The Present State of Progress Towards Clearing Glucosepane Cross-Links, a Contributing Cause of Degenerative Aging

Cross-links are in essence a type of damage resulting from metabolic waste, a natural side-effect of the normal operation of our cellular biochemistry. Many different types of sugary molecules known as advanced glycation end-products (AGEs) end up in the spaces between cells and can react with and link together the intricate structures of the extracellular matrix. The arrangement and constituents of the matrix are what gives each tissue its particular set of properties: elasticity in skin and blood vessels, the ability to bear load without brittleness in cartilage and bone, and so forth. The presence of cross-links in significant numbers sabotages these properties, such as by preventing long, parallel molecular structures from sliding freely past one another. Further, there is evidence to suggest that AGEs produce raised levels of chronic inflammation by altering cellular behavior through the receptor for AGEs, RAGE. Inflammation contributes to the pathology of all of the common age-related diseases.

Most cross-links formed by AGEs are transient, and perhaps only significant in an abnormal metabolism, such as that produced by obesity or type 2 diabetes. The present consensus is that the real problem - leading to age-related loss of skin elasticity and stiffening of blood vessels, among other issues - is produced by a single type of hardy cross-link formed by one type of AGE called glucosepane. Studies suggest that glucosepane makes up the overwhelming majority of cross-links in old humans, and our natural biochemistry is not equipped with tools that can effectively remove these chains.

This is one compound, and to greatly reduce its contribution to aging all that is needed is one moderately effective drug candidate that can break it down. This drug candidate would have as its target market more than half of the human race - pretty much everyone over the age of 30. Yet the broader research community has shown no interest in this goal, an issue we might blame on the lack of tools for working with glucosepane. Any research group diving into this problem would have to build all of the tools from scratch, and that means that near everyone who did take the time to think about it has chosen, again and again, to work on other, more accessible problems instead. This sort of situation requires philanthropy to break the log jam, and thus the only significant funding for glucosepane research in the past few years has come from the SENS Research Foundation, via philanthropists such as Jason Hope, and of course the charitable support of this community.

Nonetheless, because this is a narrow domain problem, the search for one drug candidate for one target, I believe it is the most likely SENS technology to follow on from senescent cell clearance as next in line for commercial development. A method of senescent cell clearance is currently being developed by Oisin Biotechnology, and whatever happens next after that in the SENS space is probably a race between a viable glucosepane breaker drug and transthyretin amyloid clearance, with mitochondrial DNA repair just a few years behind those. However, my knowledge of the latest activity has been getting out of date, so I recently talked to some of the people involved; Aubrey de Grey of the SENS Research Foundation, David Spiegel who runs a lab at Yale, and William Bains who collaborates with an eclectic range of researchers in numerous fields, including this one. What follows is a rough summary of their thoughts on the matter.

A Way to Make Glucosepane is a Big Step Forward

The Spiegel lab developed a reliable way to make glucosepane last year. This is a big deal because people who could not previously collaborate with this type of research can now set up their own studies and investigations. It also ensures that, at least for the foreseeable future, everyone is working from the same definition of what exactly is meant by glucosepane and its particular molecular structure.

There are Still Doubts Over the Glucosepane Consensus

The consensus on glucosepane as the overwhelming majority of relevant cross-links in the process of aging is not airtight - there are growing doubts. It is perhaps reasonable to think that it should be the primary target based on the evidence to date, and Spiegel is optimistic that useful therapies will emerge, but de Grey is cautious, and Bains somewhat unhappy about the poor quality of some past research on this topic. If there was a way to break glucosepane, then doubts could be rapidly solidified or put to rest, but that still lies in the future. The SENS Research Foundation is presently funding research with Jonathan Clark at the Babraham Institute to attempt to ratify that glucosepane is the target, determine whether or not there are other targets, and establish that the present understanding of the structure of glucosepane is in fact the right thing to aim for. Remember that a molecule made of a given set of constituent parts might have a poorly understood shape when folded, these molecules are large and complicated, and shape determines function.

A Drug Candidate Doesn't Exist Yet

There is no drug candidate to clear glucosepane at this time, and not even a speculative idea of where to look for possibilities in the enormous back catalog of existing and explored pharmacology. This lack of direction is a consequence of the lack of exploration of this type of compound in the field. Finding the drug candidate is the big gap that lies between where things stand today and the point at which someone could launch a startup company to finalize a potential glucosepane breaker therapy. The labor required to verify that such a drug candidate works or does not work is modest in comparison to the work of finding such a candidate; this would involve building fairly standard forms of assay to determine levels of glucosepane before and after treatment. One standard approach to this sort of thing would be to equip the immune system with antibodies that react to glucosepane, and then measure the response.

A Drug Candidate Will Most Likely Emerge from Mining the Bacterial World

The Spiegel lab is following the same approach as the LysoSENS research program did over the past decade, which is to search for enzymes in bacteria capable of efficiently breaking down glucosepane. We know they exist because graveyards are not sticky sumps of metabolized sugar. This might actually be discouraging to hear at first, as LysoSENS ran for a decade before transferring the first drug candidates for commercial development by Human Rejuvenation Technologies. However, an enormous advance in the ability to culture bacterial species has taken place in just the past couple of years, an advance not available to the LysoSENS researchers. One of the open secrets of the life sciences used to be that 99% of all bacterial species couldn't be cultured in the lab - but all of a sudden and with a comparatively simple technological advance, that has changed. Everything that bacterial researchers achieved in the past was accomplished with the 1% of bacterial species that were suitable to work with, but now that all bacterial species are fair game, the search space for new molecules has multiplied a hundredfold.

The researchers at the Spiegel lab have already isolated and cultured bacterial species that they are reasonably confident are consuming glucosepane. David Spiegel believes that it might plausibly take two years at the present level of funding to characterize how the bacteria are doing this and whether it involves a simple, single enzyme or something more complicated. If it is a single enzyme, then that can move fairly rapidly to becoming a drug candidate. If not, well, it is probably faster just to look for more bacteria with better candidates. This is a research project that could move faster with more money, as the activities can be carried out in parallel were there more researchers on the staff - but of course raising funds for research in this field is ever a challenge.

Note that I'm glossing over the challenges inherent in picking out enzymes from bacteria and turning them into drugs. There are often unwanted effects, such as triggering of the immune system, that have to be designed out. Many of the options for working around this problem, such as encapsulating drug molecules in a protective sheath, are not practical for something that is intended to get into the tiny spaces of the extracellular matrix. And so on. But these are all challenges that can be addressed, extra work requiring technologies and approaches from elsewhere in the research community to be pulled in.

Two Models for Future Commercial Development

There are two models for commercial development from this point. The first is for an investor with two years of patience, $2 million, and an appetite for risk and uncertainty to come in and fund a company to finish the work started by David Spiegel, William Bains, and Jonathan Clark and their research teams. This sort of thing does happen in many industries, but it is very hard to arrange without deep pockets and good connections. That is why you see this sort of arrangement more often taking the form of a partnership with a pharmaceutical company, as happened for the development of the transthyretin amyloid clearance therapy based on CPHPC.

