Evaluation of serum free fatty acids in chronic renal failure: evidence from a rare case with undetectable serum free fatty acids and population data

Background Free fatty acid (FFA) accumulation in proximal tubules plays a fundamental role in the progress of kidney disease. Here, we reported a rare case with undetectable serum FFAs and further evaluated the changes of serum FFAs in patients with chronic renal failure (CRF). Methods We analyzed the clinical data of a rare case and 574 CRF patients. The mRNA expression of lipoprotein lipase (LPL), hepatic lipase (HL) and fatty acid synthase (FASN) were determined in the rare case and 30 age-matched healthy males with qPCR. Results This rare case had serious proteinuria, hyperglycemia, lipid disorders and bilateral renal glomerular filtration dysfunction. Compared with healthy males, this case showed a 1.49-fold increase of LPL expression (P < 0.01), a 3.38-fold reduction of HL expression (P < 0.001), and no significant change of FASN expression (P > 0.05). In total, 21.6% of CRF patients showed abnormal FFAs. Biochemical parameters such as blood urea nitrogen (BUN) and creatinine (CREA) significantly differed among groups with low-, normal- or high-level-FFAs. Moreover, serum FFAs was found to be associated with BUN. FFAs decreased in the group with higher BUN (> 17.4 mmol/L) and in the group with lower estimated glomerular filtration rate (eGFR) (< 15 mL/min/1.73m2). Conclusions The proteinuria, HL low expression and renal function failure may contribute to the FFA reduction, which might imply that the renal function is severely damaged. Electronic supplementary material The online version of this article (10.1186/s12944-019-1093-5) contains supplementary material, which is available to authorized users.


Background
Chronic kidney disease (CKD) represents an everincreasing worldwide health problem [1]. Hypertension, obesity, diabetes and chronic glomerular diseases are risk factors for CKD. Simultaneously, severe CKD could lead to cardiovascular diseases, fracture, anemia, hypertension, and other complications [2,3]. Chronic renal failure (CRF) would occur if renal function failed to meet the body needs, requiring replacement therapy (i.e., renal dialysis or kidney transplantation) [4]. The causes of CKD differ by country, race and age. In the United States and the United Kingdom, end-stage kidney disease mainly results from diabetic nephropathy, while primary glomerulonephritis is the leading cause of CRF in China [5].
Free fatty acids (FFAs) are derived from triacylglycerol lipolysis governed by lipoprotein lipase (LPL) [6] and hepatic lipase (HL) [7], or de novo synthesis from acetyl-CoA, malonyl-CoA and NADPH by fatty acid synthases (FASN) [8]. In proximal tubules, FFAs act as energy sources, membrane components, and precursors of lipid mediators. FFAs could be filtered by glomeruli and reabsorbed into the proximal tubules through binding to albumin. In cases of severe proteinuria, hypoxia, or intoxication, FFAs would accumulate in proximal tubules [9]. Excessive FFAs could be esterified with glycerides and deposit in intracellular lipid droplets in the form of triglycerides, which might induce renal damage. In turn, renal dysfunction further aggravates the accumulation of FFAs in proximal tubules. FFA accumulation and overoxidation lead to podocytes structural damages, resulting in glomerulopathy and CRF [10]. Dyslipidemia may affect the kidney directly by lipotoxicity, as well as indirectly through inflammation, oxidative stress, vascular injury, and hormones changes [11]. With the development of CKD, dyslipoproteinemia becomes more pronounced and even could not be substantially improved by dialysis [12].
Abnormal FFA metabolism mediates the development of many diseases, including type 2 diabetes, cardiovascular diseases, and hypertension [13,14], which are closely related to CRF [2,3]. However, to date, only a few studies have focused on serum FFAs and CRF, and FFA profile has been only evaluated in patients undergoing dialysis or renal transplantation. Notably, very low serum FFAs are not common in clinical and laboratory observations. Therefore, it is of great significance to reveal the changes of serum FFAs in the context of renal failure. In this study, we reported a rare case with undetectable serum FFAs and further evaluated the changes of serum FFAs in 574 CRF patients.

