Identification and molecular characterization of a Zellweger disease gene homolog, PAS5, in Saccharomyces cerevisiae.
Item Type text; Dissertation-Reproduction (electronic)
Authors Gu, Kerong.
Publisher The University of Arizona.
Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
Download date 29/09/2021 12:56:15
Link to Item http://hdl.handle.net/10150/187352 INFORMATION TO USERS
This manuscript ,has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely. event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and contimJjng from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.
UMI A Bell & Howellinformalion Company 300 North Zeeb Road. Ann Arbor. MI48106·1346 USA 313!761·47oo 800:521·0600
IDENTIFICATION AND MOLECULAR CHARACTERIZATION OF A ZELLWEGER DISEASE GENE HOMOLOG, PAS5, IN Saccharomyces cerevisiae
by Kerong Gu
A Dissertation Submitted to the Faculty of the DEPARTMENT OF MOLECULAR AND CELLULAR BIOLOGY In Partial Fulfillment of the Requirements For the Degree of DOCTOR of PHILOSOPHY In the Graduate College THE UNIVERSITY OF ARIZONA
1 995 UMI Number: 9620411
UMI Microfonn 9620411 Copyright 1996, by UMI Company. All rights reserved.
This microfonn edition is protected against unauthorized copying under Title 17, United States Code.
UMI 300 North Zeeb Road Ann Arbor, MI 48103 THE UNIVERSITY OF ARIZONA ® GRADUATE COLLEGE
As members of the Final Examination Committee, we certify that we have Kerong Gu read the dissertation prepared by------IDENTIFICATION AND MOLECULAR entitled ------CHARACTERIZATION OF ZELL WEGER DISEASE GENE
HOMOLOG PAS5 IN SACCHAROMYCES CEREVISIAE
and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of ______Doctor of Philosophy ___
Carol L. Dieckmann ~. G~c=: (2/1/ 7 \ -- Date
II S Date I17t)'J
Date
Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate College.
I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.
Dissertation Director Carol L. Dieckmann 3
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the library.
Brief quotations from this dissertation are allowable without special pennission, provided that accurate acknowledgment of source is made. Request for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author. 4
ACKNOWLEDGMENTS
I want to give special thanks to my research advisor, Dr. Carol Dieckmann. She not only gave me support and guidance throughout this project, but also spent tremendous time to help me with my English. Without her patience and encouragement, all these are impossible.
I would like to thank the members of my graduate committee, Dr. Alison Adams, Dr. Thomas Lindell, Dr. Elizabeth Vireling, Dr. John Little for their timely help and advice.
I also wish to thank my colleagues in the laboratory, Kim Sparks, Doug Roberts, Wei Chen, Lorraine Marnell and former members of our laboratory, Steve Mayer, Eric Weber, Yuxiang Liu, Telsa Mittelmeier and Robin Staples, for their help during my day-to-day experiments.
Finally, I would like to thank my wife Yang Gao, for her love and support which made this possible. 5
To my mother and father 6
TABLE OF CONTENTS
LIST OF ILLUSTRATIONS ...... 10
LIST OF TABLES ...... 12
ABSTRACT ...... 13
CHAPTER ONE: INTRODUCTION ...... 15
Part One: Yeast Peroxisomes ...... 15
Part Two: Human Peroxisomal Disorders ...... 28
CHAPTER TWO: MATERIALS AND METHODS ...... 37
Strains and Plasmids ...... 37
Media ...... 37
Transformation ...... 38
Growth Curve ...... 38
Survival Rate Testing on 3-AT Medium ...... 38
Preparation of Organelles and Nycodenz Gradient Separation ..... 38
Enzyme Assays, Laemmli Gels and Western Blots ...... 39
RNA Preparation ...... 40 7 .
TABLE OF CONTENTS
Nortllem Blots ...... 41
cDNA Cloning ...... 42
CHAPTER THREE: THE YEAST PAS5 GENE IS THE FUNCTIONAL HOMOLOG OF THE MAMMALIAN PAF-l GENE REQUIRED FOR PEROXISOME ASSEMBLY ...... 47
INTRODUCTION ...... 47
RESULTS ...... 54
The Yeast PAS5 Protein Shares Sequence Similarity with the Mammalian Protein Necessary for Peroxisome Assembly '" 54
The pas5 Deletion Mutant Is Partially Defective in Peroxisomal Function ...... 59
Catalase and Acyl-CoA Oxidase Are Not Imported into Peroxisomes, but Thiolase and Peroxisomal Malate Dehydrogenase Fractionate with Peroxisomes in pas5 Deletion Mutant Cells ...... 66
Independent Transcriptional Regulatory Systems Induce Expression of Two Different PAS5 mRNAs ...... 74
Tween-40 Is Also an Inducer for the Longer PAS5 Transcript. 77
Does the pas5 Null Mutant Have a Mitochondrial Defect? " 78
DISCUSSION ...... 80 8
TABLE OF CONTENTS
CHAPTER FOUR: EXPRESSION OF HUMAN PAF-l eDNA IN THE YEAST pasS DELETION MUTANT STRAIN ...... 90
~ODUCTION ...... 90
RESULTS ...... 94
Cloning the Human PAF-I cDNA ...... 94
Making Yeast ura3- Strains ...... 96
Growth Phenotype of pas5 Deletion Mutant Cells ...... 97
Screening for Complementation of the Yeast pas5 Deletion. 100
Complementation Test ...... 102
Slow-growth on Acetate Medium Caused by Polar Effect on CBP 1 Expression ...... 103
No Complementation of Liquid Oleate Growth Defect .... 104
DISCUSSION ...... 105
CHAPTER FIVE: GENERAL CONCLUSIONS AND FUTURE DIRECTIONS ...... 109
Sequence Analysis of the 51-upstream Region of the PAS5 Gene .. 109
Transcriptional Regulation by Different Carbon Sources ...... 112 9
TABLE OF CONTENTS
Localization of Pas5p ...... 113
Overexpression of PASS and Peroxisome Proliferation ...... 114
Other Proteins That May Interact with Pas5p ...... 117
Peroxisomal Membrane Structure in pasS Deletion Mutant Cells .. 117
Pas5p and Peroxisomal Protein Import Pathways ...... 119
APPENDIX: AN STS SEARCH ALGORITHM TO FIND OVERLAPPING FRAGMENTS AND FORM CONTIGS IN PHYSICAL MAPPING .123
ABSTRACT ...... 123
INTRODUCTION ...... 123
ALGORITHM AND DISCUSSION ...... 124
CONCLUSION ...... 129
LITERATURES CITED IN APPENDIX ...... 131
REFERENCES ...... 132 10
LIST OF ILLUSTRATIONS
Figure 1.1 Primitive respiratory chain ...... 16
Figure 1.2 J3-oxidation pathway ...... 18
Figure 1.3 Speculative model of peroxisome biogenesis ...... 29
Figure 1.4 Role of the J3-oxidation pathway and the glyoxylate cycle in gluconeogenesis ...... 31
Figure 3.1 Diagram of the PAS5-CBP 1 gene cluster ...... 55
Figure 3.2 Sequence comparison of human Paf-l p, Podospora anserina Carlp and Saccharomyces cerevisiae Pas5p ...... 57
Figure 3.3 Hydropathy plot of the Pas5p protein sequence ...... 58
Figure 3.4 Alignment of the zinc finger motifs of proteins in the C3HC4 family ...... 60
Figure 3.5 Growth of the yeast pas5 deletion strain on oleic acid medium ...... 64
Figure 3.6 Growth of the yeast strains on Tween and oleic acid medium 65
Figure 3.7 Enzyme activities in the peroxisomes and mitochondria of the pas5 deletion strain ...... 69
Figure 3.8 Coomassie-stained gels and Western blots of the Nycodenz fractions ...... 73
Figure 3.9 Northern analysis of PAS5 transcripts ...... 76 11
LIST OF ILLUSTRATIONS
Figure 3.10 Acidic and basic domains ofPas5p ...... 88
Figure 4.1 Construction of the humanPAF-l cDNA ...... 95
Figure 4.2 Screening protocol for complementation of the yeast pas5 deletion mutant by using the pool of human cDNA PCR products ...... 101 12
LIST OF TABLES
Table 1.1 The pas Complementation Groups in S. cerevisiae ...... 21
Table 1.2 Peroxisomal enzymes ...... 25
Table 1.3 Classification of Peroxisomal Disorders ...... 33
Table 2.1 Yeast and E. coli strains and plasmids ...... 44
Table 3.1 Codon usage in the PAS5 gene of S. cerevisiae ...... 61
Table 3.2 Subcellular distribution of catalase ...... 72
Table 4.1 Growth of different strains on several media ...... 99 13
ABSTRACT
Peroxisome assembly in Saccharomyces cerevisiae requires the products of several genes. A gene encoding a polypeptide with shnilarity to the mammalian peroxisome assembly factor-l (PAF-J) is identified and characterized. Deletion of the PAS5 gene from the Saccharomyces cerevisiae genome results in changes in peroxisomal function. The peroxisomes are
defective in the import of catalase A, which is required for H20 2 breakdown, and acyl-CoA oxidase, which is required for fatty acid catabolism. Due to the exclusion of these enzymes from peroxisomes, the yeast mutant strain shows greater resistance to the catalase inhibitor 3-amino triazol and grows more poorly in liquid oleic acid medium than the wildtype strain. However, thiolase and peroxisomal malate dehydrogenase, two other peroxisomal proteins, migrate with peroxisomes in Nycodenz gradients of organelles prepared from the pas5 mutant strain. The product encoded by PAS5 shares 28.4% identity and 50.4% similarity along its entire length with the human PAF-J protein, which is deficient in some patients with the childhood-lethal Zellweger syndrome. Similar to the yeast pas5 mutant strain, a cell line derived from a paf J Zellweger patient 14 does not imPOlt catalase A and acyl-CoA oxidase into peroxisomes, but thiolase is associated with the peroxisomal membranes. The similarity of the defects in import into peroxisomes of the yeast and mammalian mutant cells is suggestive that the yeast PAS5 and mammalian PAF-l genes have similar functions. 15 CHAPTER ONE
INTRODUCTION
Part One: Yeast Peroxisomes
Peroxisomes (also called microbodies) are different from mitochondria and chloroplasts in many ways. Most notably, these organelles are surrounded by a single membrane, have a diameter of between 0.2 and 1 J.ll1l and do not contain DNA. The term peroxisome was proposed by de Duve for these organelles because they produce and consume hydrogen peroxide (de Duve et a!., 1966). Many oxidases that generate toxic hydrogen peroxide, as well as
catalase, the enzyme that decomposes H20 2 to water and oxygen, are residents of this compartment (Figure 1.1). Based on his pioneering work of the discovery and isolation of subcellular organelles, such as peroxisomes and lysosomes, de
Duve received the Nobel Prize in 1974 (Tolbert et aI., 1981b). Because peroxisomes have no DNA, all of their proteins must be imported from the cytosol. The current model of peroxisome biogenesis proposes that peroxisomes grow by posttranslational import of proteins into existing peroxisomes and new peroxisomes are formed by fission (Lazarow et aI., 1985; Linder et aI., 1989). Figure 1.1 Primitive respiratory chain mediated by hydrogen peroxide (H202) is characteristic of peroxisomes and gives them their name. One of the various oxidases in the particle oxidizes a substrate (RH2)' passing its electrons to oxygen to form hydrogen peroxide, which is reduced to water by catalase.
Electrons for the reduction come either from one of various small molecules
(R'H2) or, if no other donor is available, from hydrogen peroxide itself. 16 o -.r: C\I
w rn :5 < I I ()< ,, I I I I ,I, \ , , N _.0 £ 17
Peroxisomes thus resemble the endoplasmic reticulum in being a self-replicating, membrane-bound organelle that exists without a genome of its own.
Peroxisomes vary in structure and enzyme content even in different cells of a single organism. They also can adapt remarkably to changing growth conditions. Striking examples of adaptations of peroxisome function to the environment of cells have been found in studies of carbon source metabolism in yeast. During exponential growth of yeast on glucose, peroxisomes are generally very difficult to detect and their function is uncertain (van Dijken et aI., 1975;
Parish, 1975). But when methylotrophic yeasts are grown on methanol, they develop large peroxisomes that oxidize methanol (Veenhuis et aI., 1979); when baker's yeast is grown on fatty arids, they develop large peroxisomes that break down fatty acids to acetyl CoA (Veenhuis et aI., 1987; Evers et aI., 1991).
Catalase and other peroxisomal enzymes are also induced by fatty acids
(Skoneczny et al., 1988). The ability to induce the proliferation ofperoxisomes and of peroxisomal proteins has made yeasts attractive systems for the study of peroxisome biogenesis and for the cloning of genes encoding peroxisomal proteins. Moreover, the demonstration of an inducible peroxisomal f3-oxidation system in Saccharomyces cerevisiae (Figure 1.2)(Veenhuis et aI., 1987), has Figure 1.2 J3-oxidation pathway. Fatty acids are converted in peroxisomes to yield acetyl coenzyme A, the fuel for the glyoxylate cycle and hence for gluconeogenesis (the new production of glucose and other carbohydrates). The
J3-oxidation pathway that accomplishes this conversion begins with an acyl coenzyme A, an activated fatty acid that is linked to acyl CoA. Each round of 13-
oxidation removes two CH2 groups and yields a molecule of acetyl CoA. Only in peroxisomes (not in mitochondria), the first enzyme of the J3-oxidation cycle generates hydrogen peroxide, the hallmark of the peroxisomal metabolism. 18
ACYL CoA o II R-CHz-CHz-CHz-C-S-CoA
OXIDIZED H,O, FLAVO XPROTEIN Acyl-CoA Oxidase REDUCED Oz FLAVO- PROTEIN
H 0 I II R-CHz-r=IrC_C-S-CoA H . HzO Bifunctional enzyme
OHH 0 I I II R-CH.-C-C-C-S-CoA . I I H H
NAD+ Bifunctional enzyme
NADH + H+
o 0 II II R-CHz-C-CHz-C-S-CoA rC~-SH Thiolase
SHORTENED ACYL CoA + ACETYL CoA o 0 II II R-CHz-C-S-CoA + CH3-C-S-CoA 19 pennitted the exploitation of yeast genetics for studies of peroxisomal biogenesis and peroxisomal protein targeting.
Like other yeasts, S. cerevisiae does not contain a mitochondrial (3- oxidation system (Veenhuis et aI., 1987). Therefore, peroxisomes are the only organelle that contain the fatty acid metabolic pathway. Mutants that are unable to grow on oleic acid have been isolated in the laboratory of W.-H. Kunau
(Erdmann et al., 1989; Kunau et aI., 1993). Biochemical methods were applied to determine if the mutant onu-strains (oleic acid non-utilizing phenotype) were peroxisome assembly mutants (pas-mutants) or had defects in the structural genes of the (3-oxidation enzymes (fox-mutants). The catalase and (3-oxidation pathway enzyme activities were measured in either whole-cell extracts or subcellular fractions obtained from lysates of spheroplasts. These data allowed them to distinguish between individual enzyme deficiencies and mislocalization of all or some peroxisomal matrix enzymes to the cytosol. Mislocalization of peroxisomal enzymes was considered to be an indication that peroxisomes are absent or import-incompetent in the pas mutant strains. These conclusions were further verified by electron microscopy and immunocytochemistry.
H. Tabak and colleagues devised a novel positive selection procedure to 20 obtain additional peroxisome biogenesis mutants in S. cerevisiae (van Der Leij et al., 1992). This procedure exploits the fact that since catalase is inhibited by
3-amino-l,2,4-triazol, H20 2 is produced when peroxisomes oxidize fatty acids and tins toxic metabolite strongly inhibits cell growth (Figure 1.2). However,
cells that do not accumulate H20 2 as a result of a non-functional p-oxidation system or ill-assembled peroxisomes are able to survive. The selection is followed by biochenncal analysis sinrilar to that employed by Kunau. Using this selection method, they have identified tIrree new PAS genes involved in peroxisome assembly, as well as five genes that appear to down-regulate peroxisome gene expression. A sinrilar selection strategy was used by Lazarow et al. in the background of a catalase deficient strain to isolate four groups of peroxisome biogenesis mutants (peb mutants) (Zhang et aI., 1993a).
All together, as many as 18 pas complementation groups have been reported (Table 1.1). Some of these PAS genes have been cloned and studied, and some have not yet been described in detail in the primary research literature.
Sequence analysis of the identified PAS genes and their predicted gene products provides some hints for their functions. Pas 1p and Pas8p belong to a putative
ATPase protein family (Erdmann et aI., 1991; Voorn-Brouwer et aI., 1993). 21 Table 1.1 The pas Complementation Groups in S. cerevisiae
Mutant Properties of correspond Reference wild type gene products pasl A TP-binding protein (Erdmann et al., 1991) pas2 Ubiquitin-conjugating (Wiebel and Kunau, protein 1992) pas3 48 kDa integral (Rohfeld et al., 1991) peroxisomal memebrane protein pas4 Zinc finger-like motif (reviewed in Kunau et al., 1993 and Mertens and Kunau, unpublished data) pas5 Zinc finger-like motif (reviewed in Kunau et af., 1993 and Skaletz and Kunau, unpublished data) pas6 Carboxyl-tenninal SKL (reviewed in Kunau and Hartig, 1992 and Skaletz and Kunau, unpublished data) pas7 or pebl WD-40 protein (Marzioch et al., 1994; Zhang and Lazarow, 1995) pas8 ATP-binding protein (Voorn-Brouwer et al., 1993) paslO Tetratricopeptide-repeat (VanDer Leij et al., protein 1993) pasl4 Same as Snflp (Van Der Leij et al., 1992) 22 pas19 Same as Adrl p (reviewed in Elgersma et al., 1993 and Elgersma, unpublished data) pas20 N/A* (Elgersma et al., 1993) pas21 N/A (Elgersma et al., 1993) pas22 N/A (Elgersma et al., 1993) peb2 N/A (Zhang et al., 1993a,b) peb3 N/A (Zhang et al., 1993a) peb4 N/A (Zhang et al., 1993a,b) peb5 N/A (Zhang et al., 1993a,b) * N/A= not detennined. 23
Pas2p is a ubiquitin-conjugating enzyme and essential for peroxisome biogenesis
(Wiebel et al., 1992). Pas3p has two transmembrane domains and is a
peroxisomal membrane protein (Hohfeld et aI., 1991). Pas7p (peb1p) is a WD-
40 protein and essential for import ofthiolase into peroxisomes (Marzioch et aI.,
1994; Zhang et aI., 1995). Pas10p is a tetratricopeptide-repeat protein and
essential for the import of matrix proteins containing SKL-like targeting signals
into peroxisomes (van Der Leij et aI., 1993).
Development of classical and molecular genetics of other yeasts, e.g.
Pichia pastoris and Hansenula polymorpha has recently increased the number
of organisms suitable for genetic dissection of peroxisome biogenesis. This is
emphasized by the isolation of peroxisomal mutants of H. polymorpha (Cregg
et aI., 1990) and P. pastoris (Gould et aI., 1992; Liu et aI., 1992). In these
methylotrophic yeasts, peroxisomes are absolutely required for the metabolism
of methanol. Methanol-utilization defective mutants are easily isolated. These
mutants supply more information for peroxisome biogenesis. For example,
Pichia pastoris Pas5p and Pas8p are structural and functional homologs of
Saccharomyces cerevisiae Pas 1p and Pas lOp, respectively (Spong et aI., 1993;
McCollum et aI., 1993). In addition, peroxisomal mutants of Chinese hamster 24 ovary (CHO) cells provide a model system for higher eukaryotes. CHO cell mutants that resemble, both phenotypically and genotypically, the fibroblasts from patients with known genetic defects in peroxisomes, are very useful for the understanding of human peroxisomal disorders (Tsukamoto et aI., 1990).
It is now generally accepted that peroxisomes develop by importing peroxisomal proteins that are encoded by nuclear genes and synthesized on free cytosolic polyribosomes (Table 1.2)(Lazarow et aI., 1985; Borst, 1989). At present, two different peroxisomal targeting signals (PTS) have been identified which are capable of directing matrix proteins to the organelles. The first signal,
(pTS 1), resides at the extreme carboxy terminus of the protein and includes the tripeptide motif-SKL or degenerate forms of that motif e.g. S/A/C-KlR/H-LIM.
The PTS 1 motif has been shown to serve as a general microbody targeting sequence in animals, plants, and yeasts (Gould et aI., 1987,1989,1990).
However, in particular in yeast, variations of the PTSI motif were found, indicating that degenerative sequences of the original PTS 1 are functional in these organisms. Examples of these include the carboxy-terminal tripeptide AKI
(Aitchison et aI., 1991), SKI (Didion et aI., 1992), SKF (Kragler et aI., 1993),
ARF, and NKL (Hansen et aI., 1992) which act as a PTSI in different yeast 25 Table 1.2 Peroxisomal enzymes
Remarks Enzyme Reference Cell Type Catalase (Cohen et aI., 1988) S. cerevisiae (Okada et aI., 1987) C. tropicalis (Furuta et aI., 1986) Rat liver (Quan et aI., 1986) Human fibroblast Glycolate oxidase (Lazarow et aI., 1985) Cucumber Urate oxidase (Alvares et aI., 1992) Oxidases Rat liver Alcohol oxidase (van der Klei et aI., 1991) H. po/ymorpha FatlY acyl-CoA synthase (Tolbert, 1981) S. cerevisiae Acyl-CoA oxidase (Dmochowska et aI., S. cerevisiae 1990) Rat liver (Miyazawa et aI., 1987) Human fibroblast (Schram et aI., 1986) Bifunctional enzvrne (Hiltunen et aI., 1992) Fatty acid 13-oxidation S. cerevisiae (Ishii et aI., 1987) pathway Rat liver Thiolase (Glover et aI., 1994) S. cerevisiae (Hijikata et aI., 1987) Rat liver Carnitine acetyltransferase (Lazarow et aI., 1985) C. tropicalis Carnitine (Lazarow et aI., 1985) octanoyltransferase Rat liver 26
Isocitrate lyase (McCammon et aI., 1990) S. cerevisiae Malate synthetase (McCammon et al., 1990) S. cerevisiae Malate dehydrogenase (Steffan et al., 1992) S. cerevisiae (Gietl, 1990) Glyoxylate cycle Watermelon Citrate synthase (Lewin et al., 1990) S. cerevisiae
Alanine glyo~late (Takada et al., 1985) aminotransferase (Purdue et al., 1990) S. cerevisiae Human Dihydroxyacetone (Lazarow et al., 1989) QhosQhate fiHAP) acyltransferase Human Plasmalogen synthesis Alkvl-DHAP synthase (Lazarow et al., 1989) Human Alkvl-DHAP reductase (Lazarow et al., 1989) Human PMP27 S. cerevisiae (Marshall et al., 1995) (Erdmann et at., 1995) PMP31 Candida boidinii (Moreno et at., 1994) Peroxisomal membrane PMP32 Candida boidinii proteins PMP47 Candida boidinii PMP22 Rat liver (Fujiki et al., 1984) PMP70Human (Gartner et al., 1992) PMP35 Human (Tsukamoto et al., 1991) (Shimozawa et aI., 1992) 27 species. The second signal, (PTS2), was identified at the amino tenninus of rat peroxisomal thiolase (Osumi et aI., 1991; Swinkels et aI., 1991). Sequence comparisons with other peroxisomal matrix proteins revealed a consensus sequence, RLx5QIHL, which is observed at the amino tenninus of S. cerevisiae thiolase (Erdmann, 1994; Glover et aI., 1994), watermelon malate dehydrogenase (Gietl, 1990), and Hansenula polymorpha amine oxidase (Faber et aI., 1994). It is currently believed that a third type of topogenic sequence serves as an important targeting signal at internal sites. The acyl-CoA oxidase from Candida tropicalis has two redundant internal topogenic oligopeptides
(Small et aI., 1988). mtemal targeting information is also present in S. cerevisiae peroxisomal enzyme catalase A (Kragler et aI., 1993) and possibly in human alanine glyoxylate aminotransferase (purdue et aI., 1990).
