Novel patient-derived tongue squamous cell carcinoma cell lines from non-smokers: 3D and in vivo models for drug response studies.

Modeling the Non-Smoker's Tongue Cancer

Tongue squamous cell carcinoma (TSCC) is a highly aggressive malignancy of the oral mucosal epithelium. It is often characterized by rapid progression and high rates of lymph node metastasis (the spread of cancer to nearby lymph nodes). For decades, the primary drivers of this disease have been identified as tobacco use and heavy alcohol consumption. These lifestyle factors leave distinct biological footprints on the tumor's genetic and proteomic (the study of proteins) landscape. However, a growing clinical reality is emerging. A subset of patients develops TSCC without any history of smoking or heavy drinking.

To understand how these "non-smoker" tumors behave, researchers require specialized laboratory models. Currently, most commercially available TSCC cell lines are derived from patients with known histories of tobacco use. Alternatively, their exposure backgrounds remain entirely unknown. This creates a fundamental gap in oncology. We are attempting to study a diverse disease using tools biased toward a single, specific cause. This paper addresses that gap by establishing two novel cell lines, LMSCC03 and LMSCC16. These were derived from treatment-naïve, non-smoking Brazilian patients.

The limitation of tobacco-centric models

The utility of a cell line depends on its ability to mimic the biological reality of the patient. In the case of oral squamous cell carcinoma (OSCC), existing scientific models suffer from two primary deficiencies. First, there is a lack of clinical annotation. Many existing lines do not specify if the donor was a smoker. This makes it impossible to isolate tobacco-induced mutations from the intrinsic biology of the cancer.

Second, the technical difficulty of establishing primary cultures is immense. Many attempts to grow oral cancer cells in a lab result in complete culture failure. Other attempts lead to the overwhelming dominance of fibroblasts (connective tissue cells) rather than the actual cancerous epithelial cells. Without models representing non-smoking populations, researchers cannot effectively investigate the rising incidence of TSCC in younger patients. This absence limits our ability to develop personalized therapeutic strategies that account for different causes.

Isolating the epithelial signal

The researchers employed a rigorous multi-step protocol to move from raw surgical tissue to stable, pure epithelial cell lines. The goal was to bypass the common pitfall of fibroblast overgrowth. This overgrowth can mask the true characteristics of the tumor.

1. Tissue Dissociation: Fresh tumor fragments were mechanically minced. They were then subjected to enzymatic digestion using collagenase IV. This breaks down the extracellular matrix (the structural scaffolding between cells) to release individual cells.
2. Selective Depletion: To ensure the resulting lines were truly representative of the carcinoma, the authors used selective trypsinization. Trypsin is a proteolytic enzyme (an enzyme that breaks down proteins). It was used to systematically eliminate contaminating fibroblasts.
3. Verification of Identity: Once stable growth was achieved, the researchers used Short Tandem Repeat (STR) profiling. This is a method of analyzing specific, highly variable DNA sequences. They compared the new lines against the DSMZ global database. This confirmed that LMSCC03 and LMSCC16 were unique and not duplicates of existing lines.
4. Three-Dimensional Expansion: The team moved the cells into 3D environments. They utilized Poly-HEMA coated plates to grow spheroids (spherical clusters of cells). They also used Matrigel-embedded organoids (complex, 3D structures that simulate tissue organization).

Divergent responses to chemotherapy

The study characterizes the two new lines by their distinct molecular personalities. The authors report that LMSCC03 and LMSCC16 exhibit different sensitivities to standard chemotherapeutic agents. Specifically, they tested Cisplatin (a DNA-damaging platinum-based drug) and Paclitaxel (a microtubule-stabilizing agent).

The divergence is quantifiable. The authors measure the $IC_{50}$ (the concentration of a drug required to inhibit cell growth by 50%). They find that LMSCC03 is notably more sensitive to Cisplatin. Its $IC_{50}$ was $18.20 \mu\text{M}$. In contrast, LMSCC16 had an $IC_{50}$ of $44.80 \mu\text{M}$, meaning it requires much more drug to achieve the same effect. Similarly, LMSCC03 shows a lower $IC_{50}$ for Paclitaxel ($5.28 \text{ nM}$) than LMSCC16 ($11.21 \text{ nM}$).

This difference in drug response correlates with the underlying genetics. Through DNA sequencing, the authors identify two heterozygous mutations in the TP53 gene of the LMSCC16 line. This includes a missense mutation (p.R273H) that causes a substantial nuclear accumulation of the p53 protein .

This molecular profile aligns with the observed phenotype. LMSCC16 behaves as an intrinsically more resistant, mesenchymal-like model. Meanwhile, LMSCC03 maintains a more traditional epithelial profile. The tumorigenic capacity also varied. LMSCC03 achieved a 100% tumor take rate in nude mice. LMSCC16 showed a lower rate of 50% .

Unresolved questions in the model

While these lines provide a vital new toolkit, they are not exhaustive representations of the disease. The researchers note that LMSCC16 exhibited lower tumorigenicity than LMSCC03. This suggests that not all patient-derived lines translate with equal ease to animal models. There is also the question of viral involvement. Although the study provides a model for non-smokers, the HPV status of the LMSCC16 patient was not determined. This leaves a variable in the biological equation unresolved.

Crucially, the paper identifies a correlation but stops short of proving causation. While the TP53 mutations and the increased resistance to Cisplatin in LMSCC16 appear linked, the authors state their data are correlative. They have not yet performed the functional experiments required to definitively claim these mutations drive the observed resistance.

A necessary expansion of the toolkit

The establishment of LMSCC03 and LMSCC16 represents a successful shift toward more diverse cancer modeling. By providing stable, 3D-capable, and genetically characterized lines from non-smoking patients, the study offers a way to decouple lifestyle-induced damage from the core mechanics of tongue cancer.

Whether these lines become standard in drug-screening pipelines depends on their continued stability. Researchers must also observe how they replicate the complex microenvironmental cues of the human mouth. For now, they serve as a high-fidelity starting point for investigating therapeutic escape in the non-smoking population.

Figures from the paper