Large-Scale Proteomics and Phosphoproteomics of Urinary Exosomes

RESULTS

Large-Scale Proteomic Profiling of Human Urinary Exosomes

In this study, we carried out proteomic profiling of a low-density membrane fraction from human urine consisting chiefly of exosomes, using a highly sensitive LC-MS/MS system, based on an ion trap mass spectrometer (LTQ; Thermo-Finnigan; Thermo Electron, San Jose, CA). We unambiguously identified 1132 proteins including 205 proteins seen in our previous study and 927 proteins not seen in our previous study of human urinary exosomes.¹ The full list (ambiguous and unambiguous identifications) contains 1412 proteins and can be viewed in Supplemental Table 1, and the list of proteins that were unambiguously identified in both studies can be viewed at http://dir.nhlbi.nih.gov/papers/lkem/exosome/. The expanded list of exosomal proteins includes 177 proteins that are disease related, on the basis of their presence in the OMIM database (Table 1).

Table 1.

Disease-related proteins in human urinary exosomes^a

Gene	Protein Name	Pep	ID	Related to Disease [OMIM]
ABCB1	ATP-binding cassette subfamily B, member 1	23	61	Colchicine resistance [MIM: 120080] Crohn disease [MIM: 266600]
ABCC9	ATP-binding cassette, subfamily C, member 9 isoform SUR2A-δ-14	1	2	Cardiomyopathy [MIM: 608569]
ABCB11	ATP-binding cassette, subfamily B (MDR/TAP), member 11	1	3	Cholestasis, progressive familial intrahepatic 2 [MIM: 601847]
				Cholestasis, benign recurrent intrahepatic 2 [MIM: 605479]
ACAT1	Acetyl-CoA acetyltransferase 1 precursor	1	2	α-Methylacetoacetic aciduria [MIM: 203750]
ACE	Angiotensin I–converting enzyme isoform 1 precursor	23	96	Hypertension [MIM: 106180]
ACE	Angiotensin I–converting enzyme isoform 2 precurs	12	61	Renal tubular dysgenesis [267430]
ACE2	Angiotensin I–converting enzyme 2 precursor	8	17	Hypertension [MIM: 300335]
ACOT7	Acyl-CoA thioesterase 7 isoform hBACHd	1	1	Mesial temporal lobe epilepsy [MIM: 608096]
ACSL4	Acyl-CoA synthetase long-chain family member 4 isoform 2	1	2	Mental retardation, X-linked 63, MRX 63 [MIM: 300387]
ACY1	Aminoacylase 1	15	43	Aminoacylase 1 deficiency [MIM: 609924]
AHCY	S-adenosylhomocysteine hydrolase	10	28	Hypermethioninemia [MIM: 180960]
AK1	Adenylate kinase 1	4	4	Hemolytic anemia due to AK1 deficiency [MIM: 103000]
ALAD	δ-Aminolevulinic acid dehydratase isoform a	1	1	Acute hepatic porphyria [MIM: 125270]
ALB	Albumin precursor	36	139	Dysalbuminemic hyperthyroxinemia Hyperthyroxinemia, dysalbuminemic analbuminemia bisalbuminemia [MIM: 103600]
ALDOA	Aldolase A	7	14	Aldolase deficiency of red cells Myopathy and hemolytic anemia [MIM: 103850]
ALPL	Tissue nonspecific alkaline phosphatase precursor	3	4	Hypophostasia [MIM: 241500]
AMN	Amnionless protein precursor	1	1	Megaloblastic anemia 1 [MIM: 261100]
ANPEP	Membrane alanine aminopeptidase precursor	69	412	Hypertension [MIM: 151530]
APOA1	Apolipoprotein A-I preproprotein	6	17	Primary hypoalphalipoproteinemia [MIM: 604091]
APOA2	Apolipoprotein A-II preproprotein	1	1	Apolipoprotein A-II deficiency, familial
				Hypercholesterolemia, familial [MIM: 143890]
APRT	Adenine phosphoribosyltransferase isoform a	2	2	2,8-Dihydroxyadenine urolithiasis [MIM: 102600]
APRT	Adenine phosphoribosyltransferase isoform b	3	10	2,8-Dihydroxyadenine urolithiasis [MIM: 102600]
AQP1	Aquaporin 1	3	35	Aquaporin 1 deficiency, Colton-Null [MIM: 110450]
AQP2	Aquaporin 2	7	36	Autosomal recessive nephrogenic diabetes insipidus, type 1 [MIM: 222000]Autosomal dominant nephrogenic diabetes insipidus, type 1 [MIM: 125800]
ARL6	ADP-ribosylation factor–like 6	4	7	Bardet-Biedl syndrome 3 [MIM: 209900]
ARSE	Arylsulfatase E precursor	1	2	Chondrodysplasia punctata 1, X-linked recessive [MIM: 302950]
ASAH1	N-acylsphingosine amidohydrolase (acid ceramidase) 1 preproprotein isoform a	7	16	Farber disease [MIM: 228000]
ASAH1	N-acylsphingosine amidohydrolase (acid ceramidase) 1 isoform b	9	34	Farber disease [MIM: 228000]
ASL	Argininosuccinate lyase isoform 3	1	1	Argoninosuccinic aciduria [MIM: 207900]
ASS1	Argininosuccinate synthetase 1	20	59	Citrullinemia [MIM: 215700]
ATIC	5-Aminoimidazole-4-carboxamide ribonucleotide formyltransferase/IMP cyclohydrolase	1	1	Aica-ribosiduria due to ATIC deficiency [MIM: 608688]
ATP6V0A4	ATPase, H transporting, lysosomal V0 subunit a4	1	2	Renal tubular acidosis, distal, autosomal recessive [MIM: 602722]
ATP6V1B1	ATPase, H transporting, lysosomal 56/58-kD, V1 subunit B1	8	17	Renal tubular acidosis, distal, with progressive deafness [MIM: 267300]
B2M	β2-Microglobulin precursor	1	1	Hypercatabolic hypoproteinemia [MIM: 241600]
B4GALT1	UDP-Gal:βGlcNAc β 1,4- galactosyltransferase 1, membrane-bound form	1	1	Congenital disorder of glycosylation type IId [MIM: 607091]
CA2	Carbonic anhydrase II	9	25	Autosomal recessive syndrome of osteopetrosis with renal tubular acidosis [MIM: 259730]
CA4	Carbonic anhydrase IV precursor	2	2	Proximal renal tubular acidosis [MIM: 114760]
CC2D1A	Coiled-coil and C2 domain containing 1A	6	6	Mental retardation autosomal recessive 3 [MIM: 608443]
CD2AP	CD2-associated protein	14	21	Focal segmental glomerulosclerosis FSGS3 [MIM: 607832]
CETP	Cholesteryl ester transfer protein, plasma precursor	7	14	Cholesterol ester transfer protein deficiency [MIM: 607322]
CFH	Complement factor H isoform b precursor	1	1	Hemolytic uremic syndrome, atypical [MIM: 235400]
CFI	Complement factor I	1	1	Complement factor I deficiency [MIM: 610984]
CHMP2B	Chromatin modifying protein 2B	4	15	Frontotemporal dementia, chromosome 3-linked [MIM: 6000795]
CLTC	Clathrin heavy chain 1	12	24	Renal cell carcinoma [MIM: 118955]
COL18A1	α 1 type XVIII collagen isoform 1 precursor	1	1	Knobloch syndrome [MIM: 267750]
COL6A1	Collagen, type VI, α 1 precursor	6	21	Bethlem myopathy [MIM: 158810] Ullrich congenital muscular dystrophy, autosomal dominant [MIM: 254090]
COL6A3	α 3 type VI collagen isoform 5 precursor	1	2	Ullrich congenital muscular dystrophy [MIM: 254090]
CP	Ceruloplasmin precursor	6	15	Aceruloplasminemia [MIM: 604290]
