miRNA-132/212 Gene-Deletion Aggravates the Effect of Oxygen-Glucose Deprivation on Synaptic Functions in the Female Mouse Hippocampus

Cerebral ischemia and its sequelae, which include memory impairment, constitute a leading cause of disability worldwide. Micro-RNAs (miRNA) are evolutionarily conserved short-length/noncoding RNA molecules recently implicated in adaptive/maladaptive neuronal responses to ischemia. Previous research independently implicated the miRNA-132/212 cluster in cholinergic signaling and synaptic transmission, and in adaptive/protective mechanisms of neuronal responses to hypoxia. However, the putative role of miRNA-132/212 in the response of synaptic transmission to ischemia remained unexplored. Using hippocampal slices from female miRNA-132/212 double-knockout mice in an established electrophysiological model of ischemia, we here describe that miRNA-132/212 gene-deletion aggravated the deleterious effect of repeated oxygen-glucose deprivation insults on synaptic transmission in the dentate gyrus, a brain region crucial for learning and memory functions. We also examined the effect of miRNA-132/212 gene-deletion on the expression of key mediators in cholinergic signaling that are implicated in both adaptive responses to ischemia and hippocampal neural signaling. miRNA-132/212 gene-deletion significantly altered hippocampal AChE and mAChR-M1, but not α7-nAChR or MeCP2 expression. The effects of miRNA-132/212 gene-deletion on hippocampal synaptic transmission and levels of cholinergic-signaling elements suggest the existence of a miRNA-132/212-dependent adaptive mechanism safeguarding the functional integrity of synaptic functions in the acute phase of cerebral ischemia.


Introduction
Oxygen is a key determinant of mammalian life and severe oxygen deprivation is most often its antithesis. Oxygen deprivation accompanied by a sudden disruption of substrate delivery to brain tissue, as occurring in the form of ischemic stroke, triggers a devastating cascade of pathophysiological events, culminating in irreversible dysfunction and loss of CNS tissue [1]. The resulting loss of neural function in stroke patients constitutes a leading cause of disability worldwide [2], as the armamentarium of pharmacological interventions curves, square pulses of increasing stimulus intensity (duration 200 µs, 10 s inter-pulse interval, 0-9 V with 1 V increments) were delivered at the medial perforant pathway (MPP) to elicit fEPSP responses. For I/O curves, fEPSP slopes were normalized to the highest inducible slope value, for each slice recording. The stimulus intensity for the assessment of basal synaptic transmission under the OGD treatment protocol and paired pulse recordings was adjusted to elicit 80% of the maximal inducible amplitude established with I/O curves, and kept constant for the rest of the recordings. Paired-pulse inhibition (PPI) fEPSP responses were evoked by applying two consecutive stimulation pulses, with increasing interstimulus intervals (8 pulses, 20-160 ms interstimulus intervals with 20 ms increments). The effects of repeated OGD treatment on synaptic transmission were normalized to the last 5 min of normoxic and euglycemic baseline recording, preceding the OGD treatment protocol and then analyzed by comparing fEPSP slopes during OGD and recovery. Data were averaged within 30 s intervals (3 consecutive sweeps), further averaged within animals, and compared between miRNA-132/212 −/− and WT mice. A total of n = 34 slices were included in the analysis. On average, n = 2-3 slices per animal and a total of n = 7 WT and 6 miRNA 132/212 −/− animals were analyzed.

Oxygen Glucose Deprivation Protocol
Standard aCSF, of the following composition (in mM): 125 NaCl, 2.5 KCl, 20 NaHCO 3 , 2.5 CaCl 2 , 1 MgCl 2 , 25 D-glucose and 1 NaH 2 PO 4 , was modified as follows to achieve an OGD-aCSF. Glucose was replaced with equimolar sucrose and the solution was continuously gassed with 100% N 2 gas instead of a 95% O 2 /5% CO 2 gas mixture, as previously established [55]. pH was monitored and kept constant at 7.35-7.40 throughout the experiment by buffering the OGD-aCSF with 10 mM NaHCO 3 and 25 mM HEPES. Firstly, evoked potentials were recorded for 15 min under euglycemic and normoxic conditions by superfusing the slices with standard aCSF, gassed with 95% O 2 /5% CO 2 , before any treatment. The slices were then superfused with OGD-aCSF (100% N 2 ) for a total of three OGD 8 min intervals. Each OGD interval was followed by a recovery period of 10 min, in which the slices were superfused with standard, euglycemic, normoxic aCSF (95% O 2 /5% CO 2 ).
