www.sciencemag.org/content/358/6361/321/suppl/DC1

Supplementary Materials for

Adrenergic nerves activate an angio-metabolic switch in prostate cancer

Ali H. Zahalka, Anna Arnal-Estapé, Maria Maryanovich, Fumio Nakahara, Cristian D.

Cruz, Lydia W. S. Finley, Paul S. Frenette*

*Corresponding author. Email: paul.frenette@einstein.yu.edu

Published 20 October 2017, Science 358, 321 (2017)

DOI: 10.1126/science.aah5072

This PDF file includes:

Materials and Methods

Figs. S1 to S20

Table S1

References

2

Materials and Methods Animals and in vivo procedures

B6.Cg-Tg(Cspg4-cre/Esr1*)BAkik, Adrb2tm1Bkk, Adrb3tm1Lowl, Cox10tm1Ctm (all from Jackson Laboratory); C57BL/6 Charles Rivers; BALB/c nude (from Charles Rivers and Taconic); FVB-Tg(ARR2/Pbsn-MYC)7Key (from the National Cancer Institute); Adrb2fl/fl mice were a kind gift from G. Karsenty; Cdh5-CreERT2 mice were a kind gift from R.H. Adams; Csf1r-iCre mice were a kind gift from J.W. Pollard.

All in vivo experiments were approved by the Animal Care and Use Committees of the Albert Einstein College of Medicine. Orthotopic human PC3-luciferase tumors were implanted into 6- to 8-week-old male mice, and monitored by bioluminescence imaging as previously described (9). In vivo fluorescence imaging of angiosense vascular pooling agent (PerkinElmer) was performed by intravascular injection of the agent and analyzed 24 h later using an IVIS imaging system 200 series (PerkinElmer).

Orthotopic co-transplantation of culture-expanded prostate endothelial cells (Angiocrine Bioscience) and prostate tumor cells was performed by co-injection of 105 GFP-tagged endothelial cells and 5 x 104 PC-3 cells suspended in 20PL endothelial cell media (Cell Culture below) containing 5% growth factor reduced (GFR) Matrigel (Corning) into the ventral prostates of 6- to 8-week-old male mice. Animals were sacrificed two weeks after transplantation, and the tumors removed for analysis.

To assess vascular recruitment in vivo in immune-competent mice, pure type I collagen matrices (Advanced BioMatrix), were cut in half along their diameter and then carefully inserted in-between the two ventral prostate lobes through a 2.5 mm incision made in the prostate capsule using fine iris scissors (paying close attention to avoid damaging the prostate lobes). Animals were sacrificed four weeks after transplantation and the ventral prostates removed for imaging.

Chemical sympathectomy (adrenergic denervation) was performed by intraperitoneal injection of 6OHDA (Sigma), given at a first dose of 100mg/kg (freshly dissolved in PBS prior to injection, to minimize oxidation), and a second dose of 250 mg/kg 72h later. For CreERTM and CreERT2 mediated recombination, 2mg tamoxifen (Sigma) was administered by intraperitoneal injection, once daily for five consecutive days.

To assess prostate perfusion, mice were injected IV with 126mg/kg luciferin (Goldbio) dissolved in sterile saline, which was allowed to circulate for 1 minute. Mice were then immediately sacrificed, 500PL of blood was removed by intracardiac puncture and placed in Eppendorf tubes containing 20PL EDTA, and prostates were removed and dissected on ice. Prostates were then minced with scissors and digested for 30 minutes at 37qC in digestion buffer (see Cell Sorting and Flow Cytometry section), and then centrifuged for 4 minutes to pellet the cells. The supernatant was removed and saved for analysis. Blood was centrifuged at 1200 x g for 15 minutes to isolate plasma (this served as the internal control). Luciferin content was measured by addition of the luciferase enzyme with excess ATP (so that sample luciferin content would be the limiting factor in light generation; Promega) and luminosity recorded with a luminometer (Perkin Elmer). Prostate luminescence was normalized to plasma luminescence and then expressed as relative luminescence units (RLU) / mg tissue weight. Immunohistochemistry and histology

Preparation of tumor, prostate, and orthotopic matrix frozen sections: tissues were perfusion-fixed, then post-fixed in 4% paraformaldehyde overnight at 4qC, then incubated in 15% sucrose and 30% sucrose for cryoprotection, then embedded in OCT cryomedium (Tissue-Tek) and flash-frozen in 2-methylbutane (Sigma) cooled in liquid nitrogen. Tissues were warmed to -20qC, trimmed to full face,

3

and sectioned at 100Pm (thick sections) or 20Pm (thin sections) using high-profile cryotome blades (Sakura Finetek).

Immunofluorescence staining of thick tissue sections: slides were blocked/ permeabilized in PBS containing 20% normal donkey serum (Sigma) and 1% Triton X-100 for 45mins at room temperature. Slides were incubated with primary antibody in 5% normal donkey serum with 0.01% Triton X-100 overnight at 4qC. Secondary antibody was applied for 45 min at room temperature.

Immunofluorescence images were acquired with an AXIO Examiner D1 microscope (Zeiss), equipped with a confocal scanner unit CSU-X1 (Yakagawa), and automated micrometer stage (Sutter Instrument). Z-stack montage images were stitched together and reconstructed in SlideBook software (Intelligent Imaging Innovations). For quantification of vascular indices of angiogenesis, images were analyzed using the Simple Neurite Tracer plug-in (12) found in the Fiji build of ImageJ (NIH). To quantify the proximity of interaction between adrenergic nerves and the vasculature we employed a combination of morphometry and proximity analysis that has been used for studying neural organization in the central nervous system (44). Briefly, all vessels and nerves in an image were traced using Simple Neurite Tracer and converted into skeleton representations. The SWC skeleton files were then imported into the statistical programming language R, and all nerves and all vessels were converted into point vector representations with the dotprops function of the NBLAST package (45), and proximity of interaction calculated using the neuriteblast function of the same package, using a V of 10Pm. The average of the forward and reverse interaction for each nerve – vessel pair was calculated and then weighted based upon the vessel’s length as a percent of total vessel coverage in the image. Thus, the proximity scores range from 0 to 1, where 0 = complete overlap, 0.5 = close association of two structures, and 1 = no association.

For pathology analyses, paraffin-embedded prostate tissues were trimmed to full-face then sectioned at 5 Pm thickness and stained with hematoxylin and eosin. Entire cross-sectional images of slides were acquired with the P250 High Capacity Slide Scanner (PerkinElmer) and pathology manually analyzed (all lobes excluding coagulation glands) in Panoramic Viewer (3dhistotech).

To assess vascular leakage, 70,000 MW Texas Red-conjugated dextran was injected IV and allowed to circulate for 30 minutes. Tissues were then perfused, fixed, and processed. Vascular leakage was assessed by quantifying relative dextran fluorescence in the prostate interstitium.

For COA6 quantification, equal amounts of protein from cell lysates were subjected to SDS-PAGE, followed by western blot analysis using anti-E-actin and anti-COA6 antibodies (see Antibodies and Staining Reagents below). Catecholamine measurement Prostates were freshly dissected and immediately flash frozen in liquid nitrogen. Tissue samples were homogenized in 0.1M tricholoroacetic acid containing 10-2M sodium acetate, 10-4M EDTA, 5ng/mL isoproterenol (as an internal standard) and 10.5% methanol (pH 3.8), and centrifuged at 10,000 x g for 20 minutes, and the supernatant was analyzed by high pressure liquid chromatography (HPLC) by the Neurochemistry Core Laboratory at Vanderbilt University’s Center for Molecular Neuroscience Research (Nashville, TN).

Cell sorting and flow cytometry

Endothelial cells were isolated from dissected prostate and tumor tissues by mincing tissues in digestion buffer composed of Hank’s balanced salt solution (HBSS; Gibco), 2.5 mg/mL collagenase IA (Sigma), 0.6 mg/mL dispase (Gibco), 0.25 mg/mL DNase I (Sigma), 20mM HEPES (Gibco), and incubated at 37qC for 35 min on a rotating rack. All cell sorting experiments were carried out using an Aria Cell Sorter (BD Biosciences), and all flow cytometric analyses were carried out using an LSRII

4

flow cytometer running FACS Diva 8.0 software (all BD Biosciences). Samples were run in azide-free FACS buffer (so as to avoid mitochondrial damage), and cell doublets, debris, and dead cells were excluded by FCS, SSC, and DAPI (4’,6-diamidino-2-phenylindole) profiles. Data were analyzed by FlowJo (Tree Star) software.

To assess endothelial mitochondrial membrane potential '\, cell suspensions were stained on ice with cell surface markers (Antibodies and Staining Reagents below), then incubated in 100 nM TMRE (life technologies) with 50 nM Mitotracker green or Mitotracker deep red (Life Technologies and Cell Signaling Technologies, respectively) in HBSS containing 40 PM verapamil (to inhibit cell surface P-glycoprotein efflux pumps from expelling TMRE; Sigma) for 20 min light-protected at 37qC. TMRE and Mitotracker green (or deep red) fluorescence were determined by flow cytometry. To assess endothelial glucose uptake, cell suspensions were stained on ice with cell surface markers (Antibodies and Staining Reagents below), then incubated in 100 PM of the fluorescent deoxyglucose analog 2-NBDG (Cayman Chemical) suspended in glucose-free, pyruvate-free, glutamate-free Dulbecco’s Modified Eagle Medium supplemented with 1x GlutaMAX (both from ThermoFisher Scientific). 2-NBDG fluorescence was quantified by flow cytometry.

