Defective Peroxisomal Cleavage of the C27-Steroid Side Chain

Transcription

1 Defetive Peroxisomal Cleavage of the C27-Steroid Side Chain in the Cerebro-Hepato-Renal Syndrome of Zellweger Pedersen Institute for Nutrition Researh, Shool ofmediine, University of Oslo, Oslo 3, Norway; Department of Clinial Chemistry and Researh Center at Huddinge University Hospital, Karolinska Institutet, Stokholm, Sweden; and Department of Pediatris, Ullevil Hospital, Oslo 1, Norway Bengt Frode Kase, Ingemar Bjorkhem, Per High, and Jan 1. Abstrat Based on in vitro work with rat liver, we reently suggested that the peroxisomal fration is most important for the oxidation of 3a,7a,12a-trihydroxy-5,8-holestanoi aid (THCA) into holi aid. The erebro-hepato-renal syndrome of Zellweger is a fatal reessive autosomal disorder, the most harateristi histologial feature of whih is a virtual absene of peroxisomes in liver and kidneys. This disease offers a unique opportunity to evaluate the relative importane of peroxisomes in bile aid biosynthesis. A hild with Zellweger syndrome was studied in the present work. In aordane with previous work, there was a onsiderable aumulation of THCA, 3a,7a,12a,24-tetrahydroxy-5,6-holestanoi aid (24-OH-THCA), 3a,7a,12a-trihydroxy-27-arboxymethyl-5#6-holestan-26-oi aid (C2-diarboxyli aid), and 3a,7a-dihydroxy-5#-holestanoi aid in serum. In addition, a tetrahydroxylated 51e-holestanoi aid with all the hydroxyl groups in the steroid nuleus was found. 3H-Labeled 5W-holestane-3a,7a,12a-triol was administered intravenously together with 14C-labeled holi aid. There was a rapid inorporation of 3H in THCA and a slow inorporation into holi aid. The speifi radioativity of 3H in THCA was about one magnitude higher than that in holi aid. The onversion was evaluated by following the inreasing ratio between 3H and 14C in biliary holi aid. The rate of inorporation of 3H in holi aid was onsiderably less than previously reported in experiments with healthy subjets, and the maximal onversion of the triol into holi aid was only 15-2%. About the same rate of onversion was found after oral administration of 3H-THCA. Both in the experiment with 3H-5W6holestane-3a,7a,12a-triol and with 3H-THCA, there was an effiient inorporation of 3H in the above unidentified tetrahydroxylated 5#-holestanoi aid. There was only slow inorporation of radioativity into 24-OH-THCA and into the C2-diarboxyli aid. From the speifi ativity deay urve of "'C in holi aid obtained after intravenous injetion of 14Choli aid, the pool size of holi aid was alulated to be 24 mg/m2 and the daily prodution rate to 9 mg/m2 per d. This orresponds to a redution of -85 and 9%, respetively, when ompared with normal infants. It is onluded that liver peroxisomes are essential in the normal onversion of THCA to holi aid. In the Zellweger syndrome this onversion is Address orrespondene to Dr. Pedersen, Institute for Nutrition Researh. Reeived for publiation 9 January 1984 and in revised form 12 Otober J. Clin. Invest. The Amerian Soiety for Clinial Investigation, In /85/2/427/9 $ 1. Volume 75, February 1985, defetive, and as a onsequene the aumulated THCA is either exreted as suh or transformed into other metabolites by hydroxylation or side hain elongation. The aumulation of THCA, as well as the similar rate of onversion of 5#- holestane-3a,7a,12a-triol and THCA into holi aid, support the ontention that the 26-hydroxylase pathway with intermediate formation of THCA is the most important pathway for formation of holi aid in man. Introdution The erebro-hepato-renal syndrome of Zellweger is an autosomal reessive disorder that is fatal within the first year of life (1). The virtual absene of peroxisomes in liver and kidney has been desribed as the most onsistent histologi finding (2). Inreased amounts of C27-bile aid intermediates have been deteted in bile, serum, and urine of the patients (3-6). These intermediates inlude 3a,7a,dihydroxy-5(3-holestan-26-oi aid (DHCA),' 3a,7a, 1 2a-trihydroxy-5/3-holestan-26-oi aid (THCA), and 3a,7a, 12a,24-tetrahydroxy-5f3-holestan-26-oi aid (24-OH-THCA) (4-6). The aumulation of bile aid preursors with a partially oxidized side hain has been attributed to a mitohondrial defet in the leavage of the steroid side hain (4, 6). Aording to urrent onepts, the degradation of the steroid side hain in the biosynthesis of holi aid starts with a mitohondrial 26-hydroxylation and subsequent oxidation to THCA (for review see referene 7). The further onversion seems to involve 24-OH-THCA and 3a,7a, 12a-trihydroxy-24- on-5,b-holestanoi aid as intermediates, and propioni aid is leaved off in the final step. An alternative pathway has been proposed involving mirosomal 25- and 24S-hydroxylation Subsequent oxidation in this of 5#-holestane-3a,7a,12a-triol. pathway yields 3a,7a,12a,25-tetrahydroxy-5,3-holestan-24-one whih is leaved by ytosoli enzymes to holi aid and aetone (8). Aumulation of THCA in patients with Zellweger syndrome would lend support to the ontention that the major pathway in holi aid formation involves 26-hydroxylation and intermediate formation of THCA. We have reently shown that in rat liver the formation of holi aid from THCA is most effiiently arried out by the peroxisomal fration (9, 1). 1. Abbreviations used in this paper: DHCA, 3a,7a-dihydroxy-5,3- holestan-26-oi aid; GC-MS, gas hromatography-mass spetrometry; HPLC, high pressure liquid hromatography; 24-OH-THCA, 3a,7a, 12a,24-tetrahydroxy-5,B-holestan-26-oi aid; THCA, 3a,7a, 12atrihydroxy-5fl-holestan-26-oi aid; C29-diarboxyli aid, 3a,7a,12atrihydroxy-27-arboxymethyl-5#-holestan-26-oi aid. Defetive Peroxisomal Formation of Choli Aid in Zellweger Syndrome 427

