Brain Advance Access originally published online on May 6, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Brain, Vol. 126, No. 6, 1419-1429,
June 2003
© 2003 Guarantors of Brain
doi: 10.1093/brain/awg147
Expression profiling identifies responder and non-responder phenotypes to interferon-ß in multiple sclerosis
1 Neuroimmunology Branch, National Institute of Neurological Disorders and Stroke, 2 Metabolism Branch, National Cancer Institute, 3 Laboratory of Diagnostic Radiology Research, Clinical Center,National Institutes of Health, Bethesda, MD, USA, 4 SBU Therapeutics, Schering AG, Berlin, Germany
Correspondence to: Roland Martin, MD, Neuroimmunology Branch, NINDS, National Institutes of Health, Bldg 10, Room 5B-16, 10 Center DR MSC 1400, Bethesda, MD 20892-1400, USA E-mail: martinr{at}ninds.nih.gov
Received August 7, 2002. Revised November 24, 2002. Second revision January 28, 2003 Accepted February 3, 2003.
 |
Summary |
|---|
 Top Summary IntroductionPatients and methodsResultsDiscussionReferences |
|---|
Autoimmune diseases such as multiple sclerosis are characterized by complex genetic traits and pathomechanisms that translate into clinical heterogeneity. This wide heterogeneity of multiple sclerosis as well as different biological responses to immunomodulatory drugs can be expected to contribute to differential treatment responses. Strategies that dissect the relationship between the treatment response and the biological characteristics in individual patients are valuable not only as a clinical tool, but also in leading to a better understanding of the disease. Here we address the in vitro and ex vivo RNA expression profile under one approved therapy of multiple sclerosis, interferon-ß (IFN-ß, Betaseron), by cDNA microarrays and demonstrate that non-responder and responder phenotypes to IFN-ß as assessed by longitudinal gadolinium-enhanced MRI scans and clinical disease activity differ in their ex vivo gene expression profile. These findings will help to better elucidate the mechanism of action of IFN-ß in relation to different disease patterns and eventually lead to optimized therapy.
Keywords: cDNA microarrays; expression profile; interferon-ß; multiple sclerosis
Abbreviations: Gd = gadolinium; IFN = interferon; IL-8 = interleukin-8; INR = initial non-responder; NAb = neutralizing antibodies; NAbNR = non-responder due to high titres of neutralizing antibodies; PBMC = peripheral blood mononuclear cells; RR = relapsing–remitting
|
Introduction |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
Multiple sclerosis is considered a T cell-mediated autoimmune disease that is characterized by CNS inflammation, demyelination and various degrees of axonal damage (McFarlin and McFarland, 1982
; Trapp et al., 1998). Similar to diabetes and rheumatoid arthritis, genetic studies have demonstrated that major histocompatibility complex [MHC; human leucocyte antigen (HLA) in humans] class II alleles, as well as a considerable number of other as yet unidentified susceptibility genes, contribute to the quantitative genetic trait that predisposes to disease (Becker et al., 1998). Individual patients rarely carry all susceptibility genes. The genetic heterogeneity and possibly different environmental triggers, e.g. viral infections (Santoro et al., 1999), lead in turn to variability in disease patterns defined either clinically or pathologically (Lucchinetti et al., 2000). The same heterogeneity probably also contributes to differences in response to treatment.
There are currently three categories of approved therapies for the long-term therapy of relapsing–remitting (RR) multiple sclerosis, three different preparations of interferon (IFN)-ß [IFN-ß-1a (AvonexTM), IFN-ß-1b (BetaseronTM) and IFN-ß-1a (RebifTM)], glatiramer–acetate (CopaxoneTM) and mitoxantrone (NovantroneTM) [IFNB Multiple Sclerosis Study Group and the The University of British Columbia MS/MRI Analysis Group, 1995; Johnson et al., 1995; Jacobs et al., 1996; PRISMS (Prevention of Relapses and Disability by Interferon beta-1a Subcutaneously in Multiple Sclerosis) Study Group, 1998; Bielekova and Martin, 1999]. Type I IFNs, including IFN-ß, are not only part of the innate immune system and exert antiviral activities, but also cause complex immunomodulatory effects (Arnason, 1996; Wandinger et al., 2001). While their mechanism of action as an immunomodulator in the treatment of multiple sclerosis is not fully understood, speculation centres on modulation of adhesion molecule expression, inhibition of matrix metalloproteinases, regulation of apoptosis and induction of anti-inflammatory cytokines such as interleukin (IL)-10 (Stone et al., 1995; Rudick et al., 1996; Calabresi et al., 1997; Yong et al., 1998; Zipp et al., 1998; Deisenhammer et al., 2000; Wang et al., 2000). IFN-ß also upregulates a number of pro-inflammatory mediators (Wandinger et al., 2001), an observation that can not easily be reconciled with their proposed mechanism in multiple sclerosis. Previous studies have attempted to characterize the disease mechanism of multiple sclerosis and the impact of IFN-ß by measuring one or more candidate cytokines and immune markers (Gayo et al., 1999; van Boxel-Dezaire et al., 2000; Comabella et al., 2002). Because of the heterogeneity of the disease and the complex interaction of treatment and disease, it is not surprising that these studies did not provide a conclusive picture of the