To assess endothelial NADH levels, cell suspensions were stained on ice with cell surface markers (Antibodies and Staining Reagents below), then stained with the fixable viability dye eFluor 506 (eBioscience) in HBSS for 25 minutes. Since NADH exhibits a characteristic autofluorescent peak at 470nm after excitation by a UV laser (27), mitochondrial NADH was calculated as a ratio of basal NADH autofluorescence in half the sample, then maximal NADH after addition of 2 Pg of antimycin A (Sigma), a potent inhibitor of oxidative phosphorylation, in the other half. To assess endothelial cell respiratory capacity (RC), cell suspensions were prepared as described above, then basal oxidized NAD+

levels (the non-autofluorescent fraction) were measured in half the sample, then maximal levels after addition of 1PM of the oxidative phosphorylation uncoupler FCCP (Cayman Chemical), thus achieving maximal respiration. RC was calculated as a ratio of maximal NAD+ levels to basal levels.

To assess cell proliferation and apoptosis by flow cytometry, cells suspensions were stained on ice with cell surface markers (Antibodies and Staining Reagents below), then with the fixable viability dye eFluor 506, followed by cell fixation and permeabilization and stained with the primary antibodies to Ki-67 (for proliferation) and cleaved CASP3 (for apoptosis; see Antibodies and Staining Reagents below) according to the manufacturer’s instructions (BD biosciences). After exclusion of dead cells (by eFluor 506 fixable viability dye), Ki-67 and cleaved CASP3 fluorescence was measured by flow cytometry.

Antibodies and staining reagents

Alexa Fluor 647-anti-CD31 (MEC13.3), Alexa Fluor 700-anti-CD45 (30-F11), Alexa Fluor 700-anti-TER119 (TER-119), Alexa Fluor 700-anti-Ly6A/E (D7), PE-Cy7-anti-CD117 (2B8) [all Biolegend]; PerCP-eFluor 710-anti-Ki-67 (SolA15), PerCP-Cy5.5-anti-CD11b (M1/70), PE-anti-PDGFRD (APA5), APC-anti-PDGFRE (APB5), APC/APC-eFluor780/FITC-anti-EpCAM (G8.8), PE-Cy7-anti-CD31 (390), Alexa Fluor 488/eFluor660-anti-Lyve-1 (ALY7), Streptavidin APC-eFluor 780 [all eBioscience]; FITC-anti-Active cleaved Caspase-3 (C92-605), Biotin-anti-Lineage panel cocktail [TER-119 (TER-119), Ly-6G/C (RB6-8C5), CD11b (M1/70), CD3e (145-2C11)] [BD Bioscience]; anti-C1orf31 (COA6) [Protein Biotechnologies]; Alexa Fluor 488/568/647-anti-rabbit, and Alexa Fluor 594 Cick-iT£ Plus TUNEL [Thermo Fisher]; anti-tyrosine hydroxylase, anti-NG2, Alexa Fluor 555 anti-D-tubulin [Millipore]; Cy3-anti-DSMA antibody, anti-E-actin, and Hoechst 33342 [Sigma].

5

Cell culture and in vitro assays PC-3 cells stably expressing luciferase (luciferase construct, gift from W. Guo, Albert Einstein

College of Medicine) were cultured in Ham’s F-12 Nutrient Mix (Gibco), supplemented with 10% FBS (Gibco), 1.5 g/L sodium bicarbonate (Sigma), and 500 mg/mL G418 (Sigma).

Culture-expanded primary mouse prostate endothelial cells (Angiocrine Biosciences) were plated on 2Pg/mL fibronectin-coated culture dishes and cultured according to the manufacturer’s protocol. To selectively knockdown Adrb2 in VeraVec prostate endothelial cells, Adrb2/Control short hairpin RNA (shRNA) lentiviruses were generated by co-transfecting HEK-293T cells with 12Pg shRNA shuttle vector (GIPZ Lentiviral shRNA Library, OpenBiosystems, provided by John F. Reidhaar-Olson at the Einstein shRNA Facility), 0.6Pg tat vector, 0.6Pg gag/pol vector, 1Pg vsv-g vector using 45PL TransIT-X2 transfection reagent (Mirus Bio). Supernatant was collected from transfected HEK cells, filtered, and concentrated by ultracentrifugation. Concentrated virus preparations were used to transduce prostate endothelial cells, and cells expressing the construct (expressing GFP) were selected by FACS sorting. To selectively overexpress Coa6/Bckdha/or Control-GFP in VeraVec prostate endothelial cells, lentivirus vectors were made by cloning Coa6 or Bckdha gene from pCMV-SPORT6 vectors (OpenBiosystems Library provided by John F. Reidhaar-Olson at the Einstein shRNA Facility) into pHIV-IRES-ZsGreen vector (a kind gift from Chan-Jung Chang, Mount Sinai School of Medicine) using Gateway cloning System (Thermo Fisher Scientific). Then lentiviruses were generated by transfecting 12mg of pHIV/ Coa6, Bckdha or empty control vector with 0.6Pg tat vector, 0.6Pg rev vector, 0.6Pg gag/pol vector, and 1Pg vsv-g vector using 45Pl TransIT-X2 transfection reagent (Mirus Bio). Supernatant was collected from transfected HEK cells, filtered, and concentrated by ultracentrifugation.

To measure endothelial cell migration, cells were plated on fibronectin-coated culture dishes and allowed to reach confluence. Complete media was then replaced with serum free, cytokine-free media and cells were allowed to acclimate for 4 h at 37qC, before a “scratch” was created in the confluent monolayer with a p200 pipette tip. Images were acquired after the scratch was made (time 0) and imaged again 36 hours later. Migration was calculated as the percentage of the scratch area covered by migrating cells 36 h post scratch, and then normalized to control. cAMP levels were measured using a cAMP ELISA Kit according to the manufacturer’s instructions (Cell Biolabs, Inc.) after cells were stimulated for 30 minutes with 10PM noradrenaline (the adrenergic neurotransmitter) in the presence of 0.5mM of the phosphodiesterase inhibitor IMBX (3-isobutyl-1-methylxanthine; Sigma). For adenylyl cyclase experiments, cells were incubated for 36 h prior to analysis with either 10PM forskolin (Cayman) or 300PM SQ 22,536 (Cayman) in the presence of 10PM noradrenaline. Metabolic assays

For all metabolic experiments, endothelial cells were re-suspended in serum-free media containing 5mM glucose and 5% GFR Matrigel 36 h prior to analysis. For experiments involving shAdrb2 or EAdRKO-derived endothelial cells, GFR-containing media was supplemented with 10PM noradrenaline (the adrenergic neurotransmitter) and 100PM ascorbic acid (to prevent noradrenaline oxidation; both from Sigma). For galactose experiments, cells were seeded in 5mM glucose-containing media and the next day media was either replaced with fresh 5mM glucose (Sigma)-containing media or 5mM galactose (Sigma)-containing media. ATP levels were measured in lysed endothelial cells using the ATPlite Luminescent kit (PerkinElmer) according to the manufacturer’s instructions. Lactate concentration was measured in cell supernatant using the Lactate Assay Kit (BioVision) according to the manufacturer’s instructions, and levels normalized to total protein content in each sample and then displayed relative to control.

Endothelial respiration (oxygen consumption rate) was measured using the XFe96 analyzer (Seahorse Bioscience). For shAdrb2 and Coa6-GFP experiments, 3.5x104 cells were added per well in

6

XFe96 microplates. EAdRKO and EAdRWT -derived endothelial cells were obtained by FACS sorting 104 endothelial cells, freshly isolated from day 18 tumors (including the surrounding prostate tissue), directly into the center of each well in XFe96 microplates. Matrigel-containing media was replaced 1 h prior to the assay with non-buffered assay medium supplemented with 20% KnockOut Serum Replacement (ThermoFisher Scientific) and placed in a carbon-dioxide-free incubator. Oxygen consumption rate was measured sequentially under basal conditions, in response to 3PM oligomycin (an ATP synthase inhibitor), 1PM FCCP (a mitochondrial uncoupler), and 4Pg/mL of antimycin A with 3PM rotenone (both potent inhibitors of oxidative phosphorylation).

To perform metabolite profiling and isotope tracing, cells were seeded on fibronectin-coated 6-well plates dishes in pyruvate-free, glucose-free, glutamine-free DMEM, supplemented with 5mM 12C-glucose (Sigma), 2mM 12C-glutamine (Gibco), or their uniformly labeled 13C counterparts [U-13C]-glucose and [U-13C]-glutamine (Cambridge Isotope Laboratories), and 20% KnockOut Serum Replacement. Media was changed at the indicated time before analysis. Metabolites were extracted in 80% methanol containing 20μM of deuterated 2-hydroxyglutarate (d5-2HG) as an internal standard, and processed as previously described (46). To derivatize metabolites, they were methoximated and trimethylsilylated as previously described (46, 47). Metabolite profiles were analyzed by gas chromatography mass spectrometry under electron impact ionization using an Agilent 7890A GC coupled to an Agilent 5975C mass selective detector (Agilent Technologies) and peaks representing metabolites of interest were extracted and integrated using MassHunter software (Agilent). Values were then normalized to the internal standard, d5-2HG, and to duplicate sample protein content. Natural isotope abundance correction was performed by IsoCor software (48). Samples for liquid chromatography mass spectrometry (LC-MS) metabolite profiling were diluted with 10mmol ammonium acetate in methanol, containing acyl carnitine stable isotope standards (d3-C0, d3-C2, d3-C3, d3-C4, d3-C5, d3-C6, d3-C8, d3-C10, d3-C12, d3-C14, d3-C16, d3-C18), lipid internal standards (lysoPC a C9:0, PC aa C28:0, PC aa C40:0, SM C6:0), and injected into LC-MS for flow injection analysis. Microarray and Gene Set Enrichment Analysis

Total RNA from sorted endothelial cells was extracted using RNeasy Plus Micro Kit (Qiagen) according to the manufacturer’s instructions, and quality verified using the Agilent Bioanalyzer (Agilent Technologies). All further steps were performed at the Genomics Core Facility at the Albert Einstein College of Medicine according to the appropriate Affymetrix protocols, including RNA labelling with the GeneChip wild-type terminal labelling kit (Affymetrix), hybridization to Mouse Gene 2.0 ST arrays, and scanned. Raw data was normalized using the RMA algorithm in the Expression Console Software (Affymetrix), and unsupervised hierarchical clustering of genes that exhibited a t 1.6-fold change in expression and an ANOVA p-value < 0.05, was performed in the Transcriptome Analysis Console Software (Affymetrix). Gene set enrichment analysis (GSEA) was performed with C5 gene sets from the Molecular Signatures Database 5.0 (Broad Institute).

