Cryobiology 51 (2005) 152–164

www.elsevier.com/locate/ycryo

Improved preservation of human red blood cells

by lyophilization q
Ying Han *, Guo Bo Quan, Xiu Zhen Liu, En Pu Ma, An Liu, Peng Jin, Wei Cao
Institute of Transfusion Medicine, Beijing 100850, China

Received 10 January 2005; accepted 23 June 2005

Available online 10 August 2005

Abstract

The lyophilization of human red blood cells has important implications for blood transfusion in clinical medicine. In
this study, sugars, human serum albumin, polyvinylpyrrolidone, and dimethyl sulfoxide were used as protective
reagents for the lyophilization of red blood cells. Freezing temperature, shelf temperature, and the rehydration condi-
tions were optimized. The results showed that extracellular disaccharides, especially trehalose, did not increase the
recovery of hemoglobin. However, when the concentration of human serum albumin was higher than 25%, it had a
considerable protective effect on the recovery of lyophilized red blood cells; the cellular hemoglobin recovery was over
70%, which was significantly higher than that in the group without human serum albumin (P < 0.01). As the concen-
tration of polyvinylpyrrolidone was increased, the extent of vitrification also increased. But when the concentration of
polyvinylpyrrolidone was over 40%, the resulting concentration of free hemoglobin was over 1 g/L, which was signif-
icantly higher than that with 40% (P < 0.01). When lyophilization was carried out after freezing at different tempera-
tures, the recovery of cells and hemoglobin was 70–80% and there were no significant differences among the five groups.
When the shelf temperature was higher than 30
C, the samples were partly collapsed, but when the shelf temperature
was lower than 30
C, the recovery of cells in the 40 and 45
C groups was significantly higher than in the 30 and
35
C groups (P < 0.05). The recovery of cells and hemoglobin after lyophilization and rehydration in solutions con-
taining low concentrations of polymers was over 80%, which is significantly higher than the other groups (P < 0.01). In
addition, when the temperature was higher than 25
C, the concentration of free hemoglobin was significantly lower
than it was at 4
C (P < 0.01). In conclusion, our study showed the lyophilization of red blood cells is feasible. Disac-
charides have no protective effect on lyophilized cells when they are only extracellular and extensive vitrification may be
not beneficial. Although the recovery of cells after lyophilization and rehydration by our method was over 70%, the
ultrastructure of the cells may be compromised and some hemolysis does still exist. Further research is required.
 2005 Elsevier Inc. All rights reserved.

q
This work was supported by National Natural Science Foundation of China (Grant N50076046).
*
Corresponding author. Fax: +86 10 68285193.
E-mail address: hanying1001@yahoo.com.cn (Y. Han).

