The B cell helper side of neutrophils

Introduction

Neutrophils are usually viewed as terminal, short-lived, innate effector cells that remove microbes and cellular debris at sites of infection or inflammation by mediating phagocytosis and releasing proteolytic enzymes, antimicrobial proteins, and ROS. Over the last decade, a number of studies have shown that neutrophils have a lifespan longer than originally thought and mediate diverse immune functions by releasing a broad array of preformed and newly synthesized mediators, including chemokines and cytokines [1, 2]. These molecules regulate not only the mobilization and function of neutrophils but also the recruitment and activation of monocytes, DCs, NK cells, T cells, and B cells of the innate and adaptive immune systems [1, 2]. In general, neutrophils functionally interact with B cells by binding IgG and IgA, two opsonizing antibody isotypes that amplify microbial clearance by engaging powerful FcγRs and FcαRs on neutrophils [3, 4].

We found recently that human neutrophils are not only eager users but also proficient inducers of IgG and IgA, as a result of their ability to crosstalk with a unique subset of B cells lodged in the MZ of the spleen [5]. Strategically interposed between the circulatory and immune systems, MZ B cells (also known in humans as IgM memory B cells) are innate, antibody-producing lymphocytes that naturally recognize conserved microbial products and self-antigens through poorly diversified BCR (or surface Ig) molecules [6, 7]. Owing to their preactivated state and pronounced innate properties, MZ B cells rapidly mount preimmune (homeostatic) and postimmune (infection-induced) antibody responses to blood-borne antigens, including commensal antigens physiologically translocating from mucosal surfaces to the general circulation [5–8].

Our findings indicate that MZ B cells produce IgM as well as class-switched IgG and IgA antibodies after receiving helper signals from a unique subset of splenic neutrophils that are phenotypically and functionally distinct from circulating neutrophils [5]. This “mini” review will discuss the B cell helper function of splenic neutrophils in the context of recent advances on the mechanisms whereby neutrophils modulate the function of innate and adaptive immune systems.

NEUTROPHILS AS IMMUNOENHANCERS

Growing evidence shows that neutrophils enhance nonspecific innate immune responses by promoting the recruitment, activation, and maturation of monocytes, macrophages, DCs, and NK cells [2, 9]. Neutrophils also enhance specific, adaptive T and B cell responses by facilitating the differentiation of monocytes and DCs to professional APCs [2, 9]. Given the varied immunoenhancing activities of neutrophils, immunodeficient patients with quantitative or functional neutrophil disorders often develop secondary immune dysfunctions that contribute to the onset of recurrent infections [10, 11]. In general, the immunostimulating properties of neutrophils can be ascribed to their ability to produce a broad repertoire of immune mediators with pleiotropic function [1, 2].

In the initial phases of the immune response, neutrophils release the chemokines CCL3, CCL5, and CXCL10, together with the inflammatory cytokines IL-1β, IL-6 (this cytokine has been shown in mice; evidence in humans remains controversial), IL-12, and TNF, as well as a heterogeneous set of granular proteins known as alarmins [1, 12, 13]. In addition to stimulating inflammation, alarmins promote the recruitment of circulating DC precursors and stimulate their progression along a maturation program that converts them into professional APCs with T cell-stimulating capacity [1, 12, 13]. These properties are exemplified by the cationic antimicrobial peptide LL-37, an alarmin that enhances inflammatory T_H1 responses by amplifying the APC activity of DCs [13, 14]. Activated neutrophils increase DC maturation further by releasing TNF, particularly in the context of a contact-dependent crosstalk involving engagement of the integrin CD11b (macrophage antigen-1) and carcinoembryonic antigen-related cell adhesion molecule 1 (or CD66), on neutrophils with the CLR DC-specific ICAM-3-grabbing nonintegrin 1 (or CD209) on DCs [15–17]. This cell-to-cell interaction enhances the conversion of DCs into T cell-stimulating APCs in the presence of TNF production by neutrophils [15, 17].

