SPP1 promotes cancer stemness and reduces osimertinib sensitivity in non-small cell lung cancer through interactions with CD44.

Introduction
Lung cancer remains the leading cause of cancer-related mortality worldwide, imposing an enormous burden on global health [1]. Lung cancer is a heterogeneous disease with bronchial mucosal and alveolar origins and a wide range of clinicopathologic features [2]. Non-small cell lung cancer (NSCLC) accounts for 85% of all lung cancer cases [3]. According to the World Health Organization (WHO), approximately 1.8 million of patients with cancer are diagnosed with NSCLC each year [4]. However, owing to the high degree of deterioration associated with NSCLC, especially in clinically diagnosed patients who are usually in the middle to late stages of the disease, clinical treatment options are limited [5].
Although surgery, radiotherapy and chemotherapy are still the mainstays of NSCLC treatment, these have drawbacks that affect patient prognosis [6–8]. With the development of biomedicine, targeted therapies are gradually changing the status quo of NSCLC treatment.
The mechanisms of resistance to epidermal growth factor receptor–tyrosine kinase inhibitors (EGFR–TKIs) are multifactorial and can be broadly categorized as EGFR-dependent and EGFR-independent. EGFR-dependent mechanisms involve mainly secondary mutations within the kinase domain—such as T790M and C797S—that prevent drug binding. EGFR-independent mechanisms include MET or HER2 amplification, activation of bypass signaling pathways (PI3K/AKT, MAPK), and phenotypic adaptations such as epithelial–mesenchymal transition (EMT) or histologic transformation [9–12]. Nevertheless, the molecular basis of resistance to third-generation EGFR–TKIs (e.g., osimertinib) remains incompletely understood. With continuous development and progress in the biomedical field, the third-generation TKI osimertinib is being used as a first-line drug for targeted NSCLC treatment. Despite some progress in targeted drug therapy, drug resistance remains a major problem in cancer treatment. After an EGFR mutation occurs, the use of an EGFR–TKI does not improve patient survival time [13]. Therefore, exploring the mechanism of osimertinib resistance is key to overcoming drug resistance in tumors, and improving resistance targets or pathways is the most powerful way to solve NSCLC-associated drug resistance problems.
Transformation of the histology and cellular phenotype is among the main mechanisms through which osimertinib resistance is acquired [14]. Cellular phenotypic transformation is tied to cancer stem cells (CSCs). CSCs are usually found within tumors and are the cause of cancer recurrence [15]. The ability of CSCs to self-renew, differentiate and proliferate results in unfavorable patient prognosis, including treatment resistance and metastasis [16]. Therefore, exploring the association between drug resistance and CSCs is particularly important (Fig. 1).
In this study, we explored the mechanism of SPP1–CD44 resistance to osimertinib in EGFR-mutated NSCLC based on CSCs. We demonstrated that SPP1 and CD44 interactions have important effects on tumor drug resistance.

Materials and methods

Cell lines and culture
The human lung cancer cell lines HCC827 (EGFR-mutant) and H1299 (EGFR wild-type) were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Osimertinib-resistant cells (HCC827-OR) were established from parental HCC827 cells through stepwise exposure to increasing concentrations of osimertinib (AstraZeneca, UK) ranging from 0.1 to 100 nM. Cells that continued to proliferate under 100 nM osimertinib were considered resistant and designated as HCC827-OR cells.
All the cell lines were cultured in RPMI-1640 medium (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco) and 1% penicillin–streptomycin and maintained at 37 °C in a humidified incubator with 5% CO₂. All the cell lines were authenticated and routinely tested for mycoplasma contamination.
For siRNA transfection experiments, a scrambled siRNA (siNC) sequence with no homology to any known human gene was used as a negative control. Both siSPP1 and siNC were transfected under identical conditions using Lipofectamine 3000 (Invitrogen).

