A Novel C-C Chemoattractant Cytokine (Chemokine) Receptor 6 (CCR6) Antagonist (PF-07054894) Distinguishes between Homologous Chemokine Receptors, Increases Basal Circulating CCR6 1 T Cells, and Ameliorates Interleukin-23-Induced Skin Inflammation S

Blocking chemokine receptor C-C chemoattractant cytokine (chemokine) receptor (CCR) 6-dependent T cell migration has therapeutic promise in inflammatory diseases. PF-07054894

Introduction C-C chemoattractant cytokine (chemokine) receptor (CCR) 6 is expressed on key immune cells, such as memory T cells (including all Th17 cells), B cells, immature dendritic cells, and a subset of Tregs (Liao et al., 1999). Its exclusive ligand, the chemokine C-C motif ligand (CCL) 20 is produced by colonic epithelial cells, keratinocytes, dermal fibroblasts, synoviocytes, and alveolar epithelial cells. The ligand-receptor pair CCL20-CCR6 is responsible for the migration of pathogenic immune cells to sites of inflammation in a variety of autoimmune and inflammatory diseases, including psoriasis and inflammatory bowel disease (Ranasinghe and Eri, 2018). Thus, blockade of CCR6 signaling driven by CCL20 is an attractive therapeutic approach to treat inflammation associated with these chronic inflammatory disorders.
(R)-4- (1-methylcyclopentyl) methyl)amino)-3,4-dioxocyclobut-1-en-1-yl)amino)-3-hydroxy-N, N-dimethylpicolinamide (PF-07054894), (Gerstenberger et al., 2021) is a novel CCR6 small molecule antagonist. Here, we examined the pharmacological activity of PF-07054894 on CCR6 in both in vitro and in vivo models of inflammation. Chemotaxis assays using primary human immune cells were used to determine the mode of inhibition and functional selectivity of PF-07054894 for CCR6 relative to homologous chemokine receptors. Compound activity in chemotaxis assays was correlated with its receptor binding kinetics. In vivo activity of PF-07054894 was studied in two different animal models. Compound administration to naïve cynomolgus monkeys was used to interrogate the action of PF-07054894 on homeostatic circulation of CCR6 1 immune cells. The effect of PF-07054894 in inflammation was studied in a mouse interleukin-23 intradermal injection model.
Chemotaxis Assays. For CCL20/CCR6 chemotaxis experiments, human CD4 1 CCR6 1 CXCR3 -T cells (purchased from Precision for Medicine, https://www.precisionformedicine.com) were activated with anti-CD3/anti-CD28 Dynabeads (Invitrogen, https://www.thermofisher. com/us/en/home/brands/invitrogen.html) at 1.5 beads per cell ratio and 10 6 cells/ml in culture medium (RPMI, 10% fetal bovine serum, 1% penicillin/streptomycin and 4 ng/ml IL-2) at 37 C, 5% CO 2 . Three days post activation, cells complexed with beads were washed with medium and beads were removed using a Dynal MPC-2 magnet. Subsequently, fresh medium was supplied daily. Chemotaxis was carried out 12-14 days post-activation when cells had maximal expression of CCR6 (Ghannam et al., 2011). For CCL21/CCR7 chemotaxis experiments, human CD4 1 T cells were isolated from freshly drawn whole blood by immuno-magnetic negative selection using RosetteSep CD4 1 T cell Enrichment Cocktail (Stemcell Technologies, https://www.stemcell.com) followed by density gradient centrifugation using Lymphoprep (Stemcell Technologies) with Sepmate tubes (Stemcell Technologies) following the manufacturer's protocols. Remaining red blood cells were lysed with Invitrogen Ammonium-Chloride-Potassium Lysis Buffer. Prior to the chemotaxis assay, cells were incubated for 20-22 hours at 37 C, 5% CO 2 in the culture medium described above without activation beads. Primary human neutrophils were purified from freshly drawn whole blood by immune-magnetic negative selection using the EasySep Direct Human Neutrophil Isolation Kit (StemCell Technologies) following the manufacturer's protocol. Purified neutrophils were resuspended in chemotaxis buffer to 2x10 6 cells/ml and incubated for 30 minutes at 37 C with 1.5 mM calcein-AM for cell labeling. The labeled neutrophils were then washed twice with chemotaxis assay buffer and used immediately. The assay buffer for chemotaxis consisted of Hanks' Balanced Salt Solution, 20 mM HEPES pH 7.4, and 0.25% bovine serum albumin. For human chemotaxis with human T cells, the 96-well ChemoTx Disposable Chemotaxis System (Neuroprobe 101-5, https:// www.neuroprobe.com) consisting of top and bottom chambers (in a 96-well plate format) separated by a filter (3 micron and 5 micron filters were used for CCL20/CCR6 and CCL19/CCR7 chemotaxis, respectively) was used according to the manufacturer's protocol. Twenty-eight