Journalof Gerontology: BIOWGICAL SCIENCES
1999,Vol. 54A, No.4, 8137-8142
Copyright 1999 by The Gerontological Societyof America
Characterization of a Life-Extending Mutation in age-2,
a New Aging Gene in Caenorhabditis elegans
Yulong Yang and David L. Wilson
Departmentof Biology, University of Miami,Coral Gables, Florida,
We have generated a life-extending mutation, yw23, in Caenorhabditis elegans. The mutationis in what appears to be a
new aginggene, which we havedesignated age-2 When homozygous, yw23 produces an increase ofmean and maximum
life span of about 20% over that of the wild-type strain, N2. Strain HG23 [age-2(yw23)] was obtainedby screening for
longer life spans among 430 lines of nematodes two generations after exposure to the mutagen ethylmethanesulfonate.
StrainHG231 [age-2(yw23)] wasobtainedaftera singleout-crossing ofHG23to N2. When compared withN2, HG231 exhibits normal motility, slightly higher swimming rates, reducedfertility (especially at higher temperatures), somewhat
longerdevelopment times, and a slighdylargersize at the time offirst egglaying. A Gompertz analysis suggests that HG231
extendslife span by reducingthe initialmortality rate. In geneticcrosses, yw23 complements otherknown agingmutantsin
C. elegans genes-age-I, daf-2, spe-26, elk-I, elk-2, elk-3, and gro-I. A double-mutant strain, HG284, combiningmutationsin age-l and age-2, liveslongerthan animalswithindividual mutationsin eitherage-lor age-2, and exhibits a longer
life span at 25°Cthan at 20°e.
T
HE study of mutants that survive longer than wild-type
organisms may give insights into the nature of senescence
and its control. There are several such aging mutants in
Caenorhabditis elegans. Some, such as the mutants in age-I
and daf2 genes, may act by enhancing daf-16 gene expression,
which is related to the longer lived dauer state of the nematode
(1,2). Other mutants, such as those in the three elk genes and in
gro-l, appear to act by a general reduction in metabolism,
slowing development, and reducing rates of movement (3,4).
Mutations in spe-26, which produce sperm-defective worms,
may extend life span by an unknown mechanism (5).
We have generated a recessive mutation, yw23, in C. elegans
that increases mean and maximum life span by about 20%. It
appears to be in a gene that differs from those of the known
aging mutants. The singly out-crossed strain, HG231, which
contains the yw23 mutation, has a somewhat increased development time, but no reduction in swimming rate. HG231 also
has somewhat reduced fertility.
METHOD
GeneralMethods
The nematode C. elegans was cultured as described by
Brenner (6). Briefly, the nematodes were grown on NGM agar
plates seeded with Escherichia coli strain OP50, or, where indicated, were grown in liquid S medium, with E. coli at a concentration of about 109/mL (7). Animals were maintained at
15°C, and experiments were performed at 20°C or 25°C, as indicated. Wild-type was the N2, Bristol strain obtained from the
Caenorhabditis Genetics Center.
Mutagenesis
Ethylmethanesulfonate (EMS; final concentration 50 mM)
was used to generate mutations in stocks of N2-strain nematodes following the methods of Sulston and Hodgkin (7).
isolation ofyw23 Mutant
Following EMS treatment and after two generations of offspring (by self-fertilization), over 430 worms were randomly
picked and each placed on a separate agar plate with a lawn of E.
coli. These F2 worms were allowed to produce offspring, and the
life spans of each of the strains were assayed and compared with
that of N2. Our procedures generally followed those of Klass (8),
except that F2 worms were separated every other day from their
progeny. In this way we avoided the need to use a temperaturesensitive, sterile strain as the starting strain. To save some effort,
we used the survival of the individual F2 worms to guide us in
the selection of strains to examine in more detail. The strain
HG23 [age-2(yw23)] was selected for its increased life span.
