Fibre size and TTS development
The results in Fig. 1 demonstrate that the geometrical characteristics and Cm are not significantly different in muscle fibres from HSALR than those from age-matched FVB animals. In general, these results concur with previous observations that severe signs of muscle wasting are not detectable in HSALR mice (Mankodi et al. 2000). On the other hand, these data strongly suggest that, for fibres from age-matched mice, comparative differences in the magnitudes of ICl and values of Rm do not result from differences in fibres diameter, or from the relative sarcolemma/TTS membrane areas (Hodgkin & Nakajima, 1972a,b); instead, age-dependent differences between both strains must have arisen from their specific genetic differences.
Our analysis of the relationship between diameter and capacitance confirmed (for mammalian muscle fibres) that the capacitance increase can be readily predicted (Fig. 1D) by theoretical equations (section B in Appendix) previously used for amphibian fibres (Hodgkin & Nakajima, 1972a,b). This is an important result as in our case the populations of fibres used were from mice of various ages (2–6 weeks).
Age dependence of the expression of ClC-1 channels
Our data expand previous reports suggesting that ClC-1 expression is far from complete at birth. The increase in [peak ICl]max up until 6 weeks correlates well with both the maturation of TTS (Franzini-Armstrong, 1991) and the increase in capacitance illustrated in Fig. 1C, but the apparent decline after 14 weeks does not seem to have a close match in other fibre properties.
The data in Figs 2 and 4 demonstrate that the magnitude of ICl in fibres from very young (<1 month old) HSALR mice is ∼26% of that in fibres from control animals; these results confirm (and expand) previous reports from another laboratory (using the same muscle preparation) that suggest serious limitations in ICl in fibres from 9- to 22-day-old and 19-day-old HSALR mice (Lueck et al. 2007a,b). In general, results showing that both [peak-ICl]max and gCl,max increase with age differently in HSALR with respect to control animals (Figs 4 and 5) also demonstrate that the triple repeat intoxication in transgenic mutants has definite effects on the ultimate expression of functional ClC-1 channels. However, this finding is only in partial agreement with a previous report (Lueck et al. 2007a). These authors, while studying ICl in animals spanning a very narrow age range (9–22 days), reported that [peak-ICl]max increased by ∼twofold in control mice, but insignificantly in age-matched HSALR mice. In contrast, our investigations over a broader age range (up 6 months) show that the impairment in the expression of ClC-1 channels clearly ameliorates with age in mutant mice. Furthermore, the data in Figs 4 and 5 illustrate that the rate at which [peak-ICl]max and gCl,max increase with the age of the animals is significantly greater in control than in HSALR animals; this explains why Lueck and collaborators (Lueck et al. 2007a), by evaluating ICl along an insufficient age window, may have mistakenly predicted that the impairment in the expression of ClC-1 channels is sustained at all ages, a proposition in sharp contrast with our demonstration that the expression of functional ClC-1 channels progressively increases with age. Consequently, according to our data, the previous suggestion by Lueck and collaborators (Lueck et al. 2007a) that HSALR mice fail to execute the postnatal splicing transition for ClC-1 should likely be modified to indicate that the splice switching from fetal to postnatal patterns is temporarily delayed, but not prevented, in HSALR animals.
Age dependence of gCl: implications for myotonia
Our results showing that the levels of expression of functional ClC-1 channels, albeit delayed, are not permanently depressed in HSALR mice (as previously proposed) are of vital importance for the correct understanding of myotonia. Namely, our average data on [peak-ICl]max and gCl,max for very young mutant animals (up to ∼6 weeks old; Fig. 4) are readily compatible with the observation of myotonia (Mankodi et al. 2000; Lueck et al. 2007a) according to the assumption that gCl should be as low as ∼20–25% of normal to induce myotonia in normal fibres (Furman & Barchi, 1978). However, this clearly does not hold true for older animals; for example, by ∼14 weeks old, though average [peak-ICl]max was ∼20% smaller in fibres from HSALR than in those from FVB mice, the difference was not statistically significant. These average deficiencies are clearly insufficient to explain myotonia in adult HSALR mice, as claimed by other authors (Mankodi et al. 2002). In this regard, it is important to recall that while these authors correlated myotonia in HSALR mice with the (allegedly age-independent) level of transgene expression, they did report direct measurements of ICl in adult HSALR fibres. Instead, the reduction in ICl was only inferred from RIN measurements at rest (Mankodi et al. 2002). Interestingly, their report of significantly larger RIN in adult HSALR fibres, as compared with controls, is also in disagreement with our data in Fig. 6 and Table 1 (see below).
