R E S EARCH ART I C L E

K IDNEY D I S EASE

http://stm.sciencem

ag.D

ownloaded from

Tissue transcriptome-driven identification of epidermalgrowth factor as a chronic kidney disease biomarkerWenjun Ju,1,2*† Viji Nair,1† Shahaan Smith,1‡ Li Zhu,3 Kerby Shedden,2,4 Peter X. K. Song,5

Laura H. Mariani,1,6 Felix H. Eichinger,1 Celine C. Berthier,1 Ann Randolph,1 Jennifer Yi-Chun Lai,1

Yan Zhou,5§ Jennifer J. Hawkins,1 Markus Bitzer,1 Matthew G. Sampson,7 Martina Thier,8

Corinne Solier,8 Gonzalo C. Duran-Pacheco,8 Guillemette Duchateau-Nguyen,8 Laurent Essioux,8

Brigitte Schott,8 Ivan Formentini,8 Maria C. Magnone,8¶ Maria Bobadilla,8 Clemens D. Cohen,9∥

Serena M. Bagnasco,10 Laura Barisoni,11 Jicheng Lv,3 Hong Zhang,3 Hai-Yan Wang,3**Frank C. Brosius,1,12 Crystal A. Gadegbeku,13 Matthias Kretzler,1,2* for the ERCB, C-PROBE,NEPTUNE, and PKU-IgAN Consortium

Chronic kidney disease (CKD) affects 8 to 16% people worldwide, with an increasing incidence and prevalence ofend-stage kidney disease (ESKD). The effective management of CKD is confounded by the inability to identify pa-tients at high risk of progression while in early stages of CKD. To address this challenge, a renal biopsy transcriptome-driven approach was applied to develop noninvasive prognostic biomarkers for CKD progression. Expression ofintrarenal transcripts was correlated with the baseline estimated glomerular filtration rate (eGFR) in 261 patients.Proteins encoded by eGFR-associated transcripts were tested in urine for association with renal tissue injury andbaseline eGFR. The ability to predict CKD progression, defined as the composite of ESKD or 40% reduction ofbaseline eGFR, was then determined in three independent CKD cohorts. A panel of intrarenal transcripts, includingepidermal growth factor (EGF), a tubule-specific protein critical for cell differentiation and regeneration, predictedeGFR. The amount of EGF protein in urine (uEGF) showed significant correlation (P < 0.001) with intrarenal EGFmRNA, interstitial fibrosis/tubular atrophy, eGFR, and rate of eGFR loss. Prediction of the composite renal end pointby age, gender, eGFR, and albuminuria was significantly (P < 0.001) improved by addition of uEGF, with an increase ofthe C-statistic from 0.75 to 0.87. Outcome predictions were replicated in two independent CKD cohorts. Our approachidentified uEGF as an independent risk predictor of CKDprogression. Addition of uEGF to standard clinical parametersimproved the prediction of disease events in diverse CKD populations with a wide spectrum of causes and stages.

byorg/

1Department of Internal Medicine, University of Michigan, Ann Arbor, MI 48109, USA.2Department of Computational Medicine and Bioinformatics, University of Michigan, AnnArbor, MI 48109, USA. 3Renal Division, Department of Internal Medicine, Peking UniversityFirst Hospital, PekingUniversity Institute of Nephrology, Beijing 100034, China. 4Departmentof Statistics, University of Michigan, Ann Arbor, MI 48109, USA. 5Department of Biostatistics,University of Michigan, Ann Arbor, MI 48109, USA. 6Arbor Research Collaborative for Health,Ann Arbor, MI 48104, USA. 7Department of Pediatrics, University of Michigan, Ann Arbor, MI48109, USA. 8Roche Pharmaceutical Research and Early Development–Roche InnovationCenter, 4070 Basel, Switzerland. 9Division of Nephrology, Institute of Physiology, Universityof Zurich, CH-8006 Zürich, Switzerland. 10Department of Pathology, Johns Hopkins Schoolof Medicine, Baltimore, MD 21287, USA. 11Department of Pathology, Miller School ofMedicine, University of Miami, Miami, FL 33136, USA. 12Department of Molecular andIntegrative Physiology, University of Michigan, Ann Arbor, MI 48109, USA. 13Temple ClinicalResearch Institute, Temple University School of Medicine, Philadelphia, PA 19140, USA.*Corresponding author. E-mail: wenjunj@med.umich.edu (W.J.); kretzler@umich.edu(M.K.)†These authors contributed equally to this work.‡Present address: Department of Medical Oncology, University of Washington, Seattle,WA 98109, USA.§Present address: Merck Research Laboratories, North Wales, PA 19454, USA.¶Present address: Cardiovascular and Metabolic Diseases Innovative Medicines, AstraZenecaR&D, Mölndal SE-431 83, Sweden.∥Present address: Department for Kidney Disease and Hypertension and ClinicalImmunology, Harlaching Hospital, Munich 81545, Germany.**Deceased.

guest on July 7, 2020

INTRODUCTION

Chronic kidney disease (CKD) presents a major public health chal-lenge and is an important cause of global mortality (1). Eight to 16%of the population worldwide is affected by CKD (2–4) with increasedrisk for end-stage kidney disease (ESKD), cardiovascular disease(CVD), and death (5). A limited set of therapeutic approaches have beenshown to slow progression of CKD and reduce its associated CVD risk(2). Themain challenges to improve outcomes in patients withCKDarethe current inability to identify patients at high risk of loss of renalfunction in early CKD and then to provide the patients targeted, effec-tive treatments. Identifying pathobiologically important biomarkersthat can address both of these challenges would have a major impacton outcomes of patients with progressive CKD.

Multiple lines of evidence have emerged over the last four decades,supporting the concept of a common pathway activated in CKD pro-gression (6–8). This concept suggests that once kidney damage reachesa threshold, the subsequent progression is largely independent of theinitial kidney injury and uses a shared molecular mechanism for dis-ease progression (8).Combining informationongenetic risk and intrarenalgene regulation allows identification of the functional underpinningsof CKD progression, not only in model systems but also in humandisease, and strongly supports the notion of a common progressionpathway (9).

Here, we define the sharedmolecular mechanism of CKD, using fourcohorts from across three continents (Fig. 1). Our approach is rooted in

www.ScienceT

the hypothesis that both clinical phenotypes, such as baseline estimatedglomerular filtration rate (eGFR) and proteinuria, and renal tissue al-terations, such as tubular atrophy (TA), interstitial fibrosis (IF), and glo-merulosclerosis, seen inCKD are associatedwith, and a consequence of,

ranslationalMedicine.org 2 December 2015 Vol 7 Issue 316 316ra193 1

R E S EARCH ART I C L E

the dynamic molecular mechanisms re-flected in the transcriptional programsof the diseased kidney tissues. By im-plementing this strategy, we identifiedepidermal growth factor (EGF) as a bio-marker, with intrarenal mRNA and uri-nary EGF protein concentrations tightlycorrelated with eGFR at the time of bi-opsy and with the change in renal func-tion over time, independent of eGFRand proteinuria.

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from

RESULTS

Identification and validation ofintrarenal candidate markers forprediction of kidney functionAs the first step in the sequential strategyfor CKD predictor identification (Fig. 1),we quantified intrarenal transcripts andevaluated their association with clinicalphenotypes. Affymetrix GeneChip–basedsteady state transcript expression was de-rived from the tubulointerstitial compart-ment of kidney biopsies of 164 patientsfrom the European Renal cDNA Bank(ERCB) (Table 1) andcorrelatedwith eGFRat the time of biopsy. A total of 72 can-didate genes passed the selection criteriaof significant correlation [false discoveryrate (FDR) < 0.01] with eGFR and differ-ential expression compared to living do-nors. These transcripts are enriched inthe key pathogenic pathways that havebeen implicated in driving CKD progres-sion (table S1). The expression of thesetranscripts was measured in 55 CKD pa-tients from the ERCB (Table 1) by quanti-tative real-time polymerase chain reaction(qRT-PCR) (table S2) andassessed for theirability to predict patients’ eGFR at timeof biopsy (Fig. 2A). A panel of six genes[nicotinamideN-methyltransferase(NNMT),EGF, thymosin b 10 (TMSB10), tissue in-hibitor of metalloproteinases (TIMP1),

tubulin a 1a (TUBA1A), and annexin A1 (ANXA1)] showed the bestpredictive performance (correlation between predicted and observedeGFR values, r = 0.77, P < 0.0001; Fig. 2B). Candidate transcripts werethen assessed in a second, independent North American cohort, theClinical Phenotyping Resource and BiobankCore (C-PROBE), consistingof 42 patients with CKD (Table 1), with the EGF, NNMT, and TSMB10panel showing the best eGFR prediction (r = 0.62, P < 0.0001; fig. S1).

