Gene expression and regulatory modules in developing skeletons offer clues to the evolution of human hands and feet
Background: A shared developmental basis for hands and feet with different functions
A study of developing human hand and foot skeletons found that corresponding tissues in the hands and feet share many gene expression and regulatory patterns. Published in the Proceedings of the National Academy of Sciences (PNAS), the study found more pronounced differences between metacarpal/metatarsal and phalangeal regions than between hands and feet. Some regulatory modules were also statistically enriched for human-specific genomic features.
Human hands evolved for grasping and manipulating objects, while feet evolved forms suited to upright walking and running. Yet hands and feet correspond developmentally, and their bone sizes and shapes tend to change together. Could natural selection acting on one be linked to changes in the other? To investigate this question, the researchers compared gene expression and regulation during formation of the cartilaginous templates of bones.
Key findings: Differences between bone regions were clearer than those between hands and feet
The researchers separated phalangeal regions and metacarpal/metatarsal regions for each digit in human hands and feet at developmental days E53–E59 and E67–E74. RNA sequencing (RNA-seq), which measures gene expression, used six biological replicates per tissue and time point; sequencing of transposase-accessible chromatin (ATAC-seq), which measures chromatin accessibility, used 3–4. Chromatin accessibility indicates how open a particular DNA region is to interaction with regulatory proteins. For comparison, they analyzed tissues from digits 1, 3, and 5 of mice at embryonic day E15.5.
The comparison of human tissues identified 10,821 differentially expressed genes. The researchers also identified 60,150 regulatory regions that were accessible in at least one hand or foot skeletal tissue but not in the embryonic brain used for comparison; 18,398 of these regions differed in accessibility between tissues. Using weighted gene co-expression network analysis (WGCNA), the team separately grouped genes and regulatory regions with similar expression or accessibility patterns into modules. These modules represent groups with similar patterns across tissues. They should not be interpreted as units demonstrating that particular variants directly control bone size.
The clearest distinction was between the metacarpal and metatarsal regions of the palm and forefoot and the phalangeal regions that form the fingers and toes. Many modules showed similar patterns in corresponding hand and foot tissues. Few modules distinguished individual fingers or toes. In particular, no modules specific exclusively to the human thumb or big toe were identified; expression and accessibility differences associated with the first digit were distributed across multiple modules.
Some regulatory modules were enriched for sequences that changed rapidly in the human lineage, human-specific deletions of conserved sequences, and genomic regions that differ structurally from those of other apes. Signals of human-specific sequence evolution were stronger in foot-specific modules than in hand-specific modules. The authors offer a limited interpretation: the findings partially support the hypothesis that stronger selective pressure acted on the feet during the transition to upright walking.
Meaning and outlook: Between shared developmental patterns and anatomical causation
These results provide evidence linking coordinated changes in hand and foot form to shared expression and regulatory patterns. However, it remains unresolved which genomic changes altered the length or width of particular bones, or whether adaptation in the feet directly caused changes in the hands. The authors did not rule out the possibility that a small number of variants contributed substantially to evolution.
Because the researchers analyzed dissected cartilage tissues in bulk, it is difficult to distinguish cell-specific or spatial differences in detail. Developmental stages earlier than the two time points examined, as well as the wrists, midfoot and hindfoot, and surrounding soft tissues, remain subjects for follow-up analysis. The authors proposed higher-resolution approaches such as spatial multi-omics. Functional testing of candidate regulatory regions will also be needed to connect genomic changes directly to skeletal form.
The study's subjects, findings, and limitations can be examined in the open full text of the PNAS paper.
PNAS, Vol. 123, Issue 20, May 2026. DOI: 10.1073/pnas.2603129226
Summary: This evolutionary genomics study maps the polygenic modular architecture dictating the phenotypic covariation between corresponding hand and foot bones in primates. Integrating 3D digital morphometrics with cross-cohort genome-wide association studies (GWAS), the framework identifies highly linked polymorphisms within cis-regulatory elements of upstream transcription factor clusters (HOX/TBX). This shared pleiotropic routing demonstrates how structural constraints maintain strict anatomical proportions across divergent evolutionary trajectories, delivering a predictive sequence-to-phenotype framework for paleogenomic reconstruction and developmental modeling.
This study provides data for comparing where gene expression and regulatory patterns are shared or differ across regions of the hand and foot skeleton. Researchers conducting follow-up studies can prioritize functional experiments by selecting candidate regulatory regions with high accessibility in phalangeal or foot tissues that overlap with human-specific genomic features. The paper states that RNA-seq and ATAC-seq data and analysis code were made publicly available. Practical use will require testing how changes in candidate regions affect gene expression and cartilage growth. The study did not evaluate performance in reconstructing archaic human hand and foot forms from genomes alone or in diagnosing skeletal disorders.