An important challenge in uncovering evolutionary histories is to reconstruct the past based on evidence from the present. Unfortunately, there is little in the fossil record that sheds light on the origins of hemostasis. Rather, we are left with comparative analyses of morphology, function, and molecular sequences. A central assumption in comparative anatomy and physiology is that two structures that look or act alike are closely related. However, that is not always the case. Structures may look similar simply because they have evolved similar adaptations. This is termed convergent evolution. An example is seen in the case of the horseshoe crab (Tachypleus tridentatus), which contains a clotting system comprised of a protease cascade of three serine protease (SP) zymogens and a clottable protein (coagulogen). However, these components arose independently of the vertebrate clotting proteins. Among other things, they lack γ-carboxylated glutamic acid (GLA) domains, epidermal growth factor (EGF)-like domains, and kringle domains, which promote essential interactions of vertebrate clotting factors. Conversely, divergent evolution describes dissimilar phenotypes that share a common ancestor. The interpretation of molecular phylogenies for gene sequence divergence is similarly wrought with limitations. It is widely believed that the vertebrate genome has undergone two whole-genome duplication (WGD) events in addition to many local duplications, translocations, gene conversions, and gene loss. Thus, it is important to distinguish between orthologous genes (orthologs) and paralogous genes (paralogs). Orthologous genes in different species have evolved directly from a common ancestral gene, whereas paralogs have evolved from a common ancestral gene through duplication and have diverged in sequence and function from the parent gene through mutations and selection or drift. Stated another way, orthology is the relationship where sequence divergence follows speciation, whereas paralogy is defined as that condition where sequence divergence follows gene duplication.

To establish the evolutionary history of the coagulation cascade, we must turn to modern-day animals (especially early diverging vertebrates) and examine how their blood clots and/or determine whether their genome contains sequences for the various clotting factors. Functional studies of nonmammalian vertebrates have routinely employed the same clotting time assays that are used in humans, including whole-blood clotting time, prothrombin (PT) time, and activated partial thromboplastin
time (aPTT). However, the absence of factor-deficient plasma from many of these species has precluded the use of mixing studies to measure individual clotting factor levels. Functional studies of blood clotting in early vertebrates are further hampered by inter- and intraspecies variations in clotting times, small volumes of available blood (e.g., in small fish such as the zebrafish), and the nonuniform plasma composition and body temperatures of poikilotherms.

Protein purification has been used to isolate fibrinogen in different species. However, such a strategy is not feasible for many of the SPs, which circulate at trace levels. With the advent of polymerase chain reaction (PCR), it has been possible to use degenerate primers (based on multiple sequence alignment of available gene sequences) against specific domains of clotting factors to clone the cDNA for SP clotting factors. More recently, searching for DNA sequences (orthologs) has been facilitated by the availability of the whole-genome sequence of the sea squirt (Ciona intestinalis), amphioxus (Branchiostoma floridae), puffer fish (Fugu rubripes), and other higher vertebrates. Complete genome sequence is not available for lamprey or hagfish. However there is a lamprey trace database maintained by the NCBI and EBO. Once DNA sequences are available from multiple species, it is possible to align sequences and construct phylogenic trees, which are diagrams that depict the lines of evolutionary descent of different species, organisms, or genes from a common ancestor. Sequence similarities and phylogenic trees based on aligned sequences provide important information about when clotting factors arose during evolution. In addition to analyzing DNA sequences, it is valuable to examine gene organization. For example, the occurrence and arrangement of catalytic and peripheral domains yields clues about gene duplication and exon shuffling. Similarities in intron-exon structures are characteristic of orthologous genes. The presence of two similar genes (paralogs) on the same chromosome is a clue that they originated from local duplication, rather than genome-wide duplication.

Studies on the evolution of the clotting cascade have focused on the vertebrate-invertebrate divide, namely the chordates. Chordates form a phylum of animals that includes vertebrates and closely related invertebrates. There are three subphyla: urochordates or tunicata (composed of more than 3,000 species and represented by the sea squirt), cephalochordates or lancelets (composed of ˜29 species and represented by amphioxus), and the extant vertebrates including jawless fish or cyclostomes (hagfishes and lampreys) and jawed vertebrates or gnathostomes (composed of ˜53,000 species) (figure 2.1). Urochordates and cephalochordates are referred to as basal chordates, or protochordates. Among other features, they have a notochord that persists throughout life but do not have backbones. Traditionally, cephalochordates were considered to represent the closest living relatives of vertebrates, with tunicates representing the earliest chordate lineage. However, recent evidence indicates that urochordates are more closely related to vertebrates than are cephalochordates.^1,2 Jawless fish are the earliest appearing vertebrates. The two extant genera are hagfish and lamprey. Although morphologic and molecular phylogenic studies have yielded conflicting results, it is believed that hagfish and lamprey are not monophyletic, but rather represent separate divergences, with hagfish being more primitive.³ The jawed vertebrates include cartilaginous fish (also termed elasmobranchs), the bony fishes (also termed teleosts), amphibians, reptiles, birds, and mammals.

It is widely accepted that the vertebrate genome has undergone multiple rounds of WGD (reviewed in Ref.⁴) (figure 2.1). Although the timing and the genomic scale of such duplications remain poorly understood, the favored hypothesis is that the first round occurred in a common ancestor of jawed and jawless vertebrates (˜630 to 565 million years ago), with the second round occurring in a common ancestor of the jawed vertebrates (˜565 to 525 million years ago). An additional WGD may have occurred in an ancestor of the majority of ray-finned fish some 350 to 320 million years ago. The idea of genome duplication has been strengthened by the observations that model vertebrates often have multiple duplicate genes (paralogs) corresponding to a set of invertebrate orthologs.^5,6 When a gene is duplicated, one of the copies is redundant and escapes purifying selection. Natural selection cannot operate on idle proteins, but only on those that are being used. Thus, most gene duplications from WGDs are rapidly lost (this accounts for why the genome of sea squirt is smaller [330 Mb] than that of amphioxus [540 Mb]). Rarely, the mutations may provide the gene with a new function. Indeed, duplicate functional modules arising from WGDs are likely to have played a critical role in enhancing morphologic and functional complexity during vertebrate evolution.^4,7,8

Although it is widely accepted that large regional or WGD events contributed to the structure of vertebrate genomes,^9,10,11 local duplication and translocation have certainly contributed to the evolution of multigene families on separate chromosomal regions. Domain shuffling (insertion) is another important mechanism for the evolution of novel genetic material. Since the domains are often correlated with exon boundaries, exon shuffling is considered to be the driving force behind domain shuffling.¹² In this way, different domains of modular proteins may have evolved in unrelated protein families, and then been subsequently fused to previously existing domains.¹³ Other mechanisms of domain shuffling include simple fusion of genes and recruitment of mobile elements.