Transgene-free genome editing in marine algae by bacterial conjugation - comparison with biolistic CRISPR/Cas9 transformation

Abstract

The CRISPR/Cas9 technology has opened the possibility for targeted genome editing in various organisms including diatom model organisms. One standard method for delivery of vectors to diatom cells is by biolistic particle bombardment. Recently delivery by conjugation was added to the tool-box. An important difference between these methods is that biolistic transformation results in transgene integration of vector DNA into the algae genome, whereas conjugative transformation allows the vector to be maintained as an episome in the recipient cells. In this study, we have used both transformation methods to deliver the CRISPR/Cas9 system to the marine diatom Phaeodactylum tricornutum aiming to induce mutations in a common target gene. This allowed us to compare the two CRISPR/Cas9 delivery systems with regard to mutation efficiency, and to assess potential problems connected to constitutive expression of Cas9. We found that the percentage of CRISPR-induced targeted biallelic mutations are similar for both methods, but an extended growth period might be needed to induce biallelic mutations when the CRISPR/Cas9 system is episomal. Independent of the CRISPR/Cas9 vector system, constitutive expression of Cas9 can cause re-editing of mutant lines with small indels. Complications associated with the biolistic transformation system like the permanent and random integration of foreign DNA into the host genome and unstable mutant lines caused by constitutive expression of Cas9 can be avoided using the episomal CRISPR/Cas9 system. The episomal vector can be eliminated from the diatom cells by removal of selection pressure, resulting in transient Cas9 expression and non-transgenic mutant lines. Depending on legislation, such lines might be considered as non-GMOs.

Figure 1

Schematic illustration of the Myb target gene. Exons are indicated by solid filled blocks and target region Myb PAM1 (green block), and Myb PAM2 (orange block) lies within the coding region of the gene. The single Pt MYB1R domain (SHLQKYR) is indicated by a solid blue block. The intron is indicated by a thin black line. The target sequence for Myb PAM1 and PAM2 are presented below. Numbers indicate nucleotide position within the gene sequence.

Figure 2

Screening of transformants for Cas9-derived mutations in the Myb PAM1 and Myb PAM2 target regions. High resolution melting (HRM) analyses data presented as normalized melting peaks generated from PCR products from (A) primary colonies resulting from transformation with pKS diaCas9_sgRNA using biolistic bombardment (one month after transformation; n = 12) (B) primary colonies resulting from transformation with pPtPuc3m diaCas9_sgRNA using bacterial conjugation (one month after transformation; n = 12); (C) re-plated colonies resulting from transformation with pPtPuc3m diaCas9_sgRNA (three months after transformation; n = 12). WT profiles are represented by dark blue lines. Differently colored melting peaks indicate a melting behavior unlike that of the WT PCR product and implies the presence of indels in the Myb PAM1 and PAM2 target regions. Normalized melting peaks were generated by LightCycler® 96 Application Software Version 1.1.

Figure 3

Comparison of Cas9 mRNA expression levels versus genomic/episome Cas9 DNA content in biallelic mutant clones generated using biolistic bombardment or bacterial conjugation methods for CRISPR/Cas9 plasmid delivery. The grey bars represent relative Cas9 gene expression values and the black bars relative level of genomic/episome Cas9 DNA. Expression of Cas9 is the average of three biological replicates for each line and is normalized to the mutant line displaying the lowest Cas9 expression level (CON2.3). The relative amount of Cas9 DNA is normalized to CON2.3 the cell lines showing lowest Cas9 DNA content. Error bars represent the 95 percent confidence intervals. The abbreviations used are BIO: Cas9 editing lines resulting from biolistic CRISPR/Cas9 plasmid delivery; CON: Cas9 editing lines resulting from delivery of CRISPR/Cas9 plasmid by bacterial conjugation.

Figure 4

Re-editing of target regions caused by constitutively expressed Cas9 and sgRNA. Names of alleles in black font represent the first editing event, whereas names in brown font represent second editing. The target regions are indicated by bold green font with its respective PAM sequence in bold brown font. All indels are shown in bold red font. Polymorphisms present in the Myb gene (full overview in Supplementary Table 3) are indicated by blue bold font and were used to differentiate alleles for MYB1R_SHAQKYF5 gene.

Figure 5

Identification of non-transgenic mutants. 25 colonies from Myb PAM 1 (left side) and Myb PAM2 (right side) patched on 50% seawater (SW) f/2 agar plate without zeocin (A) and with zeocin (B). All patched colonies survived on non-selection plates whereas only colonies still containing pPtPuc3m-Cas9_sgRNA survived on selection plates. Colonies lacking the plasmid appeared white and dead after ten days on selection plates. Presented pictures were taken seven days after patching. Loss of pPtPuc3m-Cas9_sgRNA episome in colonies not able to survive on selection plates was verified by the inability to amplify a Cas9 DNA fragment (817 bp) from these colonies (C). Cas9 PCR products generated from colonies still able to grow on selection plates were included as positive controls. Ten months after initial transformation episome free colonies and colonies containing episomes were reanalyzed for presence of Cas9 DNA (817 bp fragment) (D). The origin of replication from the episome (855 bp fragment) and a genomic control fragment from Phatr2_52110 (188 bp) were co-amplified as a control of the samples (E). Lanes marked L contains GeneRuler 1 kb Plus DNA Ladder (Fermentas), WT is wild type P. tricornutum, PL is pPtPuc3m-Cas9_sgRNA plasmid and NTC (non-template control) is water. Numbers above lanes correspond to colony numbers indicated on the agar plates.