WILT - SENS Research Foundation

5 downloads 79 Views 129KB Size Report
disposal, cancer is arguably the hardest part of aging to combat biomedically. ... major age-related killers is a demonstration not that the war on cancer has failed ...
1

WILT: necessity, feasibility, affordability Aubrey D.N.J. de Grey, Ph.D. Methuselah Foundation PO Box 1143, Lorton, VA 22079, USA Email: [email protected]

Indexing terms: cancer, aging, nuclear DNA mutations, Whole-body Interdiction of Lengthening of Telomeres, telomere, telomerase, Alternative Lengthening of Telomeres, gene therapy, gene targeting, stem cell therapy

Abstract Despite immense investment of resources, biomedical progress in postponing death from most cancers has fallen far short of prior expert prediction. Having intercellular natural selection at its disposal, cancer is arguably the hardest part of aging to combat biomedically. WILT (Whole-body Interdiction of Lengthening of Telomeres), first suggested in 2004, is a radical proposal that seeks to address this feature of cancer head-on, by pre-emptively altering as many as possible of our mitotically competent cells in such a way that the capacity for indefinite cell division could not be achieved even by the high degree of mutagenesis and selection that a tumour harbours. WILT also incorporates proposals for addressing the severe side-effects that these alterations would certainly have. In this chapter, following an introduction to the motivation for WILT, I first evaluate the likelihood that alternative, less daunting but comparably effective approaches to controlling cancer will emerge in a similar timeframe, thereby making WILT mercifully unnecessary. Then I provide an update on the various technologies that comprise WILT, with emphasis on progress showing that WILT is likely both to be implementable within a few decades and to achieve the anti-cancer efficacy that I have previously claimed for it. Finally I address a different type of concern: that, even if technologically achievable, WILT is so complex that it may never be economically practicable.

1. Cancer as a component of aging: the challenge Those of a certain age, or a certain profession (oncology), will recall as if it were yesterday the words with which President Richard M. Nixon announced the launching of the War on Cancer (Nixon, 1971). The centrepiece of his 1971 policy initiative was a sharp rise in the budget of the National Cancer Institute, and the motivation for it was the expert advice he had received to the effect that such an investment had a good chance of leading to a comprehensive cure for cancer within as little as five years. As we all know, it didn’t happen. The fact that public and philanthropic funding for cancer research have nonetheless been robustly sustained (National Institutes of Health, 2006) is a cogent and telling rejoinder to those who claim that we must always lower expectations of the rate of biomedical progress for fear that dashed hopes will dissipate public support – but that is not my topic here. I wish to focus instead on why it didn’t happen – and we can say with hindsight that there is one overwhelming reason, a reason that biologists have in fact known for a century: in Crick’s succinct words (Dennett, 1984), evolution is cleverer than you are. The more we discover,

2 and subvert, particular tricks that cancer cells have employed to evade the assaults on them by the body’s natural defences and by our medical efforts, the more we discover that, deprived of that trick, the cancer will simply turn up a new one. 1.1. The geriatric illusion: comprehensive progress presenting as stasis But there’s more. This depressing reality is bad enough today, but there is a very clear and present danger that it will get even worse – much worse – in the decades to come. Not because of increased exposure to carcinogens, but because our progress against other causes of death may well accelerate sharply. It is incorrect to conclude that, because just as high a proportion of Americans die of cancer now as in 1971, the War on Cancer was a complete failure. There was a much cheaper way to cause a decline in cancer deaths, and one that would have been much surer of success: to close down research on cardiovascular disease, thus slowing our progress in postponing death from that cause and thereby “saving” huge numbers of people from living long enough to die of cancer. This policy was not adopted, even though death from a myocardial infarction occurs, as often as not, without a hint of prior symptoms – just the sort of “compression of morbidity” that so dominates the stated aspirations of influential geriatricians and gerontologists alike (Kalache et al, 2002). Lest you suspect that I feel we should indeed have rowed back on research into the postponement of death from heart attacks, let me reassure you that I do not: firstly because many people suffer severe symptoms from cardiovascular disease for much longer than the average time that one suffers from cancer, and secondly because such people generally seem rather to prefer still being alive, even in a diminished state of health, than being dead. But to return to my theme: the absence of a significant change in recent decades in the proportion of deaths from cancer versus other major age-related killers is a demonstration not that the war on cancer has failed, but simply that it has succeeded to roughly the same extent as have the less loudly declared wars on those other diseases. Let us, then, extrapolate these achievements of the recent past into the future. And let us do so not by simple mathematical extrapolation, but by examining the precursor technologies, currently only being explored in the laboratory but with the potential for clinical application less far in the future than Nixon’s speech is in our past, that may postpone the industrialised world’s major killers other than cancer. For, just as in decades gone, it is our achievements in those fields that will define our goal in the cancer field: if progress against cancer lags behind, the proportion of deaths from cancer will rise. 1.2. SENS: a plausible route to hugely accelerated progress Unfortunately for any complacent oncologists out there, the prognosis on this point is (from their point of view) not good. The underlying molecular and cellular causes of cardiovascular disease, neurodegeneration and type 2 diabetes (to name just three) are still very incompletely understood at the mechanistic level, but they are really rather well understood at the structural level. That is to say: in each case there are clear molecular and/or cellular changes, occurring in the body throughout life, which are confidently understood to be (jointly) necessary and sufficient precursors of these diseases. What is less (albeit increasingly) well understood are the mechanisms whereby our normal metabolism ongoingly gives rise to these changes and those whereby the changes eventually give rise to the diseases. The same applies to the “susceptibilities” that we acquire in later life – changes such as loss of muscle mass and decline in immune function, which we do not honour with the status of “disease,” but which play just as great a part in our increasing risk of imminent debilitation and death. Your first reaction to the above may be to be unimpressed: after all, one oasis of knowledge flanked by two vast expanses of ignorance does not sound much like a basis for optimism. Closer inspection, however, leads to an altogether sunnier conclusion. Since – as for any machine – the

