From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
Granulocyte-Macrophage Colony-Stimulating Factor mRNA Stabilization
Enhances Transgenic Expression in Normal Cells and Tissues
By Lakshman E. Rajagopalan, Joseph K. Burkholder, Joel Turner, Jerilyn Culp, Ning-Sun Yang, and James S. Malter
To increasetransgenicproductionofgranulocyte-macrophage colony-stimulating factor (GM-CSF),we mutated the
mRNAs 3'-untranslated region, AUUUA instability elements. Expression vectors containing
human or murine GMCSF cDNAs coding for wild-type (GM-AUUUA) or mutant
versions with reiterated AUGUA repeats (GM-AUGUA) were
transfected into cells in culture or animals using particlemediated gene-transfer
technology.
Normal peripheral
blood mononuclear cells accumulated20-fold greater levels
of
GM-CSF
mRNA
and
secreted
comparably
greater
amounts of cytokine after transfection with hGM-AUGUA
expression
vectors
versus
hGM-AUUUA. hGM-AUGUA
mRNA was fivefold more stable (tl12 = 95 minutes) than
hGM-AUUUA mRNA (tl12= 20 minutes). accounting for elevated steady-state levels. Transfection site extracts and serum samples obtained 24 hours after gene transfer ofhGMAUGUA cDNA into mouse skin contained greater than 32
ng/mL and 650 pg/mL of GM-CSF protein, respectively, compared with 0.33 ng/mL and tess than 8 pglmL for hGM-AUUUA cDNA. GM-CSF produced from mGM-AUGUA cDNA
transfected into rat abdominalepidermisinducedaprofound neutrophilinfiltrate. These data suggest a novel strategy for enhanced production of biologically active cytokines
by normal cells after in vivo gene transfer.
0 1995 by The AmericanSociety of Hematology.
I
(UTR) of most cytokine mRNAs including granulocyte-macrophage colony-stimulating factor (GM-CSF), IL-2, IFN-a,
TNF-a and IFN-y contain adenosine-uridine (AU)-rich elements (ARES)that target the mRNAs for rapid degradationlg
and inhibit translation?' The highly structured 5' UTR and
suboptimal start codon context of cytokine mRNAs are poor
initiators of translation, further impeding cytokine productiom21 We hypothesized that targeted mutations within the
3' UTR regulatory elements could stabilize cytokine mRNAs
derived from transgenes thereby increasing the availability
of templates and enhancing protein production and secretion.
We have tested this by mutating the ARES of GM-CSF
mRNA. The data show that after gene transfer of mutant
cDNAs, the steady-state level of GM-CSF mRNA was substantially higher and the production of cytokine proportionately greater than seen with wild-type cDNAs. We further
show that transgenic GM-CSF derived from mutant constructs was biologically active and able to elicit a profound
inflammatory response. These observations suggest alternative strategies to increase cytokine production after gene
transfer for the treatment of human disease.
N RECENT YEARS cytokines have shown considerable
therapeutic potential. Single clinical grade recombinant
cytokines or combinations have been systemically administered to patients with a variety of neopla~ms."~
For example,
interferon-a (IFN-a) therapy has improved the long-term
survival of patients with hairy cell leukemia2 and may be
synergistic with conventional chemotherapy for the treatment ofmultiple r n y e l ~ m anon-Hodgkin's
,~
lymphoma: and
colorectal cancer^.^
Mechanistically, exogenous cytokines, particularly interleukin-2 (IL-2), augment the proliferation and antitumor
activity of cytotoxic T lymphocytes, natural killer (NK) cells,
lymphokine-activated killer (LAK) cells, and tumor-infiltrating lymphocytes.6 Adoptive immunotherapy with IL-2
and LAK cells has been effective against malignant melanoma and renal cell carcinoma.' However, this approach is
time consuming and expensive and, thus, not yet in routine
clinical use.
Typically, cytokines have been administered systemically
by either bolus injection or continuous i n f u s i ~ n .However,
~.~
the inability to specifically target cytokines to tumor cells
requires the delivery of massive systemic doses. As such,
many patients experience significant side effects that often
prevent adequate dosage."' This has impeded the widespread
use of tumor necrosis factor-a (TNF-a)."
Daily, peritumor injection of low to moderate doses of
IL-2I2.l3or TNF-aI4 was associated with significant tumor
shrinkage without debilitating systemic side effects. However, the short serum half-life of cytokines (t1,2, 15 to 30
minutes) has encouraged alternative delivery systems including ex vivo transfer of cytokine cDNAs into tumor cells. In
such models, cytokine genes have induced impressive, local
and systemic antitumor immune activity without incapacitating side
The success of periturnor administration
has critically depended on the local cytokine concentrations
obtained after direct injection" or gene transfer.I6
Irrespective of gene delivery system, cytokine cDNAs
tend to be verypoorly expressed after gene transfer. The use
of powerful viral or endogenous gene promoters's has not
circumvented this problem, suggesting posttranscriptional
regulatory mechanisms may be limiting cytokine mRNA accumulation and translation. The 3' untranslated region
Blood, Vol 86, No 7 (October l ) , 1995: pp 2551-2558
MATERIALS AND METHODS
cDNA constructs. cDNA coding for a human GM-CSF was obtained from the American Type Culture Collection, Rockville, MD.
