Funny & Jokes

From transcriptome analysis to therapeutic anti-CD40L treatment in the SOD1 model of amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by progressive loss of motor neurons. Using unbiased transcript profiling in an ALS mouse model, we identified a role for the co-stimulatory pathway, a key
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  392 VOLUME 42 | NUMBER 5 | MAY 2010 NATURE GENETICS ARTICLES Developing drugs, from target discovery to commercialization, is inef-ficient and has a high attrition rate 1 . This problem is acute for drugs treating orphan diseases such as ALS. Completion of the human and mouse genomes paved the way for genome-wide transcriptional pro-filing and streamlining of the arduous process of drug development. Since then, genome-wide transcript profiling has profoundly affected biomarker identification 2–6 , phenotypic classification of disease 7,8  and toxicology  9,10 . It has not, however, built a strong bridge between drug target discovery and preclinical pharmacology. Here we report on efforts to prioritize molecular pathways within genome-wide transcription profiles in a transgenic mouse model of familial ALS and to determine the therapeutic benefit of an immunomodulatory biologic, anti-CD40L.To understand the interplay of molecular pathways underlying the progression of ALS, we completed an exhaustive survey of transcrip-tional changes in diseased tissues of SOD1 G93A  (B6SJL-Tg( SOD1 -G93A)1Gur) transgenic mice and applied a network-based analysis strategy  11 . An unbiased gene set enrichment analysis of all known molecular pathways activated presymptomatically revealed a T cell co-stimulatory signature in spinal cord, skeletal muscle and peripheral nerves. Activation of an immune response by antigen-presenting cells (APCs) is mediated by engagement of CD86 and CD40 receptors on APCs with CD28 and CD40L on the surface of T cells 12–15 . Beneficial effects of blocking CD40L have been observed in preclinical models of transplantation, multiple sclerosis, arthritis, systemic lupus erythema-tosus and Alzheimer’s disease 16–26 . To determine whether anti-CD40L would affect disease progression in the SOD1 G93A  mouse model, we administered a monoclonal antibody to CD40L, MR1, to SOD1 G93A  mice. Treatment with anti-CD40L delayed paralysis onset, improved body-weight maintenance and extended survival. Furthermore, the treatment with anti-CD40L antibody reduced markers of neuro-inflammation, increased the number of motor neuron cell bodies, decreased peripheral nervous system (PNS) inflammatory markers and downregulated expression of co-stimulatory genes in spinal cord. Thus, early ALS pathology is amenable to immunomodulatory strate-gies that affect disease progression in the central nervous system (CNS) and the periphery. More broadly, this study of a preclinical mouse model of familial ALS exemplifies how a high-resolution, unbiased transcriptome survey throughout the time course of a disease can lead to the identification of therapeutically tractable cellular pathways. RESULTSIdentifying molecular pathways changing at disease onset in ALS We performed an unbiased survey of the temporal changes in gene expression from the brain, spinal cord, skeletal muscle, sciatic nerve, blood and adipose tissue of SOD1 G93A  mice to identify relevant pathways. Tissues were harvested from SOD1 G93A  mice ( n  = 5) and their non-transgenic littermates ( n  = 5) at intervals of 10 d starting presymptomatically (day 30) until the end stage of disease (day 120). We identified pathways associated with onset and progression using a gene set enrichment algorithm, GlobalTest, querying a database of 498 biological pathways from the Kyoto Encyclopedia of Genes and Genomes and Biocarta 27 . As a measure of covariance in SOD1 G93A  compared with non-transgenic littermates, we calculated the Q  score statistic for each pathway, tissue and time point.Generally, the Q  score represents the difference between the geomet-ric means of two sample groups for a set of defined genes. Technically, the Q  score is a measure of how closely the covariance matrix of the set of genes is correlated with the two sample groups. The stronger the correlation between the average gene expression for a group of genes From transcriptome analysis to