1. No category

From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
1996 87: 1754-1762
A bicistronic therapeutic retroviral vector enables sorting of
transduced CD34+ cells and corrects the enzyme deficiency in cells
from Gaucher patients
JA Medin, M Migita, R Pawliuk, S Jacobson, M Amiri, S Kluepfel-Stahl, RO Brady, RK Humphries
and S Karlsson
Updated information and services can be found at:
http://www.bloodjournal.org/content/87/5/1754.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
A Bicistronic Therapeutic Retroviral Vector Enables Sorting of Transduced
CD34+ Cells and Corrects the Enzyme Deficiency in Cells From
Gaucher Patients
By Jeffrey A. Medin, Makoto Migita, Robert Pawliuk, Steven Jacobson, Mike Amiri, Stefanie Kluepfel-Stahl,
Roscoe 0. Brady, R. Keith Humphries, and Stefan Karlsson
Corrective gene transfer for therapeutic intervention in metabolic and hematopoietic disorders has been hampered by
the relatively inefficient transduction of human hematopoietic stem cells. To overcome this, a bicistronic recombinant
retrovirus has been generated that delivers both a therapeutic glucocerebrosidase (GC) cDNA for the treatment of
Gaucher disease, and a small murine cell surface antigen
(heat-stable antigen [HSAI) as a selectable marker. An amphotropic retroviral-producing cell clone was created, and
filtered supernatant was used t o transduce NIH 3T3 cells.
Sorting of transduced cells by flow cytometry enabled separation into populations based on cell surface fluorescence
intensity derived from the expressed HSA. Significant increases in GC enzyme activity were seen for the transduced
and especially the transduced and sorted cells. Similarly,
increases in GC specific activity were seen in transduced and
sorted skin fibroblasts from a patient with Gaucher disease.
To streamline future transfer and sorting protocols for hematopoietic cells, transformed B-cell lines from Gaucher patients were created. Type l B cells were transduced and
sorted, and large increases in GC specific activity occurred
with concomitant increases in integrated retroviral copy
numbers. In addition, toward the goal of using this selectable approach for corrective gene transfer t o bone marrow stem cells, CD34+ cells were isolated from normal BM
donors, transduced, and sorted based on cell surface expression of HSA. Proviral DNA was detected in approximately
40% of clonogenic progenitor colonies derived from unsorted, transduced CD34+cells, demonstrating the high titer
of the vector. However, after sorting, 100% of the colonies
had the corrective GC cDNA, demonstrating the efficiency
of this selective system for human hematopoietic progenitors. It is expected that strategies based on this approach
will allow sorting of transduced cells of many types before
implantation of transduced cells t o animals or patients. This
vector system may also be used t o simplify manipulations
and studies on retroviral-mediated gene delivery in vitro and
for in vivo models.
This is a US government work. There are no restrictions on
its use.
R
vector that enables the transduced cell to be resistant to
toxic drugs. As an example, the gene encoding the neomycin
phosphotransferase protein has been used extensively to
track vector integration and expression and as a selective
agent based on acquired resistance to added G418.” The
gene for the multidrug resistance (MDR) efflux pump protein
has also been used to protect hematopoietic cells from challenge by various drugs.’2-’6Further, in a bicistronic plasmid
construct, a therapeutic glucocerebrosidase (GC) gene was
included with the coding sequence for the MDR efflux pump
protein in transfection studies,” allowing drug selection of
transfected cells, although no actual transduction of cells by
recombinant retrovirus was demonstrated in that case.
Another strategy to recognize a distinguishing characteristic of transduced cells has been to include a cell surface
antigen cDNA in the recombinant retrovirus construct to
allow noninvasive separation of cells due to the expressed
surface marker. The practicality of this idea was demonstrated as early as 1988,” although the overall effectiveness
of this approach for corrective gene transfer was limited,
because cell surface antigens initially chosen as markers
were sizeable and addition of a therapeutic gene would have
generated a large vector and potentially adversely affected
virus titer. Effects of insert size on recombinant retroviral
titer have been demonstrated.’9 cDNA for the MDR efflux
pump has also been used as a cell surface-selectable marker
in retroviral
as has the interleukin-2 (IL-2)
recepto?’ and the low-affinity nerve growth factor receptor’’
in cell-marking studies only. All of these genes are also
relatively large, thus possibly limiting their broad utility in
therapeutic vectors containing two genes.
An ideal cell surface marker would have low expression
on the target cells to keep background reduced, and would
react well with characterized antibodies to allow rapid and
efficient separation of transduced cells. Further, the cDNA
ETROVIRAL VECTORS have been used extensively
to transfer genes into a variety of cells to investigate
the possibility of gene therapy for many disorders. The widespread utility of recombinant retroviruses has been somewhat
constrained, though, by the fact that they do not transduce
nondividing cells. Transduction of human hematopoietic
stem cells has been relatively inefficient, probably because
most of these stem cells are quiescent and possibly because
specific receptors for current viral vectors are underrepresented on these cells. Approaches to a solution to the problem of infecting early hematopoietic progenitors have included studies of fine-tuning vector delivery techniques to
increase the percentage of transduced cells. Additions of
growth factors and stromal support to increase cell cycling
have shown increased calculated transduction efficiencies of
hematopoietic progenitor and stem cells.’-“’ Another approach is to select cells that have been transduced by the
recombinant retrovirus based on a distinguishing characteristic. One way to achieve this goal is to include a gene in the
From the Developmental and Metabolic Neurology Branch, Neuroimmunology Branch, National Institute of Neurological Disease
and Stroke, National Institutes of Health, Bethesda, MD; and the
Terry Fox Laboratory, British Columbia Cancer Agency, Vancouver,
Canada.
Submitted May 22, 1995; accepted October 6, 1995.
Address reprint requests to Jeffrey A. Medin, PhD, MMGS,
DMNB, NINDS, NIH, Bldg 10, Room 3004, 9000 Rockville Pike,
Bethesda, MD 20892.
The publication costs of this article were defrayed in purt by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
This is U US government work. There are no restrictions on ifs use.