3 body’s function, including its likely remaining longevity, is determined wholly by its structure, we can make the following bold but undeniable statement: reversal of the ongoing molecular and cellular changes to which I have just referred, restoring the structure of a middle-aged adult’s body to what it was at a younger age, would comprehensively uncouple the causes of those changes from their consequences. It would postpone pathology, simply by severing the chains of events that lead to that pathology. And it would do so whatever the (incompletely understood, as just noted) details of how those molecular and cellular changes either come about or give rise to pathology. In short, reversal of the structural changes occurring in the body throughout adulthood would thoroughly sidestep our ignorance of the detailed mechanistic basis of aging. The above would, of course, be of purely academic interest if we did not in fact have any prospect of achieving this “structural rejuvenation” of the body – but we do. Some years ago I proposed (de Grey et al, 2002) that all such changes can be classified into seven major categories, for five of which a highly promising approach to reversal indeed exists. I refer the reader to my past publications for details; Table 1 provides an overview. This classification appears to be standing the test of time: indeed, it could have been made a quarter of a century ago, since all seven types of “damage” were by 1982 the subject of widespread suspicion within biogerontology with regard to their contributions to aging and age-related pathology.

Type of damage

Proposed repair strategy

Cell loss, cell atrophy

Stem cells, growth factors, exercise

Senescent/toxic cells

Ablation by immunisation or suicide gene therapy

Intracellular aggregates

Microbial hydrolases (transgenic or by enzyme therapy)

Extracellular aggregates

Immune-mediated phagocytosis

Extracellular crosslinks

AGE-breaking molecules or enzymes

Mitochondrial mutations Nuclear [epi]mutations Table 1: the seven major types of molecular and cellular damage that accumulate in aging, and foreseeable ways to repair five of them. For details, see de Grey et al. (2002). Five out of seven ain’t bad, but one might be justified in the view that it also ain’t good enough. My story is not over, however. The sixth category of damage, mitochondrial mutations, is in my view not likely to be amenable to outright repair (such as replacement of mutant mtDNA by nonmutant genomes), but there now seems to be a very strong prospect that such mutations can instead be obviated: made harmless, in other words, by the elimination of any pathways whereby they might contribute to age-related pathology. Obviation is a conceptually tricky business, since (as just mentioned) we know so little about how damage actually translates into pathology; but if an intervention targets the very first step that any such mechanism must involve, the prospects are good. Accordingly, the approach of “allotopic expression” – developing modified versions of the 13 protein-coding genes of the mtDNA, which can be introduced into the nuclear genome by gene therapy and complement any mutations that may occur in the mtDNA itself – seems highly likely to subvert any consequences of such mutations (de Grey, 2000). Six down, one to go. Mutations of the nuclear genome (or epimutations, defined as adventitious changes to the modifications of bases, histones or chromatin that determine gene expression) can of course cause cancer, and in principle they can also cause all manner of other failures of cellular

4 function. If such changes result in the cell’s death, or in its adoption of a distinctive state causing it to resist normal apoptotic signals, SENS already has them covered (see Table 1). In principle, however, such changes could be altogether more non-specific – degrading the performance of one cell in one way, the next in another, and so on. This might be daunting, or even impossible, to reverse. But here we have reason, paradoxically, to be grateful that we are susceptible to cancer. A single cell possessing the “right” constellation of mutations can kill us rather quickly by dividing uncontrollably, but dysfunction not related to cell cycle control can only be harmful if it affects a substantial proportion of the cells in a given tissue. This numerical disparity is only reduced, not eliminated, by the fact that we have many cancer-specific defence mechanisms. Thus, there is good reason to believe that cancer is the only consequence of nuclear mutations or epimutations that has pathological consequences in a currently normal lifetime (de Grey, 2007) 1.3. The unpalatable prospect: cancer as our main killer I can now, finally, return to the central topic of this section. The bottom-line conclusion of the foregoing analysis is that all aspects of age-related pathology other than cancer are – not certainly, of course, but quite probably – destined for comprehensive defeat within only a few decades. That is clearly a huge acceleration in progress when compared to the past several decades of medical advance against such scourges. Let us now suppose that, by contrast, progress in postponing the onset of and death from cancer continues only at a rate comparable to that seen in recent decades. I hardly need spell out the result that will befall us: a rather modest rise in life expectancy, combined with a calamitous rise in the proportion of our population who die of one of the most gruesome age-related diseases known to humanity. It was the appreciation of this prospect that led me to acknowledge the need to explore, now, all conceivable avenues for the avoidance of such a future by the identification of anti-cancer modalities that might allow the postponement of cancer to keep pace with the postponement of the rest of aging. The result was WILT.

