54
Biochit~tk'a ¢t Bi:~phy~'k'a Acta, I 115 ( 1991 ) 54-59 ~9 It)t)l Elsevier Science Publishers B.V. All rights reserved [1304-4165/9I/$03.51]
BBAGEN 23609
Peroxisomal enzyme activities in the human hepatoblastoma cell line HepG2 as compared to human liver R o n a i d J.A. W a n d e r s ~, Carlo W.T. van R o e r m u n d ~, M a r i e k e Griffioen 2 a n d Louis C o h e n 2 I Department of Clinical Biodlemistry, Unicersity Hospital Amsterdam, Amsterdam I The Netherland.Q and : INO htstitute for Ag~wg and Vascular Research. Gaubius Laboratory. Leiden (The Netherlands)
(Received 2 July 1991)
Key words: | lcpatoma cell; i lepaloblaslomacell line; Hep(12; Peroxisome;Hyperoxaluria;(tluman liver) In order to develop an in vitro model allowing investigation of the long-term effects of hormones and other agents on peroxisomes in liver cells, we measured the activity of a series of peroxisomal enzyme activities in HepG2 cells, a proliferating cell line derived from a human hepatoblastoma. The results obtained show that although in absolute terms peroxisomal enzyme activities are lower in HepG2 cells as compared to human liver, relative activities were comparable in HepG2 and human liver, respectively. Furthermore, it is shown that peroxisomes can easily be isolated from HepG2 cells using density gradient centrifugation. It is concluded that HepG2 cells represent a good model system to study the characteristic (long-term) regulation and control of metabolism of human liver peroxisomes.
Introduction Peroxisomes were long thought to play only a modest role in cellular metabolism, in the last few years, however, it has become clear that peroxisomes are indispensible organelles inw~lved in the /3-oxidation of a variety of compounds [2], the synthesis of ether phospholipids [3], the synthesis of bile acids [4] and other pathways. The identification of a newly recognized group of inherited diseases in man in which there is an impairment in one or more peroxisomai functions, has greatly stimulated the interest in peroxisomes [5,6]. Studies on peroxisomes have mostly been done on rat liver. The properties of peroxisomes, however, may vary from species to species even when the same organ is considered. Ataninc glyoxylate aminotransferase for instance, a deficiency of which causes hyperoxaluria type 1 [7], is exclusively peroxisomal in human liver, whereas it is entirely mitochondriai in dog and cat anu distributed equally between the two organelles in rat, mouse and hamster liver [8]. AccordingIy, conclusions drawn from experiments obtained
Correspondence: R.J.A. Wanders. Department of Clinical Bioehemistt3'. University ltospital Amsterdam, AMC. Meibergdreef 9, !105 AZ Amsterdam. The Netherlands.
with liver peroxisomes from one species, e.g., the rat, can not be extrapolated to other species. In order to study whether HepG2, a proliferating cell line derived from a human hepatoblastoma, would be a suitable model system to study the functions of peroxisomes in human liver, we measured the activity of a variety of peroxisomal enzymes and established whether or not the activity was located in peroxisomes. The results are described in this paper. Materials and Methods Cell cuhure conditions
HepG2 cells were cultured at 37°C in 75 cm-' flasks in Dulbecco's modified Eagle Medium (DMEM) supplemented with 10% (w/v) fetal calf serum. The gas phase was 5% CO.,/95% 0 2 (v/v). Preparation o f homogenates o f HepG2 cells and h u m a n
lit'er HepG2 cells and pieces of human liver were stored at - 7 0 ° C and thawed in a medium containing l0 mM sodium pyrophosphate, I mM EDTA and l0 t.tM FAD (final pH 9.0) on the day of the experiment. The thawed material was subjected to gentle homogenization, filtered through cheese-cloth and used for enzyme assays and immunoblotting (Table l and Fig. 2). When