Establishment of drug-resistant cells
The concentration escalation method was used to construct a cell line resistant to osimertinib. HCC827 cells were treated with osimertinib, and cells with resistance were screened by determining the semi-inhibitory concentration (IC50) according to the drug concentration of 0 nM → 0.1 nM → 0.5 nM → 1 nM → 5 nM → 10 nM → 50 nM → 100 nM to induce cells to develop stable drug resistance.

Western blot
Total protein was extracted using RIPA lysis buffer containing protease and phosphatase inhibitors (Beyotime, China). Equal amounts of protein (30 µg per lane) were separated by SDS–PAGE and transferred to PVDF membranes (Millipore). After blocking with 5% skim milk for 1 h, the membranes were incubated overnight at 4 °C with primary antibodies, followed by incubation with HRP-conjugated secondary antibodies. The protein bands were visualized using enhanced chemiluminescence (ECL, Thermo Fisher) and quantified by ImageJ software.
Primary antibodies against SPP1 (1:1000, Abcam, ab91655), CD44 (1:1000, Proteintech, 60224-1-Ig), and GAPDH (1:3000, Beyotime, AF1186) were used.

CCK-8
Cell viability was assessed using a CCK-8 assay (Beyotime, C0037). Briefly, 3 × 10³ cells per well were seeded into 96-well plates and treated with gradient concentrations of osimertinib for 72 h. CCK-8 reagent (10 µL) was added to each well and incubated for 1 h, after which the absorbance at 450 nm was recorded using a microplate reader.

Sphere formation assay
H1299, HCC827 and HCC827-OR cells were cultured in 24-well ultralow adhesion plates at a density of 500 cells/well under the indicated conditions. Next, 5 µg/mL insulin (Wisent, 325-043-EL) and 2 ng/mL recombinant human epidermal growth factor (Sigma, E9644) were added to the medium. The number of spheres > 50 pm in diameter was counted after 10 days.

Colony formation assay
Twenty-four hours after transfection, 500 cells were plated into 6-well plates and cultured for 2 weeks. Thereafter, the cell colonies were fixed with paraformaldehyde for 5 min and stained with 0.1% crystal violet (Beijing Suolaibao/25 g) for 15 min at room temperature. The cell colonies were then counted and photographed.

qRT–PCR
Total RNA was extracted using TRIzol reagent (Invitrogen). cDNA was synthesized using a PrimeScript RT kit (Takara), and quantitative PCR was performed using SYBR Green Master Mix (Takara) on a Bio-Rad CFX96 system.
hsa-NANOG-F: TTTGTGGGCCTGAAGAAAACT.
hsa-NANOG-R: AGGGCTGTCCTGAATAAGCAG.
hsa-OCT4-F: CTGGGTTGATCCTCGGACCT.
hsa-OCT4-R: CCATCGGAGTTGCTCTCCA.
hsa-SOX2-F: GCCGAGTGGAAACTTTTGTCG.
hsa-SOX2-R: GGCAGCGTGTACTTATCCTTCT.
hsa-GAPDH-F: GGAGCGAGATCCCTCCAAAAT.
hsa-GAPDH-R: GGCTGTTGTCATACTTCTCATGG.

Coimmunoprecipitation (Co-IP)
Coimmunoprecipitation was performed using 1 g of antibody. Protein A/G PLUS agarose immunoprecipitation reagent was added, and the samples were incubated for 8 h at 4 °C. The beads were subsequently washed 3 times with 1 ml of immunoprecipitation buffer and then subjected to Western blot analysis.

Xenograft subcutaneous implantation model
Four- to six-week-old BALB/c nude mice were housed and maintained in an SPF-level animal laboratory. All animal experiments were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (2011) and approved by the Animal Experimental Ethics Committee of The First People’s Hospital of Kunming City, Yunnan Province (no. Kmmu20241765).
The mice were randomly divided into four groups (n = 6 per group): (1) HCC827; (2) HCC827-OR; (3) HCC827-OR + siSPP1; and (4) HCC827-OR + siNC. Each mouse was injected subcutaneously with 1 × 10⁷ cells. The tumor volume was measured every 2 days using the formula V = (L × W²)/2. Osimertinib was administered orally at 5 mg/kg/day. The mice were sacrificed after 30 days for tumor collection and HE, IHC, and IF analyses.

