Dr. Speakman,

Very interesting post and questions. Not for the faint of heart! Thanks for pointing out your recently published articles. These are very interesting results. Do you know if telomerase activity is reduced in the liver compared to bone marrow, as in birds (Haussmann et al. 2007)? If so, it is interesting that your result of reduced oxidative damage in the liver seems to mirror this pattern.

My comments below are rambling. Be warned. I could simply be parroting the random literature I've read (including your own past work), but it seems to me that there would be considerable merit is examining multiple organ systems. Many have done this, obviously, but I'm unaware of studies that have done this from a life-history theoretical approach. Quantification of tissue-specific oxidative damage is a brilliant first step, as is quantifying the physiological mechanisms that are counteracting these negative effects (antioxidants, telomerase activity, etc.). No doubt that this dual quantification will create a preponderance of evidence for inferring the potential importance of some tissues relative to others. One consideration here is protective mechanisms are upregulated in response to tissue-specific damage, and I wonder how this would effect quantification of overall patterns. For instance, how might damage to the liver be associated with upregulation of telomerase or production of antioxidants in another tissue? That is, there could be regulatory trade-offs between organ systems in specific contexts. How these intrinsic trade-offs might scale to emergent life-history trade-offs might be interesting indeed. This is my long way of saying that I do think an tissue-specific approach will be necessary.

For understanding oxidative stress as one mediator of life-history trade-offs, I think it's also worthwhile to examine oxidative damage and the physiological mechanisms across contexts, which will obviously take a huge amount of clever work. Of course, because this kind of destructive sampling is mutually exclusive with longitudinal studies of individuals, a complete quantification of multiple tissues is impossible. In any case, I think doing this in one context (e.g. reproduction) is a critical beginning to understanding life-history trade-offs. But, if we're wanting to identify trade-offs and their shapes, I think we need to take a more individual-based approach across contexts. There's work coming out of long-term studies such as song sparrows showing the existence of multiple strategies for optimal reproduction (and thus tradeoffs between current and future reproduction/survival). Therefore, there's no doubt that these trade-offs can be obscured at the population level.

I'm only beginning work on telomeres in birds (using a long-term study of birds in Panama), so I'm not qualified to speak at length on this. But, I do think that damage to DNA is an important feature of intrinsic physiological aging. I think there's been overemphasis on the relationship between chronological age and telomere length and likewise relatively little emphasis on the variation within ages. Given the evidence for the Hayflick limit of chromosomal end-cap degradation, there's a seemingly direct link between DNA damage and residual reproductive value of individuals, a fundamental component of life-history theory.

Anyway, your question was multi-faceted, so my response was a bit rambling. I just wanted to spew forth for a minute or two in order to continue the discussion. You've raised some important issues.

Patrick

Thanks for your interesting post. We did not measure telomerase activity, but my understanding is that telomerase activity is much higher in rodents than in birds so impacts of oxidative stress on telomeres may be less significant in this group. Might be worth looking at in the future though.

I think you are correct that individuals may trade-off protection between different tissues. The problem I think is that at present we just don’t know what it means for long terms health, ageing and future reproductive potential if say tissue A gets more damage than tissue B. I’m not even sure how we might start to collect such information since the causes of death may not obviously link to the sites of damage. if I get lots of oxidative damage to the liver but die of pancreatic cancer what does that mean?

I agree with your idea that there may be multiple contexts where damage is important, but only one of these (ie reproduction) leads to future offspring, so that does seem a sensible starting point. Michael Garratt and I just reviewed all the literature on oxidative damage in reproduction (out next week or so online in Bioessays) and it is surprising that actually nobody seems to have yet measured damage to DNA as a consequence of reproductive effort. The focus of all the studies to date has been on lipid oxidation or damage to proteins – surprising given as you point out the seemingly obvious links. Good luck with your research.

John

Good points. You're right that the reproductive life-history stage is a good starting point. I'm not sure about the wording that "only one of these (i.e. reproduction) leads to future offspring" within the life-history trade-off context, however, since individuals must survive in order to reproduce. I just wanted to clarify this wording (which I think you already agree with), lest future readers get confused.

For the study of life-history trade-offs, we need to study factors influencing both survival and reproduction across contexts. Placing the weight of our efforts on future reproduction doesn't reflect the full range of variation in reproductive life-history trade-offs. That is, parents with higher reproductive value (a high chance of surviving and reproducing) may allocate more resources and presumably have physiological mechanisms for protection of chromosomes or tissues (to further increase survival). Allocation towards future reproduction (quantity-wise) is no doubt especially important for, say, temperate or arctic-zone passerine birds, but is no doubt relatively less important for tropical rainforest understory passerines. For instance, in a 23-gram tropical bird that has a maximum longevity of 17 years (low extrinsic mortality), it is no doubt difficult to detect the influence of the low annual reproductive output (e.g. 4-15% annual nesting success) on telomere shortening. Thus, it might be worthwhile to examine non-reproductive traits or other contexts as predictors of variation in telomere length--and oxidative "stress." For instance, over-winter or non-breeding season survival--and the kinds of physiological damage associated with this-- may directly influence future breeding output. Perhaps as a result of studying tropical birds for so long, I tend to place equal a priori weight on both survival and reproduction in terms of life-history trade-offs.

Anyway, though it sounds strange, I think we need to acknowledge that current reproductive effort is not the only life-history component that leads to future offspring. Survival is a key metric, and the context in which survival occurs will no doubt provide a nice foundation for understanding the physiological mechanisms of life-history trade-offs. Of course, in order to accomplish this, we must learn to separate the sources of both intrinsic and extrinsic of mortality. And this will require not only the continued study of our focal organisms' physiologies but also studies of the impacts of climate and predators on mortality risk.

That's it for now...I'm looking forward to your review in Bioessays!

Hi, I will be curious to see if the trend is the same with markers such as mtDna damages... Since mitochondria are the main producers of Ros, mtDna should be the first showing damages (especially attacks considering superoxide anions). However, the problem is that the majority of thesuperoxide anion is converted by the superoxide dismutase... did you ever try to measure this marker and do you know if someone try it in Sod deficent animals?

thanks for your thoughts. I agree that MtDNA does seem a likely target for superoxide radical damage. Maybe we have all been measuring the wrong thing by focussing on other macromolecular targets? A student from my lab has been to visit another lab to learn the methods for measuring MtDNA damage - but it is not an easy method to pick up. Regarding superoxide dismutase we measured total activity of this enzyme (ie not separated into the Zn-Cu and Mn forms) in both the gerbil and vole studies mentioned above and found its activity was inverse to the protein carbonylation damage. ie in the livers SOD was up and damage down, and in the serum SOD was down and damage up. This makes me think that the animals can selectively choose what organs to protect or sacrifice - and presumably they choose to do this in relation to how important they are. Hence demonstrations that damage in serum is up may not actually mean very much. I think homozygous SOD deficient mice show early lethality and hence unlikely to have been studied in reproduction, but unsure about the hets.