mechanism of action of IFN-ß. Recent advances in the development of genomics and proteomics offer opportunities to replace hypothesis-based biomarker studies by large-scale analyses that aim at capturing complex gene and protein expression responses (Alizadeh et al., 2000; Staudt and Brown, 2000; Martin et al., 2001). Studies on expression profiles employing microarray techniques have been performed (Der et al., 1998; Da Silva et al., 2002), which contributed to the notion that a more complex pattern of gene regulation is involved in the pharmacological and physiological effects of IFNs, and IFN-ß in particular. These studies were, however, limited to in vitro experiments using cell types different from the presumed target cells of IFN in multiple sclerosis [a human fibrosarcoma cell line and human umbilical vein endothelial cells (HUVEC), respectively]. Here we studied gene expression profiles in response to IFN-ß therapy in multiple sclerosis patients. We used the in vitro pretreatment gene expression profile to IFN-ß as a parameter for the baseline responsiveness and compared it with the biological response in vivo (arrays assessing the in vivo effects of IFN-ß are referred to as ex vivo experiments). The in vitro gene expression profile also served as a tool to identify genes of interest that may not be regulated strongly enough ex vivo to meet the threshold criteria of significant regulation. The main focus of our study was to identify the ex vivo gene expression pattern in six treatment responders as assessed by MRI (Stone et al., 1997), a sensitive and well-established marker for multiple sclerosis disease activity, along with clinical disease activity. As a preliminary attempt to identify differences in gene expression between responding patients and those not optimally responding to therapy we included two patients who initially responded to IFN-ß but lost their response after developing high titres of neutralizing antibodies (NAbNR) and two patients who failed to respond fully to therapy from the beginning (initial non-responders, INRs).
|
Patients and methods |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
All multiple sclerosis patients included in the current study participated in a longitudinal study with IFN-ß-1b (Betaseron, Berlex, NJ, USA) as described previously (Stone et al., 1995
; Wandinger et al., 2001) and were followed up by the National Institute of Neurological Disorders and Stroke/Neuroimmunology Branch (NINDS)/NIB outpatient clinic. They had clinically definite, RR multiple sclerosis and were monitored by monthly clinical examinations and brain MRI (1.5-T MR unit; General Electric, Milwaukee, WI, USA). New and total gadolinium (Gd)-DTPA-enhancing lesions (Magnevist; Berlex Laboratories, Cedar Knolls, NJ, USA) were assessed. The study was reviewed and approved by the NINDS Institutional Review Board, and all patients gave written informed consent. Ten female multiple sclerosis patients were identified as fulfilling the predetermined criteria for selecting their samples for cDNA array experiments. The criteria included: (i) clinically definite multiple sclerosis; (ii) monitoring before and after start of IFN therapy by monthly MRI and clinical evaluations; (iii) no immunomodulatory therapy other than IFN for 30 days prior to obtaining blood sample; (iv) no MRI activity above average baseline activity at the baseline sampling time point; (v) no acute deterioration or relapse 30 days prior to sampling time point; (vi) no concurrent infection; (vii) 6 months or more on therapy (the NAbNR samples before developing NAb were collected at earlier time points); and (viii) sufficient peripheral blood mononuclear cell (PBMC) sample available during baseline and treatment to perform cDNA array experiments. Patients were categorized based on clinical and MRI findings as: (i) treatment responders with clinically stable multiple sclerosis and a > 60% reduction in mean number of total Gd-enhancing lesions compared with baseline (six patients); (ii) loss of MRI response related to development of high titres of neutralizing antibodies (NAb) to IFN-ß (NAbNR, two patients); or (iii) failure to respond optimally from initiation of therapy as assessed by clinical and MRI measures (INR, two patients).
The indication to treat with IFN was based on MRI activity mainly in patients 1, 3, 4, 6 and NAbNR 2, and on both relapse and MRI activity in patients 2, 5, NAbNR 1 and INR 1 and 2.
PBMCs from these patients were collected as described previously (Wandinger et al., 2001) at regular intervals before and during treatment with IFN-ß, and stored in liquid nitrogen. Samples were taken within 48 h of an IFN-ß injection. For treatment responders, at least two time points were compared for the cDNA microarray experiment, i.e. baseline and treatment. For patients developing NAb, three time points were compared: baseline, treatment before NAb (or at low NAb) and treatment at time points of high NAb titres. For three of the responders (1, 2 and 3) and for the two NAb patients (treatment time point before NAb) two arrays were performed for different treatment time points that were evaluated as pairs to identify changes in gene regulation. Altogether, 10 baseline cDNA arrays were performed, eight arrays testing the in vitro response to IFN-ß, nine for testing the ex vivo profile in responders, and eight arrays testing the ex vivo response in NAbNR and INR.
