peroxisome senescence and the role of catalase import in cellular aging

Peroxisomal import of catalase derivatives in early and late passage cells .... Human peroxisomal disorders Human peroxisomal disorders are typically divided into two groups, including;

NOTE TO USERS

This reproduction is the best copy available

JAY I KOEPKE DISSERTATION

Submitted to the Graduate School

of Wayne State University,

Detroit, Michigan

in partial fulfillment of the requirements

for the degree of

3221036 2006

UMI Microform Copyright

All rights reserved This microform edition is protected against unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company

300 North Zeeb Road P.O Box 1346 Ann Arbor, MI 48106-1346

by ProQuest Information and Learning Company

I dedicate this thesis to my wife Amy, who has been a constant source of

encouragement and support Without her devotion and tolerance I do not believe

I would have been able to accomplish the work herein She is an inspiration and

I am truly grateful for having her by my side And to my unborn child, may you

benefit from your mother’s everlasting affection as I have

I would first like to offer my uppermost gratitude to my advisor, Dr Stanley

R Terlecky, whose enthusiastic approach to science as well as life’s

circumstances has left a lasting impression and will bolster my future endeavors

His encouragement, guidance, and friendship helped facilitate my graduate

career I would like to sincerely thank my committee members – Drs Nicholas

G Davis, Ladislau C Kovari, Roy B McCauley, and Raymond R Mattingly for

their helpful suggestions and supervision throughout this process I also

acknowledge collaborators such as Dr Paul A Walton and Dr Marc Fransen for

their expertise and contributions to my research project

I genuinely thank past and present members of the Terlecky laboratory –

in particular Julie Legakis, Ferdous Barlaskar, Chris Wood, and Laura Terlecky

They provided invaluable support for many of the studies herein

I am grateful to my parents Walt and Helen Koepke – who have been a

permanent fixture in my overall development I also owe appreciation to my

siblings, Jeff and Jill, and my in-laws Thomas and Wiesia Petroske

DEDICATION ii

ACKNOWLEDGEMENTS iii

LIST OF TABLES v

LIST OF FIGURES vi

CHAPTERS CHAPTER 1 – Introduction 1

CHAPTER 2 – Materials and Methods 29

CHAPTER 3 – Peroxisome Senescence in Human Fibroblasts 48

CHAPTER 4 – Catalase Inactivation Drives Cells Towards a Senescence-like Phenotype 75

CHAPTER 5 – Restoration of Peroxisomal Catalase Import 100

REFERENCES 127

ABSTRACT 155

AUTOBIOGRAPHICAL STATEMENT 157

Table 1 Human peroxisomal hydrogen peroxide-producing and

-degrading enzymes and their relative PTS1 strength 28

Figure 1 Quantitative in vitro assay for peroxisomal protein import 21

Figure 2 Senescence-associated β-galactosidase activity in human diploid fibroblasts (HDFs) 50

Figure 3 Peroxisomes in early, middle, and late passage HDFs 52

Figure 4 Ultrastructure of early and late passage HDFs 54

Figure 5 Peroxisomes in co-cultured early and late passage HDFs 56

Figure 6 PTS1(-SKL)-protein import in early, middle, and late passage HDFs 59

Figure 7 Analysis of Pex5p binding 62

Figure 8 Quantitative analysis of peroxisomal protein import 64

Figure 9 Catalase import in early and late passage cells 65

Figure 10 Pex5p’s association with organelle membranes 67

Figure 11 Hydrogen peroxide accumulates in aging HDFs 70

Figure 12 Effect of hydrogen peroxide on PTS1(-SKL)-protein import and PTS1 receptor localization 72

Figure 13 Inhibition of catalase by treatment of cells with 3-AT 77

Figure 14 Accumulation of ROS upon acute inhibition of catalase with 3-AT 80

Figure 15 Chronic catalase inhibition increases peroxisomal numbers 82

Figure 16 The effect of catalase inhibition on proliferation of HDFs 86

Figure 18 DNA damage in human fibroblasts 90

Figure 19 PTS1(-SKL)-protein import in untreated and 3-AT treated HDFs 91

Figure 20 Mitochondrial membrane potential and ROS production 94

Figure 21 Gelatin zymograpy of conditioned media from HDFs 98

Figure 22 Binding of Pex5p to catalase derivatives 104

Figure 23 Peroxisomal import of catalase derivatives 106

Figure 24 Peroxisomal import of catalase derivatives in early and late passage cells 108

Figure 25 Localization of catalase-SKL and Pmp70p in late passage cells 109

Figure 26 Association of catalase-SKL with endogenous catalase in cells 111

Figure 27 CPP-mediated catalase-SKL transduction into cultured human fibroblasts 114

Figure 28 Distribution of transduced catalase-SKL in mouse skin 116

Figure 29 Effects of transducible catalase-SKL administration on hydrogen peroxide levels in the skin of aged mice 117

Figure 30 Amplification plots of real-time PCR analysis of mouse transgenic DNA 119

Figure 31 Peroxisome deterioration spiral 122

CHAPTER 1

INTRODUCTION

I Peroxisomes

Peroxisomes, present in nearly all eukaryotic cells, are indispensable

subcellular organelles bound by a single membrane (DeDuve and Baudhuin,

1966) The study of peroxisomes began in the 1950’s with the discovery of small

(0.1-1.0 µm) spherical structures within mouse kidney cells imaged usingelectron microscopy (Rhodin, 1954) Originally termed microbodies, DeDuve and

