Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2024 Jul 30;40(6):e3501. doi: 10.1002/btpr.3501

Tangential flow filtration‐facilitated purification of human red blood cell membrane fragments and its preferential use in removing unencapsulated material from resealed red blood cell ghosts compared to centrifugation

Xiangming Gu 1, Andre F Palmer 1,✉
PMCID: PMC11659797  NIHMSID: NIHMS2011623  PMID: 39076022

Abstract

The biodistribution of many therapeutics is controlled by the immune system. In addition, some molecules are cytotoxic when not encapsulated inside of larger cellular structures, such as hemoglobin (Hb) encapsulation inside of red blood cells (RBCs). To counter immune system recognition and cytotoxicity, drug delivery systems based on red blood cell membrane fragments (RBCMFs) have been proposed as a strategy for creating immunoprivileged therapeutics. However, the use of RBCMFs for drug delivery applications requires purification of RBCMFs at large scale from lysed RBCs free of their intracellular components. In this study, we were able to successfully use tangential flow filtration (TFF) to remove >99% of cell‐free Hb from lysed RBCs at high concentrations (30%–40% v/v), producing RBCMFs that were 2.68 ± 0.17 μm in diameter. We were also able to characterize the RBCMFs more thoroughly than prior work, including measurement of particle zeta potential, along with individual TFF diacycle data on the cell‐free Hb concentration in solution and time per diacycle, as well as concentration and size of the RBCMFs. In addition to purifying RBCMFs from lysed RBCs, we utilized a hypertonic solution to reseal purified RBCMFs encapsulating a model protein (Hb) to yield resealed Hb‐encapsulated RBC ghosts (Hb‐RBCGs). TFF was then compared against centrifugation as an alternative method for removing unencapsulated Hb from Hb‐RBCGs, and the effects that each washing method on the resulting Hb‐RBCG biophysical properties was assessed.

Keywords: hemoglobin removal, red blood cell, red blood cell ghost, red blood cell membrane fragment, tangential flow filtration

1. INTRODUCTION

A critical challenge in the development of therapeutic nanoparticles (NPs) is their recognition by the immune system, and subsequent clearance from the circulation. 1 The foreign surface properties of NPs leads to macrophage uptake and rapid clearance, which significantly reduces their circulatory half‐life and therapeutic efficacy. 2 , 3 These limitations are prevalent in many biomedical NP applications including drug delivery, 4 , 5 diagnostics, and imaging. 6 , 7 Therefore, various synthetic materials have been used to shield the NP surface to evade immune recognition. 8 , 9 NP surface conjugation with poly(ethylene glycol) (PEG) is known to increase NP residence time, due to its hydrophilic surface properties. 10 , 11 Unfortunately, there are concerns regarding the immunocompatibility of PEG. Some patients have preexisting PEG antibodies, or develop PEG antibodies after repeated exposure to PEG. 12 Polydopamine (PDA) surface coatings have also been studied and seem to impart anti‐inflammatory properties by scavenging reactive oxygen species. 13 However, there is a risk of forming noncovalently bound oligomers of dopamine, 14 which do not attach to the NP surface and fail to confer benefits.

The use of red blood cell membrane fragments (RBCMFs) to surface‐treat NPs has emerged as a robust and versatile approach for integrating natural and synthetic biomaterials to form NPs that evade immune recognition, since these coated NPs mimic the surface of red blood cells (RBCs) and mitigate macrophage uptake, 15 while not requiring PEG or PDA surface conjugation. For the purpose of this work, we establish a distinction between the lysed, empty RBCMFs, and the resealed RBCMFs, also known as resealed RBC ghosts (RBCGs). An advantage of RBCGs is that, when resealed via the addition of hypertonic solution, they are shown to maintain their outer membrane as “right side out,” rather than being turned inside out, 16 , 17 the latter is linked to uptake by cellular recognition of phosphatidylserine that is traditionally present in the inner membrane leaflet. 18 Purification of RBCMFs requires RBC lysis, and subsequent removal of cell‐free hemoglobin (Hb) from the lysed RBCs, since cell‐free Hb is cytotoxic. 19 Prior work has demonstrated the ability to purify Hb from lysed RBCs at scale from lysed RBCs via tangential flow filtration (TFF). However, that work focused on purification of Hb, and treated the RBCMFs as waste material. As such, the RBCMFs were not extensively characterized. 20 This prior work also did not confirm the removal of the majority of cell‐free Hb from the RBCMFs, since their focus was on purification of Hb, and not the RBCMFs. Prior work to successfully isolate RBCMFs that did remove the majority of cell‐free Hb used very low RBC volume fractions (<5%) and took hours to isolate the membranes, which is highly inefficient when considering the low concentration of recovered RBCMFs. 21 This highlights the need for a scalable process to remove the majority of cell‐free Hb from lysed RBCs at higher volume fractions, without unduly extending the time required to purify the RBCMFs.

Once RBCMFs are purified, there is also a need to develop a scalable platform for purification of RBCGs encapsulating a therapeutic or diagnostic agent, since the unencapsulated material must be removed from the RBCGs to prevent fast clearance of the therapeutic. Therefore, scalability is a key challenge toward creating a viable RBCG manufacturing platform, which can otherwise limit translation from bench‐scale production to commercial‐scale production. Many processes for manufacturing RBCGs encapsulating a therapeutic agent are simply not scalable. 22 Previously‐cited protocols often utilize microextrusion, microfluidics, or centrifugation as key steps in the production of therapeutic or diagnsotic encapsulated RBCGs, which are difficult to scale.