Dear John

here are my general thoughts about your questions. As you can see, I am not so optimistic about the possibility to answer your questions at the moment.

1. From my point of view, there is no evidence suggesting that damage to one tissue is more important than another one. In this regard, we should also specify the meaning of ‘important’…. Important for what? If investing into reproduction pays in evolutionary terms, it might be more important to protect gonads rather than muscle. But in other contexts (e.g., escape from predators, migration) it might be just the opposite. Finally, it is quite hard to say whether the opposite response of liver and blood to treatment is reflecting selective protection. This is because it is unknown the extent to which oxidative damage measured in a certain tissue is produced locally. For example, necrosis of cells causes release of oxidized molecules into the blood stream, but there are also sources of oxidative damage in the blood itself. So any increase in damage may come from different sources, which complicates interpretation.

2. I do not believe at all that there is a ‘golden’ assay. For me, the best approach is to use a combination of assays. It is here that selection of assays may be important. For example, avoiding assays that provide redundant information (e.g., that measure only various components of protein oxidative damage) might be a good approach. I also do not see any evidence out there showing that some damage is more important than others. Bear also in mind that most of the studies are based on humans or on a few laboratory strains. And, honestly, translating those results to the over one million of species free-living out there looks to me a dangerous game. Although many mechanisms related to redox machinery appear to be quite conserved across species, there is also a lot of variation that makes almost impossible to draw general patterns. Marine invertebrates, for example, provide fantastic examples of peculiar adaptations to control oxidative stress.

Cheers

David

David

Thanks for your insightful comments. In reply..

1) I guess by important' what I meant was having an effect on future survival or fecundity. If oxidative stress really is a mediator of life history trade-offs then it has to work via effects on the life history traits. Hence I could imagine that oxidative damage to some tissues might be less likely to generate a survival effect than damage to others. However, what I dont have is an intuitive feel for what tissues might be most significant in this respect. What bits of the soma are really disposable? Your point regarding 'damage' not necessarily being local is well made and certainly makes it more complex to understand. In that sense the serum may be a global integrator of whole body damage and thus a more significant marker than damage measured in a particular tissue.

2) I agree that it seems obvious to make measurements with multiple non-redundant assays. Yet very few studies have done so. One point from our papers (cited in the original question) where we used assays of both lipid and protein damage is that the answer you get depends on the assay you use. Some assays support the oxidative damage hypothesis, but other assays in exactly the same individuals do not. So studies based on single assays, be they supportive or otherwise, should be viewed with a great deal of caution. Unfortunately that includes just about everything done so far in this area (including several papers I have co-authored).

Finally, you are very correct to point out the potentially massive diversity out there - and to find this you dont need to compare only across major taxa, Within the mammals for example due to the great work by Shelly Buffenstein we already know that naked mole rats lack glutathione peroxidase and potentially as a direct consequence suffer massive oxidative damage. Yet paradoxically when scaled to body weight are the longest lived rodent!! Maybe that is evidence enough that this idea that oxidative damage may mediate a link to survival is considerably more complex than generally envisaged

best wishes

John

Hi all,

they are several interesting points that were raised. I would like to underline some of them. The first question is why do we want to know the exact localization of oxidative damage inside the body? Is it to determine the most appropriate type of sample that should be used? For sure, in the wild, there are few doubts that blood is probably the most appropriate since it is non lethal and, given that it correlates with fitness-related traits, its meaning cannot be really questioned. If the question is more ecophysiological (as certainly John is interested in), then looking at tissues is relevant. And I would suggest that comparing different species with high and low rates of reproduction would probably produce interesting data concerning antioxidant trade-offs among tissues. Other life history situations may be of interest such as comparing naturally long-fasting and non-fasting animals to determine which tissue target is better protected or if this is a general trait at the organism level. In rodents I was thinking that kidney is particulalrly sensible in terms of lifspan, is that right? Finally, for my point of view, mitochondrial damage is a key parameter to take into account and simple proxy such as aconitase/fumarase activities (working on frozen samples) may inform on the level of ROS attack on mitochondrial target. But does anyone known about a real experimental test of the importance of the level of chronic mtDNa damage on lifespan?

Francois

thanks for your interest and your post. I was a bit confused however by some of what you wrote. Can you elaborate on what you mean by this part of your message...

" blood is probably the most appropriate since it is non lethal and, given that it correlates with fitness-related traits, its meaning cannot be really questioned."

what 'fitness related traits' are you referring to? do you mean it has been shown that individuals with high levels of oxidative damage measured in their blood have reduced survival? I am not aware of any papers that show that, so if you know of some I'd be really interested if you can post the references. or maybe I misunderstood - if so perhaps you can expand what you meant.

best wishes

John

Dear John

I am not so convinced that oxidative damage to some tissues might be less likely to generate a survival effect than damage to others. So far I personally do not see any evidence supporting that this may be the case. If we just look at human studies, we see that everyone can die because of a disease/dysfunction to a certain tissue, such as brain, blood, lung and so on. I have also the feeling that it is quite hard in many cases to separate effects of tissues on survival. This is because it might be possible that initial disfunction in tissue X spreads out across other tissues. Honestly, I think that answering most of the questions of this blog in 'in vivo' systems is likely impossible

to do. And it is probably not that important for evolutionary ecologists. In this specific context, I think that it is more important to look at organism level rather than at parts of it. Integration rather than reductionism should be our target. Clearly, if we wish to elucidate mechanisms, other approaches are certainly needed.

Best wishes

David

Sorry I am a bit late answering.

John, I think that Pierre published in his paper 2008 (Ecology) a relationship in swift between resistance to oxidative stress (KRL test) and survival (male) or fecundity (females). We have some on going studies where we also find a relationship with survival (Adélie penguins) and some where we did not find any relation with reproduction effort or survival for example (finch or king penguin chicks). But this may be blur by protocole details such as blood sampling timing.

We need to add a more closely following of oxidative blood parameters during reproduction, for instance to characterize dynamics of damage/defense over the entire cycle (pairing, incubation, rearing, post-fledging). Also another thing that count for you, differences between reproductive and non-reproductive individuals (maybe artificially kept so). What about their age-related changes in oxidative stress?

Maybe one solution for you is to shift toward proteomic analysis. We have one on going study using proteome profils of individuals facing a costly reproduction. We may detect more detailed data on oxidative balance but I am not sure of it.

Dear John and Francois

I am aware of around 10 studies (including unpublished data from me) that look at links between oxidative status parameters and some proxy variables for survival (fledging success, return rate). While some of these studies found that birds with higher antioxidant levels were more likely to 'survive' (as that Francois mentioned in his previous reply), only two studies found that a measure of oxidative damage was lower in birds that returned the next year compared to birds than did not return.

