Ig light chain expression in B cells from patients with common variable immunodeficiency

(1)

Master thesis, 30 hp

Master’s Programme in Pharmaceutical Science, 120 hp Report approved: Autumn term 2018

Supervisors: Ola Grimsholm, Anders Öhman, Examinor: Olov Nilsson

Ig light chain expression in B cells from patients with

common variable immunodeficiency

Is there a correlation to a particular clinical diagnosis?

Manjinder Singh

(2)

(3)

Abstract

B lymphocytes (B cells) are crucial for the defence against re-infecting pathogens and foreign substances as they can produce proteins called antibodies that are necessary for their neutralization. An antibody is composed of immunoglobulin (Ig) heavy chains and light chains, where the later can be either Ig kappa (к) or lambda (λ). Patients with common variable immunodeficicncy (CVID) have predominantly B-cell disorders with severe reduction in one or more Ig isotype (IgA, IgG and sometimes IgM). Even though patients with CVID have an imparied B-cell function that prevent them to elicit an effetive immune response against exogenous patoghens, 20-25% of CVID patients develop autoimmunity, and sometimes a subset of these patients develop B or T cell lymphomas over their lifetime (11-13%). The objective of this study was to perform immunophenotyping of B cells in peripheral blood from patients with CVID and healthy donors (HDs), assessing the Igк expression by flow cytometry. This was performed in order to determine whether any of these parameters could be used to differentiate between patients who develop autoimmune disease and/or lymphoma. In addition, DNA from peripheral blood mononuclear cells from HDs and CVID patients were extracted and DNA quantification and purity assessment was performed. These results will be used as the basis for the selection of patients for later whole genome sequencing that could provide genomic disease markers. Although this study included too few CVID patients (n=19) and HDs (n=5) to draw any generalizable conclusion some distinct identifiable patterns for the Igк expression could be determined. CVID patients with lymphoproliferation (LP) showed a reduction in percentage of Igк expressing B cells corresponding to a particular trend. The patterns within this trend were not observed in any other clinical patient group, which might suggest distinct patterns for this particular patient group. CVID patients with autoimmunity and lymphoma (AI+Lymphoma) displayed two patterns both of which were non-HD. The same pattern, which was observed for one of the AI+Lymphoma patients was also found for one of the patients in the autoimmunity and lymphoproliferation (AI+LP) group, which might suggest a connection between these two clinical diagnoses. Furthermore, an unidentified (UID) B-cell subpopulation (CD24^low CD27^low CD38^low) was observed, but only in patients with AI+LP and AI+lymphoma, further supporting a connection between these two diagnoses. Based on the results, CVID patients 10 and 13 from the infections only group, CVID patient 3 and 15 from the AI group and patients 1 and 6 from the AI+LP group was suggested for the whole genome sequencing along with 4 HDs.

(4)

Abbreviations

Abs – Antibodies AI – Autoimmunity

CVID – common variable immunodeficiency HD – Healthy donor

Ig – Immunoglobulin

IGRT – Immunoglobulin replacement therapy к – Kappa

λ – Lambda

LP – Lymphoproliferation MBC – Memeory B cells N – Naïve B cells

PLB – Plasmablasts (plasma cells) RR – Relative risk

Tr – Transitional B cells UID – Unidentified B cells

(5)

1. Introduction

This study was conducted to gain further knowledge about the immune disorder common variable immunodeficiency (CVID). The introduction will thus begin by focusing on the immune system and thereafter describe CVID.

1.1 The Immune system

The immune system can be divided into an early response (innate immunity) and a later response (adaptive immunity) (Fig. 1) [1-3]. The innate immunity responds in essentially the same way to repeated exposures to microbes and antigens while the adaptive immunity can increase the magnitude of its response upon repeated infections and thus, provide a long-lasting specific immunity. These two immunological systems work together in order to protect us against bacterial, viral and other types of infections and foreign substances [4]. The adaptive immune responses are mediated by cells called lymphocytes [2, 3].

Lymphocytes can be further subdivided into B and T lymphocytes (B and T cells) that have different functions. These cells rearrange their deoxyribonucleic acid (DNA) of their antigen receptors to generate a vast amount of specific receptors [1].

Fig. 1 [5]. A schematic representation of the innate and adaptive immunity. Innate immunity cells and complement system provide the early response to pathogen infection while the adaptive immunity gives the late response with the activation of B and T cells. The activation of B cells leads to the formation of memory B cells that circulate the blood looking for antigen and antibody-secreting plasma cells that home back to the bone marrow were they then secret antibodies continuously.

(8)

2

1.1.1 B cells

B cells produce and secrete proteins, called antibodies (Abs) upon activation and differentiation which provide the humoral immunity against extracellular pathogens and their toxins (Fig. 2) [2, 3]. B cells are generated in the bone marrow by stem cells that proliferate and differentiate into a B cell linage precursor called common lymphoid precursor cells (CLPs). CLPs are then stimulated for instance, by cytokines produced by stromal and macrophage cells in the bone marrow to stepwise mature into immature B cells. The immature B-cell then leaves the bone marrow to mature in the spleen into mature-naïve B-cell.

Fig. 2 [6]. Antibodies are composed of heavy (H) and light (L) chains which in turn are made up of constant (C) and variable (V) regions [2, 3]. The constant region of the H-chain is made up of CH1- CH4 domains. A membrane-bound antibody has also a transmembrane region (TM) and a cytosolic domain (CYT) which has a carboxylic acid terminal region (COOH). CH1 and CH2 on the H-chain are interconnected by a hinge which provides flexibility to the antibody when it binds antigens located close or far apart from each other on e.g.

microbe surface. The V-regions of the heavy chain (VRH) can further be subdivided into a variable (VH), diversity (DH) and a joining (JH) region as illustrated in the left part of the figure. Likewise, the V-region on the L-chain (VRL) can be divided into a variable (VL) and a joining (JL) region. Binding of antigen occurs between the VH and VL region loops. Immunoglobulin that bind antigen can however not itself generate an intracellular response signal, it has associated with it invariant signaling proteins Igα and Igβ. These signaling proteins each have an immunoreceptor tyrosine-based activation motif (ITAM) region that becomes phosphorylated and induce the signal upon binding of antigen to the B-cell receptor (BCR).

