Renal expression and urinary excretion of liver‐type fatty acid‐binding protein in cats with renal disease

Abstract Background Liver‐type fatty acid‐binding protein (L‐FABP) is a biomarker for early detection of renal disease in humans. Liver‐type fatty acid‐binding protein is cytotoxic oxidation products secreted from proximal tubules under ischemia and oxidative stress. Objective To examine renal expression and quantify urinary excretion of L‐FABP in catswith renal disease. Animals One hundred and thirty‐four client‐owned cats including 34 cats with serum creatinine (sCre) values >1.6 mg/dL and 10 other cats that died in clinics. Methods Tissue expressions of L‐FABP were examined by reverse transcription polymerase chain reaction and Western blotting. Urinary L‐FABP (uL‐FABP) and serum L‐FABP (sL‐FABP) levels were determined by enzyme‐linked immunosorbent assay. Anti‐liver‐type fatty acid‐binding protein antibody immunostained renal sections. Results Feline kidneys express L‐FABP. Strong L‐FABP signals were observed in the lumens of proximal tubular cells in 5 cats with high uL‐FABP excretion, but not in 5 cats with low uL‐FABP excretion. In 9 normal cats, uL‐FABP index was <1.2 μg/g urinary creatinine (uCre). High uL‐FABP indexes (>10.0 μg/g uCre) were detected in 7 of 100 cats with low sCre (<1.6 mg/dL) and 18 of 44 cats with high sCre (>1.6 mg/dL). There was a weak correlation between L‐FABP index and sCre, serum symmetric dimethylarginine (SDMA), or blood urea nitrogen (BUN), and these correlation coefficients were increased by analyzing only data of cats with sCre >1.6 mg/dL. There was a weak correlation between u L‐FABP index and sL‐FABP in all tested cats, but not in cats with high sCre. Conclusions and Clinical Importance This study demonstrates correlations between L‐FABP and current renal biomarkers for chronic kidney disease in cats, such as sCre and SDMA. Liver‐type fatty acid‐binding protein may be a potential biomarker to predict early pathophysiological events in feline kidneys.

Conclusions and Clinical Importance: This study demonstrates correlations between L-FABP and current renal biomarkers for chronic kidney disease in cats, such as sCre and SDMA. Liver-type fatty acid-binding protein may be a potential biomarker to predict early pathophysiological events in feline kidneys.
acute kidney injury, biomarker, chronic kidney disease, kidney

| INTRODUCTION
Renal diseases are common in domestic cats, especially elderly animals. [1][2][3] Chronic kidney disease (CKD), which results in uremia at the end stages, is a leading cause of illness and death in elderly cats. The development of CKD was reported in 31% of clinically healthy cats >9 years of age. 4 The incidence of CKD in cats is 15% at >15 years of age, 2 and more recently as 27% at >10 years of age. 5 In addition to CKD, acute kidney injury (AKI) is also an important renal disease in cats, with a case fatality rate exceeding 50% despite treatment. 6,7 Therefore, it is crucial to diagnose CKD and AKI in the early stages to prevent their progression and for clinical management.
Proteinuria is an early indicator used to identify renal dysfunction in animals. 8 Some specific proteins are present in the urine in cases of renal disease. For example, retinol binding protein 9 and cystatin C 10 levels are increased by dysfunction of proximal tubular reabsorption in cats. In contrast, Tamm-Horsfall protein excretion decreases in cats associated with decreases in its production by damage to distal tubular cells. 11 Carboxylesterase 5A (also known as cauxin), which produces a sex pheromone, 12 is a major urinary protein in normal healthy cats 13 and its excretion is decreased in cats with CKD due to decrease of number of proximal tubular cells expressing it. [14][15][16][17] The increase of urinary excretion of these proteins usually results from functional, structural, or both damages in the glomerulus and renal tubules.
Urinary excretion of liver-type fatty acid binding protein (L-FABP) has been clinically recognized as a useful biomarker for monitoring early detection of AKI and CKD in humans. 18,19 In humans, L-FABP is expressed in renal proximal tubular cells in addition to the liver, pancreas, and small intestine. 20 The functions of L-FABP include binding to and transporting fatty acids to the mitochondria or peroxisomes, where they are β-oxidized, and participation in intracellular fatty acid homeostasis. 21 Liver-type fatty acid binding protein is excreted in urine due to ischemia (or low capillary blood flow) and oxidative stress on the renal tubules before the progression of renal damage. 22,23 In the absence of renal diseases, L-FABP secreted from the liver into the blood crosses the glomerular barrier and then is reabsorbed by the proximal tubular cells, resulting that such L-FABP hardly appears in urine, in contrast, in the case of hepatitis fulminant or hepatorenal syndrome, approximately 15% of urinary L-FABPs (uL-FABPs) originate in the liver (CMIC Holdings, unpublished data).
Renal expression of L-FABP is species specific, and the expression of L-FABP is genetically suppressed in the kidneys of rodents, such as mice. 24 Therefore, it is necessary to determine whether L-FABP is expressed in the proximal tubular epithelial cells of the cat kidney as well as humans. The aim of this study was to investigate L-FABP expression in the cat kidney and the relationship between levels of L-FABP excretion and its renal distribution. We also examined the correlations between uL-FABP and current renal biomarkers in cats with renal diseases. These results will improve our understanding of the pathophysiological significance of L-FABP in cats.