The other model is to cheer on the researchers, and support them as we can with our donations, for the perhaps few years needed to iron out the doubts about glucosepane, and find a candidate bacterial enzyme. Once they are within striking distance of a proof of concept in mice or rats, then a seed-funded startup could be founded and work proceed from that point. That is much easier to swing for this community - if Oisin managed to obtain seed funding from SENS supporters, then a glucosepane-breaker company could certainly do so to the same level a few years from now.


What is particularly interesting is that AGEs have been shown to induce premature cellular senescence, suggesting that high glucose levels associated with diabetes likely accelerates ageing in other tissues that consequently manifest as disease. Cardiovascular disease and renal disease for example are accelerated in patients with diabetes and AGEs have been shown to induce senescence in cell types related to these diseases, such as endothelial progenitor cells, vascular endothelial cells and renal cells. Clearing AGEs may therefore prevent accelerated ageing mediated by senescent cells.

Posted by: DGAB at January 27th, 2016 6:26 AM

I'm less sanguine about the prospects of bacterial enzyme searches to clear glucosepane.

There are already several enzymes which are able to completely digest aged collagen, and the glucosepane links along with it. And they're nothing exotic - pepsin and trypsin are found readily in the human digestive system, which can also readily digest collagen, thank you very much. The problem is that they are enzymes which have broad-spectrum destructive capabilities, making them unsafe or impractical to simply let free in the ECM. Some also work only in toxic pH conditions, which is another issue. But bacteria and other microorganisms just need to find something which works. It's not their problem if the solution they use won't help us for the odd situation we're trying to solve.

Glucosepane is not some super-bond. It's acid-labile, and in fact one theory about conflicting results found between some studies is that the weak acids used for collagen extraction destroyed the glucosepane links in the process. It just happens to be in an awkward position to interfere with the highly specific MMP enzyme that the is designed to work in the ECM.

Another issue with enzymes is the tight packing of the collagen fibrils, making it questionable if the enzyme will be able to reliably reach its target. A tailored small molecule may be a considerably better option for that reason alone.

Posted by: AH at January 27th, 2016 7:57 AM

Thanks for the introduction today Reason, the MMTP researchers have expressed interest in testing AGE breakers as soon as something is available.

Posted by: Steve H at January 27th, 2016 9:22 AM


Hi AH ! AGEs such as glucosepane, CML, methylglyoxal, pentosidine, Amadori products, furosine, oxidized sugars and other crosslinks are in part removed, as you say by some enzymes, but especially by cell cycling dilution mechanism (the daughter cells don't inherit the dividing mother cell junk for a while, until a certain point small amount of junk is shared by new daughter cells and mother can't keep it anymore as it is dying; so transfers some to daughters; to continue making daughters cells. That's only for mitotically active cells, not for post-mitotic non-dividing/non-mitotic cells); which is also applicable for intracellular and extra-cellular lipofuscin (lipofuscin is mostly intracellular in lysosome; but also found and ends up indefinitely layed tere in extra-cellular cytosol. I think this is an internal cell problem that spills into an external cell one (lysosomes lipofuscin exocytosis and trying to evacutate it in extra-cellular milieu like cytosol).

Trypsin, chymotrypsin, serine protease, serratiopeptidase, peptidase, amylase, cellulase, papin, elastase, etc are all hydrolytic in nature and catalyze hydrolysis process. All of them cannot do anything about lipofuscin aggregates undegradability and irrevesibility (only cell cycling resolves that problem, it's the only solution; but Aubrey De Grey's team and other scientists found bacterial enzymes in nature (in sphagnum bacteria-producing lipofuscin decomposing-enzymes of decomposing peat bogs (where they found peat bogs bodies too such as Tollund mummy man who is perfect preservation) who can degrade lipofuscin).

It may be possible that some enzymes in the body are capable of degrading lipofuscin, but they are misdirected and don't degrade lipofuscin - enough - to make a difference and give slack to the lysosome, proteasome and autophagy process (which clearly can't do anything about it because lipofuscin accumulates mostly there (enlarging lysosomal mass with aggregate accrual/deposition)).

''Lipofuscin is not degradable by the host of hydrolytic enzymes contained within the lysosomal apparatus.''

''The Determination of the Concentration of Hydrolytic Enzyme Solutions: α-Chymotrypsin, Trypsin, Papain, Elastase, Subtilisin, and Acetylcholinesterase1.''

Thus, the human's hydrolitic enzymes mentioned can remove these AGEs but they can't do anything about truly undegradable aggregates accumulating in post-mitotic non-cycling cells (only those found in bacteria-producing of peat bogs can).

Posted by: CANanonymity at January 27th, 2016 10:41 AM

AH: What if a more practical target isn't glucosepane but glucosepanated collagen? One of the theories being bandied about here is that the best way to deal with glucosepane is simply to destroy collagen a little bit at a time on a regular basis, let it heal, and repeat. The problem with doing this excessively, of course, is that a rupture in the wrong place leads to a very short lifespan indeed.

It's been said multiple times that glucosepane interferes with MMP. But how much? What form does that interference take? Does it slow the process down or stop it completely? There was a highly detailed study published some months ago about where glucosepane is likely to bond. Do some of these bonds affect MMP more than others?

Posted by: Slicer at January 27th, 2016 10:42 AM


Hi Slicer !

Here is a research giving some answers:

''Advanced Glycation endproducts (AGEs) accumulate during aging in many slowly renewing tissues, skin included. In the collagen-rich extracellular matrix, glucosepane increases with age. Glucosepane is the most prevalent AGE and protein crosslink found, up to date, in the dermal extracellular matrix of aging human skin although little is known about its biological impact. To study the effects of glucosepane on skin homeostasis, an in vitro model of chronological skin aging was used, consisting of skin reconstructed on a dermal equivalent where collagen was previously modified by inserting chemically synthesized glucosepane. Analysis of some skin markers revealed several unexpected biological and morphological alterations like those induced by in vivo skin aging that could be therefore glucosepane related. Indeed, in this in vitro glucosepane modified skin model, the synthesis of pro collagen I and collagen III decreased
********whereas the level of pro-inflammatory cytokines and MMP1 but not TIMP1 increased******.
At transcriptional level, the expression of major mRNA’s for extracellular matrix components (proteoglycans and collagens) was reduced and
*******MMP9 mRNA was increased*******.
The presence of aminoguanidine (a glycation inhibitor) during glucosepane synthesis restores a normal pattern of the altered skin markers. Taken together, these results suggest a specific role of glucosepane in the skin aging process.''

