The human antimouse immunoglobulin response and the anti-idiotypic network have no influence on clinical outcome in patients with minimal residual colorectal cancer treated with monoclonal antibody CO17-1A.

Murine monoclonal antibodies (mAbs), when administered to patients, induce a human antimouse immunoglobulin immune response, especially when multiple infusions are required to obtain therapeutic efficacy. In a randomized Phase II clinical study, 83 patients with colorectal carcinoma of stage Dukes C were treated with the murine IgG2a mAb 17-1A (ab1) after curative surgery. The regimen consisted of a single infusion of 500mg of 17-1A within 2 weeks after surgery, followed by 100mg of mAbs four times every 4 weeks. Sera were taken every 2-3 weeks and screened for human antimouse antibodies (HAMA). HAMA were measured by a capture ELISA and an indirect antihuman immunoglobulin ELISA for the analysis of IgG and IgM isotypes. Anti-idiotypic antibodies (ab2) were detected by an inhibition ELISA, and anti-anti-idiotypic antibodies (ab3), recognizing the original antigen, were determined by flow cytometric analysis. About 20% of patients failed to develop HAMA; in the other patients, antibody titers were initially low after the first two infusions and reached their maximum only after a fifth infusion at 18-20 weeks after surgery. An analysis that differentiated between patients who developed recurrences and those who remained tumor-free did not show any difference in antibody titers between the two groups, neither for total HAMA nor for IgG, IgM, or ab2. The formation of ab3 was analyzed in eight patients and proved to be negative in all of them. HAMA remained detectable up to 2 years after the last treatment. In patients who experienced adverse events associated with therapy, HAMA titers tended to rise earlier; this difference, however, was not statistically significant. Thus, neither a beneficial nor a detrimental effect of HAMA formation could be determined for the clinical response to antibody therapy.

Activation of human complement by mouse and mouse/human chimeric monoclonal antibodies.

The complement (C)-activating capabilities in human serum of 32 mouse and 10 mouse/human chimeric MoAbs of different isotypes, and their fragments, were tested in vitro. Activation of C via the classical pathway (CP) was performed in 1% factor D-deficient serum in gelatin containing Veronal buffer in the presence of calcium and magnesium (GVB++), while activation of the alternative pathway of C (AP) was assessed in 10% C1q-depleted serum in the presence of 5 mM MgCl2 in GVB++. The C-activating ability of MoAbs was expressed relative to the degree of activation of complement by aggregated IgG for the CP and relative to mouse IgG1 for the AP. All of seven mouse IgG2a MoAbs were potent activators of the CP. The results of CP activation by IgG1, IgG2b and IgG3 isotypes were different for individual MoAbs. Only three (two IgG1 and one IgG3) of 32 mouse MoAbs were potent activators of the AP. IgG2a and IgG2b were relatively poor AP activators. There were a few MoAbs which activated both the AP and CP. Of 10 chimeric MoAbs, two IgG1, one IgG2 and one IgG4 were poor or non-activators of the CP. On the other hand, IgG2 and IgG4 were good AP activators. IgG3 was the most potent AP activator. Most of the F(ab')2 fragments were activators of the AP and displayed no activation of the CP. Fc fragments only activated the CP, whereas Fab' did not activate the CP or the AP. These studies suggest that the route of complement activation by class and subclass MoAbs can not always be predicted in advance and based only on their subclass identity.

Evaluation of the 323/A3 monoclonal antibody and the use of technetium-99m-labeled 323/A3 Fab' for the detection of pan adenocarcinoma.

The 323/A3 murine monoclonal antibody, initially described as reactive to breast carcinomas, is found by immunohistological analyses to have broad cross reactivity with adenocarcinomas of diverse histologic origin. The 323/A3 antigen is similar to the tumor-associated 17-1A antigen as revealed by immunoblot and cross-competition cell binding studies. We have investigated the potential use of the 323/A3 monoclonal antibody for tumor imaging as a Fab' molecule labeled with 99mTc. In vitro studies demonstrate that 323/A3 Fab' has high affinity (2-3 x 10(9) M-1) with no significant loss of immunoreactivity compared to the intact IgG. In vivo studies demonstrate that 99mTc 323/A3 Fab' can rapidly detect human breast and colon tumor xenografts growing in athymic nude mice. Distinct breast tumor visualization is observed as early as 1 h post intravenous administration with the 99mTc 323/A3 Fab'. Distinct colon tumor visualization is observed by 3 h (the earliest time point imaged). Tumor-to-blood ratios are higher for 99mTc 323/A3 Fab' than with a 99mTc-labeled nonspecific isotype-matched Fab' antibody. These results suggest that 99mTc 323/A3 Fab' can detect 17-1A antigen and may have potential clinical utility for the rapid diagnostic imaging of adenocarcinomas.

Characterization of a Gla-containing protein from calcified human atherosclerotic plaques.

