UNIVERSITY OF CINCINNATI
Date:______
I, ______, hereby submit this work as part of the requirements for the degree of: in:
It is entitled:
This work and its defense approved by:
Chair: ______
CHARACTERIZATION OF TANKYRASE STRUCTURE & FUNCTION; EVIDENCE FOR A ROLE AS A MASTER SCAFFOLDING PROTEIN
A dissertation submitted to the
Division of Research and Advanced Studies of the University of Cincinnati
in partial fulfillment of the requirements for the degree of
DOCTORATE OF PHILOSOPHY (Ph.D.)
in the Department of Molecular Genetics, Biochemistry and Microbiology
College of Medicine
2004
by
Manu De Rycker
B.S., University of Antwerp, 1995 M.S., University of Gent, 1998
Committee Chair: Carolyn Price, Ph.D. ABSTRACT
Tankyrases are novel poly(ADP-ribose) polymerases that have SAM and ankyrin protein- interaction domains. They are found at telomeres, centrosomes, nuclear pores and the Golgi- apparatus, and participate in telomere length regulation and resolution of sister chromatid association. Their other function(s) are unknown and it has been difficult to envision a common role at such diverse cellular locations. We isolated the chicken tankyrase homologs and examined their interaction partners, subcellular location and domain functions to learn more about their mode of action. Cross-species sequence comparison indicated that tankyrase domain structure is highly conserved and supports division of the ankyrin domain into five subdomains, each separated by a highly conserved LLEAAR/K motif. GST-pull down experiments demonstrated that the ankyrin domains of both proteins interact with chicken
TRF1. Analysis of total cellular and nuclear proteins showed that cells contain approximately twice as much tankyrase 1 as tankyrase 2. Although ≥90% of each protein is cytoplasmic, both tankyrase 1 and 2 were also nuclear. This nuclear location, together with its ability to interact with TRF1, point to a telomeric function for tankyrase 2. This work shows that tankyrases polymerize through their SAM domain to assemble large protein complexes. In vitro polymerization is reversible but still allows interaction with ankyrin-domain binding proteins. Polymerization also occurs in vivo, with SAM-dependent association of overexpressed tankyrase leading to the formation of large tankyrase-containing vesicles, disruption of Golgi structure and inhibition of apical secretion. Finally, tankyrase polymers are dissociated efficiently by poly(ADP-ribosy)lation. This disassembly is prevented by mutation of the PARP domain. Our findings indicate that tankyase 1 promotes both assembly and disassembly of large protein complexes. Thus, tankyrases appear to be master scaffolding proteins that regulate the formation of dynamic protein networks at different cellular locations. This implies a common scaffolding function for tankyrases at each location with specific tankyrase interaction partners conferring location-specific roles to each network, such as telomere compaction or regulation of vesicle trafficking.
ACKNOWLEDGEMENTS
First of all I would like to thank my advisor, Carolyn Price. She has not only given me the freedom to venture far away from the telomeres, but has also joined me in my excitement about the project. She has been a great guide in many ways, and I thank her for her advice, trust and support and for being a great friend.
I would also like to thank my committee members, Yoli Sanchez, Joanna Groden, Jim
Stringer and Jun Ma, for providing direction and thought-provoking questions during and outside our committee meetings.
Thanks to all the members of the Price lab, past and present, for providing a great atmosphere to work in. Special thanks to Angela and Fred, for putting up with me (and my radio) and for being great friends.
I thank the Albert J. Ryan foundation for their support.
Many friends have made these five years fun and they will be dearly missed. Special thanks to Robyn and Tim for all the great times we had together.
It is hard to express how much I am indebted to my parents, Paul and Mickie, and my sister, Kim. They have supported me in countless ways during these many years of study, and
I can’t wait to spend more time with them the coming years.
Finally I would like to thank the most important person in my life. Words are inadequate to express how much it means to me that Sandra stayed here with me for all these years. She moved almost 7000 kilometers from home to be with me, to support me and to enjoy life together. Knowing that I would see her every day has been my single greatest driving-force. TABLE OF CONTENTS
1. INTRODUCTION ...... 5
1.1. Tankyrase domain structure...... 6
1.1.1. Ankyrin repeat domain ...... 6
1.1.2. The Sterile-Alpha-Motif (SAM)...... 9
1.1.3. The poly-(ADP-ribose) polymerase (PARP) domain...... 11
1.1.4. Conclusions...... 14
1.2. Tankyrase functions...... 15
1.2.1. Function at telomeres...... 15
1.2.2. Function in vesicle trafficking and insulin signaling...... 17
1.2.3. Other functions...... 20
2. GOAL OF THESIS...... 22
2.1. Hypothesis...... 22
2.2. Overview...... 22
3. MATERIALS AND METHODS...... 24
3.1. Cell lines...... 24
3.2. Isolation of chicken tankyrase genes ...... 24
3.3. Genomic library screen and knock-out construct ...... 24
3.4. Electroporations ...... 26
3.5. Analysis of DT40 genomic DNA by Southern blot...... 27
3.6. Antibodies and protein detection ...... 28
3.7. Expression constructs...... 30
1 3.8. Protein expression and purification ...... 32
3.9. GST pull-downs...... 35
3.10. Radioligand binding assays ...... 36
3.11. Yeast two-hybrid analysis of SAM self-association...... 36
3.12. Gel filtration...... 37
3.13. Electron microscopy ...... 38
3.14. Pelleting and dilution assays...... 39
3.15. Poly-(ADP-ribose) binding assay ...... 40
3.16. Tankyrase activity assay ...... 40
3.17. Co-immunoprecipitations ...... 40
3.18. Immunofluorescence...... 41
3.19. Yeast two-hybrid screen ...... 42
4. CHARACTERIZATION OF THE CHICKEN TANKYRASES...... 44
4.1. Summary...... 44
4.2. Introduction...... 45
4.3. Results...... 47
4.3.1. Identification of chicken tankyrase genes...... 47
4.3.2. Conservation of tankyrase proteins...... 50
4.3.3. Partitioning of the ankyrin repeats into five subdomains ...... 52
4.3.4. Nucleotide binding properties of the LLEAAR/K motif ...... 54
4.3.5. Chicken tankyrases interact with cTRF1 ...... 55
4.3.6. Subcellular distribution of tankyrase 1 and 2 ...... 57
2 4.3.7. Creation of a tankyrase 2 knock-out cell line ...... 60
4.3.8. Co-regulation of tankyrase 1 and 2...... 63
4.4. Discussion...... 64
5. TANKYRASE POLYMERIZATION IS CONTROLLED BY ITS SAM AND PARP DOMAINS...... 67
5.1. Summary...... 67
5.2. Introduction...... 68
5.3. Results...... 70
5.3.1. The tankyrase SAM-domain mediates polymerization in vitro...... 70
5.3.2. The SAM domains of tankyrase 1 and 2 copolymerize in vitro ...... 75
5.3.3. SAM domain mutagenesis ...... 77
5.3.4. Self-association of full-length tankyrase ...... 78
5.3.5. In vivo formation of tankyrase complexes...... 81
5.3.6. Tankyrase self-association in vivo is mediated by the SAM domain ...... 83
5.3.7. Overexpressed tankyrase localizes to vesicle-like structures ...... 87
5.3.8. Involvement of tankyrase in protein secretion...... 91
5.3.9. Disruption of tankyrase polymers by poly-(ADP-ribosylation) ...... 93
5.4. Discussion...... 97
5.4.1. SAM domain polymerization...... 98
5.4.2. Tankyrase 1 and tankyrase 2 redundancy ...... 99
5.4.3. Tankyrase as a scaffolding molecule ...... 100
5.4.4. Functions of tankyrase scaffolds...... 102
6. YEAST TWO-HYBRID SCREEN...... 104
3 6.1. Introduction...... 104
6.2. Results...... 107
6.2.1. PARP mutagenesis and controls...... 107
6.2.2. Library screen...... 108
6.2.3. Secondary screen; regulators of tankyrase activity...... 110
6.3. Discussion...... 112
7. FINAL DISCUSSION AND FUTURE DIRECTIONS ...... 114
7.1. Tankyrase redundancy...... 114
7.2. General role for tankyrases ...... 116
7.2.1. Tankyrase and the secretory pathway...... 117
7.2.2. Tankyrase in mitosis ...... 120
7.2.3. Tankyrase at telomeres ...... 121
7.3. Conclusion ...... 122
8. REFERENCES ...... 124
4 1. INTRODUCTION
Tankyrase 1 and 2 are two recently discovered members of the poly-(ADP-ribose) polymerase family (PARP). They are unique multidomain proteins that combine a large protein interaction surface with poly-(ADP-ribose) polymerase activity, and appear to function in a variety of cellular processes.
Tankyrase 1 was identified in 1998 as a binding partner of the telomeric protein TRF1 through a yeast two-hybrid screen (Smith et al., 1998). Subsequently, it was shown to interact with a wide variety of seemingly unrelated proteins. In 2001 several groups isolated tankyrase 2, either by yeast two-hybrid screens or by serological screening of cancer patients (Kaminker et
al., 2001; Kuimov et al., 2001; Lyons et
al., 2001; Monz et al., 2001). Tankyrase
1 and 2 are 81% identical at the amino
Figure 1 Domain structure of tankyrases. HPS: acid level, and have a nearly identical homopolymeric domain, SAM: sterile-α-motif, PARP: poly- (ADP-ribose) polymerase domain structure. Both proteins contain a large ankyrin repeat region, a sterile-α-motif (SAM domain) and a catalytically active poly-
(ADP-ribose) polymerase domain. In addition, tankyrase 1 but not tankyrase 2 has an N-terminal homopolymeric domain (HPS), which contains stretches of histidines, prolines and serines
(Figure 1).
Both tankyrases are expressed in most tissues, albeit to different extents. Tankyrase 1
RNA is most abundant in adipocytes, testis and thymus while tankyrase 2 is preferentially expressed in placenta, skeletal muscle, ovary and prostate (Lyons et al., 2001; Sbodio et al.,
2002; Smith et al., 1998). This difference in tissue preference suggests that tankyrase 1 and 2
5 could have some diverging functions. However, relatively little is currently known about what the roles of these proteins are. A well tested model for tankyrase function only exists at telomeres, were just a small fraction of tankyrase resides. Most of tankyrase 1 and 2 localizes to the cytoplasm where they are thought to act in a variety of processes, most notably insulin signaling and vesicle trafficking. Nevertheless, the functional relevance of tankyrases in these processes has not yet been established. Clues towards understanding tankyrase functioning are revealed by their domain features, therefore I will start with a detailed description of the tankyrase domains, followed by a discussion of the current knowledge of tankyrase functions. To end the introduction I will present the rationale for my thesis research.
1.1. Tankyrase domain structure
1.1.1. Ankyrin repeat domain
Ankyrin (ANK) repeats are versatile protein-protein interaction modules, named after the cytoskeletal adaptor protein ankyrin, which contains 24 such repeats. ANK repeats are relatively common; they are found in more than 325 unrelated human proteins and in a wide variety of organisms, ranging from viruses to humans (Mohler et al., 2002). The number of these ~33 aminoacid motifs in different proteins varies from 2 to 33, with 2-3 being most common (Mosavi et Figure 2 Crystal structure of the al., 2004). Structurally each repeat follows the consensus: β- ANKYRIN-R ankyrin domain. α-helices are in blue, β-turns in red. hairpin – α-helix – loop – α-helix. The anti-parallel α-helices (Michaely et al., 2002).
6 pack into an extended bundle, while interactions between the sequential β-hairpins form a β- sheet, almost perpendicular in respect to the α-helices (Figure 2) (Sedgwick and Smerdon,
1999). The resulting structure has an elongated L shape, with a groove formed by the β-sheet and the α-helical bundle. The residues that define this shape are highly conserved, and mostly buried in the structure, the other, less conserved residues have a higher solvent accessibility, and mediate interactions with other proteins (Mosavi et al., 2004). Unlike protein interaction domains that bind defined sequences (e.g. SH2 domains bind phospho-tyrosine and SH3 domains bind proline rich regions), ANK repeats mediate diverse, protein-specific interactions. The crystal structures of ANK domains complexed with interacting proteins revealed that while all of the solvent-accessible surface can participate in protein interactions, most commonly interactions occur in the groove made by the β-sheet and α helix bundle. The stability of the ANK backbone and the low level of sequence conservation of non-structural residues between different proteins explain the versatility of this domain in mediating protein interactions.
Tankyrases have a large ANK repeat domain, originally thought to contain 24 repeats
(Smith et al., 1998). Alignment of these repeats showed several gaps (5) and insertions (7), which led Sbodio and colleagues (2002) to propose a new alignment without gaps, that contains only 19 ANK repeats (and two half repeats). While this alignment is not devoid of insertions, they now appear at a regular interval, after every fourth ANK repeat and interestingly, each insertion contains a conserved sequence (consensus sequence: LLEAAR), suggesting that this demarcation of the ANK repeats could indeed be meaningful. The same fourfold periodicity was shown to exist between the different ANK repeats, with for example repeats 2,6,14 and 18 sharing high pairwise similarity. Using the 24 ANK repeat alignment, Seimiya et al. (2002)
7 observed a very similar periodicity. These observations indicate that the tankyrase ANK domain
is divided into 5 subdomains of 3-4 ANK repeats, separated by a conserved insertion and have
led to the hypothesis that each subdomain might be able to independently interact with tankyrase binding proteins. The telomeric protein TRF1, the first known interaction partner of tankyrase 1, was indeed shown to be able to interact with all 5 ANK subdomains, and other binding-partners were shown to interact with several of the subdomains (Seimiya et al., 2004; Seimiya and Smith,
2002). While it remains unknown how many proteins can interact simultaneously with the entire
ANK domain, it is clear that this domain provides tankyrases with a large protein-interaction surface. Tankyrases could thus potentially be platforms for the assembly of protein complexes.
Whether there are 24 or 19 ANK repeats and whether the LLEAAR motif has any function remains unknown, and will likely only be resolved when the structure of this ANK domain is determined.
In addition to the structural characteristics of the tankyrase ANK domain, insight has also
been gained into the structural determinants of proteins that interact with this region. Several such proteins were shown to contain a tankyrase binding motif. This RxxPDG sequence is present in IRAP, TAB182, NuMA and human TRF1 (RGCADG) (Sbodio and Chi, 2002). How this sequence mediates interaction with the ANK repeats is still unknown. While this motif is helpful to identify tankyrase binding proteins, it is not absolutely required for ANK domain binding since other tankyrase binding proteins do not contain it (e.g. Grb14). Interestingly, human TRF1 seems to have two tankyrase binding regions that interact with different ANK subdomains. The first, at its N-terminal contains the RGCADG sequence and interacts with ANK subdomains 2-5, the other, in its C-teminal half, interacts with the first ANK subdomain
8 (Seimiya et al., 2004). In this respect it is interesting to note that while chicken TRF1 interacts
with tankyrase, it apparently lacks the N-terminal tankyrase binding motif.
1.1.2. The Sterile-Alpha-Motif (SAM)
SAM domains are small (~70 amino-acids), versatile interaction domains, that were
discovered through sequence alignment of 14 eukaryotic proteins, many of which are involved in
sexual differentiation (Ponting, 1995). A key role appeared to reside in the SAM domain since mutating it in several of these yeast proteins induced sterility. To reflect this, and the predicted all helical structure, the domain was named sterile-alpha-motif. More recent sequence analysis shows that SAM domains are very common, with currently nearly 1000 entries in the SMART database, that range from bacterial to human proteins (Schultz et al., 1997). The structure of several SAM domains has been determined and reveals a compact 5-helical structure with a hydrophobic core. Many of the crystal structures revealed two self-association surfaces, and thus the potential for polymerization. This was indeed shown to occur for the SAM domains of the
proteins TEL and PH,
which assemble into large
fibers, visible by electron
microscopy. Other SAM
domains are also
predicted to polymerize,
but do not form polymers
under in vitro conditions Figure 3 Crystal structure of the predicted EphB2 SAM polymer. Six SAM domains are shown. Helices are colored, 1: blue, 2: orange, 3: green, 4: (e.g. EphB2) (Figure 3) red, 5: yellow. After (Thanos et al., 1999).
9 (Kim et al., 2002; Kim et al., 2001; Thanos et al., 1999). The proposed functions of SAM
polymerization are diverse and poorly understood. In ephrin receptors, polymerization is thought to be promoted by ligand-mediated receptor clustering. The new surface created by the polymerized SAM domain could then constitute a new protein interaction surface to recruit
downstream proteins. TEL and PH are both transcriptional repressors, and SAM polymerisation
is believed to promote spreading of silent chromatin. SAM domain polymerization is also
important in many cancers where chromosomal rearrangements lead to fusions of the TEL SAM
domain to a variety of kinases involved in proliferation. This fusion results in the creation of
constitutively active kinases with oncogenic potential (Golub et al., 1996; Lacronique et al.,
1997; Peeters et al., 1997).
However, not all SAM domains polymerize, some form homodimeric associations, others
oligomeric complexes. Such complexes can include identical and/or different SAM domains
(Kim and Bowie, 2003; Ramachander and Bowie, 2004). In addition, SAM domains have been
shown to mediate interactions not only with non-SAM domain proteins, but also with RNA and
certain lipids (Aviv et al., 2003; Barrera et al., 2003). Thus, it has become clear that SAM
domains are exceptionally versatile interaction modules, that can mediate binding to a variety of
macromolecules.
The SAM domains of tankyrase 1 and 2 have so far received little attention. Sequence
comparison shows that the tankyrase 1 and 2 SAM domains are highly similar to each other (78
% identical; 90 % similar at amino-acid level), but not clearly related to any other known SAM
domains. The closest homolog is the caskin-2 SAM domain, with 33 % identity and 66 %
similarity to the tankyrase 1 SAM domain. This precludes determining their function based on
10 known SAM domain functions. A main goal of this thesis has been to determine the function of the tankyrase SAM domain and is described in depth in Chapter 5.
1.1.3. The poly-(ADP-ribose) polymerase (PARP) domain
Poly-(ADP-ribosyl)ation was discovered more than 40 years ago (Chambon et al., 1963).
This rather dramatic post-translational modification is catalyzed by poly-(ADP-ribose) polymerases (PARPs), enzymes that use NAD as a substrate to synthesize ADP-ribose polymers that are typically covalently attached to glutamic acids on target proteins. The resulting poly-
(ADP-ribose) chains (pADPr) can be branched and up to 200 units long. Since each ADP-ribose group carries two negative charges (one on each phosphate group), the resulting polymers confer a large negative charge upon modified proteins, which can affect their activity and interactions.
A common example of how poly-
(ADP-ribosyl)ation affects interactions
comes from DNA-binding proteins. Since
DNA, like pADPr, is a negatively charged
polymer, poly-(ADP-ribosyl)ation of DNA-
Figure 4 Poly-(ADP-ribosyl)ation. Left: protein modified binding proteins would be expected to with one ADP-ribose group, negative charges are indicated with red circles. Right: poly-(ADP-ribosyl)ated protein, red disrupt their interaction with DNA. This circles depict polymers. NAD : nicotinamide adenine dinucleotide. was indeed observed for a variety of DNA binding proteins, like the TATA box binding protein TBP and the transcription factor YY1 (Oei et al., 1998).
The first enzyme discovered to have PARP activity, PARP-1, is essential for genomic stability in response to genotoxic stress induced by ionising radiation and chemicals. PARP-1 is
11 activated by nicked and broken DNA, resulting in a dramatic increase in pADPr levels after
DNA damage. The importance of PARP-1 in the DNA damage response is exemplified by the hypersensitivity of PARP-1 null mice to genotoxic stress (Shall and de Murcia, 2000). PARP-1 is involved in a variety of DNA-metabolism related processes such as base-excision repair, transcription and DNA synthesis (D'Amours et al., 1999). Many other PARP enzymes have been identified, but most have not been studied to same extent as PARP-1. PARP-2 is a close homolog to PARP-1 and is also involved in DNA repair. PARP-3 localizes to the centrosome and its overexpression interferes with cell cycle progression (Augustin et al., 2003). The vault complex also contains a PARP enzyme (vPARP), but its function remains poorly understood (Kickhoefer et al., 1999).
In addition to regulation of protein function through covalent modification, some proteins bind non-covalently to pADPr (Pleschke et al., 2000). This also can alter protein function and thus provides another level of regulation by PARPs (Dantzer et al., 2004). Poly(ADP-ribose) polymers induced after DNA damage have a very short half-life (<1 min), while less dramatic modifications can exist for longer times. This bimodal behavior is explained by the characteristics of the only pADPr degrading enzyme; poly(ADP-ribose) glycohydrolase (PARG).
PARG is activated by pADPr concentrations > 5µM, such as are typically seen after PARP-1 activation (Alvarez-Gonzalez and Althaus, 1989). Less dramatic activation of PARP enzymes could thus circumvent full activation of PARG and consequently result in longer half-lives for the polymers.
The structure of a PARP-1 catalytic fragment contains two separate domains, an N- terminal all-helical domain without similarity to any known motifs, and a more complex C-
12 terminal domain containing several beta sheets. This domain shows structural similarity to
bacterial mono-ADP-ribose transferases and contains the NAD binding domain (Ruf et al.,
1996). While other PARPs show similarity to both domains, tankyrases only show homology to
the C-terminal NAD binding domain (30% identity at amino acid level between PARP and
tankyrase). Both tankyrases exhibit PARP activity towards themselves (auto-poly(ADP-
ribosyl)ation) and substrate proteins. Tankyrase 1 activity was characterized in vitro using both
the full-length protein and a SAM-PARP fragment. While full-length protein shows activity both
towards TRF1 and itself, the SAM-PARP fragment is only able to auto-ribosylate. Both proteins
nevertheless show very similar auto-ribosylation activity, indicating that all auto-ribosylation sites are present in the SAM-PARP region of the protein (Rippmann et al., 2002). Since TRF1 binds the ankyrin repeats, and does not interact with the SAM-PARP fragment it comes as no surprise that TRF1 is not modified by this fragment. Interestingly, not all ankyrin subdomains support poly-(ADP-ribosyl)ation equally; while the first 4 (most N-terminal) subdomains are, to a certain extent, dispensable, the fifth subdomain, is absolutely required for modification of tankyrase binding proteins. It is thought that this subdomain is unique in the manner that it presents binding proteins to PARP activity (Seimiya et al., 2004). Compared to PARP-1, tankyrase 1 catalytic activity appears to be much lower. When produced using a baculovirus system its catalytic efficiency is about 150 fold lower then that of unstimulated PARP-1. At physiological NAD concentrations, purified tankyrase 1 synthesizes pADPr chains of up to 100
units, with a mean length of 20 units, and unlike for PARP-1 no branched chains could be detected (Rippmann et al., 2002). The low catalytic activity of purified tankyrase suggests that, as for PARP-1, activating mechanisms might exist. While telomeric or damaged DNA do not
13 increase tankyrase 1 activity (Cook et al., 2002), MAP kinase phosporylation does, and occurs in vivo after insulin stimulation (Chi and Lodish, 2000). This suggests that tankyrase 1 has a role as an effector protein downstream of growth-factor induced signaling pathways.
Tankyrases modify several binding factors. The telomere binding protein TRF1 is modified by tankyrase 1 and tankyrase 2 (Cook et al., 2002; Smith et al., 1998) and TAB182 is modified by tankyrase 1 (Seimiya and Smith, 2002). IRAP is also modified by tankyrase but it is not known which of the two enzymes catalyzes this reaction (Chi and Lodish, 2000). Since it has not been determined whether modification also occurs in other tankyrase – substrate protein combinations, it is at this point not clear whether there is any difference in substrate modification between both tankyrases. In contrast to the abovementioned tankyrase binding factors, Mcl-1 proteins are apparently not modified by tankyrase 1. Instead, binding of Mcl-1 proteins to tankyrase 1 inhibits the modification of TRF1, and to a lesser extent, auto-modification (Bae et al., 2003).
The effect of substrate modification is only known for TRF1: poly-(ADP-ribosyl)ation disrupts its ability to bind telomeric DNA, presumably because of repulsion forces between the
DNA phosphate backbone and pADPr. Since the other substrates do not bind DNA, it is unclear how they are affected by poly-(ADP-ribosyl)ation.
1.1.4. Conclusions
Two important characteristics are apparent from the tankyrase domain structure: the ability to interact with a diverse set of proteins through the ankyrin repeat domain and the ability to dramatically alter the physico-chemical properties of interacting proteins through poly-(ADP-
14 ribosyl)ation. The existence of 5 discrete protein-binding regions in the ankyrin domain suggests that tankyrases could be platforms for the assembly of multi-protein complexes. Since only the proteins bound to the fifth ankyrin subdomain can be poly-(ADP-ribosyl)ated, the other repeats could also serve as “loading docks” to quickly replace modified proteins with unmodified proteins. Poly-(ADP-ribosyl)ation of TRF1 dramatically alters its functioning, and understanding this effect has provided important clues toward the role of tankyrase in telomere biology.
Therefore, to better understand the function of tankyrases at other locations, it will be essential to determine the effect poly-(ADP-ribosyl)ation on other tankyrase substrates. Furthermore, since tankyrases exhibit only low activity it is likely that activation mechanisms exist. Understanding these will be of great value to define the physiological roles of tankyrases. Very little is known about the tankyrase HPS and SAM domains, and, in view of gaining more insight into tankyrase behavior, it will be critical that the characteristics and role of these domains be studied.
1.2. Tankyrase functions
Tankyrases interact with a diverse set of proteins and localize to an equally diverse set of sub-cellular locations. They are found at telomeres, nuclear pores, centrosomes, Golgi-associated vesicles and other undefined cytoplasmic locations. Except for at telomeres, little is known about the role of tankyrases at these locations.
1.2.1. Function at telomeres
After its identification as a binding partner of the telomeric protein TRF1 in a yeast two- hybrid screen, tankyrase 1 presence at telomeres was confirmed by immunofluorscence experiments on metaphase chromosomes (Smith et al., 1998). Telomeres are DNA-protein
15 complexes at the ends of the chromosomes; their major role is to protect the end of the
chromosome from nucleases and from being recognized as DNA damage (Wei and Price, 2003).
TRF1 (Telomere Repeat Factor 1), and its homolog TRF2, are DNA-binding proteins that bind
specifically to telomeres (Broccoli et al., 1997b; Chong et al., 1995; Zhong et al., 1992). TRF1 is
a regulator of telomere length; its overexpression leads to a reduction in telomere length, while
overexpression of a dominant negative construct results in telomerase-mediated telomere
elongation (van Steensel and de Lange, 1997). TRF1, together with its binding partners, is
thought to package the telomere into a “closed” conformation that is less likely to be elongated
by telomerase. Both tankyrases were shown to specifically interact with the N-terminus of TRF1,
and to poly-(ADP-ribosyl)ate TRF1 (Cook et al., 2002; Kaminker et al., 2001; Smith et al.,
1998). In vitro, TRF1 looses its ability to bind telomeric sequences upon ribosylation, suggesting
that in vivo tankyrase might be able to remove TRF1 from the telomeres. Consistent with this
hypothesis, overexpression of tankyrase 1 and 2 leads to removal of the TRF1 signal from the
telomere, an effect that is dependent on tankyrase activity. As expected, the elimination of TRF1
in turn results in telomere elongation (~ 48 bp / PD) in a telomerase dependent manner (Cook et
al., 2002; Smith and de Lange, 2000). After removal from the telomere, TRF1 is degraded via
the ubiquitin-proteasome pathway (Chang et al., 2003). Together these results indicate that
tankyrases are negative regulators of TRF1 function, and can mediate the “open” telomere
structure that may be required for telomere replication and elongation.
Recently it has become clear that another TRF1 interacting protein, TIN2, is responsible
for controlling tankyrase activity towards TRF1. Ye and de Lange (2004) showed that in vitro,
TIN2 prevents TRF1 ribosylation by tankyrase without affecting auto-ribosylation, thus
16 suggesting a TRF1-specific steric hindrance mechanism that controls modification. RNAi
experiments show that upon downregulation of TIN2, there is a reduction in telomeric TRF1
signals and concomitant telomere elongation, consistent with increased tankyrase activity
towards TRF1. TIN2 is thus a telomere-specific inhibitor of tankyrase that prevents inappropriate
modification of TRF1. Whether this type of non-catalytic steric inhibition is a general mechanism for tankyrase regulation remains unclear at this time. It is also unknown how (and
when) TIN2 inhibition of tankyrase is relieved to allow for modification of TRF1.
In addition to its role in telomere length regulation, tankyrase 1 also has a role in sister chromatid separation. Reduction of tankyrase 1 levels by RNAi induced a mitotic arrest that correlated with failure to separate the telomeres of sister chromatids during metaphase.
