ZBTB38 is dispensable for antibody responses

Ethics statement

All procedures in this study were specifically approved and carried out in accordance with the guidelines set forth by the Institutional Animal Care and Use Committee at Washington University (approval 20140030) and at the University of Arizona (approval 17–266). Euthanasia was performed by administering carbon dioxide at 1.5L/minute into a 7L chamber until 1 minute after respiration ceased. After this point, cervical dislocation was performed to ensure death.

Mice

All mice were housed and bred in pathogen-free facilities. C57BL6/N mice were obtained from the National Cancer Institute. B6.Cg-Igh^aThy1^aGpi1^a (IgH^a) mice were obtained from Charles River Laboratories. Zbtb38^fl/+ mice were generated by injecting C57Bl6/J pronuclear zygotes with ribonucleoparticles of Cas9 protein and guide RNAs targeting sites flanking exon 3 of Zbtb38, alongside oligonucleotide donors as homologous recombination donors spanning these same gRNA sites. Oligonucleotide substrates contained loxP sites to interrupt gRNA target sites to prevent Cas9 re-cutting after successful recombination. A single successful founder (out of 33 tested) was identified by PCR and then bred to C57Bl6/N mice for germline transmission. Mice have been maintained by backcrossing to C57Bl6/N animals. Animals will be made available at the Mutant Mouse Resource and Research Centers upon publication of this manuscript (B6N.B6J-Zbtb38em1Dbhat/Mmucd, Strain ID 66871). ZBTB38 f/f mice were crossed to CMV-Cre (Jackson Laboratory, stock no. 006054) or VavCre (Jackson Laboratory, stock no. 008610) and maintained as ZBTB38 f/f or ZBTB38 x CMV- or Vav-Cre where littermates were used as controls. The following primers were used to confirm recombination of the targeting plasmid: SP55.mZbtb38.5'genomic.F2 5’- CCAGGGATTCAGTCCTCAGCA-3’, SP55.mZbtb38.3'genomic.R2 5’- GCCTACCCCAAACCACACTAA-3’. The following primers were used for genotyping the Zbtb38 allele: 5’LoxP forward 5’- TCTGAGTTCAAGGCCAGCTT-3’, 5’LoxP reverse 5’- TCTCCAAGCAGAAAGGGTGT-3’, and 3’LoxP reverse 5’- GGGTCGTTAGAGGATTCAGC-3’.

Immunizations

Mice were immunized intraperitoneally with 100 μg NP-OVA (Biosearch), adjuvanted with Alhydrogel (Invivogen). NP-APC used for staining was made by conjugating allophycocyanin (Sigma-Aldrich) with 4-hydroxy-3-nitrophenylacetyl-O-succinimide ester (Biosearch Technologies).

RNA extraction, cDNA synthesis, and qRT-PCR

Total RNA was extracted with TRIzol (Life technologies) and cDNA synthesized using Superscript III Reverse transcription kit with random hexamers (Life Technologies) according to manufacturer’s instructions. qRT-PCR was performed using SYBR Green PCR master mix (Applied Biosystems) on a Prism 7000 Sequence Detection System (Applied Biosystems). The primers used for Zbtb38 are: forward 5’- AGAACCAAGGATTTCCGAGTG-3’ and reverse 5’-GATGGAGAGTACTGTGTCACTG-3’. Zbtb38 transcript levels were normalized to 18S ribosomal RNA, forward 5’-CGGCTACCACATCCAAGGAA-3’ and reverse 5’-GCTGGAATTACCGCGGCT-3’ [26].

RNA-sequencing

RNA from germinal center B cells was extracted with Macherey-Nagel Nucleospin XS kits. cDNA libraries were prepared by Novogene using SmartSeq v4 kits (Takara) and processed for paired-end PE150 RNA-sequencing on an Illumina Hiseq 4000 lane. For visualization of ZBTB38 transcripts, fastq files were mapped and aligned to the mm10 genome using HiSat2 and displayed using IgV [27, 28]. For quantification of gene expression differences, fastq files were mapped using vM17 annotation files from Gencode and transcript abundances were quantified by Salmon [29]. Differentially-expressed genes quantified by DESeq2 [30]. Volcano plots were displayed using Prism software (GraphPad). Data is available on NCBI GEO (GSE152109).

