**2.2 Vertebrate models**

Projecting the human genome sequence into our understanding of human health and disease has been a new challenge faced in the post-genomic era. As described above, unicellular and invertebrate model systems can be of great value in defining the molecular components of pathways or processes that depend on the function of several interacting proteins. However, vertebrate models offer the best opportunity for defining landmarks of disease progression and for understanding the functional consequences of gene mutations. The great advantage of using vertebrates to model human diseases is clearly the possibility they offer for evaluation of new treatments. Indeed, testing new drugs on mice or other vertebrate models is often mandatory and asked by the Food and Drug Administration, for safety reasons, prior to approve new drug treatments. This step is also fundamental for paving the way towards human clinical trials, with both larger and smaller vertebrate model organisms.

Other advantages of using vertebrate models include: the ability to make efficient targeted gene knockouts by homologous recombination, they are evolutionary closer to human, the developmental overview is similar for all mammals, the availability of material at all stages of development, their brains are more similar to human, they respond to injury and can be conditioned (great advantages to study learning, neuronal connectivity and plasticity) and they provide a valuable source of primary cells for culture. The later feature will be discussed in more details in the next section of this book chapter.

On the contrary, the high maintenance cost, a relatively slow life cycle development, lower number of progeny, genetically identical offspring more difficult to obtain and the difficulty to manipulate embryos (intrauterine development) represent the major disadvantages of using vertebrate as disease model organisms.

### **2.2.1 Mice**

Over the past century, the laboratory mouse (*Mus musculus*) has become the premier mammalian model organism for experimental studies and genetic research. Scientists from a wide range of biomedical fields have used the mouse because of its close genetic and physiological similarities to humans, as well as the ease with which its genome can be manipulated and analyzed, ease of handling and it's relatively high reproductive rate. Although yeasts, worms and flies are excellent models for studying many developmental processes, mice are far better models for studying diverse physiological systems such as the immune system, the endocrine system, the nervous system, the cardiovascular system, the skeletal system and other complex physiological systems that mammals share. Like humans and many other mammals, mice naturally develop diseases that affect these systems, including cancer, atherosclerosis, hypertension, diabetes, osteoporosis and glaucoma. In addition, manipulating the mouse genome and environment can induce certain diseases that

that ALS-associated FUS/TLS mutations are toxics (Walker, et al., 2011). Through crossrescue analysis, it has been demonstrated that FUS acted together with and downstream of TDP-43 in a common genetic pathway in neurons. Furthermore, the authors found that these proteins associated with each other in an RNA-dependent complex. These results establish that FUS and TDP-43 function together *in vivo* and suggest that molecular pathways requiring the combined activities of both of these proteins may be disrupted in

Projecting the human genome sequence into our understanding of human health and disease has been a new challenge faced in the post-genomic era. As described above, unicellular and invertebrate model systems can be of great value in defining the molecular components of pathways or processes that depend on the function of several interacting proteins. However, vertebrate models offer the best opportunity for defining landmarks of disease progression and for understanding the functional consequences of gene mutations. The great advantage of using vertebrates to model human diseases is clearly the possibility they offer for evaluation of new treatments. Indeed, testing new drugs on mice or other vertebrate models is often mandatory and asked by the Food and Drug Administration, for safety reasons, prior to approve new drug treatments. This step is also fundamental for paving the way towards human clinical trials, with both larger and smaller vertebrate model

Other advantages of using vertebrate models include: the ability to make efficient targeted gene knockouts by homologous recombination, they are evolutionary closer to human, the developmental overview is similar for all mammals, the availability of material at all stages of development, their brains are more similar to human, they respond to injury and can be conditioned (great advantages to study learning, neuronal connectivity and plasticity) and they provide a valuable source of primary cells for culture. The later feature will be

On the contrary, the high maintenance cost, a relatively slow life cycle development, lower number of progeny, genetically identical offspring more difficult to obtain and the difficulty to manipulate embryos (intrauterine development) represent the major disadvantages of

Over the past century, the laboratory mouse (*Mus musculus*) has become the premier mammalian model organism for experimental studies and genetic research. Scientists from a wide range of biomedical fields have used the mouse because of its close genetic and physiological similarities to humans, as well as the ease with which its genome can be manipulated and analyzed, ease of handling and it's relatively high reproductive rate. Although yeasts, worms and flies are excellent models for studying many developmental processes, mice are far better models for studying diverse physiological systems such as the immune system, the endocrine system, the nervous system, the cardiovascular system, the skeletal system and other complex physiological systems that mammals share. Like humans and many other mammals, mice naturally develop diseases that affect these systems, including cancer, atherosclerosis, hypertension, diabetes, osteoporosis and glaucoma. In addition, manipulating the mouse genome and environment can induce certain diseases that

discussed in more details in the next section of this book chapter.

using vertebrate as disease model organisms.

ALS and FTD.

organisms.

**2.2.1 Mice** 

**2.2 Vertebrate models** 

afflict humans but normally do not strike mice, such as cystic fibrosis, obesity, blindness, anxiety, aggressive behavior, alcoholism, drug addiction and neurodegenerative disorders such as Alzheimer's disease, Parkinson's disease, Huntington's disease and ALS. Immunodeficient mice can also be used as hosts to facilitate cancer and AIDS research.

Researchers have used a collection of innovative genetic technologies to produce custommade mouse models for a broad array of specific diseases, as well as to study the function of targeted genes. One of the most important technological advances has been the ability to produce transgenic mice, in which a new gene (cloned from human or other various species, wild-type or mutated) is inserted into the mouse's germline. Indeed, the 2007 Nobel Prize in Physiology and Medicine was awarded to Drs Mario R. Capecchi, Martin J. Evans and Oliver Smithies for their discoveries of principles for introducing specific gene modifications in mice by the use of embryonic stem cells. Even more potent approaches, it is now possible to knockout or to artificially drive the expression of an inserted gene in specific tissue or at various time during development or adulthood using the Cre/Loxp system (Sauer, 1998). There are many mouse models commercially available for genetic research including thousands of unique inbred strains and genetically engineered mutants.

Researches using the laboratory mouse led to major advances in our ability to treat a number of serious diseases and conditions. Genetically, mice are more closely related to humans than invertebrates in the sense that most human genes have functional mouse counterparts and the mouse genome is organized in a very similar manner to the human genome. The mouse and human genomes are approximately the same size. They most likely contain the same number of genes and show extensive synteny and conserved gene order. Importantly, mice have also genes that are often not represented in other models, such as *C. elegans* and *Drosophila*. All these advantages make the laboratory mouse the model organism of choice to study human diseases. Although mice are widely used in research, questions remain however about their reliability as a model for human diseases. In occurrence, how to explain that many drugs worked well during preclinical trials in mice, but turned out to be ineffective when used in clinical trials on humans? Here, we will review the mouse studies that contributed toward understanding the pathogenic pathways of motor neuron disease and the testing of therapeutic approaches.

#### **2.2.1.1 Mice SOD1 models**

The use of mouse models has been of particular importance in studying the pathogenesis of Amyotrophic Lateral Sclerosis. The initial description of *SOD1* gene mutations in familial ALS patients, in 1993, first led to the hypothesis that the disease resulted from compromised enzymatic activity due to the loss of the enzyme function (Rosen, et al., 1993). However, this loss of function hypothesis was rapidly refuted, as *Sod1* KO mice in which the murine *Sod1* gene was disrupted, do not develop disease (Reaume, et al., 1996). To date, over 150 different *SOD1* missense mutations have been reported. In contrast, transgenic mice ubiquitously overexpressing various *SOD1* gene mutations with different biochemical properties, even in the presence of endogenous mouse *Sod1* gene, develop a neurodegenerative disease that is quite similar to the human illness. Of particular interest, transgenic mice overexpressing WT human SOD1 or specifically expressing mutant SOD1 only in neurons or only in glial cells do not develop disease (Bruijn, et al., 1998; Gong, et al., 2000; Pramatarova, et al., 2001). A toxic gain-of-function rather than a loss-of-function of mutant *SOD1* gene is therefore believe to be involved in ALS-linked SOD1 patients. All published mouse models have used the endogenous murine *Sod1* promoter resulting in high

