Klhl6 Deficiency Impairs Transitional B Cell Survival and Differentiation

Klhl6 Deficiency Impairs Transitional B Cell Survival and Differentiation Barbara Bertocci, Damiana Lecoeuche, Delphine Sterlin, Julius Kühn, Baptiste Gaillard, Annie De Smet, Frederique This information is current as Lembo, Christine Bole-Feysot, Nicolas Cagnard, Tatiana of September 28, 2021. Fadeev, Auriel Dahan, Jean-Claude Weill and Claude-Agnès Reynaud J Immunol 2017; 199:2408-2420; Prepublished online 14

August 2017; Downloaded from doi: 10.4049/jimmunol.1700708 http://www.jimmunol.org/content/199/7/2408

Supplementary http://www.jimmunol.org/content/suppl/2017/08/11/jimmunol.170070 http://www.jimmunol.org/ Material 8.DCSupplemental References This article cites 60 articles, 28 of which you can access for free at: http://www.jimmunol.org/content/199/7/2408.full#ref-list-1

Why The JI? Submit online. by guest on September 28, 2021 • Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2017 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Klhl6 Deﬁciency Impairs Transitional B Cell Survival and Differentiation

Barbara Bertocci,* Damiana Lecoeuche,* Delphine Sterlin,*,1 Julius Kuhn,€ † Baptiste Gaillard,*,2 Annie De Smet,* Frederique Lembo,‡ Christine Bole-Feysot,x Nicolas Cagnard,{ Tatiana Fadeev,* Auriel Dahan,* Jean-Claude Weill,* and Claude-Agne`s Reynaud*

Klhl6 belongs to the KLHL gene family, which is composed of an N-terminal BTB-POZ domain and four to six Kelch motifs in tandem. Several of these proteins function as adaptors of the Cullin3 E3 ubiquitin ligase complex. In this article, we report that Klhl6 deficiency induces, as previously described, a 2-fold reduction in mature B cells. However, we find that this deficit is centered on the

inability of transitional type 1 B cells to survive and to progress toward the transitional type 2 B cell stage, whereas cells that have Downloaded from passed this step generate normal germinal centers (GCs) upon a T-dependent immune challenge. Klhl6-deﬁcient type 1 B cells showed a 2-fold overexpression of genes linked with cell proliferation, including most targets of the anaphase-promoting complex/cyclosome complex, a set of genes whose expression is precisely downmodulated upon culture of splenic transitional B cells in the presence of BAFF. These results thus suggest a delay in the differentiation process of Klhl6-deﬁcient B cells between the immature and transitional stage. We further show, in the BL2 Burkitt’s lymphoma cell line, that KLHL6 interacts with Cullin3,

but also that it binds to HBXIP/Lamtor5, a protein involved in cell-cycle regulation and cytokinesis. Finally, we report that http://www.jimmunol.org/ KLHL6, which is recurrently mutated in B cell lymphomas, is an off-target of the normal somatic hypermutation process taking place in GC B cells in both mice and humans, thus leaving open whether, despite the lack of impact of Klhl6 deﬁciency on GC B cell expansion, mutants could contribute to the oncogenic process. The Journal of Immunology, 2017, 199: 2408–2420.

cell generation follows a stepwise process of differenti- guish T1 (CD93+CD232) from T2 B cells (CD93+CD23+) (6–8). It ation, ﬁrst in the bone marrow and later in the periphery, has been observed that immature B cells can also differentiate into B where fully mature B cells reside. After a productive T1-like and T2-like B cells within the bone marrow (9, 10). T2 rearrangement and expression of a functional BCR at the surface, B cells, which emerge in spleen with a parallel but slightly

B cells are still at an immature stage. Cross-linking of the BCR at delayed timing, could thus originate from both in situ T1 cell by guest on September 28, 2021 this stage induces cell death, a negative selection process purging differentiation and migration of T2-like cells from bone marrow the B cell repertoire from autoreactive specificities (1, 2). (11, 12). Interaction between the B cell–activating cytokine BAFF Immature B cells that survive past this checkpoint migrate to the and its specific receptor (BAFF-R) on B cells appears to play spleen and mature into follicular (FO) and marginal zone (MZ) different roles in this process, depending on the B cell differen- B cells (3). This maturation occurs in successive steps and implies tiation stage and the lymphoid organ concerned (13–19). Whereas the formation of two subsets of transitional B cells, transitional the immature to T1-like B cell progression in bone marrow seems type 1 (T1) and transitional type 2 (T2) (4), which have been to occur in a BAFF-independent mode, this is not the case in the proposed to be located in different areas of the spleen, the peri- spleen, where T1 cells require BAFF signals for differentiation arteriolar lymphoid sheath and the B cell follicles, respectively and/or survival, as shown in mixed bone marrow chimeras with (5). Several surface markers have been used to characterize T1 and BAFF-R–deficient and BAFF-R–competent cells (20). Conversely, T2 B cells, notably expression of CD93 and CD23, which distin- the differentiation of both splenic and bone marrow T1 into T2

*E´ quipe De´veloppement du Syste´me Immunitaire, Institut Necker-Enfant Malades, This work was supported by the Fondation Princesse Grace (to Team De´veloppement INSERM U1151-CNRS UMR8253, Faculte´ de Me´decine Paris Decartes, Universite´ du Syste`me Immunitaire) and the Ligue contre le Cancer (E´ quipe Labelliseé). Paris Descartes, Sorbone Paris Cite´, 75993 Paris Cedex 14, France; †Institute of The microarray data presented in this article have been submitted to the ArrayEx- Cellular and Molecular Immunology, Georg-August-University Medicine Go¨ttingen, press database (https://www.ebi.ac.uk/arrayexpress/browse.html) under accession 37073 Go¨ttingen, Germany; ‡Centre de Recherche en Cance´rologie de Marseille, x number E-MTAB-5928. INSERM U1068-CNRS UMR7258, 13273 Marseille Cedex 09, France; Plateforme de Geńomique, Imagine Institut des Maladies Geńe´tiques-Structure Fe´de´rative de Address correspondence and reprint requests to Dr. Barbara Bertocci, E´ quipe De´vel- Recherche Necker, INSERM 1163 and INSERM US24/CNRS UMS3633, 75015 oppement du Syste´me Immunitaire, Institut Necker-Enfant Malades, INSERM Paris, France; and {Plateforme de Bioinformatique, Universite´ Paris Descartes- U1151-CNRS UMR8253, Faculte´ de Me´decine Paris Decartes, Universite´ Paris Des- Structure Fe´de´rative de Recherche Necker, INSERM US24/CNRS UMS3633, cartes, Sorbone Paris Cite´, 14 rue Maria Helena Vieira Da Silva, Paris Cedex 14, 75993 Paris Cedex 14, France 75993 France. E-mail address: [email protected] 1Current address: Sorbonne Universite´s, Universite´ Pierre et Marie Curie/Universite´ The online version of this article contains supplemental material. Paris 6, INSERM, CNRS, Centre d’Immunologie et des Maladies Infectieuses Paris Abbreviations used in this article: AID, activation-induced cytidine deaminase; APC/ UMRS 1135, Assistance Publique – Hoˆpitaux de Paris, Groupement Hospitalier C, anaphase-promoting complex/cyclosome; CLL, chronic lymphocytic leukemia; Pitie´-Salpeˆtrie`re, De´partement d’Immunologie, Paris, France. Cul3, Cullin 3; FO, follicular; GC, germinal center; GSEA, gene set enrichment 2Current address: Service d’He´matologie, Hopital Robert Debre, Centre Hospitalier analysis; KO, knockout; MZ, marginal zone; PC, plasma cell; PLA, proximity liga- Universitaire de Reims, Reims, France. tion assay; T1, transitional type 1; T2, transitional type 2; wt, wild type. ORCIDs: 0000-0002-5532-9431 (T.F.); 0000-0002-4519-4473 (C.-A.R.). Copyright Ó 2017 by The American Association of Immunologists, Inc. 0022-1767/17/$35.00 Received for publication May 12, 2017. Accepted for publication July 17, 2017. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1700708 The Journal of Immunology 2409

