Gajalakshmi Dakshinamoorthy, S. Mistry, J. Kim, N. Nikulina, R. Bashier, M. Gallina, C. Hempel, S. Basak, W.M. Lee, J. Kennedy-DarlingAkoya Biosciences, Department of Research and Development, Menlo Park, CA

INTRODUCTION

Schematic of cyclical workflow. Tissues are stained with the full panel of CODEX antibodies in a single step. Iterativecycles of labelling, imaging and removing reporters are performed via a fully automated fluidics system, until allbiomarkers of interest are imaged. Images are collected and compiled across cycles to achieve single-cell resolutiondata.

CODEX: CO-Detection by indEXing

Human FFPE Antibody panel

INSTRUMENT / MICROSCOPE INTEGRATION

Seamless microscope integration. The CODEX fluidics device integrates into microscope stages through a customstage insert. The CODEX Driver Software is compatible with multiple microscope brands/types, including Keyence BZ-X710/800, Leica DMi8, & Zeiss Axio-Observer.

DEEP MULTIPARAMETRIC ANALYSIS OF METASTATIC LYMPH NODE – BREAST CANCER (FF)

Human FFPE antibody panel: Human FFPE tonsil tissue was stained with 27 markers in a single step and revealed via afull automated fluidics workflow.

Data analysis of Metastatic lymph node breast cancer. Segmented cells were clustered using X- shift clustering based on nearest neighborestimation and the cell types were annotated by phenotype marker expression.

The tumor microenvironment is comprised of numerous infiltrating immunecells, proliferating tumor cells and a variety of associated tissue cells.Immune evasion of tumor cells are collectively contributed by these cells.Hence, the spatial distribution of the associated cell types and tissuefeatures is a critical parameter of tumor biology. Current technologies arelimited to either measuring the spatial parameter or measuring high-parameter space but are not able to measure both in the same sample. Herewe present a novel multiparametric imaging technology, termed CODEX(CO-Detection by IndEXing) , a high-throughput technology to detect morethan 50 biomarkers simultaneously in a single tissue, and resolve theirrelative expression, abundance at the spatial level.

Validated & Custom Panel Designs

Simplified Microscope Integration

Automated Assay and Imaging

Image and Cytometric Analysis

CODEX Assay• Barcoded Antibodies• CODEX Reporters• Custom Barcodes

CODEX SoftwareFluidics ControlCODEX Reagents

Basic Chemistry: During each imaging cycle, three CODEX Tm Reporters are assayed to their corresponding barcodes conjugated to antibodies. Fluorescence signal captured for each biomarkers.

CD4 CD8 CD68 Ki67 Collagen IV E-cadherin Podoplanin

Helper T cells

NKs

T follicular helper cells

B cells

DCs interacting with T cells

DCs near epithelium

DCs near B cells

DCs (CD205+, vimentin+)T cells interacting

with DCs

Epithelial cells (CD63/CD9/Pan-cytokeratin)

Epithelial cells (63/9(lo))

Epithelial cells (CD9/vimentin/Pan-cytokeratin)

Epithelial cells (CD63/Pan-cytokeratin)

Combined T cell population

Plasma cells

Plasma cells near epithelial cells

Vascular endothelial cells (vWF/CD146)

Vascular endothelial cells (CD146/vWF/CD36)

Vascular endothelial cells (CD146/CD36)

DCs (CD205+)

Lymphatic vasculature

FDCs

Malignant cellsECMCAFsVascular endothelial cellsLymphatic endothelial cellsHelper T cellsKiller T cellsB cellsNK cellsDC cellsMacrophagesRed blood cellsAdipocytesMesenchymal stem cells

Custom insert holds the tissue specimen and integrates with microscope stage

Tissues are sliced onto coverslips and

stained off-line

CODEX-Cyclic workflow

CD8 (green) CD44 (green) CD45 (green) Actin CD3e CD107a CD4 CD15 (green) CD20

Ki67 Vimentin HLA-DR IDO1 CD57 CD279 CD11cPan-cytokeratin Podoplanin

CD14 CD21 CD31 CD34 CD45RO CD66c CD68 E-cadherin Histone H3

Human Renal Cell Carcinoma CODEX Tissue Analysis

FFPE-Human Breast cancer tissue analysis

Annotated cell clustersCell clusteringCell segmentation

Mapping tissue architecture

Region 3Region 2Region 1

Region 3

DAPI

E-cadherin Pancytokeratin CD31 Ki67 Beta-catenin S100a

LIF TFAM Vimentin Runx3 Keratin14 CDKN1a

/]’

Region 1Region 2

Region 3

Region1 Region2 Region3

• CODEXTm enables multiplexed, spatial analysis of tissue specimens in a fully automated workflow.• CODEXTm is compatible with a variety of tissue specimens, including FF and FFPE formats.• CODEXTm data can be analyzed using the CODEX analysis tools to characterize cell type, map the tissue architecture and cellular niches

Characterizing Human FFPE Renal Carcinoma: Human FFPE renal carcinoma was stained with 7 markers and imagedusing CODEX Tm workflow. Interactions of infiltrating immune cells (CD3 and CD8) with proliferating tumor cells (Ki67),epithelial cells (Podoplanin), and cells that may have involved in tumor invasion (E-cadherin) was characterized.

Breast carcinoma tissue analysis : Breast carcinoma (Stage IIIB: T4d N2a MX with 90% T/ 5% TCS/ 5% N) wasstained with different CODEX markers and analyzed using CODEXTM workflow. Image analysis revealed critical markersthat are involved in tumorigenesis (LIF, TFAM), Tumor invasion (E-cadherin ), Proliferation (Ki67), Prognostic marker forbreast carcinoma (Beta-catenin) and cell survival (CDKN1a).

Breast carcinoma tissue analysis : Breast carcinoma (Stage IIB: T3 NX MX with 90% T/ 20% N) was stained withdifferent CODEXTM markers and analyzed using CODEXTM workflow. Image analysis revealed critical markers that areinvolved in tumorigenesis (LIF, TFAM) and Tumor invasion (E -cadherin ).

LIF TFAM Vimentin

Tiss

ue 1

Tiss

ue 2

E-cadherin Pancytokeratin CD31More interacting

Less interacting

Tumor Microenvironment

Biological interactions of variety of cell types at the single cell level was determined..

Conclusion

Highly multiplexed single-cell spatial analysis of tissue specimens using CODEX®

Spatial correlations of multiple cell types were compared among different regions