cron-web.org

[A1CR]

How long we live may relate to stress.
http://en.wikipedia.org/wiki/Stress_%28medicine%29 may be
positive or
negative
http://en.wikipedia.org/wiki/Eustress#Distress_versus_eustress
in
its effects, it seems.

Vermeulen CJ, Loeschcke V.
Longevity and the stress response in Drosophila.
Exp Gerontol. 2006 Nov 14; [Epub ahead of print]
PMID: 17110070

The concept that lifespan is a function of the capacity to
withstand
extrinsic stress is very old. In concordance with this,
long-lived
individuals often have increased resistance against a
variety of stresses
throughout life. Genes underlying the stress response may
therefore have the
ability to affect lifespan. The progress in modern genetic
techniques has
allowed researchers to test this idea. The general stress
response involves
the expression of stress proteins, such as chaperones and
antioxidative
proteins, downregulation of genes involved in energy
metabolism and the
release of protective substances. Do these same changes in
patterns of
expression have the ability to mitigate ageing and prolong
lifespan? It
appears that parts of this response indeed are also
associated with extended
longevity, whereas some elements are not, due to their high
cost or
long-term deleterious consequences. Here we briefly review
the state of the
art of research on ageing and longevity in the model
organism Drosophila,
with focus on the role of the general stress response. We
will conclude by
contemplating some of the implications of the findings in
this research and
will suggest several directions for future research.

1. Introduction
In natural populations, most organisms regularly experience
acute threats to
survival, such as temperature extremes, food deprivation or
infection. These
and many other environmental factors may thus be considered
to constitute a
stress. A definition of stress that carries over well across
different
disciplines and levels of biological organisation is given
by Hoffmann and
Parsons (1991): "Stress is an environmental factor causing a
change in the
biological system, which is potentially injurious". Lifespan
in nature is
determined largely by the frequency with which these
stressful conditions
are encountered, the severity of the stress and the
organisms ability to
cope with it. In addition, intrinsic factors also may cause
injurious
changes to the biological system and thereby affect the
persistence of the
organism. Intrinsic stress can take on several forms, among
which
non-optimal genetic constitution (e.g., inbreeding
depression) or specific
metabolic challenges, such as accumulation of waste products
or the release
of reactive metabolites during normal metabolic activity.

To cope with stress, organisms have developed a range of
general and
specialised stress response mechanisms. Stress response can,
among many
other things, involve the expression of stress proteins or
accumulation of
protective substances. Under adverse circumstances, lifespan
will be
determined in part by the efficiency of the stress
responses. A specific
case occurs when there is a sudden alleviation of extrinsic
stress. Then we
typically observe that with time there is a gradual increase
in age-specific
mortality, a process more trivially known as senescent
ageing. Among many
other definitions, ageing is described as "the total effect
of those
intrinsic changes in an organism that adversely affect its
vitality and that
render it more susceptible to the many factors that can
cause death" (Zwaan,
1999). Ageing is thought to have evolved as a result of the
sheltering of
later ages against the effect of natural selection,
resulting in the
accumulation of mutations that depress late age survival. In
a protected
environment longevity is inversely correlated to the rate of
demographic
ageing (defined as the rate of increase in age-specific
mortality), which is
affected by many environmental and genetic variables.
According to several
ageing theories lifespan is causally related to the ability
to withstand
extrinsic or intrinsic stresses. Hence, longevity should
positively
correlate with the ability to resist stress (Kirkwood and
Austad, 2000).
Thus, genes involved in the stress response should be
relevant candidates
for lifespan determination at a mechanistic level. However,
as risk factors
do not correlate between natural and protected environments,
it is expected
that the nature of the stress changes from extrinsic to
intrinsic threats.
This may have consequences for correlations between
longevity and resistance
to extrinsic stressors. Still, many mechanistic similarities
have been
described in gene expression, protein expression and
physiology associated
with phenotypic and genetic differences in longevity and a
variety of
measures of stress resistance (discussed below). This begs
the question to
what extent the genetic architecture of stress resistance
and lifespan
overlap and for what reasons.

