132
Biochimica et Biophysics Acta, 398 (1975) 132-148 0 Elsevier Scientific Publishing Company, Amsterdam
- Printed
in The Netherlands
BBA 56621
SUBCELLULAR FRACTIONATION, PARTIAL PURIFICATION AND CHARACTERIZATION OF NEUTRAL TRIACYLGLYCEROL LIPASE FROM PIG LIVER
JOANNA
H. LEDFORD*
and PETAR
ALAUPOVIC
Lipoprotein Laboratory, Oklahoma Medical Research Foundation and Department of Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center, Oklahoma City, Okla. 73104 (U.S.A.) (Received
March 17th,
1975)
Summary
The subcellular distributions of acidic (pW 4.5) and neutral (pH 7.5) longchain t~acylglycerol lipases (glycerol ester hydrolase, EC 3.1.1.3) of pig liver have been determined. The distribution of the acidic lipase closely paralleled that of the lysosomal marker enzyme, cathepsin D. Approx. 60% of the neutral lipolytic activity resided in the soluble fraction; the distribution of this activity failed to parallel that of marker enzymes for mitochondria, lysosomes, microsomes, or plasma membranes. A method has been developed for purification of the neutral lipase from the soluble fraction by ultracentrifugation. An approximate go-fold purification was achieved, with recovery of 16% of the initial activity. The partially purified neutral lipase exhibited a pH optimum between 7.25 and 7.5. It required 30 m&I emulsified triolein for optimal activity and ceased to liberate fatty acids after 30 min of incubation. The enzymatic activity was destroyed by heating at 60°C. Neutral lipase was inhibited by sodium deoxycholate, Triton X-100 and iodoacetamide. The activity was not inhibited by sodium taurocholate, EDTA, heparin and diethyl-p-nitrophenyl phosphate. Neutral lipase failed to exhibit activity in assay systems specific for lipoprotein lipase, monoolein hydrolase, tributyrinase, and methyl butyrate esterase and showed little or no capacity to hydrolyze chyle chylomicrons or plasma very low density lipoproteins. It is suggested that the function of neutral lipase may be to supply the liver with fatty acids liberated from endogenously synthesized or stored triacylglycerols.
* Present address: Courtauld Institute London WlP 5PB. U.K.
of Biochemistry,
The Middlesex Hospital Medical School.
133
Introduction The ability of mammalian liver to hydrolyze simple organic esters was recognized at the beginning of the century [1,2] , However, the existence of a liver enzyme capable of hydrolyzing long-chain triacylglycerols was not postulated until 1965 [3-71. It has now been established that rat liver contains at least two triacylglycerol lipases. One of these enzymes possesses an acidic pH optimum and is localized in lysosomes [ 8-16 ] , The second has a neutral or alkaline pH optimum and its localization has been variously reported as soluble 161, microsomal [11,17,18] or mitochondrial [19J. Recent evidence also suggests the possib~ity of a hep~n-~nsitive lipase in plasma membranes [ll,lS] . In 1970 Multer and Alaupovic [20] noted that pig liver also contained acidic [pH 4.51 and neutral (pH 7.5) triacylglycerol lipases. The acidic enzyme was clearly lysosomal in origin while the neutral lipase showed highest specific activity in microsomes. The present communication presents in more detail the subcellular distributions of pig liver triacylglycerol lipases, describes a procedure for partial purification of the neutral lipase and reports the characteristics of the partially purified enzyme.
Materials and Methods
Substrates employed in various enzymatic de~rminations included olive oil (Sargent-Welch, Dallas, Texas, U.S.A.), triolein, sodium succinate, adenosine 5’-monophosphate and glucose 6-phosphate (Sigma, St. Louis, MO., U.S.A.), monoolein, tributyrin and methyl butyrate (Analabs, North Haven, Conn., U.S.A.), iodonitrotetrazolium violet and denatured hemoglobin (Nutritional Biochemicals, Cleveland, Ohio, U.S.A.). Standards for analysis of protein content and reaction products were human albumin (Hoechst-Behringwerke, Marburg/Lahn, G.F.R.), palmitic acid (Sigma), iodonitrotetrazolium violet formazan (Nutritional Biochemicals), L-leucine (Mann, New York, N.Y., U.S.A.) and potassium dihydrogen phosphate (J.T. Baker, Phillipsburg, NJ., U.S.A.). Other compounds utilized during the isolation or assay of liver lipases included sodium taurocholate (Maybridge Research Chemicals, Launceston, Cornwall, U.K.), Triton X-100 (Calbiochem, LaJolla, Calif., U.S.A.), and thymol blue and Nile blue A indicators (Allied Chemical, Morristown, N.J., U.S.A.). Chemicals tested for inhibitory action on lipolysis were sodium heparin (as solution, 10 000 U/ml, Upjohn, Kalamazoo, Mich., U.S.A.), EDTA (Calbiochem), sodium deoxycholate (Mann, New York, N.Y., U.S.A.), diethyl-pnitrophenyl phosphate or E-600 (K & K, Plainview, N.Y., U.S.A.), and iodoacetamide (Sigma). Purificationof oliueoil
T~a~ylglycerols were isolated from olive oil by column ~hromato~phy over silica gel G according to a modification of &he method of Crider et al. 1211. The purified preparation was a clear, nearly colorless liquid, containing only triacylglycerol by thin-layer chromatography.
