| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Molecular Genetics and Microbiology, University of Massachusetts Medical School, Worcester, MA 01655; and
Biogen, Cambridge, MA 02142
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Efficient adenovirus infection requires the presence of the
Coxsackie/adenovirus receptor (CAR) for virus attachment to the cell
surface and an integrin coreceptor,
vß3 or
vß5, which promotes
virus internalization (9, 10, 11, 12). Both human (h) and murine
(m) genes for CAR have been identified and, although they have a
different tissue expression pattern, either will allow infection by
human adenovirus when transfected into CAR-negative cell lines
(9, 10, 13, 14). Unfortunately, in the mouse, many
interesting target tissues, including cells of hemopoietic origin, do
not express CAR, limiting the use of recombinant adenoviruses for
studies of gene function or replacement in well-characterized murine
model systems of disease, particularly immunodeficiencies (9, 10, 13, 15).
To circumvent this problem, we generated transgenic mice that express hCAR. Transgene expression is controlled by the murine MHC class I H-2Kb promoter/enhancer, which has been shown to support nearly ubiquitous and constitutive transgene expression, including cells of hemopoietic origin (16, 17, 18). hCAR protein was readily detected on the surface of B and T lymphocytes from spleen, lymph nodes, bone marrow, thymus, and peritoneum. hCAR expression was sufficient to permit adenovirus infection and subsequent reporter gene expression in primary lymphocyte populations that were refractory to adenovirus infection when isolated from wild-type mice. We also found hCAR mRNA in tissues that do not express endogenous mCAR, demonstrating the potential to use adenovirus as a vehicle for gene delivery to primary cell types not previously susceptible to this method of gene transfer.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A cDNA clone encoding hCAR/pcDNA1 was generously provided by Dr.
Robert Finberg, Dana-Farber Cancer Institute (Boston, MA) (9, 10). The hCAR plasmid was digested with EcoRI
releasing a 2.5-kb fragment containing the full hCAR coding region.
This fragment was blunted, NotI linkers ligated, and cloned
into the NotI site of the H-2K-i-LTR (long terminal repeat)
expression cassette, which has a murine MHC class I promoter/enhancer
and Moloney murine leukemia virus polyadenylation signal (Fig. 1
A), and was generously provided by Dr. Irving Weissman
(Stanford University) (18). To assess protein expression
from the construct, CAR-negative Chinese hamster ovary cells were
transfected with linearized pH-2K-hCAR or empty vector using
Lipofectamine (Life Technologies, Grand Island, NY) following the
manufacturer’s protocol. Subsequently, the cell populations were
infected with adenovirus expressing ß-galactosidase (Adß-gal)
(multiplicity of infection (moi) = 10), and ß-gal expression was
determined using an Invitrogen kit (Invitrogen, Carlsbad, CA).
Twenty-four hours after infection, 40–95% of pH-2K-hCAR-transfected
Chinese hamster ovary cells stained positive for ß-gal production as
compared with only 2–5% of control cells. Endotoxin-free plasmid DNA
for transgenic production was purified using an EndoFree DNA extraction
kit (Qiagen, Valencia, CA). The H-2K-hCAR expression cassette was
excised by HindIII digestion, and purified by three phenol
extractions. Transgenic mice were produced on a C57BL/6 (B6) background
(The Jackson Laboratory, Bar Harbor, ME) by injection of pronuclei of
fertilized eggs at the UMMC transgenic facility. Transgene-positive
offspring were identified by PCR of tail DNA prepared using Qiagen DNA
prep kits. Two sets of PCR primers (Biosynthesis, Lewisville, TX) were
used for hCAR typing; neither primer set amplified mCAR sequences from
an mCAR cDNA clone or from cDNA prepared from B6 kidney. One primer set
(H-2K-hCAR) included sequences from the murine class I promoter and the
hCAR gene and produced a 600-bp fragment:
5'-AAAAGCCTCTCTCTCCACTG-3' and 5'-GGCAGTTTCCCCTTTGGCTT-3'. A second
primer set (hCAR) amplified a 365-bp fragment from within the hCAR
gene: 5'-CGCTCCTGCTGTGCTTCGTG-3' and 5'-ATCCATCAACGTAACATCTC-3'.
|
RNA was extracted from transgenic and control tissues using a Qiagen RNAeasy kit following the manufacturer’s instructions. Reverse-transcription reactions were conducted on DNase-treated RNA samples using Moloney murine leukemia virus reverse transcriptase, random hexamers (Promega, Madison, WI), and standard techniques, followed by PCR with the internal hCAR primer set. Actin was used as a positive control, and the PCR produced a 570-bp fragment. Actin: 5'-ATGGATGACGATATCGCT-3' and 5'-ATGAGGTAGTCTGTCAGGT-3'.
