Pre-transplant Transcriptional Signature in Peripheral Blood Mononuclear Cells of Acute Renal Allograft Rejection

Wenyu Xiang, Shuai Han, Cuili Wang, Hongjun Chen, Lingling Shen, Tingting Zhu, Kai Wang, Wenjie Wei, Jing Qin, Nelli Shushakova, Song Rong, Hermann Haller, Hong Jiang, Jianghua ChenFrontiers in Medicine 2022 · DOI

TL;DR

This study identifies a unique "transcriptional signature" in peripheral blood mononuclear cells (PBMCs) that can predict acute renal allograft rejection (AR) before it occurs. By analyzing gene expression patterns, researchers discovered a combination of four genes that accurately distinguishes AR patients from healthy controls, offering a promising avenue for early diagnosis and improved patient outcomes.

What Problem Does This Paper Address?

Kidney transplantation is a life-saving procedure for end-stage renal disease, but a major hurdle is acute rejection (AR). AR occurs when the recipient's immune system attacks the transplanted kidney, leading to graft dysfunction and potential failure. While current methods can detect AR, they often rely on invasive biopsies or detect changes only after significant damage has occurred. There's a critical need for non-invasive, early detection methods to identify patients at risk of AR before it manifests clinically, allowing for timely intervention and better graft survival. This paper aims to address this by investigating the molecular changes in peripheral blood cells that precede AR.

Methods

The researchers employed a multi-pronged approach combining RNA sequencing (RNA-Seq) with various bioinformatics techniques.

  • RNA Sequencing (RNA-Seq): This core technique was used to comprehensively analyze the gene expression profiles of PBMCs. Samples were collected from kidney transplant recipients who either experienced AR (classified as T cell-mediated rejection (TCMR) or antibody-mediated rejection (ABMR)) or did not (non-AR).
  • Differential Gene Expression (DEG) Analysis: Standard statistical methods were used to identify genes that were significantly up- or down-regulated in AR patients compared to non-AR patients.
  • Gene Ontology (GO) Analysis: This was used to understand the biological functions and pathways associated with the identified DEGs.
  • Pathway Analysis (Reactome, Ingenuity Pathway Analysis - IPA): These tools helped to pinpoint specific biological pathways that were altered in AR, providing insights into the underlying molecular mechanisms. IPA also allowed for the prediction of upstream regulators and downstream effects.
  • Protein-Protein Interaction (PPI) Network Analysis: This technique, using STRING and Cytoscape, helped to identify "hub genes" – genes that are highly connected and likely play central roles in the observed molecular changes.
  • Cell-Type Enrichment Analysis (xCell, CIBERSORT): Since PBMCs are a mixture of different immune cell types, these methods were used to estimate the abundance of various immune cell populations within the samples, helping to understand which cell types might be driving the observed gene expression changes.
  • Receiver Operating Characteristic (ROC) Curve Analysis: This statistical method was used to evaluate the diagnostic power of individual genes and combinations of genes for distinguishing between AR and non-AR groups.

Datasets

The study utilized peripheral blood mononuclear cell (PBMC) samples collected from:

  • 32 kidney allograft donors.
  • 42 corresponding recipients.

These recipients were categorized into:

  • Acute Rejection (AR) group:
  • T cell-mediated rejection (TCMR): n = 18
    • Antibody-mediated rejection (ABMR): n = 5
  • Non-AR group (normal/non-specific changes): n = 19

The classification of AR was based on the Banff 2017 classification of renal allograft biopsies. The study did not introduce new benchmark datasets but rather analyzed existing patient samples.

Key Insights and Ideas

  • Immune System Dysregulation is Key: The study strongly suggests that AR is driven by significant dysregulation of the immune system, with inflammatory responses and complement activation pathways being particularly prominent.
  • IL-10's Dual Role: The interleukin-10 (IL-10) signaling pathway emerged as a significant player, being enriched in both upregulated and downregulated genes. This hints at a complex, potentially context-dependent role for IL-10 in AR.
  • Histone Genes as Potential Players: The identification of several histone-related genes as "hub genes" in the PPI network is intriguing. Histones are crucial for DNA packaging, and their involvement suggests potential epigenetic modifications or altered gene regulation in AR.
  • Pro- vs. Anti-inflammatory Imbalance: The IPA analysis revealed an upregulation of pro-inflammatory pathways and a downregulation of anti-inflammatory pathways in AR patients, highlighting a critical imbalance that likely contributes to rejection.
  • A Powerful Predictive Signature: The most significant insight is the identification of a specific combination of four genes (TNFSF14, TANK, ANKRD33B, and HSPA1B) that can accurately predict AR. This suggests a potential for a non-invasive blood test for early AR detection.

