The Role of DOCK8 in Human Lymphocytes

Bethany Pillay

A thesis in fulfilment of the requirements for the degree of Doctor of Philosophy

St Vincent’s Clinical School, Faculty of Medicine UNSW Sydney

Immunology Division Garvan Institute of Medical Research

Supervisors: Professor Stuart Tangye and Associate Professor Cindy Ma

September 2019 Thesis/Dissertation Sheet

Surname/Family Name : PILLAY Given Name/s : BETHANY ANN Abbreviation for degree as give in the University : PhD calendar Faculty : MEDICINE School : ST VINCENT’S CLINICAL SCHOOL Thesis Title : THE ROLE OF DOCK8 IN HUMAN LYMPHOCYTES

Abstract 350 words maximum: (PLEASE TYPE) Dedicator of cytokinesis 8 (DOCK8) is a guanine nucleotide exchange factor that is highly expressed in lymphocytes and is involved in cytoskeletal rearrangement. Bi-allelic inactivating mutations in DOCK8 cause a combined immunodeficiency which is characterised by severe viral, bacterial and fungal infections, eczema, allergies and impaired humoral responses and which requires haematopoietic stem cell transplant (HSCT) to overcome. However, the cellular defects underlying these clinical symptoms have not been fully investigated. This thesis firstly examines the contribution of the CD4+ T cell compartment to the clinical features of DOCK8 deficiency by using cell cultures of DOCK8-deficient patients. This revealed skewed in vivo differentiation explaining patient susceptibility to infections and their propensity towards allergic conditions. To explore the cellular alterations which enable the clinical improvement of patients following HSCT, a comprehensive study of lymphocyte phenotype and function in DOCK8- deficient patients before and after HSCT was undertaken. This study found key functional differences which provided explanation for resolution of symptoms in patients post HSCT and noted persistent defects which may influence continuing patient care. Furthermore, investigation of a DOCK8-revertant patient allowed for the analysis of both DOCK8-deficient and DOCK8- revertant cells within the same individual and enabled the elucidation of the influence of environment on the identified defects of DOCK8 deficiency. Lastly, the extent of cellular impairment that was due to the involvement of DOCK8 in previously identified signalling pathways was examined with the use of immunodeficiency patients with mutations in DOCK8- related proteins. The findings of this analysis suggested that the outcomes of DOCK8 deficiency were not due solely to the role of DOCK8 in a single signalling pathway.

Declaration relating to disposition of project thesis/dissertation I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in 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 property 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 Abstracts International (this is applicable to doctoral theses only).

……………………………………………… ……………………………………..…… ……….……………………...…….… Signature Witness Signature Date The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research. FOR OFFICE USE ONLY Date of completion of requirements for Award: ii

iii

Acknowledgements

To my supervisors Stu and Cindy – thank you for agreeing to take me on despite me not having an immunology background. Thanks for all the help along the way and time spent with experiments and meetings and presentations and editing and countless emails to clinicians and collaborators to chase up details and ensure patient samples.

To Danielle – thank you for being my first port of call for almost any issue I had (including super gluing things back together) and for the tremendous number of things you do for the lab to make it a better place for us all.

To Kat and Geetha – thank you for showing me the ropes and putting up with me following you around and asking constant questions when I first started in the lab and then for agreeing to spend time with me outside the lab as a friend.

To my office buddies – thank you Simon for the interesting youtube rabbit holes we went down and thank you to my work wife Julia for being the person I talk to most.

To Emily and Frenchie – thank you to my English duo for all their vital words of encouragement and support as well as their friendship along the way.

To the rest of the lab – thanks for helping me out when I needed it, all the fun times and listening to my rants. I’m telling you, our tv show “Lab rats” would have been a hit!

To level 9 – thanks for always being happy to lend a hand or have a chat. It has been a privilege to be part of the department most obsessed with trivia.

To my family and friends – thank you for the support and for knowing to ask just enough questions to show you care but not to hassle me about when I would be finished.

To Alex – thank you for not knowing what you were getting into but for hanging in there with me anyway.

vii

Publications arising from this work

Pillay BA, Avery DT, Smart JM, Cole T, Choo S, Chan D, Gray PE, Frith K, Mitchell R, Phan TG, Wong M, Campbell DE, Hsu P, Ziegler JB, Peake J, Alvaro F, Picard C, Bustamante J, Neven B, Cant AJ, Uzel G, Arkwright PD, Casanova JL, Su HC, Freeman AF, Shah N, Hickstein DD, Tangye SG, Ma CS. JCI Insight. 2019; 25(5) Tangye SG, Bucciol G, Casas-Martin J, Pillay B, Ma CS, Moens L, Meyts I. Human inborn errors of the actin cytoskeleton affecting immunity: way beyond WAS and WIP. Immunology Cell Biol. 2019; 97(4):389-402 Béziat V, Li J, Lin JX, Ma CS, Li P, Bousfiha A, Pellier I, Zoghi S, Baris S, Keles S, Gray P, Du N, Wang Y, Zerbib Y, Lévy R, Leclercq T, About F, Lim AI, Rao G, Payne K, Pelham SJ, Avery DT, Deenick EK, Pillay B, Chou J, Guery R, Belkadi A, Guérin A, Migaud M, Rattina V, Ailal F, Benhsaien I, Bouaziz M, Habib T, Chaussabel D, Marr N, El-Benna J, Grimbacher B, Wargon O, Bustamante J, Boisson B, Müller- Fleckenstein I, Fleckenstein B, Chandesris MO, Titeux M, Fraitag S, Alyanakian MA, Leruez-Ville M, Picard C, Meyts I, Di Santo JP, Hovnanian A, Somer A, Ozen A, Rezaei N, Chatila TA, Abel L, Leonard WJ, Tangye SG, Puel A, Casanova JL. Science Immunology. 2018; 3(24) Tangye SG, Pillay B, Randall KL, Avery DT, Phan TG, Gray P, Ziegler JB, Smart JM, Peake J, Arkwright PD, Hambleton S, Orange J, Goodnow CC, Uzel G, Casanova JL, Lugo Reyes SO, Freeman AF, Su HC, Ma CS. Dedicator of cytokinesis8-deficient CD4+ T cells are biased to a TH2 effector fate at the expense of TH1 and TH17 cells. Journal of Allergy and Clinical Immunology. 2017; 139(3):933-949.

Presentations arising from this work Biennial meeting of the European Society for Immunodeficiencies, Lisbon, Portugal, 2018 Annual meeting of the Australasian Society for Immunology, Perth, Australia, 2018 ASI NSW branch meeting, Kiama, Australia, 2018 Annual meeting of the Australasian Society for Immunology, Brisbane, Australia, 2017 International Congress of Immunology, Melbourne, Australia, 2016

viii

Experiments that were not the sole work of the author

All experiments were performed by the author at the Garvan Institute of Medical Research with the following exceptions:

Fig 4.8A experiment performed by Danielle Priestley

Fig 6.4B-D (WAS and XLT patients) experiment performed by Danielle Priestley

Abbreviations

ABTS 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) AD autosomal dominant ADA adenosine deaminase deficiency APCs antigen presenting cells AR autosomal recessive BCR b cell receptor BSA bovine serum albumin CADD combined annotation dependent depletion CBA cytometric bead array CCR cc chemokine receptor CD cluster of differentiation CDC42 cell division control protein homolog cDNA complementary deoxyribonucleic acid CFSE carboxyfluorescein succinimidyl ester CMC chronic mucocutaneous candidiasis CMV cytomegalovirus CNS central nervous system CNV copy number variations CRISPR/Ca clustered regularly interspaced short palindromic repeats/ CRISPR- s9 associated protein 9 DC dendritic cells DHR dock homology region DMSO dimethyl sulfoxide DNA deoxyribonucleic acid dNTPs deoxynucleoside triphosphate DOCK2 dedicator of cytokinesis 2 DOCK8 dedicator of cytokinesis 8 DTT dithiothreitol EBV Epstein-Barr virus EDTA ethylenediaminetetraacetic acid ELISA enzyme-linked immunosorbent assay FA formaldehyde FACS fluorescence activated cell sorting F-actin filamentous actin FCS fetal calf serum FSC-A forward scatter area FSC-H forward scatter height GC germinal centre GEF guanine exchange factor gof gain of function GvHD graft vs host disease HIES hyper IgE syndrome x

HRP horse radish peroxidase HSCT haematopoietic stem cell transplant HSV herpes simplex virus ICAM-1 intercellular adhesion molecule 1 IFN interferon Ig immunoglobulin IL interleukin IS immune synapse IU international units IVIg intravenous immunoglobulin kU/l kilounit antibody per litre kUa/l kilounit of allergen-specific antibody per litre LB luria broth LCL lymphoblastoid cell line LFA-1 lymphocyte function-associated antigen 1 LN lymph nodes lof loss of function LRAP35a leucine-rich amelogenin protein 35a LRCH1 leucine rich repeats and calponin homology domain containing 1 mAb monoclonal antibody MAF minor allele frequency MAIT mucosal associated invariant t cell MFI mean fluorescence intensity MSC mutation significance cutoff NGS next generation sequencing NK natural killer NKT natural killer T cell PBMC peripheral blood mononuclear cells PBS phosphate buffered saline PCR polymerase chain reaction pDC plasmacytoid dendritic cells PID primary immunodeficiency PMA phorbol myristate acetate pSMAC peripheral supramolecular cluster qPCR quantitative polymerase chain reaction rAb rabbit antibody RBC red blood cell RIC reduced intensity conditioning RNA ribonucleic acid rpm revolutions per minute SAC heat-killed, formalin-fixed Staphylococcus aureus cells SCID severe combined immunodeficiency SEM standard error of the mean shRNA short hairpin RNA xi siRNA small interfering RNA SOC super optimal broth SSC-A side scatter area STAT3 signal transducer and activator of transcription 3 TAE T cell activation and expansion TBE tris-borate-EDTA TCM central memory T cell TCR T cell receptor TEM effector memory T cell TEMRA effector memory CD45RA expressing T cells Tfh T follicular helper cells TGF-B transforming growth factor beta Th T helper cell TLR toll-like receptor TNF tumour necrosis factor Treg regulatory T cell VZV varicella zoster virus WAS Wiskott-Aldrich syndrome WASP Wiskott-Aldrich syndrome protein WBC white blood cell WGS whole genome sequencing WT wild type XHIGM X chromosome-linked hyper IgM syndrome XLT X chromosome-linked thrombocytopenia X-SCID X chromosome-linked severe combined immunodeficiency ZNF341 zinc finger protein 341

xii

Chapter 1: Introduction and Aims ...... 1 1.1 Primary immunodeficiency (PID) ...... 1 1.1.1 History of PIDs ...... 1 1.1.2 Insights from PIDs ...... 2 1.1.3 Characterisation of PIDs ...... 3 1.1.4 Treatment of PIDs ...... 4 1.2 Hyper IgE syndrome (HIES)...... 5 1.2.1 Autosomal dominant HIES ...... 5 1.2.2 Autosomal recessive HIES ...... 6 1.3 Dedicator of Cytokinesis 8 (DOCK8) deficiency ...... 6 1.3.1 Location and nature of mutations ...... 6 1.3.2 Clinical features of DOCK8 deficiency ...... 7 1.3.3 Immunological features of DOCK8 deficiency ...... 9 1.3.4 Treatment ...... 10 1.4 DOCK8 protein ...... 11 1.4.1 Structure ...... 11 1.4.2 Expression ...... 11 1.5 DOCK8 function...... 12 1.5.1 Immune synapse formation ...... 12 1.5.2 Migration ...... 14 1.6 Aims ...... 16 Chapter 2: Materials and Methods ...... 17 2.1 Cell isolation ...... 17 2.1.1 Materials ...... 17 2.1.2 Isolation of Peripheral Blood Mononuclear Cells ...... 17 2.1.3 Isolation of lymphocyte subsets by fluorescence activated cell sorting (FACS) ...... 18 2.2 Cell culture ...... 19 2.2.1 Materials ...... 19 2.2.2 T cell cultures ...... 20 2.2.3 B cell cultures ...... 20 2.2.4 CFSE ...... 20 2.3 Flow Cytometry ...... 21 2.3.1 Materials ...... 21 2.3.2 Cell surface staining ...... 21 2.3.3 Intracellular staining ...... 23 2.3.4 Cytometric Bead Array (CBA) ...... 24 2.3.5 Cell quantification ...... 26 2.3.6 Calcium flux ...... 27 2.3.7 Analysis of flow cytometry data ...... 27 2.4 Immunoglobulin detection...... 28 2.4.1 Materials ...... 28 2.4.2 Enzyme linked immunosorbent assay (ELISA) ...... 28 2.4.3 ImmunoCAP assay ...... 29 2.5 Molecular studies ...... 30 2.5.1 Materials ...... 30 2.5.2 RNA extraction ...... 31 2.5.3 cDNA synthesis ...... 31 2.5.4 Quantitative PCR (qPCR) ...... 31 2.5.5 DNA extraction ...... 32 xiii

2.5.6 Polymerase Chain Reaction (PCR) ...... 32 2.5.7 Agarose gel electrophoresis ...... 33 2.5.8 Gene sequencing ...... 33 2.5.9 TA cloning ...... 34 2.5.10 Western blotting ...... 34 2.6 Statistical analysis ...... 35 Chapter 3: CD4+ T cell differentiation in DOCK8 deficiency ...... 36 3.1 Introduction ...... 36 3.2 Results ...... 38 3.2.1 DOCK8 is expressed in naïve and memory CD4+ T cells in human peripheral blood, tonsil and spleen ...... 38 3.2.2 DOCK8 expression is maintained in in vitro cultured CD4+ T cells ...... 38 3.2.3 DOCK8-deficient memory CD4+ T cells are skewed to a Th2 fate at the expense of Th1 and Th17 cells ...... 41 3.2.4 The skewed CD4+ T cell differentiation of DOCK8-deficient memory CD4+ T cells is not due solely to aberrant expression of lineage-specific transcription factors ...... 43 3.2.5 TCR Vβ repertoire diversity is maintained in DOCK8-deficient memory CD4+ T cells ...... 45 3.2.6 DOCK8-deficient memory CD4+ T cells exhibit signs of decreased in vitro TCR-induced activation ...... 47 3.2.7 DOCK8-deficient naïve CD4+ T cells are able to undergo Th1 and Th2 differentiation, but not Th17 differentiation in vitro ...... 49 3.2.8 DOCK8-deficient naive CD4+ T cells do not show signs of decreased in vitro TCR-induced activation ...... 51 3.2.9 DOCK8-deficient naïve and memory CD4+ T cells do not display decreased TCR-induced calcium flux ...... 53 3.3 Discussion ...... 56 Chapter 4: Haematopoietic stem cell transplant effectively rescues lymphocyte differentiation and function in DOCK8-deficient patients ...... 61 4.1 Introduction ...... 61 4.2 Results ...... 64 4.2.1 DOCK8 is constitutively expressed by lymphocytes in healthy donors and DOCK8-deficient patients post-HSCT ...... 64 4.2.2 Clinical Characteristics of DOCK8 deficient patients – impact of HSCT ..... 66 4.2.3 DOCK8-deficient lymphocytes exhibit a unique phenotype typical of aberrant in vivo differentiation ...... 71 4.2.4 Impact of HSCT on lymphocyte differentiation in vivo in DOCK8-deficient patients ...... 71 4.2.5 Memory T cells in DOCK8-deficient patients exhibit signs of exhaustion, some of which persist after HSCT ...... 74 4.2.6 Defects in proliferation, acquisition of cytotoxic effector function and cytokine secretion by CD8+ T cells in DOCK8-deficient patients are improved by HSCT ...... 77 4.2.7 Dysregulated cytokine production by CD4+ T cells in DOCK8-deficient patients is normalised by HSCT ...... 79 4.2.9 HSCT overcomes the B-cell intrinsic impairment in survival, proliferation and differentiation due to DOCK8 deficiency ...... 83 4.2.10 Elevated serum IgE and allergen specific IgE levels decrease in a time- dependent manner following HSCT of DOCK8-deficient patients...... 85 xiv

4.2.11 Impact of HSCT on lymphocyte reconstitution in other PID patients ...... 87 4.3 Discussion ...... 91 Chapter 5: Investigation of reversion in a DOCK8-deficient patient ...... 96 5.1 Introduction ...... 96 5.2 Results ...... 98 5.2.1 Clinical features and identification of bi-allelic DOCK8 variants ...... 98 5.2.2 Detection of variable levels of expression of DOCK8 protein in different immune cell subsets from the patient ...... 101 5.2.3 DOCK8 reversion occurs due to repair of the de novo missense mutation . 104 5.2.4 The extent of reversion varies depending on the lymphocyte subset and stage of differentiation ...... 106 5.2.5 Immune cell phenotyping reveals similarities with typical DOCK8 deficiency ...... 108 5.2.6 DOCK8 reversion partially restores effector function of T cells ...... 110 5.2.7 Memory, but not naïve, B cells from the DOCK8-revertant patient exhibit improved function over DOCK8-deficient B cells ...... 114 5.2.8 DOCK8+ and DOCK8- lymphocytes in the revertant patient exhibit distinct phenotypes that replicate features of healthy donors and DOCK8-deficient patients ...... 117 5.2.9 DOCK8-revertant CD8+ T cells show reduced signs of exhaustion and, unlike DOCK8-deficient cells, efficiently undergo activation and cytokine production in vitro ...... 119 5.2.10 DOCK8-revertant memory CD4+ T cells display a different cytokine production profile to DOCK8-deficient cells ...... 119 5.2.11 DOCK8-revertant CD4+ memory and CD8+ T cells have a selective advantage over DOCK8-deficient cells in vitro ...... 123 5.2.11 Broad TCR diversity is present in both DOCK8-revertant and DOCK8- deficient T cells ...... 123 5.2.12 DOCK8-revertant lymphocytes increase over time ...... 127 5.2.13 Patient IgE levels decline over time ...... 127 5.3 Discussion ...... 131 Chapter 6: Lymphocyte phenotype and function in immunodeficiency patients with mutations in DOCK8-related proteins ...... 135 6.1 Introduction ...... 135 6.2 Results ...... 137 6.2.1 Patients used in this study ...... 137 6.2.2 Phenotypic analysis of patients with mutations in DOCK8-related proteins ...... 137 6.2.3 CD8+ T cell function in patients with mutations in DOCK8-related proteins ...... 141 6.2.4 CD4+ T cell function in patients with mutations in DOCK8-related proteins ...... 143 6.2.5 Naïve B cell function in patients with mutations in DOCK8-related proteins ...... 145 6.3 Discussion ...... 147 Chapter 7: Conclusions ...... 152 7.1 Research Outcomes ...... 152 7.2 Future Directions ...... 154

Chapter 1: Introduction and Aims The immune system has developed throughout evolution to adapt to the changing threats it has faced. The human immune system provides a well-coordinated defence against pathogens and cancerous cells while also regulating responses to prevent overactivation. The underlying function of the immune system is to respond to foreign antigenic threats while simultaneously tolerating self-antigens. This is achieved by deploying effector mechanisms of two main distinct arms of the immune system. Innate immunity is considered the first line of defence as it is the initial and rapid response which involves recognition of conserved molecules shared by a broad range of pathogens. The innate immune response results in recruitment of cells such as monocytes, macrophages and dendritic cells to clear infection as well as production of molecular messengers (cytokines and chemokines) which activate the adaptive immune response. Adaptive immunity is characterised by its ability to recognise specific antigens and is a slower but more effective immune response. Additionally, adaptive immunity involves the generation of immunological memory, whereby a swift and vigorous response is following re-exposure to previously encountered pathogens. This introduction will describe patients with impaired immunity due to inborn errors and focus on a particular primary immunodeficiency.

1.1 Primary immunodeficiency (PID) Primary immunodeficiency (PID) is a condition categorised by immune dysfunction due to a genetic cause. The International Union of Immunological Societies (IUIS) Inborn Errors of Immunity (IEI) Committee, which is responsible for cataloguing immunological disorders, as of January 2018 recognised 330 distinct disorders caused by 320 monogenic defects affecting 312 genes1. Many PIDs present during childhood or even as early as birth, but many patients are also diagnosed in adulthood.

1.1.1 History of PIDs Given that PID patients are commonly identified by their susceptibility to infections, it follows that PIDs were not identified until the mid-1900s. Prior to that time, infections were common in healthy people and it was not until these infections were able to be controlled by the implementation of aseptic techniques by clinicians as well as the 1

introduction of antibiotics and vaccinations that PID patients were revealed as unusual for their recurrent and often severe infections2. The first recognised case of PID was in 1952 in a patient with a high number of episodes of pneumonia who was found to lack the gammaglobulin fraction of serum proteins and was termed agammaglobulinemia3. The next PID was reported a few years later when two papers detailed four infants who displayed a lack of lymphocytes, now termed lymphopenia, and died of severe fungal and bacterial infections and it was proposed that the combination of lymphopenia and agammaglobulinemia led to their deaths4,5. Today, we know this condition as severe combined immunodeficiency (SCID). In 1972 it was discovered that a blood sample from a SCID patient lacked adenosine deaminase (ADA) and for the first time a genetic defect responsible for a PID was identified6. This, along with advances in molecular techniques such as Sanger sequencing and the polymerase chain reaction (PCR), sparked a quest during the 1980s to map genes responsible for PID. This led to the discovery that mutations in IL2RG, encoding the IL2R γ chain, were responsible for X chromosome- linked SCID (X-SCID)7,8 and mutations in BTK resulted in X-linked agammaglobulinemia9,10 among many similar findings. Over the past 2 decades, improvements in sequencing technology, particularly the introduction of next generation sequencing (NGS), has enabled the identification of a myriad of PID-causing genetic mutations.

1.1.2 Insights from PIDs An illustration of the principle that insights into the function of the immune system can be gained from the study of rare PID but well-defined patients was in fact evidenced from the first reported case. Upon the discovery that the PID patient had no serum gammaglobulin, Bruton hypothesised that gammaglobulin prophylaxis/replacement therapy would prevent infections3. Indeed, this was elegantly demonstrated by the prevention of recurrent infections by regular injections of pooled serum fractions isolated from healthy donors. This depicted the efficacy of antibody replacement therapy, which is used widely today to treat not just individuals with agammaglobulinemia due to gene defects that severely compromise B-cell development, but also patients with antibody deficiency due to unknown genetic cause, such as common variable immunodeficiency, the most frequency PID.

The discovery of patients with severe recurrent infections who were found to actually have elevated levels of immunoglobulin11 led to the important finding that immunoglobulin alone is not responsible for protection against infections. The discovery of SCID patients who lacked T and B cells led to the insight that both cellular and humoral immunity are fundamental for a functioning immune system. Additionally, a PID with high levels of IgM but low levels of IgG and IgA and impaired antibody responses (X- linked Hyper IgM, XHIGM) was considered to be the result of an intrinsic B cell defect. However, subsequent discovery of a T cell lymphoma patient with the opposite pattern of immunoglobulin production (high IgG and IgA but absent IgM) led to further investigation. When T cells from this lymphoma patient were cultured with B cells from a patient with XHIGM, immunoglobulin production by B cells was corrected, indicating a T cell defect which subsequently affected B cell isotype switching12. Further studies revealed that mutations in CD40L, encoding CD40 ligand which is transiently expressed on activated CD4+ T cells, was the genetic basis for HIGM13-17. Critically, this discovery established that the interaction of CD40, expressed on B cells and other antigen presenting cells, with CD40L was an integral element for T-dependent B cell activation, differentiation and humoral immunity 18. Since then, PID patients have also been found with mutations in CD4019. In more recent years, the importance of various pathways in the control of distinct pathogen infections has been revealed by the study of infectious susceptibility of PID patients20 and PID research has also shed light on the pathways involved in immune regulation, including allergy, autoimmunity, autoinflammation and even malignancy21.

1.1.3 Characterisation of PIDs With the discovery of more patients affected by inborn errors of immunity and advances in laboratory techniques, characterisation of the molecular, biochemical and cellular defects underlying disease pathogenesis of PIDs has progressed at a rapid rate. The initial characterisations of PIDs were crudely based on the presence or absence of different lymphocytes. For example, patients without T and NK cells but with B cells (T-NK-B+) were classified as having mutations in either IL2RG or JAK3 which were distinguished from each other by their patterns of inheritance (i.e. X-linked [IL2RG] versus autosomal recessive [JAK3])22. Conversely, T-B-NK+ patients were considered likely to harbour mutations in RAG1 or RAG2 while T-B-NK- suggested mutations in adenylate kinase 2 3

(AK2)22. Since then, susceptibility of patients to specific pathogens has provided strong clues to enable identification of likely underlying genetic causes of their conditions. For example, Mendelian susceptibility to mycobacterial disease (MSMD) is linked to loss-of function (LOF) mutations in genes involved in production of or response to IFNγ (IFNGR1, IFNGR2, STAT2, IL12B, IL12TB1, ISG15, IRF8, NEMO, CYBB)23. Similarly, chronic mucocutaneous candidiasis (CMC) is an indicator of possible LOF mutations in STAT3, AIRE, IL17RA, IL17RC, IL17F, TYK2 or DOCK8, or gain of function (GOF) mutations in STAT124.

PIDs are currently classified into 9 groups by the IUIS IEI Committee based on pathogenesis (1. Immunodeficiencies affecting cellular and humoral immunity; 2. CID with associated or syndromic features; 3. Predominantly antibody deficiencies; 4. Diseases of immune dysregulation; 5. Congenital defects of phagocyte number, function or both; 6. Defects in intrinsic and innate immunity; 7. Auto-inflammatory disorders; 8. Complement deficiencies; 9. Phenocopies of PID) and phenotypic details are provided to assist clinician diagnosis1,25.

1.1.4 Treatment of PIDs Given that the immune dysfunction present in PIDs often leads to compromised immune defence against pathogens and hence recurrent and severe infections, the development and subsequent availability of antibiotics, antivirals and antifungals has greatly benefited the treatment of PID patients. Additionally, the development of antibody replacement in the form of intravenous immunoglobulin (IVIg), and more recently subcutaneous Ig (SCIg) which can be self-administered in a patient’s home, have continued to improve the quality of life for PID patients. In fact, IVIg is now used to treat over 100 inflammatory and autoimmune disorders26. Other therapies have also been established for specific conditions such as IFNγ therapy for chronic granulomatous disease (CGD) and IL12Rβ1-deficient patients to prevent mycobacterial infections27 and bovine-derived adenosine deaminase for ADA-deficient patients28. Haematopoietic stem cell transplant (HSCT) for PID was begun as early as 1968 for a patient with SCID and another with WAS29,30 but for quite some time success was limited to patients with a HLA-identical sibling donor. Further work on modifying donor samples prior to transplant expanded the use and more recent development of treatments to prevent graft versus host disease

(GvHD) and other post-transplant complications as well as reduced intensity conditioning (RIC) regimens has enabled much higher rates of success31. These advances, together with earlier identification of patients, the establishment of bone marrow registries and cord blood stores, have resulted in increased availability of HSCT for PID patients, which has for some patient cohorts become the standard of care.

1.2 Hyper IgE syndrome (HIES) Hyper IgE syndrome (HIES) was considered to be one of the last major PIDs without a known underlying genetic cause. It was originally reported as Job’s syndrome in 1966 in 2 patients with ‘cold’ staphylococcal abscesses. Over the next ~30 years, with the discovery of more patients, the clinical features were expanded to include eczema, eosinophilia, recurrent pneumonia and serum IgE levels increased >10 fold over the upper limit of the normal range32. Subsequent work identified patient cohorts with either a dominant33 or recessive34 inheritance pattern that appeared to be distinct disease entities.

1.2.1 Autosomal dominant HIES Autosomal dominant HIES patients (AD-HIES) were reported as unique in their manifestation of non-immunological features. A study of 30 patients revealed that a majority of the patients retained their primary teeth, and had recurrent fractures, hyperextensible joints and scoliosis (after the age of 16)33. Additionally, patients developed distinctive facial features by the age of 16 including asymmetry, prominent forehead, high palate and a broad nasal bridge. It was not until 2007 that monoallelic dominant negative mutations in signal transducer and activator of transcription 3 (STAT3) were found to be the cause of AD-HIES35,36. Since then, genetic analysis has allowed for better definition of STAT3 deficiency and associated clinical manifestation. Eczema, elevated serum IgE and eosinophilia continue to be the most frequent immunological symptoms (>90% of patients), followed by recurrent pneumonias, , CMC and newborn rash (>80% of patients)37. The non-immunological symptoms of intraoral lesions, characteristic facies and fractures are present in >70% of patients , while musculoskeletal and vascular abnormalities are also common (60-70%)37.

1.2.2 Autosomal recessive HIES Several years after the clinical characterisation of AD-HIES, 13 patients from consanguineous families were reported with an autosomal recessive form of HIES (AR- HIES). These patients shared clinical features of AD-HIES but lacked skeletal, connective tissue and dental abnormalities34. Furthermore, these patients could be distinguished from AD-HIES patients due to the high prevalence and severity of cutaneous viral infections as well as autoimmune and central nervous system complications34. Studies seeking to identify the genetic cause underlying AR-HIES identified an individual patient with a homozygous mutation in TYK238. However, genetic analysis of a larger collection of AR-HIES patients revealed mutations in TYK2 were unlikely to be the common cause of this condition39.

In 2009, biallelic recessive mutations in DOCK8 were uncovered by homozygosity mapping of patients with AR-HIES predominantly from consanguineous families and it was proposed that DOCK8 mutations accounted for the majority of AR-HIES cases40,41. Since then, recessive mutations in PGM3, ZNF341 or IL6ST, or dominant mutations in CARD11 have also been found in a limited number of patients who share clinical features of AR-HIES42-50. Despite this, DOCK8 mutations are the most prevalent cause of AR- HIES.

1.3 Dedicator of Cytokinesis 8 (DOCK8) deficiency 1.3.1 Location and nature of mutations The first reports of DOCK8 mutations relied on identifying copy number variations (CNV) and revealed homozygous large deletions of multiple exons or compound heterozygous deletions41 as well as mutations causing aberrant splicing, small deletions resulting in frameshift mutations, missense mutations resulting in single amino acid substitutions, and nonsense mutations introducing premature stop codons40. An analysis of a larger cohort of 60 patients uncovered 40 different mutations, with 72% being indels ranging from 2bp to multi-exon in length as well as several point mutations causing nonsense or splice site mutations51. Analysis by investigators at the NIH of their cohort of patients noted the prevalence of large deletions in at least one allele in 62% of unrelated patients (present in either homozygous or compound heterozygous form) which is likely due to the repetitive sequences in DOCK8 which allow for recombination52. Another

study identified a notable number of compound heterozygous patients who often had a large deletion on one allele coupled with a nonsense mutation or small indel on the other allele53. DOCK8 is encoded by 48 exons (www.enseml.org) and pathogenic mutations have been found to span the entire DOCK8 gene. There does not appear to be a particular region more likely to be targeted for mutations51 . The majority of DOCK8 mutations result in severely reduced or absent expression of DOCK8 protein, however two patients have been found to harbour loss-of function mutations in DOCK8 that do not abolish protein expression54,55.

1.3.2 Clinical features of DOCK8 deficiency Immunological DOCK8-deficient patients are highly susceptible to viral, bacterial and fungal infections. In the largest cohort study of DOCK8-deficient patients (n=136), 81% experienced viral infections most commonly with herpes viruses (herpes simplex virus [HSV], varicella zoster virus [VZV], Epstein Barr virus [EBV], cytomegalovirus [CMV])56. Viral infections were most prevalent in the skin51,57,58 especially infections with herpes simplex virus, Molluscum contagiosum40,56,57 and human papilloma virus (HPV). Bacterial infections of the skin also occur in DOCK8-deficient patients, generally due to Staphylococcus aureus which commonly gains entry when the skin barrier is compromised by eczema, a common feature of these patients52. Recurrent bacterial infections of the respiratory tract and pneumonia are also found in most DOCK8-deficient patients40,51,56,58, leading to bronchiectasis in a substantial proportion of cases (37%)51,56. Fungal infections, usually in the form of CMC, have been reported in ~30-70%51,56,58 of patients.

Allergic diseases are another hallmark of DOCK8 deficiency. Eczema or atopic dermatitis have consistently been observed in almost all patients51,56,58, is often severe and difficult to treat51,56. Serious allergies to environmental allergens and more often food allergens is a defining feature of DOCK8 deficiency compared to other immunodeficiencies52. Large cohort studies have determined the frequency of allergy to be ~65-70%in DOCK8 deficiency 51,56-58. The incidence of asthma is less than eczema and allergies but still present in substantially more patients than the general population (up to 54% vs up to10%)52,56. Increased levels of IgE are documented in almost all DOCK8-deficient

patients51,58 but span a wide range (e.g. 400-90910 IU/ml in one report40) and eosinophilia is observed in most patients41,51,56,58.

Responses to vaccination is varied amongst DOCK8-deficient patients. Protective levels of antibodies against rubella and varicella-zoster virus have been identified in patients but of those same patients fewer were responsive to vaccines against diphtheria toxoid, tetanus toxoid and H. influenzae type B41. Another report found only a limited number of patients (4/25) with poor response to pneumococcal and tetanus toxoid vaccination 58 while another included 5 patients all of whom exhibited defective antibody response to tetanus, hepatitis B, H. influenzae type B and polyvalent pneumococcal vaccines59. These varying results could be due to time since vaccination as protective antibody titers to tetanus toxoid measured in two DOCK8-deficient patients rapidly decreased over a period of 12 months59.

DOCK8-deficient patients have an increased risk of malignancy with 17% incidence reported in the largest investigated cohort56. Of these malignancies, most are of haematological or epithelial origin56 and include lymphomas and squamous cell carcinomas41,51,57. This suggests that malignancy in DOCK8-deficient patients often results from improper control of viral infections including HPV infection of the skin and EBV infection of B cells, both of which are oncogenic viruses.

Autoimmunity is rare in DOCK8-deficient patients however a small number of patients (7 cases) have been reported with autoimmune haemolytic anaemia51,56.

Non immunological Impaired growth or failure to thrive has been reported in 50-60% of DOCK8-deficient patients but this is likely a secondary effect of the recurrent infections rather than a direct effect of DOCK8 deficiency 51,56,58. DOCK8 deficiency has also been associated with non-infectious neurological manifestations. Indeed, neurological symptoms were identified in ~20-30% of patients in 1 study58 however the largest cohort study of 136 patients reported cerebral events in only 10% of patients56. One of the most commonly reported of these events is central nervous system vasculitis40,56,60 which led to stroke in a patient61. Seizures58, progressive multifocal leukoencephalopathy40 and CNS lymphoma51,62 have also been described. A small number of cases (n=3) of patients with mental retardation or autism spectrum disorder that were found to have disruptions in the chromosomal region in which DOCK8 resides have been reported but they did not display 8

many of the dysmorphic features usually present in patients with these conditions. However, the mutations were not characterised and no immune symptoms were reported in these patients63,64.

1.3.3 Immunological features of DOCK8 deficiency Numerous defects in immune cells have been defined as a result of DOCK8 deficiency. These differentiation and functional deficits likely contribute to distinct aspects of disease pathophysiology.

T cells CD8+ T cells from DOCK8-deficient patients display an increased frequency of terminally differentiated TEMRA cells at the expense of naïve and memory cells.

Furthermore, both memory and TEMRA cells exhibit a phenotype suggestive of chronic activation65. DOCK8-deficient CD8+ T cells also undergo reduced proliferation in vitro40,41,65.

CD4+ T cell frequency is reduced in DOCK8-deficient patients65 and proliferation of these cells is diminished compared to CD4+ T cells from healthy donors;40,41,65. DOCK8- deficient CD4+ T cells also exhibit impaired Th17 differentiation in vivo and in vitro and some patients have been reported to produce increased levels of Th2 cytokines54,66. There is also a reduction in regulatory T cells (Treg) in DOCK8-deficient patients41,67. Furthermore, these cells have impaired ability of suppressing proliferation of effector cells in vitro67.

B cells Previous reports found that the frequency of total B cells is increased in DOCK8-deficient patients but memory B cells are decreased59. More activated mature naïve B cells were present in DOCK8-deficient patients as were CD21lo B cells which are enriched in autoimmune patients67. Central tolerance was intact in DOCK8-deficient B cells, however there was evidence of disrupted peripheral tolerance, evidenced by increased polyreactive B cells and higher usage of self-reactive immunoglobulin genes in the antibody repertoire 67.

NK cells The frequency of NK cells and their development into mature NK cells has been reported to be normal in DOCK8-deficient patients68. However, stimulation of NK cells with target cells resulted in reduced expression of a degranulation marker, indicating reduced cytotoxicity as an outcome of functional impairment69. Indeed, when NK cells from DOCK8-deficient patients were subjected to in vitro killing assays, cytolytic function was decreased and could not be rescued by exogenous IL2, a strong activator of NK cells68. This result was replicated in NK cell lines rendered deficient of DOCK8 via siRNA as well as CRISPR/Cas9 technology70,71.