Cheers

David

Thanks David for this clarification, I was about to reply to Francois pointing out that Pierre's study did not as far as I was aware measure damage. Could you share with us what these 2 studies that did measure damage were please? plus also could you indicate of the 10 how many measured a damage parameter in blood and found no link to the proxies for survival?

Thanks

John

Hi John

you can find this info in the table of book chapter I sent you some time ago.

Cheers

David

Thanks for the message. I thought that others might like to see the data with respect to the links between oxidative damage measured in blood and survival. I hope you dont mind me posting it here. The published data extracted from the draft book chapter you sent are as follows. 'No' means no effect and 'yes' means a positive effect (ie more damage in ones that were presumed dead). In all except the last case the birds were adults and in the last refers to fledging nestlings.

Beaulieu et al (2011) No

Noguera et al (2011) Yes

Van der Commenecker (2011) No

Freeman-Gallant et al (2011) Yes

De Coster et al (2012) No

so overall 2/5 gave the anticipated result from the oxidative damage-life history theory. This doesn't seem quite so strong as Francois suggested in his post.

In your earlier post you mentioned the idea that all tissues might be equally important and that what we really need to think about are holistic approaches. I think the problem with this is to try to imagine how one would measure 'holistic' damage, when actual tissue damage varies so widely between tissues and differs in how it changes when challenged by reproduction. any suggestions on how to quantify holistic damage?

Plus I am not really so sure with the suggestion that all tissues are equal when it comes to their sensitivity to oxidative stress, and the implications of this sensitivity for survival or future reproductive performance. Bradley et al (2009) for example suggest motor neurons might be particularly sensitive to oxidative damage. and Izzotti et al (2009) indicated different parts of the eye vary in their sensitivity to oxidative damage.

best wishes

JOhn

Increased sensitivity of myoblasts to oxidative stress in amyotrophic lateral sclerosis peripheral tissues.

Bradley LJ, Taanman JW, Kallis C, Orrell RW.

Exp Neurol. 2009 Jul;218(1):92-7.

Sensitivity of ocular anterior chamber tissues to oxidative damage and its relevance to the pathogenesis of glaucoma.

Izzotti A, Saccà SC, Longobardi M, Cartiglia C.

Invest Ophthalmol Vis Sci. 2009 Nov;50(11):5251-8.

Dear John

thanks for posting these articles.

I certainly agree that sensitivity to OS varies across tissues.

I do not know if an holistic damage marker exists or it will ever exist.

When I talk about 'organism level' I refer more to interpretation than to

real measurements. I mean that, even if Bradley et al and Izzotti et al (and many others) show this variation in sensitivity, I find hard to imagine that a so localised damage may impinge on fitness traits. I would rather expect that, in order to significantly impact on fitness, the damage should spread across tissues and impact on more than one single function. But maybe I am wrong. Maybe localised damage like that to eyes might be relevant if this reduces, for example, ability to detect predators.

However, all of these questions are almost impossible to address in wild animals.

I do not do terminal research on wild animals, so I will never be able to answer satisfactorily these questions. My aim is to demonstrate that OD in blood (whatever its origin is) allows to make predictions about impact on life history and kinds of organism's responses. In the end, this sounds like a medical approach: you can sometimes predict the result, but you do not fully know the underlying mechanisms.

Best

David

That seems like a strange aim the way you have worded it

"My aim is to demonstrate that OD in blood.....allows to make predictions about impact on life history and ...organism's responses."

Is your aim not to test the hypothesis that OD mediates life history effects? rather than to demonstrate it is so in one particular type of sample? I mean if there is actually no link between OD and life history in that sample type then aiming to demonstrate that there is becomes by definition pointless. Perhaps I just misunderstood your wording?

John

yes, my wording was not correct. I meant that one of my aims is to test if OD influences life history strategies. But, given that I can do this only using blood, I guess that nobody will be ever satisfied with conclusions that can be obtained because tissues differ in sensitivity.

David

I think you are correct. I personally dont think that journals should continue to accept papers based on single tissues like blood and/or single assays of damage, as tests of the OD to life history linkage. Single tissue analyses are just not informative enough about the association, since as we have shown in the above 2 studies the directions of the effects are completely tissue and assay dependent. What we really need are studies that broaden out the range of tissues and assays used. These need to be applied across multiple species so that we can get a better feel for consistency in the trends. At the moment it isnt possible to do this without terminating the subjects involved. Another extremely useful development then would be to develop assays of tissue damage that are non-terminal so that links between tissue specific damage and hard life history endpoints like survival and future fecundity could be explored. These seem likely to me the most productive future directions in this field.

Dear John

this would mean that studies on wild animals should not be done anymore, wouldn't it? Although you find different directions of the effects, there are also studies that found similar directions across different tissues. Studies on wild animals are certainly limited, but from my point of view this is normal. Every study is at the best limited and modest. Science does not ever give final answers. What you find today can be falsified tomorrow. We just produce models that are rough approximations of reality. Clearly, on this planet, we are 7 billions of people, with a huge variation in how we see things. So opinions on this point may obviously differ. I personally see non-terminal studies on wild animals very important, but clearly conclusions should not be too strong. But this is true for every study we do. But many high impact factors journals do not like cautious science....

Thanks for sharing you thoughts with us

David

Hi again

if, as you say, we should stop to do work on wild animals because of the tissue problem, we should also stop to do work on telomeres, contaminants sexual hormones or stress hormones because in all of these cases tissues may also differ in sensitivity. Hence what we measure in the blood might be not good enough to make conclusions. If we do so, we should also stop work on behaviour and life histories of wild animals because any conclusions would not have any mechanistic support. It sounds to me like a regressus ad infinitum...

Cheers

David

Thanks for your additional comments. Actually if you read my posts again you will see that I didn't say that we should stop working on wild animals. Clearly the best context for the study of life history evolution is in the wild - so stopping work on wild animals would be a retrograde step. What I said was that I think journals should stop publishing papers based only on single tissues such as blood. It seems to me that what I said could only be translated into what you suggested I said if you hold the position that wild animals should never be terminated in the name of science - and hence that blood sampling is the only legitimate methodology in field work. This is your personal position and I respect that. However, if someone doesn't hold this view then there is no problem because it would be possible to measure both blood and multiple tissues from wild as well as domestic animals. In my opinion only collecting blood samples is not a worthwhile procedure at the moment because we have no idea what it is telling us. In some animals what happens in blood is correlated to what happens in other tissues, in other studies it isn't, and in all animals that answer depends on the assays being used. Perhaps as more data accumulate it will become evident that blood sampling is a useful approach, or maybe it will turn out to be a complete red herring. But until that time collecting data only from blood samples doesn't really move the field forwards because we don't know what the data from blood samples actually mean in the context of the hypothesis of a link between OD and life histories. As I indicated in my last post perhaps the best advances will come from finding non-terminal measures of tissue specific damage so that we can have the best of all worlds.