0011-2240/$ - see front matter  2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.cryobiol.2005.06.002
 Y. Han et al. / Cryobiology 51 (2005) 152–164
Keywords: Human; Red blood cells; Lyophilization; Rehydration; Freezing temperature; Shelf temperature; Rehydration solution
Conventional preservation methods for human
red blood cells include hypothermic storage (stor-
age at 4
C) and cryopreservation (storage at
80 or 196
C). These preservation methods
are widely used in current clinical practice, but
they cannot meet the current clinical demand for
blood supplies worldwide. The preservation time
at 4
C is short and contamination by microbes
can occur. Although cryopreservation greatly pro-
longs the storage time of red blood cells, this meth-
od needs an ultra-low-temperature refrigerator or
a liquid nitrogen container and transport of the
red blood cells is difficult and may even be com-
pletely impossible in some environmental condi-
tions [11,15]. For these reasons, a search has
begun for better methods for the preservation of
red blood cells.
Compared with conventional methods, lyophili-
zation would have many advantages, such as room
temperature storage, lower weight, and greater
convenience for transportation [15]. These advan-
tages make the lyophilization of red blood cells
more suitable for some particular environments,
such as war or disasters, by making it possible to
save lives, regardless of environmental conditions.
The idea that it may be possible to preserve mam-
malian cells by lyophilization comes from biologi-
cal adaptations that have occurred in nature [3].
Some organisms, including bacteria, yeast, and
tardigrades, can survive drying or even the com-
plete removal of water for long periods of time;
during this process, a high concentration of treha-
lose has been discovered in their bodies. In addi-
tion, some animals, for example, the tree frog,
can accumulate glucose in its red blood cells when
the temperature decreases [2]. Because sugar has
extensive protective effects on the freezing or
lyophilization of biological cells, people began to
use these sugars to preserve mammalian cells by
freezing or lyophilization [5,8,26,27]. Based on this
theory, human platelets have been successfully
lyophilized after being loaded with the disaccha-
ride trehalose; when rehydrated these cells are
capable of responding to physiological agonists
[26]. In this study, glucose was used as a membrane
153
protective reagent because glucose can be loaded
into red blood cells by simple or facilitated diffu-
sion, without special technology or treatment. In
addition, we found that dimethyl sulfoxide
(Me2SO) and human serum albumin have good
protective effects on lyophilized red blood cells.
Almost all studies reported before the 1980s
showed that no intact red blood cells were recov-
ered when lyophilization was the preservation
method [24]. Red blood cells have no nucleus
and the protection of the integrity of the plasma
membrane is a major challenge for lyophilization.
Beginning in the 1990s, research into the lyophili-
zation of red blood cells entered a new era [10].
Many researchers focused on this approach and
acquired some encouraging results [9,10,15,17–
20,24,25]. Goodrich et al. [9] reported the recovery
of metabolic functions in lyophilized red blood
cells, but the water content of the lyophilized red
blood cells was about 25–30% [24]. Although
many studies have been reported, researchers have
all faced the challenging problem of hemolysis.
Some studies showed a high concentration of free
hemoglobin in the supernatant: this may have
resulted from damage to the cell membrane during
lyophilization and/or during rehydration [15]. At
the end of 20th century, research into the lyophili-
zation of red blood cells faces many obstacles and
success was elusive [6].
Many researchers have focused primarily on the
selection of protective reagents and ignored the ef-
fects of freezing temperature and shelf temperature
on the recovery of lyophilized red blood cells. It is
sometimes thought that an ultra low temperature
causes protective solutions to vitrify and to arrest
the metabolism of cells fully, thus enabling cells
to be preserved for a long time. Following empir-
ical experiments, these investigators usually use
80 or 196
C as the freezing temperature
[15,19,24]. Rindler et al. [19] adopted hydroxyethyl
starch (HES) and maltose as protective reagents,
and systemically studied the effect of shelf temper-
ature on the lyophilization of red blood cells. They
showed that when the shelf temperature was
35
C, the hemolysis rate was lower than when
154 Y. Han et al. / Cryobiology 51 (2005) 152–164
the shelf temperature was either lower or higher
than 35
C [19]. This motivated us to evaluate
the freezing temperature, shelf temperature, and
rehydration process systematically in order to find
a better lyophilization protocol for red blood cells.
Lyophilization comprises three main steps:
freezing, sublimation in a vacuum, and rehydra-
tion. In this investigation, we evaluated the effects
on the survival of red blood cells of protective re-
agents (glucose, human serum albumin, polyvinyl-
pyrrolidone (PVP), dimethyl sulfoxide (Me2SO),
trehalose, and sucrose), freezing temperature, shelf
temperature, and rehydration solution. We found
that human serum albumin is beneficial for the
survival of red blood cells after lyophilization.
Optimized temperature and rehydration protocols
also helped to improve the preservation of red
blood cells by lyophilization.
Materials and methods
Buffer, pretreatment solution, protective solutions,
and rehydration solutions
The composition of the buffer and solutions are
given below. Unless otherwise stated, all chemicals
were analytical reagent grade and were obtained
from Sigma Chemical (St. Louis, MO, USA). Hu-
man serum albumin was purchased from Shanghai
RAAS Blood Products. 6% hydroxyethyl starch
(HES) (MW 25,000–40,000 Da) was purchased
from Leiyunshang Medicine (Changshu, Jiangsu
Province, China). All solutions were prepared in
double glass-distilled water.
Modified phosphate-buffered saline (mPBS)
(300 mOsm, pH 7.4) was used as an isotonic buff-
er: it contained 154 mM NaCl, 1.06 mM KH2PO4,
5.6 mM Na2HPO4, and 2.0 mM adenine. The pre-
treatment solution was mPBS containing 0.5 M
glucose. The protective solution contained: 0–
15% (W/V) trehalose or sucrose; 20–50% (W/V)
polyvinylpyrrolidone (PVP) (MW  40,000 Da);
5% (V/V) Me2SO; and 0–35% (W/V) human serum
albumin. Seven rehydration solutions were tested
in this study, including: (1) 6% HES, (2) mPBS
with 10% PVP (MW  40,000 Da), (3) 0.9% NaCl
with 5% carboxymethylcellulose sodium (CMS)
(MW  25,000 Da), (4) mPBS, (5) mPBS with
0.75 M glucose, (6) 5· mPBS (1560 mOsm, pH
7.4), and (7) 0.9% NaCl.
Pretreatment of red blood cells
Fresh whole blood was collected into CPDA-1
anticoagulant using healthy volunteers. The blood
was centrifuged in a MP4R low-temperature
centrifuge (International Equipment, Needham
Heights, Mass, USA) at 1000g for 10 min at 4
C
to remove the supernatant including white blood
cells, platelets, and plasma. The red blood cells
were washed three times with 0.9% NaCl under
the same conditions. The packed red blood cells
were equilibrated in mPBS containing 0.5 M glu-
cose for 4 h at 25
C after which the supernatant
was discarded. A mixture of 0.25 ml concentrated
red blood cells and 4.35 ml mPBS was used as
the control. In all studies, the volume of the lyoph-
ilized red blood cell suspension after rehydration
was 4.6 ml.
Treatment of red blood cells with disaccharides and
human serum albumin
The protective solutions containing 20% PVP,
0–15% trehalose or sucrose, and 25% human
serum albumin were used to evaluate the effect of
disaccharides on the recovery of pretreated red
blood cells following lyophilization and rehydra-
tion. In the case of human serum albumin, protec-
tive solutions containing 40% PVP, 5% Me2SO,
and 0–35% human serum albumin were used to
evaluate the effect of albumin on the survival of
red blood cells.
The red blood cells were mixed in a ratio of 1:3
with each protective solution and were placed in
10 ml plastic bottles (2.2 cm in diameter and
3.9 cm in height). Each bottle contained 1 ml of
the red cells/protective solution mixture. After
freezing at 80
C in an ultra-low-temperature
