Purification and properties of rat liver microsomal esterases.

Abstract Several rat liver microsomal esterases have been isolated. One of these was purified to homogeneity and its physical and catalytic properties were investigated. The enzyme is composed of two subunits of equal molecular weight. Each 70,000 molecular weight subunit appears to have an active site. Evidence is presented which suggests that catalysis involves the formation of an acyl-enzyme intermediate and requires the presence of an unprotonated residue with a pK of about 6.0.

SUMMARY Several rat liver microsomal esterases have been isolated. One of these was purified to homogeneity and its physical and catalytic properties were investigated. The enzyme is composed of two subunits of equal molecular weight.
Each 70,000 molecular weight subunit appears to have an active site. Evidence is presented which suggests that catalysis involves the formation of an acyl-enzyme intermediate and requires the presence of an unprotonated residue with a pK of about 6.0.
This report describes a study of the physical and catalytic properties of a purified rat liver microsomal &erase.
The purpose of this investigation was to provide a well characterized elizyme ~vith which to study the fluoride inhibition of an enzyme which does not require metal ions for catalysis.
X subsequent report (1) describes the fluoride inhibition of this enzyme.
The physical aucl catalytic properties are in general similar to those of the Iv-e11 characterized esterases from beef and pig liver.
The multiplicity of carbosyl estcrases in mammalian tissues has been lye11 cstablishcd (2). In liver, esterasc activity is found primarily in the microsomal fraction (2). There have been several recent reports regarding attempts to purify rat liver csterases. Those of Arnclt and Iirisch (3) and .+\kao and Omura (4) appeared after our purification method had been established, but during an investigation of catalytic properties and fluoride inhibition of the purified esterase. Hyasc and Tappcl (5) employed a mechanical disruption technique to solubilizc a microsomal esterase which was fractiouated further by precipitation and chrornatographic procedures.
Their investigation was primarily concerned with substrate specificity, and little data were preseiitcd regarding the homogeneity of the preparation or its physical and catalytic propertics.
Ljungquist and Augustinsson 0 To whom inquiries and requests for reprints should be addressed.
i\t least one of thpsc was honiogci~cous, but was cliaractcrizctl principally by its cliromatograpliic properties and molecular \vcight, about 170,000. 1Iost recently, ,1kao and Omura (4) rcportcd tlic purification of an acetaliilitle-li3clrolyzillg enzyme from rat liver that had a subunit molecular weight of 60,000 to 65,000.
The pliysiological role of tlw microsomnl esterascs is not understood because they arc active toward carbosylic; acid cstcrs of short chain fatty acids (2). In vitw, some liver and adipose tissue c&erases have been shown to be intcrcouvertiblc with lipasc activity (7-11) i.e. active toward esters of long chain fatty acids, but their role as lipolytic enzymes has not been &mow strated in viva (2). Rather than lipolytic activity, the cstcrascs may function primarily in the metabolism of foreign complllds (2). BIATERIALS AN11 hlETHOl)S 12eage,lls-Proteins used for standardization and reagents for esterase and protein stains were obtained from Sigma Chemical Co. Ethyl butyrate and reagents for acrylamide gels were products of I':astman. Phenyl bntyrate and diethyl p-nitrophenyl phosphate were obtained from K and K Laboratories and Baker, respectively.
Other reagents were the best grade available and were used without further purification.
Solutions were prepared in distilled water which had been passed through a Barnstead deionixer.
Il;thyl butyrate was distilled before use, and a-naphthyl acetate was recrystallized from ethanol. Phenyl butyrate was applied to a silicic acid column in hexane and eluted with 7.5yo (v/v) ether in hexane to remove free phenol. No phenol was detected by spectral analysis of the purified ester, and complete hydrolysis indicated at least 98% purity.
ProleijL Assay-Protein concentrations were routinely determined by a modification (12) of the Lowry method. The biLlret method was used in determination of an extinction of cocflicient (& = 12.8) used to calculate kinetic constants and the eqklivalent weight.
Results from the two colorimctric methods differed by about 10%. The protein standard was crystalline bovine serum albumin. the column was washed with 2 to 3 volumes of starting buffer before a linear NaCt gradient was applied using starting buffer and 0.3 1% NaCl in buffer. Fractions, 3 ml, containing esterase activity were pooled, concentrated, and stored at, 4". For esterase B, an identical gradient was applied after protein ceased, to be eluted by the starting buffer. Fractions containing the two esterase fractions (Fig. 3) were separately combined, concentrated, and stored at 4".