An obvious question that arises from the above discussion is: how is it possible that myotonia is observed in HSALR animals at ages in which average [peak-ICl]max and gCl,max are only slightly below normal? Two non-mutually exclusive explanations are supported by our data: (A) that conductive pathways, other than gCl, are implicated in the electrical response of the fibres during the genesis of myotonic runs; and (B) that electromyogram detection of myotonia may be biased towards the presence of highly impaired fibres in a population, while not reflecting the average impairment of gCl. Evidence in support of the first possibility is provided in Fig. 10. Namely, two fibres with comparable values of Rm, one from a control animal (Fig. 10A; Rm = 930 Ω cm2) and the other from an adult HSALR mouse (Fig. 10B; Rm = 740 Ω cm2), display very dissimilar electrical responses to identical current pulses. It can be observed that the control fibre shows the typical response for normal animals: a single AP. In contrast, the voltage response in the HSALR fibre is a train of APs, akin to a myotonic run, which is typical for this animal strain. Although the evidence is not conclusive, the data summarize our experience that, albeit the linear resting resistance and gCl may be normal in fibres from adult HSALR animals, alterations of other conductances manifest themselves in non-linear responses that lead to the existence of myotonic runs in fibres from these animals. Evidence in favour of the latter option was provided by analysing frequency distributions of gCl,max in fibres from animals at various ages (Fig. 5). For example, in very young HSALR mice (hatched bars, Fig. 5A), not only was the mean gCl,max value quite small but, most interestingly, unlike in its normal counterpart (filled bars, Fig. 5A), the distribution of gCl,max values around the mean was very narrow. This suggests that, in general, muscles from these animals may represent a rather homogenous population of impaired fibres. In contrast, values of gCl,max in fibres from adult HSALR mice (hatched bars, Fig. 5D) are distributed over a wide range that closely overlaps those from normal mice (filled bars, Fig. 5D); however, an important difference must be noted between these two datasets: the gCl,max distribution in HSALR includes low values that are not present in normal mice, i.e. this is a very heterogeneous distribution. This feature implies that, in principle, a fraction of, but not all, the fibres somehow evade the toxic effects of the triplet repeats. In conclusion, the presence of fibres with low gCl,max may explain the detection of myotonia in muscles from HSALR animals, while the average gCl,max is not statistically distinguishable from the normal value. It should be noted that the above explanation is compatible with published immunohistochemistry data showing that muscle cross-sections challenged with anti-ClC-1 antibodies show a mosaic pattern of ClC-1 expression in adult HSALR mice (Mankodi et al. 2002; Wheeler et al. 2007).
Figure 10. Fibres from adult HSALR mice are more excitable than from age-matched FBV mice The upper traces in A and B are voltage records acquired from FDB fibres of control and HSALR mice, respectively. The lower traces are the current pulses (50 nA, 300 ms). The records were obtained in Tyrode. The ages of the mice were 20 weeks and 17 weeks for A and B, respectively.