134
~amoge~~~~~o~. Pig livers, from animals of either sex, were obtained fresh from a local meat-packing firm and were held on ice for less than 1 h before use. Fresh livers were either processed imm~iately or divided into pieces of approximately 100 g each and frozen at -12°C. Frozen tissue was used within a l-month period. Fresh or thawed tissue, always obtained from several different lobes of the same liver, was chopped into pieces of about 1 cm3 size. 25 g of chopped tissue were mixed with 200 ml of ice-cold 0.25 M sucrose containing 0.001 M d&odium-EDTA, pH 7.2. The mixture was homogenized in a Sorvail Omnimixer (Ivan Sorvall Co., Newton, Corm., U.S.A.) for 30 s at setting 3 (gentle bomogeniz~tion) or for 30 s at setting 3 and 1 min at setting 10 (ha&r homo~e~~~ation}. Homo~ena~s were filtered through 8 layers of coarse cheesecloth into a chilled confiner, and the pH was adjusted to 7.2 with 5 M KOH. S~~ce~~~~~~~r~e~~o~~~~o~.~~rnoge~a~$ prepared from 50 g of fresh tissue by gentle homogenization were fractionated essentially according to the method of de Duve et al. [ 221 and Appelmans et al, [23]. The procedure is summarized in Fig. 1. Centrifugations were carried out near 0°C in, either a refrigerated Servall cen~ifuge (Ivan Sorvall Co,) with GSA rotor or a Spinco Model L 2-653 ultracentrifuge (Beckman Co., Palo Alto, Calif., U.S.A.) with Type 42 rotor. Final pellets were suspended in approximately 50 ml of sucrose/EDTA solution, and the suspensions were frozen immediately in small aliquots until anatyzed. F~rifi~~~o~ of ~e~~~~~~~~~~~~~yce~o~ lipme. The isolation procedure is summ~ized in Fig. 2. For each experiment a homogenate was prepared from 50 g of fresh or frozen pig liver by harsh homogenization.
Fig.
I,
Subcellular
fractionatwn
of Pig liver
homo$enate.
135 Homo
enate
7-
Centrifugatlo”
30
min,
113,000 x g,
0-c
Fig. 2. Purification of pig liver neutral lipase.
An artificial lipid emulsion was prepared by sonication of a mixture of 3.32 g of unpurified olive oil and 21.3 ml of 12 mM sodium taurocholate solution for two 1-min periods at setting 5 on a Bronwill Biosonik II (Bronwill Scientific, Rochester, N.Y ., U.S.A.) equipped with catenoidal probe. Assuming a molecular weight of triolein for the olive oil, the final concentration of triacylglycerol in the prepared emulsions was 150 E.cmol/ml. This volume of emulsion was mixed with 205 ml of soluble fraction, and centrifuged as described in Fig. 2 in a Spinco L 2-65B ultracentrifuge with Type SW-27 swinging bucket rotor. The isolated lipid cakes were suspended in 230 ml of sucroseEDTA solution for washing by ultracentrifugation. Bound lipase was separated from the washed lipid cakes by density gradient centrifugation in media containing detergent. The media were prepared by dissolving sucrose in 0.01 M phosphate buffer, pH 7.5, to give solutions of density 1.25, 1.15, 1.05, and 1.00 g/ml. Triton X-100 at a concentration of 3 mg/ml was added to each solution. The washed lipid cakes were suspended by aspiration in 17 ml of the solution of density 1.25 g/ml. 8-ml aliquots of this mixture were layered beneath discontinuous density gradients formed by sequential layering of 10 ml of each of the three lighter solutions in centrifuge tubes (1 X 3.5 inch). Centrifugation was carried out in the Type SW-27 rotor. A number of discrete bands could be visualized by their opacity. The triacylglycerol-free fraction of density 1.25 g/ml, which contained the purified activity, was collected through a hole punctured in the bottom of the tube. The 1.25 g/ml fraction was washed and concentrated by ultrafiltration in a 150 ml Amicon apparatus (Amicon Corp., Lexington, Mass., U.S.A.) equipped with a Diaflo PM-10 filter. The sample was placed in the ultrafiltration appara-
136
tus along with 3 vols ice-cold distilled water. Pressure was applied by a flow of 5-7 ljmin N, , and filtration was continued until the solution was reduced to less than 10 ml in volume. The preparation was washed 3 times with l.O-ml vols of water. The final preparation, designated purified neutral lipase, was frozen at -12°C in small aliquots until analyzed. Isolation of lipoproteins and upolipoproteins Very low density lipoproteins were isolated from a fasting hypertriacylglyceremic subject according to the method of Gustafson et al. [24]. Chyle chylomicrons were isolated from the abdominal fluid of a patient suffering from a chylous fistufa. Chylous fluid was centrifuged at 93 000 