- Email: [email protected]
A practical in vitro growth inhibition assay for the evaluation of TB vaccines
Vaccine 28 (2010) 317–322
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
A practical in vitro growth inhibition assay for the evaluation of TB vaccines Kristopher Kolibab a , Marcela Parra a , Amy L. Yang a , Liyanage P. Perera b , Steven C. Derrick a , Sheldon L. Morris a,∗ a b
Center for Biologics Evaluation and Research, United States Food and Drug Administration, Bethesda, MD, United States Metabolism Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda MD, United States
a r t i c l e
i n f o
Article history: Received 10 August 2009 Received in revised form 5 October 2009 Accepted 12 October 2009 Available online 29 October 2009 Keywords: Tuberculosis In vitro assay Vaccine Mouse Cytokines
a b s t r a c t New vaccines and novel immunization strategies are needed to improve the control of the global tuberculosis epidemic. To facilitate vaccine development, we have been creating in vitro mycobacterial intra-macrophage growth inhibition assays. Here we describe the development of an in vitro assay designed for BSL-2 laboratories which measures the capacity of vaccine-induced immune splenocytes to control the growth of isoniazid-resistant Mycobacterium bovis BCG (INHr BCG). The use of the INHr BCG as the infecting organism allows the discrimination of BCG bacilli used in murine vaccinations from BCG used in the in vitro assay. In this study, we showed that protective immune responses evoked by four different types of Mycobacterium tuberculosis vaccines [BCG, an ESAT6/Antigen 85B fusion protein formulated in DDA/MPL adjuvant, a DNA vaccine expressing the same fusion protein, and a TB Modified Vaccinia Ankara construct expressing four TB antigens (MVA-4TB)] were detected. Importantly, the levels of vaccine-induced protective immunity seen in the in vitro assay correlated with the results from in vivo protection studies in the mouse model of pulmonary tuberculosis. Furthermore, the growth inhibition data for the INHr BCG assay was similar to the previously reported results for a M. tuberculosis infection assay. The cytokine expression profiles at day 7 of the INHr BCG growth inhibition studies were also similar but not identical to the cytokine patterns detected in earlier M. tuberculosis co-culture assays. Overall, we have shown that a BSL-2 compatible in vitro growth inhibition assay using INHr BCG as the intra-macrophage target organism should be useful in developing and evaluating new TB immunization strategies. Published by Elsevier Ltd.
1. Introduction Tuberculosis remains a global public health tragedy with an estimated 2 million people dying from this disease annually. The tuberculosis epidemic has worsened in recent years primarily because of the elevated prevalence of HIV infections among TB patients in specific endemic regions, including Sub-Saharan Africa, and the increasing number of cases of multiple drug-resistant tuberculosis [1,2]. At present, the medical interventions used to treat and prevent M. tuberculosis infections are inadequate. In particular, the current vaccine against TB, Mycobacterium bovis BCG, produces incomplete and highly variable levels of protection against M. tuberculosis and is relatively ineffective in preventing adult pulmonary tuberculosis, the most infectious form of the disease [3]. To improve control of tuberculosis worldwide,
∗ Corresponding author at: Laboratory of Mycobacterial Diseases and Cellular Immunology, FDA/CBER, Building 29/Room 502, 29 Lincoln Drive, Bethesda, MD 20892, United States. Tel.: +1 301 496 5978; fax: +1 301 435 5675. E-mail address: [email protected] (S.L. Morris). 0264-410X/$ – see front matter. Published by Elsevier Ltd. doi:10.1016/j.vaccine.2009.10.047
novel TB vaccines and anti-tuberculosis medications are urgently required. The development of new vaccines against tuberculosis has been impeded by the lack of understanding of anti-tuberculosis protective mechanisms and the failure to define protective correlates of immunity. Although cell-mediated immune responses are obviously required to control M. tuberculosis infections, a complex interplay of multiple T cell subsets and immune mechanisms, and not a single specific immune activity, are likely needed to prevent tuberculous disease [4,5]. Ultimately, the capacity of the host to inhibit the growth of the TB bacilli in the macrophage, the target cell for M. tuberculosis, determines the outcome of the infection. Clearly, new tools are needed to better understand antituberculosis protective mechanisms, validate potential correlates of protective immunity, and evaluate vaccine candidates. In recent years, in vitro human and murine infection model systems have been created which can assess whether immune cells limit the growth of mycobacteria within macrophages [6–11]. These assays, which measure immune-mediated inhibition of M. tuberculosis proliferation, allow direct and relevant assessments of protective immunity. Initial studies using these in vitro mycobacterial
318
K. Kolibab et al. / Vaccine 28 (2010) 317–322
growth inhibition assays have emphasized their potential to identify acquired immune responses after BCG immunization and to dissect BCG-induced protective immune mechanisms against M. tuberculosis [6,7,9]. The relevance of these in vitro assays has been partially validated by showing that significantly enhanced inhibition of mycobacterial growth in vitro was detected after BCG vaccination of animals and humans. Despite the substantial progress in development of in vitro assays, the scope of earlier studies has been rather limited. For example, only the activity of BCG immune cells (and not immune cells induced by other candidate TB vaccines) has been evaluated in growth inhibition assays. Also, although investigators used prevention of either M. tuberculosis or BCG proliferation as assay endpoints, the vaccine-mediated in vitro control of growth of these virulent and attenuated mycobacterial strains has not been compared. Moreover, the relative mechanisms of protection against avirulent and virulent mycobacterial strains have not been analyzed in growth inhibition assays. To further evaluate these mycobacterial culture systems, we recently developed a murine in vitro growth inhibition assay for assessing the capacity of immune splenocytes to control the growth of M. tuberculosis within infected macrophages. In earlier experiments, we showed that the protective immune responses induced by different types of TB vaccine preparations (live, attenuated strains, viral vectored constructs, DNA vaccines and subunit preparations) could be detected using this murine in vitro M. tuberculosis infection assay [11]. Importantly, we demonstrated that the levels of vaccine-induced in vitro growth inhibition activity correlated with vaccine-induced in vivo protective immune responses detected at 28 days post-challenge in a mouse model of pulmonary TB. In the present study, we extended our previous findings by developing a growth inhibition assay which uses isoniazid-resistant (INHr ) BCG as the macrophage infecting organism. The substitution of BCG as the macrophage infecting organism is a significant modification because it allows the assay to be done in BSL-2 laboratories and the use of a BSL-3 lab is not required. For the INHr BCG infection assays, we again showed that levels of protective immunity detected in vitro for different candidate TB vaccines correlated with in vivo protection results in the mouse pulmonary tuberculosis model. We also demonstrated that the in vitro growth inhibition data for the BCG and the M. tuberculosis infections were generally similar. However, differences in the vaccine-induced cytokine responses were detected for the two mycobacterial infection systems suggesting that the immune mechanisms of protection against the avirulent and virulent mycobacterial strains are likely not identical. 2. Materials and methods 2.1. Animals Pathogen free C57BL/6 female mice were obtained from the Jackson Laboratories (Bar Harbour, Maine). All mice used in this study were maintained under sterile and barrier conditions at the Center for Biologics Evaluation and Research, Bethesda, MD. Mice were given sterile water, mouse chow, and bedding for the duration of the experiment. The mice were vaccinated with selected vaccine preparation at 6–8 weeks old. 2.2. Vaccine preparations The Trudeau Institute prepared the BCG Pasteur vaccine from mycobacterial culture. The SD1-DNA vaccine was constructed with the pVax DNA vector (Invitrogen, San Diego, CA) expressing an ESAT6–Antigen 85B fusion gene [12]. The pET 23B vector system (Novagen, San Diego, CA) was used to express the ESAT6–Antigen 85B (SD1) M. tuberculosis fusion protein. The SD1 protein/adjuvant
formulation was prepared by mixing the fusion protein (50 g/ml) with DDA (150 g/ml, Kodak) and MPL (250 g/ml, Avanti Polar Lipids, Alabaster, AL). The MVA vector and MVA construct (MVA4TB) expressing four tuberculosis antigens (ESAT6, Antigen 85A, Antigen 85B and Hsp65) were provided by Dr. Liyanage Perera of the National Institutes of Health, Bethesda, MD [13]. 2.3. Vaccination schedules All vaccinations were done subcutaneously except for the DNA vaccine which was given intramuscularly. Six weeks before sacrifice, C57BL/6 female mice were vaccinated once with 106 CFU of BCG. The SD1 protein/adjuvant vaccine and the adjuvant alone were given three times at 2 weeks apart with 5 g of protein per vaccination. For the DNA vaccines, 200 g of SD1-DNA and pVax vector DNA immunization groups were injected three times at 3 weeks apart. Finally, two doses of 5 × 107 pfu of the MVA vector or the MVA-4TB construct were given three weeks apart. 2.4. In vitro assay The INHr BCG strain is a highly resistant in vitro mutant of BCG Montreal (MIC for INH = 50 g/ml). In an earlier report, we had shown that the isoniazid resistance results from a deletion in the katG gene [14]. For the in vitro assay, murine bone marrow macrophages (BMM) were the target cells for INHr BCG infection. BMM were prepared by flushing the femurs of C57BL/6 female mice with Dulbecco’s modified Eagle’s medium (DMEM, Lonza, Walkersville, MD) supplemented with 10% heat-inactivated FCS, 10% L-929a conditioned medium, 1% mM l-glutamine, 1% mM HEPES buffer, 1% non-essential amino acids and 1% sodium pyruvate (complete DMEM (cDMEM)). BMM were plated at (1–1.2) × 106 per well of a 24-well plate (Costar, Corning, NY) in cDMEM and incubated at 37 ◦ C in 5% CO2 for 7 days. After the 7 days incubation, the BMM formed a confluent monolayer with a concentration of 107 cells/well. BMM were infected with INHr BCG at a multiplicity of infection of 1:100 (bacterium to BMM ratio) for 2 h at 37 ◦ C in 5% CO2 and wells were washed 5 times with PBS to eliminate non-adhering cells and extracellular bacteria. Following the last wash, PBS was replaced with 2 ml/well cDMEM and the BMM were incubated at 37 ◦ C in 5% CO2 for the remainder of the experiment. Bacterial uptake by BMM was determined by lysing a fraction of the BMM with 0.1% saponin for 5 min at room temperature. The culture lysates were serially diluted in PBS containing 0.04% Tween 80 and plated on 7H11 plates supplemented with 10% OADC (Becton Dickinson, Sparks, MD) and 50 g of isoniazid. Plates were incubated at 37 ◦ C for 2–3 weeks and colonies were counted. Growth of the INHr BCG infection within BMM was monitored by lysing cultures at days 4, 7, and 11. Again, serial dilution, plating, and counting of CFUs followed as described above. Splenocytes from mice immunized with candidate vaccines were used to determine the activity and protection of the selected vaccine against INHr BCG infection in BMM. As described previously, spleens were aseptically removed from immunized and naïve mice and smashed to obtain a single cell suspension [11]. Cell suspensions were treated with ACK lysis buffer for 4 min at room temperature, washed with cold DMEM, and added to culture flasks for 30 min incubation at 37 ◦ C to remove adherent splenic macrophages. Non-adherent splenocytes, recovered by gently pipetting the suspension, were washed and the viability was assessed by exclusion of trypan blue. A concentration of 5 × 106 of the non-adherent splenocytes were overlaid on 107 INHr BCGinfected BMM and incubated at 37 ◦ C in 5% CO2 . At the selected time point, the cells were lysed with 0.1% saponin and the diluted cell lysates were plated and counted as described above.
