Skip to main content
PMC home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2012 Jun 12;109(25):10071–10076. doi: 10.1073/pnas.1204606109

Loss of functional GABAA receptors in the Alzheimer diseased brain

Agenor Limon 1,1, Jorge Mauricio Reyes-Ruiz 1, Ricardo Miledi 1,1
PMCID: PMC3382476  PMID: 22691495

Abstract

The cholinergic and glutamatergic neurotransmission systems are known to be severely disrupted in Alzheimer's disease (AD). GABAergic neurotransmission, in contrast, is generally thought to be well preserved. Evidence from animal models and human postmortem tissue suggest GABAergic remodeling in the AD brain. Nevertheless, there is no information on changes, if any, in the electrophysiological properties of human native GABA receptors as a consequence of AD. To gain such information, we have microtransplanted cell membranes, isolated from temporal cortices of control and AD brains, into Xenopus oocytes, and recorded the electrophysiological activity of the transplanted GABA receptors. We found an age-dependent reduction of GABA currents in the AD brain. This reduction was larger when the AD membranes were obtained from younger subjects. We also found that GABA currents from AD brains have a faster rate of desensitization than those from non-AD brains. Furthermore, GABA receptors from AD brains were slightly, but significantly, less sensitive to GABA than receptors from non-AD brains. The reduction of GABA currents in AD was associated with reductions of mRNA and protein of the principal GABA receptor subunits normally present in the temporal cortex. Pairwise analysis of the transcripts within control and AD groups and analyses of the proportion of GABA receptor subunits revealed down-regulation of α1 and γ2 subunits in AD. In contrast, the proportions of α2, β1, and γ1 transcripts were up-regulated in the AD brains. Our data support a functional remodeling of GABAergic neurotransmission in the human AD brain.

Keywords: neurodegeneration, synaptic mechanism, gephyrin, glutamate receptor

Alzheimer's disease (AD) is associated with a widespread loss of synapse density and continuous degeneration of cholinergic and glutamatergic pathways (1). Although disruption of excitatory pathways is broadly accepted, inhibitory GABAergic pathways are generally thought to be well preserved in AD (reviewed in refs. 2, 3). GABA receptors (GABAARs) are pentameric complexes formed by combinations of α1–6, β1–3, γ1–2, and δ subunits; the specific combination or stoichiometry of the different GABAAR subunits contributes to their cellular localization, pharmacological profile, and function (4). Early studies of the temporal cortex, an area greatly affected by the neuropathic hallmarks of AD, showed only slight decreases (13–17%) of benzodiazepine binding (2, 5), suggesting a modest decrease of GABAARs in AD. However, unspecific binding of benzodiazepines to voltage-dependent anion channels (6) and to mitochondrial translocator proteins (7) that are increased in AD (8) confound the interpretation of results. Lower levels of GABAAR-subunits α1, α2, α4, δ, and β2 mRNAs in the prefrontal cortex of AD brains (3) and of α1, α5, and β3 mRNAs in the AD hippocampus (2, 9) suggest that some receptors may have an altered functional profile in AD. Our initial experiments, evaluating the feasibility of recording the activity of native receptors from frozen brains (10) and of studying glutamate receptors in AD (11), revealed an unexpected and substantial decrease in the amplitude of ionic responses to GABA (i.e., GABA currents) in temporal cortices of AD brains. This reduction was correlated with diminished expression of all GABA transcripts analyzed (α1, α2, α5, β2, β3, γ2) and with decreased α1 subunit protein (11). However, it remains to be determined how those reductions affected the function of GABAARs in cellular membranes. This lack of information is a critical void, considering the correlation between cortical hyperexcitability, the pathological state of patients with AD (12, 13), and the protective effects of GABAAR agonists against Aβ-induced injury (14). In the present study, we have extended our previous observations and report clear evidence of functional remodeling of native human GABAARs in AD.

Results

Temporal cortices from non-AD cases (i.e., control) and from cases with confirmed diagnosis of AD were provided by the Institute for Brain Aging and Dementia and the University of California, Irvine, Alzheimer's Disease Research Center Tissue Resources (demographic information is shown in Table S1).

GABA Receptor Subunits Expressed in AD and Non-AD Brains.

