Genetic factors (VKORC1, CYP2C9, EPHX1, and CYP4F2) are predictor variables for warfarin response in very elderly, frail inpatients.

Determining the optimal dose of warfarin for frail elderly patients is a challenging task because of the low dose requirements in such patients, the wide interindividual variability of response, and the associated risk of bleeding. The objective of this study was to address the influence of 13 common variations in eight genes on the maintenance dose of warfarin in a cohort of frail elderly inpatients. For our study, we enrolled 300 Caucasian subjects who were hospital inpatients, with a mean age of 86.7 +/- 6 years. In addition to age, genetic variants of VKORC1, CYP2C9, CYP4F2, and EPHX1 were found to be significant predictor variables for the maintenance dose of warfarin, explaining 26.6% of dose variability. Among 132 patients in whom warfarin therapy was initiated with the same low-dose regimen, we studied the relative influences of genetic and nongenetic factors. The time to first international normalized ratio (INR) > or =2 was influenced by VKORC1 and CYP2C9 genotypes (P = 0.0003 and P = 0.0016, respectively); individuals with multiple variant alleles were at highest risk for overanticoagulation (INR >4) (odds ratio, 12.8; 95% confidence interval, 2.73-60.0). In this special population of frail elderly patients with multiple comorbidities and polypharmacy, we demonstrated the main impact of genetic factors on warfarin response.

Characterization of bitiscetin-2, a second form of bitiscetin from the venom of Bitis arietans : comparison of its binding site with the collagen-binding site on the von Willebrand factor A3-domain.

BACKGROUND: Bitiscetin, a heterodimeric snake venom protein purified from Bitis arietans, binds to the A1 domain of von Willebrand factor (VWF) and induces binding of this domain to platelet glycoprotein (GP) Ib. We previously purified a distinct form of dimeric bitiscetin (herein called bitiscetin-2) that also induces the VWF A1 domain-GPIb interaction, but does not bind to the A1 domain. Instead, it interacts with the collagen-binding A3 domain of VWF. METHODS: In the current study we identify the amino terminal sequence of the bitiscetin-2 as DEGCLPDDSSRT, showing conclusively that the protein is distinct form the originally described bitiscetin. We further studied the interaction of bitiscetin-2 and VWF using DeltaA3 VWF and a series of 33 VWF point mutants previously prepared to map the collagen-binding site. RESULTS: Our results confirm that DeltaA3 VWF, even though containing the A1-domain, is unable to interact with bitiscetin-2. Mutation of VWF-A3 residues Ile975, Asp979, Pro981, Ser1020 and His1023 reduces binding by 80% while mutation of residues Val980, Glu1001 and Arg1021 reduces binding by 30-60%. A 2- to 6-fold increase of binding is caused by mutation of residues Val985, Glu987, and Arg1016. CONCLUSION: Nearly all of these mutations also affect collagen binding showing that the binding sites for bitiscetin-2 and collagen type III in the VWF-A3 domain closely overlap.

Clinicopathological significance of mitochondrial D-Loop mutations in head and neck carcinoma.

Mitochondrial DNA mutations have been reported in several types of tumours, including head and neck squamous cell carcinoma (HNSCC). The noncoding region of the Displacement-Loop (D-Loop) has emerged as a mutational hotspot and we recently found that they were associated with prognosis and response to 5 fluorouracil (5FU) in colon cancers. In order to evaluate the frequence of D-Loop mutations in a large series of HNSCC and establish correlations with clinicopathologic parameters, we sequenced the D-Loop of 109 HNSCC before a treatment by neoadjuvant 5FU-cisplatin-based chemotherapy and surgery. Then, we correlated these mutations with prognosis and response to chemotherapy. A D-Loop mutation was identified in 21% of the tumors, the majority of them were located in a C-tract (D310). The prevalence of D310 mutations increased significantly with the number of cytosines in the matched normal tissue sequence (P=0.02). Hypopharyngeal cancer was significantly more frequent (P=0.03) and tobacco consumption more important (P=0.01) in the group of patients with D-Loop mutation. The presence of D-Loop mutation was not associated with prognosis or with response to neoadjuvant chemotherapy. These results suggest that D-Loop mutations should be considered as a cancer biomarker that may be useful for the early detection of HNSCC in individuals at risk of this cancer.