NKT cells NKT cells, which are important in recognising microbial lipids and producing cytokines to prevent infections, are reduced in DOCK8-deficient patients72.

Dendritic cells It has been reported in several DOCK8-deficient patients that there is a reduction in the frequency of plasmacytoid dendritic cells (pDC) in peripheral blood 73,74. These pDCs are critical for production of type I interferon which is important for host defence against viral infections75.

1.3.4 Treatment Due to the susceptibility of DOCK8-deficient patients to a wide range of infections, they are commonly treated with prophylactic and therapeutic antibiotics, antivirals and antifungals. A report on 136 patients found 68% were treated with antibiotic prophylaxis which was deemed effective 64% of the time while 21% of patients received antiviral treatment which had an efficacy of 69%. Additionally, 10% of patients had antifungal prophylaxis which was effective in 70% of cases56. Interferon α therapy has also been found to be useful in treating severe HSV and recurrent warts which do not respond to standard treatments of salicylic acid and cryotherapy56,73,74,76. Treatment for allergic disease and asthma is usually with inhaled corticosteroids and antihistamines76. As skin infections, particularly Staphylococcus aureus, can affect eczema severity, bleach baths and antiseptics are recommended to control bacterial spread while corticosteroids can be used for eczema but run the risk of worsening viral infections77. Patients diagnosed with DOCK8 deficiency also undergo screening for potential complications including 10

assessment of pulmonary and liver function, imaging for cancers and measuring blood PCR levels of viruses (HSV, CMV, EBV)78. As DOCK8-deficient patients have compromised antibody responses, immunoglobulin replacement is used to combat sinopulmonary infections with 64% of reported patients receiving this treatment with 57% considered to be effectively treated56. However, this treatment does not improve viral infections76,77. Given the frequent and severe symptoms associated with DOCK8 deficiency, haematopoietic stem cell transplant (HSCT) has become the standard of care and is recommended at an early age56. A review of a large cohort of transplanted patients concluded that HSCT resulted in a positive outcome in the majority of cases (with 84% survival), regardless of the type of transplant (matched related donor, matched unrelated donor, mismatched donor). Improvement of symptoms was variable, with eczema improving in 99% of patients and upper airway infections and mollusca also responding well (93-94% patients showed improvement) whereas food allergies and pulmonary function did not improve as substantially79.

1.4 DOCK8 protein 1.4.1 Structure DOCK8 is a large protein of ~190kDa 80 belonging to the evolutionarily conserved DOCK180-related superfamily within the DOCK-C subfamily. This superfamily is characterised by the presence of the functional DOCK Homology Region (DHR)1 and DHR2 domains. The DHR2 domain is responsible for the guanine exchange factor (GEF) activity of DOCK8 which functions by dissociating bound GDP from its target enzyme allowing GTP to bind to the enzyme and render it active81. The DHR1 domain of DOCK8 is involved in phospholipid binding and membrane translocation and is required for cell elongation and migration82,83.

1.4.2 Expression When first described, DOCK8 expression was determined by northern blot. This revealed broad and prominent expression in many tissues, including placenta, lung, kidney and pancreas, and lower but detectable expression in the heart, brain and skeletal muscle80. These findings were subsequently confirmed by another group who also observed DOCK8 expression in the spleen and thymus64. The human protein atlas details 11

ubiquitous expression of DOCK8 across major organ groups as well as the skin, bone marrow, the immune system and endocrine tissues. RNAseq data reveals the same expression pattern but RNA expression is found to be highest in immune organs including spleen, tonsil and lymph node (www.proteinatlas.org). During investigation of immunodeficiency patients with bi-allelic inactivating mutations in DOCK8 it was revealed that DOCK8 protein was present in lysates of peripheral blood mononuclear cells (PBMC) from healthy donors 40 including both T cells and B cells 41. Expression of DOCK8 in T cells and B cells was amongst the highest reported from data encompassing numerous cell types from microarrays (biogps.org). Thus, while DOCK8 is a ubiquitously expressed protein, it is enriched in haematopoietic cells.

1.5 DOCK8 function Insight into DOCK8 function was first obtained when it was initially identified by a yeast two hybrid screen as a CDC42-interacting protein which upon cellular stimulation localised to lamellipodia along with filamentous actin, suggesting a role in actin reorganisation80. This DOCK8-CDC42 interaction was confirmed via pull down experiments in human cell lines which also highlighted the specificity of DOCK8 for CDC42 rather than other Rho GTPases84. CDC42, like other RhoGTPases, is an intracellular molecular switch which regulates numerous signalling pathways associated with organisation of the actin cytoskeleton resulting in involvement in processes such as cytokinesis, morphogenesis and cell migration85. Since its discovery, studies of various immune cells, in which DOCK8 is most highly expressed, from DOCK8-deficient humans and Dock8 mutant mice have revealed a key role for DOCK8 in two distinct processes: immune synapse (IS) formation and cell migration.

1.5.1 Immune synapse formation The IS is a specialised structure that forms between the membranes of two interacting cells and allows signalling or cytotoxic function through reorganisation of the cytoskeleton.

DOCK8 was first implicated in IS formation in B cells. These findings arose from a screen used to identify ENU mutagenesis-induced mutant mice that were unable to produce long-

lived, high-affinity antibody responses in vivo in response to T-dependent antigens. Addition of B cells to a lipid bilayer membrane coated with specific antigen and the adhesion molecule ICAM-1 results in B cell receptor (BCR)-mediated activation of the integrin LFA-1, which then binds to ICAM-1. This represents a model to investigate the dynamics of interactions between B cells and an antigen-presenting cell (APC). B cells from Dock8cpm/cpm mutant mice (containing a missense mutation leading to abnormal splicing products) demonstrated impaired LFA-1-mediated clustering of ICAM-1 into peripheral supramolecular clusters (pSMAC), an important component of the IS which surrounds the central cluster of antigen-containing BCRs and maintains IS stability86.

The observation that DOCK8 deficiency affected CD8+ T cell activation by delaying cellular division led to examination of IS formation based on the recruitment of IS components to the cell interface. When antigen-presenting DCs were co-cultured with antigen-specific naïve CD8+ T cells from Dock8pri/pri mice (containing a homozygous mutation in the DHR2 domain) the T cells displayed decreased cell interface localisation of actin, a cytoskeletal filament, and LFA-1which binds to ICAM-1 on DCs, to the immune synapse65.

Experiments focusing on CD4+ Tregs found that upon TCR stimulation, Tregs from Dock8-/- mice (homozygous mutation introducing stop codon in exon 9) did not upregulate F-actin (filamentous actin) content to the same extent as WT Tregs. Furthermore, fewer Dock8-/- Tregs adhered to a coverslip coated with anti-CD3 and ICAM-1 to mimic IS formation and interaction with an antigen presenting cell (APC). Furthermore, Dock8-/- Tregs that did achieve IS formation had decreased F-actin content. Additionally, Dock8 colocalised with F-actin to the interface in WT but not Dock8-/- Tregs. Subsequently, IS formation was more transient and Dock8-/- Tregs were less able to remove costimulatory molecules from the surface of cells via transendocytosis which is vital in rendering DCs tolerogenic87.

IS formation was also investigated in NK cells from DOCK8-deficient patients. When stimulated with target cells, DOCK8-deficient NK cells exhibited reduced accumulation of actin as well as β2 integrin and perforin, important components of the lytic process by which NK cells achieve cytotoxicity, at the IS. This result was replicated in NK cells with shRNA-mediated knockdown of DOCK8 68. Further experiments with siRNA-mediated DOCK8-suppressed NK cell lines also showed reduced polarisation of LFA-1 to the IS 13

and subsequent reduction in pSMAC formation70. Accumulation of lytic granules at the lytic pole of the cell was also impaired in NK cells lacking DOCK8 expression. Mass spectrometry of immunoprecipitated DOCK8 complexes from NK cell homogenates identified WASP and Talin as interaction partners of DOCK8, both of which had previously been shown to localise to the IS in NK and T cells. Localisation of WASP and Talin was compromised in DOCK8- suppressed-NK cells indicating DOCK8 is responsible for their polarisation to the IS70.

Overall, work investigating the IS in different lymphocytes suggests that the function of DOCK8 is twofold. First, it enables integrins to bind to and cluster adhesion molecules on the surface of the interacting cell such as the binding of LFA-1 to ICAM-1, leading to formation of pSMACs which encircle the central supramolecular clusters (cSMACs) composed of cell receptors bound to their target. Second, DOCK8 organises actin and actin regulators such as WASP and Talin to the IS to stabilise and maintain the IS interaction.

1.5.2 Migration The role of DOCK8 in cell migration was first identified in DCs in which it was already established that CDC42 was required for migration. Dock8-/- (null) mice showed significantly decreased trafficking of skin DCs to draining lymph nodes. This impaired migration occurred only when cells were moving through a 3D environment, such as the dermis of the skin which is populated by fibrillar arrays of collagen bundles. In vivo imaging revealed that Dock8-/- DCs struggled to adequately undergo the requisite morphological changes necessary to enable migration through narrow and constricted spaces. Further investigation uncovered that while in normal DCs DOCK8 and activated CDC42 was mostly present at the leading-edge membrane, this localisation of CDC42 was perturbed in Dock8-/- DCs, indicating Dock8 is required for CDC42 localisation84.

The inability of cells lacking DOCK8 to modulate their shape was investigated further in T cells. DOCK8-deficient T cells isolated displayed impaired migration into the dermal layer of human foreskin biopsies as well as into a 3D collagen gel matrix. The DOCK8- deficient T cells spent significantly more time elongated and this was replicated in cells with siRNA-mediated knockdown of DOCK8 as well as in T and NK cells from Dock8cpm/cpm mice. It was found that this prolonged elongation time resulted in 14

catastrophic fragmentation termed “cytothripsis” but only in a 3D environment. This was also recapitulated in an in vivo mouse model using epicutaneous HSV infection of wild type (WT) mice and transfer of either WT or Dock8-deficient T cells with analysis by intravital two-photon microscopy. Knockdown of CDC42 in T cells from healthy donors as resulted in cytothripsis, suggesting this defect was due to uncoordinated movement between the front and rear of cells due to impaired cytoskeletal organisation including decreased F-actin polymerisation at the poles. This work also provided an explanation for the increased incidence of viral skin infections in DOCK8-deficient patients as their lymphocytes cannot navigate the dense environment of the skin to reach the infection88.

The involvement of DOCK8 in migration of T cells from blood vessels into tissues via transendothelial migration has also been investigated. In vitro-generated Th1 cells from Dock8-/- and Dock8pri/pri mice were less able to transmigrate across an endothelial cell monolayer under low-flow conditions compared to wild type cells. This reduced transmigration ability of Dock8-deficient T cells was shown to affect homing to the lymph node which also involves migration across endothelial cells89.

In a search to identify gene alterations that could limit the migration of CD4+ T cells into the central nervous system (CNS), transfer of Dock8pri/pri encephalitogenic CD4+ T cells resulted in less CNS infiltration compared to transfer of Dock8pri/+ cells. This resulted from fewer cells translocating DOCK8 to the leading edge and CD44 to the hind part of the cell90. Immunoprecipitation of DOCK8 from a T cell line identified LRCH1, a cytoskeleton regulator in Drosophilia, as a DOCK8 binding partner. Further investigation of Lrch1 transgenic mice revealed that upon chemokine stimulation Dock8 is phosphorylated by PKCα which disrupts binding with LRCH1 allowing Dock8 to localise to the leading edge of the cell membrane and activate CDC42 leading to cytoskeleton rearrangement and migration90.

The effect of DOCK8 on macrophage migration has also been tested in a mouse model of lung inflammation. This revealed that accumulation of Dock8-deficient macrophages in the lung was impaired compared to in Dock8+/- mice. Bone marrow-derived macrophages from Dock8-/- mice displayed proper directionality in response to chemokine stimulation in vitro but a reduced average migration speed in a 2D environment. Impaired migration of Dock8-/- macrophages could be rescued by ectopic expression of WT Dock8, but not by a mutant form of Dock8 which abolished CDC42

GEF activity. This established that interactions between Dock8 and CDC42 are important for macrophage migration . Myosin, which regulates the motion of the leading edge during cell migration, did not effectively localise to the leading edge in Dock8-/- macrophages due to impaired phosphorylation of its regulatory chain (MLC2). MLC2 phosphorylation is achieved by the CDC42-effector kinase MRCK in association with the adaptor protein LRAP35a which was found to be a binding partner of Dock8. Knockdown of LRAP35a expression led to decreased MLC2 phosphorylation and reduced the migration speed of macrophages to that seen in Dock8-/- macrophages leading to the conclusion that Dock8 enables a link between CDC42 activation and myosin dynamics through interaction with LRP35a. LRP35a knockdown in bone marrow-derived DCs also resulted in decreased migration in a 3D environment, suggesting the DOCK8-LRAP35a interaction is not restricted to macrophages91.

Hence, DOCK8 appears to be involved in several processes of migration of different lymphocytes including through the skin, to the lymph node, into tissues, into the central nervous system and to the lung.

1.6 Aims Despite substantial investigation using both DOCK8-deficient humans and mouse models, the impacts of DOCK8 deficiency on the immune system remain to be fully elucidated. This thesis utilises samples from DOCK8-deficient patients to gain a detailed understanding of the effects of pathogenic recessive mutations in DOCK8 on immune phenotype and function and how these effects underly the clinical features seen in patients. Specifically, this thesis sought to achieve this aim by:

1. Examination of CD4+ T cell function in DOCK8-deficient patients 2. Analysis of lymphocyte phenotype and function in DOCK8-deficient patients before and after HSCT 3. Interrogation of the impact of DOCK8 reversion 4. Study of patients with mutations in DOCK8-related proteins

Chapter 2: Materials and Methods

2.1 Cell isolation 2.1.1 Materials - Phosphate Buffered Saline (PBS): Life technologies, Invitrogen

- Fetal Calf Serum (FCS): Life technologies, Invitrogen. Heat inactivated at 56oC for 45mins.

- Freezing media: 50% media with 40% FCS and 10% dimethyl sulphoxide (DMSO, AJAX, ThermoFisher Scientific). Filter sterilised.

- Red Blood Cell (RBC) lysis buffer: Triple distilled water with 10mM Tris-HCl (Sigma Aldrich) and 0.14M ammonium chloride (BDH, Merck), and of pH7.4. Filter sterilised.

- Ficoll-Plaque: GE Healthcare

- Trypan blue: Sigma Aldrich

- DNase: Sigma Aldrich

- Magnesium chloride (MgCl2): Sigma Aldrich

2.1.2 Isolation of Peripheral Blood Mononuclear Cells Blood was diluted 1:1 with PBS and slowly overlayed onto Ficoll-Plaque (max 35ml of blood onto 12.5ml of Ficoll) before being centrifuged at 1800rpm for 20min at room temperature, deceleration without brake. Serum was then taken from the top-most layer and stored at -20oC. Mononuclear cells were extracted from the layer above the Ficoll and washed with PBS before being incubated for 10min at 37oC with 10ml of RBC lysis buffer to lyse the red blood cells. After another PBS wash, cells were counted with 0.1% trypan blue to distinguish live cells and either used or frozen in freezing media to be stored in liquid nitrogen.

2.1.3 Isolation of lymphocyte subsets by fluorescence activated cell sorting (FACS) Cells were thawed in PBS-2% FCS and counted with 0.1% trypan blue to distinguish live cells. Monoclonal antibodies at the appropriate dilution was added at a total volume of 10ul/1x106 cells and cells were incubated for 30min on ice with no light. Cells were washed with PBS-2% FCS before being resuspended in PBS-2% FCS at a concentration 6 of 20-25x10 cells/ml (minimum volume 1mL) with DNase (20µg/ml) and MgCl2 (5mM) to prevent clumping. Cells were filtered using a 70µm nylon cell strainer and run on a FACS Aria III (BD Biosciences) flow cytometer and sorted into 5mL polypropylene tubes containing FCS at 4oC. Stained healthy control cells were used as compensation controls and sort purity of populations was usually >98%.

To isolate naïve and memory CD4+ T cells; naïve and effector memory CD8+ T cells; NK cells and naïve and memory B cells PBMCs were stained with fluorochrome-labelled mAbs against CD3, CD4, CD8, CD20, CD56, CD25, CD127, CD45RA, CCR7, CD27, CD10 and IgG (Table 1). After gating for lymphocytes based on FSC-A vs SSC-A and excluding doublets based on FSC-A vs FSC-H, CD4+ T cells were identified by expression of CD4 before exclusion of T regs (CD25hiCD127±) and naïve (CD45RA+CCR7+) and memory (CD45RA-) populations collected. CD8+ T cells were identified by expression of CD8 before collection of naïve (CD45RA+CCR7+) and effector memory (CD45RA-CCR7-) populations. B cells were identified by expression of CD20+ and the naïve (CD27-CD10-IgG-) and memory (CD27+) populations collected. NK cells were identified by exclusion of CD3+ cells and expression of CD56.

2.2 Cell culture 2.2.1 Materials - 5-(and-6)-carboxyfluorescein diacetate succinimidyl ester (CFSE): Molecular Probes Inc. Reconstituted in DMSO at 5mM.

- B and T cell media: RPM11640 (Life Technologies, Invitrogen) supplemented with 10% FCS, 10mM HEPES (Life Technologies, Invitrogen), 0.1mM non-essential amino acid solution (Life Technologies, Invitrogen), 1mM sodium pyruvate (Life Technologies, Invitrogen), 60µg/ml penicillin (Life Technologies, Invitrogen), 100µg/ml streptomycin (Life Technologies, Invitrogen), 40µg/ml bovine apo- transferrin (Life Technologies, Invitrogen) and 20µg/ml NormocinTM (InvivoGen).

- T cell activation and expansion beads (TAE; anti-CD2, -CD3, -CD28 mAbs conjugated to beads): Miltenyi Biotech

- Human recombinant CD40L/TNFSF5: R&D Systems. Crosslinked with hemagglutinin peptide (R&D Systems)

- CpG: CpG2006 (Sigma Aldrich)

- PANSORBIN Cells (heat-killed, formalin-fixed Staphylococcus aureus cells - Cowan I strain) (SAC): Millipore.

- Human recombinant IL2: Millipore

- Human recombinant IL4: DNAX

- Human recombinant IL12: R&D Systems

- Human recombinant IL1β: Peprotech

- Human recombinant IL6: Peprotech

- Human recombinant IL21: Peprotech

- Human recombinant IL23: eBioscience

- Human recombinant TGFβ1: Peprotech

- Prostaglandin E2: Sigma Aldrich

- Phorbol 12-myristate 13-acetate (PMA): Sigma Aldrich

- Ionomycin: Sigma Aldrich

- Brefeldin A: Sigma Aldrich

- Bovine Serum Albumin (BSA): Sigma Aldrich

2.2.2 T cell cultures Sorted naïve and memory CD4+ T cells were cultured at 4x105 cells/ml in human media in either round or flat bottom tissue culture plates (depending on cell number) with TAE beads attached to anti-CD2, -CD3, and -CD28 mAbs at a cell:bead ratio of 2:1 for 5 days at 37oC. Th1 cultures contained 50ng/ml of IL12, Th2 cultures contained 1unit/ml of IL4 and Th17 cultures contained 50ng/ml of IL1β, 50ng/ml of IL6, 50ng/ml of IL21, 50ng/ml of IL23, 2.5ng/ml of TGFβ and 50ng/ml of prostaglandin E2.

CD8+ T cells were cultured as above with TAE beads and with the addition of 50units/ml of IL2 to some cultures.

To examine CD4+ T cell intracellular cytokine expression, cells were restimulated on day 5 for 6 hours with 100ng/ml of PMA and 750ng/mL of ionomycin in a volume of 200µl. After 2 hours, 10µg/ml of Brefeldin A was added to prevent cytokine export via the golgi apparatus and endoplasmic reticulum.

To examine CD8+ T cell intracellular cytokine expression, cells were restimulated as above. After 1 hour 10µg/ml of Brefeldin A and 2µM monensin were added per well.

2.2.3 B cell cultures Sorted naïve B cells were cultured at 4x105 cells/mL in human media in round bottom tissue culture plates with various combinations of 200ng/ml of CD40L, 50ng/ml of IL21, 1µg/ml of CpG2006 and 0.05% SAC for 5 days at 37oC.

2.2.4 CFSE To determine cell division history, cells were resuspended in PBS-0.1% BSA at 107 cells/ml with 10µM CFSE. They were incubated for 10min at 37oC with occasional mixing before being washed with 5x volume of ice cold PBS-0.1% BSA. Cells were counted with trypan blue before being cultured.

2.3 Flow Cytometry 2.3.1 Materials - FACS buffer: PBS with 0.1% BSA and 0.1% sodium azide (Sigma Aldrich)

- 1% formaldehyde (FA): Sigma Aldrich. Made in PBS

- 0.1% Saponin: Sigma Aldrich. Made in PBS

- TCR Vβ repertoire kit: Beckman Coulter

- Indo-1AM: Molecular Probes. Reconstituted in DMSO at 1mg/ml

- Zombie Aqua cell viability dye: Biolegend

- CaliBRITE beads: BD Biosciences

- Cytometric Bead Array (CBA): BD Biosciences

2.3.2 Cell surface staining Cells were washed in FACS buffer before being incubated with fluorochrome-labelled antibodies diluted to an appropriate concentration (Table 2.1) in FACS buffer for 30min on ice without light. If a second staining step was required, cells were washed and the secondary antibodies were added for another 30min incubation after a FACS buffer wash. Cells were then washed with FACS buffer to remove unbound antibodies and fixed with 2% formaldehyde for 20min at room temperature. Cells were washed with FACS buffer and then resuspended in FACS buffer and kept at 4oC until being run on the LSRSORP or Fortessa flow cytometers (BD Biosciences).

To assess Vβ repertoire, the TCR Vβ repertoire kit was used. Cells were split 8 ways and each was surface stained with a provided antibody cocktail diluted 1/20 in FACS buffer for 30min on ice without light. Cells were then washed with FACS buffer to remove unbound antibodies, fixed with 2% formaldehyde for 20min at room temperature, washed again with FACS buffer and then resuspended in FACS buffer and kept at 4oC until being run on the LSRSORP or Fortessa flow cytometers.

Table 2.1: monoclonal Abs used for surface staining Marker Fluorochrome Specificity Clone Supplier Dilution CCR7 FITC Human 150503 R&D Systems 1:10 CCR7 PECY7 Human G043H7 BioLegend 1:10 CD3 Biotin Human OKT3 eBioscience 2.5µg/ml CD3 BV421 Human UCHT1 Becton Dickson Biosciences 1:100 CD3 BV786 Human UCHT1 Becton Dickson Biosciences 1:50 CD4 APCCY7 Human RPA-T4 Becton Dickson Biosciences 1:100 CD4 BUV395 Human SK3 Becton Dickson Biosciences 1:100 CD4 BUV737 Human SK3 Becton Dickson Biosciences 1:150 CD4 Pacific Blue Human OKT4 eBioscience 1:50 CD8 APC Human 2ST8.5H7 Becton Dickson Biosciences 1:100 CD8 BUV395 Human RPA-T8 Becton Dickson Biosciences 1:300 CD8 PECY7 Human RPA-T8 Becton Dickson Biosciences 1:200 CD8 PerCPCy5.5 Human RPA-T8 Becton Dickson Biosciences 1:80 CD10 APC Human HI10a Becton Dickson Biosciences 1:10 CD10 BUV737 Human HI10a Becton Dickson Biosciences 1:200 CD20 BUV395 Human 2H7 Becton Dickson Biosciences 1:50 CD20 PE Human L27 Becton Dickson Biosciences 1:15 CD25 FITC Human 2A3 Becton Dickson Biosciences 1:50 CD25 PE Human M-A251 Becton Dickson Biosciences 1:10 CD25 PECY7 Human M-A251 Becton Dickson Biosciences 1:80 CD27 BV786 Human L128 Becton Dickson Biosciences 1:50 CD27 PECY7 Human M-T271 Becton Dickson Biosciences 1:50 CD28 PerCPCy5.5 Human CD28.2 Becton Dickson Biosciences 1:50 CD45RA BV605 Human HI100 Becton Dickson Biosciences 1:50 CD45RA PerCPCy5.5 Human HI100 eBioscience 1:100 CD56 BV605 Human HDC56 BioLegend 1:100 CD57 FITC Human NK-1 Becton Dickson Biosciences 1:50

CD69 BV711 Human FN50 Becton Dickson Biosciences 1:40 CD95 PECF594 Human DX2 Becton Dickson Biosciences 1:50 CD107a AF647 Human H4A3 BioLegend 3:200 CD127 BV421 Human A019D5 BioLegend 1:30 CD127 BV650 Human A019D5 BioLegend 1:50 CD161 PerCPCy5.5 Human HP-3G10 eBioscience 1:30 ICOS APC Human C398.4A eBioscience 1:20 IgA+IgG+Ig unconjugated Human F(ab')2 Jackson M (H+L) ImmunoResearch 15µg/ml IgD PE Human IA6-2 Becton Dickson Biosciences 1:100 IgG APC Human G18-145 Becton Dickson Biosciences 1:10 IgG unconjugated Human EPR25A Abcam 1:50 IgG 1 PE Mouse G18-145 Becton Dickson Biosciences 1:250 IgGI k unconjugated Mouse P3.6.2.8.1 eBioscience 1:200 IgG (H+L) a647 Rabbit Polyclonal Abcam 1:200 IgM FITC Human G20-127 Becton Dickson Biosciences 1:10 PD1 Biotin Human eBioJ105 eBioscience 1:30 Streptavidin BV421 Human N/A BioLegend 1:100 TCRαβ PECY7 Human IP26 BioLegend 1:40 TCRγδ PerCPe710 Human B1.1 eBioscience 1:40 TCR Va24 FITC Human C15 Beckman Coulter 1:10 TCR Va7.2 BV421 Human 3C10 BioLegend 1:20 TCR Vb11 PE Human C21 Beckman Coulter 1:10

2.3.3 Intracellular staining To determine expression of intracellular proteins, cells were first washed with PBS. For cultured cells, incubation with Zombie Aqua dye diluted 1/500 in PBS for 10min at room temperature without light followed by washing with PBS was done. Cells were then fixed with 2% formaldehyde for 20min at room temperature. Cells were washed with 0.1% saponin three times and then incubated with fluorochrome-labelled antibodies (Table 2.2) diluted in 0.1% saponin for 30min on ice without light. If a second staining step was required, the cells were washed twice with 0.1% saponin to remove unbound mAbs and the secondary antibodies were added for another 30min incubation on ice without light. Following another two washes with 0.1% saponin, cells were washed with FACS buffer

and then resuspended in FACS buffer and kept at 4oC until being run on the LSRSORP or Fortessa flow cytometers.

Table 2.2 monoclonal Abs used for intracellular staining Marker Fluorochrome Specificity Clone Supplier Dilution DOCK8 unconjugated Human G-2 Santa Cruz Biotechnology 1:50 DOCK8 unconjugated Human EPR1251 Abcam 1 1:50 Granzyme B AF700 Human GB11 Becton Dickson Biosciences 1:50 IFNγ BV605 Human B27 Becton Dickson Biosciences 1:40 IL2 BV650 Human MQ1- Becton Dickson 17H12 Biosciences 1:50 IL2 BV711 Human 5344.111 Becton Dickson Biosciences 1:50 IL4 PECY7 Human 8D4-8 eBioscience 1:100 IL13 BV421 Human JES10- Becton Dickson 5A2 Biosciences 1:20 IL17A APCCY7 Human BL168 BioLegend 1:50 IL17F BV786 Human O33-782 Becton Dickson Biosciences 1:30 IL21 eflour660 Human eBio3A3- eBioscience N2 1:20 IL22 PE Human 22URTI eBioscience 1:80 Perforin PECY7 Human B-D48 BioLegend 1:25 TNFα BUV395 Human MAb11 Becton Dickson Biosciences 1:40

2.3.4 Cytometric Bead Array (CBA) Cytokines in culture supernatants were measured by cytometric bead array as per manufacturer’s instructions. Briefly, 25µl of supernatant was added to 25µl capture bead mix (0.25µl of each cytokine capture bead x number of samples made in capture diluent to total volume of 25µl x number of samples), mixed and incubated for 1h at room temperature in the dark. 25µl detection bead mix (0.25µl of each cytokine detection bead x number of samples made in detection diluent to total volume of 25µl x number of samples) was added, mixed and incubated for 2h at room temperature in the dark. Samples were then washed with wash buffer and resuspended in wash buffer and kept at 4oC until being run on the Fortessa flow cytometer. Individual cytokines were gated by their bead position on a plot of APC vs APC-Cy7 (Figure 2.1; Table 2.2). Standard samples (Table

2.3) were run to generate a standard curve which was used to determine sample concentrations using BD FCAP Array software.

Table 2.3: CBA cytokines Cytokines measured Bead position IL2 A4 IL4 A5 IL5 A6 IL6 A7 IL9 B6 IL10 B7 IL13 E6 IL17A B5 IL17F C6 IL21 B8 IFNγ E7 TNFα C4 Granzyme A D9 Granzyme B D7

Table 2.4: CBA standards Standards used (pg/ml) 2500 1250 625 312.5 156.25 78.13 39.06 19.53 9.77 0

Figure 2.1: CBA cytokine bead position. Example of APC vs APCCy7 plot used to identify cytokines by their bead position.

2.3.5 Cell quantification Calibrite beads were gated based on their high SSC-A and low FSC-A (Figure 2.1) and event number recorded. Live cells were identified by their lack of zombie dye staining and event number recorded. Cell quantification was determined by the following calculation: (live cells count / beads count) x number of beads added

Figure 2.1: Calibrite beads position Example of plot of SSC-A vs FSC-A used to gate Calibrite beads.

2.3.6 Calcium flux Cells were suspended in 1ml media and 1µl of 1mg/ml indo1-AM added for 30min at 37oC. Cells were washed in media and resuspended in media with diluted fluorochrome- labelled antibodies (Table 2.1) for 30min incubation on ice without light. Cells were again washed in media and resuspended in media at the appropriate volume (100µl x number of conditions to be tested) and kept on ice. For analysis, 100µl of cells were added to 300µl media for 5min at 37oC then acquired on the LSR SORP or Fortessa flow cytometer for 30 sec before addition of CD3-biotin. After 1min, streptavidin was added at a concentration of 20µg/ml and acquisition continued for another 4min before addition of 4µl of ionomycin (100mg/ml) and 90sec further acquisition. Flux was calculated based on the frequency of responding cells determined by an increase in the ratio of bound (emission UV395)/unbound calcium (emission UV525).

2.3.7 Analysis of flow cytometry data Analysis of .fcs files was performed using FlowJo 10 software (Tree Star). Mean fluorescence intensity (MFI) was determined by the geometric mean of fluorescence parameters.

2.4 Immunoglobulin detection 2.4.1 Materials - Coat buffer: 0.05M sodium carbonate (Sigma Aldrich) and 0.18M sodium bicarbonate (Sigma Aldrich) in triple distilled water at pH 9.1

- Blocking solution: 2% FCS in PBS

- Wash buffer: 0.04% Tween-20 (Sigma Aldrich) in PBS

- Secondary buffer: 0.04% Tween-20 and 1% BSA in PBS

- Avidin-HRP conjugate: 1mg/ml Peroxidase conjugated streptavidin (Jackson ImmunoResearch)

- ABTS substrate: 0.1% 2,2’-azinobis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (Sigma Aldrich), 0.03% hydrogen peroxide (AJAX), 0.1M citric acid (Sigma Aldrich) and 0.05M tri-sodium citrate in triple distilled water at pH 4.5

- ABTS stop solution: 0.2M citric acid in triple distilled water

2.4.2 Enzyme linked immunosorbent assay (ELISA) Plates were coated in antibody diluted in coating buffer and incubated overnight at 4oC. Coating buffer was removed before blocking solution was added and incubated for 1hr at 37oC. Plates were washed 3 times with wash buffer before standards, blanks and samples were added and incubated for 1-2hr at 37oC. Plates were washed 3 times with wash buffer and detection antibody diluted in secondary buffer was added and incubated for 1-2hr at 37oC. Plates were washed 3 times with wash buffer and Avidin-HRP conjugate diluted in secondary buffer at 1/15000 was added for 1hr at 37oC. Plates were washed with wash buffer 5 times and ABTS substrate added and incubated at room temperature without light until colour developed. Plates were measured on a reader (Biotek Instruments) at OD 405nm and when the top standard reached 1, stop solution was added and values recorded.

Table 2.5 Human ELISA Abs used

Antibody Use Host Specificity Dilution Supplier IgA standard 1:2 Sigma Aldrich purified goat α chain Southern IgA coat F(ab')2 specific 1:2000 biotech biotin-goat α chain Southern IgA detection F(ab')2 specific 1:2000 biotech IgM standard 1:2 Sigma Aldrich purified goat µ chain Southern IgM coat F(ab')2 specific 1:2000 biotech biotin-goat µ chain Southern IgM detection F(ab')2 specific 1:2000 biotech IgG standard 1:2 Sigma Aldrich purified goat γ chain Southern IgG coat F(ab')2 specific 1:1000 biotech biotin-goat γ chain Southern IgG detection F(ab')2 specific 1:1000 biotech

2.4.3 ImmunoCAP assay Plasma from healthy donors and patients was analysed for total and allergen-specific IgE (Sydney South West Pathology Service, Australia) using the Phadia250 ImmunoCAP platform (Thermo Scientific). IgE levels specific for a staple food mix (FX5; egg white, milk, codfish, wheat, peanut, and soybean) or mite mix (HP1; Hollister-Stier Labs, Dermatophagoides pteronyssinus, Dermatophagoides farina, German Cockroach) were determined.

2.5 Molecular studies 2.5.1 Materials - RNeasy mini prep kit: QIAGEN

- QIAamp DSP DNA Blood mini kit: QIAGEN

- QIAquick PCR purification kit: QIAGEN

- TOPO TA cloning kit for sequencing: Life technologies

- Random hexamers: Invitrogen

- PCR nucleotide mix (10mM dNTPs): Roche Diagnostics

- 5X First strand buffer: Invitrogen

- Dithiothreitol (DTT): Invitrogen

- RNaseOUT Recombinant Ribonuclease Inhibitor: Invitrogen

- SuperScript III Reverse transcriptase: Invitrogen

- Lightcycler 480 Probes master mix: Roche Diagnostics

- GoTaq reaction buffer: Promega

- 25mM MgCl2: Promega

- GoTaq DNA polymerase: Promega

- TBE buffer: Sigma Aldrich

- SyberSafe DNA gel stain: Invitrogen

- DNA molecular weight marker XIV (100-1500 base pairs): Roche Diagnostics

- MinElute gel extraction kit: QIAGEN

- One Shot TOP10 competent cells: Life technologies

- SOC medium: 2% tryptone, 0.5% yeast extract, 10mM NaCl, 2.5mM KCl, 10mM

MgCl2, 10mM MgSO4 and 20mM glucose (Invitrogen)

- NP40 lysis buffer: 1% NP40, 10mM Tris-HCl, 150mM NaCl, 0.1% NaN3 at pH 7.8

- Protease inhibitor cocktail: Aprotinin, Na3VO4, NaF, PMSF, EDTA

- NuPAGE 4-12% Tris-Bis gel: Life technologies

- SeeBlue Plus2 prestained standard: Invitrogen

- Immobilon-FL membrane: Millipore

- Odyssey blocking buffer: Licor Biosciences

- Β-mercaptoethanol (>99%): Sigma Aldrich

- Ethanol (100%): Sigma Aldrich

2.5.2 RNA extraction RNA was harvested using the RNeasy mini prep kit following the manufacturer’s instructions with centrifugation at 8000g. Briefly, cells were harvested and pellets resuspended in RLT buffer containing β-mercaptoethanol (1:100) before being added to a shredder spin column and spun for 2 min. The resulting lysate was diluted 1:1 with 70% ethanol and mixed before being added to a RNeasy column and spun for 15sec. A spin with RW1 buffer for 15sec was performed before the addition of RPE buffer and a 15sec spin which was repeated but with a 2min spin to dry the membrane. Elution was achieved with the addition of 30µl RNase free water and a 1min spin. Eluted RNA was stored at -80oC.

2.5.3 cDNA synthesis To produce cDNA, 1µl RNA was added to 0.5µl random hexamers, 1µl dNTPs and 11µl water and placed at 65oC for 5min and then on ice for 1min. 4µl 5X First strand buffer, 1µl DTT, 1µl RNAseOUT and 0.5µl SuperScript III RT were added before mixing and incubation at 25oC for 5min. Samples were then incubated at 50oC for 60min and 70oC for 15min. cDNA was stored at -20oC.

2.5.4 Quantitative PCR (qPCR) qPCR was performed using the Roche Lightcycler 480 Probe System with relative gene expression calculated based on the house-keeping gene RPL13A (Table 2.6). 2µl cDNA was added to 5µl LightCycler master mix, 0.1µl of probe, 0.2µl each of 10µM forward

and reverse primer and 2.5µl water. Samples underwent denaturation at 95oC for 10min before 45 amplification cycles of 95oC for 10sec, 60oC for 30sec and 72oC for 5sec and cooling at 40oC for 30min.