microliters of compound (or 0.1% DMSO) plus chemokine mixture was added in the bottom chambers, and 25 ll of compound (or 0.1% DMSO) plus cells (50,000 cells and 75,000 cells per chamber for CCR6/CCL20 and CCR7/CCL21 chemotaxis, respectively) mixture was added in the top chambers. The chemotaxis plates were incubated at 37 C, 5% CO 2 for 2 hours. Then the bottom chambers containing migrated cells were immediately frozen at -80 C for at least 1 hour and subsequently stained with CyQUANT dye (Life Technologies C7026, https://www. thermofisher.com/us/en/home/brands/life-technologies.html) for cell number determination following the manufacturer's protocol. Fluorescence measurements were converted to cell number using an 11-point, 2-fold dilution standard curve with the appropriate range of cell numbers. Chemotaxis with human neutrophils was carried out with a similar method as described above for human T cells with the following modifications. The FluoroBlok chemotaxis system (https://ecatalog. corning.com/life-sciences/b2c/US/en/Permeable-Supports/HTS/Corning% C2%AE-FluoroBlok%E2%84%A2-96-Multiwell-Insert-Systems%2C-PET-Membrane/p/corningFluoroBlok96WellMultiwellInsertSystemsPET MembraneInsertPlate) consisting of top and bottom chambers (in a 96-well plate format) separated by a 3-micron filter was used. Two hundred microliters of PF-07054894 (or 0.1% DMSO) plus C-X-C motif ligand (CXCL) 1 at the indicated concentrations were added to the bottom chambers, and 50 ll of PF-07054894 (or 0.1% DMSO) plus calcein-AM-labeled neutrophils (50,000 cells) mixture was added to the top chambers. The chemotaxis assay was conducted at room temperature for 1 hour. The Envision multilabel reader (Perkin Elmer) or the Typhoon Fluorescent Imager (GE Healthcare Life Sciences, 485 nm excitation/535 nm emission) were used to measure the calcein-AM-labeled primary human neutrophils that migrated to the bottom chamber. Unless stated otherwise, cells were incubated with the indicated concentrations of compounds for 30 minutes at room temperature in chemotaxis assay buffer prior to the chemotaxis assay. In the experiments shown in Fig. 3C (red symbols) and Supplemental Fig. 2 (red symbols), cells were incubated with the indicated concentrations of PF-07054894 for 20 hours at 37 C, 5% CO 2 . Then cells were centrifuged at 396 ×g for 8 minutes, the supernatant was aspirated, and cells were resuspended in chemotaxis assay buffer containing the corresponding concentrations of PF-07054894. The cell densities were adjusted to 2.2 or 3.3×10 6 cells/ml for CCR6 (Fig. 3C) or CCR7 (Supplemental Fig. 2), respectively and then the chemotaxis experiment was set up as described above. Fraction unbound of PF-07054894 in the chemotaxis assay media was determined to be 0.599% by equilibrium dialysis using HTD 96 device (https:// www.htdialysis.com/) followed by liquid chromatography tandem mass spectrometry.
Effect of PF-07054894 on Circulating Leukocytes and Gene Expression in Blood and Skin in Cynomolgus Monkeys. Male cynomolgus monkeys of Mauritian origin, 5 years of age were used in this study. All animal procedures were conducted in accordance with the current guidelines for animal welfare and according to the Institutional Animal Care and Use Committee. PF-07054894 monohydrate was formulated in vehicle (0.5% [w/v] methylcellulose in purified water) at a concentration of 10 mg/ml.
The study design was a crossover of two phases of 6 days each separated by a 9-day washout period. In each phase, animals (n 5 10/dose group) were orally (via gavage) dosed twice daily (BID, approximately 6 hours apart) with vehicle or PF-07054894 at 20 mg/kg/day (10 mg/ kg/dose, 1 ml/kg/dose) for 5 consecutive days, and once (first dose only) on day 6. In Phase 1 of the study, dose Group 1 received vehicle and dose Group 2 received PF-07054894. In Phase 2, dose Group 1 received PF-07054894 and dose Group 2 received vehicle.
Animals were given standard access to food and water throughout the duration of the study. Animals were observed daily for routine health monitoring and body weights were collected for the purpose of dose volume calculation on Day 1. Whole blood and skin biopsies for immunophenotyping, gene expression, or PF-07054894 exposure were collected as follows: Immunophenotyping and gene expression in whole blood collected before treatment, during the washout phase, and in both phases on Days 1 to 5 (1 hour prior to the second daily dose) and on Day 6 after the last dose; skin biopsy was obtained on Day 6 of Phase 2 after the last dose. Skin biopsies were collected under anesthesia.
