Anti-tumoral effect of the mitochondrial target domain of Noxa delivered by an engineered Salmonella typhimurium

Abstract

Bacterial cancer therapy relies on the fact that several bacterial species are capable of targeting tumor tissue and that bacteria can be genetically engineered to selectively deliver therapeutic proteins of interest to the targeted tumors. However, the challenge of bacterial cancer therapy is the release of the therapeutic proteins from the bacteria and entry of the proteins into tumor cells. This study employed an attenuated Salmonella typhimurium to selectively deliver the mitochondrial targeting domain of Noxa (MTD) as a potential therapeutic cargo protein, and examined its anti-cancer effect. To release MTD from the bacteria, a novel bacterial lysis system of phage origin was deployed. To facilitate the entry of MTD into the tumor cells, the MTD was fused to DS4.3, a novel cell-penetrating peptide (CPP) derived from a voltage-gated potassium channel (Kv2.1). The gene encoding DS4.3-MTD and the phage lysis genes were placed under the control of PBAD , a promoter activated by L-arabinose. We demonstrated that DS4.3-MTD chimeric molecules expressed by the Salmonellae were anti-tumoral in cultured tumor cells and in mice with CT26 colon carcinoma.

Figure 1. Schematic diagram of the plasmids used in this study.

‘Lys’ carried the locus with Salmoella lysis function of the phage iEPS5 phage.

Figure 2. Bacterial lysis phenotype of pLYS.

(A) S. typhimurium carrying pLYS were grown in LB supplemented with L-arabinose. When the optical density (A₆₀₀) of the culture reached ∼1.0, at T = 2 hrs, the LB was supplemented with L-arabinose in half of the cultures. The A₆₀₀ of the culture was measured at regular time intervals (circles) and viable bacteria were determined by CFU counting (bars). Open circles and bars represent bacteria samples without L-arabinose addition, and filled circles and bars represent bacterial samples with L-arabinose addition. (B) S. typhimurium carrying either pLYSP_BAD::lacZ or pBAD24P_BAD::lacZ grown in LB was supplemented with L-arabinose when the A₆₀₀ of the culture reached ∼1.0. The bacterial culture was taken 2 hrs after the addition of L-arabinose, separated into bacterial pellet and supernatant by centrifugation and filtration, and assayed for total β-galactosidase activity. Data represent mean ± S.D., and asterisks (*) indicate a significant difference compared between supernatant of pLYSP_BAD::lacZ and pBAD24P_BAD::lacZ (***, P<0.0001). (C) The same samples were analyzed for β-galactosidase expression by Western blot analysis. 50 µg of sonicated samples were analyzed. Bacterial whole culture not induced for bacterial lysis was included.

Figure 3. Monitoring of L-arabinose-induced bacterial lysis by a non-invasive in vivo imaging system (IVIS) and release of β-galactosidase from Salmonellae carrying pLYSP_BAD::lacZ or pBAD24P_BAD::lacZ targeted to implanted tumor tissue (CT26) in BalB/c mice.

Each time point represents the mean for n = 3 animals. (A) Bioluminescent bacteria (SKS003) carrying pLYS were injected intravenously into CT26 tumor-bearing mice. After three days, the bioluminescence signal from the Salmonellae was monitored by a non-invasive in vivo imaging system (IVIS) at the indicated times after intraperitoneal injection of L-arabinose (60 mg) in the induced group. The Y-axis indicates photons ×10^5·s^−1·cm^−2·sr⁻¹. The ROI was selected manually for quantification of photon flux. The ROI area was kept constant, and the intensity was recorded as a maximum (photons ×10^5·s^−1·cm^−2·sr⁻¹) within the ROI. (B) Photons from each mouse at 0 time was calculated 100% and compared. Data represent mean ± S.D., and asterisks (*) indicate a significant difference compared between uninduced and induced group (***, P<0.0001). (C) The number of live bacteria in the tumor tissue 8 hrs after the addition of L-arabinose was determined by CFU counting. (D) The tumor tissues were analyzed for the expression and distribution of β-galactosidase by histochemical staining with X-gal.

Figure 4. Induction of cell death by MTD fused to CPPs.

(A) Amino acid sequences of chimeric CPP-MTDs used in this study. The CPP-MTDs were arranged as follows: N′ CPP followed by a GGG linker, KLLNLISKLF (the MTD sequence), and finally the CSGT of the C-terminal end of Noxa . (B) Induction of cell death by CPP-MTD peptides as assessed by MTT assay. In vitro cultured HeLa, Hep3B, and CT26 cells were tested. (C) The expression and release of DS4.3-MTD from the Salmonellae carrying pLYSP_BAD::DS4.3-MTD at indicated time after L-arabinose addition, as analyzed by Western blot using anti-Noxa antibody. The samples were prepared as described in the legend of Figure 2. (D) Induction of cell death by the bacterial supernatant of S. typhimurium carrying pLYSP_BAD::DS4.3-MTD after L-arabinose addition. Bacterial supernatant of S. typhimurium carrying pLYS after L-arabinose addition was included. Data represent mean ± S.D., and asterisks (*) indicate a significant difference compared between pLYS and pLYSP_BAD::DS4.3-MTD (*, P<0.05).

Figure 5. Anti-tumoral effects of Salmonellae carrying either pLYSP_BAD::DS4.3-MTD or pLYS, Three days after the bacterial treatment, L-arabinose (60 mg) was intraperitoneally injected daily (n = 3 per group).

(A) Changes in tumor size after the bacterial treatment. Data represent mean ± S.D., and asterisks (*) indicate a significant difference compared between each groups with two-way ANOVA (***, P<0.0001) (B) A representative morphological change of the implanted CT26 tumor after the bacterial treatment. (C) Immuno-fluorescence examination of CT26 tumor tissue after bacterial treatment. Frozen section was obtained from the tumor tissue and stained with Noxa specific antibody (SantaCruz, sc-30209). The boxes on left show nucleus with DAPI in blue, Noxa MTD with specific antibody in green, and cytosol with TexasRed Phalloidin in red, in order from the top. The boxes on the right show merged images. (D) Histopathological examination of CT26 tumor tissue after the bacterial treatment. Tumor tissues were sliced and stained with hematoxylin-eosin (HE). Purple area represents the region with healthy cells and whitish pink the region with anucleus dead cells, the necrotic region.