Florida State University Libraries

)ORULGD6WDWH8QLYHUVLW\/LEUDULHV

2021 Investigating the Role of Gankyrin in Breast Cancer and its Potential as a Novel Therapeutic Target Jessica Margaret Jarnagin

Follow this and additional works at DigiNole: FSU's Digital Repository. For more information, please contact [email protected] THE FLORIDA STATE UNIVERSITY

COLLEGE OF ARTS & SCIENCES

INVESTIGATING THE ROLE OF GANKYRIN IN BREAST CANCER AND ITS POTENTIAL AS A NOVEL THERAPEUTIC TARGET

JESSICA MARGARET JARNAGIN

A Thesis submitted to the Department of Biological Science in partial fulfillment of the requirements for graduation with Honors in the Major

Degree Awarded: Spring, 2021 The members of the Defense Committee approve the thesis of Jessica M. Jarnagin defended on April 2nd, 2021.

______

Dr. Antonia Nemec

Thesis Director

______

Dr. Robert J. Tomko Jr.

Committee Member

______

Dr. Qian Yin

Committee Member

______

Dr. Scott Stagg

Committee Member

INVESTIGATING THE ROLE OF GANKYRIN IN BREAST CANCER AND ITS POTENTIAL AS A NOVEL THERAPEUTIC TARGET

1. ABSTRACT 4 - 5 2. INTRODUCTION: 5 - 11 2.1 BREAST CANCER 5 - 6 2.2 THE UBIQUITIN-PROTEASOME SYSTEM 6 - 8 2.3 PROTEASOME UPREGULATION IN CANCERS 8 - 11

3. METHODS: 11 - 16 3.1 CELL LINES AND PLASMID CREATION 11 - 13 3.1.1 CELL LINES 11 - 12 3.1.2 PLASMIDS - 12 3.1.3 CLONING 12 - 13 3.1.3.1 CLONING pRVYtet-off-FLAG-hsGANKYRIN 3.1.3.2 CLONING pLJM1-HA-PSMC4 AND pLJM1-HA-psmc4(337-419) 3.1.3.3 CLONING pBabe-HA-PSMC4 AND pBabe-HA-psmc4(337-419) 3.2 PROTEIN EXPRESSION 13 - 14 3.2.1 DENATURING POLYACRYLAMIDE GEL ELECTROPHORESIS 13 - 14 3.3 MAKING RETROVIRUS WITH GP2-293 CELLS 14 - 15 3.4 RETROVIRAL INFECTION 15 3.5 PEPTIDASE ACTIVITY ASSAY 15 - 16 3.6 ANCHORAGE INDEPENDENT GROWTH 16

4. RESULTS: 16 - 25 4.1 GANKYRIN IS EXPRESSED IN FLP-IN HEK293 CELLS 16 - 17 4.2 INCREASED PEPTIDASE ACTIVITY OF FLP-IN HEK293 CELLS OVEREXPRESSING GANKYRIN 18 4.3 EXPRESSION OF FLAG-GANKYRIN IN MCF10A CELLS 19 4.4 NO CHANGE IN PEPTIDASE ACTIVITY OF MCF10A CELLS OVEREXPRESSING GANKYRIN 19 - 20 4.5 ANCHORAGE INDEPENDENCE PHENOTYPE FOR MCF10A GANKYRIN AND EMPTY VECTOR CELLS 20 4.6 EXPRESSION OF HA-PSMC4 AND HA-PSMC4(337-419) IN MCF10A CELLS 21 - 22 4.7 EXPRESSION OF HA-PSMC4 AND HA-PSMC4(337-419) IN FLP-IN HEK293 CELLS 22 - 23 4.8 INCREASED PEPTIDASE ACTIVITY OF MCF10A CELLS EXPRESSING HA-PSMC4 23 - 24 4.9 ANCHORAGE INDEPENDENCE PHENOTYPE FOR MCF10A EMPTY VECTOR, GANKYRIN, PSMC4, AND GANKYRIN + PSMC4 CELLS 24 - 25

5. DISCUSSION 25 - 29 6. ACKNOWLEDGMENTS 29 - 30 7. REFERENCES 31 - 32 8. SUPPLEMENTAL INFORMATION: 33 8.1 TABLE 1 – PLASMIDS 33

1. Abstract

Breast cancer is a heterogeneous disease and is typically accompanied by the activation of

HER2, hormone receptors, and/or mutations in the BRCA genes. Treatment of breast cancer is dependent on the molecular features of the cancer, and response to treatment varies [1]. The oncogene gankyrin was found to be overexpressed in many molecular subtypes of breast cancer and correlated with poor prognosis [4]. Gankyrin is a chaperone protein that aids in the assembly of the 26S proteasome. The proteasome is a large ~70 subunit protease complex that drives cancer by destroying anti-growth and pro-death signals and is upregulated in many cancer types

[25]. Because gankyrin is critical for proteasome assembly, we aimed to determine the role that gankyrin overexpression plays in contributing to proteasome upregulation. We generated both

MCF10A breast epithelial cell lines and Flp-In HEK293 cell lines overexpressing gankyrin or empty vector which allowed us to test the effects of gankyrin overexpression on the peptidase activity of the proteasome and the occurrence of the anchorage independence phenotype. Using these cell lines, we conducted peptidase activity assays and determined that gankyrin overexpression leads to increased proteasome activity in Flp-In HEK293 cells. We also conducted anchorage independence assays, which showed that overexpression of gankyrin in

MCF10A gankyrin cells lines led to increased occurrence of anchorage independence phenotypes. Additionally, during proteasome assembly gankyrin binds to the subunit PSMC4 and functions to prevent premature docking of PSMC4 onto the core particle, so we created a minimal part of PSMC4 (psmc4(337-419)) that sequesters gankyrin to interfere with proteasome assembly in a method called gankyrin trapping [2]. We attempted to generate both MCF10A cell lines and Flp-In HEK293 cell lines that expressed either gankyrin or empty vector and psmc4(337-419), PSMC4, or an empty pBabe or pLJM1 plasmid. However, the retroviral infection of Flp-In HEK293 cells with virus containing these plasmids was unsuccessful. Also, 4

HA-psmc4(337-419) was undetectable in MCF10A cells. Using the cell lines that were

successfully created with full-length PSMC4, we conducted peptidase activity assays and anchorage independence assays which showed that overexpression of PSMC4 in MCF10A gankyrin cell lines led to increased proteasome activity and an increased occurrence of the anchorage independent growth phenotype. This data indicates that gankyrin-trapping with full length PSMC4 is not a valid mechanism for decreasing oncogenic activity. Altogether, this study indicates the significance of the role of gankyrin overexpression in proteasome upregulation, and that gankyrin is potentially a valid target for proteasome inhibition.