Generation of von Willebrand factor epitope libraries expressed in E. coli.

The von Willebrand factor (VWF) subunit is composed of several domains, often coinciding with structural regions, characterized through their specific interaction with a ligand. Since several monoclonal antibodies (MoAbs) have been shown to functionally interfere with one of the specific interactions, we have created libraries of bacterial clones expressing peptidic sequences of VWF to map antibodies directed against this protein. Randomly cleaved fragments of VWF cDNA have been cloned in a plasmid designed for the expression of small peptides as part of larger fusion proteins. The NovaTope system is a useful procedure for protein analysis, allowing screening of epitopes composed of contiguous amino acid residues. To map MoAbs with conformational discontinuous epitopes displayed on small as well as large peptidic domains, this technique had to be widely modified to obtain two VWF peptide libraries expressing two ranges of peptide length (15-70 and 100-300 amino acids). Screening with six MoAbs with an epitope in a known region was performed to control both libraries. Four MoAbs were mapped through the characterization of overlapping sequences for 5-10 different positively expressed clones respectively. Two of these mapped MoAbs had no known inhibitory effect and bind reduced VWF only. The fact that the two other MoAbs mapped VWF functional interactions with ligands, platelet GPIIb/IIIa and Factor VIII, respectively, demonstrate that our libraries are valuable tools to determine conformational epitopes.

Specific von Willebrand factor-cleaving protease in thrombotic microangiopathies: a study of 111 cases.

Retrospective studies of patients with thrombotic microangiopathies (TMAs) have shown that a deficient activity of von Willebrand factor (vWF)-cleaving protease is involved in thrombotic thrombocytopenic purpura (TTP) but not in the hemolytic-uremic syndrome (HUS). To further analyze the relevance of this enzymatic activity in TMA diagnosis, a 20-month multicenter study of vWF-cleaving protease activity was conducted in adult patients prospectively enrolled in the acute phase of TMA. Patients with sporadic (n = 85), intermittent (n = 21), or familial recurrent (n = 5) forms of TMA (66 manifesting as TTP and 45 as HUS) were included. TMA was either idiopathic (n = 42) or secondary to an identified clinical context (n = 69). vWF-cleaving protease activity was normal in 46 cases (7 TTP and 39 HUS) and decreased in 65 cases (59 TTP and 6 HUS). A protease inhibitor was detected in 31 cases and was observed only in patients manifesting TTP with a total absence of protease activity. Among the 111 patients, mean vWF antigen levels were increased and the multimeric distribution of vWF was very heterogeneous, showing either a defect of the high-molecular-weight forms (n = 40), a normal pattern (n = 21), or the presence of unusually large multimers (n = 50). Statistical analysis showed that vWF-protease deficiency was associated with the severity of thrombocytopenia (P <.01). This study emphasizes that vWF-cleaving protease deficiency specifically concerns a subgroup of TMA corresponding to the TTP entity.

Mapping and functional studies of two alloantibodies developed in patients with type 3 von Willebrand disease.

Inhibitors against von Willebrand factor (vWF) developed in two unrelated multitransfused patients (patients 1 and 2) with severe (type 3) von Willebrand disease (vWD) were analyzed. Both inhibitors were identified as antibodies of the IgG class by ELISA using immobilized purified vWF and either serum or purified Ig from the patients. Typing, mapping and functional studies of both antibodies revealed significantly distinct properties. Patient 1 antibody contained all subclasses of IgG (1, 2, 3 and 4) whereas antibody from patient 2 was a mixture of only IgG1 and 4. By ELISA using a series of immobilized purified proteolytic fragments of vWF, patient 1 antibody mainly bound to fragment SpIII and, to a lower extent, to fragments SpII and SpI; it poorly bound to P34 and the 39/34 kDa fragment. In contrast, patient 2 antibody only bound to fragments corresponding to the N-terminal portion of vWF but failed to bind to SpII. Functional studies were performed by testing the capacity of each antibody to inhibit vWF binding to its various ligands. Both antibodies blocked vWF binding to Factor VIII (FVIII), fibrillar type III collagen, bitiscetin and the subsequent induced binding to GPIb. Patient 1 antibody also blocked vWF binding to platelet GPIb when induced by ristocetin. However it failed to block vWF binding to GPIb when induced by botrocetin as well as the binding of botrocetin itself to vWF. Our data thus suggest that this inhibitor does not recognize the GPIb-binding site on vWF but the sites of vWF involved in its interaction with ristocetin. In contrast, we observed that patient 2 antibody blocked vWF binding to platelet GPIb induced by either agonist as well as vWF binding to botrocetin. Finally, the effect of the antibodies was tested on vWF binding to GPIIb/IIIa. As expected from the mapping experiments, only IgG from patient 1 blocked the interaction while IgG from patient 2 had no effect. In conclusion, we have shown that two multitransfused patients with type 3 vWD have developed alloantibodies with similar properties to those of polyclonal antibodies but with distinct effects on the functions of vWF.

