
	Version PDF

Profiling lymphocyte subpopulations in peripheral blood under efalizumab treatment of psoriasis by multi epitope ligand cartography (MELC) robot microscopy

European Journal of Dermatology. Volume 16, Number 6, 623-35, November-December 2006, Investigative report

DOI : 10.1684/ejd.2006.0005

Summary

Author(s) : Bernd Bonnekoh, Yanina Malykh, Raik Böckelmann, Sebastian Bartsch, Ansgar J Pommer, Harald Gollnick , Clinic for Dermatology and Venereology, Otto-von-Guericke-University, Magdeburg, Leipziger Str. 44, D-39120 Magdeburg, GermanyFax: (+49) 391 67 15283, MelTec GmbH & Co. KG, SkinSysTec GmbH, ZENIT Technology Park, Leipziger Str. 44, D-39120 Magdeburg, Germany.

Summary : CD11a-blocking efalizumab has recently been approved as a systemic treatment of moderate to severe chronic plaque psoriasis. When treating 6 psoriasis patients with efalizumab over 12 weeks in the present study, we observed an overall good tolerability and 5 treatment responders characterized by a decrease of PASI from 21.3 ± 5.4 to 3.9 ± 0.6. The accompanying significant increase of peripheral blood lymphocytes from 1.9 ± 0.7 to 4.3 ± 1.0 × 10 ⁹/L (p <\; 0.05) was analyzed by multi epitope ligand cartography (MELC) robot microscopy. Thereby a high-dimension simultaneous multiplex immunophenotyping was pursued using 39 fluorophore-labeled antibodies including labeled efalizumab and 3 other affinity reagents such as lectins. Due to efalizumab treatment there was a substantial decrease of the cellular expression of CD11a (detected by mab clone 25.3.1) and efalizumab binding sites (EfaBSs). This was paralleled by an increase of the number of EfaBS ^– and EfaBS ⁺ lymphocytes by a factor of 2.4× and 2.2×, respectively. The latter effect was mainly derived from a subpopulation showing a low degree of EfaBS expression. Efalizumab treatment led furthermore to an increase of the numbers of CD3 ⁺, CD4 ⁺, CD8 ⁺, CD44 ⁺, CD45 ⁺, CD45R0 ⁺, CD45 RA ⁺, CD52 ⁺, CD58 ⁺, CD247 ⁺, HLA-DR ⁺ and Sambucus nigra lectin-reactive lymphocytes (by factors from 2.0 to 3.3×). In terms of a combinatorial molecular phenotype we identified a CD3 ⁺/CD4 ⁺/CD44 ⁺/CD52 ⁺ lymphocyte subpopulation which accumulated most predominantly from 0.824 ± 0.270 × 10 ⁹/L up to 1.616 ± 0.152 × 10 ⁹/L under efalizumab treatment (p <\; 0.01). Thus, the current study extends the knowledge of efalizumab-dependent perturbations of recirculating blood lymphocyte subpopulations in psoriasis patients.

Keywords : CD11a, Biologicals, Proteomics, MELC Toponomics, Multiplex Fluorescence Immunophenotyping, Cytometry

Pictures

ARTICLE

Auteur(s) : Bernd Bonnekoh¹, Yanina Malykh², Raik Böckelmann¹, Sebastian Bartsch², Ansgar J Pommer^2,3, Harald Gollnick¹

¹Clinic for Dermatology and Venereology, Otto-von-Guericke-University, Magdeburg, Leipziger Str. 44, D-39120 Magdeburg, GermanyFax: (+49) 391 67 15283
²MelTec GmbH & Co. KG
³SkinSysTec GmbH, ZENIT Technology Park, Leipziger Str. 44, D-39120 Magdeburg, Germany

accepté le 2 Août 2006

Psoriasis represents a chronic relapsing skin disease affecting 2-3% percent of the Caucasian population. Its pathomechanism is dominated by blood pool-derived T lymphocytes which are activated in the epidermo-papillary skin compartment. This leads via TH1 reactions to endothelium activation, hyperproliferation as well as disturbed differentiation of keratinocytes which are reflected histologically by acanthosis and hyperparakeratosis and clinically by typical erythro-squamous lesions [1].In detail, it is currently still a matter of debate, whether these vicious circle-like pathomechanisms are driven by putatively existing auto-reactive T cells and auto-antigens, such as keratin 17 [2, 3], or autoimmunity-mimicking reactions critically involving e.g. NK/T cells [4]. By the known rules of MHC restriction of antigen presentation, this existence of psoriasis autoantigens would plausibly explain the phenomena of disease familiarity and its known coupling to distinct HLA alleles, such as cw6, especially in type I psoriasis with an early manifestation before the age of 40 years [5]. Moreover, autoantigen cross-reactivity with e.g. streptococci antigens might explain the clinical observation of induction and aggravation of psoriasis manifestations through bacterial infections, apart from possibly also existing causative superantigen mechanisms.Interestingly enough, psoriasis lesions may be also induced by rather unspecific inflammatory stimuli such as minor skin scratch trauma in terms of the Köbner phenomenon. These observations led to the hypothesis of an autoimmune loop or positive feedback mechanism directly intercorrelating the existence of T cell-dependent inflammation and autoantigen expression, or the absence of both of them, respectively [6, 7]. The latter would explain that psoriatic involved skin may be returned to a restrictedly stable state of clinically normal, healthy skin by appropriate treatment modalities. However, such a cure of psoriasis disease will mostly last for only a rather limited time of weeks or months after treatment withdrawal.Faced with this complex pathogenetic background, psoriasis is perceived as a disease in which a variety of treatment modalities may be highly effective although partly interfering with rather divergent primary cellular and molecular targets, connected, however, by this loop-like interdependence. The conventional armamentarium of antipsoriatic drugs comprises [6] steroids, vitamin D, its analogues (calcipotriol, tacalcitol), anthralin and tazarotene as topical agents, and methotrexate, cyclosporine, fumarates and acitretin as systemic treatment modalities. Additionally UV treatments of psoriasis such as PUVA, UVB or balneophototherapy have also been widely used. However, application of all these treatment modalities are limited by well known organ toxicities especially in a long term treatment setting [6].During the last few years psoriasis treatment has altered, due to a large impact from a new class of drugs, i.e. biologicals [8, 9], which are able to block the pathogenetic relevant positive feedback cycle of psoriasis at very precisely defined vertex points. In detail, besides i) TNF-α as a TH1 target cytokine for infliximab [10] or etanercept [11] and ii) T-cellular CD2 targeted by alefacept (containing the ligand binding site of LFA3/CD58) [12], it is especially iii) CD11a (the α-subunit of LFA-1) which has been identified as the most relevant molecule to be blocked by efalizumab, another potent antipsoriatic biological drug [13-16]. These biologicals hold the promise of a long term psoriasis treatment strategy with an advantageous benefit/risk ratio.Subcutaneously administered efalizumab, as with systemic antipsoriatic treatments in general, first passes the peripheral blood pool in order to exert its drug activity. The current study focuses on the dissection of efalizumab effects on peripheral blood lymphocyte subpopulations. These were analyzed by the innovative multi-epitope ligand cartography (MELC) robot microscopy [17, 18]. This technique allows, in an unprecedented manner, for the monitoring of single cell-related combinations of positive and negative expressions of binding sites for high numbers of divergent affinity reagents (e.g. epitopes detected by corresponding antibodies) critically involved in lymphocyte function. The highly complex results of our study contribute to the better understanding of efalizumab’s mode of antipsoriatic action.

