**3.1 The change in input resistance**

158 Amyotrophic Lateral Sclerosis

motoneurons exhibit an increase in the overall cell size and branching pattern of dendrites (Amendola et al., 2007; Amendola and Durand, 2008). These changes appear in transgenic mice at 10 days after birth (P10), long before disease onset and life end-stage (225 and 240 days, respectively) (Bruijn et al., 1997). More specifically, in mutant motoneurons the total dendritic surface area and total dendritic length are increased by 58% and 65% relative to wild-type (WT) cells, respectively. For dendritic branching, the number of dendritic branches, terminal branches, branching nodes, and maximum branching order are increased by 93%, 89%, 97%, 37%, respectively (Fig. 1). On the other hand, the soma size (diameter and surface area) and primary dendrites properties (number, diameter, and cross-sectional area) are similar in WT and mutant motoneurons. Interestingly, the increase in number of branches occurs primarily in short branches (dendritic branches with path length < 100µm) and most of the anatomical changes take place over the middle dendrites (200µm - 500µm from the soma). Surprisingly, the longest dendritic length from the soma was not increased in mutant motoneurons, indicating that dendritic overbranching occurs mainly within the cell circumjacent. In other words, the cell does not swell into larger space in the cord but develops more branching within its dendritic

Fig. 1. Anatomical alterations in mutant motoneurons of the G85R (reconstructed from the whole-cord prep1) and G93A (high expressor line) (reconstructed from the slice prep.) transgenic models relative to WT cells. 1: data from Amendola and Durand (2008).

Similar early abnormalities in the anatomical properties of motoneurons have been found in other transgenic mouse models of ALS. In the G93A (low expressor line), which expresses 8 copies of the human SOD1 gene, mutant motoneurons at P9 exhibit increased cell size, dendritic overbranching, and dendritic complexity (Filipchuk et al., 2010). The total dendritic length and number of branching nodes and terminal dendrites in mutant motoneurons were significantly increased relative to WT. Similar to the G85R transgenic model, these anatomical changes in the G93A (low expressor line) model occur early in the disease (at P9) way before symptom development (between 195 and 240 days) and end stage (between 240 and 270 days). Preliminary data from our laboratory obtained in the

spatial field.

Changes in a number of motoneuron electrical properties were reported in the transgenic mouse models of ALS. However, one challenge in the field of ALS is the large variability in data of mutant motoneuron electrical properties, which makes recording from a large number of cells is important to allow statistical tests to detect significant differences in electrophysiological properties between WT and mutant motoneuron samples. Table 1 summarizes the changes in electrical properties of mutant motoneurons observed in the various ALS transgenic mouse models. The summary shows that the changes in electrical properties involve the motoneuron passive properties (e.g., cell input resistance), the size and shape of the action potential spike and afterhyperpolarization, the amplitude of voltagesensitive ionic currents, and the motoneuron excitability and firing activity. However, some of these changes were inconsistent (i.e., not observed consistently) across the ALS transgenic models, whereas others were contradictory (i.e., opposite changes were observed). For instance, the decrease in input resistance of mutant motoneurons relative to WT is an inconsistent finding in ALS transgenic models (see table 1, column 7). This observation was seen in some studies (Bories et al., 2007; ElBasiouny et al., 2010; Quinlan et al., 2011), whereas other studies did not find a statistically significant difference between the input resistance of WT and mutant motoneurons (Pieri et al., 2003; Kuo et al., 2004; Kuo et al., 2005; van Zundert et al., 2008; Pambo-Pambo et al., 2009; Pieri et al., 2009; Filipchuk et al., 2010; Meehan et al., 2010). The changes in action potential size and width in mutant motoneurons, on the other hand, is a conflicting finding in which some studies reported an increase in spike height (van Zundert et al., 2008) and duration (Fuchs et al., 2008, 2009; Pambo-Pambo et al., 2009) relative to WT, whereas others reported reduction in spike height (Kuo et al., 2004; Pambo-Pambo et al., 2009; Filipchuk et al., 2010) and duration (Pieri et al., 2003; Quinlan et al., 2011).


Electrophysiological Abnormalities in SOD1 Transgenic

Models in Amyotrophic Lateral Sclerosis: The Commonalities and Differences 161

Table 1. Summary of the alterations in mutant motoneuron properties. SMNs: spinal motoneurons, CNs: cortical neurons, HGMNs: hypoglossal motoneurons, OMNs: ocular motoneurons, OINs: ocular interneurons. [Ca2+]o: extracellular calcium concentration, Rin: input resistance, Vrest: resting membrane potential, AP: action potential, AHP: afterhyperpolarization, PIC: persistent inward current, F-I: frequency-current relationship. VC: voltage-clamp,

motoneuron recruitment and de-recruitment on a trian

ΔI: the difference in injected current at

gular current command, No diff: no difference.


Electrophysiological Abnormalities in SOD1 Transgenic

action potential and afterhyperpolarization properties.

**Model Publication Preparation [Ca2+]o**

Quinlan et al.

Bories et al.

Pambo-Pambo et

Pambo-Pambo et

Filipchuk et al.

depending on the tissue preparation used in the experiment.

