Transdifferentiation of basal cells into luminal cells by regulating subtype-specific genes using lentiviral transfection

Transdifferentiation of basal cells into

luminal cells by regulating subtype-specific

Transdifferentiation of basal cells into luminal cells by

regulating subtype-specific genes using lentiviral transfection

Bachelor thesis

Author: Priyanca Asra

2030120

p.asra@student.avans.nl

Internship supervisors: Antoinette Tjon A Fat - Hollestelle (Ph.D.) a.hollestelle@erasmusmc.nl

Jingjing Liu (M.Sc.) j.liu@erasmusmc.nl

Mentor: Maaike Bergsma - van den Ham (M.Sc.)

mfe.vandenham@avans.nl

Erasmus MC Cancer Institute Medical Oncology

Wytemaweg 80 3015 CN Rotterdam Avans Hogeschool

Academie voor Technologie van Gezondheid en Milieu Biologie en Medisch Laboratoriumonderzoek

Lovendijkstraat 61-63 4817 AJ Breda

Preface

Hereby I present my Bachelor thesis “Transdifferentiation of basal cells into luminal cells by regulating subtype-specific genes using lentiviral transfection”.

Full of enthusiasm I started my graduate internship at the Medical Oncology department at the Erasmus MC Cancer Institute in August 2013. During my internship I have learned a lot about breast cancer and working in the lab with new techniques. Now seven months have passed away, I must say that completing my internship and writing my thesis would not have been possible without help. Therefore I want to thank the following people. I would like to thank Els Berns, for mentioning my name to Antoinette Hollestelle, otherwise I would never experienced this internship. I would like to thank Antoinette Hollestelle for giving me an interesting project, for her guidance, for answering my questions and for her support with writing my thesis. Also, I would like to thank Jingjing Liu, for helping me with my project and all the lab work. I would like to thank John Martens for giving me the opportunity to do my internship in his workgroup. A special thanks to all members of the Medical Oncology department for assisting me in the lab work and for giving me a pleasant time during my internship. Furthermore, I would like to thank my mentor Maaike van den Ham who gave me helpful advice during my internship.

Last, but not the least, I would like to thank my parents, family and friends for always supporting me and motivating me. Thank you!

Priyanca Asra Breda, March 2014

Abstract

Breast cancer is the most common cancer in women worldwide and like other cancer types, breast cancer is a genetic disease which arises by the accumulation of mutations in multiple genes. Different types of mutations allow abnormal expression of proto-oncogenes and tumor suppressor genes in the breast and these mutations are associated with different breast cancer subtypes: luminal A and B subtype, ERBB2+ subtype and basal-like subtype. In 2010, Hollestelle et al. found by gene expression analysis that 41 human breast cancer cell lines can be subdivided into luminal and basal phenotypes. Furthermore, 27 well known cancer genes were analysed for mutations. Some genes were mutated only in one specific subtype (i.e. luminal or basal), whereas other genes were mutated in both subtypes. In addition, breast cancer subtypes respond differently to anti-cancer treatments. For example basal-like breast cancer can now only be treated by non-targeted chemotherapy, which comes with many side effects. Therefore, it is important to develop new therapies for basal-like tumors.

The aim of the project was to investigate whether luminal subtype-specific mutations are causal in determining the luminal subtype of breast cancer. If so, this would allow us to transdifferentiate a basal subtype breast cancer cell line into a luminal subtype cell line and treat the transdifferentiated basal-like tumors with better targeted therapy. Therefore, we wanted to up- and downregulate luminal subtype-specific genes ERBB2, Cyclin D1 and MAP2K4 in the basal-like breast cancer cell line MDA-MB-157 (MM157). Using lentiviral transfection we transfected the MM157 cell line with respectively a cDNA or shRNA construct also containing a fluorescent marker. ERBB2 cDNA-CFP and Cyclin D1 cDNA-YFP were used to upregulate ERBB2 and Cyclin D1 and MAP2K4 shRNA-RFP was used to downregulate MAP2K4 in MM157 cells. To test whether the cDNA or shRNA constructs were integrated into the chromosomal DNA of MM157 cells, we analysed the CFP, YFP or RFP expression in cells by flowcytometry. The results showed that the MM157 cells transfected with a cDNA construct had no expression of CFP or YFP, while on the contrary the MM157 cells transfected with a shRNA construct did show RFP expression. The next step was to analyse up- or downregulation of the luminal subtype-specific genes. The qRT-PCR results showed that ERBB2 and Cyclin D1 were not overexpressed and MAP2K4 was not downregulated in the transfected MM157 cells. Moreover, the Western blotting results showed no upregulation of ERBB2 and no downregulation of MAP2K4 in the transfected MM157 cells. Based on these results we conclude that the transfection of MM157 cells, to overexpress ERBB2 and Cyclin D1 and downregulate MAP2K4, was not successful and therefore future research is needed.

A second project was performed based on a recent study, where TBX3 mutations were found in 2-6% of clinical breast tumors and were located only in exon 2 in ER-positive (i.e. luminal subtype) tumors. This suggests that TBX3 may be a new luminal subtype-specific gene. Therefore, our aim was to analyse all eight exons of the TBX3 gene for mutations in our collection of 56 human breast cancer cell lines using PCR and Sanger sequencing. Our results showed that no mutations are present in exons 2, 3, 4 and 6 of the TBX3 gene. Because exon 1 is very GC-rich, it was not possible to sequence exon 1 properly, even after several optimization tests using DMSO and betaine and increasing the annealing temperature. The other remaining exons (i.e. exon 5, 7 and 8) were not analysed, because some tests still have to be performed. Finally, at this moment we can only conclude that mutations are not present in exon 2, 3, 4 and 6 of the TBX3 gene in our collection of 56 human breast cancer cell lines.

Samenvatting

Borstkanker is de meest voorkomende kankersoort bij vrouwen en net als andere kankersoorten ontstaat borstkanker door de accumulatie van mutaties in verschillende genen. Verschillende soorten mutaties veroorzaken abnormale expressie van proto-oncogenen en tumorsuppressorgenen en worden geassocieerd met verschillende borstkanker subtypes: luminaal A en B subtypen, ERBB2+ subtype en basaal subtype. De studie van Hollestelle et al. (2010), heeft aangetoond dat met behulp van genexpressie profilering, 41 humane borstkankercellijnen onderverdeeld kunnen worden in de luminale en basale subtypen. Tevens werden hierin 27 bekende kankergenen op mutaties geanalyseerd. Hierdoor ontdekte men dat sommige genen alleen gemuteerd waren in het luminale of het basale subtype, terwijl andere genen gemuteerd waren in beide subtypen. Overigens is het bekend dat elk borstkanker subtype anders reageert op een anti-kanker behandeling. Bijvoorbeeld, een basale borsttumor kan alleen behandeld worden met de niet-specifieke chemotherapie, die met name bijwerkingen veroorzaakt. Mede hierdoor is het ontzettend belangrijk om een alternatieve behandeling te ontwikkelen voor basale tumoren.

