| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Division of Infection and Immunity, Institute of Biomedical and Life Sciences, and
Western Infirmary, Department of Immunology and Bacteriology, University of Glasgow, Glasgow, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Many studies have shown that in general, DNA vaccines injected i.m. elicit strong CTL responses and humoral responses characterized by serum IgG2a (indicating a predominant Th1-type response) (8, 9, 10, 11, 12). These responses are typically elicited following multiple doses of plasmid DNA over an extended period (often many weeks), by which time the initial cellular interactions have already occurred. Altering this default response and changing the character and magnitude of the response is in theory relatively simple using a DNA immunization approach, and many studies have shown that immune responses can be manipulated by strategies including: 1) altering the cellular location of expressed Ag (13, 14, 15, 16); 2) coexpression of cytokines or chemokines (17, 18, 19); 3) targeting Ag to MHC class II- or I-processing pathways (20, 21, 22); and 4) inducing apoptosis in and promoting phagocytosis of transfected cells (23, 24) as well as many others.
Ag form and load influence both the magnitude of initial Ag-specific T cell expansion and the size of the memory cell pool following clonal contraction (25, 26, 27, 28, 29). Hence, these Ag-related factors may be important considerations when designing DNA vaccines. Several studies have indicated that the subcellular location of plasmid-encoded Ag expression (i.e., the Ag form first seen by the APC) has both quantitative and qualitative effects on immune responses, and in some cases protective efficacy (13, 14, 15, 16). Some studies have demonstrated that differential Ag localization influences both CTL responses and serum Ab subtypes, suggesting the preferential induction and proliferation of different Th subsets (13, 14, 15, 16). However, the initial cellular events that determine these outcomes have not been elucidated. One of our objectives was to investigate whether Ag localization (i.e., form) influenced the ability of DNA vaccines to prime MHC class I- and II-restricted naive T cells.
T and B cell priming events have in the past been difficult to study due to the low precursor frequency of Ag-specific naive cells. However, adoptive transfer of small numbers of traceable naive transgenic (Tg)3 lymphocytes of known peptide specificity into normal syngeneic recipients has allowed many workers to investigate lymphocyte priming following immunization with immunogenic and tolerogenic Ags (30, 31, 32, 33). We have applied these techniques for tracking T cells in vivo following immunization with DNA constructs expressing OVA in various cellular locations. We show that DNA-expressed cell-associated Ag primes CD8+ T cells more efficiently than expressed soluble Ag, while exogenous and membrane-associated Ag primes CD4+ T cells better than cytosolic OVA forms. Furthermore, we demonstrate that simple construct manipulation, such as inclusion of an intron to increase gene expression, also influences the efficiency of T cell priming. Thus, both subcellular localization and Ag load are crucial influences on T cell priming following DNA injection. Understanding precisely how construct manipulations such as these can influence the inductive mechanisms and the cellular interactions underlying them may enhance our ability to engineer DNA vaccines that induce rapid and pathogen-appropriate recall responses.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
BALB/c and C57BL/6 (B6) mice were purchased from Harlan Olac
(Bichester, U.K.) and used between 6 and 12 wk of age. DO11.10
TCR Tg mice have been described previously (34, 35) and
contain CD4+ T cells that express a TCR that
recognizes the chicken OVA (cOVA) peptide from aa 323–339 complexed
with MHC class II molecule I-Ad (detected by the
clonotypic mAb KJ1-26; Ref. 35). OT-I
CD8+ Tgs express a V
2+
TCR that recognizes OVA257–264 complexed with
H-2Kb (36). All animals were housed
at the University of Glasgow Central Research Facility (Glasgow, U.K.),
and all procedures performed according to U.K. Home Office
regulations.
Cell lines
BALB/c RAW264.7 macrophages (European Collection of Cell Cultures, Salisbury, U.K.) were cultured in DMEM supplemented with 10% FCS, 2 mM glutamine, 100 U/ml penicillin, and 10 µg/ml streptomycin. DO11-green fluorescent protein (GFP) hybridoma (37) was cultured in RPMI supplemented as above. DO11-GFP is a T cell hybridoma with the gene encoding GFP under the control of the NFAT promoter so that upon activation through the TCR the gfp gene is transcribed and activated cells become fluorescent. Activated cells can be detected by flow cytometry using the FITC channel (FL1). COS-7 monkey kidney fibroblasts (European Cell Culture Collection) were cultured as for RAW cells. All cultures were incubated at 37°C in 5% CO2.
