Research ArticleOriginal Article
Open Access

CAMK2D isoform 15 facilitates gefitinib resistance via AKT phosphorylation in lung adenocarcinoma

Lingmei Li, Danyang Ren, Yutian Wang, Jiaojiao Zhang, Chenhui Yan, Tongyuan Qu, Ruimin He, Wenshuai Zhang, Wanghao Li, Ningrui Feng, Yalei Wang, Yuhong Guo, Lisha Qi, Lu Cao, Qianru Guo, Qiang Zhao, Zhaoxiang Ye and Wenfeng Cao
Cancer Biology & Medicine May 2026, 20250710; DOI: https://doi.org/10.20892/j.issn.2095-3941.2025.0710

Abstract

Objective: Acquired resistance to epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) is a major challenge in the treatment of lung adenocarcinoma (LUAD). Our previous study demonstrated that EGFR signaling activates calcium/calmodulin-dependent protein kinase 2 (CAMK2). However, the relationship between this mechanism and TKI resistance has not been fully elucidated.

Methods: In this study the role of calcium/calmodulin-dependent protein kinase IIδ (CAMK2D) splice variants in gefitinib resistance was investigated using clinical specimens and chronically TKI-exposed cell lines. The biological effects of CAMK2D isoform 15 were evaluated using in vitro assays and in vivo xenograft models. In addition, integrative multi-omics analysis was applied to identify potential upstream regulators of CAMK2D splicing 15.

Results: Analysis of clinical samples and cell models revealed that CAMK2D overexpression, particularly isoform 15 containing exon 7, was strongly associated with resistance to EGFR-TKIs. Upregulation of isoform 15 enhanced gefitinib resistance, while knockdown of isoform 15 restored drug sensitivity in vitro. Upregulation of isoform 15 significantly increased tumor growth in xenograft models. Mechanistically, isoform 15 directly interacted with AKT and promoted phosphorylation of AKT at Thr308, thereby activating anti-apoptotic signaling. Multi-omics analysis further identified small nuclear ribonucleoprotein polypeptide A1 (SNRPA1) as a key upstream splicing regulator. Downregulation of SNRPA1 promoted isoform 15 expression and induced a high-resistance phenotype.

Conclusions: Our findings revealed a novel mechanism underlying EGFR-TKI resistance in which SNRPA1-controlled alternative splicing drives the expression of CAMK2D isoform 15, leading to activation of AKT signaling and gefitinib resistance. These results identified CAMK2D isoform 15 as a potential prognostic biomarker and therapeutic target, providing a rationale for splicing-directed strategies to overcome EGFR-TKI resistance in LUAD.

keywords

Introduction

Lung adenocarcinoma (LUAD) is the most common histologic subtype of lung cancer with activating epidermal growth factor receptor (EGFR) mutations accounting for approximately 40% of cases worldwide1,2. Although tyrosine kinase inhibitors (TKIs) targeting mutant EGFR initially provide substantial clinical benefit, most patients eventually develop acquired resistance, leading to disease recurrence and metastasis3–6. A deeper understanding of the underlying resistance mechanisms is urgently needed to identify novel therapeutic targets and improve the efficacy of EGFR-TKI therapy.

Study flowchart
Study flowchart

In the first stage, the association between CAMK2D expression and gefitinib resistance was evaluated using lung adenocarcinoma tissues exposed to EGFR TKIs and gefitinib-resistant cell lines. In the second stage, common exon alterations of CAMK2D in gefitinib-resistant cells and classic drug-resistance pathways were examined, revealing that CAMK2D-mediated resistance occurs independently of established mechanisms. In the third stage, mechanistic analyses demonstrated that CAMK2D isoform 15 promotes gefitinib resistance by interacting with AKT and activating the AKT signaling pathway in vitro and in vivo. In the fourth stage, integrative RNA-seq and LC–MS/MS analyses of CAMK2D isoform 15 in resistant cells identified small nuclear ribonucleoprotein polypeptide A1 (SNRPA1) as a key regulator of CAMK2D isoform 15 expression.

Calcium/calmodulin-dependent protein kinase II (CAMK2) is a multifunctional serine/threonine kinase that phosphorylates diverse substrates involved in critical physiologic processes7. Increasing evidence suggests that CAMK2 has an important role in cancer progression, including regulation of cell proliferation and apoptosis8. CAMK2 exhibits tissue-specific expression patterns and CAMK2 upregulation is associated with poor prognosis in multiple cancers, suggesting a role in promoting tumor aggressiveness9. Recent studies have also linked CAMK2 to enhanced cell invasion and lymph node metastasis. Our previous data demonstrated that CAMK2 overexpression is significantly associated with advanced TNM stage and lymph node metastasis in LUAD. Notably, CAMK2 participates in EGFR signaling and promotes the expression of epithelial-mesenchymal transition (EMT)-related factors. Moreover, CAMK2 has been implicated in the regulation of cancer stemness and therapeutic resistance10. We hypothesized that CAMK2 may contribute to resistance to EGFR-TKIs through activation of downstream EGFR signaling pathways based on these findings and our previous observations.

The CAMK2 family is comprised of four genes: CAMK2A; CAMK2B; CAMK2G; and CAMK2D11. Owing to a highly variable linker region, these genes are particularly susceptible to alternative splicing, resulting in the generation of multiple isoforms with potentially distinct biological functions12. The functional significance of specific splice variants in tumorigenesis has been increasingly recognized13. Alternative splicing, an important post-transcriptional regulatory mechanism, contributes to cancer progression and metastasis by generating diverse mature mRNA transcripts through selective splice site usage14. The spliceosome regulates the recognition and processing of pre-mRNA transcripts, thereby shaping transcriptomic complexity15,16. Notably, different isoforms derived from the same gene may exert distinct or even opposing effects on cellular behavior17,18. In our previous study we demonstrated that alternatively spliced isoforms differentially regulate EMT in LUAD cells19. Beyond roles in tumor progression and metastasis, splice variants have been increasingly implicated in drug resistance, representing a major challenge for effective cancer therapy. For example, CAMK2D has been reported to enhance cisplatin resistance in ovarian cancer by suppressing apoptosis20. Given the established role of CAMK2D in chemotherapy resistance, we sought to determine whether CAMK2D also modulates the response to EGFR-TKIs in LUAD.

In the present study we demonstrated that acquired resistance to gefitinib in EGFR-mutant LUAD cells can occur independent of classic resistance mechanisms. CAMK2D isoform 15, which was generated through alternative splicing, was identified as a key driver of gefitinib resistance and the underlying molecular mechanism was further elucidated. Taken together, our findings provide new insights into the molecular basis of TKI resistance and suggest potential strategies for overcoming therapeutic resistance in LUAD.

Materials and methods

Cell culture, reagents, and data collection

Human EGFR-mutant LUAD cell lines (PC9 and HCC827) were used in this study. These cell lines were authenticated by the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) and the identity was further validated by short tandem repeat (STR) profiling (Supplementary material). Gefitinib-resistant sublines, including PC9/G1, PC9/G2, PC9/GR, and HCC827/GR, were generated using a gradient dose-escalation method. All cells were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) (Procell Biotechnology, Wuhan, China) at 37°C in a humidified incubator with 5% CO2. Mycoplasma contamination was routinely tested and not detected. Gefitinib was purchased from Selleckchem (Houston, TX, USA). A CAMK2 inhibitor was purchased from Targetmol (TP2310; Shanghai; China). RPMI-1640 medium and FBS were obtained from Wuhan Procell Biotechnology Co., Ltd. (catalog numbers: 164210-50 and PM150110, respectively; Wuhan, China). Gene expression datasets used in this study were obtained from the NCBI Gene Expression Omnibus (GEO) database.

Tissue samples

Ten LUAD tissue samples were obtained from Tianjin Medical University Cancer Institute & Hospital (Tianjin, China) between January 2012 and December 2020. Among these samples, six were collected before EGFR-TKI treatment and four were collected at disease progression during EGFR-TKI therapy. Among the six pre-treatment samples, two were derived from patients who did not develop resistance during EGFR-TKI therapy. All patients harbored EGFR mutations and received EGFR-TKI treatment. In addition, five paired LUAD samples were collected from the same patients before and after EGFR-TKI treatment. This study was approved by the Ethics Committee of Tianjin Medical University Cancer Institute & Hospital (Approved No. bc20250352). All experiments were performed in accordance with institutional guidelines and relevant regulations.

