Center for Biomedical Science and Policy

Exploring Variations in Gut Microbiome Networks among Patients with Chronic Kidney Disease (CKD)

GMU Center for Biomedical Science and Policy

Center for Biomedical Science and Policy Twitter/X

Author Information1
Caballero, Virginia2; Liao, Josephyn3; Yu, Alvin4
(Editor: Li, Meng-Hao5)

All authors are listed in alphabetical order
Walt Whitman High School, MD, Solon High School, OH, Irvington High School, CA, 5George Mason University, VA

Background

An imbalance in gut bacteria, known as gut dysbiosis, significantly increases the risk of chronic kidney disease (CKD), particularly in older adults [1]. The presence of dysbiosis in the gut bacteria not only contribute to the onset of CKD but also exacerbates its complications, such as high blood pressure, heart issues, and cognitive difficulties [1]. Dysbiosis triggers immune responses and the production of substances that cause inflammation and harm the kidneys [1]. Furthermore, treatments for CKD, such as antibiotics and specialized diets, can further disrupt the balance of gut bacteria [2]. While probiotics and targeted antibiotics are currently used to address gut imbalance, there is a need for improved methods to prevent CKD in individuals at risk of gut dysbiosis. Conversely, CKD can alter the gut bacteria in various ways. As kidney function declines, toxins like urea accumulate and promote the growth of gut bacteria [3]. These changes in gut bacteria can lead to dysbiosis and a condition known as leaky gut syndrome, where substances from the gut enter the body, exacerbating inflammation and kidney damage [3]. Additionally, CKD treatments, including antibiotics and dietary restrictions, can inadvertently worsen the balance of gut bacteria [4]. Thus, CKD and gut bacteria mutually influence each other, potentially exacerbating each other’s effects. Chronic kidney disease (CKD) and type 2 diabetes mellitus (T2DM) are prevalent chronic conditions worldwide, often occurring together and exacerbating each other’s impact on health. Patients with CKD, particularly those with end-stage renal disease (ESRD), frequently exhibit imbalanced gut bacteria, leading to altered composition [5]. Studies reveal a significant increase in specific bacteria types associated with colon inflammation, indicating a shift towards a pro-inflammatory gut environment exacerbated by heightened toxin levels due to impaired kidney function. These microbial changes contribute to local intestinal inflammation and impact systemic health, including cardiovascular complications. Cardiovascular disease (CVD), the leading cause of mortality in CKD patients, is influenced by imbalanced gut bacteria-producing inflammatory substances. Similarly, T2DM patients with CKD demonstrate similar bacterial imbalances, potentially worsening metabolic dysregulation and insulin resistance. The interaction between imbalanced gut bacteria, metabolic issues, and chronic inflammation complicates disease management, affecting kidney and heart health. Imbalanced bacteria also significantly affect cognitive function and bone health in CKD patients, exacerbating cognitive decline and contributing to bone density loss and vascular calcification, known as the “calcification paradox” [6]. Understanding these complex interactions is essential for developing targeted interventions to restore gut bacteria balance, potentially slowing CKD progression and its complications in individuals with diabetes.

Objective

The project investigated how gut microbiome networks varied among two groups of patients: patients with only chronic kidney disease and patients with both chronic kidney disease and diabetes.

Methods

Two matrices were created in R studio using information collected about hundreds of genera involved in patients with only CKD and patients with both CKD and diabetes. A threshold was necessary in order to round values to either 0 or 1. Hence if the value was less than 0.5, then it would appear as 0. Likewise if the value was greater than 0.5, then it would appear as 1. These matrices are then separately imported into Gephi, a software involved in analyzing networks. Gephi uses the information to create a graph. By changing the layout of the graph, certain clusters or communities are formed between genera. Many other statistics are then run to calculate components such as average degree (shows how high connection in a network is) or modularity (distinctness between clusters of a given network). A final comparison is done between both the graphs and the statistics to examine how they vary.

