The research team, led by Sawar Khan and Zanxian Xia of Central South University together with colleagues in China and Pakistan, focused on four families of tardigrade stress-response proteins: SAHS (secretory abundant heat-soluble), CAHS (cytoplasmic abundant heat-soluble), MAHS (mitochondrial abundant heat-soluble), and Dsup (damage suppressor). These proteins, many of them intrinsically disordered, shield cells from dehydration, radiation, and oxidative damage. Until recently, most knowledge about them came from just two well-studied species, Ramazzottius varieornatus and Hypsibius exemplaris. The new study widened the lens to four species, adding Paramacrobiotus metropolitanus and the recently characterized Hypsibius henanensis, using the only four tardigrade genomes with high-quality, well-annotated assemblies available in public databases.

The team’s approach combined several computational techniques. They compared gene structures, mapping exon and intron lengths across species; clustered proteins into orthologous groups using the OrthoMCL algorithm as implemented in OrthoVenn3; built maximum-likelihood phylogenetic trees with IQ-TREE2; identified conserved sequence motifs with the MEME Suite; and predicted three-dimensional protein structures with the AlphaFold Server, anchored by the experimentally solved crystal structure of SAHS1 from R. varieornatus. The result is a multi-layered portrait of how these protein families have diversified, and how their most important features have been preserved, across roughly 100 million years of tardigrade evolution.

The gene copy-number analysis revealed clear species-specific patterns. For the SAHS family, R. varieornatus, H. exemplaris, and P. metropolitanus each carried 13 copies, while H. henanensis had 10. The CAHS family showed even more variation: R. varieornatus had 17 genes, H. exemplaris and P. metropolitanus had 16 each, and H. henanensis had 11. In sharp contrast, the MAHS family was strictly conserved, with exactly one copy in every species examined, a pattern the authors interpret as evidence of strong functional constraint, suggesting that this mitochondrial protective protein performs an essential, possibly irreplaceable role. The Dsup family told yet another story: homologs were found in three species but were conspicuously absent from P. metropolitanus, raising the possibility that this lineage relies on alternative molecular strategies to protect its genome.

Orthologous clustering added nuance to the picture. Of 49 SAHS proteins analyzed, most fell into seven clusters comprising five conserved orthologous groups and two paralogous groups, but nearly a third, 32.65 percent, were singletons found in only one species. The CAHS family showed a similar pattern, with 60 proteins distributed into 13 clusters and 16 singletons, about 26.67 percent of the total. The authors caution that such high singleton counts could reflect lineage-specific adaptation, but they could also stem from technical limits in detecting highly divergent homologs or from gene loss under neutral evolution. Phylogenetic gene trees for both families largely followed the species tree, which the team reconstructed from five conserved single-copy orthologous proteins using the nematode Caenorhabditis elegans as an outgroup, though some individual loci showed discordant placements consistent with gene duplication and lineage-specific evolution.

Perhaps the most consequential discovery was a new member of the Dsup family. Before this study, only two Dsup proteins were known: the original identified in R. varieornatus in 2016, which made headlines when it was shown to boost radiation tolerance in human cultured cells, and a homolog in H. exemplaris. The team identified a third, H.Henanensis.Chr5.66, a 324-amino-acid protein in H. henanensis that they conclude is a likely Dsup ortholog. Pairwise comparisons showed it shares greater similarity with the H. exemplaris protein, a similarity index of 0.60, than with the R. varieornatus protein, at 0.32. Computational predictions placed all three Dsup proteins in the nucleus, with a putative nuclear localization signal near the C-terminus, consistent with their established role in shielding DNA.

What makes the Dsup findings especially intriguing is the contrast between sequence and structure. The three proteins are quite divergent in primary sequence, a pattern the authors attribute to weak selective pressure on an intrinsically disordered protein, similar to what has been observed in late embryogenesis abundant, or LEA, desiccation protectants. Yet structural modeling revealed a conserved α-helical core shared by all three, with pairwise structural superposition yielding root-mean-square deviation values of roughly 3.8 to 4.8 angstroms, indicating moderate fold conservation. All three proteins also displayed a broad central hydrophobic region and a positively charged C-terminal domain, an electrostatic profile well suited to binding nucleosomal DNA. The authors propose that Dsup operates as a flexible molecular shield, stabilizing DNA through electrostatic interactions rather than through a rigid lock-and-key architecture.

The convergence evidence extends to the SAHS and CAHS families as well. Using the ConDor webserver, which detects statistically significant excesses of independent amino-acid substitutions across protein alignments, the team identified candidate convergent sites in both families. In SAHS1, residues E102, H138, and K165 showed significant excesses of independent emergence events, with K165 exhibiting ten independent emergences to lysine and a false discovery rate of 0.00295. In CAHS1, sites S88, E167, and A201 were similarly flagged, with A201 showing seven independent emergences, predominantly to histidine, at an FDR of 0.00215. Critically, when projected onto predicted structural models, all six candidate sites localized within the conserved sequence motifs identified by the team’s motif-discovery analysis, lending biological plausibility to the idea that these recurrent substitutions are functionally relevant rather than stochastic noise.

The authors are careful about the limits of their evidence. All functional inferences in the study are computational and therefore provisional; bioinformatic signals can generate hypotheses but cannot establish molecular mechanism or physiological effect on their own. Computational models also cannot capture cellular context, expression dynamics, post-translational modifications, or the biophysical behaviors that determine protein function in living organisms. The team explicitly frames the reported sites and motifs as prioritized candidates for experimental follow-up, noting that functional assays and comparative evolutionary analyses are needed to definitively establish convergent evolution and to determine whether the recurrent substitutions they identified are truly adaptive.

Even with those caveats, the implications are far-reaching. Tardigrade stress-response proteins are already being explored for applications spanning medicine, agriculture, and space exploration: Dsup has been engineered into human cells, rice, and tobacco pollen to improve radiation tolerance, while CAHS and MAHS proteins show promise for biopreservation, pharmaceutical stabilization, and vaccine storage. By mapping the full diversity of these protein families across tardigrade lineages and pinpointing the residues and structural features under convergent selection, the new study provides a comparative framework that could guide the design of next-generation stress-protective molecules. The water bears, it seems, have not one but several molecular tricks for surviving the unsurvivable, and evolution appears to have found them more than once.

Subject of Research: Convergent evolution of stress-response protein families in tardigrades

Article Title: Integrative bioinformatics analysis of potential convergent evolution in tardigrade stress-response proteins

Article References: Khan, S., Nisar, A., Qadeer, A., Liu, S., Mehmood, S. A., Zhang, L., & Xia, Z. (2026). Integrative bioinformatics analysis of potential convergent evolution in tardigrade stress-response proteins. Stress Biology, 6(1), Article 21. https://doi.org/10.1007/s44154-026-00286-5

Image Credits: AI Generated

DOI: 10.1007/s44154-026-00286-5

Keywords: tardigrades, Dsup, SAHS, CAHS, MAHS, intrinsically disordered proteins, convergent evolution, stress tolerance, bioinformatics, anhydrobiosis, DNA protection, extremophiles