*In Vivo* and *In Vitro* Models to Study Amyotrophic Lateral Sclerosis 95

in ALS patients, Golgi fragmentation and neuroinflammation noted by the presence of reactive astroglia and microglia in diseased tissue are also observed. However, the severity of the phenotype (in term of age of onset, disease progression and disease duration) varies from one model to another and may be dependant on mouse genetic background (Heiman-Patterson, et al., 2005) and gene dosage. In deed, the life span of these ALS mice is inversely proportional to gene dosage, i.e. the number of transgene insertion within the mouse genome, leading to different steady state level of mutant SOD1 protein in the central nervous system. Such variation in the steady state protein levels must reflect different

The G93A-SOD1 mouse line is currently the most widely used experimental model in ALS research and drug testing. Unfortunately, several pharmacological approaches tested so far have produced only modest beneficial effects. Riluzole, a glutamate antagonist, extended the life span of G93A-SOD1 mice by 10 to 15 days without affecting disease onset (Gurney, et al., 1996). Today, riluzole is the only drug currently approved for ALS treatment. This treatment only produces modest beneficial effect in some ALS patients. Interestingly, is has been shown that treatment with the ER stress-protective agent salubrinal attenuated disease manifestations and delayed progression in a G93A-SOD1 mouse model (Saxena, et al., 2006). This result suggests a role of ER stress in ALS. Furthermore, it has been also shown that immunization therapy, using specific anti-misfolded SOD1 monoclonal antibodies, succeeded in reducing the level of mutant SOD1 by 23% in the spinal cord of immunized animals, in delaying disease onset and in prolonging the lifespan of G93A-SOD1 mice in proportion to the duration of treatment (Gros-Louis, et al., 2010). These results suggest that accumulation of misfolded SOD1 species is toxic, and reducing the burden of these toxic species leads to beneficial effect. Interestingly, it has been recently reported that WT SOD1 can acquire properties of ALS-linked mutant SOD1 species possibly implying a shared pathophysiological pathway between SALS and FALS (Bosco, et al., 2010b; Ezzi, et al., 2007). Another study also demonstrated that WT SOD1 may acquire toxic properties upon oxidative damage and that WT SOD1 expression dramatically exacerbated disease in transgenic mice expressing mutant SOD1 forms such as A4V, G85R, L126Z, and G93A SOD1 mutants (Wang, et al., 2009b). It is noteworthy that overexpression of WT SOD1 conferred ALS disease to unaffected A4V SOD1 mice. Based on these results, the possibility that WT

SOD1 may be a contributor of pathogenesis in sporadic ALS must be considered.

As previously mentioned, several research groups have reported that dominant mutations in the *TARDBP* gene, which encodes for TDP-43, cause ALS (Kabashi, et al., 2008; Sreedharan, et al., 2008; Van Deerlin, et al., 2008). Following these initials studies, a number of TDP-43 ALS mouse model has been described. Embryonic lethality is observed in homozygous mouse knockouts for TDP-43 (Kraemer, et al., 2010; Sephton, et al., 2010; Wu, et al., 2010). The TDP-43 deficient embryos die at embryonic day 7.5 thereby demonstrating the essential function of TDP-43 protein in development. Mice heterozygous for TDP-43 disruption only exhibit subtle muscle weakness with no evidence of motor neuron

Many of the transgenic mouse lines overexpressing WT or mutant TDP-43 reported to date have showed some ALS features including early paralysis leading to premature death (Stallings, et al., 2010; Wegorzewska, et al., 2009; Wils, et al., 2010). These pan-neuronal TDP-43 transgenic mouse models expressing high-level of either WT or mutant (A315T and

**2.2.1.2 Mice TDP-43 models** 

pathology.

stabilities and degradation of the various human SOD1 mutants.

levels of expression of the mutant transgene in all tissues. Although these models implicate mutant SOD1 in the development of motor neuron degeneration, many questions regarding the mechanism of pathogenesis remain unanswered. One of the central mysteries in ALS research is why an ubiquitously expressed gene such as *SOD1* causes selective devastation to motor neurons in the absence of pathology in other tissues. One possible explanation is that expression of the mutant protein in motor neurons may not be sufficient to lead to the development of a neurodegenerative disease in mice, suggesting that mutant *SOD1* expression in other cells may be necessary for the development of the disease. Although there are indications that neurodegeneration in ALS may not strictly result from a cellautonomous process, the selective motor neuron vulnerability observed in the disease pathogenesis remains enigmatic.

The favored hypothesis at this time of SOD1-mediated disease is that toxicity of SOD1 mutants is related to the misfolding and aggregation of SOD1 species (Gros-Louis, et al., 2009b). However, it is not clear which conformational SOD1 species and oligomers cause ALS and the exact mechanism of toxicity of the misfolded SOD1 species remains unknown. Deleterious effects could result from the interaction of misfolded SOD1 species with essential cellular components such as Bcl-2 (Pasinelli, et al., 2004), from their recruitment to outer membrane of mitochondria or from overwhelming the capacity of the protein folding chaperones and/or of ubiquitin proteasome pathway to degrade important cellular regulatory factors (Turner, et al., 2005; Urushitani, et al., 2008; Urushitani, et al., 2002). The misfolded SOD1 protein may also form aggregates that might sequester important cellular components causing cytotoxicity. Recent studies demonstrated that a fraction of SOD1 could be translocated via the ER-Golgi network and that chromogranins, which are abundant proteins in motor neurons, interneurons and activated astrocytes, may act as chaperone-like proteins to promote secretion of misfolded SOD1 mutants (Urushitani, et al., 2008; Urushitani, et al., 2006). Moreover, it has been reported that extracellular mutant SOD1 can induce microgliosis and motor neuron death (Urushitani, et al., 2006). Such ALS pathogenic mechanism based on toxicity of secreted SOD1 mutant is in line with findings that disease is not strictly autonomous to motor neurons and that toxicity can propagate from one cell to another. Interestingly, it has been proposed that endoplasmic reticulum (ER) stress response might exert a critical role in the disease pathogenesis. The ER is the site of synthesis and folding of secretory and membrane bound proteins. The capacity of the ER to process proteins is limited and the accumulation of misfolded proteins may activate different ER stress pathways. The challenges in the future are to find out which somatic insults are causing the initial protein change, and to discover ways of preventing the misfolded proteins from spreading through the nervous system.

The overexpression of either G37R, G85R, G86R, D90A, G93A, H46R/H48Q or H46R/H48Q/H63G/H120G, L126Z and G127X mutant SOD1 protein in mice leads to motor neuron degeneration (Borchelt, et al., 1994; Bruijn, et al., 1997; Gurney, et al., 1994; Jonsson, et al., 2004; Jonsson, et al., 2006; Ripps, et al., 1995; Tu, et al., 1996; Wang, et al., 2003; Wang, et al., 2002; Wong, et al., 1995). In all of these mouse models, massive death of motor neurons in the ventral horn of the spinal cord and loss of myelinated axons in ventral motor roots ultimately leads to paralysis and muscle atrophy. Histopathological findings observed in these transgenic animals include progressive accumulation of detergent-resistant aggregates containing SOD1 and ubiquitin, aberrant neurofilament accumulations in degenerating motor neurons. In addition to neuronal degeneration, similar to that observed

levels of expression of the mutant transgene in all tissues. Although these models implicate mutant SOD1 in the development of motor neuron degeneration, many questions regarding the mechanism of pathogenesis remain unanswered. One of the central mysteries in ALS research is why an ubiquitously expressed gene such as *SOD1* causes selective devastation to motor neurons in the absence of pathology in other tissues. One possible explanation is that expression of the mutant protein in motor neurons may not be sufficient to lead to the development of a neurodegenerative disease in mice, suggesting that mutant *SOD1* expression in other cells may be necessary for the development of the disease. Although there are indications that neurodegeneration in ALS may not strictly result from a cellautonomous process, the selective motor neuron vulnerability observed in the disease

