Technology Review - Published By MIT
Advertisement

Preston Estep et al. Dissent

By Preston W. Estep and Colleagues

Tuesday, July 11, 2006

smaller text tool iconmedium text tool iconlarger text tool icon

The stated goal of the SENS Challenge was to demonstrate that "SENS is so wrong that it is unworthy of learned debate." Whether or not this was accomplished by a given submission was to be determined by a panel of judges, and the final judges were self-selected from a large pool of people offered positions on the panel. Persons with relevant expertise were not included in the final panel of judges, leaving only those who know little or nothing of the existing learned debate within gerontology and life extension research. We have long objected to these and other aspects related to the judging of this challenge, and they have given us little confidence in its overall structure.

Therefore, we decided to direct all of our communications -- including this response to the decision of the SENS challenge judges -- primarily to the Technology Review readership. Since we don't regard SENS to be legitimate science or engineering, we didn't criticize it as a bad or immature example of either. We also didn't attempt to show that SENS is demonstrably wrong, since this is extremely difficult to do with an untested plan comprising legitimate science bundled together with hand-waving speculations -- even though the majority of these speculations cannot be taken seriously. Instead, we used this as an opportunity to describe general features of life extension pseudoscience and we used these general features to assess SENS. We showed that SENS is stereotypical pseudoscience, with its characteristic pervasive misrepresentations, diversionary sophistry, naïve and faulty science, and so on.

The summary of the judges' opinion states that SENS is not "demonstrably wrong," and Craig Venter says we have "not demonstrated that SENS is unworthy of discussion." It seems they suggest that SENS is highly speculative theorizing but not outside the bounds of legitimate science or engineering speculation. We strongly disagree with this assessment. Here are three primary attributes of SENS that differentiate it from the kind of nascent science or engineering described by the judges:

1. Direct contradiction of key claims by much available and generally accepted evidence.

2. Aubrey de Grey's pervasive falsehoods and misrepresentations.

3. Aubrey de Grey's demonstrated misunderstanding of relevant science and engineering.

These attributes might not make SENS "unworthy of discussion" by willfully uninformed immortalist dreamers, but they certainly make SENS "unworthy of learned debate" by people interested in real gerontological science and engineering. The judges also write: "Some scientists react very negatively toward those who seek to claim the mantle of scientific authority for ideas that have not yet been proved. Estep et al. seem to have this philosophy." We think a more appropriate statement of our philosophy is that we react very negatively toward those who seek to claim the mantle of scientific authority for ideas that are directly contradicted by a vast majority of current evidence, and are based fundamentally on misrepresentations of relevant science and engineering.

Curiously, the judges deemed our submission to be the best among three, but they disagree with our overall assessment that SENS is pseudoscience without even acknowledging the reasons why we reached our conclusion. They make no comment on any of our primary criticisms or on the presentation of evidence clearly demonstrating that SENS possesses the attributes listed above, attributes that strongly suggest that SENS is unworthy of learned debate. Why not? Is their silence an implicit acceptance of such attributes in early-stage science and engineering? If so, we strongly disagree. We think the presented evidence is clear and the judges have simply failed to further investigate the pattern of Aubrey de Grey's misconduct, even though we presented to them ample justification to do so. Therefore, we provide even more evidence that Aubrey de Grey has made overtly and transparently false claims in his response to us, and despite the ease with which these falsities are found, the judges failed to detect or consider them.

Comments

  • Response to Estep et al.'s new allegations of scientific misrepresentation
    As in the rest of this exchange I shall restrict myself to Estep et al.'s scientific points, rather than emulating their style of discourse. In their protest at the Challenge judges' failure to see SENS their way, Estep et al. reveal that not only do they not understand the word "engineering", they also don't understand the terms "deceleration" and "time point". I'm posting this in multiple parts because posting it on one go wasn't allowed.
    Rate this comment: 12345
    Guest (Aubrey de Grey)
    07/11/2006
    Posts:1
    • Response to Estep et al.'s new allegations of scientific misrepresentation
      Bennett-Baker et al.'s Figures 3 and 4, it is certainly the case that an additional increase in activation is seen between the middle age and the old age, as opposed to the plateau that is seen in the mutant mice, but that additional increase is unmistakeably smaller than the increase between the young and middle ages (which differ by the same margin, 11 months, as the middle and old mice do). Figure 4 even helpfully offers significance probabilities to support this. A deceleration that hasn't yet run to completion is still a deceleration.
      Rate this comment: 12345
      Guest (Aubrey de Grey)
      07/11/2006
      Posts:1
      • Response to Estep et al.'s new allegations of scientific misrepresentation
        Similarly, it is true that in addition to the detailed studies done on just two pairs of twins (one aged 3 and one aged 50), Fraga et al provided one table (Table 4, in their supplementary information) statistically analysing their entire dataset covering a range of ages - but the analysis they did was to split the subjects into two age groups and do ANOVA between these groups.  Thus, two and only two time points were reported. The only analysis that Fraga et al. provided covering more than two time points, Figure 5, found no statistically significant differences across a range of 19 to 66 years of age.
        Rate this comment: 12345
        Guest (Aubrey de Grey)
        07/11/2006
        Posts:1
        • Response to Estep et al.'s new allegations of scientific misrepresentation
          Estep et al. are merely digging themselves into an ever deeper hole by these repeated demonstrations of their own inability to read and understand scientific papers. It is time for them to accept that when a group of like-minded specialists can convince each other of their shared position but cannot convince highly intelligent people from other disciplines, the possibility that they are overencumbered by conventional wisdom must be seriously considered.
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/11/2006
          Posts:1
        • What the Fraga et al. data really show
          The Fraga et al. data show results from epigenetic studies on twins of many ages. They show in Table 5, for example, that for all measured variables the older twins showed higher mean differences and larger variances in epigenetic drift than younger twins of many ages. These are not "time points," these are groups stratified by chronological age. The data represent many ages. The five time points referred to by de Grey are controls demonstrating there isn’t short-term variation in an individual’s epigenetics.
          Rate this comment: 12345
          Guest (Preston Estep)
          07/11/2006
          Posts:1
          • See Table 5 of Fraga et al.
            Here is a link to the Fraga et al. paper and to some key data. Aubrey de Grey casts doubt on these data by claiming they represent only two “timepoints.” But these aren’t even timecourse data, they come from the analysis of 80 twins of many chronological ages, which in essence represents many chronological timepoints. http://www.pnas.org/cgi/reprint/102/30/10604
            Key data can be found here, particularly in the two right-most columns on “Twins distance” http://www.pnas.org/cgi/data/0500398102/DC1/5
            Rate this comment: 12345
            Guest (Preston Estep)
            07/11/2006
            Posts:1
      • Aubrey de Grey's false and misleading overinterpretation of the data
        Aubrey de Grey claims the increase between middle and old (M and O) "is unmistakeably smaller than the increase between the young and middle ages." Enough data are provided in this paper to test this claim. In the caption of Figure 4 mean expression levels are provided. The difference between M and O (relative to the difference between young(Y) and M) is less in spleen and--in clear contrast to de Grey's claim--more in kidney. So, his claim is half true, sort of, but there isn't enough statistical information provided to say these data show a significant--much less unmistakeable--deviation from linearity.
        Rate this comment: 12345
        Guest (Preston Estep)
        07/11/2006
        Posts:1
        • Aubrey de Grey's false and misleading overinterpretation of the data
          However, it appears that in kidney there MIGHT be a significant deviation ABOVE linearity at the oldest age (Fig. 4B), but rigorous claims of acceleration or deceleration are not supported by the data and statistics presented. The higher p-values in the comparisons between M and O are likely due to the obviously larger variance of expression in M and O animals relative to Y, not to a smaller mean difference between M and O, as claimed by de Grey, since the mean difference in kidney is actually larger.
          Rate this comment: 12345
          Guest (Preston Estep)
          07/11/2006
          Posts:1
        • Re: Aubrey de Grey's false and misleading overinterpretation of the data
          Progress! -- Estep has shifted his evaluation of my statements from "overtly and transparently false" to "half true, sort of". However, he's still wrong.  In my original text that he castigated, I wrote that the work reported in this paper "SUGGESTS a deceleration". This is the correct way to describe a trend that, as Estep notes, falls short of statistical significance. As to whether it is indeed a trend to acceleration or to deceleration, Estep then makes a blatant statistical error by asserting that only the means of the three sets of data matter, not the variances. P-values are required in scientific papers for a reason - namely, that they are the correct measure of confidence in what was seen. (See, for example, Miller et al. J Gerontol 56:B52.) I therefore maintain that it is indeed appropriate to describe a shift to a less significant P-value as a trend towards levelling-off, i.e. a deceleration.
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/11/2006
          Posts:1
          • Yet another correction to de Grey's misrepresentations
            De Grey continues to pretend that the data really do suggest a deceleration but there is an even clearer suggestion of an acceleration of epigenetic drift with age, since the data on kidney appear to show a much more clear trend upward than the slight downward trend of the data for spleen. Even though this is true I would object just as strongly to the claim that these data overall suggest an acceleration of epigenetic drift. Overall, these data do not suggest either an acceleration or deceleration. But, so what? Epigenetic drift increases with age.
            Rate this comment: 12345
            Guest (Preston Estep)
            07/11/2006
            Posts:1
          • And yet another correction
            I never claimed the variances don't matter. In fact, I claim the opposite. de Grey's claim that a less significant p-value means a trend toward leveling off is wrong. A higher (less significant) p-value means that the differences between the mean values of these distributions are less statistically significant, and this is completely dependent upon the sample variances. As can be seen in the kidney data the difference between the means of O and M are actually higher than between M and Y, but the p-value is also higher, which simply indicates that the large variances of the O and M groups erode statistical confidence.
            Rate this comment: 12345
            Guest (Preston Estep)
            07/11/2006
            Posts:1
            • Re: And yet another correction
              Unless my cookies are deceiving me, the Miller et al. article I referenced earlier is free online:

              http://biomed.gerontologyjournals.org/cgi/content/full/56/2/B52

              so readers can decide for themselves. Estep was indeed saying that variances don't matter, by saying that if the means are accelerating but the variances make the P values rise then the trend is an acceleration, i.e. the role of the variances in determining the trend of the P values should be ignored. This is exactly the mistake that Miller et al. pointed out.