ELISA

ELISA plates (9018, Corning) were coated overnight at 4°C in 0.1 M sodium bicarbonate buffer, pH 9.5 containing 5 ug/mL of NP₁₆- or NP₄-BSA (Bioresearch Technologies). All other incubation steps were performed at room temperature for 1 hour. Wash steps were performed between each step using PBS + 0.05% Tween-20. Plates were blocked with PBS + 2% BSA followed by serial dilutions of serum. Serum was probed with 0.1 ug/mL of biotinylated anti-mouse IgG (715-065-151, Jackson ImmunoResearch Laboratories) and then detected with streptavidin conjugated horseradish peroxidase (554066, BD biosciences). Plates were developed using TMB (Dako, S1599) and neutralized with 2N H₂SO₄. Optical density (OD) values were measured at 450 nm. Serum endpoint titer was defined as the inverse dilution factor that is three standard deviations above background using one-phase decay measurements and Prism software (GraphPad Software).

Adoptive transfer for recall responses

Splenocytes from NP-immunized ZBTB38^f/f or ZBTB38^f/f x VavCre mice were isolated and processed into single cell suspension, erythrocytes lysed using an ammonium chloride-potassium solution, and lymphocytes isolated by using a Hisopaque-1119 (Sigma-Aldrich) density gradient. Cells were washed twice prior to transfer. 10% of cells were retained for cellular analysis whereas the remaining 90% of cells were transferred into one non-irradiated IgH^a recipient mice by intravenous injection. A recall response was elicited by intravenously challenging mice with soluble NP-OVA 24 hours later.

Flow cytometry

Single cell suspensions were prepared from bone marrow or spleen, erythrocytes lysed using an ammonium chloride-potassium solution, and lymphocytes isolated by using a Hisopaque-1119 (Sigma-Aldrich) density gradient. Cells were resuspended in PBS with 5% adult bovine serum and 2 mM EDTA prior to staining with antibodies and NP-APC. The following antibodies were purchased from Biolegend: 6D5 (CD19)-Alexa Fluor 700; GL7-FITC; 281–2 (CD138)-PE; RMM-1 (IgM)-APC; 11-26c.2a (IgD)-PerCP-Cy5.5 or -Brilliant Violet 605; 16-10A1 (CD80)-PE; RA3-6B2 (B220)-FITC, -Pacific Blue, or APC-Cy7; PO3 (CD86)-FITC; PK136 (NK-1.1)-PerCP-Cy5.5; M1/70 (CD11b)-Pacific Blue; HK1.4 (Ly-6C)-Brilliant Violet 510; 1A8 (Ly-6G)-Brilliant Violet 605; A7R34 (IL-7R)-Brilliant Violet 421; E13-16.7 (Ly-6A/E)-PE; and 29-2L17 (CCR6)-PE-Cy7. The following antibodes were purchased from eBioscience: 11/41 (IgM)-PerCP-e710; 11-26c (IgD)-FITC; 2B11 (CXCR4)-PerCP-e710; 2B8 (c-Kit)-PE-Cy7; and LG.7F9 (CD27)-APC. The following antibodies were purchased from BD Pharmingen: 53–6.7 (CD8a)-PE; RM4-5 (CD4)-PE-Cy7; A2F10.1 (CD135)-PE-CF594; and 93 (CD16/CD32)-PerCP-Cy5.5.

Cells were stained on ice for 20 minutes. Germinal center B cells were enriched by staining cells with GL7-PE followed with anti-PE magnetic beads (0.5 uL/10⁷ cells, Miltenyi Biotec). Positive enrichment of GL7-expressing cells was performed using MACS LS columns (Miltenyi Biotec).

