Vitamin C as a modifier of mammalian epigenetics: implications for adaptive immunity

Department of Clinical and Experimental Medicine

Master’s Thesis

Vitamin C as a modifier of mammalian epigenetics:

implications for adaptive immunity

Jonna Håkman

LiU-IKE-EX—13/09

Department of Clinical and Experimental Medicine Linköpings universitet

Department of Clinical and Experimental Medicine

Master’s Thesis

Vitamin C as a modifier of mammalian epigenetics:

implications for adaptive immunity

Jonna Håkman

LiU-IKE-EX—13/09

Supervisor: Colm Nestor, Centre for Individualized Medicine,

Department of Pediatrics

Examiner: Jonas Wetterö

Department of Clinical and Experimental Medicine Linköpings universitet

A

BSTRACT

Ascorbic acid (AA), in popular speech vitamin C, is a commonly known nutrient. It is involved in several biological processes and deficiency can lead to scurvy. Recent publications have shown the impact of AA on epigenetic regulation in mice. Addition of AA, via enzymatic activity, enhances the generation of 5-hydroxymethylcytosine (5hmC), which is an intermediate in active demethylation of DNA.

The role of AA on epigenetic changes in humans has to our knowledge never been studied. In this study, naïve CD4+ T cells from blood donors were used as a model system to investigate AAs possible role in methylation changes in the immune system. By using dot-blot assay, hydroxymethylated DNA immunoprecipitation (hmeDIP) and qPCR, changes in methylation executed by AA could be detected. A confirmation of AAs impact on epigenetic changes in mice was observed. AA enhanced the levels of 5hmC compared to untreated cells. The Jurkat cell line, a human T lymphocyte cell line, showed an opposite result. Treatment with AA decreased the levels of 5hmC compared to untreated cells. When comparing this result with the results obtained in human naïve T cells, the same observation was made. The difference between mouse and human in the ability of producing and metabolize AA could be a reason for this opposite result.

Since AA had the ability to modify epigenetic changes in primary human CD4+ T cells, the results suggest that AA may have a function in the human immune system.

S

AMMANFATTNING

Askorbinsyra (AS), i folkmun kallad vitamin C, är ett allmänt känt näringsämne. AS är involverad i flera biologiska processer och brist kan leda till skörbjugg. Nyligen publicerade artiklar har visat att AS har en roll i epigenetisk reglering hos möss.

Vid addering av AS, via enzymatisk aktivitet, ökar produktionen av 5-hydroximetylcytosine (5hmC), som är en intermediär vid aktiv demetylering av

DNA.

Rollen hos AS vid epigenetiska förändringar hos människan har aldrig studerats. I denna studie har naiva CD4+ T-celler från blodgivare använts som ett modellsystem för att undersöka ASs eventuella roll vid förändring av metyleringsmönster i immunförsvaret. Genom att använda dot-blot analys, hydroximetylerat DNA immunoprecipitation (hmeDIP) och qPCR kunde förändringar i metylering orsakade av AS detekteras.

En bekräftelse av ASs påverkan på epigenetiska förändringar hos möss observerades. AS ökar nivåerna av 5hmC jämfört med obehandlade celler. Cellinjen Jurkat, en human T-lymfocyt cellinje, visade ett motsatt resultat. Behandling med AS minskade nivåerna av 5hmC jämfört med obehandlade celler. När man jämför detta resultat med resultaten erhållna i humana naiva T-celler gjordes samma iakttagelse. Skillnaden mellan mus och människa i förmågan att producera och metabolisera AS kan vara en anledning till detta motsatta resultat.

Eftersom AS hade förmågan att skapa epigenetiska förändringar i primära humana CD4 + T-celler, tyder resultaten på att AS kan ha en funktion i det mänskliga immunförsvaret.

T

ABLE OF CONTENTS

2.1.2.1 Ten eleven translocation (TET) enzymes  ...  4  

2.2 Tet and ascorbic acid  ...  5  

2.3 Epigenetic regulation in the adaptive immune system  ...  5  

3 Experimental Theory  ...  8  

3.1 Isolation of naïve T cells  ...  8  

3.2 Dot Blot Assay  ...  8  

3.3 Immunoprecipitation (I.P)  ...  9  

3.4 Quantitative PCR (qPCR)  ...  10  

3.4.1 SYBR Green  ...  10  

3.4.2 TaqMan  ...  11  

3.5 Mimosine treatment and Propidium iodide (P.I) staining  ...  11  

4 Material and Methods  ...  13  

4.1 Cell Culture  ...  13  

4.2 Propidium iodide (P.I)  ...  14  

4.3 Preparation of Naïve T cells  ...  14  

4.4 Vitamin C treatment  ...  15  

4.5 DNA/RNA extraction  ...  15  

4.6 cDNA  ...  15  

4.7 hmeDIP  ...  15  

4.8 Dot Blot assay  ...  16  

4.9 qPCR  ...  17  

5 Results and Discussion  ...  18  

6 Conclusions  ...  25  

7 Future work  ...  26  

8 Acknowledgement  ...  27  

1

I

NTRODUCTION

L- ascorbic acid (AA), commonly known as vitamin C, is an essential nutrient. The majority of animals and plants can synthesize ascorbic acid. Humans and primates however lack an enzyme called gulonolactone oxidase and can therefore not synthesize AA, which makes intake of AA through diet more important (Naidu 2003). AA is involved in a lot of processes in the body, and deficiency can lead to scurvy, which causes structural breakdown of the skin and epithelium (Monfort, Wutz 2013). AA is a powerful antioxidant (Padayatty, Katz et al. 2003) and is suggested to be beneficial to the immune system, but its mode of action is not yet fully described. Recent studies have shown that AA could have an impact on health and disease by affecting the epigenetic control of genome activity. It is believed that AA may act as a mediator of the interface between the genome and the environment (Minor, Court et al. 2013). In support of this theory, AA has been shown to promote maturation of T cells (Manning, Mitchell et al. 2013), promote somatic cell reprogramming (Esteban, Pei 2012) and induce DNA demethylation in mouse embryonic fibroblasts (MEF) and mouse embryonic stem (ES) cells (Minor, Court et al. 2013, Blaschke, Ebata et al. 2013).

