Guanidinylated Apolipoprotein C3 (ApoC3) Associates with Kidney and Vascular Injury

Abstract

Background: Coexistent CKD and cardiovascular diseases are highly prevalent in Western populations and account for substantial mortality. We recently found that apolipoprotein C-3 (ApoC3), a major constituent of triglyceride-rich lipoproteins, induces sterile systemic inflammation by activating the NOD-like receptor protein-3 (NLRP3) inflammasome in human monocytes via an alternative pathway.

Methods: To identify posttranslational modifications of ApoC3 in patients with CKD, we used mass spectrometry to analyze ApoC3 from such patients and from healthy individuals. We determined the effects of posttranslationally modified ApoC3 on monocyte inflammatory response in vitro, as well as in humanized mice subjected to unilateral ureter ligation (a kidney fibrosis model) and in a humanized mouse model for vascular injury and regeneration. Finally, we conducted a prospective observational trial of 543 patients with CKD to explore the association of posttranslationally modified ApoC3 with renal and cardiovascular events in such patients.

Results: We identified significant posttranslational guanidinylation of ApoC3 (gApoC3) in patients with CKD. We also found that mechanistically, guanidine and urea induce guanidinylation of ApoC3. A 2D-proteomic analysis revealed that gApoC3 accumulated in kidneys and plasma in a CKD mouse model (mice fed an adenine-rich diet). In addition, gApoC3 augmented the proinflammatory effects of ApoC3 in monocytes in vitro . In humanized mice, gApoC3 promoted kidney tissue fibrosis and impeded vascular regeneration. In CKD patients, higher gApoC3 plasma levels (as determined by mass spectrometry) were associated with increased mortality as well as with renal and cardiovascular events.

Conclusions: Guanidinylation of ApoC3 represents a novel pathogenic mechanism in CKD and CKD-associated vascular injury, pointing to gApoC3 as a potential therapeutic target.

Graphical abstract

Figure 1.

ApoC3 from CKD patients is guanidinylated. (A) Representative MALDI-TOF mass spectra of nApoC3 (i.e., unmodified ApoC3) and ApoC3 from a healthy subject and a patient with CKD. Red arrows indicate guanidinylated lysine 44. (B) gApoC3 mass-signal intensity in n=543 subjects of patients with CKD divided according to KDIGO categories of eGFR. (C) Characteristic MALDI-TOF mass spectra of plasma incubated with methylisourea, guanidine, or urea. (D) Three-dimensional structure of human ApoC3 with lysine residues marked, at which posttranslational guanidinylation was detected. (E) Guanidinylation mass-signal intensity after incubation with increasing concentrations of urea (n=4 per group). (F) Chemical pathways inducing guanidinylation of lysine residues in ApoC3.

Figure 2.

gApoC3 accumulates in mice with CKD. (A) Outline of the adenine-induced CKD mouse model. (B) Serum creatinine of mice after 2 weeks of adenine (AD) or standard diet (SD, n=6 per group). (C) Serum urea of mice after 2 weeks of AD or SD (n=6 per group). (D) Representative 2D-proteomic images showing the gApoC3 fragment FLK*GYWSK* (m/z 1113.6) in the kidneys of three individual mice subjected to AD or SD for 2 weeks. (E) Number of guanidinylated lysine residues of ApoC3 in plasma of mice subjected to AD or SD for 2 weeks (n=6 per group). (F) Representative mass fingerprint spectrum of tryptic-digested SDS-gel piece of plasma from mice after 2 weeks of AD. The arrows indicate unmodified peptide “FLKGYWSK” (1027 Da) and the modified peptide “FLK*GYWSK*” (1113 Da) of ApoC3. (G) Representative MALDI-TOF/TOF fragment mass spectrum of guanidinylated peptide of ApoC3 amino acid sequence FLK*GYWSK* (71,76) from mice after 2 weeks of AD.

