ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 268, No. 2, February 1, pp. 676-686,1989

Dihydroxyacetone Phosphate Acyltransferase and Alkyldihydroxyacetone Phosphate Synthase Activities in Rat Liver Subcellular Fractions and

Human Skin Fibroblasts

HARMEET SINGH,’ SYLVIA USHER, AND ALF POULOS

Department of Chemical Pathology, Adelaide Children k Hospital. 72 King William Road, North Adelaide, South Australia 5006

Received May 5.1988, and in revised form October 4,1988

Dihydroxyacetone phosphate acyltransferase (DHAP-AT) and alkyldihydroxyace- tone phosphate synthase (DHAP-synthase) activities were examined in subcellular frac- tions of rat liver. The results indicate that at least 80% of DHAP-AT (assays carried out at pH 5.4) activity in rat liver is in peroxisomes, and the remaining activity is mito- chondrial. In contrast, to DHAP-AT, DHAP-synthase was detected in all subcellular fractions analyzed but the activity in peroxisomes was 208-fold and 42-fold greater com- pared to mitochondria and microsomes, respectively. We estimate that at least ‘70% of the DHAP-synthase activity in rat liver is in peroxisomes. DHAP-AT and DHAP-syn- thase activities were also examined in homogenates of skin fibroblasts from patients with inherited defects in peroxisomal structure and/or function. Both the enzyme activ- ities were deficient in Zellweger syndrome whereas the activities were only partially deficient in infantile Refsum’s disease. Greater reduction in DHAP-synthase activity, but only a partial reduction in DHAP-AT activity was observed in rhizomelic chondro- dysplasia punctata. However, both DHAP-AT and DHAP-synthase activities were ei- ther normal or near normal in Refsum’s disease or X-linked adrenoleukodystrophy. The results reported suggest that various peroxisomal disease states can be identified based on DHAP-AT and DHAP-synthase activities in skin fibroblasts of patients. CC 1989Academic

Press. Inc

Acyl dihydroxyacetone phosphate (acyl DHAP)’ is a key intermediate in biosyn- thesis of glycerol ether lipids including plasmalogens. Acyl DHAP is formed by transfer of long chain fatty acids from long chain fatty acyl CoAs to DHAP by an en- zyme, dihydroxyacetone phosphate acyl- transferase (DHAP-AT) (EC 2.3.1.42) (1, 2). The ester linkage on acyl DHAP is con-

1 To whom correspondence should be addressed. * Abbreviations used: acyl DHAP, acyl dihydroxy-

acetone phosphate; DHAP-AT, dihydroxyacetone phosphate acyltransferase; DHAP-synthase, alkyldi- hydroxyacetone phosphate synthase; ALD, X-linked adrenoleukodystrophy; ZS, Zellweger syndrome; RCP, rhizomelic chondrodysplasia punctata; TCA. trichloroacetic acid.

verted enzymatically to ether linkage by an exchange reaction with long chain alco- hols. The conversion of acyl DHAP to l-0 alkyl DHAP is carried out by an enzyme, alkyldihydroxyacetone phosphate syn- thase (DHAP-synthase) (3).

The importance of peroxisomes in ether lipid biosynthesis was appreciated only re- cently after the demonstration that tissues of infants without peroxisomes (Zellweger syndrome) are deficient in plasmalogens (4). Subsequently, activities of two of the ether lipid biosynthetic enzymes, namely DHAP-AT and DHAP-synthase were found to be reduced in skin fibroblasts and leukocytes of Zellweger patients (5-7). However, we found that the published pro- cedures (5, 7) were not reliable for the as-

0003-9861/89 $3.00 Copyright 0 1989 by Academic Press. Inc. Ail rights of reproduction in any form reserved.

676

DHAP-AT AND DHAP-SYNTHASE ACTIVITIES 677

say of DHAP-synthase activity in skin fi- broblast homogenates. Also, in view of the conflicting reports published from differ- ent laboratories on the localization of DHAP-AT and DHAP-synthase (8-ll), we set out to reinvestigate the relative distri- bution of the two enzymes in subcellular fractions of rat liver, and analyzed the en- zyme activities in skin fibroblasts of pa- tients with several peroxisomal disorders. We optimized the conditions for the mea- surement of activities of DHAP-AT and DHAP-synthase and report that in rat liver both of these enzymes were mainly lo- calized in peroxisomes. Furthermore, the activities of DHAP-AT and DHAP-syn- thase were reduced in Zellweger syndrome (ZS), infantile Refsum’s disease, and rhizo- melic chondrodysplasia punctata (RCP), whereas the activities were normal or near normal in Refsum’s disease and in X- linked adrenoleukodystrophy (ALD).

MATERIALS AND METHODS

[I-‘“C]Palmitic acid (58 mCi/mmol). L-[II-“Cl- glycerol 3-phosphate (171 mCi/mmol), and L-[l-‘%I- glutamic acid (58 mCi/mmolj were supplied by Amer- sham International, Plc.. England. Sterilized solu- tions of Percoll were obtained from Pharmacia Fine Chemicals, A.B., Uppsala, Sweden, and Nycodenz from Nycomed AS, Oslo, Norway. Whatman filter pa- pers and DEAE-cellulose (DE-U) filter papers were purchased from Whatman Chemical Separation Ltd., England. Basal modified Eagle’s medium was sup- plied by Flow Laboratories, McLean, Virginia, and fe- tal calf serum by GIBCO New Zealand Ltd., New Zea- land. Dulbecco’s phosphate-buffered saline (Ca’+ and M$+ freej and trypsin-versene solutions were ob- tained from Commonwealth Serum Laboratories, Melbourne. All other chemicals and reagents were purchased from Sigma Chemical Co., St. Louis, Mis- souri. All solvents used were either from May and Baker Australia Pty. Ltd., West Footscray, Victoria, or Ajax Chemicals, Sydney, New South Wales.

