The distinct roles of calcium in rapid control of neuronal glycolysis and the tricarboxylic acid cycle

  1. Carlos Manlio Díaz-García  Is a corresponding author
  2. Dylan J Meyer
  3. Nidhi Nathwani
  4. Mahia Rahman
  5. Juan Ramón Martínez-François
  6. Gary Yellen  Is a corresponding author
  1. Department of Neurobiology, Harvard Medical School, United States
6 figures, 1 table and 1 additional file

Figures

Figure 1 with 2 supplements
Pathways for core energy metabolism and the possible influences of calcium.

Glucose (Glc) and lactate (Lac) are possible fuel molecules, and both the ATP/ADP and NADH/NAD+ pairs are compartmentalized between cytosol and mitochondria, as is [Ca2+]. For mitochondria, the …

Figure 1—figure supplement 1
Ca2+ regulation of the mitochondrial NADH shuttles.

(a) Malate-aspartate shuttle (MAS); (b) the glycerol phosphate shuttle (GPS). References in the main text.

Figure 1—figure supplement 2
How ion influx influences cytosolic energy state.

Na+ entry through the voltage-activated Na+ channels (NaV) activates the Na+ pump (Na+,K+-ATPase) in the plasma membrane, converting ATP to ADP. Ca2+ entry through the voltage-activated Ca2+

Figure 2 with 6 supplements
Inhibition of the mitochondrial pyruvate carrier reveals the independence of NADH transients in the cytosol and mitochondria.

(a) Left: Representative traces of two different readouts of the Pyronic sensor, expressed in the cytosol of a dentate granule cell (DGC). The lifetime of the donor species mTFP (black line, left …

Figure 2—figure supplement 1
The rise in [pyruvate]CYT due to MPC inhibition almost saturates the Pyronic sensor.

(a) Representative images of DGCs expressing the Pyronic sensor, pseudocolored according to the lifetime of the donor species (mTFP). Cells perfused in ACSF only (Control), or in the presence of the …

Figure 2—figure supplement 2
Inhibition of the mitochondrial pyruvate carrier slows down the TCA cycle and decreases O2 consumption at baseline.

(a) Top: Low-magnification image of an acute brain slice of the hippocampus, showing the placement of the stimulating electrode in the hilus (H) and the O2 electrode in the layer of dentate granule …

Figure 2—figure supplement 3
After MPC inhibition, O2 consumption and FADH2 production are less affected than mitochondrial NADH production in acutely stimulated DGCs.

(a) The traces represent the average normalized NAD(P)H signals (mean ± SD) before and after 2 µM UK5099 (as in Figure 2b), elicited by antidromic stimulation. The baseline before each stimulation …

Figure 2—figure supplement 4
MPC inhibition shortens the NADHCYT transient.

(a) Filmstrip of the metabolic response in the cytosol of stimulated DGCs, before and after the application of 2 µM UK5099 (as in Figure 2c). The images were pseudocolored according to the lifetime …

Figure 2—figure supplement 5
LDH inhibition prevents the effect of UK5099 on the amplitude of the NADHCYT transient, but not on its time course.

(a–b) Effects of the sequential application of 2 µM GSK-2837808A and 2 µM UK5099 (as in Figure 2d) on the parameters derived from the Peredox and RCaMP transients, respectively. The full dataset of F…

Figure 2—figure supplement 6
LDH inhibition elevates the NAD(P)H autofluorescence at baseline but does not restore the NAD(P)H overshoot after the treatment with UK5099.

(a) Representative traces of simultaneously recorded NAD(P)H and FAD+ autofluorescence signals, and RCaMP fluorescence, in a population of DGCs in an acute brain slice. (b) Effects of the sequential …

Figure 3 with 5 supplements
Calcium entrance via MCU is essential for the mitochondrial NAD(P)H overshoot.

(a) Representative traces of the NAD(P)H autofluorescence signals and the mitoRCaMP1h transients in response to different trains of depolarizing pulses. The data correspond to simultaneously …

Figure 3—figure supplement 1
Design of the mitochondrially targeted RCaMP1h.

(a) Top: Schematic representation of mitoRCaMP1h. Abbreviations indicate the mitochondrial leader sequence from the Cox8 subunit (as used in Li et al., 2014), added to the RCaMP1h sensor that …

Figure 3—figure supplement 2
Mitochondrial Ca2+ levels at rest and during stimulation are lower in DGCs from Mcufl/Δ Dock10Cre mice, compared to Mcufl/Δ control mice.

