In Fan et al. (7), it was observed that use of Amhr2^cre/+ knock-in mice (8) to generate Leydig cell-specific Tspo^cΔ/Δ mice resulted in low Mendelian ratios for homozygous cre positive mice (HO: Amhr2^cre/+Tspo^cΔ/Δ). This was interpreted as partial preimplantation embryo loss, and the authors concluded that TSPO is crucial for viability. This is a fundamental mistake because both Amhr2 and Tspo are in the same chromosome 15 and, therefore, cannot assort independently. The Tspo and Amhr2 genetic positions are just 18.18 cM apart, and the probability for chromosomal crossover between the two loci to get HO mice is calculated as 7.6% [based on Haldane (9), other estimates are similar]. Therefore, the low rate of 4.4% HO mice observed by Fan et al. (7) is anticipated and represents the precise biological value due to linkage of the two loci and is certainly not an indication of embryonic lethality (Figure 1) [Note: Actual values are not identical to calculated numbers because of differences in mouse strain, secondary regulation, specific chromosome structures, and interference at crossover sites]. Expectation of 25% HO mice based on classical Mendelian principles is not applicable in this context, and the interpretation made on this basis is highly inaccurate.

To explain their proposed case of embryonic lethality, Fan et al. (7) proposed that the Amhr2^cre/+ knock-in mice used to generate gonadal cell type-specific conditional deletions in more than 90 publications (MGI ID: 3042214), in their particular case, could induce global Tspo deletions. The justification provided was published microarray datasets that seemed to detect (with inconsistencies) an increase in Amhr2 transcription in 2-cell embryos. There was no primary data in the manuscript validating this assertion that appears highly unlikely. Even if it were to occur, global Tspo deletions would not affect the linkage and rate of HO mice as described above. In previous studies that used cre on different chromosomes, Mendelian ratios were observed during generation of viable global Tspo^−/− mice with similar mouse backgrounds (2, 3).

In Fan et al. (7), the Nr5a1^creTspo^cΔ/Δ mice showed expected Tspo deletions in Leydig cells and the adrenal cortex. Testosterone production (both baseline and after induction using hCG) was not affected, consistent with the previous report (1). Although the authors note this as “surprising,” the significance of this observation as indication that TSPO was not involved in mitochondrial cholesterol import in Leydig cells was disregarded. It is not our intention to criticize, but we are indeed under obligation to point out that the proclaimed landmark in vitro studies of Tspo disruption (10) and Tspo knockdown (11) used as foundations for asserting TSPO link to steroidogenesis were performed only using Leydig cells by this same research group. These in vitro results have since not been reproducible both in vitro (2, 4) and in vivo (1, 2) and is now also invalidated by Fan et al. (7) without an explanation.

In Nr5a1^creTspo^cΔ/Δ mice, baseline corticosterone levels were also unaffected, consistent with the previous report (2). However, when induced with ACTH, circulating corticosterone did not increase in both heterozygous (Tspo^cΔ/+) and homozygous (Tspo^cΔ/Δ) deletions of Tspo compared to Tspo^fl/fl controls, an observation that had no correlation to TSPO expression levels and was in contrast to the previous report (2). This inconsistency and differences observed with regard to lipid accumulation and changes to Lhcgr and Scarb1 expression in Nr5a1^creTspo^cΔ/Δ adrenals (7) may be linked to parallel findings showing that TSPO can affect lipid metabolism in cells (12). Although additional work is necessary, by attempting to provide explanations based on the unsubstantiated conjecture that TSPO is an essential cholesterol transport protein for steroidogenesis, the authors have missed an excellent opportunity to advance understanding of TSPO function.

Global Tspo^−/− mice are viable (2, 3) and are an excellent tool for investigating TSPO function in health and disease. Evidence refuting TSPO link to steroidogenesis is not based only on reports in Tspo^−/− mice (2, 3). Recent studies, using Tspo-deficient isolated mitochondria (3), in vitro Tspo knockdown (2), and CRISPR/Cas9-mediated Tspo knockout in cell lines (4), have all highlighted problems with reproducibility of prior work. The serious misinterpretations made in Fan et al. (7) are negatively impacting research progress across multiple fields and distracting from pursuit of the core function of TSPO.

All authors listed have made substantial, direct, and intellectual contribution to the work and approved it for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.