Posted 26 August 2003

E-mail discussion Printable version

Return to Discussion Text

Participants: Thierry Nouspikel, Stanford University, Oklahoma; Rachael Neve, McLean Hospital, Belmont, Massachusetts; Keith Crutcher, University of Cincinnati; Donna McPhie, McLean Hospital; Karl Herrup, Case Western Reserve University, Cleveland, Ohio; Sam Cicero, Case Western Reserve University; Randall D. York, Case Western Reserve University; Yan Yang, Case Western Reserve University; June Kinoshita, Alzheimer Research Forum; Greg Brewer, Southern Illinois University School of Medicine, Springfield; Tzuwei Wu, University of Southern California; Jung Ming, University of Southern California; Xiao Bing J., University of Southern California; Samantha Cicero, Case Western Reserve University; Michael Kim (YdeBong), University of Southern California; Fulya Caraman; Vickie.

Note: The transcript has been edited for clarity and accuracy.

Thierry Nouspikel
Let me put it this way: Introducing lesions into the DNA of dividing cells is a good way to trigger apoptosis.

Rachael Neve
What is the evidence for that?

Karl Herrup
The ATM mutant comes to mind. Behavior, but no cell loss. The XRCC4 and LigIV knockouts (recombination repair) are embryonic lethals. Lots of nerve cell death.

Thierry Nouspikel
As far as neurons are concerned, one of our hypotheses is that reentering the cell cycle will cause transcription of genes that were not transcribed before and are supposedly crippled with lesions. It is known that RNA PolII encountering a lot of lesions is a good trigger for apoptosis.

vickie
Did any animal models with any sort of DNA repair defects see cross-correlation with neural dysfunction?

Thierry Nouspikel
Vickie, I'm assuming transcription-coupled repair (TCR) is critical to maintain neuronal function. Patients with a specific deficit in TCR (Cockayne syndrome) generally die young due to neurological problems.

Rachael Neve
Good question.

Karl Herrup
Then why do the XRCC4 mutants have such massive cell death?

Rachael Neve
We've shown that expression of FAD mutants of APP in primary neuronal cultures causes DNA synthesis and consequent apoptosis. The primary cortical cultures, from E17 embryos, likely have not had time to accumulate DNA lesions. What are your thoughts about that?

Karl Herrup
I would propose that the accumulated DNA damage might play a role in explaining why AD doesn't start until later ages. A second hit maybe?

Rachael Neve
But Karl, you've shown that neurons can exist in the tetraploid state in AD brains for years before killing the cells. How does that jibe with Thierry's hypothesis?

Karl Herrup
Good question, Rachael. I think that there is a major difference between young neurons and old neurons. The critical period of target dependence seems a breakpoint for a nerve cell.

vickie
In adult neurons, as long as survival is concerned, is transcription-coupled DNA repair more reasonable if it is ever needed?

Thierry Nouspikel
Accumulated DNA damage can block the cell cycle through dedicated genome surveillance mechanisms. The question is, how long the cell will remain in this "suspended" state before it gives up and commits suicide.

vickie
Could you explain a little bit about target dependence?

Karl Herrup
Vickie, if a neuron doesn't receive trophic support from its target during development, it dies. This is believed to be a numerical matching function. We've shown that at least some cases of this involve the deprived neurons reentering a cell cycle.

Rachael Neve
I'm still wondering why, in our cultures, we can express FAD APPs, or overexpress APP-BP1, a cell cycle protein, and readily see DNA synthesis and consequent apoptosis. We've shown that if we prevent DNA synthesis, we don't prevent the apoptosis, but if we prevent the G to S transition, we do prevent it. That's Donna McPhie's work.

Karl Herrup
Thierry, where in the cycle does the arrest occur? And how?

Thierry Nouspikel
Karl, there may be several checkpoints. One mechanism is through the ATM/ATM Rad3-related (ATR) pathway.

Karl Herrup
Okay, but why is the neuropathology of the ATM(-/-) mouse so minor? Does ATR kick in? Does anyone know how big a role ATR plays in the CNS?

Thierry Nouspikel
Don't know. It wouldn't be the first time that mouse and man behaved differently.

Karl Herrup
I surely agree there.

Rachael Neve
Thierry, I noticed in your interview that you said you'd seen downregulation of global repair mechanisms in human fetal neurons as well as in NT2 cells. How long did you maintain the human fetal neurons before seeing this downregulation?