B cells is strongly dependent on BAFF signaling, because B cell Flow cytometry maturation in BAFF- and BAFF-R–deficient mice does not pro- Tissue samples were collected from the appropriate mice as described ceed beyond the T1 stage (13, 15, 19). in the text. After RBC lysis, cells were stained with a combination of The KLHL gene family encodes 42 different proteins, defined fluorophore-conjugated Abs (Supplemental Table I). Surface markers as being composed of an N-terminal BTB-POZ domain followed were detected and analyzed using a FACSCanto II flow cytometry ap- by a BTB and C-terminal kelch (BACK) domain and four to six paratus and the BD FACSDiva software (BD Biosciences). tandem Kelch motifs (21). The BTB domain, named by its ho- Cell culture mology with the Drosophila Bric-a-brac, Tramtrac, and Broad complex, is involved in protein–protein interaction and protein Splenic B cells, after purification (B cell isolation kit, MACS separation), were cultured in complete RPMI 1640 medium supplemented with 10% self-oligomerization (22). The BACK domain is very conserved, HyClone FetalClone I serum (Thermo Scientific), 100 U/ml penicillin and but its role is not unequivocally identified (23). The Kelch motif, 100 U/ml streptomycin, 1 mM sodium pyruvate, MEM nonessential which was discovered as a six-repeat element in the Drosophila amino acids (13), 25 mM HEPES (Invitrogen), 0.5 mM 2-ME (Sigma), kelch protein, is a segment of 44–56 aa that adopts a b-propeller and the indicated activating factors. All reagents, including anti-IgM (Jackson ImmunoResearch), LPS, CpG (Sigma), and recombinant BAFF structure and contains multiple potential protein contact sites protein (RD Biosystem), were prepared as per manufacturer’s instructions (24). Despite their shared secondary structure, the primary se- and used at the concentration indicated in Supplemental Fig. 2A. To quence of KLHL family members has little homology, suggesting monitor lymphocyte proliferation, we labeled cells with CFSE (2.5 mM; a large diversity of interacting partners and consequently multiple Miltenyi Biotec). Dye dilution was analyzed at the indicated time by flow biological functions. Although data on the molecular function of cytometry after dead cell exclusion. KLHL proteins are still fragmentary, several of them have been Immunoblotting Downloaded from identified as adaptors of Cullin 3 (Cul3)-based E3 ubiquitin ligases (25). These KLHL proteins integrate the function of ubiquitin li- Cells were washed in PBS and resuspended in Bio-Rad sample buffer or in lysis buffer (50 mM Tris-HCl [pH 8], 0.1 mM EDTA, 200 mM NaCl, 10% gase adaptor and substrate recognition by forming a complex with glycerol, Nonidet P-40 0.5%, 0.4 mM PMSF, 3 mg/ml aprotinin, 1 mg/ml Cul3 through the BTB domain and binding the substrate through leupeptin, 0.5 mg/ml pepstatin). Whole cell extracts were fractionated by the Kelch modules, mediating through ubiquitylation the degra- SDS-PAGE and transferred to a nitrocellulose membrane according to the manufacturer’s protocols (Bio-Rad). After incubation with 5% nonfat milk dation (26, 27) or relocalization of their substrate within the cell http://www.jimmunol.org/ in TBST (10 mM Tris-HCl [pH 8], 150 mM NaCl, 0.5% Tween 20) for (28). Whereas Cul3 is a ubiquitous protein, several KLHL genes 60 min, the membrane was washed once with TBST and incubated with show marked tissue-restricted expression, allowing their differ- Abs against NF-kB2 p100/p52 (1:1000), Survivin (1:1000), HBXIP ential specificity (29, 30). (1:2000), GAPDH (1:2000), or b-actin (1:10,000) at 4˚C for 12 h. Mem- We previously reported the isolation of the Klhl6 gene through branes were washed three times for 10 min and incubated with a 1:10,000 its overexpression in B cells from sheep ileal Peyer’s patches, an dilution of HRP-conjugated anti-rabbit Abs for 2 h. Blots were washed with TBST three times and visualized with the ECL system according to overexpression that is also observed in mouse and human ger- the provided protocol (Bio-Rad). Reagents are listed in Supplemental minal center (GC) B cells (31). More recently, recurrent muta- Table I. Signals were detected with a charge-coupled device camera tions in KLHL6 have been described in B cell lymphomas of GC (Fujilas 1000 Plus) and quantified with Multi Gauge software. origin (32), as well as in chronic lymphocytic leukemias (CLLs) by guest on September 28, 2021 NF-kB DNA-binding ELISA assay with mutated Ig V genes (33, 34). The group of T. Sato (35) generated Klhl6 knockout (KO) mice and reported a 2-fold re- The DNA-binding capacity of p52 was evaluated by TransAM NF-kB duction of mature B cells. The B cell deficit was linked to an Family Transcription Factor Assay kit (Active Motif), following manufacturer’s instructions. Nuclear cell extracts were prepared using the Nu- impaired proliferative response and signal transduction after clear Extract kit (Active Motif). BCR cross-linking, and a marked reduction in GC B cell formation was observed in T-dependent immune responses (35). We Gene expression profiling report in this article an independent inactivation of the Klhl6 gene. Transitional B cells from spleen of 17-d-old mice were sorted on a FAC- Although we observed the same reduction in mature B cells, we SAria II (Becton Dickinson) after staining with fluorescent Abs for CD19 identified the origin of this B cell deficit as being essentially due and CD93. Total RNA was harvested from these cells using the RNeasy to an impaired survival and maturation of transitional T1 cells RNA preparation Kit (Qiagen). RNA quality and concentration were assessed using RNA 6000 Pico LabChips with a 2100 Bioanalyzer (Agilent during B cell development. Klhl6-deficient B cells that matured Technologies). In brief, 100 ng of total RNA was reverse transcribed, and beyond this stage behave similarly to normal B cells, and ac- second-strand DNA was produced and amplified by in vitro transcription in cordingly, no impact of Klhl6 deficiency was observed on GC the presence of biotinylated ribonucleotides using the IVT Express kit responses. (Affymetrix). Gene expression analysis was performed using GeneChip Mouse Genome 430 2.0 arrays (Affymetrix), as recommended by the manufacturer. Fluorescence data were imported into two analysis software Materials and Methods packages: Affymetrix Expression Console and R Bioconductor. Gene ex- Mice pression levels were normalized using the GC-RMA algorithm, and flags were computed using MAS5. The group comparisons were done using The Comite´ National de Re´flexion Ethique sur l’Expe´rimentation Animale Student t test. To estimate the false discovery rate, we filtered the resulting and the ethics committee of French Ministry of Research and Higher p values at 5%. Enriched functional annotations for genes differentially 2 2 Education approved all animal experiments (project 014033.03). Klhl6 / expressed were analyzed with the computational method GSEA (gene set mice were generated from E14.1 embryonic stem cells and backcrossed to enrichment analysis). Microarray data are available at the ArrayExpress 2 2 C57BL/6 background for .10 generations. Klhl6 / 3 activation-induced website, accession number E-MTAB 5928. cytidine deaminase (AID)-Cre-EYFP mice were generated by back- 2 2 crossing Klh6 / and AID-Cre-EYFP knock-in mice (AID-Cre-ERT2 Inducible expression of a tagged KLHL6 protein in a Burkitt’s crossed with ROSA26-loxP-EYFP reporter mice) (36). lymphoma Chimeric mice were created by bone marrow transplantation. Bone marrow was collected from tibial and femoral bone marrow cavities of 15-d- The BTZ cell line, which is a BL2 clone harboring pTet-tTAk-zeo (37), was old C57BL/6 Ly5.1 (CD45.1) control and C57BL/6 Ly5.2 (CD45.2) Klhl62/2 cultured in RPMI 1640 supplemented with 10% HyClone FetalClone I mice. RBCs were lysed using RBC lysis buffer. Nucleated cells (107 cells/ serum, 100 U/ml penicillin, and 100 U/ml streptomycin. These cells were mouse) were injected i.v. via marginal sinus to 8-wk-old RAG-2–deficient transfected using the Amaxa technique with a pBI vector in which mice that were previously irradiated (3 Gy, 137Cs irradiator). Mice were KLHL6-EGFP fusion cDNA was cloned. Positive clones were selected and analyzed 2–3 mo after reconstitution. cultured with blasticidin (5 mg/ml; Invitrogen) and tetracycline (1 mg/ml; 2410 Klhl6 AND T1 B CELL MATURATION

Sigma). Removing tetracycline from the medium induced the expression of tion among the different B cell subsets was comparable in KO and KLHL6 fusion protein. control mice (Fig. 1B). Reduction of mature B cells was also ob- In situ proximity ligation assay served in Peyer’s patches, inguinal lymph nodes, and mesenteric

5 lymph nodes (Fig. 1C, data not shown). Induced and not induced KLHL6-expressing BL2 cells (10 ) were left We evaluated in vitro the proliferation capacity of Klhl6- attached for 1 h at 4˚C on a microscope slide (SuperFrost Ultra Plus; Thermo Scientific) and then fixed with 4% (w/v) paraformaldehyde deficient splenic FO B cells, after CFSE labeling and BCR (anti- (Sigma) in PBS for 15 min at room temperature. PLA was performed IgM) or TLR stimulation (LPS, CpG). Klhl6 deficiency did not with Duolink II Red Starter kit (Supplemental Table I) according to the alter B cell division profiles including the number of divisions manufacturer’s instructions (Sigma). The primary Abs, anti-HBXIP (1:100) performed and the cell viability (Supplemental Fig. 2A). The survival and anti-EGFP (1:500) (Supplemental Table I), were incubated at room 2/2 temperature for 2 h in a humidity chamber. Slides were mounted using capacity of Klhl6 mature B cells was also similar to controls Vectashield mounting medium with DAPI (Vector Laboratories). The during culture in the presence of BAFF, a major B cell survival samples were imaged through a Plan Apochromat 63/1.4 numeric aperture factor (Supplemental Fig. 2B). oil-immersion lens on a Zeiss LSM 710 confocal microscope (Zeiss), and The B cell deficit observed is therefore similar to the one de- Z-stacks were acquired using Zen 2009 software (Zeiss). High-resolution scribed by Kroll and coworkers (35), but our results did not reveal images were obtained with the Imaris software. The mean number of signals per cell was obtained by dividing the total number of PLA signals by the a proliferation or survival defect consecutive to BCR or TLR sig- number of nuclei (counted manually). naling in mature B cells. These data altogether suggest that Klhl6 is required for optimal differentiation, rather than for maintenance Yeast two-hybrid screen and phenotypic assay or activation of the mature B cell compartment.