Cell preparation
Spleen, lymph node, and thymus cell suspensions were prepared (19, 20), pooling cells from at least three mice euthanized by CO2 inhalation. Peritoneal cells were obtained by two cycles of lavage with ice-cold balanced salt solution (BSS) supplemented with 0.3% BSA and penicillin (10 U/ml), streptomycin (10 µg/ml), and gentamicin (10 µg/ml) (Sigma, St. Louis, MO). Macrophages were depleted from peritoneal washes by incubating the cells for 2 h at 37°C on 100-mm tissue culture dishes in complete media to allow macrophage adherence. The nonadherent cells were removed by gently washing the plates with cold BSS. In some experiments, spleen cell populations enriched for resting B cells were prepared by incubating erythrocyte-depleted spleen cell populations on anti-CD43 (S7, 5 µg/ml)-coated tissue culture plates, for 1–2 h at 4°C and harvesting the nonadherent cells in cold BSS. Bone marrow cell suspensions were obtained from femurs by flushing with ice-cold medium and aspirating the tissue clumps through sequentially smaller bored needles. RBC were lysed by treatment with Gey’s solution.
Viruses and infection
AdCMV-GFP is a replication-defective recombinant adenovirus that
expresses green fluorescent protein (GFP) under the control of a human
CMV immediate/early promoter. Adß-gal is a recombinant adenovirus
expressing ß-gal under control of the human CMV promoter. Stocks of
these viruses were generously provided by Dr. Timothy Kowalik (UMMC).
AdEF1-GFP, a recombinant adenovirus with GFP driven by a human EF1
promoter (21), was generously provided by Dr. Jay Nelson
(Oregon Health Science University, Portland, OR). AdCMV-Bex has as a
reporter a GFP variant, GFP/Bex, that has been modified for more
efficient translation in mammalian cells and was constructed in our
laboratory using a modification of the AdEasy protocol described by He
and colleagues (URL: www.coloncancer.org) (4) (Wong et
al., manuscript in preparation). A KpnI/Xho restriction
fragment containing Bex was obtained from pucBex3, generously provided
by Dr. Rachel Gerstein (UMMC) and cloned into the AdEasy pShuttle
vector. The pShuttle/Bex was electroporated into electrocompetent
BJ5183 cells carrying the pAdEasy-1 vector and recombinants isolated by
antibiotic selection. CMV-Bex recombinants were transfected into L293A
cells and virus generated. Viruses were prepared by infection of L293A
cell monolayers and purified on CsCl2 gradients
(22). Lymphocyte suspensions were exposed to AdCMV-GFP,
AdCMV-Bex, or AdEF1-GFP at an moi of 50 for 2 h at 37°C in
serum-free RPMI 1640 media (Life Technologies), washed, and, in some
cases, treated with neutralizing anti-adenovirus antiserum (Access
Biomedical, San Diego, CA) for 30 min at 37°C to inactivate any input
virus not taken up by the cells, washed, and cultured in the presence
or absence of lymphocyte activators for 48 h before FACS
analysis.
Cell culture
Cells were cultured in RPMI 1640 media supplemented with 10%
FCS (Life Technologies), 2 mM glutamine, 10 U/ml penicillin, 10 µg/ml
streptomycin, and 5 x 10-5 M 2-ME. Spleen
cells, or B cell-enriched populations were cultured with B cell
activators: 50 µg/ml of Escherichia coli 055:B5 LPS, or 10
µg/ml anti-Ig (23) plus 100 U/ml of rmIL-4
(Biological Response Modifiers Unit (BRMU), National Institutes of
Health (NIH)), or PMA (10 ng/ml) plus ionomycin (1 µg/ml). Thymus
cells were cultured with 5 µg/ml Con A plus 25 U/ml of rmIL-2 (BRMU,
NIH), and lymph node cells were cultured with the T cell activator 5
µg/ml anti-CD3
(145-2C11) (PharMingen, San Diego, CA) plus 25
U/ml of rmIL-2. Bone marrow, peritoneal cells, or macrophage-depleted
peritoneal populations were infected as described above and cultured
without stimulation.
