Identifying the molecular association between diabetic nephropathy and hepatocellular carcinoma: an <i>in-silico</i> network biology approach

Tammanna R. Sahrawat; Arnesh Saxena

doi:10.22730/jmls.2025.22.2.60

J Med Life Sci > Volume 22(2); 2025 > Article

Sahrawat and Saxena: Identifying the molecular association between diabetic nephropathy and hepatocellular carcinoma: an in-silico network biology approach

Original Article

Journal of Medicine and Life Science 2025;22(2):60-68.

Published online: May 28, 2025

DOI: https://doi.org/10.22730/jmls.2025.22.2.60

Identifying the molecular association between diabetic nephropathy and hepatocellular carcinoma: an in-silico network biology approach

Tammanna R. Sahrawat

, Arnesh Saxena

Centre for Systems Biology and Bioinformatics, UIEAST, Panjab University, Chandigarh, India

Correspondence to Tammanna R. Sahrawat Centre for Systems Biology and Bioinformatics, UIEAST, Panjab University, Sector 14, Chandigarh 160014, India
Tel: 91-172-2534603 E-mail: tammanna@pu.ac.in

Received March 3, 2025 Revised April 11, 2025 Accepted April 30, 2025

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Diabetic nephropathy (DN) and hepatocellular carcinoma (HCC) pose significant global health burdens. DN is the leading cause of end-stage renal disease and is closely associated with metabolic dysregulation, whereas HCC is the most common primary liver cancer, often arising from chronic liver conditions, including non-alcoholic fatty liver disease and cirrhosis. Although studies have indicated a clinical association between DN and HCC, the underlying molecular association remains unexplored; therefore, the present in-silico network biology study was conducted. Microarray datasets for DN and HCC were retrieved from the Gene Expression Omnibus and processed using Bioconductor Packages to identify differentially expressed genes (DEGs). For the common DEGs between DN and HCC, a protein-protein interaction (PPI) network was constructed using STRING. Topological analysis of the PPI network was performed using Cytoscape and its plug-ins, MCODE, CytoHubba, and CytoCluster, to identify hub genes, and their functional enrichment was performed using Enrichr Knowledge Graph. From the 67 common DEGs, five hub genes were identified, namely, APOA2, APOA5, APOC1, APOC3, and APOH, which belong to the apolipoprotein family. Kyoto Encyclopedia of Genes and Genomes enrichment analysis identified their involvement in cholesterol metabolism and proliferator-activated receptors (PPAR) signaling pathway, emphasizing their potential significance in lipid and other metabolic processes. Dysregulation of apolipoproteins results in altered lipid metabolism and oxidative stress in both diabetic complications and cancer progression. In conclusion, the hub genes identified in the present study can be further explored as promising biomarkers for prognosis and diagnosis in DN and HCC, as well as therapeutic interventions.

Key words: Diabetic nephropathies; Carcinoma, hepatocellular; Apolipoproteins; Comorbidity; Lipid metabolism

INTRODUCTION

Type 2 diabetes (T2D) is a multi-systemic metabolic disease characterized by hyperglycemia resulting from insufficient insulin secretion or insulin resistance. T2D is a global health challenge, giving rise to complications that affect multiple organ systems. In 2021, the International Diabetes Federation estimated that globally, 10.5% of individuals aged between 20 to 79 years have diabetes, and this rate is expected to increase to 12.2% by 2045, and globally more than half a billion individuals living with diabetes [1].

Chronic hyperglycemia in T2D leads to metabolic dysregulation, oxidative stress, macrovasculopathy (such as coronary artery disease, myocardial infarction, and congestive heart failure), and microvasculopathy (such as retinopathy, neuropathy, and nephropathy), which progressively damage tissues and organs, among other complications, diabetic nephropathy (DN), which is a primary cause of end-stage renal disease, is one of the main complication of T2D, with approximately 30-40% of individuals of T2D being diagnosed with DN. T2D progresses to DN via the production of reactive oxygen species, resulting in glomerular hyperfiltration, albuminuria, and fibrosis [2].