Much less information is available for the targeting of peroxisomal membrane proteins. This class of proteins is also synthesized on free polyribosomes and it is assumed that they are imported directly from the cytoplasm, in a manner similar to that of matrix proteins (Fujiki et aI., 1984;
Lazarow et aI., 1985). However, no PTS has been found in this kind of protein.
There is genetic evidence for the existence of three import pathways for 28 peroxisomal matrix proteins. pas7 mutants exhibit a selective import defect for thiolase, which has an N-terminal PTS, while catalase and other SKL-containing proteins are properly imported (Marzioch et al., 1994; Zhang et al., 1995). pasJ 0 mutants are selectively defective in packaging proteins targeted by -SKL but not thiolase (van Der Leij et al., 1993). In the analysis of peb mutants, a third import pathway has been proposed (Zhang et aI., 1993b). Thiolase remains in the cytosol inpebJ (pas7) mutants, whereas catalase remains in the cytosol inpeb5
(similar to pasJO) mutants. However, in addition to assaying thiolase and catalase, these authors have also assayed acyl-CoA oxidase. They found that acyl-CoA oxidase appears to be inside peroxisomes in both of the pebJ and peb5 selective-packaging mutants. These data are suggestive that the import of acyl
CoA oxidase involves a mechanism that differs from the one used by catalase and the one used by thiolase (Figure 1.3).
Part Two: Human Peroxisomal Disorders
The peroxisome is an almost ubiquitous organelle in eukaryotic cells. It serves varied functions in different species and cell types. There are two major
functions, H20 2-based respiration and fatty acid oxidation, that are shared by Figure 1.3 Speculative model of peroxisome biogenesis suggested by the
observed packaging defects. peb] and peb5 each affect just peroxisomal
enzymes whereas peb2 and peb4 prevent the packaging of all three enzymes.
One possibility is that the PEB5 and PEB] genes encode receptors for PTS1 and
PTS2 proteins. Acyl-CoA oxidase is a PTS3 protein that uses another receptor.
ThePEB2 andPEB4 genes encode parts of the translocation machinery. (From reference Zhang et aI., 1993). 29
en ~ ~ Q) .Q ~en 8. c: ... C\i"----" ... .Q ~ 0 E c.. 8. 0 .§ u
c:: -cu ...0 ...c.. ~ Q) ca ~en E 0 u .!! ~ )( ·u 0 8. ...cu en c.. cu
In mammalian cells, peroxisomes also participate ill plasmalogen synthesis, cholesterol synthesis, and bile acid synthesis (Lazarow et aI., 1989).
Plasmalogens are phospholipids containing a 1,2-unsaturated, long chain alcohol in a vinyl ether linkage to the glycerol backbone. They are one class of ether glycerolipids. Plasmalogens make up about 5 to 20 percent of most cell membranes and are especially abundant in nervous tissue. In myelin, 80 to 90 percent of the phospholipids are plasmalogens.
In human beings, normal function of peroxisomes is required for the survival of the individual. During the last decade many recessive inherited Figure 1.4 Role of the J3-oxidation pathway and the glyoxylate cycle in gluconeogenesis from (a) two-carbon compounds or (b) fatty acids. The reactions are localized within lipid bodies, peroxisomes, mitochondria, and the cytoplasm in eukaryotic cells. Ethanol Triglyceride 1 ~ Glycerol Acetate Fatty acids .;. Acetate (al Fatly
oxaloacetate , aCIds ~ '\. /" 1 (b) Malate/ '{ Acetyl CoA
Acetyl-CoA ~ Citrate /3-oxidation
Glyoxylate )
Isocitrate >-- Peroxisome Succinate
CO2 t Phosphoenol ,-,:-~ Malate ~ Oxaloacetate ~ pyruvate ~ ~ ~ - - _ ---. Glucose
Cytoplasm
Mitochondria
W I-' 32 diseases involving peroxisomes have been described (van den Bosch et aI.,
1992). Up until now, 16 complementation groups have been defined by fusion of cells from patients with peroxisomal disorders (Moser et aI., 1995). These human peroxisomal disorders have been classified into three categories
(Subramani, 1993)(Table 1.3). Group I disorders are the most severe. These diseases are caused by the absence of multiple proteins in the peroxisomes of these patients and the organelles have an abnormal structure (referred to as peroxisomal ghosts) or are undetectable. These severe disorders include
Zellweger syndrome (ZS), Neonatal adrenoleukodystrophy (NALD) and infantile
Refsum disease (IRD). Human patients born with ZS have clinical features that include severe neurologic abnormalities, dysmorphic features, hepatomegaly, and multiple renal cysts. They usually die within a couple of months of birth
(Lazarow et aI., 1989). Group II disorders are single-enzyme defects. The most common disease is X-linked ALD. Group III disorders include diseases exhibiting the loss of a smaller subset of peroxisomal functions. Persons with these disorders also have multiple enzyme defects but peroxisomes are present.
For example, Rhizomelic chondrodysplasia punctata (RCDP) and Zellweger-like syndrome belong to this category. Table 1.3 Classification of Peroxisomal Disorders
Group I: multiple loss of peroxisomal Zellweger syndrome (ZS) functions (peroxisome absent or in ghost Neonatal adrenoleukodystrophy (NALD) form) Infantile Refsum disease (IRD) Group II: single loss of peroxisomal X-linked adrenoleukodystrophy (XALD) function (peroxisome normal) Group III: multiple loss of peroxisomal Rhizomelic chondrodysplasia punctata functions (peroxisome present) (RCDP)
LV LV 34
Studies on human genetic diseases and yeast peroxisome biogenesis mutants has revealed several similarities. Fibroblasts from one complementation group of ZS patients are unable to import proteins with SKL-type sequences into peroxisomes but thiolase is packaged inside the organelle (Motley et aI., 1994).
This is a similar phenotype to that of yeast pas] 0 mutant cells (van Der Leij et aI., 1993). On the other hand, peroxisomes in fibroblasts from patients with
RCDP, which is another kind of peroxisomal disorder, have nonnal packaging of catalase, acyl-CoA oxidase and the bifunctional enzyme, but the packaging of thiolase is defective (Heikoop et aI., 1990). Yeast pas 7 mutant cells have a similar phenotype (Marzioch et aI., 1994; Zhang et aI., 1995). ZS patients from several complementation groups are defective for all peroxisomal matrix proteins, indicating that some gene products are required for the import of all peroxisomal matrix proteins, regardless of the type of topogenic infonnation
(Lazarow et aI., 1989; Motley et aI., 1994). Yeast pas] mutant cells have the same defects as the human cells lacking all peroxisomal proteins (Erdmann et aI.,
1991).
So far, three ZS disease genes have been cloned. The first one encodes a
70 kDa peroxisomal membrane protein and has been mapped to human 35 chromosome 1p21-p22 (Gartner et aI., 1992,1993). It has an ATP-binding cassette, but its function is not yet known. The second one is the peroxisome assembly factor-1 (PAP-I) gene (Shimozawa et aI., 1992). It encodes a 35 kDa peroxisomal membrane protein and is located on human chromosome 8q21.1
(Masuno et aI., 1994). The third gene encodes a receptor for PTS1, named PXRI
(Dodt et aI., 1995). It is the homolog of the Pichia pastoris PAS8 gene and the
Saccharomyces cerevisiae PASIO gene (McCollum et aI., 1993; van Der Leij et aI., 1993). Another ZS disease gene has been assigned to human chromosome
7 q 11.23 by the positional cloning method but the gene has not been found yet
(Naritomi et aI., 1988,1989; Santos et aI., 1993).
This dissertation describes the characterization of a novel yeast gene, which was originally found as an unidentified open reading frame (ORF) located on the left ann of Chromosome ten immediately upstream of the CBP I gene. The
S. cerevisiae nuclear gene, CBP I, encodes a mitochondrial protein required for the stability of cytochrome b pre-mRNA (Dieckmann et aI., 1984a,1984b,
1985,1987; Weber et aI., 1990; Mittelmeier et aI., 1990,1993). A screen of the genome database with the ORF protein sequence found that it is homolog of the human PAF-I gene. The deduced amino acid sequence of the ORF has 36 homology with the human PAF-} protein along the entire amino acid sequence.
The protein product of the ORF contains a putative zinc-finger-like motif and a putative transmembrane domain similar to those in the mammalian PAF-} protein. I present genetic and biochemical evidence showing that the ORF indeed encodes a peroxisomal protein and is a functional homolog of the mammalianPAF-} gene. We originally deposited the ORF sequence in GenBank as CRT} (carbon-source regulated transcripts) in February 1992. Recently, we learned that the CRT} has the same sequence as PAS5 (Skaletz and Kunau, unpublished data and W.-H. Kunau, personal communication). To insure that this gene has one name and all publications about it will easily be found, we renamed it PAS5. 37
CHAPTER TWO
MATERIALS AND METHODS
Strains and Plasmids
The yeast and bacterial strains used in this study as well as the plasmid
vectors employed are listed in Table 2.1.
Media
Yeast strains were grown on YNT, YNO, YNA, NM, NE and NS media.
All media contained 0.67% yeast nitrogen base without amino acids (Difco) and
0.1 % yeast extract. In addition, YNT contained 2% Tween-40, and YNO
contained 0.1 % oleic acid solubilized in 0.5% Tween-40. YNA contained 1% sodium acetate (PH 5.5). NM contained 2% methanol. NE contained 2% ethanol.
NS contained 2% sodium succinate. Also, YPD contained 0.67% yeast nitrogen base without amino acids, 2% peptone, 1% yeast extract and 2% glucose. 5-
FOA medium contained 0.67% yeast nitrogen base without amino acids, 1% yeast extract, 2% glucose, 0.1% 5-FOA, and 50 Jlg/ml uracil. Solid media contained 2% agar. Minimal salt media contain 10% glucose (Mayer et aI.,
1989), or 5% galactose or 5% raffinose are used for pre-grown yeast cells. 38
Transformation
E. coli (Hanahan, 1983) and yeast (Gietz et aI., 1991) strains were transformed using standard protocols.
Growth Curve
The cells were pre-grown in liquid YPD medium to stationary phase. 10 ml of YNO or YNT medium in 50 mI flasks were inoculated at a density between 1 and 10 x 104 cells/mI. The cultures were grown with shaking at 30°C.
At times after inoculation, cell density was determined by counting cells with a hemacytometer. The number of cells from two separate flasks were averaged.
Survival Rate Testing on 3-AT Medium
PSM medium contained 0.1 % yeast extract, 2% maltose, 0.06% lauric acid, 0.2% Tween 40 and 25 mM 3-amino-l,2,4-triazol (Sigma). Cells were seeded at 200/plate and colonies were counted after 7 days. The number of surviving colonies from two separate experiments were averaged.
Preparation of Organelles and Nycodenz Gradient Separation
Strains LL20 and LL20L\XAH were grown in a minimal salt medium containing 10% glucose (Mayer et aI., 1989) to a density of 5 x 106/mI. The 39
cells were collected, washed and inoculated at a density of 1 x 105Iml into 24
liters of YNT medium and grown for 24 hours. The cells were harvested and
organelle pellets prepared as described, except the pH was 6.0 (Bonitz et aI.,
1980). Unbroken protoplasts and nuclei were removed by a 10 min
centrifugation step at 500 x g and the resulting supernatant was centrifuged for
25 min at 25,000 x g to obtain a pellet containing peroxisomes and mitochondria.
The pellet fraction was applied to a 40 m115-36% continuous Nycodenz (Sigma)
density gradient and centrifuged at 4°C, 107,000 x g, for 90 min in a Beckman
VTi50 rotor to separate the peroxisomes from the mitochondria (Singh et aI.,
1992). 2 ml fractions were collected from the bottom.
Enzyme Assays, Laemmli Gels and Western Blots
Cytochrome c oxidase (Tzagoloff et aI., 1975), catalase (Chantrenne,
1955), acyl-CoA oxidase (Small et aI., 1985), and malate dehydrogenase
(McAlister-Henn et al., 1987), activities were measured spectrophotometric ally.
The protein concentration was determined by the bicinchoninic acid method
(Pierce Chemical Co.). 300 Jlg of protein from each sample was pelleted and suspended in 5 Jll4X Laemmli sample buffer (0.25 M Tris-HCI, pH 6.8), 40% glycerol, 20% p-mercaptoethanol, 0.024% bromophenol blue, 8% sodium 40 dodecyl sulfate) plus 3 J.ll 1.0 M Tris-HCI (PH 8.5) and 10 J.lI of water. The suspensions were heated at 100°C for 3 minutes and half of the volume (150 J.lg of protein) was loaded onto a 12% sodium dodecyl sulfate-polyacrylamide gel
(Laemmli, 1970). The gel was run overnight at 15 rnA and then stained with
Coomassie blue. For Western blots, the proteins were transferred from the gel to nitrocellulose sheets at 70 volts for 90 minutes. Catalase was detected with a 1: 100 dilution of antiserum raised to catalase of Candida tropicalis (G. Small,
Mt. Sinai School of Medicine). Thiolase was detected with a 1:10,000 dilution of antiserum raised against thiolase of Saccharomyces cerevisiae (W-H. Kunau,
U. Bochum, Germany). Acyl-CoA oxidase was detected with a I: 1000 dilution of antiserum raised against acyl-CoA oxidase of Saccharomyces cerevisiae (J.
Goodman, U. Texas, Dallas). Malate dehydrogenase was detected with a 1:750 dilution of antiserum raised against malate dehydrogenase of Saccharomyces cerevisiae (L. McAlister-Henn, U. Texas, San Antonio). Antigen-antibody complexes were detected by enhanced chemiluminescence using a recommended protocol (ECL; Amersham Corp. Arlington Heights, Ill.)
RNA Preparation
Yeast polyA(+) RNA was isolated at each time point from wildtype yeast 41 strain D273-10B (a) (Tzagoloff et al., 1976) under repressing (minimal salts,
0.3% yeast extract, 10% glucose) and derepressing (minimal salts, 0.3% yeast extract, and 1% oleic acid solubilized by 2% Tween-40) conditions according to previously described methods (Mayer et aI., 1989).
Northern Blots
Equal portions of polyA( +) RNA extracted at each time point from each strain were fractionated on a 1.2% vertical agarose gel. The gel was then blotted onto a Nytran nylon membrane (Schleicher & Schuell, Inc.). Nucleic acids were cross-linked to the blot by exposure to 254-nm UV radiation at the surface of a transilluminator (Model UV Stratalinker 1800; Stratagene, Inc.) for 2 min. The blot was prehybridized for 24 h in a buffer containing 50% formamide, 5X SCC,
50 mM sodium phosphate (PH 6.5), IX Denhardt solution, 250 Jlg of denatured, single-strand salmon sperm DNA per ml, and 0.5% sodium dodecyl sulfate
(SDS), at 45°C with constant agitation. Radioactive probe was added to the buffer and hybridized for 24 h at 45°C with constant agitation. The blots were washed in IX SCC-l % SDS at 45°C for 20 min, followed by O.IX SCC-l %SDS at 65°C for 45 min. The blots were dried briefly and exposed to X-ray film
(Kodak XAR-5) at -70°C with opposing screens (Du Pont Cronex lightning- 42
plus). To prepare PAS5 cRNA probe of high specific activity for Northem
analysis, plasmid pSH2400, which contains the PAS5 gene, was digested with
BamH I. The restriction fragments were dephosphorylated at the 5' ends with
calf intestine alkaline phosphatase (Boehringer Mannheim Biochemicals, Inc.)
and the PAS5 sequence were transcribed with T7 RNA polymerase (Boehringer
Mannheim Biochemicals, Inc.) in the presence of 700 Ci of a_32P_ATP (lCN
radiochemicals) per ml according to the specifications provided by Boehringer
Mannheim. After transcription and DNase treatment, unincorporated nucleotides
were removed from the transcription mixture by passage through a Sephadex G-
50 (Bio-Rad laboratories) column. The catalase A-specific 32P-labeled probes
were obtain by random priming of catalase A sequence as recommended by the
supplier of the labeling kit (Boehringer Mannheim Biochemicals, Indianapolis,
Inc.).
eDNA Cloning
HumanPAF-l cDNA was amplified by PCR with the primer pairs: 5380
(5'gcggatccaaaaatggcttccagaaaagag3') and 5381
(3'cctctagactaaagagcatttacttctg5'). I used 10 III (4.5XI01o/ml pfu) of a human liver cDNA library constructed in phage ",ZAP (Stratagene) as a template in a 43
PCR mixture containing 50 mM KCI, 1.5 mM MgCI2, 1 mg of bovine serum albumin per ml, 2.5 mrits of Taq DNA polymerase (promega), 1.25 mM dNTPs, and 20 ~ of primers 5380 and 5381 in a 100 J.Ll vohnne. th~ PCR reaction was run for thirty-five cycles of 1 min melting at 94°C, 1 min annealing at 44°C, and
2 min extension at 72°C in a Perkin Elmer Cetus DNA thermal cycler PCR system 480. The amplified fragments were digested with BamIll-XbaI and subcloned into the vector, pAAB85 (Adams et aI., 1995). 44 Table 2.1 Yeast and E. coli strains and plasm ids
Strain or plasmid Description Reference or source
Yeast strains LL20 «, leu2-3, leu2-112, his3-11, his3-15 (Orr-Weaver et aI., 1981)
S1S0-2B a, ura3-52, his3, leu2-3, leu2-112, (Lill, 1991) trpl-289
LU20 LL20 with ura3" mutation This study
LL20~XAH LL20 with a complete deletion of the (Liu, 1991) PAS5 coding sequence and replacement with the HIS3+ gene
LL20~XAL LL20 with a complete deletion of the (Liu, 1991) PAS5 coding sequence and replacement with the LEU2+ gene
LU20~PAS LL20~XAH with ura3" mutation This study
S1S0~XAH S1S0-2B with a complete deletion of (Liu, 1991) the PAS5 coding sequence and replacement with the HIS3+ gene
S150~XAL S 150-2B with a complete deletion of (Liu, 1991) the PAS5 coding sequence and replacement with the LEU2+ gene 45
D273-10B a,p+ (Tzagoloff et aI., 1976) DSlO a, ura3-S2, his3-11, his3-1S, leu2-112, Gift from M. leu2-3, lys2, t1trpl, GAL2+ Werner-Washburne
AI2-WT a, p+, trpl, ura3-S2, leu2-3 (Liu, 1991)
CBllp+ a, p+, adel (ten Berge et aI., 1974) KL14-4A a, hisl, trp2 (Wolf et al., 1973)
W303p+ Ct, leu2-3, leu2-112, his3-11, his3-1S, Gift from R. trpl-I, ura3-1, ade2-I, cani-iOO Rothstein
E. coli strains
RRI F-, hsdS20(rn-mB"). ara-14, proA2, (Bolivar et aI., lacYl, galK2, rpsL20(Sm1, xyl-S, mtl- 1977) 1, supE44, K-
Yeast plasmids Yep13rrlO YEp13 with an insert ofaSau3A (Dieckmann et aI., fragment of yeast nuclear DNA 1984) containing PASS, CBP 1 and NUC 1 pRS318 A yeast plasmid with LEU2 and CYH2 (Sikorski et at., markers 1990) 46
pRS318C pRS3 18 with the CAN1 gene inserted This study pRS318CU pRS3 18C with the SacI-Bgffi fragment This study ofYEp13ffl0 inserted pAAB85 YCp50 with an insert of an EeoR I and (Adams et al., 1995) BamH I fragment of the yeast GALI-IO promotor and an insert of a Xba I and Sph I fragment of the yeast actin tenninator pAAP pAAB85 with the PAS5 gene inserted This study
E. coli plasmids pBS (+) Plasmid with T3 and T7 promoters Stratagene flanking a pUC19 multiple cloning site pSH2400 pBS( +) with a 2400-bp insert This study containing the yeast sequence from the Sac I site upstream of PAS5 (at -499 relative to the ATG of PAS5) to the Hind ill site within the CBPI gene (at +793 relative to the ATG ofCBPl) pETlla T7 expression vector (Studier et al., 1990) pET-PAS 5a pETlla with the full-length PAS5 This study coding sequence amplified by PCR from pSH2400 inserted at Nde I 47 CHAPTER THREE
THE YEAST PASS GENE IS THE FUNCTIONAL HOMOLOG OF
THE MAMMALIAN PAF-l GENE REQUIRED FOR PEROXISOME
ASSEMBLY
INTRODUCTION
In all eukaryotic cells, peroxisomes compartmentalize many different cellular fimctions such as fatty acid p-oxidation, ether phospholipid biosynthesis,
cellular respiration (H20 2-forming), gluconeogenesis, bile acid synthesis, and purine catabolism (reviewed in de Duve et aI., 1966; Tolbert, 1981a; van den
Bosch et al., 1992). Contrary to earlier supposition that peroxisomes are deIived from the endoplasmic reticulum, peroxisomal proteins are translated on free
Iibosomes and imported into the organelle (reviewed in Lazarow et aI., 1985).
At least two different import systems operate to translocate proteins from the cytoplasm into peroxisomes (reviewed in Subramani, 1993).
Since biochemically and genetically Saccharomyces cerevisiae is the best characteIized eukaryotic microorganism, and, additionally the most widely 48 developed eukaryotic system for genetic engineering, it should be an excellent model system to study peroxisome biogenesis. Many different genes coding for peroxisomal proteins and regulators for peroxisomal fimctions in Saccharomyces cerevisiae have been collected by several methods (Erdmann et aI., 1989; van
Der Leij et aI., 1992; Elgersma et aI., 1993; Zhang et aI., 1993a). Some of the mutants have defects in peroxisome assembly (pas mutants) or peroxisome biogenesis (peb mutants). pas mutants mislocalize one or more peroxisomal matrix proteins to the cytosol and are defective either in the import apparatus or factors governing the proliferation of peroxisomes (reviewed in Kunau et aI.,
1992; Subramani, 1993). Several of the PAS genes have been characterized.