After undergoing maturation, DCs acquire APC activity and release the inflammatory cytokines IL-12 and TNF-α, which promote the differentiation of monocytes into macrophages, as well as the polarization of naïve CD4⁺ T cells into T_H1 cells [18]. These effector CD4⁺ T cells activate the killing function of macrophages, NK cells, and CTLs by secreting IFN-γ [2]. Neutrophils may further enhance CTL responses after migrating to draining LNs and bone marrow in response to chemotactic signals generated by microbial intruders, including signals from CCR7 ligands [19, 20]. At these sites, antigen-transporting neutrophils not only cross-present exogenous antigens to antigen-specific CD8⁺ T cell precursors of CTLs [20, 21] but also release NK cell/CTL-activating cytokines, such as IFN-γ, albeit this function is still controversial in humans [9, 22].

In addition to favoring the development of NK cell precursors in the bone marrow [11, 23], neutrophils enhance NK cell activation, including IFN-γ production, by delivering contact-dependent and contact-independent signals through the adhesion molecule ICAM-1 and the cytokine IL-18, respectively [24, 25]. Neutrophils further enhance NK cell secretion of IFN-γ by triggering IL-12 production in a subset of DCs expressing slanDCs [25]. IFN-γ from NK cells further augments slanDC release of IL-12, thereby establishing a positive loop that amplifies the activation of NK cells [25]. Conversely, DC-derived IL-12 cooperates with neutrophil-derived IL-18 to increase IFN-γ production by NK cells [24]. When activated by IL-18, NK cells enhance neutrophil survival by releasing GM-CSF [26]. However, it must be noted that activated NK cells can also cause neutrophil apoptosis through a contact-dependent mechanism involving NK expression of the death-inducing molecule Fas ligand and the killer-activating receptor NKp46 [27].

Activated neutrophils further enhance immunity by recruiting T_H1 cells via CXCL9 and CXCL10 [28]. Neutrophils also produce CCL2 and CCL20 that attract inflammatory T_H17 cells specialized in the secretion of IL-17, a pleiotropic cytokine that promotes neutrophil recruitment and activation [28]. In addition to attracting T_H1 and T_H17 cells to sites of inflammation, neutrophils enhance the differentiation of these effector T_H cells from naïve CD4⁺ T cell precursors [29]. Indeed, when exposed to inflammatory signals, some neutrophils acquire a hybrid, DC-like morphology; up-regulate the expression of APC molecules, such as MHC-II, and T cell costimulatory molecules, such as CD80 (B7.1) and CD86 (B7.2); and initiate specific T cell responses by presenting antigen to naïve CD4⁺ T cells and releasing T_H1-inducing cytokines, such as IL-12 and IFN-γ, as well as T_H17-inducing cytokines, such as IL-1β and IL-6 [29–31].

Conversely, T_H17 cells enhance the differentiation of neutrophils from bone marrow myeloid precursors by eliciting stromal cell release of G-CSF via IL-17 [32]. In addition, T_H17 cells augment the recruitment of neutrophils from the circulation by secreting CXCL8 [28]. Moreover, T_H17 and T_H1 cells augment neutrophil survival and activation by releasing GM-CSF, IFN-γ, and TNF [33]. These findings indicate that neutrophils initiate and amplify innate and adaptive immune responses by establishing bidirectional interactions with monocytes, macrophages, DCs, NK cells, and T cells through contact-dependent and contact-independent mechanisms.

Similar mechanisms are used by splenic neutrophils to activate B cells in the MZ of the spleen. Indeed, circulating neutrophils appear to acquire MZ B cell helper function after receiving contact-independent reprogramming signals via IL-10 and other STAT3-inducing cytokines produced by splenic, sinus-lining cells and macrophages [5]. Contact-independent signals are also involved in the activation of MZ B cells by neutrophils via BAFF (or B lymphocyte stimulator) and APRIL, two soluble TNF family members, structurally and functionally related to a T cell-bound, B cell-stimulating molecule, termed CD40L (or CD154) [5, 34–36].