Plasmid construction
The eukaryotic expression vector pcDNA3.1 (Invitrogen, San
Diego, CA) contains the human CMV virus Immediate/early promoter and
enhancer, the bovine growth hormone polyadenylation signal, and
neomycin and 
-lactamase genes for selection in mammalian
cells and Escherichia coli, respectively. cOVA cDNA was
kindly provided by N. Glaichenhaus (Centre National de la Recherche
Scientifique, Institut de Pharmacologie
Moléculaire et Cellulaire, Valbonne, France), and the
full-length native OVA was cloned in pcDNA3.1 to give plasmid
expressing native OVA (pNatOVA). Fig. 1
A shows all constructs used
in this study and the location of I-Ad MHC class
II- and H-2Kb MHC class I-restricted epitopes,
respectively. A plasmid designed for plasmid expressing cytosolic OVA
(pCytOVA) expression was constructed by omitting the first five codons
of the native OVA cDNA. An alternative cytosolic expression plasmid was
constructed as described previously (13) in which the
central SacI fragment (encoding aa 19–144) of the native
OVA cDNA was deleted. Because native OVA has an atypical secretion
sequence (38), we used a heterologous leader to direct OVA
to the classical secretion pathway via the endoplasmic reticulum
and Golgi. The plasmid pOVA-IFN-
(T. S. Kim, Chonnam National
University, Kwangju, South Korea) that contains the human Ig 
signal leader fused to OVA cDNA (17) was used as template
to construct the plasmid expressing secreted OVA (pSecOVA) in
pcDNA3.1. OVA expressed using this construct is secreted in mammalian
cells (17). A plasmid expressing the transmembrane region
of the human transferrin receptor (TfR) fused to aa 149–385 from
native OVA (plasmid expressing transferrin-OVA fusion protein; pTfROVA)
has been described previously (39, 40) and was
kindly donated by M. Zenke (Max-Delbück-Center for Molecular
Medicine, Berlin, Germany). This construct has been shown by others to
target OVA expression to the surface of mammalian cells in vitro and in
vivo (39, 41). The chimeric intron from pCI-neo (Promega,
Madison, WI), which is composed of the 5'-donor site from the first
intron of the human -globin gene and the branch and 3'-acceptor site
from the intron of an Ig H chain variable region was cloned at the
5'-end of the full-length OVA cDNA to give pCINatOVA. Plasmids were
prepared for transfection and immunization using Qiagen Endotoxin Free
Plasmid Maxi- and Mega-prep kits (Qiagen, Valencia, CA) according to
the manufacturer’s instructions.
|
RAW cell macrophages were seeded at 2 x
105 cells/ml in 24-well plates, activated for
48 h before transfection by the addition of murine IFN- (R&D
Systems, Minneapolis, MN) to a final concentration of 5 ng/ml, and
cultured to 90% confluence at 37°C in 5% CO2.
Transfection complexes were prepared using 15 µg/ml Lipofectamine
reagent (Invitrogen) and 3 µg/ml endotoxin-free plasmid DNA in
serum-free DMEM and allowed to form at room temperature for 30 min.
Before the addition of complexes, monolayers were washed in serum-free
DMEM. Macrophages were used 24 h following transfection in
DO11-GFP presentation assays. COS-7 fibroblasts were transfected
essentially as described for RAW macrophages. OVA protein was
demonstrated in cell lysates and cell-free culture supernatants by PAGE
and immunoblot using OVA-specific rabbit serum (Sigma-Aldrich, St.
Louis, MO), followed by alkaline phosphatase-conjugated sheep
anti-rabbit serum (Serotec, Oxford, U.K.) and
5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium
(Sigma-Aldrich) as the substrate. Serial dilutions
(log2) of cell lysates and supernatants were
spotted onto nitrocellulose membranes and probed with OVA-specific
serum to compare OVA expression levels following transfection with
different OVA constructs.
DO11-GFP presentation assay
A total of 2 x 105 DO11-GFP hybridoma cells were added to transfected RAW cell monolayers in 24-well plates 24 h posttransfection. Plates were centrifuged at 450 x g for 5 min to initiate cell contact and incubated for a further 24 h. Cells were collected for flow cytometry in PBS containing 1 mM EDTA, washed in FACS buffer (PBS, 2% FCS, 0.05% sodium azide), blocked in FcR blocking buffer (anti-CD16/32 hybridoma supernatant (clone 2.4G2), 10% mouse serum, 0.1% sodium azide), and stained with biotinylated KJ1-26 mAb (30) followed by PE-conjugated streptavidin (SA-PE; BD PharMingen, San Diego). The KJ1-26+ population was analyzed for GFP fluorescence by flow cytometry using the FL1.
Adoptive transfer
In adoptive transfer experiments, spleens and lymph nodes from
DO11.10 or OT-I mice were collected and single-cell suspensions
prepared as described previously (31, 32). Flow cytometry
was used to determine the percentage of DO11.10 and OT-I Tg cells.
