|Home | About | Journals | Submit | Contact Us | Français|
The ability of insulin like growth factor 1 (IGF-1) to prevent the pathophysiology associated with amyotrophic lateral sclerosis (ALS) is currently being explored with animal models and in clinical trials with patients. Several studies have reported positive effects of IGF-1 in reducing motor neuron death, delaying the onset of motor performance decline, and increasing life span, in SOD-1 mouse models of ALS and in one clinical trial. However, a second clinical trial produced no positive results raising questions about the therapeutic efficacy of IGF-1. To investigate the effect of specific and sustained IGF-1 expression in skeletal muscle or central nervous system on motor performance, life span, and motor neuron survival, human-IGF-1-transgenic mice were crossed with the G93A SOD-1 mutant model of ALS. No significant differences were found in onset of motor performance decline, life span, or motor neuron survival in the spinal cord, between SOD+/IGF-1+ and SOD+/IGF-1- hybrid mice. IGF-1 concentration levels, measured by radioimmunoassay, were found to be highly increased throughout life in the central nervous system (CNS) and skeletal muscle of IGF-1 transgenic hybrid mice. Additionally, increased CNS weight in SOD+ mice crossbred with CNS IGF-1 transgenic mice demonstrates that IGF-1 overexpression is biologically active even after the disease is fully developed. Taken together, these results raise questions concerning the therapeutic value of IGF-1 and indicate that further studies are needed to examine the relationship between methods of IGF-1 administration and its potential therapeutic value.
Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease which is characterized by the selective loss of spinal, cranial, and cortical motor neurons and has a mean survival life-span of 3 to 5 years after diagnosis. Although the mechanism of ALS is still unclear, it is associated with oxidative stress, protein nitration/aggregation, and excitotoxicity (Martin et al., 2007). An estimated 10% of cases are familial and result from genetic mutations (Mulder et al., 1986). A superoxide dismutase-1 (SOD1) gene mutation has been linked to primary causes of disease in familial ALS (Rosen et al., 1993). The over-expression of G93A mutant human SOD-1 (SOD+) in mice is used as model for ALS (Gurney, 1994) (Julien and Kriz, 2006).
Insulin-like growth factor 1 (IGF-1) is a neurotrophic factor which crosses the blood brain barrier and is expressed in the central nervous system (CNS, brain, spinal cord) and skeletal muscle (Lewis et al., 1993). IGF-l's potential to enhance the outgrowth of spinal motor and corticospinal neurons makes it a prospective therapeutic agent in motor neuron disease (D'Costa et al., 1998) (Ozdinler and Macklis, 2006). Studies using adeno-associated viral (AAV) retrograde delivery of IGF-1 to mouse motor neurons in vivo shows that IGF-1 increases life-span and delays disease progression in SOD+ mice (Kaspar et al., 2003). Increased life-span, delayed disease progression, and enhanced motor neuron survival has also been reported in SOD+ mice crossed with transgenic mice over-expressing human IGF-1 in muscle (Dobrowolny et al., 2005).
While these studies in mouse models are encouraging, IGF-1 studies in humans have reported mixed results. In a double-blind, placebo controlled, randomized study supported by Cephalon, and Chiron, and the Muscular Dystrophy Association, the progression of functional impairment in patients receiving high doses (0.10mg/kg/day) of rIGF-1 was reduced by 26% versus patients receiving placebo (Lai et al., 1997). However, a double-blind study supported by Cephalon and Chiron using the same inclusion/exclusion criteria, screening period, and randomization criteria showed no significant difference between patients receiving rIGF-1 and placebo (Borasio et al., 1998). To help address these discrepant results, there currently is a phase III randomized, double-blind, placebo controlled clinical IGF-1 trial underway (clinicaltrials.gov, 2007).
