Kapitel 1. Introduction


The DNA of most organisms has in addition to the four major bases; A, T, C and G, one or more minor bases. These minor bases are created by a modification of the DNA by post-replicational methylation .


Two major classes of DNA modifications exist in nature. One class methylates a ring carbon in position 5 to form C5-methylcytosine (5mC). The second class methylates exocyclic nitrogens and forms either N4-methylcytosine or N6-methyladenine.

In prokaryotes, the major role of DNA methylation is to protect host DNA against degradation by the cognate restriction enzymes. However, Escherichia coli also has two methylase systems that are not associated with restriction-modification systems. The dam gene encodes a methylase that forms 6mA and the dcm gene encodes a methylase that forms 5mC. Methylation has been shown not to be essential in Escherichia coli , but provides useful functions. The classic function is the DNA modification in a restriction-modification system. Foreign DNA from a species or strain with a different or no methylation pattern is digested by the restriction enzyme of the host restriction-modification system. Some strains of Escherichia coli (e.g. K-12) are able to restrict methylated DNA protecting it against infections by phages which escape the conventional host restriction by methylating their DNA while they replicate in the host strains. Therefore, DNA methylation, along with restriction enzymes, function to reduce the efficiency of gene transfer between unrelated species. The dam methyltransferase (methylase or MTase) is involved in a wide variety of cellular functions such as postreplicative mismatch repair, control of Escherichia coli chromosome replication and chromosome segregation, control of plasmid replication, transposition, gene expression and control of the initiation of phage P1 DNA packaging .

In eukaryotes there is no evidence for a restriction enzyme function. All potential sites are methylated in bacteria but many potentially methylatable sites are unmethylated in eukaryotes. Unmethylated sites would be sensitive to restriction if restriction modification systems similar to those in bacteria were in function. Therefore, it is likely that methylation serves other functions in eukaryotes. In these organisms, DNA methylation has been implicated in the control of several cellular processes, mainly in gene regulation and thus in processes including differentiation and development. C5 cytosine MTases can be found in both prokaryotes and eukaryotes, whereas the exocyclic MTases have been mainly isolated from prokaryotes. In general, the 5mC modification is nearly universal among eukaryotes with genomes greater than 108 bp (mammals and plants), but rare among eukaryotes with smaller genomes (yeast, flies and nematodes) . Furthermore, two C5 cytosine MTases (masc1 and masc2) with similarities to the mammalian MTases have been cloned from the filamentous fungus Ascobolus immersus . DNA methylation in Ascobolus is essential for the completion of the developmental processes underlying sexual reproduction. To date, 5mC has not been detected in the DNA of Schizosaccharomyces pombe , Saccharomyces cerevisiae , Caenorhabditis elegans and Drosophila melanogaster . Recently a MTase related Drosophila protein has been cloned, DmMT2, which is similar to that of vertebrate Dnmt2 enzymes. This work focuses on the regulation of DNA methylation by the C5 cytosine MTases in mammals.



The enzyme responsible for the methylation process is DNA MTase, which transfers a methyl group from S-adenosyl-L-methionine (SAM or AdoMet) to the 5-carbon in the pyrimidine ring of cytosine (Fig.1.1) . This process involves the formation of a transient covalent complex between the protein and the pyrimidine being modified (see Fig. 1.1). A cysteine thiol of the enzyme attacks the 6-carbon of cytosine and forms a covalent DNA-protein intermediate. The addition of the cysteine thiol activates the 5-carbon allowing transfer of the methyl group from SAM and release of S-adenosyl-L-homocysteine (SAH or AdoHcy).

Fig.1.1. Reaction pathway for C5-cytosine methyltransferase based on the mechanism proposed by Wu and Santi for thymidylate synthase and tRNA-(uracil-5)methyltransferase. The attack on carbons 5 and 6 are shown in the figure. The methyl group is transferred to carbon-5 of cytosine, and the hydrogen located on carbon-5 is released as H+. During this reaction SAM is coverted to SAH. A base (B:) that abstracts a proton from carbon-5 is needed for the elimination step .