2. How can brawn reliably defeat brain? Tumours are constantly evolving, and evolution is cleverer than we are – but we have many tools not available either to the cancer cell or to our inbuilt anti-cancer defences. How can our technological superiority overcome the cancer’s greater ingenuity? WILT is described in detail elsewhere (de Grey et al, 2004; de Grey, 2005) so I provide only a brief summary here. The central idea is that mutations which alter the level of expression of genes are frequent, but the creation of entirely new genes is a vastly more intricate process – so intricate, indeed, that it will virtually never happen in a collection of “only” a trillion cells in a period of “only” a few decades. Thus, deletion from all our cells of genes necessary for extending telomeres would render tumours unable to transcend the Hayflick limit whatever genes they turned on or off, with the result that they would be initiated just as they naturally are – indeed, very probably more often than naturally, because critically short telomeres are highly mutagenic and oncogenic – but would wilt away before reaching a clinically problematic size. Selective targeting of cancerous or precancerous cells for this modification would be inadequate, because some such cells would already possess mutations that protected them from undergoing the modification; hence, it must be done pre-emptively to as many of our cells as possible (or, at least, the mitotically competent ones). This would clearly have the rather important side-effect of also rendering mortal the stem cells of our continuously renewing tissues (blood, gut, epidermis and lung) – but such problems could in principle be averted by replenishing these stem cell pools every decade or so with autologous stem cells that also lacked telomere-extension genes but had had their telomeres extended ex vivo with exogenous telomerase. This concept is depicted in Figure 1.

5

Figure 1. The WILT concept. Metastatic cancers kill a typical untreated patient (a) rapidly and a beneficiary of contemporary treatments (b) only slightly less rapidly, on account of the genomic instability and inventiveness of such tumours. Uncompensated deletion of telomerase genes (c) prevents this but kills much sooner by preventing the maintenance of continuously renewing tissues such as the blood. In WILT, deletion of telomerase genes is combined with periodic stem cell therapy to maintain such tissues indefinitely, so that death from either cause is avoided (d). The feasibility of WILT, even with the benefit of decades of future advances in gene therapy and stem cell therapy, is clearly very far from obvious. In the next section I will discuss some of the more daunting challenges to implementing it. By way of concentrating minds, however, in this prior section I critique the “one more push” anti-cancer philosophy: the idea that we need not explore such radical approaches as WILT because altogether less scary alternatives will almost certainly emerge just as soon. Optimists on this point can be divided into two broad camps: those who pin their hopes on therapies that are truly clever enough to exhaust a cancer’s options, and those who favour a “blunderbuss” attack that requires the cancer to solve a multitude of challenges simultaneously. 2.1. Cleverer than we are – much cleverer The “cleverer therapies” approach appeals strongly to the oncologist’s self-esteem and is thus seductive – but a moment’s reflection reveals that it is precisely the approach we have been following all along and with such modest success. Current thinking on such matters, when analysed dispassionately, reveals how frighteningly cancer research has departed from all normal levels of theoretical rigour in prioritising research directions. The most prominent current example of this is the mania over so-called “cancer stem cells” – cells that, though in a minority in a typical

6 tumour, probably give rise to the bulk of that same tumour’s cellular content a few years down the road because, just like normal stem cells, they can divide without losing much of their capacity to divide again (Lobo et al, 2007). Because a tumour’s longevity relies on such cells, great effort is currently being directed at the development of therapies that specifically target them. Er… First of all, the fact that the cancer stem cell may be the required target is no reason to avoid killing the rest of the tumour. But more crucially, the whole reason why current anti-cancer therapies fail to kill cancer stem cells (and thus fail to kill cancers) is that cancer stem cells very closely resemble cells that we absolutely must not kill too many of with our therapies, namely our normal stem cells. As such, the contemporary focus on cancer stem cells is not really a new idea at all, but merely a restatement of the problem that oncologists have always faced, viz. that it’s easy to kill a cancer cell, but not to kill it selectively. A potential way out of the above might be thought to be provided by another increasingly fashionable “cleverer” approach in cancer research: the improved characterisation of individual tumours. If a therapy is chosen purely on the basis of where a tumour has arisen and how large it is (and perhaps one or two other gross histological features), there is much room for improvement: two tumours may arise in the same tissue and develop in essentially the same way as a result of quite different spectrums of mutations, and those differences may point rather clearly to differences in the likely relative efficacy of particular therapies. This has motivated the development of highly sophisticated molecular analyses of tumour gene expression, the results of which allow the subclassification of histologically hard-to-distinguish tumours into ones deemed suitable for one therapy or for another (Golub et al, 1999). While such work will indeed save lives, by addressing the genetic basis for the growth of most of the target tumour, the very fact that multiple alternative genetic misadventures can cause the same cellular phenotype is clear proof that in many cases a predominantly “type 1” tumour will simply regrow as a “type 2” tumour when the basis for type 1 growth is medically undercut. Because of this, treatment by improved characterisation can realistically provide only a modest postponement of cancer-related debilitation and death on average: quite probably not nearly enough to keep pace with therapies that combat the other major age-related killers. A particularly clear demonstration of this is provided by the drug Gleevec, which targets chronic myelogenous leukaemia on the basis of a particularly inviting mutation, the Philadelphia translocation. Gleevec’s molecular target is a fusion protein that is not encoded in our non-mutant genome at all, yet is produced by every single cell of a cancer that has suffered this particular chromosomal rearrangement – and, indeed, patients with this type of cancer have been greatly helped by this drug. After several years of use, however, relapses began to be reported: the patients’ cancers had returned, because of secondary mutations that rendered the drug ineffective (Kantarjian et al, 2006). The key point here is that this was a type of cancer that apparently did not have multiple subtypes distinguishable by additional molecular analysis, yet it still revealed itself to have multiple options at its disposal when forced to call upon them. 2.2. Name your poison: one blunderbuss or ten novice marksmen So to the other mainstream approach to radically improving our current anti-cancer arsenal: multiple simultaneous therapies. This concept comes in a number of forms, Most simply, we just apply a variety of different drugs or vaccines that seem likely to undercut different defences that the cancer cell has erected against us. A second approach is to exploit the complexity of the immune system and target multiple cell cycle control pathways via antigens presented on the cancer cell surface. A third, which has been in clinical use for some time, is to “design” an immunotherapy specific to the individual cancer, by using cells taken from a surgically removed primary tumour as antigens for the production of antibodies against residual cancer cells (in particular, metastases) (Berd, 2004). Other variations on these themes can be imagined. Unfortunately, all such “blunderbuss” approaches possess two fatal (in my view) shortcomings.