The expression of the principal GABAAR subunit mRNAs of the human temporal cortex was evaluated by quantitative PCR (qPCR). Additionally, we evaluated Gephyrin, a marker of GABAAR clustering (15), and GFAP and S100B as markers of astrogliosis (16). The expression of α1, α2, β2, β3, α5, γ2, and δ mRNAs was strongly reduced in AD. One pair of samples (C10 and AD4), prepared with a different batch of the kit (iScript; BioRad), had greater total expression than the rest and was excluded from the expression analysis because it artificially increased the difference between the control and the AD groups. After removal of these outliers, there was still a substantial decrease of α1, α2, β2, β3, α5, γ2, and δ mRNAs in AD. The levels of β1, γ1, and ρ1 mRNAs were similar to the control (Fig. 1). The levels of GABAAR subunit mRNAs were not correlated with age, postmortem interval (PMI) or storage time (ST). Despite a strong reduction of γ2 mRNA relative expression in the AD group (mean ± SEM, 1.0 ± 0.4; n = 9) compared with that in the control group (12 ± 5.9; n = 7; P = 0.008, Wilcoxon test), Gephyrin was not significantly different between both groups (0.6 ± 0.1 in AD vs. 1.2 ± 0.4 in control; P = 0.36). Similarly, although GFAP and S100B expression increased in AD, this was not statistically significant (40 ± 15 vs. 21 ± 11 for GFAP in AD and control respectively; P = 0.27; 165 ± 37 vs. 97 ± 28 for S100B; P = 0.19). Western blots were done to determine if the levels of α1, β1, γ1, and γ2 mRNAs reflected the amounts of protein in cellular membranes. In agreement with their mRNAs, α1 and γ2 were reduced in AD membranes (P = 0.034 and P = 0.036; n = 4 controls and n = 5 AD; Fig. 1 B and C) and β1 and γ1 were similar to the control (P = 0.36 and P = 0.38). The resilience of β1 and γ1 indicates a changing pool of GABAAR subunits able to coassemble in AD. When we calculated the contribution of each GABAAR subunit to the 100% of GABAARs able to coassemble together within a single case, we found that α1 and γ2 mRNAs contributed less in AD whereas α2, β1, and γ1 mRNAs were represented more (Fig. 2 A and B). As α2, β1, and γ1 are coded in the same chromosomal cluster (17), we investigated the coregulation among them and other GABAAR subunit mRNAs. Pairwise analyses show 15 significant correlations in the control group and only nine in the AD group (Table S2). In the control group, mRNAs for the most abundant GABAAR subunits in the adult brain (α1, α2, α5, β1, β3, and γ2) exhibited strong linear correlations (correlation coefficients, r, 0.77–0.99), and all except α5 mRNA covaried linearly with Gephyrin (Fig. 2 and Table S2). In the AD group, despite lower transcriptional expression, the correlations between these GABAAR subunit mRNAs were maintained (r = 0.68–0.96). Moreover, the ratio α1/γ2 in AD (0.39 ± 0.06; n = 9), which are components of the majority of GABAARs in vivo (18), was not different from that of the control (0.53 ± 0.07; n = 7; P = 0.34). Nonetheless, we observed important differences in the transcriptional profile of the AD brain. First, γ1 and β1 mRNAs were coregulated in the AD group (r = 0.84; Fig. 2C and Table S2) but not in the control group (r = 0.09). Neither γ1 nor β1 mRNAs were correlated with the expression of any of the astrogliosis markers (r < 0.3 in all cases). Second, all correlations between GABAAR subunits and Gephyrin were lost in the AD group. Taken together, these results indicate a selective decrease of GABAARs in AD, maintenance of receptors with normal α1/γ2 ratios, and a larger contribution of GABAARs containing α2, γ1, and β1. We then explored the functional impact of those changes on the receptors.

Fig. 1.

Fig. 1.

Open in a new tab

Differential expression of GABAAR subunits in AD vs. control brains. (A) Pooled qPCR data for the expression of the main GABAAR subunits present in the human temporal cortex. Relative expression is the 2−ΔCt expressed as a percentage of the reference gene (n = 9 AD and n = 7 control cases; *P < 0.05, **P < 0.01, and ***P < 0.005, Wilcoxon test). Notice the logarithmic ordinate scales. (B and C) Pooled data of the Western blots of the subset of GABAAR subunits shown in B.

Fig. 2.

Fig. 2.

Open in a new tab

GABAAR subunit contribution to the expressional stock of GABAARs in AD. (A) Examples of the contribution of each GABAAR subunit mRNA to the total expression by case. (B) Fractional (as a percentage) contribution of pooled individual GABAAR subunits mRNA in control and AD groups (n = 7 control and n = 9 AD cases). Note the reduction of α1 and γ2 mRNAs and the increment of α2, β2, and γ1 mRNAs in AD (*P < 0.05, **P < 0.01, and ***P < 0.005, Wilcoxon test). (C) Correlations between GABAARs mRNAs in AD and control groups. Each point corresponds to a single case. The coefficient of correlation (r) is shown next to its linear fit. Far left: similar couplings between α1 and γ2 in control and AD despite the lower levels of expression in AD. Center two graphs: emergence of coregulation between γ1 and β1 mRNAs in AD. Far right: coregulation of γ2 and Gephyrin mRNAs is disrupted in AD. (D) The α1/γ2 mRNAs ratio is unchanged in AD.

Decreased GABA Current Amplitudes in Normal Aging and AD.