A 10-color flow cytometry panel was designed to provide an indepth coverage of the cell types that expressed detectable levels of CCR6 in circulation. The panel used the following surface markers, fluorochromes, and clones: CD45 (APC-R700, clone D058-1283), CD3 (V500, clone SP34-2), CD4 (PerCP-Cy5.5, clone L200), CD8 (FITC, clone SK1), CD20 (APC-H7, clone 2H7), CD159a (PE, clone Z199), CCR6 (PE-Cy7, clone G034E3), CXCR3 (BV421, clone 1C6/CXCR3), CD95 (PE-CF594, clone DX2), and CD28 (APC, clone CD28.2). CCR6 (G034E3) and CD159a (Z199) were obtained from Biolegend (https:// biolegend.com) and Beckman Coulter (https://beckman.com), respectively, and the other antibodies were obtained from BD Biosciences (https://bdbiosciences.com). Antibody cocktails were generated using the manufacturers recommended concentrations. Two fluorescenceminus-one controls for CCR6 and CXCR3 were generated by omission of either CD196-PE-Cy7 or CD183-BV421, respectively. Flow cytometric analysis was conducted with a BD FACSCanto flow cytometer with DIVA software version 8.0.1. Change from baseline for fluorescence-activated cell sorting (FACS) data were calculated using the mean of two pre-dose measurements for Phase 1, and one measurement during washout period for Phase 2 for all FACS cell types and measurements (cells/ll, percent, and mean fluorescence intensity). FACS cell type data were graphed against time with both Phase 1 and 2 of the study and the average of Phase 1 and 2 to visualize treatment effects. A saturated longitudinal linear mixed effects model with fixed effects of dose (vehicle BID or PF-07054894 at 10 mg/kg BID), day, and phase with animal ID as a random effect using an AR(1) correlation structure was fit to all 61 FACS cell types. The model was fit using the lme function in R. The Tukey multiple comparison adjustment was used to control the experiment-wise error rate for least squares mean comparisons. R version 3.6.1 was used for all analyses.
Methods for gene expression in blood and skin in cynomolgus monkeys are described in the Supplemental Material (Supp. Methods for RNAseq and droplet digital polymerase chain reaction).
Effect of PF-07054894 on Mouse Interleukin-23 Intradermal Injection Model and on Mouse Splenocytes. Female C57BL/6J and CCR6-deficient mice on C57BL/6 background (Strain #:005793, N4F8), 8-10 weeks in age were purchased from Jackson Laboratory (https://www.jax.org/). All mice were housed in pathogen-free environment at a Pfizer animal facility. All protocols were approved by Pfizer Animal Care and Use Committee.
PF-07054894 was formulated weekly with 1.25% w/v hydroxypropyl cellulose grade SL, 0.05% w/v dioctyl sodium sulfosuccinate in distilled water as a nanosuspension, stored refrigerated, and stirred for 30 minutes at room temperature prior to dosing. Mice were challenged with 500 ng of murine interleukin-23 protein, carrier-free, reconstituted with 25 ml of endotoxin-free, 0.9% sterile saline, injected intradermally in the left ear for six challenges, every other day, over a 12-day period. One group of mice had 0.9% sterile saline intradermally injected into the right ear. Mice were treated with PF-07054894 or vehicle, administered orally, BID or once a day (QD) starting from the day of the first challenge, and daily for 12 days. Ear thickness was measured in triplicate using an engineer's micrometer (Mitutoyo, IL) as a means of assessing swelling, epidermal hyperplasia, and inflammation occurring in the ear during the in-life portion of the study. The measurements were performed in a blinded fashion. Treatment group comparisons were made using a longitudinal mixed effects ANOVA model with mouse as a random effect and treatment group as a fixed effect in a period of stable disease change across treatment groups over interleukin-23 challenges, 3 through 6 post murine interleukin-23 challenge. Data were tested for normality with Shapiro-Wilks test and Q-Q plots. All analyses were performed in R version 3.5.2.
To determine the effect of PF-07054894 on mouse circulating lymphocytes, 50 ll of blood was collected at the indicated time points. Red blood cells were lysed using 500 ll AKC lysis buffer at room temperature for 3 minutes. The lysis was terminated by adding 1.5 ml of a solution of Miltenyi magnetic-activated cell sorting rinsing solution (https://www.miltenyibiotec.com), phosphate buffered saline, and 1% bovine serum albumin. Cells were washed, centrifuged, and stained with the following antibody cocktail: CD45 BV786, AF700 (clone 30-F11), B220 BV605 (clone RA3-6B2), CD3 FITC (clone 145-2C11), CD4 AF647 (clone RM4-5) (BD Pharmingen), and CCR6 PE (clone 29-2L17) (Biolegend hppts://biolegend.com). Immediately before the analysis, 7ADD live dead dye was added for the live cell discrimination. Stained cells were analyzed on BD LSRFortessa cytometer. BD QuantiBrite beads were used to quantify CCR6 expression on the cell surface according to manufacturer's recommendations. Statistical analysis was carried out with either a mixed effects model with appropriate fixed, random, correlation, and variance structure or a generalized least squares model, if a random effect was not necessary, or a two-way ANOVA, if a random effect and correlation structure was not necessary. The Holm's multiple comparison adjustment was used to control the family wise error rate for multiple group comparisons in the models. All residuals were evaluated for meeting the normality assumption with the Shapiro-Wilks Test and Q-Q Plots and the homogeneity/sphericity assumption by the residual plots or Bartlett's test. All analyses performed with R version 3.5.1. The effect of PF-07054894 on B cells in vitro was examined by incubating the indicated concentrations of compound with C57Bl6/6J mice splenocytes in RPMI medium with 10% fetal bovine serum for 24 hours at 37 C. Cells were stained with CCR6 PE (clone 29-2L17) (Biolegend, hppts://biolegend.com).