So, in fact, MMPs levels (mRNA expression at least) are increased by crosslinks such as glucosepane. This makes senses it tips the balance of ECM collagen turnover vs generation, towards an extreme turnover/degradation.
Glucosepane is undegradable by proteolytic enzymes such as MMP-9. But, clearly, it creates an inflammed state where it is hyperdegradation by MMP activation (mostly likely trying to compensate from glucosepane formation and trying to degrade it, but it can' MMPs futily degrade the ECM thanks to glucosepane, messing the balance of collagen turnover and production).
Perhaps, you are right, that MMPs could degrade glucosepanated ECM collagen, that is something we don't know, still from these studies it seems not and MMPs just mess this whole up (trying to consume glucosepaned collagen - but they can't and go on and on degrading the collagen in a futile proteolytic degradation of collagen). Just my not-informed opinion of ECM dynamics.

Glucosepane cross-linked to collagen alters the homeostasis of reconstructed skin in vitro and induces alterations similar to skin aging

Posted by: CANanonymity at January 27th, 2016 12:55 PM


We're talking about glucosepane links between collagen molecules in the ECM (extracellular) not intracellular junk like lipofuscin, so I don't see how daughter cell dilution will help. I haven't explored lipofuscin much, and you might be right about that enzyme search, but I don't see how its relevant to the glucosepane enzyme search, which I think is unlikely to succeed for the reasons I set out.

Posted by: AH at January 27th, 2016 3:33 PM


AH, yes I see what you mean.
We have to remember lipofuscin is not only intracellular junk but also extracellular (it accumulates in the cytosol too) just like AGEs are.
And what's more, is that AGEs is mostly lipid peroxidation product but there is an element of AGEs - inside - lipofuscin. Lipofuscin is an amalgam of residue (lipid perodixation residues, broken DNA and also some AGEs mixed-in at times; for example A2E lipofuscin drusen is composed, in part, of AGEs).

Also, intracellular milieu (ICM) is linked to extracellular milieu (ECM), what goes inside the cells also affects ECM because they are tied together and in communication; but also ECM is created by the ICM; for example, without fibroblast or stromal cells, for example, there wouldn't be much ECM or ECM scaffold formation in the first place; for they are key elements for ECM existing in the first place. ECM and ICM both feed off of each other and can't survive without one another.

It's true that cell cyclin dilution (being an intracellular milieu effect) would not have much effect on ECM crosslinks, but we must not forget AGEs accumulate in cells'ICM too, not just the ECM. It's of course, the AGEs that create the ECM crosslinks (as glucosepane) that are the big problem.
But as said, the body found ways to remove these crosslinks, it may not be MMPs but something removes them somehow, but some form of ICM or ECM dilution mechanism or endocytosis or exocytosis mechanism to these crosslinks too.
How can we say that ? Well studies in CR showed there was a Reduction in crosslinks and AGEs. How did it achieve that, by reducing glycemia (thus reducing glucose glycation exposure creation of AGEs-crosslink collagen formation), reducing oxidative stress (improve redox), increasing proteasomal/lysosomal clearance of unfolded proteins/junk and it inhibited collagen production. And we have to remember, if calorie restriction is initiated and autophagy is inhibited at the same time, nearly all of the benefits are lost - pointing to one direction (autophagy, lysosome, proteasome clearance of AGEs, since we know that certain AGEs and crosslinks litterally block the proteasome and are undegradable to it; yet, clearly something happens and it ends up degrading - some or let's say the low oxidative stress by CR helps greatly to reduce the accumulation; but it's possible that it is undegradable in ECM).

''Caloric restriction (60% of ad libitum intake) maintained only during the 2-week experimental period did not affect collagen accumulation, but did result in decreased levels of the difunctional crosslink dihydroxylysinonor-leucine (DHLNL) in sponges implanted for 10 days.''

difunctional crosslink dihydroxylysinonor-leucine (DHLNL)

Effects of aging and caloric restriction on extracellular matrix biosynthesis in a model of injury repair in rats.

Posted by: CANanonymity at January 27th, 2016 4:43 PM


As CANanonymity noted, the current consensus is that promoting collagen destruction - for instance, by upregulating MMP's - would not be effective, since it will preferentially destroy the 'good' collagen which is not crosslinked, while the 'bad' collagen with crosslinks is unaffected, since most of the reason it *is* bad is because it's resistant to MMPs (specifically, collagenases).

How many bad ones are there? It's estimated that with advanced age (80+) there is roughly 1 glucosepane crosslink for every 5 collagen molecules. Diabetics are roughly 1 in 2. The standard figure cited is that this degree of crosslinking translates to a 30- and 45-fold decrease in collagen digestibility, respectively. This is based on an extrapolation from the work of Vater ( on lysyl-oxidase crosslinks of collagen, which found that synthesis of approximately 0.1 cross-links per molecule reduced resistance to collagenase 2-3 fold (from 80% lysis to 25% - 40% lysis in Figure 3). However, as Vater noted ad loc, "[t]here was no significant difference in resistance between fibrils with 0.13 and 0.24 cross-link/molecule." Given that nearly doubling the cross-links per molecule caused no further increase in resistance, I personally am of the opinion that resistance does not reach 30-fold, but it's still going to be pretty high.

The study from Collier that you referred to discusses your question, actually. They found 6 sites in the triple stranded collagen molecule where intramolecular crosslinking is energetically favorable. One of those sites (between a1a LYS 791 and a1b ARG 789) is within close proximity to the MMP1 (collagenase) binding site, and they theorize that it is what interferes with the MMP slicing.

Posted by: AH at January 27th, 2016 5:01 PM

PS: these ones are telling and show proteasome and ECM MMPs work together ; and both the target elements behind AGEs crosslinking problem and loss of collagen.

If CR can maintain autophagy and proteasomal activity and somehow reduce crosslinks (although it may only be New crosslinks formation/rate that is affected, rather than the cross links themselves (which are untouched but simply accumulate less) and the already existing ones are undegradable); it means autophagy has some role in that. I wonder about that ECM problem...if ICM is in perfect condition and is rejuvenated - what happens in ECM then ?
If we extrapolate using a young child or foetus, we see that they a great capability of reduced crosslinks formation; but what about current existing crosslinks - are they removed when they is biorejuvenation ?

In intracellular milieu, AGEs, can be removed through cell cycling, but in ECM, that is something hard to say; ECM remodelling by telomerase-instruction is something possible too. From this study below, it is clear telomerase is capable of instructing the ECM degradation machinery to remodell the whole thing, so most likely glucosepane gets glucose'caned' out of ECM by telomerase-dependent instructions. What is clear is that telomerase-deficiency accelerates ECM loss and AGEs/crosslink formation; and telomerase activation slows down crosslinks formation (because they are interconnected pathways, ECM machinery can be instructed by telomeres (from telomerase therapy)). And if telomeres increase - glucosepane must surely decrease too - including existing one; via ECM remodelling or some other dilution.