2. Introduction

2.1 Breast Cancer

Breast cancer is the second most common cancer found in women and is characterized by malignancies of the breast tissue. The main components of this disease are mutations of BRCA genes, activation of estrogen and/or progesterone receptors, and the activation of the human epidermal growth factor receptor 2 (HER2) encoded by the gene ERBB2 [1]. Breast cancer can

either develop sporadically or be genetically linked, however only 10% of breast cancers are

inherited [5]. If treated in the early stages of carcinogenesis prior to metastasis, there is a 70-80%

chance of full recovery. However advanced/metastatic breast cancer is currently considered

incurable [1]. The most commonly amplified genes are often targeted for treatment development,

and include TP53, PTEN, ERBB2, BRCA 1/2, and FGFR1 [6]. Over the last 20 years, treatment

of breast cancer has had improved outcomes due to the current focus on the heterogeneity of the

disease. Treatments have evolved to target the specific combination of BRCA mutations,

estrogen/progesterone receptor activation, and HER2 activation that exists in each person

diagnosed with the disease. However, studies have shown that most breast cancers are caused by 5

the accumulation of many mutations, which can make targeting treatments difficult to develop

[7]. Thus, it is crucial to develop new therapies independent of mutation status. Recently, the

oncogene gankyrin, a chaperone protein that aids in the assembly of the proteasome, was found

to be upregulated in breast tumors and its overexpression in breast epithelial cell lines induced a

tumorigenic phenotype. Gankyrin overexpression occurs in the early stages of breast cancer and is correlated to poor prognosis and metastasis [33]. Although it was found to be associated with

HER2 expression, gankyrin overexpression was found in breast cancer regardless of HER2 and

ER/PR status, including triple-negative breast cancer, an aggressive cancer with poor treatment options and prognosis [4] [34].

2.2 The Ubiquitin-Proteasome System

The ubiquitin proteasome system (UPS) is the primary mechanism for protein degradation in eukaryotic cells. Proteins that are destined for degradation are modified by the addition of a chain of the protein ubiquitin (Ub). The polyubiquitin (polyUb) chain signals the protein for delivery to the 26S proteasome.

Figure 1. Structure of the Human 26S Proteasome The proteasome consists of two main parts: the 19S regulatory particle (RP) and the 20S core particle (CP). The RP is responsible for binding incoming 6 substrates, removing the polyubiquitin chain, unfolding the substrate, and translocating it into the interior of the CP. The CP houses the peptidase active sites in its interior and cleaves the substrate into short peptides.

The proteasome is a large 2.5 mDa multi-subunit protease that can be divided into the 19S regulatory particle (RP) and 20S core particle (CP) (Figure 1). The RP can be further divided into the RP lid and RP base. The RP lid consists of nine regulatory particle non-ATPase subunits, and deubiquinates the substrate. The RP base is made up of a heterohexameric ATPase ring and four non-ATPase subunits. The base unfolds the substrate using energy derived from ATP. The

CP consists of four stacked heteroheptameric rings resulting in the formation of a barrel-like shape. The peptidase active sites are located within the CP and cleave the substrate into short peptides [2] These peptidase active sites include the PSMB5 subunit which has chymotrypsin- like activity, as it cleaves after amino acids with hydrophobic side chains; the PSMB6 subunit which has caspase-like activity, as it cleaves after amino acids with acidic side chains; and the

PSMB7 subunit which has trypsin-like activity, as it cleaves after amino acids with basic side chains [32].

The proteasome contains ~70 subunits that must assemble correctly in order to function properly. Proteasome assembly is driven by intrinsic features of several proteasome subunits themselves, as well as extrinsic, dedicated assembly chaperone proteins. This process is highly regulated, but not well understood. The assembly of the RP base subcomplex is aided by four chaperone proteins under normal growth: PAAF1, gankyrin, S5b, and p27 (Rpn14, Nas6, Hsm3, and Nas2 in yeast, respectively). These chaperones function in part to stabilize base assembly intermediates, and to regulate attachment of the lid and CP to the base. However, the mechanism by which they are evicted to form mature proteasomes is not well understood. Gankyrin has a long crescent-like shape that allows it to bind to the C-terminus of the ATPase base subunit

PSMC4. The C-terminal tail of PSMC4 docks into the CP to complete assembly. When gankyrin

is bound to PSMC4, there is steric hindrance that prevents premature docking of the RP onto the

CP, specifically preventing PSMC4 from stably associating with the CP [2] [17-19]. The eviction of gankyrin is coupled to ATP hydrolysis in the base and allows for the assembly of mature proteasomes as indicated in Figure 2 [3].

Figure 2. Base Assembly When gankyrin binds to PSMC4, the base of the RP is able to assemble. Only once gankyrin is evicted can the proteasome fully assemble. ATPase subunits: blue; assembly chaperones: tan or green

Gankyrin-trapping is a potential method of preventing proteasome assembly. If gankyrin is

sequestered and cannot interact with PSMC4, the base may not assemble properly and the base

subunits may not properly interact with the core particle, potentially leading to an accumulation

of RP and CP subcomplexes, and a decrease in mature proteasomes. Since proteasome

upregulation is a compensatory mechanism of cancer cell survival, a potential mechanism for

gankyrin-trapping is to overexpress a minimal part of PSMC4, that includes the C-terminus

necessary to bind to gankyrin, to prevent docking of PSMC4 into the CP, thus preventing

formation of mature proteasomes.

2.3 Proteasome upregulation in cancers

The proteasome is upregulated in different types of cancers, including breast cancer, and aids

in the survival of cancer cells due to antiapoptotic and proliferative signaling pathways that are

associated with proteasome activity [10]. Specifically, proteasome upregulation was found in

over 90% of women with breast cancer and is also overexpressed in other types of cancers [11].

Because of the heavy reliance on proteasomes in cancer, the proteasome is a validated target of chemotherapy. There are currently three FDA-approved proteasome inhibitors used in the treatment of cancer (bortezomib, ixazomib, carfilzomib). These proteasome inhibitors are successful in treating several blood cancers, but their use in treating solid tumors has been mostly unsuccessful due to the basic pharmacological parameters of the drugs. The distribution of bortezomib is poor due to difficulty of the drug penetrating solid tumors. This is a result of the differing vascularity and permeability of solid tumors [31]. Patients treated with proteasome inhibitors often exhibit resistance to these drugs, even if treatment was initially successful. In bortezomib, this resistance is linked to changes in expression and composition of proteasomal subunits or mutations in the PSMB5 gene which encodes the proteasome subunit PSMB5 protein, where bortezomib binds [32]. Although the second-generation inhibitor carfilzomib is improved, these drugs cause great toxicities including neuropathies and cardiovascular events that often require the termination of the treatment. In addition, although the proteasome contains

~70 subunits, all of the FDA-approved proteasome inhibitors target the PSMB5 subunit within the CP, which has redundant enzymatic activities with other subunits, PSMB6 and PSMB7 [12,

13]. Therefore, the identification of new molecular targets within the proteasome is needed in order to develop more effective proteasome inhibitors.