7 The first has already been indirectly alluded to in the previous section. What makes a cancer hard to kill is not that the individual cancer cells are hard to kill, but rather that they (and especially the “cancer stem cells” that ensure the cancer’s continued growth) are hard to distinguish from other cells on whose survival we rely for our own survival. Another way to state this problem is that cancer therapies have a rather low “therapeutic index” – range of dosage that is effective on the cancer cell but not on non-cancer cells. In reality, of course, a drug’s therapeutic index is not defined at the level of the cell but at the clinical level – and the disparity between the two is different for cancer than for other diseases, because cancer is a disease of cell proliferation. Thus, in the case of cancer we will accept a much higher “false positive” rate (non-cancer cells killed) than we might for other diseases, so long as we have a really low “false negative” rate (cancer cells surviving). But this can only go so far, hence the need to avoid excessive doses of chemotherapy and such like. Now: the whole idea of applying multiple therapies to a cancer at the same time is that they will address distinct aspects of the cancer’s methods for growth. This inevitably means that the individual therapies will incur different types of collateral damage – they will kill different (though typically overlapping) populations of non-cancer cells. Accordingly, each component therapy effectively reduces the therapeutic index of the other therapies with which it is co-administered. The second difficulty with combination therapy is, in a sense, the mirror image of the first. It turns out that some of the most formidable defences that cancers employ in resisting our attacks are highly versatile. One is so versatile that it has even been named for that versatility: it is a membrane transport protein called the multi-drug resistance protein, and it works by simply shuttling out of the cell anything that it doesn’t like the look of (Yasuhisa et al, 2007). (Why not just co-administer a drug that suppresses this transporter, along with the chemotoxic drug? Firstly, cancer cells will simply upregulate its expression to counter the suppression. Secondly and even worse, suppression of the transporter will render non-cancer stem cells more sensitive to the chemotherapy, whose therapeutic index will thus be reduced.) Similarly, though there are many ways to bring a rogue cell to the attention of the adaptive immune system, most of them rely on the target cell obligingly presenting antigens on its surface in the cleft of the major histocompatibility complex 1 (MHC1), and some cells just stop making MHC1. (This is such an obvious trick that we actually have cells, NK cells, to get rid of cells expressing no MHC1 – but cancers have counter-tricks, which exceed the scope of this chapter.) It is perhaps worth looking closely at the specific anti-cancer modality that is conceptually the most similar to WILT: reversible, drug-based inhibition of telomerase (and of genes needed for ALT, as and when they are discovered). Perhaps because of its similarity to WILT, analysis of this approach reveals particularly starkly why something as ambitious as WILT will probably be needed if we are to control cancer indefinitely. First of all, in contrast to the permanent nature of genetic modification, any reversible suppression of expression of a given gene must be actively maintained by continued application of the suppressor (unless the reversal is effected by applying a second drug rathern than just withdrawing the first, which is functionally equivalent to implementing WILT and then reintroducing telomerase by insertional gene therapy). This means that the cell constantly has the opportunity to mutate to exclude the drug. Secondly, the drug will inevitably inhibit telomerase in non-cancer cells where telomere elongation is needed (the stem cells of continuously renewing tissues), so it must be periodically withdrawn to allow those tissues to recover – and that will allow the cancer to recover too. Drug-based telomerase inhibition certainly has its place in the anti-cancer arsenal, as has been compellingly shown by recent results (Keith et al., 2007), but that place is short-term treatment, not permanent escape. I hope, in this section, to have disabused you of any confidence you may have had that gradualist approaches to defeating cancer are definitely going to see us through. I do not claim that these approaches will definitely fail: all I claim is that their prognosis is inadequate to remove the

8 incentive to look for something else. At the time of writing, WILT is the only “something else” on the table. Nothing would make me happier than the discovery that WILT can be consigned to the dustbin of unnecessarily elaborate ways to postpone aspects of aging – but nothing would make me unhappier than the failure to pursue WILT and the subsequent realisation that nothing else has turned up.

3. Feasibility of WILT: selected concerns As noted earlier, WILT consists of two types of therapy: -

deletion of genes required for telomere maintenance; and

-

regular replenishment of stem cell pools in thereby “mortalised” tissues.

Within each of these components lie many daunting challenges. I address a selection of these below. 3.1. Rendering cells telomere-elongation-incompetent in situ Our various cell types can be divided for present purposes into four classes: 1) postmitotic; 2) normally quiescent but mitotically competent: differentiated; 3) normally quiescent but mitotically competent: stem cells; and 4) rapidly dividing and/or short-lived. Postmitotic cells, being constitutively unable to undergo mitosis, do not give rise to cancers and thus are not necessary targets either for telomere maintenance removal or for compensatory stem cell therapy. (I should stress that here I refer only to cells that have undergone gross structural alterations precluding the mechanics of mitosis, such as neurons or skeletal muscle fibres: I do not include cells that are non-dividing solely because they express proteins that suppress mitosis, as in the case of replicative senescence in vitro, since that suppression could theoretically break down.) Category 4 is also a low-priority target: such cells (either the short-lived terminally differentiated cells in continuously renewing tissues, such as keratinocytes or erythrocytes, or else the rapidlydividing but developmentally committed transit amplifying cells that give rise to them) are virtually certain to die before they can accumulate the wealth of mutations necessary to turn them into a cancer, but even more conclusively they are the progeny of cells of type 3 (of which more in a moment) so are bereft of genes that have been removed from type-3 cells. Here I will, therefore, focus on cells of types 2 and 3. It should be noted that some cell types, such as hepatocytes, fall at the boundary between types 2 and 3 – they probably only divide on demand, but that demand is rather high on account of the high rate of cell death. Differentiated cells that divide only “on demand” are present in many tissues. Most are covered by the collective term “mesenchymal cells” – fibroblasts, adipocytes, glia, muscle satellite cells, chondrocytes and such like. Certain epithelial cells also fall into this category, however: hepatocytes, for example. Such cells do not typically express detectable levels of the default telomere maintenance enzyme, telomerase; this probably explains why only around 50% of tumours arising from them use telomerase as their telomere maintenance mechanism, the remainder instead using ALT, “Alternative Lengthening of Telomeres” (Johnson and Broccoli, 2007). They therefore present two main challenges. Firstly they must be genetically modified in situ rather than ex vivo, a procedure that deprives us of the ability to select post hoc for cells that have been modified in the desired way and no other. And secondly, since they so often use ALT as their telomere maintenance mechanism, we must delete genes both for telomerase and for ALT – and the latter remain obstinately undiscovered.

