Longitudinal assessment of health-span and pre-death morbidity in wild type Drosophila

Abstract

The increase in human life expectancy is accompanied by age-related cognitive and motor disability, thus raising the demand for strategies toward healthy aging. This requires understanding the biology of normal aging and late-life functional phenotypes. Genetic model organisms, such as Drosophila melanogaster, can help identifying evolutionary conserved mechanisms underlying aging. Longitudinal assessment of motor performance of more than 1000 individual flies revealed age-related motor performance decline and specific late-life motor disabilities. This allows defining heath- and ill-span and scoring late-life quality of individual flies. As in mammals, including humans, onset, duration, severity, and progression dynamics of decline are heterogenic and characterized by both, progressive worsening and sudden late-life events. Flies either become increasingly incapacitated by accumulating disability over multiple days prior to death, or they escape disability until few hours prior to death. Both late-life trajectories converge into a terminal stage characterized by stereotypical signs of functional collapse and death within 3 hours. Drosophila can now be used to evaluate life prolonging manipulations in the context of late-life quality. High sugar diet increases lifespan and late-life quality, whereas lifespan prolonging antioxidant supplementation has either no, or negative effects on late-life quality, depending on base diet and gender.

Keywords: Drosophila; aging; demography; diet; invertebrate disability; motor behavior.

Figure 1

Late-life pathophysiology of locomotor behavior. (A) Life history chart for 104 flies individually tested in the startle assay every 6 hours from the age of 60 days until death. Gray bars indicate health-span and colored bars disabilities of different categories (see colored inset, note that flies can also receive multiple scores at one testing time, e.g. leg immobility and climbing deficit). Arrows mark individuals that exhibited hyperactivity prior to death (see also Figure 3). Black and white bars on x-axis indicate day-night cycle. (B) Selective enlargement of the last 60 hours of life of flies 58 to 75. (C) Time enlargement of the last 5 days for all flies. (D) The fraction of the cohort being either fit (light gray bars) or showing climbing impairments (mild/moderate dark gray bars; severe/complete, black bars) plotted over time. The percentages of animals with impaired responsiveness (yellow), reduced moving speed (orange), and paradoxical behavior (purple) are depicted by colored lines. (E) Frequency distribution of the onset of four disability categories for all 104 flies. (F) Ethogram depicts the relative frequencies of transitions (line thicknesses) between fit (grey box), various pre-death impairments (colored boxes), and death (white box). The thickness of the connecting lines depicts the relative frequency of the occurrence of the respective state transition, and the color indicates the previous state. (G) Scores used for quantification of the degree of fitness.

Figure 2

The pathophysiology of motor performance in the startle assay. (A-B) Event-history charts depict jump (A) and flight (B) responses of 104 males tested individually every 6 hours between 60d of age and death (see also Figure 1 A). For each fly and each trial responsiveness is indicated by a dash and irresponsiveness by a blank. Insets show overlays of the impairment history (see Figure 1A-C) and the jump/flight responses for a subset of individuals. Note that impairment span was accompanied by a significant reduction in the numbers of jump or flight responses. Four individual flies (F59, F60, F71, F72) were selected as representative case-reports. Fly 59 exhibited paradoxical behavior 18h prior to death and showed no jumping or flight responses during the last 60h of life. Fly 60 showed no impairment and responded with jumping or flight at almost every single trial. Fly 71 was diagnosed with mild climbing impairment and left femur-tibia joint immobility 18 h prior to death. It performed jump but no flight responses. Fly 72 had a defective prothoracic femur-tibia joint and showed multiple climbing defects already 30 hours prior to death. It failed to respond at every late-life trial.

Figure 3

Characteristics of late-life impairments. (A) Cumulative life quality score during the last three days for each animal decreased significantly with age of death (Pearson’s correlation coefficient = 0.506; p < 0.001). (B) Average fitness scores as obtained from the startle assay reveal a marked decline in flight (green line) and jumping (red line) responses and a moderate decrease in climbing responses with age. Overall performance score (blue, sum of all scores) decreased with age. (C) Cumulative impairment scores (sum of all negative scores for each animal with disabilities) revealed no correlation between degree of impairments and age at death (Pearson’s correlation coefficient = 0.016; p > 0.2). (D) Impairment duration did not correlate with age at death (Pearson’s correlation coefficient = -0.011; p > 0.2). (E) Cohort fitness scores decreased significantly during health-span (we started scoring healthy flies at 60 days of age. Note that the oldest fly that did not show any impairment until death was 76 day old, thus resulting in a 16 days observation period, from 60-to-76 days, see x-axis in E) (Pearson’s correlation coefficient = -0.913; p < 0.001).

