BIOCHEMICAL
MEDICINE
AND
METABOLIC
BIOLOGY
46,
235-245 (1991)
Structure and Variability of Mammalian Peroxisomal Membrane Proteins GOLDER N. WILSON AND TERESA E. KING Department
of Pediatrics,
University Dallas,
of Texas Southwestern Texas 75235
Medical
School,
Received May 15, 1991 Peroxisomal membrane proteins (PMPs) from the Swiss-Webster mouse are analyzed and compared to those of rats and humans using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting. A purification procedure for fresh mouse, rat, or human biopsy liver which enriches peroxisomal/mitochondrial marker enzyme ratios over RIO-fold is characterized. When analyzed by SDS-PAGE, membranes of purified liver peroxisomes are shown to contain the same complement of 145-, 70-, 55-, 36-, and 22-kDa PMPs in rats, mice, and humans. A rabbit polyclonal antibody raised against mouse peroxisomal membranes demonstrates immunoreactivity to 145- and 70-kDa proteins in fresh liver homogenates from all three species and in control or Zellweger syndrome fibroblasts from humans. Human autopsy or placental tissues which were refrigerated before analysis exhibited 105-, 55-, and 36-kDa peptides which may be derived from the 145- and 70-kDa peptides. Such conversions, if related to degradation, may explain difficulties in purifying peroxisomes from human autopsy specimens. Variable amounts of the 55-kDa peptide also occurred in mouse adrenal and lung, and the conversion of higher to lower molecular weight PMPs could not be demonstrated by in vitro incubation of mouse liver. Further definition of the structure and variability of mammalian PMPs should be helpful in understanding polyenzymopathies such as Zehweger syndrome. o 1991 Academic PXSS. Inc.
The multiple dysmorphologic and metabolic alterations associated with human peroxisomal disorders (1,2) testify to the importance of peroxisomes in ontogeny and postnatal function. Certain of these disorders, such as acatalasemia or Xlinked adrenoleukodystrophy, involve the deficiency of a single peroxisomal enzyme which leads to a distinctive phenotype (3). Others, such as Zellweger syndrome or neonatal adrenoleukodystrophy, are associated with structural disorganization of the peroxisome and an overlapping spectrum of brain, eye, liver, adrenal, and bone disease (1). The latter disorders have multiple peroxisomal enzyme deficiencies and virtually absent peroxisomes in liver. However, residual peroxisomes in fibroblasts (4) and the immunologic detection of certain peroxisomal proteins (5) indicate that partial biogenesis occurs in these deficiency disorders. Persistence of peroxisomal membrane “ghosts” in Zellweger syndrome fibroblasts (6) suggests that import, assembly, or maintenance of peroxisomal 235 0885-4505/91 $3.00 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
236
WILSON AND KING
matrix components may be defective. A clue to protein import has recently been provided by the discovery of a specific tripeptide signal which targets proteins to peroxisomes in a wide variety of organisms (7). Of interest because of their accessibility and potential relevance to import are peroxisomal membrane proteins (PMPs). These peptides can withstand the alkaline extraction devised by Fujiki et al. (8), who showed that peroxisomes had a different membrane protein content and thus an independent pathway for biogenesis than endoplasmic reticulum or mitochondria. Several laboratories have demonstrated the presence of 70/69- and 22-kDa PMPs in rat (5,9) or human (10) tissues, and these have been visualized at normal amounts in the fibroblasts from patients with Zellweger syndrome (5,ll). Additional species of 53, 42, 36, 28, and 26 kDa have also been seen in rat liver extracts, with the 42-, 28-, and 26-kDa species postulated to be degradation products of the 70-kDa protein (9). The latter protein has been shown to be a member of the ATP-binding P-glycoprotein family through cloning of its complementary DNA from rat liver (12). Few studies of mouse PMPs have been reported, despite the potential in this organism to manipulate peroxisome structure through use of transgenic technology. Here the structure and tissue variability of mouse PMPs are described and compared to those of rats and humans. MATERIALS AND METHODS Materials. Nitrocellulose membranes from Amersham (Arlington Heights, IL), immunoblotting reagents from Sigma Chemical Corp. (St. Louis, MO), goat antirabbit peroxidase-conjugated antibody from Cappel (Malvern, PA), 32P04 from Amersham, electrophoresis reagents from Bio-Rad (Richmond, CA), and biochemicals from Sigma were obtained. Human Zellweger syndrome fibroblast lines GM 228, GM 6231, and GM 6094 were obtained from the Human Mutant Cell Repository (Camden, NJ). Tissue culture media were obtained from GIBCO (Grand Island, NY). Animal and human tissues.Sprague-Dawley white rats or Swiss-Webster white mice were obtained from Charles River Breeding Laboratories (Wilmington, MA) at weaning and maintained on regular food with standard light/dark cycles. New Zealand white rabbits were obtained at age 12 weeks and maintained on regular food. Fresh human liver was obtained from a wedge biopsy and peroxisomal purification begun immediately; autopsy human tissues were obtained from 2-16 h after death and stored at 4°C. Human fibroblasts were cultured in minimal essential medium (GIBCO) with 10% fetal calf serum, harvested by trypsinization and washing, suspended in 0.25 M sucrose, and disrupted by sonication (Branson3 X 30-set bursts small probe at 30% power) or homogenization in a small handheld homogenizer for 30 strokes. Mouse PMP purification and antibody preparation. Mouse liver peroxisomes were purified using differential centrifugation in sucrose (13) and Nycodenz (14) as described briefly in a previous publication (15). Fresh liver (9 g) was homogenized in 27 ml cold 0.25 M sucrose, IO mM Tris-HCl, pH 7.4, I mM EDTA, and centrifuged as outlined in Fig. 1 using a Beckman 52-21 centrifuge, a type JA-20 rotor, and 50-ml polyethylene tubes. At each step, the pellet was homogen&d
PEROXISOMAL
(A)
LIVER
MEMBRANE
237
PROTEINS
HOMOGENATE
10,gOO xr’” b SI 60,060
h
min xr
Post-M
Nuclei
MitochMo”dria
250,~ffnin Post-L
1.4 x 10’
g-min I
PerOxisomes
cykol
(0) ioo
MOUSE
n
Ed DHAP-AT El w
CATALASE CMOCHFiOh4E
COXIDASE
LACTATE DEHMX#XjENASE
FIG. 1. Standard purification of mammalian peroxisomes. (A) The differential centrifugation procedure of Leighton et al. (13) followed by Nycodenz step (N) or linear (No) gradients. (B) The specific activities of various marker enzymes in selected purification fractions. The values are normalized for easy comparison, with 100% corresponding to 4000 units/mg for catalase, 0.09 unit/mg for DHAPAT, 15 units/mg for cytochrome c oxidase, and 1.5 units/mg for lactate dehydrogenase. The percentage of protein relative to that in the initial liver homogenate is listed in parentheses beneath each fraction.
