Quantification of human urinary exosomes by nanoparticle tracking analysis

Abstract

Exosomes are vesicles that are released from the kidney into urine. They contain protein and RNA from the glomerulus and all sections of the nephron and represent a reservoir for biomarker discovery. Current methods for the identification and quantification of urinary exosomes are time consuming and only semi-quantitative. Nanoparticle tracking analysis (NTA) counts and sizes particles by measuring their Brownian motion in solution. In this study, we applied NTA to human urine and identified particles with a range of sizes. Using antibodies against the exosomal proteins CD24 and aquaporin 2 (AQP2), conjugated to a fluorophore, we could identify a subpopulation of CD24- and AQP2-positive particles of characteristic exosomal size. Extensive pre-NTA processing of urine was not necessary. However, the intra-assay variability in the measurement of exosome concentration was significantly reduced when an ultracentrifugation step preceded NTA. Without any sample processing, NTA tracked exosomal AQP2 upregulation induced by desmopressin stimulation of kidney collecting duct cells. Nanoparticle tracking analysis was also able to track changes in exosomal AQP2 concentration that followed desmopressin treatment of mice and a patient with central diabetes insipidus. When urine was stored at room temperature, 4°C or frozen, nanoparticle concentration was reduced; freezing at −80°C with the addition of protease inhibitors produced the least reduction. In conclusion, with appropriate sample storage, NTA has potential as a tool for the characterization and quantification of extracellular vesicles in human urine.

Key points

• Exosomes are vesicles that are released from the kidney into the urine. They contain RNA and protein from the cell of origin and can track changes in renal physiology non-invasively.

• Current methods for the identification and quantification of urinary exosomes are time consuming and only semi-quantitative.

• In this study, we applied nanoparticle tracking analysis to human urine and identified particles with a range of sizes, including a subpopulation of characteristic exosomal size that labelled positively with antibodies to exosome proteins.

• Nanoparticle tracking analysis was able to track an increase in exosomal aquaporin 2 concentration following desmopressin treatment of a kidney cell line, a rodent model and a patient with central diabetes insipidus.

• With appropriate sample storage, nanoparticle tracking analysis has potential as a tool for the rapid characterization and quantification of exosomes in human urine. This new method can be used to develop urinary extracellular vesicles further as a non-invasive tool for investigating human renal physiology.

Introduction

Exosomes are vesicles that are released from a wide range of cell types into biological fluids, including urine (Pisitkun et al. 2004). Urinary exosomes contain proteins and RNA species originating from cells of the renal glomerulus and each region of the nephron (Gonzales et al. 2010). Their cargo changes with kidney injury (Zhou et al. 2008), presenting an opportunity to track changes in intracellular pathways, which may precede a decline in renal function or represent novel therapeutic targets, without need for an invasive tissue biopsy.

At present, a panel of physicochemical properties are reported to distinguish exosomes from other extracellular vesicles present in urine. Exosomes are reported to measure 20−100 nm and appear cup shaped when visualized by transmission electron microscopy (Théry et al. 2001), have a density of 1.10−1.19 g ml⁻¹ (Keller et al. 2007) and contain proteins that are central to their production (Théry et al. 2009). These properties are, however, time consuming to measure and only semi-quantitative.

There is a pressing need for new technologies that can measure extracellular vesicles, including exosomes, in urine rapidly and accurately with minimal sample preparation. This would allow excretion in animal models and humans to be quantified and, therefore, the effect of physiological changes and disease on vesicle release to be defined. The current lack of precise quantification of urinary exosome concentration also significantly compromises RNA and protein biomarker discovery studies, because existing methods for quality control and normalization across study groups are inadequate (Dear et al. 2013).

