Research ArticleIMMUNOLOGY

A tyrosine sulfation–dependent HLA-I modification identifies memory B cells and plasma cells

See allHide authors and affiliations

Science Advances  07 Nov 2018:
Vol. 4, no. 11, eaar7653
DOI: 10.1126/sciadv.aar7653
  • Fig. 1 VLRB N8 recognizes human Bmem and PC in blood and tonsils.

    (A) Peripheral blood mononuclear cells (PBMCs) were separated into Bmem (CD19+/IgD/CD27+) and marginal zone-equivalent (MZe) cells (CD19+/IgD+/CD27+), naïve B cells (CD19+/IgD+/CD27), non–B/T cells (CD19/CD3), and T cells (CD19/CD3+). A representative of 12 examined PBMC samples is shown. Human tonsillar lymphocytes were separated into the following subpopulations: Bmem (CD19+/IgD/CD38), PC (CD19+/IgD/CD38++), naïve B cells (CD19+/IgD+/CD38), pre-GC B cells (CD19+/IgD+/CD38+), GC B cells (CD19+/IgD/CD38+), non–B/T cells (CD19/CD3), and T cells (CD19/CD3+). A representative example of 14 analyzed tonsil samples is shown. VLRB N8 reactivity by flow cytometric analysis is indicated by solid red lines, and VLR4 reactivity (specific for the BclA antigen of the exosporium of Bacillus anthracis) as a negative control is shown as solid gray histogram. (B) Ribbon model of a monomeric antigen-binding unit of VLRB N8. Parallel β sheets lining the inner concave surface encoded by the N-terminal capping LRR are shown in blue, and sequences encoded by the LRR1, variable LRRV1-4, LRRVe and connecting peptide (CP) units are depicted in orange. A variable loop protruding from the capping C-terminal LRR is shown in red. The model was generated using the IntFOLD modeling platform (49). (C) Frequencies of VLRB N8–reactive cells for each analyzed cell population in healthy blood and tonsil samples. (D) Frequencies of VLRB N8–reactive cells between FCRL4+ and FCRL4 Bmem from 14 additional tonsil specimens. Statistical significance was assessed using a Wilcoxon signed-rank test, ***P < 0.001; n = 14.

  • Fig. 2 VLRB N8 recognizes a Bmem/PC-specific epitope of HLA-I.

    (A) KMS-11 cells and (B) primary human Bmem were transfected with siRNA targeting β2-microglobulin (Δβ2) or scrambled control siRNA (−) before analysis for VLRB N8 reactivity (top row). Modulation of HLA-I cell surface expression was assessed using conventional anti–HLA-I antibodies (bottom row). Off-target effects of transfected siRNA were assessed by staining with VLRB EHT46 and conventional anti-CD138, respectively. (C) Pan–HLA-I antibodies block the recognition of HLA-I by VLRB N8. KMS-11 cells were preincubated with anti–HLA-I antibodies G46-2.6 or w6/32, anti–β2-microglobulin antibodies BBM.1 or BM-63, free HLA-I heavy chain–detecting HC-10 antibodies, or anti-CD138 antibodies before the addition of VLRB N8. The binding of VLRB N8 (left) or the blocking antibodies (right) was assessed by flow cytometric analysis. MFIs normalized to negative control VLR4 or isotype-matched control antibodies ± SD (n = 5) are shown. Ab, antibody. (D) Pan–HLA-I antibody w6/32 blocks the recognition of HLA-I by VLRB N8. Tonsillar lymphocytes were preincubated with anti–HLA-I antibodies w6/32 or HC-10, followed by evaluation of VLRB N8 binding. MFIs normalized to negative control VLR4 or isotype-matched control antibodies ± SD (n = 12) are shown. Statistically significant differences of P < 0.05 were determined with Student’s t test (A and C) and Wilcoxon signed-rank test (B and D) and were indicated by *P < 0.05, **P < 0.01, and ***P < 0.001.

  • Fig. 3 VLRB N8 reactivity does not correlate with HLA-I cell surface expression levels.

    Blood from healthy donors (HDs) and individuals diagnosed with multiple sclerosis or SLE, or tonsillar lymphocytes were analyzed for VLRB N8 binding (top row) and HLA-I binding (bottom row). Cell populations were gated as described in Fig. 1. Atypical Bmem were defined as CD19+/CD3/IgD/CD27. Median values for each population are indicated by red horizontal bars. V/H indicates the numerical values of the median of VLRB N8 signals normalized to the corresponding value of HLA-I.

  • Fig. 4 Induction of VLRB N8 binding following B cell activation.

    (A) BJAB cells were stimulated with PMA and ionomycin (P/I) for 1 hour and analyzed for VLRB N8 reactivity and HLA-I expression levels after 72 hours (red open histograms). Antibody binding obtained without stimulation is depicted in blue open histograms, and filled gray histograms represent the VLR4 and isotype control experiments. (B) Time course of BJAB response after stimulation with PMA and ionomycin. VLRB N8 binding normalized to negative control VLR4 (top), HLA-I expression normalized to isotype-matched control antibodies (middle), or the ratio of VLRB N8 relative to HLA-I (bottom) is shown. Statistical significance was determined using a multiple t test with Holm-Sidak post test. (C) BJAB cells were treated with the indicated stimuli, and VLRB N8 binding and HLA-I expression levels were assessed as in (A). Induction of VLRB N8 was determined by normalizing VLRB N8/HLA-I ratios to the corresponding unstimulated controls. Bars indicate means ± SD (n = 4). Statistical significance was determined using one-way analysis of variance (ANOVA) with Dunnett’s post test. For comparison, the VLRB N8 signals following PMA and ionomycin treatment are included in the graphic for anti-Ig responses (open bars). (D) Induction of VLRB N8 binding to BJAB cells following costimulation with anti-Ig and IFN. Induction of VLRB N8 reactivity was assessed as in (C). Statistical significance was determined using one-way ANOVA with Tukey’s post test. Statistically significant differences of P < 0.05 are indicated by *P < 0.05, **P < 0.01, and ***P < 0.001.

