Multivariate examination of response surfaces around the reaction conditions for the Scandinavian Society's recommended method for creatine kinase determinations.

We examined the reaction surfaces around five variables (imidazole, ADP, creatine phosphate, magnesium, and pH) in the Scandinavian method for determining creatine kinase, using factorial experimentation (five level, five factor) at reaction temperatures of 30 and 37 degrees C. Theoretical response surfaces were computed by fitting a quadratic polynomial equation to the experimental data by least-squares regression. Essentially no differences were apparent in the theoretical curves among the five specimens we analyzed, or between reaction temperatures. Our response-surface data showed the following: for pH and imidazole, activity optima in the region of the Scandinavian conditions; for creatine phosphate, a broad plateau over the concentration range investigated (10 to 50 mmol/L); and for magnesium and ADP, gently increasing contours with maximal creatine kinase activity at concentrations greater than those investigated in our study (magnesium 15 mmol/L, ADP 3.5 mmol/L).

Creatine kinase response surfaces explored by use of factorial experiments and simplex maximization.

We conducted a five-component, five-level response-surface experiment to optimize the pH and the concentrations of magnesium, creatine phosphate, adenosine diphosphate, and buffer in an assay for creatine kinase. Under optimal conditions, creatine kinase activity was about 5% greater than that obtained with a previously reported assay (Clin Chem 23: 1569, 1977). We also applied a simplex maximization algorithm to the response-surface equation to locate areas of maximum sensitivity. Reaction conditions for two such areas were found, each yielding approximately 11% more activity than with the previously reported method.

Stability of creatine kinase-3 in lyophilized materials.

We examined the stability of creatine kinase (EC 2.7.3.2) isoenzyme-3 (CK-3) in lyophilized bovine albumin matrices in the presence and absence of various sulfhydryl compounds and ADP. We initially purified CK-3 from human myocardium and skeletal muscle by the batch-chromatographic technique and by gradient elution column chromatography to specific activities of 293 and 93 kU/g, respectively. To assess stability, we subjected the lyophilized materials to storage studies at 4, 25, 37, 42, 56, and 65 degrees C and compared first-order rate constants for the decay of creatine kinase activity at 42 degrees C. Our most stable matrix contained, per liter, 2 mmol of ADP and 10 mmol of N-acetylcysteine, and had an extrapolated first-order half-life (Arrhenius plot) at -20 degrees C of approximately 60000 years.

Characterization and intermethod relationships of materials containing purified human pancreatic and salivary amylase.

We describe the preparation and characterization of materials containing human pancreatic and salivary alpha-amylase (EC 3.2.1.1) and examine their relationship to endogenous amylase in human serum. Amylase was purified from human pancreas and saliva by solvent- and salt-fractionation and column chromatography to specific activities of 63 and 279 kU/g, respectively. Four liquid pools, differing only in activity, were prepared from each source of amylase, each in a matrix containing, per liter: 30 g of human albumin, 50 mmol of sodium chloride, 1 mmol of calcium chloride, and 50 mmol of Tris hydrochloride buffer, pH 7.4. Characterization of the pools showed that the amylase activity in the materials was stable for at least six months at 25 degrees C; among-vial variability of amylase activity was less than or equal to 0.5% (2 CV); and the pools were free from eight possible contaminating enzymes. Plots of salivary vs pancreatic amylase activity measure in our materials with eight commercially available methods showed least-squares slopes ranging from 0.51 to 1.0. The intermethod "commutability" of the materials (i.e., how closely they mimic endogenous serum amylase) was examined in relationship to approximately 100 human sera.

Column chromatography and immunoassay compared for measuring the isoenzymes of aspartate aminotransferase in serum.

We compare a column-chromatographic method and a homogeneous immunoassay method for separately measuring the mitochondrial and cytoplasmic isoenzymes of aspartate aminotransferase. Analytical recovery for the two methods averaged 102% (SD, 2%) and 103% (SD, 4%), respectively, for 11 pools prepared by adding the purified isoenzymes to serum and 102% (SD 8.9%) and 89% (SD, 8.1%) for 26 unaltered specimens of human serum. In comparing the results of the immunoassay method (y) to the chromatographic method (x), our measurements agreed closely for the mitochondrial (y = 0.947 X + 7, r = 0.9991, standard error of estimate = 2.9 U/L) and cytoplasmic (y = 0.92x-6, r = 0.9995, standard error of estimate = 2.1 U/L) isoenzymes in pools prepared from the purified isoenzymes. Similar measurements of the 26 human serum specimens yielded the following results for least-squares evaluation; cytoplasmic isoenzyme y = 1.03x-11, r = 0.994, and standard error of estimate = 6.0 U/L; mitochondrial isoenzyme y = 0.75x+0, r = 0.927, and standard error of estimate = 3.9 U/L.

