WO2013070984A1 - Systems and integrated methods for gene expression analysis in multiple sclerosis - Google Patents

Systems and integrated methods for gene expression analysis in multiple sclerosis Download PDF

Info

Publication number
WO2013070984A1
WO2013070984A1 PCT/US2012/064247 US2012064247W WO2013070984A1 WO 2013070984 A1 WO2013070984 A1 WO 2013070984A1 US 2012064247 W US2012064247 W US 2012064247W WO 2013070984 A1 WO2013070984 A1 WO 2013070984A1
Authority
WO
WIPO (PCT)
Prior art keywords
gene
genes
seq
recited
data
Prior art date
Application number
PCT/US2012/064247
Other languages
French (fr)
Inventor
Edward Croze
Ken D. YAMAGUCHI
Volker KNAPPERTZ
Anthony T. Reder
Hugh SALAMON
Original Assignee
Bayer Healthcare, Llc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Bayer Healthcare, Llc filed Critical Bayer Healthcare, Llc
Publication of WO2013070984A1 publication Critical patent/WO2013070984A1/en

Links

Classifications

    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
    • C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
    • C12Q1/6876—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes
    • C12Q1/6883—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes for diseases caused by alterations of genetic material
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/158—Expression markers

Definitions

  • Bioinformatics is the study of information in biological systems and since the late 1980s, has been particularly critical in understanding biological processes in the context of gene expression. What sets bioinformatics apart is the focus on developing and applying computationally intensive techniques (e.g., pattern recognition, data mining, machine learning algorithms, and visualization).
  • Major research efforts in the field include sequence alignment, gene ontology, genome assembly, drug design, drug discovery, protein structure alignment, protein structure prediction, prediction of gene expression and protein-protein interactions, genome- wide association studies and the modeling of evolution.
  • MS Multiple sclerosis
  • CNS central nervous system
  • NMSS central nervous system
  • MS patients are affected during the most productive time of their lives (Lundy, J 2006).
  • MS is a progressive neurodegenerative disease resulting in increasing disability and a significant decrease in quality of life measures (Hafler et al. 2005; Weiner HL. 2009).
  • Disease progression results in permanent neurological disability.
  • MS patients and their physicians face an increasing need to determine individual therapy response to assertively manage disease progression.
  • Biomarker- optimized treatment approaches can provide such an approach (Wagner, J 2010;
  • Betaseron ® IFNb-lb was the first approved treatment for MS (The
  • IFN beta Multiple Sclerosis Study Group. 1993 was followed by other IFN-b therapies.
  • IFN-b remains an important MS drug despite the recent approval of other MS therapeutics (R Gold, 2010).
  • MS drugs currently available provide similar clinical benefit but vary considerably in severity of side effects, dosing, route of
  • IFN-b is a highly pleiotropic member of a larger family of cytokines that together play an important role in regulating both innate and adaptive immune responses (Stark G, 1998; Mogensen et al, 1999). IFN-b itself is known to have potent antiviral and immunoregulatory properties, but its exact mechanism of action in treating MS remains uncertain. Analysis of gene expression profiles for the purpose of discovering molecular mechanisms of drug action is not simply a matter of traditional statistical analysis of gene expression experimental data.
  • systems and methods are needed to create larger amounts of differential gene expression data that in predicting disease progression, and in transforming patient samples used to measure treatment effects related to positive clinical outcomes. Further, it is also desirable to provide systems and methods that can quickly and efficiently characterize and classify select genes and gene families that play a direct or indirect role in regulating biological processes suggested to be important in eliciting treatment response. Selection and measurement of specific biomarkers that provide information on disease pathophysiology during or before treatment of various diseases is also a largely unrealized promise of personalized medicine.
  • transformed populations of amplified and isolated gene segments create specific signatures that yield transformation of samples into species for quantitation of health and clinical progress and outcomes.
  • the present embodiments provide systems, methods, and compositions for improved gene expression analyses, including biomarker discovery and the creation of novel gene signatures that identify critical gene expression events in the identification, pathology or treatment of disease.
  • the systems comprise a gene database (differential gene expression) for providing a gene predictor set; a biology grouping system downstream from the gene data system for providing statistical grouping of genes for association with common biologies; and a disease grouping system downstream from the biology grouping system for analytically grouping genes for association with a particular disease and determining gene signatures.
  • the invention also enables methods for biomarker discovery based on the integrated analysis of gene expression events.
  • the methods comprise providing a gene predictor set; selecting a portion of genes from the gene predictor set and statistically grouping the genes according to effect on biological function to define a biological function gene set; selecting genes from the biological function gene set and analytically grouping them according to a particular disease to define gene signatures.
  • FIG. 1 is a general block diagram of a system for biomarker discovery.
  • FIG. 2 is the number of differentially expressed genes of treated and non-treated MS patients at various times.
  • FIG. 3 is gene sets characterized by transcription factor binding; rows show excessive differential expression in sample comparisons (columns) as indicated by FDR CERNO values. The most significant differentially expressed gene sets are distinct between short-term and long-term comparisons.
  • FIG. 4 is the 100 most significantly changed functional pathways, as detected with CERNO. Twenty-eight of the 264 long-term IFNb-lb regulated genes map to these 100 pathways. The membership matrix was ordered using hierarchical clustering, revealing a pattern that includes a cluster of functions and genes to beta- oxidation of fatty acids.
  • FIG. 5 is specific protein neighborhoods, such as those surrounding IKBKE and GPRC5C, that exhibit excessive change in a long-term IFNb-lb treatment comparison.
  • FIG. 6 is gene expression ratios showing that long-term IFNb-lb effects counter-regulate disease-related changes.
  • FIG. 7 is a general block diagram of a method for biomarker discovery.
  • biomarker discovery refers to all the processes, analysis, and steps needed in determining a gene signature.
  • biological theme refers to particular groups or sets of genes related to certain known or unknown functions or transcription factor-related or biological prosesses. In certain embodiments, these genes may be known to influence regulation of certain biological or pathophysiological functions. For each disease, various pathophysiological functions will be relevant and induced by particular genes. For instance, certain sets of genes are important and relevant in MS. In MS various genes are used in regulating the blood brain barrier (BBB), TH1, TH2, TH17
  • gene predictor set refers to all the possible known or identified genes under consideration that are derived from the gene expression and database and selected for inclusion in a predictor set by a differential analysis, such as up/down regulation, presence, deletion, duplication, copy number or any other comparison between two or more biological states having an observable basis in disease.
  • gene signature refers to a set or subset of genes that have been grouped statistically and analytically to reflect a specific biological theme.
  • FIG. 1 shows a general block diagram of the system of a biomarker discovery system 100 comprising a system for providing gene data 200 for assembly into a gene database or other platform for data assembly, retrieval or manipulation, a system for grouping genes for association with common biologies 300, a system for grouping genes for association with a particular disease 400, gene signatures 500, and a system for confirming the gene signatures 600.
  • the general block diagram shows the systems and how they interact in biomarker discovery. Other combinations, associations, modes, organizations, and sequences of the system are possible.
  • the system for providing gene data 200 comprises a source of gene data 210 that may include a gene database, a device for analyzing nucleic acids 220, and output data 230.
  • the source of gene data 210 can be derived directly from healthy individuals or diseased MS patients, ex vivo or in vitro studies using immune cell of immune cell subtypes, or other cell-based studies, disease-relevant animal models or may be derived from an established database.
  • the database or source 210 can be manipulated to yield, or may be comprised of differential data including nucleotide sequence data, gene expression data, or any measurement of protein structure or function including post-translational processing, protein-protein interaction, R A or any other regulatory or structural phenomenon but is preferably expressed as differential data, e.g., differences in expression between disease/normal,
  • the gene data 210 can be obtained prior to or after treating a source of gene data, such as a particular patient or group of patients with a particular drug or therapeutic treatment.
  • a source of gene data such as a particular patient or group of patients with a particular drug or therapeutic treatment.
  • IFN-b is a highly pleiotropic member of a larger family of cytokines that together play an important role in regulating both innate and adaptive immune responses (Stark G 1998; Mogensen et al 1999). IFN-b itself is known to have potent antiviral and immunoregulatory properties, but its exact mechanism of action in treating MS remains uncertain.
  • IFN-b The abilities of IFN-b to suppress activation of T and B cells, to inhibit translocation of immune cells from the periphery into the central nervous system, and to inhibit astrocyte and microglia cell differentiation likely play a role in MS treatment (Dhib-Jalbut and Marks, 2010; Agrawal and Yong, 2007). Given the potential of IFN-b to exert neuroprotective influences (Gok et al, 2007; Marsh et al, 2009; Yoshimura et al, 2010), evaluating how MS treatment could be enhanced by IFN-b pleiotropy could prove important in evaluating heterogeneity of patient response to treatment. Thus, a more thoroughly characterized mechanism for IFN-b treatment of MS would help guide MS patient management, especially with respect to other therapies known to work through different mechanisms of action.
  • the source of gene data 210 can be derived from collecting blood samples from MS patients and isolating a nucleic acid such as RNA or DNA from immune cells, or from gathering gene data from various population studies. Other sources known in the art or developed may be used to provide the source of gene data 210.
  • a gene database is developed from an identified clinical population. In this example, fifty-two patients with clinically defined MS, according to McDonald criteria (McDonald et al. 2001) were used. (Reder study, Table 1). For patients treated with high-frequency Betaseron® (Interferonb-lb), disease states were identified, specifically, the mean Expanded Disability Status Scale (EDSS) (Kurzke, 1983) score was between 2.5 - 5.5 , and all patients exhibited 2 or more episodes of clinically evident relapses and remissions plus characteristic MRI lesions. To obviate effects of disease activity on interferon responses, treated patients were free of exacerbations for at least 3 months before and 3 months after blood collection.
  • EDSS Expanded Disability Status Scale
  • IFNb-lb patients had received IFNb-lb for an average of 6.9 years (0.5 -10.5) and had discontinued therapy for at least 64 firs prior to beginning this study (Reder et al 2008).
  • the device for analyzing nucleic acids 220 can comprise any number of devices known or used in the art for analyzing nucleic acid data, gene expression data, or other genomics or proteomics platform as designated herein.
  • the device for analyzing nucleic acids 220 can be capable of determining expression levels of various genes.
  • the device for analyzing nucleic acids can comprise a gene chip or microarray deep sequencing, single nucleotide substitution (SNPs) analysis.
  • SNPs single nucleotide substitution
  • the device for analyzing nucleic acids 220 may be capable of determining expression levels of genes or nucleic acids in a control or when exposed to or contacted by a therapeutic drug or treatment.
  • the data output of any of these devices, individually or collectively may be a source of gene data 210 and the analytical output thereof may provide the contents of a gene database for use as designated herein.
  • EXAMPLE -2 [000032] Gene expression measurements of the disease states patient cohorts and healthy controls were grouped in this study using Affymetrix Human Exon 1.0 ST (HuEx 1.0 ST) arrays, which provide information on more than 300,000 gene transcripts. To discover biological information relevant to the mechanism of action of IFNb-lb in MS, multi-gene array analysis was employed, building on a published approach previously applied to understand IFNb-lb effects in MS patients
  • RNA samples having a 28S/18S ratio below 1.5 were excluded from the present study.
  • RNA was hybridized to GeneChip® HuEx 1.0 ST arrays (Affymetrix Inc, CA) containing approximately 1.4 million probe sets recognizing over 1 million exon clusters.
  • the resulting expression analysis comprises a gene database for use in developing a gene predictor set.
  • the Affymetrix Data Resource Center http://www.affymetrix.com/) provided a dataset of 11 tissues hybridized in triplicate on both the HuEx 1.0 ST and HG-U133 Plus 2.0 platforms.
  • Probe set (3' arrays) and transcript (HuEx 1.0 ST arrays) expression level summarization was performed using the RMA (Irizarry et al. 2003)
  • Affymetrix Power Tools Affymetrix Power Tools (APT) version 1.12.0.
  • Affymetrix's NetAffx Release 30 June 2009 was used for probe set and transcript annotations.
  • the data derived from the source for gene data 210 is then subjected to the device for analysis of nucleic acids 220 which results in output gene data 230 (See FIG. 1) maintained in a gene database.
  • the output gene data 230 can comprise a gene predictor set.
  • the predictor set of genes comprises many if not all of the possible nucleic acids or genes of interest under the defined conditions as selected from the gene database.
  • the gene predictor set can comprise many of the genes that are up or down regulated or influenced when a patient is treated with a therapeutic drug.
  • the gene prediction set can comprise those genes induced by interferon.
  • a subset of these genes may be involved in the regulation or control of various biological functions or patho-physio logical functions that result from a disease, such as multiple sclerosis.
  • the purpose of the predictor set of genes is to produce a comprehensive, but selected subset of genes that yield select biomarkers, including discrete gene signatures that impact or produce certain recognized disease states or patho-physio logical functions.
  • Some of the genes included in a predictor set and initially identified as relevant to a disease state may have only limited causal impact on creating or changing the disease state in the patient. For this reason, it is desirable to further refine the gene predictor set to select the more important biomarkers or gene signatures that are targeted for maximal impact by a therapeutic drug or treatment. This can be done by further organizing and classifying the predictor set of genes using a system for grouping genes for association with common biologies.
  • the biology grouping system is downstream of the gene database and provides a functional filter for further analyzing the gene predictor set of interferon inducible genes in mammalian cells.
  • the biology grouping system 300 further refines the gene predictor set using biology-based selection criteria, including for example, system 300 comprises CERNO software 310, at least one transcription factor database 320, or at least one ontology data base 330.
  • CERNO is defined as Coincident Extreme Ranks in Numerical Observations and uses a selection criteria on the extreme changes within a defined gene predictor set and analyzes all expression measurements rather than an arbitrarily thresholded set of differentially expressed genes.
  • CERNO encompasses a non-parametric approach that accounts for the uncertainty in expression ratio estimates and also provides a measure for strength of effect in a gene predictor set.
  • Other combinations, software, or databases, yielding selection criteria can be used with the present embodiments.
  • the system for grouping genes 300 groups genes at significant threshold expression levels and correlates these genes with various biological pathways or functions by conducting statistical analysis of the output gene data 230 provided from the system for providing gene data 200.
  • a "p" value can be assigned by assimilating all the data together. This represents the confidence level of expression of certain target genes and their expression levels.
  • each of the data provided by the CERNO software 310, the transcription factor database 320, or the ontology database 330 are compared and analyzed.
  • the gene name and the Entrez Gene ID number was used to search the NIH NCBI database (http://www.ncbi.nlm.nih.gov), and the accession number with nucleotide sequence information identified. In those cases where the database contained several variant mRNA sequences for a gene, the longest mRNA sequence was used for the mRNA sequence of the gene. In the case of the ND6 mitochondrial gene, that portion of the mitochondrial genome is identified (MT BP 12148 - 14672).
  • Data handling was performed using Perl DBI (database) and PostgreSQL, and statistical analysis was performed in R (R Development Core Team 2005).
  • Perl-DBI was used to interact with available databases.
  • Perl DBI provided complete file input and output capabilities and syntax for line-at-a-time sequential input and text manipulation/parsing.
  • PostgreSQL was used as an enterprise-level database providing data-mining and data handling properties for large texts and dynamic content.
  • Statistical analysis in R was performed to provide a programming language and environment for statistical computing and graphic compatible with analyzed databases.
  • CERNO Coincident Extreme Ranks in Numerical Observations
  • probe set T-test p-values are randomly ordered, the following condition holds for probe sets, i, ..., ⁇ mapping to a gene G, the area under the upper tail of the In a ) is a conservative p-value, p(G), for rejecting the null hypothesis, where is the rank of the T-test p-value for the i th probe set, and n a is the total number of probe sets measuring transcripts from HUGO named genes.
  • the alternative hypothesis is that the p-values for probe sets measuring transcripts from the gene G are more extreme (lower) than under the null.
  • Affymetrix 3' array probe sets were identified by review of CLIMB study data, Affymetrix tissue expression data available on both whole transcript and 3' arrays, and additional Affymetrix data as available. Selection of 3 ' array probe sets, with the intent to capture the effects measured on the HuEx 1.0 ST array, were completed prior to running tests on the Goertsches study published data. Supplementary information was provided with more in-depth descriptions of signature definition criteria, as well as the lists of the selected genes and probe sets.
  • differentially expressed genes were observed in the long-term comparison than in any short-term comparison, whether differential expression was measured by a nominal p-value cutoff or by FDR analysis (See FIG. 2).
  • differentially expressed genes at the ⁇ 0.05 cutoff are roughly three-fold more abundant in the 2-year comparison than in the short-term comparisons.
  • No probe sets exhibit a false discovery rate (FDR) of 0.05 or less in the short-term comparisons.
  • 264 genes have at least one probe set that exhibits a t-test FDR of 0.05 in the 2-year comparison (black bar). In all, 715 probe sets map to these 264 genes.
  • the overlap of the 264 long-term differentially expressed genes forming a gene signature 500 and the top- 100 differentially expressed pathways was explored.
  • a discrete twenty-eight gene signature 500 was identified in the overlap, and gene signature was used in a cluster analysis to group the pathway effects (See FIG. 4).
  • the top- 100 differentially expressed pathways as detected with CERNO point to three major effects of IFNb-lb long-term treatment.
  • Twenty-eight of the 264 long-term IFN -lb regulated genes map to these 100 pathways.
  • the membership matrix was ordered using hierarchical clustering, revealing a pattern that includes a cluster of functions and genes related to beta-oxidation of fatty acids.
  • the plotting technique identifier p- values sets as a function of set size. Three of the most differentially expressed neighborhoods of different sizes confirm the ability to isolate the gene signatures 500.
  • FDR 0.002
  • All three neighbors of GPRC5C exhibited encoding genes regulated by IFNb-lb, and individually exhibited transcripts significantly differentially expressed at /KO.001.
  • SCOT is the protein encoded by OXCT1 and is a homodimeric mitochondrial matrix enzyme that plays a central role in extrahepatic ketone body catabolism by catalyzing the reversible transfer of coenzyme A from succinyl-CoA to acetoacetate.
  • a gene signature 500 comprises the 264 long-term-regulated genes were examined by plotting the expression ratios comparing long-term treated patients to therapy-na ' ive patients against the ratios comparing therapy-na ' ive patients to healthy controls.
  • the long-term regulated gene expression for these differentially expressed probe sets tends to reverse the expression changes that characterize disease, especially when focusing on the probe sets exhibiting more significant changes in disease (p ⁇ 0.05, dark points show negative slope).
  • analytic filter 420 may comprise review and comparisons to literature data and information, can be based on observable and quantifiable patient data, observation or anecdotal information, or can be subject to further software review and/or processing.
  • Other data sources for refinement includes, but are not be limited to CERNO, clinical data, scientist review, key opinion leader observation, conventional knowledge, disease- associated animal models, genome-wide transcriptomics, disease susceptibility markers, medical records, unpublished information from scientific meetings, Advisory Boards, consultants.
  • the results of the above processing provide both statistical and analytically grouped gene signatures 500.
  • the gene signatures 500 comprise groups of genes associated with a common underlying biology. EXAMPLE -12
  • LT no short term regulation in the same direction as long-term regulation
  • LT2 more stringent, required that none of the transcripts be regulated in the short term
  • the gene signatures 500 may be subject to further optional systems and processing.
  • a system for confirming gene signatures 600 can be used to independently confirm the accuracy of the statistical and analytical grouped gene signatures 500.
  • the system for confirming gene signatures 600 can comprise literature review and confirmations, comparison to CLIMB data, validation studies, and use, application and comparison to other databases known or unknown in the art including clinical studies, disease-associated animal models, pharmacodynamic or pharmacokinetic studies, toxicity studies, pharmacoviligence studies, all available genome-wide disease susceptibility and transcriptomic data.
  • IFNb-lb signatures confirmed in independent data, and distinct from GA gene expression effects
  • IFN-b bioactivity markers have been proposed previously (Comabella et al. 2009; Van Baarsen 2008; Vosslamber et al. 2009), but few have been validated in independent studies or demonstrated to predict treatment outcomes relevant to clinical endpoints such as disability, exacerbation rate, cognition loss, brain atrophy, and EDSS.
  • IFN-b bioactivity markers e.g., transient, neutralizing antibody effects, may well predict MRI outcomes.
  • improved treatment response markers that are associated with MS-relevant regulation of biological function are needed (Arnason et al 1996;
  • therapeutically relevant molecular systems yields informed decisions regarding course of disease treatment methods and drug selection and informs a decision to adapt with another MS drug that targets a different or complementing mechanism of action.
  • HMG-CoA 3-hydroxy-3-methylglutaryl-coenzyme A
  • IFNb- lb-treated MS patients exhibited significantly higher expression of the mitochondrial encoded NADH dehydrogenase 6 gene.
  • Expression of wild-type mt-ND6 appears to help maintain normal levels of ROS, thereby reducing the likelihood of oxidative damage to cells, and mutated mt-ND6 appears causative in Leber hereditary optic neuropathy, a disease which co-occurs with MS and MS-like white matter abnormalities. Therefore, IFNb-lb may play an important role in regulating mitochondrial mediated oxidative- stress responses that likely play a key role in MS.
  • ⁇ ⁇ (encoded by IKBKE) is a link between the protein network analysis, in which ⁇ ⁇ was the hub of the most significantly regulated network neighborhood, and the canonical pathway analysis, in which mitochondrial energy metabolism was highlighted. ⁇ ⁇ regulates energy balance (Chiang et al. 2009). Moreover, the role of ⁇ ⁇ in long-term IFNb treatment is consistent with the dependence between ⁇ ⁇ and IFNb functions observed in an in vivo antiinflammatory setting (Corr et al). Recombinant IFNb induces rapid phosphorylation of ⁇ ⁇ in vitro (Tenoever et al. 2007). ⁇ is critical to IFNb's roll in response to viral infection (Tenoever et al. 2007) and likely integrates important signals from pattern recognition receptors that detect virus nucleic acids (Hiscott 2004).
  • DDX58 (aka RIG-1) is a potentially important viral detection mechanism that signals through IRF-dependent pathways, is up-regulated in three IFNb treatment studies. Proteins upstream of interferon genes are regulated by treatment, and thus the feedback mechanism of interferon response is itself modified. ⁇ ⁇ is a key part of this feedback loop and thus should play a role in multiple sclerosis treatment response.
  • the large multi-gene expression patterns disclosed herein reflect biological effects observable across platforms and patient cohorts, and to confirm smaller multi-gene signatures in independent study data.
  • the protein interaction network analysis of gene expression microarray data strongly underscored a role for ⁇ in IFNb-lb treatment effects, especially observable over long-term treatment.
  • Data analysis showed long-term and short-term effects of IFNb-lb treatment were not only distinct but actually contrasted in direction of change.
  • the long-term IFNb-lb treatment effects highlighted energy metabolism regulation and the potential for neuroprotective effects mediated by HMG-CoA-derived bio molecules and antioxidant proteins.
  • each step of a method comprises providing an expansive predictor set of genes 700, statistically grouping the predictor set of genes 700 with biological functions 800, analytically grouping genes for association with a particular disease 850, selecting genes signatures 900, and optionally confirming the gene signatures 950.
  • the expansive predictor set of genes 700 is generally provided by the system for providing the nucleic acid/gene data 200.
  • the source of gene data 210 is subjected to a device for analyzing nucleic acids 220.
  • the device for analyzing nucleic acids 220 can comprise any number of devices known in the art.
  • the device for analyzing nucleic acids 220 can comprise a micro array device.
  • the gene data is contacted to the micro array device to hybridize the nucleic acids to the micro array.
  • Various methodologies can then be performed that are well known in the art for reading and analyzing the gene data. For instance, various markers or dyes can be used for determining if the target nucleic acids have hybridized to the micro array.
  • Identified nucleic acids or genes can then be read using various readers known in the art. As a result, the gene data 230 can then be used for further processing and analysis. Other devices known in the art can be employed. The devices should be capable of identifying various nucleic acids targets of interest and quantifying their level of gene expression. Once the gene data 230 has been determined, it can then be further processed.
  • the gene data is organized into a gene predictor set and subjected to further processing.
  • Gene data can be described but not limited to the robust microchip average (RMA) algorithm with default parameters.
  • RMA microchip average
  • CERNO provides an example in which p-values derive from logarithmic expression values for a given gene chip relating probe sets P min and Pmax to a lower mean log ranking indicating a stronger strength of signal or p-value.
  • the gene predictor set is further grouped statistically using various biological functions 800. The statistical grouping according to biological functions is accomplished by subjecting the gene predictor set data to a combination of processes that can be performed serially or in parallel.
  • the gene predictor set can be processed using various software that can compare identified nucleic acid data and sequences with already characterized and identified nucleic acids and sequences that have been stored in known public or private gene databases. These gene databases and processing are particularly helpful in identifying those associated nucleic acids and genes that can be upstream from the gene predictor set.
  • nucleic acid or gene data derived from transcription factor binding by various proteins associates nucleic acid or gene data derived from transcription factor binding by various proteins to identify nucleic acids and genes that have been up or down regulated. For instance, transcription factors will have bound to those areas around the nucleic acid or target of interest. This binding provides additional details and data regarding up or down regulated genes. Further, this data provides an output pattern reaching absolute and differential gene expression. These results yield a gene predictor set.
  • Another process associates nucleic acid or gene data downstream from a gene predictor set analyzing nucleic acid and gene data linked to physiological, physical, or disease states is then compared to the gene predictor set to identify matches in sequence and/or expression data. All processes may occur simultaneously or in tandem when analyzing the predictor set.
  • data is classified according to biological functions by system 300 using statistical methods.
  • the statistical methods analyze the data and provide an associated "p" value that can be statistically analyzed. For instance, various thresholds eliminate irrelevant expression data such that meeting defined thresholds in upstream, downstream, and transcription factor binding are matched to one or more particular biological functions.
  • the results can then be subjected to further processing and analysis.
  • the gene data is then subjected to analysis for association with various diseases 850.
  • This processing takes one or more of the associated biological functions that were determined in step 800 and then subjects them individually or in combinations to determine relative significance to particular disease states.
  • the system for grouping genes for association with a particular disease 400 combines one or more of the associated biological functions 410 and subjects this data to analytical filter 420.
  • Analytic filter 420 works by using CER O 310, clinical data, known literature results, databases including research based MS patient samples (Reder, CLIMB, Goertsches databases, all published transcriptomics studies, public MS disease associated databases and disease- associated animal studies, ex vivo and in vitro studied immune cell, immune cell subtype ,other cell-based databases , or other information in combination or separately to match diseased states with associated biological function and gene data.
  • the gene data, associated biological functions and diseased states are then grouped to determine gene signatures 500. This step is accomplished by using certain threshold levels to distinguish which important correlated data is most important in diseased and non-diseased patients and which associated data meet certain "p" values of significance.
  • the final defined gene signature can then define those genes that when treated or not treated by a therapeutic treatment or drug, will have the change relative to the relative biological function. This provides for opportunities to further refine the associations for more personalized medicine.
  • the gene signature 500 can then be optionally subjected to a system for confirming gene signatures 600. This validation step can use known databases, CLIMB data, or other validation studies to further improve accuracy.
  • GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED HG- GENE ID ACCESSION HuEx ST U133 PLUS 2.0
  • GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED HG- GENE ID ACCESSION HuEx ST U133 PLUS 2.0
  • GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED GENE ID ACCESSION HuEx ST HG-U133 PLUS
  • ARPC5 10092 NM 001270439.1 SEQ ID NO 75 2017725 1569325 at 2
  • HADHA 3030 NM 000182.4 SEQ ID NO 86 2545092 208630 at
  • PROS1 5627 NM 000313.3 SEQ ID NO 107 2685268 207808 s at
  • TRAF3IP3 80342 NM 025228.2 SEQ ID NO 131 2453869 240265 at
  • HADHA 3030 NM 000182.4 SEQ ID NO 86 2545092 208630 at
  • PROS1 5627 NM 000313.3 SEQ ID NO 107 2685268 207808 s at
  • TRAF3IP3 80342 NM 025228.2 SEQ ID NO 141 2453869 240265 at
  • HADHA 3030 NM 000182.4 SEQ ID NO: 86 2545092 208630 at
  • HADHB 3032 NM 000183.2 SEQ ID NO: 146 2473735 201007 at
  • NDUFB10 4716 NM 004548.2 SEQ ID NO: 147 3644220 223112 s at
  • ARPC5 10092 NM 001270439.1 SEQ ID NO 75 2917725 211963 s at
  • PAFAH1B1 5048 NM 000430.3 SEQ ID NO 155 2500368 200816 s at
  • PAFAH1B2 5049 NM 002572.3 SEQ ID NO 156 3972241 210160 at

