Dr. Irina Kufareva, Ph.D

In health, disease, and therapy, processes that happen to humans span a wide range of scales. On the one hand, the basis for everything is in the miniscule atomic interactions governed by the first principles of physics; consequently, three-dimensional atomic-resolution molecular structure encodes in itself all the answers. However, compared to the entire proteome, structures are still a few, and even for structurally characterized proteins, extracting answers to the questions of function is nontrivial. On the other hand, even at the level of a single cell (let alone a whole organism), the tangled mess of molecular interactions is so complex that it becomes impossible to tackle with purely experimental means.

The overarching goal of our lab is to elucidate structural, molecular, and architectural principles of tumor and immune cell responses to stimuli and drugs via computational approaches. GPCRs and G proteins, the two key classes of cell signaling molecules, are the main focus of our work.

Computational structural biology of GPCRs and G proteins. Conformational plasticity of proteins, or rather the lack of methods for prediction of conformational changes that are relevant for interactions of these proteins with their ligands and effectors, are important barriers in modern computational structural biology. We develop methods for accurate computational prediction of transient interactions of proteins and chemicals with conformationally variable protein interfaces, and apply these approaches to important cellular targets. In the past, in collaboration with the Handel lab and others, we elucidated the structural basis of ligand binding and signaling in chemokine GPCRs CXCR4 and ACKR3 (both playing a central role in progression and metastasis of numerous cancers, PNAS 2014, Science 2015, PNAS 2015, Nat Comms 2017, PLoS Biol 2020, Sci Signal 2020), CCR5 (an HIV co-receptor, Immunology 2017), and CCR2 (a promising target in inflammation, autoimmunity, and immune-oncology, Nature 2016). We also made key contributions to understanding non-receptor activation of heterotrimeric G proteins, and have solved the first structure of a G protein complexed with one such activator (PNAS 2019) in collaboration with the Ghosh and Chang labs.

Computational systems biology of GPCRs. The new direction of network-based modeling and reverse-engineering of cell signaling has been initiated in the lab only recently, but has already led to exciting findings about signaling cascades downstream of several chemokine receptors. The modeling process is based on automated system identification that is informed by multiplexed datasets obtained through spatiotemporally resolved interactomics, phosphoproteomics, or Luminex multi-analyte profiling. This work seeks decoding the principles of biological information transfer and processing, and will inform the design of multimodal therapeutic strategies in cancer and inflammatory diseases.