1.1.2 Antigen binding diversity of antibodies

The antigen binding site of an antibody is composed of three loops from each heavy and light chain variable regions [2, 3]. These loops are hypervariable and called complementary

(9)

3 determining regions (CDRs). There are three CDRs (CDR1-CDR3) with CDR3 being the most hypervariable. CDRs determine antigen specificity of the antibody, thus, the diversity of antibodies binding different antigens is generated by combining different variable heavy and light chains. Additional diversity is generated by N- and P-nucleotide additions in the joints between gene segments during immunoglobulin (Ig) gene rearrangement [2, 3, 7].

1.1.3 Immunoglobulin Kappa and lambda light chain expression

During early B-cell development in the bone marrow B cells rearrange their Ig H-chain gene and after successful rearrangement, the Ig H-chain will be expressed on a pre-B-cell together with the surrogate light chain as a pre-B-cell receptor [2, 3, 8, 9]. Subsequently, the B-cell initiates the rearrangement of its Ig L-chain genes. The Ig L-chain rearrangement starts with the kappa (к) locus, if the к allele can be successfully paired with the Ig H-chain, Igк protein is expressed as the L-chain. If the cell is unable to rearrange both Igк alleles it proceeds to its Ig lambda (λ) L-chain locus. Thus, most immature B cells coming out from the bone marrow have multiple Ig L-chain gene rearrangements. Furthermore, it is estimated that around 30 percent of the B cells with Igλ expressed as their L-chain have potentially functional Igк rearrangements [10]. The inactivation and deletion of potentially functional Igк rearrangements may be essential for a B-cell to function and is performed by the к deleting element (KDE) which is located downstream of Igк constant region [9].

The KDE rearrangement is mediated by the recombination-activating gene (RAG) protein recombination.

Three separate loci encode all of the Ig H-chains, Igк L-chain and Igλ L-chain [7]. These DNA segments are located on different chromosomes (Ig H-chains on chromosome 14; Igк on chromosome 2; Igλ on chromosome 22) and are joined together only in developing B cells and not in other cell types [2, 7, 8].

There are about 45 functional VH chain genes, 35 Vк light chain and about 30 Vλ light chain genes in each locus in humans [11]. There are some indications that the chromosomal locations of Vк genes have been rearranged many times during the evolutionary process by gene duplication, deletion and transposition so that many genes have become nonfunctional due to this process [11].

Each of these locus have a different arrangement and number of constant region genes, for example, the к light chain locus has a single Cк gene while the λ light chain locus has four functional Cλ genes. The Cк and Cλ genes are made of single exons that encode for the whole C domain of the light chain.

1.1.3.1 Different expression of kappa and lambda genes

The ratio between к:λ expression in B cells has been shown to be higher in systemic lupus erythematosus (SLE) patients compared to healthy individuals (median of 1.9 vs 1.4 (p

<0.05)) [7]. This might suggest a changed selection for к and λ in patients with SLE. SLE is a systemic autoimmune disease which can result from either chronic IgG overproduction or a defect in immune complex clearance, or both simultaneously [3]. The study also found that all Vк and Vλ genes are not expressed equally in either SLE patients or healthy individuals [7]. A total of 11 genes (5 Vк and 6 Vλ genes) were identified as expressed differently.

However, no study has been found that looks at the к:λ ratio in contrast to proliferation and differentiation data to assess the relevance of к expressing B cells and lymphoproliferation.

1.1.4 Immunological tolerance

Immunological tolerance is defined as unresponsiveness to self-antigen. Self-tolerance for autologous antigens is an important trait of the immune system [2, 3]. Failure to induce

(10)

4 self-tolerance may lead to autoimmunity. Lymphocytes in normal individuals that recognize self-antigens are killed, inactivated or the lymphocyte is given one more chance to change its specificity for the antigen (receptor editing in B cells). The tolerance for self- antigens is induced in both immature lymphocytes in generative lymphoid organs (central tolerance: thymus or bone marrow) and in mature lymphocytes in peripheral sites (peripheral tolerance: lymph nodes or by regulatory T cells). The presentation of antigen to a T-cell in the absence of co-stimulator or prolonged signaling may induce anergy (an unresponsive state with downregulation of the antigen receptor). Likewise, if B cells recognize self-antigens weakly or the antigen does not crosslink enough antigen receptors the B-cell can become unresponsive.

1.1.5 Immune dysregulation

The immune system is thus a very complex system that is comprised of many types of cells and protects against external pathogens and toxins when it is working properly. But, when a person acquires some sort of relevant or major defect(s) in this immunological defense system problems start to manifest, one such problem that could arise is immunodeficiency.

1.2 Common variable immunodeficiency

Common variable immunodeficiency (CVID) is the most common symptomatic primary immunodeficiency (PID) with a prevalence of 1:25,000 to 1:50,000 and affect women more often than men [1, 4, 12-16]. It is estimated that around 200-300 individuals are affected by this disease in Sweden [4]. In contrast to most PIDs CVID is most commonly diagnosed in adults or at earliest in late childhood [3, 14]. Diagnosis is generally delayed due to limited awareness of this disorder and CVID is therefore sometimes diagnosed 15+ years after symptom onset [17]. The term variable is used to denote that symptoms and the degree of immunodeficiency are variable among patients [4]. Most probably it is a conglomerate of different diseases and not a single disease. CVID is predominantly a B-cell disorder and the most common disorder among the 20 antibody deficiencies recognized [15]. It is a heterogeneous disorder characterized by severe reduction in two or more Ig isotypes (IgA, IgG and sometimes IgM) in conjunction with normal or low B-cell numbers [1, 15, 18].

1.2.1 Symptoms in CVID

Due to reduced B-cell functionality and reduced levels of antibodies, the most common symptom is recurrent upper and lower respiratory tract infections with encapsulated bacteria, especially Streptococcus pneumoniae and unencapsulated Haemophilus influenzae which leads to chronic lung disease, bronchiectasis and can subsequently lead to mortality if untreated [10, 12, 15]. Contrary to healthy individuals these patients have fewer sickness related symptoms like fever when infected with pathogens, but they may feel inexplicably tired as the infection progresses [4]. The major factor contributing to lung disease in CVID patients is suggestively a defective antibody response against capsular polysaccharides, which is a T-cell independent process [12]. In a cohort study with 54 CVID patients, it was shown that IgM memory B cells may also play a major role in protecting against encapsulated bacteria [12]. However, the study also concluded that the reduced frequency of IgM memory B cells may not be predictive for an increased risk to acquire pneumonia infections.

Other common symptoms for patients with CVID are sinusitis, otitis and gastric tract infections with long periods of diarrhea [4]. Patients also have an increased risk for sepsis [4, 10].

