|
Are Neurons Just Too Laissez-Faire about Repair?
Thierry Nouspikel
|
|
|
|
Thierry Nouspikel led this live discussion on 14 July 2003. Readers are invited to submit additional comments by using our Comments form at the bottom of the page.
|
 View Transcript of Live Discussion — Posted 26 August 2003
View Comments By:
Ruth Itzhaki — Posted 26 August 2003
Are Neurons Just Too Laissez-Faire about Repair? Here's a Provocative Idea for the Alzheimer's Field to (Dis)Prove: Neurons Choke on Cell Cycle after Years of Letting Their DNA Fall into Disrepair
Philip Hanawalt of Stanford University has built a distinguished career studying the mechanisms of DNA repair, the medical relevance of which until recently was recognized mainly in cancer and related disorders such as Xeroderma pigmentosum and Cockayne syndrome. In recent years, however, Hanawalt and his research associate Thierry Nouspikel have turned their attention to DNA repair in differentiated neurons. This February, they proposed that postmitotic neurons may be setting themselves up for disaster by allowing DNA damage to accumulate unchecked in large swaths of their genome during a person's adult life. Melding the emerging literature on cell-cycle reentry in AD with their own recent work on DNA repair, Nouspikel and Hanawalt write that such neurons would be unable to pull off an orderly round of DNA replication. Indeed, fatal problems might arise even when RNA polymerase enzymes attempt to transcribe long-dormant cell cycle genes that have become littered with lesions (Nouspikel and Hanawalt, 2003). The authors invite researchers in the field of Alzheimer's disease and related disorders to put this hypothesis to the test.
Below is a synopsis of the essay, followed by a Q&A with the authors.
Cutting Corners on Repair…
Hanawalt and postdoc Thierry Nouspikel start out by stating that the mechanism of neuronal degeneration in AD, as in other neurodegenerative diseases, remains poorly understood. They then describe their prior in-vitro research with human NT2 cells, which showed that these cells strikingly curtail global DNA repair upon differentiation. Mature NT2 cells allowed DNA lesions to accumulate in all regions except those genes that were being transcribed (Nouspikel and Hanawalt, 2000). Other work also indicates that differentiated neurons do not efficiently repair the bulk of their genome (see, for example, Gobbel et al., 1998).
Neurons have different repair systems for different types of lesion, but current data is insufficient to generalize across these systems. Hanawalt and Nouspikel studied mostly nucleotide excision repair (NER), a versatile system involving about 30 proteins that recognize lesions, cut out a segment around it, and repair it by reading off the intact complementary strand. Reduced by 90 percent in differentiated NT2 cells, NER probably is unable to maintain an intact genome in those cells as damage accrues at the same rate as when NER was fully active, the authors write.
To prevent the crippling of needed genes, human neurons do repair their transcribed genes. For that, they probably use a mechanism called transcription-coupled repair (TCR), discovered in Hanawalt's laboratory in the mid-80's, which is thought to use RNA polymerase II as a sensor to target NER enzymes to active genes. The differentiated neurons also repair the nontranscribed strand of active genes by a still poorly understood mechanism the authors called differentiation-associated repair (DAR). Given that full-fledged genomic DNA repair consumes a lot of energy, this sparing use of repair might allow the neuron to coast through adult life quite well as long as it need not radically change its gene expression pattern or attempt to replicate its DNA.
…Costs Neurons Dearly Later On
The authors then review the growing body of evidence suggesting that neurons that have attempted to do just that are the ones which are dying in Alzheimer's (see Live Discussion). In the authors' view, the aberrant expression of different cell cycle markers in AD recorded by many laboratories—while highly suspicious—could be considered of little consequence in a postmitotic neuron until two years ago, when researchers in Karl Herrup's lab took this work a step further and showed that some neurons in AD patients do, in fact, replicate their DNA (Yang et al., 2001).
It's not just AD, either. Other labs have noted the reexpression of proliferation-related genes in a host of neurodegenerative diseases, including Pick's, Lewy body and Parkinson's diseases, supranuclear palsy, Down's syndrome, and FTDP 17. On this note, a current follow-up paper from Herrup's lab also found that all cases with mild cognitive impairment that were tested had the same percentage of neurons with cell cycle markers, even though about 30 percent of MCI cases go on to develop types of dementia other than AD. This apparently corroborates the notion that cell-cycle activation is a common theme in neurodegeneration (see ARF related news story).