Since human trials have shown varied results, we wanted to further examine the effect of IGF-1 treatment in the G93A ALS mouse model. Our approach of cross breeding SOD+ mice with transgenic mice over-expressing human IGF-1 (IGF-1) in brain or muscle should expose motor neurons to sustained and high IGF-1 concentrations throughout life. IGF-2/1 transgenic mice have high IGF-1 concentrations in the brain, cerebellum, and spinal cord (Ye et al., 1996) (Moreno et al., 2006), whereas S1/S2 transgenic mice express high levels of IGF-1 in skeletal muscle (Coleman et al., 1995; Messi and Delbono, 2003). By using both models, we have explored the effects of CNS-and-muscle derived IGF-1 on SOD-1+ mice. Additionally, we have examined the relationship between disease onset and life span with motor neuron size, motor neuron survival, and IGF-1 expression.
In the present study, we used a transgenic mouse model (S1/S2) expressing human IGF-1 (IGF-1) exclusively in the skeletal muscle (Coleman et al., 1995; Renganathan et al., 1997; Renganathan et al., 1998). These mice were screened for the presence of hIGF-1 genomic DNA via excised mouse tail segments, which were digested overnight at 55 °C in digestion buffer containing 1M Tris-NaOH, pH 8.0, 5 M NaCl, 0.5 M ethylenediamine-tetraacetic acid, 20% sodium dodecyl sulfate and 20 mg/ml proteinase K. From this mixture, DNA was extracted by phenol:chloroform:isoamyl alcohol (25:24:1). IGF-1 gene from this DNA was screened by polymerase chain reaction (PCR) with specific 25-base primers: IGF-1 5': ATT TAA gTg Ctg CTT Ttg TgA TTT C and IGF-1 3' TTC CTA CAT CCT gTA gTT CTT gTT T. Amplified DNA fragments were analyzed for the presence of a clear 450-bp fragment specific for hIGF-1 on 2.5% agarose gels. Wild-type littermate FVB mice (S1/S2-) were used as controls.
The transgenic mice over-expressing hIGF-1 exclusively in the CNS (IGF-2/1 Tg), have been described previously (Ye et al., 1996). IGF2/1 is expressed in the CNS under the control of a 5.7 kb DNA fragment of the 5' mouse IGF-II genomic regulatory region (Dai et al., 1992; Ye et al., 1996). These IGF-1 mice have been backcrossed into a C57BL/6 background for more than 10 generations and are routinely bred as heterozygotes. The IGF2/1 transgene was identified by PCR analysis of genomic DNA using the following primers: 5' end primer: 5'-GGACCGGAGACGCTCTGCGG-3' (bp 179-198) and 3' end primer: 5'-CTGCGGTGGCATGTCACTCT-3' (complementary to bp 518-537).
Amplified DNA fragments were studied for the presence of a clear 360 bp fragment specific for brain IGF-1 on 3.0% agarose gels. Wild-type littermate C57BL/6 mice, named IGF2/1(-), were used as controls.
B6SJL –TgN(SOD1-G93A)1Gur (grey mice) and B6.Cg-Tg(SOD1-G93A)1Gur/J (black mice) were acquired from the Jackson laboratory (Bar Harbor, Maine). We will refer to these SOD mutant mice strains as gSOD(+) and bSOD(+), respectively. Both mice express a high copy number of the mutant allele human SOD1 containing the Gly93 to Ala (G93A) substitution; however, bSOD+ mice live longer than gSOD+ mice, 19-23 weeks compared to 16-20 weeks, and exhibit a more homogenous phenotype (Jackson laboratory, personal communication). This variation in phenotype may be related to different backgrounds. bSOD+ mice have a C57BL/6 background while gSOD+ mice have a B6SJLF1 background. For detection of mutant human SOD-1, DNA was excised and isolated using the same procedure described for S1S2 transgenic mice except the primers used for the PCR were 5'-CAT CAg CCC TAA TCC ATC TgA-3' at 60 °C and 5'-CgC gAC TAA CAA TCA AAg TgA-3' at 59.0°C. The amplified 236bp DNA fragment representing mutant human SOD-1 was detected on 2.5% agarose gels.
For genomic DNA detection procedures in all mice, the interleukin gene was used as an internal control for proper DNA extraction and isolation. The PCR primers used for interleukin were 5'-CTA ggC CAC AgA ATT gAA AgA TCT-3' at 60°C and 5'-gTA ggT ggA AAT TCT AgC ATC ATC C-3' at 61.3 °C. Amplified DNA fragments were analyzed for the presence of a clear 324 bp fragment specific for interleukin on 2.5% agarose gels.