PAS] and PAS8 proteins are putative ATPases (Voorn-Brouwer et aI., 1993;
Erdmann et aI., 1991), PAS2 protein has similarity to a ubiquitin-conjugating enzyme (Wiebel et aI., 1992), PAS3 protein is known to be a peroxisomal membrane proetin (Hohfeld et aI., 1991), PAS7 protein is a member of the WD-
40 protein family (Marzioch et aI., 1994; Zhang et aI., 1995), and PAS] 0 protein has similarity to an outer mitochondrial membrane protein involved in import,
MAS70 (van Der Leij et aI., 1993).
Peroxisomal fimctions also play a role in human development as 49
evidenced by the lethal effects of genetic disorders, such as Zellweger syndrome
and other related diseases. Zellweger syndrome is a childhood-lethal autosomal
recessive peroxisomal ~isorder (reviewed in Lazarow et aI., 1989; Goldfischer
et aI., 1973; Santos et aI., 1988; Wiemer et aI., 1989). Zellweger patients are
defective in several functions including p-oxidation of very long chain fatty acids
and ether phospholipid biosynthesis (Heymans et aI., 1983; Datta et aI., 1984;
Singh et al., 1984). Due to the ineffective assembly ofperoxisomes, peroxisomal proteins are not imported into the organelle, even though some are synthesized normally in Zellweger fibroblasts. Most of the proteins that are not imported are rapidly degraded (Schram et aI., 1986). However, the peroxisomal enzyme
catalase, which is not imported, is stable in the cytosol after synthesis (Wanders et al., 1985; Santos et aI., 1985). The peroxisomal defects are thought to be the underlying cause of the severe neurologic abnonnalities, disfigurement, enlargement of the liver, and multiple renal cysts characteristic of Zellweger syndrome (Bowen et aI., 1964; Wilson et aI., 1986). Since the most severe peroxisomal disorder is caused by improper assembly of the organelle, the mechanism of protein targeting and assembly of this organelle has been the subject of increasingly intense research. Interestingly, the diverse biochemical 50
and morphological phenotypes of yeast pas mutants resemble many of the phenotypes reported for fibroblasts of human patients with Zellweger syndrome.
GenBank and other sequence databases are tremendously powerful tools for molecular biologists and are becoming essential to elucidate the basic relation of proteins from different organisms, such as yeast and humans. Many possibilities are available. We can work with protein or nucleic acid sequences, to search for homology, to find motifs and to predict protein structures. Through the systematic sequencing of the human and Saccharomyces cerevisiae genomes, more and more structural homologs from these two important organisms will be found by computer analysis of DNA and protein sequences.
For example, a homolog ofthe human adrenoleukodystrophy (ALD) protein was found during the effort to sequence the yeast genome (Bossier et aI., 1994).
A human eDNA for one of several peroxisomal membrane proteins,
PMP35, was isolated and a mutation in this gene, PAF-l (Peroxisomal Assembly
Factor-I), was shown to be the primary defect in two patients with Zellweger syndrome (Shimozawa et aI., 1992,1993). The subcellular phenotype of the human paf-l cells show that catalase is not localized in peroxisomes and acyl
CoA oxidase is greatly reduced, however, pre-thiolase still associates with 51
peroxisomes (Tsukamoto et aI., 1991; Shimozawa et al., 1992; Motley et aI.,
1994). In this Chapter, using computer matching algorithms to search the protein
sequence databases, I matched Saccharomyces cerevisiae Pas5p to the human
PAF-l protein. PAS5was found in the PAS5-CBPI-NUCl gene cluster located
on chromosome ten of the budding yeast (Liu, Gu, & Dieckmann, in press). The
polypeptide deduced from PAS5 DNA sequence shows strong similarity to the
mammalian PAF-l gene product. The yeast and human polypeptides display
50% similarity over -300 residues. They both contain the same kind of zinc
finger-like motif and a putative transmembrane domain. The Saccharomyces
cerevisiae PAS5 gene, with partial sequence identity to the mammalian PAF-l
gene, is also functionally similar. A deletion in the yeast gene causes a phenotype at the cellular level similar to that of mammalian paf-l cells
(Tsukamoto et aI., 1990,1991), in that catalase and fatty acyl-CoA oxidase are not imported into peroxisomes. Like the mutant paf-l Zellweger cells (Motley et al., 1994), thiolase fractionates with peroxisomes in the yeast mutant strain.
Since the pas5 mutant strain has many of the phenotypes described for human
Zellweger cells, further characterization of the PAS5 gene should provide more valuable information for study on human genetic diseases. 52
The regulation of transcription by specific repressors and activators has been well studied in Saccharomyces cerevisiae due to the ease of genetic, molecular genetic and in vitro biochemical analyses. Many different metabolic pathways are subject to transcriptional regulation appropriate to the nutritional state of the cell. For example, when cells must use galactose as a carbon source, the Gal4p activator binds upstream of the transcription start sites of several genes encoding enzymes required for the catabolism of galactose (reviewed in
Johnston et aI., 1992).
Accompanying a change in metabolism, a very dramatic change in intracellular morphology is observed when cells are transferred from fermentation on a glucose-rich medium to the non-fermentable carbon sources glycerol or fatty acids (Stevens, 1981; Evers et aI., 1991). Switching cells from a medium lacking fatty acids to the one containing oleic acid solubilized with a
Tween detergent leads to a massive proliferation of peroxisomes in yeast
(Veenbuis et al., 1987). During such a switch, peroxisomes increase in size and number and the level of peroxisomal enzymes increases dramatically. The induction of peroxisomal function may be primarily by increasing transcription of the genes coding for organellar proteins. For example, the gene coding for 3- 53
keto acyl thiolase, FOX3, is regulated transcriptionally. A cis-acting element
responsible for the increase in transcription was mapped in the 5' upstream
region of the gene, and a trans-factor that binds to that element was found
(Einerhand et aI., 1991,1993). Similar to the induction of peroxisomes,
mitochondrial mass increases ten-fold and the cells begin to respire after
switching to glycerol medium. Many nuclear genes coding for mitochondrial
proteins, PET genes (Tzagoloff et aI., 1990), are up-regulated during the switch
from fermentation to respiration. Hap1p and Hap2p/3p/4p are transcriptional
activators that are responsible for the increase in transcription of these PET
genes (reviewed in Forsburg et aI., 1989; Zitomer et aI., 1992; Guarente, 1992).
Hap 1p responds to the increase in oxygen and heme concentrations and the
Hap2p complex to the carbon source switch.
In this Chapter, I also show the characterization of the transcripts of the yeast PAS5 gene. The yeast PAS5 gene is transcribed from two different promoters, one responding to the peroxisome induction pathway and the other to the mitochondrial induction pathway. Transcription of the full-length 0.9 kb mRNA is induced by growth on media that cause peroxisome proliferation, whereas the 5'-truncated 0.8 kb mRNA increases dramatically during the 54 induction of mitochondrial function. The shorter transcripts initiate downstream
of the first in-frame ATG and therefore would encode polypeptides missing the
first 46 amino acids of the longer PAS5 protein.
RESULTS
The YeastPAS5 Protein Shares Sequence Similarity with the Mammalian
PAF-l Protein Necessary for Peroxisome Assembly
During studies of the Saccharomyces cerevisiae nuclear gene CBP 1, which encodes a mitochondrial protein required specifically for expression of the mitochondrial cytochrome b gene (Dieckmann et aI., 1984a,1984b; Weber et aI.,
1990; Staples et aI., 1993; Mittelmeier et aI., 1993; Chen et at, 1994), PAS5 was found as an unidentified open reading frame immediately upstream of the CBP 1 gene (Liu, Gu, & Dieckmann, in press). PAS5 is in the same orientation as CBP 1 and its tennination codon is 270 base pairs 5' to the ATG of CBP 1 (Figure 3.1).
PAS5 codes for a polypeptide of271 amino acid residues with a mass of 30.7 kDa. Figure 3.1 Diagram of the PAS5-CBP 1 gene cluster. CBP 1 codes for a protein required for the stability of mitochondrial cytochrome b (Dieckmann et al.,
1984a,1984b; Weber et aI., 1990; Staples et aI., 1993; Mittelmeier et aI., 1993;
Chen et aI., 1994). Open boxes represent the reading frames of the two genes.
Arrows show the transcript start sites and the direction of transcription of each gene. Restriction sites are: X, Xho I; B, BamH I; S, Sac I; A. Acc I. 55
,... a. £D o 56
A search ofthe peptide sequence databases with either the FastA (pearson
et aI., 1988), Blast (Altschul et aI., 1990) or GenQuest (Guan et aI., 1993)
matching programs revealed that yeast Pas5p has similarity to the human PAF-l
protein (Shimozawa et al., 1992), showing 28.4% identity and 50.4% similarity
along the entire amino acid sequence. Pas5p also is 28.9% identical and 50.8%
similar to the Podospora anserina CARl protein, which is a peroxisomal protein
and but also has a function related to caryogamy (Berteaux-Lecellier et aI.,
1995). A three way alignment of the yeast, Podospora and human protein
sequences calculated by the Pileup program from the Wisconsin genetics
Computer Group package (GCG) is shown in Figure 3.2.
The yeast protein has a stretch of hydrophobic amino acids in the middle
ofPas5p (Figure 3.2 and 3.3). The residues in this hydrophobic domain are well
conserved between yeast Pas5p and human Paf-Ip (44% identity and 70%
similarity over a stretch of 27 residues). This domain was found as a putative membrane spanning segment for the mammalian polypeptide, since the Paf-1 p is a peroxisomal membrane protein (Tsukamoto et aI., 1991).
Both proteins contain a cysteine-rich subregion near the carboxyl terminus. Paf-1p shows a zinc finger motif within this region. This zinc finger Figure 3.2 Sequence comparison of human Paf-1p (Tsukamoto et aI., 1991;
Shimozawa et aI., 1992), Podospora anserina Car1p (Berteaux-Lecellier et aI.,
1995) and Saccharomyces cerevisiae Pas5p (Genbank accession # M86538;
Liu, Gu, & Dieckmann, in press). The Pileup program from the Wisconsin GCG package was used to align the three sequences. Using the prescribed GAP algorithm for comparison, the calculated similarity and identity were 50.4% and
28.4% between Pas5p and human Paf-1p; and were 50.8% and 28.9% between
Pas5p and Podospora anserina Car1p. Shading show similarity. Boxes show identity. The - line show the putative transmembrane region. The * show the cysteine residues that make up a zinc-finger-like motif. ~=;~ ~ :~,~~~ ~L~ ~ ~~! ~ ~I[I;+~; ~I~~: ~.~ ~~~ ~ ~ ~ ~?m;;~i~ ~ ~~I~.tr ~~ =~ : ~; ~ ~I: :~ ~i ~::pom ~~ ~~~~rHl[: Human [) I lx( Y lfI~] c l:!:J:~:. G G ~ ~ :Ii;i~ ~ ~~~ : !P;
;:~pora ~ ~~ ~ r~::: - :~ .. Human . L :,. :~il: ~I~ ~~~ p ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~i;l~ ~1~i=~!~i;~;I:' Yeast s Podospora T Human K ,~~ ~~imiMI:i~ ~ ~~~~~~; ~ =
;::~pora :\i\\~;.::.;:.j~~ I: ~ ~ I~ ~ ~ I~ ~ I~ \E.;:;:.;:::: ~ ~ ; ~ :\:ffi.::.(.:j~ L PER ; Ir!i ~ I~ ~ ~ DQ N QAT NENE L MA Human L liJ!.£l!3 L [] s weI P LTG A P Jh S D N T ~:fi.:W T S G K E • CAL C G • •••• * * Yeast Podospora A A T S K T G V V GSA Q Human IH :I~~i~ij~ ~~ ~ !j~ ~ ;R: ~i~i~~; * * * * Yeast A ~ SSG 'If'L T ASP V Y :~::~ra !11:i~~ ~~~:~ ~ ~~I:l·~1iliiiiliiiiiii!ii~ : ~[!]~ ~ ! ~ *
U1 --.] Figure 3.3 Hydropathy plot of the Pas5p protein sequence. The relative
hydrophobicity ofPaflp was determined using the Kyte-Doolittle program (Kyte
et al., 1982) with a window size of 7 as plotted with the program MacVector 4.1
(International Biotechnologies, Inc.). Positive values represent hydrophilic regions of the protein, negative values represent hydrophobic regions of the protein. 58
00000000000 00000000000...... , ... L()~rtlN.-iO.-iNrtl~L() I I I I I 59
motif is similar to the zinc-binding domains that are present in a family of
proteins known to interact with DNA (Freemont et aI., 1991). Yeast Pas5p
matches this zinc finger motif sequence well with the critical exception that the histidine in the first finger is not conserved (see Figure 3.2 and 3.4).
Codon usage in the yeast PAS5 gene shows a high degree of bias towards the use of rare codons (Table 3.1). More than two-thirds ofthePAS5 codons are of the class that is rarely used (Sharp et aI., 1987). In general, genes that use a broader range of codons are expressed at lower levels (Montgomery et aI.,
1978). Thus, it is likely that the level ofPas5p in yeast cells is low. Indeed, a protein of the appropriate molecular weight is not readily observable by
Coomassie staining in either the organelle-enriched fractions or the post organellar supernatant of a wildtype strain (Liu, Gu, & Dieckmann, in press).
The pasS Deletion Mutant Is Partially Defective in Peroxisomal Function
To investigate whether or not the yeast gene is a fimctional homolog of the mammalian gene, a pas5 null mutant was created by disruption of the genomic
PAS5 coding sequence. The HIS3+or LEU2+ gene was inserted into the deleted region, from an Xho I site that immediately follows the first in-frame ATG to the Figure 3.4 Alignment of the zinc finger motifs of proteins in the C3HC4 family
(Freemont et aI., 1991). The conserved cysteine and histidine residues are boxed. Pas5p is the yeast protein and Paf-1 p is the mammalian protein
(Tsukamoto et aI., 1991). The bmil gene product enhances tumor development in transgenic animals (Haupt et aI., 1991; van Lohuizen et aI., 1991); RADl8 protein is involved in DNA repair in yeast (Jones et aI., 1988); RAG-l protein activates DNA recombination (Schatz et aI., 1989); Mel-l8 protein is a nuclear protein which has DNA-binding capacity and has been detected in all examined tumor cells (Tagawa et aI., 1990); RINGl is a gene in the class-II region of human major histocompatibility (MHC) complex and it is expressed in all examined tissues (Hanson et aI., 1991); rfp is expressed in a variety of mammalian tumor cell lines (Takahashi et aI., 1988); Psc and Su(z)2 maintain expression patterns in the Antp-c and bx-c loci of Drosophila (Brunk et aI.,
1991) and COPl regulates photomorphogenesis in Arabidopsis (Deng et aI.,
1992). 11 11 11 11 11 11 11 Pas5p KTVlcipR CIGGFPTNPYQ-IAIClc R ANY cIYVlcIVVKALE----WSM CIDAICIGSSG I I I I I 11 I I I I I Paf-lp GKEICIAL CIGEWPTMPH-TIGICIEIHIV-F ClyylC VKSSFLF-DMYFT CIPKICIGTEV I I I I I I I I I I I brni-l HLMICIVL CIGGYFIDATTIIElcILIHls-F CIKTlc IVRYLE---TSKY CIPIICIDVQV I I I I I I I I I I I RAIn 8 LLRlclHI CIKDFLKVPV-LTPICIGIHIT-F CISLlc IRTHLNN---QPN c PLICILFEF I I I I I I I I I I RAG-l SIslclQI CIEHILADPV-ETNlcIKIHlv-F CIRVlc ILRCLKVM--GSY c PSICIRYPC I I I I I I I I I I Mel-18 HLMICIAL CIGGYFIDATTIVElcILIHls-F CIKTlc IVRYLE---TNKY c PMICIDVQV I I I I I I I I I I RINGl ELMIClpIlcILDMLKNTMTTKEICILIHIR-FlclsDlc IVALRSG---NKEIC PTICIRKKL I I I I I I I I I I I I 1_ rfp ETTlclpvlcILQYFAEPM-MLDlcIGIHI-NIlcICAlc LARCWGTAETNVSIC PQICIRETF I I I I I I I I I I I I I I psc HIIlcIHLlcIQGYLINATTIVElcILIHls-FlcIHSlc LINHLRK---ERFIC PRlclEMVI I I I I I I I I I I I I I I COPl DLLlclpIICIMQIIKDAF-LTAICIGIHIS-FlcIYMIC IITHLRN---KSDlc PCICISQHL LJ LJ LJ LJ LJ LJ LJ LJ
(j) o 61 Table 3.1 Codon usa e in the PASS eneof S. cerevisiae* AmAcid Codon Number AmAcid Codon Number Gly GGG 6.00 Thr ACG 5.00 Gly GGA 4.00 Thr ACA 5.00 Gly GGT 2.00 Thr ACT 3.00 Gly GGC 4.00 Thr ~ 5.00
Glu GAG 4.00 Trp n!§ 7.00 Glu QM 7.00 End TGA 0.00 Asp GAT 4.00 Cys TGT 7.00 Asp GAC 1.00 Cys TGC 6.00
Val GTG 5.00 End TAG 0.00 Val GTA 3.00 End TAA 1.00 Val GTT 7.00 Tyr TAT 7.00 Val GTC 4.00 Tyr TAC 8.00
Ala GCG 7.00 Leu TTG 10.00 Ala GCA 9.00 Leu TTA 4.00 Ala g:r 3.00 Leu CTG 6.00 Ala GCC 3.00 Leu CTA 4.00
Ser AGT 2.00 Leu CTT 2.00 Ser AGC 1.00 Leu CTC 4.00 Ser TCG 3.00 Phe TIT 11.00 Ser TCA 1.00 Phe TTC 5.00
Ser :rg 9.00 Arg AGG 6.00 Ser TCC 5.00 Arg AGA 3.00 Lys ~ 5.00 Arg CGG 2.00 Lys AAA 7.00 Arg CGA 1.00
Met ATG 5.00 Arg CGT 3.00 lie ATA 1.00 Arg CGC 1.00 lie ~ 4.00 Asn AAT 9.00 lie ~ 2.00 Asn AAC 6.00
Pro CCG 2.00 Gin CAG 6.00 Pro ~ 6.00 Gin CAA 1.00 Pro CCT 0.00 His CAT 2.00 Pro CCC 2.00 His CAC 4.00 • Codon thought to be preferentially used in highly expressed genes of S. cerevlslae (Sharp et aI., 1987) are underlined. 62
Acc I site that is 38 base pairs downstream of the tennination codon of the PAS5
gene (Figure 3.1). The presence of the deleted allele in strain LL20~XAH was
confinned by Southern blot analysis (data not shown).
Yeast mutants defective in peroxisomal import show an inhibition of
growth on medium containing the oleic acid (Erdmann et aI., 1989), which is used to stimulate the proliferation of peroxisomes in Saccharomyces cerevisiae
(Veenhuis et aI., 1987). In order to metabolize oleic acid, the cells need to have a functional ~-oxidation pathway. The ~-oxidation system only exists in peroxisomes in yeaSt cells (Veenhuis et aI., 1987). To detennine the response of the pas5 deletion strain to the peroxisome-induction medium, the growth rate of the parent wildtype strain was compared to that of the mutant in oleate medium
(YNO). Because the LL20 strain cannot form colonies on oleate plates, I examined the growth rate in liquid medium to detennine whether the absence of
Pas5p would affect the cells. The pas5 deletion mutant grew more slowly and reached a lower final density than the wildtype strain in liquid YNO (see Figure
3.5, top panel). To confinn that growth inhibition was caused only by the deletion of the PAS5 gene, the mutant strain was transformed with a single-copy
CEN plasmid with and without the PAS5 gene inserted. The difference in 63
growth of the CEN-PAS5 and CEN plasmid-containing strains (Figure 3.5,
bottom panel) is similar to the difference between wildtype and the deletion mutant (Figure 3.5, top panel). The reduction in growth of the pas5 deletion
strain on liquid YNO medium is suggestive of a partial defect in peroxisomal function.
Tween-40 is needed for the solubilization of oleic acid in the medium and can be metabolized by the J3-oxidation pathway (Sloots et aI., 1991). In order to investigate the Tween-40 effect on the pas5 mutant, the wild-type and pas5 disruption cells were grown in Tween-40 medium and the growth rates were compared. The mutant grew slower than wild-type. Surprisingly, both wild-type and the disruption mutant grew faster in Tween medium than in oleate medium
(Figure 3.6). Poor growth of yeast cells in oleic acid medium has also been reported by other people (Skoneczny et aI., 1988; Thieringer et aI., 1991). This may be caused by the lack of enzymes that are necessary to eliminate the interfering cis-double bond in oleic acid. Tween-40 is mixture of saturated fatty acids, whereas oleate is pure 18: 1. However, oleate is a better peroxisomal inducer than Tween-40 according to my experimental data. The yeast cells have a five-fold higher catalase activity in peroxisomal fractions after growth in oleate Figure 3.5 Growth of the yeastpas5 deletion strain on oleic acid medium. The top panel shows the growth of the pas5 deletion strain LL20~ (~, triangles), and the isogeneic wild type strain LL20 (WT, squares). The lower panel shows the growth of the deletion strain containing either plasmid pRS318CU with no insert (t1 + pCEN) or with a PAS5 gene insert (~ + pCEN
P AF). The cells were pre-grown in liquid YPD medium to stationary phase. 10 ml of YNO (oleic acid) medium in 50 mI flasks were inoculated at a density between 1 and 10 x 104 cells/mI. The cultures were grown with shaking at 30°C.
At times after inoculation, cell density was determined by counting cells with a hemacytometer. Each point represents two separately prepared counting chambers. The log phase portion of the curve was fitted by least squares. 64
Growth on oleic acid 1000------
WT
100------~------~------4~------4 -:r------,. ~/ ~ / -- / .... / / / /4 / / • /4/ 4 10 ;:..-" o 3 6 9 12 15 18 21 24 27 30 33 Hours
Growth on oleic acid l000~j------
~ + pCEN PAF
! E 100 I ~ '; ~ .,.,.,.,. ___ ---a u '
o 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 Hours Figure 3.6 Growth of the yeast strains on Tween and oleic acid medium. Cells were grown at 30°C in 50 m1 of medium containing Tween or oleic acid as carbon source. Samples were taken at the times indicated to determine the cell titer with a hemacytometer. 0: the isogeneic wild type strain LL20 in Tween medium; ~: the pas5 deletion strain LL20LUCAH in Tween medium; .: the isogenic wild type strain LL20 in oleic acid medium; and A: the pas5 deletion strain LL20LUCAH in oleic acid medium. 65 o LO
o M
o N
o N o W w ~ ~ 66
medium than growth in Tween medium (data not shown).