After 15 min of baseline recording under normoxic conditions (95% O 2 /5% CO 2 ), slices were superfused with OGD aCSF three times for 8 min with intervals of 10 min as recovery periods under normoxic and euglycemic aCSF superfusion (95% O 2 /5% CO 2 ), following every OGD period ( Figure 1 (inset 2)). Changes in solutions reached the recording chamber within 50 s, which was taken into account in our calculations. To estimate the local oxygen saturation of the aCSF in the recording chamber an optode (Needle type housing oxygen microsensor PSt7-02, tip diameter <50 µm, OXY-1 ST transmitter (PreSens -Precision Sensing GmbH, Am BioPark 11, 93053 Regensburg, Germany) was placed approximately 100-200 µm vertically above the recording site. Beforehand, we used a two-point calibration process as per the manufacturer's instructions, with minor adjustments as previously established [56]. To achieve a nominally 0% oxygen solution (Cal 0) an 80 mM sodium sulfite solution (Na 2 SO 3 in ddH 2 O) was prepared. Cobalt nitrate Co(NO 3 ) 2 standard solution (ρ(Co) = 1000 mg/L; concentration 0.5 mol/L in nitric acid) was used to catalyze oxygen desaturation. To estimate the maximal oxygen saturation of carbogenated aCSF (Cal 100), standard aCSF was bubbled with 95% O 2 /5% CO 2 for at least 1 h before calibration. In accordance with previously published data [56], the maximal achievable PO 2 of aCSF in our experimental setup was estimated to be (720 ± 10) Torr. Oxygen saturation is reported as percentage of the maximal O 2 saturation of continuously carbogenated aCSF. Oxygen saturation measured in the recording chamber ranged from 78.450 ± 0.024% during normoxia to 10.380 ± 0.046% under OGD treatment. (Figure 1 (inset 2)) shows the local oxygen saturation in the recording chamber throughout the OGD treatment. Inset (2) Temporal course of the oxygen saturation measured in the recording chamber during repeated episodes of OGD. Superfusion of the recording chamber with 100% N2 gassed OGD aCSF led to a sharp drop of oxygen saturation below 20% within 3 min bottoming at less than 12% oxygen saturation at the end of the OGD period. Reperfusion with carbogenated aCSF (95%O2/5%CO2) increased the oxygen saturation back to values greater than 60% oxygen saturation in less than 1 min. Graph derived from optode recorded oxygen saturation data of three independent recording days. aCSF: artificial cerebrospinal fluid, OGD: Oxygen Glucose Deprivation. (A) I/O curves generated by plotting normalized fEPSP slope changes versus increasing pulses of input stimulation voltages obtained in WT (n = 7) and miRNA-132/212 −/− mice (n = 6) hippocampal slices. Two-way repeated measures ANOVA did not reveal statistically significant differences between genotypes before the implementation of the OGD treatment. Right inset: Representative raw fEPSP traces recorded from hippocampal slices obtained from WT mice (left) and miRNA-132/212 −/− mice (right) in response to increasing stimulus intensities show no apparent differences. (B) PPI ratios (second pulse/first pulse) plotted against inter-pulse intervals from Superfusion of the recording chamber with 100% N2 gassed OGD aCSF led to a sharp drop of oxygen saturation below 20% within 3 min bottoming at less than 12% oxygen saturation at the end of the OGD period. Reperfusion with carbogenated aCSF (95%O 2 /5%CO 2 ) increased the oxygen saturation back to values greater than 60% oxygen saturation in less than 1 min. Graph derived from optode recorded oxygen saturation data of three independent recording days. aCSF: artificial cerebrospinal fluid, OGD: Oxygen Glucose Deprivation. (A) I/O curves generated by plotting normalized fEPSP slope changes versus increasing pulses of input stimulation voltages obtained in WT (n = 7) and miRNA-132/212 −/− mice (n = 6) hippocampal slices. Two-way repeated measures ANOVA did not reveal statistically significant differences between genotypes before the implementation of the OGD treatment. Right inset: Representative raw fEPSP traces recorded from hippocampal slices obtained from WT mice (left) and miRNA-132/212 −/− mice (right) in response to increasing stimulus intensities show no apparent differences. (B) PPI ratios (second pulse/first pulse) plotted against inter-pulse intervals from 20 to 160 ms. No statistically significant differences ("ns" in the figure) between genotypes were detected by two-way repeated measures ANOVA. Right inset: Depiction of representative raw fEPSP traces obtained during the PPI protocol described in the materials and methods section in the dentate gyrus of WT (upper trace) and miRNA-132/212 −/− mice (lower trace), without macroscopic differences. Data is presented as mean ± SEM. fEPSP: field excitatory postsynaptic potential, I/O: Input Output, OGD: Oxygen Glucose Deprivation, PPI: Paired Pulse Inhibition.

Statistical Analysis
Statistical analyses were performed using GraphPad Prism Software, Version 8.0 (GraphPad Software, 2365 Northside Dr. Suite 560, San Diego, CA 92108, USA). Statistical analyses of more than two groups were conducted using analysis of variances (ANOVA), as appropriate. Specifically, we used two-way repeated measures ANOVA (or mixed model ANOVA) followed by Bonferroni's multiple comparisons test to analyze normalized fEPSP slopes obtained from miRNA 132/212 −/− and WT slices over time. For statistical analyses of data strictly from within subject designs, we performed one-way ANOVAs with repeated measures and subsequent Bonferroni's multiple comparisons tests, where appropriate. p-values and the number of samples used (n) are reported in the main text and represented in the figures (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p< 0.0001). All data is expressed as mean ± standard error. The threshold of statistical significance (α) was set at 0.05 in all analyses.

Dentate Gyri Basal Synaptic Transmission and Short-Term Plasticity Are Unaffected by miRNA 132/212 Gene Deletion
Before implementing the OGD treatment protocols we assessed whether miRNA 132/212 −/− and WT dentate gyri differed in their basal synaptic transmission properties. To this aim, we used standard Input/Output protocol (Materials and Methods). Recordings obtained from miRNA 132/212 −/− and WT hippocampal slices showed virtually indistinguishable synaptic responses ( Figure 1A). Two-way repeated measures ANOVA showed a statistically significant main effect of input voltage (**** p < 0.0001, F 1.342, 16.10 = 316.3), but no significant main effect of genotype ( ns P = 0.9894, F 1, 12 = 0.0001839) or input voltage × genotype interaction ( ns P = 0.8830, F 9, 108 = 0.4834). The dentate gyrus circuitry has the capacity to respond to fast repetitive stimulation via the perforant pathway with a decrease of neurotransmitter release, which can be assessed by delivering a succession of two pulses of increasing interstimulus intervals and quantifying the paired pulse inhibition (PPI) [57,58]. This conserved form of short-term plasticity has not only been associated with higher cognitive functions, such as learning and memory, but its impairment is associated with early pathophysiological hallmarks in icto-and epileptogenesis following convulsant stimulation [59]. PPI can be quantified as a paired pulse ratio (PPI-ratio) by dividing the value of the measured amplitude of the second fEPSP by the amplitude of the first fEPSP. Thus, lower PPI-ratio values indicate more pronounced inhibitory short term plasticity.

