Individual mouse analysis of the cellular immune response to tumor antigens in peripheral blood by intracellular staining for cytokines

Journal of Immunological Methods 316 (2006) 84 – 96 www.elsevier.com/locate/jim

Research paper

Individual mouse analysis of the cellular immune response to tumor antigens in peripheral blood by intracellular staining for cytokines Patrizia Giannetti, Andrea Facciabene 1 , Nicola La Monica, Luigi Aurisicchio ⁎ Istituto di Ricerche di Biologia Molecolare (IRBM) Via Pontina km 30,600 Pomezia, 00040 Italy Received 15 February 2006; received in revised form 28 July 2006; accepted 17 August 2006 Available online 14 September 2006

Abstract Among the experimental animal models, mice remain the most widely used for the evaluation of immunotherapeutic strategies. Vaccines against parasites and viral antigens are commonly administered to the appropriate mouse strain which also allows testing of the therapeutic effect. Similarly, in mice transgenic for human tumor associated antigens (TAA), cancer vaccines must lead to breakage of immune tolerance to elicit a significant effect on the tumor. However, one of the major drawbacks in the monitoring of cellular immune responses induced by vaccination is that functional immunological assays require suppression of the animals to collect the spleen or lymph nodes for analysis. Here, we report the application of a rapid intracellular staining (ICS) method to quantify antigen-specific T cells responses in small volumes of murine blood. Genetic vaccination with plasmid DNA followed by electroporation (DNA-EP) and the use of adenoviral vectors (Ad) encoding CEA as a model target antigen were applied to different strains of mice. Optimal blood volume, number of lymphocytes, sensitivity and reproducibility of intracellular staining for IFN-γ were determined both in non-tolerant/ wild type mice as well as in tolerant CEA transgenic mice upon restimulation of PBMCs with CEA peptides. Groups of vaccinated mice were then sacrificed and PBMCs and splenocytes from individual animals were compared for intracytoplasmic detection of IFN-γ and TNF-α. A significant correlation was observed between splenic and blood immune responses. Finally, the cellular immune response was followed over time in groups of vaccinated mice. The kinetics of IFN-γ producing effectors were measured after priming and successive boosting with adenoviral vectors. We show that intracellular staining for mouse PBMCs is a rapid and simple method to measure antigen-specific immune responses. It does not require animal euthanasia and mirrors the response observed in lymphoid organs such as the spleen. © 2006 Elsevier B.V. All rights reserved. Keywords: ICS; PBMC; Cytokine; Cancer vaccine; CEA

1. Introduction

⁎ Corresponding author. Tel.: +39 06 91093 233; fax: +39 06 91093 225. E-mail address: [email protected] (L. Aurisicchio). 1 Current address: Laboratory of Immunology, Center for Research on Reproduction and Women's Health, Philadelphia, PA 19104, USA. 0022-1759/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jim.2006.08.004

Vaccination has become the standard procedure for the prevention of several infectious diseases. In fact, for many viral and parasite infections, cell mediated and/or humoral responses play an important role in pathogen clearance and the clinical outcome of infection. The application of vaccines to other diseases, such as cancer, is currently feasible owing to advances in molecular engineering and a

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

better understanding of tumor immunology (Finn, 2003; Gilboa, 2004). Several tumor associated antigens (TAA) have been identified having unique expression patterns or being overexpressed by cancer cells. These antigens, under appropriate conditions, can be recognized by components of the immune system (Campi et al., 2003; Frenoy et al., 1987; Fuchs et al., 1988; Huarte et al., 2002; Kawashima et al., 1998, 1999). Therefore, many current vaccination strategies are designed to induce antibody as well as cell mediated immune responses against the antigen of interest. One of the most studied TAA is the carcinoembryonic antigen (CEA). CEA is a membrane 180 kDa glycoprotein, which is found at high levels in the fetal colon and at lower levels in the normal adult colonic epithelium (Shively and Bettay, 1985; Thompson, 1991) and is overexpressed in 90% of colorectal, 70% of gastric, pancreatic and non-small cell lung cancers and 50% of breast cancers (Shively and Bettay, 1985; Thompson, 1991). For this reason, CEA is the target of several vaccination approaches. Among them, genetic vaccines represent promising and efficient methods with which to elicit an immune response against CEA in phase I clinical trials (Aste-Amezaga et al., 2004; Marshall et al., 2005, 1999, 2000; McAneny et al., 1996). Recently, in vivo electroporation of plasmid DNA (DNA-EP) and replication-defective recombinant adenovirus (Ad) have been shown to be safe and to induce strong humoral and cellular CEA-specific immune responses in different preclinical models (Aurisicchio et al., submitted for publication; Facciabene et al., 2004; Mennuni et al., 2005). Importantly, DNA-EP and Ad vectors coding human CEA were effective in eliciting an immune-response against CEA in tolerant CEA-transgenic (CEA.Tg) mice and were shown to confer significant tumor protection. Different immunoassays have been established to measure cellular immune responses after treatment of cancer patients with tumor vaccines. Among them the classical cytotoxic T lymphocyte assay (CTL) (Engers et al., 1975), the limiting dilution assay (LDA) (Yamada et al., 1985), and proliferation assays can define the functional properties of antigen-specific T cells. Similarly, ELISPOT (Lalvani et al., 1997), FACS based MHC tetramer (Altman et al., 1996), intracellular cytokine staining (ICS) (Murali-Krishna et al., 1998) and cytokine RT-PCR (Favre et al., 1997; Hartel et al., 1999) assays are commonly used for the sensitive and quantitative monitoring of T cell epitope-specific responses. These approaches are based on T cell-specific responses that measure antigen binding (tetramer staining) or cytokine induction in response to a specific antigen. In cytokine detection assays, peripheral blood

