Expression of a humanized viral 2A-mediated lux operon efficiently generates autonomous bioluminescence in human cells

Abstract

Background: Expression of autonomous bioluminescence from human cells was previously reported to be impossible, suggesting that all bioluminescent-based mammalian reporter systems must therefore require application of a potentially influential chemical substrate. While this was disproven when the bacterial luciferase (lux) cassette was demonstrated to function in a human cell, its expression required multiple genetic constructs, was functional in only a single cell type, and generated a significantly reduced signal compared to substrate-requiring systems. Here we investigate the use of a humanized, viral 2A-linked lux genetic architecture for the efficient introduction of an autobioluminescent phenotype across a variety of human cell lines.

Methodology/principal findings: The lux cassette was codon optimized and assembled into a synthetic human expression operon using viral 2A elements as linker regions. Human kidney, breast cancer, and colorectal cancer cell lines were both transiently and stably transfected with the humanized operon and the resulting autobioluminescent phenotype was evaluated using common imaging instrumentation. Autobioluminescent cells were screened for cytotoxic effects resulting from lux expression and their utility as bioreporters was evaluated through the demonstration of repeated monitoring of single populations over a prolonged period using both a modified E-SCREEN assay for estrogen detection and a classical cytotoxic compound detection assay for the antibiotic Zeocin. Furthermore, the use of self-directed bioluminescent initiation in response to target detection was assessed to determine its amenability towards deployment as fully autonomous sensors. In all cases, bioluminescent measurements were supported with traditional genetic and transcriptomic evaluations.

Conclusions/significance: Our results demonstrate that the viral 2A-linked, humanized lux genetic architecture successfully produced autobioluminescent phenotypes in all cell lines tested without the induction of cytotoxicity. This autobioluminescent phenotype allowed for repeated interrogation of populations and self-directed control of bioluminescent activation with detection limits and EC50 values similar to traditional reporter systems, making the autobioluminescent cells amenable to automated monitoring and significantly reducing the time and cost required to perform bioluminescent workflows.

Figure 1. Schematic illustrating the construction and evaluation of autobioluminescent human cell lines.

(A) The luxCDABE genes of Photorhabdus luminescens and the frp oxidoreductase gene of Vibrio campbellii were codon optimized for expression in human cells and linked using unique viral 2A elements. Intervening stop codons were removed, leaving only the final stop codon (red octagon) at the 3′ end of the frp gene to create a single, continuous ORF. (B) The human optimized lux operon was then cloned under the control of either a CMV or EF1α promoter and transiently transfected into HEK293 (kidney), T-47D (breast cancer), and HCT116 (colorectal cancer) human cell lines, resulting in an autobioluminescent phenotype for each cell type. (C) Expression of the 2A-linked lux cassette under control of the CMV promoter produced significantly (p≤0.05) higher maximal autobioluminescent output levels in each of the cell lines relative to EF1α-based expression, making it an improved choice for the development of constitutively autobioluminescent cell lines. The CMV-based cassette was then stably transfected into the same parent cell lines and the clonal lines with the highest levels of autobioluminescent output were selected for further analysis.

Figure 2. Application of autobioluminescent human cells as target-responsive bioreporters.

(A) Human T-47D breast cancer cells naturally express estrogen receptors and proliferate in response to increased exposure to estrogenic compounds. When treated with 17β-estradiol, T-47D cells expressing the pCMV_Lux plasmid displayed increased growth rates compared to untreated control cells by day 3 for all treatments ≥1 pM, although this trend was only maintained throughout the full 6 day assay period at treatment levels ≥10 pM (inset). (B) A dose-response relationship was determined between 17β-estradiol and autobioluminescence, with an EC₅₀ value (10 pM) similar to that of traditional proliferation assays and alternative bioluminescent reporter systems. (C) When the human optimized lux operon was placed under the control of a doxycycline responsive Tet-On promoter, HEK293 cells could autoreport detected concentrations ≥5 ng/ml. (D) An up-regulation of lux gene mRNA transcript levels was observed upon doxycycline treatment, indicating that the increase in autobioluminescent production was due to activation of the Tet-On promoter and not the result of non-specific treatment effects.