1.1

A

The aim of this project is to study the clinical relevance of AA on epigenetics using human T helper cells as a model system. The project is categorized in two major parts. The first task is to confirm the very recent observations that have been reported in the paper published by Minor and colleagues (Minor, Court et al. 2013), that ascorbic acid induces demethylation via enzymatic activity in MEFs. Confirmation of previous results is essential, especially in cases were the results only have been shown once. The approach to address the experiments differs between scientists and the use of different material and techniques can affect the result. Confirmation of previous results also provides the chance to optimize techniques and protocols for further work. The second task in this project is studying the clinical relevance of AA and trying to understand the actual affect that AA has on epigenetic changes in human T helper cells.

This study is based on work with isolated primary human cells. Naïve T cells are collected from blood donors, which give a good, relevant and clean representation of the human.

2

B

ACKGROUND

2.1

E

Epigenetics is the study of changes in phenotypes or gene expression, which are independent from changes in the underlying DNA-sequence (Kanno, Vahedi et al. 2012). Epigenetic markers can be passed on through cell division and therefore also be inherited through germ cells from parent to offspring.

Much of the research involves the study of modifications of DNA and histone proteins, and how these modifications influence the chromatin structure. One of the best-characterized epigenetic modifications is DNA methylation (Figure 1), which plays an important role in a variety of cellular processes, like X-chromosome inactivation, maintenance of epigenetic memory and regulation of gene expression (Wu, Zhang 2011, Williams, Christensen et al. 2011, Bergman, Cedar 2013).

Figure 1. An overview of the building blocks that create the chromosome. Known epigenetic

changes are histone modifications and DNA methylation.

2.1.1DNAMETHYLATION

5-methylcytosine (5mC) is a methylated form of the DNA base cytosine and is called the fifth, minor base in mammalian DNA to illustrate its importance and heritability. Three enzymes called DNA methyltransferase 1 (DNMT1), DNMT3A and DNMT3B are responsible for the deposition and maintenance of 5mC, which is essential for normal development (Smith, Meissner 2013). 5mC constitutes only around 1% of all DNA bases and is found almost exclusively in a so-called CpG dinucleotide. CpG represent the base cytosine followed by guanine and the phosphate bond connecting them. The CpG sites are palindromes that are symmetrically methylated. DNA replication therefore gives you two hemimethylated CpG sites in the two new synthesized strands. To maintain the methylation pattern through cell division DNA methyltransferase 1 (DNMT1) is needed, DNMT1 methylate the nascent DNA strand (Figure 2).

Figure 2. DNA methyltransferase 3A/B (DNMT3A/B) methylate cytosine at CpG dinucleotide

palindrome sequences by addition of a methyl group. After cell division DNMT1 methylate the hemimethylated new DNA strand.

Methylated DNA is connected to gene repression through its ability to affect the binding of transcription factors and change chromatin structure (Bergman, Cedar 2013) (Figure 3). Abnormality in DNA methylation is commonly connected to disease and can for example contribute to cancer through tumor suppressor gene inactivation (Tollefsbol 2012).

Figure 3. DNA methylation can prevent transcription factors (TF) to bind to the promoter

sequence of a gene and no transcription occurs. Methylation is through this property connected to gene repression.

2.1.2DNA DEMETHYLATION

In contrast to DNA methylation there is also a mechanism called DNA demethylation. These two events are a part of the so-called DNA methylation cycle and are essential for regulation of the cell.

Recently a sixth base, called hydroxymethylcytosine (5hmC), was discovered in mouse brain tissues while determining the total amount of cytosine methylation in CpGs. They observed a presence of an unidentified spot when performing thin-layer chromatography, which turned out to be the sixth base, 5hmC. The difference between methylcytosine and hydroxymethylcytosine is the hydroxyl group attached to the methyl group (Figure 4). A family of enzymes called Ten Eleven Translocation (TET) oxygenases creates 5hmC.

Figure 4. Methylcytosine can be converted to hydroxymethylcytosine by ten eleven

translocation (TET) enzymes.

2.1.2.1 Ten eleven translocation (TET) enzymes

The TET protein family members are TET1, TET2 and TET3. These enzymes are dioxygenases and can through oxidation convert 5mC into 5hmC, were 5hmC is presumed to be an intermediate in the process of active DNA demethylation (Nestor, Ottaviano et al. 2012). 5hmC can be subsequently converted to 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC). The last step in the demethylation mechanism is the removal of 5fC and 5caC by thymine DNA glycosylase (TDG) and a cytosine is added via base excision repair (BER) (Figure 5) (Pastor, Aravind et al. 2013).

TET enzymes create a balance between the processes of methylation and demethylation and plays a role in maintaining a methylation profile of a given cell type. It also creates flexibility in cell populations where a quick response is needed to environmental signals, such as the immune system (Li, Chen et al. 2012).

The importance of TET enzymes in DNA demethylation has been shown in knockout studies of TET genes where a decreased level of 5hmC, 5fC and 5caC has been detected (Minor, Court et al. 2013, Blaschke, Ebata et al. 2013).

Figure 5. The DNA methylation cycle. 5-methylcytosine (5mC) is formed by addition of a

methyl group to cytosine by DNA methyltransferase (DNMT). The ten eleven translocation (TET) proteins can convert 5mC to 5-hydroxymethylcytosine (5hmC) by oxidation. The cycle is completed by conversion of 5hmC back to cytosine by 1) passive demethylation caused by replication, 2) TET catalyzed oxidation or 3) Deamination.

Hydroxymethylcytosine is poorly recognized by DNMT1 and during cell division there is a loss in 5hmC due to this poor recognition. This can lead to a passive demethylation that is replication-dependent (Figure 5) (Wu, Zhang 2011) . The third pathway believed to lead to demethylation of DNA is deamination of 5hmC. However there seems to be controversy about the significance of deamination as a role in the demethylation process (Pastor, Aravind et al. 2013).

