Search

We have previously designed and investigated novel allosteric effectors of hemoglobin (AEHs), as potential targeted treatment of sickle cell disease (SCD). In general, AEHs transiently covalently bind to hemoglobin (Hb), increase its affinity for O2, increasing the fraction of oxygenated sickle Hb (HbS), thus reducing HbS polymerization and countering red blood cell (RBC) sickling. In the current study, we designed a novel class of AEH molecules, incorporating a secondary mechanism of action (MOA), which is independent of Hb O2-affinity by interacting with the F-helix of deoxygenated HbS to directly destabilize its polymerization. Here, we report current results from our in-vitro and in-vivo studies with a representative AEH compound (PP-14). First, we assessed the anti-sickling properties in-vitro by incubating 0.5, 1, and 2 mM of PP-14 with whole blood suspensions from a subject with homozygous SCD (SS, hematocrit: 20%) under hypoxic conditions, with subsequent RBC sickling assessment by microscopy. Next, we subjected the samples to anoxia (100% N2 gas) to demonstrate the O2-affinity-independent antisickling mechanism. Subsequently, we tested residual samples for the degree of Hb modification (i.e., HbS-AEH adduct formation) and O2-affinity (p50) shifts. In a second experiment to further assess the secondary MOA, we subjected SS blood samples treated with various concentrations of PP-14 to hypoxia in the Hemox analyzer, which permitted us to obtain aliquot samples at defined pO2 values to establish pO2-dependent sickling. Additionally, we conducted in silico and in-vitro ADME studies to evaluate possible metabolic inhibition of a panel of CYP enzymes. Finally, we conducted a preliminary in-vivo PK/PD study in wild-type mice administered single doses of PP-14 via the oral (P.O.: 100-200 mg/kg) and intraperitoneal (I.P.: 75 mg/kg) routes. Serial blood samples were collected for up to 52 h after P.O., and up to 30 h after I.P. administration, and samples were assayed to quantify PP-14 concentrations. Residual blood samples were assayed for in-vivo Hb-AEH adduct formation, and the corresponding change in O2-affinity (Δp50, %). Our in-vitro studies demonstrated concentration-dependent inhibition of cell sickling of 25.5±11%, 44.4±3.8% and 90.8±1%, at 0.5, 1 and 2 mM of PP-14, respectively. HbS was modified correspondingly (38.9±9%, 55.7±4.9 %, and 92.4±9.8%), and was correlated linearly with the left-shift in OEC (Δp50 values of 11.3±5.1%, 29.0±13.2%, and 67.5±8.2%). Importantly, the antisickling effect was sustained under anoxic conditions (100% N2), strongly supporting the notion of a secondary, O2-affinity-independent MOA. Furthermore, we observed a dose-dependent delay in sickling, with initiation of sickling recorded at a pO2 level of 40 mmHg in absence of PP-14; and at 30 and 20 mmHg at 0.5, and 1 mM PP-14 concentrations, respectively. Complete inhibition of sickling was observed at 2 mM PP-14 through the lowest recorded pO2 level of 1.5 mmHg, a unique, effect not previously observed in any analogous AEH. In-vitro partitioning studies demonstrated that >90% of PP-14 partitioned into the RBC compartment when whole blood was incubated with 100-300 µM concentrations. Metabolic studies using pooled human liver microsomes (HLM) and isozyme-specific probe substrates suggested that up to 100 µM PP-14 did not inhibit CYP1A2, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP3A4 or CYP2B6. Wild-type mice after I.P. administration of PP-14 (75 mg/kg, n=5) showed peak concentrations in blood at 7 hrs (416.3±81.2 µM), with corresponding PD effects (Δp50 of 41.6±13.5%; modified Hb levels of 43.6±8.0%). Orally-treated mice had peak drug concentrations after 10-24 hrs, (~150 µM at 200 mg/kg, n= 2), with corresponding PD effects (Δp50 of 36.5±7.0%; modified Hb levels of 28.8±4.9%), which declined by 52 hrs. Overall, our data confirm that PP-14 is novel antisickling AEH with a secondary, O2-independent MOA in addition to the primary O2-dependent effect, as demonstrated by the inhibition of sickling under anoxic conditions. Additionally, PP-14 showed: excellent partitioning into the RBC compartment; acceptable in-silico ADME properties and in-vivo oral bioavailability; PD effects; and low liability for metabolic drug-drug interactions. Further studies to investigate formal detailed pharmacokinetic properties, and biological activity after single- or repeat-doses in a SCD mouse model are ongoing. Disclosures Safo: Sanofi: Consultancy, Research Funding; Virginia Commonwealth University: Patents & Royalties. Pagare:Virginia Commonwealth University: Patents & Royalties. Ghatge:Virginia Commonwealth University: Patents & Royalties. Rivella:Meira GTx, Ionis Pharmaceutical: Membership on an entity's Board of Directors or advisory committees; Disc medicine, Protagonist, LIPC, Meira GTx: Consultancy. Hines:Functional Fluidics: Equity Ownership. Liu:Functional Fluidics: Employment. Zhang:Virginia Commonwealth University: Patents & Royalties. Venitz:Virginia Commonwealth University: Patents & Royalties. Abdulmalik:The Children's Hospital of Philadelphia: Patents & Royalties: Provisional Patent.