Assay ofLife Spans
Five to 10 young adult hermaphrodite worms of a particular
strain were placed on an agar plate with a lawn of E. coli and
allowed to produce eggs for 4--6 hours. The worms were removed and the eggs allowed to develop. About 25 offspring
were transferred onto a new plate on the third day after egg laying. The life spans of these offspring were determined, with
transfers of the worms to a fresh plate each day during their
egg laying, and every other day during their postreproductive
period. In all experiments, zero day was the day of hatching, 1
day after egg laying.
Data Analysis
A C-Ianguage computer program was used to determine
mean and maximum life spans, and to determine best-fit parameter values for Gompertz, Weibull, and logistic survival
functions (9). Parameter values that minimize error, defined as
the sum of the squares of the differences between actual data
and calculated function values, were determined. The MannWhitney rank sum test was used to compare the life span of
different strains of nematodes (10).
B137
B138
YANG AND WILSON
Observations ofFeeding Behavior
C. elegans feeds through a bilobed pharynx, which pumps E.
coli into the intestine, crushing them as they pass through the
second lobe (11). The pumping of pharynx was observed under
a dissecting microscope.
MeasurementofSwimmingRates
About 1 day after animals reached adulthood, they were
placed in a Petri dish containing S medium. The frequency of
rhythmic thrashing of each animal was determined by counting
the number of cyclesduring 1 minute. In each case, counting was
begun about 5 secondsafterplacing the animalsin the solution.
MeasurementofDevelopmentTimes
The time from egg laying by one generation to the first egg
laying by the next generation of hermaphrodites was measured.
About 15 young, adult hermaphrodites were allowed to lay eggs
on a plate for 1 hour. Development was monitored until the time
of egg laying by newly formed young adults. There was significant variation among individualsin their time to first egg laying.
Time of first egg laying was recorded when 20 to 30 eggs were
present on the dish (from 15 to 50 individuals).
A
1.0
-
_..
0.9
0.8
0.7
c:
~ 0.6
~ 0.5
.~
~
0.4
C/)
0.3
0.2
0.1
0.0 +---+---t--t---+---+-----+-+__-+---+-----4I_ __+_---+-~___f
11
1
13
15
17
19
21
23
25
27
29
Time (days)
B
--411----_....
1.0
I--+-HG23
-+-N2
0.9
0.8
0.7
c:
Crosses Between Strains
Crosses were done with an excess of about 13 to 25 males of
one strain added to 5 to 10hermaphroditesof the other strain (7).
For the outcrossing of HG23[age-2(yw23)], male HG23
were crossed with N2 hermaphrodites. Worms that lived longer
than N2 were looked for among the F2 generation progeny.
HG231 [age-2(yw23)] was the strain selected after repeated
tests of life span.
In the case of complementation tests, strains used for males
and hermaphrodites are indicated in the text, and the life spans
of the F1 generation animals were monitored and recorded.
Development times (measured from time of egg laying in the
parent to time of first egg laying in the offspring) and fertility
also were monitored in some of the F1 progeny.
Construction ofthe DoubleMutant HG284[age-l(hx546) IT;
age-2(yw23)]
Male HG23[age-2(yw23)] were crossed with TJ401[age-l
(hx546)fer-15(b26) II] hermaphrodites. F1 hybridhermaphrodites
were allowed to produce offspring. From the offspring, 31
hermaphrodite adultswere each placedin individual dishesand allowed to produce self-progeny. Life spans of cohorts of the selfprogeny(about20 each)were determined, A line thatlivedsignificantlylonger thanTJ401 was selectedas HG284.
RESULTS
Isolationand Properties of yw23
HG23[age-2(yw23)] was isolatedfollowing EMS treatment, as
described in Methods. Figures la and b show that both HG23
hermaphrodites and HG23 males exhibitedlonger life spans than
their N2 counterparts. We have completed series of comparisons
of life spans between HG23 and N2, for hermaphrodites and
males.In a total of 29 trials, the age-2 mutant exhibited a mean increase of 17% (±3% SEM; p ~ .001 by paired t test) in median
longevity. It exhibited a mean increase in maximum longevity of
21% (±3% SEM). Each trialinvolvedthe analysis of survival for 6
~ 0.6
!!!