RNA extraction and quantitative real-time PCR

Total RNA was extracted from cells and tissue samples using TRIzol reagent (TaKaRa, Tokyo, Japan). Complementary DNA (cDNA) was synthesized using the PrimeScript RT Reagent Kit (TaKaRa) according to the manufacturer’s instructions. Quantitative real-time PCR (qRT-PCR) was performed using FastStart Universal SYBR Green Master Mix (Genstar, Beijing, China) on a real-time PCR system (Applied Biosystems, Foster City, CA, USA). The sequences of the primers used are listed in Table S1. Gene expression levels were normalized to β-actin as an internal control. All experiments were performed in triplicate.

Cell Counting Kit-8 (CCK-8) assay

Cell viability was evaluated using CCK-8 (Beijing Boxbio Science & Technology Co., Ltd., Beijing, China). Cells were seeded in 96-well plates at a density of 2000 cells per well and cultured overnight. Cells were then treated with increasing concentrations of gefitinib (0.0005–50 μM) for 48 h. Subsequently, 10 μL of CCK-8 reagent was added to each well and incubated for an additional 2 h. Absorbance was measured at 450 nm using a microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). Cell viability was calculated using the following formula:

Embedded Image

The half-maximal inhibitory concentration (IC50) of gefitinib was calculated using dose–response curves.

Colony formation assay

Cells were seeded into 6-well plates at a density of 200 cells per well and cultured overnight. The culture medium was then replaced with fresh medium containing DMSO or gefitinib (Selleck Chemicals, Houston, TX, USA). Cells were cultured for approximately 14 d to allow colony formation. Colonies were fixed and stained with Wright–Giemsa solution and counted under a light microscope.

Western blot analysis

Cells were lysed using RIPA buffer supplemented with protease and phosphatase inhibitors (Roche, Basel, Switzerland). Protein concentrations were determined using a BCA Protein Assay Kit (Thermo Scientific, Waltham, MA, USA). Equal amounts of protein were separated by 10% SDS-PAGE and transferred onto PVDF membranes. Membranes were blocked with 5% non-fat milk and incubated with primary antibodies overnight at 4°C. After incubation with HRP-conjugated secondary antibodies (K1801; Huabio, Hangzhou, China), protein bands were detected using an enhanced chemiluminescence detection system (Millipore, Burlington, MA, USA). Primary antibodies against CAMK2A (sc-13141), CAMK2B (sc-100366), CAMK2G (sc-134296), and CAMK2D (sc-100362) were obtained from Santa Cruz Biotechnology, Inc. (Santa Barbara, CA, USA). Antibodies against E-cadherin (#3195), N-cadherin (#13116), vimentin (#5741), AKT (#4691), p-AKT (Thr308) (#13038), p-AKT (Ser473) (#4060), ERK (#4067), p-ERK (#4695), cyclin E (#20808), cyclin B1 (#4138), p21 (#2947), p-RB (#8516), caspase-3 (#9662), and cleaved caspase-9 (#9509) were purchased from Cell Signaling Technology (Danvers, MA, USA). β-actin (#4970; Cell Signaling Technology) and GAPDH (#UM4002; Utibody, Beijing, China) were used as a loading control.

Co-immunoprecipitation

Cells were washed twice with cold PBS and lysed in non-denaturing lysis buffer [50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.5% NP-40, and 1 mM EDTA] containing protease and phosphatase inhibitors. Lysates were centrifuged at 12,000 g for 15 min at 4°C. The supernatants were pre-cleared with Protein A/G agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA, USA), then incubated overnight at 4°C with anti-Flag antibody or control IgG (Proteintech, Rosemont, IL, USA). Protein A/G agarose beads were subsequently added and incubated for an additional 2 h. The beads were washed 3–4 times with lysis buffer and boiled in 2× SDS loading buffer for 10 min. The immunoprecipitated proteins were then analyzed by Western blot. Flag-tagged full-length CAMK2D isoform 15 and kinase domain-deleted isoform 15 plasmids were constructed by VectorBuilder (VectorBuilder, Guangzhou, China).

Tumor xenograft model

Female BALB/c nude mice (4–6 weeks old weighint 18–22 g) were obtained from the Sipeifu Research Center of Model Animals (Beijing, China). All animal experiments were approved by the Institutional Animal Ethics Committee and performed in accordance with the National Research Council Guide for the Care and Use of Laboratory Animals (Approved No. AE-2022040). Each mouse was subcutaneously injected with 5 × 106 PC9 cells. Mice were randomly divided into three groups: Overexpression (Ov)-Isoform15 + gefitinib; Overexpression Negative Control (Ov-NC) + gefitinib; and Ov-NC + PBS. Gefitinib was administered orally at 25 mg/kg dissolved in 0.5% Tween-80 (Sigma, St. Louis, MO, USA). Tumor size and body weight were measured every 3 d. Tumor volume was calculated using the formula:

Embedded Image

Tumor growth inhibition (TGI) was calculated as:

Embedded Image

Mice were euthanized and tumors were harvested for further analysis after 35 d.

Histology and immunohistochemistry

Tissues were fixed in 10% formalin, embedded in paraffin, and sectioned (4-μm thick). Sections were deparaffinized in xylene and rehydrated through graded alcohol solutions. Immunohistochemical staining was performed using an immunohistochemistry (IHC) detection kit (Abcam, Waltham, MA, USA). Primary antibodies against CAMK2D, synaptophysin (Syn), chromogranin A (CgA), and CD56 were applied at a dilution of 1:500. Staining intensity was evaluated according to previously reported scoring methods.

Sanger sequencing

Sanger sequencing was performed to identify common resistance-associated mutations in gefitinib-resistant cells. Primers were designed using Primer-BLAST (https://www.ncbi.nlm.nih.gov/tools/primer-blast/). PCR products were purified using exonuclease I and shrimp alkaline phosphatase (Qiagen, Hilden, Germany). Purified products were sequenced using the BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific, Waltham, MA, USA) and analyzed on an ABI 3500 DNA Analyzer (Applied Biosystems, Waltham, MA, USA).

Liquid chromatography–tandem mass spectrometry proteomics and RNA sequencing

Membrane protein fractions were isolated from total cell lysates using a Membrane and Cytosolic Protein Extraction Kit (Beyotime, Shanghai, China). The extracted proteins were analyzed by liquid chromatography–tandem mass spectrometry (LC-MS/MS) using a Q Exactive Plus mass spectrometer (Thermo Fisher Scientific). Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and quantified using a Qubit 4 fluorometer (Thermo Fisher Scientific, Waltham, MA, USA) for RNA sequencing. RNA-seq libraries were prepared using the TruSeq Stranded Total RNA Library Preparation Kit (Illumina, San Diego, CA, USA) and sequenced on a HiSeq 2500 platform (Illumina, San Diego, CA, USA).

Statistical analysis

All data are presented as the mean ± standard deviation (SD) derived from at least three independent experiments, each of which was performed in triplicate. Statistical analyses were performed using SPSS 24.0 software (SPSS, Inc., Chicago, IL, USA). Prior to analysis, homogeneity of variances was evaluated using Levene’s test. For comparisons between two groups, statistical significance was determined using Student’s t-test or Welch’s t-test, depending on whether the variances were homogeneous. For multiple group comparisons, one-way analysis of variance (ANOVA) and post hoc comparisons were performed using Dunnett’s test to evaluate differences between the control and each treatment group. For instances in which multiple t-tests were applied to the same set of samples, the Bonferroni method was used to adjust the P-values. The standard pipeline was followed for RNA-seq data, including: RNA extraction and quality control; library construction; sequencing; data quality control and alignment; gene quantification; and differential expression analysis (using DESeq2/edgeR). The adopted workflow for LC-MS/MS data consisted of the following, after which statistical analysis was performed: spectral preprocessing; peak detection and alignment (using XCMS/MZmine); spectral deconvolution and compound identification (using MetaboAnalystR/MS-DIAL). The Benjamini-Hochberg false discovery rate (FDR) method was uniformly applied for multiple testing correction to control for false positives with an adjusted P-value < 0.05 considered statistically significant.

Results

CAMK2D is closely associated with the development of gefitinib resistance

Our previous studies demonstrated that CAMK2 has a critical role in the phosphorylation-dependent activation of EGFR. In vitro experiments showed that treatment with the EGFR-TKI, AG1478, which is commonly classified as a CAMK2 inhibitor, markedly reduced the level of phosphorylated EGFR (p-EGFR). These findings prompted us to investigate whether members of the CAMK2 family contribute to resistance to TKIs. The CAMK2 family consists of four isoforms: CAMK2A; CAMK2B; CAMK2D; and CAMK2G.

