Our research themes Coronary heart disease (CHD) is one of the global leading causes of death. In the UK, acute coronary disease syndromes cause c. 60% of all CHD-related deaths, leading to c. 240,000 hospitalizations each year. The vast majority of fatal CHD cases is directly related to thin-cap-fibro-atheroma (TCFA).

Atherosclerotic plaques, such as TCFA, do not present an even distribution across the arterial endothelium. Instead, they are located at predilection sites, such as side branches, curved segments and bifurcations, which are known to disturb several properties in the blood flow velocity field.

Early research on this topic suggested that biomechanical factors play an “initiation” role in atherosclerosis. Conversely, more recent studies indicate that biomechanical factors actually play a “progression” role in this pathological scenario.

In our group we are interested in studying atherosclerosis from a multi-disciplinary perspective.

To this end we combine several techniques:

  • Imaging of the arterial endothelium in atherosclerotic mice and pigs, through µCT, µMRI and confocal microscopy
  • Computational fluid dynamics (CFD), which allows us to characterize the velocity field of blood flow
  • Histology and transcriptomics of the arterial endothelium, which in conjunction with our imaging techniques allows us to create 3-D mechanical-histological-genomic reconstructions of arterial segments
  • Genetic network inference and analysis, i.e. systems biology of the blood vessel wall, using computational methods and high-throughput gene expression analytical techniques
  • Mammalian synthetic biology of small networks, with the objective of creating tools and therapies for the management of atherosclerosis.

Our research

Biomechanics

Biomechanics

Biomechanics Biomechanics is the study of the behaviour of biological materials under an applied load. The goals are to understand (1) mechanical phenomena within the body and (2) the influence of the tissue mechanical environment on cellular function, during normalcy and disease.

The power of biomechanics lies within its ability to describe complex material behaviours under specified conditions of interest. While there are numerous applications for this capability in medicine, herein we focus on the role of biomechanics in the development and instability of atherosclerotic plaques. One important translational aspect to this work is the development of diagnostic tools to predict plaques vulnerable to rupture in patients.

Most biomechanics work in our group is focused on quantifying the wall and blood flow mechanics of the normal and diseased vessel to understand how tissue mechanical perturbations correlate with the development and progression of atherosclerosis.

To do this, we employ in vivo imaging, mechanical modelling (solid, fluid, and fluid-structure interaction) using commercial software such as Abaqus and COMSOL, and quantitative physiology techniques such as 3-D histology using in-house developed MATLAB software. This approach allows us to identify novel mechanical metrics that accurately predict local regions of disease development within the vessel intima.

In addition, in our mouse model of atherosclerosis, dysfunctional endothelial cells from regions of perturbed biomechanics can then be isolated to study their pro-atherogenic genomic profile.

Research Projects

  • In vivo mechanobiology studies seek to induce perturbed hemodynamics via an implanted arterial device within hypercholesterolemic murine and porcine models of atherosclerosis. The vessels are then imaged in vivo over the time-course of plaque development, shear stress is computed with computational fluid mechanics, 3-D histology is performed, and metrics of shear are correlated to histological markers of atherosclerosis. In the mouse model, local regions of high overlap between perturbed shear and atherosclerosis, particularly plaques of a vulnerable phenotype, are used to identify and isolate dysfunctional endothelial cells for RNA sequencing.

  • In vitro mechanobiology studies of endothelial cells are performed within a stretching device and a flow chamber to study how the solid wall and blood flow mechanical environments promote pro-atherogenic cell signalling pathways.

Biomechanics

Genomics

Genomics (Systems Biology)

Genomics Cells react to their environment with a modification of their gene and protein signatures. This is regulated in their DNA, which is packed inside the nucleus of eukaryotic cells, of which our mammalian cells are a subset. The human genome consists of 20,000 to 25,000 genes and its size is not so much different from other mammals, plants and prokaryotic cells.

What makes the human genome different is its interactome (see next tab). About 10% of genes are actively transcribed, resulting in signatures of about 2,000 mRNA per cell. The number and composition of this mRNA pool depends on the environmental cues the cell encounters.

An additional level of complexity is present, because the diseased vessel wall consists of a variety of cell types – each with their dynamic genomic signatures - which vary over time due to a complex interplay of chemokine, cytokine and lipid deposition, a process called lipid-driven inflammation. Approximately 8 cell types are present inside the atherosclerotic plaque and their amount varies with time.

Since, we are primarily interested in how blood flow regulates plaque formation; we have developed methods to isolate endothelial cells from diseased blood vessels. To that end, a robot-driven laser capture was programmed to detect endothelial cells from tissue cross sections, and the laser isolated the cells, subsequently catapulting each cell into a vial.

