SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowldgment
  8. Conflict of interest
  9. References
  10. Supporting Information

Wnt5a is a non-canonical Wnt protein that is expressed at elevated levels in inflammatory conditions. Its role in inflammation remains unclear, although it is known that Wnt5a is expressed at a higher level in monocyte-derived myeloid dendritic cells (Mo-mDCs) than in monocytes and macrophages. The function of Wnt5a in dendritic cells (DCs) remains relatively unexplored. Here, we found that under Mo-mDC culture conditions, Wnt5a inhibited the generation of CD14+/low Mo-mDCs while promoting the generation of CD14+/++ CD16+ monocytes. We could further show that stimulation of monocytes with rWnt5a induced a rapid IL-6 production and that the rWnt5a treated Mo-mDC differentiation was restored upon blocking of IL-6. Also, conditioned media from Wnt5a stimulated human breast cancer cells producing IL-6, specifically inhibited Mo-mDC differentiation. These observations are strengthened by our finding that patients with sepsis, a disease involving elevated Wnt5a and IL-6 levels, also showed a significant increase in the CD14CD16++/CD14+/++ CD16+ monocyte populations, which was accompanied by a significant decrease in circulating mDCs. We finally show that under typical Mo-mDC culture conditions, monocytes isolated from patients with sepsis as compared to healthy controls, preferentially differentiated into CD14+/++ HLA-DR++ cells. We suggest that Wnt5a is a possible candidate mediator for the CD14+/++ CD16+ monocyte accumulation seen in patients with infectious disease and cancer.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowldgment
  8. Conflict of interest
  9. References
  10. Supporting Information

Dendritic cells (DCs) are antigen-presenting cells (APCs) of the immune system that play a central role in immune responses by bridging innate and adaptive immunity. Human blood DCs are a heterogeneous population of cells that mainly are divided into plasmacytoid dendritic cells (pDCs) and myeloid dendritic cells (mDCs). The mDCs are further subdivided into the mDC1 and mDC2 populations based on their surface phenotype [1-3]. mDCs can differentiate from monocytes both in vivo and in vitro [4, 5]. Exposure in vivo to stimuli from a local inflammation, an infection, or a malignant tumour induces the monocytes to migrate into the affected tissue and differentiate into either macrophages or various kinds of mDCs, and under natural conditions, this process is highly plastic and is affected by the local environment [1, 3]. Similarly, blood monocytes in vitro can differentiate into either monocyte-derived macrophages (Mo-M) or monocyte-derived mDCs (Mo-mDCs) [6, 7]. Based on their expression of CD14 and CD16 in vivo, human monocytes can be divided into three main subpopulations: the classical CD14++ CD16 monocytes, which predominate and can secrete both pro-inflammatory and anti-inflammatory mediators; the non-classical CD14CD16++ monocytes, which are assumed to be similar to the mouse Ly6Clow monocyte population and have a patrolling role [8]; the intermediate CD14+/++ CD16+ monocytes, which play a pro-inflammatory role but have also been suggested to secrete the anti-inflammatory cytokine IL-10 [9, 10].

IL-6 has been observed to inhibit DC differentiation in vivo, and this effect was attributed to activation of stat3 [11]. In diseases such as sepsis, the level of IL-6 has been directly correlated with a decrease in circulating DCs and an increase in monocytes [12]. The local tissue microenvironment can also have a profound effect on monocyte differentiation in various disease conditions [13, 14]. An example of this is the local generation of anti-inflammatory macrophages in tumours (tumour-associated macrophages, TAMs), where the tumour microenvironment skews macrophage differentiation towards tumour-promoting cells [15]. IL-6 is produced by tumour cells or stromal cells in the tumour and, together with M-CSF, also serves as a potent inhibitor of mDC differentiation. IL-6/M-CSF derived from renal cell carcinoma and pancreas cancer has previously been shown to inhibit mDC differentiation and promote the production of monocytes [16, 17].

Wnt5a is a non-canonical Wnt protein that is involved in developmental processes, cell adhesion, migration and tissue polarity [18]. Previous research has shown that Wnt5a increases the levels of IL-6 in malignant melanoma cells and in synoviocytes from patients with rheumatoid arthritis [19, 20]. Furthermore, it has been reported that expression of Wnt5a is induced in macrophages exposed to a bacterial stimulus (e.g. lipopolysaccharide, LPS) [21] and in patients with sepsis [22]. Wnt5a is also upregulated during Mo-mDC differentiation [23]. Moreover, we recently showed that Wnt5a induced tolerogenic Mo-M (Mo-M; CD14+HLA-DR-/low) in a pro-inflammatory environment [24]. In the light of these findings, we conducted this study to investigate whether Wnt5a also affects Mo-mDC differentiation.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowldgment
  8. Conflict of interest
  9. References
  10. Supporting Information
Urosepsis patients