What Goes Wrong?
Clearly, differentiated neurons can't resume proliferation successfully, Nouspikel and Hanawalt write. They die rather than divide or become tumorous (see, for example, Feddersen et al., 1992). Why is that? Completing the cell cycle would require a large-scale change in gene expression, rekindling expression of dozens of long-dormant genes. The authors propose that the lesions accumulated over many years in these silent genes trigger cell death upon cell-cycle reentry. They suggest these potential mechanisms:
1. RNA polymerase II may stall on DNA lesions; this is a strong signal for p53-mediated apoptosis (Ljungman et al, 1999).
2. The enzyme complex might be able to bypass other lesions but will, in the process, make errors in its product mRNA, which would lead to faulty proteins that impair the function of the cell.
3. Transcription-coupled repair will attempt to restore the transcribed DNA strand, but it will use the nontranscribed strand to do so. This strand has accumulated lesions at the same rate as the transcribed strand. This would likely cripple genes beyond repair, they write.
4. Genomic survey mechanisms that monitor the presence of DNA lesions—which must be suppressed while differentiated cells allow DNA damage to build—could well become reactivated when the cell cycle resumes. Quite possibly, a checkpoint mechanism holds these cells in G2 phase for a while, waiting for repair, which never happens.
What to Study?
Nouspikel and Hanawalt write that they consider both the negligence of genomic DNA repair in neurons and the reactivation of the cell cycle in neurodegenerative diseases well-established. However, the causal relationship between these two is speculative. Does lax repair lead to accumulating lesions, and do these neurons die while trying to transcribe or replicate this DNA?
Toward definitive testing of this idea, the authors offer some suggestions while inviting the field at large to come up with yet other ways to tackle the problem. First, though, a caveat. The authors performed their experiments with particular DNA-damaging agents including UV light and certain chemicals that are certainly not the physiological culprits at work in neurodegeneration. The most likely type of damage in neurons is oxidative and is largely repaired by base-excision repair, a different enzyme system whose attenuation has not yet been shown in mature neurons. However, a small subset of oxidative DNA lesions is repaired by the same pathway that deals with UV damage.
Even so, one experimental approach would be to challenge an animal model prone to neurodegeneration with low doses of a carefully selected DNA-damaging agent, and to document the accumulation of lesions in differentiated vs. dividing cells. This could then be correlated with neurodegeneration, perhaps in AbPP/PS transgenic mice (see transgenic mouse directory), SOD1 mouse models of ALS, or SCA models of polyglutamine repeat diseases. In addition, someone could devise a cell-based system to induce terminally differentiated cells to cycle again. Treating these cells with DNA-damaging agents beforehand would allow a look at whether transcription of such damaged DNA leads to cell death and whether attempted TCR would create mutations.
Whether this hypothesis proves right or wrong, rigorous testing of it will open up new perspectives on exactly how DNA damage contributes to these pathologies, the authors conclude.—Gabrielle Strobel.
Q&A with Philip Hanawalt and Thierry Nouspikel
Q:What piqued your scientific interest in Alzheimer's disease?
A:Thierry decided to study DNA repair in terminally differentiated cells and picked neurons as the epitome of such cells. While presenting that work at the 33rd Winter Brain Research Conference in Breckenridge, Colorado, Thierry heard Mark Smith talk about neurons reentering the cell cycle in AD. Thierry thought: Maybe I know why these neurons are dying; they are trying to use DNA crippled with lesions due to lack of repair.
Q: Is your lab actively pursuing AD-related angles of DNA repair?
A:Not specifically, at the present time.
Q: How could one test your hypothesis in vivo, or even in humans?
A:One could verify if DNA damage indeed accumulates in the aging brain, for example, from autopsy samples. In-vivo studies-these are not feasible in humans, but could be approximated in AD mice, for instance.
Q: Most AD mice mostly model amyloidosis or tauopathy. They have plaques, or neurofibrillary pathology, synaptic impairment, and a behavioral phenotype such as poor performance in the water maze, but they don't generally have massive neurodegeneration. Do you still consider them useful models for testing your hypothesis?