John

for those of you that are interested the paper by myself and Michael Garrett in Bioessays that addresses some of the issues raised in this question thread that was mentioned above is now available online at

http://onlinelibrary.wiley.com/doi/10.1002/bies.201300108/abstract

you can also pick up a copy on my web site at

http://www.abdn.ac.uk/energetics-research/publications/

it is paper 409.

I will post it shortly on research gate

best wishes

john

Hi John -

I really favor tissue-based assays that have been benchmarked by the clinical toxicology, cancer, and aging research communities (protein carbonyls included, but also things like isoprostanes, 8-OHdG, other markers of protein dysfolding, etc.).

These markers have been shown to respond more reliably than blood oxidative markers like TAC to genetic mutations and laboratory manipulations that change aging rate, like calorie restriction (as you know from some of your own work). They are arguably more sensitive and better 'biomarkers' of aging-related accumulation of oxidative damage. (This is true for liver even though it's a high-turnover tissue.) Non-blood tissue levels are also known to be associated with at least some disease states that could be relevant to costs and trade-offs in nature.

The problem is that they require biopsies or terminal tissue sampling, they're expensive, and they can't really be done very well by non-specialists.

As to blood-based measures of oxidative stress - not much evidence that they say anything about disease or mortality risk - although some clinical studies do relate changes in markers of lymphocyte oxidative stress or damage to disease states in humans....

These are great questions, and I've been struggling with them, too. Somebody should write an interdisciplinary methods article!

Best -

Dear John et al.

Sorry to come so late to this discussion (I have only just signed up to ResearchGate), but it seems to me that the debate is getting unnecessarily polarised. There are different approaches to this general question (e.g. model organism vs taxonomic diversity, lab vs field, and experimental vs correlational), with different objectives (e.g. whether one is interested in the overall effects or the underlying mechanisms), but all make a contribution and none is definitive. Some studies may be able to examine lots of tissues using lots of markers, while others sacrifice this depth of mechanistic detail in favour of examining longitudinal effects (so addressing the issue of fitness), or explore the effect of contrasting environmental challenges, or heritabilities of vulnerability to oxidative damage using pedigree analysis, or whatever.

For some of these approaches a non-destructive sampling protocol (which at the moment usually means a blood sample) is the only viable approach. I agree that as yet we are not sure of the processes that lead to a given level of oxidative damage in blood, but the same is true of all tissue samples, so there is no logical reason to single out blood as being a particularly inappropriate type of sample. And if (as is the case in some study systems) assays results from blood predict traits such as future survival then this is exactly the kind of assay that John was wanting when he said:

'Another extremely useful development then would be to develop assays of tissue damage that are non-terminal so that links between tissue specific damage and hard life history endpoints like survival and future fecundity could be explored.' OK, as yet we don't know exactly WHY (for instance) a blood sample predicts future lifespan in zebra finches (B.J. Heidinger et al, PNAS 2012) but there is then scope for the more detailed examination of damage across multiple tissues to explore the mechanisms that underlie this whole-organism phenomenon. I think there is room for all approaches to contribute to this complex issue.

Best wishes

Neil

Good points Neil and great to have you in the discussion. Just to clarify what I was trying to convey. Basically my feeling at present is that we don’t know enough about what is going on to place much faith in studies that are based on single assays of single tissues, be they blood or liver or whatever. I don’t have anything against blood - just the use of single tissues and single assays. You are correct that your nice PNAS study in zebra finches shows a link from an assay in blood to lifespan, but actually that was a study of telomere lengths rather than oxidative damage per se. As emphasised above, while two studies support a link of oxidative damage measured directly in blood to fitness, three other studies show no such link. This is my point. This variation makes it hard to know how one would engage in “.. more detailed examination of damage across multiple tissues to explore the mechanisms that underlie this whole-organism phenomenon“, because it isn’t actually a phenomenon – because it has no repeatability. Plus in the context of effects of reproduction on damage, the impact in blood can be completely the opposite to the impact in some tissues, in some species (but not in others, or in other tissues, or with other assays). So that was the point of raising the whole question. If the responses are completely all over the place depending on what and where you measure – what should we rely on? Is there an objective a priori way to make such a decision? I think at present we just don’t know, and consequently my own view is that studies of single assays in single tissues (which in the past I have also been responsible for publishing) do not really take us further forwards. [Plus as a somewhat different issue I suspect such studies will be prone to a large publication biases towards positive results.]

Another potential important point is temporal dynamic measures and measures that integrate events over time. For example for telomere length we know that at least in part telomere length reflects historic events of oxidative damage (to the telomere at least, likely in part reflecting total oxidative damage to a cell/soma). Measures of blood especially are likely highly variable. Repeatabilities for these measures are low (see Simons et al 2012 on carotenoids for a short review in the discussion) and may only convey short term information. The same may be true for tissues.

Measures that integrate past damage will be very valuable, telomeres are one, but other biomarkers of aging can be categorized along this continuum as well. Predictions for correlations among biomarkers and relationships with short versus long term survival will differ accordingly.

RE to Donna: Note that for oxidative stress markers there are studies that link antioxidative status to survival/fitness (in birds), but for I do not know of strong relationships with measures of oxidative damage (both from blood). "Within species antioxidant defenses have been reported to predict survival and reproduction (Alonso-Álvarez et al., 2006; Bize et al., 2008; Saino et al., 2011)."

Thanks for the interesting discussion. Congrats on your paper with Mike - John - I really enjoyed reading it. Would it be worthwile to do a formal meta-analysis on this topic (costs of reproduction and/or wider)? Investigating some moderating variables as well? The issue of publication bias can also be addressed accordingly then. Potentially include studies that simply do correlations with any fitness proxy?

that oxidative stress components are not repeatable is not quite right. Rather there are studies showing that various molecules or parameters of oxidative status show significant intra-individual repeatability over a few to many months. I think that one problem with the discussion here is that we talk about assays to measure oxidative stress. For me, this is not quite right. We should be specific and not to put altogether assays that measure different things for the simple reason that they are not comparable.

Protein carbonyls say something about the system that does not necessarily need to be the same as that it is said by, say, damage to DNA. The action of stressors, diseases and so on may be so specific that only some biomolecules are severely impacted. This clearly implies that need to use more assays (as John said), but this has been said many times in the ecological literature. But a study that is looking at a single oxidative damage parameter can still provide valuable information if inferences are made accordingly to what is being measured and if there is available background that supports the biological relevance of that specific assay. Sometimes we need compromises, and honestly I prefer to see a paper based on a single assay rather than nothing.

Best

David

Hi David, just to rephrase.. I am saying the blood measures estimating 'oxidative stress state' are in general not very repeatable. I am not talking about significance at all.. It might be statistically significantly repeatable with an r of 0.2...

