Inhibition of α-1,3-mannosyltransferase sensitizes head and neck squamous cell carcinomas to cetuximab via endoplasmic reticulum stress

Abstract

Developing effective therapeutic strategies for head and neck squamous cell carcinoma (HNSCC) remains a considerable clinical challenge. Cetuximab, a first-line targeted therapy for HNSCC, exhibits limited efficacy. The aim of this study was to explore the potential of α-1,3-mannosyltransferase (ALG3) inhibition in augmenting the therapeutic efficacy of cetuximab. We first analyzed the Cancer Genome Atlas (TCGA) data and found that ALG3 was significantly overexpressed in HNSCC tissues, correlating with worse pathological features and lower overall and disease-specific survival. Functional studies using ALG3-knockdown cells and a subcutaneous tumor model demonstrated that ALG3 inhibition markedly suppressed HNSCC proliferation both in vitro and in vivo. Furthermore, combining ALG3 inhibition with cetuximab elicited potent anti-cancer effects in vitro and in vivo. Mechanistic investigations via quantitative polymerase chain reaction, western blotting, and transmission electron microscopy revealed that ALG3 knockdown induced endoplasmic reticulum (ER) stress in HNSCC cells through the Bip/IRE1α axis. Finally, blocking N‑linked glycosylation synergistically enhanced cetuximab-mediated growth inhibition of HNSCC cells. In conclusion, ALG3 is a promising target to enhance the therapeutic efficacy of cetuximab in HNSCC.

Highlights

• •The elevated ALG3 expression of HNSCC correlates with poorer pathological parameters.
• •ALG3 is a promising target to enhance the therapeutic efficacy of cetuximab in HNSCC.
• •Depletion of ALG3 knockout induced ER stress in HNSCC cells via the Bip/IRE1α axis.

Background

Head and neck squamous cell carcinoma (HNSCC) ranks as the seventh most prevalent cancer worldwide, with over 890,000 new cases and approximately 450,000 associated deaths recorded each year.1, 2 Due to the absence of characteristic symptoms in the early stages, 60% of the patients are diagnosed in the advanced stage of the disease and present locally invasive or metastatic lesions.1, 3 HNSCC treatment primarily involves surgery or radiotherapy, combined with chemotherapy. While immune checkpoint inhibitors and targeted therapies have shown improved efficacy,4, 5, 6 the 5-year overall survival (OS) rate of advanced-stage HNSCC patients remains below 50%, with approximately 65% of patients presenting local and distant recurrence.1, 3, 7, 8

Cetuximab, a monoclonal antibody specific for epidermal growth factor receptor (EGFR), was approved by the food and drug administration as the first-line targeted drug for HNSCC in 2006.⁹ It targets the extracellular domain of EGFR and competes with natural ligands to block downstream cell growth signals.10, 11 However, the clinical efficacy of cetuximab is limited, and the objective response rate to cetuximab monotherapy in patients with recurrent or metastatic HNSCC is approximately 13%.¹² Furthermore, the combination of cetuximab with chemotherapy resulted in only a modest improvement in median OS from 7.4 to 10.1 months, thereby reflecting suboptimal clinical benefits.¹³ Given the limited efficacy of cetuximab monotherapy, there is an urgent need to identify targets that can enhance the sensitivity of HNSCC cells to cetuximab.

In the previous study,¹⁴ we employed a CRISPR-Cas9 library of 1200 highly expressed genes in HNSCC and subcutaneously implanted these cells into mice to facilitate in vivo lethality screening. The tumor tissue deoxyribonucleic acid was sequenced to identify the critical genes involved in tumor progression, along with potential targets capable of enhancing the therapeutic efficacy of cetuximab. The robust rank algorithm (RRA) negative score of alpha-1, 3-mannosyltransferase (ALG3) ranked second in the cetuximab-treated group and fifth in the placebo-treated group when compared to the control group. These results indicated that ALG3 not only facilitates HNSCC progression but may also be a prospective target for sensitizing tumor cells to cetuximab.

