Dissection of the pre-germinal center B-cell maturation pathway in common variable immunodeficiency based on standardized flow cytometric EuroFlow tools

Dissection of the Pre-Germinal Center B-Cell Maturation Pathway in Common Variable Immunodeficiency Based on Standardized Flow Cytometric EuroFlow Tools Lucı a del Pino-Molina 1 , Eduardo Lo pez-Granados 1 *, Quentin Lecrevisse 2,3 , Juan Torres Canizales 1 , Martı nPe rez-Andre s 2,3 , Elena Blanco 2,3 , Marjolein Wentink 4 , Carolien Bonroy 5 , Jana Nechvatalova 6 , Tomas Milota 7 , Anne-Kathrin Kienzler 8 , Jan Philippe  5 , Ana E. Sousa 9 , Mirjam van der Burg 10 , Tomas Kalina 11 , Jacques J.M. van Dongen 12 *and Alberto Orfao 2,3 on behalf of the EuroFlow PID Consortium 1 Clinical Immunology Department, La Paz University Hospital and Lymphocyte Pathophysiology in Immunodeficiencies Group, La Paz Institute for Health Research (IdiPAZ) and Center for Biomedical Network Research on Rare Diseases (CIBERER U767), Madrid, Spain, 2 Clinical and Translation Research Program, Cancer Research Centre (IBMCC, USALCSIC), Department of Medicine, Cytometry Service (NUCLEUS), University of Salamanca (USAL), Institute of Biomedical Research of Salamanca (IBSAL), Salamanca, Spain, 3 Biomedical Research Networking Centre Consortium of Oncology (CIBERONC) Instituto de salud Carlos III, Madrid, Spain, 4 Department of Immunology, Erasmus University Medical Center (Erasmus MC), Rotterdam, Netherlands, 5 Department of Laboratory Medicine, University Hospital Ghent, Ghent, Belgium, 6 Department of Allergology and Clinical Immunology, Faculty of Medicine, Masaryk University and St Anne’s University Hospital in Brno, Brno, Czechia, 7 Department of Immunology, Second Faculty of Medicine, Charles University and Motol University Hospital, Prague, Czechia, 8 Nuffield Department of Medicine, Experimental Medicine Division, University of Oxford, Oxford, United Kingdom, 9 Instituto de Medicina Molecular, Faculdade de Medicina, Universidade de Lisboa, Lisboa, Portugal, 10 Department of Pediatrics, Laboratory for Immunology, Leiden University Medical Center, Leiden, Netherlands, 11 CLIP - Childhood Leukemia Investigation Prague, Department of Pediatric Hematology and Oncology, 2nd Faculty of Medicine, Charles University and University Hospital Motol, Prague, Czechia, 12 Department of Immunohematology and Blood Transfusion, Leiden University Medical Center (LUMC), Leiden, Netherlands Introduction: Common Variable Immunodeficiency (CVID) is characterized by defective antibody production and hypogammaglobulinemia. Flow cytometry immunophenotyping of blood lymphocytes has become of great relevance for the diagnosis and classification of CVID, due to an impaired differentiation of mature post-germinal-center (GC) classswitched memory B-cells (MBC) and severely decreased plasmablast/plasma cell (Pb) counts. Here, we investigated in detail the pre-GC B-cell maturation compartment in blood of CVID patients. Methods: In this collaborative multicentric study the EuroFlow PID 8-color Pre-GC B-cell tube, standardized sample preparation procedures (SOPs) and innovative data analysis tools, were used to characterize the maturation profile of pre-GC B-cells in 100 CVID patients, vs 62 age-matched healthy donors (HD). Results: The Pre-GC B-cell tube allowed identification within pre-GC B-cells of three subsets of maturation associated immature B-cells and three subpopulations of mature Frontiers in Immunology | www.frontiersin.org February 2021 | Volume 11 | Article 6039721 Edited by: Sudhir Gupta, University of California, Irvine, United States Reviewed by: Hans-Hartmut Peter, University of Freiburg Medical Center, Germany Neil Romberg, Children’s Hospital of Philadelphia, United States Manisha Rajan Madkaikar, National Institute of Immunohaematology (ICMR), India *Correspondence: Jacques J.M. van Dongen [email protected] Eduardo Lo pez-Granados [email protected] Specialty section: This article was submitted to Primary Immunodeficiencies, a section of the journal Frontiers in Immunology Received: 08 September 2020 Accepted: 29 December 2020 Published: 17 February 2021 Citation: del Pino-Molina L, Lo pez-Granados E, Lecrevisse Q, Torres Canizales J, Pe rez-Andre sM,Blanco E, Wentink M, Bonroy C, Nechvatalova J, Milota T, Kienzler A-K, Philippe J, Sousa AE, van der Burg M, Kalina T, van Dongen JJM and Orfao A (2021) Dissection of the Pre-Germinal Center B-Cell Maturation Pathway in Common Variable Immunodeficiency Based on Standardized Flow Cytometric EuroFlow Tools. Front. Immunol. 