**3.2 The changes in action potential and afterhyperpolarization** 

symptomatic adult hypoglossal motoneurons (Fuchs et al., 2009).

G93A (high)

G85R

G93A (low)

Models in Amyotrophic Lateral Sclerosis: The Commonalities and Differences 163

between 2mM and 4mM, see table 1) among the various ALS studies could contribute to the discrepancies in electrical properties of mutant motoneurons. This effect is produced because the level of extracellular Ca2+ concentration modulates the magnitude of the motoneuronal Ca2+-gated (e.g., Cav 1.2, 1.3, and 2.2 channels) and Ca2+-activated (e.g., SK and Ih channels) ionic currents, which regulate the motoneuron firing activity and affect the

(2011) Slice 2mM 38.5/31.3 M<sup>Ω</sup>

(2007) Whole-cord 4mM 33.3/23.8 M<sup>Ω</sup>

al. (2009) Slice 2mM 34.5/33.1 M<sup>Ω</sup>

al. (2009) Slice 2mM 34.5/37.2 M<sup>Ω</sup>

(2010) Whole-cord 2mM 18.5/16.1 M<sup>Ω</sup>

Table 2. Comparison of the change in input resistance in mutant motoneurons (P6-P10) in the various ALS transgenic models. 1: unpublished data, 2: Amendola and Durand (2008), 3: Filipchuk et al. (2010), \*: statistically-significant (p<0.05), ns: not statistically-significant.

In sum, the reduction in input resistance of mutant motoneurons is a common, early pathological feature in the G93A (high expressor line) and G85R transgenic models, but not in the G93A (low expressor line). Also, this abnormality might sometimes not be evident

A change in the action potential properties was frequently seen in mutant motoneurons relative to WT; however, these changes were conflicting across the transgenic models (Table 1, Fig. 2). In the G93A (high expressor line) model, neonatal spinal motoneurons frequently showed faster action potential rate of rise and decay and shorter action potential duration (Pieri et al., 2003; Quinlan et al., 2011), indicating an increase in transient and persistent Na+ currents, which act to increase the excitability of mutant motoneurons. Signs of increased excitability were differently displayed in other studies as an increase in spike height (van Zundert et al., 2008), depolarization in resting membrane potential (Kuo et al., 2005), or reduction in action potential threshold (Pieri et al., 2009). Conversely, an increase in action potential duration and deceleration in rate of repolarization were observed in post-

Yes2 ElBasiouny et al. (2010) Whole-cord 2mM 16.2/11.4 M<sup>Ω</sup>

**Rin**

**WT/SOD1** 

(↓ by 29%)\*

(↓ by 30%)\*

ns

ns

ns

(↓ by 19%)\* Yes1

**Anatomical alterations** 

Yes3

### **3.1.1 Effect of tissue preparation**

The discrepancy in the changes of electrical properties of mutant motoneurons probably results partially from the differences in transgenic model types (e.g., G85R, G93A high or low expressor lines) and is compounded by the variability in experimental conditions such as animal age (e.g., neonatal, pre- or post-symptomatic adult motoneurons) and tissue preparation (e.g., cell culture, slice, whole-cord, or in-vivo), in addition to the recording conditions (e.g., extracellular Ca2+ concentrations) and measurement methods (current- or voltage-clamp). However, critical analysis of the changes in electrical properties of mutant motoneurons shows that common pathological features could be identified across the various ALS transgenic models, whereas other features could be related to experimental conditions. For instance, the decrease in input resistance was detected mainly in studies conducted on the whole-cord preparation (in which the brainstem and spinal cord are intact), but was not observed in studies on motoneurons in cell culture or slice preparations (Table 1). Because input resistance is an indirect measure of cell size, it becomes reduced in mutant motoneurons to reflect the enlargement in motoneuron anatomy discussed in the previous section. In the whole-cord preparation, the motoneuron dendrites are completely intact and their effect on input resistance is more easily detected. In the slice preparation, conversely, the motoneuron dendrites are truncated at the surface of the slice. The lack of the middle and distal portions of dendrites could mask the anatomical differences between WT and mutant motoneurons and would explain the disappearance of input resistance differences between WT and mutant motoneurons in the slice preparation. Table 2 shows a comparison of the input resistance values of neonatal motoneurons in the various transgenic models of ALS. In the G85R and G93A (low expressor line) models, the input resistance values in the whole-cord preparation (ElBasiouny et al., 2010; Filipchuk et al., 2010) were on average half of those in the slice preparation (Pambo-Pambo et al., 2009), indicating that nearly half of the dendrites were severed in the slice preparation. Strikingly, the decrease in input resistance of mutant motoneurons of the G93A (high expressor line) model was still detected in the slice preparation in Quinlan et al. (2011). This could indicate that the motoneuron anatomical alterations in the G93A (high expressor line) model are so extensive that they were still substantial in the slice preparation and/or the motoneuron membrane biophysical properties change considerably in the G93A (high expressor line) model and contribute significantly to the decrease in input resistance (see section 4). For cell culture, given that the trigger signal for the alteration in motoneuron anatomy is poorly understood and could result from the alteration in motoneuron excitability and synaptic inputs (Duch et al., 2008), mutant motoneurons in the cell culture preparation might not experience anatomical alterations due to their isolation (in the cell culture) from the pathological neural circuit (in the transgenic mouse spinal cord), leading to absence of differences in input resistance between WT and mutant motoneurons.