Material and methods

Patients, efalizumab treatment and collection of blood specimens

Adult psoriasis patients were included in the study, after giving informed consent and following a study protocol approved by the ethics committee of the Otto-von-Guericke-University, Magdeburg, Germany. We report data as part of a local amendment protocol of the Magdeburg study site to an international clinical trial entitled: “Raptiva™ study 25300 – A multicentre, open label IIIb/IV study of subcutaneously administered efalizumab in the treatment of adult patients with moderate to severe chronic plaque psoriasis who have failed to respond to, or who have a contraindication to, or are intolerant of other systemic therapies including cyclosporine, methotrexate and PUVA.”

A total of 8 psoriasis patients were included, of whom one patient turned out to be a screening failure. One patient was under methotrexate treatment when he entered the efalizumab study treatment. For statistical reasons of low, incoherent case numbers, these two patients were excluded completely from the current study which aimed at profiling in-situ-proteomics of peripheral blood lymphocytes under efalizumab treatment. The remaining six patients did not receive a systemic antipsoriatic treatment for at least 4 weeks before receiving efalizumab as study medication. These patients had not been treated by any biological drug before.

Drug dosage was 0.7 mg efalizumab per kg body weight s.c. 1× in the first week, and 1.0 mg per kg body weight s.c. 1× per week in the following 11 weeks. Screening, treatment and follow-up of these patients were performed from May 2005 to January 2006. For editorial reasons the clinical data of the patient subcohort treated at the Magdeburg study site is only presented in summary, since these data are part of the above mentioned, superimposed, large multicentre trial to be published elsewhere.

For MELC analysis 10 mL of peripheral venous blood were drawn in pre-heparinized tubes from the study participants immediately before and after 12 weeks of efalizumab treatment. Seven blood donors not affected by any systemic inflammatory disease served as appropriate healthy controls.

Sample preparation

Peripheral blood mononuclear cells (PBMC) were isolated on a Ficoll gradient (PAA Laboratories GmbH, D-35091 Cölbe, Germany; ( figure 1 )), and washed three times with RPMI-1640 medium (PAA Lab). 45,000 cells were applied on a cover slip, and dried at room temperature. The sample was fixed for 10 sec in acetone, dried at room temperature and snap frozen in liquid nitrogen-cooled 2-methylbutan (PAA Lab). Samples were kept on storage at – 20 °C for several days, or at – 80 °C for longer intervals until use.

In immediate preparation for subsequent MELC analysis the samples were submerged for 10 min in – 20 °C acetone, afterwards air dried for 10 min at RT and rehydrated for 5 min in phosphate buffered saline pH 7.4 (PBS; PAA Lab). The sample was incubated for 30 min with normal goat serum (1:30, PAA Lab), and rinsed with PBS.

MELC library

We used a MELC library of 41 fluorescence tags comprising antibodies, lectins and propidium iodide (PI) as a nucleic acid dye (table 1( Table 1 )). The appropriate working dilutions, fluorophore labels (fluorescein isothiocyanate (FITC) and phycoerythrin), incubation time (15 min) and positions within the MELC run had been established and validated in the course of systematic experiments based on conventional immunohistochemistry and FACS analysis.

Efalizumab, provided as a commercially available lyophilisate (Serono International S.A., Switzerland) was labeled with FITC and integrated into the MELC library, as described previously [18].
Table 1 MELC library of 41 informative fluorescence tags and IgG, the latter used for blocking of unspecific binding

Binding Site (Epitope)	Clone etc. & Source	Dilution, Label	MELC Run Pos.	Binding Site (Epitope)	Clone etc. & Source	Dilution, Label	MELC Run Pos.
CD2	39C1.5 ^a	1:10, FITC	4	CD52	YTH34.5 ^g	1:100, FITC	28
CD3	UCHT1 ^a	1:40, PE	15	CD54	84H10 ^a	1:10, PE	11
CD4	13B8.2 ^a	1:10, PE	10	CD56	N901 (NKH-1) ^a	1:20, PE	19
CD7	8H8.1 ^a	1:10, FITC	21	CD57	NC 1 ^a	1:30, FITC	27
CD8	B9.11 ^a	1:20, FITC	10	CD58	AICD58 ^a	1:20, FITC	7
CD11a	25.3.1 ^a	1:10, FITC	8	CD62L	SK11 ^e	1:10, PE	8
CD11b	VIM12/ ICRF44 ^b	1:30, PE	14	CD68	KP1 ^d	1:200, FITC	15
CD13	SJ1D1 ^a	1:10, FITC	20	CD79a	ZL7.4 ^g	1:50, FITC	14
CD15	AHN1.1 ^c	1:200, FITC	31	CD94	HP-3D9 ^c	1:40, PE	17
CD16	3G8 ^a	1:50, PE	5	CD138	B-B4 ^a	1:10, PE	4
CD20	B-Ly1 ^d	1:10, FITC	6	CD247	G3 ^h	1:10, FITC	23
CD26	L272 ^e	1:10, FITC	11	CLA	HECA-452^e	1:20, FITC	13
CD30	Ber-H2 ^d	1:10, FITC	26	EfaBS	Efalizumabⁱ	5 μg/mL, FITC	5
CD31	158-2B3 ^c	1:200, PE	18	HLA-DQ	SK 10 ^e	1:20, FITC	9
CD36	FA6-152 ^a	1:150, FITC	22	HLA-DR	Immu357 ^a	1:10, PE	9
CD38	T16 ^a	1:20, PE	12	IgG-BS	679.1Mc7 ^a	1:5, 1:10, 1:100 PE/FITC	1-3
CD44	J-173 ^a	1:60, FITC	19	KI-67	7B11 ^j	1:20, FITC	16
CD45	T29/33 ^d	1:10, FITC	25	MAA-BS	MAA ^k	1:80, FITC	30
CD45RA	ALB11 ^a	1:10, FITC	24	Nucl. Acids	PI ^l	1:10000	32
CD45R0	UCHL1 ^f	1:20, FITC	17	SNA-BS	SNA ^k	1:24, FITC	29
CD49d	44H6 ^g	1:20, FITC	12	TIA-1	2G9 ^a	1:50; PE	16

