The enteric nervous system is a potential autoimmune target in multiple sclerosis.

Multiple sclerosis (MS) is a chronic autoimmune disease of the central nervous system (CNS) in young adults that has serious negative socioeconomic effects. In addition to symptoms caused by CNS pathology, the majority of MS patients frequently exhibit gastrointestinal dysfunction, which was previously either explained by the presence of spinal cord lesions or not directly linked to the autoimmune etiology of the disease. Here, we studied the enteric nervous system (ENS) in a B cell- and antibody-dependent mouse model of MS by immunohistochemistry and electron microscopy at different stages of the disease. ENS degeneration was evident prior to the development of CNS lesions and the onset of neurological deficits in mice. The pathology was antibody mediated and caused a significant decrease in gastrointestinal motility, which was associated with ENS gliosis and neuronal loss. We identified autoantibodies against four potential target antigens derived from enteric glia and/or neurons by immunoprecipitation and mass spectrometry. Antibodies against three of the target antigens were also present in the plasma of MS patients as confirmed by ELISA. The analysis of human colon resectates provided evidence of gliosis and ENS degeneration in MS patients compared to non-MS controls. For the first time, this study establishes a pathomechanistic link between the well-established autoimmune attack on the CNS and ENS pathology in MS, which might provide a paradigm shift in our current understanding of the immunopathogenesis of the disease with broad diagnostic and therapeutic implications.

Loss of Aß-nerve endings associated with the Merkel cell-neurite complex in the lesional oral mucosa epithelium of lichen planus and hyperkeratosis.

The Merkel cell-neurite complex initiates the perception of touch and mediates Aß slowly adapting type I responses. Lichen planus is a chronic inflammatory autoimmune disease with T-cell-mediated inflammation, whereas hyperkeratosis is characterized with or without epithelial dysplasia in the oral mucosa. To determine the effects of lichen planus and hyperkeratosis on the Merkel cell-neurite complex, healthy oral mucosal epithelium and lesional oral mucosal epithelium of lichen planus and hyperkeratosis patients were stained by immunohistochemistry (the avidin-biotin-peroxidase complex and double immunofluorescence methods) using pan cytokeratin, cytokeratin 20 (K20, a Merkel cell marker), and neurofilament 200 (NF200, a myelinated Aß- and Aδ-nerve fibre marker) antibodies. NF200-immunoreactive (ir) nerve fibres in healthy tissues and in the lesional oral mucosa epithelium of lichen planus and hyperkeratosis were counted and statistically analysed. In the healthy oral mucosa, K20-positive Merkel cells with and without close association to the intraepithelial NF200-ir nerve fibres were detected. In the lesional oral mucosa of lichen planus and hyperkeratosis patients, extremely rare NF200-ir nerve fibres were detected only in the lamina propria. Compared with healthy tissues, lichen planus and hyperkeratosis tissues had significantly decreased numbers of NF200-ir nerve fibres in the oral mucosal epithelium. Lichen planus and hyperkeratosis were associated with the absence of Aß-nerve endings in the oral mucosal epithelium. Thus, we conclude that mechanosensation mediated by the Merkel cell-neurite complex in the oral mucosal epithelium is impaired in lichen planus and hyperkeratosis.

Cytokinesis failure and successful multipolar mitoses drive aneuploidy in glioblastoma cells.

Glioblastoma (GB) is the most frequent human brain tumor and is associated with a poor prognosis. Multipolar mitosis and spindles have occasionally been observed in cultured glioblastoma cells and in glioblastoma tissues, but their mode of origin and relevance have remained unclear. In the present study, we investigated a novel GB cell line (SGB4) exhibiting mitotic aberrations and established a functional link between cytokinesis failure, centrosome amplification, multipolar mitosis and aneuploidy in glioblastoma. Long-term live cell imaging showed that >3% of mitotic SGB4 cells underwent multipolar mitosis (tripolar>tetrapolar>pentapolar). A significant amount of daugther cells generated by multipolar mitosis were viable and completed several rounds of mitosis. Pedigree analysis of mitotic events revealed that in many cases a bipolar mitosis with failed cytokinesis occurred prior to a multipolar mitosis. Additionally, we observed that SGB4 cells were also able to undergo a bipolar mitosis after failed cytokinesis. Colchicine-induced mitotic arrest and metaphase spreads demonstrated that SGB4 cells had a modal chromosome number of 58 ranging from 23 to 170. Approximately 82% of SGB4 cells were hyperdiploid (47-57 chromosomes) or hypotriploid (58-68 chromosomes). In conclusion, SGB4 cells passed through multipolar cell divisions and generated viable progeny by reductive mitoses. Our results identified cytokinesis failure occurring before and after multipolar or bipolar mitoses as important mechanisms to generate chromosomal heterogeneity in glioblastoma cells.

