| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
*
Laboratoire de Vectorologie et Transfert de Gènes, Centre National de la Recherche Scientifique-Unité Mixte de Recherche 1582, Institut Gustave Roussy, Villejuif, France;
Unité de Biologie Moléculaire du Gène, Institut National de la Santé et de la Recherche Médicale Unité 277, Institut Pasteur, Paris, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
gene
transfer into established P815 murine tumors. While IFN- was no more
potent than the vector alone, IL-12 gene transfer promoted tumor
eradication. Despite this antitumor effect, no significant cytolytic
activity was detectable using classical cytotoxicity assays from in
vitro restimulated splenocytes. Since intratumor gene delivery may
induce a localized expansion of CTL, the presence of P815-specific
CD8+ T cells in situ was assessed. Using the Immunoscope
approach, we found a dramatic increase in clonotypic T cells at the
tumor site following IL-12, but not IFN- gene delivery. Antitumor
CD8+ T cell frequencies were then re-evaluated using this
molecular detection technique, which revealed a comparable expansion of
specific T cells in the peripheral organs, most strikingly in the
blood. These data show that local IL-12 gene transfer, in contrast to
IFN-, mediates a potent antitumor effect that correlates to clonal
tumor-specific T cell expansions in situ and in the
periphery. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Preclinical studies in rodents have shown that IL-12, a key cytokine
for the induction of Th1-oriented immune responses, is one of the most
potent molecules in cancer therapy. In initial protocols,
administration of IL-12 resulted in severe toxicity in humans (3).
Among other means, this can be circumvented by in situ delivery. The
efficiency of local secretion of IL-12 at the tumor site using
genetically modified fibroblasts (4) or tumor cells (5, 6, 7, 8, 9) or
recombinant viral vectors such as retrovirus (10), vaccinia virus (11),
or adenovirus (12, 13) has been demonstrated in a variety of tumor
models. Analysis of the antitumor mechanism of IL-12 in murine tumor
models has shown that T cells are critical immune effectors; however,
the involvement of specific T cell subsets (CD8+,
CD4+, or NK T cells) varies with the tumor model studied
(14, 15). Other immune effectors, such as NK cells or macrophages, have
been implicated in the antitumor effects of IL-12 (5, 16). An
antiangiogenic effect has also been ascribed to this cytokine (17). The
induction of IFN- production is necessary but not sufficient to
explain the anti tumor efficiency of IL-12 (18, 19).
The involvement of tumor-specific CTL has been suggested in many
systems by in vivo depletion studies showing a requirement for
CD8+ cells in IL-12-mediated antitumor effects (18, 20).
Noguchi and colleagues have further demonstrated that a protocol
combining administration of tumor-derived peptide and low doses of
IL-12 allows for the induction of specific CTL in naive mice as well as
the eradication of established tumors (21). Moreover, the infiltration
of IL-12-treated tumors by high numbers of CD8+ cells has
been observed by immunohistochemistry (18). However, while the
induction of IFN- production has often been associated with a
successful IL-12 treatment, an increase in tumor-specific cytotoxic
activity in tumor-bearing mice has rarely been reported. To our
knowledge, such an effect has been reported by three groups: in one
case, after complete tumor eradication, but compared with mice bearing
huge tumors (22); in the second, a weak cytotoxic activity was
detectable in the draining lymph node (LN) (23); and recently, IL-12
was shown to enhance the significant cytolytic activity detectable
following the intratumoral inoculation of a replication-competent
recombinant herpes simplex virus (24). In contrast, in another study
the cytolytic activity was not correlated with the antitumor effects of
IL-12 used in combination with IL-2-transfected tumor cells (25).
We attempted to correlate IL-12-mediated antitumor effects with the presence of tumor-specific CTL in the periphery and in the tumor itself. To do so, we analyzed the effects of adenovirus-mediated IL-12 gene transfer in established P815 tumors. This immunogenic tumor model was chosen because specific CTL have been shown to be involved in its rejection in a number of situations, the clearest proof of this fact being the selection in vivo of tumor variants that have lost CTL-recognized Ags (26). Furthermore, the characterization of a public repertoire among P815-specific CTL (27) allows the detection and quantification of clonotypic tumor-specific CD8+ T cells directly ex vivo from any tissue, using the Immunoscope approach (28).
We treated mice bearing established P815 tumors by direct intratumoral
injection of recombinant adenoviral vectors encoding either IL-12 or
IFN-. Paracrine delivery of IL-12 was found to cure nearly all mice,
while IFN- was no more efficient than the vector alone. Despite this
therapeutic efficacy, we could not detect significant cytolytic
activity using classical cytotoxicity assays with in vitro restimulated
splenocytes. In contrast, the Immunoscope technique revealed a dramatic
increase in specific CD8+ T cells frequency at the tumor
site following IL-12 gene delivery, which was also observed in the
peripheral organs, most strikingly in the blood. No significant
increase was observed following control virus injection or IFN-
therapy.