Microarrays and PCR
Cells stored in cryo-medium (20% dimethyl sulphoxide, 20% human serum albumin, HSA, 60% T-cell medium) were thawed and viability determined by Trypan Blue exclusion (to be in the range of >80%). After in vitro incubation and for the ex vivo experiments the cells were spun down immediately after thawing, washed in T-cell medium again, and transferred to RNeasy lysis buffer (RNeasy Kit; Qiagen, Santa Clarita, CA, USA) for further processing. For the in vitro cDNA microarray testing, 108 PBMCs from baseline time points were incubated for 24 h in 25 ml of T-cell medium (Wandinger et al., 2001). IFN-ß (100 IU/ml) was added for testing of the in vitro effects at baseline. For the ex vivo testing of in vivo effects, 2 x 106 and 108 PBMCs were processed for PCR and microarray experiments, respectively. Total RNA was prepared by the RNA MidikitTM method (Qiagen) with yields of 40–90 µg. The same amount of RNA was used on arrays to compare conditions within a patient and a reference RNA pool (prepared from stimulated and un-stimulated PBMC) was used on each array to allow comparisons of relative up- and downregulation of genes across experiments. The in vitro untreated baseline experiment was used as the comparator for both the in vitro incubation with IFN-ß and the in vivo effect.
Mini-Lymphochip (MLC) cDNA microarrays were prepared from PCR-amplified material (Staudt and Brown, 2000). cDNA probes were prepared, and microarray analysis of gene expression was essentially performed as described in Staudt and Brown (2000) and Alizadeh et al. (2000), using MLC versions that contained either 6432 or 12 672 array elements. Microarrays were analysed on a GenePix scanner (Axon Instruments, Inverurie, Scotland), and data files were entered into a custom database maintained at the National Institutes of Health (http://nciarray.nci.nih.gov). We extracted data for clustering analysis (programs ‘Cluster’ and ‘TreeView’ by M. Eisen, Stanford, CA, USA) according to the following requirements: spot size of at least 25 µm, minimum intensities of 100 relative fluorescent units (RFU) in the Cy3 and Cy5 channels or minimum intensity of at least 1000 RFU in one of the two channels; spots labeled as ‘not found’ or ‘bad’ by the software were excluded. Relative levels of gene expression were calculated as the ratio of normalized intensities of the Cy5 and the Cy3 signal. Gene expression was ‘normalized’ by calculating the difference between expression at baseline and expression after in vitro incubation or in vivo treatment and displayed as a factor for up- and downregulation and as a colour-coded matrix of genes. A 2-fold increase or decrease compared with the baseline condition was accepted as clear indication of up- or downregulation of a gene. Only annotated and sequence-confirmed genes were searched and analysed by the following algorithms.
Ex vivo and in vitro regulation of a given gene was considered relevant if at least three ex vivo (responders) or in vitro (responders and NAb patients before developing NAb) cDNA arrays had a 2-fold change in expression, respectively, and if none of the responders showed a significant regulation in the opposite direction (separately assessed for in vitro and ex vivo findings). The latter rule was applied to control for the number of chance findings; it might, however, lead to the exclusion of some true variability within the expression profiles of IFN-ß responders. If more than one out of the four arrays for non-responder status (INR and NAbNR) or one out of less than four arrays showed a significant regulation in the same direction as the responders, the gene was considered regulated also in non-responder status.
Real-time RT–PCR was employed according to a previously described method (Wandinger et al., 2001) to confirm the findings for IL-8, a key cytokine detected through cDNA arrays as being markedly downregulated ex vivo and in vitro by IFN treatment. cDNA was taken from the same sample that had been used for the cDNA array (for three responders, one INR and the two NAbNR), and additional time points at baseline and under treatment as well as additional patients were tested to allow for follow-up over several treatment time points for which not enough material was available to perform array experiments.
Statistics
To explore the ex vivo expression pattern a score sum counting genes regulated by a factor of 
2 as ‘1’ and the absence of regulation as ‘0’ was used, and two-sided t-test statistics were applied to compare responder and non-responder status. Imputation was performed for missing values using a ‘worst case’ approach in replacing missing expression values by ‘1’ for non-responders and by ‘0’ for responders. To analyse the RT–PCR findings for IL-8, mean values per patient over time were used and a two-sided t-test was performed.
|
Results |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
MRI findings and relapse activity
The responder status of the patients was defined by a reduction of at least 60% in the number of total Gd-enhancing lesions versus baseline (Stone et al., 1997
) and the marked reduction or absence of clinical disease activity (Fig. 1A–D). Accordingly, responders showed a marked reduction in the total number of Gd-enhancing lesions. Patients who developed NAb had a reduction in MRI activity preceding the development of NAb followed by reoccurrence of disease activity after developing NAb (titres of 549 and 477, respectively) (Fig. 1D). With the two INRs, both failed to demonstrate a (optimal) response with regard to the MRI criteria (Fig. 1C) from the onset of therapy. In the second INR, frequent relapses occurred in addition to fluctuating MRI activity (five relapses in the first 12 months of treatment). Two subgroups could be identified among the cohort of responders with regard to baseline MRI activity, i.e. patients with high numbers of Gd-enhancing lesions (Fig. 1A) and patients with markedly lower activity (Fig. 1B). More profound changes in gene expression appeared in those patients with high disease activity at baseline although no formal stratification was attempted due to the small number of individuals.