| Animals and sample collection
Urine and blood samples were collected from 134 cats (24 intact males, 28 castrated males, 36 intact females, and 46 spayed females, between 5 months and 21 years old), which had been treated at veterinary clinics between 2005 and 2018. Samples of 9 of the 134 cats were collected in our previous study. 15 Urine and kidney tissue samples were collected from 10 cats that died in veterinary clinics in the United States and then used for postmortem histopathological examination. We used normal kidney and liver tissues of a male cat, which were sampled in our previous study and stored at −80 C. 25 Urine samples were collected by cystocentesis using 23-gauge needles. Blood (0.5 mL) was collected from the cephalic vein using a 24-gauge needle. All urine and blood samples were centrifuged at ×500 for 5 minutes, and the supernatants were stored at −30 C until biochemical analyses, which had been done within 8 months after sampling. This study was performed in accordance with local animal ethics guidelines and was approved by the Animal Research Committee of our university. All feline owners approved of providing the urine, blood, and kidney samples of their cats for this study.

| Reverse transcription-polymerase chain reaction
Total RNA was extracted from the kidneys and liver of an intact male cat using TRIzol reagent (Gibco BRL, Grand Island, NY). cDNAs were generated from 1 μg of total RNA from the kidney and liver using ReverTra Ace (Toyobo, Osaka, Japan) with oligo(dT) 15

| Urinary excretion of L-FABP in cats
Preliminary studies were carried out to examine L-FABP excretions in cats urine using samples collected from 9 cats with CKD. They had high  (Table S1), uL-FABP values were normalized to the uCre concentrations. In the 9 cats with CKD, the uL-FABP/uCre ranged from 97.60 to 6308.25 μg/g uCre (median 225.91 μg/g uCre). The results indicated that L-FABP is excreted in the urine of cats with CKD.
Next, the correlation was examined between density of the L-

| Correlation analyses between uL-FABP and sL-FABP
The serum concentrations of L-FABP were measured using the L-FABP ELISA kit in 108 of 134 cats, which contained 12 cats whose sCre was >1.6 mg/dL (IRIS ≥ II). We could not measure sL-FABP con-