K. Kolibab et al. / Vaccine 28 (2010) 317–322
2.5. SuperArray analysis of cytokine responses from immune cells At the selected time point, non-adherent splenic cells were recovered from supernatants of the culture wells and stored in RNA later (Qiagen, Valencia, CA). The RNeasy mini kit (Qiagen) was used to isolate the total RNA from the cell suspensions. Equal amounts of RNA from these samples were reverse transcribed to cDNA using Superscript First-Strand cDNA Synthesis Kit (Invitrogen, San Diego, CA). The cytokine transcriptional responses were determined using the cDNA as the template for RT2 profiler cytokine PCR array system (SABiosciences, Frederick, MD) and an ABI Prism 7000 sequence detection system (Applied Biosystems, Foster City, CA). This system allows analysis of 84 different cytokines. The mRNA expression levels for each cytokine were then normalized according to the manufacturer’s instructions using the formula described in the RT2 profiler PCR array system user manual. The relative cytokine expression values in immune cell cultures were determined by dividing the cytokine expression levels in vaccine candidate samples by the expression values in naïve controls [15]. This value represents the mean increase or decrease of RNA expression compared to naïve controls. 2.6. Statistics The Graph Pad Prism 4 program was used to analyze the data from these experiments. Specifically, the Mann–Whitney test, t tests, and Wilcoxon matched pair analysis were used to evaluate the cytokine expression data. The Spearman correlation test and t tests were used to compare the differences between the in vitro assay results and the in vivo protection data. 3. Results 3.1. In vitro growth inhibition data In an earlier study, we established a murine in vitro co-culture system in which murine splenocytes from vaccinated mice inhibited the intra-macrophage growth of M. tuberculosis [11]. Although the M. tuberculosis assay is reproducible and relevant (the in vitro activity correlated with in vivo protection in the mouse model), an obvious shortcoming of this protocol is that it must be done in a BSL-3 laboratory. To make this assay amenable for practical use, we assessed whether immune splenocytes could consistently pre-
319
Table 1 Vaccine-induced inhibition of INHr BCG growth seen at days 4, 7, and 11 of the in vitro assay. Vaccine construct
Day 4
Day 7
BCG SD1 protein/adjuvant Adjuvant SD1-DNA DNA vector MVA-4TB MVA vector
0.82 ± 0.04a 0.72 ± 0.05 0.10 ± 0.06 0.50 ± 0.12 0.02 ± 0.01 0.52 ± 0.06 0
1.01 0.90 0.02 0.50 0.16 1.05 0.03
± ± ± ± ± ± ±
Day 11 0.11 0.11 0.01 0.18 0.10 0.18 0.02
1.05 1.01 0.07 0.25 0.01 1.01 0.09
± ± ± ± ± ± ±
0.18 0.21 0.05 0.11 0.01 0.07 0.05
a Mean ± SEM of the difference between naïve and experimental CFU for three to six experiments.
vent the growth of INHr BCG. The drug-resistant BCG was selected as the target organism for the assay because this strain can be used in a BSL-2 laboratory and it allowed us to differentiate the in vitro macrophage infecting organism from the vaccinating strains when BCG or BCG-derivatives were the immunizing vaccines. For our initial studies, murine bone marrow macrophages were infected with the INHr BCG strain and in vitro mycobacterial growth was evaluated over an 11 day period. A representative growth curve for the INHr BCG in vitro infection (BM) is shown in Fig. 1. In this assay, the CFU levels in the infected bone marrow macrophages were about 5.5 (log10 ) at day 0. Typically, the number of CFU increased by about 0.7 log10 during the 11 day culture period. Surprisingly, in contrast to the M. tuberculosis infections, a significant inhibition of growth was not seen when naïve cells were cultured with the INHr BCG-infected macrophages [11]. To investigate whether vaccine-induced anti-tuberculosis immunity can be detected in the INHr BCG infection cultures, mice were immunized with one of the following vaccines: live BCG vaccine, a TB DNA vaccine expressing an ESAT6–Antigen 85B fusion protein (SD1), the MVA-4TB construct, or a subunit preparation consisting of the SD1 protein formulated in DDA/MPL adjuvant [11,12]. As controls, mice were injected with only the adjuvant, the pVax DNA vector, or the empty MVA vector. Representative growth inhibition results are shown in Fig. 1 for co-cultures containing splenocytes taken from mice vaccinated with BCG, the SD1 protein/adjuvant preparation, or adjuvant alone and naïve animals. A summary of the temporal growth inhibition data for all of the study vaccines and controls is provided in Table 1. These results are presented as the reduction in INHr BCG concentrations (log10 ) measured in co-cultures containing immune spleen cells relative to mycobacterial CFU levels in cell cultures with naïve splenocytes. For these experiments, the maximal growth inhibition activity was generally detected at day 7 or 11 after the initiation of the cocultures. The CFU levels were about 10-fold lower in co-cultures containing immune splenocytes recovered from mice vaccinated Table 2 A comparison of the growth inhibition detected at day 7 for the INHr BCG and M. tuberculosis in vitro assays with the in vivo protection seen at day 28 post-challenge in the mouse model of pulmonary tuberculosis. In vitro (day 7)
Fig. 1. A representative growth curve for intra-macrophage INHr BCG cocultured with murine immune splenocytes. In this experiment, INHr BCG-infected macrophages (BM, open circles) were cultured with splenocytes from mice immunized with BCG (open triangles) or the SD1 protein/DDA-MPL adjuvant formulation (closed diamonds). As controls, infected macrophages were cultured with splenocytes from naïve mice (open squares) or mice injected only with the DDA-MPL adjuvant (closed triangles).
In vivo (day 28)
Vaccine construct
INHr BCG
BCG SD1 protein/adjuvant Adjuvant SD1-DNA DNA vector MVA-4TB MVA vector
1.01 0.90 0.02 0.50 0.16 1.05 0.03
± ± ± ± ± ± ±
0.11a 0.11 0.01 0.18 0.10 0.18 0.02
TB 0.87 0.91 0.05 0.40 0.07 0.47 0.02
Lung TB ± ± ± ± ± ± ±
0.05 0.18 0.01 0.13 0.02 0.07 0.01
1.04 1.20 0.10 0.70 0.20 0.79 0.08
± ± ± ± ± ± ±
0.11b 0.21 0.05 0.15 0.08 0.05 0.05
a Mean ± SEM of the difference between naïve and experimental CFU for three to six experiments. b Mean ± SEM of the difference between naïve and experimental lung CFU for two to four experiments.