Download figure to PowerPoint
Measurements of the linear electrical properties RIN, Rm and Rm*
An important methodological precaution in our RIN measurements (and the determination of Rm thereafter) is that they were made with very small current injections (<20 nA), eliciting small voltage perturbations (ΔV typically <6 mV), in order to critically assess the linear electrical properties of FDB fibres subject to their physiological resting potential (approximately −90 mV). In other words, the experimental conditions were designed to minimize the influence of intrinsic non-linearities of either ClC-1 and/or other channels that contribute to their resting conductance. Interestingly, our Rm values for fibres from FVB adult animals (668 ± 49 Ω cm2; Fig. 7B and Table 1) are in excellent general agreement with average values reported for a variety of mammalian skeletal muscle preparations; for example, average values of 642, 445 and 542 Ω cm2 have been reported for mouse extensor digitorum longus (EDL), rat diaphragm and rat EDL, respectively (Kiyohara & Sato, 1967; Farnbach & Barchi, 1977; Palade & Barchi, 1977a; Kerr & Sperelakis, 1983). But these linear Rm values are distinctly smaller than the value of 1040 ± 100 Ω cm2 reported for intercostal fibres from >2 month old FVB mice (Mankodi et al. 2002); though we do not fully understand the origin of this discrepancy, these authors’ observations were made using larger current injections that resulted in voltage perturbations possibly out of the linear range. Furthermore, our Rm values in fibres from adult HSALR mice are also significantly smaller than those reported by Mankodi and collaborators (783 ± 81 vs. 4080 ± 590 Ω cm2). It is important to highlight that the relatively large values of Rm obtained by us in fibres rendered electrically passive (Tyrode with 9-ACA + Rb) demonstrate that, in adult fibres under resting (physiological) conditions, Rm was mostly determined by contributions by K and Cl conductances rather than by unspecific residual conductances (from Table 1, ∼3.5- and 4.3-fold changes were found in fibres from control and HSALR animals, respectively; also see Fig. 11). Furthermore, the similarity of Rm between fibres from adult FVB and HSALR is compatible with our voltage-clamp result that values of gCl,max in these animals are not significantly different.
Figure 11. Conductance contributions to the resting gm in fibres from FVB and HSALR mice The proportional contributions of gCl, gK,IR and gres to gm were calculated from Rm values obtained in fibres sequentially bathed in Tyrode, Tyrode + Rb, Tyrode + 9-ACA and Tyrode + Rb + 9-ACA. The specific experiments for these calculations are a subset of those used to calculate Rm values reported in Table 1 and Fig. 7. The percentage contributions of gCl, gK,IR and gres were calculated by algebraically solving the system of linear equations using Rm data under the four experimental conditions. A and B, the data from young and adult animals, respectively. The filled and hatched columns correspond to data from control and HSALR animals, respectively. The error bars represent the SEM. The asterisks show significance at P < 0.05.
Download figure to PowerPoint
In contrast with the observations discussed above, Rm measurements in fibres from young (4 weeks old) HSALR are significantly larger (P < 0.05) than those from age-matched FVB mice (Fig. 7A and Table 1). This was a reasonable result in the light of the [peak-ICl]max and gCl,max data under voltage-clamp conditions if we accept the notion that Rm may be affected principally by ClC-1 channels (Bretag, 1987). Nevertheless, we must qualify this remark by comparing the actual Rm values in fibres from both strains. As expected, the average values of Rm in fibres rendered electrically passive are not significantly different in fibres from FVB and HSALR mice (4463 and 3735 Ω cm2, respectively; Table 1); this reinforces the idea that, other than differences in the basal conductance, the two fibre populations are almost identical. But Rm values in fibres from young HSALR animals are surprisingly small: only 32–33% larger than those of their control counterparts (Table 1), compared with the more than threefold differences in values of [peak-ICl]max and gCl,max (Fig. 4). A reasonable explanation for this important disparity is that, in fibres from these animals, conductive pathways other than gCl (e.g. the inward rectifier gK,IR conductance) play important compensatory roles in determining their relatively low linear Rm. In order to assess this option, we measured Rm in fibres treated with 9-ACA (to block gCl) and Rb (to block gK,IR) and, from these data (not shown), calculated the contributions of each conductive pathway to the total resting conductance (gm = 1/Rm, in mS cm−2). Figure 11 shows that we can account for gm with contributions from gCl and gK,IR, and from an unidentified residual conductance (gres). The black columns in Fig. 11A show that, in control young mice (3–4 weeks old; n = 7), the resting gm (∼1.1 mS cm−2; see Table 1) was contributed principally by gCl (∼53%), and less prominently (∼27%) by gK,IR. As expected from the significant reduction in gCl,max and [peak-ICl]max (Fig. 4), the ∼16% contribution of gCl to the resting conductance in fibres of age-matched HSALR mice (hatched columns, Fig. 11A, n = 6) is also significantly smaller than in control fibres. Notably, Fig. 11A also demonstrates that the reduction in gCl is associated with an exaggerated contribution of gK,IR (∼51%) to the slightly reduced overall gm (∼0.83 mS cm−2). Thus, if we focus only on the blockable conductances, the data in Fig. 11A show that the proportion of gK,IR (η), defined as η = gK,IR/(gCl + gK,IR), is ∼34% in fibres from young control animals and ∼76% in age-matched HSALR animals. To our knowledge this is the first direct demonstration of the compensatory role gK,IR may play in fibres with significantly reduced gCl. Furthermore, we believe that this result will be critically important for the understanding (on a quantitative basis) of the underlying causes of myotonia in various animal models, and in humans. Unlike the situation in young animals, Fig. 11B shows that the contributions of gCl (and/or gK,IR) to the resting gm are not significantly different between fibres of age-matched FVB and HSALR mice. Nevertheless, albeit not significant, there is hint of the reduced contribution by gCl (compensated by gK,IR) that was observed in young HSALR mice. For example, in this case η is 42% and 65% in fibres of adult control and HSALR animals, respectively. Also, in agreement with the larger variability in gCl,max observed in frequency bar graphs (Fig. 5D), the proportional contributions of gCl and gK,IR in fibres of HSALR animals is notably larger than in control fibres. This excess variability, probably arising from a mosaic impairment of the expression of ClC-1 channels, is undoubtedly a contributing factor for the lack of statistical significance between data from adult mice of both strains.