MELC robot technology: basic set-up

MELC robot technology (US-patent 6,150,173) involved distinct hardware and software components, as described earlier [17, 18]. A slide with a blood preparation was positioned onto the stage of an inverted wide-field fluorescence microscope (Leica DM IRE2; 20× air objective lens NA 0.7), equipped with fluorescence filters for FITC and phycoerythrin ( (figure 1) ). By a robotic process of on/off-pipetting, the specimen was incubated with predetermined fluorescence tags (table 1) and rinsed with wash solutions under temperature control. The phase contrast and fluorescence images were recorded by a cooled CCD camera (Apogee KX4, 1024 × 1024 pixels, 900 × 900 nm² per pixel), followed by soft bleaching (centered at 488 nm for FITC and at 546 nm for phycoerythrin). Recording of all image data and coordination of all system components were controlled by software developed by MelTec GmbH & Co. KG. All these processes (tag incubation and binding/ fluorescence detection/ soft bleaching) were part of a fully automated cycle repeated for any number of tag (incl. antibody) binding sites (incl. epitopes), respectively ( (figure 1) ). In each MELC cycle the simultaneous processing of a FITC- and a phycoerythrin-labeled tag was an option to save robot time. Two visual fields were recorded simultaneously in each MELC run. Unspecific tag binding was controlled by performing the three first MELC cycles with mouse IgG labeled to both FITC and phycoerythrin (table 1).

MELC data analysis

The computer platform of the MELC robot stored the phase contrast and raw fluorescence images for all tag binding sites and chosen visual fields. Image pre-processing comprised several steps ( (figure 1) ): The phase contrast images, taken directly before each fluorescence image of a given tag binding site, were used to overlay and merge the corresponding fluorescence images precisely, in a pixel-related manner, by determining the misalignment (cross correlation coefficient). Images were corrected for illumination faults using flat-field correction and for background with respect to unspecific IgG binding (MELC cycles 1-3).

Pre-processed image data were subjected to cell recognition as follows. Propidium iodide signal was used to detect a cell nucleus. A circle centred upon the nucleus with a maximum diameter of 15 pixels defined a foreground cell mask, corrected for the distance between centres of possibly overlapping neighboured cells. With this approach more than 99% of cells were recognized correctly ( (figure 1) ).

The expression of a tag binding site was measured to be positive in projection to a pixel when its fluorescence intensity was, with a probability of 99.7%, above background fluorescence. Cellular mean fluorescence intensity (MFI) was determined by the average of grey values of all pixels per cell. A cell was set positive when there was a defined number of positive pixels within its foreground mask ( (figure 1) ) which was a semi-automatic procedure based upon a tag binding site-specific 3D analysis correlating a) number of positive pixels per cell, b) cellular fluorescence intensity and c) cell frequencies (( figure 1 ), 3D graphic insert). Relative MELC data for the distribution of cell subpopulations were transformed to absolute values with regard to corresponding blood lymphocyte counts (in the dimension of × 10⁹/L blood).

The further analysis dealt with combinatorial molecular phenotype (CMP) motifs characterizing corresponding lymphocyte subpopulations. These CMP motifs are defined as the cell-related code of positive, negative and ambivalent expressions of tag binding sites (epitopes) in terms of an one/zero/wildcard ciphering (1/0/*). We used MelTec’s “MotifFinder” software package to search for lymphocyte subpopulations identified by CMP motifs, whose overall frequencies differed significantly in two samples, i.e. comparing the pre- versus post-stage in relation to efalizumab treatment. In detail, MotifFinder calculated the frequency of CMP motifs consisting entirely of wild cards except at four positions where this limit of search depth was imposed by computational resources.

Statistics included ANOVA multivariate analysis with post hoc Games-Howell-test and paired t-test, setting the p-value to < 0.05 (single and dual epitope analysis) or to < 0.01 (CMP motif analysis by MotifFinder), respectively. All statistical tests were performed by using a SPSS software package (version 11.0).

Results

Clinical outcome

All seven patients, having been screened successfully, terminated 12 weeks of efalizumab treatment. Overall tolerability of the treatment was good without any major or serious adverse events.

There were 6 patients who had undergone a wash-out of at least 4 weeks for any systemic psoriasis treatment before they received efalizumab. Based upon PASI evaluation there were 5 among these 6 psoriasis patients who experienced a satisfying treatment response. This overall clinical outcome was evidenced by a drop of the total PASI from 21.3 ± 5.4 (day 0) to 3.9 ± 0.6 (week 12) in the 5 efalizumab responders (data not shown). One patient did not respond satisfactorily, and even deteriorated, to be rescued successfully by subsequent cyclosporine treatment. The remaining patient who had been switched directly from MTX to efalizumab stayed under this treatment in a stable disease state.

Peripheral blood leukocyte counts

The conventional analysis of peripheral blood leukocyte counts revealed for the 5 efalizumab responders that there was a treatment-dependent increase of the total white blood cell count (table 2( Table 2 )). Breaking this effect down to granulocytes, monocytes and lymphocytes, the latter were shown to be most relevant, as evidenced by a significant increase from 1.9 ± 0.7 to 4.3 ± 1.0 × 10⁹/L (before vs. after treatment) going beyond the upper limit of the corresponding reference range (i.e. 1.0 – 4.0 × 10⁹/L, table 2). Thus, the following MELC data analysis was focused on the 5 efalizumab responders by generally gating upon lymphocytes excluding the CD3^–∩CD68⁺ monocyte fraction. The non-responder was mostly left out from our analysis for scientific statistical reasons of a low number event.
Table 2 Peripheral blood cell counts in psoriasis patients before and after 12 weeks of efalizumab treatment as compared to healthy controls. n = 6 patients having undergone a wash-out of at least 4 weeks for any systemic psoriasis treatment before efalizumab administration

Parameter (reference range)			White blood cells [× 10⁹/L] (3.9 – 10.3)	Granulocytes [× 10⁹/L] (1.9 – 8.1)	Monocytes [× 10⁹/L] (0.2 – 0.9)	Lymphocytes [× 10⁹/L] (1.0 – 4.0)
Condition		n
Healthy controls		7	7.0 ± 2.3	4.7 ± 2.1	0.5 ± 0.2	1.9 ± 0.5
Efalizumab responders	before treatment	5	7.4 ± 2.3	5.1 ± 2.4	0.4 ± 0.1	1.9 ± 0.7
after treatment	5	9.4 ± 0.8	4.7 ± 0.2	0.4 ± 0.2	4.3 ± 1.0*
Efalizumab non-responder	before treatment	1	7.6	5.4	0.7	1.5
after treatment	1	7.0	4.5	0.4	2.1

Expression of CD11a and efalizumab binding site

The first part of the MELC data analysis dealt with the cellular expression of CD11a as detected by clone 25.3.1 and efalizumab, both known to recognize divergent epitopes. When comparing the total of 5 efalizumab responders at the pre- versus the post-treatment state, we found a substantial decrease of the cellular mean fluorescence intensity (MFI) as derived from binding to either clone 25.3.1 or efalizumab (( figure 2A and B) ). To rule out the possibility that the binding site on CD11a as detected by clone 25.3.1 could have been merely blocked by the efalizumab administration for the treatment of psoriasis, we performed a series of in vitro experiments demonstrating that the preincubation of living PBMC with efalizumab at concentrations of 1 μg/mL, 0.1 mg/mL and 10 mg/mL did not impair the subsequent MELC-detection of CD11a by clone 25.3.1 as compared to control incubations with PBS or bovine serum albumin (data not shown).