Results

In terms of network connectivity and structure, the network of patients with kidney disease was more connected (with a higher average degree) than the network of patients with kidney and diabetes. The average degree of kidney disease was 43.641 while the average degree of kidney and diabetes was 23.878. The network of patients with kidney disease only has a higher network density, or proportion of possible connections that exist, (0.120) compared to the network of patients with kidney and diabetes diseases (0.07). When analyzing centrality measures, we noticed there were specific bacterial taxa (nodes) in each network with significantly higher degree centrality (more connections to other taxa); the darker clusters having higher degrees. These central taxa differ between the two networks, and seem to be the key bacteria altered in the microbiomes of patients with these diseases. The bacteria Erwinia, specifically, acts as a bridge between communities since it has high betweenness centrality in only the network for kidney-only patients. There is a slight positive correlation between degree centrality and betweenness centrality in either network in both networks. The correlation seems higher in the kidney only network. Regarding network resilience, Amycolatopsis significantly altered the structure of its community when moved. In the kidney only network, Amycolatopsis had a rather high degree. However, upon being moved in the kidney diabetes network, it had a lower degree and was no longer in the same community as Georgenia and Dickeya. When diabetes is present in addition to kidney disease, more clusters and communities are created, hence becoming less compact and abundant. While Butryicimonas has a high betweenness centrality in patients with kidney disease, it has a very low betweenness centrality in patients with kidney disease and diabetes. Conversely, while Facklmia has a high betweenness centrality in patıents with kidney disease and diabetes, it has a low one in patients with only kidney disease. We inferred that the bacteria shifted from the groups in the middle to those in the outer part of the curved network that is connected with fewer nodes as the betweenness centrality decreased. Butryicimonas have a strong influence on the kidney disease-only network, and Facklmia has a strong influence on the kidney disease and diabetes network. 

Legend

The colors of the network represent Degree Centrality by the following key:

The colors associated with higher degrees represent bacteria involved in many links. 


The size of the labels in the network represents the Betweenness Centrality. The larger labels represent bacteria that fall the most in the shortest paths of the other pairs of nodes. The larger the label, the more important the bacteria. 

Figure 1. CKD Only

Different colors represent different communities. The network for patients with CKD only has 6 communities. Genera with bigger labels have higher degrees, in other words they have higher betweenness centrality and connect to more genera. In contrast, the network for patients with both CKD and diabetes has 12 communities. However, this causes it to have less compact clusters. With the addition of comorbidity, the central taxa changes making the entire structure of the network also change.

Figure 2. CKD and Diabetes

Conclusion

Thorough data analysis and detailed research allowed us to examine the significance of certain genera as well as the effects on them as a result of comorbidity. In this case, comorbidity (patients with CKD and diabetes) caused more clusters to form leading to less connection in the network than patients with only CKD.

References

Trandafir M, Pircalabioru GG, Savu O, 2024, March. Microbiota analysis in individuals with type two diabetes mellitus and end‑stage renal disease: A pilot study. Exp Ther Med. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11000444/

Feng Z, Wang T, Dong S, Jiang H, Zhang J, Raza HK, Lei G, 2021, October. Association between gut dysbiosis and chronic kidney disease: a narrative review of the literature. J Int Med Res., https://pubmed.ncbi.nlm.nih.gov/34704483/

Amini Khiabani S, Asgharzadeh M, Samadi Kafil H, 2023, August. Chronic kidney disease and gut microbiota. Heliyon. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10440536/  

Kanbay M, Onal EM, Afsar B, Dagel T, Yerlikaya A, Covic A, Vaziri ND, 2018, August. The crosstalk of gut microbiota and chronic kidney disease: role of inflammation, proteinuria, hypertension, and diabetes mellitus. Int Urol Nephrol. https://pubmed.ncbi.nlm.nih.gov/29728993/

Wen L, Duffy A. Factors Influencing the Gut Microbiota, Inflammation, and Type 2 Diabetes, 2017, July. J Nutr. https://pubmed.ncbi.nlm.nih.gov/28615382/

Tourountzis T, Lioulios G, Fylaktou A, Moysidou E, Papagianni A, Stangou M, 2022 September. Microbiome in Chronic Kidney Disease. Life (Basel). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9604691/