Subscriber access provided by University of Florida | Smathers Libraries
Letter
Developing Pantetheinase-Resistant Pantothenamide Antibacterials: Structural Modification Impacts on PanK Interaction and Mode of Action Leanne Barnard, Konrad Johannes Mostert, Willem A. L. van Otterlo, and Erick Strauss ACS Infect. Dis., Just Accepted Manuscript • DOI: 10.1021/acsinfecdis.7b00240 • Publication Date (Web): 14 Jan 2018 Downloaded from http://pubs.acs.org on January 16, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Infectious Diseases is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Infectious Diseases
Developing Pantetheinase Pantetheinase--Resistant Pantothenamide Antibacterials: Antibacterials: StrucStructural Modification Impacts on PanK Interaction and Mode of Action Leanne Barnard†, Konrad J. Mostert†, Willem A. L. van Otterlo¶, and Erick Strauss†* †Department ¶Department
of Biochemistry, Stellenbosch University, Private Bag X1, Matieland, 7602, South Africa. of Chemistry and Polymer Science, Stellenbosch University, Private Bag X1, Matieland, 7602, South Afri-
ca. *E-mail:
[email protected] ABSTRACT: Pantothenamides (PanAms) are analogues of pantothenate, the biosynthetic precursor of coenzyme A (CoA), and show potent antimicrobial activity against several bacteria and the malaria parasite in vitro. However, pantetheinase enzymes that normally degrade pantetheine in human serum also act on the PanAms, thereby reducing their potency. In this study, we designed analogues of the known antibacterial PanAm N-heptylpantothenamide (N7-Pan) to be resistant to pantetheinase by using three complementary structural modification strategies. We show that while two of these are effective in imparting resistance, the introduced modifications impact on the analogues’ interaction with pantothenate kinase (PanK, the first CoA biosynthetic enzyme), which act as metabolic activator and/or target of the PanAms. This, in turn, directly affects their mode of action. Importantly, we discover that the phosphorylated version of N7-Pan shows pantetheinaseresistance and antistaphylococcal activity, providing a lead for future studies in the ongoing search of PanAm analogues that show in vivo efficacy.
KEYWORDS: pantothenamide, antimicrobial, coenzyme A, pantetheinase, antimetabolite, bioisostere, mode of action. The biosynthesis and utilization of the central metabolic cofactor coenzyme A (CoA) has been shown to have significant potential for development of antimicrobial targets that do not overlap with those exploited by the current arsenal of available antibiotics.1 The value of this pathway for antimicrobial development lies in CoA being essential to several key metabolic processes such as the TCA cycle and fatty acid metabolism; in organisms such as Staphylococcus aureus, which do not produce glutathione, there is a strong indication that CoA also plays a role in maintaining the intracellular redox balance through the action of a unique CoA disulfide reductase enzyme.2 The finding that the human and bacterial enzymes that are involved in CoA biosynthesis and utilization show significant differences indicates that the development of antimicrobials with low cytotoxicity should be possible.3-4 The N-substituted pantothenamides (PanAms) are a class of analogues of pantothenic acid (vitamin B5 or Pan, 1), the biosynthetic precursor of CoA, that exhibit antimicrobial activity in vitro against several organisms, including Escherichia coli, S. aureus and also the malaria parasite, Plasmodium falciparum.4-6 Among the PanAms tested, N-heptylpantothenamide (N7-Pan, 2) (Fig. 1) have specifically been shown to have potent (submicromolar MIC values) antistaphylococcal activity under certain growth conditions.5, 7-9 In E. coli, this PanAm is ineffective due to TolC-mediated efflux, but shows an MIC of 12 µM in a strain lacking this pump.10 Unfortunately, the potency of the PanAms is reduced or lost in human serum due to the presence of pantetheinase enzymes.11-12 Pantetheinases, which are members of the nitrilase superfamily and are encoded by the Vanin genes, are responsible for the degradation of pantetheine (PantSH, 3), a CoA metabolite, to Pan 1 and cysteamine (Fig. 1).13 This occurs by hydrolysis of its distal amide bond through the action of an invariant Glu-Lys-Cys catalytic triad. Pantetheinases are high-
ly specific for the pantothenoyl moiety of their substrate, but will also act on PantSH analogues in which the cysteamine moiety has been modified or replaced.13-15 Consequently, they also degrade the PanAms to form Pan 1 and the corresponding amine, leading to the observed loss of activity (Fig. 1).11-12 Advancing the PanAms as antimicrobial agents with clinical potential would therefore require interventions that reduce their pantetheinase-mediated degradation, such as those described in a recent patent application.16
Figure 1. Pantetheinase, an enzyme of the Vanin family that normally degrades the CoA-derived metabolite pantetheine (PantSH, 3) to form pantothenic acid (Pan, 1) and cysteamine, also degrades the pantothenamide antimicrobials, including N-heptylpantothenamide (N7-Pan, 2) as shown.