              Let's also not forget that Estep also described by discussion of Fraga et al. as "false and misleading" and has yet to respond to my rebuttal of that allegation.
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/11/2006
              Posts:1
              • More confusion and misunderstanding
                If one simply compares what Aubrey de Grey claims I wrote to what I wrote, they'll see what is at issue here and why SENS is not taken seriously by legitimate scientists. He continues to claim I wrote one thing while I actually wrote the opposite.
                Rate this comment: 12345
                Guest (Preston Estep)
                07/13/2006
                Posts:1
          • Yet another
            In discussions of real science and engineering I regard obviously misleading half-truths like the one at issue here as “overtly and transparently false.” Aubrey de Grey has yet to acknowledge that the data in these figures suggest an acceleration of epigenetic drift with age at least as clearly as they suggest a deceleration. In other words, the opposite of his claim. Overall, neither claim is correct.
            Rate this comment: 12345
            Guest (Preston Estep)
            07/11/2006
            Posts:1
        • Means, medians, and p-values don't support claimed deceleration
          In these figures p-values result (approximation) from Mann-Whitney U tests performed on pairs of different age groups. The p-value is simply a measure of likelihood that the two groups have the same median values, or that the two distributions are the same. The clearly higher median (and mean) difference between O and M (relative to M vs. Y) for liver suggests the opposite of a deceleration. The slightly higher p-value does not show a deceleration, it simply means that one can state with only slightly lower confidence (but still high confidence) that the two groups have different medians.
          Rate this comment: 12345
          Guest (Preston Estep)
          07/11/2006
          Posts:1
  • Guest (Luxipharos)
    07/11/2006
    Posts:1
  • The debate
    However much I would love to believe we could acheive Aubrey's SENS dream I can't help but feel this is another case of "in ten more years". Similar to the promises of abundant power from fusion power advocates.  We are just starting to piece together how ageing works at a celular level, and even were we to fully understand, model, and modify cells to make them work better, longer it may never confer the benifits desired for the organism. In fact it may make things worse.  To reverse ageing in a living organizm still smacks of science fiction, short of growing new organs from modified cell cultures and replacing them I am doubtfull at best.  While your at it whip me up one of those flying cars and personal humaniod robot to do all the house work.
    Rate this comment: 12345
    Guest (Gurthang)
    07/11/2006
    Posts:1
  • Like Kurzweil's singularity
    I'm a "Joe Public" non-scientist and most of this discussion, although interesting, is completely over my head.  I offer this observation/comparison, which may or may not have any relevance to the discussion:

    As a layman, AdG's ideas sound similar to Ray Kurzweil's concept of "the singularity," which seems to indicate that technology will solve every problem, humans will live forever, re-engineer themselves to become anything they want to be, etc.  Sounds great, but to me, those ideas are absurd on their face.  Non-scientist contributors to Kurzweil's site sound (at least to me) like new-agers with a scientific vocabulary.  I think this debate would look very similar if opened up to comments from the general public.

    Completely apart from the scientific and funding discussion is a more fundamental question:  "what would you do with all that longevity?" and "What's the matter with dying anyway?"
    Rate this comment: 12345
    Guest (Paul Schantz)
    07/11/2006
    Posts:1
    • Re: Like Kurzweil's singularity.
      What would I do with all that longevity? Continue to have fun I guess.
      What would you do with all that longevity? Use your imagination.

      What's the matter with dying anyway? Dying may not be too bad; it’s that long slow decay that precedes it that bothers me.
      Rate this comment: 12345
      Guest (Jeff Hall)
      07/11/2006
      Posts:1
    • Lack of imagination
      The problem with dying is life is over. You do not any longer exist. Do you remember anything before your were born? No! How about ever been under via anesthetic? Did you dream? Not if it was done properly. Death is nothingness forever and ever.
      Rate this comment: 12345
      Guest (FutureQ)
      07/12/2006
      Posts:1
      • Lack of imagination part 2
        My heart stopped once and I had to be shocked back alive. I had what some might call a Near Death _Non_ Experience. The difference
        between mine and the proverbial "I saw granny and Jesus" is that my lungs
        were being vacuumed at the time. The oxygen was ripped from my lungs too
        fast for the brain to ponder what was happening so it then could not trick
        me into experiencing a tunnel and lights. But I have experienced those before.
        Rate this comment: 12345
        Guest (FutureQ)
        07/12/2006
        Posts:1
        • Lack of Imagination part 3
          I was doing aerobatic maneuvers at the time and blacked out on the third
          loop of the maneuver. I saw the tunnel and hallucinated. In this case the O2
          was slowly diminishing in my brain. See the difference now using logic
          instead of emotional claptrap? I remember nothing from the heart stoppage
          event except lost time, nothing except being told I had to be coded. I was
          dead by any measure. If, for instance, they had not revived me, does anyone
          truly think I would have suddenly somehow become aware again and then
          experienced entering the after life? By what criteria could that occur?
          Rate this comment: 12345
          Guest (FutureQ)
          07/12/2006
          Posts:1
          • Lack of Imagination part 4
            My brain was simply shut off...  period.  Death is the end of the world for the individual no amoumt of wishful thinking makes any difference.  It is then best to postpone final death as long as one can.  Only our generations alive right now have an tiny inkling of that chance as unfortunately none before had.  So go on hem and haw and piss and moan fools and miss that chance if you want to but just don't get in my way!
            Rate this comment: 12345
            Guest (FutureQ)
            07/12/2006
            Posts:1
      • To dream of forever
        We all want to live longer healthier lives there is not arguement there. It is a selfish urge we are all born with but the cells that make up our bodies have no such urge.  And to "teach" them that will likely break that fragile balance that makes us work. (Systemic cancer comes to mind)  We haven't even mastered genetic engerneering or cloning yet I see little hope in this science fiction extending our lives in any meaningful way any time soon. If you want to have a real impact now:
        Eat healthy high nutrition food
        Moderate Excerse
        Never over eat
        And keep mentally active
        But that requires self-control and makes nobody any money.  It is much easier to sell people the dream of a quick fix while they press the channel up button on their remote controls.
        Rate this comment: 12345
        Guest (Gurthang)
        07/12/2006
        Posts:1
        • Forget Forever: How about one suffering person at a time
          you said: I see little hope in this science fiction extending our lives in any meaningful way any time soon.

          Regenerative medicine is hardly science fiction.  Tissue engineering, stem cells, gene therapies, etc etc.. all were science fiction less than a decade ago and are now bursting wiht promise and potential.  The only thing that will prevent these technologies from making a significant difference in relieving the suffering of aging in the near-term is the foot-dragging of pessimists.
          Rate this comment: 12345
          Guest (caerus)
          07/12/2006
          Posts:1
    • What to do with longetivity?
      Completely apart from the scientific and funding discussion is a more fundamental question:  "what would you do with all that longevity?"

      Well, to start with, the Universe seems like a really big place.  I'd like to know if their is life anywhere else in it, and what that life is like.
      Rate this comment: 12345
      Guest (Mark Bahner)
      08/14/2006
      Posts:1
  • labeling rewarded with $10,000
    It is difficult to see why Technology Review would choose to reward Estep et al.Their strenuous labeling is no substitute for logical argument.They condemn proposed projects as speculation as if we should abandon any problem which requires innovative solution.Noone should imagine that SENS will run its course without modification or addition as techniques are tested and results emerge;but SENS simply proposes repairing the various types of damage caused by ageing.Since no physical law precludes it this will happen eventually,and SENS could be the quickest,and therefore the most humane way of getting there.
    Rate this comment: 12345
    Guest (Dave Fisher)
    07/12/2006
    Posts:1
  • Fusion is trivial next to this
    There are many tens of thousands of pathways in every single cell.  They are all coupled, and they are all non-linear.  Pretending we can operate a system of tens of thousands of non-linear coupled pathways outside its normal operating range and achieve superior performance without understanding the "details" is breathtakingly naive. 

    The example of the car being modified to travel to the moon is trivial next to what AdG is proposing. 

    Biological systems "seem" simple because the details are transparent to the user.  My computer is simple because I only need to push one button to turn it on.  They talk about Moore's law doubling capacity every few years, they should just add another button that will double capacity when I press it, so I don't have to replace my hardware.  Should be easy, just add another button. 
    Rate this comment: 12345
    Guest (Dave Whitlock)
    07/16/2006
    Posts:1
    • Sidestepping our ignorance
      Dave, your absolutely correct point is precisely what led me to the SENS alternative to standard approaches to postponing aging - approaches which, gerontologists have rightly concluded for very much the reason you give, will not bear substantial fruit for many decades to come. Check www.sens.org for details.
      Rate this comment: 12345
      Guest (Aubrey de Grey)
      07/16/2006
      Posts:1
      • Huh?  Ignoring ignorance by "sidestepping" it?
        How is ignoring our ignorance of the tens of thousands of coupled non-linear pathways possibly going to work? 

        Your approach is the "add another button" approach vs the approach of the engineers at Intel of understanding the fundamentals and then designing and building something with the desired functionality. 

        Living cells are enormously more complicated than any microprocessor.  Could you double the capacity of a microprocessor by retrofitting it, while remaining ignorant of the "details"?  If you can, that isn't "engineering", it is magic. 
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/17/2006
        Posts:1
        • No, "sidestepping" by factoring it out
          An example may help.  The plethora of coupled pathways you describe cause, among other things, mutations in our mitochondrial DNA that stop it working. This may be a major contributor to aging. The SENS approach to addressing this is to make the mitochondrial DNA unnecessary by putting suitably modified copies of it into the nucleus. (Before you tell me that that's impossibly hard, check the relevant page on my website, http://www.sens.org/mtmut.htm, and the papers linked from it - it's already been done in culture for some of the genes in question and there are only 13 such genes in total.) Once this is working, the complexity that leads to the mutations is sidestepped: the mutations are now harmless, so we don't need to stop them happening, so we don't need to know how they happen.
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/17/2006
          Posts:1
          • No, ALL effects come from pathways good & bad
            I looked at your website and couldn't find any links to any papers showing any transfer of mitochondrial proteins to the nucleus, their successful expression and functional incorporation into mitochondria. 

            Give me a link, and I will try and read the paper.

            I read your papers on hydrophobicity and find the unpersuasive. 

            I think mitochondria must express those proteins for regulatory purposes while far from the nucleus, as in neurons.  Whether a mitochondrion is a meter, an inch, or a mm from the nucleus, it can't those proteins fast enough.

            Since the proteins of the respiration chain can be damaged, they have to be replacible, or the mitochondria must shut down the whole chain. 
            Rate this comment: 12345
            Guest (Dave Whitlock)
            07/17/2006
            Posts:1
            • Successful allotopic expression
              Here are the best ones:

              Nagley et al PNAS 85:2091
              Zullo et al Rejuvenation Res 8:18

              The main problem with the idea that some proteins in the respiratory chain must be synthesised near their site of function is that all the respiratory chain enzymes that have mt-coded subunits also have nuclear-coded subunits, which must necessarily be expressed at tolerably similar levels.

              Feel free to email me for more expansive discourse on this.  ag24@sens.org
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/17/2006
              Posts:1
              • isolated mitochondria synthesize proteins
                It is well known that isolated mitochondria synthesize proteins continuously.  Obviously they can only synthesize proteins coded for in mtDNA. 

                For allotopic expression to be successful in nerve cells, you would need to show that synthesis of all proteins by mitochondria was completely unnecessary, over the lifetime of a mitochondrion as it moved out the axon, and then back to the cell body, over its lifetime of about a month. 

                Yes, the respiratory chain complexes have nuclear coded components, but perhaps the only "bits" that get "damaged", are the "bits" the mitochondria can replace.  Maybe that is why all eukaryotes have the same 13 mtDNA coded proteins.
                Rate this comment: 12345
                Guest (Dave Whitlock)
                07/18/2006
                Posts:1
                • mt-coded proteins
                  I'd like to hear your hypothesis for how some subunits of an enzyme can get damaged when ones that they're right next to are immune. Not all eukaryotes have the same 13 - Chlamydomonas has only 7 of our 13, most plants have a lot more than us. A perfectly sufficient reason why all animals have the same 13 is that they all have a divergent genetic code that prevents any further transfer. This is all in my publications.
                  Rate this comment: 12345
                  Guest (Aubrey de Grey)
                  07/18/2006
                  Posts:1
                  • It is well known nitration on tyrosine is regulatory
                    of many proteins inside mitochondria.  Nitration of tyrosine can be quite specific, and can have quite specific effects. 

                    Mol Aspects Med. 2004 Feb-Apr;25(1-2):125-39.

                    Biochem Soc Trans. 2005 Dec;33(Pt 6):1399-403.

                    J Biol Chem. 2003 Sep 26;278(39):37223-30. Epub 2003 Jul 11.