238
WILSON AND KING
with fresh sucrose and recentrifuged and the supematants were pooled. The postL fraction was centrifuged 60 min at 75,000 r-pm in the Beckman TLA 100.3 rotor to yield the cytoplasmic fraction (C) for comparative purposes. The “light mitochondrial” (L) fraction was further purified as Nycodenz (N) and Nycodenz gradient (No) fractions, resulting in peroxisomes which were greater than 99% pure as examined by electron microscopy (15). Identical purification schemes were employed for the isolation of peroxisomes and peroxisomal membrane fractions from 9 g of fresh rat liver or, with proportionate modifications in volume, from almost 1 g of freshly biopsied human liver. Assays for catalase (16), dihydroxyacetone phosphate acyltransferase (DHAP-AT (17)), cytochrome c oxidase (13), and lactate dehydrogenase (19) as marker enzymes were by standard methods. Membranes were purified from various fractions by diluting 1 to 50 with 2.94 ml of 100 mM sodium carbonate, pH 11.5, and centrifuging 1 h at 75,000 rpm and 4°C in the Beckman TLA 100.3 rotor (8). Antibodies to PMPs were prepared by injecting membranes from the No fraction into rabbits as described (15). SDS-PAGE and immunoblot analysti. Protein samples (lo-30 ~1 at l-5 mg/ml) were diluted l/2-1/3 in loading buffer and applied to 12 or 7.5% SDS-PAGE gels in a Bio-Rad MiniProtean II apparatus. Electrophoresis was for 0.5 h at 200 V (50 mA), and gels were either stained in 0.1% Coomassie blue/40% methanol/10% acetic acid, by the silver-staining kit obtained from Bio-Rad, or incubated in transfer buffer prior to immunoblotting. The latter was performed as described (15) using rabbit anti-PMP antiserum diluted l/100, treatment with goat antirabbit serum, and detection by peroxidase staining (Cappel). RESULTS The purification of mouse peroxisomes as modified from prior methods (1315) is diagrammed in Fig. 1A and summarized in Table 1. Figure 1B demonstrates that l-2% of the protein in the initial S1 supematant from mouse or rat liver will be recovered in the peroxisomal N fraction, with greatly enriched ratios of peroxisomal enzyme specific activities (catalase and DHAP-AT) compared to cytoplasmic (lactate dehydrogenase) or mitochondrial (cytochrome c oxidase) enzyme specific activities. The specific activities are normalized to a 100% value for convenient comparison. The ratio of DHAP-AT to cytochrome c oxidase is some lOO-fold greater in the peroxisomal N fraction compared to the mitochondrial M fraction. Similar enrichment in the ratios of the peroxisomal enzymes to microsomal (GP-AT) or lysosomal (acid phosphatase) enzymes was observed but not shown. In quantitative terms (Table l), 25-fold enrichment of catalase specific activity with 29% yield was obtained. Since only about half of liver catalase is in the peroxisomal matrix, this represents an actual yield of over 50%. Purification of rat peroxisomes yielded virtually identical results to those shown in Fig. 1 and Table 1. A typical yield of catalase was 23% with a 27-fold enrichment of specific activity. Results with human specimens were much more variable. Purification was not successful with postmortem liver, fresh placenta, or fresh liver obtained at liver transplant. The latter sample was extremely cirrhotic and difficult to homogenize. Successful purification of human liver peroxisomes was only possible using fresh tissue from a wedge biopsy specimen obtained for trans-
w/ml
8.4 5.2 18.2 5.7 3.1 8.4 5.7 4.6
Volume (4
46 58 1.5 71 68 2 0.8 0.5
Fraction
5 Post-M M Post-L C L N NG
Protein
385 300 21 405 210 16.8 4.5 2.3
Total mg 1215 1060 720 1130 322 10,160 20,150 19,380
units/ml 56,000 61,000 1,050 81,000 22,000 20,320 16,100 9,650
Total units
Catalase
TABLE 1 Purification of Peroxisomes from Mouse Liver
145 204 58 200 105 1200 3600 4200
units/mg
100 110 145 37 29 17
1.4 8.3 25 28
Yield (%) 1 1.4
Fold
8 w
z
5 $ 3
5
2 g-
E t3
z
240
WILSON AND KING
123
123
FIG. 2. SDS-PAGE of mouse peroxisomal purification fractions. The upper panel shows discontinuous 12% SDS-PAGE containing 50 pg whole protein from fractions S, (lane l), Post-M (2), M (3), Post-L (4), C (5) L (6), and N (7) illustrated in Fig. 1. The left lane contains Sigma standard 6H with proteins of 205, 116, 97.4, 66, 45, and 29 kDa and the right lane contains standard 7 with proteins of 66, 45, 36, 29, 24, 20, and 14 kDa. The middle panel shows the corresponding membrane proteins as isolated by sodium carbonate treatment from fractions S1 (lane l), M (lane 2), L (4), and N (5). Lanes 3 and 6 show the membrane proteins extracted with 1% Triton X-100 from fractions M and N, respectively. The 6H (left) and 7 standards (right) are again included. The lower panel displays single-layer SDS-PAGE with total (left) or membrane (right) proteins from human (l), rat (2) or mouse (3) No fractions compared to the 7 standard (right lane in each group). Arrowheads indicate the positions of 145-, 70-, and 22-kDa PMPs. Gels in the upper two panels were stained with both Coomassie blue and silver; those in the lower panel were stained with Coomassie blue only.
plant evaluation-960 mg of tissue yielded 240 pg of N fraction with 20% yield of catalase and 25fold enrichment of specific activity. Purification of mouse peroxisomes is reflected by the pattern of whole or membrane proteins revealed by SDS-PAGE electrophoresis as shown in Fig. 2. The upper and middle panels confirm the observations of Fujiki et al. (8) that peroxisomal total and membrane proteins are quite different from those of other
PEROXISOMAL
123456
MEMBRANE
7
241
PROTEINS
8
910 P w
FIG. 3. Immunoblot analysis of mouse, rat, and human homogenates after 7.5% SDS-PAGE. Lanes l-3 represent mouse purification fractions (Fig. 1A) S, , L, and N; lanes 3-5 represent mouse, rat, and human N fractions prepared from liver homogenates; lane 6 is a negative control containing 40 pg protein from the cytoplasmic fraction (C in Fig. 1A) which should contain no peroxisomes. Lanes 7-10 represent control human fibroblast or Zellweger syndrome fibroblast lines GM 228, GM 6231, and GM 6094, respectively.