''Telomerase deficiency has been associated with inadequate differentiation of mesenchymal stem cells. However, the effect of telomerase deficiency on differential regulation of osteoblast specific genes, based on functional gene grouping, during in vitro osteoblast differentiation has not been reported before.ResultsTo examine these effects, Terc -/- BMSCs (bone marrow stromal stem cells) were employed which exhibited reduced proliferation ... Similarly, in Terc -/- BMSCs a marked reduction in other genes engaged in various phases of osteoblast differentiation were observed, such as Fgfr2 involved in bone mineralization, Phex and Dmp1 engaged in ossification, and Col11a1 and Col2a1 involved in cartilage condensation.
*******A similar trend was observed for genes involved in osteoblast proliferation (Tgfb1, Fgfr2 and Pdgfa) and bone mineral metabolism (Col1a1, Col2a1, Col1a2 and Col11a1). More profound changes were observed in genes engaged in *extracellular matrix production*: Col1a1, Col1a2, Mmp10, Serpinh1 and Col4a1.*******Conclusion
Taken together, these data suggest that telomerase deficiency causes impairment of BMSCs differentiation into osteoblasts affecting commitment, proliferation, matrix mineralization and maturation''

'Figure 4: Effect of telomerase deficiency on extracellular matrix (ECM) protein molecules during in vitro osteoblast differentiation of BMSCs. A) Differential expression of genes involved in cell differentiation into ostoeblasts. B-E) Differential gene expression profiles of molecules involved in extracellular matrix production in Terc -/- BMSCs – notable genetic molecules were related with ECM collagens, basement membrane constituents, ECM proteases, ECM protease inhibitors and other molecules associated with ECM. Genes that were not amplified or not detectable in the PCR array were marked as ‘n.d’ (not detectable). Osteogenic super array data are represented as fold change relative to WT controls of three independent biological replicates pooled together.'

''Cross-linking of glycated collagen in the pathogenesis of arterial and myocardial stiffening of aging and diabetes.Cross-linking of glycated collagen in the pathogenesis of arterial and myocardial stiffening of aging and diabetes.''

Arterial stiffening is reversed by telomerase therapy and this arterial ECM stiffening is due to crosslinks such as glucosepane. Steve H said that
Telocyte (telomerase rejuvenation) by DR.Fossel reverses vascular stiffening.

Aberrant gene expression profiles, during in vitro osteoblast differentiation, of telomerase deficient mouse bone marrow stromal stem cells (mBMSCs).

Downregulation of Matrix Metalloproteinases and Collagens
and Suppression of Cardiac Fibrosis by Inhibition of
the Proteasome

Proteasome subunit LMP2 is required for matrix metalloproteinase-2 and -9 expression and activities in human invasive extravillous trophoblast cell line.

Posted by: CANanonymity at January 27th, 2016 5:27 PM


I'm not sure what you mean by 'the body found ways to remove these crosslinks'. Intracellularly, yes, there are deglycation enzymes, but they require ATP to function and therefore won't work in the ECM. In the ECM, no, the body has not found ways to remove these crosslinks.

Caloric restriction may buy you some time since reducing glucose concentration should slow the *rate* of glycation, but it won't do anything to help eliminate crosslinks that have already formed.

Also, that study you linked to was dealing with enzymatic crosslinks, not nonenzymatic AGEs like glucosepane. Lysyl oxidase could be inhibited due to any number of effects from CR. If anything, though, fewer enzymatic crosslinks actually means there are more sites available for nonenzymatic crosslinks to form on later - i.e., more glucosepane!

Posted by: AH at January 27th, 2016 5:52 PM


You post too fast. In my last post, I meant the study from your post at 4:43 PM.

Posted by: AH at January 27th, 2016 5:55 PM


Regarding children, ECM is in a state of heightened remodeling until adolescence, when it slows rapidly since you stop growing. From then on, remodeling is slow, and really more of healthy replacement except for edge cases like pregnancy. This is why AGE accumulation only starts in your 20s, and follows a pretty linear accumulation from then on. Nothing really breaks with aging, it just takes a while for damage to accumulate. So I think even if rejuvenation brought you back to the health of your 20s - 40s, by your 80s - 100s you'd still be screwed by glucosepane accumulation doing whatever it's going to do.

Posted by: AH at January 27th, 2016 6:28 PM

Telomerase - or any other signaling system - can yell until it gets blue in the face, but without an enzyme or other molecule that can do the actual work of breaking down glucosepane-linked collagen, I can't see how it can accomplish anything.

Posted by: AH at January 27th, 2016 6:32 PM


''So far, no agent has been found that breaks the prevalent glucosepane and K2P crosslink structures. Enzymes that would be able to recognize and disassemble glycation products may be too big to migrate into the ECM and repair collagen or elastin in vivo. Two approaches to therapy development are presented here. ECM turnover enhancement would enhance natural processes to digest old ECM and replace it with new. It will be important to tune the collagen degradation to a rate slow enough to prevent dire side-effects, such as hemorrhage from leaky blood vessels as collagen molecules are removed and replaced.''

glomerular basement membrane (GBM)
''However, the most likely explanation for the differences in glucosepane levels is from differences in collagen turnover. First, although direct comparative measurements between skin and GBM are not available, work by Cohen and Surma (23) suggests that rat GBM turnover rate is as rapid as that of salt soluble Type I collagen, which is notoriously poorly cross-linked. Another report suggests both GBM and tendon collagen turnover are at least greater than 100 days (24). Thus, although our data are compatible with higher turnover of GBM than skin, precise measurements are needed. Second, the extent of glycation of long-lived proteins such as collagen is in the equilibrium or steady-state relationship to ambient glucose concentration and further modulated by turnover (25–28). In that regard, the extent of glycation (Amadori product) is lower in GBM when compared with tendon, aortic, and skin collagen (29, 30). Additionally, levels modestly but significantly increased with age in nondiabetic human skin collagen (26) but not GBM (31). Furthermore, GBM in the diabetic milieu undergoes increased collagen synthesis and thickening, which in itself will affect turnover (30, 32, 33). Third, similar to the current study, previous work with the pentosidine cross-link showed that levels increased exponentially to over 90 pmol/mg of collagen at age 90 years in nondiabetic skin collagen, whereas levels increased asymptotically to less than 50 pmol/mg, plateauing between the ages of 50–60 years in GBM from nondiabetics (2). Thus, these results further support the notion that GBM turns over at a faster rate when compared with skin.''

Perhaps, as Slicer suggested, ECM turnover acceleration could partially compensate
glucosepane accrual; as long as it is a maximal-safe level, not to extreme level as it has dangerous consequences like hemorraghing from leaky blood vessels. Increased ECM turnover may make no difference to existing glucosepane crosslinks since MMPs can't degrade them, but as in glomerular basement membrane that has accelerated turnover, it has lower crosslinks accumulation rate and quantity than in collagen because it has faster turnover. So I'm guessing doing like GBM could partly solve the issue.
If you say children up to adolescence have high ECM remodelling, and with aging during adulthood because of growth stop there is less ECM turnover/remodelling; well perhaps that's what's missing to solve things and we must get back to original youth levels of ECM remodelling/turnover rates.

Extracellular glycation crosslinks: prospects for removal.

Glucosepane Is a Major Protein Cross-link of the Senescent Human Extracellular Matrix.

