high-throughput precision drug screening … · acknowledgements this work was supported by funding...
TRANSCRIPT
ACKNOWLEDGEMENTSThis work was supported by funding for the CPMCRI Precision Medicine Program.
ABSTRACTMelanoma patients with B-Raf mutant tumors have been shown to respond to
treatment with B-Raf inhibitors, e.g., dabrafenib. However, over time they
become resistant to this treatment necessitating the need to discover
additional treatments to target the resistant tumor. Tumors can be treated
with a multitude of single drug or drug combinations, therefore, a platform is
needed to help guide treatment decisions for patients. Patient-derived tumor
xenograf (PDX) mouse models have been shown to have characteristics of the
patient’s original tumor. The patient-derived tumor cells (PDC) obtained from
the PDX models can be evaluated for response to antitumor agents.
Transcriptic’s cloud laboratory offers a versatile high-throughput screening
method, enabling scientists to quickly and precisely treat PDC lines against
multiple drugs. In this poster, we describe our methods for PDC drug screening
on Transcriptic’s Robotic Cloud Lab to identify a possible treatment option for
a dabrafenib resistant melanoma.
MATERIALS AND METHODS (cont . )arm moves the containers to and from the appropriate instruments without
human intervention. Drugs were supplied to Transcriptic at 2X their effective
concentrations in standard micro-1.5 tubes. These were transferred to a 384-
acoustic liquid handler source plate which was then moved to a multi-channel
liquid handler to serially dilute the drugs. This source plate was used to dose
384-well microplates containing PDC samples supplied by CPMC. Using a
simple csv well map, the acoustic liquid handler automatically transferred
each drug at each concentration from the source plate to the appropriate
location on each of the cell-containing 384-well microplates. The drugged cell
lines were then placed in an incubator for 72 hours, after which the CellTiter-
Glo assay was used to quantify cell viability/proliferation. Data from
luminescence reads were subsequently rendered graphically in Transcriptic’s
web interface (Fig. 3) and made available for download in .csv format for
further data analysis.
PDXMicrobiopsy
HTS (48 drugstargeting cancer)
PDC
Figure 1 Creating mouse avatars for melanoma at CPMC Research Institute.
INTRODUCTIONA 65 year old male with metastatic melanoma was enrolled in a clinical trial of
dabrafenib and ipilimumab, and progressed following an initial partial
response. Working with California Pacific Medical Center (CPMC)/Sutter cancer
surgeons and oncologists, we created PDX and PDC models, and confirmed
that the melanoma avatar MM-001 had a clinical and genetic profile similar to
the original tumor. To predict tumor sensitivity to targeted therapies, we used
the Transcriptic’s cloud laboratory to perform high throughput screening of
PDC samples against multiple concentrations of 48 antitumor agents
(currently being used in the clinic). The results were ranked to determine which
treatments(s) were most effective at reducing the growth of MM-001 cells.
MATERIALS AND METHODSPDX & PDC MODELS Patient tissue was digested using a collagenase /
hyaluronidase enzyme mix and 1x10 cells resuspended in Matrigel were
injected subcutaneously in a NSG mouse (Fig. 1). Once the PDX reached
1.5 cm in diameter, the tumor was harvested and a PDC was generated.
Cultures were grown as tumorspheres in ultra low attachment dishes
using DMEM and growth factors.
PHARMACOLOGY Using the Transcriptic platform (see below), PDC
were screened against a 6-point concentration-response curve of 48
drugs being used to target cancer in the clinic. After three days treatment
of PDC, the CellTiter-Glo assay was used to quantify cell viability/
proliferation. The resulting curves were fitted using non-linear regression
analysis with the program GraphPad Prism.
TRANSCRIPTIC PLATFORM A detailed protocol for dosing PDC was
submitted to and evaluated by Transcriptic. The protocol was then
translated into a python script to generate Autoprotocol, a data structure
used to instruct Transcriptic’s workcells (Fig. 2) to execute experiments.
Compatible plates and tubes are supplied to the workcell and a robotic
arm
6
H I G H - T H R O U G H P U T P R E C I S I O N D R U G S C R E E N I N G O F PAT I E N T- D E R I V E DT U M O R C E L LS O N T RA N S C R I PT I C ’ S R O B OT I C C LO U D LA BAlyssia Oh, Vanessa Biggers, Liliana Soroceanu, Sean D. McAllister, David De Semir, Mohammed Kashani-Sabet, Yin He1 12 2
RESULTS
Figure 4 PCR sequencing confirmed that PDC MM-001 has a BRAF V600E mutation similar to the patient’s tumor.
25
0
50
75
100
125
-5-67-8-9DRUG CONCENTRATION (LOG M)
VIA
BIL
ITY
(%)
Vemurafenib
25
0
50
75
100
125
-6-8 -7-9-10-11
DRUG CONCENTRATION (LOG M)
VIA
BIL
ITY
(%)
Trametinib
25
0
50
75
100
125
-5-7 -6-8-9-10DRUG CONCENTRATION (LOG M)
VIA
BIL
ITY
(%)
Selumetinib
25
0
50
75
100
125
-5-7 -6-8-9-10DRUG CONCENTRATION (LOG M)
VIA
BIL
ITY
(%)
Dabrafenib
Figure 5 Similar to the patient’s original tumor, melanoma avatar MM-001 is resistant to BRAF inhibitors.
Figure 6 Melanoma avatar MM-001 is sensitive to MEK inhibitors.
CONCLUSIONSHTS of PDC derived from a BRAF-resistant patient tumor identified
sensitivity to MEK inhibitors
•
Transcriptic’s Cloud Based Laboratory provided multiple advantages to
development of a HTS platform including:
•
The Cancer Avatar Project at CPMCRI is using a multimodal approach that
includes mouse avatars, genomics, high-throughput pharmacologic
screening, and informatics to reach the goal of fully individualized cancer
care through rapid prediction of likely response before treatment decisions
must be made.
•
Eliminating the need to develop an expensive in-house HTS platform1)
Flexibility and control of protocol development and implementation,
with the ability to control experiments from a web browser
2)
Reduce data cycle times by leveraging Transcriptic’s in-house expertise3)
CPMCRI 475 Brannan Street, Suite 220, San Francisco, CA 94107
1 1 1
1
2 TRANSCRIPTIC 3565 Haven Avenue #3, Menlo Park, CA 94025www.transcriptic.com
[email protected] / (650) 763-8432
Figure 2 Transcriptic Workcell.
Figure 3 Luminescence 384-well plate reads rendered graphically in Transcriptic’s web interface.