In a large European study in which 334 CVID patients were followed for 25.6 years (in total 9461 patient years) researchers found that subjects with infections as their only symptom survived longer than those with other severe complications such as autoimmunity, polyclonal lymphocytic infiltration and lymphoma [17].

(11)

5 However, there are asymptomatic CVID patients who do not have recurrent pneumonia, the reason for the difference among patients is still unknown [12].

1.2.2 CVID Diagnostics

Because the cause of CVID is unknown, various diagnostic criterions have been proposed [19]. The diagnostic criterion for CVID patients older than 4 years generally comprise of marked serum reduction of either IgG and IgA or IgG and IgM, or all three simultaneously [1]. In addition, the patient needs to show a defective specific antibody response to proteins and polysaccharide antigens, recurrent bacterial infections as well as no dramatic T-cell deficiency [16]. The European Society of Immunodeficiencies (ESID) and the Pan American Group of Immune deficiency (PAGID) diagnostic criterions are generally used [20]. These diagnostic criterions comprise of IgG levels <7-8 g/L (for most laboratories), weak vaccine response or lacking isohemagglutinins (antibodies against foreign blood groups) and exclusion of differential diagnoses. It is worth noting that symptoms or infection related secondary diseases are not part of the ESID/PAGID diagnostic criterion for CVID.

In Sweden five criterions should be fulfilled to diagnose a person with CVID [21].

1. The patient needs to display at least one of the following clinical signs:

autoimmunity, infection or lymphoproliferation.

2. IgG level lower than tabulated values in table 1. (IgG levels measured at least two times and with minimum 3 weeks interval).

3. IgA and IgM level lower than tabulated values in table 1.

4. Weak antibody response to vaccines or after undergone infection. (This criterion may be skipped in urgent cases).

5. Differential diagnoses must be ruled out.

Table 1. The recommended lower cut-off point for serum IgG, IgA and IgM in g/L for CVID diagnostics in Sweden at different ages.

Antibody 4-10 years 10-20 years >20 years

IgG 6.1 6.1 6.7

IgA 0.5 0.5 0.88

IgM 0.27 0.27 0.27

1.2.3 Autoimmunity in CVID

Despite the inability to elicit an effective antibody response to exogenous pathogens due to impaired B-cell function, many CVID patients respond to self-antigens, which often leads to autoimmunity [15]. Immune thrombocytopenic purpura (ITP) and autoimmune haemolytic anemia are the most common complications in a hematological context.

Systemic and organ-specific autoimmune diseases may also develop in patients with CVID.

These patients also have a propensity for lymphoid intestinal pneumonitis, granulomatous and a higher risk for lymphomas. Patients with autoimmunity have a 2.5 relative risk (RR) for mortality compared with CVID patients with only infections [17]. Furthermore, evidence suggests that the deficiency of switched IgM^- IgD^- CD27⁺ memory B cells may be correlated with autoimmunity development in CVID patients. Autoimmunity is generally treated with cortisone to suppress the immune system and thus the autoimmune disease [4].

Even though CVID is most likely not a heritable disease, different forms of autoimmunity have often been seen in relatives of these patients.

(12)

6

1.2.4 Lymphoma in patients with CVID

In the majority of CVID lymphoma cases no underlying infectious cause has been found for the disease [15]. However, some reasons have been found for an increased risk of lymphoma development in CVID, including, chronic infections, genetic variants, dysregulations of the immune system and radiation sensitivity [1]. Lymphoma appears to be more common in female CVID patients than male and appear on average 9 years after CVID diagnosis in females. Although in some patients Epstein-Barr virus (EBV), an opportunistic pathogen has been found to drive the lymphoproliferative disorder [1, 15].

The incidence of lymphoma is around 11-13% in patients with CVID and usually occurs between the age of 50-60 years and the risk of developing this malignancy is 12-18 fold higher in CVID than in the rest of the population [22]. The 12 fold increased risk for cancer development had previously been shown by a collaborative study from Denmark and Sweden using register data [23]. Relatives of these CVID patients showed no increased risk of developing any type of cancer, suggesting that the risk for morbidity in CVID may be coupled with the immunodeficiency rather than genetic traits shared within the family.

The non-infectious nature of lymphomas in patients with CVID suggests that the disease could be a manifestation of defective immune regulation contributing to poor B-cell function [15]. This was illustrated by a case report where CVID patients had antibody production recovery after acquiring human immunodeficiency virus (HIV) infection, in particular, IgG and IgM levels were higher after acquired infection [24]. The IgA levels did however not normalize suggesting separate factors predisposing a person genetically for selective IgA deficiency.

Furthermore, there is a growing number of studies that indicate that allergic reactions could also lead to lymphoma and hematologic malignancies [25].

High levels of polyclonal IgM antibodies have also been found to correlate with lymphoma development in patients with CVID [17]. Moreover, the RR for death due to lymphoma was 5.5 in comparison with subjects who only had infections as their main symptom.

1.2.4.1 Diagnosis of lymphoma in CVID

There are some difference in how lymphoma presents itself in patients with CVID compared to healthy individuals [1]. In CVID, lymphadenopathy (abnormal size, number, or consistency of lymph-nodes) is a common trait but it is not easily evaluated as lymphoma due to the fact that lymphoma in CVID is commonly extranodal and appear in not so common locations (lungs, stomach etc.). This makes it hard to detect lymphoma by routine follow-up measures in these patients. Due to the extranodal nature of lymphomas in CVID, it is generally difficult to distinguish between benign and malign lymphoma. Moreover, bone marrow biopsies seldom reveal anything.

The Diagnostic criterion for malignancy in CVID include confirmation of monoclonal expansion by immunophenotyping or by molecular methods.

1.2.5 Genetics in CVID

CVID may arise from a large number of different genetic defects and mutations involved in B-cell activation and differentiation [16]. Over the years many genetic variants have been identified in patients with CVID, for instance, TNF-like receptor transmembrane activator and CALM interactor (TACI), CD19, CD20, CD21, CD81, inducible T-cell costimulatory (ICOS), lipopolysaccharide responsive and beige-like anchor protein (LRBA) and phospholipase Cƴ2 (PLCƴ2) [1, 16]. TACI is a receptor for BAFF and APRIL cytokines which can provide co-stimulation, B-cell survival signal for activation and antibody class switching signal [3]. Moreover, DNA repair genes (MSH2, MSH5, MLH1, NBS1 and RAD50) polymorphisms have also been shown to occur in CVID [1]. The genetic instability, continuous activation of lymphocytes and the lymphoid system in event of an infection increases the risk of malignancy [25].