CRYAB	Crystallin, α B	7	12	α-B crystallinopathy [MIM: 608810]
CRYM	Crystallin, μ isoform 1	1	3	Autosomal dominant nonsyndromic deafness [MIM: 123740]
CST3	Cystatin C precursor	1	3	Icelandic-type cerebroarterial amyloidosis [MIM: 105150]
CSTB	Cystatin B	2	10	Myoclonic epilepsy of Unverricht and Lundborg [MIM: 254800]
CTSC	Cathepsin C isoform b precursor	1	1	Papillo-LeFevre syndrome [MIM: 245000]
CTH	Cystathionase isoform 2	1	1	Cystathioninuria [MIM: 219500]
CTSA	Cathepsin A precursor	3	15	Galactosialidosis [MIM: 256540]
CTSC	Cathepsin C isoform a preproprotein	1	2	Papillon-Lefevre syndrome [MIM: 245000]
CTSD	Cathepsin D preproprotein	1	2	Neuronal ceroid lipofuscinosis [MIM: 610127]
CTNS	Cystinosis, nephropathic isoform 1	1	3	Nephropathic cystinosis [MIM: 219800]
CUBN	Cubilin	104	672	Megaloblastic anemia 1, Finnish type [MIM: 261100]
CUL4B	Cullin 4B	1	1	Cabezas syndrome [MIM: 300354]
				Mental retardation-hypotonic facies syndrome [MIM: 300639]
DDAH1	Dimethylarginine dimethylaminohydrolase 1	1	5	Hypertension [MIM: 604743]
DDC	Dopa decarboxylase (aromatic l-amino acid decarboxylase)	11	31	Aromatic L-amino acid decarboxylase deficiency [MIM: 608643]
DNM2	Dynamin 2 isoform 3	1	1	Charcot-Marie-Tooth disease,
				Dominant intermediate B [MIM: 606482]
DNM2	Dynamin 2 isoform 4	1	1	Charcot-Marie-Tooth neuropathy,
				Dominant intermediate B [MIM: 606482]
DYSF	Dysferlin	1	1	Miyoshi myopathy [MIM: 254130]
DPYS	Dihydropyrimidinase	5	6	Dihydropyrimidinuria [MIM: 222748]
DSC2	Desmocollin 2 isoform Dsc2b preproprotein	1	3	Arrhythmogenic right ventricular dysplasia-11 [MIM: 610476]
DSP	Desmoplakin isoform II	10	16	Keratosis palmoplantaris striata II dilated cardiomyopathy with woolly hair and keratoderma [MIM: 605676]
ECE1	Endothelin-converting enzyme 1	1	1	Hirschsprung disease [MIM: 142623]
EFEMP1	EGF-containing fibulin–like extracellular matrix protein 1 precursor	1	3	Doyne Honeycomb retinal dystrophy [MIM: 126600]
ELA2	Elastase 2, neutrophil preproprotein	1	1	Cyclic hematopoiesis [MIM: 162800]
ENPEP	Glutamyl aminopeptidase (aminopeptidase A)	25	100	Hypertension [MIM: 138297]
FAH	Fumarylacetoacetate hydrolase (fumarylacetoacetase)	2	2	Tyrosinemia type I [MIM: 276700]
FLNB	Filamin B, β (actin-binding protein 278)	1	1	Spondylocarpotarsal synostosis syndrome [MIM: 272460]
FBP1	Fructose-1,6-bisphosphatase 1	7	16	Fructose-1,6-bisphosphatase deficiency [MIM: 229700]
FGA	Fibrinogen, α polypeptide isoform α-E preproprotein	5	17	Renal amyloidosis [MIM: 105200] Dysfibrinogenemia [MIM: 134820]
FGG	Fibrinogen, γ chain isoform γ-A precursor	1	1	Dysfibrinogenemia[MIM: 134850]
FTCD	Formiminotransferase cyclodeaminase	4	7	Glutamate formiminotransferase deficiency [MIM: 229100]
FTH1	Ferritin, heavy polypeptide 1	1	7	Iron overload, autosomal dominant [MIM: 134770]
FTL	Ferritin, light polypeptide	5	14	Hyperferritinemia-cataract syndrome [MIM: 600886]
FUCA1	Fucosidase, α-L-1, tissue	1	1	Fucosidosis [MIM: 230000]
FXYD2	FXYD domain–containing ion transport regulator 2 isoform 1	1	7	Hypomagnesemia 2, renal [MIM: 154020]
G6PD	Glucose-6-phosphate dehydrogenase isoform a	1	1	Nonspherocytic hemolytic anemia due to G6PD deficiency [MIM: 305900]
GAA	Acid α-glucosidase preproprotein	4	8	Infantile-onset glycogen storage disease Type II [MIM: 232300]
GALK1	Galactokinase 1	1	1	Galactokinase deficiency [MIM: 230200]
GBE1	Glucan (1,4-α-), branching enzyme 1	1	2	Type IV glycogen storage disease [MIM: 232500]
GCS1	Mannosyl-oligosaccharide glucosidase	1	1	Congenital disorder of glycosylation [MIM: 606056]
GK	Glycerol kinase isoform a	1	1	Glycerol kinase deficiency [MIM: 307030]
GLB1	Galactosidase, β 1 isoform a	16	70	Gangliosidosis GM1 [MIM: 230500]
GLUL	Glutamine synthetase	2	2	Congenital glutamine deficiency [MIM: 610015]
GM2A	GM2 ganglioside activator precursor	2	3	Gangliosidosis GM2 AB variant Tay-Sachs disease [MIM: 272750]
GPI	Glucose phosphate isomerase	9	19	Chronic hemolytic anemia duet to GPI deficiency [MIM: 172400]
GPR98	G protein–coupled receptor 98 precursor	1	1	Familial febrile seizures [MIM: 604352] Usher syndrome type IIC [MIM: 605472]
GSN	Gelsolin isoform b	10	21	Finnish type familial amyloidosis [MIM: 105120]
GSS	Glutathione synthetase	1	3	Glutathione synthetase deficiency [MIM: 266130]
HNMT	Histamine N-methyltransferase isoform 1	1	1	Susceptibility to asthma [MIM: 600807]
HPD	4-Hydroxyphenylpyruvate dioxygenase	1	1	Tyrosinemia type III [MIM: 276710]
HPGD	Hydroxyprostaglandin dehydrogenase 15-(NAD)	6	21	Hypertension [MIM: 601688]
HSPG2	Heparan sulfate proteoglycan 2	18	41	Schwartz-Jampel syndrome type 1 [MIM: 255800]
HSPB1	Heat-shock 27-kD protein 1	8	30	Charcot-Marie-Tooth disease, type 2F [MIM: 606595] Distal hereditary motor neuropathy IIB [MIM: 608634]
ICAM1	Intercellular adhesion molecule 1 precursor	1	1	Graves disease [MIM: 275000]
IL1RN	Interleukin 1 receptor antagonist isoform 1 precursor	2	2	Gastric cancer risk [MIM: 137215]
IRF6	Interferon regulatory factor 6	1	1	Van der Woude syndrome [MIM: 119300] Popliteal pterygium syndrome [MIM: 119500]
ITM2B	Integral membrane protein 2B	5	21	Familial dementia [MIM: 176500]
JUP	Junction plakoglobin	9	15	Naxos disease [MIM: 601214]
KALRN	Kalirin, RhoGEF kinase isoform 3	1	1	Coronary heart disease [MIM: 608901]
KHK	Ketohexokinase isoform a	2	4	Essential fructosuria [MIM: 229800]
KL	Klotho	1	1	Hyperphosphatemic tumoral calcinosis [MIM: 211900]
KLK1	Kallikrein 1 preproprotein	1	1	Decreased urinary activity of kallikrein [MIM: 147910]
LGALS3	Galectin 3	1	1	Lymphocyte function–associated antigen 1 [MIM: 116920]
LAMP2	Lysosomal-associated membrane protein 2 precursor	4	26	Danon disease [MIM: 300257]
LRRK2	Leucine-rich repeat kinase 2	4	5	Parkinson disease [MIM: 607060]
LYZ	Lysozyme precursor	1	3	Familial visceral amyloidosis [MIM: 105200]
MIF	Macrophage migration inhibitory factor (glycosylation-inhibiting factor)	1	6	Rheumatoid arthritis [MIM: 604302]
MME	Membrane metallo-endopeptidase neprilysin	48	311	HypertensionImportant cell surface marker in the diagnostic of human acute lymphocytic leukemia [MIM: 120520]