~ 0.5
.~
~
C/)
0.4
0.3
0.2
0.1
0.0
+---+-_+_--+-~-+__-+-_+_--t----+----4It--_+__+_--t-~___f
2
4
8
10
12
14
16
18
20
22
24
26
28
30
32
Time (days)
Figure 1. (A) Survival curves of hermaphrodite HG23 [age-2(yw23)] (n =
25) and the wild-type, N2, (n =25) with a temperature shift from 20°C to 25°C
on the fifth day.The differencebetween life spans was significant (p < .00(5).
(B) Survival curvesof male HG23[age-2(yw23)] (n = 25) and wild-type, N2, (n
= 26) at 20°C.The differencebetweentheirlife spanswas significant(p < .01).
to 30 individuals of each strain. Males sometimesexhibitedmore
significant increases than hermaphrodites.
HG23 was out-crossed with N2, as described in Methods, to
generate HG231[age-2(yw23)]. The out-crossed strain, HG231,
showedmuch higherfertility at 25°C than did HG23, but less than
halfthat ofN2. Fertility of HG231at 20°Cwascloserto thatofN2.
Hermaphrodites of HG231 developed more slowly than N2
hermaphrodites, requiringabout 25% more time to begin egg laying at both 20°C and 25°C. During development, body size was,
at first, somewhat smaller than N2, but HG231 hermaphrodites
reacheda largerbody size beforebeginning to lay eggs.
Interestingly, the swimming rate ofHG231 was slightly (5%)
greater (p = .(03) than that of N2. This contrasts with the elk
mutants, which, although showing longer development times,
were much slower in their movement rates (3). The locomotive
behavior and mating behaviors of male and hermaphrodite
HG231 worms appeared normal. Feeding behavior, including
pharyngeal pumping, appeared normal as well. The age-2 mutant strains,HG23 and HG231, did not form dauer larvae constitutivelyat 27°C, unlike age-l mutants (12).
age-2 GENE MUTANT
The FI hybrids, both hermaphrodite and male, of HG23 males
crossed with N2 hermaphrodites, or of HG23 hermaphrodites
crossed with N2 males, exhibited life spans that were very similar to the life span of N2 strain nematodes. Thus, yw23 is recessive to the wild-type allele. This result also indicates that the mutation is not on the X chromosome.
Complementation Tests Between yw23 and Other Known
Aging Mutants
A series of genetic crosses between HG23 and other known
aging mutants was performed. Fl hybrids of HG23[age2(yw23)] and TJ401[age-l(hx546)] had a life span similar
to that of N2 (Figure 2). Thus, the two mutants complement
one another in terms of life span, and are therefore in different
genes.
Similarly, HG23 complemented the daf-2 mutant strain
CB1370[daf2(e1370)] for life span (Figure 3). The life span of
FI hybrids was less than that of HG23 and similar to that of
, 1.0
-a:-........
BI39
N2. Furthermore, at 25°C all FI hybrids grew to adulthood
without forming dauer larvae, unlike the daf2 mutant. The FI
hybrids also showed normal fertility at 25°C, unlike daf2.
We crossed HG23 males with BA82I[spe-26] hermaphrodites. Hybrids exhibited normal fertility. Because BA82I is
sterile at 25°C and HG23 and HG23I exhibit lower fertility at
25°C than does N2, it appears that yw23 and the spe-26 mutations complement, at least in regard to fertility. In our hands,
BA82I did not exhibit a longer life span than N2.
We crossed HG231[age-2(yw23)] males with mutants elk-I
(strain CB4876), clk-2 (MQI25), clk-3 (MQI31), and gro-l
(CB45I2) hermaphrodites. In each case, the Fl hybrids developed to adulthood as fast as N2, and faster than did any of the
individual mutants. Thus, the age-2 gene appears to differ from
the four genes where these "clock" mutants are located.