Splenic neutrophils may further activate MZ B cells via contact-dependent mechanisms that include CD40L and NETs, which are extracellular projections that establish extensive interactions with MZ B cells [2, 5, 37]. NETs have antigen-binding and antimicrobial functions and originate from neutrophils undergoing “netosis”, a distinct form of apoptosis that causes massive chromatin decondensation [2, 37]. Recently, NET formation has been functionally linked to plasmacytoid DC production of IFN-α, a cytokine that plays a central role in autoimmune disorders [38, 39]. Given the autoreactive potential of MZ B cells and their constant exposure to foreign and autologous antigens, splenic neutrophils must deliver their MZ B cell helper activity in a highly controlled manner. Accordingly, splenic neutrophils express a host of immunoregulatory factors in addition to powerful B cell-activating factors [5].

NEUTROPHILS AS IMMUNOREGULATORS

In addition to delivering powerful immune-activating signals, neutrophils release immune regulatory signals that dampen immunity and inflammation [2]. For instance, neutrophils inhibit their own recruitment by releasing proresolving lipid mediators, such as lipoxin, resolvins, and protectins [2]. Neutrophils further accelerate the resolution of an ongoing immune response by terminating signaling from the CCR5 chemokine receptor ligands CCL3 and CCL5 through a process involving up-regulation of CCR5 expression on apoptotic neutrophils, followed by enhanced scavenging of CCL3 and CCL5 [2]. Furthermore, neutrophils attenuate the inflammatory activity of IL-1 by releasing soluble decoy IL-1Rs and IL-1R antagonist [2]. Clearance of apoptotic neutrophils by macrophages constitutes another important step in the resolution of inflammation, as neutrophil-engulfing macrophages trigger anti-inflammatory responses that lead to tissue repair [2, 40].

Neutrophils generate additional negative-feedback signals by up-regulating the expression of the IL-10R in response to microbial TLR ligands and inflammatory cytokines [41]. This process results in an increased responsiveness of neutrophils to the inhibitory activity of IL-10, an immunoregulatory cytokine produced by myeloid and lymphoid cells [41]. In mice, IL-10 is also produced by neutrophils after sensing microbes through TLRs and CLRs [42]. This pathway would allow neutrophils to attenuate the activation of innate and adaptive immune responses [42–44]. In humans, IL-10 production by neutrophils has also been reported, but more recent findings show that human neutrophils are unlikely to be a major source of IL-10, given that the IL-10 locus is inactivated in these cells [43–45]. Neutrophils further regulate T cell responses by promoting the formation of DCs that secrete the immunoregulatory cytokine TGF-β1 [46]. This pathway requires elastase, a neutrophil protease that activates proteinase-activated receptors on immune cells, including DCs [46]. Finally, recent studies have extended the regulatory activity of neutrophils to invariant NKT responses in humans and mice [47].

By delivering regulatory signals, neutrophils may tune down the magnitude of an immune response and contribute to its contraction after the clearance of intruding microbes. Neutrophils may also deliver regulatory signals to maintain a noninflammatory environment at sites continually exposed to antigen, such as the MZ of the spleen [6]. This enigmatic lymphoid structure mediates homeostatic and postimmune antibody responses to blood-borne antigens, including conserved microbial determinants that can activate MZ B cells in the absence of help from T cells [6, 7, 48]. The pathways underlying the regulation of T cells in the MZ of the spleen remain unclear, but splenic neutrophils deliver suppressor signals to T cells in addition to activating signals to MZ B cells [5]. This dual function may permit splenic neutrophils to induce homeostatic MZ B cell responses without eliciting the activation of inflammatory T cells.

