Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 89 (3), 583-97

IGF-1 Receptor Differentially Regulates Spontaneous and Evoked Transmission via Mitochondria at Hippocampal Synapses

Affiliations

IGF-1 Receptor Differentially Regulates Spontaneous and Evoked Transmission via Mitochondria at Hippocampal Synapses

Neta Gazit et al. Neuron.

Abstract

The insulin-like growth factor-1 receptor (IGF-1R) signaling is a key regulator of lifespan, growth, and development. While reduced IGF-1R signaling delays aging and Alzheimer's disease progression, whether and how it regulates information processing at central synapses remains elusive. Here, we show that presynaptic IGF-1Rs are basally active, regulating synaptic vesicle release and short-term plasticity in excitatory hippocampal neurons. Acute IGF-1R blockade or transient knockdown suppresses spike-evoked synaptic transmission and presynaptic cytosolic Ca(2+) transients, while promoting spontaneous transmission and resting Ca(2+) level. This dual effect on transmitter release is mediated by mitochondria that attenuate Ca(2+) buffering in the absence of spikes and decrease ATP production during spiking activity. We conclude that the mitochondria, activated by IGF-1R signaling, constitute a critical regulator of information processing in hippocampal neurons by maintaining evoked-to-spontaneous transmission ratio, while constraining synaptic facilitation at high frequencies. Excessive IGF-1R tone may contribute to hippocampal hyperactivity associated with Alzheimer's disease.