fresh homogenates were required, e.g., for subcellar fractionation studies, the procedure described below was followed. Density gradient centrifugation H e p G 2 cells were harvested and taken up in homogenization medium. 250 mM sucrose, 2 mM EDTA, 0.1% ( v / v ) ethanol, to prevent formation of inactivc catalase compound I1, and 2 mM Mops-NaOH (pH 7.4). A postnuclear supernatant was subsequently prepared by homogenization at 4°C. This was done by 50 up-and-down strokes in a glass Potter Elvehjem homogenizer equipped with a tight-fitting pestle which was rotating at 1000 rpm. Following centrifugation (600 X g,,., 10 rain, 4°C) the resulting postnuclear supernatant was loaded onto a preformed Nycodenz gradient as described before [9,10]. E n z y m e aclit'ity tllet~sttret~letlls AcyI-CoA oxidase was d e t e r m i n e d by measuring the palmitoyI-CoA-dependent production of H , O , using a modification of the method described by Van Veldhoven and Mannaerts [111. An aliquot of human Iiver
or HepG2 cell homogenate was incubated in a medium containing the following standard components (final volume: 0.25 ml): 50 mM 4-morpholinopropanesuiphonic acid (Mops)-NaOH (pH 7.6k t mM homovanillic acid, 1 m g / m l horseradish peroxidase, 0.1 mM sodium azide, 0.1 mM palmitoyI-CoA and 1).6 m g / m l bovine serum albumin. Reactions were found to proceed linearly following a lag period. Consequently, reactions were terminated at 10 and 20 min, respectively, by adding 50/.~1 of perchloric acid (1.2% (w/v)). Tubes were put on ice for 15 rain and subsequently centrifuged in an E p p e n d o r f centrifuge at 12000 x g~,, for 15 rain at 4°C. 40 g l of the clear supernatant was subsequently a d d e d to 200/11 of a solution containing
0.5 M NaHCO~ plus 10 mM E D T A (final pH 10. 7). Fluorescence was subsequently read on a COBAS Centrifugal Analyzer (Hoffmann-La Roche, Basel, Switzerland). Excitation wavelength, 327 nm, emission wavelength, 42(I nm. Quantification of fluorescence was done by incubating known concentrations of urate ( 0 - 4 g M ) in the presence of urate oxidase in the same incubation medium. The samples were subsequently processed as described above. D-Aminoacid oxidase and glycolat¢ oxidasc were measured in the same incubation medium as described for acyl-CoA oxidase except that 50 mM D-alanine and 10 mM sodium glycolate were used as substrates, respectively. Glutamate dehydrogenase, estcrase and /~-hexosaminidasc wcrc measured as described before [10]. Dihydroxyacetone phosphate acyltransferase (DHAPAT) was measured as described before [13] with the following modiqcations. [U-laC]Dihydroxyacetone phosphate was prepared from l.-[U-I'~C]glycerot-3phosphate in a medium containing 50 mM triethanolaminc-HCI (pH 7.6k 0.6 mM l-[U-~C]gtycerol3-phosphate (Ill #Ci), 5 mM sodium pyruvate, 41.25 units of lactate dehydrogenase, and 25 units of glycerol-3-phosphate dehydrogenase in a total volume of 1.0 ml. Reactions were carried out at 25°C for 60 min and terminated by addition of an equal volume of chloroform followed by vigorous mixing on a Vortex apparatus. The upper layer was collected by centrifugation. A 20 ~! aliquot was used in a standard D H A P A T assay which contained the following components: 75 mM 4-morphotinoethanesulphonic acid ( M e s ) ( p H 5.7), 8 mM NaF, !. 6 mM MgCi_,, 3. 35 m g / m l bovine serum albumin, 10 mM glycerol-3-phosphate, 0.1 mM [U-~4C]dihydroxyacetone phosphate and 0.166 mM palmitoyl-CoA in a final volume of 0.12 ml. Reactions
TABLE I Enzyme actirities m homogenates of th'p(;2 cells aml human liler cells Enzyme activity measured
Material analyzed relative activity
absolute aclivity
tlepG2 cells DHAPAT (nmo[/2 h per mg) Catatase (#mol/min per rag) AcyI-CoAoxidase (nmol/min per rag) o-Aminoacidoxidase (nmol/min per rag) Hydroxyacidoxidase A (nmol/min per rng) Aianine glyoxylateaminotransferase(nmol/min per mg) Glutamate dehydtogenase(#tool/rain per rag) Hexosaminidase(~.mol/min per rag) Esterase (p.mol/min per rag)
0,52-.2_0.03 (I,73 _-211.21)
human liver 1.49+ t}.28 2.111+ 11.45 1.117+ l).14
0.26+0.04 O, lO + 0.111
11.88 +
O.l(I _+ (I,iJ2
1.0~) _+ 0.21
31.4 + 4.3 6.(1 -L-_0.7 179 2 8
t.4 _+0.1
Data are presented as the mere:± ~ D. for three to five separate experiments.