320
K. Kolibab et al. / Vaccine 28 (2010) 317–322
Table 3 Normalized cytokine expression for the INHr BCG in vitro assay at day 7. Cytokine IFN-␥ TNF-␣ IL-21 FAS-1 IL-2 a
BCG 27.7 3.9 10.3 24.8 5.3
SD1 protein/adjuvant ± ± ± ± ±
a
12 0.9 4.5 12.6 1.7
23.7 2.2 7.5 16.4 2.9
± ± ± ± ±
14.2 0.9 5.0 7.6 1.1
Adjuvant
SD1-DNA
DNA vector
MVA-4TB
MVA vector
– – – – –
5.1 ± 2.1 – – 6.4 ± 2.2 –
– – – 3.2 ± 0.5 –
8.5 ± – 3.8 ± 18.7 ± 6.6 ±
– – – – –
3.9 2.0 9.9 2.3
Mean ± SEM for experimental gene expression/naïve gene expression at day 7 of the co-culture assay for three to six experiments.
with BCG, the SD1 protein/adjuvant formulation, and the TB MVA construct relative to the naïve controls. In contrast, moderate in vitro activity (0.5 log10 reduction relative to naïve controls) was detected in cultures containing the SD1-DNA vaccine immune splenocytes. Importantly, significant decreases (p < 0.05) in INHr BCG proliferation was seen for each vaccine candidate relevant to naïve cells and the relevant control (DDA/MPL adjuvant, DNA vector, or the MVA vector). It should be noted that the vaccine-induced growth inhibition seen in these studies is likely due to intracellular killing of mycobacteria within the infected macrophages. Only minimal release (<5%) of INHr BCG was detected in the culture supernatants before the detergent facilitated lysis of the infected macrophages (data not shown). 3.2. Comparison of in vitro growth inhibition results and the in vivo protection data Although the results in Table 1 suggested that a reproducible and practical in vitro assay for assessing vaccine effectiveness had been developed using INHr BCG intra-macrophage infections, the relevance of this system to M. tuberculosis assays was uncertain. As discussed above, we had previously established a murine M. tuberculosis co-culture system and evaluated the capacity of immune splenocytes to inhibit the growth of M. tuberculosis in bone marrow macrophages. Mice were vaccinated for that study with the same vaccines that were used in these experiments. To assess whether the in vitro growth inhibition activity detected in both mycobacterial infection systems were similar, the results from the INHr BCG and M. tuberculosis assays were compared using Spearman correlation calculations (Table 2). The non-parametric Spearman analysis showed a significant correlation between the growth inhibition data for in vitro M. tuberculosis and INHr BCG infection assays (p = 0.03). To further examine the relevance of the INHr BCG infection assay, we compared the INHr BCG in vitro growth inhibition results with in vivo protection results generated in a mouse model of pulmonary TB (Table 2). The in vivo protection of the same vaccine preparations used in this study was measured in murine vaccination/challenge studies. Immune mice were challenged by the aerosol route with 200 CFU of M. tuberculosis Erdman and sacrificed 4 weeks later to determine relative bacterial organ burdens. The in vivo protection was calculated by determining mean protective responses (naïve control lung CFU − vaccinated mice lung CFU) for two to four experiments. A comparison of the in vitro INHr BCG assay data and the in vivo protection results by Spearman analysis showed a significant correlation (p = 0.02) for the four different vaccines and control preparations. 3.3. Cytokine expression at day 7 of the INHr BCG co-culture assay Assessing the effectiveness of vaccines would be facilitated by the identification of a cytokine profile required for optimal vaccine-induced anti-tuberculosis protection. Using PCR array procedures, the expression of 84 cytokine-related genes was evaluated in non-adherent cells at 7 days after the initiation of the co-
culture assay. A list of all of the genes assessed in this study is provided in Supplementary Figure 1. The extent of expression was determined by normalizing the real-time PCR values to the expression of the Gapdh housekeeping gene and then by comparing the Gapdh-adjusted results to the levels of expression in the naïve controls. For these experiments, vaccine-induced differential regulation of cytokine expression was defined as significantly different levels of expression in naïve and immune cell cultures by the Wilcoxon matched pair analysis and expression levels at least twofold higher (or lower) than naïve controls. Table 3 shows that only five cytokine genes were consistently differentially regulated at day 7 of the co-culture with immune splenocytes and INHr BCG-infected macrophages. Not surprisingly, the expression of IFN␥ (a critical immune mediator in the anti-tuberculosis protective response) was significantly up-regulated in the vaccine co-cultures but was not elevated in the control cultures [16]. Another important protective factor, TNF-␣, was also up-regulated in the BCG and SD1 protein immune splenocyte assays at day 7. For the most effective vaccines (BCG, SD1 protein/adjuvant, and the TB MVA construct), significant enhancement of IL-2 and IL-21 expression was also detected. Finally, FAS ligand expression was substantially increased in all co-cultures containing immune splenocytes and at lower levels in the DNA vector control cultures. It should be emphasized that the expression of all five cytokines were significantly up-regulated in the most active vaccine cultures (except for TNF-␣ in the TB MVA assays) while only IFN-␥ and FAS ligand expression were increased in the co-cultures containing the immune splenocytes taken from mice immunized with the moderately active SD1-DNA vaccine. Also, the five cytokines were generally not up-regulated in control cultures. Interestingly, unlike M. tuberculosis cultures where the down-regulation of cytokine gene expression was observed, the consistent down-regulation of any of the cytokine-associated genes was not detected in the INHr BCG infection cultures [11]. 4. Discussion Relevant in vitro assays which measure the activity of vaccines designed against intracellular pathogens could be useful for assessing vaccine-induced protective immune responses, for dissecting mechanisms of protective immunity, and for evaluating vaccine potency and stability [7,17–19]. Given the potential importance of in vitro mycobacterial growth inhibition assays, our goal has been to develop a practical murine in vitro potency assay to measure the activity of TB vaccines in a BSL-2 laboratory. We hypothesized that immune-mediated in vitro mycobacterial growth inhibition activity would more directly correlate with in vivo vaccine-induced antituberculosis protective immunity than other immune response measures including IFN-␥ secretion. In an earlier study, we had validated the relevance of the murine in vitro co-culture assay by showing that in vitro vaccine-induced growth inhibition against M. tuberculosis correlated with protective responses detected in the mouse model of pulmonary tuberculosis [11]. In the current studies, we developed a reproducible and relevant mycobacterial growth inhibition assay (that could be done in BSL-2 lab) using a co-culture system containing immune splenocytes and INHr BCG-infected macrophages. The drug-resistant BCG strain was
K. Kolibab et al. / Vaccine 28 (2010) 317–322
selected as the infecting organism to allow differentiation of the intra-macrophage in vitro infection from BCG or recombinant BCG vaccinating strains used to immunize mice. Using this in vitro BSL-2 assay, the protective activity induced by four different TB vaccine types was measured. The in vitro growth inhibition results for these vaccines and relevant controls correlated with in vivo protection results from a mouse model of pulmonary TB. Importantly, the INHr BCG infection results were generally consistent with earlier data generated with the same vaccines using a M. tuberculosis infection co-culture assay. The only significant difference in growth inhibition activity between the M. tuberculosis and the INHr BCG assays was seen with the TB MVA construct immunizations (p < 0.05). Although the reasons for these differences have not been defined, the cytokine analysis demonstrated that the TB MVA vaccine did induce considerable IL-2 gene expression in the INHr BCG system while IL-2 mRNA levels were not elevated in co-cultures containing M. tuberculosis-infected macrophages and TB MVA vaccine immune splenocytes. Interestingly, in M. tuberculosis and HIV immunization studies, MVA-based vaccines have been shown to induce a high frequency of polyfunctional T cells expressing multiple cytokines including IL-2 [20–22]. The induction of polyfunctional T cells has been associated with protection against intracellular pathogens [23]. Since M. tuberculosis infections have been shown to reduce expression of specific host genes, it is possible that the relatively low levels of IL-2 expression seen in the earlier M. tuberculosis infection study for the TB MVA construct resulted from M. tuberculosis-mediated immunosupression [24,25]. The decreased levels of IL-2 and potentially reduced frequencies of polyfunctional T cells could have contributed to the limited TB MVA vaccine-induced response observed in the M. tuberculosis infection cultures. The mycobacterial growth inhibition assays which are being evaluated for use in assessing vaccine-induced protective immune responses in human clinical trials of tuberculosis vaccines have varying study designs and endpoints [6–10]. While BCG was used as the test organism in specific