From polymorphism to phenotype: FADS2 rs174538, fatty acid signatures and disease activity in Japanese Crohn’s disease

Abstract

Crohn’s disease (CD) arises from an interplay between genetic susceptibility, immune dysregulation, and environmental factors, including diet and lipid metabolism. Polyunsaturated fatty acids (PUFAs) and their desaturase enzymes have attracted interest because they generate inflammatory lipid mediators. Matsuzawa et al recently published a study in World Journal of Gastroenterology, examining how the FADS2 polymorphism rs174538 shapes fatty acid profiles and CD activity in a homogeneous cohort of Japanese patients. Using serum and erythrocyte membrane measures and refined δ-6 and δ-5 desaturase indices, the authors demonstrate that rs174538 is associated with lower erythrocyte delta-6 desaturase index (rd.d6di) and, strikingly, with decreased d5di and lower arachidonic acid levels, an effect that Matsuzawa et al largely attribute to tight linkage with FADS1 variants. Although rs174538 itself did not stratify CD activity, rd.d6di correlated positively with the CD Activity Index (CDAI) in rs174538 wild-type patients, suggesting that erythrocyte-based indices of desaturase activity may serve as integrative biomarkers of inflammatory burden and possibly treatment response. Limitations include modest sample size, predominantly mild disease, and no prospective dietary assessment, which temper inference and generalizability. Nonetheless, the work provides a genotype-stratified link between membrane PUFA metabolism and C-reactive protein or CDAI in CD, strengthening the case for metabolically informed biomarkers.

Key Words: Crohn’s disease; FADS2; rs174538; Fatty acids; Desaturase index; Arachidonic acid; Observational study

Core Tip: FADS variants shape polyunsaturated fatty acids (PUFAs) fatty-acid profiles, but it is uncertain whether this matters in Crohn’s disease (CD). Matsuzawa et al bring this question into a CD cohort by linking rs174538 to stable erythrocyte membrane fatty-acid patterns and desaturase indices. Notably, the relationship between an erythrocyte delta-6 desaturase index and CD Activity Index was present only in rs174538 wild-type carriers and disappeared in mutant carriers. This highlights that the same metabolic marker can carry different clinical information depending on genetic background, offering a clearer framework for interpreting lipid biomarkers and inconsistent omega-3 (n-3) PUFAs trial results in inflammatory bowel disease.

This editorial refers to “Association of FADS2 polymorphism rs174538 with fatty acid metabolism and disease severity in Japanese patients with Crohn's disease” by Matsuzawa et al, 2026; https://doi.org/10.3748/wjg.v32.i2.112132.

---

INTRODUCTION

In this issue of World Journal of Gastroenterology, Matsuzawa et al[1] provide new insight into how the FADS2 polymorphism rs174538 influences fatty acid metabolism and clinical expression of Crohn’s disease (CD) in a Japanese population. CD itself is a chronic, relapsing inflammatory bowel disease (IBD) characterized by transmural inflammation and heterogeneous behavior, with a rising global incidence and substantial impact on quality of life and health-care utilization[2]. In Japan, evidence-based clinical practice guidelines emphasize early risk stratification, treat-to-target strategies, and tight monitoring using clinical, endoscopic, and biomarker endpoints to improve outcomes[3]. At the same time, broader work on IBD pathogenesis has highlighted the complex interplay among host genetics, mucosal immune responses, microbiota, and environmental factors[4].