Surface biotinylation
hCAR cell surface expression was analyzed by immunoprecipitation of biotinylated surface proteins from lymphocyte subpopulations enriched by negative selection. B cell-enriched populations were prepared by adsorbing spleen cell suspensions on anti-CD43-coated plates; T cell-enriched populations were prepared similarly by adsorbing lymph node suspensions on anti-µ Ab (b7.6) (10 µg/ml)-coated tissue culture plates. In both procedures, cells were allowed to adhere for 1 h at 4°C, after which the nonadherent fraction was collected. The human kidney cell line, L293A, was used as a source of hCAR for the positive control. Populations were biotinylated using the cell-impermeable biotin derivative biotinamidocaproic acid 3-sulfo-N-hydroxy-succinaimide ester (Sigma), as described (24). Biotinylated lymphocyte populations and control 293 cells were then treated with lysis buffer: 0.5% Nonidet P-40, 0.5% DOC, 10 mM Tris, pH 8, plus protease inhibitors, as described (25). Immunoprecipitation was performed on cell lysates that were precleared with anti-IgG1 Ab-coupled agarose beads (Sigma) that had been incubated with control ascites. Precleared lysates were incubated with anti-hCAR mAb (RmcB; Ref. 26), generously provided by Dr. R. Finberg, bound to anti-IgG1 agarose beads, as previously described (19). Samples were washed and protein eluted from the beads by boiling in SDS-PAGE sample buffer and run on a 10% SDS-PAGE gel. Proteins were electrophoretically transferred to an Immobilon P membrane (Millipore, Bedford, MA) and developed with streptavidin-HRP and enhanced chemiluminescence (Amersham, Piscataway, NJ) (25).
Detection of hCAR, GFP, and integrin expression by FACS
Lymphocyte populations from transgene
(tg)+ mice were screened for hCAR expression by
cytofluorometric analysis (FACS). Briefly, erythrocyte-depleted
lymphocyte populations were incubated with a 1/100 dilution of control
or RmcB anti-hCAR ascites, washed, and incubated with anti-IgG1
FITC, fixed with 2% paraformaldehyde, and analyzed by flow cytometry
using either a FACScalibur or FACSvantage machine. GFP expression from
unfixed cell populations was determined on cells exposed to AdCMV-GFP,
AdEF1-GFP, or AdCMV-Bex and cultured for 48 h. Concurrent with the
GFP analysis, lymphoid cell subpopulations were distinguished by
staining with PE allophycocyanin or Cy5PE conjugates of: anti-B220
(CD45R) (RA3-6B2) (Caltag, Burlingame, CA), anti-Thy-1.2 (CD90)
(Caltag), anti-CD5 (53-7.3) (PharMingen), and anti-MAC-1
(CD11b) (Caltag). Integrin expression was analyzed by staining with
anti-v-PE or
anti-ß3-PE (PharMingen). FACS data were
analyzed using FlowJo software (Tree Star, San Carlos, CA). Data are
shown as contour plots at a 5% probability level. Median fluorescent
values (MFV), the 50th percentile of fluorescence of a given quadrant,
are used for comparison of GFP reporter expression.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Transgenic mice were produced using a construct in which the gene
encoding hCAR was placed under control of a MHC class I promoter
(H-2K-i-LTR) (Fig. 1
A)
(9, 18). This promoter was chosen because of the near
ubiquitous and constitutive expression of MHC class I protein on
somatic cells, especially cells of hemopoietic origin
(16). Moreover, while the constitutive amount of class I
can vary greatly between different tissues and among cell types within
a tissue, MHC expression can be modulated by cytokines
(27), potentially allowing the up-regulation of transgene
expression. Transgenic mice were generated by microinjection of
linearized plasmid into the pronucleus of fertilized eggs from B6 donor
mice and offspring typed by H-2K-hCAR- and hCAR-specific PCR of tail
biopsies. Five transgene-positive mice were identified, and four
transmitted the transgene and were used to establish founder lines by
mating with B6 mice.
