Siemens Research Paper FINAL

1

1. Abstract

The investigation carried out in this paper consisted of a pair of experiments designed to

answer the question, “Does the degree of mesenchymal phenotype affect the sensitivity of non-

small cell lung cancer cell lines to TBK1 inhibitors?” The hypothesis formulated was that if

different variants of a cell line were modified to express mesenchymal or epithelial

characteristics and were then treated with TBK1 inhibitors, then the cells that have a more

mesenchymal phenotype will show a greater sensitivity to TBK1 inhibitors than cells with a

more epithelial phenotype. Two experiments were carried out to test the hypothesis: a 96-well

plate drug response curve was generated to test the drugs’ ability to stop cellular proliferation

and a 6-well binary assay was used to test the drugs’ ability to kill off the cells. The data from

the 96-well plate was normalized twice and put into graph format, whereas the 6-well binary

assay plates were photographed and compared to determine the effects of the drugs on the cell’s

confluency. The experiments supported the hypothesis: in both experiments, data supported the

hypothesis that cells with a more mesenchymal phenotype are more sensitive to TBK1 inhibition

than are cells with a more epithelial phenotype. Further testing could involve rerunning the 6-

well assay and quantifying the amount of cells in each well, running a Western Blot to see the

protein levels of mesenchymal and epithelial indicators in the cell lines, and running tumor

studies in mice to test the efficacy of the drugs in vivo. The results from these experiments

indicate that TBK1 inhibition could be a viable option for targeting late-stage cancers as well as

mutant Ras cancers, though further experimentation is necessary to test the prospects of TBK1

inhibition as a method of combating said cancers.

2

2. Introduction

This experimental analysis seeks to look at kinase inhibition as a method of stopping

proliferation of and killing off cancer cells. This analysis utilized non-small cell lung cancer cell

lines and looked at the sensitivity of said cell lines to inhibition of TBK1, a kinase that is

involved in multiple cell processes and an integral downstream effector of the RalGEF pathway,

a downstream pathway of the oncogenic Ras protein (Cooper, et al., 2013) (Ou, et al., 2011). The

specific focus of this analysis is to determine the effect of a mesenchymal phenotype on the

sensitivity of non-small cell lung cancer cell (NSCLC) lines to TBK1 inhibitors. The question

this analysis seeks to answer is, “Does the degree of mesenchymal characteristics affect TBK1

sensitivity in non-small cell lung cancer cell lines?” The hypothesis before undertaking this

investigation is that if different variants of the same non-small cell lung cancer cell line are

treated with TBK1 inhibitors, then cell line variants displaying a more mesenchymal phenotype

will be more sensitive to TBK1 inhibition than variants of the same cell line that display a more

epithelial phenotype.

3. Background

Cancer is a prominent disease. It kills 8.2 million people every year and has 14 million

new cases every year (National Cancer Institute, 2015). Finding cures for the many different

types of cancer is a top research priority today. One of the main areas of research regarding

cancer treatment centers on the Ras oncogene and its mutated status. The Ras oncogene pathway

is part of signal transduction in cells, and it is involved in various cell processes, such as cell

cycle progression, growth, migration, apoptosis, and cell proliferation (Fernández-Medarde &

Santos, 2011). Cancers that have Ras mutations are some of the worst cancers that exist. Up to

30% of all cancers are Ras mutant (Fernández-Medarde & Santos, 2011). However, cancers with

a worse prognosis can have higher Ras-mutation rates: pancreatic cancer has a 7% survival rate

3

and up to 90% mutant Ras prevalence, lung cancer has a 17.4% survival rate for all cancers and

up to 33% mutant Ras prevalence in adenocarcinomas, and colon cancer has a 64.2% survival

rate and up to 50% mutant Ras prevalence (National Cancer Institute, 2012) (Bos, 1989). The

sheer abundance of mutant Ras cancers makes the Ras protein a prominent target for intervention

in cancers. The Ras oncogene, if able to be targeted, presents an untapped opportunity to treat

late-stage and advanced cancers, some of the most hopeless cases. However, this is much easier

than it sounds. The Ras protein itself has hardly any effective inhibitors, and its rapid mutability

make it much harder to target.

