Bone Marrow Transplantation, (1998) 21, 653–663
 1998 Stockton Press All rights reserved 0268–3369/98 $12.00
Clinical-scale human umbilical cord blood cell expansion in a novel
automated perfusion culture system
MR Koller, I Manchel, RJ Maher, KL Goltry, RD Armstrong and AK Smith
Aastrom Biosciences Inc., Ann Arbor, MI, USA
Summary:
Use of umbilical cord blood (CB) for stem cell transplantation has a number of advantages, but a major disadvantage is the relatively low cell number available.
Ex vivo cell expansion has been proposed to overcome
this limitation, and this study therefore evaluated the
use of perfusion culture systems for CB cell expansion.
CB was cryopreserved using standard methods and the
thawed unpurified cells were used to initiate small-scale
cultures supplemented with PIXY321, flt-3 ligand, and
erythropoietin in serum-containing medium. Twelve
days of culture resulted in the optimal output from most
CB samples. Frequent medium exchange led to significant increases in cell (93%), CFU-GM (82%) and LTCIC (350%) output as compared with unfed cultures. As
the inoculum density was increased from 7.5 ⴛ 104 per
cm2 to 6.0 ⴛ 105 per cm2, the output of cells, CFU-GM,
and LTC-IC increased. Cell and CFU-GM output
reached a plateau at 6.0 ⴛ 105 per cm2, whereas LTCIC output continued to increase up to 1.2 ⴛ 106 per cm2.
Because the increase in culture output did not increase
linearly with increasing inoculum density, expansion
ratios were greatest at 1.5 ⴛ 105 per cm2 for cells (6.4fold) and CFU-GM (192-fold). Despite the lack of adherent stroma, CB cultures expressed mRNA for many
growth factors (G-CSF, GM-CSF, IL-1, IL-6, LIF, KL,
FL, Tpo, TGF-␤, TNF-␣, and MIP-1␣) that were also
found in bone marrow (BM) cultures, with the exception
of IL-11 (found only in BM) and IL-3 (found in neither).
Culture output was remarkably consistent from 10 CB
samples (coefficient of variation 0.3) indicating that the
procedure is robust and reproducible. Two commercial
serum-free media were evaluated and found to support
only approximately 25% of the culture output as compared with serum-containing medium. Implementation
of optimal conditions in the clinical scale, automated cell
production system (CPS) showed that the process
scaled-up well, generating 1.7 ⴛ 107 CFU-GM (298-fold
expansion) from 1.2 ⴛ 108 thawed viable nucleated CB
cells (n = 3). The ability to generate ⬎107 CFU-GM
from ⬍15 ml of CB within this closed, automated system without the need for extensive cell manipulations
Correspondence: Dr AK Smith, Aastrom Biosciences, Inc, 24 Frank Lloyd
Wright Dr., Ann Arbor, MI 48105, USA
Received 11 June 1997; accepted 23 October 1997
will enable clinical studies to test the safety and efficacy
of expanded CB cells in the transplant setting.
Keywords: umbilical cord blood; hematopoiesis; cell
expansion; bioreactor; stem cells; serum-free medium
Human umbilical cord blood (CB) is known to contain
long-term repopulating hematopoietic stem cells, as evidenced by successful CB transplantation after high-dose
chemotherapy.1–3 CB transplants have been performed for
various malignant and non-malignant clinical indications
where no suitably matched donor was available. The use
of CB for stem cell transplantation has a number of advantages including widespread availability and reduced graftversus-host disease,4 as well as the potential to provide a
tumor-free cell source to replace current autologous cell
transplantation therapies. However, a major disadvantage is
the relatively low cell number available, which is believed
to contribute to delayed engraftment in CB transplantation
as compared with bone marrow (BM) or mobilized peripheral blood (mPB) transplantation. For example, a recent CB
study showed recovery to 500 ANC and 20 000 platelets
per mm3 required 22 and 56 days, respectively, even with
G-CSF support in the post-transplant setting.3 The time to
engraftment has been reported to be inversely related to the
cell dose given, suggesting that the availability of more
cells, or certain cell types, might lead to accelerated
engraftment.3,5 Further, because of the low cell number, the
adequacy of CB samples for adult transplantation is uncertain.6,7 Unlike BM or mPB, where the collection of more
cells is relatively straightforward, the number of cells available from a CB donor is fixed. Therefore, increasing the
number of stem and progenitor cells in a CB sample is of
great interest.
A number of laboratory studies have shown that CB cells
can be expanded ex vivo using methods similar to those
employed for the expansion of BM or mPB cells. In fact,
many studies have suggested that CB contains a more
primitive cell population, having more in vitro and in vivo
proliferative potential than BM or mPB cells,8–15 making
them a prime target for ex vivo expansion therapy or gene
therapy procedures. Most of the recent literature on CB cell
expansion has focused on the culture of CD34-enriched
cells or subsets thereof. However, perfusion bioreactor systems have been proven capable of expanding non-enriched
cells from BM16–18 and mPB,19 due to their ability to support high density accessory cell populations.17,20 In fact, in
the CB transplant setting, the potential advantages of
Umbilical cord blood cell expansion
MR Koller et al
654
CD34-enrichment prior to expansion, such as tumor purging and T cell depletion, are less relevant. The ability to
expand cells without CD34-enrichment would therefore
provide an advantage, because considerable processing
effort, cost and cell loss could be avoided.16,21
Although perfusion culture parameters for BM cell
expansion have been extensively optimized17,20,22,23 and
clinically tested,24–26 potential differences between CB and
BM cells warrant the study of several culture parameters
for CB cell expansion within the perfused culture environment. In addition, the likely clinical use of cryopreserved
and banked CB samples dictates the study of thawed cells
in such culture systems. A series of small-scale optimization studies was therefore undertaken, and led to the selection of conditions for clinical-scale expansion of CB cells
in a novel automated perfusion cell culture system, with
the ultimate goal of generating expanded CB cells under
cGMP requirements for the treatment of patients undergoing CB transplantation.
50 g for 10 min to deplete erythrocytes.28 The low density
cells were resuspended in LTC medium and were then
mixed with DMSO solution as above. Once processed by
one of these two methods, CB cell suspensions were placed
in 250 ml cryopreservation bags (Baxter) and subjected to
rate-controlled freezing (Forma Scientific, Marietta, OH,
USA). Starting at 4°C, the freezing chamber was cooled at
1°C per min to ⫺6°C, at 25°C per min to ⫺55°C to quickly
remove the latent heat of fusion, warmed at 15°C per min
to ⫺14°C to continue solid phase cooling, and then cooled
at 1°C per min to ⫺45°C and at 10°C per min to ⫺90°C,
after which bags were transferred into liquid N2 storage.