Four of the Zellweger fibroblast lines used were kindly provided by Professor David Danks (The Mur- doch Research Institute into Birth Defects, Royal Children’s Hospital, Melbourne, Victoria). One of the lines was sent by Dr. Geoffrey Sherwood (Hospital for Sick Children, Toronto, Ontario) and the ot.her by Dr. illan Clague f Royal Brisbane Hospital, Brisbane, Queensland). The diagnosis in each case was based on case histories, clinical features, and biochemical in- vestigations. Biochemical investigations were carried out in our laboratory; the Zellweger patients had in-

creased levels of very long chain fatty acids in plasma and skin fibroblasts. and reduced phytanic acid oxida- tion in skin fibroblasts.

The case histories of two of the patients with RCP have been reported earlier (12) as patient 1 and pa- tient 2. Skin fibroblasts of one of the remaining pa- tients were supplied by Dr. Les Sheffield (Royal Chil- dren’s Hospital, Melbourne, Victoria) and the other by Professor David Sillence (Royal Alexandra Hospi- tal for Children, Camperdown, New South Wales). The diagnosis in each case was based on case histo- ries, clinical symptoms, and biochemical investiga- tions. Plasmalogen content and phytanic acid oxida- tion were deficient whereas very long chain fatty acid content was normal in skin fibroblasts of RCP pa- tients.

Patients with ALD exhibited mental and motor de- generation commencing from childhood, coupled with adrenal insufficiency as reported by Schaumberg et al. (13). The diagnosis was confirmed biochemically by increased levels of very long chain fatty acids and normal levels of phytanic acid in plasma and skin fi- broblasts. The diagnosis of Refsum’s disease was based on clinical features, increased plasma levels of phytanic acid, deficient phytanic acid oxidation by skin fibroblasts, and normal levels of very long chain fatty acids in plasma and skin fibroblasts. Biochemi- cal findings of one of the Refsum’s patients have been reported previously (Case 5 of Ref. (14)). Diagnosis of infantile Refsum’s disease was based on clinical symptoms and biochemical findings including in- creased very long chain fatty acids and phytanic acid content in plasma, and increased very long chain fatty acid content and reduced phytanic acid oxida- tion in skin fibroblasts. Case histories and biochemi- cal findings of two patients (Case 2 of Ref. (15j; Pa- tient 2 of Ref. (16)) have been described. One of the patients examined was a sibling of the Case 1 de- scribed by Poulos ef al. (IT).

I.so/atio~~ of nr:qn~~eUes from rat /ire,: Adult albino rats (porton strain, purchased from Waite Institute. Adelaide) were killed by decapitation and livers were removed. Portions (6-8 g) of the livers were finely minced with scissors and hand homogenized in 9 vol of homogenization buffer (0.25 M sucrose-10 mM Tris- HCI buffer, pH 8.0-l rnbt EDTBJ using a Potter-El- vehjem-type homogenizer with a loose-fitting pestle. The resultant homogenate was centrifuged at 200g for 5 min to remove cell debris and designated as the homogenate. The subcellular fractionation was car- ried out at J°C using a fixed angle Ti-70 rotor (Beck- man Instruments, Palo Alto, CA). The homogenate was centrifuged at 2OOOg for 15 min, and the pellet gently homogenised in 25 ml of homogenization buffer and centrifuged as above. The resultant pellet (crude mitochondriaj was gently dispersed in 8 ml of homogenization buffer and used for isolation of pure mitochondria (see below). The 2000g supernatants

678 SINGH, USHER. AND POULOS

were combined and centrifuged at 16,500g for 10 min. The pellet was homogenized gently in 25 ml of homog- enization buffer and recentrifuged as above. The pel- let (crude peroxisomes) was resuspended gently in 4 ml of homogenization buffer and used for isolation of pure peroxisomes (see below). The 16,500g superna- tants were combined and centrifuged further at 130,OOOg for 60 min and the pellet (microsomes) was resuspended gently in homogenization buffer. An isoosmotic stock solution of Percoll was diluted with homogenization buffer to make a 30% Percoll solu- tion. Crude mitochondrial fraction (2 ml) was layered onto 34 ml of 30% Percoll and centrifuged at 50,OOOg for 60 min. The bottom mitochondrial protein band (18) was collected. Peroxisomes were purified from the crude peroxisomal fraction according to the method described by Ghosh and Hajra (9). Briefly, 2 ml of crude peroxisomal fraction was layered onto 15 ml of 30% Nycodenz solution containing 10 mM Tris- HCI buffer, pH 8.0, and 1 mM EDTA, centrifuged at 130,OOOg for 60 min, and the loose pellet at the bottom of the centrifuge tube was collected as pure peroxi- somes.

