Sleep phenotype in the Townes mouse model of sickle cell disease

Mouse models of sickle cell disease: Imperfect and yet very informative

The root cause of sickle cell disease (SCD) has been known for nearly a century, however, few therapies to treat the disease are available. Over several decades of work, with advances in gene editing technology and after several iterations of mice with differing genotype/phenotype relationships, researchers have developed humanized SCD mouse models. However, while a large body of preclinical studies has led to huge gains in basic science knowledge about SCD in mice, this knowledge has not led to the development of effective therapies to treat SCD-related complications in humans, thus leading to frustration with the paucity of translational progress in the SCD field. The use of mouse models to study human diseases is based on the genetic and phenotypic similarities between mouse and humans (face validity). The Berkeley and Townes SCD mice express only human globin chains and no mouse hemoglobin. With this genetic composition, these models present many phenotypic similarities, but also significant discrepancies that should be considered when interpreting preclinical studies results. Reviewing genetic and phenotypic similarities and discrepancies and examining studies that have translated to humans and those that have not, offer a better perspective of construct, face, and predictive validities of humanized SCD mouse models.

Release of hepatic xanthine oxidase (XO) to the circulation is protective in intravascular hemolytic crisis

Xanthine oxidase (XO) catalyzes the catabolism of hypoxanthine to xanthine and xanthine to uric acid, generating oxidants as a byproduct. Importantly, XO activity is elevated in numerous hemolytic conditions including sickle cell disease (SCD); however, the role of XO in this context has not been elucidated. Whereas long-standing dogma suggests elevated levels of XO in the vascular compartment contribute to vascular pathology via increased oxidant production, herein, we demonstrate, for the first time, that XO has an unexpected protective role during hemolysis. Using an established hemolysis model, we found that intravascular hemin challenge (40 μmol/kg) resulted in a significant increase in hemolysis and an immense (20-fold) elevation in plasma XO activity in Townes sickle cell phenotype (SS) sickle mice compared to controls. Repeating the hemin challenge model in hepatocyte-specific XO knockout mice transplanted with SS bone marrow confirmed the liver as the source of enhanced circulating XO as these mice demonstrated 100% lethality compared to 40% survival in controls. In addition, studies in murine hepatocytes (AML12) revealed hemin mediates upregulation and release of XO to the medium in a toll like receptor 4 (TLR4)-dependent manner. Furthermore, we demonstrate that XO degrades oxyhemoglobin and releases free hemin and iron in a hydrogen peroxide-dependent manner. Additional biochemical studies revealed purified XO binds free hemin to diminish the potential for deleterious hemin-related redox reactions as well as prevents platelet aggregation. In the aggregate, data herein reveals that intravascular hemin challenge induces XO release by hepatocytes through hemin-TLR4 signaling, resulting in an immense elevation of circulating XO. This increased XO activity in the vascular compartment mediates protection from intravascular hemin crisis by binding and potentially degrading hemin at the apical surface of the endothelium where XO is known to be bound and sequestered by endothelial glycosaminoglycans (GAGs).

A Pharmacometrics Model to Characterize a New Type of Target-Mediated Drug Disposition (TMDD) - Nonlinear Pharmacokinetics of Small-Molecule PF-07059013 Mediated By Its High-capacity Pharmacological Target Hemoglobin With Positive Cooperative Binding

In general, small-molecule target-mediated drug disposition (TMDD) is caused by the interaction of a drug with its high-affinity, low-capacity pharmacological target. In the current work, we developed a pharmacometrics model to characterize a new type of TMDD, where the nonlinear pharmacokinetics (PK) is mediated by a high-capacity pharmacological target with cooperative binding instead of target saturation. The model drug we used was PF-07059013, a noncovalent hemoglobin modulator that demonstrated promising preclinical efficacy to treat sickle cell disease (SCD), and showed complex nonlinear PK in mice with the fraction of unbound drug in blood (fu_b) decreased with an increase in PF-07059013 concentrations/doses due to the positive cooperative binding of PF-07059013 to hemoglobin. Among the various models we evaluated, the best one is a semi-mechanistic model where only drug molecules not bound to hemoglobin were allowed for elimination, with the nonlinear pharmacokinetics being captured by incorporating cooperative binding for drug molecules bound to hemoglobin. Our final model provided valuable insight on target binding-related parameters, such as the Hill coefficient γ (estimated to be 1.6), binding constant K_H (estimated to be 1450 µM), and the amount of total hemoglobin R_tot (estimated to be 2.13 µmol). As the dose selection of a compound with positive cooperative binding is tricky and challenging due to the nonproportional and steep response, our model may be valuable in facilitating the rational dose regimen selection for future preclinical animal and clinical trials for PF-07059013 and other compounds whose nonlinear pharmacokinetics are caused by similar mechanisms.