Statistical analysis

Statistical analyses were performed using GraphPad Prism 8. Specific tests used, and significance, is indicated in each panel.

ZBTB38 is highly expressed in B cell subsets

RNA-sequencing studies have reported expression of ZBTB38 in plasma cells [31]. To look more broadly across hematopoietic lineages, ZBTB38 gene expression in multiple cell types was first analyzed using available RNA-sequencing data assembled by the ImmGen Consortium [32]. A subset of cell types expressing DESeq2-normalized ZBTB38 counts greater than 800 are shown in Fig 1A [30]. High expression of ZBTB38 was observed in splenic plasma cells and plasmablasts (PC and PB), and in both light zone and dark zone germinal center B cells (LZ and DZ, Fig 1A). To confirm these data, wild-type mice were immunized with alhydrogel-adjuvanted 4-hydroxy-3-nitrophenyl-acetyl (NP) conjugated to ovalbumin (OVA) and naïve B cells (CD19⁺CD138^-GL7^-IgD⁺), NP-specific dark (CXCR4⁺) and light zone (CD86⁺) GC B cells (CD19⁺GL7⁺IgD^-), and NP-specific splenic plasma cells (CD138⁺) were sorted 11 days after immunization [33]. RNA was extracted from sorted cells and quantitative RT-PCR performed to quantify ZBTB38 transcript levels. GC B cells and splenic plasma cells contained 9- and 3-fold, respectively, higher expression of ZBTB38 compared to naïve B cells (Fig 1B; S1 Fig). We further confirmed these findings by examining published microarray data on dark and light zone GCs [33]. Both GC subsets expressed 3–4 fold higher levels of ZBTB38 than did naïve B cells across two probesets (Fig 1C). No differences in ZBTB38 expression were observed between light and dark zone GC B cells.

Generation and validation of ZBTB38 conditional knockout mice

To assess the functional role of ZBTB38 in vivo, we generated Zbtb38 floxed mice by targeting exon 3 (Fig 2A). This terminal exon contains the entire protein-coding sequence of ZBTB38. Single-stranded oligonucleotides containing loxP sites and flanking sequences of exon 3 of Zbtb38 were microinjected alongside Cas9/guideRNA ribonucleoparticles into C57Bl6/J zygotes. gRNA sites were designed to flank the endogenous Zbtb38 exon 3 and be disrupted upon homologous recombination with the targeting oligonucleotides. NdeI and EcoRI restriction sites were included in the oligonucleotides near the 5’ LoxP and 3’ LoxP sites to screen for successful homologous recombination of the targeting construct. Correct targeting of exon 3 was confirmed by PCR and restriction enzyme digestions (Fig 2B). Wild-type, targeted, and Zbtb38-deleted mice were distinguished using a set of three PCR primers flanking the 5’ and 3’ LoxP sequences (Fig 2C). To confirm deletion of Zbtb38 exon 3 at the genomic level, Zbtb38 f/f mice were crossed to mice expressing CMV-Cre to obtain germline ZBTB38 deletion. Tail genomic DNA from Zbtb38 f/f x CMV-Cre was amplified and deletion was confirmed by a 4 kb reduction in PCR amplicon size (Fig 2D). Despite genome-wide association studies linking polymorphisms in the Zbtb38 locus to human height [34–36], no differences were observed in the size of ZBTB38 deficient vs. wild-type littermates, and animals were born in expected Mendelian ratios.

