INTRODUCTION

Taurine (2-Aminoethanesulfonic acid) is a β-amino acid, synthesized de novo from the amino acid cysteine by way of cysteine dioxygenase (Stipanuk, 1986). There is increasing evidence regarding taurine as a key amino acid in both vertebrates and invertebrates, responsible for the proper function of a wide range of physiological systems (Oja and Saransaari, 2007). Taurine has been shown to be a vital nutrient in the diet of felines, which are unable to synthesize taurine, for proper development of the retina (Schmidt et al., 1977; Knopf et al., 1978). Studies have also been conducted in dogs and some species of fish which suggest that taurine supplementation may be necessary in commercial dog and fish food for proper nutrition (Backus et al., 2006; Lunger et al., 2007). The physiological roles of taurine include absorption of fat, maintenance of proper osmolarity, anti-oxidant capabilities and as a neuromodulator (Huxtable, 1992; Oja and Saransaari, 1996). In captive animals, such as lions and polar bears, taurine is an important part of the diet needed to prevent rickets by aiding in the absorption of fat soluble vitamins (Chesney and Hedberg, 2009; Chesney et al., 2009). Taurine acts as an osmotic regulator in several fish and bivalve species, allowing the marine animals to react to changes in salinity (Avella et al., 2009; Inoue et al., 2008; Hosoi et al., 2007).

Research has shown levels of taurine differ in Drosophila tissues and in hemolymph (Massie et al., 1989; Piyankarage et al., 2008). Identification of a taurine transporter, dEAAT2, in Drosophila has implicated taurine to act as a neuromodulator (Besson et al., 2005). In humans, taurine is the second most abundant amino acid within brain tissue. In the nervous system, taurine can inhibit the firing of neurons by increasing chloride conductance through activation of GABAa and glycine receptors within the nervous system (Albrecht and Schousboe, 2005). Additionally, taurine has been shown to act as a neuromodulator and neuroprotective agent by reducing CaMKII activity, providing protection from epilepsy-inducing agents and preventing excitotoxic damage by activating GABA_A receptors (Junyent et al., 2009, 2010; Louzada et al., 2004). Studies on neural tissue and insulin sensitivity in rats have shown that taurine acts as an anti-oxidant, protecting neural tissues from arsenic-induced oxidation and improving insulin sensitivity in hyperglycemic subjects (Das et al., 2009; Haber et al., 2003). Taurine has been shown to reduce the oxidative stress of the heavy metal contaminant, cadmium, in fish and mice (Kumar et al., 2009; Sinha et al., 2008). The multitude of contributions that taurine makes to physiological functions have established taurine as an important component of life and suggest that taurine supplementation can confer protection from various forms of degeneration.

The processes that ultimately contribute to cessation of life and factors that alleviate degeneration are not extensively understood. While many organisms do not experience senescence as a result of aging, most organisms including Drosophila, experience some form of functional decline (Helfand and Rogina, 2003). Drosophila melanogaster has been the model system of choice for many life span studies that examine the effects of environmental influences, genetic adaptation and behavioral changes on life span. Life span studies allow a cohort of flies to be studied in order to identify physiological problems that can become evident in a reduced life span or identify any benefit to the organism if the life span is extended. Factors such as energy consumption, energy expenditure and oxidative stress, among many others, can affect the initiation and/or rate of mortality in Drosophila (Helfand and Rogina, 2003). The goal of the present study is to examine the effect of taurine, a known anti-oxidant, anti-inflammatory agent, metabolic enhancer, osmoregulator and neuromodulator on Drosophila life span and development (Oja and Saransaari, 2007; Zamboni et al., 1993; Haber et al., 2003; Junyent et al., 2009; Murakami et al., 2002; Tsuboyama-Kasaoka et al., 2006).

MATERIALS AND METHODS

Fly stocks and conditions: The wild type strain, Canton S (Bloomington stock number 1), was reared at 23°C on cornmeal-molasses food as normal food or supplemented with the addition of specific concentrations of taurine (Acros Organics). Virgin flies were selected within 8 h of clearing the vials for all studies. Flies were passed and counted twice a week for life span studies unless otherwise specified.