The synthesis of transgene-encoded mRNA was determined by RT-PCR of
tissues from the transgenic lines. We analyzed tissues that express
endogenous mCAR, kidney, and those that do not, thymus, spleen, and
testis (13). hCAR mRNA was detected in lymphoid tissues,
spleen, and thymus, as well as in testis and kidney from founder line,
28 (Fig. 1B). In the other founders, hCAR expression was
variable; all expressed hCAR in the testis, but only three expressed
hCAR in the kidney. hCAR mRNA was not detected in any tissue from
normal B6 mice, despite the fact that kidney cells produce highly
homologous mCAR. Because all the tissues chosen for analysis express
MHC class I protein (data not shown) (27), regulatory
effects exerted on the inserted hCAR transgene by flanking genomic
sequences most likely account for the variable tissue expression
observed.
Transgenic lymphocytes express hCAR protein
hCAR protein expression on lymphoid populations from the
transgenic lines was directly assessed by surface staining with
anti-hCAR mAb and FACS analysis (RcmB; Ref. 26). hCAR
protein is readily detected on cells from both primary and secondary
lymphoid organs of the hCAR+ line 28 (Fig. 2A). hCAR protein is expressed
on all thymocytes and bone marrow cells, whereas spleens and lymph
nodes contain a small, and as yet unidentified, subpopulation of
hCAR-negative cells in addition to the major hCAR-positive population.
hCAR expression is activation independent, as surface hCAR is not
increased or decreased significantly when hCAR+
lymphocytes were activated by mitogens in vitro (data not shown).
Consistent with the RT-PCR analysis of mRNA, primary and secondary
lymphoid populations taken from normal B6 mice and the three other
founder lines, shown here for B6 bone marrow (Fig. 2A), did
not stain with anti-hCAR Ab.
|
B). hCAR protein was not detected on splenic lymphocytes
from the other founder lines (line 31 is shown) or B6 lymphocytes, a
finding consistent with our RT-PCR and FACS analyses.
Efficient infection with adenovirus is critically dependent on CAR and
facilitated by an integrin coreceptor,
Vß3 or
Vß5 (12).
We used FACS to assay for integrin because of conflicting reports
concerning v expression on resting lymphocytes
(28, 29, 30). Our analysis showed v
integrin is readily detected and is at equivalent amounts on
lymphocytes from control and transgenic animals, shown here for
transgenic and nontransgenic splenic B220+ cells
(Fig. 2C). Integrin ß3 was also
detected on resting lymphocytes by FACS analysis (Fig. 2C).
Therefore, the integrin coreceptor that facilitates adenovirus uptake
is available for adenovirus infection of lymphocyte populations.
Adenovirus infection and reporter expression in hCAR+ lymphocytes
To test whether expression of hCAR renders murine lymphocytes
susceptible to adenovirus, we prepared cells from spleen, lymph node,
thymus, bone marrow, and peritoneal cavity of hCAR line 28 and control
C57BL/6 mice and exposed them to AdCMV-GFP, a rAdV encoding a GFP
reporter gene under control of a hCMV promoter. After virus exposure,
cells were washed and cultured for 48 h in the presence or absence
of B or T cell activators. In some experiments, cells were also treated
briefly with neutralizing anti-AdV Ab and washed before culture.
Thereafter, cultured infected control or transgenic cell populations
were stained for B cell (B220), T cell (Thy-1.2), or macrophage (MAC-1)
surface Ags and analyzed by FACS. GFP+
lymphocytes were readily detected in all lymphoid tissues from the hCAR
transgenic donors, but not from the nontransgenics (Fig. 3). Furthermore, there was no difference
in this outcome when lymphocyte populations were treated with
anti-AdV Ab before activation.
|
A). Although GFP expression was detected, activation in
vitro after AdV infection was required for reporter expression, as
GFP+ cells were not found in unstimulated splenic
B cells from either transgenic or normal mice (see Fig. 5). We also
found the number of GFP-expressing cells was not increased when hCAR tg
or normal B6 splenic B cells were activated with anti-Ig plus IL-4
for 2 days before AdV exposure (data not shown), indicating that
lymphocyte activation per se does not facilitate infection. The
requirement for postinfection activation for GFP expression in B cells
was not absolute. B220+ B lineage cells from the
bone marrow of transgenic mice, cultured without stimulation,
demonstrate a high percentage of GFP+ cells after
AdCMV-GFP infection, 37%, compared with control B6 bone marrow
B220+ cells, 3% (Fig. 3B). More
impressively, unstimulated peritoneal B1a lymphocytes
(B220+, CD5+,
MAC-1low) had a high frequency of
GFP+ cells (53%, Fig. 3C) and very
robust reporter expression (MFV of 651) after infection.
|
plus
IL-2 demonstrate a marked increase in GFP fluorescence over that seen
with nontransgenic T cells, 15.6% vs 2.5% (Fig. 3D).
