Daclizumab reduces CD25 levels on T cells through monocyte-mediated trogocytosis

Patients and study design

CHOICE was a randomized, double-blind, phase 2, placebo-controlled study of patients with MS that showed a dose-dependent effect of DAC in reducing gadolinium-enhanced contrast-enhancing lesions (Gd+ CELs) on monthly brain MRI between weeks 8 and 24 in patients with relapsing–remitting MS (RRMS). ⁹ Three subcutaneous dose arms were evaluated: interferon-beta (IFNβ) + placebo (placebo/IFNβ); IFNβ + DAC low-dose (DAC Low/IFNβ); and IFNβ + DAC high-dose (DAC High/IFNβ). Sixty-four patients from CHOICE consented to participate in optional pharmacokinetic (PK) and pharmacodynamic (PD) assessments. PD assessments were performed in the same subset of patients in whom PK evaluations were performed. Samples for PD evaluations were collected at screening and at weeks 0, 2, 12, 20, 22, 28, 34, 40, and 44 (prior to dosing on dosing visits). The DAC treatment phase was weeks 0–24, and approximately eight to 12 weeks were required for antibody washout. Blood samples for PK analysis were collected in a subset of patients on days 0, 4, 7, 14, 28, 84, 140, 147, 154, 168, 196, 238, 280, 308, and 504.

Flow cytometry of MS patient blood

Flow cytometric phenotyping of patient tissues was performed using BD Biosciences (San Jose, CA, USA) antibodies on acid-citrate-dextrose (ACD)-B anticoagulated blood collected and shipped at ambient temperature for analysis next day by validated assays at Esoterix Clinical Trials Inc. CD25 expression and saturation levels were measured with antibody clones M-A251 and 2A3. An antibody against an invariant region of major histocompatibility complex (MHC) Class 2, anti-HLA-DR specific clone L243, was used to calculate percentages of activated CD4 T cells. TruCOUNT™ assay (BD Biosciences) was used to convert percentage values into cell counts per unit of blood.

Statistical analysis

Statistical and graphic analysis was performed using all results obtained from PD patients who had at least one baseline and one corresponding post-dosing laboratory measurement. Between-treatment group comparisons for differences in cell count values at nominal timepoint visits were analyzed by unpaired t test. Both parametric paired t test and nonparametric Wilcoxon signed-rank test were performed when analyzing for significant change from baseline and percentage change from baseline at each post-treatment timepoint.

Individual DAC exposure characteristics, including area under the concentration-time curve at steady state (AUC_ss (µg*hr/ml)) and steady-state trough concentration (C_min,ss (µg/ml)), were used as predictors to model changes from baseline level in individual immune subsets or changes over time. Relationship between changes from baseline level of HLA-DR⁺-activated T cells and total new or enlarged Gd+ CELs between weeks 8 and 24 was also evaluated using linear correlation, negative binomial correlation, analysis of variance, and Kruskal-Wallis Tests statistical analysis approaches.

Peripheral blood mononuclear cell (PBMC) isolation and culture

Blood from healthy volunteers was collected in sodium heparin anticoagulant vacutainer tubes and PBMC were isolated using Ficoll-Hypaque gradient centrifugation. Cells were plated at 2 × 10⁶ cells/ml in culture medium (CM) consisting of Roswell Park Memorial Institute (RPMI) medium containing 10% heat-inactivated fetal bovine serum (FBS; Sigma-Aldrich, St Louis, MO, USA), 1 mM HEPES buffer, 20 nM L-glutamine, 1 mM sodium pyruvate, and 0.05 mM 2-mercaptoethanol, 100 units/ml Penn-Strep (ThermoScientific (Hyclone), Logan, UT, USA) and maintained in a humidified CO₂ incubator.

