Isolation and characterization of human breast milk lipoamidase

The mean lipoamidase activity in human breast milk was found to be 0.073 nmol/min per mg (S.D. = 0.028, range = 0.020-0.123, n = 44). The mean lipoamidase activity is approximately 3-fold higher in milk than that in serum (0.023 nmol/min per mg, S.D. = 0.016, range = 0.001-0.059, n = 32). Lipoamidase was purified to 4400-fold by a four-step procedure from 330 ml of human breast milk. The purified enzyme was identified as a single band (Mr = 135,000) by sodium dodecyl sulfate/polyacrylamide electrophoresis. Analysis by Edman degradation indicated that the N-terminal amino acid was glycine. These results strongly suggest that milk lipoamidase is composed of a single polypeptide chain. The enzyme is considered to be a glycoprotein since it reacted positively to periodate-Schiff (PAS) staining. The isoelectric point of the enzyme was 4.2. After treatment of lipoamidase with sialidase, its position on isoelectric focusing gel moved from pH 4.2 to 4.6. This is strongly indicative that lipoamidase contains sialic acid residues. The optimum pH for the enzyme activity is 7.0. The Michaelis constant (KM) for lipoyl p-aminobenzoate is calculated as 25 microM. The enzyme activity was completely lost by heating 60 degrees C for 5 min. The effects of thiol-reactive agents, such as 2-mercaptoethanol (ME) and p-chloromercuribenzoate, were not significant. However, the enzyme activity was completely inhibited by 50 microM diisopropylfluorophosphate. Thus, this enzyme seemed to contain an essential serine residue in the active site.     

Purification of biotinidase from human plasma and its activity on biotinyl peptides

Biotinidase catalyzes the hydrolysis of N epsilon-biotinyllysine (biocytin) to form biotin and free lysine. The enzyme has been purified 4800-fold from outdated human plasma and was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis to have a molecular weight of (76 +/- 2) X 10(3). The same molecular weight was found by molecular sieve chromatography under nondenaturing conditions, indicating biotinidase is a monomer. This value is in contrast to a molecular weight of 115 000 determined by Pispa [Pispa, J. (1965) Ann. Med. Exp. Biol. Fenn., Suppl. 5, 5-39] with an impure biotinidase. The Km for biocytin was 6.2 X 10(-6) M, and biotinidase was found to be sensitive to phenylmethanesulfonamide and iodoacetamide in agreement with earlier studies by Knappe and co-workers [Knappe, J., Brümmer, W., & Bierderbick, K. (1963) Biochem. Z. 338, 599-613], who suggested that serine hydroxyl groups and sulfhydryl groups are essential for enzymatic activity. The specificity of biotinidase was examined by using synthetic and natural biotinyl peptides isolated by specific proteolytic cleavage of the biotinyl subunit of transcarboxylase. It was found that the rate of hydrolysis of biocytin was 83-fold higher than that found for biotin-containing peptides 5-13 residues in length. Removal of methionine from either side of the conserved region around the biocytin did not greatly alter the rate of cleavage. Increasing the peptide to 65-123 residues in length decreased the rate 1200-fold compared to that of biocytin.(ABSTRACT TRUNCATED AT 250 WORDS)     

Specificity of the trypanothione-dependent Leishmania major glyoxalase I: structure and biochemical comparison with the human enzyme

Trypanothione replaces glutathione in defence against cellular damage caused by oxidants, xenobiotics and methylglyoxal in the trypanosomatid parasites, which cause trypanosomiasis and leishmaniasis. In Leishmania major, the first step in methylglyoxal detoxification is performed by a trypanothione-dependent glyoxalase I (GLO1) containing a nickel cofactor; all other characterized eukaryotic glyoxalases use zinc. In kinetic studies L. major and human enzymes were active with methylglyoxal derivatives of several thiols, but showed opposite substrate selectivities: N1-glutathionylspermidine hemithioacetal is 40-fold better with L. major GLO1, whereas glutathione hemithioacetal is 300-fold better with human GLO1. Similarly, S-4-bromobenzylglutathionylspermidine is a 24-fold more potent linear competitive inhibitor of L. major than human GLO1 (Kis of 0.54 microM and 12.6 microM, respectively), whereas S-4-bromobenzylglutathione is >4000-fold more active against human than L. major GLO1 (Kis of 0.13 microM and >500 microM respectively). The crystal structure of L. major GLO1 reveals differences in active site architecture to both human GLO1 and the nickel-dependent Escherichia coli GLO1, including increased negative charge and hydrophobic character and truncation of a loop that may regulate catalysis in the human enzyme. These differences correlate with the differential binding of glutathione and trypanothione-based substrates, and thus offer a route to the rational design of L. major-specific GLO1 inhibitors.     

