Abstract
Fertility problems affect many couples. Research into male fertility commonly uses mouse models due to their availability and similar spermatogenesis to humans. A common target is the seminiferous tubules, the site of spermatozoa production, typically studied using biopsies and histological analysis. High-field Magnetic Resonance (MR) may offer a non-invasive alternative to investigate testicular function in infertility models. Here, we assess seminiferous tubules structure in sacrificed mice to determine the usefulness of MR compared to histology. Twelve mice (11 aged 35-57 days, one >9 months) were sacrificed and MR imaged at 9.4T with a Rapid Acquisition with Relaxation Enhancement sequence. Testes were scanned in situ for all mice, and excised in a subset of mice (n=4). A second subset of mice (n=4) had their testes selected for histological analysis. Seminiferous tubule diameter was measured manually from MRI and histology images. Custom image analysis scripts were created for the automated segmentation of seminiferous tubules and calculation of tissue volumes. All ex vivo and in situ images of testes exhibited clear outlines of seminiferous tubules. Ratio of total testis volume to volume of seminiferous tubules did not differ significantly between ex vivo and in situ measurements, and were similar in mature and younger mice. Both total testis volume and seminiferous tubule volume were larger in the mature animal. While histological slices trended towards larger average seminiferous tubules diameter than MRI images, we found no significant differences between MRI and histological measurements. High-field MRI can be used in a mouse model to assess testicular structure in situ. All volumetric measurements compared favourably with histological data. In situ scans also clearly showed identifiable extra-testicular tissues, such as epididymis and prostate tissues. The potential to image tissues associated with sperm maturation as well as spermatogenesis emphasises how MR could be a useful technique in mouse models of fertility, however further work is required to optimize tissue segmentation and validate this method for use in longitudinal studies. This type of measurement could be extended to human fertility studies in the future.