Landscapes

  • Chemical & Material Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Proteomics, Peptides & Aminoacids (AREA)
  • Health & Medical Sciences (AREA)
  • Organic Chemistry (AREA)
  • Wood Science & Technology (AREA)
  • Analytical Chemistry (AREA)
  • Zoology (AREA)
  • Genetics & Genomics (AREA)
  • Engineering & Computer Science (AREA)
  • Pathology (AREA)
  • Immunology (AREA)
  • Microbiology (AREA)
  • Molecular Biology (AREA)
  • Biotechnology (AREA)
  • Biophysics (AREA)
  • Physics & Mathematics (AREA)
  • Biochemistry (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)

Abstract

The present embodiments provide systems and methods for determining gene signatures relating to MS and specifically interferon-based treatments including gene/differential gene expression database for providing a gene predictor set; a biology grouping system downstream from the gene data system for providing statistical grouping of genes for association with common biologies; and a disease grouping system downstream from the biology grouping system for analytically grouping genes for association with a particular disease and determining gene signatures. Methods for drug and biomarker discovery include providing a gene predictor set; selecting a portion of genes from the gene predictor set and statistically grouping the genes according to effect on biological function to define a biological function gene set; selecting genes from the biological function gene set and analytically grouping them to define gene signatures.

Description

SYSTEMS AND INTEGRATED METHODS FOR GENE EXPRESSION ANALYSIS IN MULTIPLE SCLEROSIS
BACKGROUND
[00001] A massive number of tools and techniques are used to study gene expression. However, these tools and techniques generate voluminous data that is not easily understandable or relevant to understanding how gene expression relates to biological processes, pathophysiology, or higher order physiological functions. Much of this analysis is, therefore, left to bioinformatics.
[00002] Bioinformatics is the study of information in biological systems and since the late 1980s, has been particularly critical in understanding biological processes in the context of gene expression. What sets bioinformatics apart is the focus on developing and applying computationally intensive techniques (e.g., pattern recognition, data mining, machine learning algorithms, and visualization). Major research efforts in the field include sequence alignment, gene ontology, genome assembly, drug design, drug discovery, protein structure alignment, protein structure prediction, prediction of gene expression and protein-protein interactions, genome- wide association studies and the modeling of evolution.
[00003] Many disease states are characterized by differences in the expression levels of various genes either through changes in the copy number of the genetic DNA or through changes in levels of transcription of particular genes (e.g., through control of initiation, provision of RNA precursors, RNA processing, etc.). To date, however, limited systems or methods exist for processing molecular biology data to correlate genes with biological functions or processes. Further, there are limited systems and methods available for understanding physical or pathophysiological functions influenced by disease.
[00004] Multiple sclerosis (MS) is a chronic neurological and inflammatory disease of the central nervous system (CNS) that affects more than 2.1 million people worldwide (NMSS). MS patients are affected during the most productive time of their lives (Lundy, J 2006). MS is a progressive neurodegenerative disease resulting in increasing disability and a significant decrease in quality of life measures (Hafler et al. 2005; Weiner HL. 2009). Disease progression results in permanent neurological disability. MS patients and their physicians face an increasing need to determine individual therapy response to assertively manage disease progression. Biomarker- optimized treatment approaches can provide such an approach (Wagner, J 2010;
Croze 2010; Micheel and Ball 2010; Vandenbroeck K 2010).
[00005] Betaseron ® (IFNb-lb) was the first approved treatment for MS (The
IFN beta Multiple Sclerosis Study Group. 1993) and was followed by other IFN-b therapies. IFN-b remains an important MS drug despite the recent approval of other MS therapeutics (R Gold, 2010). MS drugs currently available provide similar clinical benefit but vary considerably in severity of side effects, dosing, route of
administration, and mechanism of action (Markowitz, 2010).
[00006] IFN-b is a highly pleiotropic member of a larger family of cytokines that together play an important role in regulating both innate and adaptive immune responses (Stark G, 1998; Mogensen et al, 1999). IFN-b itself is known to have potent antiviral and immunoregulatory properties, but its exact mechanism of action in treating MS remains uncertain. Analysis of gene expression profiles for the purpose of discovering molecular mechanisms of drug action is not simply a matter of traditional statistical analysis of gene expression experimental data.
[00007] Understanding the casual relationships that are hidden in voluminous gene expression data and the mechanisms underlying the expressed genes that are actually responsible for the existence or progression of disease is a priority for scientists. However, the most significant gene expression events can be lost or hidden in the mass of data produced in traditional genomics or proteomics analyses. For example, profound physiological changes accompany the symptoms of the disease that are unrelated to the cause of the disease and these changes may actually be part of a biochemical cascade that merely accompanies the disease, rather than being part of the fundamental cause of the disease. Furthermore, drug-based treatments also cause changes in gene expression and the treatment itself may introduce differential gene expression events that influence detection of biomarkers, including RNA and protein based serum biomarkers. For these reasons, systems and methods are needed to create larger amounts of differential gene expression data that in predicting disease progression, and in transforming patient samples used to measure treatment effects related to positive clinical outcomes. Further, it is also desirable to provide systems and methods that can quickly and efficiently characterize and classify select genes and gene families that play a direct or indirect role in regulating biological processes suggested to be important in eliciting treatment response. Selection and measurement of specific biomarkers that provide information on disease pathophysiology during or before treatment of various diseases is also a largely unrealized promise of personalized medicine.
[00008] Finally, transformed populations of amplified and isolated gene segments create specific signatures that yield transformation of samples into species for quantitation of health and clinical progress and outcomes.
SUMMARY OF INVENTION
[00009] The present embodiments provide systems, methods, and compositions for improved gene expression analyses, including biomarker discovery and the creation of novel gene signatures that identify critical gene expression events in the identification, pathology or treatment of disease. The systems comprise a gene database (differential gene expression) for providing a gene predictor set; a biology grouping system downstream from the gene data system for providing statistical grouping of genes for association with common biologies; and a disease grouping system downstream from the biology grouping system for analytically grouping genes for association with a particular disease and determining gene signatures.
[000010] The invention also enables methods for biomarker discovery based on the integrated analysis of gene expression events. The methods comprise providing a gene predictor set; selecting a portion of genes from the gene predictor set and statistically grouping the genes according to effect on biological function to define a biological function gene set; selecting genes from the biological function gene set and analytically grouping them according to a particular disease to define gene signatures.
[000011] The systems and methods are described in the context of the analysis of (beta-interferon IFNb) bioactivity and response to treatment in MS patients. The principles are applicable where gene expression databases exist and information can be extrapolated for downstream analysis as described below. DESCRIPTION OF THE FIGURES
[000012] FIG. 1 is a general block diagram of a system for biomarker discovery.
[000013] FIG. 2 is the number of differentially expressed genes of treated and non-treated MS patients at various times.
[000014] FIG. 3 is gene sets characterized by transcription factor binding; rows show excessive differential expression in sample comparisons (columns) as indicated by FDR CERNO values. The most significant differentially expressed gene sets are distinct between short-term and long-term comparisons.
[000015] FIG. 4 is the 100 most significantly changed functional pathways, as detected with CERNO. Twenty-eight of the 264 long-term IFNb-lb regulated genes map to these 100 pathways. The membership matrix was ordered using hierarchical clustering, revealing a pattern that includes a cluster of functions and genes to beta- oxidation of fatty acids.
[000016] FIG. 5 is specific protein neighborhoods, such as those surrounding IKBKE and GPRC5C, that exhibit excessive change in a long-term IFNb-lb treatment comparison.
[000017] FIG. 6 is gene expression ratios showing that long-term IFNb-lb effects counter-regulate disease-related changes.
[000018] FIG. 7 is a general block diagram of a method for biomarker discovery.
DETAILED DESCRIPTION
[000019] The present invention is not limited to the exemplary methodology, protocols, reagents, or other tool used to enable or illustrate the invention as described. The terminology used herein is for the purpose of describing particular embodiments only, and is not intended to limit the scope of the present invention which will be defined by the appended claims.
[000020] As used herein and in the appended claims, the singular forms "a," "and," and "the" include plural reference unless the context clearly dictates otherwise. Thus, for example, reference to "a gene" is a reference to one or more genes and includes equivalents thereof known to those skilled in the art, and so forth.
[000021] All publications and patents mentioned herein are hereby incorporated herein by reference for the purpose of describing and disclosing, for example, the constructs and methodologies that are described in the publications which might be used in connection with the presently described invention. The publications discussed above and throughout the text are provided solely for their disclosure prior to the filing date of the present application. Nothing herein is to be construed as an admission that the inventors are not entitled to antedate such disclosure by virtue of prior invention.
[000022] The term "biomarker discovery" refers to all the processes, analysis, and steps needed in determining a gene signature.
[000023] The term "biological theme" refers to particular groups or sets of genes related to certain known or unknown functions or transcription factor-related or biological prosesses. In certain embodiments, these genes may be known to influence regulation of certain biological or pathophysiological functions. For each disease, various pathophysiological functions will be relevant and induced by particular genes. For instance, certain sets of genes are important and relevant in MS. In MS various genes are used in regulating the blood brain barrier (BBB), TH1, TH2, TH17
(inflammatory response), oxidative stress, neurotropic repair, adhesion molecules and immune cell migration and antiviral mechanisms and antibody responses. Each of these gene sets that are interrelated by their association with a condition comprises a "biological theme". [000024] The term "gene predictor set" or "predictor set" refers to all the possible known or identified genes under consideration that are derived from the gene expression and database and selected for inclusion in a predictor set by a differential analysis, such as up/down regulation, presence, deletion, duplication, copy number or any other comparison between two or more biological states having an observable basis in disease.
[000025] The term "gene signature" refers to a set or subset of genes that have been grouped statistically and analytically to reflect a specific biological theme.
[000026] FIG. 1 shows a general block diagram of the system of a biomarker discovery system 100 comprising a system for providing gene data 200 for assembly into a gene database or other platform for data assembly, retrieval or manipulation, a system for grouping genes for association with common biologies 300, a system for grouping genes for association with a particular disease 400, gene signatures 500, and a system for confirming the gene signatures 600. The general block diagram shows the systems and how they interact in biomarker discovery. Other combinations, associations, modes, organizations, and sequences of the system are possible.
[000027] The system for providing gene data 200 comprises a source of gene data 210 that may include a gene database, a device for analyzing nucleic acids 220, and output data 230.
Source of Gene Data
[000028] The source of gene data 210 can be derived directly from healthy individuals or diseased MS patients, ex vivo or in vitro studies using immune cell of immune cell subtypes, or other cell-based studies, disease-relevant animal models or may be derived from an established database. The database or source 210 can be manipulated to yield, or may be comprised of differential data including nucleotide sequence data, gene expression data, or any measurement of protein structure or function including post-translational processing, protein-protein interaction, R A or any other regulatory or structural phenomenon but is preferably expressed as differential data, e.g., differences in expression between disease/normal,
treated/untreated, placebo/drug therapy, or other differential. The gene data 210 can be obtained prior to or after treating a source of gene data, such as a particular patient or group of patients with a particular drug or therapeutic treatment. For instance, IFN-b is a highly pleiotropic member of a larger family of cytokines that together play an important role in regulating both innate and adaptive immune responses (Stark G 1998; Mogensen et al 1999). IFN-b itself is known to have potent antiviral and immunoregulatory properties, but its exact mechanism of action in treating MS remains uncertain. The abilities of IFN-b to suppress activation of T and B cells, to inhibit translocation of immune cells from the periphery into the central nervous system, and to inhibit astrocyte and microglia cell differentiation likely play a role in MS treatment (Dhib-Jalbut and Marks, 2010; Agrawal and Yong, 2007). Given the potential of IFN-b to exert neuroprotective influences (Gok et al, 2007; Marsh et al, 2009; Yoshimura et al, 2010), evaluating how MS treatment could be enhanced by IFN-b pleiotropy could prove important in evaluating heterogeneity of patient response to treatment. Thus, a more thoroughly characterized mechanism for IFN-b treatment of MS would help guide MS patient management, especially with respect to other therapies known to work through different mechanisms of action.
[000029] In certain other embodiments the source of gene data 210 can be derived from collecting blood samples from MS patients and isolating a nucleic acid such as RNA or DNA from immune cells, or from gathering gene data from various population studies. Other sources known in the art or developed may be used to provide the source of gene data 210.
EXAMPLE - 1
Study population (Source of Gene Data)
[000030] A gene database is developed from an identified clinical population. In this example, fifty-two patients with clinically defined MS, according to McDonald criteria (McDonald et al. 2001) were used. (Reder study, Table 1). For patients treated with high-frequency Betaseron® (Interferonb-lb), disease states were identified, specifically, the mean Expanded Disability Status Scale (EDSS) (Kurzke, 1983) score was between 2.5 - 5.5 , and all patients exhibited 2 or more episodes of clinically evident relapses and remissions plus characteristic MRI lesions. To obviate effects of disease activity on interferon responses, treated patients were free of exacerbations for at least 3 months before and 3 months after blood collection.
Patients had received IFNb-lb for an average of 6.9 years (0.5 -10.5) and had discontinued therapy for at least 64 firs prior to beginning this study (Reder et al 2008). After a baseline sample was drawn, IFN-b (8 million international units (IU), specific activity 2.3 x 107 IU/mg) was self-administered subcutaneously (s.c.) under a physician's supervision. Blood was drawn prior to administration of drug (t =0), 4, 18 and 42 hrs post-administration. Additionally, 45 treatment-na'ive patients were studied. Ten R MS patients had blood draws during an exacerbation period (active RRMS), and 1 1 RRMS patients were free of exacerbations 6 months before and after sample draw and showed no EDSS progression for one year (stable RRMS). Seven treatment-na'ive patients had primary-progressive MS (PPMS), and 8 had secondary- progressive MS (SPMS). Nine healthy individuals were included. No patients had received corticosteroids or immunosuppressants for 1 year prior to or during the study.