Key Results

  • Differential Gene Expression: 975 differentially expressed genes (DEGs) were identified between AR and non-AR patients, with 776 upregulated and 199 downregulated.
  • Enriched Pathways: Significant enrichment of inflammatory response, complement activation, and IL-10 signaling pathways was observed in AR patients.
  • Hub Genes: IL-10, VEGFA, CXCL8, MMP9, and several histone-related genes were identified as key hub genes.
  • Predictive Biomarker Combination: A combination of four genes (TNFSF14, TANK, ANKRD33B, and HSPA1B) achieved an Area Under the Curve (AUC) of 92.3% (95% CI 82.8–100) in discriminating between AR and non-AR patients. This indicates a high diagnostic accuracy.

Limitations and Future Work

The authors acknowledge several limitations:

  • Small Sample Size: The study was conducted on a relatively small cohort, and larger, multi-center studies are needed to validate the findings.
  • Validation Needs: The identified transcriptional profiles require validation through independent RNA-Seq studies and analysis of kidney allograft biopsies.
  • Treatment Variability: Patients receive varying immunosuppressive drug regimens, which could influence gene expression and potentially confound results. Future work should account for these differences.

Despite these limitations, the study provides a strong foundation for future research. The authors suggest:

  • Further Investigation of Biomarkers: The identified four-gene signature warrants further investigation for its clinical utility in monitoring transplant patients.
  • Understanding Pathogenesis: The study offers a framework for a deeper understanding of the molecular mechanisms underlying acute renal allograft rejection.
  • Exploring Novel Pathways: The role of pathways like IL-10 signaling, CXCL8, and VEGFA in AR, which have been less explored, should be further investigated.

Summary updated

Abstract

Acute rejection (AR) is closely associated with renal allograft dysfunction. Here, we utilised RNA sequencing (RNA-Seq) and bioinformatic methods to characterise the peripheral blood mononuclear cells (PBMCs) of patients with acute renal allograft rejection. Pretransplant blood samples were collected from 32 kidney allograft donors and 42 corresponding recipients with biopsies classified as T cell-mediated rejection (TCMR, <i>n</i> = 18), antibody-mediated rejection (ABMR, <i>n</i> = 5), and normal/non-specific changes (non-AR, <i>n</i> = 19). The patients with TCMR and ABMR were assigned to the AR group, and the patients with normal/non-specific changes (<i>n</i> = 19) were assigned to the non-AR group. We analysed RNA-Seq data for identifying differentially expressed genes (DEGs), and then gene ontology (GO) analysis, Reactome, and ingenuity pathway analysis (IPA), protein-protein interaction (PPI) network, and cell-type enrichment analysis were utilised for bioinformatics analysis. We identified DEGs in the PBMCs of the non-AR group when compared with the AR, ABMR, and TCMR groups. Pathway and GO analysis showed significant inflammatory responses, complement activation, interleukin-10 (IL-10) signalling pathways, classical antibody-mediated complement activation pathways, etc., which were significantly enriched in the DEGs. PPI analysis showed that IL-10, VEGFA, CXCL8, MMP9, and several histone-related genes were the hub genes with the highest degree scores. Moreover, IPA analysis showed that several proinflammatory pathways were upregulated, whereas antiinflammatory pathways were downregulated. The combination of NFSF14+TANK+ANKRD 33 B +HSPA1B was able to discriminate between AR and non-AR with an AUC of 92.3% (95% CI 82.8-100). Characterisation of PBMCs by RNA-Seq and bioinformatics analysis demonstrated gene signatures and biological pathways associated with AR. Our study may provide the foundation for the discovery of biomarkers and an in-depth understanding of acute renal allograft rejection.
BibTeX
@inproceedings{xiang2022pre,
  author = {Wenyu Xiang and Shuai Han and Cuili Wang and Hongjun Chen and Lingling Shen and Tingting Zhu and Kai Wang and Wenjie Wei and Jing Qin and Nelli Shushakova and Song Rong and Hermann Haller and Hong Jiang and Jianghua Chen},
  title = {{Pre-transplant Transcriptional Signature in Peripheral Blood Mononuclear Cells of Acute Renal Allograft Rejection}},
  booktitle = {Frontiers in Medicine},
  year = {2022},
  doi = {10.3389/fmed.2021.799051},
}
Figure