www.sciencemag.org/cgi/content/full/341/6141/84/DC1

Supplementary Materials for

Monitoring Drug Target Engagement in Cells and Tissues Using the Cellular Thermal Shift Assay

Daniel Martinez Molina, Rozbeh Jafari, Marina Ignatushchenko, Takahiro Seki, E. Andreas Larsson, Chen Dan, Lekshmy Sreekumar, Yihai Cao, Pär Nordlund*

*Corresponding author. E-mail: Par.Nordlund@ki.se

Published 5 July 2013, Science 341, 84 (2013) DOI: 10.1126/science.1233606

This PDF file includes:

Materials and Methods Supplementary Text Figs. S1 to S4 Table S1 References

2

Materials and Methods.

Chemicals and Buffers. TNP-470, inhibitor of MetAP2, was from Takeda

Chemicals (Osaka, Japan). Vemurafenib (PLX4032), SB590885 and AZ628, inhibitors of

BRAF; AZD5438, inhibitor of CDK2 and CDK9; PD0332991, inhibitor of CDK4 and

CDK6; Olaparib (AZD2281) and Iniparib (BSI-201), inhibitors of PARP-1 were from

Selleck Chemicals (Houston, TX, USA). TS inhibitors Raltitrexed (Tomudex) and 5-FU

(5-Fluorouracil); DHFR inhibitor Methotrexate (Abitrexate); transport inhibitors Suramin

and S-(4-Nitrobenzyl)-6-thioinosine (NBMPR); and the TS substrate 2′-Deoxyuridine 5′-

monophosphate (dUMP) were from Sigma-Aldrich (St. Louis, MO, USA). All

abovementioned chemicals were dissolved and diluted using dimethyl sulfoxide (DMSO)

except dUMP that was dissolved and diluted in ddH2O. Phosphate-buffered saline (PBS)

was prepared using 10 mM phosphate buffer (pH 7.4), 2.7 mM potassium chloride and

137 mM sodium chloride. Kinase buffer (KB) (25 mM Tris(hydroxymethyl)-

aminomethane hydrochloride (Tris-HCl, pH 7.5), 5 mM beta-glycerophosphate, 2 mM

dithiothreitol (DTT), 0.1 mM sodium vanadium oxide, 10 mM magnesium chloride) was

from Cell Signaling Technology (Beverly, MA, USA). Tris-Buffered Saline with Tween

(TBST) buffer (150 mM NaCl, 0.05% (v/v) Tween-20, 50 mM Tris-HCl buffer (pH 7.6))

was prepared by dissolving TBS-TWEEN tablets obtained from Merck KGaA

(Darmstadt, Germany) in ddH2O. Blocking buffer was 5% (w/v) non-fat milk (Semper

AB, Sundbyberg, Sweden) diluted in TBST. Complete (EDTA-free) protease inhibitor

cocktail was from Roche (Switzerland). Isothermal calorimetry buffer (ITC buffer) (1%

DMSO, dUMP 100 µM, 20 mM Hepes pH 7.5 and 150 mM NaCl) was prepared prior to

experiments.

3

Heterologous protein expression. Genes (P26-V313 of human TS and K662-

T1011 of human PARP1) were subcloned into the pNIC28-Bsa4 vector (GenBankTM

accession number EF198106), yielding an expression construct with an N-terminal

hexahistidine tag and a TEV protease recognition site. The positive recombinant clone

was retransformed and expressed in T1 phage-resistant BL21(DE3) E. coli strain

(Merck). For expression, cells were grown at 37°C in a LEX system using 0.75 L of

Terrific Broth medium supplemented with 8 g/L glycerol, 50 µg/mL kanamycin, and 34

µg/mL chloramphenicol. When OD600 reached ca. 2, the temperature was reduced to

18°C. After 30–60 minutes, the expression of the target protein was induced by addition

of 0.5 mm isopropyl β-d-thiogalactopyranoside and incubation for 17–20 h. The cells

were harvested by centrifugation and resuspended in lysis buffer (100 mM HEPES, 500

mM NaCl, 10 mM imidazole, 10% (v/v) glycerol, 0.5 mM TCEP, pH 8.0) supplemented

with Protease Inhibitor Mixture Set III, EDTA free (Merck) and 2000 units of benzonase

(Merck), and stored at −80°C. Cells were disrupted by sonication on ice using Vibra-Cell

processor (Sonics & Materials Inc., Newtown, CT, USA). The lysate was clarified by

centrifugation at 47,000 × g for 25 min at 4°C, and the supernatant was filtered through a

1.2-µm syringe filter. Filtered lysates were loaded onto 1 mL of nickel-nitrilotriacetic

acid Superflow resin (Qiagen Inc., Valencia, CA, USA) in IMAC wash buffer 1 (20 mM