MPO	Myeloperoxidase	7	32	Myeloperoxidase deficiency [MIM: 254600]
MTHFD1	Methylenetetrahydrofolate dehydrogenase 1	4	5	Spina bifida [MIM: 601634]
MYH14	Myosin, heavy chain 14 isoform 1	1	1	Autosomal dominant nonsyndromic sensorineural deafness [MIM: 600652]
MYH3	Myosin, heavy chain 3, skeletal muscle, embryonic	1	1	Freeman-Sheldon syndrome [MIM: 193700]
MYH9	Myosin, heavy polypeptide 9, nonmuscle	19	51	Fechtner syndrome [MIM: 153640]Epstein syndrome [MIM: 153650]
MYO15A	Myosin XV	1	3	Recessive congenital deafness [MIM: 600316]
MYO6	Myosin VI	7	21	Autosomal recessive congenital sensorineural deafness [MIM: 607821] Autosomal dominant nonsyndromic sensorineural deafness [MIM: 606346]
NAGLU	α-N-acetylglucosaminidase precursor	21	63	Mucopolysaccharidosis type IIIB [MIM: 252920]
NDRG1	N-myc downstream regulated gene 1	2	5	Charcot-Marie-Tooth disease type 4D [MIM: 601455]
NEB	Nebulin	2	4	Nemaline myopathy [MIM: 256030]
NPHS2	Podocin	6	9	Autosomal recessive steroid-resistant nephrotic syndrome [MIM: 600995]
PAFAH1B1	Platelet-activating factor acetylhydrolase, isoform Ib, α subunit (45 kD)	1	1	Miller-Dieker lissencephaly syndrome [MIM: 607432]
PARK7	DJ-1 protein	1	1	Parkinson disease 7, autosomal recessive [MIM: 606324]
PCBD1	Pterin-4 α-carbinolamine dehydratase precursor	1	1	Hyperphenylalaninemia [MIM: 264070]
PDCD10	Programmed cell death 10	2	3	Cerebral cavernous malformations [MIM: 603285]
PHGDH	Phosphoglycerate dehydrogenase	2	2	Phosphoglycerate dehydrogenase deficiency [MIM: 601815]
PKD1	Polycystin 1	1	1	Polycystic kidney disease, adult, type I [MIM: 601313]
PKD2	Polycystin 2	1	2	Polycystic kidney disease, adult, type II [MIM: 173910]
PKHD1	Polyductin isoform 2	6	9	Autosomal recessive polycystic kidney disease [MIM: 263200]
PKLR	Pyruvate kinase, liver, and RBC isoform 1	1	1	Pyruvate kinase deficiency [MIM: 266200]
PLOD1	Lysyl hydroxylase precursor	1	1	Ehlers-Danlos syndrome, type VIA [MIM: 225400]
PRKCH	Protein kinase C, η	1	2	Cerebral infarction [MIM: 601367]
PROM1	Prominin 1	23	174	Autosomal recessive retinal degeneration [MIM: 604365]
PRNP	Prion protein preproprotein	1	1	Creutzfeldt-Jakob disease [MIM: 123400]
PSAP	Prosaposin isoform a preproprotein	3	6	Metachromatic leukodystrophy due to SAP1 deficiency [MIM: 249900]
				Gaucher disease, atypical due to SAP2 deficiency [MIM: 610539]
PSAP	Prosaposin isoform c preproprotein	1	4	Metachromatic leukodystrophy [MIM: 249900]
PSAT1	Phosphoserine aminotransferase isoform 1	2	4	Phosphoserine aminotransferase deficiency [MIM: 610992]
PTPRJ	Protein tyrosine phosphatase, receptor type, J precursor	1	1	Somatic colon cancer [MIM: 114500]
RAB3GAP1	RAB3 GTPase-activating protein	1	1	Warburg micro syndrome [MIM: 600118]
RBP4	Retinol-binding protein 4, plasma precursor	2	3	Retinol-binding protein deficiency [MIM: 180250]
RDX	Radixin	16	23	Autosomal recessive deafness 24 [MIM: 611022]
ROBO2	Roundabout, axon guidance receptor, homolog 2	1	1	Vesicoureteral reflux 2 [MIM: 610878]
RP2	XRP2 protein	3	5	X-linked retinitis pigmentosa 2 [MIM: 312600]
RYR1	Skeletal muscle ryanodine receptor isoform 1	1	1	Malignant hyperthermia [MIM: 145600] Central core disease [MIM: 117000] Minicore myopathy with external ophthalmoplegia [MIM: 255320]
SERPING1	Complement component 1 inhibitor precursor	7	14	Hereditary angioedema type I [MIM: 106100]
SLC3A1	Solute carrier family 3, member 1	14	25	Cystinuria [MIM: 220100]
SLC4A1	Solute carrier family 4, anion exchanger, member 1 [kAE1]	2	2	Defective kidney acid secretion leading to distal renal tubular acidosis [MIM: 179800]
SLC4A4	Solute carrier family 4, sodium bicarbonate co-transporter, member 4 [NBC1]	2	3	Renal tubular acidosis, proximal, with ocular abnormalities [MIM: 604278]
SLC5A1	Solute carrier family 5 (sodium/glucose co-transporter), member 1 [SGLT1]	2	3	Glucose/galactose malabsorption [MIM: 606824]
SLC5A2	Solute carrier family 5 (sodium/glucose co-transporter), member 2 [SGLT2]	4	9	Renal glucosuria [MIM: 233100]
SLC6A19	Solute carrier family 6, member 19	4	8	Hartnup disorder [MIM: 234500]
SLC12A1	Sodium potassium chloride co-transporter 2 [NKCC2]	25	94	Bartter syndrome, antenatal, type 1 [MIM: 601678]
SLC12A3	Solute carrier family 12 (sodium/chloride transporters), member 3 [NCC]	28	102	Gitelman syndrome [MIM: 263800]
SLC22A12	Urate anion exchanger 1 isoform a [URAT1]	1	2	Renal hypouricemia [MIM: 220150]
SLC25A3	Solute carrier family 25 member 3 isoform b precursor	1	5	Mitochondrial phosphate carrier deficiency [MIM: 610773]
SLC26A4	Pendrin	2	4	Pendred syndrome [MIM: 274600]Deafness, autosomal recessive 4 [MIM: 600791]
SLC44A4	NG22 protein isoform 1	6	59	Sialidosis 1 [MIM: 606107]
SPR	Sepiapterin reductase (7,8-dihydrobiopterin:NADP + oxidoreductase)	1	2	Dystonia, dopa-responsive, due to sepiapterin reductase deficiency [MIM: 251120]
SQSTM1	Sequestosome 1	1	2	Paget disease of bone [MIM: 602080]
SUCLA2	Succinate-CoA ligase, ADP-forming, β subunit	1	1	Mitochondrial DNA depletion syndrome [MIM: 609560]
TECTA	Tectorin α precursor	1	1	Autosomal dominant nonsyndromic sensorineural hearing loss [MIM: 601842]
TF	Transferrin	12	20	Alzheimer disease [MIM: 104300]
TPP1	Tripeptidyl-peptidase I preproprotein	8	42	Ceroid lipofuscinosis neuronal 2 [MIM: 204500]
TSG101	Tumor susceptibility gene 101	17	66	Breast cancer [MIM: 176960]
TTN	Titin isoform novex 1	4	5	Cardiomyopathy [MIM: 188840]
UMOD	Uromodulin precursor	35	1278	Medullary cystic kidney disease-2 (MCKD2) [MIM: 603860]Familial juvenile hyperuricemic nephropathy (FJHN) [MIM: 16200]
VCP	Valosin-containing protein	2	2	Inclusion body myopathy with early-onset Paget disease and frontotemporal dementia [MIM: 167320]
VAMP7	Vesicle-associated membrane protein 7	1	1	β-Ureidopropionase deficiency [MIM: 606673]
VCL	Vinculin isoform meta-VCL	3	5	Cardiomyopathy, dilated [MIM: 611407]
VWF	Von Willebrand factor preproprotein	1	4	Von Willebrand disease [MIM: 193400]
ZMPSTE24	Zinc metalloproteinase STE24	1	1	Mandibuloacral dysplasia [MIM: 608612]

