- Email: [email protected]
In vivo circulation of mouse red blood cells frozen in the presence of dextran and glucose
Cryobiology 61 (2010) 10–16
Contents lists available at ScienceDirect
Cryobiology journal homepage: www.elsevier.com/locate/ycryo
In vivo circulation of mouse red blood cells frozen in the presence of dextran and glucose q Guo Bo Quan, Min Xia Liu, Ying Han *, Lei Fang, Wei Du, Jie Xi Wang Beijing Institute of Transfusion Medicine, the Taiping Road 27#, the Haidian County, Beijing 100850, China
a r t i c l e
i n f o
Article history: Received 9 September 2009 Accepted 16 February 2010 Available online 20 February 2010 Keywords: Mouse red blood cells Glucose Freeze Dextran Life span The 24 h recovery
a b s t r a c t Loading with monosaccharide can improve the quality of human red blood cells (hRBCs) frozen with polymer. But in vivo life span of hRBCs frozen with polymer and sugar is not determined. In this study, following incubation with glucose, mouse red blood cells (mRBCs) were frozen in liquid nitrogen for 24 h using dextran as the extracellular protectant. After thawing, hemolysis, exposure of PS, and osmotic fragility of frozen mRBCs were determined in vitro. After transfusion of fluorescein isothiocyanate (FITC)labeled mRBCs, the 24 h recovery and half life span of frozen mRBCs were determined. The data indicated the postthaw hemolysis of mRBCs frozen with dextran and glucose were significantly less than that of cells frozen with dextran (17.23% ± 5.21% vs 25.96% ± 10.07%, P = 0.034). But freezing can also result in exposure of phosphatidylserine and increase of osmotic fragility of mRBCs. After transfusion, the 24 h recovery of mRBCs frozen in the absence or presence of glucose was similar to that of the control cells (P = 0.748 and 0.971). However, the half life span of mRBCs frozen in the absence or presence of glucose was significantly less than that of the control cells (P = 0.000). In addition, incubation with glucose can not increase the life span of frozen red blood cells (7.16 ± 0.93 d vs 7.15 ± 0.34 d, P = 0.982). In conclusion, incubation with monosaccharide could significantly increase the recovery of mRBCs frozen with polymer. Although freezing can significantly shorten the half life span of frozen cells, it can not influence the 24 h recovery of frozen mRBCs. In addition, incubation with monosaccharide before freezing can not increase the life span of frozen mRBCs. So according to the above data, to increase the life span of hRBCs frozen with polymer and monosaccharide, the osmotic fragility of the frozen RBCs must be decreased in the future. Ó 2010 Elsevier Inc. All rights reserved.
Introduction At present, freezing with glycerol can greatly prolong the storage time of RBCs in vitro [1]. However, freezing with glycerol has some limits which may affect activeness of people to use frozen RBCs in clinical therapy. The thawed RBCs require complicated washing procedure to remove glycerol. Furthermore, the addition and removal of glycerol can lead to transient osmotic injury and a significant loss of cells [35]. Some polymers, such as hydroxyethyl starch (HES), polyvinyl pyrrolidone (PVP), and dextran, have been used as alternatives to glycerol [14,18,23,26,28,33,37]. But freezing with polymers also faces a great challenge. Polymers can’t penetrate the membrane of red blood cells, so their capability to protect the inner membrane may be discounted. Recently, the super protection of treha-
q Statement of funding: This work was supported by the National Natural Science Foundation of China (Grant No. N30400518) and Beijing Natural Science Foundation of China (Grant No. 7072059). * Corresponding author. E-mail address: [email protected] (Y. Han).
0011-2240/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.cryobiol.2010.02.004
lose on frozen or dry cells has received attentions of many investigators [2,4,5,7,8,10,19,21,23,26,38]. But in order to obtain the optimal protective effect, trehalose must be loaded into the cytoplasm [4,38]. At present, some methods, including genetic technology [7,10], microinjection [8], fluid phase endocytosis [22,38], and so on may not be suitable for loading trehalose into RBCs. Although Satpathy et al. found osmotic imbalance and membrane lipid phase transition could introduce trehalose into RBCs [30], their method can also result in serious hemolysis [16,24,25] and oxidative stress [16] which may directly cause more cell injuries during subsequent freezing and thawing [24,26]. In the previous study, compared with trehalose, incubation with monosaccharide before freezing can efficiently increase the recovery of hRBCs frozen with polymer [24,26]. According to ‘‘the water replacement hypothesis” raised by Crowe et al., monosaccharide may also stabilize the membrane of frozen or dried RBCs through forming hydrogen bonds with phospholipids and proteins like trehalose [6]. In addition, accumulation of intracellular glucose can increase the vitrification degree of cytoplasm and decrease formation of lethally intracellular ice. Most importantly, human red blood cells can utilize glucose through glycolysis, but can not
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
metabolize trehalose owing to lack of trehalases which are required for enzymatic hydrolysis of trehalose [5,30]. So glucose may stabilize the metabolic function of frozen red blood cells [26]. In order to further evaluate the efficacy of the present freezing method, this study is to determine the in vivo 24 h recovery and half life span of mRBCs frozen with dextran and glucose. The Kunming (KM) white mouse was used as the experimental animal. FITC was used to label mRBCs. The final purpose is to obtain the data about the in vivo survival of frozen mRBCs and improve the present freezing method in the absence of glycerol. Materials and methods Reagents and solutions Unless otherwise stated, all chemicals were analytical reagent grade. Dextran (MW40,000 Da) and FITC were purchased from Sigma Chemical Co. (St. Louis, MO, USA). All solutions were prepared in double glass-distilled water. The modified phosphate buffered saline (mPBS) containing 154 mM NaCl, 1.06 mM KH2PO4, 5.6 mM Na2HPO4 and 2.0 mM adenine (300 mOsm, pH 7.4) was used as the basic buffer. The mPBS containing 40% dextran (W/V) was used as the extracellular protective solution. Preparation of mRBCs Female KM mice weighted 25–30 g were purchased from Academy of Military Medical Sciences, Beijing, China. Approximately 1 ml of blood was collected by cardiac puncture from mice anaesthetized with 0.2% pentobarbital. The citrate–phosphate-dextrose medium was used as the anticoagulant. The whole blood was not leukoreduced prior to processing. The blood was centrifuged at 1550g for 2 min at 4 °C to remove the supernatant, including white blood cells and platelets. Then the concentrated mRBCs were washed three times using mPBS. The final concentrated mRBCs were used in the study. A spun hematocrit method was used to determine the hematocrit. The hematocrits used in sugar loading and freezing were approximately 22% and 51%, respectively. The process of loading glucose into red blood cells has been described in the previous studies [24–26]. Briefly, the washed mRBCs were incubated in the mPBS buffers containing glucose for 3 h at 37 °C. After incubation, the cell suspension was centrifuged at 1550g for 2 min to remove the supernatant. Then the concentrated cells were washed using mPBS for three times to fully remove the extracellular glucose. The intracellular glucose concentration was measured by the anthrone method [24,30]. Since the anthrone method detects all sugars present in mRBCs, the control cells were treated in parallel to determine the concentration of endogenous glucose. Finally, the concentration of the endogenous glucose in the mRBCs was subtracted from the total sugar concentration. Freezing and thawing The freezing procedure was described in the previous study [24]. Briefly, the concentrated cells loaded with glucose and 40% dextran were mixed in a ratio of 1:1 (V/V). Then the cell suspension was divided into the standard 2.0 ml cryogenic tubes (Axygen Scientific, California, USA) and frozen in liquid nitrogen at approximately 2.5 °C/s. Every tube contained 1 ml of the cell suspension. The cooling rate was measured with a thermocouple (EastSun Electronic, Zhejiang, China) placed in the middle of the sample. Measurement of the cooling rate was made between 25 and 50 °C, which represents the linear changing range of the cooling rate. The frozen mRBCs were stored in the liquid nitrogen for 24 h and then thawed for 5 min in the 37 °C water bath.