This downward-shifting of the expression of CD11a and efalizumab binding site under treatment prompted us to differentiate cells with a high (hi) and low (lo) degree of efalizumab binding site (EfaBS) expression. The differentiation of EfaBS^lo and EfaBS^hi cells was based upon the selection of an additional threshold with regard to the distributions of the counts of positive pixels per cell ( (figure 2C) ).

Lymphocyte subpopulations defined by single epitope expression

The first level of the second part of the MELC data analysis allowed a precise analysis of the cell numbers of lymphocyte subpopulations as defined by the positive or negative expression of a single tag binding site (or epitope, respectively). Lymphocyte subpopulations as characterized by a positive single epitope expression were ranked according to increasing average cell numbers under the condition of the healthy controls ( (figure 3) ). For the total of the lymphocyte subpopulations defined by 41 tags of the MELC library we did not find any significant difference for the cell numbers in the comparison between i) psoriasis patients (i.e. efalizumab responders) before treatment and ii) the healthy controls. A significant increase of cell numbers was observed during efalizumab treatment for lymphocyte subpopulations (( figure 3 A and B )) showing a positive expression of CD2 (by a factor of 2.2× as related to the pre- and post-treatment averages), CD3 (2.4×), CD4 (2.0×), CD8 (2.6×), CD44 (2.2×), CD45 (2.2×), CD45R0 (2.1×), CD45RA (2.4×), CD52 (2.3×), CD58 (2.5×), CD247 (3.3×), HLA-DR (2.1×) and SNA-BS (2.2×).

Due to efalizumab treatment there was no significant change of the cell numbers for the lymphocyte subpopulation expressing CD11a as detected by clone 25.3.1 and EfaBS^hi, but a significant increase for lymphocyte subpopulations characterized by EfaBS^lo+hi (by a factor of 2.2×) and EfaBS^lo (5.4×).

Comparing efalizumab responders before vs. after treatment, a significant increase of the cell numbers of the EfaBS^neg lymphocyte subpopulation by a factor of 2.4× was observed ( (figure 3A) ). Noteworthy, among efalizumab responders and due to the treatment, there was a significant increase of the cell numbers of lymphocyte subpopulations as defined by the negative expression of the majority of the binding sites investigated, with the exception of only CD3^neg, CD44^neg, CD45^neg, CD52^neg, CD58^neg, CD138^neg, and CD247^neg (data not shown).

Corresponding relative data for the efalizumab treatment-dependent course of lymphocyte subpopulations defined by positive expression of a single epitope in relation to the total of lymphocytes (excluding the CD3^–∩CD68⁺ monocyte fraction) set to 100%, are shown in table 3( Table 3 ). These relative data seemed to be less informative than the corresponding absolute cell numbers.
Table 3 Relative data for lymphocyte subpopulations defined by single marker expression in relation to the total of lymphocytes (excluding the CD3^–∩CD68⁺ monocyte fraction) set to 100%. Results are presented in percentages (means ± standard deviations). Asterisks indicate a p-value < 0.05

Marker (Binding Site/ Epitope)	Efalizumab responders		Healthy controls (ctrl) [%]	ANOVA & Games-Howell-test
before treatment (pre) [%]	after treatment (post) [%]	pre vs. ctrl [p-value]	post vs. ctrl [p-value]	pre vs. post [p-value]
CD2	72.1 ± 8.1	70.8 ± 4.4	78.8 ± 8.5	0.384	0.138	0.953
CD3	67.2 ± 8.1	71.6 ± 6.4	74.6 ± 8.2	0.309	0.766	0.617
CD4	45.8 ± 5.4	40.1 ± 6.6	50.9 ± 8.1	0.424	0.070	0.341
CD7	45.0 ± 19.9	40.2 ± 11.3	44.2 ± 18.9	0.997	0.893	0.885
CD8	20.8 ± 4.0	24.4 ± 4.6	25.6 ± 6.6	0.299	0.930	0.417
CD11a	51.2 ± 29.5	23.3 ± 10.6	67.4 ± 16.4	0.545	0.001*	0.209
CD11b	8.4 ± 7.2	1.5 ± 0.5	8.6 ± 6.3	0.998	0.056	0.200
CD13	4.1 ± 3.9	2.9 ± 4.9	3.6 ± 4.6	0.978	0.968	0.908
CD15	4.7 ± 5.5	0.8 ± 0.5	5.3 ± 6.6	0.984	0.237	0.345
CD16	7.5 ± 4.9	1.3 ± 1.4	8.0 ± 7.0	0.984	0.096	0.095
CD20	16.1 ± 6.4	15.7 ± 5.0	10.1 ± 6.8	0.311	0.274	0.994
CD26	50.8 ± 8.0	32.6 ± 3.9	39.7 ± 11.6	0.175	0.337	0.010*
CD30	0.5 ± 0.4	0.2 ± 0.2	0.3 ± 0.2	0.563	0.791	0.371
CD31	4.5 ± 4.6	1.0 ± 0.6	1.9 ± 1.3	0.493	0.259	0.303
CD36	0.4 ± 0.3	0.4 ± 0.2	0.5 ± 0.2	0.925	0.761	0.986
CD38	17.7 ± 6.6	4.6 ± 2.0	13.9 ± 5.7	0.575	0.011*	0.021*
CD44	88.7 ± 7.8	89.6 ± 1.7	87.4 ± 7.2	0.958	0.733	0.964
CD45	93.6 ± 10.1	91.8 ± 14.2	94.8 ± 4.4	0.966	0.892	0.971
CD45RA	58.2 ± 6.9	63.6 ± 2.7	53.2 ± 16.9	0.763	0.309	0.321
CD45R0	34.2 ± 9.4	33.1 ± 3.2	40.3 ± 13.4	0.640	0.402	0.961
CD49d	41.8 ± 39.3	11.7 ± 8.4	36.3 ± 37.1	0.967	0.272	0.311
CD52	80.2 ± 11.0	83.4 ± 3.6	81.5 ± 10.4	0.976	0.897	0.815
CD54	5.3 ± 2.9	7.1 ± 4.4	4.1 ± 3.8	0.802	0.461	0.735
CD56	11.5 ± 4.5	8.7 ± 2.8	12.8 ± 4.2	0.867	0.153	0.494
CD57	7.9 ± 3.3	5.2 ± 2.8	15.5 ± 5.7	0.040*	0.006*	0.392
CD58	73.6 ± 37.6	90.3 ± 10.8	76.5 ± 32.4	0.989	0.570	0.634
CD62L	56.4 ± 26.3	59.3 ± 10.3	61.9 ± 8.8	0.896	0.894	0.971
CD79a	9.8 ± 5.9	9.2 ± 2.8	5.2 ± 2.9	0.315	0.091	0.976
CD94	16.4 ± 6.7	11.2 ± 5.9	22.6 ± 11.7	0.500	0.120	0.434
CD138	38.7 ± 21.4	39.1 ± 28.0	43.6 ± 22.2	0.921	0.952	1.000
CD247	48.2 ± 32.2	76.6 ± 15.8	39.4 ± 32.4	0.888	0.063	0.258
CLA	7.7 ± 4.1	4.9 ± 1.3	7.8 ± 8.4	0.999	0.645	0.380
EfaBS^lo+hi	74.6 ± 26.2	76.0 ± 7.4	80.3 ± 12.8	0.897	0.751	0.993
EfaBS^lo	19.4 ± 5.6	46.5 ± 4.1	29.8 ± 9.8	0.099	0.008*	0.000*
EfaBS^hi	55.2 ± 28.9	29.5 ± 10.4	50.5 ± 17.4	0.945	0.063	0.240
EfaBS^neg	25.4 ± 26.2	24.0 ± 7.4	19.7 ± 12.8	0.897	0.751	0.993
HLA-DQ	15.0 ± 3.7	15.0 ± 5.5	8.9 ± 3.0	0.041*	0.137	1.000
HLA-DR	20.0 ± 4.7	18.5 ± 3.5	15.4 ± 5.1	0.291	0.456	0.843
Ki67	33.0 ± 31.4	28.0 ± 13.7	33.6 ± 25.0	0.999	0.873	0.942
MAA-BS	40.5 ± 15.7	35.0 ± 10.7	22.8 ± 13.4	0.164	0.237	0.797
SNA-BS	88.8 ± 5.7	84.9 ± 9.3	75.6 ± 24.8	0.412	0.651	0.717
TIA-1	24.4 ± 10.5	15.5 ± 6.8	32.8 ± 7.3	0.329	0.005*	0.315