The favored hypothesis at this time of SOD1-mediated disease is that toxicity of SOD1 mutants is related to the misfolding and aggregation of SOD1 species (Gros-Louis, et al., 2009b). However, it is not clear which conformational SOD1 species and oligomers cause ALS and the exact mechanism of toxicity of the misfolded SOD1 species remains unknown. Deleterious effects could result from the interaction of misfolded SOD1 species with essential cellular components such as Bcl-2 (Pasinelli, et al., 2004), from their recruitment to outer membrane of mitochondria or from overwhelming the capacity of the protein folding chaperones and/or of ubiquitin proteasome pathway to degrade important cellular regulatory factors (Turner, et al., 2005; Urushitani, et al., 2008; Urushitani, et al., 2002). The misfolded SOD1 protein may also form aggregates that might sequester important cellular components causing cytotoxicity. Recent studies demonstrated that a fraction of SOD1 could be translocated via the ER-Golgi network and that chromogranins, which are abundant proteins in motor neurons, interneurons and activated astrocytes, may act as chaperone-like proteins to promote secretion of misfolded SOD1 mutants (Urushitani, et al., 2008; Urushitani, et al., 2006). Moreover, it has been reported that extracellular mutant SOD1 can induce microgliosis and motor neuron death (Urushitani, et al., 2006). Such ALS pathogenic mechanism based on toxicity of secreted SOD1 mutant is in line with findings that disease is not strictly autonomous to motor neurons and that toxicity can propagate from one cell to another. Interestingly, it has been proposed that endoplasmic reticulum (ER) stress response might exert a critical role in the disease pathogenesis. The ER is the site of synthesis and folding of secretory and membrane bound proteins. The capacity of the ER to process proteins is limited and the accumulation of misfolded proteins may activate different ER stress pathways. The challenges in the future are to find out which somatic insults are causing the initial protein change, and to discover ways of preventing the misfolded

The overexpression of either G37R, G85R, G86R, D90A, G93A, H46R/H48Q or H46R/H48Q/H63G/H120G, L126Z and G127X mutant SOD1 protein in mice leads to motor neuron degeneration (Borchelt, et al., 1994; Bruijn, et al., 1997; Gurney, et al., 1994; Jonsson, et al., 2004; Jonsson, et al., 2006; Ripps, et al., 1995; Tu, et al., 1996; Wang, et al., 2003; Wang, et al., 2002; Wong, et al., 1995). In all of these mouse models, massive death of motor neurons in the ventral horn of the spinal cord and loss of myelinated axons in ventral motor roots ultimately leads to paralysis and muscle atrophy. Histopathological findings observed in these transgenic animals include progressive accumulation of detergent-resistant aggregates containing SOD1 and ubiquitin, aberrant neurofilament accumulations in degenerating motor neurons. In addition to neuronal degeneration, similar to that observed

pathogenesis remains enigmatic.

proteins from spreading through the nervous system.

in ALS patients, Golgi fragmentation and neuroinflammation noted by the presence of reactive astroglia and microglia in diseased tissue are also observed. However, the severity of the phenotype (in term of age of onset, disease progression and disease duration) varies from one model to another and may be dependant on mouse genetic background (Heiman-Patterson, et al., 2005) and gene dosage. In deed, the life span of these ALS mice is inversely proportional to gene dosage, i.e. the number of transgene insertion within the mouse genome, leading to different steady state level of mutant SOD1 protein in the central nervous system. Such variation in the steady state protein levels must reflect different stabilities and degradation of the various human SOD1 mutants.

The G93A-SOD1 mouse line is currently the most widely used experimental model in ALS research and drug testing. Unfortunately, several pharmacological approaches tested so far have produced only modest beneficial effects. Riluzole, a glutamate antagonist, extended the life span of G93A-SOD1 mice by 10 to 15 days without affecting disease onset (Gurney, et al., 1996). Today, riluzole is the only drug currently approved for ALS treatment. This treatment only produces modest beneficial effect in some ALS patients. Interestingly, is has been shown that treatment with the ER stress-protective agent salubrinal attenuated disease manifestations and delayed progression in a G93A-SOD1 mouse model (Saxena, et al., 2006). This result suggests a role of ER stress in ALS. Furthermore, it has been also shown that immunization therapy, using specific anti-misfolded SOD1 monoclonal antibodies, succeeded in reducing the level of mutant SOD1 by 23% in the spinal cord of immunized animals, in delaying disease onset and in prolonging the lifespan of G93A-SOD1 mice in proportion to the duration of treatment (Gros-Louis, et al., 2010). These results suggest that accumulation of misfolded SOD1 species is toxic, and reducing the burden of these toxic species leads to beneficial effect. Interestingly, it has been recently reported that WT SOD1 can acquire properties of ALS-linked mutant SOD1 species possibly implying a shared pathophysiological pathway between SALS and FALS (Bosco, et al., 2010b; Ezzi, et al., 2007). Another study also demonstrated that WT SOD1 may acquire toxic properties upon oxidative damage and that WT SOD1 expression dramatically exacerbated disease in transgenic mice expressing mutant SOD1 forms such as A4V, G85R, L126Z, and G93A SOD1 mutants (Wang, et al., 2009b). It is noteworthy that overexpression of WT SOD1 conferred ALS disease to unaffected A4V SOD1 mice. Based on these results, the possibility that WT SOD1 may be a contributor of pathogenesis in sporadic ALS must be considered.

### **2.2.1.2 Mice TDP-43 models**

As previously mentioned, several research groups have reported that dominant mutations in the *TARDBP* gene, which encodes for TDP-43, cause ALS (Kabashi, et al., 2008; Sreedharan, et al., 2008; Van Deerlin, et al., 2008). Following these initials studies, a number of TDP-43 ALS mouse model has been described. Embryonic lethality is observed in homozygous mouse knockouts for TDP-43 (Kraemer, et al., 2010; Sephton, et al., 2010; Wu, et al., 2010). The TDP-43 deficient embryos die at embryonic day 7.5 thereby demonstrating the essential function of TDP-43 protein in development. Mice heterozygous for TDP-43 disruption only exhibit subtle muscle weakness with no evidence of motor neuron pathology.

Many of the transgenic mouse lines overexpressing WT or mutant TDP-43 reported to date have showed some ALS features including early paralysis leading to premature death (Stallings, et al., 2010; Wegorzewska, et al., 2009; Wils, et al., 2010). These pan-neuronal TDP-43 transgenic mouse models expressing high-level of either WT or mutant (A315T and

*In Vivo* and *In Vitro* Models to Study Amyotrophic Lateral Sclerosis 97

dependent loss of motor coordination revealed by rotarod and grip strength performances of the Als2 KO mice, no major motor deficits consistent with ALS or other motor neuron diseases were present in these models. All Als2-deficient mice appear to be grossly normal, viable and fertile with lifespan expectancy similar to WT littermates. However, some differences are noteworthy and may explain the heterogeneity of the phenotype, ranging from ALS to less severe PLS or IAHSP. The diversity of the apparent phenotypes among different Als2 KO mouse models may be due in part by different gene targeting strategies used to generate each mice models, ES cell lines used leading to differences in the genetic

Interestingly, thorough molecular analysis of one of these *Als2* KO mouse model revealed the presence of a number of novel *Als2* isoforms expressed in the central nervous system of these animals (Gros-Louis, et al., 2008). These results suggest that other alternatively spliced *Als2* isoforms may exist and that some of these novel *Als2* mRNA species still can be transcribed in *Als2* null animals and may compensate for the loss of the full-length protein.

Different other transgenic mice models overexpressing different human proteins either ubiquitously or specifically in neurons have been generated (for review see (Gros-Louis, et al., 2006; Julien and Kriz, 2006). These mice acquire age-dependent central nervous system pathology similar to other related motor neuron diseases such as FTDP17, progressive supranuclear palsy, spinal muscular atrophy or hereditary spastic paraplegia. Interesting pathological features, which are also associated with ALS, can be seen in these different mouse models including axonal degeneration in brain and spinal cord, progressive motor disturbance, behavioral impairment, and the presence of cellular aggregates and

As a model of human disease, the laboratory rat (*Rattus norvegicus*) offers some advantages over the mouse and other organisms. The main advantage is in fact that rats are physiologically more similar to humans compare to mice. The size of the animal also confers a valuable advantage and enhances its use as a disease model. The later advantage is especially true when performing surgical procedures, microdialysis, intravenous cannulation and for pharmacodynamic drug studies as its size enables serial blood sampling. Thereby, the rat would be a better model than the mouse to study cardiovascular disease, hypertension, diabetes, arthritis, and many autoimmune, behavioral, and addiction disorders. The rat is also a good model in neurobehavioral and stereotaxic neurological studies. The size of the rat brain offers unique possibilities for the application of microsurgical techniques, intrathecal administration of drugs, stem cell transplantation, serial sampling of the cerebrospinal fluid

However, even though its size is considered as an advantage over mice, the higher cost of maintenance (bigger cages, food, less animals can be housed per cages) and limited housing capacity in animal facilities directly related to its size also confers the principal limitation of

Transgenic rat models of ALS have also been generated. In these models overexpression of G93A or H46R mutant SOD1 led to an ALS-like phenotype (Howland, et al., 2002; Nagai, et

background, housing conditions or approaches taken to evaluate the mice.