Nanoparticle tracking analysis (NTA) is a technology that can size and count nanoparticles, such as those released from cultured cells (Soo et al. 2012) and in human plasma (Lässer et al. 2011). Nanoparticle tracking analysis is based on the principle that at any particular temperature, the rate of Brownian motion of nanoparticles in solution is determined solely by their size. In this method, laser light is directed at a fixed angle to the vesicle suspension, and the scattered light is captured using a microscope and high-sensitivity camera. By tracking the movement of individual nanoparticles over time, the software rapidly calculates their concentration and size. Published studies demonstrate that NTA can count and size specific subgroups of particles using fluorescent antibodies against surface proteins (Dragovic et al. 2011), but this has not yet been applied to urine.

In the present study, we applied NTA to human urine and identified a range of nanoparticles, including those currently classified as exosomes. Nanoparticle tracking analysis tracked increases in exosomal aquaporin 2 (AQP2) concentration following desmopressin treatment of a murine kidney cell line, a rodent model and a patient with central diabetes insipidus (CDI). We further used NTA to demonstrate that the urinary nanoparticle yield is labile and that rapid freezing to −80°C with addition of protease inhibitors proved the best approach to urine storage. Nanoparticle tracking analysis holds potential for the quantification of exosomes and other extracellular vesicles in human urine.

Methods

Urine sample collection

The second urine sample after waking was obtained from healthy volunteers (n= 5; 2 male, 3 female; mean age 25 years, range 21−29 years). Repeated urine samples were obtained over two separate time periods from a 16-year-old male with CDI secondary to a craniopharyngioma, who was being routinely treated with daily desmopressin [dDAVP nasal spray; 0.1 ml (10 μg) desmopressin acetate per spray]. The CDI patient samples were initially stored at 4°C, then frozen at −80°C. Analysis of the CDI patient samples was performed by a researcher blinded to the timing of desmopressin treatment. The protocol was agreed by the institutional ethical review body, and informed consent was obtained from the volunteers and patient included in the study. None of the volunteers was taking any medication.

Measurement of particle size and concentration distribution with NTA

Nanoparticles in the whole urine samples and isolated exosome suspensions were analysed using the NanoSight LM 10 instrument (NanoSight Ltd, Amesbury, UK). The analysis settings were optimized and kept constant between samples, and each video was analysed to give the mean, mode, median and estimated concentration for each particle size. Following published methods (Sokolova et al. 2011) and initial pilot studies comparing whole urine samples and 1:1000 dilution, all experiments were carried out at a 1:1000 dilution, yielding particle concentrations in the region of 1 × 10⁸ particles ml⁻¹ in accordance with the manufacturer's recommendations. All samples were analysed in triplicate. The concentration of particles in urine was expressed per mmol urinary creatinine. Urine creatinine concentration was measured by the creatininase/creatinase specific enzymatic method using a commercial kit (Alpha Laboratories Ltd, Eastleigh, UK) adapted for use on a Cobas Fara centrifugal analyser (Roche Diagnostics Ltd, Welwyn Garden City, UK).

Isopycnic centrifugation

This was performed as previously described (Zhou et al. 2008).

Ultracentrifugation

Exosomes were isolated from the whole urine samples as previously described (Gonzales et al. 2010).

ExoQuick reagent

Exosomes were isolated using the modified exosome precipitation method as previously described (Alvarez et al. 2012).

Fluorescent labelling with antibody conjugated to quantum dots

Anti-CD24 antibody was a kind gift of Dr P. Altevogt (German Cancer Research Center, Heidelberg, Germany). Anti-AQP2 antibody was purchased from Millipore (Billerica, MA, USA). Mouse IgG antibody was purchased from Invitrogen (Paisley, UK). Following the manufacturer's protocol, quantum dots (Qdots) were conjugated to antibodies with a Qdot 605 Antibody Conjugation Kit (Invitrogen). For fluorescent NTA analysis, a 532 nm (green) laser diode excited the Qdots with a long-pass filter (430 nm) so that only fluorescent particles were tracked and labelled particle concentration determined by NTA software.

Cell culture model of exosome release

Murine kidney collecting duct (mCCDC11) cells were grown in culture as previously described (Street et al. 2011), then stimulated with desmopressin (Sigma-Aldrich, St Louis, MO, USA), 3.16 ng ml⁻¹ for 48 and 96 h. The cell culture medium (2 ml) was then analysed by NTA following fluorescent labelling with AQP2–Qdot.

