Integrative Genomic, Transcriptomic, and Microbiome Profiles of Colon Cancer by Ancestry Provide Insights into Molecular Distinctions

This study comprehensively characterizes genomic, transcriptomic, and microbiome differences between African and European ancestry colon cancer patients in New York, revealing ancestry-specific molecular distinctions such as enriched KRAS variants and unique microbial patterns that underscore the critical need for diverse representation in cancer research to improve risk assessment and treatment strategies.

The Gist

The Big Picture: Why This Study Matters

Imagine Colon Cancer as a complex, chaotic construction site. For a long time, scientists have been trying to understand how this site gets built and how to stop it. However, most of the blueprints (genetic data) they've studied so far were drawn for a very specific type of neighborhood: European ancestry.

This study is like sending a team of investigators to a different, often overlooked neighborhood: African ancestry. They wanted to see if the construction site looks different there, if the workers (genes) are different, or if the environment (bacteria) is unique. They found that yes, it is different, and ignoring these differences means we might be missing the keys to better treatments for everyone.

---

1. The "Blueprint" Check: Genetics and Ancestry

The Analogy: Think of your DNA as a massive instruction manual for building a human. Sometimes, typos (mutations) happen in the manual that cause the building to collapse (cancer).

• What they did: The researchers looked at the instruction manuals of patients from New York City. They didn't just ask, "What race do you identify as?" (which can be like asking someone to guess their family tree based on a photo). Instead, they read the actual text of the DNA to see exactly where their ancestors came from.
• The Discovery: They found that in patients with African ancestry, a specific "typo" in the manual called KRAS happens much more often than in European patients.
  • Why this matters: Currently, there are new "magic bullets" (drugs) designed to fix the KRAS typo, but they only work on one specific version of the typo (G12C). The African ancestry patients often have different versions of the KRAS typo (like G12D or G12V) that these drugs can't fix. It's like having a lock that needs a specific key, but the current keys don't fit. We need to make new keys for these specific locks.

2. The "Construction Crew": Molecular Subtypes

The Analogy: Imagine the cancer isn't just one big blob; it's actually four different types of construction crews, each with a different style of building and different tools. Scientists call these Consensus Molecular Subtypes (CMS).

• The Discovery: The researchers found that the "African ancestry" neighborhood had a much higher number of the most aggressive and stubborn construction crews (Subtypes 2 and 4). These crews are harder to stop and often ignore standard treatments.
• The Takeaway: If doctors treat everyone the same way, they might be using a "one-size-fits-all" hammer on a crew that actually needs a sledgehammer (or a different tool entirely). Knowing the specific crew type helps pick the right weapon.

3. The "Garden": The Tumor Microbiome

The Analogy: Inside the colon, there is a garden of bacteria. Some are helpful gardeners, and some are weeds that help the cancer grow.

• The Discovery: The researchers found that in these tumors, there was a distinct "invasion" of bacteria that usually live in the mouth. Specifically, a bacteria called Fusobacterium nucleatum was found in the tumors.
• The "Oral Connection": It turns out that when this mouth bacteria shows up in the colon tumor, it brings its "friends" with it—a whole community of other oral bacteria. It's like a gang of mouth-bacteria moving into the colon and setting up a base camp.
• Why it matters: This gang seems to be more organized and aggressive in certain tumors. Interestingly, people with African ancestry have higher rates of gum disease (periodontitis), which is a breeding ground for these mouth bacteria. This suggests a link: bad oral health might be sending "reinforcements" to the colon cancer, making it worse.

4. The "Copy Number" Chaos

The Analogy: Imagine your instruction manual has pages. Sometimes, cancer cells get greedy and photocopy certain pages (amplification) or throw away important pages (deletion).

• The Discovery: The African ancestry tumors had very specific patterns of "photocopying" and "throwing away" pages. They were missing pages related to tumor suppressors (the security guards that stop cancer) and had extra copies of pages that tell the cell to grow fast.
• The Takeaway: These specific patterns of missing or extra pages help explain why the cancer behaves differently in different people.

The Bottom Line: Why Should You Care?

This study is a wake-up call. For too long, medical research has been like studying a map of only one city and assuming it applies to the whole world.