9 In situ gene therapy has had a rollercoaster ride over the past 20 years, as all readers will know. This has somewhat overshadowed the fact that it works really quite well in the laboratory, i.e. in contexts where safety is not an overriding issue. I have predicted (de Grey, 2006) that the history of gene therapy, and most particularly the year-long suspension of essentially all gene therapy clinical trials in the wake of Jesse Gelsinger’s death, will be the main trigger for society’s flight from the dictum of “first do no harm” that served the medical profession so well since Hippocrates but has now, in my view, had its day – but I digress. In the context of WILT the main requirement is not effective insertional gene therapy (introduction of a new gene in a more-or-less immaterial genomic location) but gene targeting, the alteration of specific pre-existing stretches of sequence, specifically those encoding one or both of the subunits of telomerase and its presumptive ALT counterpart. When WILT was first formulated, this seemed highly challenging: a few techniques were under development, but they all featured either a distinct lack of multi-laboratory reproducibility or an extremely low “efficiency” (proportion of cells appropriately modified) (Suzuki, 2008). Moreover, their incidence of random modification of unintended sequences was high. Fortunately, this dismal scene is brightening apace. Perhaps the most promising technology for in vivo gene targeting at present is zinc finger nucleases: fusions combining proteins that bind particular DNA sequences with high specificity and enzymes that create double-strand breaks at the site in question (Porteus et al, 2003). When such constructs are introduced into cells (using standard protein transduction techniques such as HIV-TAT, or by introducing the DNA encoding the protein using lentiviral vectors (Lombardo et al, 2007)) in concert with plasmids whose sequence is similar but appropriately modified, the cell’s homologous recombination-based double-strand break repair machinery often uses the plasmid as template rather than the native chromosomal homologue, resulting in the desired modification of the target gene. This technology has already given highly promising results (Urnov et al, 2005; Lombardo et al, 2007). And it is not alone: an unrelated approach involving in vitro evolution of bacteriophage integrases also shows great promise (Sclimenti et al, 2001; Held et al, 2005). One point that bears stressing is that, even though any mitotically competent cell in the body can potentially turn cancerous and kill us, most of them do not: hence, the postponement of cancer is a numbers game in which we (on average) gain some life from modifying a proportion of cells, more from modifying more, and so on. Thus, therapies that suffer from diminishing returns in terms of the difficulty of reaching a given proportion of cells do not so suffer in terms of the life extension that they confer. Suppose that an intervention that deletes telomere maintenance genes successfully modifies 90% of our cells. In the idealised situation where all cells were equally accessible, doubling the dose, or frequency, or potency of the intervention would be expected to modify 99% of the cells – a cell would, in essence, need to “get lucky” (be among the unaffected 10%) twice in order to retain telomere-elongation potential. In practice, however, some cells will be harder to modify than others, so maybe the double-strength therapy will only result in 95% of cells being modified. But that will still give twice the amount of postponement of death from cancer that modifying 90% of the cells would give. In this regard, a heartening feature of “gene targeting” technologies such as those mentioned above is that, because a successfully modified gene will not be further altered if targeted again, the efficiency of modification is not a major issue: so long as the fidelity of the method (avoidance of random modifications elsewhere in the genome) can be made extremely good, tissues can be treated arbitrarily often. Our ignorance of the genetic basis of ALT remains a major Achilles’ heel of WILT, but it too may be about to succumb to research efforts. It has long been known that ALT works by recombination, but this leaves many alternatives, which is the main reason why a “candidate gene” approach to characterising ALT has thus far failed. A somewhat inconspicuous result of telomere lengthening in telomerase knockout mice, reported in the discussion of a paper whose main focus was elsewhere (Herrera et al, 2001), has suggested to me that the genes encoding class switching in the spleen’s germinal centres may be the long-sought mediators of ALT. Recently this idea has