in vitro systems, others have used M. tuberculosis-based assays to evaluate vaccine-induced immunity in humans. A persistent concern in comparing growth inhibition data from the different assays is the uncertain relevance of using BCG (instead of M. tuberculosis) as the test strain in these assays. As noted above, our data from the mouse model suggests that the growth inhibition results from the BCG and M. tuberculosis infection systems are generally correlative. However, our data also suggest that the results of in vitro mechanistic studies using the two different infection organisms must be interpreted with caution. Clearly, defining mechanisms of protection has been difficult because intra-macrophage mycobacterial survival and replication is likely dependent on a complex interplay of host and bacterial factors and multiple inhibitory mechanisms probably contribute to the intracellular killing of these bacilli. In this study, we have identified several important differences between the INHr BCG and M. tuberculosis test systems which suggest that the in vitro protective mechanisms may not be identical. First, the infection concentration for INHr BCG at day 0 of the in vitro assay was about one log higher than the corresponding M. tuberculosis infections (5.5 log10 vs. 4.5 log10 ). Second, while naïve cells did not typically limit the growth of INHr BCG in infected macrophages, naïve cells often reduced mycobacterial growth by nearly 50% in our M. tuberculosis system. Sada-Ovalle et al. [26] recently demonstrated that the in vitro control of M. tuberculosis growth by naïve cells is primarily mediated by invariant NK T cells which are early and rapid producers of IFN-␥ after M. tuberculosis infections. It is uncertain whether BCG also activates NK T cells in these model systems. Finally, the cytokine profiles at day 7 of the INHr BCG and M. tuberculosis infection co-cultures were similar but not equivalent. In both systems, IFN-␥, TNF-␣, and IL-21 gene expression were up-regulated to sim-
321
ilar levels in BCG and SD1 protein vaccine co-cultures. However, for these same vaccines, FAS ligand and IL-2 expression were considerably up-regulated in the INHr BCG infection assays compared to M. tuberculosis assays. Other host genes were up-regulated (GDF15, IL-18, and IL-27) or down-regulated (IL-1 and Bmp-1) in the M. tuberculosis system but were not consistently induced above or below naïve control levels in BCG infection co-cultures. Overall, these moderate differences between the two test systems suggest that at least a subset of the immune mechanisms involved in the control of growth of the two organisms may be different. In sum, we have developed a relevant and practical in vitro assay for evaluating the preclinical activity of TB vaccines in a BSL-2 laboratory. In addition to being useful in preclinical screening of TB vaccines and for defining anti-mycobacterial protective mechanisms, this assay could be considered for the evaluation of vaccine potency or stability as manufacturing procedures for novel TB vaccines are developed. However, since splenocytes are used in the in vitro assay, it is uncertain whether a similar in vitro assay would be useful to estimate the anti-tuberculosis activity of varying types of vaccines being tested in human clinical studies where vaccine-induced immunogenicity is being assessed in peripheral blood lymphocytes (PBLs). Ongoing studies which are evaluating the capacity of PBLs from vaccinated mice to inhibit mycobacterial growth in vitro may indicate whether the assessment of vaccine-induced immunity in human PBLs is feasible using similar co-culture assays. Acknowledgements This project was funded in part with Federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services under IAA 227-06-1322. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.vaccine.2009.10.047. References [1] WHO. Global tuberculosis control - surveillance, planning, financing. WHO/HTM/TB; 2008. p. 393. [2] LoBue P. Extensively drug-resistant tuberculosis. Curr Opin Infect Dis 2009;22(2):167–73. [3] Colditz GA, Brewer T, Berkey CS, Wilson ME, Burdick E, Fineberg HV, et al. Efficacy of BCG vaccine in the prevention of tuberculosis. Meta-analysis of the published literature. JAMA 1994;271(9):698–702. [4] Flynn J. Immunology of tuberculosis and implications in vaccine development. Tuberculosis (Edinburg) 2004;84(1–2):93–101. [5] Cooper A. Cell-mediated immune responses in tuberculosis. Annu Rev Immunol 2009;27:393–422. [6] Cheon SH, Kampmann B, Hise AG, Phillips M, Song HY, Landen K, et al. Bactericidal activity in whole blood as a potential surrogate marker of immunity after vaccination against tuberculosis. Clin Diagn Lab Immunol 2002;9(4):901–7. [7] Hoft DF, Worku S, Kampmann B, Whalen CC, Ellner JJ, Hirsch CS, et al. Investigation of the relationships between immune-mediated inhibition of mycobacterial growth and other potential surrogate markers of protective Mycobacterium tuberculosis immunity. J Infect Dis 2002;186(10):1448–57. [8] Silver RF, Li Q, Boom WH, Ellner JJ. Lymphocyte-dependent inhibition of growth of virulent Mycobacterium tuberculosis H37Rv within human monocytes: requirement for CD4+ T cells in purified protein derivative-positive, but not in purified protein derivative-negative subjects. J Immunol 1998;160(5):2408–17. [9] Worku S, Hoft DF. In vitro measurement of protective mycobacterial immunity: antigen-specific expansion of T cells capable of inhibiting intracellular growth of bacille Calmette-Guérin. Clin Infect Dis 2000;30(Suppl. 3):S257–61. [10] Kampmann B, Gaora P, Snewin VA, Gares MP, Young DB, Levin M. Evaluation of human antimycobacterial immunity using recombinant reporter mycobacteria. J Infect Dis 2000;182(3):895–901. [11] Parra M, Yang A, Lim J, Kolibab K, Derrick S, Cadieux N, et al. The development of a murine mycobacterial growth inhibition assay for evaluating vaccines against Mycobacterium tuberculosis. Clin Vaccine Immunol 2009;16(7):1025–32.
322
K. Kolibab et al. / Vaccine 28 (2010) 317–322
[12] Derrick SC, Yang A, Morris SL. A polyvalent DNA vaccine expressing an ESAT6-Ag85B fusion protein protects mice against a primary infection with Mycobacterium tuberculosis and boosts BCG-induced protective immunity. Vaccine 2004;23(6):780–8. [13] Perera PY, Derrick SC, Kolibab K, Momoi F, Yamamoto M, Morris SL, et al. A multi-valent vaccinia virus-based tuberculosis vaccine molecularly adjuvanted with interleukin-15 induces robust immune responses in mice. Vaccine 2009;27(15):2121–7. [14] Rouse DA, Morris SL. Molecular mechanisms of isoniazid resistance in Mycobacterium tuberculosis and Mycobacterium bovis. Infect Immun 1995;63(4):1427–33. [15] Lim J, Derrick SC, Kolibab K, Yang AL, Porcelli S, Jacobs WR, et al. Early pulmonary cytokine and chemokine responses in mice immunized with three different vaccines against Mycobacterium tuberculosis determined by PCR array. Clin Vaccine Immunol 2009;16(1):122–6. [16] Flynn JL, Chan J, Triebold KJ, Dalton DK, Stewart TA, Bloom BR. An essential role for interferon gamma in resistance to Mycobacterium tuberculosis infection. J Exp Med 1993;178(6):2249–54. [17] Cowley SC, Elkins KE. Multiple T cell subsets control Francisella tularensis LVS intracellular growth without stimulation through macrophage interferon gamma receptors. J Exp Med 2003;198(3):379–89. [18] Cowley SC, Elkins KE. CD4+ T cells mediate IFN-gamma-independent control of Mycobacterium tuberculosis infection both in vitro and in vivo. J Immunol 2003;171(9):4689–99. [19] Cowley SC, Hamilton E, Frelinger JA, Su J, Forman J, Elkins KL. CD4-CD8- T cells control intracellular bacterial infections both in vitro and in vivo. J Exp Med 2005;202(2):309–19.
[20] Robinson HL, Sharma S, Zhao J, Kannanganat S, Lai L, Chennareddi L, et al. Immunogenicity in macaques of the clinical product for a clade B DNA/MVA HIV vaccine: elicitation of IFN-gamma, IL-2, and TNF-alpha coproducing CD4 and CD8 T cells. AIDS Res Hum Retroviruses 2007;23(12): 1555–62. [21] Hawkridge T, Scriba T, Gelderbloem S, Smit E, Tameris M, Moyo S, et al. Safety and immunogenicity of a new tuberculosis vaccine, MVA85A, in healthy adults in South Africa. J Infect Dis 2008;198(4):544–52. [22] Burgers WA, Chege G, Müller TL, van Harmelen JH, Khoury G, Shephard EG, et al. Broad, high-magnitude and multifunctional CD4+ and CD8+ T-cell responses elicited by a DNA and modified vaccinia Ankara vaccine containing human immunodeficiency virus type 1 subtype C genes in baboons. J Gen Virol 2009;90(Pt 2):468–80. [23] Darrah PA, Patel D, De Luca PM, Lindsay RW, Davey DF, Flynn BJ, et al. Multifunctional TH1 cells define a correlate of vaccine-mediated protection against Leishmania major. Nat Med 2007;13(7):843–50. [24] Mollenkopf HJ, Hahnke K, Kaufmann SH. Transcriptional responses in mouse lungs induced by vaccination with Mycobacterium bovis BCG and infection with Mycobacterium tuberculosis. Microbes Infect 2006;8(1):136–44. [25] Sinsimer D, Huet G, Manca C, Tsenova L, Koo MS, Kurepina N, et al. The phenolic glycolipid of Mycobacterium tuberculosis differentially modulates the early host cytokine response but does not in itself confer hypervirulence. Infect Immun 2008;76(7):3027–36. [26] Sada-Ovalle I, Chiba A, Gonzales A, Brenner MB, Behar SM. Innate invariant NKT cells recognize Mycobacterium tuberculosis-infected macrophages, produce interferon-gamma, and kill intracellular bacteria. PLoS Pathog 2008;4(12):e1000239.