Mouse urine study

Mice (C57/BL6; n= 6) were individually housed in metabolic cages (model 3600M021; Techniplast, Buguggiate, Italy) with free access to food and water. After acclimation, daily food and fluid intakes were measured, as well as body weight. Each mouse received a single subcutaneous injection of 0.9% NaCl (1 μl (g body weight)⁻¹) on days 1 and 2 and, after each injection, a 24 h urine collection was performed. On days 3 and 4, mice received a subcutaneous injection of desmopressin (1 μl (g body weight)⁻¹ of 10 μg ml⁻¹ drug solution), and two further 24 h urine collections were performed. A second cohort of mice (n= 5) was used as control animals, receiving subcutaneous injections of 0.9% NaCl on all 4 days. All studies were performed with the appropriate Home Office (UK) licence.

Evaluation of optimal storage conditions for nanoparticles with NTA

Freshly obtained urine samples (60 ml each) from each of the volunteers (n= 5) were subjected to four different storage protocols (6 ml 1:1000 dilution per protocol in three 2 ml aliquots), all with and without protease inhibitors (1:10 final concentration: 0.5 mm phenylmethylsulfonyl fluoride and 20 μm leupeptin, both from Sigma-Aldrich). The protocols were as follows: (i) analysed immediately with NTA; (ii) stored at room temperature (RT), 4, −20 or −80°C for 2 h; (iii) stored at room temperature, 4, −20 or −80°C for 1 day; and (iv) stored at room temperature, 4, −20 or −80°C for 1 week.

All samples were subjected to vortexing during thawing, following the recommendation of Zhou et al. (2006). Each NTA measurement for the different protocols for each subject was repeated in triplicate.

Statistical analysis

Unless otherwise indicated, analyses were performed on data generated from triplicate NTA results. Data were analysed using IBM SPSS Statistics 19 (SPSS Inc., Chicago, IL, USA) and GraphPad Prism Version 6 (GraphPad Software, La Jolla, CA, USA). The area under the curve (AUC) was determined following the trapezoidal rule. Within-subject variation (CVi) was calculated as the coefficient of variation for the AUC between the NTA triplicate measurements for each subject as a ratio between SD and mean. Univariate non-parametric Wilcoxon t tests and ANOVAs were used to determine significant differences between different storage conditions expressed as the percentage change from baseline AUC values. A value of P < 0.05 was the level of nominal significance.

Results

Nanoparticle tracking analysis identified nanoparticles in human urine

First, we investigated the size distribution of particles in human urine samples using NTA. Urine (2 ml aliquots of 1:1000 dilution) from five study participants was obtained, put on ice and applied to the sample chamber of the Nanosight LM10. The diameter of the particles ranged from 26 to 700 nm (Fig. 1A). The NTA software identified and measured particles in the expected exosomal size range of 20−100 nm (Fig. 1B). Due to the light intensity of the larger particles causing possible overestimation of the smaller particle size, following published recommendations (Dragovic et al. 2011) we focused only on particles sized <300 nm. The area under the particle size vs. concentration curve for particles in the 20−100 nm range (AUC₂₀₋₁₀₀) was 14.6 ± 2.3% of the total AUC for the 0−300 nm curve.

Figure 1. Nanoparticle tracking analysis (NTA) of human urine samples.

A, screen shot from 1:1000 diluted whole urine sample revealing a range of particle sizes. B, example NTA trace depicting the nanoparticle distribution profile for an individual subject. C, concentration of urinary particles (0–300 nm diameter) in urine samples from five study participants. The concentration is expressed as number of particles per mmol urinary creatinine.