Device for analyzing nucleic acids
[000031] The device for analyzing nucleic acids 220 can comprise any number of devices known or used in the art for analyzing nucleic acid data, gene expression data, or other genomics or proteomics platform as designated herein. For instance, the device for analyzing nucleic acids 220 can be capable of determining expression levels of various genes. In one embodiment the device for analyzing nucleic acids can comprise a gene chip or microarray deep sequencing, single nucleotide substitution (SNPs) analysis. Protein-coding gene annotation, RNA-Seq, Micro RNA, northern blots, reverse transcriptase PCR (RT-PCR) full-length cDNA (FLcDNA) sequencing, expression sequence TAG (EST) serial analysis (SAGE) noncoding RNA (miRNA, SiRNA, piRNA) or similar transcriptome technology. The device for analyzing nucleic acids 220 may be capable of determining expression levels of genes or nucleic acids in a control or when exposed to or contacted by a therapeutic drug or treatment. The data output of any of these devices, individually or collectively may be a source of gene data 210 and the analytical output thereof may provide the contents of a gene database for use as designated herein.
EXAMPLE -2 [000032] Gene expression measurements of the disease states patient cohorts and healthy controls were grouped in this study using Affymetrix Human Exon 1.0 ST (HuEx 1.0 ST) arrays, which provide information on more than 300,000 gene transcripts. To discover biological information relevant to the mechanism of action of IFNb-lb in MS, multi-gene array analysis was employed, building on a published approach previously applied to understand IFNb-lb effects in MS patients
(Yamaguchi et al, 2008). Utilizing multi-gene analysis provided stable across microarray platforms comparisons than single gene effects (Subramanian et al. 2005) and allowed multi-gene biomarker signatures to be developed and then confirmed in independently sampled, previously published differential gene expression patient data.
EXAMPLE -3
RNA isolation and Gene Chip Analysis as a Source of Gene Data 210
[000033] Heparinized blood (10-15 ml) was collected by venipuncture just prior to administration of IFN-b after the drug washout period (time = 0). Immediately after blood draw, peripheral blood mononuclear cells (MNC) were isolated using Ficoll-Paque Plus (Amersham Biosciences, NJ). RNA was isolated from MNCs using an RNeasy Midi Kit® (Qiagen, Inc., Santa Clara, CA). Quality and quantity of the resultant RNA were determined using a Bioanalyzer 2100 (Agilent Instruments, Inc., Foster City, CA). RNA samples having a 28S/18S ratio below 1.5 were excluded from the present study. RNA was hybridized to GeneChip® HuEx 1.0 ST arrays (Affymetrix Inc, CA) containing approximately 1.4 million probe sets recognizing over 1 million exon clusters. The resulting expression analysis comprises a gene database for use in developing a gene predictor set.
EXAMPLE -4
Gene expression analysis using microarrays [000034] GEO record GSE 1621 (http://www.ncbi.nlm.nih.gov/geo/query/ acc.cgi?acc=GSE16214) provided gene expression profile data from the
Comprehensive Longitudinal Investigation of MS at the Brigham & Women's Hospital (CLIMB study), which included untreated (n=82), IFNb-treated (n=94), and glatiramer-acetate-treated (n=64) (GA, Copaxone®) MS patients. GEO record GSE24427 (http://www.ncbi.
nlm.nih.gov/geo/query/acc. cgi?acc=GSE24427) provided independent gene expression profile data from IFNb- lb-treated patients as presented in Goertsches et al. 2010. The Affymetrix Data Resource Center (http://www.affymetrix.com/) provided a dataset of 11 tissues hybridized in triplicate on both the HuEx 1.0 ST and HG-U133 Plus 2.0 platforms.
[000035] Probe set (3' arrays) and transcript (HuEx 1.0 ST arrays) expression level summarization was performed using the RMA (Irizarry et al. 2003)
implementation in Affymetrix Power Tools (APT) version 1.12.0. Affymetrix's NetAffx Release 30 (November 2009) was used for probe set and transcript annotations.
Output gene data
[000036] The data derived from the source for gene data 210 is then subjected to the device for analysis of nucleic acids 220 which results in output gene data 230 (See FIG. 1) maintained in a gene database. The output gene data 230 can comprise a gene predictor set. The predictor set of genes comprises many if not all of the possible nucleic acids or genes of interest under the defined conditions as selected from the gene database. For instance, the gene predictor set can comprise many of the genes that are up or down regulated or influenced when a patient is treated with a therapeutic drug. As another example, the gene prediction set can comprise those genes induced by interferon. A subset of these genes may be involved in the regulation or control of various biological functions or patho-physio logical functions that result from a disease, such as multiple sclerosis. The purpose of the predictor set of genes is to produce a comprehensive, but selected subset of genes that yield select biomarkers, including discrete gene signatures that impact or produce certain recognized disease states or patho-physio logical functions. Some of the genes included in a predictor set and initially identified as relevant to a disease state may have only limited causal impact on creating or changing the disease state in the patient. For this reason, it is desirable to further refine the gene predictor set to select the more important biomarkers or gene signatures that are targeted for maximal impact by a therapeutic drug or treatment. This can be done by further organizing and classifying the predictor set of genes using a system for grouping genes for association with common biologies.
[000037] Referring to FIG. 1, the biology grouping system is downstream of the gene database and provides a functional filter for further analyzing the gene predictor set of interferon inducible genes in mammalian cells. The biology grouping system 300 further refines the gene predictor set using biology-based selection criteria, including for example, system 300 comprises CERNO software 310, at least one transcription factor database 320, or at least one ontology data base 330. CERNO is defined as Coincident Extreme Ranks in Numerical Observations and uses a selection criteria on the extreme changes within a defined gene predictor set and analyzes all expression measurements rather than an arbitrarily thresholded set of differentially expressed genes. CERNO encompasses a non-parametric approach that accounts for the uncertainty in expression ratio estimates and also provides a measure for strength of effect in a gene predictor set. Other combinations, software, or databases, yielding selection criteria can be used with the present embodiments. The system for grouping genes 300 groups genes at significant threshold expression levels and correlates these genes with various biological pathways or functions by conducting statistical analysis of the output gene data 230 provided from the system for providing gene data 200. A "p" value can be assigned by assimilating all the data together. This represents the confidence level of expression of certain target genes and their expression levels. In order to eliminate faulty conclusions, each of the data provided by the CERNO software 310, the transcription factor database 320, or the ontology database 330 are compared and analyzed. This can be done using parallel or serial processing. It should be noted that the CERNO software 310, the transcription factor database 320, and the ontology data base 330, analyze associated gene data upstream, downstream and similar to output gene data 230, to identify and initially group important biomarkers targets, genes, or gene signatures. EXAMPLE -5
[000038] Analysis of gene expression profiles for the purpose of discovering molecular mechanisms of drug action requires an appropriate biological or clinical context high-dimensional data interpretation. To demonstrate the feasibility of applying biological context to the output of gene expression profiles, three specific biological themes were integrated into the analysis of expression profile data to provide for hypothesis testing in a common statistical framework. The first biological theme integrated into data analysis was transcription factor (TF) binding 320, in which sets of genes annotated as bound by a common factor were used to inform regarding events upstream of gene expression; the TF binding knowledge facilitated pre-expression inference. The second theme employed gene predictor sets defined by canonical pathways and functions, such as interferon inducible genes, providing post- expression inference. Specifically, selected biological effects downstream of drug- mediated gene expression changes were identified and correlated to specific members of the gene predictor sets of interferon inducible genes. Lastly, gene predictor sets defined by genes encoding the interaction neighbors of interferon were analyzed, which allowed gene expression to implicate proteins whose activity may have been influenced by a changing molecular environment. This protein networks analysis was implemented to support peri-expression inference, or inclusion-by-association.
Formalizing the integration of gene expression profiles into these knowledge environments permitted statistical inference and thus prioritized specific effects among the myriad influences of IFNb-lb treatment on cells and physiology. The analyses allowed development of multi-gene signatures that were then rigorously tested and validated in an independent dataset. Moreover, the approach highlights specific connections among these specific biological themes and the scientific literature on MS. Tables 2-12 disclose the genes utilized and the corresponding selected probe sets to detect each gene. The SEQ ID NO for each representative mRNA sequence is also given. To determine the mRNA nucleotide sequence, the gene name and the Entrez Gene ID number was used to search the NIH NCBI database (http://www.ncbi.nlm.nih.gov), and the accession number with nucleotide sequence information identified. In those cases where the database contained several variant mRNA sequences for a gene, the longest mRNA sequence was used for the mRNA sequence of the gene. In the case of the ND6 mitochondrial gene, that portion of the mitochondrial genome is identified (MT BP 12148 - 14672).
EXAMPLE -6
Knowledgebases
[000039] Data used to define gene sets by transcription factor binding were obtained from MSigDB (http://www.broadinstitute.org/gsea/msigdb/index.jsp, Subramanian et al. 2005), specifically the c3 TFT gene sets, and by searching the Biomolecular Object Network Databank (http://bond.unleashedinformatics.com) for DNA-protein interactions. Pathway Commons (Cerami et al. 2010, 2009 download) provided canonical pathways and a network of interacting proteins. Genes in the pathways and network were identified by Entrez Gene identifiers.
EXAMPLE -7
Data handling
[000040] Data handling (microarray) was performed using Perl DBI (database) and PostgreSQL, and statistical analysis was performed in R (R Development Core Team 2005). Perl-DBI was used to interact with available databases. Perl DBI provided complete file input and output capabilities and syntax for line-at-a-time sequential input and text manipulation/parsing. PostgreSQL was used as an enterprise-level database providing data-mining and data handling properties for large texts and dynamic content. Statistical analysis in R was performed to provide a programming language and environment for statistical computing and graphic compatible with analyzed databases.
EXAMPLE -8
Short- and long-term differential gene expression [000041] The Coincident Extreme Ranks in Numerical Observations (CERNO) method was presented in Yamaguchi et al, 2008. This biology grouping system is based on an analysis of genes to investigate IFNb-lb short-term effects and MS disease gene expression. To facilitate a large number of gene set tests, a nested CERNO testing routine replaced the Monte Carlo approach previously presented. Under the null hypothesis that probe set T-test p-values are randomly ordered, the following condition holds for probe sets, i, ...,ηβ mapping to a gene G, the area under the upper tail of the
Figure imgf000015_0001
Ina) is a conservative p-value, p(G), for rejecting the null hypothesis, where is the rank of the T-test p-value for the ith probe set, and na is the total number of probe sets measuring transcripts from HUGO named genes. The alternative hypothesis is that the p-values for probe sets measuring transcripts from the gene G are more extreme (lower) than under the null. In turn, to evaluate a set of genes, j = 1, ...,ns, the area under the upper tail of the distribution with the lower limit at fs = -2^1og[r(G) . /nG] is a conservative p-value, p(S), for rejecting the null hypothesis, where r(G)j is the rank of p(G) value for the jth gene, and no is the total number of HUGO named genes. Simulation verified that the nested CERNO routine provided nearly (0-1) uniformly distributed p(S) values under the null hypothesis, and was only slightly conservative due to the discrete distribution of ranks deviating from a continuous distribution, which was a negligible effect for the thousands of probe sets and genes measured in this study.
[000042] For each predictor set gene set ¾ of interferon-inducible genes a p(S) was computed based on the T-test p-values calculated for the appropriate group or pair wise comparison in Tables 2-1 1. Benjamini-Hochberg false discovery rates (FDR) were calculated for each collection of predictor gene set test p-values (e.g., test results for all such sets defined by transcription factor binding) and for each sample comparison separately.
[000043] An integrated, cross-micro array-platform database was employed to enable flexible querying of microarray results. Queries were defined and executed to perform a biological grouping by selecting genes regulated by IFNb-lb and annotated with specific functions or pathways. Subsequent to systematic data queries,
Affymetrix 3' array probe sets were identified by review of CLIMB study data, Affymetrix tissue expression data available on both whole transcript and 3' arrays, and additional Affymetrix data as available. Selection of 3 ' array probe sets, with the intent to capture the effects measured on the HuEx 1.0 ST array, were completed prior to running tests on the Goertsches study published data. Supplementary information was provided with more in-depth descriptions of signature definition criteria, as well as the lists of the selected genes and probe sets.
[000044] Signature testing was performed using globaltest (Goeman et al. 2004, R library global test version 5.2.0), a multiple regression method that used only the gene expression results of a set of transcript measurements, and thus explicitly ignored the expression levels estimated from the entire microarray. As globaltest does not specifically model multiple transcript measurements per gene, only one transcript or probe set measurement per gene was permitted in any signature.
EXAMPLE - 9
IFNb-lb treatment differential expression and timescales overview
[000045] To characterize IFNb-lb effects on gene regulation at different timescales, differential expression in MS patient PBMCs was investigated at 4, 18, and 43 hours post-IFNb-lb dose compared to pre-dose levels ("short-term" comparisons). Pharmacokinetics of differential gene expression as measured in MS patients after a single dose of IFNb-lb using HuEx 1.0 ST arrays was similar to that previously observed using the Affymetrix 3' HG-U133 Plus 2.0 platform. These changes were contrasted with differential expression in MS patient PBMCs after approximately 2 years of IFNb-lb therapy compared to therapy-na'ive MS patient PBMCs ("long-term" comparison).
[000046] A larger number of differentially expressed genes were observed in the long-term comparison than in any short-term comparison, whether differential expression was measured by a nominal p-value cutoff or by FDR analysis (See FIG. 2). Referring to FIG. 2, differentially expressed genes at the <0.05 cutoff are roughly three-fold more abundant in the 2-year comparison than in the short-term comparisons. No probe sets exhibit a false discovery rate (FDR) of 0.05 or less in the short-term comparisons. 264 genes have at least one probe set that exhibits a t-test FDR of 0.05 in the 2-year comparison (black bar). In all, 715 probe sets map to these 264 genes. FDR analysis of t-test p-values revealed that only in the long-term comparison did any transcripts reach significance. These transcripts mapped to a gene signature of 264 genes, the majority of which are regulated by long-term IFNb- lb treatment. Although the range of expression ratios was similar for short- and long- term effects, more significant up- and down-regulation was observed in the long-term.
EXAMPLE- 10
Pathways and network analysis of long-term IFNb-lb effects in patients