The samples used in our study were collected from patients with pyelonephritis at their first consultation for the condition, and thus, the history of their disease after admission to hospital was of <3 days. The final diagnosis was based on a combination of clinical symptoms and conventional testing using Swedish national QC-approved culture. Blood from patients with gram- sepsis was used (E. coli, K. oxytoca, C. koseri). Permission for the study was obtained from the Research Ethics Committee at Lund University (Dnr 288/2007), and the participating patients or healthy volunteers gave a written informed consent. Blood samples were immediately diluted 1:2 in phosphate-buffered saline (PBS; EDTA/sucrose) and overlaid on Ficoll-Paque PLUS. Peripheral blood mononuclear cells (PBMCs) were collected for monocyte isolation, washed once in PBS and once in FACS buffer, and then stained with the indicated antibodies and analysed by flow cytometry as described below. Plasma was flash-frozen at −20 °C and then analysed for IL-6 by ELISA.

Isolation of monocytes and CD4+ T cells

Leucocytes were collected by within 2–3 h from leucocyte depletion filtration of blood from healthy blood donors performed according to a previously published method [25]. Briefly, leucocytes from filters were separated by Ficoll-Paque PLUS prior to magnetic cell sorting (MACS). Monocytes or CD4+ naïve T cells were isolated by MACS using a Monocyte Isolation kit II or a Naïve CD4+ T Cell Isolation kit II according to the instructions of the manufacturer (Miltenyi Biotec, Bergisch Gladbach, Germany). The Monocyte Isolation kit II, contains antibodies against CD3, CD7, CD16, CD19, CD56, CD123 and glycophorin A. Naïve CD4 +  T Cell Isolation kit II, contains antibodies against CD8, CD14, CD15, CD16, CD19, CD25, CD34, CD36, CD45RO, CD56, CD123, TCRγδ, HLA-DR and glycophorin A. The purity of monocytes was generally between 80 and 90% (controlled by flow cytometry using CD14/CD33/HLADR antibodies) and that of CD4+ T cells >90% (controlled by flow cytometry using CD4/CD45RA/CD3). Approval for these procedures was obtained from the Regional Ethics Committee in Lund, Sweden.

Cell culture

Monocytes isolated from leucocyte depletion filters were induced to differentiate into Mo-mDCs by culture for 7 days in 10% FBS RPMI-1640 supplemented with penicillin/streptomycin (100 U/ml and 100 μg/ml, respectively) and 10 ng/ml rh GM-CSF + 20 ng/ml rh IL-4 (Mo-mDC medium), with or without Wnt5a or Wnt3a; re-stimulation with all these agents was performed on days 3 and 5. Monocytes were induced to differentiate into Mo-M by culture for 7 days in OptiMEM supplemented with penicillin/streptomycin (100 U/ml and 100 μg/ml, respectively) and 10 ng/ml rh GM-CSF (Mo-M medium); re-stimulation was performed on days 3 and 5.

Monocytes from patients with sepsis were let to differentiate in Mo-mDC cultures or Mo-M cultures using the same procedures as described previously, with the only exception that peripheral blood from healthy volunteers was used for monocyte preparation as relevant controls to sepsis samples.

For analyses of IL-6 production, freshly isolated monocytes were kept in OptiMEM supplemented with penicillin/streptomycin and immediately stimulated with rWnt3a, rWnt5a or LPS for 3 h after which the cells and supernatants were collected and flash-frozen at −80 °C.

IL-6 blocking experiments were performed using a polyclonal goat anti-human IL-6 neutralizing antibody (R & D Systems) at 1 μg/ml added simultaneously with the Mo-mDC differentiation agents (at days 0, 3 and 5), and then harvested using non-enzymatic cell dissociation buffer prior to FACS analysis.

All media and supplements were purchased from Thermo Scientific (Logan, UT, USA) unless otherwise stated. LPS stimulations were performed day 5. For all experiments, the differentiated cells were harvested using non-enzymatic cell dissociation buffer (Sigma-Aldrich, St Louis, MO, USA).

MDA-MB-231 cells from ATCC were cultured in 10% FBS RPMI-1640 supplemented with penicillin/streptomycin. Conditioned media (CM) from MDA-MB-231 cells were harvested at subconfluency from cells either kept unstimulated (CM) or recently prestimulated with rWnt5a for 20 h (rWnt5a stim CM). Monocytes were subsequently cultured in normal Mo-mDC cultures (Mo-mDC Ctrl) with addition of (CM) or (rWnt5a stim CM) for the whole culture period (5 days) with or without addition of IL-6 blocking antibodies. In all wells, GM-CSF and IL-4 were added as usual. For analyses of IL-6 secretion, MDA-MB-231 were serum-starved for 12 h prior stimulation with rWnt3a or rWnt5a for 3 or 6 h.

