Expression of Mage-b_311-660 and Listerial proteins by 4T1 tumor cells. 4T1 tumor cells were incubated with LM-LLO or LM-LLO-Mage-b_311-660 for 2 hours, and subsequently analyzed for the expression of Mage-b_311-660 (fused with LLO) (88 kDa) and Listeria (LLO)(58 kDa) proteins by western blotting using anti-PEST antibodies (A). Spleen cells (pooled) of vaccinated (preventive/therapeutic) and control mice bearing 4T1 metastases and primary tumors were analyzed for Mage-b- and Listeria-specific immune responses by ELISPOT using BM cells (transfected with pcDNA3.1-Mage-b and pCMV1-GM-CSF) (B), and BM cells infected with LM-LLO (C), respectively. The involvement of CD8 T cells was determined by negative depletion, using magnetic beads with anti-CD8, antibodies. The results shown here are the average of two independent experiments. n=5 mice per group. The error bars represent the SEM. Significant differences between Mage-b and other groups were determined. Mann-Whitney test: p<0.05 is significant.

Effect of LM-LLO and LM-LLO-Mage-b_311-660 in vivo and in vitro. The frequency of metastases and tumor weight was determined after three preventive immunizations (A) and after a combination of one preventive and two therapeutic immunizations (B). Also Mage-b-specific immune responses (C) were analyzed after three preventive immunizations (i) and after the combination of one preventive and two therapeutic immunizations (ii) by ELISPOT. For this purpose, spleen cells were pooled per group and restimulated with bone marrow cells (transfected with pcDNA3.1-Mage-b and pCMV1-GM-CSF), and analyzed for the production of IFNγ 72 hrs later. The involvement of CD8 T cells was determined by negative depletion, using magnetic beads with anti-CD8, antibodies. Controls such as BM cells transfected with pcDNA3.1-Mage-b or pCMV1-GM-CSF, or non-transfected BM cells did not produce IFNγ (data not shown). The results shown in this figure are the average of two-three independent experiments. n=5 mice per group. The error bars in all graphs represent the standard error of the mean (SEM). Significant differences between Mage-b and other groups were analyzed using Mann-Whitney test. p<0.05 is significant.

Listeria infects and kills tumor cells. A, Both LM-LLO and LM-LLO-Mage-b_311-660, infected 4T1 and MCF7 with high efficiency after one hour of infection. Experiments were performed in triplicates, and repeated three times. The results presented in this figure are the average of three infection experiments. The error bars represent the SEM. B, Both LM-LLO and LM-LLO-mage-b_311-660 killed 4T1 and MCF7 with the same efficiency (Unpaired t test p>0.05). In this experiment, 4T1 tumor cells were incubated with the Listeria bacteria for 3 hours and then treated with gentamycin. Next day, dead and alive tumor cells were determined with trypan blue. From each sample, nine fields were blinded analyzed by two different investigators. Experiments were performed in triplicates, and repeated three times. The results presented in this figure are the average of three experiments. The error bars represent SEM. C, EM analysis shows the Listeria bacteria (LM-LLO-Mage-b_311-660) inside 4T1 and MCF7 tumor cells after one-hour infection in vitro. Bar=200μ. The Listeria bacteria also infected 4T1 primary tumor and metastases in vivo with high efficiency, as shown here after one immunization with LM-LLO-mage-b_311-660. Bar=400μ (top), Bar=50μ (bottom). In this experiment tumors and metastases of three different mice that received LM-LLO-Mage-b_311-660, were analyzed. The photographs shown here are representative for the three different mice analyzed. The Listeria bacteria are marked with white arrows.

Mechanism(s) of tumor cell kill by LM-LLO-Mage-b_311-660 and LM-LLO. A, Listeria-induced tumor cell death by activation of NADPH oxidase and production of ROS. 4T1 tumor cells were incubated with LM-LLO or LM-LLO-Mage-b_311-660 for two hours, and analyzed for activation of NADPH oxidase using antibodies to p47phox (i). NADPH oxidase-induced cell death was prevented by apocynin, while ROS-induced cell death was prevented by trolox and DPI (ii). Significant differences were found between 4T1 tumor cells treated with inhibitors compared to untreated 4T1. Unpaired t test: P<0.05 is significant. A similar result was observed with MCF7. B, Prevention of Listeria-induced ROS with apocynin. 4T1 tumor cells were incubated with LM-LLO, in the presence or absence of apocynin, and showed decreased production of ROS with H₂DFFA (cROS and mROS), or CM-H₂XRos (mROS). C, LM-LLO also induced increase in intracellular Ca²⁺ levels in 4T1 tumor cells as shown with Ca²⁺-green-1 (i). LM-LLO-induced mitochondrial disruption (ii) in 4T1 tumor cells and subsequent tumor cell death (iii), could be prevented with apocynin an/or BAPTA. While both apocynin and BAPTA were able to prevent tumor cell death by 50% when added separately, 80–90% of tumor cell death could be prevented when added combined to infected tumor cell cultures (iii). Significant differences were found between 4T1 tumor cells treated with inhibitors compared to untreated 4T1. Unpaired t test: p<0.05 is significant. D, The PEST motif in LLO of the Listeria bacteria activates NADPH oxidase and is involved in tumor cell death. This could be prevented with apocynin. LLO protein (LLO) with or without the PEST motif (LLOΔPEST) was added to 4T1 cultured, and cell death was determined the next day. Significant differences between LLO protein and other groups were analyzed. Unpaired t test: p<0.05 is significant. Experiments were performed in triplicates, and repeated two times. The results presented here are the average (Aii, Ciii, D) of or a representative (Ai, B, Ci, Cii) of two experiments performed. The error bars represent SEM.