2.2

T

The TET enzymes are Fe2+ and 2-oxoglutarate (2OG)-dependent dioxygenases, which means that their catalytic domain, consisting of a double-stranded β-helix (DSBH), binds to Fe2+ and 2OG (Monfort, Wutz 2013). The TETs are only active when iron is present in the Fe2+ state and to maintain this state an electron donor is needed for adjusting the redox state of Fe. AA has shown to be a very effective electron donor for reducing iron and therefore have a crucial role as a co- factor for the activity of the TET enzymes (Monfort, Wutz 2013).

2.3

E

The adaptive immune system (AIS), also known as the specific immune system has different features than the innate immune system. The AIS gives a response that is stimulated by exposure of different types of infectious agents and the response increases in defensive capability with every new exposure to a particular antigen due to the ability to create a memory (Abbas, Lichtman 2005).

T helper (Th) cells play a key role in the regulation of the AIS. Naïve T cells (NT) can be differentiated into a number of Th subsets called Th1, Th2, Th17 and regulatory T

cells (Tregs) via exposure to an antigen. When an antigen stimulates the NT cell it activates a set of transcription factors, which drives the cell to differentiation (Abbas, Lichtman 2005). The Th subset phenotypes are defined by the signature effector cytokines produced, the master transcription factors expressed, and the type of microbial pathogens they control (Kanno, Vahedi et al. 2012).

Figure 6. Differentiation of naïve T cells into the subsets T helper cells 1 (Th1) and Th2 by

their key transcription factors, T-bet and GATA3 of Th1 and Th2, respectively. Each Th subtype also produces cytokines: IFNγ, IL4/IL5/IL13 for Th1 and Th2 respectively, to enhance its subset commitment.

During T cell differentiation, changes in the DNA methylation patterns occur at several loci (Kanno, Vahedi et al. 2012). Interleukin 2 (IL2), a T cell growth hormone, is one of these loci. The demethylation process at the IL2 promoter started only 20 minutes after activation of murine naïve T cells (Figure 7) (Bruniquel, Schwartz 2003). To exclude the possibility of a passive demethylation process, the authors blocked DNA replication and still observed this rapid demethylation, which indicates an active enzymatic process.

Figure 7. In vitro stimulation of murine naïve T-cells resulted in 50-80% demethylation in the

As stated above, epigenetic regulation occurs in adaptive immunity. In this project we will determine the effect of AA on DNA methylation in human T-cells. We hypothesize that AA affects the adaptive immune response in humans by changing the DNA methylation dynamics in differentiating CD4+ T-cells.

3

E

XPERIMENTAL

T

HEORY

This section is meant to create a better understanding for some of the methods used in this project.

3.1

I

T

Naive T cells can be isolated from buffy coat collected from blood donors using density gradient centrifugation. By using a density gradient medium, mononuclear cells can be collected from peripheral blood because granulocytes and erythrocytes have a higher density than the mononuclear cells. This makes them sediment through the gradient medium and lymphocytes can easily be collected through suction with a pipette (Figure 8).

Figure 8. By density gradient medium and centrifugation, different layers of peripheral blood

are created. The mononuclear cells (lymphocytes) create a thin layer in the middle of the tube and ought to be collected with a pipette.

Isolation of naïve CD4+ T cells from peripheral blood mononuclear cells (PBMC) are done by labeling the unwanted cells with biotin-conjugated monoclonal antibodies followed by anti-biotin microbeads as a secondary labeling reagent. The naïve T cells distinguish from the other cells by different expression patterns on the cell surface. When cells are passing through a magnetic column in a magnetic field the magnetically labeled cells will be depleted by sticking to the magnet and naïve T cells will pass through the column and be collected for research.

3.2

D

B

A

By using a dot blot assay you can immobilize DNA on a nylon membrane and use primary and HRP-conjugated (horseradish peroxidase) secondary antibody to determine the quantity of a specific epitope of interest (Wilson, Walker 2010). To fixate DNA to the membrane, the DNA needs to be single stranded. By heat denaturation and quick cooling on wet ice double stranded DNA becomes single stranded. The sample is added to the membrane and is bound by vacuum driven microfiltration. After immobilization the membrane is treated with UV light so that the DNA cross-links properly to the membrane. To prevent non-specific binding, a blocking buffer is added before the membrane is treated with the primary antibody. The primary antibody is designed to bind to the antigen of interest and the secondary

molecule, for example (HRP). When this molecule is incubated with a chemilumiescence substrate it creates emission of light. The intensity of the light can then be translated into quantity of the epitope (Figure 9).  

Figure 9. When adding samples, the DNA is bound to the membrane. A designed primary

antibody to a specific antigen, in this case hydroxymethylcytosine, is added and binds to this molecule. A secondary antibody featured with a detection molecule is added to the membrane and binds to the primary antibody. When the membrane is treated with a chemilumiescence substrate the amount of hydroxymethylcytosine in each sample can be detected.

3.3

I

(I.P)

Commonly used techniques, such as bisulfite sequencing, do not have de ability to distinguish between 5hmC and 5mC. The development of a specific antibody to 5hmC creates the possibility to collect DNA containing 5hmC. The technique is called hydroxymethylated DNA immunoprecipitation (hmeDIP) and it is a valuable tool when determining locus-specific and genome-wide profiles of 5hmC in DNA (Nestor, Meehan 2013). Genomic DNA is sonicated to a size of approximately 200-1000 bp. Some DNA is left out to work as a comparison in qPCR and it is called input. The DNA is then treated with a primary antibody that binds specific to 5hmC. A secondary antibody equipped with a magnetic bead is added to bind to the primary (Figure 10). The DNA consisting of 5hmC is collected via a magnetic rack, were the magnetic beads is drawn to the magnet.

Figure 10. By using a primary antibody that binds to 5-hydroxymethylcytosine (5hmC) and a

secondary equipped with a magnetic bead you can separate hydroxymethylated DNA fragments from regular DNA.  

3.4

Q

PCR

(

PCR)

3.4.1SYBRGREEN

A DNA-binding dye, such as SYBR green binds to the major groove of double-stranded DNA and can during the PCR process be detected following excitation (Figure 11). This results in a real time measurement of the accumulation of DNA, allowing calculation of the initial concentration of DNA or cDNA.