Our data demonstrate that IL-12 gene delivery into established tumors
is able to induce a large tumor-specific CD8+ T cell
expansion that is associated with tumor eradication. Neither of these
effects could be attributed only to IFN- production.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
P815 is a mastocytoma of DBA/2 origin (29). The P815 subline used derived from P1.HTR (30) and was grown in RPMI 1640-GlutaMAX (Life Technologies, Cergy-Pontoise, France) supplemented with 10% FCS, 100 U/ml penicillin, and 100 µg/ml streptomycin. The tumor cell line was maintained in vitro for no longer than 1 mo before in vivo experiments. P815-B7.1 have been obtained by transfection of P1.HTR by the LL218 vector encoding murine B7.1 (a gift from Dr. L. L. Lanier, DNAX Institute, Palo Alto, CA). L1210 is a DBA/2 leukemia cell line that does not express P815 tumor Ags (31). Cell culture medium and reagents were obtained from Life Technologies.
Mice
Female DBA/2 mice were purchased from the Centre d’Élevage Janvier (Le Genest St. Isle, France) and maintained in the animal facilities of the Institut Gustave Roussy (Villejuif, France). Mice were 8–10 wk of age at the onset of each experiment. Mice were bled (200 µl) at the orbital sinus after anesthesia with ether.
Replication-defective recombinant adenoviral vectors
The murine IL-12 p35 and p40 cDNAs were obtained from the msp35
and msp40 plasmids, respectively (provided by U. Gubler, Hoffmann-La
Roche, Nutley, NJ). The p35 cDNA was first placed under the
transcriptional control of the human CMV immediate early gene promoter
by subcloning it into the pcDNA3 plasmid (Invitrogen, Leek, The
Netherlands). This expression cassette was released and ligated into
the pAd.RSVßgal expression plasmid, which contains the left end of
the adenovirus type 5 genome, resulting in pCMVp35IL-12. The p40 cDNA
was subcloned into the plasmid pAd.RSVßgal under the control of the
Rous sarcoma virus (RSV) long terminal repeat (LTR) promoter. The
RSVp40 expression cassette was inserted downstream of p35 cDNA into
pCMVp35IL-12, resulting in pCMVp35RSVp40IL-12. The recombinant
adenovirus deleted for the E1 and E3 regions was constructed by in vivo
homologous recombination in 293 cells cotransfected with the genome of
Ad.RSVßgal (32), left end deleted, and the linearized plasmid
pRSV40pCMV35IL-12. The vector AdmIL-12 was rescued, plaque purified,
and amplified using 293 cell line. The viral preparations were purified
by two CsCl density centrifugations, and viral titer (plaque-forming
units (pfu) per milliliter) was determined by plaque assay in 293 cells
(33). The recombinant adenovirus encoding murine IFN- was
constructed following the same procedure as that used for AdmIL-12 and
is described elsewhere.5
Expression of IFN- cDNA is driven by the LTR RSV promoter. The
adenoviruses used as controls were AdCO1, an adenoviral vector
identical with AdmIL-12 lacking a transgene (34), and dl324, which
contains part of the E1b region of the adenoviral genome not present in
AdCO1 or AdmIL-12 (35). Both control vectors were found to behave
similarly and are henceforth referred to as empty virus.
In vitro IFN- release assays
Spleen and LN of two mice per group were collected 3 days
following the i.v. injection of 1 x 109 pfu of
adenoviral vector diluted in 0.3 ml of PBS. Cells were harvested after
mechanical disruption of tissue, depleted from erythrocytes by osmotic
lysis, counted, and resuspended in RPMI 1640-GlutaMAX medium containing
5% heat-inactivated FCS, 100 U/ml penicillin, and 100 µg/ml
streptomycin at a final concentration of 5 x 106
cell/ml. One million cells were plated in 96-well round-bottom plates
and incubated for 48 h at 37°C, 5% CO2.
Supernatants were collected, and the amount of IFN- was assessed by
ELISA (Genzyme, Cambridge, MA). The detection limit of this assay was 5
pg/ml.
In vivo treatment of tumor-bearing mice
Mice were inoculated intradermally into the right flank with the minimal tumorigenic dose of our P815 subline (5 x 105 cells) in a volume of 0.1 ml of PBS. Nine and thirteen days following tumor cell inoculation, 5 x 108 pfu of recombinant adenoviral vector diluted in 0.1 ml of PBS was injected at the tumor site. Mice were scored for tumor growth weekly. The animals were sacrificed when the tumor exceeded a 15 mm mean diameter.