Baudhuin further characterized these organelles in 1966 – their enzymatic

description established the colocalization of hydrogen peroxide-producing

oxidases as well as catalase, an enzyme involved in the degradation of hydrogen

peroxide As a result, the organelle was designated the functional term

“peroxisome”

Mammalian cells typically contain in the range of a few hundred to a few

thousand peroxisomes; a number that can increase in response to extracellular

stimuli such as high fat diets (Ishii et al., 1980), thyroid hormones (Fringes and

Reith, 1982), and diabetes (Horie et al., 1981) Other peroxisome proliferative

agents include hypolipidemic drugs such as fibrates (Reddy, 1973), plasticizers,

and chlorinated hydrocarbons (Bremer et al., 1981) Our evidence, coupled with

studies found in the literature, indicate cells experiencing oxidative stress,

including those that are aged, specifically diseased, or xenobiotic-exposed, also

contain increased numbers of peroxisomes (Legakis et al., 2002; Feher et al.,

2005)

Formed through the concerted actions of a specific set of some 32

proteins called peroxins, peroxisomes participate in a wide range of physiological

processes depending on species and cell type (Heiland and Erdmann, 2005) In

mammals, these functions include the decomposition of lipophilic carboxylates

such as very-long-chain fatty acids (≥C24), some long-chain fatty acids (≥C14-22),

and eicosanoid lipid derivatives; conversion of cholesterol into bile acids through

β-oxidation; elimination of certain xenobiotics and glyoxylate; the initial

degradation of 3-methyl-branched fatty acids by α-oxidation; the synthesis ofcholesterol, dolichol, and other isoprenoids; the committed step in ether

phospholipids (plasmalogen) synthesis; and the catabolism of certain purines,

polyamines, and amino acids (Wanders, 2004) Many of these functions are

exclusive to the peroxisome, while others are shared with separate

compartments of the cell including the cytosol, endoplasmic reticulum, and

mitochondria In instances where function is shared, the peroxisomal

contribution may be essential, for example in the case of plasmalogen synthesis

In contrast, cholesterol synthesis appears largely complementary to the

endoplasmic reticulum-based system

Similar to mitochondria, peroxisomes contain a fatty acid β-oxidationmachinery The difference between the two systems is that they catalyze

different fatty acids and fatty acid derivatives As mentioned above,

very-long-chain fatty acids, long-very-long-chain fatty acids, and eicosanoids such as prostaglandins,

thromboxanes, and leukotrienes are oxidized within peroxisomes, whereas

mitochondria oxidize the majority of other cellular fatty acids Both peroxisomal

and mitochondrial β-oxidation involve a set of four consecutive reactions: (1)dehydrogenation; (2) hydration (of the double bond); (3) dehydrogenation again;

and (4) thiolytic cleavage Following each cycle of the peroxisomal pathway, a

2-carbon unit is split from each fatty acid in the form of an acetyl-CoA unit As a

result, peroxisomes are only able to chain-shorten fatty acids and cannot

degrade fatty acids to completion Following β-oxidation within the peroxisome,many of the products are shuttled to the mitochondria for complete oxidation to

carbon dioxide and water

The initial steps of mitochondrial β-oxidation result in the donation ofelectrons to the respiratory chain at the level of coenzyme Q (Wanders et al.,

1999) In contrast, peroxisomal oxidases donate their electrons directly to

molecular oxygen to produce hydrogen peroxide Thus, all molecular oxygen

consumed by the peroxisome results in the formation of hydrogen peroxide

Decomposition of this potentially toxic metabolite by catalase is therefore another

fundamentally critical function of the organelle (Reddy and Mannaerts, 1994;

Perichon et al., 1998; Van Veldhoven and Mannaerts, 1999; Wanders, 2004)

II Human peroxisomal disorders

Human peroxisomal disorders are typically divided into two groups,

including; (1) the peroxisome biogenesis disorders (PBD), and (2) the single

peroxisomal enzyme deficiencies Patients with PBDs suffer from a range of

pathologies, including those with effects on neurological, skeletal, hepatological,

and ocular systems

Peroxisome function is disrupted due to deletions of, or mutations in, one

of the many PEX genes, involved in the biogenesis of peroxisomes (Gould and

Valle, 2000) (PEX genes encode the peroxins defined above) Peroxisomal

enzyme activities are reduced or undetectable; and those that remain are usually

found mislocalized to the cytosol (Lazarow et al., 1985; Suzuki et al., 1988;

Wanders et al., 1984) At present, mutations in 14 different PEX genes have

been shown to be associated with disease (Shimozawa et al., 2004) That is,

human peroxisomal disease falls into 14 complementation groups

PBDs are subdivided into two broad clinical phenotypes, the first

collectively known as the Zellweger spectrum of disease and the second as

rhizomelic chondrodysplasia punctata (RCDP) type I The Zellweger spectrum

includes Zellweger syndrome (ZS), infantile Refsum’s disease (IRD), and

neonatal adrenoleukodystrophy (NALD) Of these, patients with Zellweger

(cerebro-hepato-renal) syndrome demonstrate the most severe neurological,

hepatic, and renal defects, including facial dysmorphisms, generalized hypotonia,

hepatic fibrosis/cirrhosis, adrenal insufficiency, and renal cysts These

individuals commonly die within the first few months or years after birth NALD is

of intermediate severity, and IRD is a milder form RCDP type I patients are

distinguished from other PBDs by the appearance of severe skeletal

abnormalities including disproportion in the length of the upper arms and thighs,

abnormal growth of cartilage, and spotting (punctata) visualized by X-ray near

the epiphysis of bones (Gould and Valle, 2000)