The microtransplantation method is very convenient for pharmacological and biophysical experiments of human native receptors. However, previous reports mention a large variability in the amplitude of responses among microtransplanted oocytes (10, 11, 19). Therefore, we determined the response distribution of single brains (SI Results and Fig. S1). Although the mean and the median of each distribution were linearly correlated, we chose the median as the “specific” amplitude of the GABA currents of each subject's brain to avoid effects of outliers on the mean. The mean amplitude of specific GABA currents in the AD group was approximately 70% smaller (mean ± SD, 30 ± 19 nA; n = 12 cases; 186 oocytes) than that of the control group (101 ± 76 nA; n = 13 cases; 222 oocytes; P = 0.01, Wilcoxon test; Fig. 3 A and B), confirming previous results (11). We then determined the intersubject variability among control and AD groups. This variability was larger in the control group than in the AD group (Fig. 3). Multiple regression analyses indicated that PMI, ST, and sex made negligible contributions to the variance of the specific GABA current amplitude per subject in both groups. However, in the control group, the GABA current amplitude was negatively correlated with aging (r = −0.63; P = 0.02; Fig. 3C). In AD, no correlation was observed (r = −0.14; P = 0.68). A linear regression model for the amplitude of specific GABA current showed that the diagnosis of AD or lack thereof best explained the difference in amplitude of specific GABA currents (P = 0.006), followed by age (P = 0.02) and by an interaction between age and diagnosis (P = 0.048). Taken together, these factors explained approximately 56% of the variability of GABA current amplitudes in both groups. A consequence of the differential reduction of GABA currents in AD and in normal aging is that the GABA current difference between AD and control decreases as age increases.

Fig. 3.

Fig. 3.

Open in a new tab

Responses to GABA are smaller in AD. (A) Sample currents elicited by 1 mM GABA in oocytes injected with membranes from an AD brain (AD13) or a control brain (C3). (B) Box plot of specific GABA current per case grouped by diagnosis of control or AD. Each point is the specific GABA current of a single case. The AD group (n = 12 cases, 186 oocytes) gave smaller responses than the control group (n = 13 cases, 222 oocytes). Whiskers above and below the boxes indicate the 95th and fifth percentiles. (C) Plot of the specific GABA currents vs. age. Solid line is the linear regression fit to data from the control group and the broken line is for the AD group.

Down-Regulation of GABA Currents Is Similar to That of Glutamate-Currents in AD.

Among ionotropic glutamate receptors, AMPA receptors play a crucial role in fast excitatory synapses and are very important in the processes of learning and memory. The down-regulation of AMPA receptors in AD, particularly of GluR2-containing receptors, is well documented and is now accepted as evidence of synaptic dysfunction in AD (20, 21). How does the reduction of GABA currents compare with that of AMPA-type glutamate currents in each subject? If the density and biophysical properties of the GABAARs in AD were less affected than those of AMPA-type receptors, the ratio of GABA current to AMPA current would be larger in the AD brains than in control brains. To test this hypothesis, we first determined the expression of GluR2 and GluR3 mRNAs in the control and AD groups. As previously observed, GluR2 mRNA was significantly reduced from 7.3 ± 1.6 (n = 7) in the control group to 2.9 ± 0.54 in the AD group (n = 9; P = 0.026, Wilcoxon test) (11). In contrast, GluR3 mRNA was not reduced (P = 0.15, Wilcoxon test). Importantly, the mRNA ratios of GABAAR-subunit γ2 to GluR2 and GluR3 subunits (γ2/GluR2 and γ2/GluR3) were lower in the AD group than in the control group (Fig. 4A), indicating that, in AD, γ2 mRNA is down-regulated even more than GluR2 and GluR3 mRNAs. Neither γ1/GluR3 and γ1/GluR2 were different between groups (P = 0.19, Wilcoxon test). Native glutamate currents desensitize very rapidly and are difficult to observe without allosteric modulators (11). Therefore, we determined the ratio of GABA current/AMPA-type current after potentiating glutamate currents with 10 μM cyclothiazide (CTZ), an allosteric modulator that specifically potentiates AMPA-type receptors (22). The maximal glutamate plus CTZ current was obtained after 80 s preincubation in 10 μM CTZ. The mean glutamate plus CTZ current was approximately 64% smaller in the AD group than in the control group (mean ± SD, 10 ± 9 nA vs. 28 ± 31 nA; P = 0.04, Wilcoxon test), confirming previous results (11). Glutamate plus CTZ currents also were correlated with age (Fig. S2). Only experiments in which both GABA currents and glutamate plus CTZ currents were measured in the same oocyte were used to compute the GABA/AMPA ratios. Interestingly, although the mean ratio of the AD group was smaller (3.8 ± 0.5; range, 1.4–6.7; n = 12 cases; 106 oocytes total) than that of the control group (5.1 ± 1.0; range, 2.1–13.2; n = 11 cases; 119 oocytes total), the difference was not statistically significant (P = 0.38, Wilcoxon test; P = 0.23, nested ANOVA of brains within diagnosis to account for inter- and intrasubject variability). These results indicate that the total activity or number of GABAARs in the temporal cortex of the AD brain is down-regulated to a similar extent as the activity or number of AMPA-type glutamate receptors (Fig. 4).

Fig. 4.

Fig. 4.

Open in a new tab

Comparison between GABA and AMPA-type receptors. (A) The mRNA ratios of the GABAAR-subunit γ2 to the GluR2 and GluR3 subunits (γ2/GluR2 and γ2/GluR3) were lower in the AD group, indicating that in AD the γ2 mRNA is down-regulated more than the GluR2 mRNA (n = 9 ADs, n = 7 controls). (B) Sample currents elicited by 1 mM glutamate plus 10 μM CTZ in oocytes injected with membranes from one control (C3) and one AD (AD3) cases. CTZ was preincubated for 80 s to get maximal response to glutamate. (C) The mean GABA/(Glu plus CTZ) ratio of the AD group, although smaller than the control, was not statistically significant (n = 12 ADs, 106 oocytes; n = 11 controls, 119 oocytes).

