cron-web.org

[A1CR]

CR appears to be compared for its phenotype with those of
various models of
faster or slower aging. Stress matters.

van de Ven M, Andressoo JO, Holcomb VB, Hasty P, Suh Y, van
Steeg H, Garinis
GA, Hoeijmakers JH, Mitchell JR.
Extended longevity mechanisms in short-lived progeroid mice:
Identification
of a preservative stress response associated with successful
aging.
Mech Ageing Dev. 2006 Nov 23; [Epub ahead of print]
PMID: 17126380

Semantic distinctions between "normal" aging, "pathological"
aging (or
age-related disease) and "premature" aging (otherwise known
as segmental
progeria) potentially confound important insights into the
nature of each of
the complex processes. Here we review a recent, unexpected
discovery: the
presence of longevity-associated characteristics typical of
long-lived
endocrine-mutant and dietary-restricted animals in
short-lived progeroid
mice. These data suggest that a subset of symptoms observed
in premature
aging, and possibly normal aging as well, may be indirect
manifestations of
a beneficial adaptive stress response to endogenous
oxidative damage, rather
than a detrimental result of the damage itself.

1. Definitions of aging: current focus on lifespan
Most definitions of aging attempt to capture the
irreversible, degenerative
nature of the process by focusing on quite apparent symptoms
like wrinkled
skin and gray hair that affect different people in different
ways as they
age. A more general definition of aging avoids such
individual variation and
focuses on the one constant, a time-dependent increase in
the probability of
dying. The problem with this latter definition is that
lifespan is only one
component of aging and says nothing about a potentially more
important
aspect, the quality of life up until the point of death.

In the so-called "premature aging" disorders, or segmental
progerias,
lifespan is shortened and a number of characteristics, or
"segments", of
aging (in addition to a number of disease-specific
pathologies) appear early
or in exacerbated form (Martin and Oshima, 2000). The
relation between
progeria and "normal" aging is controversial, largely
because there are so
many ways to shorten lifespan that have nothing to do with
the normal aging
process (Miller, 2004). Also poorly defined is the
connection between
"normal" and "pathological" aging. Cancer or Alzheimer's
both increase with
age, but some people become very old without either and
still die of "old
age".

On the opposite side of premature or pathological aging, the
connection
between lifespan extension and "normal" aging is most often
taken for
granted. But perhaps it should not be. An increase in
maximum lifespan can
be achieved by slowing the rate of aging or merely by
delaying its onset,
with different implications for the underlying mechanism
(Barger et al.,
2003). Restricting the diet by reducing the total amount of
food eaten
(Weindruch and Walford, 1988) achieves both. Genetic models
of lifespan
extension, however, may simply delay the onset of
aging-related pathologies,
some of which otherwise would limit lifespan (Barger et al.,
2003).

What then is aging and does broadening its definition beyond
lifespan reveal
anything useful about its nature? We gained an unexpected
insight into this
problem by performing detailed phenotypic analyses of
segmental progeroid
mice engineered with various defects in DNA damage repair.

2. Progeroid NER syndrome: longevity-associated traits in
short-lived mice
Nucleotide excision repair (NER) is an evolutionarily
conserved mechanism
for the removal of bulky, helix-distorting lesions from DNA,
such as
UV-induced damage. It functions by a "cut-and-patch"
mechanism in which the
damage is recognized, the DNA helix unwound, the damaged
strand excised, and
the remaining single-stranded region patched (Hoeijmakers,
2001). In humans,
congenital defects in NER-associated components are
uniformly associated
with UV (sun) hypersensitivity. Specific defects in a subset
of these
components can also lead to the segmental progeroid
disorders Cockayne
syndrome and trichothiodystrophy (Bootsma et al., 2002).
These recessive
disorders display postnatal onset of progressive
neurodevelopmental
pathology with overlapping progeroid features including
reduced subcutaneous
fat and small size (together known as cachectic dwarfism),
sensorineural
deafness, retinal degeneration, white matter hypomyelination
and CNS
calcification sometimes accompanied with premature
appearance of
neurofibrillary tangles (Itin et al., 2001, Nance and Berry,
1992 and Rapin
et al., 2000).