freezer (Sanyo, Guangzhou, China) for 2 h, the
samples were quickly transferred to a MINILYO
45 lyophilizer (MELITEK A/S, Hartvig Jensensvej
1 DK-4840 Nr. Alslev, Denmark) and lyophilized
for 24 h. During freezing, the cooling rate was
50–60
C/min from 25 to 80
C. The shelf tem-
 Y. Han et al. / Cryobiology 51 (2005) 152–164
perature and product temperature were measured
using the temperature probes of the lyophilizer.
During lyophilization, the shelf temperature was
35
C. The sample temperature was 25
C.
The condenser temperature was 56
C and the
vacuum pressure was 200 mbar. The rehydration
solution was 6% HES and the rehydration temper-
ature was 37
C. After lyophilization, the samples
were added to 6% HES and gently shaken until
fully rehydrated. The duration was approximately
10 s. The final volume was 4.6 ml.
Evaluation of the effect of vitrification on the
recovery of red blood cells
The PBS buffers composed of 20, 30, 40 or 50%
PVP, 5% Me2SO, and 25% human serum albumin
were used to study the effect of vitrification on the
recovery of lyophilized red blood cells. The extent
of vitrification of the protective solutions was mea-
sured by method of Rall and Zhu [16,28]. Briefly,
the protective solution was drawn into a 0.25 ml
plastic straw (IVM, lÕAigle, France) and sealed
by heat. The straw was directly plunged into liquid
nitrogen and the presence of white ice crystals in
the straw was checked. The cooling rate was about
2000
C/min. Then the straws were quickly
plunged into a 25
C water bath and the appear-
ance or absence of white ice crystals during thaw-
ing was noted. The warming rate was about
4000
C/min. The appearance of white ice crystal
indicates devitrification, while no white ice crystal
indicates vitrification. The other procedures of the
lyophilization and rehydration process were as
described above.
Lyophilization and rehydration of red blood cells at
different freezing and shelf temperatures
A solution containing 40% PVP, 25% human
serum albumin, and 5% Me2SO was used as the
protective solution in an evaluation of the effect
of freezing and shelf temperature on lyophilized
red blood cells. In the test of freezing temperature,
the samples were initially frozen at different tem-
peratures ( 20, 35, 45, 80 or 196
C) for
2 h and then quickly transferred to the lyophilizer
in which the shelf temperature was 35
C. They
155
were lyophilized for 24 h: all the other procedures
were as described above. Compared with the
lyophilization of red blood cells at different freez-
ing temperatures, the samples that were used to
measure the effect of shelf temperature on survival
were frozen at 80
C for 2 h and then lyophilized
at different shelf temperatures ( 20, 25, 30,
35, 40, and 45
C) for 24 h. During lyophili-
zation, the sample temperature was 10
C higher
than the shelf temperature. After lyophilization,
the samples were rehydrated with 6% HES and
the rehydration temperature was 37
C. The sam-
ples were washed with 6% HES, centrifuged at
1000g for 10 min at 4
C, 2 ml supernatant was re-
moved, and 2 ml of 6% HES was added and the
sample mixed. This washing procedure was repeat-
ed three times. The final colloid osmotic pressure
was 23.4 mOsm/kg.
The effect of rehydration solution on the survival of
red blood cells after lyophilization and rehydration
The protective solution was 40% PVP, 5%
Me2SO, and 25% human serum albumin. The
freezing temperature and the shelf temperature
were 80 and 35
C, respectively, and the vacu-
um pressure was 200 mbar. The lyophilization
time was 24 h. After lyophilization, samples were
rehydrated using the seven solutions described
above, all at 37
C. Then the effect of the temper-
ature of the 6% HES was tested using the same
protective solution and lyophilization method as
described above. The temperature of the 6% HES
was 4, 25 or 37
C. The rehydration procedure
was as described above.
Methods for evaluation of the residual moisture
content and cell recovery
The final moisture content of the lyophilized
samples was measured by proton-nuclear magnetic
resonance (PROTON-NMR) [12].
Owing to the high colloid osmotic pressure after
rehydration, a rehydration solution (for example,
6% HES) was used to dilute samples before count-
ing to calculate the recovery of red blood cells. The
volume of the control and of the lyophilized and
rehydrated red blood cells was 4.6 ml, respectively.
156 Y. Han et al. / Cryobiology 51 (2005) 152–164
The number of fresh and lyophilized and rehydrat-
ed red blood cells was measured by CELL-DYN
1200 hemocytometry (Abbott Diagnostics Divi-
sion, Abbott Laboratories, Abbott Park, IL,
USA). The formula used to quantify the recovery
of red blood cells was:
Concentration of red blood cells after lyophilization and rehydration
Concentration of control red blood cells
 100%.
The concentration of hemoglobin was calculated
by the Drabkin method [13] as follows: after rehy-
dration, samples were centrifuged at 2000g for
15 min to fully remove the supernatant and 6%
HES was added to a final volume of 4.6 ml. The
hemoglobin was converted to cyanmethemoglobin
using DrabkinÕs solution (0.61 mM K3Fe(CN)6,
0.77 mM KCN, and 9.43 mM Na2CO3) and the
absorbance of the cyanmethemoglobin was mea-
sured at 540 nm. The formula for the recovery of
hemoglobin was:
Concentration of hemoglobin after lyophilization and rehydration
Concentration of control hemoglobin
 100%.
Free hemoglobin was measured by the benzidine
method [1] in which free hemoglobin reacts with
benzidine and hydrogen peroxide in acidic condi-
tion and the absorbance is measured at 520 nm.
The free hemoglobin concentration was calculated
using the following formula:
OD520nm of the sample supernatant
 20 mg=100 ml.
OD520nm of standard hemoglobin
The morphology of the red cells after lyophiliza-
tion and rehydration was evaluated using a Philips
EM400T/Philips CM120 transmission electron
microscope (TEM) (Philips, Netherlands). The
protective solution contained 40% PVP, 5%
Me2SO, and 25% human serum albumin. The
freezing temperature and shelf temperature were
80 and 40
C, respectively. After lyophilization
for 24 h and rehydration with 6% HES at 37
C,
the red blood cells were mixed with 3% glutaralde-
hyde at 4
C for 2 h and then washed with sucrose
buffer (0.067 M PBS + 0.19 M sucrose) at 4
C.
The cells were then post-fixed in 1% osmium
tetroxide and washed in sucrose buffer for
15 min. After fixation, the red cells were dehydrat-
ed in ethanol and acetone according to the stan-
dard procedure and infiltrated with pure
embedding reagent for 24 h. Sections of thickness
60–70 nm were prepared and stained with uranyl
acetate for 10 min and lead citrate for 10 min in
the dark. Under the TEM, 100 cells were selected
at random to calculate the percentages of intact
cells, partly damaged cells, and ghosts,
respectively.
Statistical analysis
The recovery of red blood cells and hemoglobin
and the concentration of free hemoglobin were
analyzed by one-way ANOVA with LSD multiple
comparisons using SPSS software. All results with
a value of P < 0.05 or P < 0.01 were considered
statistically significant. The data in this study are
presented as means ± SD.
Results
The effect of trehalose or sucrose on the survival of
lyophilized red blood cells
Fig. 1 shows the effect of different concentra-
tions of sucrose or trehalose on the recovery of
hemoglobin after lyophilization. The moisture
content was measured by PROTON-NMR [12].
The moisture content after lyophilization for 24 h
was observed to increase as the disaccharide con-
centration increased, the final moisture contents
being 3.3, 3.7, 4.1, and 4.3%, respectively. The re-
sults show that disaccharides do not significantly
increase the recovery of hemoglobin. This phe-
nomenon may be due to raised osmotic pressure
and needs further research: many other studies
have reported that disaccharides are beneficial
for the preservation of mammalian cells whereas
this experiment suggested that the addition of
disaccharides to the extracellular medium, without
loading the cytoplasm, may not have any effect on
the survival of red blood cells after lyophilization.
From Fig. 1, when there were no disaccharides in
the protective solution, the recovery of hemoglo-
bin was about 80%, which was significantly higher
 Y. Han et al. / Cryobiology 51 (2005) 152–164 157