In this study, we developed a simple and scalable method for preparing Hb‐free RBCMFs from highly concentrated RBC solutions (30%–40% volume fraction). TFF was used to remove cell‐free Hb from lysed RBCs in the permeate stream, while retaining RBCMFs in the retentate. We also developed a scalable purification method to produce resealed RBCGs that could be loaded with a therapeutic or diagnostic agent for systemic delivery. As our model protein encapsulant, we utilized Hb due to the ready availability of Hb activity assays in the lab. After resealing identically‐prepared samples of Hb‐containing RBCGs, centrifugation or TFF was used to remove unencapsulated Hb from solution, leaving behind the Hb encapsulated RBCGs for subsequent biophysical analysis.

2. MATERIALS AND METHODS

2.1. Materials

Hydrochloric acid (HCl), sodium hydroxide (NaOH), sodium chloride (NaCl), potassium chloride (KCl), sodium phosphate monobasic (NaH2PO4), and sodium phosphate dibasic (Na2HPO4) were purchased from Sigma‐Aldrich (St. Louis, MO). Nonirradiated packed human RBC units were obtained from The Ohio State University Wexner Center Office of Transfusion Services (Columbus, OH) for purification of RBCMFs. The cells were expired, and thus unsuitable for transfusion. Size 16 and 17 silicone and biopharmaceutical tubing were purchased from Cole‐Parmer (Vernon Hills, IL). Hollow fiber TFF modules with 0.65 μm and 500 kDa molecular weight cut‐offs (MWCO) were obtained from Repligen (Rancho Dominguez, CA).

2.2. RBC wash and membrane fragment purification process

RBCs were washed via TFF as described by Lu et al. 23 RBCs were diluted to 45% volume fraction (hematocrit) and subjected to 10 constant volume buffer exchanges with 0.9 wt % NaCl as the wash solution using a 0.65 μm TFF module and a centrifugal pump to remove cell‐free Hb and other cell‐free proteins from the RBCs. The washed RBCs were lysed by mixing them with an equal volume of phosphate buffer (PB, 3.75 mM, pH 7.4), yielding a mixture RBCMFs and cell‐free proteins. The total volume of the lysate was reduced to ~300 mL and was subjected to 10 diacyles of constant volume diafiltration on a 0.65 μm TFF module to remove intracellular debris and cell‐free Hb using PB as the wash solution. After complete removal of cell‐free Hb, RBCMFs were concentrated to a minimum system volume of ~80 mL. This final step was performed to maximize the RBCMF concentration for resealing, which would increase the Hb encapsulation efficiency of the resealed RBCGs. The RBC wash and RBCMF purification process are shown in Figure 1.

FIGURE 1.

FIGURE 1

Open in a new tab

Red blood cell (RBC) wash and red blood cell membrane fragment (RBCMF) purification process. The tangential flow filtration system is similar for both the RBC washing and RBCMF purification processes, with only the type of pump and wash solution differing between the two processes. A centrifugal pump was used for the RBC wash process to reduce hemolysis, while a peristaltic pump was used to purify RBCMFs from lysed RBCs. Created with BioRender.com.

2.3. RBCG resealing process

Resealing RBCMFs to form resealed RBCGs was achieved by mixing a hypertonic solution with RBCMFs and purified Hb. Hb, purified via methods described in literature, 24 , 25 was added to the concentrated RBCMFs produced in Section 2.2, and used at a Hb concentration of ~100 mg/mL. Aliquots of 4 M NaCl and 4 M KCl, taken from stock solutions, were added via syringe pump to a stirring mixture of RBCMFs and purified Hb to achieve Na+ and K+ concentrations of approximately 0.77 and 0.021 M, respectively. These salt levels are significantly higher than physiological salt levels, thus creating a hypertonic environment. The mixture was cooled for 5 min at −20°C, and then incubated at 37°C for 60 min. The resulting Hb‐encapsulated resealed RBCGs (Hb‐RBCGs) had their volume fraction adjusted to ~45% before being stored at 4°C for further analysis. As a control to test various biophysical properties of the RBCGs, empty RBCGs were also prepared, without any encapsulated protein. The RBCMF resealing process used to encapsulate Hb to yield washed Hb‐RBCGs is shown in Figure 2 and shows both TFF and centrifugation modalities.

FIGURE 2.

FIGURE 2

Open in a new tab

Red blood cell membrane fragment (RBCMF) resealing process to encapsulate Hb to yield washed Hb‐RBCGs. Created with BioRender.com. RBCGs, red blood cell ghosts; TFF, tangential flow filtration.

2.4. Removal of free Hb from resealed RBCGs

2.4.1. RBCG wash via centrifugation

The first method explored for removing unencapsulated Hb from resealed Hb‐RBCGs was differential centrifugation. This method is the industry standard for washing RBCs and other micron‐scale particles. The applied centrifugal force ranges between 500 and 1500 g for differential centrifugation of RBCs due to its relative lack of detrimental effects on the RBC lipid membrane. 26 However, Hb‐RBCGs required a slightly higher centrifugal force to properly separate the pelleted Hb‐RBCGs from the supernatant, which may be due to their smaller size and relatively lower Hb content (i.e., density) than intact RBCs. Thus, Hb‐RBCGs were washed by centrifuging the solution at 2000 g for 20 min to form a pellet of Hb‐RBCGs. After supernatant removal, the Hb‐RBCG pellet was resuspended with fresh PBS. This process was repeated three times. After the final centrifugation step, the volume fraction of the Hb‐RBCGs was adjusted via the addition of PBS to achieve ~45% v/v of Hb‐RBCGs, to mimic a physiological RBC hematocrit. Hb‐RBCGs purified via centrifugation are referred to as Hb‐RBCGs‐C.