From the Department of Pathology and Laboratory Medicine and
Comprehensive Cancer Center, University of Wisconsin Medical
School; and the Department of Cancer Gene Therapy, Agracetus
lnc, Madison, Wl.
Submitted March 9, 1995; accepted May 31, 1995.
Supported by National Institutes of Health Grant No. DK4.5213
(tu J.S.M.)
Address reprint requests to James S. Malter, MD, A4fZI"SC,
Department of Pathology and Laboratory Medicine, University of
Wisconsin, Hospitals and Clinics, 600 Highiand Ave, Madison, W1
53792-2472.
The publication costs of this article were defrayed in parr 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.
0 1995 by The American Society of Hematology.
0006-4971/95/8607-0023$3.00/0
2551
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
2552
RAJAGOPALAN ET AL
Human OY-CSIf
636
wild type :
AATATTTATATATTTATATTTTTAAAATATTTATTTATTTATTTATTTA
mutant
AAT-"-
:
wild type:
mutant
:
- _ _--_- - - _ _ _ _ - _ _ATOTATOTATmATQTA---_____
685
AGTTCATATTCCATATTTATTCA
TTCA
""_"""
"
Hurino OY-CSP
7 85
wild type:
GATATTTTCTACTGATAGGGACCATTATATTTATTTATATATTTATATTTTT
mutant
GATA
:
Fig 1. Targeted mutagenesis of wild-type human
and murine GM-CSF ARES. Using overlap extension
PCR (mMaterials and Methods), the
of wild-
_ _ _ - _--_ -_____--_______---_----___-----_______
type human (nt 639 to 703) and murine (nt 789 to
880) GM-CSF (GM-AUUUAI werereplaced with four
831
wild type:
mutant
:
TAAATATTTATTTATTTATTTA~TAATTTTGcAAcTcTATTTATTGA
- - - - - ATQTATQTATaTATQTA- - _ _ _ _ - - - _ _ -
---
__--- - -TTGA
Mutagenesis of the construct (Fig 1) was performed by overlap
extension polymerase chain reaction (PCR)." Briefly, oligonucleotides complimentary to opposite strands of the most 5' and 3' regions
of GM-CSF were constructed. GM-l (nt 1 to 19) and GM-4 (nt 748
to 734) contained ~ 5 0 %GC residues (melting temperature [T,]
= 45°Cto 55°C). Mutagenic primers were constructed containing
complimentary sequences to GM-CSF immediately 5' or immediately 3' to the flanking 3' UTR AUUUA pentamers [designated GM2 (nt 620 to638) and GM-3 (nt 721 to 704)]. Atthe 5' endof
the mutagenic oligomers were 17 bases containing four ATGTA or
TACAT reiterations, respectively. PCR was performed on a GMCSF cDNA template using primer pairs of GM-UGM-2 and GM-3/
GM-4. Products were visualized by agarose gel electrophoresis and
the appropriate bands were excised and purified with a Qiagen kit
as described by
the
manufacturer (Qiagen, Chatsworth, CA).
Amplified fragments (20 ng) from the GM-1/GM-2 and GM-3/GM4 PCR products were mixed together and extended for five cycles
with Taq polymerase using an annealing temperature of 36°C for 2
minutes before denaturation at 92°C for 1 minute. Terminal primer
pairs GM-l and GM-4 were added and 35 cycles performed with
denaturation at 92°C for 1 minute, annealing at 44°C for 1 minute
and extension at 72°C for I minute. Products were visualized on an
ethidium bromide-stained agarose gel, the appropriate-sized band
of =700 bp corresponding to full-length mutant GM-AUGUA
cDNA was excised, purified by the Qiagen kit and ligated into
an EcoRV cut, T-tailed, cytomegalovirus (CMV)-driven expression
vector with a downstream SV40 poly A ~ignaI.*'~'~
Wild-type GMAUUUA cDNA was subcloned at the same site in the same vector.
After electrotransformation of competent Escherichiacoli strain
DHSa, recombinants were identified by PCR using the GM-UGM4 primer set. After the identification of appropriate recombinants,
plasmids were produced in large scale by standard methods and
purified by cesium chloride centrifugation. Murine wild-type GMAUUUA (obtained from Nicholas Gough, Walter and Elisa Hall
Institute, Melbourne, Australia) was mutagenized inan identical
manner (Fig 1). The primers used were GM-l (nt 145 to 164), GM2 (nt 769 to 788), GM-3 (nt 899 to 881), and GM-4 (nt 919 to 895).
Mutated cDNAs were sequenced by the dideoxy method2' to ensure
that no unwanted mutations were introduced.
Cells and cellculture.
Normal peripheral blood mononuclear
cells (PBMCs) were obtained after Institutional Review Board (IRB)
approval from healthy volunteer blood donors. Whole blood (200
to 500 mL) was diluted 1:l with phosphate-buffered saline (PBS)
and layered over Ficoll-Hypaque before centrifiguation at 200g for
30 minutes atroom temperature. The PBMCs were carefully re-
tandem AUGUA sequences (GM-AUGUA). Dashed
lines dmote deleted sequences in the
cDNAs.
moved, transferred to 50-mL conical tubes and washed twice with
PBS. Cells were greater than 95% viable by trypan blue exclusion
with yields in the range of 1 X lo9 cells per 500 mL of whole blood
starting material. Before transfection, cells were cultured overnight
at 37°C and 5% CO2 at a density of 5 X IO6 cells/mL in RPMI 1640
containing 10% fetal calf serum (FCS).