HEPES, 500 mM NaCl, 10 mM imidazole, 10% (v/v) glycerol, 0.5 mM TCEP, pH 7.5)

and washed with IMAC wash buffer 2 (20 mM HEPES, 500 mM NaCl, 25 mM

imidazole, 10% (v/v) glycerol, 0.5 mM TCEP, pH 7.5). Bound proteins were eluted with

500 mM imidazole and loaded onto a HiLoad 16/60 Superdex-200 column (GE

4

Healthcare, Waukesha, WI, USA) pre-equilibrated with equilibration buffer (20 mM

HEPES, 300 mM NaCl, 10% (v/v) glycerol, 0.5 mM TCEP, pH 7.5). Fractions containing

the protein of interest were pooled. TCEP was added to a final concentration of 2 mM,

and the sample was concentrated using Vivaspin 20 filter concentrators (15 kDa MW

cutoff) (GE Healthcare) at 15°C. The final protein concentration and yield was 15.5

mg/mL, 1.4 mg (PARP-1) and 20.6 mg/mL, 30 mg (TS). The protein batches were then

aliquoted, frozen in liquid nitrogen, and stored at −80°C.

Cell culture and in vitro experiments. Human cancer cell lines K562 (ATCC

No. CCL-243) and A549 (ATTC No. CCL-185) were cultured in RPMI-1640 medium

(Sigma-Aldrich); A375 cells (ATCC No. CRL-1619) were maintained in Dulbecco’s

Modified Eagle Medium (Sigma-Aldrich), and HEK-293 cells (ATCC No. CRL-1573)

were maintained in Eagle's Minimum Essential Medium (Sigma-Aldrich). All culture

media were supplemented with 0.3 g/L L-glutamine and 10% fetal bovine serum (FBS,

Gibco/Life Technologies, Carlsbad, CA, USA), 100 units/mL penicillin and 100 units/mL

streptomycin (Gibco/Life Technologies). Short-term passages (<15) were used for in

vitro cell experiments. The cell line K562 was used for the CDKs, TS and DHFR

experiments; A549 for MetAP2 experiments; HEK293 for PARP-1 and TS experiments.

A375 (BRAF V600E) and K562 (BRAF wild-type) were used for BRAF experiments

(Table I). Equal numbers of cells (0.6-1.0x106 cells per data point) were seeded in T-25

cell culture flasks (BD Biosciences, San Jose, CA, USA) or 12-well cell culture plates

(Corning Inc., Corning, NY, USA) in appropriate volume of culture medium and exposed

to a drug for 3 hours in an incubator chamber (with 5% CO2) (Memmert GmbH,

Schwabach, Germany). Control cells were incubated with an equal volume of diluent for

5

the corresponding drug. For drug concentrations see Supplementary Table I. Following

the incubation the cells were harvested (either directly or detached from the surface using

Trypsin/EDTA solution (Sigma-Aldrich)) and washed with PBS in order to remove

excess drug. Equal amounts of cell suspensions were aliquoted into 0.2 mL PCR

microtubes, and excess PBS was removed by centrifugation to leave 10 uL or less PBS in

each microtube. These cell pellets were used for CETSA as described below.

For transport inhibition experiments equal numbers of K562 cells (0.6x106 cells

per data point) were seeded in 12-well cell culture plates in appropriate volume of culture

medium and pre-incubated with a transport inhibitor (Suramin or NBMPR) for 30

minutes in an incubator chamber. The appropriate inhibitor concentrations were

determined in preliminary CETSA experiments (data not shown). Control cells were

incubated with an equal volume of DMSO. The cells were then exposed to varying

concentrations of an appropriate drug (Methotrexate or 5-FU, respectively) for 3 hours in

an incubator chamber. Following the incubation the drug-containing media were removed

by centrifugation; the cells were harvested, washed with PBS and prepared for CETSA as

described below.

For time-course experiments equal numbers of K562 cells (0.6x106 cells per data

point) were seeded in T-25 cell culture flasks and exposed to varying concentrations of

Raltitrexed. Cell culture aliquots were removed at specified times; the cells were washed

with PBS and prepared for CETSA as described below.

For re-feeding experiments equal numbers of K562 cells (0.6x106 cells per data

point) were seeded in 12-well cell culture plates in appropriate volume of culture medium

and exposed to varying concentrations of Raltitrexed for 10 minutes, 30 minutes, or 3

6

hours in an incubator chamber. Following the incubation the drug-containing media were

removed by centrifugation; the cells were harvested, washed with PBS and prepared for

CETSA as described above. The removed media were used to resuspend freshly pelleted

non-treated K562 cells (0.6x106 cells per data point). The cell suspensions were

transferred to fresh 12-well cell culture plates and incubated for additional 3 hours. The

cells were harvested, washed and prepared for CETSA as described below.