All mice were maintained in a 12:12h light:dark cycle at 22 °C. The colony was housed in a pathogen free facility of the Wake Forest University School of Medicine (WFUSM) Animal Resources Program. Mice exhibiting gross internal or external pathology at macroscopic inspection were not included in the study. Adequate measures were taken to minimize pain or discomfort and experiments were conducted in accordance with international and national standards on animal welfare. All procedures were approved by the WFUSM Animal Care and Use Committee.
Crossbreeding of gSOD and bSOD and their respective background littermates with either muscle (S1/S2) or brain (IGF2/1) IGF-1 transgenic over-expressing mice and their respective wild-type background littermates resulted in multiple combinations as described below. All SOD (+)/S1/S2(−) or SOD(+)/IGF2/1(−) and SOD(−) crossbred with S1/S2(+) or IGF2/1(+) hybrid mice were viable as demonstrated by survival beyond the 180 days reported in this work. The survival of wildtype mice was significantly shortened by crossbreeding with both SOD (+) mouse models as reported below.
IGF-1 concentration in serum, various regions of the CNS (brain, cerebellum, and spinal cord), soleus (SOL) and extensor digitorium longus (EDL) skeletal muscles, and other peripheral organs (heart and liver) was measured by radioimmunoassay (RIA) associated with IGF-1 binding protein blockade as described (Moreno et al., 2006). The RIA kit (Alpco Diagnostics, Windham, NH) used for these determinations, includes a specific, high affinity polyclonal antibody, which has a cross-reactivity with IGF-2 of less than 0.05%. The sensitivity of the assay is 0.02ng/ml.
IGF-1 overexpression results in organ enlargement (Ye et al., 1996). At age 6-7 weeks, mice were sacrificed and their brain, cerebellum, brain stem, and spinal cord blotted and weighed. Systematic anatomical references were used to separate the four CNS regions to make reliable and comparative weight determinations across and within mouse strains.
Motor performance was assessed by the rotarod test using an Economex Rota-Rod with automatic fall detection (Columbus Instruments, Columbus, Ohio). Mice were placed on the rotarod apparatus revolving at either 10 and 18 rpm for a period of 120 seconds. Animals were tested for three trials and the longest duration achieved was computed and used for statistical analysis.
Survival time was measured by recording the date that each animal reached end-stage. Mice unable to right themselves after 30 seconds after being placed on their sides were defined as end-stage (Kaspar et al., 2003). This artificial time point allows mortality to be judged in a more humane and reliable fashion.
Thionin-stained cryostat sections of spinal cord from mice were counted applying histological and morphometric procedures previously described (Li et al., 1994) (Clarke and Oppenheim, 1995) (Gould et al., 2006). The brain stem and spinal cord were cut transversely at 12-15 μm. We counted all motor neurons in the ventral horn in every 10th section throughout the entire lumbar spinal cord. Only those cells with substantial cytoplasm, a nucleus, and containing at least one nucleolus were counted (Clarke and Oppenheim, 1995). Using these rather stringent criteria for cell counting, less than 1% of the cells appear on two successive sections and, therefore, only an insignificant number of cells are counted twice. The counts presented are the raw or actual cell counts multiplied by 10 and then divided by the total number of sections per region to yield cells per section. Because motor neuron atrophy occurs in SOD1 mice, we included both large (>300μm2) and small (<300μm2) neurons within the ventral horn of mice from all genotypes. Even the smallest neurons in the ventral horn were distinct from non-neuronal cells (Fig. 1).
Thionin-stained somas were traced using a camera lucida, scanned, and cross-section areas measured using Scion (Frederick, MD) Image software (Buss et al., 2006). Because cross-sectional area underestimated cell size changes, we used cell volume for all soma size calculations (Buss et al., 2006). Only neurons with a distinct cell membrane and clear nucleus were included in these measurements.
