Therapeutic antitumor immunity depends on a highly migratory CTL population capable of activation and trafficking between lymphoid and tumor-bearing microanatomic sites. We recently adapted positron-emission tomography gene expression imaging for noninvasive, longitudinal localization and quantitation of antitumor T lymphocyte migration in vivo. In this study, we apply this system to enumerate the temporal accumulation of naive vs memory T cells. Naive or memory OT-1 CD8+ T cells, retrovirally marked with the sr39TK gene, were adoptively transferred into RAG1−/− animals bearing EL-4 or EG.7 (an OVA-expressing subline), and repetitively imaged by microPET over several weeks. Memory cells demonstrated early accumulation and apparent proliferation, with large T cell numbers at the Ag-positive tumor as early as day 1 after T cell transfer. Naive T cells did not accumulate in the E.G7 tumor until day 8, and reached only 25% of the peak levels achieved by memory T cells. Both naive and memory cells eradicated the Ag-expressing tumor at a comparable density of intratumoral T cells (2–4 × 106/g). However, due to the slower rate of T cell expansion and continued tumor growth, naive cells required ∼10-fold higher Ag-specific precursor frequency to reach a tumoricidal cell density. As recently reported, memory but not naive T cells accumulated in local lymph nodes and lungs, where they persisted as a resident population after tumor eradication. Positron-emission tomography-based immunologic imaging is a noninvasive modality providing unique and meaningful information on the dynamics of the antitumor CTL response.

Advances in understanding the nature of tumor-specific immune responses have prompted a broad-based effort to create and refine immune-mediated cancer therapies. Several modalities are under development, including molecular or dendritic cell-based vaccines, adoptive cellular immunotherapy, and immunomodulatory agents to improve the effectiveness and safety of antitumor immunity (1, 2, 3). Rational and systematic refinement of these modalities requires powerful immune monitoring methods that can accurately detect, locate, and quantitate the cell-mediated immune responses for the therapeutic intervention. Currently, immune monitoring methods are limited to invasive techniques of explanation of cell populations for in vitro measurements, such as histology, flow cytometry, or the detection of cytokines secreted from the excised lymphocytes (3, 4). These assays provide only snapshots in time and space of the immune response, and invasive approaches are not generally useful for clinical assessment. Therefore, it is crucial that methodologies that are capable of generating quantitative, noninvasive, repetitive, and spatial in vivo information about the dynamic process of immune responses be developed.

In recent years, whole-body molecular imaging has emerged as a new strategy to assess the dynamics of immune cell populations bearing reporter genes in vivo. One approach is ex vivo labeling methodologies to quantitatively localize cell trafficking in vivo. CD8+ T cells have been successfully labeled with iron oxide nanoparticle for in vivo magnetic resonance imaging tracking of cells (5). Cells have also been labeled to study cell trafficking by positron-emission tomography (PET),3 using 18F-FDG (fluorodeoxyglucose), 11CH3I, and 64Cu-pyruvaldehyde-bis(N4-methylthiosemicarbazone) (6). Such ex vivo methods are limited by the brief retention of label and viability of cells, preventing their application to the study of population dynamics over time. Bioluminescent imaging has been used for longitudinal detection of gene-marked lymphocytes. CD4+ T cells transduced with firefly luciferase have been used to assess the localization, dynamics, and response to immunotherapy in models of experimental autoimmune encephalomyelitis (7, 8) and rheumatoid arthritis (9). However, bioluminescence imaging presently cannot yield tomographic information, and in human applications is generally unfeasible due to the limitations of light penetration and scattering from interior anatomic sites.

PET-based gene expression imaging has emerged as an appealing modality for potential clinical applications because it is tomographic, quantitative, and minimally impaired by tissue absorption and scattering (10). In the antitumor response, trafficking patterns of antitumor T lymphocytes have been detected in vivo (11, 12, 13). Immune T cells were retrovirally transduced with HSV-sr39tk, and imaged by parenteral administration of the thymidine kinase (TK) substrate [18F]FHBG ([fluoro-3-(hydroxymethyl) butyl]guanine). Using this approach, transduced antitumor T cells were successfully tracked at sequential times during an unfolding immune response (11, 12). PET reporter gene imaging has also permitted in vivo assessment of TCR-induced NFAT-dependent transcriptional activity, which not only allowed monitoring of T cell trafficking, but also the assessment T cell activation (14).

In this study, we used PET-based gene expression imaging to compare the kinetics and quantitation of progeny of naive and memory CD8+ T cells, which accumulate in tumor and other tissue sites in association with tumoricidal response. Our findings were qualitatively and quantitatively concordant with these features of naive and memory cells assessed by more conventional modalities. PET-based immunologic imaging thus provides a unique and powerful modality for noninvasive evaluation of this important parameter in antitumor immunotherapy in vivo.

Male and female C57BL/6 and RAG1−/− mice, 4–6 wk old, were purchased from The Jackson Laboratory. OT-1 mice, bearing transgenes for expression of an OVA-specific TCR, were carried on the RAG1−/− background to assure exclusive T cell expression of the TCR transgene (15, 16, 17, 18). Animal husbandry and experimental procedures were approved by the University of California Los Angeles Animal Research Committee. The EL4 lymphoma cell line and E.G7, a subline transfected with an OVA expression vector (18), were obtained from the American Type Culture Collection. They were cultured in complete RPMI 1640 supplemented with 10% FBS, 100 μg/ml penicillin, and 292 μg/ml streptomycin. G418 (400 μg/ml) (Sigma-Aldrich) was also added to the media for E.G7 for transgene selection.

A total of 2 × 105 OT+CD8+ from OT-1 mice was i.p. transferred into C57BL/6 recipients. Two days after the transfer, the animals were then challenged i.p. with 2 × 107 irradiated E.G7 (5000 rads) twice at 1 wk apart. One month later, splenocytes were harvested from these immunized animals.

mAbs included FITC-conjugated 53–6.7 (anti-CD8), APC-conjugated IM7 (anti-CD44), and APC-conjugated MEL-14 (anti-CD62L), all obtained from BD Pharmingen. H-2Kb-OVA tetramer was PE-conjugated (Immunomics; Beckman Coulter). To isolate the tetramer-positive cells, we used anti-PE microbeads (Miltenyi Biotec). To delete CD8 T cells from splenocytes, we used mouse anti-CD8 (Lyt 2) Dynabeads (Dynal Biotech).

The protocol used for the retroviral infection was slightly modified from our recent studies (11, 12). The mouse stem cell virus-sr39TK-internal ribosome entry site-GFP retroviral construct was a gift from Dr. O. Witte (Howard Hughes Medical Institute and University of California Los Angeles, CA) and has been described in detail previously (11). This retrovirus permits coexpression of GFP and sr39TK, an engineered TK gene bearing the SR39 mutation, which optimizes detection due to increased avidity for our PET TK substrates, FHBG. The ecotropic retrovirus was generated by transient transfection of HEK-293T cells, and fresh viral supernatant was used for infection of lymphocytes. Mouse splenocytes were harvested and plated in 6-well plates (Costar). Cells were stimulated with one 133 μg/ml anti-CD3 and anti-CD28 (BD Pharmingen) at 107 cells per well in XVIVO15 medium (BioWhittaker). Twenty-four hours after stimulation, the cells were infected with the retrovirus supernatant at 1:1 with culture medium containing 1.6 μg/ml polybrene (Sigma-Aldrich). The infection was performed by first spin infecting the cells for 1.5 h at 2500 rpm, then overnight infection at 37°C.