85

mononuclear cells (PBMCs) collected from cancer vaccine patients are commonly restimulated in vitro with tumor antigen specific peptides or irradiated cancer cells. Interferon-γ (IFN-γ) is the cytokine most often evaluated after antigen-specific restimulation. Most of these assays have been adapted and applied to pre-clinical studies. For immunotherapeutic strategies including cancer vaccines, rodents have been found to be suitable and valid models with which to determine the safety and immunogenicity of candidate vaccines that are being developed for human applications. In particular, the most widely used pre-clinical models for immunologic and anti-tumor studies are transgenic rodents expressing the human TAA (Clarke et al., 1998; Lucchini et al., 1992) which show central and/or peripheral tolerance to the antigen of interest. The application of the above described immunoassays for the evaluation of specific T cell immune responses in murine models has been widely reported in the literature (Power et al., 1999; Taguchi et al., 1990; Tough and Sprent, 1998). Most of the assays used in murine infection and cancer models are typically conducted with cells isolated from the spleen or lymph nodes to provide sufficient numbers of antigen-responsive effector cells. For this reason, vaccinated animals need to be killed and important parameters such as kinetics of the immune response and correlation between the immune response and inhibition of infection or tumor growth cannot be easily established. As a consequence, several animals have to be sacrificed at different time points, thus limiting the possibility of assessing multiple parameters in a single mouse. Recently, Hempel et al. (2002) developed an RT-PCR assay capable of quantifying the levels of interferon-γ mRNA in the cellular immune response after vaccination with a tumor antigen using small samples of whole blood from mice that did not need to be sacrificed. The peripheral and spleen immune responses were significantly correlated. However, the RT-PCR assay is delicate and cumbersome, in that it requires several steps such as mRNA extraction, cDNA synthesis, internal controls and can result in great specimen variability. Moreover, the assay is not easily applicable to large cohorts of animals, which are often required to perform statistical analysis of the data. In this study, we have applied and adapted the intracellular cytokine staining assay to murine small blood samples. CEA was used as a model target antigen and mice were vaccinated with DNA-EP/Ad, which was previously shown to be the most effective regimen (Mennuni et al., 2005). Optimal volumes of blood and PBMC cell numbers have been determined to get sensitive and reproducible results and single animal

86

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

comparison between peripheral and spleen immune response resulted in a significant correlation for the enumeration of T cells secreting IFN-γ and TNF-α. In conclusion, we show that the ICS assay from mouse PBMCs is sensitive, easily applicable and permits the determination of cell mediated response kinetics over time for each single vaccinated animal. 2. Materials and methods 2.1. Peptides and genetic vectors Lyophilized human CEA peptides were purchased by Bio-Synthesis (Lewisville, Texas, USA) and resuspended in DMSO at 40 mg/ml. Pool A (34 peptides), pool B (45 peptides), pool C (48 peptides) and pool D (53 peptides) were assembled and final concentrations were the following: pool A = 1.176 mg/ml; pool B = 0.888 mg/ml; pool C = 0.851 mg/ml; pool D = 0.769 mg/ml. Peptides and pools were stored at −80 °C. CEA nonamer (CGIQ NSVSA, H-2Db restricted), CEA107 (TYYRPGVNLSL SCHA), CEA133 (NTTYLWWVNGQSLPV) and CEA155 (SASNPSPQYSWRING) are immunodominant epitopes identified upon immunization of C57BL/6 and CEA.Tg mice (Mennuni et al., 2005). CEA107, CEA133 and CEA155 are referred as CD4+ mix. pV1J-CEAopt and Ad-CEAopt genetic vectors contain the codon optimized version of CEA cDNA and have been described elsewhere (Mennuni et al., 2005). 2.2. Animal strains C57BL/6, CEA.Tg (H-2b), BALB/c (H-2d) and A/J (H-2a) were purchased by Charles River (Lecco, Italy) and kept in standard conditions. The animals used were between 6 and 12 weeks old. Studies were performed according to Institutional Animal Care and Use Committee-approved standard animal protocols.

resected with sterile tools and processed into single-cell suspensions for immunologic analysis. 2.4. Whole blood processing Whole blood (50–300 μl) was removed from collection tubes and placed in a fresh sterile 5 ml polystyrene FACS tube (BD Falcon). Red blood cells were lysed by the addition of 4 ml of ACK Lysing buffer (Life Technologies) to each sample. Samples were mixed briefly by vortexing and incubated at room temperature for 10 min. Samples were then centrifuged at 1200 rpm for 10 min and the supernatant was removed. The cell pellet was washed once with 4 ml of R10 medium (RPMI supplemented with 10%FBS, Pen/ Strep and L-Glutammine, Gibco) and resuspended in 300 μl R10. The typical yield of peripheral blood mononuclear cells (PBMC) from a 300 μl volume bleeding was 9.2 × 105 ± 1.3 × 105 PBMCs. Viability of the cells was assessed by Trypan blue (Sigma) exclusion and by a Guava Personal Cytometer using the Guava ViaCount Reagent (Guava Technologies Inc., Hayward, CA, USA) and was usually ≥ 95%. 2.5. Spleen cell isolation To prepare splenocytes, spleens from individual mice were removed in a sterile manner, placed in 2 ml R10 medium and disrupted by scratching them through a metallic grid, thus obtaining a single-cell suspension. Cells were centrifuged at 1200 rpm for 10 min and red blood cell osmotic lysis was carried out after incubation with 10 ml of ACK buffer as described above. After spinning at 1200 rpm for 10 min, the cell pellet was resuspended in 1 ml of R10 medium and viable cells were counted by Trypan blue exclusion. The cells were then processed for intracellular staining analysis as described below. 2.6. Intracellular Staining for IFN-γ, TNF-α and IL2

2.3. Immunization of mice DNA electroporation (DNA-EP) was carried out in mice quadriceps upon injection of 50 μg of pV1J-hCEAopt and electrically stimulated as previously described (Rizzuto et al., 1999). Mice were injected in the quadriceps with 1 × 109 vp of Ad5-hCEAopt in 50 μl PBS. Two weeks after the last injection, whole blood was obtained by retro-orbital bleeding with nonheparinized capillary tubes and immediately transferred into tubes containing EDTA (Microvette 500K3E, Sarstedt, Numbrecht, Germany) and maintained with rotation at RT until processing for immunologic analysis. Groups of mice were then sacrificed, the spleens