Dilution of HbS with non-sickling hemoglobin or hemoglobin with increased oxygen affinity is clinically beneficial in sickle cell disease. Aldehydes, including 5-HMF, tucaresol or GBT440, modify the N-terminus of HbS by reversible covalent imine formation generating modified forms of HbS that resist polymerization under low oxygen concentrations. In contrast to reversible imine formation by aldehydes, we hypothesize that stable modification of HbS will result from N-terminal retention of the initiator methionine (iMet) and subsequent N-terminal acetylation of the iMet (acetyl-iMet). MetAP2 is the methionine aminopeptidase able to cleave iMet from Val1 on α-globin and βS-globin as the unfolded N-terminal peptides emerge from the ribosome. Enzyme kinetic studies with pure MetAP2 and N-terminal octapeptides showed that βS-globin peptide is a 5-fold better substrate than α-globin peptide. Lentiviral shRNA knock-down of MetAP2 in differentiating erythroid HUDEP cells in vitro confirmed that α-globin is more extensively modified than βS-globin, consistent with the enzyme kinetic data. Selective MetAP2 inhibitors used to treat cultured human erythroid cells (HUDEP and PBMC derived CD34+) and Townes SCD mice in vivo confirmed that both α-globin and βS-globin domains of HbS are extensively modified by N-terminal iMet and acetyl-iMet. N-terminal retention of iMet and subsequent acetylation creates a mixture of modified HbS tetramers with combined modifications on both globins. Cation exchange chromatography separated nine different modified HbS variants from unmodified HbS as identified by LCMS. Purified samples of HbS modified by N-terminal iMet and acetyl-iMet had increased oxygen affinity as measured by decreased P50. Modified HbS containing the acetyl-iMet-βS-globin were found to have delayed polymerization under complete hypoxia (sodium metabisulfite triggered hypoxia in 1.8 M phosphate). Two modified HbS variants were further purified for X-ray crystallography studies (βS-globin / iMet-α-globin and acetyl-iMet-βS-globin / iMet-α-globin). Oxyhemoglobin structures of both modified HbS variants were in the R2-state previously described in structures of aldehyde modified HbS. This R2-state stabilizes the oxygenated R-state of HbS from conversion to the deoxygenated T-state that initiates HbS polymerization in sickle RBC. Treatment by selective irreversible covalent or reversible MetAP2 inhibitors resulted in high levels of HbS modification (>75%) in cultured erythroid cells (HUDEP and CD34+ cells). Dose dependent modification of HbS was observed in Townes sickle cell mouse blood RBC in vivo with total modification of HbS approaching 50%. In whole blood ex vivo studies, modification of HbS decreased RBC sickling under hypoxia (4% O2) and significantly increased the affinity of RBC for oxygen (decreased P50). Blood samples from MetAP2 inhibitor treated mice were analyzed for single-cell O2 saturation of the RBC and for the fractional flow velocity drop in whole blood rheology under decreasing partial oxygen pressures. In blood from vehicle treated sickle mice, a low-saturation peak of deoxy-HbS was observed in 7.8% O2, in contrast to blood from MetAP2 inhibitor-treated mice where the low-saturation peak was only observed in 6.4% O2. Similarly, in an assay of O2 dependent blood flow rheology, the half-maximum fractional velocity drop occurred at 5% O2 in control blood decreasing to 2% O2 in MetAP2 inhibitor treated blood. Our studies show that MetAP2 inhibition results in retention of iMet on βS-globin and α-globin and allows further acetylation of the retained iMet to create a mixture of N-terminal modified HbS tetramers. These modified HbS variants resist polymerization and RBC sickling under conditions of low O2 by delaying HbS polymerization and increasing O2 affinity. Our data suggests that MetAP2 may warrant further study as a potential therapeutic target for sickle cell disease. Disclosures Demers: Sanofi: Employment. Sturtevant:Sanofi: Employment. Guertin:Sanofi: Employment. Gupta:Sanofi: Employment. Desai:Sanofi: Employment. Vieira:Sanofi: Employment. Hicks:Sanofi: Employment. Ismail:Sanofi: Employment. Safo:Sanofi: Consultancy, Research Funding; Virginia Commonwealth University: Patents & Royalties. Wood:Sanofi: Consultancy, Research Funding. Higgins:Sanofi: Consultancy, Research Funding. Light:Sanofi: Employment.