Preparation of h.uman skin fibroblast homogenates. Human skin fibroblasts were grown under sterile con- ditions in tissue culture flasks (75 cm’) in basal modi- fied Eagle’s medium containing 10% (v/v) fetal calf serum. The cells from confluent cultures were re- moved by incubating with trypsin-versene solution (3 ml per flask) at 37°C for 2-5 min. The cells were pel- leted by low-speed centrifugation (400gfor 5 min) and were washed three times with Dulbecco’s phosphate- buffered saline (4-5 ml each time) to remove residual trypsin. The cell pellet was suspended in 0.25 M su-

crose and briefly sonicated (5-10 s) using a microtip (Ystrom) to disrupt the cells. The cell sonicate was centrifuged at 400g for 10 min to remove unbroken cells, and used for enzyme assays.

Bhydroxyacetone phosphafe acyZtru&rase. L-[U- “C]Glycerol 3-phosphate was converted enzymati- tally to [LJ-‘“Cldihydroxacetone 3-phosphate. Briefly, 10 mM L-[U-‘4C]glycerol 3-phosphate (2.5 mCi/mmol) in 0.5 ml was incubated at room temperature for 60 min in 100 mM Tris-HCl buffer, pH 8.0, in the pres- ence of pyruvate (70 mM), NAD (3 mM), 500 units/ml of lactate dehydrogenase (from rabbit muscle, Sigma Chemical Co., St. Louis, MO), and 100 Fg/ml of glyc- erol-3-phosphate dehydrogenase (from rabbit mus- cle, Boehringer-Mannheim, West Germany).

For DHAP-AT assay, the reaction mixture con- sisted of sodium acetate buffer, pH 5.4 (80 mM), MgClz (10 mM), NaF (10 mM), fatty acid-free bovine serum albumin (2 mg/assay), palmitoyl CoA (350 PM), and freshly prepared [U-“Cldihydroxyacetone 3-phos- phate (600-800 pM) in a total volume of 100 ~1. The reaction was started by the addition of 5-80 ng pro- tein and carried out at 37°C for 30 min for rat liver, or 60 min for human skin fibroblasts. The reaction was

stopped by the addition of 450 pl chloroform:metha- no1 (1:2, v/v) and the two phases were separated by the addition of 150 ~1 chloroform and 150 ~1 of 2 M KC1 containing 0.2 M HaPOd. The radioactivity extracted in the chloroform phase was spotted onto 15-mm fil- ter paper strips and the strips were transferred to glass scintillation vials. The filter paper strips were washed successively once with 5 ml of 10% trichloro- acetic acid (TCA). 5% TCA, and 1% TCA. The radio- activity retained on the filter paper strips was deter- mined. The recovery of the product, namely l-acyldi- hydroxyacetone 3-phosphate, was found to be >95%.

Alkyldihydroxyacetone phosphate synthase. The DHAP-synthase assay was carried out in the pres- ence of Tris-HCI buffer, pH 8.0 (100 mM), NaF (50 mM), NAD (0.2 mM), l-0-palmitoyldihydroxyacetone 3-phosphate (synthesized according to the method of Hajra et aL (19) (120 pM), [l-‘4C]hexadecanol (Cl60 alcohol, sp act 58 mCi/mmol) (35-40 pM), and Triton X-100 (0.04%). Hexadecanol and 1-O-palmitoyldihy- droxyacetone 3-phosphate were dispersed in Triton X-100 by sonication for 5-15 min in an ultrasonic bath and appropriate volumes were pipetted for enzyme assays. The incubations were carried for 20 min at 37°C in a total reaction volume of 100 pl. The reaction was started by adding 5-20 pg protein and terminated by the addition of 3 ml of chloroform:methanol (2:1, v/v). The two phases were separated by the addition of 0.6 ml water, and the chloroform phase was dis- carded. Twenty microlitres of 10 M HCI and 300 ~1 of chloroform were added to the upper phase, mixed, and centrifuged. The radioactivity extracted in the chloroform phase was spotted onto 15-mm strips of DE-81 filter paper, and the strips were transferred to glass scintillation vials. The filter paper strips were washed five times with 3 ml of methanol, and the ra- dioactivity retained on the filter paper strips was de- termined. The recoveries of the product through the above procedure was determined each time using l- 0-palmityldihydroxyacetone 3-phosphate and found to be 50-55%. Appropriate control experiments where the reaction was terminated at zero min, or the enzyme protein was omitted from the assay mixture were conducted every time.

Identijcation of product of alkyldihydroxyaceto phosphate synthuse. Purified peroxisomes or micro- somes were incubated as above for DHAP-synthase assay. Control experiments were also carried out where the enzyme protein was omitted from the reac- tion mixture. After 20 min incubation, the reaction was stopped with 3 ml of chloroform:methanol (2:1, v/v) and the extract was acidified using 20 ~1 10 M HCI. The two phases were separated by the addition of 0.6 ml water, and the aqueous phase was discarded. The solvent from the chloroform phase (containing the radiolabeled substrate and the product) was evap- orated under nitrogen, and the extract was applied to TLC silica gel 60 plate. The plate was developed in

DHAP-AT AND DHAP-SYNTHASE ACTIVITIES 679

hexane:diethyl etheracetic acid (80:20:1, by vol) and subjected to autoradiography for 1 week. Only two ra- dioactive spots were observed, one migrating with C16:O alcohol (radiolabeled substrate) and the other highly polar compound which migrated near the ori- gin and was absent from the control experiments. The DHAP-synthase reaction product extracted in chlo- roform (see above) was further chromatographed on TLC silica gel 60 plate using chloroform:methanol: acetic acid:5’% sodium bisulfite (25:10:3:1, by vol) as the solvent, and the plate was subjected to autoradi- ography. Only one radioactive product was observed using purified peroxisomes or microsomes as the en- zyme source. The radiolabeled product comigrated on TLC with l-0-alkyl dihydroxyacetone 3-phosphate.