### **3.1.2 Effect of extracellular Ca2+ concentration**

The measurement of input resistance values, and other electrical properties, in mutant motoneurons also depends on the extracellular concentration of Ca2+ ions in the recording solution. For instance, the input resistance value of neonatal motoneurons of the G85R model measured in 4mM of extracellular Ca2+ concentration (Bories et al., 2007) was double of that measured in 2mM of extracellular Ca2+ concentration (ElBasiouny et al., 2010) (see table 2). Although 2mM is typical, the variability in extracellular Ca2+ concentration (ranging

The discrepancy in the changes of electrical properties of mutant motoneurons probably results partially from the differences in transgenic model types (e.g., G85R, G93A high or low expressor lines) and is compounded by the variability in experimental conditions such as animal age (e.g., neonatal, pre- or post-symptomatic adult motoneurons) and tissue preparation (e.g., cell culture, slice, whole-cord, or in-vivo), in addition to the recording conditions (e.g., extracellular Ca2+ concentrations) and measurement methods (current- or voltage-clamp). However, critical analysis of the changes in electrical properties of mutant motoneurons shows that common pathological features could be identified across the various ALS transgenic models, whereas other features could be related to experimental conditions. For instance, the decrease in input resistance was detected mainly in studies conducted on the whole-cord preparation (in which the brainstem and spinal cord are intact), but was not observed in studies on motoneurons in cell culture or slice preparations (Table 1). Because input resistance is an indirect measure of cell size, it becomes reduced in mutant motoneurons to reflect the enlargement in motoneuron anatomy discussed in the previous section. In the whole-cord preparation, the motoneuron dendrites are completely intact and their effect on input resistance is more easily detected. In the slice preparation, conversely, the motoneuron dendrites are truncated at the surface of the slice. The lack of the middle and distal portions of dendrites could mask the anatomical differences between WT and mutant motoneurons and would explain the disappearance of input resistance differences between WT and mutant motoneurons in the slice preparation. Table 2 shows a comparison of the input resistance values of neonatal motoneurons in the various transgenic models of ALS. In the G85R and G93A (low expressor line) models, the input resistance values in the whole-cord preparation (ElBasiouny et al., 2010; Filipchuk et al., 2010) were on average half of those in the slice preparation (Pambo-Pambo et al., 2009), indicating that nearly half of the dendrites were severed in the slice preparation. Strikingly, the decrease in input resistance of mutant motoneurons of the G93A (high expressor line) model was still detected in the slice preparation in Quinlan et al. (2011). This could indicate that the motoneuron anatomical alterations in the G93A (high expressor line) model are so extensive that they were still substantial in the slice preparation and/or the motoneuron membrane biophysical properties change considerably in the G93A (high expressor line) model and contribute significantly to the decrease in input resistance (see section 4). For cell culture, given that the trigger signal for the alteration in motoneuron anatomy is poorly understood and could result from the alteration in motoneuron excitability and synaptic inputs (Duch et al., 2008), mutant motoneurons in the cell culture preparation might not experience anatomical alterations due to their isolation (in the cell culture) from the pathological neural circuit (in the transgenic mouse spinal cord), leading to absence of differences in input

**3.1.1 Effect of tissue preparation** 

resistance between WT and mutant motoneurons.

**3.1.2 Effect of extracellular Ca2+ concentration** 

The measurement of input resistance values, and other electrical properties, in mutant motoneurons also depends on the extracellular concentration of Ca2+ ions in the recording solution. For instance, the input resistance value of neonatal motoneurons of the G85R model measured in 4mM of extracellular Ca2+ concentration (Bories et al., 2007) was double of that measured in 2mM of extracellular Ca2+ concentration (ElBasiouny et al., 2010) (see table 2). Although 2mM is typical, the variability in extracellular Ca2+ concentration (ranging between 2mM and 4mM, see table 1) among the various ALS studies could contribute to the discrepancies in electrical properties of mutant motoneurons. This effect is produced because the level of extracellular Ca2+ concentration modulates the magnitude of the motoneuronal Ca2+-gated (e.g., Cav 1.2, 1.3, and 2.2 channels) and Ca2+-activated (e.g., SK and Ih channels) ionic currents, which regulate the motoneuron firing activity and affect the action potential and afterhyperpolarization properties.


Table 2. Comparison of the change in input resistance in mutant motoneurons (P6-P10) in the various ALS transgenic models. 1: unpublished data, 2: Amendola and Durand (2008), 3: Filipchuk et al. (2010), \*: statistically-significant (p<0.05), ns: not statistically-significant.

In sum, the reduction in input resistance of mutant motoneurons is a common, early pathological feature in the G93A (high expressor line) and G85R transgenic models, but not in the G93A (low expressor line). Also, this abnormality might sometimes not be evident depending on the tissue preparation used in the experiment.