The Y2H system was performed as previously described (38). The bait, Downloaded from human KLHL6 cDNA, was cloned into pDBa (Leu) using the Gateway Klhl6 is dispensable for GC formation technology (Invitrogen). The bait plasmid and the human spleen cDNA library [cloned into pEXP502-AD (Trp) ProQuest libraries; Invitrogen] Klhl6 expression is markedly increased in GC B cells, both in mice were transformed in MAV03 yeast strain. The cDNA of 14 selected clones and in humans. We therefore investigated the impact of Klhl6 wasamplifiedbyPCR(oligoYL159-CGCGTTTGGAATCACTACAGGG- deficiency on GC formation in chronically stimulated B cells from 39,YL259-GGAGACTTGACCAAACCTCTGGCG-39). DNA was sequenced Peyer’s patches and in spleen after a secondary antigenic chal- using the ABI Prism 3130xl Genetic Analyzer. lenge with SRBCs. http://www.jimmunol.org/ Klhl6 sequencing No significant difference was observed in the proportion of Mouse GC B cells (B220+ GL7+ PNA+) were sorted from Peyer’s patches Peyer’s patch GC B cells in Klhl6-deficient and age-matched lit- of two 7- to 9-mo-old Ung2/2/Msh22/2 mice and two littermate controls. termates (Fig. 1C). To study the SRBC recall response, we crossed Human memory B cells (IgD2CD27+) were sorted from blood of two Klhl6 null mice with AID-Cre-ERT2 3 ROSA26-EYFP reporter individuals. Amplification of a region flanking exon 1 was achieved with mice (AID-Cre-EYFP mice), in which B cells engaged in an im- Phusion polymerase from genomic DNA with the following oligonucleotides: mune response and expressing AID can irreversibly acquire EYFP for the mouse gene, MoKlhl6F 59-CCTGATCTACGCACCCACTC-39, MoKlhl6R 59-GTTAGCTTGACTCTAAGAGATG-39;forthehuman expression upon simultaneous tamoxifen feeding (36). The pro- + gene, HuKLHL6F 59-CCCCTCCCACACACCTATTC-39, HuKLHL6R portion of EYFP GC B cells was analyzed 5 d after secondary 59-GTCTCCTGACTTCAGGCCTTC-39. Sequence of the PCR products immunization with SRBC in AID-Cre-EYFP/Klhl62/2 mice and by guest on September 28, 2021 was performed after Zero-blunt cloning (Invitrogen) using the BigDye heterozygous littermates, with tamoxifen feeding at days 7 and Terminator sequencing kit and analyzed on an ABI PRISM 3130x Ge- netic Analyzer. 12 of the primary challenge and day 1 of the boost (Fig. 1D). The percentage of GL7+EYFP+ B cells was equivalent in Klhl62/2 and Statistical analysis control mice, as well as the IgG1 and IgM isotype distribution 2/2 All statistical analyses were done with GraphPad Prism software using Welch (Fig. 1D). Plasma cell (PC) formation was also similar in Klhl6 t test. Data were reported as mean 6 SD. Statistically significant differences mice and controls during this recall response (Fig. 1D). are indicated on the figures: *p , 0.05, **p , 0.01, ***p , 0.001. These data indicate that Klhl6 is dispensable for GC formation and maintenance, and therefore that Klhl6-deficient B cells do not Results show impaired activation, neither in vitro nor in vivo. B cell differentiation is impaired in Klhl6-deficient mice Klhl6 was inactivated in embryonic stem cells by replacement of Klhl6 is required for splenic transitional T1 B cell homeostasis exon 1 with a NeoR cassette, and mutant mice were generated To identify the differentiation stage affected by the absence of (Supplemental Fig. 1A). Homozygous KO animals were viable, Klhl6, the bone marrow immature B cell subset was further resolved fertile, and were obtained with normal Mendelian segregation. in two subpopulations (9, 10), newly formed/T1 (FrE, CD232)and The B cell compartment was analyzed in Klhl62/2 and age- late differentiated T2-like (FrE, CD23+)Bcells.Klhl62/2 mice matched littermates. Early stages of bone marrow B cell develop- displayed a 2-fold reduced frequency for the more mature CD23+ ment (pro-B and pre-B cells, i.e., from fraction A to fraction D, B cell population, whereas the CD232 subset was slightly increased according to Hardy’s nomenclature) were not affected (Supplemental (Fig. 2A, Table I). Fig. 1B). However, the frequency of immature (fraction E, B220+ These observations were confirmed for the bone marrow im- CD93+IgM+) and mature recirculating (fraction F, B220+CD932 mature B cell compartment of young mice (14–17 d old), a stage at IgM+) B cells differed between Klhl62/2 and control mice (Fig. 1A), which recirculation of mature B cells is minimal. Like in adults, a with an increase in the absolute numbers of immature B cells and a 2-fold reduction in the proportion and absolute numbers of late 2-fold decrease in recirculating B cells in mutant mice (Table I). differentiated T2-like cells was observed in Klhl62/2 mice com- The absolute number of splenocytes in adult mice was reduced pared with controls, together with a significant increase in the by ∼30% in Klhl62/2 compared with wild type (wt) mice, a newly formed/T1 subpopulation (Fig. 2A, Table I). These results difference entirely accounted for by a marked reduction in fre- suggest an impaired B cell survival and/or differentiation at the quency and number of the mature B cell compartment (Fig. 1B), late differentiated T2-like stage of B cell maturation in bone whereas splenic T cells appeared unaffected (data not shown). marrow, already present in newborn mice. Total numbers of splenic T1, T2, FO mature, and MZ B cells were As mentioned earlier, splenic T1 and T2 B cells showed a 2-fold all decreased ∼2-fold by Klhl6 deficiency, whereas the distribu- reduction in adult Klhl6-deficient mice compared with littermate The Journal of Immunology 2411

FIGURE 1. Klhl6 deﬁciency impacts the mature B cell pool, but not the GC B cell response. (A)Analysisof bone marrow immature B cells (FrE) (B220+IgM+CD93+) and mature recirculating B cells (B220+IgM+CD932) from 8- to 14-wk-old Klhl6-deﬁcient mice (red squares) and age-matched littermate controls (blue circles). (B) Analysis of total splenocytes (upper row) and splenic B cell subsets: T1 B cells (CD19+CD93+CD232), T2 B cells + + + + 2 + (CD19 CD93 CD23 ), FO B cells (CD19 CD93 CD23 Downloaded from CD21int), and MZ B cells (CD19+CD932CD232CD21high). (C) Analysis of Peyer’s patch B cells (B220+) and GC cells (B220+GL7+). (D) Analysis of the anti-SRBC response of AID-Cre-EYFPxKlhl62/2 mice and Klhl6 heterozygous littermates, after immunization with 5 3 109 SRBC and tamoxifen administration, performed according to the experimental protocol depicted in the upper http://www.jimmunol.org/ left panel. The lower left panels represent EYFP+ splenic B cell subsets analyzed 5 d after the boost. From left to right, EYFP+ cells, further gated into B2202 PCs and B220+GL7+ GC B cells, and EYFP+ GC B cells analyzed for IgM and IgG1 isotypes. Upper and lower middle panels display the percentage of GC B cells among EYFP+ cells and the distribution of IgM and IgG1 isotypes among EYFP+ GC B cells; upper right panel shows +

percentage of PCs among EYFP cells. Each symbol by guest on September 28, 2021 represents an individual mouse. Data (mean 6 SD) are from two independent experiments. Statistical analysis was performed with unpaired, two-tailed t test with Welch correction. *p , 0.05, **p , 0.01, ***p , 0.001.

controls, similar to the deﬁcit observed for the mature FO and MZ of control values (Fig. 2B). This observation indicates that, subsets. In newborn mice (14–17 d old), in which the splenic B cell whereas Klhl6 deﬁciency only affects B cell differentiation at the pool is mostly composed of transitional B cells, T1 and T2 cells T2-like stage in bone marrow, it appears to be required for normal were reduced to a similar extent in Klhl62/2 animals, to 50–60% homeostasis of T1 cells in spleen.