organelles. It can be seen from the middle panel that the 70-kDA PMP is selectively extracted by sodium carbonate (lane 5) and particularly by Triton X-100 (lane 6), which may explain the bias of rabbit antibodies toward this protein (see below). The lower panel demonstrates that the major PMPs obtained from the N fractions of mouse, rat, or human liver exhibit striking size conservation after 12.5% SDSPAGE and staining with Coomassie blue. Peptides of 145, 70, 55, 36, and 22 kDa are seen clearly in the liver peroxisomal membrane fractions of all three species, with the ratio of PMP 55 to 70 being much greater in the human liver fraction. Intradermal injection of highly purified mouse liver peroxisomal membranes into rabbits allowed the preparation of polyclonal antibodies to mouse PMPs. These antibodies all exhibited a bias toward the 70-kDa PMP as shown by the immunoblotting results displayed in Fig. 3. The correspondence of PMP abundance with increasing mouse peroxisome purity as depicted in lanes l-3 and the absence of reactivity in fractions depleted in peroxisomes (lane 6) support the specificity of the antibody for PMPs. The antibody to mouse peroxisomal membranes reveals PMPs of similar size in rat (lane 4) and human (lane 5) peroxisomal preparations; this antibody also identifies 145- and 70-kDa peptides in homogenates of fibroblasts from control human (lane 7) and Zellweger syndrome (lanes 8-10) fibroblast lines. Since the immunoblot shown in Fig. 3 was produced from 7.5% SDS-PAGE, the 36- and 22-kDa PMPs have migrated off the gel in this experiment. The data in Fig. 4 suggest that the 55 and 36-kDa peptides may be derived from the 145- or 70-kDa peptides. Various human tissues obtained at autopsy and stored at 4°C overnight prior to homogenization and freezing exhibited minor amounts of the 145- and 70-kDa PMPs but increased amounts of the 55- or 36kDa PMPs (lanes 5-7, lower panel Fig. 4). The difference is most evident when comparing lane 2, derived from the homogenate of fresh biopsy liver, with lanes 3 and 4 derived from autopsy liver specimens (Fig. 4, lower panel). The upper
242
WILSON AND KING
123456789
1 2 3 4
5 6 7 8 910111213
FIG. 4. PMPs detected by 7.5% SDS-PAGE and immunoblotting of mouse and human tissues. The upper panel shows mouse tissues including unfractionated liver homogenates from the same individual after 0 h (lane 2), 2 h (lane 3) 6 h (lane 4), or 24 h (lane 5) of storage at 4°C or homogenates of lung (lane 6) brain (lane 7) kidney (lane 8) or adrenal (lane 9) prepared immediately after sacrifice. The lower panel shows human liver homogenates from different individuals at biopsy (lane 2) or autopsy following 2 h (lane 3) or 16 h (lane 4) of storage at 4°C. Other lanes contain kidney (lane 5) brain (lane 6) lung (lane 7) or placenta at 16 weeks (lane 8) 18 weeks (lane 9), 22 weeks (lane lo), 23 weeks (lane 11) 26 weeks (lane 12), or 35 weeks (lane 13) of gestation. All lanes contain 20 pg of homogenate protein except for lane 1 of each panel which contains 2 fig of mouse liver No fraction as control.
panel of Fig. 4 shows an attempt to monitor the degradation of PMPs by comparing liver homogenates from a mouse at various times after sacrifice. Regardless of whether the liver was left within the sacrificed animal (lanes 3-5) or excised (not shown), the 145 and 70-kDa PMPs remained intact for up to 24 h after sacrifice. However, other mouse tissues shown in the upper panel of Fig. 4 including lung (lane 6) or adrenal (lane 9) showed considerable amounts of the 55kDa PMP even when homogenized immediately after sacrifice. Note also that the amounts of PMPs in mouse or human liver and kidney, with their larger peroxisomes, are increased relative to tissues such as brain with “microperoxisomes.” Additionally, a unique peptide of 105 kDa is seen in human placenta (lanes 8-13, lower panel, Fig. 4), and the amount of PMPs in this tissue seems to plateau at 18-26 weeks with a decrease near term.
PEROXISOMAL
MEMBRANE
PROTEINS
243
DISCUSSION Considerable size and sequence conservation of PMPs from mice, rats, and humans are indicated by the SDS-PAGE and immunoblot profiles in Figs. 2-4. This immunologic cross-reactivity demonstrates that genes encoding mammalian PMPs are substantially conserved and supports their role in a highly organized peroxisomal structure. Because denaturing gels are employed, amino acid substitutions between species will not be evident, but significant deletions or rearrangements should have affected electrophoretic migration of the peptides. Conservation over 70 million years of evolution suggests most domains of the integral membrane proteins have important functions. These could include regulation of protein import, anchoring of key enzymes, structural/functional pleiotropy as occurs with xanthine oxidase of Drosophila (22), or maintenance of an internal peroxisomal milieu with organized properties. The purification scheme for peroxisomal membrane proteins and the availability of antibody reagents for detection sets the stage for structural and functional characterization which can decide among these possibilities. The peroxisomal purification scheme described here allows the preparation of 4-5 mg of peroxisomal protein from 9 g of fresh rodent liver in about 6 h time (Fig. 1, Table 1). If very highly purified peroxisomes are needed, then the additional gradient separation produces 2-32 mg in a 12-h time span (No fraction). Treatment of highly purified peroxisomes derived from 9 g of mouse liver with sodium carbonate produces 560-800 pg of membrane protein with a peptide profile which appears identical in rats, mice, and humans (Figs. 2, 3). The 70-, 36-, and 22-kDa peptides visualized by 12% SDS-PAGE have been observed in prior studies of rat peroxisomal membrane peptides (9) and the 145kDa peptide has been demonstrated in rat and human tisues using a polyspecific antibody to rat peroxisomal membranes (5). The availability of a convenient and effective purification scheme for mammalian