Het doel van het project was het onderzoeken of de luminale subtype-specifieke genen daadwerkelijk de oorzaak zijn voor het ontwikkelen van het luminale subtype borstkanker. Als dit zo is, zou het praktisch mogelijk kunnen zijn om een basaal subtype borstkankercellijn te transdifferentiëren naar een luminaal subtype cellijn, waardoor de getransdifferentieerde basale tumoren behandeld kunnen worden met een meer gerichte therapie. In het onderzoek wilden wij de luminaal subtype specifieke genen, ERBB2 en Cyclin D1 upreguleren en MAP2K4 downreguleren in de basale borstkanker cellijn, MDA-MB-157 (MM157). Met behulp van lentivirale transfectie werd het cDNA of shRNA construct met een fluorescente marker getransfecteerd in de MM157 cellijn. ERBB2 cDNA-CFP en Cyclin D1 cDNA-YFP constructen werden gebruikt om ERBB2 en Cyclin D1 tot overexpressie te brengen en het MAP2K4 shRNA-RFP construct werd gebruikt om MAP2K4 expressie te down reguleren. Om te achterhalen of de constructen waren geïntegreerd in het genoom werd de CFP, YFP en RFP expressie geanalyseerd door middel van flowcytometrie. De resultaten toonden aan dat de getransfecteerde MM157 cellen met het cDNA construct geen CFP of YFP expressie vertoonden, terwijl de getransfecteerde MM157 cellen met het shRNA construct wel RFP expressie vertoonde. In een volgend onderzoek werd de expressie van de luminale subtype specifieke genen geanalyseerd. Uit de qRT-PCR resultaten bleek dat ERBB2 en Cyclin D1 niet tot overexpressie waren gekomen en MAP2K4 expressie niet was gedownreguleerd in de getransfecteerde MM157 cellen. Bovendien gaven de Western blotting resultaten ook weer dat ERBB2 niet was geupreguleerd en dat MAP2K4 niet was gedownreguleerd in de getransfecteerde MM157 cellen. Aan de hand van de verkregen resultaten concluderen wij dat het niet is gelukt om ERBB2 en Cyclin D1 tot overexpressie te brengen en MAP2K4 expressie te downreguleren door middel van transfectie van MM157 cellen. Uiteindelijk zullen er meerdere experimenten moeten worden uitgevoerd om het doel te behalen. Een tweede project werd uitgevoerd op basis van een recente studie, waarbij TBX3 mutaties werden gevonden in 2-6% van de klinische borsttumoren en slechts aanwezig waren in exon 2 van de ER-positieve (luminaal subtype) tumoren. Dit suggereert dat het TBX3 gen mogelijk een nieuw luminaal subtype specifiek gen is. Daarom was het doel om de acht exonen van het TBX3 gen in onze collectie van 56 borstkankercellijnen te analyseren op mutaties met behulp van PCR and Sanger sequencing. De resultaten toonden aan dat er geen mutaties aanwezig zijn in exon 2, 3, 4 en 6 van het TBX3 gen. Doordat exon 1 een GC-rijke sequentie is, was het niet mogelijk om exon 1 goed te kunnen

Table of contents

1.1 Anatomy of the normal breast and breast cancer ... 7

1.2 Genetic and environmental risk factors for breast cancer... 7

1.3 The molecular subtypes of breast cancer ... 8

1.3.1 Treatments ... 9

1.4 Genetic factors ... 9

1.4.1 Tumor suppressor genes ... 9

1.4.2 Oncogenes ... 10

2 Research goal ... 12

3 Materials and methods ... 14

3.1 Transfecting MM157 cell line and analysis ... 14

3.1.1 Puromycin kill curve ... 14

3.1.2 Lentiviral production and transduction ... 14

3.1.3 FACS analysis of transfected cells ... 15

3.1.4 Analysing mRNA expression using qRT-PCR ... 15

3.1.5 Protein isolation and Western blotting ... 16

3.1.6 STR analysis ... 17

3.2 Mutation analysis of TBX3 exons ... 18

3.2.1 PCR and Sanger sequencing ... 18

4 Results ... 19

4.1 Analysis of transfected MM157 cells with ERBB2 cDNA-CFP ... 19

4.1.1 No CFP expression in MM157 cell line transfected with ERBB2 cDNA-CFP ... 19

4.1.2 No overexpression of mRNA ERBB2 in transfected MM157 cells ... 19

4.1.3 No overexpression of ERBB2 protein in transfected MM157 cells ... 21

4.2 Analysis of transfected MM157 cells with MAP2K4 shRNA-RFP ... 22

4.2.1 Positive selection of RFP positive cells in transfected MM157 cells ... 22

4.2.2 No downregulation of MAP2K4 in transfected MM157 cells ... 23

4.3 Analysis of transfected MM157 cells with Cyclin D1 cDNA-YFP... 24

4.3.1 Transfected MM157 cells are YFP negative ... 24

4.3.2 No downregulation of Cyclin D1 in transfected MM157 cells. ... 24

4.4 DNA profiling of transfected MM157 cells versus parental MM157 cells ... 25

5 Discussion and conclusion ... 31

5.1 No overexpression of ERBB2 and Cyclin D1 in transfected MM157 cells ... 31

5.2 No downregulation of MAP2K4 in transfected MM157 cells ... 32

5.3 Mutation analysis of the TBX3 gene. ... 33

5.4 Conclusion ... 33

6 Reference list ... 34

Appendices ... 36

Appendix I: Schematic representation of lentiviral constructs ... 36

Appendix II: TBX3 genomic sequence and mutation analysis ... 39

Appendix III: cDNA sequence of ERBB2, MAP2K4 and Cyclin D1 ... 43

List of abbreviations

BCA Bicinchoninic acid

BRCA 1 Breast cancer susceptibility gene 1

BRCA 2 Breast cancer susceptibility gene 2

CCND1 Cyclin D1

CFP Cyan fluorescent protein

DCIS Ductal carcinoma in situ

ddNTPs Dideoxy nucleotides

DMSO Dimethylsulfoxide

dNTPs Deoxy nucleotides

EGFR Epidermal growth factor receptor

Env Envelop

ER Estrogen receptor

ERBB2 Epidermal growth factor receptor 2

FACS Fluorescence-activated cell sorting

FBS Fetal Bovine Serum

HEK293T/17 Human embryonic kidney 293 cells

HIV-1 Human immunodeficiency virus type 1

HPRT Hypoxanthin Phosphoribosyltransferase

IDC Invasive ductal carcinoma

IHC Immunohistochemistry

ILC Invasive lobular carcinoma

LCIS Lobular carcinoma in situ

MAP2K4 Mitogen-activated protein 2 kinase 4

MM157 MDA-MB-157

MSP Methylation-specific PCR

PBGD Porphobilinogen Deaminase

Pen/Strep Penicillin and streptomycin

PR Progesterone receptor

PTEN Phosphatase and Tensin homolog

PVDF Polyvinylidene fluoride membrane

qRT-PCR Quantitative real-time PCR

Rev Regulator of expression of virion proteins

RFP Red fluorescent protein

RPMI Roswell Park Memorial Institute medium

shRNA Short hairpin RNA

STR Short tandem repeats

Tat Trans-activator of transcription

TBX3 T-box transcription factor 3

TDLU Terminal ductal lobular unit

VSV-G Vesicular stomatitis envelope protein

1 Introduction

1.1 Anatomy of the normal breast and breast cancer

Anatomically, the breast is located in the vertical direction between the 2nd and 6th ribs and in the horizontal direction between the sternum and axillary (Figure 1A)[1]. The breast is composed of three major structures: the skin, the subcutaneous connective tissue and the corpus mammae (breast tissue). The subcutaneous tissue contains the fat and connective tissue, whereas the corpus mammae can be subdivided into stroma and parenchyma [1]. The stroma is composed of connective tissue, lymphatics, fat tissue, blood vessels and nerves. The parenchyma is the glandular tissue and each gland has 15 to 25 lobes, which are separated by the connective tissue. The gland, also called the mammary gland, consists of two epithelial cell types, the inner luminal cells and outer basal or myoepithelial cells (Figure 1B). The lobe consists of several small lobules, and these consist of branching ducts which are divided into sub segmental structures and the terminal ductal lobular unit (TDLU). The lobules are responsible for the production and secretion milk, whereas the lactiferous duct is responsible for delivering milk from the lobule to the nipple to the baby [1][2].

A. B.

Figure 1. Schematic representation of the female breast (A), and of the mammary gland (B) [3].

Approximately, 95% of breast cancers arise in the epithelia cells of the TDLU and are called carcinomas. The remaining 5% of breast cancers arise in other types of cells and are called sarcomas. Carcinoma in situ is the premalignant stage of carcinoma and is still localized to one place, whereas in invasive carcinoma the cancer starts to grow outside the original location and spread to other parts of the breast. Spreading of the tumor to other sites of the body is called metastasis. The two main histological types of carcinomas are lobular and ductal carcinoma. Ductal carcinoma in situ (DCIS) accounts for 80-85% of in situ carcinomas, while lobular carcinoma in situ (LCIS) only about 5%. For invasive carcinomas, approximately, 70-73% are invasive ductal carcinoma (IDC) and 13-16% are invasive lobular carcinomas (ILC) [4].

