Ecology and Evolution · Unit 2: Genomes, development and diversification

Evolution of Development

How can a body part shrink or disappear in a lineage while the gene that builds it still makes a normal protein? This lesson follows pelvic reduction in freshwater threespine sticklebacks and the Pitx1 gene through five kinds of evidence, from a mapped chromosome region to an engineered enhancer deletion in mice. Afterwards you can say what each method answers, read a cross and an F1 expression test, and name the questions the work leaves open.

Watch first Same Protein, Different Switch · 3:09

Marine threespine sticklebacks carry a pelvis armed with two spines. In Paxton Lake, British Columbia, the bottom-feeding benthic sticklebacks have a greatly reduced pelvis and most have no pelvic bones at all, yet the protein made by Pitx1, the gene most strongly tied to the difference, is unchanged in the fish examined. How can a body part shrink while its gene still makes a normal protein?

The answer is that a gene’s protein and a gene’s pattern of use are separate things, and evolution can change one without the other. This lesson follows the stickleback evidence step by step, because each method answers its own question, and the order in which they were applied is what turns a hunch into a case.

The puzzle: a lost pelvis with a working gene

The threespine stickleback, Gasterosteus aculeatus, lives in the sea along northern coasts and in many lakes that marine fish colonised after the last ice sheets retreated. Marine fish have a pelvic girdle with a pair of long spines. In several lakes, the populations have a reduced pelvis or almost none. The stickleback pelvic fin is homologous to your hind limb: the two structures are inherited, in modified form, from a common ancestor (see reading phylogenetic trees).

Pitx1 encodes a homeodomain transcription factor, a protein that binds DNA and switches other genes on or off. Mice lacking Pitx1 protein have malformed hind limbs, among other defects. If stickleback Pitx1 were broken outright, you would expect damage wherever the gene works, including the jaw, where it is also expressed.

The way out is the enhancer: a short stretch of noncoding DNA that binds transcription factors and raises a gene’s expression in particular tissues at particular times. A gene can have several, one for the jaw and another for the pelvis, like separate light switches for separate rooms wired to the same bulb. Break the pelvic switch and the jaw light stays on. How enhancers work is taught in eukaryotic gene regulation; here the question is how you would show that evolution used one.

Mapping and expression: one major region, four minor

Shapiro, Kingsley and colleagues reported in 2004 that pelvic reduction in their stickleback crosses mapped to one major and four minor chromosome regions. The major region sat near Pitx1. The method is linkage mapping: cross the two forms, breed the hybrids together, and look for chromosome regions whose genotype travels with the trait in the offspring.

Worked problem, part 1. These data are invented, in the style of a stickleback cross. A marine fish (pelvis complete) is crossed with a Paxton benthic fish (pelvis reduced). The F1 fish are intercrossed, giving 400 F2 fish. Each is scored for pelvic reduction and genotyped at a marker beside Pitx1: M is the marine allele, P the Paxton allele. Assume all fish are raised together and scored on the same scale.

  1. Frequency of reduction in each genotype. MM: 2 reduced of 100, or 2%. MP: 6 of 200, or 3%. PP: 32 of 100, or 32%.
  2. Read dominance. One Paxton copy (MP) barely changes the frequency from MM, while two copies raise it about tenfold. The region acts close to recessively.
  3. How much does the region explain? Overall, 2 + 6 + 32 = 40 of 400 fish, or 10%, are reduced. Yet 68 of the 100 PP fish (68%) are not reduced, and 8 reduced fish carry a marine allele. So other loci or the environment also matter, as the four minor regions in the real study suggest.

A mapped region shows that something in it is linked to the trait. It can hold many genes, and it says nothing about whether the change lies in the protein or its regulation. The 2004 paper went further on both points: the Pitx1 protein sequence was unchanged in the pelvic-reduced fish examined, and Pitx1 transcript, detected by in situ hybridization (a labelled antisense RNA probe that pairs with the transcript in a whole larva), was missing from the pelvic region of Paxton larvae and present in the jaw. The original figures had used a sense probe by mistake. The 2006 correction, with an antisense probe, reported retained mouth and lower-jaw expression, lost pelvic expression, and no tail expression in either population. Cite the correction for the expression pattern.

Cis or trans: the F1 test

Lower expression in the pelvis could come from a change near Pitx1, acting in cis (on the same DNA molecule), or from a change in some other gene that supplies Pitx1’s regulators, acting in trans (through a diffusible product). Mapping cannot separate these, because a trans regulator could sit in the same region.