Identification of a B cell-dependent subpopulation of multiple sclerosis by measurements of brain-reactive B cells in the blood.

B cells are increasingly coming into play in the pathogenesis of multiple sclerosis (MS). Here, we screened peripheral blood mononuclear cells (PBMC) from patients with clinically isolated syndrome (CIS), MS, other non-inflammatory neurological, inflammatory neurological or autoimmune diseases, and healthy donors for their B cell reactivity to CNS antigen using the enzyme-linked immunospot technique (ELISPOT) after 96 h of polyclonal stimulation. Our data show that nine of 15 patients with CIS (60.0%) and 53 of 67 patients with definite MS (79.1%) displayed CNS-reactive B cells, compared to none of the control donors. The presence of CNS-reactive B cells in the blood of the majority of patients with MS or at risk to develop MS along with their absence in control subjects suggests that they might be indicative of a B cell-dependent subpopulation of the disease.

The growth of Staphylococcus aureus and Escherichia coli in low-direct current electric fields.

Electrical potentials up to 800 mV can be observed between different metallic dental restorations. These potentials produce fields in the mouth that may interfere with microbial communities. The present study focuses on the impact of different electric field strengths (EFS) on the growth of Staphylococcus aureus (ATCC 25923) and Escherichia coli (ATCC 25922) in vitro. Cultures of S. aureus and E. coli in fluid and gel medium were exposed to different EFS. Effects were determined by calculation of viable counts and measurement of inhibition zones. In gel medium, anodic inhibition zones for S. aureus were larger than those for E. coli at all field strength levels. In fluid medium, the maximum decrease in the viable count of S. aureus cells was at 10 V⋅m(-1). Field-treated S. aureus cells presented ruptured cell walls and disintegrated cytoplasm. Conclusively, S. aureus is more sensitive to increasing electric field strength than E. coli.

Differential aspects of immune cell infiltration and neurodegeneration in acute and relapse experimental autoimmune encephalomyelitis.

Proteolipid protein (PLP)-induced experimental autoimmune encephalomyelitis (EAE) in SJL/J mice is a well-established animal model for the study of relapsing-remitting multiple sclerosis (MS). Here we investigated histomorphological differences between acute and relapse EAE in order to get further insights into the mechanisms that trigger the transition from a relapsing-remitting course to chronic-progressive disease. We found the extent of inflammation to differ in respect to the type of immune cells infiltrating the CNS and the manifestation of edema. Myelin pathology was predominated by demyelinated axons in the acute phase. In relapse EAE the extent of myelin pathology declined and was characterized by a balance between demyelinated and demyelinating nerve fibers. Axonal pathology increased with disease progression and was partly separated from myelin pathology and inflammation. The reported incisive differences between acute and relapse EAE suggest a transition from inflammatory processes to independent neurodegeneration in the disease course.

Early axonal damage and progressive myelin pathology define the kinetics of CNS histopathology in a mouse model of multiple sclerosis.

Studies of MS histopathology are largely dependent on suitable animal models. While light microscopic analysis gives an overview of tissue pathology, it falls short in evaluating detailed changes in nerve fiber morphology. The ultrastructural data presented here and obtained from studies of myelin oligodendrocyte glycoprotein (MOG):35-55-induced experimental autoimmune encephalomyelitis (EAE) in C57BL/6 mice delineate that axonal damage and myelin pathology follow different kinetics in the disease course. While myelin pathology accumulated with disease progression, axonal damage coincided with the initial clinical disease symptoms and remained stable over time. This pattern applied both to irreversible axolysis and early axonal pathology. Notably, these histopathological patterns were reflected by the normal-appearing white matter (NAWM), suggesting that the NAWM is also in an active neurodegenerative state. The data underline the need for neuroprotection in MS and suggest the MOG model as a highly valuable tool for the assessment of different therapeutic strategies.

