# RTB 2

> **NIH NIH U54** · JOHNS HOPKINS UNIVERSITY · 2022 · $345,016

## Abstract

Metastasis requires fundamental changes in cell behavior and causes most cancer deaths. Metastasis is also
an inherently 3D process involving interactions among diverse cancer cells and with the tumor microenvironment
(TME). We developed innovative 3D assays to model each step in metastasis ex vivo. We use these assays to
generate hypotheses about how cancer cells accomplish metastasis and which molecular signals should be
targeted therapeutically. In vivo validation of these hypotheses is rate limiting, technically and conceptually. We
can compare the effects of many perturbations in vitro, with real-time imaging and molecular readouts. In
contrast, in vivo validation is generally limited to measurements of tumor diameter, CTC and metastasis numbers,
and a few molecular markers in 2D sections. There is an urgent need to achieve a 3D understanding of
metastasis, including the complex interactions among cell types and transitions between cell states. The
3D imaging and spatial multi-omics approaches in TECH1 and TECH2 are ideally suited to allow us to
understand vascular invasion, the key transition from local to metastatic disease. Prior studies generally
evaluated single cell types or a few markers, largely in 2D. CODA (TECH1) will enable us to classify cell types
and their spatial relationships in 3D. DBiT-seq (TECH2) enables us to reconstruct the transcriptome and select
proteome of high-resolution regions (~10 micron) across whole sections of human tumors. We will combine these
techniques to achieve spatial multi-omics and resolve cancer cell state changes during breast cancer metastasis.
Aim 1: Adapt CODA to murine models and human breast tumors, focusing on venous invasion. We will
first supply archival human breast tumors to enable TECH to adapt their 3D deep learning algorithms to breast
cancer. We will start with a existing series of 250 human breast tumors with digitized serial sections. We will then
collect, fix, and section fresh human breast tumor samples, stained with immune and cancer cell markers. We
will use CODA to reconstruct the 3D architecture of vascular invasion and associated stromal responses. We
will also adapt CODA techniques for use with murine preclinical models. We will then leverage these insights to
reconstitute the vascular invasion niche in vitro by adapting a novel microfluidic platform we developed.
Aim 2: Adapt DBiT-seq for murine and human breast tumors, focusing on cancer cell state transitions.
We will adapt DBiT-seq to 3D human breast tumor samples to understand spatial relationships among cancer
cell states during vascular invasion. This analysis will be led from cell states and inferred state transitions we
defined in vitro using single cell RNA-seq in our 3D metastasis assays. We will then collect a staged series of
tumors and distant organs from GEMMs to define cell state transitions spatially across metastatic processes that
are difficult to sample in humans. We will then use the transcriptional and s...

## Key facts

- **NIH application ID:** 10375195
- **Project number:** 1U54CA268083-01
- **Recipient organization:** JOHNS HOPKINS UNIVERSITY
- **Principal Investigator:** Andrew Josef Ewald
- **Activity code:** U54 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2022
- **Award amount:** $345,016
- **Award type:** 1
- **Project period:** 2021-12-01 → 2026-11-30

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10375195

## Citation

> US National Institutes of Health, RePORTER application 10375195, RTB 2 (1U54CA268083-01). Retrieved via AI Analytics 2026-05-27 from https://api.ai-analytics.org/grant/nih/10375195. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*