Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease caused by selective degeneration of motor neurons in the brain and spinal cord. As motor neurons control voluntary movement and respiration, progression of the disease leads to progressive paralysis and eventually respiratory failure and death. ALS has traditionally been classified into two major categories: sporadic and familial. In general, familial ALS has an earlier onset and a slower progression, while sporadic ALS usually appears around 60 years of age (Swinnen and Robberecht, 2014). Over time it has been recognized that cognitive impairment is frequent in ALS and that frontotemporal dementia (FTD) and ALS may represent different outcomes of a common pathogenic mechanism, although the reason why some neurons, typically large spinal and cortical motor neurons, are particularly susceptible remains unclear.

It was previously reported that mutant UBQLN2^P497H mice, which reproduce a mutation found in human ALS/FTD patients, are cognitively impaired (Gorrie et al., 2014). This impairment appears linked to proteasome impairment leading to the formation of aggregates in neuronal cell bodies and dendrites, and most markedly in dendritic spines. In UBQLN2^P497H transgenic mice dendritic spines of cortical neurons are dramatically enlarged, having sizes up to 5 times the normal (Gorrie et al., 2014). Furthermore, tissue staining with ubiquilin2 antibody reveals the presence of numerous ubiquilin2-positive aggregates in all the main hippocampal subfields (CA1, CA3 and dentate gyrus), and hippocampal LTP is hindered in brain slices obtained from 3 month old UBQLN2^P497H mice. Because LTP is the consequence of precisely timed synaptic (excitatory synaptic currents) and intrinsic action potential (AP) activity (Paulsen and Sejnowski, 2000) changes in any of these parameters may underlie the observed impairment. Additionally, altered electrical excitability may affect numerous other functional properties including gene expression (Cohen and Greenberg, 2008). Dysfunctions of neuronal excitability are a constant finding in neurological disease, although the polarity of the changes may vary in different cell types or in different stages of the disease progression. For example, in Alzheimer’s disease, AB oligomers impair synaptic plasticity (Shankar et al., 2008), but intrinsic excitability appears to increase, at least in the early stages (Hall et al., 2015). In Parkinson’s disease, on the other hand, direct and indirect pathway striatal spiny neurons show alterations of intrinsic excitability of opposite polarity, while synaptic connectivity is only affected in indirect pathway neurons (Fieblinger et al., 2014). Thus, although a link between brain dysfunction and neuronal excitability is always present, the type (synaptic vs. intrinsic) and polarity of the alterations are often unpredictable.

The peculiar spinopathy that characterizes UBQLN2^P497H mice strongly suggests that the synaptic responses are affected, yet the presence of either primary or homeostatic changes in intrinsic excitability is also to be considered. Presently, however, nothing is known concerning the electrophysiological properties of hippocampal neurons in these mice at preclinical disease stage. Here we investigated the electrophysiological properties of CA1 pyramidal neurons of 1 month old UBQLN2^P497H mice. This is a particularly important time-point because at this age overt pathology is not yet detectable. We found that, at this age-point, glutamatergic signaling is severely impaired in these neurons due to a large reduction in the size of the AMPA-mediated current. Additionally, intrinsic excitability is also widely impaired. The presence of altered membrane excitability at this early stage of the disease may therefore suggest that such change might have a causal role in the progression of the disease.

Hippocampal brain slices: 29- to 35-day-old WT and transgenic UBQLN2^P497H mice were anesthetized with isoflurane and killed by decapitation. In these animals the pathology is 100% penetrant.

All experiments followed protocols approved by the Northwestern University Center for Comparative Medicine, an Association for Assessment and Accreditation of Laboratory Animal Care accredited facility, and followed guidelines issued by the National Institutes of Health and the Society for Neuroscience.

