Senior Lecturer in Organic Chemistry

B.A., D.Phil (Oxford)

Tel: 0116 2522093

email: bjr2@le.ac.uk

Research Interests

Background

Amphotericin B 1 is produced by the soil bacterium Streptomyces nodosus and, despite considerable toxicity, it has been a leading broadspectrum antifungal antibiotic for more than fifty years. Resistance has not become a problem in this long period of clinical use, unlike all other established antifungal therapies. It has also emerged as the first line of treatment against some Leishmania parasites, and has potential in the treatment of HIV infection and prion diseases.
Traditionally, 1 was injected as a solution in DMSO or as a complex with deoxycholate. Recent liposomal formulations have greatly reduced toxicity, but are sufficiently expensive to require patient selection on pharmacoeconomic considerations, and are major components of the drugs bills of some NHS Trusts.
Visceral leishmaniasis (kala azar) is caused by protozoan parasites transmitted by the sand fly, and is fatal if left untreated. It remains a ‘most neglected’ disease with 50,000 deaths each year, many in high-level poverty areas in Asia and South America.  Conventional 1 is now the first line of treatment in Bihar, India and in neighbouring Nepal.  Efficacy is close to 100% but there are nephrotoxic side effects and administration requires hospitalisation for a complicated infusion regime of fifteen injections over thirty days. Despite considerable subsidy from the WHO (US$3,000 per treatment reduced to US$200), the cost of the less toxic liposomal formulation Ambisome severely restricts it use in these regions, some of the poorest parts of the world. Thus there is a need for an affordable effective analogue with greatly reduced toxicity.

Amphotericin B toxicity is only partly alleviated by costly liposomal formulation. Improvement could be achieved by structural alteration.  Chemical studies have generated  derivatives such as N-methyl-N-fructosyl amphotericin B methyl ester (MFAME) that has been shown to be an effective non-toxic antibiotic but is much too expensive for clinical use.
The activity of MFAME shows that much improved analogues of 1 exist in chemical space, the challenge is to produce them at a low cost. Due to the complex structure of 1, chemical modification is unlikely to deliver an affordable non-toxic derivative. Such a compound might instead be obtained by engineering biosynthetic genes in S. nodosus. The parent wild-type strain can produce high levels (4 g L-1) of 1 in low cost fermentations, demonstrating the presence of a very efficient metabolic flow of the required biosynthetic building blocks, and efficient enzymatic processes. Thus, given sufficient understanding of the genetic and biosynthetic machinery, there is potential to achieve low-cost production of engineered compounds that can replace 1 in therapy.
This work is part of an ongoing programme of research with Dr Patrick Caffrey (Department of Industrial Microbiology, University College, Dublin), to achieve the in vivo production of such an affordable non-toxic analogue of 1. All molecular biology is performed in Dublin, whilst growth, isolation and purification protocols are developed in Leicester.

Recent achievements:
We have focussed on gene disruption to establish the biosynthetic pathway and the flexibility of the enzymes involved.
Disruption of the gene encoding cytochrome P450 monooxygenase that inserts the 8-hydroxyl group gave 8-deoxyamphotericin B. Disruption of the gene involved in oxidising a methyl group on C-16 to a carboxyl group gave the analogue with the  unoxidised methyl group (2) along with the polyketide synthase product aglycone. Bioassay of 2 shows that it is several fold better than 1 with reduced toxicity, indicating that the carboxyl group is not required for therapeutic activity.
A series of ketoreductases in the polyketide synthase have been disrupted. Disruption of KR16 (ketoreductase in module 16) gave 8-oxoamphotericin B in good yield. This showed greatly increased water solubility, lack of serum solubility contributes to the toxicity of 1. Disruption of KR12 gave only poor yields of the corresponding 15-oxo derivative suggesting that hemiacetal formation occurs as the acyl chain is being assembled, and that this hemiacetal ring formation is slowed by the enolisation caused by the adjacent carbonyl. Disruption of KR10 gave the truncated polyketide 3 in good yield. The pyrone is formed by cyclisation of a 3,5-dioxothioester. It is very rare for a polyketide synthase to release such large quantities of a truncated intermediate in the assembly process.
In recent work, we have generated hybrid gylcosyltransferases to substitute a different sugar group in 1. Glycosyltransferases (GT) attach the 6-deoxyaminosugar to the aglycone at C-19. A GT has two domains, one to bind to the aglycone, and one to bind to the sugar. In amphotericin B biosynthesis, this is followed by cytochrome P450 oxidation at C-8 to give 1.
Using molecular biology, Caffrey has assembled a hybrid GT.  Caffrey replaced the domain that recognises the sugar mycosamine with a domain from a related organism whose GT recognises perosamine, a regioisomer of mycosamine with an amino and hydroxyl group swopped around. The hybrid glycosyltransferase will now recognise the amphotericin aglycone, and recognise the sugar perosamine.  Other genes were inserted or disrupted to ensure the in vivo formation of activated perosamine in the absence of mycosamine. The mutant was grown in Leicester, and several milligrams of the new compound was isolated, purified and structurally characterised by MS and NMR. The product showed an improved therapeutic ratio.  This is proof of concept for future work replacing the mycosminyl sugar residue with a wide range of other sugars and oligosaccharides for biological analysis as a replacement of 1 in therapy.

Selected Publications

1. Redesign of polyene macrolide glycosylation: Engineered biosynthesis of 19-(O)-perosaminyl-amphoteronolide B. Eve Hutchinson, Barry Murphy, Terence Dunn, Ciaren Breen, Bernard Rawlings and Patrick Caffrey, Chem. Biol., 2010, 17, 174-182.
2. Engineered biosynthesis of 7-oxo and 15-deoxy-15-oxo-amphotericins: Insights into structure-activity relationships in polyene antibiotics. Patrick Power, Terence Dunne, Barry Murphy, Laura Nic Lochlainn, Dilip Rai, Charles Borissow, Bernard Rawlings and Patrick Caffrey, Chem. Biol., 2008, 15(1), 78-86.
3. Biosynthesis of amphotericin derivatives lacking carboxyl groups. Maria Carmody, Barry Murphy, Barry Bryne, Patrick Power, Dilip Rai, Bernard Rawlings and Patrick Caffrey, J. Biol. Chem., 2005, 41, 34 420-34 426.