Posted by: CANanonymity at January 27th, 2016 7:41 PM

So, first of all, I think we've moved off of the bacterial enzyme search path, yes? That was my main point.

Turning to significant, systemic changes to collagen metabolism to deal with glucosepane seems like a really risky move with far-reaching consequences. Dynamite to deal with gophers if you like analogies. Leaky blood vessels are just the start. The body is not expecting to go through constant remodeling post-adolescence like you're describing, and God alone knows what barriers you're going to come across trying to muck around with the underlying metabolism like that.

The GBM has a fast turnover rate. That's how evolution has designed it to work. Skin, bone, arteries, and all those other collagen locations ... don't. And trying to get them to change that sounds like an even bigger project than finding a small molecule to promote glucosepane breakdown.

With all due respect to Mr. Furber (in that first article), it was a nice prospective idea, but I just don't think it's feasible.

Posted by: AH at January 27th, 2016 7:55 PM

Okay, that's what I was wondering: MMPs, no matter the amount, cannot function on collagen with glucosepane crosslinks. They untwist collagen until they reach a cross-link that they can't break. And of course leaving collagen half-unmade in the human body cannot possibly be good. Got it.

Now, my question becomes: Can we develop something to finish what the MMPs started? Is there any molecule that can either usurp glucosepane's hold on its binding sites without damaging anything else or simply go around it, unmaking collagen in an entirely different way and breaking off the glucosepane-infested chunks?

Basically, if existing collagenases don't work, can someone build a better collagenase?

Note that I'm not talking about targeting glucosepane by itself; I'm talking about targeting cross-linked collagen, a much larger and seemingly (to my admittedly layman's eyes) much more accessible target. (I know that such a treatment would have to be administered very, very slowly to the very old, lest their capillaries get ripped open.)

I'm of the opinion that, if there is some small molecule that could get into collagen and destroy glucosepane without harming the underlying collagen, it's made of some very interesting atoms and structures not normally found in the human body or we would have evolved it by now. (Although this might also apply to a better collagenase.)

Posted by: Slicer at January 27th, 2016 8:05 PM

Sorry - was writing before you posted.

"So, first of all, I think we've moved off of the bacterial enzyme search path, yes? That was my main point."

Yes, wholly agreed. Bacteria eat everything, collagen and glucosepane alike. They're not picky.

"The body is not expecting to go through constant remodeling post-adolescence like you're describing"

Unfortunately, that might be the end game here, because "constant remodeling" is pretty much what you have to do to stay young, and not just for this. Out with the old (senescent cell clearance), in with the new (stem cell replacement), and a youthful signaling environment to make it work. An aged ECM? If we can't fix it, we've gotta replace it or we die. No other options.

Posted by: Slicer at January 27th, 2016 8:13 PM

In the classical model of collagen breakdown, MMP1, also known as collagenase, is the first enzyme that has to have a bite at the collagen. It breaks the collagen at one particular residue, cutting it into two halves of known size: one is 1/4 the length, the other is 3/4 the length.

After MMP1 is done, MMP2 (gelatinase) comes in and unwraps the strands the rest of the way. But MMP2 can't start until MMP1 does its job.

There are other analogues of these (MMP8, MMP9, MMP13), but they each do those same jobs to different extents.

There are some who argue that this model is wrong, and MMP2 can function by itself. They each have their own evidence. But the standard work on glucosepane - which is the one which we're following if we assume glucosepane is the issue - assumes the model I described above, which is why they use collagenase to see if it's resistant. Typically bacterial collagenase, but it makes the same 1/4-3/4 cut as the human MMP1.

So there's no half-broken collagen molecules. Collagen fibrils and microfibrils, yes, but it's not like you get up to the glucosepane link and then things fall apart. The glucosepane stops the entire breakdown.

And those tangles of half-destroyed fibrils are a big problem for the fibroblasts, yes. Getting the fibroblasts to still work by overstimulating despite the broken fibrils is one avenue people work on, particularly in skin, but you hit a plateau when too much damage accumulates that you never solved in the first place.

Posted by: AH at January 27th, 2016 8:22 PM

Someone who kept up the metabolism and growth rates that they had when they were 6 for their whole lives could end up like those sad circus freaks who get to be 9 feet tall, need crutches to walk, and die young. You can't just activate stuff like that willy nilly.

There's nothing wrong with your biology from age 20 - 50 or so. You don't need to be that radical. I see the premise and promise of SENS as external engineering interventions here and there to fix and patch this and that and let the body keep chugging along, not to reinvent us as green-skinned photosynthetic supermen who can breathe underwater. Mostly because the former is something which could happen in my lifetime, and the latter probably won't.

Posted by: AH at January 27th, 2016 8:32 PM


Nothing ventured, nothing gained.
I like that gopher analogy, makes me think of the one ''using a bazooka to kill a fly''.

I'm not saying we must test this humans but testing faster ECM turnover rates in older adult mice, cat's, dogs, pigs, naked mole rats, apes and other close-related mammals to human to see how bad it gets, and if it's possible to revert to pre-adolescence or, at the latest, very-early adulthood levels.
Then, if it didn't kill them, it would be to determine the effect in voluntary humans with lowest levels/very small increments increase of turnover and checking vitals, arterial leaking, inflammation or unforeseen complications.

I know that evolution chose that this metabolism goes down in adulhood as adulhood phenotype is different than young one and 'has changed/progressed/accumulated damaged which it adapted to', so it seems irreversible to 'go back', and 'erase the past up to now', to our young self phenotype.
I have a difficulty accepting that pre-adolescence works perfect with higher ECM turnover and growing, and none dies from that either; but in old people it would kill them because they are metabolically slow/arrested. The body will adapt, it may take a while, it could be deleterious only at the start and then, after adapting, the effects would be apparent, glucosepane would dramatically halt and plateau just like in GBM with faster turnover; I don't think 'reving up' the MMPs to adolescence levels will kill anyone below 80; above frailty may be a serious issue.

Posted by: CANanonymity at January 27th, 2016 8:42 PM

Then the question becomes "Are half-destroyed fibrils a viable target?" If we have an identifiable structure that's only still there because of glucosepane, is it feasible to remove the structure without causing physical damage and cause regrowth to replace it?

We need a handle: some way of saying "this is something that is damaged, and here's how we can remove it." And it's normally easier to get a hold of a large thing than a small one. (Nature is already getting rid of the largest things. They're called "old people".)

And, to be clear, in the long run, you have to remove senescent cells, restore cells as quickly as they're lost, repair (to stable condition) or replace the ECM when it's contaminated, remove amyloids as they accumulate, and scrub out arterial plaques. This isn't "growth" in the youthful sense, but it is growth of new cells, whether these cells are directly derived from an external source (you will have to receive stem cell injections over the course of your life or you will die of age-related something as long as you're in a biological body- there is ultimately no way around this) or not.