Intrarenal transcript-based selection of urinary biomarkersNext, noninvasive biomarker candidates were selected from the intra-renal transcript predictor panel. Markers for noninvasive evaluationwere prioritized on the basis of (i) correlation with eGFR in multiple

www.ScienceT

cohorts (table S3), (ii) kidney-specific transcript and protein expression,and (iii) compelling biology for amechanistic role in CKD. These selec-tion criteria converged on EGF as the top candidate. First, intrarenalEGFmRNAexpression consistently showed significant correlationwitheGFR (FDR≤ 0.05; r = 0.66, 0.81, and 0.42 in the discovery and the twovalidation cohorts, respectively) and contributed to the best performinggene expression predictor panel in both ERCB and C-PROBE (Fig. 2C).Second, the EGF transcript showed highly kidney-specific expressioncompared to other candidates in the human tissue panel (Fig. 2D andfig. S2). EGFmRNAwas enriched in renal cortex andmedulla comparedto glomeruli, papillary tips, and renal pelvis (Fig. 2E) (10). In situ hybrid-ization showed EGF mRNA expression in tubular cross sections in

Biopsy

Microdissectionand RNA extraction

Routine diagnosticClinical and histological data

Molecular phenotyping

Molecular to clinical phenotype correlation

Step 1

Step 2

Treatment targets

Identification of relevant signaling pathways

Renal tissue–based analysis

Validated intrarenal markers (mRNAs)

Urine

Noninvasive biomarkers

Intrarenal mRNA–urine protein correlation

Body fluid–based analysis

Step 3

Discovery cohort (ERCB): Intrarenal candidate markers (mRNAs)

Two independent validation cohorts

Clinic Histologyxx

xxx

ERCB + C-PROBE C-PROBE + NEPTUNE + PKU-IgAN

Three independent cohorts

Association with renal outcome (eGFR slope/composite end points)

Fig. 1. Schematic overviewof the tissue transcriptome-driven strategy to identify urinary biomarkersfor CKD progression.

ranslationalMedicine.org 2 December 2015 Vol 7 Issue 316 316ra193 2

R E S EARCH ART I C L E

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from

medulla and cortex (Fig. 2F, I and II), consistent with published immu-nohistochemical data demonstrating that EGF expression is restrictedto the thick ascending limb of Henle and distal tubules (11). Third, insitu hybridization showed diminished EGF transcript expression in tu-bular cells of the cortex andmedulla in CKD (Fig. 2F, III and IV versus Iand II). Finally, EGF is a knownmediator of differentiated epithelial cellfunction and regeneration. Both in vivo and in vitro studies have shown

www.ScienceT

that EGF enhances renal tubule cell regeneration and repair and ac-celerates the recovery of renal function after injury (12–14) throughits pro-proliferative and antiapoptotic action.

To test whether urinary EGF protein reflected intrarenal EGFtranscript expression, we measured EGF protein concentrations andnormalized them to urine creatinine concentration (uEGF/Cr) in pa-tients with paired urine and kidney biopsy samples. uEGF/Crcorrelated significantly with the intrarenal EGF RNA expression in34 C-PROBE (r = 0.70, P < 0.0001; Fig. 3A) and 85 Nephrotic Syn-drome Study Network (NEPTUNE) patients (r = 0.69, P < 0.0001;Fig. 3B; for cohort details, see table S4).

Next, we used urine samples of 349 C-PROBE patients (Table 2) toevaluate the association between uEGF/Cr and eGFR, independent ofthe 34 C-PROBE patients described above. uEGF/Cr correlated signif-icantly with eGFR (r = 0.77, P < 0.0001) at the time of sample procure-ment (Fig. 3C). The uEGF/Cr-eGFR correlation was significant in C-PROBE subgroups of patients with DN (n = 70) or with diabetes (n =135) (fig. S3). The correlation was further replicated in 141 NEPTUNEpatients with biopsy-proven FSGS,MCD, andMGN (Table 2) (r= 0.82,P < 0.0001, Fig. 3D) and in 452 patients from a prospective cohort studyenrolling patients with biopsy-proven primary immunoglobulin A(IgA) nephropathy at Peking University First Hospital, China (PKU-IgAN) (Table 2) (r = 0.78, P < 0.0001; Fig. 3E). In all three cohorts,uEGF/Cr showed only a weak negative correlation (r = −0.14, −0.17,and −0.31 in C-PROBE, NEPTUNE, and PKU-IgAN, respectively) withurinary albumin-to-creatinine ratio (ACR, fig. S4), suggesting that uEGF/Cr could represent a different pathophysiologic mechanism than that de-tected by albuminuria.

Association of EGF with renal biopsy outcome predictorsThe correlation of uEGF/Cr with histological scores of IF and TA wasmeasured to assess the association of this biomarker with tubuloin-terstitial lesions (15). IF/TA scores have been reported to be associatedwith cross-sectional eGFR (16) and can predict long-term kidneyfunction (17). IF and TA scores were obtained from 102 NEPTUNEpatients, and a significant inverse correlation was observed betweenuEGF/Cr and IF and TA [r = −0.75 and −0.74, respectively (P < 0.001);Fig. 4]. The correlations remained significant after adjusting for eGFRand ACR (r = −0.33 for both, P < 0.001).

EGF and CKD progressionTo determine whether EGF can predict CKD progression, we assessed theassociation between EGF and loss of renal function using eGFR slope as acontinuous variable. Both intrarenal gene expression data and three ormore eGFR values (over 4.0 ± 1.8 years) were available for 29 C-PROBEpatients, allowing correlation of renal transcript amounts with eGFRslope. Four hundred thirty-one transcripts were found to be correlatedwith eGFR slope (FDR < 0.05). A functional network analysis predictedEGF to be the top upstream regulator of 37 eGFR slope-correlatedtranscripts (P = 2.09 × 10−12; table S5 and fig. S5).

Both intrarenal EGFmRNA and urinary EGF protein/Cr showed asignificant correlation with eGFR slope in the C-PROBE cohort [r =0.58, P < 0.001, for EGF tissue mRNA (Fig. 5A) and r = 0.64, P <0.001, for uEGF/Cr (fig. S6)]. We then used uEGF/Cr to predict eGFRslope via a multivariable regression model. Adjusted for age and gen-der, the eGFR slope predicted by uEGF/Cr showed a higher correlationwith observed eGFR slope (r = 0.65, P < 0.001; Fig. 5B), compared to amodel using ACR instead of uEGF/Cr (r = 0.29, P < 0.001; Fig. 5C).

Table 1. Demographic characteristics of CKD patients in the gene ex-pression discovery and validation cohorts. Age and eGFR are presentedas means ± SD. SLE, lupus nephropathy; RPGN, rapidly progressiveglomerulonephritis; HTN, hypertensive nephropathy; MGN, membranousglomerulonephritis; DN, diabetic nephropathy; FSGS, focal segmental glo-merulosclerosis; MCD, minimal change disease; TMD, thin basement mem-brane disease.

Diseasetype

CKDpatients

eGFR(ml/min

per 1.73 m2)

Age(years)

Gender(% female)

Discovery cohort: ERCB

SLE

30 63.7 ± 29.4 34.7 ± 13.3 76.7

IgAN

24 75.9 ± 37.9 36.4 ± 14.6 25

RPGN

21 46.6 ± 31.5 58.5 ± 14.1 42.9

HTN

20 43.9 ± 25.1 57.2 ± 12.1 25

MGN

18 88.9 ± 41.4 53.4 ± 19.3 44.4

DN

17 44.3 ± 24.9 58.3 ± 10.7 29.4

FSGS

16 73.4 ± 38.4 46.2 ± 17.6 56.2

MCD

12 100.7 ± 33.9 35.8 ± 16.8 33.3

TMD

6 93.4 ± 29.4 46.0 ± 14.5 33.3

Total

164 66.4 ± 37.2 46.9 ± 17.6 43.3

Validation cohort 1: ERCB

SLE

10 60.1 ± 31.5 37.1 ± 13.8 60

IgAN

17 50.8 ± 34.1 46.0 ± 17.5 29.4

RPGN

5 20.9 ± 15.0 50.4 ± 21.7 20

MGN

4 54.4 ± 32.4 58.7 ± 23.6 50

DN

1 97.3 44.8 100

FSGS

8 55.9 ± 36.0 43.6 ± 18.6 25

MCD

4 86.0 ± 44.5 44.4 ± 19.3 0

Other

6 68.4 ± 60.7 45.2 ± 16.7 33.3

Total

55 55.7 ± 38.3 45.1 ± 17.6 32.7

Validation cohort 2: C-PROBE

SLE

16 106.4 ± 60.1 31.3 ± 9.9 87.5

IgAN

1 110.4 23 100

HTN

1 146.5 29 0

MGN

2 51.9 ± 21.9 53.0 ± 5.7 50

DN

1 50.0 80.8 100

FSGS

11 51.2 ± 21.0 41.9 ± 14.0 27.2

MCD

1 124.7 22 100

Other

9 41.3 ± 23.1 52.4 ± 18.9 55.6

Total

42 76.3 ± 50.8 39.6 ± 15.7 61.9

ranslationalMedicine.org 2 December 2015 Vol 7 Issue 316 316ra193 3

R E S EARCH ART I C L E

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from

Pre

dict

ed e

GF

R (

log 2

)

Number of markers

R

2) devr esbo . sv det ci der p(

10 122

4

6

8

10

r = 0.66, P < 0.0001

2

4

6

8

10

r = 0.42, P = 0.006

0

2

4

6

8

10

r = 0.81, P < 0.0001

Inte

nsity

val

ues

–2

0

2

4

6

Kidney

Liver

Heart

Retina Skin

Prosta

teLu

ng

Colon

Pancr

eas

Testis

Who

le br

ainUter

us

Ovary

050

100

150

Gene symbolNNMTEGFTMSB10TIMP1TUBA1AANXA1

0.8

0.6

0.7

0.5

0.4

0.3

0.1

0.2

00 5 10 15 20 25 30

A B

2 3 4

r = 0.77, P < 0.0001

C

eGF

R (

log 2

)

Observed eGFR (log2)