Fig. 1. The effect of disaccharides on the recovery of hemoglobin after lyophilization. Red blood cells were frozen and lyophilized
using trehalose and sucrose in concentrations ranging from 0 to 15%. Lyophilized red blood cells without disaccharides were used as
controls (0%). Hemoglobin concentration was analyzed using DrabkinÕs method [13]. The data points are expressed as means ± SD,
n = 10 in each group. *P < 0.01 vs controls by one-way ANOVA and LSD multiple comparison.

than in those groups that included trehalose
(P < 0.01). However, when the protective solutions
contained sucrose, the recovery of hemoglobin was
not significantly lower than it was in the group
without disaccharides (P > 0.05) (Fig. 1).
The effect of human serum albumin on the survival
of lyophilized red blood cells
The effect of human serum albumin on the sur-
vival of red blood cells was examined by adding
increasing concentrations of albumin to the pro-
tective solution. The results showed that the recov-
ery of red blood cells and hemoglobin was
dramatically improved with increasing concentra-
tion of human serum albumin (Fig. 2). The statis-
tical analysis showed that when the concentration
of human serum albumin in the protective solution
was 25%, the recovery of red blood cells and
hemoglobin after lyophilization reached 80%,
which was significantly higher than in the group
without human albumin (P < 0.01) but was not

Fig. 2. The effect of human serum albumin on the recovery of red blood cells and hemoglobin after lyophilization. Red blood cells were
lyophilized with 40% PVP and 5% Me2SO, plus human serum albumin in concentrations from 15 to 35%. The group without human
serum albumin was used as the control. The data points were expressed as means ± SD, n = 6 in each group. *P < 0.01 vs controls by
one-way ANOVA and LSD multiple comparison test.
158 Y. Han et al. / Cryobiology 51 (2005) 152–164

significantly different from the group containing
35% human albumin (P > 0.05). We also found
that the residual moisture content decreased with
the concentration of human serum albumin in
the protective solution.
The effect of vitrification of the lyophilization
solution on the survival of red blood cells
Adopting the method of Rall and Zhu [16,28],
we have evaluated the effect on cell recovery of vit-
rification of the protective solution. The results of
varying concentration of PVP are shown in Table
1. With increasing PVP concentration in the pro-
tective solution, the formation of white crystals
during freezing at 196
C decreased and there-
fore the occurrence of vitrification increased.
When the PVP concentration in the protective
solution was lower than 50%, there were visible
white crystals, but when the PVP concentration
reached 50%, there was no visible ice crystal for-
mation. As shown in Fig. 3, the concentration of
free hemoglobin in the supernatant after lyophili-
zation and rehydration first decreased as the extent
of vitrification increased and then increased. Also,
with increasing PVP concentration, the residual
moisture content of the samples, as measured by
PROTON-NMR [12], also increased, reaching
3.4, 3.7, 4.2, and 4.9%, respectively. When the
PVP concentration in the protective solution was
40%, although this did not reach the vitrifying con-
centration, the concentration of free hemoglobin
in the supernatant after lyophilization was signifi-
cantly lower than it was in the other experimental
groups (P < 0.01).