Tankyrase PARP activity is necessary for progression through metaphase, indicating that it is required for sister chromatid resolution (Dynek and Smith, 2004). These results suggest that tankyrase-1 is required at telomeres to undo some sort of telomere-specific cohesion complex that keeps sister chromatids together. Whether TRF1 or other proteins are the target of poly-
(ADP-ribosyl)ation remains to be determined, as do the upstream events that trigger tankyrase-1 activation.
1.2.2. Function in vesicle trafficking and insulin signaling
The fraction of total cellular tankyrase that localizes to telomeres is very small and the majority of these proteins is found in the cytoplasm. Several lines of evidence point towards a function for cytoplasmic tankyrases in vesicle-trafficking and insulin signaling. In an effort to identify proteins involved in insulin-stimulated translocation of GLUT4 (glucose transporter 4) vesicles, tankyrase 1 and 2 were identified as binding partners of IRAP (insulin-responsive
17 aminopeptidase). IRAP, also known as P-LAP (placental leucine aminopeptidase) and
oxytocinase, is a type II membrane spanning zinc-dependent metallo-peptidase (Rogi et al.,
1996). It is found in cytoplasmic vesicles, in the plasma membrane, and in the bloodstream and
has a role in hormone homeostasis. In vitro IRAP cleaves and inactivates a variety of small
peptide hormones such as oxytocin, vasopressin and the neuropeptides met-enkephalin and
dynorphin A (Matsumoto et al., 2000), whether these are also substrates in vivo is unknown. In
addition, IRAP is the angiotensin IV (Ang IV) receptor, a neuropeptide that is important for both
memory retention and retrieval (Albiston et al., 2001). This neuropeptide, however, is not
cleaved by IRAP, but instead blocks its activity towards other substrates. In this way, Ang IV may prevent inactivation of neuropeptides necessary for memory function. In adipocytes and myocytes IRAP localizes to GLUT4 vesicles (Kandror et al., 1994). These vesicles contain the insulin responsive glucose transporter 4. In response to insulin stimulation, they translocate from the trans-Golgi network to the plasma membrane (Martinez-Arca et al., 2000; Slot et al., 1991).
This leads to insertion of GLUT4 in the membrane and subsequent entry of glucose. IRAP also becomes inserted in the plasma membrane, but its role as a peptidase in the insulin pathway is unclear (Keller, 2003; Keller, 2004; Ross et al., 1997). It does, however, seem to have a role in
GLUT4 vesicle translocation since their insulin-responsiveness is, at least partially, conferred by the IRAP cytoplasmic domain, and overexpression of a small region (28 amino acids) of this domain induces GLUT4 vesicle translocation in absence of insulin stimulation (Waters et al.,
1997).
Both tankyrases interact with the cytoplasmic domain of IRAP (not the 28 aa necessary for translocation) through their ankyrin repeats and tankyrase 1 colocalizes with GLUT4 vesicles
18 in adipocytes. In addition, tankyrase 1 can poly-(ADP-ribosyl)ate IRAP in vitro and, poly-(ADP- ribosyl)ation activity is stimulated by MAPK phosphorylation in vivo, downstream of insulin receptor activation (Chi and Lodish, 2000). The physiological relevance of the tankyrase - IRAP interaction has not yet been determined, but tankyrase binding to the cytoplasmic domain of
IRAP together with its responsiveness to insulin signaling suggest that tankyrases could have a role in insulin promoted vesicle trafficking. Further evidence supporting a role in vesicle trafficking comes from sub-cellular localization experiments. Both in cells of epithelial and mesenchymal origin tankyrases fractionate to low density microsomes, a fraction that contains
Golgi-membranes and endosomes (Chi and Lodish, 2000; Lyons et al., 2001) indicating that the bulk of cytoplasmic tankyrases are associated with vesicles. Interestingly the localization pattern observed by immunofluorescence is quite different between these cells: in 3T3-L1 fibroblasts, tankyrase 1 colocalizes almost completely with the Golgi-marker FTCD, while there is very little overlap in epithelial cells (DU145 carcinoma cells). Thus, while tankyrases clearly associate with a membrane fraction, there may be differences in the nature of these membranes between different cell types.
Intriguingly, GRB14, another protein involved in insulin signaling, was also shown to interact with tankyrases. GRB14 is a member of the GRB7 family of adaptor proteins (Daly,
1998). These proteins contain SH2 domains and PH domains, and are involved in the assembly of protein complexes at activated receptor tyrosine kinases. In response to insulin, GRB14 is recruited to the insulin receptor, where it inhibits its catalytic activity (Bereziat et al., 2002;
Cariou et al., 2002; Kasus-Jacobi et al., 1998). GRB14 thus appears to be part of a negative feedback mechanism in insulin signaling.
19 Tankyrase 1 and 2 both form a complex with GRB14 in vivo, but only tankyrase 2 was
shown to directly interact. Like IRAP – tankyrase, this complex also resides at Golgi-membranes
or vesicles since most GRB14 cofractionates with tankyrases in the low density microsome
fraction of epithelial cells (Lyons et al., 2001). This could indicate that tankyrases recruit GRB14
to intracellular vesicles, alternatively the GRB14 PH domain could mediate membrane association. To date, the functional relevance of the tankyrase – GRB14 interaction remains
unknown, it also has not been determined yet whether tankyrases, IRAP and GRB14 all reside in
the same complex.
Together, the above observations strongly suggest that tankyrases have a role in both
insulin signaling and vesicle trafficking. To obtain insight in the role of tankyrases in these
processes it will be important to study the physiological relevance of these interactions in vivo. A
possible working model is that tankyrase activation and subsequent poly-(ADP-ribosyl)ation of
target proteins leads to both disruption of a GLUT4-vesicle tether and release of GRB14, thus
promoting vesicle translocation and providing a negative feedback mechanism for insulin
signaling.
1.2.3. Other functions
In addition to telomeres and vesicles, tankyrases are also found at nuclear pore complexes
and centrosomes (Smith and de Lange, 1999). Their function at these locations has not yet been
studied. Tankyrase colocalizes with NuMA at centrosomes, and that in vitro these two proteins
interact directly (Sbodio and Chi, 2002). NuMA (Nuclear Mitotic Apparatus) organizes the
formation of a large protein network that is essential for the establishment of the mitotic spindle.
20 Bae et al. (2003) demonstrated that tankyrase 1 can directly interact with the anti- and pro-apoptotic proteins Mcl1-L and Mcl1-S through its ankyrin repeats. Co-overexpression of tankyrase with each Mcl-1 antagonized their respective anti- and pro- apoptotic effects. This effect correlated with an apparent reduction in Mcl-1 protein levels and was not dependent on tankyrase PARP activity. While these results suggest a role for tankyrase-1 in apoptosis, it is not clear whether tankyrase-1 has any effect on endogenous Mcl-1 proteins.
A novel protein, TAB182 (182 kDa TAnkyrase-Binding protein) was identified through its interaction with the tankyrase 1 ankyrin domain (Seimiya and Smith, 2002). Both tankyrase 1 and tankyrase 2 interact with TAB182, and in vitro TAB182 is a substrate for tankyrase 1 poly-
(ADP-ribosyl)ation. TAB182 has no known conserved domains and its function is unknown.
Immunofluorescence experiments revealed the presence of TAB182 both at heterochromatin in the nucleus, and at cortical actin in the cytoplasm. At which of these locations it interacts with tankyrase was not determined.
21
2. GOAL OF THESIS
2.1. Hypothesis
The tankyrase domain structure offers clues towards a general function for these proteins.
The ankyrin repeat domain could serve as a platform to assemble protein complexes, while poly-
(ADP-ribosyl)ase activity may result in disruption of such complexes. Thus, our hypothesis is that tankyrases are master scaffolding proteins that have the intrinsic capacity to both assemble and dissassemble protein complexes. Based on the function of other SAM domains, we also hypothesize that the tankyrase SAM domain could provide a self-association surface, which would further expand tankyrase’s scaffolding ability.
2.2. Overview
Many aspects of tankyrase biology remain uninvestigated. Some key questions that require further study are:
• What are the functions of the HPS and SAM domains?
• How are tankyrases regulated?
• What are the functions of tankyrases at their different locations?
I have used a combination of biochemical, genetic and cell biology approaches, using the chicken DT40 cell line as a system to study protein function. This cell line has high levels of homologous recombination and hence allows the creation of knock-out cell lines. Since chicken tankyrases have not previously been characterized this work starts with the isolation and characterization of chicken tankyrases. After this initial characterization several avenues were followed to increase the understanding of tankyrase biology.
22 While SAM domains typically play key roles in proteins, the functions of the tankyrase
SAM domain have to date not been investigated. In this work, I have focused on addressing the role of the tankyrase SAM domain. As shown in Chapter 5, our results show that the SAM domain mediates tankyrase polymerization, and that SAM-mediated polymerization can be reversed by poly-(ADP-ribosyl)ation. We therefore propose that tankyrases are master scaffolding proteins that can both assemble and disassemble large protein complexes. The implications of this hypothesis for tankyrase functioning at its diverse locations are discussed.
Since little is known about proteins that regulate tankyrase activity I have also attempted to isolate factors that could be involved in tankyrase regulation. To date no proteins have been identified that interact with the tankyrase SAM or PARP domains, while it is exactly such proteins that could potentially alter tankyrase activity. Therefore I have performed a yeast two- hybrid screen with the SAM and PARP domains. Chapter 6 describes this screen, and shows how some of the isolated factors could potentially be regulators of tankyrase activity.
Many cells express both tankyrase 1 and 2, it is unclear however whether both proteins are completely redundant, or whether they have dedicated functions. To address this question, and to understand the function of tankyrase 2, I have attempted to knock this gene out in the chicken
DT40 cell line, as described in Chapter 4.
23 3. MATERIALS AND METHODS
3.1. Cell lines
DT40 cells (ATCC CRL-2111) and LMH cells (ATCC CRL-2117) were maintained in
RPMI-1640 (Invitrogen) supplemented with 10% fetal calf serum (Hyclone), 5% chicken serum,
50 µM β-mercaptoethanol, 100 U/l penicillin, 50 µg/ml streptomycin, and 2 mM glutamine. Sf21 insect cells were grown in SF-900 medium (Invitrogen) supplemented with 10% fetal calf serum
(Invitrogen) and 50 U/l penicillin / 25 µg/ml streptomycin (Sigma).
3.2. Isolation of chicken tankyrase genes
Chicken tankyrase genes were isolated from a λZAPII chicken embryonic fibroblast cDNA library as recommended by the manufacturer (Stratagene) (as described (Venkatesan, 2000)).
Sequence alignments were computed with the GCG Wisconsin package. RNA was isolated by
lysing DT40 cells in guanidinium thiocyanate and extracting the lysate with acidic phenol
(courtesy of Dr. C. Wei). The RNA was reverse transcribed using primer p12-1Rev3 (5’
GCAGAGGAGTGAACTGCC) and Superscript II reverse transcriptase (Gibco/BRL). First
strand cDNA was amplified by PCR with primers cTANK2F2 (5’ GTTCCTCCGGGATCCCTC)
and cTANK2cDNA (5’ GCCTCCATCATCTCGTGC) using the GC-Melt PCR kit (Clontech).
Amplified DNA was gel-purified and sequenced.
3.3. Genomic library screen and knock-out construct
A DT40 chicken genomic library (cloned into λFIX II vector (Stratagene)), was screened for
an N-terminal fragment of tankyrase 2 using standard methods. The probe used for the primary
24 screen was amplified by PCR with primers cTNK2F2 (5’ GTTCCTCCGGGATCCCTC) and
p12-1 Rev1 (5’ TCCCGGGCCGGCTCCACC), yielding a 230 bp fragment with 35 % identity to the sequence of tankyrase 1, thus making cross-hybridization unlikely. A secondary screen was
performed with a 400 bp PCR probe amplified with forward primer cTNK2F2 (5’
GTTCCTCCGGGATCCCTC) and reverse primer cTNK2cDNA1 (5’
GCCTCCATCATCTCGTGC). DNA from positive phages was purified using a Qiagen phage
miniprep kit, and subjected to restriction enzyme mapping analysis to identify colonies
containing suitable fragments. Phage 13-1 (λ13-1) contained 15 kb of genomic DNA spanning
the region from 3 kb 5’ of the tankyrase 2 ATG to 12 kb 3’ of the ATG. The lambda DNA was
cloned into pBluescript II KS (+) after digestion with Not1. Not1 releases the library DNA from
the phage vector and also cuts immediately upstream of the ATG. Two clones were obtained,
one (p12-1#2) containing 3 kb upstream of the ATG, and another (p12-1#11), containing 12 kb,
downstream of the ATG. The knock-out construct was made so that it contained 2.5 kb of
genomic DNA 5’ of the ATG and 3 kb 3’ (3’ sequence starts ~7 kb downstream of ATG).
Integration would thus result in the removal of 7 kb of genomic DNA, including the 1st exon
(Figure 12). A BamH1 site inbetween both arms allowed cloning of different drug resistance marker cassettes (LoxP flanked neomycin and blasticidin-S resistance markers, and histidinol resistance (no LoxP)). Stop codons were introduced in all three reading frames immediately 5’ to the marker cassettes to prevent expression of a truncated protein after excision of the marker cassettes.
25 3.4. Electroporations
Knock-out constructs were purified by CsCl purification, linearized with SacI, ethanol
precipitated and resuspended in 500 µl endotoxin free PBS. For the tankyrase 2 rescuing allele, pTRE2-HYG-Flag-tankyrase 2 was linearized with Apa1, and processed as described for knock- out constructs. Prior to electroporation DT40 cells were kept for a minimum of 5 days in exponential growth (cells were diluted daily to 0.5x106/ml). 1x107 cells were harvested, washed
once with endotoxin free PBS and resuspended in 300 µl endotoxin free PBS. Cells and DNA
were mixed (total 800 µl) and incubated for 10 minutes on ice in an electroporation cuvette
(Biorad, 0.4 mm gap). Electroporations were performed at 600 V / 25 µF. Parallel electroporations with a GFP expression construct (pMX-GFP) were carried out to estimate transfection efficiency, which typically was ~10%. After electroporation cells were incubated 10 minutes on ice, and dispensed gently with a sterile glass pasteur pipette into 20 ml of prewarmed
DT40 media. After overnight growth, cells were diluted to a total of 80 ml and the appropriate drugs were added (geneticin: 1.5 mg/ml, blasticidin-S: 30 µg/ml, L-histidinol: 0.75 mg/ml (96 well plate stage) – 1.5 mg/ml (24 well plate – T25 stage), hygromycin 1.5 mg/ml). Cells were dispensed in 96 well plates (200 µl / well) and single clones were isolated and frozen for storage and DNA isolation. For inhibition of transgene expression, doxycyclin was added at 100 ng / ml.
Alternative electroporation conditions can be used to get increased elecroporation efficiency: 320
V / 950 µF or 320 V / 550 µF. The higher the capacitance, the higher cell death (foam will appear in the cuvette after pulsing), so one would vary the capacitance depending on the fragility of the cell line and purpose of the experiment. Often 320 V / 550 µF is the best option for transient transfection experiments.
26
3.5. Analysis of DT40 genomic DNA by Southern blot
DT40 DNA was isolated from a minimum of 1x107 cells using the SCOOT procedure (Price
lab method: cells are lysed in buffer containing 1% SDS, RNA and protein are digested with
RNAse A and Proteinase K, proteinaceous matter is precipitated by addition of NaClO4 and
NH4Ac followed by isopropanol precipitation of the genomic DNA). Purified genomic DNA (14
µl) was digested overnight with restriction enzymes followed by fluorometric quantification.
DNA was resolved by agarose (0.5%) gel electrophoresis (12µg of each clone). For the
subsequent steps it is important to use relatively thin gels; 300 ml of agarose solution for a large
gel is sufficient. Gels were run in absence of ethidium bromide and post-stained for 10 minutes.
After imaging, gels were destained 10 minutes in distilled water, followed by depurination (10
min), denaturation (35 min), neutralization (25 min) and vacuum transfer (2 h, 50 psi) to a nylon
membrane. Probes ECO7 (3’ probe, Sph1 digestion) and 13-1#2 (5’ probe, Xba1 digestion) were
used for analysis of knock-out clones (Figure 12). ECO7 was made by PCR using primers
pECO7 F (5’ TGGAAGTTGAATTTGGGAAGTGC) and pECO7 R (5’
AGTCGACTCGATCTACTGCATCC). Probe 13-1#2 was produced by PCR with primers p13-
1#2probeF (5’ CGCGAGCTCTAATACGACTC) and p13-1#2probeR (5’
GAAAACAGGATCTGCAGAGAG). Hybridizations were carried out overnight in Church
hybridization buffer (500 mM Sodium Phosphate (pH7.2), 7% SDS, 1 mM EDTA, 0.1% BSA)
containing 0.1 mg/ml herring sperm DNA at 60 °C. Much care was taken to avoid drying out of
the membrane when replacing the hybridization buffer with wash buffer. Membranes were
washed in Church Wash (40 mM Sodium Phosphate (pH7.2), 1% SDS, 1 mM EDTA) and
27 exposed to a phosphorimager screen. A Molecular Dynamics STORM phosphorimaging system
was used for detection, and analyses were performed with ImageQuant (Molecular Dynamics).
3.6. Antibodies and protein detection
Antibody to the SAM domain of chicken tankyrase 1 was raised in rabbits as described
(Venkatesan, 2000). Briefly, the SAM domain from chicken tankyrase 1 was expressed in E. coli with a His-tag, purified by Ni2+ affinity chromatography and used to raise a polyclonal antibody
in rabbits (Covance). The antibody was affinity purified using immobilized SAM domain. To
determine the relative affinity of the antibody for chicken tankyrase 1 versus 2, the SAM
domains of tankyrase 1 and 2 were expressed as GST fusion proteins (SAM1 and SAM2) in E.
coli. The amount of each fusion protein was estimated by SDS-PAGE followed by Coomassie
blue staining and comparison to BSA standards. The signals from equal amounts of SAM1 and
SAM2 were then compared after Western blotting with affinity-purified rabbit anti-tankyrase 1 antibody. To detect endogenous proteins, whole cells or nuclei were resuspended in a suitable
volume of 150 mM NaCl, 20 mM Hepes pH 7.0. Next 3x SDS-loading dye was added to final concentration of 1x. The extracts were then briefly sonicated to reduce viscosity, boiled for 5 minutes, spun briefly, loaded onto a thick (1.5mm) SDS-PAGE gel and ran overnight at 20V. For extracts with a total volume of <100 µl no sonication was performed, instead the extracts were boiled and vortexed to reduce viscosity. Tankyrase was detected by Western blotting using the tankyrase 1 antibody (1/400 for pico detection, 1/2000 for femto detection), horseradish- peroxidase conjugated secondary antibody and Supersignal West Femto chemiluminescence substrate (Pierce). Tankyrase 1 and 2 were transcribed and translated in vitro using a Promega
28 TnT T3/T7 kit. To ascertain its specificity in immunofluorescence experiments the antibody was
incubated with an excess of purified MBP-SAM1 (SAM domain was used for antibody
production) in 150 mM NaCl, 20 mM hepes pH 7.0 for 30 minutes followed by use in an
immunofluorescence experiment as described below.
Anti-chicken TRF1 antibody production courtesy of Dr. Chao Wei. Chicken TRF1 was expressed in E. coli as a GST-fusion protein and purified on glutathione sepharose 4FF beads
(Amersham Pharmacia). The GST tag was removed by thrombin digestion, followed by gel purification. Polyclonal antibody was raised in rabbits (Covance) and purified over a cTRF1 affinity column. A DT40 cell line stably expressing FLAG-tagged cTRF1 was obtained by electroporating a TET responsive cTRF1 expression construct (based on pUHC13-3 (Gossen and
Bujard, 1992)) into DT40 cells. The cells were co-transfected with a neomycin resistance construct to allow selection of transformants with geneticin. To examine Flag-cTRF1 expression,
total cellular proteins were extracted in RIPA buffer (150 mM NaCl, 1% Nonidet P-40, 0.5%
deoxycholate, 0.1% SDS, 50 mM Tris, pH 8.0, yeast protease inhibitor cocktail, 1mM PMSF)
and analyzed by SDS-PAGE followed by Western blotting with anti-cTRF1 antibody or anti-
FLAG (M2) (Sigma) antibody. Horseradish-peroxidase conjugated secondary antibodies and
Supersignal West Pico and Femto chemiluminescence substrates were used for antibody
detection.
Other antibodies were as follows: mouse anti-poly(ADP-ribose): clone 10H
(Pharmingen). Mouse anti-MYC: clone 9E10, Rabbit anti-MYC: polyclonal (ab6, Neomarkers).
Mouse anti-FTCD: clone 58K-9 (Sigma). Mouse anti-Flag: clone M2 (Sigma). Rabbit anti-Flag:
polyclonal (Sigma). Mouse anti-α-tubulin: clone DM 1A (Sigma). Rabbit anti-histone H3
29 antibody: polyclonal (Cell Signaling). Secondary antibodies: anti-mouse and anti-rabbit
horseradish peroxidase conjugates (Pierce), RRX anti-mouse and Cy2 anti-mouse (Jackson
Immuno Labs), Cy2 anti-rabbit (Rockland), normal goat serum (Jackson Immuno Labs).
3.7. Expression constructs
For bacterial expression the following fragments were cloned as GST fusions (in pGEX-4T-
1) or MBP fusions (in pMAL-c2). GST-tankyrase 1 and 2 as used in cTRF1 pull-down experiments: amino acids 428–736 of tankyrase 1 and 332–645 of tankyrase 2. GST-LLEASK, used in nucleotide binding assays: amino acids 501-666 of tankyrase 1. The ALEASA mutant construct was created with the Stratagene Quick-change mutagenesis kit using primers
ALEASA-F and –R (GTGGATTACCGGGCGTTAGAAGCATCCGCAGCTGGAGACTTG and its reverse complement). MBP-fusion proteins for polymerization experiments: mSAM-
PARP2 (SARP2): aa 797–1168 of tankyrase 2 (with M1055 to V mutation); SAM1: aa 892–1091 of tankyrase 1 and SAM2: aa 797–1000 of tankyrase 2.
For co-expression experiments, the SAM domains of tankyrase 1 and 2 were cloned in the bicistronic bacterial expression vector pET Duet-1 (Novagen). The tankyrase 1 SAM domain
(aa 892 – 1091 of tankyrase 1) was cloned in frame with a HIS tag, the SAM 2 domain (aa 797 -
1000 of tankyrase 2) was cloned as a maltose binding protein fusion protein (NEB). Control
vectors carrying only MBP-SAM2 or HIS-SAM1 and MBP were also constructed.
For yeast expression, the SAM and PARP domains of chicken tankyrase 2 (aa 797 –
1167) and the SAM domain of chicken tankyrase 2 (aa 797 – 1000) were cloned into pGILDA
(pGILDA-SARP2 and pGILDA-SAM2). To inactivate PARP activity in the SAM-PARP
30 construct, M1055 was mutated to V, by changing 3165 A -> G using the Quick-change site-
directed mutagenesis kit (Stratagene) resulting in pGILDA-mSARP2. mSARP2 was also cloned into pJG4-5 (pJG4-5-mSARP2).
For eukaryotic expression, chicken tankyrase 1 and chicken tankyrase 2 were cloned into pcDNA3 and pTRE2-HYG using primers that added an N-terminal MYC and FLAG tag respectively. To construct pcDNA3-tankyrase 1-∆SAM, overlap PCR was performed using an
1800 bp Xba fragment of pcDNA3-tankyrase 1 as template. Primers were designed so that nucleotides 2884 – 3120, comprising the SAM domain, were removed. Successful deletion of the
SAM domain was confirmed by sequencing. The resulting 1564 bp fragment was then cloned back into the remainder of the gene using the Xba restriction sites. To create a chicken tankyrase
1 lacking PARP activity (pcDNA3-tankyrase 1 PARP†), 2 aminoacids were mutated using the
Quickchange site-directed mutagenesis kit (Stratagene). H1122-> A was made by mutating 3367
CA -> GC and E1229 -> A by changing 3689 A -> C. The SAM domains of tankyrase 1 and 2
were cloned into pcDNA3 using primers that provided an N-terminal FLAG tag and amplified aa
892 – 1091 of tankyrase 1 and aa 797 - 1000 of tankyrase 2. The marker constructs for secretion
GFP–GL–GPI and VSVG3–SP–GFP are described in (Keller et al., 2001).
Plasmids encoding tankyrase are often not stable in E.coli (even when not expressed). To
avoid plasmid loss always use reduced recombination strains such as SURE or STBL for DNA
production. Plasmids usually get lost during growth in stationary phase, it is therefore better to
grow the bacteria for a relatively short time and to inoculate with a small inoculum. Successful
conditions for a maxiprep are as follows: grow 2 ml culture from fresh colony for ~10 hours in
LB, use 1/2000 of this culture to inoculate a 100 ml culture (LB), grow for 14 hours. Place
31 bacteria on ice.
3.8. Protein expression and purification
Portions of tankyrase 1 and 2 were expressed in E. coli BL21. MBP-SAM2 was expressed in
E. coli BL21-Star-pRARE (kind gift of Carol Caparelli). Cultures were grown overnight in 5 ml
LB. In the morning 100 – 250 ml LB ( + 2 g / l glucose for MBP fusions, glucose inhibits
bacterial production of amylase, an enzyme that degrades amylose affinity matrices) were
inoculated with 4 – 10 ml of the overnight culture and grown at 37°C to OD 0.7. The cultures were then moved to 15°C and grown for 30 minutes before induction with 0.05mM IPTG.
Soluble proteins were isolated from 250 ml cultures at room temperature using 15 ml B-Per
(Pierce) in presence of 5 units of DNAse1, 0.5 mM PMSF, 1mM DTT and yeast protease inhibitor cocktail (6 µg/ml chymostatin, 1 µg/ml E64, 2µg/ml aprotinin, 0.5 µg/ml phosphoramidon, 1 µg/ml pepstatin A, 5 µg/ml leupeptin, 5 µg/ml antipain, 0.1 mM benzamidine (Sigma)). 1.5mM EDTA was added after complete resuspension of the pellet. The same concentrations of PMSF, DTT and yeast protease inhibitor cocktail were included in subsequent steps. GST-fusion proteins were coupled to glutathione 4FF beads (Amersham-
Pharmacia), washed sequentially with PBS, wash buffer A (600 mM KCl, 1.5% Triton, 20 mM
Tris-HCl pH 7.4) and PBS (each wash 10 minutes), and stored in PBS/Glycerol (2:1) at 4°C.
For GST-LLEAAR, fusion proteins were eluted from the beads with glutathione buffer
(50 mM Tris-Hcl pH8.0, 10 mM glutathione). MBP-fusion proteins were coupled to amylose beads (NEB) at 4°C for 2 hours, followed by three washes with wash buffer B (20 mM Hepes, pH 7.5, 200 mM NaCl). Fusion proteins were eluted with maltose elution buffer (20 mM Hepes, pH 7.0, 150 mM NaCl, 10 mM maltose). Protein concentrations were estimated by Coomassie
32 Blue staining and Bradford assays. Factor Xa (NEB) cleavage of the MBP fusions was
performed overnight at 4 °C. Cross-linking with bis(sulphosuccinimidyl)suberate (BS3; Pierce) was performed in 20 mM Hepes, pH 7.0, 150 mM NaCl, at 4 °C for 1 h using approx. 10 mg/ml
3 3 purified SAM domain and 0–2 mM BS . BS was stored under N2.
pET Duet constructs were expressed in E.coli strain BL21 Star (Invitrogen), containing
the plasmid pRARE to supply rare tRNAs. For protein production 25 ml cultures were grown to
stationary phase overnight in LB + 1% glucose. Subsequently, 25 ml LB + 0.2% glucose was
inoculated to an OD of 0.3. Cells were grown to OD 0.6 at 37°C, then placed at 15°C and, 30
minutes later, induced with 0.5 mM IPTG. Cells were harvested 5 hours after induction. For
purification, 10 ml cell pellets were extracted at room temperature with 2 ml B-PER (Pierce)
supplemented with 1 U DNAse1, 1 mM PMSF, 0.5 mM DTT, yeast protease inhibitor cocktail (2
µg/ml chymostatin, 1 µg/ml E64, 2 µg/ml aprotinin, 5 µg/ml leupeptin, 0.1 mM benzamidine).
1.5 mM EDTA was added after 5 minutes of extraction and extracts were spun 10 minutes at
18000xg. Clarified extracts were incubated with 200 µl amylose beads (NEB) for 2 hours at 4°C followed by three 5 minute washes with 1.5 ml 20 mM Hepes pH 7.0, 200 mM NaCl, 1 mM
PMSF, 0.5 mM DTT and yeast protease inhibitor cocktail at 4°C. Purified proteins were eluted with 2 x 40 µl 20 mM Hepes pH 7.0, 150 mM NaCl, 10 mM maltose, 0.5 mM DTT. Purified preparations were >98% pure as assayed by coommassie staining (MBP-SAM2 + SAM1).