11:603972. doi: 10.3389/fimmu.2020.603972 ORIGINAL RESEARCH published: 17 February 2021 doi: 10.3389/fimmu.2020.603972 naïve B-lymphocytes. CVID patients showed overall reduced median absolute counts (vs HD) of the two more advanced stages of maturation of both CD5 + CD38 +/++ CD21 het CD24 ++ (2.7 vs 5.6 cells/µl, p=0.0004) and CD5 + CD38 het CD21 + CD24 + (6.5 vs 17 cells/ µl, p<0.0001) immature B cells (below normal HD levels in 22% and 37% of CVID patients). This was associated with an expansion of CD21 - CD24 - (6.1 vs 0.74 cells/µl, p<0.0001) and CD21 - CD24 ++ (1.8 vs 0.4 cells/µl, p<0.0001) naïve B-cell counts above normal values in 73% and 94% cases, respectively. Additionally, reduced IgMD + (21 vs 32 cells/µl, p=0.03) and IgMD - (4 vs 35 cells/µl, p<0.0001) MBC counts were found to be below normal values in 25% and 77% of CVID patients, respectively, always together with severely reduced/undetectable circulating blood pb. Comparison of the maturation pathway profile of pre-GC B cells in blood of CVID patients vs HD using EuroFlow software tools showed systematically altered patterns in CVID. These consisted of: i) a normally-appearing maturation pathway with altered levels of expression of >1 (CD38, CD5, CD19, CD21, CD24, and/or smIgM) phenotypic marker (57/88 patients; 65%) for a total of 3 distinct CVID patient profiles (group 1: 42/88 patients, 48%; group 2: 8/88, 9%; and group 3: 7/88, 8%) and ii) CVID patients with a clearly altered pre-GC B cell maturation pathway in blood (group 4: 31/88 cases, 35%). Conclusion: Our results show that maturation of pre-GC B-cells in blood of CVID is systematically altered with up to four distinctly altered maturation profiles. Further studies, are necessary to better understand the impact of such alterations on the post-GC defects and the clinical heterogeneity of CVID. Keywords: CVID, Pre-GC B-cell tube, pre-GC maturation pathway, expression markers, EuroFlow standardization INTRODUCTION Common Variable Immunodeficiency (CVID) is the most prevalent symptomatic primary immunodeficiency (PID). It is characterized by defective antibody production that leads to hypogammaglobulinemia (1–3) with an increased susceptibility to infections, associated in some CVID patients with enteropathy, autoimmunity, lymphoproliferation, and/or risk of lymphoid malignancy due to more profound immunological dysregulation (4,5). Despite distinct monogenic defects are present in a minor fraction (<20%) of cases, and other complex oligo or polygenic genetic predisposition (6), and epigenetic alterations (e.g., impaired demethylation in genes relevant for the B cell functions) have been associated with the development of CVID (7), a well-defined pathogenic mechanism still remains to be identified in most CVID patients. Thus, assessment of the distribution of lymphocytes, particularly post-germinal center (GC) B-cells and plasmablasts/plasma cells in blood of suspicious patients by flow cytometry has become of great relevance for the diagnosis and classification of CVID (8,9). Impaired post-GC B cell maturation in the periphery (i.e., in blood and secondary lymphoid tissues) is a hallmark of CVID. However, CVID is a rather heterogeneous disease from the clinical, genetic and immunologic point of view. Thus, several classification algorithms have been proposed for CVID, which are based on the specific alterations encountered for the major B cell populations in blood (8), their proliferation history and somatic hypermutation levels (9), in combination or not with the clinical manifestations of the disease and/or more sophisticated computational (i.e., hierarchical clustering) approaches (10). Overall, impaired differentiation of mature post-GC B-cells, consisting of severely reduced circulating class-switched memory B-cells (MBC) and strongly decreased (i.e., undetectable) plasmablast/plasma cell production, are the most consistent defects in CVID. Because of this, demonstration of reduced class-switched MBC is now used among the diagnostic criteria proposed by the European Society for