5 6 7

ERCB tissue EGF mRNA (log2) in arraytraining cohort (n = 164)

ERCB tissue EGF mRNA in TLDAvalidation cohort (n = 55)

C-PROBE tissue EGF mRNA (log2) in arrayvalidation cohort (n = 42)

2 4 6 8 10 122 4 6 8

D E In

tens

ity v

alue

med

ian-

cent

ered

val

ue

Glomer

uli

Inner

rena

l cor

tex

Outer r

enal

corte

x

Outer r

enal

med

ulla

Inner

rena

l med

ulla

Papilla

ry tip

s

Renal

pelvi

s

F

7

6

5

4

Who

le blo

od

–20 –15 –10 –5 0

II

III IV

I

50 µm

50 µm50 µm

50 µm

-

.-

.l

--

/

-

-r,

,)

Fig. 2. Transcript-eGFR association andEGF expression in the kidney in patients

with CKD. (A) Correlation between observedand predicted eGFR (continuous line) on thebasis of intrarenal transcripts integrated byridge regression analysis (top 1 to 30 transcripts shown), providing eGFR prediction inthe validation cohort (n = 55). Gray arearepresents 95% confidence interval (CI)Black arrow indicates the highest correlation provided by a six-marker set, which isdepicted in (B). (B) Maximal correlation (r =0.77, P < 0.0001) was demonstrated betweenobserved eGFR and that predicted by geneexpression values of six transcripts (n = 55)Dotted line represents 95% CI. (C) IntrarenaEGF mRNA showed a significant correlationwith patients’ eGFR in discovery (n = 164) andvalidation cohorts (n = 55 and 42). TLDA, TaqMan Low Density Array. (D) EGF mRNA expression in major human organs/tissues (aselection out of a panel of 84 human organstissues and cell lines), derived from BioGPS(http://biogps.org), indicating a highly kidney-specific expression pattern. For full dataset, see fig. S2. (E) EGFmRNA expression pattern in adult kidney (glomeruli, inner and outecortex, inner and outer medulla, papillary tipsand renal pelvis); data were extracted fromHiggins et al. (10). (F) In situ hybridizationdemonstrates tubule-specific EGF mRNA inboth cortex (I and III) and medulla (II and IV)with reduced expression in CKD (III and IVcompared to healthy controls (I and II). Blackarrows indicate positive staining in pink. Scalebars, 50 mM.

www.ScienceTranslationalMedicine.org 2 December 2015 Vol 7 Issue 316 316ra193 4

R E S EARCH ART I C L E

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from

The association of baseline uEGF with renal outcome (composite endpoint of ESKD or 40% reduction from baseline eGFR) was evaluated byclassification power [comparing the area under receiver operating character-istic (ROC) curves of the corresponding models] and survival analysis(time-to-event analysis). ROC analysis showed an area under the curve(AUC) for the base model (including eGFR and ACR, adjusted for ageand gender) of 0.84 (95% CI, 0.76 to 0.91) in C-PROBE. Addition of uEGFto the base model increased AUC to 0.90 (95% CI, 0.84 to 0.95) andresulted in a statistically improved model as evaluated by the likelihood test(P = 0.001). The classification value of uEGF added to the base modelwas replicated in the NEPTUNE and PKU-IgAN cohorts (fig. S7).

In survival analysis, addition of uEGF/Cr [model 2 (M2)] to the baseprediction model M1 (including eGFR and ACR, adjusted for age andgender) demonstrated a significant (P < 0.0001) covariate effect on the

www.ScienceTranslationalMedicine.org 2 Dec

likelihood ratio test, as well as an im-provement in model fit, as R2 increasedfrom 0.15 (M1) to 0.22 (M2), and theC-statistic increased from 0.75 (M1) to0.87 (M2) (Table 3). Thus, addition ofuEGF/Cr substantially improved theability to predict renal outcomes whencompared to ACR and baseline eGFRin combination. The improved predictionability of the composite model M2 (M1plus uEGF/Cr) was replicated in bothNEPTUNE and PKU-IgAN cohorts (Ta-ble 3). Multivariable-adjusted associa-tions of uEGF/Cr with the hazard ofprogression to composite end point areshown in Fig. 6. Higher uEGF/Cr was asso-ciated with decreased risk of ESKD or 40%reduction of baseline eGFR in threeindependent cohorts: the estimated haz-ard ratios (HRs, 95% CI) were 0.27 (0.13to 0.54), 0.29 (0.15 to 0.58), and 0.53 (0.37to 0.69) for C-PROBE, NEPTUNE, andPKU-IgAN, respectively, adjusted for age,gender, eGFR, and ACR. Thus, with oneunit increase of uEGF/Cr (in log2 scale)and the other covariates withheld, thehazard of disease progression would de-crease by 73, 71, and 47%, respectively,in these three cohorts. This is equivalentto a one-unit decrease of uEGF/Cr (log2)being associated with an increased risk ofCKD progression of 3.73 (1.85 to 7.69)–fold, 3.43 (1.72 to 6.67)–fold, and 1.96(1.45 to 2.70)–fold in these three cohorts,respectively.

DISCUSSION

By applying a renal biopsy transcriptome-driven sequential marker discovery ap-proach, we demonstrated that (i) a set ofintrarenal transcripts correlates with kid-ney function across a broad range of renal

diseases; (ii) uEGF concentration showed a strong correlation with tissueEGF transcript expression andmaintained the ability to predict eGFR; (iii)uEGF concentration was associated with eGFR slope; (iv) uEGF addedpredictive power to traditional clinical prognostic markers of CKD pro-gression end points across cohorts.

The GFR-correlated transcripts are enriched in the key pathogenicpathways that have been implicated in driving CKD progression, rangingfromchronic inflammation and extracellularmatrixmodulation to tubularcell differentiation (table S1). Because these pathways are involved inCKDprogression independent of etiology, prognostic biomarkers that bestrepresent them should be applicable to a wide spectrum of causes of CKD.

Estimates of GFR are the best overall indices of kidney function in healthand disease, according to the National Kidney Foundation KidneyDisease Outcomes Quality Initiative (NKF KDOQI) Clinical Practice

0

5

10

15

r = 0.70, P < 0. 0001 r = 0. 69, P < 0.0001

–2 0 6

2

4

6

8r = 0. 77, P < 0.0001 r = 0. 82, < 0.0001

uEGF/Cr protein (log2) C-PROBE uEGF/Cr protein (log2) NEPTUNE

uEGF/Cr (log2) C-PROBE uEGF/Cr (log2) NEPTUNE

eGF

R (

log

2)

uEGF/Cr (log2) PKU-IgAN

eGF

R (

log 2)

–2 0 2

4

6

8

10

P

5

10

15

–2 8

2

4

6

8

10

r = 0.78, P < 0.0001

A

2 4 6 8 0 2 4 6–2

Tiss

ue E

GF

mR

NA

(lo

g 2)

C

4 6 82 4

E

0 2 4 6

10

B

D

Fig. 3. Correlation of uEGF/Cr with intrarenal EGFmRNA and eGFR. (A and B) uEGF/Cr is correlated withintrarenal EGF mRNA in patients with matching urineenzyme-linked immunosorbent assay (ELISA) data andtissue mRNA expression data in C-PROBE (A) (n = 34)and NEPTUNE (B) (n = 85) cohorts. uEGF/Cr is correlatedsignificantly (P < 0.0001) with eGFR at the time of bi-opsy in patients from C-PROBE [(C) n = 349], NEPTUNE[(D) n = 141], and PKU-IgAN [(E) n = 452].

ember 2015 Vol 7 Issue 316 316ra193 5

R E S EARCH ART I C L E

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from

Guidelines (18). Baseline eGFR and proteinuria are established predictorsof CKD progression and capture primary elements of glomerular function,with eGFR reflecting glomerular hemodynamics and ultrafiltration capacityand proteinuria reflecting changes in glomerular permselectivity andits impact on CKD progression. Multiple potential biomarkers for renalinflammation and fibrosis, a dominant signature in the CKD pathwayanalysis presented here, have been shown to be associated with kidneyimpairment. These include tumor necrosis factor receptors 1 and 2 (TNFR1and TNFR2) (19, 20), monocyte chemotactic protein 1 (MCP1) (21), andmatrix metalloproteinases (22). Because these molecules are foundin plasma as well, their urinary concentration is a function of bothglomerular filtration and intrarenal production in proteinuric disease,potentially affecting their utility as urinary biomarkers of tubulointerstitialdisease. Biomarkers of tubular function and reserve are not availablefor clinical use in CKD. As described in this study, uEGF has the potentialto serve this role, given its highly restricted intrarenal expression andstrong correlation between renal transcripts and urinary protein concentration.

