Page 78 - TD-4-1

Tumor Discovery Colorectal cancer: miRNA, mRNA, protein insights

originated from adenocarcinomas and were acquired 2.4.1.2. Assessment of protein expression
through surgical interventions. Following initial validation under a light microscope,
2.2. Macrodissection TMA slides were scanned at 20× magnification using a
Nanozoomer Digital Pathology scanner (Hamamatsu
To mitigate potential confounding effects from stromal Photonics, Japan). Protein expression in tumor cells was
cells, tumor specimens underwent macrodissection assessed semi-quantitatively using the H-score method,
following evaluation by a pathologist to ensure a minimum which combines the percentage of positive tumor cells with
of 50% tumor tissue content, as recommended in staining intensity (0 for negative, 1 for weak, 2 for moderate,
literature. Two 20 µm-thick serial sections were excised and 3 for strong staining). To ensure consistency, all slides
22
25
from each paraffin block. After identifying tumor regions underwent independent evaluation by a second scorer, and
on unstained sections from hematoxylin–eosin-stained the intraclass correlation coefficient (ICC) was employed
slides, macrodissection was carried out. Total RNA and to assess inter-scorer agreement.
miRNA were subsequently isolated using the miRNeasy
FFPE kit (Qiagen, Hilden, Germany). 2.5. Statistical analysis
Statistical analysis was conducted using the SPSS version 22
2.3. Quantitative reverse-transcription polymerase software package (IBM acquired SPSS Inc. USA).
chain reaction (RT-qPCR)
Categorical data were assessed for statistical significance
After synthesizing cDNA with the miScript II RT Kit and using the Chi-square test, while continuous data were
the QuantiTec Reverse Transcription Kit (Qiagen, Hilden, analyzed for differences between datasets using the
Germany) for miRNA and mRNA, respectively, the targeted Wilcoxon test. Fisher’s exact test was employed to explore
genes were quantified using the miScript SYBR Green associations between unpaired tumor groups. Spearman’s
PCR kit (Qiagen) on a 7500 Fast Real-Time PCR System correlation was used to detect correlations between targets.
(Applied Biosystems, Thermo Fisher, USA). The primer For multiple corrections testing, the Bonferroni step-down
sequences and their efficiencies, determined through assay (Holm) correction was applied. In both statistical analyses,
optimization, are provided in Supplementary File. The P < 0.05 was considered statistically significant. In this
ΔΔCt method was employed for relative quantification study, we used the Wilcoxon test, a non-parametric
of mRNA, comparing miRNA and mRNA expression in method robust to outliers and suitable for non-normally
normal versus CRC tissues. RNU6B and HPRT were used distributed data. Outliers, identified as data points outside
as reference genes for miRNA and mRNA, respectively. 23,24 the 10 and 90 percentiles in Figure 1, were retained in
th
th
the analysis without additional transformation or removal.
2.4. Evaluation of protein expression This approach provides an accurate representation of the
2.4.1. Immunohistochemistry dataset, accommodating natural data variation without the
influence of strict distributional assumptions.
2.4.1.1. TMA
After validating antibody specificity and determining 3. Results
optimal concentrations, we assessed protein expression 3.1. MiRNA quantification
in CRC tissue using Tissue Microarray (TMA) sections.
Western blotting analysis confirmed the specificity of the 3.1.1. Cutoff point for miRNAs detection
antibodies (Supplementary file Figure S1). TMAs facilitate Before analyzing mRNA expression levels, we sought
high-throughput evaluation of biomarker expression across to establish a cutoff point to distinguish high and low
numerous tissue samples, comprising paraffin blocks with expression levels. Initially, RNA was extracted from 20
minute tissue specimens arranged in an array configuration. individual pure normal colon tissues and pooled with
The antibodies employed are listed in Supplementary File. equal volumes. Subsequently, the expression levels of
TMA sections were prepared at the Nottingham Health all miRNAs and mRNAs were assessed in each normal
Science Biobank, QMC, Nottingham, UK. We stained colon tissue sample compared to the pooled sample. On
4-µm paraffin-embedded CRC TMA sections using average, the minimum fold expression for all mRNAs in
Novolink Polymer Detection Systems (Leica Microsystems, normal colon tissues was 0.6, while the maximum was 1.8.
Germany) with anti-SMAD4, anti-KLF4, anti-RASA1, Downregulation was defined as <0.6 fold, and upregulation
anti-PTEN, anti-TGFBRII, and anti-BCL2 antibodies. Each as >1.8 fold. This method provided a robust benchmark for
run included positive and negative controls to validate distinguishing between high and low mRNA expression
experimental success. Detailed immunohistochemistry in subsequent analyses. For the miRNAs, the average
staining procedures are provided in Supplementary File. minimum fold change in expression for all miRNAs in

Volume 4 Issue 1 (2025) 70 doi: 10.36922/td.4631