Several recent studies have reported different and complementary strategies to reduce PanAm breakdown in vivo. The first approach entails the use of a combination treatment consisting of a PanAm and a pantetheinase inhibitor.11 Although
ACS Paragon Plus Environment
ACS Infectious Diseases 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
this strategy has shown promise, its further development would necessitate the separate optimization of two compound sets. In addition, it is unclear what the longer-term physiological consequences of pantetheinase inhibition in the human host may be.13, 17-18 A second strategy relies on the structural modification of the PanAms to render them resistant to pantetheinase-mediated degradation, while maintaining their antimicrobial potency. We have shown that PanAm analogues in which the scissile amide bond has been displaced by modification of the β-alanine moiety indeed show reduced rates of degradation. Unfortunately, these changes also caused a loss of potency against both S. aureus and the malaria parasite, most likely due to the need for the intact pantothenoyl moiety to be retained for recognition by the molecular target(s) of the PanAms.5, 19 Indeed, an alternative modification that introduced a methyl group adjacent to the scissile amide’s carbonyl group (but retained the β-alanine backbone) decreased the rate of degradation of the most potent antiplasmodial PanAm identified to date, but reduced its activity only ~2.5-fold.20 In light of these successes, we set out to extend this work by designing and introducing other modifications to the PanAm structure that would likely offer protection against pantetheinase-mediated hydrolysis. We decided to follow three strategies towards this end (Fig. 2a): 1) exchanging the scissile amide bond with a bioisostere that is not prone to hydrolysis; 2) reducing the rate of hydrolysis by introducing methyl groups next to the scissile amide bond, making it less accessible to the pantetheinase catalytic residues by steric occlusion; and 3) modifying the PanAm’s 4'-OH group, a key structural feature that allows these compounds to be recognized as substrates by the pantetheinases (specifically through hydrogen bonding interactions with Glu and Tyr residues).13-15 To implement these strategies we used N7-Pan 2 as scaffold, due to its proven and potent antistaphylococcal activity, and its ability to inhibit E. coli strains that lack the TolC efflux pump.10 This allowed us to evaluate the activity of the new compounds in representative and well-characterized Gram-positive and Gram-negative bacteria. The modifications that were made to produce the new PanAm-based compounds are shown in Figure 2b. For the set in which the amide was exchanged (4, highlighted in green), three bioisosteres widely used in the pharmaceutical industry were employed21-23: a thioamide (N7-PanthioAm, 4a), a hydrazide (N6-PanHyd, 4b), and a sulfonamide (N7-PanSulfAm, 4c). Additionally, following the strategy disclosed in a recent patent application,16 the scissile amide bond was inverted to form C7-PanInvAm (4d), in which the amide’s orientation is opposite to that in the parent molecule. For the methylated analogues (5, highlighted in blue), the α- and β-methylated PanAms (N7-α-MePanAm 5a and N7-β-MePanAm 5b, respectively) were prepared (as the diastereomers, with the methyl group present in both configurations), as well as the Nmethylated version (N-Me-N7-PanAm, 5c). The latter was synthesized in light of the evidence that backbone Nmethylation stabilizes peptidic scaffolds and resists proteasemediated degradation.24 Finally, the set of modifications made to the 4'-OH group (6, highlighted in orange) included a dehydroxylation (4'-deoxy-N7-Pan, 6a), an exchange for an amine (4'-amino-N7-Pan, 6b) and a phosphorylation (4'-phosphoN7-Pan, 6c). Except for the phosphorylated version, the chemistry used to perform these transformations led to the loss of the stereochemical configuration of the 2'-OH group; 6a and 6b were therefore tested as racemic mixtures.
