ORIGINAL ARTICLE
Blockade of CD11a by Efalizumab in Psoriasis Patients Induces a Unique State of T-Cell Hyporesponsiveness Emma Guttman-Yassky1,3, Yulia Vugmeyster2,3,4, Michelle A. Lowes1, Francesca Chamian1, Toyoko Kikuchi1, Mark Kagen1, Patricia Gilleaudeau1, Edmund Lee1, Brisdell Hunte2, Kathy Howell2, Wolfgang Dummer2, Sarah C. Bodary2,5 and James G. Krueger1 Efalizumab (anti-CD11a) interferes with LFA-1/ICAM-1 binding and inhibits several key steps in psoriasis pathogenesis. This study characterizes the effects of efalizumab on T-cell activation responses and expression of surface markers on human circulating psoriatic T cells during a therapeutic trial. Our data suggest that efalizumab may induce a unique type of T-cell hyporesponsiveness, directly induced by LFA-1 binding, which is distinct from conventional anergy described in animal models. Direct activation of T cells through different activating receptors (CD2, CD3, CD3/28) is reduced, despite T cells being fully viable. This hyporesponsiveness was spontaneously reversible after withdrawal of the drug, and by IL-2 in vitro. In contrast to the state of anergy, Ca þ 2 release is intact during efalizumab binding. Furthermore, lymphocyte function-associated antigen-1 (LFA-1) blockade resulted in an unexpected downregulation of a broad range of surface molecules, including the T-cell receptor complex, co-stimulatory molecules, and integrins unrelated to LFA-1, both in the peripheral circulation and in diseased skin tissue. These observations provide evidence for the mechanism of action of efalizumab. The nature of this T-cell hyporesponsiveness suggests that T-cell responses may be reduced during efalizumab therapy, but are reversible after ceasing efalizumab treatment. Journal of Investigative Dermatology (2008) 128, 1182–1191; doi:10.1038/jid.2008.4; published online 31 January 2008
INTRODUCTION Lymphocyte function-associated antigen-1 (LFA-1) is a heterodimeric glycoprotein belonging to the b2-integrin family (Carlos and Harlan, 1994; Werther et al., 1996), also known as aL-b2 or CD11a/CD18, which binds intracellular adhesion molecules 1–3 (ICAM-1, ICAM-2, and ICAM-3). LFA-1 plays a pivotal role in several important homeostatic processes, including T lymphocyte recirculation, trafficking to sites of inflammation, antigen presentation by dendritic cells, and other cells such as activated keratinocytes (Stewart et al., 1996; Hogg et al., 2003; Kinashi, 2005), and T-cell
1
Laboratory for Investigative Dermatology, The Rockefeller University, New York, New York, USA and 2Genentech Inc., South San Francisco, California, USA
3
These authors contributed equally to this work
4
Current address: Wyeth-Research Inc., Andover, Massachusetts, USA
5
Current address: Schering Plough Biopharma, Palo Alto, California, USA
Correspondence: Dr JG Krueger, Laboratory for Investigative Dermatology, The Rockefeller University, 1230 York Avenue, NY, PO Box 178, New York 10065, USA. E-mail:
[email protected] Abbreviations: BD, Becton Dickinson; LFA-1, lymphocyte function-associated antigen-1; MFI, mean fluorescence intensity; PE, phycoerythrin; PMA, phorbol-12-myristate-3-acetate; RT, room temperature; TCR, T-cell receptor Received 7 June 2007; revised 1 November 2007; accepted 27 November 2007; published online 31 January 2008
1182 Journal of Investigative Dermatology (2008), Volume 128
co-stimulation (Van Seventer et al., 1990, 1991, 1992; Kanner et al., 1993). LFA-1 is a molecular target for therapeutic modulation of immune responses in disease states, as tissue inflammation in experimental models has been blocked by administration of LFA-1 antibodies (Ferrara et al., 1990; Isobe et al., 1992; Scheynius et al., 1993; Tanaka et al., 1993; Gordon et al., 1995; Moriyama et al., 1996; Willenborg et al., 1996). However, little is known about the therapeutic mechanisms of targeting integrins in human subjects. This issue is important, since the humanized anti-CD11a antibody (efalizumab) is currently an approved agent for treatment of psoriasis vulgaris (Gottlieb et al., 2002; Lebwohl et al., 2003). As psoriasis affects about 25 million people in North America and Europe, new psoriasis therapies are likely to be used in large numbers of patients. Anergy may be broadly defined as a functional state where a T cell fails to react to its cognate antigen, and it is often considered to be a mechanism to maintain peripheral tolerance (Appleman and Boussiotis, 2003; Schwartz, 2003; Macian et al., 2004). Additional features of anergic T cells include a restoration of functional inactivity by exogenous IL-2, and a lack of Ca2 þ mobilization, which normally occurs during T-cell activation. This study was designed to characterize the effects of efalizumab on T-cell activation responses and expression & 2008 The Society for Investigative Dermatology
E Guttman-Yassky et al. Hyporesponsiveness of T Cells by Efalizumab Therapy
of T-cell surface molecules during a therapeutic trial. Here we demonstrate two unexpected effects on T cells obtained from psoriasis patients. First, direct stimulation of T cells with mitogenic antibodies was significantly decreased during active treatment (anti-CD2, anti-CD3, and combined anti-CD3/CD28), but remained normal with phorbol-12-myristate-3-acetate (PMA) and ionomycin. This is in contrast to the increase in proliferative responses seen with a polyclonal stimulus in murine T cells treated with LFA-1 antibodies (Bohmig et al., 1994). Thus, only in the human experiments were T cells hyporesponsive to direct stimulation via surface molecules. Second, downregulation of numerous activation-controlling surface molecules on T cells was produced by efalizumab treatment, including those associated with the T-cell receptor (TCR) complex, co-stimulatory pathways, and integrins unrelated to LFA-1. Since both of these effects were not anticipated from the study of LFA-1 antibodies in model systems, they reinforce the need to study biologic effects of new immune antagonists in human populations receiving these agents. RESULTS In vivo, T-cell activation through anti-CD3-, anti-CD2-, and anti-CD3/CD28 is inhibited by efalizumab (anti-CD11a)