2.4.2. RBCG wash via TFF

A bench‐scale TFF module with a 500 kDa MWCO membrane was used to remove unencapsulated Hb from resealed Hb‐RBCGs, which has a molecular weight of ~64 kDa. The initial 50 mL sample of Hb‐RBCGs was concentrated down to ~40 mL, before being subjected to constant volume diafiltration using PBS as the wash solution. The constant volume diafiltration process was terminated when the permeate concentration of cell‐free Hb got below 0.25 mg/mL, at which point the cell‐free Hb levels remaining in the Hb‐RBCGs was negligible. At this end point, the washed Hb‐RBCGs were concentrated to attain a volume fraction of ~45% Hb‐RBCGs. The samples were stored in the refrigerator for further analysis. Hb‐RBCGs purified via TFF are referred to as Hb‐RBCGs‐TFF.

2.5. Characterization of RBCMFs

2.5.1. ζ potential analysis

ζ potential analysis was performed on all fresh and thawed RBCMFs using a Brookhaven Instruments ZetaPals instrument (Holtsville, NY), and followed a procedure similar to Huang et al. 27 RBCMFs were diluted 100× to a concentration of ~2 × 107 RBCMFs/mL in deionized (DI) water for analysis.

2.5.2. Size characterization

Transmission electron microscopy analysis

The morphology of RBCs and RBCMFs was analyzed using a FEI Tecnai G2 Biotwin transmission electron microscope (TEM) (FEI, Hillsboro, OR). All samples were diluted 100× in phosphate buffered saline (PBS) for analysis.

Coulter Counter size analysis

To analyze the size distribution of RBCMFs on the micron scale, Coulter Counter (CC) analysis was performed 28 on a 100× dilution of the sample, which was further diluted by adding 0.2 mL of the diluted sample into 20 mL of double‐filtered Isoton‐III buffer (Beckman Coulter, Brea, CA). Particle counts and size distribution data were collected from each sample, and each sample was prepared identically. This method is preferable over dynamic light scattering (DLS) for analyzing particles in this size range, due to the difficulty that DLS often encounters when measuring nonspherical particles that are larger than 1 μm in diameter.

2.5.3. Quantification of Hb concentration

All cell‐free Hb concentrations were measured using the Winterbourn assay. 29 Samples were diluted to <1 mg/mL Hb, and 1 mL was pipetted into a disposable cuvette. The absorbance spectrum of each sample was measured using an OLIS 8452A diode array spectrophotometer (OLIS, Athens, GA). Due to the fact that large particles scatter light, RBCMF retentate samples were additionally centrifuged and filtered through a 0.2 μm poly‐ether sulfone filter. This was acceptable, since the assay aims to quantify the amount of cell‐free Hb present in the solution and does not require the presence of the RBCMFs themselves.

2.6. RBCG analysis and characterization

2.6.1. Hb retention in Hb‐RBCGs after washing

Hb encapsulation inside Hb‐RBCGs was calculated by measuring the mass of cell‐free Hb removed from the Hb‐RBCGs after the washing procedure relative to the initial bolus of Hb added. Cell‐free Hb concentrations were measured using the Winterbourn assay 29 using a method similar to that described in Section 2.5.3. Since the starting mass of Hb is known, the remaining Hb in the washed Hb‐RBCGs must be encapsulated inside the RBCGs themselves. To account for slight variations in the initial concentration of RBCMFs used to prepare Hb‐RBCGs due to the different batches of materials produced, Hb retention inside the washed Hb‐RBCGs was normalized by dividing it by the final volume fraction occupied by the Hb‐RBCGs. This yields a normalized Hb encapsulation efficiency.

2.6.2. Viscosity analysis

The viscosity of Hb‐RBCGs‐C and Hb‐RBCGs‐TFF was compared to controls of unmodified RBCs and empty RBCGs at 45% volume fraction using a Brookfield‐Ametek LV Rheometer (Middleborough, MA). A 500 μL of sample was loaded into the rheometer, and the viscosity at a shear rate of 160 s−1 was measured at 37°C.

2.6.3. RBCG size analysis

RBCG size distributions were measured via CC using the method described in Section 2.5.2.3. A 100× dilution of the sample was taken, 28 which was further diluted by adding 0.2 mL of the diluted sample into 20 mL of double‐filtered Isoton‐III buffer (Beckman Coulter, Brea, CA). Particle counts and size distribution data were collected for each sample, and each sample was prepared identically.