RESULTS
ZSnzynze PurificationPRcsults of the purification described under "JIaterials and h3ethods" arc presented in Table I. About 70% of the csterase activity was recovered in the washed microsomal fraction.
If the pellet from the initial centrifugation of the homogenate was not washed as described under "Materials and Methods," the recovery was ouly 45%. The same observation was made during preliminary investigations when p-nitrophenyl propionate was used as the substrate. Triton X-100 effectively solubilizcd the rsterasc activity at $3 7.4. Subsequent adjustment of the pH to 5.0 allowed insoluble material to be removed by low speed ceutrifugation.
The acetone treatment of the solubilized protein served to remove the detergent, concentrate the protein, and provide additional purification.
Rcprescntative results of hydrosylapatitc column chromatography of the solubilized esterase are shown in Fig. 1. The major and minor enzyme fractions were designated esterase A and esterase 13, respectively.
The relative amounts of the two fractious were similar for each of several trials of this method. Fig. 2 shows the results of I)EAE~Sephades column chromatography of estcrase A. A single major protein fraction was eluted which contained all of the csterase activity.
Although significant additional purification was not achieved, this step removed small amouuts of contaminating proteins (including red pigments and esterase 1%) and demonstrated the purity of the esterasc A fraction.
When the esterase B fraction from the hydrosylapatite step was ~hromatographcd on I>l<AE-Scphadcs columns, two esterase fractions were resolved (Fig. 3). The first csterasc fraction, estcrase I$, was the subject of further investigation, described below.
Esterase Is1 was elutcd by the startiug buffer, but the second fraction, esterase HZ, was cluted by the NaCl gradient. Upon rechromatography of csterase El it was eluted as initially observed and no other esterase was eluted by the gradient.
The second enzyme fraction, 1s2, was uot investigated in detail, however, its elution properties and clcctrophoretic mobility were  1 (lefl). Hydroxylapatite column chromatography. Soluble protein (130 mg, 28 ml) was applied to and eluted from the column as described under "Materials and Methods." The enzyme concentration is indicated by the opera circles (O-O) and the absorbance due to protein by the closed circles (0-O). The solid line indicates the phosphate concentration in the fractions. Esterase A (Fractions 102 to 112) and esterase B (Fractions 75 to 86) were separately pooled, concentrated, and dialyzed. FIG. 2 (center). DEAE-Sephadex column chromatography of esterase A. Esterase A (20 mg) from hydroxylapatite chromatography was applied to the column and eluted as described under similar to that for c&erase A. Investigation of the fluoride inhibition of the &erase 13 fraction before and after resolution on DEAE-Sephades also suggested that the second fraction (1%) was esterase A. This purification step was valuable for the resolution of the two esterase 13 subfractions, but sometimes resulted in the loss of up to 40% of the total esterase activity.
Both esterase A and esterase IL were stable for several months when stored at 2". Esterase A precipitated slightly during storage at a concentration of 2 mg per ml, but no decrease in specific activity was detected.
Purity and Netal Ion Content of Esterase A--Disc gel electrophoresis of esterase A in the Tris-glycine system (Fig. 4) indicated a high degree of purity.
Electrophoresis in the presence of sodium dodecyl sulfate (Fig. 5) also demonstrated the purity of the preparation.
A very minor band was detected in this system with a mobility corresponding to a molecular weight twice that of the principal band. When 40 pg of protein was treated with alkaline urea and electrophoresis was carried out at pH 2.7 in the presence of urea (16), one major and two to four very faint minor bands were detected (not shown) by the sensitive Coomassie blue stain. The AN,:As=c, ratio of esterase A solutions was always 1.70 to 1.78. Values of Azso:Ablo were always greater than 80 and were as high as 250 which indicates a reasonable separation from heme containing proteins (17).