p53 alterations predict tumor response to neoadjuvant chemotherapy in head and neck squamous cell carcinoma: a prospective series.

PURPOSE: The tumor suppressor gene p53 plays a crucial role in cell cycle control and apoptosis in response to DNA damages. p53 gene mutations and allelic losses at 17p are one of the most common genetic alterations in primary head and neck squamous cell carcinoma (HNSCC). Alterations of the p53 gene have been shown to contribute to carcinogenesis and drug resistance. PATIENTS AND METHODS: In this prospective series, patients with HNSCC were treated with cisplatin-fluorouracil neoadjuvant chemotherapy. p53 status was characterized in 106 patients with HNSCC (p53 mutations, allelic losses at p53 locus, and plasma anti-p53 antibodies) to determine the existence of a relationship between p53 gene status and response to neoadjuvant chemotherapy. RESULTS: Exons 4 to 9 of the p53 gene were analyzed, and mutations were found in 72 of 106 patients with HNSCC. p53 mutations were associated with loss of heterozygosity at chromosome 17p (P <.001). The prevalence of p53-mutated tumors was higher in the group of patients with nonresponse to neoadjuvant chemotherapy than in the group of responders (81% v 61%, respectively; P <.04). When compiling p53 mutations and anti-p53 antibodies in plasma, the correlation between p53 status and response to chemotherapy was significant (87% v 57%, respectively; P =.003). A multivariate analysis showed that p53 status is an independent predictive factor of response to chemotherapy. CONCLUSION: This prospective study suggests that p53 status may be a useful indicator of response to neoadjuvant chemotherapy in HNSCC.

New assay for measuring binding of platelet glycoprotein IIb/IIIa to unpurified von Willebrand factor.

Among the numerous variants of vWD, no patient with an abnormal vWF binding to GPIIb/IIIa has been described to date. To search for such potential variants, we developed a two-site assay for measuring the binding of purified GPIIb/IIIa to vWF in biological fluids and we used it to study a large series of plasmas from various types of von Willebrand disease (vWD) and recombinant vWF (rvWF). vWF in plasma or rvWF in culture medium was immobilized onto anti-vWF monoclonal antibodies (MoAb)-coated wells of microtiter plates. After incubation with either unlabeled GPIIb/IIIa and a 125I-anti-GPIIb/IIIa MoAb or 125I-GPIIb/IIIa, binding curves and binding isotherms were respectively established. Normal pool plasma and wild-type rvWF were used as reference samples. We tested plasmas from 85 normal subjects, 115 patients with different types of vWD (64 type 1, 2 type 3, 9 type 2A, 4 type 2M, 16 type 2B, 15 type 2N, 3 type IID and 2 acquired forms) and 50 patients with various bleeding disorders. Four mutated rvWF with 2A (Glu875Lys and Pro885Ser) or 2B (Dupl.Met540 and Val551Phe) substitutions and one rvWF mutated in the RGD domain of the C-terminal part of vWF-subunit (Asp1746Gly) were also studied. Among the various samples tested, only rvWF Asp1746Gly had no affinity for GPIIb/IIIa. In contrast, GPIIb/IIIa similarly bound to the other vWF, independently of the proteic environment, the factor VIII level, the degree of multimerization or the mutation of vWF. Our results indicate that subjects with an abnormal vWF binding to GPIIb/IIIa are probably rare and difficult to target for a specific screening.