Cytotoxicity assays
Before each experiment, positive control mice (vaccinated mice) were generated by intradermal injection of 106 P815-B7.1 cells 15 days before P815 tumor graft. Spleen cells (3 x 107) from mice injected 20 days previously with 5 x 105 P815 cells were stimulated with 106 irradiated (100 Gy) P815-B7.1 cells in 20 ml of RPMI 1640 medium containing 10% FCS, 5 x 10-5 M 2-ME, and antibiotics. Cultures were maintained at 37°C for 5 days in upright 25-cm2 culture flasks. Cells were then counted and mixed at various ratios with 5000 51Cr-labeled P815 or L1210 target cells. One lytic unit is defined as the number of spleen cells that lyse 50% of 104 target cells in 4 h (36), calculated using a Microsoft Excel program provided by Dr. B. Van den Eynde (Ludwig Institute, Brussels, Belgium). The number of L1210-specific lytic units was subtracted from the number of P815-specific lytic units to yield the final result.
RNA extraction and cDNA synthesis
Tumor specimens and LN were snap-frozen immediately after removal and were kept at -80°C for no more than 1 mo. Immediately after thawing, they were homogenized and lysed in 4 M guanidium thiocyanate. For spleens and blood samples, pelleted single cell suspensions were lysed in guanidium thiocyanate before being frozen and kept at -80°C. Total RNA was isolated according to the method described by Chirgwin et al. (37). cDNA was synthesized from 10 µg of RNA, using a (dT)17 primer, AMV reverse transcriptase (Boehringer Mannheim, Mannheim, Germany), and its provided buffer, with addition of RNasin (Promega, Madison, WI).
PCR conditions
All PCR and run-off reactions were performed in a Geneamp 9600 thermocycler (Perkin-Elmer, Foster City, CA). PCR was performed in 20- to 40-µl volumes; run-off reactions were performed in a 20-µl volume. The reaction mixture for PCR contained 0.2 mM of each dNTP, 0.5 µM of each primer, 2.5 mM MgCl2, variable amounts of cDNA, 25 U/ml of Goldstar Taq DNA polymerase (Eurogentec, Seraing, Belgium), and the provided buffer. The reaction mixture for run-off reactions was the same, except 1) the single primer was added at a 0.1-µM concentration; 2) Promega Taq polymerase (Madison, WI) was used; and 3) the template was 10 µl of a PCR product diluted five times with water. Reaction cycles consisted of 25 s at 94°C, 25 s at 60°C (except for run-off reactions with the anti-A clonotypic primer P1AIR, 65°C), and 30 s at 72°C. Cycles were preceded by 3 min of denaturation at 94°C followed by 5 min of elongation at 72°C. The number of PCR cycles depended on the specific application and primers used (see below). Three cycles were used for run-off reactions.
Oligonucleotides
Oligonucleotides were purchased from Eurogentec (Seraing,
Belgium). Fluorescent oligonucleotides were directly labeled with FAM
fluorophores (Applied Biosystems, Foster City, CA). For IL-12
amplification, the following primers were used. To detect transgenic
p40, the RSV-LTR-specific primer CATTCACCACATTGGTGTGC was used along
with the p40 antisense primer GTCTTCTCTACGAGGAACGC. To detect total
p40, the RSV-LTR primer was replaced by the p40 sense primer
GAAGCACCAAATTACTCCGG. To detect transgenic p35, the CMV
promoter-specific primer CACTGCTTACTGGCTTATCG was used with the p35
antisense primer CTAGAATGATCTGCTGATGG. Primer pairs for amplification
of endogenous IFN- and hypoxanthine phosphoribosyl-transferase
(HPRT) have been described previously (38) (the 5' and run-off primers
have been used). The IFN--specific primers cited above amplify only
the endogenous IFN- cDNA because they hybridize to the 3'
untranslated region, which is not included in the AdmIFN- vector. To
amplify all IFN- messengers (either endogenous or from AdmIFN-),
the primers CACACTGCATCTTGGCTTTG and
fam-CCACATCTATGCCACTTGAG, which hybridize to the coding region,
were used. An additional primer pair has also been used for HPRT
(TTGCTGACCTGCTGGATTAC and fam-GTTGAGAGATCATCTCCACC). Primers
for TCRBV1 amplification have been described previously (27).
Quantification of HPRT and IFN- transcripts
Relative levels of HPRT and IFN- transcripts were measured in
the different cDNAs by PCR stopped in the exponential phase (39, 40).
As one of the two primers used being labeled with a FAM fluorophore,
the amount of amplified product could be quantified after loading of 2
µl of PCR product diluted four times with formamide in a 373A
automated DNA sequencer (Applied Biosystems, Foster City, CA). The
products were loaded along size standards, and the data were analyzed
with the Immunoscope software developed in the laboratory (41). The
number of cycles that provided sufficient signal without leaving the
exponential phase was first assessed in a preliminary kinetic
experiment. Typically, 24 cycles were required for HPRT, and 27 were
required for IFN-. For each sample, four tubes were amplified in
each reaction, containing decreasing volumes of cDNA as initial
template. The fact that the final signal decreased proportionally to
the initial amount of template was used as a control that the
amplification was stopped in the exponential phase. The quantification
of HPRT has been double-checked systematically with the two primer
pairs. The two quantifications always provided consistent results, and
the average value was used for normalization.