Table 2.6: Primers used for qPCR Gene Probe Forward primer Reverse Primer Standard used TBX21 34 GTCCAACAATGT AAAGATATGCGTGT Pooled Th1 GACCCAGA TGGAAGC cells GATA3 70 CGAGCAACGCAA AGTCCCTCCTCGGG JURKAT TCTGAC TCAC RORC 69 CAGCGCTCCAAC CCACATCTCCCACA F2/F7 ATCTTCT TGGACT RPL13A 28 CAAGCGGATGAA TGTGGGGCAGCATA RAMOS CACCAAC CCTC Probes from Roche and primers from Integrated DNA Technologies.

2.5.5 DNA extraction Genomic DNA was extracted using the QIAamp DSP DNA Blood mini kit as per manufacturer’s instructions but omitting the protease incubation step. Briefly, lysis buffer and 100% ethanol were added to pelleted cells and the resuspended lysate added to the spin column for a 1min spin at 6000g. Wash buffer 1 was added prior to a 1min spin at 6000g and then wash buffer 2 was added prior to a 1min spin at 20000g. After a 3min spin at 20000g to dry the membrane, 50µl elution buffer was added to the column and incubated at room temperature for 1min before a 1min spin at 6000g. Eluted DNA was stored at -20oC.

2.5.6 Polymerase Chain Reaction (PCR) 1µl gDNA was added to 5µl PCR buffer, 0.5µl dNTPs, 0.5µl each of 100ng/µl 5’ and 3’ primers, 1.5µl MgCl2, 15.88µl water and 0.125µl GoTaq polymerase and amplified as follows:

94oC for 5min x1 cycle

------

94oC for 30s

58oC for 30s x35 cycles

72oC for 45sec

------

72oC for 10min x1 cycle

PCR purification was achieved using QIAquick PCR purification kit according to the manufacturer’s instructions with centrifugation at 17900g. Briefly, buffer PB was added 5:1 to PCR product and added to a QIAquick column and spun for 1min before being washed with buffer PE containing 100% ethanol and spun for 1min. After another 1min spin to dry the column, 50µl buffer EB was added and the purified product was eluted after a 1min spin.

2.5.7 Agarose gel electrophoresis To assess PCR product size and concentration, agarose gel electrophoresis was performed with a 1.2% agarose gel containing 1:10000 diluted SyberSafe DNA gel stain and run in TBE buffer at 100V. Samples were mixed with loading dye prior to running and a 100 base pair molecular weight marker was also run. DNA bands were visualised on a UV transilluminator (Bio-Rad Laboratories).

2.5.8 Gene sequencing 30ng purified PCR product and 3.2pmol primer (Table2.7) were sent to Garvan Molecular Genetics (GMG, Darlinghurst) for sequencing. Results were analysed using Snapgene (GSL Biotech LLC).

Table 2.7: Primers used for PCR and sequencing Gene Forward primer Reverse primer Source DOCK8 CCAAATATTTCGGGAAAC ACAGTGCAAAGAAC Integrated Exon 8 TGCTCAT TCTCAAACT DNA Technologies DOCK8 CTTCTGGTCTGAAACGCT TGATGCTGAGCTTC Integrated Exon 20 GGA AGCTCAT DNA Technologies M13 GTAAAACGACGGCCAG CAGGAAACAGCTAT Sigma GAC Aldrich

2.5.9 TA cloning TA cloning was performed using the TOPO TA cloning kit for sequencing as per the manufacturer’s instructions. Briefly, fresh PCR product was gel extracted using MinElute gel extraction kit and added at a 2:1 molar ratio to 1µl TOPO vector with 1µl salt solution, and water to a volume of 6µl and incubated at room temperature for 30min before being placed on ice. 1µl of the cloning reaction was added at a 1:5 dilution to 25µl of competent cells and incubated on ice for 20min. Cells were then placed at 42oC for 30sec and returned to ice for 10min. 200µl of room temperature SOC medium was added and the sample incubated at 37oC for 1hr. Cells were spread on pre-warmed LB plates containing 50µg/ml carbenicillin and grown overnight at 37oC. Colonies were picked and added to PCR mix for amplification and purification was performed before sequencing.

2.5.10 Western blotting An equal number of PBMC or LCL cells were treated with NP40 lysis buffer and protease inhibitor cocktail on ice for 30min followed by a 20min spin at 5000g to harvest protein lysates. Samples were separated by gel electrophoresis in NuPAGE 4-12% Tris-Bis gels at 200V for 30min with SeeBlue Plus2 pre-stained standard before transfer to Immobilon- FL membrane at 30V for 90min. The membrane was blocked with Odyssey blocking buffer for 1hr at room temperature before primary antibody (Table 2.8) was added overnight at 4oC. The membrane was washed 4x with PBS-1% Tween-20 and secondary 34

antibody (Table 2.8) added for 1hr at room temperature without light before being washed again and scanned.

Table 2.8 Abs used for Western Blotting Antibody Host Specificity Clone Supplier Dilution GAPDH Mouse Human 6C5 Santa Cruz 1:1000 Biotechnology DOCK8 Mouse Human G-2 Santa Cruz 1:1000 Biotechnology DOCK8 Rabbit Human EPR12511 Abcam 1:10000 IgG IRDye Donkey Mouse Polyclonal Millenium Science 1:20000 680RD H+L IgG IRDye Goat Mouse Polyclonal Millenium Science 1:20000 800CW H+L IgG IRDye Donkey Rabbit Polyclonal Millenium Science 1:20000 800CW H+L

2.6 Statistical analysis Prism software (GraphPad) was used to determine significant differences between data sets by either unpaired t-test or unpaired t-test with Welch’s correction (when number of samples differed between groups) or paired t-test (when data was from the same individual at different times).

Chapter 3: CD4+ T cell differentiation in DOCK8 deficiency

3.1 Introduction CD4+ T cells play a vital role in the immune system’s defence against a range of threats. Their key function is to produce cytokines which activate and recruit other cells of the innate and adaptive immune system to carry out a myriad of functions, ultimately resulting in pathogen clearance. The differentiation of naïve CD4+ T cells into specialised subsets of effector cells has evolved to satisfy the diversity of required tasks. This process is mediated by the coordinated induction of expression, and subsequent function, of transcription factors by the cytokine milieu present at time of stimulation. Th1 cells are generated in response to IL12-mediated induction of expression of the transcription factor T-bet and produce the cytokines IFNγ and TNFα to provide protection against viruses and bacteria. Th2 cells protect against extracellular parasites by producing IL4, IL5 and IL13 upon induction of expression of GATA3 by IL4. Fungal infections are controlled by Th17 cells which produce a suite of cytokines including IL17A, IL17F and IL22 and express RORγt which is induced by the combination of IL1β, IL6, IL23 and TGFβ92.

While sufficient production of cytokines is required for healthy immune function, dysregulated cytokine production can have deleterious and pathogenic consequences. For instance, excessive amounts of Th1 or Th17 cytokines can lead to inflammatory or autoimmune conditions such as multiple sclerosis, rheumatoid arthritis, psoriasis and inflammatory bowel disease, while overproduction of Th2 cytokines can cause allergies and asthma93.

Although cytokine production has not been investigated extensively in CD4+ T cells in DOCK8-deficient patients, it has been examined in other immunodeficiencies that share clinical similarities with DOCK8 deficiency. These patients have loss of function (lof) mutations in STAT3 or gain of function (gof) mutations in STAT1and both exhibit increased susceptibility to recurrent and often severe bacterial and fungal infections, as seen in DOCK8-deficient patients. Analysis of memory CD4+ T cells from these patients showed a decrease in secretion of Th1 and Th17 cytokines94. STAT3lof and STAT1gof naïve CD4+ T cells have also been shown to be defective in differentiating into Th17- cytokine secreting cells in vitro under appropriate polarising culture conditions95. Thus, we wondered if DOCK8-deficient patients also had defects in Th1 and Th17 cytokine

production. To investigate this further, we performed a comprehensive analysis of cytokine production and secretion ex vivo and in vitro by naïve and memory CD4+ T cells from DOCK8-deficient patients relative to corresponding cells from healthy donors.

3.2 Results

3.2.1 DOCK8 is expressed in naïve and memory CD4+ T cells in human peripheral blood, tonsil and spleen To initiate our investigation into the function of DOCK8 in human CD4+ T cells, we first assessed DOCK8 expression by flow cytometry in naïve and memory CD4+ T cells in mononuclear cells isolated from peripheral blood, tonsil and spleen from healthy donors. Naïve cells were identified phenotypically as CD4+CCR7+CD45RA+ and memory cells as CD4+CCR7±CD45RA-. DOCK8 was highly expressed in both naïve and memory CD4+ T cells irrespective of the tissue of origin (Fig 3.1A-C). Furthermore, levels of DOCK8 expression were comparable in naïve and memory CD4+ T cells in blood and spleen, but slightly higher in tonsillar naive cells than tonsillar memory cells (Fig 3.1A-C). These findings reveal ubiquitous expression of DOCK8 during CD4+ T cell differentiation, suggestive of an important role in this process independent of cell location.

3.2.2 DOCK8 expression is maintained in in vitro cultured CD4+ T cells A key goal of this chapter is to determine the functional requirements of DOCK8 during CD4+ T cell differentiation. Thus, we next determined DOCK8 expression in in vitro activated CD4+ T cells. Naïve and memory CD4+ T cells were sort-purified from PBMCs of healthy donors and cultured with TAE beads (anti-CD2, -CD3, -CD28 mAbs conjugated to microbeads) for either 0, 3 or 5 days. Staining for DOCK8 showed that naïve and memory CD4+ T cells continued to express high amounts DOCK8 over time and DOCK8 was not downregulated during the culturing process (Fig 3.2). As in vitro cultured healthy donor cells maintained DOCK8 expression over the culture time period it was confirmed that these cells could be used as a comparison for in vitro cultured DOCK8-deficient cells in subsequent experiments.

Figure 3.1 Expression of DOCK8 in naïve and memory CD4+ T cells isolated from human peripheral blood, spleen and tonsil. (A) naïve and memory CD4+ T cells were isolated from (B) peripheral blood (C) tonsil and (D) spleen samples from healthy controls (n=3) by surface staining for CD4, CCR7 and CD45RA before being fixed, permeabilised and intracellularly stained for DOCK8 with expression determined by flow cytometry. Histograms show one representative sample. Graphs show mean±SEM. * p<0.05; ** p<0.01; *** p<0.005.

Figure 3.2 DOCK8 expression in CD4+ T cells during in vitro culture. Sorted (A) CD45RA+CCR7+ naïve and (B) CD45RA- memory CD4+ T cells from healthy donors (n=3) were cultured with TAE beads for either 0, 3 or 5 days and then intracellularly stained for DOCK8 with expression determined by flow cytometry. Contour plots show one representative sample. Graphs show mean ±SEM. **** p<0.001

3.2.3 DOCK8-deficient memory CD4+ T cells are skewed to a Th2 fate at the expense of Th1 and Th17 cells To examine the impact of DOCK8 deficiency on the in vivo differentiation of naïve CD4+ T cells into Th subsets, memory CD4+ T cells were sorted from healthy donors and DOCK8-deficient patients and cultured for 5 days with TAE beads. Supernatants were collected and cells harvested, which were then restimulated with PMA/ionomycin in the presence of Brefeldin A to enable detection of accumulated cytokines. After 6 hours, restimulated cells were harvested for intracellular cytokine staining to measure the frequency of cells expressing Th1 cytokines (IFNγ and TNFα) and Th2 cytokines (IL4 and IL13). There were significantly fewer DOCK8-deficient memory CD4+ T cells expressing Th1 cytokines and significantly higher proportions expressing Th2 cytokines than memory CD4+ T cells from healthy donors (Fig 3.3A). This resulted in a significantly skewed Th1/Th2 cytokine ratio (Fig 3.3B). Analysis of supernatants for levels of secreted Th1 (IFN, TNF), Th2 (IL4, IL5, IL13) and Th17 (IL17A, IL17F, IL22) cytokines by cytometric bead arrays and ELISAs following 5-day culture confirmed and extended the findings obtained by flow cytometric analysis. Thus, the pattern of cytokine secretion matched that of intracellular cytokine production with DOCK8-deficient memory CD4+ T cells secreting significantly lower levels of Th1 and increased levels of Th2 cytokines compared to memory CD4+ T cells from healthy donors (Fig 3.3C). These findings indicate that recessive pathogenic mutations in DOCK8 results in an increased proportion of Th2 cells and decreased proportion of Th1 cells (Fig 3.3B). This analysis also revealed that DOCK8-deficient memory CD4+ T cells produced very limited amounts of IL17A, IL17F and IL22 (Fig 3.3C) indicating a lack of Th17 cells. Collectively, these results suggest that DOCK8 deficiency impacts CD4+ T cell differentiation by selectively promoting the generation of Th2 cells from precursor naïve CD4+ T cells, at the expense of Th1 and Th17 cells.

Figure 3.3 Cytokine production and secretion by DOCK8-deficient memory CD4+ T cells. Sort-purified memory CD4+ T cells from healthy donors (n=4-5) or DOCK8-deficient patients (n=3-4) were cultured with TAE beads for 5 days. (A, B) Cells were restimulated with PMA/ionomycin in the presence of Brefeldin A and intracellularly stained for cytokines with expression determined by flow cytometry. (C) Supernatants were harvested to determine cytokine secretion by CBA. Graphs show mean±SEM. * p<0.05; ** p<0.01

3.2.4 The skewed CD4+ T cell differentiation of DOCK8-deficient memory CD4+ T cells is not due solely to aberrant expression of lineage-specific transcription factors To explore the possible mechanism underlying the skewing of DOCK8-deficient CD4+ T cells towards Th2 differentiation, expression of canonical transcription factors that mediate commitment of naïve CD4+ T cells to distinct effector fates was measured. Sorted memory CD4+ T cells were cultured for 5 days with TAE beads before RNA was harvested and qPCR performed to determine the relative expression of the transcription factors associated with Th1 (TBX21), Th2 (GATA3) and Th17 (RORC) cell differentiation. On one hand, the relative expression levels of GATA3 were significantly increased in DOCK8-deficient memory CD4+ T cells compared to memory CD4+ T cells from healthy controls (Fig 3.4). This is consistent with increased Th2 differentiation due to DOCK8 deficiency. On the other hand, relative expression of TBX21 in DOCK8- deficient memory CD4+ T cells was comparable to healthy donors, and RORC relative expression, while reduced in DOCK8-deficient memory CD4+ T cells, was not significantly different to that of normal memory CD4+ T cells (Fig 3.4). This comparable TBX21 and RORC expression is inconsistent with the reduced production of Th1 and Th17 cytokines by DOCK8-deficient memory CD4+ T cells. Hence, altered transcription factor expression alone does not explain the skewed CD4+ T cell differentiation seen in DOCK8-deficient cells, revealing an additional level of complexity to DOCK8-mediated regulation of human CD4+ T cell differentiation in vivo.

Figure 3.4 Transcription factor expression in DOCK8-deficient memory CD4+ T cells. Sorted memory CD4+ T cells from healthy donors (n=3) or DOCK8-deficient patients (n=3) were cultured with TAE beads for 5 days before RNA was harvested, cDNA made and qPCR performed to determine relative expression of TBX21, GATA3 and RORC. Graphs show mean±SEM. * p<0.05

3.2.5 TCR Vβ repertoire diversity is maintained in DOCK8-deficient memory CD4+ T cells Given that DOCK8-deficient patients are susceptible to various infections, we sought to investigate whether recurrent pathogen exposure and infection impacted CD4+ T cell differentiation, resulting in a skewed memory T cell compartment. As bacterial, viral and parasitic infections result in a biased TCR Vβ repertoire96-98, we analysed the TCR Vβ repertoire in DOCK8-deficient memory CD4+ T cells by flow cytometry which allowed for detection of ~70% of all expressed TCR Vβ chains. Staining for a wide range of TCR Vβ regions revealed a diverse polyclonal repertoire in DOCK8-deficient memory CD4+ T cells which did not differ from that of memory CD4+ T cells from healthy donors (Fig 3.5). This would suggest that any infections experienced by DOCK8-deficient patients did not result in a particular TCR Vβ chain dominating the repertoire that may preferentially favour the skewing of the CD4+ T cell memory compartment to a Th2 fate. Additionally, the presence of a diverse TCR repertoire in DOCK8-deficient patients indicates that the susceptibility of patients to infections is not due to the absence of particular Vβ clonotypes that may be responsible for generating responses to specific pathogens.

Figure 3.5 TCR diversity in DOCK8-deficient memory CD4+ T cells. PBMCs from healthy donors (n=4) or DOCK8-deficient patients (n=3) were stained for CD4, CCR7, CD45RA and TCR diversity using a TCR Vβ repertoire kit with expression determined by flow cytometry. Graphs show the mean±SEM of memory CD4+ T cells expressing the indicated TCR V chain.

3.2.6 DOCK8-deficient memory CD4+ T cells exhibit signs of decreased in vitro TCR- induced activation Previous research has shown that TCR signalling strength can influence the differentiation fate of CD4+ T cells, with stronger signalling promoting Th1 and Th17 cells and weaker signalling resulting in more Th2 cells99-101. Hence, we sought to test whether TCR signalling in DOCK8-deficient memory CD4+ T cells was impaired. To do this, we assessed induction of expression of activation markers as a downstream read-out of TCR signalling. Sorted memory CD4+ T cells were cultured with TAE beads at a cell:bead ratio of either 1:2 or 2:1 for 0 or 3 days, after which time expression of the surface markers CD25, CD69, CD95 and ICOS was assessed by flow cytometry. Significantly fewer DOCK-deficient memory CD4+ T cells upregulated CD69, CD25 and ICOS compared to corresponding cells from healthy donors in response to TAE bead stimulation (Fig 3.6). Interestingly, increasing the proportion of TAE beads four-fold did not greatly change T cell activation suggesting that a 1:2 ratio was already a strong stimulus. These results suggest that DOCK8 deficiency may result in weakened TCR signalling during CD4+ T cell activation and this would favour differentiation towards the Th2 lineage.

Figure 3.6 Activation marker expression of in vitro cultured DOCK8-deficient memory CD4+ T cells. Sorted memory CD4+ T cells from healthy donors (n=3) or DOCK8-deficient patients (n=3) were cultured for 3 days with TAE beads at bead:cell ratio of 1:2 or 2:1 and then stained for CD69, CD95, CD25 and ICOS with expression determined by flow cytometry. Graphs show mean±SEM. * p<0.05

3.2.7 DOCK8-deficient naïve CD4+ T cells are able to undergo Th1 and Th2 differentiation, but not Th17 differentiation in vitro As we had found that DOCK8 deficiency resulted in skewed in vivo CD4+ T cell differentiation through analysis of the memory T cell compartment, we next sought to determine whether this defect was intrinsic or extrinsic. To this end, we tested the ability of DOCK8-deficient naïve CD4+ T cells to undergo in vitro differentiation to specific effector fates. Sorted naïve CD4+ T cells were cultured with TAE beads for 5 days under either neutral (Th0) conditions or Th1- , Th2- or Th17- polarising conditions. Culture supernatants were collected and CBA analysis performed to measure secreted cytokines. Despite the reduction in Th1 cells in the memory T cell compartment of DOCK8-deficient patients, DOCK8-deficient naïve CD4+ T cells cultured in vitro under Th1-polarising conditions secreted comparable amounts of Th1 cytokines as naïve CD4+ T cells from healthy donors (Fig 3.7A). However, in contrast to Th1 polarisation, DOCK8-deficient naïve CD4+ T cells cultured in vitro under Th17-polarising conditions were markedly impaired in their ability to secrete the Th17 cytokines IL17A and IL17F (Fig 3.7C). Th2 cytokine production trended towards being increased in Th2-polarised DOCK8-deficient naïve CD4+ T cells (Fig 3.7B). This indicates that the Th17 differentiation defect in DOCK8-deficient CD4+ T cells is intrinsic to these cells as differentiation in vivo or in vitro yielded the same result. The Th1 differentiation defect, however, is only seen when CD4+ T cells differentiate under in vivo conditions, suggesting an element of the priming microenvironment that differs from the in vitro conditions used is involved in poor Th1 responses.

Figure 3.7 Cytokine secretion of in vitro cultured DOCK8-deficient naïve CD4+ T cells. Sorted naïve CD4+ T cells from healthy donors (n=4) or DOCK8-deficient patients (n=3) were cultured for 5 days under Th0 (TAE only), Th1 (+IL12), Th2(+IL4) or Th17 (+ IL1, IL6, IL21, IL23, TGF- , Prostaglandin E2) polarising conditions. Supernatants were collected and cytokine secretion determined by CBA. Graphs show mean±SEM.

3.2.8 DOCK8-deficient naive CD4+ T cells do not show signs of decreased in vitro TCR-induced activation Considering we had seen both decreased Th1 differentiation and decreased TCR signalling seen in DOCK8-deficient memory CD4+ T cells, we next looked at TCR signalling in naïve CD4+ T cells. As naïve CD4+ T cells cultured in vitro did not show a Th1 differentiation defect, it could be that they also do not show decreased TCR signalling. Hence, sorted naïve CD4+ T cells were cultured with TAE beads at a 2:1 bead:cell ratio for 0 or 3 days and expression of surface markers measured. Indeed, we found that DOCK8-deficient naïve CD4+ T cells had normal frequencies of cells expressing the activation markers CD25, CD69, CD95 and ICOS (Fig 3.8). The finding that naïve CD4+ T cells had normal Th1 differentiation and normal TCR-induced activation while memory CD4+ T cells had decreased Th1 differentiation and decreased TCR-induced activation indicates there is a link between decreased TCR signalling and decreased Th1 differentiation. However, while this result suggests that there is no intrinsic defect in TCR-induced activation in naïve CD4+ T cells from DOCK8-deficient patients, it may be that the in vitro conditions used are sufficiently strong to overcome any defect.

Figure 3.8 Activation marker expression of in vitro cultured DOCK8-deficient naïve CD4+ T cells. Sorted naive CD4+ T cells from healthy donors (n=3) or DOCK8-deficient patients (n=3) were cultured for 3 days with TAE beads at bead:cell ratio of 2:1 and then stained for CD69, CD95, CD25 and ICOS with expression determined by flow cytometry. Graphs show mean±SEM.

3.2.9 DOCK8-deficient naïve and memory CD4+ T cells do not display decreased TCR-induced calcium flux To further investigate the impact of DOCK8 deficiency on TCR signalling, we next measured TCR-induced calcium flux in naïve and memory CD4+ T cells from healthy donors and patients. In an immune response, engagement of the TCR results in the rapid release of calcium ions from the endoplasmic reticulum; this is then followed by entry of extracellular calcium through channels in the plasma membrane. As an outcome of this calcium release is the cessation of cell movement and stabilisation of immune synapses, it can be recognized that the greater the calcium flux, the stronger the initial TCR signalling102,103. Calcium flux has the advantage that it can be measured immediately following TCR stimulation so does not require cell purification or culture and is upstream of activation marker expression which was previously assessed (see Figs 3.6,3.8). For this assay, PBMCs were stained with the calcium indicator dye Indo-1AM and surface stained with CD4, CCR7 and CD45RA to identify naïve and memory CD4+ T cells. After recording a baseline threshold, anti-CD3-biotin and streptavidin were added at 100 seconds to stimulate and crosslink the TCR and the frequency of responding cells recorded for a further 300 seconds with the ionophore ionomycin added at 330 seconds as a positive control.

Neither naïve nor memory DOCK8-deficient CD4+ T cells displayed decreased calcium flux following TCR engagement compared to healthy donors (Fig 3.9A, B). Interestingly, 1 patient displayed a normal calcium flux in naïve and memory CD4+ T cells, another patient exhibited elevated flux in naïve CD4+ T cells and normal calcium flux in memory CD4+ T cells and a third patient demonstrated elevated calcium flux in both naïve and memory CD4+ T cells (Fig 3.9A, B). Calcium flux was consistently higher in naïve CD4+ T cells than memory CD4+ T cells from both DOCK8-deficient patients and healthy donors. This suggests that proximal TCR signalling events are unaffected by DOCK8 deficiency, and that DOCK8 mutations probably impact activation downstream of these early TCR signalling events.

As a comparison, calcium flux was also measured in patients with lof mutations in STAT3 (n=1) or ZNF341 (n=2) who, like DOCK8-deficient patients, present with hyper IgE, infectious susceptibility and Th2 skewing33,45. The STAT3lof patient displayed calcium flux that was comparable to a healthy donor in both naïve and memory CD4+ T cells (Fig 3.9 C, D). Naïve CD4+ T cells from one ZNF341-deficient patient displayed slightly 53

increased calcium flux while the other patient had normal calcium flux (Fig 3.9C). Memory CD4+ T cells from both ZNF341-deficient patients demonstrated increased calcium flux compared to memory cells from healthy donors (Fig 3.9D). Thus, overall, it would appear that these genetic mutations which manifest clinically as predisposition to infection and allergic disease due to dysregulated CD4+ T cell differentiation in vivo are likely to impact T cell function downstream of early TCR signalling events, as is the case for DOCK8 deficiency.

Figure 3.9 TCR-induced Calcium flux in CD4+ T cells. PBMCs from healthy donors (n=6), DOCK8-deficient patients (n=3; panels A and B), STAT3lof (n=1) or ZNF341-deficient (n=2) patients (panels C and D) were loaded with indo-1AM and stained for CD4, CCR7 and CD45RA. Cells were stimulated with CD3 mAb and crosslinked at 90 seconds to induce calcium flux which was measured by flow cytometry over a period of 7 minutes in naïve (A, C) and memory (B, D) CD4+ T cells. Ionomycin was added at 330 seconds to induce maximal response. Each plot shows the response of an individual unstimulated healthy donor, a stimulated healthy donor and a stimulated patient which were run in the same experiment.

3.3 Discussion Our investigation into the functionality of the CD4+ T cell compartment of DOCK8- deficient patients revealed that their memory T cells had an increased proportion of Th2 cells and a corresponding decreased proportion of Th1 and Th17 cells, compared to healthy donors. This skewed differentiation very likely contributes to some of the most frequent clinical symptoms experienced by these patients. As Th1 cells are responsible for protection against bacteria and viruses104, their decreased frequency in DOCK8- deficient patients possibly explains their susceptibility to infections by S. aureus and herpes viruses. Additionally, as Th17 cells are responsible for protection against fungal infections105, the lack of these cells in DOCK8-deficient patients would provide an explanation for the high incidence of chronic mucocutaneous candidiasis in these individuals. Furthermore, the increase in Th2 cells in DOCK8 deficiency is consistent with the increased incidence of Th2-mediated diseases such as eczema and food allergies in DOCK8-deficient patients. Interestingly, as this thesis was being undertaken another group reported some similar findings. They had previously identified impaired Th17 differentiation in autosomal recessive HIES patients but the patients’ genotypes were unknown106. They subsequently analysed naïve CD4+ T cells from confirmed DOCK8- deficient patients cultured in vitro under polarising conditions and then stimulated for 48hours with PMA/ionomycin or TAE beads. They found decreased IL17A transcripts and IL17A secretion as well as increased IL4 and IL5 secretion and normal IFN transcripts and secretion54.

The theory of TCR signalling strength playing an important role in CD4+ T cell differentiation has not been as extensively explored as the Th1/Th2 paradigm which is based on cytokine signalling. Research has investigated possible mechanisms for how TCR signalling exerts its influence on CD4+ T cell differentiation including antigen dose, TCR signalling duration and the requirement for costimulation107-109. This work revealed that strong TCR stimulation favours Th1 differentiation and a short stimulus can achieve this while weak TCR stimulation favours Th2 differentiation and this requires a prolonged stimulus as well as CD28 costimulation107-109. However, the influence of TCR signalling strength relative to the influence of cytokine signalling on the outcome of Th differentiation is yet to be determined. While it is known that in vitro culture with IL12 promotes Th1 differentiation and culture with IL4 promotes Th2 differentiation, preliminary experiments undertaken as part of this study showed that stronger in vitro 56

TCR signalling favoured Th1 differentiation while weaker TCR signalling favoured Th2 differentiation of naïve CD4+ T cells from a healthy donor. This raised the possibility that the lack of Th1 cells coupled with the overabundance of Th2 cells in the memory CD4+ T cell compartment of DOCK8-deficient patients could be due to weakened TCR signalling. This could be as a result of impaired IS formation in vivo between CD4+ T cells and APCs, a process which DOCK8 has been shown to be involved in with CD4+ Tregs from Dock8-deficient mice87. Our result of decreased activation maker expression upon TCR stimulation in DOCK8-deficient memory CD4+ T cells did suggest impaired TCR signalling but data from calcium flux experiments did not support this hypothesis. It should be noted that the TAE beads used for activation in vitro stimulate not only CD3, but also CD2 and CD28, important costimulatory molecules for T cell activation, which are not stimulated during calcium flux. This may have contributed to the differing results seen between surface marker expression and calcium flux. It may also be that TCR signalling is impacted further downstream of the early events which result in calcium flux or that calcium flux is not sufficiently sensitive to detect changes in TCR signalling due to DOCK8 deficiency. TCR signalling is a multistep process that involves the phosphorylation and recruitment of a host of different proteins110. Our finding of intact TCR signalling by calcium flux could be confirmed by further investigation of earlier TCR signalling events such as phosphorylation of proteins including Lck, ZAP70 or LAT in DOCK8-deficient memory CD4+ T cells. It is known that phosphorylation of these proteins is crucial for calcium flux as patients with mutations in LCK, ZAP70 or LAT exhibit impaired calcium flux in T cells111-113. Additionally, exploration of processes downstream of calcium flux in the TCR signalling pathway such as activation of Akt or localisation of nuclear factors such as NFAT and NF-κb may reveal that DOCK8 impacts a later stage of TCR signalling.

It is intriguing that the decrease in Th1 cells was seen only for in vivo generated memory CD4+ T cells, but not following in vitro differentiation of naïve CD4+ T cells lacking DOCK8 expression. Hence, we must consider how the elements of the in vivo environment are replicated in an in vitro system and the differences present between the two. As discussed, TCR signalling strength can influence CD4+ T cell differentiation. TAE beads were used in in vitro culture to mimic T cell activation induced via engagement of CD2, CD3 and CD28 at a bead:cell ratio of 1:2. It is possible that this T cell signalling is much stronger than that experienced under physiological conditions and 57

any TCR signalling defects in DOCK8-deficient naïve CD4+ T cells, which would result in intrinsically less Th1 differentiation, were circumvented by the strong stimulation of the TCR and costimulatory signals.

Another difference with in vitro cultured cells is that they develop in a cytokine restricted environment. The naïve CD4+ T cells cultured under Th1-polarising conditions we studied were generated in a cytokine restricted environment with the addition of exogenous IL12 only and the absence of competing cytokines, such as IL4 which would promote Th2 differentiation. The memory CD4+ T cells we analysed, on the other hand, developed in the presence of an undefined cytokine milieu in vivo. It is possible that DOCK8-deficient cells respond more strongly to IL4 than to IL12 when both cytokines are present. DOCK8-deficient CD4+ T cells may have an increased number of IL4 receptors or decreased number of IL12 receptors or they may respond to IL4 more quickly, resulting in rapid production of Th2 cells which produce IL4 which inhibits Th1 differentiation. The development of an in vitro culture which includes both IL4 and IL12 and gives rise to both Th1 and Th2 differentiation would be required to test this hypothesis. Preliminary experiments were undertaken to test the production of Th1 and Th2 cells under culture with different concentrations of IL12 and IL4. One condition was identified that resulted in a significant and approximately equal proportion of IFNγ- expressing and IL4- expressing cells from naïve CD4+ T cells isolated from a healthy donor (11.1% IFNγ positive, 9.8% IL4 positive). Unfortunately, due to limited patient samples, this was unable to be tested using cells from DOCK8-deficient patients.

An additional point of difference between the in vivo environment and in vitro culture conditions is the absence of Tregs from the latter. Tregs are CD4+ T cells that require the transcription factor Foxp3 and are able to inhibit the function of other effector CD4+ T cells114. However, previous reports established that Tregs from DOCK8-deficient patients were impaired in their suppressive ability67. Hence, when naïve CD4+ T cells were sorted, Tregs were excluded based on their high expression of CD25115 and thus were not present during in vitro culture. It has been reported that Tregs can act specifically on certain Th subsets with the expression of TIGIT causing Treg cells to suppress Th1 and Th17 cells but not Th2116. It is possible that DOCK8 deficiency impacts Tregs so that they target Th1 cells which would lead to our observed result of decreased Th1 cells in the memory compartment, which differentiated in vivo, but not naïve CD4+ T cells cultured in vitro.

It was unexpected that DOCK8-deficient patients had decreased Th17 cells, and a T cell intrinsic defect in generating Th17 cells, and yet did not have significantly decreased expression of the transcription factor RORC which is crucial in Th17 development117. Indeed, patients with mutations in RORC fail to generate Th17 cells in vivo and in vitro118. Although we only assessed RORC expression in memory cells, it has been shown by others that DOCK8-deficient naïve CD4+ T cells cultured under Th17 polarising conditions and then stimulated for 48 hours have reduced RORC transcripts54. However, there have been reports of immunodeficiency patients with decreased Th17 cells that did not have decreased RORC expression. Analysis of ZNF341-deficient memory CD4+ T cells after 5 days of culture showed decreased secretion of Th17 cytokines but only modestly reduced RORC mRNA levels45 , similar to our observations for DOCK8- deficient patients. However, when ZNF341-deficient memory CD4+ T cells were cultured in Th17 activating conditions, there was significantly less RORC expression compared to healthy controls45. As RORC expression in the healthy controls and DOCK8-deficient patients we tested was very low (relative expression in range 0.03-0.6 vs TBX21 range 3.1-16 and GATA3 range 0.5-4.4) differences may have been obscured and measuring RORC expression under Th17 conditions could shed more light on the regulation of gene expression in DOCK8-deficient CD4+ T cells.

Alternatively, impaired Th17 differentiation in DOCK8 deficiency could occur independently of RORC expression. Previous research in mice has shown that PD-L1 upregulation can inhibit Th17 differentiation119. Studies of CD4+ T cells from STAT3lof and STAT1gof immunodeficiency patients replicated this result and found that PD1- mediated inhibition of Th17 cell generation could be overcome by PD-L1 inhibition120. Investigation of PD-L1 expression on DOCK8-deficient CD4+ T cells would be required to assess whether this was applicable to DOCK8 deficiency. As STAT3lof patients also exhibit impaired Th17 cell differentiation121 it is possible that perturbations in STAT3 signalling could result in the decreased Th17 cell differentiation in DOCK8-deficient patients. Indeed, previous work has shown that STAT3 phosphorylation is reduced in T cells from DOCK8-deficient patients in response to cytokine stimulation and this results in impaired pSTAT3 translocation to the nucleus and thus reduced expression of STAT3- inducible genes54.

The work in this chapter not only provides explanation for some of the characteristic clinical features of DOCK8-deficient patients but indicates a role for DOCK8 in CD4+ T cell differentiation.

Chapter 4: Haematopoietic stem cell transplant effectively rescues lymphocyte differentiation and function in DOCK8-deficient patients

4.1 Introduction Primary immunodeficiencies (PIDs) are rare conditions caused by mutations in a single gene that cripples the development and/or function of immune cells 1,122. Currently, >350 genes have been identified that, when mutated, can result in immune dysregulation 1,123. Although these inborn errors of immunity have been typically associated with heightened susceptibility to disease due to recurrent pathogen infections, the phenotype of PIDs is much broader and can include autoimmunity, autoinflammation, allergy and malignancy 1,124-126. While PIDs due to a specific gene defect are rare, the rapid discovery of the molecular causes of novel PIDs, the ongoing appreciation of the diversity of clinical presentations of these conditions, and the application of newborn screening across several countries are revealing that collectively the incidence of PIDs is much greater than typically reported 122,123,127,128. For these reasons, it is important to understand the biology and pathogenesis of individual PIDs and have a thorough knowledge of the optimal treatments and subsequent outcomes for PIDs resulting from mutations in specific pathways.

Severe combined immunodeficiencies (SCID), due to mutations in IL2RG, JAK3, ADA, RAG, or IL7R, or combined immunodeficiencies (CIDs), due to mutations in e.g. DOCK2, STK4, MALT1, CARD11, IKBKB, are usually fatal unless early therapeutic intervention such as hematopoietic stem cell transplant (HSCT), gene therapy or enzyme replacement is applied 122. The first HSCTs for PID were performed in 1968 30,129. While results from these initial transplants for PID were disappointing 30,129, remarkable advances have been made over the past 50 years such that the overall survival of SCID/CID patients following HSCT can exceed 95% 130-135. However, depending on the age at time of transplant, incidence of infection, and nature and source of the donor, mortality post-HSCT can remain significant, with 5-10 year survival ranging from <40 to ~80% 130-132,134,135. Thus, in order to improve therapeutic outcomes, it is critical to identify correlates or biomarkers of successful immune cell reconstitution in PID patients following HSCT.

