Effect of Natalizumab on Circulating CD4+ T-Cells in Multiple Sclerosis

In multiple sclerosis (MS), treatment with the monoclonal antibody natalizumab effectively reduces the formation of acute lesions in the central nervous system (CNS). Natalizumab binds the integrin very late antigen (VLA)-4, expressed on the surface of immune cells, and inhibits VLA-4 dependent transmigration of circulating immune-cells across the vascular endothelium into the CNS. Recent studies suggested that natalizumab treated MS patients have an increased T-cell pool in the blood compartment which may be selectively enriched in activated T-cells. Proposed causes are sequestration of activated T-cells due to reduced extravasation of activated and pro-inflammatory T-cells or due to induction of VLA-4 mediated co-stimulatory signals by natalizumab. In this study we examined how natalizumab treatment altered the distribution of effector and memory T-cell subsets in the blood compartment and if T-cells in general or myelin-reactive T-cells in particular showed signs of increased immune activation. Furthermore we examined the effects of natalizumab on CD4+ T-cell responses to myelin in vitro. Natalizumab-treated MS patients had significantly increased numbers of effector-memory T-cells in the blood. In T-cells from natalizumab-treated MS patients, the expression of TNF-α mRNA was increased whereas the expression of fourteen other effector cytokines or transcription factors was unchanged. Natalizumab-treated MS patients had significantly decreased expression of the co-stimulatory molecule CD134 on CD4+CD26HIGH T-cells, in blood, and natalizumab decreased the expression of CD134 on MBP-reactive CD26HIGHCD4+ T-cells in vitro. Otherwise CD4+ T-cells from natalizumab-treated and untreated MS patients showed similar responses to MBP. In conclusion natalizumab treatment selectively increased the effector memory T-cell pool but not the activation state of T-cells in the blood compartment. Myelin-reactive T-cells were not selectively increased in natalizumab treated MS.