Effect of plasma biotinyl-peptides on biotinidase activity

Purified biotinidase (enriched 24,000-fold) from fresh human plasma exhibited reduced catalytic activity when incubated with heat-inactivated dialyzed plasma. The polypeptide fractions separated from the heat-inactivated dialyzed plasma using streptavidin-Sepharose resin showed the same effect on purified biotinidase. These inhibitory effects on biotinidase were partial (25-45%) rather than complete. The polypeptide fraction from streptavidin-Sepharose resin was analyzed by SDS-PAGE in the Laemmli system and by various types of HPLC. Analyses by ion-exchange and reversed-phase HPLC revealed the existence of three relatively small mol. wt polypeptides. Each of these peak fractions exhibited similar inhibitory effects on biotinidase activity. SDS-PAGE analysis indicated that the streptavidin affinity resin fraction was composed of four major polypeptides whose mol. wts were 120,000, 76,000, 53,000 and 27,000. The two bands of 120,000 and 76,000 corresponded to the mol. wts of the biotinyl subunit of pyruvate carboxylase, beta-methyl-crotonyl-CoA and/or propionyl-CoA carboxylase respectively. However, the polypeptides of mol. wts 53,000 and 27,000 were found to be two unique biotinyl-peptides present in human plasma. These bands on the gels were transblotted and exhibited a fluorescent activity after incubated with a FITC-avidin. These findings strongly suggest the existence of circulating plasma biotinyl-polypeptides as inhibitory factor(s) on human plasma biotinidase.     

Influence of detergents and organic solvent extraction on human gamma-glutamyltransferase activity.

Human serum and urinary gamma-glutamyltransferase (gamma-GT) have been found to react differently towards the detergents Triton X-100 and sodium lauryl sulphate (SDS). Serum gamma-GT was virtually unaffected by Triton X-100 at a concentration of 5% whereas urinary gamma-GT was 10-15% activated under similar conditions. There was a 100-fold difference in the response of serum and urinary gamma-GT to SDS. The enzyme activity in urine was completely destroyed by 0.02% SDS, whereas it required 2.0% to destroy the serum enzyme. These latter differences, however, were found to result from the environments of the serum and urinary enzyme rather than to intrinsic factors within the molecule. Extraction of serum gamma-GT from normal individuals with n-butanol resulted in little or no loss of activity from the aqueous phase, whereas more than 30% activity was lost from normal urines. Extraction using various mixtures of butanol and di-isopropyl ether (DIPE) produced largely similar results, except that 25% DIPE: 75% butanol mixture produced a marked loss of activity from serum. It is suggested that gamma-GT activity in human body fluids may be dependent on the presence of a lipid fraction.     

Purification and characterization of human serum biotinidase

Biotinidase has been purified from human serum to a specific activity of 1900 units/mg protein by a five-step procedure. After ammonium sulfate precipitation (33-55% cut) it was purified by DEAE-Sephacel, hydroxylapatite, octyl-Sepharose CL-4B, and Sephadex G-100 chromatography. The purified enzyme showed a single silver staining band with polyacrylamide gel electrophoresis under denaturing and non-denaturing conditions. Biotinidase is a glycoprotein. The sialic acid residues in the molecule are not required for enzyme activity. The Mr of human serum biotinidase estimated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Ferguson plot) and by sedimentation analysis was 68,000. Human serum biotinidase showed maximum activity in the pH range 6.0 to 7.5 with N-(d-biotinyl) p-aminobenzoate as substrate. However, with biocytin as substrate, the maximal activity of the enzyme was in the pH range 4.5 to 6.0. Using structural analogs of the substrate we have shown that biotinidase is not a general proteolytic enzyme and has specific structural requirements in the substrate for hydrolysis.     