therapeutic anti-CD40L treatment in the SOD1  model of amyotrophic lateral sclerosis John M Lincecum 1,2 , Fernando G Vieira 1,2 , Monica Z Wang  1 , Kenneth Thompson 1 , Gerald S De Zutter 1 , Joshua Kidd 1 , Andrew Moreno 1 , Ricardo Sanchez 1 , Isarelis J Carrion 1 , Beth A Levine 1 , Bashar M Al-Nakhala 1 , Shawn M Sullivan 1 , Alan Gill 1  & Steven Perrin 1 Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by progressive loss of motor neurons. Using unbiased transcript profiling in an ALS mouse model, we identified a role for the co-stimulatory pathway, a key regulator of immune responses. Furthermore, we observed that this pathway is upregulated in the blood of 56% of human patients with ALS. A therapy using a monoclonal antibody to CD40L was developed that slows weight loss, delays paralysis and extends survival in an ALS mouse model. This work demonstrates that unbiased transcript profiling can identify cellular pathways responsive to therapeutic intervention in a preclinical model of human disease. 1 ALS Therapy Development Institute, Cambridge, Massachusetts, USA. 2 These authors contributed equally to this work. Correspondence should be addressed to S.P. ( 22 December 2009; accepted 2 March 2010; published online 28 March 2010; doi:10.1038/ng.557  NATURE GENETICS   VOLUME 42 | NUMBER 5 | MAY 2010 393 ARTICLES and the expected group classification (for example, genotype), the higher the Q  score. The distribution of Q  scores for all tissue combinations and individual tissues analyzed separately, plotted as a function of age, reflects dynamic changes in pathways throughout disease progression ( Supplementary Fig. 1a – d  shows representative individual tissue distributions). We identified pathways with statistically significant Q  scores by calculat-ing a distribution of Q  scores for each time point and tissue ( Supplementary Fig. 1a – d  and Supplementary Table 1 ) and established a cutoff threshold for statistical significance by randomizing all samples and recalculating Q  scores ( Supplementary Fig. 1e – g   and Supplementary Table 1 ). Sample randomization demonstrated that Q  scores >40 with P   values <0.05 represent a statistically significant difference between the SOD1  and non-transgenic genotypes at any time point for any tissue.Filtering the Q  scores >40 with P   < 0.05 for 80 and 90 days of age across skeletal muscle, sciatic nerve, spinal cord, brain, adipose tissue and blood identified no pathways differentially regulated in all tissues. Therefore, we tested all 64 unique combinations of the tissues using the same Q  score filter. The largest unique set of tissues meeting the criteria was composed of skeletal muscle, spinal cord and sciatic nerve. We found the biological networks and genes that were significantly changing presymptomatically and at symptom onset by filtering the network for pathways having Q  scores greater than 40 in all three tissues at days 80 and 90. The resulting subnetwork contains five bio-logical pathways and 95 genes associated with these five pathways ( Supplementary Tables 2 and  3 ). Two pathways are associated with T cell activation, two with macrophage activation, and one pathway contains genes involved in co-stimulatory regulation of the adaptive and innate immune systems.The co-stimulatory pathway forms a bridge between the activa-tion of T cell responses and the amplification of the innate immune response. Using genMapp, we mapped the expression pattern of the genes in the co-stimulatory pathway for skeletal muscle, spinal cord and sciatic nerve ( Fig. 1a – c , respectively). The expression patterns are represented colorimetrically as fold changes; non-transgenic animals are compared with SOD1 G93A  animals from day 50 through end-stage disease (day 120) at 10-d intervals. There was a progres-sive increase in CD86a, a B7 receptor on APCs 20 , and in adhesion molecules (Itgam, Itgax, Cd44, Icam, Selpl and Itgb2) induced on APCs that facilitate lymphocyte recruitment and extravasation. Proinflammatory chemokines (Ccl3, Ccl6, Ccl9, Cxcl10 and Ccl12) were also upregulated in the three tissues during disease progression a B cellAPC TH cell bcd B cellAPC TH cell B cellAPC ControlALSSample groups TH cell Expression data setColor sets: gsD50gsD50 highgsD60gsD80gsD90gsD100gsD110 Legend: qsD50 gsD50 mediumgsD50 