Sickle cell disease (SCD) is a genetic hemoglobinopathy driven largely by a single codon mutation of the β-globin gene resulting in polymerization of hemoglobin S (HbS). Anti-sickling approaches that involve increasing the oxygen affinity of HbS to treat SCD are under development and offer the potential to directly prevent HbS polymerization and its downstream pathophysiology. Two such compounds, 5-hydroxymethylfurfural (5HMF) and voxelotor (GBT440) have entered clinical trials for SCD with promising results and exert their therapeutic effects by modifying the N-terminus of HbS α-globin chains to form a reversible Schiff base. Formation of this N-terminal adduct stabilizes the oxygen-bound R-state (in the R2 conformation) that increases the oxygen affinity of the altered HbS and delaying the polymerization of HbS. In addition, genetic and small molecule therapies designed to increase fetal hemoglobin (HbF) expression hold great potential for the treatment of SCD. Increasing the percentage of HbF in RBCs significantly slows sickling kinetics without affecting oxygen delivery. Combination approaches of high-O2-Hb modification with HbF inducing therapies clinically could result in increased efficacy in the treatment of SCD, but the impact of hemoglobin modifiers on fetal hemoglobin has not been reported. Our present studies investigated the effects of 5HMF and voxelotor in HbF-rich umbilical cord blood derived RBCs. HbF-rich (60-90%) RBCs were isolated from cord blood and incubated with commercially available 5HMF and voxelotor synthesized in-house. The effect of these compounds on hemoglobin oxygen affinity was determined by measuring the p50 of the oxygen saturation curve in whole cells. Sites of modification were determined directly by incubating compounds with the purified RBC lysate, stabilizing the N-terminal adduct by reduction to the amine, and analysis of the resulting modification by LC-MS. Similar to the reported p50 shifts with normal adult hemoglobin (HbA) and HbS, 5HMF and voxelotor increased the oxygen-binding affinity of HbF with an EC50 of 7.9 mM and 560 mM respectively. 1 mM voxelotor lowered cord RBC p50 to 4 mmHg in vitro. LC-MS analysis showed that 5HMF exclusively modified the N-terminus of the α-globin chain, with no modification of b-globin and g-globin chains. Unexpectedly, the α-globin, β-globin and γ-globin chains were all modified by voxelotor following incubation with cord blood. Voxelotor was also shown to modify both α-globin and the β-globin or βS-globin chains on purified HbA or HbS, respectively. These data contrast with published crystallography data demonstrating that voxelotor selectively modifies a single α-globin chain in CO-ligated HbS (Oksenberg et al 2016). Although anti-sickling aromatic aldehydes have similar effects on the oxygen binding affinity of HbA, HbS and HbF, they can vary in their selectivity for modification of the α-globin and beta-like chains of HbF, HbA, and HbS (Abraham et al. 1995). To further investigate our data with voxelotor and increase our understanding of this class of molecules, other hemoglobin modifying aldehyde molecules such as 5-formylsalicyclic acid (5FSA), tucaresol and velaresol (BW12C) will be examined. Disclosures Vieira: Bioverativ a Sanofi Company: Employment. Hong:Bioverativ a Sanofi Company: Employment, Equity Ownership. Desai:Bioverativ a Sanofi Company: Employment, Equity Ownership. Safo:Bioverativ a Sanofi Company: Consultancy; Virginia Commonwealth University: Employment. Light:Bioverativ a Sanofi Company: Employment, Equity Ownership.

Introduction: In patients with sickle cell disease,during a vasoocclusive crisis, deoxygenated sickle hemoglobin (HbS) polymerizes forming fibers of HbS in red blood cells (SS RBCs). HbS fibers distort SS RBCs, causing microvascular occlusion, increased thrombosis and inflammation, and severe pain for patients. Increasing the affinity of Hb for oxygen reduces sickling of SS RBCs, because oxygenated HbS does not form polymers. It has been reported that modification of the thiol group of Hb β-Cys93 increases Hb's affinity for oxygen by destabilizing the T-state and stabilizing the R-state. In addition, modification of HbS β-Cys93 might prevent interaction between HbS molecules and thereby reduce sickling. Herein we report a novel triazolyldisulfide compound (4,4'-Di(1,2,3-triazolyl)disulfide, designated "TD-3"), which increases the affinity of Hb for oxygen and reduces the sickling of hypoxic human SS RBCs in vitro. Intravenous administration of TD-3 to mice increases the affinity of murine Hb for oxygen. The effects of TD-3 on the affinity of Hb for oxygen may be a result of interaction with HbA-Cys93. Methods: TD-3 hydrate was dissolved in a mixture of Dulbecco's phosphate buffered saline and 30% polyethylene glycol 400. The oxygen dissociation curve (ODC) of Hb was measured at 37°C (pH 7.3) using a HEMOX analyzer. The partial oxygen pressure at which 50% of Hb is oxygenated was determined as P50 from the ODC and the P50 was used to assess the affinity of Hb for oxygen. SS whole blood was incubated with TD-3 (Hb tetramer/TD-3 = 1/1 mol/mol) for 10 min at 37°C and the P50 of TD-3 treated RBCs was determined. To evaluate the ability of TD-3 to reduce sickling in vitro, SS RBCs (Hct ≈ 20%) were incubated with TD-3 in a gas mixture of 4% oxygen and 96% nitrogen and the percentage of sickled RBCs was determined after incubation. To determine the effect of TD-3 on time-dependent change of P50 in vitro, normal human whole blood (Hct ≈ 45-50%) was incubated with TD-3 (Hb tetramer/TD-3 = 1/1 mol/mol) for 10 min and 8 h at 37°C. At both times, the P50 of RBCs was determined. To determine the effect of TD-3 on the time-dependent change of P50 in vivo, TD-3 (100 mg/kg) was administered intravenously to C57BL/6 mice and the P50 of murine Hb was determined (as hemolysate) before, and 1 and 24 h after TD-3 treatment. As a first step to investigate the mechanism of action of TD-3, normal adult hemoglobin (HbA) was treated with N-ethylmaleimide (NEM) to block HbA β-Cys93 and the P50 of NEM-treated HbA was determined. Either HbA or NEM-treated HbA was incubated with TD-3 at 37°C for 10 min (Hb tetramer/TD-3 = 1/6 mol/mol). After the incubation, the P50 of HbA and NEM-treated HbA was determined. Results: Incubation of TD-3 with SS RBCs reduced the P50 of SS RBCs from 29 mmHg to 24 mmHg. Incubation of SS RBCs with TD-3 (2 mM) in 4% oxygen decreased the percentage of sickled