characterization of the canine urinary proteome

O R I G I N A L R E S E A R C H

Characterization of the canine urinary proteomeLaura E. Brandt1, E. J. Ehrhart1,2, Hataichanok Scherman4, Christine S. Olver1, Andrea A. Bohn1,Jessica E. Prenni3,4

1Department of Microbiology, Immunology and Pathology; 2Animal Cancer Center; 3Department of Biochemistry and Molecular Biology; and4Proteomics and Metabolomics Facility, Colorado State University, Fort Collins, CO, USA

KeyWords

Biomarker, canine, exosome, proteomics,

urine

Correspondence

Jessica E. Prenni, Department of Biochemistry

and Molecular Biology, and Proteomics and

Metabolomics Facility, 2021 Campus Delivery,

Colorado State University, Fort Collins, CO

80523, USA

E-mail: [email protected]

DOI:10.1111/vcp.12147

Background: Urine is an attractive biofluid for biomarker discovery as it is

easy and minimally invasive to obtain. While numerous studies have

focused on the characterization of human urine, much less research has

focused on canine urine.

Objectives: The objectives of this study were to characterize the uni-

versal canine urinary proteome (both soluble and exosomal), to deter-

mine the overlap between the canine proteome and a representative

human urinary proteome study, to generate a resource for future

canine studies, and to determine the suitability of the dog as a large

animal model for human diseases.

Methods: The soluble and exosomal fractions of normal canine urine were

characterized using liquid chromatography tandem mass spectrometry

(LC-MS/MS). Biological Networks Gene Ontology (BiNGO) software was

utilized to assign the canine urinary proteome to respective Gene Ontology

categories, such as Cellular Component, Molecular Function, and Biologi-

cal Process.

Results: Over 500 proteins were confidently identified in normal canine

urine. Gene Ontology analysis revealed that exosomal proteins were lar-

gely derived from an intracellular location, while soluble proteins included

both extracellular and membrane proteins. Exosome proteins were

assigned to metabolic processes and localization, while soluble proteins

were primarily annotated to specific localization processes. Several proteins

identified in normal canine urine have previously been identified in

human urine where these proteins are related to various extrarenal and

renal diseases.

Conclusions: The results of this study illustrate the potential of the dog as

an animal model for human disease states and provide the framework for

future studies of canine renal diseases.

Introduction

In recent years, many researchers have utilized mass

spectrometry to characterize the human urinary prote-

ome1–7, and several of these studies have revealed

promising potential biomarkers specific for both renal

disease and extrarenal pathologic conditions.2–4,8–12

Urine is an attractive biofluid for biomarker discovery

as it is easy to obtain byminimally invasive procedures.

In addition, it is abounding with proteins, making

it ideal for diagnostic testing. The application of pro-

teomics for the analysis of canine urine so far has been

limited. Most studies have been limited in scope

and have utilized technologies that focus on the

identification of select proteins, the presence of which

has been found to differ according to environmental

states or disease.2,3,13,14 For example, two-dimensional

gel electrophoresis followed by mass spectrometry was

used to identify specific proteins related to tubulointer-

stitial injury in a canine model of progressive glomeru-

lar disease.15 In another study, surface-enhanced laser

desorption/ionization time-of-flight mass spectrome-

try was used to perform protein-level profiling of urine

from dogs with and without renal disease. 14 However,

none of these studies provided a comprehensive

proteomic profile of canine urine.

Vet Clin Pathol 0/0 (2014) 1–13©2014 American Society for Veterinary Clinical Pathology and European Society for Veterinary Clinical Pathology 1

Veterinary Clinical Pathology ISSN 0275-6382

Comprehensive proteomic analysis of any bio-

fluid, including urine, is difficult due to the qualitative

and quantitative complexity and wide range of pro-

teins. Fortunately, the analysis of urine can be focused

by targeting exosomes, which, due to their unique

acquisition of vesicular membranes and liquid

contents, provide very specific information about the

cells from which they originate. Exosomes are formed

by a stepwise process. When plasma membrane pro-

teins are endocytosed into a cell, endocytic vesicles are

generated. By inward budding, these endocytic vesi-

cles become multivesicular bodies (MVB), which con-

tain their own vesicles. After moving back to the cell

surface, the MVB bind to the cellular plasma mem-

brane, and release their internal vesicles, the exo-

somes, outside of the cell. Exosomes have a

cytoplasmic-side inward orientation, similar to intact

cells, and are thus loaded with specialized information

about their originating cells.16 They contain both

membranous and cytoplasmic constituents and are

generally excreted from cells for specific reasons, such

as cell-to-cell signaling or to act as membrane transport

proteins.16

The urinary exosomal population is thought to be

comprised largely of exosomes from renal and other

urinary tract cells. However, exosomes are also found

in serum and, due to their generally small size, can be

freely filtered by the glomerulus into the urinary

space.5 Theoretically, if a nonrenal disease causes an

increase of specific exosomes in serum, their proteins

may be found in greater abundance in urine. There-

fore, the analysis of the urine subproteome has the

potential to provide important information about the

urinary tract as well as systemic health, even though it

may be impossible to determine the actual origin of

urinary exosomes.

Thus, the goal of our current study was to adapt

proteomic methods established for the investigation of

human urine to canine urine, to characterize both ex-

osomal and soluble proteins. A comprehensive com-

parison based on protein annotation and gene

ontology was performed between our results and pub-

lished data of the human urinary proteome.1 Several

proteins previously identified in human urine and

associated with various disease states, including extra-

renal disease, renal disease, systemic hypertension,

and neoplasia, were also identified in canine urine.17–

26 While making an association between the observa-

tions of these proteins in canine urine with similar

disease processes requires further investigation, their

presence illustrates the potential for veterinary diag-

nostics as well as the use of canine models for human

conditions using investigative proteomic techniques.

Materials andMethods

Canine urine collection and protein concentration

Inclusion criteria required that dogs be between one

and 10 years of age, clinically normal, not taking any

medications or supplementations, and fed a commer-

cial adult dog food diet. Clinically normal was defined

as a healthy animal with unchanged biologic and social

functions in its known environment, and no abnor-

malities on physical examination. The health of all

dogs was further evaluated by routine CBC, including

evaluation of blood film, serum chemistry profile, uri-

nalysis, and urine protein to creatinine ratio (UPC).

Protocols for animal handling and sample collection

were reviewed and approved by the Colorado State

University Animal Care and Use Committee (ACUC).

Fifty milliliters of urine were collected by free

catch into sterile urine collection cups from all dogs.

A protease inhibitor cocktail (Sigma-Aldrich, St.