DO11.10 cells were identified in donor mice using PE-conjugated
anti-mouse CD4 (BD PharMingen) and biotin-KJ1-26 followed by
FITC-conjugated streptavidin (SA-FITC; Vector Laboratories, Burlingame,
CA). OT-I cells were identified with PE-conjugated anti-mouse CD8
(BD PharMingen) and biotinylated anti-V2 (BD PharMingen) and
SA-FITC. Approximately 3 x 106 Tg cells in
0.2 ml were transferred i.v. via the tail vein into either BALB/c or
B6-recipient mice, respectively, 1 day before immunization.
CFSE labeling
CFSE (Molecular Probes, Eugene, OR) labeling of cells from DO11.10 and OT-I mice was performed as described previously (42). Briefly, spleen and lymph node cells were suspended in HBSS at 5 x 107 cells/ml and incubated in CFSE at a final concentration of 5 µM for 10 min at 37°C. Cells were then washed twice in HBSS, stained to determine the percentage of Tg cells, and resuspended in RPMI 1640 for adoptive transfer as described above.
Immunization
Mice were immunized in both tibialis anterior muscles (i.e., two
sites) with 25 µg plasmid DNA diluted in endotoxin-free PBS in a
50-µl final volume 1 day after lymphocyte transfer and did not
receive any further boost injections. Injections were done using a
27.5-gauge tuberculin syringe fitted with a plastic needle collar to
adjust the injection depth to 
2 mm. cOVA (100 µg; Fraction V;
Sigma-Aldrich) emulsified in CFA, was administered s.c. in the scruff
of the neck 1 day after cell transfer. In some experiments,
mice were immunized with plasmid DNA as above without prior cell
transfer. Blood was collected from the tail vein at various points
postimmunization for analysis of serum Abs. In some experiments, mice
were boosted (as above) 3 wk after priming, and draining lymph
nodes and spleens were collected after a further 4 wk (day 50) for in
vitro restimulation and cytokine analysis.
FACS analysis of lymph node suspensions
Draining popliteal and inguinal lymph nodes were removed at
various times after plasmid immunization. Single-cell suspensions were
prepared and blocked in FcR blocking buffer. Clonal expansion was
analyzed following staining with PE-anti-mouse CD4, biotin-KJ1-26,
and SA-FITC for DO11.10 transfer experiments or PE-anti-mouse CD8,
biotin-anti-V2, and SA-FITC for OT-I experiments. Three-color
flow cytometry was used to demonstrate CFSE-labeled DO11.10 T cells by
staining with PerCP-conjugated anti-mouse CD4, biotin-KJ1-26, and
SA-PE. The CFSE fluorescence of 1000
CD4+KJ1-26+ events was
measured using FL1. OT-I T cells were stained using
PerCP-anti-mouse CD8, biotin-V2, and SA-PE and CFSE fluorescence
analyzed as above on gated
CD8+V2+ lymphocyte
populations. Division markers were defined using cells from mice
vaccinated with OVA in CFA. The early activation marker CD69 was
demonstrated on OT-I cells using FITC-conjugated anti-CD69 and
FITC-conjugated hamster IgG as an isotype control. The percentage of T
cell blasts was determined by increased forward scattering of
CD4+KJ1-26+ or
CD8+V2+ cell
populations.
In vitro restimulation assays
Draining lymph nodes and spleens were collected from mice at
various times postimmunization. Single-cell suspensions were prepared
in complete RPMI 1640 and 4 x 106
lymphoid cells from immunized mice were restimulated in vitro with 1
µM OVA257–264 peptide in 24-well plates.
Culture supernatants were collected after 72 h and analyzed for
IFN- production by capture ELISA.
Intracellular IFN- staining
Lymphoid cells from popliteal and inguinal lymph nodes of
OT-I-transferred plasmid-vaccinated mice were cultured for 4 h in
vitro with 1 µM OVA257–264 peptide as
described above, but in the presence of 10 µg/ml brefeldin A
(Sigma-Aldrich). Following restimulation, cells were collected and the
surface markers CD8 and V2 TCR were stained as described previously.
Cells were fixed and permeabilized using PermaCyte-FP buffers
(BioErgonomics, St. Paul, MN) according to the manufacturer’s
instructions, and aliquots were incubated in PE-conjugated
anti-IFN- or PE-conjugated rat IgG1 as isotype control (both BD
PharMingen). Three-color flow cytometry was used to demonstrate IFN-
staining in 2000
CD8+V2+-gated
lymphocytes.
ELISAs
ELISAs for OVA-specific serum IgG2a and IgG1 were performed
essentially as described previously (32) using
o-phenylenediamine (Sigma-Aldrich) as substrate. ELISAs for
quantification of IFN- in the supernatants of in vitro restimulated
lymph node and splenic cell suspensions was done as described
previously (32).