1.2 Genetic and environmental risk factors for breast cancer

Breast cancer is a genetic disease and is the most common cause of cancer death in women worldwide. About 1 out of 8 women will develop breast cancer during their lifetime, whereas the risk for getting breast cancer in men is 1 in 1000 [5][6]. Most breast cancers are diagnosed between 45

Women that have a strong familial breast cancer background, have a higher risk of developing breast cancer at a younger age [7]. This familial background of breast cancer is usually caused by an inherited mutation that causes susceptibility to breast cancer. Five to ten percent of breast cancers are hereditary breast cancer, where the genetic mutation is inherited from one of the parents. This means that the mutation is present at the germ line, thus, present in all cells of the body including reproductive cells [6][8][12]. On the contrary, most of the breast cancer cases are considered sporadic, were mutations in breast cancer genes only occur during the life time of the women. Because sporadic mutations are not inherited and mutations are not present in the reproductive cells, patients with a sporadic tumor cannot pass the mutation on to the next generation. Sporadic breast cancer can develop as a result of the lifestyle of women, several environmental factors, genetic mutations and hormones levels [11][12]. For example, early menarche, late menopause and overweight after menopause all increase the risk to develop breast cancer [8]. It have been known for a long time that high levels of or exposure to hormones such as estrogen and progesterone can increase the risk for developing breast cancer. Studies have shown that women with a high concentration of estrogen in their bodies have twice as much chance of getting breast cancer than women with a lower concentration [9]. Also the use of hormonal therapy during menopause has a great effect on the development of breast cancer. Furthermore, long term use of oral contraceptives increases the risk of breast cancer in women over 55 years, but not in younger women [10]. Moreover, studies have shown that pregnancy can also increase or reduce the risk of getting breast cancer. This is dependent of the age of women, total number of pregnancies and family history. When women are younger than 30 years old and become pregnant, the chance for developing breast cancer is lower than for women that become pregnant at the age of 30 years or older. Besides, the risk of developing breast cancer will be three-fold increased when women have a familial background of breast cancer and get pregnant at a later age [15][16]. Moreover, having more than one child and a longer duration of breastfeeding reduces the risk of breast cancer by 42% [15]. Also, environmental and lifestyle factors such as, heavy drinking of alcohol, smoking, exposures to radiation and electromagnetic fields give a low increased risk to develop breast cancer [13][14].

1.3 The molecular subtypes of breast cancer

There are various molecular subtypes of breast cancer. Each subtype can be classified based on the absence or expression of estrogen receptor (ER), progesterone receptor (PR) and overexpression of the epidermal growth factor receptor 2 (ERBB2). In most studies, immunohistochemistry (IHC) staining on paraffin-embedded tissue is used to examine the expression of ER, PR and ERBB2 in breast cancer. More recently, gene expression profiling was used to identify the molecular subtypes of breast cancer. Based on several studies, IHC and gene expression profiling have revealed that there are four major molecular subtypes of breast cancer: luminal A and B subtype, ERBB2+ subtype, and basal-like subtype [17]-[19].

Luminal breast cancer is the most common subtype of breast cancer. There are two different luminal subtypes: luminal A and luminal B. The luminal A subtype is characterized by high expression of ER and expression of PR and is associated with low expression of proliferation genes and more favorable prognosis. Luminal B subtype is ER, PR and ERBB2 positive or ERBB2 negative and has a higher expression of proliferation genes. Therefore it is associated with a poor prognosis [18][19]. Breast cancers of the luminal A and B subtype express the same markers as the luminal epithelial layer of normal breast ducts and can be treated by hormonal therapy, also called endocrine therapy [20]. The ERBB2+ breast cancer subtype has a more aggressive character than luminal breast cancers and is associated with amplification and/or overexpression of ERBB2 gene (HER2/neu) [18]. The ERBB2 gene is known as a proto-oncogene, which is able to control the expression of proliferation signals and cell growth of normal cells. Overexpression of ERBB2 activates growth factor signaling pathways and acts thereby as a oncogenic driver in breast cancer. Although ERBB2+ is an ER-negative subtype, it can be treated with Herceptin therapy [21][22].

Furthermore, 15% of breast cancers are of the basal-like subtype and are mostly diagnosed in younger patients. Basal-like breast cancer has an aggressive tumor phenotype and is therefore associated with poor prognosis. Compared to other subtypes, basal-like tumors have more losses or gains of DNA material. Basal-like tumors lack or have a low expression of ER, PR and ERBB2, but have a strong expression of genes characteristic of cellular proliferation and normal basal epithelial cells. In IHC studies, basal-like tumors can be identified based on the low of no expression of ER, PR and ERBB2 and on the expression of epidermal growth factor receptor (EGFR) and/or cytokeratin 5 (CK5). However, when no ER, PR and ERBB2 expression is identified in a tumor, this tumor is also defined as triple-negative. Therefore, some triple-negative tumors can be classified as basal-like tumors, and some basal-like tumors can be classified as triple-negative tumors because of the overlap in expression. Although, it should be emphasized that this is not a complete correlation [23].

1.3.1 Treatments

Several therapies are available for patients that are diagnosed with breast cancer and each breast cancer subtype is treated differently by either targeted or non-targeted treatment. Luminal A breast cancer can be treated with endocrine therapy whereas luminal B breast cancer can be treated with a combination of chemotherapy and endocrine therapy. Endocrine therapy is a targeted treatment against tumor cells with a high expression of ER. Most frequently in endocrine therapy, tamoxifen is used as a drug, because it has the ability to block or reduce the activity of ER. The benefit of tamoxifen is that it reduces the risk of recurrence of cancer by 47% and it decreases mortality by 26% [20]. Tumors with a high expression of ERBB2 (the ERBB2+ subtype) can be treated with targeted Herceptin therapy. The function of Herceptin (trastuzumab) is that it will bind to the extracellular domain of ERBB2 and will selectively reduce the tumor effects of amplified ERBB2. Besides, Herceptin reduces the risk of disease recurrence and death [22]. Unfortunately, tumors that lack or have low expression of ER or ERBB2, cannot be treated with either endocrine and/or Herceptin therapy. Since basal-like and triple-negative tumors lack or have low expressed ER, PR and ERBB2, patients with basal-like or triple-negative tumors can only respond to chemotherapy. Regrettably, chemotherapy causes a lot of side effects in patients, because the therapy is non-targeted it damages not only tumor cells, but also healthy cells. Moreover, not all patients with basal-like tumors respond to chemotherapy, therefore new therapies are being developed for this patient group [23].

1.4 Genetic factors

Like all cancer types, breast cancer is caused by the accumulation of genetic aberrations in the genome, which results in the activation or inactivation of multiple genes in the genome. For example, activation of an oncogene in a dominant way can cause normal cells to become cancer cells. Secondly, by recessive inactivation or mutation of a tumor suppressor gene, the protein’s function to decrease cell division, repair DNA mistakes or trigger apoptosis will be lost and tumorigenesis will occur [25][26]. Because of the interaction and cooperation between mutated gene products, the normal cells will transform into tumor cells. Thus, activation or inactivation of multiple genes causes tumorigenesis [24][25].

1.4.1 Tumor suppressor genes

The function of tumor suppressor genes is mainly to suppress tumorigenesis. Tumor suppressor genes are required to control the stability of cells. They are normal genes that slow down cell division, repair DNA mistakes or induce apoptosis in normal cells [24]. When most of their function is lost, they are unable to decrease the signal of tumor stimulation. In breast cancer, several tumor suppressor genes have lost their function, usually through mutation or deletions [25]. The most

BRCA1 and BRCA2 mutations are found in 10—15% of breast cancer cases with a familial background

(i.e. heredity breast cancer). For BRCA1 mutation carriers, the risk for developing breast cancer during their life time is 60-80%, and 45-60% for BRCA2 mutation carriers. However, it has also been reported that the expression level of BRCA1 is decreased in a subset of the sporadic breast cancers, particularly basal-like breast cancers. The inactivation of BRCA2, however, is rare in sporadic tumors [27][28]. BRCA1 and BRCA2 are involved in many biological processes and are tumor suppressor genes that maintain genomic stability and are activated in response to DNA damage [26][27]. In addition, they also play a role in the cell cycle and apoptosis. The functions of BRCA1 and BRCA2 are mediated by cellular proteins through the interaction with the BRCA proteins [28]. Many other tumor suppressor genes are involved in sporadic breast cancer, such as for example p53, PTEN and

E-cadherin. In this project, we will focus on the inactivation of MAP2K4 and TBX3 in breast cancer cell

lines.