11
In vitro evaluation of quality of frozen mRBCs In order to evaluate the quality of frozen mRBCs, hemolysis, distribution of phosphatidylserine, and osmotic fragility were measured in vitro. The hemolysis was determined using the Drabkin’s solution (Sigma Diagnostics, St. Louis, MO, USA). The thawed cell suspensions were initially mixed with equal volume of mPBS to decrease the viscosity. Then the cell suspension was used to measure the total hemoglobin concentration. The supernatant was used to measure the free hemoglobin concentration following centrifugation. Hemoglobin was transformed to cyanmethemoglobin and the absorbance of cyanmethemoglobin was measured at 540 nm using an ultraviolet spectrophotometer (PGENERAL, Beijing, China). The isotonic mPBS was used as the blank reference. The percent hemolysis was calculated using the following formula:
ðOD540 nm of the supernatant=OD540 nm of the total hemoglobinÞ ð100% Hct%Þ In addition, the hemolysis caused by glucose loading process before freezing was also evaluated. After loading with glucose, the OD values of the total hemoglobin and the supernatant were measured by the above method. The percent hemolysis was calculated using the above formula. The annexin-V-Fluos kit purchased from Roche Co. (Nonnenwald2, 82372 Penzberg, Germany) was used to label mRBCs with exposed phosphatidylserine. Briefly, the approximate 1 106 mRBCs were directly resuspended in the 200 ll of annexin binding buffer. Then the 0.2 lg of annexin-V-FITC was added to the above cell suspension and incubated for 15 min in dark environment at 25 °C. Before flow cytometry, 300 ll of binding buffer was added to the above cell suspension. Finally, the data about cells with exposed PS were analyzed by the flow cytometer (Becton–Dickinson, San Jose, CA, USA). In order to evaluate osmotic tolerance, the 0.9% NaCl diluted with the distilled water was used as the osmotic fragility solutions containing 0–0.9% NaCl. The concentration gradient of NaCl was 0.09%. After washed using mPBS for five times to remove the cryoprotectants and the free hemoglobin fully, the concentrated mRBCs were suspended in every osmotic fragility solution for 2 h at room temperature. Following centrifugation at 1550g for 5 min, the percent hemolysis was determined using the Drabkin’s solution. Determination of the 24h recovery and half life span In this study, FITC was used to label mRBCs. The procedure has been described previously [20,27]. Blood was collected from eighteen mice. The freezing study was divided into two groups according to use of glucose. Following freezing and thawing, the mRBCs were washed using mPBS for five times to fully remove the free hemoglobin and the extracellular protectants. The concentrated mRBCs were diluted with mPBS in a ratio of 1:1. Then the cell suspension was incubated with FITC for 30 min in the 37 °C water bath. The final concentration of FITC in the cell suspension was 70 lg/ml. After incubation, mRBCs were washed for three times to fully remove the extracellular FITC. The final concentrated mRBCs were suspended in mPBS. Approximately 1 1012 red blood cells labeled with FITC in 300 ll of mPBS were injected into a normovolemic mouse (allogeneic transfusion) through tail veins. In every freezing group, six mice were used as the recipients. Following transfusion for different time, 2.5 ll of blood was collected from mouse tail veins using the citrate–phosphate-dextrose medium as the anticoagulant. The blood sample was diluted using 0.5 ml mPBS. Then the cell samples were analyzed using a FACScan
12
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
flow cytometer (Becton–Dickinson, San Jose, CA, USA) equipped with an argon laser, 488 nm wave length. The percentage of FITC-labeled cells and the GEO mean were determined using FACScan software. The GEO mean can reflect the fluorescence intensity of mRBCs. The percentages of FITC-labeled mRBCs after in vivo circulation for 2 h, 1 d, 3 d, 6 d, 7 d, and 9 d were labeled as A0, A1, A3, A6, A7, and A9 respectively. The 24 h recovery was calculated using the following formulae: A1/A0 100(%). In order to calculate the half life span, the recovery of FITC-labeled cells was used as the Y-axis. The time was used as the X-axis. The Microcal (TM) Origin software (Microcal Software Inc., One Roundhouse Plaza, Northampton, MA 01060, USA) was used for regression analysis of the data and building of regression equation. When the recovery of FITC-labeled mRBCs was 50%, the corresponding time was the half life span of mRBCs. Statistical Analysis Each set of experimental conditions was replicated different times. The data were analyzed using independent-samples T test and one-way ANOVA with calculation of the least significant difference (LSD). All results with a value of P < 0.05 or P < 0.01 were considered to be statistically significant. The data were expressed as mean ± SD. Results Incubation with glucose increases the recovery of mRBCs frozen with dextran The loading efficiency of glucose in mRBCs was similar to that of hRBCs [24,25]. With increase of the extracellular concentrations of glucose, the loading efficiency of glucose in mRBCs was also increased steadily. When the extracellular concentration of glucose was increased from 200 mM to 400 mM, the intracellular concentration of glucose was increased from 3.47 ± 1.25 mM to 7.98 ± 3.21 mM steadily. Although high extracellular concentration of glucose can increase the loading efficiency of glucose, this procedure can result in hemolysis of mRBCs. The data were shown in the Fig. 1. The data was collected six mice. After washing, the concentrated cells were incubated with 200 mM or 400 mM glucose for 3 h. The Fig. 1 indicated with increase of the concentrations of glucose, the percent hemolysis of mRBCs also increased steadily. When the concentration of glucose was 200 mM or 400 mM, the percent hemolysis was 3.12% ± 1.19% and 4.97% ± 0.92%, respectively which were significantly more than that of the control cells (0.34% ± 0.21%, P = 0.000). But incubation with glucose can significantly decrease the postthaw hemolysis of mRBCs frozen with dextran. Moreover, with the increase of the concentrations of glucose, the postthaw hemolysis was firstly decreased and then increased. When the concentration of glucose was 200 mM, the postthaw percent hemolysis of mRBCs was 17.23% ± 5.21% and significantly less than that of cells frozen in the absence of glucose (25.96% ± 10.07%, P = 0.034). However, when the concentration of glucose was increased to 400 mM, the postthaw percent hemolysis was 42.57% ± 11.29%.
Fig. 1. The hemolysis of mRBCs caused by incubation with glucose and freezing. The procedures of glucose loading and freezing were described in the part of materials and methods. The percent hemolysis was determined using the Drabkin’s method. In order to evaluate the effect of glucose loading on hemolysis of mRBCs, the data were collected from six mice. The data points are expressed as means ± SD. ‘‘” P < 0.01 vs red blood cells loaded with glucose before freezing. In order to evaluate the effect of freezing on hemolysis of mRBCs, blood was collected from four mice in every study. Then in the presence or absence of glucose, the mRBCs were frozen with polymer. This whole study was repeated six times. ‘‘DD” P < 0.01 vs frozen mRBCs following incubation with 200 mM glucose or with no glucose. ‘‘D” vs frozen mRBCs following incubation with 200 mM glucose. The data were analyzed using one-way ANOVA with calculation of the least significant difference (LSD).
Fig. 2. The effects of glucose loading and freezing on PS distribution of frozen mRBCs. The Annexin V was used to label exposed PS. This study was repeated five times. In every study, blood was collected from three mice. Then in the presence or absence of glucose, the mRBCs were frozen with polymer. The figure represented the data of one study. The ‘‘M1” field represented the cells with exposed PS. ‘‘Con” represented the control cells. ‘‘Loaded with glu” represented cells loaded with glucose. ‘‘no sugar” represented mRBCs frozen with no sugar. ‘‘glu” represented mRBCs frozen with dextran and glucose.
Freezing influences distribution of phosphatidylserine of mRBCs The effect of freezing on the postthaw distribution of PS of mRBCs was shown in the Fig. 2. This study was repeated five times. In every study, blood was collected from three mice. Then in the presence or absence of glucose, the mRBCs were frozen with polymer. The Fig. 2 represented the data from one study. After incubation with glucose, the percentage of cells with exposed PS was
2.13% ± 0.96% and similar to that of the control cells (1.87% ± 1.02%, P = 0.458). However, the postthaw percentage of cells with exposed PS was increased to 12.96% ± 7.23% and significantly higher than that of the control cells (P = 0.000). In addition, similar to hRBCs, incubation with glucose can not decrease the exposure of phoshatidylserine of mRBCs frozen with polymer.
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
13
Fig. 4. Flow cytometry analysis of mRBCs labeled with FITC. (A) represented mRBCs without labeling with FITC; (B) represented mRBCs labeled with FITC. The final concentration of FITC in the cell suspension was 70 lg/ml.
Fig. 3. The effect of freezing on osmotic fragility of mRBCs. The washed cells were suspended in the dilutions of 0.9%NaCl and the hemolysis was measured. The study was repeated six times. In every study, blood was collected from three mice. Then in the presence or absence of glucose, the mRBCs were frozen with polymer. The data points are expressed as means ± SD. ‘‘con, no sugar, and glu” represented the control cells, mRBCs frozen with no sugar, and mRBCs frozen with glucose. ‘‘” P < 0.01 vs the frozen cells. ‘‘DD” P < 0.01 vs the cells frozen with dextran and glucose. The data were analyzed using one-way ANOVA with calculation of the least significant difference (LSD).
The effect of freezing on the osmotolerance of mRBCs The osmotic fragility of frozen mRBCs was more than that of the control cells (Fig. 3). The study was repeated six times. In every study, blood was collected from three mice. Then the mRBCs were frozen with polymer in the presence or absence of glucose. With increase of the concentrations of NaCl, the hemolysis decreased steadily. When the concentrations of NaCl were less than 0.54%, the percent hemolysis of frozen cells was not more than that of the control cells. However, when the concentrations of NaCl were more than 0.54%, the percent hemolysis of frozen mRBCs was significantly higher than that of the control cells (P < 0.01). In addi-
tion, this figure showed when the concentration of NaCl was increased to 0.72%, incubation with glucose could significantly decrease the percent hemolysis of mRBCs frozen with dextran.
In vivo circulation of frozen mRBCs after transfusion In this study, FITC was used to label mRBCs. The percentage of mRBCs labeled with FITC was more than 99% in vitro (seen from the Fig. 4), which showed most of mRBCs could efficiently bind with FITC. Then the fluorescent intensity of FITC-labeled red blood cells in vivo was measured. The data were shown in the Fig. 5A. Blood was collected from four mice. The washed mRBCs labeled with FITC were transfused into six recipients. The GEO mean of red blood cells after circulation for 2 h (day 0) was 68.35 ± 0.99. After circulation for 24 h (day 1), the GEO mean of mRBCs was decreased to 54.31 ± 1.83 and significantly less than that of day 0 (P = 0.007). Subsequently, the GEO means maintained constant and not significantly different from that of day 1 (P > 0.05). In addition, our study showed the autofluorescent intensity of mRBCs without labeling with FITC was approximately 4–5 and significantly less than that of cells labeled with FITC. So the adverse effect of autofluorescent intensity of mRBCs may be obviated in this study.