Among environmental exposures, diet, particularly dietary lipids, has emerged as an important modulator of intestinal inflammation. Prospective human studies have linked higher intake of linoleic acid and arachidonic acid, two major components of the n-6 pathway, with an increased risk of incident ulcerative colitis[5]. Experimental work further showed that a linoleic acid-rich soybean oil diet increased susceptibility to colitis through epithelial barrier dysfunction, gut dysbiosis, and altered bioactive lipid signaling[6]. Omega-3 (n-3) polyunsaturated fatty acids (PUFAs)-enriched diets can alter erythrocyte membrane fatty acid composition and may benefit some IBD cohorts, although reviews and meta-analyses continue to show heterogeneous clinical effects in both ulcerative colitis and CD[7-9]. These observations naturally raise the question of how endogenous PUFA metabolism, shaped by host genetics, may interact with dietary intake to influence disease course. The FADS gene cluster on chromosome 11 encodes pivotal desaturase enzymes that convert linoleic acid and α-linolenic acid into longer-chain derivatives such as arachidonic acid (AA), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). In the general population, specific FADS genotypes are associated with higher AA/Linoleic acid ratios, increased estimated desaturase activity, and greater systemic inflammation and coronary artery disease risk[10]. In Koreans, FADS polymorphisms have been linked to differences in omega-6 (n-6) PUFA composition, lipid peroxides, and coronary artery disease, highlighting ethnic differences in linkage disequilibrium (LD) and phenotypic impact[11]. A genome-wide association study from the InCHIANTI cohort further identified variants near FADS1–FADS2 as the major determinants of plasma PUFA concentrations[12]. In addition, the relationship between FADS-related genetic variation and inflammatory bowel disease has been widely explored. Table 1 summarizes representative studies showing how FADS-related variants have been linked to inflammatory bowel disease susceptibility, gene-diet interaction, and related biology[13-17].

Table 1 Representative studies of FADS-related genetic variants and loci in inflammatory bowel disease.

FADS-related variant /Locus	Disease context	Population /design	FADS-related finding	Ref.
FADS2 rs2277284	CD	Japanese case-control genetic study, 52 patients compared with Japanese reference data	The allele frequency of FADS2 rs2277284 was significantly higher in patients with CD than in controls, with an OR of 4.4	[13]
FADS1 rs174561, rs174556; FADS2 rs3834458, rs174575	CD and UC	Nested case-control study within NHS and NHS II, including 101 CD cases, 139 UC cases, and 495 controls	No clear independent association between FADS polymorphisms and CD or UC risk was observed. The diet-related association was more strongly modified by other lipid metabolism genes than by FADS	[14]
FADS1 rs174547, rs174575; FADS2 rs11230815, rs17831757, rs968567, rs174627	CD	Pediatric case-control study, 182 CD cases and 250 controls	FADS1 rs174547 and rs174575 showed no significant association with CD risk. For FADS2, rs11230815, rs17831757, rs968567, and rs174627 each exhibited an interaction with a high ω6/ω3 dietary ratio; only carriers of these specific genotypes had a significantly increased CD risk when consuming a high ratio. Notably, rs17831757 itself was not independently associated with CD	[15]
rs174537 (near FADS1); FADS2 rs174538; chr11 FADS1-FADS2-TMEM258 locus	IBD	Human eQTL fine-mapping and mechanistic study in cells, mouse DSS colitis, and colonic organoids	rs968567 is associated with rheumatoid arthritis and lipid metabolism, while IBD risk is associated with rs174537 in the same region (which is associated with increased TMEM258 expression)	[16]
rs174537 (near FADS1)	CD and UC	Large GWAS meta-analysis and validation study	In intestinal tissue eQTL analysis, the expression of the known risk site rs174537 for IBD was significantly correlated with that of FADS2	[17]

IBD: Inflammatory bowel disease; CD: Crohn’s disease; UC: Ulcerative colitis; NHS: Nurses’ Health Study; GWAS: Genome-wide association study; OR: Odds ratio.

Japan also provides a distinctive nutritional backdrop for interpreting FADS2 rs174538. Traditional dietary patterns have been relatively rich in fish and seafood, a major source of marine n-3 PUFAs[18]. However, after 2007, household fresh meat consumption surpassed fresh fish consumption in Japan[19]. In this context, a higher dietary n-6 to n-3 balance may plausibly amplify genotype-dependent differences in endogenous PUFA remodeling, consistent with an evolutionary mismatch concept in which rapid dietary change can expose previously buffered genetic vulnerabilities[20].

Within CD, Japanese work has reported relatively frequent FADS2 and ELOVL6 mutations, suggesting that genes involved in fatty acid elongation and desaturation may contribute to disease susceptibility or heterogeneity in this population[13]. Complementary studies have demonstrated that selected fatty acids can serve as useful serological markers in CD, reinforcing the clinical informativeness of detailed fatty acid profiling[21]. More broadly, a recent overview of fatty acids in CD pathophysiology emphasized that both substrate availability and enzymatic capacity may shape intestinal inflammation, barrier integrity, and response to therapy[22]. Together, these data provide a strong rationale for the present study, which moves beyond general lipidomics to dissect how a specific FADS2 polymorphism influences enzymatic indices, fatty acid signatures, and CD activity in a homogeneous Japanese cohort. More broadly, this work frames erythrocyte desaturase indices and PUFA signatures as intermediate metabolic phenotypes-a practical bridge that helps connect genetic variation to clinically relevant inflammation.

