A thesis presented for the degree of

Doctor of Philosophy

by

Rhea F Cornely

Faculty of Medicine

Centre for Vascular Research

2013

‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’

Signed: ______

Rhea F Cornely

Date: ______

‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International.

I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signed: ______

Rhea F Cornely

Date: ______

‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’

Signed: ______

Rhea F Cornely

Date: ______

The activation of T cells by an antigen presenting cell is one of the essential first steps for an effective immune response. Activation of the T cell receptor (TCR) leads to fundamental changes to T cell morphology, cytoskeletal rearrangement, membrane order and the formation of TCR microclusters which then propagate downstream signalling. TCR signalling eventually induces T cell proliferation as well as the production and secretion of interleukin 2 (IL-2) which binds to the IL-2 receptor and further stimulates T cell proliferation in an autocrine and paracrine fashion. Both signalling processes are dependent on specialised membrane domains enriched in cholesterol also termed lipid rafts and were found to sustain transient interactions with the cortical actin meshwork.

Annexin A6 (AnxA6) is a calcium-activated cytosolic phospholipid membrane binding which has been suspected to play a role in T cell development and is upregulated in a wide range of immune cells including T cells. AnxA6 is involved in receptor endocytosis, vesicle budding and cholesterol homeostasis. It has been found to preferentially bind cholesterol-rich phospholipid membranes and to interact with the cortical cytoskeleton to allow endocytosis and vesicle budding.

The aim of this study was to show that AnxA6 is crucial for an effective T cell mediated immune response. It was hypothesised that AnxA6 influences receptor signalling directly by mediating a link between the cortical cytoskeleton and cholesterol-rich membrane domains and thus AnxA6 might be part of a mechanism to target receptors to a specific membrane environment. Alternatively, AnxA6 influences signalling processes indirectly through maintaining the optimal composition of the plasma membrane due to its involvement in cholesterol homeostasis – as AnxA6 plays a role in cholesterol transport from intracellular membranes to the plasma membrane as well as low-density lipoprotein endocytosis.

In this study a delayed-type contact hypersensitivity was elicited in AnxA6 knock-out mice to generate a T cell mediated immune response in vivo. The results showed that AnxA6

mice did mount a T cell mediated immune response but the levels of proliferating CD4+ T cells were significantly lower than in wild type mice.

Investigating the source of this proliferation defect, western blot, fluorescence microscopy, qPCR and flow cytometry approaches were used to measure the response of primary murine T cells to T cell receptor as well as IL-2 receptor stimulation. Neither the early nor the late response of TCR signalling was affected in AnxA6-/- T cells. Instead, activated AnxA6-/- T cells secreted more IL-2 and the response to IL-2 stimulation was found to be impaired in AnxA6-/- T cells.

In parallel, it was investigated if and how the membrane composition of primary AnxA6-/- T cells was different from wild type T cells. The membrane order of TCR stimulated as well as naïve cells was measured with polarity-sensitive dye Laurdan. To characterise the cholesterol content and phospho- and sphingolipid composition the lipid phase was extracted from whole T cell lysates and analysed by mass spectrometry. AnxA6-/- T cells were found to have a lower degree of plasma membrane order which implies that these membranes are more fluid. In agreement with this, AnxA6-/- T cells were found to have an altered membrane lipid composition. Some of these changes potentially affect the fluidity of the T cell plasma membrane: Cholesterol, which decreases plasma membrane fluidity, is less abundant in the T cells of AnxA6-/- mice. Levels of phosphatidylethanolamine with arachidonic acid, a polyunsaturated fatty acid which increases membrane fluidity, were more abundant in AnxA6-/- T cells.

In conclusion, it could be shown that AnxA6 does play an important role in T cells and is required for efficient proliferation in a T cell mediated immune response. It is likely that the lower degree of IL-2 signalling efficiency in vitro and proliferation defect of CD4+ T cells observed in AnxA6-/- mice in vivo were a result of the changes in membrane composition.

I thank the University of NSW for supporting me with a University International Postgraduate Award (UIPA) scholarship and an Australian Postgraduate Award (APA)

A big thanks goes also to all the collaborators and university facilities that made this work possible:

Todd Mitchell from the University of Wollongong for providing me with new and improved protocol for lipid extraction and for data acquisition of same lipid extracts. Sarah Norris for MS sample acquisition and patiently answering my questions.

The various sunlight deprived facilities of the Mark Wainwright Analytical Centre:

The majority of the microscopy data was acquired in the Biomedical Imaging Facility (BMIF). I am very grateful for their microscopes and knowledgable & sociable staff, in particular Alex Macmillan and Michael Carnell.

Russell Pickford from the Bioanalytical Mass Spectrometry Facility (BMSF) provided Abbie and me with mass spec advice and new & intact phospholipid standards.

I am just as grateful for the Flow Cytometry Facility resources with ALL the colours and Chris Brownlee.

When it comes to flow cytometry the support of BD Biosciences in the person of Andrew Lim and Martin Baker deserves a mention as well.

Thanks to Sophie Pageon for reading a statistically significant amount of thesis with data that was not significantly different.

Thanks to Margaret Fennen for reading all of the thesis and add many missing commas.

Thanks to my co-supervisor Thomas Grewal for reading corrections and a lot of helpful discussions and advice over the years.

Thanks also to my supervisor Katharina Gaus for reading corrections and helpful suggestions and for showing me what it means to be successful researcher.

Thanks to my family for all the love and support despite the diameter of a whole planet between us!

Thanks to Simon for all the love and support and putting up with me the last few years despite the close proximity.

Thanks to Abbie for help, advice and for doing the PhD experience with me.

Thanks for being there near and far: Reut, Lies, Alison.

Figure 1-1: TCR signalling...... 4 Figure 1-2: Immune synapse structure...... 6 Figure 1-3: IL-2 receptor consists of three subunits – CD25, CD122 and ɣc...... 9 Figure 1-4: Structure of membrane lipids (1)...... 11 Figure 1-5: Structure of membrane lipids (2)...... 12 Figure 1-6: The effect of polar head group and chain length of fatty acid moieties on the transition temperature of phospholipids...... 13 Figure 1-7: Models of membrane organisation...... 19 Figure 1-8: AnxA6 expression is particularly high in immune cells...... 23 Figure 1-9: Structure of bovine AnxA6 with calcium (red spheres)220,221...... 24 Figure 3-1: Treatment of mice to elicit delayed-type CHS in vivo...... 54 Figure 3-2: AnxA6-/- mice display normal levels of T cell subsets...... 56 Figure 3-3: Animals used in in vivo assay do not differ in age and weight...... 57 Figure 3-4: Successful elicitation of delayed-type hypersensitivity response...... 59 Figure 3-5: T cell development and expression of surface markers on naïve and effector CD4+ T cells in wild type (green bars) AnxA6-/- (orange bars) CD4+ T lymphocytes after CHS induction...... 62 Figure 3-6: T cell development and expression of surface markers on naïve and effector CD8+ T cells in wild type (green bars) AnxA6-/- (orange bars) CD4+ T lymphocytes after CHS induction...... 64 Figure 3-7: T cell proliferation in wild type (green bars) AnxA6-/- mice (orange bars) after immune challenge...... 68 Figure 4-1: Characterisation of wild type T cells isolated from spleen...... 77 Figure 4-2: Localisation of AnxA6 and Actin at the immune synapse...... 80 Figure 4-3: dSTORM imaging of F-actin with activated primary wild type and AnxA6-/- T cells...... 83 Figure 4-4: Migration efficiency of wild type and AnxA6-/- primary T cells...... 85 Figure 4-5: Phosphorylation of signalling in early T cell activation and AnxA6 expression in AnxA6KD and CTRL Jurkat cells...... 88

Figure 4-6: Phosphorylation of signalling in early T cell activation, AnxA6 expression and IL-2 mRNA production of wild type and AnxA6-/- primary T cells...... 89 Figure 4-7: Assay principle of IL-2 secretion assay...... 92 Figure 4-8: Level of IL-2 secreting wild type and AnxA6-/- T cells...... 94 Figure 4-9: Analysis of IL-2 receptor signalling in wild type and AnxA6-/- T cells...... 98 Figure 4-10: Endocytosis of the IL-2 receptor subunit CD122 in wild type and AnxA6-/- T cells...... 101 Figure 5-1: Principle of Laurdan imaging...... 113 Figure 5-2: Laurdan imaging and analysis of plasma membrane order of CTRL and AnxA6 knock-down Jurkat cells bound to -coated beads...... 115 Figure 5-3: Laurdan imaging and analysis of plasma membrane order of primary wild type and AnxA6-/- T cells bound to antibody-coated beads...... 116 Figure 5-4: Laurdan imaging and analysis of plasma membrane order of not activated primary wild type and AnxA6-/- T cells...... 117 Figure 5-5: Isomers and isobars of PC 34:1...... 120 Figure 5-6: Schematic of triple quadrupole mass spectrometer with an ESI ion source. .. 121 Figure 5-7: Analysis of the distribution of phospholipid classes in lysates of primary T cells from AnxA6-/- (orange) and wild type mice (green)...... 124 Figure 5-8 - Cholesterol in whole T cell lysates...... 135 Figure Appendix-1: Functionality of RNAi vector and Vector map of lentiviral vector with puromycin resistance ...... 181 Figure Appendix-2: Migration efficiency of CTRL and AnxA6KD Jurkat cells...... 182

Table 2-1: Antibody and protein conjugates used for flow cytometry...... 38 Table 2-2: Activating for coating surfaces...... 40 Table 2-3: Antibodies used for detecting protein bands on western blot membranes ...... 41 Table 2-4: Lipids used in internal standard mix...... 48 Table 2-5: Scan frequencies of Qtrap5500 mass spectrometer employed in this study ...... 49 Table 5-1: Examples fatty acids found in mammals and their common names ...... 119 Table 5-2: Analysis of fatty acids in PA lipids in lysates of primary T cells from AnxA6-/- and wild type mice...... 125 Table 5-3: Analysis of fatty acids in PG lipids in lysates of primary T cells from AnxA6-/- and wild type mice...... 126 Table 5-4: Analysis of fatty acids in PI lipids in lysates of primary T cells from AnxA6-/- and wild type mice...... 127 Table 5-5: Analysis of fatty acids in PS lipids in lysates of primary T cells from AnxA6-/- and wild type mice...... 128 Table 5-6: Analysis of fatty acids in PE lipids in lysates of primary T cells from AnxA6-/- and wild type mice...... 129 Table 5-7: Analysis of isobaric PC lipid species in lysates of primary T cells from AnxA6-/- and wild type mice...... 130 Table 5-8: Analysis of SM molecular species in lysates of primary T cells from AnxA6-/- and wild type mice...... 131 Table 5-9: Analysis of PE ether lipid species in lysates of primary T cells from AnxA6-/- and wild type mice...... 132 Table 5-10: Analysis of PC ether lipid species in lysates of primary T cells from AnxA6-/- and wild type mice...... 133 Table 6-1: Primer sequences for realtime PCR assays ...... 183

1-9 7KC 7-ketocholesterol

A ACEC Animal Care and Ethics Committee

ADAP adhesion and degranulating adapter protein

AnxA6 Annexin A6

AnxA6-/- Annexin A6 knock-out

AnxA6KD stable AnxA6 knock-down cell line with Jurkat E6.1 back- ground and puromycine selection marker

AP1 activator protein 1

APC professional antigen presenting cell

ATP adenosinetriphosphate

B BCA bicinchoninic acid

BHT 2,6-Di-tert-butyl-4-methylphenol

Bis-Tris bis(2-hydroxyethyl)imino-tris(hydroxymethyl)methane

BLIMP-1 B-lymphocyte maturation protein

BrdU bromodeoxyuridine

BSA bovine serum albumin

C CD cluster of differentiation

CD25 IL-2 receptor α-subunit

CD122 IL-2 receptor β-subunit

Cdc42 cell division control protein 42 homolog

CID collision-induced dissociation

CHO cells Chinese hamster ovary cells

CHS contact hypersensitivity

cSMAC central supramolecular activation cluster

CTRL control cell line to AnxA6KD with puromycine selection marker

D DAG diacylglycerol

DC dendritic cell

DHA docosahexaenoic acid (22:6)

DIC dichroic filter

DMSO dimethylsulfoxide

DNFB dinitrofluorobenzene (or 1-fluoro-2,4-dinitrobenzene)

dSMAC distal supramolecular activation cluster

dSTORM direct stochastic optical reconstruction microscopy

DTT dithiothreitol

E ECL enhanced chemoluminescence

EDTA ethylenediaminetetraacetic

ER endoplasmic reticulum

ERK extracellular signal-regulated kinase

Elk1 ETS domain-containing protein Elk-1

EPA eicosapentaenoic acid (20:5)

ESI electrospray ionisation

F FA fatty acid

FACS fluorescence assisted flow cytometry

F-actin filamentous actin

FBS foetal bovine serum

FMO fluorescence minus one

G GADS GRB2-related adapter downstream of Shc

GAPDH glyceraldehyde-3-phosphate dehydrogenase

G6PDx murine glucose-6-phosphate dehydrogenase

GEF guanine nucleotide exchange factor

GDP guanosine diphosphate

GP generalised polarisation

GPI glycosylphosphatidylinositol

GRB2 growth factor receptor-bound protein 2

GTP guanosine-5'-triphosphate

H h hour

HCl hydrochloric acid

HDL high-density lipoprotein

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HMG-CoA 3-hydroxy-3-methyl-glutaryl-Coenzym A

HPK1 hematopoietic progenitor kinase

I ICAM intercellular adhesion molecule

IL Interleukin

IFNɣ1 interferon ɣ1

IP3 inositol 1,4,5-trisphosphate

ITAM tyrosine-based activation motif

ITK interleukin-2-inducible T cell kinase

J JAK Janus kinase

K Kd dissociation constant

L LAT linker for activation of T cells

Laurdan 6-dodecanoyl-2-dimethylaminonaphthalene

LC liqud chromatography

LCK lymphocyte-specific protein tyrosine kinase

Ld liquid disordered (phase)

LDL low-density lipoprotein

LFA-1 leukocyte function-associated antigen 1

Lo liquid ordered (phase)

M MAPK mitogen-activated protein kinase

MEK MAP/ERK kinase

MFI median fluorescence intensity

MHC major histocompatibility complex

min minutes

MOPS 3-(N-morpholino)propanesulfonic acid

MOI multiplicity of Infection

MTBE methyl tert-butyl ether

MS mass spectrometry

m/z mass-to-charge ratio

N NCK non-catalytic region of tyrosine kinase

NFAT nuclear factor of activated T cells

NFκB nuclear factor kappa-light-chain-enhancer of activated B cells

NK cell natural killer cell

NLS neutral loss scan

NPC protein Nieman-Pick type C protein

ns (statistically) not significant

N-WASP neural WASP

P PA phosphatidic acid

PAGE polyacrylamide gel electrophoresis

PBS phosphate-buffered saline

PCR polymerase chain reaction

PCNA proliferating cell nuclear antigen

PE phosphatidylethanolamine

PE-Cy7 (stain) phycoerythrin -Cy7 dye

PE-O phosphatidylethanolamine etherlipid

PFA paraformaldehyde

PG phosphatidylglycerol

PH domain pleckstrin homology domain

PI phosphatidylinositol

PI3K phosphatidylinositol-4,5-bisphosphate 3-kinase

PIP2 phosphatidylinositol 4,5-bisphosphate

PIP3 phosphatidylinositol 3,4,5-trisphosphate

PIS precursor ion scan

PKC protein kinase C

PLA2 phospholipase A2

PLCɣ1 phospholipase Cɣ1

PM plasma membrane

pMHC peptide-presenting MHC

PS phosphatidylserine

pSMAC peripheral supramolecular activation cluster

PVDF polyvinylidene difluoride

Q Q1, Q2, Q3 quadrupole 1,2,3

qPCR quantitative polymerase chain reaction

R RIPA radioimmunoprecipitation assay

V VCAM vascular cell adhesion molecule

VLA very late antigen

W WASP Wiskott-Aldrich syndrome protein

WAVE WASP family verprolin-homologous

WT wild type mouse strain C57Bl/6

S SDF-1α stromal cell-derived factor 1α

SDS sodium dodecyl sulphate

sec seconds

SLP76 SH2 domain-containing leukocyte protein of 76 kDa

SM sphingomyelin

SNARE soluble NSF attachment protein receptor

SOS son of sevenless homologs

SR-B1 scavenger receptor B1

STAT signal transducer and activator of transcription

T TBS Tris-buffered saline

TCR T cell receptor

TIRF total internal reflection

TNF tumor necrosis factor

Treg regulatory T cell

Tris tris(hydroxymethyl)aminomethane

Z ZAP70 ζ-chain associated protein kinase of 70 kDa

3-letter 1-letter Amino acid 3-letter 1-letter Amino acid Ala A Alanine Lys K Lysine Arg R Arginine Met M Methionine Asn N Asparagine Phe F Phenylalanine Asp D Aspartic acid Pro P Proline Cys C Cysteine Ser S Serine Glu E Glutamic acid Thr T Threonine Gln Q Glutamine Trp W Tryptophan Gly G Glycine Tyr Y Tyrosine His H Histidine Val V Valine Ile I Isoleucine Xaa x Unspecified Leu L Leucine

The immune system is essential to protect the body from external threats like pathogens and certain chemical agents as well as from enemies from within like cancer cells. The cells that are involved in fighting these threats are part of the innate and the adaptive immune system. The innate immune system is the evolutionarily oldest part of the immune system and constitutively active. It is very efficient in eradicating the majority of infections, but some pathogens get past this first line of defence. A further aspect of innate immunity is that it does not “learn” to recognise recurring infections caused by the same pathogen. However, the adaptive immune system is able to form an immunological memory of previous infections. The cellular components of the adaptive immune system are comprised of B and T lymphocytes. These lymphocytes develop from lymphoid progenitor cells which can differentiate into innate immune cells, like dendritic cells (DC) and natural killer (NK) cells, as well as B and T lymphocytes. T lymphocytes or T cells, which link the cell-mediated and humoral aspect of the immune response, contain very diverse subsets, most notably Treg (regulatory T cells), cytotoxic T cells and T helper cells.

The thymus is a primary lymphoid tissue which is very active in childhood and starts to shrink in adolescent individuals. Here, hematopoietic progenitor cells mature into T cells. They are submitted to a rigorous selection process to ensure all circulating T cells are able to distinguish between self (positive selection) and non-self (negative selection). The positive selection process ensures that T cells can react with, and recognise, the body’s own MHC (major histocompatibility complex). The negative selection process safeguards that only T cells that do not interact strongly with self-peptides are released from the thymus. After this positive and negative selection process, thymocytes differentiate into naïve CD4+ T helper or Treg cells and CD8+ T cells. Cognate antigen presented by an MHC will activate these T cells and induce proliferation and differentiation into effector T cells and memory T cells1.

After undergoing positive and negative selection and leaving the thymus, naïve CD4+ and CD8+ T cells require two signals to be activated: an antigenic peptide-presenting MHC (pMHC) specific to their TCR (T cell receptor) and a co-stimulatory signal2. Initially, naïve T cells will be activated by a professional antigen presenting cell (APC), like a DC. DCs express B7 receptors on their surface, which deliver a co-stimulatory signal via the CD28 molecule resident on the surface of T cells. The pMHC recognised by T cells can belong to one of two classes: MHC class I, expressed on all cells; and MHC class II, only expressed on APCs. MHC class I present samples of intracellular peptides on the outside of every cell, while MHC class II can also present antigenic peptides that have been engulfed by the APC from the surrounding environment. T cells are restricted in their ability to interact with MHC molecules. CD4+ T cells only interact with pMHC class II1,3 which limits their interactions to contacts with other immune cells. CD8+ T cells, however, can only form a productive interaction with an MHC class I. Even though MHC class I is presented on the surface of all cells, the first activation of a naïve CD8+ T cell has to occur with an APC, since they are the only cells that express MHC class I as well as the co-stimulatory molecule B7.

Naïve T cells circulate through the blood stream and the lymphatic vessels. Along the way they pass through secondary lymphatic organs like the lymph nodes and the spleen. Lymph nodes are the main location for T cells to be activated. Here, DCs present antigens to T cells4. The initial activation of naïve T cells takes 4–5 days during which they undergo clonal expansion. These clones differentiate into effector T cells and leave the lymph node5. Effector T cells do not require a co-stimulatory signal to activate and can therefore act faster on pathogens presented to them6. CD8+ T cells are destined to differentiate into cytotoxic T cells. Cytotoxic T cells destroy cells presenting pathogenic antigen through their MHC class I. Activated CD4+ cells can differentiate into different types of effector

T cells: Th1, Th2, Tfh, Th17 cells. The fate of activated CD4+ T cells is determined by the quality of the interaction between APC and T cell and the presence of different types of pathogen7. A weak APC/T cell interaction will lead to the development of Th2 functions while strong or frequent interactions are likely to create Th1 cells8. Different classes of pathogen are indicated to the T cells by specific combinations of cytokines in the immediate environment, secreted by macrophages or NK cells or other T cells. Th1 cells, for example, are induced by Interleukin (IL) 12 and IFNɣ1 (interferon ɣ1), while IL-4 and

IL-6 promote the differentiation into Th2 cells. Viruses and bacteria induce IL-12 in DCs, which stimulate NK cells to produce IFNɣ1, while helminths induce the production of IL-4 in NK1.1 T cells. Th2 cells produce IL-4 and thus induce the differentiation and proliferation of more Th2 cells9. Their main function is to stimulate B cells to differentiate

into antibody producing plasma cells. Th1 cells, on the other hand, secrete IL-2, IFNɣ1 and TNFβ (tumor necrosis factor), which will stimulate macrophages immune cells that are able to destroy bacteria by phagocytosis.

Effector T cells, that are not receiving stimulation through their TCR or via proliferation- inducing cytokines such as IL-2, undergo apoptosis. However, the immune system does not have to rely solely on activating naïve T cells if an infection with the same pathogen re- occurs, because some CD4+ and CD8+ T cells develop memory phenotype T cells. They can respond much faster than naïve T cells to pMHC stimulation, since they, like effector T cells, do not require confirmatory co-stimulation to activate. It is not clear whether these memory T cells arise from effector T cells that are allowed to survive and differentiate or whether they develop directly from activated T cells10. In the absence of co-stimulatory signals, like the CD28/B7 interaction, memory T cells proliferate rapidly upon activation to generate more effector T cells until the infection is cleared. Memory T cells usually provide lifelong immunity against their cognate antigen.

The activation of a T cell initiates dramatic reorganisation of the cell, involving the organisation of the cytoskeleton, the plasma membrane and the distribution of signalling proteins in, and attached to, the plasma membrane. The part of the T cell that interacts with the APC forms a structure which has been termed immunological synapse11,12. Here, TCR coalesce and form small clusters. The site is also enriched in co-stimulatory receptors and ligands to form a tight connection with the APC. The structure is stabilised by a ring- like assembly of actin. In a resting T cell actin is mostly depolymerised, but due to activation signals from the TCR F-actin (filamentous actin) structures form to provide stability to the contact zone of the two cells. Meanwhile, organelles and intracellular structures like the Golgi complex and microtubule organising centre are reorganised and now located closer to the activation site13, whereas the distal pole complex forms on the opposite side of the cell14. Finally, sustained signalling activity at the immunological synapse leads to fully activated T cells that release IL-2 (interleukin 2). IL-2 acts as an autocrinal cytokine by triggering proliferation and differentiation of the activated T cell it

1 Contains text from Cornely et al. 2012187

originated from. Likewise, IL-2 can induce proliferation and differentiation of other cells in the vicinity of activated T cells.

TCR signalling (Figure 1-1) starts when recognition of a pMHC by the TCR is confirmed by the T cell co-receptor CD4 or CD8 which binds to MHC. TCR ligation triggers the segregation of the T cell membrane and the assembly of multi-molecular TCR signalling clusters. During this time, TCR is phosphorylated at the ζ-chain and CD3ε as part of the TCR complex is activated. This recruits the Src family kinase, LCK (lymphocyte-specific protein tyrosine kinase), which phosphorylates the TCR at ITAMs (tyrosine-based activation motifs). The modified ITAMs are able to bind ZAP70 (ζ -chain associated protein kinase of 70kD) and ZAP70 is phosphorylated by LCK. Thus activated by LCK, ZAP70 subsequently phosphorylates the transmembrane protein LAT (linker for activation of T cells), which acts as a scaffold for adaptor proteins GRB2 (growth factor receptor-bound protein 2) and SLP76 (SH2 domain-containing leukocyte protein of 76 kDa), and signalling enzymes including PLCɣ1 (phospholipase Cɣ1) and PI3K (phosphatidylinositol-4,5- bisphosphate 3-kinase). This leads to TCR-LAT oligomerisation and the formation of TCR- LAT microclusters.

Figure 1-1: TCR signalling. TCR signalling is initiated by the interaction of pMHC with the TCR complex and induces a plethora of cellular processes. The Akt pathway promotes cell survival as well as cell proliferation. MAPK and NFκB pathways cell proliferation and cell cycle progression. IP3 release triggers Ca2+ signalling which initiates IL-2 production via NFAT. Morphological changes involving the F-actin cytoskeleton are mediated by WASP. ADAP promotes cell adhesion. The figure was reproduced from a review by Rossy and colleagues15.

TCR/LAT microclusters containing proteins like LCK, ZAP70 and SLP7616,17 initiate a multitude of signalling pathways that result in morphological and transcriptional changes

in the T cells. The restructuring of the cytoskeleton is mediated by SLP7618,19 which is connected to LAT through the adaptor GADS (GRB2-related adapter downstream of Shc)20. SLP76 forms a complex with guanine nucleotide exchange factor (GEF) Vav119,21,22 and ITK (interleukin-2-inducible T cell kinase)23,24 recruiting PLCɣ125–27. In this process PLCɣ1 is phosphorylated and activated by ITK28. Phosphorylated SLP76 also recruits NCK (non- catalytic region of tyrosine kinase)29–31 which then binds WASP (Wiskott-Aldrich syndrome protein)32,33. At the same time, SLP76 brings Vav1 in proximity with the small GTPase Cdc42 (cell division control protein 42 homolog), inducing the localized activation of Cdc4231. Active guanosine-5'-triphosphate (GTP)-Cdc42 then binds WASP and unblocks its autoinhibition34. WASP assumes a key role in restructuring the F-actin cytoskeleton as it activates Arp2/3, a nucleation protein for actin, which initiates actin polymerisation. Mutations in the WASP gene can cause different diseases, but individuals with loss of function (e.g. Wiskott-Aldritch syndrome) as well as activating mutations in WASP are more susceptible to infection34–36. SLP76 also promotes cell adhesion via ADAP (adhesion and degranulation promoting adapter protein)37–39 which ensures that the connection of T cell and APC is continuous rather than transient.

PLCɣ1-mediated signalling is facilitated by membrane as well as membrane proximal protein components. PLCɣ1 cleaves PIP2 (phosphatidylinositol 4,5-bisphosphate) into DAG

(diacylglycerol) and IP3 (inositol1,4,5-trisphosphate). DAG acts as a lipid second messenger in the plasma membrane and activates membrane-bound PKC (protein kinase C)40 signalling via NFκB (nuclear factor kappa-light-chain-enhancer of activated B cells)41. This results in transcriptional changes that promote cell division and suppress apoptosis. DAG can also recruit the GEF RasGRP to the plasma membrane where it activates Ras28 which signals via MAPK (mitogen-activated protein kinase) signalling cascade. An alternative path to Ras activation is the binding of a GRB2/SOS (Son of sevenless) complex to LAT42. As GRB2 binds to phosphorylated LAT it is bound to the GEF SOS. Thus targeted to the membrane in proximity to Ras, SOS is able to activate Ras43. In the course of the ensuing MAPK cascade, ERK1/2 (extracellular signal-regulated kinase) is activated and induces the transcription factors c-FOS and c-JUN, both of which are involved in prompting cell growth and proliferation44.

The soluble product of PIP2 hydrolysis, IP3 acts as a cytosolic second messenger. IP3 binds to Ca2+ channels in the endoplasmic reticulum (ER) and triggers the release of Ca2+ into the cytosol45. There, Ca2+ initiates NFAT (nuclear factor of activated T cells) signalling by activating the Ca2+ binding proteins calcineurin and . Calcineurin promotes the

translocation of NFAT to the nucleus where it induces the transcription and production of IL-2. Calcium signalling via PLCɣ1 also induces activation of PI3K.

The LAT supported protein signalling network utilises PIP2 for a second signalling pathway. PI3K can phosphorylate the head group to PIP3 (phosphatidylinositol 3,4,5- trisphosphate) which is the initiation of the Akt signalling pathway that furthers cell survival as well as cell proliferation.

HPK1 (hematopoietic progenitor kinase) as well as Cbl seem to be negative regulators of T cell activation and play a role in preventing hyperactivation and autoimmunity. Knock- out of HPK1 as well as Cbl family proteins led to enhanced TCR signalling and made mice more susceptible to autoimmune defects46–50.

Figure 1-2: Immune synapse structure. A signalling T cell forms a characteristic membrane structure, the immune synapse containing the concentric SMAC assemblies, each with a characteristic set of membrane proteins: cSMAC (enriched in TCR), pSMAC (LFA-1) and dSMAC (CD45). The figure was reproduced from a review by Rossy and colleagues15.

Prolonged contact to an antigen-presenting cell leads to the formation of a mature immunological synapse with a bull’s eye pattern11,51 (Figure 1-2): a central supramolecular activation cluster (cSMAC) enriched in TCR, LAT, LCK, ZAP70 and PLCɣ1 surrounded by the peripheral SMAC (pSMAC), which is mainly enriched in F-actin and the adhesion molecules LFA-1 (leukocyte function-associated antigen 1) and talin52. A third ring-like structure around the cSMAC and pSMAC is sometimes referred to as distal SMAC (dSMAC), which contains the tyrosine phosphatase CD45. CD45 is probably excluded from the centre of the immune synapse to prevent interfering with the phosphorylation processes, particular of LCK, driving T cell activation53. Although the SMAC structures are not essential for T cell activation54, the cSMAC seems to have an important function in signal modulation55 and termination56,57.

IL-2 is a 15 kDa cytokine expressed mainly by activated CD4+ T cells. Once naïve CD4+ and CD8+ T cells encounter their cognate antigen presented by a DC in a lymph node, activating T cells produce large amounts of IL-2. Upon binding to the IL-2 receptor, IL-2 induces cell proliferation and transcription of its own receptor (CD25). As a result, activated T cells that differentiate into effector T cells have an increased sensitivity to IL-2. However, prolonged exposure to IL-2 eventually reduces its own production. IL-2 functions in a paracrine as well as in an autocrine fashion and stimulates a wide variety of cells. However, an individual cell will only commit to undergo cell division once a certain number of IL-2 receptors have been triggered58. IL-2 production peaks after several hours but remains elevated for 4–5 days59. In its paracrine function IL-2 stimulates other T cells but can also stimulate proliferation in B cells60 and augment NK cell function61. Receptors for IL-2 are even found on non-immune cells like fibroblasts and endothelial cells62.

In the absence of an active immune response, naïve T cells are kept alive by a different member of the interleukin family IL-7 and interactions with self-pMHC molecules, which induce submitogenic “survival” signals63. In this way, T cell numbers remain stable despite the thymic atrophy in adult individuals. During this time of homeostatic survival, IL-2 is only produced in low amounts by CD4+ cells in secondary lymphoid organs, and CD8+ and NK cells might secrete IL-2 as well in even lower amounts64,65.

The IL-2 receptor consists of three subunits, CD25, CD122 and ɣc (Figure 1-3) that contribute to affinity, specificity and signalling activity of the complete receptor66. CD25

(also known as the α-subunit) by itself binds IL-2 (dissociation constant Kd ~ 10-8 M) but cannot by itself convey any signalling response; instead, its main function seems to be to enhance the binding of IL-2 in a trimeric receptor with the other subunits. A likely mechanism for this is a conformational change induced in IL-2 through CD25 binding that improves the association of IL-2 and CD122, which means CD25 can also improve IL-2 binding if it is presented by a neighbouring cell67. CD25 is not expressed on naïve T cells but is strongly upregulated after T cell activation and on regulatory T cells. The high levels of CD25 on Tregs suggest that they might play a role in absorbing and degrading excess IL-2 in their immediate environment to regulate the immune response and prevent autoimmune pathologies68. To enhance IL-2 binding in concert with CD122 and the ɣc subunit, CD25 does not have to be present on the same cell but can also act in trans when presented by a neighbouring cell69 or even in solution70. CD122 (IL-2 receptor β-subunit) constitutes an intermediate affinity receptor for IL-2 (Kd ~ 10-9 M) in combination with the

γc-subunit. In the presence of high levels of IL-2 CD122 and γc are sufficient to induce cell proliferation71. The combination of CD122 and ɣc is also a receptor for IL-15, a cytokine that is released by mononuclear phagocytes during viral infections. The ɣc subunit is expressed on all cells of the immune system. However, CD122 is expressed at low levels on naïve CD8+, but not on naïve CD4+ T cells62. Together with CD25, CD122 is transiently upregulated on CD8+ and CD4+ effector cells in response to TCR stimulation. γc is the common receptor subunit for the IL-2 family of interleukin receptors.

All IL-2 receptor subunits together form a high affinity IL-2 receptor (Kd ~ 10-11 M) within which CD122 and γc undergo a conformational change upon IL-2 binding that enables the phosphorylation of JAK1 (Janus kinase) and JAK3, respectively72. These kinases in turn phosphorylate STAT5a (signal transducer and activator of transcription) and STAT5b, and to a lesser degree STAT1 and STAT3. STAT5a/b, as a dimer, enters the nucleus and induces further transcription of involved in cell proliferation and cytokine production.

Once the IL-2/IL-2 receptor complex is internalised, CD25 is recycled back to the cell surface while CD122 and ɣc are degraded together with their ligand IL-264. IL-2 signalling via pSTAT5 in the cell secreting IL-2 eventually inhibits its own production. In this negative feedback loop BLIMP-1 (B lymphocyte maturation protein), induced by IL-2, is mainly responsible for repressing IL-2 transcription73.

While the JAK/STAT pathway is the most prominent IL-2 receptor signalling pathway, it is only one of the pathways activated by IL-2 (Figure 1-3). The IL-2 receptor also signals via Ras/Raf/MAPK74 and PI3K75 pathways and thus activates NFκB, c-FOS, c-Myc and c-JUN in the nucleus furthering cell cycle progression. LCK is also associated with the IL-2 receptor. LCK associates with the β-subunit of the IL-2 receptor and can phosphorylate STAT5 independently from JAK1 and 376. LCK provides a reciprocal link between TCR and IL-2 receptor signalling, since STAT5 phosphorylation has been detected in TCR stimulated T cells77, just as LCK induced STAT5 signalling was detected in resting T cells stimulated with IL-246. Unlike other kinases, the role of LCK in IL-2 receptor signalling might be related to the suppression of apoptosis rather than to inducing cell proliferation78.

Figure 1-3: IL-2 receptor consists of three subunits – CD25, CD122 and ɣc. CD25 by itself can bind IL-2 but lacks the cytoplasmic domains to elicit signalling. In the presence of high concentrations of IL-2, CD122 and ɣc can form a low affinity receptor that is able to induce proliferation due to STAT5 signalling. CD25, CD122 and ɣc subunit together form the high affinity receptor (cis), which is sensitive to IL-2 even at low concentrations. Alternatively, the trimeric receptor can also be formed by binding CD25 in trans, e.g. by contact with a CD25+ DC. IL-2 binding initiates signalling via STAT5, the PI3K/AKT and the RAS/RAF/MAPK pathway, which enhances the expression of factors involved in cell proliferation and suppression of apoptosis. IL-2 receptor activation also induces the expression of its own subunit CD25. Figure adapted from Boyman & Sprent62.

The plasma membrane defines the barrier between the inside and the outside of a cell. Made up of a semipermeable lipid bilayer, the membrane regulates import and export into the cell. While some small uncharged molecules, like dissolved gases, and some fat soluble compounds, like steroid hormones, can diffuse through the membrane, exit and entry of most molecules from and into the cytosol are highly regulated and controlled by specialised exchange protein-based machineries, like ion channels, and mechanisms of endo- and exocytosis. Injury of this barrier can result in cell death.

The membrane is made-up of three different types of amphipathic lipids: phospholipids, glycolipids and sterols. Phospholipids and glycolipids automatically assemble into lipid bilayers or vesicles due to their polar hydrophilic head group and their hydrophobic hydrocarbon tail. Phosphoglycerides (Figure 1-4 A-H) are phospholipids with a glycerol backbone esterified with two fatty acids on sn-1 and sn-2 and a phosphoester head group on the third carbon atom. The head group is bound to an eponymous small molecule, e.g. serine, ethanolamine, choline, inositol or glycerol. The two fatty acid ‘tails’ of membrane phosphoglycerides are even-numbered carbohydrate chains of dissimilar length. Other phospholipids are phosphosphingolipids, like sphingomyelin (SM, Figure 1-5 A), that composed of a sphingosine rather than a glycerol backbone and contain only one variable fatty acid. Cholesterol (Figure 1-5 B) is a sterol lipid which due to its size and small polar residue is able to fill the ‘gaps’ between the fatty acid tails making it an important regulator of bilayer permeability and fluidity. Ether lipids are another type of phospholipid found in cellular membranes, which usually carry choline or ethanolamine head groups. They are connected to the phosphoglycerol backbone through an ether linkage (Figure 1-5 D) or, in the subgroup of plasmalogens (Figure 1-5 E), through a vinyl ether. They are suspected to play a role in membrane fusion and might act as antioxidants in the membrane. However, despite their abundance (up to 18% of phospholipid mass), their role is not very well understood79. Analysis and quantification of these compounds is problematic, since their anti-oxidant properties render them unstable in vitro80. As a result their role in the cell function is not very well understood. There is no doubt, however, about their importance, since ether lipid deficient mice display defects in spermatogenesis and the central nervous system, and develop cataracts80.

Figure 1-4: Structure of membrane lipids (1). (A) The basic structure of a glycerophospholipid. R1 usually carries a saturated and R2 an unsaturated fatty acid. The polar headgroup consists either of hydrogen in (B) phosphatidic acid (PA), (C) ethanolamine, (D) serine, (E) choline or (F) glycerol (phosphatidylglycerol=PG). (H) This lipid is an example how cis-double bonds introduce a “kink” into the fatty acid residue.

Figure 1-5: Structure of membrane lipids (2). (A) Sphingomyelin consists of long chain base sphingosine (blue) linked to choline (yellow) via a phosphodiester (green). The R1 of SM usually carries a saturated, sometimes monounsaturated, fatty acid. (B) Cholesterol contains only a very small polar residue (OH, yellow). (C) The majority of phosphoglycerolipids are diacylphospholipids, but in etherphospholipids (D and E) the fatty acid at sn-1 is replaced by an ether moiety. (E) In a plasmalogen phospholipid the moiety at sn-1 is a vinyl-ether.

The ratio of different lipid classes, the length of the fatty acid chains and the nature of the polar head group are important modifiers of membrane fluidity. The degree of saturation of the fatty acid chains has important consequences for the fluidity and permeability of the plasma membrane81. Saturated fatty acids make the plasma membrane more rigid, while unsaturated fatty acids make it more liquid – akin to the saturated fatty acids in butter and unsaturated fatty acids in vegetable oil that at 20℃ result in solid or liquid aggregate state, respectively. Saturated fatty acids can extend straight into the lipid bilayer and are easily organised in a tight lattice. Longer fatty acid chain lengths result in more densely packed, rigid membranes as they can build up more hydrogen bonds and van der Waals’ forces between the fatty acid tails of the phospholipids (Figure 1-6)82. Mono- or polyunsaturated fatty acids contain “kinks” where their cis-double bonds are located (Figure 1-4 H). These bends contribute to a greater degree of membrane fluidity81 and they reduce the chances of close interaction with proximal lipid molecules resulting in lower van der Waals’ interactions. The polar head group also influences the transition temperature of the lipid and, therefore, membrane viscosity83,84 – e.g. it was found that treatment with organic solvent in bacteria shifted the lipid composition towards lipids with higher transition

temperatures85. Phosphatidylcholine (PC), -ethanolamine (PE) and -serine (PS) lipids equipped with the same chain length saturated fatty acids will have different transition temperatures (Figure 1-6). In comparison, PC generates more liquid membranes while PS and PE could contribute to a more rigid bilayer, but the influence of polar head groups on membrane fluidity has not been researched extensively.

SM usually carries a saturated or monounsaturated fatty acid. As a result, SM-rich regions of the plasma membrane are less fluid86. Apart from the length and degree of saturation of the phospholipid alkyl chain, the amount of cholesterol strongly influences the fluidity and order of the lipid bilayer. The more cholesterol and long chain saturated lipids reside in the membrane, the more rigid and ordered it becomes87.

Figure 1-6: The effect of polar head group and chain length of fatty acid moieties on the transition temperature of phospholipids. The transition temperatures from gel to liquid of PE (purple), PS (blue) and PC (green). The lipids in this graph each contain two saturated fatty acids of the same length. Longer fatty acid chain lengths lead to higher transition temperatures. The polar head groups also influence the transition temperature, which impacts the fluidity of the bilayer (i.e. lipids with a lower transition temperature increase the fluidity of the membrane)88. The transition temperature is higher in PE and PS phospholipids than in PC. Values of transition temperatures were taken from Avanti Polar Lipids (AL, United States) product information.

Essential to the integrity and function of the plasma membrane lipid bilayer is its asymmetry and the ability to segregate certain types of lipid and confine them to the inner or outer leaflet of the membrane and to form specialised domains within the bilayer. Not only the plasma membrane, but all cellular membranes, except for the ER, have an asymmetric distribution of lipids between the two leaflets of the lipid bilayer89. While the outer leaflet of the plasma membrane contains mostly PC, SM and glycosphingolipids, the inner leaflet is enriched in PS, PE and phosphatidylinositol (PI). Cholesterol is present in both leaflets and in all other eukaryotic membranes, with the lowest levels in mitochondrial membranes89. The highest amount of cholesterol is found in the plasma membrane where molar ratio of cholesterol and phospholipids is up to ~1 in some cell

types90. Adenosinetriphosphate(ATP)-dependent aminophospholipid transporters (flippases) selectively translocate PE and PS to the cytosolic leaflet91,92. Counteracting this transport are scramblases which transport phospholipids either way. The asymmetry of lipid bilayers is also essential for basic cellular processes like vesicle traffic that is abrogated in the absence of these transporters92,93.

PS is found almost exclusively on the cytosolic membrane leaflet with the exception of the ER91. PS concentration is highest in the plasma membrane and endosomes, but very low in mitochondria. Its asymmetric distribution is strictly regulated, since the exposure of PS on the outside of the cell, which is probably mediated by scramblases, is part of a mechanism that communicates apoptosis by PS exposure to other cells94. The presence of PS on the outside of the cell signals that this particular cell is entering apoptosis and serves as an “eat-me” cue for phagocytes95. The plasma membrane of blood platelets loses asymmetry when they are activated by collagen and thrombin released during injury of blood vessels. In this scenario, surface PS facilitates the binding of various factors promoting blood coagulation94. Like cholesterol, PS is not distributed homogeneously in the plasma membrane. It has been shown to be enriched in caveolae but not in clathrin-coated pits and was found in “cosegregation […] with cholesterol and sphingolipids”96. Furthermore, PS shows the second highest affinity for cholesterol out of all phospholipid classes after SM97. PS is synthesised in the ER by phosphatidylserine synthase 1 and 2 through exchanging the head group of PC or PE with serine. The knock-out of either enzyme does not affect viability in mice, but a double knock-out is lethal98,99. PS synthesised in the ER is transported to target organelles where flippases translocate PS (as well as PE) to the cytosolic leaflet91,92. PS also participates in intracellular signalling as a binding site for C2 protein domains (e.g. PKC)100, Akt signalling (through pleckstrin homology (PH) domains in Akt)101 and annexin proteins (discussed in more detail in Chapter 1.3).

Cholesterol is a major constituent of all animal membranes and a precursor for hormones, vitamin D and bile acids. As a membrane component it helps to regulate the fluidity of the plasma membrane and reduces the permeability for ions and small molecules like glucose102–104. Cholesterol consists of a rigid planar ring system with a polar OH group on one end and a hydrophobic hydrocarbon chain on the other end. The presence of (20- 25mol%) cholesterol prevents a sharp crystallisation point of phospholipid membranes and thus keeps membranes fluid in a wider temperature range. However, a further increase of cholesterol, leads to more densely packed lipids and decreases membrane fluidity105–107. The latter feature contributes to the higher degree of membrane order in

liquid ordered (Lo) phases in cellular membranes, which are enriched in cholesterol and will be discussed in more detail later.

Cholesterol can be absorbed from food or synthesised de novo by individual cells. Large- scale cholesterol production occurs in the liver from where it is transported via blood and lymphatic vessels to other tissues in the form of low-density lipoprotein (LDL) particles. These are recognised by LDL receptors, which are expressed on the cell surface according to the requirement for cholesterol in the cell, and are endocytosed via clathrin-coated vesicles108. From there they are transported to acidic endocytic compartments where cholesterol esters are hydrolysed making free cholesterol available for the cell108. Cholesterol is then delivered from late endosomes which are enriched in cholesterol to other organelles, like ER, mitochondria, recycling endosome, and plasma membrane. Imported cholesterol, which cannot be accommodated in cellular membranes, is stored as an fatty acid ester lipid in lipid droplets together with triglycerides or exported from the cell to high-density lipoproteins (HDL)108,109.

The several stages of cholesterol synthesis take place in three different cellular compartments: the cytosol, the peroxisome and the lumen of the ER. Short precursors for the synthesis are created in the cytosol. The key enzyme in this part of the synthesis is HMG-CoA-reductase (3-hydroxy-3-methyl-glutaryl-Coenzym A); the expression and activity of which is tightly regulated. After the final steps of cholesterol synthesis in the ER, cholesterol is transported rapidly to other cellular membranes, in particular the plasma membrane, or esterified for storage in liposomes108. This transport does not appear to depend on vesicles or the Golgi complex110. It is not clear how this process is regulated. However, the fact that the ER, as the main location of lipid synthesis in the cell, contains the lowest concentration of cholesterol implies that cholesterol is selectively removed and transported to other cellular locations. From the ER cholesterol is likely distributed to other membranes without the help of vesicles. It has been proposed that cholesterol could be bound to sterol-binding proteins instead, or that ER membranes engage in direct contact with other organelles for cholesterol transfer108.

Aside from lipids, proteins can span, integrate, or be tethered to the lipid bilayer and thus influence fluidity, permeability and membrane organisation as well. The localisation and orientation of membrane proteins is either intrinsic to their amino acid sequence and folding (e.g. amphiphile α-helices), or regulated by posttranslational modifications (e.g. acylation), which target them to the membrane111. Asymmetry of the lipid bilayer is created and maintained by proteins89, which illustrates the tightly linked interaction of

lipids and proteins in the plasma membrane. Furthermore, the enrichment of charged phospholipids in certain cellular membranes facilitates electrostatic targeting of cationic membrane-binding proteins to individual cellular organelles. PS, exposing one negative charge, and phosphoinositides, exposing several negative charges, are enriched in the inner leaflet of the plasma membrane and can attract cytosolic proteins with strong positive charges (e.g. by Ca2+-binding and poly-lysine tails)112. The plasma membrane and the endosomes expose the highest number of negative charges113.

Membrane lipids are not just of biophysical significance for the cell but are also providing precursors for cellular signalling112,114. PIP2 is an important intermediate of the IP3/DAG pathway, which induces Ca2+ release into the cytosol. In this pathway, PIP2 is cleaved into

DAG and IP3 by the active PLCɣ1. IP3 receptors at the ER open ion channels that release

Ca2+ into the cytosol. Other phospholipases like PLA2 (phospholipase A2) cleave arachidonic acid from the phospholipid backbone and thereby release precursors of eicosanoids, which act as inflammatory mediators once they are converted into prostaglandins or leukotrienes115.

To describe dynamics of lipids and proteins within the plasma membrane Singer and Nicolson proposed the “fluid mosaic” membrane model116 in the 1970s. In this model, lipids and proteins are more or less freely diffusing in the two-dimensional fluid of a lipid bilayer (Figure 1-7 A). At the time, the fluid mosaic membrane model displaced other models, like the unit membrane model, which placed more emphasis on the importance of proteins and assumed a more static lipid membrane with an close connection to a protein layer above and below the lipid membrane117,118. It soon became clear that the organisation of the plasma membrane was more complex than either model proposed. While lateral diffusion of transmembrane proteins could be observed in the bilayer, the diffusion rates did not match the predictions of free diffusion119 but were often slower120– 123 suggesting diffusion-hindering interactions with cytosolic components like the cortical cytoskeleton and membrane proteins124–128. The protein content of the membrane is a substantial factor for the biophysical properties of the membrane considering it comprises around 25% of the area and about 50% of the membrane mass129. However, it can vary from 20% in myelin to 75% in the inner mitochondrial membrane114.

It has also been hypothesised that the plasma membrane lipids interact with each other and certain proteins to create functional domains130 (Figure 1-7 B). In vivo these lipid

domains could form around a receptor and attract appropriate signalling proteins, not only by protein-protein interaction, but also by protein-lipid interactions. In this scenario, lipids are enriched in a lipid domain and thus create platforms for biological processes like receptor signalling. These membrane domains are often termed “membrane rafts”131 or “lipid rafts” and are defined as small (10-200 nm), dynamic cholesterol and sphingolipid- rich domains that can coalesce to form larger platforms for signalling processes132. These structures are still controversial, since, due to their size and highly dynamic nature, they elude direct observation. Early observations in artificial bilayers showed that cholesterol preferentially partitions with sphingolipids97. The saturated fatty acid residues of sphingolipids allow a higher degree of miscibility with the rigid sterol rings and saturated fatty acid tail of cholesterol than phospholipids with unsaturated fatty acids130. It was shown that cholesterol and phospholipids were not homogeneously distributed in animal cells133–135. Initially, the components of membrane rafts were analysed by submitting cells membranes to detergent extraction and differential centrifugation – with membrane rafts being characterised as detergent-resistant. Lipids and proteins that were found to be enriched in the detergent-resistant membrane fraction included SM, glycosphingolipids, glycosylphosphatidylinositol (GPI)-anchored proteins and cholesterol. The method has since been criticised as being prone to artefacts, but has only slowly been replaced by other approaches136, particularly through microscopy techniques like fluorescence correlation spectroscopy, stimulated emission depletion and photoactivation localisation microscopy which can resolve structures below 200 nm. Membrane rafts are also characterised by more densely packed lipids which results in the creation of an Lo phase compared to other parts of the membrane (liquid disordered, Ld)137,138. These differences in membrane order can be observed with the help of lipid soluble dyes like Laurdan (6- dodecanoyl-2-dimethylaminonaphthalene) and di-ANEPPDHQ with an emission spectrum that depends on the amount of water they are exposed to139. In the Lo phases more water is excluded than in Ld phases. This spectral shift can be observed and quantified through ratiometric imaging140.

The cortical cytoskeleton also contributes to the organisation of the plasma membrane, despite not being an integral part of the plasma membrane lipid bilayer. In many cells the cortical cytoskeleton consists of F-actin structures connected by spectrin tetramers and supported by a microtubule network. The structure of the cortical cytoskeleton was connected to observations that the restricted diffusion of proteins in the membrane could be influenced by the size of the cytoplasmic domain141–143. It was observed that mutants of the transmembrane protein E-cadherin diffused more freely in the bilayer the shorter its

cytoplasmic domain was143. This resulted in the creation of the “picket fence” membrane model (Figure 1-7 C)144. Kusumi et al. postulated that cytoskeletal support beneath the plasma membrane reduced the macroscopic diffusion rate within the bilayer144–146. A series of transmembrane proteins interacting with F-actin through cytosolic domains directly or indirectly would be able to create a fence-like barrier of “pickets” within the bilayer (Figure 1-7 D). The restricted diffusion within bound membrane areas was termed “hop diffusion”147,148. The theory of a cytoskeletal fence is supported by the fact that drugs and mutations interfering with F-actin polymerisation also change the dynamics of integral membrane proteins118,124,149–151, i.e. preventing actin polymerisation increases diffusion rates124,152.

The picket fence model has since been refined to accommodate raft structures (Figure 1-7 C and D). It is now suggested that cytoskeleton supported “fences” are only the first and largest (40-300 nm) tier in a hierarchy of membrane organising structures. Lipid rafts or cholesterol-rich domains might act as a second tier contained with these corrals organising lipids and proteins in dimensions of 2-20 nm, while dynamic protein complexes, as a third tier, would operate at an even smaller range (3-10 nm)144. Examples for proteins which provide a link between cytoskeletal components and membrane proteins include ezrin-radixin-moesin proteins, which contain an F-actin-binding domain as well as a domain that targets integral membrane or membrane-binding proteins153,154. Other protein classes can provide a direct link between the cytoskeleton and the membrane via lipid binding motifs (e.g. PH domain), like N-WASP (neural WASP) and spectrin155–157. The negatively charged PIP2 is a target for many F-actin binding proteins158 and the formation of actin-membrane connections itself might contribute to the formation of Lo and Ld phases in the lipid bilayer159–162. Using artificial lipid bilayers in which a PIP2- N-WASP link between bilayer and an actin network showed that actin association can induce phase-separated domains as well as stabilise them. Actin supported PIP2- containing Ld phases161. This finding suggests that in a picket-fence model F-actin might stabilise Ld phases along actin filaments while the Lo phases are located in the “holes” of the actin mesh161,162. It is not clear how the actin/N-WASP/PIP2 interaction is able to induce this separation.

Recently, it has been suggested that membrane organising principles might not just be driven by the contents of the lipid bilayer or cytosolic components, but that extracellular lattices could also contribute to the creation of membrane domains through fence-forming on the cell surface. The extracellular protein family of galectins can form lattices with

Figure 1-7: Models of membrane organisation. Singer and Nicholson proposed the fluid mosaic model in which transmembrane (blue), integral (green) and membrane-associated proteins (yellow) freely diffuse in the 2-dimensional liquid of the bilayer. (B) In the lipid raft model certain proteins (e.g. proteins with modifications like palmitoyl-residues, yellow and green), are found in Lo domains or rafts (red-shaded areas in membrane) that are enriched in cholesterol and sphingolipids while other proteins preferentially reside outside of these domains. (C) The picket fence model allows for the compartmentalisation of the plasma membrane. The grouping of certain proteins is put down to the interactions of cytosolic domains of membrane proteins with cytoskeletal structures (red) just below the plasma membrane. Figure adapted from Williamson 2011163. (D) Viewing the plasma membrane from the cytosolic side shows corrals surrounded by F- actin (red). Transmembrane and membrane-anchored proteins interacting with the actin cytoskeleton (blue) directly or indirectly and thus act like pickets of a fence, limiting the movement of proteins within this compartment. Within these compartments the membrane could be partitioned further by membrane rafts (red shaded area) with distinct lipid and protein composition. glycans on the cell surface. There is evidence to suggest that galectin-glycan lattices organise and stabilise membrane rafts through interactions with glycosylated raft proteins164.

Signalling of plasma membrane receptors and plasma membrane associated proteins is influenced by the lipid environment82,165–169. In polarised cells, the lipid composition of the plasma membrane, and therefore the lipid environment for the resident receptors, can differ between the two poles170. This was first found in rat intestinal cells in which the brush border membrane is enriched in cholesterol and SM compared to the basolateral membrane133.

Activated T cells are strongly polarised with the majority of signalling taking place at the immune synapse. The immune synapse has a distinct lipid composition, which resembles that of membrane rafts170 and exhibits a higher degree of membrane order than the distal pole of the T cell171. The lipid reorganisation of the T cell membrane is triggered by the assembly of TCR microclusters in activated T cells171. Upon activation, TCR microclusters are rearranged to generate SMACs at the immune synapse between T cell and APC172,173. Membrane rafts, which are enriched in cholesterol, also contribute to the structure of the plasma membrane required for T cell signalling174,175. Moreover, the reorganisation of the plasma membrane is supported by the actin cytoskeleton, which is crucial to establish a lasting contact following the initial interaction between T cell and APC15,131,176. The restructuring of the actin cytoskeleton is initiated by TCR signalling via the actin- regulating proteins Vav and WASP and plays a major role in the formation of T cell microclusters and the organisation of TCR signalling within the plasma membrane177,178. The importance of the interplay of membrane domains and cytoskeletal dynamics becomes apparent when the coalescence of membrane rafts is impaired in knock-out models and genetic defects where Vav121,179,180 or WASP is not expressed, or not functional, and results in immune deficiency21,35,36.

Like TCR signalling, IL-2 receptor signalling is dependent on the spatial organisation of the plasma membrane and sensitive to the local lipid composition of the plasma membrane. IL-2 receptors locate to membrane rafts and their disruption by cholesterol sequestration impairs IL-2 signalling167,181,182. More importantly, binding to IL-2 induces a conformational change of the CD122 and ɣc IL-2 receptor subunits that promotes the association of the receptor complex with membrane rafts and the actin cytoskeleton and promotes downstream signalling72.

The importance of a more rigid and more ordered plasma membrane at the immune synapse during T cell activation was noticed when cholesterol was depleted with MβCD or replaced with 7-ketocholesterol (7KC). Cholesterol depletion with MβCD can impair T cell activation183 as well as T cell migration184. However, MβCD has also been criticised because cholesterol extraction with this agent can introduce artefacts due to depletion of intracellular Ca2+ stores. This might be the underlying reason for the observed impairment of T cell activation following MβCD treatment185. 7-Ketocholesterol contains a ketone residue protruding from the sterol group and thus prevents the tight packing of lipids and specifically reduces the order of the plasma membrane. Enriching T cells with 7- ketocholesterol, rather than cholesterol, results in fewer TCR microclusters at the cell surface, the T cells fail to form actin rings and have impaired IL-2 production186. These findings demonstrate that membrane composition and fluidity are functionally important for T cell activation186,187.

Despite refinement of methods, the importance of membrane rafts as signalling platforms at the immune synapse remains debated, since lipid anchors and other post-translational modifications of signalling proteins LCK (dual acylation), LAT (palmityl anchor) previously thought to target these proteins into raft domains188–191, do not cluster upon TCR activation192. Rather, the main function of these modifications is the targeting of the protein to the plasma membrane, but not to rafts specifically15,193,194.