The single enzyme disorders are characterized by a deficiency in a single

peroxisomal function; for example, X-linked adrenoleukodystrophy (X-ALD) is

caused by a defect in the ALD protein, a peroxisomal ABC transporter for

very-long-chain fatty acids (Mosser et al., 1993) Other known single enzyme

disorders include RCDP type II (dihydroxyacetone phosphate acyltransferase)

and type III (alkyl-dihydroxyacetonephosphate synthase), Refsum’s disease

(phytanoyl-CoA hydroxylase), hyperoxaluria type I (alanine:glyoxylate

amnotransferase), the β-oxidation disorders (acyl-CoA oxidase, bifunctionalprotein, and thiolase), and acatalasemia (catalase)

Studies on patients with Zellweger spectrum disorders were the basis for

the initial findings linking peroxisomes with essential activities in human

physiology The accumulation of very-long-chain fatty acids in patients with

Zellweger syndrome led to the identification of important roles for the organelle in

fatty acid β-oxidation (Brown et al., 1982) Similarly, elevated levels of phytanicacid provided evidence for a role of peroxisomes in fatty acid α-oxidation.Furthermore, deficiencies in plasmalogen levels of Zelleweger patients pointed to

their essential role in ether phospholipid biosynthesis (Heymans et al., 1983)

III Peroxisome biogenesis

Certain yeast species contain very few peroxisomes when cultured on

glucose media Relying on their ability to proliferate peroxisomes, these

organisms have the capacity to utilize alternative carbon sources such as fatty

acids or methanol via peroxisomal metabolism (Veenhuis et al., 1987) These

observations permitted genetic approaches to be applied in an effort to identify

genes involved in peroxisome biogenesis Specifically, PEX genes were

identified by isolation of mutants lacking their ability to grow on such carbon

sources due to the inability to form functional peroxisomes (Erdmann et al.,

1991) Chinese hamster ovary (CHO) cells in culture have also been used as a

powerful model system for the identification and characterization of required

peroxisome biogenesis gene products (Fujiki, 1997) The remarkable

conservation of peroxisome biogenesis through evolution has allowed

observations using these two model systems to be applied to the understanding

of the human system Peroxisome biogenesis may be viewed as three related

and highly coordinated processes: (1) membrane lipid and protein assembly, (2)

matrix protein import, and (3) organelle growth and division

Membrane lipid and protein assembly

Peroxisomes are not capable of synthesizing their own membrane lipids of

which phosphatidycholine and phosphatidylethanolamine are the major

constituents Instead, it is widely believed that lipids destined for the peroxisome

are synthesized in the endoplasmic reticulum and transported to the organelle

How this mechanism operates remains to be discerned One way may be via a

set of specialized vesicles trafficking between the endoplasmic reticulum and

peroxisome (Purdue and Lazarow, 2001) Alternatively, a mechanism akin to the

trafficking of phospholipids from the endoplasmic reticulum to mitochondria

involving transfer from one membrane to another at specific contact sites may be

at work (Ardail et al., 1993; Achleitner et al., 1999)

Peroxisomal proteins destined for insertion into the membrane contain at

least one transmembrane region together with a basic amino acid stretch that

serves as the membrane targeting signal (mPTS) To date, only three peroxins

have been implicated in peroxisomal membrane protein (PMP) import - Pex3p,

Pex16p, and Pex19p Of these, Pex19p has received by far the most attention

This protein is predominantly cytosolic and has been found to be farnesylated,

although there is debate as to whether this modification is required for proper

function (Kammerer et al., 1997; Gotte et al., 1998; Mayerhofer et al., 2002)

Cells lacking Pex19p contain mislocalized or unstable PMPs Of debate is

whether Pex19p acts as a specific PMP import receptor or as a more nonspecific

assembly chaperone (Fransen et al., 2001; Fransen et al., 2004) Recently, a

Pex19p-binding consensus sequence was identified (Rottensteiner et al., 2004)

This sequence was found to be necessary for PMP targeting and sufficient to

target GFP to the peroxisomal membrane when located adjacent to a

transmembrane domain These results strongly favor Pex19p as a receptor for

targeting peroxisomal membrane proteins Although, there may exist a second

set of proteins including Pex3p, that localize to the peroxisomal membrane

independent of Pex19p (Jones et al., 2004)

Matrix protein import

Translation of peroxisomal matrix proteins occurs on free ribosomes in the

cytosol followed by trafficking to the surface of the peroxisome via one of two

soluble receptors These receptors, Pex5p and Pex7p, bind peroxisomal

targeting signal type 1 (PTS1) or type 2 (PTS2), respectively The majority of

matrix proteins contain a PTS1; only a select few contain PTS2 Classically, the

PTS1 consists of a conserved tripeptide located at the extreme carboxy-terminus

– most notably serine-lysine-leucine (SKL), or a biochemically related variant

(S/A/C-K/R/H-L/M) (Gould et al., 1989; Lametschwandtner et al., 1998) One

prominent exception is catalase, which contains a divergent PTS1, consisting of

lysine-alanine-asparagine-leucine (KANL) Curious is both the length of the PTS,

and the identity of the amino acids Residues as far as 12 amino acids upstream

from the carboxy-terminus may also play a role in recognition by Pex5p and thus,

in import (Eisenhaber et al., 2003; Neuberger et al., 2003; Lametschwandtner et

al., 1998) These upstream amino acids appear to contribute to the interaction

with Pex5p through their effects on accessibility/non-accessibility of the PTS1

(see http://mendel.imp.ac.at/mendeljsp/sat/pts1/PTS1predictor.jsp for “PTS1

predictor”, a powerful in silico program designed to evaluate PTS1 strength

based on these and other parameters) The generally accepted view is that for a

given enzyme, the stronger the interaction with Pex5p, the greater the capacity

for peroxisomal import

Direct binding of the 68 kDa receptor, Pex5p, to PTS1 is mediated by

seven tetratricopeptide repeats (TPR) contained within the protein Pex5p exists

in two isoforms, a short form (Pex5pS), and a long form (Pex5pL), which differ by