DNA transfections and cytokine assay. Particle-mediated gene
transfer of cDNA constructs into intact rodent skin and cultured
PBMCs was performed as previously d e ~ c r i b e d . ~Animal
~ , ' ~ use was
conducted under protocols approved by the institutional animal use
and care committee. Twenty-four hours after in vivo transfection of
rodent skin, bloodwas collected under anesthesia for analysis of
serum cytokine levels. Transfected skin was collected from animals
after death and homogenized in Dulbecco's PBS containing 1%
Triton X-l00 and 1 m m o K Pefabloc (Boehringer Mannheim, Indianapolis, IN). Transgenic cytokine levels were measured using species-specific GM-CSF enzyme-linked immunosorbent assay
(ELISA) kits (Biosource, Camarillo, CA). Standard curves were simultaneously run for the calculation of cytokine concentrations.
Northernblotting. After transfection of PBMCs, 5 X lo6 cells
were resuspended in 1 mL of complete media (RPMI 1640 with
10% FCS) and returned to culture at 37°C ina 5% CO, environment.
Where indicated, actinomycin D (final concentration, 5 pg/mL) was
added to the transfected cells to block transcription. At various times,
cells were lysed in 1 mL of TRI reagent (Molecular Research Center
Inc, Cincinnati, OH) and snap-frozen in an ethanol bath at -80°C.
After all time points for an individual experiment were taken, total
RNA was quantitatively isolated and separated by size on denaturing
formaldehyde-agarose gels as previously described.26RNA was then
transferred to nylon membranes (Micron Separations Inc, Westborough, MA) by vaccum transfer and membranes were baked at 65°C
for 30 minutes before hybridization using random-primed cDNA
probes." All probes were labeled to a specific activity of greater
than 1 X IO9 cpm/pg of DNA. Blots were washed twice in 2 X
SSC/O.l% sodium dodecyl sulfate (SDS) at room temperature for
15 minutes each and once at 60°C to 65°C for 5 to 15 minutes in
0. I x SSC/O.l% SDS before autoradiography or phosphorimaging.
GM-CSF-specific, phosphor image-derived signals were normalized
to signals for glyceraldehyde-3-phosphatedehydrogenase (GAPDH)
to accomodate differences in loading and transfer of RNA.
RESULTS
Steady-state accumulation of wild-type and mutant GMCSF mRNAs in resting PBMCs. Particle-mediated gene de-
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
2553
GM-CSF mRNA STABILIZATION
hGM-AUGUA
28
412
GM-CSF
(0.9 kb)
-
hGM-AUUUA
2 4 8 12 (h)
I
L
Fig 2. Steady-state accumulation of wild-type and mutant GMCSF mRNAs in resting PBMCs. Resting PBMCs were transfected with
either pCMV-hGM-AUUUA or pCMV-hGM-AUGUA. At the indicated
times after transfection, equal numbers of cells were procured for
Northern blotting of total RNA. The top panel shows ethidium bromide-stained 28s and 18s ribosomal RNA bands, and the middle and
lower panels show autotadiograms after hybridization with radiolabeled GAPDH or GM-CSF cDNA probes, respectively.
livery has been shown to successfully transfect normal cells
under various in vivo, ex vivo, and in vitro experimental
condition^^'.^^ providing an alternative gene transfer method
to retroviruses or liposomes. Therefore, we observed whether
ficoll purified, normal human PBMCs could be transfected
in vitro. Typically, 4% to 6% of normal PBMCs were successfully transfected by a single discharge of gold particles
loaded with 0.8 pg of a cDNA construct.24PBMCs were
transfected with either wild-type hGM-AUUUA or the modifiedhGM-AUGUA cDNAs subcloned in identical CMVdriven expression vectors. RNA was isolated at 2, 4, 8, and
12 hours posttransfection, Northern blotted, and probed using random-primed, human GM-CSF cDNA. This experiment alloweddirect comparison of the accumulation of wildtype hGM-AUUUA or mutant hGM-AUGUA mRNAs in
normalPBMCs. Examination of the ethidium bromidestained gels showed that the 28s and 18s rRNAs remained
intact and stable over the duration of the experiment (Fig
2). hGM-AUGUA mRNA increased steadily over 12 hours
posttransfection (Fig 2). whereas the steady-state level of
hGM-AUUUAmRNApeaked
at 2 hours and decreased
steadily thereafter (Fig 2). At 12 hours posttransfection, the
steady-state level of hGM-AUGUA mRNA was greater than
20-fold more abundant than that of hGM-AUUUA mRNA
(based on phosphor-imager analysis). GM-CSF-specific signals were detected only in PBMC transfected with the wildtype or mutant construct, but not with the vector control or
nakedgold beads (data not shown). Therefore, GM-CSF
signals must have originated from the transgene. ELISAs for
GM-CSF were performed on conditioned culture medium
and cell lysates from identical numbers of cells at 24 hours
after transfection. The hGM-AUGUA transfectants secreted
520 ? 12 pg of GM-CSF proteidmU1 X lo6 cells compared
with 26 2 2 pg/mUI X 10" cells for hGM-AUUUA transfectants and18 2 6 pglmLJ1 X 10" cells for vector control
transfectants. In addition, cell extracts from the hGM-AUGUA transfected cells contained an additional 220 2 16 pg
of GM-CSF proteid1 X IO6 cells. GM-CSF was not detectable in cell pellets from wild-type or vector transfected cells.