2.6.4. Oxygen equilibrium and offloading analysis

To assess any differences in the oxygen equilibria and oxygen offloading kinetics between Hb‐RBCGs‐TFF and Hb‐RBCGs‐C, the oxygen equilibrium curve (OEC) and the kinetics of oxygen offloading were measured using a Hemox Analyzer (TCS Scientific, New Hope, PA) and a stop‐flow UV–visible spectrometer (Applied Photophysics Ltd., Surrey, UK), respectively. The OEC was measured and analyzed via the Hill equation to regress the partial pressure of oxygen (pO2) where 50% of the Hb is saturated with oxygen (P50) and the Hill coefficient (n) as described in the literature. 30 The OEC for Hb‐RBCGs was compared against OECs for cell‐free Hb and fresh RBCs obtained from the literature. 31 Oxygen offloading kinetics were measured and analyzed as described in the literature to regress the oxygen offloading rate constant (koff,O2) for Hb‐RBCGs and compared against values for cell‐free Hb and RBCs obtained from the literature. 31

2.7. Statistical analysis

The IBM SPSS package (IBM, Armonk, NY) was used for all statistical analysis, and statistical significance was set at p < 0.05. Except where otherwise noted, all tables report the mean value ± one standard deviation.

3. RESULTS AND DISCUSSION

3.1. RBCMF size measurements

3.1.1. RBCMF hydrodynamic diameter and morphology

CC analysis of the unfiltered RBCMF species measured an average diameter of 2.68 ± 0.17 μm. It is important to note that the hemolysis and subsequent TFF purification process exposes RBCMFs to shear stress, which most likely contributed to the existence of RBCMF species that fall within a smaller size distribution compared with intact RBCs, which were measured to have a diameter of 4.74 ± 0.76 μm. The RBC diameter measured is smaller than reported values for RBCs in the literature, but consistent with measurements of RBCs using CC, 32 due to the unique morphology of RBCs and the assumptions made by the CC instrument software. TEM image results are shown below in Figure 3, and clearly indicate the morphological differences between RBCs, RBCMFs, and resealed Hb‐RBCGs, respectively. RBCMFs (Figure 3b) roughly retain the overall rounded morphology of native RBCs (Figure 3a), while RBCGs have more irregular shapes. While the biconcave disk shape is optimal for RBC flow in capillaries, 33 and since the RBCGs do not adhere to this morphology, their reduced size may help offset morphological changes, and other properties, such as solution viscosity, which will be important to consider when assessing their flow behavior in the microcirculation.

FIGURE 3.

FIGURE 3

Open in a new tab

Transmission electron microscopy images for expired red blood cells (a), red blood cell membrane fragments (b), and Hb‐RBCGs (c). RBCG, red blood cell ghost.

3.2. Diafiltration achieves cell‐free Hb removal from RBCMFs

After RBC lysis and during TFF purification of RBCMFs, the concentration of cell‐free Hb in the permeate decreased from an average of 19.0 ± 2.2 to 0.04 ± 0.01 mg/mL after 10 diacycles. Similarly, the cell‐free Hb concentration in the retentate decreased from 48.7 ± 4.0 to 0.3 ± 0.07 mg/mL. Figure 4 shows the reduction in cell‐free Hb with increasing diacycles. Both reductions (i.e., permeate and retentate) in cell‐free Hb levels were statistically significant (p < 0.05) when comparing initial and final cell‐free Hb concentrations to each other. The RBCMF purification protocol demonstrates the ability to remove >99% of the cell‐free Hb present in the initially lysed RBC solution. The 0.65 μm HF membrane was effective in removing cell‐free Hb and other debris below its MWCO, while retaining RBCMFs larger than the MWCO. Similar findings were observed for all replicates (N = 10). Differences in cell‐free Hb retentate concentrations were significant (p < 0.05) until after the eighth diacycle, where subsequent diacycles failed to yield significant changes in the cell‐free Hb concentration. The cell‐free Hb differences in the permeate were significantly different until after the seventh diacycle, where again subsequent diacycles failed to yield significant reduction in the cell‐free Hb concentration.

FIGURE 4.

FIGURE 4

Open in a new tab

Cell‐free Hb concentration during the tangential flow filtration facilitated red blood cell membrane fragment purification process for both retentate (a) and permeate (b) samples. ANOVA testing revealed all differences to be significant at the p < 0.05 level except for those marked with “Ψ.”

3.3. Effect of TFF on diacycle time and RBCMF concentration and size

Analysis of the individual TFF diacycles required to remove >99% Hb yielded insights into the effects of the TFF washing process on the diacycle time and RBCMF concentration and size, and these findings are summarized in Figure 5, with n = 11 replicates. As shown in Figure 5a, the time required to complete each diacycle did not significantly increase with increasing number of diacycles (p > 0.05). Outside factors such as the number of times each TFF cartridge had been previously used had an impact on batch‐to‐batch diacycle times, but overall no significant change in diacycle time was observed, both within each replicate and when combining all replicates together. Additionally, no significant change in RBCMF concentration was observed with increasing number of diacycles (p > 0.05), and particle size was only slightly impacted. Specifically, diacycles 1 and 2 had significantly larger RBCMFs than diacycles 6–10 (p < 0.05), though no other significant size differences across any of the groups were detected. Figure 5b,c highlight these trends, and suggest that the shear forces applied in the TFF circuit do not significantly alter the concentration and size of RBCMFs for the 10 diacycles required to remove 99% of the cell‐free Hb. The lack of change in RBCMF concentration and only slight decrease in particle size suggest that TFF can acceptably perform its major function of removing cell‐free Hb without significantly altering the size of the resulting RBCMFs.

FIGURE 5.

FIGURE 5

Open in a new tab

Effects of the tangential flow filtration washing process on diacycle time and red blood cell membrane fragment (RBCMF) properties. (a) Indicates the time required to complete each diacycle. (b) Shows the RBCMF concentration at each diacycle. (c) Summarizes changes in average RBCMF particle diameter as measured by Coulter Counter. Notably, diacycles 1 and 2 have significantly different (p < 0.05) mean RBCMF diameters than diacycles 6–10.