The brains were quickly removed from the skull in ice-cold low Ca²⁺, high Mg²⁺ artificial cerebrospinal fluid (ACSF), containing (in mM): 87 NaCl, 25 NaHCO₃, 2.5 KCl, 1.25 NaH₂PO₄, 0.5 CaCl₂, 7 MgCl₂, 75 sucrose and 25 glucose, bubbled with 95% O₂ and 5% CO₂ (pH 7.4). Transverse hippocampal slices of 300 μm thickness were cut using a vibroslicer (Leica VT1200) and then stored in a solution containing (in mM): 87 NaCl, 25 NaHCO₃, 2.5 KCl, 1.25 NaH₂PO₄, 0.5 CaCl₂, 7 MgCl₂, 75 sucrose and 25 glucose, bubbled with 95% O₂ and 5% CO₂ (pH 7.4) for 15 min at 35°C and subsequently at room temperature. For the recordings, slices were transferred to a recording chamber bathed in physiological ACSF (in mM: 125 NaCl, 25 NaHCO₃, 2.5 KCl, 1.25 NaH₂PO₃, 25 glucose, 2.0 CaCl₂ and 1 MgCl₂, equilibrated with 95% O₂ and 5% CO₂) at 30–32°C. Slices were visualized using an upright microscope (Scientifica) using an Olympus 60× water-immersion objective and infrared illumination. Pyramidal cells were identified on the basis of shape, location and electrophysiological properties.

Pipettes were pulled from borosilicate glass using a horizontal puller (P97, Sutter, Novato, CA, USA). For the study of synaptic currents pipettes were filled with a CsCl based internal solution consisting of (in mM): CsCl 138, NaCl 2, MgCl₂ 2, EGTA 10, HEPES 10, Na₂ATP 2, NaGTP 0.03, pH 7.3 with CsOH; by blocking voltage gated and leakage potassium currents this solution allows better clamp condition. Furthermore, for these recordings the bath solution contained picrotoxin (0.1 mM) to block GABAergic currents. For the study of intrinsic properties and excitability, no synaptic blockers were added to the extracellular solution and pipettes were filled with an internal solution containing (in mM): KGluconate 140, NaCl 8, MgCl₂ 2, EGTA 0.1, HEPES 10, Na₂ATP 2, NaGTP 0.03, pH 7.4 with KOH. Tip resistances in working solutions ranged from 3 to 8 MΩ yielding series resistances of 10–30 MΩ that were compensated by 75–90%. Recordings were performed using an Axopatch 200B amplifier. The reported membrane potential values are reported without correcting for liquid junction potentials.

Membrane potential was measured in current clamp (without any current injection) right after breaking into whole-cell configuration. Input resistance was calculated by dividing the time constant of the peak voltage response onset, obtained using a 1 s-long current injection (−300 pA) from a membrane potential of −70 mV, by the cell capacitance (obtained in voltage clamp by using the capacitance correction circuit of the amplifier). Synaptic stimulation was performed using a concentric bipolar electrode (FHC) positioned in the stratum radiatum at a distance of 125–225 μm from the soma of each patched neuron. This stimulation electrode was connected to a WPI Stimulus Isolator (A360) current stimulator that was controlled using the digital output in PClamp8. The intensity of the stimulus varied between 50 and 800 μA and was increased in 50 or 100 μA steps. For these recordings cells were voltage-clamped at −70 mV and the bath solution contained 100 μM picrotoxin (Abcam).

We previously found that LTP is impaired in hippocampal slices from 3 month old UBQLN2^P497H mice (Gorrie et al., 2014), an age when cellular pathology is clearly detectable. Here, we performed patch clamp recordings from CA1 hippocampal pyramidal neurons in acute slices obtained from 29 to 35 day-old UBQLN2^P497H mice, an age in which overt pathology is not yet detectable. As a first step, we investigated the strength of the excitatory input to the pyramidal cells. We hypothesized that the glutamatergic current is smaller in CA1 cells from transgenic mice. To this end, cells were voltage clamped at −70 mV and a postsynaptic current response was elicited by extracellular electrical stimulation of the afferent fibers (200 μs; 50–800 μA, in the stratum radiatum). The maximum current amplitude was 642.8 ± 121.6 pA in cells from control mice, while it was only 243.7 ± 72.9 pA in cells from UBQLN2^P497H mice (Figures 1A,B; 10 and 16 cells, respectively; p = 0.01 p = 0.002). The reduced postsynaptic current response may depend on presynaptic impairment, postsynaptic factors or a combination of the two. In order to test possible alterations in presynaptic release, we compared the paired-pulse ratio (PPR) of these synapses in control and UBQLN2^P497H animals. We examined the PPR at 50 and 250 ms time intervals. As previously reported (Chapman et al., 1995), in control animals the synaptic response was strongly facilitated at 50 ms intervals; the magnitude of the facilitation quickly decreased at longer time intervals and it was completely abolished at 500 ms intervals. Interestingly, no difference in PPR was observed between the control and UBQLN2^P497H animals at any of the time intervals investigated (9 and 12 cells, respectively, Figures 1C,D) thus suggesting that presynaptic mechanisms are unlikely to play an important role in the reduced glutamatergic current response. Thus, CA1 pyramidal neurons of 1-month old UBQLN2^P497H mice display an overall decreased synaptic excitation that is the result of smaller macroscopic currents, while no change was detected in presynaptic release as determined using PPR.