Many insights have been derived from studies on model
organisms including
flies (Drosophila melanogaster), worms (Caenorhabditis
elegans), mice (Mus
musculus) and yeast (Saccharomyces cerevisiae). Thus,
although our ultimate
interest might lie in describing the stressors and stress
response
mechanisms affecting lifespan in humans, we first need to
make use of more
basic animal models. Here we focus on the study of stress
responses and
their relevance to longevity in Drosophila. The stress
response is hereby
broadly taken to represent the entire change in physiology,
rather than
limited to the expression of stress proteins. Although some
proximate
mechanisms of ageing could be private to insects or
fruitflies, Drosophila
research has provided some excellent insights for ageing
studies in general.
We will first describe stress responses relevant for
longevity and then
determine to what extent the genetic basis of stress
resistance and
longevity overlap.

2. Interventions affecting longevity and the stress response
Given the importance of Drosophila in ageing research, it is
remarkable that
the major causes of fly mortality are unknown. Flies
accumulate some
mechanical damage during their lives, as they lack mitotic
tissue at
eclosion, but it is unlikely this is the main source of
mortality. Several
things can be learned from responses to interventions that
either decrease
or increase survival rate in Drosophila. Many interventions
can decrease
survival and thereby constitute a severe stress: starvation,
desiccation,
cold, heat and oxidative stress to name a few. Interventions
increasing
longevity are diet restriction and exposures to mildly
stressful conditions
(hormesis). Dietary restriction arguably may fit into the
larger category
"hormesis treatments". Remarkably, longevity-increasing
interventions do not
seem to diminish the threat to survival, whatever it may be,
but rather
activate defence mechanisms against it. Interventions that
activate the
stress response, without lowering survival themselves,
appear to be able to
extend lifespan (Hercus et al., 2003 and Rattan, 2005). It
is thought that
these treatments modulate the ageing process by activating
stress response
mechanisms at the expense of growth and reproduction.

3. Description of the stress response and its relation to
ageing processes
Many stressors evoke specific responses, but we are most
interested in
general stress responses. At present, many new studies
appear that report on
transcriptional responses. As several processes are
regulated with both
stress and age, these may indicate functional similarities.
One of the
strengths of studies on gene expression covering the entire
genome is that
they are unbiased with respect to candidate genes, and
therefore have the
potential of uncovering novel mechanisms. Transcriptional
changes in
response to ageing and stress are manifold, including
changes in transcript
abundance of genes involved in immune response,
detoxification, lipid
metabolism and protein turnover (reviewed in Pletcher et
al., 2005). Still,
there are a few common themes. First, not surprisingly, many
typical stress
response genes are transcriptionally upregulated in response
to stress.
These include genes encoding chaperone proteins (e.g., hsp),
which are
upregulated in response to heat stress and oxidative stress
(Landis et al.,
2004, Sørensen et al., 2005 and Wang et al., 2004a) and also
with age
(Landis et al., 2004 and Zou et al., 2000). In addition,
oxidative stress
response genes are included in this overall class of
stress-related genes
(e.g., Cu/Zn Sod, glutathione-S-transferase), which are
similarly
upregulated both with age (Landis et al., 2004 and Pletcher
et al., 2002)
and in response to oxidative stress and heat stress
(Girardot et al., 2004,
Landis et al., 2004, Sørensen et al., 2005, Wang et al.,
2004a and Zou et
al., 2000). Another typical feature found in response to
stress is the
genome-wide observed downregulation of mitochondrial
function (e.g.,
oxidative phosphorylation), which occurs in response to
oxidative stress
(Zou et al., 2000) and which can also be seen with age
(Landis et al., 2004,
Pletcher et al., 2002 and Zou et al., 2000). Related to
decreased
mitochondrial function is the downregulation of genes
involved in energy
metabolism (e.g., glycolysis and the citric acid cycle),
which occurs both
with age (Landis et al., 2004 and Zou et al., 2000) and in
response to
oxidative stress (Zou et al., 2000). The decrease in
mitochondrial function
and energy metabolism is especially interesting, as it may
possibly signify
decreases in the rate of ROS-production and metabolic rate.
One would expect
that changes in the abundance of mRNA would also translate
to corresponding
changes in the proteome.