Study subjects
A 43-year-old, male patient with undetectable serum FFAs and 574 CRF patients from May 2017 to August 2018 were recruited in this study. The average age for the 574 CRF patients was 62.9 ± 0.7 years and 64.1% (n = 368) were males. For genetic analysis of the case, 30 agematched healthy males (43.9 ± 0.7 years) were also enrolled at Zhongnan Hospital of Wuhan University, China. All patients were diagnosed as CRF according to the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines [15]. Healthy controls were randomly selected from physical examination population who had normal laboratory results, including liver function, kidney function, serum glucose, lipids and electrolytes, regular blood and urine tests. The exclusion criteria were diabetes, cardiovascular diseases or other serious diseases or use of any medication. This study was approved by the Ethics Committee of Zhongnan Hospital of Wuhan University and was performed according to the Declaration of Helsinki.

Total RNA extraction
Total RNA was isolated from fresh white blood cells from the case and 30 healthy males using Trizol Reagent (Invitrogen, USA) according to the manufacturer's protocol.

Gene expression determination
cDNA was synthesized using a reverse-transcription kit with DNase treatment (TOYOBO, Japan). mRNA expression of LPL, HL, FASN was evaluated in triplicates using the iTaq™ Universal SYBR GREEN Supermixes (BioRad, USA) on a CFX Connect™ Real-Time PCR Detection System (BioRad, USA) and double normalization to β-ACTIN and GAPDH. The primers are listed in Additional file 1: Table S1.

Statistical analysis
We used SPSS 20.0 to perform all statistical analyses. The flow chart of analysis is shown in Fig. 1. Continuous and normally distributed variables were presented as the mean ± standard deviation (M ± SD), and skewed variables were described by the median (interquartile range, IQR). The nonparametric Mann-Whitney U test was used to compare clinical parameters of CRF patients between different groups. The Spearman's rank correlation test was used for correlation analysis. All statistical tests were two-sided, and P-value < 0.05 was considered statistically significant.

Case description
A 43-year-old male was admitted to hospital because of CRF and persistently increased serum creatinine for 7 years, ranging from 451.0-930.8 μmol/L during his hospitalization. He had a history of hypertension for 7 years and took nifedipine controlled-release tablets, and his blood pressure was 146/86 mmHg on admission. He had no diabetes, heart disease, hepatitis B virus infection or tuberculosis, and never underwent hemodialysis. For CRF treatment, he also had calcium dobesilate to protect blood vessels and improve circulation, polysaccharide-Iron(III) complex to fight anemia, sodium bicarbonate to correct acidosis, calcium acetate tablets and vitamin D 3 to keep balance of calcium and phosphate metabolism, as well as Hushen Keli, a traditional Chinese patent medicine to improve renal function.

Serum FFAs analysis of the case and literature review
To confirm that the case indeed had undetectable FFAs, we tested his serum FFAs and commercial FFA quality controls for 3 times. The test results showed that quality controls had expected values, and serum FFAs of the patient was only 0.07 μmol/L at the 2 nd run detection. Next, to avoid that serum FFA concentration was too high to detect, we tested FFAs with diluted serum, and the results were similar to the undiluted serum. Thus, we attempted to interpret this interesting clinical observation.
In general, FFAs are mainly contained in triglycerideriched lipoproteins or bind to albumin in blood circulation [17]. Albumin-bound FFAs could be filtered through the glomeruli and accumulate in the proximal tubules in the setting of massive proteinuria [9]. Given this case had serious proteinuria (24 h-TP, 1.06 g/24 h; urine protein, positive), along with badly impaired bilateral renal GFR, one possible reason for the undetectable FFAs should be FFAs lost with urine protein. Fig 2a shows the mRNA levels of LPL, HL and FASN. Compared with healthy males, the case had a 1.49-fold increase of LPL expression (P < 0.01), a 3.38-fold reduction of HL expression (P < 0.001), and no significant change of FASN expression (P > 0.05). It appeared that the fold-change of HL expression was greater than that of LPL. The increased LPL expression might not be able to compensate for the alteration triggered by HL low expression. Thus, the significant down-regulation of HL might be also responsible for the undetectable FFAs.
Hypertension and CRF have found to be the cause and effect in FFA metabolic disorders [2,14,18]. FFAs increased in patients with hypertension, suggesting that a higher level of FFAs was an independent risk factor for hypertension [14,19]. To our surprise, the case had severe hypertension (grade-3) but with undetectable FFAs. The discrepancy between the case and previous studies might imply that FFA metabolism was complicated in CRF circumstance, and some unknown factors might also have contributed to the undetectable FFAs. Taken together, the undetectable FFA in this CRF patient was mainly caused by FFA elimination with urine protein and down-regulation of HL.

