Deep in the fast-flowing headwaters of northern Sumatra glides a fish that local communities have revered for generations as the king of the river — a powerful, bronze-flanked cyprinid that reaches imposing size among the boulders of the Batang Toru basin and around Lake Toba. Known to Batak communities as one of the ikan Batak, the Sumatran mahseer Neolissochilus sumatranus has long been woven into ceremony, cuisine and folklore, yet science has never been certain exactly what it is, how it evolved, or how it relates to the dozens of mahseer species swimming the rivers of Asia. Now researchers from Indonesia’s National Research and Innovation Agency (BRIN) and Pukyong National University in the Republic of Korea have answered part of that question by assembling the species’ first complete mitochondrial genome — the compact, maternally inherited chromosome packed into the cell’s energy factories that mothers pass, almost untouched, to their offspring. The study, published in Biochemical Genetics, turns that chromosome into a molecular archive of deep evolutionary time.

Mahseers — the large riverine carps of the genera Tor and Neolissochilus — are among Asia’s most celebrated freshwater fishes, pursued as sport fish, prized as food and protected as cultural icons from the Himalayan foothills to the Mekong and the islands of the Sunda Shelf. They are also among the continent’s most taxonomically baffling animals. Similar body shapes, variable coloration and overlapping habitats have fueled more than a century of naming disputes, and N. sumatranus, formally described by Weber and de Beaufort in 1916, sits squarely inside that confusion. Earlier DNA barcoding of North Sumatran mahseers exposed how deep the uncertainty runs, and the species has been assessed for the IUCN Red List amid mounting pressures from hydropower, overfishing and habitat degradation. Without a clear evolutionary framework, the authors argue, neither conservation planning nor sustainable fisheries management can proceed on solid ground. The genus holds evolutionary surprises of its own, including Neolissochilus pnar, the world’s largest known cave fish, described from Meghalaya in northeast India.

To build that framework, the team — led by corresponding authors Sekar Larashati of BRIN, and Hyun-Woo Kim and Shantanu Kundu of Pukyong National University — worked from a specimen morphologically identified as N. sumatranus, collected in northern Sumatra. Using next-generation sequencing, they read out the entire mitochondrial chromosome and annotated its genes with reference pipelines for fish mitogenomes. The assembled molecule measured 16,584 base pairs, squarely within the classic vertebrate range, and carried the standard complement of 37 genes: 13 protein-coding genes, two ribosomal RNA genes and 22 transfer RNAs, capped by an oversized non-coding control region that regulates replication and transcription. The sequence has been deposited in the public GenBank database under accession number PX634255, making it freely available to researchers worldwide. When the team compared the new genome with mitogenomes already published for nine other Neolissochilus species, the architecture proved strikingly conserved: the same genes, in the same order, a layout essentially frozen across millions of years of evolution in this branch of the carp family.

Within that conserved scaffold, the fine details revealed the selective pressures shaping the genome. Relative synonymous codon usage showed the protein-coding genes leaning heavily on codons for arginine, leucine and serine, a bias familiar from other fish mitochondrial genomes, where compositional pressures and mutation patterns repeatedly favor certain synonymous spellings of the same amino acid. Most of the thirteen protein-coding genes began with ATG, the canonical start instruction of vertebrate mitochondria. Several, however, ended in incomplete stop codons — a truncated TA- or a solitary T — which cannot by themselves terminate translation. The study explains the cellular workaround: during messenger RNA maturation, enzymatic polyadenylation adds a short tail of adenines to the transcript, converting the unfinished stub into a complete stop signal. The trick lets this extraordinarily compact genome, stripped of almost everything non-essential over evolutionary time, extract maximum coding value from every nucleotide it carries.

To measure the evolutionary forces acting on those genes, the researchers calculated the ratio of nonsynonymous to synonymous substitutions — dN/dS — for each protein-coding gene across the ten Neolissochilus mitogenomes. Twelve of the thirteen genes returned ratios below one: the unmistakable signature of purifying selection, in which mutations that alter the encoded protein are eliminated faster than silent changes because mitochondrial energy metabolism tolerates almost no tinkering. The lone exception was cytochrome b, or Cytb, which alone among the Neolissochilus genes failed to dip decisively below the threshold — a hint, the authors note, of lineage-specific dynamics that may reward closer functional scrutiny. Given that these genes power the oxidative phosphorylation machinery on which every heartbeat, swim and thought ultimately depends, the broader message is unambiguous: change is costly, and selection keeps the mitochondrial engine running on factory settings.