Human serum biotinidase is a thiol-type enzyme

The effects of a disulfide reducing agent and sulfhydryl blocking agents on the biotinidase activity in human serum and on the purified biotinidase have been extensively studied by using a newly developed HPLC assay method. This HPLC method directly measures the product (p-aminobenzoate, PAB), and is not interfered with by sulfhydryl-reactive agents. Further, because the substrate solution of this HPLC assay method contains only substrate (biotin 4-amidobenzoate) and phosphate buffer, accurate studies on the effects of sulfhydryl blocking reagents on the purified enzyme could be performed. Biotinidase activities in human sera (n = 83) were always enhanced by 2-mercaptoethanol (ME). The optimum concentration was found to be 1 mM. The degree of activation was variable (100 to 400% of the original) depending on the serum sample. Sulfhydryl blocking reagents such as organic mercurials were tested on fresh serum and purified enzyme. Mercuric agents were found to inhibit the activity of fresh serum and purified enzyme at 0.05 and 0.005 mM, respectively. Sulfhydryl alkylating agents, N-ethylmaleimide (NEM) and dithiobis(2-nitro)benzoic acid (DTNB), inhibited 100 and 64% of the activity of the purified enzyme at 0.1 and 1.0 mM, respectively. However, lower concentrations (less than 5 nM) of organic mercurials and mercuric ion exhibited a slight enhancement (20-30%) of the activity of the purified enzyme. These results indicate the presence of an essential sulfhydryl residue at the active center. The enzyme contains 2.5 sulfhydryls per molecule, as determined by using Ellman's assay method. Serine protease inhibitors such as phenylmethylsulfonyl fluoride (PMSF) and diisopropylfluorophosphate (DFP) did not inhibit the enzyme activity at 0.05 mM or higher concentration.(ABSTRACT TRUNCATED AT 250 WORDS)     

Comparison of profound biotinidase deficiency in children ascertained clinically and by newborn screening using a simple method of accurately determining residual biotinidase activity.

We describe a method for more accurately determining residual biotinidase activity in sera of individuals with profound biotinidase deficiency. Using this method we found that there is a statistically significant difference in the means of residual serum enzyme activities of symptomatic children and those identified by newborn screening. A subgroup of children identified by screening have activities higher than any of the symptomatic population. These children may develop mild symptoms, may develop symptoms later in life, or may not develop symptoms at all.     

Studies on the degradation of human very low density lipoproteins by human milk lipoprotein lipase

Human milk lipoprotein lipase (LPL) was purified by heparin-Sepharose 4B affinity chromatography. The time required for the purification was approximately 2 h. The acetone-diethyl ether powder of milk cream was extracted by a 0.1% Triton X-100 buffer solution and the extract was applied to the heparin-Sepharose 4B column. The partially purified LPL eluted by heparin had a specific activity of 5120 units/mg which represented a 2500-fold purification of the enzyme. The LPL was found to be stable in the heparin solution for at least 2 days at 4 °C. This enzyme preparation was found to be free of the bile salt-activated lipase activity, esterase activity, and cholesterol esterase activity. The LPL had no demonstrable basal activity with emulsified triolein in the absence of a serum cofactor. The enzyme was activated by serum and by apolipoprotein C-II. The application of milk LPL to studies on the in vitro degradation of human very low density lipoproteins can result in a 90–97% triglyceride hydrolysis. The LPL degraded very low density lipoprotein triglyceride and phospholipid without any effect on cholesterol esters. Of the partial glycerides potentially generated by lipolysis with milk LPL, only monoglycerides were present in measurable amounts after 60 min of lipolysis. These results show that the partially purified human milk LPL with its high specific activity and ease of purification represents a very suitable enzyme preparation for studying the kinetics and reaction mechanisms involved in the lipolytic degradation of human triglyceride-rich lipoproteins.     

Human serum lipoamidase

Thirty-two human serum specimens were assayed for lipoamidase (lipoyl-4-aminobenzoate hydrolase) activity. All sera had lipoamidase activities. This substrate was newly synthesized by us and had a satisfying purity as evaluated by HPLC-fluorimetric detection. Product (p-aminobenzoate) liberated was determined directly by the HPLC-fluorimetric method. Liberation of the product was linearly continued for 6 h. The pH optimum of serum lipoamidase was found to be 7.0. The effect of substrate concentration on human serum lipoamidase activity was examined and the reaction was saturated at 0.1 mmol/l. The sera obtained were from individuals aged from 1 to 8 years. The mean value of serum lipoamidase activity was found to be 1.50 U/l (SD 1.037, range 0.04-3.75, n = 32). The difference of sex effects was analyzed and no significant difference was found (males: n = 14, mean 1.48, SD 1.162, range 0.04-3.75; females: n = 18, mean 1.52, SD 0.963, range 0.48-3.51) among this age group. Biotinidase activity was also determined in these 32 serum specimens and the correlation was examined. The mean biotinidase activity was 3.16 U/l (SD 2.567, range 0.35-9.37). The correlation coefficient (r) between lipoamidase activity and biotinidase activity was 0.8931. Although the physiological significance of lipoamidase has not been known, the enzyme might play an important role in recycling of lipoate as biotinidase does.     