lowgsD50 down low    C  o  -  s   t   i  m  u   l  a   t  o  r  y  g  e  n  e  s gsD50 down medgsD50 down highNo criteria metNot found Expression data setColor sets: scd50scd50 highscd60scd80scd90scd100scd110 Legend: scd50 scd50 mediumscd50 lowscd50 NCscd50 down lowscd50 down mediumscd50 down highNo criteria metNot found Expression data setColor sets: SNd50 high Legend: SNd50 down high SNd50 medSNd50 lowSNd50 down lowSNd50 down medSNd50 down high SNd50 down highSNd80 down highSNd90 down highSNd100 down highSNd110 down high No criteria metNot found Figure 1  Co-stimulatory pathway signaling in SOD1 G93A  skeletal muscle, spinal cord and sciatic nerve is upregulated during disease progression and is increased in a subset of blood samples from individuals with ALS. We assessed the statistical differences in gene expression between SOD1 G93A  and wild-type medial gastrocnemius, spinal cord and sciatic nerve at a given time point using the nonlinear Bayes methodology in the Bioconductor LIMMA package and overlaid the resulting analysis onto the genMAPP pathway for co-stimulatory signaling (genMAPP: Mm-Std_20060628.gdb). Each colored box represents the estimated change in transcript abundance for a given gene after normalization and data summarization. Biological replicates for the SOD1 G93A  transgenic mice are n   = 5. Age-matched, non-transgenic mice are biological replicates of n   = 5. The key details the relative increase in transcript expression levels, with blue intensity representing decreasing levels and pink intensity representing increasing levels. ( a ) Gastrocnemius. ( b ) Spinal cord. ( c ) Sciatic nerve. ( d ) A heat-map visualization of the normalized expression data for genes of the human T cell co-stimulatory pathway from 27 non-ALS and 63 ALS blood samples. We normalized Affymetrix gene chip data using robust multi-array averaging and clustered samples using self-organizing maps. Gene expression intensity is represented colorimetrically, with low expression represented as blue rectangles and high expression as pink rectangles.  394 VOLUME 42 | NUMBER 5 | MAY 2010 NATURE GENETICS ARTICLES ( Fig. 1a – c  and Supplementary Tables 2 and  3 ). Increased expression of these chemokines suggests recruitment of APCs in these tissues, mediated by co-stimulatory pathway signaling, as APCs and T cells interact. The genome-wide survey of gene expression changes in mul-tiple tissues, before and during disease onset, supports the hypothesis that the co-stimulatory pathway is important in disease pathogenesis in this mouse model of ALS.We investigated the potential relevance of these findings to human patients with ALS by analyzing the co-stimulatory pathway in peripheral blood samples from affected individuals ( n  = 63) and non-ALS control samples ( n  = 27). A heat-map visualization of the normalized transcript expression data shows the co-stimulatory pathway is upregulated in 56% of ALS blood samples (35 of 63) compared to non-ALS controls ( Fig. 1d , Supplementary Table 5  and Supplementary Note ). Co-stimulatory genes in immune cells of the CNS and PNS In an effort to localize expression of co-stimulatory pathway genes to specific cell types, we performed either in situ  hybridization in the spi-nal cord or immunohistochemistry in the sciatic nerves of SOD1 G93A  mice. In situ  hybridization for representative co-stimulatory genes at day 90 in SOD1 G93A  and wild-type spinal cord showed an increase in expression of CD44 in glial cells surrounding the axial and triceps muscle motor neurons of lamina IX. Additionally, the chemokine Ccl3 is expressed in the glial cells of the external urethral sphincter, biceps femoralis and gluteal muscle motor neurons of lamina IX, as is the chemokine Cxcl10 in the biceps femoralis motor neuron pool ( Supplementary Fig. 2 ).To rule out a nonspecific inflammatory response resulting from muscle wasting or late-stage neural degeneration in the periphery, we visualized expression of CD68, an antigen associated with APCs, using immunohistochemistry in sciatic nerve. We observed spindle-shaped CD68 +  cells starting at day 50, 30 d before symptom onset ( Fig. 2a ). These cells have the morphology of resident or endoneurial macrophages. At day 60 the population of ramified CD68 +  cells was replaced with smaller numbers