10 been powerfully reinforced by a report (Liu et al, 2007) of the same phenomenon in the early embryo, where a gene central to class switching, AID, is also expressed (Morgan et al, 2004). I have argued elsewhere that ALT must, like telomerase, be encoded by genes that are firmly suppressed in most cell types, and this narrows the field of candidates dramatically. 3.2. Installing ex vivo-modified stem cells When we consider cells of “type 2” above, i.e. the stem cells of continuously renewing tissues such as the blood and the gut, we encounter a very different set of challenges. Since ALT is so rare in tumours derived from such tissues, we may be able to get away for longer (though not forever!) with only targeting telomerase. Secondly, the actual genetic modification of the requisite cells is vastly easier, because techniques can be used that achieve the desired modification only in a very small proportion of treated cells, just so long as the modification is designed to allow the suitably modified cells to be distinguished and separated from those that have been inappropriately (if at all) altered – not a great challenge with today’s molecular techniques – and then expanded to the required numbers (which, in the case of the blood, may be very small numbers (Spangrude et al, 1988)). Thirdly, in decades to come some epithelial tissues into which stem cells may not be able to migrate in sufficient numbers from the blood stream may be relatively easily maintained by wholesale transplantation of tissue-engineered organs: this might be a preferred option for the liver or kidneys, for example. But the exploitation of this opportunity requires a few additional obstacles to be overcome. The main one is to eliminate our natural, telomerase-positive stem cells. Stem cells in the bone marrow and other continuously renewing tissues occupy “niches” – locations that maintain their milieu in a state conducive to the cells’ retention of “stemness.” Engineered stem cells injected (or carried by the circulation) into the vicinity may be unable to invade such niches in adequate numbers unless the niches have been for some reason vacated by their original residents (though a resident cell’s resistance to replacement may be less than was once thought (Quesenberry et al, 2001)). An individual that had been rendered wholly telomerase-negative would probably be a fairly straightforward recipient of such treatment, because the uncompensated loss of telomeres of that person’s stem cells would progressively cause them to differentiate and/or apoptose, conveniently vacating their niches for occupation by the cells introduced in the next round of stem cell replenishment. But what about the first time a patient is treated, when their niches are full of cells with perfectly functioning telomerase that may not so readily relinquish their nest? In the case of the blood, a recent study (Czechowicz et al., 2007) reports a remarkably simple approach to this: introduction of an antibody that inactivates a cell-surface protein expressed by haematopoietic stem cells and thereby causes them to be lost from their niches (and, it seems, to die). After a carefully chosen interval (nine days in the case of mice), the replacement stem cells were injected. They engrafted rapidly enough to prevent any haematological problems, and analysis of the mice’s blood some time later revealed that over 90% of the stem cells had been replaced. A more elaborate, but possibly even more attractive, solution to this problem is to exploit a technique that has been in development for some years, being promising for anti-cancer therapies even independent of WILT. The limited efficacy of existing chemotherapeutic agents has two main causes: firstly the “therapeutic index” problem noted earlier, and secondly the ubiquitous “evolution is cleverer than you are” problem whereby the cancer cell may mutate to resist the therapy. Some examples of the latter have been molecularly characterised, and this has allowed researchers to turn cancer’s ingenuity on itself (Hobin and Fairbairn, 2002). Specifically: consider a normal, non-cancer stem cell that has (by whatever means, such as by our own hand) acquired such a chemoresistance mutation. Now suppose that the chemotherapy in question is applied. At standard dose it kills a lot of the cancer, but not all, and it also kills quite a

11 few stem cells, but not our engineered stem cell. Now let’s raise the dose. Virtually all the cancer is killed. Unfortunately, most stem cells are also killed, which – other things being equal – would kill the patient. But wait! – the patient has some chemoresistant stem cells, which happily survive even this higher dose and duly repopulate the patient. Success! The point for WILT, of course, is that this should work just as well if the chemoresistant stem cell has also been engineered to lack genes for telomerase and ALT. If such cells are injected just prior to high-dose chemotherapy, a double whammy will result: any cancer (whether or not already clinically identified) will be eliminated, and stem cell niches will be divested of their resident cells and thereby made available for the circulating engineered cells. The pre-injection part is probably appropriate only for the blood; for gut, lung and skin a post-chemotherapy approach may be favoured. If you doubt that this would be possible in practice, consider that in the case of the skin this is not materially different from standard contemporary burns therapy or cosmetic surgery (Roh and Lyle, 2006) and that for the gut it has already been demonstrated in rodents (Tait et al, 1994) and dogs (Stelzner and Chen, 2006). It should also be taken into account that there is no need to restrict the in situ gene targeting approach to mesenchymal tissues. While the blood will almost certainly be easier to render incapable of telomere elongation using ex vivo manipulation than by in situ methods, the skin may well be treatable without actually removing and replacing the epidermal layer, and so may the gut. 3.3. Putative roles of telomerase over and above telomere maintenance In recent years, a number of groups have reported evidence from mice that telomerase has roles in cell survival and/or proliferation in addition to its function in maintaining telomere length (reviewed in Sung et al., 2005). At this point, however, this work does not give grounds for appreciable concern that such roles will impact the feasibility of WILT, for several reasons: -

most, if not all, of the reported data can be explained in terms of the balance between cell division-independent telomere shortening (especially via oxidative damage) and telomerase-mediated telomere elongation;

-

mouse data may be misleading, in that organisms that express telomerase very sparingly, such as humans, will have experienced selective pressure to evolve alternative methods to achieve whatever functions telomerase may perform in telomerase-profligate organisms such as mice;

-

telomerase can be inactivated by deleting either the gene for its catalytic subunit or that for its template subunit; deletion of the latter would not affect functions performed by the catalytic subunit on its own, which are what has been reported in most cases;

-

altering six amino acids of the telomerase catalytic subunit disrupts its targeting to the telomere, and thereby abolishes its telomere-elongation activity in cellulo, without an effect on its biochemical function (Banik et al., 2002). Accordingly, this type of modification – which is easily complex enough never to be reverted by spontaneous mutation – is available as a “plan B” if entirely deleting the gene proves problematic.

4. The affordability of such an elaborate therapy I am a British citizen, and as such I have grown up with the idea that health care is naturally free at the point of delivery. I am well aware that the US citizens among my audience are not in the same boat. I will therefore begin this section by giving you the bottom line: I believe that WILT – just like the rest of SENS – will be available to all those old enough to need it, both in nations that favour a public health care system and in those that do not. This is despite the fact that, in my view, WILT will probably be every bit as expensive as initial impressions of its complexity might suggest.

