Neurodegeneration: Nicked to Death

Dispatch R55

disrupt either the monkey’s performance on the trained task or its normal behaviour in the home cage. Over several days of conditioning, the torque direction of recording site effects shifted towards the direction of stimulation site effects. The effects from a third, control (Ctrl) electrode did not change. Recording of muscle EMG confirmed that, after conditioning with the artificial connection, new muscles were activated by intracortical microstimulation at the recording site, which had previously only been activated from the stimulation site. In some cases, changes were stable for up to a week after the end of conditioning. The basic result was confirmed in 13 out of 17 tests carried out in two monkeys. In a further sequence of experiments, Jackson et al. [1] established that these changes occurred only when stimulation followed spike activity by up to 50 milliseconds: longer delays or fixed frequency stimulation did not produce the same conditioned changes. The authors interpret these changes as arising from the coincidence of stimulus-evoked activity (at the stimulation site) with synchronous firing of neurons (at the recording site) inducing plasticity in horizontal or descending motor pathways. Neurons that are synchronized at the cortical level are known to share common outputs [4], and the conditioning stimuli may have increased such synchrony. The fact that conditioning was absent unless stimuli were closely time-locked to spike events suggests that the underlying mechanism may be related to spiketiming dependent plasticity, which has previously been described at the cellular level [5]. The experiments demonstrate that natural patterns of cortical spiking in vivo during normal behaviour can lead to input-specific Hebbian plasticity when paired with stimulation at a second site. These plastic changes probably occur at multiple levels, involving not only short and long-distance connections within the cortex [6] but also at subcortical sites and in pathways descending to the spinal cord. The potential for plastic change within the motor cortex was first recognized nearly a century ago by Sherrington and Brown [7]. The

plasticity of the primary motor cortex (M1) output is undoubtedly of great importance for adaptive motor learning in primates, and particularly humans, a capacity that is essential for sophisticated behaviors as diverse as the manufacture and use of tools, sport and music making. This plasticity has been shown to be strongly influenced by motor learning, use-dependence and sensory stimulation [8]. It is also fundamental to modern ideas of neurorehabilitation and the compensatory changes that occur in the injured nervous system [9,10]. As evidenced by the recent Society for Neuroscience meeting in Atlanta, the brain–machine interface field is now a massive multimillion dollar enterprise. Many labs are working on artificial connections that could form the basis of a neural prosthesis to replace neural pathways lost through injury or disease, such as after spinal cord injury or stroke. The Neurochip used by Jackson et al. [1] links motor cortex sites just a few millimeters apart, but it could be adapted for much longer interactions, linking, for example, motor cortex to spinal cord or peripheral nerve severed from supraspinal control by injury. But in addition, the new results suggest that such a prosthesis could have additional rehabilitative roles in cases of partial injury by strengthening surviving projections between connected sites. This type

of approach might for instance be of considerable relevance for restoring further function to patients with incomplete spinal lesions; such patients constitute the majority of the spinal cord injured community. References 1. Jackson, A., Mavoori, J., and Fetz, E.E. (2006). Long-term motor cortex plasticity induced by an electronic neural implant. Nature 444, 56–60. 2. Markram, H., Lubke, J., Frotscher, M., and Sakmann, B. (1997). Regulation of synaptic efficacy by coincidence of postsynaptic APs and EPSPs. Science 275, 213–215. 3. Mavoori, J., Jackson, A., Diorio, C., and Fetz, E.E. (2005). An autonomous implantable computer for recording and stimulation in unrestrained primates. J. Neurosci. Meth. 148, 71–77. 4. Jackson, A., Gee, V.J., Baker, S.N., and Lemon, R.N. (2003). Synchrony between neurons with similar muscle fields in monkey motor cortex. Neuron 38, 115–125. 5. Hess, G., Aizenman, C.D., and Donoghue, J.P. (1996). Conditions for the induction of long term potentiation in layer II/III horizontal connections of the rat motor cortex. J. Neurophysiol. 75, 1765–1778. 6. Dan, Y., and Poo, M. (2004). Spike timing-dependent plasticity of neural circuits. Neuron 44, 23–30. 7. Brown, T.G., and Sherrington, C.S. (1912). On the instability of a cortical point. Proc. R. Soc. Lond. B 85, 250–277. 8. Nudo, R.J., Milliken, G.W., Jenkins, W.M., and Merzenich, M.M. (1996). Use-dependent alterations of movement representations in primary motor cortex of adult squirrel monkeys. J. Neurosci. 16, 785–807. 9. Nudo, R.J., Wise, B.M., SiFuentes, F., and Milliken, G.W. (1996). Neural substrates for the effects of rehabilitative training on motor recovery after ischemic infarct. Science 272, 1791–1794. 10. Ward, N.S. (2005). Neural plasticity and recovery of function. Prog. Brain Res. 150, 527–535.

