Comparative RNA-Seq Analysis of Long Interspersed Element-1 (L1) Expression in Two-and Three-Dimensional models and Patient-Derived Triple Negative Breast Cancer (TNBC)
Publication Date
July 2026
Document Type
Abstract
Start Date
24-7-2026 9:00 AM
End Date
24-7-2026 3:00 PM
Description
Long Interspersed Element-1 (L1) is the only autonomously active retrotransposon in the human genome and is frequently reactivated in cancers of epithelial origin, including breast cancer, due to epigenetic dysregulation. However, it remains unclear whether commonly used experimental models accurately reflect L1 mRNA expression in cancer patients. This study compared locus specific L1 expression in 2D and 3D breast cancer cell cultures with patient derived triple negative breast cancer (TNBC) samples to determine how accurately experimental models reflect patient L1 expression. To address this aim, L1 RNA-Seq alignment approach and L1ABA automated pipelines were used to identify expressed L1 loci from publicly available RNA sequencing datasets of breast cancer cell lines and patients with TNBC tumors. As a quality control, we confirmed that file sizes did not differ significantly among datasets and were not correlated with the number of expressed L1s. While each of the three tested 2D cell culture models (MCF7, Luminal A; HCC1806, TNBC; BT474, Luminal B) expressed at least one hundred L1 loci, only eight L1 loci were expressed in all three cell lines. In 3D models (HCC1806 and BT474 at Day 10 and Day 20), ten expressed L1 loci were shared among all the samples. L1 expression did not significantly differ between the 2D and 3D models, although some L1 loci were uniquely expressed in each model. In contrast, when the identities of L1 loci expressed in samples collected from patients with Grade 2 TNBC were compared to the HCC1806 2D and 3D cell culture models, there was a statistically significant difference in the repertoire of expressed L1s. None of the L1 loci expressed in 2D or 3D models were expressed in any of the three TNBC patient samples. Our findings establish that although 2D and 3D breast cancer models exhibit similar locus-specific L1 expression patterns, neither accurately captures the heterogeneity of L1 expression observed in human tumors.
Recommended Citation
Gambhira, Saanvi, "Comparative RNA-Seq Analysis of Long Interspersed Element-1 (L1) Expression in Two-and Three-Dimensional models and Patient-Derived Triple Negative Breast Cancer (TNBC)" (2026). Summer Research Internship Program. 9.
https://digitalscholar.lsuhsc.edu/srip/2026/undergrad/9
Comparative RNA-Seq Analysis of Long Interspersed Element-1 (L1) Expression in Two-and Three-Dimensional models and Patient-Derived Triple Negative Breast Cancer (TNBC)
Long Interspersed Element-1 (L1) is the only autonomously active retrotransposon in the human genome and is frequently reactivated in cancers of epithelial origin, including breast cancer, due to epigenetic dysregulation. However, it remains unclear whether commonly used experimental models accurately reflect L1 mRNA expression in cancer patients. This study compared locus specific L1 expression in 2D and 3D breast cancer cell cultures with patient derived triple negative breast cancer (TNBC) samples to determine how accurately experimental models reflect patient L1 expression. To address this aim, L1 RNA-Seq alignment approach and L1ABA automated pipelines were used to identify expressed L1 loci from publicly available RNA sequencing datasets of breast cancer cell lines and patients with TNBC tumors. As a quality control, we confirmed that file sizes did not differ significantly among datasets and were not correlated with the number of expressed L1s. While each of the three tested 2D cell culture models (MCF7, Luminal A; HCC1806, TNBC; BT474, Luminal B) expressed at least one hundred L1 loci, only eight L1 loci were expressed in all three cell lines. In 3D models (HCC1806 and BT474 at Day 10 and Day 20), ten expressed L1 loci were shared among all the samples. L1 expression did not significantly differ between the 2D and 3D models, although some L1 loci were uniquely expressed in each model. In contrast, when the identities of L1 loci expressed in samples collected from patients with Grade 2 TNBC were compared to the HCC1806 2D and 3D cell culture models, there was a statistically significant difference in the repertoire of expressed L1s. None of the L1 loci expressed in 2D or 3D models were expressed in any of the three TNBC patient samples. Our findings establish that although 2D and 3D breast cancer models exhibit similar locus-specific L1 expression patterns, neither accurately captures the heterogeneity of L1 expression observed in human tumors.
Comments
Mentor: Dr. Victoria Belancio, Tulane University