of foamy CD68 +  cells with prominent intracellular granules ( Fig. 2b ). The proliferation of a foamy macro-phage population is apparent at days 80 to 110. These cells have a ‘bead on a string’ appearance, suggesting that the accumulation of macrophages is specific to the axonal tracks ( Fig. 2c – e ).We quantified accumulation of macrophages in the peripheral nerves of the gastrocnemius by a morphometric analysis of CD68 +  macrophages within skeletal-muscle nerves. Briefly, we normalized CD68 +  cell counts relative to the total area of S100b +  nerve tissue in a random field for each biological replicate and each age. Controls were non-transgenic, age-matched medial gastrocnemius tissue proc-essed in parallel. Complete details of the counting protocol and the one-way analysis of variance are described in the Online Methods. At the presymptomatic stage on day 60, the average number of macro-phages was sixfold greater ( P   = 0.0069) ( Fig. 2f  ) in SOD1 G93A  than in non-transgenic tissue. At day 80, with the onset of clinical hindlimb tremor, there were fourfold more macrophages in SOD1 G93A  relative to controls ( P   = 0.0312). At day 100, 16-fold more ( P   = 0.1814) macro-phages had accumulated in distal nerves. The increased variance at this time point appears to result from higher variability between individual animals, consistent with differences in clinical progres-sion toward fulminant neurodegeneration. At the final time point of day 110, all mice displayed hindlimb paralysis, and the number of macrophages in the S100b +  nerves was 25-fold higher ( P   = 0.0077) than in controls.The progressive accumulation of CD68 +  macrophages in the peripheral nerves during all phases of the disease shows that a chronic inflammatory condition exists in ALS motor neurons and their axons. Neuroinflammation, characterized by activation of astrocytes and microglia in the central nervous system in ALS, has been well described 28–34 ; however, the accumulation of inflammatory cells of the monocyte lineage in the presymptomatic peripheral nerves of the limbs indicates that cells of the monocyte lineage are targeted to the motor axons. Effects of CD40L monoclonal antibody in vivo The co-stimulatory pathway has been implicated in autoimmune and graft-versus-host disease. Blocking the activation of the co-stimulatory pathway by inhibiting either CD40 ligand or CD86 function has therapeutic benefit in preclinical models of autoimmune disease and allograft transplantation, as well as clinical benefit in rheumatoid arthritis 35–39 . We tested the effects of blocking CD40L-dependent co-stimulation in the SOD1 G93A  model by treating transgenic mice with the CD40L-blocking monoclonal antibody MR1 (ref. 40). Pharmacokinetic analysis of anti-CD40L plasma concentration versus time in male and female SOD1 G93A  mice after a single 5-mg-per-kg (body weight) intraperitoneal injection of anti-CD40L indicated that the elimination half-life, in both genders, was about 23 d. However, the apparent volume of distribution in males was greater than in females ( Supplementary Fig. 3 ). The pharmacokinetic data differ from previ-ously published studies of Balb/C or SNF1 mice, in which the elimina-tion half-life of anti-CD40L was estimated to be approximately 7 d 41 . On the basis of these data, we designed a survival study using dosing regimens that would maintain plasma trough anti-CD40L levels above 10 µ g/ml in both males and females 42 . The rationale for maintaining plasma levels above 10 µ g/ml is based on mouse models with marked inflammatory disease, such as lupus and relapsing-remitting experi-mental autoimmune encephalomyelitis mouse models. The dosing regimen consisted of an anti-CD40L loading dose of 5.22 mg per kg (body weight) in females and 6.75 mg per kg in males, given at the a b c d e f 3.000WT2.5002.0001.5001.0000.50005060708090100110120SODAge (d)    M  a  c  r  o  p   h  a  g  e  s  p  e  r   1 ,   0   0   0      µ   m    2 Figure 2  Macrophages accumulate in peripheral nerves throughout the disease course. Shown is a time course of immunofluorescence staining for CD68 +  macrophage (green) and DAPI-counterstained nuclei (blue) in sciatic nerve. ( a ) Day 50. ( b ) Day 60. ( c ) Day 80. ( d ) Day 100. ( e ) Day 110. Scale bars in a – e , 17 µ m. ( f ) Quantification of accumulated CD68 +  macrophages in the S100b +  nerves of the gastrocnemius. WT, wild type.  