People infected with M. tuberculosis and HIV exhibit a higher risk to developing progressive diseases from both pathogens. The complex bi-directional interaction between them turns out to favour a M. tuberculosis-mediated stimulation of HIV-1 replication, and on the other hand HIV infection exerts detrimental effects on the course of TB. Such scenario is further complicated by the appearance of IRIS during HAART leading to a strong recovery of CD4 T cell responses, which may promote an overt TB or exacerbate a previously established disease.
Data about immuno-endocrine disturbances during HIV-TB co-infection are scarce and our paper provides several pieces of new information about the relationship between the HPA, in terms of adrenal steroids, and the immune response mounted by patients undergoing both infections. Extending previous findings on HIV patients 
our study demonstrates that DHEA levels are reduced in HIV-TB and HIV-TB-IRIS, the latter group also showing lower amounts of DHEA-s. Reduced levels of DHEA may be the result of a persistent immuno-inflammatory response in both groups of patients. This is in line with a previous report showing the inhibitory effects of culture supernatants from Ag-stimulated PBMC from TB patients on in vitro DHEA production by adrenal cells 
. The same is likely to apply for the lower amounts of DHEA-s seen in IRIS patients, although some changes in the conversion of DHEA-s into DHEA may be also involved. In this sense, inflammation-related compounds are known to affect DHEA sulphate metabolism 
and HIV-TB-IRIS is characterized by an increased inflammatory response 
Further support for a relative deficiency of the HPA axis during HIV infection is provided by cortisol measurements. In fact, our results showed that HIV patients had rather diminished cortisol levels and HIV-TB and IRIS patients presented modestly increased amounts of this steroid but statistically insignificant in relation to HD. As a result, both groups depicted an increased cortisol/DHEA ratio which may be inefficient for the control of inflammatory reactions, since the slightly increased cortisol levels would not compensate for the anti-inflammatory effect of DHEA 
, because its levels were markedly reduced. Since cortisol is involved with the stress response, whereas DHEA deals with the stress adaptation, the imbalance seen in both groups may reflect a more unfavorable stressful condition which may further impact on the immune responses developed by them.
Our results also expand the knowledge of regulatory influences of adrenal steroids on the immune response. Production of GCs during infections mediates not only anti-inflammatory effects but also favors an immune response that seems to be unsuitable to clear intracellular pathogens 
. The negative correlation between cortisol levels and CD4 counts in HIV-TB individuals and the inhibitory effects of this hormone on IFN-γ production in the same group of patients is in line with such statement. In contrast, DHEA augmented IFN-γ production in the same group of patients which in turn corroborates its Th1 favoring effects 
Dissimilar findings were seen when analyzing correlations. In fact, DHEA correlated positively with CD4 count from HIV-TB patients and the whole population sample, whereas a negative association between DHEA levels and frequency of Treg was found in the former group. This result fits well with the demonstration that DHEA inhibited FoxP3 expression in HIV-TB patients; a finding that provides novel information about the mechanisms by which DHEA modulates the immune response.
In tuberculosis, the potentially protective Th1 cell response may result in an immunopathological response that fails to eliminate the bacteria 
. Within this setting control mechanisms like those mediated by Treg are essential for preventing immunopathological damage without interfering with processes involved in pathogen clearance 
. In our series, Treg were only affected in HIV-LTB patients who showed decreased numbers; the implications on whether it represents a different immuno-inflammatory situation affecting both the differentiation and activity of Treg should be analyzed in future studies.
As regards FoxP3, its expression was augmented in HIV-TB and HIV-TB-IRIS patients, particularly in Mtb
-stimulated CD25+ cells from HIV-TB patients. The transcription factor Foxp3 was originally identified as a master regulator of thymus-derived natural Treg (nTreg) and constitutes the most reliable marker to specify these subsets 
. Because Foxp3 expression in these Treg is critical for the maintenance of Treg-specific functions 
, it may be assumed that HIV-TB and HIV-TB-IRIS patients have an increased Treg activity likely addressed to ameliorate the accompanying immuno-inflammatory response that usually takes place in these kind of patients.
Instead of that, an expansion of Treg with a concomitant reduced functional capacity of suppressor cells and diminished IL-10 secretion has been demonstrated in IRIS patients 
, suggesting an underlying defect in these cells, perhaps due to a decreased FoxP3 expression. Other studies in HIV negative individuals with lung TB underline the potential role of Treg since increased frequencies of both total CD4+CD25+ T cells and CD4+ CD25high
cells were shown during active disease, preferably in the involved lung 
. In the same sense, a higher percentage of circulating CD4+CD25high
T cells together with increased levels of FoxP3 mRNA in PBMC from TB patients was found, even higher at sites of active inflammation and tissue pathology 
In line with the general features of the IRIS, HIV-TB-IRIS patients had a higher IFN-γ production, which among its many activities is known to be essential for the control of M. tuberculosis
. Remarkably, these patients as well as those with HIV-TB also displayed increased levels of FoxP3+ CD25- cells, with reduced numbers of this cell population being found in HIV, HIV-LTB and HD groups. This intriguing cell population is increased in systemic lupus erythematosus patients 
, being its nature under discussion since some authors suggested that these T cells were mostly FoxP3+ CD25- non-Treg 
, whereas others have proposed that these cells were dysfunctional Treg which suppressed T cell proliferation but not IFN-γ production in vitro 
. FoxP3 expressed by Treg is relatively stable whereas that of non-Treg is not 
. Therefore, taking into account our observations on the stability of FoxP3, we believe that these cells are dysfunctional Treg. On the other hand, a recent report described a CD25low
FoxP3+ GITR+ Treg population in peripheral blood from HD with a strong suppressive activity on T cell proliferation, although the authors didn’t investigate the ability of these cells to inhibit cytokine secretion 
. Overall, we propose the analysis of these “unconventional” regulatory T cells in HIV-TB co-infection as an interesting issue for future research in the area of tuberculosis pathogenesis in the context of HIV infection.
As recorded in the case of conventional Treg, DHEA concentrations correlated negatively with FoxP3+ CD25- cells. Adrenal hormones modulate the expression of FoxP3 and the frequency of Treg in an autoimmune setting 
. Thus, we hypothesize that in co-infected patients the increased cortisol to DHEA ratio could upregulate the expression of FoxP3 and therefore the generation of dysfunctional Treg lymphocytes. Our in vitro
observations support the idea of a negative regulation of the FoxP3 transcription factor by DHEA, which was also seen in cells from Addison’s disease patients 
The HPA axis is a major constituent of the neuroendocrine system. Once the magnitude of the defensive response to a microbial agent becomes relevant, activation of the HPA axis follows with significant influences on the immune function 
. Our results associate adrenal steroids with critical components of the immune response against HIV and M. tuberculosis
, and provide a stimulating background for studying more about these links, both for a deeper comprehension of pathogenic mechanisms and the potential development of adjuvant interventions.