Fig. 5. The fluorescent intensity and in vivo recovery of FITC-labeled mRBCs after transfusion. After transfusion, the flow cytometry was used to measure the fluorescent intensity and recovery of mRBCs at different time points.(A) The GEO mean represented the fluorescent intensity. Blood was collected from four mice. The washed mRBCs labeled with FITC were transfused into six recipients. The data points were expressed as means ± SD. ‘‘” P < 0.01 vs the other time points. The data were analyzed using oneway ANOVA with calculation of the least significant difference (LSD). (B) The measuring procedures of the recovery of transfused mRBCs were described in the part of materials and methods. Briefly, the percentages of FITC-labeled mRBCs after in vivo circulation for 0, 1 d, 3 d, 6 d, 7 d, and 9 d were labeled as A0, A1, A3, A6, A7, and A9 respectively. The recovery rate was calculated using the following formulae: A1/A0 100(%).The data were collected from 18 mice equally divided into three groups (the group con, no sugar, and glu). So each group contained six mice. The data points were expressed as means ± SD. ‘‘” vs the in vivo recovery of frozen mRBCs at certain time points. The data were analyzed using the independent-samples T test.
14
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
The recovery of mRBCs labeled with FITC after transfusion was shown in the Figs. 5B and 6. The data were collected from 18 mice equally divided into three groups. Each group contained six mice. The Fig. 6 represented the data of one mouse randomly selected from the three groups. Seen from Fig. 5B, with the prolonging of time, the recovery of mRBCs labeled with FITC decreased steadily. However, the in vivo recovery of frozen mRBCs was significantly less than that of the control mRBCs after day 1. In the Table 1, the 24 h recovery of the control cells was 83.66% ± 3.01%. However, the 24 h recoveries of mRBCs frozen in the absence or presence of glucose were 82.62% ± 8.47% and 85.32% ± 3.88% respectively. The 24 h recovery of frozen mRBCs was not significantly less than that of the control cells (P = 0.748 and 0.971). But the life span of frozen mRBCs was significantly less than that of the control cells (P = 0.000). The half life span of the control cells was 10.39 ± 0.55 d. However, the half life span of red blood cells frozen in the absence or presence of glucose was decreased to approximately 7 d, which showed freezing had injured mRBCs. In addition, the half life span of the cells frozen with glucose was similar to that of the cells frozen with no glucose (7.16 ± 0.93 d vs 7.15 ± 0.34 d, P = 0.982). Discussion Polymers can replace glycerol to freeze hRBCs. Some polymers such as albumin, dextran, modified gelatine, polyvinylpyrrolidone,
Table 1 The 24 h recovery and half life span of frozen mRBCs in vivo. Groups
n
The 24 h recovery (%)
Con No sugar Glu
6 6 6
83.66 ± 3.01 82.62 ± 8.47 85.32 ± 3.88
The p value
0.748 0.971
Half life span(d) 10.39 ± 0.55** 7.16 ± 0.93 7.15 ± 0.34
The p value
0.000 0.000
The measuring procedures of the 24 h recovery and half life span of mouse red blood cells were described in the part of materials and methods. Briefly, the percentages of FITC-labeled mRBCs after in vivo circulation for 2 h, 1 d, 3 d, 6 d, and 9 d were labeled as A0, A1, A3, A6, and A9 respectively. The 24 h recovery was calculated using the following formulae: A1/A0 100(%). In order to calculate the half life span, the recovery of FITC-labeled cells was used as the Y-axis. The time was used as the X-axis. The Microcal (TM) Origin software was used for regression analysis and building of regression equation. With a recovery of FITC-labeled mRBCs was 50%, the corresponding time was the half life span of mRBCs. The data were collected from 18 mice equally divided into three groups (the group con, no sugar, and glu). So each group contained six mice. ‘‘con, no sugar, and glu” represented the control cells, mRBCs frozen with no sugar, and mRBCs frozen with glucose. The data were analyzed using independent-samples T test. ** P < 0.01 vs the frozen mRBCs.
polyethylene oxide, polyethylene glycol, and HES can not enter the cells. This property facilitates their removal after thawing. In the case of emergencies, this step could be simplified or omitted, if the additives, for example, albumin, dextran, and HES, are biodegradable and tolerated by the patient [9]. But this method can also cause in vivo hemolysis after transfusion [29,31]. The report of
Fig. 6. The recovery of FITC-labeled mRBCs after freezing and transfusion. The procedures of freezing and labeling with FITC were described in the part of materials and methods. The data were collected from 18 mice equally divided into three groups (the group C, N, and S). So each group contained six mice. This figure represented the data of one mouse randomly selected from the three groups. ‘‘C, N, and G” represented the control cells, mRBCs frozen with no sugar, and mRBCs frozen with glucose.
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
Horn et al. showed after transfusion of hRBCs frozen with HES, the plasma hemoglobin levels increased twofold to threefold, although the plasma hemoglobin could decrease to baseline levels within 24 h [14]. In order to resolve this problem, the protective effect of monosaccharides including glucose, fructose, and galactose on red blood cells frozen with polymer was evaluated in the previous study [24,26]. Loading with monosaccharide, especially galactose, could significantly increase the recovery of hRBCs frozen with dextran. However, the effect of sugar on in vivo circulation of frozen hRBCs is not known. The in vivo 24 h recovery and half life span are important index to evaluate the quality of preserved hRBCs. The minimal recovery of transfused RBCs set by the American Association of Blood Bank after circulation for 24 h is 75% [1]. It is well known that the storage procedures, including the hypothermic storage and freezing or cryopreservation, can result in decreasing recovery of in vivo transfused hRBCs [31]. After storage at 4 °C, approximately 30% and 70% of transfused RBCs disappeared from the circulation after 1 and 3 d, respectively [31]. But the reason is still unknown. Determination of RBC life span is currently performed by random labeling methods using one of the radioactive tracers: 51Cr, DF32P, or DFP-H3 (diisopropylphosphorofluoridate) [12,32]. All of these tracers have disadvantages which make them less than ideal [32]. So some investigators use non-radioactive tracers to replace the radioactive tracers, including fluorescein [3,20,27,34,36], biotin [12,15], and fluorescent lipophilic probes (PKH-3) [32]. FITC is a fluorescein with two hydrophilic hydroxyl groups and can covalently bind to the e-amino nitrogen of lysine, so FITC can stably label most of proteins [27]. Some studies used FITC to label RBCs of mouse, rat, dog, rabbit, or human [11,20,27,34,36]. Butcher et al. found FITC could stably label lymphocyte and not shift to the other cells [3]. In this study, the fluorescent intensity of FITC-labeled mRBCs remained constant from day 1 to day 9. However, autofluorescence of RBCs may influence the results of this study. The formation of autofluorescence of RBCs has been known to accompany the lipid peroxidation process. The interaction of malonyldialdehyde and the amino groups of phospholipids or proteins during lipid peroxidation can result in the formation of fluorescent chromolipids. In addition, spontaneous glycation of glycated proteins and carbohydrates may also generate autofluorescence in RBCs [17]. So the effect of autofluorescence of RBCs must be considered. In this study, the autofluorescent intensity (the GEO mean) of mRBCs was approximately 4–5. On the contrary, the fluorescent intensity of mRBCs labeled with FITC was increased to approximately 200–400. Following transfusion, the fluorescent intensity of frozen mRBCs decreased to approximately 50 and significantly higher than autofluorescent intensity of mRBCs. So the antofluorescence of mRBCs can not influence the final results in this study. Here, the KM white mouse was used as the experimental animal. The murine model has been used to study the recovery of blood cells in vivo [12,13,20,34]. Hoffmeister et al. used the murine model to study the clearance mechanism of chilled blood platelets [13]. In this study, the selection of the animal model must comply with the following standard: like hRBCs, with the increase of the concentrations of monsaccharide, the postthaw hemolysis of frozen mRBCs was firstly decreased and then increased [24]. This study showed the changing mode of postthaw hemolysis of mRBCs was similar to that of hRBCs. Moreover, the in vitro data showed loading with 200 mM glucose could significantly decrease the percent hemolysis of mRBCs frozen with dextran. But when the concentration of glucose was increased to 400 mM, the postthaw hemolysis became more serious. So it may be suitable to select KM mouse as the experimental animal. In addition, freezing can also result in exposure of phosphatidylserine and increasing osmotic fragility of mRBCs. These results were similar to hRBCs in our previous report [24]. Moreover, incubation with glucose can not