OOO6-4~71/96/8705-0013$0.00/0
1754
Blood, Vol 87, No 5 (March l), 1996: pp 1754-1762
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
1755
SORTING OF THERAPEUTICALLY TRANSDUCED CELLS
sequence would be small enough that addition of it to the
therapeutic recombinant construct would not significantly
alter the titer of the virus produced. Good candidates that
meet these criteria are the human hematopoietic cell marker
CD24 and its murine homologue, the heat-stable antigen
(HSA). These small (-30 amino acids) surface molecules
are GPI-linked cell surface glycoproteins and have been isolated based on homology of signal peptide regions, although
the mature polypeptides of CD24 and HSA have little amino
acid identity. Additionally, antibodies to one do not crossreact with the other.24Indeed, a nontherapeutic bicistronic
retroviral construct containing the human CD24 antigen has
been used to successfully sort transduced and marked primitive murine hematopoietic cells, including those with longterm repopulating ability.25
This report describes the generation of bicistronic retroviral vectors that contain the murine HSA cDNA as a selectable marker and a therapeutic human cDNA for the treatment of metabolic deficiency in Gaucher disease. Gaucher
disease is a lysosomal storage disorder manifested in hematopoietic cells that is caused by a deficiency in the lysosomal
enzyme GC.26 It is classified as type I, 11, and 111 largely
depending on the degree of neurologic involvement.*’ It is
demonstrated here that this bicistronic vector system provides for selection and sorting of transduced hematopoietic
and nonhematopoietic cells from enzyme-deficient Gaucher
patients based on cell surface HSA expression, and further
that these cells are corrected for the enzymatic defect. Transfer and selection of transduced human CD34’ cells is also
demonstrated, and the presence of the vector in all progenitor
colonies examined was observed. This methodology will
greatly facilitate identification and isolation of transduced
cells for gene transfer and marking experiments in animals
and also for gene therapy protocols for humans.
MATERIALS AND METHODS
Plasmid constructions. The human GC cDNA (herein called
LGs) was liberated from plasmid pl-GC-BS (kindly supplied by D.
Kohn, Division of Research ImmunologyBMT, Childrens Hospital
Los Angeles, Los Angeles, CA) as an XhoI-XhoI fragment and
subcloned into plasmid pGlZxto produce plasmid pLGs. A ClalClal fragment containing the encephalomyocarditis virus intemal
ribosomal entry site (IRES) sequence (from plasmid pLNSPN;
kindly supplied by A.D. Miller, Fred Hutchinson Cancer Research
Center, Seattle, WA) and the murine HSA cDNA sequencez9 was
then added to pLGs to produce vector pLGsEH (Fig I). All cloning
--
I
MoMLV
LTR
I
IRES
H
MoMLV
LTR
1 KB
Fig 1. LGsEH retroviral construct used in these studies.
A
10000
-
8000
-
6000
-
4000
-
2000
-
W AM12RGsEH X6
W AM12AGsEH #8
0
B
100r
0
100
10’
102
103
104
Fig 2. AmlZlLGsEH producer cell line. (A) GC enzyme specific
activity results for clonal isolates no. 6 and 8 in comparison to wildtype Am12 levels. Bars (in all enzyme specific activity cases) represent calculated nmollhlmg protein for a minimum of 3 points with
standard error. (B)Flow cytometric scan of the AmlZ/LGsEH no. 8
clone ( 0 )in comparison to negative AM12 cells (leftmost profile)and
GP + AM12HSATKneo positive profile (behind solid peak). In all flow
cytometric profiles, t h e ordinate is the relative cell number and t h e
abscissa is t h e relative fluorescence intensity in log scale.
orientations and junctions were confirmed by restriction digests and
by DNA sequencing.
Cell lines and virus production. BOSC 23 cells were kindly
supplied by W.Pear (Rockefeller University, New York, NY) and
were grown as previously described.” These cells were transiently
transfected with purified pLGsEH using a calcium phosphate protocol, and ecotropic virus-containing supematant was collected. This
supematant was filtered through a 0.45-pm filter and repeatedly
administered to amphotropic virus-producing GP + envAM12 cells
(referred to as AM12) that were grown in Dulbecco’s Modified
Eagle Medium (Biofluids Inc. Rockville, MD) with 10% newbom
calf serum (GIBCO-BRL, Gaithersburg, MD) and penicillin/streptomycin at 37°C with 5% CO2. The multiply transduced AM12 cells
were then diluted to single-cell density and cloned and expanded.
GC activity and flow cytometry profiles for numerous candidate
clones in comparison to background controls were obtained (Fig 2).
Supematants from high-titer AM12LGsEH clones were filtered and
used in subsequent infections with added 8 pg/mL hexadimethrine
bromide (Polybrene; Sigma, St Louis, MO).
NIH 3T3 (tk-) cells were grown under standard conditions. Titers
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
MEDIN ET AL
1756
of supernatants from AM12LGsEH clones no. 6 and 8 on 3T3 cells
were estimated to be at least IO6 infectious U/mL as measured by
comparative Southern blots (see the results).
Helper virus detection. Marker rescue assays with multiple serial
dilutions were used to attempt to detect replication-competent virus.
No replication-competent virus was found at any dilution. Briefly,
3T3/GINa cells (kindly supplied by C. Dunbar) containing a replication-defective provirus harboring the neoRmarker were used as hosts
and were infected with five serial dilutions of filtered viral supematants (assayed in duplicate) and passaged for 2 weeks to allow potential virus to spread. Supernatants were then collected and used to
infect virgin 3T3 (tk-) cells with GP + AMI2HSATKneo cell
supernatant”’ serving as a positive control. Selection was then performed with (3418 (Sigma), and drug-resistant colonies were scored
2 weeks later (data not shown).
Gaucher patient cells. Skin fibroblasts obtained from Gaucher
patients under an approved National Institutes of Health protocol
were maintained in McCoy’s 5A media (BioWhittaker, Inc, Walkersville, MD) with 10% fetal bovine serum (FBS) at 37°C with 5%
COz.The cells were infected six times overnight with AM12LGsEH
pool supernatants in the presence of 4 pg/mL protamine sulfate
(Elkins-Sinn, Inc, Cherry Hill, NJ), cultured, and then tested against
normal and infection controls for GC expression and cell surface
expression of the murine HSA (see below).