Introduction
Relapses in multiple sclerosis (MS) are caused by focal inflammatory lesions, consisting of activated macrophages and CD4 + and CD8 + T-cells, in the central nervous system (CNS). Binding of the integrin very late antigen (VLA)-4, expressed on the surface of most immune cells, to the vascular cells adhesion molecule (VCAM)-1, expressed on endothelial cells, is important for the adhesion and transmigration of systemically activated immune cells from the blood into the CNS. VLA-4 consists of an a4 (CD49d) and b1 chain (CD29). Besides being an adhesion molecule, VLA-4 has co-stimulatory effects in T-cell activation [1][2][3].
The humanized monoclonal IgG4 antibody natalizumab binds VLA-4, and the use of natalizumab as MS treatment is well established and effective, reducing clinical and MRI measures of disease activity [4]. It is assumed that the primary therapeutic effect of natalizumab relies on an inhibition of VLA-4-mediated extravasation of activated immune cells into the CNS [3]. Still, it is not fully understood if natalizumab inhibits transendothelial migration simply by blocking ligation of VCAM-1 and VLA-4, as decreased levels of VLA-4 have been detected on the T-cell surface in natalizumab-treated MS-patients [1,5,6]. This observa-tion could be an artifact due to competitive binding of natalizumab and the antibody used to visualize the VLA-4 molecule for flow cytometry, due to natalizumab induced internalization of VLA-4 or reduction of a4 or b1 integrin chain synthesis.
Previous studies have shown that natalizumab-treated MS patients have an increased pool of T-cells and other lymphocytes in the blood compartment [7][8][9] which may be enriched in T-cells with an increased activation state or an increased potential to produce pro-inflammatory cytokines [10][11][12]. A possible explanation is that pro-inflammatory T-cells may be selectively retained in blood due to VLA-4 blockade [12]. Alternatively natalizumab may interfere with immune and T-cell activation in vivo [13], as different studies showed that VLA-4-antibodies may enhance or block co-stimulatory signals provided by VLA-4 in anti-CD3 induced T-cell proliferation and cytokine production [3]. Yet, studies examining whether natalizumab interferes with the antigen-specific activation of CD4 + T-cells are rare.
To what extent minor T-cell subsets may be selectively increased in an enriched T-cell pool has so far not been studied in depth [1,6]. Memory T-cells (CD45RA 2 ) can be divided into central memory (CCR7 + ) and effector memory (CCR7 2 ) T-cells [14,15] and may have different roles in the immunopathogenesis of MS [16]. Assessing the CD45RA and CCR7 (CD197) expression by flow-cytometry, T-cells can be divided into: antigen-inexperienced naïve (CCR7 + CD45RA + ) T-cells and antigen-experienced memory (CD45RA 2 ) T-cells. Memory T-cells can further be characterized by expression of CCR7 (CD197). CCR7 is a homing receptor that enables the central memory (CCR7 + CD45RA 2 ) T cell to enter secondary lymphatic tissue. Upon antigen-specific reactivation, central memory T-cells have a high capacity to proliferate but low capacity to produce effector cytokines.
In contrast effector memory T-cells (CCR7 2 CD45RA 2 ) are able to produce high levels of effector cytokines but do only show little proliferation upon activation by local APCs in the peripheral tissues. Depending on the expression of CD27 these cells can be further divided into subsets displaying different stages of differentiation [14,15].
CD26 and CD134 are co-stimulatory molecules which are expressed on T-cells. Increased CD26 expression on T-cells has been associated with disease activity in MS and it has been suggested that CD134 has a pathogenic role in different autoimmune disease [17][18][19]. All major IL-17 producing T-cell subsets express CD161 [20] and a single nucleotide polymorphism of the CD161 gene (KLRB1) has been associated with susceptibility to develop MS [21,22]. The effect of natalizumab on CD26, CD134, CD161 and central-memory and effector-memory T-cells has not been studied yet.
We hypothesized that treatment with natalizumab would result in the retention of effector T cells, myelin reactive T-cells producing pro-inflammatory cytokines and other circulating Tcell subsets that have been associated with MS. These hypotheses were addressed by studying the phenotype of circulating T cell subsets by flow cytometry and analysis of the gene expression, and by assessing CD4 + T cell reactivity to myelin basic protein (MBP) in untreated and natalizumab-treated MS patients.

Ethics Statement
The study was approved by the regional scientific Ethics Committee of Copenhagen County (protocol KF 01-0141/95 and KF 01-314009) and all persons participating in this study provided written informed consent.

Patients
For this study we obtained blood samples from 21 untreated and 25 natalizumab-treated MS patients with a relapsing-remitting disease course. The patients included in this study did not receive treatment with glucocorticoids or other immunomodulatory drugs and did not show signs of a relapse within a period of two months prior to sampling. Among the untreated MS patients, 4 patients had previously received immunomodulatory treatment but stopped for different reasons while the remaining 17 patients were treatment naïve. 14 of the treatment naïve patients were newly diagnosed with MS (within the preceding 12 months) scheduled to start immunomodulatory treatment. The median age was 35 years (IQR = 27-42) in the untreated group and 36 years (IQR = 32-44) in the natalizumab-treated group; the degree of disability was significantly lower among untreated MS patients (median EDSS 1.5; IQR = 1-3) than among natalizumab-treated MS patients (median EDSS 2.5; IQR = 2-4; p = 0.016); the median disease duration was significantly shorter among untreated MS patients (median 2 years; IQR = 1-8 years) than among natalizumab-treated MS-patients (median 9 years; IQR = 6-12 years; p = 0.001). The effect of natalizumab treatment on the mRNA expression of genes encoding the VLA-4 subunits a4 (ITGA4) and b1 (ITGB1), was measured in whole blood samples from 25 untreated RRMS patients and from 50 RRMS patients 3, 6 and 12 months after initiation of natalizumab treatment.
For PCR analysis of the gene expression in the blood, whole blood was collected in Tempus tubes (Applied Biosystems, USA).For all other studies blood was drawn in BD VacutainerH EDTA tubes (BD Biosciences, Denmark), and PBMCs were promptly separated by density gradient centrifugation on Lymphoprep (Axis-Shield, Norway).