The Biosource GM-CSF ELISA used for these determinations has a lower limit of detection of = 10 pg/mL.
Half-lives of hGM-CSF mRNAs in resting PBMCs. Because transcription of wild-type andmutanthGM-CSF
mRNAs were both under the control of identical CMV promoters, it was unlikely that the higher steady-state mRNA
levels and increased protein production inhGM-AUGUA
transfected PBMCs was caused by differences in the transcription rates. Because we had specifically mutated the AUUUA instability elements, we expected hGM-AUGUA
mRNA to show enhanced stability. Therefore, we determined the turnover rates of these two mRNAs after blocking
transcription with actinomycin D. For this purpose, PBMCs
were isolated from single donors and cultured overnight in
RPM1 1640 with 10% FCS. Equal quantities (2.5 pg DNA/
mg gold) of each plasmid were precipitated onto gold beads
and delivered into resting PBMCs ( 1 X IO' cells/transfection) using particle-mediated gene transfer. Replicate cultures were pooled immediately after transfection to normalize differences between individual transfections. After 4
hours in culture, actinomycin D was added (5 pglmL final)
to block transcription. Equal numbers of transfected cells
were removed at the indicated times for RNA isolation and
Northern blotting. Based on absorbance at 260 nm, 2 pg of
total RNA/time point was loaded in each lane to determine
the half-life (ttn) of hGM-AUGUAmRNA (Fig 3A).Because of its lower abundance, five times as much RNA (IO
pg/time point) was Northern blotted to measure the tln of
wild-type hGM-AUUUAmRNA (Fig 3A). Ethidium bromide-stained 28s and 18s ribosomal bands were intact and
stable over the duration of the experiment. GM-CSF mRNA
signals normalized to GAPDH signals at each time point
(based upon phosphor-imager analysis) were plotted versus
time to provide a calculated half-life of 20 minutes for wildtype hGM-CSF (Fig 3B). To our knowledge, this is the first
report of the decay rate of GM-CSF mRNA in primary cell
cultures of normal PBMCs. In contrast, hGM-AUGUA
mRNA decayed with a calculated half life of 95 minutes
(Fig 3B). Therefore, the enhanced stability of mutant GMCSF mRNAs in normal cells accounts for their accumulation.
In vivo synthesis of GM-CSF protein. Although the in
vitro studies are suggestive, they do not show if our constructs will be differentially active in vivo. To address this
question, we introduced cDNAs coding for hGM-AUWA
or hGM-AUGUA mRNAs into mouse skin by particle-mediated gene transfer. Human cDNA constructs were used so
that we could employ a human GM-CSF-specific ELISA to
measure proteinmadeonlyfrom
the transgene. Basedon
control studies, murine and human GM-CSFs show no crossreactivity (data not shown). Mouse (BalblC) abdominal skin
was shaved, treated
with
a depilatory, washed,
and
transfected with gold particles coated with identical amounts
of CMV-driven hGM-AUUUA, hGM-AUGUA, or control
vector containing a luciferase reporter cDNA. Three mice
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
RAJAGOPALAN ET AL
2554
were used per treatment to evaluate reproducibility. No untoward effects were observed in treated animals other than a
slight rash at the target site. After 24 hours serum samples
were taken from animals, and transfected skin was obtained
and homogenized to obtain a total volume of 1 mL tissue
extract. Both serum and skin samples were assayed by human GM-CSF-specific ELISA to determine expression levels from the various gene constructs. Skin extracts and serum
samples obtained at 24 hours after hGM-AUGUA gene
transfer contained a 100-fold excess of human GM-CSF immunoreactive material compared with identically prepared
samples of skin and serum from hGM-AUUUA cDNA
transfected mice (Table 1). Animals that received luciferase
cDNA did not produce any detectable human GM-CSF.
hGM-AUGUA
0 20 40 60 !M
A
LGM-AUUUA
0 2040 70 (min)
hGM-AUGUA
+ hGM-AUUUA
I
I”
0
20
.
.
40
.
,
60
.
,
80
.
100
Time
(minutes)
Fig 3. Half-life measurements of wild-type and mutant GM-CSF
mRNAs in resting PBMCs. Resting PBMCswere transfected with either pCMV-hGM-AUUUA or pCMV-hGM-AUGUA. Transcriptionwas
blocked at 4 hours posltransfection with actinomycin D (5pglmL,
final). At the indicatedtimes after actinomycin D addition, equal numbers of cells were procured for the isolation of total RNA and Northern blotted. (A) The top panel shows ethidium-bromide stained 28s
and 18s ribosomal RNAbands, and the middle and lower panels
show autotadiograms after hybridization with radiolabeled GAPDH
or GM-CSF cDNA probes, respectively.