While the role of membrane rafts remains controversial, membrane composition and membrane fluidity are factors that do affect signalling processes across membranes. Treatment of T cells with unsaturated fatty acids interferes with the formation of cytoskeletal structures at, and targeting of LAT to, the immune synapse and resulted in reduced expression of T cell activation markers195. The relationship of membrane order and the integrity of plasma membrane signalling could also be demonstrated in the olfactory epithelium. 2,4,6-trichloroanisole, a membrane soluble compound present in corked wine, drastically changes the perception of smell. Takeuchi et al.196 suggest that the signalling of ion channels initiating the olfactory stimulus is changed due to altered membrane fluidity. Research from our lab supports the notion (Alexander Kross unpublished data) that ion channel activity in the olfactory epithelium and membrane order are linked. Another receptor that could be shown to be affected by a decrease in plasma membrane fluidity is the EGF receptor. EGF receptor signalling was reduced by the non-steroidal anti-inflammatory drug Licofelone via changes of the membrane properties induced by the drug197. Treatment with the drug increased the presence of saturated fatty

acids and cholesterol – both of these changes contributed to lowering the membrane fluidity. It can therefore be concluded that receptor signalling and, thus, the function of cells are sensitive to the lipid composition of the plasma membrane.

Annexin A6 (AnxA6) is part of the annexin protein family – a highly conserved group of Ca2+ and lipid-binding proteins. They are found in almost all species from protists to vertebrates198 and are expressed in the majority of mammalian tissues. All members of this protein family have four peptide repeats (“annexin repeats”) consisting of five α- helices with loop structures that are able to bind Ca2+, which triggers the translocation and association of with cell membranes. Apart from their conserved annexin repeats, annexins are also comprised of short C- and longer N-terminal domains that differ greatly in length and amino acid sequence between annexins199.

AnxA6 is a 68 kDa cytosolic protein that stands out in the annexin family due to its size and structure. It contains eight rather than four annexin repeats which are connected by a flexible linker (between domain IV and V)200,201 (Figure 1-9) and an N-terminal tail of 21 amino acids with two putative phosphorylation sites (Ser-13, Tyr-30)202,203. The doubling of annexin domains is probably the result of an evolutionary event that fused duplicates of AnxA5 and AnxA10 genes198. Studies with protein crystals suggest that this unusual structure enables AnxA6 to interact with two different membranes simultaneously201. In humans, as well as mice, AnxA6 is expressed as one of two isoforms (AnxA6-1 or AnxA6- 2)204 with AnxA6-2 being predominantly expressed205. Due to alternative splicing, AnxA6- 1 lacks a 6-amino acid sequence, VAAEIL, at the start of repeat 7 of AnxA6-2. While some functions of AnxA6 seem to be splice form specific205, the importance of the two isoforms remains to be elucidated.

2 Contains text from Cornely et al. 2011275

Figure 1-8: AnxA6 expression is particularly high in immune cells. Example of AnxA6 expression in human cells and tissues from the transcriptome project206. Mean is average level of expression over all tissues, 3xM marks expression levels 3x above average.

AnxA6 is widely expressed in lung, liver, heart, skeletal muscle and adipose tissue207. A high-throughput screen from GeneAtlas project showed AnxA6 mRNA is found in virtually all tissues with notably high levels of transcription in immune cells, including T cells

(Figure 1-8). Immune histochemistry studies showed that AnxA6 expression was not confined to cells of a certain functional phenotype208. AnxA6 expression was absent in most epithelia, with the exception of breast, sweat and salivary glands. Interestingly, AnxA6 was expressed in non-lactating breast but not in lactating breast. Furthermore, AnxA6 was expressed in endocrine organs (e.g. Langerhans cells of pancreas, Leydig cells of the testes or thyroid gland)207–209. When Clark et al. found that the AnxA6 expression pattern of B cells in lymph node germinal centres and the AnxA6 expression in thymocytes in the thymus indicated a role for AnxA6 in the development of B and T cells, they were the first ones to link AnxA6 and the immune system208.

Figure 1-9: Structure of bovine AnxA6 with calcium (red spheres)210,211. Predicted phosphorylated residues (green) and linker region (yellow) are highlighted. A represents a “side- on” view, B top view. The arrows indicate the point of view for the respective other image.

AnxA6 is a multifunctional protein with very diverse array of tasks within the cell. Owing to its ability to bind cellular membranes, AnxA6 has been found to play a regulatory role in endocytosis, exocytosis, endosomal recycling, cholesterol homeostasis and stabilising the cortical F-actin cytoskeleton. Furthermore, AnxA6 also plays an important role in EGF receptor signalling212. More recently, AnxA6 was also found to be involved in plasma membrane repair in concert with other annexin proteins213–216. All described roles for AnxA6 take place in the cytosol or at intracellular membranes, but it has also been found in exosomes217 and on the outside of invasive breast carcinoma cells218. It has been suggested that extracellular AnxA6 may mediate interaction of cells with the extracellular matrix.

The membrane-binding ability of AnxA6 originates from its affinity for membrane phospholipids, such as PS, PI, PA219,220 as well as for PE and arachidonic acid221.

AnxA6 preferentially binds to phospholipids present in the cellular membranes after increases in cytosolic Ca2+ and possibly changes of pH222,223. AnxA6 was found on early and late endosomes219,224–227, phagosomes228 and the plasma membrane213,229. The distribution of membrane-bound AnxA6 therefore coincides with the cellular membranes that contain the highest amount of PS96. In vitro studies found that low Ca2+ concentrations were required for AnxA6 to bind PS and PE, and higher Ca2+ concentrations were needed for

AnxA6 to bind PA and PI221,230, furthermore, AnxA6 also shows some affinity to PIP2 in vitro231. The positively charged Ca2+ ions bound to the annexin repeats target the protein to the negatively charged phosphate head groups of PS, PE, PI and PIP2. The charge- induced binding of AnxA6 to PI and PIP2 might be affected by the comparatively large inositol residue of these phospholipids221. Acidic pH and cholesterol also regulate the targeting and trafficking of AnxA6212,223,225,227,232,233. Membrane-bound AnxA6 has been found to be associated with both membrane rafts (including caveolae)234,235 as well as non- raft domains (e.g. clathrin-coated pits)236. AnxA6 translocates to cholesterol-rich membrane domains, which contain receptors and signal transduction machinery, in a Ca2+- dependent manner202,237,238. Apart from targeting lipids at the plasma membrane, AnxA6 also binds several components of the cytoskeleton219: spectrin239, actinin240 and F-actin241– 243.

Initially, it was believed that AnxA6 was only able to bind phospholipid-containing membranes. However, eventually reports emerged that other annexins were able to bind cholesterol (e.g. AnxA2 and AnxA5202,244–246) imparting the idea that AnxA6 might be able to do so as well. Indeed, the Ca2+-dependent binding of cellular membranes by AnxA6 seems to be augmented by the presence of cholesterol in these membranes223,237,247,248. AnxA6 also binds cholesterol-enriched membranes even in the absence of Ca2+225. In fibroblasts AnxA6 is bound to membranes in cholesterol dependent fashion. Removing cholesterol from cellular membranes with increasing amounts of MβCD reduces AnxA6 binding to these membranes233. Since AnxA6 was found to attach to a lesser degree to cholesterol-rich membranes even under Ca2+-free conditions, this interaction of AnxA6 and cholesterol is likely to be partially Ca2+- independent227,232,247,249,250. In vitro studies

with recombinant AnxA6 found that, AnxA6 might bind cholesterol with an exposed tryptophane residue (Trp343) in the linker region223, but these findings have not yet been confirmed by studies in cells.

As outlined in Chapter 1.2.1 cholesterol can be synthesised de novo or acquired from extracellular sources through LDL uptake. LDL particles are endocytosed through the LDL receptor in a clathrin-dependent fashion. After loss of the clathrin exterior, endocytic vesicles lose fuse to form larger vesicles termed early endosomes. These acidic vesicles sort specific endocytosed molecules to the trans-Golgi network, and receive hydrolases in return, maturing into late endosomes and lysosomes, with increasing luminal acidity251. LDL particles deliver cholesterol esters as well as free cholesterol into the cell. Endosomal hydrolases turn esters into free cholesterol, and unesterified cholesterol is trafficked from the endo- and lysosomal compartment to the plasma membrane and the ER108. The LDL receptor is recycled and some cholesterol might be transported to the plasma membrane in recycling endosomes. But the efflux of cholesterol from late endosomes and transport to ER, Golgi, recycling endosome and plasma membrane is not well characterised. To transport cholesterol to the ER a route via the trans-Golgi network has been described248,252. From there, cholesterol could continue to the plasma membrane as well. But this route may not exist in all cell types253 and, instead, cholesterol can be transported to the plasma membrane on an actin-dependent route via Rab8a vesicles253.

In Nieman-Pick disease defects in Nieman-Pick protein C (NPC) 1 or 2 induce a cholesterol trafficking and storage disease leading to an accumulation of unesterified cholesterol and sphingolipids in the late endosomal and lysosomal compartment. Both NPC1 and 2 are likely to be part of the same pathway to promote cholesterol egress from late endosomes. NPC1 is a large protein with a sterol-sensing domain and several transmembrane domains which is found in the late endosomes. NPC2 is a cholesterol-binding, small, and soluble protein which potentially delivers cholesterol to the membrane-resident NPC1 in the endosomal lumen which acts as an acceptor for NPC2-bound cholesterol254–256. It has been suggested that LDL cholesterol is scavenged by NPC2 to prevent the formation of cholesterol crystals inside the late endosome or lysosome256. Pharmacologically, an NPC- like phenotype can be induced in cells with U18666A. U18666A is a cholesterol analogue that inhibits cholesterol transport beyond the late endosomal compartment.

AnxA6-binding to cholesterol-enriched cellular compartments was found when the cholesterol distribution within the cell was altered with LDL-loading of cells or U18666A treatment. Through these treatments cells contained more cholesterol in the late

endosomal compartment and the binding of AnxA6 to this intracellular compartment was increased225,227. Intriguingly, AnxA6 over-expressing cells retain cholesterol in late endosomes while plasma membrane and Golgi complex have reduced levels of cholesterol implying that the cholesterol transport to those compartments is diminished248,257. U18666A treated or NPC deficient cells present a similar phenotype248. Since AnxA6 co- immunoprecipitates with NPC1248, AnxA6 could interact with NPC1 directly. Since high levels AnxA6 can lead to a retention of cholesterol in late endosomes, AnxA6 might interfere with NPC1 and assume an inhibitory role in cholesterol export to the Golgi complex212,257. Furthermore, AnxA6 was found to stimulate LDL receptor endocytosis and recycling227. These findings assigned AnxA6 a potential role in the intracellular cholesterol transport and homeostasis202,212.

AnxA6 together with AnxA2 was shown in biochemical assays to associate with membrane rafts in smooth muscle cells237,258 and in rat brain238 in a Ca2+ dependent manner and both have been identified in isolated caveolae fractions from different cells and tissues219,234,259. Similar results were obtained for AnxA6 in CHO (Chinese hamster ovary) cells202. The results are based on the detergent resistant properties of lipid rafts – an approach that has recently been criticised (see Chapter 1.2.2). Owing to the cholesterol affinity, AnxA6 could easily bind to cholesterol-enriched domains directly. It is possible that preferential binding of cholesterol-rich domains is part of the mechanism by which AnxA6 regulates EGF receptor signalling. However, the role of membrane rafts in EGF receptor signalling is controversial260 with some publications finding active EGF receptor in rafts while others find rafts decrease EGF signalling261–263. To reduce EGF receptor signalling AnxA6 binds the Ras-GTPase-activating protein p120GAP and targets it to active H-Ras and EGF receptor. H-Ras is a signalling protein promoting the activating signals of the EGF receptor which is inactivated by p120GAP202,264. This reduces H-Ras mediated signalling. AnxA6 also mediates the targeting of PKCα to EGF receptor. PKCα phosphorylates EGF receptor at a specific site (T654) which inhibits EGF receptor signalling231,265. Interestingly, suppression of EGF receptor signalling by PKCα has been reported to require caveolin-1, ganglioside GM3 and CD82266. Caveolin-1, as well as GM3, is found in lipid rafts – suggesting that lipid rafts are important for EGF receptor inactivation. Thus, AnxA6 reduces EGF receptor signalling by binding or interacting with several members of this signalling cascade212.

Caveolae are preformed ‘cup-shaped’ structures in the plasma membrane that are considered to be a raft-like membrane domain267. They are enriched in cholesterol and glycosphingolipids and also contain a specific assembly of proteins called caveolins that support the structure of individual caveolae. AnxA6 has been found to associate with caveolae and has been suspected to be involved in caveolae mediated endocytosis234,237, since AnxA6 was found to co-purify with caveolae-associated proteins234,268. High levels of AnxA6, however, interfere with the transport of caveolin-1 in the cell. AnxA6 over- expression therefore corresponds with a low occurrence of caveolae at the plasma membrane and an accumulation of caveolin-1 in the Golgi complex248. This may be an indirect effect of the involvement of AnxA6 in cholesterol homeostasis. High levels of AnxA6 lead to cholesterol accumulation in late endosomes and a depletion of cholesterol in the plasma membrane. These alterations impair the cholesterol-dependent recruitment of cytoplasmic PLA2 to the Golgi. However, export from the Golgi requires the activity of cytoplasmic PLA2269. The AnxA6-induced cholesterol imbalance in the cell has drastic consequences for the export of caveolin-1 from the Golgi to the plasma membrane248,270. Ultimately, reduced amounts of caveolin-1 at the plasma membrane diminish the formation of caveolae248. Cholesterol is also a prospective factor to facilitate the membrane curvature271,272 of caveolae together with the structural protein caveolin-1 which binds to cholesterol. A decrease of the concentration of cholesterol in the plasma membrane can therefore prevent the formation of caveolae273,274. The reduced cholesterol in the plasma membrane is a result of the altered cholesterol distribution in cells expressing high levels of AnxA6257.

Due to its ability to interact with signalling proteins and cholesterol-enriched membranes, AnxA6 has been proposed to be able to associate with membrane rafts275. The ability of AnxA6 to bind components of the cortical cytoskeleton, as well as specialised membrane domains, may reflect a general role for AnxA6 to modulate the localization of cell surface receptors in membrane domains. Many transmembrane receptors on the cell surface were found to be surrounded by specialised membrane domains – the lipid composition of which can be closely linked to the signalling capacity of these receptors (see chapter 1.2.3.). AnxA6 interacts not only with a large number of signalling proteins (e.g. LCK, H- Ras, Raf-1, p120GAP, Raf-1, EGF receptor, PKCα), but also with cytoskeletal components like F-actin and spectrin202,212,260. So, together with its affinity for phospholipids, it is possible that AnxA6 may target and organise membrane domains to assist in forming scaffolds for intricate signalling complexes275, which are triggered by calcium fluxes and directed by transient annexin-membrane-actin interactions202,275–277.

The fact that AnxA6 was found to be associated with early219 and recycling endosomes278 adds to this theory. Activated receptors are usually endocytosed and, either recycled back to the cell surface after removal of their ligand, or are degraded when late endosomes fuse with lysosomes.

AnxA6 modulates endo- and exocytosis and is associated with endosomes in the early as well as in the late stages. The mechanism of how AnxA6 affects these processes is not completely understood, but the overexpression or silencing of AnxA6 seems to affect the balance of endo- and exocytosis.

In 1992, the group of Thomas Südhof found AnxA6 in combination with ATP and Ca2+ was required for clathrin-mediated budding activity in fibroblasts236. While AnxA6-depleted cytosol showed almost no budding activity in isolated membranes, a mix of AnxA6, ATP and Ca2+ reached 80-90% of optimal budding activity, which was achieved with complete cytosol. This suggests that these three components are necessary for budding, but not sufficient, as other components in the cytosol were also contributing to the process. Later, evidence mounted that endocytosis was not dependent on AnxA6, since A431 cells deficient in AnxA6 are capable of endocytosis via clathrin-coated pits279 and AnxA6 knock- out mice appear to have an inconspicuous phenotype280. While AnxA6 does not seem to be an indispensable component of clathrin-coated vesicle endocytosis, it can modulate endocytic activity212. AnxA6 as well AnxA2 are found on clathrin-coated vesicles and are able to bind to parts of the clathrin-dependent machinery, namely the µ2 subunit281,282, which is part of the AP2 adaptor complex required for clathrin vesicle formation. Furthermore, AnxA6 was found to bind dynamin282, a GTPase necessary for the scission of clathrin-coated vesicles from the plasma membrane, as well as clathrin-independent endocytosis283–285.

Changes in AnxA6 expression levels were found to affect endocytosis and vesicle budding as well as endosomal recycling. AnxA6 over-expression in CHO cells, together with LDL receptor over-expression, increases LDL receptor endocytosis and lysosomal targeting. However, the over-expression of AnxA6 in combination with transferrin receptor over- expression leads to only mild endocytosis increase and does not affect transferrin receptor recycling227. Interestingly, both receptors are endocytosed via clathrin-coated vesicles.

This shows that AnxA6 is not tied to the clathrin-mediated endocytic route but can modulate selectively endocytic processes and recycling pathways.

After endocytosis, vesicles fuse with early endosomes. Some membrane components and receptors are recycled and bud off to return to the plasma membrane as recycling endosomes. Endosomes with components that are not recycled are either transported to the Golgi complex or mature into late endosomes that eventually fuse with lysosomes for degradation286. AnxA6 has been found to associate with early as well as late endosomes219,224–226. Taken together with the involvement of AnxA6 in cholesterol transport inside the cell, AnxA6 possibly supports not just the endocytosis of certain receptors, but also regulates the export of cholesterol-rich vesicles from the late endosomal compartment to the Golgi complex and the fusion of the lysosomal and late endosomal compartments202.

A mechanism by which AnxA6 stimulates endocytosis and lysosomal targeting of LDL could be the remodelling of the actin-spectrin cytoskeleton to release budding vesicles from the plasma membrane and late endosomes, respectively. The cortical cytoskeleton forms a tight meshwork underneath the plasma membrane and has to be locally removed to allow large membrane invaginations and vesicle budding. It was found that the interaction between AnxA6 and spectrin leads to the recruitment of the protease Calpain I, which locally hydrolyses the spectrin-actin cytoskeleton permitting vesicle entry and budding287. Along these lines, there are other publications which support the notion that AnxA6 can mediate cytoskleleton rearrangements to allow vesicle budding236,239. Similarly, the fusion or budding of late endosomes to lysosomes requires temporary destabilisation of spectrin and might also involve AnxA6 in some cases. Investigating the trafficking of LDL revealed that AnxA6 was necessary for the progression of LDL from late endosomes to the lysosomal compartment227,288.

Early in the characterisation of AnxA6 function and localisation, AnxA6 was found to be strongly expressed in non-secreting breast ductal epithelia but not expressed at all in the lactating part of the tissue208,289. In vitro experiments also showed that AnxA6 counteracts vesicle fusion mediated by AnxA2 and AnxA7 suggesting that it plays a part in negatively regulating exocytosis290,291. More recently, AnxA6 was identified together with other annexins as being part of secretory lysosomes in cytotoxic T cells292 and exosomes of cancer cells217. Moreover, AnxA6 also seems to be able to influence the machinery of the

secretory pathway indirectly through its role in intracellular cholesterol homeostasis. In AnxA6 over-expressing cells, cholesterol transport from late endosomes to the Golgi complex and plasma membrane is reduced leading to high cholesterol levels in late endosomes and low levels in the Golgi complex and plasma membrane. t-SNARE (soluble NSF attachment protein receptor) proteins necessary for vesicle trafficking of the secretory pathway localise to cholesterol-rich membrane domains. The occurrence of t- SNAREs in the cholesterol-poor plasma membrane of cells with high levels of AnxA6 is significantly reduced and the secretion of fibronectin and TNFα from these cells was diminished257.

Another form of exocytosis is the production of lipid-enveloped virus by an infected cell. Here, AnxA6 also adversely affects the successful budding of viral particles. While AnxA6 (over)expression reduces the shedding of virus and renders the progeny less infectious, silencing the expression of AnxA6 increases the production of influenza virus293,294. Ma et al. argue that AnxA6, which was shown to bind the viral protein M2, prevents the release of virions by obstructing budding events at the plasma membrane, thus reducing viral infection293. However, Musiol et al. provide evidence that AnxA6 diminishes infectivity and virion production of influenza infected cells by reducing plasma membrane cholesterol294. The weak infectivity of virus particles from AnxA6 over-expressing cells correlated with the diminished availability of cholesterol in the plasma membrane of these cells. Reduced plasma membrane cholesterol leads to lower amounts of cholesterol in the virus envelope and impedes fusion with new cell membranes for re-infection294. This demonstrates that the role or involvement of AnxA6 in exocytosis and cholesterol-homeostasis are closely linked.

To date, no genetic disease has been reported in which AnxA6 is severely mutated or not expressed. However, abrogated or reduced expression of AnxA6 has been found in many cancers295. AnxA6 is not expressed in the human squamous carcinoma cell line279, and reduces growth of these cells when it is re-expressed296,297. Recently, it has been shown that the expression of AnxA6 is silenced in gastric cancer due to promoter methylation298. Furthermore, AnxA6 is reduced in malignant melanomas299 and expressed at very low levels in estrogen-receptor negative breast cancer264. In gastric, as well as breast cancer, cells, restoring the expression of AnxA6 inhibited the growth of these cancers264,298. While

AnxA6 is generally regarded as a tumour suppressor, it seems to play a more complex role in cancer progression, since it remains expressed in invasive carcinoma cell lines and promotes anchorage-independent growth of cancer cells lines218,300.

The growth-reducing effect of AnxA6 expression in cancer cells is due to the involvement of AnxA6 in the EGF receptor activated Ras/Raf/MAPK pathway. This pathway is elemental to cell differentiation and proliferation. EGF receptor activation induces receptor dimerization autophosphorylation of distinct tyrosine residues. In the course of downstream signalling, H-Ras is activated and in turn activates Raf. In cancer cells, the small GTPase Ras is often rendered constitutively active due to mutations in the ras gene or other genes involved in Ras regulation and activation. As could be shown in breast cancer cells, AnxA6 interacts with the GTPase-activating protein p120GAP and thus induces the inactivation of Ras. AnxA6 is already bound to p120GAP when EGF receptor is activated. The calcium influx into the cytosol, triggered by EGF receptor signalling, targets AnxA6-p120GAP to the plasma membrane249,301,302. There, p120GAP and H-Ras interact and H-Ras signalling is terminated303. Alternatively, AnxA6 can bind PKCα304 while both proteins are located in the cytosol to target it to the plasma membrane after Ca2+ increase. PKCα inactivates EGF receptor by phosphorylating Thr654, which inhibits the activating tyrosine-phosphorylation of EGF receptor265,301.

Despite initial cues that AnxA6 might be involved in immunity, the research on the role of AnxA6 in the immune system remained scarce. Other annexin proteins have already been established to play a role in immunity. AnxA1, AnxA2 and AnxA5 have been reported to participate in the immune response and inflammatory processes. Even though they also were found to have intracellular functions relating to vesicle traffic, their immune-related functions are mainly extracellular. AnxA2 and AnxA5 are both present in serum and support innate immune functions. They have been identified to provide binding sites for the complement factor C1q on apoptotic cells and thus also contribute to self- tolerance305. Unlike viable cells, apoptotic cells increasingly expose the annexin ligand PS on the outside since they are no longer maintaining the asymmetric distribution of phospholipids across the cytosolic and external leaflet of the plasma membrane. Patients suffering from autoimmune diseases, like systemic lupus erythematosus, produce autoantibodies against AnxA2, AnxA5 and C1q306–308. These antibodies are likely to neutralise the functional binding sites for these proteins on the cell surface and prevent clearance of apoptotic cells. This, in turn, interferes with the protective function of these annexins. Furthermore,

AnxA2 might play a role in regulating the infectivity of human immunodeficiency virus309,310.

The role of anti-inflammatory protein AnxA1 has been widely investigated. Similarly to AnxA2 and AnxA5, the presentation of AnxA1 on the exterior of a dying cell prevents the reaction of immune cells to the self-peptides set free by this process. AnxA1 externalised on apoptotic cells averts the induction of inflammatory DCs and the generation of antigen- specific cytotoxic T cells311. AnxA1 is also externalised on activated neutrophils and interrupts their interaction with endothelial cells. In autocrine stimulations, AnxA1 counteracts cues that trigger the extravasation of neutrophils and moderates the influx of pro-inflammatory leukocytes at sites of injury or infection220. AnxA1 aids in clearing neutrophils by macrophages312. The role of AnxA1 in T cells has been less well defined. AnxA1 is only expressed in 50% of the circulating cell population, but expressed in all extravasated T lymphocytes313. Knocking-out AnxA1 in mice resulted in mixed results of increased314 or decreased T cell activation with a higher level of Th2 lineage T cells315.

A study exploring the localisation of AnxA6 found increased expression of AnxA6 in mature T cells, i.e. circulating T cells and in the medulla of the thymus, but no expression in the cortical cells of the thymus where immature thymocytes reside. These findings implied a role for this protein in T cell development208. Supporting the notion that AnxA6 is important for T cell immunity, AnxA6 was found to associate independently of Ca2+ with the membrane-bound tyrosine kinase LCK203 which is not only one of the key signalling proteins at the TCR, but it also binds to the CD122 subunit of the IL-2 receptor. However, AnxA6 knock-out (AnxA6-/-) mice have a healthy phenotype with no obvious immune deficiency in the absence of an immune challenge280. When the AnxA6-/- mouse strain was originally created, the abundance of several different immune cells was investigated280: The level of CD4+ and CD8+ T cells in the spleen and the thymus of AnxA6-/- and wild type mice were assessed by flow cytometry and no significant differences were found. Proportions of mature (CD3+) and immature (CD3-) T cells in the thymus were not significantly different between wild type and AnxA6-/- mice. Furthermore, B cells from the bone marrow were tested for the abundance of progenitor (CD45/B220+IgM-) and immature (CD45/B220-IgM+) B cells as well as the level of mature (IgM+IgD+) B cells in spleen homogenates, without showing any evidence for differences between wild type and AnxA6-/- mice. Likewise, the levels of myeloid, granulocytic, and monocytic cells in the bone marrow revealed no differences between wild type and AnxA6-/- mice. At the same time, other annexin proteins, like AnxA1, AnxA2 and AnxA5, were not notably upregulated

in spleen, heart and liver lysates in the absence of AnxA6280. This implies that these proteins that are similar in their ability to bind cellular membranes and participate in vesicle trafficking probably do not take over the role of AnxA6 in knock-out mice. A more recently developed knock-out strain (MGI database ID: 4432672) from the Mouse Genetics Project was found to be more susceptible to bacterial infection during a Citrobacter rodentium challenge as described on the Mouse Genetics Project database316,317. C. rodentium is a murine model pathogen for human enteropathogenic Escherichia coli. It primarily colonizes the lumen and mucosal surface of the large intestine (colon and cecum) and causes crypt hyperplasia and mucosal inflammation. The lack of an apparent phenotype in untreated mice is likely to be an indication that AnxA6 is a modulator of cellular functions that only rises to importance in stressed but not in healthy mice.

The goal of this study was to identify a role for AnxA6 in T cells and to establish how AnxA6 is involved in this process.

The first aim of this study was to investigate the effect of knocking-out AnxA6 to an immune response in vivo (Chapter 3). By applying a standard model to induce contact hypersensitivity (CHS), the immune system of AnxA6-/- mice was challenged and the quality and quantity of this immune response was characterised by measuring the levels of CD4+ and CD8+ effector T cells and their levels of cell division.

The second aim was to explore which T cell functions contributing to T cell activation and proliferation are affected by a lack of AnxA6 (Chapter 4). Does the genetic deletion of AnxA6 affect T cell migration, TCR signalling or IL-2 receptor signalling? An AnxA6 knock- down Jurkat cell line, as well as primary murine AnxA6-/- T cells, were subjected to migration assays in vitro. To address the involvement of AnxA6 in TCR signalling, Jurkat and primary murine T cells were activated in vitro and the levels of phosphorylated TCR signalling protein were assessed by Western blot. Moreover, the IL-2 production and secretion of murine primary T cells were measured by qPCR and an IL-2 capturing assay, respectively. The involvement of IL-2 signalling was investigated by exposing activated and cultured T cells in vitro to a dose of IL-2 and measuring the pSTAT5 response by flow cytometry. Since the activity of AnxA6 is often tied to activation by Ca2+, it was investigated how AnxA6 can affect a signalling event like IL-2 receptor ligation that is not known to

operate via increases of cytosolic Ca2+. Furthermore, the interaction of AnxA6 and F-actin during T cell activation was investigated by microscopy.

Lastly, it was investigated how AnxA6 knock-out affected membrane fluidity and lipid composition of T cells. Due to the involvement of AnxA6 in the cholesterol homeostasis of cells, it was taken into account that AnxA6 knock-out could induce changes to the cellular lipid composition and the fluidity of the plasma membrane (Chapter 5). These parameters are closely associated with the localisation and signalling efficiency of membrane proteins. To date, the role of AnxA6 in lipid composition of cellular membranes has only been investigated in cell lines with differential AnxA6 expression. But how are the membranes of T cells affected in the context of a whole organism that does not express AnxA6? The membrane fluidity of activated T cells was investigated by 2-photon microscopy of Laurdan-stained primary T cells and mass spectrometry techniques were employed to investigate the lipid composition of primary AnxA6-/- T cells.

All mice were bred at the Australian Bio Resources facility in Moss Vale, Australia, and boarded in the Lowy animal facility at the University of NSW, Sydney, under pathogen-free conditions and with a 12-h light/12-h dark cycle. Mice were fed a standard chow diet ad libitum and the drinking water sterilised by acidification. Mice were maintained for one week after receipt to acclimatise before any procedure was conducted and all procedures were carried out in accordance with the Australian code for the care and use of animals for scientific purposes, under the UNSW Animal Care and Ethics Committee (ACEC) approval numbers 10/61B and 11/13B.

The CHS reaction was induced with dinitrofluorobenzene (DNFB, Sigma, D1529). It was applied to the shaved abdomen and ear pinna of adult age-matched male mice in different concentrations through the course of the assay. 25 μl of a solution of olive oil:acetone (4:1) with 0.5% DNFB and 1 μg/ml endotoxin-free bovine serum albumin (BSA, Sigma, A-8426) were painted on the shaved abdomen of AnnexinA6 knock-out and wild type mice. After 24 hours the sensitisation treatment was repeated with a solution of 0.25% DNFB and 1 μg/ml BSA in the same vehicle. On both days the control group received the olive oil and acetone solution only. 5 days following initial sensitisation the ear pinna of the sensitised and control groups were painted with 20 μl of 0.2% DNFB and 0.8 μg/ml BSA solution (left ear) and with the olive oil and acetone solution only (right ear) for comparison. After

24 hours all mice were sacrificed by CO2 inhalation. The ear pinna swelling was measured with an engineer's micrometre. The inguinal (draining) lymph nodes were removed, weighed, and used for flow-cytometry analysis.

T cells were isolated via negative selection from the spleens of adult mice. After sacrificing the mouse, the spleen was removed and placed in a dish filled with T cell isolation buffer, a PBS (phosphate-buffered saline) solution containing 0.5% BSA, and kept on ice for transport. Shortly thereafter, in a tissue culture hood, the spleen was passed first through a 70 µm nylon mesh cell strainer (BD Biosciences, 352350) followed by a 40 µm cell strainer (BD Biosciences, 352340) to create single cell suspension. The cell suspension was then labelled with the Pan T cell isolation Kit II (Miltenyi Biotec, 130-095-130) according to the manufacturer’s instructions and depleted of all non-T cells in an autoMACS Pro Separator (Miltentyi Biotec).

Briefly, after one wash step with T cell isolation buffer, the cell concentration was determined and the required amount of cells resuspended in buffer containing an antibody cocktail, labelling all non-T cell species in the suspension with biotin-conjugated antibodies at 4℃ for 10 minutes. Subsequently, the labelled suspension was incubated with magnetic beads binding to biotin for 15 minutes at 4℃ and following another wash step the cell suspension was separated in the autoMACS Pro Separator by applying the ‘Deplete’ program.

Each pair of inguinal lymph nodes was placed on a pre-wetted 70 μm nylon cell strainer and gently pushed through the mesh with the rubber plunger of a 1 ml syringe into PBS with 1% BSA. The strainer was rinsed with 9 ml of buffer and the cell homogenate centrifuged at 230×g before resuspending in 2 ml of PBS buffer with BSA. Finally, cell densities of every cell suspension were measured with a cell counter (CasyCounter).

Whole cell populations from homogenised lymph nodes were stained with fluorochrome- coupled antibodies (Table 2-1) recognising murine CD3ε, CD4, CD8, CD62L, CD44 and proliferating cell nuclear antigen (PCNA). 106 cells from each mouse were analysed only for their surface markers while another 106 cells were fixed, permeabilised and stained for PCNA, as well as CD3ε, CD4 and CD8, to differentiate the PCNA expression of various

subsets. Relevant isotype controls or fluorescence minus one (FMO) controls were used for all experiments.

A single cell suspension was blocked with 50 µl of diluted Fc-Block solution (FcR blocking Reagent, mouse, Miltenyi, 130-092-575). After 10 minutes at 4°C, 50 µl of FACS (fluorescence assisted flow cytometry) buffer (1% bovine serum albumin (BSA) in PBS with 0.1% sodium azide) containing all antibodies for surface markers were added and incubated for 30 minutes at 4°C in the dark. After this staining step, the cells were washed twice with FACS buffer. The cells that received only surface staining were resuspended in 300 µl of FACS buffer and kept on ice in the dark until analysis. Samples which were also probed for levels of PCNA were resuspended in PBS and fixed with 2% paraformaldehyde (PFA). Afterwards cells were washed with cold PBS, permeabilised with cold methanol for 10 minutes at -20°C, washed in a PBS solution containing 0.1% Triton X-100 and resuspended in 100 μl of PBS with 0.1% BSA and 0.125 μg PE-conjugated anti-PCNA antibody. This reaction was incubated in the dark for 30 minutes at room temperature and washed twice with FACS buffer before being resuspended in 300 μl of FACS buffer for analysis.

Routinely, at least 20,000 events were acquired per sample. The flow-cytometry data was acquired with BD FACSVerse and analysed with FlowJo v10 software.

Table 2-1: Antibody and protein conjugates used for flow cytometry.

Fluorophor/ Antibody specificity Host animal clone Company, Order No Modification murine CD3ε hamster 145-2C11 PE-Cy7 BD Biosciences, 552774 murine CD4 rat RM4-5 V500 BD Biosciences, 560782 murine CD8a rat 53-6.7 Pacific Blue BD Biosciences, 558106 murine CD44 rat IM7 APC BD Biosciences, 559250 murine CD62L rat MEL-14 FITC BD Biosciences, 553150 PCNA rat 3F81 PE Abcam, ab93576 phospho-STAT5 rabbit C11C5 - Cell Signaling, 9359 rabbit IgG goat AlexaFluor 488 Invitrogen, A11070 murine CD122 rat 5H4 biotinylated eBioscience, 13-0031 Streptavidin- - - Invitrogen, S11223 AlexaFluor 488 rat IgG biotinylated eBioscience, Rat-IgG2a V500 BD 560786 Rat-IgG2a Pacific Blue BD 558109

All tissue culture practices were performed with sterile techniques in Class II biological safety cabinets.

Jurkat E6.1 cells were maintained in complete RPMI (10% foetal bovine serum (FBS) and 2 mM glutamine). Transformed Jurkat cells (AnxA6KD and CTRL) were maintained in the same medium with puromycine (Sigma, P8833) added at 1 µg/ml.

Primary murine T cells were maintained in primary T cell medium: RPMI supplemented with 10% FCS, 2 mM glutamine, 1 mM sodium pyruvate (Life Technologies, 11360-070), non-essential amino acids (Life Technologies, 11140050), 55 mM β-mercaptoethanol, penicillin-streptomycin.

Coverslips were rinsed at least 5 times with deionised water until they were completely dust-free. All water was aspirated and a 70 µl drop of anti-CD3 and anti-CD28 in PBS solution (10 µg/ml for each antibody) was carefully placed onto the centre of the coverslip and incubated over night at 4℃ or for 2 h at 37℃. The coated coverslip was then rinsed 3 times with PBS before T cells were placed on to it for activation.

Tissue culture plates were likewise coated with anti-CD3 and anti-CD28 in PBS solution (10 µg/ml for each antibody), incubated over night at 4℃ or for 2 h at 37℃, and rinsed 3 times with PBS before use.

106 Streptavidin-coated beads of 6-8 µm in diameter (Spherotec, SVP-60-5) were washed once by centrifuging the sample at 3000×g for 5 minutes and resuspending the bead pellet in PBS before centrifuging again and removing the supernatant. The beads were then incubated over night at 4℃ with 10 µg/ml of biotin-conjugated anti-CD3ε and anti-CD28 antibodies on a rotating stand. The following day the sample was washed 3 times with PBS before being incubated for 1 hour in blocking buffer consisting of 0.3% BSA in PBS. After three more PBS washes the pellet was resuspended to a concentration of 50- 100×106 beads/ml and stored refrigerated until further use.

Table 2-2: Activating antibodies for coating surfaces.

Antibody specificity clone Modification Company, Order No human CD3 OKT3 none eBioscience, 16-0037 human CD28 CD28.2 none eBioscience, 16-0289 human CD3 OKT3 biotinylated eBioscience, 13-0037 human CD28 CD28.2 biotinylated eBioscience, 13-0289 murine CD3 145-2C11 none eBioscience, 16-0031 murine CD28 37.51 none eBioscience, 16-0281 murine CD3 145-2C11 biotinylated eBioscience, 13-0031 murine CD28 37.51 biotinylated eBioscience, 13-0281

Freshly isolated murine T cells were activated on antibody-coated beads for 2 or

12 minutes in a tissue culture incubator (37℃, 5% CO2). 3-3.5×106 cells were carefully mixed with 2×106 beads in a graduated 1.5 ml micro-centrifuge tube. The non-activated control sample did not contain any beads. To stop activation all samples were cooled on ice, pelleted in a refrigerated centrifuge (5 min, 200×g) and washed with ice-cold PBS.

106 of logarithmically growing Jurkat T cells were resuspended in 600 µl of complete RPMI and incubated on antibody-coated 96-well plates (2.5×105 cells/well).

The cell pellet was lysed for 15 minutes on ice in 40 µl of radioimmunoprecipitation assay (RIPA) lysis buffer (150 mM NaCl, 50 mM Tris-hydrochloric acid (HCl) pH 7.4, 1% (v/v) NP-40, 0.25% (w/v) Na-deoxycholate) containing phosphatase (PhosSTOP, Roche Applied Science, 4906845001) and protease inhibitors (complete, Roche Applied Science, 04693116001). Cellular DNA was sheared by passing the lysate through a 29 gauge syringe 10 times. Any debris was removed by centrifuging for 15 minutes at maximum speed with the supernatant collected for analysis. Protein concentration of the supernatant was determined by a biuret assay using bicinchoninic acid (BCA) for colorimetric detection (Pierce, 23227), prepared according to the manufacturer’s instructions. Following protein quantification, cell lysates were diluted to equal protein concentration with RIPA lysis buffer.

The samples were then supplemented with Sample Reducing & Denaturing Buffer (10% (v/v) glycerol, 50 mM Tris-HCl, 2 mM ethylenediaminetetraacetic (EDTA), 2% (w/v) sodium dodecyl sulphate (SDS), 50 mM dithiothreitol (DTT), 0.002% (w/v) bromophenol blue) and heated for 5 minutes at 99℃.

600-700 µg of protein were loaded onto a 10% Bis-Tris (bis(2-hydroxyethyl)imino- tris(hydroxymethyl)methane) gel (Life Technologies, NP0315BOX) and electrophoresed for 50 minutes at 200 V in MOPS (3-(N-morpholino)propanesulfonic acid) buffer (Life Technologies, NP0001). The protein separated in the gel was blotted with the wet transfer method onto a PVDF (polyvinylidene difluoride) membrane (Millipore, IPVH00010) for 1.5 hours at 40 V.

Afterwards the membrane was blocked with a 5% BSA-TBS (Tris-buffered saline) solution and decorated first with primary antibodies (see Table 2-3) before incubating the membrane with a HRP-conjugated secondary antibody directed against the host animal (anti-rabbit, anti-mouse or anti-goat of the primary antibody). The membrane was then washed several times in TBS. The chemoluminescence reaction was started by incubating the washed membrane for 1 minute in enhanced chemoluminescence (ECL) solution (Pierce, 32209). Immediately afterwards the signal was detected in an ImageQuant LAS 4000 mini.

The detected signals were quantified using the gel analysis plug-in of ImageJ (NIH) and normalised to the intensity of the β-Actin signal of each lane.

Table 2-3: Antibodies used for detecting protein bands on western blot membranes

Antibody specificity Host animal Clone Company, Order No phospho-TCR zeta [Y142 ] rabbit EP265(2)Y Epitomics, 2280-1 phospho-ERK 1/2 [T202/Y204] rabbit n/a Cell Signaling, 9101 ERK 1/2 rabbit n/a Cell Signaling. 9102 phospho-LAT [Y191 ] rabbit polyclonal Invitrogen, 44228 29/ZAP70 ZAP70 mouse (IgG2a) BD Biosciences, 610239 Kinase phosphor-ZAP70 [Y493] rabbit polyclonal Cell Signaling, 2704 STAT 5 rabbit Cell Signaling phospho-STAT5 rabbit C11C5 Cell Signaling, 9359 phospho-PLCɣ1 [Y783] rabbit polyclonal Cell Signaling, 2821 PLCɣ1 mouse (IgG1) 10/PLCgamma BD Biosciences, 610028 AnxA6 goat N-19 Santa Cruz, sc1931 β-Actin rabbit polyclonal Abcam, ab1801 rabbit IgG HRP-conjugated goat n/a BioRad, 170-6515 mouse IgG HRP-conjugated goat BioRad, 172-1011 goat IgG HRP-conjugated rabbit polyclonal Abcam, ab6741

It should be noted here that in order to detect levels of phosphorylated, as well as total protein levels of protein membranes, were decorated first with antibodies specific to phosphorylated sites of PLCɣ1, ZAP70, LAT and TCR, the superficial protein was removed with Restore PLUS Western Blot Stripping Buffer (Pierce, 46430), blocked for 60 minutes with 5% BSA-TBS solution, and redecorated.

Jurkat cells were transfected with the NEON system from Invitrogen. Jurkat cells were grown to a density of 5-9×105 cells/ml and 2×105 cells per transfection washed once with 10 ml of PBS before resuspending in 10 µl Buffer R and mixed with 1 µg of endotoxin-free plasmid DNA. The sample was pulsed twice with a 30 ms pulse width at 1150 V. Immediately after pulsing, the sample was transferred into a well plate with well equilibrated RPMI 1640 medium (Gibco, 21870-076), supplemented with 10% FBS and 2 mM glutamine.

After 16-20 hours transfected cells were used for imaging.

A lentiviral vector containing AnxA6 knock-down shRNA and a puromycin selection marker, and a generic control shRNA with the same selection were purchased from Sigma (Mission Lentiviral Transduction particles). Jurkat cells at a density of 5×105 cells/ml per well were transduced with a multiplicity of Infection (MOI) of 1. 24 hours post transduction the media was supplemented with 0.5 µg/ml puromycine. The antibiotic containing medium was exchanged every 2 days and the surviving cells were expanded and stored as frozen stocks.

1.5-2 ×106 T cells in 150 µl of complete RPMI or primary T cell medium were stained with 25 µM Laurdan (Life Technologies, D250) for 30 minutes in a tissue culture incubator

(37℃, 5% CO2) prior to activation for 12 minutes at 37℃ on anti-CD3 and anti-CD28 coated beads (bead to cell ratio 1:2). Afterwards, cells were fixed with a 4% PFA solution

(ProSciTech, C004) for 20 minutes in a tissue culture incubator. Fixed samples were washed 3 times in PBS with 5% FBS, resuspended in 200 µl of PBS/FBS solution and spun on a glass coverslip with a Shandon Cytospin centrifuge (Thermo Scientific) (800 rpm for 2 minutes). After this coverslips were stored in PBS for a short time till they could be mounted with DAKO fluorescence mounting medium (DAKO, S3023) on coverslides for imaging.

Images were acquired with a TCS SP5 laser scanning inverted confocal microscope with

Leica LAS software (Leica), ×63 oil-immersion objective, NA = 1.4, and imaged at room temperature. Laurdan fluorescence was excited at 800 nm with a multiphoton laser system (Mai-Tai mode-locked titanium-sapphire (Ti:Sapphire) laser, Spectra-Physics). Images were acquired simultaneously. The photo multiplier tubes collected signal in the ranges of 400–460 nm and 470–530 nm.

The corrected generalised polarisation (GP) was calculated for each pixel using the

Laurdan intensity images (I400-460 and I470-530). I400-460 and I470-530 represent the intensity of the images acquired in the range of 400-460 nm and 470-530 nm, respectively. The G- factor (G) for the acquisition settings obtained for 0.5 µM Laurdan in dimethylsulfoxide (DMSO) was used as a correction factor.

GP was defined as

where

GPexp is the experimentally determined GP value of the 0.5 µM Laurdan solution and

Gtheo= 0.207, a known value of Laurdan in DMSO at 22˚C. The GP was calculated for each pixel using the Laurdan plugin of the software ImageJ 140. The GP values were calculated for the activation sites adjacent to the bead or free plasma membrane (see inset box of Figure 5-2 A and B, Figure 5-3 A and B, Figure 5-4 A and B).

4-8×105 T cells were activated in a small volume of complete RPMI or primary T cell medium (150 µl) for the indicated amount of time on antibody-coated coverslips at 37℃ in

a tissue culture incubator. Cells were fixed by adding 50 µl of 16% PFA stock solution to the medium to yield 4% final PFA. Cells were fixed for 10-15 minutes at room temperature, the PFA solution was removed and the sample was carefully washed twice with PBS, with every wash step consisting of soaking the sample in PBS for 5 minutes. Afterwards the cells were permeabilised for 5 minutes at room temperature with 0.1% Triton X-100 in PBS. After washing the samples three more times in PBS, they were incubated with 1:200 Phalloidin AlexaFluor647 (Life Technologies, A22287) or 1:300 Phalloidin AlexaFluor555 (Life Technologies) in PBS for 1 hour at room temperature, before washing again 3 times in PBS. Samples were stored in PBS and imaged within 24 hours.

Samples were imaged on a total internal reflection (TIRF) microscope (Elyra, Zeiss) following the procedure and analysis method of Williamson et al318. In brief, single molecule images were generated from a 15000 frame time-series with an algorithm identifying the centre of the light signal. Every detected event was exported to a table containing information on the x-y coordinates of each molecule. A point pattern of the molecular distribution was generated from these data. In the reconstituted image, square regions of 1.5×1.5 µm containing more than 290 molecules were selected for cluster analysis. Ripley’s K-function (Eq. 3) was then calculated using the SpPack319 Excel plugin according to following equation.

For i≠j, ( ) ∑ ∑ ( ) where if ,

, where A is the area of the analysed region (1.5×1.5 µm), n is the number of points, r is the spatial scale (radius) for the calculation of the K-function and δij is the distance between points i and j. The equation provides a count of the number of molecules surrounded by concentric circles centred on each molecule, which is then normalised to the average molecular density of the entire region. The K-function was linearised to generate the L- function according to:

( ) √ ( )⁄

The function L(r) equals r (i.e. L(r)=r) if the spatial distributions of molecules is completely random. Therefore, values ≤0 for L(r)-r indicate random distribution. Plotting L(r)-r against r reveals at which spatial scales the distribution of molecules is clustered or random. Higher values for L(r)-r indicate a more clustered distribution. Points at the edge

of the 1.5×1.5 µm distribution region were weighted to compensate for edge-related effects. Furthermore, values of L(r) generated for each point at a value of r=50 nm (L(50)) were calculated with a MATLAB (The Mathworks Inc., Natick, MA) script (Clustermappingv13.m). The script generates two-dimensional cluster heat maps of all analysed regions, which are then thresholded to produce binary cluster maps. Areas of the interpolated map with L(50)>80 were defined as clusters.

Excised spleens were placed in a dish filled with T cell isolation buffer, a PBS solution supplemented with 0.5% BSA, and kept on ice for transport to the tissue culture hood. To create a single cell suspension, the spleen was passed first through a 70 µm nylon mesh cell strainer and afterwards through a 40 µm cell strainer. The cell suspension was washed once in T cell isolation buffer and after assessing the cell number splenocytes were resuspended in primary T cell medium to a concentration of 2×106 cells/ml.

IL-2 secretion from mouse T cells was assessed with the IL-2 secretion assay kit from Miltenyi (130-090-491) according to the manufacturer’s instructions. Briefly, the cell suspension was activated by incubating for 3.5 h at 37℃/5% CO2 in an antibody-coated well plate, while the not activated control was incubated in an anti-CD28 coated well plate. Afterwards, the cells were washed with pre-warmed primary T cell medium and labelled with IL-2 catch reagent. The cells were then washed again to remove any surplus catch reagent and incubated with pre-warmed primary T cell medium (10 ml media for activated cells, 1 ml media for control cells) for 30 min at 37℃/5% CO2 with frequent mixing. Cells were then resuspended in 100 µl of T cell isolation buffer and incubated for 30 minutes with IL-2 detection, anti-CD4 and anti-CD8 antibodies (for antibody details see Table 2-1) on ice. After washing with a large volume of T cell isolation buffer, the abundance of IL-2+ CD4+ and IL-2+ CD8+ T cells was assessed by flow cytometry on a BD Biosciences FACSVerse flow cytometer. For this, a minimum of 100,000 cells were acquired. The data were analysed with FlowJo v10 (TreeStar). The IL-2+CD4+ cells were expressed as a percentage of all CD4+ cells. IL-2+CD8+ cells were analysed in the same way.

T cells were isolated from spleen, activated for 24 hours on antibody-coated surfaces, and cultured for 2 days in the presence of 5 nM IL-2 in primary T cell culture medium.

The T cells were then washed once in PBS, resuspended in 1 ml cold 0.1 M glycine buffer (pH 4), and incubated on ice for 2 minutes to remove any IL-2 bound to the outside of the cell. The solution was neutralised by adding a large excess (minimum of 10× volume) of primary T cell medium followed by another wash step with medium. The cell concentration was determined and adjusted to 2×105 cells/90 µl. The cells were left to recover in culture medium for 40 minutes at 37℃. 2×105 cells were then added to a round- bottom well plate with different amounts of murine IL-2 (PeproTech, 212-12B) yielding final concentrations from 1 nM – 10 fM. After 10 min at 37℃, IL-2 stimulation was stopped by a short incubation on ice. The medium was removed following 2 min centrifugation at 400×g, 4℃, and the cells fixed with 1.6% PFA in PBS for 10 minutes on ice. The fixed cells were permeabilised for 20 minutes on ice using ice-cold 90% methanol and washed two times with FACS buffer 2 (4% BSA in PBS with 0.2% NaN3). The samples were then incubated for 30 minutes with anti-phospho-STAT5 (1:100, for more details see Table 2-1). After two washes with FACS buffer the cells were incubated for 30 minutes in the dark with anti-rabbit-AlexaFluor488 (1:200), anti-CD4-V500 and anti-CD8a-PacificBlue. Excess fluorescent antibodies were removed with two wash steps after which the cells were resuspended in 250 µl FACS buffer 2 and transferred into FACS tubes. All incubation steps were performed at room temperature in FACS buffer 2.

T cells were isolated from spleen, activated for 24 hours on antibody-coated surfaces and cultured for 2 days in the presence of 100 U/ml IL-2 in primary T cell culture medium.

Following incubation, the T cells were washed 3 times in 10 ml RPMI, resuspended to 2×106 cells/ml in primary T cell culture medium without IL-2, and incubated for 8 hours at 37℃. For the assay ~5×106 cells were removed from culture and washed with 10 ml FACS buffer 3 (2% BSA in PBS, azide-free). The cells were again resuspended to 2×106 cells/ml in FACS buffer 3. 2×105 cells for each sample were incubated for 30 minutes with 5 µg/ml of biotinylated anti-murine CD122 antibody or a biotinylated rat IgG isotype control (Table 2-1) on ice. Afterwards they were washed 3 times with 1 ml of ice cold FACS buffer

3. The cells were then resuspended in 100 µl of primary T cell culture medium. Cell samples were kept on ice or incubated for 6 minutes at 37℃ with and without 1 nM IL-2 before being returned to the ice. After washing the cell suspension once with 1 ml of FACS buffer 3 it was resuspended in a solution containing anti-CD4-V500 (1:100), anti-CD8- PacificBlue (1:100) and streptavidin-AlexaFluor488 (200 µg/ml). The cells were incubated in this solution for 30 minutes on ice before being washed 2 times in FACS buffer 3. They were then fixed in 2% PFA for 10 minutes at room temperature, washed to remove the fixative, and then kept in FACS buffer 3 at 4℃ until flow cytometry analysis.

For qPCR analysis of IL-2 mRNA levels, isolated T cells were activated for 4 hours on antibody-coated tissue culture plates in a tissue culture incubator at 37℃.

Activated and non-activated T cells were washed in cold PBS prior to lysis. RNA was isolated using the RNeasy Mini Kit (Qiagen, 74104) and remaining DNA removed with DNA-free (Ambion, AM1906). RNA was then reverse transcribed with Superscript III First Strand Synthesis (Life Technologies, 18080-051). qPCR was then performed using the SensiMixTM SYBR kit (Quantace, QT605-02) in a Rotorgene 3000 thermocycler (Corbett Life Science) with primers specific for murine IL-2, G6PDX (murine glucose-6-phosphate dehydrogenase) and murine GAPDH (glyceraldehyde-3-phosphate dehydrogenase). The initial denaturisation stage lasted 10 minutes at 95℃, l ann al ng p k plac a 6 ℃ f 6 seconds and alternated with 15 seconds f d na u a n a 95℃ f 4 cycles.

The primer sequences were sourced from the PrimerBank online database 320. For a complete list of primers and their sequences refer to the appendix.

Individual phospholipid internal standards (Table 2-4) were reconstituted in methanol to 1 mg/ml and mixed to yield a 100 µM phospholipid standard stock in methanol/chloroform (2:1 v/v). This stock solution was aliquoted and stored at -80℃ until use. D-6 cholesterol (2,2,3,4,4,6-D6, 97-98%) was reconstituted in chloroform at a concentration of 300 µM, and stored in aliquots at -80℃ until use.

Table 2-4: Lipids used in internal standard mix.

Lipid name Abbreviation Company 1,2-dinonadecanoyl-sn-glycerol-3- PC19:0/19:0 Avanti Polar Lipids Inc. phosphatidylcholine 1,2-diheptadecanoyl-sn-glycerol-3- PE17:0/17:0 Avanti Polar Lipids Inc. phosphatidylethanolamine 1,2- diheptadecanoyl-sn-glycerol-3- PS17:0/17:0 Avanti Polar Lipids Inc. phosphatidylserine 1,2-diheptadecanoyl-sn-glycerol-3- PG17:0/17:0 Avanti Polar Lipids Inc. phosphatidylglycerol 1,2-diheptadecanoyl-sn-glycerol-3-phosphate PA17:0/17:0 Avanti Polar Lipids Inc. 2,2,3,4,4,6-D6 cholesterol Novachem

A suspension of 8×106 T cells was homogenised on ice in 150 µl of 150 mM ammonium acetate solution (Sigma, 17836) and the protein content of the lysate determined through a BCA assay. Aliquots of 650 µg protein content per replicate (~20 µl lysate) were resuspended in 30 µl methanol. A mix of 770 µl MTBE (methyl tert-butyl ether, Sigma, 650560), 172.5 μl methanol, 0.01% BHT (2,6-Di-tert-butyl-4-methylphenol, Sigma, B- 1378) and 2.5 μl 100 µM internal standard of phospholipids was combined with the sample. The mixture was rotated overnight at 4℃. The next day 200 μl of 150 mM ammonium acetate were added and mixed on a vortex for 10 minutes at 4℃. To separate the aqueous from the organic phase, the sample was centrifuged for 5 minutes at 2000×g. The lipid-containing MTBE (top) phase was removed and transferred to a fresh vial and stored at -80℃ un l analy . Before mass spectrometric analysis the sample was diluted

10–100 fold in 2:1 methanol:chloroform with 5 mM ammonium acetate.

The acquisition of mass spectrometry data was performed by Sarah Norris at Wollongong University on a hybrid triple quadrupole linear ion trap mass spectrometer (QTRAP® 5500 AB Sciex, MA, USA) equipped with an automated chip-based nano-ESI (electrospray ionisation) source (TriVersa Nanomate™, Advion Biosciences, NY, USA). PC/SM precursor ion scanning and PE neutral loss scans were performed in positive mode, while fatty acid multiple precursor ion scans were combined in negative mode. The data was analysed with Analyst QS 1.5.1 (AB Sciex) and LipidView 1.3 beta (AB Sciex) software. Uneven chain lengths were not included in the analysis. The presence of etherlipids was only analysed in PE and PC glycerophospholipids. The mass tolerance was set to 0.5, signals were identified with a minimum intensity of 0.1 and a signal to noise ratio of 5. Isotope corrected peak

areas were used for relative quantification of lipid contents and corrected for the values of blank samples.

Table 2-5: Scan frequencies of Qtrap5500 mass spectrometer employed in this study and the corresponding head groups and fatty acids detected in positive and negative ion mode (PIS=precursor ion scan, NLS=neutral loss scan, FA=fatty acid).

Positive ion mode Scan Range Lipid class PIS 184.1 640-1000 PC/SM NLS 141.0 685-950 PE NLS 185.0 755-965 PS

Negative ion mode Scan Range Lipid class PIS 153 585-950 PA/PG/PI/PS/CL PIS 241 730-1000 PI, FA 15:0 227.2 580-900 14:0' 253.2 600-900 16:1' 255.2 600-900 16:0' 269.3 560-900 17:0' 279.2 600-900 18:2' 281.3 600-900 18:1' 283.3 600-900 18:0' 297.3 600-900 19:0' 301.2 500-1000 20:5' 303.2 600-1000 20:4' 305.2 600-1000 20:3' 307.2 600-1000 20:2' 309.2 600-1000 20:1' 311.2 600-1000 20:0' 327.2 700-1000 22:6' 329.2 700-1000 22:5' 331.2 700-1000 22:4' 333.3 600-1000 22:3' 335.2 700-1000 22:2' 337.3 700-1000 22:1' 339.3 600-1000 22:0' 365.3 700-1000 24:1'

A suspension of 8×106 T cells was homogenised on ice in 1 ml homogenisation buffer (0.25 M sucrose, 20 mM HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid), 0.5 mM EDTA, complete protease inhibitor). The protein content of the lysate was determined through a BCA assay. Aliquots of 98 µg protein content per replicate (270 µl lysate) were combined with 1350 µl methanol and mixed on a vortex for 15 seconds. In a capped glass tube, a mix of 810 µl 20 mM acetic acid, 930 µl chloroform with 0.01% BHT, 1581 µl (pure) chloroform and 10 μl 300 μM d6-cholesterol internal standard was added to the sample. The mixture was agitated on a vortex for 15 seconds. To separate the aqueous from the organic phase the sample was centrifuged for 5 minutes at 3000×g. The lipid-containing chloroform (bottom) phase was carefully removed with a glass Pasteur pipette, transferred to a fresh glass tube and evaporated under vacuum (Savant SC210 SpeedVac, Thermo Scientific). Dried lipids were resuspended in 50 µl 2:1 methanol:chloroform with 5.6 mM ammonium acetate and transferred to a fresh glass vial and stored at -80℃ un l analysis. Before mass spectrometric analysis the sample was diluted 10–100 fold in MEOH:H2O (85%:15%) + 5mM ammonium acetate.

