The role of IGFBP-1 in the clinical prognosis and pathophysiology of acute kidney injury

Abstract

The ability to predict progression to severe acute kidney injury (AKI) remains an unmet challenge. Contributing to the inability to predict the course of AKI is a void of understanding of the pathophysiological mechanisms of AKI. The identification of novel prognostic biomarkers could both predict patient outcomes and unravel the molecular mechanisms of AKI. We performed a multicenter retrospective observational study from a cohort of patients following cardiac surgery. We identified novel urinary prognostic biomarkers of severe AKI among subjects with early AKI. Of 2,065 proteins identified in the discovery cohort, insulin-like growth factor binding protein 1 (IGFBP-1) was the most promising. We validated IGFBP-1 as a prognostic biomarker of AKI in 213 patients. In addition, we investigated its role in the pathophysiology of AKI using a murine model of cisplatin-induced AKI (CIAKI). Urinary IGFBP-1 concentration in samples collected from patients with stage 1 AKI following cardiothoracic surgery was significantly higher in patients who progressed to severe AKI compared with patients who did not progress beyond stage 1 AKI (40.28 ng/ml vs. 2.8 ng/ml, P < 0.0001) and predicted the progression to the composite outcome (area under the curve: 0.85, P < 0.0001). IGFBP-1 knockout mice showed less renal injury, cell death, and apoptosis following CIAKI, possibly through increased activation of the insulin growth factor receptor 1. IGFBP-1 is a clinical prognostic biomarker of AKI and a direct mediator of the pathophysiology of AKI. Therapies that target the IGFBP-1 pathways may help alleviate the severity of AKI.NEW & NOTEWORTHY The ability to predict progression to severe AKI remains an unmet challenge. Early prognostic biomarkers of AKI hold promise to improve patient outcomes by early implementation of clinical therapy, as well as unravel the pathophysiological mechanisms of AKI. Here, we present a novel urinary biomarker, IGFBP-1, that predicts the progression to severe AKI following cardiac surgery. In addition, we show that IGFBP-1 mice are protected against CIAKI, suggesting a mechanistic role for IGFBP-1 in AKI.

Keywords: IGFBP-1; acute kidney injury; biomarker; insulin-like growth factor binding protein 1.

Figure 1:. Flowchart of patient enrollment in the validation study.

Samples from 225 patient with stage 1 AKI were collected within 72 hours of cardiothoracic surgery. Twelve samples were insufficient for analysis, leaving a total of 213 patients for analysis of which 27 patients met criteria for inclusion in the composite primary outcome group.

Figure 2:. IGFBP-1 abundance and peptide sequences in the discovery proteomic analysis.

The relative abundance of spectra for IGFBP-1 (panel A) is shown for urine samples from twenty control subjects that did not progress beyond AKI stage 1 (−) and ten subjects that required renal replacement therapy (+). Quantitative value is a normalized representation of spectral counts. Each dot represents the normalized abundance of IGFBP-1 in urine for one subject. The number of peptides observed for each peptide sequence in control subjects and subjects that required renal replacement therapy are presented in panel B.

Figure 3:. Urinary concentrations of urinary biomarkers.

Shown are the measured concentrations of urinary IGFBP-1 (panel A), [TIMP-2]*[IGFBP7] (panel C), and NGAL (panel D) for patients the met the primary outcome compared to controls. Each individual data point represents a measured value (IGFBP-1 n = 213, [TIMP2]*[IGFBP7] n = 154, NGAL n = 150). Urinary concentrations of IGFBP-1 separated by stage of AKI and each individual inclusion criteria of the composite outcome are shown in (panel B). The horizontal bars represent the median value, and the surrounding boxes represent the interquartile range. The range of data for the individual groups is as follows (Stage 1: 0.111 – 439.5; Stage 2: 0.069 – 289.6; Stage 3 2.307 – 1160; RRT: 2.307 – 673.9; 30-day mortality: 4.992 – 1160). Please note that some patients qualified for primary outcome by more than one inclusion criteria. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns = not significant by unpaired T-test (Panels A, C, D) or one-way ANOVA (Panel B).

Figure 4:. Receiver operating characteristics (ROC) for biomarkers of AKI.

Shown in (Panel A) is the ROC comparison of urinary IGFBP-1 against the traditional clinical biomarker, creatinine. Values for the AUC, optimal cutoff, sensitivity, specificity, and likelihood ratio for IGFBP-1 are shown on the graph. The ROC for IGFBP-1 (n = 213), [TIMP-2]*[IGFBP7] (n = 154), and NGAL (n = 150) are shown in (Panel B). AUC values for each biomarker are displayed on the graph.

