Stroma Affects the Intratumoral Transport of Bacteria-Based Cancer Therapy

Introduction:: In the U.S., cancer is the second leading cause of death, responsible for ~600,000 deaths annually. The tumor microenvironment, comprised of stromal cells, including cancer-associated fibroblasts (CAFs) and a dense extracellular matrix (ECM), profoundly affects the efficacy of anticancer therapeutics. CAFs play a crucial role in tumor growth, progression, and metastasis in part by producing the physical scaffold of the tumor microenvironment - the ECM^[1]. Excessive expression of collagen, a primary component of ECM, in tumors promotes higher interstitial pressure and solid stress, collapses blood vessels, and increases hypoxia, limiting the penetration of anticancer therapeutics^[2]. In this work, we investigated bacteria-based drug delivery systems are a potential strategy to overcome the barriers to intratumoral transport. We and others have previously shown that attenuated tumor-targeting bacteria selectively colonize and proliferate in cancerous tissue^[3]. However, the effect of ECM content on the distribution of bacteria-based drug delivery agents remains largely unexplored. We hypothesize that self-propulsion in motile bacteria can be harnessed to reach currently inaccessible hypovascular tumor regions that are beyond the diffusion limits. Here, we use a breast cancer tumor organoid model comprising mammary CAFs(mCAFs) and epithelial breast cancer cells to assess the role of ECM production by mCAFs in the intratumoral transport and colonization of bacteria.


Materials and Methods:: The attenuated and tumor-targeting Salmonella Typhimurium VNP20009 cheY⁺ RFP^[3] was cultured in MSB medium^[4] supplemented with 35 μg/mL chloramphenicol at 37 °C and 100 rpm. Bacterial culture was harvested at OD₆₀₀ of 1.0, washed twice at 1700 ×g for 5 min, and resuspended in the CO₂-independent L-15 medium supplemented with 10% calf serum at the same OD₆₀₀. Multicellular Tumor spheroids were formed in ultra-low adhesion 96 well plates using 4T1 murine mammary carcinoma cells with or without mCAFs, as specified. Once tumor diameters reached 600 µm, up to 6 tumor spheroids were transferred into Eppendorf tubes and incubated with the bacterial suspension on an end-over-end shaker at 37 ºC to enable initial bacteria entrapment. After an hour, tumor spheroids were incubated in 2.5 µg/mL gentamicin sulfate in culture media for 30 min and washed twice with fresh culture media to eliminate planktonic bacteria. After 12, 18, 24, 36, and 48 hr incubation periods, the tumor spheroids were snap frozen, sliced, and stained to evaluate the spatial distribution of bacteria and tumor cell viability. A minimum of 3 independent experiments and 3 tumors for each experiment category were used. To quantitatively compare the effect of ECM on bacteria intratumoral transport, we used our previously developed metrics^[3] of (i) Penetration index (PI), a measure of bacterial intratumoral penetration distance; (ii) Colonization index (CI), a measure of bacteria density within a given tumor slice volume, and (iii) Distribution index (DI), which represents a combination of penetration distance and colonization density.


Results, Conclusions, and Discussions:: The 4T1 mono-culture and the 4T1/mCAF co-culture spheroids were used to investigate the influence of ECM content on bacterial colonization and penetration. Bacterial penetration in co-culture spheroids was markedly more limited than that of mono-culture spheroids at all time points (Fig.1A). In fact, at 48 hr, mono-culture tumor spheroids completely disintegrated. In contrast, the co-culture tumors were only colonized near the outer surface (Fig.1A). The significantly higher collagen content in co-culture spheroids (Fig.1B) maintains tumor integrity and limits bacterial penetration to < 50% of radial distance (Fig .1C-D).


The DI of mono-culture spheroids is significantly higher than that of co-culture spheroids at each incubation time (Fig.1E). In the mono-culture spheroid, a 3.8-fold increase in DI was observed during the 36 hr incubation with bacteria. However, in co-culture spheroids, there was no statistically significant change in DI. To further elucidate the role of ECM in the observed outcomes, we investigated PI and CI separately. There was no significant increase in PI with increasing incubation time for mono-culture and co-culture (Fig.1F). The PI value for mono-culture was about twice the PI value for co-culture. Differences in the PI trends suggest that the penetration difference might be due to dissimilar microenvironment properties caused by the presence of mCAFs (Fig.1B). Unlike the PI trends, CI of mono-culture and co-culture increased significantly, with increasing incubation time (Fig.1G). The mono-culture had a higher CI value than the co-culture at all incubation times. This can also be attributed to the dense microenvironment of the co-culture spheroids, which hinders bacterial growth. Lastly, we evaluated tumor cell viability using TUNEL staining. As seen in Fig.1H-I, the number of TUNEL-positive cells increased with incubation time. Notably, the number of TUNEL-positive cells was significantly higher at all radial locations after 48 h.
In summary, our results show that bacteria penetrated deeper and proliferated more in mono-culture tumors. The reduced performance of bacteria in co-culture spheroids is associated with higher ECM content. Many studies show the crucial role of dense ECM in tumor microenvironment-mediated drug resistance^[5–7]. In the future, we will develop ECM-targeting bacteria to overcome the barrier to bacteria intratumoral penetration.


Acknowledgements (Optional): :

References (Optional): : [1] T. Alkasalias, L. Moyano-Galceran, M. Arsenian-Henriksson, K. Lehti, Int. J. Mol. Sci. 2018, 19, 1532.

[2] R. K. Jain, Cancer Res. 1987, 47, 3039.

[3] S. Suh, A. Jo, M. A. Traore, Y. Zhan, S. L. Coutermarsh‐Ott, V. M. Ringel‐Scaia, I. C. Allen, R. M. Davis, B. Behkam, Adv. Sci. 2019, 6, 1801309.

[4] K. M. Broadway, S. Suh, B. Behkam, B. E. Scharf, J. Biotechnol. 2017, 251, 76.

[5] M. Majety, L. P. Pradel, M. Gies, C. H. Ries, PLoS One 2015, 10, e0127948.

[6] K. H. T. Paraiso, K. S. M. Smalley, Fibroblast-mediated drug resistance in cancer, Vol. 85, Elsevier, 2013, pp. 1033–1041.

[7] S. Y. Jeong, J. H. Lee, Y. Shin, S. Chung, H. J. Kuh, PLoS One 2016, 11.

[2] R. K. Jain, Cancer Res. 1987, 47, 3039.

[3] S. Wilhelm, A. J. Tavares, Q. Dai, S. Ohta, J. Audet, H. F. Dvorak, W. C. W. Chan, Nat. Rev. Mater. 2016, 1, 16014.

[4] S. Suh, A. Jo, M. A. Traore, Y. Zhan, S. L. Coutermarsh‐Ott, V. M. Ringel‐Scaia, I. C. Allen, R. M. Davis, B. Behkam, Adv. Sci. 2019, 6, 1801309.

[5] K. M. Broadway, S. Suh, B. Behkam, B. E. Scharf, J. Biotechnol. 2017, 251, 76.

[6] A. Ivascu, M. Kubbies, J. Biomol. Screen. 2006, 11, 922.

[7] M. Majety, L. P. Pradel, M. Gies, C. H. Ries, PLoS One 2015, 10, e0127948.

[8] K. H. T. Paraiso, K. S. M. Smalley, Fibroblast-mediated drug resistance in cancer, Vol. 85, Elsevier, 2013, pp. 1033–1041.

[9] S. Y. Jeong, J. H. Lee, Y. Shin, S. Chung, H. J. Kuh, PLoS One 2016, 11.