To explore this possibility, tissue samples were collected from 10 patients with LUAD. The cohort was divided into three groups: (1) two patients (brown group) who received first-line EGFR-TKI therapy without disease progression; (2) four patients (green group) enrolled at initial diagnosis who later developed acquired resistance during TKI treatment; and (3) four patients (blue group) who experienced recurrence after TKI therapy (Figure 1A). Samples from the brown and green groups were obtained before treatment, whereas samples from the blue group were collected after treatment.

Figure 1
Figure 1

CAMK2D is closely associated with the development of gefitinib resistance. (A) Schematic diagram illustrating the grouping of lung adenocarcinoma tissue samples. Pretreatment tumor tissues (marked in brown and green) and post-treatment tumor tissues (marked in blue) were collected from 10 patients. Among these, two patients showed no evidence of disease progression following treatment with EGFR-TKIs (brown group). The remaining eight patients experienced disease progression during EGFR-TKI therapy (green group and blue group). (B) Levels of CAMK2 family isoform expression in patient tissues were quantified by RT-qPCR. The color coding corresponds to that shown in (A). ***P < 0.001. (C) Representative IHC images of CAMK2D in lung adenocarcinoma tissues before (brown and green groups) and after treatment (blue group). Scale bar = 100 μm. Left panel: original magnification ×200; right panel: original magnification ×400. (D) Representative IHC staining images of CAMK2D in the same patient before and after EGFR-TKI treatment. Scale bar = 100 μm. Original magnification ×400. (E) The half-maximal inhibitory concentration (IC50) of gefitinib in parental and gefitinib-resistant cells was determined using a CCK-8 assay to evaluate the degree of drug resistance. ***P < 0.001. (F) The proliferative capacity of parental and gefitinib-resistant cells in the presence of gefitinib was evaluated using colony formation assays. Bar graphs are presented as mean ± SD. **P < 0.01; ***P < 0.001. (G) Levels of CAMK2 family isoform protein expression during the development of gefitinib-resistant cell lines were examined by Western blotting. The experiment was independently repeated three times. Independent sample t-tests were used for panels B, E, and F. CAMK2, calcium/calmodulin-dependent protein kinase II; CAMK2A, calcium/calmodulin-dependent protein kinase II alpha; CAMK2B, calcium/calmodulin-dependent protein kinase II beta; CAMK2G, calcium/calmodulin-dependent protein kinase II gamma; CAMK2D, calcium/calmodulin-dependent protein kinase II delta; EGFR-TKI, epidermal growth factor receptor tyrosine kinase inhibitor; GR, gefitinib resistance; IHC, immunohistochemistry.

RT-qPCR analysis showed that the mRNA levels of all CAMK2 family members were increased in post-treatment tissues (Figure 1B, Table S2). Transcriptomic data were analyzed from 530 LUAD cases in the TCGA database together with drug sensitivity data from the GDSC2 database to identify the CAMK2 member most closely associated with TKI resistance. The relationship between CAMK2 family members and sensitivity to TKI drugs was examined (Figure S2). Among these genes, high CAMK2D expression had the strongest association with reduced sensitivity to TKI treatment.

CAMK2D expression was next examined in clinical samples. IHC staining confirmed the RT-qPCR findings and showed low baseline CAMK2D expression in pre-treatment tissues independent of subsequent resistance status (Figure 1C). Five pairs of matched LUAD samples obtained before treatment and after the development of gefitinib resistance were collected to further evaluate treatment-associated changes. IHC staining revealed that CAMK2D expression was markedly increased in gefitinib-resistant tissues compared to the corresponding pre-treatment samples (Figure 1D).

Gefitinib-resistant subclones (PC9/GR and HCC827/GR) were established by chronically exposing EGFR-mutant LUAD cell lines (PC9 and HCC827) to gefitinib and the association between CAMK2D and TKI resistance was further validated (Figure S3). CCK-8 assays showed that the IC50 values for gefitinib were substantially higher in resistant cells (93.959 μM for PC9/GR and 35.876 μM for HCC827/GR) than the parental cells (Figure 1E, Table S2). Colony formation assays further demonstrated that resistant cells maintained stronger proliferative capacity in the presence of gefitinib (Figure 1F, Table S2), confirming the successful establishment of resistant cell models.

Western blot analysis revealed a marked increase in CAMK2D protein levels in gefitinib-resistant cells compared to parental cell lines. In contrast, the expression of other CaMK2 family members remained largely unchanged (Figure 1G).

Taken together, these results indicated that CAMK2D upregulation is closely associated with the development of TKI resistance.

CAMK2D-mediated gefitinib resistance is independent of classic resistance mechanisms

To characterize the mutational profile of the resistant cell lines, genomic DNA was extracted from the parental cell lines (PC9 and HCC827) and the gefitinib-resistant derivatives (PC9/GR and HCC827/GR), followed by Sanger sequencing analysis (Figure 2A, B). The results confirmed the presence of the EGFR exon 19 deletion in both parental and resistant cell lines, including PC9, PC9/GR, HCC827, and HCC827/GR. Notably, the EGFR exon 20 T790M mutation, a common mechanism of acquired resistance, was not detected in PC9/GR or HCC827/GR cells. Other resistance-associated mutations, including exon 21 L858R and exon 20 C797S, were also absent in these resistant cell lines.

Figure 2
Figure 2

CAMK2D-mediated gefitinib resistance is independent of classical resistance mechanisms. (A, B) DNA Sanger sequencing was performed to analyze common EGFR mutations in gefitinib-sensitive and -resistant cells. Mutation profiles were further confirmed by electrophoretic analysis. Red arrows indicate the positions corresponding to exon 19 deletions, exon 20 T790M point mutations, and exon 21 L858R point mutations. No differences were detected between parental and gefitinib-resistant cells at these sites. (C) Western blot analysis of E-cadherin, N-cadherin, and vimentin expression in gefitinib-resistant cells and the corresponding parental cells. The experiment was independently repeated three times. (D) HE and immunohistochemical staining images of CAMK2D, Syn, CgA, and CD56 in gefitinib-sensitive and -resistant cells. Representative images are shown. Scale bar = 100 μm. Original magnification ×400. EGFR, epidermal growth factor receptor; GR, gefitinib resistance; HE, hematoxylin and eosin.

In our previous study CAMK2 was shown to promote EMT in LUAD. Given the well-established association between EMT and drug resistance21, the expression of EMT-related markers in PC9/GR and HCC827/GR cells was determined next using Western blot. Unexpectedly, no significant changes in EMT marker expression were detected in either resistant cell line compared to the parental counterparts (Figure 2C).

Activation of alternative signaling pathways associated with EGFR-TKI resistance was further assessed. Western blot analysis showed that the levels of MET, phosphorylated MET (p-MET), and BRAF expression were markedly increased in PC9/GR cells, whereas no obvious changes were detected in HCC827/GR cells (Figure 2C).

In addition, cellular IHC analysis showed stronger CAMK2D staining in PC9/GR and HCC827/GR cells compared to the respective parental cells. In contrast, no positive staining was detected for the neuroendocrine markers (Syn, CgA, or CD56) (Figure 2D).

Taken together, these findings suggested that CAMK2D-mediated gefitinib resistance occurs independent of classic resistance mechanisms.

Identification and characterization of CAMK2D alternative splice variants are associated with drug resistance

The structural features of CAMK2D were analyzed to further investigate the mechanism by which CAMK2D contributes to gefitinib resistance in LUAD. CAMK2D contains a variable linker region and alternative splicing within this region can generate multiple isoforms (Figure 3A).