Conformational changes in the A3 domain of von Willebrand factor modulate the interaction of the A1 domain with platelet glycoprotein Ib.

Bitiscetin has recently been shown to induce von Willebrand factor (vWF)-dependent aggregation of fixed platelets (Hamako J, et al, Biochem Biophys Res Commun 226:273, 1996). We have purified bitiscetin from Bitis arietans venom and investigated the mechanism whereby it promotes a form of vWF that is reactive with platelets. In the presence of bitiscetin, vWF binds to platelets in a dose-dependent and saturable manner. The binding of vWF to platelets involves glycoprotein (GP) Ib because it was totally blocked by monoclonal antibody (MoAb) 6D1 directed towards the vWF-binding site of GPIb. The binding also involves the GPIb-binding site of vWF located on the A1 domain because it was inhibited by MoAb to vWF whose epitopes are within this domain and that block binding of vWF to platelets induced by ristocetin or botrocetin. However, in contrast to ristocetin or botrocetin, the binding site of bitiscetin does not reside within the A1 domain but within the A3 domain of vWF. Thus, among a series of vWF fragments, 125I-bitiscetin only binds to those that overlap the A3 domain, ie, SpIII (amino acid [aa] 1-1365), SpI (aa 911-1365), and rvWF-A3 domain (aa 920-1111). It does not bind to SpII corresponding to the C-terminal part of vWF subunit (aa 1366-2050) nor to the 39/34/kD dispase species (aa 480-718) or T116 (aa 449-728) overlapping the A1 domain. In addition, bitiscetin that does not bind to DeltaA3-rvWF (deleted between aa 910-1113) has no binding site ouside the A3 domain. The localization of the binding site of bitiscetin within the A3 domain was further supported by showing that MoAb to vWF, which are specific for this domain and block the interaction between vWF and collagen, are potent inhibitors of the binding of bitiscetin to vWF and consequently of the bitiscetin-induced binding of vWF to platelets. Thus, our data support the hypothesis that an interaction between the A1 and A3 domains exists that may play a role in the function of vWF by regulating the ability of the A1 domain to bind to platelet GPIb.

Assay of factor VIII antigen (VIII:CAg) in 294 haemophilia A patients by a new commercial ELISA using monoclonal antibodies.

Monoclonal antibodies (MoAbs 833 and D4H1) directed against human factor VIII (FVIII) have been produced on a large scale to measure VIII:CAg by two-site ELISA (Asserachrom VIII:CAg, Diagnostica Stago). F(ab')2 from MoAb 833 were used for coating and bound VIII:CAg was revealed with MoAb D4H1 coupled to peroxidase. Control plasma (100 VIII:CAg U dL-1 by comparing with the International Standard) was used as reference. The assay sensitivity was 0.1 U dL-1 VIII:CAg. No apparent effect of the plasma proteins was observed provided plasma dilution was > or = 5. Thus this ELISA allowed us to estimate VIII:CAg levels of 0.5 U dL-1 in plasma. Levels of VIII:CAg were similar to those of VIII:C (correlation coefficient r = 0.87) in plasma from normal individuals (32 cases) and in patients with von Willebrand disease of various types (30 cases). Among 294 patients with haemophilia A (HA), 161 had severe HA (VIII:C < 1 U dL-1). Among those patients, 124 were cross-reacting material (CRM) negative with undetectable VIII:CAg and 37 were CRM+ (VIII:CAg 1-31 U dL-1). In 42 patients with moderate HA (VIII:C 1-5 U dL-1), 33 were CRM reduced (VIII:CAg 0.5-8 U dL-1) and nine were CRM+ with a VIII:CAg/VIII:C ratio of 6-91 (mean 34.3). In mild HA (91 cases with VIII:C > or = 6 U dL-1), 29 patients were classified as CRM+ (VIII:C 6-57 U dL-1, VIII:CAg 17-130 U dL-1 and VIII:CAg/VIII:C ratio 1.8-13.7 (mean 4.51)). In 62 CRM reduced patients there was a linear correlation between VIII:C (6-39 U dL-1) and VIII:CAg (2-36 U dL-1) levels (r = 0.88). In conclusion, this sensitive assay allows us to distinguish the quantitative CRM reduced and negative from the qualitative (CRM+) abnormalities in haemophilia A.