Preliminary results of linkage testing with polymorphic sequence-tagged sites (13,14) suggest that age-2 may be located
on linkage group 1.
_._---.....
___ HG23
-+-HG23
___
N2
....... N2
0.9
0.9
........ Fl (HG23 M.X TJ401 H.)
-+-TJ401
0.8
0.8
0.7
0.7
0.6
i
u,
05
~ 0.5
~
04
i
c:
c:
~
rn
........ Fl (yw23 M.X daf-2 H.)
....... Fl (yw23 M.X daf-2 H.)
-+-da'-2(e1370)
0.6
.~
.
::> 0.4
rn
0.3
0.3
0.2
0.2
0.1
0.1
0.0
0.0 +--+---+--+--+---+-___+--+___1I--+--+--=>e-~.......-+---+-___+~___1
o 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
+--+-+--+-+-+-+--+---1~
a
2
4
6
---'~-+-+-+-+--+---1~-+-+-t-+--+-+-1
8 10 12 14 16 18 20 22 24 26 28 30 32 34 35 38 40 42 44 46 48 50 52 54
Time (days)
Times (days)
Figure 2. Survivalcurves of hermaphrodites of HG23[age-2(yw23)] (n = 25),
TJ401[age-l(hx546) II] (n = 25), and F1 hybrids ofHG23 males crossed with
TJ401 hermaphrodites (n = 22). There was a temperature shift from 20°C to
25°C on the fifth day. Survivalof F1 hybrids is similar to that of N2.
Figure 3. Survivalcurves of hermaphroditesof HG23[age-2(yw23)] (n = 25),
N2 (n = 25), CB1370[daf2(e1370)], and two groups of Fl hybrids of HG23
males crossed with CB1370 hermaphrodites at 25°C. Survival of both groups
ofFl hybrids is similarto that ofN2.
Table 1.TheAnalysis of Survivaland Mortality Data of the Hermaphrodites of the NewAging MutantHG231
and theWild-Type StrainN2 (at 25°C)
Parametervalues*
Error (STD)t
Typeof Function and RegressionAnalysis
Strain
Non-linear regression analysis on
Gompertz survival function (weighted)
HG231
N2
A = 0.00003965
A = 0.00033073
G = 0.4728
G = 0.4359
Non-linear regression analysis on
Weibull power survival function (weighted)
HG231
N2
A = 2.16693e-11
A = 4. 1366ge-08
g = 8.95614
g = 6.75741
0.02035
0.00945
Non-linear regression analysis on
logistic survival function (weighted)
HG231
N2
v = 18.88
v = 15.39
w = 13.3672
w = 10.0636
0.02646
0.02680
Mean and median life span
HG231
N2
Mean ± SEM = 18.8 ± 0.2 (days), median = 19.0 (days), N = 150
Mean ± SEM = 15.4 ± 0.3 (days), median = 15.6 (days), N = 100
P = .0001 [Mann-Whitney rank sum two-sample test (2-tailed)]
MRDT= 1.47
MRDT= 1.59
0.02474
0.01828
*The full definition of parameter values is given in (9). ParameterA is the Gompertz initial mortalityrate, G is the Gompertz exponentialgrowth parameter, and
MRDT is the mortality rate doubling time. Parameters a and g are for a Weibull(power) survival function (9). For logistic survival,parameter v is estimated median
life span, while w/v is a measure of the steepnessof the slope of the survivalcurve (9).
tThe errors in experimentalvalue of s versus regressionestimates are givenby the square root of [(weighted)SSR / n-2], where n = number of time intervals (data
points), and (weighted) SSR = weighted (by the square root of the number dying in each time interval)sum of the square of the differencesbetween the calculated
valuesof survivaland actual values.