peroxisomes allowed the preparation of a polyclonal rabbit antibody to mouse PMPs as was described for the rat by Lazarow and colleagues (5,6). Specificity of the rabbit antibody for PMPs is supported by correlation of reactivity with peroxisomal purity (Fig. 3)) with tissue type (Fig. 4), and with clofibrate treatment as demonstrated in a prior study (15). Immunoblotting studies reported here have focused on PMPs of higher molecular weight because of the enhanced blotting efficiency after 7.5% SDS-PAGE electrophoresis. Biased immunoreactivity toward the 70-kDa peptide has also been observed with three separate rabbit antibody preparations, and this may be explained by the greater ease with which this peptide can be extracted from the peroxisomal membrane using a variety of detergents (Fig. 2). In Fig. 3, as demonstrated in prior reports (5,6,10), antibody to rodent PMPs reacts with PMPs of 145 and 70 kDa in Zellweger syndrome and control human fibroblasts. The data in Fig. 4 explain the variable patterns of PMPs seen in some experiments and the difficulties encountered in purifying peroxisomes from human tissues obtained at autopsy. Although PMPs of 145 and 70 kDa are clearly dem-
244
WILSON AND KING
onstrated in fresh mouse or human liver (lanes 2, Fig. 4) variable amounts of 55 and 36 kDa peptides are seen in human autopsy liver (lanes 3 and 4, lower panel, Fig. 4). In placenta, a novel peptide of 105 kDa in addition to the 36-kDa peptide was detected. Origin of the 105kDa peptide from PMP 145 and origin of the 55- and 36-kDa peptides from PMP 145 or 70 by degradation is the favored explanation for these results, although simple refrigeration of mouse liver after sacrifice did not replicate these peptide conversions (lanes 1-4, upper panel, Fig. 4). Hart1 and Just (9) concluded that the 42-, 2%, and 26-kDa peptides observed in rat liver homogenates were breakdown products of PMP 70 in that organism. It will be interesting to examine whether these tissue-specific differences in PMP structure relate to degradation or alternative processing, and whether they have relevance to peroxisomal biogenesis or disease. ACKNOWLEDGMENTS Support of the Biological Humanics Foundation, Dallas Texas, and of NIH Grant ROl-HD26324 made possible certain aspects of this work, which was presented at the Society for Inherited Metabolic Disorders Annual Meeting, Sante Fe, New Mexico, April, 1991.
REFERENCES 1. Wilson GN, Holmes RD, Hajra AK. Peroxisomal disorders: Clinical commentary and future prospects. Am J Med Genet 30:771-792, 1988. 2. Schutgens RBH, Heymans HSA, Wanders RJA, Bosch Hvd, Tager J. Peroxisomal disorders: A newly recognized group of genetic diseases. Eur J Pediatr 144~430-440, 1986. 3. Moser HW. Peroxisomal disorders. J Pediutr 10&89-91, 1986. 4. Arias JA, Moser AB, Goldfischer SL. Ultrastructural and cytochemical demonstration of peroxisomes in cultured fibroblasts from patients with peroxisomal deficiency disorders. J Cell Biol 100:1789-1792, 1985. 5. Santos MJ, Imanaka T, Shio H, Lazarow PB. Peroxisomal integral membrane proteins in control and Zellweger fibroblasts. J Biol Chem 263:10,502-10,509, 1988. 6. Santos MJ, Imanaka T, Shio H, Small GM, Lazarow PB. Peroxisomal membrane ghosts in Zellweger syndrome-Aberrant organelle assembly. Science 239:1536-1538, 1988. 7. Gould SJ, Keller G-A, Schnieder, M, Howell SH, Garrard I-J, Goodman JM, Distel B, Tabak H, Subramani S. Peroxisomal protein import is conserved between yeast, plants, insects, and mammals. EMBO J 9~85-90, 1990. 8. Fujiki Y, Fowler S, Shio H, Hubbard AL, Lazarow PB. Polypeptide and phospholipid composition of the membrane of rat liver peroxisomes: Comparison with endoplasmic reticulum and mitochondrial membranes. J Cell Biol93:103-110, 1982. 9. Hart1 F-U, Just WW. Integral membrane polypeptides of rat liver peroxisomes: Topology and response to different metabolic states. Arch Biochem Biophys 255:109-119, 1987. 10. Small GM, Santos MJ, Imanaka T, Poulos A, Danks DM, Moser HW, Lazarow PB. Peroxisomal integral membrane proteins in livers of patients with Zellweger syndrome, infantile Refsum’s disease and X-linked adrenoleukodystrophy. J Inherited Metab Dis 11:358-371,1988. 11. Suzuki Y, Shimozawa N, Orii T, Hashimoto T. Major peroxisomal polypeptides are synthesized in cultured skin fibroblasts from patients with Zellweger syndrome. Pediutr Res 26~150-153, 1989. 12. Kamijo K, Taketani S, Yokota S, Osumi T, Hashimoto T. The 70-kDa peroxisomal membrane protein is a member of the Mdr (P-glycoprotein)-related ATP-binding protein superfamily. J Biof Chem 265:4534-4540, 1990. 13. Leighton F, Poole B, Beaufay H, Baudhuin P, Coffey JW, Fowler S, De Duve C. The largescale separation of peroxisomes, mitochondria, and lysosomes from the livers of rats injected with Triton WR-1339. J Cell Biol37:482-513. 1968.
PEROXISOMAL
MEMBRANE
PROTEINS
245
14. Ghosh MK, Hajra AK. A rapid method for the isolation of peroxisomes from rat liver. Anal Biochem 159:169-74, 1986. 15. Wilson GN, King T, Argyle JC, Garcia RF. Maternal clofibrate administration amplifies fetal peroxisomes. Pediatr Res 29:256-262, 1991. 16. Peters TJ, Muller M, and De Duve C. Lysosomes of the arterial wall. I. Isolation and subcellular fractionation of cells from normal rabbit aorta. J Exp Med 136:1117-1123, 1972. 17. Jones CL, Hajra AK. Properties of guinea pig liver peroxisomal dihydroxyacetone phosphate acyltransferase. J Biol Chem 25%3289-8295, 1980. 18. Datta NS, Wilson GN, Hajra AK. Deficiency of enzymes catalyzing the biosynthesis of glycerolether lipids in Zellweger syndrome. N Engl J Med 311:1080-1083, 1984. 19. Kornberg A. Lactate dehydrogenase of muscle. Methods Enzymol 1:441-443, 1955. 20. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein using the principle of protein-dye binding. Anal Biochem 72~248-254, 1976. 21. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 193~265-275, 1951. 22. Beard ME, Holtzman E. Peroxisomes in wild-type and rosy mutant Drosophila melanogaster. Proc Nat1 Acad Sci U S A 84~7433-7473, 1987.