Marker enzyme assays. Succinate dehydrogenase (succinate-INT reductase), a mitochondrial inner membrane enzyme, was assayed as described by Morre (20). a-Ketoglutarate dehydrogenase, a mito- chondrial matrix enzyme, was measured by a radio- chemical assay in which radiolabeled a-ketoglutarate was generated in sitrc from [1-“Clglutamate. The as- say conditions were those developed for the measure- ment of pyruvate dehydrogenase activity by Wicking et al. (21), but with [1-“C]glutamate rather than [1-%]alanine as the substrate. /GHexosaminidase, a lysosomal matrix enzyme, was assayed fluorometri- tally by following the release of 4-methylumbellifer- one from 4-methylumbelliferyl-2-acetamido-2- deoxy-fi-D-glucoside (22). Arylsulfatase C and glu- case-6-phosphatase, microsomal enzymes, were assayed as described (18.23). Catalase, a peroxisomal matrix enzyme, was assayed spectrophotometrically (24). Fatty acyl CoA oxidase, another peroxisomal matrix enzyme, was measured by quantifying palmi- toy1 CoA-dependent production of H202. The detec- tion method was based on the conversion of nonfluo- rescent homovanillic acid to a highly fluorescent di- mer in the presence of free radicals (25). The protein content was estimated fluorometrically using human albumin as standard (26).

RESULTS

Succinate-INT reductase (mitochondrial inner membrane enzyme) and cr-ketogluta- rate dehydrogenase (mitochondrial matrix enzyme) activities were enriched in the crude mitochondrial fraction; catalase and palmitoyl CoA oxidase (peroxisomal ma- trix enzymes) and DHAP-AT and DHAP- synthase activities were enriched in crude peroxisomal fractions of rat liver (Fig. lj. Glucose-6-phosphatase and arylsulfatase C (microsomal membrane enzymes) activ- ities were enriched in microsomal frac- tions, and P-hexosaminidase (lysosomal

matrix enzyme) activity was enriched in crude mitochondrial and crude peroxi- somal fractions of rat liver (Fig. 1).

Purified mitochondrial fraction (iso- lated by Percoll density centrifugation of crude mitochondrial fraction) contained 93-96% mitochondria, purified peroxi- somal fraction (isolated by Nycodenz gra- dient centrifugation of crude peroxisomal fraction) contained 85-87% peroxisomes, and microsomal fraction (prepared by differential centrifugation of liver homog- enate) contained 9’7-99% microsomes (Ta- ble 1). Microsomal contamination in puri- fied mitochondria and purified peroxi- somes was <5 and <15%, respectively (Table I).

DHAP-AT activity in rat liver homoge- nates was linear with protein up to 20 pg per assay and at least up to 30 min incuba- tion (Fig. 2). Similarly, DHAP-synthase activity in rat liver homogenates was lin- ear up to 15 ,ug protein and at least up to 30 min incubation time, and the activity was at least lo-fold lower compared to DHAP- AT activity (Fig. 2). Analysis of DHAP-AT activity in purified preparations of mito- chondria, peroxisomes, and microsomes indicated that DHAP-AT activity was ab- sent in microsomes but present in mito- chondria and peroxisomes. The activity in peroxisomes was 50-fold greater compared to that in mitochondria (Table II). In com- parison with DHAP-AT, DHAP-synthase activity was detectable in all three organ- elles examined but the activity in mito- chondria was very low. DHAP-synthase activity in peroxisomes was 208-fold and 42-fold greater compared to that in mito- chondria and microsomes, respectively (Table II). In purified peroxisomes DHAP- AT activity was 23-fold greater than DHAP-synthase activity.

In human skin fibroblast homogenates, the DHAP-AT activity was linear at least up to 75 pg protein and up to 60 min incuba- tion (Fig. 3). DHAP-synthase activity in human skin fibroblast homogenates was linear with protein up to 30 pg per assay and at least up to 30 min incubation and the activity was several-fold lower than DHAP-AT activity (Fig. 3). Analysis of DHAP-AT and DHAP-synthase activities

680 SINGH, USHER, AND POULOS

600 -

5600- B t

F ;400- 1

$

?I E 200- c

T

40

I

160

a -Ketoglutarate Dehydrogenase c_ 1

600

Glucose-6-Phosphatase

Horn t&t Perox MC Sup

5 5 600 t F .