12 4.1. Relative versus absolute benefits My good friend Jay Olshansky’s most prominent publications have been those that highlight how alarmingly little we can extend lifespan or healthspan by combating specific diseases, and thus how important it is to postpone their major “risk factor” (i.e., precursor), aging. One statistic he has derived is that the complete elimination of cancer would raise average age at death in the USA by less than two years (Olshansky et al, 1990). This might seem to demonstrate that the amount most of us would be willing to spend on avoiding death from cancer may be quite modest – far short of the likely cost of a therapy that involves so many state-of-the-art components as WILT does. However, a proper analysis of this question must begin not from the life-extension benefit that we would enjoy today as a result of eliminating death from cancer, but from the benefit that would result at the time that WILT could become available. As noted in the first section of this chapter, it is quite possible that all the other causes of death that remain highly prevalent in today’s industrialised world will have been (or, at least, will be in the process of being) brought under very comprehensive control by the time that WILT arrives. With these other killers removed, the difference in life expectancy that would be made by a truly complete elimination of death from cancer would be very great: decades if we consider the likely efficacy of these non-cancer therapies at that time, and centuries if we additionally take account of the rate at which those therapies will see subsequent improvement (Phoenix and de Grey, 2007). In such circumstances, the investment of a substantial proportion of an individual’s – or of society’s – wealth in a therapy that prevents cancer may seem altogether more attractive. 4.2. The necessity of regular repeat therapy A feature of WILT that has intimidated many observers is the fact that the stem cell replenishment must be repeated every decade or so. The fear is that one may be putting oneself at risk of, for whatever reason, being unable to afford the therapy on schedule, with the result that even though one would not die of cancer, one would die of starvation and anaemia as one’s gut and blood (respectively) lost the ability to self-renew. This concern suffers, in my view, from the same flaw as the one I described above: it evaluates the importance of the treatment in the context of today’s risk of death from other causes, rather than in the likely context that those causes have been essentially eliminated. Once society is faced with the choice of (a) having cancer as the only remaining age-related killer, or (b) spending quite a lot of money preventing it, it may exhibit rather less hesitation in choosing option (b) than it would today – even if there is a stringent requirement to make the prevention treatment very reliably available on schedule. 4.3. Cost of developing WILT in the first place Suppose that pharmacological telomerase/ALT inhibition is 95% effective and costs 1000 times less than WILT, whereas WILT is 100% effective but costs 1000 times more. Is the greater marginal benefit of WILT sufficient to drive the economics of developing and using it? Or could it be that the need to develop stem cell therapies for aging in general will pay for the vast majority of the cost of preparing telomerase-deficient stem cells, which could be used instead of regular stem cells at little or no additional cost, and that the need to develop gene targeting technologies for reasons entirely unrelated to WILT will pay for the vast majority of the costs of deleting telomerase and ALT in vivo? Only time will tell, but my view on the basis of current evidence is that the latter is the case: most of the technologies upon which WILT depends are needed for other purposes too, which is why they are already being pursued. Moreover, when we have 99% effective treatments for all other causes of death, a 95% effective treatment for cancer may not seem quite so adequate as it may seem today.

13

5. Conclusion WILT is a massively ambitious approach to defeating cancer. However, cancer is a hard problem – much harder than mainstream expert opinion had us believe back in 1971, and, I suspect, harder than quite a number of cancer researchers believe it to be even today. In particular, it may be a considerably harder problem than any of the other, also very hard, problems that comprise human aging. Thus, it is in grave danger of surviving as a threat to life long after those other killers have been swept aside. The way to minimise the risk of this unpalatable scenario is to think out of the box now, and identify radical new approaches, rather than to continue with the gradualist approach of attempting possibly over-simplistic refinements of therapies that have previously failed. WILT is such an approach. No one would be happier than I if it turns out to be unnecessary – but we will live (all too briefly) to regret it if we reject WILT as unnecessary until it is too late to save ourselves.

References Banik SS, Guo C, Smith AC, Margolis SS, Richardson DA, Tirado CA, Counter CM (2002) Cterminal regions of the human telomerase catalytic subunit essential for in vivo enzyme activity. Mol Cell Biol 22(17):6234-6246. Berd D (2004) M-Vax: an autologous, hapten-modified vaccine for human cancer. Expert Rev Vaccines 3(5):521-527.

Czechowicz A, Kraft D, Weissman IL, Bhattacharya D (2007) Efficient transplantation via antibody-based clearance of hematopoietic stem cell niches. Science 318(5854):12961299. de Grey ADNJ (2000) Mitochondrial gene therapy: an arena for the biomedical use of inteins. Trends Biotechnol 18(9):394-399. de Grey ADNJ (2005) Whole-body interdiction of lengthening of telomeres: a proposal for cancer prevention. Front Biosci 10:2420-2429. de Grey ADNJ (2006) Has Hippocrates had his day? Rejuvenation Res 9(3):371-373. de Grey ADNJ (2007) Protagonistic pleiotropy: why cancer may be the only pathogenic effect of accumulating nuclear mutations and epimutations in aging. Mech Ageing Dev 128(78):456-459. de Grey ADNJ, Ames BN, Andersen JK, Bartke A, Campisi J, Heward CB, McCarter RJM, Stock G (2002) Time to talk SENS: critiquing the immutability of human aging. Annals NY Acad Sci 959:452-462. de Grey ADNJ, Campbell FC, Dokal I, Fairbairn LJ, Graham GJ, Jahoda CAB, Porter ACG (2004) Total deletion of in vivo telomere elongation capacity: an ambitious but possibly ultimate cure for all age-related human cancers. Annals NY Acad Sci 1019:147-170. Dennett DC (1984) Elbow Room: The Varieties of Free Will Worth Wanting. Clarendon Press, Oxford. Golub TR, Slonim DK, Tamayo P, Huard C, Gaasenbeek M, Mesirov JP, Coller H, Loh ML, Downing JR, Caligiuri MA, Bloomfield CD, Lander ES (1999) Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286(5439):531-537. Held PK, Olivares EC, Aguilar CP, Finegold M, Calos MP, Grompe M (2005) In vivo correction of murine hereditary tyrosinemia type I by phiC31 integrase-mediated gene delivery. Mol Ther 11(3):399-408.