GABA Current Rundown and Current–Voltage Relationships in Normal Aging and AD.

Decreases of the mRNA and protein of the most abundant GABAAR subunits indicate that the smaller GABA currents of oocytes transplanted with AD membranes are the result of a reduction in the number of receptors. However, changes of GABA potency or altered receptor properties may also contribute to the production of smaller GABA currents. To determine if GABAARs in AD had enhanced rundown (23), we applied multiple pulses of 1 mM GABA to oocytes injected with control or AD membranes. There was no difference between the rundown of control and AD oocytes (Fig. 5). The sixth response to GABA, at 200 s after the first application, was reduced by 23 ± 3% in the control group (n = 9 cases, 50 oocytes) and by 29 ± 3% in the AD group (n = 7 cases, 35 oocytes; P = 0.07, Wilcoxon test; P = 0.18, nested ANOVA). Furthermore, no changes were found in the voltage dependence or in the equilibrium potential of GABAARs that could account for the smaller GABA currents from AD brains (Fig. 5). GABA currents inverted in polarity at membrane potentials of −28.3 ± 1.3 mV in oocytes transplanted with AD membranes (n = 8 cases, 27 oocytes) and −28.2 ± 1.1 mV in the control ones (n = 7 cases, 26 oocytes), with both values near the predicted equilibrium potential of chloride of −32.5 mV in Xenopus oocytes with an external Cl− concentration of 120.6 mM (Ringer solution in the present study) and an internal Cl− concentration of 33.4 mM (24).

Fig. 5.

Fig. 5.

Open in a new tab

GABA-current rundown and current–voltage relationship. (A) GABA currents elicited by repetitive 1 mM GABA applications to oocytes injected with membranes from cases C12 or AD14. (B) No difference was observed between the rundowns of the AD (n = 7 cases, 35 oocytes) and control (n = 9 cases, 50 oocytes) groups. (C) Voltage-dependence of the GABA current. GABA currents were obtained by digital subtraction of the current elicited during voltage ramp pulses (−60 to 20 mV), before and during perfusion of 30 μM GABA to oocytes injected with AD or control membranes. Representative traces obtained from C9 and AD11. No differences in voltage dependence or inversion potential were observed between AD (n = 8 cases, 27 oocytes) and control (n = 7 cases, 26 oocytes) groups.

Faster Desensitization of GABA Currents from Temporal Cortices of AD Brains.

The increase (10–90%) and decay times (90–20%) of GABA currents were measured during prolonged applications of 1 mM GABA. Fig. 6B shows that GABA currents of oocytes injected with membranes from AD brains appear to activate more slowly (307 ± 33 ms; n = 6 cases; 25 oocytes) than currents from control brains (253 ± 17 ms; n = 8 cases; 46 oocytes), but this difference was not significant (P = 0.33, Wilcoxon test; P = 0.4, nested ANOVA). In contrast, the GABA current decay in oocytes injected with AD membranes was faster (24.2 ± 1.9 s) than that of the controls (30.5 ± 1.7 s; P = 0.028, Wilcoxon test; P = 0.048, nested ANOVA). Neither the increase nor the decay of GABA currents was associated with age, PMI or ST. As the peak of the current was not correlated with its decay, increased desensitization of GABAARs cannot account for the smaller GABA currents generated by membranes from AD brains.

Fig. 6.

Fig. 6.

Open in a new tab

GABA currents desensitize faster in membranes from AD brains. (A) Superimposed 1-mM GABA current traces of oocytes injected with membranes from C12 or AD14. Inset: Same traces showing the activation of GABA currents. (B and C) Box plots of the increase and decay times, from 10% to 90% and from 90% to 20%, respectively, of GABA currents in control (n = 6 cases; 25 oocytes) and AD (n = 8 cases; 46 oocytes) groups.

Changes in GABA Sensitivity of Receptors Transplanted from Temporal Cortices of AD Brains.

GABA concentration/response curves, from 1 μM to 10 mM, were obtained for each of the AD and control cases. Peak currents to 1 mM GABA were measured before and after testing each sequentially increasing concentration of GABA at 5-min intervals to allow for full recovery of the receptors. Fig. 7B shows the log of EC50 for each case. The range of the ECs in the control group (82–119 μM) overlapped with that of the AD group (86–152 μM), and the mean EC50 of the control group was slightly, but statistically significantly, smaller (97 ± 4 μM; n = 9 cases; 33 oocytes) than the EC50 of the AD group (114 ± 6 μM; n = 10 subjects; 36 oocytes; P = 0.014, Wilcoxon test; P = 0.01, nested ANOVA). Hill numbers for both groups were similar: 1.18 ± 0.03 for the control group and 1.2 ± 0.03 for the AD group. The EC50 had no correlation with PMI or ST, and it was not correlated with the amplitude of the GABA current. As membranes from each brain contain a heterogeneous mix of GABAARs, the concentration/response curve reflects the average of different populations of GABAARs. The difference of EC50 between AD and control may indicate that the proportion of GABA receptors with less sensitivity to GABA is larger in AD brains. To test whether those receptors were up-represented in AD, we applied salicylidene salicylhydrazide (SCS), a specific blocker of GABAARs with α2β1γ1 stoichiometry (25). There was no effect on the GABA responses in any of the oocytes tested (Fig. S3), indicating that SCS does not block human GABA receptors of native membranes, or that, if it does, that particular combination was not present in the native control or AD membranes.