Dentate Gyri Synaptic Depression Following OGD Is Aggravated in miRNA 132/212 −/− Mice
The superfusion of acutely dissected brain slices with OGD-aCSF has proven to be a useful, widely established ex vivo model of ischemia employed in the search for novel therapeutic targets to treat ischemia-induced brain injuries such as stroke e.g., [60][61][62][63][64][65], as well as for the illumination of the pathophysiological responses of neuronal circuits to ischemia e.g., [55,[66][67][68][69][70]. Adapting a previously established approach [55], we used superfusion with OGD-aCSF to study the acute effects of repeated ischemia-reperfusion episodes on synaptic transmission in the dentate gyrus of acutely dissected slices obtained from miRNA 132/212 −/− mice and their respective WT littermates, who served as controls. We chose superfusion intervals with the OGD buffer for 8 min at a flow rate of 4 mL/min and a temperature of 32 • C, as in our pilot experiments this treatment robustly depressed synaptic transmission to ≤50% of its initial baseline, while allowing for recovery to 70-90% of the initial baseline within 10 min of normoxic/euglycemic reperfusion in WT slices. This protocol enabled us to study the effects of a series of multiple ischemia-reperfusion injuries on depression and recovery of synaptic transmission. We performed a two-way ANOVA analysis with repeated measures to compare synaptic transmission between the two genotypes over the course of the OGD protocol, which unveiled statistically significant main effects of time (**** p < 0.0001, F 137, 1507 = 170.4) and genotype (** p = 0.0057, F 1, 11 = 11.71) and a significant time x genotype interaction (**** p < 0.0001, F 137, 1507 = 4.523) (see Figure 2A). Subsequent Bonferroni's multiple comparisons tests for each 30 s interval revealed that fEPSP slopes recorded in miR132/212 −/− dentate gyri depressed to significantly lower values at the end of the first OGD episode (* p = 0.0442, t 1518 = 3.606) and during the first minute of the first recovery in comparison to recordings in respective WT control slices (all * p < 0.05, (t 1518 ) min = 3.603; (t 1518 ) max = 3.924, t-values referring to lowest to highest value in this time frame) (see Figure 2B1). However, this difference did not remain statistically significant up to the end of the first recovery episode ( ns P > 0.9999, t 1518 = 1.707) (see Figure 2C1). Furthermore, the genotypes also did not differ significantly at the end of the second OGD ( ns P = 0.0751, t 1518 = 3.465) or the second recovery episode ( ns P > 0.9999, t 1518 = 2.258) (see Figure 2B2,C1,C2). However, the synaptic responses from slices obtained from miRNA 132/212 −/− and WT mice differed for almost the entire third OGD interval (all * p < 0.05, for every 30 s interval between 53.5 min and 59 min, t 1518 = 3.603-4.710, t-values referring to lowest to highest value in this time frame), as well as the majority of the third recovery episode (all * p< 0.05, between 60.5 min and 64, 65, and 66 min, t 1518 = 3.642-4.790, t values referring to lowest to highest t value in this time frame, respectively) with miRNA 132/212 −/− slices recovering to significantly lower fEPSP slope values at the end of the last recovery period, compared to the WT controls (* p= 0.0473, t 1518 = 3.589) (see Figure 2A,B3,C3). Curiously, while slices obtained from WT mice robustly exhibited a small initial potentiation within the first 30 s of initial OGD, of on average of 9.441 ± 0.036% with respect to their initial prehypoxic/euglycemic baseline, this phenomenon was virtually absent in recordings obtained from miRNA 132/212 −/− mice (average depression of 0.013 ± 0.070% with respect to their initial baseline within the first 30 s of first OGD), (see Figure 2A). Collectively these data indicate that the depressive effect of repeated OGD on synaptic transmission is more pronounced in the absence of the miRNA 132/212 cluster and this difference between the genotypes becomes most pronounced with successive OGD insults.

Episodic OGD Impairs the Recovery of Dentate Gyri Synaptic Transmission in miRNA 132/212 −/−
In order to further investigate the pronounced differences in the recovery of synaptic transmission at the last ischemia-reperfusion interval (Figure 2), we analyzed the efficiency of recovery over the course of the three ischemia-reperfusion episodes within both respective groups, using a within-subject design. We thus performed one-way ANOVAs with repeated measures, as well as subsequent Bonferroni's multiple comparisons tests, to compare normalized fEPSP slope baseline values to the endpoints of each recovery interval for recordings from miRNA 132/212 −/− and WT controls, respectively. Strikingly, fEPSP slopes obtained from miRNA 132/212 −/− slices recovered to significantly reduced values at the end of each ischemia-reperfusion episode, in comparison to preceding recovery intervals, indicative of a cumulative effect of repeated OGD periods on the depression of synaptic transmission (see Figure 3A). One-way repeated measures ANOVA revealed a significant main effect of time in slices from miRNA 132/212 −/− mice (**** p< 0.0001, F 1.701, 8.503 = 182.9). Bonferroni's multiple comparisons tests showed statistically significant differences between the fEPSP slope values at normoxic, euglycemic baseline in comparison to slope values at the end of each recovery interval (**** p < 0.0001, t 5 = 17.88, 17.59, Baseline vs. first and third recovery, respectively and *** p = 0.0001, t 5 = 15.37, Baseline vs. second recovery). Furthermore, post hoc tests indicated statistically significant differences between the first vs. second recovery (* p = 0.0352, t 5 = 4.596), first vs. third recovery (** p = 0.0034, t 5 = 7.776) and second vs. third recovery (** p = 0.0040, t 5 = 7.512). In contrast, in slices obtained from WT mice within-subject analysis revealed statistically significant differences between fEPSP slope values reached at the first vs. the third recovery period, while values at the first vs. second and second vs. third recovery periods no longer differed significantly. Data is presented as mean ± SEM. fEPSP: field excitatory postsynaptic potential, OGD: Oxygen Glucose Deprivation. p < 0.05 was considered significant. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. ns: no statistical significance.