MAP2K4 is a member of the mitogen-activated protein (MAP) kinase kinase family. MAP kinase

pathway signals are important for inducing cellular responses to extracellular signals, which involve growth factors, hormones, cytokines and environmental stress. MAP2K4 is known as a tumor suppressor gene [29]. In study of Hollestelle et al. mutations of MAP2K4 were found in 3 out of 42 breast cancer cell lines (i.e. MDA-MB-415, MPE600 and MDA-MB-134VI). In addition, a deletion of

MAP2K4 was identified in the DU4475 breast cancer cell line [29][30].

TBX3 is a transcription factor of the T-box gene family. TBX3 may be associated with breast cancer,

because of its overexpression in many breast cancer cell lines as well as cancer tissue. Like the study of Yaroshet al. 2008 that has shown that the tumor suppressor gene p14ARF is repressed by over-expression of TBX3 in breast cancer cell lines [33]. In a recent study, mutations of TBX3 have been found in 6 out of 100 breast tumors, all of which located in exon 2 [34]. In our lab, TBX3 mutations were found exclusively in ER-positive breast cancer specimens (luminal breast cancer). Additionally, no TBX3 protein expression was detected in the tumors of patients with mutations in TBX3, suggesting TBX3 may be a tumor suppressor gene. However, some studies have concluded that TBX3 is an oncogene. Unfortunately, the exact function of TBX3 is still not clear in breast cancer.

1.4.2 Oncogenes

Cancers arise not only by inactivation of tumor suppressor genes, but also by activation of oncogenes. Oncogenes play a central role in tumorigenesis and are also called “mutated oncogenes”. Proto-oncogenes are the “good” or non-mutated genes. The functionalities of proto-oncogenes are controlling cell division, differentiation and cell death. When proto-proto-oncogenes are mutationally activated, amplified or overexpressed by another means, these genes are becoming “bad” genes, called oncogenes. In cancer cells, the products of oncogenes are mostly overexpressed to increase cell growth and differentiation and to inhibit cell death [24][25]. In breast cancer, several oncogenes are involved, such as ERBB2, Cyclin D1, MYC, Int and EMS1 [12].This project will focus on activation of ERBB2 and Cyclin D1 in breast cancer cell lines.

ERBB2 (also known as HER2/Neu) is a member of a tyrosine kinase family that belongs to the

epidermal growth factor receptor (EGFR) family [21]. The function of members of this family is to stimulate growth factor signaling pathways [22]. In 20-40% of breast cancers, ERBB2 is overexpressed, mostly caused by amplification of the gene. Besides breast cancer, ERBB2 overexpression is also found in other types of cancers. When ERBB2 is overexpressed, transducing signals will be disturbed, proliferation will be stimulated and apoptosis will be inhibited [31].

Cyclin D1 is a product of the CCND1 gene and is amplified in about 15% of breast cancers. However,

there is around 50% of overexpression at the mRNA and protein level of Cyclin D1 in breast cancer. This overexpression can occur in presence of gene amplification of CCND1 or on the absence, such as a mutation of CCND1 gene or by other pathway mechanisms. It has been identified that the induction

of Cyclin D1 in breast cancer cell lines reduces the G1 cell cycle, thereby leading to an increase in the number of cells progressing through G1. It has also been reported that tumor cells increase proliferation by overexpression of Cyclin D1 in breast cancer. Also, Cyclin D1 strongly binds to ER, and the overexpression of Cyclin D1 leads to a good prognosis for breast cancers [11][32].

2 Research goal

Recently, 41 human breast cancer cell lines were thoroughly characterized by expression analysis, using IHC and gene expression profiling (Figure 2) [30]. The breast cancer cell lines could be classified in two main subtypes, luminal and basal subtypes. Furthermore, 27 well-known cancer genes were analysed in these breast cancer cell lines. Some genes were mutated in cell lines of only one specific subtype, whereas other mutated genes were present in cell lines of both subtypes.

Figure 2. Molecular characterization of 41 human breast cancer cell lines. Left panel indicates the classification of the cell

lines by expression analyses; L = luminal, N = no subtype and B = Basal. Middle panel shows the gene mutation analysis of the cell lines; in red M = oncogenic mutations, D = sizeable deletions and A = amplifications. The dark pink panels indicate the cell line of interest, MDA-MB-157, and genes of interest, MAP2K4, ERBB2 and Cyclin D1 [30].

The research question of this project is, are the subtype-specific genes causal in the development of these different molecular subtypes of breast cancer. If so, we assume that it would be experimentally possible to transdifferentiate a cell line with basal/ER-negative subtype into luminal/ER-positive subtype. If this is possible, this would allow patients with basal-like tumors to be treated with targeted treatments, such as hormonal or Herceptin therapy, when chemotherapy fails.

The first goal is to up- or downregulate luminal subtype-specific genes in basal cell line. Our hypothesis is that by up-or down regulation of luminal subtype-specific genes in a basal cell line, the basal cell line will be transdifferentiated into a luminal cell line. Most of the luminal cell lines have either overexpression of oncogenes ERBB2 or Cyclin D1 and mutation of tumor suppressor gene

MAP2K4. In this research project, we will use lentiviral transfection to upregulate these two

oncogenes and downregulate this tumor suppressor gene in basal breast cancer cell line MDA-MB-157 (MMMDA-MB-157). MMMDA-MB-157 is characterized as triple-negative basal cell line, and has no amplification or overexpression of ERBB2 and Cyclin D1 and mutation of MAP2K4. Since, ERBB2 and Cyclin D1 are oncogenes, we will use a lentiviral cDNA vector to overexpress these genes. MAP2K4 is a tumor suppressor gene, therefore we will use a lentiviral shRNA construct which can cause translation inhibition and mRNA degradation of MAP2K4. All information about the cDNA and shRNA constructs can be found in Appendix I. Furthermore, the lentiviral constructs will also contain a fluorescent marker and a puromycin resistence marker. After transfection of MM157 the transfected cells will be analysed by examining the expression of the fluorescent marker by FACS, and the RNA and protein levels of the gene of interest using qRT-PCR and Western blotting. When we examine that the expression of the oncogenes and tumor suppressor gene have been up-or downregulated we will further investigate the subtype of the transfected cells, by using specific markers for basal and/or luminal cells (i.e. EpCAM, CD44, CD24) and studying ER, PR, ERBB2 expression by IHC.

Furthermore, a second project will also be performed based on a recent study, where TBX3 mutations were found in 2-6% of breast cancer tumors and were located only in exon 2 in ER-positive tumors [33][34]. In our lab TBX3 mutations were also found in ER-positive/luminal breast cancer specimens only. The question is whether TBX3 is a luminal subtype-specific gene and is it causal to develop the luminal subtype of breast cancer? Regrettably, besides exon 2 the TBX3 gene has not been screened for mutations in the other 7 exons in 56 breast cancer cell lines. The sequence of the

TBX3 gene is shown in Appendix III.

The second goal is to perform mutation analysis of the entire TBX3 gene by using PCR and Sanger sequencing.

