cron-web.org

[A1CR]

The take-home message of the paper below may be:

"The extension of life span by diet restriction in
Drosophila has been
argued to occur without limiting calories. Here we ... find
that caloric
intake is reduced on all diets that extend life span."

Min KJ, Flatt T, Kulaots I, Tatar M.
Counting calories in Drosophila diet restriction.
Exp Gerontol. 2006 Nov 22; [Epub ahead of print]
PMID: 17125951

The extension of life span by diet restriction in Drosophila
has been argued
to occur without limiting calories. Here we directly measure
the calories
assimilated by flies when maintained on full- and
restricted-diets. We find
that caloric intake is reduced on all diets that extend life
span. Flies on
low-yeast diet are long-lived and consume about half the
calories of flies
on high-yeast diets, regardless of the energetic content of
the diet itself.
Since caloric intake correlates with yeast concentration and
thus with the
intake of every metabolite in this dietary component, it is
premature to
conclude for Drosophila that calories do not explain
extension of life span.

1. Introduction
Reduced food intake without malnutrition extends life span
in many organisms
including yeast, nematode, fruit fly, and rodents (Koubova
and Guarente,
2003, Masoro, 2000 and Partridge et al., 2005). In rodents,
reduced intake
of total calories extends life span, yet limiting specific
nutrients can
also increase survivorship (Miller et al., 2005, Yu and
Masoro, 1985 and
Zimmerman et al., 2003). It is a current debate as to
whether limiting
calories is a feature of nutrition responsible for extended
life span in the
fruit fly, Drosophila melanogaster. In recent work to
address this question
Mair et al. (2005) simultaneously manipulated dietary yeast
and sugar to
vary nutrient quality and caloric value. Survivorship was
increased
substantially on diets that restricted yeast while holding
sugar constant,
while life span changed little when sugar was restricted in
diets with a
constant amount of yeast. Notably, Mair et al. reasoned that
the actual
caloric intake was equal for females given diets with equal
energetic value
because solitary test females extend their proboscis for
equal durations.
Since females on isocaloric diets with low yeast are longer
lived than those
on high yeast, the authors argue that diet restriction (DR)
mediates life
span because it limits specific nutrient components of yeast
rather than
calories.

The conclusion of Mair et al. requires actual caloric intake
to be
proportional to the energetic content of the diet media.
However, the
relationship between proboscis extension and nutrient intake
is unknown, and
recent studies confirm that food intake may not be constant
across diets
that vary in yeast or sugar concentration (Carvalho et al.,
2005 and Min and
Tatar, 2006a). Females are seen to both increase and
decrease the rate of
intake when fed a restricted diet. It is thus possible that
the actual
caloric intake will differ among females as a function of yeast
concentration when they are presented with diets of similar
energetic value.

To test whether the energetic value of diet media is
proportional to actual
caloric intake, we replicated the diet manipulation
experiments of Mair et
al. and simultaneously measured survival and caloric flux. A
complete energy
budget would measure calories from metabolism, excretion,
changes in body
mass, and total allocation to eggs. We do not account for
metabolic rate
because it does not differ among fully fed and DR D.
melanogaster (Hulbert
et al., 2004). We do not account for excreta; to the extent
this rate
differs between treatments, we will at most underestimate
the discrepancy
between actual caloric flux and diet energetic value. In
practice,
therefore, we measure caloric flux as the total energetic
value of eggs and
of body tissue across five days, and across each of the diet
types described
in Mair et al. (2005).

2. Materials and methods
2.1. Demography and media
Larvae of the Canton-S strain were grown on standard
cornmeal/sugar/yeast/agar medium (Elgin and Miller, 1980),
supplemented with
several grains of live yeast. Newly enclosed adults were
collected over 48 h
and were assigned to 1 L demography cages to a density of
200 individuals
(100 females, 100 males). Food vials (25 × 95 mm) were
attached to each cage
via a 25 mm plastic tube and changed every 2 d, at which
time dead flies
were removed, sexed, and recorded. Three replicate cages
were established
for each of four diet treatments. Cages were maintained at
25 °C, 40%
relative humidity and a 12-h light-dark cycle.

