Immune responses in the treatment of drug-sensitive pulmonary tuberculosis with phenylbutyrate and vitamin D₃ as host directed therapy

Tuberculosis remains among the top 10 leading causes of death with global estimates of 10.4 million new cases and 1.4 million deaths in 2015 [1]. Control or exacerbation of TB is dependent on host immune responses generated to combat Mycobacterium tuberculosis (Mtb) infection [2–4]. Several studies have unravelled immunological pathways that influence the outcome of Mtb infection which include cytokine-mediated signalling among T cells, macrophages and neutrophils, and phagocytes-mediated antimicrobial processes [4–9]. Studying cytokine profiles in TB patients has demonstrated its potential for use in diagnostic purposes, monitoring of treatment efficacy and development of novel treatment strategies [10–14]. Autophagy in macrophages and intracellular lysosomal degradation are important for killing of pathogens although Mtb has evolved to escape elimination by blocking phagosomal acidification and phagosome-lysosomal fusion [15–17]. Endoplasmic-reticulum (ER) stress response is triggered by mycobacterial infection and plays a critical role in the pathogenesis of TB [18].

The rise in antibiotic resistance among Mtb in the last decade rekindled attention towards alternative chemotherapies. Host-directed therapies (HDT) have emerged as a promising avenue for adjunctive treatment with the aim to modulate immune responses against the pathogen by targeting clinically relevant host biological pathways. This strategy would be beneficial in reducing the course of antibiotic treatment, preventing the spread of drug-resistant Mtb, and reducing inflammation in the lung (reviewed in [19, 20]). Phenylbutyrate (PBA) is a licensed drug indicated for the management of urea cycle disorders [21]. Vitamin D₃ (vitD₃), a dietary supplement, has diverse immune-modulatory properties. Our group has shown that PBA and vitD₃ have a strong synergistic effect on induction of antimicrobial peptides (AMPs) in lung epithelial cell lines, macrophages, and in healthy human immune cells [22–24]. In a randomized clinical trial, we have shown that PBA alone or combined with vitD₃ is a promising candidate for HDT in the treatment of drug sensitive pulmonary TB by speeding up clinical recovery [25]. Treatment with vitD₃ or the combination with PBA accelerated sputum culture conversion and enhanced expression of LL-37, the human AMP of the cathelicidin family, by immune cells. Moreover, PBA adjunctive therapy increased macrophage-mediated killing of Mtb ex vivo compared to placebo. Our in vitro study further showed that PBA can induce autophagy in a LL-37 dependent pathway and promotes intracellular killing of Mtb in human macrophages [23]. PBA is known to reduce ER stress in cells and thereby reduce inflammatory responses [26, 27]. We hypothesized that the beneficial effects imparted by HDT of PBA and vitD₃ in TB patients may be mediated through regulating expression of cytokines, chemokines and AMPs by immune cells, augmentation of autophagic responses of macrophages and reduction of chronic ER stress. Thus, in a sub-group of TB patients, we evaluated the effect of HDT on dynamics of cytokines and chemokines in culture supernatants of PBMC, HBD1 and ER stress genes and expression of LC3, an autophagy marker, in macrophages from TB patients in response to the disease and clinical improvement.

It is important to evaluate the role of immune-modulatory agents as host directed therapy in the treatment of TB that can improve efficacy of therapy, reduce or prevent detrimental outcomes of toxic drugs and improve patient compliance. Our interest in using PBA and/or vitD₃ as HDT in the treatment of TB was based on their diverse roles in modifying the host defense system. Our findings showed that adjunctive treatment with PBA or vitD₃ over 8 weeks’ time reduced concentration of cytokines/chemokines and endoplasmic reticulum stress in parallel to clinical recovery. All three adjunctive therapies increased frequency of autophagy in macrophages compared to placebo.

TB is a chronic infectious disease with a spectrum of symptoms and presentations, wherein cytokines and chemokines play important roles in manipulation of the immune responses, containment of Mtb, pathogenesis and disease manifestations. Various studies have compared serum cytokine/chemokine profiles of patients with different clinical presentations of TB, to establish the expression levels as early surrogate biomarkers of therapeutic responses, bacteriological confirmation of TB and sputum culture conversion [31–33]. We measured cytokines and chemokines in culture supernatants of PBMC instead of serum or plasma, and the findings were similar to previous studies, showing marked decline of serum cytokine/chemokine concentrations at different treatment stages [14, 33–37]. Particularly we found that compared to conventional anti-TB treatment in the placebo group, cytokine/chemokine reduction was greater in the PBA or vitD₃ groups that occurred in parallel to clinical recovery [25] (reduced TB score), suggesting a role of PBA and vitD₃ in dampening of inflammatory responses.