Trypan blue dye exclusion. Trypan blue dye exclusion was used to evaluate the

integrity of cell membranes in the heat treated cells. Approximately 0.6x106 cells from

different cell lines were heated at different temperatures for 3 minutes and allowed to

cool to RT for 3 minutes. A 10µL cell aliquot from each sample was briefly mixed with

an equal volume of 0.4% (w/v) trypan blue dye solution (Bio-Rad) and analyzed using a

TC20™ automated cell counter (Bio-Rad). Cells with the ability to exclude trypan blue

were considered retaining their cell membrane integrity whereas dye stained cells were

identified as cells with lost cell membrane integrity.

In vivo mice experiment. TNP-470 was dissolved in 100% ethanol, followed by

dilution in PBS containing 5% ethanol. The drug solution was prepared immediately

before use. For the double-blind study, 6-7 week old female C57Bl/6 mice

(Microbiology, Tumor and Cell Biology Animal Facility, Karolinska Institute,

Stockholm, Sweden) were injected subcutaneously with TNP-470 (0 - 20 mg/kg, 0.1 ml)

at the mid-dorsal location. Control animals were injected with the appropriate amount of

ethanol in PBS. All mice were kept in separate cages. After 4 hours, mice were

euthanized; liver and kidneys were removed and stored on dry ice immediately. All

7

animal procedures were carried out in accordance with the Northern Stockholm

Experimental Animal Ethical Committee approved protocols.

Cellular Thermal Shift Assay (CETSA). For the cell lysate CETSA

experiments, cultured cells from abovementioned cell lines were harvested and washed

with PBS. The cells were diluted in KB for BRAF and CDK´s and in PBS for MetAP2,

TS, DHFR and PARP-1. All buffers were supplemented with Complete protease inhibitor

cocktail. The cell suspensions were freeze-thawed three times using liquid nitrogen. The

soluble fraction (lysate) was separated from the cell debris by centrifugation at 20000 x g

for 20 minutes at 4°C. The cell lysates were diluted with appropriate buffer and divided

into two aliquots, with one aliquot being treated with drug (Supplementary Table 1) and

the other aliquot with the diluent of the inhibitor (control). After 10-30 minute incubation

at room temperature the respective lysates were divided into smaller (50µL) aliquots and

heated individually at different temperatures for 3 minutes (Veriti thermal cycler, Applied

Biosystems/Life Technologies) followed by cooling for 3 minutes at room temperature.

The appropriate temperatures were determined in preliminary CETSA experiments (data

not shown). The heated lysates were centrifuged at 20000 x g for 20 minutes at 4°C in

order to separate the soluble fractions from precipitates. The supernatants were

transferred to new microtubes and analyzed by sodium dodecyl sulfate polyacrylamide

gel electrophoresis (SDS-PAGE) followed by western blot analysis. For a schematic

procedure of CETSA see Supplementary Figure S1.

For cell lysate experiments carried out on TS, the dUMP concentration was kept

constant at 100 µM as a binding co-factor to Raltitrexed. Raltitrexed was added from

8

DMSO stocks to the final concentration of 100µM and DMSO concentration 1%. Control

samples were incubated with an equal amount of DMSO.

For cell lysate experiments carried out on PARP1, Iniparib and Olaparib were

added from DMSO stocks to the final concentration of 100µM and DMSO concentration

1%. Control samples were incubated with an equal amount of DMSO.

For the intact cell experiments the drug-treated cells from the in vitro experiments

above were heated as previously described followed by addition of KB (30µL) and lysed

using 2 cycles of freeze-thawing with liquid nitrogen. The soluble fractions were isolated

and analyzed by western blot as described above.

For the in vivo mice experiments, lysates of frozen tissues were used. Since TNP-

470 binds covalently to MetAP2, the drug dissociation from the target after cell lysis

would be limited; the target engagement will therefore be similar to that in the tissue

before lysis. The frozen organs (i.e. liver and kidneys) were thawed on ice and briefly

rinsed with PBS. The organs were homogenized in cold PBS using tissue grinders

followed by 3 cycles of freeze-thawing using liquid nitrogen. Tissue lysates were

separated from the cellular debris and lipids as mentioned above. The tissue lysates were

diluted with PBS containing protease inhibitors, divided into 50µL aliquots and heated at

different temperatures. Soluble fractions were isolated as previously described and

analyzed by western blot.

SDS-PAGE and western blot. NuPage® Novex Bis-Tris 4-12% polyacrylamide

gels with NuPAGE® MES SDS running buffer (Life Technologies) were used for

separation of proteins in the samples. Proteins were transferred to nitrocellulose

membranes using the iBlot® blotting system (Life Technologies) and to polyvinylidene

9

difluoride (PVDF) membranes using Trans-Blot® Turbo™ (Bio-Rad, Hercules, CA,

USA). Primary antibodies anti-MetAP2 (sc-365637), DHFR (sc-81844, sc-377091), TS

(sc-33679), BRAF (sc-9002), CDK2, (sc-6248), CDK4 (sc-601), CDK6 (sc-53638),

CDK9 (sc-13130), PARP-1 (sc-8007), β-actin (sc-69879); secondary goat anti-mouse

HRP-IgG (sc-2055) and bovine anti-rabbit HRP-IgG (sc-2374) antibodies (Santa Cruz