(B1 GM-CSF mRNA signals
were normalized to GAPDH mRNA signals and plotted versus time.
Table 1. Transgenic GM-CSF Production in Skin Extracts and
Serum Samples of Mice
Transfected
cDNA
Skin GM-CSF’ (ng/mL)
Serum GM-CSF. IpglmL)
hGM-AUUUA
hGM-AUGUA
0.33 -+ 0.074
>32
<E
650 2 37
A human GM-CSF-specific ELSA was used to measure protein
made only from the transgene.
These data show that the GM-CSF protein detected in these
experiments originated from the transgene. Second, they
show that mutant hGM-AUGUA cDNAs are extremely active in vivo at expression levels at least 100-fold greater than
wild-type hGM-AUUUA cDNAs.
GM-CSF produced from mutant constructs is biologically
active. To verifythattransgenicGM-CSFproteinproduced
from the mutant constructs was equivalent in biologic activity
to that produced from unmodified cDNA, we introduced both
gene constructs into adjacent regions
of rat epidermis. Because
of interspecies specificity, human GM-CSF does notinvoke
a biologicresponsein rats?’ However,murineGM-CSFis
rats, permittingvisualizationbylocal
biologicallyactivein
redness and immune cell recruitment.In addition, there is adeon ratabdomenstoperform
quatecontiguoussurfacearea
multiple, discrete transfections. The abdomens of several animalswereshavedandtreatedwith
a depilatorybeforethe
introduction by particle-mediated gene transfer ofgold particlesloadedwith CMVdriven expression vectors containing
mGM-AUUUA, mGM-AUGUA, or P-galactosidase (control)
cDNAs. These experiments were performed
in several similarly
sized animals. Twenty-four hours after gene delivery, the abdominal skins were examined grossly and punch biopsy samples were obtained from each target site. Formalin-fixed paraffin sectionswerestainedwithhematoxylinandeosinand
examined microscopically (Fig 4).
Skin regions that received P-galactosidase cDNA showed
minimal redness and were not raised (data not shown). Identical levels of redness occurred in control, shaved animals.
The sites that received wild-type mGM-AUUUA cDNA
were grossly indistinguishable from the control that received
&galactosidase cDNA. However, the sites wheremutant
mGM-AUGUA cDNA was delivered were markedly red and
raised. These differences were consistently observed at all
transfection sites and in the different experimental animals.
The suggestion that substantial inflammation was present
where mutant mGM-AUGUA wasintroduced was confirmed
by histologic examination. As shown in Fig 4, inflammation
was absent in tissues where control cDNA was introduced.
A single small focus of inflammation was observed in the
central region where wild-type mGM-AUUUA cDNA was
delivered. Polymorphonuclear cells (PMNs) were the dominant inflammatory cell type present, but the bulk of the epithelium or dermis was devoid of immune cells. However,
the total tissue site underlying the mGM-AUGUA cDNA
transfection was infiltrated with PMNs that formed a layer
of 20 to 40 cells deep, effectively separating the epidermis
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
2555
GM-CSF mRNASTABILIZATION
Fig 4. Transgenic GM-CSF recruits neutrophils to transfection site. Contiguous 3.2-cmZ regions of rat abdomen wem transfected with
pCMVpgal (top), pCMVmGMAUUUA (middle) or pCMVmGMAUGUA (bottom) by particle-mediated gene transfer. After 24 houm, punch
biopsy samples from each site were collected and futed in 10% formalin, embedded in paraffin, sectioned, and stained with hematoxylin and
eosin (original magnifications x 40 [left1 and 200 [right]).
from the underlying dermis. Infiltration into the dermis could
also be appreciated. A substantial percentage of the PMNs
had released their granules suggesting that the PMNs were
activated. These data clearly show that GM-CSF produced
from mutant transgenic constructs is indeed biologically active. Furthermore, they showed that enhanced cytokine production can profoundly increase local, immune cell recruitment.
DISCUSSION
Cytokines such as IL-4,I5 GM-CSF,I6 IFN-Y?~ and TNFa*' effectively retard tumor growth in animal models when
expressed by tumor cells after retroviral-mediated gene
transfer. Typically, cytokine gene transfer has been performed ex vivo followed by the subsequent reimplantation of
the modified malignant cells into the host. Retroviral vectormediated gene transfer has been successfully used to achieve
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
2556
constitutive expression of cytokines in explanted human mela n o m a and
~~~
renal carcinoma^.^' These experiments showed
that peritumor cytokine production recruited a range of host
immune effector cells with antitumor activity. As the growth
rate of cytokine-producing tumor cells was normal in severe
combined immunodeficient mice, antitumor effects likely require an intact immune system.32Occasionally, the local
eradication of cytokine-modified tumor cells was accompanied by a systemic antitumor response capable of destroying
unmodified parental cells at distant sites.15 Thus, local gene
therapy with cytokines may be able to induce systemic antitumor immunity.
Unfortunately, ex vivo retroviral-mediated gene therapies
are often costly and time consuming, requiring weeks for
the selection of stably transduced tumor cells.“03’In addition,
not all tumors canbe surgically explanted, cultured, and
transduced ex vivo. Retroviral vectors are unable to infect
nondividing cells or targets lacking specific cell-surface viral
Insertional mutagenesis with activation of
proto-oncogenes has been described.18,” Cytokine production cannot be modulated and continues until all stably transduced tumor cells are destroyed or expression stops for generally unknown reasons. The sequential delivery of
cytokines, whichmay be therapeutically advantageous, is
technically very difficult.