virtue of a 37 amino acid insert in the long form Six of the TPR domains interact

directly with the signal sequence, whereas TPR4 acts as a hinge region X-ray

crystallographic studies indicate a series of hydrophilic, acidic, and hydrophobic

binding pockets along the ligand interaction site of Pex5p (Gatto et al., 2000;

Maynard et a., 2004) These pockets function to secure the PTS1 residues of

the substrate within the receptor Therefore, changes in the identity of the PTS1

or upstream amino acids would be expected to be differentially recognized and

bound by Pex5p Under normal conditions, the molecule is primarily cytosolic,

with only a small amount found associated with peroxisomes It is generally

thought that Pex5p cycles between the cytosol and the peroxisome, giving rise to

the model of a shuttling receptor (Marzioch et al., 1994; Dodt and Gould, 1996)

Once complexed, the receptor and ligand move to the peroxisome membrane

and engage the membrane-associated components of the import machinery,

specifically Pex14p and Pex13p (Fransen et al., 1998) Evidence suggests an

extended shuttle mechanism in which the import receptor not only docks at the

peroxisomal membrane, but also enters the organelle lumen together with its

cargo After releasing the substrate, unloaded receptors are recycled back to the

cytosol for another round of import (Dammai and Subramani, 2001) The two

subpopulations (cytosolic and peroxisome-associated) of Pex5p may exist in

different oligomeric states The cytosolic form exists as a monomer; when

associated with the peroxisome Pex5p may form a tetramer – although there is

debate on this point (Gouveia et al., 2000; Costa-Rodrigues et al., 2005) Pex5p

has also been suggested to actually create a translocation pore in the

peroxisome membrane (Erdmann and Schliebs, 2005) although little evidence

supports this view at present

The less conserved PTS2 consensus sequence (R-L/I-X5HL) is most often

located close to the amino-terminus (Swinkels et al., 1991) Unlike PTS1, this

signal is usually removed by a peroxisomal peptidase (Subramani, 1993)

Although relatively few proteins contain a PTS2, defects in this import pathway

also lead to human disease

The PTS2 import pathway relies on the protein Pex7p as its cytosolic

receptor The molecule’s protein sequence includes six WD40 domains, each

containing a central tryptophan-aspartic acid motif (Marzioch et al., 1994) In

Saccharomyces cerevisiae, Pex7p interacts with Pex18p/Pex21p to mediate

docking to the peroxisome membrane It has been shown that the 37 amino acid

insertion in the long isoform of Pex5p can functionally replace Pex18p (Schafer

et al., 2004) Therefore, human Pex5pL mediates binding of Pex7p to the import

machinery, overcoming the requirement for Pex18p/Pex21p (Matsumura et al.,

2000; Dodt et al., 2001) Furthermore, Pex5p-deficient cell lines have been

found to be defective in both PTS2 and PTS1 import; therefore it has been

speculated that Pex5p is involved in both pathways (Wiemer et al., 1995; Dodt et

al., 1995) Indeed, both isoforms have been shown to restore PTS1 import in

Pex5p-deficient CHO cell mutants, while only Pex5pL restored both PTS1 and

PTS2 import (Otera et al., 2000)

Following the binding of Pex5p to PTS1 proteins (or Pex7p to PTS2

proteins), the complex interacts with the import machinery at the peroxisomal

membrane Pex14p is believed to mediate the initial step in peroxisomal

membrane association of Pex5p:PTS1 and Pex7p:PTS2 (Albertini et al., 1997;

Fransen et al., 1998; Schliebs et al., 1999; Will et al., 1999; Saidowsky et al.,

2001) In yeast, Pex14p has also been shown to interact with other peroxins

including Pex13p and Pex17p (Girzalsky et al., 2006) Once docked at the

membrane, the PTS1/PTS2-containing cargo translocates to the peroxisomal

matrix via a set of peroxins collectively termed the “importomer” or “import

complex” The import complex is comprised of the RING-finger proteins Pex2p,

Pex10p, and Pex12p each of which also appear to be involved in some aspect of

Pex5p recycling (Chang et al., 1999, Collins et al., 2000, Dodt and Gould, 1996)

Cells deficient in any one of these peroxins exhibit an increase in the amount of

peroxisome-associated Pex5p (Dodt and Gould, 1996) and compromised protein

import

An intraperoxisomal organizer of the import machinery, Pex8p, links the

docking and import complexes (Agne et al., 2003) Interestingly, this protein

contains both PTS1 and PTS2 targeting sequences, however, neither are

required for Pex8p import (Rehling et al., 2000; Wang et al., 2004) Rather, it

has been put forth that the PTS1 and PTS2 signals within Pex8p bind the import

receptors (Pex5p and Pex7p) on or inside the peroxisome and after cargo

release to prevent inappropriate reassociation with cargo (Wang et al., 2004;

Heiland and Erdmann, 2005)