Developmental dose curve: Cages of approximately 200 flies were acclimated to food plates with either normal or taurine supplemented food for at least 3 days. Grape juice plates with yeast paste were used to collect eggs to be transferred to food vials with the appropriate food. Eggs were collected for no more than 2 h each day until 100 eggs were collected for each concentration of taurine. For 18 days, the progress of the eggs was monitored daily to indicate the number of eggs hatched, the number of pupae produced and the number of adults that emerged.

Life span studies: Ten virgin Canton S flies, either randomly or equal numbers of males and females, were placed in cornmeal-molasses food vials. For the taurine supplemented vials, appropriate concentrations of taurine were added to the food mixture upon cooling right before filling the vials. Twice a week, the flies were counted and transferred to fresh vials throughout the life spans. Median age of flies was calculated by subtracting the number of flies on the current date from the previous date, then multiplying the result by the current number of days. These calculations from throughout the life span were added together and then divided by the total number of flies in the experiment (Dr. Blanka Rogina, personal communication).

Statistical analysis: When comparing multiple treatments, such as in the progeny development study, analysis of variance (ANOVA) was used. Multiple pairwise comparisons were conducted using the Bonferroni correction method for the significance level. This technique adjusts the significance level needed to reject the hypothesis of equality of two treatments by taking the original significance level and dividing by the number of comparisons. This technique preserves the overall error rate at the original significance level (Neter et al., 1996).

For all life span studies, survival analysis was used. Survival analysis was appropriate to study the time to death of the flies with a given treatment. Survival modeling allows the data to be censored. In fact, the data collected was actually interval censored because the event of death was not exactly observed but did occur in some interval of time between the days the vials were examined.

Three different parametric distributions for creation of the survival curves (Weibull, lognormal and logistic) were evaluated for fit to the data. The best fitting distribution was the Weibull distribution. The Weibull distribution is identified by its shape (α) and scale (β) parameters (Cassell and Berger, 2002; Allison, 1995). Once a survival curve was created for each treatment, the shape and scale parameters were statistically tested between treatment groups to determine whether or not there was a significant difference between the treatment group survivorship curves. ANOVA was used to determine if there was a statistical significance between at least two of the treatment survival curves. If the ANOVA was rejected, multiple pairwise comparisons were conducted using the Bonferroni correction to test the difference between the shape and scale parameters of the survival curves. The Bonferroni correction takes the original significance level and divides by the number of comparisons to maintain the experimental error rate. If either the shape or scale was significantly different, this would mean the survival curves were statistically significant. Using ANOVA over the traditional Kolmogorov-Smirnov test was beneficial due to the discrete nature of the data with interval censoring (Cesani et al., 2006; Mentre and Escolano, 2006). The Kolmogorov-Smirnov test should only be used when the data is continuous (NIST/SEMATECH, 2003). Also, the Kolmogorov-Smirnov test does not easily generalize to allow testing of multiple curves (Dasari et al., 2007).

Finally, a two sample t-test was used to determine the significance of the median and maximum ages between normally fed and taurine fed flies.

All analysis were conducted using the statistical software R (R Development Core Team, 2009). The survreg function in R was used to create the survival models. The ANOVA function was used in R to conduct ANOVA.

RESULTS

Effects of taurine concentration on progeny: Earlier study on the effects of taurine demonstrated adults fed 200 mM taurine is lethal to the progeny; however, the exact stage and concentration has not been identified (Massie et al., 1989). We established cages of 200 wild type Canton S mating flies and acclimated the adults to normal food and various concentrations of taurine for 4 days.

Fig. 1:

Taurine dose response of eggs developing into adults. Eggs laid on grape juice plates were transferred to vials of food with 0, 5, 10, 20, 50, or 100 mM taurine and allowed to develop for 18 days. Taurine at concentrations of 10 mM and higher affect the hatching of eggs to (a) larva. Further development of the larvae to (b) pupae and to (c) adults is only slightly affected by the presence of taurine in the food. Two time points are shown to demonstrate early and late time points for each stage. ANOVA and multiple pairwise comparisons used to determine the significance of the late time points, ***p<0.0001