|
cDNA array experiments
In vitro and ex vivo cDNA array data were obtained from eight of 10 patients (five of six responders, two of two NAb patients, and one of two INRs); in two patients, one responder and one INR, only ex vivo treatment samples could be examined due to the lack of sufficient cell material to perform the in vitro IFN incubation experiment. We identified 112 genes with a significant regulation by IFN-ß; 25 genes were found significantly regulated ex vivo and an additional 87 genes in vitro. One likely explanation for the differences between the findings in ex vivo and in vitro experiments is the dose-dependency of gene expression changes; the concentration chosen for the in vitro incubation dose was
5-fold that of peak plasma levels after IFN-ß administration in vivo. It must be taken into account that the 24 h in vitro incubation and the ex vivo expression profile will not reflect the full range of regulated genes, since a number of genes regulated early on after IFN-ß have been added/administered will no longer show a transcriptional response. Twenty of 25 (80%) of the ex vivo-regulated genes also showed a significant effect upon in vitro incubation with IFN-ß. Eighty-eight per cent of the genes regulated ex vivo in responders were not regulated during the non-responder status of the NAbNR patients or in the INRs. For 14 of 25 genes none of the non-responders showed a regulation, and for another eight genes only one out of four non-responders showed regulation.
t-tests were performed to evaluate statistically the regulation pattern of all 25 genes together, and downregulated and upregulated genes separately. P values of 0.025 for all genes and for upregulated genes support a statistical significant difference between responders and non-responders; a P value of 0.083 for downregulated genes alone is indicative of a trend. Given the semi-quantitative nature of array findings, the results are presented as direction of regulation in Tables 1 and 2. Genes regulated ex vivo are depicted in Fig. 2.
|
|
|
With regard to the NAbNR patients there are two findings of particular importance: only four genes that were significantly regulated in responders ex vivo were found in both NAbNR patients before developing antibodies, and 10 of 25 genes were not regulated in any of the NAbNR during this time period. On the other hand, the in vitro pattern of regulation does not differ from the findings in responders. Given the small number of patients and the finding that not all genes are regulated significantly in the six responders either (see Table 1), these ex vivo findings should not be misinterpreted as being predictive of later NAb development.
Out of the 87 genes that were significantly regulated in vitro, 59% were found in five or more out of the seven patients tested, thus showing a highly consistent expression profile. We decided to include the in vitro data from the NAbNR patients since their failure to respond was secondary to the development of NAb rather than a primary non-responsiveness. The in vitro response at baseline of eventually NAb-positive patients and the one INR for whom in vitro data are available do not substantially differ from the responder patients (Fig. 3).
|
Detailed analysis of gene expression changes upon IFN-ß treatment
Ex vivo findings
The cytokines IL-8 and flt-3 ligand (fms-like tyrosine kinase 3 gene) (co-stimulatory cytokine for haematopoetic progenitors) were found significantly downregulated ex vivo and in vitro. For these two genes no response was found for the INR and in one of the NAbNR patients. The marked reduction of IL-8 in responders ex vivo by a factor of 1/3 to 1/20 and the lack of an inhibitory effect in non-responders is a novel observation in multiple sclerosis. IL-8 is one of the important chemotactic mediators recruiting neutrophils to sites of inflammation, and IFN-ß has been shown to inhibit IL-8 expression in vitro via an nuclear factor (NF)-
B binding site (Oliveira et al., 1994). IL-8 has been described not to differ in its expression in unstimulated PBMC from multiple sclerosis patients and normal volunteers (Jalonen et al., 2002; Comabella et al., 2002). However, the increased expression of IL-8 in CD14+ cells after stimulation has been described to be inhibited by IFN-ß ex vivo and in vitro in multiple sclerosis patients (Comabella et al., 2002). The ex vivo effect of a 50% reduction of the proportion of IL-8 expressing cells as described by Comabella et al. is clearly less marked than observed in this study; however, it cannot be directly compared to our findings in unstimulated PBMC.
A number of genes that are involved in the regulation of proliferation were downregulated by IFN treatment ex vivo and in vitro. The most prominent effect ex vivo was found for GKLF4 (gut-enriched Kruppel-like zinc finger protein). c-fos, c-jun and flt-3 (see above under cytokines) were downregulated, indicating an anti-proliferative effect of treatment. The downregulation of GKLF AREB6 (transcription factor, zinc finger domain protein), Id2 inhibitor of DNA binding, MKP-1 (MAP kinase phosphatase 1), and of IB-alpha may have opposing effects. One gene, PCNA (proliferating cell nuclear antigen), was downregulated ex vivo but showed an opposite effect after in vitro incubation. For this group of genes the lack of regulation in the NAbNR and the INR is evident with only one of the genes showing a change in expression in one of four non-responders.
Five apoptosis-related genes were regulated ex vivo and four of five [BNIP 3, TRAIL, IEX-1L (immediate early gene, apoptosis inhibitor), TSC22-R (transforming growth factor beta stimulated clone 22 related gene)] favour a pro-apoptotic state. In the INR and the NAbNR, only one of five genes showed regulation ex vivo. CD69, as an unspecific marker of T-cell activation, was downregulated ex vivo and in vitro.
