The lupus susceptibility interval Sle3/5 confers responsiveness to prolactin in C57BL/6 (B6) mice and hyperprolactinaemia induces a lupus-like phenotype in B6.Sel3/5 mice. In this study, the immunostimulatory effects of prolactin in B6 mice containing the Sle3 portion of the Sel3/5 interval (B6.Sle3 mice) were dissected. Because of the Sle3 interval's involvement in activation of myeloid cells, the effect of dendritic cells (DCs) from prolactin-treated B6.Sle3 mice on the phenotype of B6 mice was also evaluated. B cells from prolactin-treated B6 and B6.Sle3 mice and from B6 recipients of prolactin-modulated DCs from B6.Sle3 mice were tested for DNA-reactivity and resistance to B cell receptor (BCR)-mediated apoptosis. The expression of co-stimulatory molecules on lymphocytes and myeloid cells was also evaluated. In prolactin-treated B6.Sle3 mice, transitional type 2 B cells increased while type 1 B cells decreased as a consequence of prolactin-induced resistance to BCR-mediated apoptosis leading to the survival of DNA-reactive B cells. Follicular B cells from prolactin-treated mice expressed increased levels of CD40, B7·2 and IAb, and DCs and monocytes had higher levels of CD44 and B7·2 than placebo-treated mice. Adoptive transfer of DCs from prolactin-treated B6.Sle3 mice to B6 recipients demonstrated the intrinsic ability of prolactin-modulated DCs to induce a development of lupus-like characteristics in B6 mice. Based on these results, prolactin accelerates the breakdown of immune tolerance in B6.Sle3 mice by promoting the survival, maturation and activation of autoreactive B cells, DCs and macrophages.
Systemic lupus erythematosus (SLE), the prototypical autoimmune disease, affects predominantly women and occurs with greater frequency among African Americans, Asians and Hispanics than among Caucasians, suggesting that both gender-associated factors and genetic factors are involved in its pathogenesis. Because female preponderance is especially evident during the reproductive years, when the female-to-male ratio among SLE patients reaches a striking peak of 9:1, female sex hormones are thought, at least in part, to underlie the gender bias in lupus. About 25–30% of SLE patients have mild-to-moderately increased serum prolactin levels [1-3], with the degree of hyperprolactinaemia correlating with global lupus activity or specific organ involvement in a significant number of these patients [4-7]. A review of five published clinical studies concluded that prolactin is related to lupus activity. However, it was noted that the low statistical power of the studies prevented establishing a formal causal relationship . In addition, murine studies have shown that increased serum prolactin levels accelerate disease onset and cause early mortality in lupus-prone NZB/W F1 mice [9, 10]. We have shown that mild hyperprolactinaemia similar to that seen in SLE patients can break tolerance and induce a lupus-like syndrome in strains of mice that are not predisposed to the disease [11, 12]. Sustained mild to moderately increased serum prolactin levels induce the development of lupus in mice with a BALB/c, but not in those with a C57Bl/6 (B6) genetic background. In the latter strain, the 40cM lupus-susceptibility interval Sle3/5  confers responsiveness to the immunostimulatory effects of the hormone , indicating that the interplay between genetic and hormonal factors creates a milieu permissible to the breakdown of tolerance and development of a lupus-like phenotype. In this study, we narrow down the pool of Sle3/5 genes that are responsible for the immunostimulatory effects of prolactin by demonstrating that the lupus-susceptibility locus Sle3 alone (24cM) is sufficient to confer responsiveness to the hormone. Treatment with prolactin accelerates the onset of autoimmune characteristics in B6.Sle3 mice by affecting both lymphoid and myeloid cells. We also show that prolactin-modulated dendritic cells (DCs) from B6.Sle3 mice are sufficient to induce the development of lupus-like autoreactivity in wild-type B6 mice.
Materials and methods
Six-week-old female B6.Sle3 mice and B6 mice were used in the experiments. B6.Sle3 mice are homozygous for a 24-cM congenic interval of the NZM2410 allele of Sle3 and were a kind gift from Dr E. Wakeland (University of Medicine, Texas Southwestern Medical Center, Dallas, TX, USA). B6.Sle3 mice were bred at the animal facility at Albert Einstein College of Medicine (Bronx, NY, USA). B6 mice were purchased from Taconic (Hudson, NY, USA).
