Natural Selection Against the White Allele in Drosophila
April 19, 2020
Genetics with Laboratory
This experiment uses Drosophila as a model for population genetics to provide insight as to how natural selection works on the W+ gene which is located on the X chromosome to change the frequency of the red eye (W+) allele and white eye (w) allele over three generations. The principles of Hardy-Weinberg Equilibrium (HWE) were used to determine if the frequencies of the two alleles across generations, and a Chi-square test was used to check if the population was in HWE or if evolutionary forces were at work. The Chi-square test for HWE led to the rejection of the null hypothesis which stated that the populations were in HWE. The average value of the selection coefficient against the white eye (w) allele for the three generations was estimated to be approximately s=0.224, this resulted in an overall decrease in the frequency of the white eye (w) allele from 0.5 in the parental generation to 0.26 for male Drosophila and 0.38 for female Drosophila by the third generation. The red eye (W+) allele had higher fitness, so consequently, its frequency increased from 0.5 in the parental generation to 0.74 for male Drosophila and 0.62 for female Drosophila by the third generation.
Population genetics is an emerging field of genetics that tries to explain and measure genetic variation and study its effect on molecular evolution within and between species (Casillas and Barbadilla, 2017). Thus, population genetics uses biological forces such as natural selection, migration, mutation, and genetic drift to explain the changes in allele frequencies that occur in a population (Casillas and Barbadilla, 2017).
Drosophila has commonly been used as a model organism in the field of population genetics because they have several characteristics that make it possible for genetic variation and molecular evolution to be studied in relatively short periods of time (Casillas and Barbadilla, 2017). The study of population genetics requires that genetic changes be observed across many generations, the issue with this is that the time it takes for most species to mature and reproduce is too long to be studied in a laboratory setting. However, Drosophila offers a solution to this problem because they develop rapidly and have a short life cycle, they complete the transition from an egg to an adult in as little as 9 days (Flatt, 2020). Additionally, they are capable of laying 1000 to 3000 eggs in their lifespan, with 500 to 1500 of those offspring surviving to adulthood (Flatt, 2020), this high reproductive rate makes it possible to have large sample sizes. The small size of Drosophila and the ease of their care makes it possible to breed and raise them at a low cost in laboratory settings (Flatt, 2020). Furthermore, the Drosophila was one of the first species to have its genome completely sequenced, which means there is a wealth of information available in regard to allelic variations, mutations, inheritance, and gene maps (Flatt, 2020).
This experiment will be using Drosophila and the red/white eye locus to demonstrate that natural selection and genetic drift are naturally occurring biological forces that act on a population to cause a change in allele frequencies over time. The experiment will begin with an equal proportion of the white eye (w) allele and red eye (W+) allele and will track how the proportion of each allele changes over three generations when the flies are allowed to randomly mate inside of a breeding chamber. The laws of Hardy-Weinberg Equilibrium will be applied to the Drosophila in this experiment to estimate the allele frequencies of the white (w) allele and red (W+) allele for the female flies in each generation, as well as for estimating the genotypic frequencies for the female flies. Additionally, HWE will be used to estimate the expected phenotypes of male flies in each generation, so that a Chi-square test can be used to assess if the population of flies is in HWE. If the Chi-square test reveals significant deviation from HWE in the observed population then this would indicate that an evolutionary force such as natural selection or genetic drift is acting within the population, resulting in changes to the allelic frequencies of the red eye (W+) and white eye (w) alleles over time. Furthermore, the Selection Coefficient (s) against the white eye (w) allele in males will be calculated to quantify to what degree natural selection is responsible for any observed changes in allele frequency over time.