^{Information for each protein include “Gene” name, “Protein Name”, “Pep” refers to the number of unique peptides identified in LC-MS/MS, “ID” refers to the number of spectra and “Related to Disease [OMIM]” refers to the disease with which the protein is related according to OMIM. The 34 proteins associated with kidney diseases are presented in italics.}

Predictably, a large number of proteins that were identified were integral membrane proteins involved in solute and water transport (Table 2). As seen in our previous study,¹ these proteins predominantly represent apical transporters present in every renal tubule segment, including the proximal tubule (sodium-hydrogen exchanger 3, sodium-glucose co-transporter 1 and 2, and aquaporin-1 [AQP1]), the thick ascending limb (sodium-potassium-chloride co-transporter 2 [NKCC2]), the distal convoluted tubule (thiazide-sensitive Na-Cl co-transporter [NCC]), and connecting tubule/collecting duct (AQP2, rhesus blood group C glycoprotein [RhCG, an ammonia channel], B1 subunit of vacuolar H^{-ATPase, and pendrin). Note that both polycystin-1 and polycystin-2 were detected in human urinary exosomes.}

Table 2.

Solute and water transporters^a

Ref Seq	Gene	Protein Name	Pep	ID
NP_000918	ABCB1	ATP-binding cassette, subfamily B, member 1	23	61
NP_003733	ABCB11	ATP-binding cassette, subfamily B (MDR/TAP), member 11	1	3
NP_005680	ABCB6	ATP-binding cassette, subfamily B, member 6	1	1
NP_064694	ABCC9	ATP-binding cassette, subfamily C, member 9 isoform SUR2A-δ-14	1	2
NP_149163	ABCC11	ATP-binding cassette, subfamily C, member 11 isoform a	1	1
NP_932766	AQP1	Aquaporin 1	3	35
NP_000477	AQP2	Aquaporin 2	7	36
NP_000692	ATP1A1	Na/K-ATPase α 1 subunit isoform a proprotein	19	57
NP_001001787	ATP1B1	Na/K-ATPase β 1 subunit isoform b	1	1
NP_001001937	ATP5A1	ATP synthase, H transporting, mitochondrial F1 complex, α subunit precursor	3	5
NP_001677	ATP5B	ATP synthase, H transporting, mitochondrial F1 complex, β subunit precursor	4	5
NP_001174	ATP6AP1	ATPase, H transporting, lysosomal accessory protein 1 precursor	1	1
NP_001685	ATP6V0C	ATPase, H transporting, lysosomal, V0 subunit c	1	9
NP_005168	ATP6V0A1	ATPase, H transporting, lysosomal V0 subunit a isoform 1	1	1
NP_065683	ATP6V0A4	ATPase, H transporting, lysosomal V0 subunit a4	1	2
NP_004682	ATP6V0D1	ATPase, H transporting, lysosomal, V0 subunit d1	1	1
NP_689778	ATP6V0D2	ATPase, H transporting, lysosomal 38 kD, V0 subunit D2	2	4
NP_001681	ATP6V1A	ATPase, H transporting, lysosomal 70 kD, V1 subunit A, isoform 1	22	49
NP_001683	ATP6V1B1	ATPase, H transporting, lysosomal 56/58 kD, V1 subunit B1	8	17
NP_001684	ATP6V1B2	Vacuolar H-ATPase B2	12	27
NP_001686	ATP6V1C1	ATPase, H transporting, lysosomal 42 kD, V1 subunit C1 isoform A	1	1
NP_001034451	ATP6V1C2	Vacuolar H-ATPase C2 isoform a	1	1
NP_057078	ATP6V1D	H-transporting two-sector ATPase	2	3
NP_001687	ATP6V1E1	Vacuolar H-ATPase E1 isoform a	2	2
NP_001034456	ATP6V1E1	Vacuolar H-ATPase E1 isoform c	1	1
NP_001034455	ATP6V1E1	Vacuolar H-ATPase E1 isoform b	1	1
NP_004222	ATP6V1F	ATPase, H transporting, lysosomal 14 kD, V1 subunit F	1	1
NP_004879	ATP6V1G1	Vacuolar H-ATPase G1	1	2
NP_998784	ATP6V1H	ATPase, H transporting, lysosomal 50/57 kD, V1 subunit H isoform 1	8	34
NP_036415	KCNG2	Potassium voltage-gated channel, subfamily G, member 2	1	2
NP_853514	PKD1L3	Polycystin 1–like 3	1	1
NP_001009944	PKD1	Polycystin 1 isoform 1 precursor	1	1
NP_000288	PKD2	Polycystin 2	1	2
NP_057405	RHCG	Rhesus blood group, C glycoprotein	5	8
NP_000531	RYR1	Skeletal muscle ryanodine receptor isoform 1	1	1
NP_006505	SCN10A	Sodium channel, voltage-gated, type X, α	1	11
NP_054858	SCN11A	Sodium channel, voltage-gated, type XI, α	1	1
NP_000329	SLC12A1	Sodium potassium chloride co-transporter 2	25	94
NP_000330	SLC12A3	Solute carrier family 12 (sodium/chloride transporters), member 3	28	102
NP_064631	SLC12A9	Solute carrier family 12 (potassium/chloride transporters), member 9	1	1
NP_003975	SLC13A2	Solute carrier family 13 (sodium-dependent dicarboxylate transporter), member 2	5	10
NP_073740	SLC13A3	Solute carrier family 13 member 3 isoform a	2	2
NP_001011554	SLC13A3	Solute carrier family 13 member 3 isoform b	1	3
NP_004161	SLC1A1	Solute carrier family 1, member 1	3	6
NP_066568	SLC15A2	Solute carrier family 15 (H/peptide transporter), member 2	1	1
NP_060954	SLC22A11	Solute carrier family 22 member 11	2	9
NP_653186	SLC22A12	Urate anion exchanger 1 isoform a	1	2
NP_003049	SLC22A2	Solute carrier family 22 member 2 isoform a	2	3
NP_003051	SLC22A5	Solute carrier family 22 member 5	1	1
NP_004781	SLC22A6	Solute carrier family 22 member 6 isoform a	1	3
NP_695010	SLC22A6	Solute carrier family 22 member 6 isoform d	1	1
NP_004245	SLC22A8	Solute carrier family 22 member 8	1	2
NP_005838	SLC23A1	Solute carrier family 23 (nucleobase transporters), member 1 isoform a	4	6
NP_689898	SLC23A1	Solute carrier family 23 (nucleobase transporters), member 1 isoform b	6	10
NP_005975	SLC25A1	Solute carrier family 25 (mitochondrial carrier; citrate transporter), member 1	1	1
NP_998776	SLC25A3	Solute carrier family 25 member 3 isoform b precursor	1	5
NP_775897	SLC26A11	Solute carrier family 26, member 11	1	1
NP_000432	SLC26A4	Pendrin	2	4
NP_003030	SLC2A5	Solute carrier family 2 (facilitated glucose/fructose transporter), member 5	12	22
NP_775867	SLC39A5	Solute carrier family 39 (metal ion transporter), member 5	1	1
NP_000332	SLC3A1	Solute carrier family 3, member 1	14	25
NP_001012679	SLC3A2	Solute carrier family 3 (activators of dibasic and neutral amino acid transport), member 2 isoform a	10	18
NP_002385	SLC3A2	Solute carrier family 3 (activators of dibasic and neutral amino acid transport), member 2 isoform c	3	5
NP_536856	SLC44A1	CDW92 antigen	1	1
NP_065161	SLC44A2	CTL2 protein	18	70
NP_000333	SLC4A1	Solute carrier family 4, anion exchanger, member 1	2	2
NP_003750	SLC4A4	Solute carrier family 4, sodium bicarbonate co-transporter, member 4	2	3
NP_000334	SLC5A1	Solute carrier family 5 (sodium/glucose co-transporter), member 1	2	3
NP_689564	SLC5A10	Solute carrier family 5 (sodium/glucose co-transporter), member 10 isoform 1	2	2
NP_001035915	SLC5A10	Solute carrier family 5 (sodium/glucose co-transporter), member 10 isoform 2	1	3
NP_848593	SLC5A12	Solute carrier family 5 (sodium/glucose co-transporter), member 12 isoform 2	2	4
NP_003032	SLC5A2	Solute carrier family 5 (sodium/glucose co-transporter), member 2	4	9
NP_666018	SLC5A8	Solute carrier family 5 (iodide transporter), member 8	1	1
NP_001011547	SLC5A9	Solute carrier family 5 (sodium/glucose co-transporter), member 9	1	3
NP_057699	SLC6A13	Solute carrier family 6 (neurotransmitter transporter, GABA), member 13	1	1
NP_001003841	SLC6A19	Solute carrier family 6, member 19	4	8
NP_004165	SLC9A3	Solute carrier family 9 (sodium/hydrogen exchanger), isoform 3	2	3
NP_004776	SLC9A3R2	Solute carrier family 9 isoform 3 regulator 2	1	1
NP_851322	SLCO4C1	Solute carrier organic anion transporter family, member 4C1	2	2
NP_003365	VDAC1	Voltage-dependent anion channel 1	6	43
NP_005653	VDAC3	Voltage-dependent anion channel 3	1	1

^{Table contains all of the proteins that are solute and water transporters.}

Exosomes derive from MVB and are delivered to the urine when the outer membranes of MVB fuse with the apical plasma membrane. Interestingly, 22 of the proteins identified in this study are recognized as components of the apparatus responsible for the formation of MVB (Table 3). These 22 proteins account for approximately 75% of the proteins that constitute the ESCRT-0, ESCRT-I, ESCRT-II, and ESCRT-III complexes involved in multivesicular body formation.⁵

Table 3.

Proteins of the ESCRT complex found in human urinary exosomes^a

Gene	Protein Name	Pep	ID	Ref Seq	ESCRT Complex
HGS	Hepatocyte growth factor–regulated tyrosine kinase substrate	1	1	NP_004703	ESCRT-0
TSG101	Tumor susceptibility gene 101	15	47	NP_006283	ESCRT-I
VPS28	Vacuolar protein sorting 28 isoform 1	5	8	NP_057292	ESCRT-I
VPS28	Vacuolar protein sorting 28 isoform 2	2	8	NP_898880	ESCRT-I
VPS37B	Vacuolar protein sorting 37B	4	10	NP_078943	ESCRT-I
VPS37C	Vacuolar protein sorting 37C	1	1	NP_060436	ESCRT-I
VPS25	EAP25	4	15	NP_115729	ESCRT-II
VPS36	EAP45	2	3	NP_057159	ESCRT-II
SNF8	EAP30	1	1	NP_009172	ESCRT-II
CHMP2A	CHMP2A	6	40	NP_055268	ESCRT-III
CHMP2B	CHMP2B	3	11	NP_054762	ESCRT-III
VPS24	CHMP3	1	3	NP_057163	ESCRT-III
VPS24	CHMP3	1	4	NP_001005753	ESCRT-III
CHMP4B	CHMP4B	2	6	NP_789782	ESCRT-III
CHMP5	CHMP5	2	7	NP_057494	ESCRT-III
CHMP1A	CHMP1A	1	3	NP_002759	ESCRT-III
CHMP1B	CHMP1B	1	2	NP_065145	ESCRT-III
CHMP6	CHMP6	2	3	NP_078867	ESCRT-III
VPS4A	Vacuolar protein sorting factor 4A	11	25	NP_037377	ATPase complex
VPS4B	Vacuolar protein sorting factor 4B	11	32	NP_004860	ATPase complex
PDCD6IP	ALIX	27	104	NP_037506	Accessory
C1orf58	Hypothetical protein LOC148362	11	34	NP_653296	Accessory

^{Table contains all of the proteins that are members of the ESCRT Complex (ESCRT-0, ESCRT-I, ESCRT-II, ESCRT-III, ATPase complex and accessory).}

In addition, 17 proteins identified in this study are subunits of the human vacuolar H^{-ATPase (}Table 4). Vacuolar H^{-ATPases are ATP-dependent proton pumps for proton transport into intracellular organelles.6} These proteins also mediate proton transport across the apical plasma membrane of type A intercalated cells and across the basolateral plasma membrane of type B intercalated cells.⁷ The B1 subunit is selectively expressed in intercalated cells, and its detection in urinary exosomes establish that intercalated cells secrete exosomes as do other types of epithelial cells lining the renal tubule. These proteins constitute 78% of the subunits of the V₀ and V₁ domains of the vacuolar H^-ATPase.8

Table 4.

Vacuolar H^{-ATPase subunits in human urinary exosomesa}

Ref Seq	Gene	Protein Name	Pep	ID
NP_005168	ATP6V0A1	ATPase, H transporting, lysosomal V0 subunit a isoform 1	1	1
NP_065683	ATP6V0A4	ATPase, H transporting, lysosomal V0 subunit a4	1	2
NP_001685	ATP6V0C	ATPase, H transporting, lysosomal, V0 subunit c	1	1
NP_004682	ATP6V0D1	ATPase, H transporting, lysosomal, V0 subunit d1	1	1
NP_689778	ATP6V0D2	ATPase, H transporting, lysosomal 38 kD, V0 subunit D2	2	4
NP_001681	ATP6V1A	ATPase, H transporting, lysosomal 70 kD, V1 subunit A, isoform 1	22	49
NP_001683	ATP6V1B1	ATPase, H transporting, lysosomal 56/58 kD, V1 subunit B1	8	17
NP_001684	ATP6V1B2	vacuolar H-ATPase B2	12	27
NP_001686	ATP6V1C1	ATPase, H transporting, lysosomal 42 kD, V1 subunit C1 isoform A	1	1
NP_001034451	ATP6V1C2	vacuolar H-ATPase C2 isoform a	1	1
NP_057078	ATP6V1D	H(+)-transporting two-sector ATPase	2	3
NP_001687	ATP6V1E1	vacuolar H-ATPase E1 isoform a	2	2
NP_001034455	ATP6V1E1	vacuolar H-ATPase E1 isoform b	1	1
NP_001034456	ATP6V1E1	vacuolar H-ATPase E1 isoform c	1	1
NP_004222	ATP6V1F	ATPase, H transporting, lysosomal 14 kD, V1 subunit F	1	1
NP_004879	ATP6V1G1	vacuolar H-ATPase G1	1	2
NP_998784	ATP6V1H	ATPase, H transporting, lysosomal 50/57 kD, V1 subunit H isoform 1	8	34