Quantification of public P815-specific T cells
In the tumors, the repertoire of infiltrating T cells is
oligoclonal, and the public T cells represent the majority of
BV1-expressing T cells (27). To compare the levels of infiltration by
public P815-specific T cells in the different tumor samples, the levels
of transcripts encoding TCRBV1 chains with a 9-amino acid (aa)-long
CDR3 were measured by quantitative PCR in the exponential phase (as
described above for HPRT or IFN-) with the VßS01 and Cß5'-fam
primers (27 cycles were usually required). The ratio of the TCRBV1 (9
aa) signal to the HPRT signal was used to obtain a relative level of
infiltration. To obtain an estimate of the actual frequency of the
public TIL, tumor samples were compared by the same technique with
spleen cells from naive mice. The final number was calculated by
assuming that 20% of naive splenocytes were T cells, among which 2%
used the BV1 segment.
In the peripheral organs, we took advantage of the fact that the clonal expansion of specific public T cells is observed while the repertoire profile of T cells using BV1 with other CDR3 lengths is left undisturbed (27). This allows for a rapid quantification of the frequency of the public T cells that bear a 9-aa-long CDR3 sequence recognized by clonotypic primer by assuming that the frequency of cells using other CDR3 lengths is constant, as described in detail previously (42). The cDNA from the sample of interest was amplified to saturation (40 cycles) in two PCRs, one with the VßS01 and Jß1.2far primers and one with the VßS01 and Jß2.5far primers, to quantitate anti-A and anti-E public cells, respectively. Each of these PCR was subjected to two run-off reactions, one with the corresponding fluorescent BJ-specific primer and one with the appropriate clonotypic primer (P1IR and P2IR, respectively). Equal amounts of run-off products were then electrophoresed in an automated sequencer. The intensity of the unique peak yielded by the clonotypic primer was then compared with the sum of the intensities of the peaks corresponding to CDR3 lengths of 8, 10, and 11 aa (the three major peaks, apart from the variable 9-aa peak). The final frequencies of public anti-A and anti-E cells among T cells was obtained by taking into account the specific activities of the primers and calculating that the cells represented in 8-, 10-, and 11-aa peaks of the BV1J1S2 and BV1J2S5 subpopulations constitute 1/1700th and 1/600th of the total T cell repertoire, respectively. The frequencies of anti-A and anti-E public cells were then added to obtain the final value.
Statistical analyses of data
Fisher’s exact method was performed to interpret the significance of differences between experimental groups. To interpret the survival between different groups, the log-rank test was used. Significances are indicated in the figure legends.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We constructed a replication-defective adenoviral vector harboring the two chains (p35 and p40) of mouse IL-12. Since p40 homodimers are antagonists of heterodimeric IL-12 (43), we devised a system in which more p35 than p40 is expected to be produced. Thus, p35 cDNA was transcribed under control of the CMV promoter, whereas transcription of the p40 subunit was controlled by the LTR RSV promoter, which was shown to be less potent in vivo (44).
The production of biologically active IL-12 by virally infected cells was first assessed using a PHA-blast proliferation assay and ELISA in the supernatant of various cell lines infected in vitro. With a multiplicity of infection allowing the infection of 50–70% of cells, the amount of p70 detected ranged from 15–30 ng/ml (data not shown).
Next, we tested the efficiency of in vivo IL-12 gene transfer. After
i.v. injection of 109 pfu of viral particles, mouse IL-12
p70 was detectable from days 1–6, with a peak on day 2 or 3 (1–2
ng/ml) in the sera of AdmIL-12 injected mice (data not shown).
Furthermore, since IL-12 systemic administration induces IFN-
production (18, 45), we measured IFN- levels in the supernatants of
single cell suspensions obtained from LN and spleens collected 3 days
following AdmIL-12 administration. As shown in Fig. 1
A, in vivo AdmIL-12
inoculation induced a dramatic increase in IFN- secretion from both
splenocytes and LN cells. Thus, the recombinant vector AdmIL-12 allowed
the production of biologically active IL-12 in vivo.
|
B, transcription of both IL-12 subunits, driven by the
viral vector, was almost systematically observed in the
AdmIL-12-treated tumors. Since the last adenoviral injection was on day
13, expression lasts for at least 7 days. Although significantly
weaker, some message for the endogenous IL-12 p40 was detected in
control animals. Interestingly, for one AdmIL-12-treated sample (lane 7), no transgenic p40 mRNA could be detected, while some transgene-driven p35 transcription was detectable. Coincidentally, this was the only tumor tested for which no regression had been induced by the treatment.
Similarly, in vivo expression of transgenic mIFN- following the
injection of AdmIFN- into established P815 tumors was detected using
RT-PCR (data not shown).
