Bidirectional Dysregulation of Unprocessed Pseudogene Expression in Testicular Germ Cell Tumors
Ali Kazemlou,1Argam Margosian,2Pouya Salehipour,3,*
1. Medical Genetics Department, Faculty of Medicine, Tehran University of Medical Sciences, Tehran, Iran 2. Medical Genetics Department, Faculty of Medicine, Tehran University of Medical Sciences, Tehran, Iran 3. Medical Genetics Department, Faculty of Medicine, Tehran University of Medical Sciences, Tehran, Iran
Introduction: Testicular germ cell tumors (TGCTs) arise within a tissue governed by highly specialized developmental and transcriptional programs. Although the molecular landscape of TGCT has been extensively characterized at the level of protein-coding genes, substantially less attention has been paid to unprocessed pseudogenes. Unlike processed pseudogenes, unprocessed pseudogenes retain genomic features inherited from their ancestral genes and may therefore provide a distinct view of tissue-associated transcriptional regulation. Systematic characterization of their expression may uncover molecular changes that are not captured by conventional protein-coding gene analyses. The present study examined the expression landscape of unprocessed pseudogenes in TGCT, with particular emphasis on transcripts showing consistent gain or loss of expression relative to normal testicular tissue.
Methods: RNA-sequencing expression data from 150 TGCT tumor samples from TCGA-TGCT and 450 normal testicular tissue samples from GTEx were analyzed. TCGA data were obtained using the TCGAbiolinks package in R. Following preprocessing and sample-level filtering, tumor and normal expression profiles were normalized and subjected to differential expression analysis. Genes annotated as unprocessed pseudogenes according to their Ensembl gene biotype were subsequently examined as a distinct transcript category. Candidate transcripts were prioritized according to the magnitude and direction of expression change, statistical significance after multiple-testing correction, and the proportion of tumor and normal samples in which expression was detected. Detection frequency was defined as the percentage of samples exhibiting expression above the predefined detection threshold. The use of independently generated TCGA and GTEx datasets was considered when interpreting differences between tumor and normal tissue.
Results: Unprocessed pseudogenes displayed a pronounced bidirectional expression pattern in TGCT. Three candidates showed marked gains of expression in tumor tissue: AC103563.1 (log₂FC = 8.85, adjusted P = 1.35 × 10⁻⁵¹), AC233263.2 (log₂FC = 7.60, adjusted P = 1.44 × 10⁻¹⁶), and OR7E163P (log₂FC = 6.68, adjusted P = 2.69 × 10⁻⁴⁸). OR7E163P showed a particularly broad tumor-associated distribution, being detected in 80.7% of tumor samples compared with 23.7% of normal samples. AC103563.1 and AC233263.2 were detected in 59.3% and 25.3% of tumors, compared with 9.0% and 5.6% of normal samples, respectively. A complementary pattern of expression loss was observed for GOLGA8CP (log₂FC = −11.53), AC138749.1 (log₂FC = −11.69), and GOLGA2P6 (log₂FC = −13.04), each with adjusted P-values approaching zero. GOLGA2P6 was detected in 100% of normal samples but in only 6.0% of tumors. Similarly, GOLGA8CP and AC138749.1 were detected in approximately 99% of normal samples but in fewer than 13% and 6% of tumors, respectively. These findings reveal two complementary components of the TGCT-associated unprocessed pseudogene profile: tumorassociated acquisition of expression and depletion of transcripts broadly represented in normal testicular tissue.
Conclusion: TGCT is associated with marked bidirectional alterations in unprocessed pseudogene expression relative to normal testicular tissue. The increased expression of AC103563.1, AC233263.2, and OR7E163P, together with the pronounced reduction of GOLGA8CP, AC138749.1, and GOLGA2P6, identifies a focused set of transcripts capturing both tumor-associated expression gains and loss of normal testis-associated transcriptional features. These observations support the potential value of unprocessed pseudogenes as an additional layer of the TGCT transcriptomic landscape. However, the very large tumor–normal differences should be interpreted in the context of the independently generated TCGA and GTEx cohorts. Validation in independent cohorts and experimental studies will be necessary to establish the reproducibility and biological significance of these candidates.