statistics and frequency distributions were calculated for the demographic and clinical characteristics, using the Statistical Package for the Social Sciences Version 27 (IBM Corporation, Armonk, NY). Exploratory factor analysis (EFA) was used to identify symptom clusters using Mplus Version 8.6 (Muthén & Muthén, 2019).

Methods for the EFA were reported elsewhere (Harris, Kober, Cooper, et al., 2022). In brief, using the dichotomous occurrence items, tetrachoric correlations were used to create the matrix of associations (Muthén & Muthén, 2019). The simple structure for the occurrence EFA was estimated using the method of unweighted least squares with geomin (i.e., oblique) rotation (Muthén & Muthén, 2019). Factor loadings were considered meaningful if the loading was ≥0.40 (Muthén & Muthén, 2019). Factors (i.e., symptom clusters) were adequately defined if at least two items (i.e., symptoms) had loadings of ≥0.40 (Brown, 2015). Clusters were named based on the symptoms with the highest factor loadings and the majority of the symptoms within the cluster.

With these methods, a gastrointestinal cluster (Figure 1) was identified in our previous analysis (Harris, Kober, Cooper, et al., 2022). A factor score was calculated as the sum of the occurrence ratings for the 11 symptoms within the cluster (range of 0–11). Initially, the DNA methylation analyses were conducted using the patients’ symptom cluster factor scores as a continuous value. However, the p-value distribution for the differential methylation tests across the genome was severely conservative (i.e., underabundance of low p-values; data not shown). Therefore, for the current analyses, the total factor score was dichotomized into two groups (i.e., 0 symptoms = no gastrointestinal cluster group vs. 1–11 symptoms = gastrointestinal cluster group).

Selection of Trans eCpG Loci

Candidate genes in the NF-κB signaling pathway were identified using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (Kanehisa & Goto, 2000; Supplemental Figure 1). Then, CpG sites for these genes that were known to have methylation values associated with changes in gene expression on another chromosome (i.e., trans eCpGs; Kennedy et al., 2018) were selected. These trans eCpG loci for genes within the NF-κB signaling pathway were evaluated for associations with gastrointestinal cluster group membership.

Biospecimen Processing, Quantification of Methylation Status, and Quality Control

Methods for the DNA methylation analyses are described in detail elsewhere (Kober et al., 2020). In brief, DNA was extracted from archived buffy coats using the PUREGene DNA isolation kit (Invitrogen, Carlsbad, CA); quantified using a NanoDrop UV spectrophotometer (Thermo Fisher Scientific, Waltham, MA); and normalized to a concentration of 50 ng per microliter. DNA was bisulfite converted using the Zymo EZ-96 DNA Methylation Kit (Catalog #D5004) Deep-Well Format (Zymo Research, Irvine, CA) and used as input for the Infinium HD Methylation Assay (Illumina Inc, San Diego, CA).

Of the 1071 patients in this study, DNA methylation was measured for 925 patients using the Infinium MethylationEPIC BeadChip (i.e., EPIC microarray sample) and for 146 patients using the Infinium HumanMethylation 450 BeadChip (i.e., 450K microarray sample; Illumina, Inc, San Diego, CA). The EPIC microarray sample was used as a discovery sample while the 450K microarray sample was used as a validation sample. All of the samples were scanned on the Illumina iScan (Illumina, Inc, San Diego, CA). Preliminary analysis and quality control procedures were performed using GenomeStudio (Illumina, Inc, San Diego, CA). Samples that had <90% of their targets detected at a p-value of ≤ 0.01 were flagged for review. Sample replicates and Jurkat control replicates were checked to ensure an r² value of > 0.99.

Subsequent analyses were done using well-established protocols in R (version 4.1.0; Bock, 2012). Corrections for Infinium I and II probes, balance correction, background correction, and quantile normalization were performed using the minfi package in R (version 1.38.0; Aryee et al., 2014; Du et al., 2008). Probes that contained a single nucleotide polymorphism at a CpG or flanking site and probes that aligned with multiple places on the genome were excluded (Chen et al., 2013). Methylation scores were quantified as M-values (Du et al., 2010).