Figure 11. SYBR green fluoresces when it binds to double-stranded DNA. When the DNA is

denatured, the dye is released and there is a decrease in fluorescence. During extension, when a PCR product is generated and when the polymerization is completed the dye binds to the double stranded product resulting in an increase in fluorescence that can be detected.

3.4.2TAQMAN

Another type of qPCR is the TaqMan system and it is good to use when determining expression levels (Wilson, Walker 2010). In this system, a probe consisting of an oligonucleotide is labeled with a fluorescent reporter at one end and a quencher at the other. The PCR run as normal and the probe binds during the annealing step to a targeted sequence. During extension by the Taq polymerase the probe is cleaved because Taq have 5’ nuclease activity and the reporter is released. This gives a fluorescent signal that can be detected and the signal is increasing in direct proportion to the number of starting molecules (Figure 12).

Figure 12. TaqMan assay. The PCR is done with a reporter/quencher (RQ) probe. When R-Q

is close, fluorescence is quenched. During the extension the probe is cleaved because Taq is having 5’ nuclease activity. The reporter is then released and results in a detectable increase in fluorescence.

3.5

M

P

(P.I)

The plant amino acid, mimosine, can induce a cell cycle arrest in the late G1 phase and is used when there is a desire to stop cells from proliferating (Krude 1999).   P.I is a fluorescent molecule that binds to both DNA and RNA. It is used, among other things, to analyze the different phases of the cell cycle by using flow- cytometry. The amount of DNA the cells contain correlates with the intensity of the fluorescence. The cell cycle is briefly divided into 5 stages (Figure 13).

                            Figure 13.The cell cycle is a series of events that take place in the cell, leading to its division

and duplication (replication). By using Propidium iodide the current stage in the cell can be detected.

As the DNA duplicates in the S phase the fluorescence will be twice as high in the G2/M phase then in the G0/G1 and a relative amount of cells in each phase can be

determent (Krishan 1975). Through this method, the cycle status of cells in culture can be determent. G0 (resting phase) G1 (ready for DNA synthesis) S phase(DNA synthesis) G2 (cells continue to grow) M (Mitosis-division into two daughter cells)  

4

M

ATERIAL AND

M

ETHODS

To create an overview of the methods used in this study, a flowchart illustrates the methods and in which order they were performed (Figure 14).

  Figure 14. Procedure flowchart of the study with respect to the methods that were used.

4.1

C

C

Mouse embryonic fibroblast (MEF) cell line was a generous gift from the lab of professor Karin Ollinger. The MEFs were cultured in Dulbecco's Modified Eagle Medium (DMEM) (Life Technologies, United States), with 10% fetal bovine serum (FBS) (Invitrogen, United States) and antibiotics (x100 dilution) in T75 cell culture flasks. The MEFs was put in a 37°C and 5% CO2 incubator. The cells were grown to

90% confluent before treated with AA. To a T75 flask 2ml of trypsin was added and incubated for 2 minutes in 37°C to detach the cells from the surface. 8ml of DMEM was then added for inactivating trypsin.

Jurkat cells (ATCC, United States) was cultured in RPMI Media 1640 (Life Technologies), with 10% FBS and antibiotics (x100 dilution) in T25 cell culture flasks. The cells were cultured with a cell density of 3×10! _{cells/ml. The cells were}

split every three days to prevent overpopulation in the flasks. The cells were treated with AA before harvested.

When cells were treated with mimosine, a stock solution of 40mM was made up by 25mg mimosine (Sigma-Aldrich, United States) and 3.1534ml of 0.1M NaOH. 50ml of cell culture media was mixed with 500µl of mimosine stock solution (x100 dilution). Cells were then incubated in the media consisting mimosine for 24h before treated with AA.

Propidium iodide (P.I) Cell culture - Mouse embryonic fibroblast (MEF) - Jurkat cells Naive T cells from buffy coat Vitamin C treatment DNA/RNA extraction Making cDNA from RNA. -Naive T cells hmeDIP - Naive T cells

Dot blot assay -MEFs - Jurkats - Naive T cells

Naïve T cells were cultured in RPMI Media 1640 (Life Technologies), with 10% FBS and antibiotics (x100 dilution) in T25 cell culture flasks at a density of 1×10!

cells/ml. The T cells were only cultured during the time they were treated with AA. All cells cultured was centrifuged at 300G for 10 minutes at 20°C before extraction of DNA or RNA.

4.2

P

(P.I)

MEFs and Jurkat cells were harvested and spun down at 300g for 10 minutes in 4°C, Acc/Dec 6. The pellet was resuspended in 3ml of cold absolute ethanol and fixated in -20°C over night. The cells were washed in 0,1% PBS-BSA (Phosphate Buffered Saline with Bovine Serum Albumin) three times. 20µl of P.I staining was added to the pellet and mixed well together with 1µl RNase to prevent an eventual disturbance when staining. The mixture was incubated in 4°C for 1h in the dark. Results were analyzed using a FACSAria IIITM _{(BD Bioscience, United States).}

4.3

P

N

T

For each new experiment fresh naive T cells were purified from blood donors (Blodcentralen, Linköping, Sweden). The buffy coat contained the anticoagulants CPD (Citrat Phosphate Dextros).

Buffy coat was diluted 1:1 with PBS (GE healthcare, United Kingdom). The mix was added carefully by pipetting to the surface of the density medium Lymphoprep (Stemcell Technologies, Canada) at a ratio of 1:2 in a 50ml Falcon tube. The tubes were then centrifuged with the following settings on a Universal 320R (Hettich, Germany) centrifuge:

• 20-21°C • 800G • Acc/Dec 1 • 30 minutes

Lymphocytes were then collected with a plastic Pasteur pipette and washed with cold PBS three times. The settings on the centrifuge for each washing step was:

• 4°C • 300G • Acc/Dec 6 • 10 minutes

After the last washing step the cell count of PBMC were calculated with Automated Cell Counter TC10TM (Bio-Rad, United States). Naïve CD4+ T cells were then isolated using the Naïve CD4+ T Cell Isolation Kit II Human (Miltenyi Biotec, Germany) according to manufacturers instructions.