One potential molecular target within the proteasome is gankyrin. Gankyrin overexpression is associated with poor prognosis in cancer patients [4] [14-16]. Gankyrin also has non- proteasomal roles that function in driving cancer and are associated with metastasis. In colorectal cancer, gankyrin overexpression leads to increased transcription of -catenin by activating the

PI3K/Akt signaling pathway, leading to increased angiogenesis and metastasis [15]. In breast

cancer, gankyrin overexpression leads to increased activation of the Rac1 Rho family GTPase

which results in greater rates of membrane protrusion and focal adhesion turnover, aiding in

metastasis and invasion of breast cancer cells [33]. Gankyrin also causes decreased activity of

tumor suppressors such as retinoblastoma protein (pRb) in the CDK/pRb pathway and p53 in the

MDM2/P53 pathway and these are significant in driving cancer [26]. However, as gankyrin is

also critical for proteasome assembly, we hypothesize that gankyrin overexpression may be a

contributing cause of proteasome upregulation and may also be one of the reasons why these

cancers have such poor prognoses. It is critical to evaluate the role of overexpressed gankyrin in

proteasome activity and assess whether gankyrin can be a novel therapeutic target in the

treatment of cancer.

The goal of this Honors Thesis is to understand the effects of gankyrin overexpression on the function of the proteasome, and to evaluate the potential uses of gankyrin-trapping as a therapeutic approach in solid tumors such as breast cancer. We hypothesize that gankyrin

overexpression contributes to proteasome upregulation and will lead to an increase in

proteasome activity. We also hypothesize that we will be able to trap gankyrin by

overexpressing a minimal part of PSMC4, and that this technique of gankyrin-trapping

will inhibit proteasome assembly and reduce the overall negative effects of gankyrin

overexpression in breast cancer cells. To test this hypothesis, we established MCF10A and

Flp-In HEK293 cell lines that overexpress gankyrin and evaluated the effect of this

overexpression on proteasome peptidase activity using a peptidase activity assay. To test our

second hypothesis, we attempted to establish MCF10A and Flp-In HEK293 cell lines that

expressed a minimal part of PSMC4 or overexpressed full length PSMC4 and conducted

anchorage independence assays and peptidase activity assays. This research provides a link for gankyrin overexpression to the upregulated proteasome and increased proteasome activity in

cancer cells. Additionally, this research provides a potential mechanism for the creation of a new

proteasome inhibitor by trapping gankyrin.

3. Methods

3.1 Cell lines and plasmid creation

3.1.1 Cell lines

MCF10A cells are non-transformed, immortalized breast epithelial cells derived from human

mammary tissue (ATCC). The cells were grown and maintained in DMEM/F12 medium

(ThermoFisher) supplemented with 5% horse serum (ThermoFisher), 1% penicillin-

streptomycin, epidermal growth factor (20 ng/ml) (Peptrotech), hydrocortisone (0.5 µg/ml)

(Sigma), cholera toxin (100 ng/ml) (Sigma), insulin (10 µg/ml) (Sigma-Aldrich) and grown at

37°C in a 5% CO2 humidified incubator [27].

Flp-In HEK 293 cells are human embryonic kidney cells expressing the lacZ-Zeocin™

fusion gene (ThermoFisher). In the Flp-In™ system, each cell line contains a single integrated

Flp Recombination Target (FRT) site from pFRT/lacZeo. Using Flp recombinase-mediated

recombination at the genomic FRT site in the pFRT/lacZeo plasmid, a desired DNA sequence

can be “flipped in”. Flp-In HEK293 cells were grown in Dulbecco modified Eagle's medium

(ThermoFisher) supplemented with 10% fetal bovine serum (ThermoFisher), 1% penicillin-

streptomycin (ThermoFisher), and 1% L-glutamine (ThermoFisher) at 37°C in a 5% CO2

humidified incubator and were treated with 100g/ml zeocin.

The GP2-293 virus packaging cell line (Clontech) is derived from human embryonic kidney

cells expressing the gag and pol genes and was used for retrovirus preparation. The gag gene is

necessary for the formation of the viral core, and the pol gene is necessary for expressing reverse

transcriptase. This cell line does not contain the env gene which is necessary for the formation of

the viral envelope; therefore, these cells were transfected with the pVSV-G envelope plasmid

(Clontech). These cells were maintained in the same growth media as the Flp-In HEK293 cells

[27].

3.1.2 Plasmids

Plasmids were created via standard molecular biological approaches as described below. A

full list of constructs is described in Table 1 in the supplement. Plasmids were transformed into

TOP10 F’ E. coli cells and DNA was isolated using the Qiagen DNA purification kit following

manufacturer’s instructions. Plasmid sequences are available upon request.

3.1.3 Cloning

3.1.3.1 Cloning pRVYtet-off-FLAG-hsGankyrin

The PSMD10 (Gankyrin) cDNA was amplified from pRT2054 with PacI and BamHI sites.

The amplified DNA fragments and the pRVYtet-off backbone were digested with PacI and

BamHI (NEB) for two hours at 37C. The PSMD10 cDNA was then ligated into the pRVYtet-off plasmid using T4 DNA ligase (NEB). Ligation reactions were transformed into TOP F’ E. coli and grown overnight on a LB + ampicillin plates to select for transformants. After preparation of plasmid DNA by miniprep (Qiagen), the plasmid DNA was digested with PacI and BamHI and sequenced to confirm the presence of the desired cDNA insert and to ensure there were no mutations.

3.1.3.2 Cloning pBabe-HA-PSMC4 and pBabe-HA-psmc4(337-419)

HA-PSMC4 and HA-psmc4(337-419) cDNAs were amplified from pRT2321 with EcoRI and

SalI sites. The amplified DNA fragments and the pBabe backbone were digested with EcoRI and

SalI (NEB) for two hours at 37C. The HA-PSMC4 and HA-psmc4(337-419) cDNA was then ligated into the pBabe plasmid using T4 DNA ligase (NEB). Ligation reactions were transformed into TOP F’ E. coli and grown overnight on a LB + ampicillin plates to select for transformants.

After preparation of plasmid DNA by miniprep (Qiagen), the plasmid DNA was digested with

EcoRI and SalI and sequenced to confirm the presence of the desired cDNA insert and to ensure there were no mutations.

3.1.3.3 Cloning pLJM1-HA-PSMC4 and pLJM1-HA-psmc4(337-419)

HA-PSMC4 and HA-psmc4(337-419) cDNAs were amplified from pRT2331 or pRT2332 with NheI and EcoRI sites. The amplified DNA fragments and the pLJM1 backbone were digested with NheI and EcoRI (NEB) for two hours at 37C. The HA-PSMC4 and HA- psmc4(337-419) cDNA was then ligated into the pLJM1 plasmid using T4 DNA ligase (NEB).

Ligation reactions were transformed into TOP F’ E. coli and grown overnight on a LB + ampicillin plates to select for transformants. After preparation of plasmid DNA by miniprep

(Qiagen), the plasmid DNA was digested with NheI and EcoRI and sequenced to confirm the presence of the desired cDNA insert and to ensure there were no mutations.

