If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
School of Veterinary Medicine, College of Science, Health, Engineering and Education, Murdoch University, Murdoch, Western Australia, AustraliaDepartment of Intensive Care Medicine, Royal Perth Hospital, Perth, Western Australia, AustraliaMedical School, University of Western Australia, Perth, Western Australia, Australia
Following general anaesthesia, severe hypotension was induced by phlebotomy, maintaining mean arterial blood pressure (MAP) < 40 mmHg for 60 minutes, followed by resuscitation with intravenous gelatine solution to maintain MAP > 60 mmHg for 3 hours. Following euthanasia, renal tissue was examined by light microscopy (LM) and transmission electron microscopy (TEM). Urinary and serum concentrations of neutrophil gelatinase-associated lipocalin (NGAL), cystatin C (CysC), and gamma-glutamyl transpeptidase (GGT), serum creatinine and urine output were measured at baseline and hourly until euthanasia. Data are presented as mean and 95% confidence interval and analysed using repeated measures analysis of variance with Dunnett’s adjustment, p < 0.05.
Structural damage to proximal renal tubular cells was evident on LM and TEM. Urinary biomarker concentrations were significantly elevated from baseline, peaking 2 hours after haemorrhage at 19.8 (15.1–25.9) ng mL–1 NGAL (p = 0.002), 2.54 (1.64–3.43) mg mL–1 CysC (p = 0.009) and 2043 (790–5458) U L–1 GGT (p < 0.001). Serum creatinine remained within a breed-specific reference interval in all dogs. Urinary protein–creatinine ratio (UPC) was significantly elevated in all dogs from 1 hour following haemorrhage.
Conclusions and clinical relevance
Urinary NGAL, CysC and GGT concentrations, and UPC were consistently elevated within 1 hour of severe hypotension, suggesting that proximal renal tubules are damaged in the earliest stage of ischaemia-reperfusion AKI. Measurement of urinary biomarkers may allow early diagnosis of AKI in anaesthetized dogs. Urinary GGT concentration and UPC are particularly useful as they can be measured on standard biochemistry analysers.
). Conventional diagnosis of AKI relies on surrogate markers for glomerular filtration rate (GFR) such as serum creatinine (sCr); however, sCr is insensitive and GFR has to decline significantly before AKI can be diagnosed (
). In health, NGAL is expressed by many cells throughout the body, filtered by the glomerulus and reabsorbed by the proximal tubules, resulting in minimal urinary excretion. During kidney injury, renal tubular epithelial cell expression and proximal tubular secretion of (monomeric) NGAL increases in various species (
). Of the urinary enzymes, GGT is the easiest to determine clinically as it can be measured on standard biochemistry analysers. This enzyme is found in the brush border of renal proximal tubule cells, and extra-renally in many other cells throughout the body (
). However, measurement of CysC is relevant to clinical veterinary practice because validation of a particle-enhanced immunoturbidimetric assay (PETIA) can be performed by standard biochemistry analysers (
While some studies report the use of urinary concentration of NGAL, GGT and CysC for the diagnosis of AKI in dogs prior to development of azotaemia, others investigate their diagnostic performance during experimental nephrotoxic AKI (
). Canine clinical studies are often small with limited information regarding AKI aetiology and compare biomarker performance to measurement of GFR rather than histologic evidence of acute tubular injury (
). Additionally, as CysC and some forms of NGAL are produced by non-renal cells, there is concern that urinary concentrations of these biomarkers may reflect an increased filtered load following systemic release (e.g., due to inflammation, or non-renal disease) rather than a direct consequence of renal tubular cell damage (
). The authors have previously reported findings of elevated urinary NGAL concentrations (uNGAL), measured using an enzyme-linked immunoassay, within 2 hours of renal I/R injury in dogs with histologic evidence of acute tubular injury; however, contribution of systemic NGAL release was not investigated (
The primary objective of this study was to document changes in urinary biomarker concentration in an I/R injury canine model, alongside conventional diagnostic tests for AKI. Additionally, we investigated the possible source of markers by comparing urinary excretion with serum concentrations of NGAL and CysC. We examined the type and specific location of lesions within renal tissue identified by light microscopy (LM) and transmission electron microscopy (TEM) following I/R injury. We hypothesized that urinary concentrations of NGAL, CysC (uCysC) and GGT (uGGT) would reflect changes consistent with AKI earlier than three conventional markers of reduced renal function: sCr, urine output and urinalysis. Further, we hypothesized that elevated uNGAL, and uCysC reflect structural damage to renal tubular cells.