UCL Institute of Neurology, Queen Square, London WC1N 3BG, UK. E-mail: [email protected] DOI: 10.1016/j.cub.2006.12.017

Neurodegeneration: Nicked to Death Ataxia oculomotor apraxia-1 is a neurological disorder that arises from mutations in the gene encoding the protein aprataxin. A recent study demonstrates that aprataxin is critical for the processing of obstructive DNA termini, suggesting a broader role for DNA single-strand break repair in neurodegenerative disease. David M. Wilson, III1,* and Mark P. Mattson2 Neurons are postmitotic cells that must survive and function properly for the entire lifetime of the organism. Because they cannot be

replaced and are subjected to high metabolic stress, mechanisms for coping with damaged molecules may be particularly important in these cells. Indeed, human neurodegenerative disorders, such as Alzheimer’s and Parkinson’s

Current Biology Vol 17 No 2 R56

3’-obstructive ends

5’-obstructive ends

P 3’-TOPO1 = TDP1 OH

PG P 5’-AMP = APTX

3’-PG = APE1 (TDP1)

OHOH

P P 5’-OH = PNKP

3’-P = PNKP

OH dRP

UA P 5’-dRP = Polβ

3’-UA = APE1 Wild-type cells (dividing or non-dividing) Post-mitotic neurons

Dividing cells

AOA1 or SCAN1 mutations

SSBR

aging and disease Nuclease-dependent HR

Blocked transcription

?

Neuronal cell death

Error-free repair

Current Biology

Figure 1. Outcomes of failed repair of non-conventional 30 - or 50 -obstructive ends. Trapped 30 -TOPO1–DNA intermediates (blue circle, within red box) can arise from stalled TOPO1 reactions. 30 -phosphate (P) and -phosphoglycolate (PG) damages are products of free radical attack of DNA. In some instances, abasic (AP) sites are cleaved by multifunctional DNA glycosylases, which can generate 30 -P or 30 -unsaturated aldehyde (UA) residues. 50 -AMP groups (blue lettering, within red box) are produced by failed DNA ligase reactions. On occasion, free radical attack of DNA can leave behind 50 -OH ends. 50 -dRP residues are generated by incision of DNA by an abasic endonuclease. The major enzymes for excising terminal damages or for restoring conventional termini are denoted. After termini ‘clean-up’, typical SSB repair proceeds via gap-filling by DNA Polb, and sealing of the nick by an XRCC1–LIG3a complex. In cells deficient for SSB repair, as a result of genetic mutations (AOA1 and SCAN1) or age-related disease, SSBs accumulate. In non-dividing cells, such as neuronal tissue, accumulation of SSBs leads to impaired transcription, which in turn leads to cell death. In dividing cells, SSBs are converted to double-strand breaks upon replication fork collapse, and these double-strand break intermediates are repaired efficiently and accurately by a nuclease-dependent homologous recombination (HR) pathway. Although the molecular details are presently uncertain (shown as ?), for the removal of 30 -blocking termini during HR, the ERCC1/XPF complex is a possible nuclease, while the flap endonucleases FEN1 or EXO1 are potential enzymes for processing 50 -blocking termini.