(a) Fluorescence signals from a mitochondrial matrix-targeted Ca2+ sensor (mitoRCaMP) in a population of hemizygous control DGCs (Mcufl/Δ) and MCU-KD DGCs (Mcufl/Δ Dock10-Cre), imaged using …

Figure 3—figure supplement 3
The initial dip of the NAD(P)H signal, the overall FAD+ signal, and the O2 dip in response to stimulation, are less affected than the slow NAD(P)H overshoot in DGCs from Mcufl/Δ Dock10Cre mice.

(a) Average traces of the NAD(P)H signal expressed as the ΔF/Fbaseline-i ratio (mean ± SD of number of slices, sample sizes as in Figure 3b, Right panel). The initial dip of the NAD(P)H signal was …

Figure 3—figure supplement 4
Loss of one Mcu allele slightly reduces the NADHCYT transient.

(a) Comparison of the ΔRCaMP lifetime transient among genotypes (sample size for Mcufl/fl mice: Nneurons = 86, Nslices = 9 and Nmice = 4; sample sizes for Mcufl/Δ and Mcufl/ΔDock10Cre mice as in Figu…

Figure 3—figure supplement 5
Hemizygous mice for MCU show mild impairments in the TCA cycle with respect to Mcufl/fl controls.

(a) Average traces of the NAD(P)H signal expressed as the ΔF/Fbaseline-i ratio (mean ± SD of number of slices, sample sizes for Mcufl/Δ mice as in Figure 3b, Right panel; sample size for Mcufl/fl: Ns…

Figure 4 with 3 supplements
The rise in [Ca2+]CYT, mainly caused by the activity of high-voltage-activated Ca2+ channels, makes a major contribution to the NADHCYT transients in response to stimulation.

(a) Left: Representative trace from a DGC expressing Peredox and RCaMP1h. The slice was superfused for ~20 min with the L-type Ca2+ channel inhibitor isradipine (Isra, 3 µM), and then stimulated. In …

Figure 4—figure supplement 1
A major Ca2+-dependent component of the NADHCYT also occurs under LDH inhibition.

(a) Comparison of the ΔPeredox/ΔRCaMP values obtained in the continuous presence of the LDH inhibitor GSK-2837808A, before and after the blockade of voltage gated Ca2+ channels with a combination of …

Figure 4—figure supplement 2
Calcium removal from the ACSF, without addition of EGTA, is also effective in decreasing the NADHCYT transients.

Left: Representative trace for the effect of Ca2+ removal from the ACSF on the NADHCYT transients. As in Figure 3—figure supplement 5, no EGTA was added to the solution. Right: Simple substitution …

Figure 4—figure supplement 3
DMSO control for the EGTA-AM experiments.

Left: Representative trace of the effect of DMSO (0.1%) on the Peredox and RCaMP signals. Right: The change in Peredox lifetime elicited by stimulation was unaffected by the application of 0.1% DMSO …

Figure 5 with 1 supplement
Inhibition of the Ca2+-calmodulin complex or of AMPK does not abolish the NADHCYT transients in response to stimulation.

(a) Left: Representative trace of a DGC treated with E6-berbamine (10 µM), an inhibitor of the Ca2+-calmodulin signaling pathway. Right: The magnitude of the metabolic response was expressed as the …

Figure 5—figure supplement 1
Ca2+/CaM and AMPK signaling modulates the cytosolic [Ca2+] and NADH/NAD+ ratio at baseline, and the time course of the NADHCYT transient in response to stimulation.

(a) Inhibition of the Ca2+/CaM axis with 10 µM E6-berbamine (E6-Berb) increases the RCaMP lifetime at baseline while keeping the stimulus-associated RCaMP spike intact. The Peredox lifetime at …

Figure 6 with 3 supplements
Neuronal stimulation triggers glycolysis in response to energy demand from ion pumping.

(a) Left: Representative trace of Peredox and RCaMP1h lifetimes simultaneously recorded in a DGC from an acute hippocampal slice. The slice was superfused with 2 µM GSK-2837808A for at least 30 min …

Figure 6—figure supplement 1
NADHCYT transients in 0Ca2+-ACSF with boosted Na+ influx are briefer than those in control ACSF.

