True polyploid meiosis in the human male

Abstract

Polyploidy does not usually occur in germinal cells of mammals and other higher vertebrates. We describe a unique example of mosaic autotetraploidy in the meiosis of a human male. Although the original observations were made in the late 1960s, we did not publish them at that time, because we expected to detect further examples that could be described together. However, this did not occur and we have now decided to make the observations available to demonstrate that polyploidy in mammalian male meiosis can arise at a higher frequency than expected by random polyploidization of individual meiotic cells, by either DNA duplication or cell fusion prior to synapsis. This is the first description of a population of primary spermatocytes exhibiting multivalent formation at leptotene /diakinesis in human spermatogenesis, with ring, chain, frying pan and other types of quadrivalents, typical of autotetraploidy. As many of the polyploid configurations showed apoptotic breakdown, it is likely that diploid and/or aneuploid spermatozoa would have rarely or never resulted from this mosaic autotetraploid meiosis.

Figure 1. Diploid diakinesis with extremely low chiasma count of 38. The patient exhibited a much lower average chiasma count than usually observed in human males (range 46 – 49) with either diakinesis or SC analysis. Could this have resulted from his extremely advanced age? Several other studies in cohorts of younger males have not detected a significant age-related reduction in meiotic recombination frequency.

Figure 2. Contrast enhancement to distinguish selected multivalents within an auto-tetraploid diakinetic spermatocyte. Positions of incompletely terminalised chiasmata are indicated by black arrows and the presumed position of completely terminalised chiasmata, by white arrows. Quadrivalents 2, 3, 4 and 5 are rings, quadrivalent 8 is a chain, quadrivalent 1 has a classical frying pan structure and quadrivalents 6 and 7 are intermediates. Other quadrivalents are certainly present, but were either too small or lacked interstitial incompletely terminalised chiasmata to be so identified.

Figure 3. 4n diplotene with arrowed quadrivalents.

Figure 4. Apoptotic 4n diakinesis with arrowed quadrivalents.