[000047] To further characterize long-term gene expression changes in interferon-inducible genes, three gene predictor set analyses were computed to infer protein effects upstream, downstream, and interacting with the measured expression changes. The first analysis addressed gene predictor comprised of genes bound by known transcription factors, and clearly showed the contrast between short-term and long-term transcriptional changes (See FIG. 3). Referring to FIG. 3, gene sets characterized by transcription factor binding (rows) show excessive differential expression in sample comparisons (columns) as indicated by FDR CERNO values. The most significant differentially expressed gene sets (deep shaded) are distinct between short-term and long-term comparisons. Independent data from the CLIMB study (Affymetrix 3 ' array HG-U133 Plus 2.0) show evidence of both short- and long- term treatment effects, which corroborates the multi-gene analysis of the HuEx 1.0 ST data. The multi-gene analysis approach permitted corroboration of the distinct short- term and long-term gene expression patterns using independently sampled CLIMB study data, which is comprised of patient samples drawn at various times after beginning IFN-b therapy. [000048] In the second gene set analysis, downstream effects in the long-term- IFNb-lb treatment comparison were explored. A large number of canonical pathways were influenced by treatment (224 pathways of 953 tested at FDR<0.05), as measured by differentially expressed gene predictor set of interferon-inducible genes. To aid interpretation of the changed pathways and to relate them to the single-gene changes, the overlap of the 264 long-term differentially expressed genes forming a gene signature 500 and the top- 100 differentially expressed pathways was explored. A discrete twenty-eight gene signature 500 was identified in the overlap, and gene signature was used in a cluster analysis to group the pathway effects (See FIG. 4). Referring to Figure 4, the top- 100 differentially expressed pathways as detected with CERNO point to three major effects of IFNb-lb long-term treatment. Twenty-eight of the 264 long-term IFN -lb regulated genes map to these 100 pathways. The membership matrix was ordered using hierarchical clustering, revealing a pattern that includes a cluster of functions and genes related to beta-oxidation of fatty acids.
These effects targeted energy metabolism, apoptosis, and gene transcription. Within energy metabolism, mitochondrial beta-oxidation of fatty acids comprised the intersection of a set of annotated pathways and a distinct pattern of gene membership. Prominent among the changed pathways in the long-term treatment comparison were those involved in mitochondrial function and energy metabolism, both of which have been described to play an important role in maintaining neuronal integrity (Gonsette, R. E 2007). Maintaining homeostasis within the central nervous system appears to be of critical importance in MS and other neurological diseases, and the observations herein suggested a role in that function for IFNb-lb.
[000049] To identify important proteins, and not necessarily ones in named pathways, the third gene predictor set analysis of interferon-inducible genes was applied to protein interaction neighborhoods (See FIG. 5). Referring to FIG. 5, specific protein neighborhoods, such as those surrounding IKBKE and GPRC5C, exhibit excessive change, as implied by the long-term IFN -lb treatment comparison. Each name represents a gene set of the proteins that interact with the named protein. The results of the multi-gene test for excessive differential expression are represented by the CERNO -value (y-axis) plotted against the size of the gene sets (x-axis). The sizes of the text names on the plot are scaled to emphasize more significant test results adjusted for the number of genes in the multi-gene test. The red line indicates the FDR 0.05 cutoff for the -values adjusted for the 7433 protein neighborhood tests. This peri-expression analysis revealed disrupted levels of gene transcripts
significantly associated with neighbors of particular proteins. As larger sets provide more power to detect smaller effects on test ranks, the plotting technique identifier p- values sets as a function of set size. Three of the most differentially expressed neighborhoods of different sizes confirm the ability to isolate the gene signatures 500.
[000050] The most significantly long term differentially expressed protein neighborhood gene set, IKBKE (343 loci, FDR<=2e-l 1), highlighted the ΙκΒ kinase (IK )-related kinase ΙΚ ε. Another strongly differentially expressed interaction neighborhood was labeled GPRC5C (See FIG. 4, towards lower left, 3 loci,
FDR=0.002). All three neighbors of GPRC5C exhibited encoding genes regulated by IFNb-lb, and individually exhibited transcripts significantly differentially expressed at /KO.001. One, OLA1, yielded a significant FDR (FDR = 0.02) and was thus a member of the 264 loci identified by differential expression analysis.
[000051 ] The neighborhood around OXCT 1 was notably regulated by IFNb-lb treatment (See FIG. 5 and 6 loci, FDR = 5e-5). SCOT is the protein encoded by OXCT1 and is a homodimeric mitochondrial matrix enzyme that plays a central role in extrahepatic ketone body catabolism by catalyzing the reversible transfer of coenzyme A from succinyl-CoA to acetoacetate.
EXAMPLE -11
Long-term IFNb-lb treatment effects compared to MS disease gene expression changes
[000052] To explore the relationship between treatment and MS disease, a gene signature 500 comprises the 264 long-term-regulated genes were examined by plotting the expression ratios comparing long-term treated patients to therapy-na'ive patients against the ratios comparing therapy-na'ive patients to healthy controls.
Significant long-term gene expression changes tended to reverse disease-modified gene expression, in contrast to short-term IFNb-lb effects (See FIG. 6). Referring to FIG. 6, long-term, but not short-term, IFN effects counter-regulate disease-related changes. Short-term changes: genes exhibiting the most significant changes (t-test <0.002) observed 4 hours post injection tend to be either unchanged in (stable) disease or change in the same direction as characteristic of MS patients, which is especially true of those probe sets more significantly changed in disease (p<0.05, dark points show positive slope). Long-term changes: Expression ratios in the long-term- treated patients versus therapy-na'ive patients compared to the ratios found in therapy- na'ive patients relative to healthy controls. The long-term regulated gene expression for these differentially expressed probe sets (FDR <0.05 based on t-test ) tends to reverse the expression changes that characterize disease, especially when focusing on the probe sets exhibiting more significant changes in disease (p<0.05, dark points show negative slope).
System for grouping genes for association
with a particular disease
[000053] Referring to FIG. 7, the statistical analysis and grouping of genes to a physiological state uses a suitable system 300 results in statistically grouped biology gene data associated with a particular biology or biological process 410. This data is then subject to further processing by the system for grouping genes for association with a particular disease 400. The biology gene data 410 is then subject to further filtering and refinement using an analytic filter 420. For instance, analytic filter 420 may comprise review and comparisons to literature data and information, can be based on observable and quantifiable patient data, observation or anecdotal information, or can be subject to further software review and/or processing. Other data sources for refinement includes, but are not be limited to CERNO, clinical data, scientist review, key opinion leader observation, conventional knowledge, disease- associated animal models, genome-wide transcriptomics, disease susceptibility markers, medical records, unpublished information from scientific meetings, Advisory Boards, consultants. The results of the above processing provide both statistical and analytically grouped gene signatures 500. The gene signatures 500 comprise groups of genes associated with a common underlying biology. EXAMPLE -12
Selecting genes for multi-gene signatures
[000054] The observation of long-term-specific effects motivated the decision to define five long-term expression gene signatures 500. Because gene signatures 500 were intended for cross-platform use, criteria were defined using significance of change observed in the HuEx 1.0 ST array data and in some cases a lower boundary on expression differential ratio. Three short-term effect signatures were defined to capture the short-term effects common to 4h and 18h comparisons (ST-4h+18h), as well as the effects specific to 4h (ST-4h) or 18h (ST-18h) (Tables 1-3). Two long- term IFNb-lb gene signatures 500, designated LT (no short term regulation in the same direction as long-term regulation) and LT2 (more stringent, required that none of the transcripts be regulated in the short term), were defined (Tables 5-6).
Signatures associated with biologies thought to be important in MS pathophysiology and treatment, were defined to target the mitochondrial (MT-OX), IKBKE (IKE), NRF-2, and blood-brain barrier (BBB) functions (Tables 7-10). A tenth signature for disease reversal (DR) was defined by the overlap of long-term regulation and stable RRMS changes, requiring each transcript to change in opposite directions (Table 11).
[000055] Optionally, the gene signatures 500 may be subject to further optional systems and processing. For instance, in Figure 7, a system for confirming gene signatures 600 can be used to independently confirm the accuracy of the statistical and analytical grouped gene signatures 500. The system for confirming gene signatures 600 can comprise literature review and confirmations, comparison to CLIMB data, validation studies, and use, application and comparison to other databases known or unknown in the art including clinical studies, disease-associated animal models, pharmacodynamic or pharmacokinetic studies, toxicity studies, pharmacoviligence studies, all available genome-wide disease susceptibility and transcriptomic data.
EXAMPLE- 13 Testing multi-gene signatures: IFNb-lb signatures confirmed in independent data, and distinct from GA gene expression effects
[000056] Since the ten signatures were developed using the Reder HuEx 1.0 ST dataset and adapted to the HG-U133 Plus 2.0 microarray platform using (in part) the CLIMB IFNb data, results from testing the signatures in these same data were largely unsurprising (Tables 3). Eight of the ten were significant in the CLIMB IFNb data at p <= 0.05, with signatures NRF2 (Table 8) and DR (Table 11) as the exceptions.
[000057] The Goertsches dataset (Goertsches et al.) provided an entirely independent set of IFNb- lb-treated patient data in which to test the ten signatures (Table 3). All ten signatures were significant at p <= 0.05 in at least one of the comparisons.
[000058] To investigate the specificity of the signatures for IFNb treatment, significance in the GA-treated arm of the CLIMB study was tested (Table 12). Only the IKE signature (Table 10) reached significance at p = 0.0032.
[000059] The signatures were tested in the Reder patient data grouped by MS subtype ( Tables 2-11). The DR signature (Tablel 1) reached significance in all four subtypes. Additionally, the BBB signature (Table 9) was significant in active RRMS and PPMS, and ST-18h, LT, and MT-OX (Table7) were significant in PPMS (p <= 0.05).
[000060] IFN-b bioactivity markers have been proposed previously (Comabella et al. 2009; Van Baarsen 2008; Vosslamber et al. 2009), but few have been validated in independent studies or demonstrated to predict treatment outcomes relevant to clinical endpoints such as disability, exacerbation rate, cognition loss, brain atrophy, and EDSS. A limited number of studies have suggested that certain IFN-b bioactivity markers, e.g., transient, neutralizing antibody effects, may well predict MRI outcomes. However, improved treatment response markers that are associated with MS-relevant regulation of biological function are needed (Arnason et al 1996;
Agrawal SM, Yong WW. 2007; Dhib-Jalbut S, Marks S. 2010; Hafler, DA, et al. 2005; Weiner HL. 2009). Treatment outcome biomarkers identified from IFN-b- treated MS patient differential gene expression studies have been a challenge to confirm (Croze 2010). To remedy the lack of consistent results across studies, biological modes of IFNb-lb effects were identified and then defined as multi-gene signatures targeting these effects to boost statistical power compared to single-gene models. These 10 multi-gene signatures were developed in a well-controlled differential gene expression study (Reder et al. 2008; Yamaguchi et al, 2008), confirmed in an independent gene expression study, and represent specific biological themes with relevance to clinical outcomes.
[000061] For MS, current drugs have no apparent common mechanism, which raises the possibility that a specific biomarker of response may be necessary for each MS drug. Side effects of available MS drugs can be problematic and serious, leading in some cases to death (Clifford et al. 2010). By studying a wider range of biological influences of each MS drug, patient response bio markers are analyzed by
measurement of a plurality of gene signatures. Differential drug effects on
therapeutically relevant molecular systems yields informed decisions regarding course of disease treatment methods and drug selection and informs a decision to adapt with another MS drug that targets a different or complementing mechanism of action.
EXAMPLE - 14
Relationship of treatment effects to MS gene expression
[000062] The analysis that compared drug-regulated gene expression and disease differential expression (See FIG. 3) revealed evidence of IFNb activity biomarkers in MS patients that were compared to healthy controls, which is consistent with our previously published findings (Yamaguchi et al, 2008). Moreover, the finding that long-term IFNb-lb treatment reversed some MS-characteristic gene expression changes (FIG. 3) raised the possibility that these MS patients exhibited an IFNb deficiency that was corrected by IFNb-lb treatment.
[000063] Long-term IFNb-lb treatment gene expression effects were clearly distinct from short-term effects, the latter influencing immunomodulation and antiviral responses. Long-term treatment instead influenced a range of detectable biological themes including metabolic functions, mitochondrial energetics, and a specific anti-oxidant mechanism, which are all potential influences on preservation of neuronal homeostasis. Individually or collectively, use of gene expression data further characterized by association with biological themes yields a defined group of genes characterized as a gene signature 500 defined the unique ability of that signature 500 to yield a functional biomarker for presence or progression of disease.
Metabolic functions and IFNb-lb treatment
[000064] Fatigue and cognitive function are associated with decreased brain metabolism in MS patients (Rolcke et al. 1997, Sorensen et al. 2006). The current study presented broad, multi-gene evidence for IFNb-lb regulation of mitochondrial function, in particular metabolic processes that control the balance of fatty acids and their derivative biomolecules. Importantly, mitochondrial changes could mediate neuroprotective effects of IFNb-lb via mechanisms such as influencing levels of ketone bodies (Guzman and Blazquez, 2004) or effects on carnitine regulation. (The dysregulation of the specific function "import of palmitoyl-CoA into the
mitochondrial matrix" in FIG. 4 is consistent with effects on carnitine regulation.) The proposed roles of mitochondrial dysfunction in MS were the subject of an extensive review (Mao and Reddy, 2010). These observations motivated the data- supported investigation of the enzymatic processes controlling ketone body formation.
[000065] 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) is a key intermediate in the formation of ketone bodies in mitochondria, cholesterol, and steroids in the cytoplasm. Interestingly, the cytosolic balance of HMG-CoA, and not the mitochondrial, showed the most direct evidence of regulation by IFNb-lb. The Reder array data and the CLIMB study showed no statistically significant effect of IFNb-lb on transcript levels for HMG-CoA synthase 2 (mitochondrial), the first reaction in ketogenesis. (However, the Goertsches study did show an elevation, p < 0.05 at all time points). In contrast, for HMG-CoA synthase 1 (soluble), the Reder array data showed a significant down regulation for one transcript, but the CLIMB and the Goertsches (one month) comparisons revealed up-regulation of a probe set on the 3' array (data in the public domain, ref. GEO accessions). Moreover, HMG-CoA reductase was decreased long-term by IFNb-lb treatment (p = 0.002, FDR = 0.1). Currently, this important enzyme target of statins is being considered for its potential as a multiple sclerosis treatment (Paintlia et al 2008, Willey and Elkind 2010, Wood et al. 2010, Sellner at al 2010). The ability of IFNb-lb to influence the levels of key lipid metabolism enzymes suggests that individual patient response to combination therapy may depend on specific IFNb-lb effects on such enzyme levels.