RNA extraction and qRT-PCR

mRNA was extracted and purified from cells sorted into lysis buffer (LifeTechnologies) using the Dynabeads mRNA DIRECT purification kit (LifeTechnologies), and reverse-transcribed using the RNA to cDNA EcoDry Premix (Clonetech), according to the manufacturer’s instructions. Quantitative PCR (qPCR) was performed either with SYBR GREEN master mix (primer sequences below; Roche), or TaqMan Gene Expression Master Mix (TaqMan Gene Expression Assays below; Applied Biosystems), and run on a QuantStudio 6 Real-Time PCR System (Applied Biosystems). The average threshold cycle number (Ct) for each tested gene was used to quantify relative expression: 2-(Ct(gene) –

Ct(control)). Primers sequences include: Gapdh forward, TGTGTCCGTCGTGGATCTGA; Gapdh reverse, CCTGCTTCACCACCTTCTTGA; Adrb2 forward, ATCTGAAGGAAGATTCCACGCCCA; Adrb2

7

reverse, AGAGGGTGAATGTGCCCATGATGA. TaqMan Gene Expression Assays include: Coa6 Mm00612966_m1; Bckdha Mm00476112_m1; Fh1 Mm01321349_m1; Gapdh Mm99999915_g1.

Statistical analysis

All data are represented as mean r s.e.m., except for seahorse OCR plots that are represented as mean r s.d. Comparison between two samples was performed using un-paired two-tailed Student’s t test. Comparison between multiple groups was performed using one-way ANOVA followed by Tukey’s multiple comparison test. Comparison between seahorse OCR plots was performed using a two-way ANOVA followed by Bonferroni’s multiple comparison test. Statistical analysis was performed with Prism 7 software (GraphPad).

8

Supplementary Figures:

Fig. S1. Pre-angiogenic switch (+18 days) βAdRKO and βAdRWT tumors display similarities in size and vascular function. (A) Quantification of tumor volume (using formula, volume = 0.5 length x width2; left), and luminescence by IVIS 18 days post tumor implantation (right). (B) Tumor vascular leakage assessed by in vivo angiosense fluorescence 24 h after injection. (C) Tumor hypoxia assessed in tumor cross-sections by pimonidazole positive areas. All experiments n = 4 mice/condition. Error bars = s.e.m. All p-values were not significant (> 0.05).

A

B C

0

1

2

3

4H

ypox

ic a

rea

(% o

f tum

or a

rea)

βadr+/+; +/+

βadrKO

0.0

5.0×109

1.0×1010

1.5×1010

Ang

iose

nse

fluor

esce

nce

(rad

iant

effi

cien

cy)

WTβ2-/-β3-/-

0

20

40

60

80

100Tu

mor

vol

ume

(mm

3 )

0

1×109

2×109

3×109

4×109

Rad

ianc

e (p

/sec

/cm

2 /sr)

βadr+

/+;+/+

βadrK

O0

2000

4000

6000

8000

10000

Mitoc

hond

rial m

ass

(Mito

track

er M

FI)

βadr+/+;+/+

βadrKOβAdRWTβAdRKO

9

Fig. S2. Minimal adrenergic nerve – vessel interaction is observed in the pre-neoplastic LPIN stage. (A) Quantification of adrenergic nerve density in the HPIN stage. n = 4 mice/condition. (B-C) Thick-section images and quantification of the association between adrenergic nerves and the prostate vasculature in the low-grade PIN (LPIN) stage. Representative images of wild-type (WT) prostate (top) and LPIN stage prostate (bottom). TH = tyrosine hydroxylase, CD31 = vasculature. Scale bar = 100µm (B). Quantification of LPIN stage vascular density (left), adrenergic nerve density (middle), and proximity of association between nerves and vessels (right). n = 4 mice/condition (C). (D) FACS quantification of epithelial cell proliferation during LPIN and HPIN stages. n = 4 mice/condition. *p<0.05. Error bars = s.e.m.

0

10

20

30

40

50

TH fi

ber l

engt

h (m

m) /

D

API+

area

(mm

2 )

WT

cMyc

*

Norad

rena

line

Adren

aline

Dopam

ine0

50

100

150

Cat

echo

lam

ines

(n

g/m

g pr

otei

n)

WT (8w)cMyc (8w; mPIN)

**** WT (8w)cMyc (8w; HPIN)

A B

C

D0

10

20

30

40TH

fibe

r len

gth

(mm

) /

DAP

I+ ar

ea (m

m2 )

WT (4w)

cMyc (4w; LPIN)

0.0

0.2

0.4

0.6

0.8

Ner

ve–v

esse

l ass

ocia

tion

(pro

xim

ity s

core

)

WT

cMyc

0

10

20

30

40

Tota

l ves

sel l

engt

h (m

m)

/ DA

PI+

area

(mm

2 )

WT (4w)

cMyc (4w; LPIN)

0

5

10

15

Epi

thel

ial c

ells

(% p

rolif

erat

ing)

*

TH CD31

WT

(4w

)cM

yc(4

w; L

PIN

)

WT (4

w)

WT (8

w)

cMyc

(4w)

cMyc

(8w)

0

5

10

15

Epi

thel

ial c

ells

(% p

rolif

erat

ing)

WT (4w)WT (8w)*

cMyc (4w)

cMyc (8w; mPIN)

WT (4w)WT (8w)cMyc (4w; LPIN)cMyc (8w; HPIN)

0

10

20

30

40

Tota

l ves

sel l

engt

h (m

m)

/ DA

PI+

area

(mm

2 )

WT (4w)

cMyc (4w; LPIN)WT (4w)cMyc (4w; LPIN)

10

Fig. S3. Adrb2 is expressed and efficiently deleted in prostate stromal cells, but does not contribute to cancer progression in non-endothelial stromal cells. (A) Quantification of Adrb2 expression in various angiogenesis-associated stromal populations in healthy and HPIN stage prostates isolated by FACS. n = 4 mice/condition. Stroma = CD45-, CD31-, CD140a-, CD140b-, Epcam- cells. (B) Adrb2 expression in FACS-sorted endothelial cells from HPIN-stage prostates. n = 3 mice/condition. (C) Adrb2 expression in FACS-sorted pericytes from HPIN-stage prostates. n = 4 mice/condition (left), its effect on HPIN-stage cancer progression (middle), and late stage cancer progression (right). (D) Adrb2 expression in FACS sorted myeloid cells from HPIN-stage prostates. n = 4-5 mice/condition (left), its effect on HPIN-stage cancer progression (middle), and late stage cancer progression (right). NS p>0.05. *p<0.05. **p<0.01. ***p<0.001. Error bars = s.e.m.

cMyc Adrb2fl/fl NG2-CreERTM

0

50

100

150

Pros

tate

can

cer

wei

ght (

% o

f con

trol)

cMyc

cMyc; Adrb2cKO

NS8 weeks 52 weeks

NS

cMyc Adrb2fl/fl Csf1r-iCre 12 weeksNS

52 weeksNS

A

CPr

osta

te c

ance

r w

eigh

t (%

of c

ontro

l)

Pros

tate

can

cer

wei

ght (

% o

f con

trol)

Pros

tate

can

cer

wei

ght (

% o

f con

trol)

Pros

tate

can

cer

wei

ght (

% o

f con

trol)

D

Cre-

Cre+0

50

100

150

Del

etio

n ef

ficie

ncy

(Adrb2/Gapdh

)

Csf1r-iCre Adrb2 deletion

Cre+Cre-*

Cre- Cre+

0

50

100

150

Del

etio

n ef

ficie

ncy

(Adrb2/Gapdh

) ***

Cre- Cre+

B

0

50

100

150

Pros

tate

wei

ght

(% o

f con

trol)

cMyccMyc; Adrb2cKO

cMyccMyc; Adrb2cKO

cMyc Adrb2fl/fl Cdh5-CreERT2

Stroma

Pericy

te

Endoth

elial

Myeloi

d

Stroma

Pericy

te

Endoth

elial

Myeloi

d0.00

0.05

0.10

0.15

Adrb2/Gapdh

WT (8w)

cMyc (8w; HPIN)*

Norad

rena

line

Adren

aline

Dopam

ine0

50

100

150

Cat

echo

lam

ines

(n

g/m

g pr

otei

n)

WT (8w)cMyc (8w; mPIN)

**** WT (8w)cMyc (8w; HPIN)

Cre- Cre+0

50

100

150

Del

etio

n ef

ficie

ncy

(Adrb2/Gapdh

)

Ng2-CreERTM Adrb2 deletion

Cre-Cre+

**

Cre- Cre+

0

50

100

150

200

250

Pro

stat

e w

eigh

t(%

of c

ontro

l)

Cre-

Cre+

Adrb2f/f Adrb2Δ/Δ

0

50

100

150

Pro

stat

e w

eigh

t (%

of c

ontro

l)

cMyccMyc; Adrb2cKO

0

50

100

150

Pros

tate

wei

ght

(nor

mal

ized

to c

ontro

l) cMyccMyc; Adrb2cKO

0

50

100

150

Pros

tate

can

cer

wei

ght (

% o

f con

trol)

cMyc

cMyc; Adrb2cKO

0

50

100

150

Pros

tate

wei

ght

(% o

f con

trol)

cMyccMyc; Adrb2cKO

0

50

100

150

Pros

tate

can

cer

wei

ght (

% o

f con

trol)

cMyc

cMyc; Adrb2cKO

11

Fig. S4. Adrb2 deletion in endothelial cells inhibits cancer progression and decreases nerve – vessel interaction. (A) Quantification of immunostained Ki-67+ proliferative cells in prostate sections n = 3 mice/condition. (B) FACS quantification of Ki-67+ proliferative epithelial cells n = 6 mice/condition, and (C) quantification of cleaved CASP3+ apoptotic epithelial cells n = 6 mice/condition. (D) Thick-section images of HPIN stage cMyc prostate (left, top) and cMyc; Adrb2ecKO (left, bottom). TH = tyrosine hydroxylase, CD31 = vasculature. Scale bar = 100µm. Quantification of adrenergic nerve density (middle), and proximity of association between nerves and vessels (right). n = 4 mice/condition. (E) Quantification of vascular coverage by pericytes marked by NG2. n = 4 mice/condition. (F) Quantification of vascular leakage as assessed by accumulation of IV injected fluorescent dextran in the prostate interstitium. n = 4 mice/condition. (G) Tissue hypoxia assessed in prostate cross-sections by pimonidazole positive areas. n = 4 mice/condition. *p<0.05. **p<0.01. Error bars = s.e.m.