The liquid chromatography mass spectrometry (LC/MS) was performed by an experienced operator on an Accela LC and autosampler system (ThermoFisher Scientific,USA) with a LTQ Orbitrap XL mass spectrometer (ThermoFisher Scientific)321.

The data was analysed with Analyst QS software and corrected for the values of blank samples.

All data were analysed with Prism 6 (GraphPad Software). Statistical significance is indicated by asterisks corresponding to p-values as follows: *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001. Comparisons that resulted in p-values above 0.05 were considered to be not significantly (ns) different.

The statistical significance of two data sets was assessed with an unpaired Student’s t-test unless otherwise indicated.

Multiple comparisons were assessed with the 1-way ANOVA or 2-way ANOVA and Sidak’s post-testing. In Sidak’s multiple comparisons post-tests the effect of the treatment

(activated/not activated) and the effect of the genotype (wild type/knock-out) was compared individually to answer either of these questions: Is the control group different from the treatment group? Are wild type mice different from AnxA6-/- mice?

Clark et al. were the first to link AnxA6 with the immune system after characterising the expression of AnxA6 in healthy human tissues208. They found AnxA6 to be upregulated in the development of T and B cells. In vitro approaches showing that AnxA6 interacts with LCK203, a kinase that also associates with the TCR and the IL-2 receptor, seemed to support the idea that AnxA6 might play a role in T cell function. The arrival of the AnxA6 knock-out mouse model, developed by Hawkins et al.280, eventually created the opportunity to investigate the effects of AnxA6 loss on the immune system in vivo. Due to a neomycin resistance knock-in into the ANXA6 gene, AnxA6-/- mice do not express the AnxA6 protein in any of their tissues. However, despite the complete loss of this widely expressed protein, AnxA6-/- mice are healthy, do not exhibit any apparent phenotype and breed normally. This challenged previously held views on the role of AnxA6 of being required for budding of clathrin-coated pits236 or for caveolar endocytosis268 vesicles, cardiovascular function and immunological development. Vesicle traffic is of fundamental importance for cell viability, so that an organism would not be viable if this process were severely affected by the loss of this protein. While AnxA6 is not essential for endocytosis over-expression studies since then showed that AnxA6 stimulates LDL receptor endocytosis and lysosomal targeting212. AnxA6 has also been identified as one of the major annexins in the heart322 and selective over-expression of AnxA6 in the cardiomyocytes was deleterious to heart and health323. But AnxA6 knock-out animals have generally normal cardiovascular development280, only on closer inspection cardiomyocytes were found to have altered mechanical and intracellular calcium signalling properties in cardiomyocytes324. It has been argued that, since the annexins are a large family of proteins with overlapping functions, ‘relatives’ of AnxA6, like AnxA1, AnxA2, AnxA4 or AnxA5, could take over functions of the missing one. However, protein extracts of spleen, heart and liver from

knock-out mice did not show upregulation of the tested annexins AnxA1, AnxA2 and AnxA5280. AnxA6 is the only annexin that has been implicated in regulating T cell development due to its expression profile in lymphatic tissues208, yet the immunological development of AnxA6-/- mice is normal. The level of CD4+ and CD8+ T cells in the spleen and the thymus of AnxA6-/- and wild type mice were assessed by flow cytometry and no significant differences were found280. Similarly, flow cytometric analysis of B cells, myeloid, granulytic and monocytic cells did not show and differences between AnxA6-/- and wild type mice280.

A second AnxA6 knock-out mouse was recently generated by the Wellcome Trust Sanger Institute316. The Mouse Phenotyping Project found that this mouse strain has a significantly slower clearance of bacteria from the gut (caecum) after a C. rodentium challenge than the wild type mouse strain. This finding indicates a higher susceptibility of AnxA6-/- mice to infections. However, no publication has to date examined the role of AnxA6 in T cells in depth by studying the cell-mediated immune response in mice in vivo.

Here, a delayed-type CHS was elicited in mice to assess a T cell-mediated immune response. Hypersensitivity is induced by repeated dermal exposure to an irritant. This causes an inflammation which becomes evident in the form of local tissue swelling and increased T cell proliferation. Delayed-type hypersensitivity takes several days to develop and is also known as type IV hypersensitivity. Other types of hypersensitivity elicit an antibody rather than a T cell-mediated immune response. The skin of the mouse is exposed to an irritant in the first phase (day 0 and 1), during which antigen-specific T cells are activated and consequently facilitate an immune response in all peripheral lymphatic tissues, especially the lymph nodes. The clonal expansion of activated T cells in the lymph nodes results in a large number of primed effector T cells and memory T cells. These T cells are still in circulation when, in the second phase, the animal is challenged again, locally, at a different site, with the same irritant (day 5). The effector cells are then able to respond much faster to the same antigen than the naïve T cells preceding them and can cause a local inflammation within 24 hours325. The delayed-type CHS is set apart from other types of immune responses as it mainly depends on the effector T cells generated in the initial expansion phase326,327. This means the immune challenge has to be carried out days (6-9 days) rather than weeks after the initial sensitisation. If the repeat encounter with the antigen occurs after the effector T cells have been removed from circulation by apoptosis, the response will be mediated by memory T cells and antibodies and not by recently generated effector T cells.

To elicit a CHS response, C57Bl/6 (wild type) and AnxA6-/- mice were treated with the irritant DNFB (Figure 3-1). DNFB is a small organic molecule which is a hapten and therefore needs a larger carrier protein, i.e. self-proteins from the upper skin layers or BSA, to be immunogenic. In solution, DNFB and BSA form a hapten-carrier adduct, which upon application and penetration of the skin is recognised by DCs resident in epidermis and dermis. DCs are APCs which enable activation of T cells resident in skin-draining lymph nodes that recognise the presented antigen. Before the first exposure, there are only few CD4+ and CD8+ T cells with the ability to react to this new antigen. Once activated, this population proliferates extensively over the next 4-5 days and differentiates into CD4+ or CD8+ effector and memory T cells. Repeated exposure to the hapten-carrier adduct activates these effector T cells. This results in further T cell proliferation in the lymph nodes and leads to enlarged secondary lymphatic organs. Additionally, effector

Figure 3-1: Treatment of mice to elicit delayed-type CHS in vivo. On Day 0 and Day 1 (Sensitisation phase) an acetone/olive oil solution containing 0.5% DNFB with 1 µg/ml BSA and 0.25% DNFB with 1 µg/ml BSA, respectively, was applied to the shaved abdomen of wild type and AnxA6-/- mice. The control group only received the acetone/olive oil mix. On Day 5 (Challenge) all mice received the same treatment: the left ear was painted with a solution containing 0.2% DNFB with 0.8 µg/ml BSA, while the right ear was painted with the same solution without DNFB and BSA. On day 6 all mice were culled. In the sensitisation phase DNFB/BSA compounds are engulfed by DCs and presented to naïve T cells in the lymph nodes. This activates antigen-specific T cells, which start to proliferate and differentiate into effector T cells. After reapplication of DNFB/BSA on day 5, DCs or macrophages present non-self antigen to effector T cells. This induces migration of effector T cells towards the ear pinna treated with DNFB and triggers homing of all T cell subsets towards the lymph nodes.

T cells migrate towards the inflammation site, guided by chemokines and adhesion molecules expressed by the adjacent blood vessels, and thus contribute to local tissue swelling.

A previously established method to elicit CHS through DNFB treatment328 was adopted to conduct the study in two stages. Firstly, wild type and AnxA6-/- mice were sensitised to DNFB on day 0 and day 1 by application of the compound to the abdomen and in the second stage they were challenged by repeating the exposure to the irritant on the ear pinna on day 5 (Figure 3-1). The control group only received the “challenge treatment” without being sensitised to DNFB. All animals were culled 24 hours later and their lymphocytes were analysed for activation and proliferation markers of CD4+ and CD8+ T cells by flow cytometry.

Before the in vivo T cell responses in a CHS model were examined, it was tested whether wild type (abbreviated “WT” in figures) and AnxA6-/- mice had similar T cell subpopulations. As AnxA6 has been implicated in T cell development208, there was a concern that knocking out this protein might lead to changes in the ability of naïve T cells to differentiate into effector T cells. However, when the AnxA6-/- mouse strain was generated, Hawkins et al. did not find significant differences in CD4+ and CD8+ T cell subsets compared to wild type mice280.

To confirm that the expression of T cell subsets in the AnxA6-/- was not significantly different from the T cells found in the wild type mice, a single cell suspension from lymph nodes of untreated mice was prepared for flow cytometry analysis. The T cell surface proteins CD4, CD8, CD62L and CD44 were labelled with fluorescently tagged antibodies (Figure 3-2 A1-4). Out of the lymphocyte population (Figure 3-2 A-1), the single positive CD4+ and CD8+ T cells (Figure 3-2 A-2) were identified in a flow cytometry dot plot. Of each CD4+ and CD8+ subset, the CD62L signal was plotted against the signal of CD44 (Figure 3-2 A-3 and A-4). Probing for the glycosylated cell surface proteins CD44 and CD62L is a common method to examine the levels of effector (CD62L- CD44+) and naïve (CD62L+ CD44-) T lymphocytes in mice. CD44 is a cell-adhesion molecule which is expressed on the surface of effector T cells in mice and on central memory T cells. It binds to hyaluronan on the surface of other cells to facilitate cell-cell interactions 329. CD62L or

Figure 3-2: AnxA6-/- mice display normal levels of T cell subsets. Lymph nodes from untreated mice were stained for CD4, CD8 (A-2), CD62L (A-3 and A-4) and CD44 (A-3 and A-4) and analysed by flow cytometry. An example of the gating of the FACS data is shown in panel A. Graphs in B-D show levels of lymphocytes, CD4+ and CD8+ detected in lymph nodes of wild type to AnxA6-/- mice. The presence of naïve (CD62L+CD44-), effector (CD62L-CD44+) and memory T cells (CD62L+CD44+) was investigated separately for CD4+ (C) and CD8+ T cells (D). Data of 3 experiments with 1 mouse per group in triplicate. Analysis: 2-way ANOVA, post test: compared wild type to AnxA6-/- mice with Sidak’s multiple comparisons. Error bars show SD with n=3 mice from 3 independent experiments.

L-selectin is located on the surface of naïve and central memory T lymphocytes. It is an adhesion molecule that interacts with endothelial cells and thereby enables the lymphocyte to slowly roll along the inside of the blood vessel. Furthermore, CD62L facilitates the passage of naïve and central memory T cells into peripheral lymphoid organs, but, upon activation of the T cell, L-selectin is shed from its surface330. The CD62L+ and CD44- population was considered naïve; CD62L- and CD44+ were considered to be effector T cells and double positive (CD62L+ CD44+) T cells are central memory T cells.

Staining T cells from lymph nodes of knock-out and wild type mice showed no significant differences in levels of lymphocytes, CD4+ and CD8+ T cells (Figure 3-2 B). The levels of naïve, effector and memory subsets of CD4+ (Figure 3-2 C) and CD8+ (Figure 3-2 D) T cells also did not show any significant differences between lymph nodes from wild type and AnxA6-/- mouse. These results confirm previous findings. Therefore, it could be assumed that wild type and AnxA6-/- mice had comparable levels of CD4+ and CD8+ T cells and comparable levels of naïve, effector and memory T cells of respective subsets.

In the course of the in vivo assay, an irritant is applied to the skin of the mice. The efficiency of such treatments can vary with the body mass of the mouse. Consequently, it was important to match treatment groups by body weight. Furthermore, production of naïve T cells decreases in adult mice due to thymic atrophy. Avoiding a mix of juvenile (4-8 weeks) and adult (older than 8 weeks) individuals, mice used in this in vivo study were aged between 12.4 and 15.9 weeks. The four treatment groups, wild type “control” and “sensitised” and AnxA6-/- “control” and “sensitised”, comprised a similar spread of age differences within this range. Analysis of the average age (Figure 3-3 A) and body weight (Figure 3-3 B) of all treatment groups did not reveal a significant difference between them.

This analysis demonstrates that the in vivo study with adult male mice was conducted on experimental groups maching in age and weight.

Figure 3-3: Animals used in in vivo assay do not differ in age and weight. The average age (A) and weight (B) of 5 male mice used in the in vivo study; Analysis: 1-way ANOVA, n=5, error=SD.

After establishing that untreated AnxA6-/- and wild type mice harbour comparable levels of CD4+ and CD8+ naïve, effector and memory subsets in their lymph nodes, and that the mice selected for an in vivo study are matching mice in age and body weight, an in vivo study was performed to test if AnxA6-/- T cells activate and proliferate at the same rate as wild type T cells in the response to a challenge of the immune system.

The in vivo study was conducted twice with similar results. The first experiment used four male mice per group, the second experiment used five male mice per group. All data presented are from the second experiment with five male mice per group unless otherwise indicated. In the second study, one mouse in the “AnxA6-/- control” group was found to exhibit symptoms of sickness and was removed from the analysed data.

In this in vivo study, wild type and AnxA6-/- mice were sensitised to DNFB by applying the irritant to the shaved abdomen on day 0 and 1 (Figure 3-1), while the control group, also consisting of mice from both genotypes, was only exposed to the solvent solution (acetone and olive oil, 4:1). On day 5, the exposure was repeated to elicit an immune response, but this time the irritant was applied to the ear pinnae. Mice from both treatment groups and both genotypes, were treated the same by having one ear pinna painted with the solvent solution, while the DNFB containing solution was applied to the other ear. On day 6, the mice were sacrificed and the ear pinnae were measured for swelling, spleens and inguinal lymph nodes were removed, weighed and kept for further analysis.

Comparing the thickness of the ear pinnae of the animal (Figure 3-4 A) reveals significant swelling of the DNFB treated ear compared to the solvent painted ear in the group of sensitised mice, but no differences in thickness of ear pinnae in the control groups. Similarly, the sensitised mice exhibit enlarged spleens (Figure 3-4 B) and enlarged lymph nodes (Figure 3-4 C), which yielded higher amounts of lymphocytes (Figure 3-4 D) compared to the control treatment. The differences between the two treatment groups are a strong indication, that the immune challenge was successful in both wild type and knock- out mice.

Figure 3-4: Successful elicitation of delayed-type hypersensitivity response. Wild type (green bars) and AnxA6-/- mice (orange bar) in the treatment group were sensitised to DNFB (Day 0 and 1), while mice in the control group only received the vehicle solution on their shaved abdomen. On day 5, both groups, the control group and the DNFB treated group, had one ear exposed to a DNFB containing solution, while the other ear only received the vehicle. (A) The thickness of left and right ear pinnae. Hatched bars show mean values of ears treated with DNFB on day 5, while plain bars show average thickness of ears that only received vehicle solution. (B) The weight of whole spleens relative to the body weight. (C) The relative weight of a pair of excised inguinal lymph nodes compared to the body weight of the animal. (D) The absolute cell numbers in single cell suspensions prepared from inguinal lymph nodes. A, B, C and D show data of one experiment with 5 mice per group (except for “AnxA6-/- control” n=4). Data are representative of 2 experiments. Analysis: (A) 2-way ANOVA, post test: wild type compared to AnxA6-/- mice with Sidak’s multiple comparisons, (B)-(D) 1-way ANOVA. Error bars show SD.

To compare the level CD4+ T cells and CD4+ subsets of wild type and AnxA6-/- mice after the immune challenge, single cell suspensions of the harvested lymph nodes were labelled with surface markers for flow cytometry analysis. The gating for the lymphocyte (Figure 3-5 A-1 and Figure 3-6 A-1), CD4+, CD8+ (Figure 3-5 A-2 and Figure 3-6 A-2), CD44+ and CD62L+ (Figure 3-5 A-3 and Figure 3-6 A-3) populations was conducted as before (Figure 3-2 A1-4).

The levels of CD4+ T cells (Figure 3-5 B) in the lymph nodes of the control group were the same in wild type and AnxA6-/- mice. The levels of CD4+ T cells in the lymph nodes of the control group and the sensitised group of wild type mice were the same. Similarly, the levels of CD4+ T cells from AnxA6-/- mice were the same in the control and sensitised group. This indicated that the sensitisation treatment did not change the percentage of CD4+ T cells within the T cell population. Also, there was no difference in CD4+ levels between the sensitised wild type and the sensitised group of knock-out mice. Next, the absolute cell numbers displaying a CD4+ phenotype were analysed (Figure 3-5 C). The numbers of CD4+ T cells of the control group were the same in wild type and knock-out mice. The wild type mice sensitised to DNFB prior to the immune challenge showed a significant (~6.5×) increase in CD4+ T cells compared to the control group. Similarly, AnxA6-/- mice displayed 11.5-fold amplification of CD4+ T cell numbers in the group that received sensitisation treatment compared to the control group. This shows that CD4+ T cells successfully proliferated inside the lymph nodes and/or successfully homed towards the lymph nodes of sensitised mice, while the genotype of mice had no effect on this process.

The level of CD4+ CD62L+CD44- naïve T cells was not significantly different between the wild type and AnxA6-/- mice of the control group (Figure 3-5 D) The percentage of naive T cells was slightly reduced in the sensitised group compared to the control group of wild type as well as AnxA6-/- mice, but this reduction was not significant. There was no difference in the percentage of naïve CD4+ T cells between wild type and AnxA6-/- mice of the sensitised group. Analysis of the absolute cell numbers of naïve T cells in the lymph nodes of the control group showed no difference between wild type and AnxA6-/- mice (Figure 3-5 G). The sensitised groups of wild type, as well as AnxA6-/- mice, both had a strong, significant increase of the amount of naïve T cells in the lymph node compared to the control group of wild type (6 times more) and AnxA6-/- mice (11 times more). The

number of CD62L+CD44- T cells in the sensitised group was not significantly different between wild type and AnxA6-/- mice. The strong increase in naïve T cell numbers in the sensitised group shows that naïve T cells effectively migrated towards the lymph nodes and were retained there regardless of whether they were wild type or AnxA6-/- T cells.

Analysis of the levels of CD62L-CD44+ effector T cells (Figure 3-5 E) showed that the AnxA6-/- mice control group exhibited slightly increased levels of effector T cells compared to the wild type group, but the difference did not reach significance. The lymph nodes of sensitised wild type mice contained a higher percentage of effector cells than the control group, but again, this difference was not significant. AnxA6-/- mice did not exhibit any difference in effector cell levels between the control and the sensitised group. There were also no significant differences in the effector T cell levels of sensitised wild type and sensitised AnxA6-/- mice. The absolute cell numbers in the control group lymph nodes of wild type and AnxA6-/- mice were not significantly different (Figure 3-5 H). However, the absolute amount of effector T cells in the group of sensitised wild type mice was significantly higher (8.8 times) than in the control group. Likewise, the number of effector T cells was significantly increased by 12.3 in the group of sensitised AnxA6-/- mice compared to the control treatment. The amount of effector T cells found in the lymph nodes of sensitised wild type and AnxA6-/- mice was not significantly different. These results indicate that, in wild type and AnxA6-/- mice, effector T cells homed towards the lymph nodes or proliferated there and the knock-out did not affect this process.

Memory T cells were identified by the expression of both surface markers CD62L and CD44 and they are the smallest of the analysed CD4+ T cell subsets. The control group of wild type and knock-out mice expressed the same percentages of CD62L+CD44+ T cells (Figure 3-5 F). Lymph nodes of sensitised wild type mice contained a significantly higher proportion of memory T cells than the group of the control treatment. AnxA6-/- mice of the sensitised group also displayed a significant increase in the percentage of memory T cells in comparison with their respective control group. The groups of sensitised wild type and sensitised AnxA6-/- mice, however, did not show any significant differences. The analysis of absolute numbers of CD4+ memory T cells found in the lymph nodes showed the same trends (Figure 3-5 I). There were no significant differences in the memory T cell numbers of the wild type and AnxA6-/- control groups. The group of sensitised wild type mice, on the other hand, exhibited significantly increased (11.7 times) numbers of memory T cells. Likewise, AnxA6-/- mouse lymph nodes contained significantly higher (17.8 times) numbers of memory T cells. Sensitised wild type mouse lymph nodes seemed to contain

Figure 3-5: T cell development and expression of surface markers on naïve and effector CD4+ T cells in wild type (green bars) AnxA6-/- (orange bars) CD4+ T lymphocytes after CHS induction. Flow cytometry samples were gated according to the examples shown in A-1 to A-3. Levels of CD4+ (B and C) from lymph nodes of DNFB-sensitised and control mice were acquired with flow analysis together with relative levels (D, E, F) and absolute cell numbers (G, H, I) of CD62L+ CD44-, naïve (D, G), CD62L- CD44+ effector (E, H) and CD62L+ CD44+ central memory (F, I) T cells of the CD4+ subset. All panels show data of one experiment with 5 mice per group (except for “AnxA6-/- control” n=4). Data are representative of two experiments. Analysis: 2-way ANOVA, error bars: SD. Asterisks designate significant difference compared to control group of the same genotype: **p<0.01, ***p<0.001 and ****p<0.0001. slightly less memory T cells than lymph nodes of sensitised AnxA6-/- mice, but this

difference was not significant.

In conclusion, there were no significant differences between wild type and AnxA6-/- mice in the levels or absolute numbers of CD4+ T cells and respective naïve, effector and memory subsets. The increase of naïve T cell numbers in the groups of sensitised mice as seen in Figure 3-5 C, G, H and I were a result of intensified homing of all (including not antigen specific) T cells into the lymph nodes and proliferation in the lymph nodes as a result of the hyper sensitivity reaction. This demonstrates that the final exposure to DNFB stimulated a largely T cell based response.

Flow cytometry data analysis (Figure 3-6 A1-A3) was done in the same fashion as described before (Figure 3-2 A) and CD8+ subsets were analysed for the expression of CD62L and CD44 surface markers (Figure 3-6 A-3).

The levels of all CD8+ T cells extracted from inguinal lymph nodes made up about 40% of the T cells found in lymph nodes of the control group of both wild type and AnxA6-/- mice (Figure 3-6 B). The levels of CD8+ T cells in the control group were not significantly different between wild type and AnxA6-/- mice. Comparing the CD8+ T cell levels of the control group and sensitised group of wild type mice showed a significantly decreased percentage of CD8+ T cells in sensitised wild type mice. Similarly, the percentage of AnxA6-/- CD8+ T cells was reduced in sensitised knock-out mouse lymph nodes compared to the control group. The percentage of CD8+ T cells was the same between sensitised wild type and sensitised AnxA6-/- mouse lymph nodes. The absolute cell numbers of CD8+ T cells were not significantly different between lymph nodes of wild type and AnxA6-/- mice in the control group (Figure 3-6 C). The absolute cell numbers in the groups of mice receiving the sensitisation treatment were, as expected, significantly increased compared to their respective control for both wild type (4.8-fold) and AnxA6-/- mice (7.6- fold). The absolute cell numbers of CD8+ T cells in the lymph nodes of sensitised wild type and AnxA6-/- mice were the same (Figure 3-6 C). This indicated that CD8+ T cells successfully proliferated and/or homed towards the lymph nodes of sensitised animals where the CHS was elicited. The decrease of relative levels of CD8+ T cells in the lymph nodes of sensitised wild type, as well as AnxA6-/- mice, implies that the type of immune response did not induce a very strong influx or proliferation of CD8+ T cells.

The percentage of CD8+ T cells with a naïve CD62L+CD44- phenotype was not significantly

Figure 3-6: T cell development and expression of surface markers on naïve and effector CD8+ T cells in wild type (green bars) AnxA6-/- (orange bars) CD4+ T lymphocytes after CHS induction. Cells were gated according to the examples shown in A-1 to A-3. Levels of CD8+ (B and C) from lymph nodes of DNFB-sensitised and control mice were acquired with flow analysis together with relative levels (D, E, F) and absolute cell numbers (G, H, I) of CD62L+ CD44-, naïve (D, G), CD62L- CD44+ effector (E, H) and CD62L+ CD44+ central memory (F, I) T cells of the CD8+ subset. All panels show data of one experiment with 5 mice per group (except for “AnxA6-/- control” n=4). Data are representative of two experiments. Analysis: 2-way ANOVA, error bars: SD. Asterisks indicate significant difference compared to control group of the same genotype: *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001.

different between wild type and AnxA6-/- mice of the control group. Comparing the levels of naïve T cells in the sensitised group with the control group shows that the percentage of naïve CD8+ T cells was reduced in sensitised wild type as well as sensitised AnxA6-/- mice (Figure 3-6 D), but these differences were not significant. There was no difference between the levels of CD62L+CD44- T cells in sensitised wild type and sensitised AnxA6-/- mice. The absolute numbers of naïve T cells in the lymph nodes of wild type and AnxA6-/- mice of the control group was not significantly different (Figure 3-6 G). The lymph nodes of sensitised wild type mice contained 4.3 times as many naïve CD8+ T cells as the control group. Likewise, the lymph nodes of sensitised AnxA6-/- mice contained 7.6 times as many naïve CD8+ T cells as the control group. In both cases, these increases were significant. The numbers of naïve CD8+ T cells in the lymph nodes of sensitised wild type and sensitised AnxA6-/-, however, was not significantly different. These data show that naïve CD8+ T cells successfully homed towards the lymph nodes of sensitised mice that were challenged to elicit an immune response.

The percentage of CD62L-CD44+ effector T cells was slightly, but not significantly, increased in the lymph nodes of AnxA6-/- control mice compared to the wild type control group (Figure 3-6 E). The lymph nodes of sensitised wild type mice included a significantly higher percentage of CD8+ effector T cells compared to the control group. The proportion of CD8+ effector T cells in the lymph nodes of sensitised AnxA6-/- mice was also increased compared to the control, but not significantly. Despite this, the levels of CD62L and CD44+ T cells in the lymph nodes of sensitised mice were not different between wild type and AnxA6-/- mice. The yield of CD8+ effector T cells in terms of total number of cells from lymph nodes of wild type and AnxA6-/- mice of the control group was not significantly different (Figure 3-6 H). The lymph nodes of sensitised wild type and sensitised AnxA6-/- mice, however, contained substantially more CD8+ effector cells than the respective control groups. In wild type mice, CD8+ effector cells were significantly increased by 10.7 times and in AnxA6-/- mice, they were 11-fold higher. The absolute amounts of CD62L-CD44+ T cells were not considerably different between wild type and AnxA6-/- mouse lymph nodes. These results indicate that in sensitised wild type and AnxA6-/- mice effector T cells homed towards the lymph nodes or proliferated there and the knock-out did not affect this process.

Comparing the percentage of CD8+ CD62+CD44+ central memory T cells of the control groups as well the sensitised groups of wild type and AnxA6-/- mice showed a higher amount of CD8+ memory T cells in the knock-out mice (Figure 3-6 F) but this difference

was not significant. The proportion of CD8+ memory T cells did not differ between the lymph nodes of the wild type control and the wild type sensitised group and neither did it differ between the AnxA6-/- control and sensitised group. The absolute cell numbers of CD8+ memory T cells in the lymph nodes of control wild type and control AnxA6-/- mice was not significantly different (Figure 3-6 I). In the lymph nodes of sensitised wild type and sensitised AnxA6-/- mice, however, considerably more CD8+ memory T cells were found. The cell numbers were increased 5.4-fold and 6.5-fold, respectively. This increase was significant. The numbers of CD8+ T cells in the lymph nodes of sensitised wild type and sensitised AnxA6-/- mice was not significantly different between the two genotypes.

The analysis of CD8+ T cells and their subsets in the lymph nodes of wild type and AnxA6-/- mice shows that an immune response was stimulated in CD8+ T cells, since their absolute numbers were significantly higher in the lymph nodes of sensitised mice than in lymph nodes of mice from the control group. There were no significant differences in the relative abundance or absolute number of T cells in the lymph nodes of sensitised wild type and AnxA6-/- mice. The AnxA6 knock-out therefore did not affect the ability of CD8+ T cells to home towards the lymph nodes and participate in the immune response.

In summary, the absolute cell numbers of CD4+ T cells extracted from the inguinal lymph nodes of wild type and AnxA6-/- mice, which received the DNFB sensitisation treatment before the immune challenge on day 5, were significantly increased (Figure 3-5 C). Sensitised wild type and AnxA6-/- mice also contained significantly increased numbers of all CD4+ naïve, effector and memory T cell subsets compared to the control group (Figure 3-5 G, H, I). The percentages of CD4+ T cells and CD4+ T cell naïve and effector subsets (Figure 3-5 B, D, E) found in the lymph nodes of sensitised wild type and AnxA6-/- mice were not significantly different from the control mice. Only the percentage of CD4+ central memory cells (Figure 3-5 F) was significantly increased in the lymph nodes of sensitised mice after the immune challenge. These cells constitute only a small subset (less than 6%), but they are more likely to retained in the lymph nodes than naïve T cells331.

Similar to CD4+ T cells, the absolute numbers of CD8+ T cells and CD8+ T cells subsets (Figure 3-6 C, G, H, I) were significantly increased in the inguinal lymph nodes of sensitised wild type and AnxA6-/- mice compared to the control group. But the percentage of CD8+ T cells (Figure 3-6 B) made up a smaller proportion of lymphocytes in the lymph nodes of sensitised wild type and AnxA6-/- mice compared to the control group. The reduced presence of CD8+ T cells in lymph nodes of sensitised mice was also reflected in the absolute cell numbers of CD4+ (Figure 3-5 C) and CD8+ T cells (Figure 3-6 C). The

increase of T cells in the lymph nodes of sensitised mice after the immune challenge was less intense in CD8+ T cells (4.8 and 7.5-fold increase) than in CD4+ T cells (6.5 and 11.5- fold increase). The lower percentage of CD8+ T cells in the lymph nodes of sensitised wild type and AnxA6-/- mice implies that the type of immune response did not induce an influx or proliferation of CD8+ T cells as strong as in CD4+ T cells. This observation was in agreement with data from Tedla et al328.

The absolute numbers of CD4+ and CD8+ T cells and any of their naïve, effector or memory subsets were not significantly different between wild type and AnxA6-/- mice of the sensitised group as well as the control group. The same was true for the percentages of CD4+ and CD8+ T cell and respective subsets. The CHS immune challenge therefore seemed to induce a similar immune response in wild type mice and AnxA6-/- mice.

Due to the significant increases of absolute T cell numbers in the sensitised mice after the immune challenge, it can be inferred that the immune response was elicited successfully and induced the homing to the lymph nodes and/or proliferation of the analysed T cell subsets of the sensitised mice.

The analysis of CD4+ and CD8+ T cell numbers above indicate that extensive proliferation took place in the sensitised group of mice. The labelling of surface proteins cannot be used to investigate the level of T cell proliferation by flow cytometry. There are, however, alternative methods to measuring T cell proliferation directly. In vitro, cell division is commonly analysed with fluorescent dyes that are incorporated into the cytosol of the cell of interest. With every cell division, the fluorescence signal in the cytoplasm is reduced by about 50%. In vivo this procedure is invasive and not suited to this experiment. Another method is to supplement the drinking water with DNA precursors such a bromodeoxyuridine (BrdU) which is incorporated into the cellular DNA over time. Once the supplementation of BrdU is stopped, the marker is diluted with every cell division. Naïve T cells, however, incorporate this marker very poorly even when this supplement is given over several weeks332. To still be able to analyse the proliferation activity of T cells, it is possible to measure the increase in the expression of PCNA at the end of the in vivo study. This approach requires permeabilisation and fixation of the cells isolated from the lymph nodes to label PCNA which is expressed in the cell nucleus as a co-factor of DNA polymerase δ. PCNA is upregulated during cell division, i.e. in the G1 and the S-phase (peak

expression) of the cell cycle333,334. Therefore, analysing T cells with the fluorescently labelled anti-PCNA antibody has become a means to measure proliferation activity328. Since PCNA is continuously expressed at low levels in interphase, the gating of a proliferating population in flow cytometry data is done by detecting a change in the histogram shape. Since proliferating cell populations express more PCNA, the histogram shifts to the right towards higher fluorescence intensity. Commonly, such peak shifts are analysed through median fluorescence intensity (MFI). But since the signal of PCNA stained cells (Figure 3-7) did not follow a clear unimodal distribution, the increase in MFI was not a reliable measure to detect the increase in fluorescence. Instead, a small peak at the “bright” end of the fluorescence scale (grey histogram in Figure 3-7 A-3 and A-4) which only forms in proliferating cell populations was used as an indication of dividing cells. Here, this population is referred to as “PCNA+”. In this analysis the results from two experiments were pooled to help identify outliers in the data. The “AnxA6-/- control” and

Figure 3-7: T cell proliferation in wild type (green bars) AnxA6-/- mice (orange bars) after immune challenge. Flow cytometric analyses of murine T cells isolated from lymph nodes of mice treated with DNFB. Permeabilised lymphocytes (A-1) were stained for CD4, CD8 (A-2) and PCNA (A-3, A-4). Examples for the gating strategies are provided in A1-4. Levels of PCNA expression were measured in CD8+ (B) and CD4+ T cells (C) of DNFB sensitised mice and the control treatment. B and C show data of two pooled experiments with 9 mice per group (except “AnxA6-/- control” n=7 and wild type control n=8). Analysis: 2-way ANOVA, error bars: SD, ****p<0.0001.

“WT sensitised” group each had one outlier, which was removed. Significant outliers were detected with Grubbs’ test (α=0.05) with the GraphPad outlier calculator before the data from two individual experiments was pooled.

Flow cytometry analysis of CD8+ lymphocytes showed that there were low levels of PCNA+ cells in the wild type and AnxA6-/- control group (Figure 3-7 B) that were not significantly different between the two groups. Wild type, as well as AnxA6-/-, animals sensitised to DNFB prior to the immune challenge were found to express ~5 times higher levels of PCNA than the control groups. This demonstrates that CD8+ T cells of mice that received repeated doses of DNFB harboured more proliferating CD8+ T cells than the control group. The increase in PCNA expression of CD8+ T cells from sensitised wild type and sensitised AnxA6-/- mice was not significantly different.

The CD4+ T cells (Figure 3-7: C) of the wild type and AnxA6-/- control group also expressed PCNA at low levels without any significant differences between wild type and knock-out animals. In the treatment group of sensitised wild type mice, the PCNA levels were ~5 times increased compared to the control, while in sensitised AnxA6-/- mice PCNA levels were only ~4 times higher than in the control treatment. When the PCNA expression of sensitised wild type mice was compared to the expression in sensitised AnxA6-/- mice, the PCNA levels in AnxA6-/- mice were significantly reduced.

These data show that the sensitisation treatment increased the amount of PCNA+, i.e. proliferating CD4+ and CD8+ T cells in sensitised mice compared to the control group. This increase indicates that the DNFB treatment successfully initiated T cell proliferation in the sensitised group of mice. The increased proliferation of CD8+ T cells was not significantly different between sensitised wild type and knock-out mice. Analysis of the CD4+ T cell subset, however, demonstrated that the levels of proliferating T cells in AnxA6-/- mice were significantly lower than in wild type mice. Therefore, it can be concluded that AnxA6-/- CD4+ T cells were impaired in their cell division by the lack of AnxA6.

This study aimed to examine the effect of an AnxA6 knock-out in T cells during an immune response in mice in vivo. Development of T cells in untreated AnxA6-/- mice was normal as T cells in the lymph nodes of AnxA6-/- expressed the same levels of CD4+ and CD8+ T cells and respective naïve, effector and memory subsets compared to the wild type mice. This indicated that AnxA6-/- mice were furnished with the same number of naïve CD4+ and

CD8+ T cells as wild type mice at the beginning of the in vivo study. The groups of mice in this study were also confirmed to be of a similar average age and body weight before the experiment commenced.

A T cell mediated immune response was stimulated by eliciting a CHS in wild type and AnxA6-/- mice. In this scenario only the repeated exposure of the skin of the animal to a hapten irritant can cause an immune response. To prompt a hypersensitivity reaction in mice wild type and AnxA6-/- mice were divided into a group receiving treatment with an irritant, i.e. sensitisation to the irritant DNFB on day 0 and day 1, and a control group that was not exposed to DNFB on those days. On the day of the immune system challenge (day 5) all mice, of the control as well as the sensitised group, were exposed to DNFB on one of their ear pinnae while the other ear only received the vehicle solution. It was expected that the DNFB treated ear pinna swells up in sensitised mice, but not in mice of the control group, since they had never before come in contact with the irritant. Measurements of ear pinna thickness 24 hours after the immune challenge confirmed that only the DNFB treated ear pinnae of sensitised mice were significantly thicker, i.e. swollen, than the other ear on the same animal, which only received the vehicle solution. The ear pinnae of the mice from the control group exhibited the same thickness, no swelling, regardless of whether the ear was exposed to DNFB or the vehicle solution. The DNFB treatment only elicited a local swelling in the DNFB treated ear of the mice that had been sensitised to DNFB. Analysis of the spleen and lymph node weights of mice from the sensitised and control group showed that, after the immune challenge, the sensitised mice had heavier spleens and lymph nodes than the control group. This weight increase demonstrated that the immune response was not confined to the ear of the sensitised mice, but systemic reaction had been elicited by the immune challenge.

Analysis of the T cell subsets in the inguinal lymph nodes of wild type and AnxA6-/- mice showed that the immune challenge drastically increased the number of CD4+ and CD8+ T cells and their naïve, effector and memory subsets within the lymph nodes of sensitised mice. The percentage of the different T cell subsets within the lymph node was not significantly different between control and sensitised mice - with the exception of a small relative increase in CD4+ memory cells in sensitised wild type and AnxA6-/- mice.

After the immune challenge, the levels of CD4+ T cells were the same in sensitised mice and mice of the control group. But the percentage of CD8+ T cells found in the lymph nodes of sensitised mice was reduced compared to control mice. The relative reduction of CD8+ T cells inside the lymph nodes of sensitised mice indicated that influx of CD8+ T cells

into the lymph nodes, or proliferation within the lymph nodes, was not as strong as the influx and proliferation of CD4+ T cells. The increase of total T cell numbers in sensitised mice, as well the relative reduction of CD8+ T cells found in lymph nodes of sensitised mice, occurred in wild type, as well as AnxA6-/- mice, without any significant differences between the two genotypes. Likewise, all other analysed factors, like the swelling of the ear pinna, lymph nodes and spleens, did not show any differences between wild type and AnxA6-/- mice. The lack of any significant differences in these parameters shows that the knock-out of AnxA6 did not prevent T cells and other immune cells from invading local areas of inflammation (ear pinna) or homing towards lymphatic organs (spleen and lymph nodes). This led to a general increase of T cells in the lymph nodes of sensitised mice in wild type as well as AnxA6-/- mice.

Analysing the proliferative response of T cells found in the lymph nodes of wild type and AnxA6-/- mice showed that CD4+ T cells of AnxA6-/- mice differed significantly in their proliferation response compared to their wild type equivalent. Surprisingly, the proliferation increase of CD8+ T cells was not affected by the lack of AnxA6. This could be due to the fact that, in this study, the CD8+ response was generally lower than the CD4+ T cell response and therefore proliferation was not as strong in CD8+ as in CD4+ cells. Under these circumstances, proliferation differences would not necessarily be detectable. The differential proliferative response of CD4+ and CD8+ T cells to this immune challenge was in agreement with a publication from Tedla et al328.

While the number of CD4+ T cells found in the lymph nodes of wild type and AnxA6-/- mice was the same, less of these CD4+ T cells were involved in cell division. This indicated that the influx of T cells into the lymph nodes of AnxA6-/- mice worked very efficiently, but it might result in decreased levels of CD4+ T cells in the periphery since less new CD4+ T cell are created during the immune response. The T cell numbers outside of the lymph nodes were not assessed in this study, but, if they were reduced, it can be surmised that the immune system of AnxA6-/- mice might be more vulnerable if it has to deal with more than one “threat” at a time.

The results of the in vivo study imply that the observed proliferation defect in CD4+ T cells was due to slight changes in the process of inducing T cell proliferation in AnxA6-/- mice, since other aspects of the immune response, like local swelling and homing of T cells into lymphatic organs, were not affected by the knock-out of AnxA6. There are several possibilities for how the knock-out of AnxA6 could affect the proliferation behaviour of CD4+ T cells. The fact that in this assay, only CD4+ T cells display the proliferation defect

could indicate that AnxA6-/- T cells are not sustaining a long enough contact with the APC to yield the same response as the wild type, due to a lower half-life of the TCR/pMHC interaction. If AnxA6 provides a link between the F-actin cytoskeleton and the cellular membrane275, it is possible that T cell signalling at the immune synapse is affected. Alternatively, the lack of AnxA6 could have a negative effect on IL-2 signalling. The IL-2 receptor complex was found to reside in membrane rafts and to interact with the F-actin cytoskeleton prior to endocytosis72,167,182. Finally, due to the potential involvement of AnxA6 in cholesterol homeostasis212,248, it is also possible that the knock-out of AnxA6 has a negative effect on signalling involving membrane proteins located in cholesterol-rich lipid domains or even on cell division as a whole.

The following chapters describe whether T cell signalling in AnxA6 knock-out or knock- down cells was affected at the stage of TCR signalling, IL-2 signalling or whether the lipid composition of the T cell membrane was affected by the lack of AnxA6.

Naïve T cells are programmed to remain in interphase until they recognise a pMHC on an APC and receive an activation stimulus. Until this encounter, they only receive sub- mitogenic survival signals from transient interaction with self-pMHC and IL-7 stimulation63. It has been found that naïve T cells can divide sporadically335, but this process does not provide substantial amounts of new T cells. Extensive proliferation is only induced through T cell activation and IL-2 stimulation. After the initial activation through the TCR, the cell presents more IL-2 receptors on the cell surface, which makes it susceptible to lower IL-2 concentrations66. When T cell proliferation is reduced, it can indicate that TCR or IL-2 receptor signalling processes are altered.

In the first step of T cell activation, the TCR recognises an antigen presented by an APC. When CD4 or CD8 co-receptors confirm a productive interaction with a MHC complex, TCR signalling starts a phosphorylation cascade inside the T cell. In naïve T cells an additional co-stimulatory signal is necessary to initiate this signalling cascade2. This signalling activity takes place at the immunological synapse. This structure formed at the contact site of T cell and APC sees the T cell membrane divided into concentric zones termed SMAC. Actin forms a stabilising ring on the outside around the dSMAC aiding in moving newly activated TCR into TCR microclusters and towards the pSMAC and the central cSMAC zone of the contact site. At the cSMAC, TCRs are clustered together very densely336. While signalling takes place at the cSMAC, it is also the location for signal termination due to TCR internalisation56,57. While the SMAC structure does not seem to be necessary for TCR signalling54, an intact F-actin cytoskeleton is essential for TCR microcluster formation and signal initiation55–57. In particular, the dynamics of F-actin flow from the dSMAC to the pSMAC seem to be required for signalling176.

Many proteins are involved in initiating and coordinating the polymerisation of actin at the immunological synapse at the time of T cell activation. TCR recruits the Src family kinase LCK, which phosphorylates the intracellular ζ domains of the TCR. In proximity to these domains, ZAP70 is activated by phosphorylation and in turn phosphorylates up to 4 tyrosine residues of the adapter protein LAT15,337. Membrane-anchored LAT acts as a signalling platform for many other proteins, like phosphorylated PLCɣ1 and VAV25–27. PLCɣ1 is instrumental to initiating the MAP kinase, NFκB and Ca2+ dependent pathways leading to T cell growth and differentiation. VAV is essential for actin reorganisation via WASP, WAVE (WASP family verprolin-homologous) and ARP2/3. Activated VAV allows WASP and WAVE to recruit the actin nucleation protein ARP2/3 and thus start the polymerisation of F-actin336. This process directs F-actin to the plasma membrane together with small GTPases like RAC1 and ARF that are targeted to the membrane owing to their lipid anchors31,338.

With its ability to bind F-actin, as well as PS and cholesterol-rich membranes, AnxA6 could potentially participate in linking F-actin to the plasma membrane once Ca2+ is released as part of the PLCɣ1/IP3 signalling cascade275. Since PS and cholesterol are enriched at the immunological synapse170 AnxA6 could contribute to targeting F-actin to cholesterol-rich membrane domains around TCR microclusters. However, more recent publications show that the activation site might be enriched in positively charged lipids upon activation, which could be a key mechanism for signalling proteins to unfold into their active conformation339. An abundance of positively charged head groups of phospholipids could lead to AnxA6 being excluded from the immediate environment of the TCR.

A second process promoting T cell proliferation is IL-2 signalling. TCR signalling eventually induces IL-2 expression and secretion. The released IL-2 binds to IL-2 receptors of the same or a neighbouring cell. The IL-2 receptor is a trimeric cytokine receptor consisting of CD25, CD122 and the ɣc-subunit, which is shared with many other type 1 cytokine receptors340. CD25 expression is upregulated dramatically after T cell activation341. CD122 together with the ɣc-subunit of the IL-2 receptor are responsible for JAK/STAT signalling processes72. A high affinity IL-2 receptor that will engage in signalling after binding IL-2 only forms with all subunits present342. IL-2 binding at the receptor induces conformational changes that facilitate phosphorylation of JAK1 and JAK372, which in turn phosphorylates STAT5a and STAT5b. STAT5 dimerises with itself and binds to STAT responsive elements on the genome, which promote cell differentiation or proliferation343. LCK is also a target for phosphorylation by CD122. As phosphorylated LCK

(pLCK) can contribute to STAT5 phosphorylation76,344. LCK creates a link between TCR and IL-2 signalling77; it has been suggested that the role of LCK is to prevent apoptosis rather than furthering cell proliferation78. Interestingly, AnxA6 but not A1, A2 or A4, was shown to bind to this tyrosine kinase203. The biological relevance of this interaction has not yet been identified. Other signalling pathways that are initiated through IL-2 receptor activation are the MAP kinase and the PI3K pathways345.

Like the TCR, IL-2 receptor signalling depends on the spatial organisation of the plasma membrane. IL-2 receptor subunits are found in membrane rafts and their disruption by cholesterol sequestration impairs IL-2 signalling167,181,182. More importantly, binding to IL- 2 induces a conformational change of the CD122 and ɣc subunits that promotes the association of the receptor complex with membrane rafts and the actin cytoskeleton and induces downstream signalling72.

To investigate the involvement of AnxA6 in T cell signalling in this thesis, isolated primary murine T cells were used as well as the human Jurkat T cell line. Jurkat cells are an immortalised lymphocyte line that was derived in the late 1970s from the peripheral blood of a 14-year old boy who was treated for acute lymphoblastic leukaemia346. Since then Jurkat cells have become the standard tissue culture model for studying T cell signalling. The advantage of using cell lines rather than primary T cells is that they can be grown easily in a tissue culture incubator, no animals need to be sacrificed and they can easily be manipulated to express exogenous proteins or silence the expression of endogenous proteins. An AnxA6-eGFP fusion protein was expressed in Jurkat cells to investigate co-localisation of AnxA6 and F-actin during T cell activation. Furthermore, some preliminary assays were conducted with a stable AnxA6 knock-down cell line (AnxA6KD) with a Jurkat T cell background. In this cell line, the expression of AnxA6 was reduced through shRNA complementary to AnxA6 mRNA. This shRNA construct (for sequence see Appendix) was introduced into Jurkat cells via a lentiviral vector. The vector also contained a gene for puromycine resistance, which was used to select successfully transduced Jurkat cells. A control cell line (CTRL) was created with a lentiviral vector that only contained the puromycine resistance gene. The knock-down efficiency was around 50-70% (see Chapter 4.2.5).

There are, however, certain disadvantages to conducting experiments with transfected or transformed cell lines. The expression of a fluorescent fusion protein can result in overexpression and the introduction of the fluorescent tag has the potential to alter protein function and binding. Transforming cells with a viral vector to silence gene

expression can generate artefacts originating from the knock-out technique itself. The infection with an integrating virus like lentivirus always harbours the risk of interrupting essential alleles and regulatory sequences in the DNA. In addition, the knock-down can just decrease protein expression but not entirely abolish it. Consequently, the remaining protein might still be able to cover the most important functions of the protein of interest and prevent the manifestation of a clear phenotype. These risks make it preferable to undertake experiments with cells that are known to only have the genetic defect of interest.

AnxA6-/- mice only contain a genetic defect in the ANXA6 . Therefore, observed differences to the wild type mice of the same genetic back ground are more likely to be a result of exactly this defect, i.e. the abrogation of AnxA6 expression. Furthermore, as primary T cells are isolated from animals rather than propagated in culture for years like the Jurkat lymphoma T cell line they also present a more physiologically relevant model. Primary T cells, albeit from murine rather than human origin, contain the physiological diversity of all T cells present in one individual rather than one T cell clone of a certain subset. The advantages and disadvantages of the use of cell lines were carefully considered and as a result they were both used for the experiments presented here.

Spleens and lymph nodes were excised from wild type and AnxA6-/- mice sacrificed by CO2 inhalation. Freshly isolated spleens and lymph nodes were then homogenised and T cells were isolated from the homogeneate with the AutoMACS Pan T cell isolation kit (Miltenyi). In brief, the splenocytes were washed and incubated with a solution of antibodies labelling all cells from the suspension with the exception of T cells. The non-T cell population was then removed from the cell suspension with the help of magnetic beads which bind specifically to the antibodies. In an AutoMACS cell sorter (Miltenyi) all magnetically labelled cells are retained by a special column while all unlabelled cells (T cells) are eluted. T cell fraction was then used for in vitro experiments. To characterise purity and T cell populations of the unlabelled fraction the cells were incubated with fluorescent antibodies anti-CD3, anti-CD4 and anti-CD8 or IgG isotype controls and analysed by flow cytometry.

The main lymphocyte population was selected in an FSC/SSC dot plot (Figure 4-1 A). The signal detected in this plot very loosely correlates to the size of cell and helps to exclude

debris and cell lumps. Of the lymphocyte population CD3+ cells were selected in a plot showing the intensity of the anti-CD3-PE-Cy7 staining (Figure 4-1 B). Overlaying the sample with an IgG-PE-Cy7 isotope control helped to the position the CD3+ gate and confirmed the specificity of the anti-CD3-PE-Cy7 antibody (Figure 4-1 C). 91.2% (average 90.6%) of the lymphocyte population were T cells (CD3+). CD4 and CD8 T cell populations were then detected in a dot plot visualising CD4-V500 as well as CD8-PacificBlue (Figure 4-1 D). The position of the CD4+ and CD8+ gates and the specificity of the anti-CD4 and anti-CD8 antibodies were confirmed by an overlay of the sample with an isotype control dot plot (Figure 4-1 D). The T cell population consisted of 30.4% (average 29.2%) CD8+ T cells, 62.9% (average 64%) CD4+ T cells and 6.3% double negative (CD4-CD8-) T cells.

Figure 4-1: Characterisation of wild type T cells isolated from spleen. T cells were isolated from spleen cell suspension and labelled with anti-CD3-PE-Cy7, anti-CD4-V500 and anti-CD8- PacificBlue for flow cytometry analysis. (A) The main lymphocyte population was selected according to its position in an FSC/SSC plot. (B) The CD3+ population was selected according to fluorescence intensity of the PE-Cy7 signal. (C) Overlay of an IgG-PE-Cy7 isotope control (red) with a sample (blue). (D) The CD4+ and CD8+ population were selected from CD3+ cells by plotting PacificBlue fluorescence versus V500 fluorescence. (E) Overlay of an IgG-V500 and IgG-PacificBlue isotype control (red) with a sample (blue). (F) CD3-CD4+ and CD3-CD8+ cells. Pseudo-coloured plots (A, B, D, F) show the density of detected signals (red to blue from high to low). Percentages shown in plots refer to the respective parent population with exception of (F) which shows percentage of CD3- subsets within lymphocytes rather than CD3- cells. Plots and percentages show results of one replicate of a flow cytometry analysis in triplicate. Average percentages are provided in text.

Double negative T cells outside of the thymus are usually ɣδ T cells. Analysis of the CD3- subset showed that the majority of these cells are CD4-CD8-. Only 0.9% and 0.72% of

lymphocytes are CD4+ and CD8+, respectively. Very few CD3- cells express CD4 and CD8 surface receptors, e.g. NK cells can express low levels of CD8 and CD4 is found on DC.

The analysis of the eluted cell fraction of the AutoMACS showed that ~90% of cells within the lymphocytes gate were T cells (CD3+). The T cell population of the eluted cell suspension contained ~64% CD4+ and ~29.2% CD8+ T cells. This means that the eluted fraction was indeed strongly enriched in T cells since CD4+ and CD8+ T cells together only make up 20-25% of all lymphoid cells in spleens of C57Bl/6 mice347. Analysis of the CD3- population showed that it was largely made up of CD4-CD8- cells. Due to recurring issues with the anti-CD3-PE-Cy antibody staining and the very low amount of CD4+ and CD8+ cells within the CD3- population, the gating for CD3+ cells was omitted in the following experiments.

To investigate the sub-cellular localisation of AnxA6 and F-actin during T cell activation activated T cells were imaged with TIRF microscopy. TIRF microscopy eliminates fluorescence of molecules that are not located at or proximal to the plasma membrane. This imaging technique requires the activation site to be in direct contact with the microscope slide, since TIRF illumination penetrates only to maximum of 200 nm into the specimen.

There are several ways to initiate the T cell activation process in vitro. It is possible to employ APCs, adherent pMHCs or pharmacological agents like ionomycin and phorbol esters348. pMHCs are usually receptor specific and pharmacological agents do not induce direct TCR ligation348. If APC were used for in vitro activation it would confine experimentation to confocal microscopy rather than TIRF microscopy. The most efficient way to activate cell lines, as well as primary T cells, is directly through the TCR with the use of specific antibodies that do not simply bind T cell surface proteins, but induce activation of the TCR complex in the absence of pMHC. The OKT3 antibody clone induces polyclonal T cell activation and stimulates the synthesis and release of cytokines349. Binding of OKT3 to the human CD3ε subunit activates TCR signalling. For naïve T cells and for optimal activation of Jurkat cells, a co-stimulatory ligand, like anti-CD28 (clone CD28.2) or ICAM-1 (intercellular adhesion molecule), should be provided. For murine T cell activation similar antibody clones (anti-CD3 clone 17A2 and anti-CD28 clone 37.51) need to be employed to achieve T cell activation in vitro. Glass microscope slides, as well as

plastic surfaces, can be coated with any of these antibodies and only adherent antibodies result in full T cell activation.

For determining the sub-cellular localisation of a protein it is common to use immune fluorescence. After fixation of the cells and permeabilisation of the plasma membrane with mild detergents or organic solvents, the specimen is incubated subsequently with a primary antibody specific to the protein of interest and a fluorescently labelled secondary antibody. T cells, however, often exhibit high levels of background fluorescence due to unspecific binding of the primary as well as secondary antibody. It was found that for this study the available anti-AnxA6 antibodies lacked the necessary specificity for labelling AnxA6 in T cells. Therefore Jurkat cells were transfected with a rat AnxA6-eGFP fusion protein as used previously248,264. Introducing the fusion protein into the cell by transfection harbours the risk of introducing artefacts due to over-expression. Attempts to circumvent this problem by transfecting primary AnxA6-/- with a fluorescently tagged AnxA6 construct and thereby restore the intracellular AnxA6 expression to approximately wild type levels were not successful.

To study the localisation of AnxA6 in activated T cells and to investigate whether it co- localises with F-actin, AnxA6-GFP transfected Jurkat T cells were activated for 2, 5 and 10 minutes before they were fixed with PFA and stained with fluorescently labelled phalloidin, a fungal toxin that binds F-actin with great specificity350,351 (Figure 4-2 A). Images of the same cell were acquired in two channels consecutively. The imaged cells show a characteristic ring of F-actin at the border of the activated cell, while AnxA6 extensively occupied the plasma membrane. After acquisition, the channels were overlayed and analysed in ImageJ. For all images the degree of co-localisation of the signal from the two channels (red and green) was ascertained by determining the Pearson’s Correlation Coefficient (Rr). Pearson’s Correlation Coefficient can assume a value between -1 and +1, with -1 indicating that pixels of different colours (here: red and green) are mutually excluded from each other, 0 indicating random overlap and +1 indicating 100% overlap. The image analysis (Figure 4-2 B) shows that in the early phases of activation (2 and 5 minutes) the signal of AnxA6 and F-actin are positively co-localised (Rr=0.68±0.11 and Rr=0.63±0.16) and overlap significantly more than after 10 minutes (Rr=0.39±0.22).

The images also show that both proteins formed a very different pattern during T cell activation. While AnxA6 was present in most areas with F-actin, F-actin was not present in all areas where AnxA6 was located. Nevertheless the results of the image analysis show

that AnxA6 and F-actin do co-localise to some degree during early T cell activation. Therefore, AnxA6 and F-actin can potentially interact at the immune synapse.

Figure 4-2: Localisation of AnxA6 and Actin at the immune synapse. Jurkat cells were transfected with AnxA6-eGFP, activated on anti-CD3 and anti-CD28 coated coverslips for the indicated amount of time, fixed and stained with phalloidin conjugated to AlexaFluor555. (A) Images were acquired with transmitted light and a dichroic filter (DIC); green and red fluorescence TIRF images were acquired using 488 nm (AnxA6-eGFP) and 561 nm (Phalloidin conjugated with AlexaFluor555) laser excitation. (B) Pearson’s Coefficient of the fluorescence channels was measured. Analysis: 1-way ANOVA, n≥10 cells, error=SD.

While co-localisation is a necessary condition for interaction, it does not prove that two proteins are in direct contact or interact in a meaningful way. If AnxA6 forms a critical link between F-actin, the plasma membrane and/or other components of the immune synapse, primary T cells isolated from AnxA6-/- mice might display a different organisation of F- actin during T cell activation than T cells from wild type mice.