^{Table contains proteins that are found in human urinary exosomes and are subunits of the human vacuolar H-ATPase.}

An example of the utility of exosome analysis is shown in Figure 1, describing immunoblotting in patients with Bartter syndrome type I, associated with mutations in the SLC12A1 gene, which encodes for the NKCC2 sodium-potassium-chloride co-transporter protein.⁹ The NKCC2 protein was found in the proteome of the human urinary exosomes as shown in Table 1. Urine samples were obtained from two patients (patients 1 and 2) with clinical phenotypes consistent with Bartter syndrome type I (Figure 1A).¹⁰ The clinical diagnosis for the patients with Bartter syndrome type I was confirmed by the ultrasound images showing deposits of calcium in the kidney also known as nephrocalcinosis⁹10 (Figure 1B) and other typical laboratory findings. The two urinary exosome samples obtained from the patients with Bartter syndrome type I were analyzed by immunoblotting for the presence of the NKCC2 protein (Figure 1C). Compared with the respective control samples, patients 1 and 2 showed an absence of the NKCC2 protein bands, expected at 160 kD for monomeric NKCC2 and 320 kD for dimeric NKCC2. In addition, the samples (patients 1 and 2) were probed for the thiazide-sensitive co-transporter (NCC) protein to ensure that urinary exosomes were successfully isolated and loaded properly. Strong NCC bands were obtained in samples from both patients with Bartter syndrome type I and control samples.

Figure 1.

Disease-related protein: NKCC2 and Bartter syndrome type I. (A) Details of clinical phenotype for patients with Bartter syndrome type I, patient 1 and patient 2. (B) Ultrasound images showing calcium deposits (white arrowheads) in the kidneys of patients 1 and 2. (C) Immunoblot of urinary exosomes samples from patient 1, patient 2, control 1, and control 2 using polyclonal rabbit anti-NKCC2 and NCC antibodies.

Phosphoproteomic Analysis of Human Urinary Exosomes

Protein phosphorylation is a key element of most cell regulatory processes. Recently, technical approaches that allow phosphoproteomic profiling on a large scale have been introduced.⁴1113 We used neutral loss scanning with high-stringency target-decoy analysis to identify phosphorylation sites present in exosomal proteins from human urine samples.

Nineteen phosphorylation sites corresponding to 14 phosphoproteins were identified (Table 5). These included both newly identified phosphorylation sites and sites that had been previously identified. Two orphan G-protein–coupled receptors are included in the former group, viz. GPRC5B and GPRC5C. In GPRC5B, we identified one new phosphorylation site, T389, and, in GPRC5C, we identified three new phosphorylation sites, T435, S395, and Y426. These proteins are also known as retinoic acid–induced gene 2 (GPRC5B) and retinoic acid–induced gene 3 (GPRC5C).

Table 5.

Human urinary exosome phosphopeptides^a

Ref Seq	Protein Name, Sequence	Site	Gene	Novel Site	MSn	Motif	GO Function
NP_061123	G protein–coupled receptor family C, group 5, member C isoform b		GPRC5C				Metabotropic glutamate, GABA-B–like receptor activity Protein binding Receptor activity
	R.AEDMYSAQSHQAA(T*)PPKDGK.N	T435		Yes	MS2, MS3	Proline-directed
	K.VP(S*)EGAYDIILPR.A	S395		Yes	MS2, MS3	Phosphoserine/threonine binding group
	R.AEDM(Y*)SAQSHQAATPPKDGK.N	Y426		Yes	MS2	Tyrosine kinase
NP_001035149	Secreted phosphoprotein 1 isoform c		SPP1				Cytokine activity Growth factor activity Integrin binding Protein binding
	K.AIPVAQDLNAPSDWD(S*)R.G	S192		No	MS2	Miscellaneous
	R.GKD(S*)YETSQLDDQSAETHSHK.Q	S197			MS2, MS3	Basophilic
	R.GKDSYETSQLDDQ(S*)AETHSHK.Q	S207			MS2, MS3	Acidophilic
NP_057319	G protein–coupled receptor, family C, group 5, member B precursor		GPRC5B				Metabotropic glutamate, GABA-B–like receptor activity Receptor activity Sevenless binding
	R.SNVYQPTEMAVVLNGG(T*)IPTAPPSHTGR.H	T389		Yes	MS2	Basophilic
NP_000477	Aquaporin 2		AQP2				Transporter activity Water channel activity
	R.RQ(S*)VELHSPQSLPR.G	S256		No	MS2, MS3	Basophilic
NP_004860	Vacuolar protein sorting factor 4B		VPS4B				ATP binding ATPase activity, Coupled nucleotide binding Protein binding
	K.EGQPSPADEKGND(S*)DGEGESDDPEKKK.L	S102		Yes	MS2	Acidophilic
NP_054762	Chromatin modifying protein 2B		CHMP2B				Not classified
	K.ATI(S*)DEEIER.Q	S199		No	MS2, MS3 (unfiltered)	Acidophilic
NP_687033	Proteasome α 3 subunit isoform 2		PSAM3				Protein binding Threonine endopeptidase activity
	K.ESLKEEDE(S*)DDDNM	S243		No	MS2, MS3	Acidophilic
NP_036382	Related RAS viral (r-ras) oncogene homolog 2		RRAS2				GTP binding Nucleotide binding Protein binding
	R.KFQEQECPP(S*)PEPTRK.E	S186		Yes	MS2, MS3	Proline-directed
NP_031381	Heat-shock 90-kD protein 1, β		HSP90AB1				ATP binding Nitric-oxide synthase regulator activity Nucleotide binding TPR domain binding Unfolded protein binding
	K.IEDVG(S*)DEEDDSGKDKK.K	S255		No	MS2	Acidophilic
NP_612433	Kinesin family member 12		KIF12				ATP binding Microtubule motor activity Nucleotide binding
	R.VTTRPQAPK(S*)PVAK.Q	S236		Yes	MS2, MS3	Proline-directed
NP_079119	Cytochrome b reductase 1		CYBRD1				Ferric-chelate reductase activity
	R.NLALDEAGQRS(T*)M.	T285		Yes	MS2	Basophilic
NP_001037857	Mucin 1 isoform 7 precursor		MUC1				NF-κB binding protein heterodimerization activity
	R.DTYHPMSEYPTYH(T*)HGR.Y	T118		Yes	MS2, MS3 (unfiltered)	Acidophilic	Protein homodimerization activity RNA binding Tat protein binding Unfolded protein binding
NP_000330	Solute carrier family 12 (sodium/chloride transporters), member 3		SLC12A3				Sodium ion binding Sodium:chloride
	R.GARP(S*)VSGALDPK.A	S811		Yes	MS2	Basophilic	symporter activity Symporter activity Transporter activity
NP_000329	Sodium potassium chloride co-transporter 2		SLC12A1				Potassium ion binding Sodium ion binding
	K.IEYYRN(T*)GSISGPK.V	T118		No	MS2, MS3	N/A	Sodium:potassium:chloride symporter activity
	K.IEYYRNTG(S*)ISGPK.V	S120		No	MS2, MS3	Basophilic	Symporter activity Transporter activity

^{Table contains phosphopeptides found in urinary exosomes. MS refers to spectra for phosphorylation site identification, Motif refers to phosphorylation motif site and GO Function.}

A new phosphorylation site was also identified in the COOH-terminal tail of the thiazide-sensitive co-transporter (NCC) at S811 (Figure 2). This site is distinct from the N-terminal site previously identified¹⁴ and may play a role in regulation of transport. This amino acid is conserved in humans, chimpanzees, rhesus monkeys, and horses but not in mice and rats. Simon et al.¹⁵ showed that, in rat, the amino acid sequence surrounding this site is absent owing to a difference in exon splicing.

Figure 2.

Novel phosphorylation site in the NCC. The serine-811 on the NCC protein is phosphorylated. (A) The phosphorylation site on the peptide is denoted by an asterisk (*). (B) The neutral loss peak (NL) from the +2 mass spectrum and the site-determining ions b5, b6, y7, and y8.

Novel phosphorylation sites were also identified in RRAS2 (TC21), VPS4B (an ESCRT component), cytochrome b reductase, proteasome α 3 subunit, and mucin 1. This study also revealed previously identified phosphorylation sites in AQP2 (S256),¹⁶ NKCC2 (T118 and S120),¹⁷ CHMP2B (S199),¹⁸ HSP90AB1 (S255),¹⁹ and SPP1 (S192, S197, and S207).²⁰ Phosphorylation of AQP2 at S256 was confirmed by immunoblotting human urinary exosomes samples with a phospho-specific antibody for this site (Figure 3).

Figure 3.

Detection of AQP2-S256 phosphorylation in urinary exosomes. IMCD, rat inner medullary collecting duct treated with dDAVP (V₂R-selective vasopressin analog) for 30 min; exo (10 μg), is human urinary exosomes, 10 μg; exo (72 μg), human urinary exosomes, 72 μg.

Large-Scale Proteomic Profiling of Human Urinary Exosomes

In this study, we carried out proteomic profiling of a low-density membrane fraction from human urine consisting chiefly of exosomes, using a highly sensitive LC-MS/MS system, based on an ion trap mass spectrometer (LTQ; Thermo-Finnigan; Thermo Electron, San Jose, CA). We unambiguously identified 1132 proteins including 205 proteins seen in our previous study and 927 proteins not seen in our previous study of human urinary exosomes.¹ The full list (ambiguous and unambiguous identifications) contains 1412 proteins and can be viewed in Supplemental Table 1, and the list of proteins that were unambiguously identified in both studies can be viewed at http://dir.nhlbi.nih.gov/papers/lkem/exosome/. The expanded list of exosomal proteins includes 177 proteins that are disease related, on the basis of their presence in the OMIM database (Table 1).