Louis, MO, USA) was added at a ratio of 1:10 to each

sample immediately after collection to avoid protein

degradation.27 Samples were immediately frozen to

�80°C, they were stored until all samples were col-

lected and could be analyzed in batch within

12 months of collection.27 Samples were thawed in

cool water and then kept on ice during processing.

After vortexing to reconstitute particles that may

have settled during the freezing process, all samples

were mixed together. Pooling was performed to

enable the characterization of the broad canine uri-

nary proteome across multiple breeds. While this

approach does not allow for observation of individual

differences between dogs, it was chosen to represent

a general population. Batched urine was aliquoted

into 6 ultracentrifuge tubes (25 9 89 mm; Beckman,

Brea, CA, USA), and centrifuged at 17,000g for

10 min at 4°C (L70 Ultracentrifuge, SW 28 rotor;

Beckman) to remove whole cells and debris.5 The

sample supernatants were then collected and aliquot-

ed into 6 clean ultracentrifuge tubes (Beckman).

Finally, the samples were ultracentrifuged at 113,000g

for 2 h at 4°C (L70 Ultracentrifuge, SW 28 rotor; Beck-

man) to obtain exosomal pellets.5 The supernatant

from this ultracentrifugation step was decanted and

stored at �80°C until proteomic analysis of the soluble

fraction of the canine urine could be performed.

The exosomal pellets were resuspended in isolat-

ing buffer (10 mM triethanolamine, 250 mM sucrose

at pH of 7.6; Sigma-Aldrich), and each sample was

transferred to a 1.6 mL Eppendorf tube. Dithiothreitol

(DTT; Sigma-Aldrich) was added to a final concentra-

tion of 200 mg/mL, the entire sample mix was then

Vet Clin Pathol 0/0 (2014) 1–13©2014 American Society for Veterinary Clinical Pathology and European Society for Veterinary Clinical Pathology2

The canine urinary proteome Brandt et al

heated at 95°C for 2 minutes for chemical reduction of

Tamm–Horsfall protein (THP).5 Samples were trans-

ferred to 6 ultracentrifuge tubes (14 9 89 mm; Beck-

man). The remaining volume of the ultracentrifuge

tubes was filled with isolation buffer, and the samples

were ultracentrifuged at 200,000g for one h at 4°C(L70 Ultracentrifuge, SW 41 Ti rotor; Beckman).5 The

final pellets were reconstituted in 40 lL of isolation

buffer and combined into 3 1.6 mL Eppendorf tubes,

followed by addition of 5X-SDS Laemmli buffer con-

taining Bromophenol blue and 60 mg/mL DTT at a 1:4

ratio. The sample was incubated at 60°C for 10 min

before boiling for 5 min. All exosomal pellets were

then frozen at�20°C until gel electrophoresis was per-

formed.

One-dimensional SDS-PAGE and in-gel digest ofcanine urinary proteins

Both urine supernatant and exosomal fractions were

thawed on ice. The supernatant was concentrated

using Amicon Ultra-15 centrifugal filter units with a

molecular weight cutoff of 5 kDa (Millipore, Billerica,

MA, USA) as per manufacturer’s instructions. Protein

quantification was performed on both soluble and

Bromophenol blue-dyed free exosomal urine fractions

using the Pierce Coomassie Plus (Bradford) protein

assay kit (Thermo Scientific, Waltham, MA, USA),

according to the manufacturer’s instructions. The con-

centrated supernatant had a protein content of 1.1 lg/lL, and the exosomal pellet contained 1.2 lg/lL of

protein.

Urine supernatant (24.8 lg) and exosomal pro-

teins (19.8 lg) were subjected to electrophoresis in

separate lanes on a 4–20% Mini PROTEAN TGX pre-

cast gel (BioRad, Hercules, CA, USA) with 19 Tris/

Glycine/SDS running buffer, at 175 V for approxi-

mately one hour. The gel was stained with Imperial

Protein Stain (Thermo Scientific, Rockford, IL, USA)

according to manufacturer’s instructions. Prior to

in-gel digestion, each lane was cut into 25 pieces,

which were then diced into ~1 9 1 mm squares and

stored in 0.6 mL Eppendorf tubes.

In-gel digestion was performed using a standard

trypsin digest protocol to produce peptides. Briefly, gel

pieces were destained and washed, after which DTT

reduction and iodacetamide alkylation were per-

formed. Finally, proteins were digested overnight with

mass spectrometry grade trypsin (Promega, Madison,

WI, USA) at 37°C. Tryptic peptides were extracted

from the gel pieces using 50% acetonitrile (ACN) and

0.1% trifluoroacetic acid (Sigma-Aldrich). A vacuum

centrifuge was used to evaporate any remaining sol-

vent from the samples. Samples were reconstituted in

a buffer of 0.1% acetic acid and 3%ACN.

LC-MS/MS analysis

Peptides were purified and concentrated using an

on-line enrichment column (Zorbax C18, 5 lm,

5 9 0.3 mm, Agilent, Santa Clara). Subsequent chro-

matographic separation was performed on a reverse

phase nanospray column (1100 nanoHPLC, Zorbax

C18, 5 lm, 75 lm ID 9 150 mm column; Agilent,

Santa Clara, CA, USA) using a 60-min linear gradient

from 25% to 55% buffer B (90% ACN, 0.1% formic

acid) at a flow rate of 300 nanoliters/min. Peptides

were eluted directly into the mass spectrometer (LTQ

linear ion trap; Thermo Scientific) and spectra were

collected over a m/z range of 200–2000 Da using a

dynamic exclusion limit of 2 MS/MS spectra of a given

peptide mass for 30 s (exclusion duration of 90 s).

Compound lists of the resulting spectra were generated

using Bioworks 3.0 software (Thermo Scientific) with

an intensity threshold of 5000 and one scan/group.

MS/MS spectra were searched against the NCBI

Canine protein sequence database containing 69,862

sequence entries using both the Mascot (version 2.3;

Matrix Science, Boston, MA, USA) and Sequest (Sor-

cerer-SEQUEST; Sage-N Research, Milpitas, CA, USA)

database search engines.

A peptide mass tolerance of 2 Da and a fragment

ion mass tolerance of 1.5 Da were utilized by Mascot

searches, while a peptide mass tolerance of 2 Da and a

fragment ion mass tolerance of 1.0 were used via

Sequest. The following parameters were employed in

all database searches: full tryptic digestion allowing for

one missed cleavage, variable modification of methio-

nine oxidation, and a fixed modification of cysteine

carbamidomethylation.