Expression levels of SAM1 were significantly lower than MBP-SAM2, presumably because of instability of the protein, and were therefore further analyzed by Western blotting, with antibodies for tankyrase and MBP (NEB).
33 His-tagged chicken tankyrase 1, tankyrase1-∆SAM and human TRF1 were cloned into pFastbac1 and expressed using the Bac-to-Bac baculovirus expression system (Invitrogen). Sf21 cells were infected at a MOI of 10 and harvested after 47 hours (hTRF1) or 72 hrs (FL-
tankyrase1 and tankyrase1-∆SAM). To purify TRF1 and tankyrase1-∆SAM, cells were lysed
with buffer A (500 mM NaCl, 20 mM tris-HCl pH 7.9, 5 mM imidazole, 1 mM PMSF, and yeast
protease inhibitor cocktail (6 µg/ml chymostatin, 1 µg/ml E64, 2 µg/ml aprotinin, 0.5 µg/ml
phosphoramidon, 1 µg/ml pepstatin A, 5 µg/ml leupeptin, 5 µg/ml antipain, 0.1 mM
benzamidine). The cells were sonicated briefly, extracted for 30 min on ice and centrifuged to
remove debris. The supernatant was applied to Ni++ charged chelating Sepharose (Amersham
Pharmacia) and incubated for 1 hr at 4°C with rocking. The beads were washed extensively with
buffer B (total of 70 ml) (500 mM NaCl, 20 mM tris-HCl pH 7.9, 80 mM imidazole). Protein
was eluted in three fractions using buffer C (500 mM NaCl, 20 mM tris-HCl pH 7.9, 500 mM
imidazole) and dialyzed against buffer D (300 mM KCl, 20 mM Hepes pH 7.5, 3 mM MgCl2,
20% glycerol, 1 mM DTT). hTRF1 was tested for DNA-binding activity by mobility shift assay using dsDNA containing five T2AG3 repeats (Smith et al., 1998). Cells expressing FL-tankyrase1
were washed with PBS and lysed with modified buffer A (500 mM NaCl, 20 mM tris-HCl pH
7.9, 3 mM MgCl2, 5 mM imidazole, 0.05 % Triton X-100, 1 mM PMSF and yeast protease
inhibitor cocktail). After sonication, additional modified buffer A was added, to a final volume
of ½ the culture volume. Cells were extracted for 30 min at 30°C and centrifuged to pellet debris.
Subsequent purification steps were essentially as described for hTRF1, except buffer D was
modified to 300 mM NaCl, 20 mM Hepes pH 7.5, 3 mM MgCl2, 10% glycerol, 5 mM β- mercapto-ethanol.
34 LMH cells were transfected using Transit-LT1 (Mirus). Ratio of DNA:lipid was 1µg :
3µl. For a single well in a six well plate: plate 2x105 cells, grow ~24 hours and replace media (2
ml / well). Prepare transfection complexes as follows: place 100 µl RPMI in a sterile eppendorf tube, add 6µl Transit LT-1 and mix by pipetting up and down, incubate 5 minutes at RT, add 2µg
DNA, mix by pipetting up and down, incubate 5 minutes at RT. Add mixture dropwise to well,
shake gently to mix and incubate 4 – 36 hours. Increase volumes linearly for co-transfections and
larger culture vessels.
3.9. GST pull-downs
Ankyrin repeats 10-18 (Smith et al., 1998) of chicken tankyrase 1 (aa 428 – 736) and 2 (aa
332 – 645) were expressed in pGEX-4T-1 and purified as described above. DT40 nuclear
extracts were prepared from 1x108 cells. Cells were washed once in PBS and swollen in RSB (10 mM NaCl, 1.5 mM MgCl2, 10 mM Tris-HCl, pH 7.4) for 10 min on ice. Nuclei were released by
dounce homogenization and washed three times in RSB. Nuclei were extracted with 420 mM
KCl, 20 mM Hepes-KOH, pH 7.9, 25% glycerol, 0.1 mM EDTA, 5 mM MgCl2 for 30 min at
4°C and the extract was dialyzed against buffer D (100 mM KCl, 20 mM Hepes-KOH, pH 7.9,
20% glycerol, 0.2 mM EDTA, 0.2 mM EGTA) for 2 hours at 4°C. For GST-pull downs, beads
were coupled to approximately 5 µg of fusion protein and then incubated for 90 min with 8 µg
DT40 nuclear extract in a 400 µl reaction. Beads were washed 3 times with buffer D containing
150 mM KCl, resuspended in 35 µl SDS-PAGE loading buffer and boiled. Samples were
analyzed by SDS-PAGE followed by Western blotting using anti-cTRF1 polyclonal antibody
and horseradish-peroxidase conjugated secondary. Supersignal West Pico substrate (Pierce) was
35 used for antibody detection.
3.10. Radioligand binding assays
Purified GST-LLEAAR and GST-ALEASA (3.2 µg) were incubated for 1h at room
temperature with 1 µCi radiolabeled ATP, dATP, or NAD in 100 µl binding buffer (20 mM
Hepes – KOH pH 7.5, 10 mM KCl, 3.5 mM MgCl2). Subsequently, the samples were applied to
nitrocellulose filters (Millipore MF 0.45 µm HA), and filtered on a vacuum manifold (Milliport
1225 Sampling Manifold). After two washes with 1 ml binding buffer the filters were dried and
bound nucleotides determined by liquid scintillation counting.
3.11. Yeast two-hybrid analysis of SAM self-association
For quantification of SAM Table 1 Yeast two-hybrid vectors plasmid use marker domain self-association,
bait vector, galactose induced constructs pGilda-mSARP2 pGILDA HIS3 expression and pJG4-5-mSARP2 were bait vector, expression not pEG202 HIS3 transformed into the yeast regulated target/library vector, galactose strain EGY48/pSH18-34, pJG4-5 TRP1 induced expression which harbours a β- β-galactosidase reporter pSH18-34 URA3 galactosidase reporter construct for interaction GFP reporter construct for plasmid. Expression was pGNG1 URA3 interaction confirmed by Western blotting
of yeast extracts. For β-galactosidase assays, a single colony was grown overnight in 5 ml of
36 galactose containing medium (inducing protein expression). A 600 µl aliquot of the culture was
used to measure the OD600, the remainder of the cells were pelleted and resuspended in 3 ml of buffer Z (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, 1 mM MgSO4, 50 mM 2-
mercaptoethanol). After adding 3 ml of 10% (w/v) SDS to each ml of resuspended yeast, the
sample was incubated for 30 min at 30 °C. Reactions were started by addition of 200 µl of 4
mg/ml o-nitrophenyl-β-D-galactopyranoside and stopped with 250 µl of 2 M Na2CO3. The
samples were centrifuged briefly and the absorbance measured at 420 and 550 nm. Miller units
of β-galactosidase were determined according to the following formula: (1000 × A420)–(1.75 ×
A550)/(tmin × Vml × A600 × concentration factor), where Vml is the volume of the reaction, tmin is the time of incubation in minutes, and the concentration factor is volume culture/volume resuspension buffer. Data were acquired from at least nine experiments performed in triplicate.
For statistical analysis, data were normalized by log-transformation. Repeated measures analysis of variance was carried out using the SAS (Cary, NC, U.S.A.) procedure PROC MIXED by Dr.
Linda Levin, at the University of Cincinnati Center for Biostatistical Services.
3.12. Gel filtration
Samples were separated on a Superose 6 column (Amersham Pharmacia) at 4°C. Before loading, samples were centrifuged first at 18,000 x g for 15 min and then at 100,000 x g for 30 min.The column was run at 0.35 ml/min, and 350 µl fractions were collected and analyzed by
Western blotting with tankyrase antibody. For analysis of purified MBP-SAM1 and MBP-SAM2
– SAM1 complexes the column was equilibrated with 20 mM Hepes, pH 7, 150 mM NaCl, 1 mM DTT, 200 µl purified protein was loaded on the column (from 10 ml bacterial pellet), and
37 separated at 0.35 ml/min. 20 µl of each fraction was loaded on an SDS-PAGE gel and analyzed
by Western blotting with anti-tankyrase antibody and West-femto substrate.
For analysis of endogenous tankyrase, ~7x106 LMH cells were extracted for 30 min on
ice with buffer E (150 mM NaCl, 50 mM Tris pH 8.0, 0.2 mM EDTA, 5 mM MgCl2, 1% NP40,
0.5 mM PMSF, yeast protease inhibitor cocktail) supplemented with 100 µg/ml α-2- macroglobulin. The column was equilibrated with buffer E supplemented with 100 µg/ml α-2- macroglobulin but without the protease inhibitors. One fifth of the extract was loaded on the column and fractions were TCA precipitated prior to Western blotting (entire fractions were loaded on SDS-PAGE gel). For analysis of overexpressed Myc-tankyrase 1 and Myc-tankyrase1-
∆SAM, 5x106 LMH cells were transfected with 24 µg of DNA and 72 µl TransIT-LT1 (Mirus)
according to the manufacturers instructions. Cells were harvested 24 hrs post transfection and
extracted for 30 min at 25°C with buffer E plus 1mM DTT. The column was equilibrated with
buffer E plus 1mM DTT. Analysis as for MBP-fusion proteins (above).
3.13. Electron microscopy
A drop of purified MBP-SAM1 was adsorbed for 3 minutes to formvar coated 200 mesh
copper grids (EMS). Remaining solution was gently removed with filterpaper. Grids were then
negatively stained with a drop of filtered 1% uranyl acetate for 3 minutes. Digital images were
taken on a JEOL JEM-1230 transmission electron microscope at 80kV equipped with an AMT
Advantage Plus 2K x 2K digital camera. The lengths of 40 MBP-SAM1 rods were determined
from digital electron micrographs of negatively stained MBP-SAM1. The rods were then
grouped into the length categories shown in Figure 17. For proteinase K treatment, MBP-SAM1
38 was incubated with 200 µg/ml proteinase K (Invitrogen) at 42°C for 3.5 hrs in the presence of 5
mM CaCl2.
3.14. Pelleting and dilution assays
Twenty µl of a 12.5 µg/ml solution of purified protein was centrifuged at 18,000 x g for 5
min. Pellets were separated from supernatants and resuspended in the same volume as the
supernatants. The entire samples were analyzed by SDS-PAGE and Western blotting. For the dilution experiment, purified His-tankyrase 1 (50 µg/ml) was diluted 3x, 9x, 27x, 81x and 243x with dilution buffer (300 mM NaCl, 20 mM Hepes pH 7.5, 3 mM MgCl2, 10% glycerol, 1 mM
DTT, 0.001% NP40). After 1 hr at 30°C, 20 µl (or 2 x 20 µl) of each sample was centrifuged for
15 min at 18,000 x g and the following sample volumes were analyzed by SDS PAGE: 3x
dilution, 9 µl; 9x, 27 µl; 27x, 3 µl; 81x, 9 µl,; 243x, 27 µl. Tankyrase was detected by Western
blotting using West-Pico chemiluminescent substrate (Pierce) for the 3x and 9x dilutions and
West-Femto substrate (Pierce) for the remaining dilutions. To determine whether TRF1 interacts
with polymerized tankyrase, His-tagged TRF1 was incubated with or without purified His-tagged
tankyrase1, or tankyrase1-∆SAM, for 1 hr at 37°C in modified buffer D, then centrifuged for 15
min at 18000 x g. To pellet overexpressed protein, LMH cells were transiently transfected with 2
µg FL-tankyrase1, tankyrase1-∆SAM and tankyrase1-PARP†) using TransIT-LT1 (Mirus).
Extracts were prepared at 4°C and 25°C 24 hrs after transfection and analyzed as above. To
study the effect of poly(ADP-ribosyl)ation, purified protein or cell extracts were incubated with
1 mM NAD+ and/or 10 mM 3-amino-benzamide (Sigma) prior to centrifugation.
39 3.15. Poly-(ADP-ribose) binding assay
Proteins (0.5 µg) were immobilized on nitrocellulose using a slot-blot apparatus. First, the membrane was prewetted with TBS, and the apparatus was assembled under vacuum. Proteins were applied in a total volume of 100 µl TBS under gentle vacuum. Next the membrane was removed from the apparatus, and incubated in renaturation buffer (TBS + 0.05 % Tween + 1 mM
DTT) at room temperature for 30 minutes. The membrane was next incubated with 100 µl purified poly-(ADP-ribose) (10 µg / ml in TBS-T-DTT) (Biomol SW-311) for 2 hours at room temperature (cover with parafilm to avoid drying out). Next the membrane was washed 4 times with TBS-T, and further processed as a regular Western blot for poly-(ADP-ribose) using the
West Femto reagent (anti-poly-(ADP-ribose) antibody : 1/5000).
3.16. Tankyrase activity assay
Activity assays were performed in 150 mM NaCl, 20 mM Hepes pH 7.0. 0.02 µg of purified
His-tankyrase 1 or His-tankyrase1-∆SAM were incubated with combinations of the following: 1 mM NAD+, 10 mM 3AB, 0.05 µg TRF1. After 1 hour at 37°C, reactions were stopped by addition of SDS-PAGE loading buffer and ½ of each reaction was used for analysis by Western blotting with anti-poly(ADP-ribose) antibody (1/1000 pico, 1/5000 femto).
3.17. Co-immunoprecipitations
For co-IP experiments, 70% confluent 60 mm dishes of LMH cells were transfected using 25
µl Transit-LT1 (Mirus) according to manufacturers instructions with the following combinations of plasmids: pcDNA3 alone (9.2 µg); Myc-FL-Tankyrase1 (4.6 µg) + Flag-SAM1 (4.6 µg); Myc-
40 Tankyrase1-∆SAM (4.6 µg) + Flag-SAM1 (4.6 µg); Myc-FL-Tankyrase1 (2 µg) + Flag-SAM2
(11 µg); Myc-Tankyrase1-∆SAM (2 µg) + Flag-SAM2 (11 µg). 24 hours after transfection
extracts were prepared by incubation for 30 minutes at room temperature in lysis buffer (50mM
Tris-HCl (pH 8.0), 150mM NaCl, 1% NP40, 5mM MgCl2, 1mM PMSF and yeast protease
inhibitor cocktail). Extracts were clarified by centrifugation (18,000xg, 12 minutes), and
incubated for 1 hour at 4°C with the following antibodies: anti-Myc (100 µl, culture supernatant
9E10), anti-Flag (1 µl, monoclonal M2, Sigma) or anti-α tubulin (1 µl, monoclonal, Sigma).
Next, extracts were incubated 100 µl protein-G-sepharose (10% slurry, Sigma), for 1 hour at 4°C
followed by three washes with lysis buffer. Bound proteins were released by boiling in SDS- page loading dye, and analyzed by western blotting with antibodies against the Myc and Flag tags.
3.18. Immunofluorescence
For immunofluorescence 2x105 LMH cells were grown on collagen coated glass coverslips
(Becton Dickinson) for 24 hours and transfected as described in 2.8. When ready, cells were washed 3x in PBS, fixed 10’ with 4% formaldehyde, washed 2x in PBS, permeabilized 5 minutes with 0.1% Triton X100, washed 2x with PBS, blocked for 30 minutes with either 1% BSA in
PBS (for transfected cells) or 5% normal goat serum in PBS (for detection of endogenous tankyrase), incubated with primary antibodies for 90 min, washed 5x with BSA/PBS, incubated with secondary antibodies for 45 min, stained 5’ with DAPI (2.5 µg/ml ), washed 5x with
BSA/PBS, mounted with antifading reagent (gel/mount, Biomeda), and visualized on a Nikon
E600 epifluorescence microscope using a 100x oil-immersion objective. For analysis of over-
41 expressed proteins, 0.2x106 LMH cells, grown on collagen coated glass coverslips (Becton
Dickinson) were transfected with 2 µg of each construct and 6 µl TransIT-LT1 (Mirus) for each
2 µg of DNA, according to the manufacturers instructions. 4-30 hours after transfection, cells were prepared for immuno fluorescence as described above.
3.19. Yeast two-hybrid screen
Saccharomyces cerevisiae strain EGY-48 (MATα trp1 his3 ura3 leu2::6 LexAop-LEU2)
(Origene) was used for all experiments. Growth media was either YPD (20 g/l peptone, 10 g/l yeast extract, 20 g/l glucose, pH 6.5) or SD (synthetic defined media for growth selection, contains 6.7 g/l yeast nitrogen base without aminoacids, 10 % drop out mix (contains aminoacids: Ala, Arg, Iso, Lys, Met, Phe, Thr, Tyr and Val), carbon source (20 g/l glucose for growth without protein expression or 20 g/l galactose, 10 g/l raffinose for induction of protein expression). SD was supplemented with aminoacids Trp, Ura, Leu or His as necessary. Cells were grown at 30 °C. Yeast transformations were carried out using a standard lithium acetate protocol. The library was plated out onto 24 cm x 24 cm dishes (Fisher) using sterile glass beads
(3 mm, Fisher). Yeast DNA was isolated by vortexing in equal amounts of plasmid rescue solution (2 % Triton X-100, 1% SDS, 2.5 ml, 0.1M NaCl, 10mM Tris-HCl pH 8, 0.2 ml 1mM
EDTA) and phenol-chloroform-isoamyl alcohol, in presence of glass beads, followed by ethanol precipitation. To isolate TRP1 (library) plasmids, E.coli. KC8 cells were transformed with 1 µl of crude yeast DNA and plated on M9 –Trp media, followed by standard miniprep. For mating assays, plasmids were transformed into strain RFY206 (MATa trp1∆::hisG his3∆200 ura3-52
42 lys2∆201 leu2-3) and mated with strain EGY-48 by cross streaking on YPD and replica-plating on plates selective for mated yeast.
43 4. CHARACTERIZATION OF THE CHICKEN TANKYRASES.
4.1. Summary
Tankyrase 1 and 2 are conserved, multifunctional proteins that exhibit high levels of sequence similarity. This complicates understanding their respective roles in the cell and raises questions regarding possible redundancy. To investigate these issues, we decided to use a loss- of-function approach by creating a knock-out cell line that lacks tankyrase 2. For this we used the chicken DT40 cell line because it exhibits high levels of recombination, and thus allows efficient gene targeting. Since only human tankyrases had been studied, we first characterized the chicken tankyrase proteins. Cross-species sequence comparison indicated that tankyrase domain structure is highly conserved, and supports division of the ankyrin domain into five subdomains which are each separated by a conserved LLEAAR/K motif. Glutathione S- transferase pull-down experiments demonstrated that the ankyrin domains of both tankyrase 1 and tankyrase 2 interact with chicken telomere repeat factor 1 (TRF1). Analysis of total cellular and nuclear proteins revealed that cells contain approximately twice as much tankyrase 1 as tankyrase 2. Although ≥ 90% of each protein is present in the cytoplasm, both tankyrase 1 and 2 were detected in the nucleus. The nuclear location and the ability to interact with TRF1, point to a telomeric function for tankyrase 2. Unfortunately, knocking out the tankyrase 2 gene proved problematic; the DT40 cell line contains three tankyrase 2 alleles, one of which appeared to have different sequence upstream of the start codon and could not be removed.
44 4.2. Introduction
Tankyrase 1 and tankyrase 2 show a significant functional overlap, both in vitro and when
overexpressed in cells (Cook et al., 2002; Kaminker et al., 2001; Sbodio et al., 2002). However,
it is still unclear whether they are present in identical subcellular locations and hence whether
they always function in the same cellular pathways. One approach to address these questions is
to knock-out either tankyrase 1 or tankyrase 2 in a model animal or cell line. The ensuing
phenotypes can then reveal the processes that the genes are involved in and whether they are
essential. In addition, knock-out cell-lines allow the study of structure-function relationships
through reintroduction of mutant alleles. The chicken DT40 cell-line is a powerful system for
genetic, loss of function studies. These transformed B-cells exhibit high levels of homologous
recombination, which allow for highly efficient targeting of genes (Buerstedde and Takeda,
1991; Winding and Berchtold, 2001). Because of its comparative ease of use, the DT40 system is
being used by an increasing number of investigators to create knock-outs and to study gene
function in a variety of different fields (reviewed in (Brown et al., 2003)). Since the function of
tankyrases remains poorly understood we decided to perform knock-out studies in DT40 to provide insight into their roles in the cell. Chicken tankyrases had not previously been characterized, therefore it was important to perform an initial characterization of these proteins and to compare them to their human counterparts.
To gain insight into tankyrase 1 and tankyrase 2 redundancy, an important question has been whether both proteins localize to telomeres and exert the same functions there. Such studies should ideally be carried out on endogenous proteins. However, this has turned out to be difficult because the high sequence identity between tankyrase 1 and tankyrase 2 has prevented the
45 generation of a tankyrase 2 specific antibodies. Thus, in vitro experiments were performed to
study the interaction between human tankyrase 2 and TRF1. While most reports show that both proteins can interact, others could not detect this interaction (Kaminker et al., 2001; Lyons et al.,
2001; Sbodio and Chi, 2002). In subsequent studies, overexpression of human tankyrase 2 was shown to remove TRF1 from the telomere, as had previously been shown for tankyrase 1 (Cook et al., 2002; Smith and de Lange, 2000). However, long-term experiments looking at the effect of tankyrase 2 overexpression on telomere length have not been carried out. Together, these
observations suggested that tankyrase 2 could indeed function at telomeres. To investigate this
possibility, we examined the subcellular localization of chicken tankyrases and investigated whether they can interact with TRF1.
Here, we describe the isolation of the chicken tankyrase homologs, a cross-species sequence comparison and an initial biochemical characterization of these proteins. Our analysis supports the division of the ankyrin domain into 5 sub-domains separated by a highly conserved motif which may have ATP binding activity. When fractionation experiments compared the amount of tankyrase 1 and 2 in cells and nuclei, a significant proportion of each protein was present in the nuclear fraction. Since both proteins interact with chicken TRF1, our results indicate that tankyrase 2 is likely to have a telomeric function. Overexpression studies using tankyrase 2 revealed coregulation of tankyrase 1 expression, suggesting that the two proteins interact at some level. Finally, we describe our progress towards the creation of a tankyrase 2 knock-out cell-line.