If we move to changes in translation in response to stress,
we need to first
consider an overwhelming increase in the abundance of
heat-shock proteins
(reviewed in Morrow and Tanguay, 2003). Heat-shock proteins
(or Hsp) are
molecular chaperones that protect the integrity of nascent
proteins, but in
addition are essential to a wide range of functions in the
cell, such as
protein transport, folding, assembly, signalling, secretion
and degradation
of misfolded and aggregated proteins (Morrow and Tanguay,
2003 and Sørensen
et al., 2003). Levels of the molecular chaperones Hsp22 and
Hsp70 also
increase with age, possibly indicating accumulation of
misfolded proteins or
decreased accuracy in protein synthesis (King and Tower,
1999 and Wheeler et
al., 1995).

Moving higher up in the biological organisation, we observe
changes in the
level of certain metabolites in response to stress. These
may signify
synthesis of protective substances and the release of
energy-rich compounds
from somatic stores. Malmendal et al. (2006) report an
increase in the level
of amino acids in response to heat stress. These may reflect
changes in the
rate of protein turnover, as detected in gene expression
studies, but also
suggest several specific functions. First, an increase in
the level of
alanine was observed, which may signify a shift to anaerobic
respiration
during stress. Second, an increase in tyrosine was detected,
which may be
associated with changes in the synthesis of catecholamines,
especially
dopamine. Dopamine levels go up in response to stress
(Rauschenbach et al.,
1993), and have also been shown to be associated with
longevity (Vermeulen
et al., 2006a). Also, decreases in the level of glycogen,
fatty acids and
glucose were found, which are concordant with changes in
energy metabolism.
This final point is relevant to the issue of modulation of
metabolic rate, a
putative genetic response to stress (Hoffmann and Parsons,
1991).
Furthermore, it is thought that decreased metabolic rate may
lead to
decreases in the production of ROS and hence to increased
longevity.
However, there has grown considerable controversy about a
possible causal
link between metabolic rate and longevity (discussed below).

Although several features of the stress response do not
appear to be related
with ageing, there still is considerable congruence between
the stress
response and age-related changes, which may signify a
function for stress
response mechanisms in somatic maintenance (and hence, ageing).
Alternatively, this may merely show an increase in intrinsic
stress levels
with age and not a causal relationship. We will show that
there are sound
genetic data supporting a role for stress resistance genes
in lifespan
variation.

4. Overlap between the genetic basis of stress resistance
and longevity
Associations between traits have often been revealed by genetic
manipulations. Given that lifespan is causally related to
the ability to
withstand extrinsic or intrinsic stress, longevity should
increase whenever
stress resistance is bolstered and vice versa. Many
Drosophilastudies using
artificial selection regimes report on genetic correlations
among resistance
to several stresses and longevity (Bubliy and Loeschcke,
2005, Force et al.,
1995 and Mockett et al., 2001) and this is also found for
long-lived mutants
(e.g., Lin et al., 1998). This strongly suggests the genetic
basis for
lifespan and stress resistance to overlap. This can be
investigated in more
detail using genetic analysis by mapping of Quantitative
Trait Loci (QTL).
In QTL-mapping, loci conditioning standing genetic variation
for lifespan
are mapped by their genetic linkage to molecular markers. An
interesting
twist is when these studies simultaneously assess variation
for lifespan and
stress resistance. If QTLs for these two traits colocalise
on the genome, it
is likely that the QTL contains a gene that conditions both
traits.
Surprisingly, longevity and stress resistance seem to have
largely different
genetic bases in these studies, but some potential
pleiotropic QTL are
usually found (Vieira et al., 2000 and Wang et al., 2004b).
Since genetic
variation for longevity is often associated with resistance
to several
stresses, this suggests the existence of a "general stress
mechanism" that
is associated with multiple stress resistance as well as
longevity. Naively,
one might expect this consists of a constitutive
upregulation of the
phenotypic stress response. Therefore, there has been focus
on Hsp
expression, oxidative stress response, metabolic rate and
storage of energy
compounds. We will discuss these in turn.