Relative stabilities of purified human mitochondrial and cytoplasmic isoenzymes of aspartate aminotransferase in lyophilized materials.

Eight different pools of purified human mitochondrial and cytoplasmic isoenzymes of aspartate aminotransferase were prepared, to examine the effects of the following matrix variables: the matrix support material (bovine serum albumin and polyvinylpyrrolidone), endogenous pyridoxal concentration, and azide as an antimicrobial preservation. Storage temperatures of 25 and 37 degrees C were used as a rapid and convenient means of accelerating the degradation process. Activity of the enzyme was measured with and without pyridoxal in the reaction solution. We found that the mitochondrial isoenzyme was consistently more labile than the cytoplasmic isoenzyme under identical storage conditions. Both isoenzymes were more stable in matrixes containing bovine serum albumin than in those containing polyvinylpyrrolidone. No apparent difference in the stability of either isoenzyme was observed at matrix pyridoxal concentrations of 15 micromol/L and 150 micromol/L. Only the mitochondrial isoenzyme in matrixes containing bovine serum albumin and 15 micromol of pyridoxal per liter had increased activity (about 9%) when pyridoxal was added to the enzymatic reagent. The amount of activity in reconstituted specimens did not apparently change after 72 h at 4 degrees C.

Column-chromatographic separation of isoenzymes of aspartate aminotransferase.

We describe a column-chromatographic method for separating the mitochondrial and cytoplasmic isoenzymes of aspartate aminotransferase in human serum. Bed height of the ion exchanger, pH, and salt concentrations in the eluting buffers are shown to be variables affecting the separation of the isoenzymes. Under the optimized conditions selected for this study, a 30% increase in volume was observed in one fraction, associated with changing the salt concentration of the eluting buffer and attributed to a contraction of the DEAE-Sephadex A-50. Elution profiles (enzyme activity vs. fraction number) were examined with highly purified mitochondrial and cytoplasmic isoenzymes of human origin in bovine serum albumin and human serum. Recovery of the enzyme in the eluted fractions averaged 102% (SD, 2.0%) for specimens prepared from the purified isoenzymes and 104% (SD, 10.7%) for 38 human serum specimens. The separation technique showed linearity to catalytic concentrations in excess of 200 U/liter (reaction temperature 30 degrees C) for each isoenzyme. Additional information is presented regarding among-day precision and the effect of specimen dilution.

An interlaboratory study of measurement of aspartate aminotransferase activity with use of purified enzyme materials.

The Center for Disease Control (CDC), the New York State Department of Health (NYSDH), the College of American Pathologists, and 23 manufacturers of diagnostic products participated in an interlaboratory study of aspartate aminotransferase (EC 2.6.1.1) methodologies. Six different lyophilized materials were prepared and characterized and then distributed to 293 laboratories for aspartate aminotransferase measurements. The specimens included one human serum; four catalytic concentrations of the cytoplasmic isoenzyme, two purified from human erythrocytes, and two from porcine heart; and one matrix bovine serum albumin (30 g/liter) blank. The purified isoenzymes were prepared in the matrix. We present data on Michaelis parameters (Km and Vmax), Arrhenius plots, activation with pyridoxal 5-phosphate, vial-to-vial variability, and stability on reconstitution. The 281 responses showed that most of the laboratories used NADH-detection methods (91.1%), monitored at 340 nm (79.4%), and reported results in U/liter (89.4%). The percentage of laboratories reporting use of reaction temperatures of 30 and 37 degrees C was evenly divided, i.e., 42.7 and 42%, respectively. Analytical values reported by participating laboratories were categorized by reporting temperature, instrument, and method. Results were most consistent for a selected group of laboratories that supplemented optimized reaction solutions with pyridoxal 5-phosphate.