Mechanistic and Synthetic Implications of the Diol-Ritter Reaction: Unexpected Yet Reversible Pathways in the Regioselective Synthesis of Vicinal-Aminoalcohols.

The Ritter reaction of 1,2-diolmonoesters with nitriles to 1- vic-amido-2-esters proceeds through dioxonium and nitrilium cation intermediates. To provide the basis for the reaction mechanism, novel forms of these cations were isolated, characterized, and studied by spectroscopic methods and single crystal X-ray analysis. Ground and transition state energies were determined both experimentally and theoretically. Taken together, these data suggest that the reaction proceeds via rapid formation of the dioxonium cation 9, followed by rate determining yet reversible ring opening by acetonitrile to the corresponding nitrilium cation 10 (computed Δ G⧧ = 24.7 kcal at 50 °C). Rapid, irreversible hydration of the latter affords the corresponding vic-acetamido ester. Controlled addition of H2O to the dioxonium cation 9 in acetonitrile- d3 results in near-quantitative production of deuterated acetamido ester 13a. Kinetics of this conversion (9 to 13a) are biphasic, and the slow phase is ascribed to either direct cation 9 attack by acetamide to form cation 16 via O-alkylation or by reversible ether formation. Deuterium labeling studies suggest O-alkylated cation 16 does not directly isomerize to N-alkylated cation 18; instead, it reverts to vic-amidoester 13a via the nitrilium pathway. Preliminary results indicate high regioselectivity for primary amide formation in the diol-Ritter sequence.

Studies of the extraction of bilirubin from human amniotic fluid.

The recovery of unconjugated bilirubin from human amniotic fluid was studied using dichloromethane, chloroform/isopropanol (3:1 vol/vol), and chloroform/ methanol (3:1 vol/vol) extraction of human amniotic fluid that had been supplemented with bilirubin at various concentrations. Results were compared with those obtained with conventional chloroform extraction. Mean recoveries were found to be only 28% for chloroform and 25% for dichloromethane. When the polarity of chloroform was increased by the addition of an alcohol, the mean recovery increased to only 40% for chloroform/isopropanol and 38% for chloroform/ methanol. These results suggest that extraction methods for determination of amniotic fluid "delta OD(450)" (visible spectrophotometric absorbance [optical density] of bilirubin at 450 nm) tend to underestimate the result when compared with the nonextraction (direct-scan) method, on which the Liley Prognostication Chart is based. This finding should be clinically significant, particularly if extraction and direct-scan methods are used to monitor the condition of the same patient.

Synthesis of biologically active amines via rhodium-bisphosphite-catalyzed hydroaminomethylation.

[reaction: see text] We report the use of a highly regioselective rhodium-bisphosphite catalyst for olefin hydroaminomethylation. This catalyst system was successfully applied in the synthesis of two biologically active tertiary amines, ibutilide and aripiprazole.

Studies of the inhibition of serum pseudocholinesterase activity in vitro by commonly used drugs.

We studied the effect of 17 commonly used drugs, including prescription and over-the-counter medications, on the activity of serum pseudocholinesterase (PCE) in vitro. Normal pooled human serum was incubated for 120 minutes at 37 degrees C with therapeutic serum concentrations of prescription and over-the-counter drugs, and the postincubation PCE activity was measured. Morphine, quinidine, and thioridazine depressed PCE activity by more than 5% while no or negligible effect was noted following incubation with acetaminophen, chlordiazepoxide, chlorpromazine, desipramine, doxepin, imipramine, methamphetamine, nortriptyline, phenobarbital, phenytoin, procainamide, salicylic acid, theophylline, and valproic acid. Depression of PCE activity can prolong the half-life of coadministered agents with metabolism mediated by PCE.

Valproic acid binding to human serum and human placenta in vitro.

The binding characteristics of the antiepileptic agent and teratogen valproic acid for human serum and human placenta were investigated utilizing equilibrium dialysis of the drugs in serum and in homogenates of whole placenta so that the transplacental transfer of the drug could be better defined. A low-capacity, high-affinity binder and a high-capacity, low-affinity binder for valproic acid were found in serum. However, there was only minimal, nonspecific binding of the drug to placenta. It appears that transplacental transfer of valproic acid is not mediated by binding to placenta. It is likely that such transfer occurs by passive diffusion, which may be facilitated by the lipophilicity of the drug and by development of a pH gradient across the placenta. These findings indicate that the placenta is not a depot for valproic acid.