Page 2 of 14
Figure 2. a) Strategies to reduce the rate of pantetheinasemediated degradation: 1) Replace the scissile amide bond with an amide bioisostere; 2) sterically occlude the amide carbonyl group by introduction of methyl groups adjacent to it; and 3) modification of the 4'-OH group, a key structural feature that allows pantetheinase to recognize the compound as a substrate. b) Pantetheinase-mediated hydrolysis of the N7-Pan analogues. Each N7-Pan analogue was incubated with recombinant human pantetheinase. After 24 h, the amount of amine released was determined by means of a fluorescaminebased fluorescence assay. For N7-Pan 2, heptylamine and the no substrate blank, values represent the mean of two independent experiments performed in triplicate with the error bars representing the data range/2. All other values are the means of one experiment performed in triplicate, with the error bars representing the SD (except in the case of 6b, where the background signal measured without pantetheinase-treatment was subtracted). See Supporting Information for details.
ACS Paragon Plus Environment
Page 3 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Infectious Diseases
The new N7-Pan analogues’ sensitivity to pantetheinasemediated hydrolysis was evaluated by treatment with recombinant human pantetheinase for 24 h, and subsequent measurement of the amount of amine released using an established assay.11-12, 20 The extent of degradation was then determined by comparison to the signals obtained using an equivalent amount of pure heptylamine, and a blank reaction that did not contain a substrate (Figure 2b). While the hydrolysis of N7-Pan 2 was found to be complete after 24 h, under the same conditions the bioisostere analogues gave signals comparable to the blank reaction. Among the methylated PanAms, N-Me-N7-PanAm 5c showed the highest apparent resistance to hydrolysis, with the α- and β-methylated counterparts being ~50% and ~30% degraded respectively. Interestingly, both the deoxy- and amino-versions showed relatively high levels of degradation, suggesting that interruption of the hydrogen bonding-interactions between the enzyme and the 4'-OH group (by removal or modification of the latter) is largely tolerated.13-14 However, introduction of the phosphate practically abolished pantetheinasemediated hydrolysis, most likely as the large and negatively charged group cannot be accommodated in the pantetheinase active site. This finding is consistent with recent reports of the prolonged stability of 4'-phosphopantetheine (PPantSH) in serum (which contains pantetheinase).25 It should be pointed out that the in vitro conditions used in these tests likely exaggerate those compared to what the compounds would encounter in vivo, and that these results should therefore rather be considered as a relative stability index, than as an indication of the compounds’ potential to show activity in vivo. Having demonstrated that the modifications introduced to the PanAm structure have imparted the desired resistance to pantetheinase degradation in most cases, we next turned to evaluate the effect of the changes on the compounds’ activity. In a previous study we have shown that the PanAms’ interaction with the target organism’s pantothenate kinase (PanK)— the first enzyme of CoA biosynthesis—determines their mode of action.5 In most cases, PanKs accept the PanAms as alternate substrates, phosphorylating them with a catalytic efficiency similar or comparable to that observed for its native substrate, Pan 1. This is indeed the case for the E. coli PanK (EcPanK), which shows very little kinetic distinction between Pan 1 and its analogues. The result is that PanAms that show inhibition of E. coli growth do so by being converted by EcPanK, phosphopantetheine adenylyltransferase (PPAT) and dephospho-CoA kinase (DPCK) (the subsequent CoA biosynthetic enzymes) into CoA antimetabolites that inhibit essential CoA-dependent processes, such as fatty acid biosynthesis.10, 26 Our expectation therefore was that for the new PanAm analogues to show antibacterial activity against E. coli, they would have to prove to be good substrates of EcPanK. However, the S. aureus PanK (SaPanK) is of a different type than its E. coli counterpart, and consequently its interactions with the PanAms are also distinct.5 Specifically, in a recent SAR study we discovered that while SaPanK shows a binding preference for the PanAms over Pan 1 as substrate, the resulting phosphorylated PanAms are trapped in the enzyme’s active site, causing a ~10-fold reduction in the catalytic turnover rate.27 Consequently, PanAms that have potent antistaphylococcal activity (specifically N7-Pan 2) elicit a unique SaPanK activity profile described by low KM and low kcat values, suggesting that inhibition of PanK’s phosphorylation of Pan 1 is at least one component of the PanAm mode of action in S. aureus.
