Matrix-M™ Adjuvant Induces Local Recruitment, Activation and Maturation of Central Immune Cells in Absence of Antigen

Saponin-based adjuvants are widely used to enhance humoral and cellular immune responses towards vaccine antigens, although it is not yet completely known how they mediate their stimulatory effects. The aim of this study was to elucidate the mechanism of action of adjuvant Matrix-M™ without antigen and Alum was used as reference adjuvant. Adjuvant Matrix-M™ is comprised of 40 nm nanoparticles composed of Quillaja saponins, cholesterol and phospholipid. BALB/c mice were subcutaneously injected once with, 3, 12 or 30 µg of Matrix-M™, resulting in recruitment of leukocytes to draining lymph nodes (dLNs) and spleen 48 h post treatment. Flow cytometry analysis identified CD11b+ Gr-1high granulocytes as the cell population increasing most in dLNs and spleen. Additionally, dendritic cells, F4/80int cells, T-, B- and NK-cells were recruited to dLNs and in spleen the number of F4/80int cells, and to some extent, B cells and dendritic cells, increased. Elevated levels of early activation marker CD69 were detected on T-, B- and NK-cells, CD11b+ Gr-1high cells, F4/80int cells and dendritic cells in dLNs. In spleen CD69 was mainly up-regulated on NK cells. B cells and dendritic cells in dLNs and spleen showed an increased expression of the co-stimulatory molecule CD86 and dendritic cells in dLNs expressed elevated levels of MHC class II. The high-dose (30 µg) of Matrix-M™ induced detectable serum levels of IL-6 and MIP-1β 4 h post administration, most likely representing spillover of locally produced cytokines. A lesser increase of IL-6 in serum after administration of 12 µg Matrix-M™ was also observed. In conclusion, early immunostimulatory properties were demonstrated by Matrix-M™ alone, as therapeutic doses resulted in a local transient immune response with recruitment and activation of central immune cells to dLNs. These effects may play a role in enhancing uptake and presentation of vaccine antigens to elicit a competent immune response.


Introduction
Adjuvants are compounds added to vaccine antigens to facilitate and enhance activation of the innate and adaptive immune responses, to improve the immunogenicity and efficacy of vaccines. Adjuvants exert their effects through various mechanisms, but also the nature and identity of the antigen(s) contributes to the immune response [1]. The accumulated knowledge on adjuvant mode of action is primarily based on data from studies of adjuvants formulated with vaccine antigens. However, to gain unbiased information on the specific immune stimulation elicited by the adjuvant, studies without antigen(s) are required. Basic research on adjuvant properties is central in order to better understand how adjuvants and antigens work in concert, generating a safe and efficacious immune response.
Saponins, particularly those obtained from Quillaja saponaria Molina, are known potent adjuvants and Quillaja saponins (QS) have for long been used in animal vaccines. Saponin-based adjuvants can be formulated in different ways; in free form [2], with aluminium hydroxide [3], in ISCOMs (immunostimulating complex) [4] or in ISCOM-Matrix/Matrix structures [5]. QS constitute a heterogeneous mixture of related but different chemical structures with various immunostimulatory activities, safety profiles and particle forming properties. By purification of the QS raw material, distinctive fractions with different characteristics can be defined.
The ISCOM, a potent adjuvant formulation first described in 1984 by Morein and co-workers [4], consist of stable complexes composed of saponin, cholesterol, phospholipid and incorporated antigen(s). The hallmarks of the ISCOM technology are the dosesparing potential [6], induction of high and long-lasting antibody titers and potent T cell responses [7]. However, later it was shown that antigen incorporation is not critical for these immune properties. Antigen and empty ISCOMs i.e. ISCOM-Matrix/ Matrix could simply be mixed with sustained vaccine efficacy [5].
In this study we use a novel adjuvant formulation based on two different Matrix particles made from two separate purified fractions of saponins, yielding Matrix-A TM and Matrix-C TM [8]. These Matrix particles, approximately 40 nm large, are subsequently mixed at defined ratios to get the Matrix-M TM adjuvant.
Formulated saponin-based adjuvants are thought to enhance cell trafficking and activation of immune cells, i.e. components responsible for inducing cytokine production and facilitating antigen uptake, processing and presentation on MHC class I and II [9]. Moreover, convincing data suggests that co-delivery of the saponin-based adjuvant ISCOMATRIX TM and antigen is central for CD8 + T cell induction and that their simultaneous drainage to a common lymph node is crucial for the adjuvant effect in general [10]. Interestingly though, unpublished data show that premixing of Matrix-M TM and influenza antigen is not required to mount a potent humoral immune response. Matrix-M TM and antigen could be administered up to 24 h apart with sustained IgG1 and IgG2a antibody responses.
Currently, the ISCOM-Matrix/Matrix saponin-based adjuvants Matrix-M TM [11] and ISCOMATRIX TM [12] have entered and moved forward into human clinical trials. Other ISCOM-Matrix/Matrix adjuvants are available in commercial veterinary vaccines e.g. Matrix-C TM [13]. The assembled efficacy data on these and other saponin-formulated vaccines is convincing, however, their mode of action still needs to be further investigated. In this study we have chosen to elucidate the early immunostimulatory properties of Matrix-M TM in mice without influence of antigen, in order to further characterize its mode of action. Accordingly, understanding the inherent, antigen-independent properties of adjuvants is thus essential for the development of safe and effective vaccines.  were injected subcutaneously at the base of the tail. After 48 h cells from the two draining lymph nodes (dLNs) (A) and the spleen (B) were prepared and total cell counts were performed. The relative difference in cell numbers between 3, 12 or 30 mg Matrix-M TM -and PBS-treated mice or Alumtreated and naïve mice, respectively are presented. Dotted line represent no difference in cell number between treated and control mice (fold change = 1). Data is shown as mean 6 SD (n = 6-8 mice). Significant differences between Matrix-M TM -and PBS-treated mice using Kruskal-Wallis test with Dunn's posttest or between Alum-treated and naïve mice using Mann-Whitney test are outlined with *, p,0.05; **, p,0.01; ***, p,0.001. M, Matrix-M TM . doi:10.1371/journal.pone.0041451.g002