2 Zellweger syndrome appears to offer a unique opportunity to evaluate the importane of liver peroxisomes in bile aid formation. In addition, the degree of onversion of different bile aid intermediates into THCA in suh patients may give information with respet to the relative importane of the two alternative pathways desribed for biosynthesis of holi aid in man. We here report a study on the onversion in vivo of 5(tholestane-3a,7a,12a-triol and THCA into holi aid in a patient with Zellweger syndrome. No peroxisomes ould be found in the liver of this patient. The limited apaity for onversion of these preursors into holi aid in vivo suggests that the metaboli defet resides at the peroxisomal level and that under normal onditions the major part of holi aid formation proeeds via the peroxisomal pathway. Methods Clinial. The hild was a fuilterm female of healthy nononsanguineous parents. In mother's family several members had epilepsy. Musular hypotonia prevailed from birth. She displayed a raniofaial dysmorphy harateristi of Zellweger syndrome (1). There was pes equino varus and four finger line bilaterally. With time, hepatomegaly developed and onvulsions appeared whih inreased in frequeny and duration. She was treated with 5 mg/kg phenobarbital. This floppy infant died at 4-mo-old. Cerebral omputer tomography was normal while EEG was pathologi. The levels of alanine aminotransferase and aspartate aminotransferase were highest shortly after birth (aspartate aminotransferase 1,164 U/L [normal range U/L], alanine aminotransferase 423 U/L [14-84 U/L]) and delined towards the final stage. Gamma GT (6 U/L [1-35 U/L]) and alkaline phosphatases (1,7-2, U/L [275-1,5 U/LI) remained high. Total bilirubin in serum was normal. Serum protein eletrophoresis was normal. Normotest values gradually dereased, but were almost normal at the time of the in vivo study (6% [7-13%]). Serum holesterol was 6.7 mmol/ml ( mmol/ ml) while triglyerides were in the normal range. No pipeolate was detetable in urine. Cultured fibroblasts ontained high amounts of saturated long hain fatty aids (11). At autopsy the liver was found enlarged with normal extrahepati bile duts. The liver surfae was uneven, and mirosopially marked interlobular septae and large periportal areas were seen. The lobular arhiteture was intat and irrhosis was not deteted. By eletron mirosopy the ellular arhiteture appeared normal with inreased amounts of glyogen. Normal peroxisomal strutures ould not be reognized. In the entral nervous system neuronal migration arrest typial for Zellweger syndrome (12) was observed. The ethial aspets of the present study were approved by the ethial board of the Norwegian Counil for Siene and the Humanities. Preparation of labeled steroids. 7f3-3H-5ft-holestane-3a,7a,l2atriol (7 Ci/mol) and 7#_3PH-THCA (2 Ci/mol) were prepared as desribed (13, 14). Separation on high pressure liquid hromatography (HPLC) of the latter ompound (1) showed that it ontained 28% of the 25R and 72% of the 25S isomer. 24-"C-Choli aid (5 Ci/mol) was from Amersham International pl, Amersham, England. The ompounds were purified by HPLC (see below) before use. In vivo administration oflabeled steroids and olletion ofbile and serum samples. The tritium-labeled triol (13.4 X 1 dpm) and the '"C-labeled holi aid (1.5 X 16 dpm) were dissolved in.5 ml of ethanol. The solution was passed through a Millex.22-Aim filter (Millipore Co., Bedford, MA) and slowly added under shaking to 5 ml of sterile human albumin solution (albumin 2%, Kabi AB, Stokholm, Sweden). The mixture was slowly infused intravenously during 3 min. Duodenal intubation was performed under X-ray visualization. Duodenal ontents were sampled at least 2 h after eah meal (mother's milk), 3, 6, 24, and 48 h after the infusion. The amount of aspirated duodenal fluid was small, eah sample -1 ml to avoid disturbanes in the pools of the bile aids. Venous blood samples were olleted after 24 and 48 h. Urine was olleted after 15 h. 1 mo later the patient reeived 3H-labeled THCA, 13 X 16 dpm, and "4C-labeled holi aid, 1.5 X 16 dpm, dissolved in 18,ul ethanol, mixed into the milk, and given through a gastri tube. Blood samples were olleted 6, 12, and 24 h after the meal and duodenal ontent was aspirated 11 h later. Extration, hromatography, and analytial proedures. The serum samples (.25 ml) were hydrolyzed and extrated as desribed previously (I5, 16). In this extration proedure, most but not all holesterol is removed. To the serum samples used for assay by isotope dilutionmass spetrometry, deuterium-labeled holi aid, henodeoxyholi aid, and deoxyholi aid were added before hydrolysis (15, 16). The methyl ester trimethylsilyl ether derivatives were prepared before ombined gas hromatography-mass spetrometry (I5, 16). An LKB 9 instrument equipped with a multiple ion detetor and an 1.5% SE-3 olumn was used (15, 16). The amount of holi aid was measured by use of the ions at mle 623 and 628 (orresponding to the M-15 ion in the mass spetrum of derivative of unlabeled and 2H5- labeled holi aid, respetively [16]). The amount of henodeoxyholi aid was measured by use of the ions at mle 37 and 374 (orresponding to the M-2 X 9 ion in the mass spetrum of derivative of unlabeled and 2H-labeled henodeoxyholi aid, respetively). The amount of THCA was measured by use of the ions at mle 41 and mle 374 (orresponding to the M-3 X 9 ion in the mass spetrum of derivative of unlabeled THCA and the M-2 X 9 ion in the mass spetrum of the derivative of 2H4-henodeoxyholi aid). The bile, serum, and urine samples olleted after the in vivo administration of the radioative ompounds were hydrolyzed and extrated as above. Aliquots of the extrats were separated by HPLC using a Zorbax ODS olumn (5 X 25 mm, partile size 5 Arm). Phosphori aid/potassium dihydrogen phosphate, 25 mm, ph 3.4, 24% in methanol was used as solvent at a flow rate of 1 ml/min. (Several olumns were used. The retention times shown in the figures below may therefore vary from one hromatogram to another.) l-ml frations were olleted. Parts of the frations were evaporated and after addition of ounting solution ounted at an effiieny of 