Instead of targeting the Ras protein directly, cell biologists are choosing to target the

various pathways of the Ras protein. There are three pathways regulated by the Ras protein: the

Ral/mitogen-activated protein cascade (MAPK) cascade, the phosphoinositide 3-kinase (PI3K)-

dependent phosphoinositide second messenger pathway, and the Ral guanine nucleotide

exchange factor (RalGEF)/ Ral GTPases cascade (Cooper, et al., 2013). Previous research led to

the targeting of the MAPK and PI3K pathways since they can be targeted by drugs working

alone or in combination with other drugs, but this approach is limited and these pathways have

rapidly developed resistance to targeting, leading to researchers looking at other opportunities to

target mutant Ras cancers. (Cooper, et al., 2013) This opportunity may arise in the third pathway,

the RalGEF pathway, which has not been explored as much (Cooper, et al., 2013). The RalGEF

pathway contains of RalGEFs, enzymes that activate RalA/B proteins by exchanging guanine

diphosphate for guanine triphosphate, activating the RalA/B proteins. The RalA/B proteins then

engages the exocyst complex at the RalB-Sec5 subcomplex, which then promotes the activation

of TANK-binding kinase 1 (TBK1) (Cooper, et al., 2013). TBK1 inhibition is the specific point

at which intervention efforts at the research lab were targeted. Specifically, two inhibitors were

4

used in the targeting efforts: BX795 and Compound II (CpII). BX795 is a commercially

available drug that inhibits TBK1 (Clark, et al., 2009). Compound II is a drug that was developed

at the research institution and also is a TBK1 inhibitor, but Compound II has a lower “off-target”

effect on other kinases, allowing it to target TBK1 more specifically (Ou, et al., 2011).

Both TBK1 and Compound II have been used in assays designed to check the sensitivity

of various cell lines to TBK1 inhibition. Unpublished data from assays run at the research

institution have shown various cell lines and their sensitivity to TBK1 inhibition. This assay

showed two cell lines which were on the sensitive end of the spectrum: H460 and HCC44, both

of which are K-Ras-mutant non-small cell lung cancer cell lines. Reverse phase protein assays

(RPPAs) ran at the lab showed the relative levels of various proteins present in the cells. Said

assays showed that H460 and HCC44 had lower levels of E-cadherin than cell lines on the TBK1

resistant side of the spectrum. HCC44 had lower levels of ß-catenin and increased levels of

ZEB1 protein, while H460 had moderate levels of both proteins. The levels of these proteins

function as indicators of a mesenchymal phenotype. A mesenchymal phenotype describes cells

that have lost attachment between one another and have changed their morphology from box-

shaped to spindle-like. This process is often associated with metastasis. (Davis, et al., 2014). The

epithelial to mesenchymal transition (EMT) is the change from an epithelial phenotype, in which

cells are boxy in shape and attached to one another, to a mesenchymal phenotype. The levels of

these markers raised an interesting area of research: the effects of mesenchymal status on TBK1

sensitivity in non-small cell lung cancer cells. In order to better explore this topic, the following

experiments were designed and run.

5

4. Experiment Design

The investigation consisted of two different assays: a cytostatic and cytotoxic assay.

These two different assays serve two different purposes: the cytostatic assay is designed to

measure whether or not the drugs are stopping the multiplication of cells. The cytotoxic assay

measures the ability of the drugs to kill off the cells. Both assays give insight into the TBK1

sensitivity of the cells to the drugs. The experiment used the two aforementioned cell lines: H460

and HCC44, both of which are non-small cell lung cancer cell lines derived from

adenocarcinomas. Three different variants of each line were generated: the pBABE variant,

LKB1 variant, and LKB1 kinase dead (KD) variant. The pBABE variant is a normal cancer cell

line, but it has been stably transfected with the pBABE viral vector, which does not code for any

of the genes but rather makes it easier to manipulate the genes of the cells. The pBABE variant

acts as a control for the stable transfection of the genes into the cell. The next two variants are

both genetically modified variants of these cell line. The LKB1 variant contains an

overexpressing, functional version of LKB1, a kinase and gene that serves as a tumor suppressor.

Cells that contain an overexpressed version of LKB1 have a more epithelial phenotype than the

pBABE line, allowing observation of cells that represent more towards the epithelial end of the

EMT spectrum. Thus, the epithelial-characteristic line can be contrasted with the pBABE line,

which has been described as more mesenchymal. The pBABE cell lines then serve as the

baseline against the LKB1 lines to measure the effect of LKB1 insertion on TBK1 sensitivity.

The final variant used is an LKB1 kinase dead variant of the cell line. This version of the LKB1

gene is also overexpressed but has been rendered useless, just as the gene in cells in advanced

stage cancers might be, creating a cell that will mimic a more mesenchymal phenotype. The

LKB1-KD lines help verify that the overexpression of the active LKB1 protein is causing a more

6

epithelial status and that it is causing the difference in sensitivity to TBK1 inhibition. Using these

three cell type variants from two cell types, 6 different cell lines will be tested. These cell lines

will be tested with two separate drugs: BX795 and Compound II, both of which have been

discussed in the background.

