Deconvolution of expression microarray data reveals 131I-induced responses otherwise undetected in thyroid tissue

Fourteen years of cellular deconvolution: methodology, applications, technical evaluation and outstanding challenges

Single-cell RNA sequencing (scRNA-Seq) is a recent technology that allows for the measurement of the expression of all genes in each individual cell contained in a sample. Information at the single-cell level has been shown to be extremely useful in many areas. However, performing single-cell experiments is expensive. Although cellular deconvolution cannot provide the same comprehensive information as single-cell experiments, it can extract cell-type information from bulk RNA data, and therefore it allows researchers to conduct studies at cell-type resolution from existing bulk datasets. For these reasons, a great effort has been made to develop such methods for cellular deconvolution. The large number of methods available, the requirement of coding skills, inadequate documentation, and lack of performance assessment all make it extremely difficult for life scientists to choose a suitable method for their experiment. This paper aims to fill this gap by providing a comprehensive review of 53 deconvolution methods regarding their methodology, applications, performance, and outstanding challenges. More importantly, the article presents a benchmarking of all these 53 methods using 283 cell types from 30 tissues of 63 individuals. We also provide an R package named DeconBenchmark that allows readers to execute and benchmark the reviewed methods (https://github.com/tinnlab/DeconBenchmark).

Development of a machine learning framework for radiation biomarker discovery and absorbed dose prediction

Background

Molecular radiation biomarkers are an emerging tool in radiation research with applications for cancer radiotherapy, radiation risk assessment, and even human space travel. However, biomarker screening in genome-wide expression datasets using conventional tools is time-consuming and underlies analyst (human) bias. Machine Learning (ML) methods can improve the sensitivity and specificity of biomarker identification, increase analytical speed, and avoid multicollinearity and human bias.

Aim

To develop a resource-efficient ML framework for radiation biomarker discovery using gene expression data from irradiated normal tissues. Further, to identify biomarker panels predicting radiation dose with tissue specificity.

Methods

A strategic search in the Gene Expression Omnibus database identified a transcriptomic dataset (GSE44762) for normal tissues radiation responses (murine kidney cortex and medulla) suited for biomarker discovery using an ML approach. The dataset was pre-processed in R and separated into train and test data subsets. High computational cost of Genetic Algorithm/k-Nearest Neighbor (GA/KNN) mandated optimization and 13 ML models were tested using the caret package in R. Biomarker performance was evaluated and visualized via Principal Component Analysis (PCA) and dose regression. The novelty of ML-identified biomarker panels was evaluated by literature search.

Results

Caret-based feature selection and ML methods vastly improved processing time over the GA approach. The KNN method yielded overall best performance values on train and test data and was implemented into the framework. The top-ranking genes were Cdkn1a, Gria3, Mdm2 and Plk2 in cortex, and Brf2, Ccng1, Cdkn1a, Ddit4l, and Gria3 in medulla. These candidates successfully categorized dose groups and tissues in PCA. Regression analysis showed that correlation between predicted and true dose was high with R² of 0.97 and 0.99 for cortex and medulla, respectively.

Conclusion

The caret framework is a powerful tool for radiation biomarker discovery optimizing performance with resource-efficiency for broad implementation in the field. The KNN-based approach identified Brf2, Ddit4l, and Gria3 mRNA as novel candidates that have been uncharacterized as radiation biomarkers to date. The biomarker panel showed good performance in dose and tissue separation and dose regression. Further training with larger cohorts is warranted to improve accuracy, especially for lower doses.

Age-related long-term response in rat thyroid tissue and plasma after internal low dose exposure to 131I

131I is used clinically for therapy, and may be released during nuclear accidents. After the Chernobyl accident papillary thyroid carcinoma incidence increased in children, but not adults. The aims of this study were to compare 131I irradiation-dependent differences in RNA and protein expression in the thyroid and plasma of young and adult rats, and identify potential age-dependent biomarkers for 131I exposure. Twelve young (5 weeks) and twelve adult Sprague Dawley rats (17 weeks) were i.v. injected with 50 kBq 131I (absorbed dose to thyroid = 0.1 Gy), and sixteen unexposed age-matched rats were used as controls. The rats were killed 3-9 months after administration. Microarray analysis was performed using RNA from thyroid samples, while LC-MS/MS analysis was performed on proteins extracted from thyroid tissue and plasma. Canonical pathways, biological functions and upstream regulators were analysed for the identified transcripts and proteins. Distinct age-dependent differences in gene and protein expression were observed. Novel biomarkers for thyroid 131I exposure were identified: (PTH), age-dependent dose response (CA1, FTL1, PVALB (youngsters) and HSPB6 (adults)), thyroid function (Vegfb (adults)). Further validation using clinical samples are needed to explore the role of the identified biomarkers.