YANG AND WILSON
BI40
Table2. The Fertility of HG23I, TJ40I, the Wild-Type StrainN2, and the DoubleMutantHG284 at 20°Cand 25°C
200 e
Mean (numberof progeny/worm)
STD
N (numberof animals)
25°e
N2
HG231
HG284
TJ401
N2
296.1
73.2
13
226.4
33.7
16
0.0059
249.0
40.0
8
0.0499
0.0887
19.7
6.1
3
0.0001
0.0001
166.5
48.6
17
p*
pt
HG231
HG284
TJ401
71.2
24.7
88.7
35.9
0.0
0.0
10
10
0.0001
0.0001
0.2221
10
0.0001
0.0001
*p valuefor comparisonwithN2 using the two-tailedStudent's t test (two sampleswith unequalvariance).
tp valuefor comparisonwithHG231 usingthe two-tailedStudent's t test (two sampleswithunequalvariance).
_ _ N2
Table 3. Fertility of TJ40I, the Wild-Type StrainN2, the Double
MutantHG284,and the FI Hybrids of HG284Males Crossed
WithTJ40I Hermaphrodites (at 25°C)
.......TJ401
_ _ HG284
0.9
0.8
Strain
0.7
j
c:
_
.~
::J
en
0.6
N2
0.5
Mean (number of
progeny/worm)
0.4
0.3
STD
N (number of
0.2
HG284
PI (HG284 M. X
TJ401 H.)
TJ401
166.5
88.7
94.8
0.0
48.6
17
35.9
10
22.6
6
0.0
10
animals
0.1
0.0 +-+--+--I-+-t---+-+--t--+-+~-+-+--l--+--+--+--t---+-+--+-~.....-l
o 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 36 40 42 44 46 48
Time (days)
Pigure 4. Survivalcurves of hermaphrodite TJ401 [age-l(hx546)] (n = 25),
HG284[age-l(hx546) II; age-2(yw23)] (n =25),andwild-type N2 (n =25) at 25°C.
p*
pt
0.0001
0.0001
0.6817
0.0001
0.0001
*p value for comparison with N2 using the two-tailed Student's t test (two
sampleswithunequalvariance).
tp value for comparison with HG284 using the two-tailed Student's t test
(two sampleswithunequalvariance).
SurvivalParametersforHG23i[age-2(yw23)]
A population of 150 HG23 1 animals was studied to establish
survival and mortality parameters for this strain. These are
shown in Table 1. The definitions of the parameters, and the
equations from which they come, can be found in (9). Mean
and median life span for HG231 were 22% longer than that for
N2. HG231 showed a slightly shorter mortality-rate doubling
time than N2, but also showed a lower initial (Gompertz) mortality rate. It appears that the extension of life span of HG231
may be the result of either its being more robust earlier in life,
or its exhibiting a delay in the onset of senescence. This population showed a slightly faster rate of increase of mortality late
in its life span, compared to N2. Of the three functions that
were fit to the survival data for both HG231 and N2, Gompertz,
Weibull, and logistic (9), both survival curves were better fit by
the Weibull function, which corresponds to a power function
increase in mortality rate with time. However, all three functions gave a reasonable fit to the survival data of this relatively
small population.
Characterization ofan age-I, age-2 Double Mutant
HG284, an apparent double mutant of age-i and age-2, was
isolated as described in Methods. HG284 showed a life span
that is longer than that of strains with mutations in age-I or
age-2 alone (Figure 4). On average, HG284 animals lived more
than twice as long as N2 animals.
To confirm that HG284 contains the age-2 mutation, we
crossed HG231[age-2(yw23)] males with HG284 hermaphrodites. The life span of F1 hermaphrodites was similar to that of
HG23I[age-2] and longer than that of N2 (p :;;:: .00(1), as expected. To confirm that HG284 contains the age-i mutation,
HG284 males were crossed with TJ401[age-i(hx546)]
hermaphrodites. The F1 hermaphrodites showed a life span
slightly longer than TJ401 hermaphrodites, indicating that the
age-i mutant allele still exists in HG284.