A total of six adult entire male Greyhound dogs, median (range) body weight 31 (28–32) kg, that were retired racing dogs unable to be rehomed for behavioural reasons were included in this study. Ethics approval was granted by the Murdoch University Animal Ethics Committee, Western Australia (R2726/15). The dogs were surrendered by their owners following informed consent and cared for as per the Australian code for the care and use of animals for scientific purposes. The methodology and reporting for this study were based on the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. These dogs were a control group for a separate interventional study (
); thus, group size was determined by the requirements of that study. Additionally, a sample size calculation (https://www.stat.ubc.ca/∼rollin/stats/ssize/n2.html) was performed using uNGAL data from a previous study utilizing the same model of I/R injury AKI. This calculation confirmed that a sample size of six dogs would be sufficient to identify a uNGAL increase of 110 ng mL–1 with standard deviation of 70 ng mL–1 at post haemorrhage time points, with 0.8 power and 0.05 alpha level (
). All dogs were deemed to be healthy and free of renal disease based on physical examination, renal ultrasonography, urinalysis, complete blood count, sCr, blood urea nitrogen and serum albumin concentration, with all variables within reference intervals previously reported for healthy adult Greyhounds (
). Serum and urine supernatant collected during the enrolment screening process were stored at –80 °C for later analysis (PRE samples). Dogs were acclimatized for 7 days, following enrolment, housed individually in cages incorporating an enclosed outdoor run. They were given ad libitum access to drinking water and were fed a commercial dried food (Hill’s Prescription Diet i/d Canine; Hills Pet Nutrition, NSW, Australia) twice daily. A deworming prophylactic medication (Popantel; Jurox Animal Health, NSW, Australia) was administered on admission to hospital. Physical examination was performed by a veterinarian (JD or AR) each day, and animals that developed any illness or injury were excluded.
Ischaemia-reperfusion injury model
Experimental renal I/R injury was induced via pressure-guided haemorrhage under general anaesthesia. The anaesthetic and experimental protocol is described in detail in a previous publication (
). Briefly, methadone 0.3 mg kg–1 (Ilium Methadone, 10 mg ml–1; Troy Laboratories, NSW, Australia) was administered intravenously (IV) 30 minutes prior to induction of general anaesthesia using IV alfaxalone 2.2–3.3 mg kg–1 (Alfaxan injection, 10 mg mL–1; Jurox). Anaesthesia was maintained with isoflurane (I.S.O.; VCA, NSW, Australia) in an oxygen and medical air mix (FiO2 30%) via an orotracheal tube and circle breathing system. End-tidal isoflurane was maintained at 1.3–1.4%. Mechanical ventilation was performed to maintain normocapnia [arterial carbon dioxide tension 35–45 mmHg, (4.7–6 kPa)]. Hartmann’s solution (Compound Sodium Lactate; Baxter Healthcare, NSW, Australia) 10 mL kg–1 hour–1 IV and fentanyl (Fentanyl injection, 50 μg mL–1; AstraZeneca, NSW, Australia) 2 μg kg–1 hour–1 IV were administered throughout anaesthesia. Within 60 minutes of anaesthetic induction, severe hypotension (ischaemia phase) was produced in all dogs, by removing a blood volume sufficient to maintain mean arterial pressure (MAP) ≤ 40 mmHg for 60 minutes. Fluid resuscitation was performed to increase MAP ≥ 60 mmHg for the following 3 hours via administration of 20 mL kg–1 IV succinylated gelatine solution 4% (Gelofusine; B Braun, NSW, Australia) over 30 minutes. Arterial and central venous blood were collected every 60 minutes for blood gas and co-oximetry measurement to allow calculation of oxygen