The cytostatic assay consisted of creating a drug-response curve, or DRC. In order to

create a drug-response curve, 96-well plate assay was used, using a 96-well plate (see figure 1).

Figure 1: The layout of a 96-well plate

There is a specific protocol followed during the drug-response curve assay. In order to

prevent contamination of the cells, a majority of the steps are performed in cell culture biosafety

cabinets, which protect a sterile environment for cell culture. First, cells that were being cultured

in incubators in 10 cm plates are taken and then are looked at through a microscope. If they were

not near 100% confluency (a term used to describe the amount of area the cells cover on the

plate), they were put back in the incubator to continue growing, but if they were near 100%

confluency, then the media is removed and the cells are washed 2 times with phosphate-buffered

saline (PBS). 2 mL of trypsin are then added and the plates are incubated for 5 to 10 minutes to

remove the cells from the plate. Then, 8 mL of RPMI media with 5% fetal bovine serum (FBS)

are added, which neutralizes the trypsin. This solution is then put into a tube. The cells are then

counted using an automated cell counter. If there weren’t enough cells to proceed with the assay,

the cells would be put into new 10 cm plates and then put back into incubators to grow. If there

A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12

B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 B11 B12

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12

D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12

E1 E2 E3 E4 E5 E6 E7 E8 E9 E10 E11 E12

F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12

G1 G2 G3 G4 G5 G6 G7 G8 G9 G10 G11 G12

H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12

7

are enough cells, the cells would be diluted to create a solution that had 100,000 cells per mL for

H460 and 40,000 cells per mL for HCC44; the difference in cell concentration is due to the

different proliferation rates of the cells. 11 mL of cell solution would then be put into a plastic

media reservoir, which allows for the transfer of cells into the 96-well plates. The cells are then

transferred into the 96-well plate from the reservoir using a multichannel micro-pipette. 100 µL

of cell media solution, containing 4,000 cells for HCC44 or 10,000 cells for H460, are

transferred into each well. For the purposes of this assay, each cell line has two plates. The cells

were then allowed to grow for one day in incubators set at 37 degrees Celsius and 5 percent

carbon dioxide. The next day, the drugs were added to the cells. The drugs were prepared in

another 96-well plate. Each well was filled with 164 ml of media. Then, 6 wells, labeled as wells

B11 through G11, are filled with 82 µL of extra media. Then, DMSO, the compound in which

the drugs are dissolved, is added to wells B4 to B11 and G4 to G11. Then, the drug is added:

wells C11 and D11 receive BX795, and wells E11 and F11 receive Compound II, creating

concentrations of 33.33 µM. The volume of DMSO added to each well is the same as the volume

of drug added, in order to keep the concentration of DMSO in all wells equal. 82 µL is then

taken from each well that has additional media, mixing it with the media in the well to dilute the

drug, then adding those 82 µL to the wells next to the left, creating a serial dilution of one-third.

Thus, extra media is added from wells B11 through G11 to wells B10 though G10. Each serial

dilution results in a lower concentration of the compounds, but the concentration of DMSO

remains the same in wells in the same column. This serial dilution is carried on until column 4,

creating a range on concentrations. Extra media from column 4 is thrown away. This creates a

finished drug plate (see figure 2).

8

Figure 2: The layout of the finished drug plate1

Media Media Media Media Media Media Media Media Media Media Media Media

Media Media Media DMSO DMSO DMSO DMSO DMSO DMSO DMSO DMSO Media

Media Media Media .015[1] .046[1] .137[1] .412[1] 1.23[1] 3.70[1] 11.11[1] 33.3[1] Media




Media Media Media DMSO DMSO DMSO DMSO DMSO DMSO DMSO DMSO Media

Media Media Media Media Media Media Media Media Media Media Media Media

Then, 50 µL of the media from the drug plate is added to corresponding well on the cell

plate. The cell plates are then kept back in the incubator for three days, after which they are

removed and are mixed with Cell Titer Glo, which is used to measure ATP production and gives

off light in proportion to how much ATP is present. The plate is then read by a plate reader,

determining the luminescence of each well. The luminescence values are used to determine the

relative amount of living cells in each well, demonstrating how well the drug is inhibiting

cellular proliferation. Using the absorbance values, which are standardized using methods

discussed in the Results section of this paper, a drug response curve (DRC) is generated,

showing the sensitivity of the cells to the drugs.