Similarly, thymic T cells, cultured with Con A plus IL-2, support
adenovirus infection as evidenced by a large fraction of
GFP+, Thy-1+ cells, 35%
(Fig. 3E).
Bone marrow and peritoneal cell populations are heterogeneous tissues
containing cell types other than lymphocytes. Because hCAR is expressed
on all cells within the bone marrow (Fig. 2A), we next
determined whether hCAR expression renders the nonlymphoid cells within
this complex population susceptible to infection. Cells from bone
marrow were infected with AdCMV-GFP and cultured without activators for
48 h. Nonlymphoid cells were identified by electronic gating using
forward and side scatter parameters, of populations negative for B220
and Thy-1.2. The data in Fig. 4A show that a high proportion
of these cells from hCARtg+ donors (58%) is
readily infected by AdCMV-GFP; in contrast, only 5% of a similar
population from B6 donors is susceptible to infection.
|
B)
that uninduced B6 peritoneal cells were efficiently infected (60%)
with our AdCMV-GFP construct. Likewise, peritoneal cells from
hCAR+ donors were readily infected; while a
higher proportion was GFP+ (85%, Fig. 4B), the level of GFP expression did not differ
substantially. This suggests that the presence of the transgene does
not interfere with endogenous mCAR function and/or may potentiate an
already efficient infection. The intensity of reporter expression in infected B cells is modulated by vector promoter and lymphocyte stimulus
One explanation for the finding that follicular splenic (B2) B
cells require postinfection activation for GFP expression, but
peritoneal B1a cells do not is that transcription factors necessary for
activating the hCMV promoter are not available in resting murine B2
cells. Alternatively, small resting B cells may lack the protein
synthetic capability necessary for efficient reporter expression. In
this regard, it is noteworthy that B1a cells are more biosynthetically
active and larger than resting B2 cells (32). To
distinguish between these possibilities, we infected resting splenic B
cells with AdV constructs that have GFP driven by either a CMV or
EF1 promoter, known to have different efficiencies in some cell
lines, and used an adenovirus carrying a modified GFP gene that is more
readily translated in mammalian cells (AdCMV-Bex). After infection, B
cells were stimulated with anti-Ig + IL-4, LPS, or PMA plus
ionomycin, as these activators cause cell enlargement and drive B cell
proliferation while mobilizing both overlapping and unique
transcription factors (33, 34, 35, 36).
This analysis (Fig. 5) shows that GFP
expression demonstrates a strong promoter preference. Unstimulated
hCAR+ B cells, infected with AdEF1-GFP, produced
GFP more efficiently when compared with AdCMV-Bex-infected cells (16%
vs 5% of B220+ cells), although the median level
of GFP expression was similar between the two groups (MFV of 86 vs 63,
respectively). This suggests that the higher fraction of
GFP+ cell detected is not merely a reflection of
more efficient reporter expression. As shown previously, activation
with anti-Ig plus IL-4 increased GFP+ cells;
18% of AdCMV-Bex-infected hCAR+ B cells
expressed GFP, however, at a level only slightly higher than that found
in unstimulated cells (MFV of 74). The use of the Bex variant (Fig. 5)
rather than wild-type GFP (Fig. 3) slightly increased the intensity of
fluorescence-observed (MFV of 46 in Fig. 3 vs 74 in Fig. 5), but did
not increase the fraction of B220+
GFP+ cells detected (18% vs 19%, respectively).