Mouse models of these diseases (Table 1 and references
therein) display an
overlapping set of progressive symptoms including cachectic
dwarfism,
reduced bone mineral density resembling osteoporosis,
curvature of the spine
(lordokyphosis) and failure to thrive. Cerebellar ataxia, a
disease-specific
pathology not associated with normal aging in mice, is
sometimes accompanied
by the loss of Purkinje neurons late in disease progression.
Death usually
occurs before weaning at about 3 weeks of age.

Table 1. Overlapping characteristics of excision repair
deficiencies in man
and mouse Human progeroid syndrome
==================================================
Mouse mutant DNA repair defect Lifespan Genotype specific
pathology
==================================================
XFE (Niedernhofer et al., in press) Ercc1?/? (McWhir et
al., 1993 and
Weeda et al., 1997) NER/TCR; ICLR; telomere maintenance (Zhu
et al., 2003) 3
weeks Liver/kidney polyploidy; reduced hematopoietic
reserves (Prasher et
al., 2005)
Xpf?/? (Tian et al., 2004) NER/TCR; ICLR; telomere
maintenance (Zhu et
al., 2003) 3 weeks Liver/kidney polyploidy
XPCS Xpg?/? (Harada et al., 1999 and Sun et al., 2001)
NER/TCR 2-3 weeks
Undeveloped small intestines
Xpa?/?|Csb?/? (Murai et al., 2001 and van der Pluijm et
al., 2007)
NER/TCR 3 weeks n.d.
Xpa?/?|XpdG602D/G602D (Andressoo et al., 2006b) NER/TCR
3 weeks n.d.
XPTTD Xpa?/?|XpdR722W/R722W (de Boer et al., 2002)
NER/TCR 3 weeks;
survivors 4, 12 months Cutaneous abnormalities, brittle hair
XPCS (mild) Xpa?/?|XpdR722W/G602D (Andressoo et al.,
2006b and van de Ven
et al., 2006) NER/TCR 22 weeks n.d.
? SIRT6?/? (Mostoslavsky et al., 2006) BER 3 weeks
Osteopenia,
lymphopenia
==================================================
n.d.: none determined; NER: nucleotide excision repair;
TCR:
transcription coupled repair; ICRL: interstrand crosslink
repair; BER: base
excision repair.

Recently we reported a surprising finding in Xpd/Xpa double
homozygous
mutant mice (van de Ven et al., 2006). In addition to the
progeroid features
listed above, these mice display characteristics usually
associated with
good health and extended lifespan, as in endocrine-deficient or
dietary-restricted animals (Bartke and Brown-Borg, 2004). These
characteristics, measured at 2 weeks of age when the pups
are still nursing,
include reduced weight, hypoglycemia, hypoinsulinemia,
reduced serum
insulin-like growth factor-1 (IGF-1) and reduced body
temperature. In
addition, a number of genes involved in the postnatal growth
axis are
downregulated in the livers of these animals, including
growth hormone
receptor. These features are observed in a variety of
progeroid NER mice,
including two different Xpd/Xpa (van de Ven et al., 2006)
mutants, Csb/Xpa
(van der Pluijm et al., 2007) and Ercc1 (Niedernhofer et
al., in press) and
are likely to be common to all of the progeroid NER mutants.

3. Adaptive response versus constitutive defect
Constitutive defects in endocrine-mediated insulin signaling
and dietary
restriction can both extend longevity in a number of model
organisms. In
mice, both result in reduced size, hypoglycemia,
hypoinsulinemia, reduced
serum IGF-1 and reduced temperature (Bartke and Brown-Borg,
2004 and Koubova
and Guarente, 2003). One clear difference, however, is that
these phenotypes
are permanent in endocrine-deficient animals, but reversible
in dietary
restricted animals. In short-lived progeroid NER mice,
normal pituitaries
(van der Pluijm et al., 2007) and normal growth hormone
(Niedernhofer et
al., in press, van de Ven et al., 2006 and van der Pluijm et
al., 2007) are
inconsistent with defects in hypothalamic or pituitary
function. A more
attractive hypothesis is that the alteration of energy
metabolism via
dampening of the growth hormone/IGF-1 axis in progeroid NER
mice reflects an
adaptive response in which reduction of mitochondrial
ROS-derived oxidative
DNA damage is the intended consequence. However, such a
hypothesis is
difficult to test in mice with an early onset, irreversible
condition as in
progeroid NER syndrome in which animals die within 3 weeks
after birth.

