Construction and Application of Testicular Organoid Models in Multi-Omics-Driven Male Reproductive Toxicity Assessment: A Review
- 1School of Life Science and Technology, Shandong Second Medical University, Weifang 261053, China
Abstract
Conventional male reproductive toxicity testing relies on rodent bioassays and two-dimensional cultures, which are limited by species extrapolation, low throughput, and weak mechanistic resolution. Testicular and seminiferous-tubule organoids partially restore the Sertoli–Leydig–peritubular myoid niche and permit serial, non-terminal sampling, making them compatible with multi-omics profiling in a way earlier models were not. This review examines the field across four layers: the biological hierarchy and regulatory pathways that constrain model design; organoid construction strategies together with computational simulation of morphogenesis and mass transport; multi-omics integration algorithms and toxicity feature extraction; and retrospective validation on publicly deposited datasets. We compare cell sourcing, scaffold, and culture configuration routes against tubule formation and meiotic completion, derive critical-radius and shear-stress criteria governing construct viability, and evaluate similarity network fusion, latent factor decomposition, joint matrix factorization, and graph neural network approaches. Retrospective modeling indicates that adding epigenomic input to transcriptomic features yields the largest performance gain, that gradient-boosted trees outperform deep networks at current sample sizes, and that attribution profiles align with established mechanisms. We contribute an integration framework dimensioned for testicular organoid data, a fully open-source reproducible pipeline, and an explicit mapping between quantitative organoid descriptors and toxicity endpoints. Persistent obstacles include incomplete structural reconstitution biasing omics signal, sparse and inconsistently annotated public data, simulation parameter uncertainty, and unverified cross-species transfer. Minimum reporting standards, a shared characterization vocabulary, and an open benchmark platform are proposed as prerequisites for regulatory uptake.
Related
- Retinal organoid–based evaluation of icariin-loaded extracellular vesicles for hypoxia-induced retinal injury via mitochondrial modulation
- SRY-negative 46,XX testicular disorder of sex development: Endocrine, histologic, and molecular characterization.
- Liquid‐Suspendable Metabolic Sensors Allow Real‐Time Oxygen Consumption Monitoring in Caco‐2 and Colon Organoid‐Based Gut‐on‐a‐Chip Models
- Single-cell multi-omic atlas and morphogen screening informs midbrain and hindbrain organoid engineering
- Automated spheroid and organoid image analysis from rule-based methods to deep learning
BibTeX
@article{sun2026construction,
title = {{Construction and Application of Testicular Organoid Models in Multi-Omics-Driven Male Reproductive Toxicity Assessment: A Review}},
author = {Yuting Sun and Yan Liu and Meina Xie},
journal = {International Journal of Molecular Sciences},
year = {2026},
month = nov,
doi = {10.3390/ijms27219963},
url = {https://www.mdpi.com/1422-0067/27/21/9963},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Construction and Application of Testicular Organoid Models in Multi-Omics-Driven Male Reproductive Toxicity Assessment: A Review
AU - Yuting Sun
AU - Yan Liu
AU - Meina Xie
PY - 2026
DA - 2026/11/06
JO - International Journal of Molecular Sciences
DO - 10.3390/ijms27219963
UR - https://www.mdpi.com/1422-0067/27/21/9963
AB - Conventional male reproductive toxicity testing relies on rodent bioassays and two-dimensional cultures, which are limited by species extrapolation, low throughput, and weak mechanistic resolution. Testicular and seminiferous-tubule organoids partially restore the Sertoli–Leydig–peritubular myoid niche and permit serial, non-terminal sampling, making them compatible with multi-omics profiling in a way earlier models were not. This review examines the field across four layers: the biological hierarchy and regulatory pathways that constrain model design; organoid construction strategies together with computational simulation of morphogenesis and mass transport; multi-omics integration algorithms and toxicity feature extraction; and retrospective validation on publicly deposited datasets. We compare cell sourcing, scaffold, and culture configuration routes against tubule formation and meiotic completion, derive critical-radius and shear-stress criteria governing construct viability, and evaluate similarity network fusion, latent factor decomposition, joint matrix factorization, and graph neural network approaches. Retrospective modeling indicates that adding epigenomic input to transcriptomic features yields the largest performance gain, that gradient-boosted trees outperform deep networks at current sample sizes, and that attribution profiles align with established mechanisms. We contribute an integration framework dimensioned for testicular organoid data, a fully open-source reproducible pipeline, and an explicit mapping between quantitative organoid descriptors and toxicity endpoints. Persistent obstacles include incomplete structural reconstitution biasing omics signal, sparse and inconsistently annotated public data, simulation parameter uncertainty, and unverified cross-species transfer. Minimum reporting standards, a shared characterization vocabulary, and an open benchmark platform are proposed as prerequisites for regulatory uptake.
SN - 1422-0067
ER -