The fact that DNA contains a set of modifications that is not encoded in the genetic sequence but is related to changes in gene activity, classifies DNA methylation as an epigenetic process. The recognition sequence for the mammalian DNA MTase has also been identified and is highly specific with almost all cytosine methylations occurring on 5‘-C-p-G-3‘ (CpG) and over half of CpG dinucleotides being methylated in the mammalian genome . It has also been found that cytosine methylation can occur at CpXpG trinucleotides in higher plants and also in mammals but at lower frequency .


Several aspects of the catalytic mechanism of DNA methylation have been clarified mainly for prokaryotic MTases . MTase HhaI is a C5 cytosine MTase originally isolated from the bacterium Haemophilus haemolyticus . The gene for this MTase has been cloned and sequenced and the overexpression of the protein in Escherichia coli followed by large-scale purification has allowed its crystallization . The enzyme is folded into two domains (Fig.1.2): a larger catalytic domain and a smaller DNA recognition domain which is involved in flipping the target cytosine out and


positioning it into the catalytic pocket of the enzyme (Fig.1.2).

Fig.1.2. Graphic representation of the complex of the MTase HhaI covalently bound to a 13-mer DNA duplex containing its recognition sequence. The end product of the reaction, SAH, is also present (yellow). The protein is shown in brown, the DNA backbone in magenta, the DNA bases in green, and the active-site loop and the recognition loops are represented in white (right panel). The left panel is a view looking down the DNA helix axis, in which the DNA can be seen to lie between the large domain (on the right) and the small domain (on the left). On the right panel a side view from the minor groove is shown. (Fig. from ).


The methylation of DNA helps to regulate gene expression and mistakes in this regulation can have drastic consequences leading to birth defects and cancer. Three possible ways by which DNA methylation can affect gene expression are:

Direct effect: Methylated CpG residues can directly interfere with the binding of specific transcription factors to DNA. Several transcription factors (AP-2, c-Myc/Myn, E2F and NF_B) bind to DNA sequences that include CpG sites and have been shown to be sensitive to methylation at these sites (TF1 in Fig. 1.3). Other transcription factors, however, are not sensitive to methylation (Sp1, CTF and YY1) (TF2 in Fig. 1.3).

Direct binding of specific factors: The direct binding of specific factors to methylated DNA mediates repression (Fig. 1.3). Two such factors, MeCP1 (complex of several subunits) and MeCP2, have been identified and shown to bind to methylated CpG in any sequence context and thereby preventing transcription factor binding. MeCP1 binds to DNA containing multiple symmetrically methylated CpG sites . MeCP2 is more abundant than MeCP1 and is able to bind to DNA that is asymmetrically methylated, with just a single methyl-CpG. In addition, the distribution of MeCP2 on the chromosome parallels that of methyl-CpG . Experiments on mice with a disrupted MeCP2 gene have shown that MeCP2, like DNA MTase is dispensable in embryonic stem cells but essential for embryonic development .


Chromatin structure alteration: It has been shown that methylated DNA affects the positioning of nucleosomes and influences the sensitivity of DNA to DNAase I . Further experiments using microinjection of methylated and non-methylated templates into nuclei have shown that methylation inhibits expression only after chromatin is assembled . These results support the view that methylation induces a change in conformation of chromatin to an inactive state . However, only recently it has been shown how cytosine methylation affect the structure of chromatin (Fig. 1.3) by its interaction with MeCP2 . MeCP2 binds to chromosomes in a methylation dependent manner and mediates transcriptional repression. MeCP2 contains a transcriptional repression domain which associates with a transcriptional repression complex containing mSin3a and histone deacetylase . Two possible explanations of how this interaction works have been proposed. In the first one deacetylation of lysine amino groups might allow interactions between the amino terminal histone tails and the DNA phosphate backbone. The second possibility is that deacetylation might favour interactions between adjacent nucleosomes (Fig.1.4).

Fig.1.3. A model to explain the effects of DNA methylation on transcription. The relevant parameters are proximity and methylation of CpG sites. The MeCP molecules are shown as ovals, the CpG sites as filled circles (methylated) or half circles (non-methylated) circles, the arrows represent transcription from the promoter, TF1 and 2 represent transcription factors 1 and 2, sensitive or insensitive to methylation, respectively and HDAC represents the histone deacetylase. Active transcription is indicated with an arrow.