Hub co-expression analysis of lymphocyte subpopulations centered upon efalizumab binding site

The second level of MELC data analysis dealt with the dual co-expression of given epitopes with EfaBS and vice versa. Setting the total amount of lymphocytes expressing a defined single epitope to 100% it became possible to quantify the relative percentage co-expressing EfaBS. Under these conditions T cell subpopulations defined by CD2, CD3, CD4, CD8, CD45R0, CD45RA, and CD247 showed an average EfaBS co-expression ranging from 72 to 95% under the conditions of healthy controls as well as of psoriasis patients (i.e. efalizumab responders) before and after treatment (data not shown). In contrast, B cell subpopulations as detected by CD20 or CD79a expression did show an average EfaBS co-expression only in the range from 36 to 62%.

The cell numbers of lymphocyte subpopulations showing a distinct dual co-expression of EfaBS with another defined epitope is shown as a so-called hub diagram ( (figure 4) ). For the efalizumab responders there was, due to treatment, a significant increase of the cell numbers of EfaBS^lo+hi lymphocyte subpopulations showing co-expression with CD2 (by a factor of 2.1×), CD3 (2.2×), CD8 (2.4×), CD44 (2.1), CD58 (2.3×), CD247 (3.0×), CLA (2.0×) and HLA-DR (2.4×). Comparing i) healthy controls with ii) psoriasis patients (i.e. efalizumab responders) before treatment, we did not observe a significant difference for the cell numbers of lymphocyte subpopulations defined by any hub co-expression of EfaBS with a second epitope.

Characterization of lymphocyte subpopulations by combinatorial molecular phenotypes (CMPs)

At the third level of MELC data analysis lymphocyte subpopulations were defined and analyzed in depth employing the CMP parameter. In this context a CMP represents the cell-related code for the positive or negative expression (1/0 code) of the total of tag binding sites (incl. epitopes). Given the above mentioned MELC library of 41 tags and analyzing a defined number of 1,000 lymphocytes of a specimen, we found a corresponding CMP number of 551 for the healthy controls, and of 557 and 555 for the efalizumab responders before and after treatment, respectively. Even the most dominant lymphocyte subpopulation defined by a single distinct CMP reached only a frequency of 0.6%, 1.6% and 1.1% for healthy controls, efalizumab responders before and after treatment, respectively.

We then undertook a thorough statistical comparison of the pre- and post-treatment data for the absolute lymphocyte numbers of corresponding CMP subpopulations by a MotifFinder analysis employing a p-value < 0.01. This approach relied upon so-called CMP motifs defined by a 1/0 or * code representing the presence, absence or wild card ambivalence of the expression of the tag binding sites (incl. epitopes). The search revealed a list of 51 CMP motifs which showed a statistically significant difference in frequency between the pre- and post-treatment conditions. Noteworthy, in the majority of CMP motifs there was an increase in frequency during treatment with the exception of only 5 CMP motifs. Out of this data set we hereby present the most relevant CMP motifs ranked by the most prevalent average frequency before treatment (table 4( Table 4 )).

Moreover, as the most dominant CMP motif with 4 non-wildcard positions set (p < 0.01) we identified CD3⁺/CD4⁺/CD44⁺/CD52⁺ (visualized in ( figure 5 )). The number of the corresponding lymphocyte subpopulation showed a significant increase from 0.824 ± 0.270 × 10⁹/L to 1.616 ± 0.152 × 10⁹/L in the before versus after treatment comparison for the efalizumab responders.
Table 4 Lymphocyte subpopulations defined by so-called combinatorial molecular phenotype (CMP) motifs showing a significant difference in their cell number for efalizumab responders in the pre- versus post-treatment comparison. Ranked by declining frequencies at the stage before treatment, the most relevant CMP motifis are shown out of a MotifFinder search having revealed a total of 51 significant CMP motifs. Statistical comparison was perfomed by paired t-test (p < 0.01). The sequel of binding sites in the table follows that in the MELC process

