2-4 Treating tuberculosis by targeting the Mycobacterium

2-4 Treating tuberculosis by targeting the Mycobacterium tuberculosis proteasomeby Corine K. Lau, Ph. D. IntroductionMycobacterium tuberculosis and Its ProteasomeWhat causes tuberculosis (TB) in humans? A bacterium called Mycobacterium tuberculosis (M. tuberculosis) is the culprit. Without proper treatment, TB can be fatal. Astoundingly, more than one third of the world population is infected with M. tuberculosis1. The majority of infected individuals are carriers who do not feel sick because their immune system keeps the bacteria dormant in the body. If other diseases weaken the immune system, however, M. tuberculosis can quickly attack the body. Current treatment of active TB involves a combination of four antibiotics for at least six months1. Prolonged treatment or incorrect dosage can cause M. tuberculosis to become resistant to the drugs, rendering the treatment ineffective. How can we make anti-TB drugs more effective? Many antibacterial drugs are designed to inhibit protein synthesis by the bacteria, but these drugs also negatively affect the human host cells. In recent years, researchers have found that an essential protein degradation machinery—called the proteasome—in M. tuberculosis is necessary for the bacterium to survive and thrive in experimental systems2, 3. The overall structure of the proteasome is highly conserved from archaea to eukaryotes. Could we use existing proteasome inhibitors to treat TB? The answer is not simple. Although these drugs are already in use to treat human tumors4, they also have toxic effects on the proteasomes of human host cells. In fact, some well-known proteasome inhibitors (e. g. , peptidyl epoxyketones, peptidyl aldehydes, γ-lactam-Β-lactones, and peptidyl boronate bortezomib [Velcade]) inhibit mammalian proteasomes more potently than the M. tuberculosis proteasome4, 5, 6. Clearly, these proteasome inhibitors are not ideal for treating TB. Here we’ll explore how Carl Nathan’s lab focused on the differences between the M. tuberculosis and human proteasomes to identify inhibitors that specifically bind to and inhibit the M. tuberculosis proteasome without affecting the proteasome of human host cells. Differences between Prokaryotic and Eukaryotic ProteasomesWhat functions do proteasomes carry out in cells, and how does their structure affect their function? Proteasomes degrade unwanted proteins in cells. As discussed in your textbook in Section 23. 1, three specialized enzymes (called E1, E2, and E3) recognize and sequentially add a chain of ubiquitin molecules to the doomed protein (Figure 1)7, and the ubiquitin chain signals the 26S proteasome to degrade the protein. The 26S proteasome is a multisubunit, barrel-shaped cellular protease that consists of a 20S core particle bound by 19S cap particles at each end. The 19S cap removes ubiquitin from the substrate for recycling, unfolds the substrate, and feeds it into the 20S core for degradation. Ubiquitin-mediated proteolysis plays a critical role in cell signaling, cell cycle control, and a variety of other cellular processes. Figure 1: Ubiquitin-mediated proteolysis in eukaryotes. Figure 1 Source: Illustrator produced a modified version of Figure 1 from Eldridge, A. G. & O’Brien, T. Therapeutic strategies within the ubiquitin proteasome. Cell Death and Differentiation 17, 4–13 (2010). doi: 10. 1038/cdd. 2009. 82. http: //www. nature. com/cdd/journal/v17/n1/images/cdd200982f1. jpgIn what ways are the structures of prokaryotic and eukaryotic proteasomes similar? Here, we focus on the 20S core particle. Both eukaryotic and prokaryotic 20S core particles are cylindrical structures made of a stack of four rings (Figure 2)8 that gives the 20S core particle its barrel appearance. The two outer rings are called the alpha rings, and the two inner rings are called the beta rings. Each alpha and beta ring is made of seven subunits. Figure 2: Differences between the structures of human and M. tuberculosis beta rings of the 20S core particle. Figure 2 Source: Illustrator produced a modified version of Figure 1 (panel a only) from Lin, G. et al. Distinct specificities of Mycobacterium tuberculosis and mammalian proteasomes for N-acetyl tripeptide substrates. Journal of Biological Chemistry 283, 34423–34431 (2008). doi: 10. 