The effect of treatment with α-glycosidases from Bacteroides fragilis on the survival of rat erythrocytes in the circulation

Abstract

Background

It has been demonstrated recently that α1,3-galactosidase from Bacteroides fragilis can efficiently convert human group B red blood cells (RBC) to group O cells. In addition, in vitro data indicated that the enzymatic conversion process did not affect the physiological or metabolic parameters of the RBC. The aim of this study was to investigate the lifespan of enzyme- treated RBC in vivo in the circulation.

Materials and methods

This was an experimental, randomised study. The rat was selected as the experimental subject because it expresses α-1,3galactosyl on its RBC. The efficiency of Galα1,3Gal epitope removal from RBC treated with α1,3-galactosidase was tested before the transfusion experiment to track the survival of RBC in the circulation. The animals were divided into three groups and injected via the tail vein with native, mock-treated or enzyme-treated RBC labelled with fluorescein isothiocyanate. The survival rates of the fluorescently labelled RBC were monitored by flow cytometry.

Results

Flow cytometry showed that α-galactosidase (0.02 mg/mL for RBC with a haematocrit of 30%) efficiently removed Galα1,3Gal epitopes from rat erythrocytes, although small amounts of remaining Galα1,3Gal epitopes were still detected. The in vivo data demonstrated that the half-life of enzyme-treated RBC was a little shorter than that of native RBC. However, the 24-hour survival fractions of native, mock-treated and enzyme-treated RBC were virtually identical. Most importantly, the enzyme-treated RBC, like the native RBC, were still detectable 35 days after transfusion.

Discussion

Our results indicate that α-glycosidase treatment had little effect on the in vivo survival kinetics of RBC. These data add further support to the feasibility of translating enzymatic conversion technology into clinical practice.

Introduction

The possibility of being able to produce red blood cells (RBC) that could be universally transfused into all recipients (group O RBC) would improve the blood supply while enhancing the safety of clinical transfusions¹. Enzymatic removal of blood group ABO antigens to develop universal RBC was proposed more than 25 years ago². Recently, two families of enzymes that can remove blood group A, B and AB antigens from the surface of RBC were identified by Liu et al³. The two bacterial glycosidases, which are different from other previously reported glycosidases^2,4–6, have unique properties and are so efficient that large scale conversion of blood groups A, B and AB into universal donor group O is feasible. The next phase is to subject the enzymatically converted group O RBC (ECO-RBC) to a full battery of safety tests (e.g. acute toxicity, lifespan) before they can be used in clinical transfusion. However, as there was no suitable animal model for evaluating the acute toxicity of human RBC, data on the safety of ECO-RBC have not yet been reported.

Liu et al. went on to demonstrate that α1,3-galactosidases from Bacteroides fragilis acted equally well on both branched blood group B [Galα1–3(Fucα1–2)Gal] and linear α1,3Gal structures, such as the immunodominant xenotransplantation epitope Galα1–3Galβ1–4GlcNAc (Galα1,3Gal, xenoantigen)^3,7. In addition, the enzymatic conversion process did not affect physiological or metabolic parameters of the ECO-RBC, including osmotic fragility and levels of 2,3-DPG, ATP and methaemoglobin^3,8. So far, however, there is little information on the lifetime of ECO-RBC in the circulation. In this study, rat RBC were chosen as the target cells and the rat as the animal model to investigate the lifespan of circulating ECORBC because of the expression of Galα1,3Gal epitopes on rat RBC. Using flow cytometry with Griffonia simplicifolia isolectin B4 (GS-IB4) as the probe, we first determined the efficiency of Galα1,3Gal epitope removal from rat RBC treated with different concentrations of α1,3-galactosidases. Transfusion experiments were then done to monitor the survival of α1,3-galactosidase-treated RBC in the circulation.

Materials and methods

Animals and materials

Twenty-five healthy Sprague-Dawley rats of clean grade (males, 180 to 200 g) were raised in the Experimental Animal Centre of the Academy of Military Medical Sciences (Beijing, China). α-galactosidase was obtained from Bacteroides fragilis in our laboratory. Fluorescein isothiocyanate (FITC) and fluorescein-labelled GS-IB4 were from purchased Sigma (St. Louis, Missouri, United States of America). Anti-A and anti-B blood grouping reagents were obtained from Changchun Brother Biotech Corporation Ltd. (Changchun, China), and the anti-A,B antibody came from Millipore (Livingston, United Kingdom).

Blood collection

Blood was obtained from the rats by cardiac puncture and then centrifuged. The plasma and buffy coat were discarded and the remaining packed RBC were washed three times with isotonic sterile saline solution and used for ABO blood typing, α-galactosidase treatment, flow cytometric analysis and transfusion experiments.

Blood typing and blood matching tests

Rats may have A agglutinogens⁹. Rat RBC were investigated by anti-A, anti-B and anti-A,B blood grouping reagents. Furthermore, to identify serological compatibility between recipient sera and donor rat RBC, blood matching was tested with direct agglutination studies.