extraction ratio, specifically the oxygen consumption divided by the oxygen delivery, thereby to identify the presence of anaerobic metabolism. Data collection was performed at (1) T0: 60 minutes after anaesthetic induction, prior to blood removal; (2) T1: 60 minutes after start of haemorrhage, prior to administration of gelofusine; and (3) T2–T4: 60, 120 and 180 minutes after achieving a MAP > 60 mmHg following gelofusine administration. Blood was obtained via a right jugular venous cannula (BD Angiocath IV Catheter 14 gauge × 13 cm; Becton Dickinson, NJ, USA) at each sample time. Samples were collected into plain serum collection tubes, allowed to clot, centrifuged at 1358 g for 10 minutes, and serum aliquots stored at –80 °C for later analysis. The urinary bladder was emptied completely every hour via an indwelling urinary catheter (Premium Foley Catheter (silicone) 8 Fr × 55 cm; Smiths Medical, MN, USA), and urine output recorded. At each time point, urine specific gravity was measured and dipstick analysis (Multistix; Siemens, PA, USA) performed. Urine sediment was examined using LM at low (×200) and high (×400) power. A minimum of 10 high-power fields were examined for the presence of blood cells, epithelial cells, and granular and hyaline casts. Remaining urine was centrifuged at 339 g for 5 minutes, and urine supernatant divided into multiple aliquots and stored at –80 °C for later laboratory analysis.
Urine creatinine (uCr), uCysC, sCysC, uGGT, urine protein–creatinine ratio (UPC) and sCr concentrations were measured on samples frozen at –80 °C for up to 3 months, using a Cobas Integra 400plus biochemistry analyser (Roche Diagnostics, Switzerland). Urine and sCr were measured using the Jaffe method (CREJ2; Roche Diagnostics), uGGT via an enzymatic colorimetric assay (GGTS2; Roche Diagnostics) and urinary protein using a turbidimetric method using benzethonium chloride (TPUC3; Roche Diagnostics). The uCysC and sCysC measurement via this analyser involved the use of a PETIA (Tina-quant Cystatin C Gen.2; Roche Diagnostics) previously validated for use in canine samples (
Concentrations of NGAL were measured in the urine and serum of all dogs at all sampling times using a bead-based multiplexed immunoassay (MILLIPLEX MAP Canine Kidney Toxicity Expanded Magnetic Bead Panel 1; Merck Millipore, MA, USA). While this multiplex assay is capable of simultaneously measuring other biomarkers (including CysC), published validation studies suggest measurement is precise and accurate only for NGAL in canine urine and serum; hence we report NGAL measurement only for this assay (
). Urine and serum were stored at –80 °C for up to 3 months prior to analysis. Urine was diluted by a factor of 1:2, and serum by a factor of 1:10, using assay buffer prior to analysis. For serum samples, a serum diluent (Serum Matrix; Merck Millipore) was added to wells containing the Standards and Controls in place of assay buffer in order to mitigate any matrix effects. All samples were analysed in duplicate. The assays were performed according to manufacturer instructions and fluorescence measured using a multiplex reader (Bio-Plex MAGPIX Multiplate Reader; Bio-Rad, CA, USA) with xPONENT software. The median fluorescence intensity data was analysed using a five-parameter logistic curve to calculate analyte concentrations in each sample.
Renal tissue preparation and processing
Following collection of cardiovascular data and urine/blood at T4, all dogs were euthanized using pentobarbitone (Lethabarb Euthanasia Injection, 300 mg mL–1; Virbac, NSW, Australia) 150 mg kg–1 IV. Both kidneys were removed immediately, sectioned and stored in 10% formalin. Small cubes of cortex (approximately 2–3 mm3) were placed into 3% glutaraldehyde for TEM evaluation.