Intratumor AdmIL-12 administration leads to day 9 established P815 tumor eradication
We next evaluated the effects on tumor growth of AdmIL-12 or
AdmIFN- intratumoral injection into established P815 tumors. As
shown in Fig. 2, whereas PBS has no
effect on the tumor progression, AdmIL-12 cured most mice. AdmIFN-
was not more efficient than the empty virus; both induced the
regression of about half the tumors or delayed tumor growth. All
tumor-free mice were immunized against a second lethal tumor challenge
(not shown).
|
We first investigated the systemic tumor-specific CTL activity in treated mice using a standard chromium release assay. Spleen cells of mice from various treatment groups were removed on day 20 and restimulated for 5 days. As positive controls, we included vaccinated mice rechallenged with tumor cells at the time of tumor establishment in the other mice.
Although almost all AdmIL-12-treated mice as well as half of the
AdmIFN-- or empty virus-treated mice were in the process of
rejecting their tumors, specific cytolytic activity against P815 in
spleen cells was rarely detectable (Fig. 3).
|
AdmIL-12 treatment induces a dramatic increase in IFN-
expression and in the frequency of public CD8+ T cells at
the tumor site
We next decided to measure parameters of immune activity in the
tumor, which is the most relevant location to study the effector cells.
Among surrogate markers of tumor regression is the expression level of
IFN-, since 1) it constitutes an indirect assessment of the activity
of infiltrating CTL, NK, and/or Th1-type CD4+ cells; 2) its
expression has been shown to be correlated with prognosis in human
cancer (46); and 3) its expression is induced by IL-12 (47).
Quantitative PCR analysis of tumor specimens removed on day 20 revealed
that regression of tumors upon AdmIL-12 treatment was associated with a
dramatic, nearly 50-fold increase in the levels of endogenous IFN-
mRNA (Fig. 4A). This group was
the only one found to exhibit a significant increase over both control
groups. A weak but nonsignificant increase was observed in the empty
virus-treated tumors. The AdmIFN--treated tumors displayed a wide
range of individual endogenous as well as total IFN- expression. The
total amount of IFN- expressed in AdmIFN--treated
tumors was comparable to that detected in IL-12-treated samples.
|
When this analysis was performed on AdmIL-12 treated tumors, a striking
increase (
30-fold compared with control samples) in the frequency of
tumor-specific public T cells was found (Fig. 4B). Although
empty virus or AdmIFN- induced regression or delay the growth of
most tumors, no significant increase in the frequency of the specific T
cells was observed. Thus, tumor regression induced by AdmIL-12
treatment is associated with a dramatic increase in the frequency of
P815-specific CD8+ T cells in the tumor, which cannot be
attributed to IFN- induction alone.
Finally, since NK T cells have been shown to play a major role in
IL-12-induced tumor regression in some models (15), we assessed their
presence in the P815 tumors. This could be performed with the
Immunoscope method, since these cells bear a characteristic TCRAV14AJ15
(the AJ15 segment is better known as J
281) TCR rearrangement (48).
We did detect NK T cells, but AdmIL-12 treatment only led to a twofold
increase in the frequency of these cells, which was not statistically
significant (data not shown).
A strong specific CD8+ T cell increase is also observed in the blood following AdmIL-12 therapy
Since we were able to detect a high level of specific T cells in the tumor following AdmIL-12 treatment, we decided to search for these cells in the periphery using the Immunoscope technique. Thus, we measured the frequency of public P815-specific T cells in the blood of individual mice treated with PBS, empty virus, or AdmIL-12 at various time points.
In preliminary experiments we established that during the course of
untreated tumor progression, public T cells are first detectable in the
blood around day 7, peak around days 10–12, then decrease in numbers
(J.-P. Levraud, unpublished observations). There are, however,
important interindividual variations. For these reasons, the
frequencies of P815-specific T cells had to be assessed in all groups
before therapy. As shown in Fig. 5A, the three groups did not
significantly differ on day 8.
|
B, the blood frequency of tumor-specific
CD8+ T cells was 30-fold higher in AdmIL-12-treated
compared with that in PBS-treated mice on day 20. Empty virus-treated
mice displayed a low, but not significant, increase in circulating
public T cell frequency. This frequency dropped significantly on day 40
(not shown). Thus, the increase in frequency of circulating
tumor-specific CD8+ T cells, as assessed by Immunoscope
technology, was found to reflect that observed in the tumor following
AdmIL-12 treatment. Comparative study of various sites for the frequency of public T cells
Depending on their activation status, T cells home differently to
secondary lymphoid organs. As significant amounts of tumor-specific
CD8+ T cells were found in the blood, we investigated
whether these cells could also accumulate in peripheral sites. We
analyzed by Immunoscope the frequency of public T cells in the spleen
and draining and contralateral inguinal LN, and compared it with their
frequency in the blood of the same mice on day 20 (Fig. 6). Unexpectedly, the levels estimated in
the secondary lymphoid organs were much lower than those achieved in
the blood. All mice displayed a roughly similar level of tumor-specific
CTL in the ipsilateral draining LN. On the contrary, differences were
observed between treatment groups in the spleen or contralateral LN,
where AdmIL-12-treated mice displayed higher levels. These observations
reveal that paracrine delivery of IL-12 into established tumors led to
a strong expansion of tumor-specific CD8+ T cells that can
be detected in the tumor, the blood, and the secondary lymphoid organs
using the Immunoscope technique.