Figure 3
Figure 3

Identification and characterization of CAMK2D alternative splice variants are associated with drug resistance. (A) Schematic representation of the CAMK2D protein domains. CAMK2D contains a kinase domain, a calmodulin-binding domain, a variable linker domain, and an ATP-binding domain. (B) Alternative splicing patterns of CAMK2D. The exon composition of selected CAMK2D isoforms is illustrated. Different colors represent exon compositions corresponding to different regions. For example, isoforms 20 and 22 contain exons 13, 15, and 17, whereas isoform 18 contains exons 6, 8, and 9. (C) qPCR validation of differentially expressed exons in PC9 and PC9/GR cells. The bar chart on the right shows the quantification of the qPCR results presented on the left. ns, not significant; **P < 0.01. Bar graphs are presented as the mean ± SD. The experiment was independently repeated three times. (D) Sequence differences between CAMK2D isoforms 15 and 18. The red circle indicates the differential region between the two sequences corresponding to exon 7. (E) Levels of CAMK2D isoform 15 and 18 expression in gefitinib-sensitive and -resistant cells were measured by qPCR. (F) RT-qPCR analysis of isoform 15 expression in PC9 cells during the gradual development of gefitinib resistance. (G) Levels of isoform 15 and 18 expression in tissue samples collected before (cases 1–6) and after treatment (cases 7–10) were analyzed using RT-qPCR. Data are presented as the mean ± SD in (F) and (G). *P < 0.05; **P < 0.01; *** P < 0.001. The experiment was independently repeated three times. Independent sample t-tests were used for panels C and G. Panel F was analyzed using one-way analysis of variance (ANOVA). CAMK2D, calcium/calmodulin-dependent protein kinase II delta; GR, gefitinib resistance; RT-qPCR, reverse transcription quantitative polymerase chain reaction.

CAMK2D transcript sequences were retrieved from the National Center for Biotechnology Information database to characterize these splice variants and a comprehensive alignment of 31 reported CAMK2D isoforms was performed. Sequence data were exported in FASTA format and subjected to multiple sequence alignment using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/). Comparative analysis revealed that exon variability among CAMK2D splice variants primarily involved exons 2, 3, 7, 15, 16, and 21. Alignment of these isoforms showed that different variants contained distinct combinations of these exons (Figure 3B). For example, the predominant isoform 18 lacks exon 7, isoforms 1, 2, 3, 14, and 16 lack exons 15 and 16, isoforms 4, 5, 6, 15, and 18 lack exon 15, isoforms 20 and 22 lack exon 16, and isoforms 2, 8, 10, 15, and 18 lack exon 21.

RNA was extracted from PC9 and PC9/GR cells to determine whether these variable exons were associated with gefitinib resistance and the expression of transcripts containing these six exons was examined. Exon 7 showed the most pronounced difference between parental and resistant cells (Figure 3C, Table S3), suggesting that this exon may have an important role in gefitinib resistance.

Based on this observation, we focused on isoform 15 and isoform 18, which differ only in the presence or absence of exon 7 (Figure 3D). Isoform 15 expression was markedly increased in PC9 cells following the development of gefitinib resistance, whereas isoform 18 showed a moderate decrease. Similar expression trends for isoforms 15 and 18 were also observed in HCC827 cells (Figure 3E). During the stepwise establishment of gefitinib-resistant PC9 cells, isoform 15 expression gradually increased (Figure 3F, Table S3).

The expression of isoforms 15 and 18 was further examined in tumor tissues from 10 patients with LUAD (Figure 3G, Table S3). In samples from cases 1–6 that had not received EGFR-TKI treatment, isoform 15 expression was relatively low, whereas isoform 18 expression was relatively high. In contrast, tissues from cases 7–10 that had undergone EGFR-TKI treatment had increased isoform 15 expression and decreased isoform 18 expression. Notably, the patterns of isoforms 15 and 18 expression in cases 1 and 2 were comparable to the patterns observed in cases 3–6, suggesting that primary resistance was not associated with CAMK2D splice variants.

Taken together, these findings indicated that alternative splicing of CAMK2D is associated with acquired gefitinib resistance. Specifically, upregulation of CAMK2D isoform 15 and downregulation of isoform 18 appeared to contribute to the development of gefitinib resistance in LUAD cells.

Gefitinib induces CAMK2D isoform 15 expression and isoform 15 promotes gefitinib resistance

The functional roles of isoform 15 and isoform 18 were evaluated in LUAD cells to determine which splice variant mediates the effect of CAMK2D on the gefitinib response. Isoform 15 expression was markedly upregulated in gefitinib-resistant cell lines as observed in PC9/GR and HCC827/GR cells.

Isoform 15 was knocked down in PC9/GR and HCC827/GR cells to assess the functional role (Figure 4A, Table S4). CCK-8 assays showed that knockdown of isoform 15 significantly reduced the IC50 of gefitinib in both resistant cell lines (Figure 4A). Colony formation assays demonstrated that downregulation of isoform 15 markedly suppressed the proliferative capacity of PC9/GR and HCC827/GR cells in the presence of gefitinib (Figure 4A). These findings indicated that suppression of isoform 15 partially restores the sensitivity of gefitinib-resistant cells.

Figure 4
Figure 4

Isoform 15 promotes the development of gefitinib resistance. (A–C) Level of isoform 15 or 18 expression in PC9/GR, HCC827/GR, and the respective parental cells as analyzed by qRT-PCR following knockdown of isoform 15 or overexpression of isoform 18 (left panel). Gefitinib IC50 values were determined using the CCK-8 assay after knockdown of isoform 15 or overexpression of isoform 18 or 15 (middle panel). The proliferative capacity of PC9/GR, HCC827/GR, and the parental cells as evaluated using colony formation assays in medium containing gefitinib after knockdown of isoform 15 or overexpression of isoform 18 or 15 (right panel). (D) Subcutaneous xenograft models were established by injecting PC9 cells overexpressing isoform 15 or wild-type PC9 cells into the inguinal region of 4-week-old female BALB/c nude mice. Eight days after implantation, mice were administered 25 mg/kg of gefitinib or an equivalent volume of PBS via oral gavage (n = 5 per group). Representative images of tumor growth are shown. (E) Survival curves of mice were plotted over a 35-d period under 3 treatment conditions. (F) Changes in tumor volume during the experimental period and tumor weights at the end of the experiment are shown. (G) RT-qPCR analysis of isoform 15 expression in xenograft tumor tissues from the 3 groups. Data are presented as the mean ± SD in (A–C), (F), and (G). *P < 0.05; **P < 0.01; ***P < 0.001. Data represent three independent biological replicates. Independent sample t-tests were used for panels A, B, and C. Panels F and G are analyzed using one-way analysis of variance (ANOVA). GR, gefitinib resistance; PBS, phosphate-buffered saline; RT-qPCR, reverse transcription quantitative polymerase chain reaction.

Next, whether increasing isoform 18 expression affected gefitinib responsiveness was examined. Overexpression of isoform 18 in PC9/GR and HCC827/GR cells (Figure 4B, Table S4) resulted in a modest reduction in gefitinib IC50 values. However, this effect was substantially weaker than that observed after isoform 15 knockdown (Figure 4B). In addition, colony formation assays showed that overexpression of isoform 18 did not significantly alter the proliferative capacity of resistant cells in gefitinib-containing medium (Figure 4B).

Isoform 15 was overexpressed in the parental PC9 and HCC827 cell lines to further examine the role of isoform 15 (Figure 4C, Table S4). CCK-8 assays showed that isoform 15 overexpression markedly increased the IC50 of gefitinib compared to cells transfected with the empty vector (Figure 4C), indicating reduced drug sensitivity.

The effect of isoform 15 on gefitinib resistance was next evaluated in vivo. PC9 cells overexpressing isoform 15 were mixed with Matrigel (Corning, Corning, NY, USA) and subcutaneously injected into nude mice, followed by gefitinib treatment. Survival analysis revealed that mice in the Ov-Isoform15 + gefitinib group exhibited a significantly lower percent survival compared to the Ov-NC + gefitinib group in agreement with the in vitro results (Figure 4E). Moreover, the tumor growth rate and tumor weight were significantly higher in the Ov-Isoform15 + gefitinib group than the Ov-NC + gefitinib group (Figure 4F, Table S4).

IHC staining was performed to evaluate the expression of CD56, Syn, CgA, and Ki67 in tumor tissues from each group. CD56, Syn, and CgA staining were negative in all groups, which was in agreement with the results shown in Figure 2D, indicating no evidence of neuroendocrine transformation. In contrast, Ki67 staining intensity correlated with tumor growth. The Ov-NC + gefitinib group exhibited the lowest Ki67 positivity, whereas the Ov-Isoform15 + gefitinib group showed the strongest Ki67 staining (Figure S4).

Finally, RT-qPCR analysis was performed on tumor tissues from the three groups. The results showed that the expression of isoform 15 tended to increase in tumors in the Ov-NC group after gefitinib treatment (Figure 4G, Table S4; P = 0.051).

