Defective recruitment of motor proteins to autophagic compartments contributes to autophagic failure in aging

Abstract

Inability to preserve proteostasis with age contributes to the gradual loss of function that characterizes old organisms. Defective autophagy, a component of the proteostasis network for delivery and degradation of intracellular materials in lysosomes, has been described in multiple old organisms, while a robust autophagy response has been linked to longevity. The molecular mechanisms responsible for defective autophagic function with age remain, for the most part, poorly characterized. In this work, we have identified differences between young and old cells in the intracellular trafficking of the vesicular compartments that participate in autophagy. Failure to reposition autophagosomes and lysosomes toward the perinuclear region with age reduces the efficiency of their fusion and the subsequent degradation of the sequestered cargo. Hepatocytes from old mice display lower association of two microtubule-based minus-end-directed motor proteins, the well-characterized dynein, and the less-studied KIFC3, with autophagosomes and lysosomes, respectively. Using genetic approaches to mimic the lower levels of KIFC3 observed in old cells, we confirmed that reduced content of this motor protein in fibroblasts leads to failed lysosomal repositioning and diminished autophagic flux. Our study connects defects in intracellular trafficking with insufficient autophagy in old organisms and identifies motor proteins as a novel target for future interventions aiming at correcting autophagic activity with anti-aging purposes.

Keywords: autophagosomes; autophagy; dynein; intracellular traffic; lysosomes; molecular motors; vesicles.

Figure 1

Reduced basal and inducible autophagy in old mouse cells. Primary fibroblasts derived from 4‐m and 24‐m‐old mice were maintained in either presence of serum (basal autophagy) or absence of serum for 4 hrs (induced autophagy). (a) Representative images of immunofluorescence for LC3. Insets show higher magnification. Bar: 10 μm. (b–c) Quantification of (b) number per cell and (c) fraction of cellular area occupied by LC3‐positive puncta (n = 4 and >25 cells per experiment). (d) Primary fibroblasts from 4‐m and 24‐m‐old mice were transfected with the tandem reporter mCherry‐GFP‐LC3. Top: representative images of single and merged channels. Bottom: quantification of the number of autophagosomes (APG; mCherry+ GFP+ vesicles) and autolysosomes (AUT; mCherry+ GFP‐ vesicles) (left) and percentage of APG and AUT (right) (n = 4 and >25 cells per experiment). (e) Analysis of LC3 flux in cells maintained in presence/absence of serum without additions (−) or in the presence of lysosomal protease inhibitors at saturating concentrations (PI; 20 mm NH4Cl/100 μm leupeptin) for the indicated hours. Top: representative immunoblot. Bottom: quantification of steady‐state LC3‐II levels (left) and LC3‐II flux (right) (n = 4). (f) Quantification of the rate of APG biogenesis in experiments as the one shown in E by the difference in LC3‐II levels at 2 and 4 hr after treatment with PI (n = 4). (g) Comparison of the changes in LC3‐II levels at different times of addition of PI in cells maintained in the presence (+) or absence (−) of serum (n = 4). All values are mean ± SEM. One‐way analysis of variance and Bonferroni post hoc test (for multiple comparisons) were used. Differences between 4 and 24 m are significant for *p < .05, **p < .01 and ***p < .001. Absence of symbols indicates no significative difference

Figure 2

Changes in the intracellular distribution of autophagic compartments in old mouse cells. Primary fibroblasts derived from 4‐m and 24‐m‐old mice were maintained in either presence of serum (basal autophagy) or absence of serum for 4 hrs (induced autophagy). (a) Representative images of immunofluorescence for LC3 (red) and tubulin (green). Nuclei are stained with DAPI (gray). Bottom: Dashed white lines in single red channel indicate the perinuclear region and continuous white lines the cell profile. Bar: 10 μm. (b) Quantification of LC3+ vesicles located in the perinuclear cellular region (IN) from images as the ones shown in a. Values are expressed as the cellular area occupied by LC3+ vesicles (left) or the percentage of LC3+ vesicles located in the perinuclear region in absence of serum (right) (n = 4 and >25 cells per experiment). (c) Merged channel image of the staining of the same cells for LAMP‐2A (red) and DAPI (gray). Bottom shows higher magnification regions. (d) Quantification of the fraction of the total cellular area positive for L2A+ vesicles (left) and the fraction of L2A vesicles located in the perinuclear (IN) cellular region (right) (n = 4 and >25 cells per experiment). (e–h) Cells were maintained in the absence of serum for 4 hr and incubated with LysoTracker Red DND‐99 (e) or Mitotracker (green, g). Dashed white lines indicate the perinuclear region and continuous white lines the cell profile. Insets show perinuclear regions at higher magnification to illustrate the marked differences in lysosomal density in these regions between fibroblast from 4‐m and 24‐m‐old mice. Bar: 10 μm. Quantification of the total cellular fraction positive for LysoTracker+ vesicles (f left) and the fraction of LysoTracker+ vesicles located in the perinuclear (IN) cellular region (f right) (n = 4 and >25 cells per experiment). Same values calculated for Mitotracker+ vesicles (g). Statistical analysis did not reveal significant differences in mitochondrial positioning between both age groups. All values are mean ± SEM. Two‐tailed unpaired Student's t test (for single comparisons) or one‐way analysis of variance and Bonferroni post hoc test (for multiple comparisons) were applied. Differences between 4 m and 24 m are significant for *p < .05, **p < .01 and ***p < .001. Absence of symbols indicates no significative difference