[000066] Accordingly, a defect in mitochondrial function as a biological theme exhibited by the brain in MS patients emphasizes the importance of an alternative energy source for neurons in such patients and suggests that IFNb-lb affects extrahepatic ketone body catabolism.
Antioxidant role of IFNb-lb treatment
[000067] Given the functional importance of antioxidant activity as a biological theme in both inflammatory and neurodegenerative pathophysiology, the most- significant findings from the long-term treatment comparison were reviewed from the perspective of changes in MS-relevant antioxidant activity. Three proteins, NADH dehydrogenase 6 (ND6), GPRC5C, and OLA1, and the evidence for their relevance will be discussed.
[000068] Relative to treatment-na'ive patients, IFNb- lb-treated MS patients exhibited significantly higher expression of the mitochondrial encoded NADH dehydrogenase 6 gene. Expression of wild-type mt-ND6 appears to help maintain normal levels of ROS, thereby reducing the likelihood of oxidative damage to cells, and mutated mt-ND6 appears causative in Leber hereditary optic neuropathy, a disease which co-occurs with MS and MS-like white matter abnormalities. Therefore, IFNb-lb may play an important role in regulating mitochondrial mediated oxidative- stress responses that likely play a key role in MS.
[000069] In addition to the potentially antioxidant role of ND6 in IFNb treatment, the GPRC5C neighborhood was the most differentially expressed set in its size class. GPRC5C itself was mildly differentially expressed, being increased by IFNb-lb treatment (23% increase in long-term comparison, /? = 0.015,). This G protein-coupled receptor family protein has been suggested to mediate the cellular effects of retinoic acid on the G protein signal transduction cascade
(http://www.ncbi.nlm.nih.gov/ pubmed/10945465. In addition to being a member of the gene signature 500 containing the 264 genes identified in the FDR analysis of the long-term treatment comparison, OLA1 was a contributing neighbor of GPRC5C and decreased by treatment, an observation corroborated in the CLIMB study data (p = 0.002). Recently this gene was found to play an important role in the regulation of antioxidant response (Zhang et al. 2009). The NRF2-independent, transcription-independent regulation of oxidative stress implicates a potent neuroprotective mechanism for IFNb-lb.
Linking treatment effects to signal transduction
[000070] ΙΚ ε (encoded by IKBKE) is a link between the protein network analysis, in which ΙΚ ε was the hub of the most significantly regulated network neighborhood, and the canonical pathway analysis, in which mitochondrial energy metabolism was highlighted. ΙΚ ε regulates energy balance (Chiang et al. 2009). Moreover, the role of ΙΚ ε in long-term IFNb treatment is consistent with the dependence between ΙΚ ε and IFNb functions observed in an in vivo antiinflammatory setting (Corr et al). Recombinant IFNb induces rapid phosphorylation of ΙΚ ε in vitro (Tenoever et al. 2007). ΙΚΚε is critical to IFNb's roll in response to viral infection (Tenoever et al. 2007) and likely integrates important signals from pattern recognition receptors that detect virus nucleic acids (Hiscott 2004).
Interestingly, DDX58 (aka RIG-1) is a potentially important viral detection mechanism that signals through IRF-dependent pathways, is up-regulated in three IFNb treatment studies. Proteins upstream of interferon genes are regulated by treatment, and thus the feedback mechanism of interferon response is itself modified. ΙΚ ε is a key part of this feedback loop and thus should play a role in multiple sclerosis treatment response.
Gene Signatures
[000071] All 10 gene signatures 500 identified herein were found to be significantly regulated in the Goertsches study data, indicating that both long-term and short-term effects were captured in the Goertsches patient profiles. Moreover, the gene signatures 500 targeting IFNb- lb-regulated biological functions were also significantly regulated. There is a notable trend over time toward less significant change in the Goertsches study data, as Goertsches et al). The 10 gene signatures 500 provide a family of response patterns targeting distinct molecular mechanisms of action and are to be correlated to clinical response to clarify which IFNb-lb effects are most indicative of different clinical outcomes.
[000072] Treatment of MS patients with GA reduces disease burden and progression (Carter and Keating, 2010). However, a cohort of patients appears to be non-responsive to treatment. Gene signatures 500 derived from IFNb treatment gene expression studies were not changed by GA treatment, with the exception of the IKE signature (Table 10). These observations confirm that the gene signatures 500 measure IFNb- lb-specific treatment effects, but more importantly suggest a potential advantage for combination drug use and possible criteria for guiding patient-specific care. The IKE gene signature 500 changes point to a possible common pathway of response.
[000073] The large multi-gene expression patterns disclosed herein reflect biological effects observable across platforms and patient cohorts, and to confirm smaller multi-gene signatures in independent study data. The protein interaction network analysis of gene expression microarray data strongly underscored a role for ΙΚΚε in IFNb-lb treatment effects, especially observable over long-term treatment. Data analysis showed long-term and short-term effects of IFNb-lb treatment were not only distinct but actually contrasted in direction of change. Moreover, the long-term IFNb-lb treatment effects highlighted energy metabolism regulation and the potential for neuroprotective effects mediated by HMG-CoA-derived bio molecules and antioxidant proteins.
[000074] Referring to FIG. 7, each step of a method comprises providing an expansive predictor set of genes 700, statistically grouping the predictor set of genes 700 with biological functions 800, analytically grouping genes for association with a particular disease 850, selecting genes signatures 900, and optionally confirming the gene signatures 950.
[000075] The expansive predictor set of genes 700 is generally provided by the system for providing the nucleic acid/gene data 200. As previously discussed, the source of gene data 210 is subjected to a device for analyzing nucleic acids 220. The device for analyzing nucleic acids 220 can comprise any number of devices known in the art. For instance, the device for analyzing nucleic acids 220 can comprise a micro array device. In the case a micro array is used, the gene data is contacted to the micro array device to hybridize the nucleic acids to the micro array. Various methodologies can then be performed that are well known in the art for reading and analyzing the gene data. For instance, various markers or dyes can be used for determining if the target nucleic acids have hybridized to the micro array. Identified nucleic acids or genes can then be read using various readers known in the art. As a result, the gene data 230 can then be used for further processing and analysis. Other devices known in the art can be employed. The devices should be capable of identifying various nucleic acids targets of interest and quantifying their level of gene expression. Once the gene data 230 has been determined, it can then be further processed.
[000076] Next, the gene data is organized into a gene predictor set and subjected to further processing. Gene data can be described but not limited to the robust microchip average (RMA) algorithm with default parameters. CERNO provides an example in which p-values derive from logarithmic expression values for a given gene chip relating probe sets P min and Pmax to a lower mean log ranking indicating a stronger strength of signal or p-value. For instance, the gene predictor set is further grouped statistically using various biological functions 800. The statistical grouping according to biological functions is accomplished by subjecting the gene predictor set data to a combination of processes that can be performed serially or in parallel. For instance, the gene predictor set can be processed using various software that can compare identified nucleic acid data and sequences with already characterized and identified nucleic acids and sequences that have been stored in known public or private gene databases. These gene databases and processing are particularly helpful in identifying those associated nucleic acids and genes that can be upstream from the gene predictor set.
[000077] Another process associates nucleic acid or gene data derived from transcription factor binding by various proteins to identify nucleic acids and genes that have been up or down regulated. For instance, transcription factors will have bound to those areas around the nucleic acid or target of interest. This binding provides additional details and data regarding up or down regulated genes. Further, this data provides an output pattern reaching absolute and differential gene expression. These results yield a gene predictor set. [000078] Another process associates nucleic acid or gene data downstream from a gene predictor set analyzing nucleic acid and gene data linked to physiological, physical, or disease states is then compared to the gene predictor set to identify matches in sequence and/or expression data. All processes may occur simultaneously or in tandem when analyzing the predictor set. In each case, data is classified according to biological functions by system 300 using statistical methods. The statistical methods analyze the data and provide an associated "p" value that can be statistically analyzed. For instance, various thresholds eliminate irrelevant expression data such that meeting defined thresholds in upstream, downstream, and transcription factor binding are matched to one or more particular biological functions. After the gene predictor set has been associated or correlated with various biological functions, the results can then be subjected to further processing and analysis.
[000079] In the next stage of processing, the gene data is then subjected to analysis for association with various diseases 850. This processing takes one or more of the associated biological functions that were determined in step 800 and then subjects them individually or in combinations to determine relative significance to particular disease states. The system for grouping genes for association with a particular disease 400 combines one or more of the associated biological functions 410 and subjects this data to analytical filter 420. Analytic filter 420 works by using CER O 310, clinical data, known literature results, databases including research based MS patient samples (Reder, CLIMB, Goertsches databases, all published transcriptomics studies, public MS disease associated databases and disease- associated animal studies, ex vivo and in vitro studied immune cell, immune cell subtype ,other cell-based databases , or other information in combination or separately to match diseased states with associated biological function and gene data. In the final processing step the gene data, associated biological functions and diseased states are then grouped to determine gene signatures 500. This step is accomplished by using certain threshold levels to distinguish which important correlated data is most important in diseased and non-diseased patients and which associated data meet certain "p" values of significance.
[000080] The final defined gene signature can then define those genes that when treated or not treated by a therapeutic treatment or drug, will have the change relative to the relative biological function. This provides for opportunities to further refine the associations for more personalized medicine. The gene signature 500 can then be optionally subjected to a system for confirming gene signatures 600. This validation step can use known databases, CLIMB data, or other validation studies to further improve accuracy.
[000081] The above methods steps and processes in the above mentioned embodiments need not occur in the above mentioned order. As will be clear to those of known skill in the art it may be possible to change the order of the steps of the processing or methods, combine steps, eliminate steps, slightly alter or change, or add additional steps to the above mentioned steps and processes. These changes or alterations are within the scope and spirit of the disclosed embodiments and invention.
Table 1. Reder patient characteristics.
Figure imgf000031_0001
Table 2: ST-4h + 18h (Short-Term 4h and 18h)
'Not present on HG-U133A or HG-U133B
GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED
GENE ID ACCESSION HuEx ST HG-U133 PLUS
NUMBER Transcript 2.0 PROBE
ID SET
APOL6 80830 NM 030641.3 SEQ ID NO: 1 3944243 1557236 at1
CNP 1267 NM 033133.4 SEQ ID NO: 2 3721548 208912 s at
DDX58 23586 NM 014314.3 SEQ ID NO: 3 3203086 218943 s at
DDX60 55601 NM 017631.5 SEQ ID NO: 4 2792800 218986 s at
DDX60L 91351 NM 001012967.1 SEQ ID NO: 5 2792940 228152 s at
DHX58 79132 NM 024119.2 SEQ ID NO: 6 3757602 219364 at
DTX3L 151636 NM 138287.3 SEQ ID NO: 7 2638962 225415 at
EIF2AK2 5610 NM 002759.3 SEQ ID NO: 8 2548402 213294 at
EPSTI1 94240 NM 001002264.2 SEQ ID NO: 9 3511698 235276 at
HERC5 51191 NM 016323.3 SEQ ID NO: 10 2735409 219863 at
HERC6 55008 NM 017912.3 SEQ ID NO: 11 2735362 219352 at
HSH2D 84941 NM 032855.2 SEQ ID NO: 12 3823583 1552623 at1
IFI44 10561 NM 006417.4 SEQ ID NO: 13 2343511 214453 s at
IFI44L 10964 NM 006820.2 SEQ ID NO: 14 2343504 204439 at
IFIH1 64135 NM 022168.3 SEQ ID NO: 15 2584207 219209 at
IFIT1 3434 NM 001270930.1 SEQ ID NO: 16 3257246 203153 at
IFIT2 3433 NM 001838005.2 SEQ ID NO: 17 3257192 226757 at
IFIT3 3437 NM 001549.4 SEQ ID NO: 18 3257204 229450 at
IFIT5 24138 NM 012420.2 SEQ ID NO: 19 3257268 203595 s at
IRF9 10379 NM 006084.4 SEQ ID NO: 20 3529701 203882 at
JUP 3728 NM 002230.2 SEQ ID NO: 21 3757329 201015 s at
MX1 4599 NM 001144925.1 SEQ ID NO: 22 3922100 202086 at
OAS1 4938 NM 016816.2 SEQ ID NO: 23 3432438 205552 s at
OAS2 4939 NM 002535.2 SEQ ID NO: 24 3432514 206553 at
OAS3 4940 NM 006187.2 SEQ ID NO: 25 3432467 232666 at
OASL 8638 NM 003733. SEQ ID NO: 26 3474831 205660 at
PARP12 64761 NM 022750.2 SEQ ID NO: 27 3075932 218543 s at
PARP14 54625 NM 017554.2 SEQ ID NO: 28 2639054 224701 at
PARP9 83666 NM 001146106.1 SEQ ID NO: 29 2692060 223220 s at
PRIC285 85441 NM 001037335.2 SEQ ID NO: 30 3913960 228230 at
RNF213 57674 NM 001256071.1 SEQ ID NO: 31 3737336 225929 s at
RSAD2 91543 NM 080657.4 SEQ ID NO: 32 2468351 213797 at
SAMD9 54809 NM 017654.3 SEQ ID NO: 33 3061438 219691 at
SAMD9L 219285 NM 152703.2 SEQ ID NO: 34 3061456 230036 at
STAT2 6773 NM 005419.3 SEQ ID NO: 35 3457752 205170 at
TNFSF10 8743 NM 003810.3 SEQ ID NO: 36 2705706 202688 at
UBE2L6 9246 NM 198183.2 SEQ ID NO: 37 3373962 201649 at GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED GENE ID ACCESSION HuEx ST HG-U133 PLUS
NUMBER Transcript 2.0 PROBE
ID SET
USP18 11274 NM 017414.3 SEQ ID NO: 38 3938046 219211 at
WARS 7453 NM 004184.3 SEQ ID NO: 39 3579546 200629 at
XAF1 54739 NM 017523.3 SEQ ID NO: 40 3708074 206133 at
Table 3: ST-4h (Short-Term 4h-Specific)
GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED HG- GENE ID ACCESSION HuEx ST U133 PLUS 2.0
NUMBER TRANSCRI PROBE SET PT ID
ATF3 467 NM 001040619.2 SEQ ID NO: 41 2379132 202672 s at
CXCL10 3627 NM 001565.3 SEQ ID NO: 42 2773958 204533 at
DNAJA1 3301 NM 001539.2 SEQ ID NO: 43 3166718 200880 at
FAS 355 NM 000043.4 SEQ ID NO: 44 3257098 204780 s at
HESX1 8820 NM 003865.2 SEQ ID NO: 45 2677902 211267 at
IL1RN 3357 NM 000577.4 SEQ ID NO: 46 2501204 216243 s at
RGL1 23179 NM 015149.3 SEQ ID NO: 47 2371346 209568 s at
TDRD7 23424 NM 014290.2 SEQ ID NO: 48 3181193 213361 at
Table 4: ST-18h (Short Term 18h-Specific)
GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED HG- GENE ID ACCESSION HuEx ST U133 PLUS 2.0
NUMBER TRANSCRI PROBE SET PT ID
C3AR1 719 NM 004054.2 SEQ ID NO: 49 3442941 209906 at
CECR1 51816 NM 017424.2 SEQ ID NO: 50 3951768 219505 at
CSF1R 1436 NM 005211.3 SEQ ID NO: 51 2881187 203104 at
CYBB 1536 NM 000397.3 SEQ ID NO: 52 3973839 203922 s at
FGL2 10875 NM 006682.2 SEQ ID NO: 53 3057955 204834 at
GBP4 115361 NM 052941.4 SEQ ID NO: 54 2421995 235175 at
GRN 2896 NM 002087.2 SEQ ID NO: 55 3722917 211284 s at
IFI30 10437 NM 006332.3 SEQ ID NO: 56 3824874 201422 at
KCNMB1 3779 NM 004137.2 SEQ ID NO: 57 2886679 209948 at
KDM1B 221656 NM 153042.3 SEQ ID NO: 58 2897111 227021 at
MOV10 4343 NM 020963.3 SEQ ID NO: 59 2352275 223849 s at
MTHFD2 10797 NM 006636.3 SEQ ID NO: 60 2489172 201761 at
NCF2 4688 NM 000433.3 SEQ ID NO: 61 2447414 209949 at
PLAGL2 5326 NM 002657.3 SEQ ID NO: 62 3902682 202925 s at
SH3BP2 6452 NM 001145855.1 SEQ ID NO: 63 2715629 209370 s at
SLC31A2 1318 NM 001860.2 SEQ ID NO: 64 3185498 204204 at
STAB1 23166 NM 015136.2 SEQ ID NO: 65 2623922 38487 at
TCN2 6948 NM 000355.3 SEQ ID NO: 66 3942472 204043 at
TLR2 7097 NM 003264.3 SEQ ID NO: 67 2748346 204924 at
TNFSF13B 10673 NM 006573.4 SEQ ID NO: 68 3500787 223501 at
USP32 84669 NM 032582.3 SEQ ID NO: 69 3765907 211702 s at
Table S: LT (Long-Term)
2 Not present on HG-U133A or HG-U133B
3Not present on HG-U133A or HG-U133B
"MT-ND6 (BP14149-14673)
GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED GENE ID ACCESSION HuEx ST HG-U133 PLUS
NUMBER TRANSCRIPT 2.0 PROBE SET
ID