1. Fairness: By ignoring African ancestry patients, we are missing crucial clues about how cancer works.
2. Better Drugs: We need to develop drugs that target the specific "typos" (mutations) found in African ancestry patients, not just the ones found in European patients.
3. Holistic Health: The link between mouth bacteria and colon cancer suggests that taking care of your teeth and gums might actually be a way to prevent or fight colon cancer, especially for those at higher risk.

In short: To build a better future for everyone, we need to look at the whole picture, including the parts of the population that have been left in the shadows. This study shines a light on those shadows, revealing unique differences that could save lives.

Technical Summary

1. Problem Statement

Colorectal cancer (CRC) exhibits significant disparities in incidence, severity, and mortality, with African American (AA) populations experiencing substantially higher rates and worse outcomes compared to European American (EA) populations. However, genomic and microbiome studies have historically underrepresented individuals of African ancestry (AFR). Existing studies often rely on:

• Self-reported race rather than genetic ancestry, which fails to capture the complex genetic heterogeneity within racial groups.
• Targeted gene panels rather than whole-genome (WGS) or whole-exome sequencing (WES), potentially missing low-frequency drivers or population-specific variants.
• Lack of multi-omic integration, specifically combining somatic mutations, copy number variations (CNVs), transcriptomics, and the tumor microbiome in an ancestry-stratified manner.

This study aims to comprehensively characterize the molecular landscape of colon cancer in AFR and European (EUR) ancestry populations to identify ancestry-specific drivers, mutational signatures, and microbiome compositions that may explain health disparities.

2. Methodology

The study utilized a multi-cohort, multi-omic approach involving two primary datasets:

• Cohorts:
  • P-1000 Consortium: Fresh-frozen colon tumors and matched adjacent non-tumor tissues from 52 patients (42 AFR, 10 EUR) undergoing Whole Genome Sequencing (WGS) and RNA-sequencing (RNA-seq).
  • CSHL/FFPE Cohort: Retrospective FFPE-derived tumor-only samples from 103 patients (52 AFR, 51 EUR) undergoing Whole Exome Sequencing (WES).
  • Validation: Data was compared against public datasets (TCGA-COAD, DFCI) to increase statistical power.
• Ancestry Determination: Genetic ancestry was inferred using ADMIXTURE and RAIDS software based on 1000 Genomes reference populations. Patients were assigned to AFRg (≥70% African ancestry) or EURg (≥70% European ancestry) groups, moving beyond self-reported race.
• Multi-Omic Analyses:
  • Genomics: Somatic and germline variant calling (SNVs, indels), Copy Number Variation (CNV) profiling (ASCAT), and mutational signature analysis (deconstructSigs).
  • Transcriptomics: Bulk RNA-seq analysis for Consensus Molecular Subtyping (CMS), differential gene expression, pathway activation scoring (PROGENy), and immune cell deconvolution (CIBERSORT).
  • Microbiome: 16S rRNA sequencing on matched tumor/non-tumor tissues and metagenomic analysis via WGS (Kraken2/Bracken) to identify bacterial taxa and co-occurrence networks.
  • Pre-neoplastic Analysis: Exome sequencing of 77 preneoplastic polyps (tubular adenomas and serrated polyps) to track mutation progression.

3. Key Contributions

• Ancestry-Stratified Genomics: Provided one of the first comprehensive WGS/WES analyses of CRC specifically stratified by genetic ancestry (AFR vs. EUR) rather than self-reported race.
• Discovery of Ancestry-Specific Drivers: Identified potential enrichment of functional KRAS variants and specific non-coding regulatory variants in AFR tumors that are underrepresented in European-centric databases.
• Microbiome Organization: Demonstrated that Fusobacterium nucleatum-positive tumors harbor organized oral microbial consortia, suggesting a translocation mechanism from the oral cavity to the colon.
• Methodological Rigor: Integrated WGS, WES, RNA-seq, and microbiome data to correlate somatic mutations with transcriptomic subtypes and microbial profiles, highlighting the limitations of panel-based sequencing for diverse populations.