Enzymatic treatment of rat red blood cells with α-galactosidase

The rats were transfused with one of three types of RBC: (i) native RBC, (ii) mock-treated control RBC, or (iii) α1,3-galactosidase-treated RBC. Briefly, rat RBC were divided in three samples of equal volume. The RBC in the native group were kept in isotonic saline at 4 °C until intravenous injection. The enzyme-treated RBC underwent the enzymatic reactions in a conversion buffer (200 mM glycine with 3 mM NaCl, pH 6.8) containing different concentrations of α-galactosidase (0.001, 0.005, 0.01, 0.03, 0.05, 0.1, 0.2 and 0.4 mg/mL of packed RBC with a 30% haematocrit) as described by Liu et al³. The conversion buffer and RBC were incubated together for 60 minutes under gentle rotation at 26 °C and the enzyme-treated RBC were then washed with phosphate-buffered saline (PBS). The mock-treated RBC were subjected to the same enzyme lysis procedure in the absence of α-galactosidases. The cells were tested by flow cytometry analysis using fluorescein-labelled GS-IB4 as the probe.

Flow cytometry analysis of Gal α1,3Gal expression on rat red blood cells

The rats’ native RBC and ECO-RBC were analysed by flow cytometry. The RBC (1×10⁶) were fixed with 2.0% paraformaldehyde for 30 minutes at room temperature to prevent agglutination of antigen-positive cells and were then washed with PBS and incubated with FITC-labelled GS-IB4 (10 μg/mL). The reactions were incubated for 5 minutes at room temperature under gentle agitation and then for 2 hours at 4 °C before being washed three times with cold PBS and resuspended in 500 μL of the same buffer. The percentages of fluorescently labelled RBC were analysed by a flow cytometer (Cytomics FC 500 Beckman Coulter, Brea, United States of America). All incubations were carried out in the dark. Cells incubated without the probe were used as the background control.

Preparation of fluorescein isothiocyanate-labelled rat red blood cells

In order to track the post-transfusion effects of the three groups of RBC in the circulation directly, RBC were labelled with FITC before transfusion. The RBC were incubated in PBS (pH 7.4) containing FITC (at a final concentration of 15 μg/mL) at 25 °C for 1 hour in the dark. The labelled RBC were then washed several times with PBS solution to remove the excess FITC and stored in the dark in a refrigerator at 4 °C until they were transfused (within less than 24 hours). The haematocrit was adjusted to 40% with sterilised saline for transfusion.

Survival of enzymatically treated rat red blood cells in the circulation

The Sprague-Dawley rats were randomly divided into three groups; 1 mL of FITC- labelled native, mock-treated or enzyme-treated RBC (40% haematocrit) was transfused into the recipient rats via injection into the tail vein. At least five rats were used for each group. Blood samples were taken from the recipient rats’ tail vein at different times until the FITC-labelled RBC were cleared from the circulation. The survival rates of the transfused RBC were monitored by analysing the percentages of fluorescently labelled RBC using a flow cytometer. All the experiments were performed in a second grade animal experimental room in the Experimental Animal Centre. The level of fluorescence in rats transfused with native, mock-treated and enzyme-treated RBC was measured 30 minutes after transfusion and set as 100%.

Statistical analysis

All results are shown as mean ± SE. Two-tailed paired and unpaired Student’s t tests were used to determine the statistical significance of differences between groups. P values <0.05 were considered statistically significant.

Results

Blood typing and compatibility testing

Blood typing tests showed that the rat RBC did not react with anti-A, anti-B or anti-A,B reagents (Figure 1), which indicates that there was no A agglutinogens on the RBC from these Sprague-Dawley rats. Furthermore, all donor rat RBC were compatible with recipient rats’ sera by direct agglutination (Figure 1). The transfusion experiments could, therefore, be done between the different individual rats.

Figure 1.

Testing for ABO blood group on rat RBC and compatibility testing between different rats.

Blood typing tests indicated that there were no A agglutinogens on Sprague-Dawley rat RBC. Donor rat RBC were compatible with recipient rats’ sera by direct agglutination.

Enzymatic removal of Galα1,3Gal epitopes from rat red blood cells

Native and enzyme-treated RBC were subjected to flow cytometric analysis to test the clearance rate of Galα1,3Gal epitopes from rat RBC with different concentrations of α-galactosidase using IB4 lectin to monitor the conversion (Figure 2). The results showed that Galα1,3Gal antigens could be efficiently removed from rat RBC when the concentration of α-galactosidase was 0.02 mg/mL, although some residual reactivity was still found (7.8±0.38%). Higher concentrations of enzyme did not produce greater removal of Galα1,3Gal antigens. A concentration of 0.02 mg/mL α-galactosidase was, therefore, used to treat the rat RBC for the transfusion experiment.

Figure 2.

Flow cytometric analysis of Galα1,3Gal epitopes on native rat RBC and rat RBC treated with different concentrations of α-galactosidase.

The results indicated that conversion was best with 0.02 mg α-galactosidase in a 1.0 mL reaction system (RBC with a 30% haematocrit) and that more enzyme did not improve the removal rate.