|
If the population measured by Immunoscope participates in P815
rejection, one should expect to find higher levels of specific
CD8+ T cells in mice with regressing tumors than in mice
with growing tumors, within treatment groups. As can be observed in
Fig. 7, mice whose tumor eventually
regressed displayed, on the average, a higher level of circulating
specific CD8+ T cells on day 20 than mice from the same
group that had to be sacrificed because of a large tumor burden. This
suggests that clonotypic cells detected with the Immunoscope technique
mediate a therapeutic role in vivo.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) and to a very significant increase in the
frequency of tumor-specific CD8+ T cells bearing public T
cell rearrangements in tumors (Fig. 4B), blood (Fig. 5), and
secondary lymphoid organs (Fig. 6). Nevertheless, cytotoxic
assays with in vitro-restimulated splenocytes failed to reveal
the presence of anti tumor CTL during the course of regression (Fig. 3).
Following AdmIL-12 treatment, tumor regression is associated with a
considerable increase in the level of IFN- transcripts at the tumor
site (Fig. 4A). This is in agreement with results obtained
by others with different tumor models (12, 13, 49). Although very high
levels of IFN- expression were also reached using
AdmIFN- (Fig. 4A), no improvement over the
empty virus has been achieved with this vector (Fig. 2). This is in
accordance with results of Brunda and colleagues showing that IFN-
cannot substitute for IL-12 to induce tumor regression (19).
The very high frequency of tumor-specific CD8+ T cells at
the tumor site achieved with IL-12, but not with IFN-, may explain
the difference in therapeutic efficiency between both cytokines.
Indeed, our data suggest that the ability of IL-12 to potentiate the
proliferation of activated T cells, originally described in vitro (50),
holds also true in vivo in IL-12-treated tumors.
Although a role for the public tumor-specific CD8+ T cells
in the rejection of the tumors has not been formally established, it is
supported by the fact that within treatment groups, their frequency is
higher in mice that ultimately reject their tumor (Fig. 7).
Unfortunately, since no Vß1-specific mAb is currently available, this
question cannot be addressed directly by a depletion experiment in
vivo. The mechanism by which these cells may exert their antitumor
effects is a matter of conjecture. Although no cytolytic activity was
revealed with splenocytes, these cells may still be cytolytic at the
tumor site. Other potential mechanisms of action would be regulation of
effector cells through cytokine secretion. The lack of efficiency of
AdmIFN- over empty virus argues against a major role for IFN-.
Whether other cytokines (TNF, for instance) or chemokines are involved
remains speculative.
While specific CD8+ T cells, presumably cytolytic, may play a major role in the rejection of P815 tumors following AdmIL-12 treatment, it remains possible that other effector cells are involved. No systemic NK cytotoxicity was found in our system (data not shown). However, as a direct intratumoral injection of AdmIL-12 induces a highly localized expression of the transgene (12), this result does not rule out the possibility of an activation of NK cells in situ. It has also been shown recently that NK T cells play a major role in IL-12-induced tumor regression in some models (15). Although NK T cells were detected in tumors by Immunoscope, AdmIL-12 treatment only led to a minor increase in the frequency of these cells (data not shown).
Injection of a control, "empty" virus in the tumors, although not
as efficient as AdmIL-12, had a clear impact on tumor growth (Fig. 2).
Despite this fact, only modest or incidental, but nonsignificant,
increases in specific T cells or IFN- were observed with the empty
virus compared with the vehicle (Fig. 4). The mechanisms that lead to
tumor rejection or growth delay following injection of empty virus are
also of interest. E1-deleted adenoviral vectors, such as the one we
used, are known to induce an inflammatory response at the site of gene
transfer (51, 52). This intratumor inflammatory response may induce the
activation of non-MHC-restricted effector cells (macrophages, NK
cells), leading to the killing of tumor cells without the expansion of
specific CTL. Such a mechanism, as previously hypothesized for the
effect of IL-2 gene transfer (53), may explain the anti tumor effect of
empty virus.
The lack of specific cytolytic activity in the spleen (Fig. 3) despite
the high frequency of clonotypic CD8+ T cells is an
intriguing result. This phenomenon is particularly interesting, as it
may be relevant to the paradoxical observation of positive clinical
outcomes of melanoma peptide vaccination without detectable increase in
specific CTL precursor frequency (54).
We are currently investigating the mechanisms underlying this discrepancy. At any rate, this was not due to a technical failure; it was observed in three independent experiments, and the mice vaccinated before tumor challenge, included as positive controls, displayed strong cytotoxic activities. It seems unlikely that it is due to the low sensitivity of the cytolytic assay, as the frequencies of public P815-specific CD8+ cells measured using Immunoscope in the spleens of the positive control mice were lower than those found in the spleens of AdmIL-12-treated mice (data not shown).