Nuclear expression of p27(Kip1) is associated with in vivo differentiation of adult human odontoblasts.

INTRODUCTION: Odontoblasts are terminally differentiated cells of ectomesenchymal origin that produce the dentin. Differentiated odontoblasts cannot be identified yet by a single phenotypic marker protein; therefore, a combination of markers is currently used. Up-regulation of the cyclin-dependent kinase inhibitor p27(Kip1) has been associated with exit from the cell cycle and terminal differentiation of mammalian cells. Immunoreactivity for p27(Kip1) protein was shown in many adult mouse tissues, but no information is available on the expression of p27(Kip1) in mammalian dental pulp. METHODS: Healthy and carious adult human molars with reparative dentin formation were decalcified, cryoprotected, frozen embedded, and frozen sectioned. The expression of p27(Kip1) and nestin in cells of adult human dental pulp was analyzed by immunohistochemistry using free floating sections. RESULTS: p27(Kip1) showed strong nuclear expression in many differentiated human molar odontoblasts at the odontoblastic layer. Most cells of the cell-rich zone displayed low levels of p27(Kip1) despite the fact that preodontoblasts localized in the cell-rich zone of the subodontoblastic layer have been identified as quiescent cells. The nuclear expression of p27(Kip1) in stromal cells of the dental pulp was variable, indicating that subpopulations of these cells were in distinct states of differentiation. Odontoblasts generating reparative dentin showed comparable nuclear expression of p27(Kip1) in comparison with odontoblasts synthesizing primary/secondary dentin. This result indicates that odontoblasts synthesizing primary/secondary or reparative dentin exhibit a similar differentiation status. CONCLUSIONS: Our findings show that increased expression of nuclear p27(Kip1) occurred during differentiation from preodontoblasts to odontoblasts in adult healthy and carious molars. p27(Kip1) can be used as a novel nuclear marker protein for differentiated human odontoblasts in vivo.

The complement system contributes to the pathology of experimental autoimmune encephalomyelitis by triggering demyelination and modifying the antigen-specific T and B cell response.

So far, studies of the human autoimmune disease multiple sclerosis (MS) have largely been hampered by the absence of a pathogenic B cell component in its animal model, experimental autoimmune encephalomyelitis (EAE). To overcome this shortcoming, we have previously introduced the myelin basic protein (MBP)-proteolipid protein (PLP) MP4-induced EAE, which is B cell and autoantibody-dependent. Here we show that MP4-immunized wild-type C57BL/6 mice displayed a significantly lower disease incidence when their complement system was transiently depleted by a single injection of cobra venom factor (CVF) prior to immunization. Considering the underlying pathomechanism, our data suggest that the complement system is crucial for MP4-specific antibodies to trigger CNS pathology. Demyelinated lesions in the CNS were colocalized with complement depositions. In addition, B cell deficient JHT mice reconstituted with MP4-reactive serum showed significantly attenuated clinical and histological EAE after depletion of complement by CVF. The complement system was also critically involved in the generation of the MP4-specific T and B cell response: in MP4-immunized wild-type mice treated with CVF the MP4-specific cytokine and antibody response was significantly attenuated compared to untreated wild-type mice. Taken together, we propose two independent mechanisms by which the complement system can contribute to the pathology of autoimmune encephalomyelitis. Our data corroborate the role of complement in triggering antibody-dependent demyelination and antigen-specific T cell immunity and also provide first evidence that the complement system can modify the antigen-specific B cell response in EAE and possibly MS.

The G protein Gα11 is essential for hypertrophic signalling in diabetic myocardium.