To perform the PCR reaction, specific primers will be designed that are complementary to the DNA template of interest (i.e. exon 1 and exons 3-8) and allow the specific DNA template to be amplified. Next, the size of the amplified product will be verified with gel electrophoresis. When sufficient amounts of PCR product have been formed, the amplified DNA will be used for Sanger sequencing after an Exo-SAP-IT treatment. In the sequencing reaction, one of the PCR primers will be used as the sequencing primer and dideoxy nucleotides (ddNTPs) are fluorescently labeled. After Sanger sequencing, the DNA templates will be cleaned up using ethanol precipitation and analysed on a capillary electrophoresis system. Mutations will be identified by comparing generated sequences with the reference sequence.

3 Materials and methods

3.1 Transfecting MM157 cell line and analysis

3.1.1 Puromycin kill curve

Puromycin is used in lentiviral transfection to kill all untransfected cells within 2 weeks after transfection. Because the construct that will be transfected into the cells contains a puromycin resistance gene, transfected cells will not be killed. A puromycin kill curve is performed to determine the optimal concentration of puromycin which kills all untransfected MM157 cells within two weeks. Briefly, MM157 cells were seeded in a 24 wells culture plate and cultured in standard culture medium (i.e. RPMI 1640 (1X) + glutaMAX-ITM supplemented with 10% fetal bovine serum (FBS, Biowhittaker) and 100 U/ml Penicillin/Streptomycin (Pen/Strep, Life Technologies)). The next day, cells had reached 50% confluence and puromycin (10 mg/ml, Life Technologies) was diluted in phosphate buffered saline (PBS) at concentrations: 0,1 - 0,2 – 0,3 – 0,4 – 0,5 ug/ml. The medium was replaced with 1 ml of fresh culture medium supplemented with puromycin and incubated for 1 week at 37°C. After 1 week, the medium was replaced again with fresh culture medium supplemented with puromycin and incubated for another week at 37°C. After two weeks of culturing with puromycin, cells were observed under the light microscope. The optimal concentration puromycin was selected by determining the lowest concentration of puromycin for which all cells had died.

3.1.2 Lentiviral production and transduction

The most recommended technique for overexpressing or knocking down a gene in a mammalian genome is lentiviral transduction. This technique is based on the replication of HIV-1, but modified in such a way that it is safe for use in medical research. For example, the lentiviruses which are used in transduction never carry genes that are responsible for virus replication. Furthermore, retroviral genes necessary to produce the replication deficient lentiviruses are subdivided over several vectors. For this project the third generation transduction system was used. In a third generation system, four different plasmid vectors are used, namely, a plasmid which carries the gene of interest, two packaging plasmids and an envelope plasmid. The packaging plasmid, named pMDLg/RRE encodes the capsid proteins Gag and Pol, the second packaging plasmid pRSV-REV encodes the capsid protein REV. The envelope plasmid is pMD2.g, that contains the gene VSV-g. Together, the viral proteins will made a replication deficiency lentivirus which encapsulates the gene of interest. Information on the plasmids are shown in Appendix I. The produced lentiviruses deliver the gene of interest into MM157 cells and enable incorporation in the host genome. All lentiviral experiments were performed under ML-II safety conditions.

Lentiviral production and concentration

In a 9 cm culture dish HEK293T/17 cells were cultured in standard culture medium and incubated for 24 hours to reach 75-85% confluency. Prior to transfection, the transfection mixture was prepared in a 50 ml conical tube per culture dish. For each culture dish, 20 µg construct with gene of interest (Table 1, Appendix I), 10 µg pMDL/RRE, 5 µg pRSV-REV, 5 µg pVSVG were mixed in 500 µl Hepes Buffered Saline (HBS, 20 mM Hepes, 150 mM NaCl, pH 7.0). For each transfection mixture, 13,3 µg Polyethylenimine (PEI, Sigma) solution was mixed in 500 µl HBS. 500 µl PEI solution was added dropwise to 500 µl transfection mixture, while gently shaking the tube, and incubated for 15 minutes at room temperature. In the meantime, culture medium of HEK293T/17 cells was replaced with 8 ml serum-free RPMI-glutaMAX-I medium. After incubation, the transfection mixture was added drop wise to HEK293T/17 cells and incubated for 3 hours at 37°C. After incubation, medium was replaced with 10 ml standard culture medium and incubated for 24 hours at 37°C. The next day, medium was replaced with 6 ml fresh culture medium. After 24 hours, virus-containing medium was collected with a 20 ml syringe and filtered through a 0.45 µm cellulose acetate filter and stored at 4°C. Again 6 ml fresh culture medium was added to HEK293T/17 cells and incubated for another 24 hours. Next day, the virus-containing medium was collected as before. Together, 36 ml virus solution was

collected per construct (i.e. 3 culture dishes). Before transduction, the lentiviruses were concentrated by using PEG-itTM Virus Precipitation solution (System Biosciences). Therefore 1 volume of cold PEG-it Virus Precipitation Solution was added to every 4 volumes of lentiviral containing solution and stored for up to 2 days at 4°C. After 2 days, lentiviral containing solution/PEG-it mixture was centrifuged at 1500 rcf for 30 minutes at 4°C. Supernatant was aspirated and pellet was again centrifuged at 1500 rcf for 5 minutes. All fluid was removed and lentiviral pellet was resuspended in 1/100 of original volume using cold, sterile PBS and stored at -20°C.

Lentiviral transduction of MDA-MB-157 cells

MM157 cells were seeded in 0,5 ml standard culture medium per well in 24 well culture plates and incubated for 24 hours to reach a 50% confluency. The concentrated virus solution was centrifuged before use at maximum speed for 1 minute. For each transduction, 70 µl concentrated virus solution was added to cells and incubated for 8 hours at 37°C. At the end of the day, medium was replaced with 1 ml standard culture medium supplemented with 0,3 µg/ml puromycin. Cells were cultured at 37°C and every week the standard culture medium with puromcyin was refreshed. After fourteen days cells were further cultured under ML-1 safety conditions.

Table 1. Information on constructs that were used for transfection.

Gene of interest Construct name µg/ul Positive control; construct name µg/ul

Cyclin D1 EX-B0078-Lv123 1.9 YFP; EX-EYFP-LV105 2.0

MAP2K4 pLKO.1-puro CMV-TagRFP 2.5 RFP; SHC012-pLKO.1-puro CMV-TagRFP 2.5

ERBB2 EX-Z2866-Lv127 1.8 CFP; EX-ECFP-LV105 1.9

3.1.3 FACS analysis of transfected cells

Next, parental and transfected MM157 cells were analysed by FACS, to confirm the expression of the fluorescent marker (i.e. CFP, RFP, YFP) in the puromycin-selected cells. When a substantial amount of the cells (>30%) were negative for the fluorescent marker the positive expressed cells were sorted out from the negative expressed cells. Prior to cell sorting, cells were grown to 70-80% confluence in a T75 culture flask. For determining the amount of positive cells in a cell population cells were grown to 70-80% confluency in a T25 culture flask. Further preparation of cells for both analyses were the same. Cells were washed with PBS and then trypsinized. When detached, cells were collected in RPMI medium in a total volume of 10 ml and centrifuged at 1000 rpm for 5 minutes. Cells were washed with PBS and centrifuged at 1000 rpm for 5 minutes. Then the cell pellet was resuspended in 500 µl PBS and placed in a 5 ml polystyrene falcon tube with filter top (Becton Dickinson) for analysis. Before FACS analysis, living cells were stained with Hoechst (Life Technologies), which was diluted to 1:100. FACS analysis was done by plotting the SCC against fluorescence intensity and examine the amount positive and negative cells.