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
A practical in vitro growth inhibition assay for the evaluation of TB vaccines Kristopher Kolibab a , Marcela Parra a , Amy L. Yang a , Liyanage P. Perera b , Steven C. Derrick a , Sheldon L. Morris a,∗ a b
Center for Biologics Evaluation and Research, United States Food and Drug Administration, Bethesda, MD, United States Metabolism Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda MD, United States
a r t i c l e
i n f o
Article history: Received 10 August 2009 Received in revised form 5 October 2009 Accepted 12 October 2009 Available online 29 October 2009 Keywords: Tuberculosis In vitro assay Vaccine Mouse Cytokines
a b s t r a c t New vaccines and novel immunization strategies are needed to improve the control of the global tuberculosis epidemic. To facilitate vaccine development, we have been creating in vitro mycobacterial intra-macrophage growth inhibition assays. Here we describe the development of an in vitro assay designed for BSL-2 laboratories which measures the capacity of vaccine-induced immune splenocytes to control the growth of isoniazid-resistant Mycobacterium bovis BCG (INHr BCG). The use of the INHr BCG as the infecting organism allows the discrimination of BCG bacilli used in murine vaccinations from BCG used in the in vitro assay. In this study, we showed that protective immune responses evoked by four different types of Mycobacterium tuberculosis vaccines [BCG, an ESAT6/Antigen 85B fusion protein formulated in DDA/MPL adjuvant, a DNA vaccine expressing the same fusion protein, and a TB Modified Vaccinia Ankara construct expressing four TB antigens (MVA-4TB)] were detected. Importantly, the levels of vaccine-induced protective immunity seen in the in vitro assay correlated with the results from in vivo protection studies in the mouse model of pulmonary tuberculosis. Furthermore, the growth inhibition data for the INHr BCG assay was similar to the previously reported results for a M. tuberculosis infection assay. The cytokine expression profiles at day 7 of the INHr BCG growth inhibition studies were also similar but not identical to the cytokine patterns detected in earlier M. tuberculosis co-culture assays. Overall, we have shown that a BSL-2 compatible in vitro growth inhibition assay using INHr BCG as the intra-macrophage target organism should be useful in developing and evaluating new TB immunization strategies. Published by Elsevier Ltd.
1. Introduction Tuberculosis remains a global public health tragedy with an estimated 2 million people dying from this disease annually. The tuberculosis epidemic has worsened in recent years primarily because of the elevated prevalence of HIV infections among TB patients in specific endemic regions, including Sub-Saharan Africa, and the increasing number of cases of multiple drug-resistant tuberculosis [1,2]. At present, the medical interventions used to treat and prevent M. tuberculosis infections are inadequate. In particular, the current vaccine against TB, Mycobacterium bovis BCG, produces incomplete and highly variable levels of protection against M. tuberculosis and is relatively ineffective in preventing adult pulmonary tuberculosis, the most infectious form of the disease [3]. To improve control of tuberculosis worldwide,
∗ Corresponding author at: Laboratory of Mycobacterial Diseases and Cellular Immunology, FDA/CBER, Building 29/Room 502, 29 Lincoln Drive, Bethesda, MD 20892, United States. Tel.: +1 301 496 5978; fax: +1 301 435 5675. E-mail address: [email protected] (S.L. Morris). 0264-410X/$ – see front matter. Published by Elsevier Ltd. doi:10.1016/j.vaccine.2009.10.047
novel TB vaccines and anti-tuberculosis medications are urgently required. The development of new vaccines against tuberculosis has been impeded by the lack of understanding of anti-tuberculosis protective mechanisms and the failure to define protective correlates of immunity. Although cell-mediated immune responses are obviously required to control M. tuberculosis infections, a complex interplay of multiple T cell subsets and immune mechanisms, and not a single specific immune activity, are likely needed to prevent tuberculous disease [4,5]. Ultimately, the capacity of the host to inhibit the growth of the TB bacilli in the macrophage, the target cell for M. tuberculosis, determines the outcome of the infection. Clearly, new tools are needed to better understand antituberculosis protective mechanisms, validate potential correlates of protective immunity, and evaluate vaccine candidates. In recent years, in vitro human and murine infection model systems have been created which can assess whether immune cells limit the growth of mycobacteria within macrophages [6–11]. These assays, which measure immune-mediated inhibition of M. tuberculosis proliferation, allow direct and relevant assessments of protective immunity. Initial studies using these in vitro mycobacterial
318
K. Kolibab et al. / Vaccine 28 (2010) 317–322
growth inhibition assays have emphasized their potential to identify acquired immune responses after BCG immunization and to dissect BCG-induced protective immune mechanisms against M. tuberculosis [6,7,9]. The relevance of these in vitro assays has been partially validated by showing that significantly enhanced inhibition of mycobacterial growth in vitro was detected after BCG vaccination of animals and humans. Despite the substantial progress in development of in vitro assays, the scope of earlier studies has been rather limited. For example, only the activity of BCG immune cells (and not immune cells induced by other candidate TB vaccines) has been evaluated in growth inhibition assays. Also, although investigators used prevention of either M. tuberculosis or BCG proliferation as assay endpoints, the vaccine-mediated in vitro control of growth of these virulent and attenuated mycobacterial strains has not been compared. Moreover, the relative mechanisms of protection against avirulent and virulent mycobacterial strains have not been analyzed in growth inhibition assays. To further evaluate these mycobacterial culture systems, we recently developed a murine in vitro growth inhibition assay for assessing the capacity of immune splenocytes to control the growth of M. tuberculosis within infected macrophages. In earlier experiments, we showed that the protective immune responses induced by different types of TB vaccine preparations (live, attenuated strains, viral vectored constructs, DNA vaccines and subunit preparations) could be detected using this murine in vitro M. tuberculosis infection assay [11]. Importantly, we demonstrated that the levels of vaccine-induced in vitro growth inhibition activity correlated with vaccine-induced in vivo protective immune responses detected at 28 days post-challenge in a mouse model of pulmonary TB. In the present study, we extended our previous findings by developing a growth inhibition assay which uses isoniazid-resistant (INHr ) BCG as the macrophage infecting organism. The substitution of BCG as the macrophage infecting organism is a significant modification because it allows the assay to be done in BSL-2 laboratories and the use of a BSL-3 lab is not required. For the INHr BCG infection assays, we again showed that levels of protective immunity detected in vitro for different candidate TB vaccines correlated with in vivo protection results in the mouse pulmonary tuberculosis model. We also demonstrated that the in vitro growth inhibition data for the BCG and the M. tuberculosis infections were generally similar. However, differences in the vaccine-induced cytokine responses were detected for the two mycobacterial infection systems suggesting that the immune mechanisms of protection against the avirulent and virulent mycobacterial strains are likely not identical. 2. Materials and methods 2.1. Animals Pathogen free C57BL/6 female mice were obtained from the Jackson Laboratories (Bar Harbour, Maine). All mice used in this study were maintained under sterile and barrier conditions at the Center for Biologics Evaluation and Research, Bethesda, MD. Mice were given sterile water, mouse chow, and bedding for the duration of the experiment. The mice were vaccinated with selected vaccine preparation at 6–8 weeks old. 2.2. Vaccine preparations The Trudeau Institute prepared the BCG Pasteur vaccine from mycobacterial culture. The SD1-DNA vaccine was constructed with the pVax DNA vector (Invitrogen, San Diego, CA) expressing an ESAT6–Antigen 85B fusion gene [12]. The pET 23B vector system (Novagen, San Diego, CA) was used to express the ESAT6–Antigen 85B (SD1) M. tuberculosis fusion protein. The SD1 protein/adjuvant
formulation was prepared by mixing the fusion protein (50 g/ml) with DDA (150 g/ml, Kodak) and MPL (250 g/ml, Avanti Polar Lipids, Alabaster, AL). The MVA vector and MVA construct (MVA4TB) expressing four tuberculosis antigens (ESAT6, Antigen 85A, Antigen 85B and Hsp65) were provided by Dr. Liyanage Perera of the National Institutes of Health, Bethesda, MD [13]. 2.3. Vaccination schedules All vaccinations were done subcutaneously except for the DNA vaccine which was given intramuscularly. Six weeks before sacrifice, C57BL/6 female mice were vaccinated once with 106 CFU of BCG. The SD1 protein/adjuvant vaccine and the adjuvant alone were given three times at 2 weeks apart with 5 g of protein per vaccination. For the DNA vaccines, 200 g of SD1-DNA and pVax vector DNA immunization groups were injected three times at 3 weeks apart. Finally, two doses of 5 × 107 pfu of the MVA vector or the MVA-4TB construct were given three weeks apart. 2.4. In vitro assay The INHr BCG strain is a highly resistant in vitro mutant of BCG Montreal (MIC for INH = 50 g/ml). In an earlier report, we had shown that the isoniazid resistance results from a deletion in the katG gene [14]. For the in vitro assay, murine bone marrow macrophages (BMM) were the target cells for INHr BCG infection. BMM were prepared by flushing the femurs of C57BL/6 female mice with Dulbecco’s modified Eagle’s medium (DMEM, Lonza, Walkersville, MD) supplemented with 10% heat-inactivated FCS, 10% L-929a conditioned medium, 1% mM l-glutamine, 1% mM HEPES buffer, 1% non-essential amino acids and 1% sodium pyruvate (complete DMEM (cDMEM)). BMM were plated at (1–1.2) × 106 per well of a 24-well plate (Costar, Corning, NY) in cDMEM and incubated at 37 ◦ C in 5% CO2 for 7 days. After the 7 days incubation, the BMM formed a confluent monolayer with a concentration of 107 cells/well. BMM were infected with INHr BCG at a multiplicity of infection of 1:100 (bacterium to BMM ratio) for 2 h at 37 ◦ C in 5% CO2 and wells were washed 5 times with PBS to eliminate non-adhering cells and extracellular bacteria. Following the last wash, PBS was replaced with 2 ml/well cDMEM and the BMM were incubated at 37 ◦ C in 5% CO2 for the remainder of the experiment. Bacterial uptake by BMM was determined by lysing a fraction of the BMM with 0.1% saponin for 5 min at room temperature. The culture lysates were serially diluted in PBS containing 0.04% Tween 80 and plated on 7H11 plates supplemented with 10% OADC (Becton Dickinson, Sparks, MD) and 50 g of isoniazid. Plates were incubated at 37 ◦ C for 2–3 weeks and colonies were counted. Growth of the INHr BCG infection within BMM was monitored by lysing cultures at days 4, 7, and 11. Again, serial dilution, plating, and counting of CFUs followed as described above. Splenocytes from mice immunized with candidate vaccines were used to determine the activity and protection of the selected vaccine against INHr BCG infection in BMM. As described previously, spleens were aseptically removed from immunized and naïve mice and smashed to obtain a single cell suspension [11]. Cell suspensions were treated with ACK lysis buffer for 4 min at room temperature, washed with cold DMEM, and added to culture flasks for 30 min incubation at 37 ◦ C to remove adherent splenic macrophages. Non-adherent splenocytes, recovered by gently pipetting the suspension, were washed and the viability was assessed by exclusion of trypan blue. A concentration of 5 × 106 of the non-adherent splenocytes were overlaid on 107 INHr BCGinfected BMM and incubated at 37 ◦ C in 5% CO2 . At the selected time point, the cells were lysed with 0.1% saponin and the diluted cell lysates were plated and counted as described above.
