Adult (6~8 week) C57BL/6 (B6) mice of each sex were used. All mice were kept on a 12 h light-dark cycle in a specific-pathogen-free facility with controlled temperature and humidity and had free access to food and water. All experimental procedures were conducted according to protocols approved by the directives of the Institutional Animal Care and Use Committee of KIST (Seoul, Republic of Korea, approval number: 2016-051).

Hippocampal slices were prepared as described previously [17]. Transverse slices containing hippocampus were sliced at a thickness of 300 µm using a D.S.K Linear Slicer pro7 (Dosaka EM Co., Ltd, Japan). Slices were left to recover for at least 1 h before recording in an oxygenated (95% O₂ and 5% CO₂) preparation of aCSF containing (in mM): 130 NaCl, 24 NaHCO₃, 3.5 KCl, 1.25 NaH₂PO₄, 1 CaCl₂, 3 MgCl₂, and 10 glucose (pH 7.4) at room temperature. After 1 h, the preparation of aCSF was replaced with an oxygenated recording of aCSF solution (1.5 CaCl₂ and 1.5 MgCl₂ containing aCSF) which was also used when recording was performed.

Field potentials in the CA1 stratum radiatum evoked by Schaffer-collateral stimulation were measured as previously described [1] and responses were quantified in terms of field potential amplitude measurements. Recording was performed using a Multiclamp 700B amplifier (Molecular Devices). Data was acquired and analyzed with pClamp 10.2. Recording electrodes (4~8 MΩ) were filled with NaCl (1 M).

For IOS imaging devices, an infrared (IR) light source with optical filter (775 nm wavelength, Omega Filters) was used for transillumination of brain slices and these optical signals were obtained as IOS images from the stratum radiatum of the hippocampal CA1 region using a microscope (BX50WI, Olympus) equipped with a CCD camera (ORCA-R2, Hamamatsu). Imaging Workbench software (INDEC BioSystems, CA, USA) was used for image acquisition and analyses.

In detail, we first prepared mouse brain hippocampal slices (described in the methods for electrophysiology). Next, we fixed a hippocampal slice into the recording chamber and positioned the electrical stimulator in the CA1 stratum radiatum region. A series of 80 images/s were acquired following electrical stimulation (20 Hz, 1 s, 200~300 •A). The relative change of transmittance (ΔT/T) was normalized to baseline (average of 5 images). Decay of the IOS was measured by averaging the last 10 s of the response after dividing responses with peak response and adding baseline changes.

TTX citrate (tetrodotoxin citrate, Octahydro-12-(hydroxymethyl)-2-imino-5,9:7,10a-dimethano-10aH-[1,3]dioxocino[6,5-d]pyrimidine-4,7,10,11,12-pentol citrate, Tocris, #1069), Cd²⁺ (Cadmium sulfate hydrate, Sigma-Aldrich, #25513), Concanamycin A (Folimycin, 3Z,5E,7R,8R,9S,10S,11R,13E,15E,17S,18R)-18-[(1S,2R,3S)-3 [(2R,4R,5S,6R)-4-[[4-O-(Aminocarbonyl)-2,6-dideoxy-β-D-arabino hexopyranosyl]oxy] tetrahydro-2-hydroxy-5-methyl-6-(1E)-1-propenyl-2H-pyran-2-yl]-2-hydroxy-1-methylbutyl]-9-ethyl-8,10-dihydroxy-3,17-dimethoxy-5,7,11,13-tetramethyloxacyclooctadeca-3,5,13,15-tetraen-2-one, Sigma-Aldrich, #C9705), APV (D-AP5, D-(−)-2-Amino-5-phosphonopentanoic acid, Tocris, #0106), CNQX (6-Cyano-7-nitroquinoxaline-2,3-dione, Tocris, #0190), Bicuculline ((−)-Bicuculline methobromide, Tocris, #0109), CGP-35348 ((3-Aminopropyl)(diethoxymethyl)phosphinic acid, Sigma-Aldrich, #C5851), Strychnine ((−)-Strychnine, Strychnidin-10-one, Sigma-Aldrich, #S0532), Suramin (Suramin sodium salt, Sigma-Aldrich, #S2671), DL-TBOA (DL-threo-beta-Benzyloxyaspartate, Tocris, #1223), Bumetanide (3-(Aminosulfonyl)-5-(butylamino)-4-phenoxybenzoic acid, Sigma-Aldrich, #B3023), Ba²⁺ (Barium chloride dihydrate, Sigma-Aldrich, #529591), Furosemide (5-(Aminosulfonyl)-4-chloro-2-([2-furanylmethyl]amino)benzoic acid, Sigma-Aldrich, #F4381), Ouabain (3-[(6-Deoxy-α-L-mannopyranosyl) oxy]-1,5,11α,14,19-pentahydoxycard-20(22)-enolide, Tocris, #1076), Quinine ((R)-(6-Methoxyquinolin-4-yl) ((2S,4S,8R)-8-vinylquinuclidin-2-yl)methanol, Sigma-Aldrich, #145904), DCPIB (4-(2-butyl-6,7-dichloro-2-cyclopentylindan-1-on-5-yl)oxybutyric acid, Tocris, #1540), NPPB (5-Nitro-2-(3-phenylpropyl-amino)benzoic acid, Sigma-Aldrich, #N4779).