EGF function provides an independent rationale for its evaluationas a CKD progression biomarker. EGF has been reported to be re-

www.ScienceTranslationalMedicine.org 2 De

sponsible for modulation of tissue responseto injury in the kidney after tubulointerstitialdamage (23, 24). Exogenous EGF enhancedthe tubularcell repairprocessandacceleratedtubular regeneration and recovery of renalfunction in an animal model of acute renalinjury (13), suggesting that uEGF concentra-tions are not only associatedwith concurrentrenal function but also regulate signalingpathways of tubular recovery after injury.This offers the intriguing possibility thatuEGF might serve as a surrogate markerfor regenerative functional reserveof the renaltubules, reflecting their ability to respondto future acute or chronic injury, whichare key factors in CKD progression. Thisconcept is further supportedby theunbiasedfunctional analysis of gene expression datasets, which predicted that EGF was as anupstreamregulator of eGFR slope-associatedgenes. uEGF inversely correlatedwith IF andTA, established independent morphometricpredictors of renal outcomes (25–27).These associations not only provide strongsupport for our hypothesis but also shed

light on the pathogenic role of EGF in CKD progression.Reduced concentrations of EGF protein in the urine have been pre-

viously observed in DN, IgAN, adult polycystic kidney disease, and chil-dren with chronic renal failure (28–31). For example, Torres et al.showed that in a single center cohort of 132 IgAN patients, the ratioof EGF to MCP1 in the urine was negatively associated with thecomposite outcome of ESKD and/or doubling of the baseline serum cre-atinine (23), but it was not compared to conventional predictors of ESKDin predicting renal survival.

A critical factor for the uEGF-eGFR correlation is the high degreeof tissue specificity, making uEGF robust to extrarenal events that mayaffect the accuracy of other nonspecific biomarkers. EGF is absent oronly minimally detectable in plasma (32). Urinary EGF is reported tobe derived from the ascending portion of Henle’s loop and the distalconvoluted tubule (33, 34), and in situ hybridization confirmed localsynthesis of EGF in tubular compartments. The tissue specificity mightbe responsible for the robustness of the correlation of uEGF with renalfunction and renal functional decline across a wide spectrumof systemicand primary renal diseases in diverse environments.

Table 2. Baseline characteristics of patients whose urine samples were analyzed in urine studies.

Baseline characteristics

C-PROBE (n = 349) NEPTUNE (n = 141)

cember 2015 Vol 7 Is

PKU-IgAN (n = 452)

Age

54.8 ±16.1 45.7 ± 17.3 36.0 ± 12.0

Gender (% female)

59.3 35.5 51.3

Race (% black)

47.9 23.4 0

eGFR (ml/min per 1.73 m2)

55.4 ± 32.2 67.7 ± 37.1 84.8 ± 34.1

ACR (mg/mg)

0.8 ± 1.2 2.3 ± 2.1 0.6 ± 0.6

Log2 uEGF/Cr (ng/mg)

2.5 ± 1.1 3.0 ± 1.3 3.5 ± 1.0

Inte

rstit

ial

fibro

sis

–20

20

40

60

80

100

ρ = –0.75, P < 0.001

Tub

ular

atr

ophy

–20

20

40

60

80

100

ρ = –0.74, P < 0.001

uEGF/Cr (log2)

A

20 4 6–220 4–2 6

uEGF/Cr (log2)

B

Fig. 4. Association of uEGF/Cr with tubulointerstitial damage. (A and B) Tubulointerstitial damage isreflected by the percentage of cortex affected by IF/TA. Scoring of IF/TA was based on evaluation of silver,

periodic acid–Schiff (PAS), and trichrome-stained kidney sections (n = 102) by six readers who were blindedto the uEGF/Cr value. For each section, the average scores of the six readers for IF and TAwere calculatedas indicators of tubulointerstitial damage. uEGF/Cr is significantly correlated with IF (A) and TA (B) scores(Spearman correlation, P < 0.001).

sue 316 316ra193 6

R E S EARCH ART I C L E

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from

Our study has several limitations. The discovery approach used inthis study critically depends on renal biopsy tissue, raising concerns aboutthe ability to extrapolate the renal tissue findings to CKD patients notundergoing biopsy. However, uEGF correlated as tightly with eGFRacross all C-PROBE patients as it did in the biopsy patients from thesame cohort (r=0.77 for both,P< 0.001), supporting the utility of uEGFacross CKD. A second potential limitation is that the biopsy cohortshad a relatively small proportion of patients withDN (19 of 261 acrossthe three cohorts). To determine whether the results of our study weregeneralizable to CKD patients with diabetes, who form the majorityof CKD in the United States, we specifically investigated patients withDN. In the ERCB discovery cohort, we saw a correlation between intra-renal EGF mRNA expression and eGFR in the 17 DN patients. In ad-dition, of the 349 C-PROBE patients analyzed for noninvasive markerstudies, about 39% had diabetes and 20% of patients were diagnosedwith DN. Among these patients, we found correlations of uEGF with

www.ScienceT

eGFR and eGFR slope. These data suggest that the results of our studymay be generalizable to CKD patients with diabetes, but further valida-tion in cohorts of patients with DN will be required.

To assess association with and prediction of CKD progression byuEGF, we used two principal approaches. First, eGFR slope was derivedby a mixed-effect model to capture the trajectory of the disease overtime. eGFR slope estimates assume linearity of progression and arevulnerable to acute kidney injury and treatment effects, particularly inpatients with glomerular disease at the time of presentation (35, 36).Given these constraints, we focused the assessment of EGF and itsassociation with eGFR slope in the chronic renal disease cohort (C-PROBE). The second approachusedprogression toESKDor 40%eGFRreduction as composite end points, reflecting current concepts of clin-ical trial end points in CKD (37). The three cohorts used in this studyhad relatively short prospective follow-up time periods (1.8 to 3.6 years/mean), enriching this analysis for patients with rapid progression ofCKD. The ability of uEGF to identify CKD patients with rapid progres-sion might be particularly useful to enrich clinical trials with patientswho are more likely to reach end points during the limited trial obser-vationperiod.Toassess theutility of uEGF for informingCKDmanagement

–15

–10

–5

0

5r = 0.58, P < 0.001

A

Intrarenal EGF mRNA (log2)(n = 29)

B C

–20 –15 –10 –5 5 10

–15

–10

–5

5r = 0.65, P < 0.001

–20 –15 –10 –5 5 10

–15

–10

–5

5r = 0.29, P < 0.001

eGF

R s

lope

(%

)

Observed eGFR slope (%)(n = 344)

Observed eGFR slope (%)(n = 344)

eGF

R s

lope

(%

) pr

edic

ted

by A

CR

(lo

g2)

eG

FR

slo

pe (

%)

pred

icte

d by

uE

GF

/Cr

(log 2

)

5 10

Fig. 5. Associationof EGFwith eGFR slope. (A) Intrarenal EGFRNAexpressioncorrelated significantly (P< 0.001) with eGFR slope in C-PROBE patients (n= 29).

(n= 344)with slope predicted by uEGF/Cr (B) or ACR (C) using a regressionmodel(adjusted for age andgender). eGFR slopepredictedbyuEGF/Cr (B) exhibited

(B andC) Correlation of the observed eGFR slope of CKDpatients in C-PROBE

a higher correlation with the observed value than slope predicted by ACR (C).

Table 3. The association of uEGFwith time to composite end point. Theassociation was evaluated by a Cox proportional hazards regression analysisin C-PROBE, NEPTUNE, and PKU-IgAN cohorts. Renal event is defined as thepresenceof composite endpoint of ESKDor 40% reduction in baseline eGFR.The follow-up lengths (in years) for this analysis for the three cohortswere 1.8± 0.8, 2.0 ± 0.8, and 3.6 ± 2.2, respectively. AIC, Akaike information criterion;LR, likelihood ratio test.

Cohort

Model AIC C-statistics P value(LR test)

C-PROBE (25 events/189 patients)

M1

eGFR + ACR* 206.3 0.75 (0.58–0.91)

M2

eGFR + ACR* + EGF 192.1 0.87 (0.77–0.97) <0.0001

NEPTUNE (26 events/110 patients)

M1

eGFR + ACR* 213.98 0.74 (0.61–0.86)

M2

eGFR + ACR* + EGF 204.14 0.80 (0.68–0.92) 0.0006

PKU-IgAN (68 events/428 patients)

M1

eGFR + ACR* 660.18 0.71 (0.59–0.83)

M2

eGFR + ACR* + EGF 647.94 0.75 (0.64–0.86) 0.00016

*Adjusted for age and gender.

0.0 0.5 1.0 1.5

PKU-IgAN

NEPTUNE

C-PROBE

Adjusted HR of uEGF/Cr

0.53 (0.37–0.69)

0.29 (0.15–0.58)

0.27 (0.13–0.54)

Cohort HR (95% CI)

Fig. 6. Multivariable-adjusted HRs for predicting the composite endpoint on the basis of uEGF/Cr. HRs were adjusted by age, gender, eGFR,

and ACR. Adjusted HRs and 95% CIs were obtained by separate Cox regres-sion models in each study cohort. A one-unit decrease of uEGF/Cr (in logscale) was associated with an increased risk of CKD progression of 3.73(1.85 to 7.69)–fold, 3.43 (1.72 to 6.67)–fold, and 1.96 (1.45 to 2.70)–fold in thesethree cohorts. The unadjusted HRs for EGF were 0.33 (0.21 to 0.51) (C-PROBE),0.33 (0.21 to 0.52) (NEPTUNE), and 0.57 (0.46 to 0.70) (PKU-IgAN).

ranslationalMedicine.org 2 December 2015 Vol 7 Issue 316 316ra193 7

R E S EARCH ART I C L E

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from

in routine clinical settings, further comprehensive studies with structuredlong-term follow-up will be needed.

In summary, our study used an unbiased screening strategy to iden-tify uEGF as a predictor of CKDprogression in patientswith glomerulardisease, in three independent cohorts with diverse ethnic and geographicbackgrounds. uEGF was linked to tubular differentiation and regenera-tion, a mechanism essential to retain renal function in CKD, but not wellreflected by the conventional predictors (proteinuria or baseline GFR).Addition of uEGF into a CKD biomarker panel will likely improve riskstratification of CKD patients and thereby enhance the ability to targetclinical care and limited health care resources to those inmost need, aswellas to optimize clinical trial design. In addition, the importance of EGF-dependentmechanisms inCKDprogression suggests potential therapeutictargets for patients with impaired regenerative tubular functional reserve.