Table I. Total number of cells 3106 among B220+ cells in bone marrow (hind legs)

IgM2 FrE FrE CD232 FrE CD23+ FrF Adult mice Control (n = 4) 3.09 6 1.10 1.46 6 0.43 1.08 6 0.35 0.379 6 0.11 2.23 6 0.96 Klhl62/2 (n = 4) 3.37 6 0.69 1.98 6 0.71 1.77 6 0.68 0.198 6 0.04 1.04 6 0.23 Young mice Control (n = 4) 4.38 6 2.19 3.18 6 1.49 2.02 6 1.01 0.912 6 0.38 Klhl62/2 (n = 4) 4.91 6 1.56 3.53 6 1.06 2.98 6 0.79 0.485 6 0.27 2412 Klhl6 AND T1 B CELL MATURATION Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 2. Klhl6 deficiency impairs B cell maturation and survival in bone marrow and spleen. (A) Analysis of bone marrow B cell subpopulations from adult or young Klhl6-deficient mice (14–17 d old) and age-matched littermate controls. The surface marker CD23 was used to discriminate newly formed, T1-like B cells (B220+IgM+CD93+CD232) from T2-like B cells (B220+IgM+CD93+CD23+). (B) Analysis of splenocytes from young Klhl6-deficient and age- matched control mice (14–20 d old), with percentage and absolute numbers of T1 and T2 B cell subsets. Results (mean 6 SD) represent two (A) and three (B) independent experiments. Each symbol (red squares for Klh62/2 and blue circles for controls) represents an individual mouse (n = 5–10 mice). Statistical analysis by unpaired, two-tailed t test with Welch correction. **p , 0.01, ***p , 0.001. (C) Competitive reconstitution of Rag2-deficient mice with bone marrow cells from 12- to 14-d-old Ly5.2 Klhl62/2 and Ly5.1 control mice at 1:1 ratio. The gating strategy for discriminating B cell subsets is described in Fig. 1 for spleen and blood and in (A) for bone marrow. Density plots depict the B cell subpopulations and contour plots the frequency of Klhl62/2 (red) and control cells (blue). The numbers adjacent to the outlined areas represent the percentage of cells within the gate. The bar graphs represent the ratio between Ly5.1 wt and Ly5.2 Klhl62/2 cells (mean 6 SD, n = 3 mouse chimeras). Data shown correspond to one representative experiment out of two. The Journal of Immunology 2413

To assess the behavior of Klhl6-deficient B cells in a competitive These data thus provide evidence for a role of Klhl6 in BAFF- setting, we generated mixed bone marrow chimeras by reconstitution mediated differentiation and survival of splenic transitional T1 of irradiated Rag22/2 mice with equal number of allelically marked cells, but did not reveal a functional defect in the activation of the wt (CD45.1) and Klhl6-deficient (CD45.2) bone marrow cells alternate NF-kB pathway. (Fig. 2C). In bone marrow of restored mice, this competition revealed a selective advantage of Klhl62/2 cells (a 2-fold increase) Gene expression analysis of Klhl6-deficient transitional B cells in the immature B cell subset of newly formed CD232 cells, reveals a delayed maturation profile whereas the wt/Klhl62/2 cell ratio appeared reversed at the next To understand the molecular processes underlying the impaired step of cell differentiation, in CD23+ T2-like B cells (Fig. 2C). BAFF-induced survival observed in Klhl6-deficient mice, we per- Interestingly, the proportion of control and mutant cells was un- formed gene expression profiling of splenic transitional B cells changed in immature CD232 and CD23+ B cell fractions between isolated from 17-d-old Klhl62/2 and littermate controls, ex vivo (T0) bone marrow and blood (Fig. 2C), indicating that egress from bone and after 24 h of in vitro culture in the presence of BAFF (T24). marrow is not affected by the absence of Klhl6. In spleen, Klhl6- At T0, 460 probe sets showed a 1.5 difference between KO and deficient B cells were increasingly disfavored along further B cell control cells, with 258 underexpressed genes in Klhl6-deficient cells maturation steps compared with wt cells; this difference is already and 202 overexpressed (Fig. 4A). Among underexpressed genes are starting at the T1 differentiation stage, with stronger imbalance genes related with immune cell differentiation, like CD40, CTIIA, observed for naive and MZ B cells (Fig. 2C). FCGR2B, CD274 (PDL1), and CD22; with cell migration, like These results confirm the role of Klhl6 in normal T1 homeostasis CCR6 and CXCR5; or with cytokine signaling, like LTA, EBI3, and reveal an impact on MZ B cells, which is not observed outside IL27RA,IRF7,andIFIT1-3(Fig.4A).Amongoverexpressed Downloaded from such competition context. genes, GSEA highlighted genes linked with cell proliferation, with a remarkable set of genes comprising most substrates of the Klhl6 is required for BAFF-induced survival of transitional B cells anaphase-promoting complex/cyclosome (APC/C), showing a dif- BAFF is a key cytokine involved in the survival and maturation of T1 ference in the 2-fold range (Fig. 4A, 4C) (42). This large ubiquitin cells (39). To explore further the functional role of Klhl6 in T1 ligase complex has an essential role in the control of the cell cycle,