DNA Methylation Analyses

Given that DNA methylation levels differ among blood cell types (McGregor et al., 2016), cell types were estimated using the estimateCellCounts2() function in the FlowSorted.Blood. EPIC R package (version 1.10.1; Salas & Koestler, 2021). Cell type deconvolution was performed using the IDOL L-DMR library for cluster of differentiation 8 (CD8) and CD4 T-cells, natural killer cells, B cells, monocytes, and neutrophils (Salas et al., 2018). Differences in estimates of cell type composition between the gastrointestinal cluster groups were evaluated using Welch two sample t-tests and assessed for significance at a p-value of <0.05. Any cell type composition estimates that were significantly associated with membership in the gastrointestinal cluster group were included as covariates in the final model. Given that methylation status changes over the lifespan (Jones et al., 2015), age was included as a covariate in the final regression model. Surrogate variable analysis, using the Leek method (R package version 3.4.0; Leek & Storey, 2007), was used to estimate surrogate variables for technical and non-technical variations that contributed to heterogeneity in the sample that were not due to the gastrointestinal cluster group, age, or cell type.

To evaluate for associations between gastrointestinal cluster group membership and methylation status of trans eCpG loci for NF-κB genes, tests for differentially methylated probes (DMPs) were done using a generalized linear model implemented in the limma R package using the “ls” method (version 3.48.3; Ritchie et al., 2015). The significance of the DMPs for each of the NF-κB genes was assessed using an exploratory false discovery rate (FDR) of 25% under the Benjamini-Hochberg procedure for the EPIC microarray sample (Benjamini & Hochberg, 1995). Then, candidate trans eCpG loci identified as differentially methylated in the EPIC microarray sample were evaluated for differential methylation in the 450K microarray sample (Supplemental Figure 1). For the validation assessment using the 450K microarray sample, significance of the candidate trans eCpG loci was assessed at an unadjusted p-value of 0.05.

Finally, in order to characterize potential functional roles for these eCpGs, we evaluated for evidence of regulatory elements in regions surrounding these loci using annotation data from the Encyclopedia of DNA Elements (ENCODE; Rosenbloom et al., 2013) obtained from the University of California Santa Cruz Genome Browser (Kent et al., 2002). Finally, we identified predicted functional partners for the genes with differentially methylated trans eCpGs from a protein-protein interaction network that was created using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database (Szklarczyk et al., 2019).

Demographic and Clinical Characteristics

Of the 925 patients in the EPIC microarray sample, one was excluded for insufficient phenotypic data and one for poor sample quantification. Of the remaining 923 patients, 76.2% were female, 69.4% were White, 64.1% were married or partnered, and had a mean age of 57.5 (±12.2) years (Table 1). Most patients were well-educated (16.1 ± 3.0 years), exercised on a regular basis (71.6%), and had never smoked (66.4%). Patients had an average of 2.4 (±1.4) comorbid conditions and a KPS score of 80.4 (±12.6). The most common type of cancer was breast (39.5%), followed by gastrointestinal (34.0%), gynecological (15.9%), and lung (10.5%). Patients reported 13.5 (±7.1) concurrent symptoms before their second or third cycle of chemotherapy.

Of the 146 patients in the 450K microarray sample, 100% were female, 65.5% were White, 67.6% were married or partnered, and had a mean age of 52.7 (±11.7) years (Table 2). Most patients were well-educated (16.3 ± 2.9 years), exercised on a regular basis (75.7%), and had never smoked (72.4%). Patients had an average of 2.4 (±1.4) comorbid conditions and a KPS score of 79.1 (±11.6). The most common type of cancer was breast (99.3%) followed by gastrointestinal (0.7%). Patients reported 16.0 (±7.8) concurrent symptoms before their second or third cycle of chemotherapy.

DNA Methylation Analyses

For the EPIC microarray sample, of the 90 genes that were identified in the NF-κB signaling pathway, 3785 trans eCpG loci across 56 genes were evaluated for differential methylation. Because cell type composition was not associated with gastrointestinal cluster group membership, only age and 22 surrogate variables were included as covariates in the final model. For the 450K microarray validation sample, because cell type composition was not associated with gastrointestinal cluster group membership, only age and one surrogate variable were included as covariates in the final model.

For the EPIC microarray sample, hypermethylation of one trans eCpG locus (i.e., cg03171795) for the lymphotoxin beta (LTB) gene was found to be significantly associated with the occurrence of the gastrointestinal cluster (FDR = 0.168). For the 450K microarray sample, while no association was found (p = 0.664), the direction of methylation for this locus was the same across both samples (i.e., positive log fold change; view supplemental data files at https://zenodo.org/record/6638570).

