A new approach to clustering the amino acids.

Each amino acid is represented by a vector of numerical measurements for the attributes of volume, area, hydrophilicity, polarity, hydrogen bonding, shape, and charge. Inter-residue distances are then calculated according to common metrics, and we introduce a new clustering objective function derived from information-theoretic considerations. The arguments of the function are the inter-object distances of the things to be clustered: in this case the amino acids. By means of approximating the solution of an integer programming problem, then, the residues are partitioned into clusters. The clusters obtained are compared with groups obtained in substitution/mutation studies and found to be similar. Thus, probably the strongest and most objective evidence to date is supplied for believing that physico-chemical properties account for the viability of substitutions and that the important similarities/differences are explained by a relatively small and simple set of properties.

Contributions of hippurate, indoxyl sulfate, and o-hydroxyhippurate to impaired ligand binding by plasma in azotemic humans.

We have evaluated pH, chloride, calcium and several endogenous aromatic acids as possible causes of the impaired binding of drugs by plasma albumin in renal failure. Changes in pH, chloride and calcium over the range found in renal failure had minimal or no effects on the binding of [14C]salicylate, a model probe which binds to both of the major drug-binding loci of human albumin. Hippurate and indoxyl sulfate were weak inhibitors of binding by normal plasma. Ortho-hydroxy-hippurate was undetectable or minimally elevated, except among patients with elevated plasma salicylate concentration. Although plasma hippurate and indoxyl sulfate concentrations were elevated markedly in patients with renal failure, inhibition of salicylate binding was significantly correlated only with the concentration of indoxyl sulfate. Addition of hippurate and indoxyl sulfate separately and together to normal plasma showed that these ligands could account for only 15% of the impaired binding of salicylate by azotemic plasma. The retained solutes which account for most of this binding defect remain to be identified. This uremic disorder (and perhaps others) is due not to a single chemical but to the additive effect of a family of chemicals.

Plasma hippurate in renal failure: high-performance liquid chromatography method and clinical application.

We have developed a high-performance liquid chromatography (HPLC) method for assay of hippurate in plasma of patients with renal failure. Hippurate accounts, in part, for the impaired binding of drugs and metabolites to albumin and may cause other disorders in azotemic patients. The method is precise, accurate and reproducible. Among 25 patients with acute and chronic renal failure having serum creatinine in the range of 2.9-43 mg/dl (256-3,801 mumol/l), plasma hippurate ranged from 0.11 to 16.2 mg/dl (6.1-904 mumol/l). Hippurate concentration correlated moderately closely with plasma creatinine, urea and anion gap. Its curvilinear relation to the reciprocal of serum creatinine indicated a proportional decline of GFR and tubular function or the accumulation of inhibitors of the proximal tubular anion secretory pathway. The method should be useful for further studies of abnormal albumin binding as well as other disorders in azotemic patients.

Isolation and chemical identification of inhibitors of plasma ligand binding.

The binding by serum albumin of many drugs and endogenous metabolites is impaired in humans and animals with renal failure. Unknown solute(s) retained in renal failure have been extracted from uremic fluids. When added to normal plasma they induce a similar binding defect. Similar activity can be extracted from normal urine. We have devised a series of extraction and purification techniques that yielded three binding inhibitory ligands from normal human urine in sufficient quantity and of a high degree of purity. Rigorous methods have been applied to determine chemical identity of the ligands. Purification steps consisted of: adsorption at pH 3.0 to polystyrene-divinylbenzene resin (XAD-2); elution from the resin with methanol followed by drying and solution in dilute formic acid; passage through SP-Sephadex to remove cations, especially yellow-brown pigments; adsorption to the anion exchanger QAE-Sephadex, and separation into three zones of inhibitory activity with a formic acid gradient; purification to homogeneity with C-8 or C-18 silica reversed-phase chromatography. Using this isolation procedure, followed by mass spectroscopy and nuclear magnetic resonance spectroscopy, we have shown that the binding inhibitory activity is due not to one ligand, but to a family of aromatic acids. To date hippurate, beta-(m-hydroxyphenyl)-hydracrylate and p-hydroxyphenylacetate have been identified as binding inhibitors. Other active ligands remain to be identified.

Determination of indoxyl sulfate in plasma of patients with renal failure by use of ion-pairing liquid chromatography.