Peptide identifications from both search engines

were combined using probabilistic protein identifica-

tion algorithms implemented in Scaffold software

(Proteome Software, Portland, OR, USA). Peptide and

protein probability thresholds of 95% and 90%,

respectively, were applied to the results (2.2% FDR as

calculated by Scaffold based on probability statistics).

Proteins containing shared peptides were grouped by

Scaffold to satisfy the laws of parsimony. Manual vali-

dation of MS/MS spectra was performed for all protein

identifications above these thresholds that were based

on one peptide. Selection by manual validation was

based on the following criteria: (1) The peptide must

be identified by both search engines (Sequest andMas-

cot), (2) there must be a minimum of 75% coverage of

theoretical y or b ions (at least 5 in consecutive order),


Brandt et al The canine urinary proteome

(3) there must be an absence of prominent unassigned

peaks greater than 5% of the maximum intensity, and

(4) indicative residue-specific fragmentation, such as

intense ions N-terminal to proline and immediately

C-terminal to aspartate and glutamate were used as

additional parameters of confirmation.

Enrichment analysis of GO categories

Analysis of the canine urinary proteome was modeled

after an earlier report.1 Enrichment analysis of the

total canine urinary proteome dataset was performed

using the Cytoscape plug-in tool, Biological Networks

Gene Ontology (BiNGO).28 Separate datasets were

generated for the full list of exosomal and soluble pro-

teins identified from the canine urine and were indi-

vidually searched against GO annotation terms within

the functional categories. Proteins were appropriately

grouped into GO categories and determined to be over-

or underrepresented in comparison to the complete

International Protein Index (IPI) human proteome

based on results from a hypergeometric test. A resul-

tant P-value of ≤ .001 was considered to be statistically

significant.28 Corrections for multiple term testing

were performed automatically by BiNGO using the

recommended Benjamini and Hochberg correction.28

This analysis was repeated using subset datasets for sol-

uble and exosomal proteins that were common

between dogs and people, as well as soluble and exoso-

mal proteins that were only observed in canine urine.

In all figures representing the data obtained from

BiNGO, GO terms were subjectively paired down to

include as many GO terms encompassing as many

categorized proteins as possible, while still allowing

the data to be presented in a useful manner. Given the

large amount of information, all figures illustrat-

ing the complete BINGO analysis are presented as

supplemental data.

Results

Urine was collected from 15 clinically normal, adult

pet dogs. Four dogs were later excluded due to exces-

sive numbers of acanthocytes onmicroscopic review of

the blood film (2 dogs), mildly elevated ALT activity of

435 IU/L (reference range 10–110 IU/L) and succes-

sive lymphoma (one dog), and marked struvite crys-

talluria on urine sediment examination (one dog).

The remaining 11 dogs had no abnormalities in

CBC, blood film review, serum chemistry profile, or

urinalysis, and had a UPC ≤ 0.1. This group of dogs

included 6 female spayed dogs, one female intact dog,

and 4 male castrated dogs. Dogs ranged from 2 to

9 years of age with a median of 6 years; breeds

included 5 mix-breed dogs, 3 Labrador Retrievers, one

Border Collie, one Golden Retriever, and one Pug.

Characterization of the canine urinary proteome

In total, 563 proteins were confidently identified after

the removal of redundant proteins and anionic tryp-

sinogen from the dataset. Of these, 391 were identified

in the exosome fraction (Table S1) and 214 were iden-

tified in the soluble fraction (Table S2). A total of 349

proteins were detected exclusively in the exosome

fraction (Figure 1), and 172 were detected exclusively

in the soluble fraction, while 42 proteins were identi-

fied in both fractions. A total of 74 proteins in the exo-

some fraction, and 50 proteins in the soluble fraction

were identified based on a single peptide. Manual vali-

dation of the detected spectra for these proteins was

performed as described in the Materials and Methods

section.

Themolecular weight distribution of the identified

proteins in the canine urine was highly similar to that

of people.1,4 The majority of canine urinary proteins,

from both the exosomal and soluble fractions, had

a molecular weight of approximately 10–80 kDa

(Figure 2). Only 64 canine exosomal proteins and 58

soluble proteins had amolecular weight ≥ 80 kDa.

The canine urinary proteome was further charac-

terized using BiNGO, a plug-in for Cytoscape, which

searches proteins against Gene Ontology (GO) terms

and categories.28,29 Gene Ontology was developed by

the GO Consortium in 2000 to unify the annotation of

proteins from various eukaryotic organisms by relying

on a stringently defined vocabulary of terms to be

included in the larger functional categories of Cellular

Component, Molecular Function, and Biological Pro-

cess.29 The ontologic categorization of proteins is fur-

ther enhanced through the GO’s links to various other

gene and protein databases, such as SwissPROT and

Figure 1. Venn diagram illustrating the distribution of total canine

urinary proteins within the exosome and soluble fractions identified by

liquid chromatography tandemmass spectrometry analysis.



GenBank, which helps the GO protein classifications

to stay current.29

Of the identified proteins, 6 could not be catego-

rized using BiNGO: Chain A, Grp94 N-terminal bound

to geldanamycin (exosome), Na+/K+ ATPase alpha

chain (exosome), IgG heavy chain B (exosome & solu-

ble), IgG heavy chain D (soluble), IgG heavy chain A

(soluble) and DLA class I histocompatibility antigen,

A9/A9 alpha chain (soluble) as standard gene name

abbreviations are not established for these proteins,

which are necessary to input datasets. For the remain-

ing proteins, 2 datasets were established, one com-

prised of 388 exosomal proteins, and the other of 213

soluble proteins, while 42 proteins were similar

between the 2 datasets. Within the GO categories of

Cellular Component, Molecular Function, and Biolog-

ical Process, the number of urinary proteins linked to a

minimum of one annotation term was 335, 331, and

324 from the exosomal fraction, and 190, 175, and 168

from the soluble fraction, respectively. The distribution

of over- and underrepresented proteins within the

respective GO categories representing the entire com-

plement of proteins detected in the canine urine is

shown in Figures S1–S12.

Comparison of canine vs human urinary proteomevia Gene Ontology annotation

Comparison of our canine data with recently published

human data1 revealed that 52% (205/391) of the

canine exosomal proteins and 62% (133/214) of the

canine soluble proteins were also found within the

published human urine proteome as reported in 2006

(Figure 3).1 Interestingly, 48% (186/391) of the exos-

omal and 38% (81/214) of soluble proteins identified

in the present study appear to be unique for dogs.

To further compare the canine urine proteome

with published human data, we subdivided our results

into 2 datasets, the first containing only canine pro-

teins found in common with human urine proteins,

and the second containing proteins that were only

Figure 2. Molecular weight distribution of canine urinary proteins identified in the exosome and soluble fractions by liquid chromatography tandem

mass spectrometry.

