Homer 1a uncouples metabotropic glutamate receptor 5 from postsynaptic effectors
- *Department of Pharmacology and Physiology, University of Rochester Medical Center, 601 Elmwood Avenue, Box 711, Rochester, NY 14642; and
- ‡Department of Neuroscience, Johns Hopkins School of Medicine, 905 Huntarian Building, 725 North Wolfe Street, Baltimore, MD 21205
-
Edited by Roger A. Nicoll, University of California, San Francisco, CA, and approved February 12, 2007 (received for review October 11, 2006)
Abstract
Metabotropic glutamate receptors (mGluRs) and Homer proteins play critical roles in neuronal functions including plasticity, nociception, epilepsy, and drug addiction. Furthermore, Homer proteins regulate mGluR1/5 function by acting as adapters and facilitating coupling to effectors such as the inositol triphosphate receptor. However, although Homer proteins and their interaction with mGluRs have been the subject of intense study, direct measurements of Homer-induced changes in postsynaptic mGluR–effector coupling have not been reported. This question was addressed here by examining glutamatergic excitatory postsynaptic currents (EPSCs) in rat autaptic hippocampal cultures. In most neurons, the group I mGluR agonist (S)-3,5-dihydroxyphenylglycine strongly inhibited the EPSC acutely. This modulation occurred postsynaptically, was mediated primarily by mGluR5, and was inositol triphosphate receptor-dependent. Expression of the dominant negative, immediate early form of Homer, Homer 1a, strongly reduced EPSC modulation, but the W24A mutant of Homer 1a, which cannot bind mGluRs, had no effect. (S)-3,5-dihydroxyphenylglycine-mediated intracellular calcium responses in the processes of Homer 1a-expressing neurons were reduced compared with those in Homer 1a W24A-expressing cells. However, neither the distribution of mGluR5 nor the modulation of somatic calcium channels was altered by Homer 1a expression. These data demonstrate that Homer 1a can reduce mGluR5 coupling to postsynaptic effectors without relying on large changes in the subcellular distribution of the receptor. Thus, alteration of mGluR signaling by changes in Homer protein expression may represent a viable mechanism for fine-tuning synaptic strength in neurons.
Metabotropic glutamate receptors (mGluRs) are class 3 G protein-coupled receptors expressed throughout the brain. Group I mGluRs 1 and 5 couple to the phospholipase C/calcium release cascade and to pathways regulating ion channels and synaptic currents (1, 2). In the brain, group I mGluRs are often expressed postsynaptically, where they regulate NMDA and AMPA receptor function and mediate multiple forms of plasticity (3).
The Homer family of postsynaptic scaffolding proteins was discovered by the regulated expression of the member subsequently termed Homer 1a (4). Levels of Homer 1a and variant Ania-3 are increased after periods of activity or stress, injury, or novel experience (4, 5). Other “long” Homer proteins (1b, 1c, 2, and 3) are expressed constitutively (6–8) and, unlike Homer 1a, form clusters mediated by their long C termini. Homer proteins organize postsynaptic proteins around the active site (7) by binding several targets including group I mGluRs, inositol triphosphate receptors (IP3Rs), Shank, and the TRPC1 cation channel (9–12).
Homer proteins also influence the function of their binding partners. Binding to Homers can alter the function of TRPC1 (12), and group I mGluRs couple more strongly to some effectors (e.g., ion channels) when bound to Homer 1a (13) and more strongly to others, such as IP3Rs (8), when bound to long Homer proteins. Thus, Homer proteins regulate localization of binding partners (13–18) and assemble groups of molecules into putative signaling microdomains (8, 13). However, no study to date has demonstrated the functional consequences of Homer expression on receptor–effector coupling for mGluRs located at the postsynaptic site in mammalian neurons.
In the present study, acute modulation of excitatory postsynaptic currents (EPSCs) by mGluR5 was examined in autaptic hippocampal neurons in the absence and presence of Homer 1a expression to determine its effect on mGluR signaling at the postsynapse, where critical functions of mGluRs are most relevant.
Results
EPSCs in Hippocampal Autapses.