**2.2.1.5 Mice models associated with other motor neuron diseases** 

(CSF), *in vivo* nerve recordings, and neuroimaging procedures.

intermediate filament inclusions.

**2.2.2 Rat** 

using rats as a model. **2.2.2.1 Rat SOD1 models** 

M337V) TDP-43 transgene in neurons both showed aggressive paralysis accompanied by increased ubiquitination in specific neuronal populations (Stallings, et al., 2010; Wegorzewska, et al., 2009; Wils, et al., 2010; Xu, et al., 2010). However, many concerns have been raised regarding the biological validity of these models, as they do not completely mimics disease. Since TDP-43 is a moderately and ubiquitously expressed protein, restricted non-physiological neuronal expression of TDP-43 is one of them, along with the use of the TDP-43 cDNA as transgene and the lack of cytoplasmic ubiquitinated TDP-43 inclusions. In order to better mimic the ubiquitous and moderate levels of TDP-43 expression seen in humans, another transgenic mouse models have been generated allowing the expression of genomic TDP-43 fragments (Swarup, et al., 2011). This TDP-43 mouse model exhibited agerelated phenotypic defects as seen in the human condition including both cognitive and motor deficits. Other pathological features as seen in human patients were also observed including, cytoplasmic TDP-43-positive ubiquitinated inclusions, intermediate filament abnormalities, axonopathy and neuroinflammation. These phenotypes were more severe in the transgenic TDP-43 mutant (G348C and A315T) models than the transgenic TDP-43 WT model.

### **2.2.1.3 Mice intermediate filament models**

Several transgenic mouse lines and knockout mice implicating different neurofilament subunits have been extensively studied over the past years (for complete review, see: (Julien and Kriz, 2006; Lariviere and Julien, 2004)). Even though genetic mutations in intermediate filament (IF) genes are not major causes of ALS, it is of potential relevance to ALS that transgenic mice with altered stoichiometry of neuronal intermediate filament develop pathological features of the disease (Beaulieu, et al., 2000; Beaulieu and Julien, 2003; Cote, et al., 1993; Millecamps, et al., 2006). Of particular interest was the finding that overexpression of WT peripherin, a type III intermediate filament, in *NEFL* null background mice caused age-dependant selective motor neurons degeneration (Beaulieu, et al., 1999). This mouse model is also characterized by the formation of perikaryal and axonal intermediate filament inclusions resembling spheroids in motor neurons of human ALS. The precise mechanism by which accumulation of intermediate neurofilament leads to neurodegenerative disorders is not fully understood. Neurofilament and peripherin proteins are two types of intermediate filaments detected in the majority of axonal inclusion bodies, called spheroids, in motor neurons of ALS patients (Corbo and Hays, 1992). Multiple factors can potentially cause the accumulation of intermediate filament proteins including deregulation of intermediate filament protein synthesis, proteolysis, defective axonal transport, abnormal phosphorylation, and other protein modifications.

### **2.2.1.4 Mice ALS2 models**

Truncating mutations were discovered in coding exons of a the *ALS2* gene encoding for Alsin, from patients with an autosomal recessive form of juvenile ALS, primary lateral sclerosis (PLS) and infantile-onset ascending hereditary spastic paralysis (IAHSP) (Devon, et al., 2003; Eymard-Pierre, et al., 2002; Eymard-Pierre, et al., 2006; Gros-Louis, et al., 2003; Hadano, et al., 2001; Kress, et al., 2005; Yang, et al., 2001). The pattern of inheritance and the nature of the mutations identified in this gene suggest that motor neuron degeneration seen in patients results from a loss of protein function. Six different groups have reported the generation of an alsin knockout mouse (Cai, et al., 2005; Deng, et al., 2007; Devon, et al., 2003; Gros-Louis, et al., 2008; Hadano, et al., 2006; Yamanaka, et al., 2006). Despite an agedependent loss of motor coordination revealed by rotarod and grip strength performances of the Als2 KO mice, no major motor deficits consistent with ALS or other motor neuron diseases were present in these models. All Als2-deficient mice appear to be grossly normal, viable and fertile with lifespan expectancy similar to WT littermates. However, some differences are noteworthy and may explain the heterogeneity of the phenotype, ranging from ALS to less severe PLS or IAHSP. The diversity of the apparent phenotypes among different Als2 KO mouse models may be due in part by different gene targeting strategies used to generate each mice models, ES cell lines used leading to differences in the genetic background, housing conditions or approaches taken to evaluate the mice.

Interestingly, thorough molecular analysis of one of these *Als2* KO mouse model revealed the presence of a number of novel *Als2* isoforms expressed in the central nervous system of these animals (Gros-Louis, et al., 2008). These results suggest that other alternatively spliced *Als2* isoforms may exist and that some of these novel *Als2* mRNA species still can be transcribed in *Als2* null animals and may compensate for the loss of the full-length protein.

### **2.2.1.5 Mice models associated with other motor neuron diseases**

Different other transgenic mice models overexpressing different human proteins either ubiquitously or specifically in neurons have been generated (for review see (Gros-Louis, et al., 2006; Julien and Kriz, 2006). These mice acquire age-dependent central nervous system pathology similar to other related motor neuron diseases such as FTDP17, progressive supranuclear palsy, spinal muscular atrophy or hereditary spastic paraplegia. Interesting pathological features, which are also associated with ALS, can be seen in these different mouse models including axonal degeneration in brain and spinal cord, progressive motor disturbance, behavioral impairment, and the presence of cellular aggregates and intermediate filament inclusions.

### **2.2.2 Rat**

96 Amyotrophic Lateral Sclerosis

M337V) TDP-43 transgene in neurons both showed aggressive paralysis accompanied by increased ubiquitination in specific neuronal populations (Stallings, et al., 2010; Wegorzewska, et al., 2009; Wils, et al., 2010; Xu, et al., 2010). However, many concerns have been raised regarding the biological validity of these models, as they do not completely mimics disease. Since TDP-43 is a moderately and ubiquitously expressed protein, restricted non-physiological neuronal expression of TDP-43 is one of them, along with the use of the TDP-43 cDNA as transgene and the lack of cytoplasmic ubiquitinated TDP-43 inclusions. In order to better mimic the ubiquitous and moderate levels of TDP-43 expression seen in humans, another transgenic mouse models have been generated allowing the expression of genomic TDP-43 fragments (Swarup, et al., 2011). This TDP-43 mouse model exhibited agerelated phenotypic defects as seen in the human condition including both cognitive and motor deficits. Other pathological features as seen in human patients were also observed including, cytoplasmic TDP-43-positive ubiquitinated inclusions, intermediate filament abnormalities, axonopathy and neuroinflammation. These phenotypes were more severe in the transgenic TDP-43 mutant (G348C and A315T) models than the transgenic TDP-43 WT

Several transgenic mouse lines and knockout mice implicating different neurofilament subunits have been extensively studied over the past years (for complete review, see: (Julien and Kriz, 2006; Lariviere and Julien, 2004)). Even though genetic mutations in intermediate filament (IF) genes are not major causes of ALS, it is of potential relevance to ALS that transgenic mice with altered stoichiometry of neuronal intermediate filament develop pathological features of the disease (Beaulieu, et al., 2000; Beaulieu and Julien, 2003; Cote, et al., 1993; Millecamps, et al., 2006). Of particular interest was the finding that overexpression of WT peripherin, a type III intermediate filament, in *NEFL* null background mice caused age-dependant selective motor neurons degeneration (Beaulieu, et al., 1999). This mouse model is also characterized by the formation of perikaryal and axonal intermediate filament inclusions resembling spheroids in motor neurons of human ALS. The precise mechanism by which accumulation of intermediate neurofilament leads to neurodegenerative disorders is not fully understood. Neurofilament and peripherin proteins are two types of intermediate filaments detected in the majority of axonal inclusion bodies, called spheroids, in motor neurons of ALS patients (Corbo and Hays, 1992). Multiple factors can potentially cause the accumulation of intermediate filament proteins including deregulation of intermediate filament protein synthesis, proteolysis, defective axonal transport, abnormal

Truncating mutations were discovered in coding exons of a the *ALS2* gene encoding for Alsin, from patients with an autosomal recessive form of juvenile ALS, primary lateral sclerosis (PLS) and infantile-onset ascending hereditary spastic paralysis (IAHSP) (Devon, et al., 2003; Eymard-Pierre, et al., 2002; Eymard-Pierre, et al., 2006; Gros-Louis, et al., 2003; Hadano, et al., 2001; Kress, et al., 2005; Yang, et al., 2001). The pattern of inheritance and the nature of the mutations identified in this gene suggest that motor neuron degeneration seen in patients results from a loss of protein function. Six different groups have reported the generation of an alsin knockout mouse (Cai, et al., 2005; Deng, et al., 2007; Devon, et al., 2003; Gros-Louis, et al., 2008; Hadano, et al., 2006; Yamanaka, et al., 2006). Despite an age-

model.