Statistical parameters including the exact value of n, the definition of center, dispersion and precision measures (mean±standard error of the mean (SEM)) and statistical significance are reported in the figures and figure legends. If the values come from a Gaussian distribution, all data points are tested by D'Agostino-Pearson omnibus normality test, and then appropriate statistical methods are applied. In figures, asterisks denote statistical significance as ^*p<0.05; ^**p<0.01; ^***p<0.001; ^****p<0.000, as well as non-significance with NS, p>0.05. Statistical analysis was performed in GraphPad PRISM 7 software.

The neuronal activity-dependent IOS change can be observed with a fast rise (4~6 s) followed by a slow decay (70~80 s) of light transmittance in the CA1 stratum radiatum of hippocampal slices upon electrical stimulation of the Schaffer collateral pathway for 1 s at 20 Hz as previously described [17] (Fig. 1A). Simultaneously, field excitatory post-synaptic potential (fEPSP) was measured to monitor the change in neuronal activity (Fig. 1A). To check the stimulation frequency-dependency of IOS, the IOS was measured with various stimulation frequencies from 1 to 100 Hz for 1 s. The normalized IOS peak was gradually increased in a stimulation frequency-dependent manner with an EC₅₀ at 20 Hz (Fig. 1B and 1C). These results indicate that neuronal activity-dependent light transmittance change was visualized by IOS imaging.

Many previous studies have reported that the change of light scattering in brain slices originates from water movements in astrocytes according to excitatory synaptic transmission, resulting in the light transmittance change of IOS imaging [1,11,12,14,16]. Using various inhibitors of synaptic transmission, we characterized the synaptic nature of the IOS. To check the contribution of presynaptic neurons in the generation of activity-dependent IOS, we used TTX (0.5 µM), a voltage-gated sodium channel blocker, and observed the complete abolishment of IOS peak, fEPSP, and fiber volley (Fig. 2A~E), suggesting that presynaptic action potentials are essential for the generation of activity-dependent IOSs and fEPSPs. To further confirm the involvement of neurotransmitter release in the generation of activity-dependent IOS, we used Cd²⁺ (100 µM), an inhibitor for voltage-gated calcium channels, and Concanamycin A (ConA) (0.5 µM, pre-treatment for 2 h), an inhibitor for vacuolar type H⁺-ATPase in presynaptic vesicles. We found that Cd²⁺ and ConA significantly reduced IOSs and fEPSPs (Fig. 2F~I, and 2K~N), without affecting the fiber volley (Fig. 2J and 2O). These results indicate that neurotransmitter release is also involved in the generation of activity-dependent IOS and fEPSP.

To determine the specific postsynaptic receptors for the generation of IOS, we considered various receptors for glutamate, GABA, glycine, and ATP expressed in CA1 pyramidal neurons as the putative candidates [18,19] (Fig. 3A). We found a significant reduction in IOS peak, abolishment of fEPSP, and intact fiber volley by APV (50 µM) and CNQX (20 µM) (Fig. 3B~E). However, the IOS peak was not changed by inhibitors for GABA_A and GABA_B receptors (bicuculline, 10 µM, and CGP-35348, 5 µM, Fig. 3F and 3G), even in the presence of inhibitors for glutamate receptors (APV and CNQX) (Fig. 3H and 3I). To determine if other neurotransmitters and their receptors are involved in the generation of IOS, we treated the inhibitors for glycine (strychnine, 1 µM) and ATP (suramin, 100 µM) receptors in the presence of APV and CNQX. However, we found no further inhibition of IOS peak by strychnine and suramin (Fig. 3J~M). These data suggest that activity-dependent IOS is mostly generated by excitatory synaptic transmission via postsynaptic glutamate receptors, including AMPA and NMDA receptors, rather than by other transmitters and receptors, such as GABA, glycine, and ATP.

To determine the relative contribution of each step during synaptic transmission, the IOS peak after treatment of various drugs, such as APV and CNQX, TTX, Cd²⁺, and ConA, were normalized by the peak of the control IOS (Fig. 3N). There was 5.8% of the TTX insensitive portion, and a 54.9% contribution of the postsynaptic glutamate receptor as evidenced by block percent of APV and CNQX. The contribution of action potential (AP) firing as 11.7%, vesicle fusion as 27.7% including 23.6% of RRP (readily releasable pool) vesicle fusion and 4.1% of empty vesicle fusion were calculated by taking the difference of block-percentage of Cd2+ from TTX, ConA from Cd²⁺, and APV and CNQX from ConA, respectively (Fig. 3N and 3O).