Transfer RNAs, the small adaptor molecules that read genetic codons and deliver the matching amino acids, told their own structural story. All 22 tRNA genes in N. sumatranus folded into the canonical cloverleaf configuration, with intact aminoacyl and anticodon stems and loops. The surprise came from tRNA-Ser1, the serine adaptor that in most vertebrate mitochondria loses part of its dihydrouridine arm, producing a characteristically shortened cloverleaf. In the Sumatran mahseer, that arm is fully present — no reduction, no truncation — a condition the authors highlight as a marker of structural variation within cyprinids. Because tRNA architecture can influence translation efficiency, and because tRNA genes frequently serve as hotspots for mitochondrial genome rearrangements over deep time, cataloguing these folded shapes across related species offers an evolutionary record of lineage history that single barcoding genes simply cannot match.

The control region — the genome’s largest non-coding stretch and the site where replication and transcription are orchestrated — proved the most variable part of the molecule. Comparing it across Neolissochilus species, the team identified four conserved sequence blocks, CSB-D, CSB-1, CSB-2 and CSB-3, each preserving a core that anchors the replication machinery while accumulating species-specific base changes on top. Tandem repeats, short motifs copied head-to-tail in variable copy number, appeared in only six of the eight species examined, and the region itself differed markedly in both length and nucleotide composition from one species to the next. That fluidity makes the control region a powerful comparative marker for telling closely related mahseers apart. The authors, however, add a caution: with a single specimen per species analyzed so far, the region’s usefulness as a population-level marker remains untested, since it cannot yet capture the variation that exists within each species rather than between them.

Reconstructing the family tree from the full mitogenome data, the researchers corroborated the monophyly of Torinae, the mahseer subfamily — the confirmation that all of these great river carps descend from a single common ancestor. Within that subfamily, N. sumatranus clustered closely with N. hendersoni and N. soroides, a trio the authors connect to a shared biogeographic history in Sundaland, the southeastern shelf of Asia where repeatedly falling Pleistocene sea levels exposed land bridges and rerouted rivers, letting freshwater fishes disperse between what are now separate islands and the mainland. Because mitochondrial DNA is inherited only through mothers, the tree traces matrilineal lines of descent — and the Sumatran fish’s closest maternal relatives emerge among species whose ranges lie across the drowned and emergent landscapes of the Sunda Shelf, a molecular echo of ancient river systems that today lie submerged beneath the South China Sea. The result aligns with a growing body of work showing that Pleistocene sea-level swings sculpted the evolutionary dynamics of Southeast Asia’s biodiversity hotspots.

Not everything fell neatly into place. The cladistic relationships between the two flagship genera, Neolissochilus and Tor, remain poorly resolved, and two species emerged as potential misfits: N. benasi and T. remadevii may each constitute distinct lineages outside the core clades of their assigned genera. More striking still, Tor douronensis and T. tambra — treated as taxonomic synonyms in some classifications — occupied inconsistent positions in the mitochondrial tree, a result difficult to reconcile with the idea that they are one and the same fish. Such findings matter far beyond academic bookkeeping: species limits determine which populations receive legal protection, which hatchery stocks can be moved, and how fisheries are managed. The authors argue that only integrated morphological and genetic studies spanning mainland and island Southeast Asian populations can untangle these knots, and they position the new genome as the reference point from which that revision can begin.

The genome arrives at a moment when Sumatran mahseers need it. The species is caught throughout its range, its rivers face hydropower development and land-use change, and conservation programs around Samosir and the Batang Toru tributaries have already relied on DNA barcoding to distinguish N. sumatranus from its mahseer neighbors. A complete mitochondrial genome gives those efforts a firmer backbone: sharper species identification, a template for designing population genetic surveys, and a benchmark against which the many still-unsequenced Tor and Neolissochilus mitogenomes can be assembled and compared. It clarifies, the authors write, the phylogenetic position of N. sumatranus within Torinae and lays a foundation for the systematics of a group whose evolutionary history has for too long been told in fragments. The king of the river, it turns out, was carrying its own biography inside every cell — sixteen thousand five hundred and eighty-four letters long — and scientists have now read it cover to cover.