FROM POLYMORPHISM TO PHENOTYPE: DESIGN AND KEY FINDINGS

Genetic epidemiology has begun to clarify potential causal links between circulating fatty acid levels and IBD risk. A two-sample Mendelian randomization study suggested that genetically determined variations in specific fatty acids may influence susceptibility to IBD, including CD[23]. In parallel, a detailed review of fatty acids and their lipid mediators has integrated these findings into a mechanistic framework spanning innate and adaptive immunity, epithelial biology, and microvascular responses[24]. Against this backdrop, Matsuzawa et al[1] focus on patients with established CD rather than the general population, quantifying PUFA composition in serum and erythrocytes and redefining the δ-6 desaturase index (rd.d6di = γ-linolenic acid/Linoleic acid) and δ-5 desaturase index (d5di = AA/dihomo-γ-linolenic acid) to more cleanly separate the two enzymatic steps.

Their core observation is that rs174538 mutant-allele carriage is associated with a modest reduction in rd.d6di in erythrocyte membranes, but a more pronounced decrease in d5di and AA in both serum and erythrocyte fractions, consistent with a linked haplotype spanning FADS2 and FADS1. Clinically, rs174538 genotype alone does not stratify standard markers such as CD Activity Index (CDAI). However, among rs174538 wild-type carriers, erythrocyte rd.d6di correlates positively with CDAI, whereas this relationship is lost in mutant carriers. This genotype-dependent coupling between desaturase activity and inflammatory burden illustrates how a single polymorphism embedded within a desaturase haplotype can reshape the link between metabolism and phenotype in CD. The main findings of Matsuzawa et al[1] are summarized schematically in Figure 1.

Figure 1 Schematic summary of the main findings reported by Matsuzawa et al. This schematic illustrates how the FADS2 rs174538 variant may influence n-6 polyunsaturated fatty acid (PUFA) metabolism in Japanese patients with Crohn’s disease. In the n-6 PUFA metabolic pathway, linoleic acid (LA) is converted to gamma-linolenic acid (GLA) through the catalytic activity of FADS2, whereas dihomo-gamma-linolenic acid (DGLA) is converted to arachidonic acid (AA) through the catalytic activity of FADS1. The redefined delta-6 desaturase index (rd.d6di) is calculated as GLA/LA, and the delta-5 desaturase index (d5di) is calculated as AA/DGLA. Compared with the wild-type genotype, carriers of the rs174538 mutant allele (GA/AA) showed a reduction in rd.d6di in the erythrocyte membrane (Rrd.d6di), together with lower d5di in the erythrocyte membrane and serum (Rd5di and Sd5di) and reduced AA levels. These findings suggest that the metabolic effect of rs174538 may extend beyond FADS2 itself and may partly reflect linkage disequilibrium with variants in the adjacent FADS1 locus on chromosome 11q12.2. In addition, a positive correlation between rd.d6di and the Crohn’s Disease Activity Index was observed in patients with the wild-type genotype, whereas this association was not detected in mutant allele carriers. The chromosomal positions of FADS1 and FADS2 shown in the figure are included to illustrate their close physical proximity within the FADS gene cluster. PUFA: Polyunsaturated fatty acid; CDAI: Crohn’s Disease Activity Index; LA: Linoleic acid; GLA: Gamma-linolenic acid; DGLA: Dihomo-gamma-linolenic acid. This figure was created by BioRender.com (Supplementary material).