HG284[age-i(Ju:546) /L' age-2(yw23)] animals exhibited development times that were similar to N2 at both 20 and 25°C. The
double mutant exhibited a 14% higher swimming rate than N2 (p
:;;:: .001), significantly greater than the swimming rate of
HG231[age-2(yw23)] (p:;;:: .001). The double mutant appeared normal in feeding and mating behavior. All HG284 animals formed
dauer larvae constitutivelyat 27°C, as did age-l mutant animals.
Fertility of the double mutant strain was similar to that of the
age-2 mutant strain, at both 20 and 25°C (Table 2). Its fertility,
although lower than that of N2, was much higher than that of
one of the parent strains, TJ401. TJ401 was sterile at 25°C because of the additional mutation, fer-l S, present in this strain~
This suggests that the double mutant HG284 may have lost the
fer-i5 mutant allele present in TJ40I, a somewhat surprising
result because fer-I S is located very close to the age-i gene on
linkage group II (15).
The possible loss oi fer-LS in HG284 was tested by examining fertility in the FI hybrids of the cross between HG284
males and TJ401 hermaphrodites. Only a moderate reduction in
fertility was observed (Table 3), suggesting thatfer-i5 was no
longer present in the double-mutant strain.
age-2 GENE MUTANT
B141
Table4. The Analysisof Survival and MortalityData of the Hermaphrodites of the DoubleMutantHG284 at 20°C and 25°C
Temperature
Type of Function and Regression Analysis
Parameter values*
Errors (STD)*
=
G
G
A
= 1.51128e-1O
A
= 1.2254ge-07
g = 7.13486
g =4.85179
0.03096
0.02837
=10.8127
=7.1367
0.01848
0.05159
Non-linear regression analysis on
Gompertz survival function (weighted)
A 0.00012746
A = 0.00074204
Non-linear regression analysis on
Weibull power survival function (weighted)
Non-linear regression analysis on
logistic survival function (weighted)
v
v = 33.71
=29.41
w
w
=0.2398
=0.1419
=
MRDT 2.89 days
MRDT = 4.88 days
0.04073
0.01479
Mean ± SEM = 29.6 ± 0.3 days, median = 29.3 days,
maximum =44 days, number of animals = 275
Mean ± SEM = 33.1 ± 0.6 days, median = 34.8 days,
maximum = 53 days, number of animals = 250
p s .0001 [Mann-Whitney rank sum two-sample test (2-tailed)]
Mean, median, and maximal life span
*See Table 1 for definitions of parameters and error.
0.9
0.8
0.7
§
~
0.8
l!!
~ 0.5
.~
Jl
0.4
0.3
0.2
0.1
0.0 +--+--+--+-+-t--t--t-+--+-l-+-~I--t--+--I--t--+-+-+:::::'-'...-+~""_
o 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 48 48 50 52 54 58
Time (days)
Figure 5. Survival curves of hermaphrodite HG284[age-l(hx546) 11; age275) and at 25°e (n 250). The difference between life
spans at these two temperatures was very significant (p < .0001), with the animals kept at higher temperature living longer.
2(yw23)] at 20 0 e (n
=
=
HG284[age-I(hx546) II; age-2(yw23)] animals showed a lower
initial mortality rate and a much longer mortality-rate doubling
time (smaller Gompertz parameter) than N2 animals (Table 4;
compare with Table 1 values for N2). It appears that the double
mutant has a reduced early mortality, or delayed onset of senescence, as well as a slower rate of accelerationof mortality with age.
Data in Table 4 and in Figure 5 also indicate that the double
mutant showed the unusual property of living longer at 25°C
than it did at 20°e. It showed a 10-20% longer mean, median,
and maximum life span at the higher temperature. The double
mutant exhibited a life span similar to the age-l mutant at
20°C, but their life spans at 25°C were significantly different.