The effect of freezing temperature and shelf

Table 1
Vitrification and devitrification experiments with protective temperature on the survival of lyophilized red blood
solutions during the freezing and thawing of red blood cells cells
Protective solution Appearance of ice
crystals The effect of freezing temperature on recovery
Freezing Thawing of red blood cells and hemoglobin after lyophiliza-
tion is shown in Fig. 4. After lyophilization and
20% PVP + 7% Me2SO + human albumin + +
30% PVP + 7% Me2SO + human albumin + + rehydration, the recovery of red blood cells and
40% PVP + 7% Me2SO + human albumin + hemoglobin was all over 70–80% and there was
50% PVP + 7% Me2SO + human albumin no significant difference among the five groups
+, ice crystals seen; , no ice crystals seen during freezing or (P > 0.05). The effect of shelf temperature on the
thawing. recovery of red blood cells after lyophilization is

Fig. 3. The effect of vitrification on the concentration of free hemoglobin after lyophilization. Red blood cells were lyophilized using
different concentrations of PVP (20–50%), plus 5% Me2SO and 25% human albumin. The survival rate was determined by the free
hemoglobin release after lyophilization. The concentration of free hemoglobin was measured using the benzidine method [1]. The data
points were expressed as means ± SD, n = 6 in each group. *P < 0.01 vs 20, 30, and 50% groups by one-way ANOVA and LSD
multiple comparison.
 Y. Han et al. / Cryobiology 51 (2005) 152–164 159

Fig. 4. The effect of freezing temperature on recovery of red blood cells and hemoglobin after lyophilization. After freezing at different
temperatures ( 20 to 196
C), red blood cells were lyophilized at 35
C using 40% PVP, 5% Me2SO, and 25% human albumin as the
protective solution. The data points were expressed as means ± SD, n = 10 in each temperature group. There was no significant
differences among the five groups (P > 0.05) by one-way ANOVA and LSD multiple comparison.

Fig. 5. The effect of shelf temperature on the recovery of red blood cells and hemoglobin after lyophilization, washing, and return to
isotonic solution. After freezing at 80
C, the samples were lyophilized at different shelf temperatures ( 20 to 45
C) in six groups
using 40% PVP, 5% Me2SO, and 25% human albumin as the protective solution. The data points were expressed as means ± SD,
n = 10 in each group. When the shelf temperature was 20 and 25
C, the samples could not be fully lyophilized and after the process
there was a film of water on the surface of samples. There was no significant difference in the recovery of hemoglobin in each group.
However, there was a significant difference in the recovery of red blood cells by one-way ANOVA and LSD multiple comparison,
P < 0.05 vs 30 and 35
C.

shown in Fig. 5. During lyophilization, the sample than 30
C, the recovery of red blood cells
temperatures were 10
C higher than the shelf tem- in the 40 and 45
C groups was significantly
peratures. When the shelf temperature was lower higher than that in the 30 and 35
C groups
160 Y. Han et al. / Cryobiology 51 (2005) 152–164

study, but it was found that when the shelf temper-

ature was higher than 30
C, the surface of the
samples was covered with a film of water; the sam-
ples had not been properly dried after lyophiliza-
tion for 24 h using these shelf temperatures.

The effect of rehydration solution and temperature

on the survival of lyophilized red blood cells

Seven rehydration solutions were tested (Fig. 6).

Fig. 6. The effect of different rehydration solutions on the
recovery of red blood cells and hemoglobin after lyophilization.
The lyophilized red blood cells were rehydrated by one of seven
solutions as indicated in the figure. The data points were
expressed as means ± SD, n = 10 in each group. *P < 0.01 vs
mPBS, 0.75 M glucose in mPBS, 5· mPBS and 0.9% NaCl by
one-way ANOVA and LSD multiple comparison.
(P < 0.05), but there were no significant differences
between the recovery of hemoglobin in all four
groups (P > 0.05). Shelf temperatures higher than
30
C ( 20 and 25
C) were also tested in this
When the solutions contained a low concentration
of polymers (HES, PVP, and CMS), the recovery
of red blood cells and hemoglobin after rehydra-
tion was over 80% and significantly higher than
that in the other four groups that contained no
polymers (P < 0.01, Fig. 6). A low concentration
of glucose (0.75 M) did not have any protection ef-
fect during rehydration. In addition, we have
investigated the effect of rehydration temperature
on the survival of red blood cells using 6% HES
as the rehydration solution (Fig. 7). The results
showed when the temperature was higher than
25
C, the concentration of free hemoglobin after
rehydration was significantly lower than at 4
C
(P < 0.01). This result suggests that the rehydra-
tion temperature should be closer to physiological
temperature and that a low temperature is not
beneficial.