**2.2.1.3 Mice intermediate filament models** 

phosphorylation, and other protein modifications.

**2.2.1.4 Mice ALS2 models** 

As a model of human disease, the laboratory rat (*Rattus norvegicus*) offers some advantages over the mouse and other organisms. The main advantage is in fact that rats are physiologically more similar to humans compare to mice. The size of the animal also confers a valuable advantage and enhances its use as a disease model. The later advantage is especially true when performing surgical procedures, microdialysis, intravenous cannulation and for pharmacodynamic drug studies as its size enables serial blood sampling. Thereby, the rat would be a better model than the mouse to study cardiovascular disease, hypertension, diabetes, arthritis, and many autoimmune, behavioral, and addiction disorders. The rat is also a good model in neurobehavioral and stereotaxic neurological studies. The size of the rat brain offers unique possibilities for the application of microsurgical techniques, intrathecal administration of drugs, stem cell transplantation, serial sampling of the cerebrospinal fluid (CSF), *in vivo* nerve recordings, and neuroimaging procedures.

However, even though its size is considered as an advantage over mice, the higher cost of maintenance (bigger cages, food, less animals can be housed per cages) and limited housing capacity in animal facilities directly related to its size also confers the principal limitation of using rats as a model.

### **2.2.2.1 Rat SOD1 models**

Transgenic rat models of ALS have also been generated. In these models overexpression of G93A or H46R mutant SOD1 led to an ALS-like phenotype (Howland, et al., 2002; Nagai, et

*In Vivo* and *In Vitro* Models to Study Amyotrophic Lateral Sclerosis 99

F1 heterozygous progeny and so on so forth in order to generate homozygous progeny and to facilitate the identification of gene mutation by positional cloning. Forward-genetic screens in zebrafish are greatly facilitated by the transparency of embryos and larvae, simplifying phenotypic screening and allowing large scale screening without sophisticated expansive equipments. These features provide an advantage over other vertebrate models in which aspects of organogenesis and disease pathology cannot be examined without surgical interventions. Large-scale forward-genetic screens have allowed the identification of numerous mutations in zebrafish genes (Amsterdam and Hopkins, 2006). Comparative genomic can be then used in order to identify the corresponding orthologous gene in humans and, in some instances, the mutant phenotypes have been sufficiently similar to the human pathology to allow the identification of gene mutations by a candidate gene

On the other hand, reverse genetics refers to the study of the phenotypic consequences following manipulation of gene expression. An essential tool for investigating gene function during development is the ability to perform gene knockdown, overexpression, and misexpression studies. The external egg fertilization in zebrafish enables microinjection of RNA, DNA, proteins, antisense oligonucleotides and other small molecules into the developing embryo thereby providing researchers a quick and robust assay for exploring gene function *in vivo*. Antisense morpholino oligonucleotides (AMO) are now widely used to modify gene expression by blocking translation of a targeted protein or by modifying premRNA splicing (Nasevicius and Ekker, 2000)). The ability to examine the course of a pathological process *in vivo* and in real time using AMO reverse genetic is a particular strength of zebrafish models. In particular, AMO are powerful reverse genetic tools for the biological validation of genetic variants identified in human and for defining the pathological nature of a particular gene mutation (Amsterdam and Hopkins, 2006). Finally, high-throughput screening for small-molecule chemical able to modify disease pathogenesis, firmly establishing a role for zebrafish in the field of pharmaceutical drug

In order to generate a SOD1 zebrafish model of ALS, transient overexpression of human SOD1 mutants (A4V, G37R and G93A) was achieved using mRNA microinjection into 2 to 4 cell stage blastulae (Lemmens, et al., 2007). Overexpression of mutant human SOD1 in zebrafish embryos induced a dose-dependant motor axonopathy in all studied mutations. Likewise in ALS patients, ubiquitous expression of the mutant protein produced motor neuron specific neuropathology, suggesting that this model may be useful to elucidate the mechanisms underlying specific vulnerability of motor neurons in *SOD1*-linked ALS. Interestingly, the observed motor neuron phenotype in this transient SOD1mutant zebrafish

With the goal of achieving constitutive mutant SOD1 expression in zebrafish, Ramesh and colleagues generated transgenic zebrafish models overexpressing either WT or mutant SOD1 (Ramesh, et al., 2010). Zebrafish overexpressing mutant SOD1 exhibited many hallmark phenotypes of ALS including neuromuscular junction defects, spinal motor neuron loss, muscle degeneration, decreased endurance to swim, partial paralysis and premature death. Interestingly, all of these pathological features are consistent with those seen in SOD1 transgenic mice and rats, supporting the idea that overexpression of SOD1 in vertebrate models results in the development of common pathologies. As zebrafish are

embryo expresser suggests that ALS may be a developmental disease.

approach.

discovery.

**2.2.3.1 Zebrafish SOD1 models** 

al., 2001). These transgenic rat models reproduce the major phenotypic features of human ALS, such as selective motor neuron loss, ubiquitination, hyaline inclusions, vacuolation, and neuroinflammation. However, several differences between the rat and mouse ALS models can be denoted including a more rapid progression of disease and the transient appearance of vacuoles in the transgenic SOD1 rats. As the CSF volume of a rat is 10- to 20 fold greater than that of a mouse (Nagai, et al., 2001), transgenic *SOD1* rats may facilitate to study the CSF proteome and to detect SOD1 activity or drug concentrations, even in an agedependent manner. Therefore, ALS rat models may be useful to demonstrate therapeutic efficacy, blood-brain barrier filtration and to assess the toxicity of novel therapeutic compounds in drug development.

### **2.2.2.2 Rat TDP-43 models**

In order to generate a TDP-43 rodent model that is more suitable for pharmacological studies, rat models ubiquitously overexpressing either WT TDP-43 or the M337V-associated *TARDBP* mutation were generated (Zhou, et al., 2010). In contrast to what it is observed in TDP-43 transgenic mouse models, overexpression of mutant TDP-43 only, but not the WT protein, caused widespread neurodegeneration. Despite the fact that this transgenic mutant TDP-43 rat model exhibited progressive degeneration of motor neurons, neurodegeneration was not only restricted to motor neurons. However, TDP-43 mutation affected motor neurons earlier and more severely than other neurons in the CNS at end stage of the disease. In some aspects, this rat model recapitulated TDP-43 pathological features both seen in the different mouse models and in ALS patients indicating that this model could be used in future pharmalogical studies in order to identify novel therapeutic avenues to treat TDP-43 related disorders.

### **2.2.3 Zebrafish**

In the last decade, a new vertebrate model has emerged in the study of human diseases. The principal attraction of using zebrafish (*Danio renio*) as model organisms is that its genome can be much more easily manipulated and studied in terms of genetics and development than other vertebrate model organisms. As being a vertebrate with common organs and tissues with conserved organization such as brain and spinal cord, the attractiveness of zebrafish to model human diseases lies therefore in its biology and genetics. Although there are obvious differences in the physiology of fish and humans the zebrafish offers several advantages that make it an important complement to mouse models of disease. These advantages include the embryo and larvae optical clarity (which allows easy visualization of developmental processes), the external fertilization (which facilitates gene expression manipulation), high fecundity (adult female can lay hundreds of eggs per day), and rapid development (swimming behavior appears less than 48 hours post-fertilization). As in invertebrate models, it is also relatively easy to perform forward and reverse genetic screens for gene identification and to understand specific gene function.