15
mitigate the malignant effect of freezing on distribution of phosphatidylserine and osmotolerance of mRBCs which may influence the life span of mRBCs in vivo. The in vivo study showed the 24 h recovery of frozen mRBCs after transfusion was more than 80% and similar to that of the control cells, which showed freezing with polymer and glucose could not influence the 24 h recovery of frozen cells. However, freezing can seriously influence the life span of mRBCs. In this study, the half life span of the control mRBCs was approximately 10 d. According to the report of Li et al., the half life span of KM mRBCs was approximately 12 d [20]. So our data was similar to the report of Li et al. But freezing can reduce the half life span of mRBCs. After transfusion, the half life span of frozen cells was approximately 7 d and significantly less than that of the control cells. Furthermore, loading with glucose can not increase the life span of mRBCs frozen with polymer. So the injury from freezing may be the main factor to decrease life span of mRBCs. In addition, this study also indicated the 24 h recovery could not reflect the real state of frozen RBCs in vivo. The life span must be also considered. Transfusion is an important approach to save the life of patients. But with regard to those patients having specific diseases such as disorder of hematopoiesis, it is beneficial to prolong the in vivo life span of RBCs. In conclusion, freezing with dextran and glucose can not influence the 24 h recovery of mRBCs. But freezing can seriously decrease the life span of mRBCs, which may correlate with the increase of osmotic fragility caused by freezing. Moreover, incubation with glucose before freezing can not increase the life span of mRBCs frozen with dextran. In the future, we will focus on how to decrease osmotic fragility of red blood cells frozen in the absence of glycerol. References [1] American Association of Blood Banking Technical Manual, 50th ed., Bethesda, MD, USA, 2003. [2] G.M. Beattie, J.H. Crowe, A.D. Lopez, V. Cirulli, C. Ricordi, A. Hayek, Trehalose: a cryoprotectant that enhances recovery and preserves function of human pancreatic islets after long term storage, Diabetes 46 (1997) 519–523. [3] E.C. Butcher, I.L. Weissman, Direct fluorescent labeling of cells with fluorescein or rhodamine isothiocyanate. I. Technical aspects [J], J. Immunol. Methods 37 (1980) 97–108. [4] J.H. Crowe, L.M. Crowe, Preservation of mammalian cells-learning from nature’s trick, Nat. Biotechnol. 18 (2000) 145–146. [5] J.H. Crowe, L.M. Crowe, A.E. Oliver, N.M. Tsvetkova, W. Wolkers, T. Tablin, The trehalose myth revised: introduction to a symposium on stabilization of cells in the dry state, Cryobiology 43 (2001) 89–105. [6] J.H. Crowe, A.E. Oliver, F.A. Hoekstra, L.M. Crowe, Stabilization of dry membranes by mixtures of hydroxyethyl starch and glucose: the role of vitrification, Cryobiology 35 (1997) 20–30. [7] A. Eroglu, M.J. Russo, R. Bieganski, A. Fowler, S. Cheley, H. Bayley, M. Toner, Intracellular trehalose improves the survival of cryopreserved mammalian cells, Nat. Biotechnol. 18 (2000) 163–167. [8] A. Eroglu, M. Toner, T.L. Toth, Beneficial effect of microinjected trehalose on the cryosurvival of human oocytes, Fertil. Steril. 77 (2002) 152–158. [9] B.J. Fuller, N. Lane, E.E Benson, Life in the frozen state, first ed, CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431, 2004. [10] N. Guo, I. Puhlev, D.R. Brown, J. Mansbridge, F. Levine, Trehalose expression confers desiccation tolerance on human cells, Nat. Biotechnol. 18 (2000) 168– 171. [11] M. Hamidi, H. Tajerzadeh, Carrier erythrocytes: an overview, Drug Delivery 10 (2003) 9–20. [12] G. Hoffmann-Fezer, J. Mysliwietz, W. MOrtlbauer, H.J. Zeitler, E. Eberle, U. Hiinle, S. Thierfelder, Biotin labeling as an alternative nonradioactive approach to determination of red cell survival, Ann. Hematol. 67 (1993) 81–87. [13] K.M. Hoffmeister, T.W. Felbinger, H. Falet, C.V. Denis, W. Bergmeier, T.N. Mayadas, U.H. Von Andrian, D.D. Wagner, T.P. Stossel, J.H. Hartwig, The clearance mechanism of chilled blood platelets, Cell 112 (2003) 87–97. [14] E.P. Horn, A. Sputtek, T. Standl, B. Rudolf, P. Kühnl, J. Schulte am Esch, Transfusion of autologous, Hydroxyethyl starch-cryopreserved red blood cells, Anesth. Analg. 85 (1997) 739–745. [15] P. Jilma-Stohlawetz, Biotinylation: a nonradioactive method for labeling of blood components, Infus. Ther. Transfus. Med. 27 (2000) 296–300. [16] T. Kanias, J.P. Acker, Trehalose loading into red blood cells is accompanied with hemoglobin oxidation and membrane lipid peroxidation, Cryobiology 58 (2009) 232–239.
16
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
[17] S. Khandelwal, R.K. Saxena, Age-dependent increase in green autofluorescence of blood erythrocytes, J. Biosci. 32 (2007) 1139–1145. [18] C.T. Knorpp, W.R. Merchant, P.W. Gikas, H.H. Spencer, N.W. Thompson, Hydroxyethyl starch: extracellular cryophyllactic agent for erythrocytes, Science 157 (1967) 1312–1313. [19] S.B. Leslie, E. Israeli, B. Lighthart, J.H. Crowe, L.M. Crowe, Trehalose and sucrose protect both membranes and proteins in intact bacteria during drying, Econ. Environ. Microbiol. 61 (1995) 3592–3597. [20] S.B. Li, Y.J. Jia, Z.P. Liu, Y.P. Zhang, Determining half life of mPEG-sheathed erythrocyte with FITC labeling, Bull. Acad. Mil. Med. Sci. 27 (2003) 367–369. [21] Y. Miyamoto, S. Suzuki, K. Nomura, S. Enosawa, Improvement of hepatocyte viability after cryopreservation by supplementation of long-chain oligosaccharide in the freezing medium in rats and humans, Cell Transplant. 15 (2006) 911–919. [22] A.E. Oliver, K. Jamil, J.H. Crowe, F. Tablin, Loading human mesenchymal stem cells with trehalose by fluid-phase endocytosis, Cell Preserv. Technol. 2 (2004) 35–49. [23] C. Pellerin-Mendes, L. Million, M. Marchand-Arvier, P. Labrude, C. Vigneron, In vitro study of the protective effect of trehalose and dextran during freezing of human red blood cells in liquid nitrogen, Cryobiology 35 (1997) 173–186. [24] G.B. Quan, Y. Han, M.X. Liu, F. Gao, Effects of pre-freeze incubation of human red blood cells with various sugars on postthaw recovery when using a dextran-rapid cooling protocol, Cryobiology (2009), doi:10.1016/j.cryobiol. 2009.08.001. [25] G.B. Quan, M.X. Liu, S.P. Ren, J.H. Zhang, Y. Han, Changes of phosphatidylserine distribution in human red blood cells during the process of loading sugars, Cryobiology 53 (2006) 107–118. [26] G.B. Quan, L. Zhang, Y. Guo, M.X. Liu, J.X. Wang, Y. Wang, B. Dong, A. Liu, J.G. Zhang, Y. Han, Intracellular sugars improve survival of human red blood cells cryopreserved at 80 °C in the presence of polyvinyl pyrrolidone and human serum albumin, Cryoletter 28 (2007) 95–108.
[27] S.P. Ren, E.P. Ma, X.Z. Liu, Y. Han, A. Liu, P. Jin, B. Dong, Labeling and survival studies on rabbit RBCs, Chin. J. Appl. Physiol. 19 (2003) 410–412. [28] A.P. Rinfret, Factors affecting the erythrocyte during rapid freezing and thawing, Ann. NY Acad. Sci. 85 (1960) 576–594. [29] A.W. Rowe, Cryopreservation of red cells by freezing and vitrificationsome recollections and predictions, Infus. Ther. Transfus. Med. 29 (2002) 25–30. [30] G.R. Satpathy, Z. Török, R. Bali, D.M. Dwyre, E. Little, N.J. Walker, F. Tablin, J.H. Crowe, N.M. Tsvetkova, Loading red blood cells with trehalose: a step towards biostabilization, Cryobiology 49 (2004) 123–136. [31] K.L. Scott, J. Lecak, J.P. Acker, Biopreservation of red blood cells: past, present, and future, Transfus. Med. Rev. 19 (2005) 127–142. [32] S.E. Slezak, P.K. Horan, Fluorescent in vivo tracking of hematopoietic cells. I. Technical considerations, Blood 74 (1989) 2172–2177. [33] A. Sputtek, G. Singbartl, R. Langer, W. Schleinzer, H.A. Henrich, P. Kühnl, Cryopreservation of red blood cells with the non-penetrating cryoprotectant hydroxyethyl starch, Cryo Letters 16 (1995) 283–288. [34] M. Tomita, T. Osada, I. Schiszler, Y. Tomita, M. Unekawa, H. Toriumi, N. Tanahashi, N. Suzuki, Automated method for tracking vast numbers of FITCLabeled RBCs in microvessels of rat Brain In Vivo using a high-speed confocal microscope system, Microcirculation 15 (2008) 63–174. [35] C.T. Wagner, M.L. Martowicz, S.A. Livesey, J. Connor, Biochemical stabilization enhances RBC recovery and stability following cryopreservation, Cryobiology 45 (2002) 153–166. [36] Y. Wang, P. Jin, M.X. Liu, A. Liu, G.B. Quan, W.B. Hu, Y. Guo, Y. Han, Comparison of two methods of washing frozen red blood cells in animal mode, J Clin. Transfus. Lab. Med. 18 (2006) 81–84. [37] R.J. Williams, The surface activity of PVP and other polymers and their antihemolytic capacity [J], Cryobiology 20 (1983) 521–526. [38] W.F. Wolkers, N.J. Walker, F. Tablin, J.H. Crowe, Human platelets loaded with trehalose survive freeze-drying, Cryobiology 42 (2001) 79–87.
