License: This is an open access protocol distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Protocol status: Working
We use this protocol and it's working
Created: December 12, 2023
Last Modified: May 14, 2024
Protocol Integer ID: 92220
Keywords: NanoString, GeoMx, Digital Spatial Profiler, HTAN, TMA, DSP, cell profiling, cell type
Funders Acknowledgement:
Human Tumor Atlas Network
Grant ID: NIH NCI U2C CA233280
Abstract
This protocol outlines the NanoString GeoMx Digital Spatial Profiler Whole Transcriptome Atlas (DSP WTA) assay that was applied in the Human Tumor Atlas Network (HTAN) Tissue MicroArrary (TMA) -TransNetwork Project (TNP).
The TMA-TNP evaluates various characterization and analytics methodologies on a large array of breast tumor samples representing a broad spectrum of disease state and subtype. A commercially available anonymized breast tumor TMA was purchased and serial sections were distributed. Participating HTAN Centers characterized the FFPE specimens using various imaging platforms and generated a spatially resolved cell type/state census using each center’s method of choice. Data was recorded in a common repository to enable joint analysis.
The protocol that immediately precedes this one for TMA-TNP Phase 4 can be found at: dx.doi.org/10.17504/protocols.io.ewov1o7wolr2/v1. It describes FFPE block serial sectioning, slide processing and TMA sample distribution. Two compartments (Tumor and Stroma) in each TMA core were analyzed to determine cell-to-cell interactions in the tissues.
Deparaffinize by sequential incubation with Xylene (00:03:00 twice), 100% EtOH (00:01:00 twice) and 95% EtOH (00:01:00 ).
9m
Briefly rinse the slides with PBS and incubate with 1x Tris EDTA antigen retrieval buffer (9.0) for 00:15:00 (default condition for breast cancer tissue) at 100 °C in a steamer.
15m
Wash the slides with PBS for 00:05:00 (1 time).
5m
place the slides in a coplin jar and expose RNA targets with Proteinase K at 0.1 picomolar (pM) 0.1 microgram/mL, default condition for 00:15:00 at 37 °C in waterbath
15m
Wash the slides with PBS for 00:05:00 (1 time) and proceed to the next step immediately
5m
Postfixation
Postfixation
1h 59m
Preserve tissue morphology for soft tissues
Transfer the slides in a coplin jar containing 10 % volume NBF (neutral buffered formalin) buffer and incubate them for 00:05:00
5m
Transfere the 10slides in a coplin jar containg NBF stop buffer (Tris base 24.5 g and glycine 15 g in water 2 L ) and incubate them for 00:05:00 twice
10m
Wash the slides with PBS for 00:05:00
5m
in situ hybridization
in situ hybridization
1h 18m
RNA target hybridization
Prepare hybridization solution with Buffer R (warm at Room temperature before opening) and thaw RNA detection probes on ice. Before use, mix thorouhgly by pipetting. Once thawed, refrigerate at 4 °C for up to 3 months
Mix RNA probes at 1:10 dilution (eg. mix 25 µL of RNA probe mix in 200 µL of Buffer R and 25 µL of DEPC-treated water to make up to final volume 250 µL per slide )
Clean all equipment (especially for the hybridization chamber) with RNase AWAY and allow to dry or rinse with DEPC-treated water to avoid contamination
Wet Kimwipes with 2x SSC or DEPC-treated water and place them in the hybridization chamber to keep the slides from drying out
Add 200 µL of hybridization solution to each slide and gently apply a Grace Bio-Labs hybriSlip onto the slide. Avoid the formation of air bubbles on the tissue
Incubate the slides in the hybridization chamber oven at 37 °C for 16:00:00 (up to 24:00:00 )
1d
Stringent washes
Stringent washes
Off-target probe removal
100% Formamide (FA) solution should be adjusted to Room temperature before preparing the stringent washing solution. Mix 4x SSC and 100% FA in equal volume and prewarm the solution in two coplin jars at 37 °C in water bath
Gently remove the hybriSlip covers from the slides. Transfer slides in the stringent washing solution and wash twice at 37 °C for 00:25:00 in water bath.