Cells were later thawed by placing in a 37°C water bath
for 1 min, and were then diluted 1:1 with a solution of 2.5%
human serum albumin (Michigan Dept of Public Health,
Lansing, MI, USA) and 5% dextran in saline as previously
described.4 Cells were then centrifuged and resuspended in
LTC medium.
Small-scale cell culture system
Materials and methods
Cell culture medium
Long-term culture (LTC) medium was prepared by supplementing Iscove’s modified Dulbecco’s medium (IMDM)
with 10% horse serum, 10% fetal bovine serum (FBS), 4
mm l-glutamine, 100 U/ml penicillin and 100 ␮g/ml streptomycin (all from Gibco, Grand Island, NY, USA), and 5
␮m hydrocortisone (Sigma, St Louis, MO, USA). In some
experiments, serum-free media were compared with the
serum-containing LTC medium. StemPro-34 (Gibco) was
supplemented with glutamine and antibiotics, and X-VIVO
15 (Biowhittaker, Walkersville, MD, USA) was supplemented with hydrocortisone, in accordance with the
manufacturers’ directions. All media for expansion culture
were supplemented with a previously optimized combination of 5 ng/ml PIXY321 (Immunex, Seattle, WA, USA),
0.1 U/ml erythropoietin (Epo; Amgen, Thousand Oaks, CA,
USA), and 25 ng/ml flt-3 ligand (FL; Immunex) as
described.17,27
Cell source and cryopreservation procedure
Human umbilical cords were obtained after vaginal delivery of full-term newborns (St Joseph Mercy Hospital, Ann
Arbor, MI, USA). CB was drawn from the clamped and
cannulated umbilical cord vein, and then placed into a polypropylene container containing ACD anticoagulant (Baxter,
Deerfield, IL, USA). Within 8 h at ambient temperture, the
samples were received and processed for cryopreservation.
A small sample of unmanipulated cells was first removed
for cell counts (Coulter, Hialeah, FL, USA) and assays. CB
was slowly mixed 1:5 with a solution of 50% DMSO
(Research Industries, Midvale, UT, USA), 200 ␮g/ml
DNase (Sigma), and 5% dextran (Sigma) in 0.9% saline
(Abbott, North Chicago, IL, USA) on ice as previously
described.4 Alternatively, Hespan (DuPont Merck Pharmaceuticals, Wilmington, DE, USA) was added to CB 1:5 to
give a 1.2% concentration, and cells were centrifuged at
Cells were cultured at various densities in 24 well plates
(Costar, Cambridge, MA, USA) in 0.6 ml of growth
medium using frequent manual medium exchange.17,20,29
Inoculum densities were adjusted by the measured cell
viability after thawing. All density measurements are
expressed per unit area (cm2) rather than per unit volume
in order to facilitate direct comparisons between the smallscale and clinical-scale systems. After inoculation, 50% of
the culture volume was replaced on various days to deliver
a rate of 0% (no feeding), 200% (fed on days 5, 8, 10 and
11), 250% (fed on days 4, 7, 9, 10 and 11), or 350% (fed
on days 3, 5, 7, 8, 9, 10 and 11) volume medium exchange
over 12 days of culture. Cells removed in the spent medium
at each feeding were resuspended in fresh medium and
returned to the wells. Cells were harvested after 1 to 14
days of culture using trypsinization as previously
described,17 and were then washed and resuspended in LTC
medium for counting and assays.
Clinical-scale automated cell production system (CPS)
The CPS (Aastrom Biosciences, Ann Arbor, MI, USA)
embodies a modular, closed-system process comprised of
a pre-sterilized, single-use disposable cell cassette operated
by automated, microprocessor-controlled instruments.30
The instrumentation components of the system include an
incubator unit and processor unit, along with a computerbased monitor. The cell cassette provides a functionally
closed, sterile environment in which cell production occurs.
The sterile fluid pathway contained in the cell cassette
includes the cell growth chamber, a container for medium
supply, a pump mechanism for delivery of medium to the
growth chamber, a container for the collection of waste
medium exiting the growth chamber, a container for the
collection of harvested cells, interconnected tubing between
the components, and sterile barrier elements to protect the
culture from contamination.
A dedicated small incubator controls the physical
environment and operations of each cell cassette. The incubator receives and self-engages the disposable cell cassette,
Umbilical cord blood cell expansion
MR Koller et al
and then controls the flow of medium to the growth
chamber, the temperature (4°C) of the growth medium supply compartment, the temperature (37°C) of the growth
chamber compartment, and the concentration (5% CO2 in
air) and flow rate (6 ml/min) of gases delivered to the gas
side of the culture chamber.
The processor performs the initial priming of the cell
cassette with growth medium and the controlled inoculation
and distribution of cells. The same unit also performs the
removal (harvest) of cells from the growth chamber at the
completion of the cell production process, transferring the
cells to a pre-attached harvest container (blood transfusion
bag) in a sterile manner. The heart of the processor is a
mechanical platform which provides the cell cassette with
automated tilting motions and valve actuations necessary
to achieve priming and even distribution of cells, and for
harvest of cells. The automated system design provides for
a high level of sterility assurance during each procedure
by replacing the need for operator manipulation of culture
devices within a laminar air flow hood, and is specifically
designed for production of patient-specific cells.
Flow cytometry analysis
Cells to be analyzed were washed and resuspended in phosphate buffered saline (PBS; Gibco) containing 1% bovine
serum albumin (BSA; Intergen, Purchase, NY, USA).
Mature erythrocytes were lysed with 0.16 m ammonium
chloride solution (Sigma). Tubes containing 106 cells in 0.5
ml were stained on ice with various combinations of fluorescently conjugated monoclonal antibodies. T and B lymphocytes were assessed by staining with FITC-Leu4 (antiCD3; Becton Dickinson, San Jose, CA, USA) and PELeu12 (anti-CD19; Becton Dickinson), respectively. Leukocytes and erythrocytes were assessed by staining with
PerCP-Hle-1 (anti-CD45; Becton-Dickinson) and FITCanti-glycophorin A (Dako, Carpinteria, CA, USA), respectively. Viability was determined by propidium iodide
(Molecular Probes, Eugene, OR, USA) staining. After 10
min, cells were washed and resuspended in 0.5 ml
PBS/BSA for analysis on a FACS Vantage flow cytometer
(Becton Dickinson).