0

2

4

6

8

Hyp

oxic

are

a (%

of p

rost

ate

area

)

cMyc

cMyc; Adrb2cKO

0.0

0.5

1.0

1.5

2.0

Vasc

ular

leak

age

(Fol

d ch

ange

)

cMyc

cMyc; Adrb2cKO

Fh1

Coa6

Bckdha

0

2

4

6

Fold

regu

latio

n

cMyccMyc; Adrb2cKO*

*cMyccMyc; Adrb2ecKO

0

10

20

30

40TH

fibe

r len

gth

(mm

) /

DAP

I+ ar

ea (m

m2 )

cMyc

cMyc; Adrb2ecKO

0

2

4

6

Per

icyt

e co

vera

ge(%

of v

ascu

lar s

urfa

ce a

rea) cMyc

cMyc; Adrb2ecKO

0.0

0.2

0.4

0.6

0.8

1.0

Ner

ve–v

esse

l ass

ocia

tion

(pro

xim

ity s

core

)

cMyc

cMyc; Adrb2ecKO**

A

D

E F

0

5

10

15

Epi

thel

ial c

ells

(%

pro

lifer

atin

g)

*

0.0

0.2

0.4

0.6

0.8

1.0

# Ki

67+

cells

/ D

API+

are

a (µ

m2 )

cMyc

cMyc; Adrb2cKO*

B

G

C

TH CD31

cMyc

(8w

; HPI

N)

cMyc

; Adr

b2ec

KO

0

5

10

15

Epith

elia

l cel

ls

(% C

asp3

+ )

cMyccMyc; Adrb2ecKO

Epith

elia

l cel

ls

(% C

ASP3

+ )

12

Fig. S5. Adrb2 deletion in endothelial cells inhibits angiogenesis. (A) Experimental design of orthotopic type I collagen matrix vessel migration assay. (B) Immunofluorescence analysis of vascular migration into orthotopic matrices (outlined in purple) analyzed at HPIN-stage. Cross-sectional montage of cMyc ventral prostate (left, top); Montage of cMyc; Adrb2ecKO ventral prostate (right, top). CD31 = vasculature; Epcam = epithelium; SMA = smooth muscle actin. Scale bar = 500µm. Quantification of vessel migration and branching into orthotopic type I collagen matrix after endothelial Adrb2 deletion (bottom). n = 4 mice/condition. *p<0.05. Error bars = s.e.m.

cMyc cMyc; Adrb2ecKO

A

B

PIN

0 weeks4

TAM

8

Implant matrix

0

5

10

15

20

#Ves

sels

/ mat

rix a

rea

(mm

2 )

Vessel Index

*

Vecad

-

Vecad

+0

5

10

15

20

25#

Bra

nche

s/ m

atrix

are

a(m

m2 ) *

0

1

2

3

4

5

Tota

l ves

sel l

engt

h (m

m)

/ mat

rix a

rea

(mm

2 )

Vessel Length Index*

Fh1

Coa6

Bckdha

0

2

4

6

Fold

regu

latio

n

cMyccMyc; Adrb2cKO*

*cMyccMyc; Adrb2ecKO

CD31EpcamSMA

13

Fig. S6. Transcriptome analysis revealed Adrb2 deletion increases endothelial oxidative metabolism program in the HPIN-stage and is maintained overtime. (A-C) Experimental design and microarray analysis of FACS isolated endothelial cells from cMyc and cMyc; Adrb2ecKO HPIN-stage prostates. Experimental design (A). Enriched Gene Ontology (GO) terms in cMyc; Adrb2ecKO vs. cMyc obtained by GSEA (B). Heat map of differentially expressed genes, arranged by unsupervised hierarchical clustering. Upregulated mitochondrial-associated genes that clustered, are highlighted by red bracket. n = 3 mice/condition (C). (D) Quantitative PCR validation of cMyc; Adrb2ecKO up-regulated metabolic genes. n = 3 mice/condition. (E) Expression of the mitochondrial oxidative phosphorylation genes in FACS isolated prostate endothelial cells one year after recombination. n = 3 mice/condition. *p<0.05. **p<0.01. Error bars = s.e.m.

PIN

0 Weeks4

TAM

8

A B

C D

Fh1

Coa6

Bckdha

0

2

4

6

Fold

regu

latio

n

cMyccMyc; Adrb2cKO*

*

0 2 4 6

Carbon carbon lyase actMonovalent inorg cation tm transporter act

Protein tranporter actCarbon O2 lyase act

Exonuclease actH+ tm transporter actMicrotubule binding

Double stranded DNA bindingMicrotubule motor act

Cytochrome C act

-Ln (P value)

P < 0.05

Bckdha

Coa6

Fh1

cMyc Adrb2fl/fl Cdh5-CreERT2

cMyc cMyc; Adrb2cKO

Fh1

Coa6

Bckdha

0

2

4

6

Fold

regu

latio

n

cMyccMyc; Adrb2cKO*

*cMyccMyc; Adrb2ecKO

E

Prostate

DigestFACS sort

Bladder

Coa6 Bckdha0.00

0.02

0.04

0.06

0.08

Expr

essi

on / Gapdh

(One

yea

r pos

t rec

ombi

natio

n)

*

14

Fig. S7. β-adrenergic receptor deletion in the tumor microenvironment increases expression of mitochondrial metabolism genes in endothelial cells. (A-B) Quantitative PCR of mitochondrial metabolism genes in FACS-sorted endothelial cells isolated from βAdRWT and βAdRKO peri-tumoral prostate (A) and from within the orthotopic tumor (B). n = 5 mice/condition. *p<0.05. **p<0.01. ***p<0.001. Error bars = s.e.m.

A B

Coa6

Bckdha

0.00

0.02

0.04

0.06

0.2

0.3

0.4

Exp

ress

ion

/ Gapdh

Prostate peri-tumoral endothelium

**

***

Coa6

Bckdha

0.00

0.02

0.04

0.06

0.15

0.20

0.25

Exp

ress

ion

/ Gapdh

Prostate tumor endothelium

WTβ2-/- β3-/-

*

*

βadr

+/+;+/+

βadr

KO0

2000

4000

6000

8000

10000

Mito

chon

dria

l mas

s(M

itotr

acke

r M

FI)

βadr+/+;+/+

βadrKOβAdRWTβAdRKO

15

Fig. S8. A metabolic switch occurs during the transition from the pre-neoplastic LPIN to cancer-committed HPIN stage. (A) Schematic of disease progression. (B-D) FACS analysis of metabolism in freshly isolated prostate endothelial cells. Quantification of Δψ by TMRE staining (B), NADH redox state, the main electron donor in oxidative phosphorylation (C), and mitochondrial mass by Mitotracker green staining (D). n = 4-5 mice/condition. (E) Quantification of spare respiratory capacity (SRC) by FACS in freshly isolated bone marrow hematopoietic stem and progenitor cells (cKit+ Sca-1+ CD11b– B220– CD3– Gr-1– Ter119–) as well as prostate endothelial cells isolated from the same animals. n = 4 mice/condition. *p<0.05. **p<0.01. ***p<0.001. Error bars = s.e.m.

PIN

0 Weeks4 8HPINLPIN

A

0

1000

2000

3000

4000

5000

Mito

chon

dria

l Δψ

(TM

RE

MFI

)

WT (4w)

WT (8w)

**

0

20

40

60

80

100

NAD

H (%

redu

ced)

AC

BD

*

0

5000

10000

15000

20000

Mito

chon

dria

l mas

s (M

FI) AC

BD

B

C

E

D

0255075

100125150175200

Spar

e re

spira

tory

cap

acity

(%

of b

asel

ine

NAD

H) WT basal (8w)

WT SRC (8w)cMyc basal (8w; mPIN)

cMyc SRC (8w; mPIN)

*

* WT basal (8w)WT +FCCP (8w)cMyc basal (8w; HPIN)cMyc +FCCP (8w; HPIN)

WT (4

w)

WT (8

w)

cMyc

(4w)

cMyc

(8w)

0

5

10

15

Epi

thel

ial c

ells

(% p

rolif

erat

ing)

WT (4w)WT (8w)*

cMyc (4w)

cMyc (8w; mPIN)

WT (4w)WT (8w)cMyc (4w; LPIN)cMyc (8w; HPIN)

Hematopoietic stem and

progenitor cells

0255075

100125150175200

Spar

e re

spira

tory

cap

acity

(%

of b

asel

ine

NAD

H) WT basal (8w)

WT SRC (8w)cMyc basal (8w; mPIN)

cMyc SRC (8w; mPIN)

*

*

Spar

e re

spira

tory

capa

city

(% o

f bas

elin

e N

AD+ )

0

25

50

75

100

125

Spar

e re

spira

tory

cap

acity

(%

of b

asel

ine

NAD

H) WT basal (8w)

cMyc SRC (8w; mPIN)

WT SRC (8w)cMyc basal (8w; mPIN)

Endothelial cells

NS NS

Spar

e re

spira

tory

capa

city

(% o

f bas

elin

e N

AD+ )

16

Fig. S9. Adrb2 deletion does not affect endothelial viability or mitochondrial mass. (A-C) FACS analysis of endothelial cells isolated from HPIN-stage prostates. (A) Quantification of NADH redox state (left), n = 6-7 mice/condition; and mitochondrial mass by Mitotracker green staining (right), n = 4-5 mice/condition. (B) Representative cleaved CASP3+ apoptotic endothelial cell plot (left), and quantification (right). n = 7 mice/condition. (C) Quantification of Ki-67+ proliferative endothelial cells. n = 4 mice/condition. *p<0.05. Error bars = s.e.m.