Immortalized B-cell lines were made from peripheral blood obtained from Gaucher type 1, 11, and I11 patients. Briefly, 5 X 10’
peripheral blood lymphocytes were incubated with 0.5 mL crude
supematant from an EBV-producing marmoset cell line (895-8;
American Type Culture Collection [ATCC]. Rockville, MD) in a
total of 5.0 mL media containing RPMI 1640 (Biofluids), 15% FBS,
penicillin/streptomycin, and a 1: 10,000 dilution of OKT3 (ATCC)
ascites fluid. The cells were incubated for 2 to 3 weeks at 37°C with
5% CO,, after which clumping of cells was visible. Immortalized
cell lines were expanded to 75-cmZflasks in a total of 30 mL media
(as above) and maintained in RPMI 1640 (Biofluids) with 15%
FBS and penicillin/streptomycin at 37°C with 5% CO2 and passaged
frequently. The Gaucher type I B-cell line was infected with the
filtered AM12LGsEH no. 8 supematant in the presence of 4 pgl
mL protamine sulfate for a total of four overnight applications.
CD34’ cells. Bone marrow (BM) from normal donors was collected under a National Institutes of Health protocol, and mononuclear cells were isolated by centrifugation on LSM lymphocyte separation medium (Organon Teknika-Cappel, Durham, NC). CD34’
cells were then isolated by the Celprate LC system (Cell Pro, Inc,
Bothell, WA) and transduced four times with viral supematant in
the presence of stromal cells,’* protamine sulfate (4 pg/pL), IL-3
(20 ng/mL), IL-6 (50 ng/mL), and SCF (100 ng/mL). The cells were
then subjected to fluorescence-activated cell sorter (FACS) scan and
sorted (as below).
Cell labeling a n d j o w cytometry. To analyze HSA cell surface
antigen expression, cells were washed once in a solution (HFN) that
contained Hanks balanced salt solution (GIBCO-BRL), 2% FBS,
and 0.02% sodium azide (Sigma). The cells were resuspended in
HFN supplemented with 5% human serum and incubated for 30
minutes on ice. The cells were then incubated with purified and
biotinylated M1/69 antibody (a generous gift of S. Chapel) for 40
minutes on ice, washed twice with HFN, and then incubated with Rphycoerythrin (Jackson Immunoresearch Laboratories, West Grove,
PA) for 40 minutes on ice. After two further washes with HFN, the
latter in the presence of 1 pg/pL 7-amino actinomycin D (Sigma)
to distinguish dead cells, analysis by flow cytometry was performed
using a FACScan cell analyzer (Becton Dickinson, San Jose, CA).
The data are presented as cell number versus fluorescence intensity.
Cells for sorting were incubated as above, but 2 pg/pL propidium
iodide (Sigma) was added on the last wash to distinguish dead cells.
Cells were sorted on a FACStar+ (Becton Dickinson) machine
equipped with a 5 W argon and a 30 mW helium neon laser. Sorted
cells were collected into sterile Eppendorf vials in McCoy’s 5A
medium supplemented with 10% fetal calf serum.
Enzyme assays. Cells for enzyme assay were collected, washed
with phosphate-buffered saline, and resuspended in cold extraction
buffer containing 1 1 m m o l n potassium citrate/potassium phosphate
buffer, pH 6.0, 2 mg/mL Triton X-100 (Research Products Intemational Corp, Elk Grove Village, IL), 10 mg/mL sodium taurocholate
(Sigma), and 2 mmol/L AEBSF (a serine protease inhibitor; ICN
Biomedicals, Costa Mesa, CA). The cells were disrupted by sonication twice for 8 seconds on ice with a microtip, and cellular debris
was pelleted by centrifugation in a microcentrifuge for I5 minutes
at 4°C. Protein concentration of the cell extract was measured (BioRad Protein Assay; Bio-Rad Laboratories, Hercules, CA). and various amounts of extract from the different cell types were added to
a GC assay buffer (final volume, 250 pL) that contained 15 mmol/L
4-methylumbilliferyl-~-glucopyranoside(Sigma), 1.5 mg/mL Triton
X-100, 0.1 tnol/L potassium phosphate (pH 5.9), and 1.25 m g h L
sodium taurocholate. The reaction was incubated 30 minutes to 1
hour at 37°C and stopped with 750 pL 0.4-molL glycine/0.4-mol/
L sodium hydroxide. Fluorescence was measured using a Fluorometer, and GC activity was derived from quantitation against I O pg/
mL quinine at 30% transmission as a control.
Southern blot unalysis and polymeruse chain rerrcrion. DNA
was purified from cells using a genomic DNA isolation kit (Qiagen,
Chatsworth, CA) and the manufacturer’s protocols. Genomic DNA
was digested with restriction enzyme NheI (New England Biolabs.
Beverly, MA), which cuts in both the 5’ and 3’ retroviral LTRs.
After separation by electrophoresis through agarose gels, DNA fragments with copy number controls were transferred 10 Hybond N
membranes (Amersham, Arlington Heights, 1L) and probed for the
presence of human GC cDNA sequence and/or Moloney murine
leukemia virus sequence with appropriate ”P-labeled random-primer
DNA probes.
For analysis of transduced and control skin fibroblasts, a comparative polymerase chain reaction (PCR) technique was used to establish
the presence of the cDNA insert in the genomic DNA pool, because
the amount of recovered genomic DNA was limiting. Amplification
was performed with primers located within the GC coding sequence,
allowing differentiation of the band from the added cDNA sequence
and of the genomic GC DNA sequence. PCR products, after electrophoretic separation, were transferred to membranes and probed with
specific probes (as above).
In vitro clonogenic progenitor assays. Sorted and control cells
were plated as previously
and incubated for about 2
weeks. Colonies were scored and picked into a PCR preparation
buffer’” and analyzed by PCR as above.
RESULTS
The LGsEH vector. A recombinant retroviral vector that
encodes a bicistronic message containing the therapeutic human GC cDNA and the murine HSA as a selectable marker
in a MoMLV backbone is diagramed in Fig 1. This construct
uses a “short” form of the human GC cDNA,” which i\
different from constructs previously used in this laboratory.”