Separation of blood cell subsets
CD4 + and CD8 + T-cells were separated from freshly isolated PBMCs using CD4 + T Cells Isolation Kit II, CD8 + T Cells Isolation Kit II and an autoMACS TM cell-separator (all from Miltenyi Biotec, Germany) according to the manufacturer's instructions.

Gene expression analysis
For quantitative real-time PCR (qPCR) analysis in whole blood cells mRNA was extracted with the Tempus 12-port RNA isolation kit (Invitrogen, Denmark) and cDNA was synthesized with the High Capacity cDNA RT kit (Applied Biosystems). For gene expression analysis of CD4 + and CD8 + T-cells, mRNA was isolated using Pico Pure RNA Isolation Kits (Arcturus, USA) and cDNA was synthesized with qScript cDNA SuperMix (Quanta BioSciences, USA). qPCR analyses were run on an ABI 7500 Real Time PCR System (Applied Biosystems) using predesigned and validated primers (Applied Biosystems; Table S1). Gene expression in each sample of the target mRNA, relative to an endogenous reference gene (DC t-sample ), was compared to a calibrator consisting of pooled cDNA made from PBMCs from healthy volunteers (DC t-pool ). We used CASC3 and UBE2D2 as reference genes. Gene-expression levels are given as normalization ratio (NR) calculated as: NR = 2 2DCt(sample)2DCt(pool) [23].
Flow cytometry analysis of CD4 + T-cell reactivity to MBP and TT For flow cytometry we used a BD FACSCanto II TM and the BD FACSDiva TM Software 6.1.2 (both from BD Biosciences, Denmark). Cells were harvested, washed in PBS at 4uC, and stained with anti-CD3 PacificBlue (PB), anti-CD4-PerCP-Cy5.5, anti-CD8-PeCy7, anti-CD19-APC-Cy7 and dead/live staining dye (Table S2) in a 50 ml reaction for 30 minutes in the dark at 4uC. Then the cells were washed in FACS-PBS (PBS/1% (w/v) HSA/ 2 nM EDTA (FACS-PBS)) and re-suspended in 100 ml FACS-PBS for flow cytometry. The proliferation of CD4 + T-cells was assessed as the percentage of CFSE-diluted cells within the CD3 + CD4 + population.
To measure the intracellular cytokine production in proliferating CD4 + T-cells, the cells were stained as described above using anti-CD3-PB, anti-CD8-PeCy7, anti-CD19-APC-Cy7 and live/ dead staining dye (Invitrogen, Denmark). Cells were not stained for CD4 as PMA induced a significant down-regulation of CD4 (data not shown). Cells were fixed and permeabilized with the FOXP3 permeabilization kit (BioLegend,USA) and then stained for 30 minutes at room temperature with combinations of: anti-IL-17A-PE and anti-IFN-c-APC; anti-IL-4-PE and anti-TNF-a-APC; and anti-IL-10-PE and anti-IL13-APC (Table S2). The cytokine expression was measured in proliferating and non-proliferating CD8 2 and CD8 + T-cells using flow cytometry. As control for nonspecific background fluorescence and non-specific antibody binding, TT-stimulated cells were stained with isotype controls (Table S2).