NATURE GENETICS   VOLUME 42 | NUMBER 5 | MAY 2010 395 ARTICLES start of the study on day 50, followed by weekly injections of 1.0 and 1.34 mg per kg in females and males, respectively (treatment n  = 44, vehicle n  = 45). This regimen maintained circulating levels of anti-CD40L above the target of 10 µ g/ml. To determine whether chronic administration would result in accumulation of anti-CD40L in the spinal cord, we dosed SOD1 G93A  mice with 4 weeks of maintenance dosing. Anti-CD40L levels ranged from 200–1,000 ng/ml, or 2–10% of circulating plasma levels (data not shown). All control mice received 10 ml per kg sterile saline intraperitoneally. Previously, we had dem-onstrated that a control IgG is equivalent to saline in having no ben-eficial or toxic effect relative to untreated control SOD1 G93A  (data not shown). We monitored body weight and neurological score daily to assess disease progression, survival duration and any therapeutic benefits of anti-CD40L treatment.Mutant SOD1 G93A  transgenic mice have normal body weight charac-teristics as neonates, and gain weight normally into young adulthood, compared to non-transgenic animals. Weight loss, characteristically, becomes apparent in adult animals during the onset of symptomatic disease and continues until death. Analysis of comparative weight loss in treatment and control groups can reveal the effects of treatment on disease onset and rate of progression. We examined the impact of anti-CD40L treatment on body weight using two parameters: (i) changes in body weight from study start to the attainment of peak body weight, and (ii) changes in body weight from peak body weight until death. Treatment with anti-CD40L did not affect attainment of peak body weight. However, it did produce a statistically significant 6-d delay in body-weight decline from peak body weight until death ( Fig. 3a , b  and Table 1 ).The ALS Therapy Development Institute (ALS TDI) has developed a neurological scoring system that assesses the time of disease onset and the rate of disease progression for all animals in each study group on the basis of progressive paralysis 43 . The age at onset of paralysis, for each animal, is captured by daily assessment of neurological score and is taken to be the age at which the animal reaches a neurological score of 2. Animals with a neurological score of 2 show curling of the toes and dragging of at least one limb during walking. Anti-CD40L treat-ment statistically significantly delayed paralysis onset by 8 d. Median time to attain a neurological score of 2 was 113 and 121 d in control and treatment groups, respectively ( Fig. 3c  and Table 1 ; P   = 0.0038).Time-to-event analysis of the survival data revealed that survival was significantly lengthened, by 9 d, with anti-CD40L treatment in SOD1 G93A  animals ( Fig. 3d  and Table 1 ; P   = 0.0043). Analysis of the effects of anti-CD40L treatment on female SOD1 G93A  animals showed a statistically significant delay in paralysis onset (7 d, P   = 0.038) and a 40% extension of median life duration after the first signs of paralysis ( P   = 0.030). The overall lengthening of survival for female SOD1 G93A  animals was 13 d ( P   = 0.004).To determine the significance of a 9-d lengthening of survival, we compared these data to a Monte Carlo simulation of historical control groups from a database of over 3,000 animals to estimate the chance of randomly detecting a 9-d change in survival, as previously described 44 . We randomly assigned 44 animals either to a mock control vehicle group or a mock treatment group, balancing age and gender appropri-ately, and performed a Kaplan-Meier analysis. We iteratively repeated the process 1,000 times and plotted the distributions of observed versus simulated percent change in survival. According to this analy-sis, the probability of randomly observing a 9-d change in survival (7% longer median survival) approaches 0% ( Fig. 4a ). We also com-pared our survival data with survival data, generated by ALS TDI, for mice treated with the US Food and Drug Administration–approved ALS therapeutic riluzole and with apocynin 45 . Although riluzole and apocynin subtly lengthen survival (by 3 and 5 d, respectively), the effects