The cytotoxic assay consisted of a qualitative assay, a 6-well plate assay. A majority of

the steps are performed in cell culture biosafety cabinets Again, the same procedure as listed

above was used to separate the cells from the 10 cm plates. Once the cell are put into the tube,

they are counted. If there are not enough cells, then they are replated and put into the incubator to

grow again. If there are enough cells, the cells are diluted to the needed concentrations at a total

volume of 26 mL. Each well in the 6-well plate is filled with 2 mL media (the concentrations of

cells necessary are 20,000 cells per well of HCC44 or 40,000 cells per well of H460) and two

1 [1] represent µM BX795, and [2] represents µM Compound II

9

plates are made for each cell line variant. The cells are allowed to grow until they are 100%

confluent. The media is then taken off from each plate and replaced with new media. This new

media contains one of 3 compounds: DMSO, BX795, or Compound II. The H460 lines have two

wells with 0.04% DMSO, one well with 2 µM BX795, one well with 4 µM BX795, one well

with 2 µM Compound II, and one well with 4 µM Compound II. The HCC44 lines have two

wells with 0.02% DMSO, one well with 1 µM BX795, one well with 2 µM BX795, one well

with 1 µM Compound II, and one well with 2 µM Compound II. The differences in

concentrations used results from the different IC50 values determined from the drug response

curve: HCC44 had lower IC50, so it required lesser concentrations than H460. The cells are then

put back into the incubator and are checked after every three days. If the plates have very little

cell death, change the media and add fresh media containing the appropriate amount of DMSO or

drugs. At the end point, after at least 2 rounds of treatment, when enough death has occurred, one

aspirates off the media and washes the cells twice with PBS. Then, one uses ice-cold menthol to

fix the cells to the plate for 10 minutes. After that, drain the cells and add 0.05% crystal violet to

stain the cells for 30 minutes. Then, properly dispose of the crystal violet and wash each well 6

times with water to wash off excess crystal violet. Each plate had a picture taken of it, and the

blue color of each well was compared by eye to those of other wells to determine the amount of

cell death present in each cell. Darker shades of blue indicate that there are more live cells and

thus less cell death, whereas lighter shades of blue indicate that there are less live cells and thus

more cell death.

5. Results

The results from the experiment are listed below. The cytostatic assay results are

discussed first, and then the cytotoxic assay. The cytostatic assay was normalized for various

10

different values. First, the values for wells treated with the drugs were normalized to average

values for the cells in DMSO: the absorbance values for DMSO wells (B4-B11 & G4-G11) were

averaged for each row individually and each of the bolded well’s absorbance values was divided

by this average value. For example, the values for wells C3 through F3 are divided by the

average value of B3 and G3 and so on. After these DMSO-normalized values were attained, the

values were again normalized to the cells growing in normal media: for example, the normalized

values for cells in wells C4-C11 are each divided by the value for the normalized value for C3,

and so on for rows D through F. The data is then plotted on a graph: the X-axis uses a log scale

to represent the concentration of the drug in µM, and the Y-axis represents the values obtained

from double normalization. The normalization process helps act as a control on the whole

experiment, taking out the effect that DMSO has on the cells themselves and leaving the effect of

the drugs on the cells. In addition, the normalization process provides data that is expressed in

simple numbers, less than or equal to 1.1, that can be compared to determine the effect of the

drugs on the cells. Thus, the normalized values show what the effects on the drugs themselves

were on the cells, taking out the effect of DMSO on the cells. In total, 12 tables of data were

attained (see figure 3 for which rows are used for absorbance values).

Figure 3: Layout of a 96-well plate. Cells in bold are used to generate curve

A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12

B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 B11 B12

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12

D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12

E1 E2 E3 E4 E5 E6 E7 E8 E9 E10 E11 E12

F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12

G1 G2 G3 G4 G5 G6 G7 G8 G9 G10 G11 G12

H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12

11

The tables of data listed include the double normalized data and the graphs of the data,

which include IC50 data, the concentration of drug needed to inhibit a cellular process in 50

percent of cells. The values for H460 are first, and the values for HCC44 are second.