Long-term transcriptomic and proteomic effects in Sprague Dawley rat thyroid and plasma after internal low dose 131I exposure

Background

Radioiodide (¹³¹I) is commonly used to treat thyroid cancer and hyperthyroidis.¹³¹I released during nuclear accidents, have resulted in increased incidence of thyroid cancer in children. Therefore, a better understanding of underlying cellular mechanisms behind ¹³¹I exposure is of great clinical and radiation protection interest. The aim of this work was to study the long-term dose-related effects of ¹³¹I exposure in thyroid tissue and plasma in young rats and identify potential biomarkers.

Materials and methods

Male Sprague Dawley rats (5-week-old) were i.v. injected with 0.5, 5.0, 50 or 500 kBq ¹³¹I (D_thyroid ca 1–1000 mGy), and killed after nine months at which time the thyroid and blood samples were collected. Gene expression microarray analysis (thyroid samples) and LC-MS/MS analysis (thyroid and plasma samples) were performed to assess differential gene and protein expression profiles in treated and corresponding untreated control samples. Bioinformatics analyses were performed using the DAVID functional annotation tool and Ingenuity Pathway Analysis (IPA). The gene expression microarray data and LC-MS/MS data were validated using qRT-PCR and ELISA, respectively.

Results

Nine ¹³¹I exposure-related candidate biomarkers (transcripts: Afp and RT1-Bb, and proteins: ARF3, DLD, IKBKB, NONO, RAB6A, RPN2, and SLC25A5) were identified in thyroid tissue. Two dose-related protein candidate biomarkers were identified in thyroid (APRT and LDHA) and two in plasma (DSG4 and TGM3). Candidate biomarkers for thyroid function included the ACADL and SORBS2 (all activities), TPO and TG proteins (low activities). ¹³¹I exposure was shown to have a profound effect on metabolism, immune system, apoptosis and cell death. Furthermore, several signalling pathways essential for normal cellular function (actin cytoskeleton signalling, HGF signalling, NRF2-mediated oxidative stress, integrin signalling, calcium signalling) were also significantly regulated.

Conclusion

Exposure-related and dose-related effects on gene and protein expression generated few expression patterns useful as biomarkers for thyroid function and cancer.

Transcriptional effects of 177Lu-octreotate therapy using a priming treatment schedule on GOT1 tumor in nude mice

BACKGROUND

¹⁷⁷Lu-octreotate is used for therapy of somatostatin receptor expressing neuroendocrine tumors with promising results, although complete tumor remission is rarely seen. Previous studies on nude mice bearing the human small intestine neuroendocrine tumor, GOT1, have shown that a priming injection of ¹⁷⁷Lu-octreotate 24 h before the main injection of ¹⁷⁷Lu-octreotate resulted in higher ¹⁷⁷Lu concentration in tumor, resulting in increased absorbed dose, volume reduction, and time to regrowth. To our knowledge, the cellular effects of a priming treatment schedule have not yet been studied. The aim of this study was to identify transcriptional changes contributing to the enhanced therapeutic response of GOT1 tumors in nude mice to ¹⁷⁷Lu-octreotate therapy with priming, compared with non-curative monotherapy.

RESULTS

RNA microarray analysis was performed on tumor samples from GOT1-bearing BALB/c nude mice treated with a 5 MBq priming injection of ¹⁷⁷Lu-octreotate followed by a second injection of 10 MBq of ¹⁷⁷Lu-octreotate after 24 h and killed after 1, 3, 7, and 41 days after the last injection. Administered activity amounts were chosen to be non-curative, in order to facilitate the study of tumor regression and regrowth. Differentially regulated transcripts (RNA samples from treated vs. untreated animals) were identified (change ≥ 1.5-fold; adjusted p value < 0.01) using Nexus Expression 3.0. Analysis of the biological effects of transcriptional regulation was performed using the Gene Ontology database and Ingenuity Pathway Analysis. Transcriptional analysis of the tumors revealed two stages of pathway regulation for the priming schedule (up to 1 week and around 1 month) which differed distinctly from cellular responses observed after monotherapy. Induction of cell cycle arrest and apoptotic pathways (intrinsic and extrinsic) was found at early time points after treatment start, while downregulation of pro-proliferative genes were found at a late time point.

CONCLUSIONS

The present study indicates increased cellular stress responses in the tumors treated with a priming treatment schedule compared with those seen after conventional ¹⁷⁷Lu-octreotate monotherapy, resulting in a more profound initiation of cell cycle arrest followed by apoptosis, as well as effects on PI3K/AKT-signaling and unfolded protein response.