Lamprey and Shark Embryos Reveal an Ancient Split in Hand Gene Expression

Hand genes are among the most familiar players in vertebrate development, encoding basic helix-loop-helix transcription factors that help build the lateral mesoderm, the heart, the branchial arches and the paired appendages. In most jawed vertebrates the family consists of two paralogs, hand1 and hand2, produced by ancient genome duplication. Yet the earliest branches of the vertebrate tree have remained stubbornly opaque: lampreys, which represent the jawless cyclostome lineage, were long thought to carry a single hand gene, and expression data from cartilaginous fishes were largely confined to pectoral fins. A new study published in Frontiers in Cell and Developmental Biology sets out to close that gap by examining Hand paralogs in the European river lamprey Lampetra fluviatilis and the grey bamboo shark Chiloscyllium griseum, combining phylogenetic reconstruction, local genomic synteny analysis and spatiotemporal expression profiling.

The first surprise concerns how many Hand genes lampreys actually possess. The researchers identified three Hand paralogs in the sea lamprey Petromyzon marinus, designated handα, handβ and handγ, and two corresponding genes in the hagfish Myxine glutinosa. Three matching sequences from the river lamprey were included directly in the analysis. Maximum-likelihood phylogenetic methods recovered the major gnathostome Hand1 and Hand2 groups, and the river lamprey sequences clustered consistently with their sea lamprey counterparts, while the bamboo shark sequences fell within the gnathostome Hand1 and Hand2 groups respectively. In other words, the correspondence of the lamprey α, β and γ copies between the two lamprey species is now firmly established.

What the phylogeny could not do is settle the deeper question of which, if any, cyclostome paralog corresponds to gnathostome hand1 or hand2. The relationships between cyclostome Hand paralogs and the gnathostome copies remained weakly resolved and were sensitive to the choice of substitution model and outgroup. Even when a divergent amphioxus Twist-like sequence was used to root the tree, no stable root emerged, and the authors treat the principal Hand phylogeny as unrooted. To test alternative orthology scenarios quantitatively, they compared constrained topologies with the approximately unbiased (AU) test. None of the tested placements of cyclostome Handα or Handβ with either gnathostome Hand1 or Hand2 was rejected, with AU P values ranging from 0.121 to 0.629. The sequence data simply do not carry enough signal to assign one-to-one orthology across the cyclostome-gnathostome split.

Local genomic synteny offered a complementary but equally inconclusive line of evidence. In gnathostomes, hand1 and hand2 loci sit in distinct and partly conserved neighborhoods: hand1 is frequently flanked by cnot8, faxdc2, larp1, sap30L and galnt10, whereas hand2 retains hpgd, cep44, fbxo8, scrg1, sap30 and galnt7. The paralogous combinations hand1-sap30L-galnt10 and hand2-sap30-galnt7 correspond to duplicated local genomic regions previously analysed in human, mouse and chicken. Cyclostome Hand loci retained only subsets of these neighboring genes, and their local contexts did not provide unambiguous evidence for one-to-one orthology. The authors therefore treat synteny as complementary evidence rather than an independent reconstruction of Hand duplication history, and they note that incomplete annotation of the lamprey assembly kcLamFluv1.1 around the handα and handγ loci means that missing neighbors cannot be read as gene loss.

With the gene inventory established, the team turned to expression. Using whole-mount in situ hybridization across lamprey embryogenesis, they found that handα expression begins earliest, appearing in scattered cells at stage 13 and expanding through the ventral and lateral regions of neurula-stage embryos, then alongside the neural tube, and into the ventral mesenchyme of head and trunk during head-process formation. At stage 25, handα marked the ventral branchial mesenchyme surrounding the endostyle, the medial stripe of the lower-lip mesenchyme, the cardiac region and the caudal region, and by stage 27 it was detected in branchial areas and the heart. By the latest stage examined, stage 30, handα expression was retained only in the cardiac region.

Handβ followed a different schedule. It was first detected during head-process formation in the ventral mesenchyme of the head process, then appeared in the ventral branchial mesenchyme and cardiac region at stage 25, with a weaker signal in the lower-lip mesenchyme and, unlike handα, no caudal expression at that stage. From stage 27 onward its expression intensified in the branchial areas and heart and persisted through stage 30, when a new domain appeared in the ventral caudal fin. This caudal-fin domain is the study’s most striking single observation: it extends Hand deployment to a median fin in a jawless vertebrate, something not previously described. Meanwhile, no specific handγ signal was detected at any stage or condition examined, although sequence analysis confirmed that the probe maps to a distinct handγ genomic locus. The authors caution that this negative result does not establish handγ as a pseudogene; it could reflect low or spatially restricted expression, expression outside the examined stages, transcriptional silencing, or pseudogenization not demonstrable from the present data.