X g in a Type SW-27 swinging bucket rotor for 1.5 h at 0°C. The packed lipid cake was washed twice by ultra~e~~ifugation in 0.15 M NaCl containing 0.001 M sodium-EDTA at pH 7.0. A-I and A-II polypeptides of apolipoprotein A were isolated according to the procedure of Kostner and Alaupovic [25]. Enzyme assays Lipase. Acidic (pH 4.5) triacylglycerol lipase was measured according to Miiller and Alaupovic [ 201. Neutral lipase activity (pH 7.5) was also determined by the method of these authors [ZO] in the subcellular fractionation studies. The optimal system for assay of purified neutral fipase was comprised of 0.5 ml of 0.02 M phosphate buffer (pH 7.25) containing 2 mg denatured hemoglobin, 0.5 ml of substrate emulsion (45 pm01 of triofein emulsified in 12 mM sodium taurocholate [ 33 ), and 0.5 ml of enzyme preparation. After incubation at 37°C [3] f reactions were stopped and fatty acids were extracted by the method of Dole [ 261. 1 ,umol palmitic acid in n-heptane was added to each assay tube, and standard solutions of palmitic acid in n-heptane were extracted and analyzed in identical fashion. Two methods were used for measurement of fatty acids. In the first procedure, 3-ml aliquots of heptane phase were titrated according to Schnatz [ 27 ] utilizing a Radiometer TTT-lc titrator equipped with TTA-31 titration assembly and SBR-2c recorder (Radiometer, Copenhagen, Denmark). Standard samples of palmitic acid contained 1-3 gmof of fatty acid. For the alternative procedure, 1.5 ml of a freshly prepared mixture of eth~ol~~.~2~ aqueous Nile Blue A sulfate~~.~2 M NaUH (9 : 1 : 0.2, v : v : v) were added to each 3 ml aliquot of heptane phase. Individual tubes were shaken rapidly by hand for 10 s and the phases were permitted to separate. The absorbance of the lower ethanolic phase was determined spectrophotometrically at 640 nm against a solvent blank. Standard solutions of palmitic acid in n-heptane containing 1.0-1.6 pmol fatty acid were used as references. Stock solutions of 0.02% aqueous Nile Blue A sulfate were washed with several volumes of n-heptane to remove extractable impurities and stored at room temperature. Marker enzymes. Succinic dehydrogenase, a marker for the presence of mitochondria, was determined according to the method of Pennington [ZS] . The lysosomal protease cathepsin D was assayed by the method of Beck et al. f29]. Released amino acids were quantized according to Rosen [30] with the modification of Grant f31]. ~lu~os~6*phospha~se~ a microsomal marker, was determined by the method of Hubscher and West 1321. The plasma membrane
enzyme 5’-nucleotidase was assayed according to Emmelot et al. [33]. Inorganic phosphate resulting from glucose-6-phosphatase and 5 ‘nucleotidase assays was quantitated by the method of Fiske and Subbarow [ 343. Monoolein hydroluse. Be&age’s method [ 351 for the measurement of this activity in pig liver preparations was followed except that the substrate used was unlabeled. Sodium taurodeoxycholate served as solubilizing agent for the monoacylglycerol. The reaction was quantitated by measurement of the released fatty acid as in lipase assays. Tributyrinase. The hydrolysis of tributyrin was assayed in a system adapted from Egelrud and Olivecrona [ 361. Small volumes (0.025-0.2 ml) of enzyme preparations were mixed with 0.1 M NaCl, pH 8.25, to a total volume of 1.9 ml in the sample cup of the Radiometer automatic titrating apparatus. The flow of 0.02 M NaOH required to maintain a pH of 8.25 was recorded for approximately 5 min. The reaction was started by injection of 0.1 ml of tributyrin and the flow of titrant recorded for another 5 min; reactions were adequately linear to determine rates. Results were expressed as E.tmol titrant/ min/mg protein, and specific activity was the difference between values in the presence and absence of substrate. Corrections were also made for non-enzymic hydrolysis. The tributyrin was maintained in dispersed form by rapid stirring; assays were carried out in duplicate at ambient temperature (approx. 23°C). Methyl butyrute esterase. The activity of methyl butyrate esterase was measured according to the procedure described by Arndt and Krisch [37] except that the temperature was 23°C. Results of this titrimetric procedure were expressed in the same manner as those for the hydrolysis of tributyrin. Protein determination Protein was determined by the method necessary, lipids were removed from samples analysis of protein content.
