High ACSL6 expression correlates with enhanced cytotoxicity of DHA and DHA-combined chemotherapy

To evaluate the cytotoxic effects of LCFAs and oxaliplatin (OXA) on colorectal cancer cells, we performed 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assays using Colo320, LoVo, and SW48 cell lines. The result shows that OXA exhibited potent anti-proliferative activity in all three cell lines (Fig. 1A). DHA showed stronger anti-proliferative effects (Fig. 1B) compared with EPA (Fig. 1C). Moreover, LoVo and Colo320 cells were more sensitive to DHA than SW48 cells (Fig. 1B). In combination treatments, DHA supplementation, but not EPA, enhanced OXA-induced cytotoxicity in Colo320 cells. Notably, LoVo and Colo320 cells exhibited a more pronounced decline in cell viability relative to SW48 cells (Fig. 1D–G). These findings suggests that DHA is more effective than EPA in suppressing cancer cell growth and more potently augments the cytotoxicity of oxaliplatin, particularly in Colo320 and LoVo cells.

Fig. 1: High ACSL6 expression correlates with enhanced cytotoxicity of DHA-combined chemotherapy.

A Cell viability of SW48, LoVo, and Colo320 colorectal cancer cell lines treated with increasing concentrations of oxaliplatin (OXA) for 3 days. Data are shown as mean ± SEM. B Cell viability of SW48, LoVo, and Colo320 colorectal cancer cell lines treated with increasing concentrations of DHA for 3 days. Data are shown as mean ± SEM. **p < 0.01, ****p < 0.0001 (n = 3). C Cell viability of SW48, LoVo, and Colo320 colorectal cancer cell lines treated with increasing concentrations of EPA for 3 days. Data are shown as mean ± SEM. D Colo320 cell viability after 3-day EPA and OXA combination treatment at varying doses. Data are presented as mean ± SEM (n = 3). E LoVo cell viability after 3-day DHA and OXA combination treatment at varying doses. Data are presented as mean ± SEM (n = 3). F Colo320 cell viability after treatment with DHA and OXA for 3 days at increasing concentrations. Data are shown as mean ± SEM (n = 3). G SW48 cell viability following 3-day combined treatment with DHA and OXA at different concentrations. Data are shown as mean ± SEM (n = 3). H mRNA expression levels of ACSL6 in SW48, LoVo, and Colo320 colorectal cancer cell lines. ACSL6 expression levels are normalized to SW48 with 18S rRNA as the reference gene. Data are presented as mean ± SEM. **p < 0.01, ***p < 0.001 (n = 3). I Dose-dependent effects of doxorubicin (Doxo) on MDA-MB-231 and MCF7 breast cancer cell viability over 3 days. Data are shown as mean ± SEM (n = 3). J Viability of MCF7 and MDA-MB-231 breast cancer cell lines after 3 days of DHA treatment at different concentrations. Data are presented as mean ± SEM. ***p < 0.001, ****p < 0.0001 (n = 3). K MDA-MD-231 cell viability after treatment with DHA and Doxo for 3 days at increasing concentrations. Data are shown as mean ± SEM (n = 3). L MCF7 cell viability after treatment with DHA and Doxo for 3 days at increasing concentrations. Data are shown as mean ± SEM (n = 3).

DHA uptake and its subsequent conversion to DHA-CoA are critical for its biological functions, including cytotoxicity. This process is facilitated by acyl-CoA synthase long-chain (ACSL) activity, particularly ACSL4 and ACSL6, which catalyze DHA-CoA formation and support its cellular incorporation. To investigate the potential association between these two ACSLs expression and DHA sensitivity, we conducted qPCR analyses. ACSL4 levels exhibited no significant differences among the tested colorectal cancer cell lines (Fig. S1A). In contrast, ACSL6 expression was substantially higher in LoVo and Colo320 cells compared to SW48 cells (Fig. 1H). These findings imply that a positive association between ACSL6 expression and DHA sensitivity.

Additionally, we further examined DHA sensitivity in two breast cancer cell lines, MCF7 and MDA-MB-231, both of which were sensitive to doxorubicin (Doxo) (Fig. 1I). Compared with MCF7 cells, MDA-MB-231 cells exhibited greater sensitive to DHA (Fig. 1J). Moreover, DHA enhanced Doxo-induced cytotoxicity in MDA-MB-231 cells (Fig. 1K) but not in MCF7 cells (Fig. 1L). qPCR analysis revealed that both ACSL4 and ACSL6 levels were higher in MDA-MB-231 cells than in MCF7 cells (Fig. S1B and S1C). These findings support that ACSL6 expression may be positively associated with DHA sensitivity and enhanced chemotherapy efficacy.

ACSL6 suppresses cell growth and positively associates with improved responsiveness to standard treatments

To explore the role of ACSL4 and ACSL6 in modulating the cytotoxicity of DHA-combined chemotherapy in cancer cells, we first evaluate the clinical relevance of their expression using the Kaplan–Meier platform (https://kmplot.com/analysis/). In colon adenocarcinoma, we found that higher ACSL4 mRNA expression was significantly associated with improved recurrence-free survival (P = 0.026, Fig. 2A), whereas no significant association was observed breast cancer (Fig. 2B). Additionally, patents with colon adenocarcinoma (Fig. 2C) or breast cancer (Fig. 2D) who had higher ACSL6 expression faced a significantly lower risk of recurrence following primary treatment compared with those with decreased ACLS6 levels (P = 0.0048 and P = 0.001, respectively). These suggest that ACSL6, rather than ACSL4, may serve as a predictor of treatment responsiveness in colorectal and breast cancers, with higher expression linked to improved patient outcomes. Based on these observations, we investigated the role of ACSL6 in regulating cell growth and modulating DHA sensitivity. Firstly, ACSL6 expression was reduced in LoVo cells using two independent ACSL6-specific shRNAs (#17 and #29), and western blot analysis confirmed decreased ACSL6 protein expression in these shRNA-expressing cells (Fig. 2E and Fig. S2A). MTS and long-term colony formation assays demonstrated that ACSL6 knockdown significantly enhanced LoVo cell proliferation (Fig. 2F) and colony-formation ability (Fig. 2G and Fig. S2B). Conversely, stable expression of V5-tagged ACSL6 via lentiviral transduction (Fig. 2H) reduced LoVo cell proliferation (Fig. 2I) and colony formation (Fig. 2J and Fig. S2C). Similar results were observed in MCF7 cells, where ACSL6 silencing enhanced proliferation and colony formation (Fig. 2K–M, and Figs. S1D–S2E), whereas ACSL6 overexpression suppressed cell growth (Fig. 2N–P and Fig. S2F), supporting a potential tumor-suppressive function of ACSL6.

Fig. 2: ACSL6 suppresses cell growth and positively associates with improved responsiveness to standard treatments.

A Kaplan–Meier of recurrence free survival (RFS) analysis in colorectal cancer patients stratified by ACSL4 expression. Patients with low ACSL4 expression (black line, n = 949) and high ACSL6 expression (red line, n = 387) are compared. HR = 0.76 (95% CI: 0.59–0.97), p = 0.026. B Kaplan–Meier of recurrence-free survival (RFS) analysis in breast cancer patients stratified by ACSL4 expression. Patients with low ACSL4 expression (black line, n = 2464) and high ACSL4 expression (red line, n = 2465) are compared. HR = 0.95 (95% CI: 0.91–1.11), p = 0.95. C Kaplan–Meier of RFS analysis in colorectal cancer patients stratified by ACSL6 expression. Patients with low ACSL6 expression (black line, n = 871) and high ACSL6 expression (red line, n = 296) are compared. HR = 0.66 (95% CI: 0.5–0.88), p = 0.0048. D Kaplan–Meier of recurrence-free survival (RFS) analysis in breast cancer patients stratified by ACSL6 expression. Patients with low ACSL6 expression (black line, n = 1036) and high ACSL6 expression (red line, n = 996) are compared. HR = 0.78 (95% CI: 0.67–0.9), p = 0.001. E Representative western blots of LoVo cells stably expressing either a scrambled shRNA or shRNAs specific for ACSL6. β-Actin served as a loading control. F Cell growth curves of LoVo cells stably expressing either a scrambled shRNA or ACSL6 shRNAs, measured by MTT assays. *p < 0.05, **p < 0.01 (n = 3). G Quantification of colony formation assays for ACSL6 knockdown experiments. *p < 0.05. Data are mean ± SEM. (n = 3). H Representative western blots of LoVo cells stably carrying either an empty vector or V5-tagged ACSL6. β-Actin served as a loading control. I Cell growth curves of LoVo cells stably expressing either an empty vector or V5-tagged ACSL6, measured by MTT assays. *p < 0.05 (n = 3). J Quantification of colony formation assays for ACSL6 overexpression experiments. **p < 0.01. Data are mean ± SEM. (n = 3). K Representative western blots of MCF7 cells stably expressing either a scrambled shRNA or shRNAs specific for ACSL6. β-Actin served as a loading control. L Cell growth curves of MCF7 cells stably expressing either a scrambled shRNA or ACSL6 shRNAs, measured by MTT assays. **p < 0.01, ***p < 0.001 (n = 3). M Quantification of colony formation assays for ACSL6 knockdown experiments. *p < 0.05. Data are mean ± SEM. (n = 3). N Representative western blots of MCF7 cells stably carrying either an empty vector or V5-tagged ACSL6. β-Actin served as a loading control. O Cell growth curves of MCF7 cells stably expressing either an empty vector or V5-tagged ACSL6, measured by MTT assays. *p < 0.05 (n = 3). P Quantification of colony formation assays for ACSL6 overexpression experiments. *p < 0.05. Data are mean ± SEM. (n = 3).

ACSL6 sensitizes colorectal and breast cancer cells to chemotherapy in combination with DHA

Next, we assessed whether ACSL6 sensitizes cancer cells to chemotherapy by enhancing DHA-mediated suppression of cell growth. LoVo cells expressing ACSL6 shRNAs were less sensitive to DHA treatment (Fig. 3A), and ACSL6 silencing partially reversed DHA-induced inhibition of colony formation (Fig. 3B and Fig. S3A), while exhibiting no significant effect on the response to OXA alone (Fig. 3C). When DHA was combined with OXA, ACSL6 knockdown markedly reduced the cytotoxicity of the treatments, resulting in better cell viability (Fig. 3D) and colony-forming ability (Fig. 3E and Fig. S3B) compared with control cells. In contrast, elevated ACSL6 expression increased sensitivity to DHA (Fig. 3F) and further suppressed colony growth upon DHA treatment (Fig. 3G and Fig. S3C), while having no effect on the response to OXA alone (Fig. 3H). Notably, ACSL6 overexpression led to a pronounced reduction in cell viability and colony formation following combined OXA and DHA treatment compared with empty vector controls (Fig. 3I, J, and Fig. S3D).

Fig. 3: ACSL6 sensitizes colorectal and breast cancer cells to chemotherapy in combination with DHA.

A Viability of LoVo cells stably expressing either a scrambled shRNA or ACSL6 shRNA following treatment with DHA. Data are mean ± SEM. *p < 0.05, **p < 0.01 (n = 3). B Quantification of colony formation assay for ACSL6 knockdown experiments in LoVo. Data are mean ± SEM. **p < 0.01 (n = 3). C Viability of LoVo cells stably expressing either a scrambled shRNA or ACSL6 shRNA following treatment with OXA. Data are mean ± SEM. (n = 3). D Cell viability assay of LoVo cells stably expressing either a scrambled or ACSL6 shRNAs following treatment with increasing concentrations of OXA in the presence of DHA. Data are mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 (n = 3). E Quantification of colony formation assay for LoVo cells stably expressing either a scrambled shRNA or ACSL6 shRNA following treatment with OXA and DHA. Data are mean ± SEM. *p < 0.05, ***p < 0.001 (n = 3). F Viability of LoVo cells stably carrying either an empty vector or V5-tagged ACSL6 following treatment with DHA. Data are mean ± SEM. *p < 0.05, **p < 0.01 (n = 3). G Quantification of colony formation assay for ACSL6 overexpression experiments in LoVo cells. Data are mean ± SEM. *p < 0.05, ***p < 0.001 (n = 3). H Viability of LoVo cells stably carrying either an empty vector or V5-tagged ACSL6 following treatment with OXA. Data are mean ± SEM. I Cell viability assay of LoVo cells stably carrying either an empty vector or V5-tagged ACSL6 following treatment with increasing concentrations of OXA in the presence of DHA. Data are mean ± SEM. *p < 0.05, **p < 0.01 (n = 3). J Quantification of colony formation assay for LoVo cells stably carrying either an empty vector or V5-tagged ACSL6 following treatment with OXA and DHA. Data are mean ± SEM. *p < 0.05, ***p < 0.001 (n = 3). K Viability of MCF7 cells stably carrying either an empty vector or V5-tagged ACSL6 following treatment with DHA. Data are mean ± SEM. **p < 0.01 (n = 3). L Quantification of colony formation assay for ACSL6 overexpression experiments in MCF7 cells. Data are mean ± SEM. *p < 0.05, **p < 0.01, ****p < 0.0001 (n = 3). M Viability of MCF7 cells stably carrying either an empty vector or V5-tagged ACSL6 following treatment with Doxo. Data are mean ± SEM. (n = 3). N Cell viability assays of MCF7 cells stably carrying either an empty vector or V5-tagged ACSL6 following treatment with increasing concentrations of Doxo in the presence of DHA. Data are mean ± SEM. *p < 0.05 (n = 3). O Quantification of colony formation assays for MCF7 cells stably carrying either an empty vector or V5-tagged ACSL6 following treatment with Doxo and DHA. Data are mean ± SEM. *p < 0.05, ***p < 0.001 (n = 3).

Similar effects were observed in MCF7 cells. Elevated ACSL6 expression enhanced sensitivity to DHA (Fig. 3K) and suppressed colony growth in response to DHA alone (Fig. 3L and Fig. S3E), but did not alter the response to Doxo as a single agent (Fig. 3M). Importantly, upregulated ACSL6 significantly increased the cytotoxic effect of DHA in combination with Doxo, as shown by reduced viability and colony-forming ability compared with vector controls (Fig. 3N, O, and Fig. S3F). Collectively, these findings suggest that ACSL6 potentiates the anti-proliferative effects of DHA and markedly enhances the efficacy of chemotherapeutic agents when combined with DHA, while having minimal impact on the response to chemotherapeutic agents alone.

Given that ACSL4 also has a high affinity for DHA, we examined its role in regulating cell growth in response to DHA. Similar to ACSL6 knockdown, LoVo cells expressing ACSL4 shRNAs (Fig. S3G) were less sensitive to DHA treatment (Fig. S3H); however, their sensitivity to DHA combined with OXA treatment was not significantly changed (Fig. S3I). Altogether, consistent with our clinical observations that ACSL6 may predict greater responsiveness to standard treatments, our results indicate that ACSL6 plays a crucial role in modulating the cytotoxic effects of DHA and enhances the efficacy of DHA-combined chemotherapy in colorectal and breast cancer cells.

ACSL6 attenuates ERK and AKT signaling

We next investigate the molecular mechanisms by which ACSL6 exerts its cell growth-suppressive function. The PI3K/AKT and MAPK/ERK pathways are two major intracellular signaling pathways that regulate cell growth, proliferation, survival, and metabolism. Their activities can be assessed by monitoring the phosphorylation status of AKT and ERK, respectively. Cell lysates from LoVo cells expressing ACSL6 shRNAs were subjected to western blot analysis using phospho-specific antibodies against AKT and ERK. ACSL6 knockdown led to increased phosphorylation of both AKT and ERK, indicating activation of these two pathways (Fig. 4A). Conversely, overexpression of V5-ACSL6 reduced of AKT and ERK phosphorylation (Fig. 4B). Similar effects were observed in MCF7 cells (Fig. 4C, D). These findings suggest that ACSL6 negatively regulates PI3K/AKT and MAPK/ERK pathways, with silencing enhancing AKT and ERK activity and overexpression inhibiting it, thereby supporting a role of ACSL6 as a negative regulator of cell growth, potentially through modulation of these pathways.

Fig. 4: ACSL6 attenuates ERK and AKT signaling.

A Representative western blots of LoVo cells stably expressing either a scrambled shRNA or shRNAs specific for ACSL6 (left) and quantification (right). β-Actin served as a loading control. *p < 0.05, **p < 0.01. Data are mean ± SEM. (n = 3). B Representative western blots of LoVo cells stably carrying either an empty vector or V5-tagged ACSL6 and quantification (right). β-Actin served as a loading control. *p < 0.05, **p < 0.01. Data are mean ± SEM. (n = 3). C Representative western blots of MCF7 cells stably expressing either a scrambled shRNA or shRNAs specific for ACSL6 (left) and quantification (right). β-Actin served as a loading control. *p < 0.05, **p < 0.01. Data are mean ± SEM. (n = 3). D Representative western blots of MCF7 cells stably carrying either an empty vector or V5-tagged ACSL6 and quantification (right). β-Actin served as a loading control. *p < 0.05, **p < 0.01. Data are mean ± SEM. (n = 3).

ACSL6 decreases ATP production and activates AMPK signaling

Given that ACSL family members play crucial roles in long-chain fatty acid metabolism and fatty acyl-CoA formation, processes essential for fatty acid activation, ATP metabolism and cell growth regulation [23, 34], we next assessed whether ACSL6 influences ATP turnover. Using the ATPLite Bioluminescence assay in LoVo cells stably expressing ACSL6 shRNA#17, we found that ACSL6 knockdown significantly increased intracellular ATP levels (Fig. S4A). Fluctuations in ATP levels can result from changes in either ATP consumption or production. Since ATP is primarily generated through mitochondrial oxidative phosphorylation (OXPHOS) and glycolysis, we performed a real-time ATP synthesis rate assay with Seahorse XFe24 Analyzer to measure oxygen consumption rate (OCR) and extracellular acidification rate (ECAR). ACSL6 knockdown increased both OCR and ECAR (Fig. S4B) and was associated with a trend toward increased total ATP production, without altering the relative contributions of glycolytic and mitochondrial ATP production (Fig. 5A). An energetic map plotted to compare mitochondrial ATP (mitoATP) and glycolytic ATP (glycoATP) production further revealed that ACSL6 knockdown induced metabolic potential, shifting cells toward a more energetically active state (Fig. 5B). Conversely, ACSL6 overexpression reduced intracellular ATP levels (Fig. S4C), decreased OCR and ECAR (Fig. S4D), reduced total ATP production (Fig. 5C), and shifted cells from a metabolically active to a more quiescent state (Fig. 5D). These findings imply that ACSL6 may act upstream by channeling DHA into its activated form, DHA-CoA, thereby enhancing DHA utilization and broadly suppressing cellular energy metabolism. Collectively, these results indicate that ACSL6 functions as a negative regulator of cellular bioenergetics, whereby its loss enhances ATP production and metabolic activity, while its overexpression suppresses energy metabolism, potentially contributing to its growth-suppressive effects.

Fig. 5: ACSL6 decreases ATP production and activates AMPK signaling.

A Metabolic flux analysis showing quantification of mitochondrial ATP production and glycolytic ATP production in LoVo cells stably expressing either a scrambled shRNA or ACSL6 shRNA. Data are presented as mean ± SEM. *p < 0.05 (n = 3). B Energetic map plotting mitochondrial ATP (mitoATP) versus glycolysis-derived ATP (glycoATP), illustrating overall metabolic potential in LoVo cells stably expressing either a scrambled shRNA or ACSL6 shRNA. Data are presented as mean ± SEM (n = 3). C Metabolic flux analysis showing quantification of mitochondrial ATP production and glycolytic ATP production in LoVo cells stably carrying either an empty vector or V5-tagged ACSL6. Data are presented as mean ± SEM. *p < 0.05 (n = 3). D Energetic map plotting mitochondrial ATP (mitoATP) versus glycolysis-derived ATP (glycoATP), illustrating overall metabolic potential in LoVo stably carrying either an empty vector or V5-tagged ACSL6. Data are presented as mean ± SEM (n = 3). E Representative western blots of LoVo cells stably expressing either a scrambled shRNA or shRNAs specific for ACSL6 (left) and quantification (right). β-Actin served as a loading control. **p < 0.01. Data are mean ± SEM. (n = 3). F Representative western blots of LoVo cells stably carrying either an empty vector or V5-tagged ACSL6 (left) and quantification (right). β-Actin served as a loading control. *p < 0.05, **p < 0.01. Data are mean ± SEM. (n = 3). G Metabolic flux analysis showing quantification of mitochondrial ATP production and glycolytic ATP production in MCF7 cells stably expressing either a scrambled shRNA or ACSL6 shRNA. Data are presented as mean ± SEM. *p < 0.05. H Energetic map plotting mitochondrial ATP (mitoATP) versus glycolysis-derived ATP (glycoATP), illustrating overall metabolic potential in MCF7 cells stably expressing either a scrambled shRNA or ACSL6 shRNA. Data are presented as mean ± SEM (n = 3). I Quantification of western blotting assays of MCF7 cells stably expressing either a scrambled shRNA or shRNAs specific for ACSL6. **p < 0.01 Data are mean ± SEM. (n = 3). J Metabolic flux analysis showing quantification of mitochondrial ATP production and glycolytic ATP production in MCF7 cells stably carrying either an empty vector or V5-tagged ACSL6. Data are presented as mean ± SEM. **p < 0.01 (n = 3). K Energetic map plotting mitochondrial ATP (mitoATP) versus glycolysis-derived ATP (glycoATP), illustrating overall metabolic potential in MCF7 stably carrying either an empty vector or V5-tagged ACSL6. Data are presented as mean ± SEM (n = 3). L Quantification of western blotting assays of MCF7 cells stably expressing either an empty vector or V5-tagged ACSL6. *p < 0.05. Data are mean ± SEM. (n = 3).

Furthermore, the increase in ATP levels following ACSL6 knockdown could reduce the AMP/ATP ratio, potentially leading to inhibiting AMP-activated kinase (AMPK), a known tumor suppressor [34]. To test this, we examined AMPK signaling by measuring the phosphorylation levels of AMPK at Thr172 in the activation loop of the α subunit, along with its downstream effector, acetyl-CoA carboxylase (ACC). As expected, ACSL6 knockdown reduced phospho-AMPK and phospho-ACC levels, indicating diminished AMPK activity (Fig. 5E), whereas ACSL6 overexpression increased the phosphorylation of both AMPK and ACC, consistent with enhanced AMPK signaling (Fig. 5F). Similar results were observed in MCF7 cells that ACSL6 silencing increased intracellular ATP levels (Fig. S4E), enhanced OCR and ECAR (Fig. S4F), elevated total ATP production (Fig. 5G), shifted cells toward a more energetically active state (Fig. 5H), and suppressed AMPK activation (Fig. 5I and Fig. S4G). Conversely, ACSL6 overexpression in MCF7 cells reduced ATP levels and production (Fig. S4H, I), suppressed metabolic activity (Fig. 5J, K), and promoted AMPK signaling (Fig. 5L and Fig. S4J). Altogether, these results support the notion that ACSL6 functions as a negative regulator of ATP metabolism and AMPK signaling, thereby linking it to energy homeostasis and potentially contributing to its role in suppressing cancer cell growth. Consistently, analysis of the Clinical Proteomic Tumor Analysis Consortium (CPTAC) dataset revealed significant decreases in ACSL6 protein levels in several cancers, including breast cancer (Fig. S4K), further supporting its growth-inhibitory function.

ACSL6 enhances DHA-induced ferroptosis

We observed that ACSL6 negatively regulates cell growth by suppressing cell proliferation. To determine whether ACSL6 also regulates cell death, we examined apoptosis following ACSL6 knockdown. We found silencing ACSL6 did not significantly affect the percentage of early apoptotic cells in either MCF7 or LoVo cells (Fig. 6A and Fig. S5A,). Although ACSL6 overexpression modestly increased early apoptosis compared with empty vector controls (Fig. 6B and Fig. S5B), the overall impact on basal apoptotic regulation was minimal. Given the positive association between ACSL6 expression and DHA sensitivity, and evidence that DHA triggers apoptosis in various cancer cells [35], we next assessed whether ACSL6 modulates DHA-induced apoptosis. ACSL6 knockdown attenuated, whereas ACSL6 overexpression enhanced, DHA-induced early apoptosis (Fig. 6A, B, and Fig. S5A, B), suggesting that ACSL6 alone does not strongly induce apoptosis under basal conditions but can potentiate DHA-induced apoptotic cell death.

Fig. 6: ACSL6 enhances DHA-induced ferroptosis.

A Apoptosis analysis of LoVo (left) and MCF7 (right) cells stably expressing either a scrambled shRNA or ACSL6 shRNA following treatment with 0 or 100 µM DHA for 18 h. Data are presented as mean ± SEM. n.s., not significant; *p < 0.05 (n = 4). B Apoptosis analysis of LoVo (left) and MCF7 (right) cells stably carrying either an empty vector or V5-tagged ACSL6 following treatment with 0 or 100 µM DHA for 18 h. Data are presented as mean ± SEM. *p < 0.05, **p < 0.01 (n = 4). C Representative western blots of LoVo cells stably expressing either a scrambled shRNA or shRNAs specific for ACSL6 treated with DHA (left) and quantification (right). β-Actin served as a loading control. Data are mean ± SEM. **p < 0.01, ***p < 0.001 (n = 3). D Representative western blots of LoVo cells stably expressing either a scrambled shRNA or ACSL6 shRNA treated with 100 µM DHA in the presence or absence of 3 µM RSL3 (up) and quantification (down). β-Actin served as a loading control. Data are mean ± SEM. *p < 0.05, **p < 0.01 (n = 3). E Quantification of colony formation assay for ACSL6 knockdown experiments in LoVo treated with DHA in the presence or absence of RSL3. Data are mean ± SEM. **p < 0.01, ***p < 0.001 (n = 3). F Representative western blots of LoVo cells stably carrying either an empty vector or V5-tagged ACSL6 treated with DHA (left) and quantification (right). β-Actin served as a loading control. Data are mean ± SEM. **p < 0.01, ****p < 0.0001 (n = 3). G Representative western blots of LoVo cells stably carrying either an empty vector or V5-tagged ACSL6 treated with 100 µM DHA in the presence or absence of 1 µM Fer-1 (up) and quantification (down). β-Actin served as a loading control. Data are mean ± SEM. **p < 0.01, ***p < 0.001 (n = 3). H Quantification of colony formation assay for ACSL6 knockdown experiments in LoVo cells treated with DHA in the presence or absence of Fer-1. Data are mean ± SEM. *p < 0.05 (n = 3). I Quantification of BODIPY 581/591 C11 assays in LoVo cells expressing V5-ACSL6, ACSL6 shRNAs, or empty vector controls following DHA treatment. Lipid peroxidation levels were determined by calculating the ratio of mean fluorescence intensity from oxidized C11 (FITC channel) to non-oxidized C11 (PE channel) and normalized to the mock-treated control. Data are mean ± SEM. *p < 0.05, n.s. not significant (n = 3).

Members of the ACSL family have also been implicated in DHA-induced ferroptosis in various cell types. Glutathione peroxidase 4 (GPX4), a key regulator of ferroptosis, plays a critical role in this process, with its downregulation triggering ferroptotic cell death [36]. Since we observed that ACSL6 activates AMPK signaling, and AMPK has been reported to mediate GPX4-dependent ferroptosis [37, 38], we next examined whether ACSL6 regulates DHA-induced ferroptosis by measuring GPX4 levels. In LoVo cells, ACSL6 knockdown increased GPX4 expression, partially rescuing DHA-induced GPX4 downregulation (Fig. 6C) and this effect was abolished by RSL3, a small-molecule inhibitor of GPX4 (Fig. 6D). Notably, the ACSL6 knockdown rescued DHA-suppressed colony formation (Fig. 3B), an effect abolished by RSL3 (Fig. 6E and Fig. S5C). In contrast, ACSL6 overexpression decreased GPX4 levels, further enhancing the DHA-induced reduction in GPX4 expression (Fig. 6F), and this effect was partially reversed by ferrostatin-1 (Fer-1), a ferroptosis inhibitor (Fig. 6G). Importantly, ACSL6 overexpression enhanced DHA-induced colony growth suppression (Fig. 3G), which was partially reversed by Fer-1 (Fig. 6H and Fig. S5D). A similar trend was observed in MCF7 cells (Fig. S5E–J), supporting that ACSL6 enhances DHA-induced ferroptosis, as evidenced by GPX4 reduction.

Finally, to directly quantify membrane lipid peroxidation, a biochemical hallmark of ferroptosis, in DHA-treated cells, we performed the BODIPY 581/591 C11 assay. The results showed that DHA treatment increased lipid peroxidation, which was further enhanced by ACSL6 overexpression but not by ACSL6 knockdown in both of LoVo and MCF7 cell lines (Fig. 6I and Fig. S5K, L). Collectively, these findings suggest that ACSL6 regulates cell death mainly through ferroptosis rather than apoptosis, and support that ACSL6 enhances DHA-induced ferroptosis by promoting lipid peroxidation and modulating the GPX4-associated ferroptosis pathway, thereby sensitizing cancer cells to chemotherapy in combination with DHA supplementation.

ACSL6 enhances chemotherapy sensitivity in combination with DHA in vivo

To validate the role of ACSL6 in modulating tumor sensitivity to combined oxaliplatin (OXA) and DHA treatment, we established a xenograft tumor model using LoVo cells stably expressing V5-ACSL6, ACSL6 shRNAs, or empty vector controls. We found that baseline tumor growth was not significantly affected by ACSL6 expression status (Fig. 7A and Fig. S6A). However, compared with vector controls, DHA supplementation enhanced the tumor-suppressive effects of OXA in ACSL6-overexpressing tumors, whereas ACSL6 silencing attenuated the inhibitory effects of the combination treatment (Fig. 7A, S6B–D). Mouse body weight remained stable across all treatment groups (Fig. S6E), indicating that the observed differences were treatment-specific rather than due to systemic toxicity.

Fig. 7: ACSL6 enhances chemotherapy sensitivity in combination with DHA in vivo.

A Relative tumor volume at day 50 in the mice bearing xenograft tumors derived from LoVo cells stably carrying an empty vector, expressing V5-tagged ACSL6, or ACSL6 shRNAs. Mice were treated with either mock (n = 6), DHA (30 mg/kg, 3qw, n = 6), OXA (5 mg/kg, 1qw, n = 6), or the combination of DHA and OXA (n = 6). Tumor volume was normalized to the mock-treated group. Data are mean ± SEM. n.s. not significant. *p < 0.05, **p < 0.01, ***p < 0.001. B Representative images of H&E staining and IHC analysis of GPX4 levels in xenograft tumor samples displayed in (A). Scale bar: 25 μm. n = 6.

Moreover, immunohistochemistry (IHC) analysis using antibodies against GPX4 was performed to assess ferroptosis in these tumors, and revealed that GPX4 expression was decreased in ACSL6 overexpression tumors and increased in ACSL6 silencing tumors. DHA treatment alone markedly reduced GPX4 in ACSL6-overexpressing tumors but had only a modest effect in ACSL6 knockdown tumors, consistent with our in vitro western blot results. OXA treatment alone showed little effect on GPX4 expression. Importantly, the combination of DHA and OXA further suppressed GPX4 compared with either agent alone in ACSL6-overexpressing tumors, whereas this combinatorial effect was not evident in ACSL6 knockdown tumors (Fig.7B). Collectively, these in vivo findings demonstrated that ACSL6 enhances chemotherapy sensitivity in the presence of DHA, accompanied by GPX4 downregulation as an indicator of ferroptosis. by promoting ferroptosis. This highlight ACSL6 as a potential predictive biomarker of treatment response and support the use of DHA supplementation as a strategy to improve chemotherapy efficacy in ACSL6-high tumors.