4.4

V

C

A stock solution of 50mM was made using 100mg L-ascorbic acid (Sigma) and 11.35 ml 4°C MilliQ water. The stock solution was stored at -20°C and a fresh ascorbic acid tube were always thawed before each new experiment. Calculations for the desirable concentration were made and the calculated volume of the stock solution was added to the media.

To measure the actual concentration of ascorbic acid in the media, the Ascorbic acid Assay Kit (Abcam, England) was used, following the manufacturers instructions.

4.5

DNA/RNA

Genomic DNA was extracted from cells using DNeasy Blood & Tissue Kit (Qiagen, Netherlands), according to the manufacturers instructions with the amendment of adding 10µl RNase. This is done to prevent any erroneous measurements when quantifying DNA. A NanoDrop 1000 Spectrophotometer (Thermo scientific, United States) was used to quantify DNA concentrations by adding a sample size of 1.5µl. Total RNA was extracted from cells using RNeasy Mini Kit (Qiagen), according to the manufacturers instructions. To quantify RNA concentrations, the same NanoDrop mentioned above was used, analyzing a sample size of 1.5µl.

4.6

DNA

Because RNA is highly unstable it is advantageous to turn RNA into complementary DNA (cDNA). To make complementary DNA from RNA the High capacity cDNA Reverse Transcriptase Kit (Life Technologies) was used according to the manufacturers instructions. Every time cDNA was made the following settings was applied.

1. 25°C for 10 min 2. 37°C for 2h 3. 85°C for 5min

4. 4°C until the samples were collected.

The reaction was performed in a GeneAmp PCR system 2700 (Applied Biosystems, United States).

4.7

DIP

Genomic DNA was sonicated at 4°C for 6 cycles of 30s on/off. 180ng of sonicated DNA was run on a 1% agarose gel at 95V together with a ladder (Invitrogen), to control the appropriate length of the DNA fragments. The DNA was denatured by incubation at 100°C for 10 min in hot-block, followed by quickly chilling on wet ice for 5 minutes. 45µl of the denatured DNA was collected and served as input. 45µl of 10x IP buffer was added to the remaining sample (I.P sample) together with 1µl of a polyclonal primary 5hmC antibody (ActiveMotif, United States) and incubated over night at 4°C. Incubation was done on a rotating wheel at 25rpm. 40µl Dynabeads Protein G (30mg/ml) was placed in clean Eppendorf tubes (1.5ml) and washed in

800µl cold PBS-BSA for 5 minutes on a rotating wheel at room temperature. The liquid was removed using a magnetic rack (Millipore, United States) and by pipetting remove the washing solution. The DNA/primary antibody mixture was added to the dynabeads tube and incubated for 1h on a rotating wheel at 4°C. Supernatant was removed using the magnetic rack and the Dynabeads were washed three times for 5 min on a rotating wheel with 1x IP buffer (4°C). The beads was then re-suspended in 250µl digestion buffer and 10µl Proteinase K (20mg/ml stock) and incubated in 50°C for 3h at 800rpm. This step releases the antibody complex from the DNA. (Nestor, Meehan 2013) The IP sample and the input was then purified using QIAquick PCR purification Kit (Qiagen) according to the manufacturers instructions.

4.8

D

B

90µl of a master mix made of 1M NaOH and 200mM EDTA was added to 110µl DNA (2000ng).

Genomic DNA was denatured at 100°C for 10 min and quickly cooled on wet ice. The blotting unit (Scie-Plas, United Kingdom) was connected to a vacuum pump and the nylon membrane (GE Healthcare) was washed with 2x SSC buffer and the filter (Whatman, United Kingdom) was washed with water before placed over the holes. The membrane/filter complex was washed with 200 MilliQ water and the 200µl sample was added. After the samples, 200µl of 0.4M NaOH was added. The membrane was released and washed with 2x SSC and cross-linked in UV light for 2 minutes. The membrane was left to dry for 1.5h in 37°C. After the membrane was dry, 12ml-blocking buffer (LI-COR, United States) was added for 30 minutes. A polyclonal primary 5hmC antibody (α-hmC) (ActiveMotif) was added at a concentration of 1:5000 and incubated over night at 4°C on a topple table. The α-hmC was then washed off using TBS (Tris Buffered Saline) -Tween buffer three times. 12 ml blocking buffer and a secondary antibody (Goat anti Rabbit IgG, Bio-Rad) at a 1:3000 concentration was added for 1h in room temperature in a topple table. The membrane was washed again with TBS-Tween.

When analyzing 5mC a polyclonal primary 5mC antibody (ActiveMotif) was used at similar concentrations as the one for 5hmC. The secondary antibody (Goat anti Mouse IgG, Bio-Rad) was used instead of anti Rabbit, and added at a concentration of 1:3000.

2ml of chemiluminescent substrate (Thermo Scientific) was added and incubated in the dark for 3 minutes before analyzed using the ChemiDoc™ XRS+ System (Bio-Rad). The last step was staining the membrane with methylene blue (Molecular Research center, United States) to see that an equal amount of DNA was loaded to each dot.

4.9

PCR

When performing PCR, a 7900HT Fast Real-Time PCR System (Applied Biosystems) was used. The following settings were applied when performing both SYBR Green and the TaqMan system and the number of cycles was set to 40 (Figure 15).

When analyzing gene expression from cDNA, 5µl TaqMan Gene Expression Master Mix (Applied Biosystems) was used, mixed with 0.5µl of the wanted probe (Applied Biosystems). Per sample, 4.5µl cDNA was mixed with 5.5µl of the master mix to a reaction volume of 10µl. The PCR-plate was put on a vortex, spun down and checked for any possible air bubbles. When analyzing DNA collected from hmeDIP experiments, the SYBR Green PCR Master Mix (Applied Biosystems) was used together with primers for the desired regions (Sigma). The amount of master mix, primers and sample is the same as for the TaqMan system.

50°C   2:00   95°C   10:00   95°C   00:15   _60°C   1:00  

5

R

ESULTS AND

D

ISCUSSION

The laboratory work consisted of two main parts. The first was to confirm the observations that AA enhances 5hmC generation in mouse embryonic fibroblast (Minor, Court et al. 2013). The second was to determine the clinical relevance of AA by studying its effect on the epigenetics in human, using T helper cells as a model system.