Once an immunological synapse is established in an activated T cell, newly formed TCR bound to MHC of the APC are transported from the periphery towards the centre of the immune synapse assisted by the dynamic F-actin cytoskeleton352. During this transport TCR associated signalling proteins ZAP70 and LCK are recycled. In this process, the dynamic restructuring of F-actin is critical. This is exemplified by the fact that pharmacological inhibition of actin remodelling blocks T cell activation and the formation of TCR microclusters56,353,354.

The F-actin organisation in AnxA6-/- and wild type T cells was investigated by the means of combining TIRF with dSTORM. STORM is a single molecule super-resolution fluorescent microscopy technique, which requires a series of images in which only a fraction of all

fluorescently labelled proteins emit fluorescence in each image of the series355. Post- acquisition, the centroid of the intensity profile of every detected isolated fluorescent event is analysed and the x- and y-coordinates of the single fluorescence molecule recorded (Figure 4-3 A)355. A refined version of this method that employs only one fluorescent dye label rather than a combination of two different ones is referred to as dSTORM356. The combination of dye (here: Alexa Fluor 647) and buffer (Oxygen scavenging buffer) results in “blinking” of the fluorescence dye rather than illumination of all labelled protein and bleaching over time356. The blinking appears since the fluorophore stochastically enters a stable dark state during which is susceptible to the reducing agents in the buffer. In its reduced form the fluorophore is dark for several seconds356.

Employing single molecule imaging allows the investigation of the organisation of F-actin in the early stages of T cell activation. Naïve primary T cells did not form a dense ring of actin as quickly after activation as Jurkat cells did, but did do so at later stages of activation. The F-actin appearance was more evenly spread and punctate as F-actin filaments were assisting in the organisation TCR microclusters of the immune synapse. To image primary T cells from wild type and AnxA6-/- mice T cells were isolated from the spleen as described before. Isolated T cells were activated on anti-CD3 and anti-CD28 antibody-coated coverglasses, fixed, permeabilised and stained with a phalloidin- conjugated dye suitable for dSTORM (AlexaFluor647, Molecular Probes). The patterning of F-actin filaments at the immune synapse was compared between activated wild type and AnxA6-/- T cells.

Activated cells were imaged on a TIRF microscope (Elyra, Zeiss, Figure 4-3 A-1, left) following the procedure and analysis method of Williamson et al318. In brief, single molecule images were generated from a 15000 frame time-series with an algorithm identifying the centre of the light signal of each molecule as previously described357. Every detected event was exported to a table containing information on the x-y coordinates of each molecule. From these data a point pattern of the molecular distribution was generated (Figure 4-3 A-1, right). In the reconstituted image square regions of 1.5×1.5 µm containing more than 290 molecules were selected for cluster analysis (Figure 4-3 A-1, light blue square). Ripley’s K-function was then calculated using the SpPack319 Excel plugin. The function takes into account the area of the analysed region (here 1.5×1.5 µm, Figure 4-3 A-2), the number of points, the spatial scale (radius r) and the distance between two points. It counts the number of molecules encircled by concentric rings centred on each molecule, normalised to the average molecular density of the entire region. With the

help of the results from the calculation of Ripley’s K-function the L-function is generated (equations in Methods). The function L(r) equals r (i.e. L(r)=r), if the spatial distributions of molecules is completely random. L(r)-r was then plotted against r (Figure 4-3 C). This means, within a radius where a distribution that is more clustered than a random distribution L(r)-r was positive. Values ≤0 for L(r)-r indicate random distribution. Points at the edge of the distribution region were weighted to compensate for edge-related effects. Additionally, values of L(r) generated for each point at a value of r=50 nm (L(50)) were then calculated with a MATLAB (The Mathworks Inc., Natick, MA) script (Clustermappingv13.m) and used to produce a two-dimensional cluster heat map (Figure 4-3 A-3) where red and yellow areas indicate highly clustered areas and blue symbolises least clustered areas. These cluster heat maps were thresholded to produce binary cluster maps (Figure 4-3 A-4). Areas of the interpolated map with L(50)>80 were defined as clusters (white).

The first analysis of the distribution of F-actin signals in activated wild type and AnxA6-/- T cells (Figure 4-3 B) took into account only the information of molecule coordinates. For this, 1.5×1.5 µm regions were selected from the inside of the visible cell area avoiding areas of very dense F-actin distribution like the edge of the cell. Analysing the degree of clustering over a range of distances plotting L(r)-r versus r yielded a curve that shows the clustering of F-actin at a scale from 0 to 1500 nm (Figure 4-3 C). Clustering started low for very small spatial scales, but increased radically peaking at r∼50 nm. Beyond this the molecules were still clustered but the decrease of the curve means that the distribution of molecules became closer to random (L(r)-r≤0) the larger observed area was. This analysis of F-actin clusters up to a radius of 750 nm in wild type and AnxA6-/- T cells showed that in AnxA6-/- T cells F-actin was more clustered on all spatial scales. The peak values of the curve of L(r)-r showed that the radius within which the maximum degree of clustering occurred was similar in wild type and AnxA6-/- T cells (Figure 4-3 D). The maximum degree of clustering occurred around 50 nm in wild type and AnxA6-/- cells. The same 1.5×1.5 µm regions were reanalysed by generating a heat map of molecule density which was thresholded to be converted to a binary cluster map to investigate clustering of F- actin on smaller scales (at 50 nm). The clusters of F-actin in AnxA6-/- T cells detected in the binary map contained significantly more molecules per cluster than clusters of wild type T cells (Figure 4-3 E). Furthermore, the radius of these clusters was larger in AnxA6-/- T cells than in wild type T cells (Figure 4-3 F).

Figure 4-3: dSTORM imaging of F-actin with activated primary wild type and AnxA6-/- T cells. T cells isolated from spleens of wild type and AnxA6-/- mice were activated for 12 minutes on antibody-coated coverslips, fixed, permeabilised and stained with Phalloidin-Alexa Fluor 647. Cells were imaged in STORM buffer on a TIRF microscope. (A) Single molecule images were generated from a 15000 frame time-series. 1.5x1.5µm regions were selected from the inside of the activated cell. From the detected single molecule signals of the chosen regions a heat map was generated indicating areas of high (red) and low (blue) signal density, which was converted into a binary (thresholded) image mapping areas corresponding to high signal density as a cluster (white). (B) Example images of activated wild type and AnxA6-/- T cell. (C) The randomness of the distribution of molecules in a radius r radius was analysed for regions 1.5x1.5µm in not thresholded images. Positive values for L(r)-r indicated a non-random distribution. (D) The maximum values of L(r)-r are plotted again. Higher values indicate a more clustered population. Analysis of clusters in binary image data showed (E) the number of molecules per cluster and the (F) radius of detected clusters. Data analysed with Student’s t-Test, wild type 54 regions from 7 cells and AnxA6-/- 34 regions from 6 cells, (C) error=SEM, (D-F) error=SD, ****p<0.0001.

In the analysed regions activated AnxA6-/- T cells were found to have a more clustered F- actin on all spatial scales up to a radius of at least 750 nm. The clustering of F-actin peaked at a radius of approximately 50 nm. The number of actin molecules per cluster was also higher in AnxA6-/- T cells than in wild type T cells. The size of detected actin clusters in AnxA6-/- was larger than in wild type T cells. These results indicate that F-actin, which in the early stages of T cell activation contributes to TCR cluster formation, is more densely organised in AnxA6-/- T cells than in wild type T cells.

It is not clear why AnxA6-/- T cells display a more dense appearance of F-actin endings at the activation site than cells from WT mice. If AnxA6 is involved in creating a link between the plasma membrane and F-actin, the increased density of F-actin might be necessary to compensate for the lack of a stabilising component like AnxA6.

For T cells to actively participate in an immune response, they need to be able to detect and react to chemotactic gradients in their environment and respond to homing signals, which induce the aggregation of B cells, T cells and DCs in lymph nodes. At steady state, naïve T cells circulate through the blood and the lymphatic system in a rolling fashion mediated by the L-selectin (CD62L) surface molecule of naïve T cells330. In proximity of a lymph node, endothelial cells (as well as cells inside the lymph node) secrete the chemokine CCL21358. This chemokine activates the integrin LFA-1 on the T cells and promotes adhesion to the endothelium by binding ICAM-1 expressed only by endothelial cells close to a lymph node (these cells are also called high endothelial venules)330. Attracted by the chemotactic cues, T cells migrate through the specialised endothelium of the high endothelial venules into the lymph node where they encounter mature DCs presenting samples of self and non-self antigen359. In the absence of activating signals, T cells leave the lymph node through the efferent lymphatic vessel. Effector T cells are guided towards sites of inflammation in the tissue by similar mechanisms, since they carry a different set of surface proteins than naïve T cells. In effector T cells, CD62L is shed from the surface while integrin VLA-4 (very late antigen) is expressed in large numbers330. Activated effector T cells secrete cytokines to attract other immune cells and stimulate the expression of VLA-4 and LFA-1 in T cells and the expression of binding molecules VCAM-1 (vascular cell adhesion molecule) and ICAM-1 in the surrounding vessels28,360.

The migration machinery is dependent on F-actin polymerisation. If AnxA6 provides a link between F-actin and specialised lipid domains in the plasma membrane and/or proteins in proximity to activated receptors, it is possible that chemotactic migration may be affected by the loss of AnxA6 since there are similarities between TCR and chemokine receptor signalling187. Furthermore, T cell migration and activation go hand in hand, since contact between a T cell and an APC may initially occur through the leading edge361 where TCR recognition and crawling delivers a stop signal for migration. Binding of chemokine molecules, such as SDF-1α (stromal cell-derived factor 1α), induces a switch from a spherical to a polarized T cell shape with two distinct regions: the distal pole or uropod and the leading edge, which contains chemokine receptors and F-actin and facilitates cell

interaction with the substrate. Moreover, AnxA6 expression is down-regulated or abrogated in stationary adherent cancer cells295, but it was found to be expressed in breast cancer carcinoma with an invasive phenotype and reduced expression of AnxA6 compromised motility and invasiveness of breast cancer cells218. This differential expression of AnxA6 in invasive and non-invasive cancer cells suggested that AnxA6 might play a part in cell migration.

Since defects in T cell migration caused by AnxA6 knock-out could affect the ability of T cells to mediate an effective immune response, it was necessary to confirm that any differences in immune competence, which became apparent in the in vivo study were not a result of impaired migration in AnxA6-/- lymphocytes. To this end the migration efficiency of primary murine T cells of wild type and AnxA6-/- T cells towards a chemotactic cue was assessed in a Boyden chamber assay. The same experiment was conducted with Jurkat T cells – comparing an AnxA6 knock-down cell line with an appropriate control. These cell lines are discussed in more detail in the next subchapter.

WT AnxA6-/-

SDF-1α Figure 4-4: Migration efficiency of wild type and AnxA6-/- primary T cells. The concentration of T cells in a target well of a Boyden chamber (5 µm pore size) filled with media or SDF-1α was determined after 3 hours of incubation at 37℃. The migration efficiency of wild type and AnxA6-/- primary murine T cells was compared by 2-way ANOVA using Sidak’s multiple comparison post- test. Data were pooled from two experiments performed in triplicate. Error bars show SD with n=6, ****p<0.0001.

In a Boyden chamber, a known number of T cells with media were added to the upper compartment, while the lower compartment was filled with cell-free medium. The compartments were separated by a filter with a defined pore size (5 µm). To induce migration, the chemoattractant SDF-1α was added to the bottom compartment, while the control condition (“Media”) contained only medium. After 3 hours at 37℃ in a tissue culture incubator, the cell number in the bottom well was determined.

Since naïve T cells exhibit a low activity of random migration, the lower compartment of the “Media” control contained some T cells after 3 hours of incubation even in the absence of chemoattractant (Figure 4-4). The numbers of wild type and AnxA6-/- T cells found in the medium-filled well after 3 hours of incubation were not significantly different from each other. This indicated the levels of random migration after 3 hours of incubation were the same in wild type and AnxA6-/- cells. Significantly more wild type and AnxA6-/- cells were counted in the wells containing SDF-1α than in the media-filled well. The numbers of wild type and AnxA6-/- T cells in the chemoattractant-filled wells were not significantly different. The significantly higher amount of T cells found in the chemoattractant filled wells shows that the T cells actively migrated towards the SDF-1α containing wells. The results of assessing the migration efficiency of the AnxA6 knock-down and respective control T cell line were very similar and are found in the Appendix section of the thesis.

The lack of any significant differences between wild type and AnxA6-/- T cells shows that the absence of AnxA6 did not affect the ability of naïve AnxA6-/- T cells to migrate towards chemotactic cues. The fact that AnxA6-/- T cells did not show a reduction in migration efficiency shows that any differences observed in the in vivo study are not a result of reduced T cell migration, i.e. reduced influx of T cells into secondary lymphoid organs.

The TCR is engaged for signalling by binding pMHC. The TCR-MHC interaction is confirmed by the T cell co-receptor CD4 or CD8, which binds to MHC15. In naïve T cells the binding of an APC needs to be confirmed through the stimulation of another co-receptor (e.g. CD28). TCR ligation induces the assembly of multi-molecular TCR signalling clusters. TCR is phosphorylated at the ζ-chain and CD3ε as part of the TCR complex is activated by LCK, which phosphorylates the TCR at ITAMs. ZAP70 is able to bind the phosphorylated ITAMs and is eventually activated by LCK through phosphorylation. Phospho-ZAP70 subsequently phosphorylates LAT, which provides a scaffold for several other proteins including PLCɣ1 which induces Ca2+ signalling and ERK activation. Active ERK induces the expression of the mitogenic cytokine IL-2 which usually peaks after 4-8 hours362. Most of these signalling processes are located at and within the membrane and part of dynamic protein clusters with TCR and LAT. The formation of TCR clusters at early stages of T cell (less than 5 minutes) signalling353,363 is dependent on interactions with a dynamic F-actin cytoskeleton15,176. If AnxA6 participates in facilitating a link between F-actin and TCR microclusters the phosphorylation and activation of proteins involved in this cascade might be affected275.

To investigate whether the lack of AnxA6 affects the efficiency of TCR signalling, the phosphorylation of proteins involved in early T cell signalling was tested, initially in a the AnxA6 knock-down cell line. For this, CTRL and AnxA6KD Jurkat cells were activated on anti-CD3 and anti-CD28 antibody-coated surfaces for 2, 12 and 20 minutes. The proteins of the cell lysates were separated by electrophoresis on a 10% bis-tris polyacrylamide gel to resolve a wide range of protein sizes (20 kDa – 150 kDa). The gel contents were blotted onto a membrane and the phosphorylation profiles of several T cell signalling proteins were visualised by western blot immunostaining (Figure 4-5 A). The bands in the western blot images were analysed by densitometry (Figure 4-5 B-D). The signal for total unphosphorylated protein was measured for all proteins where suitable antibodies were available (ZAP70, PLCɣ1, ERK – not shown). All bands were normalised to the β-actin band of the same sample.

Analysis of the AnxA6 expression levels revealed that the knockdown cell line expressed AnxA6 at significantly lower levels than the control cell line. On average the AnxA6 expression in AnxA6KD expressed was only 30-50% of the levels expressed in CTRL cells (Figure 4-5 B).

The amount of pZAP70 was elevated in CTRL and AnxA6KD T cells after 2 minutes of activation and decreased after that. pZAP70 levels remained slightly elevated in CTRL and AnxA6KD T cells even after 12 and 20 minutes of activation compared to the not activated controls. The levels of pZAP70 were not significantly different between CTRL and AnxA6KD cells in resting and activated T cells at any of the measured time points (Figure 4-5 C). The phosphorylation of PLCɣ1 which occurs downstream of ZAP70 activation was increased dramatically in CTRL and AnxA6KD cells after only 2 minutes of activation. pPLCɣ1 levels remained elevated for the 12 and 20 minutes time point compared to levels in resting T cells. The levels of phosphorylated pPLCɣ1 were not significantly different between CTRL and AnxA6KD cells. Further downstream of the signalling cascade, the MAP kinases ERK1 and 2 are activated by phosphorylation - a modification which enables them to activate transcription factor Elk1 (ETS domain-containing protein Elk-1) and c-Fos. After 2 minutes the phosphorylation of ERK1/2 was increased in CTRL and AnxA6KD cells. The amount of pERK1/2 in CTRL and AnxA6KD cells increased even further after 12 and 20 minutes of activation (Figure 4-5 E). The amount of ERK1/2 phosphorylation in CTRL and AnxA6KD cell was similar and not significantly different at any of the measured timepoints.

Figure 4-5: Phosphorylation of signalling in early T cell activation and AnxA6 expression in AnxA6KD and CTRL Jurkat cells. (A) T cells isolated from spleens were activated on antibody- coated surfaces for 2, 12 and 20 minutes. The amount of phosphorylation was analysed by western blot and quantified by densitometry for (B) AnxA6, (C) pPLCɣ1, (D) pZAP70 and (E) pERK1/2. β- Actin staining served as a loading control. Staining of AnxA6 confirms the knock-down of this protein in AnxA6KD cells (40% of control). Analysis: (B) Student’s t-Test, n=3 from 3 experiments with 4 replicates each; (C)-(E) 2-way ANOVA, n=3 from 3 independent experiments, error=SD.

In summary, the activation of Jurkat cells lines CTRL and AnxA6KD on surfaces coated with anti-CD3 and anti-CD28 antibody lead to a notable increase in phosphorylated ZAP70, PLCɣ1 and ERK1/2. The increase in phosphorylated signalling proteins after 2, 12 and 20 minutes showed that the method of activation was effective. CTRL and AnxA6KD T cells did not display significantly different levels of phosphorylation regarding the signalling proteins ZAP70, PLCɣ1 and ERK1/2. Depletion of roughly 60% of AnxA6 in the AnxA6KD cell line did not seem to have any disadvantageous effect on T cell signalling. However, if only small amounts of AnxA6 are involved in assisting in the coordination of F- actin at the immune synapse, the remaining 40% of expressed AnxA6 might be sufficient to create the necessary connections between the plasma membrane and the cytoskeleton. It was therefore necessary to repeat this assay with T cells that did not express any AnxA6. For this purpose, T cells were isolated from the spleens of AnxA6 knock-out mice.

To investigate whether the complete lack of AnxA6 affects the efficiency of TCR signalling, primary murine T cells from AnxA6-/- and wild type mice were activated on antibody- coated beads for 2 and 12 minutes. As a change to the previous assay antibody-coated beads were chosen instead of an antibody-coated well to activate primary T cells. This approach helped to increase the available surface area while keeping cells in a small volume of media (200 µl), since the western blot assay for primary T cells required 3-4

times as many cells as the same assay with Jurkat cells to yield a clear signal with the available antibodies. Cell lysates were separated by SDS-PAGE and the phosphorylation profiles of several T cell signalling proteins were detected by western blot immunostaining as above. Afterwards the signals on western blot images were quantified by densitometric analysis. Figure 4-6 A shows a representative western blot for all proteins analysed in this experiment. Probing for the presence of AnxA6 demonstrated the complete abrogation of AnxA6 expression in AnxA6-/- mice. The bands probed for β-Actin served as a loading control to normalise the individual lanes for densitometric analysis.

Figure 4-6: Phosphorylation of signalling proteins in early T cell activation, AnxA6 expression and IL-2 mRNA production of wild type and AnxA6-/- primary T cells. (A-F) T cells isolated from spleens were activated with antibody-coated beads for 2 and 12 minutes. The amount of phosphorylation was analysed by western Blot and quantified by densitometry. One representative experiment is shown. β-Actin staining served as a loading control. Staining of AnxA6 confirms the absence of this protein in AnxA6-/- animals. Graphs show quantified signals from western blots for (B) pTCRζ, (C) pZAP70, (D) pLAT, (E) pPLCɣ1 and (F) pERK1/2. Analysis: 2-way ANOVA, error = SD, n=3. (G) T cells were isolated from the spleen of AnxA6-/- and wild type mice, activated on antibody-coated surfaces and analysed for mRNA of IL-2. GAPDH and G6PDX were employed as housekeeping genes and assayed in 3-4 replicates. The increase in IL-2 transcription was normalised to not activated wild type and AnxA6-/- cells. Data was analysed by Student’s t-Test, error = SD, n = 3. One representative experiment of 3 is shown.

Furthermore, the transcription levels of IL-2 mRNA after 4 hours of activation were analysed by qPCR. Primary T cells were isolated from freshly excised spleens of wild type and AnxA6-/- mice as before. T cells were then incubated for 4 hours in a 37℃ incubator in well-plates with anti-CD3 and anti-CD28 coated surfaces. Activated T cells were lysed after

4 hours of incubation while the not activated control cells were lysed immediately. Cellular RNA was purified from cell lysates with a phenol-free, column-based system. Afterwards the RNA solution was additionally treated with DNase to safeguard against DNA contamination. Reverse transcription of RNA samples employed a mix containing oligo desoxy-thymidine to ensure preferential reverse-transcription of mRNA of other types of cellular RNA. Primers used in the qPCR were designed to of similar length, melting temperature and with amplicon sizes of 90-300 base pairs (see Appendix). The increase in transcription of IL-2 was determined with the ΔΔCt method364 using the housekeeping genes murine GAPDH and G6PDX as normalisers and were expressed relative to the average IL-2 mRNA signal of all not activated T cells (wild type and AnxA6-/- together).

TCRζ (Figure 4-6 B) and ZAP70 (Figure 4-6 C) did not show increased levels of phosphorylation after activation in wild type or AnxA6-/- cells. This is likely to be a result of already activated pools of protein found in primary T cells365. There was no significant difference in the levels of phosphorylated TCRζ and phosphorylated ZAP70 between wild type and AnxA6-/- T cells. LAT (Figure 4-6 D) showed increased levels of phosphorylation in wild type and AnxA6-/- T cells after 2 and 12 minutes after activation. There were no significant differences in the levels of pLAT between wild type and AnxA6-/- T cells in resting cells and after 2 and 12 minutes of activation. The levels of pPLCɣ1 (Figure 4-6 E) increased dramatically after 2 minutes of T cell activation and remained elevated after 12 minutes of activation. The levels of pPLCɣ1 were not significantly different between wild type and AnxA6-/- T cells. ERK phosphorylation was slightly increased after 2 minutes of activation in wild type T cells (Figure 4-6 F). But 2 minutes of activation did not increase the level pERK in AnxA6-/- T cells. The levels of pERK were markedly increased in wild type and AnxA6-/- T cells after 12 minutes of activation. The phosphorylation of ERK1/2 was elevated in resting AnxA6-/- T cells compared to resting wild type T cells, but this difference was not significant. Likewise the levels of pERK were not significantly different between wild type and AnxA6-/- T cells after 2 and 12 minutes of activation. IL-2 production which is initiated downstream of ERK1/2 was analysed by qPCR of T cells that were activated for 4 hours on antibody-coated surfaces. Analysis of the qPCR data showed that IL-2 mRNA levels increased almost 600-fold after 4 hours of activation of wild type as well as AnxA6-/- T cells (Figure 4-6 G). The increase of IL-2 mRNA was no significantly difference between wild type and AnxA6-/- T cells.

The analysis of activated primary T cells showed that the amount of phosphorylated LAT, PLCɣ1 and ERK did increase in wild type and AnxA6-/- T cells after 2 and 12 minutes of

activation. Surprisingly, there was no marked increase in phosphorylated TCRζ and pZAP70 which act upstream of LAT, PLCɣ1 and ERK. Comparing the level of phosphorylation of any of these T cell signalling proteins did not reveal any significant differences between wild type and AnxA6-/- T cells. The absence of any differences in the levels of phosphorylation between wild type and AnxA6-/- T cells was a strong indication that early T cell activation in AnxA6-/- T cells was just as efficient as in wild type cells. Additionally, the levels of IL-2 transcription after 4 hours of T cell activation were analysed. The intense increase of IL-2 mRNA transcription showed that wild type and AnxA6-/- T cells were successfully stimulated to engage in IL-2 transcription after T cell activation. The increase in IL-2 transcription was not significantly different between wild type and AnxA6-/- T cells. The findings of the western blot analysis showed that TCR signalling in AnxA6-/- T cells was just as efficient as in wild type T cells. This idea was confirmed by the analysis of IL-2 transcription in activated T cells, which was also not significantly different between AnxA6-/- and wild type T cells.

In conclusion, there was no significant difference between the levels of phosphorylated ZAP70, PLCɣ1 and ERK in CTRL and AnxA6KD T cells. The suspicion that the lack of any significant differences could be caused by residual AnxA6 in the knock-out cell line was not confirmed. Analysis of the levels of pTCRζ, pZAP70, pLAT, pPLCɣ1 and pERK in primary AnxA6-/- and wild type primary T cells showed also no differences in the amount of phosphorylated signalling proteins between wild type and AnxA6-/- T cell. Likewise, there was no significant difference between AnxA6-/- and wild type primary T cells in the levels of IL-2 mRNA after 4 hours of T cell activation. It can therefore be said that the investigation of early (2, 12 and 20 minutes) and sustained (4 hours) TCR signalling in vitro through activation with anti-CD3 and anti-CD28 directed antibodies revealed no significant differences between CTRL and AnxA6KD or AnxA6-/- and wild type T cells in TCR signalling. AnxA6 is thus not required for efficient T cell activation.

TCR signalling results in IL-2 and eventually IL-2 secretion. Once released into the surrounding medium it binds to IL-2 receptors on the cell surface of the same cell or a neighbouring cell. Neither TCR signalling nor IL-2 gene expression were diminished in AnxA6-/- T cells. Therefore, the finding that AnxA6-/- mice exhibited a proliferation defect in vivo might have been caused by processes downstream of IL-2 gene expression, since IL-2 is a powerful mitogenic cytokine produced by T cells shortly after activation and reaches its maximum after 5-6 hours362. Furthermore, past research implicates AnxA6 in

secretion and exocytosis. AnxA6 is expressed in resting but not in lactating (secreting) breast epithelial cells208,289. Ectopic over-expression of AnxA6 in CHO cells lead to decreased secretion of fibronectin and TNFα257. Over-expression of AnxA6 in influenza infected cells reduced production of virions, while AnxA6 knock-down increased the number of released viral particles293,294. AnxA6 knock-out could therefore affect the secretion of IL-2.

Figure 4-7: Assay principle of IL-2 secretion assay. Activated T cells and not activated control are incubated for 30 minutes with IL-2 catch reagent, which recognises T cell surface markers (the nature of the surface markers is not specified by the manufacturer of the kit) and binds IL-2. In a second incubation step, the cell surface bound IL-2 is recognised by a fluorescently labelled IL-2 detection antibody. Figure was adapted from the IL-2 secretion assay manual of Miltenyi Biotec.

To study the amount of IL-2 secreting AnxA6-/- and wild type T cells after in vitro activation a single cell suspension of splenocytes isolated from mice was activated for 3.5 hours in a tissue culture incubator on anti-CD3 and anti-CD28 coated surfaces. Subsequently, the cells were removed from the activating surface and labelled with an IL-2 capturing antibody (Figure 4-7) part of a commercial kit to detect IL-2 secreted by T cells. The catch reagent is a bifunctional antibody compound which binds T cell surface markers on one side and “catches” IL-2 with a second binding site. After removing any excess capturing antibody, the suspension of activated T cells was incubated for 30 minutes at 37℃. Once activated T cells continuously secrete IL-2. IL-2 released by any of the T cells labelled with the catch reagent, is tethered to the T cell from which it has been secreted. Due to a low cell density and constant agitation to keep the cell suspension homogeneous, the catch reagent is most likely to only bind IL-2 secreted by the cell it is attached to. Afterwards, the cells were labelled with fluorophore-conjugated anti-IL-2, anti-CD4 and anti-CD8 antibody and analysed by flow cytometry. This approach is distinct from assays

measuring the amount of secreted IL-2 in the supernatant, which would be assessed in an ELISA. With this type of assay the amount of IL-2 secreting T cells is measured. Measuring the levels of IL-2 secreting cells is a more appropriate approach considering that IL-2 secretion by T cells is a digital response; that is, they either secrete IL-2 or they do not. Scaling IL-2 secretion becomes a matter of triggering more T cells to release IL-2366.

The lymphocyte subset of the splenocyte population was selected according to scatter analysis (Figure 4-8 A-1). From the lymphocytes subset the IL-2 signal was plotted against the CD4 (Figure 4-8 A-2) and CD8 signal (Figure 4-8 A-3), respectively. The subset of IL2+CD4+ cells (blue frame) were expressed as a percentage of all CD4+ cells (purple frame). IL2+CD8+ cells were analysed in the same way.

The resting (not activated) CD4+ T cells of AnxA6-/- cells contained slightly elevated levels of IL-2 secreting cells compared to the wild type but this difference was not statistically significant (Figure 4-8 B). After 4 hours the activated wild type and AnxA6-/- CD4+ T cells both contained significantly more IL-2 secreting T cells than their not activated control. This indicated that the activation with anti-CD3 and anti-CD28 antibody to initiate IL-2 secretion was successful. Activated AnxA6-/- CD4+ T cells did contain significantly more IL- 2 secreting cells than activated CD4+ wild type T cells. Similarly, the not activated CD8+ T cells of wild type and AnxA6-/- mice did not contain significantly different levels of IL-2 secreting cells (Figure 4-8 C). The activated CD8+ wild type T cells did contain significantly more IL-2 secreting T cells than the not activated control. The same could be observed for the activated CD8+ AnxA6-/- T cells. Their level of IL-2 secreting cells was significantly increased after 4 hours. Furthermore, the activated AnxA6-/- CD8+ T cells did contain more IL-2 secreting cells than the activated wild type CD8+ T cells.

The strong increase of IL-2 secreting T cells after 4 hours of activation showed that the activation and stimulation of IL-2 production was successful and is consistent with previous findings362. The fact that activated AnxA6-/- CD4+ and CD8+ T cells did contain more IL-2 secreting T cells than activated CD4+ and CD8+ T cells from wild type mice was surprising, since AnxA6-/- T cells displayed reduced levels of proliferation compared to wild type T cells in vivo. However, this result was not completely unexpected since previous research suggested that lack of AnxA6 increases secretion208,257. Higher levels of IL-2 secreting T cells are likely to lead to an increased concentration of IL-2 in the surrounding medium. This suggests that despite producing similar amounts of IL-2 mRNA AnxA6-/- T cells might be exposed to more IL-2 cytokine into the surrounding medium. Consequently, AnxA6-/- T cells in vivo were probably exposed to higher amounts of IL-2 during the in vivo immune challenge, yet they did not proliferate as much in response to immune challenge.

Figure 4-8: Level of IL-2 secreting wild type and AnxA6-/- T cells. Splenocytes were activated for 3.5 h before being labelled with an IL-2 capture reagent. After 30 minutes cells were labelled with antibodies recognising IL-2, CD4 and CD8 for FACS analysis (A). IL-2 secretion of CD4+ T cells (B) and of CD8+ T cells (C) was calculated separately. Analysis B and C: 2-way ANOVA with Sidak’s multiple comparisons post-test, error=SD, n=6. Asterisks designate significant difference compared to not activated T cells of the same genotype, unless otherwise indicated by bars; **p<0.01, ***p<0.001, ****p<0.0001. Data was pooled from 2 independent experiments conducted in triplicate.

Above results indicate that that TCR related signalling was not affected by the lack of AnxA6. Therefore, a different process involved in T cell proliferation may be responsible for reduced proliferation and AnxA6-/- T cells. Surprisingly, activated AnxA6-/- cells contained more T cells which secreted the mitogenic cytokine IL-2 than activated wild type T cells. This lead to the conclusion that AnxA6-/- T cells were likely to be exposed to higher concentrations of IL-2 during activation events than wild type T cells. Since AnxA6-/- T cells did activate as efficiently as wild type T cells in vitro and were likely to be exposed to increased amounts of IL-2 after T cell activation in vivo, it is perplexing that AnxA6-/- T cells displayed a proliferation defect in vivo. These findings suggested that AnxA6-/- T cells might have impaired IL-2 receptor signalling which prevents them from conveying the signal received from IL-2 stimulation as effectively as wild type T cells.

The most important signalling route for the IL-2 receptor is the phosphorylation of STAT5 via JAK 1 and JAK3. The activated CD122 subunit of the IL-2 receptor initiates the phosphorylation of JAK 1 and JAK 3 and thus activated them. The activated Janus kinases then phosphorylate STAT5a and STAT5b, which form a dimer. This dimer can then pass into the nucleus and initiate proliferation as a transcription factor for mitogenic genes. The magnitude of STAT5 phosphorylation was taken as an indication of signalling activity. This process can be influenced by the availability of IL-2 receptor on the cell surface and the amount of STAT5 that is available for phosphorylation.

To test whether IL-2 signalling was affected by the knock-out of AnxA6, primary T cells were subjected to different concentrations of IL-2 and their response was measured. To analyse IL-2 receptor signalling naïve T cells could not be used as they express few if any functional receptors62. IL-2 receptor expression is triggered by TCR signalling. Once T cells are able to ‘receive’ IL-2 signals IL-2 stimulates the expression of its own receptor66. The intensity of STAT5 signalling is highest 3 days after activating and culturing of primary T cells367. T cells were isolated from wild type and AnxA6-/- mice as before. The isolated T cells were then activated on anti-CD3 and anti-CD28 coated surfaces for 24 hours and cultured for two more days in IL-2 supplemented media.

Prior to measuring STAT5 signalling in cultured primary T cells, it was established whether wild type and AnxA6-/- T cell expressed the same amounts of IL-2 receptor subunit CD122 and STAT5 under the culturing conditions of the signalling assay. To

establish that the expression of STAT5 was similar between wild type and AnxA6-/- T cells lysates of cultured cells were analysed by western blot and probed for AnxA6, STAT5 and β-actin which was used as a loading control (Figure 4-9 A). The western showed that STAT5 was expressed at similar levels in wild type and AnxA6-/- T cells 3 days after being activated and cultured with IL-2.

The amount of available CD122 on the surface of cultured wild type and AnxA6-/- T cells was assessed by flow cytometry. Cultured cells were labelled consecutively with biotinylated anti-CD122 antibody and then with fluorescent streptavidin conjugated AlexaFluor and fluorescent antibodies recognising CD4 and CD8 and analysed by flow cytometry (Figure 4-9 B). First the T cell population was selected according to their position in the FSC and SSC plot (Figure 4-9 B-1). Within this population CD4+ and CD8+ T cells were selected (Figure 4-9 B-2). The CD122 signal of CD4+ and CD8+ T cells (Figure 4-9 B-2) was not very bright but it was specific as the comparison with an IgG isotype control (grey filled histogram) shows. The signal of CD122 was more intense on CD8+ T cells, which was expected since CD8+ T cells are known to express CD122 at higher levels than CD4+ T cells. Comparing the CD122 signal of wild type and AnxA6-/- T cells did not show any marked difference.

Since the analysis of wild type and AnxA6-/- T cells by western blot and flow cytometry did not show any differences in the signals for CD122 and STAT5 it was assumed that wild type, as well as AnxA6-/- T cells, provided the same amount of catalytically active IL-2 receptor subunit CD122 on the cell surface and the same amount of STAT5 for the phosphorylation reaction.

To study the signalling elicited by IL-2, cultured primary T cells of wild type and AnxA6-/- mice were stripped of any IL-2 bound to the outside as previously367,368. Cytokine-receptor interactions were interrupted by incubating the cell suspension in ice cold glycine buffer (pH 4.0). The acidic buffer was neutralised by adding 10 volumes of media to the cell suspension. After resuspending the cells in fresh media wild type and AnxA6-/- T cells were incubated with increasing concentrations of IL-2 (10 fM - 10 nM) for 10 minutes at 37℃. IL-2 receptor activation was stopped by cooling all samples to 4℃. Fixed and permeabilised cells were labelled first with an anti-pSTAT5 (rabbit) antibody and afterwards with fluorescently labelled antibodies recognising rabbit antibody as well as CD4 and CD8 surface markers. Lastly, labelled cells were analysed by flow cytometry. T cells were gated as before (Figure 4-9 B1 and B-2). To analyse the levels of pSTAT5 in CD4+ and CD8+ T cells pSTAT5 signals were plotted as histograms rather than dot plots.

Dot plots are useful for FSC/SSC plots loosely correlating to cell size and when analysing surface markers of a cell population that are expressed at similar levels (or not expressed at all), like CD4 and CD8. Since STAT5 is expressed in all T cells and converted to pSTAT5 due to IL-2 signalling, all cells can be considered pSTAT5+, but the intensity of the pSTAT5 signal changes on a sliding scale. The pSTAT5 signal intensity was therefore determined by measuring the MFI of the histogram. As the name suggests, the MFI is the fluorescence intensity value of the median of the histogram, i.e. at the peak of the curve. As can be seen in Figure 4-9 B-3 the signal for pSTAT5 is lowest when the cells were not re-exposed to IL- 2 (filled blue histogram) while the signal was most intense in the presence of 10 nM IL-2 (blue black line). The peak of the pSTAT5 signal of IL-2 stimulated T cells shifted towards higher fluorescence intensity values when the cells were exposed to increasing concentrations of IL-2. The baseline (MFI value in the absence of IL-2 stimulation) was subtracted from the corresponding MFI values of IL-2 stimulated cells.

Analysis of the pSTAT5 signal of wild type and AnxA6-/- T cells re-exposed to increasing concentrations of IL-2 for 10 minutes showed that the levels of pSTAT5 increased together with the IL-2 concentration in the T cell medium (Figure 4-9 C and D). A sigmoid dose- response curve was fitted to the data points of CD4+ (Figure 4-9 C) and CD8+ (Figure 4-9 D) primary T cells. The dose-response curve of CD4+ wild type and AnxA6-/- T cells followed a similar shape and plateaued at a concentration of 10 nM IL-2. The levels of pSTAT5 were significantly lower in AnxA6-/- T cells exposed to high concentrations of IL-2 (1 nM and 10 nM). Similarly, CD8+ wild type and AnxA6-/- T cells responded to stimulation with increasing amounts of IL-2 with increasing levels of pSTAT5. Both wild type and AnxA6-/- T cells seemed to elicit the maximum pSTAT5 response at a dose 10 nM IL-2. Again, AnxA6-/- T cells produced significantly less amounts of pSTAT5 at the highest IL-2 concentrations used (0.1-10 nM IL-2).

In summary, the analysis of pSTAT5 signalling through the IL-2 receptor showed that activated and cultured wild type and AnxA6-/- expressed comparable levels of the IL-2 receptor subunit CD122 and the signalling protein STAT5. Wild type and AnxA6-/- T cells that were re-exposed to IL-2 for 10 minutes exhibited increased levels of pSTAT5 in response to increasing concentrations of IL-2. The pSTAT5 response of wild type and AnxA6-/- T cells plateaued at 10 nM of IL-2 which indicated that the IL-2 concentrations employed in this assay covered the full range of receptor signalling and reached receptor saturation at the highest IL-2 concentration used. The levels of STAT5 phosphorylation in response to IL-2 stimulation were significantly lower in AnxA6-/- T cells than in wild type

T cells at high IL-2 concentrations (CD4+ T cells 1-10 nM IL-2, CD8+ 0.1-10 nM IL-2). The reduced levels of pSTAT5 of IL-2 stimulated CD4+ and CD8+ AnxA6-/- T cells compared to wild type cells is a strong indication that knock-out of AnxA6 affects IL-2 signalling. Since STAT5 activation is one of the main signalling routes of the IL-2 receptor to promote T cell proliferation345, it is likely that reduced STAT5 signalling negatively affects the proliferation of IL-2 receptor expressing T cells.

Figure 4-9: Analysis of IL-2 receptor signalling in wild type and AnxA6-/- T cells. Cultured wild type (green graphs) and AnxA6-/- (orange graphs) primary T cells were incubated for 10 minutes with 0-10 ng/mL IL-2. A portion of cells was lysed and analysed by western blot (A). The blot was probed for AnxA6, STAT5 and β-actin as a loading control. The remaining cells were analysed by flow cytometry. Examples of the gating strategy are shown in (B) Samples were separately stained for CD122 (B-3) or pSTAT5 (B-4) and the surface markers CD4 and CD8 (B-2). pSTAT5 levels were determined according to the MFI of the pSTAT5+ population. Unstimulated cells (“0 M” IL-2 added) were used as a control (solid histogram) to establish a baseline which was subtracted from all samples. Progressively darker histograms are demonstrating the pSTAT5 signal shift from the lowest (10 fM) to the highest IL-2 (1 nM) concentration tested. A dose-response-curve was plotted for the baseline corrected MFI of CD4+ pSTAT5+ (C) and CD8+ pSTAT5+ (D) cells. Data was analysed by 2-way-ANOVA, error=SD, n=3 One representative expreiment of two triplicate experiments with one mouse per group is shown.

Prior to measuring the pSTAT5 response of IL-2 stimulated T cells it was established that the availability of CD122 subunit on the cell surface and STAT5 in the cytosol was not affected in AnxA6-/- T cells. But it should be noted that the process from IL-2 binding to the

receptor and initiation of STAT5 signalling could also be influenced by other components of the IL-2 receptor and signalling proteins. IL-2 binding induces conformational changes within the transmembrane and cytosolic domains of CD122 as well as the ɣc subunit and brings the two subunits closer together72. In this new conformation JAK1 and 3 which associate with CD122 and ɣc even in the absence of cytokine369,370 are able to phosphorylate each other and CD122. The phosphorylation of CD122 provides a binding site for STAT5370. The activity of JAKs which are responsible for phosphorylation of STAT5 could be affected in AnxA6-/- T cells. Furthermore, the termination of IL-2 signalling could be affected in different ways. If the IL-2 receptor complex was endocytosed faster in AnxA6-/- cells than wild type cells, the signal would be terminated sooner and less pSTAT5 will be produced by one receptor activation event.

Analysis of IL-2 signalling via STAT5 showed that IL-2 receptor signalling efficiency was reduced in IL-2 stimulated AnxA6-/- T cells compared to wild type cells. The rate with which the IL-2-activated receptor is endocytosed could be responsible for this result. AnxA6 has been implicated in clathrin-dependent as well as clathrin-independent endocytosis227,236,287. AnxA6 stimulates LDL-receptor endocytosis227, possibly through the binding and local disruption of the actin/spectrin cytoskeleton287. However, the exact mechanism by which AnxA6 influences endocytic processes is not known. The IL-2 receptor is likely to be internalised through a clathrin-independent route involving F- actin, dynamin, Rac1, Pak1/2 and cholesterol-rich membrane domains371–374372. Since AnxA6 is able to bind to cholesterol-rich membranes as well as dynamin282 in the absence of Ca2+, it is possible that AnxA6 interacts with IL-2 receptor located in cholesterol-rich membrane rafts72,167,182 and that AnxA6 knock-out might affect endocytosis.

IL-2 receptor signalling and endocytosis involves the presence of F-actin and specialised lipid domains surrounding the engaged receptor72,167,182. It is conceivable that knocking out AnxA6 could alter the endocytosis of certain receptors, since AnxA6 has been implicated in endocytic processes212,236,282. In particular AnxA6 over-expression was found to up-regulate LDL receptor endocytosis227. The endocytosis pathway of the IL-2 receptor is not fully understood, but it is known to be clathrin-independent. IL-2 receptor endocytosis has been described to depend on F-actin, dynamin, Rac1, Pak1/2 and cholesterol-rich membrane domains371–374. AnxA6, apart from being able to interact with the cytoskeleton and bind cholesterol-rich membranes was also found to bind the GTPase dynamin independently of Ca2+282.

Here, it was investigated whether the IL-2 receptor is endocytosed at different rates in wild type and AnxA6-/- T cells, be measuring the depletion of labelled IL-2 receptor after endocytosis. Due to the complexity of the IL-2 receptor, different targets were considered for their suitability. The IL-2 receptor consists of three different subunits that are not permanently linked to each other as a trimer. Instead, the subunits are likely to be assembled into a high-affinity receptor complex consisting of CD25, CD122 and ɣc once CD25 is bound to an IL-2 molecule already67. The subunit CD25 is not suitable for labelling, since it is expressed in much larger amounts after T cell activation than the other two subunits; furthermore, CD25 is not directly involved in STAT5 signalling. The signal for this subunit might not decrease significantly in the time that is suitable for IL-2 receptor monitoring (T1/2=15 min)375. Furthermore, unlike CD122 and ɣc CD25 is recycled back to the cell surface376. The ɣc subunit is typically expressed at low levels but it is not specific for the IL-2 receptor as it is involved in six other interleukin-family receptor complexes. Another possible target for labelling is IL-2 itself. However, due to the high expression of CD25, large amounts of IL-2 could be scavenged by this subunit and reduce the dynamic range of the signal. The CD122 subunit was the most suitable and convenient target to label, since it is only shared with the IL-15 receptor, which is not produced by T cells. CD122 is present in all signalling-active receptor complexes and there are monoclonal antibodies (clone 5H4)377 available directed against the extracellular epitope of the receptor. This approach has been successfully employed previously372.

To study receptor endocytosis live T cells were incubated with biotinylated anti-CD122 antibody on ice. The monoclonal antibody (5H4) used in this assay binds CD122 without preventing IL-2 binding377. After removing any excess anti-CD122 antibody the cells were then incubated at 37℃ in cell culture media with and without IL-2 – except for a control sample. The increase in temperature facilitates receptor endocytosis. After 6 minutes, the endocytosis process was stopped by cooling the cells to 4℃. Consequentially, the cells were incubated with streptavidin-conjugated AlexaFlour dye binding the biotinylated CD122 antibody and with fluorophore-conjugated anti-CD4 and anti-CD8 antibodies. Thus, any CD122 that had not been endocytosed and remained on the cell surface and was now fluorescently labelled. The labelled cells were then analysed by flow cytometry. The CD122 signal of T cells incubated for 6 minutes at 37℃ was compared to the CD122 signal of control samples, which were kept at 4℃ continuously. The control therefore contained the maximum amount of labelled CD122. The decrease of the CD122 signal in cells incubated at 37℃ for 6 minutes compared to the control shows the amount of CD122 endocytosed within 6 minutes.

Figure 4-10: Endocytosis of the IL-2 receptor subunit CD122 in wild type and AnxA6-/- T cells. Isolated T cell were activated on antibody-coated surfaces and cultured for 2 days in the presence of IL-2. Before the experiment, the cells were starved of IL-2 for 8 hours. T cells were then labelled with biotinylated anti-CD122 on ice. To induce receptor endocytosis labelled cells were incubated for 6 minutes at 37℃ and u IL-2, while control cells (no endocytosis) were kept on ice continuously. Afterwards, cells were labelled with fluorphore-conjugated strepavidin and fluorescently tagged CD4 and CD8 antibody. (A-1) Lymphocytes and their (A-2) CD4+ and CD8+ subsets were analysed by flow cytometry. The MFI of the CD122 signal was analysed separately for (A-3) CD4+ and (A-4) CD8+ T cells. (A-3 and A-4) Top: The control condition (no endocytosis) is shown for wild type (green line) and AnxA6-/- (orange) cells, grey histgram shows isotype control. (A-3 and A-4) Bottom: The CD122 signal of wild type cells is shown for the control and for a sample, that was incubated for 6 minutes in the presence of IL-2. The solid line shows CD122 signal control condition with no endocytosis. The dotted line shows CD122 signal of sample with 6 minutes of endocytosis with 1nM IL-2 added. Endocytosis (%) as plotted in (B) and (C) was the decrease in CD122 signal after 6 minutes of incubation (with and without IL-2) from the control condiction. CD4+ (B) and CD8+ (C) T cell were analysed separately. Analysis: 2-way ANOVA, error=SD, n=3 from 3 independent experiments with 2-4 replicates per experiment.

To investigate CD122 endocytosis in primary cells T cells have to be activated and cultured to induce the expression of IL-2 receptor. As described above, murine T cells from wild type and AnxA6-/- mice were isolated, activated on antibody-coated surfaces for 24 hours and cultured for two more days in the presence of IL-2. The culturing medium contained less IL-2 (100 U/ml) than in the culturing for STAT5 signalling assay, furthermore, the

cells were starved of IL-2 for 8 hours prior to the assay. These changes to the culturing conditions were made because it was found that the previous conditions did not result in a CD122 expression sufficient for endocytosis analysis by flow cytometry. After IL-2 starvation, T cells were incubated with anti-CD122-biotin on ice for 30 minutes. Any unbound antibody was then removed by washing. To induce receptor internalisation the cells were incubated for 6 minutes at 37℃ and u nM IL-2 before being returned to the ice to stop further endocytosis. Cells of control samples were kept on ice continuously to prevent any receptor internalysation. Surface CD122-biotin was read out with fluorescently labelled streptavidin in conjunction to CD4 and CD8 staining for flow cytometry analysis. Lymphocytes (Figure 4-10 A-1) were selected by scatter analysis and and CD4+ and CD8+ T cell populations were selected (Figure 4-10 A-2). The intensity of CD122 signal of CD4+ (Figure 4-10 A-3) and CD8+ T cells (Figure 4-10 A-4) was analysed in a histogram using MFI as a readout.

Figure 4-10 A-3 and A4 show the signal of CD122 of the control (no incubation at 37℃). In this condition no endocytosis of cell surface receptors occurred and T cells retained all anti-CD122-biotin labelled IL-2 receptor subunits on their surface. The control condition therefore provides the maximum signal for CD122, equivalent to 0% endocytosis. Comparison of the isotype control with the signal of CD4+ and CD8+ wild type and AnxA6- /- for CD122 confirmed the specificity of the anti-CD122 antibody (Figure 4-10 A-3 and A-4 top). Comparing the signal of wild type and AnxA6-/- CD4+ and CD8+ T cells showed that wild type and AnxA6-/- T cells expressed similar levels of CD122. Furthermore, the CD122 signal of CD4+ and CD8+ T cells shows that CD8+ T cells exhibited a much higher signal of the CD122 subunit of the IL-2 receptor. This difference is likely to be a result of the higher expression of CD122 on naïve and memory CD8+ T cells compared to CD4+ T cells62. The bottom graphs of Figure 4-10 A-3 and A-4 provide an example of the difference in CD122 signal strength in CD4+ and CD8+ T cells with no CD122 endocytosis (solid line) and after 6 minutes of incubation at 37℃ (d d l n ). The CD122 signal strength showed that after 6 minutes at 37℃ there was less anti-CD122-biotin labelled CD122 at the cell surface for the streptavidin-AlexaFluor dye to detect than in the control condition. The decrease in CD122 signal after 6 minutes at 37℃ nd ca d a CD 22 c p a ucc fully endocytosed.

The percentage of endocytosed CD122 was measured with and without IL-2 addition during 6 minutes of incubation at 37℃, since IL-2 receptor is endocytosed even in the absence of IL-2 ligand372. Analysis of CD122 internalisation in CD4+ T cells (Figure 4-10 B)

shows that wild type and well as AnxA6-/- T cells did endocytose 28.2% and 35.2%, respectively, of CD122 within 6 minutes in the absence of IL-2 receptor ligand. In the presence of 1 nM IL-2 a 31.4% (wild type) and 34% (AnxA6-/-) of CD122 were internalised by wild type and AnxA6-/- T cells within 6 minutes. In CD4+ wild type T cells the addition of IL-2 lead to a small increase (+3.2) in CD122 internalisation, but this difference was not significant. In AnxA6-/- T cells he percentage of CD122 endocytosis was very similar in the presence and absence of IL-2 during the incubation time. In both conditions, with and without IL-2, AnxA6-/- internalised more CD122 than wild type T cells (+6.9 without IL-2 and +2.6 with IL-2), but these differences were not significant.

In CD8+ T cells (Figure 4-10 C) 39% and 50.8% of CD122 were endocytosed by wild type and AnxA6-/- T cells, respectively, in the absence of IL-2. With the addition of IL-2 during 6 minutes of incubation 43.5% (wild type) and 52.3% (AnxA6-/-) of receptor were internalised by wild type and AnxA6-/- T cells. The added IL-2 slightly increased level of internalised CD122 in wild type and AnxA6-/- T cells but these differences were not significant. In both the presence and the absence of IL-2 AnxA6-/- T cells internalised more CD122 than wild type T cells within 6 minutes, however, these differences were not significant. Generally, AnxA6-/- T cells appeared to endocytose more CD122 than wild type T cells within 6 minutes (+11.8 without IL-2, +8.8 with IL-2) but these differences were not significant.

In summary, 28-35% of CD122 was internalised by CD4+ wild type and AnxA6-/- T cells at the 6 minute time point. CD8+ wild type and AnxA6-/- T cells internalised more (39-52%) CD122 in the same time frame. The addition of IL-2 did slightly increase the level of endocytosed CD122 in CD4+ and CD8+ wild type T cells, but this difference was not significant. The addition of IL-2 did not seem to affect the CD122 of CD4+ and CD8+ AnxA6-/- T cells. Under both conditions (with and without I-2) CD4+ as well as CD8+ T cells AnxA6-/- appeared to internalise more IL-2 receptor within 6 minutes than wild type T cells, but there was no significant difference of IL-2 receptor endocytosis between wild type and AnxA6-/- T cells. The percentage of endocytosed receptor at the 6 minutes time point was higher than expected, since previous publications found that half the IL-2 receptors were internalised after 15 minutes375 and Basquin et al. found that after 6 minutes only 16-20% of CD122 had been endocytosed from the surface of Kit225 cells372.

This chapter investigated whether the underlying reason for the observed proliferation defect in AnxA6-/- T cells in vivo (Chapter 3) was a result of the involvement of AnxA6 in mitogenic pathways of activated T cells. The main mitogenic pathways in T cells are initiated by the TCR and the IL-2 receptor. The signalling of both TCR15,186 and IL-2 receptor72,378 has been found to depend on cholesterol-rich membrane domains and interaction with the actin cytoskeleton. Since AnxA6 is able to bind PS, which was found to be enriched at the immune synapse170, and cholesterol-rich membranes as well cytoskeletal components, it was hypothesised that it could organise membrane domains and support receptor signalling. AnxA6 may target and organise membrane domains in the plasma membrane to assist in forming scaffolds for intricate signalling complexes275, which are triggered by calcium fluxes and directed by transient annexin-membrane-actin interactions202,275,276. In this context TCR as well as IL-2 signalling were analysed with in vitro approaches.

First, the association of AnxA6 with F-actin was examined since T cell activation critically depends on actin restructuring363,379. It was shown with TIRF microscopy that F-actin co- localised with AnxA6 during the early stages of T cell activation, but co-localisation decreased after 10 minutes of TCR activation. Furthermore, superresolution imaging indicated that AnxA6-/- primary T cells have a different molecular organisation of F-actin at the T cell activation site 12 minutes after activation. These results suggested an involvement of AnxA6 in the organisation of the cortical actin cytoskeleton at the immune synapse. F-actin polymerisation and the formation of cytoskeletal structures at the immune synapse are initiated by TCR activation29. At the same time, F-actin is an important regulator of TCR microclusters, which act as signalling platforms15,16,363. Previously, it could be shown that AnxA6 stabilises the cortical actin cytoskeleton in HEK cells, which decreases the release of Ca2+ from intracellular stores380. The lack of AnxA6 as a stabilising factor for F-actin organisation might have implications for TCR microclusters and their signalling behaviour at the immune synapse.

TCR signalling activity and activation was examined by in vitro activation of CTRL and AnxA6KD Jurkat cells, in which the AnxA6 expression was reduced by 50-70%, and primary T cells from wild type and AnxA6-/- mice. Western blot analysis of phosphorylation levels of TCR signalling proteins in CTRL and AnxA6KD Jurkat cells showed that these were similar in both cell lines. It was speculated that if AnxA6 was

involved in TCR signalling residual levels of AnxA6 expression in the knock-down cell line might be sufficient to support T cell activation. Measuring the phosphorylation levels of T cell signalling proteins in activated wild type and AnxA6-/- T cells showed no differences in wild type and AnxA6-/- T cells. This showed that even with the complete abrogation of AnxA6 expression in murine primary T cells, the ability of T cell signalling proteins to initiate and sustain the protein phosphorylation cascade from TCR to ERK was not affected. Ultimately, the TCR signalling cascade alters gene expression in T cells and via ERK and c-Jun and induces the production of the mitogenic cytokine IL-2. As established by qPCR, the mRNA levels of IL-2 in activated AnxA6-/- T cells were not different from those in wild type T cells. These data confirmed that T cell signalling and activation was not affected by the absence of AnxA6. It is therefore not likely that AnxA6 substantially contributes to T cell activation. It is however possible that other annexins expressed in T cells could be rescuing functions of AnxA6 related to TCR signalling in AnxA6-/- T cells. AnxA2 in particular has also been described to bind F-actin and specialised membrane domains242,245,381–383. However, the in vivo immune challenge of wild type and AnxA6-/- mice showed that AnxA6-/- mice exhibit a proliferation defect of CD4+ T cells. This result demonstrated that even if other annexin family proteins share functions of AnxA6 they cannot replace AnxA6 completely.

IL-2 signalling is another potent initiator of T cell proliferation. If this process was affected in AnxA6-/- T cells it could contribute to the T cell proliferation defect of AnxA6-/- mice during an immune challenge. IL-2 gene expression is induced by TCR signalling. The produced IL-2 then needs to be secreted into the extracellular medium before IL-2 signalling can occur. Since AnxA6 has been implicated in secretion it was important to confirm that IL-2 mRNA levels were reflected in IL-2 secretion. Expression and (ectopic) over-expression of AnxA6 has been implicated in negatively regulating the secretion of a diverse array of compounds, e.g. fibronectin and TNFα257, breast milk208,289 and virus particles293,294. Moreover, influenza virus exocytosis was higher in AnxA6 knock-down and cells than in cells expressing AnxA6 at normal levels293,294.

Measuring the amount of IL-2-secreting T cells by flow cytometry showed that a population of activated AnxA6-/- T cells contained more IL-2-secreting T cells than a population of activated wild type T cells. This result was in agreement with previous publications suggesting that the reduction of AnxA6 levels promotes exocytosis208,293,294. A greater abundance of IL-2-secreting T cells has been shown to result in increased IL-2 levels in T cell supernatant362. As a consequence a population of activated AnxA6-/- T cells

is exposed to higher IL-2 concentration than activated wild type T cells. Transferring these in vitro results to in vivo immune response means that T cells of AnxA6-/- mice failed to proliferate at the same rate as wild type T cells despite being exposed to higher levels of IL-2 in their environment. This suggests that AnxA6-/- T cells are not as responsive to IL-2 stimulation as wild type T cells. It is possible that AnxA6-/- T cells (partially) compensate for their reduced sensitivity to IL-2 by increasing the secretion of IL-2.