Fig. 7. The effect of temperature on the concentration of free hemoglobin during rehydration. After lyophilization, red blood cells were
rehydrated at different temperatures (4, 25, and 37
C) using 6% HES as the rehydration medium. The data points were expressed as
means ± SD, n = 5 in each group. *P < 0.01 vs 4
C group by one-way ANOVA and LSD multiple comparison.
 Y. Han et al. / Cryobiology 51 (2005) 152–164 161

Fig. 8. Transmission electron micrographs (TEM) of fresh and lyophilized red blood cells. Fresh red blood cells in panel (A) were a
normal disc shape and their hemoglobin was divided evenly in the cells. The lyophilized and rehydrated cells in panel (B) can be divided
into three types: A were normal cells; B were partially damaged cells; C were ghosts from which the hemoglobin had fully leaked.

Morphological observations of red blood cells after
lyophilization and rehydration
Red blood cells were lyophilized using 80
C
as the freezing temperature and 40
C as the shelf
temperature, and morphological changes in the
cells after rehydration were observed. The TEM
micrographs are shown in Fig. 8. The shape of
the red blood cells before lyophilization was con-
cave or discoid and their hemoglobin was distrib-
uted uniformly. Although the shape of most red
blood cells after lyophilization and rehydration
was normal and they retained an intact membrane,
some cells had an echinocytic shape and the hemo-
globin of some cells had partially leaked away
(Fig. 8). Visual counts of the percentage of intact
cells, partly damaged cells, and ghosts were 72,
20, and 8%, respectively.
Discussion
In this initial attempt to lyophilize human red
blood cells successfully, we have evaluated some
commonly used protective reagents including
sugars, polymers, and human serum albumin. In
addition, the freezing temperature, the shelf tem-
perature, and the rehydration solution were sys-
tematically evaluated for their effect on the
recovery of red blood cells after lyophilization.
Our ultimate purpose is to define an optimal
lyophilization procedure for these cells. In this
study, we have shown that when the protective
solution contained 40% PVP, 25% human serum
albumin, and 5% DMSO, the results of lyophiliza-
tion were improved with respect to other combina-
tions that were tested. Cell recovery reached 80%,
and the free hemoglobin in the supernatant was
less than 1 g/L. So far we have provided only this
in vitro index: other physiological indices, includ-
ing ATP and 2,3-DPG content and in vivo cell
survival in the circulation, need further research.
Our study showed that extracellular disaccha-
rides cannot significantly increase recovery and
may even decrease the recovery of hemoglobin.
This result was contrary to our expectation, based
on the findings of other researchers [26,27]. Wol-
kers has reported successful lyophilization of
human platelets using trehalose [26], but in this ap-
proach it is a great challenge to know how to load
the disaccharides into cytoplasm of cells [14]. Var-
ious methods have been used to load disaccharides
162 Y. Han et al. / Cryobiology 51 (2005) 152–164
into mammalian cells [5,8,14,21,26]. Endocytosis is
a fundamental process in eukaryotic cells and,
among numerous functions, this allows macromo-
lecular nutrient uptake. It has been claimed that
this is the mechanism of trehalose uptake in plate-
lets [26]. However, red blood cells have no endocy-
totic function [22,23] although trehalose has been
loaded into red blood cells by osmotic imbalance
and inducing a phospholipid phase transition
[21]. The effect of trehalose on the lyophilization
of red blood cells remains unknown. In this study,
we found that when disaccharides are present only
outside the red cell membrane, they have no
protective effect for lyophilized red blood cells.
However, because glucose can be loaded into red
blood cells easily, we incubated human red blood
cells in solutions containing a high concentration
of glucose.
CroweÕs publications show that polymers such
as HES have very high glass transition tempera-
tures (Tg) and they inhibit the fusion of liposomes
during lyophilization. However, they do not de-
press the liquid-crystalline to gel-phase transition
temperature (Tm) in dry phospholipids and do
not prevent leakage. On the other hand, glucose,
with a low Tg, does not vitrify during lyophiliza-
tion, but it depresses Tm in the dry lipid [4].
We also added a low concentration of Me2SO
to the protective solutions and found that it had
a protective effect (data not shown), but Me2SO
is potentially toxic and therefore the red blood
cells must be washed after rehydration in order
to remove it. The idea of adding Me2SO came
from work on the preservation of human platelets
[7]. In our study, some Me2SO evaporated in the
lyophilizer and when we added the rehydration
solution to the lyophilized cells, the final concen-
tration of Me2SO was further diluted, so its toxic-
ity was reduced.
Without using any disaccharide, we have
achieved a higher survival rate of lyophilized red
blood cells than others reporting their results in
the literature. One reason is the inclusion of hu-
man serum albumin in the protective solution.
Wolkers successfully lyophilized human platelets
using trehalose and human serum albumin [26].
When the concentration of human albumin was
more than 25%, the recovery of red blood cells
and hemoglobin was significantly higher than it