The gene that is being examined in this experiment is the W + gene which is found on the X chromosome, the W + protein product of this gene is responsible for loading and depositing pigment into the eyes of Drosophila (Xiao et al., 2017). The wildtype allele of this gene is the red eye (W+) allele, this is the dominant allele so flies with just one copy of the allele have a red eye phenotype (Xiao et al., 2017). There is also a recessive null mutant allele called the white eye (w) allele, this mutation disrupts the pigment pathway and results in flies with a white eye phenotype (Heil et al., 2012; Xiao et al., 2017). The white eye (w) allele is somewhat deleterious in that it lowers the fitness of white eye flies by up to 25% and causes poor vision (Heil et al., 2012; Xiao et al., 2017). The differential fitness between the red eye (W+) allele and white eye (w) allele results in selection against the white eye (w) allele, which eventually removes the mutant allele from the population (Xiao et al., 2017). The W + gene also has a critical role in the central nervous system function, the W + protein is also responsible for many molecular processes including the transport of neurotransmitters, messenger signals, and chemicals crucial to metabolic processes (Xiao et al., 2017). Originally, the eye color alone was believed to be responsible for the unsuccessful copulation of white eye Drosophila, but now it is suspected that the reduced fitness observed in white eye Drosophila is the result of this described central nervous system dysfunction (Xiao et al., 2017).
The process of natural selection, in which unique genotypes have varying levels of reproductive success, often leads to evolution if allelic frequencies change between generations (Carlini, 2020). Natural selection can only occur when two requirements are met: heritable genetic variation and differential fitness (Carlini, 2020; Flatt, 2020). In this context, fitness refers to the ability to contribute to the genetic composition of the next generation in terms of reproductive success (Carlini, 2020). Genetic variation is the result of multiple alleles existing within a population, these alleles are usually introduced by mutation (Carlini, 2020; Cook and Saccheri, 2012). These mutations can be beneficial or detrimental to a species, by either offering an advantage that increases the fitness of the individual or disadvantages by reducing the fitness of the individual (Cook and Saccheri, 2012). Together the significance of heritable genetic variation and varying fitness is that within a population there will be different alleles with corresponding levels of fitness, which means that over time the frequency of the favorable alleles will increase within a population. In essence, this change of allele frequencies over time in and across populations is the process of evolution by natural selection, and this process can be observed as the frequency of the corresponding phenotypes changes as well.
However, evolutionary change is not a process that can be seen in individuals of a species, but rather it is something that occurs in populations over many generations (Carlini, 2020). There are certain circumstances, in which environmental pressures and natural selection work together to increase the rate of allelic change between generations in a population. Some common examples of such rapid evolutionary change are observed in antibiotic-resistant bacteria, pesticide-resistant insects, and the peppered moth (Carlini, 2020). The peppered moth (Biston betularia) has been an excellent case study for the process of natural selection because of the genetic variation of its two phenotypic color forms: the wildtype phenotype is speckled black and white, while the mutant melanic phenotype is very dark almost completely black. Originally, the melanic form was uncommon and only present in low frequencies, but by the end of the 19th century, the frequency of the melanic form in industrialized London became so high that the original wildtype form almost disappeared completely (Cook and Saccheri, 2012). In the industrialized regions of England with smoke-blackened surfaces, this melanic form had higher fitness due to its ability to camouflage in the dark environment and avoid predation by birds, these selective pressures against the normal form resulted in the decrease of normal allele frequency over generations while the frequency of the melanic allele increased (Cook and Saccheri, 2012). It is estimated that the fitness of the melanic form was 30% to 50% higher than the wildtype form in industrialized areas where the dark version was common in high frequency (Cook and Saccheri, 2012). Whereas, the melanic form was only present in low frequencies in rural areas, and without the soot-darkened buildings to camouflage against the melanic form had a 10% to 12% lower fitness than the normal form (Cook and Saccheri, 2012). Eventually, when air pollution decreased and the urban environment was no longer smoke-covered, the melanic form lost its selective advantage and the frequencies began to decrease (Cook and Saccheri, 2012). The significance of these examples and one of the goals for this experiment working with Drosophila is to demonstrate that natural selection is not an invisible force, but rather an observable phenomenon.
Another force that acts on populations and results in changes to allelic frequencies over time is genetic drift. Genetic drift is the idea that within a finite population, the gametes that will constitute the next generation will be randomly sampled from the allele pool of the parental generation, which results in changes to the allele frequencies between and within populations over time due to chance alone (Casillas and Barbadilla, 2017). The effects of genetic drift are also observed as the increase in allelic frequency variation across populations, and the loss of allelic variation within a population as the heterozygous genotype declines over time leading to the fixation of one allele (Carlini, 2020). This phenomenon affects populations of all sizes but is best observed in small population sizes where it acts quickly (Carlini, 2020; Casillas and Barbadilla, 2017).