RBCs from 94% (without TD-3) to 22%. The P50 of TD-3-treated human normal RBCs was decreased from 25 mmHg (prior to adding TD-3) to 21 mmHg at 10 min. At 8 h, the P50 of TD-3-treated human normal RBCs was the same as that of RBCs treated with vehicle alone. The P50 of TD-3-treated murine Hb was reduced from 21 mmHg (before treatment with TD-3) to 18 mmHg at 1 h. At 24 h, the P50 of TD-3-treated murine Hb was the same as that of mice treated with vehicle alone. At 24 h, all of the mice that were treated with either TD-3 or vehicle alone were alive and appeared normal. Treatment of HbA with NEM reduced the P50 from 17 mmHg to 8 mmHg. Incubation of TD-3 with either HbA or NEM-treated HbA reduced the P50 of HbA from 17 mmHg to 4 mmHg, but did not alter the P50 of NEM treated HbA (8 mmHg). Conclusions: 4,4'-Di(1,2,3-triazolyl)disulfide (TD-3) increased the affinity of human normal and SS RBCs for oxygen and reduced the sickling of hypoxic human SS RBCs in vitro. Administration of TD-3 to healthy mice increased the affinity of murine Hb for oxygen and the compound was well tolerated by the mice. The effect of TD-3 on the affinity of HbA for oxygen was impaired by blocking the thiol group of HbA β-Cys93 with NEM, suggesting that the mechanism by which TD-3 increases the affinity of Hb for oxygen and reduces sickling may be through interaction with Hb β-Cys93. Our data demonstrate that TD-3 has the potential to prevent and treat sickle cell disease. Disclosures Nakagawa: Massachusetts General Hospital: Patents & Royalties: MGH filed a patent application on the use of heteroaryl disulfide compounds including TD-3 to increase the oxygen-binding affinity of hemoglobin and treat sickle cell disease and other uses of these compounds. The patent hasn't issued or licensed yet.. Zapol:Massachusetts General Hospital: Patents & Royalties: MGH filed a patent application on the use of heteroaryl disulfide compounds including TD-3 to increase the oxygen-binding affinity of hemoglobin and treat sickle cell disease and other uses of these compounds. The patent hasn't issued or licensed yet..

Allosteric effectors of hemoglobin (AEH) remain promising as a viable therapeutic approach for the management of sickle cell disease (SCD). AEH bind to hemoglobin (Hb) in a transiently covalently manner, increasing Hb affinity for oxygen (O2), with concomitant inhibition of polymerization of deoxygenated sickle Hb (HbS) and erythrocyte sickling. As part of our ongoing efforts to surmount well-known AEH druggability challenges, we have designed and synthesized VZHE-039, a novel substituted benzaldehyde with potent in-vitro anti-polymerization activity, as well as more favorable in-vivopharmacokinetic and pharmacodynamic (PK/PD) properties. Here, we report the results of our early findings. The anti-sickling properties of VZHE-039 were tested in-vitro by incubating 0.5, 1, and 2 mM concentrations with blood suspensions from a subject with homozygous SCD (hematocrit: 20%) under hypoxic conditions (4% O2/96% N2) at 37°C for 2 h. At conclusion, aliquots were drawn into a fixative (2% glutaraldehyde solution), and sickling was assessed by microscopy. Aliquot samples were also subjected to cation-exchange HPLC analyses to measure the degree of Hb modification (Hb adduct formation), as well as standard O2 equilibrium curves (OEC) to assess p50 shifts. Subsequently, we conducted in-vivo PK/PD studies in wild-type mice that received single doses of VZHE-039 via the intravenous (I.V.: 25 and 50 mg/kg), intraperitoneal (I.P.: 100-150 mg/kg) and oral (P.O.: 100-200 mg/kg) routes. Serial blood samples were collected up to 24 h after I.V. and P.O., and up to 6 h after I.P. administration, respectively; aliquots were hemolyzed, de-proteinized and subjected to reverse-phase HPLC-UV assay to quantify VZH-039 blood concentrations. Residual blood samples were also hemolyzed, and clarified lysates were assayed for in-vivoHb adduct formation, and the corresponding change in Hb oxygen affinity (Δp50, %). The results of our in-vitrostudies demonstrated a concentration-dependent inhibition of SS cell sickling: 39±1.4%, 68.6±4.9%, and 89.9±5.2%, at 0.5, 1 and 2 mM of VZHE-039, respectively. HbS was modified correspondingly (46±7.4%, 83.6±6.4%, and 96.4±6.2%), and this adduct formation was correlated linearly with the left shift in OEC (Δp50 values of 34.3±8.7%, 63.6±2.4%, and 76.6±2.5%). Noncompartmental PK analysis showed that, after I.V. administration, systemic PK was dose-independent with a CLtot of 0.5 mL/min/kg, resulting in a terminal half-life of 9 h. After I.P. administration, AUCtrap increased supra-proportionally with dose; the limited sampling schedule suggested an I.P. bioavailability of >20%. After P.O. administration, AUCtrap increased supra-proportionally with dose as well due to an increase in oral bioavailability to ~ 10%, with peak blood concentrations up to ~ 0.25 mM; the terminal half-lives were prolonged (relative to I.V.) to 14 h. The PD effect-time course of change in %Hb adduct paralleled (and was reasonably linearly related to) blood concentrations for all routes. Baseline-corrected peak PD effects after I.V. administration were 15.7% and 31.9%, after 25 and 50 mg/kg respectively; after P.O. administration of 200 mg/kg, the peak change was 17.6%. Overall, VZHE-039 exhibited a low (metabolic) clearance, presumably reflecting high-affinity and sustained Hb (and possibly albumin) binding, leading to the observed extended terminal half-life and duration of action. While its oral bioavailability is low (likely due to limited GI solubility) and there is evidence of saturable first-pass effects and possibly enterohepatic recycling, oral in-vivo exposures were associated with PD effects that were consistent with the results from the in-vitrostudies. In conclusion, the results of our current studies establish VZHE-039 as a novel, potent anti-sickling agent, confirm its proposed mechanism of action in-vitro and in-vivo, and suggest that therapeutically-relevant blood concentrations may be achieved after daily oral doses. Therefore, VZHE-039 is our new lead drug candidate that may require further modifications and studies to (a) improve its GI solubility and oral bioavailability; (b) elucidate atomic level structural interactions and kinetics of its Hb binding; and (c) formally investigate its biological activity after repeat-doses in a SCD mouse model. Disclosures No relevant conflicts of interest to declare.