Flow-cytometry of CD4 + and CD8 + T-cells
Freshly isolated PBMCs were re-suspended in staining buffer (eBiosciences, USA). In a 65 ml reaction 5610 5 PBMCs were stained with fluorochrome-conjugated antibodies for the surface markers CD3, CD4, CD8 and CD49d together with combinations of: CD26, CD134 and CD154; CD161, IL23R and CD212; CD11a and CD18; or CCR7, CD45RA and CD27. As control for non-specific antibody binding, non-specific fluorescence and spectral overlap we used the fluorescence minus one method [24] combined with Isotype-matched control antibodies (Table  S2). The expression of the stained surface molecules was measured on CD3 + CD4 + CD8 2 and CD3 + CD4 2 CD8 + T-cell subsets by flow cytometry. Depending on the expression pattern of the target molecule, expression levels were assessed as median fluorescence intensity (MFI) or percentage of positively stained cells within a defined subset. To assess absolute numbers of T-cell subsets, 50 ml blood was stained with 20 ml BD Multitest TM antibody cocktail (BD Biosciences), containing antibodies against CD3, CD4, CD8 and CD45 in Trucount TM tubes for 15 min at room temperature followed by red blood cell lysis by adding 450 ml BD FACS lysing solution (BD Biosciences) to the cells for 20 minutes at room temperature. Finally, cells were measured by flow cytometry and counts (cells/ml) of CD3 + CD4 + CD8 2 and CD3 + CD4 2 CD8 + cells were calculated according to the instructions of the Trucount TM manufacturer (BD Biosciences).

Data and statistics
Data in figures are shown as boxplots; the box represents the interquartile range (IQR); the median value is indicated by a line; the whiskers represent range. Data in tables are given as median with IQR. If not otherwise stated, differences were tested for significance by non-parametric tests: Mann-Whitney U test for unpaired samples; Wilcoxon Signed Ranks test for paired samples. P-values below 0.01 were considered statistically significant. For data processing we used Microsoft Excel and SPSS v.18.

Circulating T cells and VLA-4 expression in natalizumabtreated MS
Numbers of circulating CD8 + T-cells were significantly higher in natalizumab-treated than in untreated MS patients; numbers of CD4 + T-cells were also higher, although not significantly different ( Figure 1A). Natalizumab-treated patients had significantly lower cell surface expression of VLA-4 on CD4 + T-cell and CD8 + T-cells ( Figure 1B). This did not reflect a general decrease in integrin expression since the surface expression of the integrin LFA-1's achain (CD11a) and b-chain (CD18) on CD4 + T-cells and CD8 + Tcells did not differ in untreated and natalizumab-treated MSpatients ( Figure 1C). Neither did it appear to be due to repression of the genes encoding the VLA-4 subunits a4 (ITGA4) and b1 (ITGB1) since the expression of these genes was comparable in whole blood samples from untreated and natalizumab-treated MS patients ( Figure 1D). Finally ex vivo studies showed that natalizumab did not interfere with the binding of the anti-VLA-4 antibody used in the flow cytometry studies, suggesting that these antibodies bind to different epitopes, thereby validating the use of this antibody for flow cytometry analysis of VLA-4 expression in natalizumab-treated MS patients ( Figure 1E). Thus lower VLA-4 expression on CD4 + and CD8 + T-cells in natalizumab-treated MS patients appears to be due to internalization or shedding of VLA-4 and not decreased synthesis.
The natalizumab-treated MS patients had a significantly higher EDSS and significantly longer disease duration than the untreated MS patients; but calculating the Spearman rank correlation coefficient (SRCC) we did not find any association of EDSS or disease duration with the count of circulating naïve, centralmemory, effector-memory or effector T-cells (data not shown).