In this study, we have shown for the first time that ALG3 inhibition can sensitize HNSCC cells to cetuximab via induction of endoplasmic reticulum (ER) stress. ALG3 is markedly overexpressed in HNSCC tissues and correlates significantly with clinicopathological characteristics and survival outcomes. Furthermore, CRISPR-mediated ALG3 knockdown inhibited the proliferation of HNSCC cells in vitro and suppressed the growth of subcutaneous tumors. At the molecular level, inhibition of ALG3 induced ER stress, which decreased the proliferative capacity of HNSCC cells. Finally, ALG3 inhibition as well as induction of ER stress augmented the therapeutic efficacy of cetuximab against HNSCC.

Methods

Analysis of clinical dataset

Gene expression datasets of TCGA were downloaded from the GDC data portal (https://portal.gdc.cancer.gov). The transcriptomic data (HTSeq-counts and HTSeq-FPKM) of 548 HNSCC and 44 adjacent healthy tissue samples were converted to transcripts per million reads format. The correlation between ALG3 expression and clinical parameters, such as pathological grades, T stage, and N stage, was analyzed by the Wilcoxon rank-sum test or Kruskal-Wallis test, and P < 0.05 was considered statistically significant. The association between ALG3 expression and immune infiltration was explored through single-sample gene set enrichment analysis using the GSVA package in R (v 4.2.1).¹⁵ The HNSCC patients were stratified into the ALG3^high and ALG3^low groups based on the median ALG3 expression, and OS and disease-specific survival (DSS) were analyzed using the KM plotter (https://www.xiantaozi.com/). Hazard ratio (HR) with 95% CI and P values were calculated by the log-rank test in KM Plotter. P < 0.05 was considered statistically significant.

Cell culture and gene editing

Cal27 and SCC-25 cells were cultured in Dulbecco’s modified Eagle medium (Gibco) supplemented with 1% penicillin-streptomycin (PS) 100X solution (HyClone) and 10% fetal bovine serum (FBS; Gibco) at 37 °C under an atmosphere of 5% CO₂. The cells were harvested with 0.25% trypsin (HyClone) for passaging, and only cells from passages below 20 were used for the experiments. Routine mycoplasma test was performed every 2 months.

The CRISPR-Cas9 system¹⁴ was used for knocking down ALG3. Briefly, primer pairs containing sequences of control sgRNAs or sgRNAs targeting human ALG3 were designed and cloned into lentiCRISPR v2 vector with the puromycin resistance gene. For ALG3 overexpression, the coding DNA sequence sequence of human ALG3 was cloned into the pLV3-CMV-MCS-3×FLAG-CopGFP-Puro vector. The packaging plasmids and engineered plasmid vector were transfected into 293T cells to generate lentiviruses. The target cells were incubated with the lentivirus for 48 h in the presence of 8 μg/mL polybrene (Solarbio). The stably transduced cells were selected 24 h after removal of the lentivirus using 2 μg/mL puromycin. The validation of the ALG3^KD cell lines was confirmed by Sanger sequencing and Western blot.

Human control: GAACGUAGAAAUUCCCAUUU.

Human ALG3: GCGGCTCTTCAATGACCCAG.

Mouse models

BALB/c-nude mice (D000521) were purchased from GemPharmatech Co. Ltd., and bred at the animal center of the Second Affiliated Hospital of Chongqing Medical University. All animal procedures were reviewed and approved by the Ethics Committee of the Second Affiliated Hospital of Chongqing Medical University (Approval number: IACUC-SAHCQMU-2025-0200). To establish the xenogeneic tumor model, Ctrl-knockdown (Ctrl^KD) and ALG3-knockdown (ALG3^KD) Cal27 cells were suspended in Dulbecco’s modified Eagle medium containing 25% Corning Matrigel Matrix (v/v) and injected subcutaneously into the dorsal region; each mouse received 1×10⁶ cells. Based on the initial tumor size, the mice were divided into the control and treatment groups 3 days after engraftment and injected intraperitoneally with the control vehicle or 0.1 mg/kg tunicamycin and 50 μg cetuximab.

