(1)H Magnetic Resonance Spectroscopy of live human sperm

STUDY QUESTION: Can (1)H Magnetic Resonance Spectroscopy (MRS) be used to obtain information about the molecules and metabolites in live human spermatozoa? SUMMARY ANSWER: Percoll-based density gradient centrifugation (DGC) followed by a further two washing steps, yielded enough sperm with minimal contamination (<0.01%) from seminal fluid to permit effective MRS which detected significant differences (P < 0.05) in the choline/glycerophosphocholine (GPC), lipid and lactate regions of the (1)H MRS spectrum between sperm in the pellet and those from the 40%/80% interface. WHAT IS KNOWN ALREADY: Current methods to examine sperm are either limited in their value (e.g. semen analysis) or are destructive (e.g. immunohistochemistry, sperm DNA testing). A few studies have previously used MRS to examine sperm, but these have either looked at seminal plasma from men with different ejaculate qualities or at the molecules present in pooled samples of lyophilized sperm. STUDY DESIGN, SAMPLES/MATERIALS, METHODS: Sperm suspended in phosphate buffered saline (PBS) at 37°C were examined by (1)H MRS scanning using a (1)H excitation-sculpting solvent suppression sequence after recovery from fresh ejaculates by one of three different methods: (i) simple centrifugation; (ii) DGC with one wash; or (iii) DGC with two washes. In the case of DGC, sperm were collected both from the pellet (‘80%’ sperm) and the 40/80 interface (‘40%’ sperm). Spectrum processing was carried out using custom Matlab scripts to determine; the degree of seminal plasma/Percoll contamination, the minimum sperm concentration for (1)H MRS detection and differences between the (1)H MRS spectra of ‘40%’ and ‘80%’ sperm. MAIN RESULTS AND THE ROLE OF CHANCE: DGC with two washes minimized the (1)H MRS peak intensity for both seminal plasma and Percoll/PBS solution contamination while retaining sperm specific peaks. For the MRS scanner used in this study, the minimum sperm concentration required to produce a choline/GPC (1)H MRS peak greater than 3:1 signal to noise ratio (SNR) was estimated at ~3 × 10(6)/ml. The choline/GPC and lactate/lipid regions of the (1)H spectrum were significantly different by two-way ANOVA analysis (P < 0.0001; n = 20). ROC curve analysis of these region showed significant ability to distinguish between the two sperm populations: choline/GPC ROC AUC = 0.65–0.67, lactate/lipid ROC AUC = 0.86–0.87. LIMITATIONS, REASONS FOR CAUTION: Only 3–4 semen samples were used to assess the efficacy of each sperm washing protocol that were examined. The estimated minimum sperm concentration required for MRS is specific to the hardware used in our study and may be different in other spectrometers. Spectrum binning is a low resolution analysis method that sums MRS peaks within a chemical shift range. This can obscure the identity of which metabolite(s) are responsible for differences between sperm populations. Further work is required to determine the relative contribution of somatic cells to the MRS spectrum from the ‘40%’ and ‘80%’ sperm. WIDER IMPLICATIONS OF THE FINDINGS: (1)H MRS can provide information about the molecules present in live human sperm and may therefore permit the study of the underlying functional biology or metabolomics of live sperm. Given the relatively low concentration of sperm required to obtain a suitable MRS signal (~3 × 10(6)/ml), this could be carried out on sperm from men with oligo-, astheno- or teratozoospermia. This may lead to the development of new diagnostic tests or ultimately novel treatments for male factor infertility. STUDY FUNDING AND COMPETING INTEREST(S): This work was supported by the Medical Research Council Grant MR/M010473/1. The authors declare no conflicts of interest.