MEDICINE
AND
METABOLIC
BIOLOGY
46,
235-245 (1991)
Structure and Variability of Mammalian Peroxisomal Membrane Proteins GOLDER N. WILSON AND TERESA E. KING Department
of Pediatrics,
University Dallas,
of Texas Southwestern Texas 75235
Medical
School,
Received May 15, 1991 Peroxisomal membrane proteins (PMPs) from the Swiss-Webster mouse are analyzed and compared to those of rats and humans using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting. A purification procedure for fresh mouse, rat, or human biopsy liver which enriches peroxisomal/mitochondrial marker enzyme ratios over RIO-fold is characterized. When analyzed by SDS-PAGE, membranes of purified liver peroxisomes are shown to contain the same complement of 145-, 70-, 55-, 36-, and 22-kDa PMPs in rats, mice, and humans. A rabbit polyclonal antibody raised against mouse peroxisomal membranes demonstrates immunoreactivity to 145- and 70-kDa proteins in fresh liver homogenates from all three species and in control or Zellweger syndrome fibroblasts from humans. Human autopsy or placental tissues which were refrigerated before analysis exhibited 105-, 55-, and 36-kDa peptides which may be derived from the 145- and 70-kDa peptides. Such conversions, if related to degradation, may explain difficulties in purifying peroxisomes from human autopsy specimens. Variable amounts of the 55-kDa peptide also occurred in mouse adrenal and lung, and the conversion of higher to lower molecular weight PMPs could not be demonstrated by in vitro incubation of mouse liver. Further definition of the structure and variability of mammalian PMPs should be helpful in understanding polyenzymopathies such as Zehweger syndrome. o 1991 Academic PXSS. Inc.
The multiple dysmorphologic and metabolic alterations associated with human peroxisomal disorders (1,2) testify to the importance of peroxisomes in ontogeny and postnatal function. Certain of these disorders, such as acatalasemia or Xlinked adrenoleukodystrophy, involve the deficiency of a single peroxisomal enzyme which leads to a distinctive phenotype (3). Others, such as Zellweger syndrome or neonatal adrenoleukodystrophy, are associated with structural disorganization of the peroxisome and an overlapping spectrum of brain, eye, liver, adrenal, and bone disease (1). The latter disorders have multiple peroxisomal enzyme deficiencies and virtually absent peroxisomes in liver. However, residual peroxisomes in fibroblasts (4) and the immunologic detection of certain peroxisomal proteins (5) indicate that partial biogenesis occurs in these deficiency disorders. Persistence of peroxisomal membrane “ghosts” in Zellweger syndrome fibroblasts (6) suggests that import, assembly, or maintenance of peroxisomal 235 0885-4505/91 $3.00 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
236
WILSON AND KING
matrix components may be defective. A clue to protein import has recently been provided by the discovery of a specific tripeptide signal which targets proteins to peroxisomes in a wide variety of organisms (7). Of interest because of their accessibility and potential relevance to import are peroxisomal membrane proteins (PMPs). These peptides can withstand the alkaline extraction devised by Fujiki et al. (8), who showed that peroxisomes had a different membrane protein content and thus an independent pathway for biogenesis than endoplasmic reticulum or mitochondria. Several laboratories have demonstrated the presence of 70/69- and 22-kDa PMPs in rat (5,9) or human (10) tissues, and these have been visualized at normal amounts in the fibroblasts from patients with Zellweger syndrome (5,ll). Additional species of 53, 42, 36, 28, and 26 kDa have also been seen in rat liver extracts, with the 42-, 28-, and 26-kDa species postulated to be degradation products of the 70-kDa protein (9). The latter protein has been shown to be a member of the ATP-binding P-glycoprotein family through cloning of its complementary DNA from rat liver (12). Few studies of mouse PMPs have been reported, despite the potential in this organism to manipulate peroxisome structure through use of transgenic technology. Here the structure and tissue variability of mouse PMPs are described and compared to those of rats and humans. MATERIALS AND METHODS Materials. Nitrocellulose membranes from Amersham (Arlington Heights, IL), immunoblotting reagents from Sigma Chemical Corp. (St. Louis, MO), goat antirabbit peroxidase-conjugated antibody from Cappel (Malvern, PA), 32P04 from Amersham, electrophoresis reagents from Bio-Rad (Richmond, CA), and biochemicals from Sigma were obtained. Human Zellweger syndrome fibroblast lines GM 228, GM 6231, and GM 6094 were obtained from the Human Mutant Cell Repository (Camden, NJ). Tissue culture media were obtained from GIBCO (Grand Island, NY). Animal and human tissues.Sprague-Dawley white rats or Swiss-Webster white mice were obtained from Charles River Breeding Laboratories (Wilmington, MA) at weaning and maintained on regular food with standard light/dark cycles. New Zealand white rabbits were obtained at age 12 weeks and maintained on regular food. Fresh human liver was obtained from a wedge biopsy and peroxisomal purification begun immediately; autopsy human tissues were obtained from 2-16 h after death and stored at 4°C. Human fibroblasts were cultured in minimal essential medium (GIBCO) with 10% fetal calf serum, harvested by trypsinization and washing, suspended in 0.25 M sucrose, and disrupted by sonication (Branson3 X 30-set bursts small probe at 30% power) or homogenization in a small handheld homogenizer for 30 strokes. Mouse PMP purification and antibody preparation. Mouse liver peroxisomes were purified using differential centrifugation in sucrose (13) and Nycodenz (14) as described briefly in a previous publication (15). Fresh liver (9 g) was homogenized in 27 ml cold 0.25 M sucrose, IO mM Tris-HCl, pH 7.4, I mM EDTA, and centrifuged as outlined in Fig. 1 using a Beckman 52-21 centrifuge, a type JA-20 rotor, and 50-ml polyethylene tubes. At each step, the pellet was homogen&d
PEROXISOMAL
(A)
LIVER
MEMBRANE
237
PROTEINS
HOMOGENATE
10,gOO xr’” b SI 60,060
h
min xr
Post-M
Nuclei
MitochMo”dria
250,~ffnin Post-L
1.4 x 10’
g-min I
PerOxisomes
cykol
(0) ioo
MOUSE
n
Ed DHAP-AT El w
CATALASE CMOCHFiOh4E
COXIDASE
LACTATE DEHMX#XjENASE
FIG. 1. Standard purification of mammalian peroxisomes. (A) The differential centrifugation procedure of Leighton et al. (13) followed by Nycodenz step (N) or linear (No) gradients. (B) The specific activities of various marker enzymes in selected purification fractions. The values are normalized for easy comparison, with 100% corresponding to 4000 units/mg for catalase, 0.09 unit/mg for DHAPAT, 15 units/mg for cytochrome c oxidase, and 1.5 units/mg for lactate dehydrogenase. The percentage of protein relative to that in the initial liver homogenate is listed in parentheses beneath each fraction.
