Comparison of two quantification methods, HALO Module and ImageJ, for immunohistochemical analysis of immune markers in broiler chicken intestinal tissue
Date
2025-05
Authors
Cao, Yuyang
Major Professor
Bobeck, Elizabeth
Advisor
Committee Member
Schmitz-esser, Stephan
Kerr, Brian
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Accurate quantification of immune markers in tissue sections is critical for evaluating host responses in animal disease models. While ImageJ remains a widely used platform for image analysis due to its accessibility and flexibility, its limitations in manual region selection, parameter range, and throughput constrain its utility in complex biological studies. To date, direct comparisons of ImageJ and HALO analytical methods using the same complex dataset are lacking in published literature. Therefore, the overarching goals of this work were to: 1) directly compare ImageJ and HALO platforms using the same dataset, and 2) identify the most informative and biologically relevant output variables within the HALO analytical framework for application in complex, tissue-based disease studies. Complex model systems using poultry disease models were used to generate biological samples for analysis. The disease models involved Salmonella, Eimeria, and necrotic enteritis challenges in a factorial arrangement with dietary interventions. Since the purpose of this work was to compare immunohistochemistry methods across two separate poultry disease models, detailed outcomes from those projects outside of immunohistochemistry are reported elsewhere. To complete objective 1, jejunal sections from broiler chickens were stained for IL-10 (DAPI) and IFNγ (TXred) and analyzed using both ImageJ and HALO Area Quantification Module. Only one disease model was analyzed with both methods. ImageJ identified significant increases in DAPI (P=0.0099) and IFNγ (P=0.0002) areas at peak necrotic enteritis in response to challenge, reflecting heightened inflammatory responses. Direct comparison of ImageJ and HALO using replicate 1 data revealed important differences in analytical output. For cleared mean IL-10 area, ImageJ reported non-significant effects for diet (P=0.9968 baseline; P=0.2511 peak challenge) and challenge (P=0.2168 baseline; P=0.8244 peak challenge), while cleared mean IFN-γ area showed a significant challenge effect at peak challenge (P=0.0206) despite no significant dietary effects. In contrast, HALO's IL-10 area yielded non-significant values across all comparisons (diet: P=0.6316 baseline, P=0.1795 peak challenge; challenge: P=0.7388 baseline, P=0.2229 peak challenge), as did IFN-γ area (diet: P=0.5588 baseline, P=0.9974 peak challenge; challenge: P=0.5264 baseline, P=0.0771 peak challenge). These results suggest that HALO’s strength in high-resolution, artifact-minimized tissue analysis for complex immunological studies. To complete objective 2, two disease models were used to analyze expression of immune markers including CD3, CD4, CD8, Bu1, KUL01, TCRγδ across four timepoints (day7 (baseline), 11 (3dpi), 14 (7dpi), 21 (14dpi)) in a Salmonella Enteritidis infection model. Key metrics such as dual-positive regions, DAPI-only and TXred-only compartments where they indicate IL-10 and each of the immune marker respectively, and signal intensity classifications revealed significant and trending effects associated with both challenge and diet. IL-10 co-expression highlighted its regulatory role in modulating inflammation, especially in CD4+, CD8+, and KUL01+ cell populations. HALO provides better analytical depth, reproducibility, and parameter richness than ImageJ. While ImageJ provided limited outputs—primarily DAPI and TXred area (e.g., DAPI%: P=0.0099; TXred%: P=0.0002)—HALO identified over 30 parameters per sample, including intensity-based and co-expression metrics. In Model 1, HALO detected significant differences in TXred area (P=0.0206), TXred weak area (P=0.0195), and DAPI average intensity (P=0.0295), revealing complex immune responses to Eimeria and necrotic enteritis. In Model 2, markers such as CD3, CD4, and Bu1 exhibited consistent significance across time points, including dual-positive areas (e.g., Bu1: P=0.0169 on Day 7), intensity changes, and compartment-specific distributions. HALO’s automated cell detection, phenotype classification, and artifact exclusion reduced user bias and enabled large-scale batch analysis. These advantages position HALO as the preferred tool for tissue-based immunological studies, capable of uncovering subtle yet biologically meaningful immune signatures across diverse experimental conditions.
Series Number
Journal Issue
Is Version Of
Versions
Series
Academic or Administrative Unit
Department of Animal Science
Type
creative component
Comments
Rights Statement
Attribution-NonCommercial-NoDerivs 3.0 United States
Copyright
2025