The degree to which genetic drift changes the allele frequencies within a population depends on the effective population size (Ne), when the effective population size is small then genetic drift might be the strongest force acting on allele frequencies, but when the effective population size is large then natural selection can have a very large impact on allele frequencies by decreasing the frequency of disadvantageous alleles and increasing the frequency of advantageous alleles (Casillas and Barbadilla, 2017). The effective population size differs from the total population size because rather than including all sexually mature individuals, this value only considers the number of individuals who can contribute genetic material to the next generation (Carlini, 2020). The effective population size (Ne) can be reduced by a population bottleneck event or an unequal sex ratio (Carlini, 2020). The latter has been observed in Drosophila melanogaster, where sexual selection against the males by the female flies limits the male’s opportunities to breed, the consequences of not every male participating in the donation of genetic material to the next generation is a smaller effective population size (Begun and Whitley, 2000).
Furthermore, the nature of X-linked inheritance and the faster-X hypothesis is another factor to consider when trying to understand how natural selection and genetic drift might impact the genetic variation of the red and white eye color alleles in a population of Drosophila (Casillas and Barbadilla, 2017). The unequal distribution of the X-chromosomes between the homogametic female sex (XX) and heterogametic male sex (XY) in Drosophila can result in smaller effective population sizes for X-linked genes, because the males are essentially haploid with respect to the X chromosome, and thus do not have as many X-linked genes to contribute to future generations which makes the alleles more vulnerable to genetic drift (Singh et al., 2005). Natural selection also acts more strongly on X chromosomes because in hemizygous males the potentially deleterious X-linked mutations will be expressed and removed from the population by the selective forces against that allele (Begun and Whitley, 2000; Singh et al., 2005). The result of this is that X-linked deleterious mutations such as the white eye mutation should decrease within the population over time, more quickly than a mutation would be removed on an autosome, and eventually become removed completely (Begun and Whitley, 2000).
Hardy-Weinberg Equilibrium (HWE) is the law that dictates what should happen to a population when no evolutionary forces are acting on a population (Carlini, 2020), this makes HWE a very important tool for the study of population genetics (Namipashaki et al., 2015). The Hardy-Weinberg Equilibrium principle imagines an ideal population where after a single generation of random mating an equilibrium is achieved where genotypic frequencies can be used to estimate the allele frequencies (and vice-versus) because they are functions of one another (Carlini, 2020; (Namipashaki et al., 2015; Waples, 2014). This equilibrium is maintained once HWE is reached, allowing allele frequencies to remain constant over time (Carlini, 2020; (Namipashaki et al., 2015; Waples, 2014). In order for HWE to be applied then several assumptions must hold true. The first assumption is that there is only random mating in a population, so this assumption is violated when a species chooses its mate either by sexual selection or when reproductive opportunities are limited for some individuals (Carlini, 2020; Namipashaki et al., 2015; Waples, 2014). The second assumption is that there are no mutations, as mutations would introduce new alleles that would violate this assumption (Carlini, 2020; Waples, 2014). The third assumption is that there is no natural selection occurring, this means that all of the individuals within a population have equal evolutionary fitness, if some individuals had an advantageous genetic variation then this assumption would be violated (Carlini, 2020; Namipashaki et al., 2015; Waples, 2014). The fourth assumption is that the population is so infinitely large that genetic drift cannot occur, thus finite populations would violate this assumption (Carlini, 2020; Namipashaki et al., 2015; Waples, 2014). The fifth assumption is that there is no migration between populations because this would introduce gene flow, thus if any new alleles are introduced to or lost from a population due to migration then this assumption would be violated (Carlini, 2020; Namipashaki et al., 2015; Waples, 2014).
Obviously, the problem with Hardy-Weinberg Equilibrium is that in most populations, one or all of these assumptions are violated. Still, HWE is useful for population genetics because it can be used to predict allele frequencies from genotypic frequencies, which is how this experiment working with Drosophila will predict the frequency of the white (w) allele and red (W+) allele for the female flies in each generation. The most common way to check if a population is in equilibrium is to conduct a Chi-Square test to assess the goodness-of-fit, in which the observed genotypes are compared to what should be expected if the population were in HWE (Namipashaki et al., 2015). This test measures the amount of deviation in the observed population from HWE, significant deviation would reveal that one of the five assumptions has been violated, meaning that an evolutionary force is most likely acting on the population (Namipashaki et al., 2015; Waples, 2014).