Hematoxylin–eosin (HE) staining
Paraffin-embedded tumor tissues were HE stained for tumor observation.

Immunohistochemical evaluation of mouse tumor tissues
IHC was performed to assess the tumor tissues. The tumor tissues were sectioned, and the tissues were stained with Ki67 (Abcam, ab16667) at a dilution of 1:300, after which the staining was observed.

Immunofluorescence detection of SPP1 and CD44 expression
The tissues were fixed and sectioned, and specific antibodies [CD44 (Proteintech, 60224-1-Ig) and SPP1 (Abcam, ab91655)] were added. After the primary antibodies were washed away, a fluorescently labeled secondary antibody [Alexa Fluor 488 donkey anti-rabbit IgG (H + L) (Abcam, ab150061), Alexa Fluor 594-conjugated donkey anti-mouse IgG (H + L) (Abcam, ab150105)] was added, and the samples were observed under a fluorescence microscope (Zeiss, LSM710).

Statistical analysis
All the statistical analyses were performed using GraphPad Prism 8 software, and all the data are expressed as the mean ± standard deviation of n observations. Statistical data from more than two groups were analyzed by one- or two-way ANOVA, and multiple comparisons were performed by Tukey’s test. A p value < 0.05 was considered to indicate statistical significance.

Results

Important role of CSCs in EGFR mutation
To explore the mechanism of EGFR mutation, we investigated whether stem cells play an important role in EGFR mutation through tumor ball formation tests and cell clonogenesis tests. As shown in Fig. 2A, the sensitivity of H1299 cells to osimertinib detected by the CCK8 method was greater than that of HCC827 cells. The data in Fig. 2C confirm that compared with H1299 cells, HCC827 cells with EGFR mutation were more resistant to osimertinib, as shown in Fig. 2B and D, and the degree of tumor pellet formation and cell proliferation were greater in HCC827 cells than in H1299 cells. In addition, qRT-PCR was used to detect the expression levels of the tumor stem cell factors NANOG, OCT4 and SOX. The results revealed that the expression levels of tumor stem cell factors in HCC827 cells were greater those that in H1299 cells (Fig. 2E). These results suggest that EGFR mutation may induce CSCs.

SPP1 expression is upregulated in osimertinib-resistant cells
To identify genes potentially involved in acquired resistance, SPP1 expression was examined across NSCLC cell lines. Western blot analysis revealed markedly elevated SPP1 levels in EGFR-mutant HCC827 cells compared with H1299 cells (Fig. 3A). CCK-8 assays confirmed that HCC827-OR cells exhibited reduced sensitivity to osimertinib (Fig. 3B). Furthermore, compared with parental HCC827 cells, osimertinib-resistant HCC827-OR cells displayed a pronounced increase in SPP1 expression (Fig. 3C). These findings indicate that upregulation of SPP1 may be associated with the development of osimertinib resistance in EGFR-mutant NSCLC.

SPP1 knockdown reverses osimertinib resistance and reduces CSC-like traits
To investigate the mechanism of resistance to osimertinib, we conducted a tumor formation experiment and found that the resistant cells had a stronger ability to form spheres (Fig. 3D). To clarify the functional role of SPP1, siRNA-mediated knockdown was performed in HCC827-OR cells. Western blotting confirmed a marked reduction in SPP1 protein levels following siSPP1 transfection (Fig. 3E). At the molecular level, the expression levels of the stemness-associated transcription factors NANOG, OCT4, and SOX2 were markedly increased in resistant HCC827-OR cells compared with those in parental HCC827 cells (Fig. 3F). Functionally, SPP1 silencing significantly suppressed cell proliferation and colony-forming ability (Fig. 3G-I). Moreover, sphere formation was markedly impaired in siSPP1-treated cells (Fig. 3H). qRT–PCR analysis further revealed that the expression levels of NANOG, OCT4, and SOX2 were substantially downregulated upon SPP1 inhibition (Fig. 3J). Together, these results demonstrate that SPP1 enhances osimertinib resistance by maintaining CSC-like stemness in NSCLC cells.