In a previous detailed kinetic analysis study we had developed a mechanistic model for EcPanK and SaPanK, and therefore used that model to describe their activity profiles towards Pan 1 and N7-Pan 2 as reference parent compound and growth inhibitory PanAm respectively (Fig. 3).5 Specifically, that study found that while EcPanK does not behave in a cooperative manner towards Pan 2, SaPanK shows clear cooperativity in its activity towards its substrate. However, both enzymes demonstrated varying degrees of cooperativity towards the PanAm analogues. Most importantly, the enzyme’s activity profiles towards N7-Pan 2 respectively demonstrates how its interaction with the PanK dictates its mode of action: for EcPanK the close overlap in the activity profiles for Pan 1 and N7-Pan 2 at 50% inhibition at this concentration. With SaPanK, we found that all three compounds caused some inhibition (Fig. 4): the deoxy-variant 6a was the most potent inhibitor (IC50 ~6 µM, a value similar to that previously determined for 4'-deoxy-N5-Pan5). The amino-substituted version 6b gave a curious inhibition profile, with the reduction in activity approaching but never reaching 50%. This suggests that this analogue inhibits only one of the subunits of the dimeric SaPanK enzyme, although this remains to be confirmed. Finally, as an analogue of the product of PanK phosphorylation, 6c showed good inhibition (IC50 ~40 µM); this observation correlates with our recent finding that the PanAms inhibit SaPanK by being trapped in the enzyme as product complexes.27
Figure 3. PanK activity profiles with the new PanAm analogues as substrates. EcPanK (panel a) and SaPanK (panel c) profiles with bioisostere exchanged PanAms (4), and EcPanK (panel b) and SaPanK (panel d) profiles with methylated PanAms (5). The fitted activity profiles for Pan (1) (solid grey line) and N7-Pan (2) (dashed grey line) are shown for reference. The data are the averages of at least two independent experiments, each performed in triplicate. The error bars show SEM or range/2 as appropriate, and are smaller than the symbol when not visible. The colored solid lines are the curves that result when the data are fitted to a Hill-type equation (see Methods for details and Table S1 in Supporting Information for corresponding kinetic parameters).
Figure 4. SaPanK inhibition profiles with the 4'-OH modified PanAm analogues 6. The enzyme’s activity towards 25 µM Pan 1 was used as reference, and the relative reduction of activity in the presence of increasing amounts of inhibitor was measured. Symbols represent the average of three independent experiments, each performed in triplicate; the error bars show the SEM, and are smaller than the symbol when not visible. The dashed line indicates the 50% activity level.