Fig. 7.

Fig. 7.

Open in a new tab

GABA receptors in the AD group are less sensitive to GABA. (A) GABA currents elicited by different concentrations of GABA applied to oocytes injected with C10 or AD7 membranes. (B) Box plot of GABA logEC50s (pEC50) for the control and AD groups. These groups were different at the level of P = 0.01 (nested ANOVA) when EC50s were nested within brains and when brains were nested within diagnosis (n = 9 control, n = 10 AD; n = 3–5 oocytes per case).

Discussion

Our previous studies revealed an unexpected decrease in the amplitude of the GABA currents elicited by GABA receptors transplanted from the temporal cortices of AD brains (10, 11). Here, we extend those observations and demonstrate that GABAergic signaling is profoundly altered in the AD brain. Moreover, we found that the GABA current amplitude declines with normal aging (65–90 y) and that the GABA current decrease of AD brains is age-dependent. The loss of GABA currents in normal aging and in AD contrast sharply with the longstanding accepted view that GABA receptors are fairly well spared in senescent (26) and AD brains (27). The decline of GABA currents with aging is consistent with the soma shrinking (28, 29) and marked regressions of dendritic arbors and neuropil complexity observed in cortical areas of aged brains (30). Moreover, Loerch et al. found that the α1, α5, β3, and γ2 subunit mRNAs were among the most down-regulated transcripts in the human senescent brain (31). They also found large differences between the temporal age regulation of GABA-related genes of mice, rhesus, and humans; indeed, the comparison between brain cortex of chimpanzees and humans suggests that the marked severity of changes in neocortical structures with aging may be unique to humans (32). In the case of AD, we observed a reduction of GABA currents throughout all ages. Nevertheless, the difference from the control was greater in younger cases of AD. This result agrees with greater cortical atrophy, hypometabolism (33), and greater impairment of cognitive functions in patients with early onset of AD compared with patients who develop AD at a later age (34). The diminished GABA currents in AD cannot be explained by enhanced rundown of GABAARs, changes in their equilibrium potential, alterations of voltage/current relationship, or increased desensitization of the currents. Although we cannot exclude that changes in the unitary conductance of GABAARs could lead to smaller GABA currents, the reduction of mRNAs for synaptic (α1, α2, β2, β3, and γ2) and extrasynaptic (α5 and δ) GABAARs and the lower amounts of protein for α1 and γ2 subunits in microtransplanted cellular membranes support a decrease in the number of GABAARs as the cause of smaller GABA currents in AD brains. The profound loss of GABAARs in AD is also consistent with a total loss of GABAergic and glutamatergic terminals in perisomatic regions of cortical neurons in contact with neuritic plaques (35). Interestingly, hippocampal areas largely affected by plaques and tangles show reduced immunoreactivity for α1 and γ2 and up-regulation of γ1/3 subunits in the neuropil (9, 36), changes similar to the subunit expression profile in temporal cortex found in our study. Although principal GABAAR subunits were reduced in the AD brain, Gephyrin was similar between AD and control groups, probably as a result of compensatory increments of Gephyrin at advanced stages of AD (37). Interestingly, the strong correlations between Gephyrin and GABAAR subunits mRNAs in the control group were lost in the AD group, suggesting disruptions of Gephyrin-dependent clustering of GABAARs, and encouraging future studies focusing on the structure of GABAergic synapses in AD. The cellular localization, spatial arrangement, and kinetic and pharmacological properties of GABAARs are highly influenced by the stoichiometry of the receptors. The resilience of α2, β1, and γ1 in AD, together with the emergence of coregulation among γ1, β1, and β2, imply changes in the intrinsic properties of the GABAARs that should result in atypical GABA signaling. For instance, macroscopic GABA currents from AD brains exhibit faster desensitization and lower GABA sensitivity, indicating an increment in the number of receptors with those characteristics. In agreement with our results, receptors containing α2 and γ1 instead of α1 and γ2 are less sensitive to GABA (38, 39), and β1 accelerates the desensitization of inhibitory postsynaptic currents in reticular thalamic neurons (40). Additionally, several pharmacological properties in vivo may change. The increment in γ1-containing GABAARs is expected to reduce the sensitivity to benzodiazepines and block the effects of zolpidem (41), drugs commonly used in clinical practice. Interestingly, benzodiazepine use apparently increases the risk of dementia (42). Reduction and remodeling of GABA signaling in AD may explain, at least in part, the cortical disinhibition in several areas of the AD brain, leading to hyperexcitability and its correlation with the level of dementia (43). Future analyses of the pharmacological and physiological consequences of abnormal GABA signaling, together with the precise identification of the cellular localization of abnormal native GABAARs in the AD brain, will help to manipulate GABAergic signaling for better treatments of AD.

Materials and Methods

qPCR and Protein Analysis of GABA Subunits Expressed in AD and Control Brains.