All data are presented as mean ± s.e.m. Data were analyzed with Student's t-test, twoway repeated measures ANOVA, or with a Mann–Whitney rank sum test where appropriate. P values less than 0.01 (**) or less than 0.05 (*) were considered significant. P values greater than 0.05 were considered not significant (NS).
To determine IGF-1's effect on the clinical course of ALS, we conducted motor performance tests using the rotarod. For this analysis, gSOD+ /S1S2- and gSOD+ /S1S2+ mice and controls, gSOD−/S1S2− and gSOD−/S1S2+ mice, were tested at 10 and 18 rpm. Starting at day 107, we recorded the probability to fall in seconds with 120 seconds being the maximum value. We used day 107 as a starting point because it has been reported to be the approximate time of clinical onset for the G93A SOD+ mouse model (Kaspar et al., 2003) (Matsumoto et al., 2006). In both gSOD+ /S1S2− mice and gSOD+ /S1S2, there was a marked decrease in motor performance at day 107 at both 10 and 18 rpm compared with gSOD−/S1S2− and gSOD−/S1S2+ (Fig.1). This indicates that excesss muscle-derived IGF-1 does not alter the course of motor deterioration in the gSOD+ model.
Motor performance was also tested in CNS IGF-1 over-expressing mice; controls included gSOD−/IGF2/1− and gSOD−/IGF-2/1+ Tg. The gSOD+ /IGF2/1− mice showed a marked decrease in motor performance at day 107 at 18 rpm and at day 128 at 10 rpm compared with both gSOD−/IGF2/1− and gSOD−/IGF2/1+. The gSOD+ / IGF−2/1+ Tg mice showed decreased motor performance at day 107 at 18 rpm and at day 114 at 10rpm compared with the same control groups (Fig. 1). These results indicate that excess brain derived IGF-1 does not alter the course of motor performance deficits in the gSOD+ mouse model.
The lack of any significant beneficial effects of specific overexpression of IGF-1 in CNS and muscle on motor performance in the gSOD+ mouse model may be due to intrinsic variability among gSOD+ mice. To examine this, we used the bSOD+ strain that has a more homogenous phenotype (Jackson Labs). For the bSOD+ mice, we started the rotarod test 20 days earlier than with the gSOD+ model (initiated on day 87 vs. day 107). The rationale for this is that bSOD+ mice have not been studied as extensively and the clinical onset of the disease was unknown. In this model, we determined that deficits in motor performance started at day 87 at both 10 and 18 rpm. The bSOD+/ IGF-2/1+ mice showed a decline in motor performance at day 87 at 18 rpm and on day 141 at 10 rpm (Fig. 2). These results are similar to the bSOD+/IGF2/1− mice which showed a decrease in performance at day 87 at 18 rpm and on day 141 at 10 rpm. The bSOD−/ IGF-2/1+ Tg and bSOD−/IGF2/1− mice were used as background controls. These results indicate that IGF-1 overexpression in the CNS does not improve motor performance in bSOD+ mice.
In this experiment, we examined the effects of CNS and muscle IGF-1 overexpression on the survival of bSOD and gSOD mouse models. The average survival time for gSOD+/S1S2−, gSOD+/S1S2, gSOD+/IGF-2/1−, and gSOD+/IGF-2/1+ was 129, 120, 142, and 135 days, respectively. The same groups of bSOD mice were analyzed and their average survival was 153, 148, 155, and 143 days, respectively. While a significant difference was found between gSOD+/S1S2+ and gSOD+/S1S2− and between bSOD+/IGF2/1+ and bSOD+/IGF2/1−, this difference indicates that hybrid IGF-1 transgenic mice actually had a slightly shorter life span than their hybrid wild type counterparts. In addition, no significant differences in life span were detected between gSOD+/IGF2/1+ and gSOD+/IGF2/1− or between bSOD+/S1S2+ and bSOD+/S1S2−. These results indicate that sustained and exclusive over-expression of IGF-1 in neither muscle nor CNS prolongs life-span in the ALS mouse models used here.