Figure 3

Impact of aging and serum deprivation on the motility properties of autophagic vesicles. (a–c) Analysis of autophagic vesicle motility in primary fibroblasts from 4‐m‐old mice expressing the mCherry‐GFP‐LC3 tandem reporter and maintained in serum supplemented media alone or in the presence of the dynein inhibitor EHNA. Quantification of track mean speed (a), track length (b), and microtubule direction (c), here presented as the net movement per unit time away from (+) or toward (−) the nucleus, with the sign indicating the presumed direction of travel with respect to the microtubule plus or minus‐end of autophagosomes (APG; mCherry+ GFP+ vesicles) and autolysosomes (AUT; mCherry+ GFP‐ vesicles). (d–h) Analysis of vesicular motility in primary fibroblasts derived from 4‐m and 24‐m‐old mice expressing the mCherry‐GFP‐LC3 tandem reporter and maintained in serum supplemented media. Representative cell and sequential frames of GFP‐LC3 at the indicated times of the squared region (length, 19 microns) from 4‐m (d) and 24‐m‐old mice (e). Color scale in the image indicates degree of motility. Left: full field. Right: frames showing example movement of single vesicles (green arrows: current location; red arrows: final location; yellow arrows depict the moment when final and current location coincide (final time)). Quantification of the average of vesicular track length (f), displacement (g), and microtubule direction (h) of autophagic vacuoles (AV; mCherry+ vesicles), autophagosomes (APG; mCherry+ GFP+ vesicles) and autolysosomes (AUT; mCherry+ GFP‐ vesicles). (i–k) Comparison of APG track length (i), displacement (j), and microtubule direction (k) in the same cells but maintained in serum supplemented (+, basal) or depleted (−, induced) media for 4 hr before analysis (n = 4 and >25 cells per experiment). All values are mean ± SEM. One‐way analysis of variance and Bonferroni post hoc test (for multiple comparisons) were used. Differences are significant for *p < .05, **p < .01 and ***p < .001. Absence of symbols indicates no significative difference

Figure 4

Changes with age in motor proteins associated with autophagic compartments. (a) Immunoblot for the indicated proteins in fractions enriched on autophagosomes (APG) and autolysosomes (AUT) isolated from 4‐m‐old mouse livers. 100 μg of homogenates (HOM) and 20 μg of each fraction were loaded per lane. Molecular motor proteins are labeled in red. (b) Immunostaining of isolated LC3‐positive compartments from 4‐m‐old fibroblasts for the indicated motors (red). Yellow arrows indicate colocalization. (c,d) Analysis of the motility of APG isolated from livers of 4‐m and 24‐m‐old mice in an in vitro motility system. Liver vesicles were flowed into a 5‐μl microscopy chamber precoated with Taxol‐stabilized fluorescent microtubules (red), and after binding, these were stained with LC3 antibody (green). (c) Representative fields at the indicated times after addition of ATP. Blue arrowheads indicate examples of moving vesicles, and fuchsia arrowheads indicate examples of nonmoving vesicles. Bar: 10 μm. (d) Percentage of APG (LC3+ vesicles) moving in vitro (n = 1,328 and 1,267 tracked vesicles). (e,f) Immunoblot for the indicated proteins in APG (e) and LYS (f) isolated from fed or 24 hr starved (Stv) 4‐m or 24‐m‐old mouse livers. 100 μg of homogenates (HOM) and 20 μg of each fraction were loaded per lane. Molecular motor proteins are labeled in red. Bottom: Quantification of the levels of the motor proteins. Values are expressed as arbitrary densitometric units (ADU) per microgram of protein (n = 5). All values are mean ± SEM. Two‐tailed unpaired Student's t test (for single comparisons) or one‐way analysis of variance and Bonferroni post hoc test (for multiple comparisons) were applied. Differences are significant for *p < .05, **p < .01, and ***p < .001. Absence of symbols indicates no significative difference