Methylcellulose colony assays
Cells were inoculated in colony assay medium containing
0.9% methylcellulose (Sigma), 30% FBS, 1% BSA,
100 ␮m 2-mercaptoethanol (Sigma), 2 mm l-glutamine
(Gibco), 5 ng/ml PIXY321, 5 ng/ml G-CSF (Amgen), and
10 U/ml Epo. Cells were plated at 1500 to 20 000 per ml,
depending upon the expected clonogenicity. Cultures were
maintained for 14 days and were then scored as previously described.17
added to these stromal layers at four concentrations in
100 ␮l LTC medium per well (20 replicates each). Thawed
CB cells were plated at 2.5, 5, 10 and 20 thousand per well,
and expanded cells were plated at 7.5, 15, 30 and 60 thousand per well. The plates were then placed at 33°C in a
fully humidified atmosphere of 5% CO2 in air, and the cultures were fed weekly by replacing 100 ␮l LTC medium
per well. At week 5, cells were harvested from each well
as previously described.23 The cells from each well were
added directly to 0.25 ml of colony assay medium in nontissue culture treated 24-well plates (Falcon, Lincoln Park,
NJ, USA). After 14 days, wells were scored for colonies
as described above. For each sample, the number of LTCIC was determined through an iterative calculation procedure32 based on the maximum likelihood method.33
Analysis of endogenous growth factor mRNA by RT-PCR
The presence of transcripts encoding several hematopoietic
growth factors was assessed in both BM and CB expansion
cultures performed under identical conditions using an
inoculum density of 3 ⫻ 105 nucleated cells per cm2 with
the 250% volume medium exchange protocol. Total RNA
was isolated from parallel wells on different days of culture
using the RNeasy total RNA isolation kit (Qiagen, Santa
Clarita, CA, USA). cDNA was generated from 1 ␮g of total
RNA using 200 U of MMLV reverse transcriptase (RT;
Gibco) in 20 ␮l reactions containing 50 pmol oligo(dT),
1 mm each dNTP, 20 U RNasin RNase-inhibitor (Promega,
Madison, WI, USA), 10 mm Tris-HCl pH 8.3, 50 mm KCl,
and 5 mm MgCl2. Reactions were incubated for 10 min at
room temperature, followed by a 45 min incubation at 37°C
and 5 min at 95°C. PCR amplification was performed on
2 ␮l of each RT reaction using 1 U Taq DNA polymerase
in 25 ␮l reactions containing 200 ␮m dNTP, 25 pmol sense
primer, 25 pmol antisense primer, 10 mm Tris-HCl pH 8.3,
50 mm KCl, and 1.5 mM MgCl2. The sequences of sense
and antisense primers used to detect transcripts for each
hematopoietic growth factor are listed in Table 1. PCR
amplification was performed in a GeneAmp PCR System
(Perkin-Elmer, Foster City, CA, USA) with 35 cycles consisting of 30 s denaturation at 95°C, 30 s anneal at 62°C,
and 1 min extension at 72°C followed by a single 7 min
extension at 72°C. Reactions amplifying GM-CSF, IL-11,
and Tpo sequences underwent a second round of amplification for an additional 25 cycles using 1 ␮l of a 1:100
dilution of the initial reaction as starting material. PCR products were electrophoresed on a 2% agarose gel and visualized by ethidium bromide staining.
Results
Effect of inoculum density
Long-term culture-initiating cell (LTC-IC) assays
LTC-IC were determined by limiting dilution analysis
(LDA) on irradiated pre-formed stroma using a modification23 of a previously described technique.31 Briefly,
stroma was prepared as described17 in 96-well plates at 104
cells per well in 100 ␮l LTC medium. Test cells were then
In clinical practice, the number of CB cells available for
ex vivo expansion will likely vary from donor to donor.
Therefore, initial experiments evaluated the effect of inoculum density in 24-well plates with frequent medium
exchange over a 12 day culture. Seven cryopreserved and
thawed CB samples were used to initiate cultures with dif-
655
Umbilical cord blood cell expansion
MR Koller et al
656
Table 1
Oligonucleotide primer sequences for PCR amplification
Gene
G-CSF34
GM-CSF34
IL-1␤34
IL-334
IL-634
IL-1135
LIF36
KL37
FL38
Tpo35
TGF-␤39
TNF-␣40
MIP-1␣36
␤2-M41
Primer sequence (5⬘–3⬘)
S
AS
S
AS
S
AS
S
AS
S
AS
S
AS
S
AS
S
AS
S
AS
S
AS
S
AS
S
AS
S
AS
S
AS
CAC AGT GCA CTC TGG ACA GTG CAG GAA
CAT TCC CAG TTC TTC CAT CTG CTG CCA GAT
GAG CAT GTG AAT GCC ATC CAG GAG
CTC CTG GAC TGG CTC CCA GCA GTC AAA
CCT TGC CCT CTG GAT GGC GGC ATC
CAC GCA GGA CAG GTA CAG ATT CTT
GGA TCC TTG AAG ACA AGC TGG GTT AAC TGC TCT AAC
AAG CTT GAT ATG GAT TGG ATG TCG CGT GGG TGC
GGA TCC TCC TTC TCC ACA AGC GCC TTC GGT CCA
AAG CTT GTT CCT CAC TAC TCT CAA ATC TGT TCT G
GAC AGG GAA GGG TTA AAG GC
AGC TGT AGA GCT CCC AGT GC
AGC CCC CTC CCC ATC ACC CCT GTC AAC GCC
GAC CTC CTG CTA GAA GGC CTG GGC CAA CAC
GGA TGA CCT TGT GGA GTG CG
GCC CTT GTA AGA CTT GGC TG
TGT CAG CCC GAC TCC TCA ACC
AGT GCT CCA CAA GCA GCA GGT
CTC CTC CTG CTT GTG ACC TC
GGA CCC TCC TAC AAG CAT CA
CAA GTG GAC ATC AAC GGG TT
GCT CCA AAT GTA GGG GCA GG
GCC TGT AGC CCA TGT TGT AG
TTG GGA AGG TTG GAT GTT CG
TCC CGG CAG ATT CCA CAG AA
CAC AGT GTG GCT GTT TGG CA
CAT CCA GCG TAC TCC AAA GA
GAC AAG TCT GAA TGC TCC AC
Product size (bp)
416
343
444
279
425
338
551
416 (⫹ex6)
330 (⫺ex6)
235
437
297
468
303
165
Primer sequences specific for each message were used in PCR amplification reactions. The expected size (bp) of each PCR product is listed along with
the reference for each primer sequence.