K. Kolibab et al. / Vaccine 28 (2010) 317–322
2.5. SuperArray analysis of cytokine responses from immune cells At the selected time point, non-adherent splenic cells were recovered from supernatants of the culture wells and stored in RNA later (Qiagen, Valencia, CA). The RNeasy mini kit (Qiagen) was used to isolate the total RNA from the cell suspensions. Equal amounts of RNA from these samples were reverse transcribed to cDNA using Superscript First-Strand cDNA Synthesis Kit (Invitrogen, San Diego, CA). The cytokine transcriptional responses were determined using the cDNA as the template for RT2 profiler cytokine PCR array system (SABiosciences, Frederick, MD) and an ABI Prism 7000 sequence detection system (Applied Biosystems, Foster City, CA). This system allows analysis of 84 different cytokines. The mRNA expression levels for each cytokine were then normalized according to the manufacturer’s instructions using the formula described in the RT2 profiler PCR array system user manual. The relative cytokine expression values in immune cell cultures were determined by dividing the cytokine expression levels in vaccine candidate samples by the expression values in naïve controls [15]. This value represents the mean increase or decrease of RNA expression compared to naïve controls. 2.6. Statistics The Graph Pad Prism 4 program was used to analyze the data from these experiments. Specifically, the Mann–Whitney test, t tests, and Wilcoxon matched pair analysis were used to evaluate the cytokine expression data. The Spearman correlation test and t tests were used to compare the differences between the in vitro assay results and the in vivo protection data. 3. Results 3.1. In vitro growth inhibition data In an earlier study, we established a murine in vitro co-culture system in which murine splenocytes from vaccinated mice inhibited the intra-macrophage growth of M. tuberculosis [11]. Although the M. tuberculosis assay is reproducible and relevant (the in vitro activity correlated with in vivo protection in the mouse model), an obvious shortcoming of this protocol is that it must be done in a BSL-3 laboratory. To make this assay amenable for practical use, we assessed whether immune splenocytes could consistently pre-
319
Table 1 Vaccine-induced inhibition of INHr BCG growth seen at days 4, 7, and 11 of the in vitro assay. Vaccine construct
Day 4
Day 7
BCG SD1 protein/adjuvant Adjuvant SD1-DNA DNA vector MVA-4TB MVA vector
0.82 ± 0.04a 0.72 ± 0.05 0.10 ± 0.06 0.50 ± 0.12 0.02 ± 0.01 0.52 ± 0.06 0
1.01 0.90 0.02 0.50 0.16 1.05 0.03
± ± ± ± ± ± ±
Day 11 0.11 0.11 0.01 0.18 0.10 0.18 0.02
1.05 1.01 0.07 0.25 0.01 1.01 0.09
± ± ± ± ± ± ±
0.18 0.21 0.05 0.11 0.01 0.07 0.05
a Mean ± SEM of the difference between naïve and experimental CFU for three to six experiments.
vent the growth of INHr BCG. The drug-resistant BCG was selected as the target organism for the assay because this strain can be used in a BSL-2 laboratory and it allowed us to differentiate the in vitro macrophage infecting organism from the vaccinating strains when BCG or BCG-derivatives were the immunizing vaccines. For our initial studies, murine bone marrow macrophages were infected with the INHr BCG strain and in vitro mycobacterial growth was evaluated over an 11 day period. A representative growth curve for the INHr BCG in vitro infection (BM) is shown in Fig. 1. In this assay, the CFU levels in the infected bone marrow macrophages were about 5.5 (log10 ) at day 0. Typically, the number of CFU increased by about 0.7 log10 during the 11 day culture period. Surprisingly, in contrast to the M. tuberculosis infections, a significant inhibition of growth was not seen when naïve cells were cultured with the INHr BCG-infected macrophages [11]. To investigate whether vaccine-induced anti-tuberculosis immunity can be detected in the INHr BCG infection cultures, mice were immunized with one of the following vaccines: live BCG vaccine, a TB DNA vaccine expressing an ESAT6–Antigen 85B fusion protein (SD1), the MVA-4TB construct, or a subunit preparation consisting of the SD1 protein formulated in DDA/MPL adjuvant [11,12]. As controls, mice were injected with only the adjuvant, the pVax DNA vector, or the empty MVA vector. Representative growth inhibition results are shown in Fig. 1 for co-cultures containing splenocytes taken from mice vaccinated with BCG, the SD1 protein/adjuvant preparation, or adjuvant alone and naïve animals. A summary of the temporal growth inhibition data for all of the study vaccines and controls is provided in Table 1. These results are presented as the reduction in INHr BCG concentrations (log10 ) measured in co-cultures containing immune spleen cells relative to mycobacterial CFU levels in cell cultures with naïve splenocytes. For these experiments, the maximal growth inhibition activity was generally detected at day 7 or 11 after the initiation of the cocultures. The CFU levels were about 10-fold lower in co-cultures containing immune splenocytes recovered from mice vaccinated Table 2 A comparison of the growth inhibition detected at day 7 for the INHr BCG and M. tuberculosis in vitro assays with the in vivo protection seen at day 28 post-challenge in the mouse model of pulmonary tuberculosis. In vitro (day 7)
Fig. 1. A representative growth curve for intra-macrophage INHr BCG cocultured with murine immune splenocytes. In this experiment, INHr BCG-infected macrophages (BM, open circles) were cultured with splenocytes from mice immunized with BCG (open triangles) or the SD1 protein/DDA-MPL adjuvant formulation (closed diamonds). As controls, infected macrophages were cultured with splenocytes from naïve mice (open squares) or mice injected only with the DDA-MPL adjuvant (closed triangles).
In vivo (day 28)
Vaccine construct
INHr BCG
BCG SD1 protein/adjuvant Adjuvant SD1-DNA DNA vector MVA-4TB MVA vector
1.01 0.90 0.02 0.50 0.16 1.05 0.03
± ± ± ± ± ± ±
0.11a 0.11 0.01 0.18 0.10 0.18 0.02
TB 0.87 0.91 0.05 0.40 0.07 0.47 0.02
Lung TB ± ± ± ± ± ± ±
0.05 0.18 0.01 0.13 0.02 0.07 0.01
1.04 1.20 0.10 0.70 0.20 0.79 0.08
± ± ± ± ± ± ±
0.11b 0.21 0.05 0.15 0.08 0.05 0.05
a Mean ± SEM of the difference between naïve and experimental CFU for three to six experiments. b Mean ± SEM of the difference between naïve and experimental lung CFU for two to four experiments.