To examine group I mGluR signaling to effectors at the synapse, hippocampal neurons were grown as autapses (19–21) in which single neurons on a substrate “microisland” grow and form synapses onto themselves. Fig. 1 A shows an autaptic hippocampal neuron. When autaptic cells were patch-clamped in the whole-cell configuration, EPSCs were measured by stepping to +40 mV for 2 msec to trigger an action potential volley in the distal, unclamped processes, resulting in synchronized neurotransmitter release and a measurable postsynaptic current (Fig. 1 B, EPSC).
EPSCs in hippocampal autapses. (A) Phase contrast image of a hippocampal neuron grown in autaptic culture for 7 days. (B) Sample EPSC trace illustrating the voltage protocol used to evoke an action potential volley from a holding potential of −80 mV (above), the action current, and the EPSC (as indicated). (C) Sensitivity of the EPSC to 10 μM kynurenic acid. Sample currents before (Con), during (KYN), and after (Wash) application of KYN are shown. (D) Average + SEM inhibition by KYN and 20 μM CNQX (6-cyano-7-nitroquinoxaline-2,3-dione). The number of cells in each group is indicated in parentheses.
Although some apparent inhibitory neurons were observed (excluded from analysis), indicated by a slow inhibitory current (22), most exhibited EPSCs that were sensitive to the nonselective ionotropic glutamate receptor (iGluR) inhibitor kynurenic acid (Fig. 1 C), indicating that these neurons were glutamatergic (23). These EPSCs were inhibited 47 ± 7% by 10 μM kynurenic acid (average ± SEM; n = 6) (Fig. 1 C and D). EPSCs were rapid, reaching peak in <10 msec. The non-NMDA iGluR inhibitor 6-cyano-7-nitroquinoxaline-2,3-dione (20 μM) reduced the EPSC by 67 ± 16% (median inhibition was 81%; n = 10), suggesting that the EPSC was primarily non-NMDA, as expected because currents were recorded at −80 mV, and with 1.2 mM Mg2+ in the bath to inhibit currents through NMDA channels at −80 mV. These data indicate that autaptic hippocampal neurons were predominantly excitatory glutamatergic neurons.
EPSC Modulation by Group I mGluRs.
Application of the group I mGluR-selective agonist (S)-3,5-dihydroxyphenylglycine (DHPG) to autapses for 30–50 sec produced a reversible inhibition of the EPSC (Fig. 2 A). This effect was variable, ranging from undetectable to near total inhibition. Overall, 50 μM DHPG produced a 51 ± 6% inhibition (n = 23) of the EPSC. A dose–response curve (n = 7) demonstrated that the half-maximal DHPG dose was 3.9 μM, consistent with a group I mGluR effect (Fig. 2 B). Because of variability and run-down of the current in some cells, apparent effects less than ≈15% were in fact negligible.
The hippocampal autaptic EPSC is modulated by mGluR5. (A) Sensitivity of the EPSC to 50 μM DHPG. Sample currents before (Control), during (50 μM DHPG), and after (Wash) application of DHPG are shown. (B) Dose–response relationship for DHPG inhibition of the EPSC. The estimated half-maximal effective concentration (EC50) was 3.9 μM.
In addition to the DHPG effect, the mGluR5 selective agonist (RS)-2-chloro-5-hydroxyphenylglycine inhibited the EPSC similarly (54 ± 8%, n = 7; 1 mM). Furthermore, the specific mGluR1 antagonist (S)-(+)-α-amino-4-carboxy-2-methylbenzeneacetic acid (LY367385) did not alter the DHPG effect. In contrast, the specific mGluR5 antagonist 2-methyl-6-(phenylethynyl) pyridine (MPEP) reduced the effect of DHPG [supporting information (SI) Fig. 9 A and B]. This experiment was performed by applying 50 μM DHPG, then DHPG in the presence of 50 μM LY367385 or 30–300 nM MPEP (or both separately), followed by DHPG alone again (if the cells survived). The first DHPG application resulted in a 70 ± 7% inhibition (n = 10). In the presence of LY367385, DHPG inhibition was 61 ± 14% (n = 7), and, with MPEP, DHPG produced only a 25 ± 9% inhibition (n = 8). The final application of DHPG alone inhibited the EPSC by 72 ± 12% (n = 3) (SI Fig. 9). These data demonstrate that the predominant group I mGluR in hippocampal autapses is mGluR5 (23, 24).
mGluR5 Acts Postsynaptically.