Materials and Methods
This experiment began with a group of 40 Drosophila, which had an equal proportion of red eye males, red eye females, white eye males, and white eye females. This means that the initial frequency of the red eye\ (W^+) allele was 0.5 and the initial frequency of the white eye (w) allele was also 0.5. This population, referred to as the parental generation was housed in a modified fly breeding chamber. This chamber was constructed from a plastic shoebox that had four holes cut into the bottom to accommodate removable food bottles. The food bottles and lids were secured with putty to prevent any fly escapes. The breeding cage also had a fifth hole, which was sealed with a removable foam plug. The adult parental generation was initially transferred into the cage by anesthetizing them with carbon dioxide gas after they were placed into the enclosure the foam plug was reinserted and they were allowed to wake up by themselves. The adult flies were allowed to breed and lay their eggs in the food medium inside the removable food bottles for two weeks. After two weeks, three of the food bottles containing Drosophila eggs and larvae were removed and replaced with fresh bottles. The remaining fourth bottle was left in place to allow the first generation of flies (produced from the parental generation) to grow up and breed with one another. The three food bottles containing generation one larvae and eggs were sealed with a foam plug and a week later they were anesthetized with ether and their eye color phenotype and sex were recorded. This process of replacing jars and anesthetizing flies to record their phenotype was repeated until the third generation.
Determining Phenotypes and Genotypes
Phenotypic determinations were made by anesthetizing the flies with ether and carefully observing them with a microscope to check if they had a white or red eye color, and then by looking for identifiable sex characteristics to determine if they were male or female. This data was stored in a chart to be used for allelic and genotypic determinations. The flies were manipulated carefully with paintbrushes to prevent accidentally harming them.
Since red and white eye color are X-linked traits, female flies (XX) will have two copies of the eye color allele, while male flies (XY) will only have one copy. Since males only have one copy of the eye color allele, the observed phenotypic frequency matches the allelic and genotypic frequencies (Carlini, 2020).
Percent change was used to quantify the changes in the allelic and genotypic frequencies between generations. This was calculated by subtracting the frequency of an allele/genotype in the current generation by the frequency of an allele/genotype in the previous generation, then dividing this value by the frequency of an allele/genotype in the previous generation, and finally multiplying this value by 100 (Equation 8 below).
Table 1: Phenotypic Counts in Drosophila Across Generations
The parental generation consisted of 40 flies in equal proportion: 10 red eye females, 10 white eye females, 10 red eye males, and 10 white eye males. The parental generation mated with each other and produced generation 1 which consisted of 144 flies: 67 red eye females, 15 white eye females, 43 red eye males, and 19 white eye males. The 1st generation mated and produced generation 2 which consisted of 304 flies: 120 red eye females, 59 white eye females, 72 red eye males, and 53 white eye males. The 2nd generation was then again allowed to mate, and these mattings produced the 3rd generation which consisted of 205 flies: 101 red eye females, 17 white eye females, 64 red eye males, and 23 white eye males. This information is summarized in Table 1 above.
Table 2: Allele Frequencies for Eye Color across Drosophila Generations
The frequency of the red eye (W+) and white eye (w) alleles in each generation of flies was calculated separately for the males and females. For the male flies, the allelic frequency was equivalent to the phenotypic frequency. As shown above in Table 2, in the parental generation the allele frequency of W+ and w were both 0.5. In the first generation, the frequency of the W+ allele was higher than the frequency of the w allele (0.69 versus 0.31). In the second generation, W+ allele was still more common than the w allele, with frequencies 0.58 and 0.42, respectively. Finally, in the third generation, there was the largest difference between the two allele frequencies, the frequency of the W+ allele was 0.74 and the frequency of the w allele was 0.26.
The allelic frequencies for the female flies had to be calculated using HWE. As shown above in Table 2, the allele frequencies in the parental generation were originally in equal proportion, both W+ and w were equal to 0.5. In the first generation, the frequency of the W+ allele was 0.57 and the frequency of the w allele was 0.43. In the second generation, the W+ allele became less common than the w allele, with frequencies 0.43 and 0.57, respectively. Finally, in the third generation, W+ allele became more common with a frequency of 0.62 and the w allele became less common as the frequency fell to 0.38.