ABCE1 6059 NM 002940.2 SEQ ID NO 70 2746024 201872 s at
ABCF2 10061 NM 007189.1 SEQ ID NO 71 2632123 207623 at
ADAM9 8754 NM 003816.2 SEQ ID NO 72 3095002 1555326 a at2
AIDA 64853 NM 022831.2 SEQ ID NO 73 2749466 220199 s at
ARHGAP21 57584 NM 020824.3 SEQ ID NO 74 2962157 241701 at
ARPC5 10092 NM 001270439.1 SEQ ID NO 75 2017725 1569325 at2
CHMP2B 25978 NM 014043.3 SEQ ID NO 76 2631845 202536 at
CYFIP2 26999 NM 001037332.2 SEQ ID NO 77 2883431 215785 s at
DYNLL1 8655 NM 001037494.1 SEQ ID NO 78 3474573 200703 at
EIF4B 1975 NM 001417.4 SEQ ID NO 79 3987440 211938 at
ENPP7 339221 NM 178543.3 SEQ ID NO 80 3689249 238245 at
FAM85A 619423 XR 110590.1 SEQ ID NO 81 3403708 227917 at
FAM91A1 157769 NM 144963.2 SEQ ID NO 82 3114407 237045 at
FYB 2533 NM 001465.4 SEQ ID NO 83 2807453 211794 at
GABRA5 2558 NM 00810.3 SEQ ID NO 84 4039528 206456 at
GPX1 2876 NM 201397.1 SEQ ID NO 85 2674229 200736 s at
HADHA 3030 NM 000182.4 SEQ ID NO 86 2545092 208630 at
HCG4 54435 NR 002139.2 SEQ ID NO 87 2948024 206685 at
HK2 3099 NM 000189.4 SEQ ID NO 88 4013702 202934 at
ITPPJP 85450 NM 033397.2 SEQ ID NO 89 3305313 225582 at
KDELR3 11015 NM 006855.2 SEQ ID NO 90 3945314 207264 at
KLHL28 54813 NM 017658.3 SEQ ID NO 91 3562669 235727 at
KPNB1 3837 NM 002265.4 SEQ ID NO 92 3078594 213507 s at
LPJG2 9860 NM 014813.1 SEQ ID NO 93 2428564 238653 at
MALAT1 378938 NR 002819.2 SEQ ID NO 94 3377656 224558 s at
MMRN1 22915 NM 007351.2 SEQ ID NO 95 2735759 205612 at
MRFAP1 93621 NM 033296.1 SEQ ID NO 96 2717049 226091 s at
MTMR14 64419 NM 001077525.2 SEQ ID NO 97 2609770 222143 s at
ND6 4541 J01415.2* SEQ ID NO 98 4037708 1553575 at2
NDUFA5 4698 NM 005000.2 SEQ ID NO 99 3690325 201304 at
NLRP3 114548 NM 004895.4 SEQ ID NO 100 2390050 207075 at
OAT 4942 NM 000274.3 SEQ ID NO 101 3311157 201599 at
PDCD11 22984 NM 014976.1 SEQ ID NO 102 3304761 212424 at
PGGT1B 5229 NM 005023.3 SEQ ID NO 103 2871685 206288 at
PHLDB2 90102 NM 001134438.1 SEQ ID NO 104 2635975 225688 s at
PMAIP1 5366 NM 021127.2 SEQ ID NO 105 3790704 204285 s at
PPIA 5478 NM 021130.3 SEQ ID NO 106 3926380 226336 at
PROS1 5627 NM 000313.3 SEQ ID NO 107 2685268 207808 s at
PRR12 57479 NM 020719.1 SEQ ID NO 108 3838718 226716 at
PSMD13 5719 NM 175932.2 SEQ ID NO 109 3315549 201232 s at
RANBP6 26953 NM 012416.3 SEQ ID NO 110 3197869 213019 at
RBL2 5934 NM 005611.3 SEQ ID NO 111 3691962 212331 at
RIT1 6016 NM 001256821.1 SEQ ID NO 112 2437736 209882 at
RPL10 6134 NM 006013.3 SEQ ID NO 113 4027289 200724 at
RPL41 6171 NM 001035267.1 SEQ ID NO 114 3426092 201492 s at
RPL5 6125 NM 000969.3 SEQ ID NO 115 2423199 200937 s at
S100A12 6283 NM 005621.1 SEQ ID NO 116 2435981 205863 at
SDHC 6391 NM 003001.3 SEQ ID NO 117 2362618 215088 s at
SDHD 6392 NM 003002.2 SEQ ID NO 118 2325988 202026 at
SELP 6403 NM 003005.3 SEQ ID NO 119 2443417 206049 at
SNAPC5 10302 NM 006049.2 SEQ ID NO 120 3630178 1554093 a at3
SNRPE 6635 NM 003094.2 SEQ ID NO 121 3198043 203316 s at
SQSTM1 8878 NM 001142298.1 SEQ ID NO 122 3999965 201471 s at SRRM2 23524 NM 016333.3 SEQ ID NO 123 3677141 216629 at
SUPT4H1 6827 NM 003168.1 SEQ ID NO 124 344174 201484 at
SYCN 342898 NM 001080468.2 SEQ ID NO 125 3861881 229995 at
TBC1D2 55357 NM 001267571.1 SEQ ID NO 126 3217236 222173 s at
TFPI 7035 NM 006287.4 SEQ ID NO 127 2591482 210664 s at
TMEM170B 100113407 NM 001100829.1 SEQ ID NO 128 2895043 235798 at
TMEM200C 645369 NM 001080209.1 SEQ ID NO 129 3797283 229523 at
TMSB15B 286527 NM 194324.2 SEQ ID NO 130 4016766 1570039 at3
TRAF3IP3 80342 NM 025228.2 SEQ ID NO 131 2453869 240265 at
TRIM66 9866 NM 014818.1 SEQ ID NO 132 3361956 213748 at
TSN 7247 NM 004622.2 SEQ ID NO 133 2503618 201515 s at
UFD1L 7353 NM 005659.6 SEQ ID NO 134 3802392 209103 s at
VHL 7428 NM 000551.3 SEQ ID NO 135 2610336 203844 at
ZGPAT 84619 NM 032527.4 SEQ ID NO 136 3893659 57539 at
ZMAT2 153527 NM 144723.1 SEQ ID NO 137 2832081 224782 at
ZNF187 7741 NM 001023560.2 SEQ ID NO 138 2900449 213218 at
ZNF384 171017 NM 001135734.2 SEQ ID NO 139 3402684 212369 at
ZNF518B 85460 NM 053042.2 SEQ ID NO 140 2760624 226909 at
Table 6: LT2 (Long-Term 2)
4Not present on HG-U133A or HB-U133B
"MT-ND6 (BP14149-14673)
GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED
GENE ID ACCESSION HuEx ST HG-U133 PLUS
NUMBER TRANSCRIPT 2.0 PROBE SET
ID
ABCE1 6059 NM 002940.2 SEQ ID NO 70 2746024 201872 s at
ABCF2 10061 NM 007189.1 SEQ ID NO 71 2632123 207623 at
CHMP2B 25978 NM 014043.3 SEQ ID NO 76 2631845 202536 at
CYFIP2 26999 NM 001037332.2 SEQ ID NO 77 2883431 215785 s at
ENPP7 339221 NM 178543.3 SEQ ID NO 80 3689249 238245 at
FYB 2533 NM 001465.4 SEQ ID NO 83 2807453 211794 at
HADHA 3030 NM 000182.4 SEQ ID NO 86 2545092 208630 at
HCG4 54435 NR 002139.2 SEQ ID NO 87 2948024 206685 at
KDELR3 11015 NM 006855.2 SEQ ID NO 90 3945314 207264 at
KLHL28 54813 NM 017658.3 SEQ ID NO 91 3562669 235727 at
MALAT1 378938 NR 002819.2 SEQ ID NO 94 3377656 224558 s at
MRFAP1 93621 NM 033296.1 SEQ ID NO 96 2717049 226091 s at
ND6 4541 J01415.2* SEQ ID NO 98 4037708 1553575 at4
NDUFA5 4698 NM 005000.2 SEQ ID NO 99 3690325 201304 at
OAT 4942 NM 000274.3 SEQ ID NO 101 3311157 201599 at
PHLDB2 90102 NM 001134438.1 SEQ ID NO 104 2635975 225688 s at
PROS1 5627 NM 000313.3 SEQ ID NO 107 2685268 207808 s at
PSMD13 5719 NM 175932.2 SEQ ID NO 109 3315549 201232 s at
SDHD 6392 NM 003002.2 SEQ ID NO 118 2325988 202026 at
SNAPC5 10302 NM 006049.2 SEQ ID NO 120 3630178 1554093 a at4
SNRPE 6635 NM 003094.2 SEQ ID NO 121 3198043 203316 s at
SQSTM1 8878 NM 001142298.1 SEQ ID NO 122 3999965 201471 s at
SYCN 342898 NM 001080468.2 SEQ ID NO 125 3861881 229995 at
TRAF3IP3 80342 NM 025228.2 SEQ ID NO 141 2453869 240265 at
TSN 7247 NM 004622.2 SEQ ID NO 133 2503618 201515 s at
VHL 7428 NM 000551.3 SEQ ID NO 135 2610336 203844 at
ZGPAT 84619 NM 032527.4 SEQ ID NO 136 3893659 57539 at
ZMAT2 153527 NM 144723.1 SEQ ID NO 137 2832081 224782 at
ZNF187 7741 NM 001023560.2 SEQ ID NO 138 2900449 213218 at ZNF384 171017 NM 001135734.2 SEQ ID NO: 139 3402684 212369 at
ZNF518B 85460 NM 053042.2 SEQ ID NO: 140 2760624 226909 at
Table 7: MT-OX (Mitochondrial Fatty-Acid Beta Oxidation)
^Present on both HG-U133A andHG-U133B
"Not present on HG-U133A or HG-U133B
"MT-ND6 (BP14149-14673)
GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED
GENE ID ACCESSION HuEx ST HG-U133 PLUS
NUMBER TRANSCRIPT 2.0 PROBE SET
ID
COX4I1 1327 NM 001861.3 SEQ ID NO: 142 3672455 200086 s at5
CPT1C 126129 NM 001199752.1 SEQ ID NO: 143 3838845 227468 at
CS 1431 NM 004077.2 SEQ ID NO: 144 3824767 208660 at
ETFDH 2110 NM 004453.2 SEQ ID NO: 145 2749560 33494 at
HADHA 3030 NM 000182.4 SEQ ID NO: 86 2545092 208630 at
HADHB 3032 NM 000183.2 SEQ ID NO: 146 2473735 201007 at
ND6 4541 J01415.2* SEQ ID NO: 98 4037708 1553575 at6
NDUFB10 4716 NM 004548.2 SEQ ID NO: 147 3644220 223112 s at
NDUFS3 4722 NM 004551.2 SEQ ID NO: 148 3329904 201740 at
NDUFS7 374291 NM 024407.4 SEQ ID NO: 149 3845166 242168 at
SDHC 6391 NM 003001.3 SEQ ID NO: 117 2363618 215088 s at
SDHD 6392 NM 003002.2 SEQ ID NO: 118 2325988 202026 at
UQCRC2 7385 NM 003366.2 SEQ ID NO: 150 3652218 200883 at
UQCRFS1 7386 NM 006003.2 SEQ ID NO: 151 3946088 208909 at
Table 8: Nl F2 (NRF2 Function)
GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED
GENE ID ACCESSION HuEx ST HG-U133 PLUS
NUMBER TRANSCRIPT 2.0 PROBE SET
ID
ARPC5 10092 NM 001270439.1 SEQ ID NO 75 2917725 211963 s at
CAP1 10487 NM 006367.3 SEQ ID NO 152 2331727 200625 s at
DDOST 1650 NM 005216.4 SEQ ID NO 153 2400220 208675 s at
KEAP1 9817 NM 203500.1 SEQ ID NO 154 3850363 202417 at
PAFAH1B1 5048 NM 000430.3 SEQ ID NO 155 2500368 200816 s at
PAFAH1B2 5049 NM 002572.3 SEQ ID NO 156 3972241 210160 at
PHLDA2 7262 NM 0033113 SEQ ID NO 157 3359461 209803 s at
PSMD1 5707 NM 0028073 SEQ ID NO 158 2531712 201198 s at
PSMD13 5719 NM 175932.2 SEQ ID NO 109 3315549 201232 s at
PSMD2 5708 NM 002808.3 SEQ ID NO 159 2655650 200830 at
RTN3 10313 NM 001265589.1 SEQ ID NO 160 2746104 219549 s at
SDHD 6392 NM 003002.2 SEQ ID NO 118 2325988 202026 at
SQSTM1 8878 NM 001142298.1 SEQ ID NO 122 3999965 201471 s at
ZNF384 171017 NM 001135734.2 SEQ ID NO 139 3402684 212369 at
Table 9: Bl BB(Blood-B rain Barrier and Endothelial Function)
GENE ENTREZ EXAMPLE SEQ ID NO: SELECTED SELECTED
GENE ID ACCESSION HuEx ST HG-U133 PLUS
NUMBER TRANSCRIPT 2.0 PROBE SET
ID
AGRN 375790 NM 198576.3 SEQ ID NO: 161 4053550 212285 s at
CD 163 9332 NM 004244.5 SEQ ID NO: 162 3442706 203645 s at ICAM1 3383 NM 000201.2 SEQ ID NO: 163 3820443 202638 s at
TFPI 7035 NM 006287.4 SEQ ID NO: 127 2591482 213258 at
TYMP 1890 NM 001113756.2 SEQ ID NO: 164 3966036 217497 at
Table 10: IKE (IKBK -Related)
Figure imgf000036_0001
Table 11: PR (Disease Reversal)
Figure imgf000036_0002
Table 12. IFN -lb effects si natures confirmed b inde endent data,
Figure imgf000036_0003
Values are globaltest p-values for multi-gene signatures, Values in bold are individually significant at p < 0.05. References
1. National MS Society, http://www.nationalmssociety.org/
2. Lundy, J., B.M. Craig. The use of disease-modifying agents amoung Multiple sclerosis patients enrolled in Medicare from 1995-2002 and the impact of Medicare Part D: Analysis of claims data fronm the medicare current Beneficiary Survey.Clinical Therapeutics vol 28, 2006, p 140-145.
3. Hafler, DA, Slavik J, Anderson DE, O'Connor K, De Jager P, Baecher-Allan C.
Multiple Sclerosis. 2005. Immunological Reviews 204:208-231
4. Weiner HL. 2009. The challenge of multiple sclerosis: how do we cure a chronic heterogenous disease? Ann of Neurology 65:239-248.
5. The IFN beta Multiple Sclerosis Study Group. 1993. Interferon beta- lb is effective in relapsing-remitting multiple sclerosis. I Clinical results of a multicenter, randomized, double-blind, placebo-controlled trial. Neurology 43:655-661.
6. Gold, R. Oral Therapies for Multiple Sclerosis: A Review of Agents in Phase III Development or Recently Approved CNS Drugs2011 25 (1) pp 37-52.
7. Markowitz C, The current landscape and umet need in multiple sclerosis.
American Journal of Managed Care. V 16, 2010 pp 211-218.
8. Wagner, J., Innovations driving biomarker evaluation and use. Biomarkers in medicine 4 (6) 2010 pp 779-781.
9. Croze E. Differential gene expression and translational approaches to identify biomarkers of interferon beta activity in multiple sclerosis. J Interferon Cytokine Res. 2010 Oct;30(10):743-9.
10. Micheel, C. M., J.R. Ball. Evaluation of biomarkers and surrogate endpoints in chronic disease. National Aca Press, Washington, D.C. 2010.
Vandenbroeck K, Urcelay E, Comabella M. IFN-beta pharmacogenomics in multiple sclerosis. Pharmacogenomics. 2010 Aug;l 1(8): 1137-48.
Stark GR, Kerr, IM, Williams BR, Silverman RH, and Schreiber RD, "How cells respond to interferons," Annual Review of Biochemistry, vol. 67, pp. 227-264, 1998
Mogensen, K. E., M. Lewerenz, J. Reboul, G. Lutfalla, and G. Uze, "The type I interferon receptor: structure, function, and evolution of a family business," Journal of Interferon & Cytokine Research, vol. 19, no. 10, pp. 1069-1098, 1999. Dhib-Jalbut S, Marks S. 2010. Interferon-β mechanisms of action in multiple sclerosis. Neurology 74:S17-S24.
Agrawal SM, Yong WW. 2007. Immunopathogenesis of MS. Int Rev Neurobiol 79:99-126.
Gok B, Okutan O, Beskonakli E, Palaoglu S, Erdamar H, Sargon MF. Effect of immunomodulation with human interferon-beta on early functional recovery from experimental spinal cord injury. Spine (Phila Pa 1976). 2007 Apr 15;32(8):873- 80.
Marsh B, Stevens SL, Packard AE, Gopalan B, Hunter B, Leung PY, Harrington CA, Stenzel-Poore MP. Systemic lipopolysaccharide protects the brain from ischemic injury by reprogramming the response of the brain to stroke: a critical role for IRF3. J Neurosci. 2009 Aug 5;29(31):9839-49.
Yoshimura S, Ochi H, Isobe N, Matsushita T, Motomura K, Matsuoka T, Minohara M, Kira J. Altered production of brain-derived neurotrophic factor by peripheral blood immune cells in multiple sclerosis. Mult Scler. 2010 Oct;16(10): l 178-88. Epub 2010 Jul 23. Yamaguchi KD, Ruderman DL, Croze E, Wagner TC, Velichko S, Reder AT, et al. IFN-beta-regulated genes show abnormal expression in therapy-naive relapsing-remitting MS mononuclear cells: gene expression analysis employing all reported protein-protein interactions. J Neuroimmunol. 2008;195(1-2): 116-20. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43): 15545-50.
McDonald, W. I., A. Compston, G. Edan, et al, "Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis," Annals of Neurology, vol. 50, no. 1, pp. 121-127, 2001
Kurzke JF. Rating neurological impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983; 33: 1444-1452.
http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE16214
http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE24427
Goertsches RH, Hecker M, Koczan D, Serrano-Fernandez P, Moeller S, Thiesen H-J, et al. Long-term genome-wide blood RNA expression profiles yield novel molecular response candidates for IFN-beta-lb treatment in relapsing remitting MS. Pharmacogenomics. 2010;11(2): 147-61.
http://www.affymetrix.com/
Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Research 2003;31(4):el5. http://www.broadinstitute.org/gsea/msigdb/index.jsp
http://bond.unleashedinformatics.com/ Cerami EG, Gross BE, Demir E, Rodchenkov I, Babur O, Anwar N, et al. Pathway Commons, a web resource for biological pathway data. Nucleic Acids Res. 2011;39 (Database issue):D685-90. 2009 download.
R Development Core Team. R Foundation for Statistical Computing, Vienna, Austria. 2005. ISBN 3-900051-07-0. Available at: http://www.R-project.org. Monte Carlo ref.
Benjamini-Hochberg ref.
Goeman JJ, van de Geer SA, de Kort F, van Houwelingen HC. A global test for groups of genes: testing association with a clinical outcome. Bioinformatics. 2004;20(l):93-9. R library global test version 5.2.0. Reder, A.T., S. Velichko, K. D. Yamaguchi, et al, "IFN- ?lb induces transient and variable gene expression in relapsing-remitting multiple sclerosis patients independent of neutralizing antibodies or changes in IFN receptor RNA expression," Journal of Interferon & Cytokine Research, vol. 28, no. 5, pp. 317— 331, 2008.
Gonsette, R. E., "N euro degeneration in multiple sclerosis: The role of oxidative stress and excitotoxicity," Journal of the Neurological Sciences, vol. 274, no. 1-2, pp. 48-53, 2008.
Comabella M, Lunemann JD, Rio J, Sanchez A, Lopez C, Julia E, Fernandez M, Nonell L, Camina-Tato M, Deisenhammer F, Caballero E, Tortola MT, Prinz M, Montalban X, Martin R. 2009. A type I interferon signature in monocytes is associated with poor response to interferon-beta in multiple sclerosis. Brain. 132:3353-3365. Van Baarsen LG, Vosslamber S, Tijssen M, Baggen JM, van der Voort LF, Killestein J, van der Pouw Kraan TC, Polman CH, Verweij CL. 2008. Pharmacogenomics of interferon-β therapy in multiple sclerosis: baseline IFN signature determines pharmacological differences between patients. PloS 3:E1927.
Vosslamber S, van Baarsen L, Verweij CL. 2009. Pharmacogenomics of IFN-β in multiple sclerosis: towards a personalized medicine approach. Pharmacogenomics 10:97-108
Arnason BGW, Dayal A, Qu ZX, Jensen M, Gen? K, Reder AT. 1996. Mechanisms of action of interferon beta in multiple sclerosis. Semin Immunopathol 18: 125-148.
Clifford DB, De Luca A, Simpson DM, Arendt G, Giovannoni G, Nath A. Natalizumab-associated progressive multifocal leukoencephalopathy in patients with multiple sclerosis: lessons from 28 cases. Lancet Neurol 2010 Apr;9(4):438- 46.
Rolcke 1997
Sorensen et al. 2006
Guzman M, Blazquez C. Ketone body synthesis in the brain: possible neuroprotective effects. Prostaglandins Leukot Essent Fatty Acids. 2004;70(3):287-92.
Mao P, Reddy PH. Is multiple sclerosis a mitochondrial disease? Biochim Biophys Acta. 2010;1802(l):66-79.
Paintlia AS, Paintlia MK, Singh I, Singh AK. Combined medication of lovastatin with rolipram suppresses severity of experimental autoimmune encephalomyelitis. Exp Neurol. 2008;214(2): 168-80. Willey JZ, Elkind MSV. 3-Hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors in the treatment of central nervous system diseases. Arch Neurol. 2010;67(9): 1062-7.
Wood WG, Eckert GP, Igbavboa U, Muller WE. Statins and neuroprotection: a prescription to move the field forward. Ann N Y Acad Sci. 2010;1199:69-76. Sellner J, Weber MS, Vollmar P, Mattle HP, Hemmer B, Stuve O. The combination of interferon-beta and HMG-CoA reductase inhibition in multiple sclerosis: enthusiasm lost too soon? CNS Neurosci Ther. 2010;16(6):362-73. Zhang J, Rubio V, Lieberman MW, Shi Z-Z. OLA1, an Obg-like ATPase, suppresses antioxidant response via nontranscriptional mechanisms. Proc Natl Acad Sci U S A. 2009; 106(36): 15356-61.
Chiang S-H, Bazuine M, Lumeng CN, Geletka LM, Mowers J, White NM, et al. The protein kinase IK epsilon regulates energy balance in obese mice. Cell. 2009;138(5):961-75.
Corr M, Boyle DL, Ronacher L, Flores N, Firestein GS. Synergistic benefit in inflammatory arthritis by targeting I kappaB kinase epsilon and interferon beta. Ann Rheum Dis. 2009;68(2):257-63.
Tenoever BR, Ng S-L, Chua MA, McWhirter SM, Garcia-Sastre A, Maniatis T. Multiple functions of the IK -related kinase IKKepsilon in interferon-mediated antiviral immunity. Science. 2007;315(5816): 1274-8.
Hiscott J. Another detour on the Toll road to the interferon antiviral response. Nat Struct Mol Biol. 2004;11 : 1028-1030.
Carter NJ. Keating GM Glatiramer acetate: a review of its use in relapsing- remitting multiple sclerosis and in delaying the onset of clinically definite multiple sclerosis. Drugs. 2010 Aug 20;70(12): 1545-77.