Statistics
Group means were represented ± SEM when groups contained three or more animals and as ± range when n = 2. Differences between groups were analyzed using Student’s two-tailed t test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
OVA expression in transfected COS-7 fibroblasts was analyzed by
dot immunoblot (Fig. 1
B) and PAGE/immunoblot (data not
shown) of whole-cell lysates and cell-free culture supernatants using
an OVA-specific rabbit serum. Fig. 1B is a dot blot showing
serial doubling dilutions of cell lysates normalized for total
protein content and supernatants probed with OVA-specific serum.
pTfROVA and pCytOVA transfectants produce a predominantly
cell-associated OVA, whereas pSecOVA transfectants secrete the majority
of the OVA protein. Although this is not a quantitative assay, it
allows us to conclude that the amount of Ag expressed by the different
constructs is at least comparable in vitro. Therefore, differential
immune responses observed are likely to be due to the cellular location
of the expressed protein rather than due to differences in Ag
dose.
DNA vaccination induces OVA-specific serum IgG following a single i.m. injection
Vaccination with plasmids that targeted OVA expression to
different cellular locations induced OVA-specific IgG2a 21 days after a
single priming immunization (Fig. 2). The
plasmid pSecOVA induced more IgG1 than other OVA forms. Serum Abs were
detected as early as 13 days after priming (data not shown), and
subsequent plasmid boosts further increased the IgG2a and IgG1 levels
(data not shown). These results suggest that the subcellular
localization of OVA expression influences the magnitude and character
of the humoral response and these are determined early in the
developing immune response.
|
As we had some preliminary indication that vaccination with our OVA constructs induced an immune response, we next attempted to define some of the underlying cellular events in vivo. As activation of naive T cells in vivo is difficult to detect due to the low precursor frequency of Ag-specific cells, we have used adoptive transfer of TCR Tg T cells into normal recipients to study T cell priming following vaccination with different OVA DNA constructs. We first examined CD8+ T cell responses which DNA vaccines have been reported to prime efficiently (8, 9, 10, 11, 12).
Enumeration of OVA-specific Tg
CD8+V2+ T cells in the
draining lymph nodes of DNA-immunized mice indicated no significant
clonal expansion over a 12-day period (Fig. 3A), whereas animals immunized
with OVA in CFA showed a significant response with a peak clonal
expansion at day 6 postimmunization. However, in most experiments,
individual lymph nodes showed both an increased proportion and
increased total number of Tg cells although group means were not
statistically significant when compared with pcDNA (vector-only)
controls. Although the proportion of Tg T cells was not always
increased, we demonstrated increased numbers of
CD8+V2+ T cell blasts in
the draining popliteal lymph nodes of mice 6 days after injection with
plasmids encoding cell-associated and secreted OVA forms (Fig. 3C). Blastogenesis was accompanied by the up-regulation of
the early T cell activation marker CD69 (Fig. 3B), thus
indicating that transferred cells were indeed primed following DNA
immunization. Therefore, although Ag-specific CD8 T cells are activated
following i.m. DNA injection, the proportion of Tg CD8 T cells in the
draining lymph nodes, relative to the total number of lymphocytes, does
not increase dramatically, as is seen following protein vaccination. As
was the case following CD8+ OT-I transfer, we
could not demonstrate significant clonal expansion following
CD4+KJ1-26+ TCR Tg T cell
transfer, although individual animals showed both an increased
proportion of Tg cells and blastogenesis within the
CD4+KJ1-26+ population in
draining lymph nodes (data not shown). Thus, the peak in clonal
expansion seen for CD4 and CD8 T cells following immunization with
protein in adjuvant is not observed following i.m. DNA
vaccination.
|
As we saw little evidence of CD4+ or
CD8+ T cell clonal expansion, we wanted to
determine whether plasmid-expressed OVA was being processed to produce
the appropriate peptides for DO11.10 activation. IFN--activated
macrophages transfected with pTfROVA (Fig. 4B), pSecOVA (Fig. 4C), and all other constructs tested (data not shown)
presented OVA323–339 peptide to DO11-GFP
hybridoma cells. In these experiments, 29% of DO11-GFP cells showed
increased GFP fluorescence intensity when pSecOVA-transfected
macrophages were used as stimulator cells. All other plasmids including
pTfROVA (Fig. 4B, 15%) showed less activation (data not
shown). Therefore, we were confident that OVA expressed from all
constructs could be processed appropriately to activate DO11.10 T
cells.
|
Whereas there was no obvious increase in the total proportion of CD8+ or CD4+ Tg T cells, CFSE-labeling before adoptive transfer indicated significant T cell activation and division in vivo. CFSE segregates equally between daughter cells upon division resulting in sequential halving of fluorescence intensity with each generation (42). This technique has allowed us to follow the proliferative history of individual cells following vaccination with different DNA constructs.