One-way ANOVA with repeated measures also revealed a statistically significant main effect of time on normalized fEPSP slope values in WT slices (**** p < 0.0001, F1.688, 10.13 = 81.83, with fEPSP slopes reaching lower values after all recoveries in comparison to the initial baseline (Bonferroni's multiple comparisons tests, all *** p < 0.001, t6 = 11.51, 9.13, for Baseline fEPSP slope values vs. values at the end of the first and second recovery, respectively and **** p < 0.0001, t6 = 18.71 for Baseline vs. third recovery). fEPSP slope values recovered to significantly lower values at the end of the third vs. the first recovery interval (** p < 0.0017, t6 = 7.559). However, comparing the fEPSP slope values reached at the end of the first vs. the second ( ns P = 0.0636, t6 = 3.658), as well as the second vs. the third ( ns P > 0.9999, t6 = 0.7425) recovery intervals, no statistically significant difference was observed (see Figure 3B). In summary, these findings indicate initial reduction in followed by a stabilization of synaptic transmission in response to repeated oxygen glucose deprivation in WT dentate gyri, which could not be observed in the slices obtained from mice In contrast, in slices obtained from WT mice (white filled circles) within-subject analysis revealed statistically significant differences between fEPSP slope values reached at the first vs. the third recovery period, while values at the first vs. second and second vs. third recovery periods no longer differed significantly. Data is presented as mean ± SEM. fEPSP: field excitatory postsynaptic potential, OGD: Oxygen Glucose Deprivation. p < 0.05 was considered significant. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. ns: no statistical significance.
One-way ANOVA with repeated measures also revealed a statistically significant main effect of time on normalized fEPSP slope values in WT slices (**** p < 0.0001, F 1.688, 10.13 = 81.83, with fEPSP slopes reaching lower values after all recoveries in comparison to the initial baseline (Bonferroni's multiple comparisons tests, all *** p < 0.001, t 6 = 11.51, 9.13, for Baseline fEPSP slope values vs. values at the end of the first and second recovery, respectively and **** p < 0.0001, t 6 = 18.71 for Baseline vs. third recovery). fEPSP slope values recovered to significantly lower values at the end of the third vs. the first recovery interval (** p < 0.0017, t 6 = 7.559). However, comparing the fEPSP slope values reached at the end of the first vs. the second ( ns P = 0.0636, t 6 = 3.658), as well as the second vs. the third ( ns P > 0.9999, t 6 = 0.7425) recovery intervals, no statistically significant difference was observed (see Figure 3B). In summary, these findings indicate initial reduction in followed by a stabilization of synaptic transmission in response to repeated oxygen glucose deprivation in WT dentate gyri, which could not be observed in the slices obtained from mice lacking miRNA 132/212.

miRNA 132/212 Gene Deletion Promotes the Vulnerability of Short-Term Synaptic Depression to Deleterious Effects of OGD
We subsequently examined whether OGD insults influence dentate gyrus short-term synaptic depression and whether this form of synaptic plasticity was differentially affected by OGD in miRNA 132/212 −/− and WT dentate gyri. We thus compared PPI ratios obtained before and after the OGD treatment in both genotypes using two-way repeated measures ANOVA. In slices obtained from WT mice, repeated measures ANOVA unveiled a predicted significant main effect of the interstimulus intervals (**** p < 0.0001, F 3.734, 44.81 = 36.39), no statistically significant main effect of treatment ( ns P = 0.0954, F 1, 12 = 3.275) (see Figure 4A), and a statistically significant interstimulus interval x treatment interaction (* p = 0.0192, F 7, 84 = 2.563).
These data show that the dentate gyri inhibitory short-term plasticity is significantly impaired by repeated OGD insults in the absence of the miRNA 132/212 cluster.

