The Divide-and-Conquer Approach to Delaying Age-Related Functional Decline: Where Are We Now?

Ancient History

The reader is enthusiastically directed to the Ph.D. thesis of Stambler ¹ for a thorough account of the history of humanity's efforts to bring aging under medical control. For present purposes, it suffices to summarize the situation as having passed through four phases. In the first phase, before the early 20th century, the health problems of late life were not clearly distinguished, in their nature and thus in their amenability to medical intervention, from those of early life that in such large part yielded to the pioneering efforts of Jenner, Semmelweis, Pasteur, and others. The second phase roughly coincided with the first half of the 20th century, in which efforts were essentially descriptive and phenomenological but laid the foundations of our knowledge.

Phase three was that of “theories of aging,” when bona fide mechanistic ideas introduced by Harman, ² Hayflick, ³ Szilard, ⁴ and others competed for expert acceptance in terms of their promise as foundations of future antiaging medicines. Phase 4, beginning around 1970, was characterized by overarching disillusionment concerning the entire idea of doing anything about aging: most biogerontologists became effectively seismologists, fully aware that the phenomenon they study harms people, but lacking the remotest aspiration to stop it from happening.

This disillusionment was based on the acceptance that aging is woven inextricably into the mechanics of metabolism, such that its prevention without unacceptable side-effects is inconceivable. The discovery by Johnson's group that a single mutation could greatly extend nematode longevity ⁵ was thus so contrary to established dogma that it was virtually ignored until replicated by Kenyon et al ⁶ several years later. The last few years of the 20th century, by contrast, saw the influx into the field of numerous talented young scientists drawn by the realization that aging had unequivocally become a field in which hypotheses could be tested.

The Unattractiveness of Divide-and-Conquer

The question remained, however, as to what sort of hypotheses. Rose had shown many years previously that natural selection could greatly increase the life span of short-lived invertebrates; was the breakthrough in nematodes merely the segmental activation of an evolved response to a natural phenomenon (famine), which might thus be predicted to have minimal impact in long-lived species? ⁷ It was this line of thinking that led the author to realize, in mid-2000, that the thinking which had inspired the very origin of the field—in the terms described earlier, the transition from phase 1 to phase 2—had been taken too far. ⁸

In its essence, this transition consisted of the realization that the health problems of aging exhibited crosstalk, such that there was no realistic prospect of keeping the elderly healthy by addressing one disease at a time. But what if, rather than the existence of a unitary thing called “aging itself” that underpinned all those diseases and disabilities, the truth were somewhere in between—that a number, but a manageable number, of semi-independent lifelong processes were to blame? From a “unitarist” perspective, it was easy to view this option as too similar to the “one disease at a time” paradigm to hold any water.

Rejuvenation and Its Challenges

The “divide and conquer but not too much” concept initially suffered, in hindsight anyway, from the fact that it was coupled to a second paradigm shift: the idea that reversal of biological age, that is, removal of aging damage, might be medically easier than slowing down the body's creation of that damage. Even the (clearly appropriate, to a contemporary ear) use of the word “rejuvenation” to describe the former faced an uphill battle, as this author recalls vividly from periodic discussions with the publisher over the merits of this journal's name.

Step by step, however, the divide-and-conquer damage–repair concept (colloquially termed the “SENS paradigm”) gained adherents within the expert community, perhaps largely by powerfully rebutting unfounded challenges to it. A few years went by while the dust settled. And then, this ⁹ happened.

“The Hallmarks of Aging,” or, the Power of “Cell”

Lopez-Otin et al.'s 2013 survey ⁹ of our state of knowledge, hereafter termed “Hallmarks of Aging (HoA),” concerning the various mechanisms of aging and ways we might address them shared many features of its predecessor. One would be unjustifiably churlish to attribute its greater impact to its superior graphics or the impact factor of the journal in which it appeared; without a shadow of doubt, the overwhelming explanation was that its time had come. From this author's perspective, it drew a line under a vital but exhausting decade of debate; since the day of its publication, not once have I needed again to justify the strategies for engineered negligible senescence (SENS) paradigm to expert colleagues. Rejuvenation was, and is, here to stay.

In the remainder of this commentary, I shall attempt a comparison and contrast between various key publications that have discussed the challenge of biomedical gerontology in these terms, including not only the 2002 and 2013 papers already mentioned but also a selection of more recent ones. In doing so I will aim to highlight what such surveys can contribute to our understanding of how to develop a rejuvenation panel that truly works—that achieves the field's overarching objective of alleviating the health problems of late life and the immense suffering that goes with them.

