Page 92 - SRPSKO DRUŠTVO ISTRAŽIVAČA RAKA

Serbian Association for Cancer Research SDIRSACR

Telomere Length and hTERT Expression
Relative telomere length was measured using qPCR based on the Cawthon method. Telomere primers and IFNB1 were
used to calculate the T/S ratio. hTERT expression was quantified in patient-derived and control cell lines by RT-qPCR,
normalized to β-actin, and expressed relative to HeLa cells. Data were analyzed using one-way ANOVA with Tukey’s
post hoc test.
Chemosensitivity Testing
Drug resistance was assessed using MTT (2D cultures) and CellTiter-Glo 3D assays. Cells were treated with cisplatin,
doxorubicin (0–100 µM), or methotrexate (0–1500 µM) and viability was measured after 72 hours. In 3D conditions,
spheroids were formed before treatment. All treatments were performed in triplicates with appropriate controls.
Single-Cell RNA Sequencing and Analysis
Primary patient-derived cells (passage 3) were prepared for scRNA-seq using the 10x Genomics Chromium platform.
Libraries were generated with the Next GEM 3’ v3.1 kit and sequenced on an Illumina NovaSeq. Raw data were
processed using Cell Ranger and Seurat. Quality filtering excluded cells with <350 genes or >10% mitochondrial reads.
Doublets and low-expression genes were removed.
Dimensionality reduction and batch correction were performed using PCA, CCA, RPCA, and Harmony integration.
Clusters were identified using the Louvain algorithm (resolution 0.4) and visualized via t-SNE. Annotation was guided by
Azimuth and Human Primary Cell Atlas references. Differentially expressed genes (DEGs) were defined using Wilcoxon
rank-sum tests with the Bonferroni correction.
Gene Enrichment and Pathway Analysis
DEGs were used for over-representation analysis (ORA) of Gene Ontology (GO) and Hallmark pathways. Redundant
GO terms were filtered using the simplify function. Gene set enrichment analysis (GSEA) was performed with the
singleseqgset package. DEGs between stages within specific clusters were identified using the FindMarkers function
(FDR < 0.05 and ≥30% detection difference).
Results: Single-cell RNA sequencing of osteosarcoma cells collected at diagnosis, after chemotherapy, and at recurrence
revealed six transcriptionally distinct clusters with dynamic shifts in abundance over time. Clusters 1–3 were enriched
in proliferative and cell cycle genes (PLK1, AURKA, HIST1H1), with a reduction in cluster 1 at recurrence. Cluster 4,
predominant post-chemotherapy, expressed tumor suppressors (CITED2, PNRC1) and stress-related genes, suggesting
a transient quiescent state.
Clusters 5 and 6 were most prominent at recurrence. Cluster 5 showed a chondroblastic signature, including ACAN and
anti-apoptotic MTRNR2L12. Cluster 6 exhibited features of cancer stem cells (CSCs), expressing NES, ALDH1A1, and
metastasis-associated genes such as NCAM1 and ANKRD1. Additionally, it overexpressed myogenic markers (MYOD1,
DES, PAX7), suggesting a muscle-like migration mode distinct from classical EMT [4].
Functional assays supported these transcriptomic findings. CSC frequency increased from diagnosis to recurrence
(~0.0036%), and recurrence-stage cells demonstrated enhanced motility in wound healing assays. Live-cell imaging
revealed morphological changes, with recurrence cells showing greater perimeter, irregular shape, and reduced optical
thickness—consistent with increased invasiveness. Recurrent-stage spheroids were larger and less dense, indicating
reduced cohesion and greater structural plasticity.
Telomere maintenance also shifted during disease progression. Although telomeres shortened and hTERT expression
was undetectable, proliferative capacity was preserved via alternative lengthening of telomeres (ALT). Cluster 6 showed
elevated BLM expression, a key enzyme in ALT [5], suggesting this mechanism supports immortality in aggressive cell
populations.
Chemosensitivity varied by stage and culture model. Doxorubicin sensitivity increased overall, though ALDH1A1
expression in cluster 6 may contribute to resistance through detoxification. Cisplatin resistance was highest in 3D
cultures at recurrence, linked to low drug uptake (SLC31A1) and upregulation of DNA repair genes and the apoptosis
inhibitor XIAP. Methotrexate resistance persisted and was associated with overexpression of DHFR and ADA, despite
low transporter expression.
These results identify cluster 6 as a mesenchymal, therapy-resistant CSC population with muscle-like migratory features
and broad resistance mechanisms. The study’s longitudinal design—tracking tumor evolution within a single patient—
enabled direct observation of how subclones adapt and survive therapy. Despite limitations from in vitro culturing, this
approach allowed functional validation of transcriptomic findings. Cluster 6 emerges as a promising therapeutic target
marked by ALT activity, EMT/myogenic programs, and chemoresistance.
Conclusion: Unlike most OS studies, which are cross-sectional and compare unrelated patient samples [6], our
longitudinal, within-patient approach eliminates interpatient variability and enables direct observation of clonal shifts
and therapy-driven selection. Although this model lacks the native tumor microenvironment, it enabled functional
testing and the enrichment of CSC-like populations. This study highlights the emergence of a highly plastic, drug-
resistant CSC-like population during osteosarcoma progression. Functional and transcriptomic profiling identified
cluster 6 as a key driver of recurrence, featuring ALT-mediated immortality, myogenic remodeling, and resistance to

77