Given that exosomes are commonly accepted as being 20−100 nm in diameter, we focused on particles within this size range using AUC₂₀₋₁₀₀ to quantify their concentration. Each urine sample was analysed in triplicate. In unprocessed urine, the within-sample coefficient of variability (CVi) was 47% for AUC₂₀₋₁₀₀. Across the five participants, the median AUC₂₀₋₁₀₀ (interquartile range; IQR) was 0.07 × 10⁶ particles (mmol creatinine)⁻¹ (0.05−0.1 particles (mmol creatinine)⁻¹). The median (IQR) particle size was 74 nm (51−80 nm; Fig. 1C).

Fluorescent NTA identified antibody-labelled exosomes in human urine

To determine whether the particles of 20−100 nm diameter carried exosome-associated markers, we labelled urine with an antibody against CD24, a surface marker of human urinary exosomes (Keller et al. 2007) or AQP2, an archetypal exosomal protein (Pisitkun et al. 2004). First, we performed isopycnic centrifugation, using a sucrose gradient, of human urinary exosomes that had been concentrated by ultracentrifugation (UC; Fig. 2A). CD24 localized with the established exosomal proteins, TSG101 and flotillin-1, in the range of 1.12−1.16 g cm⁻³, which is the characteristically accepted density of exosomes. This is consistent with CD24 labelling the exosomes. Following this, the CD24 antibody was conjugated to a quantum dot fluorescent label (CD24–Qdot).

Figure 2. Exosome marker specific fluorescent labelling.

A, CD24, flotillin-1 and TSG101 localize to a density range of 1.12–1.16 g cm⁻³ on a sucrose density gradient following isopycnic centrifugation. The positive control was human urinary exosomes unseparated by isopycnic centrifugation. B, human urine samples labelled with CD24–quantum dots (Qdots) in light scatter mode (dashed line, all particles) and with the fluorescent filter in place (continuous line, CD24-labelled particles). Results from three study participants are presented. Particle concentration is expressed per mmol urinary creatinine. C, human urine samples labelled with aquaporin 2 (AQP2)–Qdots in light scatter mode (dashed line, all particles) and with the fluorescent filter in place (continuous line, AQP2-labelled particles). Results from three study participants are presented. Particle concentration is expressed per mmol urinary creatinine. D, human urine labelled with mouse IgG conjugated to Qdot as an isotype control in light scatter mode (dashed line) and with the fluorescent filter in place (continuous line). Note the absence of a ‘peak’ in the particle size range 0–100 nm with the fluorescent filter in place.

The particle size and concentration curve in urine samples labelled with CD24–Qdot antibody was measured by NTA in light scatter mode (Fig. 2B; dashed line, all particles) and again with a 430 nm fluorescent filter (Fig. 2B; continuous line, CD24-labelled particles). Light scatter mode revealed a particle size range 20−300 nm. However, with the fluorescent filter in place, the CD24–Qdot labelled particles were smaller [median size (IQR) 37 nm (31−50 nm)] consistent with CD24 binding to exosomes. The particle size and concentration curve in urine samples labelled with AQP2–Qdot antibody was also measured by NTA (Fig. 2C), and the median size (IQR) was 69 nm (51−91 nm), which is consistent with AQP2 also binding to exosomes. With regard to the CD24-labelled particles, the within-sample variability in AUC₂₀₋₁₀₀ was significantly less than unlabelled samples presented in Fig. 1 (CVi = 35%, P < 0.05). Across the five participants, the median CD24-labelled AUC₂₀₋₁₀₀ (IQR) was 0.1 × 10⁶ particles (mmol creatinine)⁻¹ (0.07−0.13 × 10⁶ particles (mmol creatinine)⁻¹) and the median AQP2-labelled AUC₂₀₋₁₀₀ (IQR) was 0.2 × 10⁶ particles (mmol creatinine)⁻¹ (0.16−0.29 × 10⁶ particles (mmol creatinine)⁻¹). As a control, urine (1:1000 dilution) was labelled with mouse IgG–Qdot (Fig. 2D). The AUC₂₀₋₁₀₀ with the fluorescent filter was ∼8-fold less than with CD24.