25m
Wash slides with 2x SSC for 00:05:00 twice
10m
Tissue staining
Tissue staining
Staining with morphological markers
Block the slides with Buffer W for 00:30:00 at Room temperature
30m
Freshly prepare visualization markers (PanCK and CD45, Alexa-fluorescent conjugated) in 1:40 and Syto13 (1:10) dilution with Buffer W to a volume of 220ul per slide.
Incubate slides and visualization markers for 01:00:00 at Room temperature
1h
Wash the slides with 2x SSC for 00:05:00 twice
10m
Sample loading to GeoMx
Sample loading to GeoMx
Sample loading onto DSP GeoMx
Log onto GeoMx software and start with “New / Continue Run”.
After loading slides with the collection plate information, select (or create) the Readout group for NGS project. DSP is ready to scan slides.
30m
A slide scan name is created and panel/visualization marker information is selected as below:
In the Probe Reagent Kit field, select Human NGS Whole Transcriptome Atlas RNA V1.0
Select the FITC/525 nm, Cy5/568 nm and Texas Red/615 nm channels.
For FITC/525 nm, select SYTO 13 as fluorophore, DNA as biological target and enter 50 as exposure time.
For Cy5/568 nm, select Alexa 532 as fluorophore, PanCK as biological target and enter 300 as exposure time.
For Texas Red/615 nm, select Alexa 594 as fluorophore, CD45 as biological target and enter 300 as exposure time.
Select FITC/525 nm as focus channel
10m
When the scan area for each slide has been adjusted with sensitivity setting, select Scan.
ROI selection and sample collection
ROI selection and sample collection
ROI selection and AOI segmentation
After scanning is done, each color channel intensity is adjusted to show visualization markers along with tissue or cell line property
Each ROI is determined and selected by pathologist's guide, and drawn with circle (maximum 660um radius), rectangle (maximum 660x785um) or polygonal shape (maximum 660x785um). Due to the limitation of scan area in the slide loading slot, two same TMA slides were used to cover half of 88 cores in each slide (88 cores were embedded and slightly shifted to either left or right side of slide)
In the segment menu, 2 segmentation classes (Tumor, stromal and others) are added and parameters are set in the following order:
For Tumor segmentation (Segment 1), Alexa 532 (PanCK) is set to positive ("+") and the others set to ignore ("0") for the tumor collection
For stromal segmentation (Segment 2), Alexa 532 (PanCK), Alexa 594 (CD45) and FITC 525 (SYTO 13) were set to ignore ("0") to collect all other regions
Then click Generate Segments.
Once all segments are automatically generated, each channel parameter needs to be manually re-adjusted with pathologist's input to confirm if the segmentation is correctly done.
Caution: less than 20 cells in each segment is removed from collection due to threshold for low signal.
Once all AOI segmentation is complete, Exit Scan Workspace button icon is clicked to approve ROI selection and samples are collected in 96-well plate.
NGS library preparation and sequencing
NGS library preparation and sequencing
1h
Sample sequencing
The sample collection plates were completely dried out at 65 °C for 01:00:00 and shipped to the genomics core at Dana Faber Cancer Institute for NGS library preparation and sample sequencing
1h
FASTQ raw sequencing files were generated and converted to DCC (Digital Count Conversion) files which can be imported to DSP for the analysis
QC and data normalization
QC and data normalization
QC DSP data and normalization
Select and queue the slides to analyze using "New Analysis" in the DSP server.
Determine the New Analysis file name and save it in the designated folder.
Open an analysis file and perform the QC with default parameters except Negative control (4 or lesser), No template PCR control (NTC count, 1,000-10,000) and minimum nuclei counts (50-100)
(Optional) QC passed samples are processed with Biological probe QC (only applying to CTA)
QC data were filtered with low expression across all segments. Choose 5-10% of segments frequency or higher of Limit of Quantification (LoQ) or both.
Upper Quartile (Q3) normalization from all target were performed after filtering out low expression targets (Q3 normalization method is recommended since larger number of targets may increase the noise of geomean normalized data) and data was exported in csv format