After 30 min, twenty eggs were transferred from the grape juice plates to vials containing the appropriate food and checked daily for developmental progress. As shown in Fig. 1a, there is a dose response in the number of eggs that are hatched. Two time points are shown in the figures: an early time point in which a number of larvae (24 h), pupae (8 days) and adults (12 days) were observed in the control treatment and corresponds to normal development of the fly and a later time point in which all of the developmental stages should have been complete (48 h, 11 and 15 days). The early time points show that taurine delays the development of all stages, particularly at 10 mM taurine and higher. As the taurine concentration increased, egg hatching at the later time point was reduced from 97% to less than 20% at 20 mM taurine (Fig. 1a). The number of eggs that developed into pupae at the early and late time points for each taurine concentration shows that 10 mM and greater taurine affects the development of pupae (Fig. 1b). From egg to adults, the lowest concentration of taurine, 5 mM, does not affect the development of adult flies (Fig. 1c). The decrease in the number of eggs hatched, eggs to pupae and eggs to adults was statistically significant for taurine concentrations of 10 mM and higher for the later time points (pairwise t-test, p<0.0001). The percentage of eggs developing into pupae and adults is approximately the same as the number of eggs that hatch for 10 mM taurine, suggesting that development beyond the first instar larvae at 10 mM taurine is not affected. Thus, exogenous taurine is most detrimental at the early stage of egg to first instar larva. These results confirm the data reported by Massie et al. (1989). We also repeated this experiment providing only normal food to the mating flies in the cages followed by the transfer of eggs to the various concentrations of taurine. We obtained similar results to our initial experiment, which provided each taurine concentration to the mating adults and the developing eggs, suggesting that the effects of taurine are not due to mating problems (inability to fertilize the eggs) or due to maternal contribution (data not shown).

Life span on taurine-supplemented food: Wild type flies (Canton S) selected within eight hours after enclosion from pupae were collected and culled 10 per vial (total of 10 vials per treatment per trial). A random mixture of males and females were placed in the vials. The flies were fed a cornmeal-molasses food alone or supplemented with 100 mM taurine. Concentrations for taurine were based on the data from the experiments performed by Massie et al. (1989) in which it was shown that 50, 100 and 200 mM taurine supplemented in Formula 4-24 Instant Drosophila medium did not affect the life span of the wild type strain Oregon R. Using the wild type strain, Canton S, we confirmed that no progeny were produced by adults fed 200 mM taurine and higher; however, the life spans of Canton S flies at 200 mM taurine and higher resulted in premature death after 2 weeks (data not shown). Thus, we continued life span studies using 100 mM taurine. Flies were transferred and counted twice a week over the life span. Red dye was added to the food to allow visual inspection of the fly abdomens for indication that flies were consuming the food. We observed both normal and taurine fed flies consumed similar amounts of the food. Three separate life spans of control and taurine fed flies (n = 3) were independently completed. The median life span for flies fed normal food was 33.68±0.41 (SE) days while supplementing the food with 100 mM taurine the median age was 38.57±0.95 (SE) days. The median age of taurine fed flies is significantly larger than normal fed flies (t-test, p = 0.0111). The extension of life span could especially be seen in the maximum life span (median age of the last 10% of flies) where taurine fed flies lived maximally to 99.98±8.73 (SE) days while normal fed flies lived for 72.56±2.56 (SE) days. The maximum age of taurine fed flies again is significantly larger than normal fed flies (t-test, p = 0.0390).

In order to compare the trend over the entire life spans of normal and taurine fed flies, survival analysis using the Weibull distribution was used. As stated in the Materials and Methods section, the Weibull, lognormal and logistic distributions were investigated for best fit. An example of the fit can be seen in Fig. 2. The Weibull does the best overall fit compared to the other curves because of its performance at the beginning and ending of the life span. It is also a decent fit for the middle section. The logistic curve does an insufficient job at the beginning time points by underestimating the percent alive. The lognormal underestimates events in the middle of the life span. All the life span experiments followed similarly in the fit of the curves to data. Therefore, the Weibull distribution was used throughout the paper to fit survival curves.

Flies fed 100 mM taurine-supplemented food over the course of the life span followed a similar decrease in survivorship as flies fed normal food over the first 30 days (Fig. 3). After 30 days, the decline in the percent of flies alive is steeper in the normal fed flies than taurine resulting in the extension of the life span. ANOVA was conducted and found that at least two of the survival curves were statistically significant in either the shape, scale, or both parameters. Therefore, pairwise comparisons were done using a Bonferroni correction for the significance level of α = 0.05/ (30 comparisons) =0.0017. First, it can be seen in Fig. 3, Table 1 and 2 that the three trials of normal (all p>0.02) are statistically the same as are all three trials of taurine (all p>0.0999) meaning that shape and scale parameters are not different within each treatment.