diseases, involve the abnormal accumulation of damaged proteins [1], and other syndromes, such as ataxia telangiectasia (AT), have been associated with defects in DNA-damage processing [2]. Work from El-Khamisy et al. [3] and a more recent study by Ahel et al. [4] have revealed that inefficient repair of DNA single-strand breaks (SSBs) can give rise to neurodegenerative disease, in particular, spinocerebellar ataxia with axonal neuropathy-1 (SCAN1) and ataxia oculomotor apraxia-1 (AOA1), respectively. These data provide evidence that non-replicating, post-mitotic neurons are particularly sensitive to the accumulation of DNA SSBs.

SSBs, one of the most common lesions formed in chromosomal DNA, are generated by the attack of reactive oxygen species [5] or as natural intermediates during certain DNA transactions, including repair and replication [6]. In both instances, SSBs can harbor non-conventional 30 or 50 termini, such as phosphates, phosphoglycolates, or trapped polypeptides, which present obstacles to polymerization and ligation activities. To remove such obstructions, cells undergo SSB repair, a process related to the more classical base excision repair (BER) pathway [7]. The SSB repair proteins excise terminal blocking groups (Figure 1), permitting gap-filling synthesis

and sealing of the final nick in DNA. AOA1 and SCAN1 are hereditary autosomal recessive ataxias affecting primarily motor coordination, i.e. gaze, speech, gait and balance [8–10]. Unlike patients suffering from other DNA-repair-related disorders characterized by neurological dysfunction, such as AT [2], patients suffering from AOA1 and SCAN1 lack non-neurological symptoms, most notably the increased cancer incidence. The recent cloning and characterization of the genes defective in AOA1 and SCAN1 has shed light on why non-replicating neuronal cells may be exquisitely sensitive to the accumulation of certain DNA intermediates. The gene mutated in SCAN1 encodes the protein tyrosyl-DNA phosphodiesterase 1 (TDP1) [8], which was shown to be the primary enzyme for excising covalently linked 30 -topoisomerase I (TOPO1) –DNA intermediates [11]. TOPO1 binds and cleaves one strand of DNA via a transient covalent protein–nucleic acid complex to relieve topological strains (i.e. supercoils) generated during repair, replication or recombination. However, the TOPO1–DNA intermediate can become ‘trapped’ either when in close proximity to DNA damage, such as oxidative lesions, or upon exposure of cells to the chemotherapeutic agent camptothecin [6]. TDP1 removes the covalently linked 30 -TOPO1 protein moiety, leaving behind a 30 -phosphate group, which is excised by polynucleotide kinase/phosphatase (PNKP) [3]. These enzymatic steps generate a normal 30 -hydroxyl end, which is suitable for polymerase extension and subsequent ligation. The studies of El-Khamisy et al. [3] indicate that cells from SCAN1 patients are unable to process specific 30 -obstructive termini efficiently. El-Khamisy and colleagues [3] found that SCAN1 cells are defective in the repair of camptothecin-induced SSBs that arise independently of DNA replication. Moreover, the authors report that TDP1 directly interacts

