Poly(ADP-ribose) polymerase-1 (Parp-1)-deficient mice demonstrate abnormal antibody responses

Mice and immunizations

Parp-1^−/− mice were produced as previously described ²⁷ and maintained in specific pathogen-free (SPF) conditions. To test T cell-dependent responses, 8–12-week-old animals were immunized intraperitoneally with 100 μg of alum-precipitated nitro-phenol-conjugated keyhole limpet hemocyanin (NP-KLH) (Biosearch Technologies, Novato, CA) with 10⁹ heat-killed Bordetella pertussis (Lee Laboratories, Grayson, GA). For T cell-independent responses, mice were immunized intraperitoneally with 100 μg of alum-precipitated NP-Ficoll (Biosearch Technologies). Serum was collected from tail bleeds before immunization and 12 days after immunization.

[³H]Thymidine incorporation

Splenic B cells were seeded in 96-well plates (2 × 10⁵ cells/well) and stimulated with lipopolysaccharide (LPS) from Escherichia coli 0111:B4 (Sigma, St Louis, MO) at 15 μg/ml. After 60 hr of culture cells were pulsed with [³H]thymidine for 12 hr. Incorporated radioactivity was measured by liquid scintillation counting. Experiments were carried out in triplicate.

Cell culture

B cells were purified from total splenic B cells by density gradient centrifugation followed by incubation with anti-Thy1.2 antibody (Sigma) and rabbit complement (Cedarlane, Burlington, ON, Canada). B cells were cultured at 10⁶ cells/ml in RPMI-1640 with 10% fetal calf serum (FCS). Splenic B cells were cultured with combinations of LPS and growth factors to measure class-switched cells. LPS (Sigma) at 50 μm was used alone or in combination with 10 ng/ml interleukin (IL)-4 (R&D Systems, Minneapolis, MN), 10 U/ml mouse interferon (IFN)-γ (R&D Systems) or 1 ng/ml human transforming growth factor (TGF)-β (R&D Systems).³⁰ Cultures were set up in triplicate and supernatants were harvested after 5 days.

Flow cytometry

Splenic B cells (10⁶) were incubated with antibody in a volume of 100 μl of phosphate-buffered saline (PBS) with 1% bovine serum albumin (BSA) for 30 min on ice. After two washes with PBS, cells were resuspended in 500 μl of PBS and analysed by flow cytometry (FACScalibur; BD Biosciences, San Jose, CA). All antibodies were purchased from (BD Biosciences).

Enzyme-linked immunosorbent assays (ELISAs)

Ninety-six-well plates (Falcon, BD Biosciences) were coated with antibody (anti-IgM, anti-IgA, anti-IgG1, anti-IgG2a, anti-IgG2b and anti-IgG3 were all purchased from BD Biosciences) in PBS at 5 μg/ml overnight at 4°. After blocking with PBS and 1% BSA, dilutions of mouse serum or culture supernatant were added, and again plates were incubated overnight at 4°. The following day biotinylated anti-immunoglobulin was added and plates were incubated for 1 hr at room temperature. After washing the plates were incubated with streptavidin-horseradish peroxidase (Dako, Glostrup, Denmark) for 1 hr at room temperature. Following washing, ortho-phenylene diamine substrate (Sigma) was added and the colour reaction was stopped with 3 N hydrochloric acid. Both unconjugated antibodies specific for each mouse isotype for coating plates and biotinylated secondary antibodies were obtained from BD Biosciences.

To detect antigen-specific antibody-secreting cells, Multiscreen-HA enzyme-linked immunosorbent spot-forming cell assay (ELISpot) plates (Millipore, Billerica, MA) were coated with NP-BSA (20 μg/ml) overnight at 4°, washed and blocked with PBS and 1% BSA. Splenic B cells (10⁵) in 100 μl of medium were added to the plates and incubated overnight at 37° in 5% CO₂. Goat anti-mouse IgG₁ conjugated to horseradish peroxidase (Southern Biotechnology, Birmingham, AL) was added after washing with PBS/0·05% Tween 20, and plates were incubated for 4 hr at room temperature. After further washes 100 μl of substrate 3-amino-9-ethyl carbozole (AEC) (Sigma) was added per well and allowed to stand for 15–30 min. AEC was then removed and the plates were extensively washed with water and spots counted under the microscope.