CMP-motif of lymphocytes from efalizumab responders																																										Lymphocyte number				t-test p
no.	CD138	CD2	CD16	EfaBS	CD58	CD20	CD62L	CD11a	HLA-DR	HLA-DQ	CD4	CD8	CD54	CD26	CD38	CD49d	CLA	CD11b	CD79a	CD3	CD68	TIA-1	KI67	CD94	CD45R0	CD31	CD11a	CD56	CD44	CD13	CD7	CD36	CD247	CD45RA	CD45	CD30	CD57	CD52	SNA-BS	MAA-BS	CD15	Before treatment [x 10⁹/L]	After treatment [x 10⁹/L]	pre- post-ratio
1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0	*	1.046 ± 0.161	2.713 ± 0.547	2.6	↑	0.001
2	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	1	*	*	*	*	*	*	*	0.898 ± 0.315	2.357 ± 0.464	2.6	↑	0.000
3	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.849 ± 0.262	1.657 ± 0.155	2.0	↑	0.000
4	*	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.806 ± 0.290	3.260 ± 0.859	4.0	↑	0.005
5	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.767 ± 0.381	2.924 ± 0.328	3.8	↑	0.001
6	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.754 ± 0.142	1.648 ± 0.314	2.2	↑	0.002
7	*	*	*	*	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	0.734 ± 0.337	2.425 ± 0.555	3.3	↑	0.001
8	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	0	*	*	*	*	0.632 ± 0.204	1.536 ± 0.284	2.4	↑	0.000
9	*	*	*	*	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	0.601 ± 0.356	2.190 ± 0.471	3.6	↑	0.001
10	*	1	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	0.469 ± 0.130	1.061 ± 0.085	2.3	↑	0.001
11	*	1	*	*	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.439 ± 0.171	1.673 ± 0.364	3.8	↑	0.001
12	*	*	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	0.390 ± 0.119	1.063 ± 0.168	2.7	↑	0.003
13	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	0.365 ± 0.107	0.961 ± 0.198	2.6	↑	0.006
14	*	*	*	*	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.327 ± 0.122	1.690 ± 0.316	5.2	↑	0.001
15	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	1	*	*	*	*	*	*	*	*	0.289 ± 0.237	0.979 ± 0.131	3.4	↑	0.004
16	*	1	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	0.253 ± 0.196	0.856 ± 0.135	3.4	↑	0.001
17	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	0.249 ± 0.075	0.636 ± 0.134	2.6	↑	0.001
18	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	1	*	*	*	0.241 ± 0.209	0.854 ± 0.110	3.5	↑	0.003
19	*	*	*	*	1	*	0	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.241 ± 0.156	0.917 ± 0.192	3.8	↑	0.003
20	*	*	*	*	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.216 ± 0.066	0.693 ± 0.150	3.2	↑	0.000
21	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	0.200 ± 0.155	0.808 ± 0.097	4.0	↑	0.002
22	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	0	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.062 ± 0.035	0.246 ± 0.045	4.0	↑	0.001
23	*	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	1	*	1	*	0	*	*	*	*	*	*	*	*	0.047 ± 0.008	0.008 ± 0.008	5.9	↓	0.002
24	*	*	*	*	*	*	*	*	*	*	*	1	*	*	*	0	*	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0.046 ± 0.042	0.328 ± 0.086	7.1	↑	0.001
25	*	1	*	*	*	*	1	*	*	*	*	*	*	*	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0	*	*	*	*	*	*	*	*	0.043 ± 0.012	0.004 ± 0.004	10.8	↓	0.003
26	*	*	*	*	*	*	*	*	*	*	*	*	*	1	1	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	*	0	1	*	*	*	*	*	*	*	0.033 ± 0.012	0.001 ± 0.002	33.0	↓	0.005

Discussion

Efalizumab is a recombinant humanized monoclonal IgG1 antibody that targets psoriasis pathogenesis at multiple levels, i.e. by inhibition of initial T cell activation in the lymph nodes, preventing endothelium-directed binding of T cells, blocking their transendothelial migration and down-regulating their reactivation in the dermal and epidermal skin layers [13]. Thus, the peripheral blood pool may be considered as a decisive first pass compartment for s.c. administered efalizumab in psoriasis to be conceived as a skin disease but also an (auto-) immune system disorder [1].

Faced with this background, we hereby describe for the first time the application of innovative MELC robot technology for the monitoring of peripheral blood lymphocyte subpopulations under the conditions of a clinical treatment study. This methodology allowed us to decipher cell-based combinatorics of 41 binding sites of affinity reagents (mostly epitopes as detected by antibodies) which by far surpasses the scope of the currently most developed state-of-the-art in flow immunocytometry, with a maximum possible simultaneous detection of 17 colours [19]. In a concomitant study we recently applied MELC robot technology employing a library of 50 affinity reagents to the analysis of psoriatic skin tissue also with respect to the efalizumab binding site [18], extrapolating conventional immunohistology [20] to a multiplex dimension. The most decisive results of our current experimental findings embedded in a clinical context may be discussed as follows.

First, although this was only a small cohort of study patients we observed 5/6 psoriasis patients with no immediate preceding systemic antipsoriatic treatment to respond to 12 weeks of efalizumab treatment by an average PASI decrease from 21 down to 4, which is in accordance with data of earlier large multicentre studies [21]. When analyzing these 5 efalizumab responders with regard to blood cell counts over the 12 weeks of treatment, we observed an increase of the number of total leukocytes from 7.4 ± 2.3 to 9.4 ± 0.8 × 10⁹/L which was mainly due to a significant increase of the lymphocytes from 1.9 ± 0.7 to 4.3 ± 1.0 × 10⁹/L. Increase of total leukocytes by about 40% from baseline, little change in granulocytes and monocytes, and more than doubling of lymphocyte counts are well recognized phenomena from earlier studies, being explained as a reflectance of the blocked binding between T-cellular LFA-1 and endothelial ICAM-1 [13]. This was the given frame of our pilot cohort of 5 efalizumab responders, being well characterized in terms of evaluation by clinical aspects and routine laboratory parameters, when we set out for a MELC analysis of blood lymphocyte subpopulations.

Second, a major object of our work was to quantify the lymphocyte-directed binding of efalizumab integrated as a ligand into the MELC process. Thereby the novel MELC-based principle of a biological-drug-binding-biochip assay as recently established for skin tissue [18] was transferred to blood lymphocyte preparations. By analyzing lymphocyte subpopulations defined by the expression of a single epitope for the relative degree of co-location with EfaBS, a corresponding co-expression between 80 and 95% was observed for CD2⁺, CD3⁺, CD4⁺, CD8⁺, CD45RA⁺, CD45R0⁺ and CD247⁺ T lymphocytes and between 49 and 57% for CD20⁺ and CD79a⁺ B cells in the efalizumab responders before treatment. Interestingly enough, we did not find any significant differences in this rate of EfaBS⁺ co-location for any epitopes studied when comparing healthy controls with efalizumab responders before and after treatment, respectively. This points to a generally relatively stable overall constitutive expression of EfaBS in i) the majority of T lymphocytes and ii) about one half of B lymphocytes in peripheral blood. To some extent contradictorily, it had been stated by others that CD11a is expressed on all circulating lymphocytes [22, 23]. However, analyzing the gradual degree of EfaBS positivity and breaking it down to low as well as high abundant EfaBS expression gave the following more detailed information.