(For example: "These repeatabilities in birds range from 0.12 (Beamonte-Barrientos & Verhulst submitted), 0.14 [231], 0.3 [127] to 0.49 [232] for antioxidant capacity ),

...but this means that the intra-individual consistency is only explaining about 4% of the variance. Which to me suggests a lot of within individual variability in these measures (measurement error ignored).

I agree with more different measures but as long as separate measures are also hard to interpret.. (as you also mention, protein carbonyls might be something different than DNA damage, but what are both telling about the physiological state of the individual..)... I think it is still useful to talk about measures of oxidative stress/damage/protection just because we interpret the underlying physiology to some degree. Apples and pears are different, but they are both fruit.

Costantini et al. 2007: repeatabilities from 0.3 to 0.6.

Saino et al. 2011: 0.49 to 0.52

Galván and Alonso-Alvarez 2009: 0.79.

In addition, I have unpublished data showing high repeatabilities, especially for some antioxidant enzymes.

But of course there is some variation anyway, as for any trait you work on.

Best

David

This intra-individual variation is what I think is interesting. Especially if some variables are more variable than others, correlations among them will be low and hard to detect - especially if there is differential lag to environmental perturbations. Hence one might falsely conclude they are unrelated causally whereas they can very well be.

Approaching this problem from a different direction, why don't these measures go up or down consistently when oxidative stress is manipulated experimentally. As in Meitern (2013) in greenfinches with the paraquat treatment and a multitude of "oxidative stress" measures.

Meitern R, Sild E, Kilk K, Porosk R, Hõrak P (2013) On the methodological limitations of detecting oxidative stress: effects of paraquat on measures of oxidative status in greenfinches. J Exp Biol 216, 2713–2721.

This point is certainly important, but not so novel. That there is a time lag is well established in the biochemical and physiological literature, as well as is a background biochemical information available in order to make interpretations about why two parameters correlate or do not. This basic biochemical knowledge is fundamental to interpret a correlation between two parameters, at least in terms of their biochemical interaction.

Hi Mirre,

I forgot to say that it would be very interesting to have a meta-analysis on this. But I guess that this should not look only at offspring rearing effort, because there might also be other features of reproduction that may induce oxidative stress. In terms of changes in OS over time, here is a my paper that looked at this over reproduction.

Costantini D., 2010. Effects of diet quality on serum oxidative status and body mass in male and female pigeons during reproduction. Comparative Biochemistry and Physiology Part A, 156: 294-299

It has not received much attention from ecologists, although it is the only one that does this on a bird species. Yes, there are only 3 parameters of serum oxidative status, but I think it provides some interesting info about variation in blood oxidative status from the incubation to the chick rearing phase.

Regretfully I doubt whether the interactions between reproduction, survival and measures of "oxidative stress" can be interpreted by 'basic biochemical knowledge'.

But I see your point.. it is still important though.. as you also state.

For example, I readily see a PC1 of different antioxidants in plasma for example regressed against a variable of interest, which in my view will likely not represent what is happening if timelags and differential intra-individual variability is important.

Also one could turn the conclusion of Meitern (2013) around. The one variable that they do find an association of treatment with is actually representing the experimental elevation of oxidative stress and the other variables are readily buffered and/or show differential temporal dynamics.

Comparing a measure that integrates "oxidative stress" together with parameters that are more variable will be an interesting approach to take. My feeling is that an integrated measure of past 'oxidative stress' will be more telling of future reproduction/survival and could also be more easily interpreted. There are more biomarkers like this for example advanced glycation end-products, which built up with age and can be measured from skin (using UV reflectance) or blood/tissue. If anyone else would know of such a measure I would be very interested. I think we have covered this point of temporal dynamics etc and integrated measures for now. Thanks for the discussion.

No, I meant that basic biochemical knowledge would be important to interpret correlations among oxidative status parameters. Have also a look at:

Costantini D., Monaghan P. and Metcalfe N., 2013. Loss of integration is associated with reduced resistance to oxidative stress. The Journal of Experimental Biology, 216: 2213-2220

Yes I know and like that paper.

Just going back to the Meitern et al (2013) et al paper in JEB I think it is a really interesting paper and Mirre's interpretation of the results across the different assays is I guess the way most people would interpret it (me included) - but playing devils advocate one could argue that the measures that didn't change are the important ones, because these were the things that it was important to protect to ensure survival. While the one measure that did change was the least important one for the same reasons.

I think to an extent this study is a missed opportunity because what actually would have been good to measure were the rates of damage via the multitude of different assays in the ones that died, relative to the ones that survived, rather than comparing survivors to controls and their pre-dose measurements. By also measuring the animals that died it might have been possible to see which markers were associated with mortality. Could be a good follow-up experiment - but I cant see it being done in the UK any time soon.

I fully agree. To me, the main actual problem in the field is that scientists do not concentrate on testing hypothesis any more but rather tailor their science towards expected outcomes. As pointed out in Selman et al. 2012 recently the use of single assays might point to misleading results.

The deaths in Meitern et al 2013 are unfortunate yes. Especially because it could be that these individuals showed the higher (ior lowest) post treatment values of all, thereby masking any effects. Change in the markers or simply a lower dose would be a good ideas yes. Note that these deaths were of course an accident, the greenfinches used were much more vulnerable to paraquat than in previous studies on other species of birds.

No such studies available in rodents? It's a shame. Maybe not in the UK, but surely you would be allowed to do this in China, no? :)

Yes I saw from the paper that they did not intend at the outset to kill the animals but since it did happen it seems to me that they should have made the most of a bad situation.

Possibly this could be done in China as a premediated experiment, knowing some of the animals were going to die from the poisoning. However, China is not a completely ethics free zone so it would need to go through ethical review and be approved. But it probably would be feasible in mice if it was well designed and thought out. I am a little wary of the accusation of ethics dodging however, so would prefer to get UK permission to do it as well - even if ultimately we did the actual experiments in China. Might not be completely impossible.

All depends on the dosing I think. At a low dosage it should not more harmful than other treatment that are ethically sound in the UK.

It was risky doing all birds at the same time. If they would have started with 2-4. Casualties would be limited and the dosage lowered

@David Sorry for not replying to your meta-analysis comments. Would be good to discuss this further. Interesting ideas. M

I'm also a bit concerned about the perception that animal experimentation is kind if unlimited in china as this sheds a wrong light on the active scientists there. Europeans who conduct experiments in china do not decide to go there to avoid ethical review. Rather, we benefit from the good research opportunities, excellent students and high recognition of science in the public. Please let's not view china in the light presented in some of the european media. This comment has nothing to do with oxidative stress as a mediator of life history trade-offs but maybe initiate new comments again....