Growth inhibition assay

The growth inhibition rate of Ctrl^KD and ALG3^KD cells treated with vehicle or cetuximab was quantified by measuring the bioluminescence signals of GFP-luciferase fusion protein. Briefly, HNSCC cells were seeded in a 96-well plate at the density of 5000 cells and incubated overnight with 2 μg/mL cetuximab. The luminescence signals were read using a SpectraMax i3 Multi-Mode Microplate Reader to quantify the surviving cancer cells. Survival rate was calculated by normalizing the luminescence of surviving cancer cells to that of the control group. Conversely, the inhibition rate was calculated as 1 minus the survival rate.

Immunohistochemistry

Tumor tissues were fixed in 10% neutral buffered formalin, embedded in paraffin, and cut into thin sections. The tissue sections were incubated overnight with anti-ALG3(1:200, Proeintech), anti-Cleaved Caspase-3(1:400, CST), anti-Ki67 (1:1000, Servicebio), anti-EGFR (1:100, Servicebio), anti-IRE1α (1:200, Proeintech), and anti-Bip (1:200, CST) antibodies at 4 °C in a wet chamber. Following incubation with horseradish peroxidase-conjugated species-specific secondary antibody at room temperature for 1 h, color was developed using diaminobenzidine. The sections were then counterstained with hematoxylin and viewed at 20x magnification using the Olympus VS120 Slide Scanning System.

Transmission electron microscopy

The ultrastructural changes in the ER of the suitably treated cells were analyzed by transmission electron microscopy (TEM). Following treatment with 0.1 µg/mL tunicamycin, the Cal27 and SCC25 cells were washed with phosphate-buffered saline, pre-fixed with 2.5% glutaraldehyde in the dark for 3-5 min, and pre-embedded with 1% agarose. The samples were then fixed in 1% osmium acid in the dark at room temperature for 2 h, dehydrated, embedded, polymerized, and cut into ultrathin sections. The sections were stained with 2% uranyl acetate and lead citrate for 10 min and examined by a HITACHI HT7800 transmission electron microscope.

RNA sequencing

Ctrl^KD and ALG3^KD Cal27 cells were seeded in 6-well plates at the density of 2 × 10⁶ cells per well in triplicate. Total RNA was extracted using TRIzol Reagent (Thermo Fisher Scientific), purified, and reverse transcribed for library construction and sequencing by Shanghai Majorbio Bio-pharm Biotechnology Co. Ltd. according to the manufacturer’s protocols (Illumina). Paired-end RNA sequencing (RNA-seq) libraries were sequenced on the Illumina NovaSeq 6000 platform, with a read length of 2 × 150 bp. Differential gene expression and functional enrichment analyses, including Kyoto Encyclopedia of Genes and Genomes pathway analysis and gene set enrichment analysis (GSEA), were conducted using the Majorbio Cloud Platform (www.majorbio.com). Raw sequencing reads have been deposited in the NCBI Sequence Read Archive database（Accession number: PRJNA1354403）.

Western blotting

Cells were lysed in RIPA lysis buffer supplemented with protease and phosphatase inhibitors. The protein content of the lysates was determined using the BCA Protein Assay kit. Protein samples were diluted in the SDS-PAGE Sample Loading Buffer and separated by SDS polyacrylamide gel electrophoresis. The membrane blots were probed with primary antibodies against ATF4 (60035-1-Ig, Proteintech), ATF6 (24169-1-AP, Proteintech), CHOP (15204-1-AP, Proteintech), IRE1α (#3294, Cell Signaling Technology), PERK (24390-1-AP, Proteintech), phospho-PERK (81251-2-RR, Proteintech), EGFR (ET1604-44, HUABIO), and GAPDH (60004-1-Ig, Proteintech), followed by horseradish peroxidase-conjugated anti-rabbit IgG antibodies.