GABAAR subunit mRNAs were evaluated by qPCR (SI Materials and Methods). iScript (Bio-Rad) was used to obtain cDNA from 4 μg of mRNA (isolated with Oligotex kit; cat. no. 70022; Qiagen). The resulting cDNA was used as a template to perform qPCR in triplicate with the SYBR Green mix from Bio-Rad and an iCycler real-time PCR machine. For data analysis, threshold cycles obtained in qPCR were analyzed with the 2−ΔCt method (44), which was used to calculate the fold change in gene expression normalized to the endogenous reference gene GAPDH. Nonparametric statistical analyses (45) were performed with JMP software (version 7.0.2; SAS Institute). Western blots to membrane protein preparations were prepared as described elsewhere by using antibodies against α1 and γ1 (Santa Cruz Biotechnology) and β1 and γ2 (Millipore). Band densities were compared by using ImageJ software (National Institutes of Health).

Electrophysiological Recording of Microtransplanted Native Plasma Membranes.

Membranes from human temporal cortices were isolated by using procedures described previously (10, 11, 46). Fifty nanoliters of a membrane preparation (protein concentration, 1–1.5 mg/mL) were injected into stage V–VI Xenopus oocytes, and, from day 1 to day 3 after injection, membrane currents were recorded from oocytes voltage-clamped at −80 mV. Results are expressed as mean ± SEM unless otherwise stated. SCS was bought from Tocris and the rest of the chemicals were from Sigma–Aldrich.

Analyses of Electrophysiological Data.

Evaluation of GABA current of the temporal cortex from a particular brain required several injections of membranes into different batches of oocytes. Therefore, the specific amplitude of GABA currents for a particular brain was taken as the median of the pooled data obtained from different experiments (Fig. S1). For comparisons of the amplitude of specific GABA currents between control and AD groups, we used the nonparametric Wilcoxon test and considered differences significant at the level of P < 0.05. Rise and decay times of GABA currents were measured by using Clampfit 10 software. GABA sensitivity (as an EC50) was determined by fitting the logistic equation (OriginPro 8.5) to response/concentration curves:

graphic file with name pnas.1204606109uneq1.jpg

where x is the concentration of GABA (in M), I is the amplitude of the agonist-response (in nA), and k is the Hill coefficient. Statistical analyses were carried out with JMP software, version 7.0.2. To compare the electrophysiological properties and the EC50 of AD and control groups, accounting for the intersubject variability and the variability within the same case, we used nested ANOVA analysis of case within diagnosis.

Supplementary Material

Supporting Information
supp_109_25_10071__index.html (984B, html)

Acknowledgments

We are extremely grateful to all the individuals who donated their brains for scientific research. The brain tissues used for this work were provided by the Institute for Brain Aging and Dementia Tissue Repository of University of California, Irvine (UCI). We are thankful to Drs. Rita Petersen and Robert Newcomb, from the UCI Center for Statistical Consulting, for advice with the statistical analyses. This work was supported by King Abdul Aziz City for Science and Technology (Saudi Arabia) Grant KACST-46749 (to R.M.).

Footnotes

The authors declare no conflict of interest.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1204606109/-/DCSupplemental.