(13)

7 Recent study results have also drawn a connection between a dramatic increase in total copy number variation (CNV) burden and CVID, the cause for this is still unknown [26].

But, when another group of researchers explored the clinical significance and origin of these findings they found no correlation between higher total CNV and incidence of malignancy or any other subphenotypes [27]. This might suggest that the increased CNV burden is static and intrinsic to CVID as a disease per se.

A monogenic cause for CVID has been found only in a minority of cases, often identified as genes involved in B-cell activation with the help of gene sequencing techniques like whole genome or exome sequencing [28]. Most forms of CVIDs are probably polygenic meaning that the patient has genetic defects in many different genes [4]. The lack of monogenic causality in the majority of CVID cases may also point to a combined environmental, genetic and dysregulated immune system as the underlying cause of the disease [14].

1.2.6 Prophylactic treatment

Immunoglobulin replacement therapy (IGRT) can be an effective prophylaxis against bacterial infections in CVID [4, 15]. IGRT is administrated most commonly as an intravenous (I.V.) or subcutaneous (S.C.) injection. I.V. administration is often repeated every three weeks while S.C. injections need to be administrated more frequently and in lower doses [4]. In a cohort study with 90 CVID patients studied for over 22 years revealed that the effective dose of IgG replacement for bacterial infection prophylaxis varied from 0.2 to 1.2 g/kg/month [29]. Furthermore, patients with confirmed bronchiectasis required higher doses. The goal of the replacement therapy should according to the study results not be to reach a particular IgG level but to improve clinical outcome. A European retrospective study with 2212 patients using the data gathered from European Society for Immunodeficiencies Database concluded that there was difference in treatment strategies among European countries [13]. The IgG replacement doses in European countries ranged from 0.13 up to 0.75 g/kg/month, the researchers also concluded that patients with trough levels (lowest concentration of drug before next dose administration) less than 4g/L spent a median of 0.8 days in the hospital per year compared with a median of 0 days for patients with higher IgG trough levels. Although, Ig therapy has been shown to substantially reduce the frequency of bacterial infections and is likely to reduce mortality it has little or no effect on inflammation and the development of lymphoma [17]. The reverse has actually been shown for patients with CVID who had a higher frequency of malignancy with Ig therapy.

IGRT has also sometimes been shown to have a limited effect in preventing the development of autoimmunity in these patients [22]. Thus, IGRT can’t replace the function of B cells and with the widespread use of IGRT, there has been a shift to autoimmune and lymphoproliferative based morbidity and mortality instead of infection based [14].

1.2.6.1 When to initiate immunoglobulin replacement therapy

The decision, when to initiate IGRT is important in many aspects, not at least if it can improve the quality of life, reduce the risk of morbidity and mortality [30]. However, the risk of exposing the patient to unnecessary treatment and the cost aspects of treatment must also be taken into account. A watch and wait approach may be advisable for patients with mild infectious manifestations who can be given antibiotic prophylaxis instead.

The most problematic decision-making scenario occurs when patients have reduced but not absent IgG level (IgG between 4-6g/L) with or without low IgA level. Patients with IgG levels of <4g/L generally require IGRT, but there are still exceptions to this rule. In Sweden CVID patients with serum IgG levels <3 g/L and IgA levels <0.07 g/L are recommended IGRT [21]. Furthermore, patients with IgG >3 g/L with high infection frequency are also recommended IGRT. Fulfilling the CVID diagnostic criterion generally mean that IGRT should be initiated [31], although, treatment of patients who do not have recurrent respiratory infections or have lymphoma is still controversial [18]. Patients with mild clinical symptoms are recommended to be followed-up every 6-12 month and in case of

(14)

8 bacterial infection be treated with antibiotics promptly [31]. One of the benefits of early IGRT is that it could reduce the risk of permanent bronchiectatic lung damage [1].

1.2.6.2 Antibiotics

One other mean of treating bacterial infection is with antibiotics [4, 30]. The duration of an antibiotic treatment is generally longer than one week for patients with CVID, antibiotics can also be prescribed as a prophylactic intervention and the treatment can then go on for several months or years [4].

2. Objective

20-25 % of CVID patients develop autoimmunity and subsequently B or T cell lymphomas (11-13%) over their lifetime [22, 23, 32, 33]. However, the reason for this disease progression is unknown. The aim of this study is to perform immunophenotyping of B cells in peripheral blood from patients with CVID and HDs, including assessing the Igк expression. This will be performed in order to determine whether any of these parameters could be used to differentiate between patients who develop autoimmune disease and/or lymphoma. To do this, PBMCs will first be isolated from whole blood. Then, B cells within PBMCs will be stained with monoclonal antibodies attached to fluorochromes, targeting different cell surface markers. The stained B cells will thereafter be analyzed with flow cytometry. These results will then be the basis for the selection of CVID patients and HDs for later whole genome sequencing, for which, DNA will be extracted from PBMCs from selected CVID patients and HDs. The extracted DNA will also be quantified and quality assessed in this study.

(15)

9

3. Methods

3.1 Isolation of PBMCs from healthy donors and CVID patients

PBMCs were isolated by density gradient centrifugation technique which separates blood cells according to their relative density, size and shape. Blood samples were diluted with sterile phosphate-buffered saline (PBS, Bacteriological laboratory, Gothenburg) with one portion blood and two portions sterile PBS to achieve a 1:3 dilution in a sterile conical tube.

3.5 mL of Ficoll (GE Healthcare, Little Chalfont, UK) density gradient medium was added to sterile 15mL conical tubes, on top of which the diluted blood was very gently layered (Fig.

3A). This was then centrifuged for 20 minutes at 400 gravitational acceleration (g) force at 20ôC with brake and acceleration setting of 1 (6 rpm/sec) in a swing-out centrifuge (4-16K, Sigma). The PBMCs were then carefully collected from the Ficoll/plasma interphase with a pipette and transferred to a new sterile 15mL conical tube (Fig. 3B). The PBMCs were then washed with PBS and centrifuged for 10 min at 400g at 20ôC with brake and acceleration setting of 9 (1000 rpm/sec) to properly remove any Ficoll present in the sample. The cells were washed once more by discarding the supernatant and resuspending the PBMCs pellet in sterile PBS, this was then centrifuged for 5 minutes at 300g at 20ôC. Thereafter, the supernatant was discarded and PBMCs pellet were resuspended in 1mL sterile PBS containing 2% fetal calf serum (FCS, Sigma-aldrich) and 1mM ethylenediaminetetraacetic acid (EDTA, Sigma-aldrich) (FACS buffer).