The IRES sequence from the EMCV was included to enhance tran~lation?~
of the downstream HSA cDNA in the
bicistronic construct. The respective orientation of the coding sequences was selected to possibly optimize the therapeutic GC enzyme expression. It follows that if protein expression from the downstream cDNA is reduced compared
with that derived from the upstream cDNA due to synthesis
of less than full-length mRNA, cells that have been selected
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
SORTING OF THERAPEUTICALLY TRANSDUCED CELLS
and sorted on the basis of highest HSA levels will likely
also have the highest expression of GC enzyme.
High-titer amphotropic virus-producing cell lines from
single-cell isolates were identified by the presence of increased GC enzyme expression and HSA surface expression
versus levels in untransduced AM 12 cells. Figure 2A shows
that in clonal isolates no. 6 and 8, GC enzyme specific activity (nanomoles per hour per milligram protein) was found to
be 9,504 5 550 and 8,270 5 72 I, respectively, representing a
more than 30-fold increase in enzyme specific activity over
background levels seen in nontransduced AMI 2 cells. Similarly, surface expression of the HSA antigen was greatly
increased over background (Fig 2B, shown for the no. 8
isolate), and these cells have the same degree of fluorescence
shift and narrow peak shape as the positive control GP +
AM 12HSATKneo viral producer cells that were originally
selected on the basis of resistance to (3418.
A Southern blot analysis was performed on genomic DNA
prepared from AM 12LGsEH isolate no. 8. A single proviral
band was seen at the expected size of 4.2 Kbp, which was
found to be present in approximately IO copies per cell (data
not shown). This AMl2/LGsEH isolate no. 8 was then used
as the amphotropic virus producer for subsequent studies.
Marker rescue assays demonstrated that this clone contained
no helper virus at any dilution.
Sorting of transduced NIH 3T3 cells. NIH 3T3 (tk-)
cells were transduced once with the AM I2LGsEH virus
supematant and then subjected to FACS. Figure 3A shows
flow cytometric analysis of cell surface HSA expression.
Cells were fractionated into categories based on HSA surface-staining intensity, expanded, and then analyzed for GC
enzyme activity and for the presence of viral DNA. Figure
3B shows results of the GC enzyme assay. Cells expanded
from the top 30% fluorescence intensity pool (MI) showed
the highest GC enzyme specific activity (2,918 5 78 nmol/
Wmg protein), representing a 4.6-fold increase over enzyme
activity observed in the unsorted pool and a full 11.6-fold
increase over background 3T3 GC enzyme activity levels.
Cells expanded from the M2 pool (top 3 1 % to 80% fluorescence intensity) also showed increased enzyme activity (twofold) over the unsorted pool.
A Southern blot analysis was performed on genomic DNA
samples isolated from the various FACS fractions (data not
shown). The intensity of the band in the transduced unsorted
fraction indicated at least one copy of vector per cell, on
average. Correll et aI2' found that neoR gene vector clones
with titers of 2 X IO6 to 1 X IOx CFU/mL produced 0.6 to
1.2 vector copies per cell in transduced 3T3 cells, and that
no signal at all was detected by Southern blot if the titer
was less than 5 X IO4 CFU/mL. This would indicate that
the clones presently described have high titer (probably > 10'
CFU/mL), although a direct measurement cannot be made,
since the vector lacks a drug-selectable marker. An increase
in band intensity in the lane loaded with genomic DNA
isolated from sorted cells was found to correlate directly with
the increased enzyme and cell surface antigen expression
observed. It is likely that the portion of cells with the greatest
fluorescence shift is derived from cells that took up and
integrated multiple copies of the recombinant retrovirus, and
not from a large number of cells in which extreme overexpression of the transgene occurred.
1757
A
100
-1
10'
I
10 2
103
104
3T3
W 3T3 with AMl2RGsEH
unsorted
El 3T3 with AM12RGsEH
M2 pool (3140%)
3T3 with AM12RGsEH
M1 pool (Top 30%)
Fig 3. Transduction of NIH 3T3 cells. (A) Flow cytometric analysis
of HSA cell surface expression on untransduced 3T3 (leftmost profile)
and transduced NIH 3T3 cells. M1, the HSA top 30% intensity pool;
M2, the next 31% to 80% pool. (BI GC enzyme specific activity in
comparison to untransduced NIH 3T3 cells.
Sorting of transduced Gaucher skinjbroblasts. Gaucher
skin fibroblasts were cultured and then transduced with the
AMI2LGsEH supernatant. Since the fibroblasts are relatively slow-growing, they were transduced multiple times
with viral supernatants to ensure that integration of the virus
occurred. Figure 4A shows FACS profiles of HSA expression on nontransduced and transduced cells. Transduced
cells with the highest fluorescence intensity were again isolated, expanded, and analyzed for GC enzyme expression.
Normal skin fibroblasts were found to have a 9.4-fold higher
enzyme activity level than those derived from this patient
(365 5 3 v 39 2 3 nmol/h/mg protein; data not shown).
Multiple infections of these Gaucher fibroblasts with the
recombinant retroviral supematant led to greatly increased GC
specific activity levels-in fact, the unsorted population of
transduced cells showed a more than 20-fold increase in enzyme levels versus the untransduced patient cell population
(data not shown). Indeed, enzyme activity was even 2.4-fold
greater than that seen in normal fibroblasts. Transduction of
these cells was extremely efficient and enzyme expression from
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
MEDIN ET AL
1758
A
8ol
0
100
10'
I
102
103
104
No added DNA
Normal Fibroblasts
Gaucher Fibroblasts
Gaucher Fibroblasts with
AMlaLGsEH unsorted
Gaucher Fibroblasts with
AMlZLGsEH sorted (Top 30%)
Gaucher Fibroblasts with
AMlWLGSEH sorted (31-70%)
Fig 4. Transductionof Gaucher patient fibroblasts. (A) Flow cytometric analysis of HSA cell surface expression on untransduced(leftmost profile) and transduced [rightmost profile) Gaucher patient skin
fibroblasts. M1, the HSA top 30% intensity pool; M2, the next 31%
to 7096 pool. [B) PCR analysis of genomic DNA isolated from pools
and conlrols. The upper band (-700 bp) corresponds to the amplified
genomic GC band, and the lower band (295 bp) results from the
integrated vector.