We determined whether in vivo exposure of lymphocytes to efalizumab reduced responsiveness through surface molecule ligation. Anti-CD3 stimulation of peripheral blood lymphocytes from 12 untreated psoriatic patients showed dosedependent activation, as measured by increased expression of CD69 on CD2-expressing lymphocytes (both mean fluorescence intensity (MFI) and percent positive cells increased) (Figure 1a). After 14 days of efalizumab treatment, CD3-mediated activation was greatly reduced at all doses in all subjects. This non-responsiveness of lymphocytes to activation by anti-CD3 was maintained throughout the treatment period with efalizumab (day 84). After washout of efalizumab (day 168), CD3-mediated T-cell activation returned to normal (Figure 1a), in line with re-expression of CD11a on T cells (Figure 1b). In a second group of 15 patients, we tested additional activating stimuli, including soluble anti-CD3 antibody, a mitogenic pair of anti-CD2 antibodies (T11.1 and T11.2), and bead-bound anti-CD3 and anti-CD28 antibodies (Figure 1c and d). Consistent with results in the first group of patients, T-cell activation (CD69 upregulation) to soluble anti-CD3 antibodies was inhibited in all 15 patients (Po0.001; Figure 1d). T-cell activation induced by CD2 ligation was also consistently inhibited during in vivo treatment with efalizumab (P ¼ 0.001). However, T-cell activation with a stronger stimulus (bead-bound anti-CD3/anti-CD28 antibodies) induced greater expression of CD69 compared with soluble CD3 or CD2 antibodies (Figure 1c and d). Although eight efalizumab-treated patients had reduced T-cell activation responses to this combined stimulus (example shown in Figure 1c), there was only about a 20% average reduction in responses to CD3/CD28 activation (P ¼ 0.37). The hyporesponsiveness of T cells under efalizumab therapy
was seen in all patients, and this was unrelated to the therapeutic efficacy. We also examined T-cell activation induced by PMA and ionomycin to determine whether T-cell hyporesponsiveness in efalizumab-treated patients was due to deletion of cells with activation potential (Figure 1e and f). We considered that this might occur even in the setting of efalizumabinduced relative leukocytosis, (Vugmeyster et al., 2004). Upon PMA/ionomycin activation, virtually all T cells showed increased expression of CD69, and neither the frequency of cells showing activation (data not shown), nor the levels of CD69 induced on individual cells, were reduced during efalizumab treatment. This indicated that the activation potential of the T cells was intact. We next determined whether a functional response of T cells was impaired by in vivo efalizumab treatment. Given that IFN-g-producing T cells are excessively expanded and activated in psoriasis patients (Szegedi et al., 2003), we examined T cells for their ability to synthesize this cytokine in response to activation by soluble anti-CD3 or anti-CD2 antibodies, by bead-bound anti-CD3/anti-CD28 antibodies, and by PMA/ionomycin activation before and during treatment with efalizumab (Figure 2). With soluble anti-CD3 antibodies, T cells from psoriasis patients treated with efalizumab for 2 weeks showed consistently reduced ability to synthesize IFN-g (Po0.02) (Figure 2a–c). As with induction of CD69, nearly complete inhibition of IFN-g synthesis in response to CD3 ligation was observed, whereas a lesser inhibition of response to CD3/CD28 co-ligation was measured (data not shown). The average frequency of IFN-g-producing cells was not altered by efalizumab treatment when PMA/ionomycin was used as the activation stimulus (P ¼ 0.2), confirming the presence of responsive T cells, cell viability, and the ability of the cells to produce this cytokine (Figure 2b and c). The differential T-cell activation response to soluble antiCD3 antibodies versus PMA/ionomycin suggested that early events in signal transduction through the TCR receptor complex might be blocked by efalizumab exposure in vivo. In contrast, T-cell anergy induced by other methods is associated with reduction in the signaling cascade that leads to intracellular calcium release (Haque et al., 1998). To test whether efalizumab induced a similar alteration in T-cell responsiveness, intracellular calcium release was monitored by Indo-1, a protein which fluoresces at different wavelengths depending on its binding to Ca þ 2. As illustrated for two patients with psoriasis in Figure 3b, T cells showed excellent Ca þ 2 flux responses to stimulation with anti-CD3 antibody plus surface cross-linking of this antibody with anti-mouse IgG. Those results were equivalent to the Ca þ 2 flux responses from normal volunteers (Figure 3a). T cells taken from patients at 2 weeks of treatment with efalizumab had no reduction in the intracellular calcium release response, including both the kinetics of release and the overall magnitude of the response. Similar results to those illustrated in Figure 3 were observed in eight sequential patients studied that were stimulated with anti-CD3 (Figure S1) or anti-CD3 and anti-CD2 (Figure S2). www.jidonline.org 1183
E Guttman-Yassky et al. Hyporesponsiveness of T Cells by Efalizumab Therapy
350
200 Day 0 Day 14 Day 84 Day 168
250 200
Mean CD11a MFI
Mean CD69 MFI
300
150 100 50 0 1,000
200
40
8
150 100 50 0
1.6
Day 0
Day 14
Day 84
Day 168
Anti-CD3 (ng ml–1) Control
Anti-CD3
Anti-CD2
Anti-CD3/CD28
Anti-CD3
Anti-CD2
Anti-CD3/anti-CD28
% Change in CD69 MFI
CD3 CD3
Day 14
Day 0
0
CD69
CD69
CD69
–20 –40 –60 –80 –100 –120
*
CD69
35
500
Number of events
30
Mean CD69 MFI
*
–140
25 20 15 10
400 300
200 100
5 0
0
Day 0
Day 14
PMA/ionomycin
101
102
103
104
CD69
Figure 1. Inhibition of ex vivo anti-CD3-, anti-CD2, and anti-CD3/anti-CD28-induced activation of circulating T cells in efalizumab-treated psoriatic patients. (a) CD69 levels (MFI) on T cells stimulated ex vivo with increasing concentrations of anti-CD3 antibody at different time points during (up to day 84) and after efalizumab treatment. (b) CD11a expression on unstimulated T cells to assess CD11a saturation throughout the treatment period (up to day 84) and upon efalizumab washout. (c) A set of representative dot plots illustrates that efalizumab treatment by week 2 inhibits T-cell activation (CD69 MFI) induced by all three cell-surface stimuli, although less so with anti-CD3/CD28 stimuli. (d) Mean percent change in CD69 MFI with various stimuli, in patients on efalizumab treatment compared with baseline. *Indicates Po0.05. (e) Mean change in CD69 MFI with PMA/ionomycin stimulation on days 0 and 14 of efalizumab treatment showing no difference in T-cell activation by this stimuli. (f) A representative CD69 histogram for PMA/ionomycin stimulation illustrates that efalizumab treatment (black line, day 14) has no effect on PMA/ionomycin-mediated upregulation of CD69, as compared with baseline (gray line, day 0). Dotted line corresponds to the baseline CD69 expression on unstimulated T cells.