Fig. 3 [34]. (A) The blood was diluted with PBS and very gently layered on the Ficoll. (B) The PBMCs were trapped in the interphase between plasma and Ficoll layer upon centrifugation while red blood cells (RBCs) sedimented to the bottom of the tube. PBMCs were then isolated by pipetting. During this isolation process, Ficoll transfer was avoided due to the cell toxicity of the Ficoll.

3.1.1 Determination of cell concentration

PBMC concentration for each sample was determined with an automated cell counter (KX- 21N, Sysmex) by pipetting 10µL of cell resuspension to a 1.5mL Eppendorf tube which was diluted with 90µL FACS buffer. This cell dilution step was varied to keep the cell

(16)

10 concentration >1 x 106 cells/mL but <20 x 106 cells/mL to ensure good measuring precision.

3.1.2 Freezing of PBMCs

For storage of PBMCs, cells were pelleted and resuspended in 1mL freezing medium (90%

FCS, 10% dimethyl sulfoxide (DMSO, Sigma-aldrich) divided into 2mL cryotubes (optimal cell concentration: 10 x 10⁶ cells/mL). Cryotubes were then placed in a freezing container containing isopropyl alcohol to achieve cooling rate close to -1°C/minute for optimal cell preservation and transferred directly to a freezer set at -80ôC. Cells were stored at -80ôC for short-term storage (up to a month) and at -150ôC for long-term storage.

3.1.3 Thawing of stored PBMCs

Frozen PBMCs were thawed before staining and flow cytometry analysis. 10mL of FACS buffer was prewarmed in a 15mL sterile conical tube to 37ôC in a water bath (GD100, Grant) before the frozen cells were placed in the same water bath for 60-90 seconds until they were approximately 70% thawed. Then, 1mL of prewarmed FACS medium was pipetted to the cryotube slowly and dropwise. The content of the cryotube was then pipetted slowly to the remaining 9mL of prewarmed FACS medium. The cell suspension was flipped once to get a homogenous cell mixture and was thereafter put aside for 10 minutes to give the cells time to adjust and rebalance. The PBMC suspension was then centrifuged for 10 minutes at 200g and 20ôC. Thereafter, the supernatant was discarded and the PBMCs were resuspended in 10mL ice-cold FACS buffer. This cell wash process was repeated once more and the suspension was centrifuged for 5 minutes at 200g at 4ôC. Finally, the supernatant was once again discarded and the PBMCs were resuspended in 1mL of ice-cold FACS buffer and stored on ice.

The cell concentration of the thawed cells was determined with the same protocol as previously described above to detect any major cell death due to the freezing and thawing process.

3.2 Staining and flow cytometry analysis

Cells and fluorochrome spillover beads were stained with antibodies attached to fluorochromes for the flow cytometric analysis.

3.2.1 Cell staining

Unspecific binding of antibody FC-domain to FC-receptors on cells was prevented by incubating the cells with 0.1% normal mouse serum for 15 minutes. Monoclonal antibodies (mAbs) were prepared for staining and diluted in FACS medium, (as shown in table 2).

Table 2. Monoclonal antibodies (mAbs) used to target molecular cell surface structures along with the used fluorochrome and dilutions ratio for cell staining.

mAb target molecule mAb clone Fluorochrome Manufacturer Dilution

CD19 SJ25C1 BV510 BD Biosciences 1:40

CD24 ML5 BV421 BD Biosciences 1:20

CD38 HB-7 PE-Cy7 BD Biosciences 1:50

CD27 M-T271 APC BD Biosciences 1:20

Kappa (к)/lambda (λ) TB28-2/

1-155-2 FITC/PE BD Biosciences 1:10/1:10

(17)

11 The cell samples were washed twice with 1mL FACS buffer and spun down for 8 minutes at 200g at 4ôC. The supernatant was pipetted off completely each time and cells were resuspended in 40µL staining mixture for 20 minutes at 4ôC in the dark (Fig. 4). The samples were then washed with 1mL FACS buffer and centrifuged for 5 minutes at 300g at 4ôC. Thereafter, the supernatant was removed almost completely (leaving 30-40µL of supernatant in the tube) by pipetting and the pellet was dissolved in 200µL of FACS medium and stored on ice. Unlabeled PBMCs were also prepared for calibration of the FACS machine lasers.

Fig. 4 [35]. Cells were stained with fluorochrome labeled antibodies against cell surface markers (CD19, CD24, CD27, CD38 and the kappa and lambda Ig light-chain). Cell staining was performed with minimum amount of light exposure which could energize unspecific antibody binding and false positive fluorochrome light emission.

3.2.1.1 Selection of cell surface target molecules and fluorochromes The target molecules were selected because CD19, CD24, CD27 and CD38 are all included in the basic B-cell panel which makes it possible to discriminate between different B-cell subpopulations. Once the subpopulations have been distinguished it is possible to determine the Igк and Igλ expression for each subpopulation. To target B cells in PBMCs CD19 was used, CD19 is expressed on almost all B-cell linage and B cells are the only known cells that express this antigen in PBMCs. CD24 is expressed on immature and mature B cells and can in combination with CD38 (a marker expressed differently in B-cell development stages) be used for instance to differentiate between different B-cell subpopulations. Likewise, CD24 can in combination with CD27 be used to distinguish between switched/unswitched memory and naive B cells.

The selection of fluorochromes was based on target molecule density on the B-cell surface.

A bright fluorochrome was selected for target molecules with low cell surface density and a lesser bright for target molecules with high density. In addition, consideration was also given to the emission spillover of each fluorochrome into other detectors.

3.2.2 Beads staining

Unlabeled mouse reference beads were used which allowed binding of the same antibody conjugates used for labeling the PBMCs. The beads were used to calibrate the fluorochrome emission spillover with the use of the internal quality control (IQC) program of the BD FACSVerse™ flow cytometer.