3.2 Protein Expression

3.2.1 Denaturing polyacrylamide gel electrophoresis

Cells were plated at a density of 2x106 cells/dish in 10 cm dishes. The following day, cells were washed with 5ml of cold PBS. These cells were harvested with 1ml of cold PBS,

13 centrifuged at 1,000 x g for 5 minutes at 4C, and the pellets were stored at -80C until lysis. The cells were thawed on ice and resuspended in 500 L of modified RIPA buffer or lysis buffer

(Flp-In HEK293 cells) or boiled lysis buffer (MCF10A cells) with protease inhibitors (100x

AEBSF, 100X Aprotinin/Leupeptin, and 1000X Pepstatin A). The lysis buffer contains 50mM

Tris-Cl (pH 7.5), 150 mM NaCl, 5mM MgCl2, 1mM EDTA, 10mM NaF, 1% Triton X-100, and diH2O. The modified RIPA buffer contains 50mM Tris-Cl (pH 8), 150 mM NaCl, 2mM EDTA,

1% NP-40, 0.1% SDS, 10mM NaF. Cells were incubated for ten minutes on ice with frequent vortexing. The debris was spun down at 21,000 x g for 10 minutes at 4C and the supernatant containing the protein was placed in a separate tube. Supernatant (60 L) was added to 15 L of

5X SDS loading buffer and kept on ice while performing a Bradford protein concentration assay.

The Bradford protein concentration assay determined the amount of supernatant necessary to load 60mg of protein. The protein/buffer mixture was boiled at 96C for 5 minutes. The samples were then loaded onto a 12% SDS-PAGE gel and run at 200V for one hour to separate proteins.

Proteins were then transferred onto a PVDF membrane (Immobilon-P) at 100V for one hour at

4C. Membranes were then subjected to immunoblotting with anti-GFP (Roche #11814460001,

1:5000), anti-FLAG (Sigma #F3165, 1:1000), or anti-HA antibodies (Covance HA.11, 1:1000) in

5% milk with 1x Tris-buffered saline with Tween-20 (TBS-T). Membranes were washed with

TBS-T in 10-minute intervals for 30 minutes then treated with secondary antibody, anti-mouse at

1:5000. Blots were then imaged using a Bio-rad Chemidoc imaging station on the chemiluminescence setting every 30 seconds for 600 seconds.

3.3 Making Retrovirus with GP2-293 Cells

The day before transfection, 2x106 GP2-293 cells were plated in 10 cm dishes. An hour

14 before transfection, the DMEM media was changed. A mixture of 500 L of 2X HEBS (Hepes buffered saline that helps maintain physiological pH), 10 g of specific DNA, and 10 g of pVSV-G was added to a 15ml conical tube. Then 500 L of CaCl2 was added dropwise to this mixture with simultaneous aeration. This mixture was incubated at room temperature for 20 minutes. The entire mixture was then added dropwise to the cells, and the plate was gently swirled to ensure complete distribution of the mixture. These cells were incubated at 37°C in a

5% CO2 humidified incubator for 7 hours. After 7 hours, the DMEM media was changed. After

72 hours, the virus was harvested by filtering the supernatant of these cells with a 0.45m syringe filter. Virus was stored at -80C in 1ml aliquots.

3.4 Retroviral Infection

The day before infection, cells were split to be 30% confluent in T25 flasks. The day of the infection, the growth media was changed 30 minutes before infection. The media was then aspirated and 1ml of virus and 1L of 4mg/ml polybrene were added to cells. For the negative control, 1ml of growth media and polybrene was added. The T25 flasks were rocked every 15 minutes for 3 hours. The next day, the cells were split to T75 flasks. The day after splitting, the media was changed to a selective media with 15g/ml of Hygromycin B for the MCF10A cells and 100g/ml of Zeocin for the Flp-In HEK293 cells, or 1g/ml (MCF10A) or 0.25g/ml (Flp-In

HEK293) of puromycin if PSMC4 vectors were used.

3.5 Peptidase Activity Assay

Flp-In HEK293 cells expressing either GFP or gankyrin, or MCF10A cells expressing

PSMC4 and/or gankyrin or empty vector were grown in 10 cm dishes with DMEM or MCF10A

media until confluent. Two 10 cm dishes each of Flp-In HEK293 cells expressing GFP or

gankyrin were treated in the presence or absence of 1g/ml doxycycline for 48 hours. The buffer

for this assay was created with activity assay buffer (containing 50 mM Tris-HCl of pH 8.0, 10

mM MgCl2), 1mM ATP, and 1mM DTT. For each culture, 4g of protein (Flp-In HEK293 cells) or 30g of protein (MCF10A cells) was added to a variable amount of buffer to create a 100l protein-buffer solution. For each culture, 75l of protein-buffer solution and 75l of 50 M suc-

LLVY-AMC solution were mixed well. AMC has an excitation wavelength of 380 nm and an emission wavelength of 460 nm. Immediately after mixing, 40l each of this mixture were loaded into 3 wells of a 384 well plate and read using a Synergy microplate reader. A negative control containing only the buffer and suc-LLVY-AMC mixture were plated and read as well.

Using the Synergy microplate reader, we measured fluorescence every 30 seconds for 1 hour.

3.6 Anchorage Independent Growth

MCF10A cells were grown in MCF10A media without EGF in T75 plates until confluent.

Cells were mixed with 0.7% noble agar and plated at a density of 10,000 cells on top of a layer

of 1% Noble Agar in 6 well plates. Cells were fed MCF10A media without EGF and 0.7% noble

agar twice every week. After 3-5 weeks, the colonies in each well were quantified.

4. Results

4.1 Gankyrin is expressed in Flp-In HEK293 cells

To evaluate the effect of gankyrin overexpression on the peptidase activity of the

proteasome, we expressed gankyrin in Flp-In HEK293 cells. Flp-In HEK293 cells are a host cell

line generated using pFRT/lacZeo. Using Flp recombinase-mediated recombination at the

genomic FRT site in the pFRT/lacZeo plasmid, a desired DNA sequence can be “flipped in”. The

expression of this DNA is inducible by treatment with doxycycline. Doxycycline is a derivative of tetracycline and is used to activate the expression of genes in tetracycline inducible plasmids

[28]. A tetracycline-inducible promoter controls the expression of the pcDNA5/FRT plasmid containing gankyrin [8]. Using this system, we developed cells that expressed either gankyrin or

GFP as the negative control.

To verify expression of gankyrin, we treated Flp-In HEK293 gankyrin or GFP cells in the presence or absence of 1g/ml doxycycline for 48 hours and conducted a western blot (Figure

3). Gankryin expression was activated by treatment with doxycycline (Figure 3), and gankyrin was not expressed in the absence of doxycycline (Figure 3). Similarly, GFP was expressed in the presence of doxycycline (Figure 3) but GFP is still expressed when not treated with

doxycycline, suggesting there was leaky expression. However, the level of expression was less in

the absence of doxycycline compared to in the presence of doxycycline.