The F1 hybrid can. Each hybrid cell holds one marine and one Paxton allele in the same nucleus, bathed by the same transcription factors. If the regulators were the problem, both alleles would be affected alike. If the difference is built into one allele’s DNA, that allele alone is down.

Worked problem, part 2. Same invented cross. In F1 pelvic buds, 90% of Pitx1 transcripts come from the marine allele and 10% from the Paxton allele. In jaw tissue the split is 50:50.

  1. Cis or trans? The two alleles share one trans environment, so a 90:10 ratio means the Paxton allele itself is less active. The change is cis-acting.
  2. Tissue-specific? The jaw ratio is 50:50, so the Paxton allele works normally there. The defect is specific to the pelvis.
  3. What has improved since part 1? Mapping alone could not exclude another gene in the region. The allele-specific result points to a cis-regulatory change at Pitx1 itself.
F1 allele test: cis or trans?⚙ original · interactive
Tissue
Hypothesis

In this sketch: invented data from the lesson, one nucleus shared by both alleles, shares of transcripts in relative units, no other source of difference between the alleles. The lesson predicts only that a cis change gives unequal shares in the pelvic bud, so no size is predicted.

Both alleles share one nucleus, so a difference built into one allele’s DNA lies in cis (that allele alone is down), while a change in a regulator would affect both alleles alike (trans). This is a sketch of the logic, not a measurement.

Chan and colleagues ran the real version in 2010: in F1 hybrids, the Paxton benthic allele was expressed less in pelvic tissue but not in head tissue.

Enhancers on trial: reporters, rescue and a mouse deletion

The same 2010 study identified Pel (later called PelA), a noncoding sequence upstream of Pitx1 that can drive expression in the developing pelvic fin. In a reporter assay, the candidate sequence is joined to a minimal promoter and a visible marker, and placed in embryos. The marine version drove pelvic expression; the Paxton version, which carried a deletion, gave pelvic expression in 0 of 8 transgenic fish tested. A reporter answers one question: can this sequence drive expression here? It does not say the sequence matters for form.

Rescue comes closer. Chan’s group put a Pitx1 minigene driven by marine Pel into fry from Bear Paw Lake, Alaska, another pelvic-reduced population, and pelvic spine development was variably enhanced compared with uninjected siblings. “Variably” is the paper’s hedge and stays.

In 2018 Thompson, Capellini, Kingsley and colleagues found a second hind-limb enhancer, PelB, downstream of Pitx1. They deleted 9,425 base pairs containing it at its native place in the mouse genome using CRISPR/Cas9 (methods in CRISPR and functional genomics). Homozygous mice were viable and fertile, and several hind-limb bones were shorter. Their hind-limb Pitx1 RNA, measured at embryonic day 12.5:

relative decrease = (control - mutant) / control = (1.00 - 0.85) / 1.00 = 0.15, or 15%

About 85% of the RNA remained. Patellae were absent only in mice carrying the PelB deletion opposite a Pitx1 null allele, a sensitized background. This enhancer tunes how much Pitx1 the limb makes; it is not an on-off switch for the whole limb.

In sticklebacks, the Paxton benthic PelB carried a 125-base-pair deletion and a 341-base-pair insertion, and its reporter drove weaker, more restricted pelvic expression than the marine one while keeping strong jaw expression. Paxton fish therefore carry changes in both PelA and PelB.

Why might Pel deletions recur? Xie and colleagues (2019) found that yeast artificial chromosomes carrying marine Pel broke about 25 to 50 times as often as an empty control, and the reverse-complement orientation broke 10 to 20 times less often than the forward one. That is a yeast assay, not a fish mutation rate, and it speaks to how deletions arise, not why they spread.

What stays open

The evidence ladder, in order: mapping localises, in situ hybridization shows where transcript is, the F1 test separates cis from trans, reporters show what a sequence can drive, and rescue or a native deletion tests whether form depends on it. Four questions remain:

  • PelA versus PelB. The fish reporters do not split the natural contributions of the two enhancers; the 2018 authors proposed engineered fish to do so.
  • Other loci. The four minor regions, and background, are not explained.
  • The agent of selection. Why reduction spread in particular lakes is not established by this work, and this lesson does not name a cause.
  • Generality. A 2023 study of sticklebacks from two small subarctic Norwegian lakes and a nearby marine site found no obvious association between its tested PelA and PelB variants and pelvic status in the lake where spineless fish occur, so the Pel story should not be assumed for every pelvic-reduced population. How form evolves at the gene level is compared across lineages in molecular evolution, and how ancient developmental toolkits were reused in how signaling evolved; the Hox genes are another case of shared developmental control.