DOCK8 is a guanine nucleotide exchange factor with key roles in regulating cytoskeletal rearrangement, cell activation, migration and survival 136,137. Despite its broad expression, DOCK8 has a critical and non-redundant role in immunity, as revealed by the discovery that bi-allelic DOCK8 mutations cause a CID characterised by recurrent mucocutaneous viral, bacterial and fungal infections (80-90% of cases), severe eczema (>95%), allergies (~70%), hyper-IgE (98%), and increased susceptibility to malignancy (HPV-induced carcinoma, EBV-lymphoma) and autoimmunity 40,41,51,56,136,137.

Numerous studies have investigated cellular defects in DOCK8 deficiency to understand both the non-redundant roles of DOCK8 in lymphocyte biology and mechanisms of disease in DOCK8-deficient patients. These investigations revealed dysregulated survival, proliferation, differentiation, migration and senescence/exhaustion of CD4+ and CD8+ T cells 41,54,65,66,88,138, decreased regulatory T cell (Treg) function 67, NK cell cytotoxicity 68,71 and NKT cell development 72, and reduced B-cell activation in vitro and memory B cell generation in vivo 59,79.

Similar to other CIDs, outcomes for DOCK8 deficiency are poor with >95% mortality by 40 years (median survival ~10-20 years), and the incidence of life-threatening infections and malignancy increases every decade 51,56. Consequently, HSCT is the standard of care for the life-threatening infections and related immune complications associated with DOCK8 deficiency 56. Several studies have examined outcomes of HSCT in DOCK8 deficiency, with generally positive results (~80% survival), but varying degrees of clinical improvement. Eczema, cutaneous viral and bacterial infections, responses to vaccines and levels of serum IgM, IgG and IgA all markedly improved post-HSCT 139-150. In contrast, allergic disease following HSCT is highly variable, either resolving 79,145,151, improving 79,139,140,142 or persisting 79,146,152. Clinical improvements in transplanted DOCK8-deficient patients have been associated with both mixed 145,149,152 or complete 139,141,146,147 donor chimerism.

In this study, we have used DOCK8 deficiency as a model to delineate mechanisms underlying disease pathogenesis pre-HSCT and improvement of clinical features of PID post-HSCT and identify correlates of immune reconstitution and function following HSCT. This allowed us to extensively catalogue cellular defects due to DOCK8 deficiency and investigate quantitative and qualitative improvement of these defects post- HSCT. Cellular improvements correlated with reconstitution of DOCK8 protein

expression and clinical outcomes in these patients. To date, this is the largest study of this kind and provides important insights into the functional changes that may predict successful immune reconstitution and guide ongoing treatments and management of DOCK8-deficient patients following HSCT. Furthermore, our study provides proof-of principle for performing high-dimensional multi-functional cellular analyses pre- and post-therapy in other PIDs to understand treatment-induced alternations in cellular behaviour and clinical outcomes and guide implementation of optimal treatments for these conditions.

4.2 Results 4.2.1 DOCK8 is constitutively expressed by lymphocytes in healthy donors and DOCK8-deficient patients post-HSCT To gain insight into the role of DOCK8 in immune function, we first determined DOCK8 expression in the major lymphocyte subsets in peripheral blood mononuclear cells (PBMCs) of healthy volunteers. DOCK8 was highly and comparably expressed in total T cells, CD4+ and CD8+ T cells, B cells and NK cells (Fig 4.1A)53,153. We also established that DOCK8 is constitutively expressed in NKT and MAIT cells (Fig 4.1A). Next, we confirmed lack of expression in patients with DOCK8 mutations and assessed restoration of DOCK8 expression following HSCT. Patients studied here exhibited near-undetectable levels of DOCK8 protein, with expression in lymphocytes (Fig 4.1B), CD4+ T, CD8+ T and CD20+ B cells (Fig 4.1C) being drastically reduced compared to healthy volunteers. Importantly, DOCK8 expression in these lymphocyte populations from transplanted patients was restored to similar levels as lymphocytes from healthy volunteers (Fig 4.1B, 1C).

Figure 4.1: DOCK8 is highly expressed in lymphocyte subsets, absent in DOCK8- deficient patients and restored following HSCT (A) PBMCs from healthy donors (n=3) were stained with Abs against CD3, CD4, CD8, CD20, CD56, CD161 and TCR Vα24, Vβ11, and Vα7.2. The cells were then fixed, permeabilised and stained with anti-DOCK8 mAb. Expression of intracellular DOCK8 in total T cells (CD3+), CD4+ T cells (CD3+CD4+CD8-), CD8+ T cells (CD3+CD4- CD8+), B cells (CD20+CD3-), NK cells (CD3-CD56+), NKT cells (CD3+TCRVα24+Vb11+) and MAIT cells (CD3+CD161+TCRVα7.2+) was then determined. Data represent the average geometric MFI ± SEM of different lymphocyte subsets from three healthy donors labelled with anti-DOCK8 mAb less the MFI of cells labelled with isotype control mAb. (B, C) PBMC from healthy donors (n=20) or DOCK8- deficient patients before (n=4) or following HSCT (pHSCT; n=15-16) were stained with Abs against CD4, CD8 and CD20 before fixing, permeabilisation and staining for DOCK8. DOCK8 expression was determined in total lymphocytes (B), as well as in CD4+ T cells, CD8+ T cells and CD20+ B cells (C). The histogram in (B) depicts DOCK8 expression in total lymphocytes from one representative healthy donor, and lymphocytes from the same DOCK8-deficient patient pre- and post-transplant as well as an isotype control. The graph in (C) represents the mean MFI ± SEM of DOCK8 expression (minus MFI of isotype control mAb). Statistics were performed using unpaired t-test with Welch’s correction *<0.05, **<0.01, ***<0.005, ****<0.001.

4.2.2 Clinical Characteristics of DOCK8 deficient patients – impact of HSCT We studied an international cohort of DOCK8-deficient patients (Table 4.1) who had either confirmed bi-allelic mutations in DOCK8 (n=18) or lacked DOCK8 protein in their leukocytes (n=2). In total, immune cells were examined in 18 DOCK8-deficient patients post-HSCT; matched PBMC samples were available from 7 patients pre- and post-HSCT, and 2 patients at two time-points post-HSCT. The source of transplant was haploidentical (n=6), matched unrelated (n=7) or matched related (n=6) donors. Consistent with a recent study of HSCT for a large cohort of DOCK8-deficient patients 79., no correlations were observed between the source of the transplant and the overall clinical outcome of the patients’ post-HSCT (Table 4.2), All DOCK8-deficient patients studied here suffered recurrent viral and bacterial infections (Table 4.1). Candidiasis was reported in 30%, allergies in 80%, and impaired vaccine responses in >90% of patients (Table 4.1). Following HSCT, infections were reduced and vaccines responses improved in all DOCK8-deficient patients (Table 4.2). Allergies improved in only 1/11 patients tested (Table 4.2) however this is an underestimate as many of the patients did not undergo formal clinical allergy testing post-HSCT. Consistent with flow cytometric analysis of DOCK8 expression, which revealed comparable expression in patients post-HSCT and healthy donors (Fig 4.1B, 1C), donor engraftment following transplant was >90% in all patients (Table 4.2).

HSCT

deficientpre patients

- Table 4.1: Clinical details of Clinical TableDOCK8 details 4.1: 67

illoma illoma

OM=Otitis Media; RSV=Respiratory Syncytial Virus; SCC=Squamous Carcinoma;Cell VZV=Varicella Zoster Virus.

Barr Barr Virus; HSCT=Haematopoietic Stem Cell Transplantation; HSV=Herpes Virus;Simplex HHV6=Human Herpes Virus 6; HPV=Human Pap

Resistant Staphylococcus aureus;

llin

Cell Lymphoma; EBV=Epstein

ge ge B

CMV=Cytomegalovirus; DLBCL=Diffuse Lar Virus; IV Ig=Intravenous Immunoglobulin; MRSA= Methici 68

deficientpost patients HSCT

- Table 4.2: Clinical details of Clinical TableDOCK8 details 4.2: 69

toxic T Lymphocytes; EBV=Epstein T toxic Lymphocytes;

Resistant Staphylococcus aureus; MTX=Methotrexate; MRD=Matched Related MTX=Methotrexate; aureus; Donor; MRD=Matched Resistant Staphylococcus

Treosulphan.

munoglobulin; MRSA= Methicillin MRSA= munoglobulin;

ter Venous Line; cGVHD=chronic Graft Versus Host Disease; Cy=Cyclophosphamide; Cys=cyclosporin; CMV=Cytomegalovirus; CTL=Cyto CMV=Cytomegalovirus; Cys=cyclosporin; Cy=Cyclophosphamide; ter Graft Disease; Host cGVHD=chronic Line; Versus Venous

d Donor; MMF=Mycophenylate Mofetil; N/A=Not Applicable; ND=Not Done; OM=Otitis Media; TAC=Tacrolimus; Thio=Thiotepa; TAC=Tacrolimus; Treo= Media; N/A=Not OM=Otitis ND=Not Done; Mofetil; d Applicable; Donor; MMF=Mycophenylate

udarabine; GI=Gastrointestinal; HSCT=Haematopoietic Stem Cell Transplantation; HSV=Herpes Simplex Virus; ImIV Simplex Transplantation; HSV=Herpes Ig=Intravenous Cell udarabine; GI=Gastrointestinal; Stem HSCT=Haematopoietic

Thymocyte Globulin; BM=Bone Marrow; Bu=Busulfan; Cam=Campath; CVL=Cathe BM=Bone Globulin; Marrow; Thymocyte Bu=Busulfan; -

ATG=Anti Flu=Fl Barr Virus; Unrelate MUD=Matched 70

4.2.3 DOCK8-deficient lymphocytes exhibit a unique phenotype typical of aberrant in vivo differentiation To elucidate effects of DOCK8 deficiency on lymphocytes and the impact of HSCT on these defects, we undertook extensive phenotypic analysis. Consistent with previous observations 40,138, we found reductions in proportions of CD3+ T cells in DOCK8- deficient patients (Fig 4.2A), largely due to reduced frequencies of CD4+ T cells (Fig 4.2A). Analysis of T cell subsets confirmed skewing of DOCK8-deficient CD8+ T cells + to effector memory (TEM) and effector memory CD45RA (TEMRA) cells at the expense of naïve and central memory (TCM) cells (Fig 4.2B). In contrast, proportions of naïve + + CD4 T cells were comparable to healthy controls, while CD4 TCM cells were decreased + and CD4 TEM cells were increased (Fig 4.2C). Tregs were proportionally increased in DOCK8-deficient patients compared to controls (not shown). DOCK8 deficiency also affected αβ and δ T cells, with reductions and increases, respectively, in these subsets in patients compared to controls, resulting in a skewed αβ/δ T cell ratio (Fig 4.2D). δ T cells have been shown to expand in response to cellular stress induced by infection, with cytomegalovirus (CMV) being the most well studied case154. The recurrent viral infections characteristic of DOCK8-deficient patients may thus contribute to the increase in δ T cells that was consistently noted in these individuals. Furthermore, DOCK8- deficient patients had ~10-fold fewer MAIT cells than healthy controls (Fig 4.2E), frequencies of NK cells and NK cell subsets were normal (Fig 4.2F, not shown, refs 40,68) and NKT cells were reduced (Fig 4.2G)72. In contrast, proportions of total (Fig 4.2A) and naïve (Fig 4.2H) B cells were significantly increased in patients compared to healthy donors, however DOCK8-deficient patients have significantly decreased proportions of total memory B cells (Fig 4.2H). Interestingly, frequencies of Ig class switched memory B cells were unaffected by DOCK8 deficiency (Fig 4.2I).

4.2.4 Impact of HSCT on lymphocyte differentiation in vivo in DOCK8-deficient patients Detailed analysis of immune cells in 18 DOCK8-deficient patients 6 to 43 months post- HSCT (mean: 15 months) revealed most of these defects in lymphocyte differentiation were improved. Specifically, CD3+ T cell proportions were significantly increased due to the recovery of total CD4+ T cells, although overall remained reduced compared to

controls (Fig 4.2A). However, proportions of CD4+ T cells tended to reach normal levels in patients ≥12 months post-HSCT (Fig 4.4A). The loss of naïve CD8+ T cells and + skewing to a CD8 TEM phenotype in untransplanted patients improved following HSCT (Fig 4.2B). While at the level of the total cohort these values were still significantly different to healthy controls (Fig 4.2B), they normalised in patients ≥12 months post + + HSCT (Fig 4.4C). CD4 TCM cells and Tregs were normalised after HSCT, but CD4 TEM cells persisted at an increased frequency at the expense of naïve CD4+ T cells (Fig 4.2C; not shown) until ≥23 months post HSCT (Fig 4.4D). Proportions of  and  T cells were promptly re-established by HSCT (Fig 4.2D; Fig 4.4E); MAIT cells in transplanted DOCK8-deficient patients were significantly increased compared to pre-HSCT levels but remained significantly reduced relative to healthy donors in patients irrespective of time of analysis post-HSCT (Fig 4.2E; Fig 4.4F). Strikingly, proportions of NKT cells remained unchanged in transplanted DOCK8-deficient, being significantly decreased at all times post-HSCT (Fig 4.2G, Fig 4.4H).

HSCT corrected the increased frequencies of total (Fig 4.2A) and naïve (Fig 4.2H) B cells in DOCK8-deficient patients. Importantly, memory B cells were significantly increased post-transplant compared to untransplanted patients but remained decreased compared to healthy donors (Fig 4.2H). The kinetics of the concurrent decline in proportions of transitional and increases in naïve and memory B cells in transplanted DOCK8-deficient patients (Fig 4.4B) is reminiscent of the temporal reappearance of these B-cell subsets in individuals undergoing HSCT for haematological malignancies 155,156. Thus, DOCK8- deficiency affects the generation and differentiation of a wide range of lymphocytes and HSCT improves most defects, but some improvements are likely to be more time- dependent.

Figure 4.2: Effect of HSCT on lymphocyte phenotype and differentiation in DOCK8- deficient patients PBMCs from healthy donors (n=22-24), untransplanted DOCK8-deficient patients (n=7-9), or DOCK8-deficient patients following HSCT (DOCK8 pHSCT) (n=18-20) were labelled with mAbs against CD3, CD4, CD8, CD20, CD56, CD45RA, CCR7, CD10, CD27, IgD, IgM, TCRαβ, TCRγδ, TCRVa24, TCRVb11, CD161 and TCR Va7.2. Proportions of (A) CD3+ cells, CD4+ T cells (CD3+CD4+), CD8+ T cells (CD3+CD8+), B cells (CD20+); (B) CD8+ naïve (CD45RA+CCR7+), central memory (TCM; CD45RA-CCR7+), effector memory (TEM; CD45RA- CCR7-) and CD45RA+ revertant memory (TEMRA; CD45RA+CCR7-) cells; (C) CD4+ naïve, TCM and TEM cell subsets; (D) αβ and γδ TCR T cells; (E) MAIT cells (CD3+TCRVa7.2+CD161+); (F) NK cells (CD3-CD56+); (G) NKT cells (CD3+TCRVa24+Vb11+); (H) transitional (CD20+CD10+CD27-), naive (CD20+CD10-CD27-) and memory (CD20+CD10-CD27+) B cell subsets; and (I) Ig class-switched memory (CD20+CD27+ IgD-IgM-) B cells were then determined by flow cytometric analysis. Contour plots depict one representative normal donor and one DOCK8-deficient patient pre- and post-HSCT. Data are the mean ± SEM. Statistics performed using Prism unpaired t-test with Welch’s correction *<0.05, **<0.01, ***<0.005, ****<0.001. 73

4.2.5 Memory T cells in DOCK8-deficient patients exhibit signs of exhaustion, some of which persist after HSCT Memory T cells from DOCK8-deficient patients exhibit phenotypic features of chronic activation or exhaustion/senescence 65,138, which likely impedes their effector function 157,158 + + . Thus, frequencies of DOCK8-deficient memory CD4 and CD8 TEM cells expressing CD57 and PD1 were significantly increased compared to healthy volunteers (Fig 4.3A, B). This was coupled with significantly decreased expression of CD127, CD27 + + and/or CD28 on DOCK8-deficient CD4 memory and CD8 TEM cells (Fig 4.3A, B). Proportions of patient memory CD4+ T cells expressing PD1 and CD127 normalised post- HSCT and CD57 was significantly decreased following HSCT but continued to exceed that of healthy volunteers (Fig 4.3A). CD27 expression on patient memory CD4+ T cells improved post-HSCT but remained less than that on memory CD4+ T cells from healthy + volunteers (Fig 4.3A). Similar phenotypic changes were observed for CD8 TEM cells, with HSCT normalising PD1, CD28, and CD127 and partially correcting CD57 expression (Fig 4.3B). Taken together, HSCT partially restores the exhausted/senescent phenotype of CD4+ and CD8+ T cells in DOCK8-deficient patients.

Figure 4.3: DOCK8-deficient memory CD4+ and CD8+ T cells exhibit signs of exhaustion which decline after HSCT PBMCs from healthy donors (n=17-22) and DOCK8-deficient patients either before (n=6-7) or after (n=16-19) HSCT were labelled with mAbs against CD4, CD8, CD45RA, CCR7, CD127, CD27, CD28, PD1 and CD57. Co-expression of CD127 and PD1, CD27 and CD57, and CD28 and CD57 by (A) memory CD4+ T cells (CD4+CD8-CD45RA- ± + + - - - CCR7 ) or (B) TEM CD8 T cells (CD8 CD4 CD45RA CCR7 ) was determined. Contour plots are representative of one healthy donor and the same DOCK8-deficient patient assessed before and 9-months following HSCT. The graphs show the mean ± SEM. Statistics performed using Prism unpaired t-test with Welch’s correction *<0.05, **<0.01, ***<0.005, ****<0.001.

Figure 4.4: Analysis of lymphocyte phenotype and differentiation stratified by time post HSCT PBMCs from healthy donors (n=22-24), untransplanted DOCK8-deficient patients (n=7-9), or DOCK8-deficient patients at 0-11 (n=10), 12-22 (n=6) or 23-43 months (n=3-4) following HSCT were labelled with mAbs against CD3, CD4, CD8, CD20, CD45RA, CCR7, CD10, CD27, CD28, CD56, CD57, CD127, PD1, TCRαβ, TCRγδ, TCRVa24, TCRVb11, CD161 and TCR Va7.2. Proportions of (A) CD3+ cells, CD4+ T cells (CD3+CD4+), CD8+ T cells (CD3+CD8+), B cells (CD20+); (B) transitional, naïve and memory B cell subsets; (C) CD4+ naïve, TCM and TEM cell subsets; (D) CD8+ naïve, TCM, TEM and TEMRA cell subsets; (E) αβ and γδ TCR+ T cells; (F) MAIT cells; (G) NK cells; (H) NKT cells; (I) memory CD4+ T cell (CD4+CD8-CD45RA-CCR7±) and (J) CD8+ TEM cell expression of PD1, CD57, CD27, CD28 and CD127 were then determined by flow cytometric analysis. Data are the mean ± SEM. Statistics performed using Prism unpaired t- test with Welch’s correction *<0.05, **<0.01, ***<0.005, ****<0.001.

4.2.6 Defects in proliferation, acquisition of cytotoxic effector function and cytokine secretion by CD8+ T cells in DOCK8-deficient patients are improved by HSCT While proliferative, cytoskeletal and survival defects of CD8+ T cells have been proposed to underlie impaired anti-viral immunity in DOCK8 deficiency 65,88, the role of DOCK8 in CD8+ T cell responses has only been assessed in a few patients 41. To investigate CD8+ T cell dysfunction due to DOCK8 deficiency, and the effect of HSCT on these functions, CD8+ T cells were labelled with CFSE and cultured for five days with anti- CD2/CD3/CD28 mAbs (TAE beads). Analysis of CFSE dilution revealed reduced proliferation of DOCK8-deficient CD8+ T cells (Fig 4.5A, left panels) 65. While exogenous IL-2 improved proliferation, the extent of division of DOCK8-deficient CD8+ T cells remained diminished compared to healthy control CD8+ T cells (Fig 4.5A, right panels).

We next measured acquisition of a cytotoxic phenotype following in vitro activation. CD107a expression, an indicator of degranulation in CD8+ T cells 159, and Granzyme A/B secretion (Fig 4.5B, C) were significantly decreased for DOCK8-deficient CD8+ T cells compared to healthy volunteers. DOCK8-deficient CD8+ T cells also exhibited generalised activation defects, with significantly reduced expression of CD25 and CD95 following in vitro stimulation (Fig 4.5D, E). IL-2, IFNγ and TNFα secretion by DOCK8- deficient CD8+ T cells was also significantly less than healthy controls (Fig 4.5F). While exogenous IL-2 increased secretion of granzymes (Fig 4.5C), TNFα and IFNγ (Fig 4.5G) and CD25 expression (Fig 4.5D) by CD8+ T cells, the overall response of IL-2 treated DOCK8-deficient CD8+ T cells remained significantly less than healthy control CD8+ T cells.

Post-HSCT, CD8+ T cells from DOCK8-deficient patients proliferated as well as those from healthy controls even without addition of IL-2 (Fig 4.5A). Furthermore, degranulation (Fig 4.5B), Granzyme B expression (Fig 4.5C), CD25 induction (Fig 4.5D), and IFN, TNF, and IL-2 secretion (Fig 4.5F) were all restored to normal levels. However, Granzyme A secretion (Fig 4.5C) and CD95 expression (Fig 4.5E) remained significantly decreased. Thus, multiple effector functions of CD8+ T cells are severely compromised in DOCK8-deficient patients, but HSCT largely restores functionality to levels similar to healthy controls.

Figure 4.5: HSCT overcomes CD8+ T cell functional defects due to DOCK8- deficiency CD8+ T cells were sorted from the peripheral blood of healthy donors (n=18-26), untransplanted DOCK8-deficient patients (n=2-7), or DOCK8-deficient patients following HSCT (DOCK8 pHSCT) (n=13-18), labelled with CFSE and then cultured with TAE (anti-CD2/CD3/CD28) beads in the absence or presence of IL-2. After five days, culture supernatants were collected, cells were harvested and then restimulated with PMA/ionomycin for six hours with Brefeldin A, monensin and anti-CD107a mAb being added after 1 hr. (A) The frequency of cells in each division was determined by dilution of CFSE. (B) Expression of CD107a and (C) secretion of Granzyme A and Granzyme B were determined by flow cytometry and cytometric bead arrays, respectively. (D) Surface expression of CD25 and (E) CD95 was determined by flow cytometry. (F, G) Secretion of IFNγ, TNFα and IL-2 were determined by cytometric bead arrays. Histograms in (A), (B), (D) and (E) are representative of one healthy donor and one paired DOCK8-defcient patient pre- and post-HSCT. Data represent the mean ± SEM. Statistics performed using Prism unpaired t-test with Welch’s correction *<0.05, **<0.01, ***<0.005, ****<0.001.

4.2.7 Dysregulated cytokine production by CD4+ T cells in DOCK8-deficient patients is normalised by HSCT Cytokine production by in vitro activated memory CD4+ T cells provides information about their differentiation in vivo 94. Memory CD4+ T cells from DOCK8-deficient patients showed significantly reduced production of Th1 (IFNγ/TNFα) and Th17 (IL- 17A/IL-17F), but increased Th2 (IL-4, IL-5, IL-13), cytokines compared to healthy controls (Fig 4.6A, Fig 4.7A)54,138.

The defect in generating Th17-type cells in vivo in DOCK8-deficient patients was intrinsic, revealed by impaired induction of IL-17A/F by DOCK8-deficient naïve CD4+ T cells cultured under Th17-polarising conditions (Fig 4.6B). Th2 cytokines (IL-5, IL- 13) were normal (Fig 4.6B). IFN expression by DOCK8-deficient naïve CD4+ T cells under Th1 conditions was intact (Fig 4.7B), despite reduced secretion (Fig 4.6B), thus suggesting an extrinsic defect underlies poor generation of Th1 cells in DOCK8-deficient 138 patients in vivo .

Consistent with our previous work, both DOCK8-deficient naïve (Fig 4.6B) and memory (Fig 4.6A) CD4+ T cells showed defects in proliferation compared to normal healthy controls cells 65,138. Since this could impact cytokine production 160, we investigated IFNγ expression by memory and Th1-stimulated naïve CD4+ T cells in the context of cell division. IFNγ production was decreased across all divisions and hence did not result from proliferative defects (Fig 4.7C, D).

Given the implication of the Th1/Th2 axis in allergy 161, we further explored Th1 and Th2 cytokine production by memory CD4+ T cells at the single cell level. We used intracellular staining to calculate the ratio of cells producing Th2 (IL-4/IL-13) vs Th1 (IFNγ/TNFα) cytokines. DOCK8-deficient memory CD4+ T cells showed a significantly increased Th2:Th1 ratio compared to controls (Fig 4.6C). By representing each donor and patient on a plot showing proportions of Th2 vs Th1 cells, DOCK8-deficient memory CD4+ T cells formed a distinct cluster away from control memory cells (Fig 4.6D). This also revealed that the perturbed Th2:Th1 ratio in each patient resulted from increased Th2 and corresponding decreased Th1 cytokine production, further supporting skewed differentiation in vivo.

HSCT greatly improved CD4+ T cell function in vivo in DOCK8-deficient patients. First, proliferation of naïve and memory CD4+ T cells from transplanted patients was 79

comparable to controls (Fig 4.6A, B). Second, Th1 cytokine production (Fig 4.6A) by memory CD4+ T cells was recovered, while there was significantly increased Th17 and decreased Th2 cytokines produced by these cells (Fig 4.6A, Fig 4.7A). CD4+CD45RA- CXCR3-CCR6+ Th17 cells 94 also increased post-HSCT (pre: 4.3 ± 2.1% [n=8]; post: 10.4 ± 6.1% [n=17]). Third, naïve CD4+ T cells from DOCK8-deficient patients post- HSCT produced normal levels of Th1, Th2, and Th17 cytokines following appropriate polarization (Fig 4.6B, Fig 4.7A). Fourth, the Th2:Th1 cytokine ratio was normalised post-HSCT (Fig 4.6C). This was due to co-incident decreases in Th2 and increases in Th1 cells, evidenced as a population clustered between DOCK8-deficient patients and controls (Fig 4.6D). Thus, CD4+ T cell differentiation defects in DOCK8-deficient patients are restored to normal or near-normal levels by HSCT.

4.2.8 Defective production of IL-21 by DOCK8-deficient CD4+ T cells IL-21 potently induces B cell activation, differentiation and Ab production 162. IL-21 production by DOCK8-deficient memory CD4+ T cells was significantly decreased compared to healthy donors (Fig 4.6E). This defect was also cell intrinsic since IL-21 induction in naïve DOCK8-deficient CD4+ T cells in vitro was also significantly impaired (Fig 4.6F). Reduced IL-21 production by DOCK8-deficient naïve CD4+ T cells was not due to diminished proliferation as fewer IL-21+ cells were detected across all divisions measured compared to those from healthy donors (Fig 4.7D). Strikingly, following HSCT, production of IL-21 (Fig 4.6E) by memory CD4+ T cells in DOCK8-deficient patients was restored, while IL-21 induction in naïve CD4+ T cells was significantly increased (Fig 4.6F).

Figure 4.6: Dysregulated cytokine production by DOCK8-deficient CD4+ T cells is greatly improved following HSCT Naïve and memory CD4+ T cells were sort-purified from the peripheral blood of healthy donors (n=7-25), untransplanted DOCK8-deficient patients (n=2-7), or DOCK8-deficient patients following HSCT (DOCK8 pHSCT) (n=6-18). The cells were labelled with CFSE and then cultured under Th0 conditions (TAE beads; naïve and memory), or Th1 (+IL-12), Th2 (+ IL-4) or Th17 (IL-1β, IL-6, IL-21, IL-23, TGF-β, prostaglandin E2) polarising conditions (naïve only) for five days. (A, B) Cells and culture supernatants were harvested to assess proliferation of (CFSE dilution) and cytokine secretion by (A) memory CD4+ T cells of IFNγ, Th2 cytokines (IL4/IL5/IL13) or Th17 cytokines (IL17A/IL17F) and (B) naïve CD4+ T cells of IFNγ, Th2 cytokines (IL5/IL13) or Th17 cytokines (IL17A/IL17F). (C, D) Cells were restimulated with PMA/ionomycin before permeabilisation and intracellular staining to determine proportions of cells expressing Th1 (IFN, TNF) and Th2 (IL-4, IL-13) cytokines. Data are depicted as (C) the ratio of cells expressing Th2 vs Th1 cytokines and (D) the combined % cells from individual donors and patients expressing Th1 (i.e. %IFN+/TNF+/ IFN+TNF+ cells) vs Th2 (i.e. %IL- 4+/IL-13+/IL-4+IL-13+ cells) cytokines. (E, F) Intracellular expression of IL-21 by memory CD4+ T cells (E) and naive CD4+ T cells cultured under Th0 or Th1-polarising conditions (F) cells was measured. Graphs depict mean ± SEM. Statistics performed using Prism unpaired t-test with Welch’s correction *<0.05, **<0.01, ***<0.005, ****<0.001.

Figure 4.7: Analysis of intracellular cytokine expression by DOCK8-deficient CD4+ T cells: effect of cell division Naïve and memory CD4+ T cells were sort-purified from the peripheral blood of healthy donors (n=13-25), untransplanted DOCK8-deficient patients (n=2-7), or DOCK8- deficient patients following HSCT (DOCK8 pHSCT; n=8-18). The cells were labelled with CFSE and then cultured under Th0 (TAE beads; memory) or Th1 (+IL-12) conditions (naïve). After five days, cells were restimulated with PMA/ionomycin before permeabilization and intracellular staining to determine (A) proportions of memory CD4+ T cells expressing Th1 (IFNγ and TNFα) or Th2 cytokines (IL4 and IL13); (B) proportions of naïve CD4+ T cells expressing IFNγ and the proportions of (C) memory or (D) naïve CD4+ T cells expressing IFNγ and IL-21 in each division interval as determined by dilution of CFSE. The data represent the mean ± SEM of cytokine positive cells, or cytokine-expressing cells in each division. Statistics performed using Prism unpaired t- test with Welch’s correction *<0.05, **<0.01, ***<0.005, ****<0.001.

4.2.9 HSCT overcomes the B-cell intrinsic impairment in survival, proliferation and differentiation due to DOCK8 deficiency Many DOCK8-deficient patients have impaired antibody responses to vaccines 41,51,59 (Table 4.1). Dock8 is required in murine B cells to generate germinal centres and long- lived humoral immunity 86, and studies of DOCK8-deficient humans reported reductions in memory B cells 59, compromised in vitro activation of naïve B cells and identified DOCK8 as an adaptor for TLR signalling 59,163,164. To extend these findings, we examined the impact of DOCK8 deficiency on naïve B cell function by assessing Ig secretion in vitro. When stimulated with mimics of T-cell help (CD40L/IL-21), TLR ligands (CpG) and BCR agonists, naïve B cells from healthy volunteers secrete IgM (Fig 4.8A). CD40L/IL-21 or CD40L/CpG/BCR stimulation also induced switching to IgG and IgA (Fig 4.8A), with CD40L/CpG/BCR being less effective than CD40L/IL-21 (Fig 4.8A). DOCK8-deficient naïve B cells secreted significantly lower levels of IgM, with a trend towards decreased IgG and IgA under most in vitro conditions (Fig 4.8A). DOCK8- deficient naïve B cells also exhibited significantly compromised survival and proliferation in vitro compared to control naïve B cells (Fig 4.8B,C). These data demonstrate that DOCK8-deficient B cells are defective in responding to not only TLR- 59,164 and BCR-mediated 163 signals, but also those delivered via CD40 and cytokines. Survival, proliferation and secretion of IgM and IgG by naïve B cells isolated from transplanted DOCK8-deficient patients were largely restored (Fig 4.8B,C). In contrast, IgA secretion by naïve B cells from transplanted patients remained significantly lower than controls (Fig 4.8D). These findings reveal DOCK8-deficiency intrinsically impairs naïve B cell survival, proliferation and differentiation, establishing that DOCK8- dependent signals are elicited in B cells downstream of numerous stimulatory receptors. However, these key functions are almost completely regained following HSCT, thus explaining improved humoral immune responses in DOCK8 deficiency post-HSCT (Table 4.2).

Figure 4.8: B cell functional defects due to DOCK8-deficiency are improved following HSCT

(A) Naïve B cells were sort-purified from healthy donors and DOCK8-deficient patients (n=7) and then cultured with CD40L/IL-21, CD40L/CpG, CpG/BCR agonist (Staphylococcus aureus Cowan I), or CD40L/CpG/BCR. After 11 days, culture supernatants were harvested and levels of secreted IgM, IgG and IgA then determined. Data represent the mean ± SEM. (B-D) Naïve B cells were sort-purified from healthy donors (n=18-23), untransplanted DOCK8-deficient patients (n=5-7), or DOCK8- deficient patients following HSCT (DOCK8 pHSCT) (n=12-15), labelled with CFSE and then cultured with combinations of CD40L, IL-21, CpG and BCR stimulus for five days. After this time, cells and culture supernatants were harvested. (B) Cell number was determined using Calibrite beads. (C) Frequency of cells in each division was determined by CFSE dilution. Histogram plots show CFSE dilution from one representative healthy donor and one paired DOCK8-deficient patient pre and post-HSCT. (D) Ig secretion of IgM, IgG and IgA was determined by ELISAs. Data in each graph represents the mean ± SEM. Statistics performed using Prism unpaired t-test with Welch’s correction *<0.05, **<0.01, ***<0.005, ****<0.001.

4.2.10 Elevated serum IgE and allergen specific IgE levels decrease in a time- dependent manner following HSCT of DOCK8-deficient patients. Most DOCK8-deficient patients have extremely high levels of total and allergen-specific IgE, consistent with severe allergies 40,41,138. Indeed, total and food allergen-specific IgE were increased 50-1000 fold in the DOCK8-deficient patients studied here compared to healthy controls (Table 4.1, Fig 4.9A). However, dust mite-specific IgE was normal or moderately increased in DOCK8-deficiency (Fig 4.9A). After HSCT, all patients showed decreased total and food allergen-specific IgE compared to pre-HSCT levels (Table 4.1, 4.2, Fig 4.9A). One patient had moderately positive dust mite allergen-specific IgE levels pre-HSCT, which was also reduced following HSCT (Fig 4.9A).

HSCT has been reported to have mixed outcomes on allergy in DOCK8 deficiency 79,151,152. We also noted variability in reductions in total and allergen-specific IgE, ranging from 2 to >1000-fold (Fig 4.9A). To investigate this further, we expressed IgE levels following HSCT as a percentage of pre-transplant levels for each patient and as a function of time post-HSCT. For this, we could also study 2 patients longitudinally. This revealed that the magnitude of the reduction in total and specific IgE levels in DOCK8-deficient patients was time-dependent post-transplant (Fig 4.9B). In general, food specific IgE levels declined at a slower rate than total IgE levels (Figure 4.9B). Hence, elevated IgE levels in DOCK8-deficient patients are decreased with HSCT and tend to normalise over time.

Figure 4.9: Elevated total and allergen specific serum IgE levels decrease in a time- dependent manner in DOCK8-deficient patients following HSCT. (A) Serum from healthy donors (n=5) and DOCK8-deficient patients (n=4-18) collected before and after HSCT were analysed for concentrations of total IgE and IgE specific for staple foods and dust mites. (B) Data in (A) expressed as a percentage of pre-transplant levels of total and allergen-specific IgE for each patient and plotted against the time post- HSCT. Points joined by a line are from the same patients (#16, #18; Table 4.1) assayed at different times post-HSCT. Dotted line indicates average of healthy control values as % of average of patient pre-transplant values.

4.2.11 Impact of HSCT on lymphocyte reconstitution in other PID patients It was possible that some of the impairments in lymphocyte differentiation in transplanted DOCK8-deficient patients were unique to DOCK8 deficiency or a general consequence of HSCT in PID. To differentiate between these possibilities, we examined lymphocytes in 8 additional PID patients who underwent HSCT due to loss-of-function mutations in UNC13D (n=2), STK4 (n=2), CYBB (n=2), CD40LG (n=1) or SH2D1A (n=1) (Table 4.3)165-167. Time post HSCT ranged from 5-84 (mean 44.5) months (Table 4.3). Proportions of CD3+, CD4+ and CD8+ T cells in these transplanted PID patients were comparable to healthy controls, while B cells were increased (Fig 4.10A). Memory B cells were reduced following HSCT (Fig 4.10B) but differentiation of CD4+ and CD8+ T cells was normal (Fig 4.10C, D). Expression of exhaustion/senescence markers on CD4+ memory T cells in these transplanted PID patients was comparable to CD4+ memory T + cells from healthy donors (Fig 4.10E) however patient CD8 TEM cells showed decreased expression of CD28 and CD127 relative to controls (Fig 4.10F). Interestingly, these PID patients exhibited decreased frequencies of MAIT (Fig 4.10G) and NKT cells (Fig 4.10H) post-HSCT.