T2D increases the risk of hepatocellular carcinoma (HCC) because the liver plays an important role in glucose metabolism. T2D is an independent risk factor for HCC in patients with diabetes, conferring a 1.8- to 4-fold increased risk [3]. T2D can also lead to the development of non-alcoholic fatty liver disease (NAFLD), a condition characterized by lipid accumulation, hepatic inflammation, and fibrosis, which can progress to non-alcoholic steatohepatitis (NASH) and HCC [4].

HCC is the most common primary liver cancer and frequently results from chronic liver disease and cirrhosis. HCC is the ninth leading cause of cancer-related death. Cirrhosis is the most important risk factor for HCC, whereas hepatitis B and C remain independent risk factors. HCC is an aggressive cancer that occurs in patients with cirrhosis and is commonly detected in advanced stages [3,5].

Several studies have reported a direct relationship between T2D and DN as well as between T2D and HCC. Notably, PTPRC, CD53, IRF8, IL10RA, and LAPTM5 genes reportedly play essential roles in the molecular mechanics of DN as a complication of T2D [6]. In an integrated bioinformatics analysis study, Maddah et al. [7] reported BUB1, CDCA8, DLGAP5, ASPM, POL1, CENPE, WDHD1, HELLS, TRIP13, and DEPDC1 genes as potential biomarkers linking T2D and HCC.

DN and HCC are prominent health issues worldwide and associated with high morbidity and mortality rates. Emerging studies have suggested a bidirectional relationship between DN and HCC, and a recent study showed an increased risk of HCC in patients with chronic kidney disease, including DN [8]. Furthermore, lipid dysregulation, oxidative stress, and chronic inflammation are shared mechanisms driving the progression of both conditions, creating a pathological link between DN and HCC [2,9].

Most genetic, transcriptomics, and bioinformatics studies have focused on the direct association of T2D with DN or T2D with HCC, leaving a gap in the understanding of the genes and pathways common in the co-occurrence of DN and HCC. To address this issue, we conducted an in-silico bioinformatics study to explore the molecular basis of the association and potential biomarkers linking DN and HCC.

Methods

The schematic layout in Fig. 1 illustrates the methodology employed to identify the molecular associations between DN and HCC.

1. Data retrieval

Microarray datasets of gene expression profiles for DN and HCC were obtained from the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) using the GEO query package in R programming language (R Foundation, Vienna, Austria) [10-12].

2. Pre-processing of microarray data and differentially expressed gene identification

The two datasets were preprocessed: the steps included for data normalization, selection, and filtering, followed by Benjamini and Hochberg statistical analysis to control the false discovery rate and identification of differentially expressed genes (DEGs) using the limma package in R Studio (R Foundation) [13]. To identify DEGs, adjusted P-value of <0.05 and log2 FC of ±1 were used. Volcano plots were constructed and visualized using ggplot2 package in R Studio [14].

DEGs obtained for DN and HC were analyzed using Venny, an online tool that generates Venn diagrams to visualize the intersections between both datasets to identify common DEGs [15].

3. Protein-protein interaction (PPI) network construction, analysis, and gene enrichment

The STRING web-based visualization software was used to construct and analyze the PPI network [16], which was further analyzed using Cytoscape [17]. For cluster analysis, MCODE and CytoCluster were used, and functional modules were identified using the CytoHubba plug-in for Cytoscape [18-20]. To obtain an integrated network of the top enriched terms from multiple libraries connected to overlapping genes, Enrichr Knowledge Graph (Enrichr-KG) was employed [21].

RESULTS

1. Data retrieval and identification of DEGs

The microarray datasets for HCC and DN had 1,368 and 481 DEGs (Table 1), respectively, which were visualized using volcano plots (Fig. 2A, B). A Venn diagram was constructed to identify the 67 common DEGs between DN and HCC (Fig. 2C).