238
WILSON AND KING
with fresh sucrose and recentrifuged and the supematants were pooled. The postL fraction was centrifuged 60 min at 75,000 r-pm in the Beckman TLA 100.3 rotor to yield the cytoplasmic fraction (C) for comparative purposes. The “light mitochondrial” (L) fraction was further purified as Nycodenz (N) and Nycodenz gradient (No) fractions, resulting in peroxisomes which were greater than 99% pure as examined by electron microscopy (15). Identical purification schemes were employed for the isolation of peroxisomes and peroxisomal membrane fractions from 9 g of fresh rat liver or, with proportionate modifications in volume, from almost 1 g of freshly biopsied human liver. Assays for catalase (16), dihydroxyacetone phosphate acyltransferase (DHAP-AT (17)), cytochrome c oxidase (13), and lactate dehydrogenase (19) as marker enzymes were by standard methods. Membranes were purified from various fractions by diluting 1 to 50 with 2.94 ml of 100 mM sodium carbonate, pH 11.5, and centrifuging 1 h at 75,000 rpm and 4°C in the Beckman TLA 100.3 rotor (8). Antibodies to PMPs were prepared by injecting membranes from the No fraction into rabbits as described (15). SDS-PAGE and immunoblot analysti. Protein samples (lo-30 ~1 at l-5 mg/ml) were diluted l/2-1/3 in loading buffer and applied to 12 or 7.5% SDS-PAGE gels in a Bio-Rad MiniProtean II apparatus. Electrophoresis was for 0.5 h at 200 V (50 mA), and gels were either stained in 0.1% Coomassie blue/40% methanol/10% acetic acid, by the silver-staining kit obtained from Bio-Rad, or incubated in transfer buffer prior to immunoblotting. The latter was performed as described (15) using rabbit anti-PMP antiserum diluted l/100, treatment with goat antirabbit serum, and detection by peroxidase staining (Cappel). RESULTS The purification of mouse peroxisomes as modified from prior methods (1315) is diagrammed in Fig. 1A and summarized in Table 1. Figure 1B demonstrates that l-2% of the protein in the initial S1 supematant from mouse or rat liver will be recovered in the peroxisomal N fraction, with greatly enriched ratios of peroxisomal enzyme specific activities (catalase and DHAP-AT) compared to cytoplasmic (lactate dehydrogenase) or mitochondrial (cytochrome c oxidase) enzyme specific activities. The specific activities are normalized to a 100% value for convenient comparison. The ratio of DHAP-AT to cytochrome c oxidase is some lOO-fold greater in the peroxisomal N fraction compared to the mitochondrial M fraction. Similar enrichment in the ratios of the peroxisomal enzymes to microsomal (GP-AT) or lysosomal (acid phosphatase) enzymes was observed but not shown. In quantitative terms (Table l), 25-fold enrichment of catalase specific activity with 29% yield was obtained. Since only about half of liver catalase is in the peroxisomal matrix, this represents an actual yield of over 50%. Purification of rat peroxisomes yielded virtually identical results to those shown in Fig. 1 and Table 1. A typical yield of catalase was 23% with a 27-fold enrichment of specific activity. Results with human specimens were much more variable. Purification was not successful with postmortem liver, fresh placenta, or fresh liver obtained at liver transplant. The latter sample was extremely cirrhotic and difficult to homogenize. Successful purification of human liver peroxisomes was only possible using fresh tissue from a wedge biopsy specimen obtained for trans-
w/ml
8.4 5.2 18.2 5.7 3.1 8.4 5.7 4.6
Volume (4
46 58 1.5 71 68 2 0.8 0.5
Fraction
5 Post-M M Post-L C L N NG
Protein
385 300 21 405 210 16.8 4.5 2.3
Total mg 1215 1060 720 1130 322 10,160 20,150 19,380
units/ml 56,000 61,000 1,050 81,000 22,000 20,320 16,100 9,650
Total units
Catalase
TABLE 1 Purification of Peroxisomes from Mouse Liver
145 204 58 200 105 1200 3600 4200
units/mg
100 110 145 37 29 17
1.4 8.3 25 28
Yield (%) 1 1.4
Fold
8 w
z
5 $ 3
5
2 g-
E t3
z
240
WILSON AND KING
123
123
FIG. 2. SDS-PAGE of mouse peroxisomal purification fractions. The upper panel shows discontinuous 12% SDS-PAGE containing 50 pg whole protein from fractions S, (lane l), Post-M (2), M (3), Post-L (4), C (5) L (6), and N (7) illustrated in Fig. 1. The left lane contains Sigma standard 6H with proteins of 205, 116, 97.4, 66, 45, and 29 kDa and the right lane contains standard 7 with proteins of 66, 45, 36, 29, 24, 20, and 14 kDa. The middle panel shows the corresponding membrane proteins as isolated by sodium carbonate treatment from fractions S1 (lane l), M (lane 2), L (4), and N (5). Lanes 3 and 6 show the membrane proteins extracted with 1% Triton X-100 from fractions M and N, respectively. The 6H (left) and 7 standards (right) are again included. The lower panel displays single-layer SDS-PAGE with total (left) or membrane (right) proteins from human (l), rat (2) or mouse (3) No fractions compared to the 7 standard (right lane in each group). Arrowheads indicate the positions of 145-, 70-, and 22-kDa PMPs. Gels in the upper two panels were stained with both Coomassie blue and silver; those in the lower panel were stained with Coomassie blue only.
plant evaluation-960 mg of tissue yielded 240 pg of N fraction with 20% yield of catalase and 25fold enrichment of specific activity. Purification of mouse peroxisomes is reflected by the pattern of whole or membrane proteins revealed by SDS-PAGE electrophoresis as shown in Fig. 2. The upper and middle panels confirm the observations of Fujiki et al. (8) that peroxisomal total and membrane proteins are quite different from those of other
PEROXISOMAL
123456
MEMBRANE
7
241
PROTEINS
8
910 P w
FIG. 3. Immunoblot analysis of mouse, rat, and human homogenates after 7.5% SDS-PAGE. Lanes l-3 represent mouse purification fractions (Fig. 1A) S, , L, and N; lanes 3-5 represent mouse, rat, and human N fractions prepared from liver homogenates; lane 6 is a negative control containing 40 pg protein from the cytoplasmic fraction (C in Fig. 1A) which should contain no peroxisomes. Lanes 7-10 represent control human fibroblast or Zellweger syndrome fibroblast lines GM 228, GM 6231, and GM 6094, respectively.
