Poster on systems pharmacology of the cholesterol biosynthesis pathway
1. 3. Results (I)
2. Pathway production
6. References
1 Deanery of Biomedical Sciences, University of Edinburgh, Edinburgh, UK. 2 Northern Ireland Centre for Stratified Medicine, University of Ulster, C-Tric, Derry,
UK (current addresses for HEB and CPM are given in reference [2])
www.guidetopharmacology.org enquiries@guidetopharmacology.org @GuidetoPHARM
A systems pharmacology study of the cholesterol
biosynthesis pathway
Supported by:
We especially thank all contributors, collaborators and NC-IUPHAR members
1. Introduction
Information on drugs, lead compounds and their pharmacological effects is expanding
in online resources, including the IUPHAR/BPS Guide to PHARMACOLOGY (GtoPdb)
[1]. This means that for key pathways and modules there is an expansion in the number
of data-supported druggable targets captured in databases. As this catalogue of
molecular interactions and our understanding of biological systems expands, it will be
advantageous to integrate these resources in order to devise new potential therapies.
Drug combinations present an opportunity for therapy development that can target
pathways more precisely than perturbing entire networks. Systems pharmacology will
also impact genomic medicine, including personalisation of treatments and stratification
of patient groups. Thus, as our understanding increases, we have opportunities to
predict, model, quantify and test combinations that may have advantages over
conventional single-drug therapies. This work explores the feasibility of such systems
pharmacology approaches with an analysis of the mevalonate branch of the cholesterol
biosynthesis pathway. This poster is a summary of a detailed paper published in Sept
2017 that also includes extensive supplementary data [2].
Updates of our extensive curation of the pathway are now in GtoPdb as below.
Figure 1. These snap shots are taken from the pathway display in:
http://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=104
The summary links specify the protein properties and detailed pages display kinetic
parameters along with substrates and inhibitors for the enzymes as ligand entries.
Figure 1. Examples of various kinase database tables
1. Harding SD, et al. (2018) The IUPHAR/BPS Guide to PHARMACOLOGY in 2018: Updates and expansion to encompass
the new Guide to IMMUNOPHARMACOLOGY. Nucl. Acids Res. 46 (Database Issue). doi: 10.1093/nar/gkx1121.
2. Benson, HE et al, (2017) Is systems pharmacology ready to impact upon therapy development? A study on the
cholesterol biosynthesis pathway. Br. J. Pharmacol. Sep 14. doi: 10.1111/bph.14037. [Epub ahead of print]
3. Mazein A et al. (2013) A comprehensive machine-readable view of the mammalian cholesterol biosynthesis pathway.
Biochem Pharmacol. 86(1):56-66. doi: 10.1016/j.bcp.2013.03.021
.
GtoPdb is an ELIXIR UK node resource
4. Model of the mevalonate arm of the cholesterol pathway
We produced a model of the mevalonate arm of the cholesterol biosynthesis pathway in
Systems Biology Graphical Notation (SBGN) for the metabolic steps from acetyl-CoA
and acetoacetyl-CoA consumption to squalene and geranylgeranyl diphosphate
production. This comprises 12 steps, 10 enzymes and 14 metabolites.
Figure 2. The full enzyme names for this pathway diagram are in Fig.1
We used the model to calculate steady-state flux profiles without inhibitors and then
tested the affects of drug combinations on fluxes via computational optimization. We
determined a combination of five inhibitors showed the desired suppression squalene
while maintaining normal geranyl diphosphate levels (full details are available in [2])
Christopher Southan1, Helen E. Benson1, Steven Watterson2, Joanna L. Sharman1,
Chido P. Mpamhanga1 and Andrew Parton2
5. Conclusions
Our initial attempts to build a systems pharmacology model of the mevalonate arm of
the cholesterol biosynthesis pathway revealed gaps and inconsistencies in the data that
prevented us from achieving a high level of confidence. In particular, we found the lack
of comprehensive and systematic parameterizations, experimental variation, ambiguity
in structural detail and inappropriate and inaccurate reporting from the primary literature
to be obstacles. That this should be the case for a pathway of such high biomedical and
commercial significance was unexpected. For this reason, our best current
parameterization represents a patchwork of values taken from multiple species and
experimental configurations. Nonetheless, by completing gaps in our knowledge with
representative values, we were able to demonstrate subtle reprogramming of pathway
dynamics that may contribute significantly to drug development. We propose that these
obstacles can be reduced through the adoption of standards and quality control.
Although we have focused on the mevalonate arm of cholesterol biosynthesis, this
approach could be applied to any pathway of interest for which targets, ligands and
kinetic parameters are known. Note also that GtoPdb expands the capture of ligand-to-
target relationships every release. However, extending modelling opportunities more
generally needs both the computational biology and the pharmacology communities to
reduce barriers to progress. The model from this work and our previous study [3] is
available from http://biomodels.org (ID 1506220000) and can freely be used and
adapted.
We combined the interrogation of multiple databases and cross-checking primary
literature to establish the enzymes involved in the pathway, the reactions they
catalyse, subcellular localization, species in which they were identified, substrate,
kinetic parameters and inhibitors. This uncovered some inconsistences and
ambiguities in database entries that we had to resolve against data specified in the
papers. We combined ordinary differential equation (ODE) kinetic models, the
pathway parameters and the inhibitor parameters to create a model describing the
dynamics of the mevalonate pathway. This incorporated Michaelis–Menten kinetics
to describe each step, except the interactions consuming isopentenyl diphosphate
and producing geranylgeranyl diphosphate and pre-squalene diphosphate. These
steps were described using mass action kinetics in order to simplify the process of
calculating the steady state of the model and hence the steady state behaviour of
the pathway. We then sought to identify the drug combination that would best
suppress the production of squalene as a precursor for cholesterol, but would also
maintain production of geranylgeranyl-diphosphate at the same levels as in the
absence of any inhibitors, thereby eliminating a significant side-effect of treatment.
After establishing the steady-state activity of the pathway in the absence of
inhibitors, we then we used computational optimization to identify a drug
combination that, at steady state, minimized squalene production, but left
geranylgeranyl diphosphate production the same as in the absence of inhibitors.