82% for 14C and 58% for 3H in a liquid sintillation spetrometer (Pakard Tri- Carb 3C). External standardization was used for determination of quenhing and the energy region settings were automatially optimized. The data were orreted for "4C ativity ounted in the tritium hannel. Frations ontaining radioativity were extrated twie with 9 ml ethyl aetate after aidifiation with HCL. The ethyl aetate was washed twie with water, taken to dryness under N2, and the residue dissolved in a small volume of methanol. The extrats were onverted to the methyl ester trimethylsilyl ether derivatives (15, 16) and analyzed by ombined gas hromatography-mass spetrometry (GC-MS) as desribed above. For identifiation of the different steroids, harateristi ions were followed by multiple ion detetion. In a few ases in whih suffiient amounts of material ould be obtained, a full mass spetrum was reorded (refer above and to Results). Assumptions and alulations. It was assumed that no loss of 3H ourred during the metabolism of 7ft-3H-labeled preursors of holi aid. The justifiation of this assumption is supported by a reent study with a mixture of 7-3H- and 24-"C-labeled holi aid. There was only a negligible loss of 3H from the holi aid during the enterohepati irulation in man (Bjorkhem, I., K. Einarsson, and K. Nilsell, unpublished observation). The onversion of the 3H-labeled triol and the 3H-labeled THCA into holi aid was based on the ratio between 3H and 14C in the mixture injeted into the patient, whih was ompared with the ratio between 3H and 14C in the holi aid isolated. It is then assumed that 1% onversion of the 3H-labeled preursor should yield holi aid with a ratio 3H/"4C, idential with that in the material injeted. The validity of this experimental approah is evident from results of a similar study by Hanson et al. (17, 18). It was reported that after 428 B. F. Kase, I. Bjdrkhem, P. Hdgd, and J. I. Pedersen

3 intravenous administration of 3H-THCA or 3H-5-holestane- 3a,7a,12a,26-tetrol together with '4C-holi aid to normal subjets, the 3H-holi aid speifi ativity deay urves were within 81-91% of the '4C-holi aid speifi ativity deay urves. Furthermore, in an in vivo study on another Zellweger patient that was performed after the ompletion of this work, we obtained a larger number of bile samples and were thus able to ompare our method with that used by Hanson and Williams (17). The perent onversion alulated by our method was 2% higher than alulated by the method of Hanson and Williams (17). From the deay of the speifi radioativity of '4C in holi aid in bile isolated after the intravenous administration of '4Choli aid, the pool size and half life of holi aid was determined aording to the Lindstedt tehnique (19). Results Serum bile aids. Fig. 1 A shows multiple ion reordings of the ion at m/e 253 (orresponding to the trihydroxyoprostanoi nuleus) and at m/e 255 (orresponding to the dihydroxyoprostanoi nuleus) obtained in analysis of a hydrolyzed and derivatized sample of a serum extrat from the patient. From the pattern obtained, it is evident that there was a predominane of trihydroxy bile aids. In the reording of the ion at m/e 253, peaks ourred with retention times idential to derivatives of holi aid, THCA, and 3a,7a,l2a,24-tetrahydroxy-5,3- holestan-26-oi aid (24-OH-THCA). In addition there was a prominent peak with a retention time typial for that of the 3a,7a, 12a-trihydroxy-27-arboxy-methyl-5(3-holestan-26-oi aid (C29-diarboxyli aid) desribed by Parmentier et al. (5). Evidene has been given that the struture of this aid is 3a,7a, 12a - trihydroxy arboxymethyl - 5(3 - holestan oi aid (2). In addition, a peak ourred with a retention time slightly longer than that of 24-OH-THCA in the reording of the ion at m/e 253. The identity of this ompound ould never be established. The peak ourring in the traing of mle 255, just in front of the henodeoxyholi aid peak, orresponds to holesterol. The trimethylsilyl ether of methyl henodeoxyholi aid partly ohromatographed with deoxyholi aid under the onditions employed. Under slightly different hromatographi onditions it was shown that there was no derivative of deoxyholi aid. There were signifiant amounts of DHCA but only traes of 24-OH-DfiCA. In a separate reording of the ion at m/e 372 (orresponding to the M-9 ion in the mass spetrum of derivative of lithoholi aid) it was shown that there were only traes of lithoholi aid. In view of the relatively small amounts of serum available, it was not possible to reord full mass spetra. The identity of holi aid was established, however, by seleted reording of the ions at mle 623 (M-1 5) and m/e 368 (M-3 X 9). The identity of henodeoxyholi aid was also onfirmed by the traing of the ion at mle 37 (M-2 X 9). The identity of THCA was onfirmed by the traing of the ions at m/e 5 and mle 41 (M-2 X 9 and M-3 X 9 in the mass spetrum of derivative of THCA), and that of 24-OH-THCA by the ions at m/e 588 and mle 498 (M-2 X 9 and M-3 X 9 in the mass spetrum of derivative of 24-OH-THCA). The identity of the C29-diarboxyli aid was established by use of the ions at mle 482 (M-3 X 9) and mle 737 (M-15) (5). The onentrations of holi aid, henodeoxyholi aid, and THCA were determined by isotope dilution-mass spetrometry using 2H5-labeled holi aid and henodeoxyholi aid as internal standards (16). The approximate onentration of 24-OH-THCA and C29-diarboxyli aid was determined from the traing at mle 253. It was assumed that the relative A B 4 U e CL a C,' U e 1 CL a m/e 253 Ak, m/e 253 p We 255 5I I I I 1I I 5 1 Retention time ( min ) Retention time ( min ) Figure 1. (A) Multiple ion detetor reording of trimethylsilyl ether methyl ester derivative of an extrat of serum from the patient at the age of 1 wk. C, holi aid; CD, henodeoxyholi aid. (B) Similar reording from a healthy infant at the same age. Defetive Peroxisomal Formation of Choli Aid in Zellweger Syndrome 429

