An effect of homology length on gene disruption in Neurospora crassa

Y. Nakanishi, C. Ishii and H. Inoue

Dept. of Regulation Biology, Saitama University, Saitama 338-8570, Japan.

Fungal Genetics Newsletter 52:5-6

For targeted gene disruption in wild-type Neurospora crassa, 1000-bp of homologous sequences on either side of the cassette used for disruption is sufficient to give more than 10 % homologous recombination. We report here that varying the length of homology on each side seems to have different effects on the homologous recombination frequency.

Gene disruption based on homologous recombination is a critical strategy for genome analysis. We analyzed the effects of homology length and transformation methods on gene disruption by a simple method. We chose ad-3A as a target gene, because loss of gene function could be identified by pigment accumulation.

A 6-kb genomic fragment which covered the whole ad-3A gene and a 5-kb adjacent region was cloned on pBluescript^® SK⁺. We inserted the 1.4-kb hyg^r gene of pCB1003 (supplied by FGSC) at the EagI site of ad-3A. DNA fragments of varying lengths were amplified by PCR from this plasmid. Primers were designed to produce constructs that had 100 to 2000-bp ad-3A genomic sequence on each side of the hyg^r gene. The albino strain (FGSC 4934: a al-2 cot-1 pan-2) was used as the transformation recipient. Transformation of the DNA fragments was done in two ways, electroporation and classical spheroplast fusion. Electroporation was done based on published methods (Ninomiya et al. 2004). Five micrograms of DNA was mixed with 100 microliter of the suspension in 2-mm electroporation cuvets. Pulse conditions were as follows. Charging voltage: 1.5 kV, Maximum voltage range/timing mode: 2.5 kV/resistance, Capacitance timing: 50 uF, Resistance timing: R6 (186 ohm). After a one-shot pulse, the suspension was immediately mixed with 1 ml of ice-cooled minimal medium supplemented with adenine and incubated for three hours at 30 C before plating. Three ml of top agar medium kept at 50 C was mixed with the suspension and poured on minimal medium containing hygromycin, Ca-pantothenate and adenine sulfate. Spheroplast formation and transformation were done as described (Tomita et al. 1993) with a few modifications. The LYSING ENZYMES preparation from Trichoderma harzianum (SIGMA, L-1412) was used at 10 mg/ml for digestion and was incubated with conidia for 30 to 60 min. Five micrograms of DNA was used for transformation of 10⁷-10⁸ conidia. In both methods, colonies were collected after three-day incubation on the selection medium and transferred to the same medium in culture tubes. Conidial and hyphal color of transformants was examined after two weeks, and DNA was isolated if needed.

The results in Table 1 show that length of homology affects homologous insertion as expected. Ratios of colored transformants (disruptants) were low in these experiments because we selected by eye, and only reliably scorable colonies were counted. Most transformants were white, which suggests that most events were ectopic insertions. However, coloration by ad-3A disruption is dependent on the ratios of nuclei containing disrupted and nondisrupted genes . If transformants are heterokaryotic for transformation, the appearance of coloration by ad-3A disruption would be less apparent. Thus, we surely underestimated insertion frequencies. This inaccuracy in scoring homologous disruptants, which would underestimate their number, did not affect the trends observed, however.

We did PCR for all transformants of the 3000-3000 group in experiment 1. Two sets of primers were constructed for checking transformation events. One contained a primer set to detect an original ad-3A region.  The two primers of this set corresponded about 3kbp upstream and 4kbp downstream, respectively, of the ad-3A translation start site. Neither of these sequences was contained in the DNA fragments used in transformation. The other set contained a primer deduced from a central region of hyg^r gene and one of the primers of the first set. We did PCR until one of the sets or both sets amplified appropriate length fragments. In addition to the 26 transformants initially identified by eye, half of the rest (42/84) were proved to have target disruption.

These results indicate that (1) At least 1000 bp of homology to both sides of a target gene is needed to get a reasonable number of disruptants. Longer homology seems better, but even 1000bp homology on both sides results in more than half of the transformants containing a disrupted gene. (2) Homology length on one side affects insertion frequency differently from that on the other side. Homology in the 5’ direction in the target gene seems more important than that in the 3’ direction. Another possibility we cannot exclude is that the direction of the marker gene with respect to the homologous flanking sequences affects homologous recombination. The hyg^r gene was a reverse direction to the ad-3A gene in this study.

In comparison of transformation methods, the both methods gave similar efficiencies of targeted disruption. However, more transformants were obtained in electroporation with a less tedious procedure. We recommend electroporation to get disruptants.

Table 1 Effect of homology length on gene disruption

References

Ninomiya, Y., K. Suzuki, C. Ishii and H. Inoue, 2004. Highly efficient gene replacements in Neurospora strains deficient for nonhomologous end-joining. Proc. Natl. Acad. Sci. USA 101: 12248-12253.

Tomita, H., T. Soshi and H. Inoue, 1993. The Neurospora uvs-2 gene encodes a protein which has homology to yeast Rad18 with unique zinc finger motifs. Mol. Gen. Genet. 238: 225-233.

Return to the FGN 52 table of contents

Return to the FGSC home page