Autofluorescence As A Noninvasive Biomarker Of Senescence And Advanced Glycation End Products in Caenorhabditis Elegans

To assess the utility of autofluorescence as a noninvasive biomarker of senescence in Caenorhabditis elegans, we measured the autofluorescence of individual nematodes using spectrofluorimetry. The fluorescence of each worm increased with age. Animals with lower fluorescence intensity exhibited longer life expectancy. When proteins extracted from worms were incubated with sugars, the fluorescence intensity and the concentration of advanced glycation end products (AGEs) increased over time. Ribose enhanced these changes not only in vitro but also in vivo. The glycation blocker rifampicin suppressed this rise in fluorescence. High-resolution mass spectrometry revealed that vitellogenins accumulated in old worms, and glycated vitellogenins emitted six-fold higher fluorescence than naive vitellogenins. The increase in fluorescence with aging originates from glycated substances and therefore could serve as a useful noninvasive biomarker of AGEs. Cistanche can serve as a new model to look for anti-AGE factors and to study the relationship between AGEs and senescence.

Cistanche Herbs For  Anti-AGE Factors

INTRODUCTION

The mean lifespan in humans has been dramatically extended in developed countries during the last half-century. However, progressive damage with aging causes detrimental effects on physiological functions, known as senescence, and senescence-related diseases also are increasing in proportion with longevity. Extension of the “healthspan” is now necessary rather than extending the lifespan alone. Interventions that slow senescence and extend the health span are desired. Brenner et al. introduced Caenorhabditis elegans, which is a small, free-living soil nematode that feeds on bacteria, for use in molecular genetics studies of differentiation and development. 

Subsequently, the worm has been used extensively as an experimental system for biological studies, including research on senescence, because of this animal’s simplicity, transparency, ease of cultivation, short lifespan, and suitability for genetic analysis2. In the worm, senescence can be examined by attenuation of locomotive ability, stress resistance, cognitive ability, and pharyngeal pumping, and by the accumulation of the so-called age pigment lipofuscin3–5 .

A variety of chemicals, including antioxidants and several drugs, are candidates to prevent senescence6,7. Furthermore, since we reported that the health span and lifespan of nematodes could be extended by feeding on probiotic bacteria8,9, the longevity effects of beneficial bacteria have also been demonstrated by a number of laboratories10–12. The longevity effects have been examined by survival curves, but such studies in worms require more than 3 weeks, despite the relatively short lifespan of C. elegans. To evaluate the effects of interventions against senescence, convenient and noninvasive indicators of senescence are desired to trace the influence in individual worms and (ultimately) in humans as the final target.

Lipofuscin, a so-called age pigment, has been extensively used as an indicator of senescence. However, the exact relationship between autofluorescence, senescence, and lifespan in the worm has remained unclear. Gerstbrein et al. reported that age pigments are valid reporters of nematode healthspan13, whereas Coburn et al. reported that blue fluorescence (excitation/emission wavelengths centered on 340/430 nm) serves as a death marker for several hours before and after death in C. elegans14. It has been suggested that the increases in blue autofluorescence over time observed in populations of aging C. elegans might reflect not the aging rate or health state of the population per se, but instead the fraction of dead or almost-dead individuals in the sample15.

Advanced glycation end products (AGEs) have attracted scientific attention16 because they are formed in high amounts in diabetes cases, but also during physiological aging, and lead to oxidative stress. AGEs are complex compounds formed by nonenzymatic reactions between reducing sugars and amino groups in proteins, lipids, or nucleic acids; Schiff bases and Amadori products then are formed through Maillard reactions, yielding AGEs. AGE content is high in patients with diabetes, rheumatoid arthritis, Alzheimer’s disease, and other diseases17–19. In patients with diabetes, AGE accumulation correlates well with the severity of the microvascular disease and rises with age. However, whether AGEs represent a cause of pathogenesis or are only secondary products remains to be elucidated. AGEs are also considered as possible causes of senescence20,21, as suggested from findings in a C. elegans model of sugar toxicity22,23.

A systematic enzyme-linked immunosorbent assay (ELISA) method has been developed by employing antibodies with specificity for several AGE compounds24–26. However, noninvasive methods to permit rough estimations of AGE levels are also required. Since several AGEs are reportedly fluorescent and their associated fluorescence intensity shows a signifificant correlation with the levels measured by ELISA27, we examined whether this fluorescence could be employed as an alternative marker for tracing the status of AGEs and senescence in a C. elegans model.

To verify whether the observed fluorescence merely reflects the death fluorescence reported by Coburn et al., worms were examined by tracing the fluorescence of individual worms while timing the demise of the respective animals. Furthermore, the effects of inhibitors of AGE formation on fluorescence and AGE accumulation were also investigated to validate the utility of C. elegans as a model animal for AGE research.