Cytospins and May-Grunwald Giemsa staining

Cytospins were prepared from non-enzymatic cell dissociation buffer-harvested Mo-mDC day 7 cultures that had or had not been treated with rWnt5a or with rWnt3a as a control. Cytospins were let to air-dry and were then May-Grunwald Giemsa stained.

Compounds

All recombinant human cytokines were obtained from R & D Systems, and the following concentrations were used in all experiments: 10 ng/ml GM-CSF, 20 ng/ml IL-4, 50 ng/ml IL-6, the standard Wnt concentrations 0.125–0.25 μg/ml Wnt3a and 0.5 μg/ml Wnt5a. Endotoxin was below detection levels (<0.1 pg/ml) in an LAL assay (GenScript). LPS was purchased from Sigma-Aldrich (St. Louis, MO, USA) and was used at a concentration of 100 ng/ml. The IL-6 neutralizing antibody (R & D Systems) was used at 1 μg/ml.

All the antibodies used in flow cytometry (i.e. CD80 clone L307.4, CD83 clone HB15e, CD1a clone HI149, CD206 clone 19.2, CD209 clone DCN46, CD14 clone M5E2, HLA-DR clone G46-6, CD33 clone WM53, CD11b-bio clone M1/70, CD163 clone GHI/61, CD16 clone 3G8, CD40 clone 5C3 and CCR7 clone 3D12) were obtained from Becton Dickinson, and analyses were performed using 7AAD (eBiosciences, San Diego, CA, USA). The pDC, mDC1 and mDC2 populations were analysed using a Blood DC Enumeration kit (Miltenyi Biotec, Bergisch Gladbach, Germany) containing antibodies against BDCA-1 (CD1c), BDCA-2 and BDCA-3.

The following antibodies were used for Western blot: anti-mouse/rat Wnt5a (R & D Systems; AF645), ERK 1/2-P (Thr 202/Tyr204, clone 197G2) and ERK 1/2 clone 137F5 (Cell Signaling Technology Inc., Danvers, MA, USA); Actin (C-4) (MP Biomedicals, Solon, OH, USA).

Pinocytosis assay

The pinocytic activity of cultured cells was assessed by FITC-dextran uptake. Monocytes were differentiated in Mo-mDC cultures as indicated above, under Wnt3a or Wnt5a stimulating conditions or not. On day 7, the differentiated cells were incubated with 1 mg/ml FITC-dextran (Sigma-Aldrich) and HLA-DR-APC at 37 °C for 20 min and were subsequently harvested using non-enzymatic cell dissociation buffer (Sigma-Aldrich) and analysed by flow cytometry. Cells incubated at 4 °C were used as a control for pinocytosis.

Allogeneic mixed lymphocyte reaction

The cultured Mo-mDCs were harvested on day 7 of culture using non-enzymatic cell dissociation buffer (Sigma-Aldrich) and then reseeded in 96-well plates and incubated with freshly isolated naive CD4+ T cells (range 10,000–50,000 T cells per experiment) at stimulator–responder ratios ranging from 1:10 to 1:100. On day 4 of co-culture, 1 mCi of [methyl-3H] thymidine was added for 18 h, and incorporation was determined in a Microbeta Counter (Perkin & Elmer, Waltham, MA, USA).

Elisa

Plasma was flash-frozen at −20 °C and then analysed for IL-6 by ELISA. Plasma was obtained from centrifuged flash-frozen blood and then analysed for IL-6 by ELISA. Supernatants from monocytes and differentiated Mo-mDCs were collected 3 h after stimulation of freshly prepared monocytes with rWnt3a, rWnt5a or LPS, or day 7 of Mo-mDC culture, respectively, and the amount of IL-6, IL-10 and IL-12 was determined using ELISA according to the instructions of the manufacturer (the minimum detectable concentrations were 0.70 pg/ml, 3.9 pg/ml and 0.5 pg/ml, respectively). All ELISA kits from R & D Systems.

Western blot

Cells were washed using PBS and then directly lysed in boiling Laemmlli sample buffer supplemented with 100 mm DTT. Protein lysates were directly run on 10% SDS-PAGE gels, transferred to PVDF membranes and blotted using PBS-Tween-BSA and ECL (Santa Cruz, CA, USA).