In adult normal muscle fibres, the ∼45% contribution of gCl to the resting conductance (gm) shown in Fig. 11B may seem too low in comparison to values of up to 85% suggested in the literature (Kerr & Sperelakis, 1983; Kwiecinski et al. 1984; Rudel & Lehmann-Horn, 1985; Bretag, 1987; De Luca et al. 2003). It should be noted that, in our case, the gCl contribution is pondered in absolute terms with respect to the total resting conductance of the cell, and goes beyond relative comparisons from changes in resistances due to ion replacements. Thus, our calculations take into account the fact that there is a finite residual conductance (gres) after gCl and gK,IR are blocked. If we compute the percentage contributions taking into account only these readily blockable conductances, as done frequently in the literature, the results are 58% and 42% for gCl and gK,IR, respectively; these values are in line with published values from other laboratories for rat muscle fibres (Bretag, 1987). Nonetheless, our data for normal mouse short muscle fibres (FDB) show that, at a resting potential of −90 mV, gCl represents significantly less than 80–85% of the resting conductance as suggested previously for long fibres (Kerr & Sperelakis, 1983; De Luca et al. 2003).
Our calculations of Rm*, based on realistic theoretical assumptions about the equivalent circuit of the muscle fibres as described in the Appendix, provide important information about the ‘true’ specific resistance of the surface and TTS membranes regardless of their area contribution. As expected, for every age group and fibre type, Rm* is approximately four–fivefold larger than Rm; this closely matches the TTS/sarcolemma area ratios, as predicted from theoretical considerations (Adrian et al. 1969a; Hodgkin & Nakajima, 1972a,b). Furthermore, Rm* values for all fibres tested under passive conditions (most conductances blocked) are indistinguishable from each other. We must also highlight that the comparison of Rm* values between animal types at different ages reinforces all the conclusions reached previously while discussing Rm data; namely, fibres from adult FVB and HSALR have similar values of Rm*, but those from young HSALR specimens have slightly (but significantly) larger values of Rm* than their control counterparts.
Is the HSALR mouse a good model for DM1?
The HSALR mouse has been a helpful model to demonstrate the toxicity of CUG expansions on the splicing of pre-mRNA of the ClC-1 channel (Mankodi et al. 2000). Although several limitations of this animal as a model of DM1 have already been recognized (Gomes-Pereira et al. 2011), the transitory impairment in the expression of ClC-1 channels, as demonstrated in this paper, has been overlooked so far. In this sense, our results have important implications, not only for the interpretation of the mechanisms of myotonia in the HSALR mouse per se, but also for the use of this animal in future therapeutic genetic trials, and as a model for the chloride channelopathy in humans. Nevertheless, our results comparing the linear electrical properties of fibres from normal and HSALR fibres, while suggesting that hyperexcitability may not exclusively result from a sarcolemma-restricted chloride channelopathy but from the non-linear behaviour of ClC-1 and/or other ion channels (especially in the TTS membranes), point out complex aspects of the mechanisms of myotonia that can be optimally studied in the HSALR mouse due to the transient nature of the impairment in the functional ClC-1 expression. In other words, new avenues of research towards the quantitative understanding of the ionic basis of myotonia, which stem from this paper, still must be undertaken in fibres from this animal model.