NEUTROPHILS AS B CELL INDUCERS

Protective antibody responses to microbial proteins follow a slow but highly specific TD pathway that generates long-lived memory B cells and plasma cells expressing hypermutated and class-switched IgG and IgA antibodies with high affinity for antigen [48]. TD antibody production occurs through a germinal center reaction involving cognate interaction of FO B cells by DC-primed T_FH cells [49]. By producing CD40L, as well as IL-21, IL-4, and IFN-γ, T_FH cells stimulate FO B cells to express AID, a DNA-editing enzyme required for SHM and CSR [48–51]. SHM increases the affinity of antibodies for antigen by introducing point mutations in the variable(diversity)joining genes encoding the antigen-binding variable region of Igs, whereas CSR modulates the effector functions of B cells by replacing IgM with IgG, IgA, or IgE [50].

Neutrophils may enhance TD antibody production by collecting antigens at sites of infection or inflammation; by transporting antigen to sites involved in antibody production; by enhancing the antigen-presenting function of DCs; and by promoting the recruitment, activation, and differentiation of CD4⁺ T cells [2, 30, 31]. Neutrophils also release the CD40L-related molecules BAFF and APRIL [5, 36, 52, 53], which facilitate the survival of plasma cells emerging from follicular TD antibody responses [5, 36, 52–54]. BAFF and APRIL production by neutrophils and other innate cell types, such as DCs and epithelial cells, also enhances extrafollicular TI antibody responses by B cells located at mucosal sites and splenic MZ [5, 55–57].

MZ B cells are naturally poised to mount IgM as well as IgG and IgA responses in a rapid TI manner, which allows MZ B cells to bridge the temporal gap (in general, 5–7 days) required for the initiation of TD antibody production by FO B cells [6, 7]. The rapid kinetics of TI antibody responses may relate to the ability of MZ B cells to use fast helper signals from DCs, macrophages, and neutrophils of the innate immune system rather than slow helper signals from T_FH cells of the adaptive immune system [5, 48, 58, 59]. Neutrophils were previously thought to colonize the MZ of the spleen only in response to blood-borne infections [58, 60]. However, we found that neutrophils occupy splenic peri-MZ areas also in the absence of infection and do so through a noninflammatory pathway that begins during fetal life and accelerates after birth—a time that coincides with the colonization of mucosal surfaces by commensal bacteria [5]. Unlike circulating neutrophils, splenic neutrophils deliver powerful antibody-inducing signals to MZ B cells (Fig. 1), and thus, we have defined these cells as N_BH cells [5].

Compared with circulating neutrophils, N_BH cells express an activated phenotype; secrete more B cell-stimulating factors, such as BAFF, APRIL, CD40L, and IL-21; produce more plasmablast-attracting factors, such as CXCL12; and strongly activate MZ but not FO B cells [5]. These unique features likely reflect the activation and reprogramming of N_BH cells by local microenvironmental signals. Consistent with this possibility, accumulation of N_BH cells coincides with postnatal exposure of splenic peri-MZ areas to microbial TLR ligands of probable mucosal origin, such as LPS and bacterial RNA [5]. In addition to activating N_BH cells, these TLR ligands stimulate perifollicular sinus-lining cells to express more neutrophil-attracting chemokines, including CXCL8, which may contribute to the recruitment of N_BH cells [5]. Accordingly, the congenital deficiency of molecules required for TLR signaling causes reduction of N_BH cells and MZ B cells [5, 61].

Microbial TLR ligands also stimulate the reprogramming of N_BH cells from conventional neutrophils by eliciting the release of noninflammatory cytokines, such as IL-10, from perifollicular sinus-lining cells and macrophages [5]. Considering that IL-10 provides regulatory signals to neutrophils [2, 41], this cytokine may be instrumental for N_BH cells to operate in a noninflammatory environment. The generation and/or maintenance of N_BH cells may also involve GM-CSF, a cytokine that promotes the recruitment, survival, and activation of neutrophils and whose production has been identified recently in a subset of splenic B cells termed IRA B cells [2, 62].

