Using experiments to test patterns of inheritance
Modern genetics began with the experiments of Gregor Mendel, an Austrian monk with an inquisitive mind. Mendel
performed crosses between garden pea plants and discovered that certain alleles can mask one another (i.e. dominant
alleles mask recessive ones described earlier). A monohybrid cross is one in which the pattern of inheritance of a single
trait (e.g., pea shape) with a pair of alleles (e.g., round or wrinkled peas) is studied. In contrast, a dihybrid cross
involves parents that are identical except for two independent traits (e.g., pea color and shape).
Morgan built upon Mendel’s findings through his discovery that traits may be sex-linked, meaning the gene is carried
on a sex-determining chromosome (X or Y). In Drosophila, the X-chromosome carries sex-linked genes. So in flies,
females have two copies of the X chromosome, and therefore carry two alleles of sex-linked genes, one on each X
chromosome. Male Drosophila, however, only have one X chromosome and therefore only carry one allele of a sexlinked gene. This means that whatever allele is passed on to males through their single X chromosome will be
expressed regardless of whether it is dominant or recessive. Since female flies inherit two alleles for sex-linked genes,
expression of the alleles follows the dominant/recessive pattern described above. Any genes that are present on the
other chromosomes—that is, not on X or Y chromosomes—are said to be autosomal. For example, all of Mendel’s pea
plant traits are inherited on autosomal chromosomes.
Alleles that are naturally common in the wild (e.g., red eyes in fruit flies) are known as wild-type alleles. Less common
alleles (e.g., white eyes in fruit flies) are known as mutant alleles. These are derived from characteristics that are
expressed naturally in the wild. It’s important to note that wild-type alleles are not always dominant, nor are mutant
alleles always recessive.
Reciprocal crosses were important to Morgan’s discovery about sex-linked genes. A reciprocal set of crosses is
composed of a forward cross (where the male parent has the mutant allele and the female parent has the wild-type
allele) and a reverse cross (where the female parent has the mutant allele and the male parent has the wild-type allele).
One project in the Duronio lab, for example, developed around the hypothesis that a mutated histone prevents DNA in a genome from properly replicating and repairing itself. To test it, the lab chose to focus on fruit fly larval tissue, which contains cells that transform into the wing and part of the thorax. Armstrong, with PhD student Taylor Penke, spent more than 10 hours dissecting upwards of 600 fruit fly larvae for analysis. “After that, we had approximately 5 million cells to manipulate,” Duronio says.
(For example number of legs, body segments, antennae, eyes etc.)
Genetics: Testing Hypotheses about Inheritance
There are also other modes that can be used to explain the inheritance of two alleles which are different than the ones
1. Both alleles are autosomal and located on the same chromosome (linked) [you still have to determine whether
each allele is dominant or recessive]
2. One allele is autosomal while the other allele is sex-linked [you still have to determine whether each allele is
dominant or recessive]
There are several ways to generate predictions from hypothesized modes of inheritance. We will use Punnett
rectangles because they are simple method and easy to remember. For the alleles we will study in lab, it is necessary to
predict the results both of parental crosses—producing the first (F1) generation—and of F1 crosses—producing the
second (F2) generation. The predicted phenotypic ratios are converted to expected frequencies and will be used to
compare with the frequencies of real (or simulated) crosses to test your hypotheses.
Each mode of inheritance (i.e., alternative hypothesis) has a unique F2 generation prediction, and therefore the results
of experiments can be compared to the predicted value to determine how alleles are passed on from parents to
offspring. The best model will be the one whose predictions do not differ statistically from the actual cross values (do
not result in the null hypothesis being rejected). We will use the Chi-Square Goodness-of-Fit test to determine
whether or not our cross data actually matches our predictions.