There is growing evidence that extracellular haemoglobin and haem mediate inflammatory and oxidative damage in sickle cell disease. Haptoglobin (Hp), the scavenger for free haemoglobin, is depleted in most patients with sickle cell disease due to chronic haemolysis. Although single infusions of Hp can ameliorate vaso-occlusion in mouse models of sickle cell disease, prior studies have not examined the therapeutic benefits of more chronic Hp dosing on sickle cell disease manifestations. In the present study, we explored the effect of Hp treatment over a three-month period in sickle mice at two dosing regimens: the first at a moderate dose of 200 mg/kg thrice weekly and the second at a higher dose of 400 mg /kg thrice weekly. We found that only the higher dosing regimen resulted in increased haem-oxygenase-1 and heavy chain ferritin (H-ferritin) expression and decreased iron deposition in the kidney. Despite the decreased kidney iron deposition following Hp treatment, there was no significant improvement in kidney function. However, there was a nearly significant trend towards decreased liver infarction.

Gene therapy for sickle cell disease is currently in active trials. Collecting hematopoietic progenitor cells safely and effectively is challenging, however, because granulocyte colony stimulating factor, the drug used most commonly for mobilization, can cause life-threatening vaso-occlusion in patients with sickle cell disease, and bone marrow harvest requires general anesthesia and multiple hip bone punctures. Plerixafor is an inhibitor of the CXCR4 chemokine receptor on hematopoietic progenitor cells, blocking its binding to SDF-1 (CXCL12) on bone marrow stroma. In support of a clinical trial in patients with sickle cell disease of plerixafor mobilization (NCT02193191), we administered plerixafor to sickle cell mice and found that it mobilizes hematopoietic progenitor cells without evidence of concomitant cell activation or brain vaso-occlusion.

Introduction: Haptoglobin (Hp), the scavenger for hemoglobin, and hemopexin (Hx), the scavenger for heme, are depleted in most patients with sickle cell disease due to chronic hemolysis. There is mounting evidence of the crucial role of free hemoglobin and/or free heme in mediating inflammatory and oxidative damage in sickle cell disease, including vaso-occlusion and acute chest syndrome. Purified Hp has been used in Japan for a variety of hemolytic conditions and has been proposed as a potential treatment for sickle cell disease. Although infusions of Hp or Hx have been shown to ameliorate vaso-occlusion, acute lung injury, and heme toxicity in sickle cell mouse models, no prior studies have examined the utility of chronic Hp treatment for amelioration of organ damage. We therefore studied the effect of 3 months of chronic Hp treatment in the Townes sickle mouse model. Methods: Male and female Townes mice (Stock number 013071, The Jackson Laboratory) were used for all experiments, starting at 1 or 3 months of age. SS genotype was confirmed by PCR and HPLC. Organ damage in the spleen, liver, and kidneys as previously described was confirmed. Human Hp solution was a kind gift from Bio Product Laboratory (BPL, Hertfordshire, UK). Hp or equivalent volume PBS control was administered intraperitoneally (IP) in the first cohort of 5 mice and then subcutaneously (SC) in the next two cohorts of 7 and 12 mice on a 48-72 hr dosing schedule of Monday, Wednesday, Friday for a period of 3 months. At the end of 3 months treatment, mice were evaluated by the following studies (with concurrent blinding to treatment group for most studies): plasma Hp (ELISA), plasma heme (QuantiChrom heme assay), urine osmolality (osmometer), urine albumin (ELISA), CBC (Advia 120), WBC differential (Advia 120 and manual count), red blood cell ektacyometry (ektacyometer), organ mass (percent of body weight), and organ histology. Results: Mouse Hp levels in SS Townes mice were confirmed to be markedly low compared to Townes AA mice (mean ± SD: SS 2 ± 1 versus AA 39 ± 4 ug/mL). Dose-finding experiments determined that a dose of 200-400 mg/kg IP or SC in SS mice resulted in a 24 hr peak concentration that was 5-14X supraphysiologic, variably physiologic at 48 hr, and absent or almost absent at 72 hr. Chronic dosing at the 400 mg/kg IP in SA mice showed no CBC or organ toxicity. Three successive cohorts of SS mice were treated with Hp (or equivalent volume of PBS): 200 mg/kg IP in 3-month old mice, 400 mg/kg SC in 3-month old mice, and 400 mg/kg SC in 1-month old mice. At the 400 mg/kg dosing levels, there was a significant decrease in iron deposition in the kidneys of both 4-month and 6-month old mice (treatment started at 1-month and 3 months, respectively) (Table 1). There was also a trend towards decreased liver infarction in 6-month old mice (Table 2). Discussion: Functional binding of the administered human Hp to the human Hb of the Townes mice likely occurred, as evidenced by the decrease in iron deposition in the kidneys, suggesting that formation of the complex prevents filtration of Hb into the kidneys. Surprisingly, kidney function as measured by urine osmolality or albumin excretion was not improved, which may be explained by continued heme-laden red cell microparticle filtration (Camus SM, Blood 2015). Encouragingly, however, a trend towards decreased liver infarction in older mice was observed. The less-than-expected effect of Hp on mouse disease severity may also be explained by: 1) continuous physiologic Hp concentrations not being maintained with the dosing frequency while continued hemolysis releases Hb every minute of the day, and 2) CD163-mediated uptake in mice seems to only account for a part of the Hb clearance as opposed to in humans (Etzerodt A, Antioxid Redox Signal 2013). Despite the limitations of the SCD mouse model, the current study suggests haptogobin infusions could be beneficial in SCD patients. Acknowledgment: The authors are grateful to Sandra Suzuka for performing the HPLC. Table 1. Table 1. Kidney iron deposition (scale 1-10) Treatment group 6-month old SS 4-month old SS 400 mg/kg Hp 4.0 ± 1.4 1 ± 1.1 PBS 9.3 ± 0.6 5 ± 2.9 p-value 0.002 0.02 Table 2. Liver infarction (scale 1-10) Treatment group 6-month old SS 4-month old SS 400 mg/kg Hp 2.6 ± 2.0 3.7 ± 2.8 PBS 6.3 ± 2.4 3.8 ± 2.3 p-value 0.07 0.91 Disclosures Belcher: Biogen Idec: Research Funding; Seattle Genetics: Research Funding; CSL Behring: Research Funding. Vercellotti:CSL Behring: Research Funding; Seattle Genetics: Research Funding; Cydan: Research Funding; Biogen Idec: Research Funding.