We investigated STK4-deficient patients in more detail, as STK4 deficiency shares clinical features with DOCK8 deficiency, including susceptibility to viral infections, elevated serum IgE, naïve T cell lymphopenia and reduced memory B cells 137,165. Naïve B cells from STK4-deficient patients exhibit impaired production of IgM, IgG and IgA following in vitro stimulation with CD40L/IL21 165(Fig 4.10I). However, Ig production by naïve B cells isolated from these patients and activated in vitro was normalised following HSCT (Fig 4.10I). Thus, analysis of patients with DOCK8 deficiency or other inborn errors of immunity allowed us to identify specific and general outcomes for reconstitution of lymphocyte development, differentiation and function following HSCT.

cells;

Cys=cyclosporin; Cys=cyclosporin;

m=Campath; CD34 sel=CD34 selection; CB=cord blood; Cy=cyclophosphamide; CD34 blood; selection; m=Campath; Cy=cyclophosphamide; CB=cord sel=CD34

HSCTpatients

MRD=Matched Related Donor; Mel=Melphalan; MUD=Matched Unrelated Donor; MMF=mycophenylate blood stem PBSCs=peripheral mofetil; MMF=mycophenylate MUD=Matched Unrelated Donor; Mel=Melphalan; Donor; Related MRD=Matched

Treo=Treosulphan.

Thymocyte Globulin; BM=bone marrow; Bu=Busulfan; Ca BM=boneGlobulin; marrow; Thymocyte

ATG=Anti Flu=Fludarabine; Pred=Prednisone; Table 4.3: Clinical details of post Clinical Tableother details 4.3:

Figure 4.10: Lymphocyte phenotype and differentiation in PID patients pHSCT PBMCs from healthy donors (n=5-7) or patients with mutations in UNC13D (n=2), STK4 (n=2), CYBB (n=2), CD40LG (n=1) or SH2D1A (n=1) who had previously undergone HSCT (PID pHSCT) (n=6-8) were labelled with mAbs against CD3, CD4, CD8, CD20, CD45RA, CCR7, CD10, CD27, CD28, CD57, CD127, PD1, TCRV24, TCRV11, CD161 and TCR V7.2 (see Figure 1 legend). Proportions of (A) CD3+ cells, CD4+ T cells (CD3+CD4+), CD8+ T cells (CD3+CD8+), B cells (CD20+); (B) transitional, naïve and memory B cell subsets; (C) + + CD4 naïve, TCM and TEM cell subsets; (D) CD8 naïve, TCM, TEM and TEMRA cell subsets; (E) + + CD4 memory and (F) CD8 TEM cells expressing PD1, CD57, CD27, CD27 and CD127; (G) MAIT cells and (H) NKT cells were then determined by flow cytometric analysis. (I) Naive B cells were sorted from healthy donors (n=3), STK4-deficient patients (n=1), or STK4- deficient patients following HSCT (STK4 pHSCT) (n=2) and cultured with CD40L+IL21 for five days. After this time, culture supernatants were harvested and Ig secretion of IgM, IgG and IgA was determined by ELISAs. Points joined by lines represent data from the same patient before and after HSCT. Data are the mean ± SEM. Statistics performed using Prism unpaired t-test *<0.05, **<0.01, ***<0.005, ****<0.001. 89

Figure 4.11: Changes observed in paired pre- and post-transplant DOCK8-deficient patients are consistent with those seen at the cohort level. PBMCs from healthy donors (red columns) and DOCK8-deficient patients before (blue dots) or following HSCT (DOCK8 pHSCT) (green dots) were labelled with mAbs against CD3, CD4, CD8, CD20, CD45RA, CCR7, CD10, CD27, PD1, CD57, CD27, CD28 and CD127. Proportions of: (A) CD3+ cells, CD4+ T cells (CD3+CD4+), CD8+ T cells (CD3+CD8+), B cells (CD20+); (B) + + transitional, naive and memory B cell subsets; (C) CD4 and (D) CD8 naïve, TCM, TEM and TEMRA + + cell subsets; and (E) memory CD4 T cells and (F) TEM CD8 T cells expressing PD1, CD57, CD27, CD28 and CD127 were determined. (G) Memory CD4+ T cells were sort-purified from healthy donors, untransplanted DOCK8-deficient patients, or DOCK8-deficient patients following HSCT, and cultured with TAE beads. After five days, the proportions of cells expressing Th1 (IFNγ, TNFα) and Th2 (IL-4, IL-13) cytokines, and the Th2/Th1 ratio were determined. (H) CD8+ T cells were sorted from healthy donors, untransplanted DOCK8-deficient patients, or DOCK8-deficient patients following HSCT and cultured with TAE beads. Secretion of IFNγ, TNFα, and IL-2 was determined after 5 days by cytometric bead arrays. Data presented is from healthy controls (n=21-25) and patients with both a pre- and post-transplant sample (n=4- 6), with data from individual patients joined by a line. Statistics performed using Prism paired t- test or unpaired t-test with Welch’s correction *<0.05, **<0.01, ***<0.005, ****<0.001.

4.3 Discussion The study of cellular defects due by DOCK8 deficiency is essential to reveal mechanisms underlying the constellation of clinical features in these patients. Previous studies showed DOCK8 deficiency compromises memory B cell generation, NK cell cytotoxicity, NKT cell development, CD8+ T cell differentiation, CD4+ T cell cytokine production and Treg function 54,59,65-68,71,72,138. We have confirmed and substantially extended these findings to produce a comprehensive catalogue of phenotypic and functional defects in DOCK8- deficient lymphocytes, including impaired CD8+ T cell cytotoxicity, reduced B-cell survival and proliferation, poor induction of Tfh-type cells, and defective generation of MAIT cells. Collectively, these defects explain poor control of pathogen infections, impaired humoral immunity and severe atopic disease in individuals with DOCK8 mutations.

HSCT is the only treatment capable of curing DOCK8 deficiency. Cellular mechanisms underlying clinical improvements post-HSCT have not previously been established. DOCK8-deficient patients assessed in our study reported poor vaccine responses prior to HSCT (Table 4.1), but normal responses post-HSCT (Table 4.2). This correlated with functional improvements in naïve B cell survival, proliferation and differentiation in vitro and a significant, albeit incomplete, increase in memory B cell formation in vivo. CD19 expression is reduced on DOCK8-deficient human B cells, and it has been proposed that this contributes to poor B-cell responses in vivo 163. We confirmed this reduction in CD19 expression and noted that HSCT increased CD19 on patient B cells (not shown); this may also facilitate improved B-cell behaviour in DOCK8-deficient patients post-HSCT. Due to the kinetics of memory B cell generation post-HSCT 155,156 and the increases in serum IgG and IgA during the first years of life 168, reconstitution of the memory compartment in DOCK8-deficient patients is expected to increase with time following HSCT. Indeed, we observed a normal frequency of memory B cells is achieved in patients who are ≥23 months post transplant (Fig 4.7B). This was also observed for the non-DOCK8 PID patients examined inasmuch that the patients who were tested ≥60 months post-HSCT had the highest proportions of memory B cells in peripheral blood (Fig 4.10B). It is likely that the increased ability of naïve and memory CD4+ T cells from HSCT- DOCK8- deficient patients to produce IL-21 would also contribute to improved B-cell function, given the potency of IL-21 in inducing human B-cell proliferation and differentiation into plasma cells (PCs)162. 91

The prevalence and severity of infections in DOCK8-deficient patients pre-transplant was dramatically reduced following HSCT. Many cellular changes are likely responsible for improved host defence, including increased frequencies, proliferation and normalised differentiation of CD4+ T cells. Increased Th1 and Th17 cytokines would contribute to effective control of bacterial, viral and fungal infections in transplanted DOCK8-deficient patients. Furthermore, restored differentiation and production of cytokines and cytolytic mediators by DOCK8-deficient CD8+ T cells would reduce viral infections post-HSCT. The mild recovery of MAIT cells may also be important in host defence in DOCK8 deficiency post-HSCT 169. Lastly, resolution of severe skin inflammation post-HSCT could in part be attributed to decreased Th2 skewing and concordant reductions in total and allergen-specific IgE. IL-21R-deficient mice and humans have increased serum IgE 170,171, and administering IL-21 to Dock8-deficient mice alleviates disease in an allergic asthma model 172. Thus, increased IL-21 production by CD4+ T cells in transplanted DOCK8-deficient patients may not only contribute to improved humoral immunity, but also mediate improvements in allergy and reductions in serum IgE. Similarly, since CpG activation suppresses IgE production by CD40L/IL-4-stimulated human naïve B cells 164, restored TLR signalling in B cells post-HSCT may also contribute to reductions in serum IgE in DOCK8-deficient patients.

Malignancies occur in 10-15% of DOCK8-deficient patients. Indeed, ~30% of patients studied here reported malignancies prior to HSCT. Incidence of malignancies in transplanted patients has not been described, likely due to insufficient follow-up times. However, the lack of malignancy amongst patients in our study and other reports 142,148,149 suggests that, by restoring CD8+ T cell cytotoxic function, HSCT also protects against tumorigenesis in DOCK8 deficiency.

Despite effective control of infections due to improved cellular function in patients’ post- HSCT, memory T cells still exhibited modest signs of chronic activation/exhaustion. This could reflect reactivation of latent viruses. Indeed, EBV reactivation occurs in ~20% of HSCT patients 173, while CMV reactivation following HSCT is associated with exhausted CD8+ T cells with fewer CD27-expressing and more PD1-expressing cells found in patients with reactivation compared to patients who did not experience viral reactivation post-HSCT174,175. Further evidence that CMV reactivation is associated with exhaustion + is that patients that experienced CMV reactivation also had expanded CD8 TEM and 176 TEMRA populations , which our patients also exhibited post transplant. Interestingly, one 92

patient we studied had CMV reactivation post-HSCT (DOCK8 #15), while another had an episode of shingles due to VZV reactivation (DOCK8 #18) following transplantation, but no further complications. Additionally, exhaustion of CD4+ and CD8+ memory T cells in DOCK8-deficient patients was reduced ≥12 months post-HSCT (Fig 4.7I, J), suggesting the effects of viral reactivation are short-lived due to functional reconstitution of immune cells.

HSCT has been reported to have variable outcomes for allergic disease in DOCK8- deficient patients. Thus, it has been reported that HSCT lead to resolution 139,142,177, improvement 140,149 or persistence 146,152,178 of allergies in these patients. However, a recent study of 56 transplanted DOCK8-deficient patients found that food allergies resolved or improved in 61% of cases 79. Furthermore, as a substantial proportion of this cohort avoided allergen exposure post-HSCT, the actual level of allergy improvement was nearly 80% (34/43 patients)79, suggesting HSCT positively impacts allergic disease in DOCK8 deficiency. Our findings of decreased production of Th2 cytokines by memory CD4+ T cells, and corresponding reductions in serum levels of allergen-specific IgE in DOCK8-deficient patients with time are consistent with decreased allergic disease in cohorts of transplanted DOCK8 deficient patients. Furthermore, the kinetics of reduction in total and allergen-specific IgE could explain conflicting results regarding the impact of HSCT on allergic disease in DOCK8-deficiency. Long-lived host PCs in bone marrow may produce IgE in DOCK8-deficient patients post-HSCT 145,149,152. They may survive conditioning but be displaced from survival niches by newly-generated PCs, resulting in apoptosis 179. Hence, the time-dependent decrease of allergen-specific IgE in transplanted patients could result from the gradual turnover of IgE+ PCs by the reconstituted donor immune system. Continued longitudinal assessment of transplanted patients will establish the long-term efficacy of HSCT on allergic disease in DOCK8-deficient patients.

Successful immune reconstitution following HSCT often depends on high donor chimerism 180. Previous reports detected >95% chimerism in CD4+ and CD8+ T cells 139,141,145,146, but varying (0-100%) levels for B cells 143,146, in HSCT DOCK8-deficient patients. Despite this, DOCK8-deficient patients with mixed B cell chimerism showed improvements comparable to patients with complete donor chimerism 145, indicating that high donor chimerism is not the sole determinant of restored immune function following HSCT of these patients. Interestingly, some DOCK8-deficient patients undergo somatic reversion and re-express functional DOCK8 protein in leukocytes 53. Interestingly, 93

DOCK8 was highly expressed in memory T cells from somatically reverted patients, but in low proportions of naïve CD4+ T cells, B cells and NK cells 53. Despite having DOCK8+ lymphocytes, clinical improvements in DOCK8-revertant patients were mild, with all continuing to have significant disease and some undergoing HSCT 53. This may be because the revertant population continues to express an exhausted/senescent phenotype 53, which may compromise their response to antigenic stimuli. These data also suggest that DOCK8 is required in most lymphocyte populations, including naïve T and B cells, to achieve significant clinical improvements. Indeed, the finding that the TCR repertoire of DOCK8-revertant cells is dominated by a few V clonotypes 53 infers that a restricted repertoire is not conducive with eliciting robust immune responses in DOCK8- revertant patients. We detected >90% donor chimerism and normal levels of DOCK8 expression in all transplanted patients. This likely explains dramatic differences in disease outcome in DOCK8-deficient patients who have undergone HSCT and those with somatic reversion in only some lymphoid cells.

A caveat of our study was that we could not analyse pre- and post-HSCT samples from all DOCK8-deficient patients. However, there were several matched patients. Examination of these individual patients pre- and post-transplant revealed identical results to those from all untransplanted and transplanted DOCK8-deficient patients (Fig 4.11). This indicates that our collective results are representative of analysis derived from matched patients.

Comparison of our results with data from other PID patients who underwent HSCT highlights outcomes that are a general consequence of the transplant process and those that are unique to DOCK8-deficiency. The shared characteristic of reduced frequencies of memory B cells across both cohorts of patients was consistent with kinetics of B cell reconstitution post-HSCT in other clinical settings 155. The lack of complete reconstitution of MAIT and NKT cells post-HSCT in both DOCK8-deficient and the other PID patients post-HSCT has similarly been reported in patients undergoing HSCT for haematological malignancies as late as 12-24 months following transplant 181,182. This may result from increased sensitivity of MAIT cells to GVHD prophylactic immunosuppressive drugs 182. These observations for MAIT and NKT cells in DOCK8-deficiency and the other PIDs examined are reminiscent of the persistent deficiency of other populations of innate lymphocytes, including NK cells and ILCs, in X-SCID and JAK/SCID patients decades post-HSCT 183. Thus, an alternative explanation is an inability of precursors of different 94

innate lymphoid cell populations to adequately seed developmental niches following HSCT to enable the reconstitution of these immune cell lineages. In contrast to these findings, DOCK8-deficient patients continued to exhibit greater signs of T-cell exhaustion/senescence post-HSCT, with CD8+ T cells in transplanted DOCK8-deficient patients remaining skewed toward TEM and TEMRA and being enriched for CD57+CD127dim cells.

In conclusion, we detail the numerous adverse effects of DOCK8 deficiency on the differentiation and function of CD8+ and CD4+ T, MAIT, NK and B cells which underlie clinical features of patients with DOCK8 mutations. By demonstrating that effective restoration of key functional defects in adaptive immune cells following HSCT corresponds to improved clinical features, we have identified cellular mechanisms for the clinical efficacy of HSCT as a treatment for DOCK8 deficiency. Comparison to other PID patients who have also undergone HSCT identified unique and general consequences of HSCT in PIDs. Combining our defined readouts of cellular function with clinical features post-HSCT may facilitate predicting long-term outcomes for DOCK8-deficient patients undergoing potentially curative HSCT. Such an approach has been applied to SCID patients, with several clinical, cellular and functional improvements being established as predictors of successful outcomes of HSCT 131,184,185.

Collectively, our study underscores the value and importance of determining the impact of monogenic mutations on immune cell function and applying these findings following therapeutic interventions in order to ensure optimal patient management and outcomes.

Chapter 5: Investigation of reversion in a DOCK8-deficient patient 5.1 Introduction Somatic reversion is when a molecular change occurs in cells which restores the expression or function of a protein which was previously affected by a pathogenic mutation and is considered “natural gene therapy”. It was first reported in primary immunodeficiency patients, specifically those with adenosine deaminase deficiency (ADA) or X-linked severe combined immunodeficiency (X-SCID), who were identified by their significantly milder clinical phenotype than individuals with mutations in ADA or IL2RG who presented with classic SCID186-188.

Similarly, somatic reversion of pathogenic mutations in DOCK8 was discovered while investigating variability in disease presentation in a cohort of DOCK8-deficient patients53. Remarkably, a significant proportion of these patients (50%; 17/34) were found to display some degree of somatic reversion53. This was more frequent in patients with compound heterozygosity, that is, when the two DOCK8 alleles contain different mutations. This is because heterozygosity allows for gene crossover or conversion events within the area harbouring the affected allele with a region of the unaffected or wild type allele. However, patients who achieved gene repair through mutation of the primary site or mutations at a secondary site have also been identified53.

Given that DOCK8 deficiency is an autosomal recessive condition and heterozygous carriers are healthy, repair of one allele could be expected to be sufficient to correct lymphocyte function and hence improve clinical disease. Comparison of cohorts of DOCK8-deficient patients with and without somatic reversions revealed that those with reversions had longer survival rates and were less likely to experience food allergies53. However, some individuals with somatic reversion were found to continue to experience severe food allergies189. Furthermore, despite the occurrence of somatic reversion, all patients still displayed high susceptibility to infections, similar to patients with unrepaired DOCK8 mutations, indicating that somatic reversion did not necessarily resolve disease. Indeed, several DOCK8-deficient patients (6/17) still required HSCT and one died despite a high proportion of their lymphocytes exhibiting somatic reversion53.

The immune phenotype that characterises DOCK8-deficiency varies in patients who harbour somatic reversion in DOCK8. For example, one patient presented with low proportions of memory B cells but normal numbers of total B and T cells189 while another

exhibited decreased total T cells and decreased CD4+ T cells, with skewing to the memory compartment190, which is more reminiscent of DOCK8-deficient patients who have not undergone reversion (see Chapter 4). Adding to this complexity, comparison of the cellular function between DOCK8-deficient patients and patients with DOCK8 reversion has not been undertaken.

The extent of DOCK8 reversion in distinct subsets of lymphocytes from DOCK8- deficient patients is also highly variable. For instance, CD8+ T cells consistently demonstrate the highest levels of reversion, with the proportion of all CD8+ T cells expressing detectable levels of DOCK8 being as high as 96% based on protein staining and flow cytometric analysis53. CD4+ T cells also show significant levels of reversion, but at levels lower than CD8+ T cells in almost all patients (on average 78% of CD8+ T cell levels)53,190. Several DOCK8-deficient patients display significant levels of DOCK8 expression due to somatic reversion in NK cells, including one patient with >95% reverted NK cells189. In contrast, in all patients examined, B cells showed much lower levels of reversion than these other lymphocyte populations, with one patient having 1% DOCK8 expression in B cells despite >90% of their CD3+ T cells expressing DOCK853. Within the T cell compartment, levels of DOCK8 reversion increased with the degree of differentiation189 which was most strikingly observed in CD4+ T cells with the frequency of reverted memory cells being on average seven times higher than that seen in naive CD4+ T cells (i.e. 7.6% of naïve cells vs 53.6% of memory cells)53.

During the course of this project, we identified a patient with compound heterozygous variants in DOCK8 who, despite recurrent infectious complications and allergy early in life, now presents as clinically healthy. We conducted extensive molecular, biochemical and cellular analyses to understand disease aetiology and pathogenesis in this individual. This led to the discovery of somatic reversion in one of the DOCK8 variants in this patient. This chapter details the identification and impact of somatic reversion in DOCK8 in this patient.

5.2 Results 5.2.1 Clinical features and identification of bi-allelic DOCK8 variants The patient is a 23 year old female with a history of recurrent bacterial, viral and fungal infections including 6 pneumonia episodes, recurrent ear infections, fungal nail infections at ages 4-6 years and oral thrush at age 15 years. The patient also exhibited elevated levels of serum IgE with pollen-specific IgE identified and had a milk allergy during infancy but serum levels of other Igs were normal. However, the patient has experienced no major infections since mid-adolescence (~7 years ago). Whole genome sequencing (WGS) was performed when the patient was 21 years of age and she was subsequently found to have compound heterozygous variants in DOCK8. The details of this are described below.

WGS of the patient identified two variants in DOCK8: A 2 base pair deletion in exon 8 (g.325693_325694del) and a missense mutation in the splice acceptor site of exon 20 (g.376975A>G; Fig 5.1A). The deletion mutation in exon 8 was found in 2/251 000 alleles recorded in the genome aggregation database (gnomAD). The deletion mutation was predicted to cause a truncating frameshift mutation (c.850_851delCT; p.Leu284ValfsTer10) with a combined annotation dependent depletion (CADD) score, a tool for scoring the deleteriousness of a variant, of 35, classifying it as one of the most deleterious variants to have been identified in DOCK8. This is seen in Fig 5.1B where each point represents an identified variant and is positioned based on their frequency in gnomAD (MAF) and CADD score. Additionally, this g.325693_325694del variant has been previously found in a homozygous state in 2 siblings with DOCK8 deficiency68. In contrast to this 2 bp deletion variant, the missense mutation in exon 20 was unique to this individual as it was not present in gnomAD and had not been previously reported. This variant (c.2206A>G) was predicted to abrogate the splice acceptor site for exon 20 and prevent the splicing that occurs at that site (CADD score 25.2). This can be seen in Fig 5.1C with the green bars, indicating splicing initiation predicted by four different programs, present in the reference sequence at the start of exon 20 (indicated by the blue box) but absent in the mutated sequence.

The c.850_851delCT deletion variant was also present in the patients’ mother, thereby revealing maternal inheritance of this allele. In contrast, the c.2206-2A>G missense mutation was absent from the germline of both parents, indicating this to be a de novo variant. Sequencing of genomic DNA isolated from a buccal swab from the patient

confirmed the de novo mutation was present in the germline, rather than being a somatic mutation arising preferentially in blood cells.

Due to the de novo nature of the c.2206-2A>G missense mutation, it was not immediately evident whether this variant was in cis (i.e. on the same chromosome) or in trans (i.e. on the other chromosome) with respect to the maternally-derived c.850_851delCT variant. To determine this, 10x chromium sequencing was employed for long-range sequencing to enable phasing of maternal and paternal SNPs with this variant. This revealed the de novo c.2206-2A>G missense mutation was indeed present on the paternal allele, thereby establishing that both copies of the patient’s DOCK8 gene carry individual deleterious mutations. Thus, this patient was diagnosed with DOCK8-deficiency due to compound heterozygous variants.

Figure 5.1 Detection of DOCK8 mutations and their predicted effects in a compound heterozygous DOCK8-deficient patient Genomic DNA was extracted from patient PBMC. WGS identified 2 variants in DOCK8; g.325693_325694del in exon 8 and g.376975A>G in exon 20. A) CADD score and minor allele frequency (MAF, as determined by gnomAD) were plotted using PopViz for the 2bp deletion variant. Mutation significance cut off (MSC) of 0.001 was applied. B) Predicted effects of the missense mutation on splicing were determined using Alamut interactive biosoftware with green bars indicating splicing initiation and the blue box indicating exon 20. C) 10x chromium sequencing confirmed these mutations to be present in trans i.e. on separate alleles. Black bars depict exons in DOCK8 and white bars are introns.

100

5.2.2 Detection of variable levels of expression of DOCK8 protein in different immune cell subsets from the patient To assess the impact of the germline DOCK8 mutations detected in this patient, we measured DOCK8 expression by intracellular staining and flow cytometry. Unexpectedly, approximately one third of the patient’s lymphocytes were positive for DOCK8 expression (Fig 5.2A). Examining lymphocyte subsets, DOCK8 expression was detected in ~50% of the patient’s CD4+ and CD8+ T cells, but less than 15% of their B cells and NK cells (Fig 5.2 B-E). Even more interestingly, within each lymphocyte compartment, memory cells had larger proportions of DOCK8-expressing cells compared to naïve cells. This was particularly striking for CD4+ T cells and B cells, with there being 4.5-fold more memory CD4+ T cells (64% vs 14%, Fig 5.2B), and ~20-fold more memory B cells (70% vs 3.5%; Fig 5.2E) expressing DOCK8 than corresponding naïve cells. CD8+ T cells also had the highest frequency of DOCK8-expressing cells in the memory compartment, however, unlike CD4+ T cells and B cells, there was also considerable + DOCK8 expression in naïve, but not TEMRA, CD8 T cells (Fig 5.2D).

The epitope in DOCK8 that is recognised by the rabbit mAb used to measure DOCK8 expression by flow cytometry is unknown. Thus, it was possible that DOCK8 detected in immune cells from this patient was a truncated variant resulting from either the indel or splice site mutations. To investigate this, whole cell lysates were prepared from PBMCs from this patient and a healthy donor, as well as B-cell lines from a confirmed DOCK8- deficient patient (c.3733_3734delAG, p.Arg1245GlufsX5; absent DOCK8 by flow cytometry) and a healthy donor as negative and positive controls respectively. The lysates were subjected to SDS-PAGE and Western blot using the rabbit mAb, as well as a mouse mAb known to bind to amino acids 119-277 in DOCK8. This revealed a protein of reduced intensity, but of normal size, corresponding to DOCK8 in the patients’ PBMCs. There were also no unique protein products, that would have resulted from a truncated protein, detected in patients’ PBMCs (Fig 5.3). This suggested that somatic reversion had occurred in this patient, allowing their cells to transcribe and express DOCK8 protein of normal size.

101

Figure 5.2 DOCK8 expression is varied in lymphocyte subsets Patient PBMC were stained for surface expression of CD4, CD8, CD20, CD56, CCR7, CD45RA, CD10 and CD27, fixed and then intracellularly stained for DOCK8. Percentage of cells positive for DOCK8 expression in (A) total lymphocytes, (B) CD3-CD56+ NK cells, (C) Total, CCR7+CD45RA+ naïve and CCR7±CD45RA- memory CD4+ T cells, (D) Total, CCR7+CD45RA+ naïve, CCR7+CD45RA- central memory, CCR7-CD45RA- effector memory and CCR7-CD45RA+ EMRA CD8+ T cells, and (E) Total, CD10+CD27- transitional, CD10-CD27- naive and CD10-CD27+ memory CD20+ B cells was determined by flow cytometry. 102

Figure 5.3 DOCK8 expression in PBMCs by Western blot Protein lysates from PBMCs from a healthy donor and the DOCK8-revertant patient (DOCK8r) as well as B-LCLs derived from a healthy control and DOCK8-deficient patient were electrophoresed and transferred to a membrane. Western blotting was undertaken with (A) rabbit monoclonal anti-human DOCK8 antibody with an unknown binding region to DOCK8 and (B) mouse monoclonal anti-human DOCK8 antibody with a known binding region of amino acids 119-277 of DOCK8.

103

5.2.3 DOCK8 reversion occurs due to repair of the de novo missense mutation To determine the molecular mechanism by which cells from this DOCK8-deficient patient had overcome the original mutations to be able to express DOCK8 protein, lymphocyte subsets were sorted and genomic DNA extracted and used as a template for PCR across the regions where the two DOCK8 mutations are located (i.e. exon 8 and exon 20 splice acceptor site). Amplified products were then sequenced. Due to the 2bp deletion in exon 8 causing a frameshift, this sequence was unable to be properly aligned and analysed, however, the missense mutation at the exon 20 splice acceptor site was able to be aligned and analysed for peak height.

At the exon 20 splice acceptor site, as the patient is heterozygous, two nucleotide peaks were visible and we could observe the peak height of each allele, i.e. wild type (A) and mutated (G), in the different lymphocyte subsets. This indicated that the peak height of the wild type allele (corresponding to the consensus sequence) appeared to be higher, indicating an increased frequency of cells with that allele, in subsets with substantial + + percentages of DOCK8-expressing cells, namely memory CD4 T cells and CD8 TEM cells. In contrast, cells with little evidence of somatic reversion (i.e. those with a low frequency of DOCK8-expressing cells such as naïve CD4+ T cells and NK cells), had comparable peak heights between the mutated and wild-type alleles (Fig 5.4). This indicates that repair of the de novo missense mutation was likely the reversion causing- event in the patient’s immune cells.

104

Figure 5.4 DOCK8 exon 20 splice acceptor site sequence in lymphocyte subsets. Genomic DNA was extracted from sorted B cells, NK cells, naïve and memory CD4+ T cells and naïve and effector memory CD8+ T cells from the patient and sequenced across the exon 20 splice site mutation-containing region. Box shows mutation site. Bases that differ from the consensus (wild type) sequence are highlighted in red. Chromatograms were visualised with Snapgene. 105

5.2.4 The extent of reversion varies depending on the lymphocyte subset and stage of differentiation Sequencing bulk PCR products generated by PCR across the exon 20 splice acceptor site mutation provided an indication that modification of this mutation was likely responsible for the expression of wild-type DOCK8 protein. However, as illustrated in Fig 5.4, the data generated by this approach lacks quantitation. To obtain greater clarity on the level of reversion occurring in different immune cell subsets, TA cloning was performed. Here, DNA from sorted lymphocyte subsets (NK cells, naïve CD4+ T cells, memory CD4+ T + + cells, naïve CD8 T cells, CD8 TEM cells, naïve B cells and memory B cells) was PCR amplified at the exon 20 mutation site, and individual amplicons were cloned into a vector, which was then transformed into E. coli, and the resulting colonies sequenced. As the patient is heterozygous at this site, if no reversion had occurred, ~50% of all colonies would be expected to contain the consensus (i.e. wild-type) sequence and 50% the mutated sequence. The percentage of repaired cells can be calculated based on the expected number of mutant colonies (i.e. 50% of total colonies) compared to the observed number.

We found that the frequency of repaired cells varied considerably amongst the analysed lymphocyte subsets, ranging from only 8.4% in naïve CD4+ T cells to 92.9% in naïve CD8+ T cells (Table 5.1). The frequency of repaired cells as determined by TA cloning differed from the frequency of repaired cells as determined by DOCK8 expression by flow cytometry. This discrepancy has also been previously noted by Jing et al53. However, in general, the same patterns for reversion seen from our TA cloning results were consistent with those observed for DOCK8 expression. Specifically, naïve CD4+ T and B cells showed fewer repaired cells than memory CD4+ T and B cells, respectively (Table 5.1). While the TA cloning results are not entirely consistent with the flow cytometry results, they do further suggest that repair of the missense mutation at the exon 20 splice acceptor site is responsible for cellular expression of DOCK8 protein in this patient. Additional evidence for this could be obtained by undertaking TA cloning of the exon 8 deletion site, which would be expected to show an equal proportion of wild type and mutant colonies, and thus indicate no repair at that site.

106

Table 5.1 TA cloning results in patient lymphocyte subsets

Cell subset Total Expected Observed Repaired cells colonies mutant mutant (%) colonies colonies NK cells 93 46.5 26 44.1 Naïve CD4+ T 83 41.5 38 8.4 Memory CD4+ T 96 48 32 33.3 Naïve CD8+ T 84 42 3 92.9 + CD8 TEM 83 41.5 25 39.8 Naïve B 91 45.5 35 23.1 Memory B 81 40.5 9 77.8

107

5.2.5 Immune cell phenotyping reveals similarities with typical DOCK8 deficiency DOCK8-deficient patients have a unique immune cellular phenotypic signature, commonly characterised by T cell lymphopenia with T cells skewed to an exhausted memory phenotype, increased total but reduced memory B cells, increased γδ T cells and reduced  T, MAIT and NKT cells (see Fig 4.2). Since this patient had detectable DOCK8-expressing cells, despite the presence of heterozygous DOCK8 mutations, we analysed the phenotype of their immune cells to compare this to previously characterised DOCK8-deficient patients who completely lack DOCK8-expressing cells. Intriguingly, this patient displayed many, but not all, of the phenotypic characteristics of DOCK8- deficient patients. Specifically, decreased frequencies of both CD3+ and CD4+ T cells, an increased frequency of B cells (Fig 5.5A), but decreased proportions of memory B cells (Fig 5.5B), and the skewing of the CD4+ and CD8+ T cell compartment to an effector memory phenotype with a corresponding loss of naïve cells (Fig 5.5C, D). However, + + while there were increases in proportions of memory CD4 T and CD8 TEM cells from this patient expressing PD1 and CD57, due to the high frequency of cells expressing CD127 and CD28 the signature of chronic activation/exhaustion of memory T cells was not as pronounced as in DOCK8-deficient patients (Fig 5.5E, F). Similarly, proportions of  T cells (Fig 5.5G), MAIT cells (Fig 5.5H), and NKT cells (Fig 5.5I) seen in this patient were comparably to healthy donors, which contrasts with typical DOCK8- deficient patients, in which these subsets are reduced. Additionally, a large increase in NK cells was observed in this patient, which is not generally seen in DOCK8-deficient patients (Fig 5.5J). Thus, the impact of the presence of DOCK8-expressing cells on the phenotype of this patient is variable, with some improvements observed relative to typical DOCK8-deficient patients, and some aspects of DOCK8 deficiency still visible.

108

Figure 5.5 The patient displays some of the phenotypic features of DOCK8 deficiency PBMC from healthy donors (n=10), the DOCK8-revertant patient (DOCK8r) and DOCK8- deficient patients (DOCK8 pt; n=7-9) were stained for CD3, CD4, CD8, CD20, CD56, CCR7, CD45RA, CD10, CD27, PD1, CD57, CD28, CD127, CD161, Vαβ, Vγδ, Vα7.2, Vα24 and Vβ11. (A) frequency of CD3+, CD4+, CD8+ T cells and B cells (CD20+); (B) frequency of transitional (CD20+CD10+CD27-), naïve (CD20+CD10-CD27-) and memory (CD20+CD10+CD27±) B cells; (C) frequency of naïve (CD4+CCR7+CD45RA+), central memory (CD4+CCR7+CD45RA-) and effector memory (CD4+CCR7-CD45RA-) CD4+ T cells; (D) frequency of naïve (CD8+CCR7+CD45RA+), central memory (CD8+CCR7+CD45RA-), effector memory (CD8+CCR7-CD45RA-) and EMRA (CD8+CCR7-CD45RA+) CD8+ T cells; (E, F) frequency of PD1, CD57, CD27, CD28 and CD127-expressing CD4+ memory (E) T cells and CD8+ effector memory (F) T cells; (G-I) frequency of (G) αβ (CD3+ Vαβ+) and γδ (CD3+Vγδ+) T cells, (H) MAIT cells (CD3+CD161+Vα7.2+), (I) NKT cells (CD3+Vβ11+ Vα24+) within the CD3+ population; and (J) frequency of NK cells (CD3-CD56+) were determined by flow cytometry. Error bars represent SEM.

109

5.2.6 DOCK8 reversion partially restores effector function of T cells To elucidate the effect of DOCK8 somatic reversion on the function of lymphocytes in this patient, we isolated their T cells and investigated their behaviour using in vitro assays.

CD8+ T cells: total CD8+ T cells were sorted from healthy donors and the DOCK8 revertant patient, labelled with CFSE, and cultured for 5 days with TAE beads with or without IL2. On day 5, supernatants were collected and cells were restimulated with PMA and ionomycin in the presence of Brefeldin A to investigate intracellular cytokine expression. Analysis of CFSE dilution showed that proliferation of patient CD8+ T cells was comparable to that of CD8+ T cells from healthy donors (Fig 5.6A). Frequency of cells expressing the activation markers CD69 and CD95 and the degranulation marker CD107a, was reduced in the patient following culture with TAE beads alone, but mostly restored by addition of exogenous IL2 (Fig 5.6B). The patient cells showed equivalent or higher expression of the intracellular cytokines IFNγ, perforin, granzyme B and TNFα when compared to healthy donor CD8+ T cells. However, IL2 expression was decreased in patient CD8+ T cells compared to CD8+ T cells from healthy donors (Fig 5.6C). Similar results were observed for cytokine secretion in culture supernatants (Fig 5.6D). Hence, this patient differs from typical DOCK8-deficient patients as they do not display the same extent of decrease in proliferation, activation, degranulation, or cytokine secretion, seen in these patients.