Results
PF-07054894 Inhibits CCL20/CCR6-Mediated T Cell Chemotaxis in an Insurmountable Manner. PF-07054894 is a novel CCR6 small molecule antagonist which blocked in vitro CCL20-dependent human T cell chemotaxis with an IC 50 of 5.7 nM (Fig. 1A). In an assay panel of 168 G proteincoupled receptors testing for agonist-dependent receptor-b-arrestin coupling, PF-07054894 at 1 lM caused negligible activation (<12% over basal activity) on all receptors tested. When examined as an antagonist, PF-07054894 caused more than 30% inhibition of agonist-dependent activity on only three receptors: CCR6, CCR7, and CXCR2 ( Fig. 1B; Supplemental Table 1).
Chemokines and their receptors mediate trafficking and migration of immune cells in homeostatic and inflammatory conditions (McCully et al., 2018). To understand the pharmacological mode of action of PF-07054894 on these receptors, chemotaxis concentration-response curves with the respective agonists were generated in the presence of various concentrations of PF-07054894. EC 50 values for the chemokines used in the chemotaxis assays (Fig. 2) were within the range of reported in vivo concentrations in inflamed tissues for human CCL20 (0.1-1 nM) (Schlenk et al., 2005;Melis et al., 2010;Kaneko et al., 2018;Riutta et al., 2018;Shi et al., 2021), human CCL19 (0.01-1 nM) (Nureki et al., 2013;Lepennetier et al., 2019), and human CXCL1 (0.1-0.2 nM) (Villard et al., 1995;Kuca-Warnawin et al., 2016) suggesting that cells used in the assay maintained physiologic sensitivity to their respective chemokines.
Human T cell chemotaxis induced by CCL20 (EC 50 5 0.09 nM) was inhibited by PF-07054894 in an insurmountable manner ( Fig. 2A). In contrast human T cell chemotaxis induced by CCL19 (EC 50 50.56 nM) and neutrophil chemotaxis dependent on CXCL1 (EC 50 5 0.17 nM) were blocked by PF-07054894 in a surmountable manner (Fig. 2, B and C). Thus, the inhibitory action of PF-07054894 on CCR6-mediated chemotaxis was relatively insensitive to ligand concentration whereas its blockade  Binding Kinetics of PF-07054894 is Important for its Pharmacological Mode of Action and Receptor Selectivity. Binding studies with [ 3 H]-PF-07054894 were carried out in CHO-K1 cells expressing human recombinant CCR6, CCR7, or CXCR2 at room temperature and at 37 C. Binding affinity of [ 3 H]-PF-07054894 to CCR6 was higher than for CCR7 and CXCR2, 3.5-and 11-fold at room temperature, and 7-and 22-fold at 37 C, respectively (Supplemental Fig. 1). The dissociation half-life of [ 3 H]-PF-07054894 for CCR6 was longer than those for CCR7 and CXCR2, 8-and 18-fold at room temperature, and 6.5-and 11-fold at 37 C, respectively (Fig. 2, D and E). Thus, PF-07054894 distinguishes between these chemokine receptors based on its binding properties. Offset kinetics were faster at 37 C than at room temperature, which is consistent with reported observations on other GPCRs (Wallace and Young, 1983;Gantzos and Neubig, 1988;Treherne and Young, 1988;Fierens et al., 2002;Isorna et al., 2005).
Faster dissociation rates of PF-07054894 from CCR7 and CXCR2 allow their chemokine ligands to surmount compound inhibition. In contrast, the slow dissociation rate and consequent prolonged residence time of PF-07054894 on CCR6 is consistent with its insurmountable pattern of chemotaxis inhibition. To provide additional evidence for the role of offset kinetics on the pharmacological mode of action of PF-07054894 on CCR6, experiments were performed examining the inhibition pattern of a CCR6 antagonist with faster dissociation rate, and the effect of compound preincubation time on chemotaxis IC 50 . SQA1 ((R)-3-Hydroxy-4-((2-((1-(4-isopropylfuran-2-yl)propyl) amino)-3,4-dioxocyclobut-1-en-1-yl)amino)-N,N-dimethylpicolinamide (Aciro et al., 2010) is a CCR6 antagonist structurally related to PF-07054894 whose CCR6 dissociation halflife was 10-fold faster than that of PF-07054894 (compare Fig. 2E with Fig. 3A). SQA1 inhibited CCL20-mediated chemotaxis in human T cells in a surmountable manner (Fig. 3B) which is consistent with the notion that its shorter residence time on CCR6 allowed CCL20 to overcome inhibition.