The effect of lovastatin and thioridazine on the degradation of cocaine in human serum in vitro.

The effect of lovastatin and thioridazine on the degradation of cocaine in human serum was studied by incubating therapeutic and toxic concentrations of either drug with cocaine in human serum at 37 degrees C. Without these other drugs, cocaine concentrations decreased by an average of 88% in 120 minutes. Lovastatin at all concentrations studied showed a negligible effect on the degradation of cocaine in human serum and did not alter the pseudocholinesterase activity of the serum. In contrast, in the presence of a therapeutic concentration of thioridazine, cocaine concentrations decreased by only 71% during this period of time, and, with a toxic concentration of thioridazine, cocaine concentrations decreased by only 55%. Thioridazine suppressed the pseudocholinesterase activity of human serum by up to 19% during the 120-minute incubation period, and this effect was more pronounced at higher thioridazine concentrations. These findings suggest that thioridazine may prolong the serum half-life of cocaine by inhibiting the pseudocholinesterase-mediated catabolism of cocaine to ecgonine methyl ester. This may be important in cocaine users who are treated with phenothiazines and other structurally similar drugs. They also may be of interest in cocaine-abusing patients who are treated with phenothiazines for schizophrenic disorders.

The binding of acetaminophen, lidocaine, and valproic acid to human milk.

The binding of acetaminophen, lidocaine, and valproic acid to pooled normal mature human milk was studied in vitro by using equilibrium dialysis. Scatchard analysis revealed high-affinity, low-capacity binding for acetaminophen (Ka [affinity constant of association], 1.47 x 10(4) L/mol; Bo [concentration of binding sites], 9.01 x 10(-4) mol/L) and some minimal, nonspecific binding. Binding ranged up to 85%. For lidocaine, low-affinity, high-capacity binding was noted (Ka, 1.42 x 10(2) L/mol; Bo, 1.69 x 10(-2) mol/L). Binding ranged up to 72%. For valproic acid, only minimal, nonspecific binding was noted at low drug concentrations with binding ranging up to 64%. The binding of these drugs to milk might enhance their excretion and subsequent ingestion by infants who are breast-fed. In addition, the low pH of the milk (6.24) may cause "ion trapping" of acetaminophen (pKa, 9.5) and lidocaine (pKa, 7.9).

Gentamicin and tobramycin binding to human serum in vitro.

The binding of gentamicin and tobramycin to human serum was studied in vitro using equilibrium dialysis of pooled human serum supplemented to various concentrations of either drug. Only minimal and variable non-specific binding was noted for each drug: gentamicin, less than 15% and tobramycin, less than 30%. Conventional Scatchard analysis conducted over an array of drug concentrations failed to identify any specific binding proteins.

The effect of parenteral nutrition fluids on the binding of therapeutic drugs to human serum in vitro.

The competitive binding of seven therapeutic drugs (carbamazepine, phenytoin, phenobarbital, procainamide, quinidine, theophylline, and valproic acid) to human serum and to five commonly used parenteral nutrition fluids in vitro was studied using equilibrium dialysis. For five of the drugs, all parenteral nutrition fluids bound less drug than human serum-phenobarbital (up to 14% less), phenytoin (up to 46% less), procainamide (up to 43% less), quinidine (up to 25% less), and valproic acid (up to 77% less)-suggesting that the presence of these fluids might increase the free fraction of these drugs in vivo. For carbamazepine, the fluids bound up to 82% more drug, suggesting that the presence of these fluids might decrease the free fraction of this drug in vivo. For theophylline, the fluids produced a minimal (no more than 5%) effect on binding to serum. The administration of parenteral nutrition fluids may significantly alter the free (active) fraction of some therapeutic drugs.

The binding of selected therapeutic drugs to human serum alpha-1 acid glycoprotein and to human serum albumin in vitro.