SENS and Hallmarks of Aging: Differences in the Categories

First let us look at the individual categories enumerated in the two accounts that originated from the rejuvenation school of thought. I should begin by noting that in the first few descriptions of SENS I listed nine categories, reducing them to seven after concluding that endocrine aging and immune aging were best described as arising from aspects of other categories. Here I refer to the seven-category version.

Cell loss (SENS) and stem cell exhaustion (HoA) are nearly identical, but they differ in that tissues, which have no replenishment from stem cells even in early adulthood (such as most regions of the brain), are included in the SENS version but not in the HoA version. Maybe that is a bit hair splitting, but it must not be overlooked.

Death-resistant cells (SENS) and cellular senescence (HoA) are nearly identical, but an important difference is that “senescence” is (conventionally, and in the HoA paper) defined in terms of gene expression markers that do not identify anergic T cells for instance, so the HoA version is not as broad as it needs to be. HoA does not include T cell anergy (or other “nonclassically death-resistant” cells) in any other category either, so this is an important omission.

Mitochondrial mutations (SENS) and mitochondrial dysfunction (HoA) are not identical, but unlike the two categories above the differences are negated by being classified elsewhere. Specifically, the HoA version includes dysfunction not arising from mutations, whereas in SENS non-mtDNA–based mitochondrial dysfunction is included in categories associated with the upstream cause of that dysfunction (e.g., impaired autophagy is the result of intracellular junk).

Intracellular and extracellular junk (SENS), in combination, line up with loss of proteostasis (HoA)—but again the SENS version includes important phenomena that the HoA version does not. (Aside: in SENS they are considered as two separate categories because the ways proposed to fix them are very different; I will return to this in a later section.) Specifically, the HoA version is overly narrow because it only refers to proteins. Proteins are indeed the main problem for extracellular junk, and in the brain also for intracellular (or so it is currently thought), but there are other key places where the junk is not protein—notably bisretinoids in macular degeneration and oxysterols in atherosclerosis. Again, HoA also does not classify these elsewhere—it overlooks them entirely.

Cancer (SENS) lines up with the combination of genomic instability and epigenetic alterations (HoA). Here, the alignment is quite complex. In SENS, we lump genomic instability and epigenetic alterations together, but then we divide them each into alterations that cause cancer and ones that do not, and we claim that the ones that do not cause cancer accumulate so slowly that they do not matter in anything like a currently normal lifetime, because of “protagonistic pleiotropy” of DNA repair and maintenance. There are good data supporting that conclusion in the case of genetic instability, but it remains controversial.

Extracellular crosslinking (SENS) is omitted entirely in HoA. Apropos, there are reasonable concerns with regard to other changes to ECM, such as noncrosslink adducts (which SENS rejects as accumulating too slowly, another case of protagonistic pleiotropy), calcification (which may be a side-effect of the body trying to minimize loss of bone mineral density, in which case it is covered by cell loss), and fraying of elastin in the largest arteries (which may just need a surgical solution, although some groups have recently identified promising alternatives).

Altered intercellular communication (HoA) does not feature in SENS, precisely because it is intercellular; that is, the source of the problem is not the communication itself but the cells that are initiating the communication. So, in SENS the therapeutic strategy is to fix the cells that are causing the problem that is being transmitted to other cells, rather than to fix the problem en route. (However, identifying the problem en route is a very good discovery strategy, and fixing it en route may be a viable crack-papering approach.)

The HoA paper highlights inflammation as a type of altered intercellular communication, but actually it is an absolutely healthy form of intercellular communication (the immune system responding to calls for help from sick cells—thus, in SENS we address the problem of an excessive inflammatory response by removing the source; that is, rendering the sick cells healthy again.)

Telomere attrition (HoA) does not appear in SENS, because fixing telomere attrition directly would undermine the SENS approach to fixing cancer. Therefore, SENS needs something else instead (since telomere attrition definitely accumulates with age in at least some cell types). Fortunately, it has already something else. There are only two types of consequences of telomere attrition—stem cell loss through apoptosis or differentiation, and cell death resistance—and both are already covered in SENS, so full implementation of SENS leaves no causal path from telomere attrition to pathology.

Deregulated nutrient sensing (HoA) does not feature in SENS, for two reasons depending on what is meant by it. In the HoA paper, the focus is on CR, and its genetic (daf-2, etc.) or pharmacological (rapamycin, etc.) emulation, which is viewed widely as fundamental to aging and in fact makes minimal (if any) difference to aging in long-lived species.