Infection efficiency was analyzed 48 h after infection by GFP expression by flow cytometry; GFP+ cells were designated as TK+ cells. Infection efficiency in most cases was 30%; there was no difference in infection efficiency or GFP fluorescence intensity distinguishing naive and memory T cell target populations. Except where indicated, naive or memory T cells after retroviral labeling were diluted with unlabeled naive or memory C57BL/6 spleen cells to achieve a final concentration of 10% OT+ cells before transfer to recipient RAG mice.

Cell were incubated with the respective Abs described above for 30 min on ice in PBS + 1% BSA, washed, fixed in 2% paraformaldehyde (w/v) in PBS, and analyzed with a FACScan flow cytometry instrument and CellQuest software (BD Biosciences). Surface expression levels were calculated as mean fluorescence intensity. Unstained cells or cells stained with isotype-matched mAbs were used as negative controls. Both gave the same background fluorescence.

A total of 4 × 105 splenocytes was plated with 105 irradiated E.G7 (5000 rads) and 2.5 × 106 irradiated filler splenocytes (1000 rads) in 96-well round-bottom tissue culture plates (Costar) in 0.2 ml of XVIVO 15 media (BioWhittaker). The cells were cultured for 24 h, the supernatant were then harvested, and IL-2 and IFN-γ concentrations were analyzed by using the Quantikine murine IL-2 and IFN-γ kit (R&D Systems).

Splenocytes (3 × 107) and 1.5 × 106 irradiated E.G7 (5000 rads) were seeded in T25 tissue culture flasks (Corning) in 10 ml of XVIVO 15 media. The cells are cultured for 5 days, and then harvested for the 51Cr-release assay. EL4 and E.G7 target cells were incubated with 51Cr-sodium chromate (Amersham) for 1 h at 37°C, washed three times, and then incubated with naive or memory cells at various ratios in 96-well round-bottom tissue culture plates (Costar). After 4 h of incubation, 51Cr in supernatants was counted. Except for sextuplet wells to determine spontaneous and maximum 51Cr release, all samples were assayed in triplicate. Specific lysis was calculated as follows: [(experimental counts − spontaneous counts)/(maximum counts − spontaneous counts)] ×100.

PET studies were performed using the microPET Primate 4-ring system (P4) as described elsewhere (19, 20). In brief, the system operates in 3D and is composed of 168 lutetium oxyorthosilicate crystal detector modules, with a 7.8-cm axial and a 19-cm transaxial field of view. Animals received injections i.v. with ∼200 μCi of [18F]FHBG or [18F]FDG at the specific activity ∼1000 Ci/mmol. After 1-h uptake, the animals were anesthetized using 2.5% isoflurane, and data were acquired for 15 min in the microPET scanner. Images were reconstructed at 2.2-mm resolution using filtered back projection.

Concentrations of labeled cells at regions of interest (ROI) were determined using the correlation of signal intensity to cell number established in our recent study (12). Briefly, the signal for uptake of parenteral ([18F]FHBG) by cells at the ROI was first measured as raw %ID/g (the percentage of the total injected dose of [18F]FHBG detected per milliliter of tissue in the ROI). To determine the specific signal above background, signal from ROIs were quantified and represented as the ratio of the %ID/g of the ROI over the %ID/g of a background region of the animal. This ratio permitted calculation of the cell number, using the following equation: X = 4.8 * 105 * (Y-1.13)/IE, where X = cell numbers/g tissue, Y = ratio of % ID/g at ROI over background %ID/g; %ID/g is the percentage of the total injected dose of [18F]FHBG detected per milliliter of tissue; and IE is retrovirus infection efficiency. The limit of detection for quantitation of a ROI using TK+ lymphocytes expressing this vector is 105 cells/g (12).

Tumor size was measured at the indicated times after inoculation using imaging or by physical methods. For imaging, mice were given injections, sedated, injected with [18F]FDG, and microPET imaged as described in the preceding section. The [18F]FDG update in the tumor ROI was used to reconstruct the volume of the tumor region. For physical measurements, tumors were measured at the indicated time points in three dimensions by the use of calipers, and tumor volumes were calculated according to the following formula: volume = width × length × height × 0.52 (0.52 is the factor used to adjust for the volume of an ellipsoid).

Lungs were harvested and minced with scissors to a fine slurry in 15 ml of digestion buffer (RPMI 1640, 5% FBS, 1 mg/ml collagenase (Boehringer Mannheim), and 30 mg/ml DNase (Sigma-Aldrich)). Lung slurry was enzymatically digested for 45 min at 37°C and further dispersed mechanically. The total cell suspension was pelleted and resuspended in FACS analysis buffer. Cell counts and viability were determined using trypan blue exclusion on a hemocytometer.

In vivo immunization was performed to generate the memory CD8 T cells. Briefly, 2 × 105 CD8+ T cells from OT-1 mice were transferred i.p. into C57BL/6 recipients. The animals were then challenged i.p. with irradiated E.G7, an EL-4 thymoma transfected with OVA, the OT-1 cognate Ag. One month later, splenocytes were harvested and analyzed for CD4, CD8, and Ag-binding (H-2Kb/SIINFEKL tetramer) by flow cytometry. Naive OT+ cells were harvested directly from splenocytes of unmanipulated OT-1 mice. For simplicity, OT-1-derived Ag-positive (CD8+tetramer+) lymphocytes are termed OT+ cells.

The abundance of naive and memory OT+ cells obtained is shown in Fig. 1,A. As expected, naive T cells harvested from unmanipulated OT-1 mice on the RAG1−/− background consisted of >90% OT+ cells. For immune mice, 10–20% of the splenocytes were OT+. These OT+ cells were increased 50- to 100-fold in cell number compared with the initially transferred OT-1 cells, indicating prominent net clonal expansion in vivo. Both naive and memory OT+ T cells were negative for activation markers CD25 and CD69 (data not shown). However, the memory OT+ T cells had higher levels of CD44 and lower levels of CD62L compared with the naive OT+ T cells (Fig. 1 B). Therefore, the naive and memory T cells had the expected surface markers for their respective differentiative state.

FIGURE 1.

Phenotype of naive and memory OT+ T cells. Naive splenocytes were harvested from OT-1 RAG1−/− mice; memory cells were harvested after transfer to and immunization of C57BL/6 mice. A, H-2Kb (SIINFEKL) tetramer and CD8 expression in lymphocyte scatter gate. B, Tetramer+CD8+ cells were gated and analyzed for CD62L and CD44 expression. Memory cells are represented as the shaded area, and the naive cells are shown as the bold line. Isotype control staining coincided with the CD44 histogram of naive cells (data not shown). C–E, Naive and memory splenocytes were stimulated by culture in vitro with E.G7 at the same precursor frequencies (10%). C, ELISA was performed after 1 day to determine the concentration of IL-2 and IFN-γ secreted. The limit of detection for both analytes was 100 ng/ml (line). D and E, 51Cr-release assay. After 5-day stimulation, activated memory (D) or naive (E) T cells were cocultured with 51Cr-loaded target cells at the indicated ratios for 4 h, and 51Cr release was assayed for cytotoxicity. Data are representative of at least three independent experiments.

FIGURE 1.

Phenotype of naive and memory OT+ T cells. Naive splenocytes were harvested from OT-1 RAG1−/− mice; memory cells were harvested after transfer to and immunization of C57BL/6 mice. A, H-2Kb (SIINFEKL) tetramer and CD8 expression in lymphocyte scatter gate. B, Tetramer+CD8+ cells were gated and analyzed for CD62L and CD44 expression. Memory cells are represented as the shaded area, and the naive cells are shown as the bold line. Isotype control staining coincided with the CD44 histogram of naive cells (data not shown). C–E, Naive and memory splenocytes were stimulated by culture in vitro with E.G7 at the same precursor frequencies (10%). C, ELISA was performed after 1 day to determine the concentration of IL-2 and IFN-γ secreted. The limit of detection for both analytes was 100 ng/ml (line). D and E, 51Cr-release assay. After 5-day stimulation, activated memory (D) or naive (E) T cells were cocultured with 51Cr-loaded target cells at the indicated ratios for 4 h, and 51Cr release was assayed for cytotoxicity. Data are representative of at least three independent experiments.