Statistical analyses

Statistical analyses of all peripheral blood cell populations were performed in spss [SPSS version 20.0, (SPSS, Inc, Chicago, IL, USA)] Statistics by nonparametric Mann–Whitney U Wilcoxon test. All other analyses statistics by Student's t-test.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowldgment
  8. Conflict of interest
  9. References
  10. Supporting Information

The CD16+ monocyte populations increase while mDCs decrease in peripheral blood of sepsis patients

Sepsis is a disease reported to be associated with high systemic levels of Wnt5a [22]. The impact of Wnt5a on myeloid blood cells is, however, unclear. In our current flow cytometric analysis of peripheral blood from patients with sepsis (Fig. 1), we found a significant increase in both the non-classical CD14CD16++ and intermediate CD14+/++ CD16+ monocyte populations (Fig. 1A, B; Fig. 1G gate A and B) and a concomitant decrease in the circulating mDC1 population (Fig. 1D), as compared to samples from healthy controls. The classical CD14++ CD16 monocyte population (Fig. 1C, G; gate C) and the mDC2 and pDC populations (Fig. 1E, F) in the patient samples showed no change compared with controls. Figure 1G shows representative CD14/CD16 dot plots and the corresponding gates used to define the different monocyte populations. Gates were set after isotype controls for CD14 and CD16 (not shown). In a previous gene expression profiling of monocyte differentiation in vitro [23], it was observed that Wnt5a was highly upregulated in the Mo-mDCs as compared to monocytes and monocyte-derived macrophages (Mo-M; also confirmed in this study using Western blot; Fig. S1A). Our results were therefore contradictory to what would be expected if Wnt5a would be beneficial for Mo-mDCs generation.

image

Figure 1. Comparison of the proportions of monocyte populations (non-classical CD14CD16++; intermediate CD14+/++ CD16+; classical CD14++ CD16) in blood samples obtained from patients with sepsis and healthy control subjects. The box plots show the variation and actual change in size of the respective cell populations as per cent of the corresponding populations in peripheral blood. It can be seen that in blood from patients with sepsis, the CD16+ monocyte populations increased (A, B), whereas the classical CD16 monocyte population was not affected (C). N = 10 and N = 10. *P < 0.05. Mann–Whitney Wilcoxon test. (D–F) Comparison of the proportions circulating dendritic cells (mDC1s, mDC2s and pDCs) in blood samples obtained from patients with sepsis and healthy control subjects. The box plots show the variation and actual change in size of the respective cell populations as per cent of the corresponding populations in peripheral blood. It can be seen that in blood from patients with sepsis, the circulating mDC1 population was decreased (D), and the circulating mDC2 (E) and pDC (F) population were not affected. N = 10 and N = 10. Mann–Whitney Wilcoxon test. (G) representative CD14/CD16 dot plots showing the gatings of the different monocyte populations (Gates A: non-classical CD14CD16++; B: intermediate CD14+/++ CD16+; C: classical CD14++ CD16). Gates were set after isotype controls for CD14 and CD16 (not shown).

Download figure to PowerPoint

Wnt5a induces CD14++ monocytes in Mo-mDC cultures

To address whether Wnt5a had an impact on Mo-mDC generation, we therefore incubated primary human monocytes for 7 days under Mo-mDC culture conditions with or without addition of rWnt3a (canonical Wnt signaling control) or rWnt5a, and the outcome was analysed by light microscopy of May-Grunwald Giemsa stained cytospins (Fig. 2A) and flow cytometry (Fig. 2B–D). The cells stained with May-Grunwald Giemsa were relatively similar, except that the rWnt3a-treated cultures contained more neutrophil/granulocyte-like cells with a polymorph nucleus (Fig. 2A, blue arrows in lower middle panel). Also, as shown in Fig. 2A, the rWnt5a-treated cultures contained a significantly increased proportion of monocyte-like cells (black arrow) compared with the rWnt3a-treated cultures (= 0.03; Ctrl 7.6% ± 3.06, rWnt3a 4.6% ± 1.34 and rWnt5a 11.2% ± 3.94).

image

Figure 2. Treatment of primary human Mo-mDC cultures with Wnt5a promotes a CD14++ CD209+ phenotype. (A) The light microscopy pictures shows May-Grunwald Giemsa staining of cytospins prepared from Mo-mDC cultures that were or were not treated with rWnt5a or with rWnt3a as a control; cells with a polymorph nucleus (blue arrows) and cells with monocyte-like morphology (black arrows) are indicated. The numbers in the micrographs represent the proportion (%) of cells with monocyte morphology in one field ± SD (N = 4), and the proportion of such cells was significantly increased in the cultures treated with rWnt5a as compared to those exposed to rWnt3a (P = 0.03). (B) Representative CD14 expression (mean fluorescence intensity, MFI) in the Mo-mDC cultures gated on all cells using 7aad as a dead cell discriminator. (C) Representative dot plots of CD14 and CD209 expression in the Mo-mDC cultures, gated on all cells using 7aad as a dead cell discriminator. The numbers (%) represent quadrant stats. (D) The relative abundance of CD14++ CD209+ (left;% ratio = % CD14++ CD209+ as compared to% CD14++ CD209+ in Ctrl cultures) and CD14low/+ CD209+ (middle;% ratio = % CD14low/+ CD209+ as compared to% CD14low/+ CD209+ in Ctrl cultures) cells (untreated Mo-mDC cultures given a value of 1). The right panel illustrates averages of the actual percentages of CD14low/+ CD209+ cells in the experiments. N = 4. Error bars represent SEM. Student's t-test **P < 0.01.