Immunodeficiencies (ESID) for CVID (11). In turn, depending on the specificdefects encountered in the post-GC MBC and Pb compartments in blood, and the severity of such defects, distinct CVID patient subgroups, associated with distinct clinical profiles, have also been identified (12). Apart from the alterations in post-GC B-cells and plasmablasts/plasma cells, an increasing number of evidences Abbreviations: APS, automated populations separator; BCR, B-cell receptor; BM, bone marrow; CVID, common variable immunodeficiency; ESID, European Society for Immunodeficiencies; FCS, flow cytometry standard; FSC, forward light scatter; GC, germinal center; HD, healthy donor; Ig, immunoglobulin; IUIS, International Union of Immunological Societies; MBC, memory B-cell; MFI, mean fluorescence intensity; Pb, plasmablast/plasma cell; PC, principal component; PCA, principal component analysis; PID, primary immune deficiency; sm, surface membrane; SOP, standard operating procedure; SSC, sideward light scatter; WBC, white blood cell. del Pino-Molina et al. Dissection of the Pre-GC B-Cell Maturation Pathway in CVID Frontiers in Immunology | www.frontiersin.org February 2021 | Volume 11 | Article 6039722 indicate that the production and maturation of B-cells in bone marrow (BM) is also altered in at least a fraction (e.g., around one third) of all CVID patients due to either a maturation blockade (13) and/or an altered bone marrow environment which is non-permissive for B-cell maturation (14). Thus, a significant proportion of CVID patients display reduced absolute B-cell counts in blood (5) and an early B-cell maturation arrest in BM (15). In addition, expansion of transitional/immature B cells, and CD21 low B-cells has also been reported in a subset of CVID patients (8,16). Altogether, these findings further support an impaired maturation of pre-GC B-cells in CVID. Herein, we investigated in detail the pre-GC B-cell maturation compartment in blood of 100 CVID patients, taking advantage of the novel and standardized flow cytometric approaches developed by EuroFlow for this purpose (17–19), e.g., the recently proposed EuroFlow PID 8-color antibody panels (20) that can be easily implemented in most diagnostic laboratories worldwide, together with the EuroFlow standard operating procedures (SOPs) for sample preparation, data acquisition and analysis, including innovative data analysis tools recently developed by EuroFlow to assess normal vs altered pre-GC B-cell maturation profiles in blood (17,20–23). MATERIAL AND METHODS Patients, Controls, and Samples Overall, 100 adult CVID patients -50 men and 50 women; median age: 41 years (y); range: 16–82yand 62 healthy donors (HD) not related to the patients (33 men and 29 women; median age: 34y; range: 19–67y) were studied in parallel, at eight different EuroFlow-PID centers. CVID was diagnosed locally at each center, according to the ESID criteria (3,24). Relevant clinical data on CVID patients was obtained from the patients’health electronic records or from national patient registries and collected at each of the 8 participating centers, including data on: patient age, gender, immunoglobulin (Ig) levels and response to vaccination at diagnosis, together with data on prior infections and type of infections (e.g., upper and lower bacterial respiratory infections, viral and fungal infections), autoimmunity (e.g., cytopenias, organ-based and systemic auto-immunity), lymphoproliferation, lymphoid interstitial pneumonitis (LIP), granulomas, splenomegaly, hepatomegaly, bronchiectasias, enteropathy, and malignancy, as well as prior therapy, including Ig replacement therapy. Blood samples were obtained, processed and measured by flow cytometry at each of the 8 participating centers after informed consent had been given by each individual participant, according to the principles of the Declaration of Helsinki. The study was approved by the local Ethics Committees of the participating centers: Hospital Universitario La Paz, Madrid, Spain (PI-2833 and 2009/3348/I); Charles University, Prague, Czech Republic (15-28541A); Erasmus MC, Rotterdam, The Netherlands (MEC-2013-026); St Anne´s University, Brno, Czech Republic (METC 1G2015); BRC-Translational Immunology Lab, University of Oxford, Oxford, United Kingdom; University of Salamanca, Salamanca, Spain (USAL/ CSIC 20-02-2013); University Hospital of Ghent, Belgium (B670201523515); and Faculdade de Medicina da Universidade de Lisboa and Centro Hospitalar Universitario Lisboa Norte, Lisbon, Portugal (937/13). Flow Cytometric Identification of B-Cells, Plasmablasts/Plasma Cells and Their Subsets in Blood Blood samples from both CVID patients and HD were processed and stained at each center with the EuroFlow 8-color PIDOT and Pre-GC B-cell tubes, following the EuroFlow SOPs for staining of cell surface membrane (sm) markers only, as previously described (20–22). Details about the specific antibody clones and fluorochrome-conjugated reagents used are provided in Supplementary Table 1. Instrument set-up and calibration were performed prior to data acquisition on ≥1x10 6 cells (range: 1 x 10 6 -5 x 10 6 cells) in FACSCanto II flow cytometers −Becton/Dickinson Biosciences (BD), San Jose, CA-, following the EuroFlow SOPs available at www.EuroFlow.org (21). Data analysis was performed centrally on pseudoanonymazed flow cytometry standard (FCS) data files deposited in the EuroFlow data repository, using the Infinicyt software (Cytognos SL, Salamanca, Spain). For data analysis, a standardized gating strategy was used for identification of all pre-GC (defined as CD19 + CD27 - sIgM + Blymphocytes) and post-GC B-cell subsets (defined as CD19 + CD27 + or CD19 + CD27 - smIgM - B-cells) present in blood, based on the EuroFlow-PID Pre-GC B cell tube as illustrated in Supplementary Figure 1.Briefly, CD19 + B-cells and plasmablasts/plasma cells were both identified by their low-tointermediate forward (FSC) and sideward (SSC) light scatter properties after excluding debris and cell doublets. Subsequently, both cell subsets were sub-classified into 11 different subsets based on their staining profile for CD19, CD38, CD24, CD21, CD27, CD5, smIgM, and smIgD: a) CD27 - CD38 hi CD24 hi CD5 + smIgM ++ D + immature/transitional B cells; b) CD27 - CD38 - CD24 het CD5 het smIgM + IgD ++ mature naive B lymphocytes; c) CD27 + CD38 lo CD5 - CD24 het smIgM ++ D + (MD + ) unswitched MBCs; d) CD27 +/- CD38 lo CD5 - CD24 het smIgM - D - (MD - ) switched MBCs; and, e) CD27 ++ CD38 hi CD5 - CD21 - CD24 - plasmablasts/PCs. Immature/transitional B cells were further sub-classified according to the pattern of expression of CD38, CD5, CD21, and CD24 into three subsets of increasingly more mature B-lymphocytes: a1) CD5 - CD38 ++ CD21 het CD24 ++ ; a2) CD5 + CD38 +/++ CD21 het CD24 ++ , and a3) CD5 + CD38 het CD21 + CD24 + immature/transitional B lymphocytes. In turn, mature naive B-lymphocytes and unswitched MBCs were also further sub-classified into three subsets each, based on the expression profile for CD21 and CD24, into CD21 + CD24 + , CD21 - CD24 ++ , and CD21 - CD24 - mature naïve B cells and unswitched MBC, respectively. For each B cell population, absolute counts were calculated using a dual platform assay based on the white blood cell (WBC) count, as assessed in a conventional hematological cell counter, and the percentage of total B cells obtained with the PIDOT tube del Pino-Molina et al. Dissection of the Pre-GC B-Cell Maturation Pathway in CVID Frontiers in Immunology | www.frontiersin.org February 2021 | Volume 11 | Article 6039723 for the same sample, as previously reported (22). Normal reference ranges were defined by the 5 th and 95 th percentile values observed in blood of 62 (ageand sex-matched) HD analyzed in parallel to the CVID patients (Supplementary Table 2). Pre-GC B-Cell Maturation Pathway in Blood A database reflecting the normal B-cell maturation pathway of pre-GC B-lymphocytes in blood was built by merging data files from 18 representative HD, using the Infinicyt software (Cytognos SL) and previously described procedures (25). For this purpose, pre-gated data files which specifically contained gated data exclusively on the three different subsets of immature blood B cells (CD5 - CD38 ++ CD21 het CD24 ++ , CD5 + CD38 +/++ CD21 het CD24 ++ and CD5 + CD38 het CD21 + CD24 + immature B cells), together with both the CD21 + CD24 + and CD21 - CD24 - mature naive B cell subsets, from blood of 18 HD stained with the Pre-GC B-cell tube were merged into a single data file. Mature naive CD21 - CD24 ++ B cells were not included in the pre-GC Bcell database since this