The catalase inhibitor 3-amino-1,2,4-triazol (3-AT) was used previously
in a positive selection for mutants defective in peroxisome function (van Der Leij
et aI., 1992). Because catalase is inactive, wildtype cells are killed by the
accumulation of H20 2 generated by acyl-CoA oxidase and other peroxisomal
oxidases, while mutant strains with a nonfunctional J3-oxidation system or ill
assembled peroxisomes survive on 3-AT. To determine if the pas5 mutant strain would show a similar increase in resistance to the drug, the survival of strains
LL20 and LL20L1XAH on plates containing 3-AT was compared (see Materials and Methods). 24% ± 2% of the pas5 mutant cells formed colonies on 3-AT plates, whereas only 0.7% ± 0.2% of the wildtype cells survived to fonn colonies. Similar to the result with oleate medium, the increased ability to survive on 3-AT plates is also suggestive of a defect in peroxisome function in the pas5 deletion strain.
Catalase and Acyl-CoA Oxidase Are Not Imported into Peroxisomes, but
Thiolase and Peroxisomal Malate Dehydrogenase Fractionate with
Peroxisomes in pasS Deletion Mutant Cells 67
To investigate whether mislocalization of key peroxisomal enzymes could
be responsible for the differences in growth on oleate and 3-AT in the pas5 null mutants, I examined the localization of four peroxisomal enzymes (acyl-CoA oxidase, thiolase, catalase A and malate dehydrogenase) in wildtype and pas5 deletion cells by fractionation of organelles on Nycodenz density gradients, followed by assays of enzyme activity (Figure 3.7) and/or immunodetection by
Western blotting (Figure 3.8). Acyl-CoA oxidase is the first and thiolase is the fourth enzyme in the fatty acid J3-oxidation cycle (Figure 1.2)(Kunau et aI.,
1992). Catalase A destroys the hydrogen peroxide fonned as a byproduct of the first step of fatty acid oxidation and the oxidation of other compounds imported for catabolism (Figure 1.1)(de Duve et aI., 1966). Peroxisomal malate dehydrogenase, found only in peroxisomes (glyoxysomes) of fungi and plants, converts oxaloacetate to malate for shuttling to the mitochondria or the cytoplasm (Figure 1.4)(Steffan et aI., 1992; Tolbert, 1981a).
First, the activity of acyl-CoA oxidase, catalase, malate dehydrogenase and cytochrome c oxidase were assayed spectrophotometric ally. Cytochrome c oxidase is a mitochondrial enzyme. Enzymes did not lose their activities on the
Nycodenz gradients, since the recovery of catalase activity after gradient 68
separation was equivalent to that applied (data not shown). Catalase activity
marks Nycodenz fractions containing peroxisomes (top left panel) and
cytochrome c oxidase activity marks the fractions containing mitochondria
(bottom left panel) as shown by the black bars (wildtype cells) in Figure 3.7. A
clear separation of peroxisomes (peak in fractions 4 and 5) and mitochondria
(peak in fractions 17 and 18) was achieved under these conditions. Acyl-CoA
oxidase (top right panel), showed a similar but broader (fractions 4 to 10)
localization to the catalase portion of the gradient. The difference in peak widths
between catalase and acyl-CoA oxidase may be explained by the fragility of
peroxisomes. The soluble matrix enzymes such as catalase A leak out during
isolation, which reduces the density of the organelle (Alexson et aI., 1985). The membrane-bound acyl-CoA oxidase remains with the organelles that have lost matrix enzymes (fractions 6-10). ill comparison to the wildtype strain, the pas5
deletion cells (hatched bars) did not have any catalase and very little acyl-CoA oxidase in the fractions equivalent to those of wildtype, which is suggestive of either the inability or reduced ability of pas5 cells to import these enzymes into peroxisomes, or perhaps the absence ofperoxisomes.
As has been shown for most yeast pas mutants and cells from human Figure 3.7 Enzyme activities in the peroxisomes and mitochondria of the pas5 deletion strain. Organellar pellets from wild type strain LL20 (WT) and strain
LL20LlXAH (il) were separated on a Nycodenz gradient. The top left panel shows the distribution of catalase A in the gradient, catalase A is a peroxisomal enzyme; the bottom left panel shows the distribution of cytochrome c oxidase, a mitochondrial enzyme; the top right panel shows the distribution of acyl-CoA oxidase, a peroxisomal enzyme; and the bottom right panel shows the distribution of malate dehydrogenase activity; there are both peroxisomal and mitochondrial malate dehydrogenases, the former is in the bottom (fraction 1) and the latter is in the top (fraction 20) of gradient. Acyl-CoA oxidase associated with broken peroxisomal membranes is in fractions 16-20 (top right panel;
Alexson et al., 1985). The panel in the following page shows the distribution of protein in the gradient. Wild type values are shown as black bars and mutant values as hatched bars. Catalase Acyl CoA oxidase 50 20
40
30 E E'" '" :; 10 ~ 20
10 • • ... o • 1 III .... o 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 234 7 8 9 10 11 .2 13 14 15 16 17 18 19 20 Cytochrome c oxidase Malate dehydrogenase 8OO.r------, 200ri------~
~I •
E'" 400 I ~ 1001 ~ ~ •
2001 ~
01 • - • BQ! o 2 3 4 5 6 9 10 11 12 13 14 15 16 17 18 19 20 2 3 5 6 8 10 11 12 13 14 15 16 17 18 19 20
Fr.clion Fraclion (J'\ ~ 70
0 C\I en
CXl ,....
CD
It)
' C") C\I C c 0 Q) U 0 - -0 C\J...... u.. a. en CXl ,.... CD It) ' C") C\I Iw/6w 71 Zellweger patients, the yeast pas5 mutant has increased levels of catalase activity in the cytoplasm; the ratio of cytoplasmic to organellar catalase specific activity increased 10-fold (Table 3.2). However, the level of the peroxisomal malate dehydrogenase activity is equivalent in the wild type and pas5 mutant strains in the gradient fractions (bottom left panel), which is suggestive that peroxisomes are still present. Both the peroxisomal and mitochondrial malate dehydrogenases are in the gradient fractions. The former is in fractions 3-9 and the latter is in fractions 11-18. This has been confirmed by Western blot data as discussed below. To verify the enzyme assay data, the presence or absence of catalase, acyl-CoA oxidase, peroxisomal malate dehydrogenase and thiolase was determined by Western blotting of Nycodenz gradients fractions. Use of antibodies against catalase A and acyl-CoA oxidase confirmed the lack of catalase A and the very low levels of acyl-CoA oxidase in the peroxisomal fractions of the deletion strain (see Figure 3.8, lower panel). The ECL antibody detection system appears to overestimate the abundance of the oxidase relative to the measured enzyme activity. As shown in the Coomassie blue-stained polyacrylamide gel of the Table 3.2 Subcellular distribution of catalase LL20 L120 I:lXAH P S Total P S Total 11.16 u 1.20 u 12.36 u 3.74 u 3.12 u 6.86 u Catalase 90% 9.7% 100% 54% 45% 100% P: organellar pellet. S: cytosol fraction. -...J N Figure 3.8 Coomassie-stained gels and Western blots of the Nycodenz fractions. The upper panels are the Nycodenz fractions separated on 10% SDS polyacrylamide gels stained with Coomassie blue. The PET URF lane contains an insoluble fraction prepared from E. coli strain BL21 overexpressing Pas5p from plasmid pET-PAS5a. WT PM and I:l PM lanes contain the post-organellar supernatant fraction. Each numbered lane contains 60 Mg of protein pooled from each of two sequential Nycodenz fractions, e.g. Lane 3 has equal amounts of protein from fractions 3 and 4. The wild type fractions are shown in the top left panel and the pas5 mutant fractions are in the top right panel. Western blot strips are shown below for the same mixed fractions. The two bands seen in catalase lanes 13-19 are non-peroxisomal proteins that cross-reacts with the antiserum. There are three bands seen in the malate dehydrogenase lanes. The top band is peroxisomal malate dehydrogenase. The bottom band is mitochondrial malate dehydrogenase and the middle one is an unknown protein, which cross-react with the peroxisomal malate dehydrogenase antiserum. u. =:la..a..a: :E :E H. ~ ~~. - -97kD -66kD - - -45kD -31kD PAS5p - WT ~ 1 3 5 7 9 11 13 15 17 19 1 3 5 7 9 11 13 15 17 19 : .: ..,."~~~~ .. ~~::;;~%~ Catalase - --: ".c. . Le OY!., ',.~~ ·;~L~·:.~02..... :~~~'.<.i ~~~i[i'~i}~·:,~r··-I!lI!III1-59kD 1 3 5 7 9 11 13 15 17 19 1 3 5 7 9 11 13 15 1719 "'"- ~~ ...... AOX ---.=t==~ ------75KD 1 3 5 7 9 11 13 15 17 19 1 3 5 7 9 1113151719 --- - - MDH3- ::!Ea.---.- - _ .. - -36KD MDH1- --===~ ~--- ...... --- -34KD 1 3 5 7 9 11 13 15 17 19 1 3 5 7 9 11 13 15 1719 . , : i": .... ~ I "t;wuc;:: ~ w 74 gradient fractions (see Figure 3.8, top panel), many major bands in wild-type peroxisomal fractions are also present in the equivalent fractions of the pas5 deletion mutant, suggestive that peroxisomes are indeed present in the mutant strain. Specifically, wildtype cells showed strong antibody staining for thiolase and peroxisomal malate dehydrogenase (Mdh3p) in the peroxisomal fractions (Figure 3.8, Fractions 3-9) and little or no staining in the mitochondrial fractions (Fractions 15-18). The pas5 deletion cells showed the same strong staining for both enzymes in the peroxisomal portion of the gradient, but some thiolase cross-reactivity is apparent also trailing into the mitochondrial fractions. Mdhl p, mitochondrial malate dehydrogenase marks the mitochondrial fractions in the wild type and the pas5 mutant strain (Fractions 11-18). Independent Transcriptional Regulatory Systems Induce Expression of Two Different PAS5 mRNAs To use fatty acids as a carbon source, yeast cells must have a functional peroxisomal p-oxidation pathway, since Saccharomyces cerevisiae does not have a redundant p-oxidation pathway in mitochondria (Veenhuis et aI., 1987). Thus, many yeast genes encoding peroxisomal proteins are transcriptionally up- 75 regulated when cells are grown in medium containing fatty acids (Einerhand et aI., 1991,1993; 19uaI et aI., 1992). A previous study showed that there are two transcripts mad~ from the PAS5 gene (Liu, Gu, & Dieckmann, in press). One is a full-length 0.9 kb mRNA and the other is a 5'-truncated 0.8 kb mRNA. To investigate whether PAS5 transcript levels are increased by induction on fatty acid medium, wild type strains were grown on glucose medium. Subsequently, cells were derepressed on oleate medium. PolyA(+) RNA was isolated at various time points after the cells were switched from glucose medium to oleic acid medium (oleic acid is solubilized in Tween 40) and analyzed by Northern blot. The PAS5 mRNA level was quantitated with a Betagen p-scope and nonnalized to the amount of actin mRNA, which does not change during the experiment. The level of 0.9-kb PAS5 transcript increased 5-fold after switching from glucose to oleic acid medium (see Figure 3.9). The 0.8 kb PAS5 transcript was not detectable. Since the activity of catalase A, an important peroxisomal enzyme, was induced in oleate medium (Skoneczny et aI., 1988), the transcript of catalase A was also studied by Northern blot experiment. The mRNA level of catalase A increased when the cells were incubated in oleate medium after switching from Figure 3.9 Northern analysis of PAS5 transcripts. PolyA(+) RNAs were separated on an agarose gel and transferred to a nylon membrane. The membrane was hybridized to an antisense RNA probe transcribed from a plasmid containing the 600-bp BamH I-BamH I fragment in the PAS5 gene. The number above each lane is the hour after the cells were switched to medium containing oleic acid. Actin mRNA level is not sensitive to changes in carbon source and was therefore used as a control for the amount of polyA( +) RNA loaded in each lane. .a .a 76 ~ ~ ~ 0) • • ,.. 0 I I N J ,... I. . I CO ..0)·1as ' 0' .!I, o I I •• .5 Lt) ... c.; C/) :J o'" The increase in the mRNA level of the 0.9 kb PAS5 transcript started after 4 hours growth in oleate medium. These results may indicate that Pas5p is a protein needed for an early stage of peroxisome proliferation, and peroxisomal matrix proteins are needed later to import into enlarged peroxisomes. Tween-40 Is Also an Inducer for the Longer PASS Transcript When Tween-40 was used as sole carbon source for the growth of Candida tropicalis, the expression of ~-oxidation pathway enzymes increased (Sloots et aI., 1991). I also observed that medium containing only Tween-40, a detergent composed mostly of palmitic acid (18:0), can induce the expression of the 0.9 kb PAS5 transcript as well as medium containing oleic acid (18:1) solubilized with Tween-40 (Northern blot not shown). Similarly, the mRNA level of catalase A also increased when the cells were grown in Tween medium after switching from glucose repression conditions. As with the induction on oleate, the dramatic induction of catalase A mRNA was also delayed 8 hours after the induction of PAS5 mRNA (Northern blot not shown). 78 Does the pas5 Null Mutant Have a Mitochondrial Defect? It had been shown previously that human fibroblasts isolated from Zellweger patients had mitochondrial defects as well as peroxisomal defects (Goldfischer et aI., 1973; Versmold et aI., 1977; Trijbels et aI., 1983; Sarnat et aI., 1983; Muller-Hocker et aI., 1984). The underlying reasons for the mitochondrial defects have not been determined. Some findings point to a defect in the electron transport chain before the step involving cytochromes or at the succinate-ubiquinone oxidoreductase level (Goldfischer et aI., 1973,1979; Trijbels et al., 1983). Versmold et al. described a different mitochondrial defect, a functional abnormality of cytochrome b, in their patients (Versmold et aI., 1977). Thus, the nature of the mitochondrial defect has been variable in different Zellweger patient cells. Molecular analysis of the pas5 deletion mutant should give more details about mitochondrial abnormalities. Therefore, I separated mitochondria on Nycodenz density gradients. Mitochondria from the pas5 deletion mutant have a broader density spectrum than wild type. In the Nycodenz separations of wild type yeast organelles, cytochrome c oxidase and mitochondrial malate dehydrogenase peak in fractions 17 and 18, indicating the location of the 79 mitochondria (see Figure 3.7). However, the pas5 deletion mutant has a broad distribution of these enzyme activities from fractions 11 to 18. The mitochondrial enzyme distribution is also coincident with the distribution of proteins in the gradient and a characteristic mitochondrial banding pattern in the Coomassie stained gel in fractions 15 to 19 for wild type (Figure 3.7 and 3.8). The mitochondrial staining pattern for the deletion mutant extends all the way from fraction 11 to fraction 19. Some thiolase is also observed in these fractions in mutant cells whereas in wild type thiolase is confined to more dense peroxisomal fractions. hnmunomicroscopy will be required to determine if the thiolase found in the mitochondrial fractions is in peroxisomes of altered density or is aberrantly localized to mitochondria. The difference between the two strains in the distribution of mitochondria has been observed for several independent Nycodenz separations, suggesting that the density differences are not an artifact of variation in gradient preparation. From a previous study, the transcription of the yeast PAS5 gene was affected by switching cells to a medium requiring mitochondrial function (Liu, Gu, & Dieckmann, in press). The cells were switched from the fermentable carbon source, glucose, to the non-fermentable carbon source, glycerol. Over 80 a 12-hr period the 0.9-kb PAS5 transcript was constitutively expressed. Surprisingly, a second O.S-kb PAS5 transcript, lUldetectable at the time when the cells were switched, appeared after four hours on glycerol mediwn. All together these findings are suggestive that the pas5 mutant has a mitochondrial defect. More specific analysis is needed, such as analysis of mitochondrial membrane composition and respiratory chain functions. DISCUSSION A search of several databases revealed that Pas5p shares extensive sequence similarity with a group of peroxisomal proteins. This group of proteins includes hwnan Paf-l p, rat Paf-l p, Chinese hamster ovary cell Paf-l p, Podospora anserina Carlp, and Hansenula polymorpha PerSp (Tsukamoto et aI., 1991; Shimozawa et aI., 1992; Thieringer et aI., 1993; Berteaux-Lecellier et aI., 1995; Tan et aI., 1995). PerSp is a peroxisomal integral membrane protein. They are all similar sized proteins and share at least 50% similarity along the entire length of the polypeptides. They all have a long hydrophobic stretch in the middle of the sequence and a putative zinc finger motif near the carboxyl 81 tenninus region. Most importantly, all of the proteins in this group are required for peroxisome biogenesis. The hydropathy profile ofPas5p based on the predicted sequence does not show evidence for a true membrane-spanning region, since a minimum of 19 amino acids is required to span a lipid bilayer CKyte et aI., 1982). The length of the hydrophobic domain in yeast Pas5p is 17 amino acids. However, the amino acids in the hydrophobic stretch are well conserved between yeast and mammalian proteins (Figure 3.2). This hydrophobic stretch in the mammalian protein is thought to anchor the polypeptide in the membrane. The rat Paf-1 p has been shown to be associated with the peroxisomal membrane, with most of its mass exposed to the cytoplasm (Tsukamoto et aI., 1991). Hansenula polymorpha Per8p also lacks a potential membrane spatming domain, but experimental data showed that it is a integral peroxisomal membrane protein (Tan et al., 1995). A Saccharomyces cerevisiae peroxisomal membrane protein, Pmp27p, does not have a predicted transmembrane domain, but is tightly bound to peroxisomal membranes (Marshall et aI., 1995). These findings suggest that Pas5p most likely has a peripheral association with the peroxisomal membrane through the short hydrophobic domains that do not span the membrane. Or, yeast 82 Pas5p may associate with other proteins embedded in the membrane. Because of many hydrophilic parts in Pas5p, it may be on the surface of the peroxisomes. The location of the protein in the cell awaits further experimental analyses. The sequences of the yeast and mammalian proteins are most similar in the zinc finger domain. This zinc finger motif is the same as the motif serving as a DNA binding domain in several other proteins (Freemont et aI., 1991). However, the motifmay be involved in protein-protein rather than protein-DNA interactions, because the mammalian PAF-l protein has been found to be located in peroxisomes, not nuclei. Yeast Pas5p is missing the histidine in the first finger which is important for binding zinc. Perhaps this finger, as for the Chinese hamster ovary PAF-l protein (Thieringer et aI., 1993), is not necessary for peroxisome biogenesis in yeast, or another residue in the polypeptide chain can substitute for the histidine in fonnation of the finger. Zellweger syndrome, adrenoleukodystrophy, and infantile Refsum disease are human genetic disorders with a primary defect in the import of proteins into peroxisomes (reviewed in Lazarow et aI., 1989). Nine complementation groups have been defined for these disorders by cell fusion (Shimozawa et aI., 1993). One human gene that has been characterized, PAF-l, complements the defects 83 of cells from two human patients in complementation group F. The same stop codon mutation has been found in the PAF-l gene of both patients (Shimozawa et aI., 1992,1993). I have verified that the yeast PAS5 gene is a functional homolog of the mammalian PAF-l gene by showing that deletion of the yeast gene results in a similar phenotype to that of the mammalian mutant at the cellular level. Mammalian paf1 mutant cells are similar to yeast pas5 cells in that they cannot import catalase and acyl-CoA oxidase into peroxisomes (Tsukamoto et aI., 1990,1991). In addition, as is true for pafl and several other Zellweger fibroblasts mutant cells, thiolase is associated with peroxisomes in the yeast pas5 mutant (Balfe et aI., 1990; Gartner et aI., 1991; Suzuki et aI., 1992; Motley et aI., 1994). Mammalian 3-ketoacyl-CoA thiolase is targeted to peroxisomes by virtue of an amino terminal extension (Swinkels et at, 1991), rather than by an internal signal or the carboxyl-terminal -SKL type signal. The mammalian N terminal targeting sequence is cleaved from the protein to produce the mature, active enzyme (Subramani, 1993). Several lines of Zellweger mutant cells have been shown to import thiolase into the peroxisomes, but the amino terminal extension is not cleaved (Balfe et aI., 1990; Gartner et aI., 1991; Suzuki et aI., 84 1992; Motley et aI., 1994), suggestive that the maturation protease is not imported into the organelle, or that it is not active. Mammalian mutant paf-l cells have been shown to accumulate the precursor fonn of thiolase, which is associated with peroxisomal ghost membranes (Tsukamoto et aI., 1990,1991; Motley et aI., 1994). Yeast thiolase has a homologous amino terminus to the mammalian protein (Swinkels et aI., 1991), which also is responsible for targeting the protein to peroxisomes (Erdmann, 1994), but the signal sequence is not cleaved after import (Swinkels et aI., 1991). Here I have shown that thiolase fractionates with peroxisomes in the yeast pas5 mutant. There is genetic evidence that yeast catalase A and acyl-CoA oxidase enter peroxisomes by a pathway separate from the thiolase import pathway. pas 10 mutants are defective in import of catalase A but not thiolase and pas 7 mutants are defective in import ofthiolase but not catalase A (van Der Leij et aI., 1992). Since catalase A and acyl-CoA oxidase are not imported into peroxisomes when the PAS5 gene is deleted from the genome, yeast pas5 mutant strains will be useful in the continued analysis of the catalase A and acyl-CoA oxidase import pathway into yeast peroxisomes (see Chapter five). In addition, these strains will be good model systems for the study of peroxisomal defects 85 associated with Zellweger syndrome. In this Chapter, I showed that the yeast PAS5 gene is unique in that it produces two transcripts, each of which is regulated by a separate control . system. Two size classes ofmRNA (0.8 and 0.9 kb) are transcribed from PAS5. The ends of the two PAS5 transcripts were mapped relative to the continuous coding sequence open reading frame by S 1 analysis (Liu, Gu, & Dieckmann, in press). The S 1 mapping data supported the approximate 100 base difference in size of the 0.9-kb and 0.8-kb transcripts obselVed in the Northern blots. The 0.9- kb transcripts are initiated upstream of the reading frame and the abundance of these mRNAs increases five-fold on oleic acid medium. Up-regulation of PAS5 transcription on glycerol medium was observed in the S 1 autoradiography as well as in the Northern blots. Beta-scope scanning of PAS5 transcript levels between the repressed and glycerol derepressed strains showed that the total level of the 0.9-kb mRNA remained relatively unchanged. However, the 0.8-kb PAS5 mRNA is initiated at multiple sites within the coding sequence and increases in abundance up to 10-fold when cells are switched from glucose to glycerol medium; use of glycerol as a carbon source requires mitochondrial function. 