Download figure to PowerPoint

We next analysed the cells by flow cytometry using the parameters CD14, HLA-DR, CD86, CD16, CD11b, CD1a, CD33, CD11c, CD83 and CD209. For representative dot plots and gatings, see Fig. S1B–C and Fig. S2. The results showed that, in the Wnt5a-stimulated wells, the proportion of CD14++-expressing cells increased, whereas the CD14+/low-expressing population decreased (Fig. 2B–D). Figure S1B shows that the cells in the Mo-mDC differentiation cultures treated with Wnt5a expressed normal levels of HLA-DR and CD86, and Fig. S1C shows that all CD14+ cells in our experiments expressed CD16 and CD11b. CD11c and CD33 were expressed at similar levels as untreated control cells (Fig. S2A or data not shown). Gates were set after isotype controls for all parameters analysed (not shown). All cells expressed CD209, which would indicate that they were Mo-mDCs (Fig. 2C), although they had a low expression of the Mo-mDC marker CD1a (Fig. S1C). The right panel in Fig. 2D shows the actual percentage of cells with a CD14+/lowCD209+ phenotype. Figure S2B–C shows the levels of CD209 expression on CD14+/low as compared to CD14++ cells in each culture, and Fig. S2D shows a representative dot plot of CD14/CD209 on monocytes cultured in Mo-mDC medium using GM-CSF only (i.e. lacking IL-4). Previous studies have demonstrated that CD14CD16+ (CD14+/++ CD16+ and CD14CD16++) monocytes in vivo can also express CD209, and the same was noted for Mo-M cultured in vitro and exposed to IL-6/IL-10 or signals causing an increase in cAMP or cGMP [17, 26-28]. Wnt5a causes an increase in cAMP and has been associated with increased IL-6 production [19, 29]. Activation of the Wnt5a-cultured cells by exposure to LPS led to a further increase in the CD14++ CD1aCD206+ population (gated on CD14+; Fig. 3A) and absence of CD83 expression (Fig. 3B), whereas CD83 was expressed by the untreated and rWnt3a-treated cells.

image

Figure 3. Monocyte-derived mDCs (Mo-mDCs) generated under Ctrl, rWnt3a or rWnt5a stimulating conditions were subjected to LPS activation for 2 days (day 5 of 7 in culture) prior to flow cytometric analysis. LPS treatment of the Mo-mDC cultures further increased the generation of the CD14++ CD206CD1a population in the rWnt5a-treated cultures as shown by the dot plots CD1a/CD206 of CD14+ gated cells (A), and those cells expressed essentially no CD83 compared with the control and rWnt3a-treated cells (B). The histograms show levels of CD83 expression, gated on all cells using 7aad as a dead cell discriminator, in unstimulated cultures (open white) and in LPS-activated cultures (filled grey). The numbers in each histogram represent the mean fluorescence intensity (MFI) of CD83 expression in the LPS-activated cultures. (C) Wnt5a induces monocytes to produce IL-6. A short (3-h) exposure to rWnt5a-induced primary human monocytes (Mo) to secrete IL-6, measured by ELISA and using LPS treatment as a positive control. ***< 0.001. N = 17. Student's t-test. (D) Mo-mDC cultures (GM-CSF + IL-4 for 7 days) were treated with IL-6-blocking antibodies, and the results were compared with controls. The CD14+/low CD209+ cell population (%) showed a relative (fold) increase upon addition of the IL-6-blocking antibodies (see Fig. S4A for CD14/CD209 dot plots and gatings). A value of 1 was given to each of the corresponding control cultures (C, Wnt3a, and Wnt5a without IL-6-blocking antibodies). = 5. Error bars represent SEM. Student's t-test *P < 0.05.

Download figure to PowerPoint

Monocyte subpopulations and Mo-mDCs in various stages of maturation have distinct functional phenotypes [8, 9, 14, 30-34]. The functional outcomes of our experiments indicate that, in Mo-mDC cultures, Wnt5a promoted generation of cells that exhibit the following (for details in procedures see 'Materials and methods'): inefficient pinocytosis of FITC-dextran (Fig. S3A, B) as measured by flow cytometric analyses of Mo-mDC cultured cells that had been exposed to FITC-dextran at 37 °C as compared to 4 °C; relatively good presentation of antigens, as measured by 3H-thymidine incorporation of allogeneic mixed lymphocyte reactions (MLRs; Fig. S3C, D) of CD4+ Naïve T cells cultured with allogeneic Mo-mDCs that had been cultured in the presence of rWnt3a/rWnt5a or not; a pro- and anti-inflammatory cytokine profile of the Mo-mDC cultures that had or had not been treated with rWnt3a or rWnt5a (measured day 7 of culture; enzyme-linked immunosorbent assay, ELISA; Fig. S3E); and normal or slightly elevated levels of HLA-DR and CD86 (Fig. S1B) but no CD1a and CD83 (Fig. S1C and Fig. 3 A, B) as measured by flow cytometry.

