7.4 Evidence for DNA as Genetic Material, Evolution & Quantitative Analysis

Evidence That DNA Is the Genetic Material

The AAMC lists Evidence That DNA Is Genetic Material as its own topic because MCAT passages frequently reproduce these experiments with a twist and ask you to interpret the controls.

1. Griffith (1928) — The Transforming Principle

Frederick Griffith injected mice with two strains of Streptococcus pneumoniae: a virulent smooth (S) strain with a polysaccharide capsule and an avirulent rough (R) strain without one.

The last row is the whole experiment: something released by dead S cells permanently converted R cells into S cells. Griffith called it the transforming principle but could not identify it.

2. Avery, MacLeod and McCarty (1944) — The Principle Is DNA

Avery's group purified the transforming extract and treated aliquots with degradative enzymes.

• Protease $\rightarrow$ transformation still occurred.
• RNase $\rightarrow$ transformation still occurred.
• DNase $\rightarrow$ transformation abolished.

Because only nuclease treatment destroyed activity, the transforming principle had to be DNA. Skeptics argued that trace contaminating protein might be responsible, which is exactly why a cleaner experiment was needed.

3. Hershey and Chase (1952) — The Blender Experiment

Hershey and Chase grew bacteriophage T2 in two separate radioactive media.

• $^{35}\text{S}$ labels protein only (cysteine and methionine contain sulfur; DNA has none).
• $^{32}\text{P}$ labels DNA only (the phosphodiester backbone contains phosphorus; amino acids do not).

Labeled phage infected E. coli; a kitchen blender sheared the empty capsids off the bacterial surface, and centrifugation separated the heavy bacterial pellet from the light phage-ghost supernatant.

• $^{35}\text{S}$ stayed in the supernatant — protein coats never entered the cell.
• $^{32}\text{P}$ appeared in the pellet and in progeny phage — DNA is the injected, heritable material.

4. Chargaff, Franklin and the 1953 Model

• Chargaff's rules: in any double-stranded DNA sample, $\%A = \%T$ and $\%G = \%C$, so purines equal pyrimidines. The A+T to G+C ratio varies between species but the pairing ratios do not.
• Rosalind Franklin's X-ray diffraction image (Photo 51) supplied the helical cross pattern, the $2\text{ nm}$ diameter, the $3.4\text{ \AA}$ base rise and the $34\text{ \AA}$ helical repeat (10 bp per turn).
• Watson and Crick (1953) combined complementarity with these dimensions to build the antiparallel double helix, immediately noting that base pairing "suggests a possible copying mechanism."

Passage cue: If a question gives you $\%G$ for a double-stranded genome, you can compute every other base. If a sample violates Chargaff's rules, the nucleic acid is single-stranded.

Evolutionary Mechanisms Beyond Hardy-Weinberg

Section 7.3 established Hardy-Weinberg equilibrium, genetic drift and the three modes of selection. Content Category 1C also names the following, which passages test directly.

Non-Random Mating: Inbreeding and Outbreeding

• Inbreeding is mating between close relatives. It increases homozygosity at every locus while leaving the allele frequencies $p$ and $q$ unchanged. Because $p^2 + q^2$ rises and $2pq$ falls, previously masked deleterious recessives are unmasked — inbreeding depression.
• Outbreeding (outcrossing) is mating between unrelated individuals; it increases heterozygosity and can produce hybrid vigor (heterosis).
• Both violate the random-mating assumption of Hardy-Weinberg, so a population may deviate from expected genotype frequencies without any selection occurring at all. Distinguishing "allele frequencies changed" (evolution) from "genotype frequencies changed" (assortative mating) is a recurring MCAT discrimination.

Polymorphism and Balanced Selection

A locus is polymorphic when two or more alleles persist at appreciable frequency. Balanced polymorphism is maintained when heterozygotes have the highest fitness:

• Sickle-cell trait ($HbA/HbS$): heterozygotes resist Plasmodium falciparum malaria while avoiding sickle-cell disease, so both alleles persist at high frequency in historically malarial regions.
• Frequency-dependent selection and heterozygote advantage at the MHC loci are the other standard examples.

Adaptation, Specialization and Fitness

Fitness ($w$) is relative reproductive contribution to the next generation's gene pool, not strength or longevity. Adaptation is a heritable trait raising fitness in a particular environment; specialization narrows a niche and raises efficiency at the cost of flexibility, which is why specialists are more vulnerable to environmental change. Selection can act at the level of the individual or, more controversially, of the group; evolutionary success is measured as an increase in percentage representation in the next generation's gene pool.

Evolutionary Time and the Molecular Clock

Neutral mutations accumulate at a roughly constant rate, so sequence divergence between two lineages is approximately proportional to time since their common ancestor. Comparing conserved proteins (cytochrome c, rRNA) across taxa builds phylogenies; the more divergent the sequence, the more distant the relationship. Homologous structures reflect shared ancestry (divergent evolution), whereas analogous structures reflect convergent evolution under similar selective pressure and must be excluded from phylogenetic inference.

Biometry: Statistical Methods for Genetic Data

The outline's Analytic Methods topic includes "biometry: statistical methods," and the MCAT's fourth scientific-inquiry skill is data-based reasoning.

• Descriptive statistics: mean, median, mode; standard deviation as spread; standard error of the mean, $SEM = s/\sqrt{n}$, shrinking as sample size grows. Non-overlapping 95% confidence intervals suggest a real difference between groups.
• Chi-square goodness of fit tests whether observed cross data match a Mendelian expectation:

$$\chi^2 = \sum \frac{(O - E)^2}{E}$$

Degrees of freedom equal the number of phenotypic classes minus one (a monohybrid cross with two classes has $df = 1$; a dihybrid $9{:}3{:}3{:}1$ has $df = 3$). If $p < 0.05$ the null hypothesis — that deviation is due to chance alone — is rejected, implying linkage, lethality or a non-Mendelian mechanism.

• Correlation is not causation, and heritability estimates apply to variance within a population, never to an individual.

Worked example. A dihybrid self-cross yields 90, 30, 28 and 12 offspring ($n = 160$). Expected $9{:}3{:}3{:}1$ values are 90, 30, 30, 10. Then $\chi^2 = 0 + 0 + 4/30 + 4/10 = 0.53$ with $df = 3$, far below the $0.05$ critical value of $7.82$, so the data are consistent with independent assortment.