Biotechnology, Santa Cruz, CA, USA) were used for immunoblotting. Rabbit

monoclonal anti-MetAP2 (RabMAb, Epitomics, Burlingame, CA, USA) was used for

western blots for the samples generated from the in vivo mice experiments. All

membranes were blocked with blocking buffer; standard transfer and western blot

protocols recommended by the manufacturers (listed above) were used. All antibodies

were diluted in blocking buffer. The membranes were developed using SuperSignal West

Dura Chemiluminescent HRP-Substrate (Thermo Scientific) according to the

manufacturer’s recommendations. Chemiluminescence intensities were detected and

quantified using a ChemiDoc™ XRS+ imaging system (Bio-Rad) with Image Lab™

software (Bio-Rad). β-actin levels were used to normalize the intensities of the in vitro

cell and in vivo mice experiments. For the CETSA curves the band intensities were

related to the intensities of the lowest temperature for the drug exposed samples and

control samples, respectively. For the ITDRFCETSA experiments the band intensities were

related to control samples.

Statistical analysis. All lysate, in vitro cell and in vivo mice CETSA data were

expressed as means ± SEM. The data (band intensities) from multiple runs (n ≥ 3) were

plotted using Graphpad Prism 6.0 software (GraphPad Software Inc., La Jolla, CA,

USA). The ITDRFCETSA data were fitted using a sigmoidal (variable slope) curve fit.

10

DSLS-experiments. Temperature-dependent aggregation was measured by using

static light scattering (StarGazer, Harbinger Biotechnology) as previously described (12)

at a constant protein concentration of (0.2 mg/mL) using a ramp rate of 1°C/min. Graphs

were plotted with Graphpad and fitted using Sigmoidal dose-response (variable slope).

CETSA-like dot-blot experiments on purified proteins. Purified protein (0.5

µg) was added to the wells of a PCR plate and the volume adjusted to 50 µL by addition

of buffer or cell lysates and ligands depending on the experimental setup. The samples

were heated for the designated time and temperature in a Veriti thermocycler (Applied

Biosystems/ Life Technologies). After heating, the samples were immediately

centrifuged for 15 min at 3000 x g and filtered using a 0.65µm Multiscreen HTS 96 well

filter plate (Merck). 3 µL of each filtrate was blotted onto a nitrocellulose membrane.

Primary antibody Penta-His Antibody BSA-free (Qiagen) and secondary goat-anti-mouse

IgG/HRP conjugate (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA,

USA) were used for immunoblotting. All membranes were blocked with blocking buffer;

standard transfer and western blot protocols recommended by the manufacturers were

used. All antibodies were diluted in blocking buffer. The dot-blot was developed using

SuperSignal West Dura chemiluminescence kit (Thermo Scientific). Chemiluminescence

intensities were detected and imaged using a Fujifilm LAS-3000 imaging system

(Fujifilm, Tokyo, Japan). Raw dot blot images were processed using ImageJ (37). The

background was subtracted and intensities quantified using the MicroArray Profile

(http://image.bio.methods.free.fr/dotblot.html) plugin. Graphs were plotted with

Graphpad software and fitted using sigmoidal dose-response (variable slope).

11

Isothermal titration calorimetry (ITC) of TS. ITC titrations on TS were

performed with Raltitrexed at two different temperatures. Protein and Raltitrexed were

transferred to ITC buffer. The Raltitrexed concentration was 400 µM and protein

concentration in the reaction chamber was 62.5 µM and 61.1 µM for the 25°C and 37°C

experiments, respectively.

12

Supplementary text

Discussion on validation data and theoretical background of CETSA.

Cell and cell membrane integrity: Trypan blue is a dye widely used to

determine integrity of mammalian cells. The blue dye only enters and stains cells when

the cell membrane integrity is lost and pores are formed whereas cells with intact

membranes do not take up this dye and remain unstained when viewed under a

microscope. In order to investigate the integrity of the cells and cell membranes after

heating, a series of dye exclusion experiments were conducted. Cell aliquots of different

cancer cell lines were heated to different temperatures, cooled, mixed with a trypan blue

solution and the cell number and dye exclusion (i.e cell membrane integrity) were

measured. The data demonstrated that for the cell lines studied, most cells are not lysed

even at high temperatures (65°-70°C) (Fig. S2) since the number of cells remains

unchanged compared to the control cells incubated at 37°C. These data confirm that the

cell membranes remain intact at temperatures up to 60-65°C depending on cell line (Fig.

S2).

Correlation of CETSA with TSA on purified proteins and lysate effects.