2. PPI Network construction and analysis

A PPI network was constructed for common DEGs using STRING and consisted of 126 nodes and 554 edges with a high confidence interaction score (Fig. 3). The network was imported into Cytoscape, and its plugin MCODE was used to detect highly interconnected clusters that represent important subnetworks within the larger PPI network (Table 2). Seven clusters were obtained, of which two highly interconnected networks (based on a significant number of nodes and edges) were merged (Fig. 4A). The resulting network was analyzed using CytoHubba to rank the genes according to significance based on the maximal clique centrality (MCC) method. The top 10 hub genes identified were APOM, APOA5, APOC3, APOA4, PON1, APOA2, APOH, APOC11, APOC4, and APOF (Fig. 4C), which were further validated using the ClusterONE algorithm of CytoCluster. Nine clusters were obtained, of which four with a P-value <0.05 were merged and further analyzed using CytoHubba (Table 2; Fig. 4B). Subsequently, APOC3, APOC2, APOH, CETP, APOA5, PLTP, APOA2, APOC1, LCAT, and APOA1 were obtained as the top 10 hub genes (Fig. 4D). Upon performing a cross-comparison between hub genes from MCODE and CytoCluster, we identified five common hub genes: APOA2, APOA5, APOC1, APOC3, and APOH.

3. Gene set enrichment analysis

The five common genes in the Enrichr-KG analysis belonged to the apolipoprotein (APO) family. Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis revealed APOA2, APOH, APOC1, and APOC3 were members of the cholesterol metabolism pathway, whereas APOC3, and APOA2 along with APOA5 were also part of the proliferator-activated receptors (PPAR) signaling pathway (Fig. 5).

DISCUSSION

Five hub genes, namely, APOA2, APOA5, APOC1, APOC3, and APOH were identified as common hub genes in the analysis of DEGs associated with DN and HCC. All hub genes belong to the APO family of proteins with molecular weights between 7 kDa and 40 kDa. These are multifunctional proteins that serve as templates for the assembly of lipoprotein particles, maintaining their structure and directing their metabolism by binding to membrane receptors and regulating enzyme activity [22]. The abnormal functioning of APOs causes dyslipidemia, and evidence suggests that dyslipidemia plays an important role in kidney disease progression in patients with diabetes [23]. Additionally, APOs are reportedly associated with tumors [24]. Notably, dyslipidemia is known to play a critical role in the pathogenesis of human malignancies such as colorectal cancer, pancreatic cancer, and HCC [25].

The APOC subfamily has been associated with HCC as APOC1, and APOC3 were expressed differently in tumor and non-tumor tissues in HCC, with APOC1 being associated with overall survival, whereas APOC3 was associated with both overall survival and recurrence-free survival [26]. APOC1 activates lecithin cholesterol acyltransferase, thus inhibiting plasma phospholipase A2, cholesteryl ester transfer protein, lipoprotein lipase, and hepatic lipase and playing important roles in lipid metabolism [27]. One study identified APOC1 as a core secretory gene in DN. The elevated expression levels in both the serum and renal tissues of DN patients contributes to DN progression, particularly proteinuria and reduced glomerular filtration rate therefore it could serve as a potential diagnostic biomarker and a therapeutic target for DN [28]. A study also reported that APOC1 was highly expressed in macrophages in HCC lesions [29].

APOC3 plays an important role in triglyceride transport and homeostasis, and its elevated levels contribute to lipid accumulation in the kidneys, exacerbating renal damage and worsening DN. Higher plasma APOC3 levels present a greater risk of renal function loss in patients with T2D. Cervantes et al. [30] reported that targeting APOC3 could help treat both DN and the associated atherosclerosis. APOA2, which is produced in the liver, is a key component of high-density lipoprotein (HDL) particles and regulates HDL structure and function. APOA2 has been reported as a biomarker in HCC and prostate cancer. It is also a marker linking dyslipidemia with the risk of kidney stones in humans and other animals. Single nucleotide polymorphism in the APOA2 promoter is reportedly associated with insulin resistance, a risk factor for T2D, diabetic kidney disease, and NASH [31,32].