Variant forms of the pig lutropin/choriogonadotropin receptor.

The cloning and sequencing of porcine lutropin/choriogonadotropin (LH/hCG) receptor messenger RNAs have shown the presence of a full-length receptor (pLHR-A) and of shorter variants lacking either the transmembrane and the intracellular domains (pLHR-B and pLHR-C) or only the transmembrane domain (pLHR-D). Moreover, immunoblotting of testicular membrane extracts has detected 85-, 68-, and 45-48-kDa proteins reacting with antireceptor antibodies. Transfection experiments were performed to assign the protein species to the various messenger RNAs and to study the function of the various receptor species. COS-7 and L-cells transfected with an expression vector encoding full-length receptor pLHR-A yielded a protein of apparent molecular mass of 105 kDa. This corresponded to the complete receptor which had undergone a different glycosylation pattern to that found in testis, since after digestion with peptide N-glycosidase F both the 105-kDa COS-7 protein and the 85-kDa testicular glycoprotein yielded a holoprotein of approximately 63 kDa. Transfection with pLHR-A also yielded a high proportion of the 68-kDa glycoprotein which was shown by digestion with endoglycosidase H to be a high-mannose precursor of the full-length receptor. The existence of a large pool of precursor species in both transfected cells and Leydig cells evokes possible physiological regulations at the level of receptor maturation.(ABSTRACT TRUNCATED AT 250 WORDS)

Pathways of internalization of the hCG/LH receptor: immunoelectron microscopic studies in Leydig cells and transfected L-cells.

Monoclonal anti-receptor antibodies were used to study the cellular traffic of the hCG/LH receptor by immunoelectron microscopy. The LHR38 antibody was shown to bind to the extracellular domain of the receptor but not to interfere with hormone binding, adenylate cyclase activation or with the rate of internalization of the receptor. Pig Leydig cells and a permanent L-cell line expressing the LH receptor were used for the study. Incubation with LHR38-gold complexes showed the LH receptors to be randomly distributed over the cell surface including the clathrin coated pits. The LH receptors were internalized via a route including coated pits, coated vesicles and multivesicular bodies to lysosomes. This route is different from that observed for beta-adrenergic, muscarinic, and yeast mating factor receptors and considered previously as possibly general for G-protein-coupled receptors. The use of [125I]LHR38 allowed precise measurement of the rate of internalization, showing the existence of a constitutive pathway which was increased 11-fold by hormone administration. Double labeling experiments suggested that the hormone (hCG-Au15nm) and the receptor (labeled with LHR38-Au5nm) have similar routes of endocytosis, both of them being degraded in lysosomes. Studies of the reappearance of LHR38-Au5nm on the surface of the cells and the use of monensin indicated that only a very small proportion of the receptor molecules were recycled to the cell surface. The distribution and the intracellular pathways of LH receptors are very similar in Leydig cells and transfected L-cells. This opens the possibility of using the latter to study, by in vitro mutagenesis, the molecular mechanisms involved in the cellular traffic of LH receptors.

A protein kinase C inhibitory activity is present in the human T lymphoblast cell line Jurkat.

The human lymphoblast cell line Jurkat is widely used as a model system for studying signal transduction pathways during lymphocyte activation. We report the presence of a potent endogenous inhibitor of protein kinase C (PKC) in the cytosolic fraction of Jurkat cells. This inhibitor is not diffusible and is thermolabile; it is assumed to be a protein. It was separated from PKC by ion-exchange chromatography on DEAE-cellulose. The inhibitory activity was partially reversed by increasing the concentration of the PKC substrate; increasing that of PKC activators (calcium and phospholipids) was without effect. PKC activity was inhibited by more than 90% in the crude cytosolic fraction but the inhibition could be completely reversed by diluting the cell extract. This inhibitory activity could not be detected in the cytosol from normal lymphocytes or from lymphoblasts from leukemic patients.

Retinoic acid inhibits the expression of cytidine deaminase linked to the differentiation of the human leukemic cell line HL-60.