Figure 1: Displaying the frequency in the red eye allele (W+) within the male population of Drosophila over three generations. Generation 0 (parental generation) had a red eye allele (W+) frequency equal to 0.5, Generation 1 had a red eye allele (W+) frequency equal to 0.69, Generation 2 had a red eye allele (W+) frequency equal to 0.58, Generation 3 had a red eye allele (W+) frequency equal to 0.74.
As shown above in Figure 1, the red eye allele (W+) frequency increased from 0.5 in the parental generation to 0.69 the 1st generation, but then decreased to 0.58 in the 2nd generation, before finally increasing to 0.74 in the 3rd generation.
Figure 2: Displaying the changes in allele frequencies between fruit fly generations, presented as present change in allele frequencies between generations of flies. These plotted points show the change from the parental generation to generation 1 (Gen 1), from generation 1 to generation 2 (Gen 2), and from generation 2 to generation 3 (Gen 3).
Shown above in Figure 2 is a graph that displays the percent change between Drosophila generations. Between generation 0 and generation 1, the red eye (W+) allele increased by 38.7% for the males and 14.5% for the females, while the white eye (w) allele decreased by 38.7% for the males and 14.5% for the females. Between generations 1 and 2, the red eye (W+) allele decreased by 16.9% for the males and 25.1% for the females, while the white eye (w) allele increased by 38.4% for the males and 34.2% for the females. Between generation 2 and generation 3 the red eye (W+) allele increased by 27.7% for the males and 44.7% for the females, while the white eye (w) allele decreased by 37.6% for the males and 33.9% for the females.
Table 3: Genotypic Frequencies for Eye Color across Drosophila Generations
Table 4: Chi-Square Test for Hardy-Weinberg Equilibrium using Male Drosophila Data
Shown above in Table 4, are the three chi-square tests for Hardy-Weinberg Equilibrium in Generations 1 through 3. These tests only looked at the male flies to see if the observed phenotypes match those that would be expected if the population were in Hardy-Weinberg Equilibrium. The observed values were simply the number of males observed with the specified phenotype, and the expected values were calculated by simulating mating outcomes between the flies in the previous generation and then multiplying those values by the number of males in the current generation. The Chi-Square value in generation 1 is 9.29 and the P-value is 0.002304, for generation 2 the Chi-square value is 0.00699 and the P-value is 0.933322, and for generation 3 the Chi-square value is 33.63 and the P-value is 0.00001.
Table 5: Estimations of the Selection Coefficient (s) Against the White (w) Allele in Male Drosophila
Figure 3: Displaying changes in the Selection Coefficient (s) against the white allele (w) across generations of male Drosophila.
The Selection Coefficient or (s) against the white allele (w) was calculated for Generations 1 through 3 for male Drosophila, this is displayed in Table 5 and Figure 3 above. In the first generation the Selection Coefficient was s=0.258, in the second generation the Selection Coefficient was s=0.004, in the third generation the Selection Coefficient was s=0.409. Across all three generations, the Selection Coefficient averaged out to s=0.224.
Figure 4: Displaying changes in allele frequency (as percent change) versus the Selection Coefficient for that generation. These plotted points show the change from the parental generation to generation 1 (Gen 1) with a selection coefficient of s=0.258, from generation 1 to generation 2 (Gen 2) with a selection coefficient of s=0.004, and from generation 2 to generation 3 (Gen 3) with a selection coefficient of s=0.409.
Shown above in Figure 4 are the changes in allelic frequencies between fruit fly generations plotted against the Selection Coefficient (s) for each generation. Showing that when selection against the white (w) allele was a large value (s=.258 and s=0.409) then the frequency of the white allele (w) decreased, and when selection against the white (w) allele is very small (s=0.004) then the frequency of the white allele (w) increased.
Figure 5: Displaying changes in genotypic frequency (as percent change) versus the Selection Coefficient for that generation of Drosophila. These plotted points show the change from the parental generation to generation 1 (Gen 1) with a selection coefficient of s=0.258, from generation 1 to generation 2 (Gen 2) with a selection coefficient of s=0.004, and from generation 2 to generation 3 (Gen 3) with a selection coefficient of s=0.409.