Close modal

The functional features of the naive and memory cell populations were analyzed in Fig. 1, C–E. Without Ag stimulation, neither memory nor naive cells secreted IL-2 and IFN-γ (data not shown), but after 24-h stimulation with irradiated E.G7, memory cells secreted more cytokine than naive cells (3-fold and 7-fold for IL-2 and IFN-γ, respectively) (Fig. 1,C). This finding validated the memory state of the immune OT+ cells, because memory cells are reported to be more efficient for Ag-induced cytokine production (21, 22, 23). After 5-day TCR stimulation, immune and memory T cell populations had comparable killing activity after in vitro stimulation (Fig. 1, D and E), consistent with other comparisons of stimulated naive and memory T cells (23, 24, 25).

RAG1−/− mice were inoculated i.m. with 1.5 × 106 tumor cells (Ag-negative and -positive cells in the left and right shoulder, respectively). One day later, the animals received naive or memory cells (5 × 106) that were 10% OT+. Because CD4 T cells are required for efficient CD8 T cell function and tumor (26, 27), all animals also received 5 × 106 CD8-depleted unmarked splenocytes as a source of help. Preliminary results demonstrated that without the adoptive transfer of T cells in these RAG1−/− hosts. Tumors became palpable ∼7 days after injection, and expanded to ∼1.5 cm3 by day 15.

Animals were repeatedly microPET scanned over a 2-wk time period using [18F]FHBG as the probe to detect the transferred T cells and their progeny. Significant signal for memory T cells in the EG.7 tumor was detected as early as day 1 post-T cell transfer (Fig. 2). Using a quantitative algorithm for lymphocyte numbers and signal intensity summarized in Materials and Methods (12), memory T cells were 4.5 × 106/g at the E.G7 tumor (Table I). No signal was detected during this time in the EL4 tumor. At day 4 posttransfer, memory cells were no longer detectable at the E.G7 tumor, but cells could be localized in local lymph nodes (LN) (0.8 × 106 and 3.7 × 106 T cells/g in cervical and mediastinal LN, respectively; Table I). The amount of cells at these sites increased by day 8, with a 6.8-fold increase of cell concentration at the cervical LN. Also, the signals detected in the thorax were no longer limited to the mediastinal LNs, but now also included the entire lungs (3.7 and 11 × 106/g at day 4 and day 8). Mediastinal LNs were positive at day 8 when imaged at the optimal anatomic plane (see Table II). We were unable to detect any signal at the Ag-negative EL4 tumor throughout this time period. In contrast to memory T cells, naive T cells transferred into tumor-bearing mice were undetectable by microPET imaging, either at the E.G7 tumor site or elsewhere, during the day-10 monitoring period (Fig. 2 B).

FIGURE 2.

MicroPET imaging of T cells in tumor-bearing mice with [18F]FHBG. A total of 5 × 106 retrovirus-labeled naive or memory T cells, adjusted to 10% OT+, was transferred i.p. into tumor-bearing RAG1−/− mice. A total of 5 × 106 CD4 T cell-depleted splenocytes was also transferred as a source of T cell help. The animals received injections i.v. with [18F]FHBG on each day indicated, and were immediately imaged with microPET. Signal intensity was represented by false-color calibration of %ID/g (injection dose per gram tissue) at the ROI over background. A, Memory T cell mice; B, naive T cell mice. Data are representative of at least three independent experiments.

FIGURE 2.

MicroPET imaging of T cells in tumor-bearing mice with [18F]FHBG. A total of 5 × 106 retrovirus-labeled naive or memory T cells, adjusted to 10% OT+, was transferred i.p. into tumor-bearing RAG1−/− mice. A total of 5 × 106 CD4 T cell-depleted splenocytes was also transferred as a source of T cell help. The animals received injections i.v. with [18F]FHBG on each day indicated, and were immediately imaged with microPET. Signal intensity was represented by false-color calibration of %ID/g (injection dose per gram tissue) at the ROI over background. A, Memory T cell mice; B, naive T cell mice. Data are representative of at least three independent experiments.

Close modal
Table I.

Quantitation of memory T cell localization by microPET imaging

Memory T Cell Transfer after Tumor Inoculationa
E.G7/axillary LNCervical LNLungs/mediastinal LN
Day 1 4.5 × 106 b.d. b.d. 
Day 4 b.d. 0.8 × 106 3.7 × 106 
Day 8 b.d. 5.4 × 106 11 × 106 
Memory T Cell Transfer after Tumor Inoculationa
E.G7/axillary LNCervical LNLungs/mediastinal LN
Day 1 4.5 × 106 b.d. b.d. 
Day 4 b.d. 0.8 × 106 3.7 × 106 
Day 8 b.d. 5.4 × 106 11 × 106 
Memory T cell transfer before tumor inoculationb
E.G7/axillary LNCervical LNLungs/mediastinal LN
Day 1 0.41 × 106 0.74 × 106 0.41 × 106 
Day 4 0.20 × 106 0.22 × 106 0.24 × 106 
Day 8 b.d. b.d. b.d. 
Memory T cell transfer before tumor inoculationb
E.G7/axillary LNCervical LNLungs/mediastinal LN
Day 1 0.41 × 106 0.74 × 106 0.41 × 106 
Day 4 0.20 × 106 0.22 × 106 0.24 × 106 
Day 8 b.d. b.d. b.d. 
a

A total of 5 × 106 memory T cells (10% OT+) was transferred i.p. into tumor-bearing RAG1−/− recipients and imaged at the indicated times after T cell transfer.

b

A total of 5 × 106 memory T cells (10% OT+) was transferred i.p. into RAG1−/− mice 10 days before tumor induction and imaged at the indicated times after tumor inoculation. In both experiments, signal was detected by microPET imaging of ROI and calculated to determine cell numbers per gram of tissue; b.d., signal below detection. Data are representative of at least three independent experiments.

Table II.

Flow cytometric enumeration of memory T cells numbers recovered from anatomic sites in tumor-bearing micea

LocationLymphocyte SubsetDay 2Day 4Day 8
Spleen OT+ 7.2 ± 3.2 6.8 ± 1.0 3.3 ± 1.2 
 CD8+ 11.0 ± 5.0 8.7 ± 0.8 9.0 ± 0.9 
E.G7/axillary LN OT+ 3.2 ± 0.8 0.16 ± 0.05 0.05 ± 0.01 
 CD8+ 3.8 ± 0.4 0.60 ± 0.05 0.07 ± 0.02 
Mediastinal LN OT+ b.d. b.d. 0.3 ± 0.2 
 CD8+ 0.56 ± 0.05 1.4 ± 0.5 0.7 ± 0.2 
Lung OT+ b.d. b.d. 1.0 ± 0.5 
 CD8+ b.d. b.d. 3.8 ± 0.3 
LocationLymphocyte SubsetDay 2Day 4Day 8
Spleen OT+ 7.2 ± 3.2 6.8 ± 1.0 3.3 ± 1.2 
 CD8+ 11.0 ± 5.0 8.7 ± 0.8 9.0 ± 0.9 
E.G7/axillary LN OT+ 3.2 ± 0.8 0.16 ± 0.05 0.05 ± 0.01 
 CD8+ 3.8 ± 0.4 0.60 ± 0.05 0.07 ± 0.02 
Mediastinal LN OT+ b.d. b.d. 0.3 ± 0.2 
 CD8+ 0.56 ± 0.05 1.4 ± 0.5 0.7 ± 0.2 
Lung OT+ b.d. b.d. 1.0 ± 0.5 
 CD8+ b.d. b.d. 3.8 ± 0.3 
a

Tumor-bearing RAG1−/− recipients of 5 × 106 memory cells (as described in Fig. 2) were sacrificed at the indicated days after T cell transfer. Lymphocytes from each anatomic site were harvested, counted, and analyzed by flow cytometry for OT+ cells (H-2Kb/SIINFEKL tetramer) and CD8+ cells. Values are absolute cell numbers (×106) from each compartment. Mean and SEM were calculated for three independent experiments. b.d., Signal below detection.