The anti- 5mC and 5hmC antibodies used in the dot blot assays specifically recognize DNA containing methylcytosine and hydroxymethylcytosine, respectively. The antibodies to 5hmC and 5mC were specific and sensitive to control DNA (Figure 16A). For every experiment using dot blot, positive controls in form of DNA known to be containing only C, 5mC or 5hmC were used to see if the antibody used was specific and sensitive. The positive control is not always shown in the result figures below, but the controls have worked every time if not stated otherwise. For an accurate result in a dot blot assay it is important to load the same amount of sample to each dot to get an equal amount of loaded DNA. The accuracy of loading can be investigated using methylene blue staining. A decreasing concentration of DNA was added and a decrease in spot intensity for each concentration is detected, showing the accurate loading of samples and the specificity of the technique (Figure 16B).

Figure 16. Dot- blot assay results from DNA extracted from mouse embryonic fibroblasts. A) Antibodies against 5- hydroxymethylcytosine (5hmC) and 5-methylcytosine (5mC) are

specific and are used to analyze DNA modification. For each experiment a positive control is present and gives a confirmation that the experiment worked. B) Methylene blue staining showed the relative amount of DNA loaded to the membrane. C) Mouse embryonic fibroblasts treated with 0, 10, 100 and 500µM of ascorbic acid for 24h and analyzed for 5hmC content. The experiment was done in two technical replicates, a and b. An increase of 5hmC was

MEFs were treated with concentrations of 0, 10, 100 and 500µM AA for 24h in duplicate to confirm recently reported results. The dot blot assay showed that mouse embryonic fibroblasts did have an increase in hydroxymethylcytosine when treated with AA. However, a consistent increase in 5hmC with increasing AA concentration was not observed (Figure 16C).

A gain in 5hmC may result in a loss of 5mC due to the methylation cycle. To investigate this theory the same DNA samples from MEFs treated for 24h were stained with a 5mC antibody. This showed a loss in 5mC even though the result was not very clear (Figure 16D) due to lack of antibody sensitivity. MEFs are highly proliferative, and as 5hmC is lost during DNA replication, the inconsistent results upon AA treatment may reflect different levels of proliferation between individual culture flasks. To control for proliferation rate, an attempt to induce cell cycle arrest using mimosine was done. Unfortunately, mimosine treatment killed the cells, confirming the previously reported toxic properties of mimosine at higher concentrations (Krude 1999). Even though a cell cycle arrest for MEFs would be desirable, the result shown still indicates a gain in 5hmC (Figure 16C), which confirms previous results on AAs impact on 5hmC levels (Minor, Court et al. 2013). Mice can synthesize AA whereas humans lack this ability. To get a better understanding of how AA affects a human cell, the Jurkat cell line was used, which is a line of human CD4+ T lymphocytes. The Jurkats were treated with different concentration of AA at different time points, but none of the conditions showed any sign of 5hmC (Figure 17A). The positive control in this assay did work, indicating that the result was not due to technical failure. The methylene blue staining also showed consistent loading of sample DNA.

Figure 17. Jurkat cells treated with vitamin C before and after cell cycle arrest by mimosine. A) Jurkat cells with no mimosine, treated with 0, 10 and 100µM of ascorbic acid

(AA) for 4, 8 and 24h. No indication of 5hmC was shown. The positive control in the dot- blot and the methylene blue staining confirm that the experiment was done correctly. B) Propidium iodide experiment comparing Jurkat cells, treated with and without mimosine for 24h. Mimosine stopped cell proliferation at the G1 phase after 24h. C) After cell cycle arrest by mimosine, Jurkats were treated with AA for 24h with 0, 100 and 500µM of AA. When the cells were not proliferating, 5hmC could be detected. The result showed a loss in 5hmC when treated with AA, compared to the untreated cells.

The Jurkat cell line proliferates very quickly. TETs probably convert 5mC to 5hmC but due to replication-dependent demethylation the amount of 5hmC is to small too be detected. In these highly proliferating cells it is even more important to induce cell cycle arrest. Trying to stop cell division with mimosine was therefore tested and analyzed using propidium iodide. The concentration of mimosine was established from literature search (Mosca, Dijkwel et al. 1992, Oppenheim, Nasrallah et al. 2000) and validated by propidium iodide experiments. A concentration of 0.4mM of mimosine successfully arrests the cells at the G1 phase after 24h. 83.2% of the cells stayed at G1, compared with no mimosine treatment where only 63.5% stayed at G1 (Figure 17B). When mimosine-treated Jurkat cells were treated with AA 5hmC was observed (Figure 17C), demonstrating the impact of proliferation. However, a decrease of 5hmC with increasing concentration of AA was observed, an opposite effect from the results shown in MEFs.

Cells that have been in culture for a long time have a methylation profile that is completely different from the cells from which they were originally derived. That is why human naïve T cells, that do not proliferate and are primary cells, are a better model system to use, as no mimosine treatment is necessary. The following results come from experiment done with human naïve T cells as a model system.

In mice, knocking down the expression of TETs decreased the effect of AA on 5hmC (Minor, Court et al. 2013). To see if AA hade any impacts on the expression of the proteins present in the methylation cycle in a NT cell, qPCR analyzes was used, using cDNA, to investigate this. To see the difference in expression between treated and untreated cells, expression of the proteins were analyzed relative to no treatment of AA. The early time point of 1h with 50µM AA reviled a relative increase in expression for TET1 and TET3, calculations made from four biological replicates (Figure 18A). The result however is not statistically significant (T-test, P value >0.05) so no robust conclusion could be made. At the later time point of 24h, no up regulation of the expression is observed, but this results is not statistically significant (T-test, P value >0.05) and no conclusion could be made. For expression of TDG and DNMTs only two biological replicates were analyzed and for a statistic evaluation more experiments needs to be done.