1–3 × 106 splenocytes or PBMCs were resuspended in 600 μl R10 and CEA nonamer or peptide pool were added at a final concentration of 10 μg/ml with Golgi Plug (1 μl/ml). SEB (Staphylococcal enterotoxin B, Sigma) at 10 μg/ml was used as an internal positive control. For cells obtained from BALB/c mice, stimulation was performed with CEA pool B. For C57BL/6 and CEA.Tg mice, CEA nonamer or CD4+ mix was used while for A/J mice, pool A, B, C and D were mixed together and used as a specific stimulus. Where indicated, CEA nonamer and CD4+ mix were mixed together. After 12–16 h incubation at 37 °C, cells were

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

Fig. 1. Overview of the method used to set up the optimal amount of murine blood. Blood was taken by retro-orbital bleeding from vaccinated mice, pooled in a single tube in the presence of EDTA and divided into two aliquots. Blood was dispensed from one aliquot to determine the optimal blood volume to perform the ICS. A second aliquot was processed with erythrocyte lysis to determine the minimal and optimal number of PBMCs to perform the ICS.

87

washed with 3 ml FACS buffer (PBS, 1% FCS) and centrifuged for 10 min at RT. Incubation with anti-mouse CD16/CD32 (Fc block, clone 2.4G2) diluted 1/50 was carried out in 100 μl FACS buffer for 15 min at 4 °C. After washing, surface antigen staining was performed with APC conjugated anti-mouse CD3ε (clone 1452C11), PE conjugated anti-mouse CD4 (clone L3T4), PerCP conjugated anti-mouse CD8α (clone 53-6.7) each diluted 1/100 in FACS buffer in 100 μl final volume and incubated for 30 min at RT in the dark. Optimal antibody dilution was routinely determined for each different new stock. After washing with PermWash, cells were resuspended in 100 μl of Cytofix–Cytoperm solution, vortexed and incubated for 20 min at 4 °C in the dark. For intracellular staining, cells were incubated with FITC conjugated anti-mouse IFN-γ (Clone XMG1.2), TNF-α (clone MP6-XT22) or IL2 (clone JES6-5H4) diluted 1/ 100 in PermWash (100 μl final volume) for 30 min at RT in the dark. After washing, cells were resuspended in 300 μl 1% formaldehyde in PBS and analyzed at FacsCalibur (Becton Dickinson) using CellQuest software. All the reagents and antibodies were purchased by Pharmingen (Becton Dickinson).

Fig. 2. Determination of the optimal blood volume and PBMC number for ICS assay. Groups of 10 BALB/c and CEA.Tg mice were immunized by DNA-EP/Ad as described in the text. Two weeks after the Ad injection, blood was processed as described in Fig. 1. PBMCs were stimulated with CEA peptides as described in the Materials and methods. The cell mediated immune response as a function of blood volume is shown in panel A and C for BALB/c (non-self) and CEA.Tg (self), respectively. Panel B and D show the suitability of the assay as a function of viable cell number for BALB/c and CEA.Tg, respectively. Each measurement was performed in triplicate and the standard deviation is shown. Stimulation with DMSO and unrelated peptides resulted in 0.01% CD8+/IFN-γ+.

88

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

2.7. IFN-γ ELISPOT assay IFN-γ ELISPOT assay was performed as previously described (Mennuni et al., 2005). Briefly, ninety-six wells MAIP plates (Millipore) were coated with 100 μl/well of purified rat anti-mouse IFN-γ (IgG1, clone R4-6A2, Pharmingen) diluted to 2.5 μg/ml in sterile PBS. After washing with PBS, blocking of plates was carried out with 200 μl/well of R10 medium for 2 h at 37 °C. Splenocytes or PBMCs were plated at the indicated number in duplicate and incubated for 20 h at 37 °C with 1 μg/ml suspension of each peptide. Concanavalin Awas used as a positive internal control for each mouse at 5 μg/ml. After washing with PBS, 0.05% Tween 20, plates were incubated O/N at 4 °C with 50 μl/well of biotinconjugated rat anti-mouse IFN-γ (RatIgG1, clone XMG 1.2, Pharmingen) diluted to 1/2500 in assay buffer (PBS, 1% FBS, 0.05% Tween 20). After extensive washing, plates were developed by adding 50 μl/well NBT/B-CIP (Pierce) until development of spots was clearly visible. The reaction was stopped by washing plates thoroughly with distilled water. Plates were air dried and spots were then counted using an automated ELISPOT reader.

of central and peripheral tolerance, CEA.Tg mice were utilized. These transgenic mice carry the entire human CEA gene and flanking sequences and express the CEA protein in the intestine, mainly in the cecum and colon. Thus, this mouse line is a useful model for studying the safety and efficacy of immunotherapy strategies directed at this self tumor antigen (Clarke et al., 1998). A group of 10 CEA.Tg was subjected to four weekly electroinjections of pV1J-CEAopt followed by a final injection AdCEAopt 2 weeks later. This immunization protocol was previously shown to be the most efficient in breaking immune tolerance against the TAA in these mice (Mennuni et al., 2005). Again, blood samples were processed as described above (Fig. 1) and PBMCs were stimulated with an H-2Db-restricted immunodominant CEA specific nonamer (Mennuni et al., 2005). ICS results are shown in Fig. 2. In BALB/c mice (non-self model), a strong immune response was obtained upon immunization with genetic vectors expressing CEA. The measured immune response was not quantitatively different when volumes of blood larger than 100 μl and viable cell number greater than 5 × 105 were subjected to the assay (Fig. 2A and B, respectively). In contrast, use of 50 μl of