Fig.1.4. Methylation and histone deacetylation. MeCP2, a protein that binds to methylated DNA, exists in a complex with histone deacetylase and mSin3a. Nucleosomes (as several connecting disks) with their histone tails (as thin lines coming out of the nucleosomes) are shown. The two mechanisms in which histone deacetylation changes the structure of the nucleosomes are also depicted. The deacetylation by the MeCP2 complex (MeCP2, mSin3a and histone deacetylase) of the lysine amino groups on the histone tails might allow the interaction of the histone tails and the DNA backbone or might lead to compaction of the chromatin by allowing the interactions between the nucleosomes (Figure from (.




A distinct change in the degree of methylation of the genome has been observed during gametogenesis and early mammalian embryogenesis. While many genes are highly methylated in sperm DNA , genes are often less methylated in oocytes (Fig. 1.5). During development, the DNA of the extraembryonal membranes (yolk sac and placenta) becomes dramatically demethylated, while the DNA of fetal tissues is subjected to a de novo methylation process after implantation .

DNA methylation in the early embryo. Dynamic changes in the methylation pattern of specific genes in the early embryo were observed. CpG islands are regions (500-2000 bases pairs) of unmethylated DNA with a high frequency of CpG dinucleotides and have been found to be associated with 5‘ ends of housekeeping genes and of some tissue specific genes . The CpG islands in housekeeping genes are unmethylated throughout development. However, the CpG islands in tissue-specific genes are heavily methylated in sperm and less methylated in the mature oocytes. Between the 4 to 16 cell stages this methylation is erased. The unmethylated state of these genes persists throughout the entire preimplantation period .

ii) Methylation during gametogenesis. Primordial germ cells (PGCs) emerge from the generally unmethylated epiblast before tissue specific de novo methylation takes place in the epiblast lineage, which gives rise to the embryo proper (somatic lineage). The trophoblast derived lineage (placenta and yolk sac) is substantially demethylated. The PGCs then migrate and colonize the developing genital ridges by 10-11 dpc . At 12.5 dpc before sexual differentiation of the primordial germ cells begins, tissue specific and imprinted genes are unmethylated . This hypomethylated state prevails until sexual differentiation of the PGCs takes place


and de novo methylation occurs in the cells populating the gonads of both sexes, forming a bimodal pattern of methylation where CpG islands remain unmethylated and non-island sequences are methylated .

Fig.1.5. Methylation during gametogenesis and early development in mammals. The overall methylation increases during gametogenesis being higher in the male gametes. During cleavage, the genome undergoes global demethylation and after the blastocyst stage the overall methylation level remains low in the germ line and increases in the somatic lineage. There are therefore two waves of de novo methylation (grey boxes), one is during gametogenesis and the other is after implantation when tissue specific patterns are being established (Fig. modified from T. H. Bestor).

This work is concerned with the regulation of DNA methylation during gametogenesis and myogenesis and focusses on the identification and characterization of alternative Dnmt1 isoforms.


Global changes in methylation levels as well as modification of methylation patterns of individual genes has been observed in a variety of tumors . However, the role of methylation in carcinogenesis is still not clear since both DNA


hypomethylation and hypermethylation have been associated with cancer .

Epigenetic changes induced by the (cytosine-5)-DNA MTase: The mRNA level and enzyme activity of MTase are higher in many tumor cells than in normal cells . However, the generally higher methyltransferase activity does not lead to an overall increase but rather to a decrease of total genomic 5mC content . When the methylation status of various genes is measured in tumor cells, both hypermethylation as well as hypomethylation become apparent . It appears that the pattern of methylation of specific genes is altered in tumor cells and normally unmethylated CpG islands may become hypermethylated .

Hypomethylation. Reduced levels of global DNA methylation have been found in a variety of malignancies. Growth inducing genes such as oncogenes can become overexpressed as a result of hypomethylation. In particular, demethylation and overexpression of the the c-fos, the c-myc and the c-H-ras proto-oncogenes are known to be involved in hepatocarcinogenesis induced by conditions such as methyl donor starvation .