In order to determine if the Drosophila population was in HWE, a Chi-square test was conducted for each generation which compared the expected and observed phenotypes of male Drosophila (shown in Table 4). If the observed data for the Drosophila populations was in accordance with the expectations of HWE then the Chi-square test would reveal very little deviation from the expectations, and thus the result would be the failure to reject the null hypothesis. On the other hand, if the deviation is great then this would mean the observed data for the Drosophila populations was not in accordance with the expectations of HWE because one of the assumptions has been violated, thus the null hypothesis should be rejected.
For generation one, the Chi-square value was chi^2= 9.29 and the P-value was 0.002304, at a level of 5% significance with one degree of freedom this is a statistically significant deviation from the expected values. Therefore, the null hypothesis is rejected because the allele frequencies in the male flies do not match that which would be expected if the population of Drosophila in generation one was in accordance with HWE.
For generation two, the Chi-square value was chi^2= 0.00699 and the P-value was 0.933322, at a level of 5% significance with one degree of freedom this is not a statistically significant deviation from the expected values. Therefore, the null hypothesis cannot be rejected because the allele frequencies in the male flies are similar to that which would be expected if the population of Drosophila in generation three was in accordance with HWE. It is important to point out that this does not mean that the null hypothesis is true because there are other factors that can explain these results, and failure to reject HWE does not necessarily rule out the possibility that one of the assumptions was violated (Waples, 2014).
For generation three, the Chi-square value was chi^2= 33.63 and the P-value was 0.00001, at a level of 5% significance with one degree of freedom this is a statistically significant deviation from the expected values. Therefore, the null hypothesis is rejected because the allele frequencies in the male flies do not match that which would be expected if the population of Drosophila in generation three was in accordance with HWE.
The rejection of the null hypothesis and deviations from HWE in generations one and three suggest that one of the five HWE assumptions has been violated, while the failure to reject the null hypothesis in generation two needs further examination to explain what happened (Namipashaki et al., 2015). Although, there is also the possibility that some of the deviations from HWE could be explained by experimental error, such as incorrectly sexing the flies which could change the allele proportions within sexes.
It possible that the random mating HWE assumption has been violated for several reasons, including an unequal sex ratio in generations one through three, mate selection based on courtship rituals and breeding behavior, and another factor is that not all of the flies become sexually mature at the same time which can prevent some flies from having an equal chance to mate (Flatt, 2020). Another possibility is that the no mutations HWE assumption has been violated, mutations could introduce new alleles or reverse existing mutations, but given that the rate of mutation is very low and that the population size is very small it is very unlikely that mutation would be observed or have any significant effect on these Drosophilas over just three generations (Waples, 2014). It is also unlikely that the no migration HWE was violated because inside the breeding chamber was sealed well with putty and foam to prevent escapes or invaders, however, it is possible that a few flies might have escaped during the phenotyping process, but this would only affect the allele frequencies of the generation being counted and not their progeny.
It seems very likely that genetic drift would have occurred in the population because this experiment used small finite Drosophila populations, which violates the HWE assumption that the population is so infinitely large that genetic drift cannot occur. The consequences of genetic drift occurring in the Drosophila populations would lead to random changes in the allele frequencies over time resulting in loss of genetic variation, and fixation of either the red or white eye allele. It is almost impossible that this population could have escaped the effects of genetic drift because the Drosophila experiment began with a parental generation consisting of just 40 flies which is a very small population size, and genetic drift acts more strongly on small populations. In subsequent Drosophila generations, the effective population size (Ne) is reduced by the initial bottleneck effect, the unequal sex ratio (62 males to 82 females in generation one, 125 males to 179 females in generation two, and 87 males to 118 females in generation three), sexual selection against male flies by females, and the unequal distribution of X-linked genes between the homogametic male (XY) and heterogametic female (XX) flies.