To determine the in vivo cytotoxicity of the transferred T cells and their progeny, tumor growth was measured by imaging with [18F]FDG (Fig. 3) and physical measurement (Fig. 4,A). As a positive control for tumor growth without T cell action, we include data for EL4, which lacks the OVA target gene; EG.7 and EL4 have the same growth characteristics in the absence of T cells (11). On day 6, when tumors were still nonpalpable for physical measurement, significant signal was imaged from both tumor sites with [18F]FDG for all animals. By day 10, E.G7 was significantly reduced compared with EL4 in mice that received memory cells (Fig. 3,A). In contrast, E.G7 was much higher in mice that received naive T cells (Fig. 3,B), similar to control mice that did not receive any T cells (data not shown). These imaging measurements were confirmed by direct calculation of tumor volume (Fig. 4 A).

FIGURE 3.

MicroPET imaging of tumor size in tumor-bearing mice with [18F]FDG. The animals used in Fig. 2 were imaged for tumor size by i.v. injection of [18F]FDG on the indicated days post-T cell transfer. Signal intensity was represented by false-color calibration as described in Fig. 2. A, Memory T cell mice; B, naive T cell mice. Data are representative of at least three independent experiments.

FIGURE 3.

MicroPET imaging of tumor size in tumor-bearing mice with [18F]FDG. The animals used in Fig. 2 were imaged for tumor size by i.v. injection of [18F]FDG on the indicated days post-T cell transfer. Signal intensity was represented by false-color calibration as described in Fig. 2. A, Memory T cell mice; B, naive T cell mice. Data are representative of at least three independent experiments.

Close modal
FIGURE 4.

Physical tumor measurements after naive and memory T cell transfer. EL4 and E.G7 tumor sizes in mice receiving naive or memory T cells were determined as described in Materials and Methods. A, Tumors in mice used in Figs. 2 and 3. B, Tumors in mice used in Fig. 8. Mean and SEM are derived from measurements of five or more mice in each group.

FIGURE 4.

Physical tumor measurements after naive and memory T cell transfer. EL4 and E.G7 tumor sizes in mice receiving naive or memory T cells were determined as described in Materials and Methods. A, Tumors in mice used in Figs. 2 and 3. B, Tumors in mice used in Fig. 8. Mean and SEM are derived from measurements of five or more mice in each group.

Close modal

FDG signal can be seen in the urinary bladder. It should be noted that FDG is excreted by renal clearance. Although mice are manually voided of bladder contents before imaging, residual urine is variably present, and is displayed when the urinary bladder is included in the imaging plane (Fig. 3).

We were interested to assess the behavior of transferred naive and memory T cells by imaging in tumor-bearing C57BL/6 mice. Accordingly, parallel experiments were performed in which C57BL/6 rather than RAG1−/− mice were used as the recipients of tumors and T cells (data not shown). Over the same 10-day time period, neither naive nor memory T cells were detectable by microPET with [18F]FHBG. Moreover, microPET imaging with [18F]FDG PET and physical tumor measurements revealed that both Ag-positive and -negative tumors grew to the same extent as control animals that had not receive any T cell transfers. This suggested that, in an immunocompetent recipient, transferred T cells were relatively impaired for activation and expansion compared with immunodeficient recipients. This is most likely due to homeostatic proliferation of donor T cells in the RAG1−/− recipients (28, 29, 30), expanding transferred cells sufficiently for detection and tumor protection under these experimental conditions.

To validate the PET findings described above, we analyzed the transferred T cells by directly harvesting cells for flow analysis from the various anatomic sites detected by microPET imaging. In these experiments, tumor-bearing mice were transferred with retrovirus-transduced memory or naive T cells (10% OT+, 5% TK+), maintained for 2, 4, or 8 days, then microPET imaged and immediately sacrificed. Lymphocytes were prepared from anatomic sites of interest, including tumors and adjacent (axillary) LNs, cervical lymph nodes, the lungs, and spleen. Cells were stained and analyzed by flow cytometry for sr39TK expression (GFP), H-2Kb/SIINFEKL tetramer, and CD8.

The results of this analysis are shown in Fig. 5. In mice transferred with memory T cells, OT+/TK+ cell (tetramer+/GFP+) lymphocytes were detected in association with the E.G7 tumor on day 2. The frequency of TK+ cells was 5–10% of the tetramer+ cells, comparable to the efficiency of retroviral TK transduction in the donor T cell population. On day 4, tumor-associated OT+ cells were substantially reduced (<10% the frequency at day 2). During these first 4 days, OT+ cells were undetectable in the EL4 tumor, lungs, and the cervical LNs (data not shown). However, by day 8 OT+ cells became detectable in the lungs and cervical LNs. In contrast, OT+ cells were below detection in naive T cell recipients at any anatomic location and time. In summary, these flow cytometry observations matched the T cell accumulation patterns observed by microPET imaging.

FIGURE 5.

Characterization of transferred T cells by flow cytometry. Tumor-bearing RAG1−/− recipients or naive or memory T cells were sacrificed at the indicated days after T cell transfer. Lymphocytes were harvested from the E.G7 tumor and adjacent axillary LNs, the cervical LNs, and the lungs, and analyzed for GFP and H-2Kb (SIINFEKL) tetramer expression by flow cytometry. Positive cells (percentage of total lymphocyte scatter gate) are tabulated in each quadrant. Unlabeled quadrants represent percentages below 0.1%. A, Memory T cell recipients; B, naive T cell recipients. No detectable cells were found in the EL4 tumor and adjacent LNs (data not shown). Data are representative of at least three independent experiments.

FIGURE 5.

Characterization of transferred T cells by flow cytometry. Tumor-bearing RAG1−/− recipients or naive or memory T cells were sacrificed at the indicated days after T cell transfer. Lymphocytes were harvested from the E.G7 tumor and adjacent axillary LNs, the cervical LNs, and the lungs, and analyzed for GFP and H-2Kb (SIINFEKL) tetramer expression by flow cytometry. Positive cells (percentage of total lymphocyte scatter gate) are tabulated in each quadrant. Unlabeled quadrants represent percentages below 0.1%. A, Memory T cell recipients; B, naive T cell recipients. No detectable cells were found in the EL4 tumor and adjacent LNs (data not shown). Data are representative of at least three independent experiments.

Close modal

Because T cell activation is associated with transient TCR down expression, we assessed the enumeration of OT-1 T cells at the tumor and remote sites by two criteria. First, GFP expression permitted detection of transferred T cells independent of tetramer binding levels. As shown in Fig. 5, GFP+ T cells rapidly dissipated at the tumor site and accumulated at remote sites, similar to the dynamics of OT+ cells. Technically, we note that there were differences in staining intensity related to tetramer preparation. However, the mean fluorescence intensity was more than a log greater than background fluorescence, and thus permitted accurate quantitation of OT+ abundance. However, when stained under identical conditions, naive and memory OT-1 T cells display the same levels of tetramer staining (Fig. 1).

