A highly potent lymphatic system‐targeting nanoparticle‐cyclosporine with efficacy in a lupus mouse model

Objective Cyclosporine A (CsA) is a powerful immunosuppressant that modulates both humoral and cell‐mediated responses, but its efficacy in treating systemic lupus erythematosus (SLE) ‐ a disease that involves lymphatic dysregulation and autoimmunity, has never been demonstrated. As a stand‐alone drug, CsA is constrained by poor gut absorption, low target‐tissue distribution, and high renal toxicity. We aim to overcome these limitations using nanoparticle delivery, with the hypothesis that improved bioavailability of nanoparticle‐encapsulated CsA will increase drug potency, leading to measurable, positive outcomes in a mouse SLE disease model. Methods Biodegradable nanoparticles were synthesized from repeating units of PLA and PEG. The polymeric backbone was further functionalized with periodically‐spaced carboxyl groups, each of which conjugated to the small molecule xanthanoid gambogic acid (GA), in order to create P2Ns‐GA nanoparticles. Unconjugated nanoparticles, referred to as P2Ns, were also created to serve as controls. GA, being exposed on the nanoparticle surface, acted as a targeting ligand for CD71 (Transferrin Receptor 1), a type II transmembrane protein expressed by enterocytes and gut lymphoid structures such as Peyer’s patches, as well as by various lymphoid and myeloid‐lineage leukocytes. Fluorescent, drug‐free nanoparticles were used to ascertain cellular and molecular binding partners based on in vitro and ex vivo assays. To determine efficacy in the mouse SLE model, CsA was encapsulated in GA‐conjugated nanoparticles, referred to as P2Ns‐GA‐CsA, and orally dosed to MRL‐ lpr mice – a strain with aggressive lupus‐like condition. P2Ns‐CsA, a nanoparticle formulation without the targeting ligand, and Neoral ® , a commercially available CsA formulation, served as comparisons. Results Fluorescent P2Ns‐GA associated more with CD3 + T cells and CD20 + B cells than P2Ns did. The GA‐conjugated nanoparticles further exhibited high affinity to intestinal lacteals and Peyer’s patch regions, suggesting that utilization of the gut lymphatic transport network was possible. With P2Ns‐GA‐CsA, we realized a 4–8 fold gain in the lymph node delivery of CsA compared to other, orally delivered drug modalities such as P2Ns‐CsA and Neoral ® . Due to the importance of lymphoid organs in the pathogenesis of lupus, increased lymphatic bioavailability of P2Ns‐GA‐CsA was paralleled by substantial reduction in virtually every aspect of SLE disease progression. Features of P2Ns‐GA‐CsA treatment included reduced lymphadenopathy and splenomegaly, near normalization of plasma cytokines, and the disappearance of nephritic hallmarks of SLE. We were also able to circumvent drug‐related renal toxicity with P2Ns‐GA‐CsA in large part because the administered dose (5 mg/kg) was significantly below the nephrotoxic dose (20 mg/kg) reported in literature. In summary, this study highlights the potential of drug repurposing made possible by nanoparticles that allow enhanced targeting of disease‐causing tissues, while sparing healthy ones from drug‐related complications.In vitro and ex vivo association of nanoparticles with peripheral blood mononuclear cells (PBMCs) and intestinal tissue sections. A) Confocal Imaging of mouse (C57BL/6J) PBMCs in vitro association with fluorescent (FITC, green) P2Ns‐GA or P2Ns nanoparticles. GA‐conjugation, which targets CD71 (Pe‐Cy7, purple), results in more robust nanoparticle association with CD3 + T cells (APC, red) and CD20 + B cells (APC, red) (Bar = 5 μm, for all images). B) Mander’s Overlap Coefficient (MOC, a measurement of fluorescent signal co‐localization) shows greater signal overlap between GA‐conjugated nanoparticle (P2Ns‐GA) and CD71, compared to either CD3 or CD20. In contrast, unconjugated nanoparticle (P2Ns) showed no preferential overlap with CD71, CD3, or CD20. C) Ex vivo , de‐paraffinized tissue sections from MRL‐lpr mouse small intestine appeared to show more P2Ns‐GA infiltration of lacteals (C, top panel) and Peyer’s Patches (D, top panel) compared to P2Ns (C & D, bottom panel). Furthermore, triple co‐localization of nanoparticles, CD3, and CD71 was qualitatively observed for P2Ns‐GA (C, top panel, zoomed image with arrows), whereas no such triple co­localization was apparent in the P2Ns ex vivo assay (C, bottom panel, zoomed image with arrows) (Bar = 50 μm, for all images). *p < 0.05, ns: not significant; comparisons were made with one‐way ANOVA followed by Tukey multiple comparison test (n = 3 PBMC replicates, n = 6 individual tissue sections).Analyses of lymphoid organ hypertrophy and concentration of plasma cytokines. Lymphoid tissues from five groups of mice were analyzed post‐euthanasia (MRL‐ lpr . mice with SLE‐like disease but not receiving any treatment; P2Ns‐CsA: MRL‐ lpr mice treated with unconjugated P2Ns‐CsA; P2Ns‐GA‐CsA: MRL‐ lpr mice treated with GA‐conjugated P2Ns‐GA‐CsA; Neoral ® : MRL‐ lpr