Masked aging

Infant mortality from inherited mutations masks early onset of aging

Polina Loseva, "Elements"
For links, see the original article.

On the human mortality curve – the dependence of mortality on age – there is a "failure" in the region of 9 years. Since the increase in mortality is considered one of the criteria for aging, it turns out that people age "negatively" before the age of 9, because mortality is falling. Recently, an article was published in the journal Cell Reports (Kinzina et al., Patterns of Aging Biomarkers, Mortality, and Damaging Mutations Illuminate the Beginning of Aging and Causes of Early-Life Mortality), the authors of which found another explanation for this phenomenon. They believe that early mortality is a manifestation of natural selection for unfavorable mutations, which has nothing to do with aging. Thus, human mortality, in their opinion, should be considered as the sum of these two factors, and the increase in mortality is not the best criterion for aging.

Fig. 1. The human mortality curve (on the left), according to the authors of the new work, is decomposed into the sum of two curves: one describes natural selection in infancy (in the center), the second – mortality "from old age" (on the right). An image from the discussed article in Cell Reports.

Aging is not an easy topic. It can definitely be said that over the years it has affected everyone in one way or another, and that, in particular, because of this, the efforts of many scientists have been devoted to its study. At the same time, the term "aging" still has no clear definition. Despite the fact that, it would seem, everyone imagines what it is, attempts to give an unambiguous definition come to a dead end. For example, if we say that aging is a gradual deterioration of the body, then what about cheerful elderly people who live an active life and do not complain about their health? And if, for example, those who have lost the ability to walk quickly are considered old, then young disabled people will have to be accepted into the ranks of the elderly.

Therefore, not only can we not reliably distinguish an old organism from a young one, but we also do not fully know how to determine that it is aging and when this process begins. One of the most widely used criteria for aging today remains a pattern that was deduced by the English mathematician Benjamin Gompertz in the XIX century: the risk of death from natural causes increases with age. Thus, if some organisms begin to die more often over time (their mortality increases), we can say that they are aging.

The Gompertz criterion is convenient because it is easy to apply it on any model objects. Most of the works devoted to life extension rely on this criterion, and their authors build survival graphs for their experimental animals that are the reverse of mortality graphs – that is, they measure what percentage of the original population remains alive at any given time (Fig. 2). The shift of the curve to the right means that mortality is "postponed" to a later age, therefore, aging is also inhibited along with this.

Fig. 2. The effect of rapamycin on the survival of male mice. Blue graph – control animals, red – taking rapamycin. On the horizontal axis, the age of mice is postponed (in days), on the vertical – the proportion of living individuals. Figure from the article by D. E. Harrison et al., 2009. Rapamycin fed late in life extends lifespan in genetically heterogeneous mice.

Of course, it is impossible to apply such a criterion to each person individually. But in general, the Gompertz curve in humans does not fully correspond to the theoretical regularity: the risk of dying does not increase throughout life. So, there are suspicions that the super-long-livers (the so-called people who are over 105 years old) have a curve that reaches a plateau (E. Barbi et al., 2018. The plateau of human mortality: Demography of longevity pioneers), that is, the risk of dying becomes permanent (Fig. 3). However, this statement is the subject of controversy, since the samples of super–long-livers are extremely small, and the result may vary depending on the methodology and accuracy of calculations.

Fig. 3. Predicted mortality plateau: the risk of death (on a logarithmic scale) depending on age. The black line is a uniform increase in the risk of dying, the blue dotted line is data on real centenarians, the orange line shows the forecast based on data on centenarians. Figure from the article by E. Barbi et al., 2018. The plateau of human mortality: Demography of longevity pioneers.

But on the other side of this curve, everything is not quite smooth (in the literal sense of the word). Gompertz was not particularly interested in children at the time, so he built his schedule starting at the age of 10. Nevertheless, according to modern data, at the beginning of life, the risk of dying also varies unevenly: in newborns, mortality is relatively high, then it falls, forming a "dip" around 9 years (Fig. 4), and then begins to grow again and does not decrease until the very end of life. This decline, in fact, is rather shallow: the risk of dying even in newborns is relatively small, so in order to consider it, you have to plot on a logarithmic scale – in normal it actually disappears.