organelles. It can be seen from the middle panel that the 70-kDA PMP is selectively extracted by sodium carbonate (lane 5) and particularly by Triton X-100 (lane 6), which may explain the bias of rabbit antibodies toward this protein (see below). The lower panel demonstrates that the major PMPs obtained from the N fractions of mouse, rat, or human liver exhibit striking size conservation after 12.5% SDSPAGE and staining with Coomassie blue. Peptides of 145, 70, 55, 36, and 22 kDa are seen clearly in the liver peroxisomal membrane fractions of all three species, with the ratio of PMP 55 to 70 being much greater in the human liver fraction. Intradermal injection of highly purified mouse liver peroxisomal membranes into rabbits allowed the preparation of polyclonal antibodies to mouse PMPs. These antibodies all exhibited a bias toward the 70-kDa PMP as shown by the immunoblotting results displayed in Fig. 3. The correspondence of PMP abundance with increasing mouse peroxisome purity as depicted in lanes l-3 and the absence of reactivity in fractions depleted in peroxisomes (lane 6) support the specificity of the antibody for PMPs. The antibody to mouse peroxisomal membranes reveals PMPs of similar size in rat (lane 4) and human (lane 5) peroxisomal preparations; this antibody also identifies 145- and 70-kDa peptides in homogenates of fibroblasts from control human (lane 7) and Zellweger syndrome (lanes 8-10) fibroblast lines. Since the immunoblot shown in Fig. 3 was produced from 7.5% SDS-PAGE, the 36- and 22-kDa PMPs have migrated off the gel in this experiment. The data in Fig. 4 suggest that the 55 and 36-kDa peptides may be derived from the 145- or 70-kDa peptides. Various human tissues obtained at autopsy and stored at 4°C overnight prior to homogenization and freezing exhibited minor amounts of the 145- and 70-kDa PMPs but increased amounts of the 55- or 36kDa PMPs (lanes 5-7, lower panel Fig. 4). The difference is most evident when comparing lane 2, derived from the homogenate of fresh biopsy liver, with lanes 3 and 4 derived from autopsy liver specimens (Fig. 4, lower panel). The upper
242
WILSON AND KING
123456789
1 2 3 4
5 6 7 8 910111213
FIG. 4. PMPs detected by 7.5% SDS-PAGE and immunoblotting of mouse and human tissues. The upper panel shows mouse tissues including unfractionated liver homogenates from the same individual after 0 h (lane 2), 2 h (lane 3) 6 h (lane 4), or 24 h (lane 5) of storage at 4°C or homogenates of lung (lane 6) brain (lane 7) kidney (lane 8) or adrenal (lane 9) prepared immediately after sacrifice. The lower panel shows human liver homogenates from different individuals at biopsy (lane 2) or autopsy following 2 h (lane 3) or 16 h (lane 4) of storage at 4°C. Other lanes contain kidney (lane 5) brain (lane 6) lung (lane 7) or placenta at 16 weeks (lane 8) 18 weeks (lane 9), 22 weeks (lane lo), 23 weeks (lane 11) 26 weeks (lane 12), or 35 weeks (lane 13) of gestation. All lanes contain 20 pg of homogenate protein except for lane 1 of each panel which contains 2 fig of mouse liver No fraction as control.
panel of Fig. 4 shows an attempt to monitor the degradation of PMPs by comparing liver homogenates from a mouse at various times after sacrifice. Regardless of whether the liver was left within the sacrificed animal (lanes 3-5) or excised (not shown), the 145 and 70-kDa PMPs remained intact for up to 24 h after sacrifice. However, other mouse tissues shown in the upper panel of Fig. 4 including lung (lane 6) or adrenal (lane 9) showed considerable amounts of the 55kDa PMP even when homogenized immediately after sacrifice. Note also that the amounts of PMPs in mouse or human liver and kidney, with their larger peroxisomes, are increased relative to tissues such as brain with “microperoxisomes.” Additionally, a unique peptide of 105 kDa is seen in human placenta (lanes 8-13, lower panel, Fig. 4), and the amount of PMPs in this tissue seems to plateau at 18-26 weeks with a decrease near term.
PEROXISOMAL
MEMBRANE
PROTEINS
243
DISCUSSION Considerable size and sequence conservation of PMPs from mice, rats, and humans are indicated by the SDS-PAGE and immunoblot profiles in Figs. 2-4. This immunologic cross-reactivity demonstrates that genes encoding mammalian PMPs are substantially conserved and supports their role in a highly organized peroxisomal structure. Because denaturing gels are employed, amino acid substitutions between species will not be evident, but significant deletions or rearrangements should have affected electrophoretic migration of the peptides. Conservation over 70 million years of evolution suggests most domains of the integral membrane proteins have important functions. These could include regulation of protein import, anchoring of key enzymes, structural/functional pleiotropy as occurs with xanthine oxidase of Drosophila (22), or maintenance of an internal peroxisomal milieu with organized properties. The purification scheme for peroxisomal membrane proteins and the availability of antibody reagents for detection sets the stage for structural and functional characterization which can decide among these possibilities. The peroxisomal purification scheme described here allows the preparation of 4-5 mg of peroxisomal protein from 9 g of fresh rodent liver in about 6 h time (Fig. 1, Table 1). If very highly purified peroxisomes are needed, then the additional gradient separation produces 2-32 mg in a 12-h time span (No fraction). Treatment of highly purified peroxisomes derived from 9 g of mouse liver with sodium carbonate produces 560-800 pg of membrane protein with a peptide profile which appears identical in rats, mice, and humans (Figs. 2, 3). The 70-, 36-, and 22-kDa peptides visualized by 12% SDS-PAGE have been observed in prior studies of rat peroxisomal membrane peptides (9) and the 145kDa peptide has been demonstrated in rat and human tisues using a polyspecific antibody to rat peroxisomal membranes (5). The availability of a convenient and effective purification scheme for mammalian