The flow cytometry overlay profile showed that almost no fluorescence signal was detected in enzyme-treated (0.02 mg/mL α-galactosidase) RBC compared with in native and mock-treated rat RBC (Figure 3): the mean fluorescence intensity (MFI) of enzyme-treated RBC decreased by about 92.2±0.38% compared to that of native and mock-treated RBC, indicating that most Galα1,3Gal epitopes on rat RBC were removed by α-galactosidase.

Figure 3.

Flow cytometric overlay profile of native and enzyme-treated rat RBC.

The FACS histograms show the relative Galα1,3Gal epitope densities as measured by IB4 lectin on native RBC (solid line), mock-treated RBC (dotted line) and enzyme-converted RBC (dashed line) RBC from rats.

In vivo survival of α-galactosidase-treated rat red blood cells

The survival fraction of RBC was detected by FITC. The in vivo data demonstrated that the half-life (t_1/2; 50% survival of RBC in circulation) of enzyme-treated RBC was a little shorter than that of native and mock-treated RBC (about 10.8 vs 14.2 and 15.3 days, respectively) (Figure 4). However, the 24-hour survival fractions of native, mock-treated and enzyme-treated RBC were virtually identical (all about 82%). Normally, RBC are cleared from the circulation approximately 40 to 50 days after allotransfusion¹⁰. Most importantly, the α-glycosidase-treated RBC were still detectable 35 days after transfusion (at which time the survival fraction was about 1.2%). These results indicate that α-glycosidase treatment had little effect on the normal survival kinetics of RBC.

Figure 4.

In vivo survival of native RBC (solid line with diamonds), mock-treated RBC (dotted line with filled circles) and a-glycosidase-treated RBC (dashed line with filled triangles) in recipient rats at different times after transfusion.

The fluorescently-labelled native and enzyme-treated RBC were still detectable 35 days after transfusion (the survival fractions were about 1.7% and 1.2%, respectively). Each point in the figure represents the mean value of five samples.

Discussion

Liu et al. previously demonstrated that α-galactosidase from Bacteroides fragilis was more efficient in converting human blood group B RBC to group O RBC than was α-galactosidase from Catimor coffee beans^3,7. Moreover, they also demonstrated that both the membrane structure and metabolic function of ECO-RBC were identical to those of normal human RBC with regards to haemolysis, osmotic fragility, levels of acetylcholinesterase, cholesterol, ATP and 2,3-DPG, oxygen affinity (P50), methaemoglobin, and deformability³. Work in our laboratory was in accordance with Liu’s results⁸. Although many in vitro results indicated that removal of the α1,3-galactosyl group did not affect the function of RBC, there was no in vivo evidence.

Gutowski et al. reported that enzymatic removal of sialyl groups from mammalian erythrocytes resulted in rapid sequestration of the cells from the circulation following autologous transfusion^11,12. The terminal β1,3-galactosyl group exposed on asialo-RBC is recognised by an autoimmune anti-galactosyl IgG and/or by a lectin-like receptor on monocytes and macrophages. Accordingly, treatment with β-D-galactosidase would enhance erythrocyte survival in circulation. Thus, a change in the chemical structure of the RBC membrane may also change the cell’s physical properties. One may, therefore, wonder whether treatment with α1,3-galactosidase from Bacteroides fragilis would affect the life-span of RBC.

We previously performed transfusion experiments using Rhesus monkeys and gibbons as the experimental subjects^6,13. Briefly, group B RBC from gibbons were subjected to treatment with α-galactosidase (from Catimor coffee beans) and then transfused into group A and AB gibbons. The results showed that transfusion of the ECO-RBC was safe. In addition, the 24-hour survival fraction of donor RBC in the recipient monkeys and gibbons was more than 95% and the t_1/2 was 16–17 days, which falls within the normal range.

In this study, we selected rats as the transfusion subjects, because they are easy and simple to work with. Furthermore, and importantly, there are Galα1,3Gal epitopes on rat RBC, whereas the RBC from many other animals lack Galα1,3Gal epitopes (e.g., goat, sheep, horse, chicken and mouse¹⁴.) Interestingly, efficient removal of the xenoantigens from rat RBC does not require a higher concentration of enzyme (0.02 mg/mL RBC) than that previously used for efficient removal of blood group B from human RBC (0.05 mg/mL RBC). Although α-glycosidase could not completely remove Galα1,3Gal from RBC even when a high concentration of the enzyme was used (the residual rate of Galα1,3Gal was about 8.5%), the enzyme could remove B antigen completely from human RBC. This discrepancy could reflect differences in structures of carbohydrates and glycoconjugates. Another possible explanation for the discrepancy could be different accessibility of the enzyme to substrates on the cell surface. In addition, the buffer system, developed for efficient conversion of human group B RBC and optimised for ionic interactions between the enzyme and human cells, may not be the optimal condition for rat cells.

We found that the t_1/2 of enzyme-treated RBC was about 10.8 days, which is a little shorter than that of untreated RBC (14.1 days). Importantly, however, the α-glycosidase-treated RBC were still detectable 35 days after transfusion (at which time, the survival fraction was about 1.2%). These results indicate that α-glycosidase treatment had little effect on the normal survival kinetics of RBC. These data add further support to the feasibility of translating enzymatic conversion technology into clinical practice.