Assessment and Clinical Relevance of Serum IL-19 Levels in Psoriasis and Atopic Dermatitis Using a Sensitive and Specific Novel Immunoassay

Because development of reliable biomarkers in psoriasis and atopic dermatitis has lagged behind therapeutic progress, we created a blood-based test to fill the void in objective methods available for dermatological assessments. Our novel interleukin-19 (IL-19) immunoassay was initially tested to determine concentrations of IL-19 serum levels, then correlated with the psoriasis activity and severity index (PASI) in psoriasis, and the eczema area and severity index (EASI) in atopic dermatitis. Not only was IL-19 increased in psoriasis and correlated to PASI, but ixekizumab administration led to rapid, sustained IL-19 decreases to normal levels, with decreases at 2-weeks correlating with PASI improvement at 16-weeks. IL-19 increased upon ixekizumab withdraw, prior to relapse, and decreased following re-treatment. In baricitinib- and etanercept-treated psoriasis patients, IL-19 decreases also correlated with improvement. Many patients with limited skin disease, including genital psoriasis and psoriatic arthritis patients, also had increased IL-19, which was reduced to normal levels upon ixekizumab treatment, correlating with PASI improvement. We also measured IL-19 in baricitinib-treated atopic dermatitis patients. In atopic dermatitis, IL-19 was significantly elevated, correlated with EASI scores, and decreased with skin improvement. Therefore, measurement of serum IL-19 provides clinicians with an objective disease-activity assessment tool for psoriasis and atopic dermatitis patients.

Generation, purification, selection, and labeling of murine monoclonal anti-human IL-19 antibodies -All experimental mouse protocols were approved by the Institutional Animal Care and Use Committee of A&G Pharmaceutical, Inc. (Columbia, MD, USA) and were conducted in accordance with their relative regulations and guidelines, with periodic review by Eli Lilly Veterinary resources. Anti-IL-19 hybridoma clones were obtained by fusing splenocytes from mice immunized with recombinant human IL-19 (prepared as described above). Clones were screened for non-overlapping epitopes and affinities determined after titrations of 0.37-270 nM of IL-19 protein using Octet Red96 (Pall ForteBio, Menlo Park, CA, USA) by global fitting of a 1:1 binding model using Data Analysis v9.0 software. The antigen-specific variable heavy (VH) and variable light (VL) gene sequences were determined from clones of interest. Sequences were transferred into separate murine IgG1 and IgK constant region expression vectors for recombinant antibody production and subsequent purification using protein A chromatography after large-scale transient transfection in HEK293 cells. Affinity maturation of the parent capture antibody was performed as previously described 1,2 with minor modifications. Briefly, VH and VL domains were combined to construct a single-chain Fragment variable (scFv) through single overlap extension PCR. The VH and VL domains were ligated into display vector, and mutagenic libraries of the VH 3 and VL complementarity determining regions were constructed by spiked mutagenesis using degenerate oligonucleotides. Libraries were screened for improved binding after iterative rounds of selection using decreasing amounts of biotinylated recombinant IL-19 by fluorescence-assisted cell sorting on a FACS Aria cell sorter (BD Biosciences, San Jose, CA, USA). Sequences of clones with enriched binding were determined and placed into expression vectors. Two monoclonal antibodies directed against separate IL-19 epitopes were selected for labeling. One milligram of each antibody was biotinylated using a Pierce kit and ruthenium-labeled using a MesoScale Discovery (MSD) kit for electrochemiluminescent (ECL) detection. Antibodies were evaluated using Matrix-Assisted Laser Desorption -Ionization-Time of Flight Mass Spectrometry (MALDI-TOF) to ensure that suitable labeling had occurred. Labeled antibodies were diluted in 50% glycerol and stored at -20 o C prior to use.
Plates were blocked with TBST plus 1% bovine serum albumin for 1 hour at room temperature.
After aspiration and washing with TBST, wells were incubated with biotinylated anti-IL-19 capture antibody (1 mg/L) for 1 hour. Following aspiration and washing with TBST, 50 μL of recombinant human IL-19 standards (serially diluted IL-19 recombinant protein ranging from 0-50,000 pg/mL), were added to the wells in assay buffer (50 mmol/L HEPES, pH 7.40, 150 mmol/L NaCl, 10 mL/L Triton X-100, 5 mmol/L EDTA, and 5 mmol/L EGTA). Serum samples were diluted 1:4 in assay buffer and added to their respective wells, and the ELISA plate was incubated overnight at 4 o C. After aspiration, wells were washed three times with TBST, and 50 μL of 4 ruthenium-labeled anti-IL-19 detection antibody (0.5 mg/L) was added for a 1-hour incubation at room temperature. Following aspiration, wells were washed three times with TBST, and 150 μL of 2X MSD read buffer was added. Ruthenium ECL in the wells was detected using a MSD plate reader. Specificity of the assay for IL-19 was confirmed by evaluating cross-reactivity to other cytokines, including additional members of the IL-10 cytokine family. Cross-reactivity was

Supplementary Results
Monoclonal anti-IL-19 antibodies were developed for use in a sandwich immunoassay to measure IL-19 in human serum ( Figures S1A and S1B). Kinetic analysis of the antibodies with nonoverlapping epitopes indicated poor affinity with one of the antibodies ( Figure S1A) precluding our ability to measure endogenous normal levels of IL-19. Subsequent in vitro affinity optimization resulted in an antibody with greatly improved binding kinetics ( Figure S1C). The optimized antibody (which became the capture antibody for the assay) had an equilibrium affinity constant (KD) of 5.76 x 10 -11 M, which was approximately 13,000-fold better than the parental antibody.
The optimal pairing used biotinylated capture antibody and ruthenium-labeled detection antibody on a streptavidin-coated MSD plate. Figure S1D shows the consistency of the assay across ten different runs. The assay showed acceptable dynamic range, background, sensitivity, and dilutional linearity. Inter-assay precision results (CVs) were less than 10% at all concentrations tested, and recoveries at all concentrations tested were 80-120%, even after six freeze-thaw cycles.
To ensure that IL-19 was not altered during the clotting process, we collected matching serum and EDTA plasma samples from healthy donors. All serum results were within 15% of EDTA plasma values. Finally, IL-17, TNFα, and members of the IL-10 family  demonstrated no cross-reactivity on the assay at concentrations as high as 25,000 pg/mL, confirming specificity for IL-19 (data not shown).

D:
A sandwich immunoassay was developed using the antibodies in Figures S1B and S1C to measure human serum IL-19. Recombinant human IL-19 was for a calibration curve starting at 50,000 pg/mL with serial dilutions. Each curve point is shown as the 95% confidence interval from 10 separate runs. The dashed horizontal line indicates the mean level of the zero calibrator.