Ursolic and oleanolic acids as antimicrobial and immunomodulatory compounds for tuberculosis treatment

At present, Tuberculosis (TB) is the only infectious disease considered by the World Health Organization (WHO) as a health emergency worldwide, because it causes nearly 2 million deaths annually [1]. TB is more frequent in developing countries and its association with human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) renders its control more difficult. In addition, the emergence of multidrug-resistant tuberculosis (MDR-TB, defined as those TB strains simultaneously resistant at least to rifampin and isoniazid) and extensively drug resistant tuberculosis strains (XDR-TB) threaten the success of the directly observed therapy short course (DOTS) and DOTS-Plus treatment programs established by the WHO [2]. Despite all the progress achieved, only one third of patients with TB receive adequate treatment; in the case of MDR, few patients have received the DOTS-Plus regimen and only about 70% of MDR-TB cases respond to the current treatment [1, 2].

Since the release of rifampicin in 1976, only rifabutin and rifapentin have been approved for TB treatment; unfortunately, these drugs are not yet widely distributed [3]. At present, a number of drugs are under investigation, but only a few compounds are found in preclinical and clinical evaluation (about 10 compounds) [1‐4]. Thus, there is an urgent need to discover new antituberculous agents that are effective in the treatment of MDR cases and also novel agents that can shorten the long conventional chemotherapy in drug-sensitive TB. Within this context, not only new synthetic drugs, but also natural products from medicinal plants are potential sources of new anti-mycobacterial products.

Chamaedora tepejilote (C. tepejilote) and Lantana hispida (L. hispida) are widely distributed plants in Mexico known as “tepejilote, palmita or palma camaedor” and “cinco negritos or verbena” respectively, both plants have been used in Mexican traditional medicine. Some of their common uses include the treatment of respiratory complaints such as cough, bronchitis, colds and pneumonia [5]. We have previously reported that the hexanic fractions from these plants had in vitro antimycobacterial activity and their bioguided fractionation showed that the triterpenic compounds ursolic acid (UA) and oleanolic acid (OA) were the specific agents involved in this activity [6‐8]. This effect has been confirmed by other authors [9‐11]. These triterpenic acids also have antibacterial [12, 13], antiviral [14], antiparasitic [13], antioxidant [15] and antitumoral activities [16], as well as hepatoprotector [17] and gastroprotector [18] effects. Interestingly, UA enhances the production of nitric oxide (NO) and tumor necrosis factor alpha (TNF-α) by activating nuclear factor-kappaB (NF-κB) in mouse macrophages [19, 20] and blocking transforming growth factor-beta 1 (TGF-β1) activity [21, 22]. The stimulation of NO and TNF-α contributes to their immunoregulatory and antitumoral effects, and could be significant in an immunotherapeutic agent against M. tuberculosis. In this study, we report the in vitro antimycobacterial activity of UA and OA isolated from the hexanic extract of the aerial parts of C. tepejilote and L. hispida, against the reference drug-sensitive M. tuberculosis strain H37Rv, monoresistant H37Rv strains, several MDR clinical isolates and a group of nontuberculous mycobacteria. The antitubercular activity of both compounds was then confirmed in a well-characterized murine model of progressive pulmonary TB. Our results show therapeutic activity attributable to a combination of bactericidal and immunotherapeutic effects.

UA (3β-hydroxy-urs-12-en-28-oic-acid) and its isomer, OA (3β-hydroxy-olea-12-en-28-oic acid) are triterpenoids compounds that are widely distributed in the plant kingdom, in medicinal herbs, and are a common component of the human diet [27]. There are comprehensive reports on their biological activities and beneficial effects in various diseases, including infectious diseases [16, 27]. In this regard, there are several reports of their significant antimycobacterial activity when they are primarily purified from diverse plants [9, 11, 28]. Indeed, the present study comprises part of a research program that involves an ethnopharmacological screening of Mexican medicinal plants in a search for activity against M. tuberculosis. Our previous studies showed that UA and OA were in part responsible for the antimycobacterial activity from L. hispida (a widely distributed, ornamental Mexican plant used in folk medicine to treat TB) and C. tepejilote (a tropical plant from Southern Mexico used to treat respiratory diseases) [6, 8]. The results presented here confirm and extend these studies, showing that purified UA and OA have in vitro antimycobacterial activity against fully drug-sensitive and monoresistant H37Rv strains, as well as several MDR clinical isolates and to a lesser degree, non-tuberculous mycobacteria. Our results on the in vitro activity of UA against M. tuberculosis H37Rv were similar to those reported previously, with a MIC value of 50 μg mL^-1 when evaluated by the radiorespirometric Bactec 460, and 31.0 and 41.9 μg mL^-1 by MABA assay; while MIC values reported for OA were 50 μg mL^-1 when tested by the radiorespirometry method and 30.0, 28.7, and 25 μg mL^-1 by MABA [9, 29‐32]. Both triterpenic acids exhibited less activity against non-tuberculous mycobacteria, with the MIC value of 100 μg mL^-1. This is in fact modest antimycobacterial activity. However, one major point of traditional medicine is the use of herb mixtures, which could be more effective than a single product for producing the desired effects [28]. UA and OA are isomers, and our results showed that the combination of both produced in vitro intracellular and in vivo synergistic effects. Although the molecular mechanism of the antimycobacterial activity has not yet been determined, it has been proposed that UA and OA can produce significant abnormalities in the bacterial cell wall [9, 13]. Both terpenoids have efficient antilipidic activity on eukaryotic cells [33], and perhaps this activity can also affect mycobacteria producing damage on the complex cell envelope, which is rich in lipids.