S = sense; AS = antisense; ␤2-M = ␤2-microglobulin control message.
ferent inoculum densities of viable nucleated cells. As the
inoculum density was increased from 7.5 ⫻ 104 per cm2 to
6.0 ⫻ 105 per cm2, the culture output of cells, CFU-GM,
and LTC-IC increased (Figure 1). Cell and CFU-GM output
generally reached a plateau at 6.0 ⫻ 105 per cm2 and did
not significantly increase further as the inoculum density
was increased to 1.2 ⫻ 106 per cm2, whereas LTC-IC output appeared to increase further. Because the increase in
culture output did not increase linearly with the increasing
inoculum density, the cell and CFU-GM expansion ratios
were highest at the inoculum density of 1.5 ⫻ 105 per cm2
(Figure 2). Expansion ratios for LTC-IC are not shown due
to the non-linear, semi-quantitative nature of the assay
when performed on populations which contain different
accessory cell mixtures, such as inoculum and expanded
cells.42
Effect of feeding interval
In the optimization of BM MNC expansion, one of the most
critical interactions between culture parameters was shown
to be that between inoculum density and perfusion rate.20
Therefore, the effect of different perfusion rates was evaluated in cultures with different CB cell inoculum densities.
Frequent medium exchange led to a significant increase in
the output of cells (93%), CFU-GM (82%), and LTC-IC
(350%) as compared with unfed cultures (Figure 3).
However, the cultures were not particularly sensitive to
variations in feeding rate within the range of 200 to 350%
volume medium exchange. This was true even with culture
inoculum densities ranging from 1.5 ⫻ 105 per cm2 to 1.2
⫻ 106 per cm2, a result significantly different from that
found with BM.20
Effect of culture duration
Because CB cells are more prolific than BM cells, it is
possible that a shorter culture duration is warranted. Conversely, because CB cells are more primitive than BM cells,
it may be that a longer culture period is required to allow
them to reach an optimal growth stage. Therefore, the duration of expansion was evaluated as a culture variable. Replicate cultures were initiated at 3 ⫻ 105 per cm2 and fed
with the 250% exchange protocol with harvest at various
times thereafter for cell analysis. The kinetics of CB cell
expansion were measured in eight experiments (Figure 4).
The results show that cell numbers increased to day 14 in
every experiment. CFU-GM numbers increased to day 10
in all experiments, to day 12 in most experiments, and from
day 12 to day 14 in some experiments. Significant differences in LTC-IC numbers on different days of culture were
not apparent, suggesting that LTC-IC were not being
Umbilical cord blood cell expansion
MR Koller et al
12
a
657
a
9
Cell expansion
Cell output ×10–6
6
4
2
6
3
0
0
0.0
4.0
8.0
0.0
12
4.0
8.0
12
4.0
8.0
12
500
b
b
9
CFU-GM expansion
CFU-GM output ×10–4
400
6
3
300
200
100
0
0
0.0
4.0
8.0
12
0.0
Inoculum density ×10–5 (per cm2)
c
Figure 2 Effect of viable nucleated cell inoculum density on expansion
ratios. Thawed cord blood cells from seven donors were inoculated at
various densities (per cm2) in parallel triplicate cultures, and the expansion
of (a) cells and (b) CFU-GM were determined. The output from each
donor is shown with a different symbol that is used throughout Figures
1, 2 and 4.
40
LTC-IC output
30
20
Flow cytometric analysis of expansion
10
0
0.0
4.0
8.0
12
Inoculum density ×10–5 (per cm2)
Figure 1 Effect of viable nucleated cell inoculum density on cord blood
cell expansion culture output. Thawed cord blood cells from seven donors
were inoculated at various densities (per cm2) in parallel triplicate cultures,
and the output of (a) cells, (b) CFU-GM, and (c) LTC-IC were determined.
The output from each donor is shown with a different symbol that is used
throughout Figures 1, 2 and 4.
During the expansion period, the cellular composition of
the culture changed dramatically, as evidenced by the significant enrichment of CFU-GM progenitors. To gain
further insight into the changes in cellular composition,
flow cytometry was used to assess the appearance and disappearance of different cell populations during the culture
period. The percentage of viable cells increased quickly to
above 98% (Figure 5a), whereas the percentage of both T
and B lymphocytes declined exponentially (Figure 5b).
Leukocytes (CD45⫹) increased in frequency, whereas
erythroid cells (glyA⫹) declined in frequency.
Use of serum-free media
depleted as the cultures progressed. Interestingly, the kinetics of expansion were not significantly affected by
changes in inoculum density or feeding rate within the
ranges tested here (not shown). Because 12 days of culture
appeared to give the most optimal culture output from the
majority of donors, subsequent experiments focused on 12day culture duration.
We have previously found a number of serum-free media
to be inadequate for the support of perfused BM MNC cultures, particularly with respect to stromal development (M
Koller et al, manuscript in preparation). Because of the lack
of adherent stroma in CB cultures, the potential for
adequate performance from serum-free media was assessed.
Both StemPro-34 and X-VIVO 15 serum-free media were
compared with serum-containing medium. Although Stem-
Umbilical cord blood cell expansion
MR Koller et al
658
6
a
4
Cell output ×10–6
Cell output ×10–6
a
4
2
3
2
1
0
0
0
100
200
300
0
400
9
5
10
15
5
10
15
10
15
b
b
CFU-GM output ×10–4
CFU-GM output ×10–4
6
6
3
4
2
0
0
0
0
100
200
300
400
50
9
c
c
LTC-IC output
LTC-IC output
40
6
3
30
20
1200 K/cm2
10
300 K/cm2
150 K/cm2
0
0
0
0
100
200
300
400
Feeding (vol %)
Figure 3 Effect of perfusion rate on cord blood cell expansion culture
output. Thawed cord blood cells from three donors were inoculated at
various densities (per cm2) in parallel triplicate cultures and received
medium perfusion at different rates. The culture output of (a) cells, (b)
CFU-GM and (c) LTC-IC (⫾s.e.m.) are shown from one of the three
experiments.
5
Day
Figure 4 Kinetics of cord blood cell expansion. Thawed cord blood cells
from eight donors were inoculated at 3 ⫻ 105 cells per cm2 in parallel
triplicate cultures, and the output of (a) cells, (b) CFU-GM and (c) LTCIC were determined over time of culture. The output from each donor is
shown with a different symbol that is used throughout Figures 1, 2 and
4.
Endogenous growth factor mRNA in CB and BM cultures
Pro-34 signficantly outperformed X-VIVO 15, both
resulted in cell and CFU-GM output that was less than one
quarter of serum-containing control cultures (Figure 6).
LTC-IC output was also significantly lower at 30–40% of
the serum-containing control.