The inadequacies of retroviral delivery systems have fueled attempts to develop alternative means of delivering cytokine genes in vivo to the vicinity of tumors. In this report,
we have used a promising new technology that utilizes a
high voltage discharge to deliver microscopic gold particles
loaded with nucleic acids into normal, nondividing cells in
culture or into animals in vivo, eliminating the need for ex
vivo manipulations. Significant advantages of the gene gun
include the ability (1) to physically target gene expression
to the site of particle delivery; (2) to transfect nondividing
cells irrespective of cell lineage; (3) to allow sequential delivery of different cytokines; and (4) to transiently produce
transgenic proteins for short duration (days) rather than permanently. Previously we have shown that reiterative cycles
of gene delivery can readily and safelybe performed on
animals, permitting modulation or long-term maintenance of
transgene dosage.34Therefore, the gene gun technology can
potentially overcome many of the deficiencies of retroviralmediated gene transfer and could be used to deliver genetic
material with therapeutic impact in vivo into normal cells
surrounding the tumor as well as into the tumor itself.
The success of gene therapy depends on the critical combination of appropriate temporal and quantitative expression
of the transfected genes. Unfortunately, cellular regulatory
mechanisms often downregulate the expression of
transfected cytokine genes.” A common feature of a number
of unstable mFtNAs is the presence of multiple reiterations
of the pentamer adenosine-uridine-uridine-uridine-adenosine
(AUUUA, ARE), in their 3’ UTRs.19 AUUUA motifs have
been identified in mostcytokine and proto-oncogene mRNAs
including GM-CSF, interleukins, interferons, TNF-a, c-fos,
c-myc, c-myb, c-sis, and c-jun. mRNAs with tandemrepeats
of this pentamer are rapidly degraded (tlR = 20 to 30 min-
RAJAGOPALAN ET AL
utes) in human or murine cells.”5.3hOur data confirms that
the presence of 3’ UTR AREs causes GM-CSF mRNA to
be rapidly degraded in the cytosol. In this report, we observed that in normal PBMCs, transgenic wild-type hGMCSF mRNA decayed with a half-life of 20 minutes. This is
more rapid than previous estimates based on data derived
from transformed
Therefore particle-mediated gene
transfer permits the study of mRNA decay in normal nontransformed cells.
However, mRNA instability can be circumvented by at
least one set of mutations. We disrupted the AUUUA boxes
in the 3‘ UTR of GM-CSF mRNA by inserting guanosines
in the third position(Fig 1). This change had previously been
shown to prevent mutant GM-CSF mRNAs from binding to
theAUUUA sequence-specific AU binding factor.” Recently, fine mutagenesis of the ARE of c-fos mRNA showed
that guanosine insertions at a comparable siteas reported
here prevented rapid decay.” Therefore, it appears likely
that RNAses that normally degrade AUUUA-containing cytokine or proto-oncogene mFWAs are also unable to recognize mutant versions.
As shown in Fig 2, mutant GM-CSF mRNA accumulated
to 20-fold greater levels than wild-type GM-CSF mRNA.
Because both mRNAs were transcribed from the same CMV
promoter it was unlikely that the difference in steady-state
levels was caused by a difference in the rates of transcription.
However, the calculated half-life (t1,2)of hGM-AUUUA
mRNAwas 20 minutes (Fig 3B), whereas hGM-AUGUA
mRNA was significantly more stable with a half-life of 95
minutes. As decay is exponential, the fivefold greater stability of hGM-AUGUA mRNA accounts for its accumulation
in PBMC after gene transfer. ELISA for GM-CSF performed
on supernatants from identical numbers of cells at 24 hours
after transfection showed that hGM-AUGUA transfectants
secreted 20- to25-foldmore
immunologically detectable
protein (550 pglmLil X IOb cells) than the hGM-AUUUA
transfectants. As approximately 5% of PBMCs are typically
transfected by the gene gun, a fully transfected population
(100%) would generate about I l ng GM-CSF/mL/ l X 10’
cells. The linear correlation between increased steady-state
levels of hGM-AUGUA mRNA and synthesis of GM-CSF
protein wassomewhat unexpected. Previous work has shown
that the 3’ UTR AREs interfere with the translatability of
IFN-p mRNA.’” Our data suggest that the U to G mutation
altered GM-CSF mRNA stability without changing its translatability. Therefore, additional ARE mutations may be effective in increasing translation as well as enhancing mRNA
stability.
We have adopted a cautious approach of targeted mutagenesis of the AREs. The substantial homology between the
3’ UTR of human and murine GM-CSFI9 suggests that regions outside of the AREs may also have functional importance. Recent workwith actin:’
bi~oid?’.~’andnanos4’
mRNAs shows that elements within the 3’ UTRs mediate
intracellular trafficking or localization. Based on these data
and the paucity of information regarding additional 3’ UTR
elements within cytokine mRNAs, we preferred targeted
ARE mutations rather than 3’ UTR deletions. The possibility
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
GM-CSF mRNA STABILIZATION
remains that other advantageous mutations can also be identified and tested. Kozak*' has shown that cytokine mRNAs
are poorly translated because of the presence of stable stemloop structures in their 5' UTR. Thus, substitution of the 5'
UTRs with less structured sequences could fuaher enhance
translational efficiencies." As many cytokine mRNAs are
under similar posttranscriptional regulation, selective mutagenesis may be beneficially applied to enhance the expression of other transgenic growth factors.