are not statisti-cally significant, whereas treatment with anti-CD40L lengthens survival by 9 d compared to historical controls, a statistically significant effect ( Fig. 4b ).In an attempt to extend these studies, we undertook fully powered efficacy studies starting after the onset of symptoms (day 80). These studies showed that anti-CD40L did not have a significant therapeutic effect Table 1 Time-to-event analysis Kaplan-Meier survival fitCox proportional hazard fitMedian time (d) P   valueEffect of treatment P   valueControlDrugChangeLog-rankWilcoxonHazard ratios.e.m.Likelihood testInitial to peak body weight4953+40.35060.15630.790.180.2860Peak body weight to death2935+60.04130.07320.640.140.0460Disease onset age113121+80.00380.00170.440.110.0010Survival (age at death)124133+90.00430.00400.490.120.0030 Drug, anti-CD40L treatment; control, vehicle alone. See Online Methods for a description of statistical techniques. 1.000.75    P  r  o  p  o  r   t   i  o  n  a   t  p  e  a   k   P  r  o  p  o  r   t   i  o  n  o  n  s  e   t  o  r  s  u  r  v   i  v  a   l Time (d) Time (d) Age (d)Age (d)0.500.2501.000.750.500.250Ctrl. ac db Drug30405060708015110951051151251351451201301401502535455565Ctrl.DrugCtrl.DrugCtrl.Drug Figure 3  Blocking CD40L with a monoclonal antibody to CD40L improves body-weight maintenance, delays disease onset and extends survival in SOD1  mice. SOD1 G93A  mice received a loading dose (5.22 mg per kg body weight for females, 6.75 mg per kg for males) followed by weekly maintenance doses (1 mg per kg for females, 1.34 mg per kg for males) given intraperitoneally beginning at 50 days of age and continuing until death. ( a ) Kaplan-Meier time-to-event analysis for time required to attain peak body weight. Time to peak was not significantly ( P   = 0.35) changed by anti-CD40L treatment. Ctrl., control; Drug, anti-CD40L. ( b ) Time-to-event analysis for the time from peak body weight until death. Body-weight maintenance was significantly ( P   = 0.0413) improved by anti-CD40L treatment. ( c ) Time-to-event analysis for disease onset, the age at which mice first showed signs of definitive neurological disease (neurological severity score of 2). Disease onset was significantly ( P   = 0.0038) delayed by anti-CD40L treatment. ( d ) Time-to-event analysis for survival, the age at which mice died. Survival was significantly ( P   = 0.0043) prolonged by anti-CD40L treatment. Results of statistical analyses are presented in Table 1 .  396 VOLUME 42 | NUMBER 5 | MAY 2010 NATURE GENETICS ARTICLES on disease onset, progression or survival. To provide further perspec-tive on the efficacy of anti-CD40L, we conducted outlier analyses relative to other drugs tested at ALS TDI, including minocycline, celecoxib and ceftriaxone, on median time to death in the SOD1 G93A  ALS model. The jackknife distance test indicates that the percent change of median survival in the SOD1 G93A  mouse model caused by anti-CD40L treatment is well above the upper confidence limit of the assay, as defined by data from 15 other drug studies in the model ( Fig. 4c  and Supplementary Table 4 ). Previous work has demonstrated that blocking CD40L signaling is efficacious in preclinical models of allograft transplant and autoimmune disease. This work now shows that blocking CD40L-dependent co-stimulation is also efficacious in a preclinical model of ALS. Pharmacodynamic markers of anti-CD40L response We investigated the mechanism of action of anti-CD40L in SOD1 G93A  animals using five approaches. First, we performed an immunohisto-chemical analysis of the effect of anti-CD40L on infiltrating CD68 +  cells in sciatic nerves. Anti-CD40L treatment significantly low-ered, by 42%, the average number of infiltrating CD68 +  cells in the sciatic nerve at day 100, compared to controls and non-transgenic littermates ( Fig. 5a – c ).It has been shown that anti-CD40L treatment in allograft trans-plant models decreases the number of cytotoxic T cells in lymph nodes 22 . In a second analysis, we harvested the sciatic lymph nodes of anti-CD40L–treated SOD1  mice, PBS-treated SOD1 G93A  mice and non-transgenic littermates at day 80, 30 d after initiation of anti-CD40L treatment. We subjected sciatic lymph node lymphocytes to FACS analysis for CD4 +  and CD8 +  T cells. SOD1 G93A  animals had a 