Dispatch R57

with DNA ligase IIIa (LIG3a), a partner of the SSB repair protein XRCC1, and exists in a complex with LIG3a, XRCC1, and PNKP that is capable of repairing tyrosyl-containing oligonucleotide substrates that mimic TOPO1–DNA SSBs. While a similar multi-protein complex was found in extracts from SCAN1 cells, the disease-causing TDP1 mutations render the phosphodiesterase incapable of excising 30 -tyrosyl blocking groups from DNA [3,8]. It was postulated that the accumulation of obstructive 30 -terminal SSBs would impair transcriptional efficacy and induce cell death. Ahel and colleagues [4] have now expanded the concept that defects in SSB repair can lead selectively to neuronal cell death. The gene defective in AOA1 encodes the protein aprataxin, which contains three conserved domains: a forkhead-associated interaction module; a histidine triad (HIT) domain found in nucleotide hydrolases and transferases; and a DNA-binding C2H2 zinc-finger motif [9,10]. While prior investigations had revealed interactions of aprataxin with proteins involved in SSB repair, namely XRCC1 and poly(ADP-ribose) polymerase 1 (PARP1) [12,13], a biochemical activity for aprataxin in this repair response had not been identified. Ahel and colleagues [4], recognizing that the aprataxin-binding proteins XRCC1 and XRCC4 are stable interaction partners of DNA ligases [13], examined whether aprataxin might release 50 -AMP intermediates that remain after failed ligation reactions. Indeed, recombinant aprataxin was found to excise AMP residues linked to the 50 -terminal phosphate group of synthetic DNA substrates, a step necessary for subsequent repair events in vitro. In addition, vertebrate cell extracts lacking aprataxin, including those from AOA1 lymphoblastoid cells, were defective in the removal of DNA-adenylates formed by abortive ligation reactions, which occur more frequently at non-conventional oxidative SSBs. The authors propose that the inability of aprataxin to act as an

AMP-hydrolase results in the accumulation of DNA strand breaks that ultimately impair normal cellular physiology and lead to neuronal cell death. Notably, the disease-causing mutations in APTX are truncating and missense in nature, and are largely confined to the HIT domain [9,10]. The current data leave two key issues unresolved. First, why do AOA1 and SCAN1 DNA repair defects lead to neurological dysfunction, yet not cancer susceptibility? One possibility is that non-dividing cells accumulate SSBs, which ultimately lead to impaired transcriptional programming and eventual cell death (Figure 1). Indeed, SSBs are effective blocks to RNA polymerase progression [14]. Furthermore, due to their unusually high rates of oxygen metabolism, neuronal cells probably accumulate more trapped 30 -TOPO1 and 50 -AMP intermediates due to the production of oxidative DNA damage. Conversely, in dividing cells, unrepaired SSBs would give rise upon replication fork collapse to DNA double-strand breaks, which would presumably be efficiently and accurately repaired by homologous recombination (Figure 1). Thus, while a broad range of DNA intermediates can drive neuronal cell loss and neurodegenerative disease, the SSBs of AOA1 and SCAN1 are apparently not particularly mutagenic in replicating tissue, unlike the double-strand breaks associated with AT and Nijmegen breakage syndrome, or the bulky, helix-distorting lesions found in xeroderma pigmentosum, or the unresolved complex DNA structures in Fanconi anemia, Werner syndrome, Bloom syndrome and Rothmund Thomson syndrome [2]. Future studies need to address specifically whether SSBs in fact accumulate in the neuronal cells of AOA1 and SCAN1 patients. Interestingly, individuals suffering from the disorder Cockayne syndrome are also not cancer-prone, yet exhibit gross developmental abnormalities and neurological deficits [15]. A defect in facilitating transcription or

repairing endogenous DNA damage, or both, may be responsible for the selective loss of neuronal tissue in this disease. Second, is it a general feature that defects in SSB repair lead to enhanced neuronal cell death? Mice devoid of the major gap-filling DNA polymerase, Polb, exhibit defective neurogenesis characterized by apoptotic cell death in the developing central and peripheral nervous systems that ultimately leads to neonatal lethality [16]. Model systems impaired in other SSB repair steps could also be examined to interrogate this hypothesis further. For instance, XRCC1 is a non-enzymatic scaffold protein that facilitates efficient BER/SSB repair [7], so deficiencies in this protein should lead to increased neuronal cell death, and accompanying neurological dysfunction. Defects in PNKP, which as noted above operates in the same pathway as TDP1, should likewise result in increased neurological deficits. Suggestive of a role for these proteins in maintaining neuronal cell viability is the fact that camptothecin-induced strand breaks accumulate in XRCC1-deficient CHO cells and in PNKP-depleted human A549 cells at levels similar to those detected in SCAN1 cells [3]. Whether or not deficits in these repair proteins will strictly affect neurodegeneration, and not cancer susceptibility, awaits investigation. In this regard, it is noteworthy that defects in the strand break response protein, PARP1, lead to both neurological abnormalities and cancer predisposition in mouse models [17]. Finally, there is emerging evidence that defects in the core BER participants, i.e. some of the DNA glycosylases, which remove endogenous base modifications, and the abasic (AP) endonuclease APE1, which initiates excision of AP sites in DNA as well as certain 30 -damages, contribute to neuronal cell survival, at least in culture [18]. We close by mentioning a few considerations relevant to the discussion above. First, DNA repair systems differ qualitatively and quantitatively between dividing and non-dividing cells [19]. In particular, global DNA repair