The bamboo shark side of the study revealed equally clear paralog-specific patterns. At the earliest stages examined, hand1 was expressed along the entire dorsoventral extent of all branchial arches, though the signal was weak, diffuse and accompanied by background staining, and additional expression appeared in the otic vesicle, sensory ganglia and the cloacal region. At stage 26, hand1 expression appeared in the pectoral fins, encompassing much of the fin mesenchyme, and by stage 27 it was also detected in the pelvic fins and became more distally restricted in paired fins. No specific hand1 signal was detected in the dorsal fins under the whole-mount conditions used. Hand2, by contrast, produced a strong, well-defined signal with little background: in the branchial arches it was confined to ventral regions and strongest in the mandibular arch, with bright punctate domains at the ventral base of each arch, and from stage 26 onward it appeared in the posterior mesenchyme of the first and second dorsal fins and the pectoral fins, later extending to the pelvic fins and contracting to posterodistal regions distal to the radials by stages 29 to 30.

Quantitative RT-PCR on dissected fins added an important nuance to the in situ picture. In the pooled samples analyzed, hand1 transcript abundance in the dorsal fin was approximately 22-fold lower than in the pectoral fin and 24-fold lower than in the pelvic fin, whereas hand2 transcript abundance in the dorsal fin was only about 1.6-fold lower than in the pectoral fin and was similar in the pectoral and pelvic fins. Low levels of hand1 transcripts were thus detectable in the dorsal fin by qRT-PCR despite the absence of a specific whole-mount signal, indicating that the difference concerns the spatial distribution and relative abundance of expression rather than an absolute presence or absence of hand1 transcription. The authors also note that because hand1 staining was weaker and more diffuse at stages 24 to 29, fine-scale comparisons of signal boundaries should be interpreted cautiously, and that in the absence of functional perturbation these expression differences are not evidence of functional divergence.

The comparative picture that emerges is one of lineage- and paralog-specific diversification built on a shared ancient foundation. Both lamprey paralogs are expressed in the heart and ventral branchial region, and the authors point out that the composite expression pattern previously attributed to a single lamprey Hand gene actually combines contributions from distinct paralogs: the gene examined by Kuraku and colleagues corresponds to the α lineage, while another earlier study examined the β lineage. Comparison of lamprey and shark expression confirms and extends previous evidence that Hand deployment in ventral and branchial regions predates the cyclostome-gnathostome split, although the authors are careful not to infer direct homology of specific branchial cell populations, since whole-mount expression does not establish germ-layer or cell-lineage identity. The hand2 expression in both paired and median fins is consistent with the recruitment of components of an older median-fin program into paired appendages, and the preferential detection of spatially distinct hand1 domains in paired rather than dorsal fins suggests differential deployment among fin types rather than strict paired-fin specificity.

The authors are equally careful about the larger evolutionary claims. The shared deployment of Hand genes in the branchial apparatus and fins of sharks and other gnathostomes is compatible with proposed developmental links between branchial arches and paired appendages, including modern versions of Gegenbaur’s hypothesis, but overlapping expression of developmental regulators does not by itself demonstrate shared developmental competence or serial homology, and the observed patterns do not discriminate among competing homology models on their own. What the study does deliver, in the authors’ own assessment, is more concrete: the resolution of distinct cyclostome Hand paralogs, the identification of a lamprey handβ caudal-fin domain, and a comparative description of paralog-specific Hand expression across paired and median fins in a cartilaginous fish. For a gene family whose vertebrate history has been shaped by whole-genome duplication, lineage-specific polyploidization and subsequent rediploidization, even a partially resolved map marks real progress.

Subject of Research: Evolutionary diversification of Hand gene paralogs in lampreys and cartilaginous fish

Article Title: The hand of evolution: conservation and divergence of hand paralogs in cyclostomes and gnathostomes

Article References: Ermakova, G. V., Meyntser, I. V., Kucheryavyy, A. V., Zaraisky, A. G., & Bayramov, A. V. (2026). The hand of evolution: conservation and divergence of hand paralogs in cyclostomes and gnathostomes. Frontiers in Cell and Developmental Biology, 14, Article 1944613. https://doi.org/10.3389/fcell.2026.1944613

Image Credits: AI Generated

DOI: 10.3389/fcell.2026.1944613

Keywords: Hand genes, lamprey, bamboo shark, cyclostomes, gnathostomes, evolutionary developmental biology, phylogenetics, gene duplication, fins, branchial arches, whole-mount in situ hybridization, vertebrate evolution