86 Since the longer PAS5 transcript is induced by media that induce peroxisome proliferation and the shorter transcript is induced on glycerol medium, which requires mitochondrial fimction, the PAS5 induction experiments may suggest that the longer transcript encodes a protein that has a peroxisomal fimction, but the shorter one encodes a protein that has a mitochondrial function. Indirect evidence supports this. The PAS5 deletion mutation, which destroys both open reading frames, show abnormalities in both peroxisomes and mitochondria. Several yeast genes generate two transcripts with different 5' ends. The short transcript is initiated within the coding sequence, so that the translation product lacks the N-terminal sequence compared to the full-length polypeptide translated from the longer transcript. In most cases, the differences at the N termini result in targeting of the proteins to different cellular compartments. For example, sue2 produces two transcripts and thus two isozymes of invertase (Carlson et aI., 1982). The shorter mRNA is transcribed constitutively and produces a cytoplasmically localized protein. The longer transcript is induced by growth on sucrose and is translated into a protein that has a secretion signal in the N-terminal extension not found in the shorter protein. Other examples 87 include TRM1 and MOD5 that code for tRNA modification enzymes and produce proteins that are targeted either to the nucleus/cytoplasm or the mitochondria (reviewed in Hopper et aI., 1992), and HTS1 and FUM1 which code for histidine tRNA synthetase (Chiu et aI., 1992) and fumarase (Wu et aI., 1987) respectively and are targeted either to the cytoplasm or the mitochondria. The two transcripts of the PAS5 gene maybe translated into two different length polypeptides. PAS5 transcripts initiated upstream of position + 1 should be translated starting at the first ATG to produce a full-length protein, whereas the transcripts initiated downstream of position + 1 should be translated starting at an internal ATG to produce a NH2-terminal truncated protein. As shown by the Sl mapping result, the first downstream in-frame ATG is at codon 46 (Faugeron-Fonty et aI., 1979). Usage of this ATG would result in translation of a 25 kDa protein. This 25 kDa protein would lack the NH2-terminal acidic domain of the full-length 30 kDa protein (see Figure 3.10) and thus may have a different function or localization in the cell. There are eight acidic residues out of a total of sixteen in the whole protein sequence that are within the 47 NH2- terminal amino acids present only in the longer protein. The rest of the protein is very basic. The difference is reflected in the calculated pI, which is 9.02 for Figure 3.10 Acidic and basic domains ofPas5p. The black bar represents the coding sequence of PAS5. The ticks above the bar show the positions of basic residues and the ticks below the bar show the positions of acidic residues. The arrows mark the positions of the first and second in-frame ATG. 88 + I 89 the longer protein and 10.06 for the shorter one. Since rat liver peroxisomal 3-ketoacyl-CoA thiolase, which contains an NH2-terminal extension, has a lower pI value when compared to the mitochondrial isozyme, which has a basic insertion region (Hijikata et aI., 1987), it is my hypothesis that the longer Pas5p polypeptide is targeted to the peroxisomes and the shorter one to the mitochondria. Consistent with this hypothesis, the NH2-tenninus of the shorter protein contains basic and hydroxylated amino acids, a signature of mitochondrial targeting sequences (reviewed in Glick et aI., 1992). COOH terminal -SKL, the best characterized peroxisomal targeting signal (Subramani, 1993), has been found in several yeast peroxisomal proteins (Kunau et aI., 1992), but is not found in Pas5p (carboxyl tenninus is -PVY). And, the NH2- terminal but not the COOH-tenninal sequences of the mammalian PAF-J protein have been shown to be critical for peroxisome function (Thieringer et aI., 1993). Further investigation of the localization of the yeast PAS5 gene products will be required. 90 CHAPTER FOUR EXPRESSION OF HUMAN PAF-l eDNA IN THE YEAST pasS DELETION MUTANT STRAIN INTRODUCTION Like other eukaryotic cell peroxisomes, these organelles in human cells contain a number of matrix proteins and perform reactions in diverse metabolic pathways including the oxidative degradation of fatty acids and the synthesis of ether lipids (Tolbert et aI., 1981a,1981b; Lazarow et aI., 1989). The metabolic importance of peroxisomes in humans has been better understood with the cloning of the PAF-l gene, a mutation in which causes the human genetic disease, Zellweger syndrome (Shimozawa et aI., 1992). Human paf-l mutant cells do not localize catalase, acyl-CoA oxidase and peroxisomal dihydroxyacetonephosphate (DHAP) acyltransferase inside peroxisomes (DHAP acyltransferase catalyzes the first step of plasmalogen synthesis). Also, in these cells, 3-ketoacyl-CoA thiolase is not processed to the mature form, but its precursor is found (Tsukamoto et aI., 1991; Shimozawa et aI., 1992). 91 Interestingly, the cells from patients of some Zellweger complementation groups do import thiolase, but not other peroxisomal proteins, into peroxisomal "ghosts" (Balfe et aI., 1990; Gartner et aI., 1991). Also, Motley et al. show that pafl mutant fibroblasts belong to one of the complementation groups that show a weak association ofthiolase with peroxisomes (Motley et aI., 1994). The PAF-l gene is well conserved in different eukaryotic organisms. Homologs of the human gene have been found in rat cells, Chinese hamster ovary cells, the fungus Podospora anserina, and the yeasts Hansenula polymorpha and Saccharomyces cerevisiae (Tsukamoto et aI., 1991; Thieringer et aI., 1993; Berteaux-Lecellier et aI., 1995; Tan et aI., 1995; Gu et aI., 1995). All of the proteins not only share sequence homology, but also contain the same kind of zinc-finger-like motif and a similar putative transmembrane domain. Furthermore, they all are required for peroxisome assembly. In Chapter three, I reported the identification and molecular characterization of the yeast gene PAS5 as a functional homolog of the human PAF-l gene. The human Paf-Ip and yeast coding sequences were shown to be 28% identical and 50% similar. My data also showed that pas5 mutant cells are unable to import catalase and acyl-CoA oxidase into peroxisomes as in the 92 mammalian paf-I mutant cells. pas5 mutant cells also import thiolase into peroxisomes. Thus, the peroxisomal protein import defects in pas5 mutant cells mimick that found in Zellweger syndrome cells. Peroxisomal proteins are conserved between yeast and human cells. Recently, another yeast peroxisomal protein's functional counterpart in humans was reported (Dodt et aI., 1995). The human homolog of P. pastoris PAS8 and S. cerevisiae PASIO genes, PXRI, was found. Mutations in the PXRI gene from fibroblasts of two Zellweger syndrome patients caused a defect in import of proteins with the peroxisomal targeting signal-1 (PTS 1). Furthennore, the human PXRI gene complemented the methanol and fatty acid growth defect of the pas8 mutant strain of P. pastoris. Since Pas8p has been demonstrated to be a PTS1 receptor in yeast, Pxr1 p is indeed a PTS 1 receptor in human. There are several published papers about the functional replacement of a yeast protein witll its human homolog. The essential function of the S. cerevisiae ALAI gene, a novel alanyl-tRNA synthetase, was rescued by expression of a human enzyme eDNA (Ripmaster et aI., 1995). Human RNA polymerase II subunit seven (hsRBP7) restored viability to S. cerevisiae strains lacking the conventional RNA polymerase gene RPB7 (Khazak et aI., 1995). Human lUfOA 93 GTPase, a member of subfamily of ras-related GTP-binding proteins, replaced its functional homolog RHO} in a yeast strain (Qadota et aI., 1994). These findings indicate that although the morphology of cells varies widely across the phylogenetic tree, important components from these diverse systems have common functions in vivo. Yeast Pas5p and human Paf-l p are functional homo logs in vivo when considering peroxisome assembly. Perhaps the most direct way to study this relationship is to examine the behavior of the human protein in yeast cells that do not have the corresponding endogenous gene. This approach is useful in helping us to understand and compare peroxisome biogenesis in both yeast and human cells. Therefore, I report here the analysis of the human PAF-} cDNA in the yeast pas5 null mutant, and the testing of whether Paf-l p from mammals and yeast Pas5p are conserved in function as well as in primary structure. 94 RESULTS Cloning the Human PAF-l eDNA I took the advantage of finding the sequence of the human PAF-l gene in GenBank to design two oligonucleotide primers for use in a polymerase chain reaction (PCR). The upstream primer contains the sequences for Bam ill and AAAAjust 5' of the ATG and the adjacent coding sequences. The AAAA is the best yeast sequence for translation initiation. The downstream primer contains coding sequences and sequences for Xba I just 3' of the stop codon. cDNA fragments were directly generated from a human liver cDNA library (see Materials and Methods). An amplified cDNA fragment with the expected size was found. Measuring the length of restriction endonuclease cutting fragments on an agarose gel, I confinned that it was the human PAF-l cDNA clone. The PCR-generated human PAF-l cDNA was digested with Bam ill and Xba I and inserted into the centromere-based yeast plasmid pAAB85 (Adams et aI., 1995), between the GALl promoter and actin terminator (Figure 4.1). I named this construct pAAP. Figure 4.1 Construction of the human PAF-l eDNA. A BamID-XbaI fragment containing the human PAF-l gene was ligated into BamID and XbaI-digested pAAB85 (Adams et aI., 1995) to produce recombinant plasmid pAAP for expression in S. cerevisiae. 95 a.. Q) -o E a.e o a. :!: z~ ow 96 Making Yeast ura3- Strains The yeast plasmid pAAP contains the URA3 gene as a selection marker, but LL20 is not a uracil auxotrophic strain. In order to select the yeast transformants with the plasmid pAAP, I needed to make ura3- mutations in LL20 wild type and the pasS deletion strains. 5-fluoro-orotic acid (5-FOA) medium can be used to select for mutant cells that carry the ura3- mutations. ura3- strains fail to utilize orotic acid as the source of the pyrimidine ring (Boeke et aI., 1984). Therefore, LL20 wild type and the pasS deletion mutant strains were grown in YPD medium overnight, and then 3xl07 cells were placed on each plate of 5- FOA medium. All of the plates were incubated at 30°C for 3 days. Five colonies were chosen from the 20-30 5_FOAresistant(R) colonies that grew on each plate. Wild type cells convert 5-FOA to an inhibitor ofthymidylate synthetase (Boeke et aI., 1984). This inhibitor is thus toxic to the cells. In the process of converting 5-FOA, there are two steps that are catalyzed by the products of the yeast genes URAS and URA3, respectively. Therefore, both ura3- and uraS mutant cells can grow on 5-FOA medium. In order to identify only the ura3- mutants, I crossed the 5-FOA Rcolonies with a DS 10 strain, which has the ura3- mutation. After testing the uracil auxotrophic phenotype of the diploid cells, I got 97 ura3- mutants of both LL20 wild type and pas5 deletion strains and renamed strains LU20 and LU20APAS, respectively. Growth Phenotype ofpas5 Deletion Mutant Cells In a previous study, the pas5 null mutant was created by deletion of the PAS5 gene from two different strains (LL20 and SI50-2B). However, the pas5 null mutant cells do not have an apparent phenotype on glucose, glycerol or galactose media, or media containing different metal ions, or at different temperatures (Liu, 1991). In Chapter three, the pas5 null mutant cells showed a slower growth rate than wild type cells when they were grown in liquid oleate medium. But in order to screen the yeast transformants with the human PAF-l PCR product pool (see reason below), I still needed a good phenotype to score on plates. Both wild type and the pas5 deletion mutant of the S 150-2B strains can grow on oleate plates without a difference in colony size. However, neither wild type nor the pas5 deletion mutant of the LL20 strains form colonies on oleate plates. Microscopic observation of microcolonies that failed to become visible colonies revealed that the spores germinated but ceased to grow after five or six cell divisions. I have 98 also surveyed several other yeast wild type strains on oleate plates. Not surprisingly, some grew and some did not (see Table 4.1). Thus, oleate plates would not make a good selection medium for complementation of the pas5 null mutant. In order to grow on methanol medium, yeast cells need to have peroxisomal enzymes alcohol oxidase, dihydroxyacetone synthase and catalase (Douma et al., 1985). The utilization of ethanol or acetate by yeast cells requires the glyoxylate cycle, at least two enzymes of which, isocitrate lyase and malate synthase, are peroxisomal proteins (Zwart et aI., 1983). To use succinate as a carbon source, yeast cells must have a functional Kreb cycle in mitochondria (Tolbert, 1981a). I tested the growth rate of yeast cells on several minimal media containing different carbon sources, such as methanol, ethanol, acetate and succinate. After five to seven days, the LL20 pas5 deletion strain showed an inhibition of growth (as judged by colony size) relative to the wild type strain when grown on methanol (NM), ethanol (NE), acetate (YNA), or succinate (NS) plates (Table 4.1). The most distinguishable phenotype of the pas5 deletion mutant cells was on YNA plates. Thus, I chose the YNA medium to test yeast transformed with the human PAF-J gene. Table 4.1 Growth of different strains on several media ------D273-10B DSIO AI2-WT CBll p+ KL14-4A W303 p+ LL20 LL20 SISO-2B SISO SISO ~XAH ~XAH ~XAL YNO 100' noneb nO!le 300 none none none none 300 300 300 Sdays big< big big big big NMd N/A" N/A N/A N/A N/A N/A 300 300 N/A N/A N/A 6days small tiny NEd N/A 400 300 N/A 300 400 300 300 300 300 300 10days big small tiny big small tiny big big big NSd N/A N/A N/A N/A N/A N/A 300 300 N/A N/A N/A 13 days small tiny YNAd N/A N/A N/A N/A N/A N/A 300 300 N/A N/A N/A 7days small tiny a) 300-400 cells were plated. The number indicates how many colonies arose. b) None means no colonies were visible by eye. c) Big> Small> Tiny indicate the sizes of colonies on the day the plates were scored. Sizes should only be used to compare strains on one medium (Rows) and not used to compare media (Columns). d) NM stands for methanol minimal medium; NE is ethanol minimal medium; NS is succinate minimal medium; YNA is acetate minimal medium. e) N/A = not determined \D \D 100 Screening for Complementation of the Yeast pasS Deletion peR and molecular cloning are powerful tools for the amplification of gene sequences from cDNA libraries. However, one limitation of the method is the fidelity of the final products. The Taq DNA polymerase produces a mutation about once per 10,000 base pairs per cycle (Lundberg et aI., 1991). The peR reactions may have generated random point mutants during amplification of the human PAF-1 cDNA products, and these mutations in human P AF-1 cDNA may cause failure to rescue the slow-growth phenotype of yeast pas5 deletion mutant cells. There are two methods to overcome this problem. One is to directly sequence the peR products to find a human PAF-1 cDNA with accurate sequence and then test the growth phenotype of the pas5 deletion strain bearing the plasmid. I saved three purified pAAPs from three single E. coli colonies which can be sequenced in the future. The other method is, instead of direct sequencing of human cDNA peR products, to use the pool of human PAF-1 cDNA peR products in a screen for complementation of the pas5 deletion (Figure 4.2). If I can get a positive clone that can rescue the slow-growth phenotype of the yeast pas5 mutant cells, then, I can purify it from yeast cells Figure 4.2 Screening protocol for complementation of the yeast pas5 deletion mutant by using the pool ofhwnan cDNA peR products. Human PAF-1 Human liver peR cloning eDNA peR cDNA library products PAAB85"\ Transgenic yeast LU20aPAS L pAAP LU20dPAS/pAAP pool Complementation test on YNA plates Purify plasmid Sequncing Positive from single colonies human cDNA clones fragments f-'o f-' 102 and sequence it. TIns peR clone may either contain an accurate human PAF-l gene sequence or have a gain of ftmction point mutation in it. I pooled the pAAP plasmids together from all of the E. coli transformants and got a "pAAP constructs pool" wmch contained tens of thousands of human PAF-l peR cDNA clones instead of a single peR clone. Complementation Test I introduced the human PAF-l expressing plasmid (PAAP) pool into the LU20&AS strain, wmch contains a pas5 gene deletion. The transformants were first selected on the medium lacking uracil to confirm that pAAP was retained in the cells. In order to control the expression of the human PAF-l gene under the GAL promoter, transformants were pre-grown overnight in several media containing different carbon sources, such as glucose, galactose and raffinose. The growth rates of the pre-grown transformants were measured (as judged by colony size) on YNA plates. However, the yeast pas5 deletion mutant cells bearing the human PAF-l expressing plasmid still formed smaller colonies compared to wild type cells., 103 Slow-growth on Acetate Medium Caused by Polar Effect on CBPl Expression The inability or slow growth of yeast cells on acetate medium is diagnostic of a defects in either the glyoxylate cycle or the tricarboxylic acid (TCA) cycle (Zwart et al., 1983; Liao et aI., 1993; McAlister-Henn et aI., 1987; Kim et aI., 1986). The enzymes of the glyoxylate cycle are residents of peroxisomes, but the enzymes of the TCA cycle are mitochondrial matrix proteins. I placed the yeast pas5 deletion mutant cells bearing the centromere based yeast plasmid with the wild type PAS5 gene on the YNA plates. After five days at 30°C, the colonies varied in size and most of them were smaller compared to wild type cells. In a recent study by Lorraine Marnell, she mutated the start codon AUGs for the long and short PasSp in the same strain. She found that the AUG- mutant strains had the same growth defect on liquid oleate medium as the pasS deletion, but the strain showed no difference in growth compared to wild type on YNA plates. I wondered whether the slow growth on acetate phenotype of the pas5 deletion strain could be due to a polar effect on the expression of CBP 1, because the deletion ends 230 base pairs upstream of CBP 1 and mitochondrial function 104 is necessary for utilization of acetate (Zwart et aI., 1983; Liao et aI., 1993; McAlister-Henn et aI., 1987; Kim et aI., 1986). To test tlus hypothesis, I transfonned the pas5 deletion strain with a plasmid bearing only the CBP 1 gene and found that it rescued the slow growth of the pas5 deletion strain on acetate. Therefore, another growth phenotype must be found for the screen to be repeated. A phenotype discernable on plates of the pas5 mutant cells that is due only to a peroxisomal defect will be a good one. No Complementation of Liquid Oleate Growth Defect Since the centromere-based yeast plasmid with wild type yeast PAS5 gene restored the slow-growth phenotype of the pas5 null mutants in oleate liquid medium, I picked up several single colonies of the transgenic yeast with the three purified pAAP plasmids, pre-grew them on YFD overnigllt and measured the growth rates in oleate liquid culture at 30°C. The human cDNA plasmid did not restore the slow growth rate of the pas5 deletion mutant cells in oleate liquid medium. Recently, I have learned that Kunau's group also failed to show the functional substitution of yeast PAS5 by human PAF-l (personal communication). 