Second, to assess the Ag specificity of the homing response, we further analyzed our flow cytometric data for absolute cell numbers of OT+ vs CD8+ lymphocytes in the various anatomic sites (Table II). In the E.G7 tumor and adjacent axillary LN, the OT+ and CD8+ cells were numerically comparable, indicating that most of the signal at this Ag-specific site was contributed by Ag-reactive T cells. In the spleen, CD8+ cells remained constant during the experimental period. OT+ cells comprised 65% of the CD8+ cells at day 2, but this reduced to 33% by day 8. Because the transferred T cell population was 10% OT+, this indicated that the recovered cells, even in populations distant from the antigenic (tumor) site, were enriched for OT+ cells. It should be noted that because of biliary/intestinal excretion of FHBG, background signal in the abdomen is too high for microPET imaging of the spleen with this probe (12). Taken together, these lines of evidence further validate that OT+ memory cells rapidly dissipate from the tumor site and accumulate at remote sites.

We wanted to determine whether the abundance and localization of T cells might be due to Ag-independent homeostatic proliferation. To do so, 5 × 106 naive or memory T cells (with 5 × 106 CD8-depleted splenocytes for T cell help) were transferred i.p. into RAG1−/− mice, and the mice were repeatedly scanned with [18F]FHBG over 10 days. Over this time period, we were unable to detect the localization of either memory or naive T cells (data not shown). We then inoculated these T cell-bearing mice with E.G7 and EL4, and imaged them over an additional 10 days by microPET for T cell localization (with [18F]FHBG) and tumor growth (with [18F]FDG).

The results of this assessment are shown in Fig. 6 and Table I (footnote b). Memory T cells were detectable at low levels in the E.G7 tumor/axillary LN 1 day after the tumor inoculation; in contrast, signal at the EL4 tumor site was below the limit of detection. T cells diminished in number at this site over time. T cells were also detectable in the mediastinal LN, lungs, and cervical LN with a similar diminution over time. No signal was detectable in mice bearing naive T cells (data not shown). The antigenic specificity and memory dependence of the parked T cells thus was comparable to T cells transferred directly into tumor-bearing mice.

FIGURE 6.

MicroPET imaging of T cells and tumors in mice transferred with T cells before tumor inoculation. A total of 5 × 106 naive or memory T cells (10% OT+) was transferred i.p. into RAG1−/− mice with 5 × 106 CD8 T cell-depleted splenocytes for T cell help. Ten days later, 2 × 106 E.G7 and EL4 were injected i.m. into the right and left shoulder, respectively. On the indicated day posttumor, mice were administered [18F]FHBG or [18F]FDG i.v., and microPET imaged for T cells and tumors, respectively. A, Memory T cell mice; B, naive T cell mice. C, cervical LN; M, mediastinal LN; A, axillary LN; G.B., gall bladder. Data are representative of three or more independent experiments.

FIGURE 6.

MicroPET imaging of T cells and tumors in mice transferred with T cells before tumor inoculation. A total of 5 × 106 naive or memory T cells (10% OT+) was transferred i.p. into RAG1−/− mice with 5 × 106 CD8 T cell-depleted splenocytes for T cell help. Ten days later, 2 × 106 E.G7 and EL4 were injected i.m. into the right and left shoulder, respectively. On the indicated day posttumor, mice were administered [18F]FHBG or [18F]FDG i.v., and microPET imaged for T cells and tumors, respectively. A, Memory T cell mice; B, naive T cell mice. C, cervical LN; M, mediastinal LN; A, axillary LN; G.B., gall bladder. Data are representative of three or more independent experiments.

Close modal

The parked T cells were distinguished in two respects: 1) their cell numbers were reduced 10- to 20-fold, and 2) they diminished rather than expanded over time. In addition, parked T cells were distinguished by a failure to control tumor growth. Thus, similar, high levels of tumor growth were observed for both E.G7 and EL4, in both memory and naive T cell mice (Fig. 6). This suggests that the number and cytolytic activity of the transferred OT+ and/or helper T cell population attenuated during the 10-day period before tumor inoculation.

To determine whether naive cells are capable of localizing and eradicating the tumor, we performed a titration experiment to increase the number of transferred naive T cells. As described before, tumor-bearing RAG1−/− mice were transferred with 5 × 106 each of naive T cells CD8-depleted helper cells. However, the naive cells were increased to 35, 70, or 100% OT+ cells (Fig. 7). Using these higher numbers of OT+ naive cells, naive T cells were undetectable by microPET imaging [18F]FHBG at days 3 or 5 (data not shown). However, by day 8 they could be observed at the E.G7 tumor site (Fig. 7, left panels). Quantitatively, the 100 and 70% OT+ groups had 1.6 and 1.0 × 106 naive T cells/g tissue, respectively. The 35% OT+ group lacked detectable T cells at the E.G7 site. Among the higher groups, T cells were no longer detectable at the E.G7 site at day 13. As expected, no signal was detected under any conditions at the EL4 site. Unlike memory T cells, naive T cells could not be detected in lungs, mediastinal LN, or cervical LN during this experimental period.

FIGURE 7.

MicroPET imaging of tumor-bearing mice transferred with high numbers of naive T cells. A total of 5 × 106 naive T cells containing 100 (A), 70 (B), or 35% (C) OT+ cells was transferred i.p. together with helper cells into tumor-bearing RAG-1−/− mice. At the indicated times, mice received injections i.v. with microPET. The highest signal values on the images are shown as red, and the lowest signal values are shown as black. The signals are represented as the ratio of %ID/g (injection dose per gram tissue) at the ROI over background. Data are representative of three independent experiments.

FIGURE 7.

MicroPET imaging of tumor-bearing mice transferred with high numbers of naive T cells. A total of 5 × 106 naive T cells containing 100 (A), 70 (B), or 35% (C) OT+ cells was transferred i.p. together with helper cells into tumor-bearing RAG-1−/− mice. At the indicated times, mice received injections i.v. with microPET. The highest signal values on the images are shown as red, and the lowest signal values are shown as black. The signals are represented as the ratio of %ID/g (injection dose per gram tissue) at the ROI over background. Data are representative of three independent experiments.

Close modal

These findings indicate that naive T cells at a higher precursor frequency were also capable of detectable expansion and homing to the Ag-expressing tumor. However, this response occurred at a much slower rate compared with memory T cells. The lack of detectable accumulation in systemic sites (lungs, mediastinal and cervical LN) may reflect an authentic difference in homing properties, or that they had not yet expanded sufficiently for detection at these sites during the period of observation.

Mice were also assessed for antitumor activity in vivo by quantitation of tumor growth with [18F]FDG and microPET imaging (Fig. 7, right panels). At day 10, both E.G7 and EL4 tumors were detectable at low but similar levels in all mice. However, at day 15 the E.G7 were undetectable in mice with 70 or 100% naive T cells, whereas the EL4 tumors had progressed to large size. Mice with 35% naive T cells showed similar growth of both E.G7 and EL4 tumors. This observation was also confirmed by physical measurements of the tumor volume monitored over time (Fig. 4 B). EL4 tumors from all animals expanded to ∼2 cm3 over the 2-wk time period, whereas the E.G7 tumors did not grow in the animals with 100 and 70% OT+ naive T cells. Tumors progressed in mice with 35% OT+ naive T cells, although there was a slight delay in growth of the E.G7 tumor compared with the EL4 tumor.

