UC Berkeley

Nucleic acid measurements and protein measurements have transformed biological inquiry, providing key insight into biological processes and disease mechanisms. Specifically, single-cell multimodal -omics assays that measure proteins and nucleic acids are becoming increasingly important to elucidate cell-to-cell heterogeneity and explore the relationships between classes of biomolecules in complex biological signaling pathways. Despite tremendous progress over the last decade, many measurement gaps still exist due to the poor sensitivity and specificity of currently available -omics assays. For example, currently available single-cell multimodal measurement tools lack the sensitivity and specificity to detect nucleic acid point mutations and protein isoforms in the same cell. In-gel -omics assays exhibit great promise to overcome these limitations due to the advantageous properties of hydrogel matrices, which result in improved sensitivity, specificity, and resolution. However, fundamental characterization of in-gel assays is lacking. The work presented in this dissertation is aimed at bridging measurement gaps in single-cell in-gel assays by (1) evaluating and characterizing various approaches for improved multiplexing and sensitivity of in-gel assays, and (2) developing novel single-cell multimodal electrophoretic assays to measure co-expression of proteins and nucleic acids.

First, we evaluated in-gel quenching of fluorophores with H2O2 as a strategy aimed at improving the multiplexing capability of in-gel immunoassays that employ fluorescence-based readouts. Significant quenching of commonly used fluorophores was observed, leading us to conclude that H2O2-based fluorophore quenching is a promising strategy for multiplexing of in-gel immunoassays. We then characterized the reaction kinetics and transport limitations of an in-gel fluorescence-based DNA readout and evaluated DNA sequence design parameters for signal amplification to develop assay design rules. When fluorescently labeled nucleotides were incorporated into single-stranded DNA overhangs immobilized in polyacrylamide hydrogels, we observed that the time for reaction completion was much greater than predicted by simple diffusion and reaction models. We attributed this discrepancy to retarded diffusion of the DNA polymerase resulting from interactions between the diffusing polymerase and the immobilized DNA. We also investigated the effects of the number of labeled nucleotides incorporated per strand and the spacing between labeled nucleotides on signal amplification. Consistent with our hypotheses, increasing the number of and spacing between incorporated labeled nucleotides resulted in increased fluorescence signal. Ultimately, our results that highlight retarded diffusion behavior of the DNA polymerase and the effects of DNA sequence design on the degree of signal amplification will inform assay design for future in-gel assays that depend on DNA polymerase elongation-based fluorescence readouts.

Next, we characterized hybridization chain reaction (HCR) in polyacrylamide gels of different densities, evaluating the degree of signal amplification achieved, the reaction kinetics, and the concentration-dependent fluorescence response. A high degree of signal amplification with a linear concentration-dependent response is crucial for improved sensitivity of in-gel assays. Our system scrutinized HCR in polyacrylamide gels that contained HCR initiator that was co-polymerized into the gels via a 5’ acrydite group. HCR resulted in three orders of magnitude of signal amplification when compared to hybridizing a fluorescently labeled single-stranded DNA oligomer. We also observed that reaction completion was achieved in 24 h across all gel densities and that in-gel HCR followed pseudo first-order DNA hybridization kinetics. Our results demonstrate that HCR is a promising signal amplification strategy across different gel densities. We then explored immuno-HCR as a strategy for improving the sensitivity of in-gel immunoassays. Immuno-HCR relies on DNA-antibody conjugates to convert protein information to DNA information for signal amplification via HCR. We began by studying the binding affinity of DNA-antibody conjugates with optical biolayer interferometry (BLI) and the partitioning behavior of DNA-antibody conjugates in polyacrylamide gels with confocal microscopy. Both the binding affinity and the partition coefficient of antibodies decreased in response to DNA conjugation, consistent with the theory that DNA conjugation results in molecules with larger hydrodynamic radii. After characterizing the binding and partitioning behavior of DNA-antibody conjugates, we applied immuno-HCR to detect proteins immobilized in a polyacrylamide gel. The result was poorer performance than standard immunoprobing with fluorescently labeled antibodies, which we attribute to an insufficient degree of signal amplification from linear HCR. Branched HCR methods that result in additional signal amplification have the potential to overcome the limitations that we observed, leading us to posit that immuno-HCR is potentially a feasible strategy for improved sensitivity of in-gel immunoassays.

Finally, we presented the concept for two assays that seek to fill the measurement gap that precludes co-detection of protein isoforms and nucleic acid point mutations in single cells. We first developed Single-Cell Protein and RNA Electrophoresis (scPREP), a single-cell multimodal in-gel electrophoretic assay for co-detection of beta-actin mRNA and protein. Our approach involved incorporating mRNA detection via in situ hybridization with padlock probes (PLPs) and rolling circle amplification (RCA) into the single-cell western blotting platform, which enables highly sensitive and specific single-cell protein measurements. With scPREP, we detected beta-actin mRNA molecules as fluorescent puncta and beta-actin protein as gaussian bands in the gel. Our preliminary results are a first step toward co-detection of nucleic acid point mutations and protein isoforms from single cells, as our chosen mRNA detection modality exhibits single-nucleotide specificity, and our protein detection modality is capable of identifying protein isoforms. Future work to expand scPREP has the potential to make a significant contribution to the -omics field and to biology. We concluded by presenting the concept for an electrophoretic assay with a qPCR readout for detection of mRNA from lysed and electrophoresed single cells. Future work is needed to demonstrate application of the concept we outlined. While the initial motivation for this assay was for use as a characterization tool to study mRNA electrophoresis and capture in polyacrylamide gel, there is potential for use as a standalone assay. Overall, both assays have the potential to introduce novel measurement capabilities.

In total, the work presented in this dissertation advances the field of single-cell -omics by providing unique insight into the behavior of chemical reactions in hydrogel-based systems and presenting novel assay concepts for co-detection of proteins and nucleic acids from single cells. We ultimately aim to be able to measure co-expression of protein isoforms and nucleic acid point mutations in single cells with the technologies introduced here. Future studies that build off the results we present in this dissertation have the potential to further enhance the knowledge of in-gel assays and solve complex problems that will unlock new fields of biological inquiry.