                    Proc Natl Acad Sci U S A. 2001 Oct 9;98(21):12056-61. Epub 2001 Oct 2

                    This paper shows quite clearly regulation of mtDNA coded genes regulating metabolic activity in retinal cells.  The time constant for inducing light resistance is about a day.  Far faster than the turnover of mitochondria. 

                    http://www.iovs.org/cgi/content/full/45/8/2489
                    Rate this comment: 12345
                    Guest (Dave Whitlock)
                    07/18/2006
                    Posts:1
                    • How does this argue against AE?
                      Your point about nitration of tyrosine is unclear to me: it is a post-translational event, which can therefore take place as normal whether the protein was synthesised in the mitochondrion or imported. Please elaborate.

                      The IOVS paper you reference does not, as far as I can see, say anything relevant to the feasibility of allotopic expression. The experiment describes the ability of eyes accustomed either to dim or to bright light to cope with very bright light. What are you inferring from this?
                      Rate this comment: 12345
                      Guest (Aubrey de Grey)
                      07/19/2006
                      Posts:1
                      • mitochondria regulation via protein synthesis
                        The coping of retinal cells to different levels of light occurs through differential expression of proteins encoded by mtDNA.  If mitochondria are remote from the nucleus, this control paradigm is unavailable for mitochodria without mtDNA. 

                        If nitration produces irreversible inhibition of proteins, that inhibition can only be "reversed" via protein synthsis.  If mitochondria are unable to synthesize proteins, they lose the ability to recover from this regulatory mechanism. 

                        The ongoing protein synthesis by mitochondria is (to me) pretty good evidence that ongoing protein synthesis is necessary for proper function.  Allotopic expression in nerve cells requires that ongoing protein synthesis by mitochondria is completely unnecessary. 
                        Rate this comment: 12345
                        Guest (Dave Whitlock)
                        07/19/2006
                        Posts:1
                      • mitochondria regulation via protein synthesis
                        The coping of retinal cells to different levels of light occurs through differential expression of proteins encoded by mtDNA.  If mitochondria are remote from the nucleus, this control paradigm is unavailable for mitochodria without mtDNA. 

                        If nitration produces irreversible inhibition of proteins, that inhibition can only be "reversed" via protein synthsis.  If mitochondria are unable to synthesize proteins, they lose the ability to recover from this regulatory mechanism. 

                        The ongoing protein synthesis by mitochondria is (to me) pretty good evidence that ongoing protein synthesis is necessary for proper function.  Allotopic expression in nerve cells requires that ongoing protein synthesis by mitochondria is completely unnecessary for optimum and long term mitochondrial function. 
                        Rate this comment: 12345
                        Guest (Dave Whitlock)
                        07/19/2006
                        Posts:1
                        • Many routes to the same regulatory result
                          OK, thanks.  Taking those one at a time:

                          The differential expression in retina, if meaningful at all, cannot be only of the proteins discussed, because the stoichiometry of the subunits in a given enzyme complex is fixed. It may be that (to use examples from the paper) COIII is regulated by mRNA degradation rates and COII and COI by translation rates, but the overall effect must be (to an approximation that the cell can handle) equivalent. And moreover, the same must be true of the nuclear-coded subunits - whose regulatory DNA is available to us to use in the engineered transgenes. It is of course possible that the coding sequences themselves are important for regulation, but that's not a common finding in biology, so hardly a reason not to try. And on top of all that, the null hypothesis is that all will be fine unless some subunit is made in inadequate quantities: making some in excess should be harmless, as it will just be degraded by (e.g.) Lon when unused.
                          Rate this comment: 12345
                          Guest (Aubrey de Grey)
                          07/19/2006
                          Posts:1
                        • Nuclear-coded proteins are nitrated too
                          As to irreversible inactivation via nitration, I've only checked the abstracts of the papers you list but I notice that the JBC and PNAS ones identify a host of nuclear-coded proteins as particularly susceptible. This rather strongly implies that mitochondrial protein import can keep up. Circumstantially, it's hardly likely that 13 out of over 1000 proteins would be substantially more in need of continuous synthesis than any of the others.
                          Rate this comment: 12345
                          Guest (Aubrey de Grey)
                          07/19/2006
                          Posts:1
                          • It depends on the organ and cell shape
                            Organs with nearly spherical cells, and mitochondria close to ribosomes and the nucleus, presumably can continue to import nuclear coded proteins into mitochondria.  Neurons can't. 

                            If AE can't work for neurons, SENS can't work for neurons either.  It takes a long time (couple of weeks?) for mitochondria to be transported out an axon to the tippy end of a motor neuron. 

                            Maybe you can redesign mitochondria to use only imported proteins in neurons, but that won't be done using the "same" proteins, and is a completely different degree of difficulty than what you are proposing.  It is many orders of magnitude more difficult.
                            Rate this comment: 12345
                            Guest (Dave Whitlock)
                            07/19/2006
                            Posts:1
                            • But how fast are neuronal mt-coded proteins replaced?
                              Certainly AE has to work in all cell types to be fully effective -- though I should mention that a good case can be made that it really only needs to work in skeletal muscle, where short segments of fibres are affected, because in other cell types the cell can simply be ablated and replaced. But you have not provided any evidence that turnover of mt-coded proteins in neurons in vivo is in fact much faster than a few weeks. If you have such references, please provide them.
                              Rate this comment: 12345
                              Guest (Aubrey de Grey)
                              07/20/2006
                              Posts:1
                              • I know - but how fast is neuronal mt protein synthesis in vivo?
                                No, of course there was no import - I was just saying that since there was no import we have no answer in that paper to the question of interest, namely whether import can keep up with synthesis. In neurons with very long axons, it may indeed be that the mitochondria at the tip get along with no proteins, or else that an adequate supply of proteins (or transcripts) is transported along the axon separately from the mitochondria so that all the mitochondrion needs to do when it needs proteins is turn on import. I think the former (the mitochondrion manages with the proteins it has, then gets transported back to the cell body for degradation or division) is much more likely, given the very large number of proteins involved. Mitochondrial biogenesis is normally perinuclear (eg J Cell Biol 135:883). Again: you have not provided any evidence that turnover of mt-coded proteins in neurons in vivo is in fact much faster than a few weeks. If you have such references, please provide them.
                                Rate this comment: 12345
                                Guest (Aubrey de Grey)
                                07/21/2006
                                Posts:1
                        • is protein synthesis really "ongoing"?
                          Now, you could (and do) note the equally powerful circumstantial argument that ongoing protein synthesis wouldn't happen for no reason. But is it really ongoing? I suspect that all we know happens is protein synthesis unsynchronised with the cell cycle. That doesn't tell us that there is an ongoing need for supply of the mt-coded proteins at a faster rate than the nuclear-coded ones, only that some mitochondria in a cell are dividing (hence both synthesising and importing proteins) at any given moment. If you know of any references that individual mitochondria perform continuous protein synthesis, please provide them.

                          Again: I didn't invent the idea of AE. Rather a lot of top mitochondriologists were interested enough in it to work on it a decade ago, and would still be doing so if grant review study sections were less conservative. That's pretty good evidence that there's no obvious reason why it's doomed.
                          Rate this comment: 12345
                          Guest (Aubrey de Grey)
                          07/19/2006
                          Posts:1
                          • Protein synthesis occurs in isolated mitochondria
                            It has nothing to do with the cell cycle or mitochondria synthesis.  Isolated mitochondria synthesize proteins, differently depending on their recent metabolic history. 

                            J Appl Physiol. 2000 May;88(5):1601-6
                            J Biol Chem. 1990 May 5;265(13):7532-8.
                            This last one, they quite clearly show protein synthesis in isolated mitochondria regulated by membrane potential and inhibited by chloramphenicol.

                            Since mitochondria synthesize some components of the respiration chain, and import others, mitochondria must have a way of assembling those various components. 

                            Different components of the same respiration chain complex have different lifetimes.

                            J Biol Chem. 1982 Apr 10;257(7):3575-80
                            Rate this comment: 12345
                            Guest (Dave Whitlock)
                            07/19/2006
                            Posts:1
                            • That paper too shows nuclear genes can keep pace
                              Yes, recent metabolic history determines synthesis rate - but also import rate. Import is limited by availability of the proteins, but that's why there's feedback to non-mt transcription and translation. Protein synthesis in isolated mitochondria is of course regulated, but since protein import cannot be measured in that experiment we are told nothing about the relevant question (whether nuclear genes can substitute for mitochondrial genes).

                              As for the JBC 1982 paper, be careful. The only complex with heterogeneous turnover rate was Complex III, and even there the sole mt-coded subunit (cytochrome b) had the same rate as three nuclear-coded ones. Bear in mind also that this experiment would have included subunits never assembled into complexes, so the differences seen may be some subunits being synthesised/imported in excess and promptly degraded. Such modest inefficiency of regulation is no surprise in a dysregulated cell population such as a hepatoma line in vitro.
                              Rate this comment: 12345
                              Guest (Aubrey de Grey)
                              07/20/2006
                              Posts:1
  • Impossibly hard versus trivially simple
    All physiological effects come from many coupled non-linear pathways.  You can't "factor out" the "bad ones" without understanding which ones are "bad", and which ones are "good".  Which ones are "bugs", and which ones are "features".  It may be that they are all "features". 

    Turning off mitochondria is "bad", but it is "better" than malignant hyperthermia. 

    Rate this comment: 12345
    Guest (Dave Whitlock)
    07/17/2006
    Posts:1
    • Medicine as an existence proof
      Dave, you appear to be holding to the irreducibility of biological complexity as a bit of an article of faith. Does your logic not imply that medicine in general ought to be impossible? How did Jenner and Pasteur save so many lives with their virtually nonexistant understanding of the immune system, if any intervention is "bad"?
      Rate this comment: 12345
      Guest (Aubrey de Grey)
      07/18/2006
      Posts:1
      • No, I am realistic about the real complexity
        Jenner and Pasteur saved lives by relying on the functionality that was already in the biological systems.  The functionality that you want to "sidestep".

        Medicine seeks to restore "normal" functionality.  You seek to operate an essentially unknown system outside its normal range, achieve superior functionality, without understanding the "details". 

        I am not saying that what you propose is "impossible".  You are saying it can be done in 10 years with modest funding.  A funding level that is on the same order as that to bring a new simple drug to market.

        I think you understimate the degree of difficulty by at least 6 orders of magnitude. 
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/18/2006
        Posts:1
        • SENS also relies on existing functionality
          No - preserving function is not the same as operating a machine outside its normal range. Vintage cars exist not because they were built to last 100 years, nor because they are simple enough that their function can be enhanced, but because preservation of function is much easier than enhancement. SENS is about preserving function by repairing or obviating damage.

          Also, I'm only saying it can be done in **mice** in 10 years with modest funding.
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/18/2006
          Posts:1
          • Huh?  preserve function by "sidestepping" ignorance?
            You are not proposing to "preserve" function, you are proposing to replace functions with "engineered" solutions, when the myriad details of those functions remain unknown. 

            Your "car" analogy is inappropriate.  Mechanical objects do not deteriorate when stored properly, or when inactive for repair.  Every component of a mechanical system can be replaced independently.  The "repair system" (mechanics + tools + information + spare parts) is external to the car and is subject to unlimited modification with no effect on the car.  The "details" of the car are completely understood.
            Rate this comment: 12345
            Guest (Dave Whitlock)
            07/18/2006
            Posts:1
            • Not replacing functions, providing them by an extra route
              No, I'm not proposing to replace functions. The proteins that would be expressed by nuclear copies of mitochondrial genes, for example, are the same proteins as are naturally expressed by the mitochondrial genes. I didn't invent the idea of putting copies of these genes in the nucleus - it's been proposed for over 20 years as a therapy for mitochondrial diseases. If it's a reasonable way to treat mitochondriopathies, why should it not be reasonable for aging?
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/18/2006
              Posts:1
            • Cars and organisms
              Vintage cars are not inactive - that's my point, they are driven regularly and the damage they incur as a result is repaired as fast as it occurs. Replacing components is indeed what we will do with the body, at the molecular and cellular level, so again the analogy is accurate. You're quite right that there's one big difference, namely that we understand the design of the car in full but not the body - but all that means is that SENS would be unrealistic if it didn't have our natural homeostatic mechanisms to ride on the coat-tails of.
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/18/2006
              Posts:1
              • There are many big differences.
                Vintage cars are inactive during repair.  The repair time can be days, weeks, months, years or decades during which time the repair process can be augmented to an arbitrary degree.  Human cells are only inactive when they are dead, after which time they cannot be repaired by any means.  There is no amount of damage to a vintage car that cannot be "repaired".  