  Figure 18. Relative expression levels of proteins involved in the methylation cycle analyzed from cDNA (naïve T cells) by qPCR. A) Expression of ten eleven translocation (TET)

calculated by four biological replicates. At 1h there was some up regulation of TET1 and TET3, not statistically significant (P value >0.05) and can therefore not be taken into calculation. B) Expression of thymine DNA glycosylase (TDG) and DNA methyltransferase (DNMTS) calculated from two biological replicates. No statistic evaluation was performed.

The impact of AA on 5hmC levels in NT cells was analyzed by dot-blot. A wide range of concentrations of AA was used. A variable, but consistent decrease of 5hmC occurred when increasing the concentration of AA (Figure 19A). The result is the opposite of that previously shown in MEFs, but shows the same pattern as found in Jurkat cells, which are also human T cells. Mouse and humans differ when it comes to being able to produce ascorbic acid, a possible reason for the opposing results seen in the two different types of cells.

To create an environment more similar to the body, physiological levels of AA were added to the NT. The physiological levels of AA in plasma is ~50µM (Naidu 2003) and are more or less consistent because of invariable intake of AA through diet. A wide span of concentration with low, physiological and high levels of AA was used to analyze a change in 5hmC. The result also showed a decrease in 5hmC, but the

amount 5hmC did not consistently follow the concentration gradient of AA (Figure 19B).

  Figure 19. Naïve T cells treated with vitamin C at different concentrations. A) Naïve T cells

were treated with a wide range of vitamin C and the result from dot-blots showed a loss in 5-hydroxymethylcytosine (5hmC) after 24h. B) A range covering physiological levels of vitamin C showed a loss of 5hmC. C) To gain physiological levels of vitamin C in the media, a higher dose then 50µM was needed when treating the cells. A small loss of 5hmC could be detected.

D) hmeDIP showed a loss of 5hmC after 24h in two biological replicates, supporting

previous observations in NT.

AA degrades rapidly at room temperature, so it was suspected that the incorrect amount of AA was present in the media, causing the non-consistent results of 5hmC. The ascorbic acid assay kit (Abcam) was used to determent the actual concentration of AA and to confirm that no AA was present in the media to begin with. A major decrease in concentration of AA occurred immediately after adding AA to 37°C media (Figure 20C), an observation coinciding with previous findings (Blaschke, Ebata et al. 2013). And after 24h, the majority of AA was gone (Figure 20B).

  Figure 20. Results using the Ascorbic acid Assay Kit (Abcam) for detection of ascorbic acid (AA) in the media. A) To control the assay and calculate the AA concentrations in the media

a standard curve was done. B) Time points of 0, 1 and 24h showed that AA decrease rapidly and a major loss is observed immediately after adding AA to media. After 24h the majority of the AA have degraded. C) AA decreases instantly and after 0h, the physiological level of AA was obtained when adding 150µM.

To generate the physiological concentration of 50µM AA in the plasma, an initial concentration of 150µM was needed. To analyze the impact of AA before total degradation, a 1h treatment was performed. Very little effect was shown, but an indication of loss in 5hmC can be observed (Figure19C). A possible explanation is that 1h is not enough time for the TET enzymes to generate sufficient 5hmC for detection by dot blot. It would be desirable to treat NT with AA for 24h and during that time maintain a constant concentration of AA in the media.

A more sensitive method to examine the amount of 5hmC is hmeDIP. For this experiment, primers for Long Interspersed Elements (LINEs) were used. LINEs are genetic elements that are found in large numbers in the genome. The human genome contains about 500 000 LINEs, which is ~17% of the genome (Cordaux, Batzer 2009). These elements are known to be highly methylated and if there is a genome wide change in methylation patterns by treatment of AA it should be picked up by hmeDIP using LINEs. Two biological replicates were compared in enrichment of 5hmC when treated with AA for 1h and 24h. When DNA was extracted from the cells treated for 1h there was not enough DNA from one of the samples, therefore only the data for 24h is shown (Figure 19D). It showed a decrease of 5hmC when treated with AA, relative to the untreated cells. The result is a further confirmation of the results from dot blots when analyzing naïve T cells.

For all dot blots analyzing 5hmC it is also desirable to see changes in 5mC. The antibody that was used for analyzing 5mC in MEFs was not specific and sensitive. For a better result a different antibody should be used. A 5mC monoclonal antibody (clone 10G4)from Zymo Research showed to be specific and sensitive at a

concentration of 1:2000 (Figure 21), but due to time constrains this antibody was never used in this study. In future research it would be advantageous to use this antibody for detecting 5mC.

Figure 21. To investigate the specificity and sensitivity of a new 5-methylcytosine (5mC)

antibody, different amounts of DNA sample were treated with different concentrations of antibody. A concentration of 1:2000 showed to be the most specific.

6

C

ONCLUSIONS

Thus, we have confirmed the recently reported effect of AA on MEFs, namely an increase in global 5hmC levels. In addition, we have shown that this effect is not replicated in either primary or transformed human cells using dot-blots and 5hmC-immunoprecipitation. This difference may reflect the differing ability of mouse and human cells to produce and metabolize AA. Our results are significant; as they are the first to show that AA can directly and rapidly alter the methylome (genome wide methylation patterns) of primary human immune cells, CD4+ T-cells. As the ability to change DNA methylation levels is critical for CD4+ T-cell differentiation, our results suggest that AA levels may directly affect the functioning of the human immune system.

7

F

UTURE WORK

During the past few years, several publications covering the importance of ascorbic acid in different biological systems have been reported. This shows that studies of this field are timely, novel and more research needs to be done.

In this study, no research regarding AAs actual influence on TET enzyme activity was measured due to time constrictions. The kit Epigenase 5mC Hydroxylase TET Activity/Inhibition Assay Kit (Colorimetric) from Epigentek can detect TET activity. To investigate AAs role in enzymatic activity, this kit is recommended for future studies.

Since there are epigenetic changes in the immune system, the role of AA in allergy would be an interesting study. By using allergen challenge in patients and controls we can investigate the possible changes of methylation with and without AA, in CD4+ T cells by gene expression microarrays. Using microarrays on NT cells not stimulated by any allergen is also an interesting experiment to determent a wide scale of methylation changes through out the genome.