Memory CD4+ T cells: next, CD4+ CD45RA- memory T cells from this patient and healthy donors were sorted, labelled with CFSE and cultured for 5 days with TAE beads. On day 5, supernatants were collected and cells were restimulated with PMA and ionomycin in the presence of Brefeldin A to investigate intracellular cytokine expression. By examining dilution of CFSE, we found that the proliferation of patient memory CD4+ T cells was reduced compared to that of memory CD4+ T cells from the healthy control (Fig 5.7A). Production of cytokines, determined either by quantifying proportions of cells expressing cytokines intracellularly (Fig 5.7B) or measuring cytokine secretion (Fig 5.7C), revealed heightened production of Th2 (i.e. IL4, IL5 and IL13) cytokines, but intact production of Th1 (i.e. IFNγ and TNFα) and Th17 (i.e. IL17A, IL17F and IL22.) cytokines as well as IL21 which is important for B cell activation. Thus, similar to CD8+ T cells, memory CD4+ T cells from the DOCK8-revertant on the whole functioned better than those of DOCK8-deficient patients, as proliferation was only slightly diminished in

110

the patient and decreases in production of Th1and Th17 cytokines was not seen. However, the skewing of memory CD4+ T cells to Th2 cells was preserved.

111

Figure 5.6 CD8+ T cell function is superior in the DOCK8-revertant patient than in typical DOCK8-deficient patients Sorted CD8+ T cells from healthy donors (n=5-9), the DOCK8-revertant patient (DOCK8r) and DOCK8-deficient patients (DOCK8 pt; n=3-7) were stained with CFSE and cultured for 5 days with TAE beads (anti-CD2, -CD3, -CD28) with and without IL2. (A) Proliferation was determined by CFSE dilution. (B) After restimulation with PMA/ionomycin, cells were stained for surface expression of CD69, CD95 and CD107a, fixed and (C) stained intracellularly for expression of the cytokines IL2, IFNγ, Perforin, TNFα and Granzyme B. (D) Supernatants were harvested on day 5 to measure cytokine secretion by CBA. Error bars represent SEM. 112

Figure 5.7: DOCK8-revertant patient memory CD4+ T cell function is not impaired to the same extent as DOCK8-deficient patients Sorted memory CD4+ T cells from healthy donors (n=8), the DOCK8-revertant patient (DOCK8r), or DOCK8-deficient patients (DOCK8 pt; n=2-7) were stained with CFSE and cultured for 5 days with TAE beads. (A) Proliferation was determined by CFSE dilution. (B) After restimulation with PMA/ionomycin, cells were intracellularly stained for cytokine expression of Th1 cytokines, Th2 cytokines, Th17 cytokines and IL21. (C) Supernatants were harvested to measure Th1, Th2 and Th17 cytokine secretion by CBA. Error bars represent SEM.

113

5.2.7 Memory, but not naïve, B cells from the DOCK8-revertant patient exhibit improved function over DOCK8-deficient B cells To compare B cell function in this DOCK8-revertant patient to that of DOCK8-deficient patients, aspects of B cell activation and differentiation (proliferation, survival, Ig secretion) were examined. CD10-CD27- naïve B cells were isolated from the DOCK8- revertant patient and healthy donors, labelled with CFSE and cultured for 5 days under various stimuli to mimic T cell help (CD40L+IL21), TLR signalling (CpG) and BCR engagement (SAC). Proliferation was measured by dilution of CFSE, cell numbers were determined by the addition of a known number of Calibrite beads and supernatants were collected to analyse Ig secretion by ELISA. We also examined responses of memory B cells, however due to limited numbers of these cells recovered from the patient, we were unable to investigate proliferation by tracking CFSE dilution.

Proliferation of naïve B cells from the DOCK8-revertant patient was markedly reduced compared to that of naïve B cells from healthy donors for all culture conditions tested. This was particularly notable following culture with the combination of CD40L, CpG and SAC (Fig 5.8A). Survival of patient naïve B cells was also diminished under a range of stimuli (Fig 5.8B). Ig secretion by patient naïve B cells was drastically reduced compared to healthy donors with no IgG being produced under any of the conditions tested, IgA levels lower in the CD40L+IL21 culture and IgM levels being several-fold lower for patient naïve B cells compared to controls in most of the in vitro cultures (Fig 5.8D). These pronounced defects exhibited by naïve B cells from the DOCK8-revertant patient in survival and Ig secretion replicate those observed in DOCK8-deficient patients while proliferation is more impaired than in typical patients.

In stark contrast to naïve B cells, responses of memory B cells from the DOCK8-revertant patient were similar to that of healthy donors, with survival (Fig 5.8C) and secretion of IgG, IgA and IgM (Fig 5.8E) reaching levels comparable to or higher than that of B cells from healthy donors under most culture conditions.

DOCK8-deficient patients have a severe reduction in memory B cells, and consequently, adequate numbers of DOCK8-deficient memory B cells could not be obtained to examine their function during the course of this thesis. However, Ig secretion by DOCK8-deficient memory B cells from 5 patients has previously been assessed in the lab – these experiments revealed an impairment in memory B-cell differentiation similar to that for

114

DOCK8-deficient naïve cells (Fig 5.8). Hence it can be concluded that the DOCK8- revertant patient investigated here has enhanced memory B cell function compared to typical DOCK8-deficient patients.

115

Figure 5.8 Restoration of function in memory, but not naïve, B cells in the DOCK8- revertant patient Naïve and memory B cells sorted from healthy donors (n=1-5), the DOCK8-revertant patient (DOCK8r) and DOCK8-deficient patients (DOCK8 pt; n=5-7) were labelled with CFSE (naïve only) and cultured for 5 days with stimuli to mimic a combination of CD4+ T cell help (CD40L, IL21), BCR engagement (SAC) and TLR signalling (CpG). (A) Proliferation of naïve B cells was determined by dilution of CFSE. Cell recovery of (B) naïve and (C) memory cells was quantified using Calibrite beads. Supernatants were harvested to determine Ig secretion by ELISA of (D) naïve and (E) memory B cells. Error bars represent SEM.

116

5.2.8 DOCK8+ and DOCK8- lymphocytes in the revertant patient exhibit distinct phenotypes that replicate features of healthy donors and DOCK8-deficient patients Since we were able to detect DOCK8-expressing and DOCK8-deficient cells in the same individual, we had the unique opportunity to evaluate the impact of DOCK8-deficiency in lymphocytes that had been exposed to the same environment. Thus, the phenotype of DOCK8-expressing (i.e. DOCK8-revertant) and DOCK8-deficient (non-revertant) lymphocyte subsets present in the reversion patient was compared to healthy donors and typical DOCK8-deficient patients. The frequencies of CD3+ T cells, CD4+ T cells (Fig 5.10A),  T and MAIT cells (Fig 5.10B, C) in DOCK8-revertant lymphocytes more closely resembled the profile of healthy donors than DOCK8-deficient lymphocytes within the patient. Accordingly, DOCK8-deficient lymphocytes within the DOCK8- revertant patient contained an increased frequency of B cells, NK cells (Fig 5.10A) and γδ T cells (Fig 5.10B) which matched well with that observed in typical DOCK8-deficient patients. As expected, the DOCK8-revertant B cells detected in the revertant patient were primarily memory cells (Fig 5.10D) and the DOCK8-revertant T cells in both CD4+ and + - - CD8 T cells were predominantly CD45RA CCR7 TEM cells (Fig 5.10E, F) while the DOCK8-deficient cells were more evenly distributed across both naïve and memory T cell subsets (Fig 5.10D, E, F). Hence, while sharing the same environment, within the same individual, DOCK8-revertant lymphocytes acquired the phenotype of lymphocytes present in healthy controls while DOCK8-deficient lymphocytes continued to resemble those seen in typical DOCK8-deficient patients.

117

Figure 5.9 Phenotype of DOCK8-revertant and DOCK8-deficient cells within the DOCK8- revertant patient PBMCS from the DOCK8-revertant patient were stained for surface expression of CD3, CD4, CD8, CD20, CD56, CCR7, CD45RA, CD10, CD27, Vαβ, Vγδ, Vα7.2, Vα24 and Vβ11 then fixed, permeabilised and intracellularly stained for DOCK8. Cells were identified as DOCK8- or DOCK8+and then analysed for frequencies of (A) CD3+, CD4+, CD8+ T cells, B cells (CD20+) and NK cells (CD3-CD56+); (B) αβ (CD3+ Vαβ+) and γδ (CD3+Vγδ+) T cells; (C) MAIT (CD3+CD161+Vα7.2+) and NKT cells (CD3+Vβ11+ Vα24+); (D) frequency of transitional (CD20+CD10+CD27-), naïve (CD20+CD10-CD27-) and memory (CD20+CD10+CD27+) B cells; (E) naïve (CD4+CCR7+CD45RA+), central memory (CD4+CCR7+CD45RA-) and effector memory (CD4+CCR7-CD45RA-) CD4+ T cells; (F) naïve (CD8+CCR7+CD45RA+), central

+ + - + - - memory (CD8 CCR7 CD45RA ), effector memory (CD8 CCR7 CD45RA ) and TEMRA (CD8+CCR7-CD45RA+) CD8+ T cells. Data from healthy controls (n=8-9) and DOCK8-deficient patients (n=7-9) was included as a comparison. Error bars represent SEM.

118

5.2.9 DOCK8-revertant CD8+ T cells show reduced signs of exhaustion and, unlike DOCK8-deficient cells, efficiently undergo activation and cytokine production in vitro To continue to investigate the biology of DOCK8-expressing revertant and DOCK8- deficient cells in this patient, CD8+ T cell function was assessed in these two populations. Patient lymphocytes were stained with CD45RA and CCR7 to identify CD8+ CD45RA- - CCR7 TEM cells and for expression of exhaustion markers and compared to TEM cells from healthy donors and DOCK8-deficient patients. This revealed that the DOCK8- + deficient CD8 TEM cells were in a more exhausted state than the DOCK8-revertant TEM CD8+ T cells present in this patient. This was indicated by increased proportions of CD57+ and PD1+ cells and fewer CD28+ and CD127+ cells within the DOCK8-deficient subset + + of CD8 TEM cells (Fig 5.11A). This is similar in phenotype to CD8 T cells from DOCK8-deficient patients (Fig 5.11A). When CD8+ T cells were sorted, cultured with TAE beads for 5 days and then restimulated with PMA/ionomycin to assess cytokine expression, the DOCK8-revertant cells contained higher frequencies of cells expressing the activation-induced markers CD69 and CD95 (Fig 5.11B and, the intracellular cytokines IFN, TNF and granzyme B (Fig 5.11C) than the DOCK8-deficient cells. Again, this resulted in DOCK8-deficient T cells in this patient resembling that of CD8+ T cells present in typical DOCK8-deficient patients, while the functionality of the DOCK8-revertant cells more matched that of healthy controls.

5.2.10 DOCK8-revertant memory CD4+ T cells display a different cytokine production profile to DOCK8-deficient cells As the microenvironment can influence CD4+ T cell differentiation and subsequent cytokine production, the cytokine profile of DOCK8-revertant and DOCK8-deficient memory CD4+ T cells was determined. Despite being exposed to identical differentiation conditions in vivo, DOCK8-revertant memory CD4+ T cells had higher frequencies of memory CD4+ T cells expressing Th1 and Th17 cytokines as well as IL21 than DOCK8- deficient memory CD4+ T cells. In fact, the frequencies of Th1 and Th17 cytokine- expressing and IL21-expressing cells by DOCK8-reverted memory CD4+ T cells were often comparable to those seen in healthy donor memory CD4+ T cells. On the other hand, DOCK8-deficient memory CD4+ T cells had an increased frequency of Th2 cytokine

119

producing cells (Fig 5.12) which mirrors DOCK8-deficient patients who exhibit a Th2 skewing at the expense of Th1 and Th17 cells.

120

Figure 5.10 Differences in DOCK8-revertant and DOCK8-deficient patient CD8+ T cells (A) Patient PBMCs were stained for surface expression of CD8, CCR7, CD45RA, PD1, CD57, CD27, CD28, CD127 and intracellularly for DOCK8. The expression of exhaustion markers on DOCK8+ and DOCK8- CD8+ CD45RA-CCR7- effector memory T cells was then determined. (B, C) Sorted patient CD8+ T cells were cultured for 5 days with TAE beads. After restimulation with PMA/ionomycin, (B) cells were stained for surface CD69 and CD95 expression and (C) intracellularly stained for expression of IL2, IFNγ, Perforin, TNFα and Granzyme B. Cells were also stained for DOCK8 which allowed for identification of DOCK8- and DOCK8+ cells. Data from healthy controls (n=6-9) and DOCK8-deficient patients (n=6- 8) were included as a comparison. Error bars represent SEM.

121

Figure 5.11 DOCK8 reversion corrects defective cytokine production by DOCK8- deficient memory CD4+ T cells Sorted memory CD4+ T cells from the DOCK8-revertant patient were cultured for 5 days with TAE beads. After restimulation with PMA/ionomycin, cells were intracellularly stained for cytokine expression as well as DOCK8 which allowed for identification of DOCK8- and DOCK8+ cells in this patient. A) Contour plot showing intracellular expression of IFNγ and IL4 by DOCK8+ and DOCK8- memory CD4+ T cells from the DOCK8-revertant patient. B) Frequency of IFNγ, TNF, IL4, IL13, IL17A, IL17F, IL22 and IL21 expressing cells. Data from healthy controls (n=7) and DOCK8-deficient patients (n=7) were included as a comparison. Error bars represent SEM.

122

5.2.11 DOCK8-revertant CD4+ memory and CD8+ T cells have a selective advantage over DOCK8-deficient cells in vitro Because environmental signals impact lymphocyte survival and proliferation, the ability of DOCK8-revertant and DOCK8-deficient total CD8+ and memory CD4+ T cells to survive and proliferate in the same environment was investigated. After 5 days of culture with TAE beads, cell viability was determined by a dead cell permeable dye (zombie). The proportion of DOCK8-revertant CD8+ T cells and memory CD4+ T cells that remained viable after this culture period was at least twice that of their DOCK8-deficient counterparts (Fig 5.13A, C). This resembles the survival of T cells from healthy donors and DOCK8-deficient patients, respectively (Fig 5.13A, C). Additionally, the frequency of DOCK8-expressing cells amongst the total live cells was measured after 5 days of culture. The frequency of DOCK8-expressing CD8+ T cells and memory CD4+ T cells was substantially higher than in cells that had not been cultured (88.4 vs 48.6; 81.1 vs 64.2), indicating a selective expansion of these revertant cells (Fig 5.12B, D; Fig 5.2C, D). Taken together, these results confirm that DOCK8 expression conveys an intrinsic selective advantage in vitro.

5.2.11 Broad TCR diversity is present in both DOCK8-revertant and DOCK8- deficient T cells Somatic reversion in DOCK8-deficient patients has previously been reported to result in only a modest clinical improvement compared to patients who did not exhibit reversion. This led to the hypothesis that this could be due to restricted TCR diversity among the revertant population of T cells. Consistent with this, one DOCK8-revertant patient had ~20% of their revertant T cells dominated by only 2 clonotypes53. Hence, we explored the diversity of the TCR V repertoire in our patient by flow cytometry using a range of mAbs that detect distinct TCR V chains, which collectively account for approximately 70% of the human TCR repertoire. Unlike the previous report53 , we found that DOCK8- revertant memory CD4+ T cells were broadly polyclonal, being distributed across the majority of TCR V chains, and did not differ from the DOCK8-deficient cells present in + this patient (Fig 5.13A). Broad diversity was also seen in CD8 TEM cells, however some differences were observed between DOCK8-revertant and DOCK8-deficient cells in their preference for particular TCR V chains. For example, V2 was present in up to 5 times

123

more DOCK8-revertant cells than DOCK8-deficient cells (Fig 5.13B). This diversity amongst DOCK8-revertant T cells may explain to some degree the lack of recurrent infections, which are characteristic to DOCK8 deficiency, in this DOCK8-revertant patient in recent years (i.e. presumably since the reversion took place).

124

Figure 5.12 Survival and proliferation of patient DOCK8-revertant and DOCK8- deficient T cells Total CD8+ and memory CD4+ T cells were sort-purified from the DOCK8-revertant patient and cultured for 5 days with TAE beads. After this time, the cells were harvested and intracellularly stained with a dead cell marker (zombie) as well as anti-DOCK8 mAb which allowed for identification of DOCK8- and DOCK8+ cells. (A, C) The frequency of live cells was determined as percentage of zombie negative cells. (B, D) The frequency of DOCK8-expressing cells was determined within the total live cell population. Data from healthy controls (n=3-7) and DOCK8-deficient patients (n=2-6) were included as a comparison. Error bars represent SEM.

125

Figure 5.13 Both DOCK8-revertant and DOCK8-deficient patient memory T cells exhibit a polyclonal TCR repertoire Patient PBMCs were stained for surface expression of CD4, CD8, CCR7 and CD45RA as well as V diversity using the Vβ repertoire kit and then intracellularly stained for DOCK8 to identify DOCK8- and DOCK8+ cells. The frequency of cells expressing different Vβ chains in (A) CD4+ CD45RA- memory and (B) CD8+ CD45RA-CCR7- effector memory T cells was determined by flow cytometry.

126

5.2.12 DOCK8-revertant lymphocytes increase over time Given that DOCK8-expressing revertant lymphocytes in this patient had a survival and proliferative advantage in vitro, we speculated that this would also occur in vivo, with the net effect being an accumulation of DOCK8-revertant cells over time. To investigate this, the frequency of DOCK8-expressing cells was determined from three different blood samples collected over a period of 35 months. The proportion of total lymphocytes expressing DOCK8 steadily increased over time, from 17% to 40% (~2.5-fold expansion, Fig 5.14A).

Within CD4+ T cells, DOCK8-revertant cells increased in the memory, but not naïve, compartment (Fig 5.14B). In contrast, the proportions DOCK8-revertant cells in all subsets of CD8+ T cells, including naïve, increased over time (Fig 5.14C). While the proportions of DOCK8-expressing cells in total B cells did not appear to accumulate over the ~3-year time frame that we were able to study the patient, the frequency of DOCK8- expressing memory B cells markedly increased over time (Fig 5.14D). Consistent with the low proportion of DOCK8-revertant expressing cells detected in the naïve and transitional B-cell subsets (Fig 5.2), there were no changes to the frequencies of revertant transitional or naïve B cells in this patient during the analysed time frame (Fig 5.14D). Thus, correction of the pathogenic DOCK8 variant by somatic reversion led to substantial normalisation of DOCK8 expression within the adaptive memory immune compartment in this patient. It is noteworthy that such large gains in DOCK8-expressing cells occurred over a relatively short period of time.

5.2.13 Patient IgE levels decline over time One clinical feature of DOCK8-deficient patients that can be measured over time is levels of total and allergen specific serum IgE and correlations with allergic disease. Hence, we wanted to test if the DOCK8-revertant patient had normalised their levels of elevated IgE over time, in accordance with the amelioration of their other symptoms as clinical reports had documented. At the first time point available, serum levels of total IgE in this patient were substantially increased at 1143 kU/l compared to healthy donors (9-34 kU/l). However, 18 months later, these levels had decreased substantially (~93%) to 76 kU/l (Fig 5.15A). Consistent with this decline in total serum IgE levels, IgE specific for staple foods was also initially high in the patient compared to healthy donors, but sharply

127

decreased over time to be comparable to healthy donors (Fig 5.16B). Dust mite specific IgE also decreased over time (Fig 5.15C). This kinetic reduction in total and allergen- specific IgE may explain why the patient experienced a milk allergy earlier in life which is now resolved.

128

Figure 5.14 DOCK8 expression in lymphocyte subsets over time. PBMCs from the DOCK8-revertant patient taken at various time points were stained for surface expression of CD4, CD8, CD20, CCR7, CD45RA, CD10 and CD27, fixed and intracellularly stained for DOCK8. Percentage of cells expressing DOCK8 in (A) total lymphocytes, (B) CD4+ T cells, (C) CD8+ T cells and (D) B cells was determined by flow cytometry. 129

Figure 5.15 Total and allergen-specific IgE levels decline over time. Serum was prepared from blood samples of the DOCK8-revertant patient at different times and was tested by ImmunoCAP assay for concentration of (A) total, (B) staple food specific and (C) dust mite specific IgE. Dotted line indicates average value of healthy donors (n=3).

130

5.3 Discussion Previous reports have presented only limited data on the phenotype and function of lymphocytes in patients with bi-allelic mutations in DOCK8 that express DOCK8 due to somatic reversion. One study documented a single patient had skewing of T cells to the memory phenotype189, similar to that observed in DOCK8-deficient patients, while another observed reduced proliferation of CD4+ T cells190, which is also a feature of DOCK8-deficient patients. Our comprehensive analysis of lymphocyte phenotype and function in a patient exhibiting somatic reversion in one variant allele of DOCK8 allowed for unprecedented comparison with typical DOCK8-deficient patients (studied in Chapter 4) and revealed both similarities and differences. The revertant patient demonstrated most of the cellular phenotypic features characteristic of DOCK8 deficiency but notably to a lesser extent than typical DOCK8-deficient patients. For example, reductions in memory B cells, MAIT cells and NKT cells in the DOCK8-revertant patient were less severe than in DOCK8-deficient patients, as was expression of markers of chronic activation/exhaustion. CD4+ and CD8+ T cell proliferation and function were not as robust as those of T cells from healthy donors but were greater than DOCK8-deficient patients. This finding also held true for memory B cells. However, naïve B cells from this revertant patient, which had a low frequency of cells expressing DOCK8, exhibited considerably compromised function suggesting that reversion to DOCK8 expression was essential for functional improvement in lymphocytes in the patient. Taken together, the DOCK8- revertant patient studied in this chapter in fact more closely resembles DOCK8-deficient patients who have undergone HSCT, who also show lower levels of impairment in immune cell differentiation and effector function post-HSCT than untransplanted DOCK8-deficient patients (see chapter 4). Hence, this DOCK8-revertant patient provides further evidence for reversion being considered ‘natural gene therapy’. It would be interesting to continue to evaluate the phenotype and function of lymphocytes in this patient, as well as their clinical progress, as the proportion of DOCK8-expressing cells continues to increase over time.

It was interesting that our results showed a large variation in the proportion of revertant cells in different lymphocyte subsets. For instance, both CD4+ and CD8+ T cell compartments contained more DOCK8-expressing cells than B cells. Indeed this has also been seen in previous reports53,189,190. It may be that DOCK8-expressing B cells do not have the same survival/proliferative advantage as DOCK8-expressing revertant T cells. 131

Our analysis of DOCK8-revertant T cells showed these cells have considerable survival and proliferative advantage over the remaining DOCK8-deficient cells present in the same patient. Due to limited cell numbers, we were not able to directly determine if this was also the case for DOCK8-expressing B cells but considering that naïve B cells from DOCK8-deficient patients have decreased survival and proliferation (see chapter 4) and other reports have shown decreased activation of memory B cells in DOCK8-deficient patients163 one could speculate that there would be some level of advantage for DOCK8- expressing B cells. Additionally, it has been demonstrated in healthy donors that CD4+ and CD8+ T cells have higher proliferative rates than B cells191-193 which would allow DOCK8-revertant T cells to expand more quickly than corresponding DOCK8-revertant B cells. Furthermore, it was striking that memory cells within the CD4+, CD8+ and B cell compartments had much higher frequencies of reverted cells than naïve cells. Previous reports have noted higher reversion frequency in memory T cells but were limited in either that they defined naïve CD8+ T cells as CD45RO- 53, a phenotype that would also capture TEMRA cells (which we have shown to be expanded in DOCK8-deficient patients; see Chapter 4) or that they studied memory cells in total T cells rather than CD4+ or CD8+ T cells189. Again, the difference in proliferative rates could be playing a role in the expansion of memory over naïve cells in the CD4+ T cell, CD8+ T cell and B cell compartments as it is well known that memory cells proliferate much more rapidly than their naïve counterparts191-193. Our analysis of revertant populations in the patient over time also supports this possibility as the frequency of DOCK8-expressing cells in naïve compartments of B and CD4+ T cells showed less expansion over time than memory compartments.

The presence of both DOCK-revertant and DOCK8-deficient immune cells within the same individual provided a unique opportunity to study a human ‘chimera’ much like mixed bone marrow chimeras in mouse studies. We were able to investigate the influence of DOCK8-deficient haematopoietic and non-haematopoietic compartments on the function of DOCK8+ and DOCK8- lymphocytes co-existing in the same environment. The increased expression of markers associated with chronic activation observed on CD8+

TEM cells in DOCK8-deficient patients, and thus premature/accelerated senescence of these cells, was hypothesised to be due to frequent and chronic encounters with viral pathogens65. However, in the DOCK-revertant patient studied here, the DOCK8- expressing cells were found to be not as exhausted as DOCK8-deficient cells which have 132

both been exposed to the same microenvironmental and pathogenic stimuli. This suggests there is an intrinsic contribution of DOCK8-deficiency to the exhaustion/senescence of + CD8 TEM cells in these patients. In Chapter 3 it was discussed how the Th1 defect in memory CD4+ T cells in DOCK8-deficient patients was due to an element of the in vivo environment not present in the in vitro cultures of DOCK8-deficient naïve CD4+ T cells which resulted in intact Th1 differentiation. Weakened TCR signalling in DOCK8- deficient cells in vivo, possibly due to impaired IS formation between CD4+ T cells and APCs, which was overcome by the method of stimulus used in vitro was suggested as this differential element. This is supported by the results from this DOCK8-revertant patient in which we observed that their DOCK8-expressing cells produced more Th1 cytokines than their DOCK8-deficient cells, despite these cells differentiating in the same cytokine milieu. Thus, it is possible that the DOCK8-expressing cells exhibited normal TCR signalling while the DOCK8-deficient cells exhibited weakened TCR signalling leading to different patterns of differentiation. This could have been due to the ability of DOCK8- expressing CD4+ T cells to form more stable IS with APCs than the DOCK8-deficient CD4+ T cells.

A particularly remarkable feature of the revertant patient studied here is that they are clinically well and have maintained this status for several years. In contrast, other DOCK8-deficient patients previously identified to harbour DOCK8-expressing lymphocytes within their peripheral immune system due to somatic reversion have continued to display severe symptoms of DOCK8 deficiency. Specifically, one patient experienced recurrent infections and severe failure to thrive despite antibacterial, antiviral and antifungal treatment and still required HSCT even though a large majority of their T cells (~80% CD4+ and ~72% CD8+) and NK cells (~98%) expressed DOCK8189. Another patient remained dependent on IVIg after antibiotics were unable to improve the occurrence and severity of infections despite the fact that, based on cDNA analysis, >60% of their T cells were deemed to express normal DOCK8190. The cohort of DOCK8- revertant patients studied by Jing et al had between 1-94% of their T cells re-expressing DOCK8. When these patients were evaluated clinically and given a total disease score, and compared to non-revertant DOCK8-deficient patients, the main clinical difference was that the disease score decreased with age for patients who exhibited somatic reversion of their DOCK8 gene but increased with age for non-revertant patients. However, of the 17 DOCK8-revertant patients studied, 6 still required HSCT and one died53. It may be 133

that DOCK8 expression in B cells accounts for this difference between patients. Previously reported patients had 0-10% DOCK8-expressing B cells53,189. The patient we studied had a larger proportion of DOCK8-expressing total B cells (~15%), which was largely due to the expansion and accumulation of DOCK8-expressing cells in the memory B-cell compartment (~88% at last evaluation). Additionally, in other reports, revertant T cells were limited to a narrower range of clonotypes53 while the revertant patient studied here displayed a broad unrestricted polyclonal TCR V repertoire in DOCK8-expressing CD4+ and CD8+ memory T cells, likely enabling them to better fight infections.

The comprehensive phenotypic and functional analysis, including investigations into DOCK8-expressing and DOCK8-deficient cells within the same lymphocyte subsets, presented here provides valuable insight into possible future treatments for DOCK8 deficiency and other PIDs. While HSCT has become the standard of care for DOCK8- deficient patients it is not without challenges including availability of a suitably matched donor and the risk of GvHD as well as viral reactivation79. The concept of gene therapy, essentially repairing a patient’s own haematopoietic stem cells with wild-type DNA via a viral vector and then reintroducing the cells to the patient, could overcome these obstacles. Indeed, gene therapy has been successfully applied to some PIDs including ADA deficiency, SCIDX1 and WAS194. To ensure success, repaired cells need to demonstrate a selective advantage over mutated cells. It has been shown in the patient studied here that this is true of DOCK8-expressing CD4+ and CD8+ T cells and memory B cells, suggesting DOCK8 deficiency could be a candidate for gene therapy. Development of improved gene therapy protocols is aiming to increase the number of cells successfully transduced with the repaired DNA and the rate of engraftment of these cells into patient bone marrow194,195. Our results here, from a revertant patient with variable levels of DOCK8 expression in different lymphocyte subsets, indicate that 100% reversion is not essential to achieve sufficient function for clinical improvement. This is an important consideration in the context of gene therapy as it suggests the procedure may be effective even without initial high levels of repaired cells present in the bone marrow

The work in this chapter further highlights the functional improvements in lymphocytes which have regained DOCK8 expression and provides insight into the influence of the cellular environment on the dysfunction of DOCK8-deficient lymphocytes.

134

Chapter 6: Lymphocyte phenotype and function in immunodeficiency patients with mutations in DOCK8-related proteins

6.1 Introduction The reported role of DOCK8 in cytoskeletal rearrangements does not fully explain the catalogue of clinical features displayed by DOCK8-deficient patients. To investigate this further, we examined the phenotype and function of lymphocytes in patients with mutations in genes that result in immune dysregulatory conditions with clinical features shared by DOCK8-deficiency or have been implicated in common signalling pathways. These genes included CDC42, as CDC42 is activated by DOCK884 ; WAS, as WASp interacts with DOCK870; and DOCK2, which like DOCK8, functions as a guanine exchange factor (GEF)81.

Autosomal dominant mutations in CDC42 have been reported in a very limited number of patients (<20) who have been categorized as having Takenouchi-Kosaki syndrome196. This syndrome is mostly characterised by thrombocytopenia, developmental delays and unique facial features197. Interestingly, similar to DOCK8-deficient patients, the majority of patients have also been reported to present with recurrent infections, especially of the respiratory tract, and several patients display T and B cell lymphopenia. Additionally, a few patients present with hypogammaglobulinemia as well as atopic disease including allergies, eczema and asthma198,199. Despite these observations, limited details exist regarding the consequences of CDC42 mutation on immune cell phenotype and function.

Hemizygous mutations in WAS can lead to either Wiskott-Aldrich syndrome (WAS) or X-linked thrombocytopenia (XLT), with indel and nonsense mutations generally resulting in WAS and missense mutations in XLT. XLT is considered a milder form of the disease seen in WAS patients200,201. WAS patients are generally identified by not only thrombocytopenia, but also susceptibility to infection and the presence of severe eczema, clinical features that are not generally seen in XLT patients although mild transient eczema has been reported in some XLT patients200,202. WAS patients also display an increased risk of autoimmune disease and malignancy and require HSCT for curative treatment203. The immune manifestations of WAS include lymphopenia, increased IgA and IgE levels but compromised Ag-specific antibody responses. Furthermore, low memory B cells, impaired suppressor function of Tregs and defective NK cell function

135

have also been reported as lymphocyte defects in WAS patients204-206. Given the milder nature of XLT symptoms, patients have a much better overall survival rate, but do still experience infections, which can be fatal, as well as autoimmunity and malignancy albeit at a lower rate and at a later age compared to WAS patients207.

To date, there have been only 12 patients reported with mutations in DOCK2208-211. These patients experience severe early-onset bacterial and viral infections and T-cell lymphopenia208, due to impaired mitochondrial respiration210. Some DOCK2-deficient patients also have low numbers of total and memory B cells and reduced NK cells208-211. T cells from DOCK2-deficient patients had impaired in vitro activation and, like DOCK8- deficent patients, are skewed to a memory phenotype. NKT numbers were decreased in DOCK2-deficient patients and NK cell degranulation was impaired in all patients tested208,210.

Given the documented interactions of CDC42 and WASP with DOCK8, and the overlapping functions of DOCK2 and DOCK8 as Rho GTPase-interacting GEFs, this chapter seeks to explore the cellular phenotype and function of lymphocytes in patients with mutations in CDC42, WAS and DOCK2. This will enable comparisons to be made and may reveal the extent of involvement of DOCK8-associated pathways in aspects of disease pathogenesis in these patients.

136

6.2 Results 6.2.1 Patients used in this study Phenotypic and functional analysis was undertaken on 1 CDC42-deficient patient, 2 WAS patients, 2 XLT patients and 1 DOCK2-deficient patient. Patient details are provided in Table 6.1. Analysis of patient phenotype was compared to age-matched healthy controls (6 months to 35 years).

6.2.2 Phenotypic analysis of patients with mutations in DOCK8-related proteins To gain insight into their clinical features and make comparisons with DOCK8-deficient patients, we investigated immune cell function in patients with mutations in genes coding for DOCK8-related proteins. Firstly, the phenotype of different lymphocyte subsets was assessed by flow cytometry. This revealed T cell lymphopenia in WAS, XLT and the DOCK2-deficient patients (Fig 6.1A). Conversely the CDC42-deficient patient had increased proportions of T cells compared to healthy donors, while frequencies of B cells and NK cells were reduced in the CDC42- and DOCK2-deficient patients compared to healthy donors (Fig 6.1A). Frequencies of CD4+ and CD8+ T cells within the CD3+ compartment were perturbed in most patients, with WAS patients exhibiting decreased CD4+ T cells, XLT and DOCK2-deficient patients displaying increased CD4+ T cells and WAS patients displaying increased CD8+ T cells. In contrast, the DOCK2-deficient patient had decreased CD8+ T cells (Fig 6.1B). This resulted in a reduced CD4/CD8 ratio in WAS patients and an increased CD4/CD8 ratio in XLT and DOCK2-deficient patients (Fig 6.1C). The differentiation of CD4+ T cells was affected in CDC42- and DOCK2- deficient patients with skewing to the effector memory subset at the expense of naïve cells (Fig 6.1D). This skewing towards more differentiated subsets was also seen in CDC42- and DOCK2-deficient and WAS patient CD8+ T cells (Fig 6.1E). On the other hand, memory B cells were decreased in all patients compared to healthy donors, with corresponding increases in transitional B cells for XLT and DOCK2-deficient patients or naïve cells for the CDC42-deficient patient (Fig 6.1F). Decreased MAIT cell frequency was observed in CDC42-deficient, WAS and DOCK2-deficient patients (Fig 6.1G) and NKT cells were absent in the DOCK2-deficient patient (Fig 6.1H). The presence of CD4+ T cells expressing an exhaustion signature was observed in WAS, XLT and DOCK2- deficient patients with high levels of CD57 expression and low levels of CD127

137

expression (Fig 6.1I). CD8+ T cells in WAS patients also displayed signs of exhaustion, with increased CD57 expression and decreased CD127 expression. CD8+ T cells from XLT patients showed decreased CD127 expression and all patients had decreased CD28 expression on CD8+ T cells, another hallmark of T cell exhaustion (Fig 6.1J).

138

Table 6.1 Clinical details of Tablepatients Clinical 6.1 details 139

Figure 6.1 Lymphocyte phenotype of patients with mutations in DOCK8-related proteins PBMCs from healthy donors (n=6) as well as a CDC42-deficient patient, WAS patients (n=2), XLT patients (n=2) and a DOCK2-deficient patient were stained with mAbs directed against CD3, CD4, CD8, CD20, CD56, CCR7, CD45RA, CD10, CD27, CD57, CD28 and CD127. (A) frequency of T cells (CD3+), B cells (CD20+) and NK cells (CD3-CD56+); (B) frequency of CD4+ T (CD3+CD4+) and CD8+ T (CD3+CD8+) cells; (C) CD4+ T cell/CD8+ T cell ratio; (D) frequency of naïve, (CD4+CCR7+CD45RA+) central memory (CD4+CCR7+CD45RA-) and effector memory (CD4+CCR7-CD45RA-) CD4+ T cells; (E) frequency of naïve (CD8+CCR7+CD45RA+), central memory (CD8+CCR7+CD45RA-), effector memory (CD8+CCR7-CD45RA-) and EMRA (CD8+CCR7-CD45RA+) CD8+ T cells; (F) frequency of transitional (CD20+CD10+CD27-), naïve (CD20+CD10-CD27-) and memory B cells (CD20+CD10+CD27+); (G) frequency of MAIT cells (CD3+CD161+Va7.2+); (H) frequency of NKT cells (CD3+V11+ V24+); (I) frequency of CD57, CD27, CD28 and CD127-expressing CD4+ memory T cells and (J) frequency of CD57, CD27, CD28 and CD127-expressing CD8+ effector memory T cells were determined by flow cytometry. Error bars represent SEM. 140

6.2.3 CD8+ T cell function in patients with mutations in DOCK8-related proteins We next analysed CD8+ T cell function in patients with mutations in DOCK8-related proteins as we had found that CD8+ T cell function was impaired in DOCK8-deficient patients. This is likely to contribute to the susceptibility of these patients to viral infections, a feature of the DOCK8-related patients in this study. Sorted CD8+ T cells were cultured for 5 days with TAE beads (anti-CD2, -CD3, -CD28) and supernatants were collected for analysis of cytokine secretion before cells were restimulated with PMA/ionomycin in the presence of Brefeldin A to determine surface marker expression and intracellular cytokine production. All patients showed substantial decreases in expression of the surface activation markers CD25, CD69 and CD95 although expression of the degranulation marker CD107a was not as impacted (Fig 6.2A). Cytokine production varied across the patients as the CDC42 patient had increased levels of IFNγ, Granzyme B and Perforin but reduced levels of TNFα, while XLT and DOCK2-deficient patients had normal levels of IFNγ, but reduced levels of all other cytokines examined (Fig 6.2B). Cytokine secretion was drastically reduced in all patients, suggesting that survival of these cells in in vitro culture was compromised (Fig 6.2C).