Important therapeutic approaches to sickle-cell disease (SCD) are based upon the observation that the abnormal properties of Hb S (Hb α2βS2) can be mitigated by exchanging the pathological βS-globin subunit for a related β-like subunit. We previously demonstrated that exchange of the non-pathological α-globin subunit for a ζ-globin subunit (a developmentally silenced globin that can be derepressed both by natural and experimental conditions) inhibits deoxyHb S polymer assembly in vitro and reverses hematological, biochemical, and physiological characteristics of SCD in mouse models in vivo. While its therapeutic potential is clear, the underlying structural basis for the profound antipolymer activity of ζ-substituted Hb S (Hb ζ2βS2) is less certain. X-ray crystallographic studies conducted at 1.95Å resolution revealed that liganded (CO-) Hb ζ2βS2 is trapped in a tense (T-state) quaternary structure, rather than in a relaxed (R-state) structure that is characteristic of nearly all liganded hemoglobins. Specifically, CO-Hb ζ2βS2 exhibited several intact T-state intersubunit salt-bridge/hydrogen-bond interactions, a preserved T-state ζ1-β2 (ζ2-β1) interface, and a characteristically enlarged T-state central water cavity and β-cleft. This structure wrongly predicts that liganded Hb ζ2βS2 will be included, rather than excluded, from the deoxyHb S polymer; and suggests that changes in the positions or the biochemical identities of individual amino acids, rather than the overall quaternary structure of liganded Hb ζ2βS2, are the chief determinants of its antipolymer activity. To define key differences in the structures of T-state deoxyHb α2βS2 and CO-Hb ζ2βS2, we superposed their corresponding globin subunits and calculated the specific displacement of individual amino-acid residues as root mean square deviation (rmsd) values. Among βS-chain residues, α→ζ exchange effects a significant 1.9Å shift in the position of the pathological βSVal6, and correspondingly large displacements of βThr4 (2.2Å) and βAsn19 (1.4Å); each repositioning predicts weakening of an intermolecular interaction that would otherwise stabilize the deoxyHb S polymer. Similar superposition analyses of the α and ζ chains reveal a significant displacement of αPro114 (1.3Å), a well-described determinant of deoxyHb S polymerization that is conserved between the two α-like subunits. Three additional α-chain residues that stabilize the deoxyHb S polymer undergo nonconservative replacement in the ζ-globin chain, but are not materially repositioned: αHis20→ζGln (basic→neutral polar), αAsn68→ζAsp (neutral polar→acidic), and αAsn78→ζGly (neutral polar→neutral). While all three replacements are predicted to weaken or ablate intermolecular contacts, the αHis20→ζGln substitution is particularly noteworthy as it reproduces the specific mutation that defines the naturally occurring anti-sickling variant αLe Lamentin. Finally, we considered the possibility that ζ-substituted Hb S is fully excluded from the deoxyHb S polymer--and therefore reduces the rate of its assembly--by comparing the crystal packing of the two hemoglobins. While deoxyHb α2βS2 packs in a familiar two-strand structure, CO-Hb ζ2βS2 assembles into a unique trimeric arrangement comprising three lateral heterotetramers, each of which interacts with an axial heterotetramer that is constituent to a separate trimer assembly. This remarkable structure is sustained by intermolecular interactions that are distinct from those observed for deoxyHb S. Moreover, the calculated buried solvent-accessible surface area for CO-Hb ζ2βS2 (4806Å2) is nearly two-fold higher than for deoxyHb α2βS2 (2510Å2), suggesting that Hb ζ2βS2 exists in solution as a stable trimer of heterotetramers, and validating the hypothesis that Hbs α2βS2 and ζ2βS2 do not co-assemble in solution. In sum, our crystal analyses indicate that the antipolymer activities of liganded Hb ζ2βS2 arise through movements in the positions of βS-chain residues, and through changes in the identities of α-chain residues. Our studies also demonstrate a novel packing structure for T-state liganded Hb ζ2βS2 that is consistent with its exclusion from the deoxyHb S polymer. These data account for the significant antipolymer activity of ζ-substituted Hb S, and recommend the utility of therapeutic approaches to SCD that are based upon α-globin subunit exchange. Disclosures Safo: Baxter and AesRx companies have licensed our patented antisickling compounds. Consulted with AesRx LLC during phase I clinical studies of the antisickling compound, 5HMF for the treatment of sickle cell disease: #7160910; #7119208 Patents & Royalties, Consultancy, Research Funding.

