Introduction

1. Introduction

1.1 Emerging of resistant pathogens

The rise of bacterial resistance against long-standing drugs of last resort, such as vancomycin, has created an urgent demand for the development of new antibiotics with enhanced or broadened antimicrobial efficacy^1-5. Recent strategy for finding new classes of antibacterial compounds based on targets identified from bacterial genomics has not yet proved successful⁶. This has led us to consider rationally manipulating potential natural products. Natural products have historically been invaluable as a source of antibacterial drugs, such as penicillins, macrolides and glycopeptides, reflecting their evolutionary origin as ‘weapons’ that bacteria use against each other. Glycopeptide, vancomycin, had long been known as the ‘last resort’ in treating serious infections caused by bacteria resistant to several other antibiotics, such as the ‘super bug’

methicillin-resistant Staphylococcus aureus (MRSA), but even vancomycin resistance has now emerged, further underlining the need for new antibiotics. Thus, compounds with activity against clinically important Gram-positive pathogens, including antibiotic-resistant strains such as MRSA, penicillin antibiotic-resistant Streptococcus peuminiae, and vancomycin resistant Enterococci (VRE) are highly demanded^{7, 8}.

1.2 Glycopeptide antibiotics

Vancomycin (Figure 1) that has outpaced multidrug resistance of methicillin-resistant Staphylococcus aureus (MRSA) for decades is facing mighty backlashes by emerging vancomycin-intermediate S. aureus (VISA) and vancomycin-resistant S. aureus (VRSA) in addition to having lost battle to vancomycin-resistant Enterococci (VRE).

Dalbavancin (Figure 1), a new member of the vancomycin family antibiotics known for active against Staphylococci (e.g. MRSA) and Streptococci while weak against

2

Enterococci (e.g. VRE), was developed based on the drug lead glycopeptide antibiotic A40926 (Figure 1). Two other renowned vancomycin-based derivatives, oritavancin and telavancin, are active against strains resistant to conventional glycopeptides (e.g. VRE) but relatively weak against methicillin-resistant Staphylococcus aureus (MRSA).

Seeking new compounds with enhanced and/or broader antimicrobial spectrum for refractory bacterial infections are never over demanded. A40926 differs from teicoplanin (Tei, 1, Figure 1) mainly in lacking the N-Ac glucosaminyl on residue 6 (r6) and the replacement of the N-acyl glucosaminyl C-6 OH substituent on residue 4 (r4) with a COOH group (Figure 1). These structural features make it an ideal system to exploit the manipulation of genes and enzymes involved in the biosynthetic pathways of both antibiotics to create hybrid structures and new analogs. The close structural similarity of vancomycin and related glycopeptide antibiotics including A40926 and teicoplanin have prompted a variety of studies to determine whether late-stage tailoring enzymes that act on one of the family members will cross-react with other family members to create new hybrid analogs; these strategies have yielded several compounds with altered and improved antimicrobial profiles, demonstrating the utility of this approach. Though many hybrid structures have been developed using these methods or have been identified from natural sources, modifications at the sugars of the residue 4 and 6 are rare.

Three new glycopeptides are currently close to common clinical usage, namely, oritavancin (Targanta Therapeutics Inc) and telavancin (Theravance Inc/Astellas Pharm Inc) which are analogs of vancomycin, and dalbavancin (Pfizer Inc) (Figure 1) which in fact is a A40926 analog. All three compounds have some commonality in structure but with difference in pharmacokinetic and pharmacodynamic properties and in vitro spectrum. In general, the three compounds all have been well tolerated in clinical trials

3

to date but with pros and cons individually. In short, oritavancin and telavancin are active against strains resistant to conventional glycopeptides (e.g. VRE) but relatively weak against strains such as methicillin-resistant Staphylococcus aureus (MRSA), while dalbavancin features active against Staphylococci (e.g. MRSA) and Streptococci but weak against Enterococci (e.g. VRE) or Staphylococci harboring the vanA gene cluster.

1.3 Modifications of glycopeptide antibiotics by biosynthetic enzymes

Previous reports have shown that even modest structural modifications of vancomycin can overcome resistance. We have elucidated the biological functions for enzymes Dbv8, 9, 21, and 29. They together were characterized to be responsible for the synthesis of the N-acyl glucosamine moiety of A40926 in the filamentous Actinomycete Nonomuraea sp. ATCC39727. Given the “diversity-oriented synthesis”

and “synthesis on privileged compound” as new concepts in the field, we are keen to build up a useful platform using native and engineered Dbv8, 9, 21, and 29 as biocatalysts. New analogs by modifications of teicoplanin will be dissimilar to any reported compound (they are almost vancomycin and A40926 derivatives). New analogs demonstrated superior in anti-VRE activity will be expected to be at least not less than oritavancin and/or telavancin to the given strains tested and they are very likely to be more effective to some oritavancin- and telavancin-insensitive strains. As a consequence, diversification of the “privileged compound” teicoplanin may help conquer bacterial resistance and ease the strong demand for new antibiotics.

1.4 Dbv29

Though many hybrid structures have been developed using these methods or have been identified from natural sources, modifications at the C6 position are rare. We have been interested in Dbv29, a hexose oxidase that catalyzes the last step in A40926 biosynthesis in the filamentous Actinomycete Nonomuraea sp. ATCC39727^{9, 10}. Our

4

previous efforts demonstrated that this flavoenzyme performs a four-electron oxidation to convert the glucosaminyl C6 alcohol found on the ‘top’ of the peptide scaffold to a carboxylic acid. Dbv29 is unusual among enzymes known to perform oxidations at the C6 site, both in the coenzyme used (for example, NAD is used in the UDP-glucose dehydrogenase family) and in lacking a typical cysteine residue in the active site^11-13. Our previous studies showed that Dbv29 activity is dependent on the presence of an acyl group on the C2 amine of the same carbohydrate residue9, 10, 14, 15

, though the exact nature of this acyl group proved somewhat flexible, as Dbv29 could oxidize both teicoplanin—containing an extended acyl chain—and analogs of A40926 with an acetyl or acyl group attached (and lacking a distant N-acetyl glucosamine substituent).

Bioinformatics analysis pointed to two tyrosine residues, Tyr165 and Tyr473, as potentially important for the reaction mechanism¹⁶, but single phenylalanine mutants did not significantly affect activity in an overnight incubation experiment. Mutagenesis of His91 and Cys151 did provide strong support that the flavin cofactor, characterized at that time as flavin mononucleotide (FMN), was covalently bound to the enzyme through these residues; unexpectedly, the double mutant of these two enzymes retained some level of activity, unlike other flavoenzymes that covalently link to their cofactor^9,

16-19

. Thus, the function of this enzyme remained enigmatic.

In this study, we sought to both understand the mechanism of this unusual enzyme and—given our prior demonstration of acyl chain promiscuity—explore whether it could be employed to generate non-natural glycopeptide analogs. We report crystallographic, biophysical and biochemical characterization of Dbv29 that demonstrates the critical role of Tyr165 and Tyr473 in the enzyme reaction.

Additionally, through the serendipitous discovery of a reaction intermediate in the crystal structure, we design a new synthetic strategy to rationally intercept the reaction

5

mechanism. This approach allows access to new classes of products that would be extremely difficult to obtain by other means and that offer improved and complementary profiles for antimicrobial drug discovery efforts.

6