Prolactin and granulocyte–macrophage colony-stimulating factor (GM-CSF) treatments
One hundred μg (in 100 μl saline solution) of prolactin from sheep pituitary gland tissue (Sigma-Aldrich, St Louis, MO, USA) or placebo (100 μl saline solution) were administered subcutaneously every day to the mice for 4 weeks. The daily treatment with 100 μg of prolactin yields a two- to threefold increase in the serum prolactin level ; these mild elevations in serum prolactin levels are similar to the prolactin levels observed in SLE patients with hyperprolactinaemia [1-3, 8].
GM-CSF injections have been shown to moderately increase the numbers of DCs in the spleen . However, as a daily dose of 1 μg GM-CSF (Gibco, Grand Island, NY, USA) per mouse injected in the presence of prolactin was not able to produce a sufficient number of DCs for adoptive transfer (5–10 × 106 DCs), we opted to implant subcutaneously GM-CSF-expressing B16-F10 melanoma cells which have been shown to have a limited effect on the immune response [16, 17]; 107 cells per mouse were injected into mice treated previously with daily injections of prolactin or saline for 2 weeks. After the implantation, the mice were treated with saline solution or prolactin for an additional 2 weeks. The percentage of DCs in the mice increased from 1 to 15% and to 30% in the placebo- and prolactin-treated mice, respectively.
Spleens from prolactin- and prolactin/GM-CSF-treated B6.Sle3 mice were ballooned with a 1-ml syringe containing collagenase D solution (400 U/ml) (Roche Diagnostics, Mannheim, Germany) and ethylenediamine tetraacetic acid (EDTA) (2 mM) (Gibco) in phosphate-buffered saline (PBS). The obtained splenocytes were incubated for 30 min at 37°C. Then, DCs were purified by using anti-mouse CD11c antibody conjugated to magnetic beads (Miltenyi Biotech, Auburn, CA, USA) resuspended in buffer [0·5% bovine serum albumin (BSA); Sigma, St Louis, MO, USA] and 2 mM EDTA in PBS. Single-cell suspensions were applied onto a magnetic affinity cell sorting (MACS) column and placed in a magnetic field (MAC separator; Miltenyi Biotech). Unlabelled cells passed through the column, the column was washed and then removed from the separator and magnetically labelled cells were flushed out. The cell suspension was run again through a new column to obtain a purity of ∼100%, which was verified by flow cytometry. CD11b and CD11c staining identified myeloid DCs (CD11bhighCD11chigh), macrophages (CD11bintermediate CD11cintermediate) and neutrophils (CD11bhighCD11c−).
Adoptive transfer of DCs
Five to 10 × 106 cells from placebo- or prolactin-treated B6.Sle3 mice were injected intravenously into 6-week-old B6 mice. Three injections were performed each week for 3 weeks. Splenic cells from the B6 recipients were collected 60 days after the adoptive transfer of DCs.
Phenotyping of immune cells
Identification of B cell subsets and expression of co-stimulatory and activation molecules on myeloid and lymphoid cells
Splenic cells were obtained from B6 and B6. Sle3 mice treated with prolactin or recipients of DCs from GM-CSF placebo-treated or GM-CSF prolactin-treated mice. Cells were stained on ice for 30 min with antibodies to CD19 to identify B cells, CD93 to identify transitional B cells and CD21 and CD23 to determine T1 and T2 transitional B cell subsets (CD19+ CD93+, CD21−, CD23− and CD19+ CD93+, CD21+, CD23+, respectively). Staining for CD3, CD4 and CD8 was performed to identify helper and cytotoxic T cells, whereas antibodies to CD11b and CD11c were used to detect DCs (CD11bhighCD11chigh), macrophages (CD11bint CD11cint) and neutrophils CD11bhighCD11c−) Also, staining for co-stimulatory molecules (CD40, CD40L and B7·2) and differentiation and activation markers (CD25, CD69 and CD44) was performed. The fluorochromes conjugated to the primary antibodies were fluorescein isothiocyanate (FITC), phycoerythrin (PE), peridinin chlorophyll (PerCP) cyanin (Cy)5·5, PE-Cy7 and allophycocyanin (APC). Biotin-conjugated antibodies were detected by Pacific Blue-conjugated streptavidin. The cells were washed after each incubation, and then fixed in 2% paraformaldehyde. Samples were analysed in an LSR II flow cytometer (BD Bioscience, San Diego, CA, USA). All staining antibodies were purchased from BD Bioscience except CD93, which was obtained from eBioscience (San Diego, CA, USA).