The composition of each diet follows the design of Mair et
al. (2005),
although our media also contains a uniform concentration of
cornmeal (Table
1). Cornmeal acts as a colloid to maintain homogeneity of
the nutritive
sugar and yeast mixture. Dry, autolysed SAF yeast was
purchased from
Lesaffre Yeast Corporation (Milwaukee, WI). Energetic
content of media is
calculated from the individual caloric value and
proportional contribution
of each component. The high yeast/low sugar and low
yeast/high sugar media
were 'isocaloric' at about 99 kcal/100 ml.

Table 1. Diet composition, caloric content and conferred
life span
==========================================
Diet treatment Composition (in 100 ml water) Energy in 100
ml media, kcal
Median life span, d (95% CI)

==========================================
High yeast/high sugar 16 g yeast, 16 g sucrose, 5.2 g
cornmeal 146.8 30
(30-34)
High yeast/low sugar 16 g yeast, 4 g sucrose, 5.2 g cornmeal
98.8 32 (30-34)
Low yeast/high sugar 4 g yeast, 16 g sucrose, 5.2 g cornmeal
98.8 46 (42-48)
Low yeast/low sugar 4 g yeast, 4 g sucrose, 5.2 g cornmeal
50.8 48 (46-48)

2.2. Body and egg collection for bomb calorimetry
Larvae of the Canton-S strain were grown as described above.
Upon eclosion,
250 females and 50 males were sorted into 1 L demography
cages; four
replicate cages plus one spare were established for each
diet treatment.
Density in the replicate cages was held constant by
replacing dead flies
with same aged adults from the spare cage of the treatment.
We measured
caloric value of females within each replicate as the
calories of all eggs
laid from eclosion (day zero) through day five, plus the
calories of all
adults at the end of day five. Since cages were established
with adults that
developed on the same diet but before they consumed any
adult diet,
differences in the energetic value of tissue at day five
reflects
differences caused by adults feeding upon the various diets.

Eggs were collected from food media dishes (60 × 15 mm) that
were attached
by a 25 mm plastic tube and funnel to each cage. Dishes were
changed daily,
eggs were washed free, freeze-dried, and weighed. At day
five, food dishes
were removed for four hours before we collected adults to
ensure that flies
did not contain undigested food. All females from each cage
were
freeze-dried and weighed.

2.3. Calorimetry
We used combustion calorimetry (e.g., Lamprecht and Schmolz,
1999 and
Schmolz et al., 2005) to determine the heat content of
bodies and eggs
combined within replicates. Each sample was mixed with
calorimetry-grade
n-tetradecane (Sigma, St. Louis, MO) at a mass ratio of 1:5
(specimen:tetradecane) and combusted in a static jacket
oxygen bomb
calorimeter (Parr Instrument, Moline, IL; model No. 1341).
We previously
determined the optimal quantity of n-tetradecane to ignite
the insect
samples to full combustion, and we calculated the net
caloric value of the
sample by subtracting the known heat content of the primer.

3. Results
Life span (Table 1) and survival (Fig. 1A) was increased in
females
maintained on diets with dilute concentrations of yeast
(Median survival:
Low yeast, 46 d; High yeast, 32 d; Log-Rank test, ?2 =
333.7, p < 0.0001),
consistent with the survival outcomes reported in Mair et
al. (2005) and
others (Chippindale et al., 1993 and Min and Tatar, 2006a).
There was a
marginal but non-significant impact of reduced dietary sugar
upon life span
(Median survival: Low sugar, 38 d; High sugar, 34 d;
Log-Rank test, ?2 =
2.01, p = 0.16). Low yeast/high sugar diet and high
yeast/low sugar diet
have the same caloric content (Table 1), but flies fed those
diets have
significantly different life span as reported by Mair et al.
We plot our
results (Fig. 1B) in the same format as Mair et al. and
recapitulate their
published outcomes.

