Covalent host targeting by thioester domains of Gram-positive pathogens

Bacterial infections represent one of the most important challenges for medicine and science. The emergence of bacteria resistant to almost all known antibiotics is a threat to global health. The urgent need for new antibacterial drugs is widely recognised but very few have been introduced in the last couple of decades. In order to develop treatments that differ significantly from known drugs like penicillin we need to understand better the ways in which bacteria cause infections. To start with, bacteria need to gain a foothold in the human body. This is achieved through a variety of molecules present on the bacterial surface. These molecules, mostly specialised proteins that often form hairlike structures called pili, bind to human cells found at the site of bacterial entry into the body (e.g. throat, airways, gums, guts, skin). Once bacteria manage to cling to human cells they either survive and multiply or they invade their host by entering cells, thus eventually spreading in the body. If it was possible to prevent the bacteria from binding to human cells, it would be possible to intervene with infections at an early stage.

Our research has uncovered a particularly intriguing mechanism that bacteria may use to bind rapidly and very tightly to human cells. This compares to known bacterial binding modes like superglue compares to Velcro. Importantly this new binding mechanism appears to be shared by some of the most important human pathogens. These include the bacterium Streptococcus pneumoniae (also known as pneumococcus), which is the most important cause of pneumonia in the UK and worldwide. It also causes a form of meningitis and septicaemia (blood poisoning), which in developing countries is responsible for one quarter of all preventable deaths in children under five. Pneumococcal diseases are very common in hospital settings and in the community and are problematic because the bacteria are resistant to many antibiotics. Other bacteria using the superglue-like binding mechanism are relatives of pneumococci. Streptococcus pyogenes causes a wide range of conditions including scarlet fever and strep throat, affecting hundreds of millions of people worldwide. These bacteria can also cause severe follow-on diseases. Of these, rheumatic heart disease (RHD) is one of the biggest killers of children and youths in India and other developing nations. Streptococcus pyogenes could develop antibiotic resistance just like the pneumococci did, which would result in the re-emergence of dangerous diseases like childbed fever or RHD. We also predict that many of the bacteria causing gum disease or teeth loss will be able to produce superglue proteins which may enable bacteria to very tightly stick to teeth surfaces and gums.

We propose to study the molecular details of the "bacterial molecular superglue". We need to determine the precise architecture of the critical proteins made by the bacteria in order to understand how they attack human cells. To achieve this, we will use a combination of two powerful methods, NMR spectroscopy and X-ray crystallography, to reveal the structures of the bacterial proteins in atomic detail. At the same time we will determine which components of human cells serve as binding targets for the bacteria. This will be accomplished by using the bacterial proteins as molecular bait, fishing for their binding partners on or in model human cells (grown in cell culture). Once the targets have been identified we will establish how exactly the bacterial proteins recognise human cell components. This information will allow us to design small molecules that would specifically react with bacterial proteins, thus forming the basis for new diagnostic tools and antimicrobial substances. We will combine our molecular studies with experiments using bacteria in order to find out if a loss of the critical superglue function on bacteria makes them less efficient invaders of the human body.

We have identified a novel family of protein domains, containing internal thioester bonds, in dozens of surface proteins of some of the most important Gram positive human pathogens including Streptococcus pyogenes, Streptococcus pneumoniae, Clostridium perfringens and several periodontal pathogens. Internal thioesters have previously only been found in complement proteins where they form covalent bonds to the surface of pathogens. The discovery of reactive thioester bonds in bacterial surface proteins known or predicted to mediate bacterial adhesion suggests the exciting possibility of a previously unrecognised principle in host-bacteria interactions, namely the covalent attachment of bacteria to host tissue.

Preliminary data confirmed the presence of thioesters in predicted thioester domains (TEDs) from S. pyogenes. Even more excitingly TEDs were found to form covalent bonds with a host protein that is known to be an important target of many Gram-positive pathogens.

We will use X-ray crystallography and NMR spectroscopy to obtain high-resolution structures of TEDs from different species, providing a molecular framework to interpret TED function. Physiologically relevant TED binding partners will be identified in pull-down assays exploiting the covalent nature of TED/target complexes. The relevance of thioesters for bacterial adhesion and cell invasion will be explored using various bacterial strains. The presence of a reactive group in bacterial surface proteins opens new avenues for diagnosis and potentially also therapeutic intervention. As a first step small molecular probes will be developed that specifically and covalently label thioesters.

Given the urgent need to develop new antimicrobial strategies to prevent the looming antibiotics crisis the discovery of a novel molecular virulence mechanism is an exciting development and needs to be addressed now in a concerted effort bringing together molecular and microbiological expertise as we propose.

Economic and societal impacts of the proposed research are in the long term the improvement of health and wellbeing and opportunities for commercialisation and exploitation (long term and more immediate).

The proposed research addresses a potential novel mechanism for bacterial virulence that is shared by some of the most important bacteria affecting humans. Streptococcus pneumoniae and Streptococcus pyogenes belong to the top ten most deadly pathogens. Despite being responsible for over 1.5 millions deaths annually, streptococcal and pneumococcal infections are neglected diseases (PolicyCures, G Finder Report 2010) and pathogenicity mechanisms are poorly understood. This hampers urgently needed progress in the development of new antimicrobial agents and vaccines, which is of paramount importance to societies in the UK and elsewhere. We must be prepared to the possibility that an extremely common and potentially devastating pathogen like Streptococcus pyogenes could develop resistance to beta lactams (there is already widespread resistance to other classes of antibiotics), as the experience with pneumococci in the late 20th century showed. Novel treatments are also required for diseases caused by multidrug resistant pneumococci that represent a major problem in community and hospital acquired infections. Our research is likely to lead to the discovery of TEDs in other Gram-positive pathogens such as staphylococci, bacilli and clostridiae. Our basic research into bacterial virulence mechanisms is highly relevant to society with a long-term impact on health.

While bacterial adhesion is a recognised new target for antimicrobial intervention, it is notoriously difficult to develop small molecules that would inhibit often multivalent interactions involving large binding interfaces. The discovery of thioesters opens the possibility to target bacterial adhesion with inhibitors that are similar to classic drugs like enzyme inhibitors. Therefore our research should be of interest to pharmaceutical companies.

While this project may not result in new therapeutic approaches in the short term, small molecules specifically reacting with bacterial thioester domains may have a more immediate impact. These can potentially serve as probes for the detection of TED-expressing bacteria with potential applications as rapid diagnostic tools. Furthermore, the spontaneous bond formation between thioester domains and target molecules may also form the basis of biotechnological applications, where TEDs could be used as covalent markers for specific detection and labelling of biomolecules.

Finally, we identified several surface proteins from periodontal pathogens likely to contain TEDs. If in the future it could be shown that these bacteria use TEDs to tightly adhere to gingival and dental surfaces, this might have interesting implications for oral hygiene and dental medicine.