Adjuvant preparation
Matrix-M TM (AbISCOH-100, Isconova AB, Uppsala, Sweden), a mixture of Matrix-A TM and -C TM at the ratio of 85:15, was used. Matrix-A TM and -C TM were prepared from separately purified fractions of QS subsequently formulated with cholesterol and phospholipid into Matrix particles. The Alum used in this study was 2% Alhydrogel (Al(OH) 3 , Brenntag Biosector, Frederikssund, Denmark).

Animals and experimental design
Eight weeks old female BALB/c mice were purchased from Scanbur (Stockholm, Sweden) and kept at the National Veterinary Institute (SVA, Uppsala, Sweden). All animal experiments were approved by Uppsala Ethical Committee (Permit numbers: C 50/ 7, C 191/11). The animals were kept in accordance with national guidelines. Mice were injected subcutaneously (s.c.) at the base of the tail with 100 ml Matrix-M TM . The therapeutic dose range of Matrix-M TM in mice is 3-12 mg, thus 3 and 12 mg were evaluated in this study. Moreover, mice were injected with a high-dose of Matrix-M TM (30 mg). Control mice received 100 ml PBS, pH 7.4. In a separate experiment mice were injected s.c. at the base of the tail with 100 ml Alum (1%) corresponding to a 500 mg Alum dose and naïve mice were used as negative controls. Matrix-M TMtreated mice were included as controls. Draining lymph nodes (dLNs, inguinal), spleen and blood were collected 4, 24 and 48 h after injection, except for 12 mg Matrix-M TM where dLN and spleen were collected at 24 and 48 h. Blood was drawn from the tail vein at all time points.

Quantification of cytokines
The presence of cytokines and chemokines in serum was measured using Cytometric Bead Array (CBA, BD Biosciences, Erembodegem, Belgium) according to the manufacturer's instructions. The protein flex sets used were IL-1b, IL-2, IL-4, IL-6, IL-10, IL-12p70, MIP-1b, IFN-c, TNF, RANTES and GM-CSF. Serum samples were diluted 1:4 with assay diluent buffer. Fluorescence measurement was performed using a FACSCanto flow cytometer (BD Biosciences) and data were analyzed using the FCAP Array software (v. 1.0.1, BD Biosciences).

Cell preparation
Spleen and dLNs were collected in cold PBS and processed to single cell suspensions by passage through a 23G needle. Cells from the two dLNs were pooled. Splenocytes were incubated in ammonium chloride for 5 min, washed with PBS containing 2% FBS and passed through a 100 mm cell strainer. Cells were resuspended in staining buffer (PBS, pH 7.4, 0.5% BSA, 2 mM EDTA, 0.1% sodium azide). The viable cell concentration was determined by staining with Trypan blue and phase contrast microscopy.