References

  • 1.Selkoe DJ. Alzheimer's disease is a synaptic failure. Science. 2002;298:789–791. doi: 10.1126/science.1074069. [DOI] [PubMed] [Google Scholar]
  • 2.Rissman RA, Mobley WC. Implications for treatment: GABAA receptors in aging, Down syndrome and Alzheimer's disease. J Neurochem. 2011;117:613–622. doi: 10.1111/j.1471-4159.2011.07237.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Luchetti S, Huitinga I, Swaab DF. Neurosteroid and GABA-A receptor alterations in Alzheimer's disease, Parkinson's disease and multiple sclerosis. Neuroscience. 2011;191:6–21. doi: 10.1016/j.neuroscience.2011.04.010. [DOI] [PubMed] [Google Scholar]
  • 4.Olsen RW, Sieghart W. International Union of Pharmacology. LXX. Subtypes of gamma-aminobutyric acid(A) receptors: Classification on the basis of subunit composition, pharmacology, and function. Update. Pharmacol Rev. 2008;60:243–260. doi: 10.1124/pr.108.00505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Lanctôt KL, Herrmann N, Mazzotta P, Khan LR, Ingber N. GABAergic function in Alzheimer's disease: Evidence for dysfunction and potential as a therapeutic target for the treatment of behavioural and psychological symptoms of dementia. Can J Psychiatry. 2004;49:439–453. doi: 10.1177/070674370404900705. [DOI] [PubMed] [Google Scholar]
  • 6.McEnery MW. The mitochondrial benzodiazepine receptor: Evidence for association with the voltage-dependent anion channel (VDAC) J Bioenerg Biomembr. 1992;24:63–69. doi: 10.1007/BF00769532. [DOI] [PubMed] [Google Scholar]
  • 7.Scarf AM, Ittner LM, Kassiou M. The translocator protein (18 kDa): central nervous system disease and drug design. J Med Chem. 2009;52:581–592. doi: 10.1021/jm8011678. [DOI] [PubMed] [Google Scholar]
  • 8.Cosenza-Nashat M, et al. Expression of the translocator protein of 18 kDa by microglia, macrophages and astrocytes based on immunohistochemical localization in abnormal human brain. Neuropathol Appl Neurobiol. 2009;35:306–328. doi: 10.1111/j.1365-2990.2008.01006.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Mizukami K, Grayson DR, Ikonomovic MD, Sheffield R, Armstrong DM. GABAA receptor β 2 and β 3 subunits mRNA in the hippocampal formation of aged human brain with Alzheimer-related neuropathology. Brain Res Mol Brain Res. 1998;56:268–272. doi: 10.1016/s0169-328x(97)00347-1. [DOI] [PubMed] [Google Scholar]
  • 10.Miledi R, Dueñas Z, Martinez-Torres A, Kawas CH, Eusebi F. Microtransplantation of functional receptors and channels from the Alzheimer's brain to frog oocytes. Proc Natl Acad Sci USA. 2004;101:1760–1763. doi: 10.1073/pnas.0308224100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Bernareggi A, Dueñas Z, Reyes-Ruiz JM, Ruzzier F, Miledi R. Properties of glutamate receptors of Alzheimer's disease brain transplanted to frog oocytes. Proc Natl Acad Sci USA. 2007;104:2956–2960. doi: 10.1073/pnas.0611513104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Pennisi G, et al. Motor cortex excitability in Alzheimer disease: One year follow-up study. Neurosci Lett. 2002;329:293–296. doi: 10.1016/s0304-3940(02)00701-2. [DOI] [PubMed] [Google Scholar]
  • 13.Di Lazzaro V, et al. Motor cortex hyperexcitability to transcranial magnetic stimulation in Alzheimer's disease. J Neurol Neurosurg Psychiatry. 2004;75:555–559. doi: 10.1136/jnnp.2003.018127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Marcade M, et al. Etazolate, a neuroprotective drug linking GABA(A) receptor pharmacology to amyloid precursor protein processing. J Neurochem. 2008;106:392–404. doi: 10.1111/j.1471-4159.2008.05396.x. [DOI] [PubMed] [Google Scholar]
  • 15.Fritschy JM, Harvey RJ, Schwarz G. Gephyrin: Where do we stand, where do we go? Trends Neurosci. 2008;31:257–264. doi: 10.1016/j.tins.2008.02.006. [DOI] [PubMed] [Google Scholar]
  • 16.Jesse S, et al. Glial fibrillary acidic protein and protein S-100B: Different concentration pattern of glial proteins in cerebrospinal fluid of patients with Alzheimer's disease and Creutzfeldt-Jakob disease. J Alzheimers Dis. 2009;17:541–551. doi: 10.3233/JAD-2009-1075. [DOI] [PubMed] [Google Scholar]
  • 17.Russek SJ. Evolution of GABA(A) receptor diversity in the human genome. Gene. 1999;227:213–222. doi: 10.1016/s0378-1119(98)00594-0. [DOI] [PubMed] [Google Scholar]
  • 18.Whiting PJ, et al. Molecular and functional diversity of the expanding GABA-A receptor gene family. Ann N Y Acad Sci. 1999;868:645–653. doi: 10.1111/j.1749-6632.1999.tb11341.x. [DOI] [PubMed] [Google Scholar]
  • 19.Roden WH, Peugh LD, Jansen LA. Altered GABA(A) receptor subunit expression and pharmacology in human Angelman syndrome cortex. Neurosci Lett. 2010;483:167–172. doi: 10.1016/j.neulet.2010.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Armstrong DM, Ikonomovic MD, Sheffield R, Wenthold RJ. AMPA-selective glutamate receptor subtype immunoreactivity in the entorhinal cortex of non-demented elderly and patients with Alzheimer's disease. Brain Res. 1994;639:207–216. doi: 10.1016/0006-8993(94)91732-9. [DOI] [PubMed] [Google Scholar]
  • 21.Gong Y, Lippa CF, Zhu J, Lin Q, Rosso AL. Disruption of glutamate receptors at Shank-postsynaptic platform in Alzheimer's disease. Brain Res. 2009;1292:191–198. doi: 10.1016/j.brainres.2009.07.056. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Partin KM, Patneau DK, Winters CA, Mayer ML, Buonanno A. Selective modulation of desensitization at AMPA versus kainate receptors by cyclothiazide and concanavalin A. Neuron. 1993;11:1069–1082. doi: 10.1016/0896-6273(93)90220-l. [DOI] [PubMed] [Google Scholar]
