Step 1: Set up what "divergence" means here.
Percent divergence just means how many bases differ between the two sister species at that spot in the genome, expressed as a percentage. A region that mutates faster, or whose mutations are not removed, ends up with higher divergence over the same span of time.
Step 2: Think about what happens to a mutation in each region.
A mutation inside the exon usually changes the protein the gene builds. Since the gene is under purifying selection, protein changing mutations tend to be harmful, and natural selection removes carriers of those mutations from the population, so most new mutations in the exon never spread and never show up as a fixed difference between species. A mutation inside the intron, or in the flanking region outside the gene, usually has no effect on any protein, since introns get spliced out and flanking DNA is not part of the coding sequence. With no protein effect, selection mostly ignores these mutations, and they drift to fixation at close to the background mutation rate.
Step 3: Compare the three regions using this logic.
The exon (E) is filtered by selection, so its divergence stays low. The intron (I) and the flanking neutral region (N) are both essentially unfiltered, so they should build up differences at a similar pace, giving $I \approx N$, and both should be higher than $E$.
Step 4: Check each answer choice against this.
Any option placing $E$ above $I$ or $N$ is backwards, since selection suppresses divergence in the exon rather than raising it. The option placing $E$ equal to $I$ ignores the fact that the exon is specifically constrained while the intron is not.
Step 5: Conclude.
Only the ranking where the exon diverges the least, and the intron and neutral region diverge about equally and more, matches how purifying selection actually works.
\[ \boxed{E \lt I = N} \]