UC Berkeley

Understanding cell-to-cell phenotypic heterogeneity is crucial for elucidating the biological mechanisms of multicellular organisms. Cytometry, or the measurement of characteristics of cells, encompasses a myriad of tools and assays that measure properties from cell size and morphology to expression levels of mRNA or protein. Conventional protein measurement assays, however, generally lack the analytical sensitivity required for single-cell analysis. To fill this gap, microfluidic design utilizes length scales and time scales that enable performing perturbations and measurements at sub-cellular resolution. In this dissertation, we describe the design, development and optimization of electrophoretic cytometry tools that advance protein and nucleic acid measurement capabilities for single-cell and single-embryo analysis (from 1 to 100 s of cells).

We first examine the technical noise and reproducibility of the single-cell polyacrylamide gel electrophoresis assay, or single-cell PAGE. We then establish a novel method for quantification of surface receptor protein localization by integrating an upstream surface immunostaining step. We detect a shift in electrophoretic migration of the antibody-antigen complex, and demonstrate agreement of results between our assay and standard methods for measuring cell surface proteins (flow cytometry and immunofluorescence).

We then develop a high-specificity, multiplexed single-cell immunoblot for screening of primary, uncultured smooth muscle cells (SMCs) for a panel of maturation markers. An emerging theory for life-threatening vascular remodeling involves the existence of a subset of SMCs with immature-like phenotype. To scrutinize healthy vessels for immature-like SMCs, we first use numerical simulations to optimize microwell volume in order to minimize initial protein dilution at the in situ lysis step. We then perform electrophoresis for separation of the 34 to 227 kDa molecular mass range of target markers, and identify a subpopulation (< 10%) of immature-like SMCs. Our results support the recently-established mechanism that only a subset of immature-like SMCs is responsible for vascular remodeling. 

Next, we focus on questions regarding the complex mechanisms governing mammalian development, such as the embryonic stage at which blastomeres first exhibit cell-fate commitment. By integrating embryo-specific sample preparation and single-embryo handling with scaled-up microfluidic immunoblotting designed for murine embryos, we measure proteins from all stages of mouse preimplantation development, from individual zygotes to single blastocyst blastomeres. Despite a lack of highly selective immunoreagents, we effectively interrogate inter-embryonic heterogeneity of embryo-specific isoforms involved in RNA-mediated gene expression. In dissociated two-cell and four-cell blastomeres, we detect significant inter-blastomeric variation in proteins implicated in cell cycle regulation, suggesting blastomere heterogeneity even in the earliest multicellular stage of preimplantation embryos. Further, 20-30 embryos recovered from a single mouse are sufficient for statistically relevant analyses, eliminating the need for sample pooling in preimplantation development studies.

Finally, we develop a novel method for dual protein-nucleic acid measurements from low starting cell numbers (from 10 s to 100 s of cells) aimed at assessing whether specific modifications in genomic DNA or alternative splicing events in preimplantation embryos translate generate different protein isoforms. To achieve the dual measurement, we fractionate cells into cytoplasmic and nuclear compartments, where cytoplasmic proteins are analyzed by electrophoretic separations and nuclei are retrieved for off-chip nuclei acid measurements. We demonstrate that the single-cell PAGE protein signal correlates strongly with protein expression prior to lysis, and measure both mRNA and DNA from retrieved nuclei. This method shows promise for determining whether the abundant splice variants and DNA modifications in preimplantation embryos translate to protein isoforms.