T cells from 1 WAS and 1 XLT patient were tested for their ability to respond to TCR- + induced stimulation by measuring calcium flux in naïve and CD8 TEM cells. Cells were loaded with indo-1am dye and calcium was detected by flow cytometry over the course of 7 minutes allowing for the frequency of responding cells to be calculated. The WAS patient exhibited slightly elevated calcium flux in both CD45RA+CCR7+ naïve and CD45RA-CCR7- effector memory cells (Fig 6.2D). Despite in vitro activated CD8+ T cells from XLT patients demonstrating reduced expression of activation markers, calcium flux of naïve CD8+ T cells was higher than a healthy donor while calcium flux in effector memory CD8 TEM cells from the XLT patient and healthy donors was comparable, indicating intact TCR-induced activation (Fig 6.2E).

141

Figure 6.2 CD8+ T cell function in patients with mutations in DOCK8-related proteins Sorted CD8+ T cells from healthy donors (n=3), a CDC42-deficient patient, an XLT patient and a DOCK2-deficient patient were cultured for 5 days with TAE beads. On day 5 supernatants were harvested and cells were restimulated with PMA/ionomycin and Brefeldin A and (A) stained for surface expression of CD25, CD69, CD95 and CD107a (B) Cells were fixed, permeabilised and intracellularly stained for IFNγ, IL2, Granzyme B, Perforin and TNFα. (C) Cytokine secretion in supernatants was measured by CBA. PBMCs from a healthy control and either a (D) WASP patient or (E) XLT patient were loaded with indo-1AM and stained for CD8, CCR7 and CD45RA. Cells were stimulated with anti-CD3 and crosslinked to induce calcium flux with percentage of responding cells measured by flow cytometry over a period of 7 minutes in naïve and effector memory CD8+ T cells. Error bars represent SEM.

142

6.2.4 CD4+ T cell function in patients with mutations in DOCK8-related proteins To further understand mechanisms underlying recurrent infections in patients with mutations in DOCK8-related proteins, the function of CD4+ T cells from these patients was examined. Sorted CD45RA- memory CD4+ T cells were cultured for 5 days with TAE beads and supernatants were collected for analysis of cytokine secretion before cells were restimulated with PMA/ionomycin in the presence of Brefeldin A to measure intracellular cytokine production. This revealed that while Th1 cytokine (i.e. IFNγ and TNFα) production and secretion was normal in the CDC42-deficient patient, memory CD4+ T cells from WAS, XLT and DOCK2-deficient patients had diminished intracellular expression and secretion of Th1 cytokines (Fig 6.3A). Intracellular expression of the Th2 cytokines IL4 or IL13 was compromised in all patients and all patients except XLT patients secreted less cumulative Th2 cytokines (Fig 6.3B). Th17 cytokine (i.e.IL17A, IL17F and IL22) production and secretion by the CDC42-deficient patient was mostly intact while the WAS and XLT patients showed reduced intracellular production, but normal cytokine secretion. The DOCK2-deficient patient displayed high intracellular cytokine expression, but reduced cytokine secretion, suggesting poor cell survival in vitro (Fig 6.3C). Measurement of calcium flux in CD4+ T cells from 1 WAS and 1 XLT patient enabled assessment of the response to TCR-induced stimulation in naïve and memory CD4+ T cells. Calcium flux was slightly elevated in both naïve and memory CD4+ T cells from the WAS patient (Fig 6.3D) while it was increased for both subsets in the XLT patient compared to a healthy donor (Fig 6.3E), demonstrating the ability of these cells to become activated by TCR signalling.

143

Figure 6.3 CD4+ T cell function in patients with mutations in DOCK8-related proteins Sorted memory CD4+ T cells from healthy donors (n=6), a CDC42-deficient patient, WAS patients (n=2), XLT patients (n=2) and a DOCK2-deficient patient were cultured for 5 days with TAE beads. On day 5 cells were restimulated with PMA/ionomycin and Brefeldin A, fixed, permeabilised and intracellularly stained for cytokine expression. Supernatants were also harvested to determine cytokine secretion. Intracellular expression and secretion of (A) Th1 (IFNγ, TNFα) cytokines, (B) Th2 (IL4, IL5, IL13) cytokines and (C) Th17 (IL17A, IL17F, IL22) cytokines was determined by flow cytometry and CBA. PBMC from a healthy control and either a WAS patient (D) or XLT patient (E) were loaded with indo-1AM and stained for CD4, CCR7 and CD45RA. Cells were stimulated with anti-CD3 and crosslinked to induce calcium flux with percentage of responding cells measured by flow cytometry over a period of 7 minutes in naïve and memory CD4+ T cells. Error bars represent SEM. 144

6.2.5 Naïve B cell function in patients with mutations in DOCK8-related proteins Naïve B cell function was assessed in patients with mutations in DOCK8-related proteins to compare to the defects in DOCK8-deficient patients. Sorted naïve B cells were cultured for 5 days with combinations of CD40L and IL21 to simulate CD4+ T cell help, CpG to act as a TLR agonist and SAC to mimic BCR engagement. Cell number was calculated using Calibrite beads and culture supernatants were collected to measure Ig secretion by ELISA. The CDC42- and DOCK2-deficient patients were assessed for cell recovery after 5 days of culture and it was found to be extremely low in all stimulus conditions tested (Fig 6.4A). Naïve B cells from healthy donor cells secreted detectable amounts of IgG after culture for 5 days with CD40L+IL21, but IgG secretion was compromised in all patients with no secretion detected by B cells from the CDC42- or DOCK2-deficient patients. This is likely due to their poor survival in vitro (Fig 6.4B). Culture with CD40L+IL21 or CD40L+CpG+SAC induced IgA production in healthy donor B cells, but again the CDC42- and DOCK2-deficient B cells secreted much less IgA under both conditions while WAS and XLT patients produced detectable, but decreased amounts of IgA (Fig 6.4C). IgM secretion was undetectable in cultures of naïve CDC42- and DOCK2-deficient B cells and reduced in WAS and XLT B cells compared to normal control B cells (Fig 6.4D).

145

Figure 6.4 Naive B cell function in patients with mutations in DOCK8-related proteins Sorted naïve B cells from healthy donors (n=2-5), a CDC42-deficient patient, WAS patients (n=2), a XLT patient and a DOCK2-deficient patient were cultured for 5 days with stimuli to mimic a combination of T cell help (CD40L, IL21), BCR engagement (SAC) and TLR signalling (CpG). (A) Cell recovery of a CDC42-deficient patient and a DOCK2-deficient patient was quantified using Calibrite beads. Supernatants from all patients were harvested to determine the secretion of (B) IgG, (C) IgA and (D) IgM by ELISA. Error bars represent SEM.

146

6.3 Discussion The analysis presented here of the phenotype and function of lymphocytes in a CDC42- deficient patient revealed mechanistic explanations for some of the clinical features reported in these patients. Reduced activation of CD8+ T cells as well as their impaired ability to produce and secrete cytokines would render patients more vulnerable to pathogens, resulting in frequent and recurrent infections198,199. The types of infections experienced have not been specified in all cases and could include fungal infections. Indeed, lower levels of Th17 cytokine secretion seen in the patient examined here would likely compromise their defence against fungal pathogens212 . One previously reported patient with the same mutation as the patient studied here (p.Y64C) exhibited low serum Ig levels and was unable to produce protective antibodies upon vaccination against measles, rubella and VZV199. These clinical observations could be explained by the findings made here. Specifically, a reduced frequency of total and memory B cells as well as impaired survival of and Ig secretion by naïve B cells following in vitro culture. Martinelli et al. suggested that CDC42-deficient patients can be divided into 3 groups according to which structural region or domain of the protein is disrupted by their mutation and that these groups segregate with the diversity of the phenotypes seen198. The patient studied here with a p.Y46C mutation would be placed in group I, a group more commonly exhibiting immunodeficiency. Thus, it would be interesting to determine whether our findings of altered lymphocyte phenotype and function are unique to group 1 CDC2-deficient patients, or whether these are also characteristic of patients classified into the other two groups.

WAS patients suffer from bacterial, viral and fungal infections, most commonly pneumonia and otitis media, varicella and yeast infections, which are often recurrent and can be fatal202. Our data on CD4+ T cells revealed reduced production of IFN and TNF as well as IL17A and IL22. Combined with a reduction in the frequency of CD4+ T cells, these cellular features of WAS could account for deficits in protection against various pathogens and the subsequent development of infectious diseases. Furthermore, while limited functional data on CD8+ T cells was able to be presented here, the observed phenotype of these cells indicates that they may also have impaired functional capacity due to chronic activation/exhaustion, as seen by the increase in terminally differentiated

TEMRA cells and acquisition of an exhausted phenotype. Remarkably, calcium flux of naïve and effector memory WAS CD8+ T cells was not diminished suggesting activation 147

via the TCR is not strictly impaired in these cells. WAS patients have also been reported to have impaired NK cell cytotoxicity which could contribute to their infectious susceptibility204,213. These repeated infections, resulting in continued cellular proliferation, likely contribute to the increased incidence of hematologic malignancy in WAS patients which accounts for the majority of malignancies diagnosed in these patients202,214. Eczema is another classic defining feature of WAS, affecting 81-92% of patients202,214. Given the link between Th2 cells and eczema, increased levels of Th2 cytokine production by WAS patient memory CD4+ T cells would be predicted. However, this was not seen. Further insight into causes of eczema could be gained by measuring serum IgE levels in these patients as it is known that WAS patients demonstrate elevated levels of IgE215. Ab responses to a range of vaccines have been tested in a cohort of WAS patients and found to be abnormal in the majority of patients across most of the immunogens examined202. Our data showing a reduction in memory B cells as well as compromised naïve B cell survival and IgG secretion could play a role in this impairment.

Despite XLT patients generally demonstrating milder clinical symptoms than WAS patients, our data do not provide a clear explanation for why this occurs. The only phenotypic difference between the WAS and XLT patients studied here was the higher frequency of CD4+ T cells in XLT patients and the increased NK cell frequency in WAS patients. Functionally, both XLT and WAS patients showed similar impairments in Th1 and Th17 cytokine production by CD4+ memory T cells and Ig secretion by naïve B cells. However, as CD8+ T cell function was not analysed in WAS patients, the level of dysfunction in this lymphocyte subset could underlie some clinical differences in WAS and XLT patients. Additionally, high expression of WASp has been reported during differentiation of different hematopoietic pathways216 and hence the effects of WAS mutations on other immune cells including monocytes and dendritic cells could also underlie clinical variation between WAS and XLT patients.

There has been debate over classifying patients as either WAS or XLT due to the variation in clinical presentation of patients. A clinical scoring system was developed ranging from 1, for patients with thrombocytopenia only, through to 5, for patients with thrombocytopenia plus severe eczema and/or frequent potentially life-threatening infections as well as autoimmunity or malignancy201,217 . XLT patients are classified as scoring 2-3 while WAS patients are scored as 4-5201,217. WASpbase (pidj.rcai.riken.jp/waspbase) is a database cataloguing 441 cases of mutations in WAS and 148

the clinical score assigned to each patient218. The majority of patients with the same mutation show the same clinical score but there are some patients with different clinical scores despite having the same mutation. This results in a differential classification as either XLT or WAS and highlights the difficulty present in segregating these patient cohorts. An alternate definition of either WASp+ (presence of normal sized protein) or WASp- (absent or truncated protein) has been proposed to correlate with severity of symptoms experienced, but exceptions have also been reported with this classification method214,219. Of the two WAS and two XLT patients studied here, one WAS patient was confirmed to lack WASp protein expression by western blot but the others have not been tested for WASp expression. Classification of all the patients studied here as either WASp+ or WASp- would provide further information into how lymphocyte phenotype and function differ between WAS and XLT patients.

The most reported clinical symptom of DOCK2-deficient patients is severe bacterial and viral infections208,209,211. Our data suggests that this could manifest from several impairments in adaptive immunity. The low numbers of total CD8+ T cells and a lack of naïve CD4+ and CD8+ T cells, coupled with defective activation and cytokine secretion of CD8+ T cells, are likely the dominant contributing factors to the compromised infectious immunity in DOCK2-deficient patients. Additionally, the reduced frequency of Th1 cells producing IFNγ and TNFα, which are important for anti-bacterial and antiviral defence, would contribute to increased susceptibility, severity and occurrence of infections. Furthermore, intracellular expression of Th17 cytokines is intact in this patient which correlates with the lack of reports of fungal infections in DOCK2 patients. The ability of patients to respond to previously encountered infection-causing pathogens would also be impaired due to the B cell deficiencies observed, including decreased total and memory B cells, poor survival of naïve B cells and lack of Ig production. Other reports have documented impaired NK cell function and antiviral interferon responses which would also contribute to patient susceptibility to infections208.

This investigation of the phenotype and function of patients with mutations in DOCK8- related proteins enables a comparison with the hallmark features of DOCK8 deficiency as uncovered in chapter 4 (Table 6.2). On the whole, all the patient cohorts shared several features with DOCK8 deficiency but also differed in some aspects. Phenotypically, the DOCK2-deficient patient most paralleled DOCK8-deficient patients with substantial T cell lymphopenia, T cell memory skewing, decreased memory B cells and decreased 149

MAIT cells but did not exhibit T cell exhaustion and had decreased rather than increased total B cells. The WAS patients and the CDC42-deficient patient displayed some of the same characteristics but to a lesser extent while the XLT patients demonstrate only moderate T cell lymphopenia and increased total B cells. Similar to DOCK8-deficient patients, all tested cohorts displayed impaired CD8+ T cell activation and cytokine secretion, as well as decreased survival of and Ig secretion by B cells. Production of all cytokines was normal in the CDC42-deficient patient while only the DOCK2-deficient patient exhibited increased Th2 cytokine production. Yet decreased Th1 cytokine production was found for the DOCK2-deficient patient as well as the WAS and XLT patients. Th17 cytokine production was decreased in the WAS and XLT patients. It is interesting that while CDC42 and WASp have been shown to interact with DOCK8 in IS formation and migration pathways, it is the DOCK2-deficient patient that most closely mimicked phenotypic and functional characteristics of DOCK8-deficient patients. DOCK2, like DOCK8, is highly expressed in hematopoietic cells and implicated in cell migration through its function as a GEF for the Rho family GTPase Rac220. The similarities between the DOCK2-deficient patient and DOCK8-deficient patients implies that their associated pathways are involved in analogous cellular processes but both are required for immune cell function. However, the fact that DOCK2 deficiency is more severe than DOCK8 deficiency, both functionally and clinically, suggests that DOCK2 may be involved in more diverse cellular processes. The discovery that the CDC42- deficient patient did not more closely replicate the features of DOCK8-deficient patients could be due to the ability of other GEFs to activate the CDC42 protein in the absence of DOCK8. In fact, DOCK9 is also expressed in hematopoietic cells and has been shown to interact with CDC42221. Additionally, CDC42 interacts with a range of proteins including kinases and scaffold proteins85 and is expressed in all tissues so the effects of mutations in CDC42 could be diverse. Similarly, the discrepancies between WAS/XLT patients and DOCK8-deficient patients could be due to the ability of WASp to function in a manner which does not rely on upstream DOCK8 function as it can be activated by factors other than CDC42 such as NCK1 and NCK2222.

The work in this chapter identifies shared and unique lymphocyte defects between DOCK8-deficient patients and patients with mutations in DOCK8-related proteins as well as shedding light on the impact of known signalling pathways on the cellular defects of DOCK8 deficiency. 150

Table 6.2 Comparison of immunological features between DOCK8-deficient patients and the patients studied in this chapter

DOCK8 patients CDC42 WAS XLT DOCK2 patient patients patients patient T cell lymphopenia NN Y Y YY T cell memory skewing Y N N YY T cell exhaustion N Y N N Phenotype Increased total B cells NN N Y NN Decreased memory B cells Y Y Y YY Decreased MAIT cells Y Y N YY CD8+ T cell Decreased activation Y ND Y Y function Decreased cytokine Y ND Y Y secretion CD4+ T cell Increased Th2 cytokines N N Y Y function Decreased Th1 cytokines N Y Y Y Decreased Th17 cytokines N Y Y N B cell Decreased survival Y Y Y Y function Decreased Ig secretion Y Y Y Y Y-conforms, YY-strongly conforms, N-doesn’t conform, NN-strongly doesn’t conform, ND-not done

151

Chapter 7: Conclusions

7.1 Research Outcomes My PhD thesis has provided an in-depth and detailed investigation into several aspects of human DOCK8 deficiency. The use of different cohorts of patients, including DOCK8- deficient patients, DOCK8-deficient patients who had undergone HSCT, a DOCK8- deficient patient with somatic reversion and patients with mutations in proteins which are functionally related to DOCK8, identified the important role of DOCK8 in a range of immune cell functions. The findings presented here reveal the value of identifying the perturbed immune cell function, due to monogenic germline variants, which causes clinical symptoms of immunodeficiency and immune dysregulation. Investigations such as this provide critical insight into the non-redundant functions of genes, molecules and pathways in immune cell development and function, elucidate mechanisms underlying disease pathogenesis, and potentially enable improved treatments for affected individuals.

The work presented here has really highlighted the importance of different CD4+ Th subsets in control of specific pathogens. The requirement for Th17 cells to prevent fungal infections is seen through multiple results. Firstly, DOCK8-deficient patients report an increased incidence and severity of fungal infections, compared to the general population, and it was shown that they had a lack of Th17 cells. Secondly, DOCK8-deficient patients who had received HSCT had an increased frequency of Th17 cells compared to untransplanted patients and experienced fewer and less severe incidences of fungal infections than those without HSCT. Thirdly, the DOCK8-revertant patient studied here was free of fungal infections and found to have normal levels of Th17 cells. Lastly, analysis of patients with mutations in DOCK8-related proteins revealed that the patients who did not acquire fungal infections (CDC42-deficient and DOCK2-deficient patients) were also the patients who had intact Th17 cell differentiation. In the same way, insight was also gained in relation to the importance of Th1 cells in defence against bacterial infections. Th1 cells were decreased in DOCK8-deficient patients who report recurrent bacterial infections and Th1 cells were shown to be increased following HSCT with less cases of bacterial infections documented in HSCT patients. The DOCK8-revertant patient did not report bacterial infections and also had normal levels of Th1 cell differentiation.

152

Identification of pathways involved in CD4+ T cell differentiation has applications well outside the context of DOCK8 deficiency. As detailed here, a balance of differentiated CD4+ T cell subsets is vital to maintaining a healthy immune system as selective impairments lead to susceptibility to infections but overabundance can also have negative consequences. This thesis uncovered that DOCK8-deficiency leads to Th2 cell skewing; thus, modulation of DOCK8 function could be used to influence CD4+ T cell differentiation. This would be advantageous in the setting of allergic conditions, which result from excess Th2 cells and include asthma, eczema and allergies, which are experienced by a large proportion of the population. Indeed, the World Allergy Organisation’s white book on allergy estimates that globally allergic rhinitis affects 10- 30% of adults, atopic eczema affects 15-30% of children, 300 million people suffer from asthma and 220-520 million experience food allergies. Manipulation of CD4+ T cell differentiation could also be beneficial in inflammatory conditions, which result from excess Th1 and Th17 cells, including rheumatoid arthritis and diabetes which affect a significant number of people (457,200 and 1.2 million Australians respectively in 2017- 2018; Australian Bureau of Statistics, National Health Survey: First Results, 2017-18). While the advantages of manipulating CD4+ T cell differentiation are clear, the work presented here also calls attention to some important considerations for this potential modulation. The results here indicate that CD4+ T cell differentiation is not driven solely by cytokine stimulation and that TCR stimulation also plays a role. Hence, while cytokine concentration could be therapeutically altered, CD4+ T cell differentiation decisions may be overridden by the strength of TCR signalling and costimulation receive by cells, which is much more challenging to control. The differing ability to undergo Th1 cell differentiation found in this thesis between in vivo and in vitro differentiated DOCK8- deficient cells also raised the possibility that cells were more sensitive to IL4 which was only evident when cells developed in the presence of a complete cytokine milieu (i.e. in vivo). This would be an important factor in developing a treatment to influence CD4+ T cell differentiation as increased sensitivity to particular cytokines in different patients would mean that the same treatment could not be used on all patients and alterations would be required for different cohorts.

The immunological outcome of HSCT presented in this thesis provides valuable information to clinicians for their patient treatment and care. A recent update on primary

153

immunodeficiency worldwide reported ~4400 patients had received HSCT in the past year223, however this does not take into account cases of HSCT for causes other than immunodeficiency i.e. cancers of the blood and bone marrow, such as multiple myeloma and leukaemia. Hence, the data presented here on immune reconstitution and function across various times post-transplant could be applied to a large number of cases to guide treatment to address remaining immune weaknesses.

7.2 Future Directions The findings of this thesis could be built upon with further investigation of the specific role of DOCK8 in immune processes in different lymphocytes. Our results currently suggest that the decreased Th1 cell differentiation seen in DOCK8-deficient memory CD4+ T cells may be due to weakened in vivo TCR signalling which is obscured during in vitro culture of naïve CD4+ T cells. This theory is supported by the results in the examined DOCK8-revertant patient CD4+ memory T cells in which DOCK8-expressing cells underwent normal Th1 cell differentiation to produce IFNγ while DOCK8-deficient cells did not despite the fact that both populations had developed in the same environment. Development of in vitro culture conditions that more closely mimic the in vivo environment may enable better characterisation of the role of DOCK8 in TCR signalling in CD4+ T cells. For example, culturing patient naïve CD4+ T cells with titrated amounts of plate-bound anti-CD3 and soluble anti-CD28 and anti-CD2 would enable more precise manipulation of the strength of TCR and costimulatory stimuli supplied. This may resolve the curious finding that Th1 cell differentiation is diminished in vivo but not in vitro in DOCK8-deficient cells.

This thesis showed an impaired B cell compartment in DOCK8-deficient patients through the analysis of phenotype, survival, proliferation and antibody secretion. Further exploration into the specific mechanisms of impairment could be taken with investigation of STAT3 phosphorylation. Reduced STAT3 phosphorylation in response to cytokine stimulation has been reported in T cells from DOCK8-deficient patients leading to reduced expression of STAT3-inducible genes54. The importance of STAT3 in the differentiation of B cells and their ability to secrete antibodies has previously been highlighted in patients with mutations in STAT3224. Hence, it would be beneficial to analyse STAT3 phosphorylation in DOCK8-deficient B cells to establish whether 154

impairments in B cell function in DOCK8-deficient patients are due to perturbations in STAT3 signalling.

Additionally, continued analysis of DOCK8-deficient HSCT patients as they progress further in time post-transplant would be of merit to extend the conclusions of the current work. This is particularly important as most of the patients analysed were investigated within the first 6-18 months of their HSCT. While we saw correction of differentiation of most lymphocyte subsets some had not recovered even in patients >23 months post HSCT (e.g. NKT and MAIT cells) although this was based on only a small number of patients. Hence, continuing to assess HSCT patient lymphocytes would strengthen our findings and also reveal whether the dynamics of lymphocyte recovery vary in different patients. Although DOCK8-deficient HSCT patient clinical outcomes were very positive with respect to infection, the impact of HSCT on malignancy remains unknown. One would expect that the improved control of viral infections would decrease the likelihood of malignancy given that there would be less cellular proliferation and hence less opportunity for oncogenic mutations to accumulate. Additionally, malignancy can result from integration of some viruses into the host cell DNA with one example being Epstein- Barr virus (EBV) leading to lymphoma which is one of the most commonly reported malignancies in DOCK8-deficient patients40,51. Similarly, the outcome of allergies in transplanted DOCK8-deficient patients could be better elucidated through continued study of patients as time post HSCT increases, as demonstrated by our data showing decreases in IgE levels over time.

Likewise, longitudinal measurement of proportions of DOCK8 revertant cells and corresponding function in the revertant patient presented in this thesis would shed further light on the kinetics and nature of the reversion process in different lymphocytes, and how restoration of DOCK8 expression in DOCK8-deficient cells results in a clinically beneficial outcome.

Further insight into the importance of the interaction of DOCK8 with CDC42 in lymphocyte function could be gained through the assessment of lymphocytes in a greater number of patients with mutations in CDC42 in order to compare to lymphocytes from DOCK8-deficient patients. While DOCK8-deficient patient T cells have been shown to

155

have less activated WASp upon TCR stimulation in vitro89, the requirement for DOCK8 in activation of WASp in other lymphocytes could be better assessed after comparison of DOCK8-deficient patients with more WAS and XLT patients.

While the use of human patients can provide a wealth of information about the role of specific genes, molecules and pathways in immune cell function, it has the inherent caveat that the pursuit of in vivo studies is limited, but ethically and logistically. This is relevant in the case of DOCK8 as it has been implicated in the processes of IS formation and immune cell migration, mostly through the results of in vivo experiments in Dock8- deficient mice or in vitro experiments using antigen-specific lymphocytes generated from genetically-modified mice (see Introduction). The processes of IS formation and migration, however, are often excluded from in vitro systems using human cells. While this thesis has detailed numerous defects in different lymphocyte populations and subsets that contribute to the clinical features of DOCK8-deficient patients, it is probable that a reduced ability of DOCK8-deficient cells to migrate to sites of interaction and form stable interactions via the IS also plays a role in the pathogenesis of DOCK8 deficiency. For example, and as discussed in Chapters 3 and 5, decreased IS formation between CD4+ T cells and APCs could impact TCR signalling strength and subsequent CD4+ T cell differentiation. Additionally, while the work presented here revealed functional defects in CD8+ T cells in vitro, that would likely result in diminished viral control in vivo, the decreased ability of these cells in vivo to migrate to and form stable interactions with target cells would contribute to the increased incidence and severity of viral infections. Furthermore, in our experiments with B cells, activation via T cell help was mimicked by CD40L and IL21. However, DOCK8-deficient B cells would likely be affected by a reduced ability to form interactions with cognate T cells and thus receive activating signals. Hence, continued work in mouse models of Dock8 deficiency and experiments with human cells that are more able to recapitulate the in vivo processes involved in achieving lymphocyte function would enable further elucidation of the hitherto unknown functions of DOCK8 in the immune system.

Ultimately, this work illustrates that the investigation of primary immunodeficiency patients, sometimes referred to as “experiments of nature”, can greatly contribute to our

156

knowledge not only of the affected patients but also of the function of the immune system with the possibility to benefit a much wider community.

157

References

1. Picard C, Bobby Gaspar H, Al-Herz W, et al. International Union of Immunological Societies: 2017 Primary Immunodeficiency Diseases Committee Report on Inborn Errors of Immunity. J Clin Immunol. 2018;38(1):96-128. 2. Casanova J-L. Human genetic basis of interindividual variability in the course of infection. Proc Natl Acad Sci U S A. 2015;112(51):E7118. 3. Bruton OC. Agammaglobulinemia. Pediatrics. 1952;9(6):722-728. 4. Hitzig WH, Biro Z, Bosch H, Huser HJ. [Agammaglobulinemia & alymphocytosis with atrophy of lymphatic tissue]. Helv Paediatr Acta. 1958;13(6):551- 585. 5. Tobler R, Cottier H. [Familial lymphopenia with agammaglobulinemia & severe moniliasis: the essential lymphocytophthisis as a special form of early childhood agammaglobulinemia]. Helv Paediatr Acta. 1958;13(4):313-338. 6. Giblett ER, Anderson JE, Cohen F, Pollara B, Meuwissen HJ. Adenosine- deaminase deficiency in two patients with severely impaired cellular immunity. Lancet. 1972;2(7786):1067-1069. 7. Noguchi M, Yi H, Rosenblatt HM, et al. Interleukin-2 receptor gamma chain mutation results in X-linked severe combined immunodeficiency in humans. Cell. 1993;73(1):147-157. 8. Puck JM, Deschenes SM, Porter JC, et al. The interleukin-2 receptor γ chain maps to Xq13. 1 and is mutated in X-linked severe combined immunodeficiency, SCIDX1. Hum Mol Genet. 1993;2(8):1099-1104. 9. Tsukada S, Saffran DC, Rawlings DJ, et al. Deficient expression of a B cell cytoplasmic tyrosine kinase in human X-linked agammaglobulinemia. Cell. 1993;72(2):279-290. 10. Vetrie D, Vorechovsky I, Sideras P, et al. The gene involved in X-linked agammaglobulinaemia is a member of the src family of protein-tyrosine kinases. Nature. 1993;361(6409):226-233. 11. Janeway CA, Craig J, Davidson M, Downey W, Gitlin D. Hypergammaglobulinemia associated with severe recurrent and chronic nonspecific infection. AMA Am J Dis Child. 1954:388-392. 12. Mayer L, Kwan SP, Thompson C, et al. Evidence for a defect in "switch" T cells in patients with immunodeficiency and hyperimmunoglobulinemia M. N Engl J Med. 1986;314(7):409-413. 13. Allen RC, Armitage RJ, Conley ME, et al. CD40 ligand gene defects responsible for X-linked hyper-IgM syndrome. Science. 1993;259(5097):990-993. 14. Aruffo A, Farrington M, Hollenbaugh D, et al. The CD40 ligand, gp39, is defective in activated T cells from patients with X-linked hyper-IgM syndrome. Cell. 1993;72(2):291-300. 15. DiSanto JP, Bonnefoy JY, Gauchat JF, Fischer A, de Saint Basile G. CD40 ligand mutations in x-linked immunodeficiency with hyper-IgM. Nature. 1993;361(6412):541- 543. 16. Fuleihan R, Ramesh N, Loh R, et al. Defective expression of the CD40 ligand in X chromosome-linked immunoglobulin deficiency with normal or elevated IgM. Proc Natl Acad Sci U S A. 1993;90(6):2170-2173. 17. Korthauer U, Graf D, Mages HW, et al. Defective expression of T-cell CD40 ligand causes X-linked immunodeficiency with hyper-IgM. Nature. 1993;361(6412):539- 541.

158

18. Noelle RJ, Ledbetter JA, Aruffo A. CD40 and its ligand, an essential ligand- receptor pair for thymus-dependent B-cell activation. Immunol Today. 1992;13(11):431- 433. 19. Ferrari S, Giliani S, Insalaco A, et al. Mutations of CD40 gene cause an autosomal recessive form of immunodeficiency with hyper IgM. Proc Natl Acad Sci U S A. 2001;98(22):12614-12619. 20. Carneiro-Sampaio M, Coutinho A. Immunity to Microbes: Lessons from Primary Immunodeficiencies. Infect Immun. 2007;75(4):1545. 21. Milner JD, Holland SM. The cup runneth over: lessons from the ever-expanding pool of primary immunodeficiency diseases. Nat Rev Immunol. 2013;13:635. 22. Ochs HD, Hitzig WH. History of primary immunodeficiency diseases. Curr Opin Allergy Clin Immunol. 2012;12(6):577-587. 23. Bustamante J, Boisson-Dupuis S, Abel L, Casanova J-L. Mendelian susceptibility to mycobacterial disease: genetic, immunological, and clinical features of inborn errors of IFN-γ immunity. Sem Immunol. 2014;26(6):454-470. 24. Puel A, Cypowyj S, Maródi L, Abel L, Picard C, Casanova J-L. Inborn errors of human IL-17 immunity underlie chronic mucocutaneous candidiasis. Curr Opin Allergy Clinical Immunol. 2012;12(6):616-622. 25. Bousfiha A, Jeddane L, Picard C, et al. The 2017 IUIS Phenotypic Classification for Primary Immunodeficiencies. J Clin Immunol. 2018;38(1):129-143. 26. Kazatchkine MD, Kaveri SV. Immunomodulation of autoimmune and inflammatory diseases with intravenous immune globulin. N Engl J Med. 2001;345(10):747-755. 27. Holland SM. Cytokine therapy of mycobacterial infections. Adv Intern Med. 2000;45:431-452. 28. Hershfield MS. PEG-ADA replacement therapy for adenosine deaminase deficiency: an update after 8.5 years. Clin Immunol Immunopathol. 1995;76(3 Pt 2):S228- 232. 29. Bach FH, Albertini RJ, Joo P, Anderson JL, Bortin MM. Bone-marrow transplantation in a patient with the Wiskott-Aldrich syndrome. Lancet. 1968;2(7583):1364-1366. 30. Gatti RA, Meuwissen HJ, Allen HD, Hong R, Good RA. Immunological reconstitution of sex-linked lymphopenic immunological deficiency. Lancet. 1968;2(7583):1366-1369. 31. Slatter MA, Gennery AR. Hematopoietic cell transplantation in primary immunodeficiency – conventional and emerging indications. Exp Rev Clin Immunol. 2018;14(2):103-114. 32. Grimbacher B, Schäffer AA, Holland SM, et al. Genetic linkage of hyper-IgE syndrome to chromosome 4. Am J Hum Genet. 1999;65(3):735-744. 33. Grimbacher B, Holland SM, Gallin JI, et al. Hyper-IgE Syndrome with Recurrent Infections — An Autosomal Dominant Multisystem Disorder. New Eng J Med. 1999;340(9):692-702. 34. Renner ED, Puck JM, Holland SM, et al. Autosomal recessive hyperimmunoglobulin E syndrome: a distinct disease entity. J Pediatr. 2004;144(1):93- 99. 35. Minegishi Y, Saito M, Tsuchiya S, et al. Dominant-negative mutations in the DNA-binding domain of STAT3 cause hyper-IgE syndrome. Nature. 2007;448(7157):1058-1062. 36. Holland SM, DeLeo FR, Elloumi HZ, et al. STAT3 mutations in the hyper-IgE syndrome. N Engl J Med. 2007;357(16):1608-1619. 159

37. Sowerwine KJ, Holland SM, Freeman AF. Hyper-IgE syndrome update. Ann N. Y. Acad Sci. 2012;1250:25-32. 38. Minegishi Y, Saito M, Morio T, et al. Human Tyrosine Kinase 2 Deficiency Reveals Its Requisite Roles in Multiple Cytokine Signals Involved in Innate and Acquired Immunity. Immunity. 2006;25(5):745-755. 39. Woellner C, Schaffer AA, Puck JM, et al. The hyper IgE syndrome and mutations in TYK2. Immunity. 2007;26(5):535. 40. Engelhardt KR, McGhee S, Winkler S, et al. Large deletions and point mutations involving the dedicator of cytokinesis 8 (DOCK8) in the autosomal-recessive form of hyper-IgE syndrome. J Allergy Clin Immunol. 2009;124(6):1289-1302 41. Zhang Q, Davis JC, Lamborn IT, et al. Combined immunodeficiency associated with DOCK8 mutations. N Engl J Med. 2009;361(21):2046-2055. 42. Sassi A, Lazaroski S, Wu G, et al. Hypomorphic homozygous mutations in phosphoglucomutase 3 (PGM3) impair immunity and increase serum IgE levels. J Allergy Clin Immunol. 2014;133(5):1410-1419. 43. Zhang Y, Yu X, Ichikawa M, et al. Autosomal recessive phosphoglucomutase 3 (PGM3) mutations link glycosylation defects to atopy, immune deficiency, autoimmunity, and neurocognitive impairment. J Allergy Clin Immunol. 2014;133(5):1400-1409. 44. Ma CA, Stinson JR, Zhang Y, et al. Germline hypomorphic CARD11 mutations in severe atopic disease. Nat Genet. 2017;49(8):1192-1201. 45. Beziat V, Li J, Lin JX, et al. A recessive form of hyper-IgE syndrome by disruption of ZNF341-dependent STAT3 transcription and activity. Sci Immunol. 2018;3(24). 46. Dadi H, Jones TA, Merico D, et al. Combined immunodeficiency and atopy caused by a dominant negative mutation in caspase activation and recruitment domain family member 11 (CARD11). J Allergy Clin Immunol. 2018;141(5):1818-1830. 47. Frey-Jakobs S, Hartberger JM, Fliegauf M, et al. ZNF341 controls STAT3 expression and thereby immunocompetence. Sci Immunol. 2018;3(24). 48. Schwerd T, Twigg SRF, Aschenbrenner D, et al. A biallelic mutation in IL6ST encoding the GP130 co-receptor causes immunodeficiency and craniosynostosis. J Exp Med. 2017;214(9):2547-2562. 49. Dorjbal B, Stinson JR, Ma CA, et al. Hypomorphic caspase activation and recruitment domain 11 (CARD11) mutations associated with diverse immunologic phenotypes with or without atopic disease. J Allergy Clin Immunol. 2019;143(4):1482- 1495. 50. Shahin T, Aschenbrenner D, Cagdas D, et al. Selective loss of function variants in IL6ST cause Hyper-IgE syndrome with distinct impairments of T-cell phenotype and function. Haematol. 2019;104(3):609-621. 51. Engelhardt KR, Gertz ME, Keles S, et al. The extended clinical phenotype of 64 patients with dedicator of cytokinesis 8 deficiency. J Allergy Clin Immunol. 2015;136(2):402-412. 52. Su HC, Jing H, Angelus P, Freeman AF. Insights into immunity from clinical and basic science studies of DOCK8 immunodeficiency syndrome. Imm Rev 2019;287(1):9- 19. 53. Jing H, Zhang Q, Zhang Y, et al. Somatic reversion in dedicator of cytokinesis 8 immunodeficiency modulates disease phenotype. J Allergy Clin Immunol. 2014;133(6):1667-1675.