Figures

Figure 1
Figure 1
Synaptic Localization of IGF-1R (A and B) Hippocampal neurons stained for IGF-1R and synaptophysin (A, presynaptic marker) or Homer1 (B, postsynaptic marker) and imaged in dual-color STED microscopy. The receptor is often found within, or close to, presynaptic terminals and also occasionally near postsynaptic structures (the arrowheads point to a few examples). The scale bar represents 2 μm. (C) A quantification of the distances from IGF-1Rs to the pre or postsynaptic structures. Overall, ∼75% of all receptors are found within the pre or postsynaptic compartments (distances 0–50 nm from the markers). The bars show means ± SEM from three independent experiments with ∼500–3,000 receptor spots analyzed in each experiment.
Figure 2
Figure 2
IGF-1R Activity Is Saturated by Basal IGF-1 in the Vicinity of Hippocampal Synapses (A) An illustration of the IGF-1R construct tagged with Citrin (Cit) or Cerulean (Cer) (top). The representative confocal images of a hippocampal neuron expressing IGF1RCit/Cer are shown (bottom). The white box in (A1) corresponds to the blow-ups in (A2) and (A3). The scale bars represent: A1, 20 μm and A2,3, 2 μm. (B) Pseudocolor-coded fluorescence images of IGF1RCit/Cer before and after Cit photobleaching. The scale bar represents 1 μm. (C) IGF1RCit/IGF1RCer shows basal levels of ETTX (n = 173 boutons) (left). The Em was reduced by AG1024 (1 μM, 20 min, n = 56), while unaffected by the addition of 50 ng/ml hIGF-1 (n = 36). The infection with shIGF1 (n = 79) reduced the Em compared with the shScr sequence infection (n = 62), the addition of hIGF1 rescued the shIGF1-induced Em reduction (n = 77) (right). (D) The AktAR FRET reporter of Akt activity (Gao and Zhang, 2008) tagged with Venus (Ven) and Cerulean (Cer) (top). The representative confocal images of a hippocampal neuron expressing AktAR are shown (bottom). The white box in (D1) corresponds to the blow-ups in (D2) and (D3). The scale bars represent: D1, 20 μm and D2,3, 2 μm. (E) Pseudocolor-coded fluorescence images of a bouton expressing AktAR, displaying donor dequenching after acceptor photobleaching. The scale bar represents 1 μm. (F) AktAR FRET in hippocampal boutons was significantly reduced from Cnt (n = 114) by 1 μM AG1024 (n = 124) and by IGF1R KD (n = 46). The AG1024 was ineffective on AktAR FRET in IGF1-KD neurons (n = 32). (p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001) (ANOVA analysis with post hoc Bonferroni’s multiple comparison tests). The error bars represent SEM.
Figure 3
Figure 3
Pharmacological Blockade and KD of IGF-1R Induces a Decrease in Pr via Inhibition of Presynaptic Ca2+ Flux (A) Representative high-magnification ΔF images before and 15 min after application of 1 μM AG1024. The stimulation protocol during FM dye staining: 30 APs at 1 Hz. The scale bar represents 2 μm. The fluorescence intensities (arbitrary units, a.u.) are coded using pseudocolor. (B) ΔF histograms before (Cnt) and after AG1024 application (AG) in a single experiment. The median ΔF decreased from 99 to 51 a.u. and the number of FM-(+) puncta decreased from 582 to 239. (C) Acute application of AG1024 reduced ΔF, D, and S across synaptic populations in 11 experiments. (D) Representative traces (left) and summary of AG1024 effect (right) on the destaining rate constant (k) during 1 Hz stimulation at hippocampal boutons. AG1024 decreased averaged k from (2.7 ± 0.02) × 10−3 to (1.8 ± 0.03) × 10−3 s−1 (n = 6 for each condition). (E) Inhibition of IGF-IR by PPP (3 hr, 1 μM, and n = 46) or αIR3 (4 hr, 1 μg/ml, and n = 42) decreased ΔF, D, and S. (F) Summary of shIGF-1R effect on presynaptic vesicle recycling (n = 24) compared to non-infected (n = 21), YFP only (n = 15) and shScr (n = 26) infected neurons. The shIGF1R induced a 2-fold reduction in S compared to shScr. The AG1024 did not further reduce S (n = 11). Infection with the WT IGF1R rescues the effect of IGF-1R KD (n = 19). (G) Effect of IGF-1 (50 ng/ml, 20 min) on presynaptic vesicle recycling (normalized to Cnt). There was no significant effect (p > 0.1) of human IGF-1 (hIGF1, n = 31), rat IGF-1 (rIGF1, n = 4), or Des-(1-3)IGF-1 analog (Des-IGF1, n = 16) as shown. (H) Representative traces of the destaining rate constant (k) during 1 Hz stimulation in cultures infected by shScr (n = 4) and shIGF-1 (n = 5). The hIGF-1 (n = 6, 50 ng/ml, and 20 min preincubation) rescued vesicle exocytosis in shIGF-1 neurons. (I) Illustration of the positive Pr regulation by basal IGF-1R activity in Pr regulation (p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001). The error bars represent SEM.
Figure 4
Figure 4
IGF-1R Blockade Induces a Reduction in AMPAR-Mediated EPSC in CA3-CA1 Hippocampal Connections (A) Experimental setup and CA3-CA1 organization. (B) Time course of AG1024 effect on fEPSP amplitude. (C) Representative fEPSP recordings before (gray) and 60 min after (green) application of AG1024 under low-frequency stimulation (0.1 Hz) (top). The scale bars represent 0.1 mV and 10 ms. The AG1024 reduced the input-output relationship between the intensity of the fiber volley amplitude and the fEPSP slope (n = 19) (bottom). (D) Representative EPSC recordings (holding potential −70 mV) before (gray) and 60 min after (green) application of AG1024 under low-frequency stimulation (0.1 Hz) (left). The summary of the effect of AG1024 on EPSC amplitude (student paired two-tailed t test and n = 7) (right) (∗∗p < 0.01 and ∗∗∗p < 0.001). The error bars represent SEM.
Figure 5
Figure 5
IGF-1R Regulates Short-Term Synaptic Facilitation under Physiological and AD-Related Conditions (A) Representative recordings of EPSC evoked by bursts (five stimuli at 50 Hz) in acute hippocampal slices before (gray) and 30 min after (green) application of AG1024. (B) Relative effect of AG1024 on peak amplitude of each EPSC in the burst normalized to the first EPSC amplitude (n = 7). (C) Representative ΔF images for single and burst stimulations in hippocampal cultures under control conditions, following IGF-1R KD and IGF-1R overexpression. The scale bar represents 2 μm. (D) During low-frequency stimulation, IGF-1R KD reduced Ssingle (n = 10), while IGF1R overexpression increased Ssingle (n = 10) compared to control (n = 8). (E) During high-frequency burst stimulation, IGF-1R KD did not alter Sburst (n = 10), whereas IGF1R overexpression increased it (n = 11) compared to control (n = 9). (F) IGF-1R KD (n = 16) increased STP (two-way ANOVA with post hoc Bonferroni tests) in comparison to control (n = 10), while IGF-1R overexpression abolished it (n = 14). (G) Representative fEPSP recordings evoked by single stimuli and bursts in CA3-CA1 connections of acute slices from WT (top) and APP/PS1 (bottom) mice before and after AG1024 application. (H) Higher levels of basal synaptic transmission in APP/PS1 (n = 13) compared to WT (n = 21) slices as indicated by increased I/O slope. The acute application of 1 μM AG1024 (30 min) reduced the I/O slope to the same level in both genotypes. (I) Peak amplitude of each fEPSP in the burst normalized to the first fEPSP amplitude. The slices from APP/PS1 (n = 13) mice exhibit lower levels of STP compared to WT ones (n = 19). The AG1024 increased STP in WT and APP/PS1 mice to the similar level (p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001). The error bars represent SEM.