B cell maintenance, we purified and cultured transitional B cells from through ubiquitylation of various protein substrates, marking them http://www.jimmunol.org/ spleen of 14- to 17-d-old mice in vitro in the presence of BAFF. for degradation at specific cell-cycle checkpoints (43). B cell survival was followed over 3 d. Klhl6-deficient B cells dis- Very strikingly, incubation in the presence of BAFF induced the played a 40% reduction in viability compared with B cells isolated downregulation of a similar set of proliferative genes (Fig. 4B, 4C), from spleen of littermate controls, whereas no difference in survival which were reduced ∼2-fold in both KO and control transitional rate was observed in unstimulated cells (Fig. 3A). B cells. This is illustrated by the reverse shape of the GSEA cell- The differentiation capacity of transitional T1 B cells was further cycle diagram of the KO/control comparison relative to the one of evaluated after purification of CD93+CD232 B cells from newborn the BAFF stimulation (compare Fig. 4D, 4E). This indicated that spleens and culture in the presence of BAFF, and differentiation into Klhl6 deficiency resulted, in transitional B cells, in a higher ex- T2 B cells was assessed by CD23 surface marker acquisition pression of genes involved in proliferation that are precise targets for by guest on September 28, 2021 (Fig. 3B). The frequency of Klhl6-deficient T2 cells was reduced BAFF-induced downmodulation. Among genes induced upon cul- during the in vitro culture compared with controls, and the impaired ture in the presence of BAFF are genes linked to cholesterol, lipid, survival of T1 transitional B cells contributed by Klhl6 deficiency and lipoprotein metabolism, involved notably in the remodeling of resulted in a 5-fold decrease in absolute T2 cell numbers (Fig. 3B). membranes through protein–lipid linkage (farnesyl transferase and Expression of the BAFF-R increases as B cells mature, and we farnesyl synthase, squalene epoxidase) (44). Klhl62/2 and control therefore analyzed BAFF-R expression by flow cytometry on bone cells presented a similar GSEA profile for both overexpressed and marrow and spleen B cell subsets in young Klhl6-deficient mice underexpressed genes during in vitro culture with BAFF, with a large and age-matched littermates, as a marker of their differentiation number of shared genes evidenced in the Venn diagrams (Fig. 4B), status. BAFF-R expression, which is first measurable on a fraction of suggesting that Klhl6 deficiency did not induce a general impair- bone marrow T1-like CD232 B cells (40), was barely detectable in ment of the response to BAFF signaling. Splenic mature B cells Klhl6-deficient mice compared with littermate controls. An increase showed minimal differences in gene expression between Klhl62/2 in BAFF-R expression was observed along B cell maturation in and controls, including upon in vitro BAFF stimulation (data not bone marrow, with lower levels observed for Klhl6-deficient T2- shown), in agreement with the normal activation profile observed for like cells compared with controls. Similar reduction of BAFF-R Klhl6-deficient cells after the transitional B cell stage. expression was observed in T1 mutant spleen cells, a difference The differential expression of cell-cycle genes was confirmed at that was still present but less pronounced at the T2 stage (Fig. 3C). the protein level for survivin (encoded by the Birc5 gene), a protein In contrast, no significant difference in BAFF-R expression was involved in cell division and antiapoptotic processes (45). Survivin observed at the mature B cell stage, in agreement with the normal expression was increased ∼2-fold in transitional Klhl6-deficient phenotype observed for Klhl6-deficient cells that survived along B cells compared with controls, a differential regulation of ex- this differentiation pathway (Supplemental Fig. 2B, right panel). pression that was restricted to T1 Klhl62/2 B cells, whereas T2 Signaling through the BAFF-R activates the noncanonical NF-kB cells did not display such change (Fig. 4F). pathway (41). Activation of this pathway was therefore evaluated Interestingly, downregulation of this set of proliferative genes is on purified transitional B cells from young Klhl62/2 and control similarly observed during the transition from bone marrow immature mice ex vivo and after 24 h BAFF stimulation by quantification of to splenic T1 B cells (ImmGen database), further documenting that the activated transcription factor p52 derived from p100 processing. progressive transcriptional extinction of such proliferation markers, Western blot analysis on total protein extracts did not show a sig- which are otherwise strictly controlled through ubiquitination/ nificant difference in p52 amounts between Klhl6-deficient and phosphorylation at the protein level, parallels maturation of cells control transitional B cells (Fig. 3D). A second approach, based on that are already nondividing (Supplemental Fig. 3). ELISA assays of nuclear extracts, designed to detect p52 binding to These results indicate that BAFF-induced maturation of tran- its DNA target, gave similar results (Fig. 3D). sitional B cells involves notably the coordinated shutdown of 2414 Klhl6 AND T1 B CELL MATURATION Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 3. Impaired BAFF-induced differentiation of Klhl6-deficient transitional B cells. (A) Survival of MACS-purified transitional B cells from young Klhl62/2 mice and controls after BAFF stimulation at the concentration of 100 ng/ml (upper panel) and unstimulated (lower panel). Data are from three to nine experiments in which cells were pooled from two to four mice. Statistical analysis was performed with multiple t tests according to the Holm–Sidak method. (B) In vitro differentiation of T1 transitional B cells from young Klhl62/2 and control mice. CD19+CD93+CD232 cells were sorted and cultured in the presence of BAFF (100 ng/ml). Differentiation and viability were evaluated by staining with CD19, CD93, CD23, and SYTOX Blue at the indicated time points. Results are displayed as density plots and are representative of two independent experiments. The number of T1 cells is reported on the right upper panel and T2 cells on the right lower panel. (C) BAFF receptor expression of bone marrow newly formed and splenic transitional B cells from Klhl62/2 and control mice. Quantification of BAFF-R mean fluorescence intensity (MFI) from bone marrow and spleen subpopulations are shown, respectively, in the upper and lower panels (mean 6 SD, n = 3–5). The p values with unpaired, two-tailed t test with Welch correction. (D) Activation of the noncanonical NF-kB pathway was monitored from purified Klhl62/2 and control transitional B cells. Cells were cultured for 22 h in the presence or absence of BAFF (150 ng/ml). The p52 transcription factor derived from the processing of p100 protein was detected by Western blot of total cell extracts (left panel). The middle and right panels show, respectively, the detection of p52 from transitional B cells ex vivo and from transitional B cells after 24-h activation in the presence of BAFF, using DNA-binding ELISA assay. The results represent two to four independent experiments in which purified B cells from two to three mice were pooled. **p , 0.01, ***p , 0.001. The Journal of Immunology 2415 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 4. Gene expression profile reveals a maturation delay of Klhl62/2 transitional B cells. (A) GSEA comparison of upregulated and downregulated genes in Klhl62/2 and control transitional B cells identified by microarray analysis. (B) GSEA comparison of upregulated and downregulated genes in Klhl62/2 (left panel) and control (middle panel) transitional B cells after 24-h BAFF stimulation. Venn diagrams (right panel) show the number of downregulated or upregulated genes shared by Klhl6-deficient and control mice during BAFF activation. (C) Heat map representation (normalized log2 expression) of selected genes representing APC/C substrates or cofactors, as well as genes involved in proliferation control, whose expression differed by .1.5-fold between Klhl62/2 and control cells. (D) GSEA plots showing enrichment of upregulated cell-cycle genes in Klhl62/2 transitional B cells compared with controls. The normalized enrichment score (NES), nominal p value, and q (false discovery rate) are given for each plot. (E) Similar GSEA plots for both control (middle panel) and Klhl62/2 cells (right panel) after BAFF stimulation, showing an inverse profile compared with the one shown in (D) for the same gene set. (F) Increased expression of survivin (encoded by Birc5)inKlhl62/2 T1 B cells. Upper left panel (Figure legend continues) 2416 Klhl6 AND T1 B CELL MATURATION proliferative genes, including most APC/C substrates, and that -proficient cells, suggesting that binding of Klhl6 to HBXIP does Klhl6 deficiency induces a delayed maturation along this gene not lead to its degradation. regulation pathway that controls exit from the cell cycle. Although HBXIP does not appear as a degradation target for Klhl6, its identification as a binding partner further links Klhl6 to Klhl6 interacts with HBXIP/Lamtor5 cell cycle and survival controls. Several members of the Kelch protein family behave as adaptor of KLHL6 is targeted by the somatic hypermutation process in the Cul3 E3 ubiquitin ligase, by binding Cul3 through the BTB normal GC B cells domain and different substrates through the Kelch domain (25). To study KLHL6 interactions, we established a Burkitt’s lymphoma Several mutations affecting the KLHL6 protein have been reported cell line (BL2, a GC B cell–derived lymphoma), transfected with an in B cell lymphomas, notably CLL with mutated VH genes, as well expression vector encoding human KLHL6 fused with EGFP at its as some other lymphomas of GC origin (32–34). Most of these C terminus, under the control of a tetracycline-regulated promoter, mutations target the first half of the BTB domain contained in inducible by tetracycline removal. KLHL6-EGFP was localized in exon 1 and are therefore located within 1 kb domain downstream the cytoplasm, and part of the overexpressed protein was distributed of the KLHL6 promoter, that is, within a domain compatible with in structures that were identified as aggresomes (Fig. 5). an AID-driven process (Fig. 6A). However, KLHL6 has not been Pull-down experiment showed that the EGFP-tagged KLHL6 reported so far as an AID off-target (50). protein coimmunoprecipitated endogenous Cul3 (Fig. 5A). KLHL6 AID off-targets can be easily revealed in UngxMsh2 double- and Cul3 interactions were also detected in the BL2-KLHL6-EGFP deficient mice in which AID-induced mutations are not corrected clone using the proximity ligation assay (PLA). Positive PLA sig- by error-free repair pathways (51). We therefore analyzed Klhl6 Downloaded from 2 2 nals (red dots, located mostly outside aggresome structures) were mutations in mouse Peyer’s patch GC B cells from wt and Ung / 2 2 present in the cytoplasm of BL2 cells at an average frequency of 12 Msh2 / mice. Klhl6 mutations were observed at a frequency of 2 2 2 2 2 interactions per cell, with a variation between 1 and .30 interac- 29 3 10 5/bp in Ung / Msh2 / mice, that is, a 4-fold lower level tions per cell (Fig. 5B). The number of interactions observed with than the Bcl6 gene, which displays the highest mutation load at non- an unrelated Ab in place of the anti-Cul3 Ab was at background Ig loci (51), whereas mutation frequency at the Klhl6 locus was close 3 25 level (0.03 interaction per cell), similar to the level observed, based to background values in wt animals (7.2 10 /bp)(Fig.6B).We http://www.jimmunol.org/ on anti-EGFP/anti-Cul3 Ab-mediated signals, in the BL2-KLHL6- also observed mutations within the same domain of the KLHL6 gene 2 EGFP clone in the presence of tetracycline. These results showed in human IgG memory B cells at a frequency of 17.5 3 10 5/bp, thus that KLHL6 is an interactor of the Cul3 ubiquitin ligase. making KLHL6 a new example of a gene targeted by hypermutation We then searched for KLHL6 substrates by a yeast two-hybrid in normal human B cells, together with BCL6 and FAS (52–54). system using the full-length human KLHL6 cDNA as bait. Although it is possible that KLHL6 may be a contributing factor Screening of a cDNA library derived from human spleen yielded to cell-cycle deregulation in lymphomas, as observed during B cell 14 clones that tested positive in two-hybrid assays. Nucleotide differentiation in the mouse, the frequency of mutagenesis ob- sequence analysis of these clones revealed a recurrent cDNA that served could also suggest that it only represents a bystander effect encoded a protein of 91 aa identified as HBXIP. HBXIP is a of hypermutation mistargeting. by guest on September 28, 2021 ubiquitous protein, showing complete conservation between mice and humans, and initially identified through its interaction with the Discussion hepatitis B virus HBx protein (46). HBXIP is endowed with Klhl6 is a member of the KLHL family that shows a preferential multiple functions, including control of cell proliferation, cyto- expression in hematopoietic and endothelial cells, and notably in kinesis, cell survival, and mTORC1 activation (47, 48). Interest- the B cell lineage. It is still a poorly characterized gene, and its ingly, HBXIP has also been reported to interact with survivin (49). description in Gene Ontology databases links it with GC formation. Accordingly, association of KLHL6 with endogenous HBXIP We report in this article, as described previously (35), that Klhl6 was detected in the BL2-KLHL6-EGFP cell line by PLA. Similar deficiency results in a 2-fold reduction in the pool of mature numbers of KLHL6-HBXIP interactions were found using two B cells in the mouse. However, whereas the highest expression of different Abs made in two different host species with an average of Klhl6 during B cell development occurs at the pre-B cell stage 25 positive PLA signals per cell, with a similar distribution fre- (ImmGen database), we identified a marked impairment in the quency (Fig. 5C). We next evaluated the abundance of HBXIP in differentiation step occurring between immature and transitional this cell line, with or without tetracycline induction. Western blot T1 and T2 cells as the major cause of this deficiency. analysis of HBXIP did not show any difference in protein levels in Competitive restoration with bone marrow cells from wt and BL2 induced or not for KLHL6 overexpression (Fig. 5D). We also Klhl62/2 mice indicated that B cell maturation up to the T1-like confirmed HBXIP–survivin interactions in the BL2-KLHL6- stage in bone marrow was not compromised by Klhl6 deficiency, EGFP cell line (Fig. 5E), but did not detect direct survivin– with T1-like cells (IgM+CD93+CD232) being even 2-fold in- KLHL6 interactions by PLA (data not shown). To study HBXIP creased compared with control cells. The proportion was re- expression in a physiologically more relevant context for Klhl6 equilibrated at the T1 stage in spleen, implying that their differ- function, we analyzed HBXIP protein expression in splenic T1, entiation is compromised in this organ. Klhl62/2 cells were T2, and mature B cells from Klhl62/2 and wt mice (Fig. 5D). clearly also disfavored at the T2-like stage in bone marrow (IgM+ HBXIP expression was altogether low, as expected from non- CD93+CD23+), as well as at the T2 and FO stages in spleen. proliferating cells, but did not differ between Klhl6-deficient and In vitro incubation with the BAFF cytokine revealed an impaired