                When vintage cars are taken out of service, continued damage stops.  Most of the damage that occurs to humans happens after the initial infarct, as in ischemia-reperfusion injury. 

                Many of the "damage" mechanisms in humans are actually "features" that are protective under other circumstances.  Apoptosis, shock, inflammation, excitotoxic death.   How will you regulate those without understanding when they are good and when they are bad?
                Rate this comment: 12345
                Guest (Dave Whitlock)
                07/18/2006
                Posts:1
                • SENS only targets initially inert damage
                  SENS doesn't propose to regulate any of the features you list.  As my website and publications make quite clear, the types of damage that SENS focuses on are strictly those that accumulate during life as side-effects of metabolism and are eventually pathogenic. The reason this is a feasible strategy is that none of these types of accumulating damage are harmful at low levels, only when they accumulate to levels seen in the second half of life, i.e. they are initially inert. We know this because 40-year-olds who have lived the way their mothers told them to exhibit very little loss of function.
                  Rate this comment: 12345
                  Guest (Aubrey de Grey)
                  07/19/2006
                  Posts:1
          • What about mice is so "simple"?
            What makes mice orders of magnitute simpler than humans?  Their genome isn't orders of magnitude smaller. 

            Metabolicly active cells generate free radicals as necessary components of their metabolic activity.  Many transcription factors are regulated by redox active free radicals.  How do you propose to repair or obviate the "damage" done by free radicals without understanding what is "damage", and what is a "signal" for transcription factors to initiate "repair"?
            Rate this comment: 12345
            Guest (Dave Whtilock)
            07/18/2006
            Posts:1
            • Simpler, yes (tohugh not that much)
              Mice are a lot easier because (a) they're not naturally as good as us at not aging, (b) we can try things out on mice that are riskier than we would do on humans, and (c) success will be quicker to validate. Mice are certainly not orders of magnitude easier, but then I didn't agree with you that I was underestimating the difficulty by orders of magnitude.
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/18/2006
              Posts:1
              • If you give mice a human lifespan Success??
                Suppose you did give a mouse a human-type lifespan?  Is that success?  What if you did so simply by transposing human lifespan systems into mice?  There is nothing "new" that can be transposed back into humans. 

                Mice with a lifespan of 80 years would be a clever parlor trick, but one that would take 80 years to demonstrate. 

                So how do you "validate" success using mice?   If you exceeded a  human lifespan you would certainly be on to something, but that will take 80+ years to demonstrate.   Something less than 80 years might not be as good as what humans already have. 
                Rate this comment: 12345
                Guest (Dave Whitlock)
                07/18/2006
                Posts:1
                • validation - timeframes and purpose
                  There are two things wrong with what you say here.  First, since SENS is about repair (rejuvenation) therapies rather than pre-emption, success can be validated in a small fraction of the organism's normal lifespan because treatment begins in middle age or later. Second, validation is not proof but it is evidence sufficient to proceed to other animals and then humans - in this respect, as in most, SENS is just the same as any other new medical intervention.
                  Rate this comment: 12345
                  Guest (Aubrey de Grey)
                  07/19/2006
                  Posts:1
            • Free radicals should indeed be left alone
              Your point about free radicals is quite correct and is exactly why I don't advocate using antioxidants as a life-extension protocol. What I propose to target instead, as you correctly say, is the damage done by free radicals - such as mtDNA mutations. I propose to leave the free radicals alone.  Again: the point of SENS is to leave metabolism alone, and to target only the initially inert consequences of metabolism.  We know that mtDNA mutations, lysosomal aggregates etc are not part of metabolism because young people get along fine without them.
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/18/2006
              Posts:1
  • There was no import of proteins!
    The 1990 JBC paper was on isolated mitochondria.  There were no external proteins for the mitochondria to import!  Mitochondria at the tippy end of an axon can't import proteins because the source of the proteins, the cell body, is a meter away.  It takes many days, usually weeks for cargo to be moved from the cell body out a 1 meter axon.  You are saying that it is ok for the mitochondria to send a transcription factor back to the cell body, wait a week for it to get to the cell body, then wait another week for the replacement proteins to arrive?

    The time constant for mtDNA protein synthesis is < 1 minute.  For SENS to work in motor neurons, there must be no adverse effects from increasing that time constant by 5 orders of magnitude, from 1 minute to 1 week. 
    Rate this comment: 12345
    Guest (Dave Whitlock)
    07/20/2006
    Posts:1
    • I know - but how fast is neuronal mt protein turnover in vivo?
      No, of course there was no import - I was just saying that since there was no import we have no answer in that paper to the question of interest, namely whether import can keep up with synthesis. In neurons with very long axons, it may indeed be that the mitochondria at the tip get along with no proteins, or else that an adequate supply of proteins (or transcripts) is transported along the axon separately from the mitochondria so that all the mitochondrion needs to do when it needs proteins is turn on import. I think the former (the mitochondrion manages with the proteins it has, then gets transported back to the cell body for degradation or division) is much more likely, given the very large number of proteins involved. Mitochondrial biogenesis is normally perinuclear (eg J Cell Biol 135:883). Again: you have not provided any evidence that turnover of mt-coded proteins in neurons in vivo is in fact much faster than a few weeks. If you have such references, please provide them.
      Rate this comment: 12345
      Guest (Aubrey de Grey)
      07/21/2006
      Posts:1
      • mtDNA expression regulated local ATP demand
        Increased ATP demand incrases expression of neuron mtRNA in 15 minutes.  Journal of Neurochemistry, 2005, 93, 850–860

        Expression of nDNA and mtDNA coded proteins is regulated differently by local ATP demand.  nRNA is confined to the cell body, complex IV activity correlates with expression of mtDNA coded COX III, not nDNA coded COX Vb.  THE JOURNAL OF COMPARATIVE NEUROLOGY 373: 139-155 (1996).

        mt coded and nuclear coded subunits of complex IV are disproportionately regulated by local ATP demand.  Complex IV activity correlates with mtDNA coded mRNA and COX I (mt coded) and not COX IV or COX VII (n coded). The Journal of Neuroscience, May 1993. 13(5): 1805-1819
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/22/2006
        Posts:1
      • This clearly shows AE and SENS will not work for neurons
        These 3 papers very clearly show local regulation of complex IV activity based on local ATP demand, due to local production of mtDNA coded proteins with time constants very short compared to transport from the nucleus. 

        Limiting the quantitiy of active complex IV in respiring mitochondira is an important control feature that equalizes O2 consumption between many mitochondria, matching O2 diffusion and consumption with ATP demand.  Attempting to substitute mitochondria fully loaded with completely active complex IV would not work as a substitute. 
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/22/2006
        Posts:1
        • Wrong again - not physiological
          Again you seem to be resorting to citing grossly non-physiological in vitro demonstrations of the rate at which mitochondrial protein synthesis *can* respond to stimuli and blithely inferring that in real life it actually *does* need to respond in this way but protein import does not need to. Re your first citation, there is not much monensin in the brain. Your second and third show only that nuclear-coded subunits can indeed be synthesised in excess with no deleterious consequences. Note that CO is regulated allosterically by ATP levels (see Eur J Biochem 249(1):350 and several later refs from the same lab), so there is no reason to suppose that there would be any problem with having all its subunits nuclear-coded.
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/23/2006
          Posts:1
          • No, you misread the papers they are in vivo!
            THE JOURNAL OF COMPARATIVE NEUROLOGY 373~139-155 (1996)

            The Journal of Neuroscience, May 1993. 13(5): 1805-1819

            are both in vivo demonstrations of cytochrome oxidase activity and correlated by immunoassay with quantity of holoenzyme, correlating with mtDNA.  IN VIVO!

            It is only the non mtDNA coded subunits that can be in excess.  Those are metal-free.  It is the mt synthesized units that are the active site, and which contain Cu and heme. 
            Rate this comment: 12345
            Guest (Dave Whitlock)
            07/23/2006
            Posts:1
            • I only said your first ref was in vitro, not those two
              I didn't say these refs (your second and third in your previous post) were not in vivo, only your first ref. For the second and third, the problem with your inference is that they only show that nuclear-coded subunits can be regulated less precisely than mt-coded ones tend to be.
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/24/2006
              Posts:1
          • Cu and Fe &nbsp;are Fenton active metals!
            The mt inner matrix has the highest H2O2 concentration in the cell.  A few % of O2 consumed ends up as H2O2 in the inner matrix.  That H2O2 will form hydroxyl radical.  "Extra" Cu and Fe cannot be stored in the inner matrix. 
            Rate this comment: 12345
            Guest (Dave Whitlock)
            07/23/2006
            Posts:1
            • Frataxin etc keep Fe/Cu safe, not the CO subunits
              Cu and Fe are indeed redox-active, but that's why mitochondria have frataxin and other systems for keeping them safe. The respiratory chain subunits aren't the way that's done.
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/24/2006
              Posts:1
          • monensin or glutamate caused ATP depletion!
            it was the ATP depletion that triggered mtDNA transcription.  They showed that glutamate did the same thing, as did 2-deoxy-D-glucose.  ATP depletion is not "non-physiologic". 

            The allosteric regulation of CO by ATP doesn't change the amount of CO present. 
            Rate this comment: 12345
            Guest (Dave Whitlock)
            07/23/2006
            Posts:1
            • Really acute ATP depletion *is* non-physiological
              Of course ATP levels fluctuate in vivo, but not so dramatically as monensin would cause. Absolutely, allosteric regulation doesn't change CO abundance - but it does change CO activity, which means that mitochondria can tolerate having more CO than it needs.
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/24/2006
              Posts:1
              • Well DUH!! of course ATP doesn't fluctuate in vivo
                Because and ONLY BECAUSE mitochondria control it!

                Take out that control loop, which AE does, and ATP is not as well controlled. 
                Rate this comment: 12345
                Guest (Dave Whitlock)
                07/24/2006
                Posts:1
                • Post-translational control clearly dominates
                  ATP levels are controlled by mitochondria, yes:

                  - chemiosmosis itself
                  - UCP opening/closing
                  - allosteric regulation

                  The idea that the loss of a far less flexible additional control loop (less flexible because it relies on protein synthesis) would be so clearly fatal to mitochondrial function that AE is a non-starter is, to say the least, not a logical conclusion.

                  Mitochondria include over 1000 proteins.  Your objection to AE relies on the proposal that neuronal mitochondria can function perfectly well importing all but 13 of those proteins from the cytosol but would be fatally affected if they had to import the other 13 too. That is a priori an extremely implausible hypothesis, and non-physiological data aren't much use in strengthening it. Yet you're using that logic as a reason not to try to develop a key component of a therapy that could defeat aging and thus save 100,000 lives a day. Where's your sense of proportion?
                  Rate this comment: 12345
                  Guest (Aubrey de Grey)
                  07/24/2006
                  Posts:1
                  • No, each of those 1000+ proteins is unique
                    They cannot be treated "independently". 

                    You assert that SENS via AE provides all the functionality of mitochondria.  That understanding the "details" are unimportant because AE can replace them all.

