Disentangling Host-Microbiome Interactomes in Colorectal Cancer: A Network-Centric Approach
Written by: Muhammad Redha Abdullah Zawawi, Siti Aishah Sulaiman, Nur Alyaa Afifah Md Shahri, and Mira Farzana Mohamad Mokhtar
Colorectal cancer (CRC) is a highly prevalent cancer type and ranks the third leading cause of mortality globally. Over the previous decades, research has provided extensive evidence showing that the gut microbiome is crucial in impacting the host by linking imbalanced gut bacteria (dysbiosis) with various health conditions, such as gastrointestinal and metabolic diseases [1]. The gut microbiome has also been discovered as a potential risk factor for colorectal cancer (CRC) due to its ability to induce inflammation, which in turn affects the development and progression of CRC. An unhealthy diet and lifestyle significantly influence the composition of the gut microbiota. Changes in the gut microbiome can contribute to chronic inflammation via bacterial adherence to the mucosal surface of the colon epithelium, leading to the release of bacterial toxins and inducing the production of hepatocyte growth factor or DNA damage by reactive oxygen species (ROS), thereby promoting tumour proliferation [2].
According to the National Strategic Plan for Colorectal Cancer (NSPCRC), approximately 70% of Malaysian patients with CRC are diagnosed at an advanced stage (III or IV), making treatment more complicated and resulting in a poor survival rate [3]. In 2021, a study led by Assoc. Prof. Dr Neoh Hui Min from UMBI made a remarkable discovery by reporting four dominant bacterial species—Parvimonas micra, Fusobacterium nucleatum, Peptostreptococcus stomatis and Akkermansia muciniphila—in our local CRC patients [4]. The increased inflammatory response in the cells was further validated in response to P. micra infection, thus confirming the bacterial role in promoting cell proliferation and inflammation [5]. Under the same host or environment, P. micra can also manipulate the production of virulence factors in co-infecting bacteria, such as P. gingivalis, which exhibits enhanced growth and increased production of secreted proteases called gingipains. This protease binds to the host cells and catalyses the cleavage of proinflammatory cytokines, later disrupting host immune signalling [6]. It is revealed that the human gut microbiome is a highly intricate ecosystem that hosts diverse ecological interactions and plays a vital role in maintaining human health [7].
Understanding the interaction between the human host and microbiome could be game-changing in medicine. Protein-protein interactions (PPIs) play a vital role in comprehending the pathogenesis or infection in all living cells. However, the mechanisms underlying the complex interaction networks within the human gut have yet to be fully understood in CRC. The application of network approaches has dramatically enhanced the study of host-microbiome interactions (HMIs) by enabling a systems-level understanding and facilitating the discovery of new insights into the complex interacting proteins between the host and microbiome. (i) Protein homology (interolog-based and (ii) the interaction of protein domain (domain-based) is commonly used to develop computational methods for predicting HMIs by leveraging the similarities of the proteins and their conserved domains [8]. Limited research has also explored structural-based and motif-based approaches as computational strategies for investigating disease-related HMI systems [9]. It is crucial to note that incorporating diverse data types and utilising multiple computational methods can lead to developing a more robust and reliable model for the HMI network.
FIGURE 1. The interplay between the gut microbiome and consecutive progression of CRC, as well as the computational approaches—namely, (i) interolog-based and (ii) domain-based methods—used for predicting the host-microbiome interaction (HPI) causing CRC.
The interolog-based method utilises sequence similarity and evolutionary conservation to identify potential protein-protein interactions, providing valuable insights into the functional relationships between proteins and their involvement in various biological processes. In this method, if a pair of proteins interact in one species (x, y) and these proteins have homologous counterparts in another species (x’, y’), the interaction between the homologous proteins is assumed to be conserved. For example, (x, y) is a template PPI pair in a reference species; x’ (host) and y’ (pathogen) are identified as homologous proteins to x and y, respectively. Thus, it is postulated that the PPI pair (x’, y’) interacts with each other [8,9]. The template PPI pair used for the interolog-based method can be retrieved from the existing known interaction from public databases like Database of Interacting Protein (DIP), IntAct Molecular Interaction Database (IntAct), Molecular INTeraction (MINT), Host-Pathogen Interaction Database (HPIDB) and Biological General Repository for Interaction Datasets (BioGrid), to generate MPI network.