To analyze motor neuron death, we examined sections of the mid-lumbar spinal cord in all four models (Fig. 4). The spinal cord showed loss and atrophy of motor neurons in both SOD+/IGF2/1+ mice and SOD+/IGF2/1− mice, indicating that the presence of IGF-1 did not deter disease progression. In contrast, SOD−/IGF2/1− and SOD−/IGF2/1+ both maintained motor neuron integrity. Similar results were recorded in the facial nucleus of the SOD/IGF2/1 hybrid mice (data not shown).
We counted motor neurons in the lumbar spinal cord of bSOD mice crossed with CNS or muscle IGF-1 transgenics (Fig. 5). In the lumbar spinal cord, the mean motor neuron counts for bSOD+/IGF2/1+, bSOD+/IGF2/1−, bSOD−/IGF2/1+, bSOD−/IGF2/1− mice were 1243, 1272, 2760, and 2558, respectively. Although no significant differences were observed between the first two and the last two groups, the SOD+ mice exhibited significantly lower motor neuron numbers than SOD- mice (Fig. 5A). Similarly, mean motor neuron counts for hybrid bSOD+ crossed with muscle IGF-1 Tg-mice, bSOD+/S1S2+, bSOD+/S1S2−, bSOD−/S1S2+, and bSOD−/S1S2− mice were 1348, 1246, 2900, and 3157, respectively (Fig. 5B). Again, no significant differences in motor neuron counts were observed between SOD+/S1S2+ and SOD+/S1S2− mice or between SOD−/S1S2+ and SOD−/S1S2− mice, whereas, the number of motor neurons was significantly lower in SOD+ compared to SOD− mice. These results indicate that localized and sustained IGF-1 overexpression either in the CNS or skeletal muscle does not prevent the loss of lumbar spinal cord motor neurons in SOD+ mice.
ALS has been associated with progressive and severe spinal motor neuron atrophy and cell death (Martin et al., 2007). To examine IGF-1's effect on spinal motor neuron atrophy, we measured and compared spinal cord motor neuron soma size across the same multiple genotypes described above. SOD mutant mice (bSOD+/IGF2/1−) exhibit a decreased number of large-sized motor neurons when compared with control (bSOD−/IGF2/1−) (Fig. 6 A, B). IGF-1 overexpressed in CNS does not modify motor neuron soma size in bSOD−/IGF2/1+ mice (Fig. 6 A, C), and does not prevent the loss of large-sized motor neurons (Fig. 6 A, D). Similarly, Figure 7 shows that SOD+ mice (bSOD+/S1S2−) have lost large motor neurons compared to SOD− mice (bSOD−/S1S2−) (Fig. 7 A, B). IGF-1 expressed in muscle does not change motor neuron soma size in bSOD−/S1S2+ mice (Fig. 7 A, C) and it does not prevent motor neuron atrophy or loss in bSOD−/S1S2− vs. bSOD+/S1S2+ mice(Fig. 7 A, D).
To determine IGF-1 expression in SOD+ mice crossbred with CNS and muscle IGF-1 transgenic mice, we measured IGF-1 concentrations in various tissues by RIA between days 116 and 203. We choose to measure IGF-1 concentrations between days 116 and 203 to verify that IGF-1 was sustained at high levels after the onset of motor deficits. Fig. 8 compares human IGF-1 concentration in brain, cerebellum, and spinal cord in bSOD+ and bSOD− mice crossbred with CNS IGF-1 Tg mice (IGF2/1). IGF-1 was highly expressed in the brain of bSOD+/IGF2/1+ mice. No significant difference was observed between this group and bSOD−/IGF2/1+ mice. Both transgenic mouse models expressed significantly higher IGF-1 levels than bSOD+/IGF2/1− mice. Similarly, in the cerebellum, IGF-1 was highly expressed in bSOD+/IGF2/1+ mice. No significant difference was observed in IGF-1 expression between this group and bSOD−/IGF2/1+ mice. However, these two groups expressed significantly more IGF-1 than bSOD+/IGF2/1− mice. In the spinal cord, IGF-1 was highly expressed in bSOD+/IGF2/1+ mice. While there was a difference in IGF-1 expression between this group and bSOD−/IGF2/1+ mice, both transgenic hybrid mouse models expressed more IGF-1 than the bSOD−/IGF2/1− mice. Figure 9 compares human IGF-1 concentration measured in the extensor digitorum longus (EDL), soleus muscles, and spinal cords of bSOD+ and bSOD− mice crossbred with muscle-derived IGF-1 Tg mice (S1S2). In the EDL muscle, IGF-1 was highly expressed in bSOD+/S1S2+ and bSOD−/S1S2+ mice. Although there was a difference in expression of IGF-1 between these groups, the bSOD+/S1S2+ transgenic mice also produced significantly more IGF-1 than the bSOD+/S1S2− and bSOD−/S1S2+ mice. In the soleus muscle, IGF-1 was highly expressed in bSOD+/S1S2+ and bSOD−/S1S2+. However, no significant difference in IGF-1 expression was found between these groups versus the bSOD−/S1S2− or bSOD+/S1S2− mice. This may result from the low level of baseline IGF-1 expression in slow twitch muscles such as the soleus (Coleman, personal communication). In addition, IGF-1 concentration in the spinal cord of S1S2 mice was measured and no significant differences were found between bSOD+/S1S2− and bSOD+/S1S2+ mice.