It is not clear why the elevated levels of IL-2 secreting AnxA6-/- T cells compared wild type cells was not reflected by increased levels of IL-2 mRNA in AnxA6-/- T cells. For the IL-2 secretion assay and qPCR primary cells were activated in vitro for a similar time span (3.5 and 4 hours, respectively). After T cell activation the synthesis of IL-2 mRNA is closely followed by the appearance of IL-2 secreting cells362 and IL-2 is not stored in T cells and kept for delayed release384.

Since AnxA6-/- T cells might be not as receptive to IL-2 stimulation as wild type T cells, the ability of AnxA6-/- T cells to respond to IL-2 stimulation was investigated. Measuring the rate at which STAT5, a main signalling mediator of the IL-2 receptor, was phosphorylated in AnxA6-/- T cells revealed significant differences between AnxA6-/- and wild type T cells. At high IL-2 concentrations, AnxA6-/- cells had significantly lower levels of pSTAT5 than wild type T cells. These data imply that the proliferation defect in vivo may be a result of reduced STAT5 signalling after IL-2 receptor triggering.

IL-2 receptor signalling and endocytosis involves the presence of F-actin and specialised lipid domains surrounding the engaged receptor72,167,182. It is conceivable that knocking out AnxA6 could alter the endocytosis of certain receptors, since AnxA6 has been implicated in endocytic processes212,236,282. In particular AnxA6 over-expression was found to upregulate LDL receptor endocytosis227. The endocytosis pathway of the IL-2 receptor is not fully understood, but it is known to be clathrin-independent. IL-2 receptor endocytosis has been described to depend on F-actin, dynamin, Rac1, Pak1/2 and cholesterol-rich membrane domains371–374. AnxA6, apart from being able to interact with the cytoskeleton and bind cholesterol-rich membranes was also found to bind the GTPase dynamin independently of Ca2+282. Consequently, the endocytosis of the IL-2 receptor subunit CD122 was determined at a fixed time point. While there seemed to be a trend for AnxA6-/- T cells to endocytose up to 11% more CD122 in 6 minutes than wild type T cells, this difference was not significant.

Taken together, these data establish a role for AnxA6 in the proliferation of activated T cells. While AnxA6 did not have any discernable influence on T cell activation itself, the absence of AnxA6 resulted in different F-actin organisation and reduced STAT5 phosphorylation upon IL-2 receptor signalling. Since this signalling pathway is one of the major contributors to the proliferation of T cells, the latter finding might present the reason for the proliferation defect of CD4+ T cells in vivo discussed in the previous chapter.

Interestingly, a peculiarity in the analysis of lymph nodes from AnxA6-/- mice in the in vivo immune challenge suggests that the in vitro findings are indeed relevant in vivo. The analysis of CD8+ T cells in lymph nodes showed that AnxA6-/- mice contained markedly (but not significantly) increased levels of CD8+ central memory T cells compared to wild type CD8+ T cells (Figure 3-2 D, Figure 3-6 F and I). It is possible that the high levels of CD8+ memory cells in AnxA6-/- mice are an indication of ineffective CD122 signalling, since it has been shown that weak CD122 signalling favours the development of CD8+ central memory T cells385.

These results do not provide an explanation for how the signalling process of the IL-2 receptor is affected by the AnxA6 knock-out. However, the affected process can be narrowed down to processes in close proximity to or integral of the T cell membrane. The reduction of STAT5 phosphorylation focused on the first protein in this signalling cascade that fulfils its main function beyond the IL-2 receptor. IL-2 receptor subunits CD122 and ɣc are transmembrane proteins with extracellular domains. IL-2 binding of the extracellular domains pulls the transmembrane and cytosolic domains of CD122 and ɣc closer together72. This enables reciprocal JAK 1 and 3 phosphorylation and phosphorylation of CD122 providing a binding site for STAT5, which can then be phosphorylated by JAK370. IL-2 signalling was measured at the level of STAT5, which was shown to be affected in AnxA6-/- T cells. Therefore the signalling cascade has to be impaired at or upstream of STAT5 phosphorylation. The CD25 subunit of the IL-2 receptor is not directly involved in signalling but supports the formation of a high affinity IL-2 receptor. It has been suggested that CD25 prolongs the time of IL-2 binding and increases therefore extends the time in which JAKs can activate STAT572. IL-2 binding also appears to promote interactions of the IL-2 receptor complex with the actin cytoskeleton, since receptor diffusion is slowed after IL-2 binding, but in cells treated with F-actin interrupting agent cytochalasin D IL-2 binding does not affect receptor diffusion as much72. The interaction of IL-2 receptor with the cytoskeleton is likely to be essential for

IL-2 signalling. T cells of WASP/WIP-/- mice which are not able to co-ordinate cytoskeletal reorganisation in response to TCR stimulation also do not phosphorylate STAT5 in response to IL-2 stimulation378. Furthermore, cytochalasin D treatment reduces STAT5 but not JAK phosphorylation72. F-actin is also required for IL-2 receptor endocytosis. CD122 and ɣc are endocytosed by a clathrin-independent pathway requiring Rac1 and Pak proteins as well as actin and dynamin371–374.

TCR and IL-2 signalling are linked through LCK and possibly engage in cross-talk under certain circumstances386. LCK interacts with TCR ITAMs as well as with the IL-2 receptor subunit CD12276 and AnxA6203. LCK is stimulated during IL-2 receptor signalling and can phosphorylate STAT576,344. In turn, STAT5 signalling can also be induced through the TCR77. However, STAT5 signalling through the TCR only occurs in naïve T cells and IL-2 stimulated T cell blasts386. Since IL-2 signalling alone does not trigger Ca2+ release into the cytosol387, AnxA6 is unlikely to directly participate in this process in the absence of simultaneous TCR stimulation. While there have been reports that imply AnxA6 in Ca2+ independent binding of signalling proteins264,388 and cholesterol-rich membranes225, it is not clear whether AnxA6 could bind CD122-associated LCK or F-actin as well as cholesterol in the absence of Ca2+. This would be necessary for AnxA6 to participate in the coordination of proteins at the plasma membrane. However, the absence of AnxA6 could impact on other processes of the cell that indirectly affect IL-2 signalling. Due to the involvement of AnxA6 in cholesterol homeostasis, it is possible that the membrane composition is affected by the knock-out of AnxA6. This will be tested in the next chapter.

The cell synthesises a vast variety of different lipid species for energy storage, as elements of signalling pathways and, most importantly, amphiphilic lipids to form bilayers providing the cell and subcellular compartments with an essential boundary. Despite efforts to catalogue the entire lipidome389,390 of cells, the reason for the vast variety of lipids and to what extent changes in the lipid composition affect the cell and its function, has not yet been fully elucidated.

The fatty acid residues of phospholipids, the phospholipid class and the cholesterol content of biological membranes influence their fluidity. As discussed in the introduction, long chain saturated lipids, high cholesterol, SM, PE or PS content result in more rigid or ordered membranes compared to membranes with short or cis-unsaturated lipid residues and a high PC content. Furthermore, the composition of cellular membranes can be adapted by the organism to changing conditions, be they temperature, lipid soluble drugs or diet391,392. These changes usually serve the purpose of maintaining the fluidity of the cellular membrane and the process has been termed homeoviscous adaptation. There are many examples of this principle which usually involve adaptation to different temperatures or exposure to certain membrane soluble compounds. E. coli bacteria that are grown overnight at 30℃ av a higher content of monounsaturated lipids than E. coli g n a 43℃392. Similarly, exposure to heat increased the occurrence of saturated lipids that help to reduce fluidity. Sudden temperature changes, however, cannot easily be compensated for. This fact is exploited in the anti-cancer treatment of melanoma. Melanoma cells exposed to a heat shock (43℃) are killed off quite efficiently, which enhances the efficiency of radio or chemotherapy. However, if the heat is not strong enough and/or some cells survive, they can change their lipid composition so that repeated heat treatment loses its effect391. The same effect can be observed after treatment with benzyl alcohol, a compound that increases membrane fluidity391.

Genetic conditions and diet can also influence lipid composition and demonstrate how flexible lipid composition of an organism can be. Diet can also impact on the membrane fluidity and lipid composition of the cells in the body. Dietary lipids and fatty acids are used by the body to assemble its own lipids. Some essential fatty acids and lipid components cannot be synthesised de novo and have to be extracted from food in large amounts (e.g. lineloic acid and choline). While increased intake of cis-polyunsaturated fatty acids like DHA (docosahexaenoic acid, 22:6) has been shown to be beneficial for human health81, the intake of trans-unsaturated fatty acids which the body inadvertently uses in place of saturated fatty acids has been linked (long-term) with coronary disease and shortened life span81. Genetic diseases like Nieman-Pick disease, which leads to an accumulation of free cholesterol and sphingolipids in endosomes, are often devastating for the individual and can result in enlarged liver and spleen, dementia, seizures and death in early childhood393.

Lipid composition does not just vary from cell to cell, there is also heterogeneity within the plasma membrane of an individual cell. On the smallest scale, this is exemplified by lipid rafts which supposedly create small dynamic domains with distinct protein and lipid composition within cellular membranes132. On a larger scale, in polarised cells, plasma membrane composition can vary along the axis of polarisation133,170. Some cells, like gut epithelia, are permanently polarised. The basolateral membrane of rat intestinal cells contains a different lipid make-up from the brush border membrane, which was found to be enriched in cholesterol and SM133. Other cells, like T cells, can be stimulated to polarise upon activation. T cells polarise when their activation is triggered by a cognate pMHC. During activation the immune synapse is formed which plays an important role in regulating T cell activation. The immune synapse is supported by a dynamic F-actin cytoskeleton supporting a retrograde movement of TCR microclusters176. The plasma membrane lipid composition of the immune synapse resembles rafts in their lipid composition and is enriched in cholesterol, SM and saturated fatty acid residues170. It also resembles the predicted lipid rafts in biophysical features like increased membrane order compared to the average membrane order of the plasma membrane171.

Dietary polyunsaturated fatty acids incorporated into the plasma membrane of immune cells are known to reduce membrane order and have been shown to reduce the overall activity of T cells394, but also exert some influence on macrophages and B cells82. This immune suppressive effect has proved beneficial for patients suffering from chronic inflammatory diseases395. The most common polyunsaturated fatty acids investigated in

this context are DHA, EPA (eicosapentaenoic acid, 20:5) and arachidonic acid. In T cells, the increased level of DHA or EPA lead to reduced T cell proliferation394 and IL-2 production by modulating TCR signalling in several ways: their presence in the plasma membrane can modify early signal transduction events within the plasma membrane and inhibit specific isoforms of PKC82,396. Furthermore EPA, DHA, and arachidonic acid reduce membrane order and T cell signalling165,170. The change of membrane order is of particular interest since the immune synapse in activated T cells displays an increased degree of plasma membrane order171. Disrupting this order by cholesterol removal, enriching the plasma membrane in 7-ketocholesterol183,186 or by incorporation of polyunsaturated fatty acids165,195 can affect the activation and signalling behaviour of T cells as well as immune synapse formation. DHA enrichment in T cells can also reduce IL-2 receptor signalling through STAT5166. Many, but not all, of the immunosuppressive effects of polyunsaturated fatty acids are suspected to result from their effect of reducing membrane order and inducing changes to the lipid rafts in the plasma membrane82,397,398.

In vivo not all types of unsaturated fatty acids are considered equally beneficial. Unsaturated fatty acids are grouped by the relative position of their last double bond in the carbon chain; DHA and EPA are part of the group of n-3 fatty acids while arachidonic acid is an n-6 fatty acid. The n-3 unsaturated fatty acids EPA and DHA may not only downregulate T cell signalling by modulating the fluidity of the plasma membrane, but also by modulating the eicosanoid metabolism. EPA and DHA exert anti-inflammatory effects since they can displace n-6 polyunsaturated fatty acids, like arachidonic acid, in the plasma membrane which are converted into more powerful proinflammatory compounds396,399,400. Making the role of polyunsaturated fatty acids in inflammation even more complex, DHA and EPA as well as arachidonic acid can downregulate surface MHC I and II molecules on the surface of various immune cells (macrophages, B cells, T cells, DCs)82; the over-expression of MHC receptors can lead to immune disorders401.

Apart from phospholipids and their fatty acid moieties, cholesterol is an important component of cellular membranes regulating fluidity and permeability – particularly of the plasma membrane. In T cells, cholesterol levels have important consequences for T cell signalling. As mentioned previously, cholesterol depletion or cholesterol substitution with a 7KC analogue reduced plasma membrane order, T cell signalling and proliferation. While cholesterol depletion negatively regulates T cell activation, increased cholesterol levels have the opposite effect, which could be shown in ɣδ T cells402 as well as the more common αβ T cells183,402–404. ɣδ T cells are a small (5% of total T cells) CD3+ but CD4- and

CD8- subset of T cells. They naturally contain higher amounts of cholesterol and display increased levels of TCR signalling and proliferative activity compared to αβ T cells402. Artificial cholesterol enrichment increased (αβ)T cell signalling402,405, fostered the development of self-reactive T cells and increased STAT5 phosphorylation through the IL- 2 receptor in mice in vivo405. Furthermore, IL-2 receptors locate to membrane rafts, which are enriched in cholesterol. The disruption of membrane rafts by cholesterol sequestration has been shown to reduce IL-2 signalling167,181,182. Once the IL-2 receptor is activated by IL- 2 binding, a conformational change is triggered in the CD122 and ɣc IL-2 receptor subunits. This conformational change can promote the association of the receptor complex with membrane rafts promoting downstream signalling72.

One of the ways in which increased membrane fluidity could contribute to changes in receptor signalling is a change of the orientation or clustering of transmembrane domains.

Fluid membranes and LD phases in bilayers are thinner than LO phases406. If hydrophobic transmembrane domain is in a bilayer that is thinner than usual it might tilt to reduce its effective length and avoid exposure of hydrophobic domains to a polar environment407,408. Alternatively, it might engage in self-aggregation or change its conformation407.

Examples of how cells react to temperature changes or the presence of lipid soluble toxins with changes in lipid composition show that the lipid composition of cellular membranes is tightly regulated by the cell. The fluidity of these membranes in particular is an important trait, since the membrane acts as a medium for proteins that need to interact with each other and be manipulated through vesicle fusion and fission at a certain rate. This study chose two different approaches to gain insight into the lipid composition of AnxA6-/- and wild type T cells and the basic biophysical property of membrane order or membrane fluidity. The membrane order of the plasma membrane was investigated by Laurdan imaging while the fatty acid and lipid class composition of whole T cells were analysed by mass spectrometry.

The fluorescence emission spectrum of the lipid soluble dye Laurdan (Figure 5-1 A) is sensitive to changes in the polarity of the surrounding medium409. Due to solvent relaxation effects, the Laurdan intensity is severely reduced with increasing solvent

polarity, and as a result being virtually non-fluorescent in water. The presence of water as a very polar substance shifts the emission spectrum to higher wavelengths while the presence of apolar compounds, i.e. lipids, shifts the emission spectrum to lower the wavelengths (Figure 5-1 B)410,411. In highly ordered Lo membranes where saturated fatty acid residues in the lipid bilayer and cholesterol pack together very tightly, the presence of water is excluded from the bilayer very efficiently. In fluid Ld membranes (poly)unsaturated fatty acid residues aggregate less densely with neighbouring lipids and leave more room for water molecules to penetrate into the lipid bilayer and alter the emission spectrum of Laurdan in the bilayer (Figure 5-1 C)412,413. With a confocal microscope the channels receiving Lo (channel 1) and Ld (channel 2) portion of the Laurdan signal are acquired simultaneously. From grey scale images of both channels, the GP values are calculated (Figure 5-1 D) and serve as a measure of membrane order. The GP can range from +1 to -1; the lower the GP value the less ordered is the observed membrane region140.

Figure 5-1: Principle of Laurdan imaging. (A) The structure of the Laurdan molecule. (B) Depending on the hydrophobicity of the surrounding molecules Laurdan emits fluorescence signals in different wavelengths. In a hydrophobic LO environment the Laurdan signal is detected mainly in Channel 2. The presence of a polar substance, like water, in an Ld environment leads to a blue shift in the Laurdan emission spectrum, which is acquired in Channel 1. (C) In a lipid bilayer made up of Ld phases containing many unsaturated fatty acid residues and Lo phases enriched in cholesterol (purple discs) and saturated lipids, water molecules (blue) can be excluded more efficiently from the Lo than the Ld phase. (D) The GP value is used as a measure for membrane order and is calculated from the intensity signal (I) of the grey scale image acquired in channel 1 (400-450nm) and channel 2 (470-530 nm). Graph in (B) was adapted from Owen et al. 2012140.

Changes in the membranes order of the T cell plasma membrane could influence the signalling behaviour of the TCR as well as the IL-2 receptor and other membrane-resident receptors82,165,405. Initially, the membrane order of T cells was analysed in a model cell line, afterwards the same method was applied to analyse primary T cells from AnxA6-/- and wild type mice.

It was investigated whether the membrane order of AnxA6KD Jurkat cells was different from the control cell line (CTRL). For this, Laurdan stained cells were mixed with anti-CD3

and anti-CD28 coated polystyrene beads and left to activate for 12 minutes at 37℃. Fixed and mounted cells were imaged with 2-photon microscopy. The images were analysed with ImageJ and grey scale GP images and pseudocoloured images were generated with the GP image analysis plugin (Figure 5-2 A, B). Plasma membrane regions in contact with the bead (“PM on bead”) and the distal plasma membrane of the same cell (“PM free”) were manually selected in the pseudocoloured images and the GP values were measured in the corresponding grey scale GP image. The GP values of CTRL and AnxA6 knock-down samples images on the same day were expressed as ΔGP values compared to the plasma membrane regions of CTRL cells in contact with the bead (relative GP=0) (Figure 5-2 C). ΔGP values below zero therefore indicate that a region of interest was less ordered than the reference region (CTRL cells “PM on bead”).

The average GP value was highest in the plasma membrane region of CTRL cells in contact with the activating bead surface (GP=0); all other GP values are expressed as the difference to this region (Figure 5-2 C). Negative ΔGP values therefore mean that the measured GP values of a particular region were lower than the GP values of the reference region. The GP values of the bead-bound membrane region (“PM on bead”) of CTRL cells were on average significantly higher than the GP values of the free plasma membrane (“PM free”) in CTRL cells. Likewise, the plasma membrane regions of AnxA6 KD cells in contact with the bead had significantly higher GP values than the free plasma membrane regions of AnxA6KD cells. Therefore, in both cell lines, the plasma membrane regions attached to the bead were more ordered with more densely packed lipids than the corresponding free plasma membrane. This indicates that the cell populations analysed were indeed activated by the antibody-coated bead.

The GP values of plasma membrane regions in contact with the bead were lower in AnxA6 KD T cells than in CTRL T cells, but the difference was not significant.

It was therefore concluded that the AnxA6 knock-down did not affect the ability of activated T cells to form an immune synapse with a Lo phase plasma membrane at the activation site.

Figure 5-2: Laurdan imaging and analysis of plasma membrane order of CTRL and AnxA6 knock-down Jurkat cells bound to antibody-coated beads. CTRL Jurkat cells (A) and AnxA6 knock-down Jurkat cells (B) were stained with Laurdan, activated for 12 minutes on antibody- coated beads, fixed, imaged and intensity images were converted into grey-scale and pseudo- coloured GP images. The GP values of the plasma membrane (PM) at the T cell activation site (‘on bead’) was measured as well as the GP values of the ‘free’ plasma membrane. The ΔGP values plotted show the difference in GP value to the “PM on bead” region of wild type T cells (C). An example of the analysed regions is given in the inset images in (A) and (B) as indicated by the red regions in the inserts. Data analysed with Student’s t-Test. “PM on bead” and “PM (free)” of the same cell were compared with a paired t-Test, “PM on bead” of CTRL and AnxA6 KD cells were compared with an unpaired t-Test, (A)-(C) error=SEM, n=51 (CTRL), n=49 (AnxA6 KD) cells pooled from two experiments; ****p<0.0001. The white scale bar in B indicates 5 µm; A and B have the same magnification. (PM=plasma membrane)

The AnxA6 knock-down cell line still expresses ~30% of the original levels of AnxA6. As a result, defects by the lack of AnxA6 might not become apparent. Therefore, the same analysis was conducted with primary T cells isolated from wild type (Figure 5-3 A) and AnxA6-/- (Figure 5-3 B) mice. The T cells were prepared as before and the acquired images were analysed in the same fashion as the Jurkat cell data.

The average GP value of the plasma membrane region of wild type cells attached to the bead surface was used as a reference for GP values of wild type and AnxA6-/- samples acquired on the same day to generate ΔGP. This region had the highest average GP values (relative GP=0) (Figure 5-3 C). As before, GP values of plasma membrane regions with and without contact to the surface of an activating bead were compared. The results showed that GP values of “PM on bead” were significantly higher than the GP values of the corresponding regions of free plasma membrane, “PM (free)” (Figure 5-3 C) in wild type T cells as well as AnxA6-/- T cells. Comparing the plasma membrane regions of wild type cells in contact with the bead with the equivalent regions of AnxA6-/- T cells revealed that AnxA6-/- T cells had significantly lower GP values at the activation site.

It can therefore be concluded that the membrane order of activated T cells is affected by a complete knock-out of AnxA6 but not by a knock-down of AnxA6 with residual expression of AnxA6.

Figure 5-3: Laurdan imaging and analysis of plasma membrane order of primary wild type and AnxA6-/- T cells bound to antibody-coated beads. T cells isolated from spleens of wild type (A) and AnxA6-/- mice (B) were stained with Laurdan, activated for 12 minutes on antibody coated beads, fixed, imaged and intensity images were converted into grey-scale and pseudo-coloured GP images. The GP values of the plasma membrane (PM) at the T cell activation site (‘on bead’) was measured as well as the GP values of the ‘free’ plasma membrane. The ΔGP values plotted show the difference to the “PM on bead” region of wild type T cells (C). The dotted line indicates 0. An example of the analysed regions is given in the inset images in (A) and (B) as indicated by the red regions in the inserts. The lipids of homogenised whole T cells from 2 mice per group were extracted and analysed for cholesterol content. Data analysed with Student’s t-Test. “PM on bead” and “PM (free)” of the same cell were compared with a paired t-Test, “PM on bead” of wild type and AnxA6-/- T cells were compared with an unpaired t-Test, (A)-(C) error=SEM, n=70 (wild type), n=56 (AnxA6-/-) cells pooled from three experiments; ****p<0.0001. The white scale bar in B indicates 5 µm; A and B have the same scale. (PM=plasma membrane)

To investigate whether plasma membrane order was affected in naïve T cells as well, GP values were measured in wild type (Figure 5-4 A) and AnxA6-/- (Figure 5-4 B) T cells that were prepared and analysed in the same way as before, but were not activated on antibody-coated beads. The GP values of samples acquired on the same day were plotted as the difference to the average GP value of the plasma membrane of wild type T cells. It was found that the GP values of the plasma membrane of naïve T cells from AnxA6-/- mice were significantly lower than of wild type T cells.

Figure 5-4: Laurdan imaging and analysis of plasma membrane order of not activated primary wild type and AnxA6-/- T cells. T cells isolated from spleens of wild type (A) and AnxA6-/- mice (B) were stained with Laurdan, fixed, imaged and intensity images were converted into grey-scale GP images and a pseudo-coloured scale. The GP values of the plasma membrane were selected and measured as indicated by the red regions in the inserts (C). The ΔGP values of samples are shown as the difference to the GP value of the plasma membrane of wild type T cells Data were pooled from four independent experiments with one mouse per group. Analysis: Student’s t-Test, n=75 (wild type), n=73 (AnxA6-/-), error = SEM, ****p<0.0001. The white scale bar in B indicates 5 µm; A and B have the same scale. (PM=plasma membrane)

In conclusion, the membrane order of plasma membrane regions bound to antibody- coated beads was significantly higher than the free plasma membrane region in Jurkat cells as well as primary T cells. The plasma membrane at the T cell activation site of CTRL and AnxA6 knock-down as well as murine wild type and AnxA6-/- T cells displayed more ordered membranes compared than the free plasma membrane. A decrease in AnxA6 expression or a complete knock-out of AnxA6 does not affect the ability of the T cell to to initiate a polarisation process that results in more ordered membrane at the immune synapse.

Comparing the GP values of the T cell activation sites of CTRL with AnxA6KD showed that membrane order at the T cell activation site of Jurkat T cells was not significantly different. In primary T cells, however, the membrane order at the activation site of AnxA6-/- T cells was significantly reduced in comparison the activation site of wild type T cells. But the activation site was not the only part of the plasma membrane where membrane order was reduced in AnxA6-/- T cells. Even in naïve T cells the plasma membrane order was lower in AnxA6-/- than in wild type T cells. This showed that the plasma membrane order was generally lower in AnxA6-/- T cells than in wild type T cells.

Surprisingly, the reduced membrane order in activated and untreated T cells did not correlate with reduced TCR signalling presented in Chapter 4, as reported previously186. However, antibody-activated AnxA6-/- T cells did retain a significant difference in the degree of membrane order between the plasma membrane region at the bead and the free plasma membrane. This showed that, in AnxA6-/- T cells, the overall membrane order is affected rather than the process that induces the condensation of the plasma membrane at

the T cell activation site. The observation that AnxA6-/- T cells do activate through the TCR normally, despite decreased membrane order, could indicate that polarisation of the T cell membrane (higher level of membrane order at the immune synapse than at the distal region) is more important for T cell activation than the actual GP value.

Since the plasma membrane order is largely a result of the different lipid classes and fatty acid moieties, the global change of membrane order is likely to be a consequence of altered lipid composition of the plasma membrane.

The fatty acid, as well as the lipid class composition of cellular membranes, influences the fluidity and permeability of the bilayer and greatly impacts on its function. Fatty acids can modulate membrane fluidity and permeability through their length and number of double bonds in the carbon chain5. The fact that the phospholipid composition impacts on membrane functions is reflected in the asymmetric distribution of lipid classes between bilayer leaflets, different composition between poles of polarised cells, and the variation of lipid class abundance among membranes of different cell types. It was therefore of interest to analyse the occurrence of different lipid classes and the length and degree of saturation of the corresponding fatty acids. The different lipid classes (e.g. PC, PE, and PS) are defined by their polar head groups, which are facing the cytosol, lumen of an organelle or the exterior of the cell. Their presence in cellular membranes can provide a target for membrane binding proteins, like annexin proteins and proteins with PI-phosphate- binding motifs102,414. The presence of certain lipids can also affect protein folding and function inside the membrane415,416. Furthermore, the lipid class ratio can impact physical aspects of the lipid bilayer as the lipid class can affect membrane fluidity and charges directly through the nature of the head group and indirectly, by fatty acid distribution since certain lipid classes are more likely to carry saturated fatty acids than others (e.g. SM417).

Due to the fatty acid synthesis pathways in the cell, only certain configurations of fatty acid chains and double bonds are commonly found in higher animals. Fatty acids found in mammals and most other metazoan organisms are usually even-numbered which is a result of the enzymatic reactions during fatty acid synthesis418. Acetyl-CoA (C2) is used to elongate the chain by two carbon atoms for every cycle of the fatty acid synthase extension

reaction419. Odd-numbered fatty acids are primarily produced by microorganisms, like fungi and bacteria. Odd-chain fatty acids usually appear in trace amounts in mammalian cells, but are detected at higher levels in ruminant animals (5%) and are presumably absorbed from lipids present in the diet or gut flora420.

Generally, myristic acid (14 carbon atoms or “C”) is the shortest chain length commonly found in phospholipids. On the other end of the scale, 22C is the longest chain length commonly found in most phospholipids, but 24C fatty acids can be found in PC and PE418,421. While the cell is capable of generating much longer carbon chains for the synthesis of cholesterol, the pathway and participating enzymes are completely separate from the synthesis of fatty acids. When the body generates unsaturated fatty acids from a saturated equivalent, the reduction of carbon-carbon bonds to introduce double bonds is usually restricted to positions up to C9. Multiple double bonds in one chain are separated by at least one methyl residue418,422. While mammalians lack the desaturases to introduce double bonds at certain positions, these fatty acids can still be abundant in the cell but are typically derived from dietary lipids found in eggs, plant oils or fish. These fatty acids are “essential” as they cannot be produced by the organism itself. However, once they enter the cellular metabolism they can be extended and further modified by the cell. This study only analysed the length and desaturation of fatty acids that are known to be abundant in mammalian phospholipids. Phospholipids containing fatty acids with odd-numbered, very short (<14C) and very long (>24C) chain fatty acids were not included in the analysis. Where it was not possible to assess the fatty acid composition (lipids analysed in positive ion mode), only lipids that generated a signal in 75% of all sample replicates were included in the analysis.

Table 5-1: Examples fatty acids found in mammals and their common names

Name Composition Position of double bonds Myristic acid 14:0 Palmitic acid 16:0 Palmitoleic acid 16:1 cis-Δ9 Stearic acid 18:0 Oleic acid 18:1 cis-Δ9 Linoleic acid 18:2 cis,cis-Δ9,Δ12 Arachidic acid 20:0 Arachidonic acid 20:4 cis,cis,cis,cis-Δ5Δ8,Δ11,Δ14 Eicasopentanoic acid (EPA) 20:5 Behenic acid 22:0 Docosahexaenoic acid (DHA) 22:6 cis,cis,cis,cis,cis,cis-Δ4,Δ7,Δ10,Δ13,Δ16,Δ19

Several fatty acids have common as well as systematic names. Common names like arachidonic acid reveal nothing about the chemical structure of the compound, while systematic names, like (5Z,8Z,11Z,14Z)-5,8,11,14-Eicosatetraenoic acid, can become long and convoluted. Therefore shorthand notations are used. Arachidonic acid can be written as 20:4 – as a fatty acid of 20 carbon atom chain length with four C=C double bonds. A phosphatidylcholine containing oleic acid (18:1) as well as palmitic acid (16:0) residues can be expressed as “PC(18:1/16:0)” or “PC 34:1” in a summary notation, which encompasses all isobaric PC lipid species (Figure 5-5). Isobaric lipids contain the same total number of carbon atoms and double bonds in the fatty acid moieties but are not necessarily made up of the same fatty acid species: PC(16:1/18:0) is an isobaric species to PC(16:0/18:1).

Figure 5-5: Isomers and isobars of PC 34:1. Positive ion analysis for PC and SM does not allow the analysis of fatty acid fragments. As a result only information of the total amount of carbon atoms and double bonds is available. Different lipids can contribute to the same signal. Lipids with the same fatty acid residues in different sn-positions are called isomers while lipids with different fatty acids but the same total amount of carbon atoms and double bonds are called isobars. This figure was adapted from Ejsing 20071.

The mass spectrometer used for phospholipid analysis was a triple quadrupole spectrometer (QTRAP 5500). The sample is dispersed through a capillary in a strong electrostatic field, which holds back ions of opposite charge and only lets ions of the same charge pass through. In positive ion analysis mode, a positive charged field holds back negatively charged ions while positively charged ions leave the capillary as micrometre sized droplets. During this process of electrospray ionisation (ESI), the solvent evaporates and the droplets break up into ever smaller units due to electrostatic repulsion. Eventually, positively charged intact molecules, also known as precursor ions, pass into

the first quadrupole (Q1) and are then accelerated towards the collision cell (Q2) filled with an inert gas. The collision of a gas particle with the lipid precursor ion breaks the lipid molecule up into smaller fragments (fragment or product ions) – a process termed collision-induced dissociation (CID). Depending on the energy of the collision and analysis type (positive or negative), different fragment sizes are generated which pass through to quadrupole 3 (Q3) and are finally registered by a detector counting the detected charges.

Figure 5-6: Schematic of triple quadrupole mass spectrometer with an ESI ion source. (A) A sample with ions of mixed polarity is dispersed through capillary nozzle in a strong electrostatic field. Here, analysis in positive ion mode is illustrated: the capillary walls are positively charged and bind negatively charged ions so only positively charged ions pass through. Micrometre-size droplets shrink as the solvent evaporates and the lipid molecules separate from each other as they are all positively charged and repel each other. (B) The charged molecules or precursor ions of a specific mass-to-charge ratio pass through quadrupole Q1 and into the collision cell. There the collision with inert gas molecules (nitrogen or argon) breaks the precursor ions up into at least two fragments. Charged fragments or product ions matching the selectivity of the quadrupole Q3 are counted by the detector.

Information about the molecular mass is resolved by the quadrupole units. They can selectively filter out ions of a specific mass-to-charge ratio (m/z). Therefore, detecting ions with a single positive or negative charge means that m/z is equivalent to the molecular mass of the intact molecule or fragment in u. If Q1 is set to let molecules of 760.5 m/z pass through and Q3 excludes everything but 184 m/z ion fragments, the detector only receives the signal of lipids with this signature – in this example “PC 34:1”. Quadrupole 2 is not used for mass selection since it contains the collision cell.

The intact PC 34:1 lipid has a mass of 759.5 u, but the positively charged molecule has attracted an extra H+ ion and is 1 u heavier. The choline fragment ion leaving the collision

cell, i.e. the polar head group of PC, has a mass of 184 m/z. To analyse several lipids, Q1 remains at each m/z only for a few seconds and scans a range of 500-1000 m/z while Q3 is “parked” to filter out fragment ions of a specific m/z. As Q1 scans through a mass range of precursor ions, this type of setup is called a precursor ion scan. In a neutral loss scan, Q1 and Q3 both change m/z selectivity but the selectivity of Q3 is offset by the mass of a neutral molecule (e.g. head group fragment phosphoethanolamine) that is expected to be lost from a charged fragment during the collision. If Q1 transmits 782.5 m/z ions in a neutral loss scan for PS, the phosphoserine fragment (141 u) known to be generated in the collision cell cannot be detected, only the positively charged diacyl product ion (found at 641.5 m/z). The mass of the product ion changes with the mass of the precursor ion.

The analysis type, i.e. positive or negative ion mode, reveals different details about the lipid molecules. Positive ion analysis only yields information about the mass of the intact molecule (through Q1) and the positively charged fragment leaving the collision cell (through Q3), i.e. the polar head group in a precursor ion scan or a positively charged diacyl fragment in a neutral loss scan. In this analysis mode, the fragmentation and detection is very efficient for PC and SM. The detected ion counts are reliable enough to normalise the sample to a matching internal standard and extract information about the concentration of lipids with the same mass. Therefore, it is possible to know the amount of “PC 34:1” in a sample but not whether this molecule contains a 18:1 and 16:0 fatty acid or one of the other possible combinations: 14:0 and 20:1, 16:1 and 18:0. These details can be detected in negative ion mode. Here, the negatively charged fragments generated in the collision cell can be analysed, i.e. negatively charged head groups (PC-acetate, PE, PS, PI, PG and PA) and fatty acids. The different fragments (head group, fatty acid of sn-1 and fatty acid of sn-2) are not all generated with the same efficiency – fatty acid 1 might “detach” more readily from the precursor ion than fatty acid 2.

Since PC and SM are more likely to form positively rather than negatively charged precursor ions and are less abundant in the negative ion mode, it makes negative ion mode analysis less suitable for quantification for these lipid classes. PI could not be reliably quantified due to the lack of a suitable internal standard, but the percentage of detected fatty acids belonging to the PI lipid class could be analysed. Ether phospholipids could not be reliably analysed in negative mode and are therefore only reported in their sum formula (e.g. PE O-40:0). The fatty acid residue of SMs can be calculated from the molecular composition of the lipids detected in positive mode. Since it is very rare that the sphingosine back bone deviates from the structure given in Figure 1-5 A, an 18C carbon

chain with one trans-double bond can be subtracted from the molecular composition. SM 34:1;2 is equivalent to SM(d18:1/16:0).

For mass spectrometric analysis, the lipids that are to be analysed have to be completely separate from other cellular compounds like nucleic acids, peptides, and other small metabolites. Lipids were extracted from whole T cell lysates with a mixture of methanol and a hydrophobic organic solvent like chloroform or MTBE. Lipids transition into the hydrophobic phase of a solvent mixture while proteins, sugars, DNA and RNA remain in the aqueous phase. Cholesterol and phospholipid analysis were performed separately on different samples. For cholesterol analysis the lipid phase was extracted with chloroform while phospholipids were extracted with MTBE. BHT was added to the mix to prevent degradation of lipids by oxidation. Lipids were extracted under agitation overnight at 4℃. On the next day, aqueous ammonium acetate solution was added, and, after 15 minutes of rigorous mixing, organic and aqueous phase were separated by centrifugation. MTBE extraction is a more recently established extraction method423 with the advantage over the more traditional extraction with chloroform424 that the lipid-containing organic layer forms at the top of the extraction mix. Therefore, the risk of contamination with substances from the aqueous layer (e.g. protein, DNA) is reduced. Before the sample was used for mass spectrometric analysis, it was diluted (10-20 times) in chloroform:methanol (2:1) with ammonium acetate. Internal lipid standards with non-physiological phospholipids were added to the sample prior to extraction to facilitate normalisation and quantification of the detected lipid class.

All phospholipid classes were analysed from the same original sample, but they were measured in different analysis modes. PC, PC-O and SM were analysed in positive ion mode and were quantified with the help of the same internal standard (PC19:0/19:0). PE, PS, PG and PA were analysed in negative mode and quantified with the help of the respective 17:0 fatty acid internal standards. PE-O was analysed in positive mode with a neutral loss scan, since the negative mode settings were not optimised to detect ether lipid species. PE-O 17:0/17:0 was the internal standard used for PE-O.

The analysis of AnxA6-/- and wild type T cells showed that PC was the most common phospholipid (~40%), followed by PE (~21%) and PS (~18%). PC-O and SM made up approximately 7-8% of the quantified phospholipids. PE-O, PG and PA made up less than 3%, less than 2% and less than 1%, respectively. Generally, the ratios of the analysed lipid

classes were very similar in wild type and AnxA6-/- T cells. PC and PS levels, however, were the notable exception: AnxA6-/- T cells contained significantly higher levels of PC (38.1% in wild type and 42.6% in AnxA6-/-) and significantly lower levels of PS (20.7% in wild type and 16.0% in AnxA6-/-).

Figure 5-7: Analysis of the distribution of phospholipid classes in lysates of primary T cells from AnxA6-/- (orange) and wild type mice (green). The quantified amounts of total lipids of one class were expressed as percentages of the total molar amount of phospholipids. The results were compared by 2-way ANOVA with Sidak’s multiple comparisons post-test, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3, *p<0.05.

The negative ion analysis of PA showed that AnxA6-/- T cell contained more short fatty acids (14C and 16C) and less long chain fatty acids (20C and 22C) than wild type T cells (Table 5-2 A). However, the variability between samples was too high for any of these differences to be significant. When the number of double bonds found in the fatty acid moieties were analysed, it was found that AnxA6-/- T cells contained significantly more unsaturated and significantly less six-fold unsaturated fatty acids (Table 5-2 B). The analysis of specific fatty acid species showed that all six-fold unsaturated fatty acids are available as DHA (22:6) (Table 5-2 C). Most PA unsaturated fatty acids found in AnxA6-/-T cells were either 16:0 or 18:0 (Table 5-2 C). Wild type T cells contained more 20:0 and 22:6, but less 16:0 than T cells from AnxA6-/- mice. Due to large variability, these differences were not significant.

Table 5-2: Analysis of fatty acids in PA lipids in lysates of primary T cells from AnxA6-/- and wild type mice. The abundance of fatty acids of a certain chain length (A), double bonds (B) and specific fatty acid species (C) was analysed by mass spectrometry. The molar quantities of lipid were calculated from the signal of the internal standard PA(17:0/17:0). The percent composition of fatty acid chain lengths, double bonds and fatty acid species was calculated for each individual sample from the quantified amount. The results were compared by 2-way ANOVA with Sidak’s multiple comparison post-test, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3, ****p<0.0001. FA = fatty acid.

A Carbon chain length B Double bonds in carbon chain

wild type AnxA6-/- wild type AnxA6-/- % SD % SD % SD % SD 14 1.23 ± 1.78 3.19 ± 3.55 ****0 74.25 ± 5.10 95.05 ± 2.31 16 12.80 ± 12.11 41.52 ± 25.34 1 6.64 ± 3.38 3.17 ± 2.74 18 43.92 ± 36.16 46.25 ± 36.73 3 0.85 ± 1.14 0.00 ± 0.00 20 23.79 ± 25.07 7.25 ± 10.06 ****6 18.25 ± 5.29 1.79 ± 3.09 22 18.25 ± 5.29 1.79 ± 3.09

C Fatty acids

wild type AnxA6 -/- % SD % SD FA 14:0 1.23 ± 1.78 3.19 ± 3.55 FA 16:0 12.80 ± 12.11 41.52 ± 25.34 FA 18:0 37.28 ± 35.29 43.08 ± 39.04 FA 18:1 6.64 ± 3.38 3.17 ± 2.74 FA 20:0 22.94 ± 25.57 7.25 ± 10.06 FA 20:3 0.85 ± 1.14 0.00 ± 0.00 FA 22:6 18.25 ± 5.29 1.79 ± 3.09

Analysing the fatty acid chain length of PG lipids in AnxA6-/- and wild type T cells in negative ion mode showed that AnxA6-/- cells contained less long chain fatty acids (20C and 22C), but significantly more 18C fatty acids (Table 5-3 A). PG fatty acid moieties were less often saturated and more often mono-unsaturated than wild type T cells (Table 5-3 B). These differences were not significant. The analysis of specific fatty acid species showed that the main reason for increased levels of 18C and mono-unsaturated fatty acids in AnxA6-/- T cells was the significantly greater abundance of 18:1 fatty acids (Table 5-3 C) in T cells of that genotype. AnxA6-/- T cells also contained less 20:0, 20:3, 22:0 and less 14:0, but these differences were not significant.

Table 5-3: Analysis of fatty acids in PG lipids in lysates of primary T cells from AnxA6-/- and wild type mice. The abundance of fatty acids of a certain chain length (A), double bonds (B) and specific fatty acid species (C) were analysed by mass spectrometry. The molar quantities of lipid were calculated from the signal of the internal standard PG(17:0/17:0). Additionally, the percent composition of fatty acid chain lengths, double bonds and fatty acid species was calculated for each individual sample from the quantified amount of total PG. The results were compared by 2-way ANOVA with Sidak’s multiple comparison post-test, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3, *p<0.05. FA = fatty acid.

A Carbon chain length B Double bonds in carbon chain

wild type AnxA6-/- wild type AnxA6-/- % SD % SD % SD % SD 14 5.88 ± 6.08 2.98 ± 3.82 0 69.44 ± 10.23 57.17 ± 17.50 16 21.05 ± 1.59 26.21 ± 1.83 1 20.91 ± 5.12 35.92 ± 17.20 *18 46.32 ± 2.39 61.49 ± 11.78 2 4.04 ± 1.58 4.11 ± 1.92 20 13.29 ± 6.81 4.41 ± 4.98 3 5.20 ± 7.55 2.80 ± 4.68 22 13.47 ± 4.49 4.91 ± 4.47 6 0.41 ± 0.70 0.00 ± 0.00

C Fatty acids

wild type AnxA6 -/- wild type AnxA6 -/- % SD % SD % SD % SD FA 14:0 5.88 ± 6.08 2.98 ± 3.82 FA 20:1 0.11 ± 0.10 0.49 ± 0.31 FA 16:0 20.29 ± 1.98 25.69 ± 1.19 FA 20:2 1.68 ± 1.16 0.37 ± 0.35 FA 16:1 0.76 ± 1.09 0.52 ± 0.89 FA 20:3 5.20 ± 7.55 2.80 ± 4.68 FA 18:0 27.44 ± 8.04 26.36 ± 11.36 FA 22:0 9.55 ± 6.28 1.39 ± 0.96 *FA 18:1 16.94 ± 7.62 31.38 ± 21.53 FA 22:1 3.09 ± 2.25 3.52 ± 3.89 FA 18:2 1.94 ± 0.88 3.75 ± 2.10 FA 22:2 0.42 ± 0.73 0.00 ± 0.00 FA 20:0 6.30 ± 3.12 0.75 ± 0.85 FA 22:6 0.41 ± 0.70 0.00 ± 0.00

PI lipids were analysed in negative ion mode. Due to the lack of a suitable PI internal standard, the composition of PI lipids was calculated from the corrected areas of ion counts obtained by the mass spectrometer without normalisation to a standard peak. Because of this different mode of analysis, PI could not be included in the comparison of different lipid classes previously in this results chapter.

In the current analysis of mouse T cells, more than half of the detected PI fatty acids were 18 carbon atoms long while 16C and 20C chain lengths made up less than 20% each (Table 5-4 A). In AnxA6-/- T cells 18C fatty acids were significantly more common than in wild type T cells. Instead, AnxA6-/- T cells had less 16C and 20C fatty acids than wild type T cells but these differences were not significant. More than 80% of the PI fatty acid moieties were unsaturated and more than 5% were mono-unsaturated or unsaturated with four double bonds (Table 5-4 B). Other polyunsaturated moieties were only detected in traces. Of the detected PI fatty acid species, the majority were 18:0, 16:0, 20:4 and 18:1 (Table 5-4 C). AnxA6-/- T cells contained significantly more PI with 18:0 acid than wild type T cells and significantly less 16:0 acid. AnxA6-/- also contained less 20:0 and more 22:6 than wild

Table 5-4: Analysis of fatty acids in PI lipids in lysates of primary T cells from AnxA6-/- and wild type mice. The abundance of fatty acids of a certain chain length (A), double bonds (B) and specific fatty acid species (C) was analysed by mass spectrometry. The percent composition of fatty acid chain lengths, double bonds and fatty acid species was calculated for each individual sample from the corrected areas of PI. The results were compared by 2-way ANOVA with Sidak’s multiple comparison post-test, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3, **p<0.01, ****p<0.0001. FA = fatty acid.

A Carbon chain length B Double bonds in carbon chain

wild type AnxA6-/- wild type AnxA6-/- % SD % SD % SD % SD 14 0.89 ± 0.65 1.59 ± 2.51 0 83.97 ± 7.00 82.79 ± 4.63 16 15.73 ± 6.35 8.88 ± 3.69 1 7.41 ± 4.40 6.37 ± 0.76 **18 65.51 ± 6.08 77.75 ± 4.28 2 0.37 ± 0.26 0.25 ± 0.19 20 13.87 ± 3.58 9.23 ± 2.61 3 0.68 ± 0.62 0.34 ± 0.28 22 4.00 ± 1.89 2.54 ± 2.57 4 7.06 ± 1.87 8.06 ± 1.52 6 0.50 ± 0.87 2.18 ± 2.14

C Fatty acids

wild type AnxA6 -/ - wild type AnxA6 -/- % SD % SD % SD % SD FA 14:0 0.89 ± 0.65 1.59 ± 2.51 FA 20:2 0.07 ± 0.12 0.04 ± 0.06 **FA 16:0 15.21 ± 6.17 7.91 ± 4.91 FA 20:3 0.68 ± 0.62 0.34 ± 0.28 FA 16:1 0.53 ± 0.20 0.98 ± 1.29 FA 20:4 7.02 ± 1.80 8.06 ± 1.52 ****FA 18:0 59.99 ± 3.90 72.44 ± 3.70 FA 22:0 2.48 ± 2.15 0.23 ± 0.24 FA 18:1 5.33 ± 3.81 5.10 ± 0.91 FA 22:1 0.85 ± 0.61 0.13 ± 0.22 FA 18:2 0.18 ± 0.23 0.21 ± 0.15 FA 22:2 0.12 ± 0.14 0.00 ± 0.00 FA 20:0 5.41 ± 4.84 0.63 ± 0.89 FA 22:4 0.04 ± 0.08 0.00 ± 0.00 FA 20:1 0.70 ± 0.71 0.16 ± 0.18 FA 22:6 0.50 ± 0.87 2.18 ± 2.14

type T cells, but these differences were not significant.

The PS fatty acid moieties of AnxA6-/- and wild type T cells were analysed in negative ion mode. The distribution of fatty acid length (Table 5-5 A) and the degree of saturation (Table 5-5 B) were not significantly different between wild type and AnxA6-/- T cells. But when the abundance of specific fatty acid species was analysed, it was found that AnxA6-/- T cells contained significantly more 18:0 than wild type T cells (Table 5-5 C). The increase of 18:0 was accompanied by a reduction of 16:0 and 18:1.

Table 5-5: Analysis of fatty acids in PS lipids in lysates of primary T cells from AnxA6-/- and wild type mice. The abundance of fatty acids of a certain chain length (A), double bonds (B) and specific fatty acid species (C) was analysed by mass spectrometry. The molar quantities of lipid were calculated from the signal of the internal standard PS(17:0/17:0). The percent composition of fatty acid chain lengths, double bonds and fatty acid species was calculated for each individual sample from the quantified amount of total PS. The results were compared by 2-way ANOVA with Sidak’s multiple comparison post-test, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3, ***p<0.001. FA = fatty acid.

A Carbon chain length B Double bonds in carbon chain wild type AnxA6-/- wild type AnxA6-/-

% SD % SD % SD % SD 14 1.79 ± 2.93 0.28 ± 0.42 0 67.32 ± 5.48 69.81 ± 6.17

16 14.62 ± 7.11 10.39 ± 2.36 1 15.20 ± 4.78 11.40 ± 6.30

18 65.61 ± 6.15 71.77 ± 3.58 2 1.31 ± 1.39 2.94 ± 1.98

20 5.35 ± 1.40 4.98 ± 4.21 3 1.09 ± 0.72 0.71 ± 0.19

22 12.62 ± 1.03 12.58 ± 1.91 4 3.55 ± 1.17 3.50 ± 4.73

6 11.54 ± 0.90 11.63 ± 1.83

C Fatty acids wild type AnxA6 -/- wild type AnxA6 -/- % SD % SD % SD % SD FA 14:0 1.79 ± 2.93 0.28 ± 0.42 FA 20:2 0.23 ± 0.11 0.26 ± 0.29

FA 16:0 14.27 ± 7.36 10.27 ± 2.42 FA 20:3 1.09 ± 0.72 0.71 ± 0.19

FA 16:1 0.36 ± 0.42 0.12 ± 0.06 FA 20:4 3.29 ± 1.14 3.34 ± 4.71

***FA 18:0 50.09 ± 1.72 58.47 ± 4.06 FA 22:0 0.62 ± 0.42 0.54 ± 0.59

FA 18:1 14.45 ± 4.16 10.63 ± 5.97 FA 22:1 0.20 ± 0.10 0.24 ± 0.32

FA 18:2 1.07 ± 1.27 2.67 ± 1.87 FA 22:2 0.01 ± 0.02 0.01 ± 0.02

FA 20:0 0.55 ± 0.37 0.25 ± 0.23 FA 22:4 0.25 ± 0.06 0.16 ± 0.10

FA 20:1 0.19 ± 0.15 0.41 ± 0.48 FA 22:6 11.54 ± 0.90 11.63 ± 1.83

Negative ion analysis of the fatty acid chain lengths found in PE showed no differences between AnxA6-/- and wild type T cells (Table 5-6 A). Polyunsaturated fatty acids with four double bonds were more common in AnxA6-/- T cells, but this difference was not significant (Table 5-6 B). The analysis of individual fatty acid species showed that 20:4 and 18:0 were the predominant species (Table 5-6 C). 16:0, 18:1 and 22:6 each contributed approximately 10% of all lipid species. AnxA6-/- T cells contained significantly more of the most abundant compound, arachidonic acid, than wild type T cells. This increase was

Table 5-6: Analysis of fatty acids in PE lipids in lysates of primary T cells from AnxA6-/- and wild type mice. The abundance of fatty acids of a certain chain length (A), double bonds (B) and specific fatty acid species (C) was analysed by mass spectrometry. The molar quantities of lipid were calculated from the signal of the internal standard PE(17:0/17:0). The percent composition of fatty acid chain lengths, double bonds and fatty acid species was calculated for each individual sample from the quantified amount of total PE. The results were compared by 2-way ANOVA, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3, **p<0.01. FA = fatty acid.

A Carbon chain length B Double bonds in carbon chain

wild type AnxA6-/- wild type AnxA6-/- % SD % SD % SD % SD 14 0.05 ± 0.06 0.05 ± 0.08 0 33.67 ± 3.42 34.06 ± 2.02 16 11.48 ± 0.57 11.81 ± 1.70 1 10.81 ± 1.43 10.47 ± 1.69 18 36.97 ± 3.84 35.61 ± 0.57 2 5.84 ± 0.55 4.14 ± 2.15 20 34.63 ± 0.22 36.55 ± 0.79 3 3.34 ± 0.45 1.68 ± 1.35 22 16.79 ± 4.47 15.93 ± 0.85 4 32.67 ± 0.99 36.90 ± 0.99 24 0.07 ± 0.07 0.04 ± 0.04 5 1.67 ± 0.82 2.09 ± 0.25 6 12.00 ± 5.50 10.66 ± 0.72

C Fatty acids

wild type AnxA6 -/- wild type AnxA6 -/ - % SD % SD % SD % SD FA 14:0 0.05 ± 0.06 0.05 ± 0.08 **FA 20:4 29.71 ± 1.00 33.80 ± 0.90 FA 16:0 11.38 ± 0.59 11.73 ± 1.69 FA 22:0 0.10 ± 0.03 0.03 ± 0.04 FA 16:1 0.10 ± 0.02 0.08 ± 0.05 FA 22:1 0.04 ± 0.02 0.04 ± 0.03 FA 18:0 21.70 ± 2.59 22.18 ± 0.33 FA 22:2 0.00 ± 0.00 0.01 ± 0.02 FA 18:1 10.12 ± 1.27 9.96 ± 1.59 FA 22:3 0.04 ± 0.02 0.00 ± 0.01 FA 18:2 5.15 ± 0.17 3.46 ± 1.85 FA 22:4 2.96 ± 0.24 3.10 ± 0.15 FA 20:0 0.44 ± 0.49 0.07 ± 0.06 FA 22:5 1.67 ± 0.82 2.09 ± 0.25 FA 20:1 0.48 ± 0.24 0.35 ± 0.18 FA 22:6 12.00 ± 5.50 10.66 ± 0.72 FA 20:2 0.69 ± 0.38 0.66 ± 0.33 FA 24:1 0.07 ± 0.07 0.04 ± 0.04 FA 20:3 3.30 ± 0.44 1.68 ± 1.36

accompanied by a reduction of other polyunsaturated fatty acids like linoleic acid and 20:3, but their levels were not significantly reduced in AnxA6-/- T cells.

The phosphocholine head group is detected very reliably in positive ion mode which makes this the preferred way to quantify PC lipids. However, positive ion mode does not reveal any detail about the specific fatty acid species of lipids with the same molecular weight and only the abundance of isobaric species was analysed.

The positive ion analysis of isobaric PC lipid species (Table 5-7) showed that 32:0, 34:1 and 36:4 were the most abundant lipid species in T cells from either genotype. There were

Table 5-7: Analysis of isobaric PC lipid species in lysates of primary T cells from AnxA6-/- and wild type mice. The abundance of PC molecular species with double bonds was analysed by mass spectrometry in positive ion mode. The molar quantities of lipid were calculated from the signal of the internal standard PC(19:0/19:0) or PC 38:0. The percent composition of fatty acid chain lengths, double bonds and fatty acid species was calculated for each individual sample from the quantified amount. The results were compared by 2-way ANOVA with Sidak’s multiple comparisons post-test, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3.

wild type AnxA6-/- wild type AnxA6-/- % SD % SD % SD % SD 28:0 0.01 ± 0.01 0.04 ± 0.03 38:4 8.96 ± 0.69 9.93 ± 2.52 30:0 1.45 ± 0.27 1.63 ± 0.46 38:5 6.08 ± 0.66 6.81 ± 1.04 32:0 19.30 ± 1.77 17.67 ± 4.12 38:6 3.14 ± 0.10 3.03 ± 0.05 32:1 1.44 ± 0.14 2.58 ± 0.39 40:0 0.01 ± 0.02 0.10 ± 0.08 32:2 0.11 ± 0.06 0.18 ± 0.06 40:1 0.07 ± 0.02 0.01 ± 0.02 34:0 2.30 ± 0.20 1.64 ± 0.96 40:2 0.14 ± 0.03 0.07 ± 0.08 34:1 13.95 ± 0.64 15.40 ± 0.62 40:3 0.22 ± 0.03 0.15 ± 0.05 34:2 5.80 ± 0.25 5.59 ± 0.68 40:4 0.69 ± 0.21 0.72 ± 0.09 34:3 0.24 ± 0.03 0.40 ± 0.07 40:5 2.40 ± 0.15 2.32 ± 0.47 34:4 0.16 ± 0.03 0.17 ± 0.01 40:6 3.02 ± 0.44 2.82 ± 0.57 34:5 0.01 ± 0.01 0.02 ± 0.02 42:0 0.02 ± 0.03 0.01 ± 0.01 36:0 0.04 ± 0.06 0.17 ± 0.09 42:2 0.03 ± 0.03 0.05 ± 0.03 36:1 3.11 ± 0.26 2.62 ± 0.15 42:3 0.08 ± 0.04 0.04 ± 0.03 36:2 6.82 ± 0.20 5.78 ± 0.42 42:4 0.05 ± 0.05 0.10 ± 0.06 36:3 4.43 ± 0.54 4.08 ± 0.40 42:5 0.63 ± 0.16 0.56 ± 0.09 36:4 11.62 ± 0.35 11.85 ± 1.32 42:6 0.32 ± 0.02 0.26 ± 0.04 36:5 0.31 ± 0.05 0.52 ± 0.14 44:5 0.10 ± 0.02 0.10 ± 0.04 38:1 0.26 ± 0.13 0.19 ± 0.12 44:6 0.11 ± 0.03 0.06 ± 0.05 38:2 0.64 ± 0.13 0.30 ± 0.13 46:1 0.00 ± 0.00 0.02 ± 0.02 38:3 1.91 ± 0.40 1.99 ± 0.16 46:5 0.01 ± 0.01 0.03 ± 0.03

no significant differences in the analysed PC lipid species of wild type and knock-out T cells.

Like PC, SM has a choline head group and has to be analysed in positive ion mode, which yields very limited details about the fatty acid composition. However, due to its sphingoid base backbone, which usually contains an 18C alkyl chain with one trans double bond, the remaining fatty acid moiety of SM can be deduced from the sum formula. An exception to this is dihydro-SM (e.g. SM 34:0;2) which is found in low amounts and contains a completely saturated sphingoid base (sphinganine)425.