46 4.3. Results
4.3.1. Identification of chicken tankyrase genes
Portions of the human tankyrase 1 cDNA were used as probes to screen a chicken embryonic fibroblast cDNA library for chicken tankyrase 1. Because the PARP domains of human tankyrase 1 and 2 are highly conserved, this region of chicken tankyrase 1 was then used as a probe to isolate the tankyrase 2 gene. Sequencing of the chicken tankyrase 1 and 2 cDNAs revealed that the encoded proteins share a high level of sequence identity with their human homologs and have the same overall domain structure (Figure 5). The 5' end of chicken tankyrase 1 is highly GC rich and like human tankyrase 1, encodes a homopolymeric region that contains tracts of histidines, prolines and serines (the HPS domain). Although the chicken tankyrase 2 gene also has a highly GC-rich stretch at its 5' end, analysis of the cDNA sequence isolated in the library screen indicated that this region was unlikely to be translated and hence to encode an HPS domain. As illustrated in Figure 6, the cDNA sequence has a stop codon at
cTank1 1 MAAPPRRSQHHHHHHGP..PPPPGPASPPAAASPPRSPSLAP...... AELGP.AAQ hTank1 1 MAA.SRRSQHHHHHHQQQLQPAPGASAPPPPPPPPLSPGLAPGTTPASPTASGLAPFASP cTank2 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ hTank2 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ cTank1 49 RHSLAGPEGEA...PPDAERPPAPE...... CSEGAAP...... GPPPG hTank1 60 RHGLALPEGDGSRDPPDRPRSPDPVDGTSCCSTTSTICTVAAAPVVPAVSTSSAAGVAPN cTank2 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ hTank2 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ cTank1 83 .SGSSSGSSASSSSSSSSTSSSVASSPA...AESPEAAG...... P.. hTank1 120 PAGSGSNNSPSSSSSPTSSSSSSPSSPGSSLAESPEAAGVSSTAPLGPGAAGPGTGVPAV cTank2 23 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~MAARRCAGGAAALAEAPGCG.SAV hTank2 12 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~MSGRRCAGGGAACASA..AA.EAV cTank1 119 SGAFRELLEACRNGDVTRVKRLVDAGNVNAKDMAGRKSTPLHFAAGFGRKDVVEHLLQTG hTank1 180 SGALRELLEACRNGDVSRVKRLVDAANVNAKDMAGRKSSPLHFAAGFGRKDVVEHLLQMG cTank2 79 EP.ARELFEACRNGDVERVKRLVRPENVNSRDTAGRKSSPLHFAAGFGRKDVVEYLLQSG hTank2 68 EPAARELFEACRNGDVERVKRLVTPEKVNSRDTAGRKSTPLHFAAGFGRKDVVEYLLQNG
47 cTank1 179 ANVHARDDGGLIPLHNACSFGHAEVVSLLLCQGADPNARDNWNYTPLHEAAIKGKIDVCI hTank1 240 ANVHARDDGGLIPLHNACSFGHAEVVSLLLCQGADPNARDNWNYTPLHEAAIKGKIDVCI cTank2 138 ANVHARDDGGLIPLHNACSFGHAEVVNLLLRHGANPNARDNWNYTPLHEAAIKGKTDVCI hTank2 128 ANVQARDDGGLIPLHNACSFGHAEVVNLLLRHGADPNARDNWNYTPLHEAAIKGKIDVCI cTank1 239 VLLQHGADPNIRNTDGKSALDLADPSAEAVLTGEYKKDELLEAARSGNEEKLMALLTPLN hTank1 300 VLLQHGADPNIRNTDGKSALDLADPSAKAVLTGEYKKDELLEAARSGNEEKLMALLTPLN cTank2 198 VLLQHGAEPTIRNTDGRTALDLADPSAKAVLTGEYKKDELLESARSGNEEKMMSLLTPLN hTank2 188 VLLQHGAEPTIRNTDGRTALDLADPSAKAVLTGEYKKDELLESARSGNEEKMMALLTPLN cTank1 299 VNCHASDGRKSTPLHLAAGYNRVRIVQLLLQHGADVHAKDKGGLVPLHNACSYGHYEVTE hTank1 360 VNCHASDGRKSTPLHLAAGYNRVRIVQLLLQHGADVHAKDKGGLVPLHNACSYGHYEVTE cTank2 258 VNCHASDGRKSTPLHLAAGYNRVKIVQLLLQHGADVHAKDKGDLVPLHNACSYGHYEVTE hTank2 248 VNCHASDGRKSTPLHLAAGYNRVKIVQLLLQHGRDVHAKDKGDLVPLHNACSYGHYEVTE cTank1 359 LLLKHGACVNAMDLWQFTPLHEAASKNRVEVCSLLLSHGADPTLVNCHGKSAVDMAPTPE hTank1 420 LLLKHGACVNAMDLWQFTPLHEAASKNRVEVCSLLLSHGADPTLVNCHGKSAVDMAPTPE cTank2 318 LLVKHGACVNAMDLWQFTPLHEAASKNRVEVCSLLLSYGADPTLLNCHNKSTIDLAPTPQ hTank2 308 LLVKHGGCVNAMDLWQFTPLHEAASKNRVEVCSLLLSYGADPTLLNCKNKSAIDLAPTPQ cTank1 419 LRERLTYEFKGHSLLQAAREADLAKVKKTLALEIINFKQPQSHETALHCAVAAVHPKRKQ hTank1 480 LRERLTYEFKGHSLLQAAREADLAKVKKTLALEIINFKQPQSHETALHCAVASLHPKRKQ cTank2 378 LKERLAYEFKGHSLLQAARESDVARIKKHLSLETVNFKHPQTHETALHCAAASPYPKRKQ hTank2 368 LKERLAYEFKGHSLLQAAREADVTRIKKHLSLEMVNFKHPQTHETALHCAAASPYPKRKQ cTank1 479 VTELLLRKGANVNEKNKDFMTPLHVAAEKAHNDVMEVLHKHGAKMNALDTLGQTALHRAA hTank1 540 VTELLLRKGANVNEKNKDFMTPLHVAAERAHNDVMEVLHKHGAKMNALDTLGQTALHRAA cTank2 438 VCELLLRKGANINEKTKDFLTPLHVASEKAHNDVVEVVVKHEAKVNALDNLGQTSLHRAA hTank2 428 ICELLLRKGANINEKTKEFLTPLHVASEKAHNDVVEVVVKHEAKVNALDNLGQTSLHRAA cTank1 539 LAGHLQTCRLLLNYGSDPSIISLQGFTAAQIGNEAVQQILSESTPVRTSDVDYRLLEASK hTank1 600 LAGHLQTCRLLLSYGSDPSIISLQGFTAAQMGNEAVQQILSESTPIRTSDVDYRLLEASK cTank2 498 HCGHLQTCRLLLSSGCDPSIVSLQGFTALQMGNESVQQLLQEGIPLGNSDADRQLLEAAK hTank2 488 YCGHLQTCRLLLSYGCDPNIISLQGFTALQMGNENVQQLLQEGISLGNSEADRQLLEAAK cTank1 599 AGDLETVKQLCSPQNVNCRDLEGRHSTPLHFAAGYNRVSVVEYLLHHGADVHAKDKGGLV hTank1 660 AGDLETVKQLCSSQNVNCRDLEGRHSTPLHFAAGYNRVSVVEYLLHHGADVHAKDKGGLV cTank2 558 AGDVDTVKKLCTVQSVNCRDIEGRQSTPLHFAAGYNRVSVVEYLLQHGADVHAKDKGGLV hTank2 548 AGDVETVKKLCTVQSVNCRDIEGRQSTPLHFAAGYNRVSVVEYLLQHGADVHAKDKGGLV cTank1 659 PLHNACSYGHYEVAELLVRHGASVNVADLWKFTPLHEAAAKGKYEICKLLLKHGADPTKK hTank1 720 PLHNACSYGHYEVAELLVRHGASVNVADLWKFTPLHEAAAKGKYEICKLLLKHGADPTKK cTank2 618 PLHNACSYGHYEVAELLVKHGAVVNVADLWKFTPLHEAAAKGKYEICKLLLQHGADPPKK hTank2 608 PLHNACSYGHYEVAELLVKHGAVVNVADLWKFTPLHEAAAKGKYEICKLLLQHGADPTKK cTank1 719 NRDGNTPLDLVKEGDTDIQDLLRGDAALLDAAKKGCLARVQKLCTPENINCRDTQGRNST hTank1 780 NRDGNTPLDLVKEGDTDIQDLLKGDAALLDAAKKGCLARVQKLCTPENINCRDTQGRNST cTank2 678 NRDGNTPLDLVKDGDTDIQDLLRGDAALLDAAKKGCLARVKKLCSPDNVNCRDTQGRHST hTank2 668 NRDGNTPLDLVKDGDTDIQDLLRGDAALLDAAKKGCLARVKKLSSPDNVNCRDTQGRHST cTank1 779 PLHLAAGYNNLEVAEYLLEHGADVNAQDKGGLIPLHNAASYGHVDIAALLIKYNTCVNAT hTank1 840 PLHLAAGYNNLEVAEYLLEHGADVNAQDKGGLIPLHNAASYGHVDIAALLIKYNTCVNAT cTank2 738 PLHLAAGYNNLEVAEYLLQHGADVNAQDKGGLIPLHNAASYGHVDVAALLIKYNACVNAT hTank2 728 PLHLAAGYNNLEVAEYLLQHGADVNAQDKGGLIPLHNAASYGHVDVAALLIKYNASLNAT
48 cTank1 839 DKWAFTPLHEAAQKGRTQLCALLLAHGADPTMKNQEGQTPLDLATADDIRALLIDAMPPE hTank1 900 DKWAFTPLHEAAQKGRTQLCALLLAHGADPTMKNQEGQTPLDLATADDIRALLIDAMPPE cTank2 798 DKWAFTPLHEAAQKGRTQLCALLLAHGADPTLKNQEGQTPLDLVTADDVSALLTAAMPPS hTank2 788 DKWAFTPLHEAAQKGRTQLCALLLAHGADPTLKNQEGQTPLDLVSADDVSALLTAAMPPS cTank1 899 ALPTCFKLQATVVSAS...... LISPASTPSCLSAASSIDNLTGPLAELAVGGASNTGD hTank1 960 ALPTCFKPQATVVSAS...... LISPASTPSCLSAASSIDNLTGPLAELAVGGASNAGD cTank2 858 ALPSCYKPQVISVSQTASSTADPLSSVPSSPSSLSAASSLDNLSGSFSELPSVVGTNSAE hTank2 848 ALPSCYKPQVLNGVRSPGATADALSSGPSSPSSLSAASSLDNLSGSFSELSSVVSSSGTE cTank1 952 GAAGTERKEGEVSGLDMNITQFLKSLGLEHLRDIFETEQITLDVLADMGHEELKEIGINA hTank1 1013 GAAGTERKEGEVAGLDMNISQFLKSLGLEHLRDIFETEQITLDVLADMGHEELKEIGINA cTank2 918 GATVLEKK..EVSGVDFSINQFVRNLGLEHLIDIFEREQITLDVLVEMGHKELKEIGINA hTank2 908 GASSLEKK..EVPGVDFSITQFVRNLGLEHLMDIFEREQITLDVLVEMGHKELKEIGINA cTank1 1012 YGHRHKLIKGVERLLGGQQGTNPYLTFHCVSQGTILLDLAPDDKEYQSVEEEMQSTIREH hTank1 1073 YGHRHKLIKGVERLLGGQQGTNPYLTFHCVNQGTILLDLAPEDKEYQSVEEEMQSTIREH cTank2 976 YGHRHKIIKGVERLISGQQGLNPYLTLNTSSSGTLLIDLSSEDKEFQSVEEEMQSTVREH hTank2 966 YGHRHKLIKGVERLISGQQGLNPYLTLNTSGSGTILIDLSPDDKEFQSVEEEMQSTVREH cTank1 1072 RDGGNAGGIFNRYNVIRIQKVVNKKLRERFCHRQKEVSEENHNHHNERMLFHGSPFINAI hTank1 1133 RDGGNAGGIFNRYNVIRIQKVVNKKLRERFCHRQKEVSEENHNHHNERMLFHGSPFINAI cTank2 1036 RDGGHAGGVFNRYNILKIQKVCNKKLWERYTHRRKEVSEENHNHANERMLFHGSPFVNAI hTank2 1026 RDGGHAGGIFNRYNILKIQKVCNKKLWERYTHRRKEVSEENHNHANERMLFHGSPFVNAI cTank1 1132 IHKGFDERHAYIGGMFGAGIYFAENSSKSNQYVYGIGGGTGCPTHKDRSCYICHRQMLFC hTank1 1193 IHKGFDERHAYIGGMFGAGIYFAENSSKSNQYVYGIGGGTGCPTHKDRSCYICHRQMLFC cTank2 1096 IHKGFDERHAYIGGMFGAGIYFAENSSKSNQYVYGIGGGTGCPIHKDRSCYVCHRQLLFC hTank2 1086 IHKGFDERHAYIGGMFGAGIYFAENSSKSNQYVYGIGGGTGCPVHKDRSCYICHRQLLFC cTank1 1192 RVTLGKSFLQFSTMKMAHAPPGHHSVIGRPSVNGLAYAEYVIYRGEQAYPEYLITYQIVK hTank1 1253 RVTLGKSFLQFSTMKMAHAPPGHHSVIGRPSVNGLAYAEYVIYRGEQAYPEYLITYQIMK cTank2 1156 RVTLGKSFLQFSAMKMAHSPPGHHSVTGRPSVNGLALAEYVIYRGEQAYPEYLITYQIVK hTank2 1146 RVTLGKSFLQFSAMKMAHSPPGHHSVTGRPSVNGLALAEYVIYRGEQAYPEYLITYQIMR cTank1 1252 PEAPSQTGTAAEQKT* hTank1 1313 PEAPSQTATAAEQKT* cTank2 1216 PEATTEA* hTank2 1206 PEGMVD.G*
Figure 5 Amino acid sequence comparison of tankyrase 1 and tankyrase 2 from chicken and humans. Dark shading delineates identical amino acids, gray shading marks similar amino acids. HPS domain, amino acids 1-115 (chicken Tankyrase 1), Ankyrin repeat domain, amino acids 120-946; SAM domain, aa 962-1027; SAM-PARP linker, aa 1028-1114; PARP domain aa 1115-1155.
49 nucleotide 120 which is 51 nt upstream of the first AUG at position 171. The apparent lack of an
HPS domain in human tankyrase 2 had been somewhat controversial because the human cDNA has a long stretch of in-frame sequence immediately 5' of the putative start codon suggesting the initiating methionine might be missing (Kaminker et al., 2001; Lyons et al., 2001). We therefore confirmed the presence of the nonsense codon in the chicken tankyrase 2 gene by performing
RT-PCR directly on chicken mRNA using primers that flanked both the putative stop codon and translation start site. Sequencing of the resulting PCR product confirmed the presence of the stop codon in frame with the start codon. The stop codon is also apparent in the recently released
102 120 171 cTank2F1 GCGGGGAGCCGCGGCCTGTGAGCGGGGCAGCGAGGGCCGCTCGGCCGCCTCCGGGGTGTCCGACCCGGCATGGCGGCG a g s r g l * a g q r g p l g r l r g v r p g M A A cTank2cDNA
Figure 6 RT-PCR strategy used to confirm the presence of a stop codon (*) upstream of the putative start codon (M). cDNA was synthesized using the cTank2cDNA primer and then amplified with the cTank2F1 and cTank2cDNA primers. The sequence of the PCR product is shown with the stop and start codons shaded. genomic sequence of tankyrase 2. Together these observations indicate that chicken tankyrase 2 does indeed lack an HPS domain. Given the high sequence identity between the chicken and human genes (see below), our results indicate that an HPS domain is also unlikely to be present in human tankyrase 2.
4.3.2. Conservation of tankyrase proteins
Comparison of the tankyrase nucleotide sequences using the GCG Gap algorithm revealed that chicken and human tankyrase 1 genes are 80% identical while the tankyrase 2 genes are 81% identical. Most of the sequence differences are at the wobble positions so the encoded tankyrase 1 proteins are 95% identical while the tankyrase 2 proteins are 93% identical
50 (Figure 7). Interestingly, the overall identity between the two members of the tankyrase family
is lower (81% between chicken tankyrase 1 and 2 and 78% between human tankyrase 1 and 2)
with the SAM domains exhibiting the lowest level of identity (77% between either pair of
paralogs). This lower sequence identity together with the lack of an HPS domain on tankyrase 2,
suggests that the two proteins may have
acquired different functions during evolution.
BLAST searches of Genbank and the
NCBI EST database identified sequences from a
number of other organisms (Mouse, Xenopus,
Pufferfish, Zebrafish, Catfish, Mosquito and Drosophila) that encoded portions of tankyrase Figure 7 Sequence conservation between human and chicken tankyrases. Numbers indicate % aa sequence identity between proteins (shown at left) or between homologs. The vertebrate homologs could individual domains. clearly be identified as tankyrase 1 or 2 based on
the level of sequence identity to the human or chicken proteins. For example, a Xenopus EST
encoded a peptide that was 94% identical to the PARP domain from chicken tankyrase 2 but only
86% identical to chicken tankyrase 1, while a Pufferfish EST encoded a peptide that was 91%
identical to the PARP domain from tankyrase 1 but only 85% identical to tankyrase 2.
Interestingly, only one Drosophila tankyrase sequence could be identified even though the entire
genome has been sequenced. The encoded protein shares a significant but much lower level of
sequence identity (~ 66%) to tankyrase 1 and 2 from humans and chickens. Like tankyrase 2, the
Drosophila protein lacks an HPS domain. However, it does not appear to be more closely related
to tankyrase 2 than to tankyrase 1 as the overall sequence identity to both vertebrate proteins is
51 quite similar. The existence of two
highly conserved tankyrase genes in
vertebrates, but only one more diverged
gene in Drosophila, suggests that the
vertebrate and insect genes evolved
from a common ancestor, with a gene
duplication occurring during vertebrate
evolution to give rise to tankyrase 1 and
2.
4.3.3. Partitioning of the
ankyrin repeats into five
subdomains
Figure 8 Sequence comparison of the LLEAAR/K motifs from In addition to providing insight tankyrase 1 and 2. Each sequence alignment corresponds to one of the four LLEAAR/K sequence motifs found in human, mouse and chicken tankyrase 1, human and chicken tankyrase 2 and Drosophila into the evolution of tankyrase genes, tankyrase. The sequence of the equivalent LLEAAR/K motifs from Xenopus, Catfish (Ictalurus punctatus) and Zebrafish (Dario rerio) are the interspecies sequence comparison included where available. Dark shading highlights identical amino acids, gray shading highlights similar amino acids. The consensus yielded information about the sequence for the motif is given at the bottom. organization of the ankyrin repeat
domain. When the human tankyrase 1 gene was first identified, sequence analysis suggested that
the ankyrin domain contained 24 ANK repeats (Smith et al., 1998). However, a subsequent
analysis indicated that the domain might instead contain 19 full ANK repeats with two half
repeats at either end (Sbodio et al., 2002). This later alignment had fewer deletions and insertions
52 and most importantly revealed a four-repeat periodicity. This periodicity results from an ~22
amino acid insertion after every fourth ANK repeat with each insertion bearing a variant of the
sequence LLEAAR/K, a motif that is a poor match to the ANK-repeat consensus (Sedgwick and
Smerdon, 1999). The insertions cause the ankyrin domain to be subdivided into five approximately equal segments that each contain approximately four repeats. These segments correspond fairly well to the subdomains of ANK repeat clusters that have been identified as separate and redundant binding sites for the tankyrase interacting proteins TRF1 and TAB182
(Seimiya and Smith, 2002). Our comparison of ANK repeat sequences from vertebrate and insect tankyrases indicates that both the sequence and position of the LLEAAR/K motif within the ankyrin domain is highly conserved (Figure 8). Thus, comparative sequence analysis supports division of the ankyrin repeat domain into five subdomains that provide alternative binding sites for tankyrase 1 and tankyrase 2-interacting proteins.
Table 2 Nucleotide binding properties of proteins containing LLEAAR/K motifs.
# of proteins in # of proteins that bind indicated nucleotide (A = NAD or Motif database with ATP or FAD or AMP) motif NAD ATP FAD cAMP A LLEAAR 14 5 1 1 0 7 LLQAAR 12 0 5 1 1 7 LLDAAK 17 2 2 0 1 5 LLEAAK 15 2 4 0 0 6 LLEASK 10 2 3 1 0 6 LLESAR 8 1 4 0 0 5 All 76 12 19 3 2 36
53 4.3.4. Nucleotide binding properties of the LLEAAR/K motif
A database-search using all variants of the LLEAAR/K motif as a query sequence
revealed that this motif is present in more than 70 proteins from organisms as diverse as bacteria and mammals, most of which are metabolic enzymes such as dehydratases, dehydrogenases and epimerases. Strikingly, close to 50% of these proteins have been shown to bind the nucleotides
NAD+, ATP, FAD or cAMP and the actual percent may be higher as little information is
available concerning the enzymatic mechanism of the remaining LLEAAR/K-containing
proteins (Table 2). This observation suggests that the LLEAAR/K-containing inserts which subdivide the ankyrin domain of 250000 ATP tankyrase 1 and 2 might provide an 200000 dATP GTP adenosine-interacting domain that is 150000 separate from the NAD-binding region of 100000 bound nucleotide (CPM) bound nucleotide the PARP domain. To determine whether 50000
0 GST GST-Tankyrase 1 GST-Tankyrase 1 this is indeed the case, we performed LLEASK ALEASA nucleotide binding assays with purified Figure 9 Radionucleotide binding assays with a tankyrase 1 fusion proteins and radiolabeled GST fusion protein containing one LLEAAR/K motif. Results are of three independent experiments, error bars represent standard deviation. nucleotides. A tankyrase 1 GST-fusion protein containing one LLEAAR/K motif (LLEASK) flanked by 2 ANK repeats at each side was used for these assays. As shown on Figure 9 the LLEASK containing protein bound ATP about
10 fold better than an equal amount of GST alone. Other nucleotides such as dATP, dGTP and
NAD (not shown) did not show strong binding. To show that the LLEASK motif was indeed
responsible for mediating ATP retention, we made a mutant construct that contained the
54 sequence ALEASA instead of LLEASK. Binding of ATP to the mutant protein was about half of binding to the wild type sequence (Figure 9), indicating that the LLEAAR/K motif indeed is involved in ATP binding. However, to unequivocally show specific binding, saturation of the binding site should be demonstrated. Unfortunately binding curve experiments to show saturation did not yield conclusive results. While a level of saturation could be observed, high levels of non-specific binding and lack of reproducibility made these results inconclusive. Hence further analyses with other types of binding assays are necessary to confirm that tankyrases indeed have an ATP binding motif.
4.3.5. Chicken tankyrases interact with cTRF1
The prominent cytoplasmic localization of human tankyrase 1 (Chi and Lodish, 2000;
Cook et al., 2002) and tankyrase 2 (Lyons et al., 2001; Sbodio et al., 2002) together with initial reports that human tankyrase 2 does not interact with hTRF1 (Lyons et al., 2001), led us to ask whether chicken tankyrase 1 and 2 interact with chicken TRF1 (cTRF1). In order to detect cTRF1 in pull down experiments it was necessary to first characterize the chicken TRF1 gene and generate antibodies to the protein. Analysis of a cTRF1 cDNA sequence (kindly provided by
D. Broccoli, B. Li and T. de Lange) showed that the human and chicken TRF1 proteins are 46% identical (56% similar) and share the same domain structure (Figure 10A) (Broccoli et al.,
1997a). In contrast, chicken TRF1 and TRF2 are only 22% identical (30% similar). We used the cDNA to produce recombinant cTRF1 in E. coli and then made antibody to the purified protein.
As shown in Figure 10B, the affinity-purified antibody detected only one band in nuclear extracts from chicken DT40 cells and this band had the same mobility as bacterially expressed
55 cTRF1 (lanes 1-4). The apparent
molecular weight of ~39 kDa corresponds
well with the predicted size of the protein
(354 amino acids or 41 kDa). To further
confirm the specificity of the cTRF1
antibody, we transfected DT40 cells with
FLAG-tagged cTRF1 and performed
Western blots on nuclear extracts from
the transfected cells. As shown in Figure
10B lanes 6 and 8, the anti-FLAG and
anti-cTRF1 antibodies detected a protein
Figure 10 Interaction of chicken tankyrase 1 and 2 with of identical mobility. Chicken TRF1 is chicken TRF1. A. Schematic showing TRF1 domain structure and sequence identity between the chicken and human proteins. The numbers show % amino acid identity significantly smaller than its human between each domain. B. Western blots showing specificity of cTRF1 antibody. Lane 1, nuclear extract from 0.25 x 106 counterpart (354 versus DT40 cells; lane 2, purified recombinant His-tagged cTRF1 expressed in E. coli; lane 3, 1 x 106 DT40 nuclei; lane 4, mix of DT40 nuclear extract and recombinant cTRF1. Lanes 5-8, 439 amino acids) with most of the size 2 x 104 DT40 cells probed with anti-cTRF1 antibody (lanes 5 & 6) or and anti-FLAG antibody (lanes 7 & 8). Lanes 5 & 7, difference concentrated in the N-terminal wild type DT40; lanes 6 & 8, DT40 cells expressing FLAG- tagged cTRF (gel courtesy of Chao Wei). C. Western blot tankyrase-interacting domain and the showing cTRF1 associated with glutathione beads coupled to GST or GST-tankyrase fusion protein after incubation with DT40 nuclear extract. Lane 1, input DT40 nuclear hinge domain (Figure 10A) (Chong et al., extract; lane 2, GST-beads + nuclear extract; lanes 3 & 4, GST-tankyrase1 fusion protein (ANK repeats 10-18) 1995). We suspect the cDNA clone is incubated without (lane 3) or with (lane 4) nuclear extract; lanes 4 & 5, GST-tankyrase 2 fusion protein (ANK repeats 10- 18) incubated without (lane 5) or with (lane 6) nuclear extract. missing the extreme 5’ end of the gene, Positions and sizes of molecular weight markers are shown to the left. cTRF1 is marked by the arrowhead. but since recombinant and endogenous cTRF1 have essentially the same mobility in an SDS gel, the missing region probably amounts to
56 only a few amino acids.
Because the cTRF1 clone may be missing part of the tankyrase-interacting domain, we chose to look for an interaction between cTRF1 and tankyrase using the endogenous chicken
TRF1 and recombinant GST-tankyrase fusion proteins. The fusion proteins contained a section of the chicken tankyrase 1 or 2 ankyrin repeats that corresponds to one of the TRF1-interacting domains in human tankyrase 1. Each fusion protein was bound to glutathione beads, the beads incubated with nuclear extracts from chicken DT40 cells, washed and analyzed by Western blotting with cTRF1 antibody. As shown in Figure 10C, cTRF1 was pulled down by both the
GST-tankyrase 1 and the GST-tankyrase 2 beads (lanes 4 & 6) but not by control beads bound to
GST alone. These results show that both chicken tankyrase 1 and 2 interact with cTRF1 through the ankyrin repeat domain. This has since been shown to also be the case for human tankyrase 2
(Cook et al., 2002; Kaminker et al., 2001).
4.3.6. Subcellular distribution of tankyrase 1 and 2
A key step towards determining whether tankyrase 2 has a telomeric function is to find out whether any of the endogenous protein is present in the nucleus. Although previous fractionation experiments had demonstrated that a large percentage of tankyrase 2 is present in low density microsomes (Lyons et al., 2001), they did not determine whether any of the protein is present in purified nuclei. To address this question, we purified nuclei from chicken DT40 cells and assayed for the presence of tankyrase 2 by Western blotting using an affinity-purified antibody made against the SAM-domain of tankyrase 1. As shown below, this antibody cross- reacts with the SAM domain of both tankyrase 1 and 2. We also used the SAM domain antibody
57
Figure 11 Relative abundance of tankyrase 1 and 2 in whole cells or nuclei. A. Demonstration of tankyrase antibody specificity. Western blots probed with tankyrase antibody. Lane 1, tankyrase 1 in vitro transcription translation product; lane 2, DT40 nuclear extract; lane 3, tankyrase 2 in vitro transcription translation product; lane 4, luciferase control in vitro transcription translation reaction. B. Detection of tankyrase 1 and 2 in chicken cells and nuclei. Upper panel; western blots of total cellular or nuclear proteins from DT40 cells probed with tankyrase antibody. The number of cells or nuclei (x 105) loaded in each lane is given at the top of each blot. Middle panel; longer exposure of the upper panel. The arrow marks tankyrase 1, the arrowhead marks tankyrase 2. Lower panel; western blots of chicken LMH hepatocellular carcinoma cells or nuclei probed with the tankyrase antibody. The same number of cells or nuclei were loaded as in the upper panels. C. Assay for tubulin contamination in purified nuclei. Western blots probed with antibody to tubulin, histone H3 or tankyrase. Lane 1, 3x105 DT40 cells; lane 2, 3x104 cells; lane 3, 3x106 nuclei. D. Affinity of chicken tankyrase antibody for the SAM domains of tankyrase 1 and 2. Western blot showing reaction of the tankyrase antibody with decreasing amounts of bacterially expressed SAM domains from tankyrase 1 or 2. The relative amount of protein loaded is shown above each lane, equal amounts were loaded in lanes 1 & 6, 2 & 7, 3 & 8. to quantify the relative amounts of tankyrase 1 and 2 in cells, as this information is needed to
assess which is the most prevalent poly(ADP-ribosyl)ating activity. To demonstrate that the
tankyrase antibody cross-reacts with both tankyrase 1 and tankyrase 2, the corresponding cDNAs
were subjected to in vitro transcription/translation and the products analyzed by Western
blotting. As shown in Figure 11A (lanes 1 & 3), the antibody detected both proteins. When total
cellular proteins and proteins from purified nuclei were separated by SDS-PAGE and Western
blotted with the tankyrase antibody, two bands in the 120-140 kDa range were detected in both
preparations (Figure 11A, lane 2 and Figure 11B). These ran at the sizes expected for tankyrase 1
58 (139 kDa) and tankyrase 2 (129 kDa) and aligned with products obtained after in vitro
transcription and translation of the chicken tankyrase 1 and 2 cDNAs (Figure 11A, lanes 1-3).
The 129 kDa band was unlikely to be a degradation product of tankyrase 1 that fortuitously ran
with the same mobility as tankyrase 2 because it was observed when whole cells were lysed
directly by boiling in SDS. Although the same relative amount of
tankyrase 2 was observed in multiple nuclear preparations, the amount was quite low.
Consequently, we were concerned that the tankyrase 2 in our preparations might merely reflect cytoplasmic contamination. To control for this, preparations of cellular and nuclear proteins were probed with antibodies to tubulin, histone H3 and tankyrase (Figure 11C). No tubulin could be detected in samples containing 3 x 106 nuclei (Figure 11C, lane 3) even though tankyrase and
histone H3 were readily apparent. Moreover, tubulin could be detected in samples containing
100 fold fewer cells (Figure 11C, lane 1), indicating that there was less than 1% cytoplasmic
contamination in the nuclear samples. Thus, our results indicate that a fraction of the endogenous
tankyrase 2 is most likely present in the nucleus.
To compare the amounts of tankyrase 1 and tankyrase 2 in whole cells, we first
determined the relative affinity of the antibody for the SAM domains of the two proteins. When
Western blots were performed with various amounts of bacterially expressed SAM domain from tankyrase 1 (SAM1) or tankyrase 2 (SAM2), a 1:2 ratio of SAM1:SAM2 was required to give the same signal intensity (Figure 11D). We therefore estimate that the antibody has approximately twice the affinity for tankyrase 1 over tankyrase 2. When we compared the number of cells required to give equivalent tankyrase 1 and 2 signals, we found that about four times more cells were required to give a tankyrase 2 signal that was of similar intensity to the
59 tankyrase 1 signal (compare Figure 11B lane 1, Tank1 signal with 0.75 x105 cells to lane 3,
Tank2 signal with 3 x 105 cells). Given the two-fold difference in antibody affinity for tankyrase
1 versus tankyrase 2, this result indicates that there is approximately two fold more tankyrase 1 than tankyrase 2 in DT40 cells.
Comparison of the tankyrase 1 signals obtained with whole DT40 cells versus nuclei indicated that approximately one tenth of the total protein is present in the nucleus (compare
Figure 11B lanes 2 and 5 or lanes 3 and 6). A similar result was obtained when we examined the amount of tankyrase 1 in whole cells and nuclei from the chicken LMH hepatocellular carcinoma
cell line (Figure 11B, lower panel), indicating that this distribution was a general phenomenon
and not specific to the DT40 B-cell line. We were not able to get an accurate estimate for the
relative amount of tankyrase 2 in the nucleus versus whole cells because the signal from the
nuclei was too weak to quantify. However, it appears that the fraction of tankyrase 2 in the nucleus is lower than the fraction of tankyrase 1.
4.3.7. Creation of a tankyrase 2 knock-out cell line
In order to determine the function of tankyrase 2 we employed a loss of function approach, using the genetically tractable DT40 system. To remove the tankyrase 2 gene from the
DT40 cell-line, a knock-out construct was made that contained a drug resistance marker cassette flanked by tankyrase 2 genomic DNA (Figure 12A). Genomic DNA for this construct was obtained by screening a DT40 genomic library with a tankyrase 2 N-terminal cDNA probe. The knock-out construct design was such that its targeted integration would result in the removal of 7 kb of the tankyrase 2 gene, including the 1st exon. The parental cells used for the knock-out
60
Figure 12 Knock-out of tankyrase 2. A. Knock-out strategy A knock-out construct containing a drug resistance marker cassette (M) flanked by tankyrase 2 genomic DNA (red) was used to remove 7 kb of the tankyrase 2 gene, including the 1st exon. B & C. Analysis of clones by Southern blot B. Two of the three tankyrase 2 alleles were knocked-out sequentially using different drug resistance markers and analyzed by Southern blot with the probe indicated in panel A (blue). Marker cassettes are BSR (Blasticidin S resistance) and HIS (Histidinol resistance). C. A single allele knock-out cell-line was probed with an upstream (green) and a downstream (blue) probe. Relative signal intensities of the bands are indicated in the bottom panel.
61 expressed the tetracycline transactivator so that regulated expression of rescuing or mutant
alleles could be achieved in later experiments. This cell line was electroporated with the knock- out construct, and clonal, drug resistant cell lines were isolated. Southern blot analysis revealed a high level of gene-targeting (70 – 90 % of the clones showed targeted integration).
Quantification of the wt and knock-out signals demonstrated a 2 to 1 ratio (wt / KO allele), indicating a total of three alleles of the tankyrase 2 gene were present in DT40 (Figure 12B).
After knocking-out the second allele a wt band remained, that now contained 1/3 of the signal, thus confirming the presence of a total of three alleles. Western blots using extract from double allele knock-out cells showed no reduction in tankyrase 2 levels. The presence of three alleles is not unusual in DT40 cells, since this generally diploid cell line has some chromosomes that are present in three copies. However, several attempts to remove the last allele failed and probing with an upstream probe revealed that the third allele likely has different sequence in the region corresponding to the 5’ arm of the knock-out construct. When a single allele knock out was probed with the probe used previously (Sph1 probe, shown in blue), which anneals downstream of the knock-out construct, the relative intensities of the bands were consistent with there being three alleles (Figure 12C). However, when the same knock-out clone was probed with an upstream probe (Xba probe, green) the intensities for the wt and knock-out alleles were now equal, indicating the existence of only two alleles. These results suggest that the sequence where the upstream probe anneals is not present in the third allele, and that an undefined amount of sequence downstream of this site could also be different. Thus, at least part of the 5’ end of our knock-out construct is probably not homologous to the third tankyrase 2 allele. This could well explain why we have been unable to remove this allele. The 3rd allele is likely the result of a
62 rearrangement, upstream of the start codon. Hence, to successfully create a complete tankyrase 2
knock-out, a new construct needs to be made containing a 5’arm that is homologous to the last
allele.