Taken together, these findings indicated that CAMK2D isoform 15, rather than isoform 18, is the functional splice variant that promotes gefitinib resistance.

Isoform 15 inhibits apoptosis by promoting AKT phosphorylation at Thr308 in PC9/GR and HCC827/GR cells

The structural features of isoforms 15 and 18 were compared to further investigate the mechanism by which isoform 15 promotes gefitinib resistance. As shown in Figure 5A, isoform 15 contains an additional 196 amino acids compared to isoform 18. Bioinformatic analysis of protein tertiary structure further indicated that isoform 15 contains a specific kinase domain that is absent in isoform 18 (Figure 5C).

Figure 5
Figure 5

Isoform 15 inhibits apoptosis by promoting phosphorylation of AKT at Thr308 in PC9/GR and HCC827/GR cells. (A) Schematic illustration of the domain structures of isoform 15 and 18. Isoform 15 contains an additional 196 amino acids. (B) RNA sequencing analysis was performed in PC9 and PC9/GR cells, followed by KEGG pathway enrichment analysis of differentially expressed genes. The top 10 significantly enriched signaling pathways are shown. (C) Three-dimensional protein structures of isoforms 15 and 18 are presented. Regions unique to isoform 15 are highlighted with red circles. (D) Molecular docking analysis of isoform 15 with AKT. In the upper left panel, the blue molecule represents isoform 15 and the yellow molecule represents AKT. The red box indicates the predicted binding region. The enlarged panel shows the interaction interface between isoform 15 (Gln165) and AKT (Thr308). (E) Western blot analysis of p-AKT (Thr308) and p-AKT (Ser473) expression in PC9/GR and HCC827/GR cells (left panel). Western blot analysis was also performed to evaluate changes in p-AKT Thr308, p-AKT Ser473, and p-ERK expression in PC9 and HCC827 cells after overexpression or knockdown of isoform 15 (middle and right panels). (F) Co-IP assays were performed in PC9 and HCC827 cells transfected with empty vector (NC), Flag-tagged full-length isoform 15 (Flag-Isoform15) or Flag-tagged kinase domain-deleted isoform 15 truncation (Flag-Δ-Isoform15). Cell lysates were immunoprecipitated with anti-Flag antibody, followed by immunoblotting with antibodies against p-AKT Thr308 and total AKT. Input controls showing levels of p-AKT Thr308, total AKT, and GAPDH expression in whole-cell lysates are shown below. (G) Western blot analysis of p-AKT and p-ERK expression in xenograft tumors. CAMK2D, calcium/calmodulin-dependent protein kinase II delta; GR, gefitinib resistance; KEGG, Kyoto Encyclopedia of Genes and Genomes; PBS, phosphate-buffered saline.

RNA-seq analysis was performed in PC9 and PC9/GR cells to explore the downstream signaling pathways involved. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis showed that differentially expressed genes were significantly enriched in the PI3K–AKT signaling pathway (Figure 5B), suggesting that isoform 15 may contribute to gefitinib resistance through activation of AKT signaling.

Structural docking analysis was performed using the Vakser Lab docking platform and the PDBePISA interface analysis tool to investigate the potential interaction between isoform 15 and AKT. Overlapping the docking results revealed a potential interaction interface between isoform 15 and AKT. Molecular docking analysis suggested that AKT Thr308 exhibited the strongest binding affinity with isoform 15 Gln165, indicating a stable interaction between the two proteins (Figure 5D).

AKT phosphorylation in gefitinib-resistant cells was examined next. Western blot analysis showed that the phosphorylation levels of AKT at Thr308 and Ser473 were markedly increased in PC9/GR and HCC827/GR cells compared to the parental counterparts (Figure 5E), whereas the total AKT levels remained unchanged.

Isoform 15 was overexpressed in parental PC9 and HCC827 cells to further determine whether isoform 15 regulates AKT activation. Like resistant cells, overexpression of isoform 15 did not alter total AKT levels but significantly increased phosphorylation of AKT at Thr308 and Ser473. In contrast, the levels of total ERK and phosphorylated ERK expression remained largely unchanged. Conversely, knockdown of isoform 15 in resistant cells led to a reduction in AKT phosphorylation at both Thr308 and Ser473.

A kinase domain–deleted truncation mutant of isoform 15 (Δ-isoform 15) was generated to further confirm the interaction between isoform 15 and AKT. Full-length isoform 15 or Δ-isoform 15 was transfected into PC9 and HCC827 cells, followed by co-immunoprecipitation (Co-IP) assays. Phosphorylated AKT at Thr308 was detected in end>immunoprecipitated complexes from cells expressing full-length isoform 15 but was absent in complexes from cells expressing the Δ-isoform 15 mutant (Figure 5F). These results indicated that isoform 15 interacts with AKT through the kinase domain and promotes AKT phosphorylation at Thr308.

Consistent with the cellular findings, analysis of protein extracts from mouse xenograft tumors showed that isoform 15 overexpression significantly increased phosphorylation of AKT at Thr308 and Ser473, whereas the ERK signaling pathway remained largely unchanged (Figure 5G).

Apoptosis-related proteins in gefitinib-resistant cells were examined following isoform 15 knockdown. Western blot analysis showed decreased expression of cyclin B1, cyclin E, p21, and phosphorylated RB, accompanied by reduced AKT phosphorylation. In contrast, the levels of cleaved caspase-3 and -9 were markedly increased (Figure S5A), indicating enhanced apoptosis. In contrast, overexpression of isoform 15 produced the opposite effect.

Rescue experiments were performed using the CAMK2 inhibitor TP2310 because a specific inhibitor targeting CAMK2D isoform 15 is not currently available. Western blot analysis showed reduced AKT phosphorylation in PC9 cells treated with TP2310. CCK-8 assays demonstrated that TP2310 treatment increased the sensitivity of PC9 cells to gefitinib (Figure S5B–D).

Taken together, these findings indicated that CAMK2D isoform 15 promotes gefitinib resistance and suppresses apoptosis in LUAD cells by activating AKT phosphorylation.

Small nuclear ribonucleoprotein polypeptide A1 regulates isoform 15 expression and is associated with gefitinib resistance

LC–MS/MS analysis was performed in gefitinib-resistant cells following isoform 15 knockdown to further investigate the regulatory mechanism underlying isoform 15-mediated gefitinib resistance. Subsequent pathway enrichment analysis using the KEGG indicated that proteins affected by isoform 15 depletion were significantly enriched in pathways related to RNA processing and alternative splicing, particularly spliceosome-associated pathways (Figure 6A). Gene Ontology enrichment analysis revealed significant enrichment in spliceosome-related biological processes (Figure 6B).

Among spliceosome components, the U2 small nuclear ribonucleoprotein complex has a key role in early spliceosome assembly. We therefore examined the expression of U2-related genes in PC9/GR and HCC827/GR cells. RT-PCR analysis showed that several U2-associated genes were differentially expressed in gefitinib-resistant cells compared to the sensitive counterparts (Figure 6C, D and Table S5).

