This study demonstrates statistical associations between immune system activation and androgen levels in infertile women with normal FOR and absence of both in association with POA/OPOI and age-dependent DFOR. The results, however, do not necessarily demonstrate the associations we expected. Data confirmed our earlier report that women with POA/OPOI and physiologic DOR, in principle, demonstrate lower testosterone levels than control patients with normal OR . The study, however, also demonstrated higher testosterone levels with than without immune system activation. Moreover, amongst women with immune system activation, androgen levels were higher in women with normal FOR (controls) than in those with DFOR (I.e., POA/OPOI and physiologic DOR). Indeed, POA/OPOI and DOR patients, whether demonstrating signs of immune activation or not, demonstrated similar androgen levels as immunologically negative controls (Figure 2).
These observations suggests that in most young women with normal FOR a degree of immune activation exists, associated with normal to mildly elevated T levels (Figure 2), while in association with low FOR, whether due to POA/OPOI or physiologic DOR, T levels are lower in absence of evidence of immune system activation. Immune system activation, thus, appears in some fashion associated with normal to mildly elevated androgens, while in absence of such immune system activation androgens plunge.
This observation may be interpreted as suggesting that immune system activation is associated with presence of an androgen-producing factor (APF), an in itself interesting observation, considering that induction of tolerance with implantation of the paternal semi-allograft, unquestionably, requires some form of maternal immune system activation. Considering the relatively small number of patients in each of the three studied patient populations, our data, however, of course, require further confirmation.
Androgens are generally believed immunosuppressive [23, 24], though also have been suggested to have an immuno-modulatory effect, either immuno-enhancing or -suppressive . Assuming immune system-induced increases androgen levels, then such immuno-modulatory effects of androgens may be able to feed back, and self-control androgen production by adrenals and/or ovaries.
Anti-adrenal autoimmune responses in Addison’s disease and autoimmune ovarian insufficiency are, indeed, similar in that both are IgG1 dominated and predominantly of Th 1 type . Albertini, only recently, described the ovary, “as something of an immunological hotspot,” since many genes, recently implicated in ovarian aging, are associated with immune pathways .
Here presented data, however, further suggest that in women with POA/OPOI and physiologic DOR the immune system’s ability to produce APF is lost. As a consequence, T levels are low (Figure 2). Two possible explanations come to mind: either the production of APF is interrupted or the effects of APF are blocked. Our results more likely suggest diminution of APF production, since blockage of APF, likely, would have to involve evidence of increased immune system activities, the opposite of what was observed.
Assuming such an androgen-regulation process in adrenals and ovaries, one also has to assume the possibility of APF overproduction. Such overproduction then could be expected to result in excessive androgen levels and, hypothetically, an ovarian PCOS- like phenotype Such phenotypes, of course, can be either hyper- or normo-androgenic . PCOS and DOR may, therefore, represent opposing extremes of, possibly, immune-mediated effects on adrenals and/or ovaries [9, 13].
We are not the first to suggest such an autoimmune etiology for at least some cases of PCOS [29, 30]. Interestingly, González et al. just recently reported that hyperandrogenism apparently exerts anti-inflammatory effects on women with PCOS . Androgen effects on early stages of follicle maturation are AR-mediated . AR-activation in benign prostatic hyperplasia was only recently demonstrated produce anti-inflammatory effects .
Our findings raise a multitude of interesting follow up questions, with the first being how an immune process, selectively, can be associated with higher androgen levels? This study does not offer an answer, but adrenals produce the majority of a woman’s DHEA, the basic substrate for T, in turn produced in ovaries . As in association with PCOS , at least a portion of observed testosterone levels, therefore, can be expected to be consequence of adrenal processes.
Immune-driven stimulation of androgen production in the zona reticularis of adrenals is conceivable. Immune-driven endocrine processes are often antibody-mediated , and even organ-specific anti-glandular autoimmunity is now increasingly recognized associated with systemic autoimmune effects . In association with Addison’s disease, an autoantibody to 21-hydroxylase, a cytochrome P450 enzyme, is diagnostic . Autoantibodies to other adrenal steroidogenic enzymes have also been described [37, 38]. The practically complete absence of anti-21-hydroxylas antibodies in here presented patients (unpublished data), however, suggests that in this study observed immune activation is distinct from Addison’s disease.
Functional autoantibodies can be either suppressive or enhancing and have also been described in non-endocrine autoimmune diseases . This led us to suggest that such regulatory autoantibodies may represent a more frequent characteristic of autoimmune diseases than is generally appreciated . This study now allows for the possibility of an autoantibody network, which controls androgen production by the adrenal zona reticularis of and/or the ovarian theca, representing APF.
Women are especially prone to autoimmune diseases, especially towards autoimmune endocrinopathies and combinations of multiple autoimmune endocrinopathies, autoimmunity towards the thyroid gland being the most frequent .
We recently reported specific genotypes and sub-genotypes of the FMR1 gene, associated with distinct ovarian aging patterns. The so-called het-norm/low sub-genotype appears associated with an ovarian PCO-like phenotype at young age, which rapidly depletes follicles, leading to early DFOR often at relatively young ages . Whether the early stages of PCO-like phenotype are associated with hyperandrogenism is unknown but the same sub-genotype of FMR1 was recently also demonstrated to convert DHEA to TT less efficiently than other FMR1 genotypes did . The specific reason why women with het-norm/low FMR1 convert so poorly is unknown but it is interesting to note that this sub-genotype is also highly associated with autoimmune risk, while its counterpart, the het-norm/high sub-genotype, is protective against autoimmunity .
In conclusion, this study reports first supportive evidence for a, possibly, immune system-associated androgen production process, likely primarily located in adrenals and/or ovaries. We hypothesize that this process, in analogy to immune processes in other endocrine organs, may be autoantibody-driven. For the interested reader a recent review on the impact of endocrine autoimmune diseases on female fertility offers additional insights .