Forward-genetic screens in zebrafish, using random mutagenesis, can be used to identify novel genes involved in embryogenesis, organogenesis or different specific biological processes. To generate mutants, male fish are exposed to the mutagen ethyl-nitroso-urea (ENU), which typically induces point mutations within zebrafish genome (Driever, et al., 1996; Haffter, et al., 1996; Henion, et al., 1996; Solnica-Krezel, et al., 1996). Random mutagenesis has also been successfully carried out in zebrafish using retroviral methods (Amsterdam, et al., 1999). Treated males are then crossed to wildtype females to produce the

al., 2001). These transgenic rat models reproduce the major phenotypic features of human ALS, such as selective motor neuron loss, ubiquitination, hyaline inclusions, vacuolation, and neuroinflammation. However, several differences between the rat and mouse ALS models can be denoted including a more rapid progression of disease and the transient appearance of vacuoles in the transgenic SOD1 rats. As the CSF volume of a rat is 10- to 20 fold greater than that of a mouse (Nagai, et al., 2001), transgenic *SOD1* rats may facilitate to study the CSF proteome and to detect SOD1 activity or drug concentrations, even in an agedependent manner. Therefore, ALS rat models may be useful to demonstrate therapeutic efficacy, blood-brain barrier filtration and to assess the toxicity of novel therapeutic

In order to generate a TDP-43 rodent model that is more suitable for pharmacological studies, rat models ubiquitously overexpressing either WT TDP-43 or the M337V-associated *TARDBP* mutation were generated (Zhou, et al., 2010). In contrast to what it is observed in TDP-43 transgenic mouse models, overexpression of mutant TDP-43 only, but not the WT protein, caused widespread neurodegeneration. Despite the fact that this transgenic mutant TDP-43 rat model exhibited progressive degeneration of motor neurons, neurodegeneration was not only restricted to motor neurons. However, TDP-43 mutation affected motor neurons earlier and more severely than other neurons in the CNS at end stage of the disease. In some aspects, this rat model recapitulated TDP-43 pathological features both seen in the different mouse models and in ALS patients indicating that this model could be used in future pharmalogical studies in order to identify novel therapeutic avenues to treat TDP-43

In the last decade, a new vertebrate model has emerged in the study of human diseases. The principal attraction of using zebrafish (*Danio renio*) as model organisms is that its genome can be much more easily manipulated and studied in terms of genetics and development than other vertebrate model organisms. As being a vertebrate with common organs and tissues with conserved organization such as brain and spinal cord, the attractiveness of zebrafish to model human diseases lies therefore in its biology and genetics. Although there are obvious differences in the physiology of fish and humans the zebrafish offers several advantages that make it an important complement to mouse models of disease. These advantages include the embryo and larvae optical clarity (which allows easy visualization of developmental processes), the external fertilization (which facilitates gene expression manipulation), high fecundity (adult female can lay hundreds of eggs per day), and rapid development (swimming behavior appears less than 48 hours post-fertilization). As in invertebrate models, it is also relatively easy to perform forward and reverse genetic screens

Forward-genetic screens in zebrafish, using random mutagenesis, can be used to identify novel genes involved in embryogenesis, organogenesis or different specific biological processes. To generate mutants, male fish are exposed to the mutagen ethyl-nitroso-urea (ENU), which typically induces point mutations within zebrafish genome (Driever, et al., 1996; Haffter, et al., 1996; Henion, et al., 1996; Solnica-Krezel, et al., 1996). Random mutagenesis has also been successfully carried out in zebrafish using retroviral methods (Amsterdam, et al., 1999). Treated males are then crossed to wildtype females to produce the

for gene identification and to understand specific gene function.

compounds in drug development.

**2.2.2.2 Rat TDP-43 models** 

related disorders.

**2.2.3 Zebrafish** 

F1 heterozygous progeny and so on so forth in order to generate homozygous progeny and to facilitate the identification of gene mutation by positional cloning. Forward-genetic screens in zebrafish are greatly facilitated by the transparency of embryos and larvae, simplifying phenotypic screening and allowing large scale screening without sophisticated expansive equipments. These features provide an advantage over other vertebrate models in which aspects of organogenesis and disease pathology cannot be examined without surgical interventions. Large-scale forward-genetic screens have allowed the identification of numerous mutations in zebrafish genes (Amsterdam and Hopkins, 2006). Comparative genomic can be then used in order to identify the corresponding orthologous gene in humans and, in some instances, the mutant phenotypes have been sufficiently similar to the human pathology to allow the identification of gene mutations by a candidate gene approach.

On the other hand, reverse genetics refers to the study of the phenotypic consequences following manipulation of gene expression. An essential tool for investigating gene function during development is the ability to perform gene knockdown, overexpression, and misexpression studies. The external egg fertilization in zebrafish enables microinjection of RNA, DNA, proteins, antisense oligonucleotides and other small molecules into the developing embryo thereby providing researchers a quick and robust assay for exploring gene function *in vivo*. Antisense morpholino oligonucleotides (AMO) are now widely used to modify gene expression by blocking translation of a targeted protein or by modifying premRNA splicing (Nasevicius and Ekker, 2000)). The ability to examine the course of a pathological process *in vivo* and in real time using AMO reverse genetic is a particular strength of zebrafish models. In particular, AMO are powerful reverse genetic tools for the biological validation of genetic variants identified in human and for defining the pathological nature of a particular gene mutation (Amsterdam and Hopkins, 2006). Finally, high-throughput screening for small-molecule chemical able to modify disease pathogenesis, firmly establishing a role for zebrafish in the field of pharmaceutical drug discovery.

### **2.2.3.1 Zebrafish SOD1 models**

In order to generate a SOD1 zebrafish model of ALS, transient overexpression of human SOD1 mutants (A4V, G37R and G93A) was achieved using mRNA microinjection into 2 to 4 cell stage blastulae (Lemmens, et al., 2007). Overexpression of mutant human SOD1 in zebrafish embryos induced a dose-dependant motor axonopathy in all studied mutations. Likewise in ALS patients, ubiquitous expression of the mutant protein produced motor neuron specific neuropathology, suggesting that this model may be useful to elucidate the mechanisms underlying specific vulnerability of motor neurons in *SOD1*-linked ALS. Interestingly, the observed motor neuron phenotype in this transient SOD1mutant zebrafish embryo expresser suggests that ALS may be a developmental disease.

With the goal of achieving constitutive mutant SOD1 expression in zebrafish, Ramesh and colleagues generated transgenic zebrafish models overexpressing either WT or mutant SOD1 (Ramesh, et al., 2010). Zebrafish overexpressing mutant SOD1 exhibited many hallmark phenotypes of ALS including neuromuscular junction defects, spinal motor neuron loss, muscle degeneration, decreased endurance to swim, partial paralysis and premature death. Interestingly, all of these pathological features are consistent with those seen in SOD1 transgenic mice and rats, supporting the idea that overexpression of SOD1 in vertebrate models results in the development of common pathologies. As zebrafish are

*In Vivo* and *In Vitro* Models to Study Amyotrophic Lateral Sclerosis 101

Consistent with this hypothesis, Gros-Louis and colleagues found that the zAls2 knocked down zebrafish led to both severe developmental abnormalities and an obvious behavioral phenotype, including swimming impairment and motor neuron disruption. This model is still the only ALS2 animal model mimicking the severe motor neuron degeneration observed in ALS2 patients. Interestingly, some of these Als2 splice variants rescued the zebrafish phenotype suggesting that these identified spliced variants, observed in Als2 null mice and in ALS2 patients, prevented the Als2 KO mice from developing severe neurodegenerative disease. These findings may also explain the variable phenotype

The domestic dog (*Canis familiaris*), with over 450 naturally-occurring hereditary diseases, is a valuable model organism for the study of human genetic diseases and complex traits. In humans, common diseases show complex modes of inheritance, and as a result, genetic analysis and gene mapping could be tedious. Rodent systems are more tractable genetically, but studies using transgenic mouse models represent an induction of a particular mutation rather than naturally arising alleles, and results are often of limited direct relevance to human disease because of profound differences in physiology. In contrast, the physiology, disease presentation and clinical response of dogs to drugs often mimic human diseases closely. The dog genome is similar in size to the human genome; containing an estimated 2.8 billion DNA base pairs (Lindblad-Toh, et al., 2005). Canine models have played an important role in advancing biomedical knowledge and techniques. Due to a long history of selective breeding, many breeds of dogs are naturally prone to genetic diseases including cancer, autoimmune and neurological disorders afflicting humans. These dog models, in which naturally occurring mutations in different genes was described, may help the study of genetic diseases difficult to study in humans. The structure of the canine population offers specific advantages for genetic mapping studies. The dog enjoys a genetic diversity unrivaled by any other mammalian species. A thousand centuries of directed breeding by humans has been responsible for that diversity leading to an unequaled variety of morphologies and behaviors, and also into a number of inherited diseases. The top 10 diseases in dogs include cancer, epilepsy, allergy, retinal disease, cataracts, and heart

disease. Several of these diseases constitute also a major health concerns to humans.

because of an obvious emotional tie toward dogs.