Fig. 4. The dynamics of mortality of people under the age of 25 with a childhood "failure" in the region of 9 years. Graph from the article by E. M. G. Milne, 2006 When does human aging begin?

Nevertheless, the "failure of childhood", unlike the "plateau of aging", appears at any calculation in any population, and not only in humans, but also in other animals (see D. Levitis, 2012. Before senescence: the evolutionary demography of ontogenesis). If we follow the definition of aging as a growing risk of dying, it turns out that the early stages of life correspond to "negative senescence", and "real" aging begins – in humans – after 9 years.

There is nothing surprising in the fact of negative aging itself: there are known animals that linger in this phase for the whole or almost the whole of their lives (such as the red-legged frog, see O. R. Jones et al., 2013. Diversity of aging across the tree of life). However, until now it was not clear what caused this "failure of childhood". For example, it can be assumed that in the early stages of development, the body accumulates less damage (such as mutations in DNA or improperly folded proteins), because cells divide frequently, and breakdowns are distributed among their descendants. Consequently, the further away, the lower the specific concentration of breakdowns in cells, and the lower the mortality rate.

At the same time, other explanations are also possible: for example, that 9 years is the safest age for a child, when he is already independent enough not to injure himself, but is still under the supervision of his parents. If so, then the mortality curve for humans does not fully correspond to aging, and an increase in mortality is not the best definition for aging.

To figure out what exactly reflects the "failure" on the graph, a group of scientists led by Vadim Gladyshev from Harvard Medical School checked to what extent the Gompertz curve for a person coincides with curves based on other definitions of aging.

For example, you can call aging a growing risk of dying from age-related diseases. This definition is not much different from the previous one (reducing aging to an increase in mortality) – after all, behind every death "from old age" actually lies death from a specific disease: in most cases, even after a seemingly sudden death, traces of a specific pathology can be found during autopsy. Nevertheless, a separate curve can be constructed for mortality from each age-related disease (Fig. 5). It turned out that the curves of mortality from a variety of diseases, the incidence of cancer and even doctor attendance in the form reproduce the overall mortality curve. Therefore, if we understand aging as the accumulation of age-related diseases, then the Gompertz curve fully reflects it.

5. Mortality curves: A – from all causes, B – from heart disease, sepsis, influenza and pneumonia and all other diseases, D – cancer incidence curve, E – curves of visits to the doctor. An image from the discussed article in Cell Reports.

Then the scientists turned to another criterion of aging: the accumulation of pathogenic mutations in cells. There are different ways to calculate the number of mutations; in this case, the authors of the work used a database of tumor genomes and built a dependence of the number of mutations in the tumor on the age at which it was detected. And here the graph turned out to be quite different from the Gompertz curve: the number of mutations increased evenly and continuously, starting from the first years of life and up to the end. It turns out that these two definitions of aging contradict each other: according to the mutation theory, aging should begin from the first years (if not from the first days of the embryo's life – there is no reason to believe that tumors do not form in the embryonic period, and according to Gompertz – from the age of 9.

Fig. 6. The occurrence of mutations in the tumor genome depending on age. From left to right: malignant lymphoma, soft tissue tumors, childhood brain cancer. An image from the discussed article in Cell Reports.

Then the researchers turned to a third way to measure aging – the epigenetic clock. Over time, DNA repackaging takes place in human cells: part of the strands is twisted, thereby blocking access to the information that is recorded on them, and the other part, on the contrary, unfolds. The cell regulates this process by attaching epigenetic markers to the DNA strands themselves and histone proteins (on which the strands are "wound") – these are chemical labels that make the thread more "sticky" or, conversely, more "slippery". Accordingly, it is possible to determine the age of cells and, with some accuracy, of the organism as a whole by a set of these labels. The most widely used model of epigenetic clocks – 353 sites where the presence/absence of a label should be assessed – was proposed in 2013 by American Steve Horvath (S. Horvath, 2013. DNA methylation age of human tissues and cell types).