Current Biology Vol 17 No 2 R58

machinery is downregulated in differentiated neurons, whereas the machinery for repairing transcribed sequences is maintained or upregulated. Second, perturbations in DNA repair more subtle than those caused by genetic mutations may contribute to the demise of neurons in age-related disorders such as Alzheimer’s disease [20]. It will be of particular importance now to determine the influence of both dramatic and subtle variation in the different DNA-damage responses, particularly the SSB repair processing enzymes (Figure 1), as well as environmental factors, such as diet and lifestyle, on the susceptibility of neurons during aging. References 1. Forloni, G., Terreni, L., Bertani, I., Fogliarino, S., Invernizzi, R., Assini, A., Ribizzi, G., Negro, A., Calabrese, E., Volonte, M.A., et al. (2002). Protein misfolding in Alzheimer’s and Parkinson’s disease: genetics and molecular mechanisms. Neurobiol. Aging 23, 957–976. 2. Rolig, R.L., and McKinnon, P.J. (2000). Linking DNA damage and neurodegeneration. Trends Neurosci. 23, 417–424. 3. El-Khamisy, S.F., Saifi, G.M., Weinfeld, M., Johansson, F., Helleday, T., Lupski, J.R., and Caldecott, K.W. (2005). Defective DNA single-strand break repair in spinocerebellar ataxia with axonal neuropathy-1. Nature 434, 108–113. 4. Ahel, I., Rass, U., El-Khamisy, S.F., Katyal, S., Clements, P.M.,

5.

6.

7.

8.

9.

10.

11.

12.

13.

McKinnon, P.J., Caldecott, K.W., and West, S.C. (2006). The neurodegenerative disease protein aprataxin resolves abortive DNA ligation intermediates. Nature 443, 713–716. Evans, M.D., Dizdaroglu, M., and Cooke, M.S. (2004). Oxidative DNA damage and disease: induction, repair and significance. Mutat. Res. 567, 1–61. Connelly, J.C., and Leach, D.R. (2004). Repair of DNA covalently linked to protein. Mol. Cell 13, 307–316. Thompson, L.H., and West, M.G. (2000). XRCC1 keeps DNA from getting stranded. Mutat. Res. 459, 1–18. Takashima, H., Boerkoel, C.F., John, J., Saifi, G.M., Salih, M.A., Armstrong, D., Mao, Y., Quiocho, F.A., Roa, B.B., Nakagawa, M., et al. (2002). Mutation of TDP1, encoding a topoisomerase I-dependent DNA damage repair enzyme, in spinocerebellar ataxia with axonal neuropathy. Nat. Genet. 32, 267–272. Moreira, M.C., Barbot, C., Tachi, N., Kozuka, N., Uchida, E., Gibson, T., Mendonca, P., Costa, M., Barros, J., Yanagisawa, T., et al. (2001). The gene mutated in ataxia-ocular apraxia 1 encodes the new HIT/Zn-finger protein aprataxin. Nat. Genet. 29, 189–193. Date, H., Onodera, O., Tanaka, H., Iwabuchi, K., Uekawa, K., Igarashi, S., Koike, R., Hiroi, T., Yuasa, T., Awaya, Y., et al. (2001). Early-onset ataxia with ocular motor apraxia and hypoalbuminemia is caused by mutations in a new HIT superfamily gene. Nat. Genet. 29, 184–188. Pouliot, J.J., Yao, K.C., Robertson, C.A., and Nash, H.A. (1999). Yeast gene for a Tyr-DNA phosphodiesterase that repairs topoisomerase I complexes. Science 286, 552–555. Gueven, N., Becherel, O.J., Kijas, A.W., Chen, P., Howe, O., Rudolph, J.H., Gatti, R., Date, H., Onodera, O., TaucherScholz, G., et al. (2004). Aprataxin, a novel protein that protects against genotoxic stress. Hum. Mol. Genet. 13, 1081–1093. Clements, P.M., Breslin, C., Deeks, E.D., Byrd, P.J., Ju, L., Bieganowski, P.,