In the experiment shown in Fig. 1A, T cells were preincubated with PF-07054894 prior to the chemotaxis assay for 30 min at room temperature which may not have been long enough to reach equilibrium (Supplemental Fig. 1). In Fig. 3C, T cells were preincubated with PF-07054894 prior to chemotaxis assay for 20 hours. Such extended equilibration time resulted in a 10-fold increase in compound potency which suggests that higher receptor occupancy by PF-07054894 was achieved upon longer preincubation. Prolonged incubation of PF-07054894 is not expected to impact its potency for CCR7and CXCR2-mediated chemotaxis because of the rapid offset kinetics for these receptors. In fact, incubation of T cells for 20 hours with PF-07054894 did not increase its potency for CCR7-dependent chemotaxis (Supplemental Fig. 2). This observation also rules out intracellular compound accumulation as the cause of the potency shift for CCR6. Taken together, these observations suggest that the pattern of chemotaxis inhibition by PF-07054894 depends on its receptor binding kinetics.
These results suggest that PF-07054894 shows functional selectivity for CCR6 relative to CCR7 and CXCR2. Upon equilibration, its potency for CCR6-mediated chemotaxis was 0.6 nM in the presence of 0.5 nM CCL20, which is within the concentration  (6) min, 19 min and 8.6 (0.6) min [mean (S.D.), n 5 2], respectively, resulting in calculated residence time values at 37 C of 21.8 min, 3.4 min, and 2 min, and at room temperature of 226 min, 27 min, and 12 min, respectively. Typical specific binding values at time zero for CCR6, CCR7 and CXCR2 were in the range of 15,000-28,000, 29,000-46,000 and 7,000-14,000 cpm, respectively. All data points and error bars represent mean and S.D. of duplicate or triplicate (room temperature) or quadruplicate (37 C) determinations of representative experiments. range observed in inflamed tissues (Schlenk et al., 2005;Melis et al., 2010;Kaneko et al., 2018;Riutta et al., 2018;Shi et al., 2021).
PF-07054894 pA2 values in chemotaxis for CCR7 and CXCR2 are similar to their respective dissociation affinity constants (K D ): pA2 5 7.41(39 nM) and K D 535-60 nM for CCR7, and pA2 5 6.79 (162 nM) and K D 5107-181 nM for CXCR2, suggesting consistency between compound binding and functional potencies for these receptors. Thus, the functional selectivity of PF-07054894 for CCR6 relative to CCR7 and CXCR2 is 50-to 150-fold, respectively, and it may be larger depending on local concentrations of CCL19 and CXCL1.
Orally Administered PF-07054894 Increases Circulating CCR6 + T Cells. In vivo activity of PF-07054894 was examined in both cynomolgus monkeys and in mice. K D and dissociation half-life values of [ 3 H]-PF-07054894 for both species are similar to those for human CCR6 (Supplemental Fig. 3), suggesting that its binding properties are evolutionarily conserved. To determine whether PF-07054894 affects the trafficking of CCR6-expressing lymphocytes in vivo, the compound was administered to naïve cynomolgus monkeys, and circulating levels of various lymphocyte populations were determined. Cynomolgus monkeys were administered PF-07054894 at 10 mg/kg PO or vehicle orally twice daily for 6 days in a two-way crossover jpet.aspetjournals.org study design. Flow cytometry analysis of whole blood, sampled daily, was used to evaluate leukocytes for differences between treatments relative to baseline. Expression of CCR6 in various cynomolgus monkey lymphocyte subsets (Table 1) was similar to human, including memory T cells (particularly in central memory CD4 1 and CD8 1 ), Th17 cells, and high expression in B cells (Liao et al., 1999;Krzysiek et al., 2000;Acosta-Rodriguez et al., 2007;Kondo et al., 2007;Singh et al., 2008). Compared with vehicle, PF-07054894 treatment resulted in an increase in the frequency of most circulating CCR6 1 T lymphocytes populations examined (CCR6 1 CD4 1 and CD8 1 cells including total, effector, memory, Th17, and Th1/17 cells) ( Table 2) that met the nominal P value threshold (P < 0.009). These observations are consistent with a role of CCR6 in basal homeostatic trafficking of T cells between peripheral blood and tissues and represents a pharmacodynamic biomarker for the action of PF-07054894 on CCR6. Regarding other lymphocyte subsets, frequency of CCR6 1 B cells, and naïve CD4 1 and CD8 1 T cells were not affected by PF-07054894. The lack of effect of PF-07054894 in frequency of peripheral CCR6 1 B cells in naïve monkey (Table 2) is consistent with the reported lack of CCR6-dependent chemotaxis in B cells in the absence of antigen activation (Liao et al., 1999(Liao et al., , 2002Reboldi et al., 2016).
Treatment with PF-07054894 caused a decrease in the percentages of CD8 1 and CD4 1 naïve T cells compared with vehicletreated animals that met the nominal P value threshold. There was a corresponding increase in percentage of memory cells in both populations which met the nominal P value threshold for CD8 1 central memory cells (Table 2). There was also a small decrease in total CD8 1 cells in the compound-treated cohort that met the nominal P value threshold. The mechanisms for these effects are not understood.