3.4. Effect of freeze–thaw cycle on the zeta potential of RBCMFs indicates their storage stability

Data for zeta potential analysis was reported as the mean ± one standard error. The ζ potential of fresh RBCMFs was −40.8 ± 1.2 mV. Thawing frozen RBCMFs at 4°C lowered the ζ potential of RBCMFs to −43.8 ± 1.5 mV. Thawing frozen RBCMFs at 20°C raised the ζ potential of the RBCMFs to −35.2 ± 1.1 mV, while thawing frozen RBCMFs at 37°C results in a ζ potential of −41.6 ± 2.0 mV. Figure 6 summarizes these changes. However, no significant differences in ζ potential were observed between groups as a result of freeze–thawing. Therefore, subjecting RBCMFs to a single freeze–thaw cycle did not significantly alter the ζ potential when compared with control values, regardless of thaw temperature. This is a promising result, as it suggests that the RBCMF precursors required to form RBCGs may be stored and thawed under conditions that would normally destroy RBCs that is being frozen without any cryoprotectant. This would allow for large batches of precursor RBCMFs to be produced, stored in a freezer, and then thawed out as needed to manufacture RBCGs.

FIGURE 6.

FIGURE 6

Open in a new tab

ζ potential of red blood cell membrane fragments (RBCMFs). The data include both freshly prepared (unfrozen) RBCMFs, as well as the effects after freezing RBCMFs at −80°C and subsequent thawing at various temperatures. ANOVA testing showed no significant differences between thawed RBCMF groups.

These ζ potentials are more negative than reported ranges of −9.3 to −15 mV for fresh RBCs, 34 but are consistent with a previous study that observed that the ζ potential is more negative for RBCMFs than intact RBCs. 35

3.5. TFF of resealed Hb‐RBCGs results in higher Hb encapsulation compared with centrifugation and places an upper bound on an acceptable number of diacycles

RBCGs encapsulating cargo must be washed in order to remove unencapsulated material and is a necessary processing step for many biomedical applications. Ideally the washing method should minimize damage to the resealed RBCGs in order to maximize the yield of resealed RBCGs encapsulating cargo. In this study, Hb was used as the model protein encapsulant, because of the ready availability of assays to measure Hb activity in our lab. Hb was encapsulated inside RBCGs and subsequently washed via TFF or differential centrifugation to remove unencapsulated protein. When normalizing the Hb encapsulation efficiency (i.e., fraction of encapsulated Hb compared with the initial Hb basis) to the volume fraction occupied by the resealed RBCGs themselves in solution (i.e., equivalent to the hematocrit), Hb‐RBCGs‐TFF retained a significantly higher amount of Hb compared with Hb‐RBCGs‐C (p < 0.05) prepared from the same batch of precursor RBCMFs. In prior work, Lu et al. 23 investigated the feasibility of using TFF to wash RBCs. In this current application of TFF for particle washing, Hb‐RBCGs were much smaller in size compared with their precursor RBCs, which may lead to a difference in the centrifugal force being required to successfully pellet out the Hb‐RBCGs. Figure 7 highlights this difference, with Hb‐RBCGs‐C possessing a normalized Hb encapsulation efficiency of 0.87 ± 0.091 versus 1.10 ± 0.051 for Hb‐RBCGs‐TFF (p < 0.05). This normalization was employed to facilitate easy comparison between replicates, to account for batch‐to‐batch differences in the initial concentration of RBCMFs used to manufacture the resealed RBCGs. While Hb is known to reversibly bind to the RBC membrane, we have demonstrated from the studies in Section 3.2 that TFF is sufficient to remove over 99% of the cell‐free Hb from RBCMFs after 10 diacycles. The observed lysis of Hb‐RBCGs and higher loss of Hb present in Hb‐RBCGs‐C compared with Hb‐RBCGs‐TFF suggests that centrifugation is more damaging to Hb‐RBCGs versus TFF. Furthermore, TFF is more amenable to scale up versus centrifugation, which makes TFF more appealing for larger‐scale applications where a higher volume of RBCGs may be needed. While this work employed 10 diacycles for the Hb‐RBCG TFF washing protocol, an increase in the concentration of cell‐free Hb was observed after the seventh diacycle for various replicates. This phenomenon is believed to be similar to the hemolysis encountered by RBCs after prolonged exposure to TFF, and thus the authors recommend that future work with this method reduce the total number of diacycles from 10 to 7 to reduce leakage of encapsulated material from the RBCGs. The resealed Hb‐RBCGs, which are prone to shear‐induced hemolysis, should not be subjected to more diacycles than is necessary.

FIGURE 7.

FIGURE 7

Open in a new tab

The effect of different wash methods (centrifugation vs. tangential flow filtration [TFF]) for purifying Hb‐RBCGs on protein (Hb) encapsulation inside Hb‐RBCGs. Hb‐RBCGs washed via centrifugation had significantly lower Hb encapsulation efficiency compared with Hb‐RBCGs washed via TFF (*p < 0.05). The Hb encapsulation efficiency was normalized by the Hb‐RBCG volume fraction to account for batch‐to‐batch variation in the precursor red blood cell membrane fragments. RBCG, red blood cell ghost.