I am not saying at all that ethics should be dodged also please read John's reply. I am also not saying all experiments in China will go. Yet, UK is very strict also in EU. Centralised EU laws soon in operation or operational already might change this (i am not up to date on this). Just to clarify I do not mean to make any ethics judgement at all. I usually try to stay as far away from that as I can..

getting back to life histories and physiological mechanisms I just found this interesting paper published a couple of weeks ago that raises another issue of potential importance in this whole area and that is timescales of consequences and measurements...

Offspring pay sooner, parents pay later: experimental manipulation of body mass reveals trade-offs between immune function, reproduction and survival

Arne Hegemann1*, Kevin D Matson1, Heiner Flinks2 and B Irene Tieleman1

Frontiers in Zoology 2013, 10:77 doi:10.1186/1742-9994-10-77

Published: 17 December 2013

Abstract

Introduction

Life-history theory predicts that organisms trade off survival against reproduction. However, the time scales on which various consequences become evident and the physiology mediating the cost of reproduction remain poorly understood. Yet, explaining not only which mechanisms mediate this trade-off, but also how fast or slow the mechanisms act, is crucial for an improved understanding of life-history evolution. We investigated three time scales on which an experimental increase in body mass could affect this trade-off: within broods, within season and between years. We handicapped adult skylarks (Alauda arvensis) by attaching extra weight during first broods to both adults of a pair. We measured body mass, immune function and return rates in these birds. We also measured nest success, feeding rates, diet composition, nestling size, nestling immune function and recruitment rates.

Results

When nestlings of first broods fledged, parent body condition had not changed, but experimental birds experienced higher nest failure. Depending on the year, immune parameters of nestlings from experimental parents were either higher or lower than of control nestlings. Later, when parents were feeding their second brood, the balance between self-maintenance and nest success had shifted. Control and experimental adults differed in immune function, while mass and immune function of their nestlings did not differ. Although weights were removed after breeding, immune measurements during the second brood had the capacity to predict return rates to the next breeding season. Among birds that returned the next year, body condition and reproductive performance a year after the experiment did not differ between treatment groups.

Conclusions

We conclude that the balance between current reproduction and survival shifts from affecting nestlings to affecting parents as the reproductive season progresses. Furthermore, immune function is apparently one physiological mechanism involved in this trade-off. By unravelling a physiological mechanism underlying the trade-offs between current and future reproduction and by demonstrating the different time scales on which it acts, our study represents an important step in understanding a central theory of life-history evolution.

Keywords: Birds; Cost of reproduction; Ecoimmunology; Ecophysiology; Immunity; Life history; Carry-over effect; Avian

Dear all,

Happy New Year! Sorry for jumping into the discussion so late, but the early discussion on single versus multiple tissues and associated objections against measuring OS in blood caught my attention.

There are indeed many good arguments for looking into several organs/tissues when it comes to oxidative stress (and I really appreciate the most recent papers about this; pointed out by John and others). I do not want to go too much into the details in which species organ-specific OS is high or low, how some of the parameters seem to be contradictory, and how all this is associated with longevity. I see some advantages of using several organs/issues, yet I also see clear disadvantages of measuring OS in multiple tissues in an invasive manner, namely by sacrificing animals. Besides ethical considerations and those related to conservation of endangered species (in case you are interested to work with legally protected wild species as I am), this approach falls short on a very important point. They are all based on a cross-sectional study design, aren’t they? Yet, when it comes to the question about how oxidative stress is related to age-dependant life-history traits or longevity per se, you may rather conduct a longitudinal study with repeated sampling and not a cross-sectional experiment. Isn't it insufficient to look at OS (even when it is done in multiple tissues) at a cross-section of the experiment. Overall, there seems to be a dilemma, because experimenters face a trade-off where they have to balance the intrinsic limitations of a cross-sectional study design against the possibe intrinsic limitations generated by single-tissue studies (when based on blood for example). Or to put it in a more provocative way, what does OS in liver, kidney and others tell about lifespan related questions when based on an invasive cross-sectional study design? What is your opinion about this?

In light of the various opinions about which tissue to pick, here is a pro-blood argument. I agree that liver, kidney, muscles among others might all be relevant, but why not blood given that blood, with all its immune cells, contains one of the prime defensive mechanisms against pathogens. An immune response is known to increase OS (see also our recent paper in JEB: Schneeberger et al.) and this could as well translate into organ-specific OS, and all this could affect life-history traits and longevity (to put it simple).

So, are we facing the described trade-off, and if yes, how should we deal with it in an optimal way.

Christian

Dear Christian

I personally share your thoughts. And I would also add that immune cells not only generate free radicals to use against pathogens, but they also have their own metabolism and redox environment. And similarly red blood cells generate their own reactive species, suffer damage and contains multiple antioxidant defences.

The blood is a tissue and it is also subject to senescence and to diseases that can be lethal. And we know at least from veterinary medicine that various deceases of blood are associated with an increase in oxidative damage.

Finally, in response to a previous comment, telomeres are extremely influenced by free radicals, hence they can work as a index of oxidative stress (already suggested by some authors). And we know that telomere shortening can vary across tissues. Hence, any argumentations against using only blood for OS should also be valid for telomeres.

However, as Neil said, measures of telomeres in blood (as those of OS) are giving valuable ecological information that cannot be overlooked.

Cheers

David

Christian –many thanks for your input. These are very good and pertinent points. You are correct. There is a trade-off at the moment between having a comprehensive picture of what is happening across all the tissues in the body – which could only be done by killing the individual and hence necessarily is a cross sectional (and in some cases ethically unacceptable) design, and making repeated measurements of a single tissue which lacks the power of cross tissue comparisons but gives a more powerful longitudinal approach. Unfortunately many past studies have managed to combine things to achieve the least powerful combination – single tissue studies in cross sectional designs. As I pointed out above, ideally what we need are assays of tissue damage across multiple tissues that are not terminal, so that we can achieve the ideal solution of a multiple tissue perspective in a longitudinal design. This may not be as utopian as it seems. There are for example stable isotope approaches that allow us to non-invasively probe liver metabolic function – by feeding an animal (or human) a labelled compound and examining the time course of its metabolism and appearance of end products in the breath. I wonder if a similar approach might be devised to probe liver damage or protection - any biochemists out there with a good idea how that could be done? Alternatively maybe there are end products of tissue specific damage that appear in urine and could be picked up by a metabolomics screen.

Following on from Davids reply. I should emphasise that I have nothing against blood as a tissue. Indeed it is the only tissue for which it is feasible in a small animal to make such repeated measurements. There is equally no doubt that diseases of the blood can be lethal. The question I raise again is what does it mean if damage goes up in blood but down in another tissue like the liver? Or goes up when we measure carbonyls or is unchanged when we measure lipid damage? Moreover what can we infer if the animal itself increases protection in one tissue, and hence damage goes down (as happens in the liver in our studies of reproduction) but sacrifices protection in another tissue and hence it allows damage there to go up (as happened in the blood in the same studies).