In vitro, T-cell activation through anti-CD3 and anti-CD3/CD28 is inhibited by efalizumab (anti-CD11a), and this hyporesponsiveness is instantly reversed by IL-2
was impaired in the presence of efalizumab and this inhibition was fully reversed by IL-2 (Figure 4c).
Anti-CD3 stimulation of T-lymphocytes in whole blood from healthy volunteers was measured by induced expression of CD69 on CD2-expressing lymphocytes. In vitro incubation with efalizumab resulted in dose-dependent inhibition of the activation response, with the IC50 of B0.1–0.3 mg ml1 (Figure 4a). Furthermore, the maximal inhibition was measured near the levels of efalizumab that would occur in vivo in psoriasis patients (B10 mg ml1) (Figure 4a). The murine monoclonal MHM24 (parent of efalizumab) and Fab fragments of efalizumab also strongly inhibited the T-cell proliferation response to CD3 ligation (Figure 4b). Similarly, T-cell proliferation triggered by CD3 and CD28 co-ligation
Peripheral blood T cells show down-modulation of cell-surface expression of CD3, CD2, TCR, CD4, CD8, VLA-4, and CD28 in efalizumab (anti-CD11a)-treated patients
1184 Journal of Investigative Dermatology (2008), Volume 128
As two types of intracellular activation systems (calcium response and PMA/ionomycin response) were functioning correctly in T cells from efalizumab-treated patients, we next analyzed the expression of cell-surface molecules that might control T-cell activation responses before and during efalizumab treatment. We examined the levels of an array of different surface molecules on both the peripheral T cells and on T cells obtained from psoriasis skin lesions by shortterm incubation.
E Guttman-Yassky et al.
1,000
50
Counts
0
600
Stimulated
400 200
–50
2
3
4
5
6
7
8
10 11 13 14 15 16 17
1,000
Patient number Anti-CD3
Anti-CD2
800
PMA+iono
CD3
Counts
Control
600
Pretreatment
OKT3 Anti-mouse IgG
1
OKT3 Anti-mouse IgG
0
–100
Day 0
Unstimulated
800
Normal
% Change in IFN-γ -producing ability at day 14
100
OKT3+T11 Anti-mouse IgG
Hyporesponsiveness of T Cells by Efalizumab Therapy
Week 2
400
Psoriasis
CD3
% Change in IFN-γ -producing ability at day 14
1,000 800
INF-γ
INF-γ
100 80 60 40 20
0
–20 –40 –60 –80 –100
0
PMA/ionomycin
Baseline
Day 14
INF-γ 100 80 60 40 20 0 –20 –40 –60 –80 Anti-CD3 –100
Baseline
Counts
Day 14
200
INF-γ
600 400 200 0
50
100 150 200 250 0
50
100 150 200 250
Seconds
Day 14
Figure 2. Inhibition of ex vivo anti-CD3- and anti-CD2-induced IFN-c production by T cells in efalizumab-treated psoriatic patients. (a) T cells’ ability to synthesize IFN-g in response to anti-CD3 antibody after 2 weeks of efalizumab treatment is highly diminished (Po0.02). (b) Representative FACS data for three different T-cell activation stimuli, anti-CD3, anti-CD2, PMA/ionomycin, at baseline and week 2 treatment with efalizumab. Efalizumab affected T-cell ability to produce IFN-g in response to CD3 ligation but not to PMA/ionomycin stimulation. (c) Percent change in IFN-g þ T cells in all study patients. Black lines correspond to individual patients.
Previous studies have demonstrated that efalizumab binds to CD11a and saturates binding sites and also has the ability to downregulate both subunits of LFA-1 (CD11a and CD18) (Vugmeyster et al., 2004). First, we studied changes in surface antigens in peripheral blood of efalizumab-treated patients (Figure 5b–g, left side). Saturation of CD11a on circulating T cells was determined using a competing custom-conjugated FITC-efalizumab antibody (Figure 5b), whereas downregulation of CD11a was assessed with CD11a (clone 25.3), which binds to a non-competitive epitope (Figure 5c). There was almost complete saturation of CD11a with efalizumab during therapy (Po0.00001) (Figure 5b). As expected, both subunits of LFA-1 were significantly downregulated in all subjects. The expression of CD11a was reduced by 85–73% (Po0.0001),
Figure 3. Assessment of intracellular signalling through TCR/CD3 complex. Normal anti-CD3-mediated intracellular Ca2 þ flux response in T cells from efalizumab-treated patient PBMCs that were loaded with Indo-1, activated with anti-CD3, and cross-linked with an anti-mouse IgG. The Ca þ 2 flux was measured as the ratio between bound and unbound Indo-1 versus time. (a) Intracellular Ca þ 2 flux from a representative normal volunteer pre- and post-stimulation with OKT3 (anti-CD3) and T11 (anti-CD2). (b and c) T cell intracellular Ca þ 2 flux responses from two representative patients are presented before and after efalizumab therapy.
whereas the expression of CD18 was reduced by 76–63% (Po0.0001). Other surface receptors unrelated to LFA-1 were also quantified (Figure 5). The levels of CD3 were not affected by efalizumab (P40.2) (Figure 5d and e). The majority of patients showed downregulation of CD28 on circulating T cells with a mean reduction of 22% (P ¼ 0.02). Both the CD28hi and CD28lo populations were reduced (Figure 5d). We were particularly interested in studying the modulation of VLA-4 (CD29 and CD49d) during efalizumab therapy, as it is an important alternative integrin for lymphocyte trafficking in the presence of LFA-1 blockade. MFI values for both components of VLA-4 on circulating CD3 þ CD8 þ cells decreased significantly at day 14 (Figure 5e and f, Figure 6). There was a 52–76% reduction in CD29 (Po0.007) and a 75–68% reduction in CD49d (P ¼ 0.001) (Figure 5c and f). The CD29hi population was largely eliminated, whereas the CD29lo population was less affected (Figure 5f and g). Similarly, the CD49dhi population was significantly decreased by approximately 80% (P ¼ 0.006) (Figure 6). www.jidonline.org 1185
E Guttman-Yassky et al.
