Here are some potential rotation projects for people interested in rotating in the lab.  Please contact me if any of these pique your interest and know that this is not an exhaustive list!

More Computational

Explore Epistasis in the Collaborative Cross

We have recently published an  article in which we look for evidence of epistasis (2 or more genes interacting to produce an effect) in a new way.  Instead of doing computationally intensive and statistically fraught combinatorial analyses of all pairs, triads, etc of possible interacting polymorphisms, we search for interactions between an individual SNP and a global genetic background measure derived from a defined founder individual or population.  This ‘polygenic epistasis’ grows stronger when more sites are interacting with one another.  The algorithm was built to work on the Recombinant Inbred panels of the HMDP or on populations called ‘advanced intercross lines,’ both of which share the fact that there are only TWO founder strains per population of mice.

UNC is the home of the Collaborative Cross, a mouse population which was designed to have 8 founders per mouse strain instead of the 2 for an RI line.

  • What changes might need to be made to the algorithm to make it work in the CC?
  • Do we see more or less evidence of epistatic effects when increasing the number of backgrounds?

Single Cell and/or Spatial Transcriptome Gene Networks

Gene co-expression networks help us to understand how groups of genes that vary between individuals act in concert with one another to create observable phenotypic differences.  We have successfully used these approaches to identify several key genes that play roles in heart failure as well as gain a deeper understanding of the pathways that are activated during heart failure progression.

Just as we can use differences across a population to explore gene networks, so too can we use differences across a tissue – how do variations in cellular neighborhoods and proximity to stressors lead to phenotypic differences?  How can we leverage the explosion of single cell and spatially-located data to improve our co-expression networks?

Spatial Transcriptomics and Cardiomyocyte Nuclearity in Healthy and Diseased Hearts

Most of the cells in our bodies have a single, diploid nucleus, yet the vast majority of cardiomyocytes in mammals are polyploid (Humans >70%) and/or multinucleated (Pigs, Rabbits > 95%).  The reasons for this are unclear, but recent studies have suggested that this multinuclearity may play a role in the protection of the heart from stress and is actually an adaptive response to the needs of the cardiomyocyte to provide a consistent, strong beat.   Single Cell and Single Nucleus RNAseq studies in cardiomyocytes published in the past year and a half have further suggested the evidence of a ‘nuclear code’ in which nuclei drawn from transcriptionally distinct subsets are combined to create the transcriptionally distinct cardiomyocyte transcriptomes.

We are waiting to get data back from a spatial transcriptomics company at which point we will be examining this problem in a myocardial infarction (heart attack) mouse model.  This rotation project would be exploring the results of this analysis:

  • Do we see evidence of gene expression changes as a function of distance from the infarction site?
  • Do we see evidence of differences in cardiomyocyte nuclearity as a function of distance from the infarction site?
  • Do we see evidence of shifts in the ‘nuclear code’ between healthy and failing hearts?
  • Do we see evidence of specialization occurring between the nuclei of a single cardiomyocyte?  (There is evidence of this happening in skeletal muscle cells)

Understand Factors that Contribute to Environmental Variability in GWAS

One benefit of the Collaborative Cross or Hybrid Mouse Diversity Panel genetic reference populations is that mice from the same strain are genetically identical to one another.  This allows us to do certain types of experiments that are impossible to do in human populations.  A potential rotation project involves studying intra-strain variation – the variation in phenotypes that occur within individual strains of genetically identical mice – to identify genes, pathways, and mechanisms that predispose individuals to greater susceptibility to daily environmental perturbations.

More Bench Science

Cell Free DNA Biomarkers of Heart Failure

Heart Failure is difficult to diagnose in its early stages as the incredible remodeling capacity of the heart means that even when it is under significant stress, there is no noticable phenotype to a patient.  To diagnose early stages of Heart Failure now requires an echocardiographic stress test that requires specialized machinery and trained technicians to operate – something not available to basically anyone at their regular doctor’s office.   Inspired by earlier work that showed that it was possible to detect islet cell death up to a year before other biomarkers detected the onset of type 1 diabetes by finding islet cell DNA released from dying cells floating in the blood, we are asking whether we can detect heart failure in mice with stressed hearts by finding cell free cardiomyocyte DNA in the blood.

In vitro Validation of Genes Associated with Heart Failure

We have identified a large number of exciting candidate genes that, in silico, appear to drive or prevent the progression of cardiac hypertrophy and other HF-associated phenotypes.  These genes span a wide variety of pathways and potential mechanisms.  In this rotation project, you would help select the genes you are most interested in validating and do so using primary isolated cardiomyocytes and fibroblasts, looking for effects on cell size, cell number and/or contraction rate.  Once we’ve identified a gene of interest in this way, we would then begin the process of understanding the mechanism by which it acts, including its subcellular localization, its effects on apoptosis, cell division, cell adhesion, etc.

Multinucleation of Developing Cardiomyocytes

Adult cardiomyocytes are highly multinucleated and polyploid, but this is not the case when you are born.  A rapid process of development that occurs just before and soon after birth rapidly transforms the heart, inducing multinucleation in most cardiomyocytes.  in vitro, we can model this progression using isolated neonatal cardiomyocytes paired with retinoic acid to force them to mature.  We are interested in exploring this process and identifying how genes we believe are associated with cardiomyocyte multinucleation affect this process.

Upcoming Projects

Upcoming projects in the lab that still need development but that you can and should discuss with Christoph if you are interested include research on scRNAseq in cardiomyocytes, cardiomyocyte nuclearity, and spatial transcriptomics.