We need to keep what we have when we're 20. And there obviously is a problem with our biology between 20 and 50 because, during that time, things are getting gradually worse! Age doesn't just happen all at once, but aspects of aging do build on each other, and we'll ultimately need to end them all or we're on an increasingly slippery slope.

CAN: We've already discussed this: extra MMPs simply will not help. Increasing turnover rates through MMP doesn't work because MMP doesn't work. And those animals you mentioned don't have the glucosepane problems that humans do, because they don't live long enough for this crud to accumulate.

I'm curious how or if the long-lived great whales deal with it, though, or if they're just as biochemically flummoxed as we are.

Posted by: Slicer at January 27th, 2016 8:59 PM


Then I guess ECM turnover increase could be dead end, if we will suffer from gigantism because our growth hormones are pre-adolescence levels. Perhaps, growth hormones inhibition or pinealectomy while doing ECM turnover increase could block excessive hormones production , just like in growth hormones knock-out mice. Or perhaps , we don't need 6 year old dire levels, but just ECM turnoverlevels from, at latest, a vigorous young 21 year old will do.

Posted by: CANanonymity at January 27th, 2016 9:08 PM

Slicer touches on a very significant problem for glucosepane research: no good animal models. Rats and mice have negligible levels of glucosepane, even when diabetic (100 and 300 pmol/mg respectively, vs 2000 and 4000 for humans). Even long lived naked mole rats aren't much better. So short of human trials, it's really hard to test this out.

Glucosepane was only identified in 2002, so we haven't seen many cross-species studies on it. They usually need to use HPLC to test for it, also (solubility, collagenase resistance, etc. are only proxies), as the immunological assays are really spotty.

Posted by: AH at January 27th, 2016 9:23 PM


We know that MMPs can't degrade glucosepane crosslinks, but GBM (glomerular basement membranes have less crosslinks than collagen types II-VII, it is closer in resemblance to collagen type I, which has the least crosslinks.But, more importantly, GBM has faster turnover than collagen and of course, it has lower crosslinks quantity and lower rate of accumulation, including glucosepane. So, MMPs or not, it's clear that increaed ECM turnover slows glucosepane and crosslinks, with or without MMPs.

Posted by: CANanonymity at January 27th, 2016 9:28 PM


What about trying in apes, monkeys, chimpanzees, rhesus macaques, bonobo, ... these models are our closest match ? Cows, horses ? Some chimps live very long too so should be good models.

Posted by: CANanonymity at January 27th, 2016 9:41 PM

Once we have removed intracellular junk and closeby extracellular junk through nanorobots, it will be to see if nanorobots are capable of reaching glucosepane in collagen or if it's too stiff to move in it, there will be a need to create a delivery method to reach glucosepane, if that glucosepane breaker must be found and if that fails too,
Well let's just there many options left,
other than redox modulation as last saving grace.

Posted by: CANanonymity at January 27th, 2016 10:09 PM

Is there any glucosepane research in dogs? Are they good models for developing glucosepane breakers?

Posted by: Antonio at January 28th, 2016 2:11 AM

@Reason or Michael, what do you think of AH's objections? Is glucosepane cleavage doomed to failure due to the danger of sufficiently powerful enzymes destroying the rest of the cell? Can these enzymes be delivered effectively to a densely packed environment like the collagen fibrils? Please say it isn't so!

Posted by: Morpheus at January 28th, 2016 3:32 AM

@AH: Given a way to create glucosepane to order, which now exists, it should be possible to generate useful mouse models. The researchers involved would have to infuse the glucosepane; if it is present, it should form cross-links.

Posted by: Reason at January 28th, 2016 5:00 AM

@Antonio - i'd hazard a guess that no one really knows if dogs are good models for human glucosepane, since decent antibodies for detecting glucosepane don't yet exist, and the chemical extraction of glucosepane is expensive without these.

Posted by: Jim at January 28th, 2016 5:38 AM


First, let me compliment you on your excellent blog. It's been a very valuable resource, and I appreciate the work you've put into advancing the cause.

Although glucosepane has recently been synthesized in vitro (I assume you mean by Draghici et al), I find it unlikely that work can have applications for animal models. Simply infusing glucosepane will not form the collagen crosslinks in question.

The simplest structure of glucosepane is a lysine residue linked to a arginine residue by a degraded form of glucose (which has turned into the imadazolium ring, etc). The formation process involves transitions through several stages, starting with the glucose bonding to the collagen lysine, going through Schiff base, Amadori rearrangment, and finally capturing the nearby collagen arginine to complete the structure. Glucosepane, by definition, must have two residues (signified by R1, R2) attached in the respective sites to be called, well, glucosepane.

Any synthesis of glucosepane is going to involve forming a structure already linked to some two residues which - since it was synthesized ex vivo - cannot be the two residues in the in vivo collagen. Simply infusing the glucosepane is not going to cause those two residues to 'pop off', and the remaining stuff in the center to reform as crosslinks on the two residues in the collagen that you'd like them too.

Posted by: AH at January 28th, 2016 7:46 AM


As I said, very little cross-species work on glucosepane. Pentosidine has been known for longer, and is easier to detect as it fluouresces. Its rates increase with age in humans at a similar rate to glucosepane, though the absolute amount is much lower than glucosepane (maybe 5%; 100 pmol/mg at 100 yrs). You can see pentosidine levels for various animals here:

As you can see, their accumulation rate is actually faster than humans (there's that metabolic mismatch problem again) but since they're all dead by their early teens, pentosidine levels never approach that of humans. Monkeys die in their twenties with pentosidine levels similar to humans in their twenties. Humans don't have serious collagen related issues, and neither will monkeys. Inducing diabetes might help, though (that sounded wrong...)

Posted by: AH at January 28th, 2016 8:02 AM

On the other hand, diabetes will kill them sooner, so you'd have less accumulation time. Oh well.

Posted by: Ah at January 28th, 2016 8:10 AM


Following up on Morpheus, and skipping the animal model sideline, do you see the issues I've brought up regarding the bacterial enzyme search project? I'm focused on small molecule screening for those reasons (and happen to think I have a new angle on it), but would like to know if there is a good counter-argument in favor of enzymes to consider.

Posted by: AH at January 28th, 2016 9:33 AM

Our project is exploring AGE breakers and are considering small molecule and Enzyme solutions. Thank you to Reason for putting us in contact with the Yale lab btw.

It is possible a mouse that is diabetic could be useful for AGE work this could be induced or selected for in a suitable strain.

The MMTP would consider either an Enzyme, small molecule or even increasing ECM turnover as possible solutions, the important thing is getting the work done and seeing what works! We will be in a position to do this in the coming year with our first funding phase starting in March if remain on track.

Posted by: Steve H at January 28th, 2016 10:32 AM

"I'm focused on small molecule screening for those reasons (and happen to think I have a new angle on it)"

If you actually do create and patent a small molecule that can safely remove glucosepane in vivo, you will instantly become unfathomably, laughably rich.

