Mary Kraft

Transport, phase-separation, and self-assembly in biological membranes

Kraft Mary_2015
Mary Kraft

Cells are tiny but complex bioreactors; the concentrations and proximities of the biomolecules within them determine which biochemical reactions will occur, which collectively determines cellular function. We are developing new approaches in which compositional signatures acquired from individual cells are used to understand and predict biological function. We use these techniques for applied research, such as detecting stem cell differentiation for tissue engineering, and for basic research on the roles of plasma membrane organization in influenza virus replication and other important biological processes.

Kraft Research Group Website Mary Kraft Faculty Profile

Single Cell Analysis for Tissue Engineering


All of the body’s blood and immune cells, which are called hematopoietic cells, are produced from the differentiation of hematopoietic stem cells (HSCs) in the bone marrow. Identification of the combinations of biochemical and biophysical cues that direct HSCs to self-renew or differentiate into distinct cell lineages would permit expanding specific blood or immune cells ex vivo for the treatment of blood cell diseases.

The scarcity of HSCs in the body has motivated the use of high-throughput screening platforms that minimize the number of rare HSCs required to screen cell response to stimuli. These platforms, and the heterogeneity of the HSC populations isolated from bone marrow, has increased the need for identifying the fate decisions of individual HSCs with location specificity. Current methods to assess the differentiation status of individual cells with location specificity have unsatisfactory reproducibility, and often employ expensive labels that may affect cell fate decisions.

We are developing an approach that uses confocal Raman microspectroscopy to acquire biochemical data from individual, living, unlabeled cells with location specificity. Because the biochemical compositions of HSCs and their more mature progeny are very similar, the challenge is to identify the Raman spectral features that encode for cell phenotype. We are addressing this challenge by developing multivariate classification models to identify the differentiation-associated combinations of peaks in the Raman spectra that can be used to identify cell phenotype. We are validating our approach in combinatorial biomaterial microarrays designed to identify the effects of cues on early HSC fate decisions.


This work will provide tissue engineers with a quantitative, objective, and label-free approach to accurately identify HSC fate decisions on the single cell level and with location specificity. This approach will greatly facilitate correlating early fate decisions to the stimuli that induced them, critical for the development of in vitro platforms for directing HSC fate.

Plasma Membrane Organization and Cell Function

The cell plasma membrane, the selectively permeable barrier that surrounds every cell, plays a critical role in both healthy biological processes and disease. The cell plasma membrane mediates the interactions between a cell and its surroundings, including cell adhesion, cell-cell recognition, nutrient uptake, and signal transduction. To coordinate these functions, the plasma membranes of mammalian cells are organized into distinct domains of differing protein and lipid composition. Although the plasma membrane is composed of hundreds of different lipid and protein species, many scientists believe that interactions between just a few components, namely cholesterol and sphingolipids, are very important for this organization.

Prior to our work, the cholesterol and sphingolipid distribution in the plasma membrane was a mystery due to a lack of methods to image cholesterol and sphingolipids in membranes without using fluorophores that may alter the lipid’s distribution and function. We have overcome this obstacle by using high-resolution secondary ion mass spectrometry (SIMS) to image metabolically incorporated, stable isotope-labeled cholesterol and sphingolipids in the plasma membrane with better than 100 nm lateral resolution. Our published work established that the plasma membranes of fibroblast cells contain sphingolipid domains that are not enriched with cholesterol, but are dependent on the cytoskeleton. This work provides a completely new understanding of plasma membrane organization.


We are now combining this technique with traditional biochemical methods and complementary imaging modalities to currently intractable biomedical problems that involve lipid-mediated biological function. One such research project is described below.

Influenza Virus Infection

The flu is a potentially life-threatening respiratory infection that is caused by the influenza virus. The influenza virus must enter lung cells and use the cell’s machinery to replicate. Towards the end of this process, the newly synthesized viral components are transported to the plasma membrane, where they assemble into new virus particles. These new virus particles are enclosed within a fragment of the plasma membrane, forming the viral envelope, when they are released from the host cell.

Intriguing features of the influenza virus are that although the plasma membrane contains many different types of lipids, the viral envelope is enriched with just a few, namely cholesterol and sphingolipids. Additionally, reductions in the cellular cholesterol or sphingolipid levels are detrimental to influenza virus replication. Current hypotheses attribute this cholesterol- and sphingolipid-sensitive influenza replication to a mechanism in which the influenza virus exploits cholesterol- and sphingolipid-enriched domains in the host cell’s plasma membrane for assembly and budding.


We are using our unique ability to image specific lipids and proteins of interest with high-resolution SIMS to test the hypothesis that the influenza virus assembles and buds from plasma membrane domains that are enriched with cholesterol and sphingolipids. We are also combining this new imaging technique with traditional biochemical methods to elucidate the roles of the host cell’s cholesterol and sphingolipids in influenza virus replication. This work may enable the development of more effective anti-influenza therapeutics.