CD8 T cells proceed through more divisions when OVA is cell-associated
Recipient B6 mice have an endogenous population (2.5–3%) of
CD8+V2+ lymphocytes that
have a CFSElow phenotype as shown in Fig. 5A. Labeled undivided cells
are characterized by CFSEhigh staining, as can be
seen following i.m. immunization with the empty vector pcDNA where the
majority of labeled cells remain undivided. In contrast, vaccination
with OVA-expressing plasmids, including pTfROVA that encodes a
cell-associated form of OVA, results in substantial cell division in
the draining inguinal lymph node. Each of the peaks between the
undivided and endogenous populations represents one division of cell
cycle-synchronized cells. The division numbers were assigned using the
clearly defined peaks for the OVA/CFA s.c. control (inguinal
nodes) and the number of cells falling within each of these division
numbers was calculated for individual mice in each group. After four
divisions, CFSElow Tg cells are obscured within
the endogenous population; thus, results may actually be an
underestimate of the number of primed/dividing cells. There was also
evidence of cell division as early as 4 days postvaccination (data not
shown).
|
CD4+ T cell division in vivo is influenced by location of DNA-expressed Ag
As was the case following transfer of CD8+
OT-I lymphocytes, there was no significant clonal expansion following
adoptive transfer of CD4+ DO11.10 T cells and
subsequent i.m. DNA immunization (data not shown). However, CFSE
staining revealed that both pTfROVA (Fig. 6) and pSecOVA (Fig. 6B)
vaccination induced the division of transferred DO11.10 cells in the
draining popliteal lymph nodes. Eight days after pTfROVA vaccination,
there was an increase in the number of CFSElow
cells in comparison to the number following pcDNA vaccination (Fig. 6A). We also observed increased forward scatter in
CD4+DO11.10 T cells from pTfROVA-vaccinated mice
(Fig. 6A), indicating T cell blastogenesis in these lymph
nodes. In comparison to the OT-I studies, fewer DO11.10 cells divided
overall; although those that did divide proceeded through at least five
divisions by 8 days postvaccination. Approximately 19.3 ± 11.9%
(group mean ± SEM; n = 3) and 8.2 ± 3.9%
of DO11.10 T cells in the popliteal nodes had divided five or
more times following pSecOVA and pTfROVA injection, respectively. In
comparison, <3% of Tg cells had divided to the same extent following
vaccination with plasmid encoding other OVA forms. Cell division was
usually limited to the popliteal lymph nodes, although later time
points revealed some dissemination of the proliferative response to the
more distal inguinal lymph nodes (data not shown).
|
The inclusion of introns in mammalian expression vectors is known to increase gene expression, and hence, protein production both in vitro and in vivo (43). Because Ag load influences the magnitude of the initial T cell clonal burst, and hence, the size of the effector and memory cell populations, we wanted to know if plasmid manipulations that increase the amount of Ag, such as including an intron, also influence T cell priming following DNA vaccination. Hence, we constructed OVA-expressing constructs with and without a chimeric intron.
Lymphoid cells from conventional B6 mice immunized (x2) with the
intron-containing construct pCI-NatOVA produced 2-fold more IFN-
than did the intronless pNatOVA, following in vitro restimulation with
the H-2Kb-associated class I-restricted
OVA257–264 peptide (Fig. 7A). This assay measures the
ability of peptide-specific memory CD8+ T cells
to respond to Ag challenge. The higher IFN- levels from the
intron-containing group suggest a quantitative difference in the size
of the day 50 memory pool between the groups with or without an
intron.
|
These differences in eventual outcome (i.e., differential IFN-
production upon in vitro restimulation at day 50 postpriming)
may be explained by variability in the efficiency of CD8 T cell priming
by the two constructs. To investigate this further, we studied T cell
priming using vaccinated mice previously transferred with CFSE-labeled
OT-I T cells. CD8+ T cells divided significantly
more times following vaccination with the intron-containing plasmid
pCI-NatOVA than after pcDNA (empty vector) injection (Fig. 7, B and C). Furthermore, the addition of an intron
at the 5'-end of the OVA cDNA significantly enhanced the proliferative
response following pNatOVA injection (p <
0.05). Eight days after pCI-NatOVA vaccination, 14.9 ± 0.6%
(group mean ± SEM; n = 3) of OT-I T cells in the
popliteal lymph nodes had divided four or more times in contrast to
7.0 ± 1.7% of cells from pNatOVA-vaccinated mice.