Design problem

Paxton fish carry changes in both PelA (upstream) and PelB (downstream). Design an experiment to measure each one’s contribution to pelvic size.

One way to do it

Work at the native locus, not with reporters, because a reporter shows what a sequence can drive in a construct, not its effect on bone. Start with marine (pelvis-complete) fish and use genome editing to install the Paxton-type change at PelA alone, at PelB alone, and at both. Make several independent lines of each so that an off-target edit in one line cannot explain the result, and keep unedited siblings from the same families as controls.

Raise all genotypes together in the same tanks and diet, so that the environment is shared, and score pelvic structures blind to genotype on a stated ordinal scale (spine length and girdle extent, summed across both sides), with a sample size chosen in advance. Measure pelvic Pitx1 RNA in the same lines to link expression to form.

Predictions: if the effects add, the double edit should be about the sum of the single edits. If they interact, the double edit will be larger or smaller than the sum; for example, each alone may do little if the other enhancer compensates. Finally, one marine background may not represent others, so repeat in a second marine population, and keep the result to the backgrounds tested.

How we measure it

Linkage mapping in experimental crosses

Two populations that differ in a trait are crossed, the hybrids are bred together, and hundreds of offspring are scored for the trait and genotyped at markers across the genome. A region whose genotype co-segregates with the trait is linked to a cause, but it can hold many genes.

Whole-mount in situ hybridization

A labelled antisense RNA probe that pairs with a gene's transcript is applied to a whole embryo or larva, showing where the transcript is present. A sense probe is the negative control; mixing the two up was the error the 2006 stickleback correction fixed.

Allele-specific expression in F1 hybrids

In a hybrid, both parental alleles share one nucleus and one set of transcription factors. Measuring how much RNA comes from each allele, tissue by tissue, tells you whether an expression difference is built into the allele's own DNA (cis) or comes from the cell's other factors (trans).

Reporter assays and engineered deletions

A candidate enhancer is placed beside a minimal promoter and a reporter such as GFP or lacZ to ask where it can drive expression. Deleting the enhancer at its native locus with CRISPR/Cas9 asks a different question: does the organism's form depend on it?

Finish line

Test yourself

8 questions

Practice and MCAT-style, with explained answers.

Next · Lesson 7

Speciation

Two monkeyflowers can be crossed by hand. Why do they almost never exchange genes in the wild?

Continue →

← Lesson 5 · Molecular Evolution

Michael D. Shapiro

Michael D. Shapiro

First author of the 2004 Nature paper that mapped pelvic reduction in threespine stickleback crosses to one major and four minor chromosome regions, with the major region near Pitx1, and that found no change in Pitx1 protein sequence in the pelvic-reduced fish examined.

The question Shapiro and his co-authors, in David Kingsley's laboratory and with collaborators including Catherine Peichel, set out to answer was whether a large change in form in the wild could be traced to particular genes. Marine threespine sticklebacks carry a pelvis with a pair of spines; some lake populations have a much reduced one. The team crossed a pelvis-complete fish with a pelvic-reduced Paxton Lake benthic fish, bred the hybrids together, and looked for chromosome regions that travelled with the trait. They found one region of major effect and four of minor effect, and the major one lay near Pitx1, a gene already known from mice to matter for hind limbs. Sequencing showed no change in the Pitx1 protein in the reduced fish they examined. In 2006 the authors published a correction: the original in situ figures had used a sense probe, and the corrected antisense experiment still showed lost pelvic expression with retained jaw expression. What the work made possible was a specific hypothesis to test: a regulatory change at Pitx1.

✓ verbatim Shapiro et al., Nature 428:717-723 (2004); correction, Nature 439:1014 (2006).
Yingguang Frank Chan

Yingguang Frank Chan

First author of the 2010 Science paper showing that the Paxton benthic Pitx1 allele is expressed less in the pelvis of F1 hybrids, that a deleted version of the upstream Pel enhancer drives no pelvic reporter expression, and that a marine Pel-driven Pitx1 transgene variably enhanced pelvic spines in Bear Paw Lake fish.

A mapped region near a gene is a strong lead, but it is a long stretch of chromosome, and it does not say whether the gene itself or something beside it carries the change. Chan and colleagues in the Kingsley laboratory tackled that gap with tests of different kinds. In F1 hybrids, where a pelvis-complete and a Paxton allele sit in the same nucleus, they compared the output of each allele and found the Paxton allele reduced in pelvic tissue but not in head tissue. They identified a noncoding sequence upstream of Pitx1, called Pel, that could drive pelvic expression in transgenic fish, and found that the Paxton version, carrying a deletion, gave pelvic expression in none of eight fish tested. Finally they put a Pitx1 minigene driven by marine Pel into fry from Bear Paw Lake, Alaska, and pelvic spine development was variably enhanced compared with uninjected siblings. That set of results turned a mapping peak into a named enhancer that could be tested directly.