4 intensity of the ion at m/e 253 was similar in the mass spetrum of THCA, 24-OH-THCA, and C29-diarboxyli aid. The results of these determinations of serum samples taken at different oasions are summarized in Table I. For the reasons of omparison, a seleted ion monitoring analysis of a serum extrat from a healthy infant of the same age is shown in Fig. 1 B. As expeted, only peaks orresponding to holi aid and henodeoxyholi aid were seen. When ompared with normal fasting referene values (21, 22), the serum onentrations of holi aid and henodeoxyholi aid ould be regarded as normal. During the further progression of the disease, both henodeoxyholi and holi aid remained within normal limits (Table I). Deoxyholi aid was not deteted in most samples. The preursors of the primary bile aids and the C29-diarboxyli aid that were measured in high onentrations are normally not deteted in infant serum. Biliary bile aids. Using the same isotope dilution tehnique as in the analyses of the serum bile aids, it was shown that the omposition of bile aids in a sample of bile from the patient at the age of 1 wk was the following: holi aid 42%, henodeoxyholi aid 53%, THCA 4%, 24-OH-THCA 1%, and C29-diarboxyli aid %. Normally, only holi aid and henodeoxyholi aid an be expeted to our in the bile of an infant of this age, and the relative amount of holi aid an be expeted to vary between 2 and 6% (23, Bjorkhem, I., unpublished observation). Conversion in vivo of 7,3-3H-S3-holestane-3a, 7a, 12a-triol into bile aids. After administration in vivo of 3H-labeled 513- holestane-3a,7a,122a-triol and '4C-holi aid, bile, serum, and urine samples were hydrolyzed and extrated as desribed in Methods. The ratio of 3H to 14C ativity of the infused material (in dpm) was 8.9. In the serum extrats this ratio was 13.3 in the 24-h sample and 15.8 in the 48-h sample. In bile extrats this ratio fell from 3.8 to 2.9. The bile extrats were analyzed by HPLC, and the hromatogram of the extrat of the sample olleted 3 h after the infusion is shown in Fig. 2. Two major peaks of radioativity Table I. Serum Bile Aid Conentrations from the Age Diagnosis Established Until the Child with Zellweger Syndrome Died Age (wk) Bile aids 4 1* LmoIIL MmoK/L zmol/l JAmol/L Choli aidt Chenodeoxyholi aidt Deoxyholi aidt..2.. THCA DHCA OH-THCA a,7a,24-Tsihydroxy-5jholestanoi aid a,6a,7a, 1 2a-Tetrahydroxy- 5,Bholestanoi aid C29-diarboxyli aid * At this age, 7-.3H-5,-holestane-3a,7a,12a-triol was infused. f Normal values for holi aid, henodeoxyholi aid, and deoxyholi aid in healthy infants are.6-18 Mmol/L, gmol/l, and <.2 Amol/L, respetively, at the age of 1 mo, and.3-2 Amol/L, gmol/l, and <.1,smol/L, respetively, at the age of 3 mo (22). Tentatively identified (see text). I a) L- Q I-.2.O- (._ : I Elution volume (ml) Figure 2. Reversed phase high pressure liquid hromatogram of a duodenal bile extrat 3 h after infusion of 3H-53-holestane- 3a,7a,12a-triol and '4C-holi aid. Hydrolysis, extration, and HPLC proedures are given in Methods. The main peaks orrespond to the following: 3a,6a,7a,12a-tetrahydroxy-51-holestanoi aid (tentatively identified, see text), 14 ml, holi aid, 18 ml, 24-OH- THCA, 2 ml, 25R-THCA, 45 ml, and 25S-THCA, 48 ml. For additional information see text. were deteted. The more polar of these, that eluted at 18 ml, ontained both 3H and 14C ativity. The retention time was idential to that of holi aid. The presene of holi aid in this material was onfirmed by GC-MS using seleted ion monitoring of the ions at m/e 253, mle 368, and mle 623. The largest and most nonpolar splitted peak that eluted from 44 to 53 ml had the same retention time as 25R and 25S THCA. Only tritium ativity was deteted in this peak. The presene of THCA in this material was onfirmed by GC-MS, using the ion at m/e 253, m/e 41, and m/e 5. Several smaller tritium-ontaining peaks were also deteted in the hromatograms. The peak that eluted at 14 ml ontained material identified by GC-MS as a derivative of THCA with a hydroxyl group in the steroid nuleus. Seleted ion monitoring showed the presene of an ion at m/e 251 (loss of four trimethylsilyl groups from the steroid nuleus) as well as presene of ions at m/e 498 (M-3 X 9) and m/e 588 (M-2 X 9). (For tentative identifiation see below.) A small peak that eluted at 2 ml ontained material that was identified as 24-OH-THCA using the ions at 253, m/e 498 and mle 588. The material orresponding to the small peak that eluted at ml was never identified. Reovery from the olumn was essentially omplete in this hromatographi system and no unmetabolized 3H-5f3-holestane-3a,7a,12a-triol was deteted. The hromatogram in Fig. 2 showed that 3H-THCA aounted for 65% of total reovered ativity in the 3-h bile sample. This relative ativity delined almost linearly to 35% after 48 h. This finding refleted rapid onversion of 3H-triol into 3H-THCA and a high degree of retention of the formed 3H-THCA. After 24 and 48 h, the 3H/4'C ratio in holi aid was 15% of the infused material (Fig. 3). The value of 2% found after 35 h was based on very low ativity (-2 dpm for tritium) and might therefore be unertain. i 4). > (- CT 6 43 B. F. Kase, I. Bjorkhem, P. Hdgd, and J. I. Pedersen

