Abstract:
The plant hormone salicylic acid (SA) has been used for millennia to relieve pain. In modern times, its acetylate form – known as aspirin – has become one of the most popular painkillers. SA production in plants has been implicated in multiple aspects of plant life including thermogenesis, response to stresses, seed viability, leaf senescence, flowering time regulation and immune responses. It has been reported that regulation of SA accumulation exhibits natural variation, for example due to the positive regulator ACCELERATED CELL DEATH 6 (ACD6). But still little is known regarding variation of the main biosynthetic pathway of SA, its involvement in flowering time regulation or its effect regarding the virulence of microbes.
To elucidate natural variation in the synthesis of SA and its effect on other processes, I utilised a newly introduced genome editing approach called CRISPR/Cas system. This tool allows the user to edit a targeted region using an endonuclease (Cas9) and an artificial RNA (sgRNA). With this method I selectively knocked out the ICS1 gene and disrupted the main SA biosynthetic pathway in seven natural accessions of A. thaliana. To efficiently screen the large number of individuals that required genotyping for this workflow (>900) I developed a preparation and analysis pipeline using deep amplicon sequencing (CRISPR-finder). Using this pipeline, individuals carrying variants at the targeted region can be identified within a few days in a cost-effective and precise manner.
Using the ics1 lines I generated, in spite of the non-functional ICS1 allele, I was able to detect residual SA in most of the lines. These results suggested that the alternative PAL biosynthetic pathway or the ICS2 gene may be responsible for appreciable levels of SA production in some genotypes. Additionally, I concluded that flowering time was not significantly affected by decreased levels of SA in these seven natural accessions, even though SA had been linked to flowering before. Significant reduction of SA accumulation was observed for all the mutant lines, but to different degrees, when compared to the corresponding wild types, during Pseudomonas syringae (Psm4326) infection. When the oomycete Hyaloperonospora arabidopsidis (14OHMLP04) was used for infecting individuals, no significant induction of SA accumulation was detected when plants exhibited resistance (flecking necrosis or trailining necrosis) or susceptibility. For the genotypes characterised with complete resistance (flecking necrosis) no effect of SA was observed. Interestingly, increased severity of trailing necrosis was observed for the ics1 mutant lines derived from partially resistant genotypes (trailing necrosis symptom). There were also accessions and their corresponding ics1 mutant lines that showed susceptibility and pathogen growth. These findings suggest that resistance to the pathogen isolate is SA-independent and that decreased levels of SA allow an increased manifestation of the potential virulence of 14OHMLP04.
Finally, the commonly used Col-0 accession was used for investigating a controversial aspect of the CRISPR/Cas system – off-target cleavage. This can occur due to non specific Cas9 activity. For assessing off-target variants, I used the same sgRNAs as the ones used for generating the ics1 mutant lines. During the investigation, I was able to detect a very small number of incidents (0, 2 and 5 depending on the line in question) that could potentially be attributed to Cas9’s cleavage activity. These numbers do not differ dramatically from the expected de novo mutation rate in A. thaliana which is approximately one mutation per generation. This makes it hard to confidently assign detected variants to Cas9 cleavage or de novo generation. These findings do not exclude that off-target cleavage events can occur when the CRISPR/Cas system is used, but they are rare and may be difficult to detect.
Thale cress
Print-on-Demand