Pathophysiological characterization of the Townes mouse model for sickle cell disease

A deeper pathophysiologic understanding of available mouse models of sickle cell disease (SCD), such as the Townes model, will help improve preclinical studies. We evaluated groups of Townes mice expressing either normal adult human hemoglobin (HbA), sickle cell trait (HbAS), or SCD (HbS), comparing younger versus older adults, and females versus males. We obtained hematologic parameters in steady-state and hypoxic conditions and evaluated metabolic markers and cytokines from serum. Kidney function was evaluated by measuring the urine protein/creatinine ratio and urine osmolality. In vivo studies included von Frey assay, non-invasive plethysmography, and echocardiography. Histopathological evaluations were performed in lung, liver, spleen, and kidney tissues. HbS mice displayed elevated hemolysis markers and white blood cell counts, with some increases more pronounced in older adults. After extended in vivo hypoxia, hemoglobin, platelet counts, and white blood cell counts decreased significantly in HbS mice, whereas they remained stable in HbA mice. Cytokine analyses showed increased TNF-alpha in HbS mice. Kidney function assays revealed worsened kidney function in HbS mice. The von Frey assay showed a lower threshold to response in the HbS mice than controls, with more noticeable differences in males. Echocardiography in HbS mice suggested left ventricular hypertrophy and dilatation. Plethysmography suggested obstructive lung disease and inflammatory changes in HbS mice. Histopathological studies showed vascular congestion, increased iron deposition, and disruption of normal tissue architecture in HbS mice. These data correlate with clinical manifestations in SCD patients and highlight analyses and groups to be included in preclinical therapeutic studies.

The relationship between pain and sleep in pediatric sickle cell disease

BACKGROUND

Pain and sleep disturbances are prevalent complications experienced by pediatric patients with sickle cell disease (SCD). This study aims to identify associations between pain and sleep, and to characterize sleep chronotype and social jetlag in children and adolescent patients with SCD.

METHODS

We performed a cross-sectional survey of 105 pediatric patients with SCD aged 8-17 years using PROMIS (Patient Reported Outcomes Measurement System) pain interference, sleep disturbance, and sleep-related impairment item banks. The μMCTQ (Ultra-short Munich Chronotype Questionnaire) assessed chronotype and social jetlag. Analyses were performed to assess associations between PROMIS measures, sleep patterns, and clinical variables.

RESULTS

Female participants reported higher T-scores for sleep-related impairment than males (females: 56.7 ± 10 vs. males 50.2 ± 9.4, p = .0009). Patients with one or more emergency department (ED) visits for pain in the last 12 months reported greater sleep disturbance (55.0 ± 8.5 vs. 50.7 ± 10, p = .046) and sleep-related impairment (57.1 ± 9.3 vs. 52.1 ± 10.2, p = .03) than patients without any ED visits for pain in the last 12 months. Pain interference was significantly associated with both sleep disturbance (r = .49, p < .0001) and sleep-related impairment (r = .46, p < .0001). The average mid-sleep time was 4:14 ± 1:44 a.m. and the average social jetlag (hh:mm) was 2:32 ± 1:35.

CONCLUSION

Our study demonstrates that pain interference is associated with both sleep disturbance and sleep-related impairment. PROMIS measures can identify patients that suffer from pain and sleep disturbances and highlights the need to conduct longitudinal prospective studies to define the directionality of pain and sleep in SCD.

A role for the circadian transcription factor NPAS2 in the progressive loss of non-rapid eye movement sleep and increased arousal during fentanyl withdrawal in male mice

RATIONALE

Synthetic opioids like fentanyl are contributing to the rise in rates of opioid use disorder and drug overdose deaths. Sleep dysfunction and circadian rhythm disruption may worsen during opioid withdrawal and persist during abstinence. Severe and persistent sleep and circadian alterations are putative factors in opioid craving and relapse. However, very little is known about the impact of fentanyl on sleep architecture and sleep-wake cycles, particularly opioid withdrawal. Further, circadian rhythms regulate sleep-wake cycles, and the circadian transcription factor, neuronal PAS domain 2 (NPAS2) is involved in the modulation of sleep architecture and drug reward. Here, we investigate the role of NPAS2 in fentanyl-induced sleep alterations.