Figure 3. Gankyrin and GFP Expression in Flp-In HEK293 Cells Western blot analysis shows expression of FLAG-tagged gankyrin in Flp-In HEK293 gankyrin cells treated in the presence of doxycycline as indicated by the presence of a band at 25 kD. The negative control, Flp-In HEK293 cells expressing GFP, is indicated with the presence of a band at 27 kD. The expression of -actin acts as a loading control and the accurate loading of protein is indicated by the presence of the band at 37 kD.

4.2 Increased peptidase activity of Flp-In HEK293 cells overexpressing gankyrin

For the peptidase activity assay, cells were treated in the presence or absence of doxycycline for 48 hours and protein was harvested. Equal amounts of protein were combined with suc-

LLVY-AMC to evaluate peptidase activity. The proteasome degrades substrates such as suc-

LLVY-AMC. When suc-LLVY-AMC has the AMC fluorophore cleaved off by the proteasome,

AMC fluoresces and the rate of fluorescence can be measured to determine the peptidase activity of the proteasome [35]. A negative control containing only the buffer and suc-LLVY-AMC

mixture was plated and read as well. Fluorescence was measured every 30 seconds for 1 hour.

Figure 4. Peptidase Activity of Flp-In HEK293 Cells Expressing GFP or Gankyrin Peptidase activity assay analysis shows a greater increase of peptidase activity of Flp-In HEK293 cells overexpressing gankyrin in the presence of doxycycline than in its absence when compared to Flp-In HEK293 cells expressing GFP in the presence of doxycycline versus in its absence.

Cells over expressing gankyrin show an increase in peptidase activity compared to the non-

induced control (Figure 4). These results indicate that the overexpression of gankyrin in Flp-In

HEK293 cells leads to increased peptidase activity of the proteasome.

4.3 Expression of FLAG-gankyrin in MCF10A cells

A pRVYtet-off inducible expression plasmid containing gankyrin with a FLAG epitope tag was prepared through cloning. GP2-293 cells were transfected with this plasmid DNA and envelope DNA and virus was harvested after 72 hours. MCF10A cells were infected with this retrovirus and expression of FLAG-tagged gankyrin was confirmed by western blot (Figure 5).

MCF10A cells that expressed empty vector were used as a negative control (Figure 5).

Figure 5. Gankyrin Overexpression in MCF10A Cells Western blot analysis shows expression of FLAG-tagged gankyrin in MCF10A gankyrin cells with the presence of a band at 25 kD. The negative control, MCF10A empty vector cells, is indicated with the absence of a band at 25 kD. The expression of -actin acts as a loading control and the accurate loading of protein is indicated by the presence of the band at 37 kD.

4.4 No change in peptidase activity of MCF10A cells overexpressing gankyrin

For the peptidase activity assay, equal amounts of protein were combined with suc-LLVY-

AMC to evaluate peptidase activity. A negative control containing only the buffer and suc-

LLVY-AMC mixture was plated and read as well. Fluorescence was measured every 30 seconds for 1 hour. Cells overexpressing gankyrin showed no significant difference in peptidase activity compared to the empty vector cells (Figure 6). These results indicate that there may not be high levels of overexpression of gankyrin in the MCF10A cells.

Figure 6. Peptidase Activity of MCF10A Cells Expressing Gankyrin or Empty Vector Peptidase activity assay analysis shows no significant difference in peptidase activity of MCF10A cells overexpressing gankyrin compared to MCF10A empty vector cells.

4.5 Anchorage independence phenotype for MCF10A gankyrin and empty vector cells

cells in each well were quantified. Figure 7 shows the increase of MCF10A gankyrin cells

compared to MCF10A empty vector cells experiencing anchorage independent growth at week 5.

Figure 7. Anchorage Independent Growth of MCF10A Gankyrin/Empty Vector Cells Analysis of MCF10A gankyrin or empty vector cells grown in soft agar shows a greater number 20 of colonies experiencing anchorage independent growth for MCF10A cells overexpressing gankyrin compared to MCF10A cells with the pRVY empty vector. 4.6 Expression of HA-PSMC4 and HA-psmc4(337-419) in MCF10A cells

To prevent proteasome assembly and to decrease peptidase activity, we hypothesized that we

could trap gankyrin by overexpressing a minimal part of PSMC4, including the C-terminus

necessary to bind to gankyrin and prevent proper proteasome assembly [2] (Figure 8). This minimal part of PSMC4, psmc4(337-419), was chosen for this study as it was co-purified with gankyrin and used to model the gankyrin-PSMC4 interaction [3].

Figure 8. Gankyrin Trapping to Prevent Proteasome Assembly A. Gankyrin interacts properly with PSMC4, allowing for proper base assembly and proteasome assembly. B. Gankyrin is sequestered by psmc4(337-419) and cannot interact with PSMC4; therefore, the base may not assemble properly, and the base subunits may not properly interact with the core particle, preventing the assembly of mature proteasomes.

The constitutively expressed retroviral plasmid, pBabe, containing either full-length PSMC4 or psmc4(337-419), each with an N-terminal HA epitope tag, was prepared through cloning.

GP2-293 cells were transfected with either PSMC4, psmc4(337-419), or pBabe plasmid DNA, and virus was harvested after 72 hours. MCF10A empty vector cells and gankyrin cells were infected with retrovirus expressing either PSMC4, psmc4(337-419), or pBabe empty vector.

Expression of HA-tagged PSMC4 was confirmed by western blot (Figure 9). Cells infected with pBabe empty vector act as a negative control to show that the results are not an off-target effect.

Figure 9. PSMC4 Expression in MCF10A Cells Western blot analysis shows expression of HA- tagged PSMC4 in MCF10A empty vector and gankyrin cells with the presence of a band at 47 kD. The negative control, MCF10A empty vector and gankyrin cells with the empty pBabe plasmid, is indicated with the absence of a band at 47 kD. HA-psmc4(337-419) is not expressed in the MCF10A cells as indicated by the absence of a band at 10 kD. Expression of FLAG-tagged gankyrin in MCF10A cells is indicated by the presence of a band at 25 kD. The expression of -actin acts as a loading control and the accurate loading of protein is indicated by the presence of a band at 37 kD.

However, the MCF10A empty vector and gankyrin cells did not express HA-psmc4(337-

419), as indicated by the lack of bands in this western blot (Figure 9). It is possible that the protein is unstable and may be degraded.