320
K. Kolibab et al. / Vaccine 28 (2010) 317–322
Table 3 Normalized cytokine expression for the INHr BCG in vitro assay at day 7. Cytokine IFN-␥ TNF-␣ IL-21 FAS-1 IL-2 a
BCG 27.7 3.9 10.3 24.8 5.3
SD1 protein/adjuvant ± ± ± ± ±
a
12 0.9 4.5 12.6 1.7
23.7 2.2 7.5 16.4 2.9
± ± ± ± ±
14.2 0.9 5.0 7.6 1.1
Adjuvant
SD1-DNA
DNA vector
MVA-4TB
MVA vector
– – – – –
5.1 ± 2.1 – – 6.4 ± 2.2 –
– – – 3.2 ± 0.5 –
8.5 ± – 3.8 ± 18.7 ± 6.6 ±
– – – – –
3.9 2.0 9.9 2.3
Mean ± SEM for experimental gene expression/naïve gene expression at day 7 of the co-culture assay for three to six experiments.
with BCG, the SD1 protein/adjuvant formulation, and the TB MVA construct relative to the naïve controls. In contrast, moderate in vitro activity (0.5 log10 reduction relative to naïve controls) was detected in cultures containing the SD1-DNA vaccine immune splenocytes. Importantly, significant decreases (p < 0.05) in INHr BCG proliferation was seen for each vaccine candidate relevant to naïve cells and the relevant control (DDA/MPL adjuvant, DNA vector, or the MVA vector). It should be noted that the vaccine-induced growth inhibition seen in these studies is likely due to intracellular killing of mycobacteria within the infected macrophages. Only minimal release (<5%) of INHr BCG was detected in the culture supernatants before the detergent facilitated lysis of the infected macrophages (data not shown). 3.2. Comparison of in vitro growth inhibition results and the in vivo protection data Although the results in Table 1 suggested that a reproducible and practical in vitro assay for assessing vaccine effectiveness had been developed using INHr BCG intra-macrophage infections, the relevance of this system to M. tuberculosis assays was uncertain. As discussed above, we had previously established a murine M. tuberculosis co-culture system and evaluated the capacity of immune splenocytes to inhibit the growth of M. tuberculosis in bone marrow macrophages. Mice were vaccinated for that study with the same vaccines that were used in these experiments. To assess whether the in vitro growth inhibition activity detected in both mycobacterial infection systems were similar, the results from the INHr BCG and M. tuberculosis assays were compared using Spearman correlation calculations (Table 2). The non-parametric Spearman analysis showed a significant correlation between the growth inhibition data for in vitro M. tuberculosis and INHr BCG infection assays (p = 0.03). To further examine the relevance of the INHr BCG infection assay, we compared the INHr BCG in vitro growth inhibition results with in vivo protection results generated in a mouse model of pulmonary TB (Table 2). The in vivo protection of the same vaccine preparations used in this study was measured in murine vaccination/challenge studies. Immune mice were challenged by the aerosol route with 200 CFU of M. tuberculosis Erdman and sacrificed 4 weeks later to determine relative bacterial organ burdens. The in vivo protection was calculated by determining mean protective responses (naïve control lung CFU − vaccinated mice lung CFU) for two to four experiments. A comparison of the in vitro INHr BCG assay data and the in vivo protection results by Spearman analysis showed a significant correlation (p = 0.02) for the four different vaccines and control preparations. 3.3. Cytokine expression at day 7 of the INHr BCG co-culture assay Assessing the effectiveness of vaccines would be facilitated by the identification of a cytokine profile required for optimal vaccine-induced anti-tuberculosis protection. Using PCR array procedures, the expression of 84 cytokine-related genes was evaluated in non-adherent cells at 7 days after the initiation of the co-
culture assay. A list of all of the genes assessed in this study is provided in Supplementary Figure 1. The extent of expression was determined by normalizing the real-time PCR values to the expression of the Gapdh housekeeping gene and then by comparing the Gapdh-adjusted results to the levels of expression in the naïve controls. For these experiments, vaccine-induced differential regulation of cytokine expression was defined as significantly different levels of expression in naïve and immune cell cultures by the Wilcoxon matched pair analysis and expression levels at least twofold higher (or lower) than naïve controls. Table 3 shows that only five cytokine genes were consistently differentially regulated at day 7 of the co-culture with immune splenocytes and INHr BCG-infected macrophages. Not surprisingly, the expression of IFN␥ (a critical immune mediator in the anti-tuberculosis protective response) was significantly up-regulated in the vaccine co-cultures but was not elevated in the control cultures [16]. Another important protective factor, TNF-␣, was also up-regulated in the BCG and SD1 protein immune splenocyte assays at day 7. For the most effective vaccines (BCG, SD1 protein/adjuvant, and the TB MVA construct), significant enhancement of IL-2 and IL-21 expression was also detected. Finally, FAS ligand expression was substantially increased in all co-cultures containing immune splenocytes and at lower levels in the DNA vector control cultures. It should be emphasized that the expression of all five cytokines were significantly up-regulated in the most active vaccine cultures (except for TNF-␣ in the TB MVA assays) while only IFN-␥ and FAS ligand expression were increased in the co-cultures containing the immune splenocytes taken from mice immunized with the moderately active SD1-DNA vaccine. Also, the five cytokines were generally not up-regulated in control cultures. Interestingly, unlike M. tuberculosis cultures where the down-regulation of cytokine gene expression was observed, the consistent down-regulation of any of the cytokine-associated genes was not detected in the INHr BCG infection cultures [11]. 4. Discussion Relevant in vitro assays which measure the activity of vaccines designed against intracellular pathogens could be useful for assessing vaccine-induced protective immune responses, for dissecting mechanisms of protective immunity, and for evaluating vaccine potency and stability [7,17–19]. Given the potential importance of in vitro mycobacterial growth inhibition assays, our goal has been to develop a practical murine in vitro potency assay to measure the activity of TB vaccines in a BSL-2 laboratory. We hypothesized that immune-mediated in vitro mycobacterial growth inhibition activity would more directly correlate with in vivo vaccine-induced antituberculosis protective immunity than other immune response measures including IFN-␥ secretion. In an earlier study, we had validated the relevance of the murine in vitro co-culture assay by showing that in vitro vaccine-induced growth inhibition against M. tuberculosis correlated with protective responses detected in the mouse model of pulmonary tuberculosis [11]. In the current studies, we developed a reproducible and relevant mycobacterial growth inhibition assay (that could be done in BSL-2 lab) using a co-culture system containing immune splenocytes and INHr BCG-infected macrophages. The drug-resistant BCG strain was
K. Kolibab et al. / Vaccine 28 (2010) 317–322
selected as the infecting organism to allow differentiation of the intra-macrophage in vitro infection from BCG or recombinant BCG vaccinating strains used to immunize mice. Using this in vitro BSL-2 assay, the protective activity induced by four different TB vaccine types was measured. The in vitro growth inhibition results for these vaccines and relevant controls correlated with in vivo protection results from a mouse model of pulmonary TB. Importantly, the INHr BCG infection results were generally consistent with earlier data generated with the same vaccines using a M. tuberculosis infection co-culture assay. The only significant difference in growth inhibition activity between the M. tuberculosis and the INHr BCG assays was seen with the TB MVA construct immunizations (p < 0.05). Although the reasons for these differences have not been defined, the cytokine analysis demonstrated that the TB MVA vaccine did induce considerable IL-2 gene expression in the INHr BCG system while IL-2 mRNA levels were not elevated in co-cultures containing M. tuberculosis-infected macrophages and TB MVA vaccine immune splenocytes. Interestingly, in M. tuberculosis and HIV immunization studies, MVA-based vaccines have been shown to induce a high frequency of polyfunctional T cells expressing multiple cytokines including IL-2 [20–22]. The induction of polyfunctional T cells has been associated with protection against intracellular pathogens [23]. Since M. tuberculosis infections have been shown to reduce expression of specific host genes, it is possible that the relatively low levels of IL-2 expression seen in the earlier M. tuberculosis infection study for the TB MVA construct resulted from M. tuberculosis-mediated immunosupression [24,25]. The decreased levels of IL-2 and potentially reduced frequencies of polyfunctional T cells could have contributed to the limited TB MVA vaccine-induced response observed in the M. tuberculosis infection cultures. The mycobacterial growth inhibition assays which are being evaluated for use in assessing vaccine-induced protective immune responses in human clinical trials of tuberculosis vaccines have varying study designs and endpoints [6–10]. While BCG was used as the test organism in specific