GC and the HSA marker in different cell types will allow
utilization of optimal conditions for infection and sorting.
Sorting of trcrnsdriccci Gnrtchcr type I U cclls. To demonstrate the feasibility of this selection-and-sorting approach
to correct the Gaucher deficiency in cells of hematopoietic
origin, B cell lines from type 1. 11, and 111 Gaucher patients
were created by EBV transformation of collected patient
lymphocytes. Figure 5 demonstrates observed GC enzyme
activity for each line. All cell lines from each type showed
GC enzyme specific activities that were approximately 30%
of normal. Calculated values of GC specific activity were
13.4 ? 0. I nmol/h/mg protein for a cell line derived from a
normal volunteer. and ranged from 3.5 to 4.0 nmolkdmg
protein for Gaucher B cells.
B cells from the Gaucher disease type I line were transduced with the AM12/LGsEH supernatant. A total of four
additions of the supernatant to the cultured cells in the presence of protamine sulfate were performed. Figure 6A shows
the resulting flow cytometric analysis of untransduced and
transduced cells. A large enhancement of the fluorescence
signal can be observed for a small number of cells. The
portion of cells marked MI (representing the top 9% in
HSA cell surface staining intensity) were then sorted and
expanded. Figure 6B shows calculated GC enzyme activity
of the transduced and sorted population of Gaucher disease
type I B cells in comparison to the unsorted population. An
enhancement of GC enzyme activity is seen in the unsorted
population versus the untransduced control, although this
enhancement still does not reach the levels of enzyme activity observed in the normal B cell line (Fig 6R). In direct
contrast, B cells selected for on the basis of HSA surface
antigen expression showed greatly enhanced GC enzyme
activity levels to more than fivefold over untransduced cell
levels. and even to levels more than twofold over those in
normal cells.
That the increased GC enzyme expression is a result of
increased retroviral copy number is shown by Southern blot
analysis in Fig 6C. Genomic DNA was prepared from the
.-c
.-5> si
n
this construct was greatly elevated. as demonstrated by the
observation that essentially the entire population of transduced
fibroblasts had a 2-log increase in observed fluorescence over
background levels (Fig 4A). Not surprisingly, differences in
enzyme activity among transduced unsorted and sorted populations were found to be relatively small (data not shown). To
ensure that the increased enzyme expression and HSA surface
fluorescence was derived from integrated retroviral copies, a
comparative PCR assay was used to examine genomic DNA
derived from these cells and from controls. Figure 4B shows
that DNA from transduced populations of cells has a band
characteristic of the integrated provirus, although, again. differences in band intensity between sorted and unsorted transduced
populations are small due to efficient transduction of these
cells. Further studies of this hicistronic retroviral construct with
respect to the kinetics of integration and protein expression of
H Gaucher Type I B Cell Line
0
'= a
q
0 2
E c
Normal B Cells
*Ol
0
10
Gaucher Type II B Cell Line
Gaucher Type 111 B Cell Line
$3
c2!
W O
n
U -
29 27 26
% of Normal Activity
Fig 5. GC enzyme specific activity of Gaucher type 1, II, and 111 B
cell lines.
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
1759
SORTING OF THERAPEUTICALLY TRANSDUCED CELLS
A
"1
Normal B cell line
Gaucher Type I B cell line
:m
MI
al5
€2
r2
20,
W O
0 Gaucher Type I B cell line
10
0 Gaucher Type I B cell line
$E
10'
100
C
102
103
104
with AMlWLGsEH unsorted
with AMlaLGsEH sorted
0
R
10 Copies
5 Copies
1 COPY
Type I B cell line
Type I with AMlULGsEH unsorted
Type I with AMlULGsEH sorted
Fig 6. Transduction of Gaucher type I B cell line. (AI Flow cytometric analysis of HSA cell surface expression on untransduced (leftmost profile) and transduced (rightmost profile) Gaucher type I B
cells. M1, the HSA-positive (top 9%) sorted population. (6) GC enzyme specific activity of sorted pools in comparison to normal B cells
and untransduced Gaucher type I 6 cells. IC1 Southern blot analysis
of sorted Gaucher type I 6 cells.
various transduced and/or sorted cell populations. After restriction enzyme digests and electrophoretic separation, the
DNA was probed for the presence of integrated provirus. A
reproducible band of faint intensity can be observed in unsorted. transduced Gaucher type I B cells (representing much
less than one copy per cell) in comparison to the copy number of controls. Figure 6C shows that sorting of B cells based
on cell surface HSA expression and expansion leads to an
increased retroviral copy number in the cells, up to one to
two copies per cell. which correlated well with the increased
enzyme expression seen. These data also demonstrate that
for this retroviral construct, even one to two copies per cell
of this recombinant retrovirus can lead to greatly increased
GC enzyme expression. higher than what is likely necessary
for corrective therapy.
FACS profiles and GC enzyme analyses were obtained
for sorted B-cell populations after substantial cell expansion.
Cell surface expression of the HSA surface marker and increased GC enzyme activity from the added cDNA were
maintained in cells in culture for an extended time (data not
shown).
Sorting of tmnshtced hiininn CD.34' cells. BM CD34'
cells were isolated from normal donors and transduced with
the AM I2LGsEH viral supernatant. Figure 7A demonstrates
the observed GC enzyme activity. Transduction with the
AM 12/LGsEH supernatant led to an increase in GC enzyme
specific activity over that seen in normal CD34' cells, and
to an increase even over that in CD34' cells transduced
with a control positive viral supernatant. LG4." Cell surface
expression of HSA was also demonstrated (Fig 7B). with
approximately IO% of the cells showing a fluorescence shift
into the MI-gated range. These cells were then sorted on
the basis of cell surface HSA expression, and clonogenic
colony assays were performed. Similar numbers of G/M and
BFU-E colonies were obtained for nontransduced. control
LG4 transduced, and AM 12/LGsEH transduced CD34' cells
(data not shown), indicating no gross perturbations in the
relative proportions of differentiated progeny cells due to
expression of the HSA.