**2.2.4.1 Dog SOD1 model** 

The disadvantages of using dog as a model are its size, cost of maintenance and housing limitations in order to provide adequate area for the dogs to run. Furthermore, there use in research has been more controversial and of public concerns than other animal models

A genome-wide association analysis reveals a *SOD1* mutation in canine degenerative myelopathy (Awano, et al., 2009). This disease has been recognized for more than 35 years as a spontaneously occurring, adult-onset spinal cord disorder of dogs (Averill, 1973). The resequencing of the canine *Sod1* gene revealed a recessive E40K missense mutation. Pathological studies of spinal cords from affected dogs showed myelin and axonal loss, SOD1-positive neuronal cytoplasmic inclusions similar to those seen in patients. The disease is also fatal in dogs and, typically, presents with progressive upper motor neuron spasticity and general proprioceptive ataxia in the pelvic limbs leading to paraplegia. Dogs develop disease generally around 8 years of age or older and disease duration will not exceed, if

observed in ALS2-linked patients.

**2.2.4 Dog** 

highly amenable to chimeric analysis, this zebrafish model will offer a tool to hopefully address in future studies the cell autonomous hypothesis seen in SOD1-linked ALS.

### **2.2.3.2 Zebrafish TDP-43 models**

In order to test the toxicity associated with various *TARDBP* mutations (A315T, G348C and A382T), transient overexpression of WT and mutant TDP-43 was achieved by *TARDBP* mRNA microinjection (Kabashi, et al., 2010). Mutant TDP-43 overexpression resulted in motor neuron defects, including shorter axons, abnormal axonal branching and swimming deficits in the absence of sensory deficits. These phenotypes were also observed in WT *TARDBP* overexpressing zebrafish, but were less severe. These results confirmed that the studied *TARDBP* mutations are toxic and suggested that TDP-43 linked mutations preferentially target motor neurons for degeneration. Interestingly, knocking down the zebrafish *Tardbp* gene (z*Tardbp*), using a specific AMO designed to block z*Tardbp* translation, led to a similar phenotype, which was rescued by co-expressing WT but not mutant human TDP-43 (Kabashi, et al., 2010). Together these two reverse genetic approaches showed that *TARDBP* mutations cause motor neuron defects and toxicity, suggesting that mutant TDP-43 contributes to disease pathogenesis by both a toxic gain-of-function and a loss-of-function molecular mechanisms.

### **2.2.3.3 Zebrafish FUS models**

To test the FUS/TLS-linked mutations toxicity, transient expression of mRNAs encoding green fluorescent protein (GFP)-FUS (WT, H517Q, R521G, R495X or G515X) fused proteins was achieved by injecting these mRNA into zebrafish eggs at the 1 to 2 cells stage of development (Bosco, et al., 2010a). Interestingly, the expression of human FUS WT and the H517Q and R521G mutants each exhibited a predominantly nuclear pattern in the spinal cord, whereas the R495X and G515X truncation mutants accumulated in the cytoplasm. Noteworthy, abnormal accumulation of protein in the cytoplasm has been frequently observed in post-mortem tissues collected from ALS patients. This phenotype was exacerbated after heat shock treatment. These results suggest that the studied FUS-linked mutations induce toxicity and that multiple factors may influence the subcellular localization of GFP-FUS *in vivo*, including various cellular stresses. Similarly, upon transient overexpression of a FUS-linked ALS mutation (R521H) in zebrafish embryos, motor deficits characterized by abnormal touch-evoke escape response was observed in 57% of the mRNA injected zebrafish larvae as compared to 23% in controls (Kabashi, et al., 2011). These results, from both studies, indicate that ALS-linked FUS/TLS mutants can cause motor neuron deficits by a toxic gain-of-function. On the other hand, loss-of-function cannot be totally excluded as knockdown of the zebrafish *Fus* (z*Fus*) gene, using antisense morpholino oligo, also yielded a motor phenotype characterized by a deficient touch-evoked escape response, reduced motor neuron outgrowth and axonal branching (Kabashi, et al., 2011). Interestingly, these phenotypes could be rescued upon co-expression of WT human FUS but not ALSlinked *FUS/TLS* mutations.

### **2.2.3.4 Zebrafish ALS2 models**

To investigate *ALS2* gene function and to elucidate if the identified mouse Als2 splicing isoforms may compensate for the loss of the full length protein in Als2 KO mice described earlier, a knock-down of the zebrafish Als2 ortholog (z*Als2*) in zebrafish was performed using an AMO directed against the start codon of the z*Als2* gene (Gros-Louis, et al., 2008). Consistent with this hypothesis, Gros-Louis and colleagues found that the zAls2 knocked down zebrafish led to both severe developmental abnormalities and an obvious behavioral phenotype, including swimming impairment and motor neuron disruption. This model is still the only ALS2 animal model mimicking the severe motor neuron degeneration observed in ALS2 patients. Interestingly, some of these Als2 splice variants rescued the zebrafish phenotype suggesting that these identified spliced variants, observed in Als2 null mice and in ALS2 patients, prevented the Als2 KO mice from developing severe neurodegenerative disease. These findings may also explain the variable phenotype observed in ALS2-linked patients.

### **2.2.4 Dog**

100 Amyotrophic Lateral Sclerosis

highly amenable to chimeric analysis, this zebrafish model will offer a tool to hopefully

In order to test the toxicity associated with various *TARDBP* mutations (A315T, G348C and A382T), transient overexpression of WT and mutant TDP-43 was achieved by *TARDBP* mRNA microinjection (Kabashi, et al., 2010). Mutant TDP-43 overexpression resulted in motor neuron defects, including shorter axons, abnormal axonal branching and swimming deficits in the absence of sensory deficits. These phenotypes were also observed in WT *TARDBP* overexpressing zebrafish, but were less severe. These results confirmed that the studied *TARDBP* mutations are toxic and suggested that TDP-43 linked mutations preferentially target motor neurons for degeneration. Interestingly, knocking down the zebrafish *Tardbp* gene (z*Tardbp*), using a specific AMO designed to block z*Tardbp* translation, led to a similar phenotype, which was rescued by co-expressing WT but not mutant human TDP-43 (Kabashi, et al., 2010). Together these two reverse genetic approaches showed that *TARDBP* mutations cause motor neuron defects and toxicity, suggesting that mutant TDP-43 contributes to disease pathogenesis by both a toxic gain-of-function and a loss-of-function

To test the FUS/TLS-linked mutations toxicity, transient expression of mRNAs encoding green fluorescent protein (GFP)-FUS (WT, H517Q, R521G, R495X or G515X) fused proteins was achieved by injecting these mRNA into zebrafish eggs at the 1 to 2 cells stage of development (Bosco, et al., 2010a). Interestingly, the expression of human FUS WT and the H517Q and R521G mutants each exhibited a predominantly nuclear pattern in the spinal cord, whereas the R495X and G515X truncation mutants accumulated in the cytoplasm. Noteworthy, abnormal accumulation of protein in the cytoplasm has been frequently observed in post-mortem tissues collected from ALS patients. This phenotype was exacerbated after heat shock treatment. These results suggest that the studied FUS-linked mutations induce toxicity and that multiple factors may influence the subcellular localization of GFP-FUS *in vivo*, including various cellular stresses. Similarly, upon transient overexpression of a FUS-linked ALS mutation (R521H) in zebrafish embryos, motor deficits characterized by abnormal touch-evoke escape response was observed in 57% of the mRNA injected zebrafish larvae as compared to 23% in controls (Kabashi, et al., 2011). These results, from both studies, indicate that ALS-linked FUS/TLS mutants can cause motor neuron deficits by a toxic gain-of-function. On the other hand, loss-of-function cannot be totally excluded as knockdown of the zebrafish *Fus* (z*Fus*) gene, using antisense morpholino oligo, also yielded a motor phenotype characterized by a deficient touch-evoked escape response, reduced motor neuron outgrowth and axonal branching (Kabashi, et al., 2011). Interestingly, these phenotypes could be rescued upon co-expression of WT human FUS but not ALS-

To investigate *ALS2* gene function and to elucidate if the identified mouse Als2 splicing isoforms may compensate for the loss of the full length protein in Als2 KO mice described earlier, a knock-down of the zebrafish Als2 ortholog (z*Als2*) in zebrafish was performed using an AMO directed against the start codon of the z*Als2* gene (Gros-Louis, et al., 2008).

address in future studies the cell autonomous hypothesis seen in SOD1-linked ALS.