2.8. Statistics Pearson product moment correlation analyses were performed with SigmaStat 3.0 (SPSS Inc., Chicago, Illinois, USA). Where indicated, results were analyzed by the Student t test. A p value b 0.05 was considered significant. 3. Results 3.1. Assay standardization To determine the optimal amount of murine whole blood and number of viable PBMCs to perform the IFN-γ ICS, a group of 10 BALB/c mice was immunized with pV1J-CEAopt plasmid by DNA-EP. Two weeks later, mice were boosted by an intramuscular injection of AdCEAopt. After 14 days, 300 μl of whole blood were obtained from each mouse and pooled in a single tube. The collected group blood was then divided into two aliquots (Fig. 1): blood was dispensed from aliquot 1 in different tubes at volumes ranging from 50 to 300 μl and subsequently processed for red cell lysis; on the other hand, blood in aliquot 2 was first processed for erythrocyte lysis, then viable cells were counted and dispensed at different number (from 2.5 × 105 to 1 × 106) in assay tubes. Samples were then stimulated with CEA peptides and processed for IFN-γ ICS. To assess the effect

Fig. 3. Minimal amount of whole blood to generate at least 1 × 104 CD3+/CD8+ gated events. PBMCs obtained using different volumes of blood from immunized and naïve mice were subjected to ICS for IFN-γ. The number of CD3+/CD8+ events gated on lymphocytes and obtained with 50, 100, 150, 200 and 300 μl of blood are shown in the graphs. Data from 25 BALB/c mice (A) and from 30 CEA.Tg mice (B) are shown. Average event numbers and standard deviations are shown.

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

blood or 2.5 × 105 viable cells resulted in a significantly lower response. A possible explanation for this observation is that the total number of cells in the tube was not sufficient to get full activation of immune T lymphocytes. No significant differences were detected in CEA.Tg mice, when analyzing different volumes of blood or viable cell number: this may have been due to the presentation of the peptide epitope by blood APCs to virtually all the immune T cells elicited by vaccination and present in low numbers in the sample. However, in order to consider the data obtained significant for statistical analysis, we fixed at 10,000 the minimum number of acquired gated CD3+/ CD8+ events for sample evaluation and this was reproducibly achieved when using volumes larger than 150 μl (Fig. 3). No significant IFN-γ background (≤0.02%) was measured with DMSO or with unrelated peptides in both mouse models (data not shown). In addition, to further confirm the reproducibility of the assay and define the background level in untreated mice,

89

200 μl of whole blood were taken out from 60 CEA.Tg naïve mice and subjected to ICS using the CEA pool of peptides. All the mice showed a CD3+/CD8+/IFN-γ background ranging from 0 to 0.03% with two single exceptions (0.07% and 0.1%, data not shown). Similar results were obtained for CD4+ cells. In the light of these observations, we fixed at 0.1% (about 5–10 fold the baseline) the threshold before considering a vaccinated mouse as a responder to the treatment. 3.2. Correlation between PBMCs and splenocytes We next performed a direct comparison of the immune response in PBMCs and splenocytes obtained from individual animals by ICS. Groups of C57BL/6 and CEA.Tg mice were vaccinated as described above and two weeks after the Ad boost, mice were bled and sacrificed for spleen removal. PBMCs and splenocytes were processed side by side and stimulated with CEA

Fig. 4. CEA specific immune response elicited in PBMCs and splenocytes in the same mouse. C57BL/6 mice were vaccinated by DNA-EP/Ad combination 2 weeks apart. Fourteen days after Ad boost, each mouse was bled from the retro-orbital sinus and the spleen was removed. A) IFN-γICS was performed using CEA nonamer peptide or DMSO as negative control. Percentages of CD8+/IFN-γ+ gated on CD3+ cells are shown. The results obtained from one mouse are presented. B) CD4+ peptide mix was used for IFN-γ-ICS. Percentages of CD4+/IFN-γ+ gated on CD3+ cells are shown. The response from a representative mouse is indicated. C) IFN-γ-ELISPOT was performed using purified PBMCs or splenocytes. Cells were counted, plated at the indicated numbers and stimulated either with CEA nonamer or CD4+ mix. Average spot forming cells (SFC) from a mouse are shown. Concanavalin A was used as a positive control (not shown).

90

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

nonamer or CD4+ mix (see materials and methods). Samples from each individual were run in duplicate to measure IFN-γ and TNF-α respectively. A typical comparative ICS is shown in Fig. 4A and B. PBMCs and splenocytes from the same animal resulted in comparable percentages of IFN-γ secreting CD3+/CD8+ cells but less CD3+/CD4+ cytokine-secreting lymphocytes in the periphery. To further confirm this observation with a different assay, an IFN-γ ELISPOT was performed and gave similar results (Fig. 4C). Importantly, for CD8+ cells a high correlation was observed between splenocytes and PBMCs both in C57BL/6 and CEA.Tg mice (Fig. 5A). In particular, Pearson product moment analyses between blood and spleen effector cell

frequency assayed by ICS from the same C57BL/6 animal resulted in significant correlation for IFN-γ and TNF-α between the two cell populations (p = 0.000025 and p = 0.000438, respectively) with correlation coefficients of r = 0.95 and r = 0.897. Similar results were obtained for individual CEA.Tg mice, with Pearson product moment analysis with p = 0.00199, r = 0.847 for IFN-γ and p = 0.00038, r = 0.9 for TNF-α. CD4+ responses in C57BL/6 were also correlated and showed a Pearson product moment analysis with p = 0.057, r = 0.617 for IFN-γ and p = 0.00078, r = 0.880 for TNFα. No significant CD4+ immune response could be measured in CEA.Tg mice (not shown) as previously observed (Mennuni et al., 2005).

Fig. 5. Correlation in CEA specific immune response between blood and spleen cells. Blood and splenocytes were collected from DNA-EP/Adtreated C57BL/6 and CEA.Tg mice as described in the text. Cells were then pulsed with CEA nonamer peptide (A) or CD4+ mix (B) and cytokinesecreting cells were measured by ICS for IFN-γ and TNF-α. Pearson product moment analysis between blood and spleen cells in the same animal is shown and resulted in a significant correlation between the two cell populations for both cytokines. Correlation coefficient (r) and p value are reported for each panel.