Table 5-8: Analysis of SM molecular species in lysates of primary T cells from AnxA6-/- and wild type mice. The abundance of SM molecular species with double bonds was analysed by mass spectrometry in positive ion mode. The molar quantities of lipid were calculated from the signal of the internal standard PC(19:0/19:0) or PC 38:0. The percent composition of fatty acid chain lengths, double bonds and fatty acid species was calculated for each individual sample from the quantified amount of total SM. The results were compared by 2-way ANOVA with Sidak’s multiple comparisons post-test, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3 *p<0.05, **p<0.01, ****p<0.0001.

wild type AnxA6-/- detected fatty % SD % SD species acid 34:0;2 FA 16:0 1.82 ± 0.68 0.55 ± 0.52 **34:1;2 FA 16:0 46.41 ± 1.27 42.48 ± 1.41 34:2;2 FA 16:1 0.37 ± 0.06 0.38 ± 0.13 **36:1;2 FA 18:0 15.12 ± 1.56 11.04 ± 0.33 36:2;2 FA 18:1 0.39 ± 0.14 0.11 ± 0.10 ****40:1;2 FA 22:0 3.23 ± 2.93 9.79 ± 0.59 *42:1;2 FA 24:0 2.51 ± 0.42 5.51 ± 0.93 42:2;2 FA 24:1 30.15 ± 2.16 30.15 ± 0.95

The analysis of SM in AnxA6-/- and wild type T cells (Table 5-8) showed that 34:1;2 (46.4% wild type and 42.5% AnxA6-/-) and 42:2;2 (both 30.1%) were the most common species. AnxA6-/- T cells contained significantly less SM 34:1;2 and SM 36:1;2 and significantly more SM 40:1;2 and SM 42:1;2. As the fatty acid residue can be deduced from the summary notation this means that AnxA6-/- T cells contained less SM with 16:0 (34:1;2) and 18:0 (36:1;2) and more SM with 22:0 (40:1;2) and 24:0 (42:1;2) fatty acids.

In mammals, the sn-1 position of an ether lipid can be an ether (prefix “O-alkyl” or “plasmanyl-“) or a vinylether (“O-alkenyl” or “plasmenyl-“) while sn-2 contains an acyl

residue like other glycerophospholipids. Both ether lipid varieties are abbreviated with the suffix “-O” to PC-O and PE-O, respectively.

Table 5-9: Analysis of PE ether lipid species in lysates of primary T cells from AnxA6-/- and wild type mice. The abundance of PE-O isobaric species was analysed by mass spectrometry in positive ion mode. The molar quantities of lipid were calculated from the signal of the internal standard PE(17:0/17:0) or PE 34:0. The percent composition of fatty acid chain lengths, double bonds and fatty acid species was calculated for each individual sample from the quantified amount of total PE- O. The results were compared by 2-way ANOVA with Sidak’s multiple comparison post-test, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3.

wild type AnxA6-/- wild type AnxA6-/- % SD % SD % SD % SD 30:0 0.18 ± 0.31 0.00 ± 0.00 40:0 34.91 ± 1.33 33.74 ± 3.62 30:1 0.10 ± 0.17 1.60 ± 2.59 40:1 3.77 ± 1.58 5.83 ± 3.13 32:1 0.18 ± 0.17 1.25 ± 1.62 40:2 2.63 ± 0.57 2.40 ± 1.70 34:0 0.06 ± 0.10 0.00 ± 0.00 40:3 0.52 ± 0.90 0.00 ± 0.00 34:1 1.41 ± 0.32 2.51 ± 1.54 40:4 2.37 ± 0.35 1.98 ± 0.24 34:2 0.60 ± 0.33 0.28 ± 0.49 40:5 1.79 ± 1.41 2.26 ± 2.36 34:3 0.00 ± 0.00 0.94 ± 1.62 40:6 5.17 ± 0.98 3.74 ± 3.40 34:4 0.00 ± 0.00 0.09 ± 0.16 42:0 2.69 ± 0.66 2.20 ± 1.92 36:0 0.03 ± 0.06 0.67 ± 1.16 42:1 2.53 ± 0.39 2.04 ± 1.77 36:3 0.22 ± 0.33 0.06 ± 0.10 42:3 1.14 ± 1.02 0.36 ± 0.63 36:4 0.62 ± 1.07 0.29 ± 0.51 42:4 0.00 ± 0.00 0.14 ± 0.24 36:5 12.61 ± 2.06 15.70 ± 3.72 42:5 0.29 ± 0.50 0.00 ± 0.00 38:0 3.97 ± 0.48 2.23 ± 2.02 42:6 0.42 ± 0.40 0.18 ± 0.16 38:2 0.16 ± 0.27 0.06 ± 0.10 44:3 0.00 ± 0.00 0.07 ± 0.12 38:3 0.00 ± 0.00 0.61 ± 0.54 44:4 0.80 ± 0.72 0.12 ± 0.21 38:4 7.10 ± 2.27 7.40 ± 2.91 46:5 0.00 ± 0.00 0.47 ± 0.81 38:5 13.62 ± 1.80 10.64 ± 2.54 46:6 0.12 ± 0.21 0.12 ± 0.21

The neutral loss scan of PE-O species in positive ion mode showed that PE-O 40:0, PE-O 36:5 and PE-O 38:5 were the most abundant species (Table 5-9). Since PE ether lipids commonly contain 16:0, 18:0 and 18:1 in the sn-1 position and polyunsaturated fatty acids in sn-279,426, the abundance of PE-O 36:5 and PE-O 38:5 indicates that PE-O species in this analysis frequently contained fatty acids of 20C chain length and were polyunsaturated. The distribution of isobaric PE-O species revealed no significant differences between AnxA6-/- and wild type T cells.

Table 5-10: Analysis of PC ether lipid species in lysates of primary T cells from AnxA6-/- and wild type mice. The abundance of PC-O molecular species with double bonds was analysed by mass spectrometry in positive ion mode. The molar quantities of lipid were calculated from the signal of the internal standard PC(19:0/19:0). The percent composition of fatty acid chain lengths, double bonds and fatty acid species was calculated for each individual sample from the quantified amount of total PC- O. The results were compared by 2-way ANOVA with Sidak’s multiple comparison post-test, the analysis was conducted in triplicates with a pooled sample from 2 mice per group. Error SD, n=3, *p<0.05, **p<0.01, ****p<0.0001. wild type AnxA6-/- wild type AnxA6-/- % SD % SD % SD % SD 30:0 0.69 ± 0.17 0.58 ± 0.23 40:1 1.84 ± 0.17 2.13 ± 0.36 ****32:0 12.49 ± 1.97 9.57 ± 2.16 40:2 0.16 ± 0.14 0.22 ± 0.14 32:1 1.54 ± 0.22 2.09 ± 0.45 40:3 0.42 ± 0.12 0.27 ± 0.26 32:2 0.09 ± 0.09 0.11 ± 0.10 40:4 1.23 ± 0.14 1.04 ± 0.49 34:0 4.11 ± 0.14 3.24 ± 0.60 40:5 1.47 ± 0.33 1.56 ± 0.36 34:1 14.82 ± 0.86 15.60 ± 2.76 42:0 2.48 ± 0.38 3.15 ± 0.11 34:2 2.29 ± 0.21 2.69 ± 0.46 42:1 3.63 ± 0.15 3.07 ± 0.31 34:3 0.28 ± 0.09 0.41 ± 0.23 42:2 0.66 ± 0.21 0.92 ± 0.43 34:4 0.16 ± 0.09 0.12 ± 0.07 42:3 0.96 ± 0.17 0.72 ± 0.19 36:0 0.75 ± 0.12 0.23 ± 0.33 42:5 0.16 ± 0.15 0.13 ± 0.20 36:1 3.08 ± 0.48 2.80 ± 0.12 42:6 0.21 ± 0.09 0.21 ± 0.03 36:2 2.38 ± 0.56 2.38 ± 0.22 44:0 0.73 ± 0.20 0.54 ± 0.47 36:3 1.47 ± 0.19 1.21 ± 0.18 44:1 2.28 ± 0.35 2.17 ± 0.33 36:4 8.09 ± 0.61 8.91 ± 0.45 44:2 0.56 ± 0.18 0.55 ± 0.49 36:5 0.69 ± 0.10 0.89 ± 0.40 44:3 0.47 ± 0.07 0.27 ± 0.13 38:0 0.93 ± 0.39 0.79 ± 0.27 44:3 0.02 ± 0.03 0.26 ± 0.08 38:1 0.65 ± 0.09 0.50 ± 0.27 44:4 0.19 ± 0.13 0.15 ± 0.16 38:2 0.71 ± 0.21 0.48 ± 0.49 44:5 0.21 ± 0.07 0.35 ± 0.08 38:3 0.42 ± 0.07 0.67 ± 0.29 44:6 0.09 ± 0.08 0.07 ± 0.09 38:4 5.44 ± 1.09 4.45 ± 0.57 46:1 0.23 ± 0.21 0.31 ± 0.13 *38:5 9.80 ± 1.25 11.54 ± 2.09 46:2 0.15 ± 0.13 0.15 ± 0.14 *38:6 3.12 ± 0.03 4.70 ± 0.77 46:3 0.22 ± 0.04 0.38 ± 0.15 40:0 7.58 ± 0.52 7.26 ± 0.47 48:2 0.02 ± 0.03 0.20 ± 0.16

The positive ion analysis of ether PC lipids of AnxA6-/- wild type T cells showed greater diversity in lipid species than PE-O ether lipids. PC-O 34:1, PC-O 32:0, 38:5 were the three most abundant isobaric species. AnxA6-/- T cells showed significantly reduced values for the saturated species PC-O 32:0 and significantly increased levels of polyunsaturated PC-O 38:5 and 38:6 (Table 5-10).

Cholesterol is present in almost all mammalian cellular membranes where it is an essential modulator of membrane fluidity and a main constituent of membrane rafts. Cholesterol can be produced by the liver and released into the bloodstream as LDL particles, which deliver cholesterol to cells throughout the body. After LDL endocytosis, cholesterol is esterified with fatty acids (mainly oleic and linoleic acid) and retained in lipid droplets in the cytosol to reduce the amount of free cholesterol and to aid storage of surplus cholesterol427. Over-expression of AnxA6 has been shown to increase LDL- receptor endocytosis in CHO cells212. Over-expression of AnxA6 also affected cholesterol homeostasis and changed the distribution of cholesterol in cellular membranes but did not change the absolute levels of cholesterol in AnxA6 over-expression cell line248,257.

In the context of T cell activation, cholesterol has been shown to modulate the reactivity of T cells. Generally, cholesterol enrichment increases signalling and proliferative activity, while cholesterol depletion reduces T cell signalling402, which could be shown in the ɣδ T cell subset402 as well as αβ T cells183,402–404. However, it has been disputed, whether the methods employed to show a reduction of T cell signalling by artificial depletion of cholesterol with MβCD are a suitable way to analyse the role of cholesterol in T cell membranes428. Still, a reduction of T cell signalling after cholesterol depletion is consistent with studies showing cholesterol-enrichment increased T cell signalling402,405 and promoted an inflammatory T cell response together with increased STAT5 phosphorylation through the IL-2 receptor405.

To analyse whether the knock-out of AnxA6 lead to changes in cholesterol levels in primary T cells of AnxA6-/- mice, the unesterified cholesterol content of whole cell lysates was measured by mass spectrometry.

For mass spectrometric analysis of cholesterol, lipids were extracted from whole T cell lysates with a mixture of methanol and chloroform. The lipid phase was extracted with chloroform-based method established by Folch et al.424. Deuterium-labelled cholesterol (D6-cholestrol) was added as an internal standard to the sample prior to extraction to facilitate quantification. After phase separation chloroform forms the bottom layer and was carefully removed. Chloroform was then completely evaporated and the dried lipids were reconstituted in chloroform:methanol (2:1) with 5 mM ammonium acetate.

Liquid chromatography mass spectrometry (LC/MS) was used to measure the cholesterol content. LC/MS analysis was performed with an Accela LC and autosampler system (ThermoFisher Scientific) with an ESI ionisation source on a LTQ Orbitrap XL mass spectrometer (ThermoFisher Scientific). In this analysis, extracted lipids were first separated by LC. The liquid is then sprayed into the device by ESI. Following this, the fine mass separation is done by a linear trap quadrupole, after which the complete molecule is detected.

The results of the cholesterol analysis (Figure 5-8) show that AnxA6-/- T cells contained significantly lower amounts of cholesterol than wild type T cells. AnxA6-/- T cells contained only 89% (± 3.8 SD) of wild type cholesterol levels.

Figure 5-8 - Cholesterol in whole T cell lysates. T cell lysates were pooled from 3 mice per group and analysed at least in triplicate. Analysis: Student’s t-Test, error = SD, n = 4 (wild type) n = 3 (AnxA6-/-), **p<0.01.

To find out whether any differences in membrane order, as observed in Laurdan imaging, could be the result of altered lipid composition in AnxA6-/- T cells compared to wild type T cells, the phospholipid classes PC, PE, PS, PG, PA, PC-O, PE-O and PI were analysed by mass spectrometry. As there was no appropriate internal standard available for PI, the level of PI could not be quantified and compared to other phospholipids. Cholesterol was also not included in the graph as the analysis used a different extraction method and a different sample of T cells.

PC was the most abundant while PA was found to be the least abundant phospholipid. In decreasing abundance, the lipids could be ranked as follows: PC>PE≥PS>PC-O, SM>PE-O, PG>PA. This distribution of phospholipid classes was generally in agreement with the literature95,429. However, none of the publications analysed exactly the same lipid classes as this study. Leidl et al. report a larger amount of PE ether lipids in human lymphocytes

than was observed in this analysis. PC ether lipids are reported at very different abundance reaching from 25% of all PC lipids in pig lymphocytes430 to a subset of 9% total plasmalogens of all phospholipids in human lymphocytes431. In this study, PC-O levels of wild type T cells made up roughly 16% of total PC lipids.

AnxA6-/- T cells showed significantly decreased levels of PC and significantly increased levels of PS. The abundance of other phospholipid classes was similar to wild type T cells. Since membrane fluidity can be influenced by the nature of the polar head groups, the significantly reduced amount of PS as well as the significantly increased amount of PC both might contribute to generating more fluid plasma membrane as observed in Laurdan imaging.

PA is the least abundant phospholipid in cellular membranes. It is the simplest of all phospholipids and serves as a precursor for the biosynthesis of other glycerophospholipids432. AnxA6-/- T cells contained significantly more saturated and significantly less polyunsaturated (6 double bonds) PA fatty acids than wild type T cells. Since the concentration of PA in total phospholipids is very low, the signal was close to the detection limit and made the analysis of this particular lipid class less reliable than others. Bearing these limitations in mind, there was a trend in AnxA6-/- cells to incorporate more short and saturated fatty acids into PA. Wild type T cells contained more long chain fatty acids and contained significantly more 22:6 than AnxA6-/- T cells. In the context of cell membrane composition, these differences are unlikely to affect membrane structure or fluidity in any way since PA was only present in very low amounts (less than 1%, Figure 5-7 B).

Like PA, PG is not a very abundant phospholipid in cellular membranes. Its most important role is to act as a pulmonary surfactant in the lung where it can make up 10% of total surfactant phospholipids. The main function of PG inside cellular membranes seems to be to act as a precursor to cardiolipin, the major lipid component of the inner mitochondrial membrane with a dimeric structure433. The analysis of PG in AnxA6-/- and wild type T cells showed an overall distribution of PG fatty acid species, consisting mainly of 16:0, 18:0 and 18:1 fatty acids, which was similar to previous findings170,434. The fatty acids in PG tended to be of shorter chain length in AnxA6-/- than in wild type T cells but these differences were not significant. AnxA6-/- T cells contained significantly more 18:1 fatty acids. However, due to the low levels of total PG in the cell (less than 2% of the total phospholipid), an increase of 18:1 fatty acids is unlikely to have an effect on the overall membrane characteristics.

PI is generally less abundant in circulating blood cells than PS, but more common than PG and PA170,429,435. It is an important precursor for signalling processes at the membrane. The inositol head group can be phosphorylated in multiple places and forms phosphatidylmonophosphate, PIP2 and PIP3. Multiple phosphorylations concentrate intense negative charges on PI-derived lipids. Cleavage of PIP2 into IP3 and DAG generates cytosolic (IP3) as well as membrane-resident (DAG) second messenger which induce or participate in Ca2+ signalling. Phosphorylated PI lipids are present in the inner leaflet in the plasma membrane in very low abundance (0.5-1% of total lipids)436 and were not analysed here. The fatty acid composition of PI in AnxA6-/- and wild type T cell was rather distinctive with 18C, 20:4 and 16:0 fatty acids the most common. This composition was in agreement with previous publications170,437–439. The PI fatty acid moieties of AnxA6-/- T cells did not follow a trend toward shorter or longer, more or less saturated fatty acid chains. Instead, the most common moiety, stearic acid, was even more abundant (significantly) in AnxA6-/- T cells. The levels of 16:0 and 20:0 fatty acids were reduced in AnxA6-/- T cells although only the reduction of 16:0 fatty acids was significant.

As discussed previously in this thesis, PS is a negatively charged phospholipid present in the cytosolic leaflet of cellular membranes89. Presentation of PS on the outside of the cell can induce clearance of apoptotic cells and in some cases blood coagulation. It is the main phospholipid binding partner for AnxA6 and is found co-segregated with caveolae, cholesterol and sphingolipids96. The analysis of PS fatty acid moieties in AnxA6-/- and wild type T cells showed that 18:0 and 18:1 fatty acids were most abundant, which is in agreement with previous analysis of PS composition in human blood platelets435 and correlates with results from human leukocytes170,429. This analysis deviates from some of previous publications in the measured levels of 20:4, that were lower than in the cited research, and the levels of 22:6 and 16:0, that were higher in this analysis. However, the levels of 20:4 and 16:0 were similar to Jurkat cells170. These differences could be a result of inter-assay variation. In this analysis AnxA6-/- T cells had reduced levels of 16:0 and 18:1 (not significant) and contained significantly more 18:0 than wild type T cells. The increase of a saturated fatty acid at the expense of a shorter and a mono-unsaturated fatty acid is likely to contribute to a decrease in membrane fluidity.

PE is synthesised in the cell either in the ER membrane from an activated glycerophospholipid precursor (CDP-DAG) or from PS in the inner mitochondrial membrane440. PE is enriched in the cytosolic layer of cellular membranes and supports the negative curvature of the inner leaflet of the plasma membrane89,441. PE accounts for about

20% of phospholipids in most mammalian cells and is present at much higher concentrations in the brain (45%)99. PE is known to contain high amounts of arachidonic acid and DHA compared to other phospholipid classes442. The analysis of AnxA6-/- and wild type T cells was in agreement with this. Publications focusing on analysing T cells and platelets also measured similar levels of fatty acids. For example, the abundance of 22:6 was high in this study compared with the fatty acids distribution in platelets435, but similar to Jurkat cells170. PE lipids of AnxA6-/- T cells were significantly more abundant in 20:4 fatty acids. Higher levels of arachidonic acid potentially contribute to higher membrane fluidity.

PC is the most abundant phospholipid in the cell and is enriched in the outer leaflet of the plasma membrane. Due to the performed positive ion analysis, only the total number of carbon atoms and double bonds in the fatty acid moiety could be reported. PC 34:1 and 32:0 were the most abundant PC species in AnxA6-/- and wild type T cells. This result was consistent with previous analysis of PC species in lymphocytes170,429. The distribution of PC species was not significantly different between AnxA6-/- and wild type T cells.

SM is enriched in the outer leaflet of the plasma membrane. It contains mainly saturated and monounsaturated fatty acids and is a component of membrane rafts417. Similarly to previous analysis of SM in lymphocytes, SM 34:1;2 and 42:2;2 were found to be the most abundant SM species429. The distribution of fatty acids in SM of AnxA6-/- and wild type T cells indicated that, in this lipid class, AnxA6-/- T cells contained more long chain saturated fatty acids (22C in and 24C in 42:1;2) and less 16:0 (in 34:1;2) and 18:0 (in 36:1;2) fatty acids. A shift towards long saturated fatty acids potentially reduces membrane fluidity.

Ether lipids are common among PC and PE lipids but their analysis is difficult since they oxidise easily80. In this analysis of AnxA6-/- and wild type T cells, only the isobaric species were reported which include O-alkyl as well as O-alkenyl ether lipids. PC ether lipids in their majority contain an O-alkyl ether connection at sn-1 while the majority of PE ether lipids contain an O-alkenyl connection at sn-1430. The percentage of PE ether lipids compared to PE in the lipid class analysis was low when compared to a previous analysis of PE lipids in lymphocytes429. Compared to previous measurements of fatty acid species in lymphocytes, there seem to be many saturated fatty acids (e.g. PE-O 40:0), while previously 38:6 and 38:5 lipids were found to most abundant79. Two saturated fatty acid moieties like in a 40:0 lipid are untypical of PE ether lipids, which contain 16:0, 18:0 and 18:1 in the sn-1 position and polyunsaturated fatty acids in sn-279,426. This makes it likely

that the ether species in the samples were in fact partially degraded. There were no differences in the fatty acid distribution on AnxA6-/- and wild type T cells. In the analysis of PC ether lipids, the most abundant species were PC-O 34:1, 38:5, 38:4 and 32:0, which is in agreement with previous findings on PC-O species in lymphocytes79. AnxA6-/- T cells contained significantly lower levels of 32:0 and higher level of 38:5 and 38:6. Possible isobaric permutations of the polyunsaturated lipids likely contain 20:4, 20:5 or 22:6. This distribution indicates that PC ether lipids of AnxA6-/- T cells are enriched in polyunsaturated fatty acids.

Cholesterol is a very abundant lipid in the cell. It is enriched in the plasma membrane compared with other cellular membranes89. Cholesterol is major regulator of plasma membrane fluidity and permeability. Measuring the levels of cholesterol in AnxA6-/- T cells was of particular interest because of the involvement of AnxA6 in cholesterol uptake and homeostasis212. The analysis of cholesterol levels in AnxA6-/- and wild type T cells showed that AnxA6-/- T cells contained significantly less cholesterol than wild type T cells. A reduction in cholesterol levels is likely to decrease membrane order in AnxA6-/- T cells compared with wild type T cells.

To investigate the underlying reason for the altered IL-2/STAT5 signalling behaviour and the proliferation defect in T cells of AnxA6-/- mice, isolated primary T cells from AnxA6-/- mice were analysed for changes in lipid composition and membrane fluidity. The membrane order of activated and resting T cells was investigated with Laurdan microscopy and the lipid composition was analysed with mass spectrometry.

Probing the membrane order by Laurdan imaging in T cells that were incubated with beads coated in activating antibodies, showed different outcomes for knock-down and knock-out T cells. The knock-down of AnxA6-/- in genetically modified Jurkat T cells did not have any effect on the membrane order of activated T cells. Primary T cells of AnxA6-/- mice, however, showed reduced levels of membrane order in the area of the plasma membrane that was engaged in activation as well as the part of the plasma membrane where the TCR was not engaged in bead-cell contact. This result indicated that the membrane order was not just affected in activated T cells but was affected globally and that the membrane fluidity of AnxA6-/- T cells was increased overall. This was confirmed by measuring the membrane order of not activated, resting T cells.

It was plausible that the reduced levels of membrane order of the increased levels of membrane fluidity were a result of an altered lipid composition in AnxA6-/- T cells. To investigate the lipid composition, the cholesterol as well as the phospholipid levels were analysed by mass spectrometry. However, with this method, whole T cells lysates were analysed rather than isolated plasma membranes.

The analysis of cholesterol and phospholipid levels in T cells of AnxA6-/- and wild type T cells showed that, generally, the distribution of lipid classes and fatty acids of this analysis was similar to previous lipid analysis170,429. There were, however, several significant differences in the lipid composition of AnxA6-/- and wild type T cells. AnxA6-/- T cells were found to have reduced levels of PS and increased levels of PC lipids. PS decrease and PC increase can potentially reduce membrane order if their fatty acid moieties remain the same (Figure 1-6). Furthermore, cholesterol levels of primary T cell lysates from AnxA6-/- and wild type mice were analysed. AnxA6-/- T cell were found to contain less cholesterol than wild type T cells. Membrane cholesterol content is known to correlate with membrane order87,443. A reduction of cholesterol in the plasma membrane is likely to contribute to a more disordered, fluid membrane.

PE, PC-O and PG of AnxA6-/- T cells contained more mono- and polyunsaturated phospholipids than wild type T cells. Of these lipid classes, PE was the most abundant lipid class. Increases in saturated lipids of longer chain lengths result in more rigid membranes while increases in poly- and monounsaturated fatty acids or shorter chain length fatty acids increase membrane fluidity87,444(Figure 1-6). PE of AnxA6-/- T cells contained increased levels of arachidonic acid. The observed changes in fatty acid profiles of SM, PS, PI and PA in AnxA6-/- T cells could potentially counteract the increase in membrane fluidity. The profile for SM showed a reduction of short and an increase of longer chain length saturated fatty acids. PS and PI contained significantly more stearic acid moieties while 16:0 and oleic acid 18:1 levels were reduced. PA also showed an increase of saturated and a decrease of polyunsaturated fatty acids but, due to its low abundance, PA is not likely to contribute to the overall fluidity of cellular membranes.

Since the analysis of T cell lipid composition was performed on whole cell lysates, it can only be speculated how the changes of lipid classes and fatty acid profiles might have affected the plasma membrane. Cholesterol as well as PS are not distributed evenly among the cellular membranes but enriched at the plasma membrane89,445,446. This observation supports the idea that a reduction in total cholesterol and PS affects the plasma membrane more than other cellular membranes. Since the total levels of these lipids are lower in

AnxA6-/- than in wild type T cells, it is likely that the part of the cell where they are most abundant is affected as well. Taken together with the results of Laurdan imaging, it is likely that the plasma membrane is indeed affected by the changes in lipid composition and that the net effect of all changes leads to an increase in membrane fluidity. Thus lower cholesterol and PS and higher PC content in AnxA6-/- T cells compared to wild type T cells outweighed changes in fatty acid composition that might have had the opposite effect.

It has been shown previously that cholesterol levels can influence IL-2 signalling. More specifically, were found to be partitioned into rafts rich in SM and cholesterol182,447 and a reduction of cholesterol resulted in a decrease of STAT5 phosphorylation167,181,182. It is therefore highly likely that the observed reduction of IL-2 signalling in AnxA6-/- T cells was caused by a changed lipid composition of the plasma membrane. The reduction of proliferation promoting IL-2 signalling could eventually lead to a proliferation deficit in CD4+ cell in AnxA6-/- mice in vivo. Adding to this theory are findings of Cho et al. showing that IL-2 signalling is more sensitive to reduction of cholesterol levels. Cholesterol depletion of CD8+ T cells affected proliferation more in cells that were IL-2 stimulated than in cells that were TCR stimulated448.

Additionally, a decrease of cholesterol levels could affect T cell proliferation beyond changes in membrane order at a more basic level. During cell division, large amounts of cholesterol are needed to generate enough membranes for a second cell rendering sufficient supply of cholesterol necessary for cell division449. The regulation of cholesterol homeostasis is therefore of great importance for activating T cells450. If cholesterol uptake or the delivery of intracellular cholesterol to target membranes is not fully functional in AnxA6-/- T cells, it could decrease the proliferation rate of activated T cells.

The aim of this study was to test whether AnxA6 is relevant for an efficient immune response in vivo. Furthermore, the goal was to establish how AnxA6 could be involved in this process.

The results presented in this thesis showed that, in response to an immune stimulus in an AnxA6 knock-out mouse model, AnxA6-/- CD4+ T cells proliferated less than wild type T cells in vivo. In vitro AnxA6-/- T cells exhibited reduced IL-2 receptor signalling but not TCR signalling. Furthermore, AnxA6-/- T cells presented an altered lipid composition, and had increased plasma membrane fluidity.

AnxA6 is a membrane binding protein that is involved in plasma membrane repair, EGF receptor signalling, LDL receptor endocytosis, lysosomal targeting and cholesterol homeostasis202,212,213,260. Due to its ability to bind cholesterol-enriched phospholipid membranes as well as cytoskeletal components, AnxA6 has been hypothesised to be an organiser of plasma membrane domains rich in cholesterol, like membrane rafts275. Membrane rafts play important roles in receptor-mediated signalling events. Many receptors, e.g. IL-2 receptor167,181,182 and TCR183,402–404, are located in cholesterol-rich domains; as a result signalling events essential for T cell activation and proliferation can be modulated by cholesterol depletion and enrichment167,181–183,402–405.

AnxA6 has been suspected to play a role in lymphocyte function since its expression pattern in lymph nodes suggested that AnxA6 was involved in T cell development208. AnxA6 is expressed in mature T cells but not in immature cortical thymocytes208. Other annexin-family proteins are involved in immune functions outside of the cell451–453, while AnxA6 is a cytosolic protein that is constitutively expressed in mature T cells and is only known to be involved in intracellular tasks.

Annexins are multifunctional proteins that have some degree of functional redundancy198,220,383. Therefore, it is sometimes difficult to identify deficits in single knock- out mice. Even mice with a double knock-out of AnxA6 and AnxA5 are viable, fertile and do not show any obvious abnormalities454. AnxA6 and other annexin proteins can be

activated to participate in signalling events by particular types of cellular stress (e.g. decrease in pH and/or Ca2+ influx)216.

Investigating the function of AnxA6 with the help of an AnxA6 knock-out mouse model required that stress is applied to a cell or the organism. In this study, the applied stressor was a treatment that prompted an immune response by the treated mice. This in vivo assay was laid out in Chapter 3. It was investigated whether the T cells of AnxA6-/- mice in vivo respond differently to an induced immune challenge than wild type mice. By eliciting a delayed-type CHS reaction with the irritant DNFB, the treated mice developed a T cell- mediated immune response resulting in swelling of the treated areas, enlarged lymph nodes and enlarged spleens. The proliferation activity of T cells from lymph nodes was analysed by measuring the levels of PCNA, the expression of which is increased during cell division. It could be shown, that in response to the induced CHS reaction, the CD4+ T cells of AnxA6-/- mice were proliferating significantly less than the CD4+ T cells of wild type mice.

The consequences of the AnxA6 knock-out are highlighted in experiments with a similar AnxA6-/- mouse strain (MGI ID: 4432672)316. When pathogenic bacteria were introduced to the digestive system of AnxA6-/- and wild type mice, AnxA6-/- mice exhibited reduced bacterial clearance from the large intestine. Outside of the sterile environment of a research facility this disadvantage is likely to result in higher incidences of infection in knock-out animals. The proliferation defect of CD4+ T cells shown in this study, as well as the delayed bacterial clearance in AnxA6-/- mice, highlights the importance of AnxA6 – despite overlapping functions with other annexin family proteins. By putting the organism under a specific stress, exposure to the irritant DNFB, this study shows that some functions are unique to AnxA6 and cannot be replicated by similar proteins.

The proliferation defect in AnxA6-/- T cells raised the question: How AnxA6 could influence cell proliferation? Given that AnxA6 expression is abrogated or reduced in some cancers, AnxA6 has been implicated in cell proliferation previously264,296–298. However, previous publications found that decreased levels of AnxA6 increased cell proliferation218,324 rather than reduced it. Proliferation of cancerous cells is often fuelled by changes in EGF receptor signalling. The current model for the role of AnxA6 in reduction of EGF signalling proposes that AnxA6 targets p120GAP and PKCα, both of which are negative modulators of EGF receptor signalling, to the plasma membrane and the EGF receptor. Then AnxA6 and stabilises activated and internalised EGF receptor in a complex with these proteins reducing EGF receptor signalling260,264,265. T cell proliferation, however, is not driven by

EGF but by the TCR and IL-2 signalling. Thus, these two pathways were the main focus for investigating the underlying cause for the proliferation defect.

Chapter 4 concentrated on elucidating the role of AnxA6 in TCR and IL-2 signalling by the means of in vitro assays. Since known signalling pathways that include a role for AnxA6 are dependent on Ca2+ signalling, it is possible that the AnxA6 knock-out leads to a reduction of TCR signalling. TCR signalling is initiated by an antigen-presenting MHC binding to a cognate TCR. Several signalling cascades are initiated after this, one of them triggers the release of IP3 from membrane phospholipids to open internal calcium ion stores from the ER. It was hypothesised that the increase of cytosolic Ca2+ could activate AnxA6, target it to the immune synapse, which is enriched in the preferred binding partners of AnxA6: PS and cholesterol. There, AnxA6 could support the targeting of F-actin to the TCR microclusters and aid in forming the immunological synapse structure and dynamic movement of TCR microclusters. Fluorescence imaging of Jurkat cells showed that F-actin and AnxA6 co-localised at the immune synapse early during T cell activation (<10 min). A dSTORM imaging approach of primary wild type and AnxA6-/- T cells showed that punctate patterns of F-actin in early stages of immune synapse formation were more densely organised in AnxA6-/- than in wild type T cells. These results indicated that AnxA6 could potentially be involved in immune synapse formation in T cell activation. Consequently, the phosphorylation of signalling proteins involved in the signalling cascade from TCR to IL-2 production was investigated in activated Jurkat and primary murine T cells. However, no reduction of phosphorylation could be observed in AnxA6 knock- down Jurkat cells or AnxA6-/- T cells compared to control and wild type T cells, respectively. It was therefore unlikely that AnxA6 was necessary for TCR signalling. However, it cannot be ruled out that AnxA6 could be involved in TCR signalling in wild type T cells. It is possible that other mechanisms, or annexin proteins, are upregulated in the absence of AnxA6 and replace the function of AnxA6 at the immune synapse, e.g. more densely organised F-actin in AnxA6-/- T cells could be part of a compensatory process which “rescues” any TCR and immune synapse related functions.

One of the main outcomes of TCR signalling is IL-2 production, which will induce auto- and paracrinal stimulation through the IL-2 receptor. Measuring the levels of IL-2 transcription did not show any differences between activated AnxA6-/- and wild type T cells. Despite this, AnxA6-/- T cells secreted significantly more IL-2 after 4 hours of activation. The increased release of IL-2, together with a reduced T cell proliferation in vivo, suggested a decreased sensitivity of IL-2 stimulation in AnxA6-/- T cells. Examining

the response of wild type and AnxA6-/- T cells to IL-2 stimulation, it could indeed be shown that activated AnxA6-/- T cells exhibited less IL-2 receptor signalling activity than wild type T cells in response to the same amounts of IL-2. Due to the involvement of AnxA6 in endocytosis, it seemed possible that IL-2 signalling may be altered by changes in IL-2 receptor endocytosis. However, there were no significant differences between the endocytosis rate of IL-2 receptor in AnxA6-/- compared to wild type T cells.

While AnxA6 knock-out clearly had an effect on IL-2 signalling, it was not clear how AnxA6 was involved in this process. Both IL-2 and TCR signalling reside in cholesterol-rich membrane domains and interact with the signalling protein LCK and indirectly the cortical cytoskeleton16,17,72,76,176. However, unlike TCR signalling, IL-2 signalling does not involve Ca2+ fluxes386,387,455. While AnxA6 is able to bind cholesterol-rich phospholipid membranes, even in the absence225,456 of Ca2+, the function of this binding has not been resolved yet. Furthermore, AnxA6 is also able to bind LCK independently of Ca2+ 203. LCK is known to interact with the IL-2 receptor78 and the TCR, but the in vivo role of the AnxA6-LCK interaction has not yet been investigated. It cannot conclusively be explained why IL-2 signalling, but not TCR signalling, is affected in AnxA6-/- T cells. Given that AnxA6 can interact with cholesterol-rich membranes and the IL-2 receptor via LCK in the absence of Ca2+ signalling, a direct involvement of AnxA6 in the signalling pathway is possible, but more evidence is needed to substantiate this theory. If AnxA6 is not part of an IL-2 pathway, IL-2 signalling in AnxA6-/- T cells may have been affected by changes to the plasma membrane caused by the knock-out. AnxA6 plays a role in cholesterol homeostasis202,212 due to its involvement in LDL receptor endocytosis and recycling225,227. It is possible that AnxA6 knock-out leads to changes in the lipid composition of membranes of AnxA6-/- T cells altering membrane properties like fluidity and permeability.

An imaging as well as mass spectrometric approach was taken to investigate whether the cell lipid composition and plasma membrane fluidity of AnxA6-/- T cells are different from wild type T cells. This was summarised in Chapter 5. Probing the plasma membrane order with Laurdan imaging showed that activated, as well as resting, AnxA6-/- T cells displayed reduced plasma membrane order compared to wild type T cells. But, despite having a generally more fluid plasma membrane, AnxA6-/- T cells, just like wild type cells, still formed a highly ordered region in the plasma membrane at the immune synapse. This supported previous findings that the mechanism of forming an immune synapse was not dysfunctional in AnxA6-/- cells, as described in Chapter 4. However, the general reduction

of membrane order in resting T cells indicated that the lipid composition of AnxA6-/- T cells may be different from wild type T cells. Analysis of the overall lipid composition of AnxA6-/- and wild type T cells showed that AnxA6-/- contain significantly less PS and cholesterol, but higher levels of PC. Furthermore, PE lipids contained higher levels of the polyunsaturated fatty acid arachidonic acid (20:4). All of these changes potentially contribute to an increase in membrane fluidity. Fatty acids in other lipid classes like SM and PI contained higher levels of long saturated fatty acids in AnxA6-/- than in wild type T cells. These types of changes contribute to decreased membrane fluidity. It is difficult to predict the net effect of all changes in the lipid composition, moreover, the plasma membrane composition was not analysed separately. Owing to these unknowns, it is not certain how the different changes in lipid composition did affect the plasma membrane in the end. However, since cholesterol as well as PS are generally enriched in the plasma membrane of cells, it is likely that changes in the levels of these lipid classes in whole T cells of AnxA6-/- mice would indeed extend to the plasma membrane of AnxA6-/- T cells. The results from the Laurdan imaging also support this conclusion as they showed reduced membrane order at the plasma membrane. It was therefore reasonable to assume that the reduced membrane order of AnxA6-/- T cells was caused by a changed lipid composition.

The makeup of the plasma membrane plays a key role in signal transduction. Significant changes in the lipid composition of the bilayer can lead to changes in signalling efficiency166,167,457. The IL-2 receptor is dependent on cholesterol-rich membrane domains and has been previously shown to be sensitive to cholesterol depletion167,182,458. Hence, a reduction in cellular cholesterol and increase in membrane fluidity might cause reduced IL-2 receptor signalling in AnxA6-/- T cells. Since IL-2 signalling delivers a very powerful proliferative stimulus to activated T cells, a reduction in IL-2 signalling was probably responsible for the observed proliferation defect of CD4+ T cells in AnxA6-/- T cells in vivo. Furthermore, a limited supply of cholesterol, which is essential for cell growth and proliferation449, could also affect the rate of cell cycle progression in AnxA6-/- T cells.

The changed lipid composition of cellular membranes and the reduced availability of cholesterol affected the ability of CD4+ T cells to produce the same proliferative response to an immune challenge as wild type mice. The theory that IL-2 signalling was affected by a decrease in membrane order well explains why IL-2 signalling was affected by knocking out AnxA6, but it does not explain why TCR signalling was not affected in a similar way. Previously, it had been reported that cholesterol depletion reduced TCR signalling183. In

this study, despite finding reduced cholesterol levels in AnxA6-/- T cells, no evidence was found that TCR signalling was affected. However, the method employed to deplete cholesterol in previous studies, i.e. with the cholesterol binding compound MβCD, has been criticised by some researchers as being prone to side-effects and artefacts459,460 and it could be demonstrated that T cells could be fully activated despite cholesterol depletion428. Cholesterol depletion with MβCD can reduce Ca2+ signalling185. Publications that demonstrated the importance of cholesterol-rich membrane domains for IL-2 signalling, also used the method of cholesterol depletion with MβCD167,458,461. Since Ca2+ does not seem to be involved in IL-2 receptor signalling, the negative side-effects of MβCD treatment found for TCR signalling did not apply to these experiments. The hypothesis that cholesterol depletion affects IL-2 signalling more severely than TCR signalling is supported by Cho et al. who tested the effect of MβCD treatment on TCR and IL-2 induced proliferation in CD8+ T cells448. MβCD dramatically decreased IL-2 stimulated proliferation but TCR induced proliferation suffered only a small reduction.

Limited availability of cholesterol can also generally reduce cell proliferation449. T cells, like other cells, receive their cholesterol either through external sources, like LDL particles, supplied in the blood stream or through de novo synthesis within the cell. Blocking cholesterol synthesis by administering statins462, culturing cells in LDL-free462,463 or delipidated media464 as well as the absence of fuctional LDL receptors462 has been shown to reduce T cell proliferation. The reduction of cholesterol levels in AnxA6-/- T cells could therefore affect the ability of CD4+ T cells to proliferate at the same rate as wild type T cells.

The results of this study highlight the importance of the “correct” lipid composition of T cell membranes for an appropriate immune response. The differences found in this study suggest that overall lipid composition of AnxA6-/- T cells is different from wild type T cells and resulted in increased plasma membrane fluidity. Ultimately, these changes can affect cell proliferation either by reducing receptor signalling via IL-2, therefore reducing the activation of mitogenic pathways. Additionally, cell division could be affected by limited availability of certain lipids like cholesterol.

This thesis could establish that the knock-out of AnxA6 does impact mice in vivo when they are exposed to the stress of an immunological challenge. CD4+ T cells do not

proliferate to the same extent during an immune response as T cells of wild type mice. The proliferation defect is likely to be a result of reduced pSTAT5 signalling during IL-2 receptor activation, which is probably an effect of the changed lipid composition of the plasma membrane of AnxA6-/- T cells.

Since only whole cell T cell lysates were analysed for their lipidome, it is not clear by what mechanism AnxA6 knock-out actually impacts on the lipid composition and whether membranes of cellular compartments are affected differently. Reduced levels of cholesterol in AnxA6-/- T cells support the model that implies AnxA6 in cholesterol transport212,257. Since AnxA6 over-expression changes the distribution of cholesterol between cellular membranes, it is likely that a knock-out of this protein not only affects the overall lipid composition of whole cells and tissues, but also the distribution within the cells. Future experimentation could investigate to what extent the plasma membrane is affected and whether lipids like cholesterol and PS, that were found to be reduced in AnxA6-/- T cells, are accumulated in some organelles and reduced in others. As AnxA6 over-expression increases LDL receptor endocytosis and recycling227, it would be interesting to investigate if this process is reduced in AnxA6-/- cells. Additionally, the path of imported cholesterol could be traced with a fluorescently modified cholesterol probe that has been established as suitable tool to investigate cholesterol transport in live cells253. Analysing the route of cholesterol could reveal if, and where, the intracellular transport is affected by AnxA6 knock-out.

As discussed before in Chapter 6 the IL-2 signalling of AnxA6-/- T cells was reduced compared to wild type T cell, but TCR signalling of AnxA6-/- T cells was not decreased, even though both signalling pathways have been shown to be sensitive to membrane fluidity and cholesterol depletion. It is possible that TCR signalling is more robust than IL- 2 signalling regarding these changes, but a further reason for this differential impairment of receptor signalling could be the culturing conditions of the cells. The investigation of TCR signalling only looked into the TCR-mediated activation of naïve T cells. The investigation of IL-2 receptor signalling, however, could not be conducted with naïve T cells as they express little if any IL-2 receptor and as a result are barely responsive to IL- 262. Activated and cultured T cells, T cell blasts, do respond to IL-2 readily as well as to TCR stimulation. If stimulated with the same cue naïve and blast cells do not signal necessarily in the same way. Beyer et al.386 showed that T cell blasts stimulated with IL-2 half an hour prior to TCR re-stimulation and naïve T cells did respond to TCR stimulation with STAT5 phosphorylation. T cells blasts that were not stimulated with IL-2 shortly

before TCR signalling was initiated did not phosphorylate STAT5. Likewise, the activation of other signalling proteins like LAT and Akt were also affected by those subtle differences in treatment. This shows that cross-talk of TCR and IL-2 signalling only occurs at certain stages in T cell differentiation with a certain combination of external stimuli. Future experimentation should therefore scrutinise potential differences of wild type and AnxA6- /- T cells in STAT5 signalling in naïve cells as well as TCR signalling in T cell blasts.

Furthermore, it would be interesting to establish whether AnxA6 is indeed associated with the IL-2 receptor complex in the absence of Ca2+. Ca2+-independent binding of cholesterol- rich membranes and LCK has been described for AnxA6203,225,282 but evidence for an LCK- AnxA6 interaction in vivo has not yet been shown and the function of this association has not been elucidated.

Apart from requiring high amounts of cholesterol to facilitate cell division, the rapid expansion of activated T cells requires a different energy metabolism based on glycolysis rather than fatty acid oxidation465. Resting T cells, however, do not express many glycolytic enzymes. The complete restructuring of the cellular metabolism is initiated during the early stages of activation. Interestingly, IL-2 receptor activation as well as co-stimulation of CD28 also initiates the Akt pathway via PI3K, which is instrumental in promoting the translocation of glucose transporter GLUT1 to the plasma membrane466,467. These transmembrane proteins import high amounts of glucose. Furthermore, Akt signalling increases the activity of two enzymes central to glycolysis (hexokinase and phosphofructokinase). It would be intriguing to investigate if IL-2-induced Akt signalling is decreased in AnxA6-/- T cells. A reduction of this pathway in AnxA6-/- T cells could also contribute to the diminished T cell proliferation in vivo.

The involvement of AnxA6 in cell proliferation raises the question of whether this finding is specific to T cells, or whether other cells might also be impacted in this way as well. There are other cells and tissue types that have to quickly form new cells, e.g. the intestinal mucosa, and regenerating tissues like the liver. Yet, AnxA6-/- mice are healthy, breed normally and do not seem to be disadvantaged by the knock-out under normal (stress- free) conditions280. Intriguingly, it has been found that AnxA6-/- mice are unable to regenerate their liver after partial hepatectomy. However, if the diet of AnxA6-/- mice is supplemented post-surgery with glucose, their regenerative capacity is close to wild type levels (unpublished results from Anna Alvarez-Guaita). Hepatocytes express IL-2 receptor468 but the function of IL-2 signalling in liver regeneration is not clear469,470. Regeneration of a part of the liver is an extreme situation that also requires the initiation

of fast proliferation to generate new cells. The proliferation defect of T cells in AnxA6-/- mice may be the symptom of a more general deficit in the cellular metabolism relating to cell proliferation.

1. Murphy, K., Travers, P., Walport, M. & Janeway, C. Janeway’s Immunobiology. Q. Rev. Biol. (Garland Science, Taylor & Francis Group: New York & London, 2012).87(3):888.

2. Acuto, O. & Michel, F. CD28-mediated co-stimulation: a quantitative support for TCR signalling. Nat. Rev. Immunol. (2003), 3(12):939–51. doi:10.1038/nri1248

3. Bryant, P. & Ploegh, H. Class II MHC peptide loading by the professionals. Curr. Opin. Immunol. (2004), 16(1):96–102. doi:10.1016/j.coi.2003.11.011

4. Théry, C. & Amigorena, S. The cell biology of antigen presentation in dendritic cells. Curr. Opin. Immunol. (2001), 13(1):45–51.

5. Von Andrian, U.H. & Mempel, T.R. Homing and cellular traffic in lymph nodes. Nat. Rev. Immunol. (2003), 3(11):867–78. doi:10.1038/nri1222

6. Croft, M. Activation of naive, memory and effector T cells. Curr. Opin. Immunol. (1994), 6(3):431–7.

7. Zhu, J., Yamane, H. & Paul, W.E. Differentiation of effector CD4 T cell populations. Annu. Rev. Immunol. (2010), 28(December 2009):445–89. doi:10.1146/annurev-immunol-030409-101212

8. Constant, S.L. & Bottomly, K. Induction of Th1 and Th2 CD4+ T cell responses: the alternative approaches. Annu. Rev. Immunol. (1997), 15297–322. doi:10.1146/annurev.immunol.15.1.297

9. Zhu, J. & Paul, W.E. CD4 T cells: fates, functions, and faults. Blood (2008), 112(5):1557–69. doi:10.1182/blood-2008-05-078154

10. Restifo, N.P. & Gattinoni, L. Lineage relationship of effector and memory T cells. Curr. Opin. Immunol. (2013), 25(5):556–563. doi:10.1016/j.coi.2013.09.003

11. Grakoui, A., Bromley, S.K., Sumen, C., Davis, M.M., Shaw, A.S., Allen, P.M. & Dustin, M.L. The Immunological Synapse: A Molecular Machine Controlling T Cell Activation. Science (80-. ). (1999), 285(5425):221–7. doi:10.1126/science.285.5425.221

12. Huppa, J.B. & Davis, M.M. T-cell-antigen recognition and the immunological synapse. Nat. Rev. Immunol. (2003), 3(12):973–83. doi:10.1038/nri1245

13. Kupfer, A., Swain, S.L. & Singer, S.J. The specific direct interaction of helper T cells and antigen- presenting B cells. II. Reorientation of the microtubule organizing center and reorganization of the membrane-associated cytoskeleton inside the bound helper T cells. J. Exp. Med. (1987), 165(6):1565– 80.

14. Cullinan, P., Sperling, A.I. & Burkhardt, J.K. The distal pole complex: a novel membrane domain distal to the immunological synapse. Immunol. Rev. (2002), 189(3):111–22.

15. Rossy, J., Williamson, D.J., Benzing, C. & Gaus, K. The integration of signaling and the spatial organization of the T cell synapse. Front. Immunol. (2012), 3352. doi:10.3389/fimmu.2012.00352

16. Bunnell, S.C., Hong, D.I., Kardon, J.R., Yamazaki, T., McGlade, C.J., Barr, V.A. & Samelson, L.E. T cell receptor ligation induces the formation of dynamically regulated signaling assemblies. J. Cell Biol. (2002), 158(7):1263–75.

17. Bunnell, S.C., Singer, A.L., Hong, D.I., Jacque, B.H., Jordan, M.S., Seminario, M.-C., Barr, V.A., Koretzky, G.A. & Samelson, L.E. Persistence of cooperatively stabilized signaling clusters drives T-cell activation. Mol. Cell. Biol. (2006), 26(19):7155–66. doi:10.1128/MCB.00507-06

18. Su, Y.W., Zhang, Y., Schweikert, J., Koretzky, G. a, Reth, M. & Wienands, J. Interaction of SLP adaptors with the SH2 domain of Tec family kinases. Eur. J. Immunol. (1999), 29(11):3702–11. doi:10.1002/(SICI)1521-4141(199911)29:11<3702::AID-IMMU3702>3.0.CO;2-R

19. Wu, J., Motto, D.G., Koretzky, G. a & Weiss, a Vav and SLP-76 interact and functionally cooperate in IL-2 gene activation. Immunity (1996), 4(6):593–602.

20. Sommers, C.L., Samelson, L.E. & Love, P.E. LAT: a T lymphocyte adapter protein that couples the antigen receptor to downstream signaling pathways. Bioessays (2004), 26(1):61–7. doi:10.1002/bies.10384

21. Paccani, S.R., Boncristiano, M., Patrussi, L., Ulivieri, C., Wack, A., Valensin, S., Hirst, T.R., Amedei, A., Del Prete, G., Telford, J.L., D’Elios, M.M. & Baldari, C.T. Defective Vav expression and impaired F-actin reorganization in a subset of patients with common variable immunodeficiency characterized by T- cell defects. Blood (2005), 106(2):626–34. doi:10.1182/blood-2004-05-2051

22. Fischer, K.D., Kong, Y.Y., Nishina, H., Tedford, K., Marengère, L.E., Kozieradzki, I., Sasaki, T., Starr, M., Chan, G., Gardener, S., Nghiem, M.P., Bouchard, D., Barbacid, M., Bernstein, A. & Penninger, J.M. Vav is a regulator of cytoskeletal reorganization mediated by the T-cell receptor. Curr. Biol. (1998), 8(10):554– 62. doi:10.1016/S0960-9822(98)70224-6

23. Dombroski, D., Houghtling, R.A., Labno, C.M., Precht, P., Takesono, A., Caplen, N.J., Billadeau, D.D., Wange, R.L., Burkhardt, J.K. & Schwartzberg, P.L. Kinase-independent functions for Itk in TCR-induced regulation of Vav and the actin cytoskeleton. J. … (2005), 174(3):1385–1392. doi:174/3/1385 [pii]

24. Woods, M.L., Kivens, W.J., Adelsman, M.A., Qiu, Y., August, A. & Shimizu, Y. A novel function for the Tec family tyrosine kinase Itk in activation of beta 1 integrins by the T-cell receptor. EMBO J. (2001), 20(6):1232–44. doi:10.1093/emboj/20.6.1232

25. Knyazhitsky, M., Moas, E., Shaginov, E., Luria, A. & Braiman, A. Vav1 oncogenic mutation inhibits T cell receptor-induced calcium mobilization through inhibition of phospholipase Cγ1 activation. J. Biol. Chem. (2012), 287(23):19725–35. doi:10.1074/jbc.M111.309799

26. Reynolds, L.F., Smyth, L.A., Norton, T., Freshney, N., Downward, J., Kioussis, D. & Tybulewicz, V.L. Vav1 transduces T cell receptor signals to the activation of phospholipase C-gamma1 via phosphoinositide 3-kinase-dependent and -independent pathways. J Exp Med (2002), 195(9):1103–1114.

27. Braiman, A., Barda-Saad, M., Sommers, C.L. & Samelson, L.E. Recruitment and activation of PLCgamma1 in T cells: a new insight into old domains. EMBO J. (2006), 25(4):774–84. doi:10.1038/sj.emboj.7600978

28. Smith-Garvin, J.E., Koretzky, G.A. & Jordan, M.S. T cell activation. Annu. Rev. Immunol. (2009), 27591– 619. doi:10.1146/annurev.immunol.021908.132706

29. Barda-Saad, M., Braiman, A., Titerence, R., Bunnell, S.C., Barr, V. a & Samelson, L.E. Dynamic molecular interactions linking the T cell antigen receptor to the actin cytoskeleton. Nat. Immunol. (2005), 6(1):80–9. doi:10.1038/ni1143

30. Gil, D., Schamel, W.W.A., Montoya, M., Sánchez-Madrid, F. & Alarcón, B. Recruitment of Nck by CD3 epsilon reveals a ligand-induced conformational change essential for T cell receptor signaling and synapse formation. Cell (2002), 109(7):901–12.

31. Matalon, O., Reicher, B. & Barda-Saad, M. Wiskott-Aldrich syndrome protein - dynamic regulation of actin homeostasis: from activation through function and signal termination in T lymphocytes. Immunol. Rev. (2013), 256(1):10–29. doi:10.1111/imr.12112

32. Saxe, C.L. WASp/Scar/WAVE. Handb. Cell Signal. (2003), 323–327.

33. O’Sullivan, E., Kinnon, C. & Brickell, P. Wiskott-Aldrich Syndrome protein, WASP. Int. J. Biochem. Cell Biol. (1999), 31(3-4):383–387. doi:Doi: 10.1016/s1357-2725(98)00118-6

34. Thrasher, A.J. & Burns, S.O. WASP: a key immunological multitasker. Nat. Rev. Immunol. (2010), 10(3):182–92. doi:10.1038/nri2724

35. Westerberg, L.S., Meelu, P., Baptista, M., Eston, M. a, Adamovich, D. a, Cotta-de-Almeida, V., Seed, B., Rosen, M.K., Vandenberghe, P., Thrasher, A.J., Klein, C., Alt, F.W. & Snapper, S.B. Activating WASP mutations associated with X-linked neutropenia result in enhanced actin polymerization, altered cytoskeletal responses, and genomic instability in lymphocytes. J. Exp. Med. (2010), 207(6):1145–52. doi:10.1084/jem.20091245

36. Ochs, H.D. & Thrasher, A.J. The Wiskott-Aldrich syndrome. J. Allergy Clin. Immunol. (2006), 117(4):725–38; quiz 739. doi:10.1016/j.jaci.2006.02.005

37. Wang, H. & Rudd, C.E. SKAP-55, SKAP-55-related and ADAP adaptors modulate integrin-mediated immune-cell adhesion. Trends Cell Biol. (2008), 18(10):486–93. doi:10.1016/j.tcb.2008.07.005

38. Burbach, B.J., Srivastava, R., Medeiros, R.B., O’Gorman, W.E., Peterson, E.J. & Shimizu, Y. Distinct regulation of integrin-dependent T cell conjugate formation and NF-kappa B activation by the adapter protein ADAP. J. Immunol. (2008), 181(7):4840–51.

39. Ophir, M.J., Liu, B.C. & Bunnell, S.C. The N terminus of SKAP55 enables T cell adhesion to TCR and integrin ligands via distinct mechanisms. J. Cell Biol. (2013), 203(6):1021–41. doi:10.1083/jcb.201305088

40. Hayashi, K. & Altman, A. Protein kinase C theta (PKCtheta): a key player in T cell life and death. Pharmacol. Res. (2007), 55(6):537–44. doi:10.1016/j.phrs.2007.04.009

41. Vallabhapurapu, S. & Karin, M. Regulation and function of NF-kappaB transcription factors in the immune system. Annu. Rev. Immunol. (2009), 27693–733. doi:10.1146/annurev.immunol.021908.132641

42. Zhang, W., Trible, R.P., Zhu, M., Liu, S.K., McGlade, C.J. & Samelson, L.E. Association of Grb2, Gads, and phospholipase C-gamma 1 with phosphorylated LAT tyrosine residues. Effect of LAT tyrosine mutations on T cell angigen receptor-mediated signaling. J. Biol. Chem. (2000), 275(30):23355–61. doi:10.1074/jbc.M000404200

43. Genot, E. & Cantrell, D. a Ras regulation and function in lymphocytes. Curr. Opin. Immunol. (2000), 12(3):289–94.

44. Brownlie, R.J. & Zamoyska, R. T cell receptor signalling networks: branched, diversified and bounded. Nat. Rev. Immunol. (2013), 13(4):257–69. doi:10.1038/nri3403

45. Zhang, J., Shehabeldin, A., da Cruz, L.A., Butler, J., Somani, A.K., McGavin, M., Kozieradzki, I., dos Santos, A., Nagy, A., Grinstein, S., Penninger, J.M. & Siminovitch, K.A. Antigen receptor-induced activation and cytoskeletal rearrangement are impaired in Wiskott-Aldrich syndrome protein-deficient lymphocytes. J. Exp. Med. (1999), 190(9):1329–42.