displays immunoblot analysis of transitional B cell extracts (106 cells) obtained by MACS enrichment, from young Klhl62/2 and littermate control mice. Quantiﬁcation of relative fold change for survivin expression, after b-actin normalization, is reported in lower left panel. Data are mean 6 SD from 11 experiments. Right panel shows Western blot analysis of total cell extracts from FACS-sorted T1 and T2 B cells. The values 1 and 1/2 indicate that protein extracts from 5 3 105 and 2.5 3 105 cells were loaded on the gel. Blot was probed with anti-survivin and anti-GAPDH. Results are representative of one of two independent experiments. The Journal of Immunology 2417 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 5. Klhl6 interacts with Culin-3 and HBXIP. (A) BL2 cells, with inducible KLHL6-EGFP expression, were lysed and subjected to immuno- precipitation with anti-EGFP Ab. Immunoprecipitated proteins were analyzed with an anti-Cul3 Ab. (B) Three-dimensional confocal images of Cul3- KLHL6 interactions detected by PLA in the inducible KLHL6-EGFP BL2 cell line. From left to right, KLHL6-EGFP protein (green), PLA signals (red), overlay, and representative section of two cells from the outlined area, showing cytoplasmic localization of signals. Nuclei are stained in blue. Graph shows the percentage of cells harboring the indicated number of interactions. Data correspond to two independent experiments in which 1398 PLA signals were detected in 115 cells. (C) Three-dimensional confocal images of KLHL6-HBXIP/LAMTOR5 interactions analyzed by PLA. From left to right, KLHL6- EGFP protein (green), PLA signals (red), overlay, and single section of two cells from outlined area, showing cytoplasmic localization of signals. The bar graph depicts the percentage of cells harboring the indicated number of interactions. PLA-positive signals were detected with two different anti-HBXIP Abs: black bar (Ab1) represents goat anti-HBXIP (Santa Cruz), and gray bar (Ab2) represents rabbit anti-HBXIP (Sigma). Results assemble two independent experiments for both anti-HBXIP Abs. 157 cells were scored detecting 3396 interactions. (D) Western blot analysis of HBXIP in the KLH6-EGFP– inducible BL2 cell line and Klhl62/2 and control mouse B cell subsets, with b-actin as loading control. Left panel, Protein extracts from 105 wt BL2 cells, EGFP-sorted cells induced for KHL6 expression (+), and repressed cells (2). Right panel, Protein extracts from mouse T1, T2, and mature (M) B cells (1.5 3 105) from Klhl6-deﬁcient and control mice. (E) HBXIP–survivin interactions in the BL2 cell line. Left panel, Representative three-dimensional confocal image shows PLA-positive signals (red dots) that were identiﬁed with rabbit anti-HBXIP and mouse anti-survivin. Right panel, Single section from outlined area showing cytoplasmic localization of signals. Data are from two independent experiments. Scale bars, 5 mm. 2418 Klhl6 AND T1 B CELL MATURATION

FIGURE 6. Somatic hypermutation targets Klhl6 in mouse and human GC B cells. Muta- tion analysis was performed on genomic DNA from B220+PNA+GL7+ cells isolated from wt or Ung2/2Msh22/2 mice (two mice), and for human samples, from IgD2CD27+ memory B cells (two individuals) and naive B cells as controls. (A) Representation of the Klhl6 locus and the region ﬂanking exon 1 that was sequenced. (B) Klhl6 mutation frequency in GC B cells from wt and Msh22/2Ung2/2 mice (left)

and human naive and memory cells (right), with Downloaded from summary of mutation data. http://www.jimmunol.org/

survival of Klhl6-deficient splenic transitional B cells, both in terms strikingly, the transcriptional profile of Klhl62/2 splenic T1 cells of T1 and T2 cell survival and of differentiation of T1 cells into T2. displayed an upregulation (around 1.5- to 2-fold) of the same set BAFF-R expression was markedly reduced in splenic T1 cells, a of cell-cycle genes, including those controlled by the APC/C difference already observed in bone marrow T1- and T2-like cells. In complex, as compared with wt cells. The upregulation of a cell contrast, naive Klhl62/2 B cells that passed this differentiation bot- proliferation signature is reminiscent of the difference in expres- by guest on September 28, 2021 tleneck displayed normal BAFF-R expression levels at their surface sion profile observed between normal immature bone marrow and showed comparable survival in vitro in the presence of BAFF B cells and splenic T1 cells, implying clearly that Klhl6-deficient and normal proliferative responses to various activation signals. transitional B cells lag behind wt cells for their maturation. Klhl6 The reduction in BAFF-R expression and the specific stages at could therefore target a protein controlling this differentiation which Klhl62/2 B cell differentiation appears to be impaired process. Moreover, similar transcriptional changes, and of similar strikingly mirrors the phenotype of BAFF/BAFF-R–deficient an- magnitude, were observed in transitional Klhl6-deficient and wt imals (13–15, 19). One hypothesis could thus be a direct or in- transitional B cells upon 24 h of in vitro culture in the presence of direct impact of the absence of Klhl6 on BAFF-R signaling or BAFF, suggesting again that absence of Klhl6 does not induce a transcription. We failed, however, to detect any changes in the specific defect in the BAFF-R signaling pathway. activation of the alternate NF-kB pathway upon in vitro culture of We showed that Klhl6 is, like several other KLHL family splenic transitional B cells in the presence of BAFF, a pathway members, a cofactor of the Cul3 ubiquitin complex and identified whose activation is jointly promoted by BAFF and BCR sig- HBXIP as its binding partner. Interestingly, HBXIP is involved in nals (55, 56). Overexpression of KLHL6 in the BL2 Burkitt’s multiple regulatory processes, controlling cell proliferation, cy- lymphoma cell line also did not induce changes in the level of tokinesis, cell survival, and as described more recently, regulation BAFF-R expression at the surface of these cells (data not shown). of the mTOR nutrient-sensing pathway (47, 48, 57). We, however, We therefore favor another hypothesis, namely that the reduc- failed to identify any modulation of the HBXIP protein performed tion in BAFF-R expression may reflect, and not cause, a delayed by Klhl6, either upon KLHL6 overexpression in a BL2 cell line or maturation process of Klhl62/2 transitional B cells, leading to in Klhl62/2 transitional B cells. Ubiquitinylation can also induce their subsequent apoptosis. Gene profiling performed on wt tran- protein relocalization, as observed for AuroraB that, after ubiq- sitional B cells before and after 24-h culture in the presence of uitinylation by Klhl9 and Klhl13, is removed from mitotic chro- BAFF revealed a striking set of 1.5- to 2-fold downregulated genes mosomes in prometaphase and relocalized at the spindle zone linked with cell proliferation. Among them were many targets of during anaphase (58). Further studies should address whether the APC/C complex, thus revealing a coordinated transcriptional relocalization of HBXIP could take place and interfere with the regulation of proteins controlled by this large E3 ubiquitin ligase maturation process of immature B cells during B cell develop- complex that regulates entry into the different phases of the cell ment. It is worth mentioning that, in differentiation pathways like cycle through its sequential binding with two distinct stage- myogenesis, in which the role of multiple Klhl genetic mutants specific interactors, Cdc20 or Fzr1 (43). Such downmodulation have been implicated, like Klhl19, Klhl40, and Klhl41, their Cul3 of genes linked with proliferation control suggests that, in cells adaptor function has been well established, with interactors de- that are already nondividing, further transcriptional reduction may scribed for some of them, but so far no ubiquitinylation targets drive mitotic exit and enforce the G0 status of naive B cells. Very have been identified (59). The Journal of Immunology 2419