CD26, CD134 and CD154 expression
Natalizumab-treated patients had significantly higher numbers of circulating CD26 NEG CD4 + T-cells and CD26 NEG CD8 + T-cells as compared to untreated MS-patients ( Figure 4A and Figure 4B). Except for CD26 NEG CD4 + T-cells, VLA-4 expression was significantly decreased on all CD26-defined CD4 + and CD8 + Tcells subsets in natalizumab-treated patients as compared to untreated RRMS ( Figure 4C and Figure 4D) patients. Globally, CD134 expression on CD4 + T-cells was not significantly different in untreated and natalizumab-treated MS-patients (data not shown), but the percentage of CD26 HIGH CD4 + T-cells which expressed CD134 was significantly lower in natalizumab-treated than in untreated MS patients ( Figure 4E). We did not find any association of disease duration or EDSS with the percentage of CD26 HIGH CD4 + T-cells which expressed CD134 (correlation analyzed by the SRCC; data not shown). CD8 + T-cells did not stain positive for CD134 and CD4 + T-cells and CD8 + T-cells did not stain positive for CD154 (data not shown). Gene expression in CD4 + and CD8 + T-cells in natalizumab treated MS Natalizumab-treated MS patients had significantly increased expression of TNF and significantly decreased expression of IL10 in circulating CD4 + and CD8 + T-cells (Table 1). We did not find any correlation between EDSS or disease duration with the mRNA expression of TNFA or IL10 in CD4 + and CD8 + T-cells (correlation analyzed by the SRCC; data not shown). The expression of TBX21, HLX1, IFNG, RORC, IL17A, IL17F, IL22, GATA3, IL4, IL5, FOXP3, TGFB1, PRF1 in circulating CD4 + and CD8 + T-cells was not different, comparing untreated and natalizumab-treated MS ( Table 1).

Reactivity of myelin-specific CD4 + T-cells and nonspecifically stimulated CD4 + T-cells from untreated and natalizumab-treated MS patients
We stimulated PBMCs from untreated and natalizumab-treated MS-patients with MBP and TT. We then assessed the frequency of proliferating CD4 + T-cells. We found no differences in the CD4 + T-cell proliferation or cytokine expression profile in untreated and natalizumab-treated MS patients ex vivo (Table 2). To examine the general potential of CD4 + and CD8 + T-cells from natalizumabtreated and untreated MS patients to express effector cytokines, we measured the cytokine production in response to PMA/ionomycin in cell cultures that had not been exposed to antigen. Even though our criteria for significance were not met, there was a trend towards a higher frequency of IL17-expressing CD3 + CD8 2 Tcells and CD3 + CD8 + T-cells in natalizumab-treated MS patients ( Figure 5)

Effect of natalizumab on myelin-specific CD4 + T-cell responses ex vivo
We examined the effect of natalizumab and IgG4-control antibody on the proliferation of CD4 + T-cells and their expression profile of CD26, CD134 and intracellular cytokines in response to MBP ex vivo. Natalizumab decreased the expression of CD134 on CD26 HIGH CD4 + T-cells which proliferated in response to MBP; the IgG4 control antibody had no significant effect on CD4 + T cell reactivity ( Figure 6A). Natalizumab had no significant effect on the fraction of proliferating CD4 + T-cells ( Figure 6B) or on the expression of IL-17A or IFN-c in MBP-reactive CD3 + CD8 2 Tcells ( Figure 6C and Figure 6D).