Fig. 1. Longevity, virtual calories, and caloric intake. (A)
Longevity.
Reducing yeast content of the diet had a much greater effect
on life span
than reducing sugar content of the diet. (B) Virtual
Calories: Plot of
median life span relative to diet energetic value. ... (C)
Assimilated
Calories: Plot of median life span (among replicate mean and
SE) relative to
calories of body and laid eggs per capita (among replicate
mean and SE).
Correlation between life span and assimilated calories is
strong and
significant.

In contrast to these results, flies fed isocaloric diets in
fact differ
markedly in assimilated calories. When energy influx is
measured by counting
the calories of fly soma and of eggs produced during 5 days,
per fly caloric
content is greater for flies fed yeast-rich diet than flies
fed yeast-poor
diet. There is a strong correlation between life span and
caloric content
(R2 = 0.98, F(1,2) = 109.99, p = 0.009, Fig. 1C).

We also observed patterns of nutrient assimilation that
address questions of
compensatory feeding. Dry weights of female body and eggs
laid during 5 days
were greater in flies fed yeast rich diet (Fig. 2), but the
extent of this
effect depended on the concentration of sugar in the diet.
Flies fed high
yeast diet produced more eggs when the diet had low sugar
than high sugar.
Similarly, flies fed high yeast/low sugar diet produced more
eggs and gained
more weight than flies fed low yeast/high sugar diet even
though these diets
were of the same energetic content.

Fig. 2. Body and egg mass of female by variation of yeast
and sugar
contents. Mass of females (ovary and immature eggs
inclusive) and produced
eggs was estimated from eclosion through 5 d old. High-yeast
induces high
egg production and weight gain. Note that low-sugar further
elevates egg
production.

4. Discussion
Food consumption has been explored across the decades of
building D.
melanogaster into an experimental model, although not with
any diet known to
extend life span (Driver et al., 1986, Edgecomb et al., 1994
and Tatar,
2007). The first study to retard aging by diet restriction
was reported only
some dozen years ago (Chippindale et al., 1993). Females
maintained with a
relatively dilute solution of dietary yeast extended life
span by 25-30% and
concomitantly laid fewer eggs. While dilute medium
presumably reduced yeast
consumption, feeding was not measured. Likewise, this early
design did not
aim to distinguish effects of calories relative from those
of specific yeast
metabolites. Studies to address these questions have only
appeared in recent
years.

Our own group simultaneously measured life span and feeding
rate of mated
females maintained on a constant agar-based diet of uniform
sugar with
varied concentrations of autolysed yeast (Min and Tatar,
2006a). Life span
was greatest upon diet of 2% yeast and progressively less as
yeast
concentration increased. We estimated feeding rate from the
uptake of a
soluble, non-digestible dye. Females on 16% yeast consumed
four-fold more
diet (dye) than those on 2% diet. This disproportionate
difference in
consumption suggests that concentrated yeast diet not only
confers a high
level of food intake but also stimulates feeding behavior,
perhaps in
response to the metabolic demands of elevated egg production.

Different outcomes were seen by Carvalho et al. (2005) where
virgin females
were presented with media that varied in both sugar and
yeast-extract;
survival was greatest upon 1% sugar-yeast diet. Consumption
was measured by
the uptake of a soluble nucleotide CTP[?-32P] tracer. In
contrast to our
observations, feeding rate increased upon progressively
dilute diets. Thus,
females upon the 1% diet ate less food (tracer) but the
quantity consumed
was only 4/10th the intake measured on 15% sugar-yeast diet.
Feeding was
stimulated at low diet concentrations, perhaps because these
media reduced
both sugar and yeast, or because compensatory feeding occurs
in the absence
of reproduction. Our new data also suggest there is
compensatory feeding
when sugar content is reduced because flies fed on high
yeast/low sugar diet
produced more eggs than flies fed on high yeast/high sugar diet.

