SHOX (skeletal atavism)

What it does

Crossing two of them

Its gene carrot

Gene key

horsegenetics.shox

Priority

91

Alleles

sa N

Outcomes

wild, shox-carrier, skeletal-atavism

Coat effect

none — every outcome is a wild type

Default allele

N

Wild population

2.0% of founders carry one copy; none carries two

Founder draws

1 × nextFloat()

Catalogue entries

1 — it never changes what a horse looks like

The combination table

On a sex chromosome, and still an autosomal recessive

The region it sits in is shared between the X and the Y, so it segregates exactly like an autosome even though it lives on a sex chromosome. That is why it needs none of the sex-linked inheritance scaffolding that sex-linked inheritance is still waiting on — and if that scaffolding ever lands, this is the gene to check it against, because it must keep behaving exactly the way it does now.

What SHOX does, and why “atavism” is a description not a mechanism

SHOX is short stature homeobox. It encodes a homeobox transcription factor — a DNA-binding developmental regulator that helps control the growth and patterning of bone, particularly during limb development.

In the horse it matters for the radius–ulna and tibia–fibula complexes. A modern horse normally has reduced, largely fused ulnae and fibulae, which is part of what gives it a long, stiff distal limb and efficient one-toed locomotion. In skeletal atavism the ulna and fibula develop more fully and stay separate and elongated, reaching toward or across the carpus and tarsus. That resembles ancestral equid anatomy — hence the name — but it is a pathological developmental defect, not a genetic reversion to an ancestral horse. The inference is that normal SHOX dosage and regulation are needed to suppress or coordinate that growth pattern; SHOX also influences cell-cycle regulation and osteogenic-cell activity, which fits the severity. Rafati et al., PMC4938674

Two deletions, not one allele

The molecular association comes from a 2016 whole-genome sequencing study of affected Shetland ponies, with long-read sequencing of BAC clones to resolve a difficult region. It found two large overlapping deletions, both of which are disease alleles.

Why a sex-chromosome locus inherits like an autosome

Both deletions sit in the pseudoautosomal region (PAR) of the X and Y. The PAR is the homologous segment the two sex chromosomes share, where recombination happens. So although the locus is physically on a sex chromosome, it transmits like an autosomal one: colts and fillies inherit and pass it the same way, and this is not a conventional X-linked disorder. The PAR recombines and is not subject to ordinary female X-inactivation the way the rest of the X is.

The clearest wording is pseudoautosomal recessive inheritance with autosomal-like transmission — which is exactly why the mod can model it as a plain autosomal recessive and needs none of the sex-linked scaffolding (roadmap) that is still unbuilt. If that scaffolding ever lands, this is the gene to check it against, because it must keep behaving the way it does now.

A technical caveat worth keeping: this region was poorly represented in older horse reference genomes, being repeat-rich and structurally complex. Long-read sequencing improved the local assembly but did not resolve every breakpoint precisely, and long tracts of repeated sequence — TGGA repeats among them — may be what generated the rearrangements in the first place.

Inheritance

A carrier pairing on either deletion gives the usual 25% affected, 50% carrier, 25% clear.

What an affected horse looks like

A disproportionate limb-skeletal disorder: markedly short legs against the body, a low rectangular outline, abnormally developed and elongated ulnae and fibulae, failure of the normal ulna/radius and fibula/tibia fusion, narrowing and deformity around the knees, severe angular and rotational deformity in fore and hind limbs, clubfoot-like conformation in some foals, and limb angles and gait that get worse as the animal grows.

Affected foals may initially survive but often cannot move normally, and the problems are not cosmetic — they compromise loading of the carpus, hock and distal limb. Many affected horses are euthanised early; UC Davis notes euthanasia is common by six months because the impairment is progressive.

Carriers: nothing detected, which is not nothing found

The study did not find a significant withers-height difference between carriers and non-carriers in its sample of 130 horses after accounting for sex — comparing N/Del-1 and N/Del-2 against N/N.

That is not proof that carriers have no small effect on conformation, limb proportion or growth. It means the published sample did not detect one, and the authors explicitly cautioned that their carrier sample was limited and may have lacked the power to see a subtle difference. So: established, two deletions cause severe skeletal atavism when biallelic; not established, that carriers are reliably smaller, shorter-legged, stronger, more athletic or more pony-like; possible but unproven, subtle carrier effects on skeletal proportion or SHOX-regulated growth.

