Use of the viral 2A peptide for bicistronic expression in transgenic mice

Abstract

Background: Transgenic animals are widely used in biomedical research and biotechnology. Multicistronic constructs, in which several proteins are encoded by a single messenger RNA, are commonly used in genetically engineered animals. This is currently done by using an internal ribosomal entry site to separate the different coding regions. 2A peptides result in the co-translational 'cleavage' of proteins and are an attractive alternative to the internal ribosomal entry site. They are more reliable than the internal ribosomal entry site and lead to expression of multiple cistrons at equimolar levels. They work in a wide variety of eukaryotic cells, but to date have not been demonstrated to function in transgenic mice in an inheritable manner.

Results: To test 2A function in transgenic mice and uncover any possible toxicity of widespread expression of the 2A peptide, we made a bicistronic reporter construct containing the coding sequence for a membrane localised red fluorescent protein (Myr-TdTomato) and a nuclear localised green fluorescent protein (H2B-GFP), separated by a 2A sequence. When this reporter is transfected into HeLa cells, the two fluorescent proteins correctly localise to mutually exclusive cellular compartments, demonstrating that the bicistronic construct is a reliable readout of 2A function. The two fluorescent proteins also correctly localise when the reporter is electroporated into chick neural tube cells. We made two independent transgenic mouse lines that express the bicistronic reporter ubiquitously. For both lines, transgenic mice are born in Mendelian frequencies and are found to be healthy and fertile. Myr-TdTomato and H2B-GFP segregate to mutually exclusive cellular compartments in all tissues examined from a broad range of developmental stages, ranging from embryo to adult. One transgenic line shows X-linked inheritance of the transgene and mosaic expression in females but uniform expression in males, indicating that the transgene has integrated into the X chromosome in this line.

Conclusion: The 2A peptide efficiently mediates co-translational cleavage in transgenic mice in which it has been inherited through the germ-line. Mice expressing it ubiquitously throughout development and into adulthood appear normal. It is therefore a viable tool for use in genetically engineered mice and represents a superior alternative to the widely used internal ribosomal entry site.

Figure 1

Testing 2A constructs in HeLa cells. (A) Constructs Tom-2A-GFP and Tom-m2A-GFP used to test the function of the 2A linker. A membrane localised red fluorescent protein (Myr-TdTomato) linked to a nuclear localised green fluorescent protein (H2B-GFP) by either a functional 2A linker (top), or by a non-functional m2A linker (bottom). The specific nucleotide and amino acid changes made in Tom-m2A-GFP are shown in red. (B-B") HeLa cells lipofected with Tom-2A-GFP. Fluorescent cells show segregation of the TdTomato and enhanced green fluorescent protein (EGFP) to mutually exclusive cellular compartments. (C-C") HeLa cells lipofected with Tom-m2A-GFP as a negative control. The EGFP and TdTomato are no longer segregated properly, consistent with them remaining linked together as a single fusion protein due to the mutations in the 2A linker. (B, C) EGFP fluorescence. (B', C') TdTomato fluorescence. (B", C") Merge of EGFP (green) and TdTomato (magenta) fluorescence. Diagram in (A) not to scale. The scale bar in (B) represents 50 μm and applies to panels B through C".

Figure 2

Testing the Tom-2A-GFP construct in chick embryos. (A-C) Three-dimensional volume rendering of a confocal image stack of embryonic chick neural crest cells electroporated with the Tom-2A-GFP construct under the control of a CMV enhancer. The enhanced green fluorescent protein (EGFP) and TdTomato segregate to the nucleus and plasma membrane respectively, indicating that the 2A peptide functions in chick cells. (A) EGFP fluorescence. (B) TdTomato fluorescence. (C) Merge of EGFP (green) and TdTomato (magenta) fluorescence. Scale indicators are rendered in perspective along with the confocal image data and do not translate reliably to a two-dimensional image, therefore, no scale bar is depicted. As an approximate indicator, the two cells at the right of the image that have just divided are, together, between 15 and 20 μm long.