Figure 3.

gApoC3 induces inflammation. (A) Release of IL-1β in the supernatant of human monocytes incubated with LPS (10 ng/ml), nApoC3, gApoC3 guanidinylated with increasing concentrations of methylisourea (1 mM, 10 mM, 100 mM, 1 M), or carbamylated ApoC (cApoC3, 50 µg/ml, 16 hours, n=6 per group). (B) Release of IL-6 in the supernatant of human monocytes incubated with LPS (10 ng/ml), nApoC3, gApoC3 guanidinylated with increasing concentrations of methylisourea (1 mM, 10 mM, 100 mM, 1 M), or cApoC3 (50 µg/ml, 16 hours, n=6 per group). (C) Superoxide production of human monocytes incubated with LPS (10 ng/ml), nApoC3, gApoC3 guanidinylated with increasing concentrations of methylisourea (1 mM, 10 mM, 100 mM, 1 M), or cApoC3 (50 µg/ml, 16 hours, n=6 per group). (D) Binding of Atto-488–labeled nApoC3 or gApoC3 (1 hour) to human monocytes as determined by flow cytometry. Representative of at least three independent experiments. (E) Internalization of Atto-488–labeled nApoC3 or gApoC3 (1 hour) into human monocytes. Representative of at least three independent experiments. (F) Fluorescence intensity of HEK293T cells transfected with Tlr2 or Tlr4 constructs incubated with Atto-488–labeled nApoC3 or gApoC3 for 1 hour. Representative of at least three independent experiments. Data are represented as mean±SEM.

Figure 4.

gApoC3 induces kidney injury in humanized mice. (A) Schematic of the unilateral ureter ligation model. (B and C) Kidney fibrosis as determined by Sirius red staining in kidneys of NOD-SCID mice 5 days after unilateral ureter ligation and injection with nApoC3 or gApoC3 guanidinylated with increasing concentrations of methylisourea (10 mM or 1 M, 50 µg/ml blood volume, n=4–6 per group). (D) Renal expression of F4/80 5 days after unilateral ureter ligation (n=4–6 per group). (E) Renal expression of Ly6G 5 days after unilateral ureter ligation (n=4–6 per group). (F) Renal expression of DKK3 5 days after unilateral ureter ligation (n=4–6 per group).

Figure 5.

gApoC3 induces vascular injury in humanized mice. (A) Schematic of the perivascular carotid injury model. (B and C) Re-endothelialized area at 1 week after perivascular carotid injury in NOD-SCID mice injected with nApoC3 or gApoC3 guanidinylated with increasing concentrations of methylisourea (10 mM or 1 M, 50 µg/ml blood volume, n=6 per group). Data are represented as mean±SEM.

Figure 6.

Guanidinylation of ApoC3 is associated with inflammation in patients with CKD. (A–G) Plasma concentrations of total cholesterol, LDL-cholesterol (LDL-C), HDL-cholesterol (HDL-C), triglycerides, ApoC3, high-sensitivity C-reactive protein (hsCRP), and serum amyloid A (SAA) according to tertiles of gApoC3 intensity (n=543). (H) gApoC3 intensity in various subgroups of participants of the prospective CARE FOR HOMe study (n=543).

Figure 7.

gApoC3 associates with renal outcomes in patients with CKD. (A) Number of participants of the CARE FOR HOMe study. (B) Restricted cubic spline plots of the association between gApoC3 mass-signal intensity and the risk of the combined renal end point in 543 patients with CKD included in the prospective CARE FOR HOMe study. The solid line indicates the risk for the combined renal end point with respective 95% CIs (light-gray area). Spikes show the individual distribution of gApoC3 mass-signal intensity. (C) Group-based trajectory modeling change of eGFR in 543 participants of the prospective CARE FOR HOMe study, which identified four distinct groups of patients (group 1, stable eGFR; group 2, declining eGFR; group 3, rapidly declining eGFR; group 4, increasing eGFR). Gray-shaded area represents 95% confidence intervals. (D) Association of gApoC3 mass-signal intensity and eGFR trajectory groups. All plots are adjusted for age, sex, prevalent CVD, body mass index, systolic BP, smoking status, diabetes mellitus, eGFR, and albuminuria.

Figure 8.

gApoC3 associates with cardiovascular outcomes in patients with CKD. (A) Restricted cubic spline plots of the association between gApoC3 mass-signal intensity and the risk of the combined cardiovascular end point in 543 patients with CKD included in the prospective CARE FOR HOMe study. The solid line indicates the risk for the combined cardiovascular end point with respective 95% CIs (light-gray area). Spikes show the individual distribution of gApoC3 mass-signal intensity. (B) Restricted cubic spline plots of the association between gApoC3 mass-signal intensity and the risk of mortality in 543 patients with CKD included in the prospective CARE FOR HOMe study. The solid line indicates the risk for mortality with respective 95% CIs (light-gray area). Black spikes show the individual distribution of gApoC3 mass-signal intensity.