Population data analysis
To provide further evidence for the undetectable FFAs in this case and understand FFA changes in CRF patients, we systematically evaluated the alterations of FFAs and clinical parameters in 574 CRF patients. Of them, 13.59% had decreased FFAs, 8.01% had increased FFAs, and 78.4% were normal.
Based on the reference interval of serum FFAs (129-769 μmol/L), we divided the 574 participants into three groups: FFA-reduced (< 129 μmol/L), FFA-normal, and FFA-increased (> 769 μmol/L). The average of FFAs in the FFA-reduced group was 82.16 μmol/L, and the lowest value was 11.28 μmol/L. Next, we examined the difference of laboratory indicators among the three groups with K-independent nonparametric test, including liver function, kidney function, serum lipids and electrolytes ( Table 2). No significant difference was found in age (P = 0.053) or gender (P = 0.072) distribution. Serum AST/ ALT, TBIL, DBIL, UBIL, GLU, BUN, CREA, TC, TG, LDL-C, ApoB, PLIP, Na + and Cl − significantly differed across the three groups. For instance, liver function parameters (i.e., AST/ALT, TBIL, DBIL and UBIL) showed an increasing trend with FFA increasing (P < 0.01); serum GLU elevated in both the FFA-reduced group and the FFA-increased group (P < 0.01) when compared to the FFA-normal group. As for renal function indicators, serum BUN in the FFA-reduced group was the highest among the three groups (P < 0.01); serum CREA was apparently higher than that in the FFA-normal group (P < 0.01). Serum lipids differed across the three groups, except for HDL-C, ApoA1 and Lp(a). Also, serum Na + and Cl − showed statistically difference across the three groups. Therefore, our data provided strong evidence that serum FFAs were closely related to the renal function, given the long-standing recommendations for clinical evaluation of renal function with serum BUN and CREA.
In addition to the BUN and CREA, the urine 24 h-TP is another important indicator for renal function. We therefore evaluated serum FFAs in 98 CRF patients who had results of urine 24 h-TP. Of the 98 patients, 19 patients had normal 24 h-TP (≤ 0.15 g/24 h) and 79 patients showed abnormal increased 24 h-TP. The median of 24 h-TP in this cohort was 1.395 g/24 h (IQR, 0.280-3.115). Consequently, CRF patients with higher urine 24 h-TP were more likely to have lower concentrations of FFAs (Fig. 2b), regardless of grouping by reference interval or the IQR range. More importantly, serum FFAs was found to be negatively associated with urine 24 h-TP (r = − 0.24, P < 0.05, Fig. 2c).
Furthermore, the Spearman's rank correlation analysis revealed that FFAs had positive correlations with AST, AST/ALT, TBIL, DBIL, UBIL, TP, GLB, CO 2 , TC, TG,  LDL-C, ApoB, PLIP, and negative correlations with TBA, GLU, BUN, ApoA1, Lp(a), K + and Cl − ( Table 3). As shown in Fig. 1, serum AST/ALT, TBIL, DBIL, UBIL, GLU, BUN, TC, TG, LDL-C, ApoB, PLIP, Cl − not only differed significantly (all P < 0.05) across the three groups, but also were significantly associated with serum FFAs. Therefore, we divided the patients into group-1 (≤ median) and group-2 (> median) on the basis of the median of the above clinical parameters to assess serum FFAs. As illustrated in Fig. 3, the Mann-Whitney U test revealed that serum FFAs were significantly increased in group-2 (AST/ALT, TBIL, DBIL, UBIL, TC, TG, ApoB, PLIP) and in group-1 (GLU, BUN, Cl − ). In agreement with that BUN was an important biomarker for renal function evaluation [20], of the 11 indicators, the abnormal rate of BUN was approximately 92.29%, while the rest was less than 50% (Additional file 1: Table S2). Particularly, the inverse correlation between FFAs and BUN further implied that the lower FFAs reflected poorer renal function.
Additionally, the eGFR was used to evaluate renal function for the vast majority of CKD patients in clinical practice due to convenience and cost-efficiency. The KDIGO guidelines also recommend that CRF patients with eGFR< 15 mL/min/1.73m 2 should be classified as CKD-5 stage [15]. We therefore divided patients into two groups by eGFR: group-1 (eGFR < 15 mL/min/ 1.73m 2 ) and group-2 (eGFR-2 ≥ 15 mL/min/1.73m 2 ). Fig. 3b demonstrates that FFAs increased in the group-1, indicating inverse correlation between FFAs and kidney function. Consistent with the population data, the case with undetectable FFAs also showed elevated GLU, BUN, and decreased eGFR, PLIP. Collectively, our data demonstrated that FFA reduction was strongly associated with glucose, lipid metabolic disorders, and severe renal damages.