Given the high expression of ZBTB38 in blood lineages, we crossed Zbtb38 f/f mice to VavCre mice (Zbtb38 f/f x VavCre, ZBTB38 KO), which express Cre recombinase primarily in hematopoietic cells [37]. To confirm loss of ZBTB38 expression in this system, we performed RNA-seq on germinal center B cells isolated 2 weeks after immunization with NP-OVA. After mapping reads and aligning to the mouse genome, we observed that transcripts within the floxed exon 3 were completely abrogated in ZBTB38 KO mice, whereas intermediate levels were observed in Zbtb38 f/+ x VavCre heterozygous littermates (Fig 3A) relative to Zbtb38 f/f controls that lack Cre recombinase (ZBTB38 WT). We next used Salmon to quantify transcript abundances and DESeq2 to identify genes differentially expressed between ZBTB38-deficient mice and controls [29, 30]. Other than ZBTB38 itself, very few genes were dysregulated in ZBTB38-deficient cells (Fig 3B; S1 Table). To increase statistical power, we compared ZBTB38-deficient samples to both wild-type and heterozygous controls. Again, very few differences were observed (Fig 3C; S2 Table). The few dysregulated genes were expressed at very low levels (e.g. CD3e) across genotypes and did not to fall into any obvious pathways or signatures.

ZBTB38 deficiency does not impair B cell responses

To determine if ZBTB38 expression is important for B cell development, we examined the frequency of B cell subsets in ZBTB38 WT and ZBTB38 KO mice. ZBTB38 deficiency did not alter the frequencies of follicular (FoB, CD11b^-CD19⁺IgD⁺IgM⁺), marginal zone (MZ, CD11b^-CD19⁺CD93^-CD21⁺CD23^-), T1 transitional (CD11b^-CD19⁺CD93⁺,CD23^-), T2/3 transitional (CD11b^-CD19⁺CD93⁺,CD23⁺), isotype-switched memory B cells (swIg MBCs, CD19⁺IgM^-IgD^-CD80⁺CCR6⁺), or total splenocyte numbers (S2A Fig). Bone marrow cellularity and plasma cell numbers in unimmunized mice were also similar across genotypes (S2B Fig).

Deficiencies in two other BTB-ZF factors, ZBTB20 and ZBTB32, do not impair B cell development at steady state but cause profound effects on plasma cell lifespan despite modest changes in gene expression [10–13]. These differences are most pronounced in specific immunization conditions and are not readily apparent in polyclonal populations of unimmunized animals. Therefore, to determine if ZBTB38 has a functional role in primary B cell responses, we first examined GC reactions. We immunized ZBTB38 WT and KO mice with NP-OVA and quantified the frequency of NP-specific GC B cells 2 weeks later, which corresponds with peak GC reactions [38]. We observed no differences in the frequencies of NP-specific GC B cells between ZBTB38 WT and KO mice (Fig 4A; S2C and S2D Fig). NP-specific serum antibodies were also similar between ZBTB38 WT and KO mice at all time points measured (Fig 4B). To specifically quantify the level of high affinity antibodies in the serum by ELISA, low density antigen (NP₄) was used to probe for antibody binding. Low density NP (NP₄) is used to capture antibodies with slow off-rates, which is correlated with antigen high affinity. This contrasts with high density NP (NP₁₆), which can capture antibodies with faster off rates due to the increased concentration of antigen. Unlike the total levels of NP-specific antibodies over time, which plateaued two weeks after immunization, the concentration of high affinity antibodies increased steadily over time and plateaued four weeks after immunization (Fig 4C). No difference in the quantity of high affinity antibodies was observed between ZBTB38 WT and KO mice (Fig 4C). Furthermore, the extent of affinity maturation, quantified as the ratio of NP₄ to NP₁₆ endpoint titers, was similar between ZBTB38 WT and KO mice (Fig 4D). Finally, we observed similar numbers and frequencies of antigen-specific bone marrow plasma cells (Fig 4E, S2E Fig).

Possible explanations for the similar serum antibody levels in ZBTB38 WT and KO mice include similar frequencies of antigen-specific LLPCs or compensatory increased antibody secretion from fewer LLPCs. To differentiate between these two possibilities, the frequency of NP-specific LLPCs was quantified by flow cytometry 8 weeks after alhydrogel-adjuvanted NP-OVA immunization of ZBTB38 WT and KO mice. The frequency and number of NP-specific LLPCs was similar between ZBTB38 WT and KO mice (Fig 4E; S2E Fig). These data demonstrate that ZBTB38 is dispensable for primary antibody responses to hapten-protein antigens.