The activity of cytidine deaminase markedly increases during the differentiation of HL-60 cells induced by dimethylsulfoxide or 1,25-dihydroxy vitamin D3, but does not increase when the inducer is retinoic acid. Here it is demonstrated that retinoic acid inhibits the increase in cytidine deaminase activity elicited by the other two inducers. This inhibitory effect of retinoic acid (i) was not the result of a direct action on the enzymatic activity; (ii) was correlated with the differentiating effect of retinoic acid, as indicated by the similar time-course and dose-dependence of both effects, and by additional studies with various retinoids and with an HL-60 variant resistant to retinoic acid-induced differentiation; (iii) required the continued presence of the drug for more than 24 h, and could not be reversed after 48 h; (iv) was manifest, after a lag-time of 24 h, at whatever time retinoic acid was added during the 5 days of treatment of the cells with the differentiation inducers; and (v) was prevented by the addition of the protein synthesis inhibitor cycloheximide. These data indicate that retinoic acid negatively regulates the expression of cytidine deaminase in HL-60 cells, and suggest that this effect is mediated by a protein, the synthesis of which should be controlled by the nuclear receptor of retinoic acid.

Monoclonal antibodies against luteinizing hormone receptor. Immunochemical characterization of the receptor.

Human CG (hCG)-receptor complexes were solubilized from porcine testicular membranes. They were chromatographed on an immunomatrix of Affi-Gel 10-D1E8 anti beta-hCG monoclonal antibody (this antibody has been shown not to interfere with hCG binding to receptor). Elution was performed at alkaline pH, a condition in which hCG-receptor complexes are relatively stable. Immunization of a mouse with these partially (approximately 15%) purified hormone-receptor complexes allowed the preparation of 20 different hybridomas, each secreting antireceptor antibodies. The latter were used for receptor characterization. Immunoblot of testicular membrane extracts or of purified receptor preparations showed the presence of a major band at approximately 85,000 mol wt and minor bands at approximately 68,000 mol wt and approximately 48-45,000 mol wt. The width of all these bands suggested some microheterogeneity possibly due to glycosylation. The same approximately 85,000 mol wt receptor was seen in ovarian membranes, but no detectable antigen was observed in liver, muscle, and kidney membranes. An immunoaffinity method (using antibody LHR 38) was devised to purify the receptor in a single step. This demonstrated that the purified receptor preparation did not contain any protein component other than those detected by immunoblot. Comparison of receptors purified by immunoaffinity chromatography using either antireceptor or antihormone monoclonal antibodies showed in both cases the presence of the 85,000 mol wt and 48-45,000 mol wt species, but the absence, in the latter case, of the 68,000 mol wt species. This suggests that the 68,000 mol wt receptor cannot bind hormone and does not form oligomers with other receptor species.

Influence of adenosine deaminase inhibition on the phosphoinositide turnover in the initial stages of human T cell activation.

An experimental model of adenosine deaminase deficiency was established on the human T cell line Jurkat by using 2'-deoxycoformycin, a strong specific inhibitor of the enzyme. When deoxyadenosine was added to the inhibited cells, the nucleotide profile was modified reproducing that found in lymphocytes from adenosine deaminase-deficient children. The metabolism of phosphoinositides, analyzed by either the release of [3H]inositol phosphates or the breakdown of 32P-prelabeled phosphatidyl inositides, was compared in normal and modified cells where dATP was accumulated. No modification in 32P labeling of phosphoinositides was detectable within the 32P-loading period. However, when the cells were stimulated by phytohemagglutinin or anti-CD3 monoclonal antibody, the phosphoinositide hydrolysis was strongly reduced in the dATP-containing lymphoblasts. This decrease was correlated with the intracellular dATP concentration.

High frequency of plasminogen activator secretion by malignant human lymphoid cell lines of T-cell type origin.

Diffuse defibrination is rarely observed in acute lymphoblastic leukemia (ALL). Clinical and immunologic data suggest that it is more likely to occur in T cell derived ALL. The current investigation involved the secretion of plasminogen activators (PA) of tissue type (t-PA) or urokinase type (U-PA) by testing supernatants of 21 permanent human leukemic cell lines originating from various hematopoietic lineages and one induced lymphoblastoid cell line. (LCL) The amount of PA in each supernatant was determined by biologic and immunoenzymologic assays. The correspondence with the expected molecular weight (MW) according to the PA type was checked by zymography. PA secretion of U-PA type was observed in the three myeloid cell lines. Except for the normal LCL, no B-lymphoid lineage related cell lines of various levels of differentiation displayed PA secretion, whereas PA activity was observed in the supernatant of six of nine malignant T-cell lines. The T-leukemic cell lines CCRF CEM, KE 37, HUT 78, and HUT 102 released U-PA-like activity. Peer released t-PA-like activity and CCRF-HSB2 supernatant showed both types of PA activity. These findings are discussed in view of the natural history of these diseases and the stage of differentiation of the cell lines.