✓ verbatim Chan et al., Science 327:302-305 (2010).
Abbey C. Thompson

Abbey C. Thompson

First author of the 2018 eLife paper that found a second Pitx1 hind-limb enhancer, PelB, downstream of the gene, deleted it at the native locus in mice, and compared marine and Paxton PelB reporters in sticklebacks.

By 2018 the stickleback story had one enhancer in it. Thompson, with Terence Capellini, Kingsley and colleagues, asked whether Pel was the whole of Pitx1's hind-limb control. Using overlapping bacterial artificial chromosome clones across an 850-kilobase stretch of the mouse Pitx1 locus, they found a second hind-limb enhancer downstream of the gene and named it PelB. They then used CRISPR/Cas9 to delete 9,425 base pairs containing PelB at its own place in the mouse genome. The mice were viable and fertile, hind-limb Pitx1 RNA fell to about 85% of control, and several hind-limb bones were shorter. In sticklebacks, the Paxton benthic version of PelB, which carried a 125-base-pair deletion and a 341-base-pair insertion, drove weaker pelvic reporter expression than the marine version while keeping strong jaw expression. Paxton fish thus carried changes in two enhancers. The paper closed by proposing engineered fish with each change alone and both together, the experiment that could separate their effects.

✓ verbatim Thompson et al., eLife 7:e38555 (2018), open access under CC BY 4.0.
Kathleen T. Xie

Kathleen T. Xie

First author of the 2019 Science paper reporting that marine stickleback Pel sequence made a yeast artificial chromosome break about 25 to 50 times more often than a control, while the reverse-complement orientation broke 10 to 20 times less often than the forward one.

Chan's 2010 work had found deletions at Pel in pelvic-reduced fish from more than one place, which raised a question about mutation rather than selection: is this stretch of DNA unusually easy to break? Xie and colleagues, again with the Kingsley laboratory, tested it in budding yeast. They placed marine Pel sequence on a yeast artificial chromosome and counted how often the chromosome broke. Against an empty-chromosome control of 3.37 breaks per million cell divisions, Pel-containing chromosomes broke roughly 25 to 50 times more often, and the effect depended strongly on which way the replication machinery moved through the sequence. In sticklebacks, they used CRISPR to make breaks at Pel and reported that these could give local deletions and pelvic loss in genetic crosses. The yeast number is not a fish mutation rate. What the study made possible was a testable account of why the same enhancer may be hit again and again, kept separate from the question of why the deletions spread.

✓ verbatim Xie et al., Science 363:81-84 (2019).
Test yourself
Question 1 of 8 Score 0 / 0

Pelvic-reduced sticklebacks examined by Shapiro and colleagues in 2004 carried a Pitx1 gene with what feature?
Sources & Integrity

Every structure, slide, simulation, and clip on this page is listed below with the source it came from. None of the 3D structures is drawn by hand or by AI: each one loads by accession ID from RCSB PDB or PubChem. This panel is generated from the lesson's own data file, so what it lists is what the page loads.

Simulations

  • Cis or trans? An allele-specific test · Original interactive — Biochemistrypedia (Deppmann Lab).

Scientist profiles & quotes

  • Michael D. Shapiro — Shapiro et al., Nature 428:717-723 (2004); correction, Nature 439:1014 (2006).
  • Yingguang Frank Chan — Chan et al., Science 327:302-305 (2010).
  • Abbey C. Thompson — Thompson et al., eLife 7:e38555 (2018), open access under CC BY 4.0.
  • Kathleen T. Xie — Xie et al., Science 363:81-84 (2019).

Quotes are verbatim from the cited source. Profile text is original to this site and was checked against the sources cited.

Animated explainer

Same Protein, Different Switch · narration: an AI narrator voice (Gemini, Charon) reading narration written from this lesson's text · animation: AI-generated in code (Manim) by Claude, a schematic with no drawn molecular structures, plots computed from their equations · captions from the narration script.