1074/jbc. M805324200 http: //www. jbc. org/content/283/49/34423/F1. expansion. htmlHow do the structures of prokaryotic and eukaryotic 20S core particle differ? In prokaryotes, all seven beta subunits that make up the beta rings are identical and have protease activity (Figure 2). Remarkably, mycobacteria such as M. tuberculosis are the only known bacterial pathogens that have proteasomes. In contrast, each beta subunit of the eukaryotic beta ring is different, and only three of the beta subunit isoforms (Β1, Β2, and Β5) have protease activity (Figure 2). Small Structural Variations Can Make a Big DifferenceIn response to growing cases of drug-resistant TB, Carl Nathan and colleagues rationalized that the small structural and biochemical differences between prokaryotic and eukaryotic proteasomes may translate into differences in substrate-binding specificity5. With this idea in mind, they set out to identify proteasome inhibitors that specifically block the activity of the M. tuberculosis proteasome, but do not affect the function of the human proteasome. Nathan and colleagues screened 20,000 chemical compounds and focused their efforts on the two most promising inhibitors, GL5 and HT11715. Notably, both of these inhibitors contain a functional group called 1,3,4-oxathiazol-2-one. In addition, GL5 and HT1171 are more than 1,000-fold more effective at inhibiting the M. tuberculosis proteasome than the human proteasome. How did these inhibitors compare with bortezomib, the most commonly used proteasome inhibitor at that time? Unlike bortezomib, which reversibly inhibits the M. tuberculosis proteasome, GL5 and HT1171 irreversibly inhibit the M. tuberculosis proteasome. More importantly, GL5 and HT1171 are less toxic to human cells than bortezomib and seem to be specific to M. tuberculosis. The even more exciting finding is that GL5 and HT1171 are able to kill nonreplicating M. tuberculosis cells, which is the state most often found in infected individuals, most of the other anti-TB drugs do not target this cell population. This exciting breakthrough holds great promise for the design of a drug to eradicate the dormant form of M. tuberculosis in infected individuals. Figure 3: Comparison of the crystal structures of the M. tuberculosis 20S core particle of the proteasome in the absence or presence of the HT1171 proteasome inhibitor. Figure 3 Source: Figure 4 (panel b only) from Lin, G. et al. Inhibitors selective for mycobacterial versus human proteasomes. Nature 461, 621–626 (2009). doi: 10. 1038/nature08357 http: //www. nature. com/nature/journal/v461/n7264/fig_tab/nature08357_F4. htmlNathan’s lab went on to determine the molecular mechanism by which GL5 and HT1171 specifically target the M. tuberculosis proteasome. Using mass spectrometry and X-ray crystallographic methods, the researchers discovered that both inhibitors bind irreversibly to the N-terminal threonine residue (called Thr1) of the beta subunit of the M. tuberculosis proteasome. Mechanistically speaking, the oxathiazol-2-one group in HT1171 forms a carbonothioated enzyme intermediate on the Thr1 residue that further reacts to become an oxazolidin-2-one molecule. This stable modification to the beta subunit of the proteasome locks it in place and prevents it from binding other substrates. How does HT1171 accomplish this feat? By comparing the crystal structures of the M. tuberculosis proteasome with or without the addition of HT1171, Nathan’s lab discovered that HT1171 induces a major conformational change in the substrate-binding pocket of the M. tuberculosis proteasome. Specifically, the modification of Thr1 by HT1171 causes an approximately 8º downward tilt of a helix in the Βsubunit, leading to a downward shift of a short loop region called S4-H1 (Figure 3). This striking change has several consequences. First, the normal interactions between amino acids 46, 47, and 48 with amino acids 101, 100, and 99, respectively, in the M. tuberculosis beta subunit become dissociated (Figure 4a). The corresponding primary amino acid sequences are not conserved in the human catalytic