4.3.8. Co-regulation of tankyrase 1 and 2
In anticipation of a complete tankyrase 2 knock-out being lethal, a DT40 cell line stably
expressing a tankyrase 2 rescuing allele was created from a single allele tankyrase 2 knock-out.
While the rescuing allele has not yet been needed to obtain the complete tankyrase 2 knock-out,
it did reveal coregulation of tankyrase 1 and tankyrase 2. Surprisingly, when expression of the
rescuing allele was analyzed by Western blot, not only
tankyrase 2 levels were increased, but also tankyrase 1
(Figure 13). Thus, it appears that the level of tankyrase 2
somehow regulates the level of tankyrase 1.
Quantification of the rescuing allele expression using
iodinated antibodies showed a ~16 fold increase in
Figure 13 Expression of tankyrase 2 tankyrase 2 levels, which resulted in a four-fold increase rescuing allele leads to increase in tankyrase 1 levels. Western blot showing in tankyrase 1 levels. The observed effect of tankyrase 2 levels of tankyrase 1 and 2 with and without expression of the tankyrase 2 rescuing allele expression on tankyrase 1 levels could be mediated at (- and + doxcycyclin (dox)). Tubulin blot shows equal loading. several different levels; transcription, RNA stability,
translation and protein stability. To investigate whether tankyrase 2 rescuing allele expression
affects transcription of tankyrase 1, we used Northern blots to compare mRNA levels in the
presence or absence of rescuing allele expression (courtesy of Dr. R. Das). These showed a
63 twofold increase in tankyrase 1 mRNA levels after induction of the rescuing allele, suggesting that the increase in tankyrase 1 at least partially results from a feedback mechanism that acts on tankyrase 1 transcription and/or mRNA stability. At this point it is not clear whether this increase in mRNA level is the sole reason for the increased amount of tankyrase 1 protein or whether other, translational / post-translational mechanisms are also involved.
4.4. Discussion
When tankyrase 1 was first identified, its presence at telomeres together with its enzymatically active PARP domain strongly implicated this protein as an effector molecule in some aspect of telomere regulation. However, the subsequent isolation of tankyrase 2, the identification of the wide variety of tankyrase 1 and 2 interaction partners, and the detection of tankyrase 1 and 2 at multiple, unrelated subcellular localizations suggested a more complex but poorly understood multifunctional behavior. By characterizing tankyrase 1 and 2 from non- mammalian cells we have been able to identify conserved aspects of tankyrase domain structure including the five subdomains within the ANK repeats. We have further shown that chicken cells contain approximately twice as much tankyrase 1 as tankyrase 2 and, although the majority of each protein resides in the cytoplasm, a significant fraction appears to be present in the nucleus.
Since both human and chicken tankyrase 2 bind to TRF1, this nuclear localization indicates that tankyrase 2, like tankyrase 1, is likely to have a telomeric function. This is supported by reports that show that overexpression of human tankyrase 2 leads to removal of TRF1 from the telomere
(Cook et al., 2002).
In order to justify the study of chicken tankyrases as a model for understanding human
64 tankyrase function, we compared the proteins of these species both at the sequence and
functional level. We observed a high level of sequence conservation, which is striking, as
telomere-associated proteins are not normally highly conserved (Konrad et al., 1999; Li et al.,
2000). This conservation fits well with tankyrase 1 and 2 being multirole effector proteins that
function not only in telomere biology but also in other cellular processes. Further support for an
evolutionary conservation of function comes from the similar localization pattern of human and
chicken tankyrases (see chapter 5) and from the observation that chicken tankyrases, like human
tankyrases, can interact with the telomeric protein TRF1. Thus, the results obtained from
tankyrase studies in chicken are likely to apply to other species.
While a wide variety of proteins are known to interact with the ankyrin domain, it is
unclear whether these proteins merely compete with each other for tankyrase binding, or if their
interaction is somehow regulated. Our finding that the LLEAAR/K motif that interrupts the ANK
repeats is both highly conserved in tankyrase proteins from different species, and a potential
ATP binding site, suggests that this motif may do more than simply serve to separate the
tankyrase ankyrin repeats into sub-domains. Interestingly, ATP binding to certain chaperones has
been shown to induce a conformational change that alters their protein binding properties
(Buchberger et al., 1995). Thus, one possibility is that the tankyrase motif acts as a regulatory element for ankyrin mediated protein binding by responding to cellular ATP.
Our findings that exogenous expression of tankyrase 2 results in increased levels of tankyrase 1 suggests some type of crosstalk between these paralogs. This is not unusual, and is often seen for proteins that heterodimerize. For example, Ku70 and Ku80 form a heterodimer which is more stable than the individual subunits (Cary et al., 1997; Errami et al., 1998). As
65 shown in chapter 5, tankyrase 1 and 2 also interact with each other, thus it is possible that higher levels of tankyrase 2 result in stabilization of an increased amount of tankyrase 1. However, the increase in tankyrase 1 mRNA levels indicates that at least some of the regulation occurs at the
RNA level. The mechanism behind the increase in tankyrase 1 mRNA nevertheless remains unclear, and could reflect an increase in transcription or RNA stability.
66 5. TANKYRASE POLYMERIZATION IS CONTROLLED BY ITS SAM AND PARP
DOMAINS
5.1. Summary
Tankyrases are novel poly(ADP-ribose) polymerases that have SAM and ankyrin protein
interaction domains. They are found at telomeres, centrosomes, nuclear pores and Golgi and
have been shown to participate in telomere length regulation and sister chromatid cohesion.
Their non-telomeric function(s) are unknown and it has been difficult to envision a common role
at such diverse cellular locations. We have shown that tankyrase 1 polymerizes through its SAM
domain to assemble large protein complexes. In vitro polymerization is reversible and still allows interaction with ankyrin-domain binding proteins. SAM-dependent polymerization also occurs in vivo, with exogenous tankyrase expression leading to disruption of Golgi structure and apical secretion. We have also shown that tankyrase polymers are dissociated efficiently by auto- poly(ADP-ribosyl)ation. This disassembly is prevented by mutation of the PARP domain. Our findings indicate that tankyrase 1 has the unique capacity to promote both assembly and disassembly of large protein complexes. Thus, tankyrases appear to be master scaffolding proteins that regulate the formation of dynamic protein networks at different cellular locations.
This implies a common scaffolding function for tankyrases at each location with specific tankyrase interaction partners conferring location-specific roles to each network; e.g. telomere compaction or regulation of vesicle trafficking.
67 5.2. Introduction
Tankyrase 1 and 2 are homologous multidomain proteins that contain ∼20 ankyrin repeats, a
SAM (sterile alpha motif) domain and a PARP (poly(ADP-ribose) polymerase) domain (Smith,
2001). Tankyrases interact with a diverse set of proteins through their ankyrin repeats. These tankyrase binding proteins participate in a wide array of cellular processes such as telomere and spindle organization, Golgi dynamics and apoptosis. Their various functions are reflected in the many cellular locations where tankyrases are found; these include telomeres, centrosomes, nuclear pores and Golgi-associated vesicles (Chi and Lodish, 2000; Smith and de Lange, 1999).
Tankyrase function is best understood at telomeres, where it appears to regulate telomere length, by altering the organization of the telomeric DNA-protein complex, and separation of the sister chromatids during mitosis (Dynek and Smith, 2004). The function of tankyrase at centromeres
and nuclear pores is unknown, but its association with Golgi vesicles is thought to be important for vesicle trafficking (Chi and Lodish, 2000; Lyons et al., 2001; Sbodio et al., 2002).
Tankyrases thus appear to be involved in a strikingly diverse range of processes, which prompts the question of what their general role is at each location. In addition, it is unclear whether tankyrase 1 and tankyrase 2 always function in the same pathway, or whether they may have diverged functions.
While the tankyrase ankyrin and PARP domains are relatively well characterized, the role
of the tankyrase SAM domain has been unclear. SAM domains are versatile interaction modules that commonly mediate homo- and heterotypic protein-protein interactions (Kim and Bowie,
2003). They exist as compact 5 helical bundles and have been found in >1000 different proteins.
While most SAM domains are thought to mediate dimerization, the SAM domains of the ephrin
68 B2 receptor (EphB2) and the transcriptional repressors TEL (translocation Ets leukemia) and PH
(polyhomeotic) have two separate interaction surfaces and can promote formation of oligomers
or polymers (Kim et al., 2002; Kim et al., 2001; Thanos et al., 1999). EphB2 oligomerization is quite weak (Smalla et al., 1999), but self-association is thought to occur when ligand binding
causes receptor clustering. This may create a new interaction surface for recruiting proteins that act downstream in the signaling pathway. In contrast, polymerization of both TEL and PH is
robust and results in the formation of long filaments. These are thought to mediate transcriptional
repression by promoting spreading of silent chromatin (Kim et al., 2002; Kim et al., 2001).
Here we show that the tankyrase SAM domain is similar to the TEL and PH SAM
domains in that it can mediate polymerization, both in vitro and in vivo. We also demonstrate
that tankyrase 1 and tankyrase 2 can interact directly with each other, to form co-polymers. We
further show that tankyrase polymers can be disrupted by auto-poly(ADP-ribosyl)ation. This
unique ability to mediate both complex assembly and disassembly suggests that tankyrases are
master scaffolding proteins, which regulate the formation of large protein networks. We thus propose a general function for tankyrases that may apply at all the different locations where these
proteins reside.
69 5.3. Results
5.3.1. The tankyrase SAM-domain mediates polymerization in vitro
Although co-immunoprecipitation experiments have shown that human tankyrase proteins can interact in vivo to form homo- and hetero-typic complexes (Sbodio et al., 2002), it
has not been determined whether
complex formation is mediated by
direct association between tankyrase
molecules or via other interacting
partners. Since SAM domains form
both dimers and longer oligomers, we
sought to determine whether tankyrase
molecules can associate via a direct
interaction between SAM domains. In
initial experiments, we used a yeast
Figure 14 Constructs used for experiments in Chapter 5. two-hybrid approach to assay for Numbers refer to aminoacids. Mutations are indicated with a yellow line, orange bars indicate LLEAAR motifs. Tags and tankyrase 2 self-association. A section fusion partners are described in the text and were placed at the N-terminus. of tankyrase 2 encompassing the SAM and PARP domains (SARP2) (Figure 14) was cloned into the DNA-binding domain (LexA-
SARP2) and activation domain (AD-SARP2) vectors of the Duplex-A yeast two-hybrid system
(Origene). However, expression of this fragment killed the yeast cells. As yeast does not have any PARP homologues, we attributed this lethality to poly-(ADP-ribosyl)ation of yeast proteins by the SARP2 polypeptide. We therefore mutated Met1055 Val, a mutation known to inactivate
70 PARP activity of human tankyrase 2 (Figure 14) (Sbodio et al., 2002). The mutated construct
(mSARP2) was not lethal and expressed well, so it was used in all further experiments.
When β-galactosidase assays were performed with LexA-mSARP2 and AD-mSARP2, a
small but reproducible activation of the β-galactosidase gene was observed (Figure 15). The level
of activation was approximately 140-
fold less than that observed for control
proteins supplied with the Duplex-A
system that exhibit a strong interaction Figure 15 Yeast two-hybrid assay for interactions between the SARP domains of tankyrase 2 The SARP domain of (results not shown). However, tankyrase 2 (Met1055 Val mutation) was expressed as a LexA DNA-binding domain or B42-activation-domain fusion protein statistical analysis by analysis of and assayed for activation of β-galactosidase activity in the presence of the corresponding activation domain/DNA binding domain fusion protein, or the activation domain/DNA-binding variance using the SAS procedure domain alone. The pairs of DNA-binding domain (DBD) and activation domain (AD) constructs are illustrated on the left, PROC MIXED indicated that signal while the resulting β-galactosidase (β-gal) activity (in Miller units) is indicated at the right. obtained with the LexA-mSARP2 and
AD-mSARP2 constructs was significantly higher than the signal obtained in control experiments
using the parental DNA-binding domain or activation domain vectors (P < 0.001). These results suggested that the SAM or PARP domain of tankyrase 2 mediates a direct self-association. The low level of β-galactosidase expression could reflect a weak interaction between the SAM domains. Alternatively, SAM-mediated self-association of the activation domain or DNA- binding domain fusion proteins could reduce productive activation domain/DNA-binding domain interactions and prevent efficient activation of the β-galactosidase gene.
71 To investigate tankyrase self- association further, we determined whether the bi-functional cross-linker bis(sulfosuccinimidyl)suberate (BS3) could cause intermolecular cross-linking of mSARP2 molecules. mSARP2 was expressed as a maltose binding protein
(MBP) fusion protein in E. coli, affinity purified and incubated with several concentrations of BS3. When the cross- linked MBP–mSARP2 (83 kDa) was analyzed by SDS-PAGE, most of the Figure 16 Crosslinking experiments indicate SAM domain polymerization. Purified tankyrase 1 and 2 SARP or SAM protein electrophoresed as a slow domain MBP fusion proteins were cross-linked with BS3 and separated on SDS/PAGE gels. A, C and D show Western blots migrating form, that barely entered the probed with antibody to tankyrase or MBP and B shows Coomassie staining. Positions of molecular mass markers are separating gel, while the rest remained in shown to the side of each gel. A. tankyrase 2 mSARP2 (Met1055 Val) domain after incubation with 0, 0.5 or 1 mM BS3. Right panel, anti-MBP; left panel, anti-tankyrase (anti-Tank) the well of the stacking gel. These high antibody. Samples in lanes 4–6 and 10–12 were cleaved with factor Xa. B. Products obtained after incubating purified MBP molecular weight forms could be detected (lanes 1–2), mSARP2 domain (mSARP; lanes 3–4) or tankyrase SAM1 domain (lanes 5–6) with (+) or without (-) 2 mM BS3. C. Products obtained with tankyrase 2 SAM domain by both Western blotting and Coomassie (SAM2) after incubation with 0, 0.1, 0.25, 0.5 and 1 mM BS3. The lower panel shows a shorter exposure of the 72 kDa staining (Figure 16A, lanes 2–3 and 5–6, monomer band. D. Products obtained with tankyrase 1 SAM domain (SAM1) after incubation with 0, 0.01, 0.05, 0.1 and 3 and Figure 16B, lanes 3–4). To ensure that 0.5 mM BS . The lower panel shows a shorter exposure of the 72 kDa monomer band. intermolecular cross-linking was not caused by the MBP domain of the fusion protein, we digested the MBP–mSARP2 with factor
72 Xa to separate the MBP and mSARP2 domains (Figure 16A, lanes 4–6 and 10–12). After cross-
linking, all the mSARP2 ran as high molecular weight complexes, whereas most of the MBP remained monomeric. At the highest concentration of cross-linker, some of the MBP was observed in the well of the stacking gel. However, a similar phenomenon was observed with
BSA when it was present during cross-linking of MBP–mSARP2 (results not shown). Therefore, we attribute this impaired mobility to the MBP becoming trapped in the well, rather than to intermolecular cross-linking. To examine whether MBP played a role in the formation of the
MBP–mSARP2 complexes, we determined whether purified MBP was cross-linked with BS3. As shown in Figure 16B (lanes 1 and 2), all of the MBP remained monomeric, indicating that the protein did not self-associate, whereas the MBP–mSARP2 clearly oligomerized to form large complexes. These results indicate that MBP-mSARP2 not only dimerizes but also has the ability to polymerize.
As PARP domains do not usually mediate dimerization or polymerization, we next asked whether the SAM domain alone is sufficient for self-association. The SAM domains of both tankyrase 1 and tankyrase 2 were expressed as MBP fusions (SAM1 and SAM2, Figure 14), purified and incubated with various concentrations of BS3. As shown in Figure 16B, C and D,
BS3 caused intermolecular cross-linking of the SAM1 and SAM2 fusion proteins, indicating that
tankyrase 1 and 2 self-association is mediated by the SAM domain. The amount of cross-linking
depended on the BS3 concentration, with low concentrations giving rise to dimers, whereas
higher concentrations resulted in dimers, trimers and larger oligomers. Overall, BS3 cross-linking
of the SAM-domain fusion proteins resulted in less of the very large polymers than cross-linking of the equivalent SAM–PARP-domain fusion proteins. This could be because the SAM-domain
73 fusion proteins folded less well than the corresponding SARP domains. Alternatively, the PARP domain might promote SAM domain interactions.
Although these studies
demonstrate that the SAM domains of
tankyrase 1 and 2 can be cross-linked into
large homotypic complexes, chemical
cross-linkers can trap transient
interactions and then drive
polymerization. Thus, this approach did
not yield information about the stability
of the SAM-domain association or the
natural size of the SAM-domain
complexes. We therefore sought to
determine the extent to which the
tankyrase SAM domain can self-associate
in the absence of cross-linker. First, we
Figure 17 Self-association of the tankyrase 1 SAM domain. analysed the size of purified SAM A. Superose 6 fractionation of MBP-SAM1. Peak fractions containing molecular weight markers are indicated; 2,000 kDa, domain by gel-filtration. When purified dextran blue; 669 kDa, thyroglobulin; 232 kDa, catalase. Fraction numbers are shown below. B & C. Transmission electron microscopy of uranyl-acetate stained MBP-SAM1 MBP-SAM1 fusion protein (65 kDa) was polymers. Arrows indicate SAM polymers. Samples in C. were treated with proteinase K prior to deposition on the EM applied to a Superose 6 size-exclusion grid. Scale bar is 100 nm. D. Histogram showing the length distribution of MBP-SAM1 rods. Mean (37 nm) and median column, it showed a broad elution profile (36 nm) lengths were calculated for 40 rods, but the shortest rods were probably underrepresented because they were difficult to distinguish from background. (Figure 17A). The peak fraction
74 corresponded to a mass of ∼930 kDa or complexes of ∼14 MBP-SAM1 molecules, but a
significant amount of protein fractionated close to the void volume of the column (>2000 kDa)
and hence must have contained complexes of >30 MBP-SAM1 molecules. Since the previous
experiments demonstrated that MBP does not mediate self-association, this result indicates that
the tankyrase 1 SAM domain can polymerize in the absence of cross-linker. Moreover, as the
polymers remain assembled during gel-filtration, the interactions mediating the SAM-SAM
association must be quite stable.
To examine the tankyrase SAM-domain complexes further, we performed transmission
electron microscopy on purified MBP-SAM1 preparations that had been negatively stained with
uranyl-acetate. As shown in Figure 17 (B and D), we observed small rod-shaped structures that
varied in length from ∼18 - 80 nm, with a mean of ∼37 nm. They were only observed in the
presence of purified MBP-SAM1 and were destroyed by treatment with proteinase K (Figure
17C), confirming that they were composed of the MBP-SAM1 protein. Thus the tankyrase 1
SAM domain resembles the SAM domains of the TEL and PH repressors which also polymerize
into ordered rods or fibre-like structures that can be observed by EM (Kim et al., 2002; Kim et
al., 2001; Tran et al., 2002).
5.3.2. The SAM domains of tankyrase 1 and 2 copolymerize in vitro
Our results show that the SAM domains of tankyrase 1 and tankyrase 2 both have the
ability to homo-polymerize. To address whether the two SAM domains can also interact directly with each other to form copolymers we performed a co-purification experiment. The SAM domain of tankyrase 2 fused to MBP (MBP-SAM2, 67 kDa) and the SAM domain of tankyrase 1
75 (SAM1, 22 kDa) were co-expressed in E.coli using a single plasmid that encodes both proteins.
To assess whether SAM1 can interact with SAM2, we affinity-purified MBP-SAM2 and assayed for co-purification of SAM1. As shown in Figure 18A, SAM1 efficiently co- eluted with MBP-SAM2 (lane 2), while purification from bacteria co-expressing MBP and
SAM1 did not yield any SAM1 (lane
4). Thus, this indicates that the SAM
domains from tankyrase 1 and 2 can
interact with each other.
Interestingly, the ratio of purified
protein to input was about the same
for MBP-SAM2 and SAM1,
suggesting no preference for homo
versus heterotypic associations. To
Figure 18 Co-polymerization of the tankyrase 1 and tankyrase determine whether the purified 2 SAM domains. A. Co-purification of tankyrase 1 SAM domain with MBP-SAM2 fusion protein. Combinations of MBP-SAM2, SAM1 – MBP-SAM2 complexes MBP and SAM1 were co-expressed in E.coli as indicated above each blot. MBP fusion proteins were purified by amylose affinity chromatography and co-purification of SAM1 was assayed by were polymeric, we analyzed their Western blot. Left panel: Western blot of input (I) and proteins bound to amylose column (B) with anti-tankyrase antibody (anti- size by gel-filtration. The broad SAM). Right panel: same blot as left panel, stripped and reprobed with anti-MBP antibody. B. Superose 6 fractionation of purified MBP-SAM2 – SAM1 complexes. Western blot with anti- elution profile of each protein (Figure tankyrase antibody. Peak fractions containing molecular weight markers are indicated; 2,000 kDa, dextran blue; 669 kDa, 18B), is consistent with the existence thyroglobulin; 232 kDa, catalase. Fraction numbers are shown below. of SAM1 – MBP-SAM2 hetero- polymers, that range in size from >2 MDa to near monomer size. The bulk of the MBP-SAM2 eluted as smaller trimeric or tetrameric complexes (peak at 242 kDa), that contained relatively
76 little SAM1. This could indicate a preference of MBP-SAM2 to form small complexes, but is
likely at least partially the result of MBP-SAM2 depolymerization during the affinity purification and gel-filtration. Another factor is that the input levels of SAM1 are much lower than the MBP-
SAM2 levels, presumably because the SAM domain is less stable without fusion to MBP. This
would result in the smaller complexes containing mostly MBP-SAM2. Unexpectedly, SAM1 had
a slightly different elution profile compared to MBP-SAM2; it had a peak in fraction 36 where
there is no corresponding peak for MBP-SAM2, instead this protein has a peak in fraction 32. A
likely explanation is that depolymerization occured during the gel-filtration run, and that the dissociated SAM1 was retarded more relative to MBP-SAM2 because of its lower molecular weight. Nevertheless, the broad elution profile of both proteins shows that they can indeed form co-polymers, with sizes exceeding 2 MDa or at least 30 monomers.
5.3.3. SAM domain mutagenesis
To determine which residues of the tankyrase 1 SAM domain mediate polymerization, we performed gel-filtration experiments with a variety of SAM domain mutants. A 3D threading program (http://www.sbg.bio.ic.ac.uk/~3dpssm/) identified the Ephrin B2 receptor SAM domain
as the closest relative, with a known crystal structure, of the tankyrase SAM domain. The
software also produced a predicted structure for the tankyrase 1 SAM domain based on the
structure of the Ephrin B2 SAM domain. The Ephrin B2 crystal structure includes two self-
association surfaces, a hydrophobic arm-exchange surface employing mostly residues in the N-
terminus of the domain, and a more hydrophilic surface using helices 3 and 5 (Thanos et al.,
1999). Residues in the tankyrase 1 SAM domain predicted to correspond to residues on the two
77 EPB3 DYTTFTT-VGDWLDAIKMGRYKESFVSAGFASFDLVAQMTAEDLLRIGVTLAGHQKKILSSIQDMRLQMNQTLP--VQV EPA4 -FSAVVS-VGDWLQAIKMDRYKDNFTAAGYTTLEAVVHVNQEDLARIGITAITHQNKILSSVQAMRTQMQQMHGRMVPV EPB2 DYTSFNT-VDEWLEAIKMGQYKESFANAGFS-FDVVSQMMMEDILRLGVTLAGHQKKILNSIQVMRAQMNQIQS--V-- Tankyrase1 EVSGLDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGHRHKLIKGVERLLGGQQGTNP--Y—-
Mutant 2 EVSGLDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADKGHEELKEIGINAYGHRHKLIKGVERLLGGQQGTNP--Y-- Mutant 3 EVSGLDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGHRHKLIKGVAALLGGQQGTNP--Y-- Mutant 4 EVSGLDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGHRHKLIKGVERKLGGQQGTNP--Y-- Mutant 5 EVSGLDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGHRHKLIKGVERLLGGQAGTNP--Y-- Mutant 6 ERSGLDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGHRHKLIKGVERLLGGQQGTNP--Y-- Mutant 7 EVSGRDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGHRHKLIKGVERLLGGQQGTNP--Y-- Mutant 8 ERSGRDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGHRHKLIKGVERLLGGQQGTNP--Y-- Mutant 9 EVSGLDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGAAAALIAGVAALLGGQQGTNP--Y-- Mutant 10 EVSGLDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGHRHALIAGVAALLGGQQGTNP--Y-- Mutant 11 EVSGLDMNITQFLKSLGLEHLRDIFETEQIT-LDVLADMGHEELKEIGINAYGHRHKLIKGVERLLGGAAGTNP--Y—
Figure 19 Mutagenesis of the tankyrase 1 SAM domain. Top panel: Alignment of the SAM domains from the Ephrin receptors B3, A4 and B2 and the chicken tankyrase 1 SAM domain. Identical residues are in red, similar residues in blue. Bottom panel: Tankyrase 1 SAM domain mutants, mutants 2,4,6,7 and 8 have mutations in the predicted arm- exchange surface, mutants 3,5,9,10 and 11 have mutations in the predicted helix 3/5 surface. Ephrin B2 surfaces were mutated (Figure 19), and polymerization was assayed by gel-filtration
of purified SAM domains. Strikingly none of the mutations affected the SAM domain elution
profile, which was essentially the same as for the wt SAM domain (data not shown), indicating
that the residues that mediate tankyrase SAM polymerization are different from those that
mediate Ephrin B2 self-association. Considering the recently appreciated flexibility of SAM
domain interactions this is not surprising (Kim and Bowie, 2003). Thus, to identify residues
critical for polymerization, more extensive and random mutagenesis experiments will have to be
carried out.
5.3.4. Self-association of full-length tankyrase
We next asked whether full-length tankyrase also undergoes SAM-dependent
polymerization. His-tagged full-length tankyrase 1 (FL-tankyrase1) and His-tagged tankyrase 1
lacking the SAM domain (tankyrase1-∆SAM) (Figure 14) were produced in Sf21 insect cells
using a baculovirus expression vector and purified by Ni2+-affinity chromatography. After
elution with imidazole and dialysis to reduce the imidazole concentration, both protein
78 preparations were subjected to centrifugation
(5 min at 18000 x g) to remove any insoluble
material. Interestingly, nearly all the FL-
tankyrase1 was pelleted, whereas most of the
tankyrase1-∆SAM remained in the
supernatant (Figure 20A). Since the FL-
tankyrase1 insolubility could reflect protein
misfolding and subsequent aggregation
rather than SAM-dependent polymerization,
we next tested whether the solubility was
concentration-dependent, as might be
expected for a specific protein-protein
interaction. Figure 20 Solubility and catalytic activity of purified full- length tankyrase 1. A. Solubility of recombinant tankyrase. Purified full-length tankyrase 1 (TNK) or tankyrase1-∆SAM To test for solubility, we made serial (∆SAM) was centrifuged and protein in the supernatant (S) or pellet (P) pellet was detected by SDS PAGE and Coomassie dilutions of FL-tankyrase1 and centrifuged staining. B. Increase in tankyrase 1 solubility with increasing dilution. Purified tankyrase was diluted to 16.7, 5.6, 1.9, 0.62, identical volumes of each dilution. The 0.2 µg/ml prior to centrifugation, protein in the supernatant (S) or pellet (P) was detected by Western blotting with tankyrase antibody. C. Effect of dialysis, salt, detergent and imidazole protein was then detected in the supernatant (1M) on FL-tankyrase1 solubility. D. TRF1 binding to tankyrase polymers. His-tagged TRF1 was incubated with or or pellet by Western blotting with tankyrase without purified His-tagged tankyrase1 (TNK), or tankyrase1- ∆SAM (∆SAM) and then centrifuged. Soluble and pelleted fractions were analyzed by Western blotting with His antibody. As shown in Figure 20B, antibody. The arrow points to tankyrase 1 or tankyrase1- ∆SAM, the arrowhead indicates TRF1. E. PARP activity of decreasing the tankyrase concentration led to recombinant tankyrase 1. Tankyrase (upper bracket) or TRF1 (lower bracket) poly(ADP-ribosyl)ation was detected by a dramatic change in the fraction of protein Western blotting with antibody to poly(ADP-ribose). The arrow marks the input TRF1 which started with a low level of modification. The proteins present in each reaction are that was soluble; essentially all the protein in indicated at the top, the presence or absence of 1 mM NAD and 10 mM 3-aminobenzamide (3AB) is indicated by (+) or (- ).