Statistical analysis

All statistical analyses were performed using GraphPad Prism Version 9.0. Data were analyzed using ANOVA and unpaired t-test to determine significant differences. P-value < 0.05 was considered statistically significant. Quantitative results are presented as mean ± SD values.

Results

ALG3 expression correlates with the clinical characteristics and prognosis of HNSCC

Based on the screening results of the CRISPR-Cas9 library (Supplemental Figure 1),¹⁴ we retrieved ALG3 expression data from the TCGA-HNSCC datasets. Compared to adjacent normal tissues, ALG3 was significantly overexpressed in HNSCC tissues (Figure 1a, P < 0.001). ALG3 expression is significantly higher in male patients compared to female patients (Supplemental Figure 2a, P < 0.001). In addition, ALG3 expression was higher in the grade 2 and 3 tumors compared to grade 1 tumors (Figure 1b, P < 0.001). Primary tumors with more extensive involvement demonstrate elevated levels of ALG3 expression within the tumor tissue (Figure 1c, P < 0.01). Furthermore, lymphovascular invasion (Figure 1d, P < 0.05) and regional lymph node metastasis (Figure 1e-f, P < 0.001) were significantly associated with higher ALG3 expression levels. On the other hand, a negative correlation was observed between ALG3 expression and tumor infiltration of CD8+ T cells, Th1 cells, and NK CD56^dim cells (Figure 1g). Finally, patients with elevated ALG3 expression had lower OS and DSS rates compared to those with low ALG3 expression (Figure 1h, P < 0.05, HR = 1.34; Figure 1i, P < 0.05, HR = 1.55). Furthermore, high ALG3 expression significantly predicted poorer survival in both male (HR = 1.39, P = 0.047) and female (HR = 1.75, P = 0.028) patients (Supplemental Figure 2b-c), demonstrating a consistent prognostic direction. Taken together, ALG3 is overexpressed in HNSCC and correlates with worse pathological parameters and clinical outcomes.

Fig. 1.

ALG3 expression correlates with different HNSCC clinical characteristics. (a) ALG3 gene expression in the HNSCC tissues (n = 503) and adjacent healthy tissues (n = 44) based on the TCGA data set. (b) ALG3 expression levels across different pathological grades of HNSCC. (c) ALG3 expression levels across different primary tumor stages of HNSCC. (d) ALG3 expression levels in HNSCC with/without lymphovascular invasion. (e) and (f) ALG3 expression levels in HNSCC with different regional lymph node metastasis grades. (g) Correlation between ALG3 expression levels and immune cell infiltration in tumor tissues. (h) and (i) Kaplan-Meier curves comparing the OS and DSS of HNSCC patients stratified by median ALG3 expression level. ns P ≥ 0.05, *P < 0.05, **P < 0.01, ***P < 0.001.

Knockdown of ALG3 inhibited the growth of subcutaneous HNSCC xenografts in mice

To determine the functional role of ALG3 in HNSCC growth, we established cell lines with stable ALG3 knockdown using the CRISPR system (Supplemental Figure 3), and the ALG3 protein abundance was reduced to about 65% of the control level. The Ctrl^KD and ALG3^KD Cal27 cells were subcutaneously injected into Balb/c-nude mice to establish xenografts. The tumors derived from ALG3^KD cells exhibited markedly slower growth and reduced volume compared to the Ctrl^KD tumors (Figure 2a and b). Consistent with the sustained knockdown of ALG3, tumors from the ALG3^KD group exhibited a significant reduction in the proliferation marker Ki-67 and in EGFR expression, along with a concomitant increase in the apoptosis marker Cleaved Caspase-3 (Figure 2c and d), collectively demonstrating potent anti-tumor effects. Interestingly, Western blot analysis of EGFR in vitro revealed comparable total protein levels between groups, but with a faster migrating band of EGFR in the ALG3^KD cells (Supplemental Figure 4). This suggests that ALG3 inhibition may primarily affect the functional modification of EGFR rather than its total expression. Taken together, ALG3 knockdown markedly diminished the proliferative capacity of the HNSCC cells and inhibited the growth of subcutaneous xenografts.