Regulatory Role of Trans eCpG Locus

Given that epigenetic modifications are dynamic and multiple factors influence gene expression, one of the challenges for methylation association studies is the establishment of a functional role for the epigenetic variation that is identified (Rakyan et al., 2011). As illustrated in Figure 2, multiple sources of independent and complementary data show that the trans eCpG locus, cg03171795, is located within a putative regulatory region of chromosome 3. Specifically, evidence of regulatory elements compiled by ENCODE suggests that this locus is situated within a region of enhancer activity (ENCODE Project Consortium, 2012; Ernst et al., 2011). Enhancer regions are areas of non-coding DNA that enhance gene transcription through the recruitment of transcription factors and RNA polymerase II and modification of chromatin accessibility (Karnuta & Scacheri, 2018). Located distal to their target genes, enhancers form loops to move in closer to the promoter region of a gene (Andersson et al., 2015). Our findings are consistent with a study that sought to identify and characterize genome-wide eCpGs in two independent datasets and found that eCpGs were enriched for enhancer annotations and transcription factor binding sites, particularly among trans eCpGs (Kennedy et al., 2018).

Role of LTB in Inflammatory Processes

LTB is a member of the TNF super family of ligands. Along with LTA, it forms a heterotrimer complex (i.e., LTA₁B₂) that exclusively binds to the LTB receptor (LTBR; Borelli & Irla, 2021). While LTA₁B₂ is expressed on the surface of lymphoid cells (e.g., B cells, natural killer cells, T cells), LTBR is expressed on stromal and myeloid cells (e.g., dendritic cells, macrophages). Signaling between these distinct cell types is important for the formation and maintenance of lymphoid tissue. While notable for its role in embryonic lymph node and Peyer’s patch formation; splenic structure maintenance; and lymph node homeostasis (Sedy et al., 2014), LTB may play a role in the regulation of mucosal immune responses of the gastrointestinal tract (Upadhyay & Fu, 2013).

From its position at the beginning of the NF-κB signaling pathway, LTB can induce NF-κB signaling and inflammatory response(s). As illustrated in Figure 3, LTB, LTA, and LTBR form a close, interacting network with three TNF receptor associated factors (TRAFs; i.e., TRAF2, TRAF3, TRAF5) and two major receptors for TNF (i.e., TNF receptor super family (TNFRSF) 1A, TNFRSF1B). These TRAFs are intracellular signaling molecules that regulate the canonical and non-canonical pathways that lead to NF-κB activation (Shi & Sun, 2018). In addition, TNFRSF1A and TNFRSF1B are receptors for LTA₃ and are involved in the canonical NF-κB signaling pathway and mediation of apoptosis (Ware, 2005).

Role of LTB in Intestinal Inflammation

While no pre-clinical or clinical study has evaluated for associations between LTB and gastrointestinal symptoms associated with chemotherapy administration, LTA, LTB, and LTBR appear to be involved in the mechanisms that underlie inflammatory bowel disease (Gubernatorova & Tumanov, 2016). For example, in a study that investigated the mechanisms by which activation of LTBR signaling influences acute inflammation in a mouse model of acute colitis induced by dextrose sulfate sodium (DSS; Jungbeck et al., 2008), comparisons of inflammatory responses in colonic tissue were performed using three models of LTBR signaling ablation (i.e., antibody binding to the receptor, LTBR-deficient mice, and LTAB-deficient mice). All three ablation models resulted in aggravation of the colitis and release of inflammatory cytokines. In addition, all of the mice lost weight. In a second study of chronic DSS-induced colitis (Stopfer et al., 2004), the expression of Ltb in colonic tissue was increased. Interestingly, inhibition of LTBR resulted in decreases in the development of inflammation as well as in the production of TNF, IL1B, and IL6 in colonic tissue. Taken together, these pre-clinical findings provide evidence of an association between LTB and inflammatory states in the bowel. In addition, these results suggest that the effects of LTB signaling may be different in the settings of acute versus chronic inflammation. Additional research is needed to determine the role of LTB signaling in the development and manifestations of chemotherapy-induced inflammatory responses in the gastrointestinal mucosa.

Two clinical studies have evaluated for differences in LTB expression in patients with and without inflammatory bowel disease (Agyekum et al., 2003; Pisani et al., 2022). In the first study (Agyekum et al., 2003), compared to healthy controls, the expression of LTB on lymphocytes and plasma cells was increased in the mucosa of the colonic tissue of patients with ulcerative colitis and in the ileum of patients with Crohn’s disease. In the second study (Pisani et al., 2022), compared to healthy controls, LTB, C-C motif chemokine ligand (CCL)19, and CCL21 were differentially expressed and upregulated in the colonic tissue of patients with microscopic colitis. These findings were confirmed in intestinal tissue from an independent sample of patients with microscopic colitis. These findings support the role of LTB signaling in gastrointestinal inflammation.