OBJECTIVES

To determine the effect of fentanyl administration and withdrawal on sleep architecture, and the role of NPAS2 as a factor in fentanyl-induced sleep changes.

METHODS

Electroencephalography (EEG) and electromyography (EMG) was used to measure non-rapid eye movement sleep (NREMS) and rapid eye movement sleep (REMS) at baseline and following acute and chronic fentanyl administration in wild-type and NPAS2-deficient male mice.

RESULTS

Acute and chronic administration of fentanyl led to increased wake and arousal in both wild-type and NPAS2-deficient mice, an effect that was more pronounced in NPAS2-deficient mice. Chronic fentanyl administration led to decreased NREMS, which persisted during withdrawal, progressively decreasing from day 1 to 4 of withdrawal. The impact of fentanyl on NREMS and arousal was more pronounced in NPAS2-deficient mice.

CONCLUSIONS

Chronic fentanyl disrupts NREMS, leading to a progressive loss of NREMS during subsequent days of withdrawal. Loss of NPAS2 exacerbates the impact of fentanyl on sleep and wake, revealing a potential role for the circadian transcription factor in opioid-induced sleep changes.

Hydroxyurea Decouples Persistent F-Cell Elevation and Induction of γ-Globin

Mechanisms that control the fetal-to-adult hemoglobin switch are attractive therapeutic targets in sickle cell disease. In this study, we investigated developmental γ-globin silencing in the Townes humanized knock-in mouse model, which harbors a construct containing the human γ-, β^A-, and β^S-globin genes, and examined the utility of this model in evaluation of pharmacologic induction of fetal hemoglobin (HbF). We studied mouse pups on the day of delivery (P0) to 28 days after birth (P28). Regardless of the hemoglobin genotype (SS, AS, or AA), the proportion of F cells in peripheral blood was 100% at P0, declined sharply to 20% at P2, and was virtually undetectable at P14. Developmental γ-globin silencing in Townes mice was complete at P4 in association with significantly increased BCL11A expression in the primary erythropoietic organs of the mouse. Hydroxyurea given at P2 significantly sustained elevated percentages of F cells in mice at P14. However, the percentage of F cells declined at P14 for treatment begun at P4. A lack of augmentation of γ-globin mRNA in erythroid tissues suggests that the apparent increase in HbF in red cells caused by hydroxyurea was not due to sustained or re-activation of γ-globin transcription, but was instead a function of erythropoiesis suppression. Thus, we provide new details of the hemoglobin switch in the Townes murine model that recapitulates postnatal γ- to β-globin switch in humans and identify the myelosuppressive toxicity of hydroxyurea as a superseding factor in interpreting pharmacologic induction of HbF.

Chlorine inhalation induces acute chest syndrome in humanized sickle cell mouse model and ameliorated by postexposure hemopexin

Triggering factors of Acute Chest Syndrome (ACS) is a leading cause of death in patients with Sickle Cell Disease (SCD) and targeted therapies are limited. Chlorine (Cl2) inhalation happens frequently, but its role as a potential trigger of ACS has not been determined. In this study, we hypothesized that Cl2 exposure resembling that in the vicinity of industrial accidents induces acute hemolysis with acute lung injury, reminiscent of ACS in humanized SCD mice. When exposed to Cl2 (500 ppm for 30 min), 64% of SCD mice succumbed within 6 h while none of the control mice expressing normal human hemoglobin died (p<0.01). Surviving SCD mice had evidence of acute hemolysis, respiratory acidosis, acute lung injury, and high concentrations of chlorinated palmitic and stearic acids (p<0.05) in their plasmas and RBCs compared to controls. Treatment with a single intraperitoneal dose of human hemopexin 30 min after Cl2 inhalation reduced mortality to around 15% (p<0.01) with reduced hemolysis (decreased RBCs fragility (p<0.001) and returned plasma heme to normal levels (p<0.0001)), improved oxygenation (p<0.0001) and reduced acute lung injury scores (p<0.0001). RBCs from SCD mice had significant levels of carbonylation (which predisposes RBCs to hemolysis) 6 h post-Cl2 exposure which were absent in RBCs of mice treated with hemopexin. To understand the mechanisms leading to carbonylation, we incubated RBCs from SCD mice with chlorinated lipids and identified sickling and increased hemolysis compared to RBCs obtained from control mice and treated similarly. Our study indicates that Cl2 inhalation induces ACS in SCD mice via induction of acute hemolysis, and that post exposure administration of hemopexin reduces mortality and lung injury. Our data suggest that SCD patients are vulnerable in Cl2 exposure incidents and that hemopexin is a potential therapeutic agent.