Fortuitously, one particular combination of mutant alleles
(XpdR722W/G602D/Xpa?/?, Table 1) resulted in mice with each
of the
longevity-associated traits of dietary restriction in
addition to all of the
pathologies of progeroid NER syndrome, save one: instead of
a 3 week
lifespan, mutants survived past weaning with 80% penetrance.
We hypothesize
this to be due to complementation between the two different
mutant Xpd
alleles (Andressoo et al., 2006a). This allowed us to study
these animals
past early development and into adulthood. We noted
additional phenotypes
including normal to elevated food intake per gram body mass
(like
hypopituitary Ames dwarf mice (Bartke et al., 2001)) despite
continuing
dwarfism, time-dependent reduction in the mass of both white
and brown
adipose tissue deposits relative to total mass, progressive
lordokyphosis,
frequent loss of balance and a mean lifespan of
approximately 5 months of
age.

Blood glucose and serum IGF-1 levels of adult mutants gave
us an unexpected
clue about the nature of the effect on growth and metabolism
observed at the
earlier age of 2 weeks. By 10 weeks of age, when control
animals are past
postnatal development and have reached sexual maturity,
blood glucose and
serum IGF-1 levels in the mutants are once again normal
despite the
remaining dwarfism (van de Ven et al., 2006). This is
further evidence
against any constitutive alteration of the growth
hormone/IGF-1 axis or
glucose homeostasis and strongly in favor of the
interpretation that the
downregulation of these components at 2 weeks of age
reflects an adaptive
response to stress.

4. SIRT6 KO mice: similar adaptive response to a common DNA
repair defect?
SIRT6 deficiency results in a phenotype strongly overlapping
progeroid NER
syndrome. Like progeroid NER mice, SIRT6 KO mice are born
normally but
display early postnatal onset of growth retardation followed by
lordokyphosis, cachexia and failure to thrive, with a
maximum lifespan of 24
days (Mostoslavsky et al., 2006). Beginning at postnatal day
12, animals
become increasingly hypoglycemic despite evidence of normal
eating (milk in
the stomach) and a normal ability to absorb glucose (R.
Mostoslavsky,
personal communication). Like in the progeroid NER mice,
reduced serum IGF-1
(Mostoslavsky et al., 2006) despite normal growth hormone
levels (D.
Lombard, personal communication), suggestive of growth hormone
insensitivity, is also present. Furthermore, ablation of the
DNA damage
response protein p53 does not affect lifespan in SIRT6
deficient mice (D.
Lombard, R. Mostoslavsky, personal communication) or
progeroid NER Csb/Xpa
mice (H. van Steeg, unpublished observation).

Differences between SIRT6 deficiency and progeroid NER
disorder also exist,
but are minor compared to the overall similarities. These
differences appear
mainly in the end-of-life pathology. SIRT6 KO mice lose
splenocytes,
thymocytes and peripheral lymphocytes as the result of a
systemic, non-cell
autonomous defect, which may be attributable in part to the
sensitivity of
these cells to hypoglycemia and low IGF-1 (Alves et al.,
2006 and
Mostoslavsky et al., 2006). The loss of Purkinje neurons
typical of
progeroid NER mice may be due to a cell-type specific
hypersensitivity to
oxidative DNA damage combined with the systemic effects of
reduced IGF-1, a
neuronal survival factor. Decreased serum IGF-1 levels are
associated with
cerebellar ataxias of various etiologies in both humans and
experimental
rodent models (Busiguina et al., 2000 and Torres-Aleman et
al., 1996), and
Purkinje neurons are hypersensitive to oxidative stress
accompanying
ischemic injury although relatively resistant to other types
of stress, such
as hypoglycemia (Mohseni, 2001). Interestingly, different
gene-specific
pathologies also exist amongst progeroid NER disorders
(Table 1). For
example, liver- and kidney-specific pathologies exclusively
in XPF- and
ERCC1-deficient mice are probably due to the particular
roles of these
proteins outside of NER, such as interstrand crosslink
repair or telomere
maintenance; brittle hair is specific to the R722W-encoding
mutation in Xpd,
probably due to transcriptional deficiencies particular to
this mutation (de
Boer et al., 1999). Despite these differences, the
overlapping phenotype of
progeroid NER syndrome and SIRT6 deficiency is consistent
with a common
adaptive response to genotoxic stress during development.