Klhl6 has been initially isolated through its marked overexpres- 11. Thomas, M. D., B. Srivastava, and D. Allman. 2006. Regulation of peripheral B cell maturation. Cell. Immunol. 239: 92–102. sion in GC B cells (31), and Sato and colleagues (35) have reported 12. Pillai, S., and A. Cariappa. 2009. The bone marrow perisinusoidal niche for an impaired GC response in Klhl6-deficient mice. We did not ob- recirculating B cells and the positive selection of bone marrow-derived serve any reduction in GC B cells in the absence of Klhl6, either B lymphocytes. Immunol. Cell Biol. 87: 16–19. 13. Schiemann, B., J. L. Gommerman, K. Vora, T. G. Cachero, S. Shulga-Morskaya, in chronically stimulated Peyer’s patches or during an anti-SRBC M. Dobles, E. Frew, and M. L. Scott. 2001. An essential role for BAFF in the response, but we cannot exclude that differences may exist in other normal development of B cells through a BCMA-independent pathway. Science immune conditions, like viral infections. We nevertheless observed 293: 2111–2114. 14. Batten, M., J. Groom, T. G. Cachero, F. Qian, P. Schneider, J. Tschopp, that active Klhl6 expression in GC B cells allows the targeting by J. L. Browning, and F. Mackay. 2000. BAFF mediates survival of peripheral AID of part of its BTB domain that lies within a promoter-proximal immature B lymphocytes. J. Exp. Med. 192: 1453–1466. region, whereas it had not been listed so far as an AID off-target 15. Sasaki, Y., S. Casola, J. L. Kutok, K. Rajewsky, and M. Schmidt-Supprian. 2004. TNF family member B cell-activating factor (BAFF) receptor-dependent and gene (50). These mutations at the Klhl6 locus were repaired in -independent roles for BAFF in B cell physiology. J. Immunol. 173: 2245–2252. 2 2 normal mouse GC B cells and could only be observed in Msh2 / 16. Yan, M., J. R. Brady, B. Chan, W. P. Lee, B. Hsu, S. Harless, M. Cancro, Ung2/2 I. S. Grewal, and V. M. Dixit. 2001. Identification of a novel receptor for repair–deficient animals. On the contrary, we observed B lymphocyte stimulator that is mutated in a mouse strain with severe B cell that KLHL6 was mutated in normal human memory B cells, deficiency. Curr. Biol. 11: 1547–1552. which adds it to the short list of mutated genes in normal human 17. Hsu, B. L., S. M. Harless, R. C. Lindsley, D. M. Hilbert, and M. P. Cancro. 2002. Cutting edge: BLyS enables survival of transitional and mature B cells through GC B cells, together with BCL-6 and FAS/CD95 (52–54). Recur- distinct mediators. J. Immunol. 168: 5993–5996. rent mutations observed in B cell lymphomas, and notably CLL 18. Rauch, M., R. Tussiwand, N. Bosco, and A. G. Rolink. 2009. Crucial role for with mutated Ig V genes, may thus be explained by such a physi- BAFF-BAFF-R signaling in the survival and maintenance of mature B cells. PLoS One 4: e5456. ological process (32–34). KLHL6 mutations observed in lympho- 19. Shulga-Morskaya, S., M. Dobles, M. E. Walsh, L. G. Ng, F. MacKay, S. P. Rao, Downloaded from mas could therefore be simple bystander products of AID off-target S. L. Kalled, and M. L. Scott. 2004. B cell-activating factor belonging to the activity, because they were not identified as driver elements of the TNF family acts through separate receptors to support B cell survival and T cell- independent antibody formation. J. Immunol. 173: 2331–2341. oncogenic process in a recent longitudinal study of CLL progres- 20. Rowland, S. L., K. F. Leahy, R. Halverson, R. M. Torres, and R. Pelanda. 2010. sion (60). In contrast, the cell division signature of Klhl6-deficient BAFF receptor signaling aids the differentiation of immature B cells into tran- transitional B cells leaves open the possible contribution of sitional B cells following tonic BCR signaling. J. Immunol. 185: 4570–4581. 21. Dhanoa, B. S., T. Cogliati, A. G. Satish, E. A. Bruford, and J. S. Friedman. 2013.

KLHL6 mutants to deregulation of proliferation in CLL. Update on the Kelch-like (KLHL) gene family. Hum. Genomics 7: 13. http://www.jimmunol.org/ 22. Albagli, O., P. Dhordain, C. Deweindt, G. Lecocq, and D. Leprince. 1995. The BTB/POZ domain: a new protein-protein interaction motif common to DNA- Acknowledgments and actin-binding proteins. Cell Growth Differ. 6: 1193–1198. We thank our colleagues Robert Weil and Jurgen€ Wienands for expertise 23. Stogios, P. J., and G. G. Prive´. 2004. The BACK domain in BTB-kelch proteins. in the study of NF-kB and B cell signaling. We thank Sandra Weller for Trends Biochem. Sci. 29: 634–637. 24. Adams, J., R. Kelso, and L. Cooley. 2000. The kelch repeat superfamily of providing human B cell samples. We acknowledge the contributions of proteins: propellers of cell function. Trends Cell Biol. 10: 17–24. Je´rome Megret (Flow Cytometry Core, Structure Fe´de´rative de Recherche 25. Canning, P., C. D. Cooper, T. Krojer, J. W. Murray, A. C. Pike, A. Chaikuad, Necker) in performing cell sorting and Nicolas Goudin (Cell Imaging Core, T. Keates, C. Thangaratnarajah, V. Hojzan, V. Ayinampudi, et al. 2013. Structural Structure Fe´de´rative de Recherche Necker) in assisting with confocal mi- basis for Cul3 protein assembly with the BTB-Kelch family of E3 ubiquitin ligases. [Published erratum appears in 2013 J. Biol. Chem. 288: 28304.] J. Biol. croscopy analysis. We thank Claire Boudet and the personnel of Labora- Chem. 288: 7803–7814. by guest on September 28, 2021 toire d’Expe´rimentation Animale et Transgeńe`se (Structure Fe´de´rative de 26. Lee, D. F., H. P. Kuo, M. Liu, C. K. Chou, W. Xia, Y. Du, J. Shen, C. T. Chen, Recherche Necker) for mouse breeding and handling. L. Huo, M. C. Hsu, et al. 2009. KEAP1 E3 ligase-mediated downregulation of NF-kappaB signaling by targeting IKKbeta. Mol. Cell 36: 131–140. 27. Lee, Y. R., W. C. Yuan, H. C. Ho, C. H. Chen, H. M. Shih, and R. H. Chen. 2010. Disclosures The Cullin 3 substrate adaptor KLHL20 mediates DAPK ubiquitination to The authors have no financial conflicts of interest. control interferon responses. EMBO J. 29: 1748–1761. 28. Maerki, S., M. H. Olma, T. Staubli, P. Steigemann, D. W. Gerlich, M. Quadroni, I. Sumara, and M. Peter. 2009. The Cul3-KLHL21 E3 ubiquitin ligase targets aurora B to midzone microtubules in anaphase and is required for cytokinesis. J. References Cell Biol. 187: 791–800. 1. Nemazee, D. A., and K. Burki.€ 1989. Clonal deletion of B lymphocytes in a 29. Yan, W., L. Ma, K. H. Burns, and M. M. Matzuk. 2004. Haploinsufficiency of transgenic mouse bearing anti-MHC class I antibody genes. Nature 337: 562–566. kelch-like protein homolog 10 causes infertility in male mice. Proc. Natl. Acad. 2. Goodnow, C. C., J. Sprent, B. Fazekas de St Groth, and C. G. Vinuesa. 2005. Sci. USA 101: 7793–7798. Cellular and genetic mechanisms of self tolerance and autoimmunity. Nature 30. Ravenscroft, G., S. Miyatake, V. L. Lehtokari, E. J. Todd, P. Vornanen, K. S. Yau, 435: 590–597. Y. K. Hayashi, N. Miyake, Y. Tsurusaki, H. Doi, et al. 2013. Mutations in 3. Allman, D. M., S. E. Ferguson, V. M. Lentz, and M. P. Cancro. 1993. Peripheral KLHL40 are a frequent cause of severe autosomal-recessive nemaline myopathy. B cell maturation. II. Heat-stable antigen(hi) splenic B cells are an immature Am. J. Hum. Genet. 93: 6–18. developmental intermediate in the production of long-lived marrow-derived 31. Gupta-Rossi, N., S. Storck, P. J. Griebel, C. A. Reynaud, J. C. Weill, and B cells. J. Immunol. 151: 4431–4444. A. Dahan. 2003. Specific over-expression of deltex and a new Kelch-like protein 4. Loder, F., B. Mutschler, R. J. Ray, C. J. Paige, P. Sideras, R. Torres, in human germinal center B cells. Mol. Immunol. 39: 791–799. M. C. Lamers, and R. Carsetti. 1999. B cell development in the spleen takes 32. Morin, R. D., M. Mendez-Lago, A. J. Mungall, R. Goya, K. L. Mungall, place in discrete steps and is determined by the quality of B cell receptor-derived R. D. Corbett, N. A. Johnson, T. M. Severson, R. Chiu, M. Field, et al. 2011. signals. J. Exp. Med. 190: 75–89. Frequent mutation of histone-modifying genes in non-Hodgkin lymphoma. 5. Chung, J. B., M. Silverman, and J. G. Monroe. 2003. Transitional B cells: step by Nature 476: 298–303. step towards immune competence. Trends Immunol. 24: 343–349. 33. Puente, X. S., M. Pinyol, V. Quesada, L. Conde, G. R. Ordoń˜ez, N. Villamor, 6.Allman,D.,R.C.Lindsley,W.DeMuth,K.Rudd,S.A.Shinton,andR.R.Hardy. G. Escaramis, P. Jares, S. Bea`, M. Gonza´lez-Dıáz, et al. 2011. Whole-genome 2001. Resolution of three nonproliferative immature splenic B cell subsets reveals sequencing identifies recurrent mutations in chronic lymphocytic leukaemia. multiple selection points during peripheral B cell maturation. J. Immunol. 167: Nature 475: 101–105. 6834–6840. 34.Puente,X.S.,S.Bea`,R.Valde´s-Mas, N. Villamor, J. Gutie´rrez-Abril, 7. Chung, J. B., R. A. Sater, M. L. Fields, J. Erikson, and J. G. Monroe. 2002. CD23 J. I. Martıń-Subero,M.Munar,C.Rubio-Pe´rez, P. Jares, M. Aymerich, et al. defines two distinct subsets of immature B cells which differ in their responses to 2015. Non-coding recurrent mutations in chronic lymphocytic leukaemia. T cell help signals. Int. Immunol. 14: 157–166. Nature 526: 519–524. 8. Rolink, A. G., J. Andersson, and F. Melchers. 1998. Characterization of im- 35. Kroll, J., X. Shi, A. Caprioli, H. H. Liu, C. Waskow, K. M. Lin, T. Miyazaki, mature B cells by a novel monoclonal antibody, by turnover and by mitogen H. R. Rodewald, and T. N. Sato. 2005. The BTB-kelch protein KLHL6 is in- reactivity. Eur. J. Immunol. 28: 3738–3748. volved in B-lymphocyte antigen receptor signaling and germinal center forma- 9. Cariappa, A., C. Chase, H. Liu, P. Russell, and S. Pillai. 2007. Naive recircu- tion. Mol. Cell. Biol. 25: 8531–8540. lating B cells mature simultaneously in the spleen and bone marrow. Blood 109: 36. Dogan, I., B. Bertocci, V. Vilmont, F. Delbos, J. Me´gret, S. Storck, 2339–2345. C. A. Reynaud, and J. C. Weill. 2009. Multiple layers of B cell memory with 10. Lindsley, R. C., M. Thomas, B. Srivastava, and D. Allman. 2007. Generation of different effector functions. Nat. Immunol. 10: 1292–1299. peripheral B cells occurs via two spatially and temporally distinct pathways. 37. Nagaoka, H., S. Ito, M. Muramatsu, M. Nakata, and T. Honjo. 2005. DNA Blood 109: 2521–2528. cleavage in immunoglobulin somatic hypermutation depends on de novo protein 2420 Klhl6 AND T1 B CELL MATURATION

