The Pioneer platform: A novel approach for selection of selective anti-cancer cytotoxic activity in bacteria through co-culturing with engineered human cells

Abstract

Most of the small-molecule drugs approved for the treatment of cancer over the past 40 years are based on natural compounds. Bacteria provide an extensive reservoir for the development of further anti-cancer therapeutics to meet the challenges posed by the diversity of these malignant diseases. While identifying cytotoxic compounds is often easy, achieving selective targeting of cancer cells is challenging. Here we describe a novel experimental approach (the Pioneer platform) for the identification and development of 'pioneering' bacterial variants that either show or are conduced to exhibit selective contact-independent anti-cancer cytotoxic activities. We engineered human cancer cells to secrete Colicin M that repress the growth of the bacterium Escherichia coli, while immortalised non-transformed cells were engineered to express Chloramphenicol Acetyltransferase capable of relieving the bacteriostatic effect of Chloramphenicol. Through co-culturing of E. coli with these two engineered human cell lines, we show bacterial outgrowth of DH5α E. coli is constrained by the combination of negative and positive selection pressures. This result supports the potential for this approach to screen or adaptively evolve 'pioneering' bacterial variants that can selectively eliminate the cancer cell population. Overall, the Pioneer platform demonstrates potential utility for drug discovery through multi-partner experimental evolution.

Fig 1. Principle of the Pioneer platform.

In the Pioneer platform, E. coli bacteria (blue) are co-cultured in soft agar containing Chloramphenicol over a mixed monolayer of human cancer cells which are labelled with ZsGreen1 fluorescence and secrete Colicin M (green), and human immortalised non-transformed cells which are labelled with mCherry fluorescence and express CAT (red) which detoxifies Chloramphenicol. Bacterial growth is repressed by Colicin M so that there is no outgrowth of bacterial colonies and culture media is not discoloured (A). However, ‘pioneering’ bacteria which exhibit or can adapt to acquire contact-independent selective cytotoxic activity against the cancer cells (represented by dashed lines and hypothetical bioactive compound H) can eliminate the source of Colicin M without affecting the production of CAT from the immortalised non-transformed cells which relieves the negative selection pressure imposed by Chloramphenicol, and bacterial outgrowth can occur leading to a detectable colour change of the culture media (B).

Fig 2. Detection and colicinogenic activity of His-tagged Colicin M in conditioned media from transduced DLD-1 Cma-ZsG1 cells.

(A) Proteins in conditioned media from DLD-1 and DLD-1 Cma-ZsG1 cells cultured for 24 hours were separated by SDS–PAGE and visualised by Western blot with the indicated antibody. (B) Relative bacterial growth of DH5α E. coli grown in liquid culture in the presence of 50% non-conditioned media (grey circles) or conditioned media from DLD-1 (blue squares) or DLD-1 Cma-ZsG1 cells (green diamonds). Error bars represent means ± standard deviations; n = 3 independent experiments; * P < 0.05 and ** P < 0.01 by two-way ANOVA Tukey’s Multiple Comparisons Test compared with the non-conditioned media control (coloured asterisks) or comparing DLD-1 and DLD-1 Cma-ZsG1 samples (black asterisks).

Fig 3. Expression of CAT in HCEC 1CT Cat-mCh cell lysates and detoxification of Chloramphenicol in conditioned media from transduced HCEC 1CT Cat-mCh cells.

(A) Proteins from cell extracts of HCEC 1CT and HCEC 1CT Cat-mCh cells cultured with (+Camp) or without (-Camp) 6.8 μg/ml Chloramphenicol for 24 hours were separated by SDS–PAGE and visualised by Western blot with the indicated antibody. (B) Relative bacterial growth of DH5α E. coli grown in liquid culture in the presence of 50% non-conditioned media (grey circles) or conditioned media from HCEC 1CT (blue squares) and HCEC 1CT Cat-mCh (orange diamonds) cells -/+Camp. Error bars represent means ± standard deviations; n = 3 independent experiments; * P < 0.05, ** P < 0.01 and **** P < 0.0001 by two-way ANOVA Tukey’s Multiple Comparisons Test compared with non-conditioned media control (coloured asterisks) or comparing HCEC 1CT +Camp and HCEC 1CT Cat-mCh +Camp samples (black asterisks).

Fig 4. Co-culture of E. coli with DLD-1 Cma-ZsG1 cells represses bacterial growth but co-culture with HCEC 1CT Cat-mCh cells promotes bacterial growth in the presence of Chloramphenicol.

Growth of DH5α E. coli colonies co-cultured in soft agar above a monolayer of human DLD-1 and DLD-1 Cma-ZsG1 cells for 3 days (A and C) or with 3.4 μg/ml Chloramphenicol above a monolayer of human HCEC 1CT or HCEC 1CT Cat-mCh cells for 2 days (B and D). Representative images from independent replicative experiments are shown (A and B). Scatter plots show error bars representing means ± standard deviations; n = 3 independent experiments; * P < 0.05 and ** P < 0.01 by unpaired student’s t-test assuming unequal variances (C and D).

Fig 5. Bacterial outgrowth is repressed by a mixed monolayer of DLD-1 Cma-ZsG1 and HCEC 1CT Cat-mCh cells in the Pioneer platform.

Growth of DH5α E. coli bacterial colonies co-cultured in soft agar with 3.4 μg/ml Chloramphenicol without human cells (row A), or above a monolayer of DLD-1 Cma-ZsG1 cells (row B) or HCEC 1CT Cat-mCh cells (row C) or a mixed monolayer of both (rows D-H) for 3 days. Empty wells (A1-A3) were used for background subtraction. A representative image from independent replicative experiments showing culture media discolouration and bacterial colony formation is shown (A), and data are represented in a bar chart showing error bars representing means ± standard deviations (B); n = 2 independent experiments.