Real-time semiquantitative polymerase chain reaction (PCR)

RNA was extracted and purified utilizing an RNeasy kit (Qiagen, Hilden, Germany). cDNA was produced from 2 μg of RNA with Omniscript (Qiagen) and an oligo dT primer (New England Biolabs, Ipswich, MA). For detection of mouse activation-induced deaminase (AID), one-sixth of the product was mixed with Taqman primer/probe mix Mm00507774_m1 (Applied Biosystems, Foster City, CA) and 2× Taqman reaction buffer (Applied Biosystems). PCR was carried out in an ABI7700 sequence detector (Applied Biosystems).

PARP assay

PARP activity in freshly isolated splenic B cells was determined as previously described.³¹ Reactions contained 0·1 μg of high-purity recombinant PARP protein (Alexis Biochemicals, San Diego, CA) as a positive control. Reactions were incubated at 25° for 30 min in assay buffer (0·1 ml) containing 50 mm Tris-HCl, pH 8·0, 4 mm MgCl₂ and 0·2 mm dithiothreitol, with or without 200 ng of activated DNA (Sigma). In radioactive reactions, 1·3 μm [³²P]NAD⁺ (4 μCi; PerkinElmer Life Sciences, Waltham, MA) was used.

Lymphocyte development in Parp-1-deficient mice

To determine the effects of Parp-1 deficiency on mature B-cell development we compared splenic B-cell populations of mutant and wild-type mice by flow cytometry (Fig. 1) using antibodies specific for the pan-B cell marker B220 (CD45R) and the pan-T cell marker CD3. Whilst the proportion of B cells in Parp-1-deficient mice was lower than in normal mice (Fig. 1a) the absolute numbers of B cells were similar (Fig. 1c) because the total number of B cells and spleen weight were slightly greater in the knockout animals. Absolute and relative numbers of T cells were significantly increased in Parp-1^−/− animals (Fig. 1a and c). There was no difference in the proportions of CD4-and CD8-expressing T cells (Fig. 1b).

We next examined the composition of the various B-cell subsets. Combined expression of IgM and IgD defined mature B cells, with IgM⁺IgD^low cells representing immature B cells and IgM⁻IgD⁻ cells being isotype-switched B cells (Fig. 2a). Parp-1^−/− mice had similar proportions of these three subsets to wild-type mice. An alternative analysis based on differential expression of the surface markers CD21 and CD23 divided splenic B cells into naïve (CD21⁻CD23⁻), follicular (CD21⁺CD23^-+) and marginal zone populations (CD21⁺CD23^low) (Fig. 2b). By this analysis there were significantly (P = 0·03, Mann–Whitney test) increased numbers of marginal zone B cells, decreased numbers of naïve B cells (Fig. 2c) and normal numbers of follicular cells. Bone marrow was stained with anti-B220 and IgM to identify early B-cell compartments (Fig. 3a). Numbers of B220^high and B220^low cells were the same in Parp-1-deficient and wild-type mice.

Long-lived plasma cells are one of the products of the germinal centre reaction, and migrate from the lymph nodes or spleen to the bone marrow. Accordingly we analysed bone marrow plasma cells defined as B220⁻IgM⁻CD138⁺.³² In Parp-1-deficient mice, 0·35 ± 0·05% (mean ± standard error of the mean) of B220⁻IgM⁻ cells expressed CD138 whereas in normal mice this population was 1·58 ± 0·25%. Therefore, bone marrow plasma cells were reduced by about 70% in Parp-1^−/− mice (P = 0·008; Mann–Whitney test) (Fig. 3b), suggesting defective production or maintenance of this population.

