Computer simulations of red blood cells associated with Gaucher disease suggest that some modeled cells would deform less, move more slowly through narrow channels and take much longer to pass through a spleen-like slit than control cells. A mixed simulated suspension also had higher viscosity, especially at low shear. The findings describe model behavior rather than direct clinical effects in patients, but they offer a framework for testing how altered cell mechanics and geometry may be connected to blood-flow problems.
The study used dissipative particle dynamics, or DPD, a computer method for representing the mechanics of individual cells and the flow of a cell suspension. The researchers compared a control red-cell model, called CTR-RBC, with three Gaucher disease models, GD-RBC1 through GD-RBC3. They examined deformation under optical-tweezer forces, the motion of the cell membrane in flowing fluid, passage through a microchannel and a modeled splenic slit, and the viscosity of a suspension.
Measurements that shaped the models
The computer work was informed by ektacytometry, a test that measures how much red blood cells elongate under shear, or flowing force. Data from 7 control subjects and 15 untreated Gaucher disease patients showed lower elongation in the disease group at 0.53 and 0.95 pascals, with a Welch's t-test giving p less than 0.05. From 1.69 pascals upward, the groups were statistically indistinguishable. The reported values were means with standard deviations, and no confidence intervals were provided.
Confocal imaging also pointed to changes in cell geometry. Ten control red blood cells from 7 healthy volunteers clustered around a projected area of 73 square micrometers. Twenty Gaucher disease red blood cells from 9 patients were smaller on average: about 66 square micrometers in the mild class and as low as 55 square micrometers in the severe class. The disease cells were also more varied in projected area than the control cells.
A separate lipidomic dataset included 11 control subjects and 16 untreated Gaucher disease subjects. Using UPLC-MS/MS, the researchers quantified glucosylceramide, glucosylsphingosine, sphingosine and sphingosine-1-phosphate. These measurements were used as inputs for the modeled disease-associated changes, rather than as measurements made on every simulated cell.
Three modeled disease profiles
The three Gaucher disease cell models were assigned a shear modulus of 42.57 micronewtons per metre, compared with 4.73 micronewtons per metre for the control model. Shear modulus describes resistance to being stretched or distorted. GD-RBC2 and GD-RBC3 also used a lower surface-to-volume ratio, 1.22 per micrometer versus 1.44 per micrometer in the control. GD-RBC3 was given a bending modulus of 4.8 x 10^-19 joules, compared with 2.4 x 10^-19 joules for the control.
Under a simulated tensile force of 100 piconewtons, GD-RBC3 showed about 27% smaller axial diameter and 42% less transverse compression than the control model. In practical terms, the modeled GD-RBC3 cell yielded less under the same pull. No inferential uncertainty estimate was reported for this simulated deformation result.
The models did not respond in a simple stepwise order. At a shear rate of 100 per second, the simulated tank-treading period was about 0.26 seconds for the control, 0.23 seconds for GD-RBC1, 0.18 seconds for GD-RBC2 and 0.28 seconds for GD-RBC3. Tank-treading is the repeated movement of the cell membrane around the cell. The pattern indicates that the modeled subtypes changed this motion non-monotonically: the first two disease models cycled faster than control, while GD-RBC3 was slower and irregular. No confidence intervals or replication uncertainty for these simulated frequencies were reported.
The largest differences appeared in tight spaces
In a simulated constriction with a pressure difference of 0.15 kilopascals, control cells reached about 2.3 millimeters per second. GD-RBC1, GD-RBC2 and GD-RBC3 reached about 1.7, 1.3 and 1.0 millimeters per second, respectively. Those speeds corresponded to reductions of about 24%, 43% and 56% relative to control. The results suggest increasing resistance across the modeled disease subtypes, with GD-RBC3 moving most slowly. The outputs were descriptive simulations without confidence intervals.
The modeled spleen-like slit produced an even sharper separation. Control passage took about 250 milliseconds, with a spread of 50 milliseconds. GD-RBC1 took about 600-700 milliseconds, while GD-RBC2 and GD-RBC3 took more than 1,200 milliseconds and stalled at the slit entrance. These are results for individual model cells, not direct measurements of red-cell retention or passage in patients.
The suspension simulations used 558 red blood cells at 36% hematocrit, with approximately 600,000 plasma particles and about 879,000 total particles. In the mixed Gaucher disease population, 96.0% of cells had the normal modeled shape and 4.0% had an abnormal shape. At shear rates of 10 per second or less, the disease suspension had a viscosity of about 18.5 centipoise, compared with about 14.3 centipoise for control, an increase of roughly 29%. The elevation remained about 10% at 100 per second and was about 27% at 1,000 per second. The exact allocation of the abnormal fraction between GD-RBC2 and GD-RBC3 was not reported.
A framework awaiting patient-specific tests
The model treats each Gaucher disease subtype as having uniform parameters, so it simplifies variation from cell to cell and from patient to patient. It also omits leukocytes, platelets and detailed splenic vasculature, while the microchannel and splenic geometries are idealized rather than patient-specific. The study therefore supports relationships among modeled red-cell properties and simulated mechanical or rheological outcomes; it does not establish clinical effects or show that changing deformability, geometry, bending modulus or adhesion would improve patient outcomes.
The next tests are direct ones: whether patient-specific red-cell properties and lipid burdens reproduce the modeled subtype hierarchy, whether independent microsphiltration experiments show the predicted delayed or failed passage, and how realistic vascular architecture and other blood-cell types alter the predictions. The supplied analysis also notes that the relationship between sphingolipid burden and membrane bending modulus remains an inference for the most affected model.
The document is a preprint identified as arXiv:2608.24021v1 and dated 25 August 2026. It states that all data supporting the findings are contained in the main text and Supporting Material. The work acknowledged support from NIH Grant R01HL154150 and France 2030 through Idex Universite Paris Cite, with computational resources at Brown University.
Paper data and sources
Original title: Quantifying the Biophysical Properties of Red Blood Cells in Gaucher Disease
Authors: Zhaojie Chai, Marine de Person, Pierre A. Buffet et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-25
DOI: Not available
Original paper · Full text