subset is barely detectable in normal blood from HD (26). Then, the merged data file was used to define the maturation pathway of pre-GC B-cells using the maturation tool developed by EuroFlow and implemented in Infinicyt (v2.0-4b for EuroFlow members only). This tool allows for automatic i) definition of vectors that reflect maturation pathways based on curve analysis algorithms, ii) classification of events into different maturation stages arbitrarily set at equal distances, iii) calculation of descriptive statistics for all events classified within each maturation stage, followed by direct visualization in a (balanced) 3-dimension (3D) APS (automated population separator) diagram, constructed using the first three principal components (PC1 to PC3) derived from PC analysis (PCA) performed with the Infinicyt software (Figure 1A). Thus, based on the maturation tool of the Infinicyt software, 10 distinct pre-GC B-cell maturation stages were (arbitrarily) defined and the normal mean fluorescence intensity (MFI) range (2SD) per maturation stage was calculated for each individual A B FIGURE 1 | Illustrating example of the normal (reference) pre-germinal-center (GC) maturation pathway. Overall phenotypes (A) and expression levels for individual markers per stage of maturation (B) are shown for each individual donor (n=18) included in the normal reference pre-GC B-cell maturation pathway used to define reference normal values, depicted as color-coded lines per marker/parameter. del Pino-Molina et al. Dissection of the Pre-GC B-Cell Maturation Pathway in CVID Frontiers in Immunology | www.frontiersin.org February 2021 | Volume 11 | Article 6039724 cell surface marker. Afterward, deconvolution of the maturation vector defined by the median MFI values per maturation stage (n=10) data was plotted for the whole set of phenotypic markers included in the EuroFlow Pre-GC B-cell tube (Figure 1A). Subsequently, median (range) and mean (+2 SD) fluorescence intensity (MFI) values per marker in HD blood pre-GC B-cells together with the corresponding B cell percentage, were plotted along the different maturation stages and used as normal reference range (Figure 1B). Subsequently, gated pre-GC Bcells from every individual CVID case were plotted against the normal reference maturation database and differences in MFI values vs. the normal blood range were recorded per marker for each of the 10 pre-established pre-GC B-cell maturation stages. MFI values below 2SD or above 2SD of the normal reference range (per maturation stage) for at least two consecutive stages of maturation of pre-GC B-cells, were considered to be altered. In parallel, a graphical display of the values per marker along the pre-GC B-cell maturation pathway, in which each patient is represented against the database, was obtained. Statistical Analyses Statistical analyses were performed with GraphPad Prism Software version 6.0 (GraphPad software, San Diego, CA). To define normal ranges for each B cell subset identified in blood, 5 th and 95 th percentile values from the 62 adult HD were used. Group comparisons were performed using the Mann-Whitney U and Kruskal-Wallis tests (for continuous variables) or the Fisher exact and X 2 tests (for categorical variables). Clustering analysis based on K-means was performed using the JMP software (free trial version 14; SAS Institute Inc., Cary, NC) based on the immunophenotypic profiles and relative distributions of pre-GC B-cells along the pre-GC maturation pathway, per maturation stage. Cluster analysis was performed by simultaneously comparing the MFI values for each surface marker and the percentage of events per stage of maturation per CVID patient against the maturation reference database. P-values<0.05 were considered to be associated with statistical significance and coded hereafter as follows: *p-value<0.05; ** p-value<0.01; *** pvalue<0.001; and, **** p-value<0.0001. RESULTS Distribution of Pre-GC B-Cell Subsets in Blood of CVID Patients Based on the data provided by the PID-orientation tube (PIDOT) (22), and the Pre-GC B-cell tube, detailed characterization of B cells in peripheral blood was achieved for a total of 11 distinct B cell subsets: i) immature/transitional B-cells (including three maturation-associated populations of CD5 - CD38 ++ CD21 het CD24 ++ ,CD5 + CD38 +/++ CD21 het CD24 ++ ,andCD5 + CD38 het CD21 + CD24 + immature B-cells); ii) mature naïve Bcells (and their three CD21 + CD24 + , CD21 - CD24 ++ , and CD21 - CD24 - subsets); iii) unswitched IgMD + , IgM + -only, and IgD + -only MBC (and their CD21 + CD24 + , CD21 - CD24 - , and CD21 - CD24 ++ subsets); iv) switched IgMD - MBC; and v) plasmablasts/ plasma cells (Figures 2A–C). Overall, the total B-cell count in blood of CVID was significantly reduced vs. age-matched HD (median: 149 vs 206 cells/µl; p=0.04). Suchdecreasewasmostlyduetoasignificantreductionofimmature B-cells (11 vs 27 cells/µl, p<0.0001), IgMD + MBC (21 vs 32 cells/µl, p=0.03) and particularly, IgMD - MBCcounts(4vs35cells/µl, p<0.0001), in the absence of virtually no plasmablasts/plasma cells (Figure 3A). Despite the overall reduced median B-cell counts observed in CVID, a significant overlap with HD was still observed with variable frequencies and patterns of alteration among CVID patients. Thus, decreased counts below normal values (<5 th percentile of age-matched HD) of IgMD - MBC were detected in 77% of the CVID patients investigated, together with undetectable plasmablasts/plasma cells in 100% of cases. In contrast, reduced immature/transitional B cells and IgMD + MBC counts were only found in 29% and 25% of cases, respectively (Figure 3A). In turn, only a small percentage of all CVID patients showed reduced mature naïve B-cell counts (11%). Altogether, the decreased numbers of the distinct B-cell subsets led to overall low total Bcell counts in blood of 21% of all CVID patients (Figure 3A). More detailed analysis of the pre-GC B-cell compartment also showed distinct patterns of alteration for different subsets of immature B-cells and mature naive B-cells. Thus, reduced counts (vs age-matched HD) of the more advanced stages of maturation of CD5 + CD38 +/++ CD21 het CD24 ++ (2.7 vs 5.6 cells/µl, p=0.0004) and CD5 + CD38 het CD21 + CD24 + (6.5 vs 17 cells/ µl, p<0.0001) immature/transitional B cells was detected in 22% and 37% of CVID patients. In contrast, the less differentiated CD5 - CD38 ++ CD21 het CD24 ++ immature/transitional B lymphocytes (0.79 vs 0.89, p>0.05) were decreased in blood in only 7% of CVID patients (Figure 3B). Regarding mature naive B-cells, an increase in CD21 - CD24 - (6.1 vs 0.74 cells/µl, p<0.0001) and CD21 - CD24 ++ (1.8 vs 0.4 cells/µl, p<0.0001) naive B cell counts was observed in CVID vs. HD, with a clear bimodal distribution (Figure 3B) due to the presence of a major subgroup of patients (73% and 94%, respectively) presenting a significant expansion of these two naive B-cell subsets, in association or not with low CD21 + CD24 + naïve B-cell counts, which were found to be reduced in only 21% of CVID patients (Figure 3B). Regarding post-GC MBC, CVID patients displayed a significant reduction (vs. HD) of CD21 + CD24 + IgMD + MBC (16 vs 31 cells/µl, p=0.0006), associated with normal or slightly increased (p>0.05) CD21 - CD24 - and CD21 - CD24 ++ IgMD + MBC numbers (Supplementary Figure 2). Plasmablasts/ plasma cells, were either not detected or severely reduced in 100% of CVID patients. Altogether, these results suggest the existence of different maturation blockades and profiles in CVID, which frequently also involve pre-GC B-cells, in addition to postGC MBC and plasmablasts/plasma cells. Pre-GC B-Cell Maturation Profile in Normal Blood Based on the innovative maturation tools developed by EuroFlow (18,25), a normal pre-GC B-cell reference del Pino-Molina et al. Dissection of the Pre-GC B-Cell Maturation Pathway in CVID Frontiers in Immunology | www.frontiersin.org February 2021 | Volume 11 | Article 6039725 maturation pathway was built as described above in the material an methods section and illustrated in Figure 1A. Detailed analysis of the normal pre-GC B-cell maturation profile showed downregulation of CD38 from stage 3 on, associated with strong CD5 and CD24 expression at early stages (stages 2–4 and stages 1–3, respectively), CD5 becoming negative from stage 7 onwards while CD24 progressively decreased from stage 4 onward; in turn, CD21 was strongly expressed from stage 1 until the last stages of maturation (stages 9 and 10), when it was downregulated. Among the other markers investigated, smIgM showed slightly higher expression levels