The bone is the major source of high circulating intact fibroblast growth factor-23 in acute murine polymicrobial sepsis induced by cecum ligation puncture

Abstract

Fibroblast growth factor-23 (FGF23), a bone-produced hormone, plays a critical role in mineral homeostasis. Human diseases associated with excessive intact circulating FGF23 (iFGF23) result in hypophosphatemia and low vitamin D hormone in patients with normal kidney function. In addition, there is accumulating evidence linking FGF23 with inflammation. Based on these studies and the frequent observation of hypophosphatemia among septic patients, we sought to elucidate further the relationship between FGF23 and mineral homeostasis in a clinically relevant murine polymicrobial sepsis model. Medium-severity sepsis was induced by cecum ligation puncture (CLP) in adult CD-1 mice of both sexes. Healthy CD-1 mice (without CLP) were used as controls. Forty-eight hours post-CLP, spontaneous urine was collected, and serum, organs and bones were sampled at necropsy. Serum iFGF23 increased ~20-fold in CLP compared to control mice. FGF23 protein concentration was increased in the bones, but not in spleen or liver of CLP mice. Despite the ~20-fold iFGF23 increase, we did not observe any significant changes in mineral homeostasis or parathyroid hormone levels in the blood of CLP animals. Urinary excretion of phosphate, calcium, and sodium remained unchanged in male CLP mice, whereas female CLP mice exhibited lower urinary calcium excretion, relative to healthy controls. In line with renal FGF23 resistance, expression of phosphate-, calcium- and sodium-transporting proteins did not show consistent changes in the kidneys of male and female CLP mice. Renal expression of the co-receptor αKlotho was downregulated in female, but not in male CLP mice. In conclusion, our data demonstrate that the dramatic, sex-independent rise in serum iFGF23 post-CLP was mainly caused by an upregulation of FGF23 secretion in the bone. Surprisingly, the upsurge in circulating iFGF23 did not alter humoral mineral homeostasis in the acutely septic mice. Hence, the biological function of elevated FGF23 in sepsis remains unclear and warrants further studies.

Fig 1. CLP-induced increase in plasma C-terminal and intact FGF23 levels.

Plasma C-terminal and iFGF23 concentration in females (A) and males (B) increased post-CLP relative to baseline (BL; 24 h prior CLP). Each bar is the mean ± SD of 5–15 mice per group. Each symbol represents an individual sample. Mixed model approach with Geisser-Greenhouse correction, followed byTukey post hoc test.*, P<0.05 vs. baseline within same sex.

Fig 2. CLP-induced increase in serum iFGF23 levels, 48 h after surgery.

Serum iFGF23 concentration increased profoundly in CLP mice of both sexes, relative to healthy controls. Each bar is the mean ± SD of 4–6 mice per group. Each symbol represents an individual sample. Inset shows results from two-way ANOVA. *, P<0.05 vs. healthy controls within same sex, Student’s t-test with Welch corrections.

Fig 3. iFGF23 protein expression in spleen, liver and bone after CLP.

(A) Splenic, (B) hepatic and (C) bony iFGF23 protein expression measured in lysates from spleen, liver and bone by ELISA, 48 h post-surgery (n = 4–5 per group). Data are mean ± SD. Each symbol represents an individual sample. Insets show results from two-way ANOVA. *, P < 0.05 vs. healthy controls within same sex by Student’s t-test with Welch correction.

Fig 4. Circulating PTH concentration, serum creatinine, and mineral homeostasis in septic CLP mice.

(A) Serum PTH level (healthy control female and male: n = 4 each; CLP female: n = 8, CLP male: n = 14), (B) serum creatinine (Crea), (C) serum phosphorus (P_i), sodium (Na) and calcium (Ca) levels (n = 4–9), (D) urinary phosphorus/creatinine (Ur P/Crea), sodium/creatinine (Ur Na/Crea) and calcium/creatinine (Ur Ca/Crea) (healthy control female and male: n = 4 each, CLP female: n = 11, CLP male: n = 22–23) in control and CLP mice, 48 h post-surgery. Bars are mean ± SD. Each symbol represents an individual sample. Inset shows results from two-way ANOVA. *, P<0.05 vs. healthy controls within same sex, Student’s t-test with Welch correction.

Fig 5. Effects of CLP-induced sepsis on the renal expression of phosphate-, sodium- and calcium-transporting proteins and of Klotho.

Quantification and original western blot images of Napi2a, NCC, TRPV5 and Klotho protein expression detected in renal total membrane fractions of (A) female and (B) male healthy control and CLP mice, 48 h post-surgery. Each bar represents the mean ± SD of 4–5 mice per experimental group. Each symbol represents the mean of an individual sample measured on two membranes. *, P<0.05 vs. healthy control, Student’s t-test with Welch correction.

Fig 6. CLP-induced sepsis does not alter renal 1α-hydroxylase mRNA expression.

Relative renal mRNA expression of 1α-hydroxylase measured by qRT-PCR in CLP mice (n = 5 female and male mice each) and healthy controls (n = 4 female and male mice each), 48 h post-surgery. Each bar represents the mean ± SD of 4–5 mice per experimental group. Each data point represents an individual animal.