In this study, we assessed the utility of microPET imaging to longitudinal monitor and quantitatively compared accumulation of transferred T cells and their progeny derived from naive vs memory during an antitumor response. Compared with naive cells, memory T cells were much more efficient in the pace and scale of T cell expansion, tumor homing, and antitumor activity in vivo. They were also distinguished by homing and persistence in local LN and in the lungs. However, both naive and memory populations were detectable at the tumor site, and at high transfer numbers, naive cells efficiently controlled tumor growth. In this discussion, we will consider how these findings compare with conventional studies comparing naive and memory cell behavior in vivo.

MicroPET imaging detected homing of memory T cells in an Ag-specific fashion to the tumor and adjacent LN as early as 1 day after transfer into tumor-bearing mice. Ag-specific tumor homing was also detected with naive T cells, but only with ∼10-fold more Ag-specific T cells, and even under these conditions after substantial delay (day 8). Others have also compared the trafficking properties of naive and memory T cells (25, 31, 32, 33, 34). Those studies show that memory T cells can directly migrate to Ag-expressing nonlymphoid tissues, and that memory T cells isolated from the nonlymphoid tissues of immunized mice have immediate ex vivo effector functions. In contrast, the trafficking of quiescent naive T cells is restricted to secondary lymphoid organs. Upon encounter with the Ags, naive T cells first need to be activated to induce clonal expansion and differentiation into effector cells (35, 36). The effector cells are competent for migration into nonlymphoid tissues, due to the up-regulation of various homing receptor systems, including selectins, chemokine receptors, and integrins, notably LFA-1 (37, 38, 39, 40, 41). The rapid tumor homing behavior of memory compared with naive T cells is consistent with the known properties of these phases of T cell differentiation.

Several observations indicated that microPET signal was attributable almost entirely to Ag-specific T cells. By flow cytometry, OT+ cells comprised most of the CD8+ cells detected in the E.G7-associated tissue and subsequently detected sites of accumulation (lung, mediastinal LN, cervical LN). TK+(GFP+) cells maintained a frequency comparable to the starting transferred population, indicating that TK expression was not selected (positively or negatively) in vivo under these experimental conditions. Notably, splenic CD8+ T cells were also enriched for OT+ cells (65 and 33% at days 2 and 4, compared with 10% in the original transferred population). This suggested that Ag-specific T cells were systemically expanded in tumor-bearing mice. Consistent with this idea, OT+ and CD8+ T cells were below the limit of detection (by microPET and flow cytometry) in tumor, lung, or LNs in mice bearing Ag-negative tumor (EL4). Similarly, the abundance of T cells was diminished when T cells were parked in mice for 10 days before tumor inoculation (presumably because of reduced survival in the absence of antigenic stimulation).

T cells transferred into immunodeficient mice also are known to undergo homeostatic proliferation (28, 29, 42, 43). There was some evidence for such expansion in this study, because OT+ T cells were minimally detectable by either microPET or flow cytometry after transfer into conventional C57BL/6 mice. However, the observation period for mice in this study (∼1 wk) was relatively short compared with reports of homeostatic proliferation, and we observed a marked reduction in cell number for T cells parked in mice before Ag-positive tumor induction. Thus, antigenic stimulation appeared to be the decisive factor in both the persistence and localization of CD8+ T cells in this study.

The difficulty imaging the tumor localization of T cells in conventional (vs “empty” immunodeficient) recipients may restrict the settings of clinical application for this methodology, because prospective patients in some cases will be relatively immunocompetent. It is possible that this limitation is overestimated in this model system, because the rapidly growing tumor permits only a brief period for T cell localization and action. In indolent tumors more characteristic of human malignancies, there are much longer surveillance times, and these may permit sufficient expansion and localization of marked T cells to permit successful imaging. This issue should be experimentally addressed in mouse models of indolent cancer formation, such as TCL1 (follicular lymphoma) and prostate membrane-specific antigen (prostate adenocarcinoma) (44, 45). We also note that cytotoxic chemotherapy remains a common treatment modality in many cancers, and after cytoreduction, such patients are lymphopenic. Because this conforms to the lymphopenic condition of the present model system, this common clinical setting may be particularly appropriate to evaluate the present strategy on a clinical basis.

MicroPET imaging permitted in situ detection of naive and memory T cell expansion after Ag stimulation. Using T cell concentrations in the tumor region (Table I) and a 0.5-g tumor size (Fig. 4), we can calculate an estimated total tumor-associated T cell numbers of 2.3 × 106 (memory T cells, day 1) and 0.8 × 106 (naive T cells, day 8). Because 5 × 105 OT+ cells were transferred, this reflects a clonal expansion of 4.6-fold for memory T cells and 1.6-fold for naive T cells. T cells were certainly localized elsewhere in the mice, so this underestimates the total systemic expansion of these T cell populations.

These estimates of memory CD8 T cell expansion are consistent with other reports (25), and similar findings with CD4 transgenic T cells in vivo have also been described (46). Several mechanisms have been proposed to explain this expansion difference between naive and memory cells: memory cells enter cell cycle with less delay, proliferate faster, and have enhanced survival (25, 35, 36, 47). Notably, the 3.5-fold greater expansion of memory compared with naive cells was comparable to a ∼3-fold expansion difference between naive and memory cells reported by Veiga-Fernandes et al. (25).

The absolute T cell numbers determined by microPET and flow cytometric quantitation were concordant. For example, at day 1–2 in the tumor site, we enumerated 4.5 × 106 to 7.2 × 106 cells, respectively. Similarly, at day 8 in the lung site, we enumerated 11 × 106 and 4.5 × 106, respectively. Such observations indicate that microPET imaging is not only a useful qualitative assessment of T cell trafficking (as reported previously; Ref.11), but actually can provide an accurate method for noninvasive quantitation of T cell accumulation in ROI. It should be emphasized that achieving such stoichiometry requires a quantitation model with empirically measured parameters for marker gene frequency and expression level, among other factors (12). Further testing in other systems should help further validate and refine this promising quantitation model.

An obvious feature of PET imaging was the early accumulation and persistence of memory cells in the LNs and lungs. We were unable to detect similar migration of naive cells, even several days after undergoing expansion and Ag-positive tumor localization. Because they had expanded to an imageable level, it seems unlikely that this observation was due to insufficient population size for detection. One explanation for this difference is a deficiency of such migratory competence in activated naive cells, as reported in some conventional homing studies (48). Alternatively, it may reflect a failure of the naive cells to survive after activation. Clonal survival of activated CD8+ cells depends on a combination of programming and maintenance functions of CD4+ T cells and APCs during CD8+ T cell memory differentiation (35, 36, 49, 50). Although CD4+ T cells were cotransferred in our experiments, their numbers or kinetics of activation may have been suboptimal to permit efficient survival programming under our experimental conditions.

Following viral or bacterial infection and effector activation, memory T cells migrate to and reside for extended periods in both infected and Ag-free nodal and noninflamed extranodal tissues for >10 mo after infection (30, 34, 51, 52). Although the homing signals for this extranodal migration are poorly understood, it is likely that the local tissue factors, notably IL-15, contribute to the survival of memory T cells at these sites (40, 53, 54, 55, 56).

The goal of this study was to assess whether microPET T cell imaging could provide a predictive endpoint for tumor immunotherapy. We observed a close correlation of tumor eradication with microPET quantitation of tumoral T cells. Thus, tumor eradication was equivalent to 4.5 × 106/g memory cells (on day 1) or 1.6 × 106/g naive T cells (day 8) at the tumor site. Because activated memory and naive T cells reach similar levels of cytotoxic activity, these findings provide a threshold density of effector CTLs needed for therapeutic efficacy. Taken together, this study demonstrates that microPET imaging robustly identifies the detailed behavior of T cells at different stages of an antitumor immune response, and can provide a quantitative biologic endpoint with therapeutic significance. PET-based immunologic imaging may offer a unique modality for longitudinal assessment of immunotherapeutic interventions targeting the CTL antitumor response.