The ex vivo upregulation of expression of five IFN marker genes (2'5' OAS, guanylate binding protein 1, STAT1, IFN-induced 17 kDa protein, IFN-induced protein kinase) in responders serves as a positive control for the expected biological effects. Only one of these genes was found regulated in non-responders.
In vitro findings
Table 2 summarizes data for gene products that were only regulated in vitro.
The upregulation of pro-inflammatory chemokines MIP-1ß, IP-10, MCP-1 and six other genes related to chemokine and cytokine pathways indicates a pro-inflammatory shift of monocyte- and lymphocyte-derived mediators, respectively. The most prominent changes are found for IP-10 (in line with our previous data) with 5.6- to 174-fold and MCP-1 with 12.6- to 160-fold increases in expression. The in vitro regulation of 13 further gene products that are involved in cell-cycle control supports the notion of an overall anti-proliferative effect of IFN-ß.The findings for IP-10 and MCP-1 should be related to the findings of Comabella et al. (2002), who found an inhibitory effect of IFN-ß on ex vivo and in vitro stimulated PBMCs. The differences may be explained by the additional stimulation used by Comabella et al. (anti-CD3 plus anti-CD28).
The in vitro findings indicate a shift toward a pro-apoptotic state of gene expression.
Gene products involved in antigen presentation (four genes) were only found to be regulated (upregulation) in vitro. Two markers of macrophage activation were downregulated. The influence on adhesion molecule expression (CD11c, CD18, CD31, CD50 and CD54) showed one gene product being upregulated and four downregulated. The pattern of regulation for cathepsin B, CD13, CD38, CD63, CD59 and CD72 is interpreted as representing a ‘balanced’ change with regard to either stimulation or inhibition of immune responses, although such conclusions are difficult to draw from array data only. The co-stimulatory molecule CD40 was upregulated, favouring a pro-inflammatory state.
The in vitro upregulation of 12 genes known as IFN marker genes (see Table 2; IFN-induced marker genes) serves as a positive control of the expected biological response to IFN.
Previous findings (Wandinger et al., 2001) regarding CCR-5 and IL-12 receptor ß2 chain were only partly confirmed here, (i) because the gene for IL-12 receptor ß2 chain was not represented on the array used and (ii) only two of the in vitro experiments showed a significant signal for CCR-5, while ex vivo arrays showed no measurable signal.
Confirmation by PCR
The marked changes in IL-8 expression were followed by quantitative real-time RT–PCR studies in four responders, two INR and three NAbNR (including three responders, the INRs and the two NAbNRs from the array series; Fig. 4) for time points under treatment from 2 to 22 months. Whereas all responders showed a consistent suppression of gene expression for IL-8 over time, an increased expression was found in the INRs and no effect in the NAbNRs once high titres of NAb occurred. Exploratory statistics based on the mean percentage change during treatment versus baseline as the measure of effect, showed a significant difference between responders and the group of INR and NAbNR (P = 0.035).
|
|
Discussion |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
Current understanding of the pathogenesis of multiple sclerosis indicates that it represents a heterogeneous group of disorders based on clinical, MRI, pathological, immunological and genetic findings (Lucchinetti et al., 2000
; Stuerzebecher and Martin, 2000). Available treatments such as IFN-ß exert a multitude of effects on various tissues and cell types, and it can be expected that not all of these are beneficial for the disease process in multiple sclerosis if considered as single markers. The lack of a uniform treatment response in the entire population of multiple sclerosis patients, even if only one stage, i.e. RR multiple sclerosis, is considered suggests that not all patients respond similarly at the molecular level to the dose levels of IFN-ß achievable in vivo, and that the net outcome of these effects will determine the clinical results and the extent to which inflammation in the CNS is inhibited. With respect to the mechanism of action of IFN-ß, other laboratories’ and our recent data demonstrate that it induces a number of anti-inflammatory, but also pro-inflammatory, activities (Wandinger et al., 2001), and it is likely that it is the balance of many moderate or even small changes in different biological pathways that leads to clinical benefit.
On the other hand, a single cytokine can have dual or multiple effects, e.g. depending on the level of expression, the activation status of the immune system or the site of action. This could apply to our findings of an increase of pro-inflammatory cytokines (in vitro) that may, if relevant in vivo, exhaust the cellular response in the periphery and thereby have a protective effect on the inflammatory response in the CNS. Tumour necrosis factor (TNF)-
is another important example (Sharief and Hentges, 1991). A wealth of evidence indicated that TNF- is involved in the pathogenetic process of multiple sclerosis and that its inhibition would benefit patients. Contrary to these expectations and different from the experiences in the treatment of rheumatoid arthritis, blocking TNF- has, however, worsened disease in multiple sclerosis patients (Lenercept Multiple Sclerosis Study Group and The University of British Columbia MS/MRI Analysis Group, 1999).