Colonies derived from infected sorted and control populations were analyzed by PCR for the presence of proviral
sequence. AM 12LGsEH-infected cells had the provirus integrated in 38%- to SO% of the colonies (Table I ) . This is
similar to the infection frequency observed using the control
LG4 supernatant (40% to 59%; Table I ) and indicates that
the titer of the AMI2/LGsEH is relatively high on human
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
MEDIN ET AL
1760
CD34+
M1
I
W CD34+ with AM12LG4
EJ CD34+ with AM12LGsEH
0
100
C
HSA
sorted colonies (#1-17)
10'
I
102
103
104
Control
uninfected colonies (#18-21)
P
-700 bp
I;
295 bp
Fig 7. Transduction of BM CD34' cells. (A) GC enzyme specific activity of transduced CD34' cells. (B) Flow cytometric analysis of HSA cell
surface expression on untransduced (0)
and transduced (rightmost line) human BM CD34' cells. M1, the top-intensity HSA sorted pool. (C)
PCR analysis of genomic DNA isolated from HSA-sorted transduced and control colonies. The upper band (-700 bp) corresponds to the
amplified genomic GC band, and the lower band (295 bp) results from the integrated vector.
hematopoietic progenitor cells. AM I2LGsEH-transduced
CD34' cells were then sorted by FACS on the basis of cell
surface expression of introduced HSA, and positive cells
were plated into methylcellulose colony cultures. All 17 colonies picked (Table I and Fig 7C) were positive for the
presence of the proviral GC sequence.
DISCUSSION
Recombinant retroviral vectors containing the GC gene
and thz murine HSA expressed in a coordinated fashion
have been used successfully to select transduced cells from
Gaucher patients by FACS. resulting in metabolic correction
of the enzyme deficiency. Infection of CD34.' human BM
cells was also demonstrated. Only a single round of immunofluorescence and flow cytometric sorting was necessary in
all cases to produce a large increase in GC enzyme specific
activity of the cells (derived from the added therapeutic GC
enzyme) in the bicistronic retroviral construct. Interestingly,
Table 1. Results of PCR Analysis for Proviral DNA Sequence
in CD34' Infected BM Colonies
Unsorted
LG4
LGsEH
Assay 1
Assay 2
HSA-Sorted
23/39
15/40
8/19
1/21
ND
17/17
Abbreviation: ND, not determined.
Only two colonies could be analyzed, due to technical difficulties.
the large increases in GC enzyme specific activity (>30fold for the AM12 vector-producing cell line, for example)
were not accompanied by any changes in growth characteristics or gross morphology of any transduced cells in companson to untransduced controls. A good correlation was also
demonstrated between virus copy number, GC enzyme levels, and HSA surface antigen expression in sorted cells.
Bicistronic retroviral vector constructs containing the GC
gene and the MDRl -selectable gene have been generated
and used to correct the metabolic deficiency in Gaucher
fibroblasts following plasmid transfection," but transduction
experiments by virus were not reported. The MDRl gene
can be selected for in vitro and in vivo by cytotoxic drugs
and can also be selected by immunoaffinity techniques, since
the MDR efflux pump protein is expressed on the cell surface. Experiments in mice have demonstrated that the MDR
gene can protect hematopoietic cells from high concentrations of cytotoxic drugs and may therefore be useful in cancer chemotherapy."." However, in consideration of constructs designed for broader utility, the MDRl gene is large
(-4,000 bp), and when an additional therapeutic gene is
added to the retroviral construct to correct a deficiency, the
recombinant vector titer may be low and transduction of
target cells. especially hematopoietic stem cells. may be very
inefficient. Recently. bicistronic vectors containing the
MDRI gene were reported to have titers in the IO' CFU/mL
range.",'" In addition, due to the size of the MDRl protein.
higher translational efficiency or increased transcription of
the cDNA may be required to generate levels of MDRl
protein on the surface of cells high enough to allow func-
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
1761
SORTING OF THERAPEUTICALLY TRANSDUCED CELLS
tional sorting with present antibodie~.~’
In contrast, the vector system described herein, using the small (243-bp) murine
HSA, generates high titers that are comparable to those obtained previously for LG4 in this laboratory and used to
transduce murine hematopoietic stem cells.”
The efficient identification and separation scheme described here can greatly facilitate studies on murine hematopoiesis and overall gene transfer technology. Using the HSA
as an expression marker or as a labeling tag, long-term hematopoietic cell-repopulation and virus-construction studies can
be expedited. Among other examples, research on long-term
hematopoietic expression from various retroviral LTRs and
also on the potential effect of retroviral promoter shutdown” can be made much easier, since detection of the
expressed surface antigen is relatively simple and only a
limited number of cells are required for the analysis. Studies
on recombinant retrovirus construction, to generate integration site-independent expression and also exploration of
regulatable promoters to modulate expression, can also benefit from the ease of detection and tracking of transduced cells
and from the limited quantitation of the expressed marker
that is possible. Furthermore, studies to examine why certain
cell types have much higher degrees of infectivity (for example, compare FACS profiles for patient fibroblasts and patient
B-cell lines in this study) than other cells can also benefit
from this rapid and convenient approach.
Since now the murine HSA and previously the human
CD24*’ have both been used to sort retrovirally transduced
cells, greater flexibility is provided for studies in murine
models and for possible future transduction into human cells.
Since no homology exists in the human for the mature peptide, it may be possible that constructs containing HSA can
be used to transduce and sort patient hematopoietic stem
cells before transplantation. Further, experiments using hematopoietic stem and progenitor cells from mice and humans
are needed to demonstrate the feasibility of using this vector
system in clinical gene therapy protocols.
It is not clear whether adverse effects from generalized
expression of HSA in most hematopoietic lineages can be
expected in mice or in humans in vivo. In transgenic mice,
significant overexpression of HSA has recently been shown
to reduce the overall numbers of thymocytes.” In preliminary experiments presented here, expression of HSA on primary human hematopoietic cells does not seem to alter the
number or type of observed hematopoietic progenitor colonies in vitro. Similar observations have been made in a study
by Conneally et al4’ using a HSA vector that also contains
a neoR gene. Therefore, HSA expression on human hematopoietic progenitors does not seem to cause abnormal differentiation or maturation of their progeny cells in vitro.