E 4oo

6 z E 200 =I

Catakise

d DHAP-Synthase

r

Patmltoyl CoA Oxtdase

5 5 600 a’ F ; 400

160

DHAP-AT

I”“1 r

Horn Mtt PerOx MIC SUP Horn Perox MIC SUP

FIG. 1. Crude subcellular fractions from rat liver were isolated as described under Materials and Methods. The enzyme assays were carried out in duplicate as described under Materials and Meth- ods, and the data are presented as mean values. Three separate experiments were conducted with similar results and the data are presented from one experiment. Horn, Mit, Peros, Mic, and Sup represent liver homogenate, crude mitochondria, crude peroxisomes, microsomes and 130,OOOy su- pernatant, respectively.

in human skin fibroblast homogenates of patients with several peroxisomal disor- ders indicates that there were different levels of residual act.ivities of the two en- zymes. For example, in Zellweger syn- drome fibroblasts there were <20% activ- ities of DHAP-AT and DHAP-synthase compared to controls (Table III). In con-

trast, in RCP patients there was greater reduction of DHAP-synthase (15% activ- ity compared to controls) activity com- pared to DHAP-AT (40% activity com- pared to controls) activity (Table III). In infantile Refsum’s disease fibroblasts much greater residual activities of both DHAP-AT and DHAP-synthase (40-45s

DHAP-AT AND DHAP-SYNTHASE ACTIVITIES 681

TABLE I

MARKERENZYMEACTIVITIESINRATLIVERSUBCELLULARFRACTIONS

Enzyme Purified

mitochondria Purified

peroxisomes Microsomes

Succinate-INT reductase (pmol. h-‘. mg protein-‘)

n-Ketoglutarate dehydrogenase (nmol. h-l. mg protein-‘)

Catalase (fimol. min-’ mg protein-‘)

Palmitoyl CoA oxidase (nmol. h-‘. mg protein-‘)

Glucose-6-phosphatase (pmol. h-‘. mg protein-‘)

Arylsulfatase C (nmol. h-‘. mg protein-‘)

/%Hexosaminidase (nmol. min-’ mg protein-‘)

Lysosomal marker Calculated

Mitochondrial markers Calculated

Peroxisomal markers Calculated

Microsomal markers Calculated

140.8 1.2 1.5

156.0 2.1 0.3

46.4 7843.1 76.2

20.1 3406.8 85.2

0.4 3.1 25.3

29.0 99.0 659.0

53.4 9.8 9.9

1.7% 0.3% 0.3%

93.4-96.2s 0%1.3% 0.2-1.0%

0.6% 83.7-86.97 0.8-2.1s

1.5-4.3%, 12.0-14.7%, 96.6-98.7’3

Note. Marker enzyme assays were performed in duplicate, as described under Materials and Methods and the data are given as mean values. Three separate experiments were carried out with similar results and the data from one experiment are presented.

activities, compared to controls) were ob- served compared to Zellweger syndrome. However, in X-linked ALD there was only a small reduction in DHAP-AT and DHAP-synthase activities, and the activit- ies of the two enzymes were normal in Ref- sum’s disease fibroblasts (Table III).

DISCUSSION

The production of 1-O-alkyldihydroxy- acetone phosphate is an obligatory step for the biosynthesis of ether lipids. The en- zymes involved in the introduction of l-O- alkyl bond in I-0-alkyldihydroxyacetone phosphate, namely DHAP-AT and DHAP- synthase, were found to be mainly present in rat liver peroxisomes. Apart from per- oxisomes, DHAP-AT activity was also de- tected in mitochondria but the specific ac-

tivity in mitochondria was 1/50th of that found in peroxisomes. The total amount of mitochondrial proteins in rat liver cells is 10 times greater compared to peroxisomal proteins (data not shown). Therefore, in rat liver cells maximal DHAP-AT in mito- chondria may be about 20% of total activ- ity found in the cell. Our results are in agreement with the calculations of De- Clercq et nZ. (lo), that under optimal assay conditions only a small proportion of rat liver DHAP-AT (assayed at pH 5.7) is ex- traperoxisomal (see Table VII of Ref. (10)). The presence of DHAP-AT activity in mi- tochondria suggests that the acyldihy- droxyacetone phosphate pathway might contribute significantly in the biosynthesis of phospholipids in mitochondria. The above observations support the earlier re- port where it was suggested that in rat liver some of the phospholipid biosynthe-

682

2400- (A)

SINGH, USHER, AND POULOS

DHAP-AT

DHAP-Synlhase

32

1 C) 20 1 KN

0 IO 20 30 40 PROTEIN (ug/reactlon)

0 10 20 30 40 INCUBATION TIME (mln)

FIGS. 2 AND 3. Dihydroxyacetone phosphate acyltransferase (A and B) and alkyldihydroxyacetone phosphate synthase (C and D) assays were carried out in duplicate using rat liver homogenate (Fig. 2) and control fibroblast homogenates (Fig. 3) as the enzyme source, and the data are presented as mean values. (2A) 30-min incubation; (2B) 69 ng protein per assay; (2C) 20-min incubation; (2D) 10 fig protein per assay; (3A) 60-min incubation; (3B) 75 ng protein per assay; (3C) 20-min incubation; (3D) 30 pg protein per assay.

sis occurs via the dihydroxyacetone phos- phate pathway (see Table IX of Ref. (10)).