APOA5 is a key regulator of plasma triglyceride levels that impacts the metabolism of triglyceride-rich lipoproteins. It also interacts with PPARα, a regulator of lipid metabolism, which influences APOA5 expression. APOA5 deficiency leads to increased plasma triglycerides and metabolic disorders. Plasma APOA5 levels are lower in patients with DN, and its polymorphisms (1131T>C and T1259>C) are associated with lipid metabolism and blood triglyceride levels, which may influence DN development [27]. While there are no direct studies linking APOA5 to HCC, its role in triglyceride regulation and liver metabolism as well as its association with NAFLD pathogenesis suggests its potential involvement in liver cancer development [33,34].

APOH encodes β2GPI, a protein essential for binding phospholipids and various negatively charged substances, playing roles in coagulation, fibrinolysis, placental homeostasis, endothelial cell activation, and apoptosis [35]. APOH is primarily expressed in the renal proximal tubules, filtered through the glomeruli, and reabsorbed into renal epithelial cells [27]. Transcriptome analysis in DN revealed downregulated APOH in diabetic tubules, with increased urinary β2GPI excretion indicating impaired tubular protein reabsorption in diabetes patients without clinical proteinuria [36]. In HCC, APOH expression was reported to be lower than that in normal tissues, and low APOH expression was correlated with poor prognosis (P<0.05) [37].

The results of the present study suggested that the APO family genes, namely APOA2, APOA5, APOC1, APOC3, and APOH, may be responsible for the molecular association between DN and HCC. The potential roles of these genes as biomarkers and therapeutic targets should be further explored to prevent one condition from leading to the other. Although our study aimed to map key genes and pathways between DN and HCC, our results were observational, based on experimental microarray datasets, and have not been validated in clinical or experimental studies. Nevertheless, our study provides valuable insights into the development of common therapeutic strategies for DN and HCC using shared molecular pathways and genes.

To the best of our knowledge, this is the first study to investigate the molecular association between DN and HCC. Our results reveal the shared significant molecular pathways between DN and HCC, primarily involving genes from the APO family, which are central to lipid metabolism and play crucial roles in both diseases. These genes may influence disease mechanisms through their effect on lipid transport and metabolic regulation, and lipid dysregulation may be a connecting factor in the pathogenesis of DN and HCC.

The identified APO genes may be potential biomarkers for the diagnosis and prognosis of both DN and HCC and may be the targets for novel therapeutic approaches to mitigate disease progression in both conditions.

Notes

ACKNOWLEDGEMENTS

The authors would like to thank their parent institute Panjab University Chandigarh, India for providing the infrastructure for carrying out the research.

CONFLICT OF INTEREST

The authors report no conflict of interest.

FUNDING

The authors did not receive any funding for this study.

Figure 1.

A comprehensive overview of the research methodology employed in this study, illustrating the step-by-step process. PPI: protein- protein interaction, DEGs: differentially expressed genes, DN: diabetic nephropathy, HCC: hepatocellular carcinoma, GEO: Gene Expression Omnibus, Enrichr-KG: Enrichr Knowledge Graph.

Figure 2.

Volcano plots of the (A) GSE29721 (HCC) and (B) GSE30528 (DN) datasets showing distribution of DEGs. Red indicates up-regulated genes, blue denotes down-regulated genes, and black represents non-significant genes. (C) Identification of 67 common DEGs between DN (yellow) and HCC (purple) obtained from Venny tool. HCC: hepatocellular carcinoma, DN: diabetic nephropathy, DEGs: differentially expressed genes.

Figure 3.

Protein-protein interaction (PPI) network constructed using the STRING database with a high-confidence interaction score of 0.900. The network consists of 126 nodes and 554 edges, representing significant interactions between the identified differentially expressed genes (DEGs). Nodes represent proteins, whereas edges denote predicted functional associations between proteins.

Figure 4.

Integrated network analysis and hub gene identification. (A) Network obtained after merging the two clusters identified by MCODE from the protein-protein interaction (PPI) network. (B) Network obtained after integrating the four clusters identified by CytoCluster from the PPI network. (C) Top 10 hub genes identified using CytoHubba (MCC algorithm) applied to the subnetwork derived from MCODE. (D) Top 10 hub genes identified using CytoHubba (MCC algorithm) applied to the subnetwork derived from CytoCluster. The networks were visualized in Cytoscape, with nodes representing genes and edges representing PPIs. Hub genes identified by CytoHubba are highlighted in red to yellow to indicate their significance within each subnetwork. MCC, maximal clique centrality.