Metabolism of pyrimidine bases and nucleosides by pyrimidine-nucleoside phosphorylases in cultured human lymphoid cells.

The anabolism of pyrimidine ribo- and deoxyribonucleosides from uracil and thymine was investigated in phytohemagglutinin-stimulated human peripheral blood lymphocytes and in a Burkitt's lymphoma-derived cell line (Raji). We studied the ability of these cells to synthesize pyrimidine nucleosides by ribo- and deoxyribosyl transfer between pyrimidine bases or nucleosides and the purine nucleosides inosine and deoxyinosine as donors of ribose 1-phosphate and deoxyribose 1-phosphate, respectively: these reactions involve the activities of purine-nucleoside phosphorylase, and of the two pyrimidine-nucleoside phosphorylases (uridine phosphorylase and thymidine phosphorylase). The ability of the cells to synthesize uridine was estimated from their ability to grow on uridine precursors in the presence of an inhibitor of pyrimidine de novo synthesis (pyrazofurin). Their ability to synthesize thymidine and deoxyuridine was estimated from the inhibition of the incorporation of radiolabelled thymidine in cells cultured in the presence of unlabelled precursors. In addition to these studies on intact cells, we determined the activities of purine- and pyrimidine-nucleoside phosphorylases in cell extracts. Our results show that Raji cells efficiently metabolize preformed uridine, deoxyuridine and thymidine, are unable to salvage pyrimidine bases, and possess a low uridine phosphorylase activity and markedly decreased (about 1% of peripheral blood lymphocytes) thymidine phosphorylase activity. Lymphocytes have higher pyrimidine-nucleoside phosphorylases activities, they can synthesize deoxyuridine and thymidine from bases, but at high an non-physiological concentrations of precursors. Neither type of cell is able to salvage uracil into uridine. These results suggest that pyrimidine-nucleoside phosphorylases have a catabolic, rather than an anabolic, role in human lymphoid cells. The facts that, compared to peripheral blood lymphocytes, lymphoblasts possess decreased pyrimidine-nucleoside phosphorylases activities, and, on the other hand, more efficiently salvage pyrimidine nucleosides, are consistent with a greater need of these rapidly proliferating cells for pyrimidine nucleotides.

Synergistic toxicity of pyrazofurin and cytidine in cytidine deaminase deficient lymphoid cells (Raji).

The intermediary metabolism of pyrimidine nucleosides was studied in a line of human B lymphoblasts (Raji) in which pyrimidine de novo synthesis deficiency was pharmacologically induced by pyrazofurin. It was found that Raji cells are cytidine deaminase deficient that cytidine has a synergistic effect on the toxicity of pyrazofurin towards these cytidine deaminase deficient cells, affecting both the proliferation and the viability of the cells. Indirect evidences suggest that this synergistic toxicity is not mediated by an effect on nucleoside diphosphate reductase nor on the first steps of pyrimidine de novo synthesis.

Cytidine deaminase activity of human normal and malignant lymphoid cells.

Cytidine deaminase activity was determined by a radioisotopic assay in extracts of peripheral blood mononuclear cells of normal individuals and of patients with acute lymphoblastic leukaemia. The normal enzyme activity had a broad pH optimum between pH 6.5 and 8.0; apparent Km values for cytidine and deoxycytidine were 3.6 +/- 0.6 mumol/l and 26.5 +/- 3.5 mumol/l, respectively; the activity was resistant to heat inactivation; of the various effectors tested, only uridine, deoxyuridine and tetrahydroxyuridine had inhibitory effects. Cytidine deaminase activity was markedly decreased in lymphoblasts of patients with acute lymphoblastic leukaemia; enzyme activity was related to the percentage of circulating blast cells, and not to the clinical, cytological or immunological characters of the leukaemia.