                    I have shown that is clearly false by presenting data where mtDNA is essential in providing a regulatory function that AE cannot possibly supply.
                    Rate this comment: 12345
                    Guest (Dave Whitlock)
                    07/25/2006
                    Posts:1
                  • SENS via AE will not work!
                    You are deluding yourself if you still think it will.  You know that there are enormous gaps in our knowledge of how mitochondria are controlled.  If any of those control systems cannot be replicated with 100.0000% fidelity via AE, then the "details" have to be understood, and replacement control systems engineered to replace those that are lost via AE.
                    Rate this comment: 12345
                    Guest (Dave Whitlock)
                    07/25/2006
                    Posts:1
                  • Sense of proportion? &nbsp;Huh? &nbsp;what is that about?
                    Your sense of proportion has clouded your judgement.

                    Your premise of SENS is that AE can replace all mitochondrial functions with no loss in functionality, so we can "sidestep" our ignorance.  I demonstrate a functionality that AE cannot provide, you simply assert that functionality is unimportant because other things are more important.  Huh?  Unless we understand how all of the thousands of coupled non-linear systems interact, we can't really say which (if any) are so unimportant they can be discarded.
                    Rate this comment: 12345
                    Guest (Dave Whitlock)
                    07/25/2006
                    Posts:1
                  • The hypothesis to be proven is SENS via AE
                    To disprove SENS via AE, I simply need to show that there is mitochondria functionality that cannot be effected via AE. 

                    I have shown that.  It is no longer an a priori situation.

                    You must now show that every functionality that cannot be provided by AE is unimportant and will have no adverse effect(s) on cell metabolism under any circumstance(s). 
                    Rate this comment: 12345
                    Guest (Dave Whitlock)
                    07/25/2006
                    Posts:1
                  • A small flaw could well be fatal
                    Mitochondria produced in the cell body must travel out the motor neuron axons to supply ATP to the tippy end perhaps a meter away.  If the ATP an individual mitochondrion can produce over its lifetime is insufficient for that journey, it will stop, clog up the axon, and the distal end will necrose.  If the round trip takes 30 days, and a mitochondria only lasts for 29, the end of that motor neuron axon will necrose.  Without growth factors from the tippy end, the whole axon will degenerate with the longest axons degenerating first.  
                    Rate this comment: 12345
                    Guest (Dave Whitlock)
                    07/25/2006
                    Posts:1
                    • Still no evidence...
                      There's nothing new in your latest series of comments. As I've noted ad nauseam, you have NOT shown ANY cases where mtDNA is essential in providing a regulatory function. You've shown cases where mtDNA protein synthesis is stimulated by microenvironmental factors, and you've identified some grossly non-physiological cases where mt transcription is regulated somewhat more rapidly than nuclear, but you have not identified a single shred of evidence that mt-coded proteins are generated faster than at least some nuclear-coded proteins of the same complex, in any physiologically relevant circumstance. The best you can possibly claim to have shown is that there are certain hypothetical circumstances in which there is an outside chance that AE will not work quite as well as the natural arrangement. That's what I meant about sense of proportion: you're treating an outside chance as a sufficient certainty that AE is not even worth trying.
                      Rate this comment: 12345
                      Guest (Aubrey de Grey)
                      07/25/2006
                      Posts:1
                      • No! you have sold AE as "sidestepping" our ignorance
                        I have shown that to be a false portrayal of the current state of the art in our knowledge of mitochondrial function.  You have said that all mitochondrial functionality can be expressed via AE.  I have shown that to be false.  You have stated we don't need to understand mitochondria because they are self-regulating without mtDNA expression.  I have shown that to be false. 
                        Rate this comment: 12345
                        Guest (Dave Whitlock)
                        07/25/2006
                        Posts:1
                        • Let's see what the referees think
                          I claim you've shown no such thing. I publish a lot of work that challenges my own views, so you can be sure I'll select highly knowledgeable, unbiased referees and respect their assessment. RR's impact factor is 8.571 so I presume you will benefit from publishing in it. Well?
                          Rate this comment: 12345
                          Guest (Aubrey de Grey)
                          07/25/2006
                          Posts:1
                      • Only cause of cell death is ATP depletion
                        and apoptosis.  Every single cell that dies will pass through the ATP level that you consider to be "non-physiologic".  Cells that pass through that level with mtDNA can express it and survive.  Nerve cells without mtDNA can't, and will die while waiting a week for AE proteins from the nucleus. 
                        Rate this comment: 12345
                        Guest (Dave Whitlock)
                        07/25/2006
                        Posts:1
                        • Sounds like a testable hypothesis
                          ... for your manuscript...
                          Rate this comment: 12345
                          Guest (Aubrey de Grey)
                          07/25/2006
                          Posts:1
                          • not "hypothesis", definition of cell death
                            Sort of like "heart failure" as a working definition of organism death.  When ATP drops below a certain level, it induces apoptosis, below that necrosis. Am. J. Physiol. 274 (Renal Physiol. 43): F315–F327, 1998.  When cells cannot sustain their ATP level they are, or soon will be dead.  If a cell can sustain its ATP level, it is still "alive", even if it does nothing else.

                            Every cell that "dies" will go through the ATP level that you consider so "non-physiological" that it can be ignored.  An obviously false assumption. 
                            Rate this comment: 12345
                            Guest (Dave Whitlock)
                            07/26/2006
                            Posts:1
                            • Relevance of AE is the hypothesis
                              Yes obviously - but what you have not shown (despite your claims to the contrary) is that the regulation of ATP level would be more than negligibly impaired if all mt proteins were made from nuclear genes, i.e. that AE would appreciably increase (or decrease) the rate of cell death.
                              Rate this comment: 12345
                              Guest (Aubrey de Grey)
                              07/26/2006
                              Posts:1
                              • SENS via AE by "sidestepping ignorance" is the hypothesis
                                You have asserted that each and every control function (known and unknown) of mt can be accomplished via AE, and that understanding the details of mt function and control are completely unnecessary. 

                                I have shown a control function that requires mtDNA expression, and which cannot be accomplished via AE. (an existance proof). 

                                You then waved your hands and said those ATP levels are unimportant.  I then showed that all cells pass through those ATP levels during their lifetime. 

                                You are the one asserting that functionality is so unimportant under each and every circumstance, that it can be ignored.  Do you have some data to back up that assertion? 
                                Rate this comment: 12345
                                Guest (Dave Whitlock)
                                07/26/2006
                                Posts:1
                                • Summary, part 1
                                  > You have asserted that each and every control function (known and
                                  > unknown) of mt can be accomplished via AE, and that understanding the
                                  > details of mt function and control are completely unnecessary.

                                  Correct.

                                  > I have shown a control function that requires mtDNA expression, and
                                  > which cannot be accomplished via AE. (an existance proof).

                                  Incorrect. None of your cited data demonstrate the need for any protein to be expressed from mtDNA.
                                  Rate this comment: 12345
                                  Guest (Aubrey de Grey)
                                  07/26/2006
                                  Posts:1
                                • Summary, part 2
                                  > You then waved your hands and said those ATP levels are unimportant.  I
                                  > then showed that all cells pass through those ATP levels during their
                                  > lifetime.

                                  Incorrect. I said that the method of induction of those ATP levels was not physiological, and I also said that the location of the genes that control ATP level is of negligible relevance to the sensitivity of that control, and that none of the data you've cited demonstrate otherwise. I still say that.

                                  > You are the one asserting that functionality is so unimportant under
                                  > each and every circumstance, that it can be ignored.  Do you have some
                                  > data to back up that assertion?

                                  Sure. Let's start with the Wong-Riley paper you cited: it showed that in each and every relevant enzyme complex some nuclear-coded proteins have the same turnover rate as the mt-coded ones.
                                  Rate this comment: 12345
                                  Guest (Aubrey de Grey)
                                  07/26/2006
                                  Posts:1
                                • Conclusion: put up or shut up
                                  I say again: if you're so sure of your position, WRITE IT UP.
                                  Rate this comment: 12345
                                  Guest (Aubrey de Grey)
                                  07/26/2006
                                  Posts:1
                                  • read another Wong-Riley paper
                                    The Journal of Neuroscience, September 1994. 14(9): 5338-5351
                                    Rate this comment: 12345
                                    Guest (Dave Whitlock)
                                    07/27/2006
                                    Posts:1
                                    • looks like more support for the feasibility of total AE
                                      Very interesting paper. Do you allege that it supports your proposition that AE of all 13 proteins cannot work in neurons? If so, please elaborate.

                                      By the way, when I said "Wong-Riley paper" earlier I actually meant the Hare and Hodges 1982 JBC paper - my apologies.
                                      Rate this comment: 12345
                                      Guest (Aubrey de Grey)
                                      07/27/2006
                                      Posts:1
                                      • It shows local control via mtDNA expression
                                        They show accumulation of precursor COIV in mitochondria (and only mitochondria) and local conversion to hCOX in response to local ATP demand. In conjunction with their other papers, that conversion is due to local expression of mtDNA. 

                                        This shows expression of mt and n COX subunits in both different time and space.  pCOIV is stable long term and makes hCOX when mtDNA is expressed.  For AE to work, pCOs from mt would need to be stable for weeks instead of minutes, and "triggerable" in response to local ATP demand.  
                                        Rate this comment: 12345
                                        Guest (Dave Whitlock)
                                        07/27/2006
                                        Posts:1
                                        • Nonsense - not via, parallel with
                                          There is control of mtDNA expression and there is control of nuclear/cytoplasmic expression and/or import, but there is no demonstration (or even suggestion) that mtDNA expression is upstream of nuclear expression (or of import) in any of these papers. Thus, there remains every reason to suppose, as the null hypothesis, that attaching the COIV leader to a COI modified to overcome the hydrophobicity barrier would confer on it the same regulation seen for COIV in this paper.  If you disagree, spell out the causal chain of events - no more scattergun.
                                          Rate this comment: 12345
                                          Guest (Aubrey de Grey)
                                          07/27/2006
                                          Posts:1
                                          • No, mtDNA is downstream but asynchronous
                                            Wong-Riley 94 clearly shows accumulation of excess pCOIV in mit and transport to axon terminals prior to formation of hCOX.  They cite work showing lifetime in cytoplasm of precursor mit proteins is short (<5 min) (I haven't read those papers yet).  That short a lifetime precludes import over long distances.  If pCOI were AE, hCOX would form immediately upon import (in cell body), precluding de novo formation of hCOX from pCOIV and new COI,II,III at sites remote in time and space as observed in response to neural activity (Wong-Riley 96,93 Chandrasekaran 2005).
                                            Rate this comment: 12345
                                            Guest (Dave Whitlock)
                                            07/27/2006
                                            Posts:1
                                            • no - hCOX assembly needs pCOIV maturation
                                              Of course not, because hCOX would not form until all the subunits were available as **matured** precursors, and since these mitochondria have a system for sequestering pCOX from maturation (and thus assembly), that would prevent assembly. It's true that pCOI **might** be matured and cause trouble by sitting around unassembled, but it's much more likely that the method for preventing pCOIV maturation would also work on other proteins that have the same leader sequence, which is the proposal for AE COI. Sequestration in the intermembrane space is an obvious candidate, and that would almost certainly be regulated by the leader sequence.
                                              Rate this comment: 12345
                                              Guest (Aubrey de Grey)
                                              07/28/2006
                                              Posts:1
                                              • No. unincorporated pCOI is rapidly degraded
                                                There is no unincorporated mt coded protein that is retained in precursor form.  Either they are incorporated into holo enzyme, or they are degraded. 