Pex5p either gains access to the peroxisomal matrix during import or is

released from the membrane after cargo release In both models it is well

supported that a single molecule of Pex5p can undergo many rounds of import

Following release of cargo, the AAA (ATPases associated with various cellular

activities) peroxins, Pex1p and Pex6p facilitate Pex5p export from the

peroxisome via an ATP-dependent mechanism Pex1p and Pex6p are both

cytosolic membrane-associated; Pex26p appears to act as their anchoring

peroxin (Miyata and Fujiki, 2005)

Pex5p (Platta et al., 2004; Kiel et al., 2004; Kragt et al., 2004) and Pex18p

(Purdue and Lazarow, 2001) have been found to be ubiquitinated during the

import process However, the physiological role of this ubiquitination is unclear

Pex18p is involved in Pex7p targeting and becomes mono- and diubiquitinated

followed by proteasomal degradation Ubiquitination of Pex5p may act as a form

of quality control (Kiel et al., 2004) or a signal for export back to the cytosol

(Platta et al., 2004; Kragt et al., 2004) The RING-finger peroxins Pex2p,

Pex10p, and Pex12p are candidates for the required, but thus far elusive

E3-ubiquitin protein ligase activity

Growth and division

The ‘growth and division’ model of peroxisome biogenesis posits that all

peroxisomal matrix and membrane proteins are post-translationally imported

from the cytosol to pre-existing organelles Peroxisomes then mature and

multiply by division rather than forming de novo (Lazarow and Fujiki, 1985).

Earlier assumptions were that each peroxisome originates from the endoplasmic

reticulum (Novikoff and Shin, 1964) In recent years it has been proposed that

both pathways may contribute to the formation of nascent peroxisomes (Heiland

and Erdmann, 2005)

Peroxisomes segregate during cell division moving on either actin

filaments in fungi and plants (Hoepfner et al., 2001; Jedd et al., 2002; Mano et

al., 2002; Mathur et al., 2002) or microtubules in mammals (Rapp et al., 1996;

Wiemer et al., 1997; Thiemann et al., 2000; Schrader et al., 2003) In response

to specific metabolic or environmental signals, many cell types regulate

peroxisome number, volume, and size Other than peroxisomal division in

constitutively dividing cells, a process called “peroxisome proliferation” rapidly

increases organelle number in response to certain extracellular stimuli For

instance, yeast cells proliferate peroxisomes when shifted to carbon sources

metabolized in peroxisomes Rodents exposed to a class of structurally diverse

chemicals called peroxisome proliferators (PP) dramatically increase the number

of their liver peroxisomes Humans are not as sensitive to PPs, although only

few studies have addressed this point Contrary to proliferation, a process called

“pexophagy” controls peroxisome turnover (Farre and Subramani, 2004)

IV In vitro systems to study peroxisomal import

Since its discovery, great strides have been made in our understanding of

the molecular mechanism of peroxisomal biogenesis In particular, much has

been learned about the import process These advances have been greatly

accelerated by the development of in vitro systems, where the import process is

reconstituted Among the techniques used and systems developed are protease

protection-based assays employing purified organelles, microinjection-based

approaches using live cells, and ELISA-based assays using semi-permeabilized

cells

Each assay has been validated using stringent controls in order to

substantiate that the methods truly recapitulate protein import into the

peroxisomal matrix All of the in vitro systems have been shown to be ligand-,

time-, temperature-, and ATP-dependent It has been demonstrated that a

number of import substrates can be used, allowing for greater flexibility in

experimentation Import of the PTS1-containing substrates has been found to be

specific and saturable Therefore, these are clearly valuable tools with which to

gain knowledge about the means proteins employ to gain access to the organelle

lumen Also, recapitulating peroxisomal protein import in vitro is important not

only from a basic science perspective, but also with respect to its ability to shed

light on human physiology and pathophysiology

Purified peroxisomes and protease protection

Much of what is known about the basic mechanism and requirements for

import was first established through the use of purified peroxisomes and

protease protection assays These early in vitro assays employed isolated

organelles from rat liver or yeast cells The purified peroxisomes are incubated

with an appropriate PTS1-containing substrate in a manner that promoted

translocation across the membrane Protection of the substrate within the

peroxisome matrix from exogenously added protease was the hallmark of import

and resolved by western blot or autoradiography (Lazarow et al., 1991)

Alternatively, a postnuclear supernatant fraction containing organelles including

peroxisomes were incubated with a radiolabeled substrate prior to protease

protection assays After import occurred, peroxisomes were separated from

other organelles on a density gradient and cosedimentation of the substrate with

peroxisomal markers measured (Fujiki and Lazarow, 1985; Miyazawa et al.,

1989)

Since its inception the protease protection assay has revealed a variety of

requirements for peroxisomal protein import Early studies not only determined

that import was dependent on time and temperature, but also the requirement of

membrane proteins and ATP hydrolysis (Lazarow et al., 1985; Imanaka et al.,

1987) In the absence of ATP, binding of the substrate to the membrane

occurred but translocation did not take place Therefore, it is believed that import

is a two-step process of binding followed by translocation across the membrane

The assay also showed that a membrane potential was not required (Imanaka et

al., 1987), whereas ions such as potassium, magnesium, and sodium did

enhance import (Imanaka et al., 1987)

An assortment of substrates have been studied using this technique

Experiments involving acyl-CoA oxidase mutants and fusion proteins along with

truncated catalase helped determine that the carboxy-terminal residues were

necessary for these proteins to be imported into rat liver peroxisomes (Miura et

al., 1992) Fusion proteins containing portions of acyl-CoA oxidase and the

cytosolic enzyme dihydrofolate reductase (DHFR) were also able to cross the

peroxisome membrane (Small et al., 1988) This was the first evidence that a

PTS could ferry a non-peroxisomal protein into the lumen of the organelle

While this assay was an important tool in the initial characterization of

peroxisomal protein import, it did have its limitations In particular, the inherent

fragility of organelles separated from their cellular context as well as certain

vagaries associated with protease protection itself caused investigators to move

to a new approach

Microinjection

This system employs live cells into which the import substrate is

microinjected and import visualized by the time-dependent redistribution of the

substrate into peroxisomes using immunofluorescence Microinjection allows the

use of intact mammalian cells, presumably better recapitulating the process as it

occurs in vivo.

