Genome-scale reconstruction of the metabolic network in Yersinia pestis, strain 91001

Ali Navid and Eivind Almaas*

Biosciences & Biotechnology Division, Lawrence Livermore National Laboratory, Livermore, California 94550-0808, USA

*Corresponding author Email address: almaas@llnl.gov

• 2 - Abstract The gram-negative bacterium Yersinia pestis, the aetiological agent of bubonic plague, is one the deadliest pathogens known to man. Despite its historical reputation, plague is a modern disease which annually afflicts thousands of people. Public safety considerations greatly limit clinical experimentation on this organism and thus development of theoretical tools to analyze the capabilities of this pathogen is of utmost importance.
  Here, we report the first genome-scale metabolic model of Yersinia pestis biovar Mediaevalis based both on its recently annotated genome, and physiological and biochemical data from literature. Our model demonstrates excellent agreement with Y. pestis’ known metabolic needs and capabilities. Since Y. pestis is a meiotrophic organism, we have developed CryptFind, a systematic approach to identify all candidate cryptic genes responsible for known and theoretical meiotrophic phenomena. In addition to uncovering every known cryptic gene for Y. pestis, our analysis of the rhamnose fermentation pathway suggests that betB is the responsible cryptic gene.
  Despite all of our medical advances, we still do not have a vaccine for bubonic plague. Recent discoveries of antibiotic resistant strains of Yersinia pestis coupled with the threat of plague being used as a bioterrorism weapon compel us to develop new tools for studying the physiology of this deadly pathogen. Using our theoretical model, we can study the cell’s phenotypic behavior under different circumstances and identify metabolic weaknesses which may be harnessed for the development of therapeutics. Additionally, the automatic identification of cryptic genes expands the usage of genomic data for pharmaceutical purposes.

• 3 - Introduction

The “black death” pandemic that ravaged Europe between the 14th to 16th centuries is the most infamous outbreak of pestilence in history. Within a short five-year period (1347-1352), thirty three million people, one out of every three Europeans perished.
The Mediaevalis biovar of the gram-negative bacterium Yersinia pestis (YP), the aetiological agent of bubonic plague, is believed to have caused this epidemic 1, 2.
The most recent outbreak of bubonic plague in Asia killed nearly 12.5 million people in India alone from 1889 to 1950. Throughout human history, a conservative estimate stipulates that 200 million people have been victims to this deadly disease in various pandemics 3. Despite modern advances in medicine, no working vaccine against the plague exists, and it is listed by the US Centers for Disease Control and Prevention (CDC) as a Category A bio-terrorism pathogen. While plague is frequently considered a disease of the past, several thousand new cases are reported each year, predominantly in Africa 4. Between 1990 to 1995, the Democratic Republic of Congo, Tanzania and Zimbabwe alone reported 4939 cases of plague 5. Hence, the recent reports of antibiotic resistant strains of YP 6-8 are cause for great alarm.
Over the past decade, the revolutionary advances in high-throughput technologies and computational approaches have led to the inception of systems biology, which aims to transform microbiology from a science which focuses on one specific cellular process or pathway, to one that the biology of the system as a whole is examined. To achieve this goal, genome-scale models of metabolism have been developed using constraint- based approaches with flux-balance analysis (FBA) being the most widely used method. FBA’s success originates from the fact that, unlike kinetic models, FBA only seeks to identify optimal metabolic steady-state activity patterns that satisfy

• 4 - constraints imposed by mass balance, the metabolic network structure, and the availability of nutrients. The most common objective function (cellular task to be optimized) is that of growth, although other choices are possible depending on the selective environment of the cell 9. FBA has been applied to several genome-scale models 10-14 with great success. Additionally, FBA has been used to examine a range of topics from the global organization of metabolic fluxes 15 to the effect of genetic knockouts 16 and the discovery of novel regulatory interactions 17. However, the overall process of system building is still highly labor intensive and requires extensive human curation to generate high-fidelity models.
  Here, we present and analyze the first genome-scale reconstruction of an organism classified by the CDC as a Category A pathogen. We demonstrate excellent agreement with known metabolic performance for YP

…(Full text truncated)…