                                                The leader sequence is much much shorter than pCOI.  There are only 2 mechanisms for import of mt proteins, electrophoresis and chaperone mediated folding in the matrix.  Proteins that span the outer and inner membrane require the proper sequence along the whole length, not just the leader.  mt coded proteins had no need to evolve the proper sequence for import.  There is no reason to suppose that they posess it.  
                                                Rate this comment: 12345
                                                Guest (Dave Whitlock)
                                                07/30/2006
                                                Posts:1
                                                • How do you know, since pCOIV is evidently not degraded?
                                                  Sigh - of course the dogma is that precursors are not retained, but the very case you're focusing on is an exception!  And if neuronal mitochondria have a trick to hold pCOIV in an unmatured (hence necessarily unincorporated) state, the chances are very good that they'll do the same for any other protein that has the same leader. Import is independent of sequence other than the leader, so long as it's not too hydrophobic.
                                                  Rate this comment: 12345
                                                  Guest (Aubrey de Grey)
                                                  07/31/2006
                                                  Posts:1
                                                  • No, import is not independent of cargo
                                                    The leader sequence is up to ~70 AA long.  pCOX1 is 514 AA long.  A single AA substitution in protoporphyrinogen oxidase (477 AA + fluorescent P) (Val 335 to Gly) completely prevented mit import (Elder GH Biochem J 2004).  Proteins are chaperoned to mt, unfolded, pulled through the outer and inner membrane translocases and folded via ATP mediated processes.  When the whole protein is long enough to span the outer and inner membranes, the "driving force" for import is the folding that happens in the matrix.   When a protein spans both, it has to unfold at one end, generate slack, and then fold at the other end.
                                                    Rate this comment: 12345
                                                    Guest (Dave Whitlock)
                                                    08/01/2006
                                                    Posts:1
                                                    • That's why the IMS model is attractive
                                                      All true, but *initiation* of import through the inner membrane is determined by the leader sequence, not the mature sequence. So if pCOIV is being stored in the intermembrane space, the same system should work for other proteins such as COI.
                                                      Rate this comment: 12345
                                                      Guest (Aubrey de Grey)
                                                      08/02/2006
                                                      Posts:1
                                                      • Huh?&nbsp; what data suggests storage in intermembrane space?
                                                        The whole sequence is required for import into the inner matrix.  How do you propse getting AE proteins to the inner matrix if they are not imported?  Engineer a new pathway? 

                                                        Initiation isn't enough, the whole protein has to be imported. 

                                                        If proteins are stored in the intermembrane space, what triggers import to matrix? 
                                                        Rate this comment: 12345
                                                        Guest (Dave Whitlock)
                                                        08/03/2006
                                                        Posts:1
                                                        • Lack of degradation, of course
                                                          Something allows pCOIV in these neuronal mitochondria to avoid degradation. As you pointed out, this is unusual. I'm merely suggesting that maybe there is an uncharacterised system that closes off TIM23 in the cell body and opens it again when the proteins are needed. If you claim to have a more economical explanation, let's hear it.
                                                          Rate this comment: 12345
                                                          Guest (Aubrey de Grey)
                                                          08/03/2006
                                                          Posts:1
                                                          • lack of degradation implies protection
                                                            I think it more likely that it is imported into the matrix, and assembled into the membrane, but without the 3 mt proteins that make up the active site.  When those are needed, mRNA is synthesized, and the pCOI-III are synthesized in ribosomes adjacent to the membrane, so as they are produced, they enter the membrane.

                                                            The number of AA in mt coded CO subunits is about 2x than it N coded.  The intermembrane space would get crowded with 3x the volume stored there via AE. 

                                                            If there is no COX without COI-III, then the newly synthesized CO's go into the matrix and are degraded. 

                                                            When hCOX gets "tired", it comes out of the membrane into the matrix and gets degraded, making room in the membrane for more hCOX.  You could store several times more enzyme this way, which translates into several times longer mt lifetime.
                                                            Rate this comment: 12345
                                                            Guest (Dave Whitlock)
                                                            08/03/2006
                                                            Posts:1
                                                            • One more time: what about the other complexes?
                                                              Your're way off with your estimates of amino acids for CO - see my other reply. But more than that, even though conceivably pCOIV is incorporated into the membrane (because it's supposed to be an early subunit involved in assembly), by your model *all* the subunits would have to be similarly protected. Further, as I said earlier, your objection to AE doesn't rest only on the behaviour of CO: every nuclear-coded subunit of *every* respiratory chain complex, and indeeed every other nuclear-coded prtein with a comparable half-life, would have to have the same problem. There's no reason to suppose CO subunits would have a shorter half-life than those of other respiratory chain complexes - quite the opposite in fact, as CO makes no superoxide.
                                                              Rate this comment: 12345
                                                              Guest (Aubrey de Grey)
                                                              08/04/2006
                                                              Posts:1
                                                              • No, SENS via AE has to "work" for all complexes
                                                                and all subunits of all complexes under all achievable metabolic conditions in all organs and tissue compartments.

                                                                If AE *can't* work for for just one of those tissue compartments, SENS *can't* work either.

                                                                The problems of AE may well be greater for complex I.  I suspect it is.  Complex I is trickier to work with than CO, which is I think why there isn't as much data on it.  ND6 is coded separately from the other mt proteins.  It is probably "special". 
                                                                Rate this comment: 12345
                                                                Guest (Dave Whitlock)
                                                                08/04/2006
                                                                Posts:1
                                                                • Precisely - which blows away your 2:1 (or even 1:1) ratio
                                                                  Of course it has to work for all complexes.  So, work out your ratios for the other complexes too, and you'll see what I mean. Quite apart from the fact that you're still ignoring the leader sequences and the accessory proteins involved in assembly, as well as your own statement (in your other most recent post) that CO is probably longer lived than the others (so the mt needs to store more copies of the others, if the Wong-Riley mechanism applies to those complexes too).
                                                                  Rate this comment: 12345
                                                                  Guest (Aubrey de Grey)
                                                                  08/05/2006
                                                                  Posts:1
                                                                  • That makes SENS more difficult not less
                                                                    I don't understand your point.  The more that needs to be stored in the mt, the fewer copies can be stored.  mt with fewer copies will wear out faster and have to be replaced sooner.  That makes SENS more difficult.  If the mt lifetime is short enough, it makes it impossible. 
                                                                    Rate this comment: 12345
                                                                    Guest (Dave Whitlock)
                                                                    08/08/2006
                                                                    Posts:1
                                                                    • No - it's the PROPORTIONAL change in cargo that matters
                                                                      My point was that if (as I think you do) you assume that all n-coded subunits of all the respiratory chain complexes are mainained in neurons in the way that W-R reported for COIV, then the addition of 13 more proteins will not change the amount stored by a particularly large factor. You originally focused on CO and inferred that the factor was 2:1 mt:n, or (after my correction) 1:1, which might indeed be problematic, but if it's more like 1:3 or 1:4 then it's hard to get particularly worried -- most biological systems have easily that much flex.
                                                                      Rate this comment: 12345
                                                                      Guest (Aubrey de Grey)
                                                                      08/08/2006
                                                                      Posts:1
                                                  • There are no "exceptions"!&nbsp;
                                                    Each and every protein in mt is unique.  You are pretending there is a simple "rule" that they all follow (except for 1 "exception" (that we know about so far)).

                                                    If we are going to "sides step our ignorance", how do we "side step" the proteins not acting the way we need them to for our "side stepping" to work? 

                                                    Assume because we can get 99.9% of the proteins in there, that the 1 we can't doesn't matter?  Even if it is pCOXI? 
                                                    Rate this comment: 12345
                                                    Guest (Dave Whitlock)
                                                    08/01/2006
                                                    Posts:1
                                                    • But there are groups, because there are more proteins than mechanisms
                                                      CO has many other nuclear-coded subunits. So does the rest of the respiratory chain. Each of these subunits somehow gets provided to neuronal mitochondria in sufficient amounts. This may happen by constructing enough holoenzyme before axonal transport, or by transporting mRNA or protein up the axon separately from mitochondria, or by storing them as precursors in transported mitochondria (the pCOIV story). But those are the only options, so at least one of those options is certainly used by dozens of different proteins. Anything that is used by dozens of different proteins is pretty damned sequence-nonspecific, and thus likely to work for another protein or 12. So if pCOIV is the only protein that behaves as Wong-Riley describes, then sure, the chances of getting COI to work the same way are low - but that's fine, because we can just have COI work the same way as other nuclear-coded respiratory chain subunits.
                                                      Rate this comment: 12345
                                                      Guest (Aubrey de Grey)
                                                      08/02/2006
                                                      Posts:1
                                                      • No mt coded proteins are like N coded ones
                                                        In CO, there are 10 subunits, 3 mt of 514, 227 and 261 AA.  The 7 N coded subunits are 147, 84, 73, 56, 56, 47, 46.  They are all quite different. COI has 12 membrane spanning helixes.  The most any N coded subunit has is 1. 

                                                        The point of the Wong-Riley result is that pCOIV is sequestered, and hCOX is created de novo by local expression of mtDNA. 

                                                        If you can't sequester pCOI-III, you can't form hCOX de novo, and if you do, the sequestered pCOI-III takes up 2x the volume of the pCOIV-VII so your mt lifetime is 1/3 as long.

                                                        Either way, SENS via AE will not work.
                                                        Rate this comment: 12345
                                                        Guest (Dave Whitlock)
                                                        08/03/2006
                                                        Posts:1
                                                        • Wrong numbers
                                                          Way wrong:

                                                          - you missed three n-coded subunits of sizes 147, 109 and 84 -- see here:
                                                            http://www-bioc.rice.edu/~graham/CcO_sequences.html

                                                          - they all have leader sequences, which if the Wong-Riley result is typical
                                                            are not cleaved

                                                          - you also have to count all the accessory subunits involved in assembly

                                                          So even if we restrict ourselves to CO, the actual ratio is closer to 1:2 than 2:1. And of course your model *can't* be restricted to CO, because you have no evidence that CO holoenzyme has a shorter halflife than any other complex.
                                                          Rate this comment: 12345
                                                          Guest (Aubrey de Grey)
                                                          08/04/2006
                                                          Posts:1
                                                          • I used bovine CO numbers, mt are same N are not
                                                            The ratio is 2.0 for b, 1.2 for h.  1/2 mt lifetime unacceptable for neurons especially. 

                                                            CO probably has the longest half life of any complex under most circumstances. Production of superoxide is a control feature so will change as local conditions change. 

                                                            How do you propose to regulate the cleaving of the leader(s) and assembly of hCO?  It looks like the trigger now is insertion of COI-III into the inner membrane, formation pre-hCO, change conformation such that leader(s) cleaved form hCO.
                                                            Rate this comment: 12345
                                                            Guest (Dave Whitlock)
                                                            08/04/2006
                                                            Posts:1
                                                            • Bovine has 13 subunits too
                                                              Total of amino acids in bovine CO is 1803, of which 1002 are the mt trio:

                                                              http://www.life.uiuc.edu/crofts/bioph354/cyt_ox.html

                                                              Moreover, if CO is relatively long-lived as you suggest, then either the other complexes regulate their n-coded subunits some other way (such as axonal transport independent of mitochondria), or else they dominate the precursor storage problem and it's their nuclear-to-mt ratio that you should be considering.

                                                              If you have any evidence that cleavage of pCOIV leader happens after insertion of COI-III into the membrane, let's hear it.
                                                              Rate this comment: 12345
                                                              Guest (Aubrey de Grey)
                                                              08/05/2006
                                                              Posts:1
                                                              • There are other possiblities too
                                                                It is not my "suggestion" that CO is the longest lived complex, it is other peoples' data.

                                                                Maybe it is only the mt coded part of the complex that has a short half life.  Maybe the n coded stuff lasts a long time.  The metal centers tend to be on the mt coded stuff.  They are likely more subject to damage.