Sickle cell disease (SCD) continues to cause significant morbidity, mortality and healthcare disparities. Despite considerable progress in understanding the underlying pathophysiology and investigating various therapeutic strategies, novel pharmacologic approaches to ameliorate SCD continue to hold immense potential and promise, especially for patients in developing countries. Our group and others have recently renewed and refocused attention to candidate drugs that directly bind to hemoglobin (Hb) and increase oxygen (O2) affinity, preventing the fundamental pathophysiology of the disease, i.e., sickle Hb (Hb S) polymerization and red blood cell (RBC) sickling. While several candidate drugs have shown biological activity in-vitro, ex-vivo and in animal studies, their ultimate success in clinical studies was hampered by toxicity concerns and/or low oral bioavailability. Recent promising reports from a phase I/II study on 5-HMF renews optimism for this therapeutic approach. We reasoned that modifications of vanillin--a previously reported antisickling agent and food constituent without known toxicities--to enhance its efficacy, would represent a feasible approach in rationally developing clinically useful candidate drugs. Consequently, we designed and synthesized two classes of compounds: INN and TD series. The former are pyridyl derivatives of vanillin, rationalized to stereospecifically inhibit deoxy-Hb S polymer formation while increasing the fraction of the soluble oxy-Hb S in regions of low O2 tension. The TD compounds represent further modification of corresponding INN compounds (with a methoxyl group on the pyridine ring), rationalized to exhibit similar dual antisickling effects, but with enhanced direct polymer destabilization properties. We subjected a prototypical compound from each class (INN-270 and TD-7) to our battery of exploratory in-vitro assays, specifically: 1) rates of Hb S binding/modification, 2) corresponding change in O2 affinity, 3) direct inhibition of Hb S polymerization, and 4) inhibition of RBC sickling under hypoxia. We incubated 0.5, 1, or 2 mM of either INN-270 or TD-7 with RBCs from patients with homozygous SCD, under hypoxia (4% O2/96% N2 gas mixture) in a shaker-incubator at 37 ˚C for 3 h. Assays were conducted in at least three replicates utilizing different samples on different days. At the conclusion of each assay, aliquot samples (~ 10 μl each) were drawn into a fixing solution under hypoxia to preserve RBC morphology for analyses. Residual RBC suspensions were washed, hemolyzed, and subjected to: cation-exchange HPLC (to determine Hb modification); P50 analyses to establish change in O2 affinity; and temperature-dependent delay time studies to establish a delay in Hb S polymerization. Our results show that both compounds permeated RBC membranes without causing hemolysis, bound to and modified intracellular Hb at high levels in a dose dependent manner, increased O2 affinity significantly, and inhibited sickling of RBCs under hypoxia. TD-7 modified Hb S in a dose-dependent manner (to 92.3 ± 5.2 %, n=4 at 2 mM), shifted O2 equilibrium to the left (Δp50 = 45.6 ± 8.2 %, n=3 at 2 mM), and inhibited RBC sickling (by 95 -100 %, n=4). Preliminary delay time analyses also showed that at 2 mM, TD-7 increased the Hb S polymerization times from 18.1 ± 1.0 min to 24.5 ± 0.5 min. INN-270 showed a similar profile, however with a lower efficacy (at 2 mM) for Hb S modification (to ~ 75 %), Δp50 of 40.3 %, sickling inhibition by ~ 70 %, and increased delay times from 15.6 ± 0.5 min to 19.7 ± 1.0 min. We have elucidated the dual antisickling mechanism of action of INN-270 and TD-7 by X-ray crystallography. Two molecules of each compound bind to Hb via Schiff-base, and a series of hydrogen-bond/hydrophobic interactions that favor a high-O2-affinity Hb state. Importantly, the methoxyl group on the pyridine ring of TD-7 forms hydrogen-bond interactions with the surface-located αF-helix, resulting in a conformational change, possibly explaining the improved potency. Based on our results, both TD7 and INN 270 exhibited greater than a 40- and 3-fold superiority in efficacy compared to vanillin and 5-HMF, respectively. We conclude that our findings justify a prospective, structure-based approach to designing novel antisickling agents with enhanced potency. In-vitro/ex-vivo murine and human PK/PD studies are currently ongoing to help guide planned in-vivo PK/PD studies in mice. Disclosures Venitz: Consulted with AesRx LLC during phase I clinical studies of the antisickling compound, 5HMF for the treatment of sickle cell disease: Consultancy. Safo:Baxter and AesRx companies have licensed our patented antisickling compounds. Consulted with AesRx LLC during phase I clinical studies of the antisickling compound, 5HMF for the treatment of sickle cell disease: #7160910; #7119208 Patents & Royalties, Consultancy, Research Funding.