0 101 102 103 104 105

0

101

102

103

104

105

0 102 103 104 105

0

101

102

103

104

105

LYVE

1

CASP3

Gated on: CD31+ CD45-

TER119-

A

B CAdrb2f/f Adrb2Δ/Δ

0

5000

10000

15000

Mito

chon

dria

l mas

s (M

FI)

0

20

40

60

80N

ADH

(% re

duce

d) VeCadCre+VeCadCre-

*

Fh1

Coa6

Bckdha

0

2

4

6

Fold

regu

latio

n

cMyccMyc; Adrb2cKO*

*cMyccMyc; Adrb2ecKO

0.0

0.2

0.4

0.6

0.8

1.0

CD

31+

Lyve

1- cel

ls(%

Cas

p3+ )

cMyccMyc; Adrb2cKO

CD

31+

LYVE

1- cel

ls(%

CAS

P3+ )

0

1

2

3

4

CD

31+

Lyve

1- cel

ls

(% p

rolif

erat

ing)

CD

31+

LYVE

1- cel

ls(%

pro

lifer

atin

g)

17

Fig. S10. β-adrenergic receptor deletion in the tumor microenvironment increases endothelial oxidative metabolism. (A) FACS analysis of Δψ by TMRE staining (left), NADH redox state (middle), and mitochondrial mass by Mitotracker green staining (right). n = 3 mice/condition. (B) Relative oxygen consumption rate in βAdRWT and βAdRKO primary tumor endothelial cells, freshly isolated by FACS sorting and incubated with noradrenaline for 36 h before analysis. n = 3 mice/condition. *p<0.05. **p<0.01. Error bars = s.e.m.

Glucoseuptakeorgraphicabouthowdidseahorse

A

B

WT

β2-/- β

3-/-

0.0

0.5

1.0

1.5

2.0

2.5

Oxy

gen

cons

umpt

ion

rate

(fold

cha

nge)

**

WT

Adrb2

-/- A

drb3

-/-

0

2000

4000

6000

8000

10000

Mito

chon

dria

l mas

s(M

itot

rack

er M

FI)

WTAdrb2 -/- Adrb3 -/-

βadr+/+;+/+

βadrKO

0

2000

4000

6000

8000

Mito

chon

dria

l Δψ

(TM

RE

MFI

)

*

βadr+/+;+/+

βadrKO

0

10

20

30

40

NAD

H (%

redu

ced)

*

βadr+/+

;+/+

βadrKO

0

2000

4000

6000

8000

10000

Mito

chon

dria

l mas

s(M

itot

rack

er M

FI)

βadr+/+;+/+

βadrKOβAdRWTβAdRKO

18

Fig. S11. Adrenergic denervation increases endothelial oxidative metabolism. (A) Experimental design (left), and representative images of PBS (middle), and 6OHDA chemically sympathectomized prostates (right). (B) FACS analysis of Δψ by TMRE staining (left), NADH redox state (middle), and mitochondrial mass by Mitotracker green staining (right). n = 4 mice/condition. **p<0.01. ****p<0.0001. Error bars = s.e.m.

TH CD31

0 days+2

6OHDA 100mg/kg i.p

+5

6OHDA 250mg/kg i.p

0

20

40

60

NAD

H (%

redu

ced)

**

0

5000

10000

15000

Mito

chon

dria

l mas

s (M

FI)

PBS 6OHDA (Denervated)

PBS

6OHDA

0

5

10

15

Tota

l ves

sel l

engt

h (m

m)

/ mat

rix a

rea

(mm

2 )

Vessel Length Index

*PBS6OHDA

A

B

0

5000

10000

15000

20000

Mito

chon

dira

l Δψ

(TM

RE

MFI

)

****

0

1000

2000

3000

4000

5000

Mito

chon

dria

l Δψ

(TM

RE

MFI

)

WT (4w)

WT (8w)

**

19

Fig. S12. Adrb2 knockdown is specific and does not affect cell viability. (A) Adrb2 expression (left) and non-targeted Adrb3 expression (right) assessed by quantitative PCR. n = 3 independent experiments. (B) FACS quantification of cleaved CASP3+ apoptotic cells (left) and quantification of Ki-67+

proliferative cells (right). n = 3 independent experiments. ****p<0.0001. Error bars = s.e.m.

0.000

0.002

0.004

0.006

0.008

0.010

Adrb2/Gapdh

shCntrlshAdrb2****

0.000

0.002

0.004

0.006

0.008

0.010Adrb2/Gapdh

shCntrlshAdrb2****

0.0

0.2

0.4

0.6

0.8

1.0

% P

rolif

erat

ing

shCntrl

shAdrb2

A

B

0.0000

0.0005

0.0010

0.0015

0.0020

0.0025

Adrb3/Gapdh

shCntrlshAdrb2

0.0

0.5

1.0

1.5

2.0

2.5

% C

asp3

+

shAdrb2

shCntrl

% C

ASP3

+

20

Fig. S13. Adrb2 knockdown increases oxidative phosphorylation metabolism. (A-C) FACS analysis of endothelial cell metabolism. Quantification of Δψ by TMRE staining (left), and NADH redox state (right) experiments (A). Quantification of mitochondrial mass by Mitotracker green staining (B), and glucose uptake by 2-NBDG (C). n = 3 independent experiments. (D) Quantification of extracellular lactate relative to shCntrl. n = 3 independent experiments. ***p<0.001. ****p<0.0001. Error bars = s.e.m.

0

1000

2000

3000

4000

Mito

chon

dira

l Δψ

(TM

RE

MFI

)

****

shAdrb2shControlshCntrlshAdrb2

A

B

D

C

0

20

40

60

80

NAD

H (%

redu

ced)

***

0

10000

20000

30000

40000

50000

Mito

chon

dria

l mas

s(M

itot

rack

er M

FI)

0

500

1000

1500

Glu

cose

(2-N

BD

G)

upta

ke (M

FI)

shControlshAdrb2***

0

50

100

150

Lact

ate

prod

uctio

n(%

of c

ontro

l)0

1000

2000

3000

4000

Mito

chon

dira

l Δψ

(TM

RE

MFI

)

****

shAdrb2

shControl

0

1000

2000

3000

4000

5000

Mito

chon

dria

l Δψ

(TM

RE

MFI

)

WT (4w)

WT (8w)

**

21

Fig. S14. Characterization of endothelial metabolism after Adrb2 knockdown. (A) Fraction of malate, fumarate, and aspartate containing three 13C-labeled carbons derived from [U-13C]-glucose. n = 3 replicates/ condition. (B-C) Panels B-C represent different analyses of the same set of experiments (n = 3 replicates/ condition. Metabolite levels were normalized to internal standard and to sample protein content). Relative intracellular levels of citrate labeling derived from [U-13C]-glutamine (B). Fraction of malate, fumarate, and aspartate containing four 13C-labeled carbons derived from [U-13C]-glutamine (C). (D) Schematic describing anaplerotic TCA metabolite labeling pattern derived from [U-13C]-glucose and [U-13C]-glutamine. (E) Metabolite profiling of acyl carnitine abundance. Dotted line represents shCntrl metabolite levels. n = 4 replicates/ conditions. *p<0.05. **p<0.01.***p<0.001. ****p<0.0001. Error bars = s.e.m.

A

C D

E

C2 C3 C4 C5C12

C12:1

C14C14

:1C14

:2C16

C16:1

C18C18

:10

5

10

15

Acy

l-car

nitin

e ab

unda

nce

(rel

ativ

e to

con

trol)

shCntrl

shAdrb2

*** *** ***

****

***

********

m+0 m+1 m+2 m+3 m+4 m+50

20

40

60

% o

f poo

l(in

50:

50 12

C/ 13

C m

edia

)

shCntrl

shAdrb2

B

shCntr

l

shAdb

rb2

0

5

10

15

Met

abol

ite le

vels

(a.u

.)

Citrate- from gln

From 13C-Gln

Unlabeled

12Carbon13Carbon (Glucose)13Carbon (Glutamine)

Pyruvate

Oxaloacetate

Aspartate(m+3; m+4)

Malate(m+3; m+4)

PC

CO2

Glucose

α-KG

Fumarate(m+3; m+4)

Citrate

Glutamine

CO2

Malate

Fumar

ate

Aspar

tate

0

10

20

30

40

m+4

(% o

f tot

al p

ool)

labe

led

by 13

C-G

ln

shCtrl

shAdrb2*** *** ***

shCntr

l

shAdb

rb2

0

5

10

15

Citr

ate

leve

ls (a

.u.)

Citrate- from gln

From 13C-Gln

Unlabeled

****

****

m+0 m+1 m+2 m+3 m+4 m+50

20

40

60

% o

f poo

l(in

50:

50 12

C/ 13

C m

edia

)

shCntrl

shAdrb2

m+0 m+1 m+2 m+3 m+4 m+50

20

40

60

% o

f poo

l(in

50:

50 12

C/ 13

C m

edia

)

shCntrl

shAdrb2

m+0 m+1 m+2 m+3 m+4 m+50

20

40

60

% o

f poo

l(in

50:

50 12

C/ 13

C m

edia

)

shCntrl

shAdrb2

m+0 m+1 m+2 m+3 m+4 m+50

20

40

60

% o

f poo

l(in

50:

50 1

2 C/ 1

3 C m

edia

)

shCntrl

shAdrb2

Malate

Fumar

ate

Aspar

tate

0

5

10

15

20

25

m+3

(% o

f tot

al p

ool)

labe

led

by 13

C-G

lc

shCtrl

shAdrb2** * *

22

Fig. S15. ADRB2 regulates endothelial Coa6 expression via cAMP signaling. (A) Quantification of ATP levels in cells cultured in the presence of glucose or galactose. n = 3 independent experiments. (B) Quantitative PCR of mitochondrial metabolism genes. n = 3 independent experiments. (C) Western blot of the mitochondrial metabolism protein COA6. (D) Quantification of the second messenger cAMP in the presence of noradrenaline. n = 3 independent experiments. (E) The effect of adenylyl cyclase inhibition (by SQ 22536), or stimulation (by forskolin), on Coa6 expression in the presence of noradrenaline. Schematic of signaling pathway (left), and quantification of expression by quantitative PCR (right). *p<0.05. **p<0.01. ***p<0.001. ****p<0.0001. Error bars = s.e.m.