  • Fig. 5 VLRB N8 recognizes a tyrosine sulfation–dependent antigen on HLA-I.

    (A) KMS-11 cells were cultured in the presence of the indicated concentrations of NaClO3 for 48 hours followed by flow cytometric assessment of VLRB N8 and HLA-I reactivity. A representative experiment is depicted in the top panel, and VLRB N8/HLA-I ratios from five independent experiments are shown in the bottom bar diagram, depicted as means ± SD. Statistical significance was determined using one-way ANOVA with Dunnett’s post test (n = 5). (B) Inhibition of VLRB N8 recognition of HLA-I on BJAB cells following PMA and ionomycin stimulation. Cells were stimulated for 1 hour with PMA and ionomycin, and VLRB N8 and HLA-I binding were assessed following a 36-hour culture with the indicated concentrations of NaClO3. Means ± SD of VLRB N8 signals normalized to HLA-I are shown. Statistical significance was determined using two-way ANOVA test with Dunnett’s post test (n = 4). (C) shRNA-mediated down-regulation of transduced BJAB cells was verified by qRT-PCR. Transcript levels of TPST1 (left) and TPST2 (right) of the indicated cell populations are depicted as means ± SD (n = 3). Statistical significance was determined using a Student’s t test. (D) shRNA-transduced BJAB cells were stimulated with anti-Ig (20 μg/ml), followed by the assessment of VLRB N8 and anti–HLA-I recognition. Numbers indicate the mean fold induction of HLA-I normalized VLRB N8. Statistical significance for induced VLRB N8 binding was determined using a one-way ANOVA test with Tukey’s post test (n = 9). (E) Tyrosine sulfation of HLA-I following antigen receptor engagement. A representative autoradiogram (left) of anti–HLA-I immunoprecipitates of unstimulated and stimulated BJAB cells and the quantitation (right) of six independent experiments are shown. 35SO4 incorporation is shown with arbitrary units (AU). Statistical significance was determined using paired Student’s t test (n = 6). (F) TPST1 and TPST2 transcript analysis of tonsillar B cell populations. Means ± SD of qRT-PCR of TPST1 or TPST2 normalized to HPRT from five independent tonsil specimens are shown. Statistically significant differences of P < 0.05 are indicated by *P < 0.05, **P < 0.01, ***P < 0.001; n.s., P > 0.05.

  • Table 1 Mass spectrometric analysis of VLRB N8 immunoprecipitates.

    Displayed are identified proteins that remain following elimination of sequences associated with intracellular molecules and with VLR4 negative control coimmunoprecipitates. Mw, molecular weight.

    RankIdentified proteinsAccession no.Mw (kDa)
    1HLA class I histocompatibility antigen, A-24 α chain1A24_HUMAN41
    2HLA class I histocompatibility antigen, B-55 α chain1B55_HUMAN40
    3Galectin-1LEG1_HUMAN15
    4Transferrin receptor protein 1TFR1_HUMAN85
    5Protein S100-A9S10A9_HUMAN13
    64F2 cell surface antigen heavy chainB4E2Z3_HUMAN56
    7UbiquitinB4DV12_HUMAN17
    8HLA class II histocompatibility antigen, DRB1-16 β chain2B1G_HUMAN30
    9HLA class I histocompatibility antigen, Cw-14 α chain1C14_HUMAN41
    10β2-microglobulinB2MG_HUMAN14
    11HLA class I histocompatibility antigen, B-51 α chain1B51_HUMAN41
    12HLA class II histocompatibility antigen, DR α chainDRA_HUMAN29

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/4/11/eaar7653/DC1

    Fig. S1. VLRB N8 binding to cell lines does not correlate with HLA-I cell surface expression levels.

    Fig. S2. VLRB N8 immunoprecipitates a prominent 42-kDa protein antigen.

    Fig. S3. Immunoprecipitation of HLA-I with VLRB N8.

    Fig. S4. Recognition of HLA-I by VLRB N8 on Bmem and PCs is independent of HLA-I expression levels.

    Fig. S5. Gating strategies for evaluation of lymphocyte populations from blood and tonsil.

  • Supplementary Materials

    This PDF file includes:

    • Fig. S1. VLRB N8 binding to cell lines does not correlate with HLA-I cell surface expression levels.
    • Fig. S2. VLRB N8 immunoprecipitates a prominent 42-kDa protein antigen.
    • Fig. S3. Immunoprecipitation of HLA-I with VLRB N8.
    • Fig. S4. Recognition of HLA-I by VLRB N8 on Bmem and PCs is independent of HLA-I expression levels.
    • Fig. S5. Gating strategies for evaluation of lymphocyte populations from blood and tonsil.

    Download PDF

    Files in this Data Supplement:

Navigate This Article