3.6. Characterization of Hb‐RBCG physical properties

In this study, the particle concentration, diameter, and viscosity of replicate RBCGs were measured and compared against native RBCs, RBCMFs, and Hb. Table 1 summarizes these data, and from it, several important details become apparent. With n = 7 replicates, there was no significant difference observed between the viscosity of RBCs and any of the RBCG samples (p > 0.05). RBCs were significantly larger in diameter compared with RBCMFs (p < 0.05) and Hb‐RBCGs‐TFF (p < 0.05), with a mean diameter of 4.74 ± 0.76 μm. Hb‐RBCGs‐C had a very wide size distribution, with a mean diameter of 8.92 ± 4.03 μm. In contrast, Hb‐RBCGs‐TFF possessed a mean diameter of 3.09 ± 1.57 μm. When measuring the size distribution of Hb‐RBCGs‐C via CC, we observed particle diameters near the upper detection limit of 18 μm for the specific CC aperture used for this analysis, as well as a very wide distribution that exceeded the size of RBCs. This suggested extreme particle aggregation. Furthermore, while no statistically significant difference was observed between the viscosities of the samples, centrifuged RBCGs exhibited a viscosity that was almost double that of the other samples, as shown in Table 1. This, again, raised concerns of RBCG aggregation resulting in larger particles, which could increase the solution viscosity, and could be potentially hazardous to patient health.

TABLE 1.

Biophysical properties of RBCGs.

Sample Concentration (109 particles/mL) Mean diameter (μm) Viscosity (cP)
Hb‐RBCGs‐C 4.37 ± 1.63 8.92 ± 4.03 7.08 ± 1.31
Hb‐RBCGs‐TFF 48.4 ± 12.45 3.09 ± 1.57 4.48 ± 0.37
Cell‐Free Hb Variable 0.005 Concentration‐dependent
RBCs 4.63 ± 0.35 4.74 ± 0.76 3.93 ± 1.77
RBCMFs 2.70 ± 0.176 2.68 ± 0.17 4.53 ± 1.27
Open in a new tab

Note: For viscosity measurements, all samples except Hb had their volume fraction adjusted to 45% to mimic physiological RBC concentrations in the blood.

Abbreviations: RBC, red blood cell; RBCG, red blood cell ghost; RBCMF, red blood cell membrane fragment; TFF, tangential flow filtration.

3.7. Oxygen‐binding equilibria and oxygen offloading kinetics of Hb‐RBCGs

It was important to determine whether the oxygen‐binding equilibria and oxygen offloading kinetics of Hb‐RBCGs were being altered by the Hb‐RBCG washing process, and those properties are summarized in Table 2 as well as Figure 8. Typical data for fresh RBCs and Hb were sourced from Banerjee et al. 31 Neither the P50, cooperativity coefficient (n), nor koff,O2, were significantly different between Hb‐RBCGs‐TFF and Hb‐RBCGs‐C (p < 0.05). Both types of Hb‐RBCGs had a lower P50 than fresh RBCs (p < 0.05), and a higher P50 than cell‐free Hb (p < 0.05). In addition, koff,O2 for both types of Hb‐RBCGs were significantly higher than RBCs (p < 0.05) and lower than cell‐free Hb (p < 0.05). For the Hb‐RBCGs‐TFF, these changes can largely be attributed to the size differences between the Hb‐RBCGs‐TFF and either native RBCs or cell‐free Hb. The path length for oxygen diffusion in RBCs is much longer than Hb‐RBCGs and cell‐free Hb due to their larger diameter. For Hb‐RBCGs‐C, while the mean particle size was larger than RBCs, the wide size distribution of particles present suggest that there were particles smaller than RBCs. The presence of these smaller particles would skew measurements of the oxygen equilibrium and oxygen offloading kinetics toward those for Hb‐RBCGs‐TFF.

TABLE 2.

Oxygen release kinetics and equilibria of Hb‐RBCGs compared with cell‐free Hb and fresh RBCs.

P50 (mm Hg) n koff,O2 (s−1)
Hb‐RBCGs‐C 13.90 ± 0.32 2.37 ± 0.04 13.05 ± 1.75
Hb‐RBCGs‐TFF 14.22 ± 0.19 2.393 ± 0.04 16.68 ± 0.89
Cell‐free Hb 12.55 ± 0.97 2.64 ± 0.09 36.51 ± 2.53
RBCs 27.16 ± 4.45 2.21 ± 0.13 8.61 ± 2.55
Open in a new tab

Note: The first‐order rate constant for oxygen offloading is reported, along with the P50 and cooperativity coefficient regressed from the oxygen equilibrium curve.

Abbreviations: RBC, red blood cell; RBCG, red blood cell ghost; TFF, tangential flow filtration.

FIGURE 8.

FIGURE 8

Open in a new tab

Oxygen binding properties of Hb‐RBCGs compared with cell‐free Hb and fresh red blood cells (RBCs). Oxygen equilibrium curve (a) and oxygen offloading kinetics (b). Representative curves were selected for each sample from a total of n = 6 different preparations for the oxygen equilibrium curves and n = 5 different preparations for the oxygen offloading kinetics. RBCG, red blood cell ghost.

Taken together, our data clearly show the feasibility of encapsulating a model protein (Hb) inside of resealed RBCMFs. Therefore, it should be possible to encapsulate other types of proteins and other types of biomolecules inside of resealed RBCMFs for potential therapeutic or diagnostic applications. It is also worth noting that there are blood group antigens on the surface of RBCMFs, therefore it is critical that for any therapeutic application that the RBCMFs be derived from group O negative RBCs, which can be universally transfused into patients with any blood type.