From Table 1 and 2 there is no statistically significant difference between normal and taurine fed flies within each trial (all p>0.0041). However, the small p-value for normal versus taurine trial 2 (scale p = 0.0048, shape p = 0.0041) suggests there is some evidence supporting a difference between taurine and normal fed flies.

Fig. 2:

Graph used to determine goodness of fit for the three distributions, Weibull (thick line), lognormal (solid, thin line) and logistic (dashed line) for the first trial of taurine supplemented flies in the normal versus taurine mixed experiment. The Weibull does the best at both the beginning and the end of the time period. The logistic curve underestimates at the early stages and the lognormal underestimates the distribution in the middle where most of the events occur and it overestimates at the end stages. Therefore, the Weibull distribution was used for all analysis

Fig. 3:

Fitted Weibull survivorship curves of random mixture of male and female fruit flies fed normal cornmeal-molasses food and flies fed food supplemented with 100 mM taurine. Solid lines represent trial 1, dashed lines represent trial 2 and the dotted lines correspond to trial 3. Normal fed trials are on the left (black lines) of the corresponding taurine fed trials (red lines). There were three independent trials for a total of approximately 300 flies per treatment. Initially, both taurine and normal fed flies followed the usual decline in the number of flies over time until about day 30. After day 30, the average number of taurine-fed flies is more likely to live as compared to the control as shown by the steeper slope in the normal curve

Table 1:	The p-values of each pairwise comparison between the means of the scale parameter β for the random mixture of flies fed normal and taurine food
*p-value of 0.0017 or less is viewed as significant after the Bonferroni correction

Table 2:	The p-values of each pairwise comparison between the means of the shape parameter α for the random mixture of flies fed normal and taurine food
*p-value of 0.0017 or less is viewed as significant after the Bonferroni correction

If not for the Bonferroni correction, normal and taurine trial 1 would also be significant (scale p = 0.0703, shape p = 0.0248). Given the borderline insignificance of the survival curves and the significance of the average median and maximum ages, we believe that taurine may be extending life span, but some confounding factors, such as the concentration, critical period for supplementation and gender, may be masking the effects of taurine.

Life span with a lower concentration of taurine-supplemented food: As noted in our initial experiments recapitulating the work from the Massie lab., with 100 and 250 mM taurine-supplemented food, flies were not surviving long in the 250 mM taurine-supplemented food and were significantly reducing the survival of the progeny. If female flies reduce fecundity due to poor conditions (i.e., toxins in the environment), this can also increase life span. Thus, we wanted to use a lower concentration of taurine that did not dramatically affect fecundity to determine whether it was effective in extending the median and maximum life spans of flies as compared to normal. Following the same feeding regimen as the higher taurine concentration (100 mM) previously used, we repeated the life span experiment using 10 mM taurine (5 vials, 10 flies per vial). At the lower taurine concentration, the median age and maximum age also increased over the normal treatment. The median age for 100 mM taurine fed flies was 35.82 days, 10 mM taurine fed flies is 35.10 days and normal fed flies is 28.80 days. The maximum age for 100 mM taurine fed flies is 81.86 days, 10 mM taurine is 96.50 days and normal is 63.50 days. Both taurine treatments increased both median and maximum age compared to normal. Survival analysis showed no significant difference among all three treatments (ANOVA, p = 0.0970). Again this p-value is borderline for the typical significance level of α = 0.05. Due to the borderline p-value, pairwise comparisons were still conducted (Table 3). It appears there is slight evidence that normal fed flies and 10 mM taurine fed flies have different scale and shape parameters (scale p = 0.0290, shape p = 0.0419).