AIMS/HYPOTHESIS: Pathological cardiac hypertrophy is an early phenotype in both types 1 and 2 diabetes. The primary stimulus for hypertrophic growth in diabetes is yet unknown and may involve neurohumoral stimulation of Gq-coupled receptors as well as direct glucose-dependent mechanisms. To discriminate between these hypertrophic stimuli we analyzed hypertrophic signalling pathways in wildtype and Gα11-knockout mice. METHODS: Experimental diabetes was induced in wildtype and knockout mice by intraperitoneal injection of streptozotocin. 8 weeks after induction of diabetes myocardial function and structure was assessed by echocardiography before sacrifice. To identify prohypertrophic signalling pathways expression and translocation of protein kinase C isoforms α, ßII, δ, ε and ζ were analyzed by immunohistochemical staining and immunoblot analysis after tissue fractionation. Changes in calcineurin signalling were identified by immunoblot analysis and functional assays. Expression levels of transcription factors GATA4 and NF-κB were quantified by real-time RT-PCR. RESULTS: Diabetic wildtype mice developed myocardial hypertrophy with preserved cardiac function. Calcineurin signalling was not different between the two groups. However, diabetic wildtype mice showed increased protein levels of PKC-α and PKC-ζ, translocation of PKC-α, -δ and -ε to cellular membranes and higher levels of NF-κB expression. In contrast, diabetic Gα11-knockout mice showed no altered phenotype and no changes in NF-κB or PKC expression, although translocation of PKC-ε occurred as in wildtypes. CONCLUSIONS: Gα11 is essential for the development of cardiac hypertrophy in type 1-diabetes. Stimulation of hypertrophic signalling through PKC-α, PKC-δ, PKC-ζ, and NF-κB appears to be receptor-dependent, whereas PKC-ε is activated by hyperglycemia, independent of Gα11.

Ablation of cyclase-associated protein 2 (CAP2) leads to cardiomyopathy.

Cyclase-associated proteins are highly conserved proteins that have a role in the regulation of actin dynamics. Higher eukaryotes have two isoforms, CAP1 and CAP2. To study the in vivo function of CAP2, we generated mice in which the CAP2 gene was inactivated by a gene-trap approach. Mutant mice showed a decrease in body weight and had a decreased survival rate. Further, they developed a severe cardiac defect marked by dilated cardiomyopathy (DCM) associated with drastic reduction in basal heart rate and prolongations in atrial and ventricular conduction times. Moreover, CAP2-deficient myofibrils exhibited reduced cooperativity of calcium-regulated force development. At the microscopic level, we observed disarrayed sarcomeres with development of fibrosis. We analyzed CAP2's role in actin assembly and found that it sequesters G-actin and efficiently fragments filaments. This activity resides completely in its WASP homology domain. Thus CAP2 is an essential component of the myocardial sarcomere and is essential for physiological functioning of the cardiac system, and a deficiency leads to DCM and various cardiac defects.

Transcription factor CREB is phosphorylated in human molar odontoblasts and cementoblasts in vivo.

A wide variety of stimuli can trigger activation of the transcription factor CREB (cAMP-responsive element binding protein), pointing toward a central role for CREB in the integration of various signaling inputs. No data are available on the expression and phosphorylation of CREB in mammalian teeth. Using immunohistochemical analysis of free-floating sections, we show here that CREB was strongly expressed and phosphorylated at Ser-133 within the nucleus of a subpopulation of adult human molar odontoblasts. Many dental pulp stromal cells and periodontal ligament fibroblasts expressed CREB and showed phosphorylation of CREB at Ser-133. In addition, cementoblasts displayed nuclear expression and phosphorylation of CREB at Ser-133. The epithelial rests of Malassez revealed strong nuclear expression of CREB, but phosphorylation at Ser-133 was variable. Our results provide the first evidence that the constitutively phosphorylated transcription factor CREB is involved in the biomineralization process of adult human molar odontoblasts and cementoblasts.

Intrinsic proinflammatory signaling in podocytes contributes to podocyte damage and prolonged proteinuria.

Inflammation conveys the development of glomerular injury and is a major cause of progressive kidney disease. NF-κB signaling is among the most important regulators of proinflammatory signaling. Its role in podocytes, the epithelial cells at the kidney filtration barrier, is poorly understood. Here, we inhibited NF-κB signaling in podocytes by specific ablation of the NF-κB essential modulator (NEMO, IKKγ). Podocyte-specific NEMO-deficient mice (NEMO(pko)) were viable and did not show proteinuria or overt changes in kidney morphology. After induction of glomerulonephritis, both NEMO(pko) and control mice developed significant proteinuria. However, NEMO(pko) mice recovered much faster, showing rapid remission of proteinuria and restoration of podocyte morphology. Interestingly, quantification of infiltrating macrophages, T-lymphocytes, and granulocytes at day 7 revealed no significant difference between wild-type and NEMO(pko). To further investigate the underlying mechanisms, we created a stable NEMO knockdown mouse podocyte cell line. Again, no overt changes in morphology were observed. Translocation of NF-κB to the nucleus after stimulation with TNFα or IL-1 was sufficiently inhibited. Moreover, secretion of proinflammatory chemokines from podocytes after stimulation with TNFα or IL-1 was significantly reduced in NEMO-deficient podocytes and in glomerular samples obtained at day 7 after induction of nephrotoxic nephritis. Collectively, these results show that proinflammatory activity of NF-κB in podocytes aggravates proteinuria in experimental glomerulonephritis in mice. Based on these data, it may be speculated that immunosuppressive drugs may not only target professional immune cells but also podocytes directly to convey their beneficial effects in various types of glomerulonephritis.

