Project Summary
It is increasingly clear that bacteria play an important role in human health. While it is natural to focus on how
intestinal bacteria affect disease, intriguing findings have elucidated the extent to which bacteria inhabit solid
tumors. Microbes have been detected in lung, pancreatic, breast, oral, gallbladder, ovarian, liver, and colorectal
cancers. Localization has been ascribed to several mechanisms, including preference for anaerobic or facultative
anaerobic bacteria to grow in the hypoxic core of tumors, presence of bacterial nutrients, lack of immune surveil-
lance, and leakiness of the often poorly structured vasculature surrounding neoplastic tissue. This tendency for
localization to solid tumors suggests that bacteria could be engineered for precise and robust drug production and
delivery from within the solid tumor environment. This dovetails with 20 years of progress in synthetic biology,
which has tended to focus on microbial engineering. However, information on how the tumor microenvironment
affects bacterial growth is largely unknown. The microenvironment will affect bacterial gene expression that ul-
timately underlies the functionality of engineered therapies, and it is difficult to imagine a predictive framework
for engineered bacterial therapies without a quantitative understanding of how bacteria react to the environment
of a growing tumor. We will use a probiotic strain of E. coli with an established safety record to develop a novel
class of biosensors to noninvasively investigate bacterial growth in the tumor microenvironment. Initially, we
will develop lysis-based biosensors that respond to specific tumor environment targets: hypoxia, pH, glucose,
and lactate (Aim 1). We will also engineer an inducible quorum sensing (QS) system that enables external control
of bacterial population dynamics, including the ability to eliminate a specific strain whenever desired (Aim 1).
Together these strains will allow us to modulate and monitor population dynamics in vivo, enabling both sens-
ing of the local environment and maintenance of an external control switch. We will test these strains using an
established in vitro organoid model (Aim 2) and in two clinically relevant animal models for solid tumor growth.
Additionally, we will use our previously developed dynOMICS technology to screen tumor extract from the two
animal models and construct a second suite of biosensors for monitoring the tumor environment (Aim 2). These
biosensors will then be tested in the animal models. We will visualize bacterial populations in a colorectal tumor
model with bacteria that are engineered to produce luciferase in order to monitor colony dynamics using our es-
tablished methods (Aim 3). We will also build on recently reported technology whereby bacteria are modified for
use with ultrasound through addition of gas vesicles that permit high resolution imaging of the engineered bac-
teria. We will use the ultrasound method to investigate NASH-induced hepatocellular carcinoma (HCC) where
a high-fat diet is used to induce HCC at 20 weeks in mice (Aim 4). This project will quantitatively characterize
how bacterial strains sense, respond, and grow in the tumors. The results will establish a platform for future
exploration of therapies that are produced and delivered by bacteria that grow within solid tumors.