Figure 3. Total urinary proteins identified by liquid chromatography

tandem mass spectrometry that are unique to canine urine (gray bars)

vs those common to human urine (black bars) in both the exosome and

soluble fractions.



detected in our canine study. BiNGO analysis was

repeated with these new datasets, searching for over-

and underrepresented terms within the GO functional

categories of Cellular Component, Molecular Func-

tion, and Biological Process (Figures S13–S30).The distributions of over- and underrepresented

proteins within the GO Cellular Component, Molecu-

lar Function, and Biological Process categories were

highly similar between datasets containing the whole

list of canine proteins and the datasets containing only

proteins common between the canine and human.

Cellular Component (canine/human commonproteins)

The majority of overrepresented proteins within the

Cellular Component group were linked to intracellular

locations (Figure S13–S15). Gene Ontology analysis of

the human urine proteome revealed many soluble

proteins annotated to extracellular terms, with fewer

numbers annotated to intracellular terms.1 Canine

urine was also observed to contain both intracellular

and extracellular components; however, intracellular

proteins were present in greater number in the exoso-

mal fraction, while extracellular proteins were more

common in the soluble fraction.

Molecular Function (canine/human commonproteins)

Many proteins common to both canine and human

urine were annotated to the same Molecular Function

(such as peptidase activity and various peptidase activ-

ity daughter terms) (Figures S16–S19). However, the

relative number of proteins annotated to various GO

terms differed between canine and human urine.1 For

example, canine urine has fewer proteins annotated to

signal transducer activity and enzyme inhibition than

human urine.1 In contrast and similar to human urine,

the majority of underrepresented Molecular Function

terms annotated to canine proteins are associated with

nuclear activities, such as DNA binding.1

Biological Processes (canine/human commonproteins)

The GO Biological Process category resulted in the

largest lists of annotated terms for statistically signifi-

cantly overrepresented proteins in canine urine (Fig-

ure S20 and S21). Again, while the distribution of

terms was similar, the relative number of proteins

annotated to terms such as cell communication,

defense response, and immune response was higher

in human urine.1 Furthermore, only a small number

of canine proteins were annotated to immunologic

terms, such as natural killer cell-mediated immunity,

leukocyte degranulation, and leukocyte-mediated

cytotoxicity.

The primary underrepresented Biological Process

GO terms in the canine dataset represented nuclear

processes, such as regulation of gene expression and

transcription (Figure S22 and S23). Conversely, while

the human dataset contains some proteins annotated

to nucleobase, nucleoside, nucleotide and nucleic acid

metabolism, themajority of underrepresented proteins

in canine urine were annotated to more general terms,

such as cell metabolism and regulation of cellular pro-

cesses.1 This may indicate a prevalence of proteins

related to nuclear biological processes in human as

comparedwith canine urine.

Cellular Component (canine-unique proteins)

As expected, Cellular Component GO terms overrepre-

sented in the exosome fraction primarily annotated to

intracellular locations (Figure S24)1 and soluble pro-

teins primarily annotated to terms related to extracel-

lular components as well as cytoskeletal elements and

components of filtrated plasma (Figure S25). No signif-

icant underrepresented GOCellular Component anno-

tations (Figure S26) were observed for proteins in the

soluble fraction that were only found in canine urine.

This is in contrast to the human data, which contain

many underrepresented proteins annotated to intra-

cellular components, such as organelle, nucleus, and

ribosome.1

Molecular Function (canine-unique proteins)

As expected, the overrepresented Molecular Function

GO terms annotated to exosomal proteins unique to

the dog are few, and include: structural molecule

activity, SNARE (soluble N-ethylmaleimide-sensitive

factor attachment receptor) binding, xylulokinase

activity, and purine-nucleoside phosphorylase activity

(Figure S27 and S28).

The underrepresented exosomal proteins unique

to the dog were annotated to similar Molecular Func-

tion terms as in the total dataset (Figure S29). The

underrepresentation of proteins annotated to signal

and molecular transducer activities suggests possible

decreased transducer activity in canine vs human

urine. No statistically significant underrepresented

terms were observed in the soluble proteins unique

to the dog within the GO Molecular Function

category.



Biological Process (canine-unique proteins)

No statistically significantly underrepresented exo-

some proteins unique to the dog were annotated to

any GO Biological Process terms. For the soluble pro-

teins unique to the dog, proteins were linked to only 2

GO Biological Process terms, ectoderm development

and epidermis development, each containing the same

annotated proteins (Figure S30). No proteins from the

human data were annotated to either of these GO

Biological Process terms.

Proteins with known disease association

Among our list of identified proteins in canine urine,

60 of these proteins (which are also reported in

human urine) are known to be related to specific ex-

trarenal diseases (Table 1).3 Many of these proteins

were linked to rare human diseases that do not have

a recognized canine counterpart, such as hyperm-

ethioninemia, Bardet-Biedl syndrome, and Papillon-

Lefevre syndrome. However, some proteins linked to

human extrarenal disease do have a canine version

of the disease. For example, many proteins on this

list are related to hypertension, including angio-

tensin 1-converting enzyme (ACE) isoforms 1 and

2, dimethylarginine dimethylaminaophydrolase 1,

glutamyl aminopeptidase, hydroxyprostaglandin

dehydrogenase 15-(NAD), and membrane metallo-

endopeptidase.3 Additionally, urinary proteins asso-

ciated with disorders such as megaloblastic anemia,

cystathioninuria, and Charcot–Marie–Tooth neurop-

athy were identified. Megaloblastic anemia can

result from acquired vitamin B6 deficiency or be

secondary to a genetic mutation; it is reported in

dogs and appears to have a similar pathogenesis as

in people.30 Cystathioninuria has been reported

in Dachshunds and a Scottish Terrier.31 Charcot–

Marie–Tooth-like neuropathies have been diagnosed

clinically in dogs; however, further research on the

pathogenesis is needed to determine if these neuropa-

thies are truly similar to those seen in people.19

Additionally, 14 known and/or potential biomar-

kers for renal disease were detected in our canine sam-

ples (Table 1), including albumin, immunoglobulin-c(IgG), and retinal-binding protein 4 (RBP or RBP4),