DISCUSSION
We conclude that yw23 is an aging mutation in a new gene,
age-2, not previously characterized as a gene in which mutations
can induce increased life span. Both males and hermaphrodites
of strain HG231[age-2(yw23)] have normal appearance and
show normal feeding and mating behaviors. HG231 strain animals develop more slowly than wild-type N2, a characteristic
that they share with the elk mutants, but unlike the elk mutants,
they do not show a reduced rate of swimming movements.
Instead, their swimming movements are slightly faster than those
of N2 strain animals. Fertility of HG231 is lower than that of N2,
especially at higher temperatures (25°C).
We cannot be certain if some of these properties, such as
lower fertility and a somewhat longer development time, exhibited by HG231, are the consequence of the yw23 mutation or of
other mutations that could be present in the strain. It is possible
that other mutations remain present after the single out-crossing
of HG23 with N2.
The somewhat larger size of HG231, compared to N2, suggests
a possible similarity to the mechanism of life span extension with
that seen in some of the longer lived strains of Drosophila (16),
which also show increased body size and somewhat longer larval
development. It may be that size or food reserves plays a role in
reducing mortality rates in these invertebrates.
The extension of mean and maximum life span by about
20%, which is caused by the yw23 mutation, seems to be the result of a lower initial mortality rate compared to N2. HG231
may actually show a slightly faster rate of acceleration of mortality with age than does N2. This is unlike TJ401 [age-I] which
shows both a lower initial mortality rate and a reduced acceleration of mortality with age (15). The survival curve for HG231 is
best fit by a Weibull (power) survival function, but the fit by
Gompertz or a logistic function also is good. The life-extending
effects of yw23 appear to be more obvious at 25°C than at 20°C.
In summary, the phenotype produced by the yw23 mutation in
age-2 is an increased life span that may be related to a more robust initial adult state. The yw23 mutation also appears to cause
a slight increase in swimming rates.
Complementation tests between HG231 and each of the other
known aging mutants indicate that the yw23 mutation is in a gene
that differs from the other known aging genes.
At 25°C, the age-l, age-2 double mutant causes a further increase in life span, more than the sum of the increases shown by
each mutant alone, and more than double the life span of N2.
This is not the only double-mutant that exhibits a synergistic enhancement of life span in C. elegans. Larsen and colleagues (17)
have reported that a combination of daf-2(eI370) with dafI2(m20) shows a quadrupling of normal life span. The OOf-2
mutant is known to double life span, but daf-12 alone does not
exhibit a longer life span than N2. Both are dauer mutants.
Bl42
YANG AND WILSON
The combination of age-I and gro-I also shows dramatically
enhanced life span over that of either mutant alone, and such
double mutants live nearly five times as long as N2 (4). Some of
the double mutants among the elk genes also show increases in
life span, especially at lower temperatures (4).
Synergistic life span increases have been shown before by
double mutants in aging genes. However, the unusual ability of
the age-L, age-2 double mutant to survive for longer times
when kept at 25°e than at 20 0 e is more unique. One or both of
the mutants may exhibit temperature sensitivity, and may more
strongly alter or block function of their respective gene products at the higher temperature. If so, the temperature sensitivity
must be dramatic because it reverses the normal situation of
shorter life spans at higher temperatures for poikilothermic animals such as C. elegans (18,19).
C. elegans has been the subject of a number of aging studies,
but a recent review pointed to the need for additional screens
for life-span mutants (20). Mutations, such as yw23, which
show a smaller increase in life span than age-lor daf-2 mutants, are more difficult to detect and study because the smaller
increase requires careful analysis of the survival curves from
larger populations of nematodes to gain significance.
Nevertheless, analysis of such mutants exhibiting smaller increases in life span may reward us with new insights into the
genetics of aging and new knowledge of how genes control or
affect senescence.