Analysis of effects of DAC on CD25 levels in vitro

PBMCs were stimulated for 47 hours with phytohemagglutinin (PHA-P at 5 μg/ml, Sigma-Aldrich, St Louis, MO, USA) followed by addition of 10 µg antibody treatment (or molar equivalents for F(ab’)2 fragments) for one hour. F(ab’)2 antibody fragments were produced through enzymatic cleavage of DAC and purified by Southern Biotech Inc (Birmingham, AL, USA). The non-FcR binding variant monoclonal antibody DAC-IgG2m3 was generated by replacing the immunoglobulin G (IgG)1 Fc portion of DAC with an IgG2a Fc domain with replacements of alanine for valine at positions 234 and 237 to render the antibody Fc domain incapable of binding FcR and leaving FcRn binding ability intact. ¹⁷ NK cells and monocytes were selectively removed at 0 hour by positive selection using antibodies specific for removal of NK cells and monocytes according to the manufacturer’s (Miltenyi Biotech, Auburn, CA, USA) suggested isolation procedures. After removal of NK and monocytes, flow cytometric analysis was performed to confirm > 95% removal of NK and monocytes from the cultures. For all in vitro culture experiments involving flow cytometric analysis, cells were cultured at 1 × 10⁶ per ml in 96-well round-bottom plates. Cells were harvested and washed with ice-cold Staining Buffer™ (BD Biosciences) prior to incubation with fluorescently conjugated, antigen-specific antibodies for 30 minutes at room temperature (RT). Antibodies obtained from BD Biosciences were used to stain CD3 (clone SK7) and CD4 (clone RPA-T4).

Images of fluorescently stained cells were also obtained using an IN Cell Image Analyzer™ (GE Healthcare, USA) after cells were cultured at 2 × 10⁶ per ml in 24-well flat-bottom plates for 47 hours followed by harvest (including plate-bound monocytes) with phosphate-buffered saline (PBS) and ethylenediaminetetraacetic acid (EDTA), repeated wash using CM, and culture in 12 × 72 mm culture tubes. Antibody treatments were added at 10 µg (or molar equivalents for F(ab’)2 fragments) per 1 × 10⁶ cells for one hour. Treatments included DAC, isotype control Ig1 (clone MSL109) antibody, and F(ab’)2 purified fragments of each respective antibody, when chemically conjugated to fluorescently labeled AF647 (Southern Biotech Inc). After incubation for 40 minutes at 37°C, the cellular contents of each tube were washed two times with CM followed by incubation for 15 minutes with ER Tracker™ (Invitrogen, Carlsbad, CA, USA), a vital fluorescent staining marker that incorporates into endoplasmic reticulum for the purposes of identifying viable cells and determining whether fluorescently labeled anti-CD25 antibodies were intracellular and not surface bound.

PBMCs from each treatment were then prepared for IN Cell analysis using methods similar to those used for flow cytometry. Monocytes were identified using an anti-CD14 specific antibody (clone M5E2, BD Biosciences). Intra-nuclear staining of the FoxP3 transcription protein was performed using clone 259D/C7 (BD Biosciences) following the manufacturer’s suggested procedures. Briefly, cells were incubated and vortexed in FoxP3 buffer A for 10 minutes at RT followed by washing twice in Staining Buffer™ and incubation in FoxP3 buffer C for 30 minutes at RT. Cells were then washed repeatedly prior to staining with cocktails of fluorescently conjugated antibodies for 30 minutes at RT. After repeated washes, cells were suspended in BD Stabilizing Fixative™ (BD Biosciences), followed by dilution 1:4 in PBS to minimize formaldehyde effects on the ER tracker™ (blue) reagent. Anti-CD14 antibody was used to identify monocytes (green); anti-CD4 (green) and anti-FOXP3 (green) antibodies were used to identify regulatory T (Treg) cells.

CD25 inhibition of IL-2-induced proliferation by lymphoblasts

Lymphoblast cultures were generated using three days’ stimulation of PMBCs with 1 ng/ml IL-2. Proliferation of lymphoblasts was measured by quantifying ³H thymidine incorporation after pulsing with 1 μCi per well methyl-³H thymidine (GE Healthcare, Pittsburgh, PA, USA) in RPMI medium for the last 16 hours of cell culture. Cells were harvested onto FilterMate™ membranes using a Packard Harvester (PerkinElmer, Waltham, MA, USA), and beta particle emission was measured by scintillation counting. Individual replicates were performed on separate days. Statistical analysis using unpaired t test were performed for between-treatment group comparison after normalizing mean ³H thymidine counts per minute (cpm) for each antibody treatment condition of each day’s experiment to the stimulation only, no antibody, control treatment condition. Concentrations of antibody evaluated included: 0, 0.001, 0.005, 0.01, 0.05, 0.1, 1, 10, and 20 µg/ml.