MATERIALS AND METHODS

Study designThe objective of this study was to identify and validate noninvasivemarkers for CKD progression. We applied a transcriptome-driven se-quential strategy. For tissue transcriptome analysis, gene expressionprofiles from 164 consecutive biopsies from the ERCB (38) were usedas a discovery cohort. Subsequent profiles of 55 patients from ERCBserved as the first validation cohort, and 42 patients from C-PROBEwhose samples were available for tissue gene expression analysis servedas the second validation cohort (Table 1). Thirty-two living donor trans-plant biopsies obtained at the timeof transplantationwereused as controls.Urinary biomarker analysis was performed on all samples available at thetime of the study from baseline visits from three independent cohortstudies (C-PROBE, NEPTUNE, and PKU-IgAN), representing 349,141, and 452 CKDpatients, respectively (Table 2). In the above cohorts,only patients aged 18 years or older were included in this study. All bio-specimens were procured after informed consent and with approval ofthe local ethics committees. The details of cohort designs and sampleprocurement are provided in Supplementary Materials and Methods.

Gene expression analysisTranscriptome analysis was performed on microdissected tubuloin-terstitial components of human renal biopsies prospectively procuredfor molecular analysis, using Affymetrix GeneChip and TaqMan LowDensity Arrays as previously published (39). Data processing details areprovided in Supplementary Materials and Methods. Normalized expres-sion data were log2-transformed and batch-corrected. FDR was appliedto account for multiple testing. The Cel files are available at Gene Expres-sionOmnibus (www.ncbi.nlm.nih.gov/geo/) under reference nos. GSE32591,GSE37455 (5), GSE35488 (6), GSE47185 (7), and GSE69438.

In situ hybridizationIn situ hybridization was performed by branched-DNA signal ampli-fication with QuantiGene ViewRNA ISH Tissue Assay Kit (Affymetrix/Panomics Solutions) on formalin-fixed paraffin-embedded human kid-ney tissue using the specific probe set for human EGFmRNA. Furtherdetails are provided in Supplementary Materials and Methods.

Measurements of urinary EGFuEGF concentration was measured in spot urine samples with HumanEGF Immunoassay Quantikine ELISA (R&D Systems). Samples (1:150dilution) and standards were run in duplicate, absorbance was measured

www.ScienceT

with a VersaMax ELISA plate reader, and results were calculated withSoftMax Pro (Molecular Devices). Detailed assay validation informationis provided in Supplementary Materials and Methods. uEGF was nor-malized to urine creatinine concentration and is referred to as uEGF/Cr.

GFR estimatesGFR was estimated by the four-variable Modification of Diet in RenalDisease (MDRD) study equation (40) inERCB,C-PROBE, andNEPTUNE.BothMDRD- andCKD-EPIGFRestimateswere determined forC-PROBEsubjects, and their log2-transformed values showed a high correlation(r = 0.993), indicating limited impact of the GFR estimation on the datapresented. InPKU-IgANcohort,GFRwas estimatedbyamodifiedMDRDequation (c-aGFR4) specifically adapted forChineseCKDpatients (41, 42).Log2-transformedc-aGFR4andMDRD-eGFRwerehighlycorrelated(r=0.996).

Outcome measureTwo outcomes were used to evaluate CKD progression: (i) eGFR slopeand (ii) a composite end point of ESKD or 40% reduction of baselineeGFR (37). eGFR slope was calculated for patients with at least threeeGFR records over a minimum follow-up of 1.5 years, including eGFRsrecorded before the baseline visit, if available. To account for irregularlyspaced creatinine measurements, a linear mixed-effects model (43) wasapplied to calculate subject-specific eGFR slope. More details areprovided in Supplementary Materials and Methods.

Evaluation of tubulointerstitial damage using IF/TAHistopathology was assessed in NEPTUNE biopsy cohort with digitalimages obtained from paraffin-embedded kidney tissue stained withhematoxylinandeosin, PAS, silver-based, and trichromestains (44).Wholeslide images of glass slides from 102 cases stored in the NEPTUNE digitalpathology repository were assessed for percentage of cortex involved byIF/TAby six pathologists. The percentage of cortex involved by IF/TAwasdetermined in each individual stain and averaged for an overall % value.More details are provided in Supplementary Materials and Methods.

Statistical analysisThe correlations of log2-transformed transcript expression values withboth baseline eGFR and eGFR slope were calculated using Pearson cor-relation. FDR was applied to account for multiple testing. The selectioncriteria for candidate transcript markers included (i) significant corre-lation of transcript levels with baseline eGFR (|r| > 0.4, FDR < 0.01) and(ii) significant differential expression of transcripts compared to livingdonor kidney biopsy controls (FDR≤ 0.05, fold change≥1.8 or≤0.56)in at least three of eight kidney diseases studied in the discovery cohort.

Ridge regression was performed on the discovery cohort to com-bine transcripts showing associations with baseline eGFR into a multi-marker signature for better prediction. In the analysis, the transcriptswere ranked by their strength of marginal association with log2 eGFR.The top-ranked k geneswere first standardized, then combined as linearpredictors withweights of correlation coefficients, and then carried overinto two validation cohorts.

Linear regression was applied to investigate the association betweeneGFR slope and the noninvasive proteinmarker uEGF, adjusted for ageand gender. The overall quality of fit was evaluated by an F test at the0.05 level of significance.

To evaluate the classification power of the candidate marker, weused logistic regression and AUC values from the ROC curves to com-pare nested models (www.inside-r.org/packages/cran/epicalc/docs/

ranslationalMedicine.org 2 December 2015 Vol 7 Issue 316 316ra193 8

R E S EARCH ART I C L E

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from

lrtest). To assess the effect of the candidate marker on the hazard ofreaching composite end points, Cox proportional hazards models werefit to the available prospective data. We examined nested models, start-ing with the base model (M1) including baseline eGFR, ACR, age, andgender. Samples with missing data were excluded from the analysis. Inthe subsequent model (M2), uEGF/Cr was added, and the goodness offit and improved prediction ability of the additional parameters were as-sessed by likelihood ratio tests, C-statistics, and AIC. All statistical analysiswasperformedusingR (www.R-project.org) andSASsoftware (version9.3).

Our approach to minimize overfitting included rigorous correctionfor multiple testing of the initial gene expression analysis, sequentialvalidation of intrarenal markers in two test cohorts with two differentmRNA quantification technologies (microarray and qRT-PCR), andreplication of the uEGF analysis (association with eGFR and predictionof the composite renal end point) in two independent cohorts. Further-more, we limited the number of covariates in Cox models with respectto observations. All patients who passed inclusion criteria and hadsamples available at the time of the study were included in the analysis.

SUPPLEMENTARY MATERIALS

www.sciencetranslationalmedicine.org/cgi/content/full/7/316/316ra193/DC1Materials and MethodsFig. S1. Baseline eGFR prediction by a three-marker panel in C-PROBE.Fig. S2. EGF and two other candidate mRNAs’ expression patterns in a panel of 84 humanorgans, tissues, and cell lines.Fig. S3. Correlation of EGF mRNA and uEGF/Cr with eGFR and eGFR slope in DN patients andCKD patients with diabetes.Fig. S4. Correlation of uEGF/Cr with ACR.Fig. S5. EGF as the top upstream regulator of genes correlated with eGFR slope.Fig. S6. Correlation of uEGF/Cr with eGFR slope.Fig. S7. ROC curve and corresponding AUC statistics for models with and without uEGF.Fig. S8. mRNA localization by in situ hybridization: Negative and positive control images.Table S1. Significantly enriched canonical pathways in intrarenal marker set.Table S2. qRT-PCR assays used to validate expression of the intrarenal transcripts.Table S3. Correlations of identified intrarenal transcripts with log2 eGFR of patients from thediscovery and two validation cohorts.Table S4. Demographic characteristics of NEPTUNE patients with intrarenal EGF expressiondata available.Table S5. Top 10 upstream regulators of transcripts correlated with eGFR change over time(eGFR slope).References (45–53)