We confirmed that presynaptic release of glutamate and postsynaptic activation of the glutamate receptor are the main sources of IOSs (Fig. 3O). This finding raised the possibility that the extracellular glutamate and K⁺ resulting from excitatory synaptic transmission could be detected by IOS (Fig. 4A). To test this idea, we treated the glutamate (10 mM) and K⁺ (5 mM) in a 4% agarose gel to mimic the brain slice and avoid the excitotoxicity that occurs in real brain slices (Fig. 4B). It has been reported that electrical stimulation in the hippocampus induces millimolar changes in glutamate [20] and K⁺ [21]. Interestingly, we found that IOS change by treatment of glutamate and K⁺ in agarose gel (Fig. 4C and 4D), suggesting the extracellular ionic concentration changes as a source of IOS.

Astrocytes rapidly take-up glutamate resulting from presynaptic glutamate release during synaptic transmission via glutamate transporters, mostly expressed in astrocytes [22]. Recently, the astrocytic glutamate transporter EAAT2 has been reported to contribute to IOS as evidenced by a significant inhibition of IOS by dihydrokainic acid (DHK), an inhibitor for glutamate transporter EAAT2 [16,23]. However, a long-term treatment with glutamate transporter inhibitors could cause an undesirable side-effects such as desensitization of glutamate receptors such as AMPA, kainate, and NMDA receptors by accumulated glutamate at the synaptic junctions, giving rise to a misinterpretation of the results. Therefore, to test the precise contribution of glutamate uptake in IOS generation, we applied TBOA (100 µM), a non-selective glutamate transporter inhibitor to brain slices and monitored IOS upon repetitive stimulations for a long-period of time. Interestingly, TBOA initially increased the amplitude of IOS upon first stimulation, but then significantly decreased the amplitude upon following repetitive stimulations (Fig. 4E). These two different responses seem to be caused by further activation and desensitization of postsynaptic glutamate receptors including AMPA and NMDA receptors in response to accumulated extracellular glutamate. In contrast, the baseline IOS was increased from 10 min after treatment of TBOA, sustained, and suddenly decreased following stimulations (Fig. 4E). This baseline change might be caused by slower glutamate accumulation and diffusion in the extracellular space. It was difficult to interpret the involvement of glutamate uptake via transporter in IOS generation, which might be due to side effects such as desensitization of glutamate receptors by TBOA. To solve this issue, we administered the TBOA in the presence of APV and CNQX. However, we did not find any significant difference in IOS peak between before and during TBOA applications (Fig. 4F and 4G), indicating that there is no direct involvement of glutamate uptake in IOS generation.

Astrocytes govern extracellular K⁺ uptake, K⁺ buffering, or K⁺ clearance to maintain the extracellular K⁺ concentration. Previously, many K⁺ channels or transporters including Na-K-Cl cotransporter (NKCC1), inward-rectifier K⁺ channel subtype (Kir4.1), K-Cl cotransporter (KCC2), Na⁺/K⁺-ATPase (Na⁺/K⁺ pump), and two-pore-domain K⁺ channels (K2P) have been proposed as the molecular targets for K⁺ buffering in astrocytes [21,24,25,26,27,28]. We tested these various molecular candidates for K⁺ uptake using pharmacological tools (Fig. 5A). We found no significant inhibition of IOS peak by inhibitors for NKCC1 (bumetanide, 10 µM), Kir4.1 (Ba²⁺, 100 µM), KCC2 (Furosemide, 3 mM), and Na⁺/K⁺ pump (ouabain, 100 µM) (Fig. 5B~J). For the last candidate, we considered two-pore domain K⁺ channels, K2P, which have been known to majorly contribute to background leaking K⁺ current, or “passive conductance” in astrocytes [29,30]. To test the involvement of K2P channels for the source of K⁺ uptake, we used quinine, an inhibitor of the K2P channel. Despite decreases in fEPSP by quinine, quinine significantly inhibited the IOS peak (Fig. 5N~P). These results indicate that the K2P channel mediates astrocytic volume change through K⁺ uptake rather than NKCC1, Kir4.1, KCC2, or the Na⁺/K⁺ pump.

Next, we explored the molecular mechanisms of the volume decrease following K⁺ uptake. Many previous studies have reported that the regulatory volume decrease is mediated by the extrusion of anions via volume-regulated anion channel (VRAC) [31,32]. The molecular identity which governs the volume decrease following K⁺ uptake has yet to be established. Therefore, we first investigated the possible involvement of anion channels using the general anion channel blocker, NPPB (50 µM). NPPB increased the baseline and eliminated the decay of IOS, an indication of the inhibition of a volume decrease, upon repetitive stimulations. However, DCPIB, a specific inhibitor for VRAC [33], did not affect these (Fig. 6A and 6B). These results suggest that the volume decrease is mediated by DCPIB-insensitive anion channels.

Taken together, the sources of activity-dependent IOS mostly originate from postsynaptic glutamate receptors, including AMPA and NMDA receptors, but not from GABA, ATP, and Glycine receptors. Our results further reveal that both extracellular glutamate and K⁺ are taken up by astrocytes near the synapse during synaptic transmission. Furthermore, we found that the decay of IOS is mediated by the DCPIB-insensitive anion channels expressed in astrocytes (Fig. 6C). Our results demonstrated that the functional coupling between synaptic activity and astrocytic volume transient in hippocampal slices during excitatory synaptic transmission are the major source of IOS imaging.