This matters because small stature in horses is genetically crowded. ACAN and the size loci LCORL/NCAPG, HMGA2 and TBX3 all bear on normal and extreme size variation and on other dwarfing syndromes. SHOX skeletal atavism should not be folded in with miniaturisation generally — and in the mod it is not: it sits on the health layer, not among the size genes.

Coat

No evidence that these deletions affect base colour, black/red pigment production, dilution, grey, white spotting, appaloosa patterning, or mane, tail, feather and hair texture. SHOX is a developmental skeletal regulator; contemporary equine genomic summaries file it under body size and skeletal development, away from MC1R, ASIP, KIT, STX17, PMEL and MITF. If a colour looks linked to the disorder inside one family or breed, that is pedigree structure or chance, not a SHOX coat effect.

CRLF2, the unresolved passenger

Both deletions also remove part of CRLF2, which encodes a cytokine-receptor-related protein in haematopoietic and immune signalling, with pathways touching bone metabolism. The evidence still favours SHOX loss or misregulation as the driver: Del-1 removes the SHOX coding sequence outright, Del-2 leaves it intact and probably removes the regulatory elements it needs, and human SHOX deficiency produces short stature and limb deformity, which is strong cross-species support.

What is unresolved is whether CRLF2 loss modifies the skeletal phenotype or adds effects of its own in affected horses or carriers. The original paper says this cannot be excluded and that possible non-skeletal effects remain unexplored. That is a stated uncertainty, not evidence that affected horses have a known immune disease — and the mod models none.

Jumping and speed

For affected horses the answer is not interesting: the limb geometry, joint deformation, rotational and angular deviation and progressive locomotor impairment make jumping or speed work medically inappropriate.

For carriers, no published evidence demonstrates any advantage or disadvantage in jump height or scope, gallop speed, racing performance, stride length, turning, endurance, gait quality or soundness under sport workloads. Athletic performance is highly polygenic and heavily shaped by training, conformation, management, rider and injury history; reviews associate speed, stamina, gait and jumping with MSTN, DMRT3, COX4I2 and PDK4, not SHOX.

Breeds and frequency

Shetland Pony. The breed with the strongest molecular evidence; Del-1 and Del-2 were discovered and validated there, and OMIA lists Shetland Pony as the documented breed for both. OMIA 002013 In a random sample of 94 Swedish Shetlands born between 1968 and 2000, Del-1 allele frequency was 4.79%, Del-2 was 1.06%, and the combined observed carrier frequency was 11.7% — usually rounded to about 12%.

Those are historical and population-specific figures and should not be read as the current frequency in every country, registry or line. Testing could push them down over time; founder effects or a popular sire could raise them locally.

Miniature Horse. Skeletal atavism has been clinically reported in miniatures, including an 8-month-old filly with severe rotational and angular deformity and radiographically complete ulnae and fibulae. PMID 15373256 The 2016 molecular study included 18 US horses, four of them apparent cases found while investigating another congenital dwarfism disorder, and laboratories now list Miniature Horse alongside Shetland Pony for testing. But there is no peer-reviewed, representative allele-frequency estimate for American Miniatures comparable to the 94-pony Swedish survey — the phenotype is documented, testing is recommended, and a breed-wide carrier frequency is not established. UC Davis VGL

Everything else. The specific Del-1 and Del-2 deletions are firmly documented as causal in Shetland Ponies, with testing guidance extending to Miniatures, likely on shared miniature-pony ancestry. That should not be generalised to all pony breeds, drafts, sport horses or full-sized horses without breed-specific evidence — skeletal malformation and dwarfism-like phenotypes have different causes in different populations. The mod’s WILD_CARRIER_PERCENT of 2.0 is once again a world-wide founder number standing in for a ~12% figure inside one small breed.

What is settled, and what is not

The breeding answer is testing before pairing at-risk animals, not indiscriminate culling: a carrier bred to a clear mate cannot produce an affected foal from this disorder. Test both deletions, because either contributes; do not read athletic ability, coat or ordinary height off carrier status; and remember that a negative result for these two deletions does not rule out every other cause of dwarfism, chondrodysplasia, congenital angular limb deformity or orthopaedic disease.