Abstract 822 Sickle cell disease (SCD) is a genetic hemolytic disease with high morbidity and mortality affecting millions of individuals worldwide. Although SCD was identified a century ago, we still lack effective mechanism-based therapies to treat this disease. Using non-biased metabolomic screening, we found that sphingosine-1-phosphate (S1P) is significantly elevated in the blood of SCD mice. Further analysis revealed that the activity of sphingosine kinase 1 (Sphk1, the enzyme that produces S1P) is significantly elevated in erythrocytes of SCD mice. Chronic treatment of SCD mice with a SphK1 inhibitor significantly attenuated sickling, hemolysis, inflammation and multiple tissue damage by reducing erythrocyte and plasma S1P levels. Erythrocyte S1P levels were further elevated following hypoxia/reoxygenation-induced acute sickle crisis (ASC) in SCD mice and blocking its elevation by a Sphk1 specific inhibitor significantly reduced hallmark features associated with ASC and increased survival rates. To extend our pharmacological findings, we utilized a genetic approach to inhibit the synthesis of Sphk1 specifically in hematopoietic cells (HCs). For this purpose we infected bone marrow cells (BMCs) isolated from SCD mice with recombinant lentivirus encoding shRNA specific for Sphk1 or a scrambled shRNA sequence. Following viral transduction, we transplanted the genetically modified BMCs from SCD mice to lethally irradiated WT recipients to generate SCD chimeras. Similarly, we found that knockdown of SphK1 by shRNA in SCD chimeras led to a significant reduction of erythrocyte and plasma S1P and subsequent decreased sickling, hemolysis and inflammatory cells. Strikingly, we found that splenomegaly was substantially reduced in SCD chimera mice with the specific knockdown of SphK1 in hematopoietic cells. As with SCD mice, we found that erythrocyte Sphk1 activity and erythrocyte and plasma S1P levels were significantly elevated in humans with SCD compared to normal individuals. Inhibition of SphK1 in cultured primary human erythrocytes isolated from SCD patients inhibited hypoxia-induced elevation of erythrocyte S1P levels and reduced sickling. Thus, we have revealed for the first time that SphK1-mediated S1P elevation in SCD erythrocytes is a key contributor to sickling in SCD and that Sphk1 inhibition can attenuate both acute and chronic sickling events and disease progression. In an effort to determine the molecular mechanism underlying S1P-induced sickling, we unexpectedly found that S1P directly binds with Hb and results in a reduced Hb-O2 affinity. This finding led us to further discover that 2,3-diphosphoglycerate, another erythrocyte specific allosteric modulator, is required for S1P-mediated allosteric modulation and that these two endogenous heterotropic modulators work cooperatively to induce a substantial reduction in Hb-O2 affinity. Thus, our discovery adds a significant new chapter to erythrocyte physiology by revealing S1P is a novel allosteric modulator of Hb-O2 affinity and also provides a mechanism underlying S1P-mediated sickling by promoting the formation of deoxyHbS. Thus, the work reported here could be the foundation leading to future human trials and a possible therapy for SCD, a life-threatening hemolytic disorder for which the current treatment is extremely limited. Disclosures: No relevant conflicts of interest to declare.

Abstract LBA-3 Sickle cell disease (SCD) is a debilitating hemolytic disorder with high morbidity and mortality affecting millions of individuals worldwide. Although SCD was first identified a century ago, we still lack effective mechanism-based safe therapies to treat this disease. Thus, identification of specific molecules triggering sickling, the central pathogenic process of the disease, is extremely important to advance our understanding of the molecular basis for the pathogenesis of SCD and to develop novel therapeutics. Using non-biased metabolomic screening, we found that sphingosine-1-phosphate (S1P) is significantly elevated in the blood of SCD mice. Further analysis revealed that the activity of sphingosine kinase 1 (Sphk1, the enzyme that produces S1P) is significantly elevated in erythrocytes of SCD mice. Chronic treatment of SCD mice with a SphK1 inhibitor significantly attenuated sickling, hemolysis, inflammation and multiple tissue damage by reducing erythrocyte and plasma S1P levels. Erythrocyte S1P levels were further elevated following hypoxia/reoxygenation-induced acute sickle crisis (ASC) in SCD mice and blocking its elevation by a Sphk1 specific inhibitor significantly reduced hallmark features associated with ASC. As with SCD mice, we found that erythrocyte Sphk1 activity and erythrocyte and plasma S1P levels were significantly elevated in humans with SCD compared to normal individuals. Inhibition of SphK1 in cultured primary human erythrocytes isolated from SCD patients inhibited hypoxia-induced elevation of erythrocyte S1P levels and reduced sickling. Thus, we have revealed for the first time that SphK1-mediated S1P elevation in SCD erythrocytes is a key contributor to sickling in SCD and that Sphk1 inhibition can attenuate both acute and chronic sickling events and disease progression. S1P is an important signaling molecule regulating diverse biological processes. Although S1P is predominantly produced and stored in RBCs, nothing was known about the physiological role of S1P in normal RBCs or the pathophysiological role of S1P in SCD until we conducted a metabolomic screen. In an effort to determine the molecular mechanism underlying S1P-induced sickling, we unexpectedly found that S1P directly binds with Hb and results in a reduced Hb-O2 affinity. This finding led us to further discover that 2,3-diphosphoglycerate, another erythrocyte specific allosteric modulator, is required for S1P-mediated allosteric modulation and that these two endogenous heterotropic modulators work cooperatively to induce a substantial reduction in Hb-O2 affinity. Supporting the biochemical and functional findings, molecular modeling predicts that S1P binds near the water filled central cavity of HbA at a site that is different from the Hb-2,3-DPG binding site. Thus, our discovery adds a significant new chapter to erythrocyte physiology by revealing S1P is a novel allosteric modulator of Hb-O2 affinity and also providing a mechanism underlying S1P-mediated sickling by promoting the formation of deoxyHbS. Thus, the work reported here could be the foundation leading to future human trials and a possible therapy for SCD, a life-threatening hemolytic disorder for which the current treatment is extremely limited. The significance of our findings extends well beyond SCD. Our findings reveal a previously unrecognized important role for S1P in erythrocyte physiology and indicate a new concept for the regulation of O2 release from Hb under normal and sickle cell disease conditions. For SCD, elevated S1P is detrimental because reduced Hb-O2 affinity leads to more deoxygenation of HbS, increased sickling and subsequent multiple life-threatening complications. However, for normal erythrocytes, elevated S1P is likely beneficial by decreasing Hb-O2 affinity allowing for more O2 release to hypoxic tissues. Thus, for humans with normal Hb, if elevated S1P can induce O2 release to hypoxic tissues it may be a novel therapeutic target for a range of disorders, from chronic heart failure to diabetic retinopathy, traumatic blood loss, pulmonary disease and even cancer. In this way our findings reveal important novel opportunities to treat and prevent not only SCD but also multiple cardiovascular and pulmonary diseases associated with hypoxia. Thus, the impact of our novel finding is significant and enormous. Disclosures: No relevant conflicts of interest to declare.