DAC treatment in vivo decreased CD25⁺ T cells

As seen in Figure 1, a reduction of approximately 30% in the percentages of CD25-expressing CD4⁺ T cells was observed two weeks after a single dose of DAC for patients in DAC Low/IFNβ and DAC High/IFNβ treatment groups. Percentages of CD25⁺ cells generally returned to baseline levels at antibody washout between weeks 28 and 40. ⁹ No treatment-associated changes were seen for the Placebo/IFNβ treatment group. The sustained reductions in percentages of CD25-expressing CD4⁺ cells were observed throughout the treatment phase and in the absence of significant changes in CD4 cell counts for the two DAC treatment groups. ⁹

OPEN IN VIEWER

Figure 1.

Reductions in percentages of CD4⁺ T cells that stained positive for CD25 expression were observed for DAC Low/IFNβ and DAC High/IFNβ patients, throughout the dosing period. Shown are mean and SEM measurement results for patients in the Placebo/IFNβ (open circles, n = 17), Low DAC/IFNβ (upward triangles, n = 25), and High DAC/IFNβ (closed circles, n = 21) treatment groups. DAC and placebo dosing visits are indicated by downward arrows. Surface CD25 protein levels generally returned to baseline levels following DAC washout and CD25 desaturation beginning around week 28. ⁹ Reductions in surface CD25 levels were not observed for the placebo/IFNβ group patients.

DAC treatment in vivo reduced activated HLA-DR⁺ T cells

In contrast to the reduction in CD25⁺ T cells that had reached maximum levels at the week 2 timepoint, steady reductions in HLA-DR⁺-activated CD4 T cell counts that eventually attained statistical significance (p < 0.05) compared with placebo treatment group levels were observed in the DAC treatment groups at weeks 12 and 20 (Figure 2). Both DAC dose groups showed statistically significant (p < 0.01) reduced levels of activated HLA-DR⁺ T cells at two, 12, 20, and 22 weeks (and week 28 for DAC Low/IFNβ only) when the comparisons were within treatment groups, i.e. change from baseline and percentage change from baseline. The timing of these reductions in activated HLA-DR⁺ T cells, during the treatment phase, approximated when DAC dosing groups were associated with maximum reductions in the inflammatory lesion activity, as determined by MRI. ⁹ Significant reductions compared to baseline levels in activated HLA-DR+ T cells were not observed for the placebo/IFNβ dose group. Statistically significant increases in absolute HLA-DR⁺CD4⁺ T cell counts were observed during the washout period, a time of diminished clinical benefit in both DAC dose groups according to results of within-treatment comparisons (DAC Low/IFNβ, p < 0.05, week 34, 40, 44; and DAC High/IFNβ, p < 0.01, week 44), while significant increases were not observed when the analysis was compared with levels in the placebo/IFNβ group.

OPEN IN VIEWER

Figure 2.

Decreases from baseline in levels of absolute HLA-DR⁺-activated CD4⁺ T cell counts were observed for both DAC dosing groups. Shown are mean and SEM measurement for patients in the Placebo/IFNβ (open circles, n = 17), Low DAC/IFNβ (upward triangles, n = 25), and High DAC/IFNβ (closed circles, n = 21) treatment groups. Statistically significant reductions (p < 0.05) compared with placebo group counts were observed at week 12 (*p < 0.05) and week 22 (**p < 0.01).

Because a dose-response was observed in the CHOICE study in terms of efficacy, i.e. the DAC High/IFNβ and not the DAC Low/IFNβ treatment group achieved significance as defined by the study primary efficacy endpoint, an exposure versus PD response correlation was assessed to examine the relationship between patients’ serum DAC exposure levels (AUC_ss or C_min,ss) and activated T cell counts. Statistically significant, but biologically weak, correlations were observed between serum DAC antibody C_min,ss and average HLA-DR⁺ cell counts (Figure 3(a)) and change from baseline levels in average activated T cell counts during the treatment phase. When examining the effect of DAC treatment on average HLA-DR⁺CD4⁺ T cell count during the treatment period using analysis of variance (Figure 3(b)), a treatment effect was significant only for the DAC high-dose group (DAC high dose vs placebo: p = 0.0103; DAC low dose vs placebo: p = 0.1044).

OPEN IN VIEWER

Figure 3 (a) and (b).