Concentration (µM) 0 0.0152 0.0457 0.1372 0.4115 1.2346 3.7037 11.111 33.333

H460

pBABE

1

BX795 neg cont norm 1 0.98356 1.02172 1.03830 1.05783 0.50617 0.36360 0.14879 0.01204

BX795 neg cont norm 1 1.00798 1.03116 0.99805 1.03706 0.49582 0.35152 0.09567 0.00853

Cmpd II neg cont norm 1 1.01276 1.06584 1.08121 1.10837 0.93586 0.56510 0.40067 0.00440


H460

pBABE

2

BX795 neg cont norm 1 0.97120 0.95902 0.96257 1.02974 0.55591 0.34979 0.13218 0.01084

BX795 neg cont norm 1 0.95161 0.96986 0.98848 1.03755 0.50707 0.37192 0.11489 0.00740



H460

LKB1

1

BX795 neg cont norm 1 1.01422 1.02927 1.00693 0.99611 0.52344 0.40430 0.20741 0.02633

BX795 neg cont norm 1 1.03661 1.00059 1.04080 1.00590 0.50739 0.40683 0.14840 0.02312



H460

LKB1

2

BX795 neg cont norm 1 1.01769 1.01595 0.96492 1.01540 0.57621 0.39369 0.18274 0.02690

BX795 neg cont norm 1 1.01668 1.04073 1.00959 1.03882 0.62567 0.41842 0.15120 0.02303



H460

LKB1-

KD 1

BX795 neg cont norm 1 0.93621 1.01006 1.01286 0.97351 0.36270 0.28845 0.12150 0.01349

BX795 neg cont norm 1 0.96843 1.00819 0.90525 0.95963 0.37362 0.28003 0.09391 0.01053



H460

LKB1-

KD 2

BX795 neg cont norm 1 0.98647 0.99820 1.04713 1.01077 0.39464 0.30070 0.15149 0.01505

BX795 neg cont norm 1 0.99364 1.00755 1.01656 0.95136 0.36995 0.28234 0.11008 0.00932



H460 DRC - BX795 H460 DRC – Compound II

-2 -1 0 1 2 0.0

0.5

1.0

Log10 Drug [ ] µM

H460 + pBABE + BX795 H460 + LKB1 + BX795 H460 + LKB1-KD + BX795

BX795 IC50 (+ pBABE) = 1.70 µM

BX795 IC50 (+ LKB1) = 1.98 µM

BX795 IC50 (+ LKB1-KD) = 0.95 µM

-2 -1 0 1 2 0.0

0.5

1.0 H460 + pBABE + Compound II H460 + LKB1 + Compound II

H460 + LKB1-KD + Compound II

Cmpd II IC50 (+ pbabe) = 5.88 µM Cmpd II IC50 (+ LKB1) = 7.73 µM Cmpd II IC50 (+ LKB1-KD) = 2.94 µM

Log10 Drug [ ] µM

12

Concentration (µM) 0 0.01524 0.04573 0.13717 0.41152 1.23456 3.70370 11.1111 33.3333

HCC44

pBABE

1

BX795 neg cont norm 1 0.95182 0.91345 0.87610 0.64323 0.14863 0.02815 0.00780 0.02512

BX795 neg cont norm 1 0.99263 0.96508 0.89113 0.59972 0.16480 0.01608 0.00389 0.02076



HCC44

pBABE

2

BX795 neg cont norm 1 0.88060 0.84192 0.83030 0.58139 0.15285 0.02600 0.00962 0.02681

BX795 neg cont norm 1 0.97348 0.94232 0.90512 0.66977 0.16682 0.03164 0.00478 0.02817



HCC44

LKB1

1

BX795 neg cont norm 1 0.93656 0.88222 0.85490 0.74851 0.52492 0.24442 0.01584 0.11927

BX795 neg cont norm 1 0.94189 0.91926 0.85040 0.74517 0.51320 0.22661 0.01046 0.11903



HCC44

LKB1

2

BX795 neg cont norm 1 0.83700 0.84675 0.84353 0.69053 0.51115 0.17409 0.01533 0.11717

BX795 neg cont norm 1 0.89465 0.87673 0.82222 0.70354 0.54096 0.30614 0.01508 0.11078



HCC44

LKB1-

KD 1

BX795 neg cont norm 1 0.92521 0.88598 0.83302 0.63865 0.14958 0.03460 0.00882 0.01505

BX795 neg cont norm 1 0.97602 0.84927 0.78791 0.60570 0.1153 0.02806 0.00392 0.01115



HCC44

LKB1-

KD 2

BX795 neg cont norm 1 0.78187 0.84149 0.76611 0.64879 0.14569 0.02464 0.010604 0.01580

BX795 neg cont norm 1 0.85985 0.93369 0.81514 0.69233 0.15743 0.03568 0.003437 0.01669