clusterin, aquaporin-1, hemopexin, fetuin-A, and

ubiquitin A-52 (UBA52), among others. Furthermore,

proteins known to have altered regulation during renal

disease, localized to the kidney, or involved in genetic

renal diseases were detected and include chloride

intracellular channel 1, solute carrier family 12

(Na+K+Cl� transporters) member 1, and myo-inositol

oxygenase.32–35 Notably, multiple known canine bio-

markers of renal disease, such as C-reactive protein

(CRP), GGT, and N-acetyl-b-glucosaminidase (NAG),

were not identified in this study due to our selection of

clinically normal dogs.36–38

Discussion

Adaptation of protocols developed for human urine

resulted in the identification 563 proteins in pooled

urine from clinically healthy dogs. Dogs are thought to

have more THP in their urine than people. The higher

levels of THP and the incompleteness of the canine

protein database both probably contributed to the

reduced number of proteins identified in our study

compared with recent proteomic studies of human

urine.39 Chemical reduction of THP was performed

using previously described methods; however, these

methods were developed for human urine and are

probably not optimized for the high THP levels found

in canine urine. As carnivores, dogs excrete vitamin A

in their urine, and THP has been identified as the reti-

nol carrier protein in canine urine.39 Development of

additional and/or alternative methods to more effi-

ciently remove THP from canine urine could improve

detection coverage of proteins. Canine urine also con-

tains relatively high levels of albumin. Commercially

available kits have been developed for albumin

removal and could be optimized for their use with

canine urine; however, these methods often result in

the nonspecific depletion of additional proteins of

potential biologic interest.1

The molecular weight distribution of canine uri-

nary proteins was similar to that of human urinary

proteins.1 It is widely accepted that proteins filtered via

the healthy glomerulus are generally < 68 kDa.40 The

majority of canine and human urinary proteins are in

the range of 10–79 kDa, while a smaller number of

proteins with molecular weights ≥ 80 kDa are present

in normal urine from both people and dogs.1 These

higher molecular weight proteins may therefore repre-

sent postglomerular filtration proteins released or

secreted by renal epithelial cells, and an in-depth study

of these proteins could provide valuable information

about real-time renal health.

In regard to cellular components, the highest

number of soluble proteins was allocated to the extra-

cellular region and exosomal proteins to the cytoplasm

and organelles. Plasma membrane also contained

many soluble and exosomal proteins similar to human

urine where all urinary proteins were analyzed

together.1 Our sub datasets of proteins (those in



Table 1. Canine urinary proteins with known associations with human diseases.

Gene Name Protein Name Related Human Disease

Soluble(S)/

Exosome (EX)

ACE Angiotensin 1-converting enzyme isoform 1 Hypertension2 S & EX

ACE 2 Angiotensin 1-converting enzyme 2 Hypertension2 S & EX

ACY1 Aminoacylase 1 Aminoacylase 1 deficiency EX

AGT Angiotensinogen Hypertension, chronic kidney disease (CKD),

IgA nephropathy11S

AHCY S-adenosylhomocysteine Hypermethioninemia EX

AHSG Fetuin-A Acute renal injury10 S

ALB Albumin precursor Dysalbuminemic hyperthyroxinemia,

hyperthyroxinemia–dysalbuminemic analbuminemia

bisalbuminemia glomerular damage, proximal tubular damage2

S & EX

ALDOA [Fructose-bisphosphate] aldolase A

(aka “lung cancer antigen”)

Aldolase deficiency of red cells; myopathy & hemolytic anemia2 EX

AMN Amnionless protein precursor Megaloblastic anemia 12 EX

APOA 1 Apolipoprotein A-1 precursor Primary hypoalphalipoproteinemia2 EX

AQP 1 Aquaporin 1 AQP1 deficiency, Colton-Null, renal neoplasms, ischemic

reperfusion injury to the kidneys5,23EX

ARL6 ADP-ribosylation factor-(like) 6 Bardet–Biedl syndrome 32 EX

ASL Argininosucinate lyase isoform 3 Argininosuccinic aciduria 2 EX

ASS1 Argininosucinate synthetase 1 Citrullinemia2 EX

ATP6V0A4 ATPase, H+ transporting, lysosomal

V0 subunit a4

Renal tubular acidosis, distal, autosomal recessive2 EX

ATP6V1B1 Atpase, H+ transporting, lysosomal

56/58 kDa, V1 subunit B1

Renal tubular acidosis, distal, w/progressive deafness2 EX

B4GALT1 UDP-Gal:bGlcNac b 1,4-galactosyltransferase

1, membrane-bound form

Congenital disorder of glycosylation type IId2 S

CA2 Carbonic anhydrase II Autosomal recessive syndrome of osteopetrosis with

renal tubular acidosis2EX

CA4 Carbonic anhydrase IV precursor Proximal renal tubular acidosis2 EX

CHMP2B Chromatin modifying protein 2B Frontotemporal dementia, chromosome 3-linked2 EX

CLU Clusterin General renal injury. Glomerular, tubular, and renal

papillary damage. Proximal tubular24S

COL6A3 a 3 type VI collagen (isoform 5 precursor) Ullrich congenital muscular dystrophy2 S & EX

CRYM Crystalline, l isoform (1) Autosomal dominant nonsyndromic deafness2 EX

CTH Cystathionase isoform (2), [we got 1] Cystathioninuria2 EX

CTSA Cathepsin A (precursor) Galactosialidosis2 S

CTSD Cathepsin D Papillon-Lef�evre syndrome2 S & EX

DDAH1 Dimethylarginine dimethylaminohydrolase 1 Hypertension2 EX

DNM2 Dynamin 2 (isoform 4) Charcot–Marie–Tooth neuropathy2 EX

DPYS Dihydropyrimidinase [isoform 1] Dihydropyrimidinuria2 EX

DYSF Dysferlin Miyoshi myopathy2 EX

EFEMP1 EGF-containing fibulin-like extracellular matrix Doyne Honeycomb retinal dystrophy2 S

EGF Epidermal growth factor Acute renal injury39 S & EX

ELA2 Elastase (2), neutrophil Cyclic hematopoiesis2 S

ENPEP Glutamyl aminopeptidase (aminopeptidase A) Hypertension2 S & EX

FBP1 Fructose-1, 6-bisphosphatase 1 Fructose-1, 6-bisphosphatase deficiency2 EX

FGA Fibrinogen, a (polypeptide isoform a-E) Renal amyloidosis, Dysfibrinogenemia2 S

FTL Ferritin, light (polypeptide), [chain] Iron overload, autosomal dominant EX

GAA (Acid) a-glucosidase preprotein Infantile-onset glycogen storage disease Type II2 S

GGT/GGT1 c-glutamyltransferase Proximal tubular damage38,39 EX

GPI Glucose [6-]phosphate isomerase Chronic hemolytic anemia due to GPI deficiency2 EX

GSS Glutathione synthetase Glutathione synthetase deficiency2 EX

HPD 4-Hydroxyphenylpyruvate dioxygenase Tyrosinemia type III2 EX

HPGD Hydroxyprostaglandin dehydrogenase 15-(NAD) Hypertension2 EX

HPX Hemopexin Glomerular disease/damage40 S

(continued)



common with human and those unique to the canine

urine) followed a similar distribution.