We did not detect any significant differences in IGF-1 expression in the heart, liver, and serum, between transgenic and nontransgenic mice (data not shown). These data are consistent with a specific overexpression of IGF-1 in the CNS and muscle.
The lack of effect of IGF-1 overexpression on pathophysiology in ALS mice may result from a reduced effect of the trophic factor in hybrid mice. Therefore, we also examined whether IGF-1 has any other biological effects on the brain, cerebellum, brain stem, or spinal cord. For example, it has been reported that increased IGF-1 expression is associated with increased organ weight due to increased neuron numbers (Ye et al., 1996) (Hodge et al., 2005) (Ozdinler and Macklis, 2006). As shown in Figure 10, brain, cerebellum, brain stem, and spinal cord weights in SOD−/IGF2/1+ and SOD+/IGF2/1+ were significantly increased when compared to SOD−/IGF2/1− and SOD+/IGF2/1− mice. This increase in organ weight in IGF-1 transgenic mice indicates that IGF-1 is in fact biologically active in the transgenic hybrid mice.
Despite the sustained overexpression of IGF-1 in either skeletal muscle or the CNS of mice expressing a mutant form (G93A) of the human SOD-1 gene, these animals were indistinguishable from control SOD-1 mice in several measures of pathophysiology that characterize this mouse model of familial ALS. Deficits in motor performance, decreased life span, motor neuron atrophy and loss of motor neurons were all unaffected by IGF-1 overexpression. This failure to detect any effects of CNS or muscle overexpression of IGF-1 in the SOD-1 mouse model of ALS was surprising in that several previous studies have reported beneficial effects of IGF-1 in this same mouse model.
In one study intramuscular injection of an AAV construct expressing IGF-1 was retrogradely transported to the spinal cord of SOD-1 mice and shown to increase lifespan, improve motor performance and reduce motor neuron atrophy and degeneration (Kaspar et al., 2003). In another study SOD-1 mice that overexpress IGF-1 in skeletal muscle were also reported to have enhanced motor performance, increased life-span and reduced degeneration of motor neurons (Dobrowolny et al., 2005). Finally, continuous intrathecal delivery of exogenous IGF-1 to SOD-1 mice delayed deficits in motor performance, extended life-span and reduced motor neuron loss (Nagano et al., 2005). However, despite the positive effects of IGF-1 in these studies, the benefits were transient and the IGF-1 treated SOD-1 mice in all three studies died prematurely compared to WT mice. Nonetheless, the positive effects observed in these three studies differ significantly from the uniformly lack of effect of IGF-1 in the present study.
Some possible differences between the present study and these three studies that could potentially account for the discrepant results include: genetic background; mode of IGF-1 administration; promoters used for muscle-specific overexpression; differential access of motor neurons to IGF-1 overexpression in muscle or CNS; and lower levels of IGF-1 overexpression in the present study.