Figure 5

Loss of KIFC3 affects the motility of lysosomal compartment. (a,b) Coimmunostaining for KIFC3 (red) and the lysosomal enzyme Cathepsin D (a) or the endolysosomal marker LAMP1 (b) in primary mouse fibroblasts. Single and merged channels (top) and 3D reconstruction of the boxed regions (bottom). (c) Immunoblot for the indicated proteins of lysates from NIH 3T3 control (Ctr) or knocked‐down for KIFC3 (−). (d,e) Track length (d) and displacement (e) for autophagosomes (APG) (n = 438, 638), autolysosomes (AUT) (n = 472, 596), BSA (n = 1,233, 2,074), or LysoTracker (LysTk) (n = 1,835, 2,367) positive vesicles in the same cells stably expressing the mCherry‐GFP‐LC3 tandem reporter or incubated with Alexa647‐BSA or LysTk (n = 5 independent experiments). (f,g) Track displacement (f) and microtubule direction (g) of BSA (n = 1,887, 1,661) or LysTk (LYS) (n = 2,352, 2,494) positive vesicles in primary fibroblasts isolated from 4‐m or 24‐m‐old mice incubated with Alexa647‐BSA or LysTk. (h) Microtubule direction of BSA and LysTk (LYS) positive vesicles in Ctr or KIFC3 (−) fibroblasts maintained in serum supplemented (+, basal) or serum‐free media (−, induced) for 4 hr and incubated with Alexa647‐BSA or LysTk (n = 2,150, 2,367, 2,269, 1,835, 1,800, 2,017, 1,324, 1,233 from top to bottom). All values are mean±SEM. Two‐tailed unpaired Student's t test (for single comparisons) or one‐way analysis of variance and Bonferroni post hoc test (for multiple comparisons) were applied. Differences are significant for *p < .05, **p < .01 and ***p < .001. Absence of symbols indicates no significative difference

Figure 6

Deficient autophagy upon KIFC3 knockdown. (a–c) Autophagic flux in NIH 3T3 cells control (Ctr) or knocked‐down for KIFC3 (−). Immunoblot for LC3 (a) and p62 (b) in cells maintained in the presence (+, basal) or absence of serum (−, induced) and treated or not (−) with lysosomal protease inhibitors (PI) for the indicated times. (c) Quantification of flux as LC3‐II (left) and p62 (right) degradation after 4 hr. Values are expressed relative to Ctr serum + (n = 4). (d–h) Ultrastructure of Ctr and KIFC3 (−) cells. (d) Representative whole‐cell electron micrographs. Bar: 5 μm. (e) Quantification of the number of vesicles (left), percentage of cytoplasm occupied by vesicles (middle), and average size of vesicles (right) in Ctr and KIFC3 (−) cells by morphometric quantification of electron micrographs. (n > 10 micrographs). (f) Higher magnification of the vacuolar compartments in KIFC3 (−) cells. Bottom shows several vacuoles with apparently connected lumen, suggestive of homotypic fusion events. Bar: 0.5 μm (g) Examples of multivesicular bodies (MVB) (arrows) in Ctr and KIFC3 (−) cells. Fusion events between MVB and the vacuolar compartments are observed at higher frequency in KIFC3 (−) cells. Bar: 0.5 μm. (h) Higher magnification images of mitochondria sequestered inside double‐membrane vesicles in KIFC3 (−) cells. Bar: 0.5 μm. (i) Ctr and KIFC3 (−) cells transfected with mt‐Keima. Images of merged channels or only red field (FL Lyso). Nuclei are stained with DAPI (gray). Boxed areas are shown at higher magnification to show mitochondria present in acid compartments. Bar: 5 μm right: quantification of the fraction of red‐labeled mitochondria (mitophagy index). Values are shown as fold values in Ctr cells (n = 3 and >25 cells per experiment). Two‐tailed unpaired Student's t test (for single comparisons) or one‐way analysis of variance and Bonferroni post hoc test (for multiple comparisons) were applied. Differences are significant for *p < .05, **p < .01 and ***p < .001. Absence of symbols indicates no significative difference