46. Shui, J.-W., Boomer, J.S., Han, J., Xu, J., Dement, G. a, Zhou, G. & Tan, T.-H. Hematopoietic progenitor kinase 1 negatively regulates T cell receptor signaling and T cell-mediated immune responses. Nat. Immunol. (2007), 8(1):84–91. doi:10.1038/ni1416

47. Di Bartolo, V., Montagne, B., Salek, M., Jungwirth, B., Carrette, F., Fourtane, J., Sol-Foulon, N., Michel, F., Schwartz, O., Lehmann, W.D. & Acuto, O. A novel pathway down-modulating T cell activation involves HPK-1-dependent recruitment of 14-3-3 proteins on SLP-76. J. Exp. Med. (2007), 204(3):681–91. doi:10.1084/jem.20062066

48. Bachmaier, K., Krawczyk, C., Kozieradzki, I., Kong, Y.Y., Sasaki, T., Oliveira-dos-Santos, a, Mariathasan, S., Bouchard, D., Wakeham, a, Itie, a, Le, J., Ohashi, P.S., Sarosi, I., Nishina, H., Lipkowitz, S. & Penninger, J.M. Negative regulation of lymphocyte activation and autoimmunity by the molecular adaptor Cbl-b. Nature (2000), 403(6766):211–6. doi:10.1038/35003228

49. Naramura, M., Jang, I.-K., Kole, H., Huang, F., Haines, D. & Gu, H. C-Cbl and Cbl-b regulate T cell responsiveness by promoting ligand-induced TCR down-modulation. Nat. Immunol. (2002), 3(12):1192–9. doi:10.1038/ni855

50. Murphy, M.A., Schnall, R.G., Venter, D.J., Barnett, L., Bertoncello, I., Thien, C.B.F., Langdon, W.Y. & Bowtell, D.D.L. Tissue Hyperplasia and Enhanced T-Cell Signalling via ZAP-70 in c-Cbl-Deficient Mice Tissue Hyperplasia and Enhanced T-Cell Signalling via ZAP-70 in c-Cbl-Deficient Mice. (1998),

51. Monks, C.R., Freiberg, B.A., Kupfer, H., Sciaky, N. & Kupfer, A. Three-dimensional segregation of supramolecular activation clusters in T cells. Nature (1998), 395(6697):82–6. doi:10.1038/25764

52. Freiberg, B.A., Kupfer, H., Maslanik, W., Delli, J., Kappler, J., Zaller, D.M. & Kupfer, A. Staging and resetting T cell activation in SMACs. Nat. Immunol. (2002), 3(10):911–7. doi:10.1038/ni836

53. Rossy, J., Owen, D.M., Williamson, D.J., Yang, Z. & Gaus, K. Conformational states of the kinase Lck regulate clustering in early T cell signaling. Nat. Immunol. (2013), 14(1):82–9. doi:10.1038/ni.2488

54. Schubert, D.A., Gordo, S., Sabatino, J.J., Vardhana, S., Gagnon, E., Sethi, D.K., Seth, N.P., Choudhuri, K., Reijonen, H., Nepom, G.T., Evavold, B.D., Dustin, M.L. & Wucherpfennig, K.W. Self-reactive human CD4 T cell clones form unusual immunological synapses. J. Exp. Med. (2012), 209(2):335–52. doi:10.1084/jem.20111485

55. Cemerski, S., Das, J., Giurisato, E., Markiewicz, M.A., Allen, P.M., Chakraborty, A.K. & Shaw, A.S. The balance between T cell receptor signaling and degradation at the center of the immunological synapse is determined by antigen quality. Immunity (2008), 29(3):414–22. doi:10.1016/j.immuni.2008.06.014

56. Varma, R., Campi, G., Yokosuka, T., Saito, T. & Dustin, M.L. T cell receptor-proximal signals are sustained in peripheral microclusters and terminated in the central supramolecular activation cluster. Immunity (2006), 25(1):117–27. doi:10.1016/j.immuni.2006.04.010

57. Vardhana, S., Choudhuri, K., Varma, R. & Dustin, M.L. Essential role of ubiquitin and TSG101 protein in formation and function of the central supramolecular activation cluster. Immunity (2010), 32(4):531– 40. doi:10.1016/j.immuni.2010.04.005

58. Cantrell, D.A. & Smith, K.A. The interleukin-2 T-cell system: a new cell growth model. Science (1984), 224(4655):1312–6.

59. Kalia, V., Sarkar, S., Subramaniam, S., Haining, W.N., Smith, K.A. & Ahmed, R. Prolonged interleukin- 2Ralpha expression on virus-specific CD8+ T cells favors terminal-effector differentiation in vivo. Immunity (2010), 32(1):91–103. doi:10.1016/j.immuni.2009.11.010

60. Zubler, R.H., Lowenthal, J.W., Erard, F., Hashimoto, N., Devos, R. & MacDonald, H.R. Activated B cells express receptors for, and proliferate in response to, pure interleukin 2. J. … (1984), 160(October):1170–1183.

61. Henney, C.S.C., Kuribayashi, K., Kern, D.E. DE & Gillis, S. Interleukin-2 augments natural killer cell activity. Nature (1981), 291(5813):335–338. doi:10.1038/291335a0

62. Boyman, O. & Sprent, J. The role of interleukin-2 during homeostasis and activation of the immune system. Nat. Rev. Immunol. (2012), 12(3):180–90. doi:10.1038/nri3156

63. Boyman, O., Purton, J.F., Surh, C.D. & Sprent, J. Cytokines and T-cell homeostasis. Curr. Opin. Immunol. (2007), 19(3):320–326. doi:10.1016/j.coi.2007.04.015

64. Malek, T.R. The biology of interleukin-2. Annu. Rev. Immunol. (2008), 26453–79. doi:10.1146/annurev.immunol.26.021607.090357

65. Setoguchi, R., Hori, S., Takahashi, T. & Sakaguchi, S. Homeostatic maintenance of natural Foxp3(+) CD25(+) CD4(+) regulatory T cells by interleukin (IL)-2 and induction of autoimmune disease by IL-2 neutralization. J. Exp. Med. (2005), 201(5):723–35. doi:10.1084/jem.20041982

66. Malek, T.R. & Castro, I. Interleukin-2 receptor signaling: at the interface between tolerance and immunity. Immunity (2010), 33(2):153–65. doi:10.1016/j.immuni.2010.08.004

67. Wang, X., Rickert, M. & Garcia, K.C. Structure of the quaternary complex of interleukin-2 with its alpha, beta, and gammac receptors. Science (2005), 310(5751):1159–63. doi:10.1126/science.1117893

68. Sharma, R., Zheng, L., Deshmukh, U.S., Jarjour, W.N., Sung, S.-S.J., Fu, S.M. & Ju, S.-T. A regulatory T cell- dependent novel function of CD25 (IL-2Ralpha) controlling memory CD8(+) T cell homeostasis. J. Immunol. (2007), 178(3):1251–5.

69. Wuest, S.C., Edwan, J.H., Martin, J.F., Han, S., Perry, J.S. a, Cartagena, C.M., Matsuura, E., Maric, D., Waldmann, T. a & Bielekova, B. A role for interleukin-2 trans-presentation in dendritic cell-mediated T cell activation in humans, as revealed by daclizumab therapy. Nat. Med. (2011), 17(5):604–9. doi:10.1038/nm.2365

70. Rubin, L.A. & Nelson, D.L. The soluble interleukin-2 receptor: biology, function, and clinical application. Ann. Intern. Med. (1990), 113(8):619–27.

71. Taniguchi, T. & Minami, Y. The IL-2/IL-2 receptor system: a current overview. Cell (1993), 73(1):5–8.

72. Pillet, A.-H., Lavergne, V., Pasquier, V., Gesbert, F., Thèze, J. & Rose, T. IL-2 induces conformational changes in its preassembled receptor core, which then migrates in lipid raft and binds to the cytoskeleton meshwork. J. Mol. Biol. (2010), 403(5):671–92. doi:10.1016/j.jmb.2010.08.056

73. Martins, G. a, Cimmino, L., Liao, J., Magnusdottir, E. & Calame, K. Blimp-1 directly represses Il2 and the Il2 activator Fos, attenuating T cell proliferation and survival. J. Exp. Med. (2008), 205(9):1959–65. doi:10.1084/jem.20080526

74. Zmuidzinas, A., Mamon, H.J., Roberts, T.M. & Smith, K.A. Interleukin-2-triggered Raf-1 expression, phosphorylation, and associated kinase activity increase through G1 and S in CD3-stimulated primary human T cells. Mol. Cell. Biol. (1991), 11(5):2794–803.

75. Moon, J.J., Rubio, E.D., Martino, A., Krumm, A. & Nelson, B.H. A permissive role for phosphatidylinositol 3-kinase in the Stat5-mediated expression of cyclin D2 by the interleukin-2 receptor. J. Biol. Chem. (2004), 279(7):5520–7. doi:10.1074/jbc.M308998200

76. Zhou, Y.J., Magnuson, K.S., Cheng, T.P., Gadina, M., Frucht, D.M., Galon, J., Candotti, F., Geahlen, R.L., Changelian, P.S. & O’Shea, J.J. Hierarchy of protein tyrosine kinases in interleukin-2 (IL-2) signaling: activation of syk depends on Jak3; however, neither Syk nor Lck is required for IL-2-mediated STAT activation. Mol. Cell. Biol. (2000), 20(12):4371–80. doi:10.1128/MCB.20.12.4371-4380.2000

77. Welte, T., Leitenberg, D., Dittel, B.N., Al-Ramadi, B.K., Xie, B., Chin, Y.E., Janeway, C.A., Bothwell, A.L., Bottomly, K. & Fu, X.Y. STAT5 interaction with the T cell receptor complex and stimulation of T cell proliferation. Science (1999), 283(5399):222–5. doi:10.1126/science.283.5399.222

78. Miyazaki, T., Liu, Z.J., Kawahara, A., Minami, Y., Yamada, K., Tsujimoto, Y., Barsoumian, E. Lo, Permutter, R.M. & Taniguchi, T. Three distinct IL-2 signaling pathways mediated by bcl-2, c-myc, and lck cooperate in hematopoietic cell proliferation. Cell (1995), 81(2):223–31.

79. Wallner, S. & Schmitz, G. Plasmalogens the neglected regulatory and scavenging lipid species. Chem. Phys. Lipids (2011), 164(6):573–89. doi:10.1016/j.chemphyslip.2011.06.008

80. Gorgas, K., Teigler, A., Komljenovic, D. & Just, W.W. The ether lipid-deficient mouse: tracking down plasmalogen functions. Biochim. Biophys. Acta (2006), 1763(12):1511–26. doi:10.1016/j.bbamcr.2006.08.038

81. Stillwell, W. Membranes and Human Health. An Introd. to Biol. Membr. (2013), 339–355. doi:10.1016/B978-0-444-52153-8.00015-5

82. Shaikh, S.R. & Edidin, M. Polyunsaturated fatty acids, membrane organization, T cells, and antigen presentation. Am. J. Clin. Nutr. (2006), 84(6):1277–89.

83. Segura, A., Duque, E., Mosqueda, G., Ramos, J.L. & Junker, F. Multiple responses of Gram-negative bacteria to organic solvents. Environ. Microbiol. (1999), 1(3):191–198. doi:10.1046/j.1462- 2920.1999.00033.x

84. Weber, F.J. & de Bont, J.A.M. Adaptation mechanisms of microorganisms to the toxic effects of organic solvents on membranes. Biochim. Biophys. Acta - Rev. Biomembr. (1996), 1286(3):225–245. doi:10.1016/S0304-4157(96)00010-X

85. Pinkart, H. & White, D. Phospholipid biosynthesis and solvent tolerance in Pseudomonas putida strains. J. Bacteriol. (1997), 179(13):

86. Lippincott-Schwartz, J. & Phair, R.D. Lipids and cholesterol as regulators of traffic in the endomembrane system. Annu. Rev. Biophys. (2010), 39559–78. doi:10.1146/annurev.biophys.093008.131357

87. Te Vruchte, D., Lloyd-Evans, E., Veldman, R.J., Neville, D.C.A., Dwek, R.A., Platt, F.M., van Blitterswijk, W.J. & Sillence, D.J. Accumulation of glycosphingolipids in Niemann-Pick C disease disrupts endosomal transport. J. Biol. Chem. (2004), 279(25):26167–75. doi:10.1074/jbc.M311591200

88. Goldstein, D.B. The effects of drugs on membrane fluidity. Annu. Rev. Pharmacol. Toxicol. (1984), 2443–64. doi:10.1146/annurev.pa.24.040184.000355

89. Van Meer, G., Voelker, D.R. & Feigenson, G.W. Membrane lipids: where they are and how they behave. Nat. Rev. Mol. Cell Biol. (2008), 9(2):112–24. doi:10.1038/nrm2330

90. Cooper, R.A. Influence of increased membrane cholesterol on membrane fluidity and cell function in human red blood cells. J. Supramol. Struct. (1978), 8(4):413–30. doi:10.1002/jss.400080404

91. Daleke, D.L. Phospholipid flippases. J. Biol. Chem. (2007), 282(2):821–5. doi:10.1074/jbc.R600035200

92. Muthusamy, B.-P., Natarajan, P., Zhou, X. & Graham, T.R. Linking phospholipid flippases to vesicle- mediated protein transport. Biochim. Biophys. Acta (2009), 1791(7):612–9. doi:10.1016/j.bbalip.2009.03.004

93. Pomorski, T. & Menon, a K. Lipid flippases and their biological functions. Cell. Mol. Life Sci. (2006), 63(24):2908–21. doi:10.1007/s00018-006-6167-7

94. Bevers, E.M. & Williamson, P.L. Phospholipid scramblase: an update. FEBS Lett. (2010), 584(13):2724– 30. doi:10.1016/j.febslet.2010.03.020

95. Verhoven, B., Schlegel, R.A. & Williamson, P. Mechanisms of phosphatidylserine exposure, a phagocyte recognition signal, on apoptotic T lymphocytes. J. Exp. Med. (1995), 182(5):1597–601.

96. Fairn, G.D., Schieber, N.L., Ariotti, N., Murphy, S., Kuerschner, L., Webb, R.I., Grinstein, S. & Parton, R.G. High-resolution mapping reveals topologically distinct cellular pools of phosphatidylserine. J. Cell Biol. (2011), 194(2):257–75. doi:10.1083/jcb.201012028

97. Demel, R.A., Jansen, J.W., van Dijck, P.W. & van Deenen, L.L. The preferential interaction of cholesterol with different classes of phospholipids. Biochim. Biophys. Acta (1977), 465(1):1–10.

98. Arikketh, D., Nelson, R. & Vance, J.E. Defining the importance of phosphatidylserine synthase-1 (PSS1): unexpected viability of PSS1-deficient mice. J. Biol. Chem. (2008), 283(19):12888–97. doi:10.1074/jbc.M800714200

99. Vance, J.E. & Tasseva, G. Formation and function of phosphatidylserine and phosphatidylethanolamine in mammalian cells. Biochim. Biophys. Acta (2013), 1831(3):543–54. doi:10.1016/j.bbalip.2012.08.016

100. Verdaguer, N., Corbalan-Garcia, S., Ochoa, W.F., Fita, I. & Gómez-Fernández, J.C. Ca(2+) bridges the C2 membrane-binding domain of protein kinase Calpha directly to phosphatidylserine. EMBO J. (1999), 18(22):6329–38. doi:10.1093/emboj/18.22.6329

101. Huang, B.X., Akbar, M., Kevala, K. & Kim, H.-Y. Phosphatidylserine is a critical modulator for Akt activation. J. Cell Biol. (2011), 192(6):979–92. doi:10.1083/jcb.201005100

102. Papahadjopoulos, D., Cowden, M. & Kimelberg, H. Role of cholesterol in membranes. Effects on phospholipid-protein interactions, membrane permeability and enzymatic activity. Biochim. Biophys. Acta (1973), 330(1):8–26.

103. Grunze, M. & Deuticke, B. Changes of membrane permeability due to extensive cholesterol depletion in mammalian erythrocytes. Biochim. Biophys. Acta - Biomembr. (1974), 356(1):125–30.

104. Papahadjopoulos, D., Nir, S. & Oki, S. Permeability properties of phospholipid membranes: effect of cholesterol and temperature. Biochim. Biophys. Acta - Biomembr. (1972), 266(3):561–83.

105. Yeagle, P.L. Cholesterol and the cell membrane. Biochim. Biophys. Acta (1985), 822(3-4):267–87.

106. Estep, T.N., Mountcastle, D.B., Biltonen, R.L. & Thompson, T.E. Studies on the anomalous thermotropic behavior of aqueous dispersions of dipalmitoylphosphatidylcholine-cholesterol mixtures. Biochemistry (1978), 17(10):1984–9.

107. Maxfield, F.R. & van Meer, G. Cholesterol, the central lipid of mammalian cells. Curr. Opin. Cell Biol. (2010), 22(4):422–9. doi:10.1016/j.ceb.2010.05.004

108. Ikonen, E. Cellular cholesterol trafficking and compartmentalization. Nat. Rev. Mol. Cell Biol. (2008), 9(2):125–38. doi:10.1038/nrm2336

109. Hu, J., Zhang, Z., Shen, W.-J. & Azhar, S. Cellular cholesterol delivery, intracellular processing and utilization for biosynthesis of steroid hormones. Nutr. Metab. (Lond). (2010), 7(Table 1):47. doi:10.1186/1743-7075-7-47

110. Mesmin, B. & Maxfield, F.R. Intracellular sterol dynamics. Biochim. Biophys. Acta (2009), 1791(7):636– 45. doi:10.1016/j.bbalip.2009.03.002

111. Resh, M.D. Trafficking and signaling by fatty-acylated and prenylated proteins. Nat. Chem. Biol. (2006), 2(11):584–90. doi:10.1038/nchembio834

112. Bohdanowicz, M. & Grinstein, S. Role of phospholipids in endocytosis, phagocytosis, and macropinocytosis. Physiol. Rev. (2013), 93(1):69–106. doi:10.1152/physrev.00002.2012

113. Fairn, G.D. & Grinstein, S. Cell biology. Precursor or charge supplier? Science (2012), 337(6095):653– 4. doi:10.1126/science.1227096

114. Berg, J.M., Tymoczko, J.L. & Stryer, L. Biochemistry. (W. H. Freeman & Co: New York, 2002).

115. Calder, P.C. Immunomodulation by omega-3 fatty acids. Prostaglandins. Leukot. Essent. Fatty Acids (2007), 77(5-6):327–35. doi:10.1016/j.plefa.2007.10.015

116. Singer, S.J. & Nicolson, G.L. The fluid mosaic model of the structure of cell membranes. Science (1972), 175(4023):720–31.

117. Robertson, J.D. Membrane structure. J. Cell Biol. (1981), 91(3 Pt 2):189s–204s.

118. Nicolson, G.L. The Fluid-Mosaic Model of Membrane Structure: Still relevant to understanding the structure, function and dynamics of biological membranes after more than 40years. Biochim. Biophys. Acta (2013), doi:10.1016/j.bbamem.2013.10.019

119. Saffman, P.G. & Delbrück, M. Brownian motion in biological membranes. Proc. Natl. Acad. Sci. U. S. A. (1975), 72(8):3111–3.

120. Edidin, M., Zagyansky, Y. & Lardner, T.J. Measurement of membrane protein lateral diffusion in single cells. Science (1976), 191(4226):466–8.

121. Jacobson, K., Wu, E. & Poste, G. Measurement of the translational mobility of concanavalin A in glycerol-saline solutions and on the cell surface by fluorescence recovery after photobleaching. Biochim. Biophys. Acta (1976), 433(1):215–22.

122. Schlessinger, J., Axelrod, D., Koppel, D.E., Webb, W.W. & Elson, E.L. Lateral transport of a lipid probe and labeled proteins on a cell membrane. Science (1977), 195(4275):307–9.

123. Fowler, V. & Branton, D. Lateral mobility of human erythrocyte integral membrane proteins. Nature (1977), 268(5615):23–6.

124. Sheetz, M.P., Schindler, M. & Koppel, D.E. Lateral mobility of integral membrane proteins is increased in spherocytic erythrocytes. Nature (1980), 285(5765):510–1.

125. Nicolson, G.L. Transmembrane control of the receptors on normal and tumor cells. I. Cytoplasmic influence over surface components. Biochim. Biophys. Acta (1976), 457(1):57–108.

126. Elgsaeter, A. & Branton, D. Intramembrane particle aggregation in erythrocyte ghosts. I. The effects of protein removal. J. Cell Biol. (1974), 63(3):1018–36.

127. Edelman, G.M. Surface modulation in cell recognition and cell growth. Science (1976), 192(4236):218– 26.

128. Ash, J.F., Louvard, D. & Singer, S.J. Antibody-induced linkages of plasma membrane proteins to intracellular actomyosin-containing filaments in cultured fibroblasts. Proc. Natl. Acad. Sci. U. S. A. (1977), 74(12):5584–8.

129. Dupuy, A.D. & Engelman, D.M. Protein area occupancy at the center of the red blood cell membrane. Proc. Natl. Acad. Sci. U. S. A. (2008), 105(8):2848–52. doi:10.1073/pnas.0712379105

130. Lingwood, D. & Simons, K. Lipid rafts as a membrane-organizing principle. Science (2010), 327(5961):46–50. doi:10.1126/science.1174621

131. Viola, A. & Gupta, N. Tether and trap: regulation of membrane-raft dynamics by actin-binding proteins. Nat. Rev. Immunol. (2007), 7(11):889–96. doi:10.1038/nri2193

132. Pike, L.J. Rafts defined: a report on the Keystone Symposium on Lipid Rafts and Cell Function. J. Lipid Res. (2006), 47(7):1597–8. doi:10.1194/jlr.E600002-JLR200

133. Chapelle, S. & Gilles-Baillien, M. Phospholipids and cholesterol in brush border and basolateral membranes from rat intestinal mucosa. Biochim. Biophys. Acta (1983), 753(2):269–71.

134. Humphries, G.M. & Lovejoy, J.P. Lateral phase separation of phospholipids as a basis for increased permeability of membranes towards fluorescein and other chemical species. J. Membr. Biol. (1984), 80(3):249–56.

135. Lange, Y., Matthies, H. & Steck, T.L. Cholesterol oxidase susceptibility of the red cell membrane. Biochim. Biophys. Acta (1984), 769(3):551–62.

136. Heerklotz, H. Triton promotes domain formation in lipid raft mixtures. Biophys. J. (2002), 83(5):2693– 701. doi:10.1016/S0006-3495(02)75278-8

137. Rietveld, A. & Simons, K. The differential miscibility of lipids as the basis for the formation of functional membrane rafts. Biochim. Biophys. Acta (1998), 1376(3):467–79.

138. Schuck, S., Honsho, M., Ekroos, K., Shevchenko, A. & Simons, K. Resistance of cell membranes to different detergents. Proc. Natl. Acad. Sci. U. S. A. (2003), 100(10):5795–800. doi:10.1073/pnas.0631579100

139. Parasassi, T., Conti, F. & Gratton, E. Time-resolved fluorescence emission spectra of Laurdan in phospholipid vesicles by multifrequency phase and modulation fluorometry. Cell. Mol. Biol. (1986), 32(1):103–8.

140. Owen, D.M., Rentero, C., Magenau, A., Abu-Siniyeh, A. & Gaus, K. Quantitative imaging of membrane lipid order in cells and organisms. Nat. Protoc. (2012), 7(1):24–35. doi:10.1038/nprot.2011.419

141. Edidin, M., Zúñiga, M.C. & Sheetz, M.P. Truncation mutants define and locate cytoplasmic barriers to lateral mobility of membrane glycoproteins. Proc. Natl. Acad. Sci. U. S. A. (1994), 91(8):3378–82.

142. Sako, Y. & Kusumi, a Barriers for lateral diffusion of transferrin receptor in the plasma membrane as characterized by receptor dragging by laser tweezers: fence versus tether. J. Cell Biol. (1995), 129(6):1559–74.

143. Sako, Y., Nagafuchi, a, Tsukita, S., Takeichi, M. & Kusumi, a Cytoplasmic regulation of the movement of E-cadherin on the free cell surface as studied by optical tweezers and single particle tracking: corralling and tethering by the membrane skeleton. J. Cell Biol. (1998), 140(5):1227–40.

144. Kusumi, A., Suzuki, K.G.N., Kasai, R.S., Ritchie, K. & Fujiwara, T.K. Hierarchical mesoscale domain organization of the plasma membrane. Trends Biochem. Sci. (2011), 36(11):604–15. doi:10.1016/j.tibs.2011.08.001

145. Kusumi, A., Fujiwara, T.K., Chadda, R., Xie, M., Tsunoyama, T. a, Kalay, Z., Kasai, R.S. & Suzuki, K.G.N. Dynamic organizing principles of the plasma membrane that regulate signal transduction: commemorating the fortieth anniversary of Singer and Nicolson’s fluid-mosaic model. Annu. Rev. Cell Dev. Biol. (2012), 28215–50. doi:10.1146/annurev-cellbio-100809-151736

146. Kusumi, A., Shirai, Y.M., Koyama-Honda, I., Suzuki, K.G. & Fujiwara, T.K. Hierarchical organization of the plasma membrane: investigations by single-molecule tracking vs. fluorescence correlation spectroscopy. FEBS Lett. (2010), 584(9):1814–1823. doi:S0014-5793(10)00151-1 [pii] 10.1016/j.febslet.2010.02.047

147. Kusumi, A. & Sako, Y. Cell surface organization by the membrane skeleton. Curr. Opin. Cell Biol. (1996), 8(4):566–74.

148. Fujiwara, T., Ritchie, K., Murakoshi, H., Jacobson, K. & Kusumi, A. Phospholipids undergo hop diffusion in compartmentalized cell membrane. J. Cell Biol. (2002), 157(6):1071–81. doi:10.1083/jcb.200202050

149. Tsuji, A. & Ohnishi, S. Restriction of the lateral motion of band 3 in the erythrocyte membrane by the cytoskeletal network: dependence on spectrin association state. Biochemistry (1986), 25(20):6133–9.

150. Poste, G., Papahadjopoulos, D. & Nicolson, G.L. Local anesthetics affect transmembrane cytoskeletal control of mobility and distribution of cell surface receptors. Proc. Natl. Acad. Sci. U. S. A. (1975), 72(11):4430–4.

151. Lenne, P.-F., Wawrezinieck, L., Conchonaud, F., Wurtz, O., Boned, A., Guo, X.-J., Rigneault, H., He, H.-T. & Marguet, D. Dynamic molecular confinement in the plasma membrane by microdomains and the cytoskeleton meshwork. EMBO J. (2006), 25(14):3245–56. doi:10.1038/sj.emboj.7601214

152. Umemura, Y.M., Vrljic, M., Nishimura, S.Y., Fujiwara, T.K., Suzuki, K.G.N. & Kusumi, A. Both MHC class II and its GPI-anchored form undergo hop diffusion as observed by single-molecule tracking. Biophys. J. (2008), 95(1):435–50. doi:10.1529/biophysj.107.123018

153. Yin, H.L. & Janmey, P. a Phosphoinositide regulation of the actin cytoskeleton. Annu. Rev. Physiol. (2003), 65(1):761–89. doi:10.1146/annurev.physiol.65.092101.142517

154. Fehon, R.G., McClatchey, A.I. & Bretscher, A. Organizing the cell cortex: the role of ERM proteins. Nat. Rev. Mol. Cell Biol. (2010), 11(4):276–87. doi:10.1038/nrm2866

155. Macias, M.J., Musacchio, A., Ponstingl, H., Nilges, M., Saraste, M. & Oschkinat, H. Structure of the pleckstrin homology domain from beta-spectrin. Nature (1994), 369(6482):675–7. doi:10.1038/369675a0

156. Miki, H., Miura, K. & Takenawa, T. N-WASP, a novel actin-depolymerizing protein, regulates the cortical cytoskeletal rearrangement in a PIP2-dependent manner downstream of tyrosine kinases. EMBO J. (1996), 15(19):5326–35.

157. Rohatgi, R., Ma, L., Miki, H., Lopez, M., Kirchhausen, T., Takenawa, T. & Kirschner, M.W. The interaction between N-WASP and the Arp2/3 complex links Cdc42-dependent signals to actin assembly. Cell (1999), 97(2):221–31. doi:10.1016/S0092-8674(00)80732-1

158. Kapus, A. & Janmey, P. Plasma membrane--cortical cytoskeleton interactions: a cell biology approach with biophysical considerations. Compr. Physiol. (2013), 3(3):1231–81. doi:10.1002/cphy.c120015

159. Dinic, J., Ashrafzadeh, P. & Parmryd, I. Actin filaments attachment at the plasma membrane in live cells cause the formation of ordered lipid domains. Biochim. Biophys. Acta (2013), 1828(3):1102–11. doi:10.1016/j.bbamem.2012.12.004

160. Head, B.P., Patel, H.H. & Insel, P.A. Interaction of membrane/lipid rafts with the cytoskeleton: Impact on signaling and function: Membrane/lipid rafts, mediators of cytoskeletal arrangement and cell signaling. Biochim. Biophys. Acta (2014), 1838(2):532–45. doi:10.1016/j.bbamem.2013.07.018

161. Liu, A.P. & Fletcher, D. a Actin polymerization serves as a membrane domain switch in model lipid bilayers. Biophys. J. (2006), 91(11):4064–70. doi:10.1529/biophysj.106.090852

162. Edidin, M. Switching sides: the actin/membrane lipid connection. Biophys. J. (2006), 91(11):3963. doi:10.1529/biophysj.106.094078

163. Williamson, D.J. The role of LAT clusters in early immune cell signalling. Unsworks.unsw.edu.au (2011), Retrieved from http://unsworks.unsw.edu.au/fapi/datastream/unsworks:10201/SOURCE02

164. Garner, O.B. & Baum, L.G. Galectin-glycan lattices regulate cell-surface glycoprotein organization and signalling. Biochem. Soc. Trans. (2008), 36(Pt 6):1472–7. doi:10.1042/BST0361472

165. Stulnig, T.M., Berger, M., Sigmund, T., Raederstorff, D., Stockinger, H. & Waldhäusl, W. Polyunsaturated fatty acids inhibit T cell signal transduction by modification of detergent-insoluble membrane domains. J. Cell Biol. (1998), 143(3):637–44.

166. Li, Q., Wang, M., Tan, L., Wang, C., Ma, J., Li, N., Li, Y., Xu, G. & Li, J. Docosahexaenoic acid changes lipid composition and interleukin-2 receptor signaling in membrane rafts. J. Lipid Res. (2005), 46(9):1904– 13. doi:10.1194/jlr.M500033-JLR200

167. Goebel, J., Forrest, K., Morford, L. & Roszman, T.L. Differential localization of IL-2- and -15 receptor chains in membrane rafts of human T cells. J. Leukoc. Biol. (2002), 72(1):199–206.

168. Heiner, A.L., Gibbons, E., Fairbourn, J.L., Gonzalez, L.J., McLemore, C.O., Brueseke, T.J., Judd, A.M. & Bell, J.D. Effects of cholesterol on physical properties of human erythrocyte membranes: impact on susceptibility to hydrolysis by secretory phospholipase A2. Biophys. J. (2008), 94(8):3084–93. doi:10.1529/biophysj.107.118356

169. Allsopp, R.C., Lalo, U. & Evans, R.J. Lipid raft association and cholesterol sensitivity of P2X1-4 receptors for ATP: chimeras and point mutants identify intracellular amino-terminal residues involved in lipid regulation of P2X1 receptors. J. Biol. Chem. (2010), 285(43):32770–7. doi:10.1074/jbc.M110.148940

170. Zech, T., Ejsing, C.S., Gaus, K., de Wet, B., Shevchenko, A., Simons, K. & Harder, T. Accumulation of raft lipids in T-cell plasma membrane domains engaged in TCR signalling. EMBO J. (2009), 28(5):466–76. doi:10.1038/emboj.2009.6

171. Gaus, K., Chklovskaia, E., Fazekas de St Groth, B., Jessup, W. & Harder, T. Condensation of the plasma membrane at the site of T lymphocyte activation. J. Cell Biol. (2005), 171(1):121–31. doi:10.1083/jcb.200505047

172. Yokosuka, T., Sakata-Sogawa, K., Kobayashi, W., Hiroshima, M., Hashimoto-Tane, A., Tokunaga, M., Dustin, M.L. & Saito, T. Newly generated T cell receptor microclusters initiate and sustain T cell activation by recruitment of Zap70 and SLP-76. Nat. Immunol. (2005), 6(12):1253–62. doi:10.1038/ni1272

173. Varma, R., Campi, G., Yokosuka, T., Saito, T. & Dustin, M.L. T cell receptor-proximal signals are sustained in peripheral microclusters and terminated in the central supramolecular activation cluster. Immunity (2006), 25(1):117–27. doi:10.1016/j.immuni.2006.04.010

174. Gaus, K., Chklovskaia, E., Fazekas de St Groth, B., Jessup, W. & Harder, T. Condensation of the plasma membrane at the site of T lymphocyte activation. J. Cell Biol. (2005), 171(1):121–31. doi:10.1083/jcb.200505047

175. Rodgers, W., Farris, D. & Mishra, S. Merging complexes: properties of membrane raft assembly during lymphocyte signaling. Trends Immunol. (2005), 26(2):97–103. doi:10.1016/j.it.2004.11.016

176. Babich, A., Li, S., O’Connor, R.S., Milone, M.C., Freedman, B.D. & Burkhardt, J.K. F-actin polymerization and retrograde flow drive sustained PLCγ1 signaling during T cell activation. J. Cell Biol. (2012), 197(6):775–87. doi:10.1083/jcb.201201018

177. Campi, G., Varma, R. & Dustin, M.L. Actin and agonist MHC-peptide complex-dependent T cell receptor microclusters as scaffolds for signaling. J Exp Med (2005), 202(8):1031–1036. doi:10.1084/jem.20051182

178. Yu, Y., Smoligovets, A.A. & Groves, J.T. Modulation of T cell signaling by the actin cytoskeleton. J. Cell Sci. (2013), 126(Pt 5):1049–58. doi:10.1242/jcs.098210

179. Holsinger, L.J., Graef, I.A., Swat, W., Chi, T., Bautista, D.M., Davidson, L., Lewis, R.S., Alt, F.W. & Crabtree, G.R. Defects in actin-cap formation in Vav-deficient mice implicate an actin requirement for lymphocyte signal transduction. Curr. Biol. (1998), 8(10):563–72. doi:10.1016/S0960- 9822(98)70225-8

180. Villalba, M., Bi, K., Rodriguez, F., Tanaka, Y., Schoenberger, S. & Altman, A. Vav1/Rac-dependent actin cytoskeleton reorganization is required for lipid raft clustering in T cells. J. Cell Biol. (2001), 155(3):331–8. doi:10.1083/jcb.200107080

181. Hoessli, D.C., Poincelet, M. & Rungger-Brändle, E. Isolation of high-affinity murine interleukin 2 receptors as detergent-resistant membrane complexes. Eur. J. Immunol. (1990), 20(7):1497–503. doi:10.1002/eji.1830200714

182. Lamaze, C., Dujeancourt, A., Baba, T., Lo, C.G., Benmerah, A. & Dautry-Varsat, A. Interleukin 2 receptors and detergent-resistant membrane domains define a clathrin-independent endocytic pathway. Mol. Cell (2001), 7(3):661–71.

183. Kabouridis, P.S., Janzen, J., Magee, A.L. & Ley, S.C. Cholesterol depletion disrupts lipid rafts and modulates the activity of multiple signaling pathways in T lymphocytes. Eur. J. Immunol. (2000), 30(3):954–63. doi:10.1002/1521-4141(200003)30:3<954::AID-IMMU954>3.0.CO;2-Y

184. Mañes, S. & Martínez-A, C. Cholesterol domains regulate the actin cytoskeleton at the leading edge of moving cells. Trends Cell Biol. (2004), 14(6):275–8. doi:10.1016/j.tcb.2004.04.008

185. Pizzo, P., Giurisato, E., Tassi, M., Benedetti, A., Pozzan, T. & Viola, A. Lipid rafts and T cell receptor signaling: a critical re-evaluation. Eur. J. Immunol. (2002), 32(11):3082–91. doi:10.1002/1521- 4141(200211)32:11<3082::AID-IMMU3082>3.0.CO;2-2

186. Rentero, C., Zech, T., Quinn, C.M., Engelhardt, K., Williamson, D.J., Grewal, T., Jessup, W., Harder, T. & Gaus, K. Functional implications of plasma membrane condensation for T cell activation. PLoS One (2008), 3(5):e2262. doi:10.1371/journal.pone.0002262

187. Cornely, R., Grewal, T. & Gaus, K. The Actin Cytoskeleton and Membrane Organisation in T Lymphocytes. Cytoskelet. Hum. Dis. (2012), 103–121. doi:10.1007/978-1-61779-788-0_5

188. Brown, D.A. Lipid rafts, detergent-resistant membranes, and raft targeting signals. Physiology (Bethesda). (2006), 21(1):430–9. doi:10.1152/physiol.00032.2006

189. Del Pozo, M.A., Alderson, N.B., Kiosses, W.B., Chiang, H.-H., Anderson, R.G.W. & Schwartz, M.A. Integrins regulate Rac targeting by internalization of membrane domains. Science (2004), 303(5659):839–42. doi:10.1126/science.1092571

190. Lin, J. Localization of LAT in Glycolipid-enriched Microdomains Is Required for T cell Activation. J. Biol. Chem. (1999), 274(41):28861–28864. doi:10.1074/jbc.274.41.28861

191. Zhang, W., Trible, R.P. & Samelson, L.E. LAT palmitoylation: its essential role in membrane microdomain targeting and tyrosine phosphorylation during T cell activation. Immunity (1998), 9(2):239–46.

192. Glebov, O.O. & Nichols, B.J. Lipid raft proteins have a random distribution during localized activation of the T-cell receptor. Nat. Cell Biol. (2004), 6(3):238–43. doi:10.1038/ncb1103

193. Hundt, M., Harada, Y., De Giorgio, L., Tanimura, N., Zhang, W. & Altman, A. Palmitoylation-dependent plasma membrane transport but lipid raft-independent signaling by linker for activation of T cells. J. Immunol. (2009), 183(3):1685–94. doi:10.4049/jimmunol.0803921

194. Tanimura, N., Saitoh, S., Kawano, S., Kosugi, A. & Miyake, K. Palmitoylation of LAT contributes to its subcellular localization and stability. Biochem. Biophys. Res. Commun. (2006), 341(4):1177–83. doi:10.1016/j.bbrc.2006.01.076

195. Geyeregger, R., Zeyda, M., Zlabinger, G.J., Waldhäusl, W. & Stulnig, T.M. Polyunsaturated fatty acids interfere with formation of the immunological synapse. J. Leukoc. Biol. (2005), 77(5):680–8. doi:10.1189/jlb.1104687

196. Takeuchi, H., Kato, H. & Kurahashi, T. 2,4,6-Trichloroanisole is a potent suppressor of olfactory signal transduction. Proc. Natl. Acad. Sci. (2013), 2–7. doi:10.1073/pnas.1300764110

197. Tavolari, S., Munarini, A., Storci, G., Laufer, S., Chieco, P. & Guarnieri, T. The decrease of cell membrane fluidity by the non-steroidal anti-inflammatory drug Licofelone inhibits epidermal growth factor receptor signalling and triggers apoptosis in HCA-7 colon cancer cells. Cancer Lett. (2012), 321(2):187–94. doi:10.1016/j.canlet.2012.02.003

198. Moss, S.E. & Morgan, R.O. The annexins. Genome Biol. (2004), 5(4):219. doi:10.1186/gb-2004-5-4-219

199. Lizarbe, M.A., Barrasa, J.I., Olmo, N., Gavilanes, F. & Turnay, J. Annexin-phospholipid interactions. Functional implications. Int. J. Mol. Sci. (2013), 14(2):2652–83. doi:10.3390/ijms14022652

200. Kawasaki, H., Avila-Sakar, A.J., Creutz, C.E. & Kretsinger, R.H. The crystal structure of annexin VI indicates relative rotation of the two lobes upon membrane binding. Biochim. Biophys. Acta (1996), 1313(3):277–82.

201. Avila-Sakar, A.J., Kretsinger, R.H. & Creutz, C.E. Membrane-bound 3D structures reveal the intrinsic flexibility of annexin VI. J. Struct. Biol. (2000), 130(1):54–62. doi:10.1006/jsbi.2000.4246

202. Enrich, C., Rentero, C., de Muga, S.V., Reverter, M., Mulay, V., Wood, P., Koese, M. & Grewal, T. Annexin A6 - linking Ca(2+) signalling with cholesterol transport. Biochim. Biophys. Acta (2010), 1813(5):935– 47. doi:10.1016/j.bbamcr.2010.09.015

203. Dubois, T., Soula, M., Moss, S.E., Russo-Marie, F. & Rothhut, B. Potential interaction between annexin VI and a 56-kDa protein kinase in T cells. Biochem. Biophys. Res. Commun. (1995), 212(1):270–8.

204. Moss, S.E. & Crumpton, M.J. Alternative splicing gives rise to two forms of the p68 Ca2(+)-binding protein. FEBS Lett. (1990), 261(2):299–302.

205. Fleet, A., Ashworth, R., Kubista, H., Edwards, H., Bolsover, S., Mobbs, P. & Moss, S.E. Inhibition of EGF- dependent calcium influx by annexin VI is splice form-specific. Biochem. Biophys. Res. Commun. (1999), 260(2):540–6. doi:10.1006/bbrc.1999.0915

206. Su, A.I., Wiltshire, T., Batalov, S., Lapp, H., Ching, K. a, Block, D., Zhang, J., Soden, R., Hayakawa, M., Kreiman, G., Cooke, M.P., Walker, J.R. & Hogenesch, J.B. A gene atlas of the mouse and human protein- encoding transcriptomes. Proc. Natl. Acad. Sci. U. S. A. (2004), 101(16):6062–7. doi:10.1073/pnas.0400782101

207. Kaetzel, M.A., Pula, G., Campos, B., Uhrin, P., Horseman, N. & Dedman, J.R. Annexin VI isoforms are differentially expressed in mammalian tissues. Biochim. Biophys. Acta (1994), 1223(3):368–74.

208. Clark, D.M., Moss, S.E., Wright, N.A. & Crumpton, M.J. Expression of annexin VI (p68, 67 kDa- calelectrin) in normal human tissues: evidence for developmental regulation in B- and T-lymphocytes. Histochemistry (1991), 96(5):405–12.

209. Silva, F.G., Sherrill, K., Spurgeon, S., Südhof, T.C. & Stone, D.K. High-level expression of the 32.5- kilodalton calelectrin in ductal epithelia as revealed by immunocytochemistry. Differentiation. (1986), 33(2):175–83. doi:10.1111/j.1432-0436.1987.tb01554.x

210. Avila-Sakar, A.J., Creutz, C.E. & Kretsinger, R.H. Crystal structure of bovine annexin VI in a calcium- bound state. Biochim. Biophys. Acta (1998), 1387103–116.

211. Moll, A., Hildebrandt, A., Lenhof, H.-P. & Kohlbacher, O. BALLView: an object-oriented molecular visualization and modeling framework. J. Comput. Aided. Mol. Des. (2005), 19(11):791–800. doi:10.1007/s10822-005-9027-x

212. Grewal, T., Koese, M., Rentero, C. & Enrich, C. Annexin A6-regulator of the EGFR/Ras signalling pathway and cholesterol homeostasis. Int. J. Biochem. Cell Biol. (2010), 42(5):580–4. doi:10.1016/j.biocel.2009.12.020

213. Potez, S., Luginbühl, M., Monastyrskaya, K., Hostettler, A., Draeger, A. & Babiychuk, E.B. Tailored Protection against Plasmalemmal Injury by Annexins with Different Ca2+ Sensitivities. J. Biol. Chem. (2011), 286(20):17982–91. doi:10.1074/jbc.M110.187625

214. Draeger, A., Monastyrskaya, K. & Babiychuk, E.B. Plasma membrane repair and cellular damage control: the annexin survival kit. Biochem. Pharmacol. (2011), 81(6):703–12. doi:10.1016/j.bcp.2010.12.027

215. Monastyrskaya, K., Babiychuk, E.B., Hostettler, A., Rescher, U. & Draeger, A. Annexins as intracellular calcium sensors. Cell Calcium (2007), 41(3):207–19. doi:10.1016/j.ceca.2006.06.008

216. Monastyrskaya, K., Babiychuk, E.B. & Draeger, A. The annexins: spatial and temporal coordination of signaling events during cellular stress. Cell. Mol. Life Sci. (2009), 66(16):2623–42. doi:10.1007/s00018-009-0027-1

217. Shtam, T.A., Naryzhny, S.N., Landa, S.B., Burdakov, V.S., Artamonova, T.O. & Filatov, M. V. Purification and in vitro analysis of exosomes secreted by malignantly transformed human cells. Cell tissue biol. (2012), 6(4):317–325. doi:10.1134/S1990519X12040116

218. Sakwe, A.M., Koumangoye, R., Guillory, B. & Ochieng, J. Annexin A6 contributes to the invasiveness of breast carcinoma cells by influencing the organization and localization of functional focal adhesions. Exp. Cell Res. (2011), 317(6):823–37. doi:10.1016/j.yexcr.2010.12.008

219. Pol, A., Ortega, D. & Enrich, C. Identification of cytoskeleton-associated proteins in isolated rat liver endosomes. Biochem. J. (1997), 327741–6.

220. Gerke, V., Creutz, C.E. & Moss, S.E. Annexins: linking Ca2+ signalling to membrane dynamics. Nat. Rev. Mol. Cell Biol. (2005), 6(6):449–61. doi:10.1038/nrm1661

221. Edwards, H.C. & Crumpton, M.J. Ca(2+)-dependent phospholipid and arachidonic acid binding by the placental annexins VI and IV. Eur. J. Biochem. (1991), 198(1):121–9.

222. Podszywalow-Bartnicka, P., Strzelecka-Kiliszek, A., Bandorowicz-Pikula, J. & Pikula, S. Calcium- and proton-dependent relocation of annexin A6 in Jurkat T cells stimulated for interleukin-2 secretion. Acta Biochim. Pol. (2007), 54(2):261–71. doi:20071453 [pii]

223. Domon, M.M., Matar, G., Strzelecka-Kiliszek, A., Bandorowicz-Pikula, J., Pikula, S. & Besson, F. Interaction of annexin A6 with cholesterol rich membranes is pH-dependent and mediated by the sterol OH. J. Colloid Interface Sci. (2010), 346(2):436–41. doi:10.1016/j.jcis.2010.03.015

224. Jäckle, S., Beisiegel, U., Rinninger, F., Buck, F., Grigoleit, A., Block, A., Gröger, I., Greten, H. & Windler, E. Annexin VI, a marker protein of hepatocytic endosomes. J. Biol. Chem. (1994), 269(2):1026–32.

225. De Diego, I., Schwartz, F., Siegfried, H., Dauterstedt, P., Heeren, J., Beisiegel, U., Enrich, C. & Grewal, T. Cholesterol modulates the membrane binding and intracellular distribution of annexin 6. J. Biol. Chem. (2002), 277(35):32187–94. doi:10.1074/jbc.M205499200

226. Pons, M., Ihrke, G., Koch, S., Biermer, M., Pol, A., Grewal, T., Jäckle, S. & Enrich, C. Late endocytic compartments are major sites of annexin VI localization in NRK fibroblasts and polarized WIF-B hepatoma cells. Exp. Cell Res. (2000), 257(1):33–47. doi:10.1006/excr.2000.4861

227. Grewal, T., Heeren, J., Mewawala, D., Schnitgerhans, T., Wendt, D., Salomon, G., Enrich, C., Beisiegel, U. & Jäckle, S. Annexin VI stimulates endocytosis and is involved in the trafficking of low density lipoprotein to the prelysosomal compartment. J. Biol. Chem. (2000), 275(43):33806–13. doi:10.1074/jbc.M002662200

228. Desjardins, M., Celis, J.E., van Meer, G., Dieplinger, H., Jahraus, A., Griffiths, G. & Huber, L.A. Molecular characterization of phagosomes. J. Biol. Chem. (1994), 269(51):32194–200.

229. Skrahina, T., Piljić, A. & Schultz, C. Heterogeneity and timing of translocation and membrane-mediated assembly of different annexins. Exp. Cell Res. (2008), 314(5):1039–47. doi:10.1016/j.yexcr.2007.11.015

230. Blackwood, R.A. & Ernst, J.D. Characterization of Ca2(+)-dependent phospholipid binding, vesicle aggregation and membrane fusion by annexins. Biochem. J. (1990), 266(1):195–200.

231. Hoque, M., Rentero, C., Cairns, R., Tebar, F., Enrich, C. & Grewal, T. Annexins - scaffolds modulating PKC localization and signalling. Cell. Signal. (2014), 26(6):1213–1225. doi:10.1016/j.cellsig.2014.02.012

232. Golczak, M., Kirilenko, A., Bandorowicz-Pikula, J., Desbat, B. & Pikula, S. Structure of human Annexin A6 at the air-water interface and in a membrane-bound state. Biophys. J. (2004), 87(2):1215–1226. doi:10.1529/biophysj.103.038240

233. Sztolsztener, M.E., Strzelecka-Kiliszek, A., Pikula, S., Tylki-Szymanska, A. & Bandorowicz-Pikula, J. Cholesterol as a factor regulating intracellular localization of annexin A6 in Niemann-Pick type C human skin fibroblasts. Arch. Biochem. Biophys. (2010), 493(2):221–33. doi:10.1016/j.abb.2009.11.001

234. Sprenger, R.R., Speijer, D., Back, J.W., De Koster, C.G., Pannekoek, H. & Horrevoets, A.J.G. Comparative proteomics of human endothelial cell caveolae and rafts using two-dimensional gel electrophoresis and mass spectrometry. Electrophoresis (2004), 25(1):156–72. doi:10.1002/elps.200305675

235. Domon, M.M., Besson, F., Bandorowicz-Pikula, J. & Pikula, S. Annexin A6 is recruited into lipid rafts of Niemann-Pick type C disease fibroblasts in a Ca2+-dependent manner. Biochem. Biophys. Res. Commun. (2011), 405(2):192–6. doi:10.1016/j.bbrc.2010.12.138

236. Lin, H.C., Südhof, T.C. & Anderson, R.G. Annexin VI is required for budding of clathrin-coated pits. Cell (1992), 70(2):283–91.

237. Babiychuk, E.B. & Draeger, A. Annexins in cell membrane dynamics. Ca(2+)-regulated association of lipid microdomains. J. Cell Biol. (2000), 150(5):1113–24.

238. Orito, A., Kumanogoh, H., Yasaka, K., Sokawa, J., Hidaka, H., Sokawa, Y. & Maekawa, S. Calcium- dependent association of annexin VI, protein kinase C alpha, and alpha on the raft fraction derived from the synaptic plasma membrane of rat brain. J. Neurosci. Res. (2001), 64(3):235–41.

239. Watanabe, T., Inui, M., Chen, B.Y., Iga, M. & Sobue, K. Annexin VI-binding proteins in brain. Interaction of annexin VI with a membrane skeletal protein, calspectin (brain spectrin or fodrin). J. Biol. Chem. (1994), 269(26):17656–62.

240. Mishra, S., Chander, V., Banerjee, P., Oh, J.G., Lifirsu, E., Park, W.J., Kim, D.H. & Bandyopadhyay, A. Interaction of annexin A6 with alpha actinin in cardiomyocytes. BMC Cell Biol. (2011), 127. doi:10.1186/1471-2121-12-7

241. Kobayashi, R. & Tashima, Y. Purification, biological properties and partial sequence analysis of 67-kDa calcimedin and its 34-kDa fragment from chicken gizzard. Eur. J. Biochem. (1990), 188(2):447–53.

242. Babiychuk, E.B., Palstra, R.J., Schaller, J., Kämpfer, U. & Draeger, A. Annexin VI participates in the formation of a reversible, membrane-cytoskeleton complex in smooth muscle cells. J. Biol. Chem. (1999), 274(49):35191–5.

243. Hosoya, H., Kobayashi, R., Tsukita, S. & Matsumura, F. Ca(2+)-regulated actin and phospholipid binding protein (68 kD-protein) from bovine liver: identification as a homologue for annexin VI and intracellular localization. Cell Motil. Cytoskeleton (1992), 22(3):200–10. doi:10.1002/cm.970220307

244. Dillon, S.R., Mancini, M., Rosen, a & Schlissel, M.S. Annexin V binds to viable B cells and colocalizes with a marker of lipid rafts upon B cell receptor activation. J. Immunol. (2000), 164(3):1322–32.

245. Lambert, O., Cavusoglu, N., Gallay, J., Vincent, M., Rigaud, J.L., Henry, J.-P. & Ayala-Sanmartin, J. Novel organization and properties of annexin 2-membrane complexes. J. Biol. Chem. (2004), 279(12):10872– 82. doi:10.1074/jbc.M313657200

246. Ayala-Sanmartin, J., Henry, J.P. & Pradel, L.A. Cholesterol regulates membrane binding and aggregation by annexin 2 at submicromolar Ca(2+) concentration. Biochim. Biophys. Acta (2001), 1510(1-2):18– 28.

247. Ayala-Sanmartin, J. Cholesterol enhances phospholipid binding and aggregation of annexins by their core domain. Biochem. Biophys. Res. Commun. (2001), 283(1):72–9. doi:10.1006/bbrc.2001.4748

248. Cubells, L., Vilà De Muga, S., Tebar, F., Wood, P., Evans, R., Ingelmo-Torres, M., Calvo, M., Gaus, K., Pol, A., Grewal, T., Enrich, C. & Vila de Muga, S. Annexin A6-induced alterations in cholesterol transport and caveolin export from the Golgi complex. Traffic (2007), 8(11):1568–89. doi:10.1111/j.1600- 0854.2007.00640.x

249. Grewal, T., Evans, R., Rentero, C., Tebar, F., Cubells, L., de Diego, I., Kirchhoff, M.F., Hughes, W.E., Heeren, J., Rye, K.-A., Rinninger, F., Daly, R.J., Pol, A. & Enrich, C. Annexin A6 stimulates the membrane recruitment of p120GAP to modulate Ras and Raf-1 activity. Oncogene (2005), 24(38):5809–20. doi:10.1038/sj.onc.1208743

250. Bandorowicz-Pikula, J., Kirilenko, A., van Deursen, R., Golczak, M., Kühnel, M., Lancelin, J.-M., Pikula, S. & Buchet, R. A putative consensus sequence for the nucleotide-binding site of annexin A6. Biochemistry (2003), 42(30):9137–46. doi:10.1021/bi034359m

251. Storch, J. & Xu, Z. Niemann–Pick C2 (NPC2) and intracellular cholesterol trafficking. Biochim. Biophys. Acta - Mol. Cell Biol. Lipids (2009), 1791(7):671–678. doi:10.1016/j.bbalip.2009.02.001

252. Urano, Y., Watanabe, H., Murphy, S.R., Shibuya, Y., Geng, Y., Peden, A.A., Chang, C.C.Y. & Chang, T.Y. Transport of LDL-derived cholesterol from the NPC1 compartment to the ER involves the trans-Golgi network and the SNARE protein complex. Proc. Natl. Acad. Sci. U. S. A. (2008), 105(43):16513–8. doi:10.1073/pnas.0807450105

253. Kanerva, K., Uronen, R.-L., Blom, T., Li, S., Bittman, R., Lappalainen, P., Peränen, J., Raposo, G. & Ikonen, E. LDL cholesterol recycles to the plasma membrane via a Rab8a-Myosin5b-actin-dependent membrane transport route. Dev. Cell (2013), 27(3):249–62. doi:10.1016/j.devcel.2013.09.016

254. Kwon, H.J., Abi-Mosleh, L., Wang, M.L., Deisenhofer, J., Goldstein, J.L., Brown, M.S. & Infante, R.E. Structure of N-terminal domain of NPC1 reveals distinct subdomains for binding and transfer of cholesterol. Cell (2009), 137(7):1213–24. doi:10.1016/j.cell.2009.03.049

255. Vance, J.E. Transfer of cholesterol by the NPC team. Cell Metab. (2010), 12(2):105–6. doi:10.1016/j.cmet.2010.07.004

256. Xie, X., Brown, M.S., Shelton, J.M., Richardson, J. a, Goldstein, J.L. & Liang, G. Amino acid substitution in NPC1 that abolishes cholesterol binding reproduces phenotype of complete NPC1 deficiency in mice. Proc. Natl. Acad. Sci. U. S. A. (2011), 108(37):15330–5. doi:10.1073/pnas.1112751108

257. Reverter, M., Rentero, C., de Muga, S.V., Alvarez-Guaita, A., Mulay, V., Cairns, R., Wood, P., Monastyrskaya, K., Pol, A., Tebar, F., Blasi, J., Grewal, T. & Enrich, C. Cholesterol transport from late endosomes to the Golgi regulates t-SNARE trafficking, assembly, and function. Mol. Biol. Cell (2011), 22(21):4108–23. doi:10.1091/mbc.E11-04-0332

258. Babiychuk, V.S., Draeger, A. & Babiychuk, E.B. Smooth muscle actomyosin promotes Ca2+-dependent interactions between annexin VI and detergent-insoluble glycosphingolipid-enriched membrane domains. Acta Biochim. Pol. (2000), 47(3):579–89.

259. Calvo, M. & Enrich, C. Biochemical analysis of a caveolae-enriched plasma membrane fraction from rat liver. Electrophoresis (2000), 21(16):3386–95. doi:10.1002/1522-2683(20001001)21:16<3386::AID- ELPS3386>3.0.CO;2-L

260. Grewal, T. & Enrich, C. Annexins--modulators of EGF receptor signalling and trafficking. Cell. Signal. (2009), 21(6):847–58. doi:10.1016/j.cellsig.2009.01.031

261. De Laurentiis, A., Donovan, L. & Arcaro, A. Lipid rafts and caveolae in signaling by growth factor receptors. Open Biochem. J. (2007), 112–32. doi:10.2174/1874091X00701010012

262. Ringerike, T., Blystad, F.D., Levy, F.O., Madshus, I.H. & Stang, E. Cholesterol is important in control of EGF receptor kinase activity but EGF receptors are not concentrated in caveolae. J. Cell Sci. (2002), 115(Pt 6):1331–40.