Splenic cells from prolactin-treated B6.Sle3 mice and B6 recipients of DCs from B6.Sle3 mice treated with GM-CSF/prolactin and respective controls were cultured in medium containing 10 μg/ml of anti-IgM (Southern Biotech, Birmingham, AL, USA) for 16–18 h at 37°C. Then, the cells were washed and stained on ice for the surface markers CD19, CD93, CD21 and CD23 for identification of B cell subsets, as described above. After that, the cells were washed and incubated with annexin V-FITC (BD Bioscience, San Diego, CA, USA) for 15 min at room temperature. The cells were analysed by the LSR II (BD Bioscience) within an hour in the presence of TO-PRO-3 iodide, a viability probe for discrimination of necrotic cells (Invitrogen, Carlsbad, CA, USA).
Enzyme-linked immunospot (ELISPOT) assay
Splenocytes from placebo- and prolactin-treated B6.Sle3 mice and B6 recipients of DCs from placebo and prolactin-treated B6.Sle3 mice and control mice were cultured in dsDNA-coated plates for 48 h with media alone [RPMI-1640 plus 5% fetal calf serum (FCS)] or in media containing anti-CD40 (10 μg/ml; BD Bioscience) and interleukin (IL)-4 (300 U/ml; Sigma-Aldrich) or anti-CD40 and anti-mouse IgG (10 μg/ml; Jackson ImmunoResearch, West Grove, PA, USA). The cells were washed and then incubated with biotin–goat anti-mouse IgG (1 : 500) and alkaline phosphatase (AP)-streptavidin (1 : 1000; Southern Biotechnology) overnight at 4°C and 1 h at room temperature, respectively. AP substrate and 5-bromo-4-chloro-1H-indol-3-yl) dihydrogen phosphate (BCIP) (Amresco, Solon, OH, USA) were added to generate spots in a colorimetric reaction. The spots were counted under a dissecting microscope.
Formalin-fixed, paraffin-embedded kidney sections from placebo- and prolactin-treated mice were deparaffinized in alcohol and then stained with biotinylated anti-mouse IgG antibody, as described previously  The alkaline phosphatase ABC kit (Vector Laboratories, Burlingame, CA, USA) was used to develop the slides.
The evaluation of mean values, standard deviations and statistical significance (P-values) was performed by using Student's two-tailed t-test. Absolute numbers were calculated from the FACS percentage data and converted to log10 to calculate P-values by Student's two-tailed t-test.
DNA-reactive B cells
B6.Sle3 mice have been derived from NZM2410 mice. Sle3 locus is on mid-chromosome 7, and confers the development of age-dependent low levels of anti-nuclear antibodies. To determine whether prolactin provides an additive immunostimulatory effect on B cells, young 6-week-old B6.Sle3 mice were treated with prolactin for 4 weeks. Prolactin-treated mice displayed a significantly higher number of DNA-reactive B cells upon exposure to anti-CD40 and IL-4 than did placebo-treated mice (P = 0·01) (Fig. 1a). The same prolactin treatment did not induce an increased number of DNA-reactive B cells in B6 mice .
B6.Sle3 mice treated with prolactin developed IgG deposition in the glomeruli (Fig. 1b), indicating that hyperprolactinaemia leads to the production of high-affinity nephritogenic anti-DNA antibodies. Thus, the treatment with prolactin and the lupus susceptibility locus Sle3 separately do not induce lupus in B6 mice, but together they can break B cell tolerance and induce an autoimmune phenotype.