8

A

CKNOWLEDGEMENT

First I want to thank my supervisor Colm Nestor for introducing me to this project and for being a great supervisor. Your guidance and advice during this time have been invaluable.

Thanks to my examiner Jonas Wetterö for your thoughts and input on my report. Many thanks to the people in Mikael Benson’s research group at the center for individualized medicine for all the help in the lab and for creating a great work environment.

Thanks to my friend and opponent, Ida Grandell, for valuable discussions and input regarding this project.

Finally, I want to thank my wonderful friends, family and beloved Niklas for all the pep talk and support.

9

R

EFERENCES

ABBAS, A. and LICHTMAN, A., 2005. Cellular and Molecular Immunology. 5 edn. Philadelphia: Elsevier Saunders.

BERGMAN, Y. and CEDAR, H., 2013. DNA methylation dynamics in health and disease. NATURE STRUCTURAL & MOLECULAR BIOLOGY, 20(3), pp. 274-281.

BLASCHKE, K., EBATA, K. and KARIMI, M., 2013. Vitamin C induces Tet-dependent DNA demethylation and a blastocyte- like state in ES cells. Nature, 500, pp. 222-228.

BRUNIQUEL, D. and SCHWARTZ, R., 2003. Selective, stable demethylation of the interleukin-2 gene enhances transcription by an active process. Nature, 4(3), pp. 235-240.

CORDAUX, R. and BATZER, M.A., 2009. The impact of retrotransposons on human genome evolution. Nature reviews.Genetics, 10(10), pp. 691-703.

ESTEBAN, M.A. and PEI, D., 2012. Vitamin C improves the quality of somatic cell reprogramming. Nature genetics, 44(4), pp. 366-367.

KANNO, Y., VAHEDI, G., HIRAHARA, K., SINGLETON, K. and O'SHEA, J.J., 2012. Transcriptional and epigenetic control of T helper cell specification: molecular mechanisms underlying commitment and plasticity. Annual Review of Immunology,

30, pp. 707-731.

KRISHAN, A., 1975. Rapid flow cytofluorometric analysis of mammalian cell cycle by propidium iodide staining. The Journal of cell biology, 66(1), pp. 188-193.

KRUDE, T., 1999. Mimosine Arrests Proliferating Human Cells before Onset of DNA Replication in a Dose-Dependent Manner. Experimental Cell Research, 1(247), pp. 148-159.

LI, Y., CHEN, G., MA, L., OHMS, S., SUN, C., SHANNON, F. and FAN, J., 2012. Plasticity of DNA methylation in mouse T cell activation and differentiation. BMC Molecular Biology, 13(16), pp. 1-19.

MANNING, J., MITCHELL, B., APPADURAI, D.A., SHAKYA, A., PIERCE, L.J., WANG, H., NGANGA, V., SWANSON, P.C., MAY, J.M., TANTIN, D. and

SPANGRUDE, G.J., February 5, 2013. Vitamin C Promotes Maturation of T-Cells.

http://online.liebertpub.com.lt.ltag.bibl.liu.se/doi/pdf/10.1089/ars.2012.4988 edn. United States: Mary Ann Liebert, Inc.

MINOR, E.A., COURT, B.L., YOUNG, J.I. and WANG, G., 2013. Ascorbate Induces Ten-Eleven Translocation (Tet) Methylcytosine Dioxygenase-mediated Generation of 5-Hydroxymethylcytosine. The Journal of biological chemistry,

288(19), pp. 13669-13674.

MONFORT, A. and WUTZ, A., 2013. Breathing-in epigenetic change with vitamin C. EMBO reports, 14(4), pp. 337-346.

MOSCA, P.J., DIJKWEL, P.A. and HAMLIN, J.L., 1992. The plant amino acid mimosine may inhibit initiation at origins of replication in Chinese hamster cells. Molecular and cellular biology, 12(10), pp. 4375-4383.

NAIDU, A., 2003. Vitamin C in human health and disease is still a mystery? An overview. http://www.nutritionj.com/content/2/1/7 edn. Mysore, India: Nutrition Journal.

NESTOR, C. and MEEHAN, R., 2013. Hydroxymethylated DNA

Immunoprecipitation (hmeDIP). Experimental protocol edn. Methods in molecular biology. Forthcoming.

NESTOR, C.E., OTTAVIANO, R., REDDINGTON, J., SPROUL, D., REINHARDT, D., DUNICAN, D., KATZ, E., DIXON, J.M., HARRISON, D.J. and MEEHAN, R.R., 2012. Tissue type is a major modifier of the 5-hydroxymethylcytosine content of human genes. Genome research, 22(3), pp. 467-477.

OPPENHEIM, E.W., NASRALLAH, I.M., MASTRI, M.G. and STOVER, P.J., 2000. Mimosine is a cell-specific antagonist of folate metabolism. The Journal of biological chemistry, 275(25), pp. 19268-19274.

PADAYATTY, S., KATZ, A. and WANG, Y., 2003. Vitamin C as an Antioxidant: Evaluation of Its Role in Disease Prevention. Journal of the American College of Nutrition, 22(1), pp. 18-35.

PASTOR, W.A., ARAVIND, L. and RAO, A., 2013. TETonic shift: biological roles of TET proteins in DNA demethylation and transcription. Nature reviews.Molecular cell biology, 14(6), pp. 341-356.

SMITH, Z. and MEISSNER, A., 2013. DNA methylation: roles in mammalian development. Nature genetics, 14, pp. 204-220.

TOLLEFSBOL, T.O., 2012. Epigenetics in human disease. 1 edn. Amsterdam ; New York: Academic Press.

WILLIAMS, K., CHRISTENSEN, J. and HELIN, K., 2011. DNA methylation: TET proteins-guardians of CpG islands? EMBO reports, 13(1), pp. 28-35.

WILSON, K. and WALKER, J., eds, 2010. Principles and Techniques of

Biochemistry and Molecular Biology. 7 edn. New York: Cambridge University Press. WU, H. and ZHANG, Y., 2011. Mechanisms and functions of Tet protein-mediated 5-methylcytosine oxidation. Genes & development, 25(23), pp. 2436-2452.