Intrinsic properties are as important as the synaptic ones to determine overall neuronal function. Therefore, we also compared the intrinsic excitability in CA1 pyramidal cells of WT and UBQLN2^P497H mice. First, we investigated the basic membrane properties of these cells. No difference was detected in the resting membrane potential (−63.8 ± 0.7 mV and −63.3 ± 1.0 mV, n = 12, 11), capacitance (184.6 ± 7.7 pF and 186.5 ± 17.1 pF n = 12, 13), or input resistance (71.7 ± 3.5 MΩ and 70.2 ± 8.2 MΩ in 12 WT and 13 UBQLN2^P497H cells respectively; Figure 2). Next, we measured the ratio of the peak to steady-state response to an injection of 1 s long hyperpolarizing current steps (Russo et al., 2007) to estimate the size of the hyperpolarization-activated cationic current (Ih). Both genotypes showed only modest voltage sag in response to hyperpolarizing pulses; the sag ratio, however, was slightly but consistently decreased in UBQLN2^P497H mice (Figure 3), suggesting that Ih density was moderately decreased in cells from these mice, although this did not significantly affect the anode break depolarization (Figure 3C). We then measured the firing phenotype of these neurons. Cells were recorded in current clamp and their resting membrane potential was set at –70 mV with injection of negative bias current, when necessary. The input/output function was studied by measuring the voltage responses to injection of square depolarizing current pulses (1 s; 100 pA steps; Figures 4A,B). We found that the maximum firing frequency elicited by saturating current injections was significantly lower in pyramidal cells of UBQLN2^P497H mice compared to control cells (Figure 4C), The instantaneous frequency was also reduced and declined significantly faster with each spike in UBQLN2^P497H compared with control cells (Figure 4D). In keeping with this finding, the fast afterhyperpolarization (fAHP) that follows the first AP in a train was also significantly reduced in UBQLN2^P497H cells over a range of different current injections (in response 400 pA current injections the peak-to-trough voltage change was 104.06 ± 2.9 mV vs. 110.80 ± 2.5 mV in mutant and wild type control animals, respectively (12 cells in each group; p = 0.0496); with 700 pA current injections the peak-to-trough voltage change was 99.28 ± 2.7 mV vs. 108.81 ± 2.0 mV in control, 12 and 11 cells, respectively, p = 0.01, Figure 5A). Moreover, the amplitude of the fAHP in cells from UBQLN2^P497H mice decayed faster within an AP train (the ratio of the first to the last fAHP with 700 pA current injections was 0.64 ± 0.05 and 0.77 ± 0.02, in UBQLN2^P497H and WT n = 12 and 11, respectively; p = 0.03, Figure 5B). The slow afterhyperpolarization (sAHP) at the end of a long spike train was also significantly reduced in cells from UBQLN2^P497H mice, but only following current injections ≥800 pA (Figures 5C,D), a level comparable with those at which differences in firing frequency become apparent (see Figure 5C). This suggests that the differences in sAHP may be attributed by and large to the different numbers of spikes within each train rather than to differences in channel density or gating.

Finally, we examined the properties of the AP upstroke. We found that in cells from UBQLN2^P497H mice the AP threshold was more depolarized compared to wild type (−40.5 ± 0.87 mV vs. −43.6 ± 1.2 mV, n = 14 and 12, respectively, p = 0.03, Figures 6A–C). Consistent with this finding, the maximum dV/dt was also significantly smaller in UBQLN2^P497H mice (442.29 ± 52.2 V*s^-1 vs. 610.0 ± 39.2 V*s^-1, n = 14 and 12 respectively, p = 0.03, Figure 6D). Thus, intrinsic excitability is decreased in a number of relevant parameters in UBQLN2^P497H cells, possibly contributing to the cognitive deficits.

DR: electrophysiological recordings and data analysis; drafting the manuscript. EL: creating the mouse model. H-XD: creating the mouse model; critical discussions of the project. TS: general development of the project; providing financial support; drafting the manuscript. MM: general development of the project; financial support; experimental design, writing the manuscript.

This work was supported by NIH grant NS078504 (TS).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer AM and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.