Tertiary lymphoid organ development coincides with determinant spreading of the myelin-specific T cell response.

While the role of T cells has been studied extensively in multiple sclerosis (MS), the pathogenic contribution of B cells has only recently attracted major attention, when it was shown that B cell aggregates can develop in the meninges of a subset of MS patients and were suggested to be correlates of late-stage and more aggressive disease in this patient population. However, whether these aggregates actually exist has subsequently been questioned and their functional significance has remained unclear. Here, we studied myelin basic protein (MBP)-proteolipid protein (PLP)-induced experimental autoimmune encephalomyelitis (EAE), which is one of the few animal models for MS that is dependent on B cells. We provide evidence that B cell aggregation is reflective of lymphoid neogenesis in the central nervous system (CNS) in MBP-PLP-elicited EAE. B cell aggregation was present already few days after disease onset. With disease progression CNS B cell aggregates increasingly displayed the phenotype of tertiary lymphoid organs (TLOs). Our results further imply that these TLOs were not merely epiphenomena of the disease, but functionally active, supporting intrathecal determinant spreading of the myelin-specific T cell response. Our data suggest that the CNS is not a passive "immune-privileged" target organ, but rather a compartment, in which highly active immune responses can perpetuate and amplify the autoimmune pathology and thereby autonomously contribute to disease progression.

The extent of ultrastructural spinal cord pathology reflects disease severity in experimental autoimmune encephalomyelitis.

Experimental autoimmune encephalomyelitis (EAE) has been studied for decades as an animal model for human multiple sclerosis (MS). Here we performed ultrastructural analysis of corticospinal tract (CST) and motor neuron pathology in myelin oligodendrocyte glycoprotein (MOG) peptide 35-55- and MP4-induced EAE of C57BL/6 mice. Both models were clinically characterized by ascending paralysis. Our data show that CST and motor neuron pathology differentially contributed to the disease. In both MOG peptide- and MP4-induced EAE pathological changes in the CST were evident. While the MP4 model also encompassed severe motor neuron degeneration in terms of rough endoplasmic reticulum alterations, the presence of intracytoplasmic vacuoles and nuclear dissolution, both models showed motor neuron atrophy. Features of axonal damage covered mitochondrial swelling, a decrease in nearest neighbor neurofilament distance (NNND) and an increase of the oligodendroglial cytoplasm inner tongue. The extent of CST and motor neuron pathology was reflective of the severity of clinical EAE in MOG peptide- and MP4-elicited EAE. Differential targeting of CNS gray and white matter are typical features of MS pathology. The MOG peptide and MP4 model may thus be valuable tools for downstream studies of the mechanisms underlying these morphological disease correlates.

Adaptive immune response to model antigens is impaired in murine leukocyte-adhesion deficiency-1 revealing elevated activation thresholds in vivo.

Absence of ß2 integrins (CD11/CD18) leads to leukocyte-adhesion deficiency-1 (LAD1), a rare primary immunodeficiency syndrome. Although extensive in vitro work has established an essential function of ß2 integrins in adhesive and signaling properties for cells of the innate and adaptive immune system, their respective participation in an altered adaptive immunity in LAD1 patients are complex and only partly understood in vivo. Therefore, we investigated adaptive immune responses towards different T-dependent antigens in a murine LAD1 model of ß2 integrin-deficiency (CD18⁻/⁻). CD18⁻/⁻ mice generated only weak IgG responses after immunization with tetanus toxoid (TT). In contrast, robust hapten- and protein-specific immune responses were observed after immunization with highly haptenated antigens such as (4-hydroxy-3-nitrophenyl)21 acetyl chicken γ globulin (NP21-CG), even though regularly structured germinal centers with specificity for the defined antigens/haptens in CD18⁻/⁻ mice remained absent. However, a decrease in the hapten/protein ratio lowered the efficacy of immune responses in CD18⁻/⁻ mice, whereas a mere reduction of the antigen dose was less crucial. Importantly, haptenation of TT with NP (NP-TT) efficiently restored a robust IgG response also to TT. Our findings may stimulate further studies on a modification of vaccination strategies using highly haptenated antigens in individuals suffering from LAD1.