Figure 4

Flies die in two different modes: Sudden versus silent death. (A) Life history chart for 30 flies filmed continuously from age 60 days until death. Health-span is indicated by horizontal gray bars and ill-span by black bars. Circles mark animals without impairments but with a short period of hyperactivity during late-life (see below). Black arrows demarks an example animal with multiple days of ill-span. (B) Impairment duration was statistically similar in flies dying younger (black bar) versus older than 70 days (grey bar, students T-test, p > 0.9). (C) Legend for (Ci), which shows the occurrence of different spontaneous behaviors during the last 11 hours of life plotted as cohort averages in time bins of 10 minutes (x- axis). Percent time spent (left y-axis) walking (red), standing on the wall or the bottom of the vial (green line), being immobile or in a supine position (gray). Black bars depict the number of climbing efforts (right y-axis). Horizontal bars depict cohort averages and SEMs for the time of the last feeding (blue) and for entering a permanent supine position (black). (D-E) Same data presentation as in C but for a representative animal with high locomotor activity and no impairments until the last day (D, sudden death) as compared to a representative animal with impairments and low locomotor activity (E, silent death). (F) Principle component analysis revealed two distinct clusters of animals. (G, H) Number of climbing efforts (G) and time spent walking (H) (averages, solid lines; SD, dotted lines) in bins of 10 minutes during the last 11 hours of live for all cluster 1 animals (blue) versus all cluster 2 animals (green). (I) Fitness scores of every illderly fly from the startle assay cohort of 104 flies (see also Figure 1) plotted over age (upper panel). Fitness to impairment state transition can be sudden or gradual as exemplified by the score changes of four individuals (middle and lower panels). (H) The average score of all wellderly versus all illderly flies from the startle assay (see figure 1) plotted over the last 10 days of life follows different trajectories.

Figure 5

Oregon-R and Lausanne-S strains display similar late-life pathophysiology. (A, B) Life history chart of Oregon-R and Lausanne-S male (A) and female (B) flies individually tested in the startle assay daily from the age of 60 days (Oregon-R) and 20 days (Lausanne-S) until death. Gray bars indicate health-span and black bars ill-span. Arrows mark individuals that exhibited hyperactivity prior to death. (C) Survivorship curves of male and female Oregon-R and Lausanne-S flies of A and B. (D) Trajectories of disability are plotted separately for Oregon-R and Lausanne-S males and females as average performance scores during the last 10 days prior to death.

Figure 6

High sugar diet and antioxidants increase lifespan. (A, B) Survivorship curves of males (A) and females (B) of six cohorts of flies, each raised on a different diet (C:P 2/1, blue; C:P 2/1 + curcumin, green; C:P 2/1, + super fruits, red, C:P 8/1, purple, C:P 8/1 + curcumin, black; C:P 8/1 + super fruits, orange). For statistics see Table 2. (C, D) Life history charts of all six male (C) and all six female (D) populations. Horizontal gray bars depict health-span and black bars ill-span. (F) Schematic representation of the different feeding paradigms used.

Figure 7

Dietary interventions can change late-life quality substantially. (A, B) Box plots depict ill-span duration for all 6 male (A) and all 6 female (B) populations raised on different diets. *** indicates statistical significance with p < 0.001, Kruskal Wallis ANOVA with Dunn’s posthoc testing. (C, D) Average disability scores of illderly flies plotted over the last 10 days of life for all 6 male (C) and all 6 female (D) populations. ** indicates statistical significance (p < 0.01, one way repeated measures ANOVA). (E, F) Duration of health-span (light gray bars), mild/moderate ill-span (dark gray bars), and severe/complete ill-span (black bars), as a percentage of lifespan are plotted for each diet and separately for males (E) and females (F). * indicates statistically significant difference (Kruskal Wallis ANOVA with Dunn’s posthoc testing (G, H). The relative percentages of animals without impairments (light gray bars), with mild impairments (medium gray bars) and with severe impairments (dark gray bars) during the last day of life plotted separately for each diet and males (G) and females (H). * indicates statistically significant differences (p < 0.05, Chi-square test). (I) The cumulative startle response score through the last ten days of life (late-life functional signature) plotted for each diet (color coded) and separately for females (front row) and males (back row).