Contents lists available at ScienceDirect
Cryobiology journal homepage: www.elsevier.com/locate/ycryo
In vivo circulation of mouse red blood cells frozen in the presence of dextran and glucose q Guo Bo Quan, Min Xia Liu, Ying Han *, Lei Fang, Wei Du, Jie Xi Wang Beijing Institute of Transfusion Medicine, the Taiping Road 27#, the Haidian County, Beijing 100850, China
a r t i c l e
i n f o
Article history: Received 9 September 2009 Accepted 16 February 2010 Available online 20 February 2010 Keywords: Mouse red blood cells Glucose Freeze Dextran Life span The 24 h recovery
a b s t r a c t Loading with monosaccharide can improve the quality of human red blood cells (hRBCs) frozen with polymer. But in vivo life span of hRBCs frozen with polymer and sugar is not determined. In this study, following incubation with glucose, mouse red blood cells (mRBCs) were frozen in liquid nitrogen for 24 h using dextran as the extracellular protectant. After thawing, hemolysis, exposure of PS, and osmotic fragility of frozen mRBCs were determined in vitro. After transfusion of fluorescein isothiocyanate (FITC)labeled mRBCs, the 24 h recovery and half life span of frozen mRBCs were determined. The data indicated the postthaw hemolysis of mRBCs frozen with dextran and glucose were significantly less than that of cells frozen with dextran (17.23% ± 5.21% vs 25.96% ± 10.07%, P = 0.034). But freezing can also result in exposure of phosphatidylserine and increase of osmotic fragility of mRBCs. After transfusion, the 24 h recovery of mRBCs frozen in the absence or presence of glucose was similar to that of the control cells (P = 0.748 and 0.971). However, the half life span of mRBCs frozen in the absence or presence of glucose was significantly less than that of the control cells (P = 0.000). In addition, incubation with glucose can not increase the life span of frozen red blood cells (7.16 ± 0.93 d vs 7.15 ± 0.34 d, P = 0.982). In conclusion, incubation with monosaccharide could significantly increase the recovery of mRBCs frozen with polymer. Although freezing can significantly shorten the half life span of frozen cells, it can not influence the 24 h recovery of frozen mRBCs. In addition, incubation with monosaccharide before freezing can not increase the life span of frozen mRBCs. So according to the above data, to increase the life span of hRBCs frozen with polymer and monosaccharide, the osmotic fragility of the frozen RBCs must be decreased in the future. Ó 2010 Elsevier Inc. All rights reserved.
Introduction At present, freezing with glycerol can greatly prolong the storage time of RBCs in vitro [1]. However, freezing with glycerol has some limits which may affect activeness of people to use frozen RBCs in clinical therapy. The thawed RBCs require complicated washing procedure to remove glycerol. Furthermore, the addition and removal of glycerol can lead to transient osmotic injury and a significant loss of cells [35]. Some polymers, such as hydroxyethyl starch (HES), polyvinyl pyrrolidone (PVP), and dextran, have been used as alternatives to glycerol [14,18,23,26,28,33,37]. But freezing with polymers also faces a great challenge. Polymers can’t penetrate the membrane of red blood cells, so their capability to protect the inner membrane may be discounted. Recently, the super protection of treha-
q Statement of funding: This work was supported by the National Natural Science Foundation of China (Grant No. N30400518) and Beijing Natural Science Foundation of China (Grant No. 7072059). * Corresponding author. E-mail address: [email protected] (Y. Han).
0011-2240/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.cryobiol.2010.02.004
lose on frozen or dry cells has received attentions of many investigators [2,4,5,7,8,10,19,21,23,26,38]. But in order to obtain the optimal protective effect, trehalose must be loaded into the cytoplasm [4,38]. At present, some methods, including genetic technology [7,10], microinjection [8], fluid phase endocytosis [22,38], and so on may not be suitable for loading trehalose into RBCs. Although Satpathy et al. found osmotic imbalance and membrane lipid phase transition could introduce trehalose into RBCs [30], their method can also result in serious hemolysis [16,24,25] and oxidative stress [16] which may directly cause more cell injuries during subsequent freezing and thawing [24,26]. In the previous study, compared with trehalose, incubation with monosaccharide before freezing can efficiently increase the recovery of hRBCs frozen with polymer [24,26]. According to ‘‘the water replacement hypothesis” raised by Crowe et al., monosaccharide may also stabilize the membrane of frozen or dried RBCs through forming hydrogen bonds with phospholipids and proteins like trehalose [6]. In addition, accumulation of intracellular glucose can increase the vitrification degree of cytoplasm and decrease formation of lethally intracellular ice. Most importantly, human red blood cells can utilize glucose through glycolysis, but can not
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
metabolize trehalose owing to lack of trehalases which are required for enzymatic hydrolysis of trehalose [5,30]. So glucose may stabilize the metabolic function of frozen red blood cells [26]. In order to further evaluate the efficacy of the present freezing method, this study is to determine the in vivo 24 h recovery and half life span of mRBCs frozen with dextran and glucose. The Kunming (KM) white mouse was used as the experimental animal. FITC was used to label mRBCs. The final purpose is to obtain the data about the in vivo survival of frozen mRBCs and improve the present freezing method in the absence of glycerol. Materials and methods Reagents and solutions Unless otherwise stated, all chemicals were analytical reagent grade. Dextran (MW40,000 Da) and FITC were purchased from Sigma Chemical Co. (St. Louis, MO, USA). All solutions were prepared in double glass-distilled water. The modified phosphate buffered saline (mPBS) containing 154 mM NaCl, 1.06 mM KH2PO4, 5.6 mM Na2HPO4 and 2.0 mM adenine (300 mOsm, pH 7.4) was used as the basic buffer. The mPBS containing 40% dextran (W/V) was used as the extracellular protective solution. Preparation of mRBCs Female KM mice weighted 25–30 g were purchased from Academy of Military Medical Sciences, Beijing, China. Approximately 1 ml of blood was collected by cardiac puncture from mice anaesthetized with 0.2% pentobarbital. The citrate–phosphate-dextrose medium was used as the anticoagulant. The whole blood was not leukoreduced prior to processing. The blood was centrifuged at 1550g for 2 min at 4 °C to remove the supernatant, including white blood cells and platelets. Then the concentrated mRBCs were washed three times using mPBS. The final concentrated mRBCs were used in the study. A spun hematocrit method was used to determine the hematocrit. The hematocrits used in sugar loading and freezing were approximately 22% and 51%, respectively. The process of loading glucose into red blood cells has been described in the previous studies [24–26]. Briefly, the washed mRBCs were incubated in the mPBS buffers containing glucose for 3 h at 37 °C. After incubation, the cell suspension was centrifuged at 1550g for 2 min to remove the supernatant. Then the concentrated cells were washed using mPBS for three times to fully remove the extracellular glucose. The intracellular glucose concentration was measured by the anthrone method [24,30]. Since the anthrone method detects all sugars present in mRBCs, the control cells were treated in parallel to determine the concentration of endogenous glucose. Finally, the concentration of the endogenous glucose in the mRBCs was subtracted from the total sugar concentration. Freezing and thawing The freezing procedure was described in the previous study [24]. Briefly, the concentrated cells loaded with glucose and 40% dextran were mixed in a ratio of 1:1 (V/V). Then the cell suspension was divided into the standard 2.0 ml cryogenic tubes (Axygen Scientific, California, USA) and frozen in liquid nitrogen at approximately 2.5 °C/s. Every tube contained 1 ml of the cell suspension. The cooling rate was measured with a thermocouple (EastSun Electronic, Zhejiang, China) placed in the middle of the sample. Measurement of the cooling rate was made between 25 and 50 °C, which represents the linear changing range of the cooling rate. The frozen mRBCs were stored in the liquid nitrogen for 24 h and then thawed for 5 min in the 37 °C water bath.