We thank Dr. Waldemar Ladno and Judy Edwards for assistance with microPET imaging and the chemists and cyclotron crew for production of radioisotopes. We acknowledge David Stout and Andy Loening for data analysis, Dr. Bo Wei for assistance in the isolation of LNs, and Dr. Carrie Miceli for the OT-1 transgenic mice.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by National Institutes of Health Grants CA86306 (to S.S.G. and J.B.), CA09056 (to H.S.), and CA16042, Grant AI28697 from Jonsson Comprehensive Cancer Center Flow Cytometry core, and a grant from the Ruzic Medical Research Foundation (to J.B.).

3

Abbreviations used in this paper: PET, positron-emission tomography; FDG, fluorodeoxyglucose; TK, thymidine kinase; FHGB, [fluoro-3-(hydroxymethyl) butyl]guanine; ROI, region of interest; TK, thymidine kinase; LN, lymph node.

1
Rosenberg, S. A..
1999
. A new era for cancer immunotherapy based on the genes that encode cancer antigens.
Immunity
10
:
281
-287.
2
Yeh, H., C. H. June.
2003
. Use of cell-based therapies for modification of host immune responses.
Dev. Biol.
112
:
99
-104.
3
Ribas, A., L. H. Butterfield, J. A. Glaspy, J. S. Economou.
2003
. Current developments in cancer vaccines and cellular immunotherapy.
J. Clin. Oncol.
21
:
2415
-2432.
4
Clay, T. M., A. C. Hobeika, P. J. Mosca, H. Lyerly, M. A. Morse.
2001
. Assays for monitoring cellular immune responses to active immunotherapy of cancer.
Clin. Cancer Res.
7
:
1127
-1135.
5
Kircher, M. F., J. R. Allport, E. E. Graves, V. Love, L. Josephson, A. H. Lichtman, R. Weissleder.
2003
. In vivo high resolution three-dimensional imaging of antigen-specific cytotoxic T-lymphocyte trafficking to tumors.
Cancer Res.
63
:
6838
-6846.
6
Adonai, N., K. N. Nguyen, J. Walsh, M. Iyer, T. Toyokuni, M. E. Phelps, T. McCarthy, D. W. McCarthy, S. S. Gambhir.
2002
. Ex vivo cell labeling with 64Cu-pyruvaldehyde-bis(N4-methylthiosemicarbazone) for imaging cell trafficking in mice with positron-emission tomography.
Proc. Natl. Acad. Sci. USA
99
:
3030
-3035.
7
Costa, G. L., M. R. Sandora, A. Nakajima, E. V. Nguyen, C. Taylor-Edwards, A. J. Slavin, C. H. Contag, C. G. Fathman, J. M. Benson.
2001
. Adoptive immunotherapy of experimental autoimmune encephalomyelitis via T cell delivery of the IL-12 p40 subunit.
J. Immunol.
167
:
2379
-2387.
8
Hardy, J., M. Edinger, M. H. Bachmann, R. S. Negrin, C. G. Fathman, C. H. Contag.
2001
. Bioluminescence imaging of lymphocyte trafficking in vivo.
Exp. Hematol.
29
:
1353
-1360.
9
Nakajima, A., C. M. Seroogy, M. R. Sandora, I. H. Tarner, G. L. Costa, C. Taylor-Edwards, M. H. Bachmann, C. H. Contag, C. G. Fathman.
2001
. Antigen-specific T cell-mediated gene therapy in collagen-induced arthritis.
J. Clin. Invest.
107
:
1293
-1301.
10
Herschman, H. R..
2003
. Molecular imaging: looking at problems, seeing solutions.
Science
302
:
605
-608.
11
Dubey, P., H. Su, N. Adonai, S. Du, A. Rosato, J. Braun, S. S. Gambhir, O. N. Witte.
2003
. Quantitative imaging of the T cell antitumor response by positron-emission tomography.
Proc. Natl. Acad. Sci. USA
100
:
1232
-1237.
12
Su, H., A. Forbes, S. S. Gambhir, J. Braun.
2004
. Quantitation of cell number by a positron emission tomography reporter gene strategy.
Mol. Imaging Biol.
6
:
139
-148.
13
Koehne, G., M. Doubrovin, E. Doubrovina, P. Zanzonico, H. F. Gallardo, A. Ivanova, J. Balatoni, J. Teruya-Feldstein, G. Heller, C. May, et al
2003
. Serial in vivo imaging of the targeted migration of human HSV-TK-transduced antigen-specific lymphocytes.
Nat. Biotechnol.
21
:
405
-413.
14
Ponomarev, V., M. Doubrovin, C. Lyddane, T. Beresten, J. Balatoni, W. Bornman, R. Finn, T. Akhurst, S. Larson, R. Blasberg, et al
2001
. Imaging TCR-dependent NFAT-mediated T-cell activation with positron emission tomography in vivo.
Neoplasia
3
:
480
-488.
15
Carbone, F. R., M. J. Bevan.
1990
. Class I-restricted processing and presentation of exogenous cell-associated antigen in vivo.
J. Exp. Med.
171
:
377
-387.
16
Brossart, P., M. J. Bevan.
1997
. Presentation of exogenous protein antigens on major histocompatibility complex class I molecules by dendritic cells: pathway of presentation and regulation by cytokines.
Blood
90
:
1594
-1599.
17
Bennett, S. R. M., F. R. Carbone, T. Toy, J. F. A. P. Miller, W. R. Heath.
1998
. B cells directly tolerize CD8+ T cells.
J. Exp. Med.
188
:
1977
-1983.
18
Carbone, F. R., M. J. Bevan.
1989
. Induction of ovalbumin-specific cytotoxic T cells by in vivo peptide immunization.
J. Exp. Med.
169
:
603
-612.
19
Chatziioannou, A. F., S. R. Cherry, Y. Shao, R. W. Silverman, K. Meadors, T. H. Farquhar, M. Pedarsani, M. E. Phelps.
1999
. Performance of microPET: a high-resolution lutetium oxyorthosilicate PET scanner for animal imaging.
J. Nucl. Med.
40
:
1164
-1175.
20
Yu, Y., A. J. Annala, J. R. Barrio, T. Toyokuni, N. Satyamurthy, M. Namavari, S. R. Cherry, M. E. Phelps, H. R. Herschman, S. S. Gambhir.
2000
. Quantification of target gene expression by imaging reporter gene expression in living animals.
Nat. Med.
6
:
933
-937.
21
Whitmire, J. K., R. Ahmed.
2000
. Costimulation in antiviral immunity: differential requirements for CD4+ and CD8+ T cell responses.
Curr. Opin. Immunol.
12
:
448
-455.
22
Zinkernagel, R. M., M. F. Bachmann, T. M. Kundig, S. Oehen, H. Pirchet, H. Hengartner.
1996
. On immunological memory.
Annu. Rev. Immunol.
14
:
333
-67. 333–367.
23
Zimmermann, C., K. Brduscha-Riem, C. Blaser, R. M. Zinkernagel, H. Pircher.
1996
. Visualization, characterization, and turnover of CD8+ memory T cells in virus-infected hosts.
J. Exp. Med.
183
:
1367
-1375.
24
Cho, B. K., C. Y. Wang, S. Sugawa, H. N. Eisen, J. Z. Chen.
1999
. Functional differences between memory and naive CD8 T cells.
Proc. Natl. Acad. Sci. USA
96
:
2976
-2981.
25
Veiga-Fernandes, H., U. Walter, C. Bourgeois, A. McLean, B. Rocha.
2000
. Response of naive and memory CD8+ T cells to antigen stimulation in vivo.