The results presented suggest that in rat liver at least 80% of the DHAP-AT is asso- ciated with peroxisomes, and that the en- doplasmic reticulum (microsomes) does not contain DHAP-AT activity (assays carried at pH 5.4). The results on DHAP- AT activity in peroxisomes support the findings of Hajra et a.1. (27) who calculated that 90% of the DHAP-AT activity mea- sured in rat liver was peroxisomal, and

suggested that the remaining activity may be microsomal. However, they did not en- tertain the possibility that the remaining activity could be mitochondrial. Bell and associates have claimed previously that DHAP-AT activity in rat liver is present in microsomes (29, 30). It may be pointed out that they measured DHAP-AT activity at pH 7.0. Based on kinetic parameters, Schlossman and Bell (30) suggested that glycerol-&phosphate acyltransferase and DHAP-AT activity in rat liver microsomes

DHAP-AT AND DHAP-SYNTHASE ACTIVITIES 683

DHAP-AT

“‘-I (A)

800 - @I

600 - 600-

;; ti a g 400- a E a a

200-

20 40 SO 80 0 20 4b so 80

DHAP-Synthase

16

1 @)

t;

i t 5 E a

0 8 1 1 t 0 10 20 30 40

PROTEIN (uglreachon) 3

8-

0 :0

INCUBATION TIME (mm)

FICURE~

is due to a single enzyme. Microsomal glycerol-3-phosphate acyltransferase (28) which is involved in phospholipid biosyn- thesis may, to a certain extent, transfer fatty acids from fatty acyl CoA to dihy- droxyacetone 3-phosphate. Furthermore, two DHAP-AT activities with distinct pH optima, namely pH 5.5 and pH 7.5, have been described in rat liver (8,31). The pres- ence of two DHAP-AT activities has also been described in Chinese hamster ovary cells (32). The DHAP-AT activity detected in rat liver microsomes by Bell and associ- ates (29, 30) and DeClercq et aZ. (10) may be the pH 7.0 enzyme and not the pH 5.5 enzyme. Alternatively, the activity de-

tected in liver microsomes by Bell and co- workers (29,30) may be the result of break- age of peroxisomes and sedimentation of peroxisomes and peroxisomal membranes along with microsomes.

We confirm the earlier report (27) that DHAP-synthase activity is enriched in peroxisomes, and find comparable specific activity in peroxisomes (compare values in Table II with Table I of Ref. (27)). Hajra et al. (27) calculated that DHAP-synthase activity in microsomes and peroxisomes is equally distributed. However, Rabert et al. reported higher specific activity of DHAP- synthase in microsomes compared to per- oxisomes (see Table 3 of Ref. (11)). Fur-

684 SINGH, USHER, AND POULOS

TABLE II

DIHYDROXYACETONE PHOSPHATE ACYLTRANSFERASE (DHAP-AT) AND ALKYLDIHYDROXYACETONE PHOSPHATE (DHAP)-SYNTHASE ACTIVITIES IN RAT LIVER SUBCELLULAR FRACTIONS

Enzyme Purified

mitochondria Purified

peroxisomes Microsomes

DHAP-AT (nmol. h-l. mg protein’) Observed Calculated”

DHAP-synthase (nmol. h-‘. mg protein -‘) Observed Calculated”

36.1 1227.3 7.8 29.0 1440.5 0

0.7 53.1 2.4 0.3 62.3 1.5

hbte. The enzyme assays were performed in duplicate and the data are presented as mean values. Three separate experiments were conducted with similar results.

’ Calculated values represent enzyme activities expected if organelles were pure. Calculations were based on the activities of the marker enzymes observed in each organelle (see Table I) to adjust for the amount of protein and the enzyme activities contributed by the contaminating organelles.

thermore, they reported that the products 30 times lower specific activities than re- of DHAP-synthase reaction in microsomes ported here of DHAP-synthase in purified and peroxisomes were different, and in mi- preparations of peroxisomes (compare Ta- crosomes the major product was dephos- ble II with Table 3 of Ref. (11)). From the phorylated derivative, namely, l-o-alkyl- results described we calculate (on the basis dihydroxyacetone. We investigated the re- of total amounts of microsomal:peroxi- action product and found that under the somal protein ratio of lO:l, data not given) assay conditions described the product of that ~25% of the total DHAP-synthase ac- DHAP-synthase reaction in peroxisomes tivity in rat liver cell is microsomal. Sim- and microsomes was the same, namely, l- ilarly, we calculate that ~5% of the total 0-alkyldihydroxyacetone phosphate. We DHAP-synthase activity in the rat liver did not find any evidence of the presence of cell is mitochondrial. Overall >70% of the 1-0-alkyldihydroxyacetone. It. may be DHAP-synthase activity in rat liver is per- point.ed out that Rabert et al. found at least oxisomal. Furthermore, we find significant

TABLE III

DIHYDROXYACETONE PHOSPHATE AULTRANSFERASE (DHAP-AT) AND ALK~LDIHYDROXYACETONE PHOSPHATE (DHAPJ-SYNTHASE ACTIVITIES IN HUMAN SKIN FIBROBLAST HOMOGENATES

Diagnosis

DHAP-AT DHAP-synthase

(nmol. h-‘. mg protein-‘)

Control (n = 10) 6.43 f 2.93 1.67 + 0.61 Zellweger’s syndrome (n = 6) 1.03 f 0.37 0.35 f 0.12

Rhizomelic chondrodysplasia punctata (n = 4) 2.13 f 0.11 0.24 f 0.05 Infantile Refsum’s disease (7~ = 6) 2.58 + 0.64 0.74 * 0.35

Refsum’s disease (n = 3) 6.41 f 3.26 2.15 rt_ 0.26 X-linked adrenoleukodystrophy (n = 6) 4.39 f 0.36 1.17 f 0.38

Note. The assays for each cell line were performed in duplicate and experimental blanks without fibroblast homogenates carried out each time. The data represent means f standard deviation, and n is the number of cell lines tested.