**2.2.3.2 Zebrafish TDP-43 models** 

molecular mechanisms.

**2.2.3.3 Zebrafish FUS models** 

linked *FUS/TLS* mutations. **2.2.3.4 Zebrafish ALS2 models**  The domestic dog (*Canis familiaris*), with over 450 naturally-occurring hereditary diseases, is a valuable model organism for the study of human genetic diseases and complex traits. In humans, common diseases show complex modes of inheritance, and as a result, genetic analysis and gene mapping could be tedious. Rodent systems are more tractable genetically, but studies using transgenic mouse models represent an induction of a particular mutation rather than naturally arising alleles, and results are often of limited direct relevance to human disease because of profound differences in physiology. In contrast, the physiology, disease presentation and clinical response of dogs to drugs often mimic human diseases closely. The dog genome is similar in size to the human genome; containing an estimated 2.8 billion DNA base pairs (Lindblad-Toh, et al., 2005). Canine models have played an important role in advancing biomedical knowledge and techniques. Due to a long history of selective breeding, many breeds of dogs are naturally prone to genetic diseases including cancer, autoimmune and neurological disorders afflicting humans. These dog models, in which naturally occurring mutations in different genes was described, may help the study of genetic diseases difficult to study in humans. The structure of the canine population offers specific advantages for genetic mapping studies. The dog enjoys a genetic diversity unrivaled by any other mammalian species. A thousand centuries of directed breeding by humans has been responsible for that diversity leading to an unequaled variety of morphologies and behaviors, and also into a number of inherited diseases. The top 10 diseases in dogs include cancer, epilepsy, allergy, retinal disease, cataracts, and heart disease. Several of these diseases constitute also a major health concerns to humans.

The disadvantages of using dog as a model are its size, cost of maintenance and housing limitations in order to provide adequate area for the dogs to run. Furthermore, there use in research has been more controversial and of public concerns than other animal models because of an obvious emotional tie toward dogs.

#### **2.2.4.1 Dog SOD1 model**

A genome-wide association analysis reveals a *SOD1* mutation in canine degenerative myelopathy (Awano, et al., 2009). This disease has been recognized for more than 35 years as a spontaneously occurring, adult-onset spinal cord disorder of dogs (Averill, 1973). The resequencing of the canine *Sod1* gene revealed a recessive E40K missense mutation. Pathological studies of spinal cords from affected dogs showed myelin and axonal loss, SOD1-positive neuronal cytoplasmic inclusions similar to those seen in patients. The disease is also fatal in dogs and, typically, presents with progressive upper motor neuron spasticity and general proprioceptive ataxia in the pelvic limbs leading to paraplegia. Dogs develop disease generally around 8 years of age or older and disease duration will not exceed, if

*In Vivo* and *In Vitro* Models to Study Amyotrophic Lateral Sclerosis 103

from post-mortem tissues is limited due to the difficulty to isolate living cells from adult

Thus, such postmortem tissue biopsies are mostly used to perform histological and immunohistochemical, genetic as well as proteomic studies. More recently, they were also used to extract neural precursor cells that were further differentiated into motor neurons

The best way to preserve all the cellular content and the complex electrophysiological and biochemical organization of cells in the spinal cord is to maintain a whole tissue slice in an organotypic culture. In addition, organotypic slice cultures can be obtained from both embryos and postnatal animals, conferring a major advantage when using transgenic mice from which the disease genotype has to be ascertained after birth (Kosuge, et al., 2009; Mazzone and Nistri, 2011). After dissection of the lumbar spinal cord and removal of the meninges, 200 to 400µm-thick transversal sections are sectioned and transferred into membrane inserts fitting six-well or 12-well culture plates (Caldero, et al., 2010). These organotypic cultures can be used for more than 2 months (Delfs, et al., 1989). Various types of molecules can be added in the culture medium, such as kainate or lithium, to modulate

These tissue sections can then be used to perform various analysis like recording the neuron rhythmic activity by placing the tissue on a multielectrode array (Tscherter, et al., 2001; Young, et al., 2007), or studying real-time glutamate release using a biosensor and redox reaction current using a multichannel potentiostat (Mazzone and Nistri, 2011), in addition to the conventional immunohistochemical staining and electrophysiological recordings (Caldero, et al., 2010; Young, et al., 2007). Neonatal brainstem slices can be used to analyze excitotoxicity on hypoglossal motor neurons that control the respiratory drive. These neurons are particularly useful for electrophysiological studies because they exhibit a range of rhythmic patterns that

Organotypic culture is a convenient culture system that closely reproduces the *in vivo* situation and allows dynamic studies with various drugs and useful analysis systems. However, it does not always fully recapitulate what is happening *in vivo* (Tovar, et al., 2009). In addition, it does not easily allow the study of different mixed cell combinations by removing or by adding specific cell types to the tissue slices. These organotypic spinal cord slice cultures are difficult to obtain from postmortem patients due to limited availability and it is obviously impossible to study disease progression using human spinal cord biopsies as

The isolation of individual cells is extremely useful to study various intracellular mechanisms from proliferation to mRNA expression, mitochondrial function, protein aggregation, intermediate filament assembly or axonal transport. However, neurons cannot proliferate and are difficult to maintain in culture. In addition, adult nervous tissues are difficult to access and the viability of the cells extracted from them is very low. Actually, most of the cells that can be maintained in culture after extraction are neural precursor cells (Haidet-Phillips, et al., 2011). To overcome this limitation, embryos are preferred for nervous tissue harvesting. Indeed, at the E12-13 embryonic developmental stage, the spinal cord of mice embryos is easy to access

will be altered during early onset of ALS (Cifra, et al., 2011a; Cifra, et al., 2011b).

brain or spinal cord, especially motor neurons.

**3.1 Organotypic cultures of spinal cord slices** 

and glial cells to develop innovative models of the disease.

neurotoxicity (Caldero, et al., 2010; Mazzone and Nistri, 2011).

opposed to biopsies obtained from animal models.

**3.2 Spinal cord cell cultures** 

euthanasia is delayed, 3 years. These findings suggest that this disease afflicting dogs is closely related to ALS and identify canine degenerative myelopathy to be the first recognized spontaneously occurring animal model for ALS. The canine ALS model may be particularly valuable for evaluating therapeutic interventions as the environmental conditions and the SOD1 level of expression mimic better the human ALS situation.

### **2.2.5 Pig**

Pigs, although not easily kept for laboratory research, are readily available for biomedical research through the large-scale industrial production of pigs produced for human consumption. Recent research has facilitated the biological experimentation with pigs, and helped develop the pig into a novel model organism for biomedical research.

The domesticated pig (*Sus scrofa*) shares several similarities with human, in particular the size of organs and various aspects of anatomy and physiology. The development of somatic cloning technology and the merger with techniques of targeted genetic modification and conditional gene expression will enhance the possibilities for creating useful models for human diseases in pigs. The pig has also evolved as the major target species for producing xenografts in order to provide appropriate human organs. The sequencing of the domestic pig genome has not yet been fully completed. However, initial draft revealed that the size and composition of the porcine genome is comparable to that of humans; comprising about 2.7 billion base pairs (Hart, et al., 2007). Furthermore, both gene content and sequence are highly conserved between pig and human. Detailed information on the porcine genome together with emerging transgenic technologies, such as siRNA or conditional knockouts will enhance our possibilities to create useful pig models. Other advantages of using domestic pigs to model human diseases include high fertility, great abundance, rapid growth, anatomy and physiology not too different from human and the possibility to introduce genetic modifications in its genome. The pig has been an essential and very successful model in biomedical research and is particularly suited to close the gap between basic research in current models and clinical application. The future will certainly see several promising porcine models for human diseases.

The high resemblance between the central nervous systems of humans and pigs makes the pig an ideal model organism for studying human neurodegenerative diseases. For neurodegenerative disorders such as ALS, Parkinson's disease and Alzheimer's disease, the pig may represent a model superior to other models presently available. Large animals, including pigs and non-human primates in neuroscience enable the use of conventional clinical brain imaging and the direct testing of surgical procedures. The evaluation of novel therapeutic avenues in an animal model with higher brain complexity will allow a more direct translation to human diseases.
