The cDNA for human gamma-glutamyl carboxylase, which accomplishes the post-translational modification required for the activity of all of the vitamin K-dependent proteins, was cloned. The enzyme is a 758-residue integral membrane protein and appears to have three transmembrane domains near its amino terminus. The hydrophilic COOH-terminal half of the carboxylase has 19.3 percent identity with soybean seed lipoxygenase. Expression of the cloned cDNA resulted in an increase in carboxylase activity in microsomes of transfected cells compared to mock-transfected cells.

We used recombinant DNA techniques to map a panel of six monoclonal antibodies (MoAbs) to regions of the human factor IX molecule. A-2 maps to 17 amino acids at the amino terminus of the heavy chain of IXa; 2D5, an inhibitor of clotting, is defined to 36 amino acids of the first EGF- like domain of human factor IX. A-4, A-5, C10D, and FXC008 all map to a region of the heavy chain containing amino acids 180 through 310, suggesting an immunodominant site. FXC008 has been reported to interfere with binding of factor IXa to factor VIII:Ca.

As an approach to the study of structure-function relationships in the normal and defective forms of human coagulation factor IX, we have begun to develop a series of monoclonal antibodies against specific sites on the protein. Zymogen and activated forms of normal factor IX were used initially as antigen for the preparation of monoclonal antibodies. Recombinant phage were prepared by cloning small (50- to 500-nucleotide) random DNA fragments from the coding region of a factor IX cDNA clone into the expression vector lambda gt11. Immunological screening of these recombinants with mixtures of monoclonal antibodies identified several immunoreactive phage. Further analysis showed that the monoclonal antibody designated IX-30 was generating the positive signals at a frequency of approximately 1/2,500 recombinants. Subcloning and sequence analysis of the inserted DNA in the immunoreactive phage revealed overlapping in-frame insertions, from which it could be inferred that the site in factor IX recognized by IX- 30 is confined to residues 111 through 132 in the light chain. Similar mapping with other monoclonal antibodies should provide additional probes for the protein structure of human factor IX.

In order to map the regions of human factor IX recognized by monoclonal antibodies we have inserted random fragments of the coding region of the cDNA for human factor IX into the lambda phage expression vector, lambda gtll. The resultant recombinant phage were screened with monoclonal antibodies, the immunoreactive phage were isolated, and the DNA of the inserted fragment was sequenced to determine which amino acids were immunoreactive. This data, coupled with data obtained from the use of specific fragments of human factor IX expressed in E. coli from a T7 phage promoter, has allowed us to map the location of several epitopes on the surface of the factor IX molecule. In those cases where the antibodies are specific for human factor IX, additional narrowing of the epitope is possible by comparing the amino acid sequence of human factor IX to the bovine molecule. Six monoclonal antibodies from 3 different laboratories have been mapped. The immunodominant epitopes appear to be the amino terminus of the activation peptide, the amino terminus of the heavy chain and the epidermal growth factor domains.

with a monoclonal antibody (A-1) detect a prevalent dimorphism in plasma coagulation factor IX. The antibody was shown to react with a dimorphic segment of the normal factor IX sequence as follows. First, A-1 bound to isolated activation peptide (residues 146 through 180) prepared from activated factor IX from a normal plasma pool. Second, binding of recombinant factor IXs with A-1 or factor IX from normal individuals was strong only when they had Threonine (Thr) at position 148; factor IXs with the Alanine (Ala) allele at that position were far less reactive. Third, immunoblot reactivity of Escherichia coli fusion proteins containing known segments of the factor IX sequence restricted the epitope to residues 147 through 153. In 120 hemophilia B pedigrees, the Ala immunoassay allele frequency was 0.19 and did not differ from the Ala frequency in normal males. In 22 of 49 families, immunoassay testing was informative for classification of obligate or possible carriers. In one large family, 4 obligate carriers were heterozygous for the dimorphism and 3 of their 7 daughters were classified as carriers. In other families, when the affected member had less than 1 nmol/L factor IX antigen, heterozygosity for Thr/Ala alleles excluded the carrier state even when DNA studies were not informative. Strong linkage disequilibrium of Thr/Ala alleles with the common TaqI DNA polymorphism was found. Nineteen of 75 normal and hemophilic factor IX genes had the 1.3- kilobase (kb) fragment and coded for the Ala allele; the rest had the 1.8-kb fragment and coded for Thr. In selected families, the A-1 immunoassay is an inexpensive and rapid method to confirm and supplement restriction fragment length polymorphism analyses of DNA for carrier testing.