4. CONCLUSION

We demonstrate that TFF is a scalable process for the production of purified RBCMFs. Potential adverse effects due to the presence of cell‐free Hb should be mitigated by the removal of >99% of the cell‐free Hb present in the RBCMF preparations via TFF. The resealing of TFF‐purified RBCMFs in the presence of the model protein Hb via ionic reversal was shown to be a viable approach for the production of Hb‐RBCGs. Upon formation of Hb‐RBCGs, we demonstrated the ability of TFF remove unencapsulated Hb, and observed that TFF‐washed Hb‐RBCGs possessed a higher protein encapsulation efficiency and much lower degree of aggregation than Hb‐RBCGs washed via centrifugation.

AUTHOR CONTRIBUTIONS

Xiangming Gu: Validation; visualization; methodology; writing – review and editing; project administration; conceptualization. Andre F. Palmer: Conceptualization; validation; supervision; writing – review and editing; funding acquisition; project administration; resources; methodology.

CONFLICT OF INTEREST STATEMENT

A provisional patent application was filed based on the materials described in this work.

ACKNOWLEDGMENTS

The authors thank Shuwei Lu for assisting with the RBC washing process and associated cell‐free Hb quantification assay. The authors would also like to extend their gratitude to the Dr. Jeffrey Chalmers lab for allowing use of their Coulter Counter instrument, and Dr. L. James Lee lab for allowing use of their BI‐200SM goniometer and ZetaPals. This work was supported by National Institute of Health grants R01HL126945, R01HL138116, R01HL156526, R01HL158076, R01HL159862, R01HL162120, and R01EB021926 and Department of Defense grant W81XWH‐18‐1‐0059.

Gu X, Palmer AF. Tangential flow filtration‐facilitated purification of human red blood cell membrane fragments and its preferential use in removing unencapsulated material from resealed red blood cell ghosts compared to centrifugation. Biotechnol. Prog. 2024;40(6):e3501. doi: 10.1002/btpr.3501

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.

