Gene signature and prediction model of the mitophagy-associated immune microenvironment in renal ischemia-reperfusion injury

Abstract

Background: Renal ischemia-reperfusion injury (IRI) is an inevitable occurrence during kidney transplantation. Mitophagy, ferroptosis, and the associated immune microenvironment (IME) have been shown to play important roles in renal IRI. However, the role of mitophagy-associated IME genes in IRI remains unclear. In this study, we aimed to construct a prediction model of IRI prognosis based on mitophagy-associated IME genes.

Method: The specific biological characteristics of the mitophagy-associated IME gene signature were comprehensively analyzed using public databases such as GEO, Pathway Unification, and FerrDb. Correlations between the expression of prognostic genes and immune-related genes and IRI prognosis were determined by Cox regression, LASSO analysis, and Pearson's correlation. Molecular validation was performed using human kidney 2 (HK2) cells and culture supernatant as well as the serum and kidney tissues of mice after renal IRI. Gene expression was measured by PCR, and inflammatory cell infiltration was examined by ELISA and mass cytometry. Renal tissue damage was characterized using renal tissue homogenate and tissue sections.

Results: The expression of the mitophagy-associated IME gene signature was significantly correlated with IRI prognosis. Excessive mitophagy and extensive immune infiltration were the primary factors affecting IRI. In particular, FUNDC1, SQSTM1, UBB, UBC, KLF2, CDKN1A, and GDF15 were the key influencing factors. In addition, B cells, neutrophils, T cells, and M1 macrophages were the key immune cells present in the IME after IRI. A prediction model for IRI prognosis was constructed based on the key factors associated with the mitophagy IME. Validation experiments in cells and mice indicated that the prediction model was reliable and applicable.

Conclusion: We clarified the relationship between the mitophagy-related IME and IRI. The IRI prognostic prediction model based on the mitophagy-associated IME gene signature provides novel insights on the prognosis and treatment of renal IRI.

Keywords: gene signature; immune microenvironment; mitophagy; prediction model; renal ischemia-reperfusion injury.

Figure 1

Heatmaps and correlation analysis of differentially expressed MRGs/FRGs. (A) Venn diagram of differentially expressed MRGs/FRGs in DBD, DCD and LD. All 6 MRGs and 6 FRGs were differentially expressed in the 3 subject populations. (B–D) Heatmaps of MRGs/FRGs expression in samples from (B) DBD, (C) DCD, and (D) LD. (E–G) Correlation matrices of MRGs/FRGs expression in DBD, DCD and LD. Red indicates positive correlation and blue indicates negative correlation. A darker indicates a higher correlation (Pearson coefficient), and a black X indicates no statistical significance. MRGs, mitophagy-related genes; FRGs, ferroptosis-related genes; DBD, brain-dead donor; DCD, cardiac dead donor; LD, living donor; IR, ischemia reperfusion.

Figure 2

Molecular subtyping. (A) Process of NMF subtype clustering. (B) Heatmap of subtypes. (C) PCA of subtypes. (D) Survival curves of subtypes. (K–M method). (E) Box plot of immune cell infiltration in each subtype. (F) Box plot of MRGs/FRGs expression in each subtype. Gene expression in the box plot was compared using the Wilcoxon test. ns, no significance, *P < 0.05, **P < 0.01. ***P < 0.001, ****<0.0001. NMF, non-negative matrix factorization; PCA, principal component analysis.

Figure 3

Risk score and evaluation. (A) ROC curves of RS in predicting the 1-, 2- and 3-year survival of kidney transplant patients. (B) Survival curves (K–M method) of the low and high RS groups. (C) Box plot of immune cell infiltration in the low and high RS groups. Data were compared using the Wilcoxon test. **P <0.01, ****P < 0.0001, ns, no significance.

Figure 4

HK2 cells under hypoxic environment would cause the occurrence of ferroptosis, and accompanied by the secretion of inflammatory factors. (A) HK2 cells were cultured in hypoxia for 24 h and reoxygenated for 4 h, and the apoptotic and necrotic ratios were determined using flow cytometry, (B) Mitochondrial membrane potential alteration was detected by JC-1 probes, Scale bar, 50 um and 25 um. (C) Inflammatory factors secreted by HK2 within cell culture supernatants were examined, (D) Cell culture supernatant were assayed for markers of ferroptosis. n = 3, *P < 0.05, **P < 0.01 between groups as indicated.

Figure 5

Renal IRI causes renal insufficiency and in mice. (A) Secrum creatinine was detected in different groups of mice and found that IR was higher than Sham group, (B) HE staining of mouse kidney tissue, can see that IR can lead to shedding necrosis and vacuole-like degeneration of tubular epithelial cells, Scale bar, 50 um and 25um. (C) Using single cell flow technology, (D) PCR to determine the expression of kidney tissue expression of MRGs and FRGs, (E) Electron microscopic photography showed significant changes in mitochondrial structure after IR. Scale bar, 200 nm. n = 6, **P < 0.01 between groups as indicated.