Figure 6
Figure 6
IGF-1R Blockade Increases mEPSC Frequency (A) Representative traces of mEPSCs for control and following AG1024 application. The scale bars represent 40 pA, 1 s. (B) Cumulative histograms of mEPSC amplitudes in control (n = 21) and following AG1024 application (n = 23). (C) Summary of data in (B). The mean of mEPSC amplitude was unaffected by AG1024. (D) Cumulative histogram of mEPSC interevent intervals showing a shift to smaller values after IGF-1R blockade (the same experiments as in B). (E) Summary of data in (D). The mEPSC frequency is increased 2.5-fold after IGF-1R blockade (∗∗p < 0.01 and non-significant, ns). The error bars represent SEM.
Figure 7
Figure 7
Basal IGF-1R Activity Enhances mEPSC Frequency by Reducing Resting [Ca2+]mito (A) Representative images of OGB-1 AM Ca2+ indicator before and 30 min after AG1024 application (top). The FM4-64 dye staining that labels the functional presynaptic boutons in the same area is shown (bottom). The scale bar represents 2 μM. (B) Time course of AG1024-mediated increase in resting Ca2+ at presynaptic boutons under control (n = 282), FCCP (1 μM, 30 min preincubation, and n = 361), and BAPTA-AM (10 μM, 30 min pre incubation, and n = 198) conditions. (C) Summary of AG1024 effect (% of t = 0) on resting [Ca2+]cyto at presynaptic boutons under TTX in non-treated (NT, n = 707), thapsigargin (Thapsi, 10 μM and n = 219), dantrolene (Dant, 10 μM and n = 137), and FCCP (1 μM and n = 904) treated cultures. In all the conditions, the blockers were incubated for 30 min prior to AG1024 application. (D) Representative confocal images of a hippocampal neuron expressing mCherry-mito (left), 2mtGCaMP6m (middle), and the merged image of both markers showing overlapping expression (right). The scale bar represents 10 μm. (E) Time course of resting [Ca2+]mito following application of AG1024 (n = 53). (F) Summary of AG1024 (n = 58), FCCP (n = 62), or FCCP+AG (n = 39) effects (30 min incubation) on resting [Ca2+]mito. (G) Summary of mEPSC frequency and amplitude under Cnt (n = 14), FCCP (n = 14), and FCCP+AG (n = 10) conditions. (H) Representative confocal images of boutons of a hippocampal neuron infected with SypI-A.Team1.03. The scale bar represents 2 μm. (I) Pseudocolor-coded fluorescence images of mseCFP and cpmVen showing increased cpmVen fluorescence following mseCFP photobleaching. The scale bar represents 2 μm. (J) ETTX in Cnt (n = 292) and following application of AG1024 (n = 210) or FCCP (n = 287). (K) ETTX in Cnt (n = 180) and following application of oligomycin (n = 154) or dGlu (n = 48) (∗∗p < 0.01 and ∗∗∗p < 0.001) (non-significant, ns). The error bars represent SEM.
Figure 8
Figure 8
Basal IGF-1R Activity Inhibits AP-Evoked [ATP]pre and Slows Ca2+mito Transients (A) Representative high-magnification confocal images of an axon of hippocampal neuron expressing mito-tracker (1) and 2mtGCaMP6m without stimulation (2), following single AP (3), and following burst of 5 APs at 100 Hz (4). The scale bar represents 2 μm. (B) Traces of Ca2+mito transients at different boutons evoked by single AP at 0.1 Hz (top) or by burst of 5 APs at 100 Hz, inter-burst interval 30 s (bottom), quantified as ΔF/F (average of ten traces). (C) Averaged Ca2+mito transients for single (top) or burst (bottom) stimulations under control (n = 89, 80), AG1024 (1 μM and n = 34, 43), oligomycin (1 μg/ml and n = 96, 94), and shMCU (65% KD efficiency and n = 22, 30) conditions. (D) Peak amplitude summary of Ca2+mito transients for the data presented in (C). (E) Effect of AG1024 on Em under spontaneous spiking activity under non-treated conditions (ncnt = 431 and nAG = 165), following preincubation with FCCP (ncnt = 198 and nAG = 70), or oligomycin (ncnt = 191 and nAG = 73). (F) shMCU does not affect Em under spontaneous spiking activity (nshScr = 75, nshMCU = 47, and p = 0.77)
Figure 9
Figure 9
Basal IGF-1R Activity Suppresses AP-Evoked Ca2+cyto Transients and Synaptic Vesicle Recycling by Inhibiting Mitochondrial ATP Synthesis (A) Representative traces of Ca2+ transients evoked by 0.1 Hz stimulation during 500 Hz line scan at boutons and quantified as ΔF/F (average of ten traces) showing the effect of AG1024 under non-treated (NT) and in the presence of FCCP or oligomycin. The scale bars represent 20%, 200 ms. (B) AG1024 reduces the charge transfer of Ca2+cyto transients in non-treated conditions (ncnt = 36 and nAG = 32), but not in the presence of FCCP (ncnt = 28 and nAG = 25) or oligomycin (ncnt = 23 and nAG = 24). (C) Representative color-coded ΔF images showing the effect of AG1024 in FCCP-treated (left) and oligomycin-treated (right) cultures on presynaptic vesicle recycling measured by FM1-43 staining (30 stimuli at 1 Hz). (D) FCCP (n = 25) reduced total S compared to Cnt (n = 18) and prevented AG1024 effect (n = 15) (left). Oligomycin (n = 14) reduced the total S compared to Cnt (n = 23) and prevented AG1024 effect (n = 9) (right) (∗∗p < 0.01 and ∗∗∗p < 0.001). The error bars represent SEM. (E) Time course of the effect of FCCP and oligimycin on fEPSP amplitude before and after application of AG1024 in CA3-CA1 connections in acute slices. The representative fEPSP recordings evoked by 0.1 Hz stimuli in control (gray), following co-application of 1 μM FCCP and 1 μg/ml oligomycin (orange), and after subsequent 1 μM AG1024 application (brown) (insert). The scale bars represent 0.2 mV, 10 ms. (F) FCCP and oligomycin reduced fEPSP amplitude and occluded the effect of AG1024 in CA3-CA1 connections (n = 4). (G) FCCP and oligomycin increased short-term facilitation evoke by bursts (five stimuli at 50 Hz) and occluded the effect of AG1024 in CA3-CA1 connections (n = 4).