Morphological analysis
Single cell suspensions were centrifuged onto microscope slides (10 5 cells/slide, 500 rpm, 5 min) using a cytocentrifuge (Shandon, Cheshire, United Kingdom). Slides were air-dried at RT, stained with May-Grünwald-Giemsa (Merck, Darmstadt, Germany) and analyzed using a Leica CTRMIC microscope (Leica Microsystems, Wetzlar, Germany). The analysis of the slides was done in an unblinded fashion.

Flow cytometry analysis
Cell suspensions were prepared as described above and incubated 20 min at 4uC with anti-mouse CD16/CD32 antibody (Mouse BD Fc Block TM ). Cells were then transferred to a 96-well microtiter plate and incubated with antibodies for 30 min at 4uC (5610 5   and dendritic cells (DCs). Appropriate fluorochrome-conjugated isotype antibodies were used as controls. Cells were washed and incubated 10 min at RT with 7-AAD to stain non-viable cells. All staining procedures were conducted on ice and reagents were purchased from BD Biosciences unless otherwise stated. Cells were analyzed using a FACSCanto flow cytometer (BD Biosciences), 50 000 events per sample were collected. Retrieved data was analyzed using the FACSDiva software (v. 5.0.2, BD Biosciences).

Statistical analysis
Data were analyzed using the non-parametric tests Mann-Whitney or Kruskal-Wallis test with Dunn's post test (GraphPad Prism 5.01 for Windows, GraphPad Software, San Diego, California, USA). Statistical significance was assigned at a p value of ,0.05.

High-dose Matrix-M TM increases IL-6 and MIP-1b levels in serum
Mice were injected s.c. with 3, 12 or 30 mg of Matrix-M TM or Alum and the cytokine/chemokine levels in serum were analyzed using CBA after 4 and 24 h. Thirty mg of Matrix-M TM significantly increased serum levels of IL-6 (4466319 pg/ml, p,0.001) and to a lesser degree also MIP-1b (65650 pg/ml, p,0.05) (Figure 1). The IL-6 level was also increased in 12 mg Matrix-M TM -treated mice compared to PBS control (110649 versus 0.3861.3, p,0.01). IL-6 and MIP-1b levels returned to near background levels at 24 h (data not shown). Increased levels of circulating IL-6 and MIP-1b were not observed with 3 mg Matrix-M TM or Alum (Figure 1 and data not shown). The levels of IL-1b, IL-2, IL-4, IL-10, IL-12p70, IFN-c, TNF and GM-CSF were below detection levels and the levels of RANTES retrieved after administration of Matrix-M TM or Alum did not differ from the levels in the control mice (data not shown).