79 a ~120 nM solution appeared in the pellet (lanes 1 and 2), whereas most of the protein in a ~15 nM solution remained in the supernatant (lanes 5 and 6). Thus, the insolubility of full-length tankyrase appeared to be caused by specific SAM-domain mediated polymerization into a high molecular weight form. This behavior is similar to what was previously observed for the purified
TEL repressor where SAM-mediated polymerization leads to the formation of large, insoluble complexes (Kim et al., 2001; Tran et al., 2002). We were also able to cycle full-length tankyrase between soluble and insoluble forms by the addition and removal of imidazole but not by adding salt and detergent (500 mM NaCl, 0.05 % Triton X-100) (Figure 20C lanes 5-8). Purified FL- tankyrase 1 was soluble when first eluted from the Ni+ column with imidazole (lanes 1-2), it became insoluble when the imidazole was removed by dialysis (lanes 3-4), but returned to the soluble form upon re-addition of imidazole (lanes 7-8). The reversible nature of the tankyrase insolubility again indicates that solubility is determined by a specific polymerization/depolymerization reaction.
To further assess the properties of the polymerized tankyrase, we determined whether it retained the capacity to interact with TRF1, a known ankyrin domain binding protein, and whether it was catalytically active. To test for interaction with TRF1, we incubated FL- tankyrase1 or tankyrase1-∆SAM with purified TRF1 and then centrifuged the sample to monitor
TRF1 solubility. As shown in Figure 20D, all the TRF1 was soluble when incubated without tankyrase or with tankyrase1-∆SAM, however, it became partially insoluble in the presence of
FL-tankyrase1, indicating that polymeric tankyrase has the ability to interact with TRF1.
To examine the catalytic activity of the tankyrase polymers we assayed for auto- ribosylation and poly(ADP-ribosyl)ation of TRF1. Purified tankyrase was incubated with its
80 substrate, NAD+, in the presence or absence of purified TRF1 and the PARP inhibitor 3-amino-
benzamide (3AB), PARP activity was then assessed by Western blot analysis using an antibody
to poly(ADP-ribose) (Figure 20E). FL-tankyrase1 showed robust auto-ribosylation (Figure 20E,
lane 2) and poly(ADP-ribosyl)ation of TRF1 (lane 6). The activity was dependent on the
presence of NAD+ and inhibited by 3AB. Interestingly, tankyrase1-∆SAM was much less active
than FL-tankyrase 1 (lanes 9 & 11). This decrease in PARP activity suggests that the SAM
domain is required for optimal catalytic activity, perhaps because dimerization/oligomerization
promotes catalysis as has been observed for the related enzyme PARP-1 (Mendoza-Alvarez and
Alvarez-Gonzalez, 1999; Mendoza-Alvarez and Alvarez-Gonzalez, 2004). Overall, our results
show that full-length, catalytically competent tankyrase 1 molecules can undergo a reversible
SAM-mediated association that results in assembly of large high molecular weight complexes.
As the isolated SAM domain of tankyrase 2 can also self-associate, we would expect tankyrase 2
to also oligomerize in a SAM-dependent manner.
5.3.5. In vivo formation of tankyrase complexes
To evaluate whether tankyrases can also polymerize in vivo, we assayed for tankyrase complexes in cells that overexpressed tankyrase 1. Initially we carried out a simple pelleting experiment using extracts from chicken LMH cells that had been transiently transfected with
Myc-tagged FL-tankyrase1 or tankyrase1-∆SAM (Figure 14). The cells were lysed with NP40 24 hrs after transfection and the cell lysates were centrifuged to assess the solubility of the tagged protein. As shown in Figure 21A, essentially all the FL-tankyrase 1 was pelleted when extracts were prepared at 4°C and only a small fraction was soluble at 25°C. In contrast, tankyrase1-
81 ∆SAM was fully soluble at both temperatures. The difference in solubility between the full- length protein and tankyrase1-∆SAM in cell extracts is strikingly similar to what was observed with the purified protein, suggesting that tankyrase can also undergo SAM-mediated polymerization in vivo. While the SAM- dependent insolubility could be caused by the SAM domain mediating interactions with other insoluble proteins (e.g. components of the cytoskeleton), the high level of tankyrase expression would have been expected to saturate such interactions Figure 21 In vivo complex formation. A. Solubility of leaving some protein free in solution. overexpressed tankyrase 1 in LMH cell lysates. Cells expressing FL-tankyrase 1 (TNK) or tankyrase1-∆SAM The small amount of soluble (∆SAM) were lysed at 4°C or 25°C, centrifuged and tankyrase was detected in the supernatant (S) or pellet (P) by Western blotting with tankyrase antibody. B. Superose 6 tankyrase 1 obtained after room fractionation of soluble tankyrase extracted from cells expressing FL-tankyrase 1 (TNK), tankyrase1-∆SAM temperature extraction suggested that (∆SAM) or untransfected cells (endo). Total amount of protein loaded on the column was ~250 µg from TNK and tankyrase polymers may disassemble at ∆SAM expressing cells and ~650 µg from wild type cells. The input material is shown in the lane to the left; amounts loaded were 1% of total for TNK, 2% for SAM and 10% higher temperatures. To assess the size of wild type extracts. The peak fractions containing marker proteins are shown at the top and fraction numbers are the soluble FL-tankyrase1 complexes and shown below. Detection was by Western blotting with tankyrase antibody. tankyrase1-∆SAM, we fractionated extracts from tankyrase-transfected LMH cells on a Superose 6 column (Figure 21B). FL- tankyrase1 had a broad elution profile with the largest complexes fractionating close to the void volume (>2,000 kDa) and the smallest with a mass of 250-350 kDa. In contrast, tankyrase1-
∆SAM eluted in a sharp peak with an apparent mass of 350 kDa. Thus, even under conditions
82 that appear to cause partial dissociation of tankyrase polymers, much of the full-length tankyrase still exists in large complexes that are dependent on the SAM domain. It is unclear whether the tankyrase1-∆SAM exists as a monomer or as a small complex, as the deviation from the expected mass of 130 kDa for the monomer could result from the elongated shape of the ankyrin repeat domain (Sedgwick and Smerdon, 1999).
To compare the size of the complexes containing over-expressed tankyrase to those formed by endogenous tankyrase, we also performed gel-filtration with extracts from untransfected LMH cells. The solubility of endogenous tankyrase depended on the volume of buffer used to extract the cells (data not shown), thus for these experiments we used sufficient buffer to render most of the tankyrase soluble. When fractionated on a Superose 6 column, endogenous tankyrase 1 displayed a similar elution profile to the overexpressed FL-tankyrase1
(Figure 21B). Thus, like overexpressed tankyrase, endogenous protein also exists in very large complexes. However, we would expect these complexes to contain a higher ratio of tankyrase interaction partners than complexes from cells that overexpress tankyrase because the level of overexpressed tankyrase is likely to greatly exceed that of the interaction partners. In both cases, the complexes that can be extracted from cells are likely to be considerably smaller than the full- sized complexes that exist in vivo.
5.3.6. Tankyrase self-association in vivo is mediated by the SAM domain
The previous results indicate that tankyrase 1 forms large SAM-domain-dependent complexes in cells, and suggest that these complexes are the result of SAM-mediated polymerization. To show that direct SAM-SAM associations occur in vivo, we next performed
83 co-immunoprecipitation experiments using cells overexpressing a variety of tankyrase constructs
(Figure 14). To assay for tankyrase 1 SAM-mediated self-association, we transfected LMH cells with Flag-tagged SAM domain of tankyrase 1 (SAM1) in combination with Myc-tagged FL- tankyrase1 or tankyrase1-∆SAM and prepared detergent extracts at room temperature. These extracts contain a small fraction of partially depolymerized, soluble FL-tankyrase1. Since tankyrase1-∆SAM does not polymerize, it was completely soluble. These differences in solubility result in much higher levels of tankyrase1-∆SAM in the soluble fraction compared to
FL-tankyrase1 (Figure 22A top panel). Flag-SAM1 expressed well, and ran consistently as a doublet, which may indicate a post-translational modification. Immunoprecipitations were performed with anti-Myc antibody to precipitate FL-tankyrase1 and tankyrase1-∆SAM, anti-Flag antibody to precipitate SAM1 and anti-tubulin antibody as a control. As expected, both FL- tankyrase1 and tankyrase1-∆SAM were efficiently precipitated with anti-Myc antibody (Figure
22A, top panel, lanes 4 & 5). A small fraction of Flag-SAM1 co-precipitated with tankyrase1-
∆SAM, while a much more substantial amount co-precipitated with FL-tankyrase1 (note the relatively low levels of FL-tankyrase1 in the extracts)(bottom panel). Essentially the same results were obtained when anti-Flag antibody was used for immunprecipitation (Figure 22A, lanes 6 &
7). These results indicate that the tankyrase 1 SAM domain can indeed self-associate in vivo, and, in addition, suggest that residues present in Flag-SAM1 but not in the core SAM region that was deleted from tankyrase1-∆SAM, can also mediate a weak self-association. Other observations that substantiate this results are the higher than expected molecular weight of tankyrase1-∆SAM when analyzed by gel-filtration (Figure 21) or crosslinking (data not shown).
84 To show that SAM-mediated
interactions can occur between
tankyrase 1 and tankyrase 2 we
performed essentially the same co-
immunoprecipitation experiment with
the SAM domain of tankyrase 2 (Flag-
SAM2). As for SAM1, FL-tankyrase1
and tankyrase1-∆SAM could be
Figure 22 Detection of SAM-mediated self-assocation by precipitated using anti-Flag antibodies co-immunoprecipitation. A. Myc-tagged FL-tankyrase1 (FL) or tankyrase1-∆SAM (∆SAM) were cotransfected with Flag- (Figure 22B, lanes 6 & 7) indicating tagged tankyrase 1 SAM domain (SAM1) in LMH cells. Lysates were immunoprecipitated with anti-Myc, anti-Flag or anti-tubulin antibodies, and precipitated fractions were that SAM1-SAM2 interactions can analyzed by sequential Western blots with anti-Myc and anti- Flag antibodies. Input represents 10% of extract used for occur in cells. The comparatively small immunoprecipitation reactions. B. Same experiment as A. but LMH cells were transfected with Flag-tagged tankyrase 2 amount of precipitated tankyrase1- SAM domain (SAM2) instead of SAM1. ∆SAM again suggests that an interaction surface outside of the core SAM region exists. These interactions could however not be detected by anti-Myc immunoprecipitation. Several factors may explain this: first, the Myc antibody was less efficient than the Flag antibody at precipitating tagged protein, second, Flag-
SAM2 expression was significantly lower than Flag-SAM1. Thus, the Myc immunoprecipitation was probably too inefficient to allow detection of the co-precipitated SAM2 domain.
85
Figure 23 In vivo tankyrase localization. A. LMH cells expressing Myc-tagged FL-tankyrase1 (Myc-tankyrase 1), Myc-tagged tankyrase1-∆SAM (Myc-∆SAM), or Flag-tagged tankyrase 1 SAM domain (Flag-SAM1) were fixed and incubated with antibody to the Myc or Flag tags. Nuclei were counterstained with DAPI. B. Immunofluorescence of LMH cells expressing Flag-tagged tankyrase 2 (Flag-tankyrase 2) or Flag-tagged tankyrase 2 SAM domain (Flag-SAM2), detection with anti-Flag antibody. C. Cells co-expressing Myc-tagged FL-tankyrase 1 and Flag-tagged tankyrase 2. Panel 1, staining with Myc antibody; panel 2, staining with Flag antibody; panel 3, combination of the Myc and Flag signals. D. Untransfected cells were stained with antibody to the tankyrase SAM domain (panel 1), tankyrase antibody that had been preincubated with purified SAM domain (panel 2), or secondary antibody (panel 3).
86 5.3.7. Overexpressed tankyrase localizes to vesicle-like structures
To analyze the nature of the tankyrase complexes formed in vivo, we used indirect
immunofluorescence to examine the location of tankyrase in LMH cells transiently transfected
with Myc-tagged FL-tankyrase1 or tankyrase1-∆SAM. FL-
tankyrase1 showed an unexpected localization pattern with
many cells exhibiting staining of large vesicle-like structures
rather than the punctate staining pattern observed with the
endogenous protein (compare Figure 23A panel 1 and Figure
23D panel 1). The SAM domain was required for this
localization as cells expressing tankyrase1-∆SAM instead showed a diffuse cytoplasmic staining (Figure 23A panel 2). Figure 24 Tankyrase vesicles in Hela cells. Hela cells were transfected with Myc-tagged chicken tankyrase 1 and This finding suggests that the vesicle-like structures resulted processed for immunofluorescence with anti-myc antibody (green). Nuclei from SAM-mediated association between tankyrase were stained with DAPI. molecules. Similar vesicle-like structures were seen when
chicken tankyrase 1 was expressed in HeLa cells and when chicken tankyrase 2 was expressed in
LMH cells (Figure 24 and Figure 23B panel 1). When FL-tankyrase1 and tankyrase 2 were co-
expressed in LMH cells, both proteins colocalized to the vesicle-like structures, further
supporting an in vivo interaction between tankyrase 1 and tankyrase 2 (Figure 23C).
Examination of cells overexpressing Flag-tagged SAM1 or SAM2 domains revealed a diffuse cytoplasmic staining similar to that observed with tankyrase1-∆SAM, indicating that the isolated domain was not sufficient to form the vesicle-like structures (Figure 23A, panel 3 and
Figure 23B panel 2). However, when the SAM domain of tankyrase 1 or tankyrase 2 was co-
87
Figure 25 In vivo detection of SAM domain self-association by immunofluorescence. A. LMH cells co-expressing Myc-tagged FL-tankyrase 1 and Flag-tagged SAM1 domain. B. Cells co-expressing Myc-tagged tankyrase1-∆SAM and Flag-tagged SAM1 domain. C. Cells co-expressing Myc-tagged FL-tankyrase 1 and Flag-tagged SAM2 domain. D. Cells co-expressing Myc-tagged tankyrase1-∆SAM and Flag-tagged SAM2 domain. Panels 1, staining with Myc antibody; panels 2, staining with Flag antibody; panels 3, combination of the Myc and Flag signals.
88
Figure 26 Confocal microscopy shows vesicular nature of tankyrase structures. LMH cells were transfected with Myc-tagged FL-tankyrase1, and prepared for immunofluorescence with anti-Myc antibody (green). Nuclei were stained with propidium iodide (red). Sequential images are shown for one transfected cell. Distance between sequential images is ~0.3 µm along the Z-axis. The series shows opening and closing of tankyrase structures, confirming they are spherical.
expressed with FL-tankyrase 1, its distribution changed and it became also localized to the vesicle-like structures (Figure 25A and C). When the SAM domains were co-expressed with tankyrase1-∆SAM, no vesicles were formed and both proteins exhibited a diffuse localization
(Figure 25B and D). These observations support the results from the co-immunoprecipitation
experiments indicating that tankyrase SAM-SAM interactions take place in vivo as well as in
vitro. They also show that regions outside the SAM domain are required for formation of the
tankyrase vesicle-like structures. One possibility is that the ankyrin domain mediates recruitment
of tankyrase molecules by a membrane-associated protein (e.g. IRAP), with SAM-mediated
polymerization then leading to the formation of tankyrase vesicles. Although the tankyrase
89 vesicles were larger than typical cytoplasmic vesicles, optical sectioning by confocal microscopy revealed that they had the same characteristic spherical structure with a central lumen (Figure
26). The actual size of the vesicles varied considerably between cells (Figure 27A). In some cells the FL-tankyrase1 showed a punctate cytoplasmic distribution similar to the pattern obtained with endogenous tankyrase (compare panel 1 in Figure 27A with panel 1 in Figure 22D), other cells had intermediate size vesicles with a clear lumen, while others had very large vesicles
(Figure 27A, panels 2-3). The size of the structures in any one cell is determined by the amount of tankyrase DNA taken up by that cell, as transfection of cells with increasing amounts of tankyrase DNA increased the fraction of cells with large or very large vesicles (Figure 27B).
FL-tankyrase1 could be detected within 4 hours of transfection in time-course experiments to look at the initial size and shape of the vesicles (data not shown). At this stage, most cells showed a punctate staining, but some already contained vesicles with a visible lumen.
At subsequent time, points the fraction of cells with large vesicles gradually increased, suggesting that the vesicles grow over time.
Although the tankyrase vesicles are clearly unusual structures that result from abnormally high levels of tankyrase expression, their dependence on the SAM-domain, together with their capacity to expand as tankyrase levels increase, provides strong support for SAM-mediated polymerization of tankyrase occuring in vivo.
90
Figure 27 Effect of tankyrase concentration on vesicle size. A. Examples of the different size of vesicle- like structures observed in FL-tankyrase1 expressing cells. Detection with anti-Myc antibody. B. LMH cells were transfected with 0.25, 1 and 2 µg of Myc-tagged FL-tankyrase 1, supplemented with empty vector (pcDNA3) to a total of 2 µg DNA. Cells were processed for immunofluorescence 30 hours after transfection as described in Materials and Methods. Transfected cells were classified into three categories according to their vesicle size; (1) punctate, no visible lumen; (2) medium, visible lumen of up to ~15 pixels diameter; (3) large, large lumen > 15 pixels). A total of 3457 cells were analysed in three independent experiments and standard deviations were calculated.
5.3.8. Involvement of tankyrase in protein secretion
The established co-fractionation of tankyrase with low density microsomes (Chi and
Lodish, 2000; Lyons et al., 2001) prompted us to investigate whether the tankyrase vesicles might be involved in the secretory pathway. We did this by looking for tankyrase co-localization with established secretory pathway and Golgi markers. Previous studies had indicated that the extent to which endogenous tankyrase co-localizes with the Golgi marker formiminotransferase cyclodeaminase (FTCD) (Gao et al., 2002) is dependent on the cell type. In fibroblasts, tankyrase and FTCD show an almost complete overlap in their distribution (Chi and Lodish, 2000), while in cells of epithelial origin there is almost no overlap (Lyons et al., 2001). When we examined
91
Figure 28 Effects of tankyrase overexpression on the secretory pathway. A & B. Co-localization of tankyrase vesicles with FTCD. Control LMH cells (A) or LMH cells expressing Myc-tagged FL-tankyrase1 (B) were fixed and incubated with tankyrase antibody (panel A1), Myc antibody (panel B1), or FTCD antibody (Panels A2 and B2). Panels A3 and B3 show an overlay of the tankyrase and FTCD signals. Cells were counterstained with DAPI. C. Localization of GPI-GFP and VSVG3-GFP in wild type cells. LMH cells were transfected with GPI-GFP or VSVG3- GFP and analyzed by fluorescence microscopy after 30 hours at 37°C (GPI-GFP) or 32°C (permissive temperature for temperature sensitive VSVG3-GFP). D & E. Effect of Myc-FL-tankyrase1 on secretion of marker proteins. Cells were transfected with Myc-tagged FL-tankyrase 1 and GPI-GFP or VSVG3-GFP. Panels D1 & E1, exogenous tankyrase detected with Myc antibody; D2 & E2, GPI-GFP and VSVG3-GFP fluorescence; D3 & E3 the overlay of panels 1 & 2. F. GPI-GFP localizes to the periphery of the tankyrase vesicles. Confocal microscopy of LMH cells co-transfected with Myc-tagged FL-tankyrase 1 and GPI-GFP. Arrow indicates vesicle were GPI-GFP can be clearly seen at the periphery of the vesicle.
92 the distribution of endogenous tankyrase and FTCD in LMH hepatocytes, we also found little co-
localization between the two proteins (Figure 28A). However, the opposite proved true for cells
overexpressing FL-tankyrase1. In these cells, most of the normal Golgi staining was disrupted
and much of the FTCD now co-localized with the tankyrase vesicles (Figure 28B). To study
whether protein secretion was affected by this apparent disruption in Golgi integrity, we
performed co-localization experiments with two secretory pathway markers; GPI-anchored GFP
for the apical secretion pathway and VSVG-3-GFP for the basolateral secretory pathway (Keller
et al., 2001). When either marker construct was transfected into LMH cells in the absence of FL- tankyrase1, they showed the expected membrane localization pattern (Figure 28C). Likewise, when VSVG3-GFP and FL-tankyrase1 were co-transfected into the cells the VSVG3-GFP distribution remained normal. However, co-transfection of GPI-GFP and FL-tankyrase1caused the GPI-GFP to redistribute to the tankyrase vesicles (Figure 28D & E). Confocal microscopy showed that the GPI-GFP co-localized with tankyrase on the periphery of the vesicles (Figure
28F). Since the GPI anchor targets GPI-GFP to lipid bilayers, this suggests that the tankyrase vesicles are surrounded by a lipid membrane. Moreover, no GPI-GFP could be detected at the plasma membrane, indicating that normal trafficking of GPI-GFP was disrupted. These results confirm previous studies linking tankyrase to Golgi dynamics (Chi and Lodish, 2000; Sbodio et al., 2002,) and suggest that it may play a role in the apical secretion pathway.
5.3.9. Disruption of tankyrase polymers by poly-(ADP-ribosylation)
As poly(ADP-ribosyl)ation confers a large negative charge to modified proteins, charge build-up during tankyrase auto-ribosylation may cause tankyrase polymers to dissociate. To test
93 this possibility, we examined the effect of
auto-ribosylation on tankyrase solubility.
Purified tankyrase was incubated with NAD+
for various periods of time, the reactions
were then centrifuged and the soluble and
insoluble fractions were monitored for
reaction products by Western blot analysis
using antibodies to tankyrase and poly(ADP-
ribose). Incubation of tankyrase with NAD+
for only 10 minutes resulted in substantial
auto-ribosylation and a dramatic change in Figure 29 Dissociation of tankyrase complexes by tankyrase autoribosylation. A. Solubility of purified full- tankyrase solubility, as shown in Figure 29A. length tankyrase 1 after incubation with NAD+. Samples + were incubated with or without 1 mM NAD for 10 or 30 In the absence of NAD+, tankyrase was mins at 25ºC and centrifuged to separate soluble (S) from insoluble (P) material. Tankyrase was detected by Western blotting with antibody to tankyrase (α-TNK) or poly(ADP- largely insoluble, whereas after addition of ribose) (α-PAR). The arrow marks the position of + unmodified tankyrase B. Solubility of overexpressed NAD most of it became soluble and tankyrase 1 in LMH cell extracts. Cells expressing FL- tankyrase1 (TNK) or catalytically inactive FL-tankyrase1 remained in the supernatant. Thus, tankyrase (TNK-PARP†) were lysed at 25ºC in the presence (+) or absence (-) of 1 mM NAD+ and/or 10 mM 3AB, centrifuged, and tankyrase detected in the supernatant (S) or polymers are indeed dissociated by auto- pellet (P) by Western blotting with tankyrase antibody. C. Solubility of tankyrase TRF1 complexes after incubation poly(ADP-ribosyl)ation. with 1mM NAD+ . Experiment as A., but in presence of purified HIS tagged human TRF1. Arrowhead indicates position of unmodified TRF1. Anti-HIS antibody only Given that tankyrase overexpression detects unmodified TRF1. D. Poly-(ADP-ribose) binding assay. Purified TRF1 (1), FL-tankyrase1 (3), tankyrase1- causes PARP-dependent release of TRF1 ∆SAM (2) and BSA (4) were applied to a nitrocellulose membrane and incubated with purified poly-(ADP-ribose) from the telomere (Cook et al., 2002; Smith (PAR). After washing, binding of poly-(ADP-ribose) was detected by Western Blotting with anti-poly-(ADP-ribose) antibody (anti-PAR). Input TRF1 and FL-tankyrase1 both and de Lange, 2000), we expected that started with a low level of modification (- PAR panel).
94 interactions between tankyrase and its various binding partners would also be disrupted
poly(ADP-ribosyl)ation. We therefore formed complexes between tankyrase and TRF1 and
looked for an increase in TRF1 solubility after addition of NAD+ (Figure 29C). While TRF1 solubility did increase with time, a substantial amount of both TRF1 and tankyrase remained insoluble. A poly-(ADP-ribose) binding assay (Figure 29D) suggested that TRF1, like TRF2, can bind to poly(ADP-ribose) chains (Dantzer et al., 2004). In addition, both FL-tankyrase1 and tankyrase1-∆SAM bound poly-(ADP-ribose), albeit to a lesser extent. Thus, following dissociation from tankyrase, TRF1 may form secondary complexes with the poly(ADP-ribose) chains that are pelleted. Such complexes would probably be unstable in vivo as they would be degraded by PARG, the enzyme that hydrolyzes poly-(ADP-ribose) (Davidovic et al., 2001).
To test whether poly(ADP-ribosyl)ation disrupts tankyrase complexes formed in vivo, we next examined the solubility of overexpressed tankyrase in cell extracts made in the presence of
NAD+. As before, LMH cells were transiently transfected with Myc-tagged FL-tankyrase1, the
cells were lysed 24 hrs later, and the lysate was centrifuged to separate insoluble from soluble
material. In contrast to what had been observed in previous experiments where most of the FL-
tankyrase1 was insoluble, addition of NAD+ to the extraction buffer caused all of the tankyrase
to become soluble (Figure 29B). The NAD+ also caused the tankyrase to run as a smear in an
SDS gel, indicating that it had been poly(ADP-ribosyl)ated. Addition of 3AB to the extraction
buffer prevented both poly(ADP-ribosyl)ation and conversion to the soluble form. We therefore
conclude that poly(ADP-ribosyl)ation results in dissociation of tankyrase complexes that have
been formed in vivo.
95 To ensure that the ADP-ribosylation had been carried out by tankyrase rather than another PARP activity in the cell lysate, we performed the same experiment with catalytically inactive tankyrase. Mutation of two residues in the PARP domain of human tankyrase 1 has been shown to completely abrogate its activity (Cook et al., 2002). Since the PARP domains of human and chicken tankyrase 1 are 99% identical, we mutated the same residues to obtain an inactive chicken tankyrase 1. The mutant gene was transfected into LMH cells and cells were processed as before. As expected, the inactive tankyrase fractionated to the insoluble fraction after extraction in absence of NAD+. However, the protein also remained insoluble when NAD+ was
added to the extraction buffer, indicating that dissociation of tankyrase complexes containing
catalytically active tankyrase had been caused by auto-poly(ADP-ribosyl)ation (Figure 29B).
96 5.4. Discussion
Tankyrase was first identified as a telomere protein and its role in telomere length regulation
is now well established (Cook et al., 2002; Smith and de Lange, 2000; Smith et al., 1998). Thus,
the discovery that tankyrase 1 and 2 reside at diverse cellular locations and have multiple
interaction partners has raised considerable debate over their non-telomeric function(s) (Chi and
Lodish, 2000; Smith and de Lange,
1999). Because telomeres, Golgi and
centrosomes have very different
functions, it has been unclear how
tankyrases could play a common
role at each location. We now show
that tankyrases can polymerize
through their SAM domains both in
vivo and in vitro to form very large
protein complexes; the complexes Figure 30 Model for tankyrase lattice assembly and disassembly. Lattice assembly occurs through SAM-domain mediated self- formed in vitro contain >25 association of tankyrase molecules and binding of tankyrase interaction partners to the individual tankyrase ankyrin repeat modules. Tankyrase auto-ribosylation and poly(ADP-ribosyl)ation of monomers. We also show that interaction partners causes lattice disassembly. Subsequent removal of the poly(ADP-ribose) chains by PARG would then allow lattice tankyrase can use its PARP activity reassembly. to efficiently disrupt these
complexes by auto-ribosylation. Thus, tankyrases have a unique capacity to both promote
formation and dissociation of large protein complexes. We therefore propose that tankyrases are
master scaffolding molecules that use their SAM domains and PARP activity, first to assemble
97 and next to disassemble large protein lattices at different locations in the cell (Figure 30). These
lattices are likely to perform different functions at each cellular location. At telomeres, they may
regulate opening and closing of the telomeric complex, at the centrosome they may promote
spindle assembly or disassembly, and at Golgi they may regulate Golgi dynamics and vesicle
trafficking. Different signals may activate/repress tankyrase PARP activity at specific cellular
locations, thus causing a location-specific response.
5.4.1. SAM domain polymerization
SAM domains are exceptionally versatile interaction domains (Kim and Bowie 2003).