It all comes back to which tissues and damage targets are the most important drivers of the important functional end points (like survival and future fecundity), and I think at present we don’t actually know the answer to that – nor given the trade-off Christian has highlighted do we have a good method by which it might be studied.

Hi John,

Thanks for your response. I think you are right. We should strive to develop OS/tissue-damage markers or tissue sampling protocols (possibly minimal-invasive biopsies) that facilitate longitudinal studies based on multiple tissues and samples. This should be the direction to go. However, we may have to follow compromises at this point as long as the compromise advances science. From my (limited) experience with senescence studies, cross-sectional study designs are not really appreciated in the scientific community (to put it mildly), and, frankly, they may be indeed of limited use in the study of age-related traits. You stated earlier that you wouldn’t accept a single tissue study (particularly when it is based on a cross-sectional study design). Yet, I see some potential for single-tissue studies when the (biological) relevance of a single-tissue approach (e.g. blood) has been validated, e.g by using artificial challenge experiments, parasite infestation experiments, or by looking at rates of reproduction and how they effect OS in the selected tissue. In particular, I see some relevance for blood as a target tissue when you are interested in immuno-senescence and the interplay between OS and immune defense. Based on such validations and research focus a single-tissue approach would have scientific merit, wouldn’t it? Lastly, single-tissue studies may have particular merit when ethical and conservation arguments prevent an invasive multiple-tissue approach (provided that validation experiments support the significance of the selected tissue for the given research question). You are certainly right in that the choice of tissue and damage targets may vary between study questions and also between organisms, and that much has to be learned in the future. Therefore I would plead that single-tissue studies should still be considered when there is good reason for not following a multiple tissue approach and if appropriate validations demonstrate their biological relevance. I hope that non-invasive markers for OS and tissue damage will be developed at any point soon. Any suggestions how to do this?

Christian

Christian and John. Interesting discussing on longitudinal versus cross-sectional. It will be possible to detect selective disappearance from the population assuming the cross-sectional slope is the within-individual slope (and/or assuming selective disappearance acts on absolute level of trait and not slope, in other words assuming heterogeneity of age against trait is negligible) and comparing the variance at each timepoint. Samples sizes will have to be considerable to do this but it is possible. I will have to think on the specific statistics to do it, but it should be feasible. I think that would be a way to go.

A more experimental approach would be to within-individual sample blood and correlate the resultant change with age to cross-sectional measures of tissues.

One more thought. The within individual consistency in oxidative stress markers across different tissues in cross-sectional sampling will be very interesting. Actually most models of aging will predict that with aging tissues will start to differ from each other due to stochastic damage. A cross-sectional (age) experiment with considerable sample size would be able to test that prediction.

Relevant to the question of whether there is individual consistency in oxidative stress markers across different tissues in cross-sectional sampling is the new attached paper by Reichert et a., which shows pretty high within-individual correlations between telomere lengths in red blood cells and in other tissues (even those differing in rate of cell turnover, such as liver, bone marrow, spleen, brain).Given that telomere length is strongly affected by levels of oxidative stress, this is quite an encouraging result.

Neil

Thanks for that paper Neil. Interesting, the lack of correlation with bone marrow and RBC, actually a negative correlation once this outlier is removed. Also the RBC TL is longer than in bone marrow. So telomerase kicks in during differentiation or so (different celtypes with different lengths)? Probably there is quite a bit out there about it, maybe not.. Although the authors say it is possible that their bone marrows samples were "contaminated" with other cell types. No mentioning which parts of bone they used as well, might well make a difference..

I would be very interested in seeing these correlations at differently aged subgroups.

Thanks for posting this Neil. Interesting paper. I'd be interested to know your, and others, views on the severely non-normal distributions of the telomere lengths of different individuals. To me this looks like a small cluster of uncorrelated data, with various outliers dotted around that must completely drive the regressions and estimates of significance, particularly since they used Pearson rather than Spearman correlations. Given the distributions - rank correlations would seem much more appropriate. My guess is that if they had ditched the three outliers with very high RBC telomere lengths instead of just one, all the cross tissue correlations would disappear.

John is right. It is quite tricky and although this is excellent work and a very much needed study.. I would say that correlations between RBC and tissue might be there but they seem rather weak ~0.2 or so. Warrants some new studies I would say. I measured figure 2 and re-analysed using Spearman rank and Pearson (omitted top 2 outliers for consistency). The only one that seems robust is the correlation between RBC and Brain TL. (r=pearson, rs= spearman rank, number thereafter is P)

spleen RBC

r =0.79 0.0003

rs=0.29 0.284

omit highest 2

r=0.18 0.53

muscle RBC

r =0.57 0.014

rs=0.28 0.26

omit highest 2

r=0.46 0.07

heart RBC

r=0.48 0.046

rs=0.23 0.37

omit highest 2

r=0.044 0.87

Liver RBC

r=0.64 0.032

rs=0.46 0.15

omit highest 2

r=0.37 0.33

Brain RBC

r=0.72 0.004

rs=0.72 0.004

omit highest 2

r=0.63 0.027

Dear John

I noted in your FE article, where you find this variation between liver and serum, that

there is no mention of treatment of samples with streptomycin sulphonate before running the assays for protein carbonyls. This is a very important step because it allows to remove interference from carbonyls on nucleic acids. I also noted that you did not specify if samples were diluted in order to have 1mg of proteins before the assays. So I wonder in which way you calculated the concentration of protein carbonyls per mg proteins.

Cheers

David

Hi all

You are absolutely right and we will go back to the tissue telomere data. You should take our article for what it is, just a try to sort out what are the telomere length links between some tissues but when using a small sample size . A definitive conclusion will need a much bigger number of individuals. I will be happy to do it again with mice for example, where telomerase activity is higher. At least it suggested that some relationships exist and that following red blood cell telomere erosion may give some indications on the way other organs aged. But it also rises another question that one may be confronted with while having several oxidative stress measurements in different tissues and willing to relate them with fitness traits. They may give contrasting relationships going in different directions. How to sort out what is important and what is not? On which other metabolic and cell pathways may oxidative stress variation be related or having an effect? This probably depends on the question but a way to tackle it may be to use proteomics and screen protein profiles of tissues of interest. It will allow to look at biological changes as a whole (oxidative balance, cell metabolism, energy metabolism, others...the holistic approach of David). We used it in mice while studying cost of reproduction and surprisingly (or not) found no relation between blood oxidative stress and liver proteins (but that other proteins are actually affected). Does anyone already try this?