Discussion
Matrix-M TM is a potent adjuvant and its immunomodulatory activities have been proven in a number of species, however, the mechanisms behind these properties are not yet fully understood [11], [14]. In this study, we show that Matrix-M TM treatment results in a local and transient immune stimulation with recruitment of lymphocytes, macrophages and granulocytes to dLNs and spleen. The number of cells in dLNs increased about 3fold after 48 h in Matrix-M TM -treated mice compared to PBS control. These results are in agreement with previous findings with a similar adjuvant where recruitment of lymphocytes to dLNs was reported in sheep after administration of ISCOMATRIX TM [15]. Further, Matrix-M TM administration also leads to activation of lymphocytes, DCs and granulocytes in dLNs as shown by increased expression of CD69. The biological impact of upregulation of this early activation marker is not entirely elucidated however it reflects functionality of the immune cells [16]. Additionally, Matrix-M TM induces maturation of antigen presenting cells, detected by elevated levels of CD86 and MHC class II. CD86 is central for priming and activation of naïve T cells. The MHC class II expression was primarily affected on DCs in dLNs 24 h after Matrix-M TM treatment and the expression was downregulated after 48 h. The down-regulation of MHC class II on DCs after 48 h may be due to lack of an antigen to process and present to T helper (Th) cells. Although Matrix-M TM administration increased the number of cells in spleen these were not at all as activated compared to cells in the dLNs. This could indicate that the splenocytes are primarily not activated by Matrix-M TM directly, contrary to the lymph node cells. The spleen cells are most likely activated secondarily by activated migrating cells as shown by the increased expression of late activation markers such as CD86. Alum, in contrast, did not induce recruitment or activation of studied immune cells in dLNs or spleen at the chosen time points, except for an increased expression of CD69 and CD86 on DCs in dLN after 48 h. On the contrary, in Matrix-M TM -treated mice the up-regulation of CD86 occurred already at 24 h suggesting a different kinetics in immune cell activation. The Alum findings are in agreement with previously published data showing no increased CD69 expression on B-, T-or NK-cells as well as no increased CD86 expression on B cells or increased CD86/MHC class II expression on DCs at 24 h post injection [17]. The observed differences in early immune cell activation and recruitment between Matrix-M TM and Alum clearly implicate distinct adjuvant mechanisms of action.
The murine therapeutic dose interval of Matrix-M TM is between 3 and 12 mg. Three mg Matrix-M TM still recruited and activated immune cells, although the kinetics was to some extent delayed compared to higher doses tested. An administration of 12 mg Matrix-M TM induced a somewhat elevated serum level of IL-6 and the high-dose of Matrix-M TM induced a higher and a more heterogeneous secretion of IL-6 and also MIP-1b. Circulating levels of IL-6 and MIP-1b most likely represent spillover of locally produced cytokines and the secretion of these cytokines at the injection site may facilitate e.g. antigen presentation. IL-6 has been shown both in vitro and in vivo to be induced by a number of saponinbased adjuvants [18], [19]. IL-6 is primarily secreted by tissueresident macrophages and provides signals for migration and activation of innate and adaptive immune cells. In particular, IL-6 is essential for induction of IL-21 production by naïve CD4 + T cells upon antigen stimulation. IL-21 promotes production of most IgG subclasses [20], [21]. MIP-1b, also produced by macrophages, is a potent chemoattractant for human granulocytes, NK cells and activated T cells. Previously, Alum and MF59 have been shown to induce MIP-1b in vitro [22]. However in this study, as for the 3 mg Matrix-M TM treatment, Alum treatment did not result in increased serum levels of the analyzed cytokines/chemokines. The CD11b + Gr-1 high granulocytes, predominantly neutrophils [23] was the population increasing the most in dLNs after Matrix-M TM treatment relative to the PBS control. This recruitment was detected even for the lowest dose evaluated nevertheless, the kinetics of the granulocyte migration was dose dependent. The highest number of cells was detected at the earliest time point analyzed, 4 h. Thus, it would be interesting to study the cell recruitment to the dLNs at an even earlier time point to further investigate the kinetics of granulocyte migration. The large recruitment of granulocytes to dLNs could influence the immune response in different ways. Several studies in mice show that neutrophils can transport live bacteria and antigen to dLNs [24]-[] [27]. This indicates a role for neutrophils in the adaptive immune response, possibly by delivering antigens to DCs [28]. It is also reported that CD11b on activated neutrophils interact with the c-type lectin DC-SIGN present on DCs leading to activation, providing an important link between innate and adaptive immunity [29]. These DCs can thereafter activate T cells and induce Th1 polarization. Interestingly, granulocytes with ingested microbes have shown to be involved in antigen presentation and to serve as substrate for in vivo cross-priming of CD8 + T cells by DCs [30], [31]. In addition, the adjuvant MF59 targets granulocytes, monocytes and macrophages, inducing an increased migration of these cells to the injection site [22]. However, it was shown that neutrophil depletion have no impact on the adjuvant effects of MF59, thus neutrophils and other cell populations possess overlapping functions [32]. Further, neutrophils have a negative impact on Th-and B-cell responses towards antigen formulated with e.g. Alum [33]. As Matrix-M TM also augments recruitment of neutrophils and DCs to the injection site, indicated by elevated cell numbers in dLNs, an increase in early antigen uptake and transportation may be achieved. In contrast, Alum treatment did not increase the neutrophil or DC numbers in dLNs nor in spleen.  The nanoparticle property of Matrix-M TM is likely essential for the ability to increase antigen uptake by phagocytosis, enabling antigen presentation. It has been shown that induction of innate immune responses by other particulate adjuvants e.g. aluminiumcontaining adjuvants is due to activation of the NLRP3 inflammasome [34]-[] [36]. However, if this activation is essential for the adjuvant effects on the adaptive immunity is yet a matter of debate [37], [38].
NK cells also increased in dLNs after Matrix-M TM treatment. Interestingly, it has been shown that DCs activated with e.g. LPS or CpG, activates and recruits NK cells to antigen-stimulated lymph nodes. Consequently, due to IFN-c produced by NK cells there is a polarization towards a Th1 immune response [39], [40].
Taken together, the data presented show that low doses of Matrix-M TM induce a local transient proinflammatory response with recruitment, activation and maturation of important immune cells. Most likely the recruitment and activation of neutrophils and DCs by Matrix-M TM represents an important mechanism for the adjuvants' capacity to induce cytotoxic T lymphocytes. As an emerging body of evidence show that immune cells recruited after administration of several adjuvants seems to have overlapping functions, further investigations to decipher the mode of action of Matrix-M TM are required. In conclusion, we suggest that the described properties of Matrix-M TM may play a central role for efficient uptake and presentation of antigens, being significant for design and development of effective vaccines.