263. Jang, I., Ho, J., Dae, B., Sik, S., Hwan, M., Suh, P. & Ho, S. Localization of phospholipase Cgamma1 signaling in caveolae: importance in EGF-induced phosphoinositide hydrolysis but not in tyrosine phosphorylation. FEBS Lett. (2001), 491(1-2):4–8. doi:10.1016/S0014-5793(01)02165-2

264. Vilá de Muga, S., Timpson, P., Cubells, L., Evans, R., Hayes, T.E., Rentero, C., Hegemann, A., Reverter, M., Leschner, J., Pol, A., Tebar, F., Daly, R.J., Enrich, C. & Grewal, T. Annexin A6 inhibits Ras signalling in breast cancer cells. Oncogene (2009), 28(3):363–77. doi:10.1038/onc.2008.386

265. Koese, M., Rentero, C., Kota, B.P., Hoque, M., Cairns, R., Wood, P., Vilà de Muga, S., Reverter, M., Alvarez- Guaita, A., Monastyrskaya, K., Hughes, W.E., Swarbrick, A., Tebar, F., Daly, R.J., Enrich, C. & Grewal, T. Annexin A6 is a scaffold for PKCα to promote EGFR inactivation. Oncogene (2013), 32(23):2858–72. doi:10.1038/onc.2012.303

266. Wang, X., Yan, Q., Sun, P., Activation, K.C.-A., Cd, R. & Paller, A.S. Suppression of Epidermal Growth Factor Receptor Signaling by Protein Kinase C- α Activation Requires CD82 , Caveolin-1 , and Ganglioside. Cancer Res. (2007), 679986–9995. doi:10.1158/0008-5472.CAN-07-1300

267. Bastiani, M. & Parton, R.G. Caveolae at a glance. J. Cell Sci. (2010), 123(Pt 22):3831–6. doi:10.1242/jcs.070102

268. Schnitzer, J.E., Liu, J. & Oh, P. Endothelial caveolae have the molecular transport machinery for vesicle budding, docking, and fusion including VAMP, NSF, SNAP, annexins, and GTPases. J. Biol. Chem. (1995), 270(24):14399–404. doi:10.1074/jbc.270.24.14399

269. Grimmer, S., Ying, M., Wälchli, S., van Deurs, B. & Sandvig, K. Golgi vesiculation induced by cholesterol occurs by a dynamin- and cPLA2-dependent mechanism. Traffic (2005), 6(2):144–56. doi:10.1111/j.1600-0854.2005.00258.x

270. Cubells, L., Vilà de Muga, S., Tebar, F., Bonventre, J. V, Balsinde, J., Pol, A., Grewal, T., Enrich, C. & de Muga, S.V. Annexin A6-induced inhibition of cytoplasmic phospholipase A2 is linked to caveolin-1 export from the Golgi. J. Biol. Chem. (2008), 283(15):10174–83. doi:10.1074/jbc.M706618200

271. Chen, Z. & Rand, R.P. The influence of cholesterol on phospholipid membrane curvature and bending elasticity. Biophys. J. (1997), 73(1):267–76. doi:10.1016/S0006-3495(97)78067-6

272. Van Duyl, B.Y., Meeldijk, H., Verkleij, A.J., Rijkers, D.T.S., Chupin, V., de Kruijff, B. & Killian, J.A. A synergistic effect between cholesterol and tryptophan-flanked transmembrane helices modulates membrane curvature. Biochemistry (2005), 44(11):4526–32. doi:10.1021/bi047937n

273. Fielding, C.J. & Fielding, P.E. Cholesterol and caveolae: structural and functional relationships. Biochim. Biophys. Acta (2000), 1529(1-3):210–22.

274. Breen, M.R., Camps, M., Carvalho-Simoes, F., Zorzano, A. & Pilch, P.F. Cholesterol depletion in adipocytes causes caveolae collapse concomitant with proteosomal degradation of cavin-2 in a switch- like fashion. PLoS One (2012), 7(4):e34516. doi:10.1371/journal.pone.0034516

275. Cornely, R., Rentero, C., Enrich, C., Grewal, T. & Gaus, K. Annexin A6 is an organizer of membrane microdomains to regulate receptor localization and signalling. IUBMB Life (2011), 63(11):1009–17. doi:10.1002/iub.540

276. Buzhynskyy, N., Golczak, M., Lai-Kee-Him, J., Lambert, O., Tessier, B., Gounou, C., Bérat, R., Simon, A., Granier, T., Chevalier, J.-M., Mazères, S., Bandorowicz-Pikula, J., Pikula, S. & Brisson, A.R. Annexin-A6 presents two modes of association with phospholipid membranes. A combined QCM-D, AFM and cryo- TEM study. J. Struct. Biol. (2009), 168(1):107–16. doi:10.1016/j.jsb.2009.03.007

277. Alvarez-Guaita, A., Vilà de Muga, S., Owen, D., Williamson, D., Magenau, A., García-Melero, A., Reverter, M., Hoque, M., Cairns, R., Cornely, R., Tebar, F., Grewal, T., Gaus, K., Ayala-Sanmartín, J., Enrich, C. & Rentero, C. In vivo evidence for Annexin A6-dependent plasma membrane remodelling. Br. J. Pharmacol. (2014),

278. Ortega, D., Pol, A., Biermer, M., Jäckle, S. & Enrich, C. Annexin VI defines an apical endocytic compartment in rat liver hepatocytes. J. Cell Sci. (1998), 111 ( Pt 2261–9.

279. Smythe, E., Smith, P.D., Jacob, S.M., Theobald, J. & Moss, S.E. Endocytosis occurs independently of annexin VI in human A431 cells. J. Cell Biol. (1994), 124(3):301–6.

280. Hawkins, T.E., Roes, J., Rees, D., Monkhouse, J. & Moss, S.E. Immunological development and cardiovascular function are normal in annexin VI null mutant mice. Mol. Cell. Biol. (1999), 19(12):8028–32.

281. Creutz, C.E. & Snyder, S.L. Interactions of annexins with the mu subunits of the clathrin assembly proteins. Biochemistry (2005), 44(42):13795–806. doi:10.1021/bi051160w

282. Turpin, E., Russo-Marie, F., Dubois, T., de Paillerets, C., Alfsen, a & Bomsel, M. In adrenocortical tissue, annexins II and VI are attached to clathrin coated vesicles in a calcium-independent manner. Biochim. Biophys. Acta (1998), 1402(2):115–30.

283. Howes, M.T., Mayor, S. & Parton, R.G. Molecules, mechanisms, and cellular roles of clathrin- independent endocytosis. Curr. Opin. Cell Biol. (2010), 22(4):519–27. doi:10.1016/j.ceb.2010.04.001

284. Maldonado-Báez, L., Williamson, C. & Donaldson, J.G. Clathrin-independent endocytosis: a cargo- centric view. Exp. Cell Res. (2013), 319(18):2759–69. doi:10.1016/j.yexcr.2013.08.008

285. Doherty, G.J. & McMahon, H.T. Mechanisms of endocytosis. Annu. Rev. Biochem. (2009), 78857–902. doi:10.1146/annurev.biochem.78.081307.110540

286. Gould, G.W. & Lippincott-Schwartz, J. New roles for endosomes: from vesicular carriers to multi- purpose platforms. Nat. Rev. Mol. Cell Biol. (2009), 10(4):287–92. doi:10.1038/nrm2652

287. Kamal, A., Ying, Y. & Anderson, R.G. Annexin VI-mediated loss of spectrin during coated pit budding is coupled to delivery of LDL to lysosomes. J. Cell Biol. (1998), 142(4):937–47.

288. Pons, M., Grewal, T., Rius, E., Schnitgerhans, T., Jäckle, S. & Enrich, C. Evidence for the Involvement of annexin 6 in the trafficking between the endocytic compartment and lysosomes. Exp. Cell Res. (2001), 269(1):13–22. doi:10.1006/excr.2001.5268

289. Lavialle, F., Durand, G. & Ollivier-Bousquet, M. Localisation of Annexin V and Annexin VI in Lactating Mammary Epithelial Cells. Intercell. Signal. Mammary Gland (1995), 275–276. doi:10.1007/978-1- 4615-1973-7_53

290. Zaks, W.J. & Creutz, C.E. Annexin-chromaffin granule membrane interactions: a comparative study of synexin, p32 and p67. Biochim. Biophys. Acta (1990), 1029(1):149–60.

291. Donnelly, S.R. & Moss, S.E. Annexins in the secretory pathway. Cell. Mol. Life Sci. (1997), 53(6):533–8.

292. Schmidt, H., Gelhaus, C., Nebendahl, M., Lettau, M., Lucius, R., Leippe, M., Kabelitz, D. & Janssen, O. Effector granules in human T lymphocytes: the luminal proteome of secretory lysosomes from human T cells. Cell Commun. Signal. (2011), 9(1):4. doi:10.1186/1478-811X-9-4

293. Ma, H., Kien, F., Manière, M., Zhang, Y., Lagarde, N., Tse, K.S., Poon, L.L.M. & Nal, B. Human annexin A6 interacts with influenza a virus protein M2 and negatively modulates infection. J. Virol. (2012), 86(3):1789–801. doi:10.1128/JVI.06003-11

294. Musiol, A., Gran, S., Ehrhardt, C., Ludwig, S., Grewal, T., Gerke, V. & Rescher, U. Annexin A6-balanced late endosomal cholesterol controls influenza A replication and propagation. MBio (2013), 4(6):e00608–13. doi:10.1128/mBio.00608-13

295. Mussunoor, S. & Murray, G.I. The role of annexins in tumour development and progression. J. Pathol. (2008), 216(2):131–40. doi:10.1002/path.2400

296. Theobald, J., Hanby, A., Patel, K. & Moss, S.E. Annexin VI has tumour-suppressor activity in human A431 squamous epithelial carcinoma cells. Br. J. Cancer (1995), 71(4):786–8.

297. Theobald, J., Smith, P.D., Jacob, S.M. & Moss, S.E. Expression of annexin VI in A431 carcinoma cells suppresses proliferation: a possible role for annexin VI in cell growth regulation. Biochim. Biophys. Acta (1994), 1223(3):383–90.

298. Wang, X., Zhang, S., Zhang, J., Lam, E., Liu, X., Sun, J., Feng, L., Lu, H., Yu, J. & Jin, H. Annexin A6 is down- regulated through promoter methylation in gastric cancer. Am. J. Transl. Res. (2013), 5(5):555–62.

299. Francia, G., Mitchell, S.D., Moss, S.E., Hanby, A.M., Marshall, J.F. & Hart, I.R. Identification by differential display of annexin-VI, a gene differentially expressed during melanoma progression. Cancer Res. (1996), 56(17):3855–8.

300. Koumangoye, R.B., Nangami, G.N., Thompson, P.D., Agboto, V.K., Ochieng, J. & Sakwe, A.M. Reduced annexin A6 expression promotes the degradation of activated epidermal growth factor receptor and sensitizes invasive breast cancer cells to EGFR-targeted tyrosine kinase inhibitors. Mol. Cancer (2013), 12(1):167. doi:10.1186/1476-4598-12-167

301. Rentero, C., Evans, R., Wood, P., Tebar, F., Vilà De Muga, S., Cubells, L., de Diego, I., Hayes, T.E., Hughes, W.E., Pol, A., Rye, K.-A., Enrich, C. & Grewal, T. Inhibition of H-Ras and MAPK is compensated by PKC- dependent pathways in annexin A6 expressing cells. Cell. Signal. (2006), 18(7):1006–16. doi:10.1016/j.cellsig.2005.08.008

302. Pons, M., Tebar, F., Kirchhoff, M., Peiró, S., de Diego, I., Grewal, T. & Enrich, C. Activation of Raf-1 is defective in annexin 6 overexpressing Chinese hamster ovary cells. FEBS Lett. (2001), 501(1):69–73.

303. Grewal, T. & Enrich, C. Molecular mechanisms involved in Ras inactivation: the annexin A6-p120GAP complex. Bioessays (2006), 28(12):1211–20. doi:10.1002/bies.20503

304. Schmitz-Peiffer, C., Browne, C.L., Walker, J.H. & Biden, T.J. Activated protein kinase C alpha associates with annexin VI from skeletal muscle. Biochem. J. (1998), 330 ( Pt 2675–81.

305. Martin, M., Leffler, J. & Blom, A.M. and A5 serve as new ligands for C1q on apoptotic cells. J. Biol. Chem. (2012), 287(40):33733–44. doi:10.1074/jbc.M112.341339

306. Ao, W., Zheng, H., Chen, X.-W., Shen, Y. & Yang, C.-D. Anti-annexin II antibody is associated with thrombosis and/or pregnancy morbidity in antiphospholipid syndrome and systemic lupus erythematosus with thrombosis. Rheumatol. Int. (2011), 31(7):865–9. doi:10.1007/s00296-010- 1379-4

307. Hrycek, A. & Cieślik, P. and anti-annexin antibodies in patients with systemic lupus erythematosus. Rheumatol. Int. (2012), 32(5):1335–42. doi:10.1007/s00296-011-1793-2

308. Trouw, L.A. & Daha, M.R. Role of anti-C1q autoantibodies in the pathogenesis of lupus nephritis. Expert Opin. Biol. Ther. (2005), 5(2):243–51. doi:10.1517/14712598.5.2.243

309. Rai, T., Mosoian, A. & Resh, M.D. Annexin 2 is not required for human immunodeficiency virus type 1 particle production but plays a cell type-dependent role in regulating infectivity. J. Virol. (2010), 84(19):9783–92. doi:10.1128/JVI.01584-09

310. Harrist, A. V, Ryzhova, E. V, Harvey, T. & González-Scarano, F. Anx2 interacts with HIV-1 Gag at phosphatidylinositol (4,5) bisphosphate-containing lipid rafts and increases viral production in 293T cells. PLoS One (2009), 4(3):e5020. doi:10.1371/journal.pone.0005020

311. Weyd, H., Abeler-Dörner, L., Linke, B., Mahr, A., Jahndel, V., Pfrang, S., Schnölzer, M., Falk, C.S. & Krammer, P.H. on the surface of early apoptotic cells suppresses CD8+ T cell immunity. PLoS One (2013), 8(4):e62449. doi:10.1371/journal.pone.0062449

312. Dalli, J., Jones, C.P., Cavalcanti, D.M., Farsky, S.H., Perretti, M. & Rankin, S.M. Annexin A1 regulates neutrophil clearance by macrophages in the mouse bone marrow. FASEB J. (2012), 26(1):387–96. doi:10.1096/fj.11-182089

313. Perretti, M., Wheller, S.K., Flower, R.J., Wahid, S. & Pitzalis, C. Modulation of cellular annexin I in human leukocytes infiltrating DTH skin reactions. J. Leukoc. Biol. (1999), 65(5):583–9.

314. Yang, Y.H., Song, W., Deane, J.A., Kao, W., Ooi, J.D., Ngo, D., Kitching, A.R., Morand, E.F. & Hickey, M.J. Deficiency of annexin A1 in CD4+ T cells exacerbates T cell-dependent inflammation. J. Immunol. (2013), 190(3):997–1007. doi:10.4049/jimmunol.1202236

315. D’Acquisto, F., Paschalidis, N., Sampaio, A.L.F., Merghani, A., Flower, R.J. & Perretti, M. Impaired T cell activation and increased Th2 lineage commitment in Annexin-1-deficient T cells. Eur. J. Immunol. (2007), 37(11):3131–42. doi:10.1002/eji.200636792

316. Mouse Genome Informatics and the Wellcome Trust Sanger Institute Mouse Genetics Project (MGP) Obtaining and Loading Phenotype Annotations from the Wellcome Trust Sanger Institute (WTSI) Mouse Resources Portal (AnxA6 knock-out). Database Release (2011),

317. Bhinder, G., Sham, H.P., Chan, J.M., Morampudi, V., Jacobson, K. & Vallance, B.A. The Citrobacter rodentium mouse model: studying pathogen and host contributions to infectious colitis. J. Vis. Exp. (2013), (72):e50222. doi:10.3791/50222

318. Williamson, D.J., Owen, D.M., Rossy, J., Magenau, A., Gooding, J.J., Gaus, K. & Wehrmann, M. Pre-existing clusters of the adaptor Lat do not participate in early T cell signaling events. Nat. Immunol. (2011), 12(7):655–62. doi:10.1038/ni.2049

319. Perry, G.L.W. SpPack: spatial point pattern analysis in Excel using Visual Basic for Applications (VBA). Environ. Model. Softw. (2004), 19559–569.

320. Spandidos, A., Wang, X., Wang, H. & Seed, B. PrimerBank: a resource of human and mouse PCR primer pairs for gene expression detection and quantification. Nucleic Acids Res. (2010), 38(Database issue):D792–9. doi:10.1093/nar/gkp1005

321. Palmgrén, J.J., Töyräs, A., Mauriala, T., Mönkkönen, J. & Auriola, S. Quantitative determination of cholesterol, sitosterol, and sitostanol in cultured Caco-2 cells by liquid chromatography-atmospheric pressure chemical ionization mass spectrometry. J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. (2005), 821(2):144–52. doi:10.1016/j.jchromb.2005.04.029

322. Doubell, A.F., Lazure, C., Charbonneau, C. & Thibault, G. Identification and immunolocalisation of annexins V and VI, the major cardiac annexins, in rat heart. Cardiovasc. Res. (1993), 27(7):1359–67.

323. Gunteski-Hamblin, A.M., Song, G., Walsh, R.A., Frenzke, M., Boivin, G.P., Dorn, G.W., Kaetzel, M.A., Horseman, N.D. & Dedman, J.R. Annexin VI overexpression targeted to heart alters cardiomyocyte function in transgenic mice. Am. J. Physiol. (1996), 270(3 Pt 2):H1091–100.

324. Song, G., Harding, S.E., Duchen, M.R., Tunwell, R., O’Gara, P., Hawkins, T.E. & Moss, S.E. Altered mechanical properties and intracellular calcium signaling in cardiomyocytes from annexin 6 null- mutant mice. FASEB J. (2002), 16(6):622–4. doi:10.1096/fj.02-0216fje

325. Luo, Y. & Dorf, M.E. Delayed-type hypersensitivity. Curr. Protoc. Immunol. (2001), Chapter 4(1993):Unit 4.5. doi:10.1002/0471142735.im0405s55

326. Zehn, D., King, C., Bevan, M.J. & Palmer, E. TCR signaling requirements for activating T cells and for generating memory. Cell. Mol. Life Sci. (2012), 69(10):1565–75. doi:10.1007/s00018-012-0965-x

327. Kaplan, D.H., Igyártó, B.Z. & Gaspari, A.A. Early immune events in the induction of allergic contact dermatitis. Nat. Rev. Immunol. (2012), 12(2):114–24. doi:10.1038/nri3150

328. Tedla, N., Wang, H.W., McNeil, H.P., Di Girolamo, N., Hampartzoumian, T., Wakefield, D. & Lloyd, A. Regulation of T lymphocyte trafficking into lymph nodes during an immune response by the chemokines macrophage inflammatory protein (MIP)-1 alpha and MIP-1 beta. J. Immunol. (1998), 161(10):5663–72.

329. Galandrini, R., Galluzzo, E., Albi, N., Grossi, C.E. & Velardi, A. Hyaluronate is costimulatory for human T cell effector functions and binds to CD44 on activated T cells. J. Immunol. (1994), 153(1):21–31.

330. Wedepohl, S., Beceren-Braun, F., Riese, S., Buscher, K., Enders, S., Bernhard, G., Kilian, K., Blanchard, V., Dernedde, J. & Tauber, R. L-selectin--a dynamic regulator of leukocyte migration. Eur. J. Cell Biol. (2012), 91(4):257–64. doi:10.1016/j.ejcb.2011.02.007

331. Mackay, C., Marston, W. & Dudler, L. Naive and memory T cells show distinct pathways of lymphocyte recirculation. J. Exp. … (1990), 171(March):801–817.

332. Sprent, J., Zhang, X., Sun, S. & Tough, D. T-cell turnover in vivo and the role of cytokines. Immunol. Lett. (1999), 65(1-2):21–5.

333. Kurki, P., Vanderlaan, M., Dolbeare, F., Gray, J. & Tan, E.M. Expression of proliferating cell nuclear antigen (PCNA)/cyclin during the cell cycle. Exp. Cell Res. (1986), 166(1):209–19.

334. Bravo, R., Frank, R., Blundell, P.A. & Macdonald-Bravo, H. Cyclin/PCNA is the auxiliary protein of DNA polymerase-delta. Nature 326(6112):515–7. doi:10.1038/326515a0

335. Boyman, O., Létourneau, S., Krieg, C. & Sprent, J. Homeostatic proliferation and survival of naïve and memory T cells. Eur. J. Immunol. (2009), 39(8):2088–94. doi:10.1002/eji.200939444

336. Gomez, T.S. & Billadeau, D.D. T cell activation and the cytoskeleton: you can’t have one without the other. Adv. Immunol. (2008), 97(08):1–64. doi:10.1016/S0065-2776(08)00001-1

337. Billadeau, D.D., Nolz, J.C. & Gomez, T.S. Regulation of T-cell activation by the cytoskeleton. Nat. Rev. Immunol. (2007), 7(2):131–43. doi:10.1038/nri2021

338. Koronakis, V., Hume, P.J., Humphreys, D., Liu, T., Hørning, O., Jensen, O.N. & McGhie, E.J. WAVE regulatory complex activation by cooperating GTPases Arf and Rac1. Proc. Natl. Acad. Sci. U. S. A. (2011), 108(35):14449–54. doi:10.1073/pnas.1107666108

339. Gagnon, E., Schubert, D.A., Gordo, S., Chu, H.H. & Wucherpfennig, K.W. Local changes in lipid environment of TCR microclusters regulate membrane binding by the CD3ε cytoplasmic domain. J. Exp. Med. (2012), 209(13):2423–39. doi:10.1084/jem.20120790

340. Liao, W., Lin, J.-X. & Leonard, W.J. Interleukin-2 at the crossroads of effector responses, tolerance, and immunotherapy. Immunity (2013), 38(1):13–25. doi:10.1016/j.immuni.2013.01.004

341. Leonard, W.J., Krönke, M., Peffer, N.J., Depper, J.M. & Greene, W.C. Interleukin 2 receptor gene expression in normal human T lymphocytes. Proc. Natl. Acad. Sci. U. S. A. (1985), 82(18):6281–5.

342. Takeshita, T., Asao, H., Ohtani, K., Ishii, N., Kumaki, S., Tanaka, N., Munakata, H., Nakamura, M. & Sugamura, K. Cloning of the gamma chain of the human IL-2 receptor. Science (1992), 257(5068):379– 82.

343. Nosaka, T., Kawashima, T., Misawa, K., Ikuta, K., Mui, A.L. & Kitamura, T. STAT5 as a molecular regulator of proliferation, differentiation and apoptosis in hematopoietic cells. EMBO J. (1999), 18(17):4754–65. doi:10.1093/emboj/18.17.4754

344. Hatakeyama, M., Kono, T., Kobayashi, N., Kawahara, A., Levin, S.D., Perlmutter, R.M. & Taniguchi, T. Interaction of the IL-2 receptor with the src-family kinase p56lck: identification of novel intermolecular association. Science (1991), 252(5012):1523–8.

345. Benczik, M. & Gaffen, S.L. The interleukin (IL)-2 family cytokines: survival and proliferation signaling pathways in T lymphocytes. Immunol. Invest. (2004), 33(2):109–42. doi:10.1081/IMM-120030732

346. Schneider, U., Schwenk, H.U. & Bornkamm, G. Characterization of EBV-genome negative “null” and “T” cell lines derived from children with acute lymphoblastic leukemia and leukemic transformed non- Hodgkin lymphoma. Int. J. Cancer (1977), 19(5):621–6.

347. The Jackson Laboratory & JAX Mice Database Physiological Data Summary – C57BL/6J (000664). (2007),

348. Kruisbeek, A.M., Shevach, E. & Thornton, A.M. Proliferative assays for T cell function. Curr. Protoc. Immunol. (2004), Chapter 3Unit 3.12. doi:10.1002/0471142735.im0312s60

349. Schwab, R., Crow, M.K., Russo, C. & Weksler, M.E. Requirements for T cell activation by OKT3 monoclonal antibody: role of modulation of T3 molecules and interleukin 1. J. Immunol. (1985), 135(3):1714–8.

350. Dancker, P., Löw, I., Hasselbach, W. & Wieland, T. Interaction of actin with phalloidin: Biochim. Biophys. Acta - Protein Struct. (1975), 400(2):407–414. doi:10.1016/0005-2795(75)90196-8

351. Wulf, E., Deboben, A., Bautz, F.A., Faulstich, H. & Wieland, T. Fluorescent phallotoxin, a tool for the visualization of cellular actin. Proc. Natl. Acad. Sci. U. S. A. (1979), 76(9):4498–502.

352. Kaizuka, Y., Douglass, A.D., Varma, R., Dustin, M.L. & Vale, R.D. Mechanisms for segregating T cell receptor and adhesion molecules during immunological synapse formation in Jurkat T cells. Proc. Natl. Acad. Sci. U. S. A. (2007), 104(51):20296–301. doi:10.1073/pnas.0710258105

353. Campi, G., Varma, R. & Dustin, M.L. Actin and agonist MHC-peptide complex-dependent T cell receptor microclusters as scaffolds for signaling. J. Exp. Med. (2005), 202(8):1031–6. doi:10.1084/jem.20051182

354. Yokosuka, T., Sakata-Sogawa, K., Kobayashi, W., Hiroshima, M., Hashimoto-Tane, A., Tokunaga, M., Dustin, M.L. & Saito, T. Newly generated T cell receptor microclusters initiate and sustain T cell activation by recruitment of Zap70 and SLP-76. Nat. Immunol. (2005), 6(12):1253–62. doi:10.1038/ni1272

355. Rust, M.J., Bates, M. & Zhuang, X. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat. Methods (2006), 3(10):793–5. doi:10.1038/nmeth929

356. Heilemann, M., van de Linde, S., Schüttpelz, M., Kasper, R., Seefeldt, B., Mukherjee, A., Tinnefeld, P. & Sauer, M. Subdiffraction-resolution fluorescence imaging with conventional fluorescent probes. Angew. Chem. Int. Ed. Engl. (2008), 47(33):6172–6. doi:10.1002/anie.200802376

357. Thompson, R.E., Larson, D.R. & Information, W.W.W. author contact Precise Nanometer Localization Analysis for Individual Fluorescent Probes. Biophys. J. (2002), 82(5):2775–2783. doi:10.1016/S0006- 3495(02)75618-X

358. Ziegler, E., Oberbarnscheidt, M., Bulfone-Paus, S., Förster, R., Kunzendorf, U. & Krautwald, S. CCR7 signaling inhibits T cell proliferation. J. Immunol. (2007), 179(10):6485–93.

359. Moreau, H.D. & Bousso, P. Visualizing how T cells collect activation signals in vivo. Curr. Opin. Immunol. (2014), 2656–62. doi:10.1016/j.coi.2013.10.013

360. Bromley, S.K., Mempel, T.R. & Luster, A.D. Orchestrating the orchestrators: chemokines in control of T cell traffic. Nat. Immunol. (2008), 9(9):970–80. doi:10.1038/ni.f.213

361. Krummel, M.F. & Macara, I. Maintenance and modulation of T cell polarity. Nat. Immunol. (2006), 7(11):1143–9. doi:10.1038/ni1404

362. Sojka, D.K., Bruniquel, D., Schwartz, R.H. & Singh, N.J. IL-2 secretion by CD4+ T cells in vivo is rapid, transient, and influenced by TCR-specific competition. J. Immunol. (2004), 172(10):6136–43.

363. Choudhuri, K. & Dustin, M.L. Signaling microdomains in T cells. FEBS Lett. (2010), (October): doi:10.1016/j.febslet.2010.10.015

364. Radonic, A. Guideline to reference gene selection for quantitative real-time PCR. Biochem. Biophys. Res. Commun. (2004), 313(4):856–862. doi:10.1016/j.bbrc.2003.11.177

365. Rozdzial, M. & Kubo, R. Developmental regulation of the TCR zeta-chain. Differential expression and tyrosine phosphorylation of the TCR zeta-chain in resting immature and mature T. J. … (1994), 153(4):1563–1580.

366. Huang, J., Brameshuber, M., Zeng, X., Xie, J., Li, Q., Chien, Y., Valitutti, S. & Davis, M.M. A single peptide- major histocompatibility complex ligand triggers digital cytokine secretion in CD4(+) T cells. Immunity (2013), 39(5):846–57. doi:10.1016/j.immuni.2013.08.036

367. Cotari, J.W., Voisinne, G., Dar, O.E., Karabacak, V. & Altan-Bonnet, G. Cell-to-cell variability analysis dissects the plasticity of signaling of common γ chain cytokines in T cells. Sci. Signal. (2013), 6(266):ra17. doi:10.1126/scisignal.2003240

368. Krutzik, P.O. & Nolan, G.P. Intracellular phospho-protein staining techniques for flow cytometry: monitoring single cell signaling events. Cytometry. A (2003), 55(2):61–70. doi:10.1002/cyto.a.10072

369. Tanaka, N., Asao, H., Ohbo, K., Ishii, N., Takeshita, T., Nakamura, M., Sasaki, H. & Sugamura, K. Physical association of JAK1 and JAK2 tyrosine kinases with the interleukin 2 receptor beta and gamma chains. Proc. Natl. Acad. Sci. U. S. A. (1994), 91(15):7271–5.

370. Leonard, W.J. & O’Shea, J.J. Jaks and STATs: biological implications. Annu. Rev. Immunol. (1998), 16293–322. doi:10.1146/annurev.immunol.16.1.293

371. Grassart, A., Dujeancourt, A., Lazarow, P.B., Dautry-Varsat, A. & Sauvonnet, N. Clathrin-independent endocytosis used by the IL-2 receptor is regulated by Rac1, Pak1 and Pak2. EMBO Rep. (2008), 9(4):356–62. doi:10.1038/embor.2008.28

372. Basquin, C., Malardé, V., Mellor, P., Anderson, D.H., Meas-Yedid, V., Olivo-Marin, J.-C., Dautry-Varsat, A. & Sauvonnet, N. The signalling factor PI3K is a specific regulator of the clathrin-independent dynamin- dependent endocytosis of IL-2 receptors. J. Cell Sci. (2013), 126(Pt 5):1099–108. doi:10.1242/jcs.110932

373. Sauvonnet, N., Dujeancourt, A. & Dautry-Varsat, A. Cortactin and dynamin are required for the clathrin-independent endocytosis of gammac cytokine receptor. J. Cell Biol. (2005), 168(1):155–63. doi:10.1083/jcb.200406174

374. Basquin, C. & Sauvonnet, N. Phosphoinositide 3-kinase at the crossroad between endocytosis and signaling of cytokine receptors. Commun. Integr. Biol. (2013), 6(4):e24243. doi:10.4161/cib.24243

375. Smith, K.A. The structure of IL2 bound to the three chains of the IL2 receptor and how signaling occurs. Med. Immunol. (2006), 53. doi:10.1186/1476-9433-5-3

376. Subtil, A., Delepierre, M. & Dautry-Varsat, A. An alpha-helical signal in the cytosolic domain of the interleukin 2 receptor beta chain mediates sorting towards degradation after endocytosis. J. Cell Biol. (1997), 136(3):583–95.

377. Malek, T.R., Furse, R.K., Fleming, M.L., Fadell, A.J. & He, Y.W. Biochemical identity and characterization of the mouse interleukin-2 receptor beta and gamma c subunits. J. Interferon Cytokine Res. (1995), 15(5):447–54.

378. Le Bras, S., Massaad, M., Koduru, S., Kumar, L., Oyoshi, M.K., Hartwig, J. & Geha, R.S. WIP is critical for T cell responsiveness to IL-2. Proc. Natl. Acad. Sci. U. S. A. (2009), 106(18):7519–24. doi:10.1073/pnas.0806410106

379. Sechi, A.S. & Wehland, J. Interplay between TCR signalling and actin cytoskeleton dynamics. Trends Immunol. (2004), 25(5):257–65. doi:10.1016/j.it.2004.03.003

380. Monastyrskaya, K., Babiychuk, E.B., Hostettler, A., Wood, P., Grewal, T. & Draeger, A. Plasma membrane-associated annexin A6 reduces Ca2+ entry by stabilizing the cortical actin cytoskeleton. J. Biol. Chem. (2009), 284(25):17227–42. doi:10.1074/jbc.M109.004457

381. Rescher, U., Ruhe, D., Ludwig, C., Zobiack, N. & Gerke, V. Annexin 2 is a phosphatidylinositol (4,5)- bisphosphate binding protein recruited to actin assembly sites at cellular membranes. J. Cell Sci. (2004), 117(Pt 16):3473–80. doi:10.1242/jcs.01208

382. Draeger, A., Wray, S. & Babiychuk, E.B. Domain architecture of the smooth-muscle plasma membrane: regulation by annexins. Biochem. J. (2005), 387(Pt 2):309–14. doi:10.1042/BJ20041363

383. Gerke, V. & Moss, S.E. Annexins: from structure to function. Physiol. Rev. (2002), 82(2):331–71. doi:10.1152/physrev.00030.2001

384. Huse, M., Lillemeier, B.F., Kuhns, M.S., Chen, D.S. & Davis, M.M. T cells use two directionally distinct pathways for cytokine secretion. Nat. Immunol. (2006), 7(3):247–55. doi:10.1038/ni1304

385. Castro, I., Yu, A., Dee, M.J. & Malek, T.R. The basis of distinctive IL-2- and IL-15-dependent signaling: weak CD122-dependent signaling favors CD8+ T central-memory cell survival but not T effector- memory cell development. J. Immunol. (2011), 187(10):5170–82. doi:10.4049/jimmunol.1003961

386. Beyer, T., Busse, M., Hristov, K., Gurbiel, S., Smida, M., Haus, U.-U., Ballerstein, K., Pfeuffer, F., Weismantel, R., Schraven, B. & Lindquist, J. a Integrating signals from the T-cell receptor and the interleukin-2 receptor. PLoS Comput. Biol. (2011), 7(8):e1002121. doi:10.1371/journal.pcbi.1002121

387. Lu, Y., Morley, P. & Durkin, J.P. Signalling events mediating the activation of protein kinase C by interleukin-2 in cytotoxic T cells. Cell. Signal. (1999), 11(4):275–85.

388. Campbell, K.A., Minashima, T., Zhang, Y., Hadley, S., Lee, Y.J., Giovinazzo, J., Quirno, M. & Kirsch, T. Annexin 6 interacts with p65 and stimulates NF-κB activity and catabolic events in articular chondrocytes. Arthritis Rheum. (2013), doi:10.1002/art.38182

389. Ejsing, C.S. Molecular characterization of the lipidome by mass spectrometry. (2007), (October 2006):1–105.

390. Fahy, E., Cotter, D., Sud, M. & Subramaniam, S. Lipid classification, structures and tools. Biochim. Biophys. Acta (2011), 1811(11):637–47. doi:10.1016/j.bbalip.2011.06.009

391. Csoboz, B., Balogh, G.E., Kusz, E., Gombos, I., Peter, M., Crul, T., Gungor, B., Haracska, L., Bogdanovics, G., Torok, Z., Horvath, I. & Vigh, L. Membrane fluidity matters: hyperthermia from the aspects of lipids and membranes. Int. J. Hyperthermia (2013), 29(5):491–9. doi:10.3109/02656736.2013.808765

392. Shigapova, N., Török, Z., Balogh, G., Goloubinoff, P., Vígh, L. & Horváth, I. Membrane fluidization triggers membrane remodeling which affects the thermotolerance in Escherichia coli. Biochem. Biophys. Res. Commun. (2005), 328(4):1216–23. doi:10.1016/j.bbrc.2005.01.081

393. Ikonen, E. & Hölttä-Vuori, M. Cellular pathology of Niemann-Pick type C disease. Semin. Cell Dev. Biol. (2004), 15(4):445–54. doi:10.1016/j.semcdb.2004.03.001

394. Thies, F., Nebe-von-Caron, G., Powell, J.R., Yaqoob, P., Newsholme, E.A. & Calder, P.C. Dietary supplementation with gamma-linolenic acid or fish oil decreases T lymphocyte proliferation in healthy older humans. J. Nutr. (2001), 131(7):1918–27.

395. Calder, P.C. N-3 Polyunsaturated Fatty Acids, Inflammation, and Inflammatory Diseases. Am. J. Clin. Nutr. (2006), 83(6 Suppl):1505S–1519S.

396. Calder, P.C. N-3 Fatty Acids, Inflammation and Immunity: New Mechanisms To Explain Old Actions. Proc. Nutr. Soc. (2013), 72(3):326–36. doi:10.1017/S0029665113001031

397. Shaikh, S.R. & Teague, H. N-3 fatty acids and membrane microdomains: from model membranes to lymphocyte function. Prostaglandins. Leukot. Essent. Fatty Acids (2012), 87(6):205–8. doi:10.1016/j.plefa.2012.09.007

398. Stillwell, W. & Wassall, S.R. Docosahexaenoic acid: membrane properties of a unique fatty acid. Chem. Phys. Lipids (2003), 126(1):1–27. doi:10.1016/S0009-3084(03)00101-4

399. Wada, M., DeLong, C.J., Hong, Y.H., Rieke, C.J., Song, I., Sidhu, R.S., Yuan, C., Warnock, M., Schmaier, A.H., Yokoyama, C., Smyth, E.M., Wilson, S.J., FitzGerald, G. a, Garavito, R.M., Sui, D.X., Regan, J.W. & Smith, W.L. Enzymes and receptors of prostaglandin pathways with arachidonic acid-derived versus eicosapentaenoic acid-derived substrates and products. J. Biol. Chem. (2007), 282(31):22254–66. doi:10.1074/jbc.M703169200

400. Issazadeh-Navikas, S., Teimer, R. & Bockermann, R. Influence of dietary components on regulatory T cells. Mol. Med. (2012), 1895–110. doi:10.2119/molmed.2011.00311

401. Nagaraju, K., Raben, N., Loeffler, L., Parker, T., Rochon, P.J., Lee, E., Danning, C., Wada, R., Thompson, C., Bahtiyar, G., Craft, J., Hooft Van Huijsduijnen, R. & Plotz, P. Conditional up-regulation of MHC class I in skeletal muscle leads to self-sustaining autoimmune myositis and myositis-specific autoantibodies. Proc. Natl. Acad. Sci. U. S. A. (2000), 97(16):9209–14.

402. Cheng, H.-Y., Wu, R., Gebre, A.K., Hanna, R.N., Smith, D.J., Parks, J.S., Ley, K. & Hedrick, C.C. Increased cholesterol content in gammadelta (γδ) T lymphocytes differentially regulates their activation. PLoS One (2013), 8(5):e63746. doi:10.1371/journal.pone.0063746

403. Murai, T., Sato, C., Sato, M., Nishiyama, H., Suga, M., Mio, K. & Kawashima, H. Membrane cholesterol modulates the hyaluronan-binding ability of CD44 in T lymphocytes and controls rolling under shear flow. J. Cell Sci. (2013), 126(Pt 15):3284–94. doi:10.1242/jcs.120014

404. Mahammad, S., Dinic, J., Adler, J. & Parmryd, I. Limited cholesterol depletion causes aggregation of plasma membrane lipid rafts inducing T cell activation. Biochim. Biophys. Acta (2010), 1801(6):625– 34. doi:10.1016/j.bbalip.2010.02.003

405. Surls, J., Nazarov-Stoica, C., Kehl, M., Olsen, C., Casares, S. & Brumeanu, T.-D. Increased membrane cholesterol in lymphocytes diverts T-cells toward an inflammatory response. PLoS One (2012), 7(6):e38733. doi:10.1371/journal.pone.0038733

406. Elson, E.L., Fried, E., Dolbow, J.E. & Genin, G.M. Phase separation in biological membranes: integration of theory and experiment. Annu. Rev. Biophys. (2010), 39207–26. doi:10.1146/annurev.biophys.093008.131238

407. Killian, J.A. & Nyholm, T.K.M. Peptides in lipid bilayers: the power of simple models. Curr. Opin. Struct. Biol. (2006), 16(4):473–9. doi:10.1016/j.sbi.2006.06.007

408. Park, S.H. & Opella, S.J. Tilt angle of a trans-membrane helix is determined by hydrophobic mismatch. J. Mol. Biol. (2005), 350(2):310–8. doi:10.1016/j.jmb.2005.05.004

409. Parasassi, T., Di Stefano, M., Loiero, M., Ravagnan, G. & Gratton, E. Cholesterol modifies water concentration and dynamics in phospholipid bilayers: a fluorescence study using Laurdan probe. Biophys. J. (1994), 66(3 Pt 1):763–8.

410. Weber, G. & Farris, F.J. Synthesis and spectral properties of a hydrophobic fluorescent probe: 6- propionyl-2-(dimethylamino)naphthalene. Biochemistry (1979), 18(14):3075–8.

411. Parasassi, T., Krasnowska, E., Bagatolli, L. & Gratton, E. Laurdan and Prodan as Polarity-Sensitive Fluorescent Membrane Probes. J. Fluoresc. (1998), 8(4):365–373. doi:10.1023/A:1020528716621

412. Parasassi, T., De Stasio, G., D’Ubaldo, A. & Gratton, E. Phase fluctuation in phospholipid membranes revealed by Laurdan fluorescence. Biophys. J. (1990), 57(6):1179–86. doi:10.1016/S0006- 3495(90)82637-0

413. Gaus, K., Zech, T. & Harder, T. Visualizing membrane microdomains by Laurdan 2-photon microscopy. Mol. Membr. Biol. (2006), 23(1):41–8. doi:10.1080/09687860500466857

414. Moravcevic, K., Oxley, C.L. & Lemmon, M. a Conditional peripheral membrane proteins: facing up to limited specificity. Structure (2012), 20(1):15–27. doi:10.1016/j.str.2011.11.012

415. Becker, T., Horvath, S.E., Böttinger, L., Gebert, N., Daum, G. & Pfanner, N. Role of phosphatidylethanolamine in the biogenesis of mitochondrial outer membrane proteins. J. Biol. Chem. (2013), 288(23):16451–9. doi:10.1074/jbc.M112.442392

416. Yeagle, P.L. Non-covalent binding of membrane lipids to membrane proteins. Biochim. Biophys. Acta (2013), doi:10.1016/j.bbamem.2013.11.009

417. Slotte, J.P. Biological functions of sphingomyelins. Prog. Lipid Res. (2013), 52(4):424–37. doi:10.1016/j.plipres.2013.05.001

418. Stillwell, W. Membrane Lipids : Fatty Acids. An Introd. to Biol. Membr. (2013), 43–56. doi:10.1016/B978-0-444-52153-8.00004-0

419. Chan, D.I. & Vogel, H.J. Current understanding of fatty acid biosynthesis and the acyl carrier protein. Biochem. J. (2010), 430(1):1–19. doi:10.1042/BJ20100462

420. Christie, W.W. Fatty Acids: Straight-Chain Saturated. Structures, Occurrence and Biosynthesis. AOCS Lipid Libr. (2014), 1–6.

421. Sud, M., Fahy, E., Cotter, D., Brown, A., Dennis, E.A., Glass, C.K., Merrill, A.H., Murphy, R.C., Raetz, C.R.H., Russell, D.W. & Subramaniam, S. LMSD: LIPID MAPS structure database. Nucleic Acids Res. (2007), 35(Database issue):D527–32. doi:10.1093/nar/gkl838

422. Christie, W.W. Fatty Acids: Methylene-interrupted double bonds. Structures, Occurrence and Biochemistry. AOCS Lipid Libr. (2013), 1–10.

423. Matyash, V., Liebisch, G., Kurzchalia, T. V, Shevchenko, A. & Schwudke, D. Lipid extraction by methyl- tert-butyl ether for high-throughput lipidomics. J. Lipid Res. (2008), 49(5):1137–46. doi:10.1194/jlr.D700041-JLR200

424. Folch, J., Lees, M. & Sloane Stanley, G.H. A simple method for the isolation and purification of total lipides from animal tissues. J. Biol. Chem. (1957), 226(1):497–509.

425. Ramstedt, B., Leppimäki, P., Axberg, M. & Slotte, J.P. Analysis of natural and synthetic sphingomyelins using high-performance thin-layer chromatography. Eur. J. Biochem. (1999), 266(3):997–1002.

426. Mueller, H.W., O’Flaherty, J.T., Greene, D.G., Samuel, M.P. & Wykle, R.L. 1-O-alkyl-linked glycerophospholipids of human neutrophils: distribution of arachidonate and other acyl residues in the ether-linked and diacyl species. J. Lipid Res. (1984), 25(4):383–8.

427. Guo, Y., Cordes, K.R., Farese, R. V & Walther, T.C. Lipid droplets at a glance. J. Cell Sci. (2009), 122(Pt 6):749–52. doi:10.1242/jcs.037630

428. Rouquette-Jazdanian, A.K., Pelassy, C., Breittmayer, J.-P. & Aussel, C. Full CD3/TCR activation through cholesterol-depleted lipid rafts. Cell. Signal. (2007), 19(7):1404–18. doi:10.1016/j.cellsig.2007.01.015

429. Leidl, K., Liebisch, G., Richter, D. & Schmitz, G. Mass spectrometric analysis of lipid species of human circulating blood cells. Biochim. Biophys. Acta (2008), 1781(10):655–64. doi:10.1016/j.bbalip.2008.07.008

430. Sugiura, T., Masuzawa, Y. & Waku, K. Alkenyl and alkyl ether phospholipids in pig mesenteric lymph node lymphocytes. Lipids (1980), 15(6):475–8.

431. Sze, D.Y. & Jardetzky, O. Characterization of lipid composition in stimulated human lymphocytes by 1H-NMR. Biochim. Biophys. Acta - Mol. Cell Res. (1990), 1054(2):198–206. doi:10.1016/0167- 4889(90)90241-5

432. Christie, W.W. Phosphatidic acid, Lysophosphatidic acid and related lipids. AOCS Lipid Libr. (2013), 1– 8.

433. Christie, W.W. Phosphatidylglycerol and related lipids. AOCS Lipid Libr. (2013), 1–7.

434. Okano, G. & Akino, T. Variations in the molecular species of lung phosphatidylglycerol. Lipids (1979), 14(6):541–6.

435. Mahadevappa, V.G. & Holub, B.J. The molecular species composition of individual diacyl phospholipids in human platelets. Biochim. Biophys. Acta (1982), 713(1):73–9.

436. Christie, W.W. Phosphatidylinositol and related lipids. AOCS Lipid Libr. (2013), 111(1976):1–7.

437. Thompson, W. & MacDonald, G. Isolation and characterization of cytidine diphosphate diglyceride from beef liver. J. Biol. Chem. (1975), 250(17):6779–85.

438. Wood, R. & Harlow, R.D. Structural analyses of rat liver phosphoglycerides. Arch. Biochem. Biophys. (1969), 135(1):272–81.

439. Thompson, W. & MacDonald, G. Cytidine diphosphate diglyceride of bovine brain. Positional distribution of fatty acids and analysis of major molecular species. Eur. J. Biochem. (1976), 65(1):107– 11.

440. Kelly, K. & Jacobs, R. Phospholipid biosynthesis. AOCS LipidLibrary (2011), 1–9.

441. Morris, R.J. Ionic control of the metastable inner leaflet of the plasma membrane: Fusions natural and artefactual. FEBS Lett. (2010), 584(9):1665–9. doi:10.1016/j.febslet.2009.11.017

442. Christie, W.W. Phosphatidylethanolamine and related lipids. AOCS Lipid Libr. (2013), 501

443. Chin, J.H. & Goldstein, D.B. Membrane-Disordering of Ethanol Variation with Membrane Cholesterol Content and Depth of the Spin Label Probe. Mol. Pharmacol. (1981), 19425–431.

444. Storch, J. & Kleinfeld, A. The lipid structure of biological membranes. Trends Biochem. Sci. (1985), (November):418–421.

445. Vance, J.E. & Steenbergen, R. Metabolism and functions of phosphatidylserine. Prog. Lipid Res. (2005), 44(4):207–34. doi:10.1016/j.plipres.2005.05.001

446. Zachowski, A. Phospholipids in animal eukaryotic membranes: transverse asymmetry and movement. Biochem. J. (1993), 294 ( Pt 11–14.

447. Vámosi, G., Bodnár, A., Vereb, G., Jenei, A., Goldman, C.K., Langowski, J., Tóth, K., Mátyus, L., Szöllösi, J., Waldmann, T. a & Damjanovich, S. IL-2 and IL-15 receptor alpha-subunits are coexpressed in a supramolecular receptor cluster in lipid rafts of T cells. Proc. Natl. Acad. Sci. U. S. A. (2004), 101(30):11082–7. doi:10.1073/pnas.0403916101

448. Cho, J.-H., Kim, H.-O., Surh, C.D. & Sprent, J. T cell receptor-dependent regulation of lipid rafts controls naive CD8+ T cell homeostasis. Immunity (2010), 32(2):214–26. doi:10.1016/j.immuni.2009.11.014

449. Singh, P., Saxena, R., Srinivas, G., Pande, G. & Chattopadhyay, A. Cholesterol biosynthesis and homeostasis in regulation of the cell cycle. PLoS One (2013), 8(3):e58833. doi:10.1371/journal.pone.0058833

450. Bensinger, S.J., Bradley, M.N., Joseph, S.B., Zelcer, N., Janssen, E.M., Hausner, M.A., Shih, R., Parks, J.S., Edwards, P. a, Jamieson, B.D. & Tontonoz, P. LXR signaling couples sterol metabolism to proliferation in the acquired immune response. Cell (2008), 134(1):97–111. doi:10.1016/j.cell.2008.04.052

451. D’Acquisto, F., Perretti, M. & Flower, R.J. Annexin-A1: a pivotal regulator of the innate and adaptive immune systems. Br. J. Pharmacol. (2008), 155(2):152–69. doi:10.1038/bjp.2008.252

452. Iaccarino, L., Ghirardello, A., Canova, M., Zen, M., Bettio, S., Nalotto, L., Punzi, L. & Doria, A. Anti- annexins autoantibodies: their role as biomarkers of autoimmune diseases. Autoimmun. Rev. (2011), 10(9):553–8. doi:10.1016/j.autrev.2011.04.007

453. Munoz, L., Frey, B., Pausch, F., Baum, W., Mueller, R., Brachvogel, B., Poschl, E., Rodel, F., von der Mark, K., Herrmann, M. & Gaipl, U. The Role of Annexin A5 in the Modulation of the Immune Response Against Dying and Dead Cells. Curr. Med. Chem. (2007), 14(3):271–277. doi:10.2174/092986707779941131

454. Belluoccio, D., Grskovic, I., Niehoff, A., Schlötzer-Schrehardt, U., Rosenbaum, S., Etich, J., Frie, C., Pausch, F., Moss, S.E., Pöschl, E., Bateman, J.F. & Brachvogel, B. Deficiency of annexins A5 and A6 induces complex changes in the transcriptome of growth plate cartilage but does not inhibit the induction of mineralization. J. Bone Miner. Res. (2010), 25(1):141–53. doi:10.1359/jbmr.090710

455. Sawami, H., Terada, N., Franklin, R. a, Okawa, H., Uchiyama, T., Lucas, J.J. & Gelfand, E.W. Signal transduction by interleukin 2 in human T cells: activation of tyrosine and ribosomal S6 kinases and cell-cycle regulatory genes. J. Cell. Physiol. (1992), 151(2):367–77. doi:10.1002/jcp.1041510218

456. Hulce, J.J., Cognetta, A.B., Niphakis, M.J., Tully, S.E. & Cravatt, B.F. Proteome-wide mapping of cholesterol-interacting proteins in mammalian cells. Nat. Methods (2013), 10(3):259–64. doi:10.1038/nmeth.2368

457. Jastrzebska, B., Debinski, A., Filipek, S. & Palczewski, K. Role of membrane integrity on G protein- coupled receptors: Rhodopsin stability and function. Prog. Lipid Res. (2011), 50(3):267–77. doi:10.1016/j.plipres.2011.03.002

458. Vereb, G. & Matko, J. Cholesterol-dependent clustering of IL-2Rα and its colocalization with HLA and CD48 on T lymphoma cells suggest their functional association with lipid rafts. Proc. … (2000), 97(11):

459. Kenworthy, A.K. Have we become overly reliant on lipid rafts? Talking Point on the involvement of lipid rafts in T-cell activation. EMBO Rep. (2008), 9(6):531–5. doi:10.1038/embor.2008.92

460. Zidovetzki, R. & Levitan, I. Use of cyclodextrins to manipulate plasma membrane cholesterol content: evidence, misconceptions and control strategies. Biochim. Biophys. Acta (BBA)- … (2007), 1768(6):1311–1324.

461. Marmor, M.D. Role for lipid rafts in regulating interleukin-2 receptor signaling. Blood (2001), 98(5):1489–1497. doi:10.1182/blood.V98.5.1489

462. Cuthbert, J. & East, C. Detection of familial hypercholesterolemia by assaying functional low-density- lipoprotein receptors on lymphocytes. New Engl. J. … (1986), 314(14):879–883.

463. A, C.J. & Lipsky, P.E. Provision of Cholesterol to Lymphocytes by High Density and Low Density Lipoproteins. J. Biol. Chem. (1987), 262(16):7808–7818.

464. Chyu, K.-Y., Lio, W.M., Dimayuga, P.C., Zhou, J., Zhao, X., Yano, J., Trinidad, P., Honjo, T., Cercek, B. & Shah, P.K. Cholesterol lowering modulates T cell function in vivo and in vitro. PLoS One (2014), 9(3):e92095. doi:10.1371/journal.pone.0092095

465. Wang, R. & Green, D.R. Metabolic checkpoints in activated T cells. Nat. Immunol. (2012), 13(10):907– 15. doi:10.1038/ni.2386

466. Frauwirth, K. a, Riley, J.L., Harris, M.H., Parry, R. V, Rathmell, J.C., Plas, D.R., Elstrom, R.L., June, C.H. & Thompson, C.B. The CD28 signaling pathway regulates glucose metabolism. Immunity (2002), 16(6):769–77.

467. Rathmell, J.C., Vander Heiden, M.G., Harris, M.H., Frauwirth, K. a & Thompson, C.B. In the absence of extrinsic signals, nutrient utilization by lymphocytes is insufficient to maintain either cell size or viability. Mol. Cell (2000), 6(3):683–92.

468. Tinel, M., Elkahwaji, J., Robin, M. a, Fardel, N., Descatoire, V., Haouzi, D., Berson, a & Pessayre, D. Interleukin-2 overexpresses c-myc and down-regulates cytochrome P-450 in rat hepatocytes. J. Pharmacol. Exp. Ther. (1999), 289(2):649–55.

469. Jeon, M., Kwon, H.J., Kim, Y.H., Han, K.-I., Nam, K.-W., Baik, Y., Lee, S., Kim, W.-J. & Han, M.-D. Pretreatment with recombinant human interleukin 2 (rhIL-2) Up-regulates PCNA-positive cells after partial hepatectomy in rat liver. Biotechnol. Bioprocess Eng. (2014), 19(1):43–51. doi:10.1007/s12257-013-0667-8

470. Wadamori, K., Oka, M., Tokuda, N., Fujikura, Y., Hazama, S., Fukumoto, T. & Suzuki, T. Influence of continuous interleukin-2 administration via the portal vein on liver regeneration following partial hepatectomy in rats. Hepatology (1996), 23(6):1578–83. doi:10.1053/jhep.1996.v23.pm0008675180

Cornely, R., Grewal, T., & Gaus, K. (2012). The Actin Cytoskeleton and Membrane Organisation in T Lymphocytes. In M. Kavallaris (Ed.), Cytoskeleton and Human Disease (pp. 103–121). Humana Press. doi:10.1007/978-1-61779-788-0_5

Cornely, R., Rentero, C., Enrich, C., Grewal, T., & Gaus, K. (2011). Annexin A6 is an organizer of membrane microdomains to regulate receptor localization and signalling. IUBMB life, 63(11), 1009–17. doi:10.1002/iub.540

For generating a stable knock-down of AnxA6 in Jurkat T cells a construct was purchased from Mission Lentiviral Transduction Particles were purchased from Sigma (Clone ID TRCN0000011461). shRNA sequence for AnxA6 knock-down: CCGGCGGGCACTTCTGCCAAGAAATCTCGAGATTTCTTGGCAGAAGTGCCCGTTTTT

Figure Appendix-1: Functionality of RNAi vector and Vector map of lentiviral vector with puromycin resistance gene. It is made up of the following sites: cppt (Central polypurine tract), hPGK (Human phosphoglycerate kinase eukaryotic promoter), puroR (Puromycin resistance gene for mammalian selection), SIN/LTR 3' (self inactivating long terminal repeat), f1 ori (f1 origin of replication), ampR (Ampicillin resistance gene for bacterial selection), pUC ori (pUC origin of replication), 5' LTR (5' long terminal repeat), Ψ Psi (RNA packaging signal), RRE (Rev response element).

Figure Appendix-2: Migration efficiency of CTRL and AnxA6KD Jurkat cells. The concentration of T cells in a target well of a Boyden chamber (5 µm pore size) filled with media or SDF-1α was determined after 3 hours of incubation. The migration efficiency of Control and AnxA6KD Jurkat cells were compared by 2-way ANOVA using Sidak’s multiple comparison post test. Data were pooled from 3 experiments with six replicates per experiment and analysed, error bars are SD with n=18, ****p<0.0001.

Since T cells exhibit a low activity of random migration, the lower compartment of the “Media” control contained some T cells after 3 hours of incubation even in the absence of chemoattractant (Figure Appendix-2). The numbers of Jurkat CTRL and AnxA6KD T cells found in the medium filled well after 3 hours of incubation were not significantly from each other. This indicated the levels of random migration after 3 hours of incubation were the same in CTRL and AnxA6KD cells. Significantly more CTRL and AnxA6KD cells were counted in the wells containing SDF-1α than in the media filled well. The numbers of CTRL and AnxA6KD Jurkat cells in the chemoattractant filled wells were not significantly different. The significantly higher amount of T cells found in the chemoattractant filled wells shows that the T cells actively migrated towards the SDF-1α containing wells. The lack of any significant differences between CTRL and AnxA6KD cells shows that the knock- down of AnxA6 did not reduce the efficiency of T cell migration towards a chemotactic cue.

Table 6-1: Primer sequences for realtime PCR assays

Mouse Primer Name Sequence (5’->3’) Full name of gene Accession number Source (Primerbank ID) ms IL-2 fwd (3) TCTGCGGCATGTTCTGGATTT Interleukin 2 (IL-2) NM_008366 PBID 291575157b1 ms IL-2 rev (3) ATGTGTTGTCAGAGCCCTTTAG Interleukin 2 (IL-2) NM_008366 PBID 291575157b1 ms GAPDH fwd (2) TGCACCACCAACTGCTTAG glyceraldehyde-3-phosphate dehydrogenase (GAPDH) NM_008084.2 Saada ms GAPDH rev (2) GGATGCAGGGATGATGTTC glyceraldehyde-3-phosphate dehydrogenase (GAPDH) NM_008084.2 Saada ms G6PDX fwd CTCCAATCAACTGTCGAACCA glucose-6-phosphate dehydrogenase (G6PDX) NM_008062 PBID 50053914b2 ms G6PDX rev TTGTCTCGATTCCAGATGGGG glucose-6-phosphate dehydrogenase (G6PDX) NM_008062 PBID 50053914b2

.