Accessory cells within perfused BM cultures are known to
produce a number of growth factors which likely improve
culture performance.16,17,43 Because the accessory cell composition of CB cultures is different from BM cultures (ie
lacking an adherent stroma), the presence of mRNA encoding a number of growth factors was assessed by RT-PCR
(Figure 7). Surprisingly, transcripts for most growth factors
tested (G-CSF, GM-CSF, IL-1␤, IL-6, LIF, KL, FL, Tpo,
Percent
β 2-M
100 bp ladder
MIP-1α
TGF-β
TNF- α
TPO
FL
LIF
KL
IL-11
IL-6
IL-3
75
GM-CSF
IL-1β
a
G-CSF
100
100 bp ladder
Umbilical cord blood cell expansion
MR Koller et al
Viability
50
CD45+
Bone
marrow
GlyA+
25
0
0
4
8
12
100
b
Cord
blood
CD3+
CD19+
Percent
10
Figure 7 Growth factor mRNA in bone marrow and cord blood expansion cultures. Total RNA was isolated from day 6 BM and CB cultures.
RT-PCR reactions were performed with primers specific for each message
(Table 1). PCR products were separated on a 2% agarose gel and visualized by ethidium bromide staining. The results are representative of
four donor BM cultures and three donor CB cultures.
1
0.1
0
4
8
12
Days
Figure 5 Flow cytometric analysis of cord blood cell expansion. The
percentage of (a) viable cells, leukocytes (CD45⫹) and erythroid cells
(GlyA⫹) is shown over time of culture from a representative experiment.
The decline of (b) T cells (CD3⫹) and B cells (CD19⫹) is also shown.
TGF-␤, TNF-␣, and MIP-1␣) were detected in both BM
and CB cultures throughout the 12-day culture period. In
contrast, IL-11 was detected only in BM cultures whereas
IL-3 was not detected in either system. Although KL message was detected in both CB and BM, the ratio of soluble
(416 bp) to membrane-bound (330 bp) message appeared
to be decreased in CB cultures, similar to the decreased
ratio observed from PB cells.44 As expected, negative control RT reactions performed in the absence of the RT
enzyme did not result in detectable products with any of
the primer pairs when used as a template in the PCR reactions (not shown).
1.0
Normalized culture output
StemPro-34
X-VIVO 15
0.8
0.6
0.4
0.2
0.0
Cell
CFU-GM
LTC-IC
Figure 6 Use of serum-free media for cord blood cell expansion culture.
Thawed cord blood cells from three donors were inoculated at 3 ⫻ 105
cells per cm2 in parallel triplicate cultures with serum-containing medium,
StemPro-34 and X-VIVO 15. Data within each of the three experiments
were normalized with respect to the serum-containing control (ie 1.0 represents the serum-containing control), and the mean (⫾s.e.m.) normalized
output of cells, CFU-GM, and LTC-IC is shown.
Automated clinical-scale CB cell expansion
The small-scale experimental results allowed the selection
of culture conditions for clinical-scale expansion of CB
cells in the automated Aastrom CPS. Thawed CB cells from
three donors were inoculated at an average of 1.2 ⫻ 108
viable nucleated cells (1.5 ⫻ 105 per cm2) per clinical-scale
device, and were cultured for 12 days with continuous perfusion in parallel with small-scale cultures. In the clinicalscale device, cells and CFU-GM were expanded an average
of 3.9-fold and 298-fold, respectively (Table 2). These data
are consistent with the parallel small-scale culture data,
indicating that the perfusion culture process was scaled-up
without loss of productivity. These experiments showed
that a relatively small inoculum of thawed CB cells could
be used to generate, on average, 4.4 ⫻ 108 cells, 1.7 ⫻
107 CFU-GM, and 4.6 ⫻ 103 LTC-IC in a single clinicalscale device.
Analysis of donor-to-donor variability
Over the course of these experiments, 10 CB samples were
processed and expanded under a standard set of small-scale
659
Umbilical cord blood cell expansion
MR Koller et al
Table 2
Clinical-scale expansion of CB cells in the automated Aastrom cell production system
Experiment
Cell expansion
CFU-GM expansion
CFU-GM number
LTC-IC number
Small-scale 1
Clinical-scale 1
5.7
3.4
187
180
43072
18616736
35
12647
Small-scale 2
Clinical-scale 2
4.5
2.6
518
338
41401
12160760
4
819
Small-scale 3
Clinical-scale 3
4.7
5.6
171
192
37775
19247451
ND
471
Whole CB was cryopreserved and thawed, and then expanded using the small-scale and automated clinical-scale culture systems in parallel. Experiments
1, 2 and 3 were initiated with inoculum densities of 1.6 ⫻ 105, 1.8 ⫻ 105, and 1.3 ⫻ 105 thawed viable nucleated cells per cm2, respectively.
ND = not determined.
culture conditions employing 3 ⫻ 105 cells per cm2 with
medium perfusion for 12 days (Figure 8). Culture output
was remarkably consistent from donor-to-donor, covering
a 0.3-log range in cell output (coefficient of variation, Cv
= 0.31), a 0.4-log range in CFU-GM output (Cv = 0.35),
and a 1.0-log range in LTC-IC output (Cv = 0.70). CFUGM expansion ratios were somewhat more variable covering the range from 48- to 309-fold, with a mean value of
126-fold.
Discussion
The utilization of CB holds great promise as an alternative
cell source for stem cell transplantation as well as other
new procedures such as ex vivo gene therapy. However, as
experience accumulates in the clinical setting, limitations
for the more widespread application of CB are emerging.
One limitation is the low number of cells available from
most cryopreserved CB samples. This low cell number may
contribute to delayed engraftment of neutrophils and platelets,3,5 and may limit the general use of CB transplantation
107
Cells
106
Culture output
105
104
CFU-GM
103
102
101
LTC-IC
1417
1410
1389
1364
1346
1342
1326
1322
1291
100
1281
660
Donor number
Figure 8 Donor-to-donor variability in cord blood cell expansion.
Thawed cord blood cells from 10 donors were inoculated at 3 ⫻ 105 cells
per cm2 in triplicate cultures perfused with 250% volume exchange for
12 days. The output of cells, CFU-GM and LTC-IC were remarkably consistent from donor-to-donor.
in adults and large pediatric patients.6,7 The use of ex vivo
cell expansion has been proposed to overcome these potential limitations in CB transplantation. In addition, the successful expansion of CB cells would likely allow the use
of smaller CB samples, thereby increasing the available
donor pool and facilitating the use of better HLA-matched
samples for many transplants.
For adult BM cell expansion, a perfusion culture technique has been previously developed and optimized,20,22,23,27 and the technique has since been implemented
at the clinical-scale in a novel, automated cell production
system (CPS).30 The CPS has been used to generate patientspecific expanded BM cells used to augment normal BM
transplantation,24,25 and as the sole means of hematopoietic
support following cytoablative chemotherapy in breast cancer patients.26 The current study addressed the optimization
of the perfusion culture technique for the expansion of CB
cells, and its subsequent clinical-scale implementation in
the CPS.