5 C U Y U 12 o Time (hours) Figure 3. Time ourse of holi aid formation after infusion of 3H- 5#3-holestane-3a,7a,l2a-triol and '4C-holi aid. The values plotted are derived from hromatograms as shown in Fig. 2. The ratio of 3H/ 14C in the holi aid peak fration (elution volume 18 ml) has been expressed as perentage of the 3H/'4C ratio of the infused 3H-triol and '4C-holi aid tritium ativity. Fig. 4 shows the speifi radioativity deay urves of 14Choli aid, 3H-holi aid, and 3H-THCA. From the speifi radioativity deay urve of '4C-holi aid, it ould be alulated that the pool size of holi aid in the patient was only 24 mg/m2. The syntheti rate of holi aid (19, 24) was 9 mg/m2/d. The speifi radioativity of 3H in holi aid inreased between 3 and 6 h and then dereased. Due to the small amounts of radioativity, the auray in the latter measurement may be low, and no speifi information an therefore be obtained from the deay of 3H speifi radioativity between 6 and 24 h. The speifi radioativity of 3H-THCA was about one magnitude higher than that of 3H-holi aid and remained onstant between 3 and 24 h. The most likely explanation of this fairly onstant speifi ativity is a ontinuous formation of 3H-THCA from an intrahepati pool of 3H-triol. The serum samples drawn 24 and 48 h after the infusion of 3H-5,3-holestane-3a,7a,12a-triol and 14C-holi aid were hydrolyzed and extrated as desribed in Methods. The extrats were separated on HPLC and the material in the tritiumontaining peaks was analyzed by GC-MS as above. The HPLC profile of the extrat of the serum sample taken 24 h after the infusion is shown in Fig. 5. The least polar splitted peak (elution volume 34-4 ml) was shown to ontain THCA. A major tritium-ontaining peak eluted in front of holi aid (elution volume 12 ml). The material had the same elution volume relative to holi aid as that in the orresponding hromatogram obtained in the analysis of the bile samples (refer above). Smaller radioative peaks were shown to ontain holi aid (elution volume 15 ml) and 24-OH-THCA (elution volume 19 ml). The material that eluted at 2 ml was identified as the C29-diarboxyli aid. A full mass spetrum was thus idential with that reported by Parmentier et al. (5) and ontained prominent peaks at 253, 281, 343(M-2 X 9-229), 482(M-3 X 9), 572(M-2 X 9), and 662(M-9). The peak that eluted at 5 ml was not identified. A small amount of the injeted radioativity was not reovered from the olumn in this hromatographi system. When the samples were analyzed on the same Zorbax ODS olumn but eluted with 19% aetate (ph 4.37) in methanol (8), an additional peak amounting to 15% of the tritium ativity was obtained. The retention time was idential to that of 5,3-holestane-3a,7a, 12a,26-tetrol. The material in this peak was not suffiient for omplete identifi- 4o 4 - IN E -W.4.._Q (I H-THCA - LI I- 4) Q. 'a (o I (7) 8. _. 6 4_. 4 -, 2 I u 2- b 3H-Choli Aid I ~ 14C-Choli Aid Time ( h ) Figure 4. Speifi radioativity deay urves of 3H-THCA, 3H-holi aid and '4C-holi aid obtained after intravenous administration of a mixture of 3H-5#-holestane-3a,7a,12a-triol and '4C-holi aid. Elution volume (ml) Figure 5. Reversed phase high pressure liquid hromatogram of serum extrat 24 h after infusion of 3H-5fl-holestane-3a,7a, 12a-triol. Hydrolysis, extration, and HPLC proedures are given in Methods. The most polar peak with elution volume 5 ml was not identified. Other peaks orrespond to the following: 3a,6a,7a,12a-tetrahydroxy- 5#-holestanoi aid (tentatively identified, see text), 12 ml, holi aid, 15 ml, 24-OH-THCA, 19 ml, 25R and 25S-THCA, 35 and 37 ml, respetively. C29-diarboxyli aid was identified in the fration that eluted at 2 ml. For additional information see text. Defetive Peroxisomal Formation of Choli Aid in Zellweger Syndrome 431

6 ation, but the presene of a trihydroxy holestane nuleus was onfirmed by seleted ion monitoring of the ion at m/e 253. The tritium ativity orresponding to THCA amounted to 35 and 26% of the total ativity reovered after 24 and 48 h, respetively. In the overnight urine sample olleted 15 h after the infusion the 3H ativity was 5,5 dpm/ml. After extration and hromatography, this ativity ould entirely be aounted for as THCA or more polar produts. In partiular, 55% of the ativity was reovered in frations orresponding to tetrahydroxylated 5fl-holestanoi aid with all hydroxyl groups in the steroid nuleus (see below). Conversion in vivo of 7,3-3H-THCA into bile aids. After administration per os of 3H-THCA and '4C-labeled holi aid, bile and serum samples were hydrolyzed, extrated, and analyzed by HPLC as desribed in Methods. The material in the tritium-ontaining peaks was identified by GC-MS as desribed above. The HPLC profile of the extrat of the bile olleted 11 h after administration of the radioative ompounds is shown in Fig. 6. About 5% of the tritium ativity was reovered in the peak orresponding to THCA (elution volume ml), and most of the remaining was reovered in the holi aid fration. The material in the small peak that eluted in front of holi aid (elution volume i 1-12 ml) was not identified, but orresponded to the similar peak observed in the hromatogram of the serum extrat (see below). The ratio of 3H to 14C in the holi aid peak was 7.7% of that in the radioative mixture given per orally. This shows that after 11 h the formation of holi aid from THCA is onsiderably redued ompared with what is found when 3H-THCA is administered to normal subjets (17). In the hromatogram of the extrat of the pooled serum samples (Fig. 7), a peak orresponding to THCA was identified as above (elution volume ml). The major tritiumontaining peak that eluted in front of holi aid (elution 5-fl- volume ml) was identified as a tetrahydroxylated holestanoi aid, with all hydroxyl groups in the steroid nuleus. A full mass spetrum of this ompound ontained I CK 4, L- u a) CLI.. -6 'I Ix Elution volume (ml) Figure 6. Reversed phase high pressure liquid hromatogram of duodenal bile extrat 11 h after the administration of 3H-THCA per os. Hydrolysis, extration, and HPLC proedures are given in Methods. The main radioative peaks orrespond to the following: 3a,6a,7a,12a-tetrahydroxy-5,-holestanoi aid (tentatively identified, see text), ml, holi aid, 15 ml, 24-OH-THCA, 21 ml, and THCA, 43 ml. For additional information see text. a ai '4._ '4.O- a: u._ t 13 (a LI L. Q 'a -I Elution volume (ml) Figure 7. Reversed phase high pressure liquid hromatogram of pooled serum extrats 6, 12, and 24 h after the administration of 3H- THCA per os. Hydrolysis, extration, and HPLC proedures are given in Methods. The main radioative peaks orrespond to the following: 3a,6a,7a,12a-tetrahydroxy-513-holestanoi