3.1.4 Analysing mRNA expression using qRT-PCR

To examine upregulation of ERBB2 and Cyclin D1 and downregulation of MAP2K4, quantitative real-time PCR (qRT-PCR) was used. First total RNA of parental and transfected MM157 cells was isolated using RNeasy Mini Kit (Qiagen), see protocol in Appendix IV. RNA was then reverse transcribed into cDNA to perform the qRT-PCR method as written below.

cDNA synthesis

cDNA was generated from 2 µg of total isolated RNA. In the primer hybridization step, RNA was mixed in equal amounts with 0,5 µg oligo(dT)18 primers and 0,2 µg random hexamer primers, incubated for 5 minutes at 70°C and subsequently cooled down to 0°C. Next, for each sample a cDNA

H (Ambion) for 30 minutes at 37°C and then kept at 0°C. Finally, 21 µl DNase, RNase free water was added to each sample and samples were stored at -20°C.

qRT-PCR

A qRT-PCR reaction included 1x ABgene SYBR Mastermix, 0.25 µM forward primer, 0.25 µM reverse primer and 5.0 µl cDNA (20x diluted) in a total volume of 25 µl. The qRT-PCR reaction started with 15 minutes for activation of Taq polymerase at 95°C. Then followed by 40 PCR cycles of 15 sec at 95°C, 30 sec at 62°C, 30 sec at 72°C, and 30 sec at 79°C. qRT-PCR was performed using a Stratagene Mx3000P Real-time PCR system (Agilent Technologies). The data was analysed using the MxPro-Stratagene Mx3000P qPCR software v4.10. First, SYBR green fluorescent was normalized by passive reference dye, ROX, which was included in the SYBR green PCR master mix. For normalization the ROX fluorescent was to remain constant in each PCR cycle. Next, by plotting the ΔRn against the PCR cycle number and setting the threshold at 0.02 in the exponential phase, the threshold cycle (Ct) value was determined. For normalization, housekeepers genes porphobilinogen deaminase (PBGD) and hypoxanthin phosphoribosyl transferase (HPRT) were taken. HPRT was normalized to PBGD, by taking the value of the average HPRT Ct value (Ct calculated by 72°C) minus the Ct average of PBGD (Ct calculated by 79°C), and sum that value in each sample with Ct value of HPRT. Then the average of HPRT and PBGD were taken to normalize the Ct value of the gene of interest for each sample. Several primer sets were designed for each gene of interest based on the protocol Rules of PCR primer design (Appendix IV). An overview of the used qPCR primer sequences is shown in Table 2, the remaining primer sets for the gene of interest are shown in Appendix III. The primer combinations were tested by examining the primer specificity and formation of primer dimers, and calculating the PCR efficiency. In a dissociation plot, whereby the fluorescence is plotted against the temperature (i.e. melting curve), we examined specificity and if primer dimers were formed, by looking out for multiple peaks in the melting curve. If only one peak was observed in the dissociation curve the primers were specific for the gene of interest. Next, to verify the PCR efficiency, a mixture of cDNA of 41 human breast cell lines (P1) was taken along, and a four-fold dilution series of P1 (i.e. P2-P5). For calculating the PCR efficiency we had used the following formula: E = [10 (-1/slope)] – 1. The PCR efficiency should between 90% and 110% which corresponds to a slope between -3.58 and -3.10.

Table 2. Sequences of forward and reverse primers for qRT-PCR.

Primer name Primer sequence (5’-3’) Length (bp)

ERBB2_Exon22F ERBB2_Exon24R GTCTACAAGGGCATCTGGATC CATCACGTATGCTTCGTCTAAG 123 MAP2K4_Exon9F MAP2K4_Exon11R CGAGTTTCATCAACTTTGTCAAC CGACCTCAACGGCACGTTC 120 Cyclin D1_Exon1F Cyclin D1_Exon2R CCTGCCGTCCATGCGGAAG AGAGGCCACGAACATGCAAG 178

3.1.5 Protein isolation and Western blotting

Protein expression in the transfected cell lines was determined by Western blotting. The proteins of parental and transfected MM157 cells were isolated by using M-PER Mammalian Protein extraction reagent. The total isolated protein concentration was determined spectrophotometrically using the Pierce BCA protein assay kit (Thermo Scientific). The BCA assay was performed according to manufacturer’s instructions, see Appendix I.

Prior to Western blotting, samples were prepared with NuPAGE LDS sample buffer and Reducing Agent and then incubated at 95°C for 15 minutes. After incubation, 6 µg of each protein sample was loaded on a 4-12% NuPAGE Novex Bis-Tris Gel with 1.0 mm wells (Life Technologies) and 5 µl of Spectra Multicolor Broad Range Protein ladder was taken as size marker. The gel was placed in a Xcell SureLock electrophoresis system, whereby the inner compartment was filled with 200 ml of 1x MOPS SDS Running Buffer (Life Technologies) with 250 µl NuPAGE Antioxidant (Life Technologies). The

outer compartment was filled with 1x MOPS SDS Running buffer only for cooling. Next, the gel was ran at 125 V for 1,5 hours. The size separated proteins in the gel were transferred to a polyvinylidene fluoride membrane (PVDF) by electroblotting for 1,5 hours at 225 mA using an Owl separation system. The PVDF membrane was then blocked with NFDM blocking solution (5% non-fat dry milk in 10 ml 1x TBS + 0.1% Tween-20 (TBS-T)) for 1 hour at room temperature. Then, the membrane was incubated overnight at 4°C with primary antibody in BSA blocking solution (5% BSA in 10 ml 1x TBS-T). Information on antibodies can be found in Table 3. After incubation, the membrane was washed 5 times for 5 minutes with TBS-T. Then secondary antibody diluted in NFDM blocking solution was added to the membrane and incubated for 1 hour at room temperature. The membrane was washed 5 times for 5 minutes in TBS-T and 5 minutes with TBS. Finally, the membrane was incubated for 1 minute with a 1:1 mixture of ECL (Thermo Scientific) and then covered in plastic foil. The membrane was then placed in a film cassette and developed in the KODAK X-OMAT 1000 processor. The developed film was scanned and the bands were quantified using the 1D scana EX 3.0 program. Using 1D scana EX 3.0 program, the picture was edited as a black and white picture, and pictures were saved as TIF files. Furthermore, housekeeper protein β-actine was used for normalization. Therefore, the band intensity of the protein of interest (e.g. MAP2K4, ERBB2 and Cyclin D1) was divided against the band intensity of β-actine.

Table 3. Information on the primary and secondary antibodies which were used.

Primary antibody Type Dilution Company

anti MAP2K4 Rabbit polyclonal 1:250 Cell Signalling

anti ERBB2 Mouse polyclonal 1:1000 Dako

anti Cyclin D1 Mouse polyclonal 1:2000 Cell Signalling

anti β-actine Mouse polyclonal 1:2000 Sigma-Aldrich

Secondary antibody Type Dilution Company

Goat anti rabbit Polyclonal 1:2000 Dako

Rabbit anti mouse Polyclonal 1:5000 Dako

3.1.6 STR analysis

After transfecting the MM157 cell line, we wanted to confirm that the transfected cells are still identical to the parental cell line to exclude cell line cross-contamination. Therefore STR profiling was performed. The human genome contains thousands of short tandem repeats (STR), which are stretches of DNA that vary in length between individuals. By looking at multiple STR loci simultaneously a unique STR profile is revealed for each individual and thus for every breast cancer cell line. In a STR analysis, various STR loci are amplified using various specific primers which contain different fluorescent labels. The advantage of using different fluorescent labels is that many loci can be detected in one single reaction. After amplification of the STR loci, PCR products will be separated by length in a capillary gel electrophoresis system and the fluorescent labels can be detected. Furthermore, an allelic ladder is also included in the reaction, which is a mix of common STR alleles, whereof the number of repeat units is already known.

Prior to STR analysis method, the DNA of untransfected and transfected MM157 cell lines was isolated by using the DNA isolation DNeasy minikit (Qiagen, Appendix IV) and the quantity of DNA was measured on a Nanodrop (Spectrophotometer, ND100). Briefly, each 10 ng of DNA sample was prepared by adding 1x Gold Star Buffer (Powerplex 16, Promega), 1x primer pair mix (Powerplex 16, Promega) and 1.65 U AmpliTaq Gold (Applied Biosystems) in a total volume of 10 ul. The PCR reaction was as follows, incubation of 11 minutes at 95°C, 1 minute at 96°C, followed by 10 cycles of

parental MM157 cell line and the transfected MM157 cells, Genemarker software (Softgenetics) was used. The data was presented by plotting the fluorescent signal against the length of DNA in an electropherogram. The separated size fragments of each sample were converted into alleles by comparison to the allelic ladders (positive control). The function of the allelic ladder is also to ensure that the peaks in the software are correctly demonstrated. Then, the number of repeats of each loci of transfected MM157 cell lines as compared to the parental MM157 cell line as a reference.