In addition, treatment with PF-07054894 resulted in increases in CCR6 mean fluorescence intensity (an indicator of CCR6 cell surface density) in central memory and Th1/17 CD4 1 cells, and in B cells where this effect was particularly pronounced meeting the nominal P value threshold (P < 0.0001) ( Table 2). These findings suggested target engagement of PF-07054894 with CCR6. The effects of PF 07054894 treatment on cell frequency and mean fluorescence intensity relative to vehicle were observed by the second day of dosing and were sustained through Day 6 of treatment (Fig. 4).
To examine the effect of PF-07054894 administration on gene expression, total RNAseq analysis and CCR6 droplet digital polymerase chain reaction were performed in blood samples throughout the study and in skin biopsies at the end of the study, in both vehicle-and compound-treated groups. There were no substantial differences in CCR6 expression in blood and skin samples between treatment groups (Supplemental Table 2). The analysis of blood RNAseq data between treated and vehicle groups revealed 60 genes that were significantly and differentially expressed across treatment groups. However, these genes did not segregate into any biologic grouping.
PF-07054894 is Efficacious in Interleukin-23-Induced Inflammation. To study the effect of PF-07054894 in a skin inflammation model, the compound was tested in a mouse ear skin inflammation/edema model induced by intradermal injection of interleukin-23 which was previously reported to require CCR6 by mediating the trafficking and positioning of skin resident cd-T cells within the dermis and epidermis (Hedrick et al., 2009;Cai et al., 2011;Mabuchi et al., 2013). In addition, CCR6-expressing dendritic cells in the skin have also been reported to contribute to inflammation in this model (Singh et al., 2016). Female C57BL/6J mice were challenged with murine interleukin-23 protein intradermally in the ear and PF-07054894 at 100 mg/kg or vehicle was administered by oral gavage either once or twice per day. As shown in Fig. 5A and Table 3, both dose regimens caused a statistically significant reduction in ear swelling relative to vehicle (P < 0.001). Twice daily dosing was more efficacious than once daily (P < 0.05) and showed higher compound systemic exposure (Table 3;  Supplemental Table 4). CCR6 deficient mice were protected from interleukin-23 challenge as previously reported and were not different from either dose regimen of PF-07054894 jpet.aspetjournals.org and did not meet the nominal P value threshold (Table 3). Thus, PF-07054894 inhibited mouse ear swelling induced by interleukin-23 intradermal injection to an extent similar to genetic ablation of CCR6, suggesting that pharmacological blockade rather than any effect in immune system development is sufficient to account for the role of CCR6 in this model. Flow cytometry analysis of peripheral blood revealed that PF-07054894 caused an increase in CCR6 mean fluorescence intensity in B cells, similar to the observation in cynomolgus monkeys (Fig. 5B). This effect was recapitulated in vitro. PF-07054894 caused a concentration-dependent increase in CCR6 mean fluorescence intensity in B cells upon incubation with mouse whole blood with an EC 50 5 0.9 nM (0.07 nM fraction unbound) (Fig. 5C).

Discussion
Pharmacological Mode of Action of PF-07054894 in Chemotaxis Assay. Insurmountable behavior of antagonists having slow dissociation rates has been described previously (Rang, 1966;Olins et al., 1995;Mathiesen et al., 2006;Riddy et al., 2015), particularly under "hemi-equilibrium" assay conditions where receptor, agonist and antagonist do not reach equilibrium during the assay timeframe (Gaddum, 1957;Nickerson, 1957;Kenakin et al., 2006). During the short-term chemotaxis assay, cells migrate toward a chemical gradient of CCL20 from the top to the bottom chambers, so equilibration between CCR6, CCL20, and PF-07054894 is unlikely to occur. The insurmountable inhibition of PF-07054894 on CCR6 may enable PF-07054894 to block CCR6-mediated responses regardless of the local concentration of CCL20 in inflamed tissue.
One possible factor that could contribute to differences in patterns of chemotaxis inhibition is the extent of agonistreceptor reserve which depends on receptor number and efficiency of receptor-effector coupling. High receptor reserve contributes to surmountable inhibition because activation of a relatively small fraction of receptors is sufficient to achieve maximal response. Conversely, low receptor reserve will favor insurmountable inhibition (James et al., 1991;Kenakin et al., 2006;Riddy et al., 2015). Reported whole cell binding affinities for CCL20 (0.4-0.7 nM) (Baba et al., 1997;Hieshima et al., 1997;Liao et al., 2002), CCL19 (1.6 nM) (Corbisier et al., 2015), and CXCL1 (1.2-1.4 nM) (Lee et al., 1992;Pat Cerretti et al., 1993) are 4-to 7-fold, 3-fold, and 13-fold higher than their respective chemotaxis EC 50 values in this study. This suggests that there was receptor reserve in all three systems examined, and that the insurmountable behavior of PF-07054894 for CCL20/CCR6 was unlikely due to low receptor reserve. Whether high receptor reserve is relevant to the surmountable behavior of PF-07054894 on CCL19/CCR7 and CXCL1/CXCR2 chemotaxis remains a possibility. Based on the reported tissue concentrations of these chemokines (see above), it is expected that receptor reserve is present in vivo.
Effect of PF-07054894 on Circulating Lymphocytes in Cynomolgus Monkey. A role of CCR6 in normal T cell migration has been proposed based on CCR6 knockout mice which showed altered T cell composition in intestinal mucosa but normal hematologic and bone marrow profiles (Cook et al., 2000;Varona et al., 2001). CCR6-dependent migration of CD4 1 and CD8 1 T cells in a variety of inflammatory conditions has been described (Nistala et al., 2008;P ene et al., 2008;Diani et al., 2017;Diani et al., 2019). To our knowledge, this is the first report demonstrating a role of CCR6 in homeostatic trafficking of circulating T cells in a naïve animal.
PF-07054894 caused an increase in cell surface CCR6 mean fluorescence intensity in monkey and mouse B cells and in certain monkey T cell populations (Table 2; Figs. 4 and 5). This effect is unlikely due to the increase in receptor transcription because no detectable effect of PF-07054894 on CCR6 mRNA was observed by droplet digital polymerase chain reaction (Supplemental Table 2). CCL20-stimulated CCR6 internalization and recycling has been reported including in B cells (Meissner et al., 2003;Lu et al., 2018), which is dependent on receptor coupling to b-arrestin. Although CCL20-dependent CCR6 internalization is unlikely to occur in peripheral blood because of the very low levels of circulating CCL20 (1 pM, Agalliu et al., 2013), ligand-independent activation of b-arrestin by CCR6 has been reported (Julian et al., 2017). Because PF-07054894 blocks CCR6-b-arrestin coupling (Fig. 1B), it is tempting to speculate that basal constitutive b-arrestin-mediated CCR6 intracellular recycling between plasma membrane and cytoplasm is blocked by PF-07054894 resulting in the observed increase in cell surface CCR6 mean fluorescence intensity, particularly in B cells. Fig. 4. Effect of PF-07054894 on MFI and cell frequencies of selected lymphocyte populations in cynomolgus monkeys. Animals were treated with vehicle (N 5 10) or PF-07054894 (N 5 10) at 10 mg/kg twice a day in a crossover study design as described in Methods. Change from baseline by treatment averaged over Phase 1 and 2 of the study is shown for CCR6 1 B cell MFI and CD4 1 CCR6 1 CXCR3 À (Th17 cells), CD4 1 CCR6 À CXCR3 1 , and CD4 1 CCR6 À CXCR3 À cell frequencies (as % of CD4 1 cells). A difference between vehicle and compound treated cohorts was observed for CCR6 1 B cell MFI and Th17 cell frequencies, but not for CD4 1 CCR6 À cells frequencies ( Table 3) that met the nominal P value threshold. Each dot represents an individual animal, and the lines are average determinations. Vehicle-treated and PF-07054894 groups are depicted in red and blue, respectively. 88 Li et al. at ASPET Journals on August 6, 2023 jpet.aspetjournals.org

Downloaded from
Efficacy of PF-07054894 in Mouse Interleukin-23 Induced Inflammation Model. The effect of PF-07054894 in mouse interleukin-23 induced skin inflammation is consistent with the reported efficacy of other CCL20/CCR6 blocking agents in the same model: a dual equipotent CCR6/CXCR2 antagonist (Campbell et al., 2019) and an engineered CCL20 locked dimer (CCL20LD) partial agonist of CCR6 (Getschman et al., 2017). Furthermore, CCL20LD was also efficacious in an interleukin-23 induced psoriatic arthritis model (Shi et al., 2021). Thus, blockade of CCR6 with PF-07054894 may have promise for the treatment of human interleukin-23 mediated inflammatory diseases such as psoriasis, psoriatic arthritis, and ulcerative colitis (Erichsen et al., 2020;Noviello et al., 2021;Atzeni et al., 2022). The effects of deleting or blocking CCR7 and CXCR2 in the interleukin-23 induced skin inflammation model are not known.
Functional Selectivity of PF-07054894 for CCR6 Relative to CCR7 and CXCR2. In vivo effects of PF-07054894 in mouse and cynomolgus monkeys were observed at trough plasma unbound concentrations of 2-8 nM (Supplemental Tables 3 and 4). The activity of PF-07054894 in chemotaxis assays is consistent with such exposures because PF-07054894 unbound concentrations greater than 1 nM cause a profound inhibition in human CCR6 chemotaxis ( Fig. 2A) at 0.5 nM CCL20 which is in the range of the concentrations observed in inflamed tissues (0.1-1 nM, see above). In contrast, no appreciable inhibition of CCR7 and CXCR2 chemotaxis was observed at such compound concentrations in the presence of relevant concentrations of CCL19 (0.01-1 nM) and CXCL1 (0.1-0.2 nM) (Fig. 2, B and C).
Functional selectivity of PF-07054894 for CCR6 versus CCR7 and CXCR2 has different possible implications. The potential impact of blocking CCR7 is not well understood. There are no clinical studies reporting the effects of inhibiting CCR7 in human subjects. CCR7 knockout mice reportedly develop higher numbers of T cells in the peripheral blood and lymphocytic infiltration in peripheral organs (F€ orster et al., 1999;Davalos-Misslitz et al., 2007). Under normal homeostatic  Table 4. Both dose regimens of PF-07054894 were efficacious as shown by the statistical analysis in Table 3. Number of animals per group: CCR6 knockout 5 8, BID cohorts 5 15; QD cohorts 5 10. Results are representative of three similar experiments. Data shown as mean and SD. (B) PF-07054894 at 100 mg/kg BID caused an increase in MFI in B cells (B220 1 lymphocytes). Cells from individual animals were stained for CCR6 expression. There were 10 animals per group, except in the naïve group, which had five animals. The graph shows CCR6 expression on B220 1 B cells as box plots; black diamonds indicate the mean values of each group. Statistics bars show the comparisons between the indicated groups performed using two-way ANOVA on the change from baseline with Holm's adjusted P values (* 5 P < 0.05; **** 5 P < 0.0001). The difference between indicated groups are as follows (estimate, 95% confidence interval): panel I, PF-07054894/no interleukin-23 versus Naïve 5 24.80 (4.17, 45.4); panel II, PF-07054894/no interleukin-23 versus Naïve 5 49.84 (29.21, 70.5) and PF-07054894/interleukin-23 versus Vehicle/interleukin-23 5 34.98 (18.1, 51.82). (C) Mouse splenocytes were incubated for 24 hours at 37 C with the indicated concentrations of PF-07054894, and cell surface expression of CCR6 was measured by flow cytometry. PF-07054894 caused an increase in CCR6 MFI on B cells (CD45 1 CD3 À B220 1 cells). Each data point shows the mean and S.D. of triplicate determinations. Results are representative of two similar experiments.
CCR6 Selective Antagonist in Homeostasis and Inflammation 89 at ASPET Journals on August 6, 2023 jpet.aspetjournals.org conditions CCR7 has been reported to mediate the exit of T lymphocytes from peripheral tissues (Debes et al., 2005). However, cd T cells that express CCR6 and interleukin-17 have been shown to enter the skin via the CCR6/CCL20 axis and to egress the skin in a CCR7-independent manner (Geherin et al., 2012). The egress of lymphocytes from sites of inflammation has been shown to be CCR7-dependent in some studies (Bromley et al., 2005;McNamee et al., 2013;G omez et al., 2015), and CCR7-independent in other studies (Brown et al., 2010;Vander Lugt et al., 2013), depending on the animal model and the experimental conditions. In addition, CCR7 has been reported to contribute to the homing of lymphocytes to secondary lymph organs and migration of immune cells to sites of inflammation. CCR7 mediated the migration of pathogenic dendritic cells into inflamed joints in a collagen induced arthritis mouse model (Moschovakis et al., 2019). In a mouse experimental autoimmune encephalomyelitis model, CCR7 played a key role in the priming of autoimmune T cells (Belikan et al., 2018). Deleting or neutralizing CCR7 was effective in treatment of an acute graft versus host disease mouse model by reducing lymphoid organ infiltration of donor CCR7 1 cells (Cuesta-Mateos et al., 2020). Thus, the literature is inconclusive regarding the potential impact of inhibiting CCR7mediated T cell egress from tissues, and blockade of CCR7 might be beneficial in inflammation.
CXCR2 antagonists have been pursued to block inflammatory neutrophil infiltration and tested in asthma and chronic obstructive pulmonary disease clinical trials, which showed reduction in neutrophil migration to the lung (Nair et al., 2012;Todd et al., 2016;Watz et al., 2017) with modest clinical efficacy (Nair et al., 2012;Rennard et al., 2015;O'Byrne et al., 2016;Lazaar et al., 2020). CXCR2 antagonist-related lowering of blood neutrophil counts was observed in several independent studies (Nair et al., 2012;Seiberling et al., 2013;Hastrup et al., 2015;Jurcevic et al., 2015;Rennard et al., 2015;Todd et al., 2016;Watz et al., 2017), which is consistent with the role of CXCR2 in releasing neutrophils from bone marrow into the circulation (Eash et al., 2010). Thus, blockade of CXCR2 might have beneficial anti-inflammatory effects but it may also cause undesired neutropenia. The risk of observing neutropenia upon administration of PF-07054894 at therapeutic doses is minimal because the compound shows at least 150-fold selectivity for CCR6 relative to CXCR2.
In conclusion, PF-07054894 is a novel CCR6 antagonist with slow offset kinetics and insurmountable inhibition of CCL20/CCR6 chemotaxis. This attractive pharmacological profile may ensure sustained CCR6 blockade leading to inhibition of pathogenic cell migration to inflammatory sites driven by high local levels of CCL20.