Figure 5.

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of the top five hub genes identified from the protein-protein interaction (PPI) network, performed using Enrichr Knowledge Graph (Enrichr-KG). The analysis highlights significantly enriched pathways associated with the hub genes, providing insights into their biological functions.

Table 1.

Concise overview of the datasets and dysregulated genes

Disease	Dataset	Normal samples	Patient samples	Sample tissue	Dysregulated genes	Up-regulated genes	Down-regulated genes
Diabetic nephropathy	GSE30528	13	9	Kidney	481	355	126
Hepatocellular carcinoma	GSE29721	10	10	Liver	1,368	662	706

Table 2.

Highly connected subnetworks obtained from MCODE and Cytocluster

REFERENCES

1. Sun H, Saeedi P, Karuranga S, Pinkepank M, Ogurtsova K, Duncan BB, et al. IDF diabetes atlas: global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract 2022;183:109119.

2. Wang N, Zhang C. Oxidative stress: a culprit in the progression of diabetic kidney disease. Antioxidants (Basel) 2024;13:455.

3. Balogh J, Victor D 3rd, Asham EH, Burroughs SG, Boktour M, Saharia A, et al. Hepatocellular carcinoma: a review. J Hepatocell Carcinoma 2016;3:41-53.

4. Chalasani N, Younossi Z, Lavine JE, Charlton M, Cusi K, Rinella M, et al. The diagnosis and management of nonalcoholic fatty liver disease: practice guidance from the American Association for the Study of Liver Diseases. Hepatology 2018;67:328-57.

5. Llovet JM, Zucman-Rossi J, Pikarsky E, Sangro B, Schwartz M, Sherman M, et al. Hepatocellular carcinoma. Nat Rev Dis Primers 2016;2:16018.

6. Guo M, Dai Y, Jiang L, Gao J. Bioinformatics analysis of the mechanisms of diabetic nephropathy via novel biomarkers and competing endogenous RNA network. Front Endocrinol (Lausanne) 2022;13:934022.

7. Maddah R, Shariati P, Arabpour J, Bazireh H, Shadpirouz M, Kafraj AS. Identification of critical genes and pathways associated with hepatocellular carcinoma and type 2 diabetes mellitus using integrated bioinformatics analysis. Inform Med Unlocked 2022;30:100956.

8. Yeh H, Chiang CC, Yen TH. Hepatocellular carcinoma in patients with renal dysfunction: pathophysiology, prognosis, and treatment challenges. World J Gastroenterol 2021;27:4104-42.

9. Buechler C, Aslanidis C. Role of lipids in pathophysiology, diagnosis and therapy of hepatocellular carcinoma. Biochim Biophys Acta Mol Cell Biol Lipids 2020;1865:158658.

10. Davis S, Meltzer PS. GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics 2007;23:1846-7.

11. Woroniecka KI, Park AS, Mohtat D, Thomas DB, Pullman JM, Susztak K. Transcriptome analysis of human diabetic kidney disease. Diabetes 2011;60:2354-69.

12. Stefanska B, Huang J, Bhattacharyya B, Suderman M, Hallett M, Han ZG, et al. Definition of the landscape of promoter DNA hypomethylation in liver cancer. Cancer Res 2011;71:5891-903.

13. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47.

14. Tyner S, Briatte F, Hofmann H. Network visualization with ggplot2. R J 2017;9:27-59.

15. Oliveros JC. VENNY. An interactive tool for comparing lists with Venn diagrams [Internet]. Barcelona: BioinfoGP; c2024 [cited 2025 Feb 13]. Available from: https://bioinfogp.cnb.csic.es/tools/venny/.

16. von Mering C, Huynen M, Jaeggi D, Schmidt S, Bork P, Snel B. STRING: a database of predicted functional associations between proteins. Nucleic Acids Res 2003;31:258-61.

17. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 2003;13:2498-504.