Discussion
In the present study we show that: 1) decreased VLA-4 surface expression on circulating CD4 + and CD8 + T-cells in natalizumab treated MS patients is not explained by decreased transcriptional activity of ITGA4 and ITGB1, encoding the a4 and b1 subunits of VLA-4, or associated with decreased expression of LFA-1 integrin. In addition we demonstrate that the observed decrease of VLA-4 surface expression is not caused by competitive binding of natalizumab and the antibody used for detection of VLA-4; 2) numbers of circulating effector memory CD4 + and CD8 + T-cells are increased in natalizumab treated MS-patients; 3) CD4 + and CD8 + cells from natalizumab treated patients express increased levels of mRNA encoding TNF-a transcript, however, examination of a broad panel of effector T-cell markers reveals only little evidence for a generally increased activation state of circulating Tcells; 4) reactivity of MBP-specific CD4 + T-cells from natalizumabtreated and untreated MS-patients does not differ ex vivo but natalizumab-treated patients have decreased expression of CD134 on CD26 HIGH CD4 + T cells in vivo and natalizumab decreases the frequency of MBP-reactive CD4 + T-cells that are CD26 HIGH CD134 + in vitro.
Explorative studies, as our own, may yield false positive results due to mass significance. We tried to meet this problem by choosing a 1% significance level and by examining groups of functionally associated targets. Furthermore we compared small cohorts which may limit the ability to show significance in some differences we observed. Our T cell reactivity assay is biased towards measuring HLA class II-dependent responses, induced by whole protein preparations, rather than CD8 + T-cell responses, as the latter are better studied by analyzing the reactivity to HLA class I-binding peptides. Data interpretation is hampered by differences in disability (EDSS) and disease duration between the group of untreated and natalizumab treated MS patients. Nonetheless, the major positive findings did not correlate with the duration of the MS disease or the disability of the MS patients. There may also be a selection bias regarding the disease course: natalizumab treated patients must have had a history of high disease activity, as this is a requirement to start treatment with natalizumab, whereas the disease course of the mainly newly diagnosed untreated MS patients cannot be predicted. A longitudinal study design had been desirable but was not realizable as most natalizumab treated patients, according to treatment guidelines, receive immunomodulatory treatment with glatiramer acetat or interferon-beta until the day natalizumab-treatment is initiated.
In MS the role of the circulating effector memory T cells, which was the memory T-cell subset that was most profoundly increased in natalizumab treated patients, is uncertain. Treatment with fingolimod, which mainly results in the retention of central memory T cells in lymph nodes, is an efficacious MS therapy [25][26][27]. Indeed, one study has indicated that it is mainly CCR7 + T cells that enter the CNS from the blood stream in MS, and that local reactivation within the CNS is needed for the differentiation of these cells to effector memory cells that subsequently invade the CNS parenchyma [16]. Nevertheless, effector memory T-cells have a high capacity to produce pro-inflammatory cytokines [14,15] and may contribute to an enriched pool of proinflammatory cells.
In natalizumab-treated patients, increased frequencies of systemic T-cells and PBMCs that produce pro-inflammatory cytokines such as IL-2, IFN-c, TNF-a and IL-17 were found, but often only transiently [10,12]. Even though statistical significance was not reached, our own and another study also showed trends towards increased frequencies of pro-inflammatory, IL-17 secreting, cells in patients treated with natalizumab [5]. In the present study the ex vivo response of CD4 + T-cells to MBP did not differ between natalizumab-treated and untreated MS-patients. This argues against a selective accumulation of potentially autoaggressive MBP reactive CD4 + T-cells in the circulation in natalizumab-treated MS patients. However, we cannot exclude that other pathogenic auto-reactive T-cells may be increased. Furthermore, it is uncertain to what extent the CFSE assay used in the present study measures antigen-reactivity in central memory and effector memory T cell subsets [14,15].