% Inhibitation of anti-CD3-induced CD69 upregulation
Hyporesponsiveness of T Cells by Efalizumab Therapy
100 80
In vivo efalizumab concentration
60
40
20
0 0.0001
0.01
1
100
Efalizumab (μg ml–1) 160,000
80,000
120,000
Mean CPM±SD
Mean CPM±SD
Anti-CD3+activation
80,000 40,000 0 IgG
Efalizumab
MHM24
60,000 40,000 20,000 0
Fab of Efalizumab
Anti-CD3+CD28 activation
– IgG
– IL-2 (50 U ml–1) IL-2 (50 U ml–1) Efalizumab Efalizumab Efalizumab
Figure 4. Efalizumab-mediated inhibition of in vitro anti-CD3- and anti-CD3/anti-CD28-induced T-cell activation and the reversal of this inhibition by exogenous IL-2. (a) Anti-CD3-induced T-cell activation in the presence of the indicated concentration of efalizumab was measured by monitoring CD69 expression (geometric MFI) on CD2 þ lymphocytes in whole blood obtained from normal healthy volunteers. A bar indicates efalizumab in vivo levels in psoriatic patients during therapeutic administration. (b) Anti-CD3-dependent PBMC proliferation in the presence of control IgG, efalizumab, monoclonal antibody MHM24 (murine parent of efalizumab), or Fab fragment of efalizumab was measured. (c) Anti-CD3/anti-CD28-dependent T-cell proliferation in the presence of efalizumab (1 mg ml1) with and without IL-2 was assessed. Panels (b) and (c) show mean data (n ¼ 3), with error bars showing standard deviations.
CD2 levels were significantly downregulated (although still quantifiable) with efalizumab treatment by 32–50% (Po0.03) (Figure 5a and g). The expression of the CD8 co-receptor on CD3 þ CD8 þ T lymphocytes was significantly decreased as early as day 14, with 58–65% reduction (Po0.003). CD8 MFI values showed a clear trend of recovery to pretreatment levels after therapy termination (Figure 6c). Finally, there was also a significant down-modulation of CD4 and TCR to about 64–82% of baseline values (data not shown). Psoriatic lesional T cells show down-modulation of cell-surface expression of LFA-1, CD28 þ , CD29 þ , CD8 þ and CD3 þ in efalizumab (anti-CD11a þ )-treated patients
significantly reduced to 83% of baseline (P ¼ 0.01). CD8 expression was significantly reduced to about 78% of baseline (P ¼ 0.03). However, unlike peripheral blood lymphocytes, CD3 expression on T cells in psoriatic lesions was significantly decreased in 8 out of 14 cases. Overall the mean reduction in the MFI for the CD3 marker was 22% (P ¼ 0.0004). There were no significant changes in the expression of TCR, CD2, CD4, and CD49d. Hence, although efalizumab saturates T-cell CD11a and downregulates LFA-1 in both circulating and skin compartments, there are interesting differences in the effects on other surface molecules. For example, the mean reduction in CD28 is higher in skin infiltrating T cells, whereas changes in CD2 and TCR are apparent only in circulating cells.
T cells obtained from psoriatic skin lesions before and during efalizumab therapy were examined for the expression of cellsurface markers involved in T-cell trafficking and activation for comparison to circulating T cells (Figure 5a and b–g, right side). Not all the changes observed in circulating T-cell surface antigens described above were present in psoriatic lesional T cells. CD11a and CD18 median levels were 55 and 44% of baseline, respectively (P ¼ 0.001), which was a pattern similar to circulating T cells, but greater in magnitude. There was also an extremely consistent down-modulation of CD28 expression to approximately 68% of the baseline level (P ¼ 0.001), which was also a greater reduction relative to that on circulating T cells. CD29 levels were moderately but
DISCUSSION In this study, we describe a unique T-cell hyporesponsiveness induced by anti-CD11 binding through the humanized monoclonal therapeutic antibody efalizumab. We postulate that efalizumab directly induces an alteration of the T-cell activation threshold for CD3 or CD2. This altered activation threshold induced by LFA-1 antagonism in humans is distinct from anergy described in mouse models (Appleman and Boussiotis, 2003), since (a) the suppression is antigenindependent, as demonstrated by CD2 and CD3 effects; (b) the hyporesponsiveness is spontaneously reversible upon washout of the drug; and (c) Ca2 þ flux is preserved after CD3 ligation. However, similar to true anergy, hyporesponsiveness
1186 Journal of Investigative Dermatology (2008), Volume 128
E Guttman-Yassky et al. Hyporesponsiveness of T Cells by Efalizumab Therapy
120
Mean MFI at day 14 relative to baseline
Tissue
Blood 2
Blood 1
100 80
* ***
* ** *
**
*
** 60
* * ** * ** * *
40
**
20 0 CD11a
CD18
CD2
CD28
CD29
CD49d
Blood T cells Day 0
CD3
Skin T cells Week 2
96.7
Day 0 15.1
Week 2 97.6
5.98
98.7
22.6
CD18
CD18
CD8
CD11a E 88.5
CD18
50.7
CD3
CD11a 25.3
MFI=212
MFI=62
MFI=99
MFI=41
CD28 18.8
44.6
15.8
91
77.4
CD3
54.5
MFI=175
MFI=50
MFI=26
MFI=40
CD29
CD49d
MFI=263
MFI=87
CD29
CD49d
MFI=139
MFI=65
CD2
Figure 5. Down-modulation of multiple cell-surface molecules in T cells from peripheral blood and lesional skin in efalizumab-treated psoriatic patients. (a) Mean level (MFI) of surface molecules on T cells at baseline and after 2 weeks of treatment for peripheral blood (study 1, open square; study 2, gray square) and lesional skin (black square). Asterisks (*) indicate that relative MFI values were statistically different compared with baseline, *Po0.05, **Po0.01, ***Po0.001. (b–g) A representative set of FACS dot plots from an individual patient (gated on T cells, except for panels (d) and (e)) is shown for peripheral blood (left) and lesional skin (right) at baseline and day 14. Quadrants were set on the basis of appropriate isotype controls. A fraction of positive cells in the indicated quadrants or an MFI value is shown. In panel (g), MFI values were calculated for T cells in the circled gate (CD29hi).
in efalizumab-treated patients was instantly reversible with IL-2. We excluded other explanations for this reduced T-cell activation, such as depletion of reactive T-cell populations or
reduced T-cell activation potential. Efalizumab-treated psoriasis patients have increased circulating T cells, so it is unlikely that there would be depletion of reactive T cells (Vugmeyster et al., 2004), and furthermore, the cells were www.jidonline.org 1187
E Guttman-Yassky et al. Hyporesponsiveness of T Cells by Efalizumab Therapy
100
80
80
60
60
40
40
20
20
% of max
100
0 100
101
102
103
104
0 100
101
CD49d 440 CD49dhigh All CD3+CD8+cells
240
140
Mean CD8 MFI±SD
Mean CD49d MFI±SD
103
104
500
CD49dlow
340
102
CD49d
400 300 200 100 0
40 0
14
28
56
Day
0
14
28
56
56+
56++
Last
Day
Figure 6. Down-modulation of VLA-4 and CD8 in peripheral blood CD8 þ T cells in efalizumab-treated psoriatic patients. (a) Histograms for a representative patient illustrate bimodal distribution of VLA-4 (CD49d/CD29) at baseline (gray line) on CD8 þ cells, and a subsequent shift in VLA-4hi/VLA-4lo ratio after 2 weeks of treatment. (b) Individual patient data were used to calculate the mean CD49d MFI for each time point and for each CD3 þ CD8 þ subset, as indicated. Dotted line corresponds to baseline CD49d MFI value for the CD49dlo subset. (c) CD8 levels on all CD3 þ CD8 þ cells were quantified throughout the treatment period (12 weeks) and upon efalizumab washout.