Claims

CLAIMS We Claim:
1. A system for determining gene signatures, comprising:
(a) a gene database comprising a gene predictor set comprised of genes whose expression is induced by interferon;
(b) a biology grouping system comprising means for associating the gene predictor set with a biological state to yield a biological function gene set; and
(c) a disease grouping system for analytically grouping the biological function gene set with multiple sclerosis to yield a gene signature comprising at least one gene selected from Tables 2-12.
2. The system of claim 1, wherein the means for associating the gene
predictor set with a biological state is CER O.
3. The system of claim 1, wherein the means for associating the gene
predictor set with a biological state is a transcription factor database.
4. The system of claim 1, wherein the means for associating the gene
predictor set with a biological state is an ontology database.
5. A system as recited in claim 1, wherein the gene database further
comprises a device for analyzing nucleic acids.
6. A system as recited in claim 1, wherein the gene database further
comprises a source of gene data.
7. The system as recited in claim 6, wherein the source of gene data
comprises a micro array designed for detection of gene expression.
8. The system as recited in claim 1, wherein the disease grouping system is CERNO.
9. The system as recited in claim 1, wherein the biology grouping system further comprises multiple sclerosis medical literature.
10. The system as recited in claim 1, wherein the disease grouping system is comprised of clinical data.
11. The system as recited in claim 1 , wherein the gene signatures are
comprised of short term expression selected from Tables 2-4.
12. The system as recited in claim 1, wherein the gene signatures are
comprised of gene exhibiting long term expression selected from Tables 5- 6.
13. The system as recited in claim 1, wherein the gene signatures are
comprised of genes exhibiting mitochondrial fatty acid beta oxidation selected from Table 7.
14. The system as recited in claim 1, wherein the gene signatures are
comprised of genes exhibiting NRF2 functions selected from Table 8.
15. The system as recited in claim 1, wherein the gene signatures are
comprised of genes exhibiting blood brain barrier and endothelial function selected from Table 9.
16. The system as recited in claim 1, wherein the gene signatures are
comprised of genes exhibiting IKBKE related expression selected from Table 10.
17. The system as recited in claim 1, wherein the gene signatures are
comprised of genes exhibiting disease reversal selected from Table 11.
18. A system for determining gene signatures for character of a multiple
sclerosis treatment, comprising:
(a) a gene database comprising a gene predictor set comprising genes whose expression is induced by interferon;
(b) means for selecting a first set of genes from the predictor set having a biological function; (c) means for grouping the first set of genes having the biological function with a second set of genes that encode the interaction neighbors of a protein expressed by a gene in the first set; and
(d) means for comparing expression of genes from the first set with the second set.
19. A method for analyzing gene signatures during therapeutic treatment of multiple sclerosis, comprising:
(a) providing a gene predictor comprising genes whose expression is induced by interferon;
(b) selecting a set of differentially expressed genes from the predictor set having a common biological function;
(c) selecting gene sets encoding interaction neighbors of the genes
expressed in step (b) in a sample from a therapeutic treatment;
(d) analyzing the differential expression from the therapeutic treatment in gene signatures selected from the genes in Tables 2-12.
20. A method as recited in claim 19, wherein the gene predictor set comprises genes in patients treated with IFN-b.
21. A method as recited in claim 21, wherein the gene predictor set comprises genes in patients treated with IFNb-b.
22. A method as recited in claim 20, wherein the affected genes are
upregulated.
23. A method as recited in claim 21, wherein the affected genes are down regulated.
PCT/US2012/064247 2011-11-08 2012-11-08 Systems and integrated methods for gene expression analysis in multiple sclerosis WO2013070984A1 (en)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
US201113292015A 2011-11-08 2011-11-08
US13/292,015 2011-11-08
US201213531992A 2012-06-25 2012-06-25
US13/531,992 2012-06-25

Publications (1)

Publication Number Publication Date
WO2013070984A1 true WO2013070984A1 (en) 2013-05-16

Family

ID=47222317

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2012/064247 WO2013070984A1 (en) 2011-11-08 2012-11-08 Systems and integrated methods for gene expression analysis in multiple sclerosis

Country Status (1)

Country Link
WO (1) WO2013070984A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2017033886A1 (en) * 2015-08-24 2017-03-02 国立大学法人東京大学 Oxytocin sensitivity prediction marker
US10650031B2 (en) 2009-02-10 2020-05-12 Ayasdi Ai Llc Systems and methods for visualization of data analysis

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050064483A1 (en) * 2003-08-28 2005-03-24 Baylor College Of Medicine Gene expression profiling technology for treatment evaluation of multiple sclerosis
WO2010033624A1 (en) * 2008-09-16 2010-03-25 Bayer Healthcare Llc Interferon response in clinical samples (iris)

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050064483A1 (en) * 2003-08-28 2005-03-24 Baylor College Of Medicine Gene expression profiling technology for treatment evaluation of multiple sclerosis
WO2010033624A1 (en) * 2008-09-16 2010-03-25 Bayer Healthcare Llc Interferon response in clinical samples (iris)

Non-Patent Citations (53)

