Search

Mutations at CpG dinucleotides were delineated in the factor IX gene of 38 hemophilia B patients. When transitions at CpG were considered with those previously reported by us and those compiled in the factor IX mutation database, the following patterns emerged. Many CpG sites were mutated with high frequency, while two CpG sites were infrequently mutated (R29-->Q and R116-->TGA). Of the 6 possible nonsense mutations and the 14 missense mutations that would produce a nonconservative change at conserved amino acids, all have been observed to cause hemophilia B except A-10-->T and R338-->Q. By contrast, none of the 6 missense changes at nonconserved amino acids have been observed to cause hemophilia B. At those CpG sites that are frequently mutated, the rate of transitions is estimated to be 20-fold higher than transitions at non-CpG sites. Point mutations in close proximity to CpG dinucleotides did not seem elevated.

Some missense changes are compatible with normal protein function while others compromise essential aspects of protein maturation, specific activity, or stability. For those missense changes that alter function in the intact organism, how likely is it for the mutated protein to retain appreciable residual activity? By genetic analysis of patients with hemophilia B of known severity, this question can be addressed for missense mutations that reduce factor IX activity by fourfold or more below the average. We estimate that missense changes cause only 59% of moderate and severe disease, but these mutations are almost always (95%) of independent origin. In contrast, missense mutations are found in virtually all (97%) families with mild disease, but only a minority of these (41%) are of independent origin. From the aggregate data, we estimate that most (71%) of the independent deleterious missense mutations cause at least a 20-fold decrease in factor IX activity.

By direct genomic sequencing, we have delineated the causative mutation in 64 families of European decent with hemophilia B. Six (9%) had a C----T transition at base 31008, which substitutes methionine for threonine 296 (T296----M) in the catalytic domain of factor IX. Five of the patients had the same haplotype (frequency of 16% in the northern European population). These individuals are of Amish/German descent and they are likely to share a common ancestor. The sixth patient had a different haplotype, which indicates that his mutation had an independent origin. The data highlight the importance of clinical criteria for the classification of hemophilia B. All six patients had clinically mild disease and their factor IX coagulant activities were in the range of 3%-6% when tested simultaneously in one laboratory, yet the factor IX activities provided with patient records varied 40-fold. Due to the high frequency of this mutation, we have utilized the technique of polymerase chain reaction amplification of specific alleles (PASA) to perform rapid and inexpensive carrier diagnoses in the families with this mutation. This is of particular importance for the Amish since the mutation should account for much of, if not all, the mild hemophilia B that is commonly found in this population.

Direct sequencing of the regions of the factor IX gene of likely functional significance was performed in four patients with severe hemophilia B. In two of the individuals, a transition at the dinucleotide CpG caused a nonsense mutation at arginine 333. In the other two individuals, a transition at CpG caused a nonsense mutation at arginine 29. Since these patients are all unrelated, as shown by differing alleles of the TaqI polymorphism in intron four or extensive nonoverlapping pedigrees, the mutations arose independently. In addition, the origin of one arginine 333 mutation in one family has been traced to the germline of the maternal grandfather. The frequent occurrence of mutations at arginine codons that contain the sequence CGN can be explained by the dramatic elevation of transitions at CpG. As a result, approximately one in four individuals with hemophilia B is expected to have a mutation at arginine and nonsense mutations at one of six arginine residues should be common causes of severe hemophilia.