Chemical modification of sickle hemoglobin (Hb S) to form stable high affinity Schiff-base adducts has been an attractive approach towards finding a potential therapeutic option for sickle cell disease (SCD). An ideal candidate drug should rapidly enter the bloodstream, permeate red blood cell membrane, bind specifically with intracellular Hb S and inhibit cell sickling with minimal adverse effect. In an effort to find drugs that satisfy these criteria, we recently designed, synthesized and studied three novel benzaldehydes (INN-296, INN-298 and INN-312) with enhanced potency. The compounds are pyridyl derivatives of benzaldehyde, and hence, combine structural features of two previously determined antisickling agents: vanillin and pyridoxal. All three compounds shifted the allosteric equilibrium of Hb S toward the oxy- or R-state by destabilizing the deoxy- or T-state. The results of in vitro studies of the antisickling effects of a representative compound (INN-312) are reported. Upon incubation of suspensions of sickle erythrocytes (SS cells) with 0.5, 1 or 2 mM of INN-312 under hypoxia (4% O2 :96% N2) at 37°C, sickling of SS cells was inhibited in a dose-dependent manner (15 ± 2, 44 ± 10 and 81 ± 8% inhibition, respectively). Cation-exchange HPLC analysis of lysates from the pre-incubated SS cells revealed a new peak in addition to the original Hb S peak, indicative of formation of Schiff-base adducts of Hb. Oxygen equilibrium curves (OECs) of SS-cell suspensions and lysates were shifted toward the left in a dose-dependent manner. X-ray crystal structures of these derivatives revealed their symmetric binding to the two N-terminal αVal1 of Hb S, and seem to indicate that their superior antisickling activity may arise from effector-induced interference with Hb S polymerization, as well as shifting the OEC to the high affinity state. In vitro studies on INN-296 and INN-298 showed similar results. Studies in vivo were performed using transgenic sickle mice (3 mice per group). The mice were treated intraperitoneally with single doses of 50, 100 or 150 mg/kg of INN-312. To study pharmacokinetic profiles of INN-312 in treated mice, blood samples (~20 μl each) were collected under anesthesia via retro-orbital venipuncture into EDTA tubes at 30 min, 1 h and every hour afterwards for 5 hours. Plasma from each sample was de-proteinized and analyzed by reversed-phase HPLC for quantification of INN-312 present in the blood. A non-compartmental pharmacokinetic model with first-order elimination rate was used to determine the plasma concentration-time data using PK Solutions 2.0 software (SUMMIT Research Services, Montrose, CO, USA). The area under the plasma concentration curve (AUC) increased in a dose-dependent manner (314 ± 22 μg/ml/min, 648 ± 33 μg/ml/min and 1044 ± 63 μg/ml/min in mice treated with 50, 100 and 150 mg/kg, respectively). The terminal half-life (T1/2= 0.75 ± 0.15 h), peak concentration time (Tmax= 0.5 h), and mean resident time (MRT= 1.2 ± 0.2 h) values were consistent for all three dosage groups. The observed maximum plasma concentration (Cmax)was also increased in a dose-dependent manner. These novel pyridyl derivatives of benzaldehyde shifted the position of Hb OEC toward the left most strongly among various compounds reported to date. Further detailed studies are necessary to validate this approach to developing better antisickling agents.

Since its inception in 1997, the NHLBI Sickle Cell Disease (SCD) Reference Laboratory (Ref Lab) has received over 150 candidate drugs from over 30 researchers, universities and companies for evaluation. All candidate drugs are evaluated for both their beneficial and adverse effects in vitro using more than 10 sophisticated methods. Specifically, each candidate drug is studied to see whether it inhibits cell sickling, prolongs the delay time prior to deoxy-hemoglobin (Hb) S polymerization, increases the solubility of deoxy-Hb S, hydrates red blood cells (RBCs), prevents dehydration of RBCs, shifts the oxygen equilibrium curve (OEC) toward the left, reduces the adhesion of RBCs to endothelial cells, promotes the synthesis of Hb F, causes hemolysis, promotes formation of met-Hb and increases the denaturation of intracellular Hb S. With this new approach, we have found more than 10 new antisickling agents that showed beneficial effects without significant adverse effects in the in vitro tests. Of these, five drugs were further studied in vivo using transgenic mice that produce human Hb S, and all five drugs inhibited the formation of sickled cells in the blood and significantly prolonged their survival time under acute hypoxic conditions. The five drugs are NIPRISAN (plant extracts) [Brit. J. Haematol. (BJH) 122:1001,2003], MX-1520 (prodrug of vanillin) (BJH125: 788,2004), 5-hydroxymethyl-2-furfural (5HMF) (BJH128:552,2005), NS3623 (Blood 97: 1461,2001, in vivo results not shown), and FLOCOR (surfactant) (Blood94:420a,1999). Of these 5 drugs, FLOCOR and NS3623 did not have an antisickling effect in vitro. It is quite interesting that FLOCOR and NS3623 which showed no direct antisickling effect in vitro, showed a strong antisickling effect in vivo. Some of our results suggest that combined use of some of these drugs may have a synergistic effect in vivo. Studies on the combined use of these drugs are planned. Researchers with candidate drugs that may have beneficial effects for SCD are welcome to contact the SCD Ref Lab for free in vitro evaluation of their agents (for more information, please see our website at www.tatcom.com/sickle-cell).