Coa6

Bckdh

a0.00

0.05

0.10

0.15

0.20

0.25

Exp

ress

ion

/ Gapdh

shCntrlshAdrb2

**

Mr (K)

42 β-ACTIN

10 COA6

shC

ntrl

shA

drb2

Vehicl

e

Forsk

olin

SQ 2253

6

Vehicl

e

Forsk

olin

SQ 2253

60.00

0.05

0.10

0.15

0.20

Coa6/Gapdh

shCntrlshAdrb2

***

*

A B

0

200

400

600

800

1000

cAM

P (p

M/ µ

g pr

otei

n)

shCntrl

shAdrb2

< 5.0

C

EForskolin

Adrb2

Gs-protein

cAMP

ATP

Coa6

SQ 22536

D

0

10

20

30

ATP

(pM

/cel

l)

shCntrl (5mM gluc)

shAdrb2 (5mM gluc)

shCntrl (5mM galactose)

shAdrb2 (5mM galactose)**

***

*******

Glucose (5mM)Galactose (5mM)

+ - + -- + - +

23

Fig. S16. Adrb2 knockdown inhibits endothelial cell migration. (A-D) Endothelial migration and microtubule organization in a scratch wound assay. Representative images of shCntrl and shAdrb2 at 0 hours after the scratch was made (top) and 36 hours later (bottom) (A). Immunofluorescent images of microtubule orientation in cells that migrated into the scratch. Scale bar = 50µm. (B). Quantification of endothelial migration (left) and microtubule orientation (right). n = 3 independent experiments. *p<0.05. ***p<0.001. Error bars = s.e.m.

B

α-tubulin Hoescht

0

50

100

150

Endo

thel

ial m

igra

tion

(% o

f con

trol)

shCntrl

shAdrb2*

36 h

our

0 ho

ur

shCntrl shAdrb2

0

50

100

150

End

othe

lial m

igra

tion

(% o

f con

trol)

shCntrl

shAdrb2*

A

C D

0

20

40

60

80

Cen

troso

me

orie

ntat

ion

tow

ard

scra

tch

(% o

f cel

ls) Cntrl-GFP

shAdrb2***

Mic

rotu

bule

orie

ntat

ion

tow

ard

scra

tch

(% o

f cel

ls)

24

Fig. S17. Coa6 expression increases endothelial oxidative phosphorylation. (A) Validation that Coa6 overexpression (Coa6-GFP) in primary prostate endothelial cells increases both Coa6 mRNA expression

by quantitative PCR (left) and COA6 protein levels by western blot (right). (B) FACS analysis of of Δψ

by TMRE staining (left), NADH redox state (middle), and mitochondrial mass (right). n = 3 independent

experiments. (C) Quantification of endothelial cell migration in a scratch wound assay. n = 3

independent experiments. (D) FACS quantification of Ki-67+ proliferating endothelial cells (left) and

cleaved CASP3+ apoptotic cells (right). n = 3 independent experiments. *p<0.05. **p<0.01.

****p<0.0001. Error bars = s.e.m.

A

B

C D

Cnt

rl-GFP

Coa

6-G

FP

Bckdh

a-GFP

0

10000

20000

30000

40000

50000

Mito

cho

nd

ria

l mass

(Mit

otr

ac k

er

MFI)

Cntrl-

GFP

Coa6-

GFP

Bckdh

a-GFP

0

20

40

60

80

NA

DH

(%

reduce

d)

**

Mr (K)

42 β-ACTIN

10 COA6

Cn

trl-

GF

P

Co

a6

-GF

P

0.0

0.2

0.4

0.6

0.8

% P

rolif

era

ti ng

Cntrl-GFP

Coa6-GFP

*

0

50

100

150

Endoth

elia

l mig

ratio

n

(% o

f co

ntr

ol)

Cntrl-GFP

Coa6-GFP

*

0.0

0.1

0.2

0.3

0.4

% C

asp

3+

Cntrl-GFP

Coa6-GFP

% C

AS

P3

+

Cnt

rl -G

FPCoa

6-G

FP

-G

FP

0

2000

4000

6000

Mito

chondiral D

y(T

MR

E M

FI)

******

0

1000

2000

3000

4000

5000

Mito

chondrial

(TM

RE

MF

I )

WT (4w)

WT (8w)

**

Cnt

rl-GFP

Coa

6-GFP

Bckdh

a-GFP

0.0

0.1

0.2

0.3

Coa

6 /G

apdh

****

****

ΔΨ

Bckdh

a

25

Fig. S18. Cox10 co-deletion restores endothelial metabolism and cancer progression. (A-E) HPIN-stage analysis of cMyc, cMyc; Adrb2ecKO, cMyc; Cox10ecKO, and double cMyc; Adrb2ecKO; Cox10ecKO showing that deletion of both Adrb2 and Cox10 restores glucose uptake (A; n = 6-7 mice/condition), and NADH redox ratio (B; n = 6-7 mice/condition), without affecting mitochondrial mass (C; 6-7 mice/condition), and restores epithelial proliferation (D; n = 6-7 mice/condition), without affecting apoptosis (E; n = 6-7 mice/condition). *p<0.05. **p<0.01. ***p<0.001. ****p<0.0001. Error bars = s.e.m.

A B

C D

E

0

2000

4000

6000

Mito

chon

dria

l Δψ

(TM

RE

MFI

)

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

*** cMyccMyc; Adrb2ecKO

cMyc; Cox10ecKO

cMyc; Adrb2ecKO; Cox10ecKO

0

5000

10000

15000

Mito

chon

dria

l mas

s (M

FI) cMyc

cMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

0

200

400

600

800

1000

Glu

cose

(2-N

BDG

)up

take

(MFI

)

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

********

***

0

20

40

60

80

NAD

H (%

redu

ced)

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

********

****

0

5

10

15

20

Epith

elia

l cel

ls

(% p

rolif

erat

ing)

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO**

*****

0

2

4

6

8

Epith

elia

l cel

ls

(% C

asp3

+ )

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

Epith

elia

l cel

ls(%

CAS

P3+ )

26

Fig. S19. Cox10 co-deletion restores prostate perfusion and angiogenesis. (A) Schematic summarizing luciferin-based prostate perfusion assay (left) and quantification of prostate perfusion (right). n = 4 mice/condition. (B) Quantification of adrenergic nerve density (left), and proximity of association between nerves and vessels (right). n = 4 mice/condition. (C) Quantification of vessel migration, density, and branching into orthotopic type I collagen matrices after endothelial Adrb2, Cox10, or Adrb2 + Cox10 deletion. n = 4 mice/condition. *p<0.05. **p<0.01. Error bars = s.e.m.

0

5000

10000

15000

20000

Pro

stat

e pe

rfusi

on (R

LU/m

g) cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

****

*

0

20

40

60

TH fi

ber l

engt

h (m

m) /

D

API+

area

(mm

2 )

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

0

2000

4000

6000

Mito

chon

dria

l Δψ

(TM

RE

MFI

)

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

*** cMyccMyc; Adrb2ecKO

cMyc; Cox10ecKO

cMyc; Adrb2ecKO; Cox10ecKO

0.0

0.2

0.4

0.6

0.8

1.0

Ner

ve–v

esse

l ass

ocia

tion

(pro

xim

ity s

core

)

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

****

*

A

B

C

Inject luciferin IVDissect prostate

Digest prostateand pellet

Isolatesupernatant

Add luciferase +

ATP and measure

luminescence

0

5

10

15

20To

tal v

esse

l len

gth

(mm

)/ m

atrix

are

a (m

m2 )

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO**

**

0

10

20

30

40

#Ves

sels

/ mat

rix a

rea

(mm

2 )

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

***

*

0

20

40

60

80

#Bra

nche

s/ m

atrix

are

a (m

m2 )

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

*

*

*

Prostate

Luciferin containingfraction

Bladder

27

Fig. S20. Cox10 co-deletion restores epithelial proliferation, but does not affect endothelial proliferation or viability. (A-E) Immunofluorescence analysis of epithelial and endothelial proliferation and apoptosis. Thin cross-sectional images of cMyc, cMyc; Adrb2ecKO, cMyc; Cox10ecKO, and cMyc; Adrb2ecKO; Cox10ecKO prostates. Scale bar = 100µm (A). Quantification of Ki-67+ proliferating epithelial cells (B), TUNEL+ apoptotic epithelial cells (C), Ki67+ proliferating endothelial cells (D), and TUNEL+ apoptotic endothelial cells (E). *p<0.05. **p<0.01. Error bars = s.e.m.