SPP1 promotes osimertinib resistance through upregulation of CD44 expression
Because CD44 is a well-recognized CSC marker implicated in therapeutic resistance, we next examined whether SPP1 regulates CD44 expression. Western blot analysis revealed that CD44 expression was significantly higher in HCC827 cells than in H1299 cells (Fig. 4A) and further elevated in HCC827-OR cells compared with parental cells (Fig. 4B). Notably, SPP1 knockdown markedly decreased CD44 protein levels (Fig. 4C). These findings suggest that SPP1 may promote osimertinib resistance by upregulating CD44 expression, thereby maintaining CSC properties.

SPP1 targets CSC stemness by interacting with CD44
Given that both SPP1 and CD44 are key regulators of CSC maintenance, we investigated their potential molecular interactions. Coimmunoprecipitation assays confirmed the direct physical interaction between the SPP1 and CD44 proteins (Fig. 5A). In addition, we assayed SPP1 and CD44 expression levels after overexpressing CD44 in HCC827-OR cells in which SPP1 was knocked down, and the results revealed that when SPP1 expression decreased, CD44 expression also decreased (Fig. 5C). Furthermore, the results of a tumor spheroid formation assay revealed that the levels of tumor stem cell marker molecules (NANOG, OCT4, and SOX2) were also decreased (Fig. 5B and D). Cell cloning experiments revealed that the overexpression of CD44 in cells with low SPP1 expression increased the proliferation of tumor cells (Fig. 5E). These results indicate that SPP1 and CD44 synergistically promote the development of drug resistance in tumors.

SPP1 enhanced osimertinib drug resistance in vivo
To validate the in vitro findings, xenograft models were established using H1299, HCC827, HCC827-OR, and HCC827-OR + siSPP1 cells in nude mice. Tumor growth was markedly accelerated in mice bearing HCC827 xenografts compared with those bearing H1299 tumors, confirming that EGFR mutation promotes tumorigenicity (Fig. 6A–D). Compared with mice injected with H1299 cells, mice injected with HCC827 cells had significantly larger tumors, indicating that the EGFR mutation promotes tumor growth. After analyzing the tumor tissues, we found that the levels of SPP1 and CD44 were greater in EGFR-mutant lung cancer than in normal lung cancer (Fig. 6F) and that the levels of tumor marker molecules were significantly greater (Fig. 6E). HE and Ki67 immunohistochemistry were used to stain the tumor tissue, and the results revealed that the tumor tissue from mice with NSCLC with EGFR mutation deteriorated and proliferated more seriously (Fig. 6G and H). Immunofluorescence staining revealed that the expression levels of SPP1 and CD44 in HCC827 cells were greater than those in H1299 cells when these cells were injected into the body to form tumors (Fig. 6I).
In osimertinib-resistant xenografts, SPP1 knockdown significantly reduced tumor growth, weight, and proliferation (Fig. 7A–D). After SPP1 was knocked down, the proliferation of drug-resistant tumors decreased, and the expression levels of SPP1 and CD44 in tumor tissues also decreased (Fig. 7F); moreover, the expression levels of tumor stem cell marker molecules decreased (Fig. 7E). HE staining and Ki67 immunohistochemistry were used to stain the tumor tissues, and the results revealed that the degree of deterioration and proliferation of NSCLC tumor tissues was reduced in mice with SPP1 knockout (Fig. 7G and H). Immunofluorescence staining revealed that the expression levels of SPP1 and CD44 in HCC827-OR-siSPP1 cells decreased after the injection of HCC827-OR-siSPP1 cells into the body to form tumors (Fig. 7I).
The results of the above in vivo experiments revealed that SPP1 enhanced the dryness of tumor cells through a synergistic effect with CD44 to promote osimertinib resistance in NSCLC.