105 DISCUSSION The result that the human PAF- J cDNA failed to complement th~ growth deficiency on oleate of the yeastpas5 deletion mutant could be caused by three problems. First, the human PAF- J cDNA only encodes one protein, but two yeast PAS5 transcripts might be translated into two different proteins. One human protein may not be able to complement the two knocked out in the yeast pas5 deletion mutant strain. Second, the heterologously expressed human PAF- J protein may be unstable inside the yeast cells. Third, the human homolog may not fit in the protein complex that involves Pas5p, because only 50% of the protein sequence is similar between human Paf-Ip and yeast Pas5p. I need to make several changes to this cross-species replacement experiment. First, I should use pas5 mutant host cells that have a point mutation in the first start codon ATG to knock out only the longer transcript, which may encode a protein that has a peroxisomal function, and no change in the second ATG to keep the shorter transcript intact, which may encode a protein that has mitochondrial function. This strain has been made by Lorraine Marnell and it shows slow-growth on liquid oleate medium, but is like wild type for growth on 106 acetate plates. Second, in order to examine these transgenic yeast strains for fimctional complementation, I still need to find an easily scorable phenotype for the yeast pas5 mutant cells, such as lethality instead of a slow growth phenotype. Because the heterologously expressed human Paf-1 p may only show partial recovery for the yeast pas5 mutant cells, I could verify this by a viability test. Partial fimctional complementation of a yeast mutant by a human homolog has been reported in a number of cases. The yeast rpb 7 deletion mutant cells bearing human homolog hsRPB7 showed temperature sensitivity and lost viability that was significantly different from the wild type yeast cells (Khazak et al., 1995). The yeast strains that relied solely on expression of human RHOA in place of the yeast RHOl were able to grow at 23°C but grew neither at 37°C nor in the presence of high concentration Ca 2+, even at 23 ° C (Qadota et aI., 1994). Similarly, there were 60% fewer detectable actin cables in yeast mutant cells bearing nematode homo logs when compared to wild type strains (Waddle et aI., 1993). Therefore, restored viability of yeast mutants by transformation with the homolog from another organism is a sufficient proof for cross-species complementation, since partial restoration of the lost function caused by novel gene mutations is acceptable. Third, in order to make sure the human protein is 107 stable in yeast cells, I need to verify the expression of the human PAF-J protein in transgenic yeast cells by immunoblotting analysis. Fourth, to minimize possible problems in the control of expression of the human PAF-J gene in yeast cells, the GAL promoter may be changed to the yeast PAS5 promoter. Galactose is not a good carbon source to put in the medium for the complementation test. A previous study showed that the pas5 deletion mutant cells did not have an apparent phenotype on galactose media and also the overexpression of the PAS5 gene had a negative effect on cell growth (Liu, 1991). There was also a report that peroxisomal enzymes are reduced in their amount or become inactive after shifting from oleic acid medium to galactose medium (Evers et aI., 1991). Fifth, if there is not any positive indication for the cross-species functional replacement after all of these changes, use of a reciprocal human/yeast chimera, encoding the NH2-tenninal of the protein of one organism and the COOH-terminal half of the other, may be an alternative approach. Several functional complementations of a yeast mutant by a human/yeast chimera have been reported. Mutations in yeast genes encoding the RNA-binding protein Snp 1p, the nucleotide exchange factor Cdc25p, the transcription factor Swi2p, and ATP-binding cassette transporter Ste6p have all been complemented by chimeric constructs consisting of fused 108 regions of the wild type yeast gene and its putative mammalian homolog (Smith et al., 1991; Wei et al., 1992; Khavari et aI., 1993; Teem et aI., 1993). 109 CHAPTER FIVE GENERAL CONCLUSIONS AND FUTURE DIRECTIONS I have identified PAS5 as a yeast homolog of the mammalian PAF-J gene and examined its role in peroxisome assembly. Given the sequence similarity between Pas5p and the human peroxisomal assembly factor, the induction of PAS5 expression by the peroxisomal inducer oleate, and the observation that the pas5 null mutation causes defects in import of peroxisomal proteins, I conclude that PAS5 is an important peroxisomal assembly factor in yeast. Furthermore, two mRNAs are produced from the PAS5 gene. These two transcripts are independently regulated by separate control systems and may produce two different polypeptides. Thus, PAS5 may be a multiple-function gene encoding two different proteins. Sequence Analysis of the 5'-upstream Region of the PASS Gene Inspection of the 51-upstream region (5 1-UTR) of the PAS5 gene revealed several interesting features. First, a TATA box for the longer mRNA was found in the upstream region and a TATA-like site for the shorter mRNA was found 110 inside the open reading frame. Second, both oleate and glycerol response elements were found in the upstream region of the PAS5 gene. The presumptive T ATA box for the longer transcript of PAS5, TAT AA, starts at position -88 base pair from the start codon (label as +1). The most prevalent induced transcription start site for the shorter transcript is at position +73 as shown by SI mapping data (Liu, Gu, & Dieckmann, in press). A canonical TATA box is not apparent for these downstream transcriptional initiation sites, but as TATA binding protein binds to the minor groove of AT rich sequences (Starr et aI., 1991), perhaps the ATAAA starting at +38 serves as a TATA-box. Another yeast gene induced by growth on fatty acids is FOX3, which encodes thiolase, the fourth enzyme in the ~-oxidation pathway. A partially palindromic oleic acid response element (ORE) in the upstream region of FOX3 has been defined by deletion analyses and gel retardation as 5' CGGGGATAATAGTATTAACACCG 3' (Einerhand et aI., 1991,1993). Using the Wisconsin GCG package program "Findpattems", a partial match to the FOX3 half-element was found at position -423 of PAS5 with 7 matching base pairs out of 11,5' ATTAACA 3'. The possibility of this sequence or others for 111 the increase in the 0.9 kb PAS5 transcript levels will require further investigation. The dependence of increased message levels with oleic acid medium on sequen~es in the 5' UTR is supported by the result that a fusion of the upstream region of PAS5 to lacZ increases production of J3-galactosidase when the host cells are switched from glucose to fatty acid medium (Bell and Dieckmann, unpublished results). Two different types of response elements and binding proteins have been shown to be responsible for up-regulating the transcription of several genes that code for mitochondrial proteins (Forsburg et aI., 1989; Schneider et aI., 1991). Hap 1p binds to elements that have the consensus sequence 5' WWNNWCCGWWWWNWCCG 3' (where W is a purine and N is any base) and induces transcription in response to increased oxygen and heme concentration. The Hap2p/3p/4p complex binds to the consensus 5' TNATTGGT 3' and induces transcription in a response to switching from glucose medium which represses respiration to non-repressing carbon sources such as glycerol. Again using the Wisconsin GCG package program "Findpattems", two sequences in the PAS5 5' UTR that fit well to a Hap 1p binding site were found; both have four mismatches, but only one mismatch in the critical CCG elements. 112 One sequence starts at -240 and the other starts at -45. There are ten sites that partially match the Hap2p complex binding site; they all contain two mismatches. As for the elements that govern the response to fatty acids, the cis elements responsible for the up-regulation of the 0.8 kb mRNA are in the 5'UTR; expression of the PAS5-1acZ fusion increases when cells are switched from glucose to glycerol medium (Bell and DieckmruID, unpublished results). Further analysis of the 5'UTR of PAS5 will explore the information of the cis-acting regulation elements and trans-acting activators or repressors and provide more knowledge about why one gene can be transcriptionally regulated independently by two different systems. Transcriptional Regulation by Different Carbon Sources Northern blot experiments described in Chapter three demonstrated that the two independent transcripts of PAS5 are regulated by different carbon sources. A full-length 0.9 kb transcript is induced five-fold when cells are switched from glucose medium to media containing fatty acids as sole carbon source; peroxisomal function is required for utilization of fatty acids as a carbon source. In contrast, the 0.8 kb transcript is initiated downstream of the putative 113 start codon and is not induced by fatty acids. However, it increased 10-fold by a switch to glycerol medium; mitochondrial function is required for utilization of glycerol as a carbon source. The transcription pattern of the PAS5 gene suggests that the gene may encode two polypeptides that serve functions related to peroxisomes and mitochondria respectively. Further analysis of the localization of these protein products of the PAS5 gene will reveal the reason for the dual regulation systems. Localization of Pas5p Pas5p is a protein of very low abundance. It is not detectable on stained gels or Western blots ofNycodenz gradient subcellular fractions obtained from wild-type cells grown in peroxisome induction medium. However, Pas5p was seen on a stained gel when it was overexpressed in E. coli (Liu, 1991 ). According to the northern blot result, even at the highest level observed, PAS5 transcripts are approximately 50-fold lower in abundance than actin mRNA. Consistent with the conclusion that Pas5p is oflow abundance, rare codon usage is found for the PAS5 gene. Thus, the intracellular localization of Pas5p has not yet been determined by conventional methods. 114 To study the intracellular localization of Pas5p, extensions can be added to the protein that will be recognized immlll1ologically. The fusion proteins could be observed in subcellular fractions ofNycodenz gradients by Western blots or in whole cells by electron microscopic analysis, using gold-labeled antibodies which recognize the fusion epitopes. Several other yeast peroxisomal proteins have been localized by using epitope tags. In the study of location of Pas7p in yeast cells, a fusion protein was constructed in which the amino-terminus of Pas7p was tagged with a human c-Myc epitope. The epitope-tagged Pas7p was found to be associated with peroxisomes by immlll1oblotting (Marzioch et aI., 1994). Epitope tagging was also used to study the subcellar location ofPeb1p. The influenza virus hemagglutinin (HA)-tagged PEBl gene product was fOlll1d to be associated with peroxisomes by immlll1ofluorescence (Zhang et aI., 1995). Furthermore, the H. Polymorpha Pas8p and S. cerevisiae Pmp27p were localized on peroxisomal membranes by electron microscopy (Tan et aI., 1995; Erdmann et aI., 1995). Overexpression of PASS and Peroxisome Proliferation Peroxisome proliferation can be induced on medium containing oleic acid 115 (Evers et aI., 1991). M. Veenhuis has shown that the first signs of peroxisome development were observed 2 hours after the transfer of cells from glucose medium into oleic acid medium (Veenhuis et aI., 1987). The organelles gradmllly increased in size and number. After 12 hours many mature peroxisomes were found inside cells. The distinct increase in catalase activity also happened at that time (Veenhuis et aI., 1987; Thieringer et aI., 1991; Evers et aI., 1991). In Chapter three, I showed the induction of PAS5 mRNA as soon as 4 hours after incubation in oleate medium. Prolongation of incubation to 12 hours led to a distinct increase of catalase A mRNA. The induction of PAS5 before the induction of catalase A is suggestive that Pas5p is involved in the proliferative response of peroxisomes to environmental changes. Three stages of peroxisome proliferation in yeast have been reported (Veenhuis et aI., 1992). There is an early stage in which several peroxisomes develop from a preexisting organelle after switching from glucose medium to peroxisome induction medimn. The second stage is the division of these peroxisomes into more small organelles. Finally, in the enlargement phase, the volume per peroxisome increases dramatically. Since the overexpression of peroxisomal matrix enzymes only caused an increase in organelle size but not number (Distel et aI., 1988; 116 Roggenkamp et aI., 1989; Godecke et aI., 1989), it appears that peroxisome size may be a reflection of the amount of matrix proteins, while peroxisome proliferation is controlled by a separate mechani~m. H. polymorpha peroxisomal membrane protein Per8p shows strong similarity to Pas5p (22% similar and 49% identity over the entire length)(Tan et aI., 1995). Per8p also has a carboxyl-terminal C3HC4 motif. Like pas5 mutant cells, peroxisomal proteins are located in the cytosol instead of in organelles in per8 mutant cells. These results are suggestive that Per8p may be a structural and functional homolog of Pas5p. Overexpression of PER8 when cells were grown in methanol medium resulted in markedly increased peroxisome numbers. Thus, Per8p was proposed to be an important component that controls the proliferation of this organelle in H. polymorpha. In a previous study, overexpression of PAS5 when cells were grown in glucose or galactose medium resulted poor growth (Liu, 1991). To investigate whether Pas5p is a component that controls peroxisome proliferation in S. cerevisiae, I may try to overexpress PAS5 in cells growing in oleic acid medium and check the number of peroxisomes by electron microscopic analysis or immunofluorescence. 117 Other Proteins That May Interact with Pas5p The identification of proteins that interact with Pas5p will provide further insight into the molecular mechanism that regulates peroxisomal assembly and proliferation. The zinc-finger-like motif in Pas5p has also been found in Pas4p, a protein needed for peroxisome formation (Erdmann et aI., 1992). This motif was suggested to involve protein-protein rather than DNA-protein interactions in these non-nuclear proteins. Expressing Pas4p from a multicopy vector led to massive increases in the number of organelles (Erdmann et al., 1992). This result suggested that this protein is required for peroxisome proliferation. Overexpression of another S. cerevisiae peroxisomal membrane protein, Pmp27p, also promotes peroxisomal proliferation (Marshall et al., 1995). Similar to PAS5, Pmp27 mRNA was induced by oleate 3 hours after switching from glucose to oleic acid medium. It will be interesting to probe for functional and physical interactions between Pas5p, Pas4p and Pmp27p. Peroxisomal Membrane Structure inpas5 Deletion Mutant Cells Zellweger cells have a few number of abnormal peroxisomes. They are larger in size and contain at least the enzyme thiolase. Tllis peroxisomal 118 membrane structure has been referred to as a "ghost" (Santos et aI., 1988; Wiemer et aI., 1989; Balfe et aI., 1990; Gartner et aI., 1991). Deletion of the PAS5 gene in yeast yielded cells with normal growth on glucose, but with impaired growth on oleate. This phenotype has also been observed inpmp27 deletion mutant cells (Erdmann et aI., 1995). When grown under noninducing conditions, wild type and pmp27 deletion mutant cells did not differ with respect to size and number of peroxisomes, suggesting that Pmp27p is not required for the division and inheritance of peroxisomes under noninducing conditions. During oleic acid induction, a progressive decrease in the number of peroxisomes per mutant cells was observed. However, a few giant peroxisomes were detected in the mutant cells. Although the peroxisomal membrane structures have not yet been investigated by microscopy in the pas5 mutant, I would like to postulate that they do exist and are an additional feature of the pas5 phenotype. This conclusion is primarily based on the following observations. Transformation of pas5 mutants with plasmids bearing the PAS5 gene resulted in recovering the growth rate on oleate to the wild type rate. Accumulated evidence has strongly suggested that peroxisomes do not form by a de novo assembly mechanism, but 119 arise from preexisting organelles by growth and division (Lazarow et aI., 1985). Consistent with this model, there must be some peroxisome membrane structures in pas5 deletion mutant cells in order to regenerate normal peroxisomes. Pas5p and Peroxisomal Protein Import Pathways Many mutant strains of the yeast Saccharomyces cerevisiae have been collected that display defective import of proteins into peroxisomes (Erdmann et aI., 1989; Van Der Leij et aI., 1992; Elgersma et aI., 1993; Zhang et aI., 1993a). The pas mutants describe as many as 18 different complementation groups (Elgersma et al., 1993). For example, the pas 7 mutant cells are unable to import thiolase, which has the PTS2 sequence (Marzioch et aI., 1994; Zhang et aI., 1995); the pas]O mutant cells are defective in the transport of proteins containing PTSI into peroxisomes (van Der Leij et aI., 1992,1993); the pas] mutant cells mislocalize all peroxisomal proteins to the cytosol. Unlike Pmp27p, Pas5p may not only have a role in peroxisome proliferation, but may be also required for the import of peroxisomal matrix proteins. The pas5 mutant is most like a pas] 0 mutant in that they both can import thiolase but not catalase A. However, the PAS]O sequence is different from that of PAS5 and has been 120 suggested to encode a protein that acts as a PTS 1 receptor in the PTS 1 import pathway (Subramani, 1993; McCollum et aI., 1993). Pas5p and Pasl0p could cooperate in the import pathway that is specific for catalase A and other proteins. Catalase A from S. cerevisiae is the best-characterized peroxisomal protein and has been reported to contain a C-tenninal PTS (SSNSKF) bearing some resemblance to PTS1, as well as another internal PTS3 (Kragler et aI., 1993). Yeast peroxisomal malate dehydrogenase requires the carboxyl-tenninal -SKL signal (pTS 1) for import into peroxisomes (Steffan et aI., 1992, and L. McAlister-Henn, personal communication), but it is still associated with peroxisomes in the pas5 deletion mutant, as shown in Chapter three. This result suggests that there may be an -SKL import pathway in yeast peroxisomes that is ftmctionally different from the pathway for catalase A. This hypothesis can be tested further in the yeast pas5 deletion strain by analysis of other -SKL proteins, such as the bifunctional enzyme required for the second and third enzymatic steps in ~-oxidation (Hiltunen et al., 1992), and the peroxisomal citrate synthase required for the glyoxylate cycle (Lewin et aI., 1990). Since catalase A and acyl-CoA oxidase are not imported into peroxisomes 121 III pas5 mutant cells, and both catalase A of Saccharomyces cerevisiae and acyl-CoA oxidase of Candida tropicalis have been shown to have peroxisomal targeting signals that are located in the middle part of the coding sequences (Kragler et aI., 1993; Small et aI., 1988), it is suggestive that Pas5p may be important for the PTS3 import pathway. It has been reported that the two independent PTSs in catalase A are sufficient to direct reporter proteins into peroxisomes (Kragler et aI., 1993). Deletion of the PAS5 gene from the genome cause catalase A import defect, even which has both PTS 1 and PTS3. However, the PTS 1 import pathway used by peroxisomal malate dehydrogenase is intact in pas5 mutant cells. This leads to the question of whether blocking one of the catalase import signals will affect the other one. I may find out this by dissecting the two PTSs in catalase A. I can use the fusion proteins consisting of the cytochrome c oxidase, which is a mitochondrial protein, and either one of the catalase A PTSs, and then test which PTS of catalase A can lead the reporter protein into peroxisomes in pas5 mutant cells. If the PTS 1 of catalase A works but PTS3 does not, it means blocking one PTS will affect the other one. If the PTS3 of catalase A works but PTS 1 does not, it is suggestive that catalase A uses a different PTS 1 import pathway from the one used by peroxisomal malate 122 dehydrogenase. If both PTSs of catalase A do not work, it is sugestive that catalase A import requires Pas5p. The relationship between peroxisomal proliferation and peroxisomal protein import is poorly understood. This PAS5 gene is therefore an ideal target for the elucidation of the molecular mechanism of the control of peroxisomal proliferation and peroxisomal protein import in S. cerevisiae. 123 APPENDIX AN STS SEARCH ALGORITHM TO FIND OVERLAPPING FRAGMENTS AND FORM CONTIGS FOR PHYSICAL MAPPING ABSTRACT The application of computer techniques for physical mapping of chromosomes has proven to be important for genome analysis. An improved STS search algorithm for computing STS physical maps is described. This algorithm provides a more efficient map analysis by reducing the contribution of false positive and false negative matches, and should provide accurate fragments with less data than conventional STS match algorithms. INTRODUCTION Sequence-tagged site (STS) matching algorithms find overlaps between any two cloned fragments that were prepared by cutting genomic DNA with restriction endonuclases and inserted into yeast (YAC), PI (PAC), or bacterial 124 artificial chromosomes (BAC) (Burke et aI., 1987; Pierce et aI., 1992; Shizuma et al., 1992). These cloned fragments are analyzed by measuring STS markers. The various fragments are then compared to find potential overlaps (Alizadeh et al., 1991; Chumakov et al., 1992; Foote et aI., 1992). However, the STS match algorithms have two disadvantages. 1) If the number of these fragments are very large, the computing time will become very long. 2) Iftwo fragments share one STS marker, false negative and false positive STS markers will produce incorrect matches and, as a result, predict incorrect overlaps. ALGORITHM AND DISCUSSION Here, I present a more efficient STS search algorithm that only needs three steps to bring two overlapping fragments together (Figure 1). Fn (n=ij,k,l...) stands for the fragment nand Sm (m=ij,k,l...) for the STS marker m. Fn(Sm) is defined as the presence of Sm on Fn. The algorithm starts with Fi(Si) and searches for the other marked fragment, Fj(Si), among the fragment 125 data pool. After finding Fj(Si), the algorithm searches for a second marker, Sj, on the same Fj fragment. Fi is then searched for Fi(Sj). If all three searching steps give positive answers, Fi and Fj are placed in the same contig group. In more general terms, if Fi(Si) is found, then Fi(Si)=I, and if not found, Fi(Si)=O. If the three search steps, Fi(Sm) x Fj(Sm)=I, then Fi and Fj are placed in the same contig group. Fi __.Si. _____Sj __ I /\ 11 I I 2 13 v------» I Fj __Si Sj ____Sk FIG 1. Three steps to find overlapping fragments. The reasoning behind the previously used STS matching algorithms was as follows. The STS marker is a specific DNA fragment amplified by two 20 base pair polymerase chain reaction (PCR) primers. If a particular fragment is produced, then the two PCR primers sites must be present. If there is an Si on two different fragments, the chance that these two fragments overlap is very high. The probability (P) that they are not overlapping fragments is very low, (1/4~ x (1/4~ = 6.8*10-49. Since the insertion fragment size of a YAC or BAC 126 is around 100,000-500,000 bp, then the probability that a pair of the same STS sites happen by chance is less than 3 x 10-43. Therefore, it is almost impossible that Si occurs by chance on two different fragments. However, these previously used STS matching algorithms may incorrectly predict overlaps if an STS marker is incorrectly predicted from a fragment. The new STS search algorithm that I present here solves this problem. Since two STS markers (Si, Sj) are used on two different fragments, the accuracy offinding fragment overlaps is much higher. It does not simply add one more STS marker to the search, but the search for the second STS is limited to the two fragments that share the first STS. The ordering of this search will reduce the computing time compared to matching two STS markers on two different fragments. Also, the only situation in which this algorithm can be wrong is that there are two false positive STS markers in the two fragments. The false positive STS markers are either generated by misreading of a blot, which can be eliminated by repeating the experiment, or generated by amplification of a different DNA fragment flanked by similar primary sites. Many of the latter kind of false positives can be eliminated by gel electrophoresis of the products, since tlle lengths should be different. However, this procedure is time consuming. The 127 STS search algorithm complements this experimental step, because a simple false positive signal would not be confinned by a second STS site search (Figure 2). If we start with Si which is present on both fragments, then another STS marker Sk needs to be found to confirm an alignment. The only chance of making false overlap is if both false positives occured on these two fragments. Ifthe frequency of false positives is 1%, the frequency of false alignment will be reduced to 0.1 % x 0.1 %=0.01 % by using this algorithm. A false negative STS marker is a particular STS marker that is not present on a particular fragment due to deletion of that region during cloning, or to peR errors. The frequency of false negatives is much higher than that of false positives. It can be as high as 40%. However, false negatives do not affect the STS search algorithm. False Positive Fi___ Si ____(Sj) _____ /\ 1 Yes 1 1 No 1 Yes 1 ,,------:>1 Fj _____(Si) Sk. ___ False Positive FIG2. Simple false positive result will not make overlap. 