Mycobacterial infections are controlled by the activation of macrophages through type 1 cytokine production by T cells [34‐36]. IFN-γ and TNF-α are essential for this process because they promote macrophage activation and iNOS expression. This is clearly observed in our BALB/c mouse model, which is based on infection via the trachea with a high dose of M. tuberculosis H37Rv [26, 37]. In this model, there is an initial phase of partial resistance dominated by Th1 cytokines plus TNF-α and the expression of iNOS, followed by a late phase of progressive disease after 1 month of infection, characterized by a lower expression of IFN-γ, TNF-α, or iNOS, progressive pneumonia, extensive interstitial fibrosis, high bacillary counts and very high levels of immunosuppressive factors such as TGF-β1 and Prostaglandin E-2 (PGE2) [26, 37, 38]. This BALB/c tuberculosis model has been used extensively to test different forms of therapy [39‐41], thus confirming that it is highly suitable for exploring the efficiency of new natural drugs or immunotherapy based on the airway infection route, which is the most common pathway of infection in humans and the highest rate of bacterial multiplication in the lung correlates with the extent of tissue damage (pneumonia) and death of infected animals [26].

Although contrasting differences in immune responses have been observed that depend on triterpenic concentrations and the biological status of the target cells used in different experimental systems [42], it has been reported that UA and OA stimulate IFN-γ production [43], and also upregulate iNOS and TNF-α expression through NF-kB transactivation in murine resting macrophages [19, 20]. More recently, it has been demonstrated that UA modulates human dendritic cells via activation of IL-12, polarizing the Th-1 response [44]. Tuberculous animals treated with both triterpenic acids showed a higher expression of IFN-γ, TNF-α, and iNOS than non-treated control animals, or even than sick mice successfully treated with conventional chemotherapy, suggesting that UA and OA exert an effect as immunostimulating factors that can restore the protective antimycobacterial cytokine pattern during advanced disease, producing a significant decrease of bacillus loads and tissue damage.

Suppression of T-cell responses to mycobacterial antigens is a consistent feature of TB [45], and in vitro and in vivo observations indicate that TGF-β participates in these effects [46‐50]. It is well established that M. tuberculosis and its components are efficient inducers of the TGF-β1 production by macrophages and this cytokine is a significant factor in the suppression of cell-mediated immunity (CMI) and in the induction of fibrosis [49]. Another molecule that is also produced in high amounts during progressive TB and has CMI suppressing activities is PGE-2. In fact, TGF-β and PGE2 share many immunomodulatory functions, such as the inhibition of IFN-γ, interleukin 2 (IL-2) and IL-12 production [50, 51] and macrophage deactivation, suppressing TNF-α and iNOS production [52, 53]. We have shown, in this experimental model of pulmonary TB, that by blocking TGF-β activity by the administration of its soluble receptor type 3 or betaglycan, while simultaneously suppressing PGE-2 production by the administration of niflumic acid, a specific cyclooxygenase type 2 (COX-2) blocker, we can produce a significant therapeutic benefit associated with restoration of the protective cytokine pattern (41). Interestingly, UA and OA antagonize TGF-β1 activity by blocking the binding of its specific receptor [21, 22], which is the same function as the soluble receptor type 3 or betaglycan. Moreover, UA and OA also suppress prostaglandin production by blocking the binding of c-Jun to the response element of the COX-2 promoter, thus preventing the transcription of this enzyme [54], or by irreversible inhibition of secretory phospholipase A2 [55, 56]. Thus, the restoration of the protective cytokine pattern observed in animals treated with UA or OA could be attributable to the modulating effect that they have on TGF-β and COX-2 activity. However, there are published reports indicating contrary activities that are receptor- and mouse strain- dependent [57]. Thus, as mentioned previously [58] with respect to the control of the inflammatory response, these triterpenoids can have both positive and negative effects, and further evaluations of their effect on the biological status of target cells or tissues in health and disease are necessary.

It is noteworthy that to date, there are no studies that describe the antituberculous effect of the pure compounds of medicinal plants. Thus, to our knowledge, this study constitutes the first that focuses on evaluating the antituberculous activity in vivo of this type of compound. Moreover, we recently published data on the toxicity in-vivo of the UA/OA mixture; it is practically innocuous when evaluated by the s.c. route in acute (LD₅₀ > 300 mg kg^-1) and subacute (13 mg kg^-1 repeated-dose during 28-day) assays in BALB/c mice [59]. Therefore, these compounds can be considered potential candidates to the treatment of TB.

Our results showed that UA and OA obtained from medicinal plants used in Mexican traditional medicine to treat TB have modest antimycobacterial and some immunoregulatory activities that permit the control of pulmonary TB in mice, indicating that research on natural products can produce novel antibiotic and/or immunotherapeutic agents useful for the treatment of this significant infectious disease.