Exosomal proteins yielded significantly lower

numbers of proteins that are integral and intrinsic to

the membrane. These types of proteins are buried

inside, penetrate one side of the plasma membrane’s

lipid bilayer, or span both sides of the lipid bilayer,

respectively. It is somewhat surprising that proteins

embedded in the plasmamembranewould not be pres-

ent in higher numbers in the exosomal fraction. How-

ever, perhaps integral and intrinsic membrane

proteins are preferentially retained by the plasma

membrane during exocytosis of exosomal vesicles.

Nuclear and intracellular proteins were underrepre-

sented in the soluble protein fraction. This is an

expected result as it is unlikely that nuclear and

intracellular proteins would escape from normal,

healthy cells into the renal tubular lumen.

Proteins involved in various Molecular Functions

were found in both the exosomal and soluble urine

fractions. The canine urinary exosomal fraction con-

tains many proteins related to various types of protein

binding and catalase activity. In metabolically active

tissue such as the kidneys, it is not surprising that these

intracellular molecular activities would be amplified in

the exosomal fraction of urine. A similar distribution of

annotated GO Molecular Function terms was found in

the datasets of common proteins and the dataset com-

prised of proteins observed only in canine urine. Pro-

tein binding and hydrolase activity (a GO daughter

term of catalytic activity) are both significantly over-

represented in human urine.1 Proteins involved in

Table 1. (continued).

Gene Name Protein Name Related Human Disease

Soluble(S)/

Exosome (EX)

IgG Immunoglobulin-c Glomerular damage S & EX

KLK1 Kallikrein 1 preprotein Decreased urinary activity of kallikrein2 S

LGALS3 Galectin 3 [binding precursor] Lymphocyte function-associated antigen 12 S & EX

LRRK2 [Ig superfamily containing] Leucine-rich

repeat (kinase 2)

Parkinson’s disease2 S

LYZ Lysozyme precursor Familial visceral amyloidosis, general renal disease6,38 S

MME Membrane metallo-endopeptidase Hypertension2 EX

MYH14 Myosin[-6] (heavy chain 14) isoform 1 Autosomal dominant nonsyndromic sensorineural deafness2 EX

NDRG1 N-myc downstream regulated gene 1 Charcot–Marie–Tooth disease type 4D2 EX

NEB Nebulin [related anchoring protein

isoform S isoform 2]

Nemalin myopathy2 EX

PARK7 DJ-1 protein [isoform 1] Parkinson’s disease2 EX

PHGDH [D-3] Phosphoglycerate dehydrogenase Phosphoglycerate dehydrogenase deficiency2 EX

PKLR Pyruvate kinase (liver, and RBC isoform)

[we got isozymes M1/M2]

Pyruvate kinase deficiency2 EX

PODXL Podocalyxin Glomerular disease, diabetic nephropathy, lupus nephritis,

and IgA nephropathy2EX

PRKCH Protein kinase C (ή) [& caseine kinase

substrate in neurons 3 isoforms]

Cerebral infarction2 EX

PROM1 Prominin 1 [precursor] Autosomal recessive retinal degeneration2 EX

RAB3GAP1 (RAB3)

GTPase-activating protein

Warburg micro syndrome2 EX

RBP4 Retinol-binding protein 4, plasma precursor Retinal-binding protein deficiency, proximal tubular

dysfunction12S

RDX Radixin [moesin-binding phosphoprotein 50] Autosomal recessive deafness2 EX

SCL12A3 Solute carrier family 12, member 3 Gitelman syndrome2 EX

SCL3A1 Solute carrier family 3, member 1 Cystinuria2 EX

TF Transferrin Alzheimer disease2 S

THP Tamm–Horsfall Protein Chronic renal disease, medullary cystic kidney disease,

and familial juvenile hyperuricemic nephropathy16S & EX

TPP1 Tripeptidyl-peptidase 1 Ceroid lipofuscinosis neuronal 22 EX

TSG101 Tumor susceptibility gene Breast Cancer2 EX

UBA52 Ubiquitin A-52 ribosomal protein fusion product 1 Diabetic nephropathy7 EX

UMOD Uromodulin precursor Medullary cystic kidney disease-2, Familial juvenile

hyperuricemic nephropathy2S & EX



protein binding were also present in large numbers in

the canine urine soluble fraction. Additionally, multi-

ple proteins associated with the binding of other sub-

stances, such as carbohydrates, glycoproteins, lipids,

and vitamins, and involved in enzyme regulation and

inhibition were identified within the canine soluble

fraction.

Proteins that were underrepresented in all canine

datasets were largely related to nuclear activities, such

as nucleic acid binding. This is similar to findings in

human urine, and again, not surprising as nuclear

material and activities should be restricted to the

nucleus in intact, healthy cells.1 In contrast to human

urine, where signal transducer activity is significantly

overrepresented with 275 annotated proteins1, pro-

teins related to signal transducer activity were signifi-

cantly underrepresented in the dataset of unique

soluble canine proteins. Cellular signal transduction,

according to the GO definition, involves conveying a

signal across a cell to trigger a response, such as a

change in cellular function or state. The significance of

this startling difference between the protein composi-

tions of canine vs human urine is not yet understood.

A major GO daughter term of signal transduction is

receptor binding, to which 23 proteins in the canine

urinary soluble fraction are annotated. It may simply

be that fewer, more specialized metabolic functions

occur in canine renal epithelium and glomerular fil-

trate. Further studies are required to fully understand

the basis for this observed difference.

Amongst GO Biological Process terms, canine

urine proteins were heavily annotated to terms related

to metabolic process, response to stimulus, transport,

and anatomical structural development. These terms,

and related terms, were likewise heavily linked to

human urinary proteins. However, human urine also

contained overrepresented proteins annotated to

terms related to the immune response, which was not

observed in the canine data. This may indicate a higher

degree of immunologic activity in normal human

urine vs normal canine urine.

Conclusions

Our proteomic analysis of normal canine urine has

yielded an extensive list of proteins, some of which are

already being used as urinary biomarkers for renal dis-

ease in people. Moreover, proteins such as fetuin-A

and UBA52 have not been previously identified in

canine urine and may prove to be useful indicators of

acute renal injury in dogs. Additionally, proteins asso-

ciated with extrarenal disease, as mentioned above,

were also identified. Thus, the results of this study pro-

vide a framework for future studies of canine renal and

extrarenal diseases, and also suggest that the dog may

represent a viable large animal model to study genetic

diseases, systemic hypertension, neoplasia, and other

human diseases.