(a) The Peredox baseline remained unaltered during the recordings from Figure 6a, where stimulations were delivered ~10 min after changing solutions. The data were compared using a repeated measures …

Figure 6—figure supplement 2
The late overshoot in the NAD(P)H signal disappears in a Ca2+-free solution and is not recovered by α-pompilidotoxin application.

(a) Left: Representative trace of the autofluorescence and RCaMP signals in the DGC layer of an acute hippocampal slice. These experiments were performed with samples from juvenile wild-type mice. …

Figure 6—figure supplement 3
Spontaneous oscillations in the Peredox signal may occur during the prolonged application of zero Ca2+-ACSF, in the presence of EGTA and LDH inhibition.

Prolonged exposure to a nominal Ca2+-free solution, plus LDH inhibition, caused spontaneous elevations of the Peredox lifetime in 25 ± 13% of the cells in each slice (slices = 10, mice = 5), …

Tables

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Chemical compound, drugNBQX (6-Nitro-7-sulfamoylbenzo[f]quinoxaline-2,3-dione, Disodium Salt)Toronto Research ChemicalsCat#N550005; CAS:479347-86-9
Chemical compound, drugD-AP5 (D-(-)−2-Amino-5-phosphonopentanoic acid)AbcamCat#ab120003; CAS:79055-68-8
Chemical compound, drugIsradipineAbcamCat#ab120142; CAS:75695-93-1
Chemical compound, drugα-pompilidotoxinAlomone LabsCat#P-170
Chemical compound, drugEGTA-AMAnaspec IncCat#AS-84100; CAS:99590-86-0
Chemical compound, drugCalmidazoliumCayman ChemicalCat#14442; CAS:57265-65-3
Chemical compound, drugDorsomorphin dihydrochlorideTocris BioscienceCat#3093; CAS:1219168-18-9
Chemical compound, drugE6-berbamineSanta CruzCat#sc-221573; CAS:73885-53-7
Chemical compound, drugUK5099Santa CruzCat#sc-361394; CAS:56396-35-1
Chemical compound, drugUK5099Tocris BioscienceCat#5185
Chemical compound, drugGSK-2837808ATocris BioscienceCat#5189; CAS:1445879-21-9
Chemical compound, drugMgCl2 (1M solution)TeknovaCat#M0304
Chemical compound, drugPicrotoxinSigma-AldrichCat#P1675; CAS:124-87-8
Chemical compound, drugSodium pyruvateSigma-AldrichCat#P8574; CAS:113-24-6
Chemical compound, drugCdCl2Sigma-AldrichCat#202908; CAS:10108-64-2
Chemical compound, drugStrophanthidinSigma-AldrichCat#S6626; CAS:66-28-4
Strain, strain background Mus musculusC57BL/6NCrl miceCharles RiverRRID:IMSR_CRL:27
Strain, strain background Mus musculusEdil3Tg(Sox2-cre)1Amc/J mice (Sox2-Cre)The Jackson LaboratoryRRID:IMSR_JAX:004783
Strain, strain background Mus musculusMcufl/fl miceKwong et al., 2015RRID:IMSR_JAX:029817
Strain, strain background Mus musculusDock10Cre miceKohara et al., 2014Tonegawa lab
Strain, strain background Mus musculusMcufl/Δ miceThis paperObtained by germline deletion of the floxed allele, followed by backcrossing; Yellen lab
OtherAAV9.Syn.RCaMP1h.WPRE.SV40Penn Vector CoreDiscontinuedViral vector
Recombinant DNA reagentAAV.CAG.Peredox.WPRE.SV40Mongeon et al., 2016RRID:Addgene_73807
Recombinant DNA reagentAAV.Syn.RCaMP1h.WPRE.SV40Akerboom et al., 2013
Recombinant DNA reagentpZac2.1-CaMKII-mito-GCaMP6sLi et al., 2014
Recombinant DNA reagentAAV.CAG.Pyronic.WPRE.SV40San Martín et al., 2014Derived from RRID:Addgene_51308
Recombinant DNA reagentAAV.Syn.mito-RCaMP1h.WPRE.SV40This paperProduced by addition of a mitochondrial targeting signal; Yellen lab
Software, algorithmMATLAB R2014bMathworksRRID:SCR_001622
Software, algorithmOrigin 9.1OriginLabRRID:SCR_002815
Software, algorithmGraphPad Instat version 3.06GraphPad SoftwareRRID:SCR_000306
Software, algorithmMicrosoft Excel version 2009MicrosoftRRID:SCR_016137

Additional files

Download links