Mutational analysis of N-ethyl-N-nitrosourea (ENU) in the fission yeast Schizosaccharomyces pombe

Forward genetics has boosted our knowledge on genic function in a multitude of biological models and it has significantly contributed to the understanding of genetic bases of development, ageing and human diseases. With the advent of the next generation sequencing and use of powerful bioinformatic tools, this traditional genetic strategy has acquired a new impulse. At present, whole genome sequencing assisted by in silico analysis allows the rapid and efficient identification of gene variants that are responsible for a particular phenotype. In this experimental pipeline, it is crucial to start by provoking a large number of random changes in the genome of the model organisms to be screened. A range of chemical mutagens are used to this end because most of them display particular reactivity properties and act differently over DNA. Here we use N-ethyl-N-nitrosourea (ENU) as a mutagen for the first time to our knowledge in the fission yeast Schizosaccharomyces pombe. By comparison to the extensively used Ethyl methanesulfonate (EMS) in a phenotype-based study, we conclude that ENU is a very potent and easy-to-use mutagen. Judging from DNA sequence analysis of the identified mutants, ENU induces base changes rather than indels and the mutational spectrum in the fission yeast seems similar to that found in mice but different to that described in other single-celled organisms such as budding yeast and E. coli. Using ENU in S. pombe, we have gathered a collection of 49 auxotrophic mutants including two deleterious alleles of ATIC human ortholog. Defective alleles of this gene are causative of AICA-Ribosiduria, a severe genetic disease. We have also identified 5 aminoglycoside-resistance inactivating mutations in APH genes. All these mutations reported here may be of interest in the metabolism and antibiotic resistance research fields.

7 133 ENU for 20 minutes at 30º. To serve as untreated controls, Aliquots 1 and 3 were 134 processed and washed as aliquots 2 and 4 respectively; except for the lack of the 135 alkylating agent. Afterwards, aliquots 1 and 2 were added into 8 ml of 5% Na 2 S 2 O 3 to 136 inactivate EMS reactivity and washed immediately with fresh media. We had also found 137 in previous trials that KOH solution, which is used in other systems to inactivate ENU´s 138 reactivity, resulted rather toxic for yeast cells. Thus, aliquots 3 and 4 were washed three 139 times with only minimal medium lacking nitrogen source to avoid cell division before 140 plating. After the last wash, cells were resuspended in fresh media and the number of 141 cells per millilitre was scored in all tubes, considering the average of three independent 142 counts in Neubauer´s chamber as the reference cell number for each tube (Fig 1,  174 Loss-of-function mutation frequencies 175 In order to compare the mutagenic potential of ENU and EMS in fission yeast, we 176 screened for auxotrophy-causing mutations as a gene loss-of-function readout over the 177 same number of genomic targets. After EMS and ENU mutagenesis of prototrophic 178 cells explained above (see materials and methods and Fig 1C), we plated them onto 179 YES media. Resulting colonies were counted up and replica-plated onto synthetic 180 minimal media. We then searched for any auxotrophic mutant growing in rich media 181 but unable to proliferate in media with no supplements (EMM). We found 21 such 182 mutants out of 3840 colonies (0,54%) in the case of EMS, and 28 mutants out of 4353 183 colonies for ENU treatment (0,64%) ( Table 2). We further checked if some of these 184 mutants really interrupted specific metabolic pathways and whether any of these could 185 have mutations in more than one metabolic pathway. Mutants were plated onto minimal 186 media lacking just the final product for one of the most common auxotrophic markers 187 used in this yeast: leucine, adenine, uracil and histidine respectively. We found 188 particular auxotrophs for all these metabolites, except for uracil in the case of ENU 189 (Table 2).
190 198 Thus, these could represent ideal candidates for going down to the specific mutation in 199 the DNA to discover new deleterious alleles for both or any of such genes, contributing 200 also to assess the ENU´s mutational spectrum in fission yeast. Therefore, we first 201 crossed both mutants to a wild type strain to check out whether the two metabolic 202 deficiencies segregate together. In both cases, in 16 pulled tetrads, 100% of adenine 203 requiring spores need histidine as well. We then crossed both mutants to each other and 204 did not find any prototroph within the offspring (10 tetrads). These data indicate that 205 there is only one single locus affected and this one is the same in both mutants.

206
To distinguish among the two putative mutated loci (ade9 and ade10), we 11 208 (www.pombase.org). Neither mutation showed linkage to csi1 deletion marker, leaving 209 ade10 as the only candidate. We therefore sequenced ade10 coding locus in both 210 mutants. We found a different single base pair substitution in each strain (Table 3), 211 confirming that both mutants are allelic to ade10 (named ade10.68 and ade10.424 212 respectively) and that the double auxotrophic deficiency is due to a single defective 213 gene rather than two independent mutations in different pathways.
214 Table 3. Mutational spectrum of ENU in fission yeast.

Marker Base pair change
Coding nucleotide change (pFA6a numbering)

Amino acid change
215 Base pair substitutions found after ENU treatment are summarized along with the 216 change that abolish function of corresponding protein product.
12 218 Mutational spectrum of ENU in fission yeast 219 The reference strain used in this study was originally chosen to bear both G418 and 220 Hygromycin B resistance markers integrated into chromosome I and II respectively. To 221 survey for the type of mutations induced by ENU in the fission yeast genome, we 222 selected loss-of-function mutations in those dominant markers after the treatment. To 223 this end, we obtained 45000 colonies in regular YES rich media after ENU treatment 224 described before (tube 4) ( Fig 1D). These were replica-plated back to regular YES and 225 YES containing either G418 or Hygromycin B antibiotics (50 mg/ml). Sensitive 226 colonies (three G418 s and two Hph s ) were picked for marker sequencing.
227 Sequence comparison to wild type control markers identified mutations leading 228 to antibiotic resistance loss (listed in Table 3). Both enzymes conferring G418 and    356 Therefore, these experimental conditions would give a rough estimation between 14520 357 and 18000 colonies to be screened to find a knock-out hit in a given average gene.
358 Slight variations are to be considered by factors such as G-C content or gene´s size.

359
Taking all this together, we propose that this molecule can be very efficiently