Using his method, the authors of the article under discussion assessed how the epigenetic age of a person grows, and found that this dependence also differs from the Gompertz curve: at the beginning of life they noticed a slow growth, and then the degree of methylation began to rise sharply (Fig. 7). Thus, it turned out that if you believe the Gompertz curves and mortality from various diseases, then aging begins at 9 years old, and if you rely on mutations and epigenetics, then – from the very first days of life.

Fig. 7. The relative degree of methylation of 353 DNA sections from the "Horvath clock" depending on age. An image from the discussed article in Cell Reports.

To resolve this contradiction, Gladyshev and colleagues suggested that the early decline in mortality is caused not by negative aging, but by selection for unfavorable mutations, that is, by the fact that carriers of "non-working" variants of vital genes die at an early age. They were led to this idea by a recent study on mice (M. E. Dickinson et al., 2016. High-throughput discovery of novel developmental phenotypes). Its authors worked with mice knocked out by different key developmental genes (deprived of both copies of the gene). They calculated mortality at different stages of embryogenesis and noticed that it continues the "failure of childhood", that is, at the early stages of development is higher than at the later stages. Gladyshev and colleagues constructed the same curve for humans: in humans, too, mortality was maximal in the second trimester of pregnancy and decreased several dozen times during the third trimester.

Then the authors of the work under discussion found among the human genes orthologs of those mouse genes, knockouts for which lead to mortality at different stages of embryogenesis. Scientists have suggested that homozygotes for these genes in humans should be extremely rare, and selection will go mainly on heterozygous mutations. Having found selection coefficients in databases (a parameter that reflects the strength of selection against the least viable individuals) for these genes, the researchers found that they also fall depending on the stage at which carriers of unfavorable mutations are "eliminated". This is probably due to when genes start working during development: it would be logical to assume that the force of selection that acts on a particular gene, the less the later it "turns on" during life.

Thus, scientists propose to consider mortality as the sum of two factors: early selection for unfavorable mutations and age-related mortality. It turns out that mortality is in any case associated with mutations, but they have a different nature: at the beginning of life, mainly those that the child inherited from his parents are rejected, and during life – those that, along with other injuries, arose independently in the cells of an adult organism. These two mortality factors act in the opposite phase, and at 9 years there is a minimum for their total effect: the selection is almost over, and age mortality has not yet gained strength.

Figure 8. The mortality curve as the sum of two factors: early selection and aging. An image from the discussed article in Cell Reports.

There are several important consequences of this model. The first is that the increase in mortality is not the best and not a universal definition of aging. And if in the middle of life the pattern noticed by Gompertz remains fair and not disputed by anyone, then at least at the beginning of life it is incorrect: early mortality, apparently, has nothing to do with aging. Therefore, no matter how convenient it is to measure aging by the risk of dying, other criteria – for example, the accumulation of mutations and other damage or a change in DNA methylation – may be more correct.

One could argue here that we rarely encounter the need to measure the aging of young children, and even more so embryos (although there are exceptions here: for example, rapidly aging children with Hutchinson–Guilford progeria). Nevertheless, it would be good if the definition of aging was universal and applicable to any organisms at any stage of life. This, among other things, will allow us to answer the question that still remains unresolved: when does the body begin to age?

The second important consequence of the model of Gladyshev and his colleagues is that early mortality, paradoxically, masks the early onset of aging. If we believe the molecular criteria of aging – the number of mutations or epigenetic marks – then it begins at the first stages of embryogenesis. However, at this stage, the mortality that these changes provoke is so small that it becomes invisible against the background of mortality caused by selection for mutations inherited from parents.

Finally, the third consequence is that there is no negative aging at the physiological level – at least in humans. The decline in mortality that we see on the Gompertz curve means that the least viable embryos or newborns have died, and does not mean at all that the survivors have become younger or stronger. Whether this is the case in other animals and whether it is still possible to find any signs of negative aging in humans remains to be seen.

Portal "Eternal youth" http://vechnayamolodost.ru