Wnt5a-induced IL-6 inhibits differentiation of Mo-mDCs

As mentioned above, Wnt5a causes an increase in cAMP and also that cAMP has been associated with increased IL-6 production per se [19, 29]. We therefore set out to investigate whether Wnt5a could induce IL-6 secretion from freshly prepared primary human monocytes. We could show that Wnt5a did indeed induce a rapid [3] production of IL-6 in primary human monocytes as measured by ELISA and using Wnt3a or LPS as controls (Fig. 3C). We also found that addition of an IL-6-blocking antibody to the various cultures from day 0 restored the formation of Mo-mDCs (CD14+/lowCD209+) in the Wnt5a-treated Mo-mDC cultures (Fig. 3D and Fig. S4A for dot plots and gatings). When we analysed IL-6 levels in serum from patients with sepsis, we found that IL-6 was elevated in serum from the sepsis patient group as compared to healthy controls (Table 1). Although the patients in Table 1 are distinct from the patients analysed in Fig. 1, the level of IL-6 has previously been directly correlated with a decrease in circulating DCs and an increase in monocytes in patients with sepsis [12].

Table 1. Serum from sepsis patients (gram-) display elevated IL-6 levels
Serum IL-6 levels (pg/ml)Healthy controls (N = 13)Sepsis patients (N = 8)
  1. Mann–Whitney U test *P = 0.035 Asymp. Sig. (2-tailed); P = 0.076 Exact Sig. as compared to control group. The minimum detectable concentrations were 0.70 pg/ml.

Subject
1025
200
300
47423
52044
600
7032
8050
90 
100 
110 
120 
130 
Mean Serum IL-6 levels (pg/ml)7.23 (SD ± 20.0)21.7 (SD ± 18.7)*

Wnt5a inhibits ERK1/2-P activity in monocytes

It is known that differentiation of Mo-mDCs is inhibited by activation of STAT3-PY but is promoted by activation of ERK1/2-P [11, 35]. IL-6 signalling can lead to activation of both STAT3-PY and ERK1/2-P in monocytes, depending on the cellular context and the intracellular signalling regulators being expressed [36, 37]. We recently observed that stimulation of primary human monocytes by exposure to rWnt5a-induced STAT3-PY activation [24]. Figure 4A further illustrates that a short [3] exposure of primary human monocytes to Wnt5a led to an inhibition of ERK1/2-P (especially ERK1-P). Interestingly, pretreatment of primary human monocytes with rWnt5a delayed ERK1/2-P elicited by rIL-6 (Fig. 4B lower). This would suggest that Wnt5a could hinder generation of mDCs both through the inhibition of ERK1/2-P and concurrent induction of IL-6.

image

Figure 4. Wnt5a inhibits ERK1/2-P in monocytes. (A) Levels of ERK1/2-P were decreased in primary human monocytes exposed to rWnt5a (3 h) as compared to controls (untreated or treated with rWnt3a). (B) Kinetics of IL-6 induced ERK1/2-P in primary human monocytes pretreated or not with rWnt5a [1] show that Wnt5a delays the IL-6 induced ERK1/2-P. (C–D) Monocytes derived from healthy controls or patients with sepsis were differentiated under Mo-mDC or Mo-M conditions as indicated. CD14 and HLA-DR expression levels were analysed, gated on all cells using 7aad as a dead cell discriminator. (C) Dot plots representing the diverse CD14 and HLA-DR expression in differentiated sepsis as compared to control samples. Dotted lines indicate gatings used to define HLA-DR MFI. The numbers in each gate represent the size (%) of each population. (D) The histograms represent HLA-DR MFI levels (HLA-DR MFI) on CD14+/low (left histogram) or CD14++ (right histogram) differentiated monocytes derived from healthy controls (white bars) as compared to patients with sepsis (black bars). N = 8. Error bars represent SEM. Student's t-test *P < 0.05.