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

To better characterize the correlation in CEA immune response between blood and spleen cells at different time points with different cytokines, C57BL/6 mice were vaccinated by DNA-EP/Ad as described above. Groups of 6 mice were sacrificed at day 0 (at the time of DNA-EP), 6, 14 (Ad injection), 18, 21, 28 and 35 post-vaccination. At each time point PBMCs and splenocytes from the same mouse were stimulated with a peptide pool containing CEA nonamer and CD4+ mix peptides, then ICS for IFN-

91

γ, TNF-α and IL2 was performed. For CD8+ cells, the detection of a measurable immune response started 14 days after DNA-EP (Fig. 6A and C). Four days later, upon Ad boost, a contraction of the response was observed followed by a marked increase from day 21 onward. No significant difference between PBMCs and splenocytes was measured at all the analyzed time points both for IFNγ and TNF-α secreting cells. A similar trend of the response was noted also for CD4+ cells (Fig. 6B and D).

Fig. 6. Correlation between blood and spleen cells during the development of CEA-specific immune response. C57BL/6 mice were vaccinated by DNA-EP at day 0 and by Ad injection at day 14. Groups of 6 mice were sacrificed at 0, 6, 14, 18, 21, 28 and 35 post-vaccination and blood and spleen were collected. A) percentages of CD3+/CD8+ cells measured by IFN-γ-ICS; B) percentages of CD3+/CD4+ cells measured by IFN-γ-ICS; C) percentages of CD3+/CD8+ cells measured by TNF-α-ICS; D) percentages of CD3+/CD4+ cells measured by TNF-α-ICS; E) percentages of CD3+/ CD4+ cells measured by IL2-ICS. Each symbol corresponds to values from a single mouse. Time points where responses measured with PBMCs or with splenocytes were significantly different (Student t test, p b 0.05) are indicated by the redial symbol (#).

92

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

Fig. 7. Kinetics of PBMC IFN-γ cytokine production in a vaccination schedule. BALB/c mice were immunized at day 0 by DNA-EP and at day 21 and 120 with Ad, both genetic vectors were expressing CEA. The immune response was measured on days 35, 63, 84, 120, 136 and 160 after priming using a pool of CEA peptides. The background of the group was consistently ≤0.02% at all the analyzed time points.

However, as previously observed, a higher percentage of CD4+/IFN-γ+ and CD4+/TNF-α+ cells was reproducibly measured in the spleen at all time points. This observation was further confirmed by IL2 ICS (Fig. 6E). Differently from IFN-γ and TNF-α kinetics, the detection of CD4+/ IL2+ cells started at day 6 post-vaccination. No significant detection of CD8+/IL2+ cells was observed at any time point (data not shown). These results show that the immune response measured in the periphery is comparable to that in the spleen over time during its development and expansion

and suggest that CD8+ and CD4+ cells may have different homing dynamics in lymphoid organs. 3.3. Kinetics of the in vivo immune response against CEA Finally, the kinetics of the immune response in a typical group of mice vaccinated with genetic vectors expressing CEA was followed over time. BALB/c mice were immunized at day 0 by DNA-EP with pV1JCEAopt and at day 21 with Ad-CEAopt. Immune

Fig. 8. Kinetics of PBMC IFN-γ cytokine production in A/J mice. A/J mice were immunized on day 0 by DNA-EP and on day 21 with Ad, both genetic vectors were expressing CEA. The immune response was measured on days 35 and 63 after priming using a pool of peptides encompassing the entire CEA antigen. A) kinetics of the CD8+ specific immune response; B) kinetics of the CD4+ specific immune response. The background of the group was consistently ≤0.02% at the analyzed time points.

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

response was measured at day 35, 63, 84, 120, 136 and 160 after priming using a pool of CEA peptides (Fig. 7). The CD8+ cell mediated immune response reached its peak between 2 and 6 weeks after Ad injection and gradually returned to basal levels at day 120, when mice received a further boost with Ad-CEAopt. The immune response was promptly restored and reached maximal levels 2 weeks later. No significant CD4+ against CEA was measured in BALB/c and this was observed also in the spleen (data not shown). To further confirm the suitability of ICS from PBMCs in a mouse strain with a different genetic background, we immunized A/J mice by DNA-EP/Ad. The vaccinated mice showed a strong CD8+ immune response which peaked 2 weeks after Ad boost (Fig. 8A). In addition, a significant CD4+ immune response was also measured (Fig. 8B). These data show that ICS from PBMCs permits the monitoring over time of the cell mediated immune response in mice and is applicable to different murine strains. 4. Discussion Over the past years, murine models have become essential in the study of various immunological activities directly translatable to patients enrolled in clinical trials. Several mouse strains as well as mice transgenic for human antigens (Hance et al., 2005; Ostrand-Rosenberg, 2004; Piechocki et al., 2003) or devoid of molecules involved in the regulation of the immune system (Ronchetti et al., 2002; Sharma et al., 2005) have been successfully utilized for the characterization of novel mechanisms connected with the immune response or the identification of optimal vaccination strategies able to exert therapeutic effects in viral diseases and cancer occurrence and/or progression. However, one major limitation is that the mouse has not been satisfactorily adapted to longitudinal studies, and for conducting timedependent immunologic studies euthanasia is required at different intervals to assess the immune response in lymphoid organs such as the spleen or the lymph nodes. Of course, these cross-sectional studies require a large number of animals for each time point and must take into account the variation between individuals. Therefore, it has been difficult to monitor the effective kinetics of the cellular immune response within the same animal or to directly correlate the specific immune response and the ability of a treated animal to inhibit viral infections or tumor growth. In this study, we have applied and set up the optimal conditions for an easy and reproducible assay based on intracellular staining of cytokines to quantify specific cellular immune responses using small samples of blood