REFERENCES AND NOTES

1. R. Lozano, M. Naghavi, K. Foreman, S. Lim, K. Shibuya, V. Aboyans, J. Abraham, T. Adair,R. Aggarwal, S. Y. Ahn, M. Alvarado, H. R. Anderson, L. M. Anderson, K. G. Andrews, C. Atkinson,L. M. Baddour, S. Barker-Collo, D. H. Bartels, M. L. Bell, E. J. Benjamin, D. Bennett, K. Bhalla,B. Bikbov, A. Bin Abdulhak, G. Birbeck, F. Blyth, I. Bolliger, S. Boufous, C. Bucello, M. Burch,P. Burney, J. Carapetis, H. Chen, D. Chou, S. S. Chugh, L. E. Coffeng, S. D. Colan, S. Colquhoun,K. E. Colson, J. Condon, M. D. Connor, L. T. Cooper, M. Corriere, M. Cortinovis, K. C. de Vaccaro,W. Couser, B. C. Cowie, M. H. Criqui, M. Cross, K. C. Dabhadkar, N. Dahodwala, D. De Leo,L. Degenhardt, A. Delossantos, J. Denenberg, D. C. Des Jarlais, S. D. Dharmaratne, E. R. Dorsey,T. Driscoll, H. Duber, B. Ebel, P. J. Erwin, P. Espindola, M. Ezzati, V. Feigin, A. D. Flaxman,M. H. Forouzanfar, F. G. Fowkes, R. Franklin, M. Fransen, M. K. Freeman, S. E. Gabriel, E. Gakidou,F. Gaspari, R. F. Gillum, D. Gonzalez-Medina, Y. A. Halasa, D. Haring, J. E. Harrison, R. Havmoeller,R. J. Hay, B. Hoen, P. J. Hotez, D. Hoy, K. H. Jacobsen, S. L. James, R. Jasrasaria, S. Jayaraman,N. Johns, G. Karthikeyan, N. Kassebaum, A. Keren, J.-P. Khoo, L. M. Knowlton, O. Kobusingye,A. Koranteng, R. Krishnamurthi, M. Lipnick, S. E. Lipshultz, S. L. Ohno, J. Mabweijano,M. F. MacIntyre, L. Mallinger, L. March, G. B. Marks, R. Marks, A. Matsumori, R. Matzopoulos,B. M. Mayosi, J. H. McAnulty, M. M. McDermott, J. McGrath, G. A. Mensah, T. R. Merriman,C. Michaud, M. Miller, T. R. Miller, C. Mock, A. O. Mocumbi, A. A. Mokdad, A. Moran, K. Mulholland,M. N. Nair, L. Naldi, K. M. Narayan, K. Nasseri, P. Norman, M. O’Donnell, S. B. Omer, K. Ortblad,R. Osborne, D. Ozgediz, B. Pahari, J. D. Pandian, A. P. Rivero, R. P. Padilla, F. Perez-Ruiz, N. Perico,D. Phillips, K. Pierce, C. A. Pope III, E. Porrini, F. Pourmalek, M. Raju, D. Ranganathan, J. T. Rehm,

www.ScienceT

D. B. Rein, G. Remuzzi, F. P. Rivara, T. Roberts, F. Rodriguez De León, L. C. Rosenfeld, L. Rushton,R. L. Sacco, J. A. Salomon, U. Sampson, E. Sanman, D. C. Schwebel, M. Segui-Gomez,D. S. Shepard, D. Singh, J. Singleton, K. Sliwa, E. Smith, A. Steer, J. A. Taylor, B. Thomas,I. M. Tleyjeh, J. A. Towbin, T. Truelsen, E. A. Undurraga, N. Venketasubramanian, L. Vijayakumar,T. Vos, G. R. Wagner, M. Wang, W. Wang, K. Watt, M. A. Weinstock, R. Weintraub, J. D. Wilkinson,A. D. Woolf, S. Wulf, P.-H. Yeh, P. Yip, A. Zabetian, Z.-J. Zheng, A. D. Lopez, C. J. Murray,M. A. AlMazroa, Z. A. Memish, Global and regional mortality from 235 causes of death for20 age groups in 1990 and 2010: A systematic analysis for the Global Burden of DiseaseStudy 2010. Lancet 380, 2095–2128 (2012).

2. A. S. Levey, R. Atkins, J. Coresh, E. P. Cohen, A. J. Collins, K. U. Eckardt, M. E. Nahas, B. L. Jaber,M. Jadoul, A. Levin, N. R. Powe, J. Rossert, D. C. Wheeler, N. Lameire, G. Eknoyan, Chronickidney disease as a global public health problem: Approaches and initiatives—A positionstatement from Kidney Disease Improving Global Outcomes. Kidney Int. 72, 247–259 (2007).

3. J. Coresh, D. Byrd-Holt, B. C. Astor, J. P. Briggs, P. W. Eggers, D. A. Lacher, T. H. Hostetter,Chronic kidney disease awareness, prevalence, and trends among U.S. adults, 1999 to2000. J. Am. Soc. Nephrol. 16, 180–188 (2005).

4. V. Jha, G. Garcia-Garcia, K. Iseki, Z. Li, S. Naicker, B. Plattner, R. Saran, A. Y.-M. Wang, C.-W. Yang,Chronic kidney disease: Global dimension and perspectives. Lancet 382, 260–272 (2013).

5. I. R. Whitman, H. I. Feldman, R. Deo, CKD and sudden cardiac death: Epidemiology, mecha-nisms, and therapeutic approaches. J. Am. Soc. Nephrol. 23, 1929–1939 (2012).

6. B. M. Brenner, T. W. Meyer, T. H. Hostetter, Dietary protein intake and the progressivenature of kidney disease: The role of hemodynamically mediated glomerular injury inthe pathogenesis of progressive glomerular sclerosis in aging, renal ablation, and intrinsicrenal disease. N. Engl. J. Med. 307, 652–659 (1982).

7. A. B. Fogo, Mechanisms of progression of chronic kidney disease. Pediatr. Nephrol. 22, 2011–2022 (2007).

8. K. S. Hodgkins, H. W. Schnaper, Tubulointerstitial injury and the progression of chronickidney disease. Pediatr. Nephrol. 27, 901–909 (2012).

9. S. Martini, V. Nair, B. J. Keller, F. Eichinger, J. J. Hawkins, A. Randolph, C. A. Böger, C. A. Gadegbeku,C. S. Fox, C. D. Cohen, M. Kretzler; European Renal cDNA Bank; C-PROBE Cohort; CKDGenConsortium, Integrative biology identifies shared transcriptional networks in CKD. J. Am.Soc. Nephrol. 25, 2559–2572 (2014).

10. J. P. T. Higgins, L. Wang, N. Kambham, K. Montgomery, V. Mason, S. U. Vogelmann, K. V. Lemley,P. O. Brown, J. D. Brooks, M. van de Rijn, Gene expression in the normal adult human kidneyassessed by complementary DNA microarray. Mol. Biol. Cell 15, 649–656 (2004).

11. E. C. Salido, J. Lakshmanan, D. A. Fisher, L. J. Shapiro, L. Barajas, Expression of epidermalgrowth factor in the rat kidney. An immunocytochemical and in situ hybridization study.Histochemistry 96, 65–72 (1991).

12. H. D. Humes, D. A. Cieslinski, T. M. Coimbra, J. M. Messana, C. Galvao, Epidermal growthfactor enhances renal tubule cell regeneration and repair and accelerates the recovery ofrenal function in postischemic acute renal failure. J. Clin. Invest. 84, 1757–1761 (1989).

13. J. Lechner, N. A. Malloth, P. Jennings, D. Heckl, W. Pfaller, T. Seppi, Opposing roles of EGF inIFN-a-induced epithelial barrier destabilization and tissue repair. Am. J. Physiol. Cell Physiol.293, C1843–C1850 (2007).

14. R. Safirstein, A. Z. Zelent, P. M. Price, Reduced renal prepro-epidermal growth factor mRNAand decreased EGF excretion in ARF. Kidney Int. 36, 810–815 (1989).

15. A. B. Farris, R. B. Colvin, Renal interstitial fibrosis: Mechanisms and evaluation. Curr. Opin.Nephrol. Hypertens. 21, 289–300 (2012).

16. M. M. Diaz Encarnacion, M. D. Griffin, J. M. Slezak, E. J. Bergstralh, M. D. Stegall, J. A. Velosa,J. P. Grande, Correlation of quantitative digital image analysis with the glomerular filtrationrate in chronic allograft nephropathy. Am. J. Transplant. 4, 248–256 (2004).

17. M. L. Nicholson, T. A. McCulloch, S. J. Harper, T. J. Wheatley, C. M. Edwards, J. Feehally,P. N. Furness, Early measurement of interstitial fibrosis predicts long-term renal functionand graft survival in renal transplantation. Br. J. Surg. 83, 1082–1085 (1996).

18. National Kidney Foundation, K/DOQI clinical practice guidelines for chronic kidney disease:Evaluation, classification, and stratification. Am. J. Kidney Dis. 39 (Suppl. 1), S1–S266 (2002).

19. T. Kurashina, S. Nagasaka, N. Watanabe, D. Yabe, N. Sugi, K. Nin, M. Hosokawa, Y. Nomura,M. Fukushima, Y. Nakai, F. Nishimura, A. Taniguchi, Circulating TNF receptor 2 is closelyassociated with the kidney function in non-diabetic Japanese subjects. J. Atheroscler.Thromb. 21, 730–738 (2014).

20. T. Gohda, M. A. Niewczas, L. H. Ficociello, W. H. Walker, J. Skupien, F. Rosetti, X. Cullere,A. C. Johnson, G. Crabtree, A. M. Smiles, T. N. Mayadas, J. H. Warram, A. S. Krolewski, Circulating TNFreceptors 1 and 2 predict stage 3 CKD in type 1 diabetes. J. Am. Soc. Nephrol. 23, 516–524 (2012).

21. S. M. Titan, J. M. Vieira Jr., W. V. Dominguez, S. R. S. Moreira, A. B. Pereira, R. T. Barros, R. Zatz,Urinary MCP-1 and RBP: Independent predictors of renal outcome in macroalbuminuricdiabetic nephropathy. J. Diabetes Complications 26, 546–553 (2012).

22. W. He, R. J. Tan, Y. Li, D. Wang, J. Nie, F. F. Hou, Y. Liu, Matrix metalloproteinase-7 as a surrogatemarker predicts renal Wnt/b-catenin activity in CKD. J. Am. Soc. Nephrol. 23, 294–304 (2012).