A:Any change in the spectrum of expressed genes might show an effect on cell function if damage had accumulated in those genes while previously dormant. Of course, the time scale is much shorter for these experiments with rodents—perhaps the damage accumulation in humans takes years, not months.
Q: There are several broad categories of DNA repair. Can you give us a quick primer on which ones are at play here? Are they largely identical between neurons and, say, fibroblasts?
A:DNA is subject to a lot of damage, from its own instability (spontaneous depurination, and deamination of cytosines), from products of the cellular metabolism (methylation by S-adenosyl-methionine, oxidation by products of the oxygen metabolism), and from environmental insults, both physical (UV, ionizing radiations) and chemical (food carcinogens, cigarette smoke, pollution, anticancer drugs). All in all, it is estimated that each cell has to deal with thousands of DNA lesions per day.
We have evolved a number of DNA repair mechanisms to deal with these:
- Direct damage reversal. Rare in human cells. An example is the suicide enzyme MGMT, which can remove an illicit methyl group from a guanine and transfer it onto itself.
- Mismatch repair. Repairs single nucleotide mismatches and small insertion loops.
- Base-excision repair. Repairs damage to a single base (oxidation, methylation, etc.). A collection of specific enzymes called glycosylases each recognize a given lesion, or a small subset of lesions, and detaches the base from the deoxyribose. An AP-endonuclease can then recognize these abasic sites, remove the sugar, and a DNA polymerase will fill the gap.
- Nucleotide excision repair. A more versatile system that repairs a wide array of lesions, probably because it recognizes the change in DNA geometry rather than the lesion itself. It then excises a chunk of about 30 nucleotides spanning the lesion. Essentially the same set of enzymes (about 30 polypeptides if you count all the subunits) takes care of all types of lesions.
- Strand break repair. Works either by homologous recombination, or by nonhomologous end-joining (the V(D)J recombination system used for the generation of antibodies and T cell receptors).
- Daughter-strand gap repair. After replication, it uses the sister chromatid to repair gaps left by DNA polymerase(s) opposite noncoding lesions.
Of these, base-excision repair and nucleotide excision repair can be coupled to transcription in that the transcribed strand of active genes is repaired faster than the rest of the genome, including the nontranscribed strand. This is probably due to RNA polymerase encountering a blocking lesion, and calling for help. The basic processes of base-excision repair and nucleotide excision repair operate in essentially all cell types, but with different efficacy. See Lindahl and Wood, 1999 for an excellent review on the matter.
Q: Can you venture a guess on what might prompt neurons in AD to reenter the cell cycle? Knowing that would help in the design of experiments to test your hypothesis.
A:Not really. But the same issue of BioEssays contains a hypothesis by Lu et al.) discussing the role of Pin1-mediated prolyl isomerization in AD, which might provide an answer to your question.
Q: How well do differentiated neurons repair oxidative damage to DNA?
A:That we don't know yet, but we'd love to. Unfortunately, it's technically difficult to measure low amounts of oxidative damage, especially in a gene-specific (and strand-specific) manner. In order to see oxidative damage, Thierry had to treat neurons with so much peroxide that it killed the cells, so he wouldn't see any repair anyway…. But that's a purely technical problem, so there is hope we can solve it as technology evolves.
Q: Oxidative stress has been implicated as a culprit in many neurodegenerative diseases, though there is no agreement on how early it comes into play, how specific it is to Alzheimer's, and what its primary mechanism of action is. Where do reactive oxygen species fit into your hypothesis?
A:Not yet in a specific way, except as noted above that some ROS-induced damage is subject to nucleotide excision repair, for example, cyclopurines.
Q: You write that nucleotide excision repair, whose deficiency you have studied in differentiated neurons, is less important in repairing oxidative stress than base-excision repair. Should its efficiency be studied in differentiated neurons, and are you doing so?
A:Yes, we would love to. But, as mentioned above, there are technical problems that have prevented us from doing it so far.
Q: Is this supposedly more relevant repair mechanism more difficult to study than NER? Some work on 8-oxoguanine glycosylase in AD exists in the literature; see, for example, Lovell et al., 2000.