Is there any evidence that this hypothetical shared adaptive
response is
triggered by a common genotoxic stress? SIRT6-deficient
cells are
hypersensitive to the effects of ROS generated by ionizing
radiation or
H2O2, and monoadducts by the alkylating agent MMS, but not
to ultraviolet
radiation, consistent with a defect in the base excision
repair (BER) system
(Mostoslavsky et al., 2006) or in other DNA damage response
pathways.
Although BER is functionally distinct from NER, there is
recent genetic and
biochemical evidence of a partial functional overlap between
components
previously thought to be specific to one system (NER
components XPG and CSB)
or the other (BER components OGG-1 and PARP-1) (Dianov et
al., 2000, Licht
et al., 2003, Osterod et al., 2002, Thorslund et al., 2005
and Tuo et al.,
2002). Although much further evidence is required, it is
tempting to
speculate that such a shared DNA repair defect can elicit a
common adaptive
response.

5. Adaptive stress response is not a general characteristic
of genome
instability
In addition to progeroid NER syndrome and SIRT6 deficiency,
a number of
genetically engineered mice have progeroid characteristics
(reviewed in
Lombard et al., 2005). At face value, these progeroid
conditions appear to
have much in common. The engineered mutations are mostly in
genes involved
in nucleic acid metabolism, for example, other types of DNA
repair (Espejel
et al., 2004), telomere maintenance (Lee et al., 1998),
chromosome
segregation (Baker et al., 2004), DNA methylation (Sun et
al., 2004),
mitochondrial DNA replication fidelity (Trifunovic et al.,
2004) and the DNA
damage response (Maier et al., 2004 and Tyner et al., 2002).
Also, the
progeroid phenotypes have an overlapping set of
characteristics including
cachectic dwarfism, reduced fertility, hair loss and
graying, curvature of
the spine, cancer predisposition and shortened lifespan.

We asked whether other forms of genome instability in
addition to excision
repair defects could trigger a preservative organismal
response through
effects on postnatal growth and energy metabolism. We chose
KU80-deficient
mice, with a defect in repairing DNA double-strand breaks
via the
non-homologous endjoining pathway. These animals are
cachectic dwarfs
throughout their lives and display many characteristics of
premature
senescence on both the cellular and organismal levels (Vogel
et al., 1999).
In 2-week-old animals, however, we found no difference in
blood glucose,
serum IGF-1, or gene expression from the postnatal growth
axis in the liver
as there is in progeroid NER syndrome (van de Ven et al., 2006).

Together these data suggest that whatever the apparent
similarities amongst
genomically unstable progeroid mice, on both the molecular
and organismal
levels different mechanisms, or possibly similar mechanisms
with very
different kinetics, are at work. In support of this
conclusion, a different
form of genome instability related to short telomeres
produces a related
subset of progeroid symptoms, including hair loss and
graying, osteoporosis
and fingernail atrophy, in a variety of otherwise unrelated
progerias (Hofer
et al., 2005).

6. Evolution of the preservative stress response
An adaptive stress response involving downregulation of
growth and
alteration of energy metabolism in favor of conservation may
have evolved to
cope with periods of reduced food availability or
life-threatening disease.
It may be thus best defined as a preservative stress
response rather than a
longevity stress response, because its primary purpose is to
help animals
through a period of stress that could occur during any stage
of life (and
thus could be selected for) rather than to extend lifespan
past the
reproductive years. A developmental stage-specific version
of this response
is also conserved in the worm C. elegans. During early
larval development,
food inadequacy triggers an adaptive response known as dauer
formation in
which metabolism is altered via the insulin-signaling
pathway in an attempt
to survive until environmental conditions are once again
favorable (Kenyon
et al., 1993).

The phenotypes of progeroid NER syndrome and SIRT6
deficiency suggest
another way to trigger this response: a particular type of
oxidative
genotoxic stress during early development. The rapid
postnatal onset
suggests that birth stress, which involves an increase in
ROS levels
(Randerath et al., 1997a and Randerath et al., 1997b) may
trigger it and
that rapid postnatal growth may further exacerbate it.
Combined with the
inability to repair certain endogenous lesions as in SIRT6
or NER
deficiency, oxidative genotoxic stress may trigger an
adaptive response
intended to reduce generation of ROS through mitochondrial
respiration and
thus prevent further damage.