Third, comparing the efalizumab responders at the stage before and after treatment there was a substantial reduction of cellular expression levels for EfaBS and CD11a (as detected by mab clone 25.3.1) in lymphocytes. This is in good agreement with literature data describing a reduction in CD11a expression on circulating T lymphocytes to approximately 15-30% of baseline following efalizumab treatment of 1 mg/kg/week for 12 weeks [13]. The phenomenon has been named as “leukocyte CD11a saturation by efalizumab” and may persist even when a patient misses an efalizumab dose, as shown in an individual case [24]. During efalizumab treatment, we observed a significant increase of the total number of EfaBS⁺ lymphocytes by a factor of 2.2×, but not of the CD11a⁺ lymphocytes. The effect and divergence may be explained primarily by a concomitant significant increase of the number of EfaBS^lo lymphocytes (5.4×). And in this context it has to be underlined that efalizumab recognizes an epitope of CD11a distinct from that of mab clone 25.3.1 [23], which may show a lower affinity to CD11a than efalizumab.

Our observations are in good agreement with in vivo and in vitro studies, having shown that efalizumab induces CD11a receptor down-modulation from the plasma membrane of lymphocytes [25, 26], which has been explained by a CD11a-mediated internalization and lysosomal targeting of efalizumab as one of its clearing mechanisms [23]. Interestingly enough, under in vitro conditions this mechanism was dependent upon so-called cross-linking conditions involving the addition of an anti-mouse IgG antibody [23].

Fourth, in the group of efalizumab responders the 12 weeks of treatment led to a significant increase of CD3⁺, CD4⁺, CD8⁺, CD44⁺, CD45⁺, CD45R0⁺, CD45RA⁺, CD52⁺, CD58⁺, CD247⁺, HLA-DR⁺ and Sambucus nigra lectin-reactive lymphocytes, i.e. by factors from 2.0 to 3.3×. Thereby the maximum factor of 3.3× was observed for the lymphocyte subpopulation bearing the T cell receptor zeta chain as detected by the anti-CD247 antibody. Until now, to the best of our knowledge, the reactivity of peripheral blood lymphocytes to NeuAc-alpha2-6Gal-R-specific Sambucus nigra lectin [27] has not yet been investigated in psoriasis. Our results confirm and extend former blood data of flow immunocytometry from others that efalizumab treatment leads to an increase of i) T lymphocytes [28], and ii) CD8⁺ cells of the naive and memory phenotype [13]. And again in accordance with blood data of flow immunocytometry of others [28], we found a statistical trend for an increase of also B-lymphocytes and NK-cells (see data for CD20 and CD56 in ( figure 3A )). Thus, although not having had the opportunity in the current study to perform conventional flow immunocytometry in parallel (see below), our MELC data for single epitope expression fit well with literature data. A fact which we consider as an additional, at least indirect, external validation of our methodology.

Fifth, looking into the details of CMP motifs defining corresponding lymphocyte subpopulations by means of a MotifFinder strategy (p < 0.01) we identified a CD3⁺/CD4⁺/CD44⁺/CD52⁺ core motif which was the most dominant to separate the stages before (0.824 ± 0.270 × 10⁹/L) and after treatment (1.616 ± 0.152 × 10⁹/L) in the group of efalizumab responders. Therefore this CD4⁺ lymphocyte subpopulation expressing H-CAM (syn. CD44) critically involved in leukocyte rolling may be considered as predominantly accumulating in the peripheral blood under successful efalizumab treatment of psoriasis.

Our above mentioned hypothesis-free finding of an efalizumab treatment-dependent increase of the CD3⁺/CD4⁺/CD44⁺/CD52⁺ lymphocyte subpopulation may await future i) re-evaluation by MELC and ii) alternate methodological, hypothesis-based a posteriori validation by conventional flow immunocytometry in larger patient populations in a prospective comparative study design. It is noteworthy that, in contrast to flow immunocytometry which is still highly limited in colour number, it is the unique methodological and conceptual strength of MELC to follow hypothesis-free searches only with the limitation of the choice of the appropriate tags (antibodies/lectins etc.).

Generally, the accumulation of obviously recirculating lymphocytes in the blood periphery under efalizumab treatment, as analyzed in detail in the current study, may be regarded as one possible cause for intercurrent disease worsening [8, 21] or rarely occurring rebound phenomena after sudden treatment withdrawal. This assumption is based upon the putative effector cell role for the induction and sustaining of psoriatic skin lesions, as contributing to these recirculating lymphocytes.

Sixth, a major task for the future may be the predictive identification of efalizumab non-responders. Given our small patient cohort comprising 5 responders and 1 non-responder, valid statistical statements in this regard were impossible. Nevertheless, we would like to hypothesize that for future larger cohorts multiplex immunophenotyping of peripheral blood lymphocytes by MELC robot microscopy might be promising to approximate the ultimate goal of individual treatment response prediction. Such predictive strategies could be based upon CMP motif analyses as well as on simple canonic discrimination analyses of single epitope expressions. The latter approach was exemplarily performed for our cohort of 6 patients and the corresponding 7 healthy controls which identified, as a preliminary result, a cluster of CD4, CD7, CD36, CD45, CD45R0, CD247, HLA-DQ and SNA as relevant markers for a maximum statistical separation of the efalizumab non-responder from the responders (at the stage before treatment) and the healthy controls (data not shown).

Taken together, we have hereby confirmed that the clinical response to efalizumab treatment of psoriasis is paralleled by a down-modulation of cellular CD11a expression in peripheral blood lymphocytes, which show a concomitant increase in cell numbers of a hitherto unrecognized broad spectrum of diverse subpopulations.

Acknowledgements

The excellent technical assistance of Mandy Könnecke and Kathrin Brennecke is gratefully appreciated. We acknowledge the merit of Dr. Walter Schubert (Institute of Medical Neurobiology) and the Molecular Pattern Recognition Research (MPRR) Group (Otto-von-Guericke-University, Magdeburg, Germany), who invented the basic principles of MELC robot technology. The outlicensing of MELC technology to MelTec GmbH & Co. KG and SkinSysTec GmbH allowed us to develop an advanced application platform in the field of dermatology including skin and blood biochip methodology (patents pending). Moreover we recognize gratefully the special contribution of Peter Karcher (MelTec GmbH & Co. KG) as the developer of the MotifFinder statistics tool.

Financial Support: The study was supported in part by Serono International S.A., Switzerland

References

1 Krueger JG. The immunologic basis for the treatment of psoriasis with new biologic agents. J Am Acad Dermatol 2002; 46: 1-23.

2 Johnston A, Gudjonsson JE, Sigmundsdottir H, Love TJ, Valdimarsson H. Peripheral blood T-cell responses to keratin peptides that share sequences with streptococcal M proteins are largely restricted to skin homing CD8(+) T-cells. Clin Exp Immunol 2004; 138: 83-93.

