2.1. Strain and Media:
The strain utilized in this study, Schizosaccharomyces pombe strain or fission yeast, designated as SPY5101, was acquired from Moazed Danesh's group, as documented in Ragunathan's article [44]. In their work, 10XtetO sites were inserted upstream of a 114bp fragment of the ura4+ promoter at the endogenous locus of the ura4+ gene. Moreover, a GFP fusion protein was inserted downstream of the ura4+gene. Under such conditions, TetR-Clr4-I can bind with 10XtetO, inducing ectopic heterochromatin formation (H3K9me2) and subsequent loss of ura4+ gene function upon silencing, as shown in Fig. 2A. However, adding (tet) to the cell culture is crucial in controlling ura4+ gene expression, facilitating the removal of ectopic heterochromatin and subsequent activation of the ura4 + gene, as evidenced in Fig. 4B and the red bar in Figure S3 of Ragunathan's study [44]. Consequently, the (10XtetO- ura4 -GFP) marker shows sensitivity to (5-FOA) drug in the presence of (tet).
By utilizing this system in our experiments, the absence of tetracycline or (- tet condition) signifies ura4+ gene silencing due to H3K9-methylated heterochromatin presence (H3K9me2). Conversely, the presence of (tet) in the cell culture or (+ tet condition) indicates ura4+ gene activation in the euchromatin region, resulting in cell death because the active ura4+gene is sensitive to (5-FOA) drug.
Employing both conditions within a single experiment, mutant colonies may arise on solid media with (5-FOA drug + tet) only if there is a mutation within the ura4+ gene. This methodology enables the estimation of phenotypic mutation rates in the absence and presence of (tet) at the same gene locus (ura4+ gene). Thus, we could investigate whether H3K9-methylated heterochromatin (H3K9me2) elevates or diminishes mutation rates of a specific gene and thoroughly analyze its isolated effect to ensure precise results.
Regarding the media for cell growth, we cultivated this strain using a fully synthetic medium EMMG [45] during fluctuation assays. Drug concentrations were adjusted to (5-FOA) at (1 g/L) and (tet) at (10 µM/L).
2.2. Silencing Assay:
Before estimating the mutation rate, it is essential to confirm the strain's sensitivity to media containing both (5-FOA + tet) drugs. Therefore, the strain with the modified reporter (10XtetO- ura4+-GFP) was subjected to a four- or five-fold serial dilution and then spotted on plates containing yeast extract (YE), (5-FOA) (1 g/L), and (tet) (10 µM).
2.3. Developing a Fluctuation Assay for Increased Precision:
Accurately estimating mutation rates is crucial for understanding the mechanisms of evolution and disease. Choosing the correct method is essential to accurately estimate the parameter m, representing the number of mutation events per culture. Although several methods are available to estimate mutation rates, fluctuation assays are the most accurate and straightforward for detecting parameter m. We employed this method to ensure the accuracy of our findings and avoid the pitfalls of other methods, such as mutation accumulation and mutant accumulation assays.
To achieve precise results, a rigorous approach to experimental design and data analysis is essential. Our proposed improvements, based on the latest
[46] and original fluctuation assay protocols [47], enabled us to obtain reliable and compelling data.
Fluctuation Assays:
Fluctuation assays were conducted on two isogenic clones of the previously mentioned strain after single-clone isolation, as illustrated in Fig. 3. For each assay, an isogenic clone was grown in liquid EMMG media at 32°C until saturation. The preculture was then diluted into fresh media to achieve an initial inoculum (No) of approximately 200 cells and split into two different media types. The first media condition contained only liquid EMMG media, representing the absence of tetracycline (-tet condition). The second media condition contained EMMG with (tet) (10 µM/L), representing the presence of tetracycline (+ tet condition).
Using 96-well plates, we dispensed 10 µL of diluted cultures for each experimental condition, thereby enhancing the overall precision of the assay. This 96-well format allowed us to maintain consistent culture volumes, use foil covers to prevent evaporation, and keep the samples in the dark to avoid the decomposition of tetracycline. The cultures in the 96-well plates were grown until saturation. Subsequently, 24 parallel cultures were taken for each experimental condition to determine the average number of cells per culture before plating (Nt). The remaining 72 parallel cultures were plated separately for each condition, with and without (tet), on 9 cm Petri dishes containing EMMG, (5-FOA) (1 g/L), and (tet) (10 µM/L).
After incubation at 32°C for 3–4 days in the dark, the number of mutants r for each parallel culture was recorded. Some of these colonies were picked and replated on 9 cm Petri dishes containing EMMG, (5-FOA) (1 g/L), and (tet) (10 µM/L) to confirm their phenotypic resistance to these drugs.
2.4. Statistical Analysis of Fluctuation Assay Results:
To calculate the parameter m representing the number of mutational events per culture in R software, we individually calculated the number of mutants or colonies on the agar plates for each parallel culture. Following Eq. 18 from [48], the phenotypic mutation rates µ were determined using (µ = parameter m/Nt), where Nt represents the average number of cells per culture before plating on the agar plates. Subsequently, we calculated the 95% confidence intervals (CI) for the combined phenotypic mutation rates of clone 1 and clone 2 using the Ma-Sandri-Sarkar Maximum Likelihood Estimator (MSS-MLE) [49], a method that has proven to be the most reliable over the years.
Our attempt to apply the MSS-MLE approach in R software using the recursive likelihood function encountered a significant challenge due to the large number of mutants. To overcome this limitation, we developed a non-recursive version of the likelihood function and introduced non-parametric bootstrapping [50]. This approach yielded substantial findings, addressing a gap in the literature where no alternative non-recursive likelihood function appears to be available. For detailed insights into the calculation of 95% confidence intervals through bootstrapping, please refer to [50].
2.5. Computational Analysis:
To ensure the accuracy of our fluctuation assay data, we compared the predicted cumulative frequency distribution of mutants with the experimental data and created a corresponding plot using MATLAB. Leveraging a MATLAB file from a similar study [51], we performed computational simulations for our study. According to Lang's study, the one-parameter model accounts for the Luria-Delbrück distribution [47]. Additionally, the two-parameter model combines the Luria-Delbrück distribution (parameter m) with the Poisson distribution (parameter n = md) to form the joint distribution for the number of mutants per culture. This approach allowed us to assess mutation events occurring after the plating process (d), which represents the total number of cell divisions capable of producing mutants post-plating on selective media containing both (5-FOA) and tetracycline.
By fitting our fluctuation assay data into these one-parameter and two-parameter models, we were able to simulate the outcomes of the assay, thereby visually validating the accuracy of our experimental results