To determine whether the data obtained by CETSA correlate with data obtained by

conventional TSA, we overexpressed and purified TS and the poly (ADP-ribose)

polymerase domain of PARP-1. We subsequently compared the thermal melt behavior of

the purified proteins using the CETSA protocol to differential static light scattering

(DSLS) data obtained using Stargazer instrument, routinely used for TSA of purified

proteins (12). CETSA of purified PARP-1 with cognate inhibitors replicated the essential

13

features of the DSLS experiment, including the approximate melting temperature of the

ligand-free protein and the shift sizes induced by ligand binding (Fig. S3A).

To allow direct comparison of thermal behavior of a heterologously expressed vs.

endogenous protein from mammalian cells, we performed experiments on purified

polyHis-tagged TS. We obtained DSLS melting curves of the purified TS, as well as its

CETSA profiles upon addition of high concentrations of the folate analog Raltitrexed.

We also performed a CETSA assay on a sample containing an E. coli lysate added to the

purified protein and on a HEK-293 cell lysate containing endogenous TS. The substrate

dUMP is required for Raltitrexed binding and was therefore added at constant

concentration with the purified protein (100µM). Comparative analysis of these two

experiments yielded similar melting temperatures for the unliganded protein as well as

significant shifts for all the proteins upon addition of the compound. The data also

confirmed that addition of the lysate did not have a substantial effect on the experimental

outcome (Fig. S3B).

To shed further light on the lysate effect, we performed ITDRFCETSA on BRAF in

A375 cell lysates. Three cognate inhibitors were used, and each experiment was

performed in a high lysate concentration as well as a low, 10-fold diluted, lysate (Fig.

S3C). Very small effects were caused by two of the inhibitors (AZ628, SB590885), while

PLX4032 was apparently less effective at the higher lysate concentration than at the

lower one, which could be due to protein binding or metabolic factors at the higher lysate

concentration.

Monitoring of parameters in antifolate transport and activation. As for TS

with Raltitrexed (Fig. 2C), a similar shift of ITDRFCETSA was obtained for DHFR with

14

Methotrexate (Fig. S4A). A time-course experiment with Raltitrexed indicated that after

2-3 hours, the ITDRFCETSA of TS was saturated (Fig. 2D); this saturation was not due to

depletion of the drug in the medium (Fig. S4B). By contrast, a similar experiment using

starved cells resulted in a limited accumulation of the drug (Fig. S4C).

Theoretical background for CETSA. Thermal shift assays with purified

proteins are widely used to monitor maximum shifts at high compound concentrations, as

the melting temperature (Tm) correlates to affinities to the target proteins (13, 14). Dose-

response experiments have only been shown to directly yield absolute binding constants

in the specific cases when a fully reversible system can be established (38, 39). Dose-

response dependent shifts start above Tm for the ideally behaving systems. However,

many proteins precipitate upon unfolding; hence the system is not reversible. When

precipitation following unfolding is used for detection, as in DSLS experiments and

CETSA, the systems are clearly not in equilibrium. However the determined aggregation

temperatures at high concentrations of compound, Tagg, are often close to Tm and can be

used as a relative measure to rank affinities (38, 39). When the system is not in

equilibrium, dose-response curves based on Tagg or Tm measurements will be complex

and based on multiple parameters. In these cases the dose-response curves will be

smoother; significantly higher concentrations of compounds will be needed to achieve a

significant response.

To exemplify the temperature dependence of binding constants for one of the targets, we

performed Isothermal calorimetry titrations on TS with Raltitrexed at two different

temperatures. The ITC data was fitted to give a Kd at 25°C and 37°C to 25.8 and 38 nM

respectively. By extrapolation via the integrated van’t Hoff relation, assuming a constant

15

Cp, the Kd at 50°C is estimated to be below 100 nM. Ki values for TS has been reported

to be in the range of 90-1000 nM using different activity assays at 30-37°C (Pubchem;

Raltitrexed CID104758).

Due to the complications discussed above, in the present work we do not attempt to

determine absolute affinities, but instead use an isothermal dose-response curve as a

fingerprint of the target engagement along a range of drug concentrations, ITDRFCETSA.

These experiments are informative as long as a reasonable constant set of measurement

parameters is used (cell or lysate condition, temperature and time), and can be utilized to

compare concentration effects of two different experiments. For example, when it can be

assumed that the target protein is in the same activation state in the two experiments, i.e.,

have the same affinity to the drug, the observed difference in concentration between two

ITDRFCETSA will directly correspond to the difference in the concentration of the added

compound required to establish the same local effective concentration of compound at the

drug target in the lysate or cell. In cases where only the activation state of the protein is

changed, ITDRFCETSA will reflect changes in affinity which will correlate with absolute

affinities. Due to the higher temperatures of the experiment and the continuous loss of

protein due to precipitation during the heating stage, this value will need further de-

convolution to allow better determination of drug affinity at 37°C. We envisage that

detailed studies of the CETSA behavior of a specific protein in lysates at different

conditions, as well as correlation with other data such as IC50 values obtained in lysates

or for purified proteins, could potentially allow direct correlation of CETSA affinities to

absolute affinities in cells. In practice, during drug development programs, the relative

16

ranking of the behavior of compounds within compound series is the main goal, and

ITDRFCETSA should constitute a valuable tool for such prioritizations.