To determine if ZBTB38 expression is required for secondary responses and MBC differentiation, the frequency of NP-specific memory B cells (CD19⁺GL7^-IgM^-IgD^-CD80⁺CCR6⁺) was quantified in ZBTB38 WT and KO mice 8 weeks after immunization with alhydrogel-adjuvanted NP-OVA. No difference was observed in the frequency or number of NP-specific MBCs (Fig 5A; S3 Fig). To assess MBC function, splenocytes from ZBTB38 WT and KO mice were adoptively transferred into allotype-distinct naïve IgH^a recipients and mice challenged with soluble NP-OVA 24 hours later. Donor IgH^b NP-specific antibodies originating from ZBTB38 WT and KO mice were tracked over time. NP-specific antibody titers were not altered by ZBTB38 deficiency (Fig 5B). Thus, ZBTB38 is also dispensable for secondary B cell responses.

ZBTB38 deficiency does not alter the development of hematopoietic cells

Given ZBTB38 expression in hematopoietic progenitors (Fig 1A), we next assessed if ZBTB38 deficiency alters the development of lymphoid and/or myeloid lineages by quantifying the frequencies of different cell populations by flow cytometry. We first focused on hematopoietic stem cells (HSCs, cKit⁺Sca1⁺Flk2^-CD27⁺) and progenitors with varying degrees of lineage commitment in the bone marrow [39]. HSCs differentiate into multipotent progenitors (MPPs, cKit⁺Sca1⁺Flk2⁺CD27⁺) that can give rise to both the myeloid and lymphoid lineages. MPPs can then differentiate into common myeloid progenitors (CMPs, cKit⁺Sca1^-Flk2⁺FcγR^-) or common lymphoid progenitors (CLPs, cKit^-/loSca1^-CD27⁺FcγR^-Flk2⁺IL7Rα⁺) [40, 41]. CLPs give rise to B, T, natural killer (NK), and innate like cells (ILCs). CMPs give rise to basophils, eosinophils, mast cells, dendritic cells as well as granulocyte monocyte progenitors (GMPs, cKit⁺Sca1^-Flk2^-FcγR⁺) [41]. GMPs can then give rise to neutrophils and monocytes. We identified no statistically significant differences in the frequencies of these progenitor populations between ZBTB38 WT and KO mice (Fig 6A; S4 Fig).

Mean ± SEM are shown; each symbol represents an individual mouse. Statistical significance was determined by Mann-Whitney test; n.s. indicates no significance (p > 0.05). (B) Frequencies of B cells (B220⁺), natural killer (NK, NK1.1⁺) cells, CD4⁺ and CD8⁺ T cells (B220^-CD11b^-NK.1^-), monocytes (CD11b⁺Ly6C^hiLy6G^-), and neutrophils (CD11b⁺Ly6C⁺Ly6G⁺) of peripheral blood mononuclear cells (PBMCs) from ZBTB38 WT and KO mice were quantified by flow cytometry. Mean ± SEM are shown; each symbol represents an individual mouse. Statistical significance was determined by Mann-Whitney test; n.s. indicates no significance (p > 0.05).

To assess if mature hematopoietic lineages require ZBTB38 for their development or maintenance, we analyzed the frequencies of different peripheral blood mononuclear cell (PBMCs) populations. We observed no differences between ZBTB38 WT and KO mice in the frequencies of B cells (B220⁺), NK cells (NK1.1⁺), T (CD4⁺ or CD8⁺ B220^-NK1.1^-CD11b^-), monocytes (CD11b⁺Ly6C^hiLy6G^-), or neutrophils (CD11b⁺Ly6C⁺Ly6G⁺) (Fig 6B). Thus, ZBTB38 is not required for the development or maintenance of these hematopoietic lineages.