Nausea and a Gastrointestinal Cluster

While additional research is needed to evaluate the role of LTB in a gastrointestinal cluster in oncology patients receiving chemotherapy, our findings build on previous studies of patients who overlapped with the samples used in this analysis (Papachristou et al., 2019; Singh et al., 2021; Singh, Kober, et al., 2020; Singh, Dhruva, et al., 2020). Two studies evaluated for differentially perturbed pathways between oncology patients with and without chemotherapy-induced nausea (Singh et al., 2021; Singh, Dhruva, et al., 2020). In addition to the NF-κB signaling pathway (Singh, Dhruva, et al., 2020), two additional pathways that are implicated in gastrointestinal inflammation (Cinausero et al., 2017) were differentially perturbed between patients with and without nausea (i.e., apoptosis (Singh et al., 2021), cytokine-cytokine signaling (Singh, Dhruva, et al., 2020)).

While these previous studies focused on a single symptom (Singh et al., 2021; Singh, Dhruva, et al., 2020), nausea is associated with the co-occurrence of several other gastrointestinal symptoms. For example, in another study by our team (Singh, Kober, et al., 2020), compared to patients without nausea, patients with nausea were more likely to report the occurrence of 11 additional gastrointestinal symptoms. Of note, nine of these symptoms (i.e., abdominal cramps, change in the way food tastes, constipation, diarrhea, difficulty swallowing, dry mouth, lack of appetite, weight loss, vomiting) were identified in our gastrointestinal cluster. While these findings are not surprising given that patients overlapped across the different analyses, similar findings were reported elsewhere. For example, in a study that evaluated the severity of 22 symptoms in patients with ovarian cancer receiving chemotherapy (Donovan et al., 2016), nausea was associated with five symptoms that were identified in our gastrointestinal cluster (i.e., bowel disturbances, dizziness, lack of appetite, vomiting, weight loss).

In another study by our team that identified a gastrointestinal cluster using network analysis (Papachristou et al., 2019), nausea was identified as the most important symptom within the network (i.e., theoretically has the greatest impact on other symptoms). We suggested that alleviating nausea may reduce the occurrence of the other symptoms within the network. Taken together, these findings suggest that nausea may be a sentinel symptom that drives the occurrence of other gastrointestinal symptoms.

Limitations and Future Directions

While these data provide preliminary evidence to support the hypothesis that the trans eCpG locus cg03171795 is involved in regulatory processes, it is not entirely clear how hypermethylation of this trans eCpG locus regulates the expression of LTB. An association between the methylation state of cg03171795 and expression of LTB was identified by Kennedy et al. (2018) in their genome-wide association study that tested for associations between CpG methylation and gene expression using data from the Multi-Ethnic Study of Atherosclerosis. Notably, trans eCpGs were identified more often than was expected and were more common than cis or distal eCpGs. The authors hypothesized that methylation regulates gene expression largely through secondary regulatory mechanisms, such as enhancer CpGs rather than promoter CpGs (Kennedy et al., 2018). Given that the regulatory role of trans eCpGs is an understudied area of epigenomic research, ongoing research is warranted. In vitro analyses may shed light on the indirect regulatory role(s) that hypermethylation of cg03171795 has on the expression of LTB (Rakyan et al., 2011). In addition, future research is warranted to evaluate the functional effect of LTB and other genes in the NF-κB signaling pathway in patients receiving chemotherapy. A multi-staged data-integrated multi-omics analysis (Harris et al., 2021) that evaluates for associations between a gastrointestinal cluster and differentially expressed genes (i.e., RNA, proteins) can shed light on the functional effect of these genes.

While an association between the trans eCpG locus cg03171795 and LTB was identified in the EPIC sample, this association was not found in our 450K sample. This lack of validation may be due to heterogeneity between the samples (e.g., gender, cancer type); the relatively small sample size (n = 146); and/or too small an effect size to identify a relationship. Additional research is needed to validate this association in patients receiving chemotherapy. In addition, given the study’s cross-sectional design, evaluation of changes in levels of methylation for cg03171795 and LTB expression throughout chemotherapy is warranted. Given the statistical challenges with the distribution of the gastrointestinal cluster factor scores, the best methods to incorporate the use of symptom cluster factor scores in methylation analyses need to be determined.