Although CETSA data in many aspects is highly related to the data obtained from

purified proteins (as discussed above), the melting curves in cells and lysates are more

dispersed than for purified proteins. The exact causes for the topology of these curves for

each protein will differ, but it is likely that some of the topology features reflect the

presence of multiple forms of the protein in the cell. Detailed CETSA studies might

therefore allow the convolution of the behavior of different isoforms or activation states

of specific proteins in the cell.

Finally, there is room for significant improvements in the quantification and collection of

CETSA data. To allow rapid establishment of assays for many proteins, and to verify that

the right protein is probed, we used western blots in the present feasibility study.

However, western blots are prone to significant quantification errors. When fully

developed, high-quality and ultra-sensitive immune techniques (40) could assist in

generating improved data whereby smaller samples could be used (currently.

ca. 106 cells). In the present study we used a relatively long temperature exposure time (3

minutes) to make sure that the temperature is equilibrated in the entire sample and that

there is no doubt that the proteins have had time to unfold and aggregate. However, as

most proteins unfold in the second or sub-second time range (41, 42), it should be

possible to shorten the temperature stage significantly in order to minimize other

potential global cellular effects in the cell of the heating stage.

Considering the tissues experiment, which were made after lysis, they are also highly

feasibly to make directly by heating tissue aliquots, to get in cell data for non-covalent

17

binders. However, the establishment of high reproducibility of such experiments requires

some technical developments, which we felt was outside the scope of this initial

feasibility study.

18

Fig. S1.

Figure S1. Schematic illustration of CETSA melt curve and ITDRFCETSA

procedure.

19

Fig. S2.

Figure S2. Cell count and cell membrane integrity. Cell count and dye

exclusion by four different cell lines after heating to different temperatures. Data

presented as mean ± SEM, n ≥ 3.

20

Fig S3.

Figure S3. Correlation of CETSA with TSA on purified proteins and lysate

concentration effects. A) Comparison of thermal shifts (apparent Tm) of

21

purified PARP-1 domain derived from the DSLS light scattering experiment and

the CETSA process. The top graph shows an overlay of melting curves of the

catalytic domain of PARP-1. Two experiments on purified protein were performed

in the same buffer using Stargazer (solid lines) and a CETSA-type denaturation

experiment (symbols and dashed lines) with a set of PARP-1 inhibitors. These

data show that the drop in signal in the centrifugation experiments follows the

increase of Stargazer signal (which monitors the light scattering of the

aggregates formed upon precipitation). B) Correlation of CETSA thermal shifts of

purified proteins in buffer, with lysate added, and in human HEK-293 cells. Bar

graph showing a comparison of TS TAgg data with and without 100uM of

Raltitrexed from Stargazer and CETSA-type denaturation experiments in buffer,

E. coli lysate and HEK-293 lysate. C) ITDRFCETSA at 52°C for BRAF with different

inhibitors in high lysate concentration and in lysate diluted 10 times with buffer

(“Low”) (n=3).

22

Fig S4.

23

Figure S4. Monitoring of parameters in antifolate transport and activation.

A) ITDRFCETSA at 52°C of DHFR with Methotrexate in intact cells vs. lysate (n=4).

B) Saturation of ITDRFCETSA at 52°C of TS. Responses after 10, 30 and 180

minutes of incubation with Raltitrexed. C) The media from these cells were added

to fresh cells followed by incubation for 180 minutes. The response of TS in the

cell shows that Raltitrexed is not depleted in the media. D) ITDRFCETSA at 52°C of

TS in cells grown under starvation conditions (day 2 and 3) indicate decreased

24

transport of Raltitrexed (n=4).

Table S1. Targets, drugs, drug concentrations and cell lines used in this study. ITDRFCETSA Melt curve

Cell lines used

Temperature °C Drug concentration

Target/Drug Lysate Intact Cell

Tissue lysate Lysate

Intact Cell

Tissue lysate

BRAF

A375/K562 PLX4032 52/56 56

100µM -

SB590885 52/56 52

100µM - AZ628 52/56 56

100µM -

CDK2/4/6/9

K562 PD0332991 - 45

100µM 10µM

AZD-5348 52 -

- -

TS

HEK-293/ K562

Raltitrexed 52 52

100µM 10µM Methotrexate - 52

- 10µM

5-FU - 52

100µM 150µM MetAP2

A549

TNP-470 72 - 72 1µM 10µM 0-

20mg/kg DHFR

K562

Methotrexate 52 52

100µM 10µM PARP-1

HEK-293

Olaparib 50 -

100µM - Iniparib 50 -

100µM -

Supplementary Table 1. Temperatures of the isothermal dose-response fingerprint

(ITDRFCETSA) and the drug concentrations used in the saturated CETSA curve

experiments. CDK: Cyclin dependent kinase; BRAF: v-Raf murine sarcoma viral

oncogene homolog B1; MetAP2: methionine aminopeptidase-2; DHFR: dihydrofolate

reductase; TS: thymidylate synthase; PARP-1: Poly [ADP-ribose] polymerase 1. 5-FU: 5-

Fluorouracil.