Really the difference is that HoA describes the short-lived setting as “deregulated”—actually it is optimized for maximum evolutionary fitness in the presence of plentiful food, and shorter lived species just need to make bigger changes to that set point when they are starved than long-lived ones do.

But there is a distinct phenomenon that can also be viewed as deregulated nutrient sensing, namely insulin resistance arising from an EXCESS of caloric intake. That certainly DOES affect humans, the structural basis being excess fat and glycogen storage in adipose tissue and muscle mainly, and one way to fix it could be SENS-esquely to view it as intracellular junk and/or death-resistant fat cells, but in practice it is so easy to fix just by fasting that arguably it should not really be viewed as part of aging at all.

Successors to HoA

A number of publications in the past decade have sought to refine the HoA classification. Here, rather than try and fail to be comprehensive, we briefly examine three examples.

Kennedy et al ¹⁰ defined seven “pillars of aging” using terminology that more accurately describes pillars of being alive, but more substantively it is not so much a survey of needed biomedical research as biological research—with a focus on understanding aging rather than on doing anything about it, even though most of the literature cited reported successful interventions.

This was perhaps to be expected in a paper that was essentially a report on an NIH-funded conference, but it was nonetheless disappointing given that most of the authors are well known to be ardently focused on intervention. Its central call to action, echoing the founding principles of the Geroscience Interest Group, was to have more research that straddles the artificial research boundaries between aging and so-called age-related diseases, but just like the GSIG it had negligible impact in that regard.

Another meeting report ¹¹ constituted arguably the first concerted attempt to plug the more serious gaps in HoA, listing 14 rather than 9 categories. It only partially succeeded; it included extracellular matrix stiffening but not accumulation of nonproteinaceous waste, and the addition of impaired autophagy is questionable, while alterations in splicing are in this author's view best viewed as a narrow, although with broad consequences, aspect of other hallmarks. It also included microbiome alterations, on whose status as a primary aspect of aging the jury is out.

On its 10th anniversary, the HoA authors reconvened to update their opus ¹² and presented 12 hallmarks, adding 3 that had also been added by Schmauck-Medina et al. but unaccountably still omitting extracellular matrix stiffening.

Was this what was needed? In this author's view, no. These refinements were laudable as far as they went, but they continued to miss the main point, to which I now turn.

The Need to Classify Damage Repair As Well As Damage

The HoA paper, and to a somewhat greater extent its recent successors, classifies the damage of aging in a manner broadly agreeing with what was proposed in 2002. However, in all those publications, the ways in which this or that category of damage might be addressed by foreseeable medicine have been, frankly, relegated to a series of afterthoughts—or, if we want to be more charitable, have not been similarly classified.

This is a severe shortcoming. The core purpose of this sort of work is not to provide grant administrators or journal editors with guidelines for partitioning applications or manuscripts among their staff; it is to identify gaps or misprioritizations in biomedical research, which harbor the risk of a future in which most but not all major examples of damage can be eliminated, and thus the age of onset of age-related debilitation is little changed even though the relative prevalence of its various manifestations may be redistributed.

In contrast to HoA and its successors, SENS was founded on the concept of a classification that lines up each category of damage with a generic strategy for its elimination. Initially each category was matched with just one such strategy, but over the years more and more have emerged, and in most cases progress has been sufficient to attract the interest of private-sector investors.

Combination Rejuvenation Studies: A Grossly Neglected Research Area

By definition, a divide-and-conquer strategy consists of two phases: first develop the components, then combine them. The past decade's lack of emphasis on the need to align damage categories with intervention modalities has had the practical, real-world consequence that the need for the latter phase is only now becoming recognized, and even now only in a severely limited way. This author's foundation remains, as of this writing, the only organization giving such work sufficient priority to embark on large mouse life span studies that involve gene and cell therapies, rather than being restricted to orally available interventions. ¹³

Conclusion

The ubiquitous, although initially slow, acceptance of the idea that aging is a “rope of events”—a loosely coupled set of processes with distinct, though overlapping, causes—has allowed biomedical gerontology to mature into a discipline that possesses genuine promise for alleviating an astronomical amount of human suffering by greatly postponing the health problems of late life.

The parallel realization that rejuvenation, that is, repair of aging damage, is more practical and promising than mere slowing of the rate at which the body inflicts that damage upon itself has fueled many research areas to the point where they are now richly supported by private investors. However, these massive welcome developments will fail to fulfill their ultimate potential, or at least will take unnecessarily long to do so, unless the community maintains a clear understanding of the integrated and synergistic nature of the overall research program that will be required.