subunits (Β1, Β2, and Β5), therefore, HT1171 would not be capable of disrupting the same set of amino acid pairs in the human proteasome. Second, a short loop in the M. tuberculosis beta subunit becomes disordered (Figures 3 and 4a, curved black dashes), further constricting the substrate-binding pocket. Last, the S4-H1 loop forms a new hydrogen bond and three water-mediated hydrogen bonds with a neighboring beta subunit (Figure 4b). This finding is quite unexpected because the interactions with the neighboring beta subunit stabilize the new position of the S4-H1 loop. The corresponding amino acid pairs involved in these interactions also are not conserved in the human Β1, Β2, and Β5 catalytic subunit isoforms. The lack of sequence conservation between the human proteasome and the M. tuberculosis proteasome at these critical amino acids residues likely serves as the basis for the specific inhibition of the M. tuberculosis proteasome by GL5 and HT1171. Figure 4: Crystal structures of the M. tuberculosis 20S proteasome active site. Figure 4 Source: Figure 4 (panels c and d only) from Lin, G. et al. Inhibitors selective for mycobacterial versus human proteasomes. Nature 461, 621–626 (2009). doi: 10. 1038/nature08357 http: //www. nature. com/nature/journal/v461/n7264/fig_tab/nature08357_F4. htmlSummaryPreviously developed antibacterial drugs target nucleic acid synthesis, protein synthesis, cell wall synthesis, or folic acid metabolism. Recent discoveries from Nathan and colleagues add protein degradation to the list. Their insightful research design allowed them to identify compounds that specifically inhibit the M. tuberculosis proteasome without affecting the human proteasome. This breakthrough discovery has important implications for the design of more selective and effective anti-TB drugs, but we need to gain an understanding of the molecular action of these proteasome inhibitors in the context of a complete organism to validate the biochemical studies. Another exciting avenue for future investigation will be to combine two or more drugs that target different M. tuberculosis cellular processes, such as combining a proteasome inhibitor with a protein synthesis inhibitor to create an even more powerful cocktail to kill the bacterial pathogen. References1. Centers for Disease Control and Prevention. Cdc. gov. Accessed May 6, 2011. http: //www. cdc. gov/tb/2. Darwin, K. H. et al. The proteasome of Mycobacterium tuberculosis is required for resistance to nitric oxide. Science 302, 1963–1966 (2003). doi: 10. 1126/science. 10911763. Gandotra, S. et al. In vivo gene silencing identifies the Mycobacterium tuberculosis proteasome as essential for the bacteria to persist in mice. Nature Medicine 13, 1515–1520 (2007). doi: 10. 1038/nm16834. Hoeller D. & Dikic, I. Targeting the ubiquitin system in cancer therapy. Nature 458, 438-444 (2009). doi: 10. 1038/nature079605. Lin, G. et al. Inhibitors selective for mycobacterial versus human proteasomes. Nature 461, 621–626 (2009). doi: 10. 1038/nature083576. Borissenko, L. & Groll, M. 20S proteasome and its inhibitors: Crystallographic knowledge for drug development. Chemical Reviews 107, 687–717 (2007). doi: 10. 1021/cr05025047. Eldridge, A. G. & O’Brien, T. Therapeutic strategies within the ubiquitin proteasome. Cell Death and Differentiation 17, 4–13 (2010). doi: 10. 1038/cdd. 2009. 828. Lin, G. et al. Distinct specificities of Mycobacterium tuberculosis and mammalian proteasomes for N-acetyl tripeptide substrates. Journal of Biological Chemistry 283, 34423–34431 (2008). doi: 10. 1074/jbc. M805324200Arguments AGAINST covalent drugsi. Irreversible inhibitors can be toxic and cause severe side effects when they covalently modify off-target proteins. ii. Once a potential covalent drug is tested in laboratory animals or humans, it may react with off-target proteins and thus not be of clinical use. iii. The examples of successful covalent drugs are too few, so it is unlikely that the pursuit of this class of drugs will have much success. iv. Techniques such as computer modeling are of more use for designing non-covalent drugsA. i and ii onlyB. iii and iv onlyC. i, ii, and iii onlyD. i, iii, and iv onlyE. i, ii, iii, and iv