25

References and Notes

1. K. I. Kaitin, Deconstructing the drug development process: The new face of innovation. Clin. Pharmacol. Ther. 87, 356 (2010).doi:10.1038/clpt.2009.293 Medline

2. J. Orloff et al., The future of drug development: Advancing clinical trial design. Nat. Rev. Drug Discov. 8, 949 (2009). Medline

3. T. A. Yap, S. K. Sandhu, P. Workman, J. S. de Bono, Envisioning the future of early anticancer drug development. Nat. Rev. Cancer 10, 514 (2010).doi:10.1038/nrc2870 Medline

4. M. Rask-Andersen, M. S. Almén, H. B. Schiöth, Trends in the exploitation of novel drug targets. Nat. Rev. Drug Discov. 10, 579 (2011).doi:10.1038/nrd3478 Medline

5. A. Ruiz-Garcia, M. Bermejo, A. Moss, V. G. Casabo, Pharmacokinetics in drug discovery. J. Pharm. Sci. 97, 654 (2008).doi:10.1002/jps.21009 Medline

6. J. P. Gillet, M. M. Gottesman, Mechanisms of multidrug resistance in cancer. Methods Mol. Biol. 596, 47 (2010).doi:10.1007/978-1-60761-416-6_4 Medline

7. R. Barouch-Bentov, K. Sauer, Mechanisms of drug resistance in kinases. Expert Opin. Investig. Drugs 20, 153 (2011).doi:10.1517/13543784.2011.546344 Medline

8. D. S. Auld, N. Thorne, W. F. Maguire, J. Inglese, Mechanism of PTC124 activity in cell-based luciferase assays of nonsense codon suppression. Proc. Natl. Acad. Sci. U.S.A. 106, 3585 (2009).doi:10.1073/pnas.0813345106 Medline

9. C. Schmidt, GSK/Sirtris compounds dogged by assay artifacts. Nat. Biotechnol. 28, 185 (2010).doi:10.1038/nbt0310-185 Medline

10. M. Guha, PARP inhibitors stumble in breast cancer. Nat. Biotechnol. 29, 373 (2011).doi:10.1038/nbt0511-373 Medline

11. B. I. Kurganov, Kinetics of protein aggregation. Quantitative estimation of the chaperone-like activity in test-systems based on suppression of protein aggregation. Biochemistry (Mosc.) 67, 409 (2002).doi:10.1023/A:1015277805345 Medline

12. M. Vedadi et al., Chemical screening methods to identify ligands that promote protein stability, protein crystallization, and structure determination. Proc. Natl. Acad. Sci. U.S.A. 103, 15835 (2006).doi:10.1073/pnas.0605224103 Medline

13. E. Wahlberg et al., Family-wide chemical profiling and structural analysis of PARP and tankyrase inhibitors. Nat. Biotechnol. 30, 283 (2012).doi:10.1038/nbt.2121 Medline

14. O. Fedorov et al., A systematic interaction map of validated kinase inhibitors with Ser/Thr kinases. Proc. Natl. Acad. Sci. U.S.A. 104, 20523 (2007).doi:10.1073/pnas.0708800104 Medline

15. S. Marsh, Thymidylate synthase pharmacogenetics. Invest. New Drugs 23, 533 (2005).doi:10.1007/s10637-005-4021-7 Medline

16. Y. G. Assaraf, Molecular basis of antifolate resistance. Cancer Metastasis Rev. 26, 153 (2007).doi:10.1007/s10555-007-9049-z Medline

17. M. Visentin, R. Zhao, I. D. Goldman, The antifolates. Hematol. Oncol. Clin. North Am. 26, 629, ix (2012).doi:10.1016/j.hoc.2012.02.002 Medline

18. N. Hagner, M. Joerger, Cancer chemotherapy: Targeting folic acid synthesis. Cancer Manag. Res. 2, 293 (2010). Medline

19. M. Huang, Y. Wang, S. B. Cogut, B. S. Mitchell, L. M. Graves, Inhibition of nucleoside transport by protein kinase inhibitors. J. Pharmacol. Exp. Ther. 304, 753 (2003).doi:10.1124/jpet.102.044214 Medline

20. S. Q. Yin, J. J. Wang, C. M. Zhang, Z. P. Liu, The development of MetAP-2 inhibitors in cancer treatment. Curr. Med. Chem. 19, 1021 (2012).doi:10.2174/092986712799320709 Medline

21. J. D. Moore et al., Phase I dose escalation pharmacokinetics of O-(chloroacetylcarbamoyl) fumagillol (TNP-470) and its metabolites in AIDS patients with Kaposi’s sarcoma. Cancer Chemother. Pharmacol. 46, 173 (2000).doi:10.1007/s002800000149 Medline