It is noteworthy that we performed cytotoxic assays with spleen cells collected 20 days after tumor graft on tumor-bearing hosts, while positive control mice were tumor free at that time. Possibly, the presence of a tumor may influence the outcome of the assay. In accordance with this hypothesis, cytotoxic assays performed on day 40, after total tumor eradication, resulted in strong cytolytic activity in two of three AdmIL-12-treated mice, while the frequency of public CD8+ T cells assessed by Immunoscope dropped significantly between days 20 and 40 (data not shown). Similarly, Bramson et al. reported cytotoxic activity following AdmIL-12 treatment of polyoma middle T Ag-expressing tumors in completely cured mice 42 days after the initial tumor graft (22). In contrast, Toda et al. found strong cytolytic activities with spleen cells collected from mice bearing tumors (24); however, the administration of a replicative herpes simplex virus vector to induce tumor necrosis renders the comparison difficult.
Comparison of the numbers of specific T cells by direct physical techniques and functional limiting dilution assays has shown that the activation status of the T cells critically affects the issue of procedures requiring in vitro restimulation (55). The lack of cytolytic activity in regressing tumor-bearing hosts might be attributed to the fact that tumor-specific CD8+ T cells, having recently encountered their cognate Ag, are activated effector cells prone to apoptosis upon further stimulation. In contrast, in cured mice, tumor-specific cells would be memory cells ready to proliferate in vitro in the presence of their Ag.
The observation that P815-specific T cells are more abundant in the
blood than in the draining LN (Fig. 6) was unexpected. It should be
stressed that our comparative study concerned rather late time points,
i.e., 20 days after grafting of tumor cells. Initial expansion of
tumor-specific precursors is supposed to take place in the draining LN.
Activated T cells, once they loose expression of homing receptors such
as CD62L, might leave the node to seed the tumor and the periphery
(56). Since the frequency of specific CD8+ T cells in the
draining LN is not modulated by immunotherapy, in contrast to what is
observed in the tumor, it is likely that the major specific T cell
expansion occurring under AdmIL-12 treatment takes place at the tumor
site itself. The observed differences in public T cell frequencies in
the blood, contralateral LN, and spleen are consistent with the
re-entry of cells expanded at the tumor site into the bloodstream. More
detailed analysis of the kinetics of specific T cell populations in the
various organs should clarify this point.
The development of reliable methods to evaluate the biological effects of cancer immunotherapeutic approaches is required for their optimization. The questions of where and how to evaluate the antitumor immune response elicited by immunotherapy are still a matter of debate. The fact that among peripheral organs, the blood contains the highest levels of specific CD8+ T cells following AdmIL-12 treatment is an interesting observation, since PBL constitute the most readily accessible source of T cells in humans. In mice, by contrast, it is far easier to obtain large number of T cells from the spleen or LN. However, the higher frequency of precursors among PBL probably makes it worth testing this location more systematically. Furthermore, our data stress the necessity of direct methods to evaluate the frequency of tumor-specific CTL, rather than systems based on in vitro culture. Techniques must be sensitive and resolutive, and the Immunoscope is clearly well adapted to such applications, as shown in various studies dealing with human cancer (57, 58, 59, 60, 61). However, this approach has some intrinsic limitations due to the fact that the TCR rearrangements of the relevant T cells first need to be characterized. The usage of tetrameric MHC-peptide TCR ligands, shown to be able to bind specific T cells for FACS analysis (62), will probably alleviate this problem. Furthermore, it will allow functional analysis of tumor-specific cells. Flow cytometry is limited in sensitivity, but, ultimately, combination of MHC-peptide-tetramers with Immunoscope analysis (63) should provide the advantages of both techniques. It would now be of major interest to test with such technology whether local IL-12 delivery in human tumors is able to induce the expansion of tumor-specific CD8+ T cells among the PBL of patients.
To our knowledge, this is the first report of a direct observation of an increase in tumor-specific CD8+ T cells frequency following IL-12-mediated therapy of cancer. This study formally demonstrates that IL-12 is able to induce the expansion and possibly the reactivation of specific T cells in hosts bearing progressively growing tumors. Adenovirus-mediated intratumoral delivery of IL-12 is able to induce not only a potent local immune response, but also the recirculation of large numbers of tumor-specific CD8+ T cells. Thus, local delivery of IL-12 into tumors should be implemented to induce systemic immune responses and exert effects on distant metastases.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Both authors contributed equally to this work.