4. Key Results

A. Somatic Mutations and Genomic Architecture

• KRAS Enrichment: KRAS mutations were significantly more frequent in AFRg tumors compared to EURg tumors (approx. 45–50% in AFR vs. 33–34% in EUR). This suggests a potential need for therapies targeting non-G12C KRAS variants (e.g., G12D, G12V) which are more common in AFR populations.
• Pathway Differences: AFRg MSS tumors showed a lower frequency of mutations in the HIPPO pathway compared to EUR cohorts.
• Novel Drivers: Several genes (e.g., SETD5, FLCN, RPL22) previously identified in large European studies were also mutated in AFR cohorts. Additionally, genes like ZBTB7A and BCL9L showed potential enrichment in AFR.
• Germline Variants: AHNAK2 showed increased mutation frequency in AFRg cohorts. AHNAK2 overexpression is linked to poor survival and metastasis.
• Mutational Signatures: Both groups showed signatures associated with aging (SBS1, SBS5) and mismatch repair deficiency (MSI-H). AFR samples showed higher prevalence of SBS18 (oxidative damage) and DBS20 (aristolochic acid exposure), though the latter did not correlate with self-reported tobacco use, suggesting environmental toxin exposure.
• Copy Number Variation (CNV): CNV profiles clustered primarily by CMS subtype rather than ancestry. However, specific recurrent deletions (e.g., chr18q involving SMAD2/4) and amplifications (e.g., chr8q, chr20q) were identified.

B. Transcriptomics and Molecular Subtyping

• CMS Distribution: In the P-1000 cohort, CMS2 and CMS4 (associated with aggressive phenotypes and immune exclusion) were predominantly found in AFR patients, whereas CMS1 was more common in EUR patients. This contrasts with TCGA data, suggesting cohort-specific dynamics or the impact of genetic ancestry on subtype distribution.
• Gene Expression: Three pseudogenes (PSPHP1, RP11-208G20.2, RP11-208G20.3) were significantly overexpressed in AFR tumors. PSPHP1 (a pseudogene of PSPH) showed the strongest association with African admixture.
• KRAS Signaling: GSEA revealed upregulation of "KRAS signaling" in AFR tumors. In CMS4 AFR tumors, KRAS mutation status correlated with higher pathway activation scores.
• Immune Infiltration: AFR tumors showed distinct immune cell infiltration patterns, including differences in memory B cells, plasma cells, and M1 macrophages compared to EUR tumors.

C. Microbiome Analysis

• Diversity: Tumor tissues had significantly lower microbial diversity than adjacent non-tumor tissues. Right-sided tumors showed reduced diversity compared to left-sided tumors.
• Oral Microbial Translocation: Fusobacterium nucleatum (Fn) was exclusively detected in tumor tissues. Fn-positive tumors showed a 4-fold increase in oral-origin taxa (e.g., Porphyromonas, Treponema, Prevotella).
• Co-occurrence Networks: Co-abundance network analysis revealed distinct clusters of oral bacteria (including Fusobacterium, Porphyromonas gingivalis, Prevotella spp.) that co-assemble in tumors, suggesting organized biofilm-like structures similar to periodontal plaque.
• Ancestry-Specific Taxa: 83 taxa were significantly enriched in AFR tumors, including multiple oral-associated species, whereas no taxa were uniquely enriched in EUR tumors.

D. Pre-neoplastic Progression

• Analysis of polyps vs. tumors confirmed APC as an early "gatekeeper" mutation.
• KRAS, TP53, and PIK3CA mutations were enriched in tumors compared to polyps, indicating later drivers.
• Serrated polyps in EUR patients frequently harbored BRAF mutations as early events, whereas AFR polyps showed different patterns.

5. Significance and Implications

• Addressing Health Disparities: The study provides critical evidence that CRC biology differs by genetic ancestry, challenging the "one-size-fits-all" approach in genomic medicine. The enrichment of KRAS variants and specific CNVs in AFR patients suggests that current clinical panels and targeted therapies (often focused on G12C) may miss actionable targets in this population.
• Microbiome and Etiology: The finding of organized oral microbial consortia in tumors, particularly in AFR populations with higher rates of periodontitis, supports the hypothesis that oral dysbiosis contributes to CRC pathogenesis via translocation.
• Need for Diverse Cohorts: The results underscore the urgent need to expand AFR representation in genomic databases. Relying on European-centric data leads to an incomplete understanding of CRC drivers and hinders the development of equitable treatment strategies.
• Future Directions: The authors advocate for the use of Whole Genome Sequencing (WGS) over gene panels to capture structural variants and non-coding drivers, and suggest that long-read sequencing could further elucidate complex genomic regions relevant to racial disparities.

In conclusion, this integrative study reveals distinct molecular, transcriptional, and microbial landscapes in African ancestry colon cancer patients, providing a foundation for developing ancestry-specific diagnostic and therapeutic interventions.