HCC44 DRC – BX795 HCC44 DRC – Compound II

The cytotoxic assay does not provide a number output, but rather one can tell the relative

amount of dead cells through image analysis. By looking at the shade of the stain, one can tell

qualitatively how much the cells died and determine the effectiveness of the drug, as darker

shade indicates less cell death and a lighter shade indicates greater cell death. The pictures that

-2 -1 0 1 2 0.0

0.5

1.0

Log10 Drug [ ] µM

HCC44 + pBABE + BX795 HCC44 + LKB1 + BX795 HCC44 + LKB1-KD + BX795

BX795 IC50 (+ pBABE) = 0.56 µM BX795 IC50 (+ LKB1) = 1.40 µM BX795 IC50 (+ LKB1-KD) = 0.62 µM

-2 -1 0 1 2 0.0

0.5

1.0

Log10 Drug [ ] µM

HCC44 + pBABE + Compound II HCC44 + LKB1 + Compound II HCC44 + LKB1-KD + Compound II

Cmpd II IC50 (+ pBABE) = 0.61

µM Cmpd II IC50 (+ LKB1) = 1.35 µM Cmpd II IC50 (+ LKB1-KD) = 0.48 µM

13

were taken of the plates are listed, along with a caption describing the variant of the cell line and

when it was stained. The pictures are below, with H460 first and then HCC44. The layout of the

plates is shown in the tables below (first for H460 and then for HCC44):

Figure 4: The layouts for the plates (H460 and HCC44)

0.04% DMSO 4 µM BX795 4 µM Compound II




6. Analysis

The H460 DRC had interesting results. The IC50 values of the pBABE lines for BX795

(1.70 µM) was not much different than that of the LKB1 variant (1.98 µM), and the curves for

both lines seem to indicate that they have similar sensitivity to BX795. The LKB1-KD line,

however, has a much lower IC50 value (0.95 µM), almost half of the pBABE and LKB1

variants. Comparing the more epithelial variant (LKB1) and the more mesenchymal variants

(LKB1-KD and pBABE), one can see that the variants displaying a more mesenchymal

phenotype are more sensitive to TBK1 inhibition through BX795 than the variant with a more

H460 Set 1

pBABE

H460 Set 1

LKB1

H460 Set 1

LKB1-KD

H460 Set 2

pBABE

H460 Set 2

LKB1

H460 Set 2

LKB1-KD

HCC44 Set

1 pBABE

HCC44 Set

1 LKB1

HCC44 Set 1

LKB1-KD

HCC44 Set 2

pBABE

HCC44 Set

2 LKB1

HCC44 Set

2 LKB1-KD

14

epithelial phenotype is, though the pBABE variant is not that much more sensitive to TBK1

inhibition than the LKB1 variant. This could be because the overexpression of LKB1 (as in the

LKB1 lines) does not affect the levels of LKB1 that much from the base state and thus does not

affect the sensitivity of the cell to TBK1 as much, but the overexpression of an inactive form of

LKB1 (as in the LKB1-KD lines) causes the cells to become much more sensitive to TBK1

inhibition.

The results from the Compound II DRC for H460 also show similar results: there is not

that much difference in between the IC50 values for the pBABE (5.88 µM) and LKB1 (7.73

µM). However, the LKB1-KD line has an IC50 (2.94 µM), more than half that of the LKB1 line

(7.73 µM). So, just as the values for the H460 BX795 DRC shows, the LKB1 overexpression

does not have that much of an effect on TBK1 sensitivity, though the values for the LKB1-KD

line are much lower, indicating that the overexpression of the inactive form has a much greater

effect on the phenotype than the overexpression of the active form. The data from both DRCs

supports the hypothesis, since the values for the more mesenchymal lines (LKB1-KD and

pBABE) are less that the values for the more epithelial lines (LKB1), even though the pBABE

values are not much lower than those of LKB1.

The HCC44 lines show a much greater sensitivity than do the H460 lines, and a different

pattern exists in the IC50 values. The IC50 for BX795 of the pBABE line (0.56 µM) was

extremely close to that of the LKB1-KD line (0.62 µM). However, there is a large disparity

between the BX795 IC50 the pBABE line (.56 µM) and the LKB1 line (1.40 µM). The IC50 of

the LKB1 variant is much higher than that of the pBABE or LKB1-KD variants. Thus, the data

supports the hypothesis, since the cells overexpressing LBK1, which have a more epithelial

phenotype, are less sensitive to TBK1 inhibition. The Compound II DRC data matches up with

15

that of the BX795 DRC: both the pBABE line (0.62 µM) and LKB1-KD line (0.48 µM) have

similar IC50s, whereas the LKB1 line has a much higher IC50 (1.35 µM). The data show that the

more epithelial variant is much less sensitive to TBK1 inhibition than the more mesenchymal

variants. In addition, LKB1 overexpression has a much greater effect in the HCC44 cells, which

have a much higher IC50 value for the LKB1 lines, whereas the overexpression of an inactive

form of LKB1 does not have much of an effect on TBK1 sensitivity, as seen through the similar

IC50 values of the pBABE and LKB1-KD lines.