Predictably, a large number of proteins that were identified were integral membrane proteins involved in solute and water transport (Table 2). As seen in our previous study,¹ these proteins predominantly represent apical transporters present in every renal tubule segment, including the proximal tubule (sodium-hydrogen exchanger 3, sodium-glucose co-transporter 1 and 2, and aquaporin-1 [AQP1]), the thick ascending limb (sodium-potassium-chloride co-transporter 2 [NKCC2]), the distal convoluted tubule (thiazide-sensitive Na-Cl co-transporter [NCC]), and connecting tubule/collecting duct (AQP2, rhesus blood group C glycoprotein [RhCG, an ammonia channel], B1 subunit of vacuolar H^{-ATPase, and pendrin). Note that both polycystin-1 and polycystin-2 were detected in human urinary exosomes.}

Exosomes derive from MVB and are delivered to the urine when the outer membranes of MVB fuse with the apical plasma membrane. Interestingly, 22 of the proteins identified in this study are recognized as components of the apparatus responsible for the formation of MVB (Table 3). These 22 proteins account for approximately 75% of the proteins that constitute the ESCRT-0, ESCRT-I, ESCRT-II, and ESCRT-III complexes involved in multivesicular body formation.⁵

In addition, 17 proteins identified in this study are subunits of the human vacuolar H^{-ATPase (}Table 4). Vacuolar H^{-ATPases are ATP-dependent proton pumps for proton transport into intracellular organelles.6} These proteins also mediate proton transport across the apical plasma membrane of type A intercalated cells and across the basolateral plasma membrane of type B intercalated cells.⁷ The B1 subunit is selectively expressed in intercalated cells, and its detection in urinary exosomes establish that intercalated cells secrete exosomes as do other types of epithelial cells lining the renal tubule. These proteins constitute 78% of the subunits of the V₀ and V₁ domains of the vacuolar H^-ATPase.8

An example of the utility of exosome analysis is shown in Figure 1, describing immunoblotting in patients with Bartter syndrome type I, associated with mutations in the SLC12A1 gene, which encodes for the NKCC2 sodium-potassium-chloride co-transporter protein.⁹ The NKCC2 protein was found in the proteome of the human urinary exosomes as shown in Table 1. Urine samples were obtained from two patients (patients 1 and 2) with clinical phenotypes consistent with Bartter syndrome type I (Figure 1A).¹⁰ The clinical diagnosis for the patients with Bartter syndrome type I was confirmed by the ultrasound images showing deposits of calcium in the kidney also known as nephrocalcinosis⁹10 (Figure 1B) and other typical laboratory findings. The two urinary exosome samples obtained from the patients with Bartter syndrome type I were analyzed by immunoblotting for the presence of the NKCC2 protein (Figure 1C). Compared with the respective control samples, patients 1 and 2 showed an absence of the NKCC2 protein bands, expected at 160 kD for monomeric NKCC2 and 320 kD for dimeric NKCC2. In addition, the samples (patients 1 and 2) were probed for the thiazide-sensitive co-transporter (NCC) protein to ensure that urinary exosomes were successfully isolated and loaded properly. Strong NCC bands were obtained in samples from both patients with Bartter syndrome type I and control samples.

Figure 1.

Disease-related protein: NKCC2 and Bartter syndrome type I. (A) Details of clinical phenotype for patients with Bartter syndrome type I, patient 1 and patient 2. (B) Ultrasound images showing calcium deposits (white arrowheads) in the kidneys of patients 1 and 2. (C) Immunoblot of urinary exosomes samples from patient 1, patient 2, control 1, and control 2 using polyclonal rabbit anti-NKCC2 and NCC antibodies.

Phosphoproteomic Analysis of Human Urinary Exosomes

Protein phosphorylation is a key element of most cell regulatory processes. Recently, technical approaches that allow phosphoproteomic profiling on a large scale have been introduced.⁴1113 We used neutral loss scanning with high-stringency target-decoy analysis to identify phosphorylation sites present in exosomal proteins from human urine samples.

Nineteen phosphorylation sites corresponding to 14 phosphoproteins were identified (Table 5). These included both newly identified phosphorylation sites and sites that had been previously identified. Two orphan G-protein–coupled receptors are included in the former group, viz. GPRC5B and GPRC5C. In GPRC5B, we identified one new phosphorylation site, T389, and, in GPRC5C, we identified three new phosphorylation sites, T435, S395, and Y426. These proteins are also known as retinoic acid–induced gene 2 (GPRC5B) and retinoic acid–induced gene 3 (GPRC5C).

A new phosphorylation site was also identified in the COOH-terminal tail of the thiazide-sensitive co-transporter (NCC) at S811 (Figure 2). This site is distinct from the N-terminal site previously identified¹⁴ and may play a role in regulation of transport. This amino acid is conserved in humans, chimpanzees, rhesus monkeys, and horses but not in mice and rats. Simon et al.¹⁵ showed that, in rat, the amino acid sequence surrounding this site is absent owing to a difference in exon splicing.

Figure 2.

Novel phosphorylation site in the NCC. The serine-811 on the NCC protein is phosphorylated. (A) The phosphorylation site on the peptide is denoted by an asterisk (*). (B) The neutral loss peak (NL) from the +2 mass spectrum and the site-determining ions b5, b6, y7, and y8.

Novel phosphorylation sites were also identified in RRAS2 (TC21), VPS4B (an ESCRT component), cytochrome b reductase, proteasome α 3 subunit, and mucin 1. This study also revealed previously identified phosphorylation sites in AQP2 (S256),¹⁶ NKCC2 (T118 and S120),¹⁷ CHMP2B (S199),¹⁸ HSP90AB1 (S255),¹⁹ and SPP1 (S192, S197, and S207).²⁰ Phosphorylation of AQP2 at S256 was confirmed by immunoblotting human urinary exosomes samples with a phospho-specific antibody for this site (Figure 3).

Figure 3.

Detection of AQP2-S256 phosphorylation in urinary exosomes. IMCD, rat inner medullary collecting duct treated with dDAVP (V₂R-selective vasopressin analog) for 30 min; exo (10 μg), is human urinary exosomes, 10 μg; exo (72 μg), human urinary exosomes, 72 μg.

DISCUSSION

Large-Scale Proteomic Profiling of Human Urinary Exosomes

One of the objectives of this study was to expand the existing human urinary exosome database by using a higher sensitivity LC-MS/MS mass spectrometer and improved computational tools for matching spectra to proteins in the human proteome. The LTQ mass analyzer has an increased trapping efficiency, ion capacity, and ion ejection rate compared with the LCQ mass analyzer¹ used in our previous study. We identified the peptide sequences using the SEQUEST program and analyzed them using the target-decoy database search strategy and the InsPecT tool. The target-decoy database search strategy allows adjustment of SEQUEST search parameters to ensure a given false-discovery rate (FDR).¹³ The InsPecT tool uses de novo sequencing to generate tag filters, which are then used to search the database to “look for any peptide that matches the tag.”²¹ The data have been made available to the general public and can be downloaded from our laboratory's website (http://dir.nhlbi.nih.gov/papers/lkem/exosome/). In addition, the database can be searched using the BLAST algorithm.

As illustrated in Figure 1, analysis of human urinary exosomes by mass spectrometry and immunoblotting can provide information with regard to genetic diseases involving apical proteins as shown by the qualitative assessment of urinary exosome samples from patients with Bartter syndrome type I. The urinary exosome patient samples showed a complete absence of NKCC2 protein bands. Mutations in the SLC12A1 gene cause Bartter syndrome type I.⁹ Many such mutations presumably result in misfolding of the NKCC2 protein, preventing the apical delivery of the protein and therefore preventing incorporation into urinary exosomes.

Barriers to Clinical Use of Human Urinary Exosome Analysis

We previously reviewed the potential of urinary exosome analysis as a route to biomarker discovery in renal diseases and delineated barriers to success with the approach.²2122 One important barrier is the lack of standard protocols for collection, processing, and storage of urine samples to allow reproducible measurements to be made in any clinical laboratory. We have proposed a set of procedures that can serve as a beginning point in the development of such techniques (http://intramural.niddk.nih.gov/research/uroprot/). Our current approach includes an ultracentrifugation step, which requires expensive instrumentation and long processing times. Filtration methods have been proposed to replace the ultracentrifugation step.³ A particular knotty problem is removal of Tamm-Horsfall protein, an extraordinary abundant urinary protein that interferes with successful mass spectrometry and immunoblotting.²³ In the long run, the most important technical challenge may be to develop quantification approaches that allow detection of changes in excretion rates of particular biomarker candidates. Both labeling and nonlabeling methods have been developed to make protein mass spectrometry quantitative²⁴; however, the biggest barrier to quantification lies in development of adequate normalization techniques providing surrogates for timed collections of urine, which are notoriously inaccurate.²¹ Use of creatinine as a normalizing variable may be inadequate because of high subject-to-subject variability in its rate of excretion.²¹ Even without quantification, urinary exosome analysis can be valuable in situations such as genetic diseases (e.g., Bartter syndrome type I [Figure 1]), where a protein may be entirely absent from urinary exosomes.