14 Herrera E, Martinez-A C, Blasco MA (2000) Impaired germinal center reaction in mice with short telomeres. EMBO J 19(3):472-481. Hobin DA, Fairbairn LJ (2002) Genetic chemoprotection with mutant O6-alkylguanine-DNAalkyltransferases. Curr Gene Ther 2(1):1-8. Johnson JE, Broccoli D (2007) Telomere maintenance in sarcomas. Curr Opin Oncol 19(4):377382. Kalache A, Aboderin I, Hoskins I (2002) Compression of morbidity and active ageing: key priorities for public health policy in the 21st century. Bull World Health Organ 80(3):243244. Kantarjian HM, Talpaz M, Giles F, O'Brien S, Cortes J (2006) New insights into the pathophysiology of chronic myeloid leukemia and imatinib resistance. Ann Intern Med 145(12):913-923. Keith WN, Thomson CM, Howcroft J, Maitland NJ, Shay JW (2007) Seeding drug discovery: integrating telomerase cancer biology and cellular senescence to uncover new therapeutic opportunities in targeting cancer stem cells. Drug Discov Today 12(15-16):611-621. Liu L, Bailey SM, Okuka M, Muñoz P, Li C, Zhou L, Wu C, Czerwiec E, Sandler L, Seyfang A, Blasco MA, Keefe DL (2007) Telomere lengthening early in development. Nat Cell Biol 9(12):1436-1441. Lobo NA, Shimono Y, Qian D, Clarke MF (2007) The biology of cancer stem cells. Annu Rev Cell Dev Biol 23:675-699. Lombardo A, Genovese P, Beausejour CM, Colleoni S, Lee YL, Kim KA, Ando D, Urnov FD, Galli C, Gregory PD, Holmes MC, Naldini L (2007) Gene editing in human stem cells using zinc finger nucleases and integrase-defective lentiviral vector delivery. Nat Biotechnol 25(11):1298-1306. Morgan HD, Dean W, Coker HA, Reik W, Petersen-Mahrt SK (2004) Activation-induced cytidine deaminase deaminates 5-methylcytosine in DNA and is expressed in pluripotent tissues. J Biol Chem 279(50):52353-52360. National Institutes of Health. Appropriations (2006) NIH Publication No. 06-5:436-438. Nixon RM (1971) Annual Message to the Congress on the State of the Union. http://www.presidency.ucsb.edu/ws/index.php?pid=3110 Olshansky SJ, Carnes BA, Cassel C (1990) In search of Methuselah: estimating the upper limits to human longevity. Science 250(4981):634-640. Phoenix CR, de Grey ADNJ (2007) A model of aging as accumulated damage matches observed mortality patterns and predicts the life-extending effects of prospective interventions. AGE 29(4):133-189. Porteus MH, Cathomen T, Weitzman MD, Baltimore D (2003) Efficient gene targeting mediated by adeno-associated virus and DNA double-strand breaks. Mol Cell Biol 23(10):35583565. Quesenberry PJ, Stewart FM, Becker P, D'Hondt L, Frimberger A, Lambert JF, Colvin GA, Miller C, Heyes C, Abedi M, Dooner M, Carlson J, Reilly J, McAuliffe C, Stencel K, Ballen K, Emmons R, Doyle P, Zhong S, Wang H, Habibian H (2001) Stem cell engraftment strategies. Ann N Y Acad Sci 938:54-61. Roh C, Lyle S (2006) Cutaneous stem cells and wound healing. Pediatr Res 59(4 Pt 2):100R103R.

15 Sclimenti CR, Thyagarajan B, Calos MP (2001) Directed evolution of a recombinase for improved genomic integration at a native human sequence. Nucleic Acids Res 29(24):5044-5051. Spangrude GJ, Heimfeld S, Weissman IL (1988) Purification and characterization of mouse hematopoietic stem cells. Science. 241(4861):58-62. Stelzner M, Chen DC (2006) To make a new intestinal mucosa. Rejuvenation Res 9(1):20-25. Sung YH, Choi YS, Cheong C, Lee HW (2005) The pleiotropy of telomerase against cell death. Mol Cells 19(3):303-309. Suzuki T (2008) Targeted gene modification by oligonucleotides and small DNA fragments in eukaryotes. Front Biosci. 2008 Jan 1;13:737-744. Tait IS, Evans GS, Flint N, Campbell FC (1994) Colonic mucosal replacement by syngeneic small intestinal stem cell transplantation. Am J Surg 167(1):67–72. Urnov FD, Miller JC, Lee YL, Beausejour CM, Rock JM, Augustus S, Jamieson AC, Porteus MH, Gregory PD, Holmes MC (2005) Highly efficient endogenous human gene correction using designed zinc-finger nucleases. Nature 435(7042):646-651. Yasuhisa K, Shin-Ya M, Michinori M, Kazumitsu U (2007) Mechanism of multidrug recognition by MDR1/ABCB1. Cancer Sci 98(9):1303-1310.