aid (tentatively identified, see text), 13 ml, holi aid, ml, 24-OH- THCA, 21 ml, C29-diarboxyli aid, 21 ml, 25R- and 25S-THCA, 48 and 51 ml. prominent ions at m/e 251 (loss of four trimethylsilyl groups from the steroid nuleus), 341( ), 393(M-4 X 9-15), 48(M-4 X 9), 498(M-3 X 9), and 588(M-2 X 9). The ompound is thus a tetrahydroxylated 5,B-holestanoi aid with all the hydroxyl groups in the steroid nuleus. Smaller peaks were identified as holi aid (elution volume ml), 24-OH-THCA (elution volume 21 ml), and the C29- diarboxyli aid (elution volume ml). Disussion Inreased amounts in serum and bile of C27-steroids with a partially oxidized side hain is a harateristi finding in Zellweger syndrome (4-6). This was onfirmed in the present work (Fig. 1 and Table I). A defetive mitohondrial side hain leavage has been suggested as an explanation for this aumulation (4, 6). Abnormal struture and funtion of liver mitohondria have been desribed in Zellweger syndrome (2), but suh a finding does not seem to be obligate (25). More remarkable, however, is that no peroxisomes an be reognized in liver and kidneys of the patients (2), and the possibility has been disussed that the mitohondrial hanges are seondary to the lak of peroxisomes (25, 26). Reently, we ould demonstrate that in the rat liver the peroxisomal fration was by far the most ative in atalyzing onversion of THCA into holi aid (9, 1), and the same seems to be the ase in human liver (unpublished experiments), In view of this, it seems likely that the aumulation of the C27-steroid bile aid preursors, in partiular of THCA in Zellweger syndrome, is due to a redued apaity for peroxisomal 13-oxidation of the steroid side hain. The results of our in vivo experiments showed that radioativity given as 5ft-holestane-3a,7a, 12a-triol rapidly appeared in more polar produts with partially oxidized C27-steroid side hain, mainly as THCA (Fig. 2). This shows that the mitohondrial 26-hydroxylation is funtioning in Zellweger syndrome and supports the previous suggestion that the metaboli a QaI L- 4 ao -._ ; C- 432 B. F. Kase, I. Bjirkhem, P. Hdgd, and J. I. Pedersen

7 defet is loalized at the final side hain leavage reation (4). In addition, the rapid onversion of 513-holestane-3a,7a,12atriol into THCA seems to exlude the possibility that mirosomal 25- and 24S-hydroxylation of 513-holestane-3a,7a, 12atriol is an important pathway for formation of holi aid in man (8, 27). If this had been the ase, there would be no reason for aumulation of THCA. The results thus support the ontention that the 26-hydroxylase pathway is the most important one in the normal biosynthesis of holi aid (28, 29). The blok in bile aid synthesis was not omplete, sine tritiated holi aid was deteted after in vivo administration both of 3H-5fl-holestane-3a,7a, 12a-triol and 3H-THCA. The onversion into holi aid was, however, very slow and inomplete ompared with what has been observed when bile aid preursors have been administered in vivo to adult humans. When suh labeled preursors are administered to either bile fistula patients or to normal subjets, there is a rapid (within minutes) and almost omplete onversion to bile aids (17, 18, 3-32). When 3H-51-holestane-3a,7a,12a-triol was given to bile fistula patients, -75% was onverted to holi aid (31). The peak speifi ativity in holi aid ourred already after 3 min and the tritium ativity orresponding to the preursor was ompletely eliminated after 13 h (31). Similarly, when 3H-THCA was administered intravenously to either normal subjets (17) or to bile fistula patients (32), there was a rapid and nearly omplete (81-97%) onversion into holi aid. These results ontrast with our finding of a maximal onversion of 513-holestane 3a,7a, 12a-triol to holi aid of 15-2% (Fig. 3). 11 h after administration, the degree - of onversion to holi aid was only 9% for 3H-5fl-holestane- 3a,7a, 12a-triol and 8% for 3H-THCA (Figs. 3 and 6). In the evaluation of these results it should be kept in mind that the THCA experiment was performed only a few weeks before the patient died.,8-oxidation of the steroid side hain is not a limiting step under normal onditions, and there is likely to be a onsiderable overapaity of enzyme ativity in the peroxisomes. The low onversion of 3H-THCA into holi aid observed here may be due to residual peroxisomal enzyme ativity, or to the presene of alternative pathways to holi aid bypassing the peroxisomal system. The similar rate of onversion of labeled 513-holestane-3a,7a,12a-triol and of THCA into holi aid during the first 11 h (refer above) makes it unlikely that the 25-hydroxylase pathway (see Introdution) plays an important role in the Zellweger syndrome as has been suggested (4). If alternative pathways exist from the triol, it is therefore more likely that these involve metaboli modifiations of THCA. The very low speifi radioativity of 3H in the holi aid as ompared with that of 3H-THCA may suggest that part of the holi aid in the patient had been formed by an unknown pathway bypassing both 513-holestane-3a,7a,12a-triol and THCA. More likely, the low ratio of speifi ativities may be explained by a very slow influx of tritium from THCA into a muh larger holi aid pool in bile (refer to Results). Differenes in slopes of the deay urves ould tell if alternative pathways are involved. Suh information would require a longer olletion period and more data points than we were able to obtain in Fig. 4. Also, Hanson et al. (33), who studied the metabolism of labeled THCA in two siblings with holestasis due to intrahepati bile dut anomalies, reported similar findings as ours. Also, in that ase, there was a higher speifi radioativity in THCA than in holi aid isolated from urine. The differenes were, however, less marked than in the present study. The pool size of holi aid (24 mg/m2) was drastially redued in our patient. Watkins et al. (23) reported a pool size of -3 mg/m2 for newborn healthy full-term infants (24) and _9 mg/m2 for premature infants. The daily prodution rate of holi aid orresponds to -'/I of that formed in normal newborns (24), whih indiates that the defiient onversion of THCA to holi aid is not ompensated for by alternative pathways. A main produt in serum and urine from both 513- holestane-3a,7a,12a-triol and THCA was identified as a tetrahydroxylated 5f3-holestanoi aid with all hydroxyl groups in the steroid nuleus. 