The positive (antibody conjugate binding beads) and negative (non-antibody binding beads) were vortexed for a few seconds and 3 drops of each bead suspension were sampled into a 1.5mL Eppendorf tube. These beads were then washed twice with 1mL FACS medium and centrifuged for 5 minutes at 500g. The wash medium was pipetted off completely each time and the beads were finally resuspended in 140µL of FACS buffer. 20µL of bead

(18)

12 suspension was then pipetted into six new 1.5mL Eppendorf tubes and antibodies were added to each tube separately (Table 3). This was then incubated for 10 minutes in the dark.

The beads were then washed with 1mL FACS medium and centrifuged for 5 minutes at 500g. The medium was then pipetted off completely and the pellet was dissolved in 200µL of FACS medium and stored on ice.

Table 3. Monoclonal antibodies (mAbs) used to target molecular cell surface structures along with the used fluorochrome and dilutions ratio for beads staining.

mAb target

molecule mAb

clone Fluorochrome Manufacturer Dilution

CD19 SJ25C1 BV510 BD Biosciences 1:40

CD24 ML5 BV421 BD Biosciences 1:20

CD38 HB-7 PE-Cy7 BD Biosciences 1:50

CD27 M-T271 APC BD Biosciences 1:20

Kappa (к) G20-193 FITC BD Biosciences 1:20

Lambda (λ) 1-155-2 PE BD Biosciences 1:20

3.2.3 Analysis of flow cytometry data

The flow cytometric data was analyzed with FlowJo v.10.4. The difference between HDs and CVID patient subpopulations and Igк and λ expression for different subgroups of B cells was then statistically analyzed with t-tests in GraphPad prism v.6 to detect any statistical significant differences. The aim of the flow cytometric analysis was to acquire cell phenotypic information for patients and aid the selection of patients that would undergo WGS and subsequently GWAS.

3.3 Gating strategy

A schematic representation of the gating strategy used for the flow cytometry data is outlined in fig. 5. First mononuclear cells were gated to exclude any cell debris from the analysis that could affect the percentage of gated lymphocytes with using the side scatter- area (SSC-A) and forward scatter-area (FSC-A). Secondly, duplicates were excluded from the analysis by gating single cells, this was done by first using the forward scatter-height (FSC-H) and FSC-A and subsequently, setting the side scatter-height (SSC-H) against side SSC-A. To gate B cells a histogram was used for the pan B-cell marker (CD19). The B cells were then further gated using CD24 and CD38 as well as CD24 and CD27 to analyze known B-cell subpopulations. Finally, Igк and Igλ L-chains were analyzed for all subpopulations using к and λ. The unidentified and plasmablast subpopulations could be separated as two populations with CD24/CD27 but were observed as one population with CD24/CD38.

(19)

13

Fig. 5. Schematic representation of the gating strategy used for the flowcytometry analysis. The lymphocytes were gated as a percentage of mononuclear cells to eliminate any cell debris from the analysis. Secondly, single cells were gated to eliminate any duplicate cell data. Thereafter, B cells were gated using histogram for the pan B-cell marker - CD19. The B cells were then further gated using CD24 and CD38 as well as CD24 and CD27 into known/unknown B-cell subpopulations. Finally, immunoglobulin Kappa and Lambda light chain percentage of the isolated B-cell subpopulations were gated. Unidentified and plasmablast subpopulations were observed as one population with CD24/CD38 but could be separated into two distinct subpopulations with CD24/CD27 as seen in the figure (dotted line).

3.4 DNA extraction

DNA was isolated from PBMCs with the help of Quick-DNA™ Miniprep Plus Kit (Zymo research). The purified DNA was RNA free and the method was suitable for subsequent DNA-sequencing by SciLifeLab. A total of two DNA extractions were performed, one from a healthy donor (HD 1) and one from a patient with infections and lymphoproliferation (CVID 8).

(20)

14 Approximately 1-5 x 10⁶ cells/200µL of the suspended PBMCs in PBS was added to a microcentrifuge tube to which 200µL of Biofluid & Cell Buffer was added along with 20µL Proteinase K. The tube was then vortexed thoroughly and incubated for 10 minutes at 55^oC.

Thereafter, 420µL of Genomic Binding Buffer was added to the tube and mixed properly.

The mixture was then transferred over to a Zymo-Spin™ IIC-XL Column inside a collection tube. Subsequently, the column was centrifuged at 12,000g for 1 minute, and thereafter, the collection tube was discarded along with its flow-through.

The column was then placed inside a new collection tube and 400µL DNA Pre-Wash Buffer was added to the column. This was then spun down for one minute at 12,000g. Thereafter, the collection tube was emptied and the column was once again loaded into the same collection tube and washed twice with 700µL and subsequently 200µL of g-DNA Wash Buffer and centrifuged for one minute each time. The column was then transferred to a new microcentrifuge tube and 50µL of DNA Elution Buffer was added to the column. Finally, the tube was incubated for 5 minutes at 55^oC and then spun down for one minute. The eluted DNA in 50µL of DNA Elution Buffer was then stored at -20^oC.

3.4.1 DNA quantification and purity assessment

Nucleic acid quantification was performed by the use of NanoDrop spectrophotometer (ND-1000, Thermo Fisher Scientific). The software running the NanoDrop spectrophotometer calculated the sample concentration, which was measured for each sample.

The software also provided the DNA sample purity. The purity was assessed by absorbance measurements at 230nm, 260nm and 280nm. For each of the samples the ratio of absorbance maxima at 260/230nm and 260/280nm was calculated (reference value: 2.0- 2.2 and ~1.8, respectively). The trough of the spectrum in each sample and peak absorbance was also assessed (~230nm and ~260nm, respectively) as an extra purity measure.

To assess the genomic DNA integrity the DNA was run on a 0.8% agarose gel electrophoresis and checked for any fragment smearing. 20-60 ng of DNA was loaded into each well and run at 90V for ~1 hour. The DNA ladder used had a range of 0.1-12 kilobase pairs.

3.5 Ethical approval

The study was approved by the Central Ethical Review Board in Gothenburg (ref.no: 727- 17) on the 15^th of November 2017.

(21)

15

4. Results

In order to determine if the immunophenotyping and/or the Igк expression of B cells in peripheral blood from patients with CVID could be used as marker(s) defining patients with or without autoimmunity and/or lymphoma, these parameters were analyzed by flow cytometry. Subsequently, these results along with clinical data obtained from Östra hospital in Gothenburg were used in an attempt to order CVID patient subgroups within common patterns related to different clinical diagnoses.

The results section will thus first present the flow cytometry gating plots. Followed by, graph plots for immunophenotyping of B-cell subpopulations and identified patterns.