Figure 6
Figure 6
Figure 6

SNRPA1 regulates isoform 15 expression and contributes to gefitinib resistance. (A) Liquid chromatography–tandem mass spectrometry (LC–MS/MS) analysis was performed in PC9 cells after knockdown of isoform 15. (B) KEGG pathway enrichment and Gene Ontology (GO) analyses indicated that differentially expressed genes were significantly enriched in RNA splicing-related pathways and spliceosome-associated processes. (C, D) Levels of U2-related gene expression in PC9/GR and HCC827/GR cells were measured by RT-qPCR. (E) Effects of isoform 15 overexpression on the expression levels of U2-related genes in PC9 cells were evaluated by RT-qPCR. (F) The left panel shows a volcano plot of differentially expressed genes between control and isoform 15-overexpressing cells derived from RNA sequencing data. The Venn diagram illustrates the overlap between LC–MS/MS and RNA sequencing results, identifying SNRPA1 as the only gene common to both datasets. The red dot in the volcano plot indicates SNRPA1. (G) Enrichment analysis revealed that several U2-related genes (SNRPA1, SNRPB, SNRPD2, SNRPD3, and DDX46) were preferentially downregulated in gefitinib-resistant cells. Gene expression ratios are shown. (H, I) RT-qPCR analysis of isoform 15 expression in PC9 cells transfected with si-SNRPA1. CCK-8 assays showed that si-SNRPA1 partially promoted gefitinib resistance in PC9 cells. DDX46 was used as a negative control. Data are presented as the mean ± SD in (A), (D), (E), (H), and (I). ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001. Data represent three independent biological replicates. Independent sample t-tests were used for panels C, D, E, and I. Panel H was analyzed using one-way analysis of variance (ANOVA). DEAD-Box Helicase 46; GR, gefitinib resistance; GO, gene ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; LC–MS/MS, liquid chromatography–tandem mass spectrometry; RT-qPCR, reverse transcription quantitative polymerase chain reaction; RNA-seq, RNA sequencing; SNRPA1, small nuclear ribonucleoprotein polypeptide A1; SNRPB, small nuclear ribonucleoprotein polypeptide B; SNRPB2, small nuclear ribonucleoprotein polypeptide B2; SNRPD2, small nuclear ribonucleoprotein polypeptide D2; SNRPD3, small nuclear ribonucleoprotein polypeptide D3; SF3A1, splicing factor 3A subunit 1; SF3A2, splicing factor 3A subunit 2; SF3A3, splicing factor 3A subunit 3; SF3B1, splicing factor 3B subunit 1; SF3B2, splicing factor 3B subunit 2; SF3B3, splicing factor 3B subunit 3; SF3B4, splicing factor 3B subunit 4; PHF5A, PHD finger protein 5A; SNRPE, small nuclear ribonucleoprotein polypeptide E; SNRPF, small nuclear ribonucleoprotein polypeptide F; TAT-SF1, tat-specific factor 1.

To further explore the relationship between US-associated genes and isoform 15, the expression in PC9 cells overexpressing isoform 15 was analyzed (Figure 6E, Table S5). Five genes were significantly altered, including small nuclear ribonucleoprotein polypeptide A1 (SNRPA1), small nuclear ribonucleoprotein polypeptide B (SNRPB), small nuclear ribonucleoprotein D2 (SNRPD2), small nuclear ribonucleoprotein D3 (SNRPD3), and DEAD-box helicase 46 (DDX46). Notably, the expression of these genes tended to decrease under conditions of isoform 15 overexpression or gefitinib resistance (Figure 6G).

To further identify key regulators, RNA sequencing analysis was performed in parental PC9 and PC9/GR cells. Integration of RNA-seq data with LC–MS/MS results using Venn diagram analysis identified SNRPA1 as a common differentially expressed gene. RNA-seq analysis further confirmed that SNRPA1 expression was significantly reduced in gefitinib-resistant cells compared to parental cells (Figure 6F).

Whether SNRPA1 regulates isoform 15 expression was examined next. Knockdown of SNRPA1 led to a marked increase in isoform 15 expression, whereas knockdown of DDX46 had little effect (Figure 6H, Table S5).

Although DDX46 exhibited the highest abundance in the LC–MS/MS dataset, these results suggested that SNRPA1, rather than DDX46, functions as a potential regulator of isoform 15.

Finally, the functional impact of SNRPA1 on gefitinib sensitivity was assessed. CCK-8 assays showed that the IC50 value of gefitinib was significantly higher in cells transfected with si-SNRPA1 than cells transfected with si-NC. In contrast, DDX46 knockdown did not significantly affect gefitinib sensitivity (Figure 6I, Table S5). These findings indicated that SNRPA1 downregulation contributes to isoform 15 upregulation and promotes gefitinib resistance in EGFR-mutant LUAD cells.

Discussion

Acquired resistance to EGFR-TKIs are a major challenge in the treatment of EGFR-mutant LUAD22–24. The underlying mechanisms are complex and not fully understood. Classic resistance pathways, including secondary EGFR mutations, activation of bypass signaling pathways, and histologic transformation, have been extensively documented25. However, the role of post-transcriptional regulation, particularly alternative splicing, has received relatively little attention.

Alternative splicing generates multiple transcript isoforms from a single precursor mRNA (pre-mRNA), thereby expanding protein diversity and enabling tissue-specific and dynamically regulated gene expression26,27. This process has an important role in tumor proliferation, differentiation, and metastasis28,29. Increasing evidence suggests that splicing alterations contribute to tumor adaptation under therapeutic pressure30. For example, recent studies have reported differential expression of specific isoforms during resistance to PD-1 blockade therapy, highlighting the potential of isoform-specific targets for therapeutic intervention and biomarker development31. In addition, several reports have shown that alternative splicing variants are differentially expressed in EGFR-TKI–resistant tissues and cells, suggesting that splicing regulation may participate in the development of acquired resistance in non-small cell lung cancer32,33.

CAMK2D is a multifunctional serine/threonine kinase that generates multiple isoforms through alternative splicing. CAMK2D has been implicated in the regulation of cell proliferation, apoptosis, and stress responses34. Recent studies have demonstrated that CAMK2D contributes to cisplatin resistance in ovarian cancer and may serve as a potential therapeutic target in BRCA1-associated protein-1-deficient malignant mesothelioma, indicating an important role in regulating drug responsiveness35,36.

Building on these observations, the present study systematically investigated the role of CAMK2D in gefitinib resistance in LUAD. Isoform 15 was identified as a key variant associated with drug resistance through comprehensive analysis of CAMK2D splice variants. Isoform 15, which contains an exon-7–encoded kinase domain, was markedly upregulated in gefitinib-resistant cells and correlated with increased IC50 values in PC9 and HCC827 cells. Functional experiments demonstrated that knockdown of isoform 15 restored gefitinib sensitivity, whereas overexpression promoted resistance both in vitro and in vivo, indicating that isoform 15 acts as a critical molecular determinant of gefitinib resistance. Mechanistically, isoform 15 was shown to directly interact with AKT through the kinase domain and promote phosphorylation of AKT at Thr308. Activation of this signaling pathway suppressed apoptosis and enabled sustained proliferative capacity in the presence of gefitinib. These findings revealed a previously unrecognized mechanism linking CAMK2D alternative splicing to activation of the AKT signaling pathway in TKI resistance.

Analysis of clinical specimens further supported the relevance of this mechanism. Evaluation of tissue CAMK2D from the same patients and across different patients before and after TKI treatment demonstrated a marked increase in CAMK2D protein levels following TKI therapy. Furthermore, the correlation between CAMK2D isoform 15 and various clinicopathologic features were analyzed. Notably, higher isoform 15 expression was significantly associated with an increased frequency of lymph node metastasis (P = 0.031). However, several limitations should be acknowledged. In our cohort, fewer than one-third of patients had complete survival information and the limited sample size precluded robust statistical analysis of overall survival (OS) and progression-free survival (PFS). Nevertheless, our previous study involving 160 LUAD patients demonstrated that overall CAMK2 expression was significantly correlated with TNM stage, lymph node metastasis, and patient survival outcomes. Given that isoform 15 represents a specific CAMK2D splice variant, it is reasonable to speculate that isoform 14 may also have prognostic significance. Future studies with larger patient cohorts will be required to validate this possibility.

To further elucidate the regulatory mechanism responsible for isoform 15 upregulation, proteomic data obtained by LC–MS/MS were integrated with transcriptomic profiles from gefitinib-resistant cells. This integrative analysis identified SNRPA1 as a potential upstream regulator. SNRPA1 is an essential component of the U2 small nuclear ribonucleoprotein complex and has a critical role in pre-mRNA splicing. Previous studies have also implicated SNRPA1 in tumor invasion and metastatic colonization37–39. In the present study, SNRPA1 expression was markedly reduced in gefitinib-resistant cells and inversely correlated with isoform 15 expression. Functional experiments further demonstrated that SNRPA1 knockdown promoted isoform 15 upregulation and increased gefitinib resistance. These findings suggest that SNRPA1 may regulate CAMK2D alternative splicing and thereby influence the development of TKI resistance.

In summary, this study provides the systematic evidence that CAMK2D isoform 15 promotes acquired gefitinib resistance through SNRPA1-mediated regulation of alternative splicing and subsequent activation of AKT signaling. These findings highlight post-transcriptional regulation as an important and previously underappreciated mechanism in TKI resistance. Importantly, the results identified the following two potential therapeutic targets: CAMK2D isoform 15; and the upstream regulator, SNRPA1. Future studies will focus on elucidating the detailed molecular mechanism by which SNRPA1 controls CAMK2D splicing and exploring whether targeting this regulatory axis could restore drug sensitivity in EGFR-TKI–resistant LUAD.