The expansion of cells from cryopreserved and thawed
CB samples was found to be enhanced through the use of
frequent medium exchange or perfusion. Interestingly, CB
cultures initiated over a wide range of inoculum densities
were relatively insensitive to variations in the medium perfusion rate, a result that contrasts with BM cultures.20 This
difference is likely due to the lack of stromal formation in
CB cultures, because much of the effect of medium perfusion has been shown to be stromal cell-mediated.17
Nevertheless, the high density of non-stromal accessory
cells present in the CB cultures did benefit from medium
perfusion. These accessory cells are likely important for the
endogenous production of many growth factors, as is the
case in perfusion BM cultures.17,43
Despite the lack of adherent stroma, CB cultures, like
BM cultures, expressed mRNA for a wide variety of
important growth factors. The most significant difference
between BM and CB cultures was the lack of IL-11 transcripts in CB cultures. This novel observation forms the
basis for further investigations regarding the effect of IL11 on CB expansion. Previous studies on CB cell growth
factor expression have shown that CD34⫹CD45RAloCD71lo
cell cultures express IL-1␤ mRNA,45 and that stimulated
MNC can be induced to express mRNA for a number of
growth factors including G-CSF, GM-CSF, M-CSF, MIP1␣ and TGF-␤, although transcript levels were often found
to be less than levels from stimulated adult PB MNC.46,47
Umbilical cord blood cell expansion
MR Koller et al
The issue of inoculum density is particularly important
for CB due to the potential limitation of cells in the clinical
setting. As inoculum density was increased, culture output
increased, although in a significantly non-linear fashion.
This resulted in lower cell and CFU-GM expansion ratios
at both the lowest and highest inoculum densities, although
total culture output was greatest at the higher inoculum densities. The maximization of expansion ratio vs the maximization of total culture output is an important balance to
strike, because the competing objectives of providing a sufficient cell dose, but from the smallest possible inoculum,
are both very compelling.
The clinical-scale CPS results indicate that ⬎107 CFUGM can be generated from a relatively small inoculum of
approximately 108 thawed viable nucleated cells which is
easily obtained from ⬍15 ml of CB, representing ⬍20%
of the cells available from a typical CB sample.3 This
expanded CFU-GM dose is approximately 10-fold greater
than that typically used in CB transplantation,3 and might
be beneficial for accelerating engraftment when given in an
augmentation setting along with unmanipulated cells from
the same sample. In order to implement this approach, it
would be necessary to aliquot a small fraction of each CB
sample prior to cryopreservation. Alternatively, the entire
CB sample could be expanded and used for transplant, as
has been done with adult BM.26 Although some recent
literature suggests that progenitor dose is not necessarily
predictive of short-term engraftment potential,48,49 other
studies have shown that cells at different stages of differentiation contribute to engraftment over different time scales
in vivo in both mice50–52 and humans.53 Further, other studies have shown that ex vivo expanded cells can accelerate
engraftment in mice.52,54–56 Ultimately, the clinical utility
of expanded CB cells can only be resolved through human
clinical trials designed to test their safety and efficacy. The
simplicity of operation and closed-system design of the
CPS offer significant advantages over manual culture techniques and will enable clinical studies to test the efficacy
of expanded CB cells in the transplant setting.
Although GVHD has not been a major limitation in CB
transplantation to date, some severe reactions have
occurred,2,3,5 and the incidence of GVHD might be
expected to increase as more adults are transplanted. Flow
cytometric analysis indicated a significant reduction in T
and B lymphocytes during the expansion procedure, and
this may be of potential benefit in the reduction of GVHD.
Despite the success of serum-free media formulations in
CD34-enriched cell culture,10 serum-free media are not
generally used in accessory cell-containing cultures, likely
due to the lack of support for BM stromal cell populations
(M Koller et al, manuscript in preparation). Unlike BM
cells, CB cells in vivo do not proliferate in association with
an adherent stromal cell population, and therefore might
not be expected to be as dependent on stroma as has been
shown for BM cells.17 In fact, CD34⫹CD38⫺ cells from
CB are reported to be less dependent on stroma than
CD34⫹CD38⫺ cells from BM.14 Nevertheless, the two
commercial serum-free media supported less than one quarter the level of culture output with serum-containing
medium, perhaps due to the support by serum of high den-
sity non-stromal accessory cell populations which were
present in these cultures.
Many studies have suggested that CB cells are more
primitive than BM or mPB cells by in vitro proliferative
measures.8,10–12 It has also been shown that the mean telomere length, a putative indicator of cellular proliferative
history, is longer in CB cell populations than in BM cells.13
Furthermore, long-term human cell engraftment in xenogeneic transplant models has been more successfully achieved with CB than with BM cell populations.15 Finally,
gene transduction protocols have been more successful with
CB cells than with BM or mPB cells.9 In the current study,
CFU-GM expansion (126-fold) obtained with CB cells at
an inoculum density of 3 ⫻ 105 per cm2 was significantly
greater than that obtained with adult BM cells (13-fold)
under identical culture conditions.27 These different lines
of evidence indicate that CB cell populations are significantly more primitive and have more in vitro and in vivo
proliferative potential than BM or mPB cells. The ability to
effectively manipulate and expand CB cells ex vivo offers a
significantly improved opportunity for the successful
implementation of a number of cell and gene therapy57
protocols.
The data presented indicate that the perfusion culture
technique was robust, yielding significant expansion from
all donor samples tested, and resulted in a predictable output dose of cells, CFU-GM and LTC-IC. The level of reproducibility with CB was better than that obtained with adult
BM MNC in terms of CFU-GM output (Cv 0.35 vs 0.50)
and LTC-IC output (Cv 0.70 vs 1.08), and this was considerably better than the greater variability observed from
BM CD34-enriched cell expansion.58 It is uncertain
whether donor-to-donor variability is due to genetic factors,
or to other factors such as donor health, history or aspiration technique. Because CB is collected rather easily without aspiration or mobilization, the effect of collection technique is minimized. Further, the effect of donor health or
history may also be minimized in these neonatal cell
sources. Therefore, the donor-to-donor variability in CB
cell expansion may give a more accurate assessment of the
role of genetic factors on this phenomenon.
Acknowledgements
We thank the Clinical Research Office and the Labor and Delivery
Unit at St Joseph Mercy Hospital (Ann Arbor, MI, USA) for collection of CB specimens, and Christopher Parrish and David A
Brott for flow cytometry analysis. This work was supported in part
by National Institutes of Health grant No. 1-R43-DK-53112-01.