A:There are several methods to detect 8-oxo-guanine (electrochemical cell, antibodies, glycosylases), but it's difficult to adapt them to a gene-specific, strand-specific assay. So far, we had no luck with either of these methods.
Q: Folic acid deficiency has been mentioned as impairing DNA repair in hippocampal neurons; see Kruman et al., 2002. Any thoughts on that?
A:Not yet.
Q: A majority of Alzheimer's researchers consider the amyloid hypothesis as the best framework to explain AD pathogenesis, though the evidence for late-onset, sporadic AD is weaker than that for familial AD. Where would the DNA repair/cell cycle hypothesis fit in with the amyloid hypothesis? Or is it an entirely separate explanation?
A:Not sure. But the two hypotheses are not necessarily mutually exclusive. Amyloid might account for the differences between AD and other neurodegenerative diseases in which neurons also resume the cell cycle before they die.
Q: Could a buildup of Ab simply worsen an existing problem by causing additional DNA damage?
A:Perhaps.
Q: Any links between DNA repair and neurofibrillary tangle pathology that you'd like to comment on?
A:We'll pass on that one.
Q: What kind of evidence would disprove your hypothesis?
A:Suppose we determine how neurons downregulate global repair (which is what Thierry is working on right now). Suppose we can disable this mechanism, in a transgenic mouse, for instance, that could be crossed with an AD mouse. If such mice still displayed AD-like pathology and symptoms, it would be a severe blow to our hypothesis.
Q: If you had $10 million, what sort of study would you fund to obtain definitive proof or refutation?
A:Can't describe that in a few sentences, but we'll take it!
Q: I realize you may be speaking largely from the perspective of an outside observer. As such, what appears to you to be the most vexing unresolved question in AD research today?
A:What's the initial event that triggers the cascade of events leading to the disease? What causes neurons to enter the cell cycle? Why are AD symptoms different from other neurodegenerative diseases where neurons also reenter the cell cycle before dying?
Additional Reading:
Lindahl T, Wood RD. Quality control by DNA repair. Science 1999 Dec 3;286(5446):1897-905 Abstract
Takashima H, Boerkoel CF, John J, Saifi GM, Salih MA, Armstrong D, Mao Y, Quiocho FA, Roa BB, Nakagawa M, Stockton DW, Lupski JR. Mutation of TDP1, encoding a topoisomerase I-dependent DNA damage repair enzyme, in spinocerebellar ataxia with axonal neuropathy. Nat Genet 2002 Oct;32(2):267-72. Abstract
Nagano I, Murakami T, Manabe Y, Abe K. Early decrease of survival factors and DNA repair enzyme in spinal motor neurons of presymptomatic transgenic mice that express a mutant SOD1 gene. Life Sci. 2002 Dec 20 ; 72(4-5):541-8.
Abstract
Caldecott KW. DNA single-strand break repair and spinocerebellar ataxia. Cell. 2003 Jan 10;112(1):7-10
Abstract
Ho PI, Ortiz D, Rogers E, Shea TB. Multiple aspects of homocysteine neurotoxicity: Glutamate excitotoxicity, kinase hyperactivation and DNA damage. J Neurosci Res. 2002 Dec 1 ; 70(5):694-702
Abstract
Lee DH, O'Connor TR, Pfeifer GP. Oxidative DNA damage induced by copper and hydrogen peroxide promotes CG-->TT tandem mutations at methylated CpG dinucleotides in nucleotide excision repair-deficient cells. Nucleic Acids Res. 2002 Aug 15;30(16):3566-73
Abstract
Ren K, de Ortiz SP. Non-homologous DNA end joining in the mature rat brain. J Neurochem. 2002 Mar ; 80(6):949-59.
Abstract
Kruman II, Kumaravel TS, Lohani A, Pedersen WA, Cutler RG, Kruman Y, Haughey N, Lee J, Evans M, Mattson MP. Folic acid deficiency and homocysteine impair DNA repair in hippocampal neurons and sensitize them to amyloid toxicity in experimental models of Alzheimer's disease. J Neurosci. 2002 Mar 1;22(5):1752-62.
Abstract
Shaikh AY, Martin LJ. DNA base-excision repair enzyme apurinic/apyrimidinic endonuclease/redox factor-1 is increased and competent in the brain and spinal cord of individuals with amyotrophic lateral sclerosis. Neuromolecular Med. 2002 ;2(1):47-60.