Biotinidase in the porcine cerebrum

Biotinidase activities found in porcine brains (n = 3) were as follows: cerebrum, 4.4 +/- 0.2 pmol/min per milligram of protein; cerebellum, 7.6 +/- 0.3 pmol/min per milligram of protein; medulla, 2.9 +/- 0.3 pmol/min per milligram of protein. These values are relatively high compared with the activities in rat or guinea pig brains. Subcellular distribution of biotinidase was found mainly in the soluble cytoplasmic fraction (S3), i.e., in the supernatant of 0.32 M sucrose S2 solution after ultracentrifugation at 105,000g for 90 min. This is in contrast to the guinea pig livers, in which the subcellular distribution of biotinidase is mainly found in the microsomal fraction. After a seven-step purification (22,200-fold enrichment), porcine brain biotinidase is identified as a single polypeptide by the sodium dodecyl sulfate-polyacrylamide gel electrophoresis system, and its molecular weight is determined as 68,000 Da. The isoelectric point of the enzyme was 4.3. Sialidase treatment strongly suggests the presence of sialyl residues in this enzyme. Amino acid analysis indicates relatively high hydrophilicity and high content of glycine and serine. The enzyme activity is inhibited by organic mercurials, but not by diisopropylfluorophosphate. Abundant soluble biotinidase in brain cytoplasm may play an important role which has not been discovered yet.     

Determination of specific activities and kinetic constants of biotinidase and lipoamidase in LEW rat and Lactobacillus casei (Shirota)

Enzyme kinetic parameters, such as K(m), V(max) (or V), k(cat)/K(m), and K(i) (by biotin or lipoic acid) for biotinidase and lipoamidase were determined in Lewis (LEW) rat and Lactobacillus casei (Shirota) using fluorimetric high-performance liquid chromatography (HPLC). It was found that the final protein concentration below 0.1mg/ml is sufficient to obtain linear hydrolytic reaction and to determine the Michaelis-Menten type kinetic parameters (K(m), V, K(i)). We applied this HPLC enzyme assay method onto the rat and some bacteria. The highest specific activities (Vs) for biotinidase were found in Lactobacillus casei (Shirota) and rat kidney. It was also found that the largest K(i) by product for biotinidase and lipoamidase were present in the Lactobacillus casei (Shirota). There has been found specie (between rat and mouse) differences and tissue (organ) differences, together with tissue region differences and sex differences in some tissues. Summary of the distributions of both enzymes in LEW rat was also presented. Therefore, this HPLC determination method for the enzyme kinetic parameters in tissues is expected to be an indispensable tool for the investigation of the various diseases in humans.     

Partial biotinidase deficiency is usually due to the D444H mutation in the biotinidase gene

Newborn screening for biotinidase deficiency has identified children with profound biotinidase deficiency (<10% of mean normal serum activity) and those with partial biotinidase deficiency (10%–30% of mean normal serum activity). Children with partial biotinidase deficiency and who are not treated with biotin do not usually exhibit symptoms unless they are stressed (i.e., prolonged infection). We found that 18 of 19 randomly selected individuals with partial deficiency have the transversion missense mutation G1330>C, which substitutes a histidine for aspartic acid444 (D444H) in one allele of the biotinidase gene. We have previously estimated that the D444H mutation results in 48% of normal enzyme activity for that allele and occurs with an estimated frequency of 0.039 in the general population. The D444H mutation in biotinidase deficiency is similar to the Duarte variant in galactosemia. The D444H mutation in one allele in combination with a mutation for profound deficiency in the other allele is the common cause of partial biotinidase deficiency. Received: 8 December 1997 / Accepted: 22 January 1998     

Two-domain structure of alanyl-tRNA synthetase from Bombyx mori: isolation of the N-terminal catalytic domain

