Reduced insulin signaling maintains electrical transmission in a neural circuit in aging flies

Abstract

Lowered insulin/insulin-like growth factor (IGF) signaling (IIS) can extend healthy lifespan in worms, flies, and mice, but it can also have adverse effects (the "insulin paradox"). Chronic, moderately lowered IIS rescues age-related decline in neurotransmission through the Drosophila giant fiber system (GFS), a simple escape response neuronal circuit, by increasing targeting of the gap junctional protein innexin shaking-B to gap junctions (GJs). Endosomal recycling of GJs was also stimulated in cultured human cells when IIS was reduced. Furthermore, increasing the activity of the recycling small guanosine triphosphatases (GTPases) Rab4 or Rab11 was sufficient to maintain GJs upon elevated IIS in cultured human cells and in flies, and to rescue age-related loss of GJs and of GFS function. Lowered IIS thus elevates endosomal recycling of GJs in neurons and other cell types, pointing to a cellular mechanism for therapeutic intervention into aging-related neuronal disorders.

Fig 1. Ubiquitously reduced insulin/insulin-like growth factor signaling (IIS) prevents age-associated decline in transmission through the giant fiber system (GFS).

(A) The GFS, showing insertion sites for recording and stimulating electrodes. The monosynaptic tergotrochanteral muscle (TTM) pathway involves the large electrochemical giant fiber (GF)- tergotrochanteral muscle motor neuron (TTMn) synapse. The electrochemical GF-peripherally synapsing interneurons (PSI), PSI, and chemical (cholinergic) PSI-dorsal longitudinal muscle motor neuron (DLMn) synapses comprise the bisynaptic dorsal longitudinal muscle (DLM) pathway. All neuromuscular synapses are chemical (glutamatergic). (B) Response latencies significantly increased with age when recorded from the TTMs (left) and DLMs (right) (n = 8–9). Representative TTM and DLM traces are shown below. Red bars indicate the time between brain stimulus and muscle response. (C) Reduced IIS (da-GAL4/UAS-InR^dn) prevented age-associated decline in GFS transmission (age x genotype interaction between the control genotypes and da-GAL4/UAS-InR^dn is significant, P < 0.03; n = 6–16; S1 Table). Error bars denote SEM.

Fig 2. Reduced insulin/insulin-like growth factor signaling (IIS) prevents age-associated loss of gap junctions (GJs) in the giant fiber system (GFS).

(A) Insulin receptor (IR) immunolabeling in the giant fiber (GF) interneurons. The 5-day-old GFs were injected with rhodamine-dextran (magenta). Scale bar: 8 μm. (B) Representative confocal images (z-series projections) of thoracic regions enriched in GJs. (Shaking-B protein [SHAK-B] staining): anterior midline area (arrow) and bilateral tracts (arrowheads). The scale bar represents 10 μm for all panels. A and B denote anterior and posterior directions. (C) Quantification of SHAK-B signal intensities (left) and area (right) in the thoracic bilateral tracts (n = 5–6 per genotype per age; age x genotype interaction for integrated density: P = 0.0328). Error bars denote SEM.

Fig 3. Insulin signaling manipulation in the adult nervous system.

(A) Nervous system-specific insulin/insulin-like growth factor signaling (IIS) down-regulation (RU+) prevented age-associated loss of transmission in the tergotrochanteral muscle (TTM) pathway (n = 7–10). (B) TTM response latency deteriorated faster with age in flies over-expressing IIS (RU+) (age x treatment interaction: P = 0.0105; n = 6–9). Both panels: error bars denote SEM.

Fig 4. Insulin signaling regulates giant fiber system (GFS) function during aging, and gap junctional density.

(A) GFS-specific over-expression of InR^dn abolished the age-related response latency decline (tergotrochanteral muscle [TTM] pathway age x genotype interaction: P = 0.0004; n = 6–9). (B) Representative images of thoracic shaking-B protein (SHAK-B) staining in 45-day-old control (A307-GAL4/+) flies (left), and 45-day-old A307-GAL4/UAS-InR^dn flies (right). Scale bar: 15 μm. (C) Quantification of SHAK-B signal intensities in the bilateral tracts of the GFS (n = 4–7). (D) Top: Over-expression of SHAK-B(N+16) prevented functional decline in the GFS with age (interaction P = 0.015; n = 6–14 per genotype/age). Bottom: Representative TTM traces from 45-day-old control (left) and SHAK-B(n+16)-over-expressing (right) flies. Red arrows indicate response latency periods. All panels: error bars denote SEM.