This stress response may not be limited to aging or
aging-related pathology,
but may also occur in response to acute stress. In support
of this, we have
data implicating downregulation of IGF-1 and growth hormone
receptor on the
mRNA level in response to the acute oxidative stress of
renal ischemia
reperfusion injury (J. Mitchell, unpublished observation).
Furthermore,
chronic exposure of mice to a peroxisome proliferator
(resulting in elevated
oxidative DNA damage (de Waard et al., 2004)) induces a
similar response in
the gene expression profile of livers of wild-type mice (van
der Pluijm et
al., 2007). A reduction of the growth hormone/IGF-1 may thus
be a more
general marker of both chronic and acute oxidative stress.
However, as is
clear from the shortened lifespan of progeroid NER and SIRT6
deficient mice,
reduced IGF-1 on its own cannot ensure a beneficial outcome;
the nature and
duration of the stress must also be taken into consideration
(Fig. 1).

Fig. 1. Correlations between serum IGF-1, genotoxic stress
and lifespan. In
dietary restricted and endocrine-deficient mice, reduced
serum IGF-1 (black
bars; y axis on the left) correlates with increased
longevity. In progeroid
NER and SIRT6 knockout (excision repair deficient) mice,
this correlation
does not hold. To explain this apparent paradox, we add the
presumed cause
of the reduced IGF-1 in excision repair deficient and
dietary restricted
animal: stress (red bars; y axis on the right). Different
shadings indicate
the different nature of the stressors (unrepaired endogenous
DNA damage,
dark red; reduced energy intake, light red; baseline stress
in control and
endocrine-deficient animals, empty). Constitutive unrepaired
genotoxic
stress may overrule the efficacy of reduced IGF-1 signalling
in excision
repair-deficient progeroid mice, while in dietary
restriction or endocrine
deficiency the stress is either of a different nature, or
absent relative to
controls as in constitutive endocrine deficiency.

7. Terminal senescent weight loss: an adaptive stress
response to normal
aging?
If the reaction to certain types of genome instability
resembles a stress
response, does the oxidative DNA damage that accumulates
with age also
trigger such a beneficial response? There is some evidence
that normal aging
evokes a similar response to stress. In the absence of
age-related disease,
such as cancer or diabetes, people at advanced age often
enter a period of
weight loss culminating in death. This syndrome is known as
geriatric
failure to thrive (Sarkisian and Lachs, 1996). The cause of
the weight loss
is currently unknown. Rats that live to advanced age also
display a related
phenomenon known as senescent terminal weight loss. Although
originally
believed to be caused by decreased food intake, new data
indicate this not
to be the case (Black et al., 2003). The growth
hormone/IGF-1 axis is
already greatly reduced by this age, and thermoregulation
and blood glucose
may also be altered in this terminal senescent period (Black
et al., 2003
and van de Ven et al., 2006). In light of the data reviewed
here, perhaps
its not surprising then that rats (Black et al., 2003) (and
probably mice;
Miller et al., 2005) that experience this senescent terminal
weight loss
"syndrome" actually live significantly longer than those
that do not. In
other words, we propose that terminal senescent weight loss
and geriatric
failure to thrive, despite their foreboding names, are
probably components
of a beneficial, adaptive response triggered very late in
life in response
to the accumulated oxidative damage to macromolecules
including DNA.

8. Conclusions
Everything changes with age, and few of them for the better.
Most things,
from our muscles to our short-term memory, deteriorate over
time. Here we
emphasize the potential importance of another component of
the aging
process: the body's own response to deterioration. This
adaptive response
probably evolved to combat other stresses such as starvation
early in life,
but may be activated in cases of premature, pathological and
natural aging.
A more nuanced and useful definition of aging should thus
include at least
four basic components: underlying mechanisms including
molecular oxidation
(Harman, 1988), genetic background (from progeroid to
centenarian) defining
susceptibility to such damage, time-dependent primary
effects of stochastic
oxidative macromolecular damage, and secondary adaptive
organismal attempts
to counterbalance these effects. In the future, a better
understanding of
these components and their interactions will tell us more,
not just about
how we age, but about how we function at all stages of our
lives.