of Lowry by solvent
et al. [38]. Where extraction prior to
Results Calorimetric assay for long-chain fatty acids Attempts to quantitate long-chain fatty acids by manual titration in a two-phase system employing Nile Blue indicator [39] were not considered satisfactory due to difficulties in visual detection of the end point. However, the fact that a color change occurred in this reaction suggested the possibility that the assay might be adapted to spectrophotometric quantitation. In the presence of alkali, ethanolic Nile Blue indicator exhibited a pink color which failed to display an absorption maximum within the wavelength range 220-800 nm. In the absence of base, however, the blue indicator solution absorbed maximally at 640 nm. It was reasoned that the mixing of samples of fatty acid with indicator solution previously taken to its alkaline end point should produce blue color by downward titration of the indicator. Experiments were set up to determine conditions necessary for the reaction and to establish whether the color change could be quantitated spectrophotometrically. Known amounts of fatty acid, dissolved in 4 ml n-heptane, were assayed as described in Methods, utilizing color reagents in which the volume ratios of ethanol, 0.02%
1600
0600
0
1
2
3
4
5
pmoles PALMITIC ACID Fig. 3. Relationship of color yield to fatty acid concentration with several experimental color Color reagent mixtures contained absolute ethanol: 0.02% Nile Blue: 0.02 M NaOH in volume 9.0 : 1.0 : 0.5 (.---. ) and 9.0 : 1.0 : 0.2 (“- - - - - -0).
reagents. ratios of
aqueous Nile Blue, and 0.02 M NaOH were varied. The results of two representative assays are shown in Fig. 3, Development of blue color was sigmoidal with respect to fatty acid concentration. A range of linearity occurred in each plot; this range was determined primarily by the ratio of NaOH to Nile Blue present in the color reagent. On the basis of these results, the method was considered to be quite satisfactory for quantitation of long-chain fatty acids released during lipolysis and was adopted for use under the following specific conditions. Standard solutions of palmitic acid in n-heptane contained from 1.0 to 1.6 pmol of fatty acid; all assay samples contained 1.0 pmole fatty acid dissolved in n-heptane and added during the extraction procedure, and the enzyme concentration in assay samples was adjusted to maintain fatty acid release at less than 0.6 pmol.
Subcellular fractionation Subcellular fractionation of fresh pig liver resulted in quantitative isolation of five subcellular fractions: the nuclear (N), the mitochondrial (M), the light mitochondrial (L), the microsomal (P), and the soluble (S). The homogenate and each of the isolated fractions were assayed for their protein content and activities of six different enzymes: succinic dehydrogenase, cathepsin D, glucose-6-phosphatase, 5’-nucleotidase, triacylglycerol lipase at pH 4.5 and triacylglycerol lipase at pH 7.5. The data reported are based on the means of three separate experiments. The recovery of protein in the five subcellular fractions averaged 92% of that in the homogenate. Likewise, recoveries of the enzymatic activities cathepsin D (89%), glucose-6-phosphatase (91%) and lipase at pH 7.5 (95%) were nearly complete. The recoveries of succinic dehydrogenase (79%), lipase at pH 4.5 (80%) and 5’nucleotidase (68%) activities were somewhat less satisfactory. The latter enzyme, especially, appeared rather labile under the experimental
139
Cathcpsm
D
. % TOTAL PROTEIN Fig. 4. Distribution profiles of enwmes in pig liver subcellular fractions. N. nuclear fraction: dzial fraction; L. tight mitochondrial fraction; P, microsomal fraction: S. soluble fraction.
M. mitochon-
conditions. To correct for this variability, the amounts of protein and activities actually recovered in the five subcellular fractions have been taken as 100% in the summary of the data presented in Fig. 4. Purificationof neutraltriacylglycerollipase Determinations of the pH optimum of the neutral lipolytic activity in several preparations of soluble fractions (S) yielded curves similar to that presented by MtiIler and Alaupovic [20] for the pig liver microsomal enzyme. In one instance, the optima of the soluble fraction and the particulate fraction
140 TABLE
I
PURIFICATION S, soluble fraction; Number
OF PIG LIVER EBL-3,
of experiments
NEUTRAL
emulsion-bound
LIPASE bpase fraction
Fraction
8
S EBL-3
6
Purifwd
Protein
neutral
lipase
* Defined as specific activity of fraction/specific ** Numbers in parentheses give range of values.
3. (%)
100 37 ( 22-49) 16 ( 12-26)
100 1 ( 0.4-1.3j** 0.2 ( 0.1-0.2) activity
Lipase activity
(%)
Purification
’
1 41 (2656) 92 (84-111)
of S.
isolated by centrifugation at 113 000 X g for 30 min from a single homogenate were determined. The shapes of the curves were quite similar with peak activity occurring at pH 7.5 in both fractions. Purification of neutral triacylglycerol lipase was carried out according to the scheme presented in Fig. 2. The soluble fraction was chosen as starting material because of its high percentage of total lipase activity and its lack of membranous particles. Results of the purification procedure are shown in Table I. Successful isolation of emulsion-bound lipase fractions depended on maintenance of a smooth, unbroken emulsion. The sodium taurocholate-stabilized emulsions, however, tended to coagulate under certain conditions, including the presence of ions in the starting material or variation of the temperature during centrifugation by more than one degree from 0°C. The bound lipase was separated from the emulsion by density gradient
Fig. 5. PH dependence of neutral triacylglycerol lipase. Enzyme concentration was 14 fig protein/test. Substrate was purified olive oil at concentration of 60 I.cmol/test based on molecular weight of triolein. The buffer system was 0.02 M phosphate/O.1 M glycyl-glycine. Incubations were carried out for 15 min.