Download figure to PowerPoint

Differentiation of monocytes from sepsis patients promote generation of disparate cells as compared to healthy controls

It is thought that classical monocytes preferentially differentiate into TAMs in a tumour microenvironment [14]. Whether intermediate CD14+/++ CD16+ monocytes preferentially differentiate into certain myeloid subtypes is not yet explored. To evaluate whether differentiation of sepsis patient monocytes would lead to generation of dissimilar cell types, as compared to healthy control monocytes, we performed Mo-mDC or Mo-M cultures using the corresponding samples. Interestingly, as shown in Fig. 4C–D, we found that using CD14 and HLA-DR as phenotypic markers, and gating on CD14+/low or CD14++ cells, sepsis monocytes cultured under Mo-mDC conditions promoted generation of cells with a CD14++HLA-DR+/++ phenotype, while those cultured in Mo-M conditions promoted cells with a CD14+/lowHLA-DRlow or CD14++HLA-DRlow phenotype as compared to healthy controls (Fig. 4C–D). It is interesting to note that in the circulation of both healthy controls and patients with sepsis, the intermediate CD14+/++ CD16+ monocytes display higher levels of HLA-DR expression than both classical CD14++ CD16 and non-classical CD14CD16++ monocytes, but it is the classical CD14++ CD16monocytes that down regulate HLA-DR the most upon sepsis (Fig. 4SB; gates defined in Fig. 1G).

IL-6 derived from Wnt5a-stimulated breast cancer cells inhibit Mo-mDC generation

To evaluate the Wnt5a-induced effects on Mo-mDC differentiation in a different cellular context, we next conducted experiments to treat a Wnt5a-negative but IL-6-expressing breast cancer cell line (MDA-MB-231) with rWnt5a. Thereafter, we added the conditioned medium (CM) from control-stimulated MDA-MB-231 cells (carrier; CM) and compared it with rWnt5a-stimulated MDA-MB-231 cells (20 h; rWnt5a CM), with regard to possible effects on the Mo-mDC differentiation of primary monocytes. Figure 5A shows that in Mo-mDC cultures, addition of CM from MDA-MB-231 cells inhibited Mo-mDC differentiation, while promoting monocytes, as judged by flow cytometric analysis of CD14 and CD1a expression. Interestingly, when the breast cancer cells had been prestimulated with rWnt5a, the inhibition was even more pronounced. The effect was normalized upon blocking IL-6 (Fig. 5A). Increased levels of IL-6 upon rWnt5a stimulation of the already potent IL-6-producing MDA-MB-231 cells were confirmed (Fig. S4C). Also, the level of CD14 expression was affected as shown in Fig. 5B, where addition of rWnt5a prestimulated breast cancer cell media specifically induced cells expressing CD14 at high levels (CD14++) in an IL-6 dependent manner (indicated by dotted line; CD14++ CD206+). IL-6 block antibody control shown in Fig. S4D.

image

Figure 5. Conditioned media (CM) from rWnt5a-stimulated IL-6-producing human breast cancer cells inhibited Mo-mDC differentiation. (A) Primary human monocytes cultured in Mo-mDC cultures with or without addition of conditioned media (CM) from IL-6-producing MDA-MB-231 breast cancer cells prestimulated with rWnt5a [3] (rWnt5a stim CM) or not (CM). The dot plots show CD14 and CD1a expression of the Mo-mDC differentiated cells, gated on all cells using 7aad as a dead cell discriminator. Numbers in the dot plots (%) indicate quadrant stats. Addition of CM inhibited CD1a expression while promoting CD14 expression. This was even more pronounced when the breast cancer cells had been prestimulated with rWnt5a for 3 h. The effect was brought back to relatively normal levels by blocking IL-6. The histograms represent percentage (%) of CD14+ cells (grey bars) and CD1a+ cells (white bars) in the cultures. N = 3. Error bars represent SEM. Student's t-test *P < 0.05; **P < 0.01. (B) As in (A) but analysing CD14 and CD206, which showed that the monocytes expressed CD14 at high levels primarily when the MDA-MB-231 cells had been prestimulated with rWnt5a. The dotted line represents the rWnt5a-induced CD14++CD206+ cell population. The numbers (%) in the dot plots represent quadrant stats. The histogram represents the ratio of CD206+ CD14++ cells (%)/CD206+ CD14−/low cells (%) (dark grey bars) in the cultures. N = 3. Error bars represent SEM. Student's t-test *P < 0.05.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowldgment
  8. Conflict of interest
  9. References
  10. Supporting Information

In this study, we show that Wnt5a can induce IL-6 in monocytes and that this leads to a decreased proportion of Mo-mDC and an increased proportion of CD14+/++CD16+ monocytes generated, in in vitro human Mo-mDC cultures. We correlate this with the increased levels we find of CD14+/++ CD16+ monocytes and serum IL-6, as well as decreased circulating mDC1, in patients with sepsis, a disease with high systemic levels of Wnt5a. Wnt5a is a non-canonical Wnt protein that is expressed at high levels in inflammatory conditions such as cancer, infection [21, 22] and inflammation/auto-immunity [38-40]. Its role in inflammation remains unclear; however, it is known that Wnt5a expression is induced upon TLR-signalling in macrophages [22, 40], and also, Mo-mDCs have been associated with even higher expression levels of Wnt5a [23].