Disclosure: The authors have indicated that they have no

affiliations or financial involvement with any organization

or entity with a financial interest in, or in financial competi-

tion with, the subject matter or materials discussed in this

article.

References

1. Adachi J, Kumar C, Zhang Y, Olsen JV,MannM. The

human urinary proteome containsmore than 1500 pro-

teins, including a large proportion ofmembrane pro-

teins. Genome Biol. 2006;7:R80.

2. Coon JJ, Zurbig P, DaknaM, et al. CE-MS analysis of

the human urinary proteome for biomarker discovery

and disease diagnostics. Proteomics Clin Appl.

2008;2:964.

3. Gonzales PA, Pisitkun T, Hoffert JD, et al. Large-scale

proteomics and phosphoproteomics of urinary exo-

somes. J Am Soc Nephrol. 2009;20:363–379.

4. Kentsis A,Monigatti F, Dorff K, Campagne F, Bachur R,

Steen H. Urine proteomics for profiling of human dis-

ease using high accuracymass spectrometry. Proteomics

Clin Appl. 2009;3:1052–1061.

5. Pisitkun T, Shen RF, KnepperMA. Identification and

proteomic profiling of exosomes in human urine. Proc

Natl Acad Sci USA. 2004;101:13368–13373.

6. SunW, Li F,Wu S, et al. Human urine proteome analy-

sis by three separation approaches. Proteomics.

2005;5:4994–5001.

7. Thongboonkerd V,McLeish KR, Arthur JM, Klein JB.

Proteomic analysis of normal human urinary proteins

isolated by acetone precipitation or ultracentrifugation.

Kidney Int. 2002;62:1461–1469.

8. Cutillas PR, Chalkley RJ, Hansen KC, et al. The uri-

nary proteome in Fanconi syndrome implies specific-

ity in the reabsorption of proteins by renal proximal

tubule cells. Am J Physiol Renal Physiol. 2004;287:

F353–F364.

9. Hoorn EJ, Pisitkun T, Zietse R, et al. Prospects for uri-

nary proteomics: exosomes as a source of urinary bio-

markers.Nephrology. 2005;10:283–290.

10. Kentsis A, Lin YY, Kurek K, et al. Discovery and valida-

tion of urinemarkers of acute pediatric appendicitis

using high-accuracymass spectrometry. Ann Emerg

Med. 2010;55:62–70.



11. OettingWS, Rogers TB, Krick TP,Matas AJ, Ibrahim

HN. Urinary beta(2)-microglobulin is associated with

acute renal allograft rejection. Am J Kidney Dis.

2006;47:898–904.

12. Theodorescu D,Wittke S, RossMM, et al. Discovery

and validation of new protein biomarkers for urothelial

cancer: a prospective analysis. Lancet Oncol. 2006;7:230–

240.

13. Miller I, Teinfalt M, LeschnikM,Wait R, GemeinerM.

Nonreducing two-dimensional gel electrophoresis for

the detection of Bence Jones proteins in serum and

urine. Proteomics. 2004;4:257–260.

14. Forterre S, Raila J, Schweigert FJ. Protein profiling of

urine from dogs with renal disease using ProteinChip

analysis. J Vet Diagn Invest. 2004;16:271–277.

15. NabityMB, Lees GE, Dangott LJ, Cianciolo R, Suchod-

olski JS, Steiner JM. Proteomic analysis of urine from

male dogs during early stages of tubulointerstitial injury

in a caninemodel of progressive glomerular disease. Vet

Clin Pathol. 2011;40:222–236.

16. Thery C, Zitvogel L, Amigorena S. Exosomes: composi-

tion, biogenesis and function.Nat Rev Immunol.

2002;2:569–579.

17. Dieterle F, Perentes E, Cordier A, et al. Urinary cluster-

in, cystatin C, beta2-microglobulin and total protein as

markers to detect drug-induced kidney injury.Nat Bio-

technol. 2010;28:463–469.

18. Dihazi H, Muller GA, Lindner S, et al. Characteriza-

tion of diabetic nephropathy by urinary proteomic

analysis: identification of a processed ubiquitin form

as a differentially excreted protein in diabetic

nephropathy patients. Clin Chem. 2007;53:1636–

1645.

19. Granger N. Canine inheritedmotor and sensory neur-

opathies: an updated classification in 22 breeds and

comparison to Charcot-Marie-Tooth disease. Vet J.

2011;188:274–285.

20. Jang HR, Kim SM, Lee YJ, et al. The origin and the clini-

cal significance of urinary angiotensinogen in proteinu-

ric IgA nephropathy patients. AnnMed. 2012;44:448–

457.

21. KwonO, AhnK, Zhang B, et al. Simultaneousmonitor-

ing of multiple urinary cytokines may predict renal and

patient outcome in ischemic AKI. Ren Fail.

2010;32:699–708.

22. Morrissey JJ, London AN, Luo J, Kharasch ED. Urinary

biomarkers for the early diagnosis of kidney cancer.

Mayo Clin Proc. 2010;85:413–421.

23. OmbrellinoM,Wang H, Yang H, ZhangM, et al. Fetuin,

a negative acute phase protein, attenuates TNF synthe-

sis and the innate inflammatory response to carra-

geenan. Shock. 2001;15:181–185.

24. RosenbergME, Silkensen J. Clusterin and the kidney.

Exp Nephrol. 1995;3:9–14.

25. Varghese SA, Powell TB, BudisavljevicMN, et al. Urine

biomarkers predict the cause of glomerular disease. J

Am Soc Nephrol. 2007;18:913–922.

26. Zhou H, Pisitkun T, Aponte A, et al. Exosomal Fetuin-A

identified by proteomics: a novel urinary biomarker for

detecting acute kidney injury. Kidney Int.

2006;70:1847–1857.

27. Zhou H, Yuen PS, Pisitkun T, et al. Collection, storage,

preservation, and normalization of human urinary exo-

somes for biomarker discovery.Kidney Int.

2006;69:1471–1476.

28. Maere S, Heymans K, KuiperM. BiNGO: a Cytoscape

plugin to assess overrepresentation of gene ontology

categories in biological networks. Bioinformatics.

2005;21:3448–3449.

29. AshburnerM, Ball CA, Blake JA, et al. Gene ontology:

tool for the unification of biology. The Gene Ontology

Consortium.Nat Genet. 2000;25:25–29.

30. He Q, Fyfe JC, Schaffer AA, et al. Canine Imerslund-

Grasbeck syndromemaps to a region orthologous to

HSA14q.MammGenome. 2003;14:58–764.