Basal Synaptic Transmission Remains Unaffected in the miRNA 132/212 −/− Hippocampus after Episodic OGD Insults
Next, we sought to analyze the consequence of multiple ischemia-reperfusion insults on basal synaptic transmission responses to a range of stimulus intensities. We therefore compared standard I/O curves recorded before and after the OGD treatment protocol for both groups respectively, and compared I/O curves obtained from both genotypes after the OGD treatment using separate two-way ANOVAs with repeated measures. Comparing I/O curves obtained from WT slices before and after repeated OGD, we observed a statistically significant main effect of input voltage (**** p <0.0001, F 1.214, 14.57 = 570.3), but no significant effect of treatment (before/after OGD treatment) ( ns P = 0.8042, F 1, 12 = 0.06426) and also no significant input-voltage x treatment interaction ( ns P = 0.9607, F 9, 108 = 0.3371, Figure 5A). Analyzing I/O curves from miRNA 132/212 −/− slices from before and after the OGD treatment repeated measures ANOVA revealed a significant main effect of input voltage (**** p < 0.0001, F 1.170, 11.70 = 177.8), but not treatment (before vs. after OGD treatment) ( ns P = 0.3928, F 1, 10 = 0.7975), nor a significant input voltage x treatment interaction ( ns P = 0.9758, F 9, 90 = 0.2904) (see Figure 5B). F3.734, 44.81 = 36.39), no statistically significant main effect of treatment ( ns P = 0.0954, F1, 12 = 3.275) (see Figure 4A), and a statistically significant interstimulus interval x treatment interaction (* p = 0.0192, F7, 84 = 2.563). . Repeated OGD impaired short term inhibitory plasticity in the absence of miRNA 132/212. Normalized and averaged PPI ratios (second pulse/first pulse) plotted against inter-pulse intervals from 20 to 160 ms, comparing responses from WT slices before and after the OGD treatment protocol: (A) PPI ratios, at ascending interstimulus intervals derived from WT slices and (B) miRNA-132/212 −/− slices before and after the OGD treatment protocol. (C) Comparison of PPI ratios, at ascending interstimulus intervals obtained from WT and miRNA-132/212 −/− slices after the OGD treatment. (D) Representative raw fEPSP traces obtained from WT (upper traces, black) and miRNA-132/212 −/− slices (lower traces, gray), before (left) and after (right) OGD treatment, applying a 20 ms inter-pulse interval paired pulse stimulation. Two-way repeated measures ANOVA revealed no statistically significant differences in the comparison of PPI ratios recorded in WT slices before and after OGD treatment (A). In contrast, PPI ratios increased significantly in miRNA-132/212 −/− slices after the OGD treatment, most pronounced at the 40 ms interstimulus interval (Bonferroni's multiple comparisons tests) (B). Twoway repeated measures ANOVA revealed a significant difference between the genotypes after the OGD treatment (C). Data from n = 7 WT and n = 6 miRNA 132/212 −/− is presented as mean ± SEM. p < 0.05 was considered significant. * p < 0.05. ns: no statistical significance.
In contrast, repeated measures ANOVA analyses of PPI ratios obtained from miRNA 132/212 −/− slices comparing PPI ratios from before and after OGD treatment yielded a significant main effect of inter-pulse interval (**** p < 0.0001, F4.068, 40.68 = 29.84), as well as significant main effect of treatment (before vs. after) (* p = 0.0276, F1, 10 = 6.634), with no statistically significant inter-pulse interval x treatment interaction ( ns P = 0.1173, F7, 70 = 1.724). Bonferroni post hoc tests revealed a statistically significant increase in the PPI ratio . Repeated OGD impaired short term inhibitory plasticity in the absence of miRNA 132/212. Normalized and averaged PPI ratios (second pulse/first pulse) plotted against inter-pulse intervals from 20 to 160 ms, comparing responses from WT slices before and after the OGD treatment protocol: (A) PPI ratios, at ascending interstimulus intervals derived from WT slices and (B) miRNA-132/212 −/− slices before and after the OGD treatment protocol. (C) Comparison of PPI ratios, at ascending interstimulus intervals obtained from WT and miRNA-132/212 −/− slices after the OGD treatment. (D) Representative raw fEPSP traces obtained from WT (upper traces, black) and miRNA-132/212 −/− slices (lower traces, gray), before (left) and after (right) OGD treatment, applying a 20 ms inter-pulse interval paired pulse stimulation. Two-way repeated measures ANOVA revealed no statistically significant differences in the comparison of PPI ratios recorded in WT slices before and after OGD treatment (A). In contrast, PPI ratios increased significantly in miRNA-132/212 −/− slices after the OGD treatment, most pronounced at the 40 ms interstimulus interval (Bonferroni's multiple comparisons tests) (B). Two-way repeated measures ANOVA revealed a significant difference between the genotypes after the OGD treatment (C). Data from n = 7 WT and n = 6 miRNA 132/212 −/− is presented as mean ± SEM. p < 0.05 was considered significant. and miRNA-132/212 −/− mice (n = 6) hippocampal slices. Separate two-way ANOVAs with repeated measures comparing basal synaptic transmission recorded before and after OGD treatment in WT dentate gyri (A) and miRNA-132/212 −/− dentate gyri (B) did not reveal statistically significant differences. (C) Normalized fEPSP slopes obtained from WT and miRNA-132/212 −/− slices after OGD treatment did not differ significantly. (D) Representative raw fEPSP traces recorded from hippocampal slices obtained from WT mice (upper row) and miRNA-132/212 −/− mice (lower row) in response to ascending stimulus intensities before (left) and after (right) OGD treatment. In both cases, the differences in the peak amplitudes before and after OGD are emphasized in the right panels by vertical bar indicators. Data is presented as mean ± SEM. fEPSP: field excitatory postsynaptic potential, OGD: Oxygen Glucose Deprivation. ns: no statistical significance.
In summary, the sensitivity of basal synaptic transmission in response to a range of different stimulus intensities remained unaffected after repeated OGD insults in either genotype. Figure 5. Repeated OGD did not significantly alter basal synaptic transmission in either genotype. I/O curves generated by plotting normalized fEPSP slope changes versus increasing pulses of input stimulation voltages obtained in WT (n = 7) and miRNA-132/212 −/− mice (n = 6) hippocampal slices. Separate two-way ANOVAs with repeated measures comparing basal synaptic transmission recorded before and after OGD treatment in WT dentate gyri (A) and miRNA-132/212 −/− dentate gyri (B) did not reveal statistically significant differences. (C) Normalized fEPSP slopes obtained from WT and miRNA-132/212 −/− slices after OGD treatment did not differ significantly. (D) Representative raw fEPSP traces recorded from hippocampal slices obtained from WT mice (upper row) and miRNA-132/212 −/− mice (lower row) in response to ascending stimulus intensities before (left) and after (right) OGD treatment. In both cases, the differences in the peak amplitudes before and after OGD are emphasized in the right panels by vertical bar indicators. Data is presented as mean ± SEM. fEPSP: field excitatory postsynaptic potential, OGD: Oxygen Glucose Deprivation. ns: no statistical significance.
