Sensitive, High-throughput Single-cell Secretion Analysis via a Living Cell Surface Immunosorbent Assay for Metabolic Engineering

Abstract

Yeast has long been a cornerstone of synthetic biology and metabolic engineering, serving as a versatile chassis for the sustainable production of pharmaceuticals, biofuels, and fine chemicals. Yet a persistent challenge in advancing yeast-based biomanufacturing is the ability to efficiently screen and isolate superior strains from vast genetic libraries. Conventional approaches—including microplate-based assays and droplet microfluidics—are constrained by fundamental limitations. Microplate assays are typically built upon genetically encoded biosensors or colorimetric readouts, which provide only indirect and moderately sensitive measurements of extracellular metabolite secretion. Droplet-based biosensors, in turn, are often restricted to specific metabolites and suffer from high background noise, compromising detection sensitivity. Both approaches are further limited by low screening throughput and slow sorting speed, restricting their applicability to large-scale library analysis. These bottlenecks have hindered the effective implementation of directed evolution and the optimization of complex metabolic pathways in yeast.

To address these challenges, this thesis develops a living cell surface immunosorbent assay that enables sensitive and ultrahigh-throughput single-cell secretion analysis for strain engineering and directed evolution. The work is structured into three main parts.

First, the design, fabrication, and characterization of the sensing assay were systematically investigated using Saccharomyces cerevisiae as a model (Chapter 3). Aptamer-based molecular probes were immobilized on the yeast surface through a biotin–streptavidin strategy, forming dense and uniform coatings (~1.4 × 10⁷ probes per cell). A unique feature of the system is its selective anchoring on mother yeast cells, with negligible transfer to daughter cells during budding, ensuring long-term stability across multiple generations for up to three days. Flow cytometry confirmed that the assay sensitively quantified low-abundance extracellular analytes—including small molecules and proteins—with detection limits below 1 μM. Time-lapse confocal imaging further revealed rapid capture of secreted ATP at the single-cell level. These results establish the sensing platform as a robust, flexible, and highly sensitive platform for single-cell secretion monitoring.

Second, the assay was applied to demonstrate high-throughput screening and rapid sorting of large yeast populations (Chapter 4). By releasing cells from droplets prior to cytometric analysis, the system bypassed the fundamental limitation of inefficient single-cell encapsulation. This innovation enabled the screening of libraries exceeding 10⁷ cells in a single run and sorting at standard cytometry rates of 10³–10⁴ events per second, representing more than a 30-fold improvement over conventional droplet-based approaches. Capturing metabolites directly on the cell surface further allowed resolution of distinct secretion profiles within heterogeneous populations. In addition, multiplexed functionalization with different aptamers enabled simultaneous monitoring of multiple metabolites and precise identification of rare subpopulations in mixed cultures. Collectively, these advances demonstrate that the living cell surface sensor integrates quantitative single-cell secretion profiling with ultrahigh-throughput screening and rapid sorting, overcoming key limitations of existing methods.

Third, the platform was applied to the directed evolution of metabolic enzymes and transporters in the vanillin biosynthetic pathway (Chapter 4). Screening of an error-prone PCR library of enoyl-CoA hydratase/aldolase (ECH) yielded high-producing variants, with the best strain achieving a 2.7-fold increase in vanillin output compared with the parental strain. Sequencing revealed four substitutions (I90N, Y169C, N212Q, and P213L) that enhanced enzyme stability and catalytic efficiency. In parallel, screening of more than 2 × 10⁷ transporter variants identified top-performing mutants, including a variant carrying K40R and V79D substitutions, which significantly improved transporter efficiency. These results highlight the versatility of the assay in evolving both catalytic enzymes and membrane transporters.

In summary, this thesis establishes the living cell surface immunosorbent assay as a versatile and generalizable strategy for sensitive, multiplexed, and ultrahigh-throughput analysis of extracellular secretion at the single-cell level. Its successful application to the directed evolution of both metabolic enzymes and membrane transporters in the vanillin biosynthetic pathway demonstrates its broad utility in synthetic biology and industrial biomanufacturing. Beyond vanillin production, this platform can be readily extended to other metabolites and secreted proteins, as well as to diverse host systems, enabling more comprehensive screening of secretion yield and quality. As such, the living cell surface immunosorbent assay provides a powerful foundation for strain optimization, precision biomanufacturing, and the development of robust microbial cell factories for sustainable bio-based production.

Date of Award	21 Jan 2026
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Chia-hung CHEN (Supervisor)