INTERPRETING FATTY ACID SIGNATURES IN THE CONTEXT OF THE FADS GENE CLUSTER

Beyond single polymorphisms, several complementary lines of evidence support the notion that CD is associated with characteristic remodeling of circulating and tissue fatty acid pools. A conceptual review of fatty acids in CD pathophysiology has emphasized that alterations in PUFA metabolism can influence epithelial barrier function, immune-cell polarization, and production of bioactive eicosanoids and specialized pro-resolving mediators. Observational data indicate that fatty acid profiles differ between IBD patients and healthy controls, and that these patterns may relate to disease duration and activity[25]. Although Matsuzawa et al[1] primarily interrogated the n-6 axis, the translational relevance of these findings extends to n-3 interventions because both n-6 and n-3 PUFAs draw on shared desaturase and downstream enzymatic machinery. Interventional studies using highly purified EPA free fatty acid have demonstrated reductions in fecal calprotectin and a lower relapse rate in ulcerative colitis[26], whereas a Cochrane analysis of n-3 PUFAs for maintenance of remission in CD concluded that benefits, if present, are modest and inconsistent across trials[9]. Reviews of diet in IBD have suggested that heterogeneity in study design and inter-individual differences shaped by background diet and host genetics may contribute to the conflicting evidence on n-3 PUFAs supplementation[27,28].

These observations resonate with the genetic insights surrounding FADS variants. Matsuzawa et al[1] extended this concept to CD by showing that rs174538 shapes a genotype-dependent link between an intermediate metabolic phenotype and disease activity: Rd.d6di correlates with CDAI only in wild-type carriers, not in mutant carriers. This pattern provides a testable genetic hypothesis for why n-3 PUFAs supplementation trials in IBD have yielded conflicting results: If baseline desaturase flux and membrane PUFA remodeling capacity differ by FADS background, then the biological impact of n-3 PUFAs interventions may vary in magnitude across genotype strata and dietary n-6/n-3 contexts. Consequently, genotype-unselected trials may dilute subgroup-specific benefits, contributing to null or inconsistent average effects. If FADS haplotypes materially influence the capacity to generate AA and downstream mediators, then patients with different FADS backgrounds may occupy distinct “metabolic niches” despite apparently similar diets. The loss of correlation in mutation carriers may not indicate that inflammation does not affect lipid pathways, but rather that the dynamic range or sensitivity of rd.d6di is reduced in this genetic context, rendering it no longer a valid proxy for changes in CDAI. This mechanism requires validation using lipid mediator data at the tissue or cellular level, for example through stable isotope tracer studies or stimulated immune-cell lipid mediator production assays. Additional Mendelian randomization work has suggested a potential causal relationship between specific fatty acid traits and IBD risk[29], lending further weight to the idea that PUFA metabolism lies upstream of, rather than merely downstream from, intestinal inflammation. Complementing this, targeted lipidomic studies have shown that calprotectin, a widely used non-invasive biomarker of intestinal inflammation, is associated with oxidized linoleic and AA derivatives (HODE and HETE), underscoring the tight coupling between fatty acid-derived lipid mediators and mucosal inflammatory activity[30]. Together, these genetic and mechanistic insights provide a rationale for translating PUFA signatures from descriptive biomarkers into clinically testable tools, motivating the clinical and research implications discussed next.

CLINICAL AND RESEARCH IMPLICATIONS: TOWARD GENOTYPE-INFORMED FATTY ACID PHENOTYPING

From a clinical perspective, the most immediate implication of the Matsuzawa et al[1] study is the potential for genotype-informed metabolic biomarkers. Importantly, the novelty here is not simply that FADS genotypes influence PUFA profiles-a well-established concept in nutritional genetics-but that, within an established CD cohort, rs174538 defines a genotype-dependent relationship between an intermediate metabolic phenotype and inflammatory burden. Specifically, rs174538 genotype alone does not stratify CDAI, yet erythrocyte rd.d6di tracks CDAI only in wild-type carriers, and this coupling is lost in mutant carriers, suggesting that the same metabolic readout may be clinically informative in one genetic background but uninformative in another. Erythrocyte-based desaturase indices and fatty acid signatures may complement clinical indices, endoscopy, and fecal calprotectin, particularly when repeated invasive assessment is impractical or when conventional biomarkers provide discordant signals[3]. In this context, erythrocyte membrane fatty acids may offer a more stable readout of PUFA metabolism than serum or plasma fractions, because they are less sensitive to recent dietary intake, have slower turnover, and better reflect longer-term fatty acid exposure[31,32].