3.2 Mutation analysis of TBX3 exons

For the second goal, mutation analysis of the TBX3 gene, the main techniques are PCR and Sanger sequencing. Using the method of Sanger sequencing allows for the analysis of the nucleotide order of a DNA sequence and subsequent screening for mutations in DNA sequences. First a PCR reaction will be performed and the PCR product will be analysed with gel electrophoresis. After PCR reaction an Exo-SAP-IT treatment was performed to remove the remaining dNTPS and primers of the PCR reaction. Then the treated Exo-Sap-IT PCR product can be sequenced.

3.2.1 PCR and Sanger sequencing

First the DNA samples of the 56 breast cancer cell lines will be amplified. For DNA amplification, the PCR reaction mixture contained 20 ng of DNA sample, 1x GoTaq PCR buffer (Promega), 1.5 mM MgCl2, 200 µM dNTPs (GE Healthcare), 1 µM forward and 1 µM reverse primer in a total volume of 10 ul. DNA was denatured at 94°C for 2 minutes. After denaturation, Taq polymerase enzyme mixture containing 1x GoTaq PCR buffer (Promega) and 0.25 U GoTaq (Promega) in a total volume of 5 ul, was added to each reaction. The cycling conditions were 35 cycles of 94°C for 30 sec, 58°C for 1 minute and 72°C for 1 minute. After cycling, samples were incubated at 72°C for 5 minutes and then kept at 25°C. Finally, 2 µl of each PCR product, including 9 µl loading dye (OG2X), was loaded on an 1,5% agarose gel and ran for about 1 hour at 100 volt. After the PCR reaction, the samples were purified before moving to sequencing. Therefore, for removing left over primers and dNTPs, Exo-SAP-IT treatment was performed. 1x GoTaq PCR buffer (Promega) and 0,5 U Exo-SAP-IT (Affymetrix), in a final volume of 10 µl was added to 2,5 µl PCR product. The mixture was incubated for 15 minutes at 37°C and then the enzymes were inactivated for 15 minutes at 80°C.

For sequencing, 2,0 µl Exo-SAP-IT treated PCR product was mixed with 1x Big Dye Terminator Reaction mix (Applied Biosystems), 1x Sequencing buffer (Applied Biosystems) and 0.16 µM forward or reverse primer in a final volume of 10 µl. Samples were denatured at 96°C, followed by 25 cycles of 96°C for 30 sec, 58°C for 30 sec and 60°C for 2 minutes, and then kept at 4°C. After sequencing, ethanol/sodium acetate precipitation was performed. 10 µl of sequencing product was mixed with 90 µl sequencing precipitation mixture (i.e. 100 ml milli-Q, 30 ml 3M NaAc (pH 4.6), 770 ml 95% EtOH) and centrifuged at 3000 g for 5 minutes at room temperature. Supernatant was discarded by shortly spinning the inverted plate. The pellet was washed with 150 µl 70% ethanol, centrifuged at 3000 g for 1 minute and again supernatant was discarded. Pellet was resuspended in 20 µl Hi-Di formamide and denatured at 94°C for 2 minutes and then directly stored at 4°C until samples were processed by the ABI Prism 3130xl Genetic analyzer (Applied Biosystems).

For analysing mutations in the samples, the software program Mutation Surveyor (Softgenetics) was used. The reference sequence of each exon was obtained from Ensembl. The software alignes the electropherograms of each sample to the reference sequence and automatically selects one sample as the reference sample (i.e. sample with the highest quality). Mutations are then identified by comparing the peak heights for each base in the electropherogram of the sample with the peaks height for each base in the electropherogram of the reference sample. The differences between the sample and the reference sample will be displayed, whereby a mutation is assigned when this difference reaches a certain threshold. Differences below the threshold are considered normal variation between experiments.

4 Results

4.1 Analysis of transfected MM157 cells with ERBB2 cDNA-CFP

4.1.1 No CFP expression in MM157 cell line transfected with ERBB2 cDNA-CFP

The results of the puromycin kill curve showed that the transfected MM157 cells should be cultured in standard culture medium with 0,3 µg/ml puromycin. To overexpress the ERBB2 in the MM157 cell line, cells were transfected with a ERBB2 cDNA-CFP construct. As a control, MM157 cells were transfected with a CFP construct. After selection of transfected MM157 cells by puromycin, we investigated CFP expression in the transfected MM157 cells, using flow cytometry. The results are shown in Figure 3 were the untransfected parental MM157 cells were taken as a reference and were CFP negative. Next, we examined 100% CFP positive cells in the transfected MM157 CFP cells, which clearly indicates that the cells were successfully transfected. However, when we examine the CFP expression in MM157 cells transfected with ERBB2 cDNA-CFP, all cells were CFP negative. This suggests that the ERBB2 cDNA-CFP construct is not transfected properly into the MM157 cells or that the transfection was successful but the CFP protein is not detectable by the flow cytometry. To examine why the cells are not CFP positive we further analysed the expression of ERBB2 in the transfected MM157 cells.

Parental MM157 cells MM157 ERBB2 cDNA-CFP cells MM157 CFP cells

Figure 3. MM157 cells transfected with CFP control construct are CFP positive but MM157 cells transfected with ERBB2 cDNA-CFP construct are CFP negative. In the dotplot the Alexa Fluor 430-A (CFP) is plotted against the side scattered light

(SSC-A). The high expressed CFP are gated, CFP+ (Viogreen-SSC).

4.1.2 No overexpression of mRNA ERBB2 in transfected MM157 cells

Our FACS data revealed that no CFP positive cells were determined in transfected MM157 cells with ERBB2 cDNA-CFP. However, the FACS data gives no information about the expression of ERBB2 in transfected MM157 cells. To examine whether the mRNA of ERBB2 was upregulated or not in the transfected MM157 cells, qRT-PCR was performed. We tested several primer combinations designed against ERBB2. In this test we investigated the specificity of the primers and the formation of primer dimers. Next, we also examined the PCR efficiency through a standard curve. Figure 4 shows an example of two primer sets for ERBB2: primer set 4 (X19F+X20R) and primer set 5 (X22F+X24R). In the amplification plot, a doubling of the Ct value by every 4-fold dilution step is shown, which indicates that the PCR efficiency is around 100%. Furthermore, a genomic DNA sample was taken as a negative control for which we expected no amplification of DNA. Although genomic DNA is amplified by primer set 4 (shown by black arrow, Figure 4A) and not by primer set 5. This result indicates that primer set 5 is a more specific primer set to amplify the ERBB2 cDNA. Next, we examined the dissociation curve and observed multiple peaks are formed with primer set 4 (see red arrow, Figure

100 % 0.7 %

Amplification plot Dissociation curve A. Primer set 4 (X19F+X20R)

B. Primer set 5 (X22F+X24R)

Figure 4. An example of two primer sets for ERBB2. Amplification plot and dissociation curve of dilution series P1 to P5

which has been amplified with primer set 4 (A) and 5 (B) are shown. The amplification plot is the plot of the number of cycles (x-as) versus the fluorescence (y-as). In the dissociation plot the temperature (x-as) is plotted against the fluorescence (y-as). The black arrow shows the amplified genomic DNA and the red arrow the primer dimers.

Several other primer sets for ERBB2 were also tested by measuring the PCR efficiency and examining the dissociation curve. All designed primers can be found in Appendix III. Based on these data, primer set 5 was selected as the best primer set for amplification of ERBB2 cDNA. The primer sequence is: exon 22/23 forward primer 5’-GTCTACAAGGGCATCTGGATC-3’ and exon 23/24 reverse primer 5’CATCACGTATGCTTCGTCTAAG-3’. Furthermore, based on the dissociation curve (Figure 4B), we examined that the melting temperature of the PCR products is 82°C. Since the standard measuring temperatures in the PCR program are 72°C and 79°C, and we observe no primer dimers or aspecific products below 82°C in the dissociation curve, the Ct value of ERBB2 could also be measured at 72°C or 79°C.