This non-contact technique enabled us to isolate their RNA. As only 5% of total RNA, the next step is to isolate the mRNA from the total RNA pool. In order to measure these changes in the mRNA pool one need to have very precise measurements, as the large size of the gene pool has a mass of only 10 picograms (1 picogram = 10-12 gram). Equipment available in Imperial College (Bioanalyzer, Qbit) enabled to measure their content.

Next, we transformed the mRNA pool into complementary DNA or cDNA. The resulting cDNA library will be used to be analysed by a new RNA sequencer (Illumina 5700) which enables to screen the composition of the library in terms of base composition. This pipeline currently needs 1000-2000 endothelial cells as input.

A recent discovery in our group was that endothelial cells are increasingly heterogenic during the development of the plaques. This added a new level of complexity as it means we have to obtain sufficient RNA from smaller numbers of endothelial cells as currently possible.

Cell heterogeneity is considered an important phenomenon in many diseases, and our discovery adds new information to these observations. As stated above measuring the size of the mRNA pool in single cell needs specialised equipment as only 10 pico-gram of mRNA is present per cell. Recently, a new field of microfluidics has been developed that enables to measure single cell genomics.

Single cell microfluidics uses water droplets in oil to capture individual cells into tiny compartments, creating a two-phase system. These small compartments reduce the need for large amount of material and minimize diffusion distances, reducing cost and increasing single cell analysis speed.

We have recently developed microfluidics unit which might capture cells obtained from the laser capture device. Capture efficiencies varied between 40-70% depending on flow of the water phase relative to the oil phase. We are now interested in developing ways to merge droplets as to lyse cells, destroy DNA, rRNA, tRNA and in this way create a pipeline for cDNA library generation on a chip.

Available research projects

  • Optimization of RNA isolation techniques and cDNA libraries
  • Developing of new microfluidics systems
  • Developing new tools for single cell cDNA library generation

Genomics

Bioinformatics

Bioinformatics

Bioinformatics The large genomic signatures present in individual cells are organised in gene networks, and this field is often termed interactome.

The size of the human interactome - in contrast to that of the human genome - is vast and much higher than that of other species. It seems that the biological complexity is better reflected in measurements of the interactome than in size of the human genome or transcriptome.

The interactome is a reflection of the interactions of individual genes into large networks. The inference of these gene networks from high throughput data, like RNA sequencing has shown a large interest and development over the last decade.

As part of an ongoing collaboration, we are currently developing new gene network inference methods, based upon abductive logic (in close collaboration with Dr. Russo from department of Computing). This gene inference method has shown to be more accurate in detecting complex gene networks from high throughput techniques than other, existing statistical methods.

We are testing this network inference method now by generating a robust meta-set of mechanosensitive gene network, based upon our work and that reported in the literature. This robust data set will be used as input for the gene inference technique. The outcome of the gene inference code, which consists of a list of possible gene networks needs to be tested and evaluated.

To that end we have developed a novel highly accurate, highly modular, high throughput platform which is capable of performing 2000 independent experiments. This platform allows a systematic testing of the list of possible gene networks in a high throughput manner. We are currently aiming at finding ways for the gene inference software to predict the experiments to discriminate between the several gene networks.

Available research projects

  • Developing software for analysis for RNA sequencing
  • Developing software for merging of platforms
  • Developing software for network analysis
  • Coupling network to dynamical systems
  • Coupling dynamical systems to synthetic biology

Bioinformatics

Synthetic biology

Synthetic Biology (Therapies)

Synthetic Biology It is anticipated that the genomics-bioinformatics module will discover that several signalling modules are "diseased" and we are exploring novel techniques to "heal" these diseased modules.

What could happen is that entire signalling modules or parts are down-regulated. As a consequence, the sub-module may become rewired partially bypassing the diseased pathway. Or, other networks may become up-regulated to compensate for the loss of another pathway.

To restore entire signalling modules that are disrupted, we are developing methods to replace these modules applying synthesis networks. However, when parts of the network are down regulated than we aim to rebalance the network with smart interventions, either replacing single gene-single protein, or interfering with multiple sites in the network. The choice on where to interfere strongly depends on a comprehensive flux balance analysis of the entire network.

In order to perform these actions we borrow techniques from the synthetic biology. Synthetic biology aims at designing, categorizing and implementing custom biological parts that can be used to elicit controllable functions in a cellular environment. Synthetic gene networks can highly contribute to the diagnosis and treatment of a wide variety of diseases, including atherosclerosis.

Available research projects

  • Designing a siRNA network
  • Designing an assay for the activity of a mechanosensors
  • Constructing a CripsR/Cas9 stable cell line
  • Constructing an endothelial microRNA detector
  • Constructing an hypoxia detector for endothelium

Synthetic Biology