viable (annexin V/propidium iodide staining; data not shown) and fully activatable by PMA/ionomycin. Peripheral CLA þ T cells were not investigated, although it might be interesting to consider this T cell subset in a future study. Efalizumab is efficacious in treating psoriasis and it is assumed that key mechanisms of action include reduced migration of T cells into the skin, as well as reduced T-cell activation through interference with immune synapse formation between T cells and antigen-presenting cells, as suggested by animal models (Ferrara et al., 1990; Isobe et al., 1992; Scheynius et al., 1993; Tanaka et al., 1993; Gordon et al., 1995; Moriyama et al., 1996; Willenborg et al., 1996). In superantigen T-cell activation experiments from psoriasis patients, T-cell activation was also reduced (unpublished data), possibly reflecting interference with immune synapse formation. Here, we demonstrate reduced T-cell responsiveness independent of immune synapse formation, suggesting a direct effect on T cells. The mechanism of this reduced responsiveness is important to understand. If conventional anergy and possibly tolerance can be induced by ‘‘signal 1 without signal 2’’ (Schwartz, 1990; Huppa et al., 2003; Jenkinson et al., 2005; Lin et al., 2005), it is possible that efalizumab-induced effects are caused by ‘‘signal 2 without signal 1’’. One model of T-cell activation via the immune synapse has LFA-1 serving to increase initial adhesion, with main signal transduction occurring via the TCR complex (Huppa et al., 2003). However, it is now clear that LFA-1 also is capable of directly transducing distinct activation signals. Hence, ligation of LFA-1 by efalizumab might deliver signals (signal 2) that make a T-cell hyporesponsive to later T-cell activation signals delivered via the TCR/CD3 complex (signal 1). 1188 Journal of Investigative Dermatology (2008), Volume 128
Broad re-organization of activation-controlling surface molecules on T cells is also produced by efalizumab. In vivo efalizumab administration leads to down-modulation not only of CD11a/CD18 as expected (Gottlieb et al., 2002; Vugmeyster et al., 2004), but also of multiple surface molecules, which may be involved in T-cell activation, including CD3, TCR, CD4, CD8, CD28, and the integrin VLA-4. This occurs in T cells both in the peripheral circulation as well as in the tissue, although the effect is more restricted in the tissue. Possible mechanisms of downregulation of such an extensive array of surface molecules include capping of surface molecules by efalizumab (Mortensen et al., 2005), alteration of cytoskeletal components that are involved in immune synapse formation such as talin, or incorporation of surface molecules into lipid rafts (Hogg et al., 2003; Kinashi, 2005). T-cell activation often leads to incorporation of integrins into lipid rafts and other membrane proteins may complex with the integrins for colocalization within the rafts (Meiri, 2005). There is now evidence that efalizumab clearance is, at least partially, mediated by internalization via CD11a, which supports such a hypothesis (Coffey et al., 2005). These data provide insight into the mechanism of action of efalizumab contributing to its effect on T-cell-mediated autoimmune diseases such as psoriasis. Moreover, there are interesting conclusions to be made with regards to the response to self- and foreign neo- and recall antigens. However, the mechanism of downregulation of these surface molecules is difficult to assess because we are unable to reproduce these findings in vitro by incubation of peripheral blood with efalizumab (unpublished data). If efalizumab induced specific antigen-dependent anergy, it could be
E Guttman-Yassky et al. Hyporesponsiveness of T Cells by Efalizumab Therapy
speculated that long-lasting tolerance even beyond efalizumab presence could be induced. Our data do not support this hypothesis, but rather suggest a transient decrease in T-cell activation, which is reversible by IL-2 exposure or by antibody withdrawal. Indeed, in a recent human trial (Genentech ACD2244g) investigating the immune response to model antigens during efalizumab in psoriasis patients, T-cell-dependent antibody responses to the specific model neoantigen bacteriophage fX174 were greatly reduced during treatment, but responses to repeated administration of the same antigen normalized after withdrawal without tolerance induction, contradicting previous data from murine studies (Fischer et al., 1986; Benjamin et al., 1988; Krueger et al., 2005) ( JG Krueger et al. Effect of therapeutic integrin (CD11a) blockade with efalizumab on immune responses to model antigens in humans. Manuscript submitted for J Immunol, 2007). Taken together, these data emphasize the extent to which T-cell responses are blocked in efalizumab-treated patients and are in good agreement with the results of this study. We observed that T-cell responses to CD3/CD28 co-ligation were less suppressed than cell activation via CD3 or CD2 alone, suggesting that secondary immune responses in efalizumab-treated patients could be less impacted by treatment. Consistent with this observation, the human immunization trial demonstrated only a moderate reduction, but not abrogation, of responses to skin recall antigens measured by delayed-type hypersensitivity testing. The unique set of cellular effects seen with efalizumab is clearly distinct from the genetic defects of b2-integrins in humans (leukocyte adhesion deficiency) or mice (Hogg and Bates, 2000; Shaw et al., 2001; Bunting et al., 2002). Overall we have observed unexpected cellular immune effects of LFA-1 antagonism, which are different from those reported in any model system, and could not have been predicted from animal studies. Given the emerging role of the IL-17 pathway in psoriasis (Lowes et al., 2007), future studies to evaluate the therapeutic effects of efalizumab on this pathway will be considered. MATERIALS AND METHODS Study design Fourteen patients (11 males, 3 females, ages 28–59 years, median 48 years) and 17 patients (4 males, 13 females, ages 29–71, median 47 years) with moderate to severe psoriasis were enrolled in two consecutive studies, study 1 and study 2, respectively, which were approved by the Rockefeller University Hospital Institutional Review Board. All patients were consented and the study was conducted according to the Declaration of Helsinki Principles. The patients were treated with weekly, subcutaneous efalizumab at 1 mg kg1 per week for 12 weeks, and followed for an additional 12 weeks. A range of clinical responses to efalizumab therapy was observed, from complete clearing of psoriasis to minimal improvement. One and two patients withdrew from the first and second studies, respectively, and did not have final biopsies. Five healthy volunteers (2 males, 3 women, ages 20–68, median 46 years) were also enrolled for assessing the dose-dependent activation response to efalizumab treatment in vitro. Response was determined both histologically and
clinically, as previously described (Wittkowski, 2004; Lowes et al., 2005). In study 1, 11 patients were categorized as responders and two as non-responders. In study 2, 11 patients were categorized as responders and four as non-responders.