A significant relationship was observed between changes in individual patients’ activated HLA-DR⁺ T cell counts and DAC serum exposure levels (Figure 3). Placebo/IFNβ (open circles), DAC Low/IFNβ (triangles), DAC High/IFNβ (closed circles), and 95% confidence interval for Pearson correlation are indicated. Although weak (r² = 0.127), the negative linear correlation observed between individual DAC trough concentration at steady state (C_min,ss) and patients’ average reductions from baseline level absolute HLA-DR⁺CD4⁺ T cell counts during the dosing period (Figure 3(a)) was statistically significant when the high DAC dose was compared to placebo (Figure 3(b)).

Dose-dependent down-modulation of CD25 on activated T cells by DAC in vitro

In attempting to understand how DAC exposure might lead to in vivo reductions in CD25 levels on T cells, DAC effects were studied using PHA-activated PBMC cultures of T, B, NK and monocytes in vitro (Figure 4). DAC, DAC-IgG2M3 and DAC-F(ab’)2 were capable of blocking CD25 on the antigen-rich CD4 T cell population in a dose-dependent manner (Figure 4(a)); however, only the FcR binding CD25 blocking antibody (i.e. DAC) reduced CD25 cell surface expression (Figure 4(b)). The experiment was repeated after selective removal of NK cells and monocytes prior to adding the anti-CD25 antibody or antibody fragments to the remaining PBMC cultures. Complete removal of monocytes abolished CD25 reduction by DAC antibody (Figure 5(a)). In contrast, complete removal of NK cells had no effect on DAC-mediated CD25 reduction (Figure 5(b)).

OPEN IN VIEWER

Figure 4 (a) (b).

DAC rapidly reduced CD25 levels on PHA-activated T cells in vitro. The DAC, DAC-F(ab’)2, and DAC-IgG2M3 were capable of saturating CD25 receptors on antigen-rich CD4 T cells (Figure 4(a)) as evidence of blocking binding of competing anti-CD25 binding clone 2A3. Only the Fc containing DAC antibody was associated with CD25 reduction (Figure 4(b)), control antibodies, and the non-FcR binding DAC variants DAC-F(ab’)2 and DAC-IgG2M3 and had no effect on CD25 levels, as determined by fluorescence intensity of the noncompeting anti-CD25 antibody clone M-A251. Actual concentrations of F(ab’)2 fragments reflect molar equivalents of complete IgG1 antibody expressed as µg/ml. CD25 levels were expressed as MFI relative to the level in the absence of antibody. Mean and standard error of the mean (SEM) from the four experiments are shown.

OPEN IN VIEWER

Figure 5 (a) (b).

DAC antibody-dependent reduction in CD25 levels on antigen-rich CD4 T cells was prevented after selective monocyte removal prior to addition of 2 µg/ml DAC antibody to remaining PHA-stimulated PBMC cultures (Figure 5(a)). In contrast, prior selective removal of NK cells did not prevent significant (p = 0.032) reductions in CD25 levels when DAC was added to the remaining PBMC in the cultures. Mean and SEM shown (n = 5).

To confirm DAC-mediated CD25 reduction was not caused by killing of CD25⁺ lymphocytes, percentages of dead cells in the cultures were measured using two sensitive reagents, 7-amino-actinomycin D (7AAD) and AnnexinV-FITC. Despite reductions in CD4⁺CD25⁺ cells of roughly 60% or greater, in the presence of DAC antibody the percentages of dead cells were in low single digits (data not shown), suggesting that a cell-killing mechanism could not account for the observed reduction in CD4⁺CD25⁺ cells.

Monocyte-mediated trogocytosis removes DAC from the cell surface

Further experiments were conducted to determine whether the Fc-mediated, monocyte-dependent removal of CD25 from the surface of DAC antibody-treated lymphocytes shared characteristic morphological features of trogocytosis, and whether antibody Fc-dependent removal of CD25 also took place on FoxP3 expressing Treg cells (Figures 6 and 7). Two-day PHA-stimulated PBMC cultures were incubated one hour with AF647-labeled DAC antibody and corresponding isotype controls (all staining red). Only cultures treated with the Fc containing full-length DAC antibody were associated with a punctate cytoplasmic staining pattern corresponding to internalized fluorescently labeled DAC-AF647 antibody, and such a staining pattern was observed only for CD14⁺ monocytes (Figures 6(b) and (d)).

OPEN IN VIEWER

Figure 6.