Posted by: Slicer at January 28th, 2016 10:39 AM


Diabetic mice have glucosepane levels of 300 pmol/mg. Following the accumulation curves for humans, those are the levels you'd expect to see in an 8-year old (non-diabetic). So we're not talking about extensive cross-linking, certainly nothing which would have physiological manifestations.

What is the MMTP?

Posted by: AH at January 28th, 2016 10:48 AM


You'd be surprised how difficult it is to get people to beat a path to your door, even when you do have a better mousetrap

Posted by: AH at January 28th, 2016 10:49 AM

@AH: On the small molecule versus enzyme thing, I'm absolutely unqualified to comment, though I understand the issues you raise. This is really something I'd like others with more knowledge of the practical side of getting these things to work to comment on.

Posted by: Reason at January 28th, 2016 11:36 AM


Thanks for the link. I suppose you could still measure glucosepane clearance if you found something that works, but given the animal-human mismatch I don't think you'd see any clear effects on morbidity or mortality, like they're looking for in that project.

Maybe it's a good plan for other areas of SENS research. I don't know much about those areas.

Posted by: AH at January 28th, 2016 11:36 AM

@Michael or Aubrey:
Reason has thrown in the towel on AH's question of whether bacterial enzymes capable of cleaving glucosepane would also destroy the entire cell. AH proposes that small molecule drugs be used and be more easily delivered into tightly packed collagen fibrils. He states that he is screening small molecules and may have some sort of improved method. Do you agree with him that such enzymatic failure may be likely? In the manner of SteveH and the Major Mouse Project, perhaps SENS should talk to AH or at least be aware of a possible alternative method that is superior to the enzymatic approach?

@AH, would you care to elaborate on your findings? Or if not, perhaps direct communication with SENS officials could aid you similar to grants to Oisin Biotechnology, Organovo, etc. to spur your research.

Posted by: Morpheus at January 28th, 2016 5:51 PM


I should have been clearer, as I don't mean to overstate my progress... I've got an idea or two, but it's still pretty early stage. No physical screening going on yet, and no concrete suggestions for things the MMTP could run on their murine subjects.

If I ever get a candidate small molecule I'm confident in, but am short on capital, I may give that a try.

Or maybe the enzyme guys will show me up. I'll have no complaints if they come through. (Ok, some complaints, but overshadowed by the fact that they crossed off a SENS target)

Posted by: AH at January 28th, 2016 7:12 PM


'' I'm curious how or if the long-lived great whales deal with it, though, or if they're just as biochemically flummoxed as we are.''

It seems it's more the latter, they too accumulate crosslinks and face MMP crosslink block problem but the impact seems not as substantial although that depends on which type of whale, age, nutrition, body size, blubber fat mass and total mass (increased ECM mass with total mass increase). For example, beluga whales live rough 50 some years, they are short-lived whales, some of them get type II diseases and as such, show diabetic human-like features of accelerated collagen crosslinking by AGEs production during glycation/glycoxidation to their hyperglycemia.

In contrast, Bowhead whales are the longest-lived mammal at 211 years MLSP. A study of the transcriptome of this whale's organs showed that it has better insulin and whole-body glucose disposal (this alters longevity through the IGF-1 axis) as such its pancreatic Beta cells secrete enough dosed insulin to reduce fasting and post-prandial blood glucose levels. This careful normoglycemia, nearing border controlled hypoglycemia, render blood glycation/glycoxidation almost to very low levels (we can infer its glycated albumin and glycated hemoglobin HbA1c are kept low, another study verified blood glycemia in pregnant and non-pregnant bowhead adult females and their levels fell within non-diabetic normal human levels (4 to 6 mmol/L, 72 to 108 mg/dL glucose). This means no diabetes in Bowhead whales, obviously since diabetes is accelerated aging of the ECM, such as species pentosidine accrual age curves show and same thing for crosslinks like glucosepane. Maintaining normo or mild hypoglycemia is necessary to avoid most glycation/glycoxidation and slow aging, such is Bowhead whales obvious trick to reach 200 years. Same thing in giant turtles who normoglycemic or hypoglycemic for their entire life; Centenarians have reduced blood glucose levels and reduced blood glucose increase over the years, their blood glucose is monotonic plateau, where as elders dying below 80s show blood glucose rise each decade until their death, their pancreatic Beta cells can't secrete enough insulin and they are mildly pre-diabetic/insulin resistant.

Here is a bit that shows whales are no different on collagen crosslinks dynamics, common minke whales live 30 to 50 years max. What is clear is that redox-insulin axis is better in bowhead whales and that transcriptome study showed they have better vascular endothelial function and thus, lower arterial collagen ECM stiffening (by reduced crosslinks and AGEs glycation/glycoxidation of their vasculature :

'' Additionaly, starving fish have more crosslinked-collagen than do well-fed ones (Love et al., 1975, Sikorski et al., 1990). However, fish skin-collagen has an extremely low crosslinking rate and a highly crosslinked molecule is rarely found (Cohen-Solel et al., 1961).
It is to be noted that (alpha1)2alpha2 heterotrimer [a crosslink] is seen in the collagen of the common minke whale (Nagai et al., 2008) ''

Posted by: CANanonymity at January 28th, 2016 8:55 PM

We could almost certainly create a useful mouse model for AGE and measure it's levels if we had a breaker. We benefit from favourable rates in Germany for testing and housing so it would be easy for us to run a breaker study given something to work with. I mentioned diabetic predisposed mice as a possible approach but I am certain something better could be created.

Posted by: Steve h at January 29th, 2016 12:57 AM

To clarify we would be testing mice specifically for glucosepane cleaving in such a test rather than lifespan. We know that AGE does not reach pathological levels in mice making lifespan work pointless.

Posted by: Steve h at January 29th, 2016 1:07 AM

So, there's a lot here, of course ... I'm going to address what seem to be the main points of confusion:

First, AH, thanks for engaging here: you're clearly highly knowledgeable in this area, and your input is welcome (as would be direct communication with us, particularly if you're engaged in intervention-oriented research on glucosepane, especially since you're doing so using an approach consistent with the "damage-repair" strategy of SENS).

It is of course correct that glucosepane is actually pretty labile: as you probably know, the reason it went undetected for so long was exactly that the harsh treatments and broad-spectrum proteases normally used to break down aging collagen to look for nonenzymatic crosslinks destroy it in the process. It was only identified through a painstaking process of sequential enzymatic digestion of the native collagen. This is one of the many reasons why having a ready source of synthetic glucosepane is valuable: by incorporating it into collagen peptides, it can be used to generate antibodies for the convenient labeling of glucosepane in vitro and in vivo.

A viable glucosepane-cleaving rejuvenation biotechnology, then, must to selectively cleave glucosepane crosslinks without damaging the native collagen. This would both directly restore its youthful elasticity, and should also allow for normal, physiological turnover of the tissue (such as by restoring MMP1's ability to access to its binding site, which the molecular dynamics simulation of glucosepane formation linked by Slicer suggests is likely impeded by glucosepane because its favorable formation site is only two amino acid residues away).