  • 23.Palma E, et al. Phosphatase inhibitors remove the run-down of gamma-aminobutyric acid type A receptors in the human epileptic brain. Proc Natl Acad Sci USA. 2004;101:10183–10188. doi: 10.1073/pnas.0403683101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Barish ME. A transient calcium-dependent chloride current in the immature Xenopus oocyte. J Physiol. 1983;342:309–325. doi: 10.1113/jphysiol.1983.sp014852. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Thompson SA, et al. Salicylidene salicylhydrazide, a selective inhibitor of beta 1-containing GABAA receptors. Br J Pharmacol. 2004;142:97–106. doi: 10.1038/sj.bjp.0705689. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Allen SJ, et al. Biochemical evidence of selective nerve cell changes in the normal ageing human and rat brain. J Neurochem. 1983;41:256–265. doi: 10.1111/j.1471-4159.1983.tb11837.x. [DOI] [PubMed] [Google Scholar]
  • 27.Owen F, Poulter M, Waddington JL, Mashal RD, Crow TJ. [3H]R05-4864 and [3H]flunitrazepam binding in kainate-lesioned rat striatum and in temporal cortex of brains from patients with senile dementia of the Alzheimer type. Brain Res. 1983;278:373–375. doi: 10.1016/0006-8993(83)90276-7. [DOI] [PubMed] [Google Scholar]
  • 28.Coleman PD, Flood DG. Neuron numbers and dendritic extent in normal aging and Alzheimer's disease. Neurobiol Aging. 1987;8:521–545. doi: 10.1016/0197-4580(87)90127-8. [DOI] [PubMed] [Google Scholar]
  • 29.Hof PR, Morrison JH. The aging brain: Morphomolecular senescence of cortical circuits. Trends Neurosci. 2004;27:607–613. doi: 10.1016/j.tins.2004.07.013. [DOI] [PubMed] [Google Scholar]
  • 30.Dickstein DL, et al. Changes in the structural complexity of the aged brain. Aging Cell. 2007;6:275–284. doi: 10.1111/j.1474-9726.2007.00289.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Loerch PM, et al. Evolution of the aging brain transcriptome and synaptic regulation. PLoS ONE. 2008;3:e3329. doi: 10.1371/journal.pone.0003329. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Sherwood CC, et al. Aging of the cerebral cortex differs between humans and chimpanzees. Proc Natl Acad Sci USA. 2011;108:13029–13034. doi: 10.1073/pnas.1016709108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Rabinovici GD, et al. Increased metabolic vulnerability in early-onset Alzheimer's disease is not related to amyloid burden. Brain. 2010;133:512–528. doi: 10.1093/brain/awp326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Marshall GA, Fairbanks LA, Tekin S, Vinters HV, Cummings JL. Early-onset Alzheimer's disease is associated with greater pathologic burden. J Geriatr Psychiatry Neurol. 2007;20:29–33. doi: 10.1177/0891988706297086. [DOI] [PubMed] [Google Scholar]
  • 35.Garcia-Marin V, et al. Diminished perisomatic GABAergic terminals on cortical neurons adjacent to amyloid plaques. Front Neuroanat. 2009;3:28. doi: 10.3389/neuro.05.028.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Iwakiri M, et al. An immunohistochemical study of GABAA receptor gamma subunits in Alzheimer's disease hippocampus: Relationship to neurofibrillary tangle progression. Neuropathology. 2009;29:263–269. doi: 10.1111/j.1440-1789.2008.00978.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Agarwal S, Tannenberg RK, Dodd PR. Reduced expression of the inhibitory synapse scaffolding protein gephyrin in Alzheimer's disease. J Alzheimers Dis. 2008;14:313–321. doi: 10.3233/jad-2008-14305. [DOI] [PubMed] [Google Scholar]
  • 38.Fraser DD, et al. GABAA/benzodiazepine receptors in acutely isolated hippocampal astrocytes. J Neurosci. 1995;15:2720–2732. doi: 10.1523/JNEUROSCI.15-04-02720.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Bonnert TP, et al. θ, a novel gamma-aminobutyric acid type A receptor subunit. Proc Natl Acad Sci USA. 1999;96:9891–9896. doi: 10.1073/pnas.96.17.9891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Huntsman MM, Huguenard JR. Fast IPSCs in rat thalamic reticular nucleus require the GABAA receptor beta1 subunit. J Physiol. 2006;572:459–475. doi: 10.1113/jphysiol.2006.106617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Khom S, et al. Pharmacological properties of GABAA receptors containing gamma1 subunits. Mol Pharmacol. 2006;69:640–649. doi: 10.1124/mol.105.017236. [DOI] [PubMed] [Google Scholar]
  • 42.Gallacher J, et al. Benzodiazepine use and risk of dementia: Evidence from the Caerphilly Prospective Study (CaPS) J Epidemiol Community Health. 2011 doi: 10.1136/jech-2011-200314. [DOI] [PubMed] [Google Scholar]
  • 43.Liepert J, Bär KJ, Meske U, Weiller C. Motor cortex disinhibition in Alzheimer's disease. Clin Neurophysiol. 2001;112:1436–1441. doi: 10.1016/s1388-2457(01)00554-5. [DOI] [PubMed] [Google Scholar]
  • 44.Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Δ Δ C(T)) Method. Methods. 2001;25:402–408. doi: 10.1006/meth.2001.1262. [DOI] [PubMed] [Google Scholar]
  • 45.Yuan JS, Reed A, Chen F, Stewart CN., Jr Statistical analysis of real-time PCR data. BMC Bioinformatics. 2006;7:85. doi: 10.1186/1471-2105-7-85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Limon A, Reyes-Ruiz JM, Miledi R. Microtransplantation of neurotransmitter receptors from postmortem autistic brains to Xenopus oocytes. Proc Natl Acad Sci USA. 2008;105:10973–10977. doi: 10.1073/pnas.0804386105. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information
supp_109_25_10071__index.html (984B, html)
1204606109_pnas.201204606SI.pdf (216.4KB, pdf)

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES