Recruitment-MP were formulated to produce CCL22 release kinetics such that a physiological gradient of CCL22 could be established for effective T reg recruitment in vivo. These microparticles were made out of a biocompatible, biodegradable poly(lactic-co-glycolic acid) (PLGA) polymer using a well-established double emulsion solvent evaporation technique (). Scanning electron micrographs of MP indicate that they are spherical and slightly porous (fig. S1A, inset). The surface of Recruitment-MP was formulated to be porous to allow continuous release (without periods of lag) of chemokine. Further, particles were designed to be large enough to avoid uptake by phagocytic cells (i.e., >10 μm) and to prohibit their movement across the vascular endothelium, with consequent immobilization at the site of placement (fig. S1B). Last, Recruitment-MP released CCL22 in a linear manner over a period of 40 days (fig. S1A).
To test the ability of Recruitment-MP to prevent graft rejection in a clinically relevant model of VCA, vascularized hindlimbs were transplanted from Brown Norway (BN) rat donors to Lewis (LEW) recipients (complete MHC mismatch). Animals receiving only the baseline immunosuppression protocol of FK506/ALS (anti-rat lymphocyte serum) (timing presented in fig. S1C) served as controls. These animals consistently rejected grafts 2 to 3 weeks after FK506 was discontinued at postoperative day (POD) 21 ( Fig. 1A). Three doses of Recruitment-MP (9, 50, and 100 mg) were tested to determine an effective gradient necessary for recruiting T reg and prolonging graft survival. Minutes after hindlimb transplants were completed, particles were administered by a single subcutaneous injection on the lateral aspect of the hindlimb (to avoid the vascular anastomosis medially) after blunt dissection of the subcutaneous plane with an 18-gauge needle (to allow for better distribution of the particles). MP were administered similarly 21 days after transplant. Animals receiving 9 and 100 mg of Recruitment-MP had a graft mean survival time (MST) of 41.5 and 44.0 days, respectively ( Fig. 1A). Treatment with 50 mg of Recruitment-MP significantly prolonged graft survival indefinitely, with long-term survival >200 days in six of eight animals (75%). Furthermore, animals treated with 50 mg of blank MP (MST, 39 days), soluble CCL22 (injected subcutaneously in the allograft) (MST, 37.5 days), and 50 mg of Recruitment-MP injected in the contralateral (nontransplanted) limb (MST, 36.0 days) did not exhibit prolonged graft survival compared to animals receiving only the baseline immunosuppression protocol.
With Recruitment-MP promoting allograft survival, as observed macroscopically, we next examined allograft tissues histologically. Rejecting grafts exhibited sloughing of the epidermis and substantial mononuclear cell infiltration in the dermis and perivascular regions ( Fig. 2A). Rejecting grafts also exhibited substantial myositis as evidenced by mononuclear cell infiltration in muscle tissue and disruption of tissue architecture ( Fig. 2B). Conversely, biopsies from Recruitment-MP-treated (50 mg) animals with long-term surviving allografts (>200 days) showed minimal cellular infiltration and intact tissue architecture, similar to muscle and skin biopsies from normal animals ( Fig. 2, A and B). At earlier time points when allografts rejected in control animals, Recruitment-MP-treated hindlimbs exhibited noticeable mononuclear cell infiltrates in the dermis (albeit less than in rejecting limbs) and low to moderate epidermal hyperplasia ( Fig. 2A); however, the epidermis remained intact, and there was substantially less evidence of cellular infiltrates in muscle tissue ( Fig. 2B).
To characterize cellular infiltrates, skin from Recruitment-MP-treated and rejecting control allografts, as well as autologous skin from contralateral hindlimbs, was dissociated into single-cell suspensions and analyzed by flow cytometry. Rejecting allograft skin contained slightly more total leukocytes (live CD45+ cells) and CD8+ cytotoxic T cells than skin from Recruitment-MP-treated allografts, although the differences were not significant. CD4+ FoxP3 − effector T cells were significantly reduced in Recruitment-MP-treated grafts, while CD4+ FoxP3+ T reg were found at comparable levels. Autologous skin had significantly fewer total leukocytes and T cells compared to rejecting and Recruitment-MP-treated allografts. The proportion of T reg (% FoxP3+ of total CD4+ or CD8+ T cells) in skin from Recruitment-MP-treated allografts was significantly increased compared to that in both rejecting allografts and autologous skin ( Fig. 2C).
To determine whether treatment with Recruitment-MP could suppress inflammation locally in the context of VCA, intragraft full-thickness skin samples and draining lymph nodes were harvested from actively rejecting animals (grade III to IV rejection) and Recruitment-MP-treated VCA recipients (experimental end point; POD >200). We then measured expression of proinflammatory genes: TNF-α, IFN-γ, IL-17A, Perforin-1, and Serglycin. Expression of all five genes (normalized to expression in normal tissue) was decreased significantly in skin biopsies from Recruitment-MP-treated VCA recipients when compared to corresponding tissue samples from actively rejecting grafts ( Fig. 3A). In draining lymph nodes of Recruitment-MP-treated VCA recipients, expression of TNF-α, IL-17A, Perforin-1, and Serglycin was also significantly decreased compared to actively rejecting animals ( Fig. 3B).
To assess potential mechanisms underlying the enhanced allograft survival associated with Recruitment-MP, local phenotypic changes in the draining lymph nodes of animals with long-term surviving grafts were examined. At the experimental end point (grade III rejection or POD > 200 for long-term survivors), draining and nondraining inguinal lymph nodes were harvested, and the local CD4+ T helper cell phenotype was analyzed. Allograft draining and nondraining lymph nodes from Recruitment-MP-treated animals exhibited a decreased incidence of inflammatory CD4+IFN-γ+ cells (normalized to percentages from naïve animals) when compared to draining lymph nodes from actively rejecting animals. Further, draining lymph nodes from actively rejecting animals demonstrated a higher percentage of CD4+IFN-γ+ cells than from nondraining lymph nodes of actively rejecting animals ( Fig. 4B). Allograft draining lymph nodes from Recruitment-MP-treated animals exhibited an increased incidence of CD4+CD25 hiFoxP3+ cells (normalized to percentages from naïve animals) when compared to both nondraining lymph nodes from Recruitment-MP-treated animals and draining lymph nodes from actively rejecting animals ( Fig. 4A).
The suppressive and proliferative capacity of T reg and conventional T cells (T conv) isolated from Recruitment-MP-treated VCA recipients and naïve LEW rats were measured in mixed leukocyte reaction (MLR). Specifically, splenocytes from both Recruitment-MP-treated VCA recipients and naïve animals were flow-sorted into two groups: CD4+CD25 hi (T reg) or CD4+CD25 − (T conv). T conv from both sets of animals were then cultured with irradiated BN (donor) stimulator splenocytes. There was no observed immune hyporesponsiveness with T conv isolated from Recruitment-MP-treated animals when compared to normal controls ( Fig. 4C). To assess the suppressive function of T reg from Recruitment-MP-treated animals, T reg from naïve or Recruitment-MP-treated animals were cocultured with T conv from naïve LEW (syngeneic) rats and irradiated BN (donor) splenocytes. T reg isolated from Recruitment-MP-treated animals were more effective than naïve T reg at inhibiting proliferation of naïve T conv stimulated with BN splenocytes ( Fig. 4D).
To test the donor antigen specificity of CD4+CD25 hi T reg isolated from Recruitment-MP-treated VCA recipients, additional MLRs were performed. Naïve LEW CD4+CD25 − T conv were cocultured with CD4+CD25 hi T reg from Recruitment-MP-treated VCA recipients and stimulated with irradiated BN (donor) or Wistar Furth (WF; third party, complete MHC mismatch) splenocytes. T reg from long-term surviving grafts showed enhanced suppressive function ( P < 0.05) against BN stimulation compared to WF stimulation on a per-cell basis ( Fig. 5A).
To test whether Recruitment-MP could impart donor antigen-specific dominant tolerance, animals with long-term surviving allografts (>200 days) were challenged with nonvascularized skin allografts from LEW (syngeneic), BN (donor), and WF (third party). These animals received no further immunosuppression or microparticle treatment beyond POD 21. All three animals acutely rejected skin grafts from WF animals, as evidenced by a lack of graft take, characterized by necrosis, wound contraction, and scarring. However, animals accepted both LEW and BN grafts, with minimal wound contracture and eventual hair regrowth ( Fig. 6, B and C).
Given promising results with CCL22 delivery in the rat hindlimb VCA model, we sought to determine whether chemically synthesized CCL22 (with potential advantages for clinical translation) could serve as an alternative to recombinant CCL22 to selectively enrich human T reg. Migration of human CD3+ T cells from peripheral blood toward human CCL22 was evaluated using in vitro transwell migration assays. Both recombinant and synthetic CCL22 promoted specific migration of human CD4+ CD25+ CD127 low T cells from a diverse pool of total CD3+ T cells in a dose-dependent manner ( Fig. 6), and selective FoxP3 expression by these cells confirmed their identity as T reg ( Fig. 6C). Compared to CD4+ CD25+/− CD127+ helper T cells and CD3+ CD4 − T cells, which include CD8+ T cells and a smaller population of γδ T cells, T reg expressed higher levels of CCR4 ( Fig. 6C). This differential expression of CCR4 resulted in enrichment of T reg in the migrating fraction ( Fig. 6A), with T reg frequencies among migrating T cells enhanced by as much as 1.6-fold compared to starting populations of total CD3+ T cells. While the greatest T reg enrichment corresponded to the greatest CCL22 concentration (1000 ng/ml) ( Fig. 6A), T reg migration indices, which represent CCL22-induced T reg migration, appear to peak at a lower concentration of 100 ng/ml ( Fig. 6B). Human T reg migrated comparably toward both recombinant and synthetic CCL22.
Six- to 8-week-old male LEW, BN, and WF rats (Charles River Laboratories, Wilmington, MA) were used. Animals were maintained under an Institutional Animal Care and Use Committee-approved protocol in a specific pathogen-free environment at the University of Pittsburgh.
Using techniques developed in the University of Pittsburgh's Department of Plastic Surgery (), hindlimbs from donor BN rats were transplanted orthotopically to recipient LEW rats. Briefly, donor femoral vessels were anastomosed end-to-end to recipient femoral vessels using 10-0 nylon sutures, and femoral osteosynthesis was performed with an 18-gauge intramedullary rod. Attachment of muscles and skin closure were achieved with simple interrupted sutures. Surgical time was approximately 2 to 3 hours. Anesthesia was achieved perioperatively with inhaled isoflurane, and buprenorphine was used for analgesia.
PLGA MPs containing recombinant mouse CCL22 (R&D Systems, Minneapolis, MN) were prepared using a standard water-oil-water double emulsion procedure, as described (). Briefly, PLGA (RG502H; Boehringer Ingelheim, Petersburgh, VA) MP were prepared by mixing 200 μl of an aqueous solution containing 25 μg of rmCCL22 and 2 mg of bovine serum albumin and 15 mmol NaCl with 200 mg of polymer dissolved in 4 ml of dichloromethane. The first water-in-oil emulsion was prepared by sonicating this solution for 10 s. The second oil-in-water emulsion was prepared by homogenizing (Silverson L4RT-A) this solution with 60 ml of an aqueous solution of 2% polyvinyl alcohol (molecular weight, ~25,000 Da; 98% hydrolyzed; PolySciences, Warrington, PA) for 60 s at 3000 rpm. This solution was then mixed with 1% polyvinyl alcohol and placed on a stir plate agitator for 3 hours to allow the dichloromethane to evaporate. The microparticles were then collected and washed four times in deionized (DI) water, to remove residual polyvinyl alcohol, before being resuspended in 5 ml of DI water, frozen, and lyophilized for 72 hours (VirTis BenchTop K freeze dryer, Gardiner, NY; operating at 100 mtorr).
Surface characterization of microparticles was conducted using scanning electron microscopy (JEOL JSM-6510LV/LGS), and microparticle size distribution was determined by volume impedance measurements on a Beckman Coulter counter (Multisizer 3, Beckman Coulter, Fullerton, CA). CCL22 release from microparticles was determined by suspending 7 to 10 mg of microparticles in 1 ml of phosphate-buffered saline placed on an end-to-end rotator at 37°C. CCL22 release sampling was conducted by centrifuging microparticles and removing the supernatant for CCL22 quantification using enzyme-linked immunosorbent assay (ELISA) (R&D Systems, Minneapolis, MN) ().
All hindlimb recipients in all groups received the same baseline immunosuppression protocol consisting of 21 days of tacrolimus (LC Laboratories, Woburn, MA) at a dose of 0.5 mg/kg per day, injected intraperitoneally from the time of transplant. Rats also received two doses of rabbit ALS (Accurate Chemical, Westbury, NY) injected intraperitoneally on day −4 and POD 1 (fig. S1C). MP were injected via an 18-gauge needle after subcutaneously undermining dermal tissue to allow for even spreading of the particles. Further, MP were injected in the lateral aspect of the transplanted limb (unless otherwise noted) immediately following hindlimb transplantation and again on POD 21. Animals given transplants were allocated into groups consisting of the following treatments: (i) FK506/ALS baseline immunosuppression only ( N = 4); (ii) 9 mg of Recruitment-MP (10 mg/ml; N = 6); (iii) 50 mg of Recruitment-MP (50 mg/ml; N = 8); (iv) 100 mg of Recruitment-MP (50 mg/ml; N = 6); (v) 50 mg of blank MP ( N = 3); (vi) soluble CCL22 assuming 100% encapsulation efficiency of group 3 ( N = 4); (vii) 50 mg of Recruitment-MP injected in the contralateral, native naïve (nontransplanted) limb ( N = 4).
To assess rejection, hindlimbs were monitored daily and scored for rejection (appearance grading) based on physical examination (). Limbs were given a daily score using the following scale: grade 0 (no rejection), grade I (edema), grade II (erythema and edema), grade III (epidermolysis), and grade IV (necrosis and "mummification"). Grafts were considered rejected when displaying signs of progressive grade III rejection.
Skin and muscle samples were obtained from the transplanted limbs at their experimental end point: progressive grade III rejection, early nonrejecting (POD 29 to 33), or long-term survival (>200 days). Samples were fixed in 10% neutral buffered formalin, paraffin-embedded, sectioned at 5 μm, and stained with hematoxylin and eosin (H&E) for microscopic examination of tissue architecture and mononuclear cell infiltration.
Draining and nondraining (contralateral) inguinal lymph nodes were harvested at the experimental end point (progressive grade III rejection or long-term survival >200 days). Lymph nodes were then processed to form single-cell suspensions. Cells were stained with fluorescently labeled antibodies for CD4 (clone OX-35), CD25 (OX-39), FoxP3 (FJK-16s), and IFN-γ (DB-1) (BD Biosciences, San Jose, CA; eBioscience, San Diego, CA). The FoxP3/Transcription Factor Staining Buffer Set (eBioscience) was used for intracellular staining. Before IFN-γ cytokine staining, cells were cultured overnight with Cell Stimulation Cocktail (plus protein transport inhibitor; eBioscience). Skin samples (1 cm × 1 cm) were cut into thin strips for enzymatic digestion in culture medium supplemented with collagenase D (1 mg/ml) and deoxyribonuclease (DNase) (1 mg/ml). Tissue was incubated for 2 hours at 37°C and 100 rpm shaking and then mechanically dissociated through 70-μm nylon filters. Cells were stained with antibodies for CD4 (OX-35), CD8a (OX-8), CD45 (OX-1), and FoxP3 (FJK-16s), as well as Fc block (anti-CD32, D34-485) and a fixable viability dye (eBioscience). Stained lymph node and skin cells were analyzed on a BD LSR Fortessa cytometer, and CountBright Absolute Counting Beads (Thermo Fisher Scientific, Waltham, MA) were added to skin cell suspensions to determine total cell counts. Results were analyzed using FlowJo (Ashland, OR).
Gene expression profiles of inflammatory markers were evaluated in the skin and lymph nodes of long-term graft survivors, actively rejecting animals, and naïve rats. Total RNA was extracted from samples using TRI Reagent according to the manufacturer's instructions and quantified using a NanoDrop 2000 spectrophotometer. For each reverse transcriptase assay, 4 μg of RNA was converted to complementary DNA using a QuantiTect Reverse Transcription kit. Quantitative real-time polymerase chain reaction (qRT-PCR) was then performed using VeriQuest Probe qPCR Mastermix, according to the manufacturer's instructions, with 5′ nuclease PrimeTime qPCR assays specific for IFN-γ (Rn00594078_m1 Dye: VIC-MGB_PL), TNF-α (Rn99999017_m1 Dye: VIC-MGB_PL), Perforin-1 (Rn00569095_m1 Dye: VIC-MGB_PL), Serglycin (Rn00571605_m1 Dye: VIC-MGB_PL), IL-17 (Rn01757168_m1 Dye: VIC-MGB_PL), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH; endogenous control, Rn99999916_s1). Duplex reactions (target gene + GAPDH) were run and analyzed on the StepOnePlus Real-Time PCR System. Relative fold changes of IFN-γ, TNF-α, Perforin-1, Serglycin, and IL-17 expression were calculated and normalized based on the 2 method, and then further normalized to normal control. Skin biopsies from naïve animals or contralateral limbs served as untreated controls.
Spleens from rats with long-surviving hindlimb grafts and naïve LEW rats were processed into single-cell suspensions. Red blood cells (RBCs) were lysed using RBC lysis buffer (Thermo Fisher Scientific). CD4+ T cells were isolated using CD4 T cell enrichment columns according to the manufacturer's instructions (Miltenyi Biotec, Auburn, CA). CD4+ enriched cells were then stained with anti-CD4 (OX-35) and anti-CD25 (OX-39). CD4+CD25 − (T conv) and CD4+CD25 hi (T reg) populations were sorted using a BD FACSAria cell sorter. To assess proliferative function, CD4+CD25 − (T conv) from long-term surviving and naïve rats were stained with violet proliferation dye (VPD450; BD Biosciences) and each was cocultured/stimulated with irradiated splenocytes harvested from donor strain rats at 2:1 T conv:stimulators. At the end of the 7-day MLR period, proliferation was measured via VPD450 dilution by flow cytometry. The proliferative capacity of T conv from long-term surviving rats was normalized to that of naïve rats.
To quantify the suppressive cell function of CD4+CD25 hi (T reg) isolated from long-term surviving and normal control rats, we tested their ability to suppress T conv proliferation in MLR. CD4+CD25 − T conv from naïve rats were stained with VPD450 and cocultured/stimulated with irradiated splenocytes from BN rats and CD4+CD25 hi T reg from either long-term surviving or naïve rats at 2:1 T conv:T reg. At the end of the 7-day culture period, proliferation was measured via VPD450 dilution by flow cytometry. Percent suppression was calculated using the following formula: {1 - [(% Proliferation of naïve T conv cultured with BN splenocytes and T reg)/(% Proliferation of naïve T conv cultured with BN splenocytes)]} × 100%.
MLRs were also set up to test for antigen specificity of the CD4+CD25 hi T reg isolated from long-term surviving rats. CD4+CD25 − T conv from naïve rats were stained with VPD450 and stimulated with either BN or WF (third party) irradiated splenocytes in the presence of CD4+CD25 hi T reg isolated from long-term surviving rats. At the end of the 7-day culture period, proliferation was measured via VPD450 dilution using FlowJo (Ashland, OR). Percent suppression was calculated using the following formula: {1 - [(% Proliferation of naïve T conv cultured with BN or WF splenocytes and T reg)/(% Proliferation of naïve T conv cultured with BN or WF splenocytes)]} × 100%.
Donor antigen-specific tolerance was assessed in vivo in long-surviving graft recipients (>200 days) from the 50-mg Recruitment-MP group via skin graft challenge. Skin allografts were harvested from normal donor strain (BN) or third-party strain rats (WF) and transplanted to the dorsal thoracic area of long-term survivors >200 days after VCA. Grafts were supported in place with xeroform and surgical gauze for 5 days and subsequently evaluated daily for signs of rejection. Rejection was defined as 80% necrosis of the skin graft.
Healthy donor human peripheral blood was obtained from the Central Blood Bank (Pittsburgh, PA), and peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Paque Plus (1.077 g/ml; GE Healthcare, Chicago, IL) gradient centrifugation. Total CD3+ T cells were isolated from PBMCs by immunomagnetic negative selection with an EasySep Human T Cell Isolation kit (StemCell Technologies, Cambridge, MA). Chemotaxis assays were conducted in 24-well plates with 5-μm pore polycarbonate transwell filters (Corning Costar, Corning, NY), as previously reported (). Serum-free AIM V medium (600 μl; Thermo Fisher Scientific) containing recombinant human CCL22 (0, 0.1, 1, 10, 100, or 1000 ng/ml) (R&D Systems), or synthetic human CCL22 (Almac, Souderton, PA), was added to the lower receiver wells. T cells suspended in serum-free AIM V medium were added to the top chambers of the transwells (100 μl, 5 × 10 5 cells per well). For total cell "input" controls, cells were placed directly in receiver wells. After incubating at 37°C with 5% CO 2 for 2 hours, migrating cells in the lower wells were recovered and transferred to fluorescence-activated cell sorting (FACS) tubes for staining. Cells were stained with fluorescently labeled antibodies for CD4 (RPA-T4; BD Biosciences), CD25 (2A3; BD Biosciences), CD127 (hIL-7R-M21; BD Biosciences), and a fixable viability dye (eBioscience). Some cells were also stained for CCR4 (1G1; BD Biosciences) and FoxP3 (236A/E7; eBioscience), or mouse IgG1,κ isotype controls, using the eBioscience FoxP3/Transcription Factor Staining Buffer Set. Immediately before analysis on a BD LSR II flow cytometer, CountBright Absolute Counting Beads were added to each tube. Frequencies of T reg (CD4+ CD25+ CD127 low) in the migrating populations were calculated. To determine total numbers of migrating T cells, cell counts were normalized to total counting beads. T reg chemotaxis migration indices were calculated by normalizing numbers of T reg migrating in the presence of CCL22 to the average number of spontaneously migrating T reg in the absence of CCL22.
Statistical analyses were performed in GraphPad Prism v8. All data are expressed as means ± SD, and significant differences between experimental groups were determined by two-tailed Student's t test, Welch's t test, or Mann-Whitney U test for two independent samples, or one-way analysis of variance (ANOVA) followed by Tukey's post hoc tests for comparison of multiple groups. For human T reg migration assays with two types of CCL22 and multiple doses, results were analyzed by two-way ANOVA, followed by Dunnett's pairwise comparisons for dose effect. A one-sample t test was used to compare fold changes in an experimental group to a value of one, and the effects of various treatments on VCA survival were analyzed using a log-rank test. Differences were considered significant if P < 0.05.
Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/6/11/eaax8429/DC1
Fig. S1. Characterization of Recruitment-MP and experimental timeline for hindlimb transplantation.
Fig. S2. Skin flow cytometry gating strategy used for Fig. 2C.
Fig. S3. Representative images of a hindlimb VCA recipient challenged with LEW, BN, and WF nonvascularized skin grafts.
All Rights Reserved © 2024