4.7 Expression of HA-PSMC4 and HA-psmc4(337-419) in Flp-In HEK293 cells

The constitutively expressed retroviral plasmid, pLJM1, containing PSMC4 or psmc4(337-

419), each with an HA epitope tag was prepared through cloning. GP2-293 cells were transfected with either PSMC4, psmc4(337-419), or pLJM1 plasmid DNA, and virus with was harvested 22

after 72 hours. Flp-In HEK293 GFP cells and Flp-In HEK293 gankyrin cells were infected with retrovirus expressing either PSMC4, psmc4(337-419), or pLJM1 empty vector. This retroviral infection was unsuccessful, as the cells containing the pLJM1 plasmid were selected for with

100g/ml zeocin and 0.25g/ml puromycin and none survived the selection process. The exact reason for the failed infection is unknown and requires further investigation. The pLJM1 plasmid was chosen for the Flp-In HEK293 cell line because previous data showed that this cell line did not express the pBabe plasmid after transient transfection or infection but was successful in expressing the pLJM1 plasmid (data not shown).

4.8 Increased Peptidase activity of MCF10A cells expressing HA-PSMC4

For the peptidase activity assay, MCF10A cells expressing HA-PSMC4 or empty vector were grown to confluency and protein was harvested. Equal amounts of protein were combined with suc-LLVY-AMC to evaluate peptidase activity.

Figure 10 shows data for the fold increase of the rate of fluorescence of the MCF10A gankyrin HA-PSMC4, MCF10A gankyrin pBabe, and MCF10A empty vector HA-PSMC4 cells over MCF10A empty vector cells. MCF10A cells overexpressing gankyrin and the empty vector pBabe did not have a significant increase in peptidase activity when compared to the MCF10A cells with the empty vector pRVY plasmid and the empty vector pBabe plasmid. This indicates that the level of gankyrin overexpression in these MCF10A cells does not increase proteasome activity, which is consistent with previous experiments. However, MCF10A cells overexpressing gankyrin and PSMC4 have an increase in peptidase activity compared to MCF10A cells with the empty vector pRVY and empty vector pBabe. This indicates that the overexpression of both gankyrin and PSMC4 leads to increased proteasome activity. We suspect that the cell may be

23 using the full length PSMC4 as a functional protein that works properly in the proteasome, which is likely to lead to increased proteasome activity.

Figure 10. Peptidase Activity of MCF10A Cells Expressing PSMC4 and/or Empty Vector pBabe Peptidase activity assay analysis shows no significant increase of peptidase activity of MCF10A cells overexpressing gankyrin compared to MCF10A cells with the pRVY and pBabe empty vectors. MCF10A cells overexpressing gankyrin and PSMC4 showed an increase in peptidase activity compared to MCF10A cells with the pRVY and pBabe empty vectors.

4.9 Anchorage Independence Phenotype for MCF10A empty vector, gankyrin, PSMC4,

and gankyrin + PSMC4 cells

Anchorage independence assays (soft agar assays) were prepared to evaluate the appearance of the anchorage independence phenotype, a cancerous phenotype. MCF10A cells were grown in media without EGF in T75 plates until confluent. Cells were mixed with 0.7% noble agar and plated at a density of 10,000 cells on top of a layer of 1% Noble Agar in 6 well plates. Cells were fed MCF10A media without EGF and 0.7% noble agar twice every week. After weeks 3-5, the cells in each well were quantified. Figure 11 shows the increase of MCF10A gankyrin cells compared to MCF10A empty vector cells experiencing anchorage independent growth at week 3, which is consistent with the results shown in Figure 7. Additionally, Figure 11 shows the increase of the MCF10A gankyrin + PSMC4 cells compared to the MCF10A empty vector, gankyrin, and PSMC4 cells experiencing anchorage independent growth. Figure 11 also shows

that the MCF10A PSMC4 cells are experiencing an increase in the anchorage independent

growth phenotype compared to the MCF10A empty vector cells.

Figure 11. Anchorage Independent Growth of MCF10A Empty Vector/Gankyrin/PSMC4/Gankyrin + PSMC4 Cells Analysis of MCF10A empty vector, gankyrin, PSMC4, or gankyrin + PSMC4 cells grown in soft agar shows a greater number of colonies experiencing anchorage independent growth for MCF10A cells overexpressing both gankyrin and PSMC4 compared to MCF10A cells with the pRVY and pBabe empty vectors, overexpressing just gankyrin, or overexpressing just PSMC4. Data also shows a greater number of colonies of the MCF10A cells overexpressing just PSMC4 experiencing anchorage independent growth compared to the MCF10A cells with the pRVY and pBabe empty vectors.

5. Discussion

In this study, we evaluated the relationship between gankyrin overexpression and increased

proteasome activity. We successfully expressed gankyrin in Flp-In HEK293 cells and MCF10A

cells (Figure 3 and 5). Using a peptidase activity assay, we demonstrated that gankyrin overexpression leads to increased peptidase activity of the proteasome in Flp-In HEK293 cells

(Figure 4). However, gankyrin overexpression did not lead to increased activity in MCF10A

cells (Figure 6). The Flp-In HEK293 cell line allowed us to ensure that the PSMD10 gene encoding gankyrin was integrated into a single FRT site and that gankyrin is highly expressed

(ThermoFisher). The Flp-In HEK293 cell line came from a single clone that was verified to only have one single integrated FRT site. The MCF10A cells did not use the Flp-In system; therefore,

we are unsure where the PSMD10 gene encoding gankyrin was integrated and how many times it was integrated, which may affect the level of gankyrin expression. Also, the MCF10A cell line did not come from a single clone; therefore, there is the possibility for greater variability of gankyrin expression amongst the MCF10A cells. Overall, these results indicate that when gankyrin is highly overexpressed, the role of gankyrin as a chaperone protein leads to increased peptidase activity of the proteasome.

Additionally, the results from the anchorage independence assay showed that MCF10A cells overexpressing gankyrin had an increased occurrence of developing the anchorage independence phenotype than MCF10A empty vector cells (Figure 7). Because the MCF10A cells

overexpressing gankyrin did not have an increase in peptidase activity but had an increased occurrence of developing an anchorage independence phenotype, a cancerous phenotype, it is likely that the other oncogenic roles that gankyrin plays outside of proteasome assembly are causing this carcinogenic effect.

Because the MCF10A breast epithelial cells are more representative of cells that could be mutated to endogenously overexpress gankyrin and become malignant, it is imperative that this experiment be repeated using MCF10A cells that use the Flp-In system. This will allow for greater control over gankyrin expression level and will lead to more accurate results.

In this study, we also evaluated the potential of using either PSMC4 or a minimal part of

PSMC4 to prevent proteasome assembly. The proteasome inhibitors that are currently approved for use can only be used in certain blood derived cancers, have poor distribution, and all target the same PSMB5 subunit. Therefore, there is a great need for the development of new proteasome inhibitors with diverse targets [12, 13]. In order complete this analysis, we successfully expressed HA-PSMC4 in MCF10A gankyrin and MCF10A empty vector cells

(Figure 9). However, expression of the minimal part of PSMC4 (HA-psmc4(337-419)) was not detected when MCF10A cells were infected with a retrovirus containing the DNA encoding psmc4(337-419) (Figure 9). The MCF10A cells were successfully infected with the pBabe-puro-

HA-psmc4(337-419) plasmid, as indicated by their survival when treated with 1g/ml puromycin and 15g/ml Hygromycin B compared to the MCF10A cells with no plasmid. Despite this, HA- psmc4(337-419) was not detected. There may be multiple reasons for this, but we suspect that

HA-psmc4(337-419) was degraded as it is not the natural protein, may be unstable, and may not be folded correctly. In the future, we will recreate the virus containing HA-psmc4(337-419) to ensure that it was not the low level of expression in the particular virus that was obtained in this experiment that resulted in our inability to detect it in MCF10A cells. If HA-psmc4(337-419) is still not expressed, we will consider creating the HA-psmc4(337-419) or a smaller part of

PSMC4 into a cell-pentrating peptide to minimze the chance of it being degraded due to instability or incorrect folding.