Discussion
In the current study, we found that a rare case with undetectable FFAs had severe renal dysfunction, hypertension, hyperglycemia, proteinuria and HL down-regulation. Further population data analysis demonstrated that 21.6% CRF patients had abnormal serum FFAs. Specifically, more CRF patients had decreased FFAs when compared to patients with increased FFAs (13.6% vs. 8.0%). FFA reduction could be caused by multiple factors, including loss with proteinuria, hyperglycemia, lipid metabolic disorders, severely impaired renal function and dysregulation of FFA-related enzymes.
In line with a previous report that FFAs showed renoprotective potential, renal transplant recipients with higher plasma FFAs had lower risk of graft failure [21], our population data also suggested that lower serum FFAs indicated worse kidney function in CRF patients. The care with undetectable FFAs had dramatically reduced eGFR (12 mL/ min/1.73m 2 ), proteinuria, and significantly increased BUN, CREA. By telephone follow-up, we noticed that this patient had been taking hemodialysis owing to out of controlled Notably, different individual FFAs had various effects on renal function. In most cells such as podocytes, the overload of saturated fatty acid could induce insulin resistance and cell death via lipid toxicity, whereas monounsaturated fatty acid could reverse this lipotoxicity [22]. Older individuals with a higher level of plasma polyunsaturated fatty acids were at lower risk of developing renal insufficiency [23]. CKD patients under hemodialytic treatment had an increase of monounsaturated fatty acids and a decrease of n-3 polyunsaturated fatty acids [18]. Earlier studies also observed that polyunsaturated fatty acids intake showed protective effects on renal function [3,24].
Hyperglycemia was a risk factor for CKD [25]. In this study, serum GLU abnormally elevated in the rare case, similar results were also found in CRF patients with increased or reduced FFAs. Additionally, CRF patients with higher serum GLU and lower serum FFAs exhibited apparent reduction of eGFR (Additional file 1: Figure S1). Our data implied that, to some extent, serum lower FFAs and hyperglycemia promoted renal dysfunction in CRF patients.
Besides, the relation of population data between FFAs with AST/ALT, TBIL, DBIL and UBIL suggested normal liver function was vital for FFA metabolism. Serum FFAs depended on the balance between FFA release from adipose tissue and FFA uptake/oxidation by liver and muscle tissues [26]. In our study, FFA alterations were closely related to serum levels of TC, TG, LDL-C and ApoB, which were consistent with the source and metabolism of FFAs. In terms of serum electrolytes, we only found that Cl − was inversely correlated with FFAs. Interestingly, a recent study has proved that elevated FFAs and triacylglycerol levels directly reduced blood Mg 2+ levels in metabolic disorders [27].
Nevertheless, this study could not rule out the limitations and challenges. Firstly, we only determined the mRNA expression of LPL, HL and FASN in the case and 30 healthy males, these results should be further confirmed in a larger sample size study and covered more genes related to FFA metabolism. Secondly, owing to treatment requirements, the patients were not under standard diet, exercise and same medications, we therefore were unable to adjust for these factors in the statistical analyses. It is not surprising that all these limitations did not pose a serious problem to our findings that low FFAs indicated severe renal damages based on the large population data analysis.