93

collected without requiring euthanasia of the animal. CEA was used as a model antigen and genetic vectors used as the vaccination platform to elicit an anti-CEA immune response. Since the level and the quality of the immune response are affected by central and peripheral tolerance, in addition to wild type mice as models where CEA represents a non-self antigen, CEA.Tg mice (Clarke et al., 1998) were used as a self model. To standardize the assay, the optimal amounts of murine blood and viable PBMCs required to accurately and reproducibly measure the immune response elicited in vaccinated mice were first determined. Volumes larger than 100 μl and viable cell numbers greater than 5 × 105 were required for optimal results both in non-self and self models (Fig. 2): for this reason and to get a number of events sufficient for statistical analysis (Fig. 3), we fixed at 200 μl the amount of blood which contains sufficient PBMCs and can be routinely obtained from an individual mouse without significant alteration of blood parameters (data not shown). The background level for IFN-γ positive PBMCs both for CD8+ and CD4+ cells in naïve and mice stimulated with DMSO or unrelated peptides was reproducibly lower than 0.02–0.03%. On this basis, we fixed at 0.1% the threshold for considering a treated mouse as a responder to the vaccine. To ask the question whether the peripheral blood lymphocytes would mirror the immune response measured in the spleen, we compared side-by-side PBMCs and splenocytes obtained from the same immunized animal for the antigen-specific secretion of two cytokines: IFN-γ and TNF-α (Fig. 4A and B). We chose these Th1type cytokines since they are produced by T cells after antigenic or mitogenic stimulation, play important roles in the regulation of the immune response (Gajewski et al., 1989) and are often connected with effective anti-viral and anti-tumoral immunologic responses. A significant linear regression correlate was found between the CD8+ CEAspecific immune response detected in the peripheral blood and spleen from the same animal (p ≤ 0.002) for both cytokines (Fig. 5A), thus indicating that PBMCs provide a good estimate of the systemic, antigen-specific immune response that is typically measured in the spleen. Interestingly, the CD4+ measured immune response was correlated but it was higher in the spleen than in the periphery (Fig. 5B). These data were also confirmed by mean of a different assay, the IFN-γ ELISPOT (Fig. 4C). The observation that the number of immune-reactive cells appear to be lower in the peripheral blood than in the spleen was shown also by others with different methods (Hempel et al., 2002) and could be due to homing dynamics in lymphoid organs at the time point chosen for analysis (Kroemer et al., 1993). To address this issue in

94

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

our system, we conducted a vaccination protocol in C57BL/6 mice and compared PBMCs and spleen kinetics utilizing a pool of well characterized immunodominant peptides to measure in the same animal both the CD8+ and CD4+ specific response. In addition to IFN-γ and TNF-α we also measured IL2 to further validate the assay for the detection of a different cytokines. Interestingly, the CD8+ specific response in the periphery and in the spleen did not significantly differ at all the analyzed time points (Student t test, p b 0.05, see Fig. 4A and C, Fig. 5A) while CEAreactive CD4+ cell count was correlated but was significantly higher in the lymphoid organ (Fig. 4B, C and Fig. 5B). These data suggest that CD4+ and CD8+ activated T cells have different circulation dynamics in the periphery. Finally, we applied the murine blood ICS to vaccinated mice and measured the immune response over time in different mouse strains. The kinetics of CD8+ (in BALB/c and A/J mice) and CD4+ (in A/J) IFNγ secretion upon antigen stimulation could be measured for each single individual over time (Figs. 7 and 8). The results show that the immune response in vaccinated mice can be highly variable though utilizing the same vaccination methods with inbred mouse strains which are supposed to show similar responses. These variations were also observed using splenocytes (Fig. 5). There are two possible explanations for this observation: 1) technical reasons such as different transduction of muscle fibers by DNA-EP and/or Ad may result in different vaccination efficacy; 2) albeit genetically identical, different individuals may mobilize a different repertoire of T cells, thus eliciting different levels of measurable immune response. Importantly, the level and the quality of the immune response may be strictly correlated with prophylactic and/or therapeutic effects on tumors. In the light of this, the murine blood ICS constitutes an easy and valid tool with which to identify predictive biomarkers for the efficacy of cancer vaccines in mouse models. To date, several methods for measuring cell mediated immune responses in mouse peripheral blood have been described such as proliferation (Peterson, 1987), RTPCR assays (Hempel et al., 2002) and, here, the ELISPOT assay (Fig. 4C). These assays are potentially applicable but present several disadvantages: 1) they are cumbersome and not easily transferred to large numbers of animals; 2) RT-PCR cannot discriminate between the vigorous secretion activity of a few cells and weak responses of a large number of lymphocytes thus appearing to be greatly dependent on the optimal time point of stimulation, which is specific to each cytokine for a given antigen (Listanova et al., 2003); 3) the ELISPOT assay requires accurate cell counting for

reliable cytokine-secreting cell enumeration; 4) among the assays available to monitor the immune response, they are less sensitive and efficient as compared to ICS (Tassignon et al., 2005) and cannot permit focusing on a specific T cell subpopulation (CD4+ , CD8+, etc.). Likewise, methods to find correlates between immunologic responses and anti-tumor efficacy have been based on invasive methods such as partial or total splenectomy (Rosato et al., 2003). A potential drawback to the application of murine blood ICS to vaccinated tumor-bearing animals is that the presence of antigenic neoplastic lesions might dramatically alter lymphocyte recirculation and homing to lymphoid tissues, thus modifying their number in peripheral blood and changing the existing correlation between blood and spleen. Indeed, preliminary data from our laboratory obtained with tumor-bearing mice vaccinated with DNA-EP/Ad show that the immune response is severely hampered and the number of responder animals drastically reduced. However, the correlation between the periphery and the spleen is still maintained and most importantly, immunologic responders show a better anti-tumor effect (Aurisicchio et al., submitted for publication). 5. Conclusion In conclusion, the murine blood ICS described here provides a practical method to optimize immunization regimens to find potential correlates between the quality and the potency of the immunologic response and therapeutic effects in settings comparable to those used in clinical studies in humans. Acknowledgements This work was supported in part by FIRB Grant RBME017BC4 from Italian MIUR. We thank Prof. Gennaro Ciliberto for critical reading of the manuscript and IRBM animal house personnel for excellent technical assistance. References Altman, J.D., Moss, P.A., Goulder, P.J., Barouch, D.H., McHeyzerWilliams, M.G., Bell, J.I., McMichael, A.J., Davis, M.M., 1996. Phenotypic analysis of antigen-specific T lymphocytes. Science 274 (5284), 94. Aste-Amezaga, M., Bett, A.J., Wang, F., Casimiro, D.R., Antonello, J.M., Patel, D.K., Dell, E.C., Franlin, L.L., Dougherty, N.M., Bennett, P.S., Perry, H.C., Davies, M.E., Shiver, J.W., Keller, P.M., Yeager, M.D., 2004. Quantitative adenovirus neutralization assays based on the secreted alkaline phosphatase reporter gene: application in