The cytotoxic binary assay also shows the effects of BX795 and Compound II on the

cells by showing how much they are killed rather than how much they stop growing. The H460

lines presented a challenge in that they did not show much death upon one week of drug

treatment (as shown by Set 1). There is some cell death present in the cells, but only discernible

in the 4 µM BX795 wells, and only by a little bit. A second week of treatment with fresh media

produced greater results, allowing for images which can be better distinguished, though only

slightly so. Though it is still hard to tell, the Set 2 plates show greater cell death in the pBABE

and LKB1-KD plates versus those in the LKB1 plates in the wells treated with 4 µM BX795,

since the pBABE and LKB1-KD plates are lighter, indicating greater cell death. The other wells

are still too close in color to be able to make a distinction regarding the effects of the drugs on

the cell lines. This cell line might be more resilient to the drug treatment, and the cytotoxic

effects of the drug may be limited. However, the greater resistance to treatment suggested by the

BX795 DRC and Compound II DRC of the cell line may account for this phenomenon, thus

meaning that the cell line may require more treatment with a higher concentration of drugs

before showing much cell death.

16

The HCC44 cytotoxic assay results mirrored those from the DRC: the pBABE and

LKB1-KD lines are much lighter in color than the LKB1 cell line in the wells treated with

BX795 and 2 µM Compound II, again showing that the mesenchymal type cells are more

sensitive to TBK1 inhibition than are the epithelial type cells. At the same time, the HCC44 lines

show a much greater difference in color from being treated with the drugs than did the H460

lines, a result of HCC44’s lower IC50 when compared to H460, thus showing that the line is

more sensitive to TBK1 inhibition. Still, both plates do have wells that were not able to be

distinguished very well from one another. The data from the H460 cytotoxicity assay is

inconclusive: the wells are so close to each other in color that they cannot be distinguished.

However, the H460 assay does not contradict the hypothesis, and the HCC44 assay supports the

hypothesis.

7. Discussion, Extension, and Conclusion

The results from the experiments support the hypothesis, that cells with a more

mesenchymal phenotype are more sensitive to TBK1 inhibition than are cells with a more

epithelial phenotype. The data from the DRC show that the LKB1-KD and pBABE lines, which

are more mesenchymal in nature, show greater sensitivity to BX795 and Compound II than the

LKB1 line, which is more epithelial in nature. Also, the cytotoxic assays tends support the

hypothesis, as the cytotoxic assays show a lighter shade in the LKB1-KD and pBABE lines than

the LKB1 lines, which are a much darker shade of blue, even though the results from the H460

assay are inconclusive.

One of the improvements that could be made to this investigation is performing another

trial of the cytotoxic assay. The H460 lines could be treated with higher concentrations of the

drug to see a greater effect of the drug on the H460 lines. This would create a lighter shade of

blue that would be much easier to read and make it easier to see the differences between the

17

various drug concentrations and different cell line variants. The same improvement could be

made to the HCC44 lines as well, as some of the wells are very similar in color across all plates,

such as the 1 µM Compound II wells. This would help provide more conclusive results for the

experiment. These are the improvements that could be made to the current experimental design.

However, further studies and experiments are needed to determine whether or not TBK1

inhibition is a viable method for treating more mesenchymal-type cancers. Further experiments

would also be required to determine the efficacy of TBK1 inhibition at killing Ras mutant cancer

cells. One further extension of the cytotoxic assay would be to quantify the results obtained from

the experiment. This would involve repeating the cytotoxic assay, but once the plates are stained,

the stains would be dissolved in acid and quantified, providing numerical values that could be

used to compare the cell death present in the wells. This method would make it easier to check

the amount of cell death present on the plates, providing more certainty to the results obtained

from this assay. Another set of assays that could be used to determine the mesenchymal status of

the cells are Western Blots. A Western Blot involves taking protein lysate from the cells,

separating the protein lysate from the cells based on size of protein fragments in a

polyacrylamide gel, then transferring the protein onto a membrane and blotting it with primary

antibodies to detect the presence of proteins on the membrane. The membrane is then treated

with secondary antibodies, which are tagged with a chemiluminescent indicator. The membrane

is then put next to photography film in a dark room, and this film is developed to present a

readout of the various protein levels. This assay could be used to detect the relative levels of

various proteins within the cells. There are two different Western blots that could be run: one

would involve using the cell line variants without any drug treatment and measuring the various

levels of mesenchymal-state and epithelial-state markers, such as E-cadherin, vimentin, ß-