REFERENCES
1. Takaku F Clinical application of cytokines for cancer treatment. Oncology 51:119, 1994
2. Smith J W , Longs DL, Urba WJ, Clark J W , Watson T, Beveridge J, Colon KC, Sznol M, Creekmore SP, Alvord WG, Lawrence
JB, Steis R Prolonged, continuous treatment of hairy cell leukemia
patients with recombinant interferon-a2a. Blood 78:1664, 1991
3. Montuoro A, DeRosa L, DeBlasio A, Pacilli L, Petti N, DeLaurenzi A: Alph-2a-interferon/me/melphalan/prednisoneversus melphaladprednisone in previously untreated patients with multiple myeloma. Br J Haematol 76365, 1990
4. Smalley RV, Anderson J W , Hawking MJ, Bhide V, O'Connell
MJ, Oken MM, Borden EC: Interferon-alpha combined with cytotoxic chemotherapy for patients with non-Hodgkin's lymphoma. N
Engl J Med 327:1336, 1992
5. Wadler S, Schwartz EL, Goldman M, Lyver A, Rader M,
Zimmerman M, Itri L, Weinberg V, Wiernik PH: Fluorouracil and
recombinant alpha-2a-interferon: An active regimen against advanced colorectal carcinoma. J Clin Oncol 7:1769, 1989
6. Foon K: Biological response modifiers: The new immunotherapy. Cancer Res 49:1621, 1989
7. Rosenberg SA, Lotze MT, Muul LM, Chang AE, Avis FP,
Leitman S , Linehan WM, Rosenberg CN, Lee RE, Rubin JT, Seip
CA, Simpson CG, White DE: A progress report on the treatment of
157 patients with advanced cancer using lymphokine-activated killer
cells and interleukin-2 or high-dose interleukin-2 alone. N Engl J
Med 3162389, 1987
8. Perez EA, Scudder SA, Meyers FA, Tanaka MS, Paradise C,
Gandara DR: Weekly 24-hour continuous infusion interleukin-2 for
metastatic melanoma and renal cell carcinoma: A phase I study. J
Immunother 1057, 1991
9. Schiller JH, Storer BE, Witt PL, Alberti D, Tombes MB, Arzoomanian R, Proctor RA, McCarthy D, Brown RR, Voss SD, Remick SC, Grem JL, Borden EC, Trump DL: Biological and clinical
effects of intravenous tumor necrosis factor-alpha administered three
times weekly. Cancer Res 51:1651, 1991
10. Geertsen P, Hermann GG, Maase H, Steven K: Treatment of
metastatic renal cell carcinoma by continuous intravenous infusion
of recombinant interleukin-2: A single-center phase Il study. J Clin
Oncol 10753, 1992
11. Rosenberg SA, Lotze ML, Yang JC, Abersold PM, Linehan
WM, Seipp CA, White DE: Experience with the use of high-dose
interleukin-2 in the treatment of 652 cancer patients. Ann Surg
210:474, 1989
12. Cortesina G, Stefani AD, Giovarelli M, Barioglio MG, Cavallo GP, Jemma C, Forni G: Treatment of recurrent squamous cell
carcinoma of the head and neckwithlow doses of interleukin-2
injected perilymphatically. Cancer 62:2482, 1988
13. Vaage J: Local and systemic effects during interleukin-2 therapy of mouse mammary tumors. Cancer Res 47:4296, 1987
14. Balkwill FR, Lee A, Aldam G, Moodie E, Thomas JA, Tavernier J, Fiers W: Human tumor xenografts with recombinant human
2557
tumor necrosis factor alone or in combination with interferons. Cancer Res 46:3990, 1986
15. Golumbek R,Lazenby AJ, Levitsky HI, Jaffee LM, Karasuyama H,Baker M, Pardoll DM: Treatment of established renal cancer
by tumor cells engineered to secrete interleukin-4. Science 254713,
1991
16. Dranoff G, Jaffee E, Lazenby A, Golumbek P, Levitsky H,
Brose K, Jackson V, Hamada H, Pardoll D, Mulligan RC: Vaccination withirradiated tumor cells engineered to secrete murine granulocyte-macrophage colony-stimulating factor stimulates potent, specific, and long-lasting anti-tumor immunity. Proc Natl Acad Sci USA
90:3539, 1993
17. Forni G, Giovarelli M, Santoni A: Lymphokine-activated tumor inhibition in vivo. The local administration of interleukin 2
triggers nonreactive lymphocytes from tumor-bearing mice to inhibit
tumor growth. J Immunol 134:1305, 1985
18. Mulligan RC: The basic science of gene therapy. Science
260:926, 1993
19. Shaw G, Kamen R: A conserved AU sequence from the 3'
untranslated region of GM-CSF mRNA mediates selective mRNA
degradation. Cell 46:659, 1986
20. Kruys V, Marinx 0, Shaw G , Deschamps J, Huez G: Translation blockade imposed by cytokine-derived AU-rich sequences. Science 245:852, 1989
21. Kozak M: An analysis of vertebrate mRNA sequences: Intimations of translational control. J Cell Biol 115:887, 1991
22. Higuchi R: Using PCR to engineer DNA, in Erlich HA (ed):