DHAP-AT AND DHAP-SYNTHASE ACTIVITIES 685

amounts (29-37s of total) of DHAP-syn- thase activity in the high-speed (130,OOOg for 60 min) supernatant. In comparison, 31-33s of the total palmitoyl CoA oxidase was released in the high-speed superna- tant under those conditions. In contrast, DHAP-AT activity could not be detected in the high-speed supernatant. Therefore, we postulate that the localization of DHAP- AT and DHAP-synthase within peroxi- somes is different. As palmitoyl CoA oxi- dase (peroxisomal matrix enzyme) and DHAP-synthase are released under sim- ilar conditions, we suggest that in rat liver the peroxisomal DHAP-synthase is lo- cated either in the peroxisomal matrix or loosely attached to the peroxisomal mem- brane.

We confirm earlier observations that DHAP-AT and DHAP-synthase activities are deficient in skin fibroblast homoge- nates of Zellweger patients (5-7). The re- sults in skin fibroblast homogenates of Zellweger patients support the findings in rat liver and also provide indirect evidence that most of the activities of DHAP-AT and DHAP-synthase in skin fibroblasts are also peroxisomal. It mav be pointed out that we found higher activities of DHAP- AT but lower activities of DHAP-synthase in homogenates of normal skin fibroblasts than those reported earlier (compare val- ues in Table III with Table 3 of Ref. (,6) and Table 1 of Ref. (7)). However, Datta et al. (5) reported much higher activities of DHAP-AT and DHAP-synthase in homog- enates of control fibroblasts than those re- ported here (compare values in Table III with Table 1 of Ref. (5)). These differences in activities could not be interpreted as methodological differences because we find comparable activities of DHAP-AT and DHAP-synthase in purified preparations of rat liver peroxisomes as described ear- lier by the same group (27).

Our observations in skin fibroblasts of patients with RCP support earlier obser- vations that RCP patients have a partial deficiency of DHAP-AT (33). However, we found slightly higher residual DHAP-AT activity in RCP patients than those re- ported earlier (33). In addition to partial deficiency of DHAP-AT, DHAP-synthase

activity is deficient in RCP, and the reduc- tion in activity of DHAP-synthase is even greater than that observed in Zellweger syndrome (Table III). DHAP-AT and DHAP-synthase activities are reduced in infantile Refsum’s disease but the residual activities for both enzymes are substan- tially higher than those reported for two patients by Wanders et ul. (3435). Extrap- olating the observations on DHAP-AT and DHAP-synthase on rat liver to human skin fibroblasts, we speculate that at least some of the residual activities of DHAP-AT ob- served in RCP, ZS, and infantile Refsum’s disease is due to mitochondrial DHAP-AT. Similarly, some of the residual DHAP- synthase activities measured in RCP, ZS, and infantile Refsum’s disease is due to microsomal DHAP-synthase. We postu- late that the higher residual activities of DHAP-AT and DHAP-synthase in infan- tile Refsum’s disease compared to ZS is due to the presence of residual peroxi- somes in infantile Refsum’s disease. The above hypothesis is supported by our ear- lier observations on very long chain fatty acids (greater than 22 carbons in length) B-oxidation which is known to occur mainly in peroxisomes (18,37j. The partial reduction of DHAP-AT activity in infan- tile Refsum’s disease confirms our earlier report (36). In addition, we report that DHAP-AT and DHAP-synthase activities in Refsum’s disease were normal and those in X-linked ALD were near normal. From the results, we conclude that DHAP-AT and DHAP-synthase activities can be used as peroxisomal markers both in rat liver and in skin fibroblasts. Furthermore, the results presented suggest that peroxi- somal disease states can be distinguished based on DHAP-AT and DHAP-synthase activities in homogenates of skin fibro- blasts.

ACKNOWLEDGMENTS

The authors are thankful to Mrs. Greta Richardson and Mrs. Kathy Nelson for growth and maintenance of cell cultures, Dr. David Johnson for synthesizing l- O-palmitoyldihydroxyacetone 3-phosphate, and Ms. Peta Knapman and Ms. Sophie Lazenkas for typing the manuscript. This work was supported by grants from the Adelaide Children’s Hospital Research

686 SINGH, USHER, AND POULOS

Foundation and the National Health and Medical Re- search Council of Australia.

REFERENCES

1. HAJRA, A. K. (1968) J. BioL ChenL 243,3458-3465. 2. LABELLE, E. F., JR., AND HAJRA, A. K. (19’72) J.

BioL Chum. 247,5825-5834. 3. HAJRA, A. K. (1970) Biochen~. Bioph.ys. Res. Com-

mun. 39,1037-1043. 4. HEYMANS, H. S. A., SCHUTGENS, R. B. H., TAN, R.,

VAN DEN BOSCH, H., AND BORST, P. (1983) Nu- twre (London) 306,69-70.

5. DATPA, N. S., WILSON, G. N., AND HAJRA, A. K. (1984) New EngL J. Med 311,1080-1083.

6. SCHUTGENS, R. B. H., ROMEYN, G. J., WANDERS, R. J. A., VAN DEN BOSCH, H., SCHRAKAMP, G., AND HEYMANS, H. S. A. (1984) Biochem. Bic- phys. Res. Commun. 120,179-184.

7. SCHRAKAMP, G., ROSENBOOM, C. F. J., SCHUT- GENS, R. B. H.. WANDERS, R. J. A., HEYMANS, H. S. A., TAGER, J. M., AND VAN DEN BOSCH, H. (1985) J. Lipid Res. 26,867-873.