Thierry Nouspikel
Twenty-nine weeks. Given the age of the donor, that would correspond to 42-week-old neurons. The downregulation appeared somewhat earlier, though (16 weeks, if I recall correctly).

junming
What are the check points that drive cell division activity to apoptosis?

Rachael Neve
Junming, we've shown that if you block the G1 to S transition, you can block neuronal apoptosis in our model.

Karl Herrup
Junming, to Rachael's comments I'd add that I think my young vs. old concept applies. In the young there may be specific checkpoint-related apoptosis pathways. In the old, I think it's more a slow atrophy ending in an apoptotic crisis.

Randall D. York
To Rachael/Karl, it is also possible that attempts to replicate in the presence of DNA damage may signal death checkpoints in the majority of cells, but a subset escape these checkpoints and are able to progress to the tetraploid state and undergo G2 arrest, similar to what is seen in Drosophila-replication mutants. In this case, the tetraploid neurons that are surviving for prolonged periods would represent the minority.

Rachael Neve
So Karl, do you think that explains why we so readily get DNA synthesis and apoptosis in our model expressing FAD APPs in embryonic neurons?

Karl Herrup
Rachael, I think it's a big contributing factor. Plus, what age of cortex are you looking at?

Rachael Neve
Karl, we use E17 cortices and do our experiments on day five in vitro.

Randall D. York
Rachael, how is the G1 to S transition blocked in this case??

Rachael Neve
Donna, what did we use? Mimosine and what else?

Donna McPhie
Desferoxamine.

junming
Karl, for the stem cells, a lot of new cells go into apoptosis if they do not differentiate to a certain destination. Does the cell density play a role there, also?

Karl Herrup
Junming, do you think that's a DNA damage problem or a problem of apoptotic pathways?

junming
Karl, I think there may somehow be a link between cell cycle proteins and apoptosis.

vickie
Is that likely due to DNA repair mechanism or otherwise?

Rachael Neve
Thierry, is there a way to assay whether DNA from aged human brains, in fact, has accumulated global lesions relative to DNA from younger human brains?

Karl Herrup
Good question, Rachael! I've wondered about that myself. How do you tease out neurons from glia (and other stuff)?

Rachael Neve
Maybe you could do a microdissection from a part of the brain that is rich in neurons, like maybe from the hippocampal CA layer?

Thierry Nouspikel
There are ways. We just need to know what lesions we should look at. There are so many, repaired by various repair systems. I was using UV-induced lesions and benzopyrene, for reason of convenience (good test lesions for the nucleotide excision repair (NER) pathway). But I'm aware that these may not be the best physiological lesions, at least in the brain.

Rachael Neve
I was wondering if it would be useful to cross, say, an FAD APP + FAD PS transgenic with Stratagene's Big Blue mouse, and see if more mutations occur than if you crossed the Big Blue with a control mouse.

Karl Herrup
Rachael, while you're crossing, what about putting in a DNA repair glitch such as making the mice ligase IV heterozygotes?

Rachael Neve
Oooh, I like that.

Karl Herrup
I'll bet Peter McKinnon would be willing to part with some animals.

Thierry Nouspikel
Karl, why not put a glitch in the NER pathway, too, just for the fun of it?

Rachael Neve
You know, it really would be a fun experiment to do. Our transgenic mouse facility is a bit overcrowded at the moment, so I imagine that there wouldn't be a lot of rejoicing among the animal care staff, but it would be VERY interesting.

vickie
What are the insults used in these DNA damage models?

Thierry Nouspikel
Oxidative lesions may be more relevant for neurons, but most are repaired by another pathway (BER) and I don't know whether it's repressed in neurons. There are a few oxidative lesions repaired by NER, though (cyclopurines).

Randall D. York
Thierry, how difficult would it be to determine whether the upregulation in DNA repair enzymes and activity reported with acute oxidative stress and/or ischemia is associated with transcribed genes or global repair that may accompany attempts to replicate in neurons that have reentered a cell cycle??

Thierry Nouspikel
Randall, I'm not sure I understand your question. Aren't we talking about two different models here? There are several mice deficient in one or the other NER enzyme.

Randall D. York
Perhaps, but there appears a common attempt at cell cycle reentry in both models.

Greg Brewer
We can get two-year-old old rat neurons to divide many times. Does that mean they do it with mistakes or maybe fix them first?

Rachael Neve
That's interesting, Greg. Maybe because they continue to divide, they maintain their DNA in good condition.

Karl Herrup
Greg, do you know whether your neurons might be GABAergic interneurons? We find that they seem to get around all of the restrictions we've been talking about. Those neurons might even be where some of the adult neurospheres are coming from.