Based on the results of this experiment and widely available Drosophila literature it seems possible that the last HWE that there is no natural selection occurring has also been violated. Existing research into the W+ gene has revealed that flies with the red eye (W+) allele breed faster and are more successful at mating than their white eyed (w allele) counterparts (Xiao et al., 2017). The red eyed flies not only possess superior breeding abilities, but they also have an increased central nervous system function in comparison to the white eyed flies (Xiao et al., 2017). The red eye flies have functioning W+ gene, which is advantageous because this allows them to recover quickly from anoxia-induced comas (such as when they are anesthetized by carbon dioxide gas or ether), and this gene is also linked to increased learning and memory, and preference for having a large exploratory territory (Xiao et al., 2017). The white eye flies have the null mutated version of the W+ gene which is disadvantageous, leading to reduced ability to recover from anoxia-induced comas, a reduced exploratory territory size, reduced learning ability to learn memory, and light-sensitivity (Xiao et al., 2017). The consequence of having these two alleles in the population is that the advantageous red eye (W+) allele has higher fitness so the frequency is expected to increase over time, while the disadvantageous white eye (w) allele has lower fitness, and thus the frequency of this allele is removed from the population over time (Waples, 2014).
The process of natural selection in these populations of Drosophila was observed as changes to the allele frequencies between the generations. As shown in Figure 2, the red eye (W+) allele increased by 38.7% for the males and 14.5% for the females from the parental generation to the first generation, while the white eye (w) allele decreased by 38.7% for the males and 14.5% for the females. The Selection Coefficient for generation one was s=0.258 (shown in Table 5 and Figure 3), which means selection against the white allele was moderately high. It also makes sense that the white (w) allele was selected against more strongly in the male flies than in the female flies (38.7% vs 14.5% decrease) because selection does not act directly on the heterozygote female (X^W+,X^w).
The data for the second generation showed that the Selection Coefficient was very low (s=0.004), and an opposite trend was observed when the red eye (W+) allele decreased by 16.9% for the males and 25.1% for the females between the first and second generation while the white eye (w) allele increased by 38.4% for the males and 34.2% for the females (shown in Figures 2 & 3, and Table 5). This was surprising because it appears to support that the population was in HWE as discussed above. However, this is not necessarily the case, especially given the selection observed in the two other generations and the rejection of HWE in those situations because one of the principles of HWE is that once it is achieved it is maintained. So, if the second generation was actually in accordance with the assumptions of HWE then generation three should have been as well, but this was not the case. Therefore, this is data abnormality has several possible explanations. The first possible explanation is that some of the original homozygous red eye females in the parental generation were injured when they were transferred into the breeding container if such injuries impaired their ability to reproduce then the first generation would have a higher level of red eye heterozygotes, which would, in turn, could have resulted in the higher frequency of white (w) allele seen in the second generation. Another possible explanation for these results is that the second generation had the largest population which could have reduced the effect of genetic drift, being made up of 304 flies, in contrast to the first generation which only had 144 flies and the third population which had 205 (shown in Table 1) (Casillas and Barbadilla, 2017). On the contrary, genetic drift is a random occurrence so maybe it just happened to act more strongly on this generation and negated the effects of natural selection, which temporarily increased the proportion of the white allele. This would be possible if for whatever reason the effective population size was made up of more flies carrying the white eye (w) allele. Another possibility is that for whatever reason it took longer for the red eye flies in generation one to hatch and reach sexual maturity, which would mean that the white eye flies could have mated and laid their second-generation eggs before the red eye flies (Flatt, 2020). The effects of this could translate into proportionally more of the white eye flies having had completed the egg to adult transformation by the time phenotypic counts were taken. Finally, the last factor to take into consideration is that some of the red eye flies escaped during the phenotypic counting process despite being anesthetized with ether, which resulted in some of the red eye flies not having been reported in the data. Ironically, this ability to quickly recover from anoxia-induced comas is one of the very reasons that the red eye (W+) allele has been reported to have a higher fitness than the white eye (w) allele.
The trends in the third generation mirrored that of the first generation. As shown in Figure 2, the red eye (W+) allele increased by 27.7% for the males and 44.7% for the females, from the parental generation to the first generation, while the white eye (w) allele decreased by 37.6% for the males and 33.9% for the females. The Selection Coefficient for generation one was s=0.409 (shown in Table 5 and Figure 3), which means selection against the white allele was very high. The increase in the Selection Coefficient from generation one to three is most likely due to the nature of X-linked traits, where selection first initially acted against the hemizygous white males but then as time went on also acted more strongly against the female flies. This idea is supported by the fact that the white eye (w) allele and homozygous white eye genotype (X^w,X^w) decreased most rapidly in female flies as the duration of the experiment went on (Figures 2, 4 & 5).