Alanyl-tRNA synthetase of 115K daltons from Bombyx mori was cleaved into two fragments of 62K and 47K daltons by trypsin. The 47K fragment was active in aminoacylation of tRNA, whereas the 62K fragment was inactive. The 47K and 62K fragments were found to be located at the N- and C-terminal ends, respectively, in the intact enzyme. The intact enzyme was protected from trypsin-attack by the cognate tRNA. The Km value of the 47K fragment for tRNA was 22 microM which is about 16-fold higher than that for the intact enzyme (1.4 microM). The molecular activities of the fragment and the intact enzyme were 2.2 s-1 and 16.8 s-1, respectively. This indicates that the 62K domain enhances affinity for tRNA and it is responsible for the full activity of tRNA aminoacylation. These results do not support the "covalently linked dimer" hypothesis, but indicate that the alanyl-tRNA synthetase is a functional monomer consisting two large domains.     

Maternal and neonatal somatomedin C/insulin-like growth factor-I (IGF-I) and IGF binding proteins during early lactation in the pig

RIA for insulin-like growth factor-I (IGF-I) was performed on Tris-neutralized, acid-ethanol extracts of porcine, bovine, ovine and human mammary secretions, and porcine maternal and neonatal sera. High levels (50-500 ng/ml) of immunoreactive IGF-I were present in the colostrum of all three animal species. IGF-I was also identified in porcine milk, though at levels 10- to 100-fold reduced relative to that in colostrum. Maternal (pig) sera was characterized by IGF-I concentrations intermediate between that in colostrum and that in milk. IGF-I levels were relatively low in serum of newborn pigs and exhibited an approximately 1.4-fold increase between Days 3 and 7 of postnatal life. Fractionation of pig colostrum in nondenaturing, gel-filtration columns demonstrated association of endogenous IGF-I with two prominent binding proteins (Mr's of 150,000 and 50,000 for the complexes). A third immunoreactive component was also observed to elute in the column void volume fractions (Mr greater than 158,000). The 150,000 and 50,000 Mr complexes were also present in serum obtained from sows at term. In contrast, IGF-I immunoreactivity in porcine milk was localized exclusively in the 150,000 Mr complex. Incubation of porcine colostrum and milk with 125I-IGF-I revealed a prominent, unoccupied IGF binding protein corresponding to that of the 150,000 Mr complex, whereas serum obtained from sows at term displayed both the 150,000 and 50,000 Mr unoccupied forms. Fractionation of (pooled) serum obtained from porcine neonates immediately at birth revealed a heterogeneous pattern of IGF-I immunoreactivity which included both the 150,000 and 50,000 Mr forms. Qualitative differences in this chromatographic pattern were apparent in serum at 6 hr postnatal and after ingestion of colostrum had occurred. The unoccupied IGF binding proteins in newborn pig serum were solely of the small size class. These results demonstrate that mammary secretion of IGF-I and its binding proteins are temporally regulated during the period immediately surrounding parturition. Physiologic alterations in the serum IGF-I profile during early postnatal life may reflect in part the uptake and/or response of the neonate to maternal IGF-I.     

Characterization of Quinolinic Acid Phosphoribosyltransferase in Human Blood and Observations in Huntington''s Disease

Quinolinic acid (QUIN), an excitotoxic compound present in the mammalian CNS and periphery, has been hypothetically linked to human neurodegenerative disorders such as Huntington's disease and epilepsy. Quinolinic acid phosphoribosyltransferase (QPRT), the catabolic enzyme of QUIN, is found in the CNS and peripheral organs where it may be a major influence on the tissue levels of QUIN. We have measured QPRT activity in human blood as a means of assessing one aspect of QUIN metabolism in humans. The enzyme was present in blood cells, platelets having a sixfold greater activity than erythrocytes, but was essentially absent from the plasma. In a blood cell fraction, enzyme activity was potently inhibited by phthalic acid (IC50 = 6.1 microM). Kinetic analyses conducted over a range of QUIN concentrations yielded Km values of 1.89-3.75 microM and Vmax values of 33.4-72.5 fmol nicotinic acid mononucleotide/h/mg protein. Enzyme activity varied 2.2-fold between normal individuals, was reasonably constant over a series of sampling intervals, and showed some diminution when blood was stored for 1 month at -20 degrees C. No differences of enzyme activity in erythrocytes or platelets were apparent between three Huntington's disease patients and their unaffected spouses. These data indicate that measurements of QPRT activities in blood are a convenient means to monitor QUIN metabolism in human subjects and that a deficiency of the enzyme is not apparent in Huntington's disease.     