Similar articles

See all similar articles

Cited by 32 PubMed Central articles

See all "Cited by" articles

References

    1. Abramov E., Dolev I., Fogel H., Ciccotosto G.D., Ruff E., Slutsky I. Amyloid-beta as a positive endogenous regulator of release probability at hippocampal synapses. Nat. Neurosci. 2009;12:1567–1576. - PubMed
    1. Adams T.E., Epa V.C., Garrett T.P.J., Ward C.W. Structure and function of the type 1 insulin-like growth factor receptor. Cell. Mol. Life Sci. 2000;57:1050–1093. - PubMed
    1. Alnaes E., Rahamimoff R. On the role of mitochondria in transmitter release from motor nerve terminals. J. Physiol. 1975;248:285–306. - PMC - PubMed
    1. Amaducci L., Tesco G. Aging as a major risk for degenerative diseases of the central nervous system. Curr. Opin. Neurol. 1994;7:283–286. - PubMed
    1. Atasoy D., Ertunc M., Moulder K.L., Blackwell J., Chung C., Su J., Kavalali E.T. Spontaneous and evoked glutamate release activates two populations of NMDA receptors with limited overlap. J. Neurosci. 2008;28:10151–10166. - PMC - PubMed

Publication types

MeSH terms

LinkOut - more resources

Feedback