Rachael Neve
Good point, Karl, and that brings up the issue of why do we see regional specificity in AD neurodegeneration.

Thierry Nouspikel
Greg, can you get them to resume dividing after they go quiescent? Or do you have to keep them growing continuously?

vickie
Greg, how did you get your old neurons to divide? If it is not a secret.

junming
Greg, I have also noticed that at a certain density, a lot of them enter apoptosis.

Greg Brewer
Karl, I don't know if they're GABAergic. Do you think they keep dividing throughout life so that's why they can still do it?

vickie
Greg, are you using some kind of stem cell growth conditions?

Greg Brewer
Vickie, they divide readily at low density in Neurobasal/B27 + FGF2 (published in Brewer, 1999).

vickie
Thanks, Greg.

Karl Herrup
Greg, I think they retain the potential. The other thing I'm not sure about is whether there may be a true "stem" cell somewhere (in response to Vickie's query).

Greg Brewer
Karl, the percentages seem too high to be stem cells: 50 percent neurofilament + BrdU positive after just five days in culture.

Karl Herrup
Greg, point taken.

vera
What do you think about birth order's role in AD?

Rachael Neve
Vera, why did you bring up that question? Is there a higher AD risk for certain positions in the birth order? I was not aware of that.

junming
Are there some checkpoints that can change the direction of a cell from dividing to apoptosis pathway?

Rachael Neve
Junming, I think that's the $64,000 question.

junming
Ha ha!

Thierry Nouspikel
There was a hypothesis (by Meyn, 1995) that it's a matter of time. First, the cell cycle is stopped pending damage to be repaired. If this does not occur within a certain time frame (?), then apoptosis is triggered.

Rachael Neve
Thierry, at what point is the cell cycle paused?

Thierry Nouspikel
G1, I think.

Rachael Neve
That might explain why, if we block G1 to S, we can block apoptosis following DNA synthesis in neurons.

Thierry Nouspikel
At least it would make sense to allow time for repair before replication occurs. Although stops in G2 could permit repair by recombination-like events.

Greg Brewer
We have unpublished data that the neuron division restores the cytochrome oxidase levels of old neurons to that of young ones. So maybe a lot of repair is happening, not just DNA. We intend to measure mtDNA mutation levels.

Rachael Neve
Greg, that's fascinating! It will be interesting to see the results of your assays of mtDNA mutations levels.

Tzuwei Wu
Does anyone know if mtDNA can be methylated?

Karl Herrup
If we set aside the problems of chromosome mechanics, would we be wise or foolish to try to get the neurons to finish the job and get through G2 and M. The upside of this is that we'd have two cells instead of one. The first one was doing fine without repair. Would it hurt to have two?

Thierry Nouspikel
Probably not...assuming that replication went smoothly. How many mutations would be introduced by trying to replicate through DNA lesions?

Rachael Neve
Karl, did you have any ideas about how to get the neurons through G2 and M without them dying?

Karl Herrup
Thierry, good point. Rachael, no.

Rachael Neve
Darn. It would be cool to do. Well, considering the tours de force you've achieved so far, Karl, I wouldn't be surprised if you found a way to do it!

Karl Herrup
Rachael, I can't even type without dyslexia.

Randall D. York
Are you able to estimate the degree of damage in neuronal cells? It also seems unlikely that conventional replication could proceed without signaling checkpoints in the absence of a novel, error-prone replication mechanism.

Thierry Nouspikel
Randall, an idea is that signaling checkpoints were disabled in nondividing neurons. They may be slow to restart....

Randall D. York
I like that idea; is there any evidence for this?

Thierry Nouspikel
Randall, that's a question for the cell-cycle gurus. Are some critical cell cycle-related proteins downregulated in neurons? I'm sure somebody must have looked....

Rachael Neve
Thierry, how much DNA do you need to assess DNA damage in older human brains vs. younger? Can one do it with microdissections?

Thierry Nouspikel
Rachael, for global genome, about one microgram. For gene-specific repair, 50 micrograms per time point.

Rachael Neve
Hmm, so you don't need that much DNA. It seems that there are neuron-rich regions of the brain that one could microdissect to do the experiment. If I find a way to get the tissue and isolate the DNA, will you do the assay?

Thierry Nouspikel
Sure. But be aware that we can only look for a few specific lesions: UV-induced and some chemicals.

Karl Herrup
What about specific amplification processes? Could we envision maybe using cre-lox to create neuron-specific methods?