The rapid progress in the area of genomics, particularly the study of the expression of thousands of genes by cDNA or oligonucleotide microarrays, now enables us to examine these questions at a different level than previously. Rather than looking at single markers or cytokines in a hypothesis-driven way and based on previous knowledge, we can now analyse responses of very complex systems and start to assess the response of cells and tissues in their interactions. In the present study, we therefore used gene expression profiling to capture not only the complex immunomodulatory activity of IFN-ß in individual multiple sclerosis patients, but furthermore to relate the biological response of a number of informative genes to the therapeutic response to IFN-ß treatment. These experiments are a first attempt to determine whether individual responder and non-responder profiles or sets of marker genes can be identified in IFN-ß-treated multiple sclerosis patients. Based on a well accepted disease activity marker that is useful for visualizing the response to IFN-ß (Stone et al., 1995, 1997), the total number of Gd-enhancing MRI lesions, and the clinical disease activity, we classified patients as: (i) IFN-ß responders (>60% reduction of MRI disease activity); (ii) patients who initially responded, but later developed NAb and then lost IFN-ß responsiveness; and (iii) two patients who never fully responded to IFN-ß (INR). The ex vivo response of patients of group (ii) and (iii) was meant to provide a ‘negative control’; there was a clear trend for ex vivo gene regulation that distinguishes this group of non-responding patients from the responders with most of the genes of interest being not regulated. The in vitro response to treatment in patients later developing NAbs documents that these patients are capable of responding to IFN-ß and do not differ at baseline from the responders. The same appears to be correct in INRs. However, the in vitro expression profile from only one of the INRs, which does not reveal a different pattern from those of the responders, does not exclude with certainty that differences even at baseline could be predictive of later non-response, and further studies are needed to exclude a pharmacogenetic variance (changes in primary IFN receptor interaction, faster downregulation of receptor) as a potential explanation.
The following conclusions can be drawn from our study. First, in agreement with the literature on the subject and our previous, limited observations, the gene expression profile in response to IFN-ß in vitro and in vivo is complex and includes, influences on cell migration, matrix degradation, proliferation, cell cycle control, cell differentiation, antigen processing and presentation, apoptosis, and cytokine and chemokine regulation. Secondly, even though the ex vivo response pattern reflects only part of the in vitro effects of IFN-ß on gene expression, 25 out of the 112 genes that are modified are regulated by at least a factor of two ex vivo. In the future, the evaluation of the ex vivo response assessed after a short treatment interval may enable us to allow prediction of long-term treatment response in the context of, for example, frequent MRI examination early during treatment. If confirmed in larger studies the ability to predict treatment response to a particular treatment early would be extremely helpful in patient management. Finally, our data with IL-8 represent an example of how single genes potentially related to treatment response can be identified using the cDNA array approach. However, even though the results on IL-8 appear rather clear with regard to potentially distinguishing responders from non-responders, we believe that the definition of the responder state should not rely on individual genes, but rather on groups of genes, given the variability of single gene expression versus the desired predictive value of a method being used to identify responders early during treatment. Future studies should examine this question in larger cohorts of patients. We expect that arrays capturing the expression of large numbers of genes, enabling the examination of all relevant biological responses to a treatment, will lead to more individualized therapies, e.g. treatment at higher doses of IFN-ß-1b, with different drugs, or with combination therapies similar to the strategies that have been tested in cancer.
|
Acknowledgements |
|---|
We wish to thank Joan Ohayon, Jennifer McCartin, Helen Griffith and Roger Stone (NIB, NINDS, NIH) for their assistance in patient care and data retrieval, and Drs C. Bash and N. Richert, Laboratory for Diagnostic Radiology Research, Clincal Center, NIH, for reading MRI films. S.St. was a National Institutes of Health postdoctoral fellow. K.P.W. was a postdoctoral fellow of the Deutsche Forschungsgemeinschaft (Wa 1343/1-1). A.R. was supported by a grant of the Mildred Scheel Stiftung.
|
References |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
Alizadeh AA, Eisen MB, Davis RE, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 2000; 403: 503–11.[CrossRef][Medline]
Arnason BG. Interferon beta in multiple sclerosis. Clin Immunol Immunopathol 1996; 81: 1–11.[CrossRef][ISI][Medline]
Becker KG, Simon RM, Bailey-Wilson JE, et al. Clustering of non-major histocompatibility complex susceptibility candidate loci in human autoimmune diseases. Proc Natl Acad Sci USA 1998; 95: 9979–84.
Bielekova B, Martin R. Multiple sclerosis: immunotherapy. Curr Treat Options Neurol 1999; 1: 201–20.[Medline]
Calabresi PA, Tranquill LR, Dambrosia JM, et al. Increases in soluble VCAM-1 correlate with a decrease in MRI lesions in multiple sclerosis treated with interferon beta-1b. Ann Neurol 1997; 41: 669–74.[CrossRef][ISI][Medline]
Comabella M, Imitola J, Weiner HL, et al. Interferon-beta treatment alters peripheral blood monocytes chemokine production in MS patients. J Neuroimmunol 2002; 126: 205–12.[CrossRef][ISI][Medline]
DaSilva A, Brickelmaier M, Majeau GR, et al. Comparison of gene expression patterns induced by treatment of human umbilical vein endothelial cells with IFN-2b vs. IFN-ß1a: understanding the functional relationship between distinct type I interferons that act through a common receptor. J Interferon Cytokine Res 2002; 22: 173–88.[CrossRef][ISI][Medline]
Deisenhammer F, Mayringer I, Harvey J, et al. A comparative study of the relative bioavailability of different interferon beta preparations. Neurology 2000; 54: 2055–60.