Role of human serum biotinidase as biotin-binding protein.

Biotinidase shows two binding sites for biotin, with Kd = 59 and 3 nM respectively, and requires tryptophan and cysteine residues of the biotinidase protein for biotin-binding activity. Analysis of human serum by various column-chromatographic techniques indicates that biotinidase is the only protein which exchanges with labelled (+)-biotin. It was shown previously that epileptic patients receiving a high average dose of anticonvulsants (containing a carbamide group) have lower biotin concentrations than those receiving a low dose. We have shown in human serum and with purified biotinidase that these anticonvulsant drugs compete with biotin for binding to the protein moiety.     

n-acetyl-beta-hexosaminidase activity in human breast milk

1. The lysosomal enzyme, N-acetyl-beta-hexosaminidase (HEX) is present in human breast milk. It is composed predominantly of "A" (heat-labile) and "B" (heat-stable) isozymes which coelute with the corresponding major serum isozymes on DE-52 ion-exchange chromatography. 2. Total HEX activity in "early" milk obtained at 2.8 +/- 1.4 weeks post partum, is approx. 2.5-fold higher (87 +/- 29 nmol/60'/mg protein. n = 10) than that of pregnancy serum (35.7 nmol/60'/mg protein) prior to delivery. 3. These levels increase to greater than 3-fold (110 +/- 20 nmol/60'/mg protein, n = 13) as the milk matures (10.3 +/- 4.2 weeks). 4. The specific activity of HEX A in the milk changes little with time post partum, because absolute 5. In contrast, HEX B specific activity is increased, as absolute levels (per volume) remain constant in the face of decreasing milk protein content, 5. These changes result in a high degree of correlation (r = 0.81) between time of lactation and % HEX A observed.     

Gammaherpesviruses encode functional dihydrofolate reductase activity

We overexpressed and purified from Escherichia coli the dihydrofolate reductase (DHFR) of the gammaherpesviruses human herpesvirus 8 (HHV-8), herpesvirus saimiri (HVS), and rhesus rhadinovirus (RRV). All three enzymes proved catalytically active. The K(m) value of HHV-8 DHFR for dihydrofolate (DHF) was 2.02+/-0.44 microM, that of HVS DHFR was 4.31+/-0.56 microM, and that of RRV DHFR is 7.09+/-0.11 microM. These values are approximately 5-15-fold higher than the K(m) value reported for the human DHFR. The K(m) value of HHV-8 DHFR for NADPH was 1.31+/-0.23 microM, that of HVS DHFR was 3.78+/-0.61 microM, and that of RRV DHFR was 7.47+/-0.59 microM. These values are similar or slightly higher than the corresponding K(m) value of the human enzyme. Methotrexate, aminopterin, trimethoprim, pyrimethamine, and N(alpha)-(4-amino-4-deoxypteroyl)-N(delta)-hemiphthaloyl-L-ornithine (PT523), all well-known folate antagonists, inhibited the DHFR activity of the three gammaherpesviruses competitively with respect to DHF but proved markedly less inhibitory to the viral than towards the human enzyme.     

Protein-O-carboxylmethyltransferase from cytosol and membranes of chicken erythrocytes

Cytosolic protein-O-carboxylmethyltransferase was purified more than 4,000-fold in specific activity and membrane-associated protein-O-carboxylmethyltransferase carboxymethylase about 900-fold from chicken erythrocytes by use of a combination of affinity chromatography on immobilized S-adenosyl-L-homocysteine and gel filtration on Sephacryl S-200 (Pharmacia), together with 3-((3-cholamidopropyl)-dimethylammonio)-1-propane-sulfonate as a detergent to solubilize the membrane-associated enzyme. The two enzymes were characterized by examining the dependence of their activity on pH and on concentration of S-adenosyl-L-methionine using fetuin as an exogenous methyl-acceptor substrate, and were found to differ somewhat. The cytosolic enzyme had a pH optimum of 6.0 with an apparent Km value of 2.1 microM for S-adenosyl-L-methionine, whereas corresponding values for the membrane-associated enzyme were 6.5 and 0.71 microM. This report deals with the biochemical differences between purified cytosolic and membrane-associated protein carboxymethylase from the same cell source.