RESULTS

Autofluorescence of extracted proteins as a biomarker of aging and AGEs in worms

Fluorescence spectrophotometry of extracted proteins indicated that excitation at 325–365 nm made 17-day-old worms fluoresce with the highest emission in the 400–430 nm (peak 420 nm) interval (Fig. 1a). The blue fluorescence was comparatively lower during young adulthood (3-day-old to 5-day-old) but rose over time (in 7-day-old animals) (Fig. 1b). Lipofuscin, a so-called age pigment, emits red fluorescence (excitation at 530–560 nm, emission at 585–645 nm)15; in our previous study, this marker was detected in late-life (more than 13-day-old) worms9. 

Since the auto fluorescence found in the present study appeared at an earlier stage and increased over time, blue auto fluorescence was expected to be a better biomarker for tracing senescence. Since several AGE compounds are known to be auto-fluorescent27, we examined the relationship between AGE contents and blue fluorescence by western blotting in which an antibody to Nε- (carboxymethyl) lysine (CML), a representative AGE, was used. The blotting showed that an outstanding band was amplified over time (Fig. 1c and Supplementary Fig. 1). Indeed, the quantification of the region indicated increased AGE (Fig. 1d), although CML is not auto fluorescent.

To determine if the auto fluorescence indirectly indicated an increase in AGEs formed by the Maillard reaction, proteins extracted from worm homogenates were aseptically incubated for up to 4 weeks with reducing sugars such as glucose, ribose, and fructose. The fluorescence of the proteins rose over time in the presence of sugars, while there was no marked increase in fluorescence in reactions performed without sugars (Fig. 2a). Worm proteins became fluorescent via glycation even in vitro. Among the reactions performed with each of the three sugars, those with ribose produced the highest fluorescence, followed by the two other sugars. This finding could be explained by a report by Bunn and Higgins in which the reactivity of each mono-saccharide with amino groups to form Schiff base linkages is dependent on the extent to which it exists in the open (carbonyl) structure rather than in a ring (hemiacetal or hemiketal) structure, and the rates of reactivity increased in the order of ribose (10.0), fructose (4.5), and glucose (0.6) (×10−3 mM−1 h−1 ), respectively28.

Fig. 1 Fluorescence spectrophotometry and the AGEs of worm protein samples. an Excitation emission matrix (EEM) plots of autofluorescence from lysed worms at 3-day-old and 17-day-olds. The fluorescence levels at the emission wavelength in the EEM plots are displayed as colored heatmaps. One hundred worms (wild-type N2) each of 3-day-old and 17-day-old were lysed in 3 µL of lysis buffer and 4 µL of 15% SDS. Proteins were released by freeze-thaw cycles, and the photoluminescent spectra were collected with a fluorescence spectrophotometer (Hitachi FL-4500). The fluorescence properties of lysed nematodes were analyzed using EEM fluorescence spectroscopy. Excitation at 325–365 nm made 17-day-old worms fluoresce with the highest emissions in the 400–430 nm interval corresponding trajectories as a series of fluorescence intensities, whereas 3-day-old young worms did not fluoresce. Multivariate analysis (single-wavelength excitation with multiple-wavelength emission, and synchronous-scanning fluorometry) yielded EEM plots consisting of single-scan excitation. b Levels of auto fluorescence (ex 340/em 360–600) by worm protein samples extracted from animals of each age group (3–13 days old). Worms (wild-type N2) were collected in tubes, washed five times, and lysed in each tube. Samples were then ground using a Mini Cordless Grinder (Funakoshi, Tokyo, Japan) to release protein. Protein contents were quantitatively measured as described in the Methods. A fluorescence spectrum was determined for each 30-µL sample (containing 1.5 μg of total protein) using a multimode grating microplate reader model SH-9000Lab (Corona Electric, Ibaraki, Japan). Each measurement was carried out three times. c Western blotting analysis of the AGE CML in samples extracted from worm (wild-type N2) populations of different ages. A representative photograph of three reproducible experiments is shown. Blots were derived from different parts of the same gel (shown in Supplementary Fig. 1) and were processed in parallel. d Quantification of AGEs in western blots using densitometry. The bands surrounded by rectangular frames in Fig. 1c were quantified, and data from three reproducible experiments is shown as mean ± SE.

respectively.

To directly assess if the increases in fluorescence in the presence of sugars were due to AGE formation, ELISA was used to detect the representative AGE compound CML. CML increased in these reactions, and the level in samples incubated with ribose was up to nearly seven-fold higher than those in controls incubated without sugars for 4 weeks (Fig. 2b). Compared to ribose, the amounts of CML produced with glucose or fructose were low, similar to CML in the control; this finding is concordant with a previous report29.

