Chromosome 2
Human chromosome 2 carries the telomere and centromere remnants of a fusion of two ancestral ape chromosomes.
6 min readUpdated
Chimpanzees, gorillas, and orangutans have 24 pairs of chromosomes. Humans have 23. If humans share an ancestor with the other great apes, that ancestor almost certainly had 24 pairs, since three of the four living lineages still do.
A chromosome pair cannot simply be lost — chromosomes carry hundreds of genes and deleting a pair is fatal. Two pairs fusing into one is survivable, and fusion leaves physical traces that can be looked for at specific locations.
What a fusion would leave behind
Two features of chromosome structure make the prediction testable.
Telomeres are repetitive caps on chromosome ends. In vertebrates the sequence is TTAGGG, repeated thousands of times. They occur at the ends and nowhere else; their function is to stop the cell from treating the chromosome end as broken DNA in need of repair.
Centromeres are the constricted regions where spindle fibers attach during cell division. A normal chromosome has exactly one, marked by its own repetitive sequence — alpha satellite DNA in primates. A chromosome with two active centromeres is pulled toward opposite poles during mitosis and torn apart.
If two chromosomes fused end to end, the result should show:
- Telomere repeats stranded in the middle of the chromosome, at the join, running head to head — because the two ends met facing each other.
- A second centromere region that has been deactivated, since two working centromeres are lethal.
- Gene order matching two specific chimpanzee chromosomes laid end to end.
None of these has a function. A telomere in the middle of a chromosome does nothing. A dead centromere does nothing.
What was found
| Prediction | Result | Source |
|---|---|---|
| Head-to-head telomere repeats at the join | Found at band 2q13: arrays of TTAGGG repeats in inverted orientation, in the middle of the chromosome | IJdo et al., PNAS 88 (1991) |
| A deactivated second centromere | Alpha satellite DNA at 2q21.3-q22.1, non-functional, at the position corresponding to the chimpanzee 2B centromere | Avarello et al., Human Genetics 89 (1992) |
| Gene order matching two chimp chromosomes | Human chromosome 2 corresponds along its full length, in order, to chimpanzee chromosomes 2A and 2B joined at the fusion point | Chimpanzee Sequencing Consortium, Nature 437 (2005) |
The 1991 paper is titled "Origin of human chromosome 2: an ancestral telomere-telomere fusion."
The sequence order matters. Banding studies in the 1970s and Yunis and Prakash's 1982 comparison proposed that human chromosome 2 corresponded to two ape chromosomes, before the region was sequenced. The molecular confirmation came in 1991, and the full genome comparison in 2005.
Why fusion spreads through a population
An individual carrying a fusion has 23 pairs while everyone else has 24. Their offspring would be expected to have pairing problems during meiosis, which looks like a barrier to the fusion ever becoming universal.
This process is observable in living populations.
| Case | Detail |
|---|---|
| Humans today | Robertsonian fusions occur in roughly 1 in 1,000 births. Carriers are typically healthy with reduced but real fertility |
| House mice | Populations in different Alpine valleys carry different fusion arrangements, with chromosome counts from 22 to 40 |
| Horses | Przewalski's horse has 33 pairs, the domestic horse 32. They interbreed and produce fertile offspring |
The fertility cost is real and survivable, and fixation is easiest in small isolated populations — the situation of early hominin groups.
Common objections
"The fusion site is inside a functional gene, so it isn't a fusion scar"
Jeffrey Tomkins of the Institute for Creation Research argued this in the Answers Research Journal in 2013, in a paper titled "Alleged Human Chromosome 2 'Fusion Site' Encodes an Active DNA Binding Domain Inside a Complex and Highly Expressed Gene." His claim is that the site sits within DDX11L2, that this region is transcribed, and that a functional gene is not a fusion scar.
The transcription is real. The inference does not follow.
DDX11L2 belongs to a family of DDX11L sequences located near telomeres across many human chromosomes, and they are generally regarded as pseudogenes — degraded copies rather than protein-coding genes. Transcription is not the same as function; a large fraction of the human genome is transcribed at low levels, including regions known to be non-functional.
More importantly, the argument addresses only one of the three lines of evidence. It says nothing about the inverted head-to-head orientation of the repeats, nothing about the deactivated centromere at 2q21, and nothing about the gene order matching chimpanzee 2A and 2B end to end. A later insertion into a fusion site does not explain why the site is there.
"The telomere repeats are degenerate, so they aren't real telomeres"
Accurate as an observation. The repeats at the fusion site are degraded, and the region contains interspersed non-telomeric sequence.
This is what the model predicts. A junction sitting in a genome for roughly six million years accumulates mutations, and there is no selection maintaining a telomere that no longer caps anything. Pristine repeats would be harder to explain than degraded ones, since they would imply the sequence was being actively preserved.
"A designer could have created humans with 23 pairs from the start"
The common design argument, made by Answers in Genesis and others: similarity reflects a shared designer rather than shared ancestry.
The argument has some force for functional similarities. Two aircraft from different manufacturers both have wings, and that is not evidence of a shared factory.
It does not extend to these features, because they are not functional. Holding that chromosome 2 was created as it stands requires holding that a designer placed a non-functional telomere array in the middle of a chromosome, in head-to-head orientation, at exactly the point where two chimpanzee chromosomes would meet if joined — and then placed a disabled centromere at exactly the position of the chimpanzee 2B centromere.
Neither does anything. Both are what descent with fusion predicts.
"This just shows humans and chimps are similar, not related"
Similarity alone would be weak evidence, and common design would be a live explanation for it.
The evidence here is not similarity but a specific shared error at a specific location. The distinction is the same one used in plagiarism detection: matching correct answers prove little, while matching mistakes in the same places are difficult to explain independently.
The pattern also extends. Inversions and translocations distinguish human and chimpanzee chromosomes at several other loci, and which apes share which rearrangements produces the same family tree as anatomy, endogenous retroviruses, and other genetic markers.
"This means humans evolved from chimpanzees"
Not a creationist objection but a common misunderstanding.
The evidence indicates humans and chimpanzees share a common ancestor. Chimpanzees are not ancestral to humans; their lineage has been evolving for exactly as long as ours, and the ancestor was neither a human nor a chimpanzee.
What the evidence shows
Human chromosome 2 carries three independent traces of an end-to-end fusion: inverted telomere repeats at the join, a deactivated centromere at the expected position, and gene order matching two chimpanzee chromosomes joined end to end.
The general prediction was made from chromosome banding in the 1970s and early 1980s, before the region was sequenced. The molecular evidence was found where it was predicted to be.
None of the three features is functional, which is what makes them evidence of history rather than of design. The mechanism by which such fusions spread is observed in living populations, including humans.