Fluorochrome-labeled DAC-AF647 was internalized by monocytes after PHA-stimulated PBMC cultures were incubated with DAC-AF647 antibody. Monocytes were identified by anti-CD14 antibody (a: green). CD25 was identified with DAC-AF647 (b: red). Viable cells were identified with the endoplasmic reticulum-specific ER-Tracker™ reagent (c: blue). Overlay images with the labeled DAC-AF647 antibody, the ER-Tracker™ and CD14⁺ monocytes are shown (d). Negligible or faint staining was observed for CD14+ monocytes following treatments with AF647-labeled DAC-F(ab’)2 fragment and AF647-labeled isotype control antibody.

OPEN IN VIEWER

Figure 7.

Fluorochrome-labeled DAC-AF647 was faint or undetectable on the surface of CD4⁺ T cells and FoxP3⁺ T cells when PHA-stimulated PBMC cultures were incubated with the DAC-AF647-labeled antibody. CD4 T cells were identified by anti-CD4 antibody (a: green) and putative Treg cells were identified in separate staining by anti-FOXP3 antibody (a: green). Viable cells were identified using ER-Tracker™ (c: blue). Abundant staining was observed for CD4⁺ T cells after incubation with the AF647-labeled DAC-F(ab’)2 fragment. The DAC-F(ab’)2-AF647 staining pattern on T cells was punctuate in appearance on the cell surface. Following trogocytosis by monocytes, the AF647-labeled DAC antibody was strikingly absent from surfaces of FoxP3⁺ Treg cells.

Longer exposure times revealed the DAC-F(ab’)2-AF647 antibody bound to surfaces of CD14⁺ monocytes (data not shown); in contrast, longer exposures could not reveal surface staining with DAC-AF647 antibody, suggesting the monocytes in culture express CD25, but FcR-dependent monocyte on monocyte trogocytosis and internalization of CD25 was taking place in the presence of DAC antibody. After the incubation to allow trogocytosis by the monocytes, the AF647-labeled DAC antibody was remarkably devoid from the surfaces of FoxP3⁺ Treg cells (Figure 7(d)).

More potent IL-2 inhibition by trogocytosis competent anti-CD25 antibody

Both the CD25 blocking antibodies DAC and DAC-IgG2M3 inhibited IL-2-induced proliferation of lymphoblasts in a dose-dependent manner (Figure 8(a)), but the maximum inhibition was reached at different concentrations of 1.0 (DAC) and 10 µg/ml (DAC-IgG2M3), respectively. Although the CD25 epitope binding domains for each of the antibodies are identical, the DAC-IgG2M3 antibody exhibited significantly less inhibition of IL-2-mediated proliferation in comparison to DAC antibody at concentrations near the inflection points for the inhibition assay. As a control, the DAC and DAC-IgG2M3 antibodies were evaluated for ability to dose-dependently inhibit IL-2-mediated proliferation of a T leukemia cell line, Kit225. No significant difference was observed for the DAC-IgG2M3 antibody’s ability to inhibit IL-2-mediated proliferation in the FcγR1-expressing cell null culture conditions of a T leukemia monoculture (Figure 8(b)). In the presence of monocytes, the trogocytosis-competent DAC IgG1 antibody has significantly greater ability to block IL-2-dependent proliferation in comparison to the trogocytosis-null DAC-IgG2M3 antibody.

OPEN IN VIEWER

Figure 8 (a) (b).

The CD25-blocking antibodies DAC and DAC-IgG2M3 inhibited IL-2-induced proliferation of lymphoblasts in a dose-dependent manner. Data are presented as mean and SEM measurements of total counts per minute of ³ H thymidine-associated emission for three independent experiments.  At concentrations near the inflection points for the inhibition curves, the non-FcR-binding DAC-IgG2M3 antibody had significantly (0.05 µg/ml, p < 0.045 and 0.1 µg/ml, p < 0.014) reduced ability to inhibit IL-2-mediated proliferation of PBMCs compared with DAC (a). Both DAC and DAC-IgG2M2 antibodies inhibited IL-2-induced proliferation of a T cell leukemia cell line, Kit223, in a dose-dependent fashion; however, under the trogocytosis-null culture condition of a tumor cell line, there is no significant difference in proliferation inhibition by the IgG1 versus the IgG2M3 antibodies (b).