23. D. D. Torres, M. Rossini, C.Manno, F.Mattace-Raso, C. D’Altri, E. Ranieri, P. Pontrelli, G. Grandaliano,L. Gesualdo, F. P. Schena, The ratio of epidermal growth factor tomonocyte chemotactic peptide-1 in the urine predicts renal prognosis in IgA nephropathy. Kidney Int. 73, 327–333 (2008).

ranslationalMedicine.org 2 December 2015 Vol 7 Issue 316 316ra193 9

R E S EARCH ART I C L E

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from

24. G. Grandaliano, L. Gesualdo, F. Bartoli, E. Ranieri, R. Monno, A. Leggio, G. Paradies, E. Caldarulo,B. Infante, F. P. Schena, MCP-1 and EGF renal expression and urine excretion in human con-genital obstructive nephropathy. Kidney Int. 58, 182–192 (2000).

25. D. Serón, F. Moreso, J. Bover, E. Condom, S. Gil-Vernet, C. Cañas, X. Fulladosa, J. Torras,M. Carrera, J. M. Grinyó, J. Alsina, Early protocol renal allograft biopsies and graft outcome.Kidney Int. 51, 310–316 (1997).

26. D. Serón, Risk factors associated with the deterioration of renal function: The role ofprotocol biopsies. Prilozi 28, 291–302 (2007).

27. Y. An, F. Xu, W. Le, Y. Ge, M. Zhou, H. Chen, C. Zeng, H. Zhang, Z. Liu, Renal histologic changes andthe outcome in patients with diabetic nephropathy. Nephrol. Dial. Transplant. 30, 257–266 (2015).

28. Y.-K. Tsau, C.-H. Chen, Urinary epidermal growth factor excretion in children with chronicrenal failure. Am. J. Nephrol. 19, 400–404 (1999).

29. E. R. Mathiesen, E. Nexø, E. Hommel, H.-H. Parving, Reduced urinary excretion of epider-mal growth factor in incipient and overt diabetic nephropathy. Diabet. Med. 6, 121–126(1989).

30. E. Ranieri, L. Gesualdo, F. Petrarulo, F. P. Schena, Urinary IL-6/EGF ratio: A useful prognostic markerfor the progression of renal damage in IgA nephropathy. Kidney Int. 50, 1990–2001 (1996).

31. T. Weinstein, D. Hwang, A. Lev-Ran, Y. Ori, A. Korzets, J. Levi, Excretion of epidermal growthfactor in human adult polycystic kidney disease. Isr. J. Med. Sci. 33, 641–642 (1997).

32. A. L. Mattila, L. Viinikka, I. Saario, J. Perheentupa, Human epidermal growth factor: Renalproduction and absence from plasma. Regul. Pept. 23, 89–93 (1988).

33. P. E. Jørgensen, S. D. Hilchey, E. Nexø, S. S. Poulsen, C. P. Quilley, Urinary epidermal growthfactor is excreted from the rat isolated perfused kidney in the absence of plasma. J. Endocrinol.139, 227–234 (1993).

34. W. B. H. Melenhorst, G. M. Mulder, Q. Xi, J. G. Hoenderop, K. Kimura, S. Eguchi, H. van Goor,Epidermal growth factor receptor signaling in the kidney: Key roles in physiology anddisease. Hypertension 52, 987–993 (2008).

35. L. G. Hunsicker, S. Adler, A. Caggiula, B. K. England, T. Greene, J. W. Kusek, N. L. Rogers,P. E. Teschan, Predictors of the progression of renal disease in the Modification of Dietin Renal Disease Study. Kidney Int. 51, 1908–1919 (1997).

36. L. Li, B. C. Astor, J. Lewis, B. Hu, L. J. Appel, M. S. Lipkowitz, R. D. Toto, X. Wang, J. T. Wright Jr.,T. H. Greene, Longitudinal progression trajectory of GFR among patients with CKD. Am. J.Kidney Dis. 59, 504–512 (2012).

37. J. Coresh, T. C. Turin, K. Matsushita, Y. Sang, S. H. Ballew, L. J. Appel, H. Arima, S. J. Chadban,M. Cirillo, O. Djurdjev, J. A. Green, G. H. Heine, L. A. Inker, F. Irie, A. Ishani, J. H. Ix, C. P. Kovesdy,A. Marks, T. Ohkubo, V. Shalev, A. Shankar, C. P. Wen, P. E. de Jong, K. Iseki, B. Stengel,R. T. Gansevoort, A. S. Levey; CKD Prognosis Consortium, Decline in estimated glomerularfiltration rate and subsequent risk of end-stage renal disease and mortality. JAMA 311,2518–2531 (2014).

38. Y. Yasuda, C. D. Cohen, A. Henger, M. Kretzler; European Renal cDNA Bank (ERCB) Consortium,Gene expression profiling analysis in nephrology: Towards molecular definition of renal dis-ease. Clin. Exp. Nephrol. 10, 91–98 (2006).

39. H. Schmid, A. Boucherot, Y. Yasuda, A. Henger, B. Brunner, F. Eichinger, A. Nitsche, E. Kiss,M. Bleich, H.-J. Gröne, P. J. Nelson, D. Schlöndorff, C. D. Cohen, M. Kretzler; European RenalcDNA Bank (ERCB) Consortium, Modular activation of nuclear factor-kB transcriptionalprograms in human diabetic nephropathy. Diabetes 55, 2993–3003 (2006).

40. A. S. Levey, J. Coresh, T. Greene, L. A. Stevens, Y. L. Zhang, S. Hendriksen, J. W. Kusek, F. Van Lente;Chronic Kidney Disease Epidemiology Collaboration, Using standardized serum creatininevalues in the modification of diet in renal disease study equation for estimating glomerularfiltration rate. Ann. Intern. Med. 145, 247–254 (2006).

41. Y.-C. Ma, L. Zuo, J.-H. Chen, Q. Luo, X.-Q. Yu, Y. Li, J.-S. Xu, S.-M. Huang, L.-N. Wang, W. Huang,M. Wang, G.-B. Xu, H.-Y. Wang, Modified glomerular filtration rate estimating equation forChinese patients with chronic kidney disease. J. Am. Soc. Nephrol. 17, 2937–2944 (2006).

42. L. Zhang, F. Wang, L. Wang, W. Wang, B. Liu, J. Liu, M. Chen, Q. He, Y. Liao, X. Yu, N. Chen,J.-e. Zhang, Z. Hu, F. Liu, D. Hong, L. Ma, H. Liu, X. Zhou, J. Chen, L. Pan, W. Chen, W. Wang, X. Li,H. Wang, Prevalence of chronic kidney disease in China: A cross-sectional survey. Lancet 379,815–822 (2012).

43. L. H. Ficociello, E. T. Rosolowsky, M. A. Niewczas, N. J. Maselli, J. M. Weinberg, A. Aschengrau,J. H. Eckfeldt, R. C. Stanton, A. T. Galecki, A. Doria, J. H. Warram, A. S. Krolewski, High-normalserum uric acid increases risk of early progressive renal function loss in type 1 diabetes:Results of a 6-year follow-up. Diabetes Care 33, 1337–1343 (2010).

44. L. Barisoni, C. C. Nast, J. C. Jennette, J. B. Hodgin, A. M. Herzenberg, K. V. Lemley, C. M. Conway,J. B. Kopp, M. Kretzler, C. Lienczewski, C. Avila-Casado, S. Bagnasco, S. Sethi, J. Tomaszewski,A. H. Gasim, S. M. Hewitt, Digital pathology evaluation in the multicenter Nephrotic Syn-drome Study Network (NEPTUNE). Clin. J. Am. Soc. Nephrol. 8, 1449–1459 (2013).

45. C. A. Gadegbeku, D. S. Gipson, L. B. Holzman, A. O. Ojo, P. X. K. Song, L. Barisoni, M. G. Sampson,J. B. Kopp, K. V. Lemley, P. J. Nelson, C. C. Lienczewski, S. G. Adler, G. B. Appel, D. C. Cattran,M. J. Choi, G. Contreras, K. M. Dell, F. C. Fervenza, K. L. Gibson, L. A. Greenbaum, J. D. Hernandez,S. M. Hewitt, S. R. Hingorani, M. Hladunewich, M. C. Hogan, S. L. Hogan, F. J. Kaskel, J. C. Lieske,K. E. C. Meyers, P. H. Nachman, C. C. Nast, A. M. Neu, H. N. Reich, J. R. Sedor, C. B. Sethna,H. Trachtman, K. R. Tuttle, O. Zhdanova, G. E. Zilleruelo, M. Kretzler, Design of the Nephrotic

www.ScienceTr

Syndrome Study Network (NEPTUNE) to evaluate primary glomerular nephropathy by a multi-disciplinary approach. Kidney Int. 83, 749–756 (2013).

46. C. D. Cohen, K. Frach, D. Schlöndorff, M. Kretzler, Quantitative gene expression analysis inrenal biopsies: A novel protocol for a high-throughput multicenter application. Kidney Int.61, 133–140 (2002).

47. C. C. Berthier, R. Bethunaickan, T. Gonzalez-Rivera, V. Nair, M. Ramanujam, W. Zhang,E. P. Bottinger, S. Segerer, M. Lindenmeyer, C. D. Cohen, A. Davidson, M. Kretzler, Cross-speciestranscriptional network analysis defines shared inflammatory responses in murine and humanlupus nephritis. J. Immunol. 189, 988–1001 (2012).