In summary, the sensitivity of basal synaptic transmission in response to a range of different stimulus intensities remained unaffected after repeated OGD insults in either genotype.
Importantly, the confirmed miRNA 132/212 target Acetylcholinesterase (AChE) has been independently implicated in cognitive deficits in the aftermath of ischemic stroke [77]. Intriguingly, the expression of enzymes that regulate the availability of ACh at its site of action, such as AChE, have been shown to be involved in the regulation of the expression and function of nicotinergic and muscarinergic acetylcholine receptors [78,79]. Interestingly, the alpha-7 nicotinic receptor (nAChR-α7) and the muscarinic acetylcholine receptor M1 (mAChR-M1) have also been independently implicated in adaptive and maladaptive responses to hypoxia and ischemia [80][81][82][83][84][85]. Moreover, miRNA 132 has been shown to be involved in the regulation of methyl-CpG binding protein 2 (MeCP2) during ischemia preconditioning [37]. Furthermore, emerging evidence has recently started to unveil how crosstalk between the miRNA 132/212 cluster and cholinergic signaling influences synaptic transmission in the dentate gyrus [22], as well as cognitive correlates of synaptic communication in the hippocampus, such as learning and memory [33]. On the grounds of these observations and the above described aggravation of the depressive effect of repeated ischemia-reperfusion insults on synaptic transmission, we next explored the impact of miRNA 132/212 deletion on the expression of α7-nAChR, AChE, mAChR M1 and MeCP2 in the rodent hippocampus.

Modeling Ischemia Reperfusion Injuries Ex Vivo to Study the Effects of OGD on Synaptic Transmission
Studying the effects of OGD on synaptic transmission in brain slices ex vivo has proven to be a versatile approach in unveiling pathophysiological hallmarks of ischemiareperfusion induced insults to neuronal circuits and identifying novel therapeutic targets, e.g., [46,55,[61][62][63]67,68,86]. Although this methodology largely neglects the delicate interplay of all elements of the neurovascular unit in its broader in vivo context [87], it is particularly suited to study the acute functional responses of the isolated, intact neuronal circuitry to ischemia reperfusion insults, with high spatial and temporal resolution, in a controlled experimental setting. By adapting this widely established approach we here provided the first electrophysiological characterization of an involvement of the miRNA 132/212 cluster in the acute response of the dentate gyrus neuronal circuitry to repeated ischemia reperfusion insults. We showed that miRNA 132/212 deletion aggravates ischemia-induced depression of synaptic transmission in the mammalian dentate gyrus. These observations expand on the recent pioneering studies which demonstrated a protective role of miRNA 132 expression in the context of OGD in several cell lines [34][35][36]. Indeed, miRNA 132 overexpression has been shown to elicit antiapoptotic and neuroprotective effects in ischemia-challenged hippocampal neurons [35,36]. Moreover, miRNA 132 has been shown to be upregulated after OGD challenges [36], as well as in the course of ischemia preconditioning in isolated hippocampal neurons, which provided a protective effect in a subsequent OGD episode [35]. Here we explored how synaptic transmission in the dentate gyrus responds to a rapid succession of ischemia-reperfusion insults.
We here observed that in the absence of miRNA 132/212 the depression of synaptic transmission in the dentate gyrus, elicited by repeated OGD, is aggravated and the recovery from these insults is impaired. Interestingly, we observed a steady decrease in the efficiency of the synaptic transmission recovery from OGD induced depression in the absence of miRNA 132/212, as compared to WT controls. We thus propose the existence of a miRNA 132/212 dependent rapid adaptive response to ischemia-reperfusion insults, which might be involved in safeguarding the functional integrity of synaptic transmission in the dentate gyrus from ischemia-induced disruption.
While many studies have examined the influence of OGD on synaptic transmission in the hippocampal CA1 region, e.g., [55,[60][61][62][63][64]66,70,[88][89][90], the dentate gyrus has been studied less extensively in this paradigm, arguably because it has been shown to respond more resiliently to ischemia [86,[91][92][93], making the study of ischemia-reperfusion induced damage towards it functional integrity more tedious. Nevertheless, several properties of the dentate gyrus merit further exploration of the mechanisms and consequences of ischemic insults to this region. The dentate gyrus has been conceptualized to serve a "gatekeeper" role in filtering cortical excitatory inputs to the hippocampus proper and its failure to effectively limit high frequency activity of excitatory inputs was associated with ictogenesis [94]. Additionally, we observed that repeated OGD insults significantly compromise the inhibitory short-term plasticity response of the dentate gyrus to high frequency paired pulse stimulation in the absence of the miRNA 132/212 cluster, particularly in the range of short inter-pulse intervals (40 ms). It is of note that we have not studied PPI in response to paired stimuli delivered at intervals longer than 160 ms, to which the observed effects might not necessarily extrapolate. Interestingly, the impairment of PPI in the dentate gyrus molecular layer has been observed to be an early consequence of seizure inducing perforant-path stimulation ex and in vivo and hence has been discussed as a pathological hallmark in icto-and epileptogenesis [59,95]. It has been widely established that repeated episodes of ischemia increase neuronal excitability and the occurrence of epileptiform discharges in the hippocampus [63,88,89,96]. Moreover, hypoxia-induced seizures have been linked to a lasting increase in dentate gyrus granule cell excitability [97]. Therefore, it appears that the delicate homeostasis of neuronal excitability in the hippocampal formation is particularly vulnerable to hypoxia and ischemia. Interestingly, while we here observed an OGD induced compromise of dentate gyrus inhibitory short term plasticity in the absence of the miRNA 132/212 cluster, previous research has in contrast indicated that miRNA 132 upregulation aggravates epileptiform discharges in an in vitro low Mg 2+ neuronal culture model [26]. Moreover, the silencing of miRNA 132 decreased spontaneous recurrent seizures in an in vivo rodent model of lithium-pilocarpine induced temporal lobe epilepsy [24]. The contribution of miRNA 132/212 mediated signaling on neural excitability might thus depend highly on the precise context of the icto-and epileptogenic insult and differ across different neural circuits. Our findings thus invite further research on the putative role of the disruption of short term plasticity in the dentate gyrus in post ischemic icto-and epileptogenesis and a putative role of miRNA 132/212 mediated regulation of neural excitability, a topic of clinical relevance in light of the burden of early post stroke seizures [98].  [45].