To determine whether the EPSC modulation by mGluR5 arose pre- or postsynaptically, the ratio of EPSC amplitudes using a paired-pulse protocol was examined. Evoking two EPSCs with a 50-msec interval produces paired EPSCs (Fig. 3) with the second current ranging in relative size from depression to facilitation. This ratio (P2/P1), influenced by factors including extracellular Mg2+ and Ca2+, and the size of the readily releasable pool of transmitter (25, 26), is a function of transmitter release. Therefore, only a presynaptic modulatory mechanism would normally alter the ratio.
Modulation of the EPSC by DHPG occurs postsynaptically. (A) Sample paired-pulse current traces illustrating the paired-pulse ratio in uninhibited (Con) and DHPG-inhibited currents. (B) Average + SEM P2/P1 ratio before (open bar) and during (filled bar) EPSC inhibition by the indicated drug.
Paired EPSCs were examined before and during DHPG inhibition, and the P2/P1 ratio was examined. Fig. 3 A shows sample currents from such a cell. Although the EPSC was strongly reduced by DHPG, the P2/P1 ratio was unchanged. On average, P2/P1 was 0.82 ± 0.07 and 0.83 ± 0.13 (n = 23) before and during DHPG inhibition, respectively (Fig. 3 B). By contrast, EPSC inhibition by group III mGluRs, which act predominantly presynaptically (2), with 300 μM L-AP4, increased P2/P1 from 0.53 ± 0.08 to 0.75 ± 0.11 (n = 6). Finally, direct block of postsynaptic iGluRs with kynurenic acid left P2/P1 unchanged (0.60 ± 0.12 in control and 0.61 ± 0.10 in kynurenic acid) (Fig. 3 B). These data demonstrate that DHPG-mediated inhibition was at least largely postsynaptic, consistent with previous reports (27).
Homer 1a Uncouples Postsynaptic mGluR5 from the EPSC.
The long Homer proteins (Homer 1b, 1c, 2, and 3) are constitutively expressed in hippocampal neurons (4). Homer 1a, a dominant negative subtype, is expressed at high levels after neuronal insult or stress and less severe activity such as that during novel experience (5). To assess its role in regulating mGluR5 coupling to effectors at the postsynapse, Homer 1a was expressed in autapses by using the Sindbis viral expression system (28–30). Fig. 4 B illustrates a fluorescence image of an autaptic neuron infected with Sindbis encoding GFP and, after an internal ribosome entry sequence (IRES), Homer 1a.
Expression of Homer 1a uncouples mGluR5 from EPSC modulation. (A) Scatter plot illustrating the maximal inhibition by DHPG of the EPSC in each uninfected (Con; filled circles) and Homer 1a-expressing (H1a; open circles; neurons expressing either Homer 1a IRES GFP and GFP IRES Homer 1 were used) autaptic neuron. The mean (heavy black line) and median (gray line) inhibition for each group are also shown. The Inset shows DHPG inhibition in a sample Homer 1a-expressing neuron. (B) Epifluorescence image showing GFP expression in one GFP IRES Homer 1a-expressing neuron. (C) Scatter plot illustrating the maximal inhibition by DHPG of the EPSC in each uninfected (Con; filled squares) and Homer 1a W24A-expressing (W24A; open squares) autaptic neuron. The mean (black lines) and median (gray lines) inhibition for each group are also shown. (D) Epifluorescence image showing GFP expression in one Homer 1a W24A IRES GFP-expressing neuron.
Cells expressing Homer 1a (Homer 1a IRES GFP or GFP IRES Homer 1a) showed reduced EPSC inhibition by DHPG (Fig. 4 A). Uninfected cells were inhibited 52 ± 7% (n = 17, median = 52%) by 50 μM DHPG, whereas those expressing Homer 1a showed 8 ± 9% inhibition (n = 8; median = 18%). By contrast, EPSC inhibition by presynaptic group III mGluRs using the selective agonist L-AP4 was unaltered by Homer 1a expression, as was EPSC modulation by the muscarinic acetylcholine receptor agonist oxotremorine-M (SI Fig. 10). These data demonstrate that Homer 1a expression selectively uncouples mGluR5 from effector targets in the postsynaptic density.