Hypermethylation. There have been reports of regional increases in DNA methylation levels and regional hotspots for hypermethylation on chromosomes 3p, 11p and 17 p in a variety of human tumors have been shown . Genes involved in growth arrest and terminal differentiation such as the MyoD1 gene and tumor suppressor genes such as pRb , bcr-abl , pVHL , pWT , estrogen receptor , p57KIP2 , MDG1 and p16 are often found inactivated by either mutation or hypermethylation in tumor cells. Hypermethylation of tumor suppressor genes may lead to gene inactivation and result in a selective growth advantage of affected cells.

ii) Genetic changes at the target site of (cytosine-5) DNA MTase: Cytosine methylation is also responsible for the induction of a high percentage of disease-causing point mutations in tumor suppressor genes in somatic and germline cells. Mutations occur by spontaneous deamination of 5mC . Both cytosine (C) and 5mC deaminate in single- and double-stranded DNA to form uracil and thymine, respectively . However, it is more difficult for the cell to correct the resulting T:G mismatch since thymine, unlike uracil, is a normally occurring DNA base. In aqueous solution at 37oC, a 4- to 9- fold higher deamination frequency of 5-mC relative to C in single stranded DNA and in double stranded DNA a ~ 2-fold higher deamination frequency was measured for 5mC when compared to C . A sophisticated DNA repair apparatus has evolved in mammals which catalyzes the selective repair of the T-G mismatches (resulting from 5mC deamination) back to C-G pairs . The remaining unrepaired C to T transitions cause many gene mutations in humans leading to hereditary diseases and cancer .


Vertebrates and other organisms with DNA methylation show a depletion in the frequency of occurrence of the CpG dinucleotide, which is the predominant site at which 5mC is found. This depletion is thought to be due to increased mutagenesis as described above.

The target cytosines of MTase in prokaryotes are mutated to thymine with an increased frequency when compared with non target cytosines. In eukaryotes, the enhanced mutability of 5mC contributes significantly to carcinogenesis and inherited disease . In humans, several genes such as p53 and p16 are frequently mutated at CpGs in tumor cells and 30% of all inherited mutations are believed to occur by mutations at CpG sites in the germ line .

Metabolism of the cofactor SAM and its involvement in carcinogenesis: Several studies indicate that aberration of the metabolism of the cofactor SAM may also play an important role in carcinogenesis . Aberrations of the SAM metabolism may not only affect DNA methylation but also other methyl-transfer reactions (e.g. synthesis of spermine and spermidine). Insufficient supply of methionine, folate and choline leads to hypomethylation of liver DNA, increased expression of the c-H-ras, c-jun and c-myc genes and to the generation of liver tumors in rats . In humans, methyldonor deficiency is correlated with an increased risk for liver and colon tumors .


Imprinted genes are those genes whose expression is determined by their parental origin, and it has been reported that cancer cells sometimes show loss of imprinting which might be explained by changes in DNA methylation that are known to accompany tumorigenesis. Other diseases that have been reported to be associated with imprinting include Prader-Willi , Angelman and Rett syndromes.

Furthermore, X-inactivation in female mammals is an event associated with a methylation change in CpG islands. The CpG islands within the promoter regions of the X-linked housekeeping genes become highly methylated at or soon after gene inactivation. The methylation of CpG islands on inactive X-chromosome has important medical implications with regard to the fragile X-chromosome, and it has been found that a CpG island near to the breakage site is methylated in most of the affected people . In contrast to X-linked genes, the CpG islands associated with autosomal genes remain unmethylated, with some exceptions. These exceptions include Alu elements and L1 retrotransposons which are frequently methylated in the human genome. Although methylation is inherently mutagenic and has led to genomic alterations, it is tolerated possibly because of its contribution to the regulation of gene expression and its necessity during embryonic development.



DNA methylation patterns are known to be maintained over multiple cell generations, however, are not the same in different cell types. The particular patterns are established during early development and can change during development and disease. A model to explain the maintenance and change of cell specific patterns of methylation has been presented . A maintenance MTase was postulated that has a strong preference for hemimethylated sites and thus maintains a given methylation pattern after DNA replication. In addition, multiple de novo MTases were postulated, that are expressed and active at specific stages during development.