1). ! l
270 4:_34 8.2 +_ 1.2 44
± 9
3.3 +_ 0.7
HepG2 cells
human li~er
1.00
1.00
1.40 (I.50 O, 19 o. 19 6{} II .5
1.35 fl,72 0.5~
344 2.7
1.37
18 !
5.5 29 2.2
56 were started by the addi!ion of 5-20 /11 aliquols of homogcnate protein. Incubations ~cre carried out at 37°C and reactions terminated by adding 6()() /a.l chlor o l b r m / m e t h a n o l ( l : l by vo[ume) and 15() /.tl of a solution containing 2 M KCl plus (I.2 M H.~PO4 followed by vigorous mixing and centrifugation (5 rain. 700 × g,,,. 4°C). "I'ile upper layer was subsequently removed and 201),ul of the lower layer containing radiolabelled palmitoyl-DHAP spotted on small paper disks (Wllatman 3 mm cellulose filter papers (No. 1030021 )). The filters were subsequently dried and washed in 12(1 ml of 10% (twice), 5% and 1~ ( w / v ) T C A followed by quantificalion of radioactivity.
All other reagents were of the purest analytical grade available. Resu Its
Peroxisomal enzyme uctit'ities in HepG2 cells: compatqson with human lit'er In order to characterize peroxisomcs in HepG2 cells, we first measured the activity of catalase, .)-amino acid oxidase, hydroxy acid oxidase A, alanine glyoxylaie •"lminotransferase. acyl-CoA oxldase, and dihydroxyacetonephosphate acyltransferase (DHAPAT). The latter enzyme has been regarded to be the marker of choice fl)r peroxJsomes, at least when measured at slightly
Pr¢Tmrution qf antilnulie.s ,rid immttnot~h~tting procedttres Antibodies were raised in ral'tbils using the rid[owing antigens: acvl-('oA oxidase [14]. bifunclion;t[ protein [15] arid pcroxisonl;ll thi~)lase [lfq as isolaled from di-(¢thylhexyl)phthahlt¢-indticcd r;it livers, caltilase ;is purified from bovine liver I171 and afaninc glyoxyhite aminolr;lnsferas¢ ;i.~ purified from htinlan I i w r l isl. I m m u n o b h l i l i n g was chine :is described before l l
Biochit~tk'a ¢t Bi:~phy~'k'a Acta, I 115 ( 1991 ) 54-59 ~9 It)t)l Elsevier Science Publishers B.V. All rights reserved [1304-4165/9I/$03.51]
BBAGEN 23609
Peroxisomal enzyme activities in the human hepatoblastoma cell line HepG2 as compared to human liver R o n a i d J.A. W a n d e r s ~, Carlo W.T. van R o e r m u n d ~, M a r i e k e Griffioen 2 a n d Louis C o h e n 2 I Department of Clinical Biodlemistry, Unicersity Hospital Amsterdam, Amsterdam I The Netherland.Q and : INO htstitute for Ag~wg and Vascular Research. Gaubius Laboratory. Leiden (The Netherlands)
(Received 2 July 1991)
Key words: | lcpatoma cell; i lepaloblaslomacell line; Hep(12; Peroxisome;Hyperoxaluria;(tluman liver) In order to develop an in vitro model allowing investigation of the long-term effects of hormones and other agents on peroxisomes in liver cells, we measured the activity of a series of peroxisomal enzyme activities in HepG2 cells, a proliferating cell line derived from a human hepatoblastoma. The results obtained show that although in absolute terms peroxisomal enzyme activities are lower in HepG2 cells as compared to human liver, relative activities were comparable in HepG2 and human liver, respectively. Furthermore, it is shown that peroxisomes can easily be isolated from HepG2 cells using density gradient centrifugation. It is concluded that HepG2 cells represent a good model system to study the characteristic (long-term) regulation and control of metabolism of human liver peroxisomes.