To ensure that the uncoupling observed with Homer 1a was not the result of protein overexpression or the indirect result of Sindbis virus infection, a W24A point mutant of Homer 1a, which cannot bind mGluRs (8, 31, 32), was expressed in hippocampal autapses by using the Sindbis system (Fig. 4 C and D). Homer 1a W24A had no effect on DHPG-mediated EPSC modulation. The EPSC was inhibited 49 ± 12% (n = 6; median = 43%) and 58 ± 8% (n = 8; median = 57%) in uninfected and Homer 1a W24A-expressing cells, respectively. These data demonstrate that Homer 1a dramatically reduces coupling of mGluR5 to postsynaptic targets.
Modulation of the EPSC by mGluR5 Requires IP3R Activation.
The precise mechanism of this EPSC inhibition by mGluR5 is not known. Acute iGluR modulation via group I mGluRs may depend on second messengers in the Gq/phospholipase C pathway (33–35). To better understand the Homer-dependent effects described above, it was important to determine whether the mechanism of modulation required activation of the IP3R, because Homer proteins bind the IP3R and regulate mGluR coupling to it (36). Thus, sensitivity of EPSC modulation by DHPG to low-molecular-weight heparin (HEP), an IP3R inhibitor, was examined. Fig. 5 A shows EPSC traces from a control autaptic neuron (Fig. 5 A Left) and one in which 100 μg/ml HEP was included in the patch pipette (Fig. 5 A Right). EPSC modulation by DHPG was inhibited by HEP, reducing the inhibition from 44 ± 10% (n = 7; median 48%) to 15 ± 7% (n = 7; median 10%) in neurons with intracellular HEP. In each cell, DHPG was applied at least 2 min after achieving whole-cell mode to allow for HEP dialysis into cells. However, inclusion of 3 μM bisindolylmaleimide, a protein kinase C inhibitor, in the pipette did not alter DHPG-mediated EPSC modulation (Fig. 5 B). These data suggest that mGluR5-mediated EPSC inhibition in hippocampal autapses requires activation of the IP3R, but not activation of PKC. Therefore, overexpression of Homer 1a, which can disrupt mGluR5 association with the IP3R, uncouples postsynaptic mGluR5 from EPSC modulation, perhaps by preventing efficient signaling through IP3Rs (37).
Modulation of the EPSC by DHPG requires the IP3R but not PKC. (A) Sample current traces showing DHPG modulation of the autaptic EPSC in control cells (Left) and cells recorded in the presence of 100 μg/ml intracellular low-molecular-weight heparin (Right). (B) Scatter plot illustrating the maximal EPSC inhibition in each cell by 50 μM DHPG in control cells (open circles), cells recorded with intracellular heparin (filled circles), or 3 μM intracellular bisindolylmaleimide (filled squares). The mean (black lines) and median (gray lines) inhibition for each group are also shown.
Homer 1a Expression Reduces the mGluR5-Mediated Rise in [Ca2+]i.
The data above demonstrate that Homer 1a expression reduces IP3R-dependent EPSC modulation by postsynaptic mGluR5. A likely mechanism for this uncoupling is disruption of efficient signaling between mGluR5 and the IP3R (7), both of which bind the EVH1 domain of Homer proteins and indirectly interact in the presence of long Homer proteins. This model predicts that expression of Homer 1a will weaken the mGluR5-dependent rise in intracellular calcium in hippocampal neurons, which express long Homer proteins constitutively. To measure this directly, intracellular calcium levels in the processes of cultured hippocampal neurons expressing either Homer 1a IRES GFP or Homer 1a W24A IRES GFP were monitored with Fura2 during DHPG application (Fig. 6). Fig. 6 A shows baseline-subtracted calcium levels from several regions of interest on resolved processes of infected neurons. The peak of the calcium response was smaller in neurons expressing Homer 1a than in those expressing the W24A point mutant. Neurons expressing the W24A point mutant showed an ≈50% greater peak response to DHPG than those expressing wild-type Homer 1a, and the enhanced response persisted for several seconds after peak. Slowing of the peak of the calcium response was not detected.
