Lottrup, G1; Nielsen, J E1; Maroun, L L2; Møller, L M A4; Yassin, M.1; Leffers, H1; Skakkebæk, N E1; Rajpert-De Meyts, E1
1 Department of Growth and Reproduction, Juliane Marie Centre, Rigshospitalet, The Capital Region of Denmark2 Patologiafdelingen, Diagnostisk Center, Rigshospitalet, The Capital Region of Denmark3 Obstetrics and Gynaecology, Herlev and Gentofte Hospital, The Capital Region of Denmark4 Research Centre for Prevention and Health, FCFS, The Capital Region of Denmark
STUDY QUESTION: What is the differentiation stage of human testicular interstitial cells, in particular Leydig cells (LC), within micronodules found in patients with infertility, testicular cancer and Klinefelter syndrome? SUMMARY ANSWER: The Leydig- and peritubular-cell populations in testes with dysgenesis contain an increased proportion of undifferentiated cells when compared with control samples, as demonstrated by increased delta-like homolog 1 (DLK1) and decreased insulin-like peptide 3 (INSL3) expression. WHAT IS KNOWN ALREADY: Normal LC function is essential for male development and reproduction. Signs of LC failure, including LC micronodules, are often observed in patients with reproductive disorders. STUDY DESIGN, SIZE, PARTICIPANTS: In this retrospective study, a panel of markers and factors linked to the differentiation of LCs was investigated in 33 fetal and prepubertal human specimens and in 58 adult testis samples from patients with testicular germ cell tumours, including precursor carcinoma in situ (CIS), infertility or Klinefelter syndrome. PARTICIPANTS/MATERIALS, SETTING, METHODS: The expression patterns of DLK1, INSL3, chicken ovalbumin upstream promoter transcription factor 2 (COUP-TFII), cytochrome P450, family 11, subfamily A, polypeptide 1 (CYP11A1) and smooth muscle actin (SMA) were investigated by immunohistochemistry and quantitative RT-PCR. The percentage of positive LCs was estimated and correlated to total LC numbers and serum levels of reproductive hormones. MAIN RESULTS AND THE ROLE OF CHANCE: DLK1, INSL3 and COUP-TFII expression changed during normal development and was linked to different stages of LC differentiation: DLK1 was expressed in all fetal LCs, but only in spindle-shaped progenitor cells and in a small subset of polygonal LCs in the normal adult testis; INSL3 was expressed in a subset of fetal LCs, but in the majority of adult LCs; and COUP-TFII was expressed in peritubular and mesenchymal stroma cells at all ages, in fetal LCs early in gestation and in a subset of adult LCs. CYP11A1 was expressed in the majority of LCs regardless of age and pathology and was the best general LC marker examined here. SMA was weakly expressed in peritubular cells in the fetal and infantile testis, but strongly expressed in the adult testis. In pathological testes, the numbers of DLK1-positive interstitial cells were increased. The proportion of DLK1-positive LCs correlated with total LC numbers (R = 0.53; P < 0.001) and was higher in testis with enlargement of the peritubular layers (P < 0.01), which was also highly associated with DLK1 expression in the peritubular compartment (P < 0.001). INSL3 expression was absent in some, but not all LC micronodules, and in the majority of LCs, it was mutually exclusive of DLK1. LIMITATIONS, REASONS FOR CAUTION: The number of samples was relatively small and no true normal adult controls were available. True stereology was not used for LC counting, instead LCs were counted in three fields of 0.5 µm(2) surface for each sample. WIDER IMPLICATIONS OF THE FINDINGS: The population of LCs, especially those clustered in large nodules, are heterogeneous and comprise cells at different stages of differentiation. The study demonstrated that the differentiation and function of LCs, and possibly also peritubular cells, are impaired in adult men with testicular pathologies including testis cancer and Klinefelter syndrome. STUDY FUNDING/COMPETING INTERESTS: This work was funded by Rigshospitalet's research funds, the Danish Cancer Society and Kirsten and Freddy Johansen's foundation. The authors have no conflicts of interest.
Human Reproduction (oxford, England), 2014, Vol 29, Issue 8, p. 1637-50