ACKNOWLEIXJMENTS
We gratefully acknowledge research support from the American Federation
for Aging Research. We thank the Caenorhabditis Genetics Center and Tom
Johnson for C. elegans strains. Address correspondence to David Wilson,
Department of Biology, University of Miami, P. O. Box 249118, Coral Gables,
FL 33124-0421. E-mail: [email protected]
REFERENCES
1. Lin K, Dorman JB, RodanA, Kenyon, C. daf-16: An HNF-3/forkhead
family member that can function to double the life-span of Caenorhabditis
elegans. Science. 1997;278:1319-1322.
2. Ogg S, Paradis S, Gottlieb S, et al. The fork head transcription factor DAF16 transduces insulin-like metabolic and longevity signals in C. elegans.
Nature. 1997;389:994-999.
3. Wong A, Boutis P, Hekimi S. Mutations in the elk-I gene of
Caenorhabditis elegans affect developmental and behavioral timing.
Genetics. 1995;139:1247-1258.
4. Lakowski B, Hekimi S. Determination oflife-span in C. elegans by four
clock genes. Science. 1996;272:1010-1013.
5. Van Voorhies WA. Production of sperm reduces nematode lifespan.
Nature. 1992;360:456-458.
6. Brenner S. The genetics of Caenorhabditis elegans. Genetics. 1974;77:
71-94.
7. Sulston J, Hodgkin J. Methods. In: Wood WB, ed. The Nematode
Caenorhabditis elegans. Cold Spring Harbor, NY: Cold Spring Harbor
Laboratory Press; 1988:587-606.
8. Klass MR. A method for isolation of longevity mutants in the nematode
Caenorhabditis elegans and initial results. Ageing Dev. 1983;22:279-286.
9. Wilson DL. The analysis of survival (mortality) data: fitting Gompertz,
Weibull, and logistic functions. MechAgeing Dev. 1994;74:15-33.
10. Fisher LD, Van Belle G. Biostatistics: A Methodology for the Health
Sciences. New York: John Wiley and Sons; 1993.
11. Wood, WB. Introduction to C. elegans biology. In: Wood WB, ed. The
Nematode Caenorhabditis elegans. Cold Spring Harbor, NY: Cold Spring
Harbor Laboratory Press; 1988:1-16.
12. Malone EA, Inoue T, Thomas JR. Genetic analysis of the roles of daf28
and age-1 in regulating Caenorhabditis elegans dauer formation.
Genetics. 1996;143:1193-1205.
13. Williams BD, Schrank B, Huynh C, Shownkeen R, Waterston RH. A genetic mapping system in Caenorhabditis elegans based on polymorphic
sequence-tagged sites. Genetics. 1992;131:609-624.
14. Williams BD. Genetic mapping with polymorphic sequence-tagged sites.
In Epstein HF, Shakes DC, eds. Caenorhabditis elegans: Modern
Biological Analysis of an Organism. San Diego, CA: Academic Press;
1995:81-96.
15. Johnson TE. Increased lifespan of age-l mutants in Caenorhabditis elegans and lower Gompertz rate of aging. Science. 1990;249:908-912.
16. Partridge L, Fowler K. Direct and correlated responses to selection on age
at reproduction in Drosophila melanogaster. Evolution. 1992;46:76-91.
17. Larsen PL, Albert PS, Riddle DL. Genes that regulate both development
and longevity in Caenorhabditis elegans. Genetics. 1995;139:1567-1583.
18. Klass MR. Aging in the nematode Caenorhabditis elegans: major biological and environmental factors influencing life span. Ageing Dev. 1977;6:
413-429.
19. Finch CEo Longevity, Senescence, and the Genome. Chicago: Univ. of
Chicago Press; 1990.
20. Kenyon C. Environmental factors and gene activities that influence life
span. In Riddle DL, Blumenthal T, Meyer BJ, and Priess JR, eds. C. elegans II. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press;
1997:791-813.
Received June 5, 1998
Accepted November 9, 1998