11
In vitro evaluation of quality of frozen mRBCs In order to evaluate the quality of frozen mRBCs, hemolysis, distribution of phosphatidylserine, and osmotic fragility were measured in vitro. The hemolysis was determined using the Drabkin’s solution (Sigma Diagnostics, St. Louis, MO, USA). The thawed cell suspensions were initially mixed with equal volume of mPBS to decrease the viscosity. Then the cell suspension was used to measure the total hemoglobin concentration. The supernatant was used to measure the free hemoglobin concentration following centrifugation. Hemoglobin was transformed to cyanmethemoglobin and the absorbance of cyanmethemoglobin was measured at 540 nm using an ultraviolet spectrophotometer (PGENERAL, Beijing, China). The isotonic mPBS was used as the blank reference. The percent hemolysis was calculated using the following formula:
ðOD540 nm of the supernatant=OD540 nm of the total hemoglobinÞ ð100% Hct%Þ In addition, the hemolysis caused by glucose loading process before freezing was also evaluated. After loading with glucose, the OD values of the total hemoglobin and the supernatant were measured by the above method. The percent hemolysis was calculated using the above formula. The annexin-V-Fluos kit purchased from Roche Co. (Nonnenwald2, 82372 Penzberg, Germany) was used to label mRBCs with exposed phosphatidylserine. Briefly, the approximate 1 106 mRBCs were directly resuspended in the 200 ll of annexin binding buffer. Then the 0.2 lg of annexin-V-FITC was added to the above cell suspension and incubated for 15 min in dark environment at 25 °C. Before flow cytometry, 300 ll of binding buffer was added to the above cell suspension. Finally, the data about cells with exposed PS were analyzed by the flow cytometer (Becton–Dickinson, San Jose, CA, USA). In order to evaluate osmotic tolerance, the 0.9% NaCl diluted with the distilled water was used as the osmotic fragility solutions containing 0–0.9% NaCl. The concentration gradient of NaCl was 0.09%. After washed using mPBS for five times to remove the cryoprotectants and the free hemoglobin fully, the concentrated mRBCs were suspended in every osmotic fragility solution for 2 h at room temperature. Following centrifugation at 1550g for 5 min, the percent hemolysis was determined using the Drabkin’s solution. Determination of the 24h recovery and half life span In this study, FITC was used to label mRBCs. The procedure has been described previously [20,27]. Blood was collected from eighteen mice. The freezing study was divided into two groups according to use of glucose. Following freezing and thawing, the mRBCs were washed using mPBS for five times to fully remove the free hemoglobin and the extracellular protectants. The concentrated mRBCs were diluted with mPBS in a ratio of 1:1. Then the cell suspension was incubated with FITC for 30 min in the 37 °C water bath. The final concentration of FITC in the cell suspension was 70 lg/ml. After incubation, mRBCs were washed for three times to fully remove the extracellular FITC. The final concentrated mRBCs were suspended in mPBS. Approximately 1 1012 red blood cells labeled with FITC in 300 ll of mPBS were injected into a normovolemic mouse (allogeneic transfusion) through tail veins. In every freezing group, six mice were used as the recipients. Following transfusion for different time, 2.5 ll of blood was collected from mouse tail veins using the citrate–phosphate-dextrose medium as the anticoagulant. The blood sample was diluted using 0.5 ml mPBS. Then the cell samples were analyzed using a FACScan
12
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
flow cytometer (Becton–Dickinson, San Jose, CA, USA) equipped with an argon laser, 488 nm wave length. The percentage of FITC-labeled cells and the GEO mean were determined using FACScan software. The GEO mean can reflect the fluorescence intensity of mRBCs. The percentages of FITC-labeled mRBCs after in vivo circulation for 2 h, 1 d, 3 d, 6 d, 7 d, and 9 d were labeled as A0, A1, A3, A6, A7, and A9 respectively. The 24 h recovery was calculated using the following formulae: A1/A0 100(%). In order to calculate the half life span, the recovery of FITC-labeled cells was used as the Y-axis. The time was used as the X-axis. The Microcal (TM) Origin software (Microcal Software Inc., One Roundhouse Plaza, Northampton, MA 01060, USA) was used for regression analysis of the data and building of regression equation. When the recovery of FITC-labeled mRBCs was 50%, the corresponding time was the half life span of mRBCs. Statistical Analysis Each set of experimental conditions was replicated different times. The data were analyzed using independent-samples T test and one-way ANOVA with calculation of the least significant difference (LSD). All results with a value of P < 0.05 or P < 0.01 were considered to be statistically significant. The data were expressed as mean ± SD. Results Incubation with glucose increases the recovery of mRBCs frozen with dextran The loading efficiency of glucose in mRBCs was similar to that of hRBCs [24,25]. With increase of the extracellular concentrations of glucose, the loading efficiency of glucose in mRBCs was also increased steadily. When the extracellular concentration of glucose was increased from 200 mM to 400 mM, the intracellular concentration of glucose was increased from 3.47 ± 1.25 mM to 7.98 ± 3.21 mM steadily. Although high extracellular concentration of glucose can increase the loading efficiency of glucose, this procedure can result in hemolysis of mRBCs. The data were shown in the Fig. 1. The data was collected six mice. After washing, the concentrated cells were incubated with 200 mM or 400 mM glucose for 3 h. The Fig. 1 indicated with increase of the concentrations of glucose, the percent hemolysis of mRBCs also increased steadily. When the concentration of glucose was 200 mM or 400 mM, the percent hemolysis was 3.12% ± 1.19% and 4.97% ± 0.92%, respectively which were significantly more than that of the control cells (0.34% ± 0.21%, P = 0.000). But incubation with glucose can significantly decrease the postthaw hemolysis of mRBCs frozen with dextran. Moreover, with the increase of the concentrations of glucose, the postthaw hemolysis was firstly decreased and then increased. When the concentration of glucose was 200 mM, the postthaw percent hemolysis of mRBCs was 17.23% ± 5.21% and significantly less than that of cells frozen in the absence of glucose (25.96% ± 10.07%, P = 0.034). However, when the concentration of glucose was increased to 400 mM, the postthaw percent hemolysis was 42.57% ± 11.29%.
Fig. 1. The hemolysis of mRBCs caused by incubation with glucose and freezing. The procedures of glucose loading and freezing were described in the part of materials and methods. The percent hemolysis was determined using the Drabkin’s method. In order to evaluate the effect of glucose loading on hemolysis of mRBCs, the data were collected from six mice. The data points are expressed as means ± SD. ‘‘” P < 0.01 vs red blood cells loaded with glucose before freezing. In order to evaluate the effect of freezing on hemolysis of mRBCs, blood was collected from four mice in every study. Then in the presence or absence of glucose, the mRBCs were frozen with polymer. This whole study was repeated six times. ‘‘DD” P < 0.01 vs frozen mRBCs following incubation with 200 mM glucose or with no glucose. ‘‘D” vs frozen mRBCs following incubation with 200 mM glucose. The data were analyzed using one-way ANOVA with calculation of the least significant difference (LSD).
Fig. 2. The effects of glucose loading and freezing on PS distribution of frozen mRBCs. The Annexin V was used to label exposed PS. This study was repeated five times. In every study, blood was collected from three mice. Then in the presence or absence of glucose, the mRBCs were frozen with polymer. The figure represented the data of one study. The ‘‘M1” field represented the cells with exposed PS. ‘‘Con” represented the control cells. ‘‘Loaded with glu” represented cells loaded with glucose. ‘‘no sugar” represented mRBCs frozen with no sugar. ‘‘glu” represented mRBCs frozen with dextran and glucose.
Freezing influences distribution of phosphatidylserine of mRBCs The effect of freezing on the postthaw distribution of PS of mRBCs was shown in the Fig. 2. This study was repeated five times. In every study, blood was collected from three mice. Then in the presence or absence of glucose, the mRBCs were frozen with polymer. The Fig. 2 represented the data from one study. After incubation with glucose, the percentage of cells with exposed PS was
2.13% ± 0.96% and similar to that of the control cells (1.87% ± 1.02%, P = 0.458). However, the postthaw percentage of cells with exposed PS was increased to 12.96% ± 7.23% and significantly higher than that of the control cells (P = 0.000). In addition, similar to hRBCs, incubation with glucose can not decrease the exposure of phoshatidylserine of mRBCs frozen with polymer.
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
13
Fig. 4. Flow cytometry analysis of mRBCs labeled with FITC. (A) represented mRBCs without labeling with FITC; (B) represented mRBCs labeled with FITC. The final concentration of FITC in the cell suspension was 70 lg/ml.
Fig. 3. The effect of freezing on osmotic fragility of mRBCs. The washed cells were suspended in the dilutions of 0.9%NaCl and the hemolysis was measured. The study was repeated six times. In every study, blood was collected from three mice. Then in the presence or absence of glucose, the mRBCs were frozen with polymer. The data points are expressed as means ± SD. ‘‘con, no sugar, and glu” represented the control cells, mRBCs frozen with no sugar, and mRBCs frozen with glucose. ‘‘” P < 0.01 vs the frozen cells. ‘‘DD” P < 0.01 vs the cells frozen with dextran and glucose. The data were analyzed using one-way ANOVA with calculation of the least significant difference (LSD).
The effect of freezing on the osmotolerance of mRBCs The osmotic fragility of frozen mRBCs was more than that of the control cells (Fig. 3). The study was repeated six times. In every study, blood was collected from three mice. Then the mRBCs were frozen with polymer in the presence or absence of glucose. With increase of the concentrations of NaCl, the hemolysis decreased steadily. When the concentrations of NaCl were less than 0.54%, the percent hemolysis of frozen cells was not more than that of the control cells. However, when the concentrations of NaCl were more than 0.54%, the percent hemolysis of frozen mRBCs was significantly higher than that of the control cells (P < 0.01). In addi-
tion, this figure showed when the concentration of NaCl was increased to 0.72%, incubation with glucose could significantly decrease the percent hemolysis of mRBCs frozen with dextran.
In vivo circulation of frozen mRBCs after transfusion In this study, FITC was used to label mRBCs. The percentage of mRBCs labeled with FITC was more than 99% in vitro (seen from the Fig. 4), which showed most of mRBCs could efficiently bind with FITC. Then the fluorescent intensity of FITC-labeled red blood cells in vivo was measured. The data were shown in the Fig. 5A. Blood was collected from four mice. The washed mRBCs labeled with FITC were transfused into six recipients. The GEO mean of red blood cells after circulation for 2 h (day 0) was 68.35 ± 0.99. After circulation for 24 h (day 1), the GEO mean of mRBCs was decreased to 54.31 ± 1.83 and significantly less than that of day 0 (P = 0.007). Subsequently, the GEO means maintained constant and not significantly different from that of day 1 (P > 0.05). In addition, our study showed the autofluorescent intensity of mRBCs without labeling with FITC was approximately 4–5 and significantly less than that of cells labeled with FITC. So the adverse effect of autofluorescent intensity of mRBCs may be obviated in this study.