Rachael Neve
Thierry, we see a lot of upregulations of cell cycle proteins in AD; I'm not sure if anyone's seen downregulation of any of them. Anyone know?

Thierry Nouspikel
Not in AD. In normal, terminally differentiated neurons.

Karl Herrup
Rachael, as far as I know, everything is up. Cyclins, CDKs, CKIs. And they're everywhere.

Rachael Neve
Thierry, are you planning to continue this work of looking at DNA lesions and mechanisms by which they're produced in neurons?

Thierry Nouspikel
Yes, but for reasons of convenience I have switched to a different system (macrophages). Once I have dissected the mechanism in these cells, I'll go back and look if it's the same in neurons.

Rachael Neve
Neurons are so much prettier than macrophages!

Karl Herrup
Thierry, don't stay in the periphery too long. We need your talents in the CNS. And neurons ARE prettier.

Thierry Nouspikel
The reason I went to macrophages is that I can grow billions of precursor cells, then differentiate over 90 percent of them to macrophages. This allows me to do some hardcore biochemistry. Can't do that with neurons.

Rachael Neve
Thierry, we do some pretty hardcore biochemistry with our neuronal cultures. I don't know that we get billions, but we certainly get hundreds of millions.

Thierry Nouspikel
I was thinking of purifying an activity missing in terminally differentiated cells, to microsequence the protein. You need a lot of cell extract for that....

Rachael Neve
True.

Rachael Neve
How are you assaying for that activity, by the way, Thierry?

Thierry Nouspikel
Cis-platinum cross-links in an in-vitro assay. Tenfold differences between macrophages and replicating precursors.

Randall D. York
Can you tell us anything about the specialized polymerases/enzymes to bypass DNA damage before a replicative enzymes take over (pol n; hREV-1,) and their role in neuronal cells.

Thierry Nouspikel
There are several DNA polymerases that can sail through DNA lesions, with or without introducing mutations in the process. There is also a repair pathway known as "daughter-strand gap repair" that can fill the gaps left by a DNA polymerase puzzled by a DNA lesion in its template.

Rachael Neve
Thierry, are those specific pathways downregulated in neurons?

Thierry Nouspikel
I don't know.

Randall D. York
It may be interesting to consider these in differentiated neurons as well.... In the event that checkpoint pathways are not downregulated, there must be mechanisms in place to avoid apoptotic signals, particularly in the tetraploid AD neurons.

Karl Herrup
I wonder if there is something to be gained from looking at primary neurons vs. neuroblastomas. To be fair, it would have to be a PNS/neuroblastoma comparison. But clearly cell division kills the primary PNS neuron but leaves the neuroblastoma untouched. Prettier than macrophages, anyway.

Rachael Neve
Yes, it's always fun to work with pretty cells. It is something to look forward to when you come into the lab every day. I always head straight to the incubator to admire the neurons :-).

vickie
Mitochondria are so heterogenous and capable of rapidly reproducing to replace themselves, is mtDNA repair crucial for neuron survival or is mtDNA repair limited because mitochondria can rapidly replicate?

Rachael Neve
Vickie, that's a very, very good point. Thierry, do you know the answer to that?

Thierry Nouspikel
It's certainly important. Several people work on the topic. Susanne Ledoux jumps to mind, but there are others.

vickie
Thanks!

Greg Brewer
Glenn Wilson is another working on mtDNA repair. I suspect neurons will keep getting more plastic, the more we study them.

June Kinoshita
Hi all, as the editor of the Alzforum, I want to thank you all for participating in today's chat. We are at the end of our allotted time (although you are all welcome to continue using the chat room as long as you like). I was wondering if you would all like to submit some closing thoughts regarding the scientific opportunities offered by the convergence of Thierry's work and AD research. What next?

Rachael Neve
I like the idea of crossing transgenic mice using Big Blue, as Karl and I discussed earlier....

vickie
So, neurons are working hard on DNA repair even in adult, is that a consensus?

Karl Herrup
I think this has been a very good session. I agree with Rachael that there are some interesting crosses to do to compromise cells in different ways and then stress their neurons with cell cycle stressors. To keep Thierry engaged, we should compare these with macrophages (and other cells) from the same animals. Could be very good.

Rachael Neve
I also want to thank Thierry for presenting a most provocative hypothesis. I was quite excited when I read your hypothesis in BioEssays.

Thierry Nouspikel
Thanks, Rachael. Bye, everyone.

June Kinoshita
Special thanks to Thierry! Goodbye!