Homer 1a reduces mGluR5-mediated rise in intracellular calcium. (A) Homer 1a expression reduces the mGluR5-induced intracellular calcium signal in the processes of hippocampal neurons. Ratio of Fura 2 fluorescence (340:380 nm excitation) time course (average ± SEM) for hippocampal neurons expressing either Homer 1a W24A (gray) or Homer 1a (black). Data were baseline-subtracted from the average of all points 10 sec prior to DHPG application. The data include measurements from 61 points on the processes of six infected neurons and 59 points on the processes of five infected neurons in Homer 1a W24A and Homer 1a-expressing cells, respectively. (B) Peak calcium responses minus baseline in response to 50 μM DHPG and 10 μM oxo-M, as indicated in dSRed-transfected neurons (filled bars) and neurons transfected with dSRed plus Homer 1a (open bars). Calcium measurements were obtained as in A.
Similar experiments were also carried out by using neurons transfected with Homer 1a using electroporation and coexpressing the fluorescent protein dSRed, because GFP can interfere with detection of changes in Fura fluorescence. These data (Fig. 6 B) show a similar impairment of the DHPG-induced calcium signal, consistent with the Sindbis expression data. Furthermore, Homer 1a expression in these same cells failed to reduce calcium signaling in response to oxotremorine-M, a muscarinic acetylcholine receptor agonist, confirming that the reduction in signaling observed with DHPG was not due to a general reduction in IP3R activity. Together, these data are consistent with data obtained from cerebellar Purkinje neurons expressing either Homer 1a or Homer 1b in a previous study (8) and demonstrate that Homer 1a expression selectively disrupts efficient coupling between mGluR5 and IP3Rs in the processes of cultured hippocampal neurons.
Effect of Homer 1a on mGluR5 Distribution.
Expression of short vs. long Homer proteins may alter the subcellular distribution of mGluR5 in central neurons (14, 31, 37, 38), which may explain the data observed here. Expression of Homer 1a may lead to localization of mGluR5 at the soma or other regions away from the postsynaptic site, thus uncoupling the receptor from EPSC modulation. To determine whether such a change in mGluR5 distribution occurred here, the mGluR5 was examined by using immunofluorescence.
Fig. 7 illustrates a field of mGluR5-labeled (Alexa Fluor 555, in red) uninfected neurons (Fig. 7 A) and a neuron exposed to the Sindbis GFP IRES Homer 1a construct (Fig. 7 B). Similar cultures exposed to the fluorescently conjugated secondary antibody without prior exposure to the primary anti-mGluR antibody showed no detectable red fluorescence (data not shown). Most, but not all, neurons appear to be mGluR5-positive, consistent with the observation that most neurons responded to DHPG in patch-clamp experiments (Figs. 2–5). Second, mGluR5 staining was apparent on the soma and processes of both uninfected and Homer 1a-expressing neurons (Fig. 7 B and SI Fig. 11). Indeed, no qualitative change in distribution was detectable between these two groups. Thus, although mGluR5 may be selectively excluded from association with the postsynaptic density in Homer 1a-expressing neurons, the data in Fig. 7 do not support a mechanism in which uncoupling of mGluR5 from postsynaptic targets results from exclusion from the dendritic compartment.
Expression of Homer 1a in hippocampal neurons does not detectably alter the distribution of mGluR5. (A) GFP-fluorescence image of a GFP-infected hippocampal neuron (Left) and AF555 fluorescence indicating mGluR5 expression (Right) from the same field. (B) GFP-fluorescence image of a Homer 1a IRES GFP-infected hippocampal neuron (Left) and AF555 fluorescence indicating mGluR5 expression (Right) from the same field. Both images illustrate mGluR5 expression in the cell body and processes of these representative cells. (C) Control fluorescence images illustrating the lack of signal in the GFP channel in an uninfected, mGluR5-positive cell (Left) and the lack of AF555 fluorescence in a GFP-transfected, mGluR5-negative neuron (Right), confirming the lack of bleed-over across the two channels.
Homer 1a Does Not Affect Coupling to Somatic Calcium Channels.
Recent work has shown clear differences in mGluR5 distribution when various Homer proteins are expressed (14). Data in Fig. 7 suggest that, under the conditions used here, a dramatic rearrangement of mGluR5 is not observed. Coupling of somatic mGluR5 to voltage-dependent calcium channels was therefore examined in 7- to 9-day-old hippocampal cultures. To avoid the technical (space clamp) problems of voltage clamping neurons with processes, current measurements were made from outside-out patches pulled from the soma of hippocampal neurons. Barium (25 mM) was used as the charge carrier to enhance currents through high-voltage activated calcium channels.