Skin biopsies Skin biopsies were obtained at baseline and 2 weeks after the first dose and processed as previously described (Ferenczi et al., 2002). Briefly, skin samples were cut into small pieces and placed in 5 mM EDTA (Fisher Scientific, Pittsburgh, PA) acid at 4 1C (pH 7) for 4 hours with continuous gentle agitation. The resultant cell suspension was filtered through a nylon mesh filter, centrifuged, washed with phosphate-buffered saline, and resuspended at 105 cells ml1 for staining and analysis by FACS.
Peripheral blood samples Peripheral blood draws were taken at baseline and at either 2, 12, and 24 weeks (study 1) or at 24 hours, 2, 6, and 12 weeks (study 2). Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized samples by using standard Ficoll–Hypaque (Pharmacia Biotech, Uppsala, Sweden) density-gradient sedimentation. For study 1, whole blood was used for activation assay, whereas PBMCs were used for T-cell phenotype by FACS, with the exception of CD11a expression for which both whole blood and PBMCs were analyzed. For study 2, PBMCs were used for activation assays, FACS, and Ca2 þ flux experiments, whereas whole blood was used to determine cytokine production and T-cell phenotype by FACS.
Whole blood and PBMC activation assays Heparinized whole blood was incubated with either 1 mg ml1 of mouse anti-CD3 antibody (clone UCHT1; PharMingen, San Diego, CA) overnight at 37 1C in the presence of varying dilutions (0.0001–100 mg ml1) of efalizumab (in vitro assay; Figure 1a); 1.6–1,000 ng ml1 of mouse anti-CD3 antibody (clone UCHT1) overnight at 37 1C (ex vivo assay; Figure 2a); or 25 ng ml1 PMA, 2 mg ml1 ionomycin, and 10 mg ml1 brefeldin A for 4 hours at 37 1C (ex vivo assay; Figure 2e and f). Upon red blood cell lysis (168 mM NH4Cl, 10 mM KHCO3, 0.089 mM Na4EDTA), cells were stained with anti-CD2 FITC and anti-CD69 phycoerythrin (PE)-cy5 (Becton Dickinson, San Jose, CA (BD)). Geometric mean fluorescent intensity (GMFI) or MFI of CD69 on CD2 þ lymphocytes was determined by FACS analysis for the in vitro assay (Figure 4a) or ex vivo assay (study 1; Figure 4a, e, f), respectively. For the in vitro assay, percent inhibition was calculated. The IC50 was calculated using a four-parameter curve fit in KaleidaGraph Software. For study 2 (Figure 1c and d), PBMCs were activated for 4 hours at 37 1C, using 1 mg ml1 of the OKT3 antibody (anti-CD3) (OrthoBiotech, Bridgewater, NJ), 0.5 mg ml1 T11.1 þ 0.5 mg ml1 T11.2 (anti-CD2) (Immunotech, Cedex, France), or anti-CD3/CD28 Dynabeads (Dynal Biotech, Invitrogen Corp., Carlsbad, CA) (three beads per T cell) in the presence of 10 mg ml1 brefeldin A (Sigma Aldrich, St Louis, MO). After removing Dynabeads on Dynal magnetic particle concentrator, PBMCs were washed with FACSWash (2% fetal bovine serum in phosphatebuffered saline, 0.1% sodium azide (Sigma Aldrich)), and permeabilized using FACSPerm (BD). MFI of CD69 on CD3 lymphocytes (with anti-CD3 clone SK7 that does not directly compete with OKT3 for T-cell binding) was determined by FACS analysis. Two out of 14 and two of 17 patients in study 1 and study 2, respectively, were excluded www.jidonline.org 1189
E Guttman-Yassky et al. Hyporesponsiveness of T Cells by Efalizumab Therapy
in calculations of the mean, as they had baseline (day 0) T-cell activation below the acceptance criteria (less than 15% CD69 T cells) or had missing samples.
Peripheral blood lymphocyte proliferation assay using healthy volunteers Cell proliferation was determined using 3H-labeled thymidine incorporation as previously described (van Kooyk et al., 1996; Khanna and Hosenpud, 1999). Briefly, 96-well plates were coated for 1 hour at 37 1C with suboptimal concentration of anti-CD3 (T3b, 0.07 mg ml1) or anti-CD3 (0.07 mg ml1) þ anti-CD28 (0.125 mg ml1), followed by goatanti-human Fc (400 ng well1, 1 hour at 37 1C), 1% BSA (100 ml well1, 30 minutes at 37 1C), and efalizumab (100 ng well1, 1 hour at 37 1C), as described previously (van Kooyk et al., 1996). Resting peripheral blood lymphocytes were added (100 000 cells well1) and cultured for 3 days, as described previously (van Kooyk et al., 1996). On day 3, cells were pulsed for 16 hours with 3H-labeled thymidine (1.52 TBq mmol1, 0.5 mCi well1; Amersham Corp., Arlington Heights, IL), and uptake was quantified to measure cell proliferation in the presence of either efalizumab, MHM24 (efalizumab parent monoclonal murine non humanized antibody), or the Fab fragment of efalizumab (all at 0.07 mM, from Genentech Inc., South San Francisco, CA).
Whole-blood cytokine analysis Heparinized whole blood was activated for 4 hours using either antiCD3 (1 mg ml1 OKT3), anti-CD2 (0.5 mg ml1 T11.1 þ 0.5 mg ml1 T11.2), anti-CD3/CD28 Dynabeads, or PMA/ionomycin (25 ng ml1 PMA, 2 mg ml1 ionomycin), and 10 mg ml1 brefeldin A. Samples were treated with 2 mM EDTA for 10 minutes at room temperature (RT). Red blood cells were then lysed with FACSLyse solution (BD) and centrifuged and washed to obtain PBMCs, which were frozen in media containing RPMI medium 1640 (Gibco/BRL, Rockville, MD), with 1 mM HEPES buffer (Sigma Aldrich), 0.1% gentamycin (Gibco/ BRL), and 5% normal human serum (C-Six Diagnostic, Germantown, WI) with 10% dimethyl sulfoxide. Unactivated controls were processed in parallel. Before cytokine staining, PBMCs were thawed, washed, and permeabilized with FACSPerm. The fraction of IFNg-producing T cells (SK7 þ lymphocytes) was determined by flow cytometry. Percent change in IFNg-producing ability was calculated.