In contrast, AdEF1-GFP-infected hCAR+ B cells had
a very high proportion of GFP-expressing cells, 67% of
B220+ cells, with a significant increase in
reporter intensity (MFV of 282). hCAR B cells stimulated with PMA and
ionomycin also demonstrated differences in promoter efficiency. The
percentage of GFP+ B cells was lower in AdCMV-Bex
vs AdEF1-GFP-infected cells (24 vs 51% of B220+
cells, respectively), although the MFV was similar between these groups
(133 vs 149). It should be noted that the differences in GFP expression
are host cell dependent; in a permissive cell line, L293A, both AdV
constructs efficiently express GFP (data not shown).
The data in Fig. 5 also show that B cell activation per se is not
sufficient for reporter production, as LPS fails to substantially
increase either the number of GFP+ cells or the
level of GFP over that seen in unstimulated cells infected with either
the CMV or EF1 construct. This despite the fact that both cell cycle
analysis (data not shown) and [3H]thymidine
incorporation show LPS is as potent an activator of B cells as are
anti-Ig plus IL-4 or PMA plus ionomycin (Fig. 5). As discussed
previously, B6 B cells were not susceptible to infection by either
construct, as GFP+ cells were not
found in either unstimulated or anti-Ig plus IL-4-stimulated
populations (Fig. 5). Taken together, the data are consistent with the
notion that activation is not the critical variable, rather, the
available transcription factors limit the efficiency of promoter
expression.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MHC class I proteins are constitutively expressed on almost all adult somatic cells, although, among different tissues, protein levels can differ by as much as 100-fold (16, 17). Because cells in the hemopoietic lineage have the highest level of class I protein, it was unexpected that hCAR driven by a murine class I promoter was found on lymphocytes from only one of four founder tg+ lines. The H-2Kb promoter in our H-2K-i-hCAR construct has been used previously to produce transgenic mice expressing human growth hormone; in these studies, both tissue distribution and the relative abundance of transgene mRNA closely followed that of H-2Kb mRNA (17). Moreover, the same expression cassette we used was used previously to direct hBcl-2 expression in T and B lymphocytes in other transgenic mice (18). Because hCAR mRNA was found in testis tissue from all founder lines, the restricted expression of hCAR that we find is most likely the result of tissue-specific regulation. This regulation could result from a positional effect modifying transgene accessibility or through close acting negative regulatory elements. Because the H-2K promoter is responsive to transcription factors induced in cells exposed to cytokines (27), hCAR protein may be induced on lymphocytes or other cells from founder lines that do not constitutively express the hCAR transgene. This possibility is under investigation.
In our studies, AdV-infected murine T cells require stimulation before reporter expression. In contrast, it has been reported that resting human peripheral blood T lymphocytes (PBTL) will express an AdV-delivered ß-gal reporter gene driven by a hCMV promoter (37). In this study, PBTL were transduced by a Ad-ßgal vector targeted to T cells by a bispecific Ab that recognized a CD3 chain on the PBTL TCR, and a FLAG epitope engineered into the Ad penton base coat protein (AdZ.FLAG). There are a number of possible reasons for the apparent differences between our results. First, the hCMV promoter may be activated by transcription factors present in resting human lymphocytes, but absent from murine lymphocytes. It has been shown that the mCMV promoter is more efficient in rodent cells than the hCMV promoter (38, 39, 40). Second, the conversion of ß-gal may be a more sensitive detected product than fluorescence of GFP because of the amplification of the readout by the enzymatic activity of ß-gal. Third, CD3 molecules targeted by the bispecific anti-CD3/FLAG Ab could be cross-linked upon AdZ.FLAG binding, thereby activating the T cells. In our experiments, activation by anti-CD3 cross-linking induces GFP expression in murine T cells. It is noteworthy that in agreement with our results, phorbol ester markedly increases reporter expression, in tranduced PBTL.
Other lymphocyte subpopulations also require activation for GFP
expression. We posit that activation is required to drive reporter gene
expression from the vector, rather than changing the susceptibility of
lymphoid cells to infection or increasing the protein synthetic
capability of small resting lymphocytes. This position is based on a
number of observations. First, resting transgenic B and T cells express
both hCAR and the vß3
integrin necessary for AdV attachment, uptake, and uncoating. Second,
when unstimulated lymphocytes were exposed to AdV and treated
subsequently with anti-AdV Ab to neutralize any uninternalized
virus before activation, no difference in the percentage of
GFP+ cells was observed between Ab-treated and
untreated populations. Third, there was no substantial increase in the
number of GFP+ cells when B cells were activated
for 2 days before AdV infection. Fourth, B6 lymphocytes failed to
express GFP, irrespective of activator exposure, despite the fact that
they express vß3
integrin involved in AdV internalization. These results show that only
transgenic cells have all the requisite molecules for efficient AdV
attachment and uptake, and the activation of resting cells does not
increase the efficiency of infection. We also demonstrated that
lymphocyte activation per se is not sufficient to drive GFP expression.
