In vivo and in silico mapping of cell-cell interactions in the haematopoietic stem cell niche

Every second of our adult lives we produce some 2 million red blood cells; over the course of a day in total 10^12 new blood cells are generated in healthy individuals. The process which produces this extraordinary number of cells begins in the bone marrow where a specific set of stem cells, so called haematopoietic stem cells (HSCs), reside. These cells give rise to all the different cells that make up our blood and the immune system. Understanding of this process is of fundamental importance, and the ability to rationally affect the dynamics of the haematopoietic system will also have major consequences for ageing research, regenerative medicine and clinical haematology.

In order to function properly, HSCs rely on the support of other cells inside the bone marrow, but these are as yet not known with certainty. In the proposed research we will use experiments in mice to determine which cells allow HSCs to function properly. The experiments we will conduct use a very powerful type of microscopy allowing us to visualise the HSCs inside the bone marrow of living mice. We will obtain high-resolution 3D images over time showing the location of HSCs and the identity of their neighbouring cells. We will use computational method to perform several measurements on these images and the resulting data will be analysed statistically and provide the basis for the development of computer models of the cells and their interactions inside the bone marrow. We can run these computer models to simulate the events happening in vivo and by comparing the simulation output with the experimental data we will be able to test, improve and validate our understanding of the cellular interactions responsible to support the stem cells in the bone marrow. The mathematical model here serves as a summary of our understanding of the mechanisms acting within the bone marrow; any disagreement between the simulations and the observed data points to gaps in our understanding and will motivate further analysis.

Based on preliminary analyses and modelling we will then study mice that lack certain cell-types. This in turn will provide us with more detailed insights about the effect that these cells have on the fate of HSCs. In addition to measuring the spatial distributions of cells in different types of bones, we will also determine the differences in gene expression in HSCs that result from deletion of other cell types.

Finally, we plan to use new microscopy-based techniques which allow us to directly kill individual cells in the neighbourhood of HSCs and study the response of HSCs to such perturbations. If we can successfully predict that HSCs migrate towards other cells of the same type as the deleted cell then this would substantially increase our confidence in our models.

Computer models of the cell population dynamics inside the HSC niche will be used to systematically probe our understanding; but they can also be used in the future to replace experiments in mice.

Correct functioning of haematopoietic stem cells (HSCs) depends on their interaction with complex niches in the bone marrow (BM) and the question is open whether different BM cell combinations form functionally distinct niches. Endosteal and perivascular niches are suggested to support quiescent and active HSCs respectively, but a clear comparison between the two is missing. We propose to combine sophisticated mouse genetics, cutting edge in vivo imaging and statistical and computational analyses to map the cellular interactions between HSCs and other mouse bone marrow cell types in the BM space. We will use statistical modelling to determine how different cell types maintain healthy HSCs and test the hypothesis that the lineage of niche cells is critical for their role. We will complement live imaging of HSC niches in the calvarium BM with immunohistological analysis of cross sections from both calvarium and femur to further expand the array of identifiable niche cells and to compare HSC niches located in anatomically distinct environments. Mathematical modelling will implement mechanistic hypotheses computationally, and we will test the fit of the resulting predictions with the experimental observations. We will use genetic approaches and photo-ablation to eliminate both entire niche lineages and individual/few niche cells and we will use intravital imaging to monitor the consequences on HSC behaviour. Transcriptomic analysis on purified HSC populations will determine how interactions (or lack thereof) with different cell types affect gene expression in HSCs. By comparing our results with known signatures of healthy, expanding or impaired HSCs we will investigate the cause-effect link between the nature of HSC-niche interaction and HSC function. With this work we will obtain a comprehensive map of the cellular processes contributing to HSC maintenance and enabling haematopoiesis in healthy individuals, which are lost when haematological diseases develop.

Multi-scale problems and in vivo analyses are all-pervasive in biology and especially in biomedical research. The biggest immediate and mid-term impact of this research is the development of an integrative framework for the quantitative analysis of the in vivo haematopoietic stem cell (HSC) niche. There is tremendous scope for applying such tools in fundamental and applied biological and biomedical research. The HSC niche is of direct biomedical importance, and maintenance of a healthy niche environment is also pivotal for healthy ageing (in humans as well as animals).

Other application areas include:
- tissue engineering and stem cell biology;
- regenerative medicine.

To maximize short-term impact we will release software in a suitable licensing framework that will allow easy and free access to academic stake-holders. The commercial scope for such software will be explored with the Imperial College Technology Transfer office, but free access to academic researchers is important to us. We will also disseminate all the in vivo imaging and niche modification protocols through publications and direct teaching. All data sets generated from intravital microscopy, immunohistological analysis and transcriptomic analysis will be deposited in appropriate, open access databases.

Understanding the factors that control the health of the HSC niche is of fundamental importance and many of the major implications will be realized only over longer time-scales. In the medium term we will, however, also discuss the application of our findings with clinical researchers and stakeholders in the biomedical sector, including cancer charities. Over shorter time-scales we hope to use our in silico models of the HSC niche for the 3R purposes and we will investigate and maximize the likely impact in conversations with the NC3Rs.

In addition we will address the distinct lack of individuals trained and conversant in both computational and laboratory techniques. The need for such individuals in academia and industry is likely to increase and a suitably trained individual will be able to lead innovative new research programmes. It is one of our essential aims to aid the RAs to become recognized researchers at the wet/dry interface of stem cell systems biology.