For LM, formalin-fixed tissues were processed by the Comparative Pathology & Digital Imaging Shared Resource (The Ohio State University, Columbus, OH, USA). A single board-certified veterinary pathologist (RC), blinded to the results of the biomarker analysis, scored haematoxylin and eosin and periodic acid Schiff–stained 3 μm thick histologic sections of both kidneys. Prior to microscopic evaluation, lines were drawn to delineate the superficial cortex, corticomedullary junction and medulla. Within each region, 20 randomly selected fields entirely of tubulointerstitium (e.g., lacking glomeruli or large calibre arteries) were examined at 200× and the average number of injured tubules per field was calculated for each specimen/region. Tubular injury was defined as loss of the apical brush border, denudation of tubular basement membranes, singly dead tubular epithelial cells and tubules with intraluminal detached cells/cellular debris. After enumeration of injured tubules in each region, the greatest difference among samples was for the cortex, so degree of tubular injury was graded based on these quantitative data as follows: grade 0/normal (no lesions observed); grade 1/minimal lesions (mean ≤ 1 injured tubule per field); grade 2/mild lesions (mean ≤ 3 injured tubules per field); grade 3/moderate lesions (mean > 3 injured tubules per field). The pathologist did not identify any specimens with severe tubular lesions.
For TEM, after post-fixation of glutaraldehyde-fixed tissue in 1% osmium tetroxide, the specimens were serially dehydrated, infiltrated in an acetone/epoxy plastic, and embedded in plastic. The plastic blocks were sectioned to a silver-grey interference colour (55–60 nm) and placed on copper mesh grids. The sections were stained with filtered lead citrate/sodium citrate solution (Electron Microscopy Sciences). Grids were imaged on a JEOL JEM-1400 TEM (JEOL USA, Inc., MA, USA) and representative images were photographed with an Olympus SIS Veleta 2K camera (Olympus Soft Imaging Solutions GmbH, Germany).
Data are presented as mean or geometric mean (for transformed data), and 95% confidence intervals (CIs) of the mean (or geometric mean). All data were tested for normality using the Shapiro-Wilk statistic and Q-Q plots with rejection of the null hypothesis of normality at p < 0.05. Where marker concentrations were not normally distributed, logarithmic (base-10) transformation was used to generate a normal data distribution. The raw or transformed concentrations were tested for an effect of time using a repeated-measurements, one-way analysis of variance with the F test considered significant at p < 0.05; significance then directed post hoc multiple comparisons of concentrations at T1–T4 to those at T0 using Dunnett’s adjustment with type I error controlled at 0.05 (Prism8 software; Graphpad, CA, USA). Magnitude of change in biomarker concentration from T0 for individual dogs at T1–T4 was calculated as: TX biomarker concentration/T0 biomarker concentration, and used to produce a radar plot (Microsoft Excel; Microsoft Corporation, WA, USA).
All six dogs enrolled in the study completed it. Target MAP (≥60 mmHg at T0, T2, T3 and T4; and ≤ 40 mmHg at T1) were achieved in all dogs (Fig. 1). Mean (95% CI) volume of blood withdrawn to achieve MAP ≤ 40 mmHg for 60 minutes was 51 (39–63) mL kg–1. Oxygen extraction ratio at T1 was 0.58 (0.45–0.70), consistent with anaerobic metabolism (
), and returned to baseline (<0.3) in all dogs from T2–T4 (Fig. 1).
Oliguria (urine output < 1.0 mL kg–1 hour–1) was present in all six dogs at T1, five dogs (83%) at T2, and four dogs (67%) at T3 and T4 (Fig. 1). Urinalysis revealed haematuria in some dogs from PRE–T2, cylinduria in some dogs from T1 onwards, and proteinuria in all dogs from T2 onwards (Table 1). UPC was significantly elevated from T2 onwards (Table 2).
Table 1Frequency of abnormal urinalysis findings in six isoflurane anaesthetized Greyhounds before anaesthesia (PRE), following induction of anaesthesia with intravenous (IV) alfaxalone (T0), after 60 minutes of haemorrhage and severe hypotension (T1), and 1 (T2), 2 (T3) and 3 hours (T4) after fluid resuscitation. For detailed information, see Fig. 1 legend. A dipstick reading > trace was regarded as a ‘positive’ result. Casts were granular and/or hyaline. Haematuria is > 5 erythrocytes/high-power field. Pyuria is > 5 leucocytes/high-power field.
For urine specific gravity, when above measurable range of refractometer recorded as 1.060. Note that from T2 onwards, administration of IV synthetic colloid solutions will have interfered with refractometry (Yam et al. 2019).