3 Address correspondence and reprint requests to Prof. Philippe Kourilsky, Unité de Biologie Moléculaire du Gène, Institut National de la Santé et de la Recherche Médicale Unité 277, Institut Pasteur, 25 rue du Dr. Roux, 75724 Paris Cedex 15, France. E-mail address:
4 Abbreviations used in this paper: TIL, tumor-infiltrating lymphocytes; LN, lymph node(s); RSV, Rous sarcoma virus; LTR, long terminal repeat; AdmIL-12, recombinant adenovirus vector encoding mouse IL-12; pfu, plaque-forming units; AdmIFN-, recombinant adenovirus vector encoding mouse IFN-; HPRT, hypoxanthine phosphoribosyltransferase.
5 L. Cordier, N. C. Fernandez, J.-C. Sabourin, P. Opolon, N. Di Falco, M. Perricaudet, J. Bignon, A. Kane, and H. Haddada. Adenoviral mediated cytokine or costimulatory gene transfer into sarcomatous malignant mesothelioma: tumor regression following intratumoral delivery of IFN- gene. Submitted for publication.
Received for publication June 25, 1998. Accepted for publication September 22, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
production. Blood 90:2541.14 NKT cells in IL-12-mediated rejection of tumors. Science 278:1623. production. J. Immunol. 153:1697.[Abstract]
in mediating the antitumor efficacy of interleukin-12. J. Immunother. Immunol. 17:71.
thymocytes secreting a distinct pattern of cytokines and expressing a very restricted T cell receptor repertoire. Eur. J. Immunol. 27:544.[Medline]
production by natural killer cell stimulatory factor: characterization of the responder cells and synergy with others inducers. J. Exp. Med. 173:869.This article has been cited by other articles:
|
|
 |
|
  F. Siddiqui, C.-Y. Li, S. M. LaRue, J. M. Poulson, P. R. Avery, A. F. Pruitt, X. Zhang, R. L. Ullrich, D. E. Thrall, M. W. Dewhirst, et al. A phase I trial of hyperthermia-induced interleukin-12 gene therapy in spontaneously arising feline soft tissue sarcomas Mol. Cancer Ther., January 1, 2007; 6(1): 380 - 389. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X.-c. Chen, R. Wang, X. Zhao, Y.-q. Wei, M. Hu, Y.-s. Wang, X.-w. Zhang, R. Zhang, L. Zhang, B. Yao, et al. Prophylaxis against carcinogenesis in three kinds of unestablished tumor models via IL12-gene-engineered MSCs Carcinogenesis, December 1, 2006; 27(12): 2434 - 2441. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Harada, M. Shimada, S. Okano, T. Suehiro, Y. Soejima, Y. Tomita, and Y. Maehara IL-12 Gene Therapy Is an Effective Therapeutic Strategy for Hepatocellular Carcinoma in Immunosuppressed Mice J. Immunol., December 1, 2004; 173(11): 6635 - 6644. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Shi, A. L. Rakhmilevich, C. P. Heise, K. Oshikawa, P. M. Sondel, N.-S. Yang, and D. M. Mahvi Intratumoral Injection of Interleukin-12 Plasmid DNA, Either Naked or in Complex with Cationic Lipid, Results in Similar Tumor Regression in a Murine Model Mol. Cancer Ther., September 1, 2002; 1(11): 949 - 957. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Wan, L. Lu, J. L. Bramson, S. Baral, Q. Zhu, A. Pilon, and K. Dayball Dendritic Cell-Derived IL-12 Is Not Required for the Generation of Cytotoxic, IFN-{gamma}-Secreting, CD8+ CTL In Vivo J. Immunol., November 1, 2001; 167(9): 5027 - 5033. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Peng, J. C. Krauss, G. E. Plautz, S. Mukai, S. Shu, and P. A. Cohen T Cell-Mediated Tumor Rejection Displays Diverse Dependence Upon Perforin and IFN-{gamma} Mechanisms That Cannot Be Predicted From In Vitro T Cell Characteristics J. Immunol., December 15, 2000; 165(12): 7116 - 7124. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Bousso, F. Lemaitre, J. Bilsborough, and P. Kourilsky Facing Two T Cell Epitopes: A Degree of Randomness in the Primary Response Is Lost Upon Secondary Immunization J. Immunol., July 15, 2000; 165(2): 760 - 767. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Mortarini, A. Borri, G. Tragni, I. Bersani, C. Vegetti, E. Bajetta, S. Pilotti, V. Cerundolo, and A. Anichini Peripheral Burst of Tumor-specific Cytotoxic T Lymphocytes and Infiltration of Metastatic Lesions by Memory CD8+ T Cells in Melanoma Patients Receiving Interleukin 12 Cancer Res., July 1, 2000; 60(13): 3559 - 3568. [Abstract] [Full Text]
|
|
|
|
|
|
C. S. Schmidt and M. F. Mescher Adjuvant Effect of IL-12: Conversion of Peptide Antigen Administration from Tolerizing to Immunizing for CD8+ T Cells In Vivo J. Immunol., September 1, 1999; 163(5): 2561 - 2567. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
0.01) from the PBS
group. Only the AdmIL-12 group exhibits a significant
(p < 0.05) increase in survival compared with the
empty virus group.