This method was used to demonstrate, for the first time, that cells from

patients afflicted with Zellweger syndrome do not import a PTS1-containing

substrate This assay was also used to begin the identification of cytosolic

factors necessary for import (Walton et al., 1994), as well as the landmark

observation that substrates can be imported into the peroxisome while in a folded

or oligomeric state (Walton et al., 1995)

Semi-permeabilized cell-free assays

These systems make use of semi-permeabilized cells, where the plasma

membrane is perforated, but the organelle membranes are left intact Import is

assessed by immunofluorescence microscopy of cells immobilized on coverslips

or by enzyme-linked immunosorbent assay (ELISA) These are discussed in

turn

Unlike import into purified peroxisomes or microinjection, the development

of the immunofluorescence-based import assay allowed investigators to utilize

semi-intact cells to identify specific factors required for peroxisomal matrix protein

import Specifically, the plasma membrane of the cells grown on coverslips is

permeabilized with streptolysin-O, a sulfhydryl-activated bacterial cytolysin which

forms pores up to 30 nm in cholesterol-rich membranes (Bhadki et al., 1985)

SLO forms these pores in the plasma membrane only, leaving the organelle

membranes intact This allows the PTS-containing import substrate to gain

access to the interior of the cell After an incubation to allow import to occur, the

cells are then analyzed for import of the substrate by immunofluorescence

microscopy

This assay revealed that cytosolic components are essential for the import

of both PTS1- and PTS2-containing proteins (Wendland and Subramani, 1993;

Legakis and Terlecky, 2001) It was established, through a variety of assay

permutations, that Hsp70 and Hsp40, as well as Pex5p and Pex14p are

absolutely required for both PTS1 (Walton et al., 1994; Fransen et al., 1998), and

PTS2 import (Legakis and Terlecky 2001), and that Pex5p and Pex14p play a

role in PTS1 import (Wiemer et al., 1995; Fransen et al., 1998; Legakis and

Terlecky, 2001) Use of this assay also confirmed that oligomeric structures are

imported into peroxisomes via the PTS2 import pathway (Legakis and Terlecky,

2001)

The ELISA assay (Figure 1) is derived from a cell-free system first

developed for analyzing receptor-mediated endocytosis (Smythe et al., 1992)

Quantitation of peroxisomal import in human cells is assessed using a

biotinylated PTS1-containing substrate captured on antibody-coated microtiter

wells (Terlecky, 2002) The importance of being able to precisely quantitate

import offers an advantage over previously mentioned non-quantitative methods

The assay is a reliable multiple sample system, designed to examine a number

of parameters within a single experiment

Tissue culture cells are mechanically disrupted to perforate the plasma

membrane During an import reaction, biotinylated PTS-containing substrates

are able to access the intact peroxisomes within the semi-intact cells

Biotinylated luciferase is typically used as the import substrate as it contains a

known PTS1 targeting signal (SKL) (Gould et al., 1987) and antibodies are

available However, in this thesis report the assay was also modified to evaluate

the import of human peroxisomal catalase

After the incubation period unimported substrate is removed by

centrifugation and residual amounts are inactivated with an excess of avidin,

which blocks free biotin groups located on the unimported substrate Similar to

protease protection, the hallmark of import is the ability of an imported substrate

to resist inactivation, in this case masking by avidin Unbound avidin is

quenched by the addition of biocytin, a biotin-like molecule At this point

detection of the imported substrate can be assessed by two different protocols

(Terlecky, 2002) The first calls for cells to be solubilized in a sodium

dodecylsufate salt (SDS) and Triton X-100 containing buffer The imported

substrate is released from the peroxisome and captured by anti-luciferase

antibodies on microtiter wells For the second approach, an organelle pellet is

prepared by homogenization of the cells and subsequent centrifugation steps

The organelles are solubilized and the released substrate captured on microtiter

wells (Terlecky, 2002)

The imported substrate trapped by anti-luciferase (or catalase) antibodies

is recognized by streptavidin labeled with horseradish peroxidase which forms a

colored precipitate This molecule detects the unmasked biotin groups on the

imported substrate only, while biotin on unimported substrate is not recognized

A microplate absorbance reader is utilized to measure the signal, completing the

assay

This system has made many important contributions to the field of

peroxisomal protein import Time and temperature dependencies as well as the

requirement of ATP hydrolysis have been recapitulated Cytosol has also been

shown to stimulate import As expected, fibroblasts from a Zellweger patient

were found to be unable to import the PTS1 ligand in this assay An advantage

of this system is the ability to add antibodies, ions or other molecules to the

reaction For example, it has been shown that zinc has a stimulatory effect on

import Furthermore, antibodies to the zinc-finger protein, Pex2p, were found to

have an inhibitory effect (Terlecky et al., 2001)

The ELISA-based assay has the ability to quantitate import or the lack

thereof in a number of human cell lines such as epidermoid carcinoma (A431)

cells, human diploid fibroblasts, and peroxisomal import defective (Zellweger) cell

lines When compared to other in vitro assays this system is rapid and has the

capacity to examine a number of parameters Another advantage is the ability to

evaluate the kinetics of import by implementing multiple time points

The assay is ELISA-based and employs semi-permeabilized human cells and abiiotinylated import substrate Import is either assessed directly in cells or afterisolation of cellular organelles/peroxisomes A, avidin; B, biotin; HRP,horseradish peroxidase; P, peroxisome; PTS, peroxisomal targeting signal; SA,

streptavidin Figure is taken from Current Protocols in Cell Biology (Terlecky,

2002)

V Role of peroxisomes in aging

Among the contributing factors to cellular aging are telomere shortening,

DNA damage and related genomic instability, modified expression of certain

genes, alterations in membrane lipid homeostasis, and the accumulation of

reactive oxygen species (ROS) (reviewed in Johnson et al., 1999) The free

radical theory of aging dates back to 1956, when Denham Harman put forth the

notion that free radicals formed during aerobic respiration have the capacity to

contribute to the process of aging (Harman, 1956) Although other factors are

clearly at work, oxygen free radicals do appear to play a significant role in the

degenerative process associated with aging through indiscriminate damage to

macromolecular components including protein, lipid, and nucleic acids (Beckman

and Ames, 1998) Mitochondria are widely regarded as the chief cellular

generators of ROS and ironically, a major focus of free radical assault (Beckman

and Ames, 1998; Lee and Wei, 2001) However, mitochondria are not the only

source of cellular ROS

Consumption of oxygen occurs in different compartments of the cell, in

particular mitochondria, the endoplasmic reticulum, and peroxisomes (Moldovan

and Moldovan, 2004) The vast majority, some ninety percent, of the oxygen

metabolized in mitochondria is converted to water and the rest to superoxide

(O2-) Peroxisome fatty acid β-oxidation produces a large amount of hydrogenperoxide and relatively small amounts of superoxide Importantly, the highly

reactive hydroxyl radicals (OH·) can be formed through reactions involving the

overproduction of O2- and H2O2

Consistent with the theory that membrane homeostasis affects aging

(Shinitzky, 1987), alterations in peroxisomal lipid metabolism may result in

detrimental changes to overall lipid composition and the function of

biomembranes Genetic defects in peroxisomal lipid metabolism are known to

cause severe pathologies during human development, particularly involving

neurological degeneration Conceivably, mild peroxisomal insufficiency during

aging may result in similar manifestations

The single limiting peroxisomal membrane provides an isolated

environment in which hydrogen peroxide can be readily decomposed by the

tetrameric, heme-containing enzyme, catalase (Chance et al., 1979) The role of

catalase as a key component of a cell’s antioxidant defenses is well established

(Beckman and Ames, 1998; Chance et al., 1979; Halliwell and Gutteridge, 1989;

Masters and Crane, 1995; Perichon et al., 1998) Defects in its expression

(Eaton and Ma, 1995; Wen et al., 1988), stability (Crawford et al., 1988), or

peroxisomal localization are associated with oxidative stress, disease, and aging

Mislocalized catalase is associated with accumulation of hydrogen peroxide and

perhaps other reactive oxygen species (ROS) in cells, and with seriously

compromised neurological function in patients (Kawada et al., 2004; Legakis et

al., 2002; Sheikh et al., 1998)

The ability to balance the generation and decomposition of ROS prevents

oxidative damage to the peroxisome and other cellular components of the cell

Disruption of the ability to decompose hydrogen peroxide appears to lead to a

net increase in the production of cellular ROS

This balance in the production and degradation of hydrogen peroxide and

other ROS prevents accumulation of the potentially harmful reactants and

accompanying downstream effects on cellular constituents In contrast,

perturbations in peroxisomal catalase levels, as seen in certain pathological

situations (Takahara, 1952; Yildirim et al., 2004) or known to accompany aging

(Legakis et al., 2002; Rao et al., 1990; Beier et al., 1993; Brown-Borg et al.,

2000), result in oxidative damage to the cell Hence, the peroxisome may

participate in the aging process through global effects on membrane function due

to its own compromised function, as well as through the generation of ROS

What role such oxidative stress plays in the initiation or progression of disease or

in the process of aging is only beginning to be explored

Investigators working with a variety of model systems have documented a

clear relationship among cellular and peroxisomal catalase levels, ROS,

oxidative stress, and aging Rodent liver has been the focus of many of these

early studies What has emerged is that mice and rats consistently display an

age-related decline in cellular hepatic catalase levels (for a review see Youssef

and Badr, 2005)