The proportion of B cells expressing GFP after infection with AdCMV-GFP
or AdEF1-GFP was not substantially increased over unstimulated cells by
activation with LPS, despite the fact that LPS induced more
proliferation and a higher fraction of cells in cycle than any of the
other activators used in these experiments (Fig. 5).
There is a strong promoter preference for reporter expression in infected B2 cells. Unstimulated B2 cell populations infected with a construct in which the reporter was driven by the EF1 promoter had as many GFP+ cells as did activated B2 cells infected with a construct using the CMV promoter. This promoter preference was also demonstrable after activation; AdEF1-GFP-infected hCAR+ B cells had more GFP+ cells, with a higher amount of GFP per cell than the AdCMV-Bex-infected B2 population comparably activated. In contrast, promoter preference was not seen in B1a cells; a majority of unstimulated peritoneal B1a cells express GFP, at high levels after infection with either the AdCMV-GFP or AdEF1-GFP constructs. Although B1a cells are considered resting by numerous criteria, they are capable of self renewal, and unlike B2 cells, they have STAT-3 mobilized to the nucleus (41, 42). These characteristics support the possibility that differences in important and as yet unidentified transcription factors exist between the B1a and B2 subpopulations and may explain why unstimulated B1 cells express reporter more efficiently than resting B2 cells. Moreover, the finding that both AdCMV-Bex and AdEF1-GFP constructs give equivalent GFP expression in permissive cells, peritoneal B1 cells, and the L293 cell line shows there is no intrinsic difference between the efficiency of reporter expression provided the appropriate transcription factors are available.
Primary lymphocytes are notoriously poor targets for gene transfer. We believe the hCAR transgenic will complement retroviral and lentiviral gene transfer systems. Some advantages of the AdV transfer system include: the potential for superinfection with different AdV constructs, vectors that accommodate large inserts, are produced in high titer, and can direct the production of large amounts of protein from the inserted gene(s). Moreover, gene dose can be titered by changing the moi. For in vitro studies, this allows a direct analysis of gene function in biosynthetic and regulatory pathways controlling proliferation, differentiation, and death of normal cells using freshly isolate lymphocytes rather than transformed cell lines or clones selected for long-term survival in culture. Because receptor expression in heterozygotes is sufficient to support AdV infection and the founder is an inbred B6, AdV-susceptible lymphocytes can be easily generated on defined genetic backgrounds by a single cross, and animals with recessive genetic defects or targeted mutations can, in most cases, be generated by a simple backcross. This provides the opportunity to rapidly model corrective gene therapy approaches in vitro. For example, we are currently examining how infection with AdV-BTK (Bruton’s tyrosine kinase) constructs alters the function of B cells with the X-linked immune defect (xid). We appreciate one potential impediment to a fully realized in vitro model is the efficiency of AdV gene expression in resting lymphocytes. While many insights can be developed on stimulated lymphocytes, which we have shown express AdV-encoded genes efficiently, other analytical approaches, for example the dissection of a regulatory pathway using dominant-negative mutants, would initially require expression of the mutant protein in quiescent cells. We think it will be a simple matter to identify the appropriate promoters to engineer into AdV constructs by their association with proteins expressed in resting cells. MHC class I, CD20, and Eµ/LCK promoters/enhancers are all currently being evaluated for their ability to direct gene expression in resting lymphocytes.
The transgenic model also provides an opportunity to model gene
replacement strategies in an in vivo environment more closely
approximating the clinical setting. Gene replacement could be affected
by transfer of cells infected ex vivo with AdV-encoding drugs or
enzymes targeting metabolic disorders, as well as by systemic delivery
of an AdV vector. In the former case, we are intrigued by our
observation that B1a cells are efficiently infected and express
AdV-encoded genes robustly from both CMV and EF1 promoters without
the need for lymphocyte activation. These properties combined with the
self-renewing potential of these cells and their sequestered anatomical
location (41) may make them ideal test vehicles for
delivery of therapeutic genes. Because AdV does not readily integrate,
the durability of AdV-mediated gene expression in replicating cells
may, in some cases, be a practical concern, as the therapeutic gene
would be diluted out during cell division. However, AdV vectors are
available that can replicate their genome during cell division without
the production of infectious particles (8), thus allaying
this concern.