0
15
Biochimica et Biophysics Acta, 398 (1975) 132-148 0 Elsevier Scientific Publishing Company, Amsterdam
- Printed
in The Netherlands
BBA 56621
SUBCELLULAR FRACTIONATION, PARTIAL PURIFICATION AND CHARACTERIZATION OF NEUTRAL TRIACYLGLYCEROL LIPASE FROM PIG LIVER
JOANNA
H. LEDFORD*
and PETAR
ALAUPOVIC
Lipoprotein Laboratory, Oklahoma Medical Research Foundation and Department of Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center, Oklahoma City, Okla. 73104 (U.S.A.) (Received
March 17th,
1975)
Summary
The subcellular distributions of acidic (pW 4.5) and neutral (pH 7.5) longchain t~acylglycerol lipases (glycerol ester hydrolase, EC 3.1.1.3) of pig liver have been determined. The distribution of the acidic lipase closely paralleled that of the lysosomal marker enzyme, cathepsin D. Approx. 60% of the neutral lipolytic activity resided in the soluble fraction; the distribution of this activity failed to parallel that of marker enzymes for mitochondria, lysosomes, microsomes, or plasma membranes. A method has been developed for purification of the neutral lipase from the soluble fraction by ultracentrifugation. An approximate go-fold purification was achieved, with recovery of 16% of the initial activity. The partially purified neutral lipase exhibited a pH optimum between 7.25 and 7.5. It required 30 m&I emulsified triolein for optimal activity and ceased to liberate fatty acids after 30 min of incubation. The enzymatic activity was destroyed by heating at 60°C. Neutral lipase was inhibited by sodium deoxycholate, Triton X-100 and iodoacetamide. The activity was not inhibited by sodium taurocholate, EDTA, heparin and diethyl-p-nitrophenyl phosphate. Neutral lipase failed to exhibit activity in assay systems specific for lipoprotein lipase, monoolein hydrolase, tributyrinase, and methyl butyrate esterase and showed little or no capacity to hydrolyze chyle chylomicrons or plasma very low density lipoproteins. It is suggested that the function of neutral lipase may be to supply the liver with fatty acids liberated from endogenously synthesized or stored triacylglycerols.
* Present address: Courtauld Institute London WlP 5PB. U.K.
of Biochemistry,
The Middlesex Hospital Medical School.
133
Introduction The ability of mammalian liver to hydrolyze simple organic esters was recognized at the beginning of the century [1,2] , However, the existence of a liver enzyme capable of hydrolyzing long-chain triacylglycerols was not postulated until 1965 [3-71. It has now been established that rat liver contains at least two triacylglycerol lipases. One of these enzymes possesses an acidic pH optimum and is localized in lysosomes [ 8-16 ] , The second has a neutral or alkaline pH optimum and its localization has been variously reported as soluble 161, microsomal [11,17,18] or mitochondrial [19J. Recent evidence also suggests the possib~ity of a hep~n-~nsitive lipase in plasma membranes [ll,lS] . In 1970 Multer and Alaupovic [20] noted that pig liver also contained acidic [pH 4.51 and neutral (pH 7.5) triacylglycerol lipases. The acidic enzyme was clearly lysosomal in origin while the neutral lipase showed highest specific activity in microsomes. The present communication presents in more detail the subcellular distributions of pig liver triacylglycerol lipases, describes a procedure for partial purification of the neutral lipase and reports the characteristics of the partially purified enzyme.
Materials and Methods
Substrates employed in various enzymatic de~rminations included olive oil (Sargent-Welch, Dallas, Texas, U.S.A.), triolein, sodium succinate, adenosine 5’-monophosphate and glucose 6-phosphate (Sigma, St. Louis, MO., U.S.A.), monoolein, tributyrin and methyl butyrate (Analabs, North Haven, Conn., U.S.A.), iodonitrotetrazolium violet and denatured hemoglobin (Nutritional Biochemicals, Cleveland, Ohio, U.S.A.). Standards for analysis of protein content and reaction products were human albumin (Hoechst-Behringwerke, Marburg/Lahn, G.F.R.), palmitic acid (Sigma), iodonitrotetrazolium violet formazan (Nutritional Biochemicals), L-leucine (Mann, New York, N.Y., U.S.A.) and potassium dihydrogen phosphate (J.T. Baker, Phillipsburg, NJ., U.S.A.). Other compounds utilized during the isolation or assay of liver lipases included sodium taurocholate (Maybridge Research Chemicals, Launceston, Cornwall, U.K.), Triton X-100 (Calbiochem, LaJolla, Calif., U.S.A.), and thymol blue and Nile blue A indicators (Allied Chemical, Morristown, N.J., U.S.A.). Chemicals tested for inhibitory action on lipolysis were sodium heparin (as solution, 10 000 U/ml, Upjohn, Kalamazoo, Mich., U.S.A.), EDTA (Calbiochem), sodium deoxycholate (Mann, New York, N.Y., U.S.A.), diethyl-pnitrophenyl phosphate or E-600 (K & K, Plainview, N.Y., U.S.A.), and iodoacetamide (Sigma). Purificationof oliueoil
T~a~ylglycerols were isolated from olive oil by column ~hromato~phy over silica gel G according to a modification of &he method of Crider et al. 1211. The purified preparation was a clear, nearly colorless liquid, containing only triacylglycerol by thin-layer chromatography.