Maintenance of methylation. Replication results in the generation of a hemimethylated double-stranded DNA, composed of a methylated parental strand and unmethylated nascent strand. Only this hemimethylated DNA serves as a substrate for the maintenance MTase (Fig. 1.6). This model proposes that DNA methylation patterns are restored after replication because DNA MTase is very efficient in methylating hemimethylated CpG sites (maintenance methylation) but relatively inefficient in methylating nonmethylated substrates.

ii) De novo methylation. DNA MTases can also methylate certain unmethylated CpG sites that are not in a hemimethylated configuration, a process referred to as de novo methylation (Fig.1.6). De novo methylation occurs extensively during gametogenesis and after implantation (Fig1.5) . Several studies have shown de novo methylated cytosines in genomic regions containing preexisting methylated cytosines (methylation spreading), such as it occurs at sites where viral DNA integrates in the genome . Since cytosine methylation can directly and indirectly affect the DNA binding of certain transcriptional regulatory factors (Fig. 1.3), the introduction of additional methylated cytosines within gene regulatory sequences may influence gene expression . This spreading of cytosine methylation in gene regulatory sequences has been implicated in the gene silencing characteristic of fragile X syndrome , cellular senescence , and X-chromosome inactivation .

Demethylation. Demethylation may occur throughout development but is more pronounced during preimplantation development. Demethylation processes have also been shown during gene activation in resting cells . Two possible pathways have been described for the loss of DNA methylation:

Passive demethylation: This occurs by a failure to methylate newly replicated DNA (Fig. 1.6). This type of `demodification´ takes place after inactivation of cellular MTases with 5-azacytidine . Furthermore, it has been shown that a correlation exists between the overall decrease in the genomic methylation level and the active retention of Dnmt1 in the cytoplasm from the oocyte to the blastocyst stage .


Active demethylation: It has been shown that methyl groups are removed when DNA is not even undergoing replication, such as that seen on inducing the vitellogenin gene in the chick liver or the globin gene in stimulated erythroleukemia cells . Active demethylation has been described in transiently transfected myoblasts , in postmeiotic spermatocytes and in preimplantation mouse embryos . This active process of demethylation could involve 5mC DNA glycosylase activity which has been detected in HeLa nuclear extracts . Other studies have also shown that the demethylation process is at least partly mediated by RNA molecules and that demethylation takes place through the removal of DNA nucleotides and their conversion to an RNase-sensitive form . Recently, a human cDNA that encodes a DNA demethylase (dMTase) activity that can catalyze the removal of a methyl residue from 5mC and its release as methanol has been cloned .

Fig.1.6. Establishment and maintenance of DNA methylation patterns. The establishment of DNA methylation patterns involves a number of processes. Left, de novo methylation: when a potential methylatable site (indicated as a circle) that is not methylated on both strands of the DNA undergoes methylation (indicated as M outside the circles). Center, maintenance methylation: the DNA molecule is composed of a parental methylated strand (thick line) and a nascent unmethylated strand (thin line). Maintenance DNA MTases add methyl groups to the nascent unmethylated cytosines that reside opposite to methylated sites on the parental strand and thus the pattern of methylation is replicated. Right, demethylation: when a methyl group (M) is removed from methylated CpG sites.



The first mammalian DNA MTase reported was the Dnmt1. The Dnmt1 cDNA from mouse has been cloned and its nucleotide sequence determined . Partial loss of function mutation of the Dnmt1 gene results in embryonic lethality showing that Dnmt1 is essential for mouse development . Dnmt1 contains a C-terminal (570 amino acids) domain joined to a much larger N-terminal (1051 amino acids) domain by a region of 13 alternating Gly and Lys residues; this region is encoded in the cDNA by 39 consecutive purine residues. This hydrophilic sequence is likely to form a flexible link between the N- and C-terminal domains. The C-terminal domain contains ten motifs (I-X) which are also present in most prokaryotes 5mC MTases. Motif I (PheXGlyXGly) was proposed to be part of the cofactor (SAM) binding site. This assignment was based on the presence of this consensus sequence in a wide variety of SAM dependent MTases, including N6-adenine, N4-cytosine, RNA and protein MTases . Motif IV contains a dipeptide (ProCys) known to take part in the catalytic reaction. In the bacterial MTase HhaI this reaction involves the formation of a covalent intermediate between the cysteine from motif IV and the target cytosine (see section 1.1.2). The region between motifs VIII (GlnXArgXArg) and IX (ArgGlu) contains the target recognition domain (TRD) which determines the DNA sequence specificity as well as the base to be methylated within the target sequence . Several functions have also been localized to the N-terminal domain. The N-terminal domain contains a cysteine-rich region that is similar to metal-binding sites found in several regulatory proteins and it was shown to bind Zn . The location of some of the relevant features in Dnmt1 are shown in Fig.1.7.