is in the group without human albumin (P < 0.01).
We also evaluated the influence of vitrification
on the lyophilization of red blood cells. The results
showed that increasing the PVP concentration in
the protective solution increased the extent of vit-
rification: when the PVP concentration was 50%,
no ice crystals were apparent during freezing, but
there was more hemolysis after lyophilization than
there was when the PVP concentration was 40%.
This result was counter-intuitive; we, and others,
have thought that vitrification is beneficial for
the preservation of mammalian cells but in fact it
may result in injury, perhaps due to the high
osmotic pressure and water absorptive properties
of PVP.
Lyophilization is a complicated process that in-
cludes cryopreservation and sublimation, as a re-
sult of which conventional cryopreservation
protocols cannot simply be extrapolated to lyoph-
ilization. The freezing temperature had a major ef-
fect on the survival of lyophilized red blood cells
and the effect of shelf temperature was also impor-
tant. Spieles et al. [24] and Rindler et al. [19] sys-
temically studied the effects of shelf temperature
on the lyophilization of red blood cells. Rindler
et al. [19] found that although sufficient drying
was achieved at 35
C, the hemolysis rate was
as high as 85% and there were no intact red blood
cells. Spieles et al. [24] showed that decreasing the
shelf temperature and prolonging the drying time
caused red blood cells to show a continuous
increase in hemolysis during the drying process.
Using our newly developed lyophilization meth-
od and protective solution we found that there was
no significant difference in the recovery of red
blood cells and hemoglobin between all the freez-
ing temperatures tested ( 20 to 196
C)
(P > 0.05). When the shelf temperature was higher
than 30
C, and the sample temperature was
higher than 20
C, the gross hemolysis of red
blood cells was most likely due to partial collapse.
We observed that the surface of the samples was
coated with water. When the shelf temperatures
were lower than 30
C, and with further decreas-
es in temperature, the recovery of red blood cells
and hemoglobin after lyophilization and washing
in isotonic medium was higher than 60%. From
 Y. Han et al. / Cryobiology 51 (2005) 152–164
this study, we can conclude that the best protective
effect is found only when the shelf temperature is
lower than 30
C.
Because the shelf temperature is higher than the
freezing temperature it follows that when frozen
samples are transferred from a low-temperature
medium to the lyophilizer, a significant ‘‘secondary
warming’’ occurs and this may injure the cells. One
may need to consider avoiding liquid nitrogen
temperatures when we select the freezing tempera-
ture for lyophilization. We think that the freezing
temperature should be close to the shelf tempera-
ture. But in this study, we found that even when
the freezing temperature was 20
C and the tem-
perature of samples was 10
C, lyophilization
was still relatively effective. We think that the sam-
ples can be fully solidified by a second freezing
after the samples have been transferred to the shelf
of the lyophilizer. So, during lyophilization the
selection of shelf temperature is very important.
When the rehydration solution contained a low
concentration of polymers, the recovery of red
blood cells and hemoglobin was over 80%, which
is significantly higher than in the groups without
polymer (P < 0.01). When the rehydration temper-
ature was too low, the hemolysis rate was in-
creased; temperatures higher than 25
C
improved the effectiveness. The mechanism of
these effects needs further research but we think
that polymers can decrease injury due to the
osmotic pressure gradients during rehydration.
After lyophilization, the Tg of the samples was rel-
atively high and if the red blood cells were rehy-
drated at a low temperature (for example, 4
C),
the temperature could be lower than the Tg of
the lyophilized red blood cells which could cause
devitrification and recrystallization, thereby dam-
aging the cells.
Although we have achieved some success in the
lyophilization of red blood cells by our present
method, the cell membrane is still damaged. After
rehydration, the recovery of cells was over 80%,
and most of the cells appeared normally concave,
but when washed and returned to the isotonic
state, the recovery of red blood cells and hemoglo-
bin was only about 60%, which suggests that
hemoglobin leakage occurs during washing. By
TEM, the recovery of intact cells after lyophiliza-
163
tion and rehydration was about 72%. This suggests
that the ultrastructure of the red blood cells may
be compromised. The reason for this is not clear
and needs further research.
PVP cannot be loaded into red blood cells and
the concentration in our protective solution was
too high, leading to a high osmotic pressure that
may have damaged the cells. We therefore used
permeable cryoprotective agents—Me2SO and glu-
cose [4]—rather than PVP and disaccharides
[26,27]. Although Me2SO can easily permeate
through the cell membrane and protect the inner
surface of the cell membrane it is toxic to cells. It
will be important to find a cryoprotectant to
replace Me2SO. Glucose can easily be loaded into
red blood cells without special methods and it may
stabilize the inner surface of the red cell
membrane.
In conclusion, our results demonstrated that