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96 epidemiologic studies and in the design of adenovector vaccines. Hum. Gene Ther. 15 (3), 293. Aurisicchio, L., Mennuni, C., Giannetti, P., Calvaruso, F., Nuzzo, M., Cipriani, B., Palombo, F., Monaci, P., Ciliberto, G., La Monica, N., submitted for publication. Immunization with plasmid DNA and Adenoviral vectors encoding the rhesus homologue of human Carcinoembryonic antigen breaks tolerance in nonhuman primates without side effects. Campi, G., C.M., Consogno, G., et al., 2003. CD4+ T cells from healthy subjects and colon cancer patients recognize a carcinoembryonic antigen-specific immunodominant epitope. Cancer Res. 63, 8481. Clarke, P., Mann, J., Simpson, J.F., Rickard-Dickson, K., Primus, F.J., 1998. Mice transgenic for human carcinoembryonic antigen as a model for immunotherapy. Cancer Res. 58, 1469. Engers, H.D., Thomas, K., Cerottini, J.C., Brunner, K.T., 1975. Generation of cytotoxic T lymphocytes in vitro. V. Response of normal and immune spleen cells to subcellular alloantigens. J. Immunol. 115 (2), 356. Facciabene, A., Aurisicchio, L., La Monica, N., 2004. Baculovirus vectors elicit antigen-specific immune responses in mice. J. Virol. 78 (16), 8663. Favre, N., Bordmann, G., Rudin, W., Kammula, U.S., Marincola, F.M., Rosemberg, S.A., 1997. Comparison of cytokine measurements using ELISA, ELISPOT and semi-quantitative RT-PCR. J. Immunol. Methods 204, 57. Finn, O.J., 2003. Cancer vaccine: between the idea and reality. Nat. Rev., Immunol. 3, 360. Frenoy, N., S.J., Cahour, A., Burtin, P., 1987. Natural antibodies against the carcinoembryonic antigen (CEA) and a related antigen, NCA, in human sera. Anticancer Res. 7, 1229. Fuchs, C., K.F., Kern, P., Hoferichter, S., Jager, W., Kaldn, J.R., 1988. CEA-containing immune complexes in sera of patients with colorectal and breast cancer: analysis of complexed immunoglobulin classes. Cancer Immunol. Immunother. 26, 180. Gajewski, T., Schell, S.R., Nau, G., Fitch, F.W., 1989. Regulation of T cell activation: differences among T-cell subsets. Immunol. Rev. 111, 79. Gilboa, E., 2004. The promise of cancer vaccine. Nat. Rev., Cancer 4, 401. Hance, K.W., Zeytin, H.E., Greiner, J.W., 2005. Mouse models expressing human carcinoembryonic antigen (CEA) as a transgene: evaluation of CEA-based cancer vaccines. Mutat. Res. 576 (1–2), 132. Hartel, C., Bein, G., Kirchner, H., Kluter, H., 1999. A human wholeblood assay for analysis of T-cell function by quantification of cytokine mRNA. Scand. J. Immunol. 49 (6), 649. Hempel, D.M., Smith, K.A., Claussen, K.A., Perricone, M.A., 2002. Analysis of cellular immune responses in the peripheral blood of mice using real-time RT-PCR. J. Immunol. Methods 259 (1–2), 129. Huarte, E., S.P., Lasarte, J.J., et al., 2002. Identification of HLA-B27restricted cytotoxic T lymphocyte epitope from carcinoembrynic antigen. Int. J. Cancer 97, 58. Kawashima, I., H.S., Tsai, V., et al., 1998. The multiepitope approach for immunotherapy for cancer: identification of several CTL epitopes from tumor-associated antigens expressed on solid epithelial tumors. Hum. Immunol. 59, 1. Kawashima, I., T.V., Southwood, S., Takesako, K., Sette, A., Celis, E., 1999. Identification of HLA-A3-restricted cytotoxic T lymphocyte epitopes from carcinoembryonic antigen and HER-2/neu y primary in vitro immunization with peptide pulsed dendritic cells. Cancer Res. 59, 431.