Fig.1.7. Structure of the mouse Dnmt1 protein. The Dnmt1 protein contains an N-terminal domain of 1051 amino-acids which is joined to the 570 amino-acid C-terminal domain by a run of 13 alternating lysyl and glycyl residues, shown in the figure as a thick connecting line. Binding site for PCNA: proliferating cell nuclear antigen ; NLS: nuclear localization sequence ; Ser: phosphorylation site at serine 514 ; RFTS: replication foci targeting sequence ; Zn binding: in vitro zinc binding (Cys-rich) region ; PBHD: polybromo-1 protein homologous domain ; ProCys: dipeptide part of the catalytic site ; TRD: target recognition domain ; filled boxes: highly conserved sequence motifs (I, II, IV, V, VI, VII, IX, X); grey boxes: less conserved sequence motifs (III, V, VII) between the mammalian and the bacterial MTases.

A major phosphorylation site of the Dnmt1 has been reported , indicating that the murine Dnmt1 is posttranslationally modified. The site lies in a region of the protein that is required for targeting of the MTase to the replication foci during S phase of the cell cycle . Furthermore, it has been proposed that several factors might be involved in maintaining the methylation patterns in the


mammalian genome by interacting with Dnmt1. It has been reported that Dnmt1 binds to the proliferating cell nuclear antigen (PCNA) in vitro and this could mediate the association of Dnmt1 with the DNA replication machinery in vivo.

Proteolytic cleavage between the N- and C-terminal domains stimulates the de novo methylation activity of the murine Dnmt1 in vitro, suggesting that this enzyme may also be responsible for de novo methylation occurring in vivo. This was proposed to be the consequence of separation of the catalytic domain from the inhibitory N-terminal domain. This complex structure of Dnmt1 raises the possibility that Dnmt1 might carry out both, maintenance and de novo, methylation and these activities might be controlled by the expression of alternative isoforms or by posttranslational modification of the enzyme. Another alternative but not mutually exclusive possibility is that an independently encoded and developmentally regulated DNA MTase is responsible for de novo methylation. A null allele of the Dnmt1 gene showed that embryonic stem (ES) cells homozygous for the null mutation still had very low levels of genomic DNA methylation. Infection of mutant ES cells with the Moloney murine leukaemia virus (MoMuLV) resulted in de novo methylation of the integrated provirus DNA with a similar rate to that seen in wild type ES cells, suggesting that de novo methylation is not totally impaired by loss of Dnmt1. These results provide evidence that an independently encoded DNA MTase is expressed in early embryonic cells. Recently three additional mammalian MTases genes were cloned, Dnmt2 , Dnmt3a and Dnmt3b . Dnmt3a and Dnmt3b are essential for de novo methylation and for mouse development and Dnmt3b is required for methylation of centromeric minor satellite repeats . However, the expression of Dnmt3a and Dnmt3b has not yet been characterized.

Evidence for the presence in some tissues of alternative Dnmt1 isoforms first came from the identification of a transcript longer than the one present in somatic cells during spermatogenesis . This longer transcript could be due to additional sequences at the 5‘ end of the Dnmt1 gene or to new alternatively spliced Dnmt1 isoforms. Thus, different Dnmt1 isoforms with different enzymatic properties might be expressed in specific tissues at particular stages and might play a role in the regulation of methylation during gametogenesis and tissue differentiation. The focus of this study was, therefore, to search for alternative Dnmt1 isoforms and to study their expression during development, as well as to try to elucidate the Dnmt1 exon-intron boundaries and its relation with the protein structure.


The purpose of this section is to describe the principles and practical details of the techniques used in this study. Specific details of the methods used are described in the respective results section.

© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.

DiML DTD Version 2.0
Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML - Version erstellt am:
Tue Sep 17 12:11:03 2002