11 and 6a-hydroxylated holi aid has been identified in human neonatal urine (34) and 213- and 6ahydroxylated holi aid has been found in duodenal aspirates from neonates with high intestinal obstrution (35). It thus appears likely that the tetrahydroxylated THCA ould be hydroxylated in 113, 2,B, or 6a position. Sine a harateristi ion at mle 217 or m/e 243 was laking, the first two possibilities may be exluded. We are left with the possibility that the struture orresponds to 3a,6a,7a,12a-tetrahydroxy-513-holestanoi aid, but further work is needed to establish this. Sine this ompound is relatively polar, it may represent an important route for urinary exretion of aumulated THCA. The retention time of the hydroxylated THCA on gas hromatography is very similar to that of 24-OH-THCA, and by mass fragmentography they share several fragments. This may have aused overestimation of the amount of aumulated 24- OH-THCA in Zellweger syndrome in a previous report (4). After injetion of 5#-holestane-3a,7a,12a-triol, and after administration of THCA per os, we ould detet only small amounts of tritium ativity in the peak orresponding to 24- OH-THCA (Figs. 5 and 7). Aumulation of 24-OH-THCA has been taken as evidene for a metaboli blok in the onversion of 24-OH-THCA into holi aid in mitohondria (4). Another possibility is that the 24-OH-THCA identified may be a produt of hydroxylation in the endoplasmi retiulum. Determination of the stereoisomeri struture of the ompound may differentiate between these possibilities. It should be emphasized that our patient was treated with phenobarbital, and it annot be exluded that the hydroxylase ativity observed here in part may have been indued by the treatment. Very reently, however, we were able to analyze serum from another infant with the Zellweger syndrome that had never been treated with phenobarbital. The pattern of bile aids in serum from that infant was almost idential to that obtained here. After administration of the holi aid preursors, we ould detet only small amounts of radioativity that orresponded to the C29-diarboxyli aid. This appeared surprising in view of the high onentration of this ompound in serum. The explanation may be that this metabolite is formed very slowly by hain elongation of THCA. Sine there is very little exretion of this ompound in bile and urine (5) it aumulates in plasma. The hain elongation of THCA may be analogous to the elongation of long-hain fatty aids in fibroblasts ultured from skin biopsies from patients with Zellweger syndrome (36). Chain shortening of very long-hain fatty aids is one Defetive Peroxisomal Formation of Choli Aid in Zellweger Syndrome 433

8 I 11No~~~~~~~~I III VI H 'OH Figure 8. Proposed metabolism of 5ft-holestane-3a,7a, 12a-triol (I) in Zellweger patient. The triol is oxidized to THCA (II) via 5ftholestane-3a,7a,12a,26-tetrol (not shown). THCA aumulates as a onsequene of impaired peroxisomal formation of holi aid (III) and is hydroxylated in the 24-position (VI) and in a nulear position, possibly 6a (V), or side hain elongated to C29-diarboxyli aid (IV). important funtion of peroxisomes (37). Aumulation of C29- diarboxyli aid (5) and very long-hain fatty aids in Zellweger syndrome (36) may thus both be the expression of ompensating mehanisms to restrited peroxisomal metaboli proesses. Adrenal leuodystrophy i's another linial entity with aumulation of very long-hain fatty aids and depletion of peroxisomes (11, 36). Thus, both this disease and Zellweger syndrome probably represent a new lass of peroxisomal diseases with different linial expressions. We onlude from the present findings that liver peroxisomes are essential in the normal formation of holi aid from THCA. In Zellweger syndrome, there is a defetive onversion of THCA to holi aid (Fig. 8), and as a onsequene the aumulated THCA is either exreted as suh or transformed to other metabolites by hydroxylation or by hain elongation. Aknowledgment We are grateful to Dr. Anne-Lise Borresen, Institute for Medial Genetis, University of Oslo, for the ulture of fibroblasts and to Dr. Hugo W. Moser, The John F. Kennedy Institute, Baltimore, for the analysis of very long-hain fatty aids in fibroblasts. The skillful tehnial assistane of Eva Torma Grabner and Anita L6vgren is appreiated. This work was supported by the Norwegian Researh Counil for Siene and the Humanities, by the Anders Jahres Foundation, and by the Swedish Medial Researh Counil (3X-3141). Referenes 1. Bowen, P., C. S. N. Lee, H. Zellweger, and R. Lindenberg A familial syndrome of multiple ongenital defets. Bull. Johns Hopkins Hosp. 114: Goldfisher, S., C. L. Moore, A. B. Johnson, A. J. Spiro, M. P. Valsamis, H. K. Wisniewsi, R. H. Rith, W. T. Norton, I. Rapin, and L. M. Gartner Peroxisomal and mitohondrial defets in the erebro-hepato-renal syndrome. Siene (Wash. DC). 182: Eyssen, H., G. Parmentier, F. Compernolle, J. Boon, and E. Eggermont Trihydroxyoprostanoi aid in the duodenal fluid of two hildren with intrahepati bile dut anomalies. Biohim. Biophys. Ata. 273: IV 9H "I" 4. Hanson, R. F., P. S. VanLeeuwen, G. C. Williams, G. Grabowski, and H. L. Sharp Defets of bile aid synthesis in Zellweger's syndrome. Siene (Wash. DC). 23: Parmentier, G. G., G. A. Janssen, E. A. Eggermont, and H. J. Eyssen C2rbile aids in infants with oprostani aidemia and ourrene of a 3a,7a,12a-trihydroxy-5f-C2q diarboxyli bile aid as a major omponent in their serum. Eur. J. Biohem. 