To determine whether isolating exosomes from human urine would improve the purity (fewer non-CD24-positive particles of size >100 nm) and reduce variability of NTA measurements (lower CVi) we used two established methods, i.e. UC and Exoquick™ reagent. Figure 3 demonstrates particle size vs. concentration curves for the CD24–Qdot-conjugated UC-concentrated exosomes. In light scatter mode, UC-concentrated samples still contained non-exosomal sized particles. However, with the fluorescent filter in place, the CD24–Qdot-labelled particles were smaller [median size (IQR) 62 nm (21−91 nm)], consistent with CD24 binding to exosomes. CD24–Qdot-conjugated UC-isolated exosomes had a smaller coefficient of variation within sample (CVi = 24%) compared with whole urine samples (P= 0.02). Across the five participants, the median CD24 labelled AUC₂₀₋₁₀₀ (IQR) was 0.23 × 10⁶ particles (mmol creatinine)⁻¹ (0.19−0.32 × 10⁶ particles (mmol creatinine)⁻¹; Fig. 3A).

Figure 3. Comparison of different exosome isolation methods.

A, human urinary exosomes concentrated by ultracentrifugation conjugated with CD24–Qdots in light scatter mode (dashed line, all particles) and with the fluorescent filter in place (continuous line, CD24-labelled particles). Results from three study participants are presented. Particle concentration is expressed per mmol urinary creatinine. B, human urinary exosomes concentrated by Exoquick™ reagent conjugated with CD24–Qdots in light scatter mode (dashed line, all particles) and with the fluorescent filter in place (continuous line, CD24-labelled particles). Results from three study participants are presented. Particle concentration is expressed per millimolar urinary creatinine.

Figure 3B shows particle size vs. concentration curves for the CD24–Qdot-conjugated Exoquick™-isolated exosomes. Similar to the data from the UC-isolated exosomes, the light scatter trace for Exoquick™-treated samples revealed a range of particle sizes of 20−300 nm, consistent with the presence of non-exosomal particles. However, with the fluorescent filter in place, the CD24–Qdot-labelled particles were smaller [median size (IQR) 60 nm (41−98 nm)], again consistent with CD24 selectively binding to exosomes. Compared with the UC isolation method, CD24–Qdot-conjugated Exoquick™-isolated exosomes had a high coefficient of variation within samples (CVi = 56%, P < 0.05). Across the five participants, the median CD24-labelled AUC₂₀₋₁₀₀ (IQR) was 0.04 × 10⁶ particles (mmol creatinine)⁻¹ (0.02−0.08 × 10⁶ particles (mmol creatinine)⁻¹) for Exoquick™-isolated exosomes.

Nanoparticle tracking analysis detected changes in exosome composition with minimal sample processing

We have previously demonstrated that murine kidney collecting duct (mCCDC11) cells release AQP2-containing exosomes following stimulation with the vasopressin analogue desmopressin (Street et al. 2011). However, this prevous study used UC processing of large volumes of cell culture medium and Western blotting to demonstrate AQP2 upregulation, which is a time-consuming, labour-intensive approach. We tested whether NTA could detect AQP2 upregulation in exosomes without any preprocessing of the cell culture medium. Nanoparticle tracking analysis detected significant differences in AQP2-expressing exosomal concentrations following desmopressin stimulation for 96 h in our cell line (Fig. 4A). Following this, NTA was applied to urine samples collected from mice injected over consecutive days first with saline and then with desmopressin. Desmopressin significantly reduced urine flow rate (data not shown). Nanoparticle tracking analysis detected a significant increase in urinary AQP2-expressing exosomes following desmopressin treatment (Fig. 4B). Urine samples from a CDI patient treated with desmopressin were analysed, and there was a clear increase in AQP2-expressing exosomes immediately following administration of desmopressin (Fig. 5).

Figure 4. Changes in AQP2-positive exosomes following desmopressin stimulation.