FACS analysis FACS was used to analyze heparinized whole blood, PBMCs, or single cell suspensions from skin biopsies. When whole blood was used, cells were stained with antibodies for 10 minutes at RT, lysed with FACSLyse for 10 minutes at RT, and washed with FACSWash. PBMCs were stained for 15 minutes at RT. The following antibodies were used: CD18 (PE), CD3 (PerCP), CD2 (FITC), TCR (FITC), CD49d (PE), CD29 (FITC), CD4 (FITC), CD8 (Per) CD25 (PE), CD28 (PE), CD69 (FITC), IFN-g (FITC), and IL-2 (FITC) (BD). Staining for CD11a was performed with two different anti-CD11a mAbs: efalizumab–FITC (Genentech, custom design) and clone 25.3-FITC (Immunotech). Appropriate IgG isotype controls were added (BD). Cells were washed with FACSWash and resuspended in 1.3% formaldehyde (Fisher Scientific) in FACSWash. Samples were analyzed with fourcolor staining using a FACSCalibur Flow Cytometer and CELLQuest software after calibration with CaliBRITE beads and FACSComp software (BD). 1190 Journal of Investigative Dermatology (2008), Volume 128
Ca þ 2 flux PBMCs (107 per milliliter) were loaded with 2 mM Indo-1 (Molecular Probes, Invitrogen Corp., Carlsbad, CA) for 30 minutes at 37 1C in RPMI. The cells were washed with warm RPMI, resuspended in RPMI at RT, and acquired at baseline using a BD LSR flow cytometer. OKT3 (anti-CD3) (OrthoBiotech) activator was added and after an additional 20 seconds, a cross-linking mouse IgG2a was added. The Ca flux was measured as the ratio between bound and unbound Indo-1 versus time.
Statistical analysis Student’s t-test and two-sided sign test were performed, with the assumption of equal probability of an increase or decrease from day 0 value. Significance was accepted as Po0.05. CONFLICT OF INTEREST YV and SB were, and WD, BH, and KH are, employed by Genentech Inc. JGK has served as a consultant for Genentech. All other authors have no conflicting financial interests.
ACKNOWLEDGMENTS This research was supported in part by a General Clinical Research Center grant (M01-RR00102) from the National Center for Research Resources at the National Institutes of Health. JGK is supported by the following National Institutes of Health (NIH) grants: R01 AI-49572 and AI-49832. MAL is the recipient of the NIH/NIAMS grant K23 AR052404-01A1.
SUPPLEMENTARY MATERIAL Figure S1. Anti-CD3-mediated intracellular Ca2 þ flux response is normal in T cells from efalizumab-treated patients. Figure S2. Anti-CD3 and anti-CD2-mediated intracellular Ca2 þ flux response is normal in T cells from efalizumab-treated patients.
REFERENCES Appleman LJ, Boussiotis VA (2003) T cell anergy and costimulation. Immunol Rev 192:161–80 Benjamin RJ, Qin SX, Wise MP, Cobbold SP, Waldmann H (1988) Mechanisms of monoclonal antibody-facilitated tolerance induction: a possible role for the CD4 (L3T4) and CD11a (LFA-1) molecules in self-non-self discrimination. Eur J Immunol 18:1079–88 Bohmig GA, Kovarik J, Holter W, Pohanka E, Zlabinger GJ (1994) Specific downregulation of proliferative T cell alloresponsiveness by interference with CD2/LFA-3 and LFA-1/ICAM-1 in vitro. J Immunol 152:3720–8 Bunting M, Harris ES, McIntyre TM, Prescott SM, Zimmerman GA (2002) Leukocyte adhesion deficiency syndromes: adhesion and tethering defects involving beta 2 integrins and selectin ligands. Curr Opin Hematol 9:30–5 Carlos TM, Harlan JM (1994) Leukocyte–endothelial adhesion molecules. Blood 84:2068–101 Coffey GP, Fox JA, Pippig S, Palmieri S, Reitz B, Gonzales M et al. (2005) Tissue distribution and receptor-mediated clearance of anti-CD11a antibody in mice. Drug Metab Dispos 33:623–9 Ferenczi K, Fuhlbrigge RC, Pinkus J, Pinkus GS, Kupper TS (2002) Increased CCR4 expression in cutaneous T cell lymphoma. J Invest Dermatol 119:1405–10 Ferrara JL, Mauch P, Van Dijken PJ, Crosier KE, Michaelson J, Burakoff SJ (1990) Evidence that anti-asialo GM1 in vivo improves engraftment of T cell-depleted bone marrow in hybrid recipients. Transplantation 49: 134–8 Fischer A, Durandy A, Sterkers G, Griscelli C (1986) Role of the LFA-1 molecule in cellular interactions required for antibody production in humans. J Immunol 136:3198–203
E Guttman-Yassky et al. Hyporesponsiveness of T Cells by Efalizumab Therapy
Gordon EJ, Myers KJ, Dougherty JP, Rosen H, Ron Y (1995) Both anti-CD11a (LFA-1) and anti-CD11b (MAC-1) therapy delay the onset and diminish the severity of experimental autoimmune encephalomyelitis. J Neuroimmunol 62:153–60
Scheynius A, Camp RL, Pure E (1993) Reduced contact sensitivity reactions in mice treated with monoclonal antibodies to leukocyte functionassociated molecule-1 and intercellular adhesion molecule-1. J Immunol 150:655–63
Gottlieb AB, Krueger JG, Wittkowski K, Dedrick R, Walicke PA, Garovoy M (2002) Psoriasis as a model for T-cell-mediated disease: immunobiologic and clinical effects of treatment with multiple doses of efalizumab, an anti-CD11a antibody. Arch Dermatol 138:591–600
Schwartz RH (1990) A cell culture model for T lymphocyte clonal anergy. Science 248:1349–56
Haque S, Dumon H, Haque A, Kasper LH (1998) Alteration of intracellular calcium flux and impairment of nuclear factor-AT translocation in T cells during acute Toxoplasma gondii infection in mice. J Immunol 161:6812–8 Hogg N, Bates PA (2000) Genetic analysis of integrin function in man: LAD-1 and other syndromes. Matrix Biol 19:211–22
