A drug-selectable acoustic reporter gene system for human cell ultrasound imaging

Abstract

A promising new field of genetically encoded ultrasound contrast agents in the form of gas vesicles has recently emerged, which could extend the specificity of medical ultrasound imaging. However, given the delicate genetic nature of how these genes are integrated and expressed, current methods of producing gas vesicle-expressing mammalian cell lines requires significant cell processing time to establish a clonal/polyclonal line that robustly expresses the gas vesicles sufficiently enough for ultrasound contrast. Here, we describe an inducible and drug-selectable acoustic reporter gene system that can enable gas vesicle expression in mammalian cell lines, which we demonstrate using HEK293T cells. Our drug-selectable construct design increases the stability and proportion of cells that successfully integrate all plasmids into their genome, thus reducing the amount of cell processing time required. Additionally, we demonstrate that our drug-selectable strategy forgoes the need for single-cell cloning and fluorescence-activated cell sorting, and that a drug-selected mixed population is sufficient to generate robust ultrasound contrast. Successful gas vesicle expression was optically and ultrasonically verified, with cells expressing gas vesicles exhibiting an 80% greater signal-to-noise ratio compared to negative controls and a 500% greater signal-to-noise ratio compared to wild-type HEK293T cells. This technology presents a new reporter gene paradigm by which ultrasound can be harnessed to visualize specific cell types for applications including cellular reporting and cell therapies.

Keywords: acoustic reporter genes; cell engineering; ultrasound imaging.

FIGURE 1

Design of mARG_ds HEK293T clones. (a) Schematic of the transposase integrating mARG_ds construct designs. From left to right, the XLone‐mARG1 cassette possesses the Blasticidin resistant gene and the Tet‐On 3G system downstream of the constitutive EF‐1α promoter, and the mARG GvpB gene and mCherry downstream of the Dox inducible TRE3GS promoter. The XLone‐mARG2 cassette possesses the Puromycin resistant gene and the Tet‐On 3G system downstream of the constitutive EF‐1α promoter, and the mARG GvpF, GvpG, GvpL, GvpS, GvpK, GvpJ, GvpU genes (separated by P2A) and GFP downstream of the Dox inducible TRE3GS promoter. Finally, PB‐UbC‐mARG3 cassette possesses the mARG GvpF, GvpG, GvpL, and GvpK genes downstream of the constitutive hUbC promoter and Geneticin resistant gene downstream of the constitutive PGK promoter. (b) Flow cytometry of mARG_ds HEK293T clonal lines against GFP (XLone‐mARG2) and mCherry (XLone‐mARG1).

FIGURE 2

Single cell mARG_ds HEK293T cells express gas vesicles and produce ultrasound contrast. (a) Process flow for generating gas vesicle‐producing single‐cell clone HEK293T cell lines. (b) Clone 6 HEK293T mARG_ds gene expression validated by fluorescence microscopy. Scale bar = 10 μm. (c) Ultrasound contrast is produced by doxycycline‐treated HEK293T mARG_ds single cell clones after 3 days of treatment (left panel) whereas cells not treated with doxycycline produce no ultrasound contrast (right panel). (d) Signal to noise ratio of ultrasound imaging is significantly stronger in doxycycline‐treated cells compared to non‐doxycycline treated cells and non‐mARG_ds cells, indicative of gas vesicle production (n = 6). (e) Phase contrast image of HEK293T mARG_ds cells after 3 days of doxycycline treatment with small puncta spread throughout the cell (white arrows), indicative of gas vesicle presence. (f) Phase contrast image of HEK293T mARG_ds cells undergoing no doxycycline treatment.

FIGURE 3

Mixed population HEK293T mARG_ds cells produce sufficient ultrasound contrast after several days of antibiotic selection and doxycycline treatment. (a) Process flow for generating gas vesicle‐producing mixed population HEK293T cells. (b) Antibiotic selected and doxycycline‐treated HEK293T mARG_ds mixed population cells properly integrated the mARG_ds genes. Scale bar = 10 μm. (c) Doxycycline‐treated HEK293T mARG_ds mixed population cells produce stronger ultrasound contrast (left panel) compared to non‐doxycycline treated mARG_ds mixed population cells. (d) mARG_ds mixed population HEK293T cells exhibit a stronger signal to noise (SNR) ratio after 3 days of doxycycline treatment compared to non‐doxycycline treated mARG_ds cells and non‐mARG cells (n = 7). (e) Phase contrast image of HEK293T mARG_ds cells after 24 h of doxycycline treatment with small puncta spread throughout the cell (white arrows), indicative of gas vesicle presence. (f) Phase contrast image of HEK293T mARG_ds cells undergoing no doxycycline treatment. Small puncta are still present (white arrows), though in much smaller quantities than doxycycline‐treated cells.