peroxisomes allowed the preparation of a polyclonal rabbit antibody to mouse PMPs as was described for the rat by Lazarow and colleagues (5,6). Specificity of the rabbit antibody for PMPs is supported by correlation of reactivity with peroxisomal purity (Fig. 3)) with tissue type (Fig. 4), and with clofibrate treatment as demonstrated in a prior study (15). Immunoblotting studies reported here have focused on PMPs of higher molecular weight because of the enhanced blotting efficiency after 7.5% SDS-PAGE electrophoresis. Biased immunoreactivity toward the 70-kDa peptide has also been observed with three separate rabbit antibody preparations, and this may be explained by the greater ease with which this peptide can be extracted from the peroxisomal membrane using a variety of detergents (Fig. 2). In Fig. 3, as demonstrated in prior reports (5,6,10), antibody to rodent PMPs reacts with PMPs of 145 and 70 kDa in Zellweger syndrome and control human fibroblasts. The data in Fig. 4 explain the variable patterns of PMPs seen in some experiments and the difficulties encountered in purifying peroxisomes from human tissues obtained at autopsy. Although PMPs of 145 and 70 kDa are clearly dem-
244
WILSON AND KING
onstrated in fresh mouse or human liver (lanes 2, Fig. 4) variable amounts of 55 and 36 kDa peptides are seen in human autopsy liver (lanes 3 and 4, lower panel, Fig. 4). In placenta, a novel peptide of 105 kDa in addition to the 36-kDa peptide was detected. Origin of the 105kDa peptide from PMP 145 and origin of the 55- and 36-kDa peptides from PMP 145 or 70 by degradation is the favored explanation for these results, although simple refrigeration of mouse liver after sacrifice did not replicate these peptide conversions (lanes 1-4, upper panel, Fig. 4). Hart1 and Just (9) concluded that the 42-, 2%, and 26-kDa peptides observed in rat liver homogenates were breakdown products of PMP 70 in that organism. It will be interesting to examine whether these tissue-specific differences in PMP structure relate to degradation or alternative processing, and whether they have relevance to peroxisomal biogenesis or disease. ACKNOWLEDGMENTS Support of the Biological Humanics Foundation, Dallas Texas, and of NIH Grant ROl-HD26324 made possible certain aspects of this work, which was presented at the Society for Inherited Metabolic Disorders Annual Meeting, Sante Fe, New Mexico, April, 1991.
REFERENCES 1. Wilson GN, Holmes RD, Hajra AK. Peroxisomal disorders: Clinical commentary and future prospects. Am J Med Genet 30:771-792, 1988. 2. Schutgens RBH, Heymans HSA, Wanders RJA, Bosch Hvd, Tager J. Peroxisomal disorders: A newly recognized group of genetic diseases. Eur J Pediatr 144~430-440, 1986. 3. Moser HW. Peroxisomal disorders. J Pediutr 10&89-91, 1986. 4. Arias JA, Moser AB, Goldfischer SL. Ultrastructural and cytochemical demonstration of peroxisomes in cultured fibroblasts from patients with peroxisomal deficiency disorders. J Cell Biol 100:1789-1792, 1985. 5. Santos MJ, Imanaka T, Shio H, Lazarow PB. Peroxisomal integral membrane proteins in control and Zellweger fibroblasts. J Biol Chem 263:10,502-10,509, 1988. 6. Santos MJ, Imanaka T, Shio H, Small GM, Lazarow PB. Peroxisomal membrane ghosts in Zellweger syndrome-Aberrant organelle assembly. Science 239:1536-1538, 1988. 7. Gould SJ, Keller G-A, Schnieder, M, Howell SH, Garrard I-J, Goodman JM, Distel B, Tabak H, Subramani S. Peroxisomal protein import is conserved between yeast, plants, insects, and mammals. EMBO J 9~85-90, 1990. 8. Fujiki Y, Fowler S, Shio H, Hubbard AL, Lazarow PB. Polypeptide and phospholipid composition of the membrane of rat liver peroxisomes: Comparison with endoplasmic reticulum and mitochondrial membranes. J Cell Biol93:103-110, 1982. 9. Hart1 F-U, Just WW. Integral membrane polypeptides of rat liver peroxisomes: Topology and response to different metabolic states. Arch Biochem Biophys 255:109-119, 1987. 10. Small GM, Santos MJ, Imanaka T, Poulos A, Danks DM, Moser HW, Lazarow PB. Peroxisomal integral membrane proteins in livers of patients with Zellweger syndrome, infantile Refsum’s disease and X-linked adrenoleukodystrophy. J Inherited Metab Dis 11:358-371,1988. 11. Suzuki Y, Shimozawa N, Orii T, Hashimoto T. Major peroxisomal polypeptides are synthesized in cultured skin fibroblasts from patients with Zellweger syndrome. Pediutr Res 26~150-153, 1989. 12. Kamijo K, Taketani S, Yokota S, Osumi T, Hashimoto T. The 70-kDa peroxisomal membrane protein is a member of the Mdr (P-glycoprotein)-related ATP-binding protein superfamily. J Biof Chem 265:4534-4540, 1990. 13. Leighton F, Poole B, Beaufay H, Baudhuin P, Coffey JW, Fowler S, De Duve C. The largescale separation of peroxisomes, mitochondria, and lysosomes from the livers of rats injected with Triton WR-1339. J Cell Biol37:482-513. 1968.
PEROXISOMAL
MEMBRANE
PROTEINS
245
14. Ghosh MK, Hajra AK. A rapid method for the isolation of peroxisomes from rat liver. Anal Biochem 159:169-74, 1986. 15. Wilson GN, King T, Argyle JC, Garcia RF. Maternal clofibrate administration amplifies fetal peroxisomes. Pediatr Res 29:256-262, 1991. 16. Peters TJ, Muller M, and De Duve C. Lysosomes of the arterial wall. I. Isolation and subcellular fractionation of cells from normal rabbit aorta. J Exp Med 136:1117-1123, 1972. 17. Jones CL, Hajra AK. Properties of guinea pig liver peroxisomal dihydroxyacetone phosphate acyltransferase. J Biol Chem 25%3289-8295, 1980. 18. Datta NS, Wilson GN, Hajra AK. Deficiency of enzymes catalyzing the biosynthesis of glycerolether lipids in Zellweger syndrome. N Engl J Med 311:1080-1083, 1984. 19. Kornberg A. Lactate dehydrogenase of muscle. Methods Enzymol 1:441-443, 1955. 20. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein using the principle of protein-dye binding. Anal Biochem 72~248-254, 1976. 21. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 193~265-275, 1951. 22. Beard ME, Holtzman E. Peroxisomes in wild-type and rosy mutant Drosophila melanogaster. Proc Nat1 Acad Sci U S A 84~7433-7473, 1987.