in vitro systems, others have used M. tuberculosis-based assays to evaluate vaccine-induced immunity in humans. A persistent concern in comparing growth inhibition data from the different assays is the uncertain relevance of using BCG (instead of M. tuberculosis) as the test strain in these assays. As noted above, our data from the mouse model suggests that the growth inhibition results from the BCG and M. tuberculosis infection systems are generally correlative. However, our data also suggest that the results of in vitro mechanistic studies using the two different infection organisms must be interpreted with caution. Clearly, defining mechanisms of protection has been difficult because intra-macrophage mycobacterial survival and replication is likely dependent on a complex interplay of host and bacterial factors and multiple inhibitory mechanisms probably contribute to the intracellular killing of these bacilli. In this study, we have identified several important differences between the INHr BCG and M. tuberculosis test systems which suggest that the in vitro protective mechanisms may not be identical. First, the infection concentration for INHr BCG at day 0 of the in vitro assay was about one log higher than the corresponding M. tuberculosis infections (5.5 log10 vs. 4.5 log10 ). Second, while naïve cells did not typically limit the growth of INHr BCG in infected macrophages, naïve cells often reduced mycobacterial growth by nearly 50% in our M. tuberculosis system. Sada-Ovalle et al. [26] recently demonstrated that the in vitro control of M. tuberculosis growth by naïve cells is primarily mediated by invariant NK T cells which are early and rapid producers of IFN-␥ after M. tuberculosis infections. It is uncertain whether BCG also activates NK T cells in these model systems. Finally, the cytokine profiles at day 7 of the INHr BCG and M. tuberculosis infection co-cultures were similar but not equivalent. In both systems, IFN-␥, TNF-␣, and IL-21 gene expression were up-regulated to sim-
321
ilar levels in BCG and SD1 protein vaccine co-cultures. However, for these same vaccines, FAS ligand and IL-2 expression were considerably up-regulated in the INHr BCG infection assays compared to M. tuberculosis assays. Other host genes were up-regulated (GDF15, IL-18, and IL-27) or down-regulated (IL-1 and Bmp-1) in the M. tuberculosis system but were not consistently induced above or below naïve control levels in BCG infection co-cultures. Overall, these moderate differences between the two test systems suggest that at least a subset of the immune mechanisms involved in the control of growth of the two organisms may be different. In sum, we have developed a relevant and practical in vitro assay for evaluating the preclinical activity of TB vaccines in a BSL-2 laboratory. In addition to being useful in preclinical screening of TB vaccines and for defining anti-mycobacterial protective mechanisms, this assay could be considered for the evaluation of vaccine potency or stability as manufacturing procedures for novel TB vaccines are developed. However, since splenocytes are used in the in vitro assay, it is uncertain whether a similar in vitro assay would be useful to estimate the anti-tuberculosis activity of varying types of vaccines being tested in human clinical studies where vaccine-induced immunogenicity is being assessed in peripheral blood lymphocytes (PBLs). Ongoing studies which are evaluating the capacity of PBLs from vaccinated mice to inhibit mycobacterial growth in vitro may indicate whether the assessment of vaccine-induced immunity in human PBLs is feasible using similar co-culture assays. Acknowledgements This project was funded in part with Federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services under IAA 227-06-1322. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.vaccine.2009.10.047. References [1] WHO. Global tuberculosis control - surveillance, planning, financing. WHO/HTM/TB; 2008. p. 393. [2] LoBue P. Extensively drug-resistant tuberculosis. Curr Opin Infect Dis 2009;22(2):167–73. [3] Colditz GA, Brewer T, Berkey CS, Wilson ME, Burdick E, Fineberg HV, et al. Efficacy of BCG vaccine in the prevention of tuberculosis. Meta-analysis of the published literature. JAMA 1994;271(9):698–702. [4] Flynn J. Immunology of tuberculosis and implications in vaccine development. Tuberculosis (Edinburg) 2004;84(1–2):93–101. [5] Cooper A. Cell-mediated immune responses in tuberculosis. Annu Rev Immunol 2009;27:393–422. [6] Cheon SH, Kampmann B, Hise AG, Phillips M, Song HY, Landen K, et al. Bactericidal activity in whole blood as a potential surrogate marker of immunity after vaccination against tuberculosis. Clin Diagn Lab Immunol 2002;9(4):901–7. [7] Hoft DF, Worku S, Kampmann B, Whalen CC, Ellner JJ, Hirsch CS, et al. Investigation of the relationships between immune-mediated inhibition of mycobacterial growth and other potential surrogate markers of protective Mycobacterium tuberculosis immunity. J Infect Dis 2002;186(10):1448–57. [8] Silver RF, Li Q, Boom WH, Ellner JJ. Lymphocyte-dependent inhibition of growth of virulent Mycobacterium tuberculosis H37Rv within human monocytes: requirement for CD4+ T cells in purified protein derivative-positive, but not in purified protein derivative-negative subjects. J Immunol 1998;160(5):2408–17. [9] Worku S, Hoft DF. In vitro measurement of protective mycobacterial immunity: antigen-specific expansion of T cells capable of inhibiting intracellular growth of bacille Calmette-Guérin. Clin Infect Dis 2000;30(Suppl. 3):S257–61. [10] Kampmann B, Gaora P, Snewin VA, Gares MP, Young DB, Levin M. Evaluation of human antimycobacterial immunity using recombinant reporter mycobacteria. J Infect Dis 2000;182(3):895–901. [11] Parra M, Yang A, Lim J, Kolibab K, Derrick S, Cadieux N, et al. The development of a murine mycobacterial growth inhibition assay for evaluating vaccines against Mycobacterium tuberculosis. Clin Vaccine Immunol 2009;16(7):1025–32.
322
K. Kolibab et al. / Vaccine 28 (2010) 317–322
[12] Derrick SC, Yang A, Morris SL. A polyvalent DNA vaccine expressing an ESAT6-Ag85B fusion protein protects mice against a primary infection with Mycobacterium tuberculosis and boosts BCG-induced protective immunity. Vaccine 2004;23(6):780–8. [13] Perera PY, Derrick SC, Kolibab K, Momoi F, Yamamoto M, Morris SL, et al. A multi-valent vaccinia virus-based tuberculosis vaccine molecularly adjuvanted with interleukin-15 induces robust immune responses in mice. Vaccine 2009;27(15):2121–7. [14] Rouse DA, Morris SL. Molecular mechanisms of isoniazid resistance in Mycobacterium tuberculosis and Mycobacterium bovis. Infect Immun 1995;63(4):1427–33. [15] Lim J, Derrick SC, Kolibab K, Yang AL, Porcelli S, Jacobs WR, et al. Early pulmonary cytokine and chemokine responses in mice immunized with three different vaccines against Mycobacterium tuberculosis determined by PCR array. Clin Vaccine Immunol 2009;16(1):122–6. [16] Flynn JL, Chan J, Triebold KJ, Dalton DK, Stewart TA, Bloom BR. An essential role for interferon gamma in resistance to Mycobacterium tuberculosis infection. J Exp Med 1993;178(6):2249–54. [17] Cowley SC, Elkins KE. Multiple T cell subsets control Francisella tularensis LVS intracellular growth without stimulation through macrophage interferon gamma receptors. J Exp Med 2003;198(3):379–89. [18] Cowley SC, Elkins KE. CD4+ T cells mediate IFN-gamma-independent control of Mycobacterium tuberculosis infection both in vitro and in vivo. J Immunol 2003;171(9):4689–99. [19] Cowley SC, Hamilton E, Frelinger JA, Su J, Forman J, Elkins KL. CD4-CD8- T cells control intracellular bacterial infections both in vitro and in vivo. J Exp Med 2005;202(2):309–19.
[20] Robinson HL, Sharma S, Zhao J, Kannanganat S, Lai L, Chennareddi L, et al. Immunogenicity in macaques of the clinical product for a clade B DNA/MVA HIV vaccine: elicitation of IFN-gamma, IL-2, and TNF-alpha coproducing CD4 and CD8 T cells. AIDS Res Hum Retroviruses 2007;23(12): 1555–62. [21] Hawkridge T, Scriba T, Gelderbloem S, Smit E, Tameris M, Moyo S, et al. Safety and immunogenicity of a new tuberculosis vaccine, MVA85A, in healthy adults in South Africa. J Infect Dis 2008;198(4):544–52. [22] Burgers WA, Chege G, Müller TL, van Harmelen JH, Khoury G, Shephard EG, et al. Broad, high-magnitude and multifunctional CD4+ and CD8+ T-cell responses elicited by a DNA and modified vaccinia Ankara vaccine containing human immunodeficiency virus type 1 subtype C genes in baboons. J Gen Virol 2009;90(Pt 2):468–80. [23] Darrah PA, Patel D, De Luca PM, Lindsay RW, Davey DF, Flynn BJ, et al. Multifunctional TH1 cells define a correlate of vaccine-mediated protection against Leishmania major. Nat Med 2007;13(7):843–50. [24] Mollenkopf HJ, Hahnke K, Kaufmann SH. Transcriptional responses in mouse lungs induced by vaccination with Mycobacterium bovis BCG and infection with Mycobacterium tuberculosis. Microbes Infect 2006;8(1):136–44. [25] Sinsimer D, Huet G, Manca C, Tsenova L, Koo MS, Kurepina N, et al. The phenolic glycolipid of Mycobacterium tuberculosis differentially modulates the early host cytokine response but does not in itself confer hypervirulence. Infect Immun 2008;76(7):3027–36. [26] Sada-Ovalle I, Chiba A, Gonzales A, Brenner MB, Behar SM. Innate invariant NKT cells recognize Mycobacterium tuberculosis-infected macrophages, produce interferon-gamma, and kill intracellular bacteria. PLoS Pathog 2008;4(12):e1000239.