Finally, we also find that the hCAR transgene is expressed in tissues that do not normally express endogenous mCAR. This suggests to us that AdV-mediated gene delivery can be extended to organs and tissues previously refractory to transduction.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Madelyn R. Schmidt, Department of Molecular Genetics and Microbiology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655.
3 Abbreviations used in this paper: AdV, adenovirus V; ß-gal, ß-galactosidase; BSS, balanced salt solution; CAR, Coxsackie/adenovirus receptor; EF1, elongation factor 1; GFP, green fluorescent protein; h, human; LTR, long terminal repeat; m, murine; MFV, mean fluorescent value; moi, multiplicity of infection; PBTL, peripheral blood T lymphocyte.
Received for publication May 3, 2000. Accepted for publication July 12, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
vß3 and vß5 promote adenovirus internalization but not virus attachment. Cell 73:309.[Medline]
v integrins in adenovirus cell entry and gene delivery. Micro. Mol. Biol. Rev. 63:725.v and ß3 integrin chains on murine lymphocytes. Proc. Natl. Acad. Sci. USA 93:14698.
B components in primary B cells. Int. Immunol. 10:285.This article has been cited by other articles:
|
|
 |
|
  C. A. Shaw, N. Larochelle, R. W.R. Dudley, H. Lochmuller, G. Danialou, B. J. Petrof, G. Karpati, P. C. Holland, and J. Nalbantoglu Simultaneous Dystrophin and Dysferlin Deficiencies Associated with High-Level Expression of the Coxsackie and Adenovirus Receptor in Transgenic Mice Am. J. Pathol., December 1, 2006; 169(6): 2148 - 2160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.-L. Toh, S.-S. Hong, F. v. d. Loo, L. Franqueville, L. Lindholm, W. v. d. Berg, P. Boulanger, and P. Miossec Enhancement of Adenovirus-Mediated Gene Delivery to Rheumatoid Arthritis Synoviocytes and Synovium by Fiber Modifications: Role of Arginine-Glycine-Aspartic Acid (RGD)- and Non-RGD-Binding Integrins J. Immunol., December 1, 2005; 175(11): 7687 - 7698. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Zhang and J. M. Bergelson Adenovirus Receptors J. Virol., October 1, 2005; 79(19): 12125 - 12131. [Full Text] [PDF]
|
|
|
|
|
|
F. Kuhnel, B. Schulte, T. Wirth, N. Woller, S. Schafers, L. Zender, M. Manns, and S. Kubicka Protein Transduction Domains Fused to Virus Receptors Improve Cellular Virus Uptake and Enhance Oncolysis by Tumor-Specific Replicating Vectors J. Virol., December 15, 2004; 78(24): 13743 - 13754. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Bhattacharya, E. C. Logue, S. Bakkour, J. DeGregori, and W. C. Sha Identification of gene function by cyclical packaging rescue of retroviral cDNA libraries PNAS, June 25, 2002; 99(13): 8838 - 8843. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Cohen, Z. Q. Xiang, G.-P. Gao, H. C. J. Ertl, J. M. Wilson, and J. M. Bergelson Chimpanzee adenovirus CV-68 adapted as a gene delivery vector interacts with the coxsackievirus and adenovirus receptor J. Gen. Virol., January 1, 2002; 83(1): 151 - 155. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Vos, A. Neubert, E. Pommerening, T. Muller, L. Dohner, L. Neubert, and K. Hughes Immunogenicity of an E1-deleted recombinant human adenovirus against rabies by different routes of administration J. Gen. Virol., September 1, 2001; 82(9): 2191 - 2197. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Tallone, S. Malin, A. Samuelsson, J. Wilbertz, M. Miyahara, K. Okamoto, L. Poellinger, L. Philipson, and S. Pettersson A mouse model for adenovirus gene delivery PNAS, July 3, 2001; 98(14): 7910 - 7915. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