18

catenin, and N-cadherin. This would help establish whether the cell line variants are more

epithelial or more mesenchymal, allowing for a better interpretation of the data from this

analysis. Another Western Blot assay would involve the same set-up as the cytotoxic assay, but

instead of dying the cells, the cells would be lysed and protein lysate would be collected. The

protein lysate could be used to determine the effects that TBK1 has on a cellular level, showing

which pathways are being inhibited and how this relates to the mesenchymal or epithelial state of

the cell. Currently, the data suggests a correlation between the mesenchymal status of a cell and

its sensitivity to TBK1 inhibitors, but there is not an understanding of what makes the cells more

sensitive to TBK1 inhibition, so such an assay would illuminate what is occurring at the cellular

level. Finally, this experimental analysis only looked at two cell lines that had been selected

based on their characteristics. Expanding the research to include variants of other cell lines

would provide even greater data into this phenomenon, providing greater insight. In order to test

the real world efficacy of TBK1 inhibition and its correlation to mesenchymal status, mouse

studies would need to be conducted. Mice would need to be implanted with tumors of different

mesenchymal and epithelial characteristics and treated with TBK1 inhibitors such as BX795 and

Compound II, in order to see the correlation of the data in an animal setting

Though these are all further extensions of the experiment, the data from this assay

suggests a new method of treating cancer. The cell lines studied in the assay were all K-Ras

mutant cancer cell lines, and this assay seems to show that TBK1 inhibition may work in

combating K-Ras mutant cancers, which have low survival rates, as mentioned in the

introduction. At the same time, mesenchymal type cells are associated with late-stage and

advanced cancers, which have undergone metastasis. TBK1 inhibition, based off the results of

this experiments, may hold a promise in treating such types of cancers. The possibilities for this

19

method of treatment are bright: more experiments need to be done to ascertain the efficacy of

TBK1 inhibition for treating mesenchymal-type cancers.

In conclusion, this investigation has provided evidence supporting the hypothesis, that

cells with more mesenchymal phenotype are more sensitive to TBK1 inhibition than are cells

with a more epithelial phenotype. This research suggests that there could be applications to treat

advanced and late-stage cancers, though further research is required to test the applications of the

drug in this setting. The results of this experiment may lead to new discoveries to combat mutant

Ras cancers, helping to provide hope to those who have very little.

20

References

Bos, J. L., 1989. ras Oncogenes in Human Cancer: A Review. Cancer Research, 1 September,

pp. 4682-4689.

Clark, K., Plater, L., Peggie, M. & Cohen, P., 2009. Use of the Pharmacological Inhibitor BX795

to Study the Regulation and Physiological Roles of TBK1 and IKB Kinase. The Journal of

Biological Chemistry, 22 May, 284(21), pp. 14136-14146.

Cooper, J. M., Bodemann, B. O. & White, M. A., 2013. The RalGEF/Ral Pathway: Evaluating

an Intervention Opportunity for Ras Cancers. The Enzymes, pp. 137-156.

Davis, F. M., Stewart, T. A., Thompson, E. W. & Monteith, G. R., 2014. Targeting EMT in

cancer: opportunities for pharmacological intervention. Trends in Pharmacological Sciences,

September, pp. 479-488.

Fernández-Medarde, A. & Santos, E., 2011. Ras in Cancer and Developmental Diseases. Genes

and Cancer, pp. 344-358.

Kim, H. S. et al., 2013. Systematic Identification of Molecular Subtype-Selective Vulnerabilities

in Non-Small-Cell Lung Cancer. Cell, 24 October, pp. 552-566.

National Cancer Institute, 2012. Cancer Survival Rates. [Online]

[Accessed 11 August 2015].

National Cancer Institute, n.d. Cancer Statistics: National Cancer Institute. [Online]

Available at: http://www.cancer.gov/about-cancer/what-is-cancer/statistics

[Accessed 10 August 2015].

Ou, Y.-H.et al., 2011. TBK1 Directly Engages Akt/PKB Survival Signaling to Support

Oncogenic Transformation. Molecular Cell, 18 February, pp. 458-470.

Siemens Research Paper FINAL

Documents