REFERENCES

  • 1. Schaer DJ, Alayash AI, Buehler PW. Gating the radical hemoglobin to macrophages: the anti‐inflammatory role of CD163, a scavenger receptor. Antioxid Redox Signal. 2007;9(7):991‐999. [DOI] [PubMed] [Google Scholar]
  • 2. Kuhn DA, Vanhecke D, Michen B, et al. Different endocytotic uptake mechanisms for nanoparticles in epithelial cells and macrophages. Beilstein J Nanotechnol. 2014;5(1):1625‐1636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Roy R, Tripathi A, Das M, Dwivedi PD. Cytotoxicity and uptake of zinc oxide nanoparticles leading to enhanced inflammatory cytokines levels in murine macrophages: comparison with bulk zinc oxide. J Biomed Nanotechnol. 2011;7(1):110‐111. [DOI] [PubMed] [Google Scholar]
  • 4. Feng X, Lv F, Liu L, et al. Conjugated polymer nanoparticles for drug delivery and imaging. ACS Appl Mater Interfaces. 2010;2(8):2429‐2435. [DOI] [PubMed] [Google Scholar]
  • 5. Cheng J, Teply BA, Sherifi I, et al. Formulation of functionalized PLGA‐PEG nanoparticles for in vivo targeted drug delivery. Biomaterials. 2007;28(5):869‐876. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Xia Q, Zhang Y, Li Z, Hou X, Feng N. Red blood cell membrane‐camou fl aged nanoparticles: a novel drug delivery system for antitumor application. Acta Pharm Sin B. 2019;9(4):675‐689. doi: 10.1016/j.apsb.2019.01.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Feng X, Tang Y, Duan X, Liu L, Wang S. Lipid‐modified conjugated polymer nanoparticles for cell imaging and transfection. J Mater Chem. 2010;20(7):1312‐1316. [Google Scholar]
  • 8. Dong F, Hall CH, Golech SA, et al. Immune effects of resuscitation with HBOC‐201, a hemoglobin‐based oxygen carrier, in swine with moderately severe hemorrhagic shock from controlled hemorrhage. Shock. 2006;25(1):50‐55. [DOI] [PubMed] [Google Scholar]
  • 9. Rodriguez PL, Harada T, Christian DA, Pantano DA, Tsai RK, Discher d E. Minimal “self” peptides that inhibit phagocytic clearance and enhance delivery of nanoparticles pia. Science. 2013;339(6122):971‐975. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3624763/pdf/nihms412728.pdf [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. El‐Gogary RI, Rubio N, Wang JTW, et al. Polyethylene glycol conjugated polymeric nanocapsules for targeted delivery of quercetin to folate‐expressing cancer cells in vitro and in vivo. ACS Nano. 2014;8(2):1384‐1401. [DOI] [PubMed] [Google Scholar]
  • 11. Madan J, Pandey RS, Jain V, Katare OP, Chandra R, Katyal A. Poly (ethylene)‐glycol conjugated solid lipid nanoparticles of noscapine improve biological half‐life, brain delivery and efficacy in glioblastoma cells. Nanomedicine. 2013;9(4):492‐503. [DOI] [PubMed] [Google Scholar]
  • 12. Armstrong JK, Hempel G, Koling S, et al. Antibody against poly(ethylene glycol) adversely affects PEG‐asparaginase therapy in acute lymphoblastic leukemia patients. Cancer. 2007;110(1):103‐111. [DOI] [PubMed] [Google Scholar]
  • 13. Tang Y, Tan Y, Lin K, Zhu M. Research progress on polydopamine nanoparticles for tissue engineering. Front Chem. 2021;9:1‐9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Hong S, Na YS, Choi S, Song IT, Kim WY, Lee H. Non‐covalent self‐assembly and covalent polymerization co‐contribute to polydopamine formation. Adv Funct Mater. 2012;22(22):4711‐4717. [Google Scholar]
  • 15. Fang RH, Kroll AV, Gao W, Zhang L. Cell membrane coating nanotechnology. Adv Mater. 2018;30:e1706759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Parker, JC. Stimulation of calcium‐sodium exchange in dog redblood cells by hemolysis and resealing. BBA ‐ Biomembranes. 1988;943(3):463–470. [DOI] [PubMed] [Google Scholar]
  • 17. Hudec R, Lakatoš B, Orlický J, Varečka L. Reconstitution of the basal calcium transport in resealed human red blood cell ghosts. Biochem Biophys Res Commun. 2004;325(4):1172‐1179. [DOI] [PubMed] [Google Scholar]
  • 18. Millner A, Ekin A‐GG. Lipid players of cellular senescence. Metabolites. 2020;10(9):1‐17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Alayash AI. Mechanisms of toxicity and modulation of hemoglobin‐based oxygen carriers. Shock. 2019;52(1S):41‐49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Sheffield C, Spates G, Droleskey R, Green R, DeLoach J. Preparation of lipid‐free human hemoglobin by dialysis and ultrafiltration. Biotechnol Appl Biochem. 1987;9(3):230‐238. doi: 10.1111/j.1470-8744.1987.tb00474.x [DOI] [PubMed] [Google Scholar]
  • 21. Rosenberry TL, Chen JF, Lee MML, Moulton TA, Onigman P. Large scale isolation of human erythrocyte membranes by high volume molecular filtration. J Biochem Biophys Methods. 1981;4(1):39‐48. [DOI] [PubMed] [Google Scholar]
  • 22. Froude VE, Zhu Y. Dielectrophoresis of functionalized lipid unilamellar vesicles (liposomes) with contrasting surface constructs. J Phys Chem B. 2009;113(6):1552‐1558. doi: 10.1021/jp808454w [DOI] [PubMed] [Google Scholar]
  • 23. Lu S, Allyn M, Weigand M, Chalmers JJ, Palmer AF. Tangential flow filtration facilitated washing of human red blood cells: a proof‐of‐concept study. Vox Sang. 2022;117(6):803‐811. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Palmer AF, Sun G, Harris DR. Tangential flow filtration of hemoglobin. Biotechnol Prog. 2009;25(1):189‐199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Elmer J, Harris DR, Sun G, Palmer AF. Purification of hemoglobin by tangential flow filtration with diafiltration. Biotechnol Prog. 2009;25(5):1402‐1410. doi: 10.1002/btpr.217 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Kiss F, Toth E, Miszti‐Blasius K, Nemeth N. The effect of centrifugation at various g force levels on rheological properties of rat, dog, pig and human red blood cells. Clin Hemorheol Microcirc. 2016;62(3):215‐227. [DOI] [PubMed] [Google Scholar]
  • 27. Huang YX, Wu ZJ, Mehrishi J, et al. Human red blood cell aging: correlative changes in surface charge and cell properties. J Cell Mol Med. 2011;15(12):2634‐2642. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Rhyner, M.N . The Coulter Principle for Analysis of Subvisible Particles in Protein Formulations. AAPSJ. 2011;13:54‐58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Winterbourn C.C. Oxidative reaction of hemoglobin. Methods in Enzymology. 1990;186:265‐272. [DOI] [PubMed] [Google Scholar]
  • 30. Cuddington CT, Wolfe SR, Palmer AF. Biophysical properties of tense quaternary state polymerized human hemoglobins bracketed between 500 kDa and 0.2 μm in size. Biotechnol Prog. 2022;38(1):1‐8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Banerjee U, Wolfe S, O'Boyle Q, Cuddington C, Palmer AF. Scalable production and complete biophysical characterization of poly(ethylene glycol) surface conjugated liposome encapsulated hemoglobin (PEG‐LEH). PLoS One. 2022;17(7):1‐24. doi: 10.1371/journal.pone.0269939 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Shvalov AN, Soini JT, Chernyshev AV, Tarasov PA, Soini E, Maltsev VP. Light‐scattering properties of individual erythrocytes. Appl Optics. 1999;38(1):230‐235. [DOI] [PubMed] [Google Scholar]
  • 33. Mesarec L, Góźdź W, Iglič A, Kralj‐Iglič V, Virga EG, Kralj S. Normal red blood cells' shape stabilized by membrane's in‐plane ordering. Sci Rep. 2019;9(1):1‐11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Fontes A, Fernandes HP, de Thomaz AA, Barbosa LC, Barjas‐Castro ML, Cesar CL. Measuring electrical and mechanical properties of red blood cells with double optical tweezers. J Biomed Opt. 2008;13(1):014001. [DOI] [PubMed] [Google Scholar]
  • 35. Luk BT, Jack Hu CM, Fang RH, et al. Interfacial interactions between natural RBC membranes and synthetic polymeric nanoparticles. Nanoscale. 2014;6(5):2730‐2737. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Articles from Biotechnology Progress are provided here courtesy of Wiley

RESOURCES