Images & figures

This lesson carries no raster figures. Every structure is rendered live from RCSB PDB or PubChem, and every plot is drawn from its equation in the browser.

integrity Machines check the mechanics. The build confirms that every local file exists, that every question's answer key points at a real choice, and that any AI-generated scientific figure names the person who checked it; a weekly job re-resolves every accession ID against its database. Questions are original; MCAT-style items are written for this site and are not real AAMC questions. Banned sources (the Tymoczko textbook, publisher decks, copyrighted figures) are not used.

Like I'm five

A gene is like a recipe, and the body has switches that say where to use it. Some lake fish have the same recipe for a hip-bone helper as sea fish, but the switch for using it in the hips is broken. So the recipe still gets used in the mouth, but not much in the hips, and those fish grow only a small hip.
High school

Threespine sticklebacks from the sea have a pelvis with two spines, but some lake populations have a much smaller one. Scientists crossed the two kinds and found that the biggest difference sat near a gene called Pitx1. The protein it makes was unchanged. Instead, a stretch of nearby DNA called an enhancer, which turns Pitx1 on in the developing pelvis, was damaged in the lake fish. Pitx1 was still switched on in their jaws. Changing where a gene is used, rather than the protein it makes, let one body part change without disturbing others.
College

Pelvic reduction in Paxton Lake benthic sticklebacks maps to one major and four minor regions, the major one near Pitx1, a homeodomain transcription factor needed for hind-limb development. The Pitx1 coding sequence was unchanged in the fish examined. In F1 hybrids the Paxton allele is expressed less than the pelvis-complete allele in the pelvis but not in the head, which places the change in cis. A deleted Paxton version of the upstream enhancer Pel drove no pelvic reporter expression, and a marine Pel-driven Pitx1 transgene variably enhanced spines in Bear Paw Lake fish. Because enhancers are modular, a tissue-specific enhancer lesion can change one structure while the gene's other roles, which loss of the protein would also remove, continue.
Graduate

Each method answers its own question: linkage localises, in situ hybridization shows where transcript is detectable, F1 allele-specific expression separates cis from trans, reporters test sufficiency in a construct, and rescue or native-locus deletion test dependence of form. Thompson et al. (2018) added PelB downstream of Pitx1. A native mouse PelB deletion left about 85% of bulk hind-limb Pitx1 RNA, shortened several bones, and removed patellae only opposite a Pitx1 null allele, so enhancers here tune dose rather than act as all-or-none switches. Paxton fish carry changes in both PelA and PelB, and a reporter cannot apportion their natural effects. Xie et al. (2019) found Pel fragile in a yeast chromosome assay, offering a mutational reason for recurrent deletions, which is separate from why the deletions spread.

Open questions: what are the separate and combined skeletal effects of the Paxton PelA and PelB changes at the native fish locus, which other loci explain the minor regions, what ecological agent favoured reduction in particular lakes, and why some populations, such as the subarctic Norwegian lake sampled in 2023, show no obvious association with the tested variants.
The probe that was the wrong way round

In situ hybridization uses an antisense RNA probe, which pairs with a gene's transcript; the matching sense probe should give little specific signal and serves as a control. In 2006 the authors of the 2004 stickleback paper published a correction stating that the original Pitx1 expression figures had been made with a sense probe. With the correct antisense probe, they reported Pitx1 expression in the mouth and lower jaw of Paxton benthic larvae but none in the pelvic region, and no expression in the tail of either marine or Paxton larvae. The central conclusion survived; the published pictures did not.

— Shapiro et al., Correction, Nature 439:1014 (2006).
Eighty-five per cent of what?

The mouse PelB result is easy to misquote. Thompson and colleagues measured Pitx1 RNA in whole hind limbs at embryonic day 12.5, normalised to the housekeeping gene Pgk1, in seven wild-type and eight homozygous mutant samples. The mutant limbs kept about 85% of the control level. That is a 15% drop in a bulk average. It does not say every cell lost 15%, and it does not translate into a 15% shorter bone. The skeletal effect was a set of modest reductions in the lengths of several hind-limb bones.

— Thompson et al., eLife 7:e38555 (2018), qRT-PCR of E12.5 mouse hind limbs.
A fragile stretch of DNA, measured in yeast

To ask whether Pel breaks easily, Xie and colleagues did not count mutations in fish families, which would take many generations. They moved the sequence into budding yeast on an artificial chromosome, where breakage can be scored over millions of cell divisions. An empty chromosome broke at 3.37 per million divisions, and chromosomes carrying marine Pel broke roughly 25 to 50 times as often, and the reverse-complement orientation broke 10 to 20 times less often than the forward one. The result describes that yeast assay, not the rate at which wild sticklebacks gain deletions.

— Xie et al., Science 363:81-84 (2019).

 ↑ Top