Here, we focused on the previously unexplored role of miRNA 132/212 signaling in the immediate response of synaptic transmission to transient OGD in the intact dentate gyrus circuitry. Given the emerging role of the miRNA 132/212 cluster as a key regulator of cholinergic signaling in the nervous [22,71,73,77] and other organ systems [72,[74][75][76], we focused on how miRNA 132/212 disruption alters key mediators in cholinergic signaling, heavily implicated in the resilience to ischemic injuries. We here showed that the deletion of miRNA 132/212 results in a marked upregulation of AChE and a seemingly inverse downregulation of mAChR-M1 in female rodent hippocampi. AChE inactivates synaptic ACh signaling by rapid hydrolysis, discontinuing cholinergic transmission [99]. The herein observed combined upregulation of AChE together with mAChR-M1 downregulation bears the potential to critically blunt cholinergic signaling, the integrity of which appears crucial in the adaptive response to ischemia. Importantly, hypoxia has been shown to induce cognitive impairments and morphological damage in the rodent hippocampus in vivo (see also [100]), accompanied by increased AChE expression, while the administration of AChE inhibitors was able to alleviate hypoxia-induced cognitive impairments and neurodegeneration [101,102]. Moreover, recent data obtained from cerebrospinal fluid samples of human subjects not only demonstrated a robust negative regulatory relationship between miRNA 132 and its confirmed target AChE, but diminished miRNA 132 and heightened AChE levels were also associated with the development of post-stroke dementia [77]. Likewise, it has been shown that mAChR-M1 mediated signaling can be influenced by ischemic challenges, as well as consequences of ischemia in the central nervous system. Transient hypoxia has been shown to induce casein kinase 1 alpha (CK1α K46R)-dependent phosphorylation and sequestration of mAChR-M1 in vitro [103]. Correspondingly, mAChR-M1 mRNA was downregulated as a consequence of hypoxia in the rodent cortex and hippocampus in vivo [102]. Interestingly, mAChR-M1 mediated cholinergic signaling has been demonstrated to promote the accumulation and transcriptional activation of hypoxia inducible factor 1-alpha (HIF-1 α), a master regulator of the adaptive response to hypoxia in vitro [81] (see also [104]). This proposed role of mAChR-M1 mediated adaptation to hypoxia is underscored by the neuroprotective effect of hexahydropyrimidine derivatives, which were found to be putative mAChR-M1 ligands, in an in vivo rodent model of hypoxia [105].

Limitations and Perspectives
Although miRNA 132 and miRNA 212 share identical seed sequences, exhibit similar mature sequences and hence considerable overlap in their mRNA targets, the degree of functional redundancy of the mature miRNAs 132 and 212 in OGD and other physiological and pathophysiological contexts has not yet been fully elucidated [16,106]. Thus, although we here employed previously characterized total miR-132/212 double knockout mice [21], the data herein presented cannot provide conclusions on whether the observations reported are carried by the absence of either one of these miRNAs or their joint deletion. Our research therefore encourages further independent exploration of the individual contributions of both miRNA 132 and 212 to the early synaptic response to OGD. Moreover, we here employed a constitutive full-body knockout model in which this miRNA cluster of interest was absent during the entire ontogeny of the animal. Thus, it remains to be explored whether the observed aggravation of synaptic depression in response to OGD is due to a disruption of an acute, adaptive miRNA 132/212 dependent regulatory mechanism, or whether the chronic absence of the miRNA 132/212 cluster has primed the synaptic response prior to the insult.
Other authors [35] have also reported an absence of upregulation of miRNA 132 after 1 h following a presentation of 30 min of OGD as examined in primary dissociated neuronal cultures, whereas this miRNA was upregulated only 24 h after this OGD insult. Therefore, given that a timeframe of <2 h was used here for electrophysiological recordings as pertaining to the OGD protocol, we believe that the methodological limitations inherent to using acute hippocampal slices hamper the possibility to effectively examine the levels of miRNA 132 and miRNA 212 prior to and after OGD. Further research is therefore encouraged using alternative methods, including-for example-the use of organotypic slices or in vivo cerebral ischemia. Moreover, further in vivo experiments will be necessary to elucidate the possibility of an miRNA 132/212 dependent dynamic regulation of the expression of proteins involved in cholinergic signaling over the acute and subacute phases of OGD insults and its long term consequences.