Here, we show that the tankyrase SAM domain is similar to the TEL and PH SAM domains in
that they all mediate polymerization. Structural studies indicate that other SAM domains, like the
EphB2 SAM domain, could also polymerize, but do not form large complexes in vitro (Thanos
1999, Kim 2001 and 2002). The crystal structures of the SAM domains from the EphB2 receptor and the TEL and the PH transcriptional repressors have revealed two separate interaction surfaces that can associate to form long polymers. However, the exact mechanisms of SAM- domain polymerization appear to be very variable. In the case of the TEL and PH proteins, the two surfaces interact in a head-to-tail manner to form long helical polymers (Kim et al., 2002;
Kim et al., 2001). In contrast, the EphB2 SAM domains seem to oligomerize via alternating head-to-head and tail-to-tail interactions (Smalla et al., 1999; Thanos et al., 1999). Furthermore, while the TEL and PH SAM domains form a similar quaternary helical structure, they use different residues to mediate this self-association (Kim, 2002). Our failure to create polymerization deficient SAM-domain mutants exemplifies the variability of SAM domain interactions Thus, extensive mutagenesis and structural studies will be required to determine the
98 interaction surfaces and orientation of tankyrase SAM polymers. Interestingly, our co- immunoprecipitation experiments point to the existence of an additional tankyrase self- association surface in the vicinity of the core SAM domain (Figure 22). In support of this, tankyrase 1 lacking its SAM domain behaved like a dimer or trimer rather than a monomer when analyzed by gel-filtration (Figure 21), and weak self-association of tankyrase1-∆SAM was also observed in cross-linking experiments (data not shown). The exact location of this surface is unknown, however, a 12 residue alpha-helix is predicted just downstream of the core SAM domain, and is a possible candidate to mediate this interaction.
5.4.2. Tankyrase 1 and tankyrase 2 redundancy
In Chapter Four, we described the problems associated with assessing potential functional differences between tankyrase 1 and tankyrase 2. We showed that both proteins have a similar distribution between the cytoplasm and the nucleus and that both can interact with TRF1, suggesting that the proteins are at least partially redundant. The observation that both tankyrases can interact directly with each other through their SAM domains indicates that they may cooperate in some or all tankyrase mediated processes. Whether they have the same role in these processes is unclear, and since tankyrase 1 has a domain (HPS) that is not present in tankyrase 2 it is conceivable that they have both shared and distinct functions. Knock-down of tankyrase 1 by
RNAi has been shown to induce a mitotic block, suggesting tankyrase 2 cannot substitute for tankyrase 1 function (Dynek 2004). However, since tankyrase 2 levels are typically lower than those of tankyrase 1, it is possible that the RNAi effect is merely the result of reducing the amount of total tankyrase (1+2) below a critical level necessary for proper sister chromatid resolution. To pinpoint potential differences between these proteins, it will be essential to
99 uncover the function of the HPS domain and to acquire unequivocal information on the intracellular localization of tankyrase 1 versus tankyrase 2.
5.4.3. Tankyrase as a scaffolding molecule
The ability to bring proteins together into large complexes is a characteristic that is shared by many different scaffolding proteins. These proteins are important for defining cellular structures (Bennett and Baines, 2001; Gascard and Mohandas, 2000) and they function in signal transduction pathways where they bring together groups of signalling molecules or anchor them at specific locations within a cell (Boeckers et al., 2002; Garrington and Johnson, 1999; Michel and Scott, 2002). Because tankyrases have 5 separate ankyrin repeat clusters that each interact independently with tankyrase-binding proteins, they had previously been proposed to act as scaffolding molecules (Seimiya et al., 2004; Seimiya and Smith, 2002). However, our findings add a new dimension to the proposed scaffolding function. First, we have shown that the complexes are much larger than previously thought and may actually take the form of a lattice- like structure with tankyrase polymers linked together by interaction partners that form dimers or have multiple tankyrase-binding sites (e.g. TRF1). Second, our findings indicate that tankyrases differ from other scaffolding proteins in that they have a catalytic effector function that they can use to disassemble these large lattices. It is this unique combination of scaffolding plus effector activity that leads us to suggest that their in vivo function is to act as a regulated tether that controls the formation and dissociation of large protein lattices in response to cellular signals.
The precise composition and hence role of each lattice would depend on its location in the cell.
Although tankyrases are unique in that they have both scaffolding and effector activities combined within one molecule, they share some traits with other scaffolding proteins such as the
100 Shank proteins, Sans and the AKAP (A-kinase anchoring proteins) family (Boeckers et al., 2002;
Kikkawa et al., 2003; Michel and Scott, 2002). The Shank and Sans proteins resemble tankyrases
in that they contain both ankyrin repeats and a SAM domain and are thought to oligomerize
through their SAM domain. Shank proteins are found at excitatory synapses where they organize
the postsynaptic density by linking postsynaptic receptors to the cytoskeleton (Boeckers et al.,
2002). Sans is critical for formation of the protein networks that organize hair bundles in the
inner ear and deletion of the SAM domain leads to disorganised hair bundles and deafness
(Kikkawa et al., 2003). The finding that multiple scaffolding proteins contain both ankyrin
repeats and a SAM domain suggests that this arrangement of domains is well suited for
assembling large protein networks because it allows the scaffolding protein to both self-associate
and interact with other proteins.
Although AKAPs differ in structure from tankyrases, they also target a general effector
enzyme activity to a specific location in a cell. AKAPs themselves do not have a catalytic
activity, but each AKAP binds protein kinase A (PKA) and anchors it at a different site within
the cell. This compartmentalization of PKA provides a mechanism for the cell to respond to a
localized cAMP gradient by activating PKA at that site, thus ensuring a directed response to a
general signal. It seems likely that tankyrase PARP activity, and hence complex dissociation,
will be regulated in a similar location specific manner. The localization of each AKAP-PKA
complex is achieved through a targeting domain on the AKAP that interacts with molecules at
the desired location. In some ways, this parallels the mechanism for tankyrase localization where
factors such as TRF1 or the centrosomal protein NuMA, bind the ankyrin repeats, thus tethering the tankyrase scaffold at a specific location.
101 5.4.4. Functions of tankyrase scaffolds
The ability of tankyrases to assemble and disassemble protein networks indicates that
tankyrases are unlikely to form permanent structural scaffolds but rather more dynamic
structures that are associated with processes that require rapid changes in cellular organization.
The intracellular distribution of tankyrase fits well with this hypothesis as telomeres, mitotic
spindles and the Golgi network are all structures that can undergo dramatic changes during the
cell cycle. Telomeres are large DNA-protein complexes that are thought to cycle between a
closed compact structure that protects the DNA terminus and a more open structure that makes
the terminus accessible to telomerase during S phase (Blackburn, 2001). As the telomere protein
TRF1 is an important architectural component of the telomeric complex (van Steensel and de
Lange, 1997), its removal by tankyrase would on its own be expected to cause some
decompaction of the telomeric complex. However, our finding that tankyrase forms protein
scaffolds that could effectively cross-link distant regions of the telomeric tract suggests that
disruption of the whole tankyrase-TRF1 protein lattice may have a much more profound effect
on telomere opening.
The finding that tankyrase co-localizes with Golgi and secretory vesicles, has led to the suggestion that tankyrase is somehow involved in vesicle trafficking (Chi and Lodish, 2000;
Lyons et al., 2001; Sbodio et al., 2002). Again, vesicle trafficking is a dynamic process that requires coordinated budding, storage, release, transport and fusion with the plasma-membrane, so it is fairly easy to envision a role for tankyrase scaffolds in this process. In insulin-starved adipocytes tankyrase 1 co-localizes with GLUT4 vesicles, but after insulin stimulation tankyrase remains in the cytoplasm while the GLUT4 vesicles translocate to the cell membrane (Chi and
102 Lodish, 2000). This loss of co-localization suggests that tankyrase scaffolds may serve as the
tether that prevents vesicle translocation. Tankyrase binds to the N-terminus of IRAP, the region
that protrudes from the GLUT4 vesicle, thus a tankyrase scaffold could anchor the vesicle until
the insulin response activates tankyrase PARP activity and hence scaffold dissociation.
Our studies with overexpressed tankyrase also support a role for tankyrase in secretion
because increasing the extent of SAM-mediated scaffold formation by protein overexpression led
to redistribution of a Golgi marker and inhibition of the apical but not the basolateral secretory pathway. While the mechanism underlying these effects is unclear, they might reflect a role for tankyrase scaffolds in vesicle budding from the trans-Golgi stacks, or in the translocation or fusion of specific vesicle types. Regulation of protein secretion is critical for a wide array of key cellular events but many aspects of the various secretory pathways are still poorly understood.
Thus, our finding that tankyrase scaffolds may underlie the mechanism that regulates some of these pathways is both exciting and may provide novel ways to study this complex area.
103 6. YEAST TWO-HYBRID SCREEN
6.1. Introduction
Tankyrases can poly-(ADP-ribosyl)ate a variety of proteins involved in many different processes. This modification strongly alters the physico-chemical properties of target proteins and has therefore major effects on their behavior (e.g. loss of DNA-binding activity of TRF1).
To prevent inappropriate modification of tankyrase target proteins, tankyrase activity must be regulated in the cell. Two mechanisms that control tankyrase target modification are known but these are either not generally applicable or cause only small changes in tankyrase activity.
Therefore, other regulatory processes likely exist. Several known and potential regulatory mechanisms are discussed below:
• steric hindrance
Certain proteins could block the activity of tankyrase towards its targets by acting as a steric
block preventing the PARP domain from reaching the target protein. Such proteins could
theoretically interact with either tankyrase or the target protein, with the latter providing
higher specificity. This method of regulation is thought to occur at telomeres; TIN2, a protein
that localizes to telomeres through its interaction with TRF1, prevents modification of TRF1
while not affecting the tankyrase - TRF1 interaction or intrinsic tankyrase catalytic activity
(Kim et al., 1999; Ye and de Lange, 2004). While proteins that interact directly with
tankyrase could also provide a level of regulation, currently no such proteins have been
identified. TIN2 is thought to act only on telomeric tankyrase, however, at different cellular
locations other proteins could exert the same role, thus conferring location-specific control
over tankyrase mediated protein modification. At this point it is not known how TIN2
104 inhibition on tankyrase is relieved. Plausibly, the TRF1 – TIN2 interaction is controlled by an upstream pathway.
• post-translational modifications
Often post-translational modifications, such as phosphorylation, are used in cells to alter
protein activity. Tankyrase 1 was shown to be modified by MAP kinase in response to insulin
signaling, which resulted in a small increase in poly-(ADP-ribosyl)ation activity (Chi and
Lodish, 2000). Interestingly all the tankyrase 1 consensus MAPK sites are in its N-terminal
HPS domain, far away from the catalytic domain. It is currently not clear whether indeed
these sites are phosphorylated, but if they are, it would suggest that interactions between the
HPS and PARP domains exist.
• regulation of self-association
The enzyme PARP needs to form homodimers to be active (Mendoza-Alvarez and Alvarez-
Gonzalez, 2004). The results described in chapter 5 suggest that tankyrase 1 also needs to
self-associate for optimal activity and that this is mediated by the SAM domain. Regulation
of SAM polymerization could thus control tankyrase activity. A compelling example for this
type of regulation exists; Yan (Drosophila homolog of TEL), a transcriptional repressor,
polymerizes through its SAM domain, and acts by facilitating the spreading of silent
chromatin (Kim et al., 2001; Rebay and Rubin, 1995). The protein Mae is required for
inhibition of Yan, and was recently shown to interact with the Yan SAM domain with high
affinity. This interaction results in depolymerization of Yan and abolishes its repressor activity (Qiao et al., 2004), thus providing the first example of a regulatory protein that acts
105 through inducing depolymerization of SAM polymers. Whether this type of regulation also
exists for tankyrases is unknown.
While many proteins have been identified that interact with the tankyrase ankyrin repeats, to date no proteins that bind the SAM or PARP domains have been found. For the reasons explained above, such proteins would be good candidates to regulate tankyrase activity. Thus, in order to identify potential tankyrase regulators we have carried out a yeast two-hybrid screen using the
SAM-PARP region of tankyrase 2, and have started assessing whether the identified proteins can control tankyrase activity.
106 6.2. Results
6.2.1. PARP mutagenesis and controls
To isolate proteins that interact with the tankyrase 2 SAM and PARP domains we cloned
the C-terminus of tankyrase 2 into the yeast two-hybrid bait vector pGilda (pGilda-SARP2, Lex-
A DNA binding domain fusion). However, expression of this construct proved lethal to yeast.
Since yeast does not contain any poly-(ADP-ribose) polymerase activity, and hence no poly-
(ADP-ribose) glycohydrolase activity, this toxicity is not unexpected, and was likely the result of irreversible poly-(ADP-ribosyl)ation of yeast proteins. To overcome this problem, we created an
inactive SAM-PARP protein (pGilda-mSARP2) that contained a single amino acid mutation in
the PARP domain. This construct was successfully expressed in yeast and was used for the
screen.
The library screen was performed using the Duplex-A system (Origene). The target, an
activated chicken T-cell library, was cloned into the galactose inducible vector pB42AD (B42
activation domain fusion) (Delaware Biotechnology Institute). The yeast strain used for the
screen (EGY-48) contains the essential leucine biosynthesis gene 3-isopropylmalate dehydratase
(LEU2) downstream of 6 Lex-A sites (instead of its regular upstream activating sequences).
Under normal conditions this gene is thus not expressed and the strain is auxotrophic for leucine.
However, when a Lex-A binding transcriptional activator is assembled through the interaction of
the bait with a target protein, gene expression is induced and the yeast can grow in absence of
leucine. As a second marker for interaction we used a GFP expression construct that has several
Lex-A binding sites in its promoter region (plasmid pGNG1, kind gift of Dr. I. Somssich,
(Cormack et al., 1998)). Interaction between the bait and target proteins thus results in GFP
107 expression, which can be easily detected by UV illumination. The use of this marker, instead of
the more traditional expression of β-galactosidase, provides several benefits, most notably time
and cost savings (no X-gal).
Several controls were performed preceding the library transformation. Autoactivation
controls showed that the bait by itself did not induce growth on media lacking leucine, neither
did it induce expression of GFP. A repression assay was carried out to show that LexA-mSARP2
is expressed and can enter the nucleus. In this assay, a reporter plasmid is used that expresses β-
galactosidase. Several Lex-A binding sites in the promoter region of this construct allow for
inhibition of transcription by Lex-A fusion proteins. LexA-mSARP2 could efficiently inhibit
transcription of the β-galactosidase gene, thus confirming the gene is expressed and enters the
nucleus.
6.2.2. Library screen
A total of 40 µg of library DNA was transformed into yeast strain EGY-48 containing the
GFP reporter construct and pGilda-mSARP2. Transformation yielded approximately 1.2x107 colonies that were pooled and frozen. For interaction screening, 8x107 cells were plated out on
media lacking leucine. Of the clones that grew, a total of 100 green fluorescent clones were
picked and restreaked to assure purity of the clones. An overview of the screen is given in Figure
31. Plasmid DNA was isolated from 33 clones and used in mating assays to verify their
interaction with Lex-A-mSARP2. All but one of the proteins showed interaction with Lex-A-
mSARP2 and none of the candidates was able to activate the reporters in absence of Lex-A-
mSARP2. Redundancy was tested among the 32 remaining candidates, and 3 proteins proved to
108 occur twice. A total of 29 clones remained and further mating experiments showed that they all
Figure 31 Scheme of yeast two-hybrid screen. Lex-A is Lex-A DNA binding domain, B42 is B42 activation domain, yellow section of the fusion proteins indicates the 1½ ankyrin repeats, MAX is control protein supplied with system. interact with the SAM domain of tankyrase 2.
A selection of the isolated plasmids was sequenced and their potential identity was determined by BLAST searches (Table 3). Verification of these interactions by biochemical and in vivo approaches has not been carried out and is critical to determine whether these proteins genuinely interact with tankyrase 2.
109 Table 3 Predicted identities of 6 sequenced candidates Predicted identity of Description sequenced candidates 1 SET member of nucleosome assembly protein (NAP) family, involved in apoptosis, transcription and nucleosome assembly, phosphatase 2A inhibitor 2 HMGB2 non-histone chromosomal protein, bends DNA during replication, transcription and recombination 3 growth hormone involved in regulation of postnatal body receptor (homology growth to human 5'UTR) 4 ribosomal protein structural constituent of ribosome S11 5 MTO1 mitochondrial translation optimization factor, involved in protein biosynthesis 6 RRS1 ribosome biogenesis regulatory protein homolog, involved in ribosome biogenesis
6.2.3. Secondary screen; regulators of tankyrase activity
The lethality of PARP activity in yeast provides a tool for the identification of proteins that inhibit PARP activity. Thus, in order to identify proteins that interact with tankyrase 2 and inhibit its activity, we tested whether any of the candidates could relieve PARP toxicity. Strains expressing the putative interactors were mated with a strain expressing catalytically active SAM-
PARP fused to Lex-A (Lex-A-SARP2*) followed by testing of viability and interaction. To assay for viability, the mated yeast were plated on synthetic media containing leucine; growth would thus indicate relief of PARP toxicity but not whether interaction between bait and target occurred. 24 out of 29 candidates allowed yeast expressing Lex-A-SARP2* to grow and could thus either reduce PARP activity, or serve as acceptors for poly-(ADP-ribosyl)ation and as such protect yeast proteins from being modified (the probable cause of the PARP induced lethality).
110 Of these 24, only five induced GFP expression, and none allowed growth on media lacking
leucine, suggesting that most candidates do not interact with Lex-A-SARP2*. The 5 GFP
inducing proteins might interact weakly with Lex-A-SARP* and potentially inhibit PARP
activity. The weakness of the interaction could result from small levels of poly-(ADP-
ribosyl)ation, which would hamper DNA binding. Currently the identity of these 5 candidates is unknown.
111 6.3. Discussion
A yeast two-hybrid screen was performed to isolate novel tankyrase 2 interaction partners
and potential tankyrase regulators by using the C-terminus of tankyrase 2 as bait. This region
contains the SAM and PARP domains in addition to 1½ ankyrin repeats and a SAM-PARP linker
domain. A large number of candidate tankyrase 2 interacting proteins were identified, and when
a subset was tested for interaction with the SAM domain alone, all proved to interact with this region. This is not unexpected since the SAM domain is a known protein binding motif, while
PARP catalytic domains generally do not mediate protein interactions. In addition, the SAM
domain expression construct as used in this screen contains 1½ ankyrin repeats, and while
ankyrin repeat interactions typically require several repeats, it is possible that a set of the
identified proteins actually interact with this short ankyrin region. Both SAM and ankyrin
sequences could potentially be “sticky”, which would result in the isolation of many, non-
specific interaction partners. It is therefore essential that the interactions be verified using
alternate methods. With this in mind, two of the proteins identified in the screen, SET and
HMGB2, are particularly interesting since they interact with each other (in the SET complex)
and localize to chromatin (Fan et al., 2002). The SET complex is involved in a variety of processes, including nucleosome assembly, chromatin modification and DNA nicking during apoptosis (Lieberman and Fan, 2003). The fact that two proteins of the same complex were found to interact with tankyrase 2 suggests that it is also a member of the SET complex, and in addition, the known ability of tankyrases (and PARPs in general), to dissociate DNA-binding proteins from DNA, predicts a potential function for tankyrase in this complex.
112 A secondary screen to identify candidate tankyrase 2 inhibitors was carried out by exploiting the lethality of PARP activity in yeast. Many of the candidate interactors rescued the yeast from PARP activity. This indicates that these proteins prevent yeast proteins from being
ADP-ribosylated, but does not show that the candidates actually inhibit PARP activity. Since many of the candidates do not seem to interact with active Lex-A-SARP2* in the yeast two- hybrid system, it is possible that they are poly-(ADP-ribosyl)ated and that this results either in the disruption of their interaction with tankyrase 2 or in inability to transactivate the reporters. A few candidates show weak interaction with active Lex-A-SARP2*, and could potentially inhibit its activity. The identity of these proteins is not known at this point, and further in vitro experiments are required to verify the interaction and to study their effect on tankyrase 2 PARP activity.
113 7. FINAL DISCUSSION AND FUTURE DIRECTIONS
The high level of sequence homology between tankyrase 1 and tankyrase 2 and their
presence at a wide variety of intracellular locations has prompted two fundamental questions:
• Are tankyrase 1 and tankyrase 2 redundant or do they have separate functions?
• Do tankyrases have common functions at each of their various cellular locations? If so,
how are they implemented in a location-specific manner?
The goal of this dissertation has been to address these questions by using both in vitro and in vivo
approaches to explore tankyrase biochemistry and cell biology. Below I will discuss our current
understanding of each issue, propose specific models and future directions.
7.1. Tankyrase redundancy
Because of the high homology between the two tankyrases, the redundancy issue has been
hard to resolve. Currently no tools exist to visualize tankyrase 1 and tankyrase 2 separately in
vivo. Alternative, indirect approaches that we and others have employed suggest that tankyrase 1 and tankyrase 2 act in the same compartments, interact with the same partners and associate directly with each other, all supporting possible redundancy. However, these experiments did not provide information on whether both proteins have the same functions at the various locations.
Evidence for functional redundancy comes from experiments with overexpressed tankyrase showing that both tankyrases can have a similar effect on the localization of the telomeric protein
TRF1 (Cook et al., 2002). On the other hand, RNAi knock-down of tankyrase 1 was recently shown to result in mitotic arrest, providing the most compelling evidence to date that differences in function exist between the two tankyrases (Dynek and Smith, 2004). Nevertheless, rescue of
114 the mitotic arrest by increasing the expression of tankyrase 2 was not tested, leaving open the
possibility that the effect was quantitative rather than qualitative.
To gain insight into tankyrase redundancy it will be critical to understand whether both
proteins always act together, or whether they also occur in separate spatio-temporal
compartments. A genetic tagging strategy could be pursued to overcome antibody problems.
Different epitope tags would be targeted to each tankyrase gene in a genetically pliable system,
like DT40 or mouse, and would allow for specific detection of each gene-product. We have been
working on knocking-in such a tag into the tankyrase 2 gene in DT40 and expect to obtain this
cell line in the near future. Localization of each protein will provide additional information on
differences between tankyrase 1 and tankyrase 2, but will not address intrinsic functional
differences between these two molecules. The HPS domain is an immediate target for potential
functional diversity, since this domain is only present in tankyrase 1. The role of the HPS domain
has remained elusive. Although it has been implicated in the regulation of tankyrase 1 PARP activity, the evidence was indirect (Chi and Lodish, 2000). However, its effect on tankyrase activity could easily be tested through the creation of HPS deletions and HPS - tankyrase 2 chimaeras, and by investigating whether the MAPK consensus phosphorylation sites in the HPS are phosphorylated as predicted. Functional differences might also result from sequence divergence in domains present in both proteins. This is most likely to occur for the SAM domain and the surrounding areas as these have the lowest level of conservation. Tankyrase 1 and tankyrase 2 appear to copolymerize efficiently through their SAM domains, and no other functions have yet been attributed to this region. However, the relatively high sequence divergence between the tankyrase 1 and tankyrase 2 SAM domains may indicate that
115 polymerization of tankyrase 1 and tankyrase 2 are regulated in different ways, eg. by regulatory factors specific for one or the other SAM domain.
Thus, addressing the potential redundancy of tankyrase 1 and tankyrase 2 is not straightforward, but understanding their differences could well provide new insights into their functions and could eventually be important if tankyrases prove to be potential targets in disease treatment.
7.2. General role for tankyrases
To gain insight into the general cellular role of tankyrases, we and others have studied the properties of the individual tankyrase domains and have integrated the results into a working model for tankyrase function. The ankyrin repeat domain provides tankyrases with a large and versatile protein interaction surface, allowing them to recruit up to 5 binding partners for each tankyrase molecule (Seimiya et al., 2004; Seimiya and Smith, 2002). We show that the SAM domain can mediate tankyrase polymerization, thus creating a tankyrase scaffold that can bind a large number of proteins. In addition, we show that the activated PARP domain is able to disassemble these extensive protein complexes. Thus, we propose that tankyrases are a unique type of scaffolding protein that have the ability to assemble large and complex protein networks,
through ankyrin-mediated protein recruitment and SAM polymerization, and, unlike any other
scaffolding protein, also have the intrinsic ability to dissociate such complexes. How this general
function would be implemented at the different locations where tankyrases reside is a key
question, and will only be resolved by studying tankyrase function and interactions at each
location.
116 7.2.1. Tankyrase and the secretory pathway
While tankyrases are most abundant in the cytoplasm, only their nuclear functions are understood in any detail. However, several observations suggest that their cytoplasmic functions may include participation in the secretory pathway:
• most tankyrase co-fractionates with the membrane fraction that contains Golgi-
membranes
• tankyrase colocalizes with Golgi-apparatus markers in fibroblasts
• tankyrase interacts with the cytoplasmic tail of a secreted transmembrane protein (IRAP)
• tankyrase overexpression disrupts the Golgi-apparatus and blocks secretion of apically
targeted proteins.
Clearly the above observations leave many questions unanswered. For example, we do not know if tankyrases are only targeted to IRAP containing membranes, or if are there other recruitment factors. If the first scenario is true, then tankyrases must be specifically involved in the mechanisms that pertain to controlled secretion of IRAP (and other proteins that co-secrete with IRAP). However, if tankyrases are recruited to a variety of transmembrane proteins, then they may have a much broader function in the secretory pathway (e.g. the apical secretory pathway).
Clearly a key question concerns the precise function of tankyrases in secretion. With the limited functional data that is currently available it is hard to narrow down a potential role.
However, based on comparisons to other proteins, and our model that tankyrases are polymeric master scaffolding proteins several models can be proposed.
117 The secretory pathway is responsible for the processing and transport of membrane and secreted proteins. Interestingly, polymeric proteins are common in this pathway. For their formation, transport vesicles depend on a cytoplasmic coat that consists of large protein complexes, often organized by polymeric scaffolding proteins (e.g. clathrin, Sec13p-Sec31p, …)
(Bonifacino and Glick, 2004; Kirchhausen, 2000). After vesicle budding and scission from the trans-Golgi network, uncoating occurs to allow fusion with the plasma membrane. The complexity and number of separate secretory pathways is only starting to being appreciated, and for many types of vesicles the coat proteins are unknown (Bonifacino and Glick, 2004; Elsner et al., 2003; Roth, 2004). Tankyrases display certain characteristics of coat proteins: they polymerize, they co-fractionate with the membrane fraction containing TGN membranes, and their overexpression blocks apical protein secretion and leads to formation of exceedingly large vesicles that contain Golgi markers. A speculative model for how tankyrase might act as a coat protein in the secretory pathway is as follows. First, transmembrane proteins such as IRAP recruit tankyrases to the Golgi apparatus, SAM-mediated polymerization could then participate in the budding and formation of transport vesicles (e.g. GLUT4 vesicles). Uncoating of these vesicles could next be induced by activation of tankyrase PARP activity which would disassemble the protein coat. Regulation of tankyrase activity would thus provide a mechanism to control the uncoating which is required for vesicle fusion, and potentially trafficking of the vesicles. This hypothesis is supported by the observation that insulin stimulation causes both an increase in tankyrase activity and GLUT4 vesicle translocation in adipocytes.
An equally plausible model is that tankyrases simply retain GLUT4 vesicles by assembling a tethering complex, rather than acting as a true coat protein. Tankyrases could
118 immobilize vesicles that budded from the Golgi-apparatus by linking the vesicle to an anchor site
(e.g. the cytoskeleton). Upon activation of tankyrase (through a signaling pathway) poly-(ADP-
ribosyl)ation could then lead to disruption of this tether, and subsequent release of the vesicle.
These two models are not mutually exclusive, tankyrases could be involved both in the
coat formation, and as a coat protein also provide a tether. To investigate whether tankyrases act
as coat proteins, the endogenous localization of tankyrases will have to be determined by
ultrastructural studies. Immunogold electron microscopy could reveal whether endogenous
tankyrase indeed forms a coat around vesicles. In vitro reconstitution experiments of sections of
the secretory pathway could address the function of tankyrase in vesicle budding in a controlled
environment. In the past such experiments have provided key insights into the molecular mechanisms of the secretory pathway (Balch et al., 1984; Bonifacino and Glick, 2004; Sato and
Nakano, 2004). In vivo experiments disrupting the tankyrase-IRAP interaction or tankyrase
polymerization could reveal whether tankyrase is required to form IRAP vesicles (i.e. coat
function), or for IRAP vesicle tethering. In both models, tankyrase activation is required for release / fusion of the vesicles, which is consistent with our overexpression experiments where large vesicles are formed that inhibit delivery of their load to the plasma membrane. In this situation, tankyrase levels are presumably so high that polymerization outweighs depolymerization, thus trapping the secretory load.
These models for tankyrase function at the Golgi-apparatus are only two of many alternatives; tankyrases could also be involved in processes such as the assembly of signaling complexes on the surface of the Golgi-apparatus or in Golgi fragmentation and reassembly during and after mitosis (Colanzi et al., 2003).
119 The significance of tankyrase interactions with non-transmembrane proteins in the cytoplasm (GRB14, Mcl1, TAB182) is not clear at this point, and experiments specifically disrupting these interactions will be required to understand their functional relevance.
Interestingly both GRB14 and Mcl1 are known signaling proteins, supporting the notion that tankyrase might be involved in assembling signaling complexes.