* Cited by examiner, † Cited by third party
Title
"Interferon beta-lb is effective in relapsing-remitting multiple sclerosis. I Clinical results of a multiccntcr, randomized, double-blind, placebo-controlled trial", NEUROLOGY, vol. 43, 1993, pages 655 - 661
AGRAWAL SM; YONG WW: "Immunopathogenesis of MS", INT REV NEUROBIOL, vol. 79, 2007, pages 99 - 126
AMASON BGW; DAYAL A; QU ZX; JENSEN M; GENE K; REDER AT: "Mechanisms of action of interferon beta in multiple sclerosis", SEMIN IMMUNOPATHOL, vol. 18, 1996, pages 125 - 148
CARTER NJ; KEATING GM: "Glatiramer acetate: a review of its use in relapsing-remitting multiple sclerosis and in delaying the onset of clinically definite multiple sclerosis", DRUGS, vol. 70, no. 12, 20 August 2010 (2010-08-20), pages 1545 - 77
CCRAMI EG; GROSS BE; DCMIR E; RODCHCNKOV I; BABUR O; ANWAR N: "Pathway Commons, a web resource for biological pathway data", NUCLEIC ACIDS RCS., 2011, pages 39
CHIANG S-H; BAZUINE M; LUMENG CN; GELETKA LM; MOWERS J; WHITE NM ET AL.: "The protein kinase IKKepsilon regulates energy balance in obese mice", CELL, vol. 138, no. 5, 2009, pages 961 - 75
CLIFFORD DB; DE LUCA A; SIMPSON DM; ARENDT G; GIOVANNONI G; NATH A.: "Natalizumab-associated progressive multifocal leukoencephalopathy in patients with multiple sclerosis: lessons from 28 cases", LANCET NEUROL, vol. 9, no. 4, April 2010 (2010-04-01), pages 438 - 46
COMABELLA M; LÜNEMANN JD; RIO J; SÁNCHEZ A; LOPEZ C; JULIÀ E; FERNÁNDEZ M; NONELL L; CAMINA-TATO M; DEISENHAMMER F: "A type T interferon signature in monocytes is associated with poor response to interferon-beta in multiple sclerosis", BRAIN, vol. 132, 2009, pages 3353 - 3365, XP055072343, DOI: doi:10.1093/brain/awp228
CORR M; BOYLE DL; RONACHER L; FLORES N; FIRESTEIN GS: "Synergistic benefit in inflammatory arthritis by targeting I kappaB kinase epsilon and interferon beta", ANN RHEUM DIS., vol. 68, no. 2, 2009, pages 257 - 63
CROZE E.: "Differential gene expression and translational approaches to identify biomarkers of interferon beta activity in multiple sclerosis", J INTERFERON CYTOKINE RES., vol. 30, no. 10, October 2010 (2010-10-01), pages 743 - 9
DHIB-JALBUT S; MARKS S.: "Interferon-? mechanisms of action in multiple sclerosis", NEUROLOGY, vol. 74, 2010, pages S 17 - S24
GOEMAN JJ; VAN DE GEER SA; DE KORT F; VAN HOUWELINGEN HC.: "global test for groups of genes: testing association with a clinical outcome", BIOINFORMATICS, vol. 20, no. 1, 2004, pages 93 - 9
GOERTSCHES RH; HECKER M; KOCZAN D; SERRANO-FERNANDEZ P; MOELLER S; THIESEN H-J ET AL.: "Long-term genome-wide blood RNA expression profiles yield novel molecular response candidates for IFN-beta-lb treatment in relapsing remitting MS", PHARMACOGENOMICS, vol. 11, no. 2, 2010, pages 147 - 61, XP009182717
GOK B; OKUTAN O; BESKONAKLI E; PALAOGLU S; ERDAMAR H; SARGON MF.: "Effect of immunomodulation with human interferon-beta on early functional recovery from experimental spinal cord injury", SPINE, vol. 32, no. 8, 15 April 2007 (2007-04-15), pages 873 - 80
GOLD, R: "Oral Thcrapics for Multiple Sclerosis: A Review of Agents in Phase III Development or Recently Approved CNS", DRUGS, vol. 25, no. 1, 2011, pages 37 - 52
GONSETTE, R. E.: "Neurodegeneration in multiple sclerosis: The role of oxidative stress and excitotoxicity", JOURNAL OF THE NEUROLOGICAL SCIENCES, vol. 274, no. 1-2, 2008, pages 48 - 53, XP025587925, DOI: doi:10.1016/j.jns.2008.06.029
GUZMAN M; BLAZQUEZ C.: "Ketone body synthesis in the brain: possible neuroprotective effects", PROSTAGLANDINS LEUKOT ESSENT FATTY ACIDS, vol. 70, no. 3, 2004, pages 287 - 92
HAFLER, DA; SLAVIK J; ANDERSON DE; O'CONNOR K; DE JAGER P; BAECHER-ALLAN C.: "Multiple Sclerosis", IMMUNOLOGICAL REVIEWS, vol. 204, 2005, pages 208 - 231
HILPERT J ET AL: "Biological response genes after single dose administration of interferon beta-1b to healthy male volunteers", JOURNAL OF NEUROIMMUNOLOGY, ELSEVIER SCIENCE PUBLISHERS BV, XX, vol. 199, no. 1-2, 13 August 2008 (2008-08-13), pages 115 - 125, XP023178133, ISSN: 0165-5728, [retrieved on 20080618], DOI: 10.1016/J.JNEUROIM.2008.04.036 *
HISCOTT J.: "Another detour on the Toll road to the interferon antiviral response", NAT STRUCT MOL BIOL., vol. 11, 2004, pages 1028 - 1030
IRIZARRY RA; BOLSTAD BM; COLLIN F; COPE LM; HOBBS B; SPEED TP.: "Summaries of Affymetrix GeneChip probe level data", NUCLEIC ACIDS RESEARCH, vol. 31, no. 4, 2003, pages EL5, XP002460628, DOI: doi:10.1093/nar/gng015
KURZKE JF.: "Rating neurological impairment in multiple sclerosis: an expanded disability status scale (EDSS", NEUROLOGY, vol. 33, 1983, pages 1444 - 1452
LUNDY, J; B.M. CRAIG: "Analysis of claims data fronm the medicare current Beneficiary", SURVEY.CLINICAL THERAPEUTICS, vol. 28, 2006, pages 140 - 145
MAO P; REDDY PH.: "Is multiple sclerosis a mitochondrial disease?", BIOCHIM BIOPHYS ACTA, vol. 1802, no. 1, 2010, pages 66 - 79, XP026793174
MARKOWITZ C: "The current landscape and umet need in multiple sclerosis", AMERICAN JOURNAL OF MANAGED CARE, vol. 16, 2010, pages 211 - 218
MARSH B; STEVENS SL; PACKARD AE; GOPALAN B; HUNTER B; LEUNG PY; HARRINGTON CA; STENZEL-POORE MP.: "Systemic lipopolysaccharide protects the brain from ischemic injury by reprogramming the response of the brain to stroke: a critical role for lRF3", J NEUROSCI., vol. 29, no. 31, 5 August 2009 (2009-08-05), pages 9839 - 49
MCDONALD, W. I.; A. COMPSTON; G. EDAN ET AL.: "Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis", ANNALS OF NEUROLOGY, vol. 50, no. 1, 2001, pages 121 - 127, XP009053064, DOI: doi:10.1002/ana.1032
MICHEEL, C. M.; J.R. BALL: "Evaluation of biomarkers and surrogate endpoints in chronic disease", 2010, NATIONAL ACA PRESS
MOGENSEN, K. E.; M. LEWERENZ; J. REBOUL; G. LUTFALLA; G. UZE: "The type I interferon receptor: structure, function, and evolution of a family business", JOURNAL OF INTERFERON & CYTOKINE RESEARCH, vol. 19, no. 10, 1999, pages 1069 - 1098
PAINTLIA AS; PAINTLIA MK; SINGH I; SINGH AK: "Combined medication of lovastatin with rolipram suppresses severity of experimental autoimmune encephalomyelitis", EXP NEUROL., vol. 214, no. 2, 2008, pages 168 - 80
R FOUNDATION FOR STATISTICAL COMPUTING, 2005, ISBN: 3-900051-07-0, Retrieved from the Internet <URL:http://www.R-project.org>
REDER, A.T.; S. VELICHKO; K. D. YAMAGUCHI ET AL.: "IFN-?1b induces transient and variable gene expression in relapsing-remitting multiple sclerosis patients independent of neutralizing antibodies or changes in IFN receptor RNA expression", JOURNAL OF INTERFERON & CYTOKINE RESEARCH, vol. 28, no. 5, 2008, pages 317 - 331, XP008143134, DOI: doi:10.1089/jir.2007.0131
SATOH J I ET AL: "T cell gene expression profiling identifies distinct subgroups of Japanese multiple sclerosis patients", JOURNAL OF NEUROIMMUNOLOGY, ELSEVIER SCIENCE PUBLISHERS BV, XX, vol. 174, no. 1-2, 1 May 2006 (2006-05-01), pages 108 - 118, XP027901471, ISSN: 0165-5728, [retrieved on 20060501] *
SATOH JUN-ICHI ET AL: "Microarray analysis identifies a set of CXCR3 and CCR2 ligand chemokines as early IFN[beta]-responsive genes in peripheral blood lymphocytes in vitro: an implication for IFN[beta]-related adverse effects in multiple sclerosis", BMC NEUROLOGY, BIOMED CENTRAL LTD., LONDON, GB, vol. 6, no. 1, 19 May 2006 (2006-05-19), pages 18, XP021015931, ISSN: 1471-2377, DOI: 10.1186/1471-2377-6-18 *
SELLNER J; WEBER MS; VOLLMAR P; MATTLE HP; HEMMER B; STUVE O.: "The combination of interferon-beta and HMG-CoA reductase inhibition in multiple sclerosis: enthusiasm lost too soon?", CNS NEUROSCI THER., vol. 16, no. 6, 2010, pages 362 - 73
STARK GR; KERR, IM; WILLIAMS BR; SILVERMAN RH; SCHREIBER RD: "How cells respond to interferons", ANNUAL REVIEW OF BIOCHEMISTRY, vol. 67, 1998, pages 227 - 264, XP002543752, DOI: doi:10.1146/annurev.biochem.67.1.227
STÜRZEBECHER S ET AL: "Expression profiling identifies responder and non-responder phenotypes to interferon-beta in multiple sclerosis", BRAIN, OXFORD UNIVERSITY PRESS, OXFORD, GB, vol. 126, 3 May 2006 (2006-05-03), pages 1419 - 1429, XP008126516, ISSN: 0006-8950, DOI: 10.1093/BRAIN/AWG147 *
SUBRAMANIAN A; TAMAYO P; MOOTHA VK; MUKHERJEE S; EBERT BL; GILLETTE MA ET AL.: "Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles", PROC NATL ACAD SCI U S A., vol. 102, no. 43, 2005, pages 15545 - 50, XP002464143, DOI: doi:10.1073/pnas.0506580102
TENOEVER BR; NG S-L; CHUA MA; MCWHIRTER SM; GARCIA-SASTRE A; MANIATIS T.: "Multiple functions of the IKK-related kinase IKKepsilon in interferon-mediated antiviral immunity", SCIENCE, vol. 315, no. 5816, 2007, pages 1274 - 8, XP009150759, DOI: doi:10.1126/science.1136567
VAN BAARSCN LG; VOSSLAMBCR S; TIJSSCN M; BAGGCN JM; VAN DCR VOORT LF; KILLCSTCIN J; VAN DCR POUW KRAAN TC; POLMAN CH; VCRWCIJ CL: "Pharmacogenomics of interferon-(3 therapy in multiple sclerosis: baseline IFN signature determines pharmacological differences between patients", PLOS, vol. 3, 2008, pages E1927
VAN BAARSEN LISA G M ET AL: "Pharmacogenomics of interferon-beta therapy in multiple sclerosis: baseline IFN signature determines pharmacological differences between patients", PLOS ONE, PUBLIC LIBRARY OF SCIENCE, US, vol. 3, no. 4, 2 April 2008 (2008-04-02), pages e1927 - 1, XP002506468, ISSN: 1932-6203, DOI: 10.1371/JOURNAL.PONE.0001927 *
VANDCNBROCCK K; URCCLAY E; COMABCLLA M.: "IFN-bcta pharmacogcnomics in multiple sclerosis", PHARMACOGCNOMICS, vol. L1, no. 8, August 2010 (2010-08-01), pages 1137 - 48
VOSSLAMBER S; VAN BAARSEN L; VERWEIJ CL.: "Pharmacogenomics of IFN-? in multiple sclerosis: towards a personalized medicine approach", PHARMACOGENOMICS, vol. 10, 2009, pages 97 - 108, XP008102751, DOI: doi:10.2217/14622416.10.1.97
WAGNER, J., INNOVATIONS DRIVING BIOMARKER EVALUATION AND USE. BIOMARKERS IN MEDICINE, vol. 4, no. 6, 2010, pages 779 - 781
WANDINGER K-P ET AL: "COMPLEX IMMUNOMODULATORY EFFECTS OF INTERFERON-BETA IN MULTIPLE SCLEROSIS INCLUDE THE UPREGULATION OF T HELPER 1-ASSOCIATED MARKER GENES", ANNALS OF NEUROLOGY, JOHN WILEY AND SONS, BOSTON, US, vol. 50, no. 3, 1 September 2001 (2001-09-01), pages 349 - 357, XP008010479, ISSN: 0364-5134, DOI: 10.1002/ANA.1096 *
WEINER HL.: "The challenge of multiple sclerosis: how do we cure a chronic heterogenous disease?", ANN OF NEUROLOGY, vol. 65, 2009, pages 239 - 248
WILLCY JZ; ELKIND MSV: "3-Hydroxy-3-mcthylglutaryl-cocnzymc A rcductasc inhibitors in the treatment of central ncrvous system diseases", ARCH NCUROL., vol. 67, no. 9, 2010, pages 1062 - 7
WOOD WG; ECKERT GP; IGBAVBOA U; MULLER WE.: "Statins and neuroprotection: a prescription to move the field forward", ANN N Y ACAD SCI., vol. 1199, 2010, pages 69 - 76
YAMAGUCHI K D ET AL: "IFN-beta-regulated genes show abnormal expression in therapy-naive relapsing-remitting MS mononuclear cells: Gene expression analysis employing all reported protein-protein interactions", JOURNAL OF NEUROIMMUNOLOGY, ELSEVIER SCIENCE PUBLISHERS BV, XX, vol. 195, no. 1-2, 1 March 2008 (2008-03-01), pages 116 - 120, XP022612054, ISSN: 0165-5728, [retrieved on 20080214], DOI: 10.1016/J.JNEUROIM.2007.12.007 *
YAMAGUCHI KD; RUDCRMAN DL; CROZC E; WAGNER TC; VCLICHKO S; RCDCR AT: "IFN-bcta-rcgulatcd genes show abnormal expression in thcrapy-naivc rclapsing-rcmitting MS mononuclcar cells: gcnc expression analysis employing all reported protein-protein interactions", J NEUROIMMUNOL., vol. 195, no. 1-2, 2008, pages 116 - 20
YAMAGUCHI, K ET AL.: "Supplementary material", 14 February 2008 (2008-02-14), XP002691509, Retrieved from the Internet <URL:http://www.sciencedirect.com/science/article/pii/S0165572807004456> [retrieved on 20130131] *
YOSHIMURA S; OCHI H; ISOBE N; MATSUSHITA T; MOTOMURA K; MATSUOKA T; MINOHARA M; KIRA J.: "Altered production of brain-derived neurotrophic factor by peripheral blood immune cells in multiple sclerosis", MULT SCLCR., vol. 16, no. 10, 23 July 2010 (2010-07-23), pages 1178 - 88
ZHANG J; RUBIO V; LIEBERMAN MW; SHI Z-Z: "OLAl, an Obg-like ATPase, suppresses antioxidant response via nontranscriptional mechanisms", PROC NATL ACAD SCIU S A., vol. 106, no. 36, 2009, pages 15356 - 61

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10650031B2 (en) 2009-02-10 2020-05-12 Ayasdi Ai Llc Systems and methods for visualization of data analysis
US11263244B2 (en) 2009-02-10 2022-03-01 Ayasdi Ai Llc Systems and methods for visualization of data analysis
US11868376B2 (en) 2009-02-10 2024-01-09 Symphonyai Sensa Llc Systems and methods for visualization of data analysis
WO2017033886A1 (en) * 2015-08-24 2017-03-02 国立大学法人東京大学 Oxytocin sensitivity prediction marker

Similar Documents

Publication Publication Date Title
JP7297015B2 (en) epigenetic chromosomal interactions
Hillen et al. Plasmacytoid DCs from patients with Sjögren's syndrome are transcriptionally primed for enhanced pro-inflammatory cytokine production
JP5670615B2 (en) Diagnosis, prognosis and monitoring of disease progression in systemic lupus erythematosus via microarray analysis of blood leukocytes
JP2010508826A (en) Diagnosis of metastatic melanoma and monitoring of immunosuppressive indicators via blood leukocyte microarray analysis
JP2013526845A (en) Genes and combinations of genes that predict an initial response or non-response of a subject suffering from an inflammatory disease to a cytokine targeted drug (CyTD)
AU2015279621A1 (en) Methods for diagnosing risk of renal allograft fibrosis and rejection
WO2015179777A2 (en) Gene expression profiles associated with sub-clinical kidney transplant rejection
JP2012527895A (en) Characteristics of B cells associated with immune tolerance in transplant recipients
Han et al. Identification of diagnostic mRNA biomarkers in whole blood for ankylosing spondylitis using WGCNA and machine learning feature selection
Liu et al. Genetic and functional associations with decreased anti-inflammatory tumor necrosis factor alpha induced protein 3 in macrophages from subjects with axial spondyloarthritis
Kupfer et al. Microarray characterization of gene expression changes in blood during acute ethanol exposure
AU2008209462A1 (en) Gene expression profiling for identification, monitoring, and treatment of lupus erythematosus
WO2013070984A1 (en) Systems and integrated methods for gene expression analysis in multiple sclerosis
Xuan et al. Do lncRNAs and circRNAs expression profiles influence discoid lupus erythematosus progression?—a comprehensive analysis
US20060115826A1 (en) Gene expression profiling for identification monitoring and treatment of multiple sclerosis
US20110300542A1 (en) Gene Expression Profiling For Identification, Monitoring And Treatment Of Multiple Sclerosis
AU2007272824A1 (en) Gene expression profiling for identification, monitoring and treatment of multiple sclerosis
US20050181516A1 (en) Bioequivalence determination using expression profiling
EP2024514B1 (en) Assessment of effect of an agent on a human biological condition using rodent gene expression panels
JP7491847B2 (en) Precision medicine for pain: diagnostic biomarkers, pharmacogenomics, and repurposed drugs
EP3825418A2 (en) Molecular signatures for distinguishing liver transplant rejections or injuries
Jeon et al. Gene signatures and associated transcription factors of allergic rhinitis: KLF4 expression is associated with immune response
US20130345086A1 (en) Cd4+ t-cell gene signature for rheumatoid arthritis (ra)
Moore et al. Evaluation of methods incorporating biological function and GWAS summary statistics to accelerate discovery
Fazel-Najafabadi et al. A multilayered post-GWAS analysis pipeline defines functional variants and target genes for systemic lupus erythematosus (SLE)

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12791039

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12791039

Country of ref document: EP

Kind code of ref document: A1