at early (stages 1–3) vs later stages with stable levels from stage 4 to 10, while IgD and CD19 showed similarly stable expression levels along all stages of maturation of pre-GC B-cells. By definition CD27 was not expressed in pre-GC B-cells (Figure 1B). Altered Maturation Profiles of Blood Pre-GC B-Cells in CVID Direct comparison of the phenotype of maturation-associated blood pre-GC B-cell subsets from CVID patients vs. the normal maturation was performed in 88 patients by plotting phenotypic data from each CVID against the normal reference maturation database (Figures 4 and 5). For better visualization of the phenotypic deviations from normal, a normalized scale (Supplementary Figure 3) was used. Overall, distinct patterns of alteration (cell counts and/or MFI values per marker in ≥2/10 consecutive stages of maturation of pre-GC B-cells falling ≥2SD apart from the normal distribution) of the pre-GC B-cell maturation were detected in every case (88/88; 100%). Thus, four distinct patterns of alterations were identified: a) CVID with normally-appearing maturation pathways (groups 1 to 3) but AB C FIGURE 2 | Distribution of the major subsets of immature, naive and memory B cells (MBC) and plasmablasts/plasma cells (A), including the different subsets of pre-germinal-center (GC) B-cells (B), in blood of a representative adult healthy donor (HD) and their maturation-associated relationship (C). Three-dimensional principal component (PC) analysis (PCA) plots in (A, B) were built based on PC1 (A: mean fluorescence intensity of CD27, 29.25%; IgD, 22.56%; IgM, 20.45%; CD21, 11.16%; CD38, 9.63% and CD5, 6.95%; B: mean fluorescence intensity of CD21, 30.15%; CD24, 27.64%; CD38, 23.07% and CD5, 19.14%), PC2 (A: mean fluorescence intensity of IgM, 43.89%, CD27, 34.62%, CD38, 10.18%, IgD, 9.96%; CD21, 1.16% and CD5, 0.18%; B: mean fluorescence intensity of CD21, 43.66%, CD38, 23.62%, CD24, 17.09% and CD5, 15.63%) and PC3 (A: mean fluorescence intensity of CD38, 42.02%, CD21, 19.56% CD27, 12.19%, CD5, 12.03%, IgD, 11% and IgM, 3.20%; B: mean fluorescence intensity of CD5, 56.16%, CD24, 27.80%, CD38, 14.85% and CD21, 1.19%) vectors using the (balanced) automated population separator (APS1-2) 3-D view of Infinicyt. In both (A, B), the distinct color-coded cell populations displayed were gated as described in Supplementary Figure 1. del Pino-Molina et al. Dissection of the Pre-GC B-Cell Maturation Pathway in CVID Frontiers in Immunology | www.frontiersin.org February 2021 | Volume 11 | Article 6039726 distinct patterns of alteration on the levels of expression of individual phenotypic markers (57/88, 65%) and b) CVID patients with a clearly altered pre-GC B-cell maturation pathway (31/88 cases, 35%; CVID group 4) (Figures 4A and B). In turn, unsupervised clustering analysis further revealed the presence of three different profiles among the former CVID patient groups 1 to 3, depending on the pattern of deviation in the number of cells per maturation stage and the levels of expression of individual markers, from the normal reference maturation pathway (Figure 5 and Supplementary Figure 3). In group 1 a large fraction of CVID patients (42/88, 48%) was included, who showed significantly reduced total B-cell (128 vs 206 cells/µl, p=0.008) and both CD5 + CD38 +/++ CD21 het CD24 ++ (2.6 vs 5.6 cells/µl, p=0.0013) and CD5 + CD38 het CD21 + CD24 + (7.3 vs 17 cells/µl, p<0.0001) immature/transitional B-cell numbers, in the absence of overall relevant phenotypic deviations from the normal maturation profile of blood preGC B-cells (Supplementary Figure 3). Group 2 consisted of 8/88 (9%) CVID patients that showed overexpression of smIgM at intermediate stages of maturation of pre-GC B-cells and of CD38 at early and late stages of maturation of pre-GC B-cells; compared to HD, CVID cases classified in group 2 also showed significantly higher counts of the most immature CD5 - CD38 ++ CD21 het CD24 ++ (7.6 vs 0.89 cells/µl, p=0.0009) pre-GC B-cell subset, associated with reduced numbers of more differentiated CD5 + CD38 het CD21 + CD24 + immature B lymphocytes (5.1 vs 17 cells/µl, p=0.0183) and IgMD + (particularly CD21 + CD24 + ) MBC (4 vs 31 cells/µl, p<0.0001); in addition, group 2 patients also showed decreased levels of CD21 at stages 7–8, reflecting the parallel increase in