N_BH cells induce AID expression, IgG and IgA CSR, and antibody-secreting plasmablasts by activating MZ B cells through a TI mechanism involving BAFF, APRIL, and somewhat surprisingly, IL-21 [5]. These molecules also stimulate B cell and plasma cell survival [48], raising the possibility that neutrophils sustain MZ B cell responses through multiple mechanisms in vivo. Remarkably, human N_BH cells include N_BH1 and N_BH2 subsets that show distinct phenotypic and functional features [5]. Indeed, compared with N_BH1 cells, N_BH2 cells express lower levels of surface CD15 and CD16 and stimulate more effective MZ B cell responses as a result of a more pronounced release of BAFF, APRIL, and IL-21 [5]. Consistent with this, immunodeficient patients with quantitative or functional neutrophil disorders or defective BAFF, APRIL, or IL-21 signaling have fewer MZ B cells and reduced steady-state production of IgG and IgA to TI carbohydrate but not TD protein antigens [5]. Of note, antibodies to TD antigens are produced by plasma cells lodged in bone marrow niches that contain eosinophils secreting the plasma cell-survival factor APRIL [63], which suggests that there is a division of labor between neutrophils and eosinophils for the maintenance of plasma cells at distinct lymphoid sites.

The stimulation of MZ B cells by N_BH cells may involve the formation of NET-like structures, which indeed appear to trap microbial products, such as bacterial RNA, and establish extensive interactions with MZ B cells [5]. By trapping blood-borne commensal antigens, possibly originating from mucosal surfaces, NETs would facilitate the provision of BCR ligands to MZ B cells under steady-state conditions [5, 8, 48, 64, 65]. NETs may also deliver TLR ligands to MZ B cells, and indeed, there is some evidence pointing to an important role of TLRs in the maturation, maintenance, activation, and function of MZ B cells [39, 66, 67]. Together with BAFF, APRIL, and IL-21, microbial and possibly autologous TLR ligands from N_BH cells may contribute to the induction of AID expression and initiation of CSR and SHM in a fraction of MZ B cells under homeostatic conditions [5]. In humans, MZ B cells initiate SHM during fetal life through a TI pathway that remains active after birth [7, 48, 68]. Indeed, spleens from adult individuals contain mutated MZ B cells but few or no T_FH cells and active germinal centers [7, 69, 70]. Furthermore, mutations are decreased in MZ B cells from patients with congenital neutropenia but not in patients with a congenital impairment of TD antibody production as a result of a deficiency of CD40L [5, 71, 72].

The TI nature of the helper signals provided by N_BH cells is suggested further by the observation that N_BH cells not only induce CSR, SHM, and antibody production in the absence of CD4⁺ T cells but also suppress the activation of CD4⁺ T cells, at least in vitro [5]. By exerting this dual B cell helper–T cell suppressor function, N_BH cells may skew antibody responses to commensal blood-borne antigens toward a MZ-based TI pathway at the expense of a follicle-based TD pathway. Overall, the crosstalk between N_BH cells and MZ B cells may be instrumental to generate a second line of innate antibody defense against microbial antigens that break the first line of defense provided by the mucosal barrier.

CONCLUSIONS

In contrast to the traditional view of neutrophils as unsophisticated cells of the phagocytic system, neutrophils have emerged recently as an important component of the effector and regulatory circuits that control the magnitude and quality of an immune response. By establishing bidirectional interactions with DCs, monocytes, macrophages, and T cells, neutrophils modulate innate and adaptive immune responses in health and disease states [2]. Our studies indicate that neutrophils crosstalk further with MZ B cells to enhance the generation of ready-to-use antibodies against conserved microbial antigens [48]. A possible implication of our findings is that adjuvants capable to harness the B cell helper activity of neutrophils could improve protective antibody responses to antigens recognized by MZ B cells, including poorly immunogenic TI antigens. Another implication of our studies is that the pathogenesis of recurrent infections arising in immunodeficient patients with neutrophil disorders may be more complex than anticipated originally and could involve decreased production of preimmune antibodies to specific sets of conserved microbial antigens.