Although expression of the minimal part of PSMC4 was not detected, we continued with a peptidase activity assay to test the effects of the overexpression of the full length PSMC4 protein on proteasome peptidase activity and determined that there was an increase trending toward significance in peptidase activity of MCF10A gankyrin cells expressing HA-PSMC4 compared to MCF10A cells expressing both empty vectors (Figure 10). Additionally, we conducted an anchorage independence assay to test the effects of the overexpression of the full length PSMC4 protein on the occurrence of the anchorage independence phenotype, a cancerous phenotype. We

determined that there were a greater number of colonies of the MCF10A cells that overexpressed both gankyrin and PSMC4 experiencing anchorage independent growth compared to the

MCF10A cells with both the pRVY and pBabe empty vectors, overexpressing gankyrin, or

overexpressing PSMC4 (Figure 11). We suspect that the overexpression of HA-PSMC4 increased proteasome activity and the occurrence of the anchorage independence phenotype

because it may have been used as a functional subunit in the proteasome; therefore, aiding in

proteasome function and increasing proteasome activity.

The peptidase activity assay of MCF10A gankyrin and empty vector cells expressing either

HA-PSMC4 or empty vector showed that there was no significant increase in peptidase activity

of the MCF10A gankyrin cells compared to the MCF10A empty vector cells (Figure 10), further

supporting the data found in our previous experiment (Figure 6). Additionally, the anchorage

independence assay of the MCF10A gankyrin and empty vector cells expressing either HA-

PSMC4 or empty vector showed that there was no significant increase in the occurrence of the

anchorage independent growth phenotype of the MCF10A gankyrin cells compared to the

MCF10A empty vector cells (Figure 11), again further supporting the data found in our previous

experiment (Figure 7).

Based on the results of this study, we can conclude that gankyrin overexpression in Flp-In

HEK293 cells leads to increased proteasome activity. We can also conclude that HA-PSMC4

does not trap gankyrin to prevent it from aiding in proteasome assembly and leads to increased

proteasome activity. In the future, in order to increase the possibility that a minimal part of

PSMC4 can be expressed in MCF10A cells and function to trap gankyrin, we will recreate the

virus containing HA-psmc4(337-419) to ensure that it was not the low level of expression in the

particular virus that was obtained in this experiment that resulted in our inability to detect it in

MCF10A cells. If HA-psmc4(337-419) is still not expressed, we will consider creating the HA-

psmc4(337-419) or a smaller part of PSMC4 into a cell-pentrating peptide to minimze the chance

of it being degraded due to instability or incorrect folding. If we are able to successfully express

a minimal part of PSMC4 and it proves to decrease peptidase activity of the proteasome, we will

conduct immunoprecipitation assays to determine if the minimal part of PSMC4 and gankyrin are bound together. This will ensure that our results are not caused by an off-target effect. To

further evaluate the effects of a minimal part of PSMC4 expression on the development of

cancerous phenotypes, we will conduct invasion assays. As gankyrin overexpression is indicated

as aiding in metastasis and invasion, these studies will determine if gankyrin-trapping can

prevent or limit this cancerous phenotype [4] [14-16]. Lastly, there is a need to evaluate the

effects of a gankyrin-trapping subunit on proteasome activity and the development of cancerous phenotypes in other cancer-prone tissues, as gankyrin overexpression occurs in multiple cancer types.

Our current knowledge that gankyrin overexpression leads to increased proteasome activity provides essential information regarding the role of gankyrin overexpression in the oncogenic activity of the proteasome. This information validates our hypothesis that gankyrin may be an appropriate target for limiting the increased proteasome activity that occurs in cancer cells.

Additionally, our results which indicate that HA-PSMC4 is unsuccessful in decreasing proteasome activity and that HA-psmc4(337-419) is not expressed in MCF10A cells show that more research must be done to determine if overexpressing a minimal part of PSMC4 is a valid mechanism for gankyrin trapping and proteasome inhibition. Breast cancer is the second leading cause of cancer-related deaths in women, and it is essential that we learn more about the tools these cancer cells use to survive and how we can prevent them from functioning [1].

6. Acknowledgements

I would like to express my sincere gratitude to Dr. Antonia Nemec. Dr. Nemec has guided

me throughout my time working on my Honors Thesis and throughout my entire college career. 29

Not only has Dr. Nemec given me generous support, but she has also taught me how to think,

plan, and work independently. I will always cherish the time and tremendous effort that Dr.

Antonia Nemec has put into my development as a scientist and a person. I would also like to

thank Dr. Robert J. Tomko Jr. for providing his expertise, excellent feedback, and a helping hand. I would like to thank Randi Reed and Jenny Warnock for helping me in the laboratory and always giving me encouragement. Lastly, I would like to thank Dr. Scott Stagg and Dr. Qian Yin for being on my Honors Thesis committee and taking an interest in my academic career.