B cell maturation during the transitional stage
Treatment with prolactin induced changes in the maturation pattern of transitional T1 (CD19+CD93+CD21−CD23−) and T2 (CD19+CD93+CD21+CD23+) B cells in B6.Sle3 mice. The T1 B cell subset was decreased and T2 subset was increased in prolactin-treated mice when compared to placebo-treated B6.Sle3 mice (P = 0·01 and 0·003, respectively) (Fig. 1c,d). Similarly, the T1/T2 ratio was lower in B6. Sle3 mice treated with prolactin than in the placebo-treated mice (data not shown). T1 and T2 subsets in B6.Sle3 mice were also compared to the transitional subsets in congenic wild-type B6 mice that are not affected by treatment with prolactin; there was no difference in the T1/T2 ratio (data not shown).
B cell receptor (BCR)-mediated apoptosis of transitional T1 B cells in B6.Sle3 mice
To understand prolactin-mediated alterations of the maturation pattern of transitional B cells (decreased numbers of T1 and increased numbers of T2 B cells in B6.Sle3 mice treated with prolactin), the effect of prolactin on anti-IgM-induced apoptosis, a mechanism for negative selection of T1 B cells [18, 19] was studied. Prolactin protected T1 B cells from BCR-mediated apoptosis, but in placebo-treated mice the T1 B cell subset under the same conditions displayed a higher number of apoptotic cells (P = 0·002, Fig. 1e); a flow-cytometry histogram shows the inhibition of apoptosis of T1 B cells from prolactin-treated B6.Sle3 mice (Fig. 1f, dark line).
Activated T cells
B6.Sle 3 mice show an elevated CD4/CD8 ratio  and the treatment with prolactin increased this ratio further (P = 0·01, Fig. 2a). In addition, prolactin induced higher numbers of CD4 T cells expressing CD40L (P = 0·05, Fig. 2b); activated helper T cells are recognized as a major driver of autoimmunity in lupus.
Over-expression of activation and co-stimulatory molecules; CD40-, B7·2- and MHC II-expression on B cells of B6.Sle3 mice
The effect of prolactin on the expression of a variety of activation and co-stimulatory markers was compared between prolactin- and placebo-treated B6.Sle3 and B6 mice. In B6.Sle3 mice, treatment with prolactin increased the numbers of follicular (Fo) B cells expressing CD40, and B7·2 (P = 0·03) and induced an increased expression of IAb (P = 0·01) on Fo B cells; all three molecules are necessary for the B–T cell interactions, and their up-regulation may lead to enhanced antibody responses (Fig. 3a–d). However, CD44 and CD69 were not affected by prolactin in B6.Sle3 mice (data not shown).
Enhanced macrophage activation and maturation of DCs
B6.Sle3 mice contain more mature and activated macrophages, DCs and neutrophils than do B6 mice [21, 22]. Treatment with prolactin further increased the maturation and activation status of DCs and macrophages, as indicated by the over-expression of B7·2 (P = 0·01 and 0·04, respectively) and CD44 (P = 0·01 for both cell types) on their surface (Fig. 3e–h). Treatment with prolactin had no effect on the activation status of neutrophils in B6.Sle3 mice, nor did it affect the expression of B7·2 and CD44 on myeloid cells in B6 mice (data not shown).
Prolactin-modulated- DCs from B6.Sle3 mice exert immunostimulatory effects in B6 mice
DNA-reactive B cells
Given that the Sle3 locus generates mature DCs and that the carrying phenotype is intrinsic to APCs , we speculated that prolactin-modulated DCs would be able to induce a lupus-like phenotype when they are transferred adoptively in B6 mice. The B6 recipients of DCs from prolactin-treated B6.Sle3 mice developed a similar B cell phenotype to that of B6.Sle3 mice treated with prolactin. Two of the three recipients of prolactin-modulated DCs showed mild–moderate IgG deposits, and one recipient showed only mild deposits. The recipients of prolactin-modulated DCs had more DNA-reactive B cells than did the recipients of DCs from placebo-treated B6.Sle3 mice (P = 0·02) (Fig. 4a).
B cell maturation pattern
The recipients of prolactin-modulated DCs displayed an increased number of transitional T2 B cells (P = 0·05). T1 B cells were decreased in mice which received prolactin-modulated DCs when compared to the recipients of DCs from placebo-treated mice; although the difference was not significant, there was a decrease of the T1/T2 ratio in the prolactin treated DC group (Fig. 4b, c).