Relevance to Renal Biology

Several of the proteins newly identified in urinary exosomes in this study may have considerable relevance to renal biology and the mechanism of renal disease. Our previous study¹ identified proteins that were characteristic of most of the cell types facing the urinary space from podocytes through transitional epithelial cells of the urinary drainage system. In this study, we identified markers of two additional cell types, type A and B intercalated cells. Specifically, the B1 subunit of the H^{-ATPase is apically located in type A intercalated cells,25} and the anion transporter pendrin is present in type B intercalated cells.²⁶ Previously, we showed that urinary exosomes are derived from the apical endosomal pathway so that, although the B1 subunit of the H^{-ATPase is also expressed in type B intercalated cells, its basolateral location probably precludes delivery to urinary exosomes. Overall, we identified 17 different vacuolar H-ATPase subunits in urinary exosomes, 78% of the whole V}₀–V₁ complex.⁸

We identified all of the subunits of the four ESCRT complexes (ESCRT-0 through ESCRT-III) in urinary exosomes in this study. The ESCRT complexes play a central role in the formation of MVB and the secretion of exosomes.⁵

Four different orphan G-protein–coupled receptors were identified in urinary exosomes in this study, namely GPR98, GPRC5A, GPRC5B, and GPRC5C. These receptors are presumably apically located in one or more renal tubule cells. GPR98 (also known as very large G-protein–coupled receptor 1 or Neurepin) has more than 6000 amino acids. The three GPRC5 proteins are members of the metabotropic glutamate family, but their natural ligands are unknown. It will be of interest in future studies to discover the role of these proteins in renal development and regulation.

Phosphoproteomic Analysis of Human Urinary Exosomes

Posttranslational modifications (PTM) of proteins play an important role in protein function. Among the most important PTM is phosphorylation. Protein phosphorylation regulates cellular signaling processes and may determine protein structure, function, and subcellular localization.¹² The ability to detect PTM, such as phosphorylation, in urinary exosomes may provide an additional level of information that could aid in diagnosis and treatment of a variety of renal disorders. Furthermore, discovery of PTM in urinary exosomes can provide clues about physiologic and pathophysiologic mechanism. In this study, we identified 14 phosphoproteins. The specific phosphorylation sites identified included six that were previously identified and eight that had not been previously identified. Among the novel sites was serine-811 in the NCC protein. This amino acid is conserved in humans, chimpanzees, rhesus monkeys, and horses but not in mice and rats. The amino acid sequence surrounding this site is absent in rodents owing to a difference in exon splicing.¹⁵ Finally, AQP2 phosphorylated at serine-256 was readily detectable in urinary exosomes. Because this phosphorylation event is increased by vasopressin-stimulated activation of adenylyl cyclase,¹⁶ measurements of the amount of serine-256–phosphorylated AQP2 in urine may provide an improved means of assessing the state of vasopressin activation using phospho-specific antibodies.

CONCISE METHODS

Urinary Exosome Isolation

Urine was collected from eight healthy humans: Four men (aged 22 to 33) and four women (aged 24 to 35) (National Institute of Diabetes and Digestive and Kidney Diseases Clinical Research Protocol 00-DK-0107). Fifty milliliters per subject was collected and mixed together. The urinary exosome isolation procedure is shown in Figure 4. Protease inhibitors were added (1.67 ml of 100 mM NaN₃, 2.5 ml of 11.5 mM 4-[2-aminoethyl] benzenesulfonyl fluoride, and 50 μl of 1 mM leupeptin). The mixed sample was centrifuged at 17,000 × g for 10 min at 4°C. The 17,000 × g supernatant was ultracentrifuged at 200,000 × g for 1 h at 25°C. The ultracentrifugation step was repeated 3 additional times, adding new 17,000 × g supernatant volume each time to each of the 12 tubes. Each of the 12 pellets was suspended with 50 μl of “isolation solution” (10 mM triethanolamine and 250 mM sucrose). The suspensions were pooled together.

Figure 4.

Differential centrifugation procedure for the isolation of urinary exosomes from urine.

The abundant urinary protein uromodulin or Tamm-Horsfall protein forms very high molecular weight complexes through disulfide linkages. These complexes sediment in the 200,000 × g spin unless denatured. To denature the zona pellucida domains in the Tamm-Horsfall protein, we mixed the resuspended pellet with 200 mg/ml dithiothreitol (DTT) at 95°C for 2 min. The resuspended pellet was added to an ultracentrifuge tube, and isolation solution was added to increase the volume to 8 ml. The sample was centrifuged at 200,000 × g for 1 h at 25°C. The pellet was suspended in 50 μl of isolation solution and frozen at −80°C.

InsPecT

We performed an additional analysis of the tandem mass spectrometry data using the InsPecT tool.²⁸ InsPecT uses de novo sequencing to generate sequence information (tag filters) from the experimental data. The tag filters are used to search the human protein database, nonredundant NCBI Reference Sequences (RefSeq) human protein database released on January 26, 2007, and identify peptide sequences that match with the experimental data. The size of the tag filters are three peptides in length on average. As shown in Figure 5, the tag filter generated for the protein CHMP1A matches the experimental data accurately. The peptide sequences identified using the tag filters are then scored to estimate that the top match is correct.²⁸ The score procedure computes the P value for each peptide sequence by “comparing the match quality score to the distribution of quality scores for incorrect matches.” For these data, we accept only peptide matches with P ≤ 0.05.

Figure 5.

Spectrum generated by InsPecT for CHMP1A protein (NP_002759). The peptide sequence is RVYAENAIRK. The tag region for the b ions and the y ions are shown by the black solid lines.

Elimination of Ambiguous Protein Identifications

Once proteins were identified using the approaches described, we needed to determine whether all identifications corresponded to unique gene products. An “ambiguous identification” is defined as an identification for which the peptide sequence that is used to determine the protein identity is found in multiple proteins that are not splice variants of the same gene (Figure 6).

Figure 6.

Criteria to disambiguate data set. (A) An unambiguous identification when a peptide sequence was a 100% match without gaps to one and only one protein. (B) An unambiguous identification when a peptide sequence was a 100% match without gaps to more than one protein, but these proteins are splice-variant products of one unique gene. (C) An ambiguous identification when a peptide sequence was a 100% match without gaps to more than one protein deriving from more than one gene, and the identification was based only on that single peptide.

To disambiguate the data set, we generated software that automates the comparison of each peptide sequence to the protein sequences in the RefSeq Human Protein Database using the BLAST algorithm. An identification was considered unambiguous when the sequence was a 100% match without gaps to one and only one protein (Figure 6A). An identification was also considered unambiguous when the sequence was a 100% match without gaps to more than one protein but these proteins are splice-variant products of one unique gene (Figure 6B). An identification was considered ambiguous when a peptide sequence was a 100% match without gaps to more than one protein deriving from more than one gene and the identification was based only on that single peptide (Figure 6C). The proteins identified from at least one unambiguous peptide were considered unambiguous proteins. The proteins that contained only ambiguous peptides were considered ambiguous proteins.

Urinary Exosome Isolation

Urine was collected from eight healthy humans: Four men (aged 22 to 33) and four women (aged 24 to 35) (National Institute of Diabetes and Digestive and Kidney Diseases Clinical Research Protocol 00-DK-0107). Fifty milliliters per subject was collected and mixed together. The urinary exosome isolation procedure is shown in Figure 4. Protease inhibitors were added (1.67 ml of 100 mM NaN₃, 2.5 ml of 11.5 mM 4-[2-aminoethyl] benzenesulfonyl fluoride, and 50 μl of 1 mM leupeptin). The mixed sample was centrifuged at 17,000 × g for 10 min at 4°C. The 17,000 × g supernatant was ultracentrifuged at 200,000 × g for 1 h at 25°C. The ultracentrifugation step was repeated 3 additional times, adding new 17,000 × g supernatant volume each time to each of the 12 tubes. Each of the 12 pellets was suspended with 50 μl of “isolation solution” (10 mM triethanolamine and 250 mM sucrose). The suspensions were pooled together.

Figure 4.

Differential centrifugation procedure for the isolation of urinary exosomes from urine.

The abundant urinary protein uromodulin or Tamm-Horsfall protein forms very high molecular weight complexes through disulfide linkages. These complexes sediment in the 200,000 × g spin unless denatured. To denature the zona pellucida domains in the Tamm-Horsfall protein, we mixed the resuspended pellet with 200 mg/ml dithiothreitol (DTT) at 95°C for 2 min. The resuspended pellet was added to an ultracentrifuge tube, and isolation solution was added to increase the volume to 8 ml. The sample was centrifuged at 200,000 × g for 1 h at 25°C. The pellet was suspended in 50 μl of isolation solution and frozen at −80°C.

In-Gel Trypsin Digestion

The protein concentration was determined using the Bradford Assay. This sample was solubilized in Laemmli sample buffer (1.5% SDS, 6% glycerol/10 mM Tris HCl, and 60 mg/ml DTT). Proteins in the exosome sample were separated by 1D SDS-PAGE using a Bio-Rad Ready Gel 4 to 15% polyacrylamide gradient gel with 125 μg distributed among two lanes. The gel was stained with Colloidal Coomassie Blue (GelCode Blue Stain Reagent; Pierce, Rockford IL) for 10 min and destained using ddH₂O (2 × 30 min). The gel was divided from top to bottom into 40 1-mm strips over the entire molecular weight range of the gel. Each strip was diced into small pieces (1 mm^{) and placed into labeled centrifuge tubes.}

The gels pieces were destained by adding 100 μl of 25 mM ammonium bicarbonate (NH₄HCO₃)/50% acetonitrile (ACN) for 10 min and were dried using a SpeedVac. The samples were reduced in a solution of 10 mM DTT and 25 mM NH₄HCO₃ at 56°C for 1 h. The samples were alkylated in a solution containing 55 mM iodoacetamide and 25 mM NH₄HCO₃ in the dark at room temperature for 45 min. The gel pieces were washed with 25 mM NH₄HCO₃ and dehydrated in a solution containing 25 mM NH₄HCO₃ and 50% ACN. The samples were dried using the SpeedVac. The samples were rehydrated in a solution containing 12.5 ng/μl trypsin (V5113; Promega, Madison, WI) in 25 mM NH₄HCO₃ and digested overnight at 37°C. Peptides were extracted using 50% ACN/0.1% formic acid (FA). The extracted samples were dried using the SpeedVac to remove ACN and then reconstituted with 0.1% FA. All 40 peptide samples were desalted using C₁₈ Zip Tips (Millipore, Billerica, MA) before analysis by mass spectrometry.