                                                                Cleavage happens after transport of mt out axon, and before formation of hCO.  Obviously some change in the mt trigger it.  Many different stimuli are known to trigger expression of mtDNA, which is known to precede formation of hCO.  Proteases are not coded by mt, so what causes the cleavage of pCOIV leader is a structural/conformational/location change.  Formation of hCO from subunits is a big conformation change.  Why assume another one is necessary?
                                                                Rate this comment: 12345
                                                                Guest (Dave Whitlock)
                                                                08/08/2006
                                                                Posts:1
                                                                • Guest (Aubrey de Grey)
                                                                  08/08/2006
                                                                  Posts:1
                                                                • Maybe this, maybe that (again)
                                                                  If you have data that show that CO is longest-lived, cite it, then we can see whether the study examined the mt-coded subinits.

                                                                  Now, your new suggestion is altogether more interesting -- that the synthesis of the mt-coded subunits could be the trigger for maturation and assembly of the n-coded ones. For the first time in this immense thread I see no reason why you are necessarily wrong. However, as you say, many stimuli are known to trigger mtDNA transcription and translation... and thus these stimuli too are available to trigger pCOIV maturation and assembly. Thus, I don't need to propose a new trigger for maturation - there are ample already. I do need to propose a new system for *delay* of maturation until this trigger occurs, but so do you.
                                                                  Rate this comment: 12345
                                                                  Guest (Aubrey de Grey)
                                                                  08/08/2006
                                                                  Posts:1
                                                                  • Huh?&nbsp; How if no mtDNA to transcribe?
                                                                    How can triggers of mtDNA transcription "trigger" pCOIV maturation if there is no mtDNA to be transcribed? 
                                                                    Rate this comment: 12345
                                                                    Guest (Dave Whitlock)
                                                                    08/13/2006
                                                                    Posts:1
                                                                    • Two answers
                                                                      1) The trigger doesn't have to act via triggering transcription - the two consequences can be independent.

                                                                      2) In vivo, OXPHOS-negative cells hardly ever have no mtDNA; rather, they have large deletions. Thus, there still is mtDNA to be transcribed.
                                                                      Rate this comment: 12345
                                                                      Guest (Aubrey de Grey)
                                                                      08/14/2006
                                                                      Posts:1
                                                  • Not "dogma", it is data
                                                    6-12% of exhisting and imported proteins are proteolysed (Langer T JBC 2005).  Most of that is respiratory chain. 

                                                    If mt degrade all these proteins, there are enzymes to do so, and also mechanisms to distinguish which proteins to proteolyse and which ones not to.  It isn't like the proteosome, where proteins get tagged, the specificity has to do with something else. 

                                                    It can't be just the leader, because there are lots of soluble proteins in the matrix.  
                                                    Rate this comment: 12345
                                                    Guest (Dave Whitlock)
                                                    08/01/2006
                                                    Posts:1
                                                    • Data say most, dogma says all
                                                      It may well be like the proteasome in fact - Ub-tagging in the cytosol is sometimes triggered by exposure of hydrophobic residues, and that may be a trigger for Lon too. But I digress. The main thing is, the Wong-Riley result tells us that some proteins don't get degraded. Maybe only one or two, but your whole objection to AE relies on it not being one or two but rather all nuclear-coded respiratory chain subunits.
                                                      Rate this comment: 12345
                                                      Guest (Aubrey de Grey)
                                                      08/02/2006
                                                      Posts:1
                                                  • Maybe it's a unique "feature" of pCOIV?
                                                    Maybe it has multiple configurations, the cytoplasm form, the unfolding at the outer membrane, the folding at the inner membrane, the sequester for a long time in neurons, and the hCOX form.  All forms that are protected from proteolysis.

                                                    Forms that depend on the specific sequence and how those specific sequences interact with all the Hsps, other chaperones all the proteases, and all the other proteins so they don't agglomerate. 

                                                    Why would pCOI just "happen" to act precisely the same way?  There is no reason to suppose that it would.  It is more likely that unincorporated COI is a danger signal and is rapidly degraded. 
                                                    Rate this comment: 12345
                                                    Guest (Dave Whitlock)
                                                    08/01/2006
                                                    Posts:1
                                                    • Maybe this, maybe that
                                                      Obviously since the experinment has yet to be done we don't know what will happen - this is all about which outcome is more likely. You're not taking al the available data into account in making that determination.
                                                      Rate this comment: 12345
                                                      Guest (Aubrey de Grey)
                                                      08/02/2006
                                                      Posts:1
                      • Never said AE wouldn't work
                        Only that SENS via AE won't work.

                        Patients with mitopathies have functional mitochondria that don't work well.  Augmenting their function via AE is completely different than SENS.

                        You are pretending that SENS via AE is "engineering".  It isn't. 

                        "Proportional" doesn't work for high n systems that are coupled and non-linear, which all biological systems are.  Such systems are chaotic.  
                        Rate this comment: 12345
                        Guest (Dave Whitlock)
                        07/25/2006
                        Posts:1
                        • Aha - progress
                          I claim your distinction is an admission that AE doesn't have to reproduce natural protein synthesis regulation 100% after all. But let's see what the referees say. Well?
                          Rate this comment: 12345
                          Guest (Aubrey de Grey)
                          07/25/2006
                          Posts:1
                          • You claim SENS can "sidestep" ignorance
                            AE for 1 mtDNA coded protein without Cu or Fe  (ie complex V subunit) is completely different than replacing all 13 with no loss in functionality with nearly complete ignorance of how the system is regulated. 

                            AE is not SENS.  SENS is replacement of all 13 mtDNA coded proteins via AE while remaining ignorant of the details.

                            SENS is going to the moon.  AE is gaining some altitude.  I can gain some altitude by jumping, I can't get to the moon that way.  
                            Rate this comment: 12345
                            Guest (Dave Whitlock)
                            07/25/2006
                            Posts:1
                            • Top people are trying AE on COX subunits too
                              Your blunderbuss approach to this topic is becoming tedious. Write it up for publication or admit you can't. I've challenged you to do this about half a dozen times now -- your failure to respond is becoming conspicuous.
                              Rate this comment: 12345
                              Guest (Aubrey de Grey)
                              07/25/2006
                              Posts:1
                              • Why would I switch to a forum you control?
                                So you can find "reviewers" who will accept your handwaving?  You have posted nothing that refutes the data showing that expression of mt coded proteins is important in short term regulation of ATP production.  You have simply asserted it. 

                                I know there is no data to support your assertion because there are no techniques for measuring the local ATP level on the length scale of mt spacing and correlating it with gene expression. 

                                We know that cells die, we know that when they do, their ATP level goes to near zero.  Obviously passing through the level that you consider unimportant. 
                                Rate this comment: 12345
                                Guest (Dave Whitlock)
                                07/26/2006
                                Posts:1
                                • Because my track record shows I waive that control
                                  I don't edit a journal for fun - I edit it so as to progress the relevant science, including discussions of that science. Besides, you also have the option to submit such a manuscript elsewhere. I publish my ideas in peer-reviewed fora (and not in RR!) precisely so that they are properly evaluated by the world leaders in the relevant field. Don't you?
                                  Rate this comment: 12345
                                  Guest (Aubrey de Grey)
                                  07/26/2006
                                  Posts:1
                    • Write your skepticism up properly if you're so sure
                      As I seem to need to remind you, AE is a long-established concept which many eminent mitochondriologists consider feasible for therapeutic purposes, including neurodegenerative mitochondriopathies. If you really think that not only I but they are so deluded, it's high time you wrote up your views for publication. I would certainly consider a manuscript on the topic for Rejuvenation Research. This staccato forum is not ideal for exploring complex matters; who knows, you might even have a point if you would just express it properly. Presumably you do have the courage of your convictions...?
                      Rate this comment: 12345
                      Guest (Aubrey de Grey)
                      07/25/2006
                      Posts:1
          • each mitochondria must be regulated independantly!
            In a neuron, the only concentrations that matter to a mitochondrion are those local to it.  O2, NO, ATP, pyruvate, etc.  In a neuron, these can change by large amounts along the length of an axon.  The operating point of the mitochondria has to change too. 
            Rate this comment: 12345
            Guest (Dave Whitlock)
            07/23/2006
            Posts:1
            • Quite - hence allosteric regulation...
              Of course - and that's why mitochondria have things like allosteric regulation of CO activity, which allow autonomous regulation of each mitochondrion.
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/24/2006
              Posts:1
          • Huh? blithely assuming observed functionality unimportant?
            Who is making blithe assumptions?  It is known mitochondria move in neurons.  It is known in different regions of a neuron, mitochondria will be subjected to different conditions O2, ATP, NO, etc.  Mitochondria can move ~30 microns/min (Journal of Cell Science 117, 2791-2804).  Makes sense to me, that adjusting operating conditions rapidly is important.  Assuming it isn't is extremely presumptious.  
            Rate this comment: 12345
            Guest (Dave Whitlock)
            07/23/2006
            Posts:1
            • See above, ad nauseam
              See above - they can and do do that, without relying on protein synthesis over the timeframe of a few weeks that a mitochondrion is distant from the cell body, by using mechanisms like allosteric regulation.

              One more time: you have provided zero evidence that mitochondria in neurons in vivo actually turn over their mt-coded proteins faster than every few weeks. Unless and until you can show that they do, your argument that AE is hopeless totally collapses. Please stop citing papers that do not bear on the question on which your conclusion unambiguously depends.
              Rate this comment: 12345
              Guest (Aubrey de Grey)
              07/24/2006
              Posts:1
              • The papers I cited quite clearly show it
                If you had read them, you would see that they do. 
                The Journal of Neuroscience, May 1993. 13(5): 1805-1819  Quoting from their abstract:  "acute regulatory changes in CO activity occurring over periods of days are controlled mainly by regulation of the mitochondrial genes that encode the catalytic subunits of the enzyme."
                Rate this comment: 12345
                Guest (Dave Whitlock)
                07/24/2006
                Posts:1
                • I've read more than the abstract
                  I did read them.  The J Neurosci paper looks at whole tissue, so it mainly tells us what's happening in the cell body (where most of the mitochondria are), so we're told nothing about your problem (travel up and down the axon limiting turnover rates). Also, protein levels were measured with polyclonal antibodies against the whole enzyme - for individual subunits. only mRNA levels were assayed. But also, as I said earlier, it's non-physiological -- it examines the response to an acute stimulus (TTX), so even if protein levels of individual subunits in synapses had been examined it would STILL tell us nothing about whether in normal physiological circumstances there would be any disparity between the regulation of mt-coded versus nuclear-coded proteins.
                  Rate this comment: 12345
                  Guest (Aubrey de Grey)
                  07/24/2006
                  Posts:1
      • The only hope for AE and SENS is engineered substitutes
        for mitochondria, but those will not be achieved simply by transfering mtDNA coded proteins into the nucleus.  The functionality of local control of O2 consumption matched to local O2 diffusion, matched to local ATP demand must be maintained. 

        "Maybe" the mtDNA proteins now coded as precursors, designed to be stable (in quite a difficult environment), soluble and triggerable "just in time" can stored, but if the volume of a mitochondria's "lifetime" supply, exceeds the volume of mtDNA and ribosomes, you will never get the same energy density or functionality. 
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/22/2006
        Posts:1
  • Engineered substitute for mitochondria not SENS
    A substitute for mitochondria that completely changes the design paradigm(s) is so far beyond our level of understanding that we can't even guess as to how difficult it would be. 