Then, plots for Igк expression of each B-cell subpopulation will be presented along with each identified pattern and overall trend. Finally, the clinical data plots will be presented.

4.1 Age groups for patients with CVID

A total of 19 patients with CVID and 5 HDs were analyzed in this study. The age groups for patients with CVID are presented in table 4 below. No age or sex data was available for the 5 HDs due to the use of buffy coats and therefore no matching was performed between patients with CVID and HDs.

Table. 4. Age groups for patients with CVID.

Age group: < 30 years 31-40 years 41-50 years 51-60 years >60 years CVID

patient: 1, 2, 5 12, 16, 17 4, 7, 9, 10, 11,

13, 15, 19 3, 6, 8, 18 14

4.2 Gating of flow cytometry data

The flow cytometry data from HDs and patients with CVID were gated according to the strategy outlined in fig. 5 (above). The cell frequency results etc. are presented below in each plot.

4.2.1 Healthy donors

The results from one representative HD (HD 1) are presented in fig. 6. The HDs did not present any unidentified or plasmablast cell populations, hence, the к/λ plots are omitted.

(22)

16

Fig. 6. The flow cytometry plots for one of the healthy donors (HD 1), the same gating strategy was used as outlined in fig. 5 above.

4.2.2 CVID patients lacking the unidentified and plasmablast cell population

The results from one representative CVID patient (CVID 10) are presented in fig. 7 below.

This particular patient did not present any unidentified or plasmablast cell population, hence, those plots are omitted.

(23)

17

Fig. 7. Flow cytometry plots for one of the CVID patients (CVID 10).

4.2.3 CVID Patients displaying an unidentified cell population

The flow cytometry plots for one representative CVID patient (CVID 6) that, in addition to transitional, naïve and memory B cells displayed an unidentified (CD24^low CD27^lowCD38^low) B-cell subpopulation but no plasmablasts are presented in fig. 8 below. The unidentified subpopulation was also shown to correspond to a distinct B-cell subpopulation when CD24/CD38 was used for gating.

(24)

18

Fig. 8. Flow cytometry plots for one of the CVID patients (CVID 6). This patient displayed in addition to transitional, naïve and memory B cells an unidentified B-cell subpopulation.

4.3 Cell frequencies

Based on the flow cytometry data analysis, graphs were made in order to detect any relevant patterns that could be compared and tied to a particular clinical diagnosis. All cell frequencies are given as a percentage of the previously gated cells and are given in table form in appendix 1 along with Ig к:λ ratios for all B-cell subpopulations.

4.3.1 Percent lymphocytes of PBMCs

The percentage of lymphocytes within PBMCs from HDs and patients with CVID are shown in fig. 9. The mean value and standard deviation (SD) for HD as a group was 77 ± 6.4%

while for the CVID group the same values were 62 ± 17.6%. The highest measured percentage of lymphocytes for HD and CVID was the same in both groups (84 percent for HD 1 and CVID 4) while the lowest percentage per group was 66% for HD 2 and 17% for

(25)

19 CVID 18. Even though no statistical significant difference could be detected between the HD and CVID group when an unpaired t-test was performed (p-value 0.081) there was a larger variation in patients with CVID compared to HDs.

(A) (B)

Fig. 9. (A) Graph showing the percentage of lymphocytes within PBMCs for healthy donors (HDs) and patients with CVID. (B) Scatter plot for visualizing the spread of HD and CVID data points, mean and standard deviation bars for the two groups are shown in the graph.

4.3.2 Percent B cells of lymphocytes

The percentage of B cells (CD19+ lymphocytes) are shown in fig. 10 for HDs and patients with CVID. The mean and SD for HD and CVID was 3.6 ± 1.5 and 4.5 ± 3.6 %, respectively.

The highest percentage of B cells for HD was 6% for HD 1 and the lowest 2% for HD 5, for patients with CVID it was 15% for CVID 1 and 1% for CVID 2, 4 and 11. Even if no statistical significant difference was detected between the groups (p-value 0.58) the variation in the CVID patient group appears to be larger than for HDs.

1 2 3 4 5 1 2 3 4 5 6 7 8 9 10

11 12

13 14

15 16

17 18

19 0

1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 1 0 0

% Lymphocytes of PBMCs

H D C V ID

HD

CVID 0

2 0 4 0 6 0 8 0 1 0 0

% Lymphocytes of PBMCs

(26)

20

(A) (B)

Fig. 10. (A) Percentage of B cells of gated lymphocytes for healthy donors (HDs) and patients with CVID. (B) Graph representing the spread of data points for HD and CVID patients along with mean values and standard deviation bars for each group.

4.3.3 B-cell subpopulations in healthy donors

The percentage of transitional (Tr), naïve (N) and memory B cells (MBC) as a proportion of CD19+ B cells are shown in fig. 11 for HDs along with their mean values. The mean value for transitional B cells was 3.2%, for naïve 53.2% and 32.6% for memory B cells.

Fig. 11. Percent of B-cell subpopulations for healthy donors: transitional (Tr), naïve (N) and memory B cells (MBC) as a proportion of CD19+ B cells.

4.3.4 B-cell subpopulations in patients with CVID

The percentage of Tr, N and MBC as a proportion of CD19+ B cells for patients with CVID are shown in fig. 12 along with the mean values for the subpopulations. For four CVID patients (2, 4, 6 and 11) an unidentified (UID) B-cell population was observed in addition to Tr, N and MBCs, and in three CVID patients (4, 5 and 12) plamablasts (PLB) (plasma cells) were detected. The spread of the data points was generally wider for CVID compared to HDs. A statistical significant difference could be detected between transitional B cells from HD and transitional B cells from patients with CVID (p-value 0.046).

1 2 3 4 5 1 2 3 4 5 6 7 8 9 1011 12 131415 16 1718 19 0

5 1 0 1 5 2 0

% B cells of lymphocytes

H D C V ID

HD

CVID 0

5 1 0 1 5 2 0

% B cells of lymphocytes

Tr N

MBC 0

1 0 2 0 3 0 4 0 5 0 6 0 7 0

% Subpopulation of B cells for HD H D 1

H D 2 H D 3 H D 4 H D 5

(27)

21

Fig. 12. Percentage of transitional (Tr), naïve (N), memory (MBC), unidentified (UID) and plasmablast (PLB) CD19+ B-cell subpopulations for patients with CVID.