References
1 Gluckman E, Broxmeyer HE, Auerbach AD et al. Hematopoietic reconstitution in a patient with Fanconi’s anemia by
means of umbilical-cord blood from an HLA-identical sibling.
New Engl J Med 1989; 321: 1174–1178.
2 Wagner JE, Kernan NA, Steinbuch M et al. Allogeneic sibling
umbilical cord blood transplantation in children with malignant and non-malignant disease. Lancet 1995; 346: 214–219.
3 Kurtzberg J, Laughlin M, Graham ML et al. Placental blood
661
Umbilical cord blood cell expansion
MR Koller et al
662
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
as a source of hematopoietic stem cells for transplantation into
unrelated recipients. New Engl J Med 1996; 335: 157–166.
Rubinstein P, Rosenfield RE, Adamson JW et al. Stored placental blood for unrelated bone marrow reconstitution. Blood
1993; 81: 1679–1690.
Rubinstein P, Carrier C, Adamson JW et al. New York Blood
Center’s program for unrelated placental/umbilical cord blood
(PCB) transplantation: 243 transplants in the first 3 years.
Blood 1996; 88: 142a (Abstr.).
Newton I, Charbord P, Schaal JP et al. Toward cord blood
banking: density separation and cryopreservation of cord
blood progenitors. Exp Hematol 1993; 21: 671–674.
Fritsch G, Stimpfl M, Buchinger P et al. Does cord blood
contain enough progenitor cells for transplantation? J Hematother 1994; 3: 291–298.
Koller MR, Bender JG, Miller WM et al. Reduced oxygen
tension increases hematopoiesis in long-term culture of human
stem and progenitor cells from cord blood and bone marrow.
Exp Hematol 1992; 20: 264–270.
Lu L, Xiao M, Clapp DW et al. High efficiency retroviralmediated gene transduction into single isolated immature and
replatable CD34(3⫹) hematopoietic stem/progenitor cells
from human umbilical cord blood. J Exp Med 1993; 178:
2089–2096.
Lansdorp PM, Dragowska W, Mayani H. Ontogeny-related
changes in proliferative potential of human hematopoietic
cells. J Exp Med 1993; 178: 787–791.
Hirao A, Kawano Y, Takaue Y et al. Engraftment potential
of peripheral and cord blood stem cells evaluated by a longterm culture system. Exp Hematol 1994; 22: 521–526.
Traycoff CM, Kosak ST, Grigsby S et al. Evaluation of ex
vivo expansion potential of cord blood and bone marrow hematopoietic progenitor cells using cell tracking and limiting
dilution analysis. Blood 1995; 85: 2059–2068.
Vaziri H, Dragowska W, Allsopp RC et al. Evidence for a
mitotic clock in human hematopoietic stem cells: loss of telomeric DNA with age. Proc Natl Acad Sci USA 1994; 91:
9857–9860.
Hao Q-L, Shah AJ, Thiemann FT et al. A functional comparison of CD34⫹CD38⫺ cells in cord blood and bone marrow.
Blood 1995; 86: 3745–3753.
Vormoor J, Lapidot T, Pflumio F et al. Immature human cord
blood progenitors engraft and proliferate to high levels in severe combined immunodeficient mice. Blood 1994; 83: 2489–
2497.
Koller MR, Manchel I, Newsom BS et al. Bioreactor expansion of human bone marrow: comparison of unprocessed, density-separated and CD34-enriched cells. J Hematother 1995;
4: 159–169.
Koller MR, Palsson MA, Manchel I et al. LTC-IC expansion
is dependent on frequent medium exchange combined with
stromal and other accessory cell effects. Blood 1995; 86:
1784–1793.
Zandstra PW, Eaves CJ, Piret JM. Expansion of hematopoietic
progenitor cell populations in stirred suspension bioreactors
of normal human bone marrow cells. Biotechnol 1994; 12:
909–914.
Sandstrom CE, Bender JG, Papoutsakis ET et al. Effects of
CD34⫹ cell selection and perfusion on ex vivo expansion of
peripheral blood mononuclear cells. Blood 1995; 86: 958–970.
Koller MR, Manchel I, Palsson MA et al. Different measures
of human hematopoietic cell culture performance are optimized under vastly different conditions. Biotechnol Bioeng
1996; 50: 505–513.
Laver J, Traycoff CM, Abdel-Mageed A et al. Effects of
CD34⫹ selection and T cell immunodepletion on cord blood
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
hematopoietic progenitors: relevance to stem cell transplantation. Exp Hematol 1995; 23: 1492–1496.
Palsson BO
/ , Paek S-H, Schwartz RM et al. Expansion of
human bone marrow progenitor cells in a high cell density
continuous perfusion system. Bio/Technology 1993; 11:
368–371.
Koller MR, Emerson SG, Palsson BO
/ . Large-scale expansion
of human stem and progenitor cells from bone marrow mononuclear cells in continuous perfusion culture. Blood 1993; 82:
378–384.
Silver SM, Adams PT, Hutchinson RJ et al. Phase I evaluation
of ex vivo expanded hematopoietic cells produced by perfusion
cultures in autologous bone marrow transplantation. Blood
1993; 82: 296a (Abstr.).
Champlin R, Mehra R, Gajewski J et al. Ex vivo expanded
progenitor cell transplantation in patients with breast cancer.
Blood 1995; 86: 295a (Abstr.).
Stiff PJ, Oldenberg D, Hsi E et al. Successful hematopoietic
engraftment following high-dose chemotherapy using only ex
vivo expanded bone marrow grown in Aastrom (stromalbased) bioreactors. Proc Am Soc Clin Oncol 1997; 16: 88a
(Abstr.).
Koller MR, Oxender M, Brott DA et al. flt-3 ligand is more
potent than c-kit ligand for the synergistic stimulation of ex
vivo hematopoietic cell expansion. J Hematother 1996; 5:
449–459.
Yurasov SV, Flasshove M, Moore MAS. Density enrichment
and characterization of hematopoietic progenitors and stem
cells from umbilical cord blood. Bone Marrow Transplant
1996; 17: 517–525.
Schwartz RM, Palsson BO
/ , Emerson SG. Rapid medium perfusion rate significantly increases the productivity and longevity of human bone marrow cultures. Proc Natl Acad Sci
USA 1991; 88: 6760–6764.
Armstrong RD, Koller MR, Maluta J et al. Clinical systems
for the production of cells and tissues for human therapy. In:
Oxender D, Post L (eds). Handbook of Experimental Pharmacology. Springer Verlag: Berlin, 1998 (in press).
Sutherland HJ, Lansdorp PM, Henkelman DH et al. Functional characterization of individual human hematopoietic
stem cells cultured at limiting dilution on supportive marrow
stromal layers. Proc Natl Acad Sci USA 1990; 87: 3584–3588.
de St Groth SF. The evaluation of limiting dilution assays. J
Immunol Meth 1982; 49: R11–R23.