1.060 (all > 1.060)
∗ At T1, data are missing from one dog (n = 5) for dipstick analysis, and data from two dogs are missing (n = 4) for sediment examination (owing to insufficient sample volume).
† For urine specific gravity, when above measurable range of refractometer recorded as 1.060. Note that from T2 onwards, administration of IV synthetic colloid solutions will have interfered with refractometry (
Table 2Mean (95% confidence intervals) of serum and urine biochemistry and renal biomarker concentrations of six Greyhounds before anaesthesia (PRE), following anaesthetic induction (T0), after 60 minutes of haemorrhage and severe hypotension (T1), and 1 (T2), 2 (T3) and 3 hours (T4) after fluid resuscitation. For detailed information, see Fig. 1 legend. Data are not reported for uCysC at T1 owing to insufficient sample size (data only recorded from three dogs). For multiple comparison testing of uCysC data, a concentration of 0.4 [i.e., the limit of quantification (LOQ)] was used for T0. Cr, creatinine; CysC, cystatin C; GGT, gamma-glutamyl transpeptidase; NGAL, neutrophil gelatinase-associated lipocalin; UPC, urine protein–creatinine ratio.
p (F test)
Serum Cr (mg dL–1)
Serum CysC (mg mL–1)
Serum NGAL (ng mL–1)
Urine CysC (mg mL–1)
all < 0.4 (LOQ)
all < 0.4 (LOQ)
Urine CysC/Cr (mg:g)
Urine GGT (U L–1)
Urine GGT/Cr (U:g)
Urine NGAL (ng mL–1)
Urine NGAL/Cr (ng:g)
Urine protein/Cr (mg:mg)
Urine/serum Cr (mg:mg)
Urine/serum CysC (mg:mg)
Urine/serum NGAL (ng:ng)
∗Significantly greater (p < 0.05) than T0 (Dunnett’s test).
There was a significant elevation in sCr from baseline at T1 onwards (Table 2 & Fig. 2); however, no dog had an increase in sCr above a published reference interval for the Greyhound breed of 1.1–2.0 mg dL–1 (
). Of the six dogs, three were classified as IRIS grade 2 AKI, one IRIS grade 1 AKI, and two non-AKI (Table 3).
Table 3Renal tubule histologic lesions grade, International Renal Interest Society (IRIS) acute kidney injury (AKI) grade, and maximum magnitude of change in urinary biomarker concentration (calculated as maximum biomarker concentration achieved post-haemorrhage/biomarker concentration following 60 minutes of anaesthesia but before haemorrhage) for six isoflurane anaesthetised greyhounds that underwent a haemorrhage and fluid resuscitation model of AKI. Cr, creatinine; CysC, cystatin C; GGT, gamma-glutamyl transpeptidase; NGAL, neutrophil gelatinase-associated lipocalin; UPC, urine protein creatinine ratio.
Maximum magnitude of change in urinary biomarker concentration
Results of the biomarker analysis are detailed in Table 2, Table 3 and displayed in Fig. 2. The uNGAL concentration was significantly elevated from T2–T4, and uNGAL/Cr elevated from T3 onwards. The urine/serum NGAL ratio was significantly elevated from T2 onwards. uCysC was significantly elevated from T2 onwards, but the urine/serum CysC ratio was elevated at T2 only. The uGGT concentration and uGGT/Cr were significantly elevated from T2 onwards. No significant change was detected for serum NGAL or sCysC concentration during the study.
Histologic examination demonstrated renal tubular lesions consistent with acute ischaemic injury in four dogs: three dogs with grade 1 lesions and one dog with grade 2 lesions (Table 3). The histologic lesions were identified in the cortical proximal tubular cells, predominantly at the tubuloglomerular junction. Lesions included tubular epithelial cell vacuolation, detached intratubular necrotic cells, cellular debris and attenuated epithelial cells (Fig. 3). There were no abnormal findings on histologic examination of renal tissue of the other two dogs.
