Abstract
<title>Abstract</title> <p>Transcriptional regulation in Archaea is described almost entirely through nineteen canonical one-component regulator families (HTH, RHH, and metal-responsive) adapted from bacterial and eukaryotic paradigms. Whether this reference set captures the true diversity of archaeal DNA-binding proteins, or only the fraction resembling other domains of life closely enough to be annotated, has not been assessed at genome scale. Scanning a phylogenetically balanced 331-genome panel (804,755 proteins across all seven major archaeal lineages), we find that only 43.3% of 38,153 DNA-binding-domain (DBD) proteins fall into the canonical families; the remainder carry non-canonical recognizable domains (37.9%) or lack annotation (18.8%). Clustering the non-canonical fraction yields 250 candidate families (nArcR-1 to nArcR-250; 13,237 proteins). We stress these are defined relative to the nineteen-family reference: 171 are non-canonical rather than unknown, and 79 are genuinely uncharacterized—so archaeal regulators are far more diverse than the canonical set implies, not that 250 wholly new families exist. Projecting families across 3,526 genomes (62.5% from uncultivated lineages) via family-specific HMMs recovers correct assignments for 97.3% of held-out members. Fourteen families are broadly distributed; forty-two are lineage-restricted. Genomic-context and operator-motif analysis generates testable regulon hypotheses for 178 families, implicating toxin–antitoxin, CRISPR/defense, signalling, DNA-repair, and transport systems. Structurally, 175 of 184 families adopt recognizable known DBD folds (median TM-score 0.81), with nine matching no reference folds as candidate structural novelties; deeper annotation leaves just nine genuinely unannotated families as priority experimental targets. Motif and RNA-seq validation confirm family-specific operators while tempering direct-regulator claims. Together these results define a large, systematically characterized layer of non-canonical archaeal DNA-binding proteins and a ranked set of experimentally testable hypotheses.</p>