Fig 5. Lowered insulin/insulin-like growth factor signaling (IIS) increases connexin 43 (Cx43) gap junction (GJ) formation in human cells.

(A and B) Confocal pictures of human retinal pigment epithelial (RPE1) cell monolayers upon elevated or reduced IIS, stimulated or not with insulin (1 hour) or insulin receptor (IR)/insulin-like growth factor-1 receptor (IGF1R) dual inhibitor (“IR inhib”), as indicated, and immunostained for Cx43 (green), integrin α3 (ITGA3, red), and DNA (DAPI, blue), images are representative of at least 30 captures from 3 independent experiments). Bar, 10 μm. (C, D, G) Quantification from high-throughput microscopy images of the total levels of Cx43 in RPE1 cells upon elevated or reduced IIS, stimulated or not with insulin, IR/IGF1R dual inhibitor (“IR inhib”), protein kinase C (PKC) activator or inhibitor or lysosomal inhibitors (NH₄Cl or bafilomycin A [BafA]) or transfected with the indicated wild-type (WT), constitutively active (CA) or dominant-negative (DN) Rab constructs, and normalized as indicated. (Data are shown as means ± SEM from 3 independent experiments (over 12,000 Cx43 punctae per condition); n.s., not significant; *P < 0.05, **P < 0.1, ***P < 0.001; 1-way ANOVA and Dunnett test versus “−insulin,” “+insulin,” or enhanced green fluorescent protein [EGFP], as appropriate). (E and H) Percentage of Cx43 colocalizing with ITGA3 (“Surface”), early endosome antigen 1 (EEA1; “Early Endosomes”) or lysosomal-associated membrane protein 1 (Lamp1; “Lysosomes”) in cells treated as indicated (Data are shown as means ± SEM from 3 independent experiments (over 12,000 Cx43 punctae per condition); n.s. = not significant; *P < 0.05, **P < 0.1, ***P < 0.001; 1-way ANOVA and Dunnett test versus “−insulin,” “+insulin,” or EGFP, as appropriate). (F) Transferrin efflux (endosomal recycling) measured by flow cytometry from RPE1 cells upon elevated or reduced IIS, stimulated or not with insulin (1 hour), as indicated. (Data are shown as means ± SEM from 3 independent experiments and normalized to “elevated IIS” at 7.5 minutes. Over 10,000 cells were analyzed per condition and per experiment; n.s., not significant; ***P < 0.001; 1-way ANOVA and Dunnett test.

Fig 6. Over-expression of recycling Rabs rescues age-related loss of gap junctions (GJs) and giant fiber system (GFS) function.

(A and B) Over-expression of Rab4 (wild type [WT]) or Rab11(WT) in the GFS led to increased levels of shaking-B protein (SHAK-B) in the thorax of old flies. Representative confocal images for Rab4(WT) are shown in (A), the quantification in (B) (n = 5–10). (C) Tergotrochanteral muscle pathway (TTM) response latencies from young ([y] days 5–7) and old ([o] days 45–50) flies of various genotypes. WT and constitutively active (CA) construct over-expressed WT or CA forms of Rab4 and Rab11, respectively. Bars with different first letters indicate significant difference (irrespective of the subscript). The letters in the parentheses indicate a lack of significance with the specified bar (n = 4–8). (D) Quantification of the SHAK-B signal intensity in the bilateral tracts of young (7-day-old) flies with silenced Rab4 or Rab11 expression (n = 8–10). (E) Rab11 is indispensable for the effect of reduced signaling on the transmission through the TTM branch of the GF circuit (n = 5–13). Error bars denote SEM.

Fig 7. Model for the age-related effect of reduced insulin/insulin-like growth factor signaling (IIS) on gap junctional density and electrophysiological properties of the giant fiber system (GFS).

Attenuated insulin signaling prevents the loss of shaking-B protein (SHAK-B)-containing gap junctions (GJs) and response latency increase with age by stimulating plasma membrane recycling of GJ subunits via Rab4 and/or Rab11.