Thanks to Mirre to have highlighted us about the brain -RBC telomere stronger relationship. I wonder whether this is of particulalr interest for John questioning. I remember a paper from Fridell surexpressing UCP2, a mitochondrial protein supposed to decrease ROS production, in brain of fly. It induced an increase in lifespan in those flies, and given that (1) telomere lengths seem to be related between RBC and brain, and that (2) RBC telomere length is a proxy of lifespan (at least in early life in ZF), it putatively points the brain as a ikey tissue for oxidative stress study. Either altered cognition or overall metabolic dysregulation insured by the CNS may render the brain particularly interesting to study as a target of oxidative stress. Also, the CNS is involved in circadian rhythmic regulation, and we are looking at the relationships between circadian desynchronisation and oxidative stress, but without great success for the moment...

Hi Francois, thanks for your comments, very interesting. The nice thing with flies is tissue specific - conditional - knockouts which allow us to delve into mechanism faster. It also depends where the difference in TL of bone marrow and RBC comes from and the "lack" of correlation. (Note that actually it seems the relationships are all positive and thus across the board it seems that there is a relationship between RBC and tissues, but maybe weaker in some tissues than others). Maybe RBC TL is a reflection of shortening in the blood stream which is possible, or that RBC originating from bone marrow get elongated differentially between individuals or so or with age..

Actually circadian desynchronization is one of my interests in relation to aging (mostly because my wife is a chronobiologist). How are you going about this, multiple phase-shifts, predictable or unpredictable and/or constant lighting. If you would like to discuss this we can talk via email or so.. it is a bit of topic here I guess. :)

Yes that is a good idea. I am working with E Challet on desynchronisation (jetlag of 10 hours for 3 months) on mice and diurnal rats. Let's continue this discussion by email. I will have more to say in February-March when I will get the final results on telomeres.

Apologies for the delay in responding to these comments. Hectic weekend preparing for trip back to the UK and travelling all day today. In reply to the recent comments

Mirre. Thanks for the reanalysis of the data for the telomeres paper along the lines I suggested might be useful. It certainly seems that using non-parametric statistics the conclusions are somewhat weakened. In fact if you teach statistics this is a textbook example for students about why conforming to the assumptions of normality are important, and how the choice of statistics (parametric or non-parametric) can have a big impact on the conclusions. That said the continued link of RBC telomeres to brain telomeres is of great interest, and perhaps points to the brain as a tissue that would be worth focussing on. As I noted in a comment above, as far as I am aware there is no study looking at brain oxidative damage due to female reproduction in mammals to date (although there is one for males).

Francois. You picked up the exact problem. How do we interpret it when different tissues respond to the treatment in different ways? I think however you have also highlighted another issue here and that is when we talk about damage in a given tissue there are many potential targets in that tissue and the consequences of damage to them are not always equal. In fact, the gist of Shelly Buffenstein’s observations in naked mole rats, which show phenomenal levels of protein damage and lack one of the main antioxidant enzymes (glutathione peroxidise) but are the longest lived rodent (for their size), is that they use certain proteins to sort of soak up damage and protect the more important ones. So as we get more stuck into the details of this it will certainly become more and more important to know what exactly got damaged – because paradoxically some damage may actually be a form of protection. It also seems likely for example that damage to some sections of non-coding DNA would be irrelevant, while other damage (e.g. to promoters or within exons) might be devasating. Presumably then repair is targeted at the important bits and damage accumulates where it is least significant. My guess is that there may be a ton of literature on this already out there. However, it means that simply knowing oxidative damage has gone up may not really be very useful knowledge at the end of the day. For proteins, proteomics seems a good route to get a handle on this but while it is possible to detect phophorylated proteins I am unsure if it is feasible to detect oxidative damage on proteins by MS-MS. Would be very interested to learn more about your attempts to do this. Of course, repeating the zebra finch study in mice would be very welcome. Especially if you were to also measure direct markers of oxidative damage in the tissues as well. In fact we have a ton of mouse tissues already collected from completely phenotyped individuals so if you want to save some time and collaborate on making such measurements (ie you do the teleomeres and we will do the other oxidative damage markers) then please don’t hesitate to get in touch.

David. Thanks for the specific comments on the methods in our paper(s). I am sorry it wasn’t more clear. The commercial kit we used is specifically for protein carbonyls. During the assay the livers were homogenized with ice-cold saline. The homogenate was centrifuged for 10 min at 1500 g at 4 °C. The supernatants were used for the assay, and were treated with with streptomycin sulphonate to remove interference from carbonyls on nucleic acids. The samples were not diluted to 1mg protein. This is because before the measurement, the protein concentrations for all the samples were determined using Bradford method, allowing us to correct for protein concentration in the final values. I can send the calculation details if you want. Hope that is clear but please get back if you need more information.

Francois, Mirre – regarding circadian desynchrony and ageing: the attached paper we published a couple of years back may be of interest.

Wyse, C.A., Coogan, A.N., Selman, C., Hazlerigg, D.G. and SPEAKMAN, J.R. (2010) Association between Mammalian Lifespan and Circadian Free-Running Period: The Circadian Resonance Hypothesis Revisited. Biology Letters 6: 696-698

Best wishes

John

Dear John

thank you very much for the clarifications.

Cheers

David

Dear John

I have to say that it is quite curious that oxidative damage goes down in the liver.

In fact, if we look at ecotoxicological research or at the effects of stress hormones, liver is actually a target tissue, showing strong increases in oxidative damage or changes in local expression of genes related to oxidative stress. Might it be that, for some reason, artificial selection of voles led to phenotypes that prioritize protection of liver? Do you have any chance to replicate the study using wild-caught individuals maintained under mesocosm conditions?

David

I agree that the decrease in damage in the liver in reproduction is a surprise and completely the reverse of expectations from life history theory.

But it is important to note that this is not an isolated observation unique to these voles. Michael Garratt observed it in wild and domesticated mice ahead of our studies, and we have also observed it in gerbils. (pawel koteja also observed reduced damage to kidneys and muscle in reproducing bank voles).

We cannot disprove with our current data the possibility that this is an artefact of domestication. It seems however unlikely that selection of the voles has acted to dramatically change their responses relative to what happens in the wild, as the colony is not long enough in captivity, and regularly replenished with fresh wild stock. Plus some of Michael's mice are effectively wild - and show the same response.

John

Hi John:

In a recent search of the literature on clinically relevant oxidative damage biomarkers, I saw a notice that in 2013 the U.S. National Institute of Environmental Health Sciences launched an initiative to do "a comparative study of biomarkers of oxidative stress (BOSS) to find out whether a fingerprint for measurement of oxidative stress exists." The comparative emphasis of the project extends from humans to nonhuman primates and rodents.

The research group listed here (and publications to date) are very clinically (and human-) oriented, but maybe this effort will produce some results that are relevant to evolutionary and comparative zoologists. Here's a link to the site:

http://www.niehs.nih.gov/research/resources/databases/bosstudy/

Best -

Donna Holmes