| Correlation analyses between uL-FABP and sCre or SDMA
We examined the correlations between uL-FABP index and sCre, BUN, or SDMA in 134 cats of which detail information is shown in F I G U R E 2 Urinary excretion of L-FABP in cats. A, Western blotting of urine (aliquots of 10 μL) collected from 9 cats with CKD was performed using anti-L-FABP antibody. Other detail information of the 9 cats is described in Table S1. B, Correlation between urinary L-FABP contents quantified by L-FABP ELISA and density of the L-FABP band detected in Figure 2A. CKD, chronic kidney disease; ELISA, enzyme-linked immunosorbent assay; L-FABP, liver-type fatty acid-binding protein Table S1. Serum SDMA is derived from intranuclear methylation of L-arginine by protein arginine methyltransferase and primarily eliminated by renal excretion, suggesting that SDMA is a potential endogenous marker of glomerular filtration rate (GFR). 27  Spearman's rank correlation coefficient analysis showed that there are significant correlation between Log uL-FABP/uCre and Log sCre (Spearman's rho: 0.2880; P = .0007, Figure 5A) and between Log F I G U R E 3 Representative immunohistochemical images of L-FABP in paraffin-embedded renal sections of 10 cats. A-E and F-J are renal sections of 5 cats with low urinary L-FABP index and 5 cats with high uL-FABP index, respectively. Their individual information, urinary L-FABP indexes, pathological diagnosis, and clinical history are described in Table S2. L-FABP, liver-type fatty acid-binding protein; uL-FABP, urinary L-FABP uL-FABP/uCre and Log BUN (Spearman's rho: 0.4420; P = .0001, Figure 5B). These correlation coefficients were increased between Log uL-FABP/uCre and Log sCre (Spearman's rho: 0.3944; P = .0210) and between Log uL-FABP/uCre and Log BUN (Spearman's rho: 0.6637; P < .0001) by analyzing data only of 34 cats whose sCre was >1.6 mg/dL.
Thirty-four of the 134 cats was diagnosed as azotemic CKD based on sCre concentration >1.6 mg/dL in conjunction with USG < 1.035. 28 According to International Renal Interest Society (IRIS) guidelines, 29  and CKD in cats would be 4 μg/g uCre, which is lower than in humans whose values are <8.4 μg/g uCre. 38 The present study suggests that it will be valuable to find cats with high values of uL-FABP index without increases in other renal biomarkers for diagnosis of early AKI in cats. There was a weak correlation, but significantly between uL-FABP and SDMA in 106 cats, and between uL-FABP and sCre in 134 cats, in addition to the positive correlation between SDMA and sCre as described in previous reports. 39,40 To predict early pathophysiological events in feline kidneys, we are now focusing on cats that exhibit high L-FABP/uCre (>10 μg/g uCre) but low sCre (<1.6 mg/dL) and low serum SDMA (<14 μg/dL). This might be because the mechanisms underlying increases in uL-FABP excretion are different with sCre or SDMA. sCre and SDMA are biomarkers to assess renal function and they increase when GFR has declined. In contrast, L-FABP excretion reflects the tubular response due to ischemia and oxidative stress, and increases from the early stages of pathological events before structural damage of the glomerular, tubular, and both cells. 33,37 Other biomarkers, such as urinary albumin 41 and N-acetyl-β-D-glucosaminidase (NAG), 16  This suggests that proteins that increase pathophysiological events but not structural and functional damage to the kidney might be suitable as biomarkers for early prediction of renal diseases in cats. Therefore, we expect to find early stages of AKI and CKD from cats with high uL-FABP index.
It is crucial to diagnose the early stages of CKD and AKI in cats for their clinical managements. Ischemic damage is a common cause of AKI in human patients. 43 AKI and CKD are bidirectionally connected, as the presence of CKD represents a predisposing factor for AKI and the occurrence of AKI is a risk factor for subsequent development of CKD. 44 In cats, AKI can be caused by toxic, ischemic, inflammatory, obstructive, or infectious insults. 6,7,45 AKI initiates a cascade that results in chronic and progressive tubulointerstitial inflammation and fibrosis closely mimicking those observed in naturally occurring CKD. 46 Although the bidirectional nature of this relationship was not investigated, AKI could be a factor in the development or progression of CKD in cats as well as in humans. Therefore, we propose that L-

| CONCLUSIONS
The present study indicated that L-FABP is expressed in the kidneys of cats as in humans. We propose that L-FABP is a potential bio-

OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
This study was performed in accordance with local animal ethics guidelines and was approved by the Animal Research Committee of the Faculty of Agriculture of Iwate University (#A201432).