Intron influences IFN- production by CD8 T cells early after
priming
Seven days after OT-I cell transfer and vaccination with
pCI-NatOVA and pNatOVA, the number of IFN- producing
CD8+ Tg T cells in draining lymph nodes following
brief peptide restimulation in vitro was higher for the
intron-containing group (Fig. 7D). This assay measures
primary and rapid recall responses, and the data suggest that even at
this early time point, there are measurable differences between groups,
not only in cell division but also in terms of effector molecules,
i.e., cytokines. Based on these results, we can conclude that the
efficiency of T cell priming is enhanced by the inclusion of a
mammalian intron in the construct design.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
i.m. injection of constructs encoding cell-associated forms of OVA
(pTfROVA and pCytOVA) induced greater and more rapid
CD8+ T cell division in the draining lymph nodes
than when secreted OVA constructs were injected. Cell-associated forms
of OVA induced CD8+ T cell division within 4 days
and many cells (e.g., 40% for pTfROVA; Fig. 5B) had
divided by 8 days postvaccination. Secreted Ag including OVA targeted
to either the classical secretory pathway (pSecOVA) or secreted via a
nonclassical mechanism (pNatOVA) was also capable of priming
CD8+ T cells. However, in both of these cases,
<20% of cells had undergone one or more divisions by 8 days after
vaccination (Fig. 5B). Furthermore, following vaccination
with pSecOVA, the initiation of the response was somewhat delayed in
comparison to that for cell-associated forms. We were also able to
detect cell division following vaccination with cell-associated forms
at sites more distal to the site of injection (i.e., the inguinal lymph
nodes) when the response to pSecOVA was restricted to the local
popliteal lymph nodes. These results suggest that cell-associated OVA
induces a more rapid and disseminated response and indicated that
although CD8+ T cells can divide in response to
DNA vaccination with constructs encoding soluble Ag, cell-associated
forms greatly enhance the speed and magnitude of this type of
response.
Any discussion of mechanisms by which these different OVA forms are processed for presentation to MHC class I-restricted CD8+ T cells is complicated by the fact that both nonpresenting muscle cells and APCs may be transfected by injected DNA and both may express Ag. If APCs take up the plasmid directly and express and process Ag, then cytosolic Ag can readily access the MHC class I presentation pathway, whereas Ag targeted to the classical secretory pathway is perhaps less efficient at accessing this pathway. However, there is increasing evidence that direct transfection of bone marrow-derived APCs (such as dendritic cells (DCs)) may be less important, but rather the major role of the APC is in presenting Ag (or peptide) produced by transfected myocytes to naive T cells (cross-priming). Recent studies have shown that cross-priming is also a major mechanism for priming CD8+ T cells following transdermal gene gun immunization (7). Although our study was not designed to follow the fate of Ag targeted to different subcellular locations, it would be interesting to identify where precisely immunologically relevant peptide was being generated in vivo for each of these constructs. MHC class I-associated peptide on the surface of muscle cells may be directly transferred to DCs, or chaperone:peptide complexes may be taken up by a receptor-mediated mechanism (6). APCs may acquire Ag (or peptide) by phagocytosis of apoptotic or necrotic muscle cells, and secreted protein may enter APCs by micropinocytosis, be processed, and presented. Studies of CD8 T cell cross-priming using OT-I T cells have shown that cell-associated OVA (OVA-coated splenocytes) is presented much more efficiently than soluble OVA, with 50,000-fold more soluble Ag required to stimulate equivalent cell proliferation (25). Our data suggest that although soluble exogenous Ag (secreted OVA) can access the MHC class I-processing pathway, this is less efficiently presented (cross-presented) to CD8 T cells than cell-associated forms.
We next examined the ability of OVA constructs to induce CD4+ T cell responses. All forms of OVA could be processed by macrophages and presented to DO11.10-GFP T cell hybridomas in in vitro presentation assays. In these assays, macrophages transfected with constructs designed to secrete OVA were best for activating the DO11.10-GFP hybridoma, thus exogenous Ag may enter the MHC class II-processing pathway more efficiently than other Ag forms. Studies by others have also demonstrated the differing abilities of OVA targeted to different cellular locations of transfected DCs to stimulate proliferation of CD4+ DO11.10 T cells (40). In these experiments, OVA targeted to the cell surface using the same TfR fusion construct used in our experiments, or as a fusion with murine invariant chain stimulated proliferation of T cells, whereas cytosolic Ag expression was ineffectual. Based on our in vitro results, we were confident that plasmid-encoded OVA could be processed and presented to CD4+ DO11.10 T cells, at least in vitro.
In our in vivo studies, secreted and surface/membrane-bound forms of
OVA were efficient at stimulating the proliferation of adoptively
transferred DO11.10 T cells, suggesting that the optimal way to prime
CD4+ T cells is to target Ag for secretion or
membrane association. In some mice, 56% of DO11.10 T cells had
divided, although the group mean was significantly lower (Fig. 6B). Exogenous protein is known to efficiently enter the MHC
class II-processing pathway after being taken up by APCs, processed and
Ag-derived peptide presented with MHC class II molecules. The fact that
pTfROVA was efficient for priming DO11.10 T cells is somewhat
surprising, particularly if it is assumed that myocytes are the
predominant cell type transfected and expressing the Ag. If
professional APCs such as DCs are directly transfected, the TfR
component should direct the TfROVA fusion proteins to the cell surface
and recycling vesicles, which may then intersect with
endosomes/lysosomes where intact fusion protein may be
proteolytically degraded for loading on MHC class II molecules. In
contrast, cytosolic Ag expression was inefficient at inducing division
of DO11.10 T cells, at least at the time points examined in this study.