Heat shock proteins are attractive candidates for a general
stress
mechanism. They are expressed in response to a wide variety
of stresses and
their beneficial effect on longevity is plausible, as many
age-specific
diseases are associated with protein aggregation (e.g.,
Alzheimer disease in
humans). In concordance with a protective function, an
increased expression
of hsp22 and hsp 23 transcripts has been verified in lines
artificially
selected for increased lifespan (Kurapati et al., 2000).
However, selection
on increased survival can also lead to a decrease in the
level of
stress-inducible Hsp70 expression, suggesting other
mechanisms become
important, such as more stable proteins or a higher level of
constitutive
protection (Norry and Loeschcke, 2003). Wang et al. (2004a)
reported that a
screening for transcripts that are upregulated by multiple
stresses in adult
Drosophila uncovered thirteen genes, among which hsp 26 and
hsp 27 were
subsequently shown to be capable of increasing adult lifespan.
Overexpression of heat-shock proteins has been shown to have
beneficial
effects on lifespan, age-specific survival and stress
resistance (Morrow et
al., 2004 and Tatar et al., 1997), but detrimental effects
on these
characters have also been observed (e.g., Bhole et al.,
2004). Presumably,
these discrepancies arise because respective Hsp have
specific functions
that are sensitive to timing and location of expression.
Ectopic expression
may easily give rise to deleterious effects, as the
expression of Hsp is
also known to elicit a large cost in terms of growth,
developmental rate and
fertility (reviewed in Sørensen et al., 2003).

For several reasons, oxidative stress response is likely
also to be involved
in ageing processes. Oxidative stress is present in
virtually all tissues
throughout life, as normal oxygen metabolism gives rise to
the release of
reactive oxygen species (ROS) from the mitochondria. This is
battled in most
organisms by detoxifying enzymes, such as superoxide
dismutase (Sod),
catalase and glutathione peroxidase. In some studies,
increased expression
of Cu/Zn Sod was found in lines artificially selected for
increased
longevity (e.g., Arking et al., 2000), but not in others
(Mockett et al.,
2001). Overexpression of Cu/Zn Sod can, in some instances,
increase lifespan
(Parkes et al., 1998 and Sun and Tower, 1999), but also
often fails to have
an effect (Orr and Sohal, 1993 and Seto et al., 1990).
Again, the timing and
location of expression appear to be critical determinants as
to whether the
modification will have a beneficial effect. In addition, it
is suggested
that increasing antioxidative capacity by means of
overexpression of oxygen
scavenger proteins only will increase lifespan in relatively
short-lived
genetic backgrounds (Orr and Sohal, 2003).

Related to the topic of oxidative stress response is the
regulation of
metabolic rate. A decrease in metabolic rate may attenuate
the production of
ROS. In addition, it is a plausible mechanism for the
lifespan increasing
effect of dietary restriction. Concordant with this, it was
found in several
studies that lines selected for either increased longevity
or stress
resistance have decreased metabolic rate (Hoffmann and
Parsons, 1989 and
Service, 1987). However, this finding could not be
replicated in several
other studies (Djawdan et al., 1997, Van Voorhies et al.,
2003 and Vermeulen
et al., 2006a), nor could decreased metabolic rate be
detected in dietary
restricted Drosophila (Hulbert et al., 2004). This indicates
that decreased
metabolic rate is not a prerequisite for increased longevity
or stress
resistance.

A genetically correlated response frequently found in
long-lived individuals
is increased stores of lipids and/or carbohydrates (e.g.,
Djawdan et al.,
1996), matching findings from long-lived mutants of the
insulin-like
signalling pathway in C. elegans and Drosophila (Böhni et
al., 1999 and
Tatar et al., 2001). In addition, lipid fraction is known to
correlate
strongly with resistance to starvation stress in Drosophila
(reviewed in
Hoffmann and Harshman, 1999). This finding is often
interpreted as an
adaptation for increased energy demands during stressful
periods. However,
again, increased lipid or carbohydrate stores are not
invariably linked to
increased longevity, and therefore do not seem to be a
prerequisite for
increased longevity (Force et al., 1995 and Vermeulen et
al., 2006b).