22. E. Lounkine et al., Large-scale prediction and testing of drug activity on side-effect targets. Nature 486, 361 (2012). Medline

23. J. Cicenas, M. Valius, The CDK inhibitors in cancer research and therapy. J. Cancer Res. Clin. Oncol. 137, 1409 (2011).doi:10.1007/s00432-011-1039-4 Medline

24. M. Malumbres, M. Barbacid, Cell cycle, CDKs and cancer: A changing paradigm. Nat. Rev. Cancer 9, 153 (2009).doi:10.1038/nrc2602 Medline

25. D. W. Fry et al., Specific inhibition of cyclin-dependent kinase 4/6 by PD 0332991 and associated antitumor activity in human tumor xenografts. Mol. Cancer Ther. 3, 1427 (2004). Medline

26. H. Davies et al., Mutations of the BRAF gene in human cancer. Nature 417, 949 (2002).doi:10.1038/nature00766 Medline

27. J. T. Lee et al., PLX4032, a potent inhibitor of the B-Raf V600E oncogene, selectively inhibits V600E-positive melanomas. Pigment Cell Melanoma Res. 23, 820 (2010).doi:10.1111/j.1755-148X.2010.00763.x Medline

28. G. Bollag et al., Clinical efficacy of a RAF inhibitor needs broad target blockade in BRAF-mutant melanoma. Nature 467, 596 (2010).doi:10.1038/nature09454 Medline

29. X. Liu et al., Iniparib nonselectively modifies cysteine-containing proteins in tumor cells and is not a bona fide PARP inhibitor. Clin. Cancer Res. 18, 510 (2012).doi:10.1158/1078-0432.CCR-11-1973 Medline

30. E. Dean et al., Phase I study to assess the safety and tolerability of olaparib in combination with bevacizumab in patients with advanced solid tumours. Br. J. Cancer 106, 468 (2012).doi:10.1038/bjc.2011.555 Medline

31. B. Lomenick, R. W. Olsen, J. Huang, Identification of direct protein targets of small molecules. ACS Chem. Biol. 6, 34 (2011).doi:10.1021/cb100294v Medline

32. J. Luo, N. L. Solimini, S. J. Elledge, Principles of cancer therapy: Oncogene and non-oncogene addiction. Cell 136, 823 (2009).doi:10.1016/j.cell.2009.02.024 Medline

33. G. P. Hussmann, K. J. Kellar, A new radioligand binding assay to measure the concentration of drugs in rodent brain ex vivo. J. Pharmacol. Exp. Ther. 343, 434 (2012).doi:10.1124/jpet.112.198069 Medline

34. B. Lomenick et al., Target identification using drug affinity responsive target stability (DARTS). Proc. Natl. Acad. Sci. U.S.A. 106, 21984 (2009).doi:10.1073/pnas.0910040106 Medline

35. G. M. West, L. Tang, M. C. Fitzgerald, Thermodynamic analysis of protein stability and ligand binding using a chemical modification- and mass spectrometry-based strategy. Anal. Chem. 80, 4175 (2008).doi:10.1021/ac702610a Medline

36. P. M. Matthews, E. A. Rabiner, J. Passchier, R. N. Gunn, Positron emission tomography molecular imaging for drug development. Br. J. Clin. Pharmacol. 73, 175 (2012).doi:10.1111/j.1365-2125.2011.04085.x Medline

37. C. A. Schneider, W. S. Rasband, K. W. Eliceiri, NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671 (2012).doi:10.1038/nmeth.2089 Medline

38. J. K. Kranz, C. Schalk-Hihi, Protein thermal shifts to identify low molecular weight fragments. Methods Enzymol. 493, 277 (2011).doi:10.1016/B978-0-12-381274-2.00011-X Medline

39. D. Matulis, J. K. Kranz, F. R. Salemme, M. J. Todd, Thermodynamic stability of carbonic anhydrase: Measurements of binding affinity and stoichiometry using ThermoFluor. Biochemistry 44, 5258 (2005).doi:10.1021/bi048135v Medline

40. M. Hammond, R. Y. Nong, O. Ericsson, K. Pardali, U. Landegren, Profiling cellular protein complexes by proximity ligation with dual tag microarray readout. PLoS ONE 7, e40405 (2012).doi:10.1371/journal.pone.0040405 Medline

41. A. R. Fersht, V. Daggett, Protein folding and unfolding at atomic resolution. Cell 108, 573 (2002).doi:10.1016/S0092-8674(02)00620-7 Medline

42. U. Mayor, C. M. Johnson, V. Daggett, A. R. Fersht, Protein folding and unfolding in microseconds to nanoseconds by experiment and simulation. Proc. Natl. Acad. Sci. U.S.A. 97, 13518 (2000).doi:10.1073/pnas.250473497 Medline