On the other hand, the domain-based method refers to the recognition and response mechanisms mediated by specific protein domains. Protein domains are integral components that are critical in specific HMIs and functions. During a pathogen infection, the host immune system is triggered and activated through a series of complex molecular interactions. These interactions involve protein domains derived from both the host and the pathogen, where host proteins containing specific domains could recognise the virulence factors produced by the pathogen [8,9]. In the computational approach context, protein domains act as the mediator of interactions. The evaluation of HMI involves predicting the interaction between protein pairs (x, y) by considering the presence of functional domains (d1, d2) within each pair. Each pair of host and pathogen proteins that share at least one domain will be considered protein interacting partners. Like the interolog-based method, template domain-domain interaction (DDI) pair can be retrieved from the current DDI information, including Three-Dimensional Interacting Domains (3DID), Database of Protein Domain Interactions (DOMINE) and Integrated Domain-Domain Interaction (IDDI). By incorporating interolog- and domain-based information, the risk of overpredicting interactions can be mitigated and false positives in the predictions can be reduced. This network-centric analysis provides a glimpse into the interactions between hosts and microbiomes that could lead to a better understanding of infection mechanisms and the impact of the gut microbiome on CRC.
With the growing number of generated host-microbiome interaction datasets, proper management to store the data is essential. In UMBI, Total Research Information Management System (TRIMS) dan High-Performance Computer (HPC) were set up between 2016 to 2017 for efficient storage, organisation, and retrieval of our in-house research data. The users have actively utilised the TRIMS-HPC system, particularly involving a big data analysis project using an HPC server. In the coming years, a microbiome study funded by Dana Langganan Sukuk Pakej Rangsangan Ekonomi Prihatin Rakyat (PRIHATIN), led by Professor Datuk Dr A Rahman A Jamal (UMBI’s Principal Research Fellow) will be carried out by investigating the association between the microbiome and the non-communicable diseases like cancer, obesity, and Alzheimer’s. The interactions between gut microbiomes and human hosts are becoming a hot topic in this area of research as a new target to tackle NCD. With shotgun sequencing technology, the genome of the gut microbiome from Malaysian patients will be extracted and subsequently processed and stored in TRIMS-HPC, making UMBI one of the leading local institutes to study disease-causing microbiomes. While the causal relationship between microbes and NCD is still in its early stages, advancing the molecular understanding of these disease-modulating interactions may hold significant scientific impact and clinical value for precise treatment in the future.
References:
[1]. Zou, S., Fang, L., & Lee, M. H. (2018). Dysbiosis of gut microbiota in promoting the development of colorectal cancer. Gastroenterology report, 6(1), 1-12.
[2]. Kim, J., & Lee, H. K. (2022). Potential role of the gut microbiome in colorectal cancer progression. Frontiers in immunology, 12, 5786.
[3]. Ministry of Health Malaysia 2021. National Strategic Plan for Colorectal Cancer 2021-2025.
[4]. Osman, M. A., Neoh, H. M., Ab Mutalib, N. S., Chin, S. F., Mazlan, L., Raja Ali, R. A., … & Jamal, R. (2021). Parvimonas micra, Peptostreptococcus stomatis, Fusobacterium nucleatum and Akkermansia muciniphila as a four-bacteria biomarker panel of colorectal cancer. Scientific reports, 11(1), 2925.
[5]. Hatta, M. N. A., Mohamad Hanif, E. A., Chin, S. F., Low, T. Y., & Neoh, H. M. (2023). Parvimonas micra Infection Enhances Proliferation, Wound Healing, and Inflammation of a Colorectal Cancer Cell Line. Bioscience Reports, BSR20230609.
[6]. Higashi, D. L., Krieger, M. C., Qin, H., Zou, Z., Palmer, E. A., Kreth, J., & Merritt, J. (2023). Who is in the driver’s seat? Parvimonas micra: An understudied pathobiont at the crossroads of dysbiotic disease and cancer. Environmental Microbiology Reports.
[7]. Priya, S., & Blekhman, R. (2019). Population dynamics of the human gut microbiome: change is the only constant. Genome biology, 20, 1-4.
[8]. Loaiza, C. D., Duhan, N., Lister, M., & Kaundal, R. (2021). In silico prediction of host–pathogen protein interactions in melioidosis pathogen Burkholderia pseudomallei and human reveals novel virulence factors and their targets. Briefings in Bioinformatics, 22(3), bbz162.
[9]. Nourani, E., Khunjush, F., & Durmuş, S. (2015). Computational approaches for prediction of pathogen-host protein-protein interactions. Frontiers in microbiology, 6, 94.