Schwartz RH (2003) T cell anergy. Annu Rev Immunol 21:305–34 Shaw JM, Al-Shamkhani A, Boxer LA, Buckley CD, Dodds AW, Klein N et al. (2001) Characterization of four CD18 mutants in leucocyte adhesion deficient (LAD) patients with differential capacities to support expression and function of the CD11/CD18 integrins LFA-1, Mac-1 and p150,95. Clin Exp Immunol 126:311–8
Hogg N, Laschinger M, Giles K, McDowall A (2003) T-cell integrins: more than just sticking points. J Cell Sci 116(Part 23):4695–705
Stewart MP, Cabanas C, Hogg N (1996) T cell adhesion to intercellular adhesion molecule-1 (ICAM-1) is controlled by cell spreading and the activation of integrin LFA-1. J Immunol 156:1810–7
Huppa JB, Gleimer M, Sumen C, Davis MM (2003) Continuous T cell receptor signaling required for synapse maintenance and full effector potential. Nat Immunol 4:749–55
Szegedi A, Aleksza M, Gonda A, Irinyi B, Sipka S, Hunyadi J et al. (2003) Elevated rate of Thelper1 (T(H)1) lymphocytes and serum IFN-gamma levels in psoriatic patients. Immunol Lett 86:277–80
Isobe M, Yagita H, Okumura K, Ihara A (1992) Specific acceptance of cardiac allograft after treatment with antibodies to ICAM-1 and LFA-1. Science 255:1125–7
Tanaka Y, Kobayashi K, Takahashi A, Arai I, Higuchi S, Otomo S et al. (1993) Inhibition of inflammatory liver injury by a monoclonal antibody against lymphocyte function-associated antigen-1. J Immunol 151: 5088–95
Jenkinson SR, Williams NA, Morgan DJ (2005) The role of intercellular adhesion molecule-1/LFA-1 interactions in the generation of tumorspecific CD8+ T cell responses. J Immunol 174:3401–7 Kanner SB, Grosmaire LS, Ledbetter JA, Damle NK (1993) Beta 2-integrin LFA-1 signaling through phospholipase C-gamma 1 activation. Proc Natl Acad Sci USA 90:7099–103 Khanna AK, Hosenpud JD (1999) Cyclosporine induces the expression of the cyclin inhibitor p21. Transplantation 67:1262–8 Kinashi T. (2005) Intracellular signalling controlling integrin activation in lymphocytes. Nat Rev Immunol 5:546–59 Krueger JG, Ochs H, Patel P, Gilkerson E, Dummer W (2005) Impact of efalizumab T cell modulation on immune response in psoriasis patients. J Invest Dermatol 124:A44 Lebwohl M, Tyring SK, Hamilton TK, Toth D, Glazer S, Tawfik NH et al. (2003) A novel targeted T-cell modulator, efalizumab, for plaque psoriasis. N Engl J Med 349:2004–13 Lin J, Miller MJ, Shaw AS (2005) The c-SMAC: sorting it all out (or in). J Cell Biol 170:177–82 Lowes MA, Bowcock AM, Krueger JG (2007) Pathogenesis and therapy of psoriasis. Nature 445:866–73 Lowes MA, Turton JA, Krueger JG, Barnetson RS (2005) Psoriasis vulgaris flare during efalizumab therapy does not preclude future use: a case series. BMC Dermatol 5:9 Macian F, Im SH, Garcia-Cozar FJ, Rao A (2004) T-cell anergy. Curr Opin Immunol 16:209–16 Meiri KF (2005) Lipid rafts and regulation of the cytoskeleton during T cell activation. Philos Trans R Soc Lond B Biol Sci 360:1663–72 Moriyama H, Yokono K, Amano K, Nagata M, Hasegawa Y, Okamoto N et al. (1996) Induction of tolerance in murine autoimmune diabetes by transient blockade of leukocyte function-associated antigen-1/intercellular adhesion molecule-1 pathway. J Immunol 157:3737–43 Mortensen DL, Walicke PA, Wang X, Kwon P, Kuebler P, Gottlieb AB et al. (2005) Pharmacokinetics and pharmacodynamics of multiple weekly subcutaneous efalizumab doses in patients with plaque psoriasis. J Clin Pharmacol 45:286–98
van Kooyk Y, Binnerts ME, Edwards CP, Champe M, Berman PW, Figdor CG et al. (1996) Critical amino acids in the lymphocyte function-associated antigen-1 I domain mediate intercellular adhesion molecule 3 binding and immune function. J Exp Med 183:1247–52 Van Seventer GA, Bonvini E, Yamada H, Conti A, Stringfellow S, June CH et al. (1992) Costimulation of T cell receptor/CD3-mediated activation of resting human CD4+ T cells by leukocyte function-associated antigen-1 ligand intercellular cell adhesion molecule-1 involves prolonged inositol phospholipid hydrolysis and sustained increase of intracellular Ca2+ levels. J Immunol 149:3872–80 van Seventer GA, Newman W, Shimizu Y, Nutman TB, Tanaka Y, Horgan KJ et al. (1991) Analysis of T cell stimulation by superantigen plus major histocompatibility complex class II molecules or by CD3 monoclonal antibody: costimulation by purified adhesion ligands VCAM-1, ICAM-1, but not ELAM-1. J Exp Med 174:901–13 Van Seventer GA, Shimizu Y, Horgan KJ, Shaw S (1990) The LFA-1 ligand ICAM-1 provides an important costimulatory signal for T cell receptormediated activation of resting T cells. J Immunol 144:4579–86 Vugmeyster Y, Kikuchi T, Lowes MA, Chamian F, Kagen M, Gilleaudeau P et al. (2004) Efalizumab (anti-CD11a)-induced increase in peripheral blood leukocytes in psoriasis patients is preferentially mediated by altered trafficking of memory CD8+ T cells into lesional skin. Clin Immunol 113:38–46 Werther WA, Gonzalez TN, O’Connor SJ, McCabe S, Chan B, Hotaling T et al. (1996) Humanization of an anti-lymphocyte function-associated antigen (LFA)-1 monoclonal antibody and reengineering of the humanized antibody for binding to rhesus LFA-1. J Immunol 157: 4986–95 Willenborg DO, Staykova MA, Miyasaka M (1996) Short term treatment with soluble neuroantigen and anti-CD11a (LFA-1) protects rats against autoimmune encephalomyelitis: treatment abrogates autoimmune disease but not autoimmunity. J Immunol 157:1973–80 Wittkowski KM (2004) Effects and non-effects of paired identical observations in comparing proportions with binary matched-pairs data. Stat Med 23:2317
www.jidonline.org 1191