160

54. Keles S, Charbonnier LM, Kabaleeswaran V, et al. Dedicator of cytokinesis 8 regulates signal transducer and activator of transcription 3 activation and promotes T(H)17 cell differentiation. J Allergy Clin Immunol. 2016;138(5):1384-1394. 55. Ruiz-García R, Lermo-Rojo S, Martínez-Lostao L, et al. A case of partial dedicator of cytokinesis 8 deficiency with altered effector phenotype and impaired CD8+ and natural killer cell cytotoxicity. J Allergy Clin Immunol. 2014;134(1):218-221. 56. Aydin SE, Kilic SS, Aytekin C, et al. DOCK8 deficiency: clinical and immunological phenotype and treatment options - a review of 136 patients. J Clin Immunol. 2015;35(2):189-198. 57. Chu EY, Freeman AF, Jing H, et al. Cutaneous Manifestations of DOCK8 Deficiency Syndrome. JAMA Dermat. 2012;148(1):79-84. 58. Alsum Z, Hawwari A, Alsmadi O, et al. Clinical, Immunological and Molecular Characterization of DOCK8 and DOCK8-like Deficient Patients: Single Center Experience of Twenty Five Patients. J Clin Immunol. 2013;33(1):55-67. 59. Jabara HH, McDonald DR, Janssen E, et al. DOCK8 functions as an adaptor that links Toll-like receptor–MyD88 signaling to B cell activation. Nat Immunol. 2012;13(6):612-620. 60. Sabry A, Hauk PJ, Jing H, et al. Vaccine strain varicella-zoster virus-induced central nervous system vasculopathy as the presenting feature of DOCK8 deficiency. J Allergy Clin Immunol. 2014;133(4):1225-1227. 61. AlKhater SA. CNS vasculitis and stroke as a complication of DOCK8 deficiency: a case report. BMC Neurol. 2016;16:54-54. 62. Sanal O, Jing H, Ozgur T, et al. Additional Diverse Findings Expand the Clinical Presentation of DOCK8 Deficiency. J Clin Immunol. 2012;32(4):698-708. 63. Vinci G, Chantot-Bastaraud S, El Houate B, Lortat-Jacob S, Brauner R, McElreavey K. Association of deletion 9p, 46,XY gonadal dysgenesis and autistic spectrum disorder. Mol Hum Reprod. 2007;13(9):685-689. 64. Griggs BL, Ladd S, Saul RA, DuPont BR, Srivastava AK. Dedicator of cytokinesis 8 is disrupted in two patients with mental retardation and developmental disabilities. Genomics. 2008;91(2):195-202. 65. Randall KL, Chan SS, Ma CS, et al. DOCK8 deficiency impairs CD8 T cell survival and function in humans and mice. J Exp Med. 2011;208(11):2305-2320. 66. Lambe T, Crawford G, Johnson AL, et al. DOCK8 is essential for T-cell survival and the maintenance of CD8+ T-cell memory. Eur J Immunol. 2011;41(12):3423-3435. 67. Janssen E, Morbach H, Ullas S, et al. DOCK8 deficient patients have a breakdown in peripheral B cell tolerance and defective regulatory T cells. J Allergy Clin Immunol. 2014;134(6):1365-1374. 68. Mizesko MC, Banerjee PP, Monaco-Shawver L, et al. Defective actin accumulation impairs human natural killer cell function in patients with dedicator of cytokinesis 8 deficiency. J Allergy Clin Immunol. 2013;131(3):840-848. 69. Qin T, An Y, Liu C, et al. Novel DOCK8 gene mutations lead to absence of protein expression in patients with hyper-IgE syndrome. Immunol Res. 2016;64(1):260- 271. 70. Ham H, Guerrier S, Kim J, et al. Dedicator of cytokinesis 8 interacts with talin and Wiskott-Aldrich syndrome protein to regulate NK cell cytotoxicity. J Immunol. 2013;190(7):3661-3669. 71. Kearney CJ, Vervoort SJ, Ramsbottom KM, et al. DOCK8 Drives Src-Dependent NK Cell Effector Function. J Immunol. 2017;199(6):2118-2127. 72. Crawford G, Enders A, Gileadi U, et al. DOCK8 is critical for the survival and function of NKT cells. Blood. 2013;122(12):2052-2061. 161

73. Al-Zahrani D, Raddadi A, Massaad M, et al. Successful interferon-alpha 2b therapy for unremitting warts in a patient with DOCK8 deficiency. Clin Immunol. 2014;153(1):104-108. 74. Keles S, Jabara HH, Reisli I, et al. Plasmacytoid dendritic cell depletion in DOCK8 deficiency: rescue of severe herpetic infections with IFN-alpha 2b therapy. J Allergy Clin Immunol. 2014;133(6):1753-1755. 75. Liu Y-J. IPC: Professional Type 1 Interferon-Producing Cells and Plasmacytoid Dendritic Cell Precursors. Ann Rev Immunol. 2004;23(1):275-306. 76. Yong PFK, Freeman AF, Engelhardt KR, Holland S, Puck JM, Grimbacher B. An update on the hyper-IgE syndromes. Arthritis Res Ther. 2012;14(6):228-228. 77. Su HC, Jing H, Zhang Q. DOCK8 deficiency. Ann N Y Acad Sci. 2011;1246:26- 33. 78. Biggs CM, Keles S, Chatila TA. DOCK8 deficiency: Insights into pathophysiology, clinical features and management. Clin Immunol. 2017;181:75-82. 79. Aydin SE, Freeman AF, Al-Herz W, et al. Hematopoietic Stem Cell Transplantation as Treatment for Patients with DOCK8 Deficiency. J Allergy Clin Immunol Pract. 2019;7(3):848-855. 80. Ruusala A, Aspenström P. Isolation and characterisation of DOCK8, a member of the DOCK180-related regulators of cell morphology. FEBS Letters. 2004;572(1- 3):159-166. 81. Côté J-F, Vuori K. Identification of an evolutionarily conserved superfamily of DOCK180-related proteins with guanine nucleotide exchange activity. J Cell Sci. 2002;115(24):4901-4913. 82. Côté J-F, Motoyama AB, Bush JA, Vuori K. A novel and evolutionarily conserved PtdIns(3,4,5)P3-binding domain is necessary for DOCK180 signalling. Nat Cell Biol. 2005;7(8):797-807. 83. Premkumar L, Bobkov AA, Patel M, et al. Structural basis of membrane targeting by the Dock180 family of Rho family guanine exchange factors (Rho-GEFs). J Biol Chem. 2010;285(17):13211-13222. 84. Harada Y, Tanaka Y, Terasawa M, et al. DOCK8 is a Cdc42 activator critical for interstitial dendritic cell migration during immune responses. Blood. 2012;119(19):4451- 4461. 85. Bishop AL, Hall A. Rho GTPases and their effector proteins. Biochem J. 2000;348 Pt 2(Pt 2):241-255. 86. Randall KL, Lambe T, Johnson AL, et al. Dock8 mutations cripple B cell immunological synapses, germinal centers and long-lived antibody production. Nat Immunol. 2009;10(12):1283-1291. 87. Janssen E, Kumari S, Tohme M, et al. DOCK8 enforces immunological tolerance by promoting IL-2 signaling and immune synapse formation in Tregs. JCI insight. 2017;2(19):e94298. 88. Zhang Q, Dove CG, Hor JL, et al. DOCK8 regulates lymphocyte shape integrity for skin antiviral immunity. J Exp Med. 2014;211(13):2549-2566. 89. Janssen E, Tohme M, Hedayat M, et al. A DOCK8-WIP-WASp complex links T cell receptors to the actin cytoskeleton. J Clin Invest. 2016;126(10):3837-3851. 90. Xu X, Han L, Zhao G, et al. LRCH1 interferes with DOCK8-Cdc42–induced T cell migration and ameliorates experimental autoimmune encephalomyelitis. J Exp Med. 2017;214(1):209-226. 91. Shiraishi A, Uruno T, Sanematsu F, et al. DOCK8 Protein Regulates Macrophage Migration through Cdc42 Protein Activation and LRAP35a Protein Interaction. J Biol Chem. 2017;292(6):2191-2202. 162

92. Deenick EK, Ma CS, Brink R, Tangye SG. Regulation of T follicular helper cell formation and function by antigen presenting cells. Curr Opin Immunol. 2011;23(1):111- 118. 93. Jiang S, Dong C. A complex issue on CD4+ T-cell subsets. Immunol Rev 2013;252(1):5-11. 94. Ma CS, Wong N, Rao G, et al. Monogenic mutations differentially affect the quantity and quality of T follicular helper cells in patients with human primary immunodeficiencies. J Allergy Clin Immunol. 2015;136(4):993-1006. 95. Ma CS, Wong N, Rao G, et al. Unique and shared signaling pathways cooperate to regulate the differentiation of human CD4+ T cells into distinct effector subsets. J Exp Med. 2016;213(8):1589-1608. 96. Choi Y, Lafferty JA, Clements JR, et al. Selective expansion of T cells expressing V beta 2 in toxic shock syndrome. J Exp Med. 1990;172(3):981-984. 97. Hodara VL, Jeddi-Tehrani M, Grunewald J, et al. HIV infection leads to differential expression of T-cell receptor V beta genes in CD4+ and CD8+ T cells. AIDS. 1993;7(5):633-638. 98. Clarêncio J, de Oliveira CI, Bomfim G, et al. Characterization of the T-Cell Receptor Vβ Repertoire in the Human Immune Response against Leishmania Parasites. Infect Immun. 2006;74(8):4757-4765. 99. Constant S, Pfeiffer C, Woodard A, Pasqualini T, Bottomly K. Extent of T cell receptor ligation can determine the functional differentiation of naive CD4+ T cells. J Exp Med. 1995;182(5):1591-1596. 100. Hosken NA, Shibuya K, Heath AW, Murphy KM, O'Garra A. The effect of antigen dose on CD4+ T helper cell phenotype development in a T cell receptor-alpha beta-transgenic model. J Exp Med. 1995;182(5):1579-1584. 101. van Panhuys N, Klauschen F, Germain RN. T-cell-receptor-dependent signal intensity dominantly controls CD4(+) T cell polarization In Vivo. Immunity. 2014;41(1):63-74. 102. Lewis RS. Calcium Signaling Mechanisms in T Lymphocytes. Ann Rev Immunol. 2001;19(1):497-521. 103. Oh-hora M, Rao A. Calcium signaling in lymphocytes. Curr Opin Immunol. 2008;20(3):250-258. 104. Murphy KM, Reiner SL. The lineage decisions of helper T cells. Nat Rev Immunol. 2002;2(12):933-944. 105. Weaver CT, Harrington LE, Mangan PR, Gavrieli M, Murphy KM. Th17: An Effector CD4 T Cell Lineage with Regulatory T Cell Ties. Immunity. 2006;24(6):677- 688. 106. Al Khatib S, Keles S, Garcia-Lloret M, et al. Defects along the T(H)17 differentiation pathway underlie genetically distinct forms of the hyper IgE syndrome. J Allergy Clin Immunol. 2009;124(2):342-348. 107. Iezzi G, Scotet E, Scheidegger D, Lanzavecchia A. The interplay between the duration of TCR and cytokine signaling determines T cell polarization. Eur J Immunol. 1999;29(12):4092-4101. 108. Tao X, Constant S, Jorritsma P, Bottomly K. Strength of TCR signal determines the costimulatory requirements for Th1 and Th2 CD4+ T cell differentiation. J Immunol. 1997;159(12):5956-5963. 109. Pfeiffer C, Stein J, Southwood S, Ketelaar H, Sette A, Bottomly K. Altered peptide ligands can control CD4 T lymphocyte differentiation in vivo. J Exp Med. 1995;181(4):1569-1574.

163

110. Huse M. The T-cell-receptor signaling network. J Cell Sci. 2009;122(9):1269- 1273. 111. Arpaia E, Shahar M, Dadi H, Cohen A, Rolfman CM. Defective T cell receptor signaling and CD8+ thymic selection in humans lacking Zap-70 kinase. Cell. 1994;76(5):947-958. 112. Hauck F, Randriamampita C, Martin E, et al. Primary T-cell immunodeficiency with immunodysregulation caused by autosomal recessive LCK deficiency. J Allergy Clin Immunol. 2012;130(5):1144-1152. 113. Keller B, Zaidman I, Yousefi OS, et al. Early onset combined immunodeficiency and autoimmunity in patients with loss-of-function mutation in LAT. J Exp Med. 2016;213(7):1185-1199. 114. Miyara M, Sakaguchi S. Natural regulatory T cells: mechanisms of suppression. Trends Mol Med. 2007;13(3):108-116. 115. Baecher-Allan C, Brown JA, Freeman GJ, Hafler DA. CD4+CD25high Regulatory Cells in Human Peripheral Blood. J Immunol. 2001;167(3):1245-1253. 116. Joller N, Lozano E, Burkett Patrick R, et al. Treg Cells Expressing the Coinhibitory Molecule TIGIT Selectively Inhibit Proinflammatory Th1 and Th17 Cell Responses. Immunity. 2014;40(4):569-581. 117. Unutmaz D. RORC2: The master of human Th17 cell programming. Eur J Immunol. 2009;39(6):1452-1455. 118. Okada S, Markle JG, Deenick EK, et al. Impairment of immunity to Candida and Mycobacterium in humans with bi-allelic RORC mutations. Science. 2015;349(6248):606-613. 119. Hirahara K, Ghoreschi K, Yang XP, et al. Interleukin-27 priming of T cells controls IL-17 production in trans via induction of the ligand PD-L1. Immunity. 2012;36(6):1017-1030. 120. Zhang Y, Ma CA, Lawrence MG, et al. PD-L1 up-regulation restrains Th17 cell differentiation in STAT3 loss- andSTAT1 gain-of-function patients. J Exp Med. 2017;214(9):2523-2533. 121. Milner JD, Brenchley JM, Laurence A, et al. Impaired T(H)17 cell differentiation in subjects with autosomal dominant hyper-IgE syndrome. Nature. 2008;452(7188):773- 776. 122. Fischer A, Notarangelo LD, Neven B, Cavazzana M, Puck JM. Severe combined immunodeficiencies and related disorders. Nat Rev Dis Primers. 2015;1:15061. 123. Meyts I, Bosch B, Bolze A, et al. Exome and genome sequencing for inborn errors of immunity. J Allergy Clin Immunol. 2016;138(4):957-969. 124. Fischer A, Provot J, Jais JP, Alcais A, Mahlaoui N. Autoimmune and inflammatory manifestations occur frequently in patients with primary immunodeficiencies. J Allergy Clin Immunol. 2017;140(5):1388-1393. 125. Lyons JJ, Milner JD. Primary atopic disorders. J Exp Med 2018;215(4):1009- 1022. 126. Mayor PC, Eng KH, Singel KL, et al. Cancer in primary immunodeficiency diseases: Cancer incidence in the United States Immune Deficiency Network Registry. J Allergy Clin Immunol. 2018;141(3):1028-1035. 127. Puck JM. Newborn screening for severe combined immunodeficiency and T-cell lymphopenia. Immunol Rev. 2019;287(1):241-252. 128. Seleman M, Hoyos-Bachiloglu R, Geha RS, Chou J. Uses of Next-Generation Sequencing Technologies for the Diagnosis of Primary Immunodeficiencies. Front Immunol. 2017;8:847.

164

129. Hong R, Cooper MD, Allan MJ, Kay HE, Meuwissen H, Good RA. Immunological restitution in lymphopenic immunological deficiency syndrome. Lancet. 1968;1(7541):503-506. 130. Brown L, Xu-Bayford J, Allwood Z, et al. Neonatal diagnosis of severe combined immunodeficiency leads to significantly improved survival outcome: the case for newborn screening. Blood. 2011;117(11):3243-3246. 131. Haddad E, Logan BR, Griffith LM, et al. SCID genotype and 6-month posttransplant CD4 count predict survival and immune recovery. Blood. 2018;132(17):1737-1749. 132. Heimall J, Logan BR, Cowan MJ, et al. Immune reconstitution and survival of 100 SCID patients post-hematopoietic cell transplant: a PIDTC natural history study. Blood. 2017;130(25):2718-2727. 133. Myers LA, Patel DD, Puck JM, Buckley RH. Hematopoietic stem cell transplantation for severe combined immunodeficiency in the neonatal period leads to superior thymic output and improved survival. Blood. 2002;99(3):872-878. 134. Pai S-Y, Logan BR, Griffith LM, et al. Transplantation Outcomes for Severe Combined Immunodeficiency, 2000–2009. New Eng J Med. 2014;371(5):434-446. 135. Dvorak CC, Hassan A, Slatter MA, et al. Comparison of outcomes of hematopoietic stem cell transplantation without chemotherapy conditioning by using matched sibling and unrelated donors for treatment of severe combined immunodeficiency. J Allergy Clin Immunol. 2014;134(4):935-943. 136. Su HC, Jing H, Angelus P, Freeman AF. Insights into immunity from clinical and basic science studies of DOCK8 immunodeficiency syndrome. Immunol Rev. 2019;287(1):9-19. 137. Tangye SG, Bucciol G, Casas-Martin J, et al. Human inborn errors of the actin cytoskeleton affecting immunity: way beyond WAS and WIP. Immunol Cell Biol. 2019;97(4):389-402. 138. Tangye SG, Pillay B, Randall KL, et al. Dedicator of cytokinesis 8-deficient CD4(+) T cells are biased to a TH2 effector fate at the expense of TH1 and TH17 cells. J Allergy Clin Immunol. 2017;139(3):933-949. 139. Boztug H, Karitnig-Weiss C, Ausserer B, et al. Clinical and immunological correction of DOCK8 deficiency by allogeneic hematopoietic stem cell transplantation following a reduced toxicity conditioning regimen. Pediatr Hematol Oncol. 2012;29(7):585-594. 140. Metin A, Tavil B, Azik F, et al. Successful bone marrow transplantation for DOCK8 deficient hyper IgE syndrome. Pediatr Transplant. 2012;16(4):398-399. 141. Cuellar-Rodriguez J, Freeman AF, Grossman J, et al. Matched related and unrelated donor hematopoietic stem cell transplantation for DOCK8 deficiency. Biol Blood Marrow Transplant. 2015;21(6):1037-1045. 142. Kuskonmaz B, Ayvaz D, Tezcan I, Yuce A, Sanal O, Cetinkaya DU. Successful hematopoietic stem cell transplantation after myeloablative conditioning in three patients with dedicator of cytokinesis 8 deficiency (DOCK8) related Hyper IgE syndrome. Bone Marrow Transplant. 2018;53(3):339-343. 143. Shah NN, Freeman AF, Su H, et al. Haploidentical Related Donor Hematopoietic Stem Cell Transplantation for Dedicator-of-Cytokinesis 8 Deficiency Using Post- Transplantation Cyclophosphamide. Biol Blood Marrow Transplant. 2017;23(6):980- 990. 144. Ghosh S, Schuster FR, Adams O, et al. Haploidentical stem cell transplantation in DOCK8 deficiency - Successful control of pre-existing severe viremia with a

165

TCRass/CD19-depleted graft and antiviral treatment. Clin Immunol. 2014;152(1-2):111- 114. 145. Al-Herz W, Chu JI, van der Spek J, et al. Hematopoietic stem cell transplantation outcomes for 11 patients with dedicator of cytokinesis 8 deficiency. J Allergy Clin Immunol. 2016;138(3):852-859 146. Freeman AF, Shah NN, Parta M, et al. Haploidentical related donor hematopoietic stem cell transplantation with post-transplantation cyclophosphamide for DOCK8 deficiency. J Allergy Clin Immunol Pract. 2016;4(6):1239-1242. 147. Uygun DFK, Uygun V, Reisli I, et al. Hematopoietic stem cell transplantation from unrelated donors in children with DOCK8 deficiency. Pediatr Transplant. 2017;21(7):e13015. 148. Gatz SA, Benninghoff U, Schutz C, et al. Curative treatment of autosomal- recessive hyper-IgE syndrome by hematopoietic cell transplantation. Bone Marrow Transplant. 2011;46(4):552-556. 149. Bittner TC, Pannicke U, Renner ED, et al. Successful long-term correction of autosomal recessive hyper-IgE syndrome due to DOCK8 deficiency by hematopoietic stem cell transplantation. Klin Padiatr. 2010;222(6):351-355. 150. Al Shekaili L, Sheikh F, Al Gazlan S, et al. Novel mutation in DOCK8-HIES with severe phenotype and successful transplantation. Clin Immunol. 2017;178:39-44. 151. Azik F, Azkur D, Avci Z, et al. Resolution of food-induced anaphylaxis in DOCK8-deficient patients following bone marrow transplantation. Turk J Pediatr. 2015;57(1):112-115. 152. Happel CS, Stone KD, Freeman AF, et al. Food allergies can persist after myeloablative hematopoietic stem cell transplantation in dedicator of cytokinesis 8- deficient patients. J Allergy Clin Immunol. 2016;137(6):1895-1898. 153. Pai SY, de Boer H, Massaad MJ, et al. Flow cytometry diagnosis of dedicator of cytokinesis 8 (DOCK8) deficiency. J Allergy Clin Immunol. 2014;134(1):221-223. 154. Khairallah C, Déchanet-Merville J, Capone M. γδ T Cell-Mediated Immunity to Cytomegalovirus Infection. Front Immunol. 2017;8:105-105. 155. Cuss AK, Avery DT, Cannons JL, et al. Expansion of functionally immature transitional B cells is associated with human-immunodeficient states characterized by impaired humoral immunity. J Immunol. 2006;176(3):1506-1516. 156. Suryani S, Fulcher DA, Santner-Nanan B, et al. Differential expression of CD21 identifies developmentally and functionally distinct subsets of human transitional B cells. Blood. 2010;115(3):519-529. 157. Wherry EJ, Kurachi M. Molecular and cellular insights into T cell exhaustion. Nat Rev Immunol. 2015;15(8):486-499. 158. Edwards ESJ, Bier J, Cole TS, et al. Activating PIK3CD mutations impair human cytotoxic lymphocyte differentiation and function and EBV immunity. J Allergy Clin Immunol. 2018: doi: 10.1016/j.jaci.2018.1004.1030. 159. Betts MR, Brenchley JM, Price DA, et al. Sensitive and viable identification of antigen-specific CD8+ T cells by a flow cytometric assay for degranulation. J Immunol Methods. 2003;281(1-2):65-78. 160. Gett AV, Hodgkin PD. Cell division regulates the T cell cytokine repertoire, revealing a mechanism underlying immune class regulation. Proc Natl Acad Sci U S A. 1998;95(16):9488-9493. 161. Romagnani S. Immunologic influences on allergy and the TH1/TH2 balance. J Allergy Clin Immunol. 2004;113(3):395-400. 162. Moens L, Tangye SG. Cytokine-Mediated Regulation of Plasma Cell Generation: IL-21 Takes Center Stage. Front Immunol. 2014;5:65. 166

163. Sun X, Wang J, Qin T, et al. Dock8 regulates BCR signaling and activation of memory B cells via WASP and CD19. Blood Adv. 2018;2(4):401-413. 164. Massaad MJ, Cangemi B, Al-Herz W, et al. DOCK8 and STAT3 dependent inhibition of IgE isotype switching by TLR9 ligation in human B cells. Clin Immunol. 2017;183:263-265. 165. Moran I, Avery DT, Payne K, et al. B cell-intrinsic requirement for STK4 in humoral immunity in mice and human subjects. J Allergy Clin Immunol. 2019. 166. Gray PE, O'Brien TA, Wagle M, et al. Cerebral Vasculitis in X-linked Lymphoproliferative Disease Cured by Matched Unrelated Cord Blood Transplant. J Clin Immunol. 2015;35(7):604-609. 167. Gray PE, Shadur B, Russell S, et al. Late-Onset Non-HLH Presentations of Growth Arrest, Inflammatory Arachnoiditis, and Severe Infectious Mononucleosis, in Siblings with Hypomorphic Defects in UNC13D. Front Immunol. 2017;8:944. 168. Buckley RH, Dees SC, O'Fallon WM. Serum immunoglobulins. I. Levels in normal children and in uncomplicated childhood allergy. Pediatrics. 1968;41(3):600-611. 169. Wilson RP, Ives ML, Rao G, et al. STAT3 is a critical cell-intrinsic regulator of human unconventional T cell numbers and function. J Exp Med. 2015;212(6):855-864. 170. Ozaki K, Spolski R, Feng CG, et al. A critical role for IL-21 in regulating immunoglobulin production. Science. 2002;298(5598):1630-1634. 171. Kotlarz D, Zietara N, Milner JD, Klein C. Human IL-21 and IL-21R deficiencies: two novel entities of primary immunodeficiency. Curr Opin Pediatr. 2014;26(6):704- 712. 172. Wu J, Zhang S, Qin T, et al. IL-21 alleviates allergic asthma in DOCK8-knockout mice. Biochem Biophys Res Commun. 2018;501(1):92-99. 173. Lankester AC. Management of Epstein-Barr virus reactivation following allogeneic stem cell transplantation. Rep Pract Oncol Radiother. 2007;12(3):163-165. 174. Scherrenburg J, Pietersma F, Jacobi R, Schuurman R, Meijer E, van Baarle D. CMV-reactivation influences T-cell differentiation and CMV-specific T-cell reconstitution after stem cell transplantation. J Clin Cell Immunol 2012;S9:001. 175. Gallez-Hawkins GM, Thao L, Palmer J, et al. Increased Programmed Death-1 Molecule Expression in Cytomegalovirus Disease and Acute Graft-versus-Host Disease after Allogeneic Hematopoietic Cell Transplantation. Biol Blood Marrow Tr. 2009;15(7):872-880. 176. Lugthart G, van Ostaijen-ten Dam MM, Jol - van der Zijde CM, et al. Early Cytomegalovirus Reactivation Leaves a Specific and Dynamic Imprint on the Reconstituting T Cell Compartment Long-Term after Hematopoietic Stem Cell Transplantation. Biol Blood Marrow Tr. 2014;20(5):655-661. 177. Barlogis V, Galambrun C, Chambost H, et al. Successful allogeneic hematopoietic stem cell transplantation for DOCK8 deficiency. J Allergy Clin Immunol. 2011;128(2):420-422. 178. Al-Herz W, Ragupathy R, Massaad MJ, et al. Clinical, immunologic and genetic profiles of DOCK8-deficient patients in Kuwait. Clin Immunol. 2012;143(3):266-272. 179. Radbruch A, Muehlinghaus G, Luger EO, et al. Competence and competition: the challenge of becoming a long-lived plasma cell. Nat Rev Immunol. 2006;6(10):741-750. 180. Khan F, Agarwal A, Agrawal S. Significance of chimerism in hematopoietic stem cell transplantation: new variations on an old theme. Bone Marrow Transplant. 2004;34(1):1-12. 181. Haraguchi K, Takahashi T, Hiruma K, et al. Recovery of Valpha24+ NKT cells after hematopoietic stem cell transplantation. Bone Marrow Transplant. 2004;34(7):595- 602. 167

182. Solders M, Erkers T, Gorchs L, et al. Mucosal-Associated Invariant T Cells Display a Poor Reconstitution and Altered Phenotype after Allogeneic Hematopoietic Stem Cell Transplantation. Front Immunol. 2017;8:1861. 183. Vely F, Barlogis V, Vallentin B, et al. Evidence of innate lymphoid cell redundancy in humans. Nat Immunol. 2016;17(11):1291-1299. 184. Recher M, Berglund LJ, Avery DT, et al. IL-21 is the primary common gamma chain-binding cytokine required for human B-cell differentiation in vivo. Blood. 2011;118(26):6824-6835. 185. Miggelbrink AM, Logan BR, Buckley RH, et al. B-cell differentiation and IL-21 response in IL2RG/JAK3 SCID patients after hematopoietic stem cell transplantation. Blood. 2018;131(26):2967-2977. 186. Hirschhorn R, Yang DR, Israni A, Huie ML, Ownby DR. Somatic mosaicism for a newly identified splice-site mutation in a patient with adenosine deaminase-deficient immunodeficiency and spontaneous clinical recovery. Am J Hum Genet. 1994;55(1):59- 68. 187. Hirschhorn R, Yang DR, Puck JM, Huie ML, Jiang CK, Kurlandsky LE. Spontaneous in vivo reversion to normal of an inherited mutation in a patient with adenosine deaminase deficiency. Nat Genet. 1996;13(3):290-295. 188. Stephan V, Wahn V, Le Deist F, et al. Atypical X-linked severe combined immunodeficiency due to possible spontaneous reversion of the genetic defect in T cells. N Engl J Med. 1996;335(21):1563-1567. 189. Hagl B, Spielberger BD, Thoene S, et al. Somatic alterations compromised molecular diagnosis of DOCK8 hyper-IgE syndrome caused by a novel intronic splice site mutation. Sci Rep. 2018;8(1):16719. 190. Kienzler AK, van Schouwenburg PA, Taylor J, et al. Hypomorphic function and somatic reversion of DOCK8 cause combined immunodeficiency without hyper-IgE. Clin Immunol. 2016;163:17-21. 191. Macallan DC, Asquith B, Irvine AJ, et al. Measurement and modeling of human T cell kinetics. Eur J Immunol. 2003;33(8):2316-2326. 192. Macallan DC, Wallace DL, Zhang Y, et al. B-cell kinetics in humans: rapid turnover of peripheral blood memory cells. Blood. 2005;105(9):3633-3640. 193. Westera L, van Hoeven V, Drylewicz J, et al. Lymphocyte maintenance during healthy aging requires no substantial alterations in cellular turnover. Aging Cell. 2015;14(2):219-227. 194. Fischer A, Hacein-Bey Abina S, Touzot F, Cavazzana M. Gene therapy for primary immunodeficiencies. Clin Genet. 2015;88(6):507-515. 195. Thrasher AJ, Williams DA. Evolving Gene Therapy in Primary Immunodeficiency. Mol Ther. 2017;25(5):1132-1141. 196. Takenouchi T, Kosaki R, Niizuma T, Hata K, Kosaki K. Macrothrombocytopenia and developmental delay with a de novo CDC42 mutation: Yet another locus for thrombocytopenia and developmental delay. Am J Med Genet A. 2015;167(11):2822- 2825. 197. Takenouchi T, Okamoto N, Ida S, Uehara T, Kosaki K. Further evidence of a mutation in CDC42 as a cause of a recognizable syndromic form of thrombocytopenia. Am J Med Genet A. 2016;170(4):852-855. 198. Martinelli S, Krumbach OHF, Pantaleoni F, et al. Functional Dysregulation of CDC42 Causes Diverse Developmental Phenotypes. Am J Hum Genet. 2018;102(2):309- 320.

168

199. Motokawa M, Watanabe S, Nakatomi A, et al. A hot-spot mutation in CDC42 (p.Tyr64Cys) and novel phenotypes in the third patient with Takenouchi-Kosaki syndrome. J Hum Genet. 2018;63(3):387-390. 200. Villa A, Notarangelo L, Macchi P, et al. X-linked thrombocytopenia and Wiskott- Aldrich syndrome are allelic diseases with mutations in the WASP gene. Nat Genet. 1995;9(4):414-417. 201. Zhu Q, Zhang M, Blaese RM, et al. The Wiskott-Aldrich syndrome and X-linked congenital thrombocytopenia are caused by mutations of the same gene. Blood. 1995;86(10):3797-3804. 202. Sullivan KE, Mullen CA, Blaese RM, Winkelstein JA. A multiinstitutional survey of the Wiskott-Aldrich syndrome. J Pediatr. 1994;125(6, Part 1):876-885. 203. Ochs HD, Thrasher AJ. The Wiskott-Aldrich syndrome. J Allergy Clin Immunol. 2006;117(4):725-738; quiz 739. 204. Orange JS, Ramesh N, Remold-O'Donnell E, et al. Wiskott-Aldrich syndrome protein is required for NK cell cytotoxicity and colocalizes with actin to NK cell- activating immunologic synapses. Proc Natl Acad Sci U S A. 2002;99(17):11351-11356. 205. Park JY, Shcherbina A, Rosen FS, Prodeus AP, Remold-O'Donnell E. Phenotypic perturbation of B cells in the Wiskott-Aldrich syndrome. Clin Exp Immunol. 2005;139(2):297-305. 206. Adriani M, Aoki J, Horai R, et al. Impaired in vitro regulatory T cell function associated with Wiskott-Aldrich syndrome. Clin Immunol. 2007;124(1):41-48. 207. Albert MH, Bittner TC, Nonoyama S, et al. X-linked thrombocytopenia (XLT) due to WAS mutations: clinical characteristics, long-term outcome, and treatment options. Blood. 2010;115(16):3231-3238. 208. Dobbs K, Dominguez Conde C, Zhang SY, et al. Inherited DOCK2 Deficiency in Patients with Early-Onset Invasive Infections. N Engl J Med. 2015;372(25):2409-2422. 209. Alizadeh Z, Mazinani M, Shakerian L, Nabavi M, Fazlollahi MR. DOCK2 Deficiency in a Patient with Hyper IgM Phenotype. J Clin Immunol. 2018;38(1):10-12. 210. Alosaimi MF, Shendi H, Beano A, et al. T-cell mitochondrial dysfunction and lymphopenia in DOCK2-deficient patients. J Allergy Clin Immunol. 2019;144(1):306- 309. 211. Moens L, Gouwy M, Bosch B, et al. Human DOCK2 Deficiency: Report of a Novel Mutation and Evidence for Neutrophil Dysfunction. J Clin Immunol. 2019;39(3):298-308. 212. Mengesha BG, Conti HR. The Role of IL-17 in Protection against Mucosal Candida Infections. J Fungi. 2017;3(4):52. 213. Gismondi A, Cifaldi L, Mazza C, et al. Impaired natural and CD16-mediated NK cell cytotoxicity in patients with WAS and XLT: ability of IL-2 to correct NK cell functional defect. Blood. 2004;104(2):436-443. 214. Imai K, Morio T, Zhu Y, et al. Clinical course of patients with WASP gene mutations. Blood. 2004;103(2):456-464. 215. Peacocke M, Siminovitch KA. Wiskott-Aldrich syndrome: New molecular and biochemical insights. J Am Acad of Dermatol. 1992;27(4):507-519. 216. Parolini O, Berardelli S, Riedl E, et al. Expression of Wiskott-Aldrich Syndrome Protein (WASP) Gene During Hematopoietic Differentiation. Blood. 1997;90(1):70-75. 217. Ochs HD, Filipovich AH, Veys P, Cowan MJ, Kapoor N. Wiskott-Aldrich Syndrome: Diagnosis, Clinical and Laboratory Manifestations, and Treatment. Biol Blood Marrow Tr. 2009;15(1):84-90. 218. Schwarz K, Nonoyama S, Peitsch MC, et al. WASPbase: a database of WAS- and XLT-causing mutations. Immunol Today. 1996;17(11):496-502. 169

219. Jin Y, Mazza C, Christie JR, et al. Mutations of the Wiskott-Aldrich Syndrome Protein (WASP): hotspots, effect on transcription, and translation and phenotype/genotype correlation. Blood. 2004;104(13):4010-4019. 220. Fukui Y, Hashimoto O, Sanui T, et al. Haematopoietic cell-specific CDM family protein DOCK2 is essential for lymphocyte migration. Nature. 2001;412(6849):826-831. 221. Meller N, Irani-Tehrani M, Kiosses WB, Del Pozo MA, Schwartz MA. Zizimin1, a novel Cdc42 activator, reveals a new GEF domain for Rho proteins. Nat Cell Biol. 2002;4(9):639-647. 222. Thrasher AJ, Burns SO. WASP: a key immunological multitasker. Nat Rev Immunol. 2010;10:182. 223. Modell V, Orange JS, Quinn J, Modell F. Global report on primary immunodeficiencies: 2018 update from the Jeffrey Modell Centers Network on disease classification, regional trends, treatment modalities, and physician reported outcomes. Immunol Res. 2018;66(3):367-380. 224. Avery DT, Deenick EK, Ma CS, et al. B cell-intrinsic signaling through IL-21 receptor and STAT3 is required for establishing long-lived antibody responses in humans. J Exp Med. 2010;207(1):155-171.

170