Measurement of membrane integrity in canine spermatozoa using a fluorescent computer-assisted spermatozoal...

The aim of this study was to compare measurement of spermatozoal membrane status using computer-assisted spermatozoal quantification (CASQ) and eosin-nigrosin (EN) staining with manual counting after CFDA/PI staining. Analysis was performed on both fresh and thawed cryopreserved canine semen. Membrane-disrupted spermatozoa (MDS) were counted using CASQ (n = 311) in an untreated sample and a completely membrane-disrupted sample, and the percentage of membrane-intact spermatozoa (MIS) calculated: (Total count - Untreated sample count) ÷ Total count × 100. Spermatozoa were stained with a one-step EN stain (n = 501), and then, at least 100 spermatozoa were manually examined under ×1,000 magnification and classified as MDS (stained with eosin) or MIS (non-stained). Spermatozoa from the same samples were also stained with CFDA/PI, and then, at least 200 spermatozoa were manually examined under ×1,000 magnification and classified as MIS (completely stained by CFDA) or MDS. The percentage of progressively motile spermatozoa (PMS) was determined by both computer-assisted semen analysis (CASA) and subjective methodologies, and the data were subsequently analysed to measure the agreement between the CASQ and EN methods with the CFDA/PI technique using Bland-Altman methodology. Pearson's correlation was measured between the MIS and PMS percentage samples and correlation coefficients compared. The mean MIS percentage was lower for CASQ and higher for EN than in CFDA/PI for all comparisons. The agreement of MIS percentage between CASQ and CFDA/PI was -20.2% to 32.0%, and between EN and CFDA/PI was -32.9% to 14.9%. In all methods, the MIS and PMS percentages were correlated (p < .001). Measurement of CFDA/PI appeared to be the most reliable and accurate method of determining MIS percentage in dogs. Further investigation is required to determine whether the CASQ technique can be improved. Eosin-nigrosin staining also appeared to be unreliable at MIS <80% and overestimated the MIS percentage.