Interestingly, SaPanK did not show the N7-Pan-type kinetic profile towards any of the PanAm analogues. In fact, with one exception the enzyme showed some level of activity towards all the compounds, albeit with increased KM and/or reduced kcat values that corresponded to specificity constants (kcat/KM) less than half that of Pan. This finding would suggest that none of these new analogues would inhibit S. aureus in a manner similar to N7-Pan 2, but instead could serve as antimetabolite precursors. This is particularly the case for the sulfonamide 4c, which had an SaPanK activity profile very similar to that of Pan 1. This would suggest that analogues such as 4c could potentially act as inhibitors of S. aureus with a mode of action similar to that found in E. coli, where antimetabolite formation predominates as the basis for inhibition. The PanAm analogues 6 that had their 4'-OH group modified, and therefore cannot act as PanK substrates (or as antimetabolite precursors), were evaluated as inhibitors of EcPanK and SaPanK’s ability to phosphorylate Pan 1. This was done by introducing the compounds to reaction mixtures containing 25 µM Pan 1, a concentration in the range of both
Finally, we set out to determine if the observed interactions with the PanK enzymes correlated with their antibacterial potential. Growth inhibition assays were performed with an E. coli tolC knockout strain (JW5503 from the Keio collection28) and S. aureus RN4220 using minimal medium (i.e. with no Pan 1 present) to provide conditions that would lead to maximum sensitivity toward the potential inhibitors. The bacteria were incubated for 24 h at 37 °C using 200 µM of each PanAm analogue; if inhibition was observed at this concentration, the MIC (minimum inhibitory concentration) was determined. The inhibition experiment was repeated in a rich medium (1% tryptone) that contains ~1–5 µM Pan 1 for the most potent inhibitors, to also evaluate their activity when the vitamin is present as a potential antagonist. The results show that for E. coli only treatment with C7PanInvAm 4d resulted in >80% inhibition at 200 µM (Table 1). Although some of the other analogues acted as substrates of EcPanK with similar or improved activity compared to Pan, none of these showed inhibition. This suggests that the modifications to these compounds either reduce cell permeability
ACS Paragon Plus Environment
Page 5 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Infectious Diseases
and uptake, or that the corresponding antimetabolites are either not formed or that they show poor activity to the various CoA-dependent targets. Interestingly, the only other two compounds that did show some (20–50%) inhibition at 200 µM are the 4'-OH modified analogues 6a and 6b. As these compounds cannot act as substrates of EcPanK and therefore cannot be converted into CoA antimetabolites, the observed growth inhibition is most likely due to their inhibition of EcPanK. This suggests that these compounds could act as leads for the development of pantetheinase-resistant PanAm antibacterials that specifically target EcPanK and PanKs that are similar to it (so-called Type I PanKs1, 3), i.e. with a mode of action that is currently only observed in S. aureus. Table 1. Growth inhibition of E. coli JW5503 (ΔtolC) and S. aureus RN4220 by the N7-Pan analogues.
N7-Pan analogue
E. coli a % inhibition at 200 µM
S. aureusb MIC (µM)
N7-Pan 2
>99%
200
N7-PanSulfAm 4c
none
>200
C7-PanInvAm 4d
>80%
~12.5 (>200)
N7-α-MePanAm 5a
none
>200
N7-β-MePanAm 5b
none
~50 (>200)
N-Me-N7-PanAm 5c
none
>200
4'-Deoxy-N7-Pan 6a
~30%
~25 (>200)
4'-Amino-N7-Pan 6b
20-45%
> 200
4'-Phospho-N7-Pan 6c
none
~1.8 (~24)
aAn
E. coli ΔtolC strain (JW5503) was grown in minimal medium for 24 h at 37 °C in the presence of 200 µM of the indicated N7-Pan analogues; the reported values give the amount of growth inhibition observed at this concentration relative to an untreated control. bS. aureus RN4220 was grown in minimal medium for 24 h at 37 °C in varying concentrations of the N7Pan analogues. The MIC was taken as the lowest concentration of compound that showed full growth inhibition relative to a control. Compounds that showed inhibition under these conditions were tested by growing for 20 h at 37 °C in 1% tryptone (values in parentheses). The reported values represent the mean of two or more independent experiments. cData from a previous published study.5
For S. aureus, four N7-Pan analogues—4d, 5b, 6a and 6c— had MIC values lower than 200 µM, with 4'-phospho-N7-Pan 6c surprisingly showing quite potent activity (MIC ~1.8 µM, corresponding to
