A transient CRISPR/Cas9 expression system for genome editing in Trypanosoma brucei

Objective Generation of knockouts and in situ tagging of genes in Trypanosoma brucei has been greatly facilitated by using CRISPR/Cas9 as a genome editing tool. To date, this has entailed using a limited number of cell lines that are stably transformed to express Cas9 and T7 RNA polymerase (T7RNAP). It would be desirable, however, to be able to use CRISPR/Cas9 for any trypanosome cell line. Results We describe a sequential transfection expression system that enables transient expression of the two proteins, followed by delivery of PCR products for gRNAs and repair templates. This procedure can be used for genome editing without the need for stable integration of the Cas9 and T7RNAP genes.


Introduction
The establishment of genome editing by CRISPR/Cas9 in trypanosomatids has greatly increased the ease with which knockouts can be generated, as two copies of a non-essential gene can often be deleted in a single round of transfection. The system most widely used for Trypanosoma brucei entails creating cell lines which express (either constitutively or inducibly) both Cas9 to generate a DNA double-strand break and T7RNAP for the transcription of guide RNAs from a DNA template [1][2][3][4]. The original pTB011 plasmid generated for T. brucei [4] encodes Cas9 flanked by tubulin sequences, enabling the construct to be integrated into the corresponding multicopy array. We modified this plasmid by replacing the puromycin acetyltransferase gene with that of T7RNAP, giving rise to pTB011_Cas9_T7RNAP_blast (Fig. 1a). This plasmid allows generation of CRISPR/Cas9-competent cell lines, constitutively expressing T7RNAP and Cas9 from the tubulin locus, in a single round of transfection [5]. Further modification of this plasmid, and optimisation of the transfection protocol for transient expression of the proteins and guide RNAs, are described below.

Results
In order to expand the range of cell lines that can be genetically modified by CRISPR/Cas9, without the need for stable transformation, we developed a protocol for transient expression of all three components (Cas9, T7RNAP and guide RNAs) in T. brucei. The plasmid Allin-one-Cas9 (pAi1C9; Fig. 1b) was constructed by replacing the α-tubulin sequence upstream of T7RNAP with 366 bp of the EP procyclin promoter and 5′ UTR [6].
To test the functionality of the plasmid, we used the LeishGEdit software [4] to design a guide RNA that allows tagging of the C-terminus of phosphodiesterase B1 (PDEB1) with mNeonGreen (mNG). The repair template, in the form of a purified PCR product, including homology arms of 30 bp on each side, was amplified from pPOTv7 [7]. As a positive control we transfected a derivative of T. brucei Lister 427 procyclic forms that already

BMC Research Notes
*Correspondence: isabel.roditi@izb.unibe.ch 1 Institute of Cell Biology, University of Bern, Bern, Switzerland Full list of author information is available at the end of the article expresses Cas9 and T7RNAP constitutively (Lister 427/ Cas9) [5]. This gave rise to cells with a fluorescent flagellum as expected (tryptag.org) (Fig. 2a).
For genome editing by transient expression of Cas9 and T7RNAP, we tested two different transfection schemes. Initially, all components were transfected simultaneously: the circular plasmid pAi1C9 encoding Cas9 and T7RNAP, the DNA template for in vivo transcription of a guide RNA, and the DNA repair template harbouring mNG for C-terminal tagging and a hygromycin resistance gene. This procedure gave few hygromycinresistant clones, only 10 clones in total, from 4 separate transfections. These clones either did not express mNG (4 clones), or expressed it as a cytoplasmic protein (2 clones) or expressed it correctly localised to the flagellum (4 clones). Examples are shown in Fig. 2b. We hypothesise that, due to insufficient expression of Cas9 and T7RNAP, the amounts of site-specific guide RNAs and double-strand breaks at the correct locus were too low to drive efficient integration by homology-directed repair.
We therefore tested sequential transfections in which we first used pAi1C9 to enable expression of Cas9 and T7RNAP. A second transfection 20 h later provided the templates for the gRNA and the mNG repair construct. The same electroporation conditions were used for the plasmid and the templates (see Methods and Fig. 3). This procedure yielded 20 clones. There were 16 clones in which mNG was correctly localised to the flagellum, one in which it was cytoplasmic, 2 clones with mixed populations of mNG localised either to the flagellum or the cytoplasm, and one that was negative (see Fig. 2b for representative examples). One clone with flagellar mNeonGreen had a fluorescence intensity twice that of the others (clone B1) and might have been tagged on both alleles.
We have also used this procedure of sequential transfections to simultaneously knock out both copies of trypanin or GPI-anchor transamidase subunit 8 (GPI8), which are non-essential genes in cultured procyclic form T. brucei [8,9]. Genotyping data is provided in Additional file 1.

Transfection of pooled PCR products for repair constructs and sgRNA templates
The entire culture from the first transfection was centrifuged at 1700 g for 5 min, the supernatant discarded and the cells resuspended in 100 µl TbBSF transfection buffer containing the pooled PCR products (see protocol for PCRs below). For tagging one sgRNA template (targeted to the 3′ end of the ORF) and one repair template (hygromycin resistance) were provided; to generate the knockouts, we provided two sgRNA templates (targeted to the 5′ and 3′ ends of the ORF) and two repair templates (hygromycin and neomycin resistance genes). Transfection was performed as described above and the cells were transferred to 10 ml SDM79 supplemented with 10% FBS. The cells were diluted 1:5, 1:50 To remove particles of subcellular size, a cut-off of 3 x 10 4 was applied to the forward scatter. A total of 10 4 events were recorded and analysed using FlowJo software without gating. TF transfection and 1:500 in conditioned medium (fresh medium + 20% supernatant of a log-phase culture) and distributed into 24-well plates (1 ml in each well). Transformants were selected using 25 µg ml −1 Hygromycin B and/ or 15 µg ml −1 Geneticin; stable clones were obtained 2 weeks post selection.

Limitations
Transient transfection with pAi1C9 gives rise to fewer clones than cell lines that are stably transformed with the T7RNAP and Cas9 genes. It has the advantage, however, that it can be applied to any T. brucei cell line. Moreover, in contrast to stable integration of the CRISPR/Cas9 machinery, transient transfection does not require additional selectable markers and it has the added advantage that it circumvents possible Cas9 toxicity. In addition to using Cas9 for deletion, mutation, tagging or integration of ectopic copies, this procedure could also be adapted for nuclease inactive Cas9 variants for targeting RNAs or epigenetic modifications [14,15].