31. Frimpter GW. Cystathioninuria in a patient with cystin-

uria. Am JMed. 1969;46:832–836.

32. Kemter E, Rathkolb B, Bankir L, et al. Mutation of the

Na+-K+-2Cl(-) cotransporter NKCC2 inmice is associ-

ated with severe polyuria and a urea-selective concen-

trating defect without hyperreninemia. Am J Physiol

Renal Physiol. 2010;298:F1405–F1415.

33. Lokman FE, Seman NA, Ismail AA, et al. Gene expres-

sion profiling in ethnicMalays with type 2 diabetes

mellitus, with andwithout diabetic nephropathy. J

Nephrol. 2011;24:778–789.

34. Tulk BM, Edwards JC. NCC27, a homolog of intracellu-

lar Cl- channel p64, is expressed in brush border of

renal proximal tubule. Am J Physiol Renal Physiol.

1998;274:F1140–F1149.

35. Xie P, Sun L, Oates PJ, Srivastava SK, Kanwar YS.

Pathobiology of renal-specific oxidoreductase/

myo-inositol oxygenase in diabetic nephropathy: its

implications in tubulointerstitial fibrosis.Am J Physiol

Renal Physiol. 2010;298:F1393–F1404.

36. Clemo FA. Urinary enzyme evaluation of nephrotoxi-

city in the dog. Toxicol Pathol. 1998;26:29–32.

37. Maddens B, Daminet S, Smets P, Meyer E. Escherichia

coli pyometra induces transient glomerular and tubular

dysfunction in dogs. J Vet Intern Med. 2010;24:1263–

1270.

38. Smets PM,Meyer E,Maddens BE, Duchateau L, Dami-

net S. Urinarymarkers in healthy young and aged dogs



and dogs with chronic kidney disease. J Vet Intern Med.

2010;24:65–72.

39. Schweigert FJ, Raila J, Haebel S. Vitamin A excreted in

the urine of canines is associated with a Tamm-Horsfall

like protein. Vet Res. 2002;33:299–311.

40. Stockham SL, ScottMA. Fundamentals of Veterinary

Clinical Pathology. Ames, Iowa: Blackwell; 2008.

Supporting Information

Additional Supporting Information may be found in

the online version of this article:

Figure S1. Gene Ontology (GO) Cellular Component

terms significantly overrepresented (P < .001) in the

canine urine exosome fraction in comparison to the

total IPI human proteome. The annotation ratio is the

number of canine exosomal proteins associated with a

specific GO term divided by the total number of canine

exosomal proteins annotated to the specific functional

annotation group by GO. The annotation ratio is calcu-

lated in the samemanner for IPI human proteins and is

defined as the number of IPI proteins associated with a

specific GO term divided by the total number of IPI

human proteins annotated by GO within a functional

group.

Figure S2.GOCellular Component terms significantly

overrepresented (P < .001) in the canine urinary solu-

ble fraction in comparison to the total IPI human pro-

teome.

Figure S3. GO Cellular Component terms significantly

underrepresented (P < .001) in the canine exosomal

fraction in comparison to the total IPI human proteome.

Figure S4.GOCellular Component terms significantly

underrepresented (P < .001) in the canine soluble

fraction in comparison to the total IPI human prote-

ome.

Figure S5. GOMolecular Function terms significantly

overrepresented (P < .001) in canine urinary exosome

fraction in comparison to the total IPI human.


overrepresented (P < .001) in canine urinary soluble


ome.




ome.




ome.

Figure S9. GO Biological Process terms significantly

overrepresented (P < .001) in the canine urinary exo-

some fraction in comparison to the total IPI human

proteome.


overrepresented (P < .001) in the canine soluble frac-

tion in comparison to the total IPI human proteome.




ome.




ome.

Figure S13. Canine/human common exosome pro-

teins. Shown is a selection of major annotated GO Cel-

lular Component terms that are significantly

overrepresented (P < .001) as compared to the total IPI

human proteome. The annotation ratio is calculated as

previously described in Figure S1.

Figure S14. Canine/Human common soluble proteins

overrepresented. A selection of major annotated GO

Cellular Component terms significantly overrepre-

sented (P < .001) as compared to the total IPI human

proteome here.

Figure S15. Canine/Human common soluble proteins

underrepresented. All annotated GO Cellular Compo-

nent terms that are significantly underrepresented

(P < .001) as compared to the total IPI human

proteome.

Figure S16. Canine/Human common exosome pro-

teins. A selection of major annotated GO Cellular

Component terms significantly overrepresented

(P < .001) as compared to the total IPI human prote-

ome.

Figure S17. Canine/Human common soluble proteins.

A selection of major annotated GO Cellular Com-

ponent terms significantly overrepresented (P < .001)

as compared to the total IPI human proteome.


teins. All annotated GO Molecular Function terms sig-



nificantly underrepresented (P < .001) as compared to

the total IPI human proteome.

Figure S19. Canine/Human common soluble pro-

teins. All annotated GO Molecular Function terms sig-

nificantly underrepresented (P < .001) as compared to



teins. A selection of major annotated GO Biological

Process terms significantly overrepresented (P < .001)



teins. A selection of major annotated GO Biological

Process terms significantly overrepresented (P < .001)



teins. All annotated GO Biological Process terms signif-

icantly underrepresented (P < .001) as compared to










Figure S25. Canine unique soluble proteins overrep-

resented. A selection of major annotated GO Cellular

Component terms that are significantly overrepre-

sented (P < .001) as compared to the total IPI human

proteome.

Figure S26. Canine unique exosome proteins under-

represented. All annotated GO Cellular Component

terms that are significantly underrepresented

(P < .001) as compared to the total IPI human prote-

ome.

Figure S27. Canine unique exosome proteins. A selec-

tion of major annotated GO Molecular Function terms

\significantly overrepresented (P < .001) as compared

to the total IPI human proteome.

Figure S28. Canine unique soluble proteins. A selec-

tion of major annotated GO Molecular Function terms

significantly overrepresented (P < .001) as compared

to the total IPI human proteome.

Figure S29. Canine unique exosome proteins. All

annotated GO Molecular Function terms significantly

underrepresented (P < .001) as compared to the total

IPI human proteome.

Figure S30. Canine unique soluble proteins. All anno-

tated GO Biological Process terms that are significantly

overrepresented (P < .001) as compared to the total IPI

human proteome.

Table S1. Proteins identified in the exosome fraction

of canine urine.

Table S2. Proteins identified in the soluble fraction of

canine urine.



characterization of the canine urinary proteome

Documents