Fig. 2.

Inhibition of ALG3 inhibited tumor growth in vivo. (a) Representative tumor images on day 23 after engraftment. (b) Tumor growth curve (n = 4); (c) Representative IHC images showing ALG3, Cleaved Caspase-3, Ki-67 and EGFR expression in tumor tissues. Scale bar, 25-50 µm. (d) ALG3-positive area, Cleaved Caspase-3-positive area, Ki-67-positive area and EGFR-positive area in the immunostained sections. ns P ≥ 0.05, *P < 0.05, **P < 0.01, ***P < 0.001.

ALG3 promoted the proliferation and migration of HNSCC cells in vitro

RNA sequencing of the Ctrl^KD and ALG3^KD HNSCC cells (Supplemental Figure 5a) and subsequent GSEA revealed a positive correlation between ALG3 expression and the cell cycle and proliferation gene set signatures (Supplemental Figure 5b-c). Furthermore, ALG3 knockdown significantly reduced proliferation rates of the Cal27 and SCC25 cell lines over a period of 72 h in the CCK-8 assay, while ALG3 overexpression had the opposite effect on their proliferative capacity (Figure 3a-d, P < 0.001). Comparable outcomes were observed in the colony formation assay (Supplemental Figure 5d). Consistent with this, ALG3 knockdown led to a marked decrease in the percentage of EdU-positive proliferative cells (Figure 3e). In addition, wound healing assay was performed using the Ctrl^KD, ALG3^KD, and ALG3-overexpressing (ALG3^OE) cells, and the results showed that inhibition of ALG3 impaired the migration of HNSCC cells, while its overexpression enhanced migration capacity (Figure 3f-g). These findings collectively demonstrated that ALG3 expression promotes the proliferation and migration of HNSCC cells in vitro.

Fig. 3.

ALG3 promoted the proliferation and migration of HNSCC cells in vitro. (a-d) The percentage of viable Ctrl^KD, ALG3^KD, and ALG3^OE Cal27 and SCC25 cells in the CCK-8 assay (n = 6). (e) Representative images showing EdU-labeled cells in the indicated groups (blue: DAPI; purple: Edu), scale bar, 50 µm. (f) and (g) Representative images showing wound coverage by the Ctrl^KD, ALG3^KD, and ALG3^OE Cal27 and SCC25 in the wound healing assay. Scale bar, 200 µm. ns P ≥ 0.05, *P < 0.05, **P < 0.01, ***P < 0.001.

Knockdown of ALG3 augmented the cytotoxic effect of cetuximab against HNSCC cells

To determine whether targeting ALG3 enhanced the efficacy of cetuximab against HNSCC, we treated ALG3^KD HNSCC cell lines with different concentrations of cetuximab and observed growth inhibition (Supplemental Figure 6). The results indicated that ALG3 knockdown in combination with cetuximab significantly inhibits HNSCC cell growth. Knockdown of ALG3 significantly enhanced the cytotoxicity of cetuximab (2 μg/mL) in terms of proliferative inhibition, and this inhibitory effect increased in a time-dependent manner (Figure 4).

Fig. 4.

Knockdown of ALG3 augmented the cytotoxic effect of cetuximab against HNSCC cells (a-d) Growth inhibition of Ctrl^KD and ALG3^KD Cal27 and SCC25 cell lines following 24 h/48 h treatment with 2 μg/mL cetuximab, quantified by luminescence (n = 3). (e) and (f) Percentage of viable Ctrl^KD and ALG3^KD HNSCC cells following treatment with vehicle or cetuximab (n = 6). ns P ≥ 0.05, *P < 0.05, **P < 0.01, ***P < 0.001.