A

B

0

2000

4000

6000

Mito

chon

dria

l Δψ

(TM

RE

MFI

)

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

*** cMyccMyc; Adrb2ecKO

cMyc; Cox10ecKO

cMyc; Adrb2ecKO; Cox10ecKO

0.0

0.2

0.4

0.6

Epi

thel

ial a

popt

otic

inde

x (%

TU

NE

L+ )

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

C

D E

0.0

0.1

0.2

0.3

0.4

End

othe

lial a

popt

otic

inde

x (%

TU

NE

L+ )

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

cMyc cMyc; Adrb2ecKOcMyc; Adrb2ecKO;

Cox10ecKOcMyc; Cox10ecKO

CD31 Epcam Ki-67

0

5

10

15

20

25

Epi

thel

ial p

rolif

erat

ion

inde

x (%

Ki6

7+ )

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

***

*

Epith

elia

l pro

lifer

atio

n in

dex

(% K

i-67+ )

0

5

10

15

End

othe

lial p

rolif

erat

ion

inde

x (%

Ki6

7+ )

cMyccMyc; Adrb2cKO

cMyc; Cox10cKO

cMyc; Adrb2cKO; Cox10cKO

Endo

thel

ial p

rolif

erat

ion

inde

x(%

Ki-6

7+ )

28

Table S1. Isotopologue distribution of citrate derived from [U-13C]-glutamine

Fraction labeled by 13C-Gln Isotopologue of Citrate shCntrl shAdrb2 p-value

n = 3 n = 3 m+0 0.55 0.51 <0.0001 m+1 0.11 0.09 <0.0001 m+2 0.11 0.10 0.0004 m+3 0.04 0.07 <0.0001 m+4 0.14 0.16 <0.0001 m+5 0.02 0.07 <0.0001 m+6 0.01 0.00 0.0097

References and Notes

1. J. Folkman, K. Watson, D. Ingber, D. Hanahan, Induction of angiogenesis during the

transition from hyperplasia to neoplasia. Nature 339, 58–61 (1989).

doi:10.1038/339058a0 Medline

2. G. Bergers, L. E. Benjamin, Tumorigenesis and the angiogenic switch. Nat. Rev. Cancer 3,

401–410 (2003). doi:10.1038/nrc1093 Medline

3. Y. S. Mukouyama, D. Shin, S. Britsch, M. Taniguchi, D. J. Anderson, Sensory nerves

determine the pattern of arterial differentiation and blood vessel branching in the skin.

Cell 109, 693–705 (2002). doi:10.1016/S0092-8674(02)00757-2 Medline

4. W. Li, H. Kohara, Y. Uchida, J. M. James, K. Soneji, D. G. Cronshaw, Y.-R. Zou, T.

Nagasawa, Y. S. Mukouyama, Peripheral nerve-derived CXCL12 and VEGF-A regulate

the patterning of arterial vessel branching in developing limb skin. Dev. Cell 24, 359–371

(2013). doi:10.1016/j.devcel.2013.01.009 Medline

5. A. J. Ekstrand, R. Cao, M. Bjorndahl, S. Nystrom, A.-C. Jonsson-Rylander, H. Hassani, B.

Hallberg, M. Nordlander, Y. Cao, Deletion of neuropeptide Y (NPY) 2 receptor in mice

results in blockage of NPY-induced angiogenesis and delayed wound healing. Proc. Natl.

Acad. Sci. U.S.A. 100, 6033–6038 (2003). doi:10.1073/pnas.1135965100 Medline

6. P. Martin, Wound healing—Aiming for perfect skin regeneration. Science 276, 75–81 (1997).

doi:10.1126/science.276.5309.75 Medline

7. S. W. Cole, A. S. Nagaraja, S. K. Lutgendorf, P. A. Green, A. K. Sood, Sympathetic nervous

system regulation of the tumour microenvironment. Nat. Rev. Cancer 15, 563–572

(2015). doi:10.1038/nrc3978 Medline

8. D. He, A. Manzoni, D. Florentin, W. Fisher, Y. Ding, M. Lee, G. Ayala, Biologic effect of

neurogenesis in pancreatic cancer. Hum. Pathol. 52, 182–189 (2016).

doi:10.1016/j.humpath.2016.02.001 Medline

9. C. Magnon, S. J. Hall, J. Lin, X. Xue, L. Gerber, S. J. Freedland, P. S. Frenette, Autonomic

nerve development contributes to prostate cancer progression. Science 341, 1236361

(2013). doi:10.1126/science.1236361 Medline

10. J. L. Saloman, K. M. Albers, D. Li, D. J. Hartman, H. C. Crawford, E. A. Muha, A. D. Rhim,

B. M. Davis, Ablation of sensory neurons in a genetic model of pancreatic ductal

adenocarcinoma slows initiation and progression of cancer. Proc. Natl. Acad. Sci. U.S.A.

113, 3078–3083 (2016). doi:10.1073/pnas.1512603113 Medline

11. C. M. Zhao, Y. Hayakawa, Y. Kodama, S. Muthupalani, C. B. Westphalen, G. T. Andersen,

A. Flatberg, H. Johannessen, R. A. Friedman, B. W. Renz, A. K. Sandvik, V. Beisvag, H.

Tomita, A. Hara, M. Quante, Z. Li, M. D. Gershon, K. Kaneko, J. G. Fox, T. C. Wang, D.

Chen, Denervation suppresses gastric tumorigenesis. Sci. Transl. Med. 6, 250ra115

(2014). doi:10.1126/scitranslmed.3009569 Medline

12. M. H. Longair, D. A. Baker, J. D. Armstrong, Simple Neurite Tracer: Open source software

for reconstruction, visualization and analysis of neuronal processes. Bioinformatics 27,

2453–2454 (2011). doi:10.1093/bioinformatics/btr390 Medline

13. T. Twardowski, A. Fertala, J. P. Orgel, J. D. San Antonio, Type I collagen and collagen

mimetics as angiogenesis promoting superpolymers. Curr. Pharm. Des. 13, 3608–3621

(2007). doi:10.2174/138161207782794176 Medline

14. J. A. Tuxhorn, G. E. Ayala, M. J. Smith, V. C. Smith, T. D. Dang, D. R. Rowley, Reactive

stroma in human prostate cancer: Induction of myofibroblast phenotype and extracellular

matrix remodeling. Clin. Cancer Res. 8, 2912–2923 (2002). Medline

15. N. Burns-Cox, N. C. Avery, J. C. Gingell, A. J. Bailey, Changes in collagen metabolism in

prostate cancer: A host response that may alter progression. J. Urol. 166, 1698–1701

(2001). doi:10.1016/S0022-5347(05)65656-X Medline

16. W. J. Huss, C. F. Hanrahan, R. J. Barrios, J. W. Simons, N. M. Greenberg, Angiogenesis and

prostate cancer: Identification of a molecular progression switch. Cancer Res. 61, 2736–

2743 (2001). Medline

17. K. Ellwood-Yen, T. G. Graeber, J. Wongvipat, M. L. Iruela-Arispe, J. Zhang, R. Matusik, G.

V. Thomas, C. L. Sawyers, Myc-driven murine prostate cancer shares molecular features

with human prostate tumors. Cancer Cell 4, 223–238 (2003). doi:10.1016/S1535-

6108(03)00197-1 Medline

18. M. Ittmann, J. Huang, E. Radaelli, P. Martin, S. Signoretti, R. Sullivan, B. W. Simons, J. M.

Ward, B. D. Robinson, G. C. Chu, M. Loda, G. Thomas, A. Borowsky, R. D. Cardiff,

Animal models of human prostate cancer: The consensus report of the New York meeting

of the Mouse Models of Human Cancers Consortium Prostate Pathology Committee.

Cancer Res. 73, 2718–2736 (2013). doi:10.1158/0008-5472.CAN-12-4213 Medline

19. T. Iwata, D. Schultz, J. Hicks, G. K. Hubbard, L. N. Mutton, T. L. Lotan, C. Bethel, M. T.

Lotz, S. Yegnasubramanian, W. G. Nelson, C. V. Dang, M. Xu, U. Anele, C. M. Koh, C.

J. Bieberich, A. M. De Marzo, MYC overexpression induces prostatic intraepithelial

neoplasia and loss of Nkx3.1 in mouse luminal epithelial cells. PLOS ONE 5, e9427

(2010). doi:10.1371/journal.pone.0009427 Medline

20. P. Carmeliet, R. K. Jain, Molecular mechanisms and clinical applications of angiogenesis.

Nature 473, 298–307 (2011). doi:10.1038/nature10144 Medline

21. E. Hinoi, N. Gao, D. Y. Jung, V. Yadav, T. Yoshizawa, M. G. Myers Jr., S. C. Chua Jr., J. K.

Kim, K. H. Kaestner, G. Karsenty, The sympathetic tone mediates leptin’s inhibition of

insulin secretion by modulating osteocalcin bioactivity. J. Cell Biol. 183, 1235–1242

(2008). doi:10.1083/jcb.200809113 Medline

22. L. Deng, J.-F. Zhou, R. S. Sellers, J.-F. Li, A. V. Nguyen, Y. Wang, A. Orlofsky, Q. Liu, D.

A. Hume, J. W. Pollard, L. Augenlicht, E. Y. Lin, A novel mouse model of inflammatory

bowel disease links mammalian target of rapamycin-dependent hyperproliferation of

colonic epithelium to inflammation-associated tumorigenesis. Am. J. Pathol. 176, 952–

967 (2010). doi:10.2353/ajpath.2010.090622 Medline

23. X. Zhu, R. A. Hill, D. Dietrich, M. Komitova, R. Suzuki, A. Nishiyama, Age-dependent fate

and lineage restriction of single NG2 cells. Development 138, 745–753 (2011).

doi:10.1242/dev.047951 Medline

24. I. Sörensen, R. H. Adams, A. Gossler, DLL1-mediated Notch activation regulates endothelial

identity in mouse fetal arteries. Blood 113, 5680–5688 (2009). doi:10.1182/blood-2008-

08-174508 Medline

25. K. De Bock, M. Georgiadou, S. Schoors, A. Kuchnio, B. W. Wong, A. R. Cantelmo, A.

Quaegebeur, B. Ghesquière, S. Cauwenberghs, G. Eelen, L.-K. Phng, I. Betz, B.