CD21 - A B FIGURE 3 | Distribution of distinct subsets of pre-germinal-center (GC) and post-GC B-cells in blood of Common Variable Immunodeficiency (CVID) patients (n=100) vs age-matched HD (n=62). (A) absolute counts of major pre-GC and post-GC B-cell subsets are shown using box and whiskers plots separately for healthy donor (HD) (n=62) and CVID patients (n=100) where horizontal lines and vertical lines represent the median and both 5 th and 95 th percentile values, respectively. Percentage of CVID patients with decreased or increased counts below or above normal values (<5 th and >95 th percentile) detected in age-matched HD are shown as percent values. (B) absolute pre-GC CD5 - CD38 ++ CD21 het CD24 +++ , CD5 + CD38 +/++ CD21 het CD24 ++ , and CD5 + CD38 het CD21 + CD24 + immature B-cell and CD21 + CD24 + , CD21 - CD24 - , and CD21 - CD24 ++ naïve B-cell subset counts in CVID vs HD. N.S. not statistically significant differences detected, *p-value<0.05; ***p-value<0.001; ****p-value<0.0001 (Mann Whitney U test). del Pino-Molina et al. Dissection of the Pre-GC B-Cell Maturation Pathway in CVID Frontiers in Immunology | www.frontiersin.org February 2021 | Volume 11 | Article 6039727 CD24 - naïve B cell counts (8.8 vs 0,74 cells/µl, p=0.02) (Table 1). Finally, group 3 included 7/88 (8%) CVID patients characterized by showing under-expression of CD38 together with higher levels of CD24 at the last stages of maturation of pre-GC Bcells, in line with the underlying increased counts of CD21 + CD24 + mature naïve B cells observed among these cases vs. the other CVID patient groups 1 to 3: 137 vs 76, 42, and 59 cells/µl in CVID groups 1, group 2, and group 4, respectively (Table 1). As referred above, group 4 CVID patients presented clearly altered pre-GC B-cell maturation profiles, the most common alteration (31/31 cases) in their blood pre-GC B-cells consisting of absence/decreased number of cells at the earliest stages of maturation (stages 1-3). In addition, group 4 CVID patients showed abnormally lower counts in blood of CD5 - CD38 ++ CD21 het CD24 ++ -0.36 vs 0.98, 7.6, 1.6, and 0.89 cells/µl in group 1 (p=0.015), group 2 (p<0.0001) group 3 (p=0.003) patients and HD (p=0.004), respectivelyand CD5 + CD38 +/++ CD21 het CD24 ++ -1.4 vs 2.6, 12 and 5.6 cells/µl, group 1 (p=0.03), group 2, (p=0.0117) CVID patients and HD (p<0.0001)-, immature/transitional B cells, together with significantly higher counts of CD21 - CD24 - -9.9 vs 4, 0.74 cells/µl in CVID group 1 A B FIGURE 4 | Illustrating 2-dimensional and 3-dimensional dot plot graphical examples of the maturation pathway of pre-GC B-cells in representative Common Variable Immunodeficiency (CVID) patients showing normal-appearing (A; CVID groups 1–3) vs severely disturbed maturation pathways (B; CVID group 4). In all plots, the reference pre-germinal-center (GC) B-cell maturation pathway defined for 18 (individual) HD green lines is shown together with a gray/black line corresponding to the maturation pathway of the two individual CVID patients shown in (A, B), respectively. Colored dots correspond to the median values obtained for the 10 different maturation stages (color code defined in the right) where stage 1 is colored as blue and stage 10 is depicted in red. del Pino-Molina et al. Dissection of the Pre-GC B-Cell Maturation Pathway in CVID Frontiers in Immunology | www.frontiersin.org February 2021 | Volume 11 | Article 6039728 FIGURE 5 | Maturation diagrams per marker of all Common Variable Immunodeficiency (CVID)patients included in each of the three CVID groups of patients displaying a normal-appearing pre-GC maturation profile in blood (groups 1–3). In each diagram the (normalized) reference maturation pathway is shown as a grey area for the ±2 SD of normal MFI values per marker. The pre-GC B-cell maturation profile found in CVID patients from each group is displayed as ±2 SD bars colorcoded by CVID patient group (CVID group 1 is shown in blue, group 2 in green and group 3 in pink). Values below or above the normal 2SD limit for ≥2 consecutive maturation stages were considered to reflect an altered marker expression profile. del Pino-Molina et al. Dissection of the Pre-GC B-Cell Maturation Pathway in CVID Frontiers in Immunology | www.frontiersin.org February 2021 | Volume 11 | Article 6039729