134
~amoge~~~~~o~. Pig livers, from animals of either sex, were obtained fresh from a local meat-packing firm and were held on ice for less than 1 h before use. Fresh livers were either processed imm~iately or divided into pieces of approximately 100 g each and frozen at -12°C. Frozen tissue was used within a l-month period. Fresh or thawed tissue, always obtained from several different lobes of the same liver, was chopped into pieces of about 1 cm3 size. 25 g of chopped tissue were mixed with 200 ml of ice-cold 0.25 M sucrose containing 0.001 M d&odium-EDTA, pH 7.2. The mixture was homogenized in a Sorvail Omnimixer (Ivan Sorvall Co., Newton, Corm., U.S.A.) for 30 s at setting 3 (gentle bomogeniz~tion) or for 30 s at setting 3 and 1 min at setting 10 (ha&r homo~e~~~ation}. Homo~ena~s were filtered through 8 layers of coarse cheesecloth into a chilled confiner, and the pH was adjusted to 7.2 with 5 M KOH. S~~ce~~~~~~~r~e~~o~~~~o~.~~rnoge~a~$ prepared from 50 g of fresh tissue by gentle homogenization were fractionated essentially according to the method of de Duve et al. [ 221 and Appelmans et al, [23]. The procedure is summarized in Fig. 1. Centrifugations were carried out near 0°C in, either a refrigerated Servall cen~ifuge (Ivan Sorvall Co,) with GSA rotor or a Spinco Model L 2-653 ultracentrifuge (Beckman Co., Palo Alto, Calif., U.S.A.) with Type 42 rotor. Final pellets were suspended in approximately 50 ml of sucrose/EDTA solution, and the suspensions were frozen immediately in small aliquots until anatyzed. F~rifi~~~o~ of ~e~~~~~~~~~~~~~yce~o~ lipme. The isolation procedure is summ~ized in Fig. 2. For each experiment a homogenate was prepared from 50 g of fresh or frozen pig liver by harsh homogenization.
Fig.
I,
Subcellular
fractionatwn
of Pig liver
homo$enate.
135 Homo
enate
7-
Centrifugatlo”
30
min,
113,000 x g,
0-c
Fig. 2. Purification of pig liver neutral lipase.
An artificial lipid emulsion was prepared by sonication of a mixture of 3.32 g of unpurified olive oil and 21.3 ml of 12 mM sodium taurocholate solution for two 1-min periods at setting 5 on a Bronwill Biosonik II (Bronwill Scientific, Rochester, N.Y ., U.S.A.) equipped with catenoidal probe. Assuming a molecular weight of triolein for the olive oil, the final concentration of triacylglycerol in the prepared emulsions was 150 E.cmol/ml. This volume of emulsion was mixed with 205 ml of soluble fraction, and centrifuged as described in Fig. 2 in a Spinco L 2-65B ultracentrifuge with Type SW-27 swinging bucket rotor. The isolated lipid cakes were suspended in 230 ml of sucroseEDTA solution for washing by ultracentrifugation. Bound lipase was separated from the washed lipid cakes by density gradient centrifugation in media containing detergent. The media were prepared by dissolving sucrose in 0.01 M phosphate buffer, pH 7.5, to give solutions of density 1.25, 1.15, 1.05, and 1.00 g/ml. Triton X-100 at a concentration of 3 mg/ml was added to each solution. The washed lipid cakes were suspended by aspiration in 17 ml of the solution of density 1.25 g/ml. 8-ml aliquots of this mixture were layered beneath discontinuous density gradients formed by sequential layering of 10 ml of each of the three lighter solutions in centrifuge tubes (1 X 3.5 inch). Centrifugation was carried out in the Type SW-27 rotor. A number of discrete bands could be visualized by their opacity. The triacylglycerol-free fraction of density 1.25 g/ml, which contained the purified activity, was collected through a hole punctured in the bottom of the tube. The 1.25 g/ml fraction was washed and concentrated by ultrafiltration in a 150 ml Amicon apparatus (Amicon Corp., Lexington, Mass., U.S.A.) equipped with a Diaflo PM-10 filter. The sample was placed in the ultrafiltration appara-
136
tus along with 3 vols ice-cold distilled water. Pressure was applied by a flow of 5-7 ljmin N, , and filtration was continued until the solution was reduced to less than 10 ml in volume. The preparation was washed 3 times with l.O-ml vols of water. The final preparation, designated purified neutral lipase, was frozen at -12°C in small aliquots until analyzed. Isolation of lipoproteins and upolipoproteins Very low density lipoproteins were isolated from a fasting hypertriacylglyceremic subject according to the method of Gustafson et al. [24]. Chyle chylomicrons were isolated from the abdominal fluid of a patient suffering from a chylous fistufa. Chylous fluid was centrifuged at 93 000 