Decision Making: Don’t Risk a Delay Decisions under risk and choices between delayed outcomes are usually treated as two separate problems. A new study suggests that these two classes of decision making are more related than previously thought, and that delay discounting may tune an animal’s attitude towards risky choices. Tobias Kalenscher Delay not; swift the flight of fortune’s greatest favours Seneca What you risk reveals what you value Jeanette Winterson

Have you ever lost money playing the lottery? And you still

haven’t arranged your private retirement provision? If you said yes to the first question, you are in good company, as approximately 40% of my German countrymen occasionally try their luck with the lottery, and even 11% do this on a regular basis [1]. If you answered the latter question with yes, too, you are likewise not alone: in 2001, only one third of all adult German

14.

15.

16.

17.

18.

19.

20.

Brenner, C., Moreira, M.C., Taylor, A.M., and Caldecott, K.W. (2004). The ataxiaoculomotor apraxia 1 gene product has a role distinct from ATM and interacts with the DNA strand break repair proteins XRCC1 and XRCC4. DNA Repair (Amst) 3, 1493–1502. Kathe, S.D., Shen, G.P., and Wallace, S.S. (2004). Single-stranded breaks in DNA but not oxidative DNA base damages block transcriptional elongation by RNA polymerase II in HeLa cell nuclear extracts. J. Biol. Chem. 279, 18511–18520. Cleaver, J.E. (2005). Cancer in xeroderma pigmentosum and related disorders of DNA repair. Nat. Rev. Cancer 5, 564–573. Sugo, N., Aratani, Y., Nagashima, Y., Kubota, Y., and Koyama, H. (2000). Neonatal lethality with abnormal neurogenesis in mice deficient in DNA polymerase beta. EMBO J. 19, 1397–1404. Koh, D.W., Dawson, T.M., and Dawson, V.L. (2005). Poly(ADP-ribosyl)ation regulation of life and death in the nervous system. Cell Mol. Life Sci. 62, 760–768. Wilson, D.M., III, and McNeill, D.R. (2006). Base excision repair and the central nervous system. Neuroscience, Epub ahead of print. Nouspikel, T., and Hanawalt, P.C. (2002). DNA repair in terminally differentiated cells. DNA Repair (Amst) 1, 59–75. Davydov, V., Hansen, L.A., and Shackelford, D.A. (2003). Is DNA repair compromised in Alzheimer’s disease? Neurobiol. Aging 24, 953–968.

1

Laboratory of Molecular Gerontology, and 2Laboratory of Neurosciences, National Institute on Aging Intramural Research Program, Baltimore, Maryland, USA. *E-mail: [email protected] DOI: 10.1016/j.cub.2006.12.012

citizens had a voluntarily provided private retirement provision [2]. When playing lotto, you show a certain risk-proneness, as you prefer to invest money into a gamble whose actual outcome is uncertain, instead of using that money to buy a commodity that you could obtain with certainty. On the other hand, when hesitating to contract a retirement plan, you prefer using your budget to afford things that you fancy today, instead of investing it to obtain benefits that are yet to come. These scenarios exemplify two classes of decision making that are extensively discussed in the choice literature: decisions under risk, and inter-temporal decisions. The first class, decisions under risk, involve choosing between an option with