Xanthine Oxidase Drives Hemolysis and Vascular Malfunction in Sickle Cell Disease

OBJECTIVE

Chronic hemolysis is a hallmark of sickle cell disease (SCD) and a driver of vasculopathy; however, the mechanisms contributing to hemolysis remain incompletely understood. Although XO (xanthine oxidase) activity has been shown to be elevated in SCD, its role remains unknown. XO binds endothelium and generates oxidants as a byproduct of hypoxanthine and xanthine catabolism. We hypothesized that XO inhibition decreases oxidant production leading to less hemolysis. Approach and Results: Wild-type mice were bone marrow transplanted with control (AA) or sickle (SS) Townes bone marrow. After 12 weeks, mice were treated with 10 mg/kg per day of febuxostat (Uloric), Food and Drug Administration-approved XO inhibitor, for 10 weeks. Hematologic analysis demonstrated increased hematocrit, cellular hemoglobin, and red blood cells, with no change in reticulocyte percentage. Significant decreases in cell-free hemoglobin and increases in haptoglobin suggest XO inhibition decreased hemolysis. Myographic studies demonstrated improved pulmonary vascular dilation and blunted constriction, indicating improved pulmonary vasoreactivity, whereas pulmonary pressure and cardiac function were unaffected. The role of hepatic XO in SCD was evaluated by bone marrow transplanting hepatocyte-specific XO knockout mice with SS Townes bone marrow. However, hepatocyte-specific XO knockout, which results in >50% diminution in circulating XO, did not affect hemolysis levels or vascular function, suggesting hepatocyte-derived elevation of circulating XO is not the driver of hemolysis in SCD.

CONCLUSIONS

Ten weeks of febuxostat treatment significantly decreased hemolysis and improved pulmonary vasoreactivity in a mouse model of SCD. Although hepatic XO accounts for >50% of circulating XO, it is not the source of XO driving hemolysis in SCD.

Smooth muscle cytochrome b5 reductase 3 deficiency accelerates pulmonary hypertension development in sickle cell mice

Pulmonary and systemic vasculopathies are significant risk factors for early morbidity and death in patients with sickle cell disease (SCD). An underlying mechanism of SCD vasculopathy is vascular smooth muscle (VSM) nitric oxide (NO) resistance, which is mediated by NO scavenging reactions with plasma hemoglobin (Hb) and reactive oxygen species that can oxidize soluble guanylyl cyclase (sGC), the NO receptor. Prior studies show that cytochrome b5 reductase 3 (CYB5R3), known as methemoglobin reductase in erythrocytes, functions in VSM as an sGC heme iron reductase critical for reducing and sensitizing sGC to NO and generating cyclic guanosine monophosphate for vasodilation. Therefore, we hypothesized that VSM CYB5R3 deficiency accelerates development of pulmonary hypertension (PH) in SCD. Bone marrow transplant was used to create SCD chimeric mice with background smooth muscle cell (SMC)-specific tamoxifen-inducible Cyb5r3 knockout (SMC R3 KO) and wild-type (WT) control. Three weeks after completing tamoxifen treatment, we observed 60% knockdown of pulmonary arterial SMC CYB5R3, 5 to 6 mm Hg elevated right-ventricular (RV) maximum systolic pressure (RVmaxSP) and biventricular hypertrophy in SS chimeras with SMC R3 KO (SS/R3KD) relative to WT (SS/R3WT). RV contractility, heart rate, hematological parameters, and cell-free Hb were similar between groups. When identically generated SS/R3 chimeras were studied 12 weeks after completing tamoxifen treatment, RVmaxSP in SS/R3KD had not increased further, but RV hypertrophy relative to SS/R3WT persisted. These are the first studies to establish involvement of SMC CYB5R3 in SCD-associated development of PH, which can exist in mice by 5 weeks of SMC CYB5R3 protein deficiency.