This is also the resolution of the conundrum highlighted by Slicer: [Since] MMPs, no matter the amount, cannot function on collagen with glucosepane crosslinks ... Can we develop something to finish what the MMPs started? Is there any molecule that can either usurp glucosepane's hold on its binding sites without damaging anything else or simply go around it, unmaking collagen in an entirely different way and breaking off the glucosepane-infested chunks?

Neatly cleaving glucosepane from the native structure will achieve this aim in a more direct and durable way than trying to unbind it or work around it, and (as noted) will not only allow for the turnover of the previously-crosslinked collagen fibril, but also restore its motion and function, likely rendering turnover unnecessary unless other damage has accumulated.

It's quite true that glucosepane accumulates much more slowly in normally-aging and even diabetic rats than it does in human aging. This isn't really an impediment to using them for basic testing of candidate glucosepane-cleaving agents, however. Even if glcosepane's low tissue burden in rats means that cleaving it will not have dramatic rejuvenating effects in these animals (which is a reasonable prediction, but might be happily disproven in the event), its high prevalence in aging and diabetic human collagen, and its implication in the complications of diabetes, will make the mere demonstration of a candidate's ability to cleave glucosepane crosslinks in vivo a sufficient proof-of-concept to spur further work to move it down the therapeutic pipeline into human testing.

(It's perhaps worth noting that even if glucosepane did accumulate in rats in a way that more closely scales to their rate of aging, cleaving it wouldn't necessarily extend their lifespan, because of the "weakest link in the chain" problem: to extend the lives of otherwise-healthy, normally-aging mice or humans by definition requires addressing all of the cellular and molecular damage driging aging to some degree, using a comprehensive panel of rejuvenation biotechnologies. As Chief Science Officer de Grey wrote to address a canard along these lines from critics:

[It is true] that no SENS intervention—in isolation—has ever been shown to extend any organism's lifespan. I do not recall Henry Ford alerting potential customers that the components of a car—in isolation—remain obstinately stationary when burning petrol is poured on them, nor do I recall his being castigated for this omission.

And in the meantime, the free availability of the anticipated glucosepane-detecting antibodies generated by the research that SENS Research Foundation funded at Yale will enable the identification of more suitable animal models for the demonstration of a hard health outcome. It's even conceivable (though not likely) that the mouse might be one such: the 300 pmol/mg figure cited by AH is for rats, and conceivably mice might prove a closer parallel to human aging and diabetes in this regard). On this front:

AH, you noted that Although glucosepane has recently been synthesized in vitro (I assume you mean by Draghici et al), I find it unlikely that work can have applications for animal models. Simply infusing glucosepane will not form the collagen crosslinks in question.

That wouldn't be the idea. The ability to synthesize glucosepane is now enabling the Yale crosslink team to incorporate glucosepane into synthetic, chemically-uniform crosslinked peptides. These can be used to create antigens with which to immunize rabbits, inducing the formation of antibodies targeting glucosepane-containing peptides. Labeled monoclonal antibodies generated from hybridomas derived from B-cells from such animals could then be used for the convenient detection of glucosepane crosslinks in tissue samples, and then evaluate the effects of candidate glucosepane-cleaving rejuvenation therapies in vitro and in vivo.

You (AH) also express concern that bacteria and other microorganisms just need to find something which works [to degrade glucosepane]. It's not their problem if the solution they use won't help us for the odd situation we're trying to solve."

Right ... the idea, however, is to find bacteria that can grow under physiologic conditions (temperature, pH, etc) with glucosepane as the sole available energy source, or alternatively as the sole available source of lysine and arginine for strains auxotrophic for these amino acids. In the wild, a community of bacteria and other microorganisms will often cooperate or compete for energy sources, leading to strains that are specialized in attacking specific bonds within a larger structure to carve out a niche for themselves (or a role within a larger network). It is then trivial to negatively screen candidate enzymes identified through such methods against physiologic collagen and later other proteins.

In addition to wild-derived bacteria, the Yale group is working to construct a cosmid library (see also here) to screen the huge diversity of gut microbe enzymes in easily-cultured E. coli lines, which would then similarly be screened against glucosepane as the sole source of energy and/or lysine and arginine.

Another issue with enzymes is the tight packing of the collagen fibrils, making it questionable if the enzyme will be able to reliably reach its target. A tailored small molecule may be a considerably better option for that reason alone.

Entirely possible, yes. The advantage of a microorganism-based enzyme-discovery program is that the enormous diversity of solutions generated by evolution (both in nature and later through directed enzyme evolution) allows for the hypothesis-neutral testing of an enormous range of candidate therapies generated by the engines of evolution, which are much cleverer than I am ;) . At the moment, I don't think anyone has much of an idea on how to identify good small molecule candidates with conventional medicinal chemistry methods (though I'm intrigued to hear that you have some thoughts, AH!).

I'm a also going to quote from a recent email from Dr. de Grey, from a conversation on this subject in which I was a participant:

the access issue for enzymes reaching crosslinks in tightly-wound collagen is one that has been on our mind since forever, and is in tension with the converse problem of specificity of small molecules. The fact that glucosepane looks so weird, seven-membered ring etc, does not translate as well as I had initially hoped to the conclusion that drugs to cleave it will often be harmless to structures that the body synthesises on purpose. For this reason one approach that we are looking at is organocatalysts, which are small peptides with enzymatic activity - they are a compromise between size and specificity, and thus may be the best of both worlds.

But there is a huge amount not known about all this; it may just be that regular enzymes work fine because breaking the subset of glucosepane that they can get to will loosen things up enough that they can get to a bit more, and so on.

No matter what the strategy one favors for identifying therapeutic candidates, the glucosepane reagents generated by David Spiegel's group through the SENS Research Foundation funding will for the first time provide the soil that will allow a thousand flowers to bloom. Like you, AH, and I'm sure like everyone here, I'd be delighted to make progress toward a viable glucosepane-degrading therapeutic based on on any strategy, though it would sure be nice to claim a part in the bragging rights for that, too ;) . Solvitur ambulando!

Posted by: Michael at January 29th, 2016 1:53 PM

Thank you, Michael for another great response to everyone's concerns and keeping the hope alive.

Posted by: Morpheus at January 29th, 2016 6:37 PM

I wonder if you can get organocatalysts produced in bacteria, and if you could apply the same high throughput automated scan for one that breaks glucosepane. And if you could use directed evolution to improve the organocatalysts?

Maybe put the glucosepane behind a barrier in the wells that is to big for enzymes to pass through but not to big for organocatalysts to pass through.

Posted by: Jim at January 31st, 2016 9:18 PM

I'm working on an ex-vivo tissue engineering project to provide a model environment for testing glucosepane breakers.

Posted by: Chris Rose at June 15th, 2016 4:42 PM

Post a comment; thoughtful, considered opinions are valued. Comments incorporating ad hominem attacks, advertising, and other forms of inappropriate behavior are likely to be deleted.

Note that there is a comment feed for those who like to keep up with conversations.