Ncounter advanced analysis 2

Total RNA was extracted from conjugate 40a-treated or vehicle-treated tumor tissues of HCC-bearing
mice with the RNA-easy kit (QIAGEN). The concentration (absorbance
at 260 nm) and purity (A260/280 and A260/230 ratios) of the extracted
RNA were measured by spectrophotometry, and the integrity of the RNA
was further determined by a 2100 Bioanalyzer system (Agilent Technologies).
The RNA, hybridized with barcoded probes (NanoString Technologies)
provided in the nCounter Mouse PanCancer Immune Profiling panel kit,
was then used for measurement of the mRNA expression of 770 genes
related to immune responses. Nanostring nSolver 4.0 and nCounter advanced
analysis 2.0 software (NanoString Technologies) were used for data
processing, immune cell profiling, and pathway scoring according to
developer’s instructions.

In this study, the nCounter PanCancer Pathways panel that included 770 genes from 13 canonical pathways (see Supplementary Table S1 for gene and probe information). These gene sets covered diverse biological pathways such as Notch, Wnt, Hedgehog, chromatin modification, transcriptional misregulation, DNA damage repair, TGFβ, MAPK, JAK-STAT, PI3K, Ras, cell cycle, and apoptosis. The samples were read at 555 FOV (Field of view) and resulting RCC data files were analyzed for QC in nSolver 3.0. Subsequent analyses were performed using the nCounter Advanced Analysis 2.0 plug-in (NanoString Technologies, Inc. 530 Fairview Ave N, Seattle, Washington, USA). The gene expression normalization was performed using the geNorm algorithm that selected the best housekeeping genes from the initial list of 40 genes (attached). To visualize the results, unsupervised clustering was used to generate heatmap based on the QC passed, normalized data counts of individual genes. Differential expression was graphed as a volcano plot with individual genes −log₁₀ (p-value) and log₂ fold change compared to the control group. Pathview module was used to display overexpressed genes (gold color) or downregulated genes (blue color) overlaid on KEGG pathways.

We calculated the T cell‐inflamed GEP score using the weighted sum of normalized expression values of 18 genes related to antigen presentation, chemokine expression, cytotoxic activity, and adaptive immune resistance: PSMB10, HLA‐DQA1, HLA‐DRB1, CMKLR1, HLA‐E, NKG7, CD8A, CCL5, CXCL9, CD27, CXCR6, IDO1, STAT1, CD274 (PD‐L1), CD276 (B7‐H3), LAG3, PDCD1LG2 (PDL2) and TIGIT, as previously described.⁸, ⁹ The continuous score was then standardized to have a mean of 0 and a standard deviation (SD) of 1. The categorical score was dichotomized into T cell‐inflamed GEP^Hi (upper tertile) and T cell‐inflamed GEP^Low (middle and lower tertile). From the previous work,⁸, ⁹ most people with a score below the top tertile had rapid disease progression. The abundance of immune cell populations was quantified using the average of the log₂‐transformed expression values for sets of marker genes that are expressed stably and specifically in given cell types¹³ and implemented in the nCounter Advanced Analysis 2.0 module of Nanostring's nSolver software.

Samples from four high (Treg level at D64 > 250 cell/µl blood), four low-Treg-responders (Treg level at D64 < 150 cell/µl) and four placebo patients at D1, D8 and D64 were investigated using the NanoString platform. Patient classification is illustrated in Table 1. Briefly, 300 ng of total RNA was mixed with capture and reporter probes from the auto-immune discovery panel. A hybridization period was allowed for 16 h at 65°C. Samples were scanned using nCounter^® SPRINT profiler (NanoString Technologies Inc.). The autoimmune discovery panel contains 755 mRNA targets: 740 immune-related transcripts and 15 housekeeping genes. NanoString data were analysed using nSolver^TM4.0 and nCounter Advanced Analysis 2.0 (NanoString Technologies Inc.). Moreover, to identify which selection of transcripts were responsible for patient group differences, data were imported into Qlucore Omics Explorer (Qlucore) and multigroup comparison statistical analysis was performed.

Data were shown as mean ± standard deviation (SD). Data were analyzed with unpaired two-tailed Student’s t-test when comparing between two groups, in GraphPad Prism (GraphPad software, La Jolla, CA, www.graphpad.com). Statistical significance among three groups were analyzed using two-way ANOVA test. p-values of <0.05 were considered as statistically significant (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001). Statistical analysis for NanoString data was performed by R-based statistic program by either loglinear regression or simplified negative binomial model and adjusted with the Benjamini–Yekutieli method in nCounter Advanced Analysis 2.0 (NanoString Technologies). Samples were considered significant if p < 0.01 (−log₁₀ > 2).