Next, we analysed the ERBB2 expression of transfected MM157 cells by qRT-PCR. Before analysing the ERBB2 mRNA expression, we verified the PCR efficiency and then normalized each sample against

HPRT and PBGD. In Figure 5, the amplification plot and normalized Ct value of ERBB2 for the

untransfected and transfected MM157 cells with ERBB2 cDNA-CFP and CFP are shown. Examining the amplification curve, we observe that the mRNA expression of ERBB2 in the untransfected and transfected MM157 cells are not different. However, this value is not normalized against the housekeeper genes. After normalization, the Ct values of each sample are plotted in a graph, see Figure 5B. The untransfected parental MM157 cell line was taken as the reference. The data represents that the mRNA expression of transfected MM157 cells with CFP are approximately equal to the untransfected parental MM157 cell line. Our hypothesis was that if ERBB2 is overexpressed in the transfected MM157 cell line, the expression should be at least 5 fold increased compared to the parental and CFP transfected MM157 cell line. However, when we examine the Ct value of the transfected MM157 cell line with ERBB2 cDNA-CFP, the data represents that the Ct value is one Ct lower compared to the references, which indicates that the expression is only 2 fold increased. Thus, the results indicate that the mRNA expression of ERBB2 is doubled in transfected MM157 ERBB2

cDNA-CFP. Unfortunately, we assume that the ERBB2 is not sufficiently overexpressed by the transfection of MM157 cells with ERBB2 cDNA-CFP.

A. B.

Figure 5. ERBB2 mRNA levels are slightly increased in transfected MM157 cells with the cDNA ERBB2-CFP construct. A, In

the amplification plot the untransfected parental and transfected MM157 cell lines are shown (in triplicate) by plotting the fluorescent against the cycle number. The PCR efficiency of qRT-PCR was approximately 110%. B, The graph shows the different MM157 cell lines plotted against the Ct value. The Ct value of ERBB2 was normalized against the average of Ct value of HPRT and PBGD. The Ct value of ERBB2 was measured at 72°C.

After qRT-PCR, the next step was to analyse if also the protein level of ERBB2 in transfected MM157 cells are equal or different as compared to the parental MM157 cells.

4.1.3 No overexpression of ERBB2 protein in transfected MM157 cells

To analyse the ERBB2 protein level in the parental and transfected MM157 cells, Western blotting was performed. Based on the mRNA expression level in the transfected cells, our expectation was that the protein level of ERBB2 was also not overexpressed in the transfected MM157 cell line with ERBB2 cDNA-CFP. In Figure 6A we observe that ERBB2 and β-actin protein is expressed in untransfected parental MM157 cells and in both transfected MM157 cell lines. At first glance, the band densities of ERBB2 are equal in the parental and transfected MM157 cells. After ERBB2 was normalized against β-actin (Figure 6B), data represents that the ERBB2 protein expression is slightly increased in transfected MM157 CFP. However, our hypothesis was that for an overexpression of ERBB2 in MM157 cells the protein expression should be at least 5 times higher than the parental MM157 cells. Regrettably, these results indicate that the ERBB2 protein expression is not increased in the transfected MM157 cells.

A. B. 23,58 _22,57 23,32 0 5 10 15 20 25 30 35

Parental ERBB2 cDNA-CFP CFP

C t val ue MM157 cell lines 100% 111% 149% 0% 20% 40% 60% 80% 100% 120% 140% 160% 180%

Parental ERBB2 cDNA-CFP CFP

Ex pr e ss ion (% )

Based on the qRT-PCR and Western blotting results we have confirmed that the transfection of MM157 cells with ERBB2 cDNA-CFP was not successful, because no overexpression of ERBB2 was observed in the transfected MM157 cells with MAP2K4 shRNA-CFP construct.

4.2 Analysis of transfected MM157 cells with MAP2K4 shRNA-RFP

4.2.1 Positive selection of RFP positive cells in transfected MM157 cells

To downregulate MAP2K4 in MM157 cells, cells were transfected with a MAP2K4 shRNA-RFP construct and control RFP construct. To confirm expression of RFP in transfected MM157 cells, puromycin selected cells were analysed by flow cytometry. The RFP fluorescence was analysed in the PE channel of the FACS. The results are shown in Figure 7. There is no RFP expression detectable in parental MM157 cells, as expected. However, there is expression of RFP detectable in both transfected MM157 cell lines (MAP2K4 shRNA-RFP and RFP). The results indicate that 94.5% of MM157 cells with RFP are RFP positive, whereas MM157 cells with MAP2K4 shRNA-RFP are 86.3% RFP positive. Next, we selected cells with the highest RFP expression (shown as red dots, Figure 7A) by sorting the cells with the FACS and cultured these cells again. After one month, a second FACS analysis was performed to investigate if the transfected MM157 cells are still RFP positive or not. Results are shown in Figure 7B. Our expectation was that both the transfected MM157 cell lines would be 90-100% RFP positive, but instead we observe again low expressed RFP cells in the transfected cells. This may indicate that the transfected MM157 cells with a low RFP expression are in a different stage of cell cycle than the transfected MM157 cells with a higher RFP expression.

Parental MM157 cells MM157 MAP2K4 shRNA-RFP cells MM157 RFP cells

A.

B.

Figure 7. Expression of RFP in transfected MM157 MAP2K4 shRNA-RFP cells and MM157 RFP cells. The dotplot represents

RFP expression of cells by plotting the FITC versus RFP fluorescence. A, First analysis of the parental and transfected MM157 cells. B, After one month of culturing, a second FACS analyse was performed of the sorted RFP positive MM157 cells. 69.5% 94.5% 83.1% 54.3% 86.3% 57.3%

4.2.2 No downregulation of MAP2K4 in transfected MM157 cells

Based on the FACS analysis, we known that the transfected MM157 cells are RFP positive. This suggests that the construct of MAP2K4 shRNA-RFP and RFP has been inserted into the genome, and that the MAP2K4 expression in transfected MM157 cells with MAP2K4 shRNA-RFP should be downregulated. By performing a qRT-PCR and Western blotting analysis, we examined the mRNA expression and protein expression of MAP2K4 in the parental MM157 cell line and both transfected MM157 cells.

For qRT-PCR analysis, several MAP2K4 qRT-PCR primers were designed and tested for primer dimers and PCR efficiency as described in paragraph 4.1.2. From the 5 different primer sets, see Appendix III, primerset 5 (X9F1+X11R1) was selected as the best primerset to amplify the MAP2K4 cDNA. The primer sequence is: exon 9 forward primer 5’-CGAGTTTCATCAACTTTGTCAAC-3’ and exon 11 reverse primer 5’-CGACCTCAACGGCACGTTC-3’. In Figure 8A the results of mRNA expression are shown. We observe that there is a variation in Ct value between the parental MM157 cells and transfected MM157 cells (Figure 8A). However, the next step was to normalized the Ct value against PBGD and

HPRT. These results are plotted in Figure 8B and indicate that the mRNA expression level of

untransfected parental MM157 cells and transfected MM157 with RFP are approximately equal to each other. The company (Sigma) had guaranteed that the shRNA-MAP2K4-RFP construct will knock-down MAP2K4 by 88%, which is equal to a 2 Ct change and 4 fold decreasing. However, we examine that the MAP2K4 expression of MM157 cells with MAP2K4 shRNA-RFP transfected is only 2 fold decreased compared to the parental and the RFP transfected MM157 cell line. Thus, the results indicate that MAP2K4 is not sufficiently down regulated in the transfected MM157 cells with MAP2K4 shRNA-RFP. A. qRT-PCR B. Western blotting 23,47 24,91 23,58 0 5 10 15 20 25 30 35

Parental MAP2K4

shRNA-RFP RFP C t val ue MM157 cell lines 100% 84% 87% 0% 20% 40% 60% 80% 100% 120%

Parental MAP2K4 shRNA- RFP

Ex pr e ss ion (% )