128 The advantages of this STS search algorithm are: 1) to reduce computing time compared to the previously used STS match algorithms; 2) to reduce the chance of false positive STS mark~rs producing false overlaps; 3) to eliminate incorrect alignments caused by false negative STS markers; 4) to give feedback for additional fragment analysis in the laboratory. Examples of feedback are: (a) false positive data can be retested by designing new primers to amplify a region adjacent to the STS site (Figure 2); (b) cloned fragments, such as FI, have just one STS marker (Figure 3) can be identified by the STS search algorithm, and then analyzed for presence of other matches. Fi. ____Si. ____Sj ___.Sk~ __ FJ· Si S· ------~- Fk S· Sk -~--- "---- FI_(Si)___ Sj ___(Sk) ___ need retest need retest FIG3. Contig Fi, Fj and Fk give feedback to Fl. Another situation in which this algorithm can be wrong and is not caused by experimental errors, is that there are two repeated regions in two non- overlapping fragments and Si, Sj are located in both of these regions. Obviously, 129 this kind of chance is very small. We can design peR primers to avoid repeat regions. Ifwe use the STS markers which are located in repeat regions, we can search for an additional STS marker and overcome this problem. If we use three STS markers as input to the same algorithm, the number of computing steps increases from three to five (Figure 4) and computing time almost does not change. However, the chance of obtaining false overlaps, P, is reduced another 2 4 °= 1.1 *10 12 times. Especially, there are lots of repeat regions that we do not know yet in the human genome. STS markers easily fall in these regions. Fi___ Si Sj Sk. ____ 1 "'------>1 1 1 1 4 15 1 3 1 1 ,,------> 1 " Fj ______Si 2 Sj Sk___ _ FIG4. Use of three STS markers to find overlapping fragments. CONCLUSION In summary, physical mapping is widely used in genome analysis. However, the limitation is that in order to produce more accurate detailed 130 genomic maps, more and more STS markers are needed to find overlapping fragments and to make large contigs. Each new STS marker needs to be tested against all of the collected cloned fragments. This procedure will make more laboratory work as the collection pool becomes bigger. The use of more than one STS marker to do the comparison by the existing STS match algorithms means that more computing steps are required and the computing time will be increased enormously. The use of more STS markers also means that more false positives will cause more problems for these STS match algorithms. The contigs would include more false overlaps. My new STS search algorithm is a significant improvement over the existing STS match algorithms by requiring fewer calculation steps and by discarding false negatives. More importantly, using this STS search algorithm, only two STS markers are needed on each fragment to place it in a contig group and these contigs have very high quality. Reducing the required number of STS markers means fewer experiments and increases the efficiency of using the available data. 131 LITERATURES CITED IN APPENDIX Alizadeh,F., Karp,R.M., Newberg,L.A., & Weisser,D.K. (1991). Physical mapping of chromosomes: A combinatorial problem in molecular biology, Proc. 4th ACM-SIAM Symoosium on Discrete Algorithms. Burke,D.T., Carle,G.F., & Olson,M.V. (1987). Cloning oflarge segments of exogenous DNA into yeast by means of artificial chromosome vectors, Science 236, 806-812. Chumakov,!., Rigault,P., Guilou,S., Ougen,P., Billaut,A., Guasconi,G., Gervy,P., LeGall,!., Soularue,P., Grinas,L., Bougueleret,L., Bellanne Chantelot,C., Lacoix,B., Barilot,E., Gesnouin,P., Pook,S., Vaysseix,G., Frelat,G., Schitz,A., Sambucy,J., Bosch,A., Estivill,X., Weissenbach,J., Vignal,A., Reithman,H., Cox,D., Patterson,D., Gardiner,K., Hattori,M., Sakaki,Y., Ichikawa,H., Ohki,M., Paslier,D.L., Heilig,R., Antonarakis,S., & Cohen,D. (1992). Continuum of overlapping clones spanning the entire human chromosome 21q, Nature, 359, 380-386. Foote,S., Vollrath,D., Hilton,A., & Page,D.C. (1992). The human Y chromosome overlapping DNA clones spanning the euchromatic region, Science 258,60-66. Pierce,lC., Sauer,B., & Sternberg,N. (1992). A positive selection vector for cloning high molecular weight DNA by the bacteriophage PI system: improved cloning efficacy, Proc. Natl. Acad. Sci. USA, 89,2056-2060. Shizuma,H., Birren,B., Kim,UJ., Mancino,V., Slepak,T., Tachiiri,Y., and Simon,M. (1992). Cloning and stable maintenance of300-kilobase-pair fragments of human DNA in Escherichia coli using an F-factor-based vector, Proc. Natl. Acad. Sci. USA, 89, 8794-8797. 132 REFERENCES Adams, A. E. M., Shen, W., Lin, C., Leavitt, J., & Matsudaira, P. (1995). Isofonn-specific complementation of the yeast sac6 null mutation by human fimbrin. Mol Cell BioI. 15,69-75. Aitchison, J. D., lVlurray, W. W., & Rachubinski, R. A. (1991). The carboxyl-tenninal tripeptide Ala-Lys-Ile is essential for targeting Candida tropicalis trifunctional enzyme to yeast peroxisomes. J BioI Chern, 266(34), 23197-23203. Alexson, S. E. H., Fujiki, Y., Shio, H., & Lazarow, P. B. (1985). Partial disassembly ofperoxisomes. J Cell BioI. 101,294-305. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., & Lipman, D. J. (1990). Basic local alignment search tool. J Mol BioI. 215,403-410. Alvares, K., Widrow, R. J., Abu-Jawdeh, G. M., Schmidt, J. V., Yeldandi, A. V., Rao, S. R., & Reddy, J. K. (1992). Rat urate oxidase produced by recombinant baculovirus expression: formation of peroxisome crystalloid core-like structures. Proc Natl Acad Sci USA, 89,4908-4912. Balfe, A., Hoefler, G., Chen, W. W., & Watkins, P. A. (1990). Aberrant subcellular localization of peroxisomal 3-ketoacyl-CoA thiolase in the Zellweger syndrome and rhizomelic chondrodysplasia punctata. Pediatric Research, 27, 304-310. Berteaux-Lecellier, V., Picard, M., Thompson-Coffe, C., Zickler, D., Panvier-Adoutte, A., & Simonet, J.-M. (1995). A nonmammalian homolog of the PAF 1 gene (Zellweger syndrome) discovered as a gene involved in caryogamy in the fungus Podospora anserina. Cell. 81, 1043-1051. Boeke, J. D., Lacroute, F., & Fink, G. R. (1984). A positive selection for mutants lacking orotidine-5'-phosphate decarboxylase activity in yeast: 133 5-fluoro-orotic acid resistance. Mol Gen Genet. 197, 345-346. Bolivar, F., Rodriguez, R. L., Greene, P. J., Betlach, M. C., Heyneker, H. L., Boyer, H. W., Crosa, J. H., & Falkow, S. (1977).· Construction and characterization of new cloning vehicles II. A multi-purpose qloning system. Gene, 2, 95-113. Bonitz, S. G., Coruzzi, G., Thalenfeld, B. E., Tzagoloff, A., & Macino, G. (1980). Assembly of the mitochondrial membrane system. Physical map of the Oxi310cus of yeast mitochondrial DNA. J BioI Chern, 255, 11922-11926. Borst, P. (1989). Peroxisome biogenesis revisited. Biochim Biophys Acta, 1008, 1-13. Bossier, P., Fernandes, L., Vilela, C., & Rodrigues-pousada, C. (1994). The yeast YKL7 41 gene situated on the left arm of chromosome XI for a homologue of the human ALD protein. Yeast. 10,681-686. Bowen, P., Lee, C. S. N., Zellweger, H., & Lindenberg, R. (1964). A familiar syndrome of multiple congenital defects. Bulletin of Johns Hopkins Hospital, 114,402-414. Brunk, B. P., Martin, E. C., & Adler, P. N. (1991). Drosophila genes Posterior Sex Combs and Suppressor two ofzeste encode proteins with homology to the murine bmi-l oncogene. Nature, 353, 351-355. Carlson, M., & Botstein, D. (1982). Two differentially regulated mRNAs with different 51 ends encode secreted and intracellular forms of yeast invertase. Cell, 28, 145-154. Chantrenne, H. (1955). Effets dlun inhibiteur de la catalase sur la formation induite de cet enzyme chez la levure. Biochim Biophys Acta, 16, 410-417. Chen, W., & Dieckmann, C. L. (1994). Cbp1p is required for message 134 stability following 5'-processing of COB mRNA. J BioI Chern, 269(24), 16574-16578. Chiu, M. 1., Mason, T. L., & Fink, G. R. (1992). HTS1 encodes both the cytoplasmic and mitochondrial histidyl-tRNA synthetase of Saccharomyces cerevisiae: mutations alter the specificity of compartmentation. Genetics, 132, 987-1001. Cohen, G., Rapatz, W., & Ruis, H. (1988). Sequence of the Saccharomyces cerevisiae CTA] gene and amino acid sequence of catalase A derived from it. Eur J Biochem, 176, 159-163. Cregg,l M., van Klei, 1. 1, Sluter, G. 1, Veenhuis, M., & Harder, W. (1990). Peroxisome-deficient mutants of Hansenula polymorpha. Yeast. 6, 87-97. Datta, N. S., Wilson, G. N., & Hajra, A. K. (1984). Deficiency of enzymes catalyzing the biosynthesis of glycerol-ether lipids in Zellweger syndrome. A new category of metabolic disease involving the absence of peroxisomes. New England Journal of Medicine, 311, 1080-1083. de Duve, C., & Baudhuin, P. (1966). Peroxisomes (microbodies and related particles). Physiology Reviews, 46, 323-357. Deng, X., Matsui, M., Wei, N., Wagner, D., Chu, A. M., Feldmann, K. A., & Quail, P. H. (1992). COP], an Arabidopsis regulatory gene, encodes a protein with both a zinc-binding motif and a Gp homologous domain. Cell, 71, 791-801. Didion, T., & Roggenkamp, R. (1992). Targeting signal of the peroxisomal catalase in the methylotrophic yeast Hansenula polymorpha. FEBS Lett, 303(3), 113-116. Dieckmann, C. L., Homison, G., & Tzagoloff, A. (1984a). Assembly of the mitochondrial membrane system. Nucleotide sequence of a yeast nuclear 135 gene (CBP}) involved in 5' end processing of cytochrome b pre-mRNA. J BioI Chern, 259(8), 4732-4738. Dieckmann, C. L., Koerner, T. 1., & Tzagoloff, A. (1984b). Assembly of the mitochondrial membrane system. CBP}, a yeast nuclear gene involved in 5' end processing of cytochrome b pre- mRNA. J BioI Chern, 259(8), 4722-473l. Dieckmann, C. L., & Tzagoloff, A. (1985). Assembly of the mitochondrial membrane system. CBP6, a yeast nuclear gene necessary for synthesis of cytochrome b. J BioI Chern, 260, 1513-1520. Dieckmann, C. L., & Mittelmeier, T. M. (1987). Nuclearly-encoded CBP1 interacts with the 5' end of mitochondrial cytochrome b pre-mRNA. CUff Genet, 12,391-397. Distel, B., Ley, I. V. D., Veenhuis, M., & Tabak, H. F. (1988). Alcohol oxidase expressed under nonmethylotrophic conditions is imported, assembled, and enzymatically active in peroxisomes of Hansenula polymorpha. J Cell BioI, 107, 1669-1675. Dmochowska, A., Dignard, D., Maleszka, R., & Thomas, D. Y. (1990). Structure and transcriptional control of the Saccharomyces cerevisiae POX} gene encoding acyl-coenzyme A oxidase. Gene, 88,247-252. Dodt, G., Braverman, N., Wong, C., Moser, A., Moser, H. W., Watkins, P., Valle, D., & Gould, S. J. (1995). Mutations in the PTS1 receptor gene, PXR}, define complementation group 2 of the peroxisome biogenesis disorders. Nature Genetics, 9, 115-125. Douma, A. C., Veenhuis, M., de Konig, W., & Harder, W. (1985). Dihydroxyacetone synthase is localized in the peroxisoma matrix of methanol-grown Hansenula polymorpha. Arch MicrobioL 143,237-243. Einerhand, A. W. C., Voorn-Brouwer, T. M., Erdmann, R., Kunau, W., & 136 Tabak, H. F. (1991). Regulation of transcription of the gene coding for peroxisomal 3-oxoacyl-CoA thiolase of Saccharomyces cerevisiae. Eur J Biochem. 113, 122 Einerhand, A. W. C., Kos, W. T., Distel, B., & Tabak, H. F. (1993). Characterization of a transcriptional control element involved in proliferation of peroxisomes in yeast in response to oleate. Eur J Biochem. 214, 323-331. Elgersma, Y., Van Den Berg, M., Tabak, H. F., & Distel, B. (1993). An efficient positive selection procedure for the isolation of peroxisomal import and peroxisome assembly mutants of Saccharomyces cerevisiae. Genetics. 135, 731-740. Erdmann, R., Veenhuis, M., Mertens, D., & Kunau, W. (1989). Isolation of peroxisome-deficient mutants of Saccharomyces cerevisiae. Proc Natl Acad Sci USA. 86, 5419-5423. Erdmann, R., Wiebel, F. F., Flessau, A., Rytka, J., Beyer, A., Frohlich, K., & Kunau, W. (1991). PAS], a yeast gene required for peroxisome biogenesis, encodes a member ofa novel family of putative ATPases. CelL 64, 499-510. Erdmann, R., & Kunau, W. (1992). A genetic approach to the biogenesis ofperoxisomes in the yeast Saccharomyces cerevisiae. Cell Biochem Func. 10, 167-174. Erdmann, R. (1994). The peroxisomal targeting signal of3-oxoacyl-CoA thiolase from Saccharomyces cerevisiae. Yeast. 10,935-944. Erdmann, R., & Blobel, G. (1995). Giant peroxisomes in oleic acid-induced Saccharomyces cerevisiae lacking the peroxisomal membrane protein Pmp27p. J Cell BioI, 128(4),509-523. Evers, M. E., Hohfeld, J., Kunau, W., Harder, W., & Veenhuis, M. (1991). Physioloical studies on the utilization of oleic acid by Saccharomyces cerevisiae in relation to microbody development. FEMS Microbiology 137 Letters, 90, 73-78. Faber, K. N., Haima, P., Gietl, C., Harder, W., AB, G., & Veenhuis, M. (1994). The methylotrophic yeast Hansenula polymorpha contains an inducible import pathway for peroxisomal matrix proteins with an N-terminal targeting signal (pTS2 proteins). Proc Natl Acad Sci USA, 91, 12985-12989. Faugeron-Fonty, G., Culard, F., Baldacci, G., Goursot, R., Prunell, A., & Bernardi, G. (1979). The mitochondrial genome of wild-type yeast cells VIII. The spontaneous cytoplasmic "petite" mutation. J Mol BioI, 134,493-537. Forsburg, S. L., & Guarente, L. (1989). Communication between mitochondria and the nucleus in regulation of cytochrome genes in the yeast Saccharomyces cerevisiae. Annu Rev Cell BioI. 5, 153-180. Freemont, P. S., Hanson, I. M., & Trowsdale, J. (1991). A novel cysteine-rich sequence motif. Cell. 64, 483-484. Fujiki, Y., Rachubinski, R. A., & Lazarow, P. B. (1984). Synthesis ofa major intergral membrane polypeptide of rat liver peroxisome on free polysomes. Proc Natl Acad Sci USA, 81, 7127-7131. Furuta, S., Hayashi, H., Huikata, M., Miyazawa, S., Osumi, T., & Hashimoto, T. (1986). Complete nucleotide sequence of cDNA and deduced amino acid sequence of rat liver catalase. Proc Natl Acad Sci USA. 83, 313-317. Gartner, J., Chen, W. W., Kelley, R. I., Mihalik, S. J., & Moser, H. W. (1991). The 22-kD peroxisomal integral membrane protein in Zellweger Syndrome - presence, abundance, and association with a peroxisomal thiolase precursor protein. Pediatric Research, 29, 141-146. Gartner, J., Moser, H., & Valle, D. (1992). Mutations in the 70K peroxisomal membrane protein gene in Zellweger syndrome. Nature Genetics, 1, 16-23. 138 Gartner, J., Kearns, W., Rosenberg, C., Pearson, P., Copeland, N. G., Gilbert, D. J., Jenkins, N. A., & Valle, D. (1993). Localization of the 70-kDa peroxisomal membrane protein to human 1p21-p22 and mouse 3. Genomics. 1.2.,412-414. Gietl, C. (1990). Glyoxysomal malate dehydrogenase from watermelon is synthesized with an amino-terminal transit peptide. Proc Natl Acad Sci USA. 87, 5773-5777. Gietz, R. D., & Schiestl, R. H. (1991). Applications of high efficiency lithium acetate transformation of intact yeast cells using single-stranded nucleic acids as carrier. Yeast. 7, 253-263. Glick, B. S., Beasley, E. M., & Schatz, G. (1992). Protein sorting in mitochondria. Trends Biochem Sci, 17, 453-459. Glover, J. R., Andrews, D. W., Subramani, S., & Rachubinski, R. A. (1994). Mutagenesis of the amino targeting signal of Saccharomyces cerevisiae 3-ketoacyl-CoA thiolase reveals conserved amino acids required for import into peroxisomes in vivo. J BioI Chern. 269, 7558-7563. Godecke, A., Veenhuis, M., Roggenkamp, R., Janowicz, Z. A., & Hollenberg, C. P. (1989). Biosynthesis of the peroxisomal dihydroxyacetone synthase from Hansenula polymorpha in Saccharomyces cerevisiae induces growth but not proliferation ofperoxisomes. CUff Genet. 16, 13-20. Goldfischer, S., Moore, C. L., Johnson, A. B., Spiro, A. J., Valsamis, M. P., Wisniewski, H. K., Ritch, R. H., Norton, W. T., Rapin, L, & Gartner, L M. (1973). Peroxisomal and mitochondrial defects in the cerebro-hepato-renal syndrome. Science. 182,62-64. Goldfischer, S. (1979). Peroxisomes in disease. J Histoch Cytoch. 27(10), 1371-1373. Gould, S. J., Keller, G. A., & Subramani, S. (1987). Identification ofa 139 peroxisomal targeting signal at carboxy tenninus of firefly luciferase. J Cell BioI, 105,2923-2931. Gould, S. J., Keller, G. A., Hosken, N., Wilkinson, J., & Subramani, S. (1989). A conserved tripeptide sorts proteins to peroxisomes. J Cell BioI, 108, 1657-1664. Gould, S. J., Keller, G.-A., Schneider, M., Howell, S. H., Garrard, L. J., Goodman, J. M., Distel, B., Tabak, H., & Subramani, S. (1990). Peroxisomal protein import is conserved between yeast, plants, insects and mammals. EMBO J, 9, 85-90. Gould, S. J., McCollum, D., Spong, A. P., Heyman, J. A., & Subramani, S. (1992). Development of the yeast Pichia pastoris as a model organism for a genetic and molecular analysis of peroxisome assembly. Yeast. 8, 613-628. Gu, K. L., & Dieckmann, C. L. The Saccharomyces cerevisiae gene PAS5 is required for peroxisome assembly. (unpublish) Guan, X., Mural, R., Petrov, S., & Uberacher, E. C. (1993). A sensitive sequence comparison server for DNA and PROTEINS. In Anonymous, Proceedings, Genome Sequencing and Analysis Conference V. Hilton Head, South Carolina: Guarente, L. (1992). Messenger RNA transcription and its control in Saccharomyces cerevisiae. In J. Broach, E. W. Jones, & J. Pringle (Eds.), The Molecular and Cellular Biology of the Yeast Saccharomyces: Gene Expression. (pp. 49-98). Plainview, N.Y. Cold Spring Harbor Press. Hanahan, D. (1983). Studies on transformation of Escherichia coli with plasmids. J Mol BioI, 166,557-580. Hansen, H. T., Didion, T., Thiemann, A., Veenhuis, M., & Roggenkamp, R. (1992). Targeting sequences of two major peroxisomal proteins in the methylotrophic yeast Hansenula polymorpha. Mol Gen Genet. 235, 269-278. 140 Hanson, 1., Rogoussis, J., & Trowsdale, J. (1991). Organization of the human HLA-class-ll region. International Journal of Cancer, 6, 18-19. Haupt, Y., Alexander, W. S., Barri, G., Klinken, S. P., & Adams, J. M. (1991). Novel zinc finger gene implicated as myc collaborator by retrovirally accelerated lymphomagenesis in E~-myc transgenic mice. CelL 65, 753-763. Heikoop, J. C., van Roermund, C. W. T., Just, W. W., Ofinan, R., Schutgens, R. B. H., Heymans, H. S. A., Wanders, R. J. A., & Tager, J. M. (1990). Rhizomelic Chondrodysplasia Punctata: Deficiency of 3-oxoacyl-coenzymes A thiolase in peroxisomes and impaired processing of the enzyme. J Clin Invest. 86, 126-130. Heymans, H. S., Schutgens, R. B., Tan, R., van den Bosch, H., & Borst, P. (1983). Severe plasmalogen deficiency in tissues of infants without peroxisomes (Zellweger syndrome). Nature, 306, 69-70. Hijikata, M., Ishii, N., Kagamiyama, H., Osumi, T., & Hashimoto, T. (1987). Structure analysis of cDNA for rat peroxisomaI3-ketoacyl-CoA thiolase. J BioI Chern, 262(17), 8151-8158. Hiltunen, J. K., Wenzel, B., Beyer, A., Erdmann, R., Fossa, A., & Kunau, W. (1992). Peroxisomal multifunctional p-oxidation protein of Saccharomyces cerevisiae. J BioI Chern, 267,6646-6653. Hopper, A. K., & Martin, N. C. (1992). Processing of yeast cytoplasmic and mitochondrial precursor tRNAs. In E. W. Jones, J. R. Pringle, & J. R. Broach (Eds.), The Molecular and Cellular Biology of the Yeast Saccharomyces: Gene Expression. (pp. 99-141). Plainview, NY: Cold Spring Harbor Laboratory Press. Hohfeld, J., Veenhuis, M., & Kunau, W. (1991). PAS3, a Saccharomyces cerevisiae gene encoding a peroxisomal integral membrane protein essential for peroxisome biogenesis. J Cell BioL 114, 1167-1178. 141 Igual, J. C., Gonzalez-Bosch, C., Franco, L., & Perez-Ortin, J. E. (1992). The POT1 gene for yeast peroxisomal thiolase is subject to three different mechanisms of regulation. Mol Microbiol. 6, 1867-1875. Ishii, N., Hijikata, M., Osumi, T., & Hashimoto, T. (1987). Structure organization of the gene for rat enoyl-CoA hydratase: 3-hydroxyacyl-CoA dehydrogenase bifunctional enzyme. J BioI Chern. 262(17), 8144-8150. Johnston, M., & Carlson, M. (1992). Regulation of carbon and phosphate utilization. In E. W. Jones, J. R. Pringle, & J. R. Broach (Eds.), The Molecular and Cellular Biology of the Yeast Saccharomyces: Gene Expression. (pp. 193-281). Plainview, N.Y. Cold Spring Harbor Laboratory Press. Jones, J. S., Weber, S., & Prakash, L. (1988). The Saccharomyces cerevisiae RAD 18 gene encodes a protein that contains potential zinc finger domains for nucleic acid binding and a putative nucleotide binding sequence. Nucleic Acids Research. 16, 7119-7131. Khavari, P. A., Peterson, C. L., Tamkun, J. W., Mendel, D. B., & Crabtree, G. R. (1993). BRG1 contains a conserved domain of the SWI2/SNF2 family necessary for normal mitotic growth and transcription. Nature. 