BCR-induced apoptosis of T1 B cells
A higher percentage of apoptotic T1 B cells was observed in the recipients of placebo DCs than in the recipients of DCs from prolactin-treated B6.Sle3 mice. The difference in apoptosis between untreated and anti-IgM stimulated T1 B cells from placebo mice was significant (P = 0·004); in contrast, the difference in the prolactin treated group was not significant (Fig. 4d,e). This observation indicates that prolactin-modulated DCs are initiating the altered B cell maturation pattern depicted by the decreased T1/T2 ratio in the B6 recipients as a consequence of the inhibition of apoptosis. Similar T1/T2 alterations were observed in B6.Sle3 mice which received a 4-week prolactin treatment, as shown in Fig. 1c,d.
B cell activation markers
The B6 recipients of DCs from prolactin-treated B6.Sle3 mice expressed higher levels of IAb on Fo B cells than the Fo B cells from recipients of placebo DCs (P = 0·01). The fact that B cells express more IAb enables them to be more efficient APCs, which may contribute to the development of a lupus-like phenotype (Fig. 5a,b). As shown previously, Sle3 DCs carry intrinsic abnormalities that induce certain autoimmune phenotypes associated with lupus . DCs from prolactin-treated B6.Sle3 mice were able to transfer the prolactin-mediated lupus characteristics to B6 mice, which do not develop a lupus-like disease when treated with prolactin .
Twenty-five to 30% of patients with SLE have increased serum prolactin, and in a significant number of these patients prolactin levels correlate with lupus activity and/or system organ involvement . These data raised interest in the role of prolactin in lupus and have led to numerous murine studies [1, 10-12, 23]. It has been shown that treatment with prolactin (100 μg/day) exerts immunostimulatory effects in BALB/c mice by altering BCR-mediated tolerance-induction during B cell maturation and, consequently, ‘rescuing’ autoreactive B cells from deletion . The same prolactin treatment did not affect B cells in B6 mice , but the NZM240-derived Sle3/5 genetic interval provided responsiveness to prolactin in these mice. B6.Sle3/5 mice developed a lupus-like phenotype when treated with 100 μg prolactin daily for 4 weeks , indicating that the Sle3/5-encoding proteins confer sensitivity to the hormone. The work presented here demonstrates that a portion of the Sle3/5 interval, the lupus susceptibility locus Sle3, is sufficient for conferring responsiveness to prolactin and prolactin-modulated DCs from B6.Sle3 mice can induce the development of DNA-reactivity in B6 mice.
Using B6.Sle3 mice , we endeavoured to expand our evaluations of the immune effects of prolactin in order to understand more clearly their implications in the pathogenesis of lupus. The Sle3 locus induces minor immunological changes in B6 mice, such as a slightly increased CD4/CD8 ratio and activated myeloid cells. Myeloid cell hyperactivity results in the secondary activation of lymphocytes leading to an age-dependent development of anti-nuclear antibodies when B6.Sle3 mice reach 9–12 months [20-22]. The Sle3 mice require additional factors, such as the presence of the Sle1 locus for progression to full-blown lupus .
In the current study, prolactin is defined as an immunostimulator that exacerbates and accelerates the Sle3 phenotype towards lupus pathology when the mice are only 2·5–3 months old. A doubling of serum prolactin levels hastens the occurrence of IgG-secreting DNA-reactive B cells in young B6.Sle3 mice, which indicates that prolactin enhances the previously reported effectiveness of Sle3 in mediating Ig class-switching to IgG . These IgG anti-dsDNA antibodies are pathogenic, and deposit in the kidneys of B6.Sle3 mice.
Increased prolactin levels exert a protective effect on B cell maturation from the transitional T1–T2 stage, allowing higher numbers of autoreactive B cells to escape negative selection at the T1/T2 interphase [16,17], which is a check-point for B cell tolerance induction [25-28].
Autoreactive T2 cells ‘rescued’ by prolactin in B6.Sle3 mice mature further into fully functional B cells, which differentiate into anti-dsDNA antibody-secreting plasma cells capable of initiating the development of a lupus phenotype.
Also, prolactin induced an over-expression of B7·2 and MHC class II on B cells and DCs, and also CD40 on B cells. Murine and human data indicate that DCs can provide B cells with CD40-dependent survival signals and CD40-independent proliferative signals , both of which could play a role in DNA reactivity observed in prolactin-treated B6.Sle3 mice. Reports of increased B7·2 expression on B cells of SLE patients  indicate that prolactin-induced up-regulation of B7·2 may play an important role in the acceleration of autoimmunity in the B6.Sle3 murine model.