Figure 3

Widespread expression of Tom-2A-GFP in transgenic embryos. An E11.5 mouse embryo transgenic for Tom-2A-GFP under the control of a CAG promoter is shown. (A) Brightfield image, (B) Enhanced green fluorescent protein (EGFP) fluorescence and (C) TdTomato fluorescence. EGFP and TdTomato can be seen to be expressed throughout the embryo. Fluorescence intensity of both EGFP and TdTomato is higher in regions of relatively greater tissue density. The scale bar in (A) represents 1 mm and applies to all panels.

Figure 4

Function of 2A peptide in pregastrulation embryos. (A-A") E2.5, (B-B") E4.0 and (C-C") E5.5 embryos expressing Tom-2A-GFP ubiquitously under the control of a CAG promoter. The embryos shown represent the F2 generation with respect to the founder F0 transgenic mouse. In all the embryos shown, the transgene was inherited from a heterozygous father. Cells in both embryonic as well as extra-embryonic tissues show segregation of TdTomato and enhanced green fluorescent protein (EGFP) to mutually exclusive cellular compartments, indicating 2A function in both lineages. (A, B, C) EGFP fluorescence. (A', B', C') TdTomato fluorescence. (A", B", C") Merge of EGFP (green) and TdTomato (magenta) fluorescence. The scale bar in (A) represents 50 μm and applies to all panels.

Figure 5

Function of 2A peptide in E8.5 embryos. (A-A") Transverse section through an E8.5 transgenic embryo. (B-B") Transverse section through the neural tube of an E8.5 transgenic embryo. Note the 'ring' of dividing cells lining the inside of the neural tube. (C-D") High magnification view of neural tube cells expressing Tom-2A-GFP, stained with 4',6-diamidino-2-phenylindole (DAPI) (nucleus) and phalloidin (sub-cortical actin marking plasma membrane), confirming that enhanced green fluorescent protein (EGFP) and TdTomato are expressed in mutually exclusive compartments, in the nucleus and plasma membrane respectively. (A, B, C) EGFP fluorescence. (A', B', C') TdTomato fluorescence. (A", B", C") Merge of EGFP (green) and TdTomato (magenta) fluorescence. (D) DAPI fluorescence. (D') Atto647N-phalloidin fluorescence. (D") Merge of DAPI (green) and Atto647N (magenta). The scale bar in (A) represents 63 μm in (A-A"), 31 μm in (B-B") and 10 μm in (C-D").

Figure 6

Function of 2A peptide in various adult tissues. Sections through the heart (A, A'), lung (B, B'), liver (C, C'), kidney (D, D') and intestine (E, E') of adult CAG-TAG1 and CAG-TAG2 transgenic mice. The enhanced green fluorescent protein signal is depicted in green and TdTomato signal is depicted in magenta. No overlap of the H2B-GFP and Myr-TdTomato signal is seen, demonstrating efficient functioning of the 2A peptide. The scale bar in (A) represents 20 μm and is applicable to all panels.

Figure 7

Mosaic expression of CAG-TAG2 due to X inactivation. Sections through the heart of an adult female (A, A', and A") and male (B, B' and B") CAG-TAG2 mouse are shown. (A, B) Td-Tomato fluorescence showing expression of the transgene. (A', B') Atto647N-phalloidin fluorescence used as a 'counter-stain' to visualise all the tissue in the field of view. (A", B") Merge of TdTomato (red) and phalloidin (green) fluorescence, illustrating the mosaic expression of the CAG-TAG transgene in females due to X inactivation, but uniform expression in males. The scale bar in (A) represents 100 μm and is applicable to all panels.

Figure 8

Expression levels of CAG-TAG1 transgene across four generations. TdTomato fluorescence intensity was used as a measure of transgene expression levels. TdTomato fluorescence was quantified in cardiac tissue from the CAG-TAG1 male founder (F0), and from his F1, F2, and F3 male adult offspring. The plot represents the distribution of average TdTomato fluorescence intensity per pixel for 10 different fields of view for each sample. F1, F2 and F3 mice express the transgene at equivalent levels. They do not show a significant difference in expression levels (Student t-test, P > 0.1 for all three combinations of pair-wise comparisons of the three samples). The founder has a broader distribution of average intensity per pixel and shows a lower minimum value than the others because he was mosaic for the transgene, and some of the fields of view measured had regions that did not express the transgene, causing a lowering of the average pixel intensity.