Although endogenous Ags can be presented with MHC class II, the
mechanism by which cytosolic Ag in myocytes is acquired by APCs and
presented with MHC class II is unclear but may involve phagocytosis of
cellular material, which then enters the MHC class II-processing
pathway. Although we have not shown in this study that vaccination
using secretion and surface association constructs is better at
providing help to B cells for Ab synthesis, we are interested to
extrapolate the finding that these constructs are better for initial T
cell proliferation to whether these also are superior for enhancing T
cell differentiation and cytokine production and for helping B cells
(i.e., effector function). If this is the case, then an explanation for
secreted Ag vectors inducing higher Ab titers (13, 14, 15) may
not only be by increasing the availability of Ag in the lymph node for
priming B cells, but also by increasing the activation, proliferation,
and differentiation of Th cells.
Others have shown significant clonal expansion of DO11.10 T cells following gene gun immunization of OVA-expressing plasmids (33). In contrast, we were unable to demonstrate significant and reproducible increases in the total percentage of DO11.10 cells in the draining lymph nodes, although we could identify individual mice that had elevated numbers of Tg T cells. i.m. plasmid injection results in the production of small amounts of Ag (typically picogram-low nanogram/muscle; Ref. 44 and data not shown), and this may explain the absence of a significant rapid increase in the number of CD4+ (or CD8+) T cells. Immunogenic and tolerogenic forms of protein Ag both induce significant DO11.10 clonal expansion within 5 days as measured by the percentage of Tg cells (30, 32). Decreasing the amount of Ag has been shown to decrease clonal burst size, although the use of more sensitive techniques such as CFSE labeling has demonstrated that even with a very low amount of Ag, cells still divide and differentiate to effector and memory phenotypes (26). Following protein immunization there is an initial clonal burst, influenced by cell recruitment and cell division, after which many cells die while others progress through the cell cycle and differentiate into effectors or memory cells. Thus, the resulting pool of memory and effector cells after immunization is directly related to the initial clonal burst size. i.m. DNA vaccination, due to the small amounts of Ag and the continued Ag expression, does not appear to result in a single rapid clonal burst, but rather delayed and sustained T cell division.
Recent data indicate that CD8+ and CD4+ T cells are fundamentally different in their requirements for activation and clonal expansion (45). Recent studies have suggested that CD8+ T cell proliferation requires less Ag for activation and is not influenced by the duration of Ag presentation. Upon activation, these cells enter a developmental program that instructs them to continue division and differentiation into effectors and memory cells in the absence of further Ag stimulation (26, 27, 28, 45). It is hypothesized that this is an adaptation for situations, for example, at initial stages of virus infection when very low amounts of Ag are present. This may partially explain why DNA vaccines are particularly good at inducing CTLs. In contrast, CD4+ Th1 cells appear to require repeated Ag exposure and increased amounts of Ag for the survival of proliferating cells and for differentiation into cytokine-producing effector cells, although not for initial proliferation (28). In our study, CD4+ DO11.10 T cells only divided following injection of constructs expressing soluble and membrane-associated Ag, and in contrast to the OT-I studies, relatively fewer cells divided. This may be related to insufficient quantities of Ag to stimulate extensive CD4+ T cell division, at least at the time points examined, although different sensitivities of the different T cell clones may also explain this observation. Therefore, theoretically, increasing Ag amount should determine whether or not CD8+ T cells differentiate into CTLs/memory cells, but may increase the absolute number of dividing cells and subsequently effectors and memory cells. To evaluate the influence of Ag amount on the efficiency of T cell priming following DNA immunization, we constructed vectors that incorporated elements for altering the level of Ag expression.
The inclusion of an intron in the expression constructs (which is known
to increase gene expression and thus the total amount of Ag; Ref.
43) indicated that higher Ag doses increased the number of
OT-I cells undergoing division. Following pNatOVA injection, 18% of
OT-I T cells in the popliteal lymph nodes had divided one or more times
(i.e., 82 ± 2, 2% remained undivided; Fig. 7C),
whereas the inclusion of a chimeric intron preceding the NatOVA cDNA
(pCI-NatOVA) increased the number of cells that had divided to 35%
(i.e., 65.1 ± 2.1% undivided). This result demonstrated that
increasing the amount of a poorly immunogenic Ag increases
CD8+ T cell division, which may compensate for
the location of Ag expression. In addition to its effect on
CD8+ T cell division, we were also able to
demonstrate increased numbers of Ag-specific
IFN-+ CD8 T cells in the first days after
vaccination. Additional studies are needed to further characterize the
influence of introns (and hence Ag load) on the resulting memory and
effector populations generated by DNA immunization.