Although the muscle IGF-1 mice used here were on a FVB background, the CNS IGF-1 mice were on a C57BL/6 background similar to the SOD-1 mice in the three studies cited above. Our mode of IGF-1 administration also may be a source for differences in that it is possible that continuous exposure to IGF-1 throughout development could alter receptor expression and hence dampen IGF-1's effect. However, muscle IGF-1 transgenic mice used by Dobrowolny et al. (2005) were also exposed to IGF-1 from birth and showed a positive effect from IGF-1. In the study by Dobrowolny et al., (2005), a myosin light chain (mlc) promoter was used to drive skeletal muscle expression of IGF-1, whereas we used an alpha actin skeletal muscle promoter (Dobrowolny et al., 2005). However, both promoters result in muscle-specific overexpression of IGF-1 in primarily fast twitch-muscles and both induce skeletal muscle hypertrophy. Although exogenous IGF-1 injected into skeletal muscle of the chick embryo (Rind and von Bartheld, 2002) and adult rat (Caroni, 1993) is retrogradely transported, the retrograde transport of IGF-1 wasn't specifically examined in our study or that of Dobrowolny and colleagues. In the present study, IGF-1 levels in the spinal cord as measured by RIA were similar in bSOD+/S1S2- and bSOD+/S1S2+ mice. Even in the absence of retrograde transport, however, target-derived neurotrophic factors can induce retrograde signaling cascades (Zweifel et al., 2005). For example, IGF-1 receptors are expressed on motor axons and terminals (Caroni, 1993) and in a previous study we demonstrated that muscle-specific overexpression of IGF-1 can affect motor terminals and innervation (Messi and Delbono, 2003). Additionally, the application of exogenous IGF-1 at the site of peripheral axotomy in the sciatic and facial nerve of neonatal mice and rats also rescues motor neurons from injury induced cell death (Hughes et al., 1993; Li et al., 1994). Finally, intramuscular injection of IGF-1 binding proteins that sequester and reduce the biological activity of endogenous IGF-1 in muscle, perturbs neuromuscular innervation, consistent with a role for endogenous muscle derived IGF-1 on motor neurons (Caroni, 1993; D'Costa et al., 1998; Rind and von Bartheld, 2002). Accordingly, it seems unlikely that our failure to observe any effect of IGF-1 overexpression in muscle on disease progression in SOD-1 mice is due to a total lack of motor neuron access to IGF-1. However, we cannot exclude the possibility that our use of the alpha actin promoter precluded the retrograde transport of IGF-1 and that retrograde transport may be critical for preventing motor neuron deficits in SOD-1 mice. For example, in a related study comparing muscles vs CNS overexpression of GDNF, it was indicated that retrograde transport of GDNF may be critical for neuroprotection in SOD-1 mice (Li et al., 2007); GDNF was neuroprotective following muscle but not CNS overexpression. Although this might also explain why CNS overexpression of IGF-1 was ineffective in the present study, it does not explain why CNS (intrathecal) infusion of IGF-1 was neuroprotective in SOD-1 mice (Nagano et al., 2005) or why in our study, CNS overexpression of IGF-1 resulted in increased growth of brain and spinal cord, a clear index of biological activity in the CNS.
Although we are unable to provide a definitive explanation for our failure (in contrast to previous studies) to detect a neuroprotective effect of IGF-1 in a mouse model of ALS, these negative results, together with a negative results of one of two clinical trials using IGF-1 with ALS patients (Borasio et al., 1998), indicate a note of caution in the future use of IGF-1 to treat motor neuron disease. However, since SOD+ mouse models do not always represent human disease, it is still possible that IGF-1 may have some beneficial effects in humans, the current clinical trial will be crucial to understanding how the route of IGF-1 effects the disease course in humans.
The present study was supported by grants from the National Institutes of Health/National Institute on Aging (AG13934 and AG15820) and the Muscular Dystrophy Association to Osvaldo Delbono, and by the Wake Forest University Claude D. Pepper Older Americans Independence Center (P30-AG21332). Additionally, support was from NIH grant NS 48982, a grant from the Robert Packard Center for ALS Research, and the A. Tab Williams and Family Endowment Fund to R.W. Oppenheim.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.