A, the difference in exosomal concentration in the cell culture media is expressed as the area under the curve (AUC) for particles sized 20–100 nm that labelled with an antibody to AQP2. The cells were stimulated with desmopressin (3.16 ng ml⁻¹) for 48 or 96 h. *P < 0.05 T-TEST. B, difference in exosomal concentration in the urine samples from desmopressin-treated (n= 6) or control mice (n= 5), expressed as the AUC for particles sized 20–100 nm that labelled with an antibody to AQP2. Particle concentration is expressed per mmol urinary creatinine. *P < 0.05 T-TEST.

Figure 5. Nanoparticle tracking analysis tracked changes in AQP2-positive exosome concentration following desmopressin treatment of a patient with central diabetes insipidous.

For A and B, urine aliquots were collected over 2 separate days. The AQP2-exosome concentration in the urine samples is expressed as the AUC for particles sized 20–100 nm that labelled with an antibody to AQP2. Particle concentration is expressed per mmol urinary creatinine. The time of administration of desmopressin treatment is indicated by the dashed line.

Evaluation of optimal storage methods for urine

Nanoparticle tracking analysis was used to assess the effect of different urine storage protocols on urinary exosome concentation. Urine (1:1000 dilution) was analysed immediately after collection, which acted as the baseline for comparison of the different storage conditions with and without addition of protease inhibitors (RT, 4, −20 and −80°C; data not shown for all storage temperatures with added protease inhibitors). For each condition, the nanoparticle concentration in the AUC₂₀₋₁₀₀ range was assessed again after 2 h, 1 day and 1 week. Importantly for future study design, a significant decrease in the nanoparticle yield in the AUC₂₀₋₁₀₀ range was observed with time, regardless of storage condition (P < 0.05 for all protocols). Indeed, our data suggest exosomal degradation within 2 h of urine collection. In this context, storage at −80°C with addition of protease inhibitors resulted in substantially less reduction in AUC₂₀₋₁₀₀ compared with other storage conditions (Fig. 6).

Figure 6. Different storage protocols and urine particle concentration.

Nanoparticle tracking analysis was used to measure the particle concentration between 20 and 100 nm in diameter (AUC₂₀₋₁₀₀) after storage at various temperatures for 2 h, 1 day or 1 week. The baseline was immediate measurement after sample collection. n= 5 per group. *P < 0.05 (ANOVA) for significant differences between protease inhibition and none. Abbreviations: RT, room temperature; and PI, protease inhibitors.

Discussion

The potential of urinary extracellular vesicles, such as exosomes, as a biomarker reservoir for studying kidney physiology and disease has been extensively reviewed (Dear et al. 2013). However, the inability to identify and quantify the presence of exosomes in clinical specimens rapidly and accurately remains a translational roadblock (Hosseini-Beheshti et al. 2012). Current, standardized methods for investigating the distributions of exosomal size and concentration are time consuming and only semi-quantative. The aim of our study was to apply a recent technological advance, NTA, to determine whether this holds potential for the identification of nanoparticles, including exosomes, in human urine samples. Also, we optimized storage conditions for urinary exosomes by using NTA to determine particle loss in different storage conditions. We demonstrated that when whole urine samples were applied to the Nanosight chamber, NTA could determine the size and concentration distribution of nanoparticles sized between 26 and 700 nm, and a significant percentage of the overall particle distribution was within the expected exosomal size range (20−100 nm). Labelling with two different exosomal proteins confirmed the presence of exosome-sized particles in unprocessed urine, which provides ‘proof-of-concept’ data demonstrating that NTA can rapidly quantify exosome concentration without time-consuming sample processing. However, the intra-assay variability was substantially reduced by using an antibody specific fluorescent label compared with unlabelled urine samples. With the fluorescent filter in place, we were able to visualize a larger concentration of particles sized between 20 and 100 nm compared with the light scatter mode. This may be due to the intense light scatter from larger particles interfering with the accurate and reproducible measurement of smaller particles (Dragovic et al. 2011). For the purpose of the present study, a sample size of five was used to provide ‘proof of concept’; this number of subjects was sufficent to provide data with relatively small interquartile ranges and will provide the necessary information concerning variability that can now be used in future power calculations.