20% reduction in the frequency of CD8 +  T cells compared to non-transgenic littermates at day 80 ( Fig. 5d , e ). Treatment with anti-CD40L lowered the frequency of CD8 +  T cells in SOD1 G93A  sciatic lymph node by 51% com-pared to controls ( Fig. 5f  ). This suggests that anti-CD40L treatment shifts the CD4/CD8 ratio and thus affects T cell function.Third, we investigated the impact of anti-CD40L treatment on neuroinflammation in the central nervous system by immunohisto-chemical analysis for markers of astrocytosis and microgliosis in the spinal cords of mice receiving 50 d of anti-CD40L, compared to control mice ( Fig. 6 ) The control group ( Fig. 6a ) had a substantial amount of glial fibrillary acidic protein (Gfap)–positive stain-ing in what appear to be activated astrocytes. In contrast, the spinal cord tissue from the anti-CD40L–treated group had lower levels of Gfap staining ( Fig. 6b ). We assayed microglial activation by staining for the marker protein galectin-3 (Mac-2). Figure 6c  shows Mac-2 staining of microglia in untreated control spinal cord tissue, with the cells extending 600 a b c 60%50%40%30%20%10%Effect frequency% false positives50040030020010010043210DrugHistorical female controlRiluzole treated femaleApocynin treated femaleMR1 treated femaleanti-CD40L    D   i  s   t  a  n  c  e 100120140Age (days)160500   –   3   0   t  o  –   2   5   %   –   2   5   t  o  –   2   0   %   –   2   0   t  o  –   1   5   %   –   1   5   t  o  –   1   0   %   –   1   0   t  o  –   5   %   –   5   t  o   0   %   0   t  o   5   %   5   t  o   1   0   %    1   0   t  o   1   5   %    1   5   t  o   2   0   %    2   0   t  o   2   5   %    2   5   t  o   3   0   % 00%    P  e  r  c  e  n   t  s  u  r  v   i  v  a   l Figure 4  Meta-analysis of anti-CD40L treatment compared with riluzole, apocynin and historical controls. ( a ) Monte Carlo analysis of historical control SOD1  animals to assess the probability of randomly detecting a 9-d lengthening of survival. We randomly assigned 44 untreated SOD1 G93A  animals to either a mock control group or a mock treatment group and performed Kaplan-Meier survival analysis. We performed 1,000 iterations and plotted the frequency distribution as a function of percent change in survival. ( b ) The Kaplan-Meier survival data from historical control groups compared with a group treated with weekly intraperitoneal injection of 1 mg per kg (body weight) anti-CD40L. ( c ) Outlier analysis shows that anti-CD40L results do not fall within the distribution of results from other drugs tested in the SOD1 G93A  model. The jackknife outlier distance for each observation is calculated using estimates of the mean, s.d. and correlation matrix that do not include the observation itself. Extreme multivariate outliers are identified as points that exceed the upper distance value limit (dotted line). Results are tabulated by drug in Supplementary Table 4 . a bd e fc (–)MR1 SOD1 (+)MR1WT(–)MR1 400 n   = 5 n   = 4 n   = 3300    C   D   6   8  +  c  e   l   l  c  o  u  n   t  s 200100010 4 10 4 10 3 10 3 510.793711.210 2 10 2 10 1    F   L   2  -   H  :   C   D   8   P   E   F   L   2  -   H  :   C   D   8   P   E FL3-H: CD4 PerCP-Cy5.5FL3-H: CD4 PerCP-Cy5.510 1 10 0 10 0 10 4 10 4 10 3 10 3 54.20.5529.515.710 2 10 2 10 1 10 1 10 0 10 0    F   L   2  -   H  :   C   D   8   P   E FL3-H: CD4 PerCP-Cy5.510 4 10 4 10 3 10 3 56.11.6114.527.810 2 10 2 10 1 10 1 10 0 10 0 Figure 5  MR1 treatment lowers the frequency of CD68 +  cells in sciatic nerve and CD8 +  T cells in sciatic lymph node. ( a ) CD68 +  macrophage (green) in S100b +  sciatic nerve (red), 100-day-old controls. ( b ) CD68 +  macrophage (green) in S100b +  sciatic nerve (red), 100-day-old anti-CD40L–treated mice. ( c ) Quantification of reduction of CD68 +  macrophage by anti-CD40L treatment, day 100. White bar, control; gray bar, anti-CD40L–treated; black bar, untreated age-matched non-transgenic mice. WT, wild type. ( d – f ) FACS analysis of CD4 and CD8 expression in CD3 +  lymphocytes isolated from sciatic lymph node in age-matched (day 80) non-transgenic SOD1 G93A  ( d ), untreated control ( e ), and SOD1 G93A  anti-CD40L–treated mice ( f ). Full flow cytometry details and machine settings are described in the Online Methods.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!