In humans, an immune response could be generated, and
animal experiments, particularly in primates or other large
outbred animals, are necessary to explore this potential complication. The HSA was originally isolated from a hematopoietic cell line,*’ and the cDNA encodes a glycosylated
protein with a core of only 30 amino acids that is expressed
on developing cells of earlier hematopoietic lineages and on
some neural cells. Varied roles for HSA have been proposed,
including that of a costimulator for antigen-presenting
~ e l l s ? ‘ ~a ~modulator
*
of cell-cell a d h e ~ i o n , ~and
? . ~a~factor
in the protection of cells from self-complement depletion.45
As inducible and regulatable promoter systems become more
it may be possible to induce
characterized and
HSA expression only during a short time, long enough to
allow the efficient sorting the system affords, and then to
turn off long-term expression. This would circumvent some
of the potential immunologic hazards of using a murine antigen in human systems.
ACKNOWLEDGMENT
The authors thank Nick Flerlage for excellent technical assistance,
Drs Jacob Reiser and Raphael Schiffmann for critical reading of the
manuscript, and Dr Jim Nagle for sequencing the plasmid constructs.
REFERENCES
1. Hughes PF, Eaves CJ, Hogge DE, Humphries RK: High-efficiency gene transfer to human hematopoietic cells maintained in
long-term marrow culture. Blood 74: 1915, 1985
2. Bodine DM, Karlsson S, Nienhuis AW: Combination of interleukin 3 and 6 preserves stem cell function in culture and enhances
retrovirus-mediated gene transfer into hematopoietic cells. Proc Natl
Acad Sci USA 86:8897, 1989
3. Nolta JA, Kohn DB: Comparison of the effects of growth
factors on retroviral vector-mediated gene transfer and the proliferative status of human hematopoietic progenitor cells. Hum Gene Ther
1:257, 1990
4. Hughes PF, Thacker JD, Hogge D, Sutherland HJ, Thomas TE,
Lansdorp PM, Eaves CJ, Humphries RK: Retroviral gene transfer to
primitive normal and leukemic hematopoietic cells using clinically
applicable procedures. J Clin Invest 89: 1817, 1992
5 . Luskey BD, Rosenhlatt M, Zsebo K, Williams DA: Stem cell
factor, interleukin-3 and interleukin-6 promote retroviral mediated
gene transfer into murine hematopoietic stem cells. Blood 80:396,
1992
6. Moore KA, Deisseroth AB, Reading CL, Williams DE, Belmont JW: Stromal support enhances cell-free retroviral vector transduction of human bone marrow long term culture-initiating cells.
Blood 79:1393, 1992
7. Crooks GM, Kohn DB: Growth factors increase amphotropic
retrovirus binding to human CD34+ bone marrow progenitor cells.
Blood 82:3290, 1993
8. Bodine DM, Seidel NE, Gale MS, Nienhuis AW, Odic D:
Efficient retrovirus transduction of mouse pluripotent hematopoietic
stem cells mobilized into the peripheral blood by treatment with
granulocyte colony-stimulating factor and stem cell factor. Blood
84: 1482, 1994
9. Nolta JA, Hanley MB, Kohn DB: Sustained human hematopoiesis in immunodeficient mice by cotransplantation of marrow stroma
expressing human IL-3: Analysis of gene transduction of long-lived
progenitors. Blood 83:3041, 1994
10. Xu L, Stahl SK, Dave HPG, Schiffmann R, Correll PH, Kessler S, Karlsson S: Correction of the enzyme deficiency in hematopoietic cells of Gaucher patients using a clinically acceptable retroviral transduction protocol. Exp Hematol 22:223, 1994
11. Eglitis MA, Anderson WF: Retroviral vectors for introduction
of genes into mammalian cells. Biotechniques 6:608, 1988
12. Ueda K, Cardarelli C, Gottesman MM, Pastan I: Expression
of a full-length cDNA for the human “MDRI” gene confers resistance to colchicine, doxorubucin, and vinblastine. Proc Natl Acad
Sci USA 84:3004, 1987
13. Guild BC, Mulligan RC, Gros P, Houseman DE: Retroviral
transfer of a murine cDNA for multidrug resistance confers pleiotropic drug resistance to cells without prior drug selection. Proc Natl
Acad Sci USA 85:1595, 1988
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
1762
14. Podda S, Ward M, Himelstein A, Richardson C, de la FlorWeiss E, Smith L, Gottesman M, Pastan I, Bank A: Transfer and
expression of the human multiple drug resistance gene into live
mice. Proc Natl Acad Sci USA 89:9676, 1992
15. Sorrentino BP, Brandt SJ, Bodine D, Gottesman M, Pastan
I, Cline A, Nienhuis AW: Selection of drug-resistant bone marrow
cells in vivo after retroviral transfer of human MDR1. Science
257:99, 1992
16. Ward M, Richardson C, Pioli P, Smith L, Podda S, Goff S,
Hesdorffer C, Bank A: Transfer and expression of the human multiple
drug resistance gene in human CD34+ cells. Blood 841408, 1994
17. Aran JM, Gottesman MM, Pastan I: Drug-selected coexpression of human glucocerebrosidase and P-glycoprotein using a bicistronic vector. Proc Natl Acad Sci USA 91:3176, 1994
18. Strair RK, Towle MJ, Smith BR Recombinant retrovirusese n d ing cell surface antigens as selectable markers. J Vim1 624756, 1988
19. Lynch CM, Israel DI, Kaufman RI,Miller AD: Sequences in
the coding region of clotting factor VI11 act as dominant inhibitors
of RNA accumulation and protein production. Hum Gene Ther
4259, 1993
20. Choi K, Frommel TO, Stem RK, Perez CF, Kriegler M,
Tsuruo T, Roninson IB: Multidrug resistance after retroviral transfer
of the human MDRI gene correlates with P-glycoprotein density in
the plasma membrane and is not affected by cytotoxic selection.