Additionally, compared to the wide range of information available in the scientific literature, here we explored only a very narrow section of putative targets directly and indirectly implicated in miRNA 132/212 dependent signaling. Further research will be therefore necessary to advance towards a mechanistic understanding of how synaptic transmission is influenced by the miRNA 132/212 cluster in response to OGD. Moreover, while the OGD protocol used for this study was optimized to allow for the induction of a sequence of repeated OGD intervals, with intermittent recoveries, this protocol does not model the irreversible loss of synaptic transmission (as in e.g., [64]) and accompanying massive neuronal death as observed in the core of a ischemic lesion [107]. Whereas previous research has also utilized various cell culture-based assays to elaborate on the putative tissue-protective and antiapoptotic role of the miRNA 132/212 cluster in the neuronal response to severe OGD, in the time frame of 12 to 48 h [35,36] we here explored the role of this cluster in the acute response of synaptic transmission to a comparatively mild succession of brief OGD intervals. The long-term consequences of the reported observations, their relevance in the in vivo context of the intact neurovascular unit and whole brain circuitry, and their translational significance remain to be elucidated.
An additional and very important limitation of the work described here pertains to the potential physiological relevance of the sex of the subjects. Here we have exclusively focused on the study of female subjects and, consequently, the scope of our findings must be toned down accordingly to avoid generalizations. We have previously reported that α7-nAChR expression is upregulated in the absence of miRNA 132/212 in the hippocampi of male mice [22]. However, we did not observe an effect of miRNA 132/212 deletion on hippocampal α7-nAChR protein levels in female hippocampi. Given that sex-specific effects on the outcome of ischemic insults have recently been established for the effects of the miRNAs Let7f and mir363-3p [108], it may appear tempting to speculate on sex differences in miRNA 132/212 dependent regulations of cholinergic targets, and their putative role in the context of OGD. However, as our observations were derived from separate paradigms, any conclusions on putative sex-specific miRNA dependent regulations would be premature, as this research question merits a thorough independent investigation in its own right.
Moreover, previous research on the role of the miRNA 132/212 cluster in the modulation of central circadian rhythms unveiled highly strain-and light-regimen-specific effects of miRNA 132/212 deletion when comparing miRNA 132/212 −/− mice bred on C57BL/6N and 129/Sv backgrounds [109]. Independent replication of our findings in miRNA 132/212 −/− mice bread on different background strains might thus help to exclude the possibility of similar strain-specific roles of this cluster in the synaptic response to OGD.
Hippocampal synaptic plasticity events, including Long Term Potentiation (LTP) and Long Term Depression (LTD), have also been extensively investigated as potential mechanisms critical for the formation and regulation of learning and memory functions. We here observed that the miRNA 132/212 gene deletion promotes the vulnerability of short-term synaptic depression to repeated oxygen glucose deprivation. We, however, have not yet explored the susceptibility of LTP and LTD to OGD. As the miRNA 132/212 cluster has previously been shown by others (as well as by our group) to be involved in the regulation of hippocampal synaptic plasticity [21,22], the exploration of the putative involvement of the miRNA 132/212 cluster in the susceptibility of physiological synaptic plasticity to OGD as well as its possible involvement in aberrant forms of LTP (such as ischemic long term potentiation (iLTP) [110]) merits thus further investigations.

Final Conclusions
Here we described for the first time an involvement of the miRNA 132/212 cluster in the acute response of synaptic transmission to repeated ischemia reperfusion insults in the mammalian hippocampus, using an ex vivo electrophysiological approach. While previous research has implicated anti-apoptotic and regenerative effects of the miRNA 132/212 cluster via the silencing of FOXO3 [36] and SOX2 [45] in the hours and days following an OGD insult, our research focused on the early, acute functional effects of miRNA 132/212 dependent signaling on the integrity of synaptic transmission during repeated OGD events. Together with previous observations, focusing on OGD induced miRNA 132/212 regulation and tissue protection in the time frame of hours to days [35,36,45], our findings implicated diverse and pleiotropic miRNA 132/212 mediated adaptations to ischemic and hypoxic challenges.
This proposition might warrant further exploration of the manifold mechanisms regulated by miRNA 132/212 dependent signaling in the temporal course of ischemic challenges and their consequence on molecular signaling and cell viability.
We further explored the miRNA 132/212 dependent regulation of key mediators in the cholinergic signaling underlying both the physiological basis of higher order cognitive functions as well as the adaptive response to ischemia. We propose that the herein observed alterations in AChE and mAChR-M1 expression and aggravation of ischemia induced synaptic dysfunction in the dentate gyrus elicited by miRNA 132/212 deletion might help in the establishment of a link between miRNA 132/212, cholinergic signaling and ischemia associated cognitive impairment. Specifically, we speculate that the previously established link of concomitant miRNA 132 downregulation and AChE upregulation to post-ischemic dementia might be partially mediated by the failure of a putative miRNA 132 mediated mechanism to safeguard the functional integrity of hippocampal cholinergic synaptic signaling from ischemic disruption. Future research, addressing how miRNA 132/212 dependent signaling is involved in short-and long-term synaptic plasticity under OGD in vivo in various brain circuits over time, might help in the exploration of this research question. Our findings furthermore invite the elucidation of whether the expression of key players of cholinergic signaling such as ACh recpetors, or enzymes mediating the functional availability of ACh might be regulated by the miRNA 132/212 cluster over different phases during and after ischemic insults, in the in vivo context. However, as the herein chosen paradigm can only illuminate a narrow aspect of miRNA 132/212 dependent signaling under OGD, any translational considerations must be phrased cautiously. For example, it has been recently shown that miRNA 132/212 overexpression is able to compromise blood-brain barrier integrity via regulation of tight