Inhibition of ALG3 induced ER stress in HNSCC cells

ALG3 is localized within the ER and the Golgi apparatus,¹⁶ and catalyzes Dol-P-Man-dependent mannosylation of proteins at the luminal side of the ER.¹⁷ The results of GSEA showed that ALG3 expression correlated negatively with the unfolded protein response gene set signature (Supplemental Figure 7), suggesting a potential link between ER stress and ALG3. ER stress is an adaptive response that restores ER and intracellular homeostasis following oxidative stress, genetic variation, or nutrient deprivation and is mediated by IRE1α, PERK, ATF6, and BiP/GRP78. To validate the effect of ALG3 on ER stress, the cells were treated with the N-glycosylation inhibitor tunicamycin, and the expression levels of factors related to ER stress were analyzed.

The expression of IRE1α and BiP mRNAs was 1-2-fold higher in the ALG3^KD Cal27 cells compared to the Ctrl^KD Cal27 cells (Figure 5a, P < 0.001), while the CHOP, ATF6, IRE1α, and BiP transcripts were upregulated in the ALG3^KD SCC25 cells (Figure 5b, P < 0.005). We then examined the levels of ER-related proteins in Ctrl^KD and ALG3^KD Cal27 and SCC25 cell lines, with tunicamycin-treated cells as the positive control group. The results showed that ALG3 knockdown upregulated the IRE1α, Bip in both Cal27 and SCC25 cells (Figure 5c and d, Supplemental Figure 8). The above results indicated that inhibition of ALG3 may activate ER stress in HNSCC cells through the Bip/IRE1α axis. Notably, ALG3 inhibition in SCC25 cells potently induces ER stress, resulting in a high basal BiP level that is not further augmented by subsequent tunicamycin treatment under these conditions (Figure 5d). To specifically evaluate IRE1α activation, we measured XBP1 splicing. The sXBP1/uXBP1 mRNA ratio was significantly elevated in ALG3^KD cells compared to Ctrl^KD cells (Supplemental Figure 9). This confirms that ALG3 inhibition activated the IRE1α-XBP1 branch of the ER stress. The structural changes in subcellular organelles were observed by TEM. The ALG3^KD Cal27 and SCC25 cells exhibited a marked dilatation of the ER lumen compared to control cells (Figure 5e and f, marked with a red rectangle).

Fig. 5.

Inhibition of ALG3-induced ER stress in HNSCC cells. (a) and (b) Relative expression of IRE1α, BiP, CHOP, and ATF6 mRNAs in the indicated groups (n = 3). (c) and (d) Immunoblot showing relative expression of PERK, IRE1α, Bip, CHOP, and ATF6 proteins in the indicated groups (n = 3). The right panels showing densitometric quantification of IRE1α and Bip expression from three independent biological replicates, normalized to GAPDH. (e) and (f) Representative TEM images showing changes in the ER (red rectangle) of Ctrl^KD and ALG3^KD HNSCC cells. Scale bar, 200 nm. The HNSCC cells were treated with 0.1 µg/mL tunicamycin (Tun) or vehicle for 24 h. ns P ≥ 0.05, *P < 0.05, **P < 0.01, ***P < 0.001.

N‑linked glycosylation blockade and cetuximab synergistically inhibited the growth of HNSCC xenografts in vivo

Given the key role of ALG3 in the early stages of N-linked glycosylation ¹⁷ and the impact of ALG3 inhibition on the therapeutic efficacy of cetuximab in vitro (Figure 4), we next evaluated the effect of blocking N-terminal glycosylation in the cetuximab-treated HNSCC cells. While ALG3 inhibition significantly enhanced the anti-tumor effects of cetuximab, the combination of tunicamycin and cetuximab exerted the strongest cytotoxic activity in vitro (Figure 6a and b). We assessed that both perturbations sensitized cells to cetuximab, their effects are not additive when combined (Supplemental Figure 10), likely due to the profound ER stress already triggered by tunicamycin. To validate the results in vivo, Balb/c-nude mice were subcutaneously injected with Ctrl^KD or ALG3^KD Cal27 cells and treated with tunicamycin and cetuximab (Figure 6c). As shown in Figure 6d and e, tunicamycin or ALG3 knockdown synergistically suppressed tumor growth with cetuximab. Furthermore, no significant loss in body weight was recorded for any of the groups (Figure 6f). In vivo tumor context, tumors from the ALG3^KD group showed increased expression of both Bip and IRE1α (Supplemental Figure 11). Taken together, blocking N-linked glycosylation through genetic inhibition of ALG3 or pharmacological intervention could enhance the therapeutic efficacy of cetuximab.

Fig. 6.

N-linked glycosylation inhibitor tunicamycin synergized with cetuximab in HNSCC treatment. (a) and (b) The percentage of viable Ctrl^KD, ALG3^KD Cal27, and SCC25 cells treated with vehicle or 0.1 µg/mL tunicamycin, 2 μg/mL cetuximab for 24 h (n = 6), the pairwise comparison details were provided in Supplemental Tables 1 and 2. (c) Schematic illustration of the treatment of syngeneic tumor model with tunicamycin and cetuximab (n = 6). (d) Tumor growth curves of the indicated groups. (e) Representative tumor images on day 18 after engraftment. (f) Body weight of mice from the indicated groups. ns P ≥ 0.05, *P < 0.05, **P < 0.01, ***P < 0.001.

Discussion

HNSCC is a significant contributor to cancer-related mortality on a global scale. Given that approximately 65% of patients are diagnosed at locally advanced stages, as well as the limited efficacy of the first-line agent cetuximab, the 5-year survival rate of HNSCC patients remains below 50%. Based on preliminary screening conducted using a CRISPR pooled library in mice, we had previously identified ALG3 as a potential target to improve the efficacy of cetuximab. In the present study, functional assays demonstrated that ALG3 promoted the proliferative activity of HNSCC cells. Furthermore, ALG3 knockdown synergistically enhanced the anti-tumor efficacy of cetuximab by blocking cellular N‑linked glycosylation. Inhibition of ALG3 also induced ER stress in the tumor cells through the Bip/IRE1α axis. This study has shown for the first time that selective ablation of ALG3 can enhance the efficacy of cetuximab in HNSCC.

Analysis of TCGA data showed that ALG3 is significantly upregulated in HNSCC tissues and correlates with shorter OS and DSS, thereby highlighting its potential as a prognostic biomarker. In addition, ALG3 expression also showed a significant association with tumor pathological grades, tumor stages, lymphovascular invasion, and regional lymph node metastasis, all of which are key predictors of HNSCC progression and prognosis. Furthermore, ALG3 has been identified as a potential prognostic biomarker and a therapeutic target for multiple cancers, such as ovarian cancer,¹⁸ bladder cancer,¹⁹ breast cancer,16, 20 lung cancer,²¹ and hepatocellular carcinoma.22, 23 Consistent with the findings of Shao et al,²⁴ we found that ALG3 promotes the proliferation of HNSCC cells by regulating the cell cycle pathway, indicating that it is an independent risk factor for HNSCC. In our previous study,¹⁴ ALG3 emerged as one of the top five hits from the HNSCC CRISPR-Cas9 library when comparing tumor growth in the control, placebo-treated, and cetuximab-treated groups. Therefore, ALG3 is not only an independent prognostic factor for HNSCC but also a potential therapeutic target.