Nanospray LC-MS/MS

A high-sensitivity linear ion trap mass spectrometer, LTQ (Thermo Electron Corp.) equipped with a nanoelectrospray ion source was used to acquire m/z ratios in both precursor ions (MS1) and fragmented ions (MS2) scans. To reduce further the sample complexity before mass analysis, we injected the tryptic peptides extracted from each gel slice using an Agilent 1100 nanoflow system (Agilent Technologies, Palo Alto, CA) into a reversed-phase liquid chromatographic column (PicoFrit, Biobasic C₁₈; New Objective, Woodburn, MA). This LC-MS/MS method allows the acquisition of raw data files that are the MS/MS scans of the five highest intensity peaks after fragmentation with collision-induced dissociation in the LTQ mass analyzer.

Analysis of Data

The raw data files were searched against the NCBI Reference Sequences (RefSeq) human protein database by using BIOWORKS software (Thermo Finnigan). BIOWORKS utilizes SEQUEST, which is a program that “finds database candidate sequences whose theoretical spectra are compared with the experimental spectrum.”²⁷ To identify thoroughly peptide sequences, we searched the raw data files using the target-decoy approach and InsPecT.

In addition, we analyzed the data in a two-step process. The first step was to assess and minimize false-discovery peptide identifications using the target-decoy approach, manual inspection of spectra, and InsPecT. The second step was to assess and eliminate ambiguous protein identifications.

Target-Decoy

To apply the target-decoy database searching strategy,¹³ we used the NHLBI Proteomics Core Facility in-house software to create a composite database containing the forward and reverse sequences of the nonredundant NCBI Reference Sequences (RefSeq) human protein database released on January 26, 2007. We used the forward sequences as the target database and the reversed sequences as the decoy database. We searched the raw data files against this composite database. After the search, we assessed the FDR by the number of peptides matched from the reversed sequences. The parameters that determine the stringency of the filtering criteria include XCorr, Sp rank, and delta Cn. These parameters were incrementally adjusted, thereby reducing the false-discovery identifications until a target FDR was achieved. In our case, the data were filtered to a target of 2% FDR, and the actual FDR was 1.91%. The filter settings used were min Xcorr rank 1, min Sp rank 10, min delta Cn 0.08, charge + 1 min Xcorr 2.37, charge + 2 min Xcorr 2.87, and charge + 3 min Xcorr 3.37.

InsPecT

We performed an additional analysis of the tandem mass spectrometry data using the InsPecT tool.²⁸ InsPecT uses de novo sequencing to generate sequence information (tag filters) from the experimental data. The tag filters are used to search the human protein database, nonredundant NCBI Reference Sequences (RefSeq) human protein database released on January 26, 2007, and identify peptide sequences that match with the experimental data. The size of the tag filters are three peptides in length on average. As shown in Figure 5, the tag filter generated for the protein CHMP1A matches the experimental data accurately. The peptide sequences identified using the tag filters are then scored to estimate that the top match is correct.²⁸ The score procedure computes the P value for each peptide sequence by “comparing the match quality score to the distribution of quality scores for incorrect matches.” For these data, we accept only peptide matches with P ≤ 0.05.

Figure 5.

Spectrum generated by InsPecT for CHMP1A protein (NP_002759). The peptide sequence is RVYAENAIRK. The tag region for the b ions and the y ions are shown by the black solid lines.

Minimizing False-Discovery Peptide Identifications

In addition to the target-decoy approach the InsPecT analysis, we validated the quality of proteins identified by manually checking the spectra of those proteins with one unique peptide. We filtered out the proteins that did not have the expected molecular weight that matched to the corresponding regions in the 1-D SDS PAGE.

Elimination of Ambiguous Protein Identifications

Once proteins were identified using the approaches described, we needed to determine whether all identifications corresponded to unique gene products. An “ambiguous identification” is defined as an identification for which the peptide sequence that is used to determine the protein identity is found in multiple proteins that are not splice variants of the same gene (Figure 6).

Figure 6.

Criteria to disambiguate data set. (A) An unambiguous identification when a peptide sequence was a 100% match without gaps to one and only one protein. (B) An unambiguous identification when a peptide sequence was a 100% match without gaps to more than one protein, but these proteins are splice-variant products of one unique gene. (C) An ambiguous identification when a peptide sequence was a 100% match without gaps to more than one protein deriving from more than one gene, and the identification was based only on that single peptide.

To disambiguate the data set, we generated software that automates the comparison of each peptide sequence to the protein sequences in the RefSeq Human Protein Database using the BLAST algorithm. An identification was considered unambiguous when the sequence was a 100% match without gaps to one and only one protein (Figure 6A). An identification was also considered unambiguous when the sequence was a 100% match without gaps to more than one protein but these proteins are splice-variant products of one unique gene (Figure 6B). An identification was considered ambiguous when a peptide sequence was a 100% match without gaps to more than one protein deriving from more than one gene and the identification was based only on that single peptide (Figure 6C). The proteins identified from at least one unambiguous peptide were considered unambiguous proteins. The proteins that contained only ambiguous peptides were considered ambiguous proteins.

Patients with Bartter Syndrome Type I

We collected spot urine samples from two patients with clinically diagnosed Bartter syndrome type I. The patients were enrolled in the institutional review board–approved protocol 76-HG-0238. We obtained written informed consent from the parents and/or patient. We collected urine samples from healthy humans and used them as controls. We processed all samples using the differential centrifugation method to isolate human urinary exosomes described already. Each sample was prepared for immunoblotting by solubilizing in Laemmli buffer (1.5% SDS, 6% glycerol, 10 mM Tris HCl, and 60 mg/ml DTT). The samples, patient 1 and patient 2, and the control samples, control 1 and control 2, were loaded onto a 1-D SDS-PAGE gel on the basis of time as measured by creatinine excretion. The proteins were transferred to Immobilon-P (Millipore) membranes, blocked, and probed with antigen-specific NKCC2 and NCC primary antibodies. We incubated the blots with species-specific fluorescence secondary antibodies (Alexa 688) and visualized them using the Odyssey Infrared Imaging System (LiCor, Lincoln, NE).

Phosphopeptide Enrichment and LC-MS/MS Analysis

We collected urine specimens (200 ml) from six healthy humans, three men and three women. We processed the specimens 400 ml/d for 3 d and pooled them. The exosome isolation was as described previously except that phosphatase inhibitors 10 mM NaF (Sigma, St. Louis, MO), 20 mM β-glycerol phosphate (Fluka, St. Louis, MO), and 1 mM sodium orthovanadate (Sigma) were added. The pellet was resuspended in 6 M guanidine HCl/50 mM NH₄HCO₃.

The sample was concentrated using a Centricon tube at 13,500 × g, with a starting volume of 420 μl and a final volume of 55 μl. The sample was reduced with 50 mM DTT for 1 h at 56°C. The sample was alkylated by addition of 100 mM iodoacetamide for 1 h (dark) at room temperature and was digested with trypsin overnight at 37°C. The sample was centrifuged at 16,000 × g for 20 min. The supernatant was kept, and 100% FA was added to inactivate the trypsin. The sample was desalted on a 1-ml HLB column (Waters Oasis, Milford, MA) by positive displacement via a syringe with a luer adapter. The sample was eluted with two elution buffers. Elution buffer 1 contained 50% ACN and 0.1% FA, and elution buffer 2 contained 90% ACN and 0.1% FA. The eluents, 50 and 90%, were dried using the SpeedVac.

Phosphopeptides were enriched from the samples using the Pierce Phosphopeptide Isolation Kit (cat. no. 89853) according to the manufacturer's protocol. Phosphopeptide samples were desalted using C₁₈ ZipTips (Millipore) before analysis by mass spectrometry.

Phosphopeptide samples were analyzed on an Agilent 1100 nanoflow system (Agilent Technologies) LC connection to a Finnigan LTQ FT mass spectrometer (Thermo Electron) equipped with a nanoelectrospray ion source as described previously.¹¹ The five most intense ions were sequentially isolated and fragmented (MS2) in the linear ion trap using collision-induced dissociation. The data-dependent neutral loss algorithm in XCALIBUR software was used to trigger an MS3 scan when a neutral loss of 98.0, 49.0, or 32.7 Da was detected among the two most intense fragment ions in a given MS2 spectrum.

Analysis of Phosphopeptide Data Sets

We searched MS raw data files against a composite database containing the forward and reversed peptide sequence of the Human RefSeq Database from January 26, 2007. Putative phosphopeptides were selected and filtered to produce MS2 and MS3 data sets with target FPR of 2% (high stringency) and 20% (low stringency) via the PhosphoPIC program.⁴ This software was also used to merge MS2 and MS3 data sets into a single file to facilitate subsequent data analysis. Phosphopeptides identified in MS2 spectra were submitted for automated phosphorylation site assignment using the Ascore algorithm.¹³ A site with an Ascore ≥19 (>99% confidence) was considered to be unambiguously assigned. Phosphopeptides present only in MS3 spectra were checked manually. We used Scansite (http://scansite.mit.edu/motifscan_seq.phtml) to determine the phosphorylation motif for the identified sites. We searched the PhosphoSite database (http://www.phosphosite.org) to determine whether the sites were novel or previously identified.

Disclosures

None.