Serum immunoglobulin levels

Serum immunoglobulins in unimmunized mice were measured by ELISA (Fig. 4a). Concentrations of IgM, IgG1 and IgG3 in Parp-1^−/− mice were within the normal range, but concentrations of IgA and IgG2a were increased and IgG_2b levels were decreased.

Class-switch recombination can be activated in vitro by various stimulators. ^1,30 LPS and IL-4 induce IgG1, LPS and IFN-γ induce IgG2a and LPS, TGF-β and IL-5 induce switching to IgA. Accordingly we tested the effects of these combinations on splenic B cells from Parp-1-deficient animals (Fig. 4b).

There were no significant differences in IgM or IgG1 but both IgG2a (P = 0·041; Mann–Whitney test) and IgA (P = 0·009; Mann–Whitney test) were significantly elevated. Therefore, together the abnormalities in serum immunoglobulin levels and the in vitro induction of class switching suggest that Parp-1 has a role in regulating isotype expression.

Antibody responses

To determine immune responses in Parp-1^−/− mice we first immunized with the T cell-independent antigen NP-Ficoll. Parp-1^−/− mice had normal responses to this antigen (Fig. 5a). To assess T cell-dependent immune responses animals were immunized with NP-KLH and, in contrast to T cell-independent responses, the T cell-dependent responses of Parp-1^−/− mice were markedly reduced (Fig. 5b). In order to measure differences in the size or number of germinal centres following immunization we carried out immunohistochemistry (Fig. 5d). Germinal centres were of similar size in wild-type and Parp-1-deficient animals and the percentages of follicles containing germinal centres were also similar (76 ± 5% in wild-type mice and 71 ± 7% in Parp-1-deficient mice).

Overall, therefore, Parp-1^−/− mice had reduced T cell-dependent immune responses with normal germinal centre size and numbers. Additionally, Parp-1-deficient animals had reduced numbers of anti-NP-producing plasma cells (P = 0·02; Mann–Whitney test) (Fig. 5c).

Whilst abnormal germinal centre function could explain the observed defects in isotype expression, T cell-dependent immune responses and plasma cell numbers, we cannot exclude the possibility of defects in survival or homing of post-germinal centre B cells. In order to try to demonstrate a specific germinal centre defect we measured induction of AID expression in splenic B cells in response to LPS.³³ AID was detectable at day 3 in normal mice and was induced further by day 7. However, in Parp-1-deficient animals AID was not detectable at day 3, and induction at day 7 was markedly reduced as compared with normal mice (Fig. 6a).

[³H]Thymidine incorporation and in vitro IgM secretion in response to LPS

Regulated B-cell proliferation is important for normal antibody responses. We therefore measured proliferative and immunoglobulin secretory responses to LPS. Purified B cells from normal and Parp-1-deficient mice were purified and proliferation was measured. [³H]Thymidine incorporation following 5 days of stimulation with LPS was not significantly different between groups of Parp-1-deficient and wild-type mice (Fig. 6b), demonstrating normal proliferative responses to this stimulant.

Therefore, unlike CD19^−/− and CD40 ligand^−/− B cells, Parp-1^−/− B cells proliferated normally but there was a failure of germinal centre function (Figs 4 and 5).

PARP activity in B cells

To determine whether the phenotype of Parp-1^−/− mice could be attributable to the enzymatic function of Parp-1 or to a non-enzymatic effect we measured poly(ADP-ribosylation) in purified B cells from normal, Parp-1-deficient and Parp-2-deficient mice (Fig. 6c). PARP activity was similar in B cells from normal and Parp-2-deficient animals, but was significantly decreased in Parp-1-deficient animals, suggesting that defects in poly(ADP-ribosylation) have a role in generating abnormalities of class switching and antibody production.