    If you can't even guess how difficult it is, it isn't "engineering".  For it to be "engineering", you need to have a defined path. 
    Rate this comment: 12345
    Guest (Dave Whitlock)
    07/22/2006
    Posts:1
    • Absolutely correct
      Subject says it all.  You won't find me making that mistake!
      Rate this comment: 12345
      Guest (Aubrey de Grey)
      07/23/2006
      Posts:1
  • If SENS is "science" it must be falsifiable
    What hypothetical data would falsify SENS? 
    Rate this comment: 12345
    Guest (Dave Whitlock)
    07/27/2006
    Posts:1
    • It's as falsifiable as any engineering project
      SENS can be falsified by implementing it in mice and seeing them die on schedule, just as the Wright Brothers' design for their plane could have been falsified by its not flying, or ITER could be falsified by being built and not generating sustained fusion, etc. The experiment is hard and expensive to do, but is still judged worth doing. (Only basic scientists ever seem to need to be told this...)
      Rate this comment: 12345
      Guest (Aubrey de Grey)
      07/27/2006
      Posts:1
      • An engineering project always has a finite budget
        If something does not have a finite budget, it is not an engineering project. 

        If the Wright brothers' designs had not worked, it would only have falsified their hypotheses for those specific designs.  The hypothesis that heavier than air flight was possible could not be falsified because it was already known to be true because birds did it. 
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/27/2006
        Posts:1
        • $1b over 10 years for mice (as I've published many times)
          (In case you havent noticed...)
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/27/2006
          Posts:1
        • existence proof
          Following your logic that heavier than air flight was not falsifiable because birds already flew, one could point to very long lived bowhead whales, sea turtles and bristlecone pines as existence proof of extreme longevity. A basic science approach to using birds as the discovery model for flight would've taken us down the blind alley of ornithopters. Engineering took a different route and satsifyingly successful route.
          Rate this comment: 12345
          Guest (Dave)
          07/31/2006
          Posts:1
          • I could have written that!
            Ah - different Dave I suddenly realise :-)
            Rate this comment: 12345
            Guest (Aubrey de Grey)
            07/31/2006
            Posts:1
          • Hypothesis of SENS is not longevity
            The hypothesis (that I have spent most time knocking) is that the 13 proteins coded in mt can be replaced via AE coded proteins with no change in mt function, and this can happen while remaining ignorant of the details of mt physiology, and this can occur in living humans while they are alive. 

            I have no doubt that longer lived mt exist, and no doubt that in principle longer lived mt can be created that are human compatable. 

            None of those long lived organisms use AE.  They all code every single one of those 13 mt coded proteins in their mt.  Bristle cone pines code a lot more besides.  So I don't think your analogy is apt. 
            Rate this comment: 12345
            Guest (Dave Whitlock)
            08/04/2006
            Posts:1
    • it's also "partially falsifiable"
      SENS can also be "partly falsified" by a demonstration that a specific approach to it won't work. For example, if there actually *were* evidence that AE could not substitute for mitochondrially encoded proteins in some cell types, and if there were also evidence that mtDNA mutations accumulation in those cell types was a contributor to age-related functional decline in a currently normal lifetime, that component of SENS would need to be replaced by something else (such as protofection, or selective ablation and replacement of affected cells, etc). The essence of SENS as a composite, divide-and-conquer approach to combating aging would not be falsified in this circumstance.

      Another "partial falsification" would be if a type of age-related change other than those that SENS enumerates (eg, aspartate racemisation of long-lived proteins) were shown to contribute to aging. Then a way to combat that (e.g. stimulating turnover) would also be needed as an additional SENS component.
      Rate this comment: 12345
      Guest (Aubrey de Grey)
      07/27/2006
      Posts:1
  • No, that is not falsifiability
    That a mechanic is unable to repair a classic car does not falsify the hypothesis that such a car can be repaired.

    Unless all possible combinations are tried, a hypothesis is not falsified by the absence of a success. 

    Are you saying SENS can only be falsified by trying all possible combinations of all possible proteins with all possible sequences?   
    Rate this comment: 12345
    Guest (Dave Whitlock)
    07/27/2006
    Posts:1
    • I said "as falsifiable as any engineering project"
      Are you saying engineering innovation is always a waste of money until it works? If not, what distinguishes SENS in your eyes from (for example) ITER?
      Rate this comment: 12345
      Guest (Aubrey de Grey)
      07/27/2006
      Posts:1
      • ITER is not an engineering project
        The Big Dig is an engineering project.

        "ITER is a joint international research and development project that aims to demonstrate the scientific and technical feasibility of fusion power."  http://www.iter.org/

        We know that thermonuclear fusion happens.  There have been multiple devices that have produces energy from fusion, just not "efficiently enough", or "economically enough".  If SENS is to be compared to fusion projects, it should be compared to those of the 1950s.  When they thought it was trivial and would bring power "too cheap to meter" in just a few years. 

        Trying to build ITER in 1950 would have been a waste of money.  Trying to build the Big Dig in 1850 would have been a waste of money too. 
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/27/2006
        Posts:1
      • Any engineering project is "falsified" if it exceeds budget
        Going to the moon for a week with a budget of $500 billion may be an "engineering project" because it has already been done for less.   

        Going to the moon for a week for $100 is trivially falsified because a week's labor at minimum wage is more than that. 

        Research on SENS may be justified, but it is research, not engineering.  You can't "sidestep" ignorance in engineering.  Either your design accomodates ignorance successfully or your design fails. 

        Trial and error shots in the dark is the Edisonian approach to research, he sized distribution wire by experiment.  Sprague engineered the size of wires by calculation.   
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/27/2006
        Posts:1
  • Is $1 billion a budget, or a down payment?
    One of the reasons the Big Dig was so much over budget was because they made the "budget" before they did soil borings and before they knew what soil conditions they had to design for.  They tried to "side step" their ignorance by making assumptions (aka guesses).  They chose the guesses that gave the lowest estimate, and so the project was doomed to be over budget. 

    You have made a number of assumptions (aka guesses), regarding the degree of difficulty of AE.  Those assumptions are difficult to evaluate because AE hasn't been done before. 

    At the end of 10 years and $1 billion will SENS be either successful or falsified?  Or will we know that $1 billion worth of "magic" bullets don't work, but $1 billion of something else might? 
    Rate this comment: 12345
    Guest (Dave Whitlock)
    07/27/2006
    Posts:1
    • a budget - which might indeed overrun - so what?
      Now we're getting somewhere. Your comparison of SENS with 1950s fusion is, I respectfully point out, a guess. My guess is that SENS resembles 2000s fusion. Call it a "research and development project" rather than an engineering project if you wish; it's still worth trying now, even if there's a fair chance that it will overrun in time and money, just as the Big Dig was worth starting 15 years ago. Likewise, call SENS "accommodating ignorance" (having plan B's if parts of the design hit trouble, as I've already given examples of) rather than sidestepping it if you wish; it's still goal-directed rather than curiosity-driven, which is the key thing that distinguishes technology from basic science.

      Bottom line: your argument against pursuing SENS now is a guess, and a considerably less educated guess than my contrary one, because I've been studying the component technologies, publishing on them, and discussing them with the leaders in the relevant fields for several years.
      Rate this comment: 12345
      Guest (Aubrey de Grey)
      07/28/2006
      Posts:1
      • Guesswork is not "engineering"
        Fusion work in the 1950s was goal oriented.  Their goal was to make power too cheap to meter, not to do research.   

        No amount of “expertise” would allow an engineer to simply walk the Big Dig site and “estimate” project cost without doing soil borings.  Pretending you can, is not “engineering”, it is gambling.  Those who "budgeted" the Big Dig knew how to get accurate budgets, they chose not to.  You don't know how to get an accurate budget for SENS. 

        My argument against persuing SENS now is that there are plenty of other projects that will save more lives in the nearer term than SENS will.
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/30/2006
        Posts:1
        • near term vs long term
          I don't see your point - SENS is obviously also goal-oriented. There is research involved in figuring out the details of getting SENS working, but so there was (and still is) with fusion.

          > Those who "budgeted" the Big Dig knew how to get accurate budgets, they
          > chose not to.  You don't know how to get an accurate budget for SENS.

          Quite true - but as I already said, you an call SENS and fusion R&D rather than engineering and the argument for pursuing them still applies, because they're goal-directed.

          > My argument against persuing SENS now is that there are plenty of other
          > projects that will save more lives in the nearer term than SENS will.

          What's so important about the nearer term? If we thought the way you suggest, we'd never have started work on fusion, because the money could always have been spent making energy a bit cheaper in easier ways.
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/31/2006
          Posts:1
      • SENS is a high tech Pascal's wager
        Immortality has an infinite value, so any expense is negligible by comparison, even if the probablity of success is very small (but non-zero). 

        However, it is of infinite value to those people that are alive today only if it can be attained during their lifetime.  The SENS gamble makes sense in that context.

        It is not clear if immortality is of infinite value to society.  Many people are already dying of starvation and diseases that can be prevented.  SENS will not change that, it may even make it worse.  We do not currently posess the technology or political will to provide every human with an acceptable lifestyle in a sustainable manner. 

        When global warming melts Greenland it will raise sea level by 7 meters.  Many hundreds of millions will be displaced.  I consider that a more important problem than SENS.    
        Rate this comment: 12345
        Guest (Dave Whitlock)
        07/30/2006
        Posts:1
        • We may have just exceeded this system's space limit
          I've had "Runtime Error" on several attempts to post a reply to your last post. Trying this short post instead...
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/31/2006
          Posts:1
        • other people matter too
          > Immortality ... is of infinite value to those people that are alive today
          > only if it can be attained during their lifetime.

          Wrong, because people have value to each other. I, for example, feel that I'll die happy if I've helped bring the defeat of aging closer.
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/31/2006
          Posts:1
        • Empowering humanity
          > Many people are already dying of starvation and diseases that can be
          > prevented.  SENS will not change that, it may even make it worse.

          No - defeating aging could well empower humanity and give us the determination to prevent preventable things. Preventable things persist now mainly because of ubiquitous fatalism, which I think results mainly from the inevitability of aging.
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/31/2006
          Posts:1
        • Displacement of millions worse than death of billions??
          > When global warming melts Greenland it will raise sea level by 7
          > meters.  Many hundreds of millions will be displaced.  I consider that
          > a more important problem than SENS.

          That's the first time I've heard displacement rated as more serious than death... But also, it's highly implausible that combating global warming and combating aging will impinge on each other's budgets.
          Rate this comment: 12345
          Guest (Aubrey de Grey)
          07/31/2006
          Posts:1
          • False alarm I guess
            Hm - well, the original post was smaller than some of my past ones but chopping it into three parts worked, ho hum.
            Rate this comment: 12345
            Guest (Aubrey de Grey)
            07/31/2006
            Posts:1
  • Excellent debate
    Hi Dave Whitlock, this is a fascinating debate. You may find the following of interest:
    http://www.imminst.org/forum/index.php?s=&act=SF&f=173
    Rate this comment: 12345

    prometheus
    08/19/2006
    Posts:2
  • AE alternatives - cheaper, faster to implement using today's technology
    1. Increase the activity of mt DNA repair enzymes (or direct other DNA repair enzymes to mt)
    2. Increase the rate of turnover (autophagy) in faulty (ROS leaky or poor ATP producers)
    Rate this comment: 12345

    prometheus
    08/19/2006
    Posts:2

Log In

Forgot your password?     Register »
Advertisement

Videos

The Marcellus Shale Gas Rush
Technology Review November/December 2009

Current Issue

Natural Gas Changes the Energy Map
The United States has vast supplies of this cleaner fossil fuel. But how should we use it?
Advertisement
Advertisement
Subscribe to Technology Review's daily e-mail update. Enter your e-mail address

TECHNOLOGY RESOURCES

More Technology News from Forbes

Advertisement
MIT Massachusetts Institute of Technology © 2009 Technology Review. All Rights Reserved.