12: Mathis, R. K., J. B. Watkins, P. Szzepanik-Van Leeuwen, and I. T. Lott Liver in the erebro-hepato-renal syndrome: defetive bile aid synthesis and abnormal mitohondria. Gastroenterology. 79: Bjorkhem, I Mehanism of bile aid biosynthesis in mammalian liver. In Comprehensiv Biohemistry. Elsevier Sientifi Publishing Co., Amsterdam. In press. 8. Salen, G., and S. Shefer Bile aid synthesis. Annu. Rev. Physiol. 45: Pedersen, J. I., and J. Gustafsson Conversion of 3a,7a,12atrihydroxy-5f-holestanoi aid into holi aid by rat liver peroxisomes. FEBS (Fed. Eur. Biohem. So.) Lett. 121: Kase, B. F., I. Bjorkhem, and J. I. Pedersen Formation of holi aid from 3a,7a,12a-trihydroxy-5,-holestanoi aid by rat liver peroxisomes. J. Lipid Res. 24: Kawamura, N., A. B. Moser, H. W. Moser, T. Ogino, K. Suzuki, H. Shaumbury, A. Milunsky, J. Murphy, and V. Kishimoto High onentration of hexaosanoate in ultured skin fibroblast lipids from adrenoleuko-dystrophy patients. Biohem. Biophys. Res. Commun. 82: Evrard, P., V. S. Caviness, Jr., J. Prats-Vinas, and G. Lyon The mehanism of arrest of neuronal migration in the Zellweger malformation: an hypothesis based upon ytoarhitetoni analysis. Ata Neuropathol. (Berlin). 41: Bjorkhem, I., and J. Gustafsson w-hydroxylation of the steroid side hain in the biosynthesis of bile aids. Eur. J. Biohem. 36: Gustafsson, J Biosynthesis of holi aid in rat liver: formation of holi aid from 3a,7a,12a-trihydroxy- and3a,7a,12a,24- tetrahydroxy-5#-holestanoi aids. Lipids. 15: Angelin, B., I. Bjorkhem, E. Einarsson, and S. Ewerth Hepati uptake of bile aids in man. Fasting and postprandial onentrations of individual bile aids in portal venous and systemi blood serum. J. Clin. Invest. 7: Bjorkhem, I., and. Falk Assay of the major bile aids in serum by isotope dilution-mass spetrometry. Sand. J. Clin. Lab. Invest. 43: Hanson, R. F., and G. C. Williams Metabolism of 3a,7a, 12a-trihydroxy-5,-holestan-26-oi aid in normal subjets with an intat enterohepati irulation. J. Lipid Res. 18: Hanson, R. F., A. B. Staples, and G. C. Williams Metabolism of 5#-holestane-3a,7a,12a-26-tetrol and 5g#-holestane- 3a,7a,12a,25-tetrol into holi aid in normal human subjets. J. Lipid Res. 2: Lindstedt, S The turnover of holi aid in man. Bile aids and steroids. 51. Ata Physiol. Sand. 4: Janssen, G., S. Toppet, and G. Parmentier Struture of the side hain of the C29 diarboxyli bile aid ourring in infants with oprostani aidemia. J. Lipid Res. 23: Heikura, S., S. Simila, K. Finni,. Maentausta, and. Janne Choli aid and henodeoxy-holi aid onentrations in serum during infany and hildhood. Ata Paediatr. Sand. 69: Finni, K Serum bile aids during infany and hildhood. Thesis. Ata Universitatis Ouluensis Serie D. Media No Watkins, J. B., P. Szzepanik, J. G. Gould, P. Klein, and R. Lester Bile salt metabolism in the human premature infant. Preliminary observation of pool size and synthesis rate following prenatal administration of dexamethasone and phenobarbital. Gastroenterology. 69: Watkins, J. B., D. Ingall, P. Szzepanik, P. Klein, and R. Lester Bile salt metabolism in the newborn. Measurements of 434 B. F. Kase, I. Bjorkhem, P. Hdgd, and J. I. Pedersen

9 pool size and synthesis by stable isotope tehnique. N. Engl. J. Med. 288: Borst, P Animal peroxisomes (mirobodies), lipid biosynthesis and the Zellweger syndrome. Trends Biohem. Si. 8: Trijbels, J. M. F., J. A. Berden, L. A. H. Monnens, J. L. Willems, A. J. M. Janssen, R. B. H. Shutgens, and M. Vanden Broek- Van Essen Biohemial studies in the liver and musle of patients with Zellweger syndrome. Pediatr. Res. 17: Shefer, S., F. W. Chemg, B. Dayal, S. Hauser, G. S. Tint, G. Salen, and E. H. Mosbah A 25-hydroxylation pathway of holi aid biosynthesis in man and rat. J. Clin. Invest. 57: Oftebro, H., I. Bjorkhem, S. Skrede, S. Shreiner, and J. I. Pedersen Cerebrotendinous xanthomatosis. A defet in mitohondrial 26-hydoxylation required for normal biosynthesis of holi aid. J. Clin. Invest. 65: Bjorkhem, I.,. Fausa, G. Hopen, H. Oftebro, J. I. Pedersen, and S. Skrede Role of the 26-hydroxylase in the biosynthesis of bile aids in the normal state and in erebrotendinous xanthomatosis. An in vivo study. J. Clin. Invest. 71: Swell, L., J. Gustafsson, C. C. Shwartz, L. G. Halloran, H. Danielsson, and Z. R. Vlahevi An in vivo evaluation of the quantitative signifiane of several potential pathways to holi and henodeoxyholi aids from holesterol in man. J. Lipid Res. 21: Vlahevi, Z. R., C. C. Shwartz, J. Gustafsson, L. G. Halloran, H. Danielsson, and L. Swell Biosynthesis of bile aids in man. Multiple pathways to holi aid and henodeoxyholi aid. J. Biol. Chem. 255: Swell, L., J. Gustafsson, H. Danielsson, C. C. Shwartz, and Z. R. Vlahevi Biosynthesis of bile aids in man. An in vivo evaluation of the onversion of R and S 3a,7a,12a-trihydroxy-51% holestanoi and 3a,7a,12a,24-tetrahydroxy-5gB-holestanoi to holi aid. J. Biol. Chem. 256: Hanson, R. F., J. N. Isenberg, G. C. Williams, D. Hahey, P. Szzepanik, P. Klein, and H. L. Sharp The metabolism of 3a,7a,12a-trihydroxy-5(@-holestanoi aid in two siblings with holestasis due to intrahepati bile dut anomalies. J. Clin. Invest. 56: Strandvik, B., and S. A. Wikstrom Tetrahydroxylated bile aids in healthy human newborns. Eur. J. Biohem. 12: Clayton, P. T., D. P. R. Muller, and A. M. Lawson The bile aid omposition of gastri ontents from neonates with high intestinal obstrution. Biohem. J. 26: Brown, F. R., III, A. J. MAdams, J. W. Cummins, R. Konkol, I. Singh, A. B. Moser, and H. W. Moser Cerebro-hepato-renal (Zellweger) syndrome and neonatal adrenoleuko-dystrophy: similarities in phenotype and aumulation of very long hain fatty aids. Johns Hopkins Med. J. 151: Bremer, J., and K. R. Norum Metabolism of very longhain monounsaturated fatty aids (22:1) and the adaptation to their presene in the diet. J. Lipid Res. 23: Defetive Peroxisomal Formation of Choli Aid in Zellweger Syndrome 435