7. References

1. Harbeck, N., Penault-Llorca, F., Cortes, J. et al. Breast cancer. Nat Rev Dis Primers, 2019. 5, 66 2. Howell, L.A., Tomko, R.J. & Kusmierczyk, A.R. Putting it all together: intrinsic and extrinsic mechanisms governing proteasome biogenesis. Front. Biol, 2017. 12 (1): p.19-48 3. Nemec A. A., Peterson A. K., & Warnock J. L. et al. An Allosteric Interaction Network Promotes Conformation State-Dependent Eviction of the Nas6 Assembly Chaperone from Nascent 26S Proteasomes. Cell Rep, 2019. 26(2): 483–495. 4. Kim H. Y., Kim J. H., Choi Y. W. et al. Gankyrin is frequently overexpressed in breast cancer and is associated with ErbB2 expression. Experimental and Molecular Pathology, 2013. 94(2): p. 360-365 5. Shiovitz, S. & Korde, L. A. Genetics of breast cancer: a topic in evolution. Ann. Oncol, 2015. 26: p. 1291–1299 6. Nik-Zainal, S. et al. Landscape of somatic mutations in 560 breast cancer whole-genome sequences. Nature, 2016. 534: p. 47–54 7. Yates, L. R. & Desmedt, C. Translational genomics: practical applications of the genomic revolution in breast cancer. Clin. Cancer Res, 2017. 23: 2630–2639 8. Li Z., Li Y., Gu Z., Development and verification of an FLP/FRT system for gene editing in Bacillus licheniformis. Sheng Wu Gong Cheng Xue Bao, 2019. 35(3): p. 458-471 9. Tomko, R.J., Jr. and M. Hochstrasser, Molecular architecture and assembly of the eukaryotic proteasome. Annu Rev Biochem, 2013. 82: p. 415-45. 10. Hoeller, D., Dikic, I. Targeting the ubiquitin system in cancer therapy. Nature, 2009. 458: p. 438– 444 11. Chen L., & Madura K. Increased Proteasome Activity, Ubiquitin-Conjugating Enzymes, and eEF1A Translation Factor Detected in Breast Cancer Tissue. Cancer Res, 2005. 65(13): p. 5599- 5606 12. Richardson P.G., Zweegman S., O’Donnell E.K. et al. Ixazomib for the treatment of multiple myeloma. Expert Opinion on Pharmacotherapy, 2018. 19(17): p. 1949-1968 13. Teicher B. A. & Tomaszewski J. E. Proteasome Inhibitors. Biochemical Pharmacology, 2015. 96(1): p. 1-9 14. Fu, J., Chen, Y., Cao, J. et al., p28GANK overexpression accelerates hepatocellular carcinoma invasiveness and metastasis via phosphoinositol 3kinase/AKT/hypoxiainducible factor1 pathways. Hepatology, 2013. 53: 181-192. 15. Feng H., Chen H., Yang P., et al., Gankyrin sustains PI3K/GSK-3/-catenin signal activation and promotes colorectal cancer aggressiveness and progression. Oncotarget, 2016. 7(49): p. 81156–81171 16. Riahi M. M., Sistani N. S., Zamani P. et al., Correlation of Gankyrin oncoprotein overexpression with histopathological grade in prostate cancer. Neoplasma, 2017. 64(5): p. 732-737 17. Funakoshi, M., et al., Multiple assembly chaperones govern biogenesis of the proteasome regulatory particle base. Cell, 2009. 137(5): p. 887-99. 18. Schmidt, M., et al., Proteasome-associated proteins: regulation of a proteolytic machine. Biol Chem, 2005. 386(8): p. 725-37. 19. Nandi, D., Tahiliani P, Kumar A and Chandu D, The ubiquitin-proteasome system. J. Biosci., 2006. 31(1): p. 137-155. 20. Mori, S., Chang, J., Andrechek, E. et al., Anchorage-independent cell growth signature identifies tumors with metastatic potential. Oncogene, 2009. 28: p. 2796–2805 21. Wu S. H., Sheng S. R., Liang X. H., et al., The role of tumor microenvironment in collective tumor cell invasion. Future Oncol, 2017. 13(11): p. 991-1002 22. Ataollahi M. R., Sharifi J., Paknahad M. R., et al., Breast cancer and associated factors: a review. J Med Life, 2015. 84: p. 6–11.

23. Sha Z., & Goldberg A. L., Proteasome-Mediated Processing of Nrf1 Is Essential for Coordinate Induction of All Proteasome Subunits and p97. Current Biology, 2014. 24(14): p. 1573-1583. 24. Dawson S., Apcher S., Mee M. et al., Gankyrin Is an Ankyrin-repeat Oncoprotein That Interacts with CDK4 Kinase and the S6 ATPase of the 26 S Proteasome. The Journal of Biological Chemistry, 2002. 277(13): p. 10893–10902 25. Chen, Y., Zhang, Y. & Guo, X. Proteasome dysregulation in human cancer: implications for clinical therapies. Cancer Metastasis Rev, 2017. 36: p. 703–716 26. Zamani P., Riahi M. M., Momtazi-Borojeni A. A., et al., Gankyrin: a novel promising therapeutic target for hepatocellular carcinoma. Artificial Cells, Nanomedicine, and Biotechnology, 2018. 46(7): p. 1301-1313 27. Yamtich, J., Nemec, A. A., Keh, A., & Sweasy, J. B. A germline polymorphism of dna polymerase beta induces genomic instability and cellular transformation. PLoS Genetics, 2012. 8(11) 28. Das, A. T., Zhou, X., Metz, S. W., Vink, M. A., & Berkhout, B. Selecting the optimal tet-on system for doxycycline-inducible gene expression in transiently transfected and stably transduced mammalian cells. Biotechnology Journal, 2015. 11(1): p. 71-79 29. Chen, D., Frezza, M., Schmitt, S., Kanwar, J., & Dou, Q., Bortezomib as the first proteasome inhibitor anticancer drug: current status and future perspectives. Current Cancer Drug Targets, 2011. 11(3): p. 239-253 30. Holzman, T. Svedberg unit (S). Encyclopedia of Molecular Biology, 2002. 31. Williamson, M. J., Silva, M. D., Terkelsen, J., Robertson, R., Yu, L., Xia, C., et al. The relationship among tumor architecture, pharmacokinetics, pharmacodynamics, and efficacy of bortezomib in mouse xenograft models. Molecular Cancer Therapeutics, 2009. 8(12): p. 3234- 3243 32. Thibaudeau, T. A., & Smith, D. M. A practical review of proteasome pharmacology. Pharmacological Reviews, 2019. 71(2): p. 170-197. 33. Zhen, C., Chen, L., Zhao, Q., Liang, B., Gu, Y., Bai, Z., et al.,. Gankyrin promotes breast cancer cell metastasis by regulating Rac1 activity. Oncogene, 2012. 32(29): p. 3452-3460 34. Chavez, K. J., Garimella, S. V., & Lipkowitz, S. Triple negative breast cancer cell lines: one tool in the search for better treatment of triple negative breast cancer. Breast Disease, 2011. 32(1-2): p. 35-48. 35. Powell, S. R., Davies, K. J., & Divald, A. Optimal determination of heart tissue 26s-proteasome activity requires maximal stimulating ATP concentrations. Journal of Molecular and Cellular Cardiology, 2007. 42(1): p. 265-269. 36. Yamtich, J., Nemec, A. A., Keh, A., & Sweasy, J. B. A germline polymorphism of DNA polymerase beta induces genomic instability and cellular transformation. PLoS Genetics, 2012. 8(11).

8. Supplement

8.1 Table 1- Plasmids

Name Genotype Source pRT 1213 pRVYtet-off [36] pRT 2054 pcDNA/FRT/TO-FLAG-hsGankyrin RJT Jr. unpublished pRT 2320 pBabe Addgene pRT 2321 pGEM-PSMC4 This study pRT 2331 pBabe-puro-HA-PSMC4 This study pRT 2332 pBabe-puro-HA-psmc4(337-419) This study pRT 2462 pLMJ1 Addgene pRT 2486 pLMJ1-HA-psmc4(337-419) This study pRT 2492 pLMJ1-HA-PSMC4 This study pRT 2513 pRVYtet-off-FLAG-hsGankyrin This study