While the dye and tracer studies confirm that flies consume
less food upon
dilute diet they do not address whether life span is
modulated by reduced
calories or by specific metabolites of the diet. To solve
this specific
problem Mair et al. (2005) independently varied the
concentration of sugar
and of yeast to produce diets with similar caloric content
but varied
composition. The survival of fecund females was markedly
increased on diets
that restricted yeast but not when sugar was limited.
Comparing life span
across diets that varied in quality but not in energetic
content provided a
way to assess the relative importance of specific nutrients
and calories.
The low yeast/high sugar and the high yeast/low sugar diets
were
energetically equivalent, yet females lived about 13 days
longer on low
yeast diet. Based on the frequency of proboscis extension in
undisturbed
conditions, Mair et al. (2005) argued that aging females
feed at the same
rate on each of the tested diets, and therefore that they
consume the same
amount of calories upon diets of similar energetic content.
Since females on
diets with low yeast concentration are longer-lived than
those on diets with
high yeast diet of the same energetic content, Mair et al.
(2005) concluded
that diet restriction mediates life span independent of
caloric intake.

This inference requires that nutrient acquisition is
proportional to
nutrient concentration in the diet. Our data here do not
support this
assumption. As reported, life span was strongly increased on
diets with
reduced yeast concentration, and there was little impact on
survival from
diets with reduced sugar, and when we plot our life span
data relative to
diet energetic value we recapitulate the results of Mair et
al. (2005).
However, females on iso-caloric diets did not assimilate the
same amount of
calories. Females fed low-yeast diet consume almost half the
calories of
females on high-yeast diet, and life span was strongly
correlated with
assimilated calories. Thus, females on low-yeast diet
consumed fewer
calories, less yeast and aged slowly. From this experimental
design it is
premature to exclude calories as a determinant of Drosophila
life span
because the intake of calories from yeast is confounded with
consumption of
all metabolites within yeast.

Ultimately it may be necessary to treat the fly more like we
do rodents:
directly control food intake or explicitly measure
assimilated nutrients.
Experimental regulation of food intake, in fact, has been
applied to the
Mediterranean fruit fly (Carey et al., 2002) and to the
housefly (Cooper et
al., 2004). In these relatively large flies there was no
positive effect of
reduced nutrient intake on life span, contrary to
precedence. Scaling such
protocols to study the effect of defined food intake with
Drosophila should
be a high priority.

Likewise, dyes and tagged nucleotides at best are partial
proxies of
consumption because they only mark solute intake. Adult
Drosophila does not
have chewing mouthparts and insoluble metabolites from yeast
embedded below
the media surface are relatively inaccessible. Measuring
stable isotopes of
carbon and nitrogen from diet can remedy this problem and
precisely measure
nutrient acquisition and metabolic flux. The butterfly
Heliconius
charitonius was shown through such methods to acquire
essential amino acids
from dietary pollen to produce eggs (O'Brien et al., 2003).
We recently
adapted this approach for Drosophila to determine how adults
use sugar of
the larval diet (Min et al., 2006). Turnover of carbon from
larval acquired
dietary sugar is rapid and nearly complete in the metabolite
pool used to
produce eggs and in adult somatic tissue itself. Applying
this approach with
the carbon and nitrogen from dietary yeast has the potential
to identify how
specific metabolites are acquired and allocated to somatic
maintenance when
diet restriction extends life span.

Efforts to clarify how the practice of diet restriction with
Drosophila
affects nutrient uptake are progressing but incomplete. It
is clear that
dilution of dietary yeast reduces food consumption and
extends life span. On
the other hand, we cannot yet resolve the effects of
calories from those of
metabolites specific to yeast, let alone distinguish the
effects of
components within yeast, such as carbohydrates, sterols,
fatty acids,
vitamins, minerals and amino acids. Amino acids deserve
attention because
reduced methionine extends life span in rats and in mice
(Miller et al.,
2005 and Zimmerman et al., 2003). Whether D. melanogaster
survival can be
improved by limiting dietary amino acids remains difficult
to assess because
defined diets optimized for larvae are not suitable for
adults (Lamb, 1978
and Min and Tatar, 2006b). It is perhaps ironic that while
the tools to
dissect the molecular basis of metabolism and aging in
Drosophila have
advanced tremendously in recent years we are currently stuck
on a
fundamental problem of gastronomy: How do we measure and
control what is
eaten by a fly?