The magnitude of the antigen-specific T cell response is separated from the severity of spinal cord histopathology in remitting-relapsing experimental autoimmune encephalomyelitis.

Multiple sclerosis (MS) is an autoimmune disorder of the central nervous system. The remitting-relapsing experimental autoimmune encephalomyelitis (EAE) in the SJL mouse strain is a common animal model for MS and similar to the human disease it is considered to be T helper cell mediated. Besides interferon-γ secreting T(H)1 cells in particular the T(H)17 subset is believed to be highly pathogenic. Spreading of the T(H)1 and T(H)17 response to newly emerging determinants has been used to explain clinical disease relapse, but if the magnitude of the T(H)1/T(H)17 response is linked to clinical relapse severity has remained unresolved. Here, we assessed clinical EAE severity, the extent of spinal cord histopathology and the magnitude of the antigen-specific T helper cell and autoantibody response in proteolipid protein peptide 139-151 (PLP:139-151)-immunized SJL mice in clinical remission and relapse. We demonstrate that spinal cord histopathology comprised inflammation, demyelination as well as axonal loss and correlated well with clinical disease severity. Although the degree of spinal cord histopathology and clinical severity was separated from the PLP:139-151-specific T(H)1/T(H)17 cell and antibody response, it was linked to the number of infiltrating macrophages and activated microglia. In particular, there was a correlation between their secretion product interleukin-1ß and the degree of axonal loss. Although CD4(+) T cells seem to be mainly involved in disease initiation, we suggest that it is the downstream activation of the innate immune response that defines the magnitude of the disease outcome.

Histone deacetylases 2 and 9 are coexpressed and nuclear localized in human molar odontoblasts in vivo.

Histone deacetylases (HDACs) are components of nuclear multiprotein complexes that deacetylate histones and perform important roles in repression of transcription.Using specific rabbit mAbs, we analyzed by immune histochemistry and confocal immunofluorescence analysis the expression and subcellular localization of HDAC1­4 and HDAC9 in sections of adult human third molars. HDAC2 and HDAC9 were expressed in some pulpal cells and strongly expressed in the majority of mature odontoblasts.In contrast, only weak expression of HDAC1, HDAC3 and HDAC4 was observed. Confocal immunofluorescence analysis together with the DNA stain DRAQ5 revealed that HDAC2 and HDAC9 were coexpressed within the odontoblast nucleus, but localized to distinct subnuclear structures.In contrast to the current point of view, HDAC2 is strongly expressed in a terminally differentiated cell type.Our results imply that class I and II HDACs are involved in the transcriptional regulation of human odontoblasts in vivo.

Experimental autoimmune encephalomyelitis--achievements and prospective advances.

Multiple sclerosis (MS) is an autoimmune disorder of the CNS. Different subtypes of the disease have been noted, and characterized by distinct clinical courses and histopathologic manifestations. The most intensively studied animal model of MS, experimental autoimmune encephalomyelitis (EAE), classically leads to deficits in motor functions, and is mediated by T helper cells. Recently, T(H)17 cells were ascribed an even greater pathogenic impact than T(H)1 cells, but new findings render this view controversial. Although classic EAE has been an invaluable tool, it does not cover the entire pathogenic entity of MS. Especially B-cell contribution and autoantibody-dependence are not mirrored adequately: therefore, new B-cell-dependent models, such as MP4-induced EAE, have been introduced. Furthermore, certain symptoms and the spontaneous onset of MS are not featured in classic EAE. Herein, atypical and spontaneous EAE models can be used for investigation of common symptoms, such as tremor and ataxia, as well as spontaneous disease development. MS displays a marked inter-individual heterogeneity, and no single model will be able to cover all features. Thus, depending on the objective of one's study, the appropriate EAE model has to be carefully chosen. In addition, refined models should be designed to gain a more complete understanding of MS.