In this article we describe the isolation of a 4-carboxyglutamic acid (Gla)-containing protein from calcified human atherosclerotic plaques. The protein was extracted from pulverized calcified plaques by demineralization with ethylenediaminetetraacetate and was subsequently purified by anion-exchange fast protein and high-performance liquid chromatography by using ion-exchange and gel-filtration columns. The protein was designated as plaque Gla protein (PGP) and has an apparent mass of 23 kD as estimated from sodium dodecyl sulfate-polyacrylamide gel analysis. By determining the sequence of its first six amino acid residues, the protein was unequivocally demonstrated to be not related to any other known protein. Moreover, no immunological relationship (as tested by Western blot analysis) was found between PGP and other known Gla-containing proteins.

The quantification of gammacarboxyglutamic acid residues in plasma-osteocalcin.

In this paper we describe an assay procedure for determining the amount of gammacarboxyglutamic acid (Gla) residues in serum- or plasma-osteocalcin. The test includes removing by ethanol precipitation the majority of the proteins from the plasma (e.g., the Gla-containing coagulation factors) and the specific extraction of osteocalcin with the aid of immobilized immunopurified antibodies. It is demonstrated that the Gla-content of circulating osteocalcin from normal cows is similar to that of osteocalcin obtained from bone. Hence, the fact that part of the newly synthesized osteocalcin does not bind to the hydroxyapatite matrix in bone cannot be explained by an undercarboxylation of the molecule.

Circulating osteocalcin during oral anticoagulant therapy.

In this paper we present the following observations: 1) In sheep vitamin K-antagonists like phenprocoumon induce a decrease of the serum levels of osteocalcin (bone Gla-protein) and of the affinity of the circulating osteocalcin for hydroxyapatite. 2) In sheep vitamin K counteracts the effect of phenprocoumon on the blood coagulation system, but not that on the osteocalcin production. 3) In human subjects vitamin K-antagonists also lead to decreased levels of serum osteocalcin and a low affinity of the protein for hydroxyapatite. 4) These two variables reached steady-state levels within 24 h after the start of oral anticoagulant treatment and--at continuation of the therapy--they remained low for at least several years.

Isolation and partial characterization of a vitamin K-dependent carboxylase from bovine aortae.

Vitamin K-dependent carboxylase activity has been demonstrated in the crude microsomal fraction of the intima of bovine aortae. The procedure for the isolation of vessel wall carboxylase is a slight modification of the general preparation procedure for tissue microsomes. The highest activity of the non-hepatic enzyme was observed at 25 degrees C and hardly any NADH-dependent vitamin K reductase could be demonstrated. The optimal reaction conditions for both vessel wall as well as liver carboxylase were similar: 0.1 M-NaCl/0.05 M-Tris/HCl, pH 7.4, containing 8 mM-dithiothreitol, 0.4% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulphonic acid (CHAPS), 0.4 mM-vitamin K hydroquinone and 2 M-(NH4)2SO4. Warfarin inhibits the hepatic and non-hepatic carboxylase/reductase enzyme complex more or less to a similar degree. We have measured the apparent Km values for the following substrates: Phe-Leu-Glu-Glu-Leu ('FLEEL'), decarboxylated osteocalcin, decarboxylated fragment 13-29 from descarboxyprothrombin and decarboxylated sperm 4-carboxyglutamic acid-containing (Gla-)protein. The results obtained demonstrated that liver and vessel wall carboxylase may be regarded as isoenzymes with different substrate specificities. The newly discovered enzyme is the first vitamin K-dependent carboxylase which shows an absolute substrate specificity: FLEEL and decarboxylated osteocalcin were good substrates for vessel wall carboxylase, but decarboxylated fragment 13-29 and decarboxylated sperm Gla-protein were not carboxylated at all.

Substrate recognition by vitamin K-dependent carboxylase.

Decarboxylated osteocalcins were prepared and purified from bovine, chicken, human and monkey bones and assayed for their ability to serve as a substrate for vitamin K-dependent carboxylase from bovine liver. Substantial differences were observed, especially between bovine and monkey d-osteocalcin. Since these substrates differ only in their amino acid residues 3 and 4, it seems that these residues play a role in the recognition of a substrate by hepatic carboxylase.

The effect of Gla-containing proteins on the precipitation of insoluble salts.

The precipitation of insoluble salts containing divalent metal ions is inhibited by Gla-containing proteins of various origin. In this paper we demonstrate that: Gla-residues are required for the inhibitory activity; the inhibition is effected by a protein which in vivo is bound to calcified tissue (osteocalcin) as well as by proteins occurring in blood plasma (factor X) and urine (the urinary Gla-protein); The inhibitor concentration required for 50% precipitation-inhibition varied slightly from one salt to the other, but no marked differences were observed between the effects of the various Gla-containing proteins used; Precipitation-inhibition occurred in all phosphates (Be, Ca, Mn and Zn) and in all calcium salts (phosphate, oxalate and carbonate) tested.