Other explorative studies suggested that expression levels of mRNA encoding transcription factors and cytokines associated with T-cells are increased in circulating immune cells in natalizumab-treated patients. One study found transiently increased expression of IL17 and CCR6 and persistently increased  Table 1. mRNA expression of genes associated with T-helper (T H ) cell immune activation and regulatory T-cell activity in CD4+ and CD8+ T-cells from untreated and natalizumab treated MS-patients.  CD161 is a transmembrane cell surface molecule which is expressed on CD4 + and CD8 + T-cells, and a single nucleotid polymorphism of the CD161 gene (KLRB1) has been associated with susceptibility to develop MS [21]. In MS, surface expression of CD161 is increased on circulating CD8 + T-cells, and CD161 + CD8 + T-cells are present in MS lesions. How CD161 potentially contributes to the pathogenesis of MS is uncertain [28][29]. All major IL-17 producing T-cell subsets express CD161 [20]. In agreement with this we found, compared to CD161 NEG cells, higher expression of the IL-23-receptor (IL-23R) on CD161 + CD4 + T-cells and on CD161 LOW CD8 + T-cells. Still, CD161 is not a T H 17-marker, as CD161 + T-cells' production of IFN-c by far exceeds that of IL-17, and in particular CD161 HIGH CD8 + T-cell in MS patients are sensitive to IL-12-induced proliferation and IFN-y production [28]. This again is consistent with significantly higher CD212 (IL-12 receptor-beta 2 chain) expression on CD161 HIGH CD8 + T-cells as compared to CD161 NEG CD8 + Tcells in the present study. We assumed that assessment of IL23R and CD212 on CD161-subsets could reveal differences in the activation state of CD4 + and CD8 + T-cells. However, expression of both receptors was comparable in untreated and natalizumabtreated patients.
The co-stimulatory molecule CD26 (dipeptidyl peptidase IV) is expressed on memory CD4+ and CD8+ T cells [17,18]. Myelin basic protein (MBP)-specific T-cell-clones, derived from MS To assess antigen-induced CD4 + T-cell proliferation we incubated CFSE labeled peripheral blood mononuclear cells (PBMCs) from untreated and natalizumab treated MS patients with myelin basic protein and tetanus toxoid for 7 days. Using flow-cytometry CD4 + T-cell proliferation was assessed as the percentage of CFSE diluted cells within the CD4 + T-cell proliferation. After activation with phorbol 12-myristate 13-acetate and staining of intracellular cytokines the percentage of cytokine expressing cells within the proliferating CD8 2 T-cell population was determined by flow-cytometry. Data are given as median with interquartile range in the parenthesis. There were no significant differences in antigen-specific CD4 + T-cell reactivity between untreated and natalizumab treated MS patients using Mann-Whitney U Test with a level of significance p,0.01. IL = interleukin; IFN = interferon; TNF-a = tumor necrosis factor-a. doi:10.1371/journal.pone.0047578.t002 patients express high levels of CD26 [30], and T-cells with high levels of CD26 express IFN-c and TNF-a [31]. Increased frequencies of circulating CD4 + T-cells and CD8 + T-cells, expressing CD26 have been associated with clinical and MRI measures of disease activity in MS [31][32][33][34]. CD134 is expressed on activated CD4+ T cells (including T H 1, T H 2 and T H 17 and T REG cells) and CD8 + T-cells [19] and appears to promote activation, clonal expansion and cytokine secretion of these cells, when challenged with antigen [19,35]. Stronger CD134 expression on CD26 HIGH than on CD26 NEG and CD26 INT CD4 + T-cells suggests that the expression of these molecules may be associated. In a study of IFN-b treatment effects in RRMS, we found that treatment with IFN-b reduced the expression of CD134 on CD26 HIGH CD4+ T cells [36], in accordance with what we found in natalizumab-treated patients in our study. Thus decreased CD134 expression on CD26 HIGH CD4 + T-cells may reflect a decrease in T-cell activation. Indeed, we also found that the expression CD134 was reduced on CD26 HIGH CD4 + T-cells that proliferated in response MBP in the presence of natalizumab in vitro. Depending on the epitope specificity, different VLA-4 antibodies are able to block or stimulate VLA-4 mediated costimulatory signals in non-specifically or MBP-activated (CD4 + ) Tcells [1,2,[37][38][39][40][41]. Likewise, natalizumab may prevent VLA-4 engagement by endogenous costimulatory molecules in our system. Overall there is little evidence that natalizumab itself has a pro-inflammatory effect on PBMCs and CD4 + T-cell responses in vitro [6,12,13].
In conclusion our study suggests that effector memory CD4 + and CD8 + T-cells but not MBP-reactive CD4 + T-cells are selectively enriched in natalizumab treated patients. Our own data and other previous publications on the activation state of CD4 + and CD8 + T-cells are ambiguous and indeed we showed that the T-cell activation marker CD134 is expressed significantly lower on CD4 + CD26 HIGH T-cells in natalizumab-treated RRMS. Our finding that natalizumab also decreased the expression of CD134 + CD26 HIGH which proliferated in response to MBP in vitro suggests that natalizumab may not only affect the recirculation of CD4 + T-cells but may also directly induce phenotypical changes of the CD4 + T-cell. The relevance of this finding for the clinical efficacy of natalizumab remains to be established.

Supporting Information
Table S1 Targets used in gene expression analysis by PCR. (DOCX)