Interestingly, calorically restricted animals exhibit increased catalase

expression and activity (Beier et al.,1993; Rao et al., 1990) Studies by

Brown-Borg and Rakoczy (Brown-Brown-Borg et al., 2000) showed that in the long-lived Ames

dwarf mouse catalase concentrations are generally elevated In contrast, they

found that in short-lived growth hormone- (over-expressing) transgenic mice,

catalase amounts are reduced Furthermore, mice with little (hypocatalasemic)

or no catalase (acatalasemic) are at increased risk for the formation of tumors

(Ishii et al., 1996; Ito et al., 1986) and perhaps other pathologies

Co-overexpression of catalase and superoxide disumutase in the fruit-fly

Drosophila melanogaster increases lifespan (Orr et al., 1994) – as do catalase/superoxide dismutase-mimetics in the nematode, Caenorhabditis

elegans (Melov et al., 2000) Catalase in these worms exists in several forms

-including cytosolic (CTL1) and peroxisomal (CTL2) versions More recent

studies demonstrate that genetic disruption of the cytosolic form is without effect;

however, disruption of peroxisomal form results in a progeric (i.e prematurely

aged) phenotype (Petriv and Rachubinski, 2004) In the latter case, the

organism exhibits decreased egg-laying capacity (interpreted as a malfunctioning

of the organism’s developmental program), altered peroxisome morphology

perhaps secondary to ROS accumulation, and a shortened lifespan In the same

work, the authors also show that in the yeast Saccharomyces cerevisiae,

knocking-out peroxisomal catalase (T), but not cytosolic catalase (A),

compromises viability As the authors succinctly summarize, “a shorter lifespan

may be a general consequence of a lack of peroxisomal catalase” (Petriv and

Rachubinski, 2004)

Humans lacking cellular catalase were first identified by Takahara and

colleagues (Takahara, 1952; Takarhara et al., 1948) They described patients as

suffering from oral ulcerations and accompanying gangrene (caused either by i

hydrogen peroxide applied as an antiseptic; ii infection by hydrogen

peroxide-producing bacteria; or iii hydrogen peroxide produced by inflammatory

phagocytic cells) but were otherwise asymptomatic This is consistent with

recent observations in catalase-knock-out mice, where the animals show no

major developmental abnormalities although they do possess certain

tissue-specific deficiencies in ROS metabolism (Ho et al., 2004) These and other

observations (reviewed in Eaton and Ma, 1995) have led to acatalasemia having

long been considered a relatively “benign” disease (however, see also Goth and

Eaton, 2000)

As only a few cases of true human acatalasemia have been described –

perhaps a more relevant condition is hypocatalasemia, where catalase levels are

reduced from 25 to 80% of normal Here, the number of afflicted individuals is far

greater – estimates are that over 1 in 500 are affected to varying extents (Eaton

and Ma, 1995; Goth and Vitai, 2003) Once again, obvious developmental

defects are absent; however, pioneering work by Goth and colleagues and others

(Yildirim et al., 2004) has shown an accelerated onset of age-related diseases

including type 2 diabetes, atherosclerosis, cataracts, as well as macular

degeneration, tumors, and anemia in these patients The suggestion is that

oxidant-induced cellular damage accumulates over years – facilitating the

development and progression of age-related disease

Human Peroxisomal Enzyme

PTS1 (Carboxy-terminal 12 residues)

PTS1 Strength

H2O2-generating enzymes

H2O2-degrading enzymes

Table 1: Human peroxisomal hydrogen peroxide-generating and

–degrading enzymes and their relative PTS1 strength (according to PTS1

Predictor).

CHAPTER 2

MATERIALS AND METHODS

Cell culture

CHO and A431 cells were obtained from American Type Culture

Collection (ATCC) (Rockville, MD) Early passage IMR90 human diploid

fibroblasts were obtained from the National Institutes of Aging, Aging Cell

Repository/Coriell Institute for Medical Research (Camden, NJ) Hs68, WI38 and

Hs27 human diploid fibroblasts were purchased from ATCC (Manassas, VA)

Hypocatalasemic fibroblasts were obtained from Coriell Cell Repositories and

called “acatalasemic” by the supplier All cells were cultured in DMEM

supplemented with 10% fetal bovine serum or 10% serum supreme (Gibco,

Grand Island, NY), penicillin, and streptomycin The cells were maintained at

37oC in humidified incubators supplemented with 5% CO2 To achieve higher

passage levels for IMR90, Hs68, WI38 and Hs27 cells, the cells were expanded

through sub-cultivation Late passage cells were confirmed to be at or near

replicative senescence by staining for senescence-associated β-galactosidase.Where indicated, cells were grown on glass coverslips pretreated with ProNectin

F (Biosource International, Camarillo, CA)

Reagents and antibodies

Firefly (Photinus pyralis) luciferase and human erythrocyte catalase were

purchased from Sigma (St Louis, MO) Three recombinant hexahistidine- (His)6

-tagged human catalase proteins with different carboxy-terminal PTS1 sequences

were expressed in bacteria and purified with nickel-nitriloacetic acid agarose as

per manufacturer’s protocols (Qiagen, Valencia, CA) Glutathione

S-transferase-(GST)-tagged human Pex5p was expressed and purified as described (19)

Rabbit polyclonal antibodies directed against human catalase were obtained from

Calbiochem (La Jolla, CA) or custom ordered from Sigma-Genosys (St Louis,

MO); rabbit polyclonal antibodies directed against PMP70 were purchased from

Affinity Bioreagents (Golden, CO); and mouse monoclonal antibodies to the

Xpress™ epitope were procured from Invitrogen (Carlsbad, CA) Rabbit

polyclonal antibodies to GST were a kind gift from Dr Marc Fransen

Fluorescently conjugated secondary antibodies were purchased from Jackson

ImmunoResearch Laboratories, Inc (West Grove, PA) or KPL (Gaithersburg,

MD) 2,7-dichlorofluorescin diacetate was purchased from Acros Organics

(Fisher Scientific)

5,5’,6,6’-tetrachloro-1,1’,3,3’-tetraethylbenzimidazolyl-carbocyanine iodide (JC-1) and MitoTracker Red CM-H2Xros were purchased

from Molecular Probes (Eugene, OR) The catalase inhibitor

3-Amino-1,2,4-triazole and all other reagents were obtained from standard sources