Autofluorescence as a biomarker of aging and AGEs in vivo 

We next tested whether the fluorescence could be detected not only from extracted proteins but also in living worms. When the autofluorescence of individual worms was read on plastic wrap film stretched on 384-well plates with a multimode grating microplate reader, the spectrophotometry results were similar to those of extracted proteins (Fig. 3a). The individual intensity of blue fluorescence increased over time (Fig. 3b and Supplementary Fig. 2), while no red fluorescence was detected with this method. Since blue fluorescence (340/430 or 350/460 nm) has been previously reported to be associated with death14,15, the analysis was performed by excluding data obtained within 2 days prior to an animal’s death. However, the blue fluorescence of the worms still increased over time: the increase of fluorescence must therefore be partially independent of the death fluorescence. To test if the autofluorescence could be a biomarker of senescence, we examined how many days the worms could survive after the autofluorescence was measured on 13-day-old worms. The life expectancy of each worm was inversely related to the intensity of autofluorescence (Fig. 3c).

The fluorescence intensity was also elevated in worms grown with ribose (Fig. 4a); the lifespan of these worms was signifificantly shorter than that of control worms (Fig. 4b). In contrast, worms grown on medium containing rifampicin, an AGEs blocker, emitted less fluorescence compared to control worms grown without rifampicin (Fig. 4c); the pro-longevity effect of rifampicin has been previously reported30. Furthermore, daf-2, a representative pro-longevity mutant, also auto-fluoresced less compared to the wild-type (Fig. 4d), while the fluorescence intensity of the wild-type increased with aging.

A portion of the fluorescence is expected to be derived from compounds produced in association with animal death. The blue fluorescence could be the death marker that Coburn et al. observed in C. elegans for several hours before and after death; the blue light would come from anthranilic acid that is formed immediately after worms die14. Reportedly, death fluorescence is emitted by the glycosylated form of anthranilic acid produced by the kynurenine pathway14. kyu-1 mutant worms, which lack the kynureninase, were accordingly expected not to emit the death fluorescence. However, even in the presence of this mutation, aged worms still appeared to emit higher fluorescence than younger worms (Fig. 5a), and worms grown on nematode growth medium (NGM) supplemented with ribose exhibited signifificantly higher fluorescence than those grown in the absence of supplemental ribose or with the control sugar sorbitol (Fig. 5b). The lifespan of the worms maintained with ribose was obviously shorter than that of the others (Fig. 5c), suggesting pro-aging effects of AGEs in worms.

Identification of the autofluorescent materials To identify substances that fluoresced in aging animals, an LCMALDI assay was performed on the extracted proteins. The six most-abundant proteins found in each sample are listed in Table 1; notably, vitellogenins and elongation factors were enriched among the top-45fluorescing proteins in older worms. Furthermore, using high-resolution MS and the Mascot computational environment, we searched for proteins that were more abundant in 15-day-old worms than in 3-day-old worms and identified 180 proteins (Supplementary Tables 1–3). Although vitellogenins again were enriched (present at six-fold higher levels in older worms than in young adults), the levels of elongation factors were similar in the two groups(Supplementary Fig. 3). To verify whether these proteins could be glycated to form autofluorescent AGEs, purified elongation factor and vitellogenin were incubated with ribose in vitro. The fluorescence emitted from glycated vitellogenins exhibited spectra similar to those obtained with crude extracts recovered from the 17-day-old worms (Fig. 6a). Although riboflavin is known to fluoresce, we found that the emission wavelength of riboflavin was distinct from that of worm extracts. The FL fluorescence of glycated vitellogenin and elongation factor were six- and three-fold (respectively) those of the non-glycated proteins by 23 days in vitro (Fig. 6b).

The AGEs detected by western blotting with anti-CML antibodies also increased over time from 3-day-old to 13-day-old worms in vivo (Fig. 1c). By simultaneous western blotting with anti-vitellogenin antibodies, we found that the anti-CML signals were the same sized bands as vitellogenins (Fig. 6c). Although the vitellogenins YP170 and YP115 accumulated with age during this period, the intensity of the AGE signals was enhanced higher than the increase in the amount of vitellogenin YP170 (Fig. 6d). Fluorescence microscopy revealed that 13-day-old worms (Fig. 7d–f, columns 2 and 3) emitted signifificantly more intense blue light than 3-day-old young adults (Fig. 7a–c, columns 2 and 3; Fig. 7g); the gonads also looked brighter (Fig. 7d–f, columns 1 and 2). CML is a representative AGE, although CML is not itself ft fluorescent. Pentosidine, another AGE, is known to be fluorescent. However, immunostaining with neither anti-CML nor anti-pentosidine anti-bodies provided prominent differences between old and young worms (data not shown). Anti-CML however showed diffuse spreading across tissues, while anti-pentosidine hardly showed this.