18. Bader GD, Hogue CW. An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics 2003;4:2.

19. Li M, Li D, Tang Y, Wu F, Wang J. CytoCluster: a cytoscape plugin for cluster analysis and visualization of biological networks. Int J Mol Sci 2017;18:1880.

20. Chin CH, Chen SH, Wu HH, Ho CW, Ko MT, Lin CY. cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol 2014;8 Suppl 4:S11.

21. Evangelista JE, Xie Z, Marino GB, Nguyen N, Clarke DJB, Ma'ayan A. Enrichr-KG: bridging enrichment analysis across multiple libraries. Nucleic Acids Res 2023;51:W168-79.

22. Kim J, Basak JM, Holtzman DM. The role of apolipoprotein E in Alzheimer's disease. Neuron 2009;63:287-303.

23. Rutledge JC, Ng KF, Aung HH, Wilson DW. Role of triglyceride-rich lipoproteins in diabetic nephropathy. Nat Rev Nephrol 2010;6:361-70.

24. Mendivil CO, Zheng C, Furtado J, Lel J, Sacks FM. Metabolism of very-low-density lipoprotein and low-density lipoprotein containing apolipoprotein C-III and not other small apolipoproteins. Arterioscler Thromb Vasc Biol 2010;30:239-45.

25. Huang HT, Zhang XY, Zhang C, Ling Q, Zheng SS. Predicting dyslipidemia after liver transplantation: a significant role of recipient metabolic inflammation profile. World J Gastroenterol 2020;26:2374-87.

26. Darwish NM, Al-Hail MK, Mohamed Y, Al Saady R, Mohsen S, Zar A, et al. The role of apolipoproteins in the commonest cancers: a review. Cancers (Basel) 2023;15:5565.

27. Pan X. The roles of fatty acids and apolipoproteins in the kidneys. Metabolites 2022;12:462.

28. Yu K, Li S, Wang C, Zhang Y, Li L, Fan X, et al. APOC1 as a novel diagnostic biomarker for DN based on machine learning algorithms and experiment. Front Endocrinol (Lausanne) 2023;14:1102634.

29. Gao J, Li Z, Lu Q, Zhong J, Pan L, Feng C, et al. Single-cell RNA sequencing reveals cell subpopulations in the tumor microenvironment contributing to hepatocellular carcinoma. Front Cell Dev Biol 2023;11:1194199.

30. Cervantes J, Koska J, Kramer F, Akilesh S, Alpers CE, Mullick AE, et al. Elevated apolipoprotein C3 augments diabetic kidney disease and associated atherosclerosis in type 2 diabetes. JCI Insight 2024;9:e177268.

31. Malik G, Ward MD, Gupta SK, Trosset MW, Grizzle WE, Adam BL, et al. Serum levels of an isoform of apolipoprotein A-II as a potential marker for prostate cancer. Clin Cancer Res 2005;11:1073-85.

32. Bougarne N, Weyers B, Desmet SJ, Deckers J, Ray DW, Staels B, et al. Molecular actions of PPARα in lipid metabolism and inflammation. Endocr Rev 2018;39:760-802.

33. Xu Y, Lu LL, Liu SS, Du SX, Zhu HL, Dong QJ, et al. Apolipoprotein A5 gene polymorphisms are associated with non-alcoholic fatty liver disease. Hepatobiliary Pancreat Dis Int 2018;17:214-9.

34. Geh D, Anstee QM, Reeves HL. NAFLD-associated HCC: progress and opportunities. J Hepatocell Carcinoma 2021;8:223-39.

35. Miyakis S, Giannakopoulos B, Krilis SA. Beta 2 glycoprotein I--function in health and disease. Thromb Res 2004;114:335-46.

36. Lapsley M, Flynn FV, Sansom PA. Beta 2-glycoprotein-1 (apolipoprotein H) excretion and renal tubular malfunction in diabetic patients without clinical proteinuria. J Clin Pathol 1993;46:465-9.

37. Li X, Wang L, Wang L, Feng Z, Peng C. Single-cell sequencing of hepatocellular carcinoma reveals cell interactions and cell heterogeneity in the microenvironment. Int J Gen Med 2021;14:10141-53.