Conclusions

This study demonstrated that alternative splicing of CAMK2D, particularly the upregulation of isoform 15, represents a key mechanism underlying acquired gefitinib resistance in LUAD. Isoform 15 promotes drug resistance independent of canonical EGFR mutations by enhancing AKT phosphorylation at Thr308, thereby suppressing apoptosis and sustaining tumor cell survival. In addition, downregulation of the spliceosome regulator, SNRPA1, significantly increases expression of isoform 15. In vitro and in vivo experiments further confirmed that isoform 15 overexpression markedly accelerates tumor growth under gefitinib treatment, whereas knockdown restores drug sensitivity. Collectively, these findings identified CAMK2D isoform 15 as a potential therapeutic target and suggested that modulation of alternative splicing may represent a promising strategy for overcoming EGFR-TKI resistance in LUAD (Figure 7).

Figure 7
Figure 7

Schematic model: Alternative splicing of CAMK2D drives EGFR-TKI resistance in lung adenocarcinoma. This model illustrates how alternative splicing of CAMK2D pre-mRNA determines cellular response to EGFR-TKI therapy in lung adenocarcinoma. Left panel: Under normal conditions, pre-CAMK2D is processed by the spliceosome to generate isoform 18. This isoform does not activate the AKT signaling pathway, allowing EGFR tyrosine kinase inhibitors, such as gefitinib, to effectively induce apoptosis and suppress proliferation. Cells remain sensitive to EGFR-TKI treatment. Right panel: CAMK2D isoform 15 is preferentially produced when the splicing balance is altered. This isoform retains exon 7 and contains a unique kinase domain. This domain directly interacts with and phosphorylates AKT, and the AKT signaling pathway is continuously activated. AKT signaling promotes cell survival, proliferation, and stemness properties while inhibiting apoptosis, ultimately conferring resistance to gefitinib and other EGFR-TKIs. Aberrant expression of spliceosome components, such as SNRPA1, may contribute to this splicing switch, representing a non-canonical pathway of acquired resistance independent of secondary EGFR mutations. CAMK2D, calcium/calmodulin-dependent protein kinase II delta; EGFR, epidermal growth factor receptor; SNRPA1, small nuclear ribonucleoprotein polypeptide A; TKI, tyrosine kinase inhibitor.

Conflict of interest statement

No potential conflicts of interest are disclosed.

Author contributions

Conceived and designed the analysis: Wenfeng Cao, Zhaoxiang Ye, Qiang Zhao.

Collected the data: Jiaojiao Zhang, Tongyuan Qu, Ruimin He, Wenshuai Zhang, Yalei Wang, Yuhong Guo.

Contributed data or analysis tools: Danyang Ren, Chenhui Yan, Wanghao Li, Ningrui Feng, Lisha Qi.

Performed the analysis: Yutian Wang, Lu Cao, Qianru Guo.

Wrote the paper: Lingmei Li.

Data availability statement

The data generated in this study are available upon request from the corresponding author.

  • Received November 11, 2025.
  • Accepted March 31, 2026.