Der S, Zhou A, Williams BRG, et al. Identification of genes differentially regulated by interferon , ß or 
using oligonucleotide arrays. Proc Natl Acad Sci USA 1998; 95: 15623–8.
Gayo A, Mozo L, Suarez A, et al. Interferon beta-1b treatment modulates TNFalpha and IFNgamma spontaneous gene expression in MS. Neurology 1999; 52: 1764–70.
11 IFNB Multiple Sclerosis Study Group and The University of British Columbia MS/MRI Analysis Group. Interferon beta-1b in the treatment of multiple sclerosis: final outcome of the randomized controlled trial. Neurology 1995; 45: 1277–85.[Abstract]
Jacobs LD, Cookfair DL, Rudick RA, et al. Intramuscular interferon beta-1a for disease progression in relapsing multiple sclerosis. The Multiple Sclerosis Collaborative Research Group (MSCRG). Ann Neurol 1996; 39: 285–94.[CrossRef][ISI][Medline]
Jalonen TO, Pulkkinen K, Ukkonen M, et al. Differential intracellular expression of CCR5 and chemokines in multiple sclerosis subtypes. J Neurol 2002; 249: 576–83.[CrossRef][ISI][Medline]
Johnson KP, Brooks BR, Cohen JA, et al. Copolymer 1 reduces relapse rate and improves disability in relapsing-remitting multiple sclerosis: results of a phase III multicenter, double-blind placebo-controlled trial. The Copolymer 1 Multiple Sclerosis Study Group. Neurology 1995; 45: 1268–76.[Abstract]
15 Lenercept Multiple Sclerosis Study Group and The University of British Columbia MS/MRI Analysis Group. TNF neutralization in MS: results of a randomized, placebo-controlled multicenter study. Neurology 1999; 53: 457–65.
Lucchinetti C, Bruck W, Parisi J, et al. Heterogeneity of multiple sclerosis lesions: implications for the pathogenesis of demyelination. Ann Neurol 2000; 47: 707–17.[CrossRef][ISI][Medline]
Martin R, Stürzebecher CS, McFarland HF. Immunotherapy of multiple sclerosis: where are we? Where should we go? Nat Immunol 2001; 2: 785–8.[CrossRef][ISI][Medline]
McFarlin DE, McFarland HF. Multiple sclerosis (second of two parts). New Engl J Med 1982; 307: 1246–51.[ISI][Medline]
Oliveira IC, Mukaida N, Matsushima K, et al. Transcriptional inhibition of the interleukin-8 gene by interferon is mediated by the NF-B site. Mol Cell Biol 1994; 14: 5300–8.
20 PRISMS (Prevention of Relapses and Disability by Interferon beta-1a Subcutaneously in Multiple Sclerosis) Study Group. Randomised double-blind placebo-controlled study of interferon beta-1a in relapsing/remitting multiple sclerosis. Lancet 1998; 352: 1498–504.[CrossRef][ISI][Medline]
Rudick RA, Ransohoff RM, Peppler R, et al. Interferon beta induces interleukin-10 expression: relevance to multiple sclerosis. Ann Neurol 1996; 40: 618–27.[CrossRef][ISI][Medline]
Santoro F, Kennedy PE, Locatelli G, et al. CD46 is a cellular receptor for human herpesvirus 6. Cell 1999; 99: 817–27.[CrossRef][ISI][Medline]
Sharief MK, Hentges R. Association between tumor necrosis factor-alpha and disease progression in patients with multiple sclerosis. New Engl J Med 1991; 325: 467–72.[Abstract]
Staudt LM, Brown PO. Genomic views of the immune system. Annu Rev Immunol 2000; 18: 829–59.[CrossRef][ISI][Medline]
Stone LA, Frank JA, Albert PS, et al. The effect of interferon-beta on blood-brain barrier disruptions demonstrated by contrast-enhanced magnetic resonance imaging in relapsing–remitting multiple sclerosis. Ann Neurol 1995; 37: 611–9.[CrossRef][ISI][Medline]
Stone LA, Frank JA, Albert PS, et al. Characterization of MRI response to treatment with interferon beta-1b: contrast-enhancing MRI lesion frequency as a primary outcome measure. Neurology 1997; 49: 862–9.