X g in a Type SW-27 swinging bucket rotor for 1.5 h at 0°C. The packed lipid cake was washed twice by ultra~e~~ifugation in 0.15 M NaCl containing 0.001 M sodium-EDTA at pH 7.0. A-I and A-II polypeptides of apolipoprotein A were isolated according to the procedure of Kostner and Alaupovic [25]. Enzyme assays Lipase. Acidic (pH 4.5) triacylglycerol lipase was measured according to Miiller and Alaupovic [ 201. Neutral lipase activity (pH 7.5) was also determined by the method of these authors [ZO] in the subcellular fractionation studies. The optimal system for assay of purified neutral fipase was comprised of 0.5 ml of 0.02 M phosphate buffer (pH 7.25) containing 2 mg denatured hemoglobin, 0.5 ml of substrate emulsion (45 pm01 of triofein emulsified in 12 mM sodium taurocholate [ 33 ), and 0.5 ml of enzyme preparation. After incubation at 37°C [3] f reactions were stopped and fatty acids were extracted by the method of Dole [ 261. 1 ,umol palmitic acid in n-heptane was added to each assay tube, and standard solutions of palmitic acid in n-heptane were extracted and analyzed in identical fashion. Two methods were used for measurement of fatty acids. In the first procedure, 3-ml aliquots of heptane phase were titrated according to Schnatz [ 27 ] utilizing a Radiometer TTT-lc titrator equipped with TTA-31 titration assembly and SBR-2c recorder (Radiometer, Copenhagen, Denmark). Standard samples of palmitic acid contained 1-3 gmof of fatty acid. For the alternative procedure, 1.5 ml of a freshly prepared mixture of eth~ol~~.~2~ aqueous Nile Blue A sulfate~~.~2 M NaUH (9 : 1 : 0.2, v : v : v) were added to each 3 ml aliquot of heptane phase. Individual tubes were shaken rapidly by hand for 10 s and the phases were permitted to separate. The absorbance of the lower ethanolic phase was determined spectrophotometrically at 640 nm against a solvent blank. Standard solutions of palmitic acid in n-heptane containing 1.0-1.6 pmol fatty acid were used as references. Stock solutions of 0.02% aqueous Nile Blue A sulfate were washed with several volumes of n-heptane to remove extractable impurities and stored at room temperature. Marker enzymes. Succinic dehydrogenase, a marker for the presence of mitochondria, was determined according to the method of Pennington [ZS] . The lysosomal protease cathepsin D was assayed by the method of Beck et al. f29]. Released amino acids were quantized according to Rosen [30] with the modification of Grant f31]. ~lu~os~6*phospha~se~ a microsomal marker, was determined by the method of Hubscher and West 1321. The plasma membrane
enzyme 5’-nucleotidase was assayed according to Emmelot et al. [33]. Inorganic phosphate resulting from glucose-6-phosphatase and 5 ‘nucleotidase assays was quantitated by the method of Fiske and Subbarow [ 343. Monoolein hydroluse. Be&age’s method [ 351 for the measurement of this activity in pig liver preparations was followed except that the substrate used was unlabeled. Sodium taurodeoxycholate served as solubilizing agent for the monoacylglycerol. The reaction was quantitated by measurement of the released fatty acid as in lipase assays. Tributyrinase. The hydrolysis of tributyrin was assayed in a system adapted from Egelrud and Olivecrona [ 361. Small volumes (0.025-0.2 ml) of enzyme preparations were mixed with 0.1 M NaCl, pH 8.25, to a total volume of 1.9 ml in the sample cup of the Radiometer automatic titrating apparatus. The flow of 0.02 M NaOH required to maintain a pH of 8.25 was recorded for approximately 5 min. The reaction was started by injection of 0.1 ml of tributyrin and the flow of titrant recorded for another 5 min; reactions were adequately linear to determine rates. Results were expressed as E.tmol titrant/ min/mg protein, and specific activity was the difference between values in the presence and absence of substrate. Corrections were also made for non-enzymic hydrolysis. The tributyrin was maintained in dispersed form by rapid stirring; assays were carried out in duplicate at ambient temperature (approx. 23°C). Methyl butyrute esterase. The activity of methyl butyrate esterase was measured according to the procedure described by Arndt and Krisch [37] except that the temperature was 23°C. Results of this titrimetric procedure were expressed in the same manner as those for the hydrolysis of tributyrin. Protein determination Protein was determined by the method necessary, lipids were removed from samples analysis of protein content.