Neutron activation analysis of 15 mg of exhaustively dialyzed esterase A did not reveal the presence of any significant quantity of metal ions. Both EDTA and 1 ,lO-phenanthroline had no effect on enzyme activity at pH 7 to 7.5 either during extensive dialysis or during catalysis. This observation was important since we wished to investigate fluoride inhibition of an enzyme that did not contain nor require the presence of added metal ions. Molecular and Equivalent Weights of Esterase A-The molecular weight was estimated to be 135,000 to 140,000 by gel filtration on Sephadex G-200 as shown in Fig. 6. The subunit molecylar weight, 66,000 to 72,000, was determined by disc gel electrophoresis in the presence of sodium dodecyl sulfate (Fig. 7). The equivalent weight, about 73,000, was estimated from the enzymedependent amount of p-nitrophenol released from diethyl p-nitrophenyl phosphate (19, 20). The reaction at pH 8.0, 25", was complete in 10 to 15 s and did not proceed further during the The relative values of the molecular, subunit, and equivalent weights are 2 : 1: 1, respectively. The molecular and equivalent weights of esterase A are summarized in Table II.  (Fig. 6), and the second value was determined from the slope of the calibration curve and the observed elution volume relative to the serum albumin dimer.
The error indicated is based on the probable uncertainty of the calibration curve. *The first value was determined from the calibration curve (Fig. 7), and the second value from the slope of the calibration curve and the observed mobility relative to serum albumin. Esterase A was added to a pa-stat reaction vessel containing phenyl butyrate (1.8 mM), KNOI (50 mM), and various concentrations of methanol in a total volume of 11.2 ml at pH 8.3,23". After 0.5 min, a 3.0-ml aliquot was transferred to a cuvette containing 0.2 ml of Tris chloride (final buffer concentration 50 mM, pH 8.3,23").
The hydrolysis rate (pH-stat) was measured simultaneously with the total reaction rate (spectrophotometer, 270 nm), and appropriate corrections were applied for the changes in volume.
The methanolysis rate was calculated by subtracting the hydrolysis rate from the total reaction rate. FIG. 9 ( ysis of phenyl butyrate hydrolysis did not reveal any deviation from I\Iichaelis-i\lIerlteil kinetics when the substrate conccntration was as high as 2.2 rnbr. The hydrolysis of ethyl butyrate was accompanied by substrate inhibition at concentrations greater than 0.6 rnhf, (Fig 10). Extrapolation of the linear portion of the curve indicated that V,,,, was 770 units per mg and Zi, was 0.9 m&f.
Properties of Z&erase &--Elcctrophorcsis of the esterase I& fraction in the Tris-glycine system indicated that the fraction contained two or three electrophoretic species (Fig. 11) which were distinct from &erase A. Each protein band had esterase activity, but the most intcnsc cstcrase stain was associated with the slowest minor protein compoiicnt.
Elcctrophorcsis of esterase 13, preparations in the presence of sodium dodecyl sulfate revealed a single protein band with the same mobility as &erase A (Fig. 5). Thus, in the presence of sodium dodrcyl sulfate both esterasc A and the csterase 131 proteins had the same electrophoretic mobility, but were distinguished in the nondcnaturing Tris-glycine system. Thercforc, the subunit molecular weight of esterase A and the esterase 11, proteins must be similar. Estimates of the molecular weight of esterabe IsI by gel filtration were inconclusive.
An analysis of the pH dependence of the enzyme activity of the esterasc 131 fraction revealed a more complicated relationship than observed for esterase A. The prescncc of morc than one esterasc species in the fraction probably contributed to the complexity of the pII dcpcndcncc.
The esterase 13, fraction could be distinguished from csterase A by the observation that it did not appreciably catalyze methanolysis of phenyl butyrate, and 2721 ethyl butyrate hydrolysis was not accompanied by substrate inhibition even at 9 mM ethyl butyrate (Zi, = 15 m&f).

I)ISCUSSION
This report describes some definitive physical and catalytic properties of a purified rat liver microsomal csterase. Attempts to isolate distinct esterase species from rat liver have been made only in recent years and sufficient) descriptive data are not available t,o allow conclusive comparison of t,lre various enzymes isolated and assayed by a variety of techniques.
The esterases described by Akao and Omura (4) and Ljmiquist and August,insson (6) are similar in some respects to the estcrases described in this report.
The microsomal fraction of liver homogenates was found to contain 70% of the esterase activity.
The smaller recoveries of enzyme in this fract,ion reported by ot,hers (6, 21) may have resulted from lack of washing, or incomplete washing of the initial low speed pellet (22). When microsomal suspensions containing Triton X-100 (0.35 mg per ml) were centrifuged at 105,000 x g for GO min, 80% of the csterasc activity and 20% of the protein were recovered in the supernatant fraction.
