Ionophore-mediated swelling of erythrocytes as a therapeutic mechanism in sickle cell disease

Abstract

Sickle cell disease (SCD) is characterized by sickle hemoglobin (HbS) which polymerizes under deoxygenated conditions to form a stiff, sickled erythrocyte. The dehydration of sickle erythrocytes increases intracellular HbS concentration and the propensity of erythrocyte sickling. Prevention of this mechanism may provide a target for potential SCD therapy investigation. Ionophores such as monensin can increase erythrocyte sodium permeability by facilitating its transmembrane transport, leading to osmotic swelling of the erythrocyte and decreased hemoglobin concentration. In this study, we treated 13 blood samples from patients with SCD with 10 nM of monensin ex vivo. We measured changes in cell volume and hemoglobin concentration in response to monensin treatment, and we perfused treated blood samples through a microfluidic device that permits quantification of blood flow under controlled hypoxia. Monensin treatment led to increases in cell volume and reductions in hemoglobin concentration in most blood samples, though the degree of response varied across samples. Monensin-treated samples also demonstrated reduced blood flow impairment under hypoxic conditions relative to untreated controls. Moreover, there was a significant correlation between the improvement in blood flow and the decrease in hemoglobin concentration. Thus, our results demonstrate that a reduction in intracellular HbS concentration by osmotic swelling improves blood flow under hypoxic conditions. Although the toxicity of monensin will likely prevent it from being a viable clinical treatment, these results suggest that osmotic swelling should be investigated further as a potential mechanism for SCD therapy.

Figure 1.

Data collection and analysis. An overview of methods of sample preparation, data collection, and defining rheological variables used throughout the study. (A) Schematic of the monensin treatment workflow. (B) Representative image of raw velocity data (below) as it relates to oxygen tension (above) from a single sickle cell disease (SCD) patient sample. In the bottom panel of (B), blood flow velocity is compared between monensin treatment (red) and the untreated control (blue). The average oxygenated shear rate during experiments was 355 s^-1, within physiologic range for channel dimensions. In this sample, it appears that the velocity at normoxia of the monensin-treated condition is lower than that of the untreated condition. In order to address the differences in normoxic velocities between treatment conditions, conductance of all 13 samples was calculated to determine if additional variables were present contributing to normoxic velocities (Online Supplementary Figure S3). There were no significant differences in conductance at normoxia between treatment conditions in all samples, indicating velocity differences at normoxia were related to driving pressure. (C) The oxygenated (160 mmHg) and deoxygenated (0 mmHg) sections of the collected velocity data in (B), normalized by the average oxygenated steady state velocity for the representative sample. The representative single patient data in (C) demonstrates a 13% velocity response for a monensin-treated sample and a 33% response for the untreated control. This corresponds to a velocity recovery of 20% after monensin treatment. Velocity response is calculated using the difference between oxygenated (160 mmHg) and deoxygenated (0 mmHg) velocities and velocity recovery is calculated by the difference in the control and monensin treated response. SN: supernatant; PBS: phosphate-buffered saline, SS: steady-state.

Figure 2.

Dependent parameter analysis. Correlative data from 13 untreated sickle cell disease (SCD) samples to determine the relationship, if any, between mean corpuscular volume (MCV) and mean cell hemoglobin concentration (MCHC) and each variable’s relation to sample velocity response. A Pearson correlation coefficient analysis was used to determine the strength of the linear relationship and a two-tailed analysis of the Pearson coefficient was used to determine significance of the correlation. (A) No correlation was identified between MCV and MCHC (r=-0.008, P=0.982), establishing MCV and MCHC as independent variables. (B) MCV as it relates to velocity response. No correlation was identified between MCV values and velocity response (r=-0.13, P=0.660). A slope of -0.001 and -0.1 are found in (A) and (B) respectively. The slope for these figures is provided for clarification of the scale for (A and B). (C) MCHC and velocity response had a significant correlation (r= 0.83, P<0.001).

Figure 3.

Monensin treatment efficacy and variability. Summary data of monensin treatment. (A and B) The effect of monensin on mean cell hemoglobin concentration (MCHC) and mean corpuscular volume (MCV). The monensin-treated group had significantly higher MCV and lower MCHC compared to the control group. (C) There was a significantly lower velocity response in the monensin-treated group compared to that in the control group. Significance between control and monensin-treated groups was determined using a Wilcoxon signed-rank test and indicated by the asterisks (*) denoting P<0.01. Sample ID represents the de-identified patient ID corresponding to the sample. Error bars indicate the standard deviation in velocity response over 3 oxygenation/deoxygenations cycles. SS: steady state.

Figure 4.

Linear correlation analysis. In order to determine which effect of monensin was driving sample velocity recovery, a correlation analysis to define the relationship between velocity recovery and monensin-induced mean corpuscular volume (MCV) or mean cell hemoglobin concentration (MCHC) change was completed. A Pearson correlation coefficient analysis was conducted to determine the strength of the linear relationship. (A) Change in MCV and change in MCHC was significantly positively correlated (r=0.91, P<0.001). (B) Velocity recovery and MCV change also were positively significantly correlated (r=0.87, P<0.001). (C) The strongest correlation was found between velocity recovery and MCHC change (r=0.96, P<0.01).