8. HAJRA, A. K., AND Wu, D. (1985) AnaL BiochenL 148,233-244.

9. GHOSH, M. K., AND HAJRA, A. K. (1986) AnaL Bie them. 159,169-174.

10. DECLERCQ, P. E., HAAGSMAN, H. P., VAN VELD- HOVEN, P., DEBEER, L. J., VAN GOLDE, L. M. G., AND MANNAERTS, G. P. (1984) J. BioL Chem. 259,9064-9075.

11. RABERT, U., VOLKL, A., AND DEBUC% H. (1986) BioL Chem. Hoppe-Seylor 367,215-222.

12. POULOS, A., SHEFFIELD, L., SHARP, P., SHER- WOOD, G., JOHNSON, D., BECKMAN, K., FEL- LENBERG, A. J., WRAITH, J. E., Qrow, C. W., USHER, S.. AND SINGH, H. (1988) J. Fediatr. 113, 685-690.

13. SCHAUMBERG, H. H., POWERS, J. M., .~ND RAINE, C. S. (1975) Arch. NeuroL 32,577-591.

14. Pou~os, A., SHARP, P., FELLENBERG, A. J., AND DANKS, D. M. (1985) Hum. Genet. 70,172-177.

15. POULOS, A., AND SHARP, P. (1984) Neurology 34. 1606-1609.

16. BUDDEN, S. S., KENNAWAY, N. G., BUIST, N. R. M., Pou~os, A., AND WELEBER, R. G. (1986) J. Fedi- atr. 108,33-39.

17. POLJLOS, A., POLLARD, A. C., MITCHELL, J. D., WISE, G., AND MORTIMER, G. (1984) Arch.. Dis. Child 59,222-229.

18. SINGH. H., DERWAS, N., AND Pou~os, A. (1987) Arch Biochem. Biophys. 259,382-390.

19. HAJRA, A. K., SARASWATHI, T. W., AND DAS, A. K. (1983) C&m. Fhys. Lipids 33,179-193.

20. MORRE, D. J. (1971) in Methods in Enzymology (Jakoby, W. B., Ed.), Vol. 22, pp. 134-148, Aca- demic Press, New York.

21. WICKING, C. A., SCHOLEM, R. D., HUNT, S. M., AND BROWN, G. K. (1986) B&hem. J 239,89-96.

22. KOLODNY, E. H., AND MUMFORD, R. A. (1976) Clin Chim. Acta 70,247-257.

23. RICKETTS, T. R. (1963) Clin. Chim. Acta 8,160-162. 24. BERGMEYER, H. U., GAWEHN, K., AND GRASSL, M.

(1974) in Methods of Enzymatic Analysis (Bergmeyer, H. U., Ed.), 2nd ed., Vol. 1, pp. 425- 522, Academic Press, New York.

25. GUILBAULT, G. G., BRIGMAC, P., JR., AND ZIMMER, M. (1968) Anal. Chem. 40,190-196.

26. UDENFREIND, S., STEIN, S., BOHLEN, P., DAIRMAN, W., LEIMGRUBER, W., AND WEIGELE, M. (1972) S&ewe 178,871-872.

27. HAJRA, A. K., GHOSH, M. K., WEBBER, K. O., AND DATTA, N. S. (1986) in Enzymes of Lipid Me- tabolism II (Freysz, L., Dreyfus, H., Massa- relli, R., and Ghatt, S., Eds.), pp. 199-207, Ple- num, New York.

28. BALLAS, L. M., AND BELL, R. M. (1980) B&him. Biophys. Acta 602,578-590.

29. BALLAS, L. M., AND BELL, R. M. (1981) Biochim Biophys. Acta 665,586-595.

30. SCHLOSSMAN, D. M., AND BELL, R. M. (1977) Arch. Biochem. Biophys. 182,732-742.

31. BALLAS, L. M., LAZ~ROW, P. B., AND BELL, R. M. (1984) B&him. Biophys. Acta 795,297-300.

32. ZOELLER, R. A., AND RAETZ, C. R. H. (1986) Froc. NatL rlcad. Sci. USA 83,5170-5174.

33. HEYMANS, H. S. A., OORTHUYS, J. W. E., NELCK, G., WANDERS, R. J. A., AND SCHUTGENS, R. B. H. (1985) NezLv EngL J. Med 313.187-188.

34. WANDERS, R. J. A., SCHUTGENS, R. B. H., SCHRA- KAMP, G., VAN DEN BOSCH, H., TAGER, J. M., POLL-THE, B. T., AND SAUDUBRAY, J. M. (1986) J Inherited Metab. Dis. 9, Suppl. 2,325-328.

35. WANDERS, R. J. A., SCHUTGENS, R. B. H., SCHRA- KAMP. G., VAN DEN BOSCH, H., TAGER, J. M., SCHR~M, A. W., HASHIMOTO, T., POLL-THE, B. T., AND SAUDUBRAY, J. M. (1986) Eur. J. Fed- iutr. 145,172-175.

36. VAN CRUGTEN, J. T., PATON, B., AND POULOS, A. (1986) J. Ixherited Metab. Dis. 9,163-168.

37. SINGH, H., DERWAS, N., AND POULOS, A. (1987) Arch. B&hem. Biophys. 254,526-533.