SCD1 expression is altered in ALS muscle
On the basis of our previous microarray data, obtained from a transgenic mouse model of mutant SOD1-linked familial ALS [11], in this study we investigated the significance of the down-regulation of SCD1 for the metabolic capacity of muscles and their response to injury. The expression of SCD1 in the gastrocnemius of SOD1(G86R) mice, which are affected by a progressive denervation atrophy [19], was already diminished at 60 days of age. In this respect, it is noteworthy to mention that our previous electromyography studies on this mouse line revealed that the amplitudes of the compound muscle action potentials, a reduction of which typically reflects a decrease in the number of functional motor units, were normal at the age of 75 days. In addition, mice did not present at this age any abnormal spontaneous electrical activity, which would have reflected the common response of muscle to loss of innervation [16]. According to these findings, we can conclude that SCD1 down-regulation occurred precociously in our SOD1(G86R) mouse model. We then showed here that the decrease in SCD1 expression also persisted during the course of the disease, at 90 days of age, when muscle denervation becomes detectable and motor deficits usually arise, and at about 105 days of age, when hind legs start to be paralysed. At that moment, the decrease in SCD1 expression was also noticeable in the tibialis anterior, which is another muscle in the mouse hind leg displaying less oxidative metabolism than the gastrocnemius (Figure 1A). As a consequence of the repression of muscle SCD1 expression, we observed that the C18:1/C18:0 fatty acid ratio, an index of the desaturation activity of the enzyme [20], was slightly reduced in presymptomatic muscle extracts but significantly diminished at the end stage in both gastrocnemius and tibialis anterior (Figure 1B). It is noteworthy to mention that our previous studies had shown that SOD1(G86R) mice typically exhibit decreased postprandial lipidemia and increased peripheral clearance of lipids, both of which can be ascribed to muscle hypermetabolism [10]. Therefore, an excess of uptake of exogenous lipids in this tissue could mask otherwise earlier and more robust differences in the index of SCD activity.
10.1371/journal.pone.0064525.g001 Figure 1  SCD1 expression and activity in ALS mouse muscle.
(A) Time course of SCD1 expression in gastrocnemius (GT, brown columns) and tibialis anterior (TA, orange columns) from SOD1(G86R) mice at indicated ages. Wild-type expression is represented by 100% baseline. ***P<0.001 (One sample t-test, n = 5–11). (B) C18:1/C18:0 fatty acid ratio in gastrocnemius and tibialis anterior from SOD1(G86R) mice (brown columns) and wild-type littermates (white columns) at indicated ages. *P<0.05 (1-way ANOVA followed by Bonferroni's multiple comparison test for gastrocnemius, and unpaired t-test for tibialis anterior, n = 3–10).   To obtain independent evidence that SCD1 down-regulation is a typical feature of ALS, we took advantage of our transcriptome database composed of deltoid biopsies from patients with the sporadic form of the disease [12]. The expression of not only SCD1 but also SCD5, a primate-specific enzyme variant with identical function [21], was lower in ALS patients, as compared to normal control subjects. Furthermore, the repression of SCD1 expression was much more remarkable in a muscle not clinically or electromyography affected than in a muscle at an advanced stage of pathology, characterized at the clinical level by reduced strength and neurogenic electromyography pattern (Figure 2A). That SCD1 down-regulation could be observed both in presymptomatic SOD1(G86R) mouse muscle and in relatively healthy human ALS muscle prompted us to speculate that such a pattern of expression might not be solely related to the loss of muscle innervation characteristic of the disease. To address this question, we compared SCD1 expression in gastrocnemius submitted to acute denervation, as obtained by cutting and removing several millimeters of the sciatic nerve, or transient denervation followed by re-innervation, as obtained by crushing the sciatic nerve for several seconds. Under these conditions, the expression of SCD1 was increased after axotomy but significantly reduced after crush (Figure 2B). Overall, these findings provide evidence for the implication of SCD1 in the pathological process triggering ALS, and suggest that SCD1 down-regulation could be involved in the restoration of muscle function in response to injury.
10.1371/journal.pone.0064525.g002 Figure 2  SCD1 expression in ALS patient muscle and after nerve injury.
(A) Expression of SCD1 and SCD5 in deltoid muscle biopsies from ALS patients and healthy subjects (CT, white columns), as identified by microarray analysis of the database deposited at http://www.ebi.ac.uk/arrayexpress/(accession number E-MEXP-3260) [12]. ALS samples were obtained from muscle not clinically or electromyography affected (Unaff, orange columns) and from muscle with advanced pathology, characterized by reduced strength and neurogenic electromyography pattern (Aff, brown columns). *P<0.05 (1-way ANOVA followed by Tukey's multiple comparison test, n = 4–10). (B) Expression of SCD1 in gastrocnemius following sciatic nerve axotomy (Axo) or crush at indicated post-operation days. Contralateral muscle expression is represented by 100% baseline. **P<0.01, ***P<0.001 (One sample t-test, n = 4–10).

S