Figure 5:. Effect of IGFBP-1 knockout on CIAKI.

Serum creatinine values (Panel A) and blood urea nitrogen (BUN) (Panel B) measured from mouse serum following cisplatin or saline (control) injection are displayed. Each dot (control) or triangle (cisplatin injection) represent a measurement for an individual mouse. Control = saline injected wild type, Control-KO = saline injected IGFBP-1 knockout, Cisplatin-WT = cisplatin injected wild type, Cisplatin-KO = cisplatin injected IGFBP-1 knockout. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns = not significant by one-way ANOVA.

Figure 6:. Effect of IGFBP-1 KO on renal tubular cell damage during CIAKI.

Representative images of PAS staining of mice kidney tissue are shown at 200x magnification. Wild type sham mice are represented in panel (A), wild type mice that received cisplatin injection are found in panel (B), and IGFBP-1 KO mice that received cisplatin injections are represented in panel (C). Slides from each animal were scored blinding by a renal pathologist (panel D) for degree of tubular injury. Histology scoring was evaluated by Kruskal-Wallis with Dunn’s multiple comparison test. Each dot (control) or triangle (cisplatin injection) represent scoring for an individual mouse. Control = saline injected wild type, Control-KO = saline injected IGFBP-1 knockout, Cisplatin-WT = cisplatin injected wild type, Cisplatin-KO = cisplatin injected IGFBP-1 knockout. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns = not significant.

Figure 7:. Effect of IGFBP-1 KO on cell death and apoptosis during CIAKI.

Representative imagines of TUNEL staining from wild type saline injected mice (panel A), wild type cisplatin injected mice (panel B), and IGFBP-1 KO cisplatin injected mice (panel C) are shown. The percentage of TUNEL positive cells (panel D), caspase 3/7 activity in renal tissue homogenates (panel E) and BCL-xl protein concentrations of renal tissue homogenates (panel F) are shown below the representative images. Each dot (control) or triangle (cisplatin injection) represent a measurement for an individual mouse, (n = 6–12). Control = saline injected wild type, Control-KO = saline injected IGFBP-1 knockout, Cisplatin-WT = cisplatin injected wild type, Cisplatin-KO = cisplatin injected IGFBP-1 knockout. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns = not significant by one-way ANOVA.

Figure 8:. Proteomic analysis of mouse renal tissue homogenates following cisplatin injection.

Changes in the renal proteome between saline injected control mice (grey dots and columns), saline injected IGFBP-1 KO control (orange dots and columns), cisplatin injected wild type (blue dots and columns) and cisplatin injected IGFBP-1 KO mice are shown. IGFBP-1 was not detected in the analysis. Known biomarkers of AKI including NGAL and AGT. IGFBP7 was not changed. IGFBP4 was significantly decreased in cisplatin injected IGFBP-1 KO mice. AGT = angiotensinogen; IGFBP4 = insulin-like growth factor binding protein 4; IGFBP7 = insulin-like growth factor binding protein 7. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns = not significant by one-way ANOVA.

Figure 9:. Measurement of phosphorylate AKT and FAK in renal tissue homogenates.

The percentage of total AKT that is phosphorylated is displayed in (Panel A). The total concentration of p-FAK is displayed in (Panel B). Values are normalized for tissue homogenate protein concentrations. Each dot (control) or triangle (cisplatin injection) represent a measurement for an individual mouse. Control = saline injected wild type, Control-KO = saline injected IGFBP-1 knockout, Cisplatin-WT = cisplatin injected wild type, Cisplatin-KO = cisplatin injected IGFBP-1 knockout. * p < 0.05 and ns = not significant by one-way ANOVA.

Figure 10:. Schematic representation of proposed mechanism of IGFBP-1 signaling in AKI.

IGFBP-1 modulates cellular signaling in renal proximal tubular epithelial cell by one of two mechanisms: sequestration of IGF-1 or direct binding of the C-terminus RGD domain to the α₅β₁ integrin receptor. Binding of IGF-1 to its receptor activates a phosphorylation cascade through AKT via Phosphoinositide 3-kinase (PI3K) ultimately leading to increased synthesis of the anti-apoptotic BCL-xl protein. Binding of IGFBP-1 to the α₅β₁ integrin receptor activates a separate phosphorylation cascade through FAK via SRC kinases ultimately leading to increased cell migration and increased luminal pressure worsening AKI. The decreased apoptotic markers, increased BCL-xl, and increased ratio of p-AKT in our studies suggest that IGFBP-1 KO mice are protected against CIAKI due to inability to sequester IGF-1 and subsequent increased activation of the IGF-1r.