Based on the reports that CD40-over-expressing B cells can become plasma cells in the presence of DCs without T cell intervention [31, 32], prolactin-induced up-regulation of CD40 seems to be an important contributor to the accelerated autoantibody production in B6.Sle3 mice. Thus, prolactin-mediated over-expression of CD40 on B cells with or without CD40L expression on T cells might be a contributory factor for the increased differentiation of B cells into autoantibody-secreting plasma cells [33, 34].
Nine to 12-month-old B6.Sle3 mice exhibit myeloid cell hyperactivity that is intrinsic to the Sle3 genes [22, 25]. In-vivo treatment with prolactin up-regulates the activation status of myeloid cells in 2–3-month-old B6.Sle 3 mice by increasing the expression of B7·2 and CD44 molecules on APCs. Prolactin-induced over-expression of B7·2 shows that the hormone can accelerate the maturation status of DCs and increase their co-stimulatory properties, and also enhances their capacity to adhere to the extracellular matrix components  in target tissues, potentially initiating the inflammatory organ-specific injuries observed in lupus [36, 37].
The adoptive transfer of DCs from prolactin-treated B6.Sle3 mice into B6 recipients induced significantly more activated MHC class II over-expressing B cells and DNA-reactive B cells than did the adoptive transfer of DCs from placebo-treated B6.Sle3 mice. A recent study reported a link between the DC hyperactivity in B6.Sle3 mice and the X chromosome-encoded gene IL-1 receptor-associated kinase-1 (IRAK-1) , but this association has been questioned by others . It seems that because of the high linkage disequilibrium in the Xq28, both IRAK1 and methyl–cytosine–phosphate–guanosine (CpG)-binding protein 2 (MECP2) could be considered candidate genes for lupus.
The DC-mediated elevation of the number of DNA-reactive B cells in B6 mice could be accomplished by direct interaction between DCs and B cells or indirectly by secretion of cytokines. CD40-over-expressing DCs, known for their role in the early stages of lupus , can help autoreactive immature B cells to escape negative selection at the T1/T2 interphase. The survival of autoreactive specificities can be facilitated by prolactin-modulated DCs presenting unprocessed autoantigen to B cells . Also, prolactin-modulated DC may induce differentiation of autoreactive B cells into autoantibody-secreting plasma cells by secretion of cytokines such as IL-6, IL-12 and B lymphocyte stimulator (BlyS) . However, the lower anti-DNA reactivity in B6 recipients of prolactin-modulated DCs from B6.Sle3 mice than the DNA reactivity observed in prolactin-treated B6.Sle3 mice underscores the importance of the direct effects of prolactin on autoreactive B cells such as the over-expression of CD40 . Down-regulation of CD40 on B cells occurs in SLE patients after successful treatment with rituximab .
In conclusion, in young 2–3-month-old B6.Sle3 mice, mild hyperprolactinaemia, similar to that observed in almost 30% of SLE patients, induces a lupus-like illness characterized by activated myeloid and lymphoid cells, an increased number of DNA-reactive B cells and autoantibody deposits in the kidneys. The effects of prolactin on B cells are mediated partially by accelerated maturation of DCs which can activate the self-reactive CD40-over-expressing B cells directly or indirectly via CD40L+CD4+T cells. Increased CD40 expression on B cells seems to be a direct effect of prolactin, while the over-expression of MHC II on B cells could be induced by prolactin-modulated DCs. In addition, DCs from prolactin-treated B6.Sle3 mice have the capacity to induce autoreactivity in B6 mice, which do not respond to treatment with prolactin.
This study was supported by the The Lupus Foundation. We would like to thank Dr E. Wakeland (University of Texas Southwestern Medical Center, Dallas) for his generous gift of the B6.Sle3 mice. We also thank Gabriel Rosenfeld and Yu Zhang for their technical help, and Joel Correa da Rosa from The Rockefeller University, CCTS, for the statistical analysis.
The authors affirm that they have no actual or potential conflicts of interest to disclose.