Abstract
Santiard-Baron D, Lacoste A, Ellouk-Achard S, Soulié C, Nicole A, Sarasin A, Ceballos-Picot I. The amyloid peptide induces early genotoxic damage in human preneuron NT2. Mutat Res. 2001 Aug 8 ; 479(1-2):113-20.
Abstract
Culmsee C, Bondada S, Mattson MP. Hippocampal neurons of mice deficient in DNA-dependent protein kinase exhibit increased vulnerability to DNA damage, oxidative stress and excitotoxicity. Brain Res Mol Brain Res. 2001 Mar 5;87(2):257-62.
Abstract
Duker NJ, Sperling J, Soprano KJ, Druin DP, Davis A, Ashworth R. beta-Amyloid protein induces the formation of purine dimers in cellular DNA. J Cell Biochem. 2001 ;81(3):393-400.
Abstract
Cardozo-Pelaez F, Brooks PJ, Stedeford T, Song S, Sanchez-Ramos J. DNA damage, repair, and antioxidant systems in brain regions: a correlative study. Free Radic Biol Med. 2000 Mar 1;28(5):779-85.
Abstract
Lovell MA, Xie C, Markesbery WR. Decreased base excision repair and increased helicase activity in Alzheimer's disease brain. Brain Res. 2000 Feb 7 ; 855(1):116-23.
Abstract
Schmitz C, Materne S, Korr H. Cell-Type-Specific Differences in Age-Related Changes of DNA Repair in the Mouse Brain - Molecular Basis for a New Approach to Understand the Selective Neuronal Vulnerability in Alzheimer's Disease. J Alzheimers Dis. 1999 Dec ; 1(6):387-407
Abstract
Hermon M, Cairns N, Egly JM, Fery A, Labudova O, Lubec G. Expression of DNA excision-repair-cross-complementing proteins p80 and p89 in brain of patients with Down Syndrome and Alzheimer's disease. Neurosci Lett. 1998 Jul 17;251(1):45-8.
Abstract
|
Comments on Live Discussion |
|
|
| |
Comment by: Ruth Itzhaki
|
|
|
Submitted 24 July 2003
| Permalink
|
Posted 26 August 2003
|
|
|
I was sorry not to be able to participate in the forum on DNA repair
on
Monday, but I wonder if anybody was aware that we worked on that topic
some years ago (published between about 1987 and 1993). It was on
gamma-irradiation of lymphocytes, looking at UDS, extent of
replication
after stimulation, single-strand and double-strand breaks, and
chromosome aberrations. We found a small but statistically
significant
difference in the latter (dicentrics) between ADs and age-matched
normals. Our work is summarised in my review in Molecular
Neurobiology,
1994, 9, 1-13. I would gladly send reprints of that or of the
individual papers (Mutation Res., J. Med. Genetics, Int. J Rad. Biol.,
etc), if anybody is interested; I do not have any in pdf form though.
References: Smith TA, Neary D, Itzhaki RF. DNA repair in lymphocytes from young and
old individuals and from patients with Alzheimer's disease.
Mutat Res. 1987 Sep;184(2):107-12.
Abstract
Tobi SE, Moquet JE, Edwards AA, Lloyd DC, Itzhaki RF. Gamma-radiation
induced chromosome aberrations in Alzheimer lymphocytes. Biochem Soc
Trans. 1990 Jun;18(3):393-4. No abstract available.
Abstract
Smith TA, Itzhaki RF. Repair of DNA single-strand breaks in lymphocytes
from Alzheimer's disease patients. Gerontology. 1991;37(4):193-8.
PMID: 1916309
Abstract
Tobi SE, Itzhaki RF DNA double-strand breaks measured by pulsed-field
gel electrophoresis in irradiated lymphocytes from normal humans and
those with Alzheimer's disease. Int J Radiat Biol. 1993
May;63(5):617-22.
Abstract
View all comments by Ruth Itzhaki |
|
|
|
|
|
|
Submit a Comment on this Live Discussion
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Home
Webinar: A Webinar is a seminar conducted remotely over the Web. Attendees view the slides through their Web browser and hear the presentations over their own telephones. 
