The skeleton, teeth, and otoconia are normally the only mineralized tissues or organs in the human body. We describe physiological biomineralization in collagenous matrices as well as a more derived noncollagenous matrix. The origin of the collagenous matrices used in mineralized skeletal tissues can be traced to a soft tissue in early Metazoa. In early vertebrates, a genetic system coding for ancient soft collagenous tissue was co-opted for biomineralization using redundant genes resulting from whole genome duplication. However, genes more specific to mineralized tissues arose subsequent to the genome duplication by genomically local tandem duplication. These new genes are the basis for a novel genetic system for various mineralized tissues in skeleton and teeth. In addition, any tissue can be abnormally mineralized, and many pathologies of mineralization in humans are known.

Caseins are among cardinal proteins that evolved in the lineage leading to mammals. In milk, caseins and calcium phosphate (CaP) form a huge complex called casein micelle. By forming the micelle, milk maintains high CaP concentrations, which help altricial mammalian neonates to grow bone and teeth. Two types of caseins are known. Ca-sensitive caseins (α_s- and β-caseins) bind Ca but precipitate at high Ca concentrations, whereas Ca-insensitive casein (κ-casein) does not usually interact with Ca but instead stabilizes the micelle. Thus, it is thought that these two types of caseins are both necessary for stable micelle formation. Both types of caseins show high substitution rates, which make it difficult to elucidate the evolution of caseins. Yet, recent studies have revealed that all casein genes belong to the secretory calcium-binding phosphoprotein (SCPP) gene family that arose by gene duplication. In the present study, we investigated exon–intron structures and phylogenetic distributions of casein and other SCPP genes, particularly the odontogenic ameloblast–associated (ODAM) gene, the SCPP-Pro-Gln-rich 1 (SCPPPQ1) gene, and the follicular dendritic cell secreted peptide (FDCSP) gene. The results suggest that contemporary Ca-sensitive casein genes arose from a putative common ancestor, which we refer to as CSN1/2. The six putative exons comprising CSN1/2 are all found in SCPPPQ1, although ODAM also shares four of these exons. By contrast, the five exons of the Ca-insensitive casein gene are all reminiscent of FDCSP. The phylogenetic distribution of these genes suggests that both SCPPPQ1 and FDCSP arose from ODAM. We thus argue that all casein genes evolved from ODAM via two different pathways; Ca-sensitive casein genes likely originated directly from SCPPPQ1, whereas the Ca-insensitive casein genes directly differentiated from FDCSP. Further, expression of ODAM, SCPPPQ1, and FDCSP was detected in dental tissues, supporting the idea that both types of caseins evolved as Ca-binding proteins. Based on these findings, we propose two alternative hypotheses for micelle formation in primitive milk. The conserved biochemical characteristics in caseins and their immediate ancestors also suggest that many slight genetic modifications have created modern caseins, proteins vital to the sustained success of mammals.

Diverse hard tissues constituted a tooth-like skeletal element in extinct jawless vertebrates. Today, similar tissues are found in our teeth. These tissues mineralize in the extracellular matrix and involve various macromolecules. Among these molecules are secretory calcium-binding phosphoproteins (SCPPs) coded by genes that arose by duplication. Although the repertoire of SCPPs may vary in different lineages, some SCPPs are unusually acidic and are thought to participate in the mineralization of a collagenous matrix, principally either bone or dentin. Other SCPPs are rich in Pro and Gln (P/Q) and are employed to form the tooth surface. In tetrapods, the tooth surface is usually covered with enamel which develops in a matrix comprised of P/Q-rich SCPPs. By contrast, the tooth surface tissue in teleosts is called enameloid and it forms in a dentin-like collagenous matrix. Despite the difference in their matrix, both enamel and enameloid mature into hypermineralized inorganic tissues. Notably, some P/Q-rich SCPP genes are primarily expressed at this stage and their proteins localize between the tooth surface and overlying dental epithelium. Moreover, an orthologous gene is used for maturation of these 2 different tissues. These findings suggest distinct roles of acidic and P/Q-rich SCPPs during the evolution of hard tissues. Acidic SCPPs initially regulated the mineralization of bone, dentin, or a similar ancient collagenous tissue through interaction with calcium ions. P/Q-rich SCPPs arose next and originally assembled a structure or a space that facilitated the hypermineralization of dentin or a dentin-like tissue. Subsequently, some P/Q-rich SCPPs were coopted for the mineralizing enamel matrix. More recently, however, many SCPP genes were lost in toothless birds and mammals. Thus, it appears that, in vertebrates, the phenotypic complexity of hard tissues correlates with gain and loss of SCPP genes.

The bowfin (Amia calva) is a ray-finned fish that possesses a unique suite of ancestral and derived phenotypes, which are key to understanding vertebrate evolution. The phylogenetic position of bowfin as a representative of neopterygian fishes, its archetypical body plan and its unduplicated and slowly evolving genome make bowfin a central species for the genomic exploration of ray-finned fishes. Here we present a chromosome-level genome assembly for bowfin that enables gene-order analyses, settling long-debated neopterygian phylogenetic relationships. We examine chromatin accessibility and gene expression through bowfin development to investigate the evolution of immune, scale, respiratory and fin skeletal systems and identify hundreds of gene-regulatory loci conserved across vertebrates. These resources connect developmental evolution among bony fishes, further highlighting the bowfin’s importance for illuminating vertebrate biology and diversity in the genomic era.

The mineralized skeleton is a critical innovation that evolved early in vertebrate history. The tissues found in dermal skeletons of ancient vertebrates are similar to the dental tissues of modern vertebrates; both consist of a highly mineralized surface hard tissue, enamel or enameloid, more resilient body dentin, and basal bone. Many proteins regulating mineralization of these tissues are evolutionarily related and form the secretory calcium-binding phosphoprotein (SCPP) family. We hypothesize here the duplication histories of SCPP genes and their common ancestors, SPARC and SPARCL1. At around the same time that Paleozoic jawless vertebrates first evolved mineralized skeleton, SPARCL1 arose from SPARC by whole genome duplication. Then both before and after the split of ray-finned fish and lobe-finned fish, tandem gene duplication created two types of SCPP genes, each residing on the opposite side of SPARCL1. One type was subsequently used in surface tissue and the other in body tissue. In tetrapods, these two types of SCPP genes were separated by intrachromosomal rearrangement. While new SCPP genes arose by duplication, some old genes were eliminated from the genome. As a consequence, phenogenetic drift occurred: while mineralized skeleton is maintained by natural selection, the underlying genetic basis has changed.