Table 3:	The p-values of each pairwise comparison between the means of the scale parameter β and shape parameter α for the normal, 10 and 100 mM taurine experiment
*p-value of 0.0083 or less is viewed as significant after the Bonferroni correction

Fig. 4:

Fitted Weibull distributions for each treatment group (normal, 10 and 100 mM taurine concentrations). From left to right, the survival curves are normal food (solid line), 100 mM taurine (dotted line) and 10 mM taurine (dashed line). Flies fed normal food follow the typical Weibull curve and show a faster decline in the number of living flies. The two concentrations of taurine shift the curve to the right increasing the survivorship of flies throughout the life span

This suggests there is a possible difference in the survival curves. A Bonferroni correction of α = 0.05/ (6 comparisons) =0.0083 was used to determine significance. Figure 4 shows the Weibull fitted survival curves of the three treatments, normal, 10 mM taurine and 100 mM taurine. Normal and 100 mM taurine curves look similar in shape, but 100 mM taurine fed flies are shifted to the right, increasing the number of flies surviving throughout the life span. The 10 mM taurine curve takes on a different shape compared to the normal curve. Throughout most of the life span (until about day 60), 10 mM taurine appears to be beneficial to survival, keeping a larger number of flies alive.

Critical period for taurine supplementation: After 30 days, flies fed taurine-supplemented food appeared to have an advantage that allowed the extension of the maximum life span (Fig. 3). Based on this result, we investigated whether a chronic diet of taurine-supplemented food was necessary for the extension of maximum life span or if there is a critical time period (within or after the first four weeks of adulthood) that taurine must be supplemented in the diet to confer the life span extension. As shown in Fig. 3 and 4, chronic taurine-supplemented food (both high and low concentrations) compared to normal food extends the maximum life span of the flies.

Fig. 5:

Fitted Weibull distributions to each treatment group that switches food after 4 weeks (normal-10 mM taurine, 10 mM taurine-normal, normal-100 mM taurine, 100 mM taurine-normal and normal-normal). Solid lines represent normal to high concentration, normal to normal and normal to low concentration (respectively), dashed lines represent low concentration to normal treatment and high concentration to normal food (respectively)

In these experiments, flies were given normal food for the first 4 weeks followed by either 10 mM taurine or 100 mM taurine for the remainder of the life span. The reciprocal experiments beginning with either 10 mM taurine or 100 mM taurine followed by normal food for the remainder of the life span were also conducted.

An overall comparison of normal, normal to 10 mM taurine, normal to 100 mM taurine, 10 mM taurine to normal and 100 mM taurine to normal found no statistical significance among the groups (ANOVA, p = 0.5200). A graphical representation of the Weibull survival curves are shown in Fig. 5. All curves look very similar to normal in both scale and shape.

Life spans of mated male versus female on taurine: We also performed life span studies in which the sex of mated flies were taken into account. Vials of 10 flies consisting of 5 females and 5 males were fed either normal cornmeal-molasses food or food supplemented with 100 mM taurine throughout the life span (total of 100 flies per sex). When the flies were passed and counted, the gender of the dead flies were identified and recorded. The median age of normal males was 32.64 days, taurine males was 55.62 days, normal females was 36.34 days and taurine females was 25.18 days. The maximum age for normal males was 64.67 days, taurine males was 119.00 days, normal females was 74.57 days and taurine females was 63.40 days. Both normal males and females have similar median and maximum ages. The most interesting component is that 100 mM taurine dramatically decreases both median and maximum ages of females while the males benefit from exogenous taurine. Taurine fed females also have a median and maximum age less than that of normal females.

Survival analysis of the life spans of normal and 100 mM taurine fed sexes show that taurine fed sexes are affected differently. Testing the equality of all the survival curves was rejected (ANOVA, p<0.0001). The survival curves of the life spans show that normal male and female curves are similar (scale p = 0.1576, shape p = 0.1581) while the curve for taurine fed males is shifted far to the right compared to normal males (scale p<0.0001, shape p = 0.0091) increasing the likelihood of survival throughout the life span (Fig. 6).

Fig. 6:	Fitted Weibull distributions to each treatment group (male-normal, male-taurine, female-normal, female-taurine). Solid lines represent normal food. The bottom dashed line is the female-taurine group and the top dashed line is the male-taurine group. Throughout the entire life span, all treatment groups respond similarly to the control group

Table 4:	P-values of each pairwise comparison between the means of the scale parameter β and shape parameter α for each sex on normal and 100 mM taurine food
*p-value of 0.0042 or less is viewed as significant after the Bonferroni correction

The p-values were compared to a Bonferroni correction of α = 0.05/ (12 comparisons) = 0.0042. Taurine decreases the survival of females (scale p = 0.0001, shape p = 0.0098) demonstrated by the shift to the left in the curve compared to normal females (Table 4).