The binding of acetaminophen, lidocaine, phenobarbital, phenytoin, theophylline, and valproic acid to human serum alpha-1 acid glycoprotein (orosomucoid) and to human serum albumin separately in vitro was investigated using equilibrium dialysis of the unlabeled drugs. Each drug was studied at a therapeutic concentration. Alpha-1 acid glycoprotein was studied at one elevated and two physiological concentrations, whereas albumin was studied at one physiological and two low concentrations. The nonphysiological concentrations were consistent with those that might be seen in a variety of clinical conditions. Acetaminophen, phenobarbital, theophylline, and valproic acid showed negligible binding to alpha-1 acid glycoprotein. However, lidocaine and phenytoin demonstrated binding to this protein, and increases in the alpha-1 acid glycoprotein concentration produced decreases in the unbound (free) or "active" concentration of these two drugs. All drugs but acetaminophen bound to albumin, and decreases in the albumin concentration yielded increases in the unbound (free) or "active" concentration of the remaining 5 drugs. These findings are significant when lidocaine, phenytoin, phenobarbital, theophylline, or valproic acid are used in patients with clinical conditions that may affect the concentration of these two binding proteins.

The effect of ethanol and pH on the adsorption of drugs from simulated gastric fluid onto activated charcoal.

The effect of ethanol and pH on the adsorption of acetaminophen (ACET), phenobarbital (PHB), phenytoin (PHY), salicylic acid (SA), and theophylline (THEO) from simulated gastric fluid onto activated charcoal was studied. For the ethanol study, each drug was prepared at a concentration of 10 g/L in ethanol; in hydrochloric acid (HCl), 1.2 mol/L; and in HCl, 1.2 mol/L, containing 75% ethanol, 50% ethanol, and 25% ethanol (v/v), respectively. For the pH study, each drug was prepared at a concentration of 10 g/L in HCl, 1.2 mol/L, pH 1.0, and in buffers of pH 1.7, 3.0, 4.0, 4.8, 5.8, 6.5, 7.4, and 9.4. After the addition of 1 g of activated charcoal to 10 mL of each solution, it was incubated for one hour at 37 degrees C. For comparison, in each experiment a blank consisting of the solution without charcoal was also incubated. With increasing concentrations of ethanol, there were substantial decreases in the adsorption of ACET, PHB, and PHY to charcoal. Ethanol-induced decreases in the adsorption of SA and THEO were less pronounced. Changes in pH did not affect the adsorption of ACET, PHB, PHY, or THEO. However, the adsorption of SA was decreased slightly at pH 1.0 and 3.0.

Procainamide and quinidine inhibition of the human hepatic degradation of meperidine in vitro.

Procainamide and quinidine inhibition of the degradation of meperidine in human liver was investigated by incubation of two concentrations of either drug with meperidine in homogenates of human liver over 24 and 36 h. Meperidine concentrations declined by 26% after incubation for 24 h and by 42% after incubation for 36 h. In the presence of procainamide, however, they decreased by only 15% to 18% at 24 h and by only 26% to 28% at 36 h. In the presence of quinidine, they declined by only 18% to 19% at 24 h and by only 27% to 28% at 36 h. Procainamide and quinidine may inhibit human hepatic carboxylesterase hCE-1, which is responsible for catalyzing the hydrolysis of meperidine. This inhibition may prolong the biological half-life of meperidine in patients receiving the drug together with either procainamide or quinidine.

Carisoprodol: an unrecognized drug of abuse.

During a 6-month monitoring period, carisoprodol was detected in the urine specimens of 19 patients for whom drug screening had been ordered for purposes of patient care. The clinical history suggested that in 7 cases the drug was abused or implicated in a suicide attempt or gesture. In another 7 cases, the drug was used primarily for medical purposes, and in 5 cases the reason for use could not be determined. One patient ingested homemade tablets that were found to contain carisoprodol. In an additional case, the drug was detected in breast milk. Physical findings, clinical history, and treatment are described, and the profile of a typical carisoprodol user is discussed. It seems that carisoprodol has become an unrecognized drug of abuse, at least in our community. This drug and its metabolite, meprobamate, should be included in comprehensive drug screening.