of Lowry by solvent
et al. [38]. Where extraction prior to
Results Calorimetric assay for long-chain fatty acids Attempts to quantitate long-chain fatty acids by manual titration in a two-phase system employing Nile Blue indicator [39] were not considered satisfactory due to difficulties in visual detection of the end point. However, the fact that a color change occurred in this reaction suggested the possibility that the assay might be adapted to spectrophotometric quantitation. In the presence of alkali, ethanolic Nile Blue indicator exhibited a pink color which failed to display an absorption maximum within the wavelength range 220-800 nm. In the absence of base, however, the blue indicator solution absorbed maximally at 640 nm. It was reasoned that the mixing of samples of fatty acid with indicator solution previously taken to its alkaline end point should produce blue color by downward titration of the indicator. Experiments were set up to determine conditions necessary for the reaction and to establish whether the color change could be quantitated spectrophotometrically. Known amounts of fatty acid, dissolved in 4 ml n-heptane, were assayed as described in Methods, utilizing color reagents in which the volume ratios of ethanol, 0.02%
1600
0600
0
1
2
3
4
5
pmoles PALMITIC ACID Fig. 3. Relationship of color yield to fatty acid concentration with several experimental color Color reagent mixtures contained absolute ethanol: 0.02% Nile Blue: 0.02 M NaOH in volume 9.0 : 1.0 : 0.5 (.---. ) and 9.0 : 1.0 : 0.2 (“- - - - - -0).
reagents. ratios of
aqueous Nile Blue, and 0.02 M NaOH were varied. The results of two representative assays are shown in Fig. 3, Development of blue color was sigmoidal with respect to fatty acid concentration. A range of linearity occurred in each plot; this range was determined primarily by the ratio of NaOH to Nile Blue present in the color reagent. On the basis of these results, the method was considered to be quite satisfactory for quantitation of long-chain fatty acids released during lipolysis and was adopted for use under the following specific conditions. Standard solutions of palmitic acid in n-heptane contained from 1.0 to 1.6 pmol of fatty acid; all assay samples contained 1.0 pmole fatty acid dissolved in n-heptane and added during the extraction procedure, and the enzyme concentration in assay samples was adjusted to maintain fatty acid release at less than 0.6 pmol.
Subcellular fractionation Subcellular fractionation of fresh pig liver resulted in quantitative isolation of five subcellular fractions: the nuclear (N), the mitochondrial (M), the light mitochondrial (L), the microsomal (P), and the soluble (S). The homogenate and each of the isolated fractions were assayed for their protein content and activities of six different enzymes: succinic dehydrogenase, cathepsin D, glucose-6-phosphatase, 5’-nucleotidase, triacylglycerol lipase at pH 4.5 and triacylglycerol lipase at pH 7.5. The data reported are based on the means of three separate experiments. The recovery of protein in the five subcellular fractions averaged 92% of that in the homogenate. Likewise, recoveries of the enzymatic activities cathepsin D (89%), glucose-6-phosphatase (91%) and lipase at pH 7.5 (95%) were nearly complete. The recoveries of succinic dehydrogenase (79%), lipase at pH 4.5 (80%) and 5’nucleotidase (68%) activities were somewhat less satisfactory. The latter enzyme, especially, appeared rather labile under the experimental
139
Cathcpsm
D
. % TOTAL PROTEIN Fig. 4. Distribution profiles of enwmes in pig liver subcellular fractions. N. nuclear fraction: dzial fraction; L. tight mitochondrial fraction; P, microsomal fraction: S. soluble fraction.
M. mitochon-
conditions. To correct for this variability, the amounts of protein and activities actually recovered in the five subcellular fractions have been taken as 100% in the summary of the data presented in Fig. 4. Purificationof neutraltriacylglycerollipase Determinations of the pH optimum of the neutral lipolytic activity in several preparations of soluble fractions (S) yielded curves similar to that presented by MtiIler and Alaupovic [20] for the pig liver microsomal enzyme. In one instance, the optima of the soluble fraction and the particulate fraction
140 TABLE
I
PURIFICATION S, soluble fraction; Number
OF PIG LIVER EBL-3,
of experiments
NEUTRAL
emulsion-bound
LIPASE bpase fraction
Fraction
8
S EBL-3
6
Purifwd
Protein
neutral
lipase
* Defined as specific activity of fraction/specific ** Numbers in parentheses give range of values.
3. (%)
100 37 ( 22-49) 16 ( 12-26)
100 1 ( 0.4-1.3j** 0.2 ( 0.1-0.2) activity
Lipase activity
(%)
Purification
’
1 41 (2656) 92 (84-111)
of S.
isolated by centrifugation at 113 000 X g for 30 min from a single homogenate were determined. The shapes of the curves were quite similar with peak activity occurring at pH 7.5 in both fractions. Purification of neutral triacylglycerol lipase was carried out according to the scheme presented in Fig. 2. The soluble fraction was chosen as starting material because of its high percentage of total lipase activity and its lack of membranous particles. Results of the purification procedure are shown in Table I. Successful isolation of emulsion-bound lipase fractions depended on maintenance of a smooth, unbroken emulsion. The sodium taurocholate-stabilized emulsions, however, tended to coagulate under certain conditions, including the presence of ions in the starting material or variation of the temperature during centrifugation by more than one degree from 0°C. The bound lipase was separated from the emulsion by density gradient
Fig. 5. PH dependence of neutral triacylglycerol lipase. Enzyme concentration was 14 fig protein/test. Substrate was purified olive oil at concentration of 60 I.cmol/test based on molecular weight of triolein. The buffer system was 0.02 M phosphate/O.1 M glycyl-glycine. Incubations were carried out for 15 min.
0
15