4.3.4.1 Pattern 1 for patients with CVID

Subsequent to determining the proportions of B-cell subpopulations, based on flow cytometry results, an attempt was made to identify patterns that could be used to classify patients into groups that would later be used for clinical data comparison. The cutoff value for what was considered an increase or decrease was set to equal or greater than five percent. This cutoff was set to limit the number of patterns to a manageable level and to cluster relatively similar patterns together without losing to much data details in the process.

The first CVID pattern observed was a resemblance of what was observed in HDs (HD 1, 2 ,4 and 5) with an increasing percentage of B-cell subpopulation in the following order: Tr

< MBC < N (Fig. 13). For CVID 18 an UID B-cell population was observed and for CVID 5 and 12 a low percentage of PLBs were detected.

1 2 3 4 5 6 7 8 9 1 0

1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9

Tr N

MBC

UID

PLB 0

1 0 2 0 3 0 4 0 5 0 6 0 7 0

% Subpopulation of B cells for CVID patients

(28)

22

Fig. 13. Pattern 1 for patients with CVID displayed an increasing percentage of cells in the following order:

transitional (Tr) < memory (MBC) < naïve (N) B cells which resemble what was observed in healthy donors (HD 1, 2, 4 and 5). CVID patient 18 also displayed an unidentified (UID) B-cell subpopulation while patients 5 and 12 displayed low percentage of plasmablasts (PLBs).

The second pattern observed had an increasing percentage of cells in the following order:

Tr < N < MBCs (Fig. 14) which is a relatively similar pattern to what was observed for HD 3.

Fig. 14. Pattern 2 for patients with CVID displays an increasing percentage of cells in the following order:

transitional (Tr) < naïve (N) < memory B cells (MBC) which resembles what was observed in healthy donor 3 (HD 3) in fig. 11.

1 2 3 4 5 6 7 8 9 1 0

1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9

Tr N

MBC 0

1 0 2 0 3 0 4 0 5 0 6 0 7 0

1 2 3 4 5 6 7 8 9 1 0

1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9

Tr N

MBC

UID

PLB 0

1 0 2 0 3 0 4 0 5 0 6 0 7 0

(29)

23 4.3.4.3 Pattern 3 for patients with CVID

The third CVID pattern observed had an increasing percentage of B-cell subpopulations in the following order: Tr < N/MBCs (Fig. 15). For one of the CVID patients (CVID 4) UID and PLB subpopulations were also observed. This pattern could potentially represent a slight deviation from the HDs.

Fig. 15. Pattern 3 for patients with CVID displays an increasing percentage of cells in the following order:

transitional (Tr) < naïve (N)/memory B cells (MBC). One of these CVID Patients (CVID 4) also displayed an unidentified (UID) B-cell subpopulation as well as a plasmablast (PLB) subpopulation.

The forth pattern observed was an decreasing percentage of B-cell subpopulations in the following order: Tr > N > MBCs (Fig. 16). For CVID patients 6 and 11 an unidentified B-cell population was also observed. This pattern was not observed in HDs.

Tr N

MBC

UID

PLB 0

5 1 0 1 5 2 0 2 5 3 0 3 5

1 2 3 4 5 6 7 8 9 1 0

1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9

(30)

24

Fig. 16. Pattern 4 for patients with CVID. Subpopulations displayed a decreasing percentage of cells in the following order: transitional (Tr) > naïve (N) > memory B cells (MBC). CVID patients 6 and 11 also displayed an unidentified (UID) B-cell subpopulation.

4.3.5 Igк expressing B-cell subpopulations in healthy donors

The percentage of cells expressing either Igк or Igλ L-chain were determined after gating on CD19+ Tr, N and MBCs. For the ease of comparison, only the proportion of B cells expressing Igк is presented below because there were no double-expressing cells (expressing both Igк and Igλ L-chains). Furthermore, the number of Igк plus Igλ producing cells within each gated subpopulation corresponded to ~100%.

The results for HDs are shown in fig. 17. The mean values and SDs for Igк expressing Tr, N and MBCs were 57.2 ± 3.3%, 58.8 ± 3.0% and 57.2 ± 2.5%, respectively. These mean values correlated well with current literature values of around 60% for Igк for healthy individuals.

Fig. 17. Percentage of immunoglobulin kappa light chain expression for healthy donors (HD) along with the mean values for transitional (Tr), naïve (N) and memory B cells (MBC).

1 2 3 4 5 6 7 8 9 1 0

1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9

Tr N

MBC

UID 0

1 0 2 0 3 0 4 0 5 0 6 0

Tr N

MBC 5 0

5 5 6 0 6 5 7 0

% Kappa of B cell subpopulations for HD

H D 1 H D 2 H D 3 H D 4 H D 5

Ig light chain expression in B cells from patients with common variable immunodeficiency

Ig light chain expression in B cells from patients with

common variable immunodeficiency

Is there a correlation to a particular clinical diagnosis?

Abstract

Abbreviations

Table of contents

1. Introduction

1.1 The Immune system

1.1.1 B cells

1.1.2 Antigen binding diversity of antibodies

1.1.3 Immunoglobulin Kappa and lambda light chain expression

1.1.4 Immunological tolerance

1.1.5 Immune dysregulation

1.2 Common variable immunodeficiency

1.2.1 Symptoms in CVID

1.2.2 CVID Diagnostics

1.2.3 Autoimmunity in CVID

1.2.4 Lymphoma in patients with CVID

1.2.5 Genetics in CVID

1.2.6 Prophylactic treatment

2. Objective

3. Methods

3.1 Isolation of PBMCs from healthy donors and CVID patients

3.1.1 Determination of cell concentration

3.1.2 Freezing of PBMCs

3.1.3 Thawing of stored PBMCs

3.2 Staining and flow cytometry analysis

3.2.1 Cell staining

3.2.2 Beads staining

3.2.3 Analysis of flow cytometry data

3.3 Gating strategy

3.4 DNA extraction

3.4.1 DNA quantification and purity assessment

3.5 Ethical approval

4. Results

4.1 Age groups for patients with CVID

4.2 Gating of flow cytometry data

4.2.1 Healthy donors

4.2.2 CVID patients lacking the unidentified and plasmablast cell population

4.2.3 CVID Patients displaying an unidentified cell population

4.3 Cell frequencies

4.3.1 Percent lymphocytes of PBMCs

4.3.2 Percent B cells of lymphocytes

4.3.3 B-cell subpopulations in healthy donors

4.3.4 B-cell subpopulations in patients with CVID

4.3.5 Igк expressing B-cell subpopulations in healthy donors