Introduction Peroxisomes were long thought to play only a modest role in cellular metabolism, in the last few years, however, it has become clear that peroxisomes are indispensible organelles inw~lved in the /3-oxidation of a variety of compounds [2], the synthesis of ether phospholipids [3], the synthesis of bile acids [4] and other pathways. The identification of a newly recognized group of inherited diseases in man in which there is an impairment in one or more peroxisomai functions, has greatly stimulated the interest in peroxisomes [5,6]. Studies on peroxisomes have mostly been done on rat liver. The properties of peroxisomes, however, may vary from species to species even when the same organ is considered. Ataninc glyoxylate aminotransferase for instance, a deficiency of which causes hyperoxaluria type 1 [7], is exclusively peroxisomal in human liver, whereas it is entirely mitochondriai in dog and cat anu distributed equally between the two organelles in rat, mouse and hamster liver [8]. AccordingIy, conclusions drawn from experiments obtained
Correspondence: R.J.A. Wanders. Department of Clinical Bioehemistt3'. University ltospital Amsterdam, AMC. Meibergdreef 9, !105 AZ Amsterdam. The Netherlands.
with liver peroxisomes from one species, e.g., the rat, can not be extrapolated to other species. In order to study whether HepG2, a proliferating cell line derived from a human hepatoblastoma, would be a suitable model system to study the functions of peroxisomes in human liver, we measured the activity of a variety of peroxisomal enzymes and established whether or not the activity was located in peroxisomes. The results are described in this paper. Materials and Methods Cell cuhure conditions
HepG2 cells were cultured at 37°C in 75 cm-' flasks in Dulbecco's modified Eagle Medium (DMEM) supplemented with 10% (w/v) fetal calf serum. The gas phase was 5% CO.,/95% 0 2 (v/v). Preparation o f homogenates o f HepG2 cells and h u m a n
lit'er HepG2 cells and pieces of human liver were stored at - 7 0 ° C and thawed in a medium containing l0 mM sodium pyrophosphate, I mM EDTA and l0 t.tM FAD (final pH 9.0) on the day of the experiment. The thawed material was subjected to gentle homogenization, filtered through cheese-cloth and used for enzyme assays and immunoblotting (Table l and Fig. 2). When
fresh homogenates were required, e.g., for subcellar fractionation studies, the procedure described below was followed. Density gradient centrifugation H e p G 2 cells were harvested and taken up in homogenization medium. 250 mM sucrose, 2 mM EDTA, 0.1% ( v / v ) ethanol, to prevent formation of inactivc catalase compound I1, and 2 mM Mops-NaOH (pH 7.4). A postnuclear supernatant was subsequently prepared by homogenization at 4°C. This was done by 50 up-and-down strokes in a glass Potter Elvehjem homogenizer equipped with a tight-fitting pestle which was rotating at 1000 rpm. Following centrifugation (600 X g,,., 10 rain, 4°C) the resulting postnuclear supernatant was loaded onto a preformed Nycodenz gradient as described before [9,10]. E n z y m e aclit'ity tllet~sttret~letlls AcyI-CoA oxidase was d e t e r m i n e d by measuring the palmitoyI-CoA-dependent production of H , O , using a modification of the method described by Van Veldhoven and Mannaerts [111. An aliquot of human Iiver
or HepG2 cell homogenate was incubated in a medium containing the following standard components (final volume: 0.25 ml): 50 mM 4-morpholinopropanesuiphonic acid (Mops)-NaOH (pH 7.6k t mM homovanillic acid, 1 m g / m l horseradish peroxidase, 0.1 mM sodium azide, 0.1 mM palmitoyI-CoA and 1).6 m g / m l bovine serum albumin. Reactions were found to proceed linearly following a lag period. Consequently, reactions were terminated at 10 and 20 min, respectively, by adding 50/.~1 of perchloric acid (1.2% (w/v)). Tubes were put on ice for 15 rain and subsequently centrifuged in an E p p e n d o r f centrifuge at 12000 x g~,, for 15 rain at 4°C. 40 g l of the clear supernatant was subsequently a d d e d to 200/11 of a solution containing