In summary, the allele frequencies changed the most quickly at the beginning of the experiment because the initial exposure of the white allele in hemizygous male flies to selection removed the white allele from the population. After this generation, the ability of the red eye (W+) allele to increase was limited by the existence of the heterozygote form in females which remained high throughout the experiment (shown in Table 3). Furthermore, as shown in Figure 4, a decrease in the frequency of the white eye (w) allele was highest when natural selection acted against it in generations one and two, when the Selection Coefficient was moderately strong (s=.258 and s=0.409). When the Selection Coefficient was very weak in generation two, the frequency of the white allele (w) increased because the selective pressures against it were not strong enough to remove it from the population. However, as discussed above, there are many explanations other than the absence of natural selection that explains this abnormality. Overall the across all three generations the Selection Coefficient against the white eye (w) allele averaged out to s=0.224, which is similar to reports of 25% selection against the white eye (w) allele in other experimental Drosophila populations (Xiao et al., 2017).
In conclusion, the increase in the frequency of the red eye (W+) allele from 0.5 in the parental generation to 0.74 in male Drosophila and 0.62 in female Drosophila by the third generation shows that the red eye (W+) allele has increased fitness in comparison to the white eye (w) allele that decreased from 0.5 in the parental generation to 0.26 in male Drosophila and 0.38 in female Drosophila by the third generation. This observed change in allele frequencies supports the idea that natural selection and evolution are observable phenomenons. The overall Selection Coefficient s=0.224 supports the claim that the white eye (w) allele has reduced fitness and is removed by natural selection over time. Finally, the results of the Chi-square test for fitness of Hardy-Weinberg Equilibrium show that there is a significant deviation from the expectations of HWE which reveals that one or more of the five HWE assumptions have been violated.
Begun, D. J., and P. Whitley. (2000). Reduced X-linked nucleotide polymorphism in Drosophila simulans. Proceedings of the National Academy of Sciences, 97(11):5960-5965.
Bronstein, O., A. Kroh, and E. Haring. (2018). Mind the gap! The mitochondrial control region and its power as a phylogenetic marker in echinoids. BMC Evolutionary Biology, 18(80).
Carlini, D. (2020) Natural Selection Experiment. BIO-356 Genetics with Laboratory. American University Department of Biology: Washington, D.C.
Carlini, D. (2020) Lecture #25: Migration & Genetic Drift. BIO-356 Genetics with Laboratory. American University Department of Biology: Washington, D.C.
Carlini, D. (2020) Lecture #22: Population Genetics I Hardy-Weinberg Equilibrium. BIO-356 Genetics with Laboratory. American University Department of Biology: Washington, D.C.
Casillas, S., and A. Barbadilla. (2017). Molecular Population Genetics. Genetics, 205(3): 10003-1035.
Cook, L. M., and I. J. Saccheri. (2012). The peppered moth and industrial melanism: evolution of a natural selection case study. Heredity (Edinb), 110(3): 207-212.
Flatt T. (2020). Life-History Evolution and the Genetics of Fitness Components in Drosophila melanogaster. Genetics, 214(1), 3–48.
Heil, C. S. S., M. J. Hunter, J. KF. Noor, K. Miglia, B. Manzano-Winkler, S. R. McDermott, and M. AF. Noor. (2012). Witnessing Phenotypic and Molecular Evolution in the Fruit Fly. Evolution (N Y), 5(4): 629-634).
Namipashaki, A., Z. Razaghi-Moghadam, and N. Ansari-Pour. (2015). The Essentiality of Reporting Hardy-Weinberg Equilibrium Calculations in Population-Based Genetic Association Studies. Cell Journal, 17(2): 187-192.
Singh, N. D., J. C. Davis, and D. A. Petrov. (2005). X-Linked Genes Evolve Higher Codon Bias in Drosophila and Caenorhabditis. Genetics, 171(1): 145-155.
Xiao, C., S. Qiu, and R. M. Robertson. (2017) The white gene controls copulation success in Drosophila melanogaster. Scientific Reports, 7(7712).
Waples, R. S. (2014). Testing for Hardy–Weinberg Proportions: Have We Lost the Plot? Journal of Heredity, 106(1): 1-19.