95

Kroemer, G., Cuende, E., Martinez, C., 1993. Compartmentalization of the peripheral immune system. Adv. Immunol. 53, 157. Lalvani, A., Brookes, R., Hambleton, S., Britton, W.J., Hill, A.V., McMichael, A.J., 1997. Rapid effector function in CD8+ memory T cells. J. Exp. Med. 186 (6), 859. Listanova, S., Temmerman, S., Stordeur, P., Verscheure, V., Place, S., Zhou, L., Locht, C., Mascart, F., 2003. Optimal kinetics for quantification of antigen-induced cytokines in human peripheral blood mononuclear cells by real-time PCR and by ELISA. J. Immunol. Methods 281, 27. Lucchini, F., S.M., Hu, N., Villa, A., Brown, J., Cesano, L., Mangiarini, L., Rindi, G., Kindl, S., Sessa, F., et al., 1992. Early and multifocal tumors in breast, salivary, harderian and epididymal tissues developed in MMTY-Neu transgenic mice. Cancer Lett. 64, 203. Marshall, J.L., Hawkins, M.J., Tsang, K.Y., Richmond, E., Pedicano, J.E., Zhu, M.-Z., Schlom, J., 1999. Phase I study in cancer patients of a replication defective avipox recombinant vaccine that expresses human carcinoembryonic antigen. J. Clin. Oncol. 17, 332. Marshall, J.L., Hoyer, R.J., Toomey, M.A., Faraguna, K., Chang, P., Richmond, E., Pedicano, J.E., Gehan, E., Peck, R.A., Arlen, P., Tsang, K.Y., Schlom, J., 2000. Phase I study in cancer patients of a diversified prime and boost vaccination protocol using recombinant vaccinia virus and recombinant nonreplicating avipox virus to elicit anti-carcinoembryonic antigen immune responses. J. Clin. Oncol. 18, 3964. Marshall, J.L., Gulley, J.L., Arlen, P.M., Beetham, P.K., Tsang, K.Y., Slack, R., Hodge, J.W., Doren, S., Grosenbach, D.W., Hwang, J., Fox, E., Odogwu, L., Park, S., Panicali, D., Schlom, J., 2005. Phase I study of sequential vaccinations with fowlpox-CEA(6D)TRICOM alone and sequentially with vaccinia-CEA(6D)-TRICOM, with and without granulocyte-macrophage colony-stimulating factor, in patients with carcinoembryonic antigen-expressing carcinomas. J. Clin. Oncol. 23 (4), 720. McAneny, D., Ryan, C.A., Beazley, R.M., Kaufman, H.L., 1996. Results of a phase I trial of a recombinant vaccinia virus that expresses carcinoembryonic antigen in patients with advanced colorectal cancer. Ann. Surg. Oncol. 3 (495–500). Mennuni, C., Calvaruso, F., Facciabene, A., Aurisicchio, L., Storto, M., Scarselli, E., Ciliberto, G., La Monica, N., 2005. Efficient induction of T-cell responses to carcinoembryonic antigen by a heterologous prime-boost regimen using DNA and adenovirus vectors carrying a codon usage optimized cDNA. Int. J. Cancer 117 (3), 444. Murali-Krishna, K., Altman, J.D., Suresh, M., Sourdive, D.J., Zajac, A.J., Miller, J.D., Slansky, J., Ahmed, R., 1998. Counting antigenspecific CD8 T cells: a reevaluation of bystander activation during viral infection. Immunity 8 (2), 177. Ostrand-Rosenberg, S., 2004. Animal models of tumor immunity, immunotherapy and cancer vaccines. Curr. Opin. Immunol. 16 (2), 143. Peterson, W.J., 1987. A method to assess the proliferative activity of small numbers of murine peripheral blood mononuclear cells. J. Immunol. Methods 96 (2), 171. Piechocki, M.P., Ho, Y.S., Pilon, S., Wei, W.Z., 2003. Human ErbB-2 (Her-2) transgenic mice: a model system for testing Her-2 based vaccines. J. Immunol. 171 (11), 5787. Power, C.A., Grand, C.L., Ismail, N., Peters, N.C., Yurkowski, D.P., Bretscher, P.A., 1999. A valid ELISPOT assay for enumeration of ex vivo, antigen-specific, IFNgamma-producing T cells. J. Immunol. Methods 227 (1–2), 99. Rizzuto, G., Capelletti, M., Maione, D., Savino, R., Lazzaro, D., Costa, P., Mathiesen, I., Cortese, R., Ciliberto, G., Laufer, R., La

96

P. Giannetti et al. / Journal of Immunological Methods 316 (2006) 84–96

Monica, N., Fattori, E., 1999. Efficient and regulated erythropoietin production by naked DNA injection and muscle electroporation. Proc. Natl. Acad. Sci. U. S. A. 96, 6417. Ronchetti, S., Nocentini, G., Riccardi, C., Pandolfi, P.P., 2002. Role of GITR in activation response of T lymphocytes. Blood 100 (1), 350. Rosato, A., Zoso, A., Milan, G., Macino, B., Dalla Santa, S., Tosello, V., Di Carlo, E., Musiani, P., Whalen, R.G., Zanovello, P., 2003. Individual analysis of mice vaccinated against a weakly immunogenic self tumor-specific antigen reveals a correlation between CD8 T cell response and antitumor efficacy. J. Immunol. 171 (10), 5172. Sharma, R., Bagavant, H., Jarjour, W.N., Sung, S.S., Ju, S.T., 2005. The role of Fas in the immune system biology of IL-2R alpha knockout mice: interplay among regulatory T cells, inflammation, hemopoiesis, and apoptosis. J. Immunol. 175 (3), 1965. Shively, J.E., Bettay, J.D., 1985. CEA-related antigen: molecular and clinical significance. Crit. Rev. Oncol./Hematol. 2, 355. Taguchi, T., McGhee, J.R., Coffman, R.L., Beagley, K.W., Eldridge, J.H., Takatsu, K., Kiyono, H., 1990. Detection of individual mouse splenic

T cells producing IFN-gamma and IL-5 using the enzyme-linked immunospot (ELISPOT) assay. J. Immunol. Methods 128 (1), 65. Tassignon, J., Burny, W., Dahmani, S., Zhou, L., Stordeur, P., Byl, B., De Groote, D., 2005. Monitoring of cellular responses after vaccination against tetanus toxoid: comparison of the measurement of IFN-gamma production by ELISA, ELISPOT, flow cytometry and real-time PCR. J. Immunol. Methods 305, 188. Thompson, J.A.E.A., 1991. Carcinoembryonic antigen gene family: molecular biology and clinical perspective. J. Clin. Lab. Anal. 5, 344. Tough, D.F., Sprent, J., 1998. Anti-viral immunity: spotting virusspecific T cells. Curr. Biol. 8 (14), R498. Yamada, A., Ziese, M.R., Young, J.F., Yamada, Y.K., Ennis, F.A., 1985. Influenza virus hemagglutinin-specific cytotoxic T cell response induced by polypeptide produced in Escherichia coli. J. Exp. Med. 162 (2), 663.