Tembuyser, K. Brepoels, J. Welti, I. Geudens, I. Segura, B. Cruys, F. Bifari, I. Decimo,

R. Blanco, S. Wyns, J. Vangindertael, S. Rocha, R. T. Collins, S. Munck, D. Daelemans,

H. Imamura, R. Devlieger, M. Rider, P. P. Van Veldhoven, F. Schuit, R. Bartrons, J.

Hofkens, P. Fraisl, S. Telang, R. J. Deberardinis, L. Schoonjans, S. Vinckier, J. Chesney,

H. Gerhardt, M. Dewerchin, P. Carmeliet, Role of PFKFB3-driven glycolysis in vessel

sprouting. Cell 154, 651–663 (2013). doi:10.1016/j.cell.2013.06.037 Medline

26. K. Wilhelm, K. Happel, G. Eelen, S. Schoors, M. F. Oellerich, R. Lim, B. Zimmermann, I.

M. Aspalter, C. A. Franco, T. Boettger, T. Braun, M. Fruttiger, K. Rajewsky, C. Keller, J.

C. Brüning, H. Gerhardt, P. Carmeliet, M. Potente, FOXO1 couples metabolic activity

and growth state in the vascular endothelium. Nature 529, 216–220 (2016).

doi:10.1038/nature16498 Medline

27. Y. H. Wang, W. J. Israelsen, D. Lee, V. W. C. Yu, N. T. Jeanson, C. B. Clish, L. C. Cantley,

M. G. Vander Heiden, D. T. Scadden, Cell-state-specific metabolic dependency in

hematopoiesis and leukemogenesis. Cell 158, 1309–1323 (2014).

doi:10.1016/j.cell.2014.07.048 Medline

28. F. Kocabas, J. Zheng, C. Zhang, H. A. Sadek, Metabolic characterization of hematopoietic

stem cells. Methods Mol. Biol. 1185, 155–164 (2014). doi:10.1007/978-1-4939-1133-

2_10 Medline

29. M. Maryanovich, Y. Zaltsman, A. Ruggiero, A. Goldman, L. Shachnai, S. L. Zaidman, Z.

Porat, K. Golan, T. Lapidot, A. Gross, An MTCH2 pathway repressing mitochondria

metabolism regulates haematopoietic stem cell fate. Nat. Commun. 6, 7901 (2015).

doi:10.1038/ncomms8901 Medline

30. A. Krützfeldt, R. Spahr, S. Mertens, B. Siegmund, H. M. Piper, Metabolism of exogenous

substrates by coronary endothelial cells in culture. J. Mol. Cell. Cardiol. 22, 1393–1404

(1990). doi:10.1016/0022-2828(90)90984-A Medline

31. M. Seandel, J. M. Butler, H. Kobayashi, A. T. Hooper, I. A. White, F. Zhang, E. L. Vertes,

M. Kobayashi, Y. Zhang, S. V. Shmelkov, N. R. Hackett, S. Rabbany, J. L. Boyer, S.

Rafii, Generation of a functional and durable vascular niche by the adenoviral E4ORF1

gene. Proc. Natl. Acad. Sci. U.S.A. 105, 19288–19293 (2008).

doi:10.1073/pnas.0805980105 Medline

32. T. Cheng, J. Sudderth, C. Yang, A. R. Mullen, E. S. Jin, J. M. Matés, R. J. DeBerardinis,

Pyruvate carboxylase is required for glutamine-independent growth of tumor cells. Proc.

Natl. Acad. Sci. U.S.A. 108, 8674–8679 (2011). doi:10.1073/pnas.1016627108 Medline

33. J. Munger, B. D. Bennett, A. Parikh, X.-J. Feng, J. McArdle, H. A. Rabitz, T. Shenk, J. D.

Rabinowitz, Systems-level metabolic flux profiling identifies fatty acid synthesis as a

target for antiviral therapy. Nat. Biotechnol. 26, 1179–1186 (2008). doi:10.1038/nbt.1500

Medline

34. K. A. Hotchkiss, A. W. Ashton, R. Mahmood, R. G. Russell, J. A. Sparano, E. L. Schwartz,

Inhibition of endothelial cell function in vitro and angiogenesis in vivo by docetaxel

(Taxotere): Association with impaired repositioning of the microtubule organizing center.

Mol. Cancer Ther. 1, 1191–1200 (2002). Medline

35. R. Suzuki, K. Hotta, K. Oka, Spatiotemporal quantification of subcellular ATP levels in a

single HeLa cell during changes in morphology. Sci. Rep. 5, 16874 (2015).

doi:10.1038/srep16874 Medline

36. A. Ghosh, P. P. Trivedi, S. A. Timbalia, A. T. Griffin, J. J. Rahn, S. S. L. Chan, V. M. Gohil,

Copper supplementation restores cytochrome c oxidase assembly defect in a

mitochondrial disease model of COA6 deficiency. Hum. Mol. Genet. 23, 3596–3606

(2014). doi:10.1093/hmg/ddu069 Medline

37. F. Diaz, C. K. Thomas, S. Garcia, D. Hernandez, C. T. Moraes, Mice lacking COX10 in

skeletal muscle recapitulate the phenotype of progressive mitochondrial myopathies

associated with cytochrome c oxidase deficiency. Hum. Mol. Genet. 14, 2737–2748

(2005). doi:10.1093/hmg/ddi307 Medline

38. M. Potente, P. Carmeliet, The link between angiogenesis and endothelial metabolism. Annu.

Rev. Physiol. 79, 43–66 (2017). doi:10.1146/annurev-physiol-021115-105134 Medline

39. S. Liu, Y. Soong, S. V. Seshan, H. H. Szeto, Novel cardiolipin therapeutic protects

endothelial mitochondria during renal ischemia and mitigates microvascular rarefaction,

inflammation, and fibrosis. Am. J. Physiol. Renal Physiol. 306, F970–F980 (2014).

doi:10.1152/ajprenal.00697.2013 Medline

40. S. Schoors, K. De Bock, A. R. Cantelmo, M. Georgiadou, B. Ghesquière, S. Cauwenberghs,

A. Kuchnio, B. W. Wong, A. Quaegebeur, J. Goveia, F. Bifari, X. Wang, R. Blanco, B.

Tembuyser, I. Cornelissen, A. Bouché, S. Vinckier, S. Diaz-Moralli, H. Gerhardt, S.

Telang, M. Cascante, J. Chesney, M. Dewerchin, P. Carmeliet, Partial and transient

reduction of glycolysis by PFKFB3 blockade reduces pathological angiogenesis. Cell

Metab. 19, 37–48 (2014). doi:10.1016/j.cmet.2013.11.008 Medline

41. A. Eichmann, I. Brunet, Arterial innervation in development and disease. Sci. Transl. Med. 6,

252ps9 (2014). doi:10.1126/scitranslmed.3008910 Medline

42. J. B. Aragon-Ching, R. A. Madan, W. L. Dahut, Angiogenesis inhibition in prostate cancer:

Current uses and future promises. J. Oncol. 2010, 361836 (2010).

doi:10.1155/2010/361836 Medline

43. G. Bergers, D. Hanahan, Modes of resistance to anti-angiogenic therapy. Nat. Rev. Cancer 8,

592–603 (2008). doi:10.1038/nrc2442 Medline

44. J. Kohl, A. D. Ostrovsky, S. Frechter, G. S. Jefferis, A bidirectional circuit switch reroutes

pheromone signals in male and female brains. Cell 155, 1610–1623 (2013).

doi:10.1016/j.cell.2013.11.025 Medline

45. M. Costa, J. D. Manton, A. D. Ostrovsky, S. Prohaska, G. S. Jefferis, NBLAST: Rapid,

sensitive comparison of neuronal structure and construction of neuron family databases.

Neuron 91, 293–311 (2016). doi:10.1016/j.neuron.2016.06.012 Medline

46. B. W. Carey, L. W. Finley, J. R. Cross, C. D. Allis, C. B. Thompson, Intracellular α-

ketoglutarate maintains the pluripotency of embryonic stem cells. Nature 518, 413–416

(2015). doi:10.1038/nature13981 Medline

47. T. Kind, G. Wohlgemuth, D. Y. Lee, Y. Lu, M. Palazoglu, S. Shahbaz, O. Fiehn, FiehnLib:

Mass spectral and retention index libraries for metabolomics based on quadrupole and

time-of-flight gas chromatography/mass spectrometry. Anal. Chem. 81, 10038–10048

(2009). doi:10.1021/ac9019522 Medline

48. P. Millard, F. Letisse, S. Sokol, J. C. Portais, IsoCor: Correcting MS data in isotope labeling

experiments. Bioinformatics 28, 1294–1296 (2012). doi:10.1093/bioinformatics/bts127

Medline

31

42. J. B. Aragon-Ching, R. A. Madan, W. L. Dahut, Angiogenesis inhibition in prostate cancer: current uses and future promises. Journal of oncology 2010, 361836 (2010).

43. G. Bergers, D. Hanahan, Modes of resistance to anti-angiogenic therapy. Nature reviews. Cancer 8, 592-603 (2008).

44. J. Kohl, A. D. Ostrovsky, S. Frechter, G. S. Jefferis, A bidirectional circuit switch reroutes pheromone signals in male and female brains. Cell 155, 1610-1623 (2013).

45. M. Costa, J. D. Manton, A. D. Ostrovsky, S. Prohaska, G. S. Jefferis, NBLAST: Rapid, Sensitive Comparison of Neuronal Structure and Construction of Neuron Family Databases. Neuron 91, 293-311 (2016).

46. B. W. Carey, L. W. Finley, J. R. Cross, C. D. Allis, C. B. Thompson, Intracellular alpha-ketoglutarate maintains the pluripotency of embryonic stem cells. Nature 518, 413-416 (2015).

47. T. Kind et al., FiehnLib: mass spectral and retention index libraries for metabolomics based on quadrupole and time-of-flight gas chromatography/mass spectrometry. Analytical chemistry 81, 10038-10048 (2009).

48. P. Millard, F. Letisse, S. Sokol, J. C. Portais, IsoCor: correcting MS data in isotope labeling experiments. Bioinformatics (Oxford, England) 28, 1294-1296 (2012).