Introduction: Gene therapy for sickle cell disease (SCD) is currently in active trials. Finding a safe and effective method for collection of hematopoietic stem cells (HSC) in SCD remains a challenge. Granulocyte colony stimulating factor (G-CSF), used most commonly for collecting HSC, can cause life-threatening vaso-occlusion in SCD. Bone marrow harvest requires general anesthesia and multiple punctures. Plerixafor is an inhibitor of the CXCR4 chemokine receptor on HSC and interferes with binding to SDF-1 on bone marrow stroma. As pre-clinical data in support of a clinical trial in SCD patients studying plerixafor mobilization (NCT02193191), we administered plerixafor to SCD mice to assess the risk of cell activation and vaso-occlusion. Methods: 3-6 month old SS Berkeley (n=12) or SS Townes mice (n= 18) were used. Littermate mice were randomized to subcutaneous treatment with plerixafor (Genzyme-Sanofi) 10 mg/kg once, G-CSF (Amgen) 250 ug/kg daily for 5 days, or equivalent volume (5 uL/g) normal saline daily for 5 days. Peripheral blood was harvested at 1-2 hr (plerixafor) or 4-5 hr (G-CSF and normal saline) after the last dose for the following studies: CBC (Advia), enumeration of HSC mobilization (Lin-Sca-1+ c-kit+ Flt3- (LSKF) cells by flow cytometry), neutrophil activation (L-selectin shedding by flow cytometry), and endothelial activation (soluble P-selectin by ELISA). Berkeley mice underwent MRI imaging before and after completion of treatment. Results: CBC showed the mean WBC and platelet counts of both plerixafor and G-CSF groups to be significantly different from saline, but the WBC differential was only significantly different (in % neutrophils and lymphocytes only) from saline in the G-CSF group (Table 1). The percentages of HSC subsets were significantly higher in both plerixafor and G-CSF groups compared to saline, with no significant differences between plerixafor and G-CSF (Table 2). L-selectin was low and soluble P-selectin high only in the G-CSF group, with both markers significantly different from both plerixafor and saline (Table 3). MRI imaging showed no significant differences in cerebral blood flow (measures oxygen delivery), mean diffusivity (measures vasogenic swelling), or fractional anisotropy (measures axonal integrity) pre- compared to post-treatment in any group. Tables show mean ± SD and significant p-values compared to saline. Discussion: Plerixafor and G-CSF were effective as evidenced by expected changes in WBC and platelet counts with treatment compared to saline. Both plerixafor and G-CSF significantly mobilized HSC subsets. There was a trend towards higher mobilization with G-CSF of the more primitive LSKF subset, but clinical data indicate that addition of plerixafor to G-CSF mobilizes a higher number of more primitive CD34+CD38- than G-CSF alone (Fruehauf S, Cytotherapy 2009). In support of potential safety of plerixafor in SCD patients, there was no evidence of neutrophil or endothelial activation with plerixafor, in contrast to G-CSF. Despite the evidence of neutrophil and endothelial activation with G-CSF, there was no evidence of perfusion-related organ damage as measured by MRI parameters. These findings suggest that plerixafor canbe safely and effectively used for HSC mobilization from SCD patients for use in gene therapy. Acknowledgments: We are grateful to Farid Boulad, MD and Tsiporah Shore, MD for providing plerixafor. Table 1. Significant CBC and WBC differential parameters Treatment group Saline (n=6) Plerixafor (n=8) G-CSF (n=8) WBC (103/uL) 21.7 ± 5.7 39.1 ± 15.5 (p=0.02) 56.3 ± 25.8 (p=0.006) Platelet (103/uL) 1,118 ± 394 554 ± 255 (p=0.007) 582 ± 280 (p=0.01) % neutrophils 20 ± 7 20 ± 6 51 ± 20 (p=0.003) % lymphocytes 76 ± 8 74 ± 7 40 ± 16 (p=0.0003) Table 2. Percent of hematopoietic stem cell subsets in lineage-negative population Treatment group Saline (n=6) Plerixafor (n=8) G-CSF (n=8) SKF 0.04 ± 0.08 0.72 ± 0.58 (p=0.01) 1.33 ± 1.15 (p=0.02) SK 0.12 ± 0.24 1.89 ± 1.42 (p=0.009) 1.42 ± 1.17 (p=0.003) Table 3. Neutrophil and endothelial activation markers Treatment group Saline (n=6) Plerixafor (n=8) G-CSF (n=8) L-selectin fluorescence intensity 8483 ± 5216 8106 ± 3987 2833 ± 1470 (p=0.04) Soluble P-selectin (ng/mL) 197 ± 61 180 ± 36 277 ± 45 (p=0.005) Disclosures No relevant conflicts of interest to declare.