Fig. 5. The fluorescent intensity and in vivo recovery of FITC-labeled mRBCs after transfusion. After transfusion, the flow cytometry was used to measure the fluorescent intensity and recovery of mRBCs at different time points.(A) The GEO mean represented the fluorescent intensity. Blood was collected from four mice. The washed mRBCs labeled with FITC were transfused into six recipients. The data points were expressed as means ± SD. ‘‘” P < 0.01 vs the other time points. The data were analyzed using oneway ANOVA with calculation of the least significant difference (LSD). (B) The measuring procedures of the recovery of transfused mRBCs were described in the part of materials and methods. Briefly, the percentages of FITC-labeled mRBCs after in vivo circulation for 0, 1 d, 3 d, 6 d, 7 d, and 9 d were labeled as A0, A1, A3, A6, A7, and A9 respectively. The recovery rate was calculated using the following formulae: A1/A0 100(%).The data were collected from 18 mice equally divided into three groups (the group con, no sugar, and glu). So each group contained six mice. The data points were expressed as means ± SD. ‘‘” vs the in vivo recovery of frozen mRBCs at certain time points. The data were analyzed using the independent-samples T test.
14
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
The recovery of mRBCs labeled with FITC after transfusion was shown in the Figs. 5B and 6. The data were collected from 18 mice equally divided into three groups. Each group contained six mice. The Fig. 6 represented the data of one mouse randomly selected from the three groups. Seen from Fig. 5B, with the prolonging of time, the recovery of mRBCs labeled with FITC decreased steadily. However, the in vivo recovery of frozen mRBCs was significantly less than that of the control mRBCs after day 1. In the Table 1, the 24 h recovery of the control cells was 83.66% ± 3.01%. However, the 24 h recoveries of mRBCs frozen in the absence or presence of glucose were 82.62% ± 8.47% and 85.32% ± 3.88% respectively. The 24 h recovery of frozen mRBCs was not significantly less than that of the control cells (P = 0.748 and 0.971). But the life span of frozen mRBCs was significantly less than that of the control cells (P = 0.000). The half life span of the control cells was 10.39 ± 0.55 d. However, the half life span of red blood cells frozen in the absence or presence of glucose was decreased to approximately 7 d, which showed freezing had injured mRBCs. In addition, the half life span of the cells frozen with glucose was similar to that of the cells frozen with no glucose (7.16 ± 0.93 d vs 7.15 ± 0.34 d, P = 0.982). Discussion Polymers can replace glycerol to freeze hRBCs. Some polymers such as albumin, dextran, modified gelatine, polyvinylpyrrolidone,
Table 1 The 24 h recovery and half life span of frozen mRBCs in vivo. Groups
n
The 24 h recovery (%)
Con No sugar Glu
6 6 6
83.66 ± 3.01 82.62 ± 8.47 85.32 ± 3.88
The p value
0.748 0.971
Half life span(d) 10.39 ± 0.55** 7.16 ± 0.93 7.15 ± 0.34
The p value
0.000 0.000
The measuring procedures of the 24 h recovery and half life span of mouse red blood cells were described in the part of materials and methods. Briefly, the percentages of FITC-labeled mRBCs after in vivo circulation for 2 h, 1 d, 3 d, 6 d, and 9 d were labeled as A0, A1, A3, A6, and A9 respectively. The 24 h recovery was calculated using the following formulae: A1/A0 100(%). In order to calculate the half life span, the recovery of FITC-labeled cells was used as the Y-axis. The time was used as the X-axis. The Microcal (TM) Origin software was used for regression analysis and building of regression equation. With a recovery of FITC-labeled mRBCs was 50%, the corresponding time was the half life span of mRBCs. The data were collected from 18 mice equally divided into three groups (the group con, no sugar, and glu). So each group contained six mice. ‘‘con, no sugar, and glu” represented the control cells, mRBCs frozen with no sugar, and mRBCs frozen with glucose. The data were analyzed using independent-samples T test. ** P < 0.01 vs the frozen mRBCs.
polyethylene oxide, polyethylene glycol, and HES can not enter the cells. This property facilitates their removal after thawing. In the case of emergencies, this step could be simplified or omitted, if the additives, for example, albumin, dextran, and HES, are biodegradable and tolerated by the patient [9]. But this method can also cause in vivo hemolysis after transfusion [29,31]. The report of
Fig. 6. The recovery of FITC-labeled mRBCs after freezing and transfusion. The procedures of freezing and labeling with FITC were described in the part of materials and methods. The data were collected from 18 mice equally divided into three groups (the group C, N, and S). So each group contained six mice. This figure represented the data of one mouse randomly selected from the three groups. ‘‘C, N, and G” represented the control cells, mRBCs frozen with no sugar, and mRBCs frozen with glucose.
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
Horn et al. showed after transfusion of hRBCs frozen with HES, the plasma hemoglobin levels increased twofold to threefold, although the plasma hemoglobin could decrease to baseline levels within 24 h [14]. In order to resolve this problem, the protective effect of monosaccharides including glucose, fructose, and galactose on red blood cells frozen with polymer was evaluated in the previous study [24,26]. Loading with monosaccharide, especially galactose, could significantly increase the recovery of hRBCs frozen with dextran. However, the effect of sugar on in vivo circulation of frozen hRBCs is not known. The in vivo 24 h recovery and half life span are important index to evaluate the quality of preserved hRBCs. The minimal recovery of transfused RBCs set by the American Association of Blood Bank after circulation for 24 h is 75% [1]. It is well known that the storage procedures, including the hypothermic storage and freezing or cryopreservation, can result in decreasing recovery of in vivo transfused hRBCs [31]. After storage at 4 °C, approximately 30% and 70% of transfused RBCs disappeared from the circulation after 1 and 3 d, respectively [31]. But the reason is still unknown. Determination of RBC life span is currently performed by random labeling methods using one of the radioactive tracers: 51Cr, DF32P, or DFP-H3 (diisopropylphosphorofluoridate) [12,32]. All of these tracers have disadvantages which make them less than ideal [32]. So some investigators use non-radioactive tracers to replace the radioactive tracers, including fluorescein [3,20,27,34,36], biotin [12,15], and fluorescent lipophilic probes (PKH-3) [32]. FITC is a fluorescein with two hydrophilic hydroxyl groups and can covalently bind to the e-amino nitrogen of lysine, so FITC can stably label most of proteins [27]. Some studies used FITC to label RBCs of mouse, rat, dog, rabbit, or human [11,20,27,34,36]. Butcher et al. found FITC could stably label lymphocyte and not shift to the other cells [3]. In this study, the fluorescent intensity of FITC-labeled mRBCs remained constant from day 1 to day 9. However, autofluorescence of RBCs may influence the results of this study. The formation of autofluorescence of RBCs has been known to accompany the lipid peroxidation process. The interaction of malonyldialdehyde and the amino groups of phospholipids or proteins during lipid peroxidation can result in the formation of fluorescent chromolipids. In addition, spontaneous glycation of glycated proteins and carbohydrates may also generate autofluorescence in RBCs [17]. So the effect of autofluorescence of RBCs must be considered. In this study, the autofluorescent intensity (the GEO mean) of mRBCs was approximately 4–5. On the contrary, the fluorescent intensity of mRBCs labeled with FITC was increased to approximately 200–400. Following transfusion, the fluorescent intensity of frozen mRBCs decreased to approximately 50 and significantly higher than autofluorescent intensity of mRBCs. So the antofluorescence of mRBCs can not influence the final results in this study. Here, the KM white mouse was used as the experimental animal. The murine model has been used to study the recovery of blood cells in vivo [12,13,20,34]. Hoffmeister et al. used the murine model to study the clearance mechanism of chilled blood platelets [13]. In this study, the selection of the animal model must comply with the following standard: like hRBCs, with the increase of the concentrations of monsaccharide, the postthaw hemolysis of frozen mRBCs was firstly decreased and then increased [24]. This study showed the changing mode of postthaw hemolysis of mRBCs was similar to that of hRBCs. Moreover, the in vitro data showed loading with 200 mM glucose could significantly decrease the percent hemolysis of mRBCs frozen with dextran. But when the concentration of glucose was increased to 400 mM, the postthaw hemolysis became more serious. So it may be suitable to select KM mouse as the experimental animal. In addition, freezing can also result in exposure of phosphatidylserine and increasing osmotic fragility of mRBCs. These results were similar to hRBCs in our previous report [24]. Moreover, incubation with glucose can not