Discussion
The process of cancer development is often accompanied by a series of genetic mutations, some of which may lead to abnormalities in cell signaling, thereby promoting tumor growth, spread, and immortality. EGFR is a receptor of the tyrosine kinase family that influences cell proliferation, differentiation, division, survival and cancer development [17]. EGFR mutations can lead to the aberrant activation of signaling pathways that promote tumor cell proliferation, survival and metastasis, making EGFR mutations important targets in NSCLC treatment.
In recent years, an increasing number of studies have shown that the aberrant expression of SPP1 is closely related to the occurrence, development and metastasis of a variety of tumors [18]. SPP1 is abnormally highly expressed in a variety of tumors and has been used as an important marker [19–24]. Importantly, our data suggest that SPP1/CD44 signaling connects extracellular cues with CSC transcriptional regulation, enabling the persistence of a slow-cycling, drug-tolerant subpopulation. CD44 is considered a marker of CSCs, and cells that overexpress CD44 exhibit resistance. Silencing CD44 in lung cancer inhibits cell growth and induces apoptosis by inactivating EGFR signaling [25]. One study reported the upregulation of CD44 expression after EGFR–TKI treatment [25], which is consistent with our findings. Other studies have also shown that CD44 expression was significantly higher in EGFR–TKI-resistant lung cancer cells than in parental cells (HCC827) [26]. However, several limitations should be considered. Our experiments were primarily conducted in HCC827 and HCC827-OR cells, which may not fully capture the diversity of resistance mechanisms across different EGFR-mutant NSCLC subtypes. Future validation in other EGFR-mutant cell lines (such as PC9 and H1975), as well as patient-derived organoid or xenograft models, will be essential to confirm whether this mechanism is broadly conserved [27, 28]. Additionally, the role of the SPP1-CD44 signaling axis warrants further investigation in terms of other mechanisms of resistance to EGFR–TKIs (e.g., first-generation gefitinib and erlotinib; second-generation afatinib and dacomitinib; and third-generation lazertinib and olmutinib).
Although our results focus on tumor cell–intrinsic SPP1, the tumor microenvironment (TME) may also play a critical role. SPP1 is abundantly secreted by tumor-associated macrophages and fibroblasts, which can reinforce CD44 activation through paracrine signaling [29]. Exploring how TME-derived SPP1 contributes to osimertinib tolerance may help clarify the contextual complexity of this pathway.
From a translational perspective, SPP1 and CD44 may serve as potential biomarkers to predict or monitor osimertinib resistance. Elevated SPP1/CD44 expression in tumors or plasma has been associated with poor therapeutic response and may identify patients at risk for early adaptive resistance [30–32]. Moreover, our findings raise the possibility of targeting the SPP1–CD44 axis as a therapeutic strategy. Preclinical data suggest that CD44-blocking antibodies, SPP1-neutralizing agents, or inhibitors of STAT3 and YAP/TAZ signaling can suppress CSC traits and restore EGFR–TKI sensitivity [33, 34]. Combining such agents with osimertinib may offer a rational approach to delay or overcome resistance. In the future, integrating SPP1/CD44 expression analysis into patient biopsy or liquid biopsy samples before and after treatment could help determine its prognostic value and feasibility as a noninvasive monitoring tool.
Our results reveal that SPP1–CD44-mediated CSC activation represents a novel, nongenetic mechanism of osimertinib resistance. These findings increase the mechanistic depth of EGFR–TKI resistance biology; however, further studies incorporating diverse models and clinical samples are warranted to validate the role of this pathway as both a therapeutic target and a biomarker for acquired resistance in EGFR-mutant NSCLC.

Conclusion
We demonstrate that the overexpression of SPP1 is one of the mechanisms underlying acquired resistance to EGFR–TKIs. The mechanism by which SPP1 confers acquired resistance to EGFR–TKIs may involve CD44 and CSCs.

Supplementary Information