References

  1. 1.↵
    2. Soltis AR,
    3. Bateman NW,
    4. Liu J,
    5. Nguyen T,
    6. Franks TJ,
    7. Zhang X, et al.
    Proteogenomic analysis of lung adenocarcinoma reveals tumor heterogeneity, survival determinants, and therapeutically relevant pathways. Cell Rep Med. 2022; 3: 100819.
  2. 2.↵
    2. Eltayeb K,
    3. La Monica S,
    4. Tiseo M,
    5. Alfieri R,
    6. Fumarola C.
    Reprogramming of lipid metabolism in lung cancer: an overview with focus on EGFR-mutated non-small cell lung cancer. Cells. 2022; 11: 413.
    OpenUrlCrossRef
  3. 3.↵
    2. Tumbrink HL,
    3. Heimsoeth A,
    4. Sos ML.
    The next tier of EGFR resistance mutations in lung cancer. Oncogene. 2021; 40: 1–11.
    OpenUrlCrossRefPubMed
  4. 4.
    2. Wang H,
    3. Xiong A,
    4. Chen X,
    5. Guo J,
    6. Tang Z,
    7. Wu C, et al.
    CXCR1+ neutrophil infiltration orchestrates response to third-generation EGFR-TKI in EGFR mutant non-small-cell lung cancer. Signal Transduct Target Ther. 2024; 9: 342.
    OpenUrlPubMed
  5. 5.
    2. Boch T,
    3. Köhler J,
    4. Janning M,
    5. Loges S.
    Targeting the EGF receptor family in non-small cell lung cancer—increased complexity and future perspectives. Cancer Biol Med. 2022; 19: 1543–64.
    OpenUrlAbstract/FREE Full Text
  6. 6.↵
    2. Liang X,
    3. Yao Y,
    4. Li X,
    5. Gao T,
    6. Dai X.
    Cancer mortality trends in China from 2013–2021 and projections to 2030. Cancer Biol Med. 2025; 22: 1223–39.
    OpenUrlAbstract/FREE Full Text
  7. 7.↵
    2. He Q,
    3. Li Z.
    The dysregulated expression and functional effect of CaMK2 in cancer. Cancer Cell Int. 2021; 21: 326.
    OpenUrlPubMed
  8. 8.↵
    2. Brzozowski JS,
    3. Skelding KA.
    The multi-functional calcium/calmodulin stimulated protein kinase (CaMK) family: emerging targets for anti-cancer therapeutic intervention. Pharmaceuticals (Basel). 2019; 12: 8.
    OpenUrlPubMed
  9. 9.↵
    2. Peng JM,
    3. Tseng RH,
    4. Shih TC,
    5. Hsieh SY.
    CAMK2N1 suppresses hepatoma growth through inhibiting E2F1-mediated cell-cycle signaling. Cancer Lett. 2021; 497: 66–76.
    OpenUrlCrossRefPubMed
  10. 10.↵
    2. Lin Y,
    3. Huang Y,
    4. Yang B,
    5. Zhang Y,
    6. Ji N,
    7. Li J, et al.
    Precision therapy targeting CAMK2 to overcome resistance to EGFR inhibitors in FAT1-mutated oral squamous cell carcinoma. Chin Med J (Engl). 2025; 138: 1853–65.
    OpenUrlPubMed
  11. 11.↵
    2. Coultrap SJ,
    3. Bayer KU.
    Nitric oxide induces Ca2+-independent activity of the Ca2+/calmodulin-dependent protein kinase II (CaMKII). J Biol Chem. 2014; 289: 19458–65.
    OpenUrlAbstract/FREE Full Text
  12. 12.↵
    2. Senga Y,
    3. Ishida A,
    4. Shigeri Y,
    5. Kameshita I,
    6. Sueyoshi N.
    The phosphatase-resistant isoform of CaMKI, Ca2+/calmodulin-dependent protein kinase Iδ (CaMKIδ), remains in its “primed” form without Ca2+ stimulation. Biochemistry. 2015; 54: 3617–30.
    OpenUrlPubMed
  13. 13.↵
    2. Martinez-Canton M,
    3. Gallego-Selles A,
    4. Galvan-Alvarez V,
    5. Garcia-Gonzalez E,
    6. Garcia-Perez G,
    7. Santana A, et al.
    CaMKII protein expression and phosphorylation in human skeletal muscle by immunoblotting: isoform specificity. Free Radic Biol Med. 2024; 224: 182–89.
    OpenUrlPubMed
  14. 14.↵
    2. Chen Y,
    3. Huang M,
    4. Liu X,
    5. Huang Y,
    6. Liu C,
    7. Zhu J, et al.
    Alternative splicing of mRNA in colorectal cancer: new strategies for tumor diagnosis and treatment. Cell Death Dis. 2021; 12: 752.
    OpenUrlCrossRefPubMed
  15. 15.↵
    2. Deschenes M,
    3. Durand M,
    4. Olivier MA,
    5. Pellerin-Viger A,
    6. Rodier F,
    7. Chabot B.
    A defective splicing machinery promotes senescence through MDM4 alternative splicing. Aging Cell. 2024; 23: e14301.
  16. 16.↵
    2. Li K,
    3. Cheng C,
    4. Piao Q,
    5. Zhao Q,
    6. Yi J,
    7. Bao Y, et al.
    Genome-wide identification of pan-cancer common and cancer-specific alternative splicing events in 9 types of cancer. Genomics. 2024; 116: 110917.
  17. 17.↵
    2. Kanemoto Y,
    3. Sanada F,
    4. Shibata K,
    5. Tsunetoshi Y,
    6. Katsuragi N,
    7. Koibuchi N, et al.
    Expression of periostin alternative splicing variants in normal tissue and breast cancer. Biomolecules. 2024; 14: 1093.
    OpenUrlPubMed
  18. 18.↵
    2. Fajac A,
    3. Simeonova I,
    4. Leemput J,
    5. Gabriel M,
    6. Morin A,
    7. Lejour V, et al.
    Mutant mice lacking alternatively spliced p53 isoforms unveil Ackr4 as a male-specific prognostic factor in Myc-driven B-cell lymphomas. Elife. 2024; 13: RP92774.
  19. 19.↵
    2. Li L,
    3. Qi L,
    4. Liang Z,
    5. Song W,
    6. Liu Y,
    7. Wang Y, et al.
    Transforming growth factor-β1 induces EMT by the transactivation of epidermal growth factor signaling through HA/CD44 in lung and breast cancer cells. Int J Mol Med. 2015; 36: 113–22.
    OpenUrlPubMed
  20. 20.↵
    2. Xu X,
    3. Zheng Z,
    4. Jia L,
    5. Suo S,
    6. Liu B,
    7. Shao T, et al.
    Overexpression of SMARCA2 or CAMK2D is associated with cisplatin resistance in human epithelial ovarian cancer. Oncol Lett. 2018; 16: 3796–804.
    OpenUrlPubMed
  21. 21.↵
    2. Chen R,
    3. Wang J,
    4. Huang S,
    5. Hu X,
    6. He X,
    7. Zhang T, et al.
    Carnitine palmitoyltransferase 1C promotes EMT-associated cisplatin resistance in non-small cell lung cancer cells. Cancer Biol Med. 2025; 22: 48–66.
    OpenUrlAbstract/FREE Full Text
  22. 22.↵
    2. Zhen S,
    3. Jia Y,
    4. Zhao Y,
    5. Wang J,
    6. Zheng B,
    7. Liu T, et al.
    NEAT1_1 confers gefitinib resistance in lung adenocarcinoma through promoting AKR1C1-mediated ferroptosis defence. Cell Death Discov. 2024; 10: 131.
    OpenUrlPubMed
  23. 23.
    2. Shen NX,
    3. Luo MY,
    4. Gu WM,
    5. Gong M,
    6. Lei HM,
    7. Bi L, et al.
    GSTO1 aggravates EGFR-TKIs resistance and tumor metastasis via deglutathionylation of NPM1 in lung adenocarcinoma. Oncogene. 2024; 43: 2504–16.
    OpenUrlPubMed
  24. 24.↵
    2. Zhang D,
    3. Tian X,
    4. Wang Y,
    5. Liu F,
    6. Zhang J,
    7. Wang H, et al.
    Polyphyllin I ameliorates gefitinib resistance and inhibits the VEGF/VEGFR2/p38 pathway by targeting HIF-1a in lung adenocarcinoma. Phytomedicine. 2024; 129: 155690.
  25. 25.↵
    2. Leonetti A,
    3. Sharma S,
    4. Minari R,
    5. Perego P,
    6. Giovannetti E,
    7. Tiseo M, et al.
    Resistance mechanisms to osimertinib in EGFR-mutated non-small cell lung cancer. Br J Cancer. 2019; 121: 725–37.
    OpenUrlCrossRefPubMed
  26. 26.↵
    2. Chen MX,
    3. Tian Y,
    4. Zhu FY,
    5. Fan T,
    6. Yan HX,
    7. Sun PC, et al.
    Alternative splicing of VRF1 acts as a molecular switch to regulate stress-induced early flowering. Cell Rep. 2024; 43: 114918.
  27. 27.↵
    2. Hou JY,
    3. Wang XL,
    4. Chang HJ,
    5. Wang XX,
    6. Hao SL,
    7. Gao Y, et al.
    PTBP1 crotonylation promotes colorectal cancer progression through alternative splicing-mediated upregulation of the PKM2 gene. J Transl Med. 2024; 22: 995.
    OpenUrlPubMed
  28. 28.↵
    2. Wang PS,
    3. Liu Z,
    4. Sweef O,
    5. Xie J,
    6. Chen J,
    7. Zhu H, et al.
    Long noncoding RNA ABHD11-AS1 interacts with SART3 and regulates CD44 RNA alternative splicing to promote lung carcinogenesis. Environ Int. 2024; 185: 108494.
  29. 29.↵
    2. Wei L,
    3. Li Y,
    4. Chen J,
    5. Wang Y,
    6. Wu J,
    7. Yang H, et al.
    Alternative splicing in ovarian cancer. Cell Commun Signal. 2024; 22: 507.
    OpenUrlPubMed
  30. 30.↵
    2. Bonnal SC,
    3. López-Oreja I,
    4. Valcarcel J.
    Roles and mechanisms of alternative splicing in cancer — implications for care. Nat Rev Clin Oncol. 2020; 17: 457–74.
    OpenUrlCrossRefPubMed
  31. 31.↵
    2. Wang X,
    3. Liu T,
    4. Li Y,
    5. Ding A,
    6. Zhang C,
    7. Gu Y, et al.
    A splicing isoform of PD-1 promotes tumor progression as a potential immune checkpoint. Nat Commun. 2024; 15: 9114.
    OpenUrlCrossRefPubMed
  32. 32.↵
    2. He S,
    3. Shi J,
    4. Zhou H,
    5. Li Q,
    6. Wu L.
    Lnc-ABCA12-8 confers acquired resistance to gefitinib in non-small cell lung cancer by regulating the alternative splicing of fibronectin 1 in the IIICS region. Cancer Gene Ther. 2022; 29: 1686–96.
    OpenUrlPubMed
  33. 33.↵
    2. Lu H,
    3. Ai J,
    4. Zheng Y,
    5. Zhou W,
    6. Zhang L,
    7. Zhu J, et al.
    IGFBP2/ITGA5 promotes gefitinib resistance via activating STAT3/CXCL1 axis in non-small cell lung cancer. Cell Death Dis. 2024; 15: 447.
    OpenUrlCrossRefPubMed
  34. 34.↵
    2. Huangfu L,
    3. He Q,
    4. Han J,
    5. Shi J,
    6. Li X,
    7. Cheng X, et al.
    MicroRNA-135b/CAMK2D axis contribute to malignant progression of gastric cancer through EMT process remodeling. Int J Biol Sci. 2021; 17: 1940–52.
    OpenUrlPubMed
  35. 35.↵
    2. Karnan S,
    3. Ota A,
    4. Murakami H,
    5. Rahman ML,
    6. Wahiduzzaman M,
    7. Hasan MN, et al.
    CAMK2D: a novel molecular target for BAP1-deficient malignant mesothelioma. Cell Death Discov. 2023; 9: 257.
    OpenUrlPubMed
  36. 36.↵
    2. Chai B,
    3. Wu Y,
    4. Yang H,
    5. Fan B,
    6. Cao S,
    7. Zhang X, et al.
    Tau aggregation-dependent lipid peroxide accumulation driven by the hsa_circ_0001546/14-3-3/CAMK2D/Tau complex inhibits epithelial ovarian cancer peritoneal metastasis. Adv Sci (Weinh). 2024; 11: e2310134.
  37. 37.↵
    2. Fish L,
    3. Khoroshkin M,
    4. Navickas A,
    5. Garcia K,
    6. Culbertson B,
    7. Hänisch B, et al.
    A prometastatic splicing program regulated by SNRPA1 interactions with structured RNA elements. Science. 2021; 372: eabc7531.
  38. 38.
    2. Chen Y,
    3. Hao Q,
    4. Wang S,
    5. Cao M,
    6. Huang Y,
    7. Weng X, et al.
    Inactivation of the tumor suppressor p53 by long noncoding RNA RMRP. Proc Natl Acad Sci U S A. 2021; 118: e2026813118.
  39. 39.↵
    2. Yuan P,
    3. Ling L,
    4. Gao X,
    5. Sun T,
    6. Miao J,
    7. Yuan X, et al.
    Identification of RNA-binding protein SNRPA1 for prognosis in prostate cancer. Aging (Albany NY). 2021; 13: 2895–911.
    OpenUrlPubMed
Back to top

In this issue

Print
Download PDF
Email Article
Citation Tools

Jump to section

Keywords