The ER is a central organelle in eukaryotic cells where approximately more than a third of all cellular proteins are synthesized and folded. The uncontrolled proliferative capacity of malignant cells in growing tumors generates a hostile microenvironment characterized by high metabolic demand, hypoxia, nutrient limitation, and acidosis. These adverse conditions alter the protein folding capacity of the ER in cancer cells, consequently inducing ER stress.²⁵ In mammalian cells, ER stress is initiated by three ER transmembrane proteins that function as sensors of protein-folding stress: IRE1α, ATF6, and PERK. During proteostasis, the molecular chaperone BiP/GRP78 binds to these sensors and maintains them in an inactive state.26, 27 However, during ER stress, BiP is titrated away from sensors due to its ability to bind misfolded proteins with higher affinity, leading to the activation of ER stress sensors—such as IRE1α in this study. IRE1α can further activate nuclear factor-κB (NF-κB) pathways that regulate cell death.²⁸ A productive, non-lethal ER stress response restores ER homeostasis, thereby ensuring survival. In contrast, unresolved or extreme ER stress can lead to cell death.²⁶ We found that inhibition of ALG3 induced ER stress through the Bip/IRE1α axis, resulting in proliferative arrest of HNSCC cells. Intriguingly, the accompanying BiP induction was markedly more pronounced in SCC25 cells compared to Cal27 cells, highlighting that the amplitude of this compensatory response varies across cellular contexts and may influence the ultimate fate of stressed cells.

ALG3 is an ER-resident glycosyltransferase that plays a key role in the N-linked glycosylation of proteins, a common post-translational modification that predominantly occurs within the ER and Golgi apparatus.²⁹ N-linked glycosylation contributes to protein stabilization, accurate intracellular localization, and the promotion of correct protein folding through molecular chaperones.³⁰ In the study, ALG3 functions as a dedicated glycosyltransferase for N-linked glycosylation. Its inhibition thus primarily impairs this process, as evidenced by the aberrant electrophoretic mobility of the heavily N-glycosylated EGFR. The subsequent ER stress and synergistic interaction with cetuximab are likely consequences of this specific glycoproteomic alteration. Recent studies have shown that dysregulated protein glycosylation in tumor cells maintains proliferative signaling, prevents programmed cell death and induces chemoresistance.³¹ In fact, ALG3 has been identified as a potential marker of radiosensitivity in breast cancer and may promote radioresistance and cancer stemness through the glycosylation of TGFBR2.¹⁶ Inhibition of ALG3 has been shown to disrupt N-linked glycosylation, facilitate ferroptosis in tumor cells, and contribute to a pro-inflammatory microenvironment, thereby augmenting anti-tumor immune responses.³² The absence of ALG3 was shown to reduce the glycosylation of PD-L1, which in turn improved the therapeutic efficacy of anti-PD-1 antibody in triple-negative breast cancer.³³ Furthermore, a recent study demonstrated that increased expression and glycosylation of PD-L2 in tumors are associated with cetuximab resistance.³⁴ Thus, the combination of glycosylation inhibition and cetuximab is a promising therapeutic strategy for HNSCC.

In summary, our study highlights the significant potential of ALG3 as a therapeutic target for HNSCC and the feasibility of glycosylation inhibition or ALG3 inactivation as potential strategies for boosting the efficacy of cetuximab. Future research should elucidate the pathways downstream of the ALG3 modulating ER stress via the Bip/IRE1α axis to identify more effective therapeutic targets for HNSCC.

Funding and support

This work was supported by the China Postdoctoral Science Foundation [grant number 2024M763890 (Yu Hao)]; Special Funding for Postdoctoral Research Projects in Chongqing [grant number 2024CQBSHTB3106 (Yu Hao)]; Kuanren Talents Program of the second affiliated hospital of Chongqing Medical University (Shengguo Wang).

CRediT authorship contribution statement

Yu Hao: Writing – review & editing, Writing – original draft, Validation, Supervision, Resources, Project administration, Funding acquisition, Conceptualization. Guangchuan Deng: Writing – review & editing, Project administration. Bolei Li: Writing – review & editing, Project administration, Conceptualization. Lei Cheng: Writing – review & editing, Conceptualization. Shengguo Wang: Supervision, Investigation. Zhenzhou Yang: Writing – original draft, Investigation, Conceptualization.

Declarations of interest

The authors declare that they have no competing interests that could have appeared to influence the work reported in this paper. All authors are consent for publication.

Appendix A. Supplementary material

Supplementary material

.

Data availability

Data will be made available on request.