As exemplified by the large number of processes known to be
involved in
lifespan determination, it is clear that ageing has a
multifactorial nature.
As none of these processes alone can satisfactorily explain
all intra- and
inter-specific variation in lifespan, and as the involvement
of many of them
in lifespan determination is very plausible, we need to
consider an
additional level of regulation. Presumably, all respective
responses are
tied together by higher-order signalling networks. The most
likely
candidates coordinating these processes are in the
insulin/IGF-like
signalling (IIS) pathway, in conjunction with endocrine
hormones and
neuropeptides (reviewed in Partridge and Gems, 2002 and
Tatar et al., 2003).
C. elegans worms that carry mutations in genes that function
in the IIS
pathway (e.g., age-1 and daf-2) are known to have increased
stress
resistance and increased lipid stores in addition to greatly
increased
longevity (Gems et al., 1998). These findings have been
partly duplicated in
Drosophila (Böhni et al., 1999 and Tatar et al., 2001),
indicating that
control of lifespan by insulin/IGF-like signalling is
possibly a public
mechanism that is conserved in many organisms. The IIS
pathway has been
shown to be able to control ageing of organs in an
autonomous fashion, as
demonstrated in the Drosophila heart, but importantly also
shows
non-autonomous control of lifespan (Wessells et al., 2004).
Tuning
organismal responses, such as stress responses, to
environmental and
internal cues is accomplished by a complicated network of
signalling
proteins. At the top of the cascade are proteins involved in
sensing
nutrient availability and cellular indicators of stress,
which are able to
activate signalling cascades (e.g., by endocrine release of
insulin-like
messengers). This will result in the activation of response
elements, such
as transcription factors. One of the key players is the
forkhead
transcription factor FOXO, which is thought to mediate
longevity promoting
signals by transcriptionally activating proteins involved in
stress
protection (e.g., aforementioned heat shock proteins and
oxygen scavenger
proteins), as well as proteins involved in damage repair and
cell cycle
arrest. Both the IIS pathway, that responds to nutritional
cues, and the
stress-responsive Jun-N-terminal kinase (JNK) pathway
converge on FOXO, thus
making the latter a central integrator controlling cellular
and organismal
adaptations, including stress resistance and longevity (Wang
et al., 2005).
FOXO activity also can be modulated at a translational level
by interaction
with TOR, that responds to changes in growth factors, amino
acids, oxygen
tension and energy status (Luong et al., 2006). It thus
appears that
regulator loci that have transcriptional or translational
control over genes
can potentially affect all aforementioned traits. It is
tempting to envisage
signalling mechanisms at the top of the hierarchy to be
preserved, whereas
downstream response genes are free to diverge to fit
specific requirements
for each species.

5. Future directions
What are the implications of some of these findings? First,
we note that the
age-specificity of mortality is an understudied aspect. If
we assume that
there is a specific timing to the respective (intrinsic and
extrinsic)
sources of mortality, this will result in complex
age-specific patterns in
the requirement of corresponding sets of resistance genes
(Pletcher et al.,
2005). For example, some alleles may ensure proper function
of vital
processes and therefore are required, but not sufficient,
for high longevity
as is suggested for genes involved in oxidative stress
response (Orr and
Sohal, 2003). As a result, respective "longevity genes" may
develop temporal
epistatic relationships to each other. Only few studies have
addressed
age-specific properties of QTL affecting lifespan and
mortality rates (e.g.,
Curtsinger and Khazaeli, 2002 and Nuzhdin et al., 2005). A
second reason to
study age-specific patterns is that it becomes possible to
verify that the
ageing process is attenuated, as opposed to mere transient
improvements in
survival. For example, demographic studies show that the
beneficial effects
of dietary restriction are not due to a decrease in the rate
of demographic
ageing, but rather by decreases in the short-term risk of
death (Mair et
al., 2003). On a similar note, several mutants and genetic
constructs are
capable of prolonging lifespan, but it is unclear whether
this occurs by
modulating demographic ageing or by improving overall
condition, expressed
as better survival at all ages. Among other things,
information on
age-specific patterns of mortality in mutants is relevant to
evolutionary
theories of ageing, as it addresses assumptions on how
mutations affect
age-specific survival.