We have demonstrated that the location and level of plasmid-encoded Ag expression influence the efficiency of both CD4+ and CD8+ T cell priming, and that simple construct manipulations influence the initiation of the immune response. Using sensitive techniques such as adoptive transfer of Tg lymphocytes, we have identified some key factors that influence the developing immune response following DNA injection. The exploitation of these new technologies may allow us to understand many of the key inductive mechanisms related to lymphocyte priming, cell interactions, cell trafficking, and the longevity of the immune response following DNA vaccination at the level of the individual cell. This may ultimately allow us to produce "designer" DNA vaccines that have been engineered to produce the most relevant type and magnitude of response at the most appropriate location for a particular pathogen.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Catherine Rush, Division of Infection and Immunity, Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow, U.K. G12 8QQ. E-mail address: C.Rush{at}bio.gla.ac.uk
3 Abbreviations used in this paper: Tg, transgenic; cOVA, chicken OVA; DC, dendritic cell; GFP, green fluorescent protein; SA-FITC, FITC-conjugated streptavidin; TfR, transferrin receptor; FL1, FITC channel; pCytOVA, plasmid expressing partial cytosolic OVA; pNatOVA, plasmid expressing native OVA; pSecOVA, plasmid expressing secreted OVA; pTfrOVA, plasmid expressing transferrin-OVA fusion protein; SA-PE, PE-conjugated streptavidin.
Received for publication March 11, 2002. Accepted for publication August 27, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-helical domain of PspA elicits protective immunity against Streptococcus pneumoniae. Infect. Immun. 69:5456. hybrid construct. Immunology 94:135.[Medline]
1,4-galactosyltransferase is in the transmembrane domain. J. Biol. Chem 267:4084.This article has been cited by other articles:
|
|
 |
|
  J. Gao, J. M. Coulson, A. Whitehouse, and N. Blake Reduction in RNA Levels Rather than Retardation of Translation Is Responsible for the Inhibition of Major Histocompatibility Complex Class I Antigen Presentation by the Glutamic Acid-Rich Repeat of Herpesvirus Saimiri Open Reading Frame 73 J. Virol., January 1, 2009; 83(1): 273 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Solvason, Y.-P. Lou, W. Peters, E. Evans, J. Martinez, U. Ramirez, A. Ocampo, R. Yun, S. Ahmad, E. Liu, et al. Improved Efficacy of a Tolerizing DNA Vaccine for Reversal of Hyperglycemia through Enhancement of Gene Expression and Localization to Intracellular Sites J. Immunol., December 15, 2008; 181(12): 8298 - 8307. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Kadaoui and B. Corthesy Secretory IgA Mediates Bacterial Translocation to Dendritic Cells in Mouse Peyer's Patches with Restriction to Mucosal Compartment J. Immunol., December 1, 2007; 179(11): 7751 - 7757. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Werner-Klein, C. Dresch, P. Marconi, and T. Brocker Transcriptional Targeting of B Cells for Induction of Peripheral CD8 T Cell Tolerance J. Immunol., June 15, 2007; 178(12): 7738 - 7746. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. I. M. Loeffler, C. U. Schoen, W. Goebel, and S. Pilgrim Comparison of Different Live Vaccine Strategies In Vivo for Delivery of Protein Antigen or Antigen-Encoding DNA and mRNA by Virulence-Attenuated Listeria monocytogenes Infect. Immun., July 1, 2006; 74(7): 3946 - 3957. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Lauterbach, A. Gruber, C. Ried, C. Cheminay, and T. Brocker Insufficient APC Capacities of Dendritic Cells in Gene Gun-Mediated DNA Vaccination. J. Immunol., April 15, 2006; 176(8): 4600 - 4607. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. W. Stone, S. Barzee, V. Snarsky, K. Kee, C. A. Spina, X.-F. Yu, and R. S. Kornbluth Multimeric Soluble CD40 Ligand and GITR Ligand as Adjuvants for Human Immunodeficiency Virus DNA Vaccines J. Virol., February 15, 2006; 80(4): 1762 - 1772. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Cassataro, S. M. Estein, K. A. Pasquevich, C. A. Velikovsky, S. de la Barrera, R. Bowden, C. A. Fossati, and G. H. Giambartolomei Vaccination with the Recombinant Brucella Outer Membrane Protein 31 or a Derived 27-Amino-Acid Synthetic Peptide Elicits a CD4+ T Helper 1 Response That Protects against Brucella melitensis Infection Infect. Immun., December 1, 2005; 73(12): 8079 - 8088. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-J. Ko, S.-Y. Ko, Y.-J. Kim, E.-G. Lee, S.-N. Cho, and C.-Y. Kang Optimization of Codon Usage Enhances the Immunogenicity of a DNA Vaccine Encoding Mycobacterial Antigen Ag85B Infect. Immun., September 1, 2005; 73(9): 5666 - 5674. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