3 Böckelmann R, Horn T, Gollnick H, Bonnekoh B. Interferon-γ-dependent in-vitro model for the putative keratin 17 autoimmune loop in psoriasis: exploration of pharmaco- and gene-therapeutic effects. Skin Pharmacol Physiol 2005; 18: 42-54.

4 Nickoloff BJ, Nestle FO. Recent insights into the immunopathogenesis of psoriasis provide new therapeutic opportunities. J Clin Invest 2004; 113: 1664-75.

5 Henseler T, Christophers E. Disease concomitance in psoriasis. J Am Acad Dermatol 1995; 32: 982-6.

6 Bonnekoh B, Gollnick H. Psoriasis. In: Katsambas AD, Lotti TM, eds. European Handbook of Dermatological Treatment. 2. Ed. Berlin: Springer, 2003: 433-50.

7 Bonish BK, Nickoloff BJ. Immunopathogenesis - Psoriasis. In: Gordon KB, Ruderman EM, eds. Psoriasis and psoriatic arthritis - an integrated approach. Berlin: Springer, 2005: 23-35.

8 Smith CH, Anstey AV, Barker JNWN, Burden AD, Chalmers RJG, Chandler D, Finlay AY, Grifitths CEM, Jackson K, McHugh NJ, McKenna KE, Reynolds NJ, Ormerod AD. British association of dermatologists guidelines for use of biological interventions in psoriasis 2005. Br J Dermatol 2005; 153: 486-97.

9 Gottlieb AB, Bos JD. Recombinantly engineered human proteins: transforming the treatment of psoriasis. Clin Immunol 2002; 105: 105-16.

10 Chaudhari U, Romano P, Mulcahy LD, Dooley LT, Baker DG, Gottlieb AB. Efficacy and safety of infliximab monotherapy for plaque-type psoriasis: a randomized trial. Lancet 2001; 357: 1842-7.

11 Mease PJ, Goffe BS, Metz J, Van der Stoep A, Finck B, Burge DJ. Etanercept in the treatment of psoriatic arthritis and psoriasis: a randomized trial. Lancet 2000; 356: 385-90.

12 Ellis CN, Krueger GG. Treatment of chronic plaque psoriasis by selective targeting of memory effector T lymphocytes. N Engl J Med 2001; 345: 248-55.

13 Jullien D, Prinz JC, Langley RGB, Caro I, Dummer W, Joshi A, Dedrick R, Natta P. T-cell modulation for the treatment of chronic plaque psoriasis with efalizumab (Raptiva™): mechanisms of action. Dermatology 2004; 208: 297-306.

14 Leonardi CL. Efalizumab, an overview. J Am Acad Dermatol 2003; 49(suppl 2): S98-S104.

15 Marecki S, Kirkpatrick P. Efalizumab. Nat Rev Drug Discov 2004; 3: 473-4.

16 Papp K, Bissonnette R, Krueger JG, Carvey W, Gratton D, Gulliver WP, Lui H, Lynde CW, Magee A, Minier D, Ouellet JP, Patel P, Shapiro J, Shear NH, Kramer S, Walicke P, Bauer R, Dedrick RL, Kim SS, White M, Garovoy MR. The treatment of moderate to severe psoriasis with a new anti-CD11a monoclonal antibody. J Am Acad Dermatol 2001; 45: 665-74.

17 Schubert W, Bonnekoh B, Pommer AJ, Philipsen L, Böckelmann R, Malykh Y, Gollnick H, Friedenberger M, Bode M, Dress AWM. Analyzing proteome topology and function by automated multi-dimensional fluorescence microscopy. Nat Biotechnol 2006; 24: 1270-8.

18 Bonnekoh B, Böckelmann R, Pommer AJ, Malykh Y, Philipsen L, Gollnick H. The CD11a binding site of efalizumab in psoriatic skin tissue as analysed by multi epitope ligand cartography (MELC) robot technology: introduction of a novel biological drug binding biochip assay. Skin Pharmacol Physiol 2007; 20: 96-111.

19 Perfetto SP, Chattopadhyay PK, Roederer M. Seventeen-colour flow cytometry: unraveling the immune system. Nat Rev Immunol 2004; 4: 648-55.

20 Bovenschen HJ, van de Kerkhof PCM, Gerritsen WJ, Seyger MMB. The role of lesional T cells in recalcitrant psoriasis during infliximab therapy. Eur J Dermatol 2005; 15: 454-8.

21 Menter A, Kosinski M, Bresnahan BW, Papp KA, Ware Jr. JE. Impact of efalizumab on psoriasis-specific patient-reported outcomes. Results from three randomized placebo-controlled clinical trials of moderate to severe plaque psoriasis. J Drugs Dermatol 2004; 3: 27-38.

22 Desroches CV, Andreoni C, Rigal D. Differential expression of the LFA-1 molecule on the human peripheral blood mononuclear cell subpopulations. Immunol Lett 1990; 24: 13-20.

23 Coffey GP, Stefanich E, Palmieri S, Eckert R, Padilla-Eagar J, Fielder PJ, Pippig S. In vitro internalization, intracellular transport, and clearance of an anti-CD11a antibody (Raptiva) by human T-cells. J Pharmacol Exp Ther 2004; 310: 896-904.

24 Lowes MA, Turton JA, Krueger JG, Barnetson R. StC. Psoriasis vulgaris flare during efalizumab therapy does not preclude future use: a case series. BMC Dermatol 2005; 5: 9.

25 Bauer RJ, Russel DL, White RL, Murray MJ, Garovoy MR. Population pharmacokinetics and pharmacodynamics of the anti-CD11a antibody hu1124 in human subjects with psoriasis. J Pharmacokinet Biopharm 1999; 27: 397-420.

26 Gottlieb AB, Krueger JG, Bright R, Ling M, Lebwohl M, Kang S, Feldman S, Spellman M, Wittkowski K, Ochs HD, Jardieu P, Bauer R, White M, Dedrick R, Garovoy M. Effects of administration of a single dose of a humanized monoclonal antibody to CD11a on the immunobiology and clinical activity of psoriasis. J Am Acad Dermatol 2000; 42: 428-35.

27 Halliday JA, Franks AH, Ramsdale TE, Martin R, Palant E. A rapid, semi-automated method for detection of Galbeta1-4GlcNAc alpha2,6-sialyltransferase (EC 2.4.99.1) activity using the Sambucus nigra agglutinin. Glycobiology 2001; 11: 557-64.

28 Mortensen DL, Walicke PA, Wang X, Kwon P, Kuebler P, Gottlieb AB, Krueger JG, Leonardi C, Miller B, Joshi A. Pharmacokinetics and pharmacodynamics of multiple weekly subcutaneous efalizumab doses in patients with plaque psoriasis. J Clin Pharmacol 2005; 45: 286-98.