lyophilization of human red blood cells is feasible,
but there are still some questions to resolve. Our
results were better than those of Rindler et al.
[20] and Spieles et al. [24]. This study suggests that
extracellular disaccharides do not have any benefit
for the lyophilization of red blood cells. On the
other hand, human serum albumin plays an
important role in the protection of red blood cells.
Glucose and Me2SO also have a protective effect.
Future study will focus on the prevention of ultra-
structural damage to the cell membrane during the
rehydration process and the search for an efficient
and safe substitute as the membrane protectant. In
addition, the physiological function of the recon-
stituted cells in vivo, and other biochemical, enzy-
matic, and hematological indices need to be
studied in future research.
References
[1] N.Q. Bai, Blood Preservation, first ed., Shanghai Science
and Technology Publishing Company, Shanghai, 1984, pp.
252–253.
[2] S.P.J. Brooks, B.A. Dawson, D.B. Black, K.B. Storey,
Temperature regulation of glucose metabolism in red
blood cells of the freeze-tolerant wood frog, Cryobiology
39 (1999) 150–157.
[3] J.H. Crowe, L.M. Crowe, Preservation of mammalian
cells—learning from natureÕs trick, Nat. Biotechnol. 18
(2000) 145–146.
164 Y. Han et al. / Cryobiology 51 (2005) 152–164
[4] J.H. Crowe, A.E. Oliver, F.A. Hoekstra, L.M. Crowe,
Stabilization of dry membranes by mixtures of hydroxy-
ethyl starch and glucose: the role of vitrification, Cryobi-
ology 35 (1997) 20–30.
[5] A. Eroglu, M.J. Russo, R. Biaganski, A. Fowler, S. Cheley,
H. Bayley, M. Toner, Intracellular trehalose improves the
survival of cryopreserved mammalian cells, Nat. Biotech-
nol. 18 (2000) 168–171.
[6] F. Franks, Freeze-drying blood: reality or trick? Cryo
Letters 17 (1996) 1.
[7] D.Y. Gao, K. Neff, H.Y. Xiao, H. Matsubayashi, X.D.
Cui, P. Bonderman, D. Bonderman, K. Harvey, J.A.
McIntyre, J. Crister, C.C. Miraglia, T. Reid, Development
of optimal techniques for cryopreservation of human
platelets. I. Platelet activation during cold storage (at 22
and 8
C) and cryopreservation, Cryobiology 38 (1999)
225–235.
[8] A. Garciade Castro, A. Tunnacliffe, Intracellular treha-
lose improves osmotolerance but not desiccation
tolerance in mammalian cells, FEBS Lett. 487 (2000)
199–202.
[9] R.P. Goodrich, S.O. Sowemimo-Coker, C.R. Zerez, K.
Tanaka, Preservation of metabolic activity in lyophilized
human erythrocytes, Proc. Natl. Acad. Sci. USA 89 (1995)
967–971.
[10] R.P. Goodrich, C.M. Williams, R.S. Franco, M. Weiner,
Lyophilization of red blood cells, United States Patent 4,
874, 690 (1989)..
[11] Y. Han, E.P. Ma, A new strategy to preserve red blood
cells—lyophilization, Chin. J. Blood Transfus. 16 (2000)
210–211.
[12] G.Y. Li, R. Feng, J.G. Tian, H.B. Chen, Z.H. Du, Y. Han,
M.D. Liu, Content of water in freezed-dried red blood cells
determined by Proton-NMR, Chin. J. Magn. Reson. 16
(1999) 165–168.
[13] J.Z. Li, H.L. Wang, Z.C. Han, Xue Ye Shi Yan Xue, first
ed., Shanghai Science and Technology Publishing Compa-
ny, Shanghai, 1997, pp. 3–4.
[14] A.E. Oliver, K. Jamil, J.H. Crowe, F. Tablin, Loading
human mesenchymal stem cells with trehalose by fluid-
phase endocytosis, Cell Preservation Technol. 2 (2004) 35–
49.
[15] G.B. Quan, Y. Han, X.Z. Liu, Effect of vitrification state of
protective solutions on recovery of red blood cells after
lyophilization preservation, J. Exp. Hematol. 11 (2003)
308–311.
[16] W.F. Rall, Factors affecting the survival of mouse embryos
cryopreserved by vitrification, Cryobiology 24 (1987) 387–
402.
[17] V. Rindler, I. Heschel, G. Rau, Freeze-drying of red blood
cells: how useful are freeze/thaw experiments for optimi-
zation of the cooling rate?, Cryobiology 39 (1999) 228–235.
[18] V. Rindler, I. Heschel, G. Rau, Freeze-drying of red blood
cells: various strategies to improve cell survival, Infusions-
ther. Transfusionsmed. 5 (1997) 368.
[19] V. Rindler, S. Luneberger, P. Schwindke, Freeze-drying of
red blood cells at ultra-low temperatures, Cryobiology 38
(1999) 2–15.
[20] V. Rindler, B.T. Weaver, I. Heschel, Freezing-drying of red
blood cells: the influence of the cooling rates, Cryobiology
37 (1998) 430–431.
[21] 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.
[22] S.L. Schrier, D.T. Chiu, M. Yee, K. Sizer, B. Lubin,
Alteration of membrane phospholipid bilayer organization
in human erythrocytes during drug-induced endocytosis,
J. Clin. Invest. 72 (1983) 1698–1705.
[23] S.L. Schrier, I. Junga, J. Krueger, M. Johnson, Require-
ments of drug-induced endocytosis by intact human
erythrocytes, Blood Cells 4 (1978) 339–353.
[24] G. Spieles, I. Heschel, G. Rau, An attempt to recover
viable human red blood cells after freeze-drying, Cryo
Letter 17 (1996) 43–52.
[25] A.M. Tometsko, S. Dertinger, D. Torous, K. Tomesko,
Method and composition for lyophilizing red blood cells,
United States Patent 5,750, 330 (1998).
[26] W.F. Wolkers, N.J. Walker, F. Tablin, J.H. Crowe,
Human platelets loaded with trehalose survive freeze-
drying, Cryobiology 42 (2001) 79–87.
[27] W.F. Wolkers, N.J. Walker, Y. Tamari, F. Tablin, J.H.
Crowe, Towards a clinical application of freeze-dried
platelets, Cell Preservation Technol. 1 (2003) 175–188.
[28] S.E. Zhu, T. Sakurai, K. Edashige, T. Machida, M. Kasai,
Optimization of the procedures for the vitrification of
expanded mouse blastocysts in glycerol-based solutions,
J. Reprod. Dev. 40 (1994) 293–300.