48. H. N. Reich, D. Tritchler, D. C. Cattran, A. M. Herzenberg, F. Eichinger, A. Boucherot, A. Henger,C. C. Berthier, V. Nair, C. D. Cohen, J. W. Scholey, M. Kretzler, A molecular signature of protein-uria in glomerulonephritis. PLOS One 5, e13451 (2010).

49. W. Ju, C. S. Greene, F. Eichinger, V. Nair, J. B. Hodgin, M. Bitzer, Y.-s. Lee, Q. Zhu, M. Kehata,M. Li, S. Jiang, M. P. Rastaldi, C. D. Cohen, O. G. Troyanskaya, M. Kretzler, Defining cell-typespecificity at the transcriptional level in human disease. Genome Res. 23, 1862–1873 (2013).

50. M. Frericks, C. Esser, A toolbox of novelmurine house-keeping genes identified bymeta-analysisof large scale gene expression profiles. Biochim. Biophys. Acta 1779, 830–837 (2008).

51. K. Matsushita, B. K. Mahmoodi, M.Woodward, J. R. Emberson, T. H. Jafar, S. H. Jee, K. R. Polkinghorne,A. Shankar, D. H. Smith, M. Tonelli, D. G. Warnock, C.-P. Wen, J. Coresh, R. T. Gansevoort,B. R. Hemmelgarn, A. S. Levey; Chronic Kidney Disease Prognosis Consortium, Comparison ofrisk prediction using the CKD-EPI equation and the MDRD study equation for estimated glo-merular filtration rate. JAMA 307, 1941–1951 (2012).

52. B. DeSilva, W. Smith, R. Weiner, M. Kelley, J. Smolec, B. Lee, M. Khan, R. Tacey, H. Hill, A. Celniker,Recommendations for the bioanalytical method validation of ligand-binding assays to supportpharmacokinetic assessments of macromolecules. Pharm. Res. 20, 1885–1900 (2003).

53. P. Zhang, P. X.-K. Song, A. Qu, T. Greene, Efficient estimation for patient-specific rates ofdisease progression using nonnormal linear mixed models. Biometrics 64, 29–38 (2008).

Acknowledgments: We acknowledge all participating centers of the ERCB–Kröner-FreseniusBiopsy Bank (ERCB-KFB), the C-PROBE, the NEPTUNE, the PKU-IgAN study cohort, and their par-ticipants for their cooperation. We thank the support of George M. O’Brien Michigan KidneyTranslational Core Center and theMichigan Diabetes Research Center at the University of Michigan.Funding: This study was supported by the Else Kröner-Fresenius Foundation (for ERCB); by theEuropean Consortium for High-Throughput Research in Rare Kidney Diseases (EURenOmics; Eu-ropean Union FP 7:305608); by NIH (R01DK079912, P30DK081943, DK083912, P30DK020572,and UL1RR000433); by Office of Rare Diseases Research, National Center for Advancing Transla-tional Sciences, National Institute of Diabetes and Digestive and Kidney Diseases, University ofMichigan andNephCure Kidney International (U54DK083912); and by the University ofMichiganHealth System and Peking University Health Sciences Center Joint Institute for Translational andClinical Research. Analysis of urine samples of C-PROBE patients was supported by Hoffman–LaRoche. Author contributions:M.K., W.J., and V.N. participated in the study design. W.J., S.S., L.Z.,A.R., M.T., C.S., and B.S. (for C-PROBE), W.J., S.S., S.M.B., and L.B. (for NEPTUNE), L.Z. (for PKU-IgAN),andC.D.C. (for ERCB) participated in datageneration. K.S., V.N., L.Z.,W.J., S.S., F.H.E., C.C.B., J.Y.-C.L., Y.Z.,P.X.K.S., L.H.M., L.E., I.F., G.C.D.-P., G.D.-N., B.S., M.C.M., M. Bobadilla, H.-Y.W., H.Z., J.L., L.Z., andM.K.participated in data analysis. W.J., V.N., L.Z., H.Z., L.H.M., K.S., M. Bitzer, S.M.B., L.B., M.G.S., F.C.B.,and M.K. participated in data interpretation; J.J.H., C.A.G., H.-Y.W., H.Z., and C.D.C. provided studymaterials; W.J. and C.C.B. participated in figure preparation; and W.J., V.N., L.H.M., P.X.K.S., F.C.B.,K.S., andM.K. participated in writing the paper. The contribution of V.N. in this study is, in part, tofulfill the thesis requirement at the Ludwig Maximilian University of Munich. W.J. and V.N. sharefirst authorship. W.J. and M.K. share correspondence. Competing interests:M. Bobadilla, G.C.D.-P.,G.D.-N., L.E., M.C.M., M.T., I.F., C.S., M.K., V.N., andW.J. hold a patent PCT/EP2014/073413 “Biomarkersand methods for progression prediction for chronic kidney disease” related to this work. M.K. re-ports grants from Hoffman–La Roche during the conduct of the study; research support fromAbbVie, AstraZeneca, Boehringer Ingelheim, and Eli Lilly outside the submitted work. M.K. is onthe Board Advisory Committee of AbbVie, Eli Lilly, and Pfizer (honoraria paid to institution). C.D.C.received speaker honoraria from Hoffman–La Roche. Data and materials availability: The geneexpression Cel files are available at Gene ExpressionOmnibus (www.ncbi.nlm.nih.gov/geo/) underreference nos. GSE32591, GSE37455, GSE35488, GSE47185, and GSE69438.

Submitted 12 June 2015Accepted 21 October 2015Published 2 December 201510.1126/scitranslmed.aac7071

Citation: W. Ju, V. Nair, S. Smith, L. Zhu, K. Shedden, P. X. K. Song, L. H. Mariani, F. H. Eichinger,C. C. Berthier, A. Randolph, J. Y.-C. Lai, Y. Zhou, J. J. Hawkins, M. Bitzer, M. G. Sampson, M. Thier, C. Solier,G. C. Duran-Pacheco, G. Duchateau-Nguyen, L. Essioux, B. Schott, I. Formentini, M. C. Magnone,M. Bobadilla, C. D. Cohen, S. M. Bagnasco, L. Barisoni, J. Lv, H. Zhang, H.- Y. Wang, F. C. Brosius,C. A. Gadegbeku, M. Kretzler, for the ERCB, C-PROBE, NEPTUNE, and PKU-IgAN Consortium, Tissuetranscriptome-driven identification of epidermal growth factor as a chronic kidney diseasebiomarker. Sci. Transl. Med. 7, 316ra193 (2015).

anslationalMedicine.org 2 December 2015 Vol 7 Issue 316 316ra193 10

disease biomarkerTissue transcriptome-driven identification of epidermal growth factor as a chronic kidney

the ERCB, C-PROBE, NEPTUNE, and PKU-IgAN ConsortiumBarisoni, Jicheng Lv, Hong Zhang, Hai-Yan Wang, Frank C. Brosius, Crystal A. Gadegbeku, Matthias Kretzler and forBrigitte Schott, Ivan Formentini, Maria C. Magnone, Maria Bobadilla, Clemens D. Cohen, Serena M. Bagnasco, Laura Sampson, Martina Thier, Corinne Solier, Gonzalo C. Duran-Pacheco, Guillemette Duchateau-Nguyen, Laurent Essioux,Celine C. Berthier, Ann Randolph, Jennifer Yi-Chun Lai, Yan Zhou, Jennifer J. Hawkins, Markus Bitzer, Matthew G. Wenjun Ju, Viji Nair, Shahaan Smith, Li Zhu, Kerby Shedden, Peter X. K. Song, Laura H. Mariani, Felix H. Eichinger,

DOI: 10.1126/scitranslmed.aac7071, 316ra193316ra193.7Sci Transl Med

tracked over time without requiring invasive biopsies.demonstrated that the amount of EGF in the urine is just as useful, providing a biomarker that can be easily epidermal growth factor (EGF) in the kidneys as a marker of kidney disease progression. Moreover, the authorsfour independent cohorts totaling hundreds of patients from around the world to identify the expression of

. have now drawn onet alare more likely to progress to end-stage disease and need aggressive management. Ju Chronic kidney disease is a common medical problem worldwide, but it is difficult to predict which patients

Urine marker to the rescue

ARTICLE TOOLS http://stm.sciencemag.org/content/7/316/316ra193

MATERIALSSUPPLEMENTARY http://stm.sciencemag.org/content/suppl/2015/11/30/7.316.316ra193.DC1

CONTENTRELATED

http://science.sciencemag.org/contenthttp://science.sciencemag.org/content/sci/361/6406/eaat9819.fullhttp://stm.sciencemag.org/content/scitransmed/10/441/eaan4886.fullhttp://stm.sciencemag.org/content/scitransmed/5/175/175ra29.fullhttp://stm.sciencemag.org/content/scitransmed/6/256/256ra136.fullhttp://stm.sciencemag.org/content/scitransmed/7/279/279ra36.full

REFERENCES

http://stm.sciencemag.org/content/7/316/316ra193#BIBLThis article cites 53 articles, 13 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

registered trademark of AAAS. is aScience Translational MedicineScience, 1200 New York Avenue NW, Washington, DC 20005. The title

(ISSN 1946-6242) is published by the American Association for the Advancement ofScience Translational Medicine

Copyright © 2015, American Association for the Advancement of Science

by guest on July 7, 2020http://stm

.sciencemag.org/

Dow

nloaded from