Simultaneous knockout of multiple LHCF genes using single sgRNAs and engineering of a high-fidelity Cas9 for precise genome editing in marine algae

Abstract

The CRISPR/Cas9 system is an RNA-guided sequence-specific genome editing tool, which has been adopted for single or multiple gene editing in a wide range of organisms. When working with gene families with functional redundancy, knocking out multiple genes within the same family may be required to generate a phenotype. In this study, we tested the possibility of exploiting the known tolerance of Cas9 for mismatches between the single-guide RNA (sgRNA) and target site to simultaneously introduce indels in multiple homologous genes in the marine diatom Phaeodactylum tricornutum. As a proof of concept, we designed two sgRNAs that could potentially target the same six light-harvesting complex (LHC) genes belonging to the LHCF subgroup. Mutations in up to five genes were achieved simultaneously using a previously established CRISPR/Cas9 system for P. tricornutum. A visible colour change was observed in knockout mutants with multiple LHCF lesions. A combination of pigment, LHCF protein and growth analyses was used to further investigate the phenotypic differences between the multiple LHCF mutants and WT. Furthermore, we used the two same sgRNAs in combination with a variant of the existing Cas9 where four amino acids substitutions had been introduced that previously have been shown to increase Cas9 specificity. A significant reduction of off-target editing events was observed, indicating that the altered Cas9 functioned as a high-fidelity (HiFi) Cas9 nuclease.

Keywords: Phaeodactylum tricornutum; High-Fidelity (HiFi) Cas9 nuclease; Light-harvesting complex proteins; diatom; lhcf mutants; off-target gene editing; precision genome editing.

Figure 1

Overview of LHCF genes and target sequences of LHCF mPAM1 and mPAM2 sgRNAs. The localization of targeted LHCF genes in chromosome 2 and 24 is schematically presented. Three predicted on target (LHCF1, LHCF3, LHCF4) regions and one target region with mismatch (LHCF2) are located in chromosome 2, whereas two possible target regions with mismatches (LHCF5 and LHCF11) are located in chromosome 24. The target sites of LHCF mPAM1 and LHCF mPAM2 are highlighted with dark green and light green rectangles, respectively. Brown font colour of genes names before the target region indicates target sites without mismatches, whereas blue font colour indicates target sites with mismatches. Mismatched nucleotides are labelled red. PAM sites (NGG) are underlined. Cas9 cut site at the target region is indicated by a black arrow. More information about target regions is given in Table S1.

Figure 2

Disruption of multiple LHCF genes. (a) Phenotypes of high light (280 µmol photons/m²/s, HL) and low light (50 µmol photons/m²/s, LL) acclimated P. tricornutum WT, lhcf mPAM1 15.1 and lhcf mPAM1 6.1.11 cultures. All cell cultures were adjusted to equal cell density (27 million cells/mL) and imaged with an iPhone 6s. (b) Schematic overview of the genetic composition of mutated LHCF genes in lhcf mPAM1 15.1 and lhcf mPAM1 6.1.11 mutants. WT genes are white. Knocked out (KO) genes are light grey and annotated by indel (i/d) and the base length of the indel (#bp) if known. In‐frame gene fusion products are labelled dark grey. Confirmed DNA sequences are labelled with a solid border. Target regions that could not be amplified by PCR have dashed borders. (c) Western blot analysis of LHCFs proteins from WT and lhcf mPAM1 mutant lines acclimated to HL or LL conditions. P. tricornutum WT, lhcf mPAM1 mutants 15.1 and 6.1.11 were grown for two weeks in HL and LL before harvesting and isolation of total protein. Proteins separated on an SDS‐PAGE gel were blotted on a nitrocellulose membrane, and polyclonal rabbit antibodies were used to target LHCFs, and D1‐protein (psbA/D1) of photosystem II. MagicMarkTM XP Western Protein standards report the protein mass (in kDa).

Figure 3

Effects of knocking out multiple LHCF genes at pigment level. WT and multiple lhcf mutants acclimated to HL (280 µmol photons/m²/s) and LL (50 µmol photons/m²/s). Fuco per Chl a (Fuco:Chl a) ratios are presented as the mean of n replicates of the relative unit mol:mol, with the standard error of the mean (±SEM, nWT = 9, n6.1.11 = 4, n15.1 = 4). Statistical significance was inferred using two‐sided t‐test assuming heteroskedastic variance. Asterisk (**) denotes significant difference with P < 0.001.

Figure 4

Schematic representation of CRISPR/Cas9 vectors used in this study. Standard diaCas9 (top, pKS diaCas9_sgRNA) and HiFi‐diaCas9 (bottom, HiFi‐pKS diaCas9_sgRNA) vectors differ in four amino acid residue positions. The substitutions are indicated in the HiFi‐diaCas9 gene. The U6 promoter drives sgRNA expression while expression of Cas9 protein is driven by the LHCF2 promoter.