mice treated with commercially available CsA formulation, from Norvatis International AG; MRL mice without SLE‐like disease but otherwise under the same parental background as MRL‐ lpr . All treatments were delivered orally via gavage at 5mg/kg/24HRs for Neoral ® and 5mg/kg/48HRs for P2Ns‐GA‐CsA and P2Ns‐CsA) Significant lymphoadenopathy (A&C, blue arrows) and splenomegaly (B&D) were associated with untreated MRL‐ lpr in addition to thymic hypertrophy (E). SLE‐like lymphoproliferation in MRL‐ lpr was accompanied by drastic increases in the concentrations of multiple plasma cytokines/chemokines (F) essential to the survival, differentiation, and localization of not only T‐cells, but also B‐cells, granulocytes, monocytes, and macrophages. Aggregated increase in plasma cytokines (represented as log 2 FC from baseline) was plotted in (G) to illustrate near‐global increase in the concentrations of analyzed cytokines/chemokines in the MRL‐ lpr group (F, left‐most column). Treatment with Neoral ® (F, 4 th column from left) was generally not effective in reducing elevated cytokine/chemokine concentrations, but P2Ns‐GA‐CsA (F, middle column) resulted in almost complete normalization of cytokine/chemokine in the plasma. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant; comparisons were made with one‐way ANOVA followed by Tukey multiple comparison test (n = 6–8 tissues, and n = 3 multiplex cytokine/chemokine plasma samples).Reductions in systemic inflammatory signaling and lymphoid dysregulation was paralleled by improved CsA drug delivery to the lymph nodes. Treatment groups were analyzed post‐euthanasia for the concentration of CsA delivered to lymphatic tissue. Because of relative enlargement of lymph nodes in the MRL‐ lpr background, avascular cervical and axillary lymph nodes were used to estimate lymphatic drug delivery in mice. CsA concentrations were determined with mouse CsA ELISA in plasma, lymph nodes, kidney and spleen (A). Significant, 4–8 fold increase in the drug’s plasma and lymph node bioavailability was achieved with P2Ns‐GA‐CsA nanoparticles compared to other oral CsA treatments. Secondary complications of SLE were also prominent in MRL‐ lpr untreated mice, these included: higher plasma anti‐dsDNA igG level, a predictor of lupus nephritis (B), higher lymphocyte absolute count, and gradually worsening renal function as indicated by BUN and serum creatinine (D), as well as abundant igG deposition at the mesangial and periglomerular regions of the kidneys (E) (Bar = 50 μm, for all images). Treatments with Neoral ® and P2Ns‐CsA appeared not to have exacerbated renal pathologies (D and E), but they were marginally beneficial in slowing down nephritis progression. In contrast, P2Ns‐GA‐CsA was the only treatment that significantly normalized all of the assayed parameters (B–E). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant; comparisons were made with one‐way ANOVA followed by Tukey multiple comparison test (n = 4–8 individual plasma, serum, lymph node, kidney, and spleen)Further analyses of renal damage resulting from SLE‐like MRL‐ lpr mutation. (A) Histological images showing that renal damage in the untreated M RL‐ lpr appeared to have a strong correlation with lymphocyte infiltration, suggesting that a bulk of the damage can be attributed to maladaptive acute and chronic inflammatory response. Both CD20 + and CD3 + cells, (which P2Ns‐GA nanoparticles had demonstrated strong association with), were enriched in the glomeruli (CD20 and CD3) as well as the interstitial space (CD3), covering large portions of proximal and distal tubules, and the collecting ducts; thereby likely causing increased necrosis/apoptosis‐related tissue loss (A–C) (Bar = 50 μm, for first two rows of images; Bar = 200 μm, for third row of images). SDF‐1α/CXCL12 and BCA‐1/CXCL13, two strongly upregulated chemokines in the plasma, appeared to have helped recruit immune cells with cognate receptors CXCR4 and CXCR5 to the renal parenchyma. Literature evidence point to lupus nephritis of both high severity and high mortality when CXCR4 + B cells, and CXCR5 + double‐negative T‐cells become part of the infiltrate. Here, we observed elevated transcript levels of CXCR4 and CXCR5 in the MRL‐ lpr untreated mice (D&E). Renal histology also showed significant upregulation of CXCR4 (brown) among infiltrating leukocytes as well as endogenous expression in the proximal tubules. CXCR5 + population (red) appeared to decorate the migrating front of infiltrating leukocytes (F) (Bar = 200 μm, for all images). All of these severe hallmarks of nephritis were effectively attenuated by P2Ns‐GA‐CsA treatment, based only qPCR and histology data (A–F). By contrast, comparable oral treatments with Neoral ® and P2Ns‐CsA were much less effective (A–F) in counteracting renal inflammation and consequent tissue damage. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.000; comparisons were made with one‐way ANOVA followed by Tukey multiple comparison test (n = 5–8 individual kidneys).