Similarly, Akao and Omura (4) found t,hat rat liver microsomal esterase was readily solubilized by nonionic detergents, but not by sodium chloride solutions.
The est,erase A preparat,ion appeared to be essentially homogeneous when protein stains were used in three elcctrophoretic systems, including two that involve prot,ein denaturation.
However, in the Tris-glycine system, the very sensitive esterase stain sometimes revealed several very minor bands with both greater and less mobility than the principal band. Some of these minor bands had the same mobility as that of the two to three esterase species in the esterase 131 preparat,ion.
The high degree of purit,y of esterase A was also indicated by its elution from DEAE-Sephades with essentially constant specific activity. In the catalytic studies, the esterase concentrations were in the range 0.02 to 0.2 pg per ml so that any very minor esterase contaminants woulcl be espected to have no significant effect on the observed properties of &erase A.
The purified carbosylesterases from pig and beef liver are reversibly dissociable into active subunits (2). Most analyses have indicated a dimeric structure, but recent work by  and Aune (24) has suggested that the native beef and pig liver esterases may be trimeric structures.
The results presented here for csterase A from rat liver parallel the earlier reports for the pig and beef liver estcrases. Cert'ainly more precise techniques'could be applied, but the data presented here support the conclusion that t'he enzyme consists of two subunits of equal molecular weight, and t,hat t)he native enzyme contains two active sites. Although each subunit probably has an active site, no unequivocal evidence is available to support this conclusion.
It has been well established that the catalytic mechanism of the pig and beef liver &erases involves the format,ion of an enzyme bound acyl-serine intermediate (2). Esterase A react's stoirhiometrical1y wit,li diethyl p-nitrophenyl phosphate and it readily catalyzes the methanolysis of phenyl but'yratc.
These two lines of evidence support the conclusion t'hat esterase A is a serinc hydrolasc.
If t,he criteria described by Greenzaid and Jcncks (25) arc applied to the data describing mcthanolysis catalyzed by esterase A (Fig. S), the results indicate t,hat in the absence of methanol, acylation and dcacylation occur at comparable rates. If dcacylation wcrc t,hc rate limiting step in catalysis, the V,,,, for et'hyl-and phenyl but)yrate hydrolysis should be the same. The Vnlax for these two substrates at 25" was 770 and 440 units per mg, respccntivclg, whicah suggests that cleacylation is not entirely rate limiting and supports the cotlelusion tnade from the data rcgartlitrg mrtltattolysis. TllCSC results, however, are not considcrctl prccisc enough to bc the basis for any definite conclusions rcgartling rate litniting reaction because measurements of V,,,,, for ethyl butgrate hydrolysis were made at substrate concentrations no more than 0.6 I<, due to sub&ate inhibition.
Although the role of a serine rcsiduc at the active site of pig and beef liver esterases has been established, only very recently has evidence been reported that suggests that, a hist,idittc residue also participates in catalysis (26). The data presented here clearly indicate that the catalytic: activity for the rat liver enzytnc depends on the presence of an uttprotonatrd residue with pK about 6.0 at 25". The ionization of this residue also critically affects the fluoride inhibition of the esterasc. In association with an investigation (1) of the fluoride inhibition, it was found that the heat. of ionization of the rcsiduc is about 8.0 Cal per mole. The pK and heat of ionization of this critical residue suggest that it is hi&dine.
Although the data arc not definitive, this represents the first attempt to csplain the pH dependence of rat liver esterase activity, and is an approach not previously reported for the well characterized beef and pig liver enzymes.
Disc gel electrophoresis of the estcrase 13, fraction in the Trisglycine system revealed the presence of two or three principal proteins, each with e&erase activity.
It was usually noted that the most intense esterase stain corresponded to one of the least abundant proteins.
When the satne esterase 1st fraction was subjected to disc gel electrophoresis in the presence of sodium dodecyl sulfate, a single protein band was observed which comigrated with the esterase A subunit.
Although the data may suggest that esterase A and the esterase 131 fraction are related as dimer and monomer, respectively, no direct evidence is available to support this conclusion.
In fact their catalytic properties are quite distinct.
Estcrasc 131 did not readily catalyze the methanolysis of phenyl butyrate, it did not undergo substrate inhibition by ethyl butyrate, and its inhibition by fluoride was not time dependent (1).