PCR Technology. Principles and Applications for DNA Amplification. New York, N Y , Stockton, 1989, p 61
23. Cheng L, Ziegelhoffer PR, Yang N-S: In vivo promoter activity and transgene expression in mammalian somatic tissues evaluated
by using particle bombardment. Proc Natl Acad Sci USA 90:4455,
1993
24. Burkholder JK, Decker J, Yang N-S: Rapid transgene expression in lymphocyte and macrophage primary cultures after particle
bombardment-mediated gene transfer. J Immunol Methods 165:149,
1993
25. Sanger F, Nicklen S, Coulson AR: DNA sequencing with
chain-terminating inhibitors. Proc Natl Acad Sci USA 745463, 1977
26. Rajagopalan LE, Malter JS: Modulation of granulocyte-macrophage colony-stimulating factor mRNA stability in vitro by the
adenosine-uridine binding factor. J Biol Chem 269:23882, 1994
27. Shanafelt AB, Johnson KE, Kastelein RA: Identification of
critical amino acid residues in human andmouse granulocyte-macrophage colony stimulating factor and their involvement in species
specificity. J Biol Chem 266:13804, 1991
28. Gansbacher B, Bannerji R, Daniels B, Zier K, Cronin K,
Gilboa E: Retroviral vector-mediated interferon-gamma gene transfer into tumor cells generates potent and long lasting antitumor
immunity. Cancer Res 50:7820, 1990
29. Asher AL, Mule JJ, Kasid A, Restifo NP, Salo JC, Reihert
CM, Jaffe G, Fendly B, Kriegler M, Rosenberg SA: Murine tumor
cells transduced with the gene for tumor necrosis factor-alpha. J
Immunol 146:3227, 1991
30. Gansbacher B, Zier K, Cronin K, Hantzopoulos PA, Bouchard
B, Houghton A, Gilboa E, Golde D: Retroviral gene transfer induced
constitutive expression of interleukin-2 or interferon-gamma in irradiated human melanoma cells. Blood 802817, 1992
31. Gastl G, Finstad CL, Guarini A, Bosl G, Gilboa E, Bander
NH, Gansbacher B: Retroviral vector-mediated lymphokine gene
transfer into human renal cancer cells. Cancer Res 52:6229, 1992
32. Taylor CW, Grogan TM, Lopez MH, Leong SP, Odeleye A,
Feo-Zuppardi FJ, Hersh EM: Growth and dissemination of human
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
2558
malignant melanoma cells in mice with severe combined immune
deficiency. Lab Invest 67:130, 1992
33. Miller AD: Human gene therapy comes of age. Nature
357:455, 1992
34. Sun WH, Burkholder JK, Sun J, Culp J, Turner J, Lu XG,
Pugh TD, Ershler WB, Yang N-S: Roc Natl AcadSci USA 92:2889,
1995
35. Lindsten T, June CH, Ledbetter JA, Stella G, Thompson CB:
Regulation of lymphokine messenger RNA stability by a surfacemediated T-cell activation pathway. Science 244:339, 1989
36. Bickel M, Cohen RB, Pluznik DH: Post-transcriptional regulation of granulocyte-macrophage colony-stimulating factor synthesis in murine T cells. J Immunol 1452340, 1990
37. Gillis P, Malter JS: The adenosine-uridine binding factor recognizes the AU-rich elements of cytokine, lymphokine and oncogene
mRNAs. J Biol Chem 266:3172, 1991
RAJAGOPALAN ET AL
38. Zubiaga AM, Belasco JG, Greenberg ME: The nonamer UUAUUUAUU is the key AU-rich sequence motif that mediates mRNA
degradation. Mol Cell Biol 15:2219, 1995
39. Kislauskis EH, Zhu X, Singer RH: Sequences responsible for
intracellular localization of @-actinmessenger RNA also affect cell
phenotype. J Cell Biol 127:441, 1994
40. Ferrandon D, Elphick L, Nusslein-Volhard C, St. Johnson D:
Staufen protein associates with the 3’UTR of bicoid mRNA to form
particles that
move
in a microtubule-dependent manner. Cell
79:1221, 1994
41. Macdonald PM: Bicoid mRNA localization signal: Phylogenetic conservation of function and RNA secondary structure. Development 110:161, 1990
42. Smith JL, Wilson JE, MacDonald PM: Overexpression of
oskar directs ectopic activation of nanos and presumptive pole cell
formation in drosophila embryos. Cell 70:849, 1992
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
1995 86: 2551-2558
Granulocyte-macrophage colony-stimulating factor mRNA
stabilization enhances transgenic expression in normal cells and
tissues
LE Rajagopalan, JK Burkholder, J Turner, J Culp, NS Yang and JS Malter
Updated information and services can be found at:
http://www.bloodjournal.org/content/86/7/2551.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.