7.2.2. Tankyrase in mitosis
Tankyrases are thought to play a role in mitosis because they localize to centrosomes, where they interact with NuMA (Nuclear Mitotic Apparatus). NuMA is essential for the formation of the spindle and seems to form a protein network that is thought to stabilize the spindle (Zeng, 2000). Although tankyrase 1 interacts and co-localizes with NuMA at the centrosomes, RNAi of tankyrase 1 did not appear to interfere with spindle assembly, while it nevertheless induced a mitotic block due to unresolved sister chromatids (Dynek and Smith,
2004; Sbodio and Chi, 2002; Smith and de Lange, 1999). Thus tankyrase 1 is likely not required for the formation of the spindle (unless tankyrase 2 has a redundant function). The ability of tankyrases to dissociate large protein complexes could indicate a role in disassembly of the spindle after separation of the chromatids, a process that is poorly understood. Addressing the following questions should provide initial information about the role of tankyrases at the spindle:
• can NuMA and tankyrase form co-polymeric complexes (in vitro or in vivo), i.e.
could tankyrase be part of the centrosomal NuMA protein lattice
• can tankyrase poly-(ADP-ribosyl)ate NuMA and does this result in disruption of
NuMA self-association, if so this would support a role of tankyrases in spindle
disassembly
120 • does disruption of the tankyrase – NuMA interaction affect spindle assembly
and/or disassembly. By specifically targeting this interaction, tankyrase functions
at the spindle could be separated from its role in telomere resolution.
7.2.3. Tankyrase at telomeres
The working model for tankyrase function at telomeres is that upon activation, tankyrase poly(ADP-ribosyl)ates TRF1, which results in release of both tankyrase and TRF1 from the telomere. This in turn allows telomerase to access the telomere and results in telomere elongation. While opening of the telomeric structure might be one function of tankyrases at telomeres, in addition they could play a role in the assembly of the telomeric chromatin. As a scaffolding protein that can recruit a lot of TRF1, tankyrases may mediate the assembly of higher order telomeric chromatin. If tankyrases indeed play a role in packaging the telomere (as compared to un-packaging), then their absence should lead to a more open telomere without affecting the presence of TRF1. This could then result in increased access for telomerase and subsequent telomere elongation. However, since knock-down of tankyrase 1 leads to cell cycle arrest it will be difficult to study such effects on telomere length. Other approaches will thus be necessary to study the role of tankyrase in telomeric chromatin assembly. For example, one could perform electron microscopy experiments on in vitro assembled telomeric DNA-protein complexes; such experiments were performed with TRF1 and telomeric DNA and showed that
TRF1 can pair telomeric tracts (Griffith et al., 1998). By adding tankyrase or polymerization deficient tankyrase (lacking the SAM domain) one might observe differences in the structure of these in vitro assembled complexes and in their ability to pair multiple telomeric DNA tracts.
121 Addition of NAD to these reactions could provide additional information on the effect of poly-
(ADP-ribosylation) on telomere packaging. Alternatively, one could look at nuclease sensitivity
of telomeric DNA in presence or absence of tankyrase. The recent discovery that tankyrase activity is required for sister chromatid resolution at telomeres during mitosis suggests that telomere specific protein complexes hold the telomeres together, and that poly-(ADP- ribosyl)ation can disrupt these tethering complexes (Dynek and Smith, 2004). These observations point to yet another role for tankyrases and are consistent with our model that tankyrases are master scaffolding proteins. It will be exiting to uncover the molecular nature of this telomere specific tether, and the role of tankyrase in its formation and dissociation.
7.3. Conclusion
Tankyrases reside at many locations in the cell and may participate in a variety of processes, such as telomere elongation, mitosis, apoptosis and secretion. Functional studies at telomeres have shown that tankyrases can mediate the disassembly of protein-DNA complexes by applying their poly-(ADP-ribose) polymerase activity, and that tankyrase 1 activity is essential for progression through mitosis. However, since most tankyrase is cytoplasmic, it has been unclear what the common function of tankyrases is at all its locations. We now show that tankyrases polymerize through their SAM domains to form a platform that promotes the assembly of large protein complexes. We also show that they are efficiently depolymerized by auto-poly-(ADP- ribosyl)ation. Thus, we propose that a general role for tankyrases, both in the nucleus and the cytoplasm, is to act as a scaffolding protein that can assemble and disassemble location-specific protein complexes in a regulated manner.
122 The involvement of tankyrases in a very diverse set of processes makes their study both exciting and challenging. Our model that tankyrases act as master scaffolding proteins now provides testable hypotheses for their various location-specific functions. Key experiments to resolve these functions will involve the in vivo disruption of interactions between tankyrases and specific partner molecules, and inhibition of tankyrase polymerization. In addition, in vitro reconstitution assays could be used to pinpoint the molecular role of tankyrases in defined processes, such as vesicle budding and telomere packaging. Together, these experiments should reveal the physiological relevance of our observation that tankyrases can mediate the assembly and disassembly of large protein networks. In addition, the issues of tankyrase regulation (see
Chapter 5) and potential involvement in diseases such as diabetes and cancer, need to be addressed.
In conclusion, tankyrases appear to be multifunctional master scaffolding enzymes that are highly conserved in the animal kingdom. Thus it is likely that they play key roles both in the cytoplasm and nucleus. Understanding these different cellular functions will provide new insights, not only into tankyrase biology, but also cellular processes such as telomere packaging, mitosis and protein secretion.
123
8. REFERENCES
Albiston, A.L., McDowall, S.G., Matsacos, D., Sim, P., Clune, E., Mustafa, T., Lee, J., Mendelsohn, F.A., Simpson, R.J., Connolly, L.M. and Chai, S.Y. (2001) Evidence that the angiotensin IV (AT(4)) receptor is the enzyme insulin-regulated aminopeptidase. J Biol Chem, 276, 48623-48626.
Alvarez-Gonzalez, R. and Althaus, F.R. (1989) Poly(ADP-ribose) catabolism in mammalian cells exposed to DNA-damaging agents. Mutat Res, 218, 67-74.
Augustin, A., Spenlehauer, C., Dumond, H., Menissier-De Murcia, J., Piel, M., Schmit, A.C., Apiou, F., Vonesch, J.L., Kock, M., Bornens, M. and De Murcia, G. (2003) PARP-3 localizes preferentially to the daughter centriole and interferes with the G1/S cell cycle progression. J Cell Sci, 116, 1551-1562.
Aviv, T., Lin, Z., Lau, S., Rendl, L.M., Sicheri, F. and Smibert, C.A. (2003) The RNA-binding SAM domain of Smaug defines a new family of post-transcriptional regulators. Nat Struct Biol, 10, 614-621.
Bae, J., Donigian, J.R. and Hsueh, A.J. (2003) Tankyrase 1 interacts with Mcl-1 proteins and inhibits their regulation of apoptosis. J Biol Chem, 278, 5195-5204.
Balch, W.E., Dunphy, W.G., Braell, W.A. and Rothman, J.E. (1984) Reconstitution of the transport of protein between successive compartments of the Golgi measured by the coupled incorporation of N-acetylglucosamine. Cell, 39, 405-416.
Barrera, F.N., Poveda, J.A., Gonzalez-Ros, J.M. and Neira, J.L. (2003) Binding of the C-terminal sterile alpha motif (SAM) domain of human p73 to lipid membranes. J Biol Chem, 278, 46878-46885.
Bennett, V. and Baines, A.J. (2001) Spectrin and ankyrin-based pathways: metazoan inventions for integrating cells into tissues. Physiol Rev, 81, 1353-1392.
Bereziat, V., Kasus-Jacobi, A., Perdereau, D., Cariou, B., Girard, J. and Burnol, A.F. (2002) Inhibition of insulin receptor catalytic activity by the molecular adapter Grb14. J Biol Chem, 277, 4845-4852.
Blackburn, E.H. (2001) Switching and signaling at the telomere. Cell, 106, 661-673.
124 Boeckers, T.M., Bockmann, J., Kreutz, M.R. and Gundelfinger, E.D. (2002) ProSAP/Shank proteins - a family of higher order organizing molecules of the postsynaptic density with an emerging role in human neurological disease. J Neurochem, 81, 903-910.
Bonifacino, J.S. and Glick, B.S. (2004) The mechanisms of vesicle budding and fusion. Cell, 116, 153-166.
Broccoli, D., Chong, L., Oelmann, S., Fernald, A.A., Marziliano, N., van Steensel, B., Kipling, D., Le Beau, M.M. and de Lange, T. (1997a) Comparison of the human and mouse genes encoding the telomeric protein, TRF1: chromosomal localization, expression and conserved protein domains. Hum Mol Genet, 6, 69-76.
Broccoli, D., Smogorzewska, A., Chong, L. and de Lange, T. (1997b) Human telomeres contain two distinct Myb-related proteins, TRF1 and TRF2. Nat Genet, 17, 231-235.
Brown, W.R., Hubbard, S.J., Tickle, C. and Wilson, S.A. (2003) The chicken as a model for large-scale analysis of vertebrate gene function. Nat Rev Genet, 4, 87-98.
Buchberger, A., Theyssen, H., Schroder, H., McCarty, J.S., Virgallita, G., Milkereit, P., Reinstein, J. and Bukau, B. (1995) Nucleotide-induced conformational changes in the ATPase and substrate binding domains of the DnaK chaperone provide evidence for interdomain communication. J Biol Chem, 270, 16903-16910.
Buerstedde, J.M. and Takeda, S. (1991) Increased ratio of targeted to random integration after transfection of chicken B cell lines. Cell, 67, 179-188.
Cariou, B., Perdereau, D., Cailliau, K., Browaeys-Poly, E., Bereziat, V., Vasseur-Cognet, M., Girard, J. and Burnol, A.F. (2002) The adapter protein ZIP binds Grb14 and regulates its inhibitory action on insulin signaling by recruiting protein kinase Czeta. Mol Cell Biol, 22, 6959-6970.
Cary, R.B., Peterson, S.R., Wang, J., Bear, D.G., Bradbury, E.M. and Chen, D.J. (1997) DNA looping by Ku and the DNA-dependent protein kinase. Proc Natl Acad Sci U S A, 94, 4267-4272.
Chambon, P., Weill, J.D. and Mandel, P. (1963) Nicotinamide mononucleotide activation of new DNA-dependent polyadenylic acid synthesizing nuclear enzyme. Biochem Biophys Res Commun, 11, 39-43.
Chang, W., Dynek, J.N. and Smith, S. (2003) TRF1 is degraded by ubiquitin-mediated proteolysis after release from telomeres. Genes Dev, 17, 1328-1333.
125 Chi, N.W. and Lodish, H.F. (2000) Tankyrase is a golgi-associated mitogen-activated protein kinase substrate that interacts with IRAP in GLUT4 vesicles. J Biol Chem, 275, 38437- 38444.
Chong, L., van Steensel, B., Broccoli, D., Erdjument-Bromage, H., Hanish, J., Tempst, P. and de Lange, T. (1995) A human telomeric protein. Science, 270, 1663-1667.
Colanzi, A., Suetterlin, C. and Malhotra, V. (2003) Cell-cycle-specific Golgi fragmentation: how and why? Curr Opin Cell Biol, 15, 462-467.
Cook, B.D., Dynek, J.N., Chang, W., Shostak, G. and Smith, S. (2002) Role for the related poly(ADP-Ribose) polymerases tankyrase 1 and 2 at human telomeres. Mol Cell Biol, 22, 332-342.
Cormack, R.S., Hahlbrock, K. and Somssich, I.E. (1998) Isolation of putative plant transcriptional coactivators using a modified two-hybrid system incorporating a GFP reporter gene. Plant J, 14, 685-692.
Daly, R.J. (1998) The Grb7 family of signalling proteins. Cell Signal, 10, 613-618.
D'Amours, D., Desnoyers, S., D'Silva, I. and Poirier, G.G. (1999) Poly(ADP-ribosyl)ation reactions in the regulation of nuclear functions. Biochem J, 342 ( Pt 2), 249-268.
Dantzer, F., Giraud-Panis, M.J., Jaco, I., Ame, J.C., Schultz, I., Blasco, M., Koering, C.E., Gilson, E., Menissier-de Murcia, J., de Murcia, G. and Schreiber, V. (2004) Functional interaction between poly(ADP-Ribose) polymerase 2 (PARP-2) and TRF2: PARP activity negatively regulates TRF2. Mol Cell Biol, 24, 1595-1607.
Davidovic, L., Vodenicharov, M., Affar, E.B. and Poirier, G.G. (2001) Importance of poly(ADP- ribose) glycohydrolase in the control of poly(ADP-ribose) metabolism. Exp Cell Res, 268, 7-13.
Dynek, J.N. and Smith, S. (2004) Resolution of sister telomere association is required for progression through mitosis. Science, 304, 97-100.
Elsner, M., Hashimoto, H. and Nilsson, T. (2003) Cisternal maturation and vesicle transport: join the band wagon! (Review). Mol Membr Biol, 20, 221-229.
Errami, A., Finnie, N.J., Morolli, B., Jackson, S.P., Lohman, P.H. and Zdzienicka, M.Z. (1998) Molecular and biochemical characterization of new X-ray-sensitive hamster cell mutants defective in Ku80. Nucleic Acids Res, 26, 4332-4338.
126 Fan, Z., Beresford, P.J., Zhang, D. and Lieberman, J. (2002) HMG2 interacts with the nucleosome assembly protein SET and is a target of the cytotoxic T-lymphocyte protease granzyme A. Mol Cell Biol, 22, 2810-2820.
Gao, Y.S., Vrielink, A., MacKenzie, R. and Sztul, E. (2002) A novel type of regulation of the vimentin intermediate filament cytoskeleton by a Golgi protein. Eur J Cell Biol, 81, 391- 401.
Garrington, T.P. and Johnson, G.L. (1999) Organization and regulation of mitogen-activated protein kinase signaling pathways. Curr Opin Cell Biol, 11, 211-218.
Gascard, P. and Mohandas, N. (2000) New insights into functions of erythroid proteins in nonerythroid cells. Curr Opin Hematol, 7, 123-129.
Golub, T.R., Goga, A., Barker, G.F., Afar, D.E., McLaughlin, J., Bohlander, S.K., Rowley, J.D., Witte, O.N. and Gilliland, D.G. (1996) Oligomerization of the ABL tyrosine kinase by the Ets protein TEL in human leukemia. Mol Cell Biol, 16, 4107-4116.
Gossen, M. and Bujard, H. (1992) Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proceedings of the National Academy of Sciences USA, 89, 5547-5551.
Griffith, J., Bianchi, A. and de Lange, T. (1998) TRF1 promotes parallel pairing of telomeric tracts in vitro. J Mol Biol, 278, 79-88.
Kaminker, P.G., Kim, S.H., Taylor, R.D., Zebarjadian, Y., Funk, W.D., Morin, G.B., Yaswen, P. and Campisi, J. (2001) TANK2, a new TRF1-associated poly(ADP-ribose) polymerase, causes rapid induction of cell death upon overexpression. J Biol Chem, 276, 35891- 35899.
Kandror, K.V., Yu, L. and Pilch, P.F. (1994) The major protein of GLUT4-containing vesicles, gp160, has aminopeptidase activity. J Biol Chem, 269, 30777-30780.
Kasus-Jacobi, A., Perdereau, D., Auzan, C., Clauser, E., Van Obberghen, E., Mauvais-Jarvis, F., Girard, J. and Burnol, A.F. (1998) Identification of the rat adapter Grb14 as an inhibitor of insulin actions. J Biol Chem, 273, 26026-26035.
Keller, P., Toomre, D., Diaz, E., White, J. and Simons, K. (2001) Multicolour imaging of post- Golgi sorting and trafficking in live cells. Nat Cell Biol, 3, 140-149.
Keller, S.R. (2003) The insulin-regulated aminopeptidase: a companion and regulator of GLUT4. Front Biosci, 8, s410-420.
127 Keller, S.R. (2004) Role of the insulin-regulated aminopeptidase IRAP in insulin action and diabetes. Biol Pharm Bull, 27, 761-764.
Kickhoefer, V.A., Siva, A.C., Kedersha, N.L., Inman, E.M., Ruland, C., Streuli, M. and Rome, L.H. (1999) The 193-kD vault protein, VPARP, is a novel poly(ADP-ribose) polymerase. J Cell Biol, 146, 917-928.
Kikkawa, Y., Shitara, H., Wakana, S., Kohara, Y., Takada, T., Okamoto, M., Taya, C., Kamiya, K., Yoshikawa, Y., Tokano, H., Kitamura, K., Shimizu, K., Wakabayashi, Y., Shiroishi, T., Kominami, R. and Yonekawa, H. (2003) Mutations in a new scaffold protein Sans cause deafness in Jackson shaker mice. Hum Mol Genet, 12, 453-461.
Kim, C.A. and Bowie, J.U. (2003) SAM domains: uniform structure, diversity of function. Trends Biochem Sci, 28, 625-628.
Kim, C.A., Gingery, M., Pilpa, R.M. and Bowie, J.U. (2002) The SAM domain of polyhomeotic forms a helical polymer. Nat Struct Biol, 9, 453-457.
Kim, C.A., Phillips, M.L., Kim, W., Gingery, M., Tran, H.H., Robinson, M.A., Faham, S. and Bowie, J.U. (2001) Polymerization of the SAM domain of TEL in leukemogenesis and transcriptional repression. Embo J, 20, 4173-4182.
Kim, S.H., Kaminker, P. and Campisi, J. (1999) TIN2, a new regulator of telomere length in human cells. Nat Genet, 23, 405-412.
Kirchhausen, T. (2000) Clathrin. Annu Rev Biochem, 69, 699-727.
Konrad, J.P., Mills, W., Easty, D.J. and Farr, C.J. (1999) Cloning and characterisation of the chicken gene encoding the telomeric protein TRF2. Gene, 239, 81-90.
Kuimov, A.N., Kuprash, D.V., Petrov, V.N., Vdovichenko, K.K., Scanlan, M.J., Jongeneel, C.V., Lagarkova, M.A. and Nedospasov, S.A. (2001) Cloning and characterization of TNKL, a member of tankyrase gene family. Genes Immun, 2, 52-55.
Lacronique, V., Boureux, A., Valle, V.D., Poirel, H., Quang, C.T., Mauchauffe, M., Berthou, C., Lessard, M., Berger, R., Ghysdael, J. and Bernard, O.A. (1997) A TEL-JAK2 fusion protein with constitutive kinase activity in human leukemia. Science, 278, 1309-1312.
Li, B., Oestreich, S. and de Lange, T. (2000) Identification of human Rap1: implications for telomere evolution. Cell, 101, 471-483.
Lieberman, J. and Fan, Z. (2003) Nuclear war: the granzyme A-bomb. Curr Opin Immunol, 15, 553-559.
128 Lyons, R.J., Deane, R., Lynch, D.K., Ye, Z.S., Sanderson, G.M., Eyre, H.J., Sutherland, G.R. and Daly, R.J. (2001) Identification of a novel human tankyrase through its interaction with the adaptor protein Grb14. J Biol Chem, 276, 17172-17180.
Martinez-Arca, S., Lalioti, V.S. and Sandoval, I.V. (2000) Intracellular targeting and retention of the glucose transporter GLUT4 by the perinuclear storage compartment involves distinct carboxyl-tail motifs. J Cell Sci, 113 ( Pt 10), 1705-1715.
Matsumoto, H., Rogi, T., Yamashiro, K., Kodama, S., Tsuruoka, N., Hattori, A., Takio, K., Mizutani, S. and Tsujimoto, M. (2000) Characterization of a recombinant soluble form of human placental leucine aminopeptidase/oxytocinase expressed in Chinese hamster ovary cells. Eur J Biochem, 267, 46-52.
Mendoza-Alvarez, H. and Alvarez-Gonzalez, R. (1999) Biochemical characterization of mono(ADP-ribosyl)ated poly(ADP-ribose) polymerase. Biochemistry, 38, 3948-3953.
Mendoza-Alvarez, H. and Alvarez-Gonzalez, R. (2004) The 40 kDa carboxy-terminal domain of poly(ADP-ribose) polymerase-1 forms catalytically competent homo- and heterodimers in the absence of DNA. J Mol Biol, 336, 105-114.
Michaely, P., Tomchick, D.R., Machius, M. and Anderson, R.G. (2002) Crystal structure of a 12 ANK repeat stack from human ankyrinR. Embo J, 21, 6387-6396.
Michel, J.J. and Scott, J.D. (2002) AKAP mediated signal transduction. Annu Rev Pharmacol Toxicol, 42, 235-257.
Mohler, P.J., Gramolini, A.O. and Bennett, V. (2002) Ankyrins. J Cell Sci, 115, 1565-1566.
Monz, D., Munnia, A., Comtesse, N., Fischer, U., Steudel, W.I., Feiden, W., Glass, B. and Meese, E.U. (2001) Novel tankyrase-related gene detected with meningioma-specific sera. Clin Cancer Res, 7, 113-119.
Mosavi, L.K., Cammett, T.J., Desrosiers, D.C. and Peng, Z.Y. (2004) The ankyrin repeat as molecular architecture for protein recognition. Protein Sci, 13, 1435-1448.
Oei, S.L., Griesenbeck, J., Schweiger, M. and Ziegler, M. (1998) Regulation of RNA polymerase II-dependent transcription by poly(ADP-ribosyl)ation of transcription factors. J Biol Chem, 273, 31644-31647.
Peeters, P., Raynaud, S.D., Cools, J., Wlodarska, I., Grosgeorge, J., Philip, P., Monpoux, F., Van Rompaey, L., Baens, M., Van den Berghe, H. and Marynen, P. (1997) Fusion of TEL, the ETS-variant gene 6 (ETV6), to the receptor-associated kinase JAK2 as a result of t(9;12) in a lymphoid and t(9;15;12) in a myeloid leukemia. Blood, 90, 2535-2540.
129 Pleschke, J.M., Kleczkowska, H.E., Strohm, M. and Althaus, F.R. (2000) Poly(ADP-ribose) binds to specific domains in DNA damage checkpoint proteins. J Biol Chem, 275, 40974- 40980.
Ponting, C.P. (1995) SAM: a novel motif in yeast sterile and Drosophila polyhomeotic proteins. Protein Sci, 4, 1928-1930.
Qiao, F., Song, H., Kim, C.A., Sawaya, M.R., Hunter, J.B., Gingery, M., Rebay, I., Courey, A.J. and Bowie, J.U. (2004) Derepression by depolymerization; structural insights into the regulation of Yan by Mae. Cell, 118, 163-173.
Ramachander, R. and Bowie, J.U. (2004) SAM domains can utilize similar surfaces for the formation of polymers and closed oligomers. J Mol Biol, 342, 1353-1358.
Rebay, I. and Rubin, G.M. (1995) Yan functions as a general inhibitor of differentiation and is negatively regulated by activation of the Ras1/MAPK pathway. Cell, 81, 857-866.
Rippmann, J.F., Damm, K. and Schnapp, A. (2002) Functional characterization of the poly(ADP- ribose) polymerase activity of tankyrase 1, a potential regulator of telomere length. J Mol Biol, 323, 217-224.
Rogi, T., Tsujimoto, M., Nakazato, H., Mizutani, S. and Tomoda, Y. (1996) Human placental leucine aminopeptidase/oxytocinase. A new member of type II membrane-spanning zinc metallopeptidase family. J Biol Chem, 271, 56-61.
Ross, S.A., Herbst, J.J., Keller, S.R. and Lienhard, G.E. (1997) Trafficking kinetics of the insulin-regulated membrane aminopeptidase in 3T3-L1 adipocytes. Biochem Biophys Res Commun, 239, 247-251.
Roth, M. (2004) New candidates for vesicle coat proteins. Nat Cell Biol, 6, 384-385.
Ruf, A., Mennissier de Murcia, J., de Murcia, G. and Schulz, G.E. (1996) Structure of the catalytic fragment of poly(AD-ribose) polymerase from chicken. Proc Natl Acad Sci U S A, 93, 7481-7485.
Sato, K. and Nakano, A. (2004) Reconstitution of coat protein complex II (COPII) vesicle formation from cargo-reconstituted proteoliposomes reveals the potential role of GTP hydrolysis by Sar1p in protein sorting. J Biol Chem, 279, 1330-1335.
Sbodio, J.I. and Chi, N.W. (2002) Identification of a tankyrase-binding motif shared by IRAP, TAB182, and human TRF1 but not mouse TRF1. NuMA contains this RXXPDG motif and is a novel tankyrase partner. J Biol Chem, 277, 31887-31892.
130 Sbodio, J.I., Lodish, H.F. and Chi, N.W. (2002) Tankyrase-2 oligomerizes with tankyrase-1 and binds to both TRF1 (telomere-repeat-binding factor 1) and IRAP (insulin-responsive aminopeptidase). Biochem J, 361, 451-459.
Schultz, J., Ponting, C.P., Hofmann, K. and Bork, P. (1997) SAM as a protein interaction domain involved in developmental regulation. Protein Sci, 6, 249-253.
Sedgwick, S.G. and Smerdon, S.J. (1999) The ankyrin repeat: a diversity of interactions on a common structural framework. Trends Biochem Sci, 24, 311-316.
Seimiya, H., Muramatsu, Y., Smith, S. and Tsuruo, T. (2004) Functional subdomain in the ankyrin domain of Tankyrase 1 required for poly(ADP-ribosyl)ation of TRF1 and telomere elongation. Mol Cell Biol, 24, 1944-1955.
Seimiya, H. and Smith, S. (2002) The telomeric poly(ADP-ribose) polymerase, tankyrase 1, contains multiple binding sites for telomeric repeat binding factor 1 (TRF1) and a novel acceptor, 182-kDa tankyrase-binding protein (TAB182). J Biol Chem, 277, 14116-14126.
Shall, S. and de Murcia, G. (2000) Poly(ADP-ribose) polymerase-1: what have we learned from the deficient mouse model? Mutat Res, 460, 1-15.
Slot, J.W., Geuze, H.J., Gigengack, S., James, D.E. and Lienhard, G.E. (1991) Translocation of the glucose transporter GLUT4 in cardiac myocytes of the rat. Proc Natl Acad Sci U S A, 88, 7815-7819.
Smalla, M., Schmieder, P., Kelly, M., Ter Laak, A., Krause, G., Ball, L., Wahl, M., Bork, P. and Oschkinat, H. (1999) Solution structure of the receptor tyrosine kinase EphB2 SAM domain and identification of two distinct homotypic interaction sites. Protein Sci, 8, 1954-1961.
Smith, S. (2001) The world according to PARP. Trends Biochem Sci, 26, 174-179.
Smith, S. and de Lange, T. (1999) Cell cycle dependent localization of the telomeric PARP, tankyrase, to nuclear pore complexes and centrosomes. J Cell Sci, 112 ( Pt 21), 3649- 3656.
Smith, S. and de Lange, T. (2000) Tankyrase promotes telomere elongation in human cells. Curr Biol, 10, 1299-1302.
Smith, S., Giriat, I., Schmitt, A. and de Lange, T. (1998) Tankyrase, a poly(ADP-ribose) polymerase at human telomeres. Science, 282, 1484-1487.
Thanos, C.D., Goodwill, K.E. and Bowie, J.U. (1999) Oligomeric structure of the human EphB2 receptor SAM domain. Science, 283, 833-836.
131 Tran, H.H., Kim, C.A., Faham, S., Siddall, M.C. and Bowie, J.U. (2002) Native interface of the SAM domain polymer of TEL. BMC Struct Biol, 2, 5. van Steensel, B. and de Lange, T. (1997) Control of telomere length by the human telomeric protein TRF1. Nature, 385, 740-743.
Venkatesan, R.N. (2000) Vertebrate telomeres and telomere-interacting proteins (Ph.D. Thesis). University of Nebraska--Lincoln, Lincoln, Nebraska.
Waters, S.B., D'Auria, M., Martin, S.S., Nguyen, C., Kozma, L.M. and Luskey, K.L. (1997) The amino terminus of insulin-responsive aminopeptidase causes Glut4 translocation in 3T3- L1 adipocytes. J Biol Chem, 272, 23323-23327.
Wei, C. and Price, M. (2003) Protecting the terminus: t-loops and telomere end-binding proteins. Cell Mol Life Sci, 60, 2283-2294.
Winding, P. and Berchtold, M.W. (2001) The chicken B cell line DT40: a novel tool for gene disruption experiments. J Immunol Methods, 249, 1-16.
Ye, J.Z. and de Lange, T. (2004) TIN2 is a tankyrase 1 PARP modulator in the TRF1 telomere length control complex. Nat Genet, 36, 618-623.
Zeng, C. (2000) NuMA: a nuclear protein involved in mitotic centrosome function. Microsc Res Tech, 49, 467-477.
Zhong, Z., Shiue, L., Kaplan, S. and de Lange, T. (1992) A mammalian factor that binds telomeric TTAGGG repeats in vitro. Mol Cell Biol, 12, 4834-4843.
132