Hemorrhagic disorder due to an isoniazid-associated acquired factor XIII inhibitor in a patient with Waldenström's macroglobulinemia.

A case is described of a 75-year-old woman with a history of pulmonary tuberculosis and Waldenström's macroglobulinemia who developed an inhibitor of coagulation factor XIII while taking isoniazid. The patient presented with a subcutaneous hematoma of the abdominal wall that extended from the xiphoid process to the symphysis pubis and measured 20 cm in diameter. Results of routine coagulation studies were normal with the exception of an increased solubility of the patient's plasma clot in 5M urea consistent with a deficiency of factor XIII activity. Persistence of the deficiency following a 1:2 dilution of the patient's plasma in normal plasma indicated the presence of an inhibitor. A sample of the patient's plasma was depleted of IgG by streptococcal protein G adsorption. The IgG-depleted plasma did not inhibit factor XIII activity, indicating that the inhibitory activity was not attributable to the underlying IgM paraprotein. The patient's purified IgG, on the other hand, inhibited factor XIII activity and the inhibitory activity could be neutralized by anti-IgG antibody. The patient's IgG also inhibited factor XIII-mediated incorporation of fluorescent monodansylcadaverine into casein. Binding of the patient's IgG to factor XIII concentrate was demonstrated by enzyme-linked immunosorbent assay and the IgG that bound to the factor XIII was demonstrated to be polyclonal. Isoniazid was discontinued after the patient was admitted to the hospital. Cryoprecipitate infusion controlled bleeding and reduced the inhibitor titer by 50%. Treatment with cyclophosphamide and prednisone, followed by extracorporeal immunoadsorption over a staphylococcal protein A column, did not reduce the inhibitor titer further. Plasma exchange therapy reduced the inhibitor titer to undetectable levels but failed to restore factor XIII activity. Infusions of factor XIII concentrate reproducibly restored factor XIII activity and were not associated with an anamnestic rise in the inhibitor titer. This represents the seventh reported case of an acquired inhibitor to factor XIII associated with the ingestion of isoniazid.

New methods for the study of folate coenzymes: endogenous polyglutamate patterns of subcellular hepatocyte fractions and of regenerating rat liver.

A procedure has been developed that permits the quantitation of three different pools of one-carbon-substituted folates with the simultaneous determination of their corresponding poly-gamma-glutamyl chain lengths. Pool 1 is made up of N5,N10-methylene-tetrahydrofolates and unsubstituted tetrahydro and dihydrofolates. Pool 2 is made up solely of N5-methyl-tetrahydrofolates and pool 3 includes the folates with one-carbon-substituents at the formyl oxidation level: N5,N10-methenyl-tetrahydrofolates, N10- and N5-formyl-tetrahydrofolates, and N5-formimino-tetrahydrofolates. Conditions have been found that permit the selective cleavage of the C9-N10 bond of the folates of pools 1, 1 + 2, and 1 + 2 + 3. The cleaved folates are quantitated as the Bratton-Marshall azo-dyes of the p-amino-benzoyl-poly-gamma-glutamates (AzoGlun) after converting the uncleaved pools to Bratton-Marshall negative products. Determination of the poly-gamma-glutamyl chain length of the cleaved pools is carried out by high pressure liquid chromatography of the AzoGlun comparing the elution position of each peak with that of authentic synthetic markers. Quantitation of the individual peaks is obtained by integration and by reference to an internal synthetic standard (usually AzoGlu3). To increase the sensitivity of the procedure and make it applicable to plasma or needle biopsy specimens we have developed a synthesis for [3H]-N'-naphthylethylenediamine. This compound is the one coupled to the diazotized p-aminobenzoyl-poly-gamma-glutamates to form the AzoGlun in the Bratton-Marshall test. The new methods do not require labeling the animal's folates by the prior injection of radioactive folic acid thus permitting the study of endogenous folate patterns. Applied to rat liver these methods have shown that pool 3 accounts for 44.3 +/- 7.7%, pool 2 for 36.7 +/- 4.7%, and pool 1 for 18.5 +/- 5.8% of the total folates. The pentaglutamates made up about 52% and the hexaglutamates about 42% of the folates. Hepta and tetraglutamates account for approximately 2.8 and 3.5% respectively. The cytosolic folates are predominantly hexaglutamates whereas pentaglutamates predominate in the heavy mitochondrial fraction. Profound changes in the relative proportions of the polyglutamates were observed 24 h after partial hepatectomy. There are marked increases in hepta and hexaglutamates and decreases in penta and tetraglutamates. These results support the hypothesis that changes in polyglutamate chain length are related to the regulation of one-carbon metabolism.

The proteoglycan decorin binds C1q and inhibits the activity of the C1 complex.

Decorin, a small collagen-binding dermatan sulfate proteoglycan, is widely distributed as a component of extracellular matrices. Using a solid phase binding assay, we showed that decorin bound C1q at physiologic pH and ionic strength. The interaction did not require divalent cations and was time and temperature dependent reaching equilibrium in 4 h at 37 degrees C. Binding was specific and saturable with an apparent dissociation constant of 7.6 x 10(-9) M. Decorin was shown to bind pepsin-derived fragments containing the collagenous domain of C1q and collagenase-derived fragments containing the globular domain of C1q. Because these fragments share a short sequence of amino acids, this finding suggests that decorin binds to a region of C1q located near the junction of the two domains. Competition studies using purified preparations of the decorin core protein and the glycosaminoglycan chains showed that only the former inhibited binding of decorin to C1q indicating that the interaction is mediated by the decorin core protein. Decorin was shown to inhibit the hemolytic activity of purified C1 as well as C1 in normal human serum. Approximately 50% inhibition was observed at a decorin concentration of 2 micrograms/ml. Inhibition was not observed if C1 was bound to Ag-complexed antibody. Furthermore, neither the core protein nor the glycosaminoglycan chain of decorin inhibited C1, indicating that the intact proteoglycan is necessary for functional activity.

Molecular bases for inherited human complement component C6 deficiency in two unrelated individuals.

Deficiency of the sixth component of complement (C6D) is frequently associated with recurrent neisserial infections, especially meningitis caused by Neisseria meningitidis. We here report the molecular bases of C6D in two unrelated subjects, one African American (case 1) and the other Japanese (case 2). Screening all 17 exons of the C6 gene and their boundaries by exon-specific PCR/single strand conformation polymorphism demonstrated aberrant single stranded DNA fragments in exon 12 of case 1 and exon 2 of case 2. Nucleotide sequencing of the amplified DNA fragments revealed a homozygous single-base deletion (G1936) in exon 12 case 1 and a heterozygous single base deletion (C291/C292/C293/C294) in exon 2 of case 2. Both mutations resulted in frame shifts and premature termination of the C6 polypeptide. Sequence-specific oligonucleotide probe hybridization and direct sequencing of exon 12 amplified from genomic DNA further supported the homozygosity of the mutation in case 1. Case 2 is apparently compound heterozygote, but the putative mutation in the other allele of the C6 gene remains unknown. Both case 1 and case 2 were homozygous for the C6A allotype. These data indicate that at least three distinct mutational events can cause C6D, single nucleotide deletions in exons 2 and 12, and a mutation yet unidentified. Thus, similar to other complement protein deficiencies, the pathogenesis of C6D appears to be heterogeneous.