Stürzebecher S, Martin R. Neuroimmunology of multiple sclerosis and experimental allergic encephalomyelitis. Neuroimaging Clin N Am 2000; 10: 649–68.[ISI][Medline]
Trapp BD, Peterson J, Ransohoff RM, et al. Axonal transection in the lesions of multiple sclerosis. New Engl J Med 1998; 338: 278–85.
vanBoxel-Dezaire AH, van Trigt-Hoff SC, Killestein J, et al. Contrasting responses to interferon beta-1b treatment in relapsing-remitting multiple sclerosis: does baseline interleukin-12p35 messenger RNA predict the efficacy of treatment? Ann Neurol 2000; 48: 313–22.[CrossRef][ISI][Medline]
Wandinger KP, Stürzebecher CS, Bielekova B, et al. Complex immunomodulatory effects of interferon-beta in multiple sclerosis include the upregulation of T helper 1-associated marker genes. Ann Neurol 2001; 50: 349–57.[CrossRef][ISI][Medline]
Wang X, Chen M, Wandinger KP, et al. IFN-beta-1b inhibits IL-12 production in peripheral blood mononuclear cells in an IL-10-dependent mechanism: relevance to IFN-beta-1b therapeutic effects in multiple sclerosis. J Immunol 2000; 165: 548–57.
Yong VW, Chabot S, Stuve O, et al. Interferon beta in the treatment of multiple sclerosis: mechanisms of action. Neurology 1998; 51: 682–9.
Zipp F, Weller M, Calabresi PA, et al. Increased serum levels of soluble CD95 (APO-1/Fas) in relapsing–remitting multiple sclerosis. Ann Neurol 1998; 43: 116–20.[CrossRef][ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  F Sellebjerg, P Datta, J Larsen, K Rieneck, I Alsing, A Oturai, A Svejgaard, P Soelberg Sorensen, and L. Ryder Gene expression analysis of interferon-{beta} treatment in multiple sclerosis Multiple Sclerosis, June 1, 2008; 14(5): 615 - 621. [Abstract] [PDF]
|
|
|
|
|
|
E. Wiesemann, M. Deb, C. Trebst, B. Hemmer, M. Stangel, and A. Windhagen Effects of interferon-{beta} on co-signaling molecules: upregulation of CD40, CD86 and PD-L2 on monocytes in relation to clinical response to interferon-{beta} treatment in patients with multiple sclerosis Multiple Sclerosis, March 1, 2008; 14(2): 166 - 176. [Abstract] [PDF]
|
|
|
|
 |
|
 A. E. Baird Blood Genomics in Human Stroke Stroke, February 1, 2007; 38(2): 694 - 698. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. R. Sharp, H. Xu, L. Lit, W. Walker, M. Apperson, D. L. Gilbert, T. A. Glauser, B. Wong, A. Hershey, D.-Z. Liu, et al. The Future of Genomic Profiling of Neurological Diseases Using Blood Arch Neurol, November 1, 2006; 63(11): 1529 - 1536. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F Gilli, F Marnetto, M Caldano, A Sala, S Malucchi, M Capobianco, and A Bertolotto Biological markers of interferon-beta therapy: comparison among interferon-stimulated genes MxA, TRAIL and XAF-1 Multiple Sclerosis, February 1, 2006; 12(1): 47 - 57. [Abstract] [PDF]
|
|
|
|
 |
|
 T. Chtanova, R. Newton, S. M. Liu, L. Weininger, T. R. Young, D. G. Silva, F. Bertoni, A. Rinaldi, S. Chappaz, F. Sallusto, et al. Identification of T Cell-Restricted Genes, and Signatures for Different T Cell Responses, Using a Comprehensive Collection of Microarray Datasets J. Immunol., December 15, 2005; 175(12): 7837 - 7847. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F Sellebjerg, C Ross, N Koch-Henriksen, P S. Sorensen, J L Frederiksen, K Bendtzen, and T L Sorensen CD26+CD4+T cell counts and attack risk in interferon-treated multiple sclerosis Multiple Sclerosis, December 1, 2005; 11(6): 641 - 645. [Abstract] [PDF]
|
|
|
|
 |
|
 T. Ishizu, M. Osoegawa, F.-J. Mei, H. Kikuchi, M. Tanaka, Y. Takakura, M. Minohara, H. Murai, F. Mihara, T. Taniwaki, et al. Intrathecal activation of the IL-17/IL-8 axis in opticospinal multiple sclerosis Brain, May 1, 2005; 128(5): 988 - 1002. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K Morgen, A L. Crawford, R D Stone, R Martin, N D Richert, J A Frank, and H F McFarland Contrast-enhanced MRI lesions during treatment with interferon{beta}-1b predict increase in T1 black hole volume in patients with relapsing-remitting multiple sclerosis Multiple Sclerosis, April 1, 2005; 11(2): 146 - 148. [Abstract] [PDF]
|
|
|
|
 |
|
 D. F. Moore, H. Li, N. Jeffries, V. Wright, R. A. Cooper Jr, A. Elkahloun, M. P. Gelderman, E. Zudaire, G. Blevins, H. Yu, et al. Using Peripheral Blood Mononuclear Cells to Determine a Gene Expression Profile of Acute Ischemic Stroke: A Pilot Investigation Circulation, January 18, 2005; 111(2): 212 - 221. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Bielekova and R. Martin Development of biomarkers in multiple sclerosis Brain, July 1, 2004; 127(7): 1463 - 1478. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||