At present, the data do not support routine FADS genotyping or desaturase profiling in everyday practice. However, they strongly justify incorporating such measurements into prospective cohorts and interventional trials. A key translational implication is interpretability: Genotype-stratified analyses may reveal metabolism-activity relationships that are diluted or missed entirely in pooled cohorts. Genotype-stratified nutritional studies could test whether patients with wild-type FADS haplotypes and higher δ-6 activity are particularly susceptible to linoleic acid–rich diets or derive greater benefit from n-3 PUFAs-enriched regimens, whereas patients carrying rs174538-linked haplotypes might respond differently. For genotypes consistent with reduced delta 6 desaturation, precursor-only strategies based on alpha-linolenic acid may produce only modest increases in EPA and little change in DHA, making direct EPA and DHA supplementation a more direct interventional strategy to test[33]. Integration of lipidomic and genetic data could also help explain why some trials of fish oil or n-3 PUFAs supplementation in IBD have shown benefit while others have not.

LIMITATIONS AND FUTURE DIRECTIONS

As with any single-center observational study, the work by Matsuzawa et al[1] has limitations. The sample size is modest, limiting power to detect small genotype-phenotype effects or to explore interactions with specific therapies or disease behaviors. The cross-sectional design provides only a snapshot of disease activity and fatty acid profiles, precluding firm conclusions about temporal dynamics-whether shifts in desaturase indices precede flares, accompany them, or reflect downstream consequences of inflammation. Detailed information on dietary intake and background nutritional status was not collected prospectively, leaving potential confounding by total fat intake, n-6/n-3 ratios, and specific food patterns. Importantly, the interpretation that rs174538 acts mainly through FADS1 linkage remains tentative. Because rs174538 tags a highly correlated regulatory region, disentangling its contribution from linked functional variants across TMEM258–FADS1–FADS2 will require larger cohorts with fine-mapping and conditional haplotype-based analyses. Moreover, the desaturase indices and erythrocyte PUFA percentages are intermediate metabolic phenotypes-informative proxies of pathway flux, but not direct readouts of mucosal effector biology. Without tissue- or cell-level lipid mediator profiling, it remains uncertain how rs174538-tagged haplotypes translate into the balance of cyclooxygenase/Lipoxygenase-derived eicosanoids and pro-resolving mediators within the inflamed gut. This question is now technically approachable, as LC-MS/MS-based studies have already characterized mucosal lipid mediator patterns from paired inflamed and non-inflamed colonic biopsies in ulcerative colitis and from fasting plasma together with colonic biopsies in Crohn’s disease, revealing disease-linked shifts in prostaglandins, HETEs, oxylipins, and endocannabinoids[34,35]. Future studies combining FADS genotyping with intestinal (or immune-cell) lipidomics under standardized dietary conditions would be particularly powerful.

Finally, LD and haplotype structure across the TMEM258–FADS1–FADS2 region are strongly ancestry-dependent, with distinct LD blocks and common haplotypes whose tagging properties may not be portable across populations[36]. For example, in Japanese reference data, rs174538 shows very high LD with rs174537 near FADS1 (R² = 0.97), whereas Korean data implicate rs174537 with lower AA and lower AA/dihomo-γ-linolenic acid-supporting multi-ethnic validation using locus-wide genotyping/fine-mapping rather than extrapolating from rs174538 alone. Nevertheless, despite these limitations, the study provides a valuable framework for future biomarker and therapeutic development in CD. By linking FADS genotype with intermediate metabolic phenotypes and clinical disease activity, it offers a useful basis for future validation and translation.

CONCLUSION

The study by Matsuzawa et al[1] advances our understanding of how the FADS gene cluster-and FADS2 rs174538 in particular-shapes fatty acid phenotypes and clinical expression of CD in a Japanese population. By linking a specific desaturase polymorphism to quantitative indices of δ-5 and δ-6 activity, erythrocyte PUFA signatures, and disease activity, it illustrates the power of intermediate metabolic phenotypes to illuminate the path from polymorphism to phenotype. For practicing gastroenterologists, these findings do not yet mandate changes in routine care, but they should heighten awareness that patients with CD differ not only in their immunologic and microbiome profiles but also in their intrinsic capacity to generate pro- and anti-inflammatory lipid mediators. For researchers, the study underscores the need to integrate FADS genotyping, detailed fatty acid phenotyping, and longitudinal clinical outcomes into future work. Ultimately, such efforts may enable truly genotype-informed strategies for dietary counseling and host-directed therapy in CD, bringing us closer to the goal implied in the title of this editorial: Moving from polymorphism to phenotype in CD.