Mycobacterial infection has been shown to induce ER stress and damage ER. Cells deal with ER stress by activation of the unfolded protein response (UPR), but chronic activation of these pathways can eventually result in apoptosis and lung injury [18, 38–40]. Several biomarkers of ER stress have been used in various studies by assessing activation of the UPR; spliced X-box binding protein-1 (XBP1spl) mRNA has been shown to be a reliable marker in revealing dose-dependent changes in the transcript level of this gene [28]. Mycobacterial secreted antigens such as ESAT-6 has been shown to increase XBP1spl as well as other ER stress genes in epithelial cell lines [38]. Toll-like receptor activation was also shown to increase splicing of the transcription factor XBP1 in the absence of other signs of activation of the UPR, and this was found to regulate innate immune responses in macrophages [41]. PBA is a classic ER stress inhibitor; it is a chemical chaperone that improves ER folding capacity and trafficking of mutant proteins out of the ER, stabilization of protein conformation and thereby reduces accumulation of misfolded proteins in ER lumen. In cell lines, experimental models and tissue samples, PBA has been shown to reduce expression of GRP-78, CHOP and other UPR-related markers such as ATF6, XBP1 and phosphorylation of eIF2α [27, 42–44]. Vitamin D is also considered a natural ER stress reliever [30, 45]. Down-regulation of GADD34 transcripts after 8 weeks of therapy was evident in all 4 arms, however down-regulation of spliced XBP1 was evident only in PBA and vitD₃ arms compared to placebo. Thus, in vivo effect of PBA and vitD₃ adjunct treatments was demonstrated in diminishing ER stress in MDM during TB infection that may have a role in improving clinical outcome. Moreover, diminished inflammatory cytokine responses as we have shown here may be linked to suppression of ER stress [46, 47].

Autophagy is an important physiological process that is applied by macrophages to control and eliminate intracellular pathogens. We have earlier demonstrated in a mechanistic study that PBA, active vitD₃ separately and in combination induced autophagy in macrophages from healthy donors when infected in vitro with Mtb [23], increased the intracellular killing of Mtb compared to untreated MDM and induced LL-37 in cell supernatants of macrophages and lymphocytes after oral ingestion by healthy adults [22]. The novel activity of PBA as an inducer of autophagy was LL-37-dependent. We further showed in the present study that, all three adjunctive therapies (PBA, vitD₃ and PBA + vitD₃) in the clinical trial [25] promoted autophagy in ex vivo infected macrophages from the patients, indicating that the inducers can prime the macrophages for autophagy response. Moreover, the autophagy process occurred in parallel to induction of LL-37 in macrophage/PBMC [25], continuing up to an additional week after the adjunct therapy was completed. MDM-mediated elimination of Mtb in the vitamin D and PBA + vitamin D₃ arms in the trial thus strongly suggest a role of the autophagy process in the eradication. Autophagy also plays a housekeeping role in removing misfolded, unfolded or aggregated proteins and clearing damaged organelles, such as ER and mitochondria. In view of the involvement of the UPR in activating autophagy, it is interesting to note that whereas treatment increased autophagy, it appeared to decrease activation of the UPR to ER stress. This suggests that mechanisms distinct from the UPR are involved in the observed treatment-induced autophagy.

Human beta defensin-1 (HBD1) is produced constitutively by all human epithelia and some immune cells. Monocyte-derived-macrophages and -dendritic cells both express HBD1 mRNA, showing increased expression after activation with IFN-γ and/or lipopolysaccharide in a dose- and time-dependent fashion [48]. When expression of HBD1 transcripts was studied in MDMs from the patients we did not observe any increase or down-regulation of the peptide in the treatment groups compared to placebo during the course of TB disease. Thus, PBA or vitD₃ did not seem to have any modulating effect on HBD1 expression in peripheral MDM as seen with LL-37 [25].

One of the limitations of this study was the small size used per treatment group for evaluation of ER stress genes and HBD1 genes. The expression of transcripts was studied only in macrophages, not other cells in PBMC. This was done because of unavailability of adequate number of matched/paired samples from different time points and suboptimum concentrations of specimens remaining from the clinical trial. Our analysis of activation of the UPR was restricted to measurement of GADD34 and XBP1spl because of limitations in the availability of RNA, and therefore other UPR markers (including protein markers) should be assessed in future studies. Only one autophagy marker LC3 was used instead of a combination of markers such as ATG 5 and Beclin-1; however, our earlier studies showed the LC3 marker as reliable for assessment of autophagy in macrophages [23]. The combined adjunct therapy of PBA and vitD₃ did not exhibit any effect on cytokines/chemokines or ER stress reduction in TB patients and require further investigations. Notably, the combined treatment with PBA and vitD₃ of the patients exhibited the fastest sputum culture conversion [25]. Another limitation was that we measured cytokines and chemokines in saponin-treated cell culture fluid instead of serum or plasma as traditionally done [10–14, 31–33]. Measurement of cytokines in the serum does not accurately reflect active secretion of cytokines due to an ongoing infection; moreover it is affected by clearance from the circulation. Our hypothesis is that the PBMC from TB patients are in vivo activated and they do not require ex-vivo stimulation; release of cytokines/chemokines from unstimulated cells into the culture supernatant indicate active secretion and thus are more relevant to the ongoing disease process. However, use of saponin allowed excretion of intra-cellular cytokines into the culture supernatant which was not a spontaneous release of cytokines.

demonstration of clinically beneficial immunomodulation that occur in parallel to improved treatment outcomes suggest that use of repurposed agents such as Phenylbutyrate and vitamin D₃ can be a valuable strategy in HDT against drug-sensitive TB. The toxicities and poor treatment outcomes of current therapies against MDR-TB necessitate newer approaches to improve the management and control of MDR-TB. Development of host-directed therapies integrated with Mtb-targeted chemotherapy provides a complicated challenge because temporal events of infection and host immunity may play a critical role in determining HDT efficacy. Thus, the therapeutic potential of PBA and vitD₃ HDT against multidrug resistant TB warrants urgent clinical evaluation in well-designed multi-center clinical trials in endemic settings.