Barium currents recorded in patches from uninfected cells were reversibly inhibited 73 ± 7% (n = 4) by 50 μM DHPG (Fig. 8), confirming that mGluR5 resides on the cell soma in the absence of Homer 1a expression. In contrast to mGluR5-mediated modulation of EPSCs, calcium channel modulation by somatic mGluR5 was unaffected by Homer 1a expression. In patches from neurons expressing Homer 1a, barium currents were inhibited 68 ± 16% (n = 5), similar to control, suggesting that a large-scale redistribution of mGluR5 does not occur after Homer 1a expression, at least not during the time scale and within the parameters of this study. Thus, Homer 1a expression can alter the coupling of mGluR5 to certain effectors, such as those at the postsynaptic density, by disrupting efficient signaling domains that are constitutively assembled when long Homer proteins are predominant. This change in coupling need not rely on a large-scale redistribution of mGluR5.
Expression of Homer 1a does not alter DHPG-mediated modulation of currents through somatic calcium channels. (A) Sample current traces from the indicated voltage protocol (Inset) illustrating DHPG modulation of the barium currents from somatic outside-out patches in uninfected (Upper) and Homer 1a-expressing (Lower) hippocampal neurons. (B) Average + SEM inhibition of the somatic, outside-out patch barium current in uninfected and Homer 1a-expressing neurons, as indicated. Inhibitions were calculated from peak currents during a voltage ramp from −80 to +80 mV (see Methods). The number of cells in each group is indicated in parentheses.
Discussion
This study provides the first demonstration of Homer-dependent regulation of postsynaptic mGluR5–effector coupling in mammalian neurons. EPSC modulation in hippocampal autapses was reduced in cells expressing Homer 1a, but not Homer 1a W24A, a mutant that cannot bind mGluRs. Furthermore, the effect of Homer 1a did not appear because of dramatic changes in subcellular distribution of mGluR5. These data suggest that Homer 1a induces a subtle molecular rearrangement to reduce coupling to effectors at the postsynapse, such as the IP3R, but not those at the soma, such as voltage-dependent channels.
Ango et al. (14) examined the effect of Homer expression on mGluR5 distribution in neurons. In that study, mGluR5 was seen only in the soma when expressed alone. Coexpression of Homer 1b/c induced dendritic expression of mGluR5 whereas Homer 1a was associated with global mGluR5 distribution. Thus, effects in the present study may be due to relocation of mGluR5 away from the postsynapse. However, a recent study showed that overexpression of Shank 1B enhanced mGluR coupling to BK channels in hippocampal neurons, although expression of Homer 1b and Shank did not alter dendritic mGluR5 localization (38). Those data, obtained from culture conditions similar to this study, suggest that changes in mGluR5 coupling more likely result from rearrangement of the postsynaptic environment than from movement of mGluR5 from the synaptic site. This may occur in part because of inclusion or exclusion of IP3Rs from the synapse (38). Data in the present study show that Homer 1a can also influence mGluR coupling to postsynaptic iGluRs. The strength of this pathway may also depend on proximity of IP3Rs and the endoplasmic reticulum, because an IP3R-dependent mechanism is likely (33–35, 39), and EPSC modulation in hippocampal autapses appears to require IP3R activation (Fig. 5), although iGluR modulation via group I mGluRs by an apparent direct G protein mechanism has been observed in cortical neurons (40).
Hippocampal neurons constitutively express long Homer proteins (4), which are associated with weakened coupling to membrane calcium channels (13). So it was surprising that Homer 1a did not enhance coupling of somatic mGluR5 to calcium channels. It was also surprising that, in uninfected cells, mGluR5-mediated calcium channel modulation was strong (see Fig. 8). Group I mGluR agonists have been reported ineffective at modulating calcium currents in hippocampal neurons (27). In our hands, acutely isolated hippocampal neurons exhibited no calcium current inhibition by DHPG (data not shown), but modulation was seen in neurons 7–9 days in culture (Fig. 8). It is unclear why we saw strong coupling to calcium channels but they did not (27). Nonetheless, in our hands, mGluR5/calcium channel coupling suggests that, although long Homer proteins are constitutively expressed, they may be restricted to the dendritic compartment and/or the postsynaptic region (37), such that somatic mGluRs may be unassociated with Homer proteins. Thus, overexpression of Homer 1a would not alter this coupling.