0.5 M NaHCO~ plus 10 mM E D T A (final pH 10. 7). Fluorescence was subsequently read on a COBAS Centrifugal Analyzer (Hoffmann-La Roche, Basel, Switzerland). Excitation wavelength, 327 nm, emission wavelength, 42(I nm. Quantification of fluorescence was done by incubating known concentrations of urate ( 0 - 4 g M ) in the presence of urate oxidase in the same incubation medium. The samples were subsequently processed as described above. D-Aminoacid oxidase and glycolat¢ oxidasc were measured in the same incubation medium as described for acyl-CoA oxidase except that 50 mM D-alanine and 10 mM sodium glycolate were used as substrates, respectively. Glutamate dehydrogenase, estcrase and /~-hexosaminidasc wcrc measured as described before [10]. Dihydroxyacetone phosphate acyltransferase (DHAPAT) was measured as described before [13] with the following modiqcations. [U-laC]Dihydroxyacetone phosphate was prepared from l.-[U-I'~C]glycerot-3phosphate in a medium containing 50 mM triethanolaminc-HCI (pH 7.6k 0.6 mM l-[U-~C]gtycerol3-phosphate (Ill #Ci), 5 mM sodium pyruvate, 41.25 units of lactate dehydrogenase, and 25 units of glycerol-3-phosphate dehydrogenase in a total volume of 1.0 ml. Reactions were carried out at 25°C for 60 min and terminated by addition of an equal volume of chloroform followed by vigorous mixing on a Vortex apparatus. The upper layer was collected by centrifugation. A 20 ~! aliquot was used in a standard D H A P A T assay which contained the following components: 75 mM 4-morphotinoethanesulphonic acid ( M e s ) ( p H 5.7), 8 mM NaF, !. 6 mM MgCi_,, 3. 35 m g / m l bovine serum albumin, 10 mM glycerol-3-phosphate, 0.1 mM [U-~4C]dihydroxyacetone phosphate and 0.166 mM palmitoyl-CoA in a final volume of 0.12 ml. Reactions
TABLE I Enzyme actirities m homogenates of th'p(;2 cells aml human liler cells Enzyme activity measured
Material analyzed relative activity
absolute aclivity
tlepG2 cells DHAPAT (nmo[/2 h per mg) Catatase (#mol/min per rag) AcyI-CoAoxidase (nmol/min per rag) o-Aminoacidoxidase (nmol/min per rag) Hydroxyacidoxidase A (nmol/min per rng) Aianine glyoxylateaminotransferase(nmol/min per mg) Glutamate dehydtogenase(#tool/rain per rag) Hexosaminidase(~.mol/min per rag) Esterase (p.mol/min per rag)
0,52-.2_0.03 (I,73 _-211.21)
human liver 1.49+ t}.28 2.111+ 11.45 1.117+ l).14
0.26+0.04 O, lO + 0.111
11.88 +
O.l(I _+ (I,iJ2
1.0~) _+ 0.21
31.4 + 4.3 6.(1 -L-_0.7 179 2 8
t.4 _+0.1
Data are presented as the mere:± ~ D. for three to five separate experiments.
1). ! l
270 4:_34 8.2 +_ 1.2 44
± 9
3.3 +_ 0.7
HepG2 cells
human li~er
1.00
1.00
1.40 (I.50 O, 19 o. 19 6{} II .5
1.35 fl,72 0.5~
344 2.7
1.37
18 !
5.5 29 2.2
56 were started by the addi!ion of 5-20 /11 aliquols of homogcnate protein. Incubations ~cre carried out at 37°C and reactions terminated by adding 6()() /a.l chlor o l b r m / m e t h a n o l ( l : l by vo[ume) and 15() /.tl of a solution containing 2 M KCl plus (I.2 M H.~PO4 followed by vigorous mixing and centrifugation (5 rain. 700 × g,,,. 4°C). "I'ile upper layer was subsequently removed and 201),ul of the lower layer containing radiolabelled palmitoyl-DHAP spotted on small paper disks (Wllatman 3 mm cellulose filter papers (No. 1030021 )). The filters were subsequently dried and washed in 12(1 ml of 10% (twice), 5% and 1~ ( w / v ) T C A followed by quantificalion of radioactivity.
All other reagents were of the purest analytical grade available. Resu Its
Peroxisomal enzyme uctit'ities in HepG2 cells: compatqson with human lit'er In order to characterize peroxisomcs in HepG2 cells, we first measured the activity of catalase, .)-amino acid oxidase, hydroxy acid oxidase A, alanine glyoxylaie •"lminotransferase. acyl-CoA oxldase, and dihydroxyacetonephosphate acyltransferase (DHAPAT). The latter enzyme has been regarded to be the marker of choice fl)r peroxJsomes, at least when measured at slightly
Pr¢Tmrution qf antilnulie.s ,rid immttnot~h~tting procedttres Antibodies were raised in ral'tbils using the rid[owing antigens: acvl-('oA oxidase [14]. bifunclion;t[ protein [15] arid pcroxisonl;ll thi~)lase [lfq as isolaled from di-(¢thylhexyl)phthahlt¢-indticcd r;it livers, caltilase ;is purified from bovine liver I171 and afaninc glyoxyhite aminolr;lnsferas¢ ;i.~ purified from htinlan I i w r l isl. I m m u n o b h l i l i n g was chine :is described before l l