Furthermore, TEM was performed on the four dogs with histologic lesions (Fig. 3). There was ultrastructural evidence of acute tubular epithelial injury characterized by loss of the apical brush border, nuclear pyknosis, cytoplasmic vacuolation and mitochondrial swelling. In some distal tubular lumens, there was detached cellular debris. Also, two dogs had frequent membrane whorls in proximal tubular cell cytoplasm. Notably, the glomeruli had mild vacuolation of podocyte cytoplasm and mild segmental effacement of their foot processes, but other components of the glomeruli were within normal limits. Taken together, all these features are indicative of acute epithelial cell injury in dogs that had otherwise normal kidneys.
The results support our hypothesis that the urinary biomarkers NGAL, CysC and GGT show changes consistent with AKI earlier than conventional diagnostic methods following I/R injury in anaesthetized dogs. While sustained elevations in urinary concentrations from T2 were documented for all three biomarkers, conventional tests failed to show consistent changes. Identification of renal tubular lesions on LM and TEM, and a lack of temporal changes in serum concentration of NGAL or CysC despite sustained increases in urinary excretion, supports our second hypothesis. We hypothesized that increased excretion of the biomarkers results directly from proximal renal tubular injury, consistent with a recent mechanistic study showing a distinct proinflammatory and profibrotic proximal tubule cell state that fails to repair in AKI (
). During tubular injury, uNGAL is thought to increase via two mechanisms: (1) increased expression and production of NGAL by the distal nephron and (2) reduced ability of damaged proximal tubules to reabsorb circulating NGAL filtered by the kidney. Calculation of the urine/serum ratio of a biomarker may assist in determining whether the biomarker present in the urine is of renal or extra-renal origin. For example, an increased filtered load due to a higher circulating concentration of a biomarker will result in a low urine/serum ratio. While increased renal production/release of biomarker will result in a higher ratio (
). Thus, the increased urine/serum NGAL ratio observed from T2, in combination with histologic evidence of proximal tubule damage, suggests a renal origin for the increased uNGAL during our study. It should be noted that many dogs developed bilirubinuria and had blood in their urine during the study (Table 1), which might have led to falsely high uNGAL concentrations (
). However, bilirubinuria was present in these dogs from T0 and had largely resolved by T4, so is unlikely to explain the large increases in uNGAL observed from T2 onwards. Blood was most frequently present at T1 and T2, resolving by T3, so again is unlikely to explain the large increases in uNGAL that persisted until T4.
Limited data are available on the use of uCysC as a biomarker of AKI in dogs;
reported higher uCysC concentration in dogs with renal disease than in healthy dogs; however, only two dogs with AKI were included in that study. Increased urinary excretion of CysC during AKI is thought to occur as a result of impaired reabsorption by damaged proximal tubule cells. Indeed, uCysC can differentiate between humans with glomerular and tubular lesions (
). In this study, elevated urine/serum CysC ratio at T3 supports a renal origin for the increased uCysC. However, the lack of change in urine/serum CysC at T2 and T4 suggests that increased systemic concentrations of CysC may have been at least partially responsible for elevated uCysC. It is important to note that proteinuria may inhibit uCysC reabsorption, so the increased UPC from T2 onwards may be responsible for the uCysC changes (
) found no difference in uGGT/Cr ratio between dogs with AKI and healthy controls; however, most dogs included in that study were diagnosed as IRIS AKI grade 3 or higher. It is possible that uGGT peaks very early following renal injury, as it is released at the stage of cellular injury and death, then declines during the maintenance phase of AKI when structural cellular injury stabilizes (
). While this early peak may limit usefulness of uGGT for the identification of later stages of AKI, it represents a useful marker at this early, perioperative stage. A significant benefit of uGGT over NGAL or CysC in a clinical setting is the ability to measure it cheaply and quickly using in-house biochemistry analysers.