Finney DJ. The estimation of bacterial densities from dilution
series. J Hygiene 1951; 49: 26–35.
Guba SC, Sartor CI, Gottschalk LR et al. Bone marrow stromal fibroblasts secrete interleukin-6 and granulocyte–macrophage colony-stimulating factor in the absence of inflammatory stimulation: demonstration by serum-free bioassay,
enzyme-linked immunosorbent assay, and reverse transcriptase polymerase chain reaction. Blood 1992; 80: 1190–
1198.
Rozen S, Skaletsky HJ. Primer3. Code available at
http://www-genome.wi.mit.edu/genomesoftware/other/primer
3.html. 1996.
Cluitmans FHM, Esendam BHJ, Landegent JE et al. Constitutive in vivo cytokine and hematopoietic growth factor
expression in the bone marrow and peripheral blood of healthy
individuals. Blood 1995; 85: 2038–2044.
Aye MT, Hashemi S, Leclair B et al. Expression of stem cell
factor and c-kit mRNA in cultured endothelial cells, monocytes and cloned human bone marrow stromal cells (CFU-RF).
Exp Hematol 1992; 20: 523–527.
Lisovsky M, Braun SE, Ge Y et al. Flt3-ligand production by
human bone marrow stromal cells. Leukemia 1996; 10:
1012–1018.
Umbilical cord blood cell expansion
MR Koller et al
39 Zhou L-J, Tedder TF. A distinct pattern of cytokine gene
expression by human CD83⫹ blood dendritic cells. Blood
1995; 86: 3295–3301.
40 Kaashoek JGJ, Mout R, Falkenburg JHF et al. Cytokine production by the bladder carcinoma cell line 5637: rapid analysis
of mRNA expression levels using a cDNA-PCR procedure.
Lymph Cyto Res 1991; 10: 231–235.
`
41 Le Bousse-Kerdiles M-C, Chevillard S, Charpentier A et al.
Differential expression of transforming growth factor-␤, basic
fibroblast growth factor, and their receptors in CD34⫹ hematopoietic progenitor cells from patients with myelofibrosis and
myeloid metaplasia. Blood 1996; 88: 4534–4546.
42 Koller MR, Maher RJ, Brott DA. LTC-IC assays performed
in the presence of accessory cells are non-linear and semiquantitative. Blood 1996; 88: 457a (Abstr.).
43 Koller MR, Bradley MS, Palsson BO
/ . Growth factor consumption and production in perfusion cultures of human bone
marrow correlates with specific cell production. Exp Hematol
1995; 23: 1275–1283.
44 Ramenghi U, Ruggieri L, Dianzani I et al. Human peripheral
blood granulocytes and myeloid leukemic cell lines express
both transcripts encoding for stem cell factor. Stem Cells 1994;
12: 521–526.
45 Watari K, Lansdorp PM, Dragowska W et al. Expression of
interleukin-1 beta gene in candidate human hematopoietic
stem cells. Blood 1994; 84: 36–43.
46 Lee SM, Knoppel E, van de Ven C et al. Transcriptional rates
of granulocyte–macrophage colony-stimulating factor, granulocyte colony-stimulating factor, interleukin-3, and macrophage colony-stimulating factor genes in activated cord vs
adult mononuclear cells: alteration in cytokine expression may
be secondary to post-transcriptional instability. Pediatr Res
1993; 34: 560–564.
47 Chang M, Suen Y, Lee SM et al. Transforming growth factor␤1, macrophage inflammatory protein-1␣, and interleukin-8
gene expression is lower in stimulated human neonatal compared with adult mononuclear cells. Blood 1994; 84: 118–124.
48 Uchida N, Aguila HL, Fleming WH et al. Rapid and sustained
hematopoietic recovery in lethally irradiated mice transplanted
with purified Thy-1.1loLin⫺Sca-1⫹ hematopoietic stem cells.
Blood 1994; 83: 3758–3779.
49 Hanania EG, Giles RE, Kavanigh J et al. Results of MDR-1
50
51
52
53
54
55
56
57
58
vector modification trial indicate that granulocyte–macrophage
colony-forming unit cells do not contribute to post-transplant
hematopoietic recovery following intensive systemic therapy.
Proc Natl Acad Sci USA 1996; 93: 15346–15351.
Ploemacher RE, van der Sluijs JP, van Beurden CAJ et al.
Use of limiting dilution type long-term marrow cultures in
frequency analysis of marrow-repopulating and spleen colonyforming hematopoietic stem cells in the mouse. Blood 1991;
78: 2527–2533.
Kiefer F, Wagner EF, Keller G. Fractionation of mouse bone
marrow by adherence separates primitive hematopoietic cells
from in vitro colony-forming cells and spleen colony-forming
cells. Blood 1991; 78: 2577–2582.
Szilvassy SJ, Weller KP, Chen B et al. Partially differentiated
ex vivo expanded cells accelerate hematologic recovery in
myeloablated mice transplanted with highly enriched longterm repopulating stem cells. Blood 1996; 88: 3642–3653.
Robertson MJ, Soiffer RJ, Freedman AS et al. Human bone
marrow depleted of CD33-positive cells mediates delayed but
durable reconstitution of hematopoiesis: clinical trial of MY9
monoclonal antibody-purged autografts for the treatment of
acute myeloid leukemia. Blood 1992; 79: 2229–2236.
DiGiusto DL, Lee R, Moon J et al. Hematopoietic potential of
cryopreserved and ex vivo manipulated umbilical cord blood
progenitor cells evaluated in vitro and in vivo. Blood 1996;
87: 1261–1271.
Muench MO, Firpo MT, Moore MAS. Bone marrow transplantation with interleukin-1 plus kit-ligand ex vivo expanded
bone marrow accelerates hematopoietic reconstitution in mice
without loss of stem cell lineage and proliferative potential.
Blood 1993; 81: 3463–3473.
Serrano F, Varas F, Bernad A et al. Accelerated and longterm hematopoietic engraftment in mice transplanted with ex
vivo expanded bone marrow. Bone Marrow Transplant 1994;
14: 855–862.
Kohn DB, Weinberg KI, Nolta JA et al. Engraftment of genemodified umbilical cord blood cells in neonates with adenosine deaminase deficiency. Nature Med 1995; 1: 1017–1023.
Koller MR, Manchel I, Brott DA et al. Donor-to-donor variability in the expansion potential of human bone marrow cells
is reduced by accessory cells but not by soluble growth factors. Exp Hematol 1996; 24: 1484–1493.
663