366, 170-174. Khazak, V., Sadhale, P. P., Woychik, N. A., Brent, R., & Golemis, E. A. (1995). Human RNA polymerase II subunit hsRPB7 functioons in yeast and influences stress survival and cell morphology. Melocular Biology of the Cell. §(7), 759-775. Kim, K., Rosenkrantz, M. S., & Guarente, L. (1986). Saccharomyces cerevisiae contains two functional citrate synthase genes. Mol Cell BioI. 6, 1936-1942. Kragler, F., Langeder, A., Raupachova, J., Binder, M., & Hartig, A. (1993). Two independent peroxisomal targeting signals in catalase A of 142 Saccharomyces cerevisiae. J Cell BioI, 120,665-673. Kunau, W., & Hartig, A. (1992). Peroxisome biogenesis in Saccharomyces cerevisiae. Antonie van Leeuwenhoek. 62, 63-78. Kunau, W., Beyer, A., Franken, T., Gotte, K., Marzioch, M., Saidowsky, J., Skaletz-Rorowski, A., & Wiebel, F. F. (1993). Two complementary approaches to study peroxisome biogenesis in Saccharomyces cerevisiae: Forward and reversed genetics. Biochimie, 75, 209-224. Kyte, J., & Doolittle, R. F. (1982). A simple method for displaying the hydropathic character ofa protein. J Mol BioI. 157, 105-132. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227, 680-685. Lazarow, P. B., & Fujiki, Y. (1985). Biogenesis ofperoxisomes. Annual Review of Cell Biology, 1,489-530. Lazarow, P. B., & Moser, H. W. (1989). Disorders of peroxisome biogenesis. In C. R. Scriver, A. L. Beaudet, W. S. Sly, & D. Valle (Eds.), The metabolic basis of inherited disease. (pp. 1479-1509). New York: McGraw-Hill. Lewin, A. S., Hines, V., & Small, G. M. (1990). Citrate synthase encoded by the CIT2 gene of Saccharomyces cerevisiae is peroxisomal. Mol Cell BioI, 10(4), 1399-1405. Liao, X., & Butow, R. A. (1993). RTGI and RTG2: two yeast genes required for a novel path of communication from mitochondria to the nucleus. Cell, 72, 61-71. Linder, P., & Slonimski, P. P. (1989). An essential yeast protein, encoded by duplicated genes TIF 1 and TIF2 and homologous to the mammalian translation initiaition factor eIF-4A, can suppress a mitochondrial missense 143 . mutation. Proc Nati Acad Sci USA, 86,2286-2290. Liu, H., Tan, X., Veenhuis, M., McCollum, D., & Cregg, J. M. (1992). An efficient screen for peroxisome-deficient mutants of Pichia pastoris. J BacterioI. 174(15), 4943-4951. Liu, Y. (1991). Ph.D. Dissertation: Characterization and functional analysis of the CRTI-CBPI-NUCI cluster in the yeast Saccharomyces cerevisiae. University of Arizona. Liu, Y., Gu, K. L., & Dieckmann, C. L. (in press). Independent regulation of full-length and 5'-truncated PAS5 mRNAs in Saccharomyces cerevisiae. Yeast. Lundberg, K. S., Shoemaker, D. D., Adams, M. W. W., Short, J. M., Sorge, J. A., & Mathur, E. J. (1991). High-fidelity amplification using a thennostable DNA polymerase isolated from Pyrococcus furiosus. Gene, 108, 1-6. Marshall, P. A., Krimkevich, Y. L, Lark, R. H., Dyer, J. M., Veenhuis, M., & Goodman, J. M. (1995). Pmp27 promotes peroxisomal proliferation. J Cell BioI. 129(2),345-355. Marzioch, M., Erdmann, R., Veenhuis, M., & Kunau, W. (1994). PAS7 encodes a novel yeast member of the WD-40 protein family essential for import of 3-oxoacyl-CoA thiolase, a PTS2-containing protein, into peroxisomes. EMBO 1, 13,4908-4918. Masuno, M., Shimozawa, N., Suzuki, Y., Kondo, N., Orii, T., Tsukamoto, T., Osumi, T., Fujiki, Y., Imaizumi, K., & Kuroki, Y. (1994). Assignment of the human peroxisome assembly factor-l gene (PXMP3) responsible for Zellweger syndrome to chromosome 8q21.1 by flurescence in situ hyridization. Genomics, 20, 141-142. Mayer, S. A., & Dieckmann, C. L. (1989). The yeast CBP 1 gene produces 144 two differentially regulated transcripts by alternative 3'-end fonnation. Mol Cell BioI. 9(10), 4161-4169. McAlister-Henn, L., & Thompson, L. M. (1987). Isolation and expression of the gene encoding yeast mitochondrial malate dehydrogenase. Journal of Bacteriology, 169(11),5157-5166. McCammon, M. T., Veenhuis, M., Trapp, S. B., & Goodman, J. M. (1990). Association of glyoxylate and beta-oxidation enzymes with peroxisomes of Saccharomyces cerevisiae. J BacterioL 172,5816-5827. McCollum, D., Monosov, E., & Subramani, S. (1993). The pas8 mutant of Pichia pastoris exhibits the peroxisorneal protein import deficiencies of Zellweger syndrome cells--The P AS8 protein binds to the COOH-terminal tripeptide peroxisomal targeting signal, and is a member of the TPR protein family. The journal of Cell Biology, 121(4), 761-774. Mittelmeier, T. M., & Dieckmann, C. L. (1990). CBPI function is required for stability of a hybrid cob-olil transcript in yeast mitochondria. CUff Genet. 18,421-428. Mittelmeier, T. M., & Dieckmann, C. L. (1993). In vivo analysis of sequences necessary for CBPI-dependent accumulation of cytochrome b transcripts in yeast mitochondria. Mol Cell BioL 13(7),4203-4213. Miyazawa, S., Hayashi, H., Hijikata, M., Ishii, N., Furuta, S., Kagamiyama, H., Osumi, T., & Hashimoto, T. (1987). Complete nucleotide sequence of cDNA and predicted amino acid sequence of rat acyl-CoA oxidase. J BioI Chern, 262, 8131-8137. Montgomery, D. L., Hall, B. D., Gillam, S., & Smith, M. (1978). Identification and isolation of the yeast cytochrome c gene. CelL 14,673-680. Moreno, M., Lark, R., Campbell, K. L., & Goodman, J. M. (1994). The peroxisomal membrane proteins of Candida boidinii: gene isolation and 145 expression. Yeast. 10, 1447-1457. Moser, A. B., Rasmussen, M., Naidu, S., Watkins, P., McGuinness, M., Hajra, A. K., Chen, G., Raymond, G., Liu, A., Gordon, D., Gamaas, K., Walton, D., Skjeldal, O. H., Guggenheim, M. A., Jack~on, L. G., Elias, E. R., & Moser, H. W. (1995). Phenotype of patients with peroxisomal disorders subdividied into sixteen complemetation groups. The Journal of Pediatrics. 127(1), 13-22. Motley, A., Hattema, E., Distel, B., & Tabak, T. (1994). Differential protein import deficiencies in human peroxisome assembly disorders. J Cell BioI. 125, 755-767. Muller-Hocker, J., Walther, J. U., Bise, K., Pongratz, D., & Hubner, G. (1984). Mitochondrial myopathy with loosely coupled oxidative phosphorylation in a case of Zellweger syndrome. Virchows Archiv B, 45, 125-138. Naritomi, K., hyakuna, N., Suzuki, Y., Orii, T., & Hirayama, K. (1988). Zellweger syndrome and a microdeletion of the proxisomallong arm of chromosome 7. Human Genetics. 80,201-202. Naritomi, K., lzumikawa, Y., Ohshiro, S., Yoshida, K., Shimozawa, N., & Suzuki, Y. (1989). gene assignment of Zellweger syndrome to 7 q 11.23: report of the second case associated with a pericentric inversion of the chromosome 7. Human Genetics, 84, 79-80. Okada, H., Ueda, M., Sugaya, T., Atomi, H., Mozaffar, S., Hishida, T., Teranishi, Y., Okazaki, K., Kamiryo, T., & Tanaka, A. (1987). Catalase gene of the yeast Candida tropicalis. Eur J Biochem, 170, 105-110. Orr-Weaver, T. L., Szostak, J. W., & Rothstein, R. J. (1981). Yeast transformation: a model system for the study of recombination. Proc Natl Acad Sci USA, 78, 6354-6358. 146 Osumi, T., Tsukamoto, T., Hata, S., Yokota, S., Miura, S., Fuyiki, Y., Hijikata, M., Miyazawa, S., & Hashimoto, T. (1991). Amino-tenninal presequence of the precursor of peroxisomal 3-ketoacyl-CoA thiolase is acavable signal peptide for peroxisomal targeting. Biochem Biophys Res Commun, 181,947-954. Parish, R. W. (1975). The isolation and charcterization ofperoxisomes (microbodies) from baker's yeast Saccharoyces cerevisiae. Arch MicrobioL 105, 187-192. Pearson, W. R., & Lipman, D. J. (1988). Improved tools for biological sequence comparison. Proc Natl Acad Sci USA, 85, 2444-2448. Purdue, P. E., Takada, Y., & Danpure, C. J. (1990). Identification of mutations associated with peroxisome-to-mitochondriion mistargeting of Alanine/Glyoxylate aminotransferase in primary hyperoxaluria type 1. J Cell BioI. 111(6),2341-2351. Qadota, H., Anraku, Y., Botstein, D., & Ohya, Y. (1994). Conditional lethality of a yeast strain expressing human RHOA in place of RHO 1. Proc Natl Acad Sci USA, 91, 9317-9321. Quan, F., Komeluk, R. G., Tropak, M. B., & Gravel, R. A. (1986). Isolation and characterization of the human catalase. Nucleic Acids Research, 14(13),5321-5335. Ripmaster, T. L., Shiba, K., & Schimmel, P. (1995). Wide cross-species amino acyl-RNA synthetase replacement in vivo: yeast cytoplasmic alanine enzyme replaced by hmnan polymyositis serum antigen. Proc Natl Acad Sci USA, 92(11), 4932-4936. Roggenkamp, R., Didion, T., & Kowallik, K. (1989). Formation of irregular giant peroxisomes by overproduction of the crystalloid core protein methanol oxidase in the methylotrophic yeast Hansenula polymorpha. Mol Cell BioI. 9(3), 988-994. 147 Santos, M. J., Ojeda, J. M., Garrido, J., & Leighton, F. (1985). Peroxisomal organization in normal and cerebrohepatorenal (Zellweger) syndrome fibroblasts. Proc Natl Acad Sci USA, 82, 6556-6560. Santos, M. J., Imanaka, T., Shio, H., Small, G. M., & Lazarow, P. B. (1988). Peroxisomal membrane ghosts in Zellweger syndrome - aberrant organelle assembly. Science, 239, 1536-1538. Santos, M. J., Moser, A. B., Drwinga, H., Moser, H. W., & Lazarow, P. B. (1993). Analysis ofperoxisomes in lymphoblasts: Zellweger syndrome and a patient with a deletion in chromosome 7. Pediatric Research, 33(5), 441-444. Sarnat, H. B., Machin, G., Darwish, H. Z., & Rubin, S. Z. (1983). Mitochondrial myopathy of cerebro-hepato-renal (Zellweger) syndrome. Canadian Journal of Neurological Science, 10, 170-177. Schatz, D. G., Oettinger, M. A., & Baltimore, D. (1989). The V(D)J recombination activating gene, RAG-i. Cell. 59, 1035-1048. Schneider, J. C., & Guarente, L. (1991). Regulation of the yeast CfTi gene encoding cytochrome c1 by HAPI and HAP2/3/4. Mol Cell BioI. 11, 4934-4942. Schram, A. W., Strijland, A., Hashimoto, T., Wanders, R. J. A., Schutgens, R. B. H., van den Bosch, H., & Tager, J. M. (1986). Biosynthesis and maturation of peroxisomal p-oxidation enzymes in fibroblasts in relation to the Zellweger syndrome and infantile Refsum disease. Proc Natl Acad Sci USA, 83,6156-6158. Sharp, P. M., & Li, W. (1987). The codon adaptation index - a measure of directional synonymous codon usage bias, and its potential application. Nucleic Acids Research, 15, 1281-1295. Shimozawa, N., Tsukamoto, T., Suzuki, Y., Orii, T., Shirayoshi~ Y., Mori, 148 T., & Fujiki, Y. (1992). A human gene responsible for Zellweger syndrome that affects peroxisome assembly. Science, 255, 1132-1134. Shimozawa, N., Suzuki, Y., Orii, T., Moser, A., Moser, H. W., & Wanders, R. J. A. (1993). Standardization of complementation grouping of peroxisome-deficient disorders and the second Zellweger patient with Persoxisomal Assembly Factor-l (PAF-l) defect. American Journal of Human Genetics, 52, 843-844. Sikorski, R. S., & Boeke, J. D. (1990). In vitro mutagenesis and plasmid shuffling: From cloned gene to mutant yeast. Methods in Enzymology, 194, 302-318. Singh, I., Moser, A. E., Goldfischer, S., & Moser, H. W. (1984). Lignoceric acid is oxidized in the peroxisome: implications for the Zellweger cerebro-hepato-renal syndrome and adrenoleukodystrophy. Proc Nat! Acad Sci USA, 81,4203-4207. Singh, K. K., Small, G. M., & Lewin, A. S. (1992). Alternative topogenic signals in peroxisomal citrate synthase of Saccharomyces cerevisiae. Mol Cell BioI. 12, 5593-5599. Skoneczny, M., Chelstowska, A., & Rytka, J. (1988). Study of the coinduction by fatty acids of catalase A and acyl CoA oxidase in standard and mutant Saccharomyces cerevisiae strains. Eur J Biochem, 174,297-302. Sloots, J. A., Aitchison, J. D., & Rachubinski, R. A. (1991). Glucose-responsive and oleic acid-responsive elements in the gene encoding the peroxisomal trifunctional enzyme of Candida tropcalis. Gene, 105, 129-134. Small, G. M., Burdett, K., & Connock, M. J. (1985). A sensitive spectrophotometric assay for peroxisomal acyl-CoA oxidase. Biochem J, 227, 205-210. 149 Small, G. M., Szabo, L. J., & Lazarow, P. B. (1988). Acyl CoA oxidase contains two targeting sequences each of which can mediate protein import into peroxisomes. EMBO J, 7(4), 1167-1173. Smith, V., & Barrell, B. G. (1991). Cloning ofa yeast Ul snRNP 70k protein homologue: functional conservation of an RNA-binding domain between humans and yeast. EMBO J, 10,2627-2634. Spong, A. P., & Subramani, S. (1993). Cloning and characterization of PAS5: A gene required for peroxisome biogenesis in the methylotrophic yeast Pichia pastoris . J Cell BioI. 123(3),535-548. Staples, R. R., & Dieckmann, C. L. (1993). Generation of temperature-sensitive cbpJ strains of Saccharomyces cerevisiae by PCR mutagenesis and in vivo recombination: Characteristics of the mutant strains imply that CBPI is involved in stabilization and processing of cytochrome b pre-mRNA. Genetics, 135,981-991. Starr, D. B., & Hawley, D. K. (1991). TFIID binds in the minor groove of the TATA box. Cell, 67, 1231-1240. Steffan, J. S., & McAlister-Henn, L. (1992). Isolation and characterization of the yeast gene encoding the MDH3 isozyme of malate dehydrogenease. J BioI Chern, 267(34), 24708-24715. Stevens, B. (1981). Mitochondrial structure. In J. N. Strathem, E. W. Jones, & J. R. Broach (Eds.), The Molecular Biology of the Yeast Saccharomyces. (pp. 471-504). Cold Spring Harbor: Cold Spring Harbor Press. Studier, F. W., Rosenberg, A. H., Dunn, J. J., & Dubendorff, J. W. (1990). Use ofT7 RNA polymerase to direct expression of cloned genes. Methods in Enzymology, 185,60-89. Subramani, S. (1993). Protein import into peroxisomes and biogenesis of 150 the organelle. Annu Rev Cell BioI, 9,445-478. Suzuki, Y., Shimozawa, N., Yajima, S., Orii, T., Yokota, S., Tashiro, Y., Osumi, T., & Hashimoto, T. (1992). Different intracellular localization of peroxisomal proteins in fibroblasts from patients with aberrant peroxisome assembly. Cell Structure and Function, 17, 1-8. Swinkels, B. W., Gould, S. J., Bodnar, A. G., Rachubinski, R. A., & Subramani, S. (1991). A novel, cleavable peroxisomal targeting signal at the amino-terminus of the rat 3-ketoacyl-CoA thiolase. EMBO J. 10,3255-3262. Tagawa, M., Sakamoto, T., Shigemoto, K., Matsubara, H., Tamura, Y., Ito, T., Nakamura, A., Okitsu, 1., Imai, K., & Taniguchi, M. (1990). Expression of novel DNA-binding protein with zinc finger structure in various tumor cells. J BioI Chern, 262, 20021-20026. Takada, Y., & Noguchi, T. (1985). Characteristics of alanine: glyoxylate aminotransferase from Saccharomyces cerevisiae, a regulatory enzyme in the glyoxylate pathway of glycine and serine biosynthesis from tricarboxylic acid cycle intermediates. Biochem, 231, 157-163. Takahashi, M., Inaguma, Y., Hiai, H., & Hirose, F. (1988). Developmentally regulated expression of a human "finger" -containing gene encoded by the 5' half of the ret transforming gene. Mol Cell BioI. 8, 1853-1856. Tan, X., Waterham, H. R., Veenhuis, M., & Cregg, J. M. (1995). The Hansenala polymorpha PER8 gen encodes a novel perixosmal integral membrane protein involved in proliferation. J Cell BioI. 128(3),307-319. Teem, J. L., Berger, H. A., Ostedgaard, L. S., Rich, D. P., Tsui, L. C., & Welsh, M. J. (1993). Indentification ofrevertats for the cystic fibrosis DF508 mutation using STE6-CFTR chimeras in yeast. CelL 73, 335-346. ten Berge, A. M. A., Zoutewelle, G., & Needleman, R. B. (1974). 151 Regulation of maltose fermentation in Saccharomyces carisbergensis. Mol Gen Genet. 131, 113-121. Thieringer, R., Shio, H., Han, Y., Cohen, G., & Lazarow, P. B. (1991). Peroxisomes in Sa~charomyces cerevisiae: Immunofluorescence analysis and import of catalase A into isolated peroxisomes. Mol Cell BioI. 11, 510-522. Thieringer, R., & Raetz, C. R. H. (1993). Peroxisome-deficient Chinese hamster ovary cells with point mutations in peroxisome assembly factor-I. J BioI Chern, 268, 12631-12636. Tolbert, N. E. (1981a). Metabolic pathways in peroxisomes and glyoxysomes. Annual Review of Biochemistrv, 50, 133-157. Tolbert, N. E., & Essner, E. (1981b). Microbodies: peroxisomes and glyoxysomes. J Cell BioI. 91(3 Pt.2), 271s-283s. Trijbels, J. M. F., Berden, J. A., Monnens, L. A. H., Willems, J. L., Janssen, A. J. M., Schutgens, R. B. R., & Van Den Broek-Van Essen, M. (1983). Biochemical studies in the liver and muscle of patients with Zellweger Syndrome. Pediatric Research, 17,514-517. Tsukamoto, T., Yokota, S., & Fujiki, Y. (1990). Isolation and characterization of Chinese hamster ovary cell mutants defective in assembly ofperoxisomes. J Cell BioI. 110,651-660. Tsukamoto, T., Miura, S., & Fujiki, Y. (1991). Restoration by a 35K membrane protein of peroxisome assembly in a peroxisome-deficient mammalian cell mutant. Nature, 350, 77-81. Tzagoloff, A., Akai, A., & Needleman, R. B. (1975). Assembly of the mitochondrial membrane system: isolation of nuclear and cytoplasmic mutants of Saccharomyces cerevisiae with specific defects in mitochondrial functions. J Bacteriol. 122, 826-831. 152 Tzagoloff, A., Akai, A., & Foury, F. (1976). Assembly of the mitochondrial membrane system XVI. Modified form of the ATPase proteolipid in oligomycin-resistant mutants of Saccharomyces cerevisiae. FEBS Lett, 65, 391-395. Tzagoloff, A., & Dieckmann, C. L. (1990). PET genes of Saccharomyces cerevisiae. Microbiol Rev, 54(9),211-225. van den Bosch, H., Schutgens, R. B. H., Wanders, R. J. A., & Tager, J. M. (1992). Biochemistry ofperoxisomes. Annu Rev Biochem, 61, 157-197. van der Klei, I. J., SuIter, G. J., Harder, W., & Veenhuis, M. (1991). Assembly of Alcohol oxidase in the cytosol of a peroxisome-deficient mutant of Hansenula polymorpha-prperties of the protein and arichiticture of the crystals. Yeast 7,15-24. van Der Leij, I., Van Den Berg, M., Boot, R., Franse, M., Distel, B., & Tabak, H. F. (1992). Isolation of peroxisome assembly mutants from Saccharomyces cerevisiae with different morphologies using a novel positive selection procedure. J Cell BioI, 119, 153-162. van Der Leij, I., Franse, M. M., Elgersma, Y., Distel, B., & Tabak, H. F. (1993). PAS lOis a tetratricopeptide-repeat protein that is essential for the import of most matrix proteins into peroxisomes of Saccharomyces cerevisiae. Proc Natl Acad Sci USA, 90, 11782-11786. van Dijken, J. P., Veenhuis, M., Kreger-Van RIJ, N. J. W., & Harder, W. (1975). Microbodies in methanol-assimilating yeast. Arch MicrobioL 102, 41-44. van Lohuizen, M., Verbeek, S., Scheijen, B., Wientjens, E., & van der Gulden, H. (1991). Identification of cooperating oncogenes in EJ.l-myc transgenic mice by provirus tagging. Cell, 65, 737-752. Veenhuis, M., Keizer, I., & Harder, W. (1979). Characterization of 153 peroxisomes in glucose-grown Hansenula polymorpha and their development after the transfer of cells in methanol-containing media. Arch MicrobioL 120, 167-175. Veenhuis, M., Mateblowski, M., Kunau, W., & Harder, W. (1987). Proliferation of microbodies in Saccharomyces cerevisiae. Yeast 3, 77-84. Veenhuis, M., & Goodman, J. M. (1992). Peroxisomal assembly: membrane proliferation precedes the induction of the abundant matrix proteins in the methylotrophic yeast Candida biodinii. J Cell Sci, 96, 583-590. Versmold, H. T., Bremer, H. J., Herzog, V., Siegel, G., Bassewitz, D. B. v., Irle, 1., Voss, H. v., Lombeck, 1., & Brauser, B. (1977). A metabolic disorder similar to Zellweger Syndrome with hepatic acatalasia and absence of peroxisomes, altered content and redox state of cytochromes, and infantile cirrhosis with hemosiderosis. European Journal of Pediatrics, 124,261-275. Voom-Brouwer, T., Van Der Leij, 1., Hemrika, W., Distel, B., & Tabak, H. F. (1993). Sequence of the PASS gene, the product of which is essential for biogenesis ofperoxisomes in Saccharomyces cerevisiae. Biochim Biophys Acta, 1216,325-328. Waddle, J. A., Cooper, J. A., & Waterston, R. H. (1993). The Ci. and P subunits of nematode actin capping protein function in yeast. Mol BioI Cell, 1,907-917. Wanders, R. J., Schutgens, R. B., & Tager, J. M. (1985). Peroxisomal matrix enzymes in Zellweger syndrome: activity and subcellular localization in liver. Journal of Inherited Metabolic Disease, 8 Suppl. 2, 151-152. Weber, E. R., & Dieckmann, C. L. (1990). Identification of the CBPI polypeptide in mitochondrial extracts from Saccharomyces cerevisiae. J BioI Chern, 265(3), 1594-1600. 154 Wei, W., Mosteller, R. D., Sanyal, P., Gonzales, E., McKinney, D., Dasgupta, C., Li, P., Liu, B. X., & Broek, D. (1992). Identification ofa mammalian gene structurally and functionally related to the CDC25 gene of Saccharomyces cervisiae. Proc Nat! Acad Sci USA. 89, 7100-7104. Wiebel, F. F., & Kunau, W. (1992). The Pas2 protein essential for peroxisome biogenesis is related to ubiquitin-conjugating enzymes. Nature, 359, 73-76. Wiemer, E. A. C., Brul, S., Just, W. W., van Driel, R., Brouwer-Kelder, E., Van Den Berg, M., Weijers, P. J., Schutgens, R. B. H., van den Bosch, H., Schram, A., Wanders, R. 1. A., & Tager, J. M. (1989). Presence of peroxisomal membrane proteins in liver and fibroblasts from patients with the Zellweger syndrome and related disorders: evidence for the existence of peroxisomal ghosts. European Journal of Cell Biology, 50, 407-417. Wilson, G. N., Holmes, R. G., Custer, J., Lipkowitz, 1. L., Stover, J., Datta, N., & Hajra, A. (1986). Zellweger syndrome: Diagnostic assays, syndrome delineation, and potential therapy. American Journal of Medical Genetics, 24, 69-82. Wolf, K., Dujon, B., & Slonimski, P. P. (1973). Mitochondrial genetics V. Multifactorial mitochondrial crosses involving a mutation conferring paromomycin-resistance in Saccharomyces cerevisiae. Mol Gen Genet 125, 53-90. Wu, M., & Tzagoloff, A. (1987). Mitochondrial and cytoplasmic fumarases in Saccharomyces cerevisiae are encoded by a single nuclear gene FUM!. J BioI Chern, 262, 12275-12282. Zhang, J. W., Han, Y., & Lazarow, P. B. (1993a). Novel peroxisome clustering mutants and peroxisome biogenesis mutants of Saccharomyces cerevisiae. J Cell BioI. 123, 1133-1147. Zhang, J. W., Luckey, C., & Lazarow, P. B. (1993b). Three peroxisomal 155 protein packaging pathways suggested by selective penneabilization of yeast mutants defective in peroxisome biogenesis. Molecular Biology of the CelL 4, 1351-1359. Zhang, J. W., & Lazarow, P. B. (1995). PEB1(PAS7) in Saccharomyces cerevisiae encodes a hydrophilic, intra-peroxisomal protein that is a member of the WD repeat family and is essential for the import of thiolase into peroxisomes. J Cell BioI. 129(1),65-80. Zitomer, R. S., & Lowry, C. V. (1992). Regulation of gene expression by oxygen in Saccharomyces cerevisiae. Microbiol Rev, 56, 1-11. Zwart, K. B., Veenhuis, M., Plat, G., & Harder, W. (1983). Characterization of glyoxsomes in yeasts and their transformation into peroxisomes in response to changes in environmental conditions. Arch Microbiol. 136,28-38.