Using our newly developed ion-pairing reversed-phase liquid-chromatographic method for assay of indoxyl sulfate, we measured its concentration in plasma of normal subjects and patients in various degrees of renal failure. Response was linear over the range of 50 to 25 000 pmol of indoxyl sulfate injected into the chromatograph. We demonstrated the specificity of the assay for azotemic plasma by using enzymatic conversion with a sulfatase. For a moderately above-normal indoxyl sulfate concentration in azotemic plasma of 134 mumol/L (29 mg/L), the within-day CV was 1.6%, the day-to-day CV 2.8%. Mean analytical recovery was 101.0% (CV = 2.8%). Over a range of 29 to 192 mg of creatinine per liter of plasma (x), indoxyl sulfate (y) concentration (in mumol/L) was positively correlated (y = 1.30x + 0.43). This method should prove valuable for further study of uremic toxins.

Applications of clustering theory to cancer mortality data.

A clustering problem is one which requires the partitioning of a finite set of objects into subsets so as to optimize a given objective function. The present paper applies such techniques to a well-known set of data on U.S. cancer mortality over the years 1950 to 1967. Two varieties of clustering solutions are obtained: first, states of the United States are clustered according to similarities in their patterns of cancer mortality; then, tumor types are clustered according to mortality similarities over the set of all states. In both, a number of clustering experiments are conducted and these involve the application of a variety of metrics to the raw data, as well as both heuristic and exact solutions of the mathematical problems. The results are displayed graphically, discussed, and in the case of cancer clusters, compared to observed clinical associations.

Suppression of para-aminohippurate transport in the isolated perfused kidney by an inhibitor of protein binding in uremia.

Serum from patients with advanced renal failure contains substances that inhibit binding of small ligands to albumin. The extractable inhibitors of binding to albumin (lx), which were previously shown in rat kidney slices to inhibit para-aminohippurate (PAH) transport, were added to the in vitro perfused rat kidney in an attempt to investigate effects on organic acid transport in the intact organ. Following addition of extract to the perfusate there was an immediate and profound decrease in whole kidney PAH secretion. Mean PAH clearance fell an average of 60% +/- 7% after addition of extract from human uremic pleural exudate and an average of 62% +/- 6% after addition of similarly prepared extract from normal human urine. There was no change in inulin clearance, vascular resistance, urine flow, or fractional reabsorption of sodium. The effect was shown to be as marked when bovine serum albumin (4.0 g/dL) was added to the perfusate, clearances falling 66% +/- 7%. Controls showed no change in PAH or inulin clearance during 90 minutes of in vitro perfusion. An increase in ultraviolet (UV) absorbance by the urine within minutes after addition of inhibitor to the perfusate suggested tubular secretion of a major portion of the extract. Partial recovery of PAH clearance was observed after 20 to 40 minutes in most kidneys and near complete recovery occurred in some kidneys. The inhibitory effect was duplicated by addition of hippurate at 24 mg/dL (1.2 mmol/L) to the perfusate.(ABSTRACT TRUNCATED AT 250 WORDS)

Isolation of inhibitors of ligand: albumin-binding from uremic body fluids and normal urine.

Unknown ligands that are retained in renal failure bind to albumin and impair binding of endogenous metabolites and many drugs. We previously showed that an extract from uremic serum can induce a similar abnormality in normal plasma. These binding inhibitors are also present in normal urine. In the present studies we showed by analytical reversed-phase HPLC that the extract of uremic serum contained over 50 UV absorbing peaks. Preparative anion-exchange chromatography of extracts of normal urine and uremic pleural fluid yielded three to four peaks of phenytoin-binding inhibitory activity. The largest peak from normal urine was shown by proton NMR and mass spectroscopy to contain hippuric acid. An earlier eluting peak was found by HPLC to contain several inhibitors-one was definitely identified as p-hydroxyphenylacetic acid and a second as probably beta-(m-hydroxyphenyl)-hydracrylic acid. These methods may be useful for identifying other uremic toxins. The binding abnormality and perhaps other uremic derangements result from the total effect of a family of toxins rather than from the action of a single retained solute.

Impaired plasma phenytoin binding in uremia. Effect of in vitro acidification and anion-exchange resin.

Phenytoin binding to uremic plasma was studied in vitro. Inhibition of binding independent of albumin concentration was demonstrated. Acidification from pH 8 to pH 3 produced a small decrease in binding by normal plasma but no change or a small increase in binding by uremic plasma. After plasma was acidified to pH 3.0, passed through an anion-exchange resin and realkalinized to pH 8, the binding by uremic plasma was restored to near normal. These studies indicate that the uremic abnormality causing impaired plasma drug binding is reversible and supports the concept that a competing ligand, tightly bound at physiologic pH, is responsible for impaired drug binding.