synthesis but not on uracil DNA glycosylase. Proc. Natl. Acad. Sci. USA 102: identiﬁcation of the genomic targets of the cytidine deaminase AID and its co- 2022–2027. factor RPA in B lymphocytes. Nat. Immunol. 12: 62–69. 38. Walhout, A. J., and M. Vidal. 2001. High-throughput yeast two-hybrid assays for 51. Liu, M., J. L. Duke, D. J. Richter, C. G. Vinuesa, C. C. Goodnow, large-scale protein interaction mapping. Methods 24: 297–306. S. H. Kleinstein, and D. G. Schatz. 2008. Two levels of protection for the B cell 39. Miller, J. P., J. E. Stadanlick, and M. P. Cancro. 2006. Space, selection, and genome during somatic hypermutation. Nature 451: 841–845. surveillance: setting boundaries with BLyS. J. Immunol. 176: 6405–6410. 52. Pasqualucci, L., A. Migliazza, N. Fracchiolla, C. William, A. Neri, L. Baldini, 40. Tussiwand, R., M. Rauch, L. A. Fluck,€ and A. G. Rolink. 2012. BAFF-R ex- R.S.Chaganti,U.Klein,R.Kuppers,€ K. Rajewsky, and R. Dalla-Favera. pression correlates with positive selection of immature B cells. Eur. J. Immu- 1998. BCL-6 mutations in normal germinal center B cells: evidence of so- nol. 42: 206–216. matic hypermutation acting outside Ig loci. Proc.Natl.Acad.Sci.USA95: 41. Claudio, E., K. Brown, S. Park, H. Wang, and U. Siebenlist. 2002. BAFF- 11816–11821. induced NEMO-independent processing of NF-kappa B2 in maturing B cells. 53. Shen, H. M., A. Peters, B. Baron, X. Zhu, and U. Storb. 1998. Mutation of BCL- Nat. Immunol. 3: 958–965. 6 gene in normal B cells by the process of somatic hypermutation of Ig genes. 42. Manchado, E., M. Eguren, and M. Malumbres. 2010. The anaphase-promoting Science 280: 1750–1752. complex/cyclosome (APC/C): cell-cycle-dependent and -independent functions. 54. Muschen,€ M., D. Re, B. Jungnickel, V. Diehl, K. Rajewsky, and R. Kuppers.€ Biochem. Soc. Trans. 38: 65–71. 2000. Somatic mutation of the CD95 gene in human B cells as a side-effect of 43. Sivakumar, S., and G. J. Gorbsky. 2015. Spatiotemporal regulation of the the germinal center reaction. J. Exp. Med. 192: 1833–1840. anaphase-promoting complex in mitosis. Nat. Rev. Mol. Cell Biol. 16: 82–94. 55. Stadanlick, J. E., M. Kaileh, F. G. Karnell, J. L. Scholz, J. P. Miller, W. J. Quinn, 44. Karnell, F. G., R. J. Brezski, L. B. King, M. A. Silverman, and J. G. Monroe. III, R. J. Brezski, L. S. Treml, K. A. Jordan, J. G. Monroe, et al. 2008. Tonic 2005. Membrane cholesterol content accounts for developmental differences in B cell antigen receptor signals supply an NF-kappaB substrate for prosurvival surface B cell receptor compartmentalization and signaling. J. Biol. Chem. 280: BLyS signaling. Nat. Immunol. 9: 1379–1387. 25621–25628. 56. Sasaki, Y., E. Derudder, E. Hobeika, R. Pelanda, M. Reth, K. Rajewsky, and 45. Altieri, D. C. 2015. Survivin - the inconvenient IAP. Semin. Cell Dev. Biol. 39: M. Schmidt-Supprian. 2006. Canonical NF-kappaB activity, dispensable for 91–96. B cell development, replaces BAFF-receptor signals and promotes B cell pro- 46. Melegari, M., P. P. Scaglioni, and J. R. Wands. 1998. Cloning and character- liferation upon activation. Immunity 24: 729–739. ization of a novel hepatitis B virus x binding protein that inhibits viral repli- 57. Garcia-Saez, I., F. B. Lacroix, D. Blot, F. Gabel, and D. A. Skouﬁas. 2011. cation. J. Virol. 72: 1737–1743. Structural characterization of HBXIP: the protein that interacts with the anti- Downloaded from 47. Bar-Peled, L., L. D. Schweitzer, R. Zoncu, and D. M. Sabatini. 2012. Ragulator apoptotic protein survivin and the oncogenic viral protein HBx. J. Mol. Biol. is a GEF for the rag GTPases that signal amino acid levels to mTORC1. Cell 405: 331–340. 150: 1196–1208. 58. Sumara, I., M. Quadroni, C. Frei, M. H. Olma, G. Sumara, R. Ricci, and 48. Wen, Y., V. S. Golubkov, A. Y. Strongin, W. Jiang, and J. C. Reed. 2008. In- M. Peter. 2007. A Cul3-based E3 ligase removes Aurora B from mitotic chro- teraction of hepatitis B viral oncoprotein with cellular target HBXIP dysregu- mosomes, regulating mitotic progression and completion of cytokinesis in hu- lates centrosome dynamics and mitotic spindle formation. J. Biol. Chem. 283: man cells. Dev. Cell 12: 887–900. 2793–2803. 59. Gupta, V. A., and A. H. Beggs. 2014. Kelch proteins: emerging roles in skeletal

49. Marusawa, H., S. Matsuzawa, K. Welsh, H. Zou, R. Armstrong, I. Tamm, and muscle development and diseases. Skelet. Muscle 4: 11. http://www.jimmunol.org/ J. C. Reed. 2003. HBXIP functions as a cofactor of survivin in apoptosis sup- 60. Nadeu, F., J. Delgado, C. Royo, T. Baumann, T. Stankovic, M. Pinyol, P. Jares, pression. EMBO J. 22: 2729–2740. A. Navarro, D. Martı´n-Garcı´a, S. Bea`, et al. 2016. Clinical impact of clonal and 50. Yamane, A., W. Resch, N. Kuo, S. Kuchen, Z. Li, H. W. Sun, D. F. Robbiani, subclonal TP53, SF3B1, BIRC3, NOTCH1, and ATM mutations in chronic K. McBride, M. C. Nussenzweig, and R. Casellas. 2011. Deep-sequencing lymphocytic leukemia. Blood 127: 2122–2130. by guest on September 28, 2021