In the current study, sCr remained within a breed-specific reference interval and, when used according to the IRIS grading criteria, failed to classify two of the six dogs as having AKI despite histologic evidence of tubular injury in these two dogs. The increase in sCr from baseline to T1 probably reflects the impact of haemorrhage and fluid administration on GFR and volume status, rather than a sustained reduction in renal function. Oliguria is a well-recognized symptom of AKI and utilized in the IRIS AKI grading criteria, but not a consistent finding following I/R injury in our study (
). Of the variables measured during urinalysis, sediment examination is arguably one of the most useful indicators of acute tubular damage, the presence of epithelial cells and casts in urine reflecting renal tubular cell injury and death (
). However, in this study, epithelial cells were not present in the urine following I/R injury, and cylinduria occurred in only two dogs. Of interest to us, UPC was consistently elevated in all dogs following I/R injury. Failure of the glomerular filtration barrier allows large proteins such as albumin to enter the kidney and subsequently to be excreted in the urine as tubular reabsorption pathways become saturated (
). As glomerular dysfunction is not a feature of early AKI, UPC is not always considered during diagnostic testing for AKI. Given the lack of glomerular lesions on LM or TEM, the increased UPC observed may reflect failure of the damaged tubules to reabsorb filtered proteins, and increased production and release of proteins by the kidneys themselves (e.g., NGAL, GGT). While IV infusion of gelatine solutions causes false proteinuria with some protein measurement techniques, the turbidimetric method we used should not have been affected by colloid infusion (
In the current study, in the two dogs with no histologic evidence of acute tubular injury, elevations in uNGAL, uCysC and uGGT were still observed. This is probably because the urinary profile represents the output of both entire kidneys, whereas only small portions of the kidney can be evaluated microscopically. Additionally, evidence indicates that uNGAL is a more sensitive marker of AKI than urinalysis in humans, suggesting that increased biomarker expression may commence prior to significant cellular injury (
). At this early stage of AKI, sublethal damage to renal tubular cells might occur without any obvious change on LM, for example, the redistribution of cell surface receptors results in tubular dysfunction owing to altered cell polarity (
A limitation of this study was the use of a gelatine infusion for resuscitation. Renal tubular epithelial cell lesions have been identified in animals following large infusions of synthetic colloids, including gelatine. It is possible that the histologic lesions and biomarker changes identified in this study may not result from I/R injury alone. However, lesions commonly associated with synthetic colloid administration are osmotic nephrosis such as severe vacuolation and microvesiculation (
). The reason for this is unclear, but the authors hypothesized that induction of NGAL by a gelatine substrate may occur. However, elevated uNGAL subsequent to experimental I/R injury AKI in the absence of gelatine administration is well documented in other species, so the uNGAL changes we observed probably reflect I/R injury AKI in some part (
). Regardless, further investigation of uNGAL during early AKI in dogs that have not received synthetic colloid infusions are required.
To conclude, our model of I/R injury AKI showed that urinary concentrations of NGAL, CysC and GGT were elevated before an elevation in sCr or reduction in urine output, which are currently used to diagnose AKI. Furthermore, serum concentrations of these biomarkers did not change, confirming their renal origin. Further work is required to document longer-term changes in these markers, and their ability to predict long-term outcomes including renal recovery after severe AKI.
JD: study design, data collection, laboratory analysis, data interpretation, statistical analysis, preparation of manuscript. GR and RC: study design, laboratory analysis, data interpretation, review and editing of manuscript. KH and DM: study design, review and editing of manuscript. GH: study design, statistical analysis, review and editing of manuscript. AR: study design, data collection, data interpretation, review and editing of manuscript.
Conflict of interest statement
Authors declare no conflict of interest.
The authors would like to acknowledge and thank the contribution of Dr Mary Nabity for assistance with study design, and Dr Andrew Currie and Dr Kirsty Townsend for assistance with Luminex equipment. Renal histopathology was performed by the Comparative Pathology & Digital Imaging Shared Resource, Department of Veterinary Biosciences and the Comprehensive Cancer Center, The Ohio State University, Columbus, OH, USA. Renal electron microscopy was performed by the division of Nephropathology, Department of Pathology, The Ohio State University, Columbus, OH, USA. Funding for this work was assisted by a Murdoch University School of Veterinary and Life Sciences Small Grant (2016) and through support received through an Australian Government Research Training Program Scholarship. Renal histopathology was supported in part by grant P30 CA16058 , National Cancer Institute , MD, USA .
Structure and expression of the human cystatin C gene.