Proc Natl Acad Sci USA 88:7386, 1991
21. Sugimoto Y, Aksentijevich I, Murray GJ, Brady RO, Pastan
I, Gottesman MM: Retroviral co-expression of a multidrug resistance
gene (MDRI) and human a-galactosidase A for the gene therapy of
Fabry’s disease. Hum Gene Ther 6:905, 1995
22. Olsen JC, Johnson LG, Wong-Sun ML, Moore KL, Swanstrom R, Boucher RC: Retrovirus-mediated gene transfer to cystic
fibrosis airway epithelial cells: Effect of selectable marker sequences
on long-term expression. Nucleic Acids Res 21:663, 1993
23. Mavilio F, Ferrari G, Rossini S, Nobili N, Bonini C, Casorati
G, Traversari C, Bordignon C: Peripheral blood lymphocytes as
target cells of retroviral vector-mediated gene transfer. Blood
83:1988, 1994
24. Kay R, Takei F, Humphries RK: Expression cloning of a
cDNA encodmg M1169-JI Id heat-stable antigens. J Immunol
145:1952, 1990
25. Pawliuk R, Kay R, Lansdorp P, Humphries RK: Selection of
retrovirally transduced hematopoietic cells using CD24 as a marker
of gene transfer. Blood 842868, 1994
26. Brady RO, Kanfer JN, Shapiro D: Metabolism of glucocerebrosides. 11. Evidence of an enzymatic deficiency in Gaucher’s disease. Biochem Biophys Res Commun 18221, 1965
27. Beutler E, Grabowski CA: Gaucher disease, in Scriver CR,
Beaudet AL. Sly WS, Valle D (eds): The Metabolic Basis of Inherited Disease. New York, NY, McGraw-Hill, 1995, p 2641
28. Correll PH, Colilla S, Karlsson S: Retroviral vector design for
long-term expression in murine hematopoietic cells in vivo. Blood
84:1812, 1994
29. Kay R, Takei F, Humphries RK: Expression cloning of a
cDNA encoding M1/69-J1 Id heat-stable antigens. J Immunol
145:1952, 1990
30. Pear WS, Nolan GP, Scott ML, Baltimore D: Production of
high-titer helper-free retroviruses by transient transfection. Proc Natl
Acad Sci USA 90:8392, 1993
MEDIN ET AL
3 1. Conneally E, Bardy P, Chappel S, Eaves CJ, Humphries RK:
Expression of murine heat stable antigen on human hematopoietic
cells as a marker of retroviral infection. Blood 84:414a, 1994 (abstr)
32. Xu L, Kluepfel-Stahl S, Blanco M, Schiffman R, Dunbar C,
Karlsson S: Growth factors and stromal support generate very efficient retroviral transduction of peripheral blood CD34+ cells from
Gaucher patients. Blood 86:141, 1995
33. Tsuji S, Choudary PV, Martin BM, Winfield S, Barranger
JA, Ginns EI: Nucleotide sequence of cDNA containing the complete
coding sequence for human lysosomal glucocerebrosidase. J Biol
Chem 261:50, 1986
34. Sorge J, West C, Westwood B, Beutler E: Molecular cloning
and nucleotide sequence of human glucocerebrosidase cDNA. Proc
Natl Acad Sci USA 827289, 1985
35. Adam MA, Ramesh N, Miller AD, Osborne WRA: Internal
initiation of translation in retroviral vectors carrying Picorna virus
5‘ nontranslated regions. J Virol 65:4985, 1991
36. Sugimoto Y, Aksentijevich I, Gottesman MM, Pastan I: Efficient expression of drug-selectable genes in retroviral vectors under
control of an internal ribosome entry site. Biotechnology 12694,
1994
37. Sorrentino BP, McDonagh KT, Woods D, Orlic D: Expression
of retroviral vectors containing the human multidrug resistance I
cDNA in hematopoietic cells. Blood 86:491, 1995
38. Challita P-M, Kohn DB: Lack of expression from a retroviral
vector after transduction of murine hematopoietic stem cells is associated with methylation in vivo. Proc Natl Acad Sci USA 91:2567,
1994
39. Hough MR, Takei F, Humphries RK, Kay R: Defective development of thymocytes overexpressing the costimulatory molecule,
heat-stable antigen. J Exp Med 179:177, 1994
40. Conneally E, Bardy P, Chappel S, Eaves CJ, Humphries RK:
Identification and selection of transduced human hematopoietic cells
using murine heat stable antigen (HSA) as a marker of retroviral
infection. J Hematother 4:3, 1995
41. Liu Y, Jones B, Aruffo A, Sullivan KM, Linsley PS, Janeway
JA Jr: Heat-stable antigen is a costimulatory molecule for CD24 T
cell growth. J Exp Med 175:437, 1992
42. Hubbe M, Altevogt P: Heat-stable antigedCD24 on mouse T
lymphocytes: Evidence for a costimulatory function. Eur J Immunol
24:731, 1994
43. Kadmon G, Eckert M, Sammar M, Schachner M, Altevogt
P: Nectadrin, the heat-stable antigen, is a cell adhesion molecule. J
Cell Biol 5:1245, 1992
44. Sammar M, Aigner S, Hubbe M, Schinmacher V, Schachner
M, Vestweber D, Altevogt P: Heat-stable antigen (CD24) as ligand
for mouse P-selectin. Int Immunol 6:1027, 1994
45. Hitsumoto Y, Ohnishi H, Nakano A, Hamada F, Sahelu S,
Takeuchi N: Inhibition of homologous complement activation by
the heat-stable antigen. Int Immunol 5:805, 1993
46. Gossen M, Bujard H: Tight control of gene expression in
mammalian cells by tetracycline-responsive promoters. Proc Natl
Acad Sci USA 895547, 1992
47. Wang Y, O’Malley BW, Tsai SY, O’Malley BW: A regulatory system for use in gene transfer. Proc Natl Acad Sci USA
91:8180, 1994
48. Gossen M, Freundlieb S, Bender G, Muller G, Hillen W,
Bujard H: Transcriptional activation by tetracyclines in mammalian
cells. Science 268:1766, 1995