Using two commonly used methods of isolating urinary exosomes, i.e. UC and Exoquick™, we demonstrated that only UC resulted in a further reduction in intra-assay variability, but it is also a labour-intensive method. It is important to note that both UC and Exoquick™ isolated non-exosomal particles from human urine, and caution must be exercised not to assume incorrectly that these techniques result in a pure exosomal preparation. This may lead to incorrect conclusions regarding the protein or RNA content of exosomes or their biological activity.

The protein composition of exosomes can track changes in the proteome of the cell. We previously demonstrated this using mCCDC11 cells. However, the measurement of exosomal AQP2 upregulation relied on Western blotting, preceded by UC to concentrate exosomes from approximately 20 ml of culture medium. In the present study, using only 2 ml of culture medium, NTA clearly identified AQP2 upregulation in exosomes with no sample processing. Following on from the results from our cell line, we applied NTA to urine samples collected from mice before and after desmopressin treatment. Nanoparticle tracking analysis was able to report differences in urine AQP2-positive exosomal concentrations between treated and non-treated conditions. The average urine flow rate in the mouse is ∼1 ml (24 h)⁻¹, and this small urine volume results in low exosome yields following current protocols, such as ultracentrifugation. Nanoparticle tracking analysis can rapidly detect exosome protein changes in small volumes of mouse urine without extensive sample processing, which represents a significant advance for non-invasive, longitudinal physiological studies in the mouse.

Applied to urine samples from a CDI patient treated with desmopressin, NTA was able to track changes in AQP2-positive exosome concentrations over time. Desmopressin resulted in a rapid increase in AQP2-expressing exosomes, which is consistent with studies that have reported rapid increases in total urinary AQP2 following subcutaneous administration of desmopressin (Kanno et al. 1995). This increase is too rapid to represent new protein synthesis and is likely to reflect cytoplasmic AQP2 transfer to the cell membrane. Nanoparticle tracking analysis represents a rapid-throughput alternative to the current Western blot protocols. Further in vivo studies are now needed to determine whether NTA can quantify the upregulation of other proteins in urinary exosomes and how this change relates to tubular protein expression.

Finally, by using NTA to evaluate storage methods for urinary nanoparticles, we found that −80°C with protease inhibition was the optimal approach for storage of whole urine samples, resulting in the maximal preservation of urinary nanoparticles compared with the other temperatures (RT, 4 and −20°C) with or without protease inhibition. This is consistent with previous published work that used Western blotting for exosomal marker proteins as a read-out for exosome concentration (Zhou et al. 2006). Importantly, however, we found a significant loss of nanoparticles for all storage conditions, even within 2 h of obtaining the sample.

We have demonstrated the potential of NTA in the identification of urinary exosomes. Development of this technology could allow rapid quantification of kidney tubular exosome excretion in health and disease and provide a crucial quality control for protein or RNA biomarker discovery studies.

Glossary

Abbreviations

AQP2

aquaporin 2

AUC

area under the curve

CD24

cluster of differentiation 24

CDI

central diabetes insipidus

CVi

within-subject coefficient of variability

IQR

interquartile range

mCCDC11

murine cortical collecting duct cell line

NTA

nanoparticle tracking analysis

Qdot

quantum dots

RT

room temperature

TSG101

tumour susceptibility gene 101

UC

ultracentrifugation

Additional information

Competing interests

None declared.

Author contributions

Experiments were performed in the laboratory of J.W.D. W.O. performed the majority of the studies. Immunoblotting was performed by J.M.S. All animal work and urine collection was performed by N.E.L.S. and J.R.I. Exoquick™ exosome isolation was performed by E.J.T. The studies were designed by W.O., M.A.B. and J.W.D. Most of the data analysis was performed by W.O. The article was drafted by J.W.D. and revised critically by all authors. All the authors approved the final version.

Funding

The authors acknowledge the contribution of the British Heart Foundation Centre of Research Excellence Award. J.W.D. acknowledges the financial support of NHS Research Scotland (NRS), through NHS Lothian. Funding was received from the Diabetes Research and Wellness Foundation and NC3R.