Homer proteins and mGluRs are implicated in processes such as synaptic plasticity (41), addiction (42, 43), nociception (44), schizophrenia (45), and epilepsy (46, 47). In each case, regulation of synaptic strength via Homer expression appears to be a critical regulator. The data above show that regulation of postsynaptic mGluR signaling is an important aspect of Homer 1a-dependent regulation of synaptic strength. Furthermore, Homer 1a expression alters mGluR signaling even when large-scale changes in mGluR distribution are not seen, suggesting that this mechanism may be used to fine-tune synaptic strength on a short-term basis, rather than the more protracted time required to redistribute mGluRs between cellular compartments.
Methods
Hippocampal Neuronal Cultures.
Hippocampal autapses were generated similar to that described previously (19). Detailed methods on culturing and transfection are contained in SI Text.
Electrophysiology, Calcium Imaging, and Data Analysis.
Patch-clamp and calcium imaging experiments were performed by using standard methods (see ref. 13) and are described in detail in SI Text.
The bath solution for EPSC recording had 140 mM NaCl, 5.4 mM KCl, 10 mM 4-Hepes, 1.2 mM MgCl2, 1.2 mM CaCl2, and 15 mM glucose. The bath solution for outside-out patches had 140 mM Tris hydroxymethyl aminomethane, 20 mM Hepes, 10 mM glucose, 25 mM BaCl2, and 0.0003 mM tetrodotoxin to pH 7.4 with methanesulfonic acid. The pipette solution for EPSC recording had 130 mM KCl, 1.1 mM EGTA, 10 mM Hepes, 0.1 mM CaCl2, 4 mM MgATP, and 0.1 mM NaGTP (pH 7.2). The pipette solution for barium currents had 120 mM N-methyl-d-glucamine, 20 mM TEA methanesulfonic acid, 11 mM EGTA, 10 mM Hepes, 10 mM sucrose, 1 mM CaCl2, 4 mM MgATP, 0.3 mM Na2GTP, and 14 mM Tris creatine phosphate (pH 7.2).
Immunofluorescence.
Neurons were fixed in 4% paraformaldehyde/4% sucrose PBS, permeabilized in 0.05% Triton X-100, then exposed to rabbit polyclonal anti-mGluR5 (Chemicon International, Temecula, CA) antibody in PBS with 1% BSA for 25 min, washed in PBS, exposed to an Alexa Fluor 555 goat anti-rabbit secondary antibody (Invitrogen, Carlsbad, CA) for 30 min, and washed in PBS before imaging.
Acknowledgments
We thank D. I. Yule and L. E. Wagner (University of Rochester) for help with the calcium measurement experiments, S. Ikeda and H. Chen (National Institute on Alcohol Abuse and Alcoholism) for advice on hippocampal culture techniques, and T. B. Begenisich (University of Rochester) for providing HEP in a pinch. This work was supported by a grant from the Epilepsy Foundation of America (to P.J.K.).
Footnotes
- †To whom correspondence should be addressed. E-mail: paul_kammermeier{at}urmc.rochester.edu
-
Author contributions: P.J.K. and P.F.W. designed research; P.J.K. performed research; P.J.K. and P.F.W. contributed new reagents/analytic tools; P.J.K. analyzed data; and P.J.K. and P.F.W. wrote the paper.
-
The authors declare no conflict of interest.
-
This article is a PNAS Direct Submission.
-
This article contains supporting information online at www.pnas.org/cgi/content/full/0608991104/DC1.
- Abbreviations:
- mGluR,
- metabotropic glutamate receptor;
- iGluR,
- ionotropic glutamate receptor;
- IP3R,
- inositol triphosphate receptor;
- DHPG,
- (S)-3,5-dihydroxyphenylglycine;
- HEP,
- low-molecular-weight heparin;
- EPSC,
- excitatory postsynaptic current;
- IRES,
- internal ribosome entry sequence.
- © 2007 by The National Academy of Sciences of the USA