15
mitigate the malignant effect of freezing on distribution of phosphatidylserine and osmotolerance of mRBCs which may influence the life span of mRBCs in vivo. The in vivo study showed the 24 h recovery of frozen mRBCs after transfusion was more than 80% and similar to that of the control cells, which showed freezing with polymer and glucose could not influence the 24 h recovery of frozen cells. However, freezing can seriously influence the life span of mRBCs. In this study, the half life span of the control mRBCs was approximately 10 d. According to the report of Li et al., the half life span of KM mRBCs was approximately 12 d [20]. So our data was similar to the report of Li et al. But freezing can reduce the half life span of mRBCs. After transfusion, the half life span of frozen cells was approximately 7 d and significantly less than that of the control cells. Furthermore, loading with glucose can not increase the life span of mRBCs frozen with polymer. So the injury from freezing may be the main factor to decrease life span of mRBCs. In addition, this study also indicated the 24 h recovery could not reflect the real state of frozen RBCs in vivo. The life span must be also considered. Transfusion is an important approach to save the life of patients. But with regard to those patients having specific diseases such as disorder of hematopoiesis, it is beneficial to prolong the in vivo life span of RBCs. In conclusion, freezing with dextran and glucose can not influence the 24 h recovery of mRBCs. But freezing can seriously decrease the life span of mRBCs, which may correlate with the increase of osmotic fragility caused by freezing. Moreover, incubation with glucose before freezing can not increase the life span of mRBCs frozen with dextran. In the future, we will focus on how to decrease osmotic fragility of red blood cells frozen in the absence of glycerol. References [1] American Association of Blood Banking Technical Manual, 50th ed., Bethesda, MD, USA, 2003. [2] G.M. Beattie, J.H. Crowe, A.D. Lopez, V. Cirulli, C. Ricordi, A. Hayek, Trehalose: a cryoprotectant that enhances recovery and preserves function of human pancreatic islets after long term storage, Diabetes 46 (1997) 519–523. [3] E.C. Butcher, I.L. Weissman, Direct fluorescent labeling of cells with fluorescein or rhodamine isothiocyanate. I. Technical aspects [J], J. Immunol. Methods 37 (1980) 97–108. [4] J.H. Crowe, L.M. Crowe, Preservation of mammalian cells-learning from nature’s trick, Nat. Biotechnol. 18 (2000) 145–146. [5] J.H. Crowe, L.M. Crowe, A.E. Oliver, N.M. Tsvetkova, W. Wolkers, T. Tablin, The trehalose myth revised: introduction to a symposium on stabilization of cells in the dry state, Cryobiology 43 (2001) 89–105. [6] J.H. Crowe, A.E. Oliver, F.A. Hoekstra, L.M. Crowe, Stabilization of dry membranes by mixtures of hydroxyethyl starch and glucose: the role of vitrification, Cryobiology 35 (1997) 20–30. [7] A. Eroglu, M.J. Russo, R. Bieganski, A. Fowler, S. Cheley, H. Bayley, M. Toner, Intracellular trehalose improves the survival of cryopreserved mammalian cells, Nat. Biotechnol. 18 (2000) 163–167. [8] A. Eroglu, M. Toner, T.L. Toth, Beneficial effect of microinjected trehalose on the cryosurvival of human oocytes, Fertil. Steril. 77 (2002) 152–158. [9] B.J. Fuller, N. Lane, E.E Benson, Life in the frozen state, first ed, CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431, 2004. [10] N. Guo, I. Puhlev, D.R. Brown, J. Mansbridge, F. Levine, Trehalose expression confers desiccation tolerance on human cells, Nat. Biotechnol. 18 (2000) 168– 171. [11] M. Hamidi, H. Tajerzadeh, Carrier erythrocytes: an overview, Drug Delivery 10 (2003) 9–20. [12] G. Hoffmann-Fezer, J. Mysliwietz, W. MOrtlbauer, H.J. Zeitler, E. Eberle, U. Hiinle, S. Thierfelder, Biotin labeling as an alternative nonradioactive approach to determination of red cell survival, Ann. Hematol. 67 (1993) 81–87. [13] K.M. Hoffmeister, T.W. Felbinger, H. Falet, C.V. Denis, W. Bergmeier, T.N. Mayadas, U.H. Von Andrian, D.D. Wagner, T.P. Stossel, J.H. Hartwig, The clearance mechanism of chilled blood platelets, Cell 112 (2003) 87–97. [14] E.P. Horn, A. Sputtek, T. Standl, B. Rudolf, P. Kühnl, J. Schulte am Esch, Transfusion of autologous, Hydroxyethyl starch-cryopreserved red blood cells, Anesth. Analg. 85 (1997) 739–745. [15] P. Jilma-Stohlawetz, Biotinylation: a nonradioactive method for labeling of blood components, Infus. Ther. Transfus. Med. 27 (2000) 296–300. [16] T. Kanias, J.P. Acker, Trehalose loading into red blood cells is accompanied with hemoglobin oxidation and membrane lipid peroxidation, Cryobiology 58 (2009) 232–239.
16
G.B. Quan et al. / Cryobiology 61 (2010) 10–16
[17] S. Khandelwal, R.K. Saxena, Age-dependent increase in green autofluorescence of blood erythrocytes, J. Biosci. 32 (2007) 1139–1145. [18] C.T. Knorpp, W.R. Merchant, P.W. Gikas, H.H. Spencer, N.W. Thompson, Hydroxyethyl starch: extracellular cryophyllactic agent for erythrocytes, Science 157 (1967) 1312–1313. [19] S.B. Leslie, E. Israeli, B. Lighthart, J.H. Crowe, L.M. Crowe, Trehalose and sucrose protect both membranes and proteins in intact bacteria during drying, Econ. Environ. Microbiol. 61 (1995) 3592–3597. [20] S.B. Li, Y.J. Jia, Z.P. Liu, Y.P. Zhang, Determining half life of mPEG-sheathed erythrocyte with FITC labeling, Bull. Acad. Mil. Med. Sci. 27 (2003) 367–369. [21] Y. Miyamoto, S. Suzuki, K. Nomura, S. Enosawa, Improvement of hepatocyte viability after cryopreservation by supplementation of long-chain oligosaccharide in the freezing medium in rats and humans, Cell Transplant. 15 (2006) 911–919. [22] A.E. Oliver, K. Jamil, J.H. Crowe, F. Tablin, Loading human mesenchymal stem cells with trehalose by fluid-phase endocytosis, Cell Preserv. Technol. 2 (2004) 35–49. [23] C. Pellerin-Mendes, L. Million, M. Marchand-Arvier, P. Labrude, C. Vigneron, In vitro study of the protective effect of trehalose and dextran during freezing of human red blood cells in liquid nitrogen, Cryobiology 35 (1997) 173–186. [24] G.B. Quan, Y. Han, M.X. Liu, F. Gao, Effects of pre-freeze incubation of human red blood cells with various sugars on postthaw recovery when using a dextran-rapid cooling protocol, Cryobiology (2009), doi:10.1016/j.cryobiol. 2009.08.001. [25] G.B. Quan, M.X. Liu, S.P. Ren, J.H. Zhang, Y. Han, Changes of phosphatidylserine distribution in human red blood cells during the process of loading sugars, Cryobiology 53 (2006) 107–118. [26] G.B. Quan, L. Zhang, Y. Guo, M.X. Liu, J.X. Wang, Y. Wang, B. Dong, A. Liu, J.G. Zhang, Y. Han, Intracellular sugars improve survival of human red blood cells cryopreserved at 80 °C in the presence of polyvinyl pyrrolidone and human serum albumin, Cryoletter 28 (2007) 95–108.
[27] S.P. Ren, E.P. Ma, X.Z. Liu, Y. Han, A. Liu, P. Jin, B. Dong, Labeling and survival studies on rabbit RBCs, Chin. J. Appl. Physiol. 19 (2003) 410–412. [28] A.P. Rinfret, Factors affecting the erythrocyte during rapid freezing and thawing, Ann. NY Acad. Sci. 85 (1960) 576–594. [29] A.W. Rowe, Cryopreservation of red cells by freezing and vitrificationsome recollections and predictions, Infus. Ther. Transfus. Med. 29 (2002) 25–30. [30] G.R. Satpathy, Z. Török, R. Bali, D.M. Dwyre, E. Little, N.J. Walker, F. Tablin, J.H. Crowe, N.M. Tsvetkova, Loading red blood cells with trehalose: a step towards biostabilization, Cryobiology 49 (2004) 123–136. [31] K.L. Scott, J. Lecak, J.P. Acker, Biopreservation of red blood cells: past, present, and future, Transfus. Med. Rev. 19 (2005) 127–142. [32] S.E. Slezak, P.K. Horan, Fluorescent in vivo tracking of hematopoietic cells. I. Technical considerations, Blood 74 (1989) 2172–2177. [33] A. Sputtek, G. Singbartl, R. Langer, W. Schleinzer, H.A. Henrich, P. Kühnl, Cryopreservation of red blood cells with the non-penetrating cryoprotectant hydroxyethyl starch, Cryo Letters 16 (1995) 283–288. [34] M. Tomita, T. Osada, I. Schiszler, Y. Tomita, M. Unekawa, H. Toriumi, N. Tanahashi, N. Suzuki, Automated method for tracking vast numbers of FITCLabeled RBCs in microvessels of rat Brain In Vivo using a high-speed confocal microscope system, Microcirculation 15 (2008) 63–174. [35] C.T. Wagner, M.L. Martowicz, S.A. Livesey, J. Connor, Biochemical stabilization enhances RBC recovery and stability following cryopreservation, Cryobiology 45 (2002) 153–166. [36] Y. Wang, P. Jin, M.X. Liu, A. Liu, G.B. Quan, W.B. Hu, Y. Guo, Y. Han, Comparison of two methods of washing frozen red blood cells in animal mode, J Clin. Transfus. Lab. Med. 18 (2006) 81–84. [37] R.J. Williams, The surface activity of PVP and other polymers and their antihemolytic capacity [J], Cryobiology 20 (1983) 521–526. [38] W.F. Wolkers, N.J. Walker, F. Tablin, J.H. Crowe, Human platelets loaded with trehalose survive freeze-drying, Cryobiology 42 (2001) 79–87.