Another aspect that is of potential interest is the study of
pleiotropic
patterns that will be produced by regulatory signalling
networks. We have
established that longevity often correlates with resistance
against several
stresses, although none of the suggested mechanisms are
entirely successful
in explaining all variation in lifespan. Intuitively, a
plausible
evolutionary explanation of this fact presents itself: If
the integrity of
the soma is threatened by a multitude of threats, preserving
normal lifespan
will depend on the simultaneous recruitment of a wide
variety of stress
responses. This has two major implications. First, a genetic
correlation
between stress resistance or gene expression and longevity
does not
necessarily signify causality. More powerful genetic
techniques (e.g.,
overexpression of candidate genes) are required to show that
causal
relations exist, although these suffer from problems
associated with timing
and location effects. Recruiting the stress response by
applying mild stress
treatments, as suggested by advocates of hormetic treatments
(Rattan, 2005),
will have a better coordinated effect, but may suffer from
unwanted
pleiotropic effects. This leads us to the second
implication. Manipulation
of lifespan, which is mediated by attenuating signalling
pathways can
possibly elicit undesirable side-effects on other traits,
most notably
fertility (Leroi et al., 2005). Evolutionary studies have
repeatedly shown
that increments in lifespan through artificial selection
protocols or
environmental manipulations nearly always have a cost in
terms of growth and
fecundity (Zwaan, 1999). Such antagonistic regulation has
likely evolved to
survive periods of adverse condition. If such antagonistic
pleiotropy is
also found in humans, this may be an impediment in the
application of ageing
interventions.

6. Conclusion
Indisputably a link between stress resistance and longevity
exists in
Drosophila, as there is an appreciable overlap between gene
expression and
physiological responses to stress and ageing. Often, there
is a correlation
between expression of stress response genes and longevity
and overexpression
of stress genes can occasionally, although not invariably,
lead to increases
in longevity. These findings show that the study of the
stress response
remains a worthwhile field to explore with respect to ageing
research. The
mechanistic aspects of variation in stress resistance and
longevity have not
been fully elucidated yet, but large progress has been made
in identifying
genes that condition these traits. It is to be expected that
study of their
properties will lead to a better understanding of the
physiological
mechanisms underlying variation in lifespan and stress
resistance.

We feel that several topics warrant increased attention in
future studies.
Biodemography, the study of age-specific patterns can become
a valuable tool
to overcome the high complexity posed by the multifactorial
nature of
ageing. Age-specific patterns are already studied in
QTL-mapping studies
(e.g., Nuzhdin et al., 2005) but other type of studies could
profit from
this approach. For example, overexpression studies using
inducible promotors
(e.g., Bhole et al., 2004) could be greatly enriched by data
on the time
window in which genetic processes are relevant to survival,
which may not
necessarily be throughout life.

Many laboratories use careful genetic studies to test the
link of molecular
function to the phenotype, by studying several biological
levels
simultaneously. As discussed, the genes affecting lifespan
are, for
evolutionary reasons, expected to generate pleiotropic
effects, especially
on fertility (Leroi et al., 2005). Many studies support this
finding, as the
expression of stress response genes is often costly
(Sørensen et al., 2003).
In Drosophila, these effects are consistently found to be
present and they
are undoubtedly also present in other organisms. However,
the implications
of such pleiotropy for interventions in human ageing are barely
investigated. Up till now, this has been a neglected
direction of research,
which will grow rapidly more important, as prospects for ageing
interventions in humans come close.