Nat. Immunol.
1
:
47
-53.
26
Shedlock, D. J., J. K. Whitmire, J. Tan, A. S. MacDonald, R. Ahmed, H. Shen.
2003
. Role of CD4 T cell help and costimulation in CD8 T cell responses during Listeria monocytogenes infection.
J. Immunol.
170
:
2053
-2063.
27
Janssen, E. M., E. E. Lemmens, T. Wolfe, U. Christen, M. G. von Herrath, S. P. Schoenberger.
2003
. CD4+ T cells are required for secondary expansion and memory in CD8+ T lymphocytes.
Nature
421
:
852
-856.
28
Gudmundsdottir, H., L. A. Turka.
2001
. A closer look at homeostatic proliferation of CD4+ T cells: costimulatory requirements and role in memory formation.
J. Immunol.
167
:
3699
-3707.
29
Seddon, B., G. Legname, P. Tomlinson, R. Zamoyska.
2000
. Long-term survival but impaired homeostatic proliferation of naive T cells in the absence of p56lck.
Science
290
:
127
-131.
30
Kim, S. K., K. S. Schluns, L. Lefrancois.
1999
. Induction and visualization of mucosal memory CD8 T cells following systemic virus infection.
J. Immunol.
163
:
4125
-4132.
31
Sprent, J., C. D. Surh.
2002
. T cell memory.
Annu. Rev. Immunol.
20
:
551
-579.
32
Weninger, W., M. A. Crowley, N. Manjunath, U. H. von Andrian.
2001
. Migratory properties of naive, effector, and memory CD8+ T cells.
J. Exp. Med.
194
:
953
-966.
33
Weninger, W., N. Manjunath, U. H. von Andrian.
2002
. Migration and differentiation of CD8+ T cells.
Immunol. Rev.
186
:
221
-233.
34
Masopust, D., V. Vezys, A. L. Marzo, L. Lefrancois.
2001
. Preferential localization of effector memory cells in nonlymphoid tissue.
Science
291
:
2413
-2417.
35
van Stipdonk, M. J., E. E. Lemmens, S. P. Schoenberger.
2001
. Naive CTLs require a single brief period of antigenic stimulation for clonal expansion and differentiation.
Nat. Immunol.
2
:
423
-429.
36
van Stipdonk, M. J., G. Hardenberg, M. S. Bijker, E. E. Lemmens, N. M. Droin, D. R. Green, S. P. Schoenberger.
2003
. Dynamic programming of CD8+ T lymphocyte responses.
Nat. Immunol.
4
:
361
-365.
37
De Winter, H., H. Cheroutre, M. Kronenberg.
1999
. Mucosal immunity and inflammation. II. The yin and yang of T cells in intestinal inflammation: pathogenic and protective roles in a mouse colitis model.
Am. J. Physiol.
276
:
G1317
-G1321.
38
Potsch, C., D. Vohringer, H. Pircher.
1999
. Distinct migration patterns of naive and effector CD8 T cells in the spleen: correlation with CCR7 receptor expression and chemokine reactivity.
Eur. J. Immunol.
29
:
3562
-3570.
39
Sallusto, F., D. Lenig, R. Forster, M. Lipp, A. Lanzavecchia.
1999
. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions.
Nature
401
:
708
-712.
40
Lodolce, J. P., D. L. Boone, S. Chai, R. E. Swain, T. Dassopoulos, S. Trettin, A. Ma.
1998
. IL-15 receptor maintains lymphoid homeostasis by supporting lymphocyte homing and proliferation.
Immunity
9
:
669
-676.
41
Butcher, E. C., L. J. Picker.
1996
. Lymphocyte homing and homeostasis.
Science
272
:
60
-66.
42
Ploix, C., D. Lo, M. J. Carson.
2001
. A ligand for the chemokine receptor CCR7 can influence the homeostatic proliferation of CD4 T cells and progression of autoimmunity.
J. Immunol.
167
:
6724
-6730.
43
Barthlott, T., G. Kassiotis, B. Stockinger.
2003
. T cell regulation as a side effect of homeostasis and competition.
J. Exp. Med.
197
:
451
-460.
44
Hoyer, K. K., S. W. French, D. E. Turner, M. T. Nguyen, M. Renard, C. S. Malone, S. Knoetig, C. F. Qi, T. T. Su, H. Cheroutre, et al
2002
. Dysregulated TCL1 promotes multiple classes of mature B cell lymphoma.
Proc. Natl. Acad.Sci. USA
99
:
14392
-14397.
45
Watabe, T., M. Lin, H. Ide, A. A. Donjacour, G. R. Cunha, O. N. Witte, R. E. Reiter.
2002
. Growth, regeneration, and tumorigenesis of the prostate activates the PSCA promoter.
Proc. Natl. Acad. Sci. USA
99
:
401
-406.
46
Garcia, S., J. DiSanto, B. Stockinger.
1999
. Following the development of a CD4 T cell response in vivo: from activation to memory formation.
Immunity
11
:
163
-171.
47
Grayson, J. M., A. J. Zajac, J. D. Altman, R. Ahmed.
2000
. Cutting edge: increased expression of Bcl-2 in antigen-specific memory CD8+ T cells.
J. Immunol.
164
:
3950
-3954.
48
Cerwenka, A., T. M. Morgan, R. W. Dutton.
1999
. Naive, effector, and memory CD8 T cells in protection against pulmonary influenza virus infection: homing properties rather than initial frequencies are crucial.
J. Immunol.
163
:
5535
-5543.
49
Sun, J. C., M. A. Williams, M. J. Bevan.
2004
. CD4+ T cells are required for the maintenance, not programming, of memory CD8+ T cells after acute infection.
Nat. Immunol.
5
:
927
-933.
50
Smith, C. M., N. S. Wilson, J. Waithman, J. A. Villadangos, F. R. Carbone, W. R. Heath, G. T. Belz.
2004
. Cognate CD4+ T cell licensing of dendritic cells in CD8+ T cell immunity.
Nat. Immunol.
5
:
1143
-1148.
51
Klonowski, K. D., K. J. Williams, A. L. Marzo, D. A. Blair, E. G. Lingenheld, L. Lefrancois.
2004
. Dynamics of blood-borne CD8 memory T cell migration in vivo.
Immunity
20
:
551
-562.
52
Turner, S. J., G. Diaz, R. Cross, P. C. Doherty.
2004
. Analysis of clonotype distribution and persistence for an influenza virus-specific CD8+ T cell response.
Immunity
18
:
549
-559.
53
Becker, T. C., E. J. Wherry, D. Boone, K. Murali-Krishna, R. Antia, A. Ma, R. Ahmed.
2002
. Interleukin 15 is required for proliferative renewal of virus-specific memory CD8 T cells.
J. Exp. Med.
195
:
1541
-1548.
54
Judge, A. D., X. Zhang, H. Fujii, C. D. Surh, J. Sprent.
2002
. Interleukin 15 controls both proliferation and survival of a subset of memory-phenotype CD8+ T cells.
J. Exp. Med.
196
:
935
-946.
55
Schluns, K. S., K. Williams, A. Ma, X. X. Zheng, L. Lefrancois.
2002
. Cutting edge: requirement for IL-15 in the generation of primary and memory antigen-specific CD8 T cells.
J. Immunol.
168
:
4827
-4831.
56
Waldmann, T. A., S. Dubois, Y. Tagaya.
2001
. Contrasting roles of IL-2 and IL-15 in the life and death of lymphocytes: implications for immunotherapy.
Immunity
14
:
105
-110.