Human monocytes can be divided into three main subpopulations: the classical CD14++ CD16 monocytes; the non-classical CD14CD16++ monocytes; and the intermediate CD14+/++ CD16+ monocytes [8]. It has been reported that CD14+/++ CD16+ and CD14CD16++ cells are induced in patients with both sepsis and breast cancer [12, 41], and reduced levels of circulating DCs have also been observed [42]. An increased proportion of CD16+/++ blood monocytes in patients with cancer was even associated with a worse overall survival [43, 44]. Similarly, increased plasma levels of IL-6 has been associated with a worse overall survival in patients with sepsis and cancer [45, 46].

IL-6 is a difficult cytokine to interpret due to its dual pro- and anti-inflammatory properties [47]. Indeed, IL-6 has been shown to be produced by classically activated macrophages and to under certain conditions induce alternatively activated M2 macrophage differentiation in vitro [48]. The blood from patients with sepsis has high systemic levels of Wnt5a, increased plasma IL-6 and increased proportions of CD14+/++ CD16+ monocytes. This correlation is strengthened by the fact that other patients groups displaying enhanced levels of Wnt5a have increases in IL-6 and either CD14++ CD16+ monocytes or alternatively activated macrophages [38-40, 43]. One explanation to the dual functions of IL-6 has been assigned its downstream signalling cascade leading to activation of either STAT3 or ERK1/2, depending on context. In this study, we show that Wnt5a actually inhibits ERK1/2 activity in monocytes. Furthermore, Wnt5a has previously been correlated with increased levels of IL-6 in malignant melanoma cells and synovial fibroblasts from patients with rheumatoid arthritis [26, 40].

The extravasation of monocytes, and their further differentiation into various myeloid effector cells, is determined by signals from tissues that may differ between pathological situations [13, 14]. One example is cancer, where IL-6 and M-CSF produced by cancer cells, has been shown to inhibit mDC differentiation while promoting the generation of monocytes [16, 17]. In this study, we present that treatment of Mo-mDC cultures with rWnt5a, hinders generation of Mo-mDCs while promoting CD14+/++ CD16+ monocytes. We also show that this is dependent on the Wnt5a-induced production of IL-6 in the monocyte cultures. As Wnt5a has previously been correlated with induction of inflammatory mediators [40], we propose that Wnt5a is an important cytokine inducer that depending on cell context or the environment will have different effects depending on the local mediators induced by Wnt5a. In support of this, we have previously published that in a pro-inflammatory Mo-M1 environment (monocyte-derived macrophages in M1 conditions), Wnt5a promoted generation of tolerogenic Mo-M with a CD14+/++HLA-DR−/low phenotype [24]. This is reflected by the findings that in blood from patients with sepsis, a disease displaying a profound systemic cytokine release, we find an increase in both non-classical CD14CD16++ and intermediate CD14+/++ CD16+ monocytes expressing HLA-DR at high versus low levels, respectively. During the last year, a novel myeloid derived cell population has been described, the neutrophil-DC hybrid [49, 50]. We cannot exclude that some of the cells we find upon rWnt5a or especially rWnt3a treatment might be similar to these hybrid cells as observed by both morphology and phenotype.

In conclusion, we propose that Wnt5a-induced IL-6 and concurrent inhibition of ERK1/2 activity contributes to the inhibition of Mo-mDC differentiation. This introduces Wnt5a as a possible candidate mediator for the CD14+/++ CD16+ monocyte accumulation in patients with infectious disease and cancer.

Acknowldgment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowldgment
  8. Conflict of interest
  9. References
  10. Supporting Information

The authors thank Ms Elise Nilsson for May-Grunwald Giemsa and IHC staining. We also thank Patricia Ödman for linguistic revision. This work was supported by the Swedish Cancer Foundation, the Medical Research Council, the SUS Research Foundations, MAS Cancer research foundation, Gunnar Nilssons Cancer Foundation, Ollie och Elof Ericssons Foundation, Gyllenstiernska Krapperup foundation, Kocks foundations, Alfred Österlunds foundation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowldgment
  8. Conflict of interest
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowldgment
  8. Conflict of interest
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
sji12075-sup-0001-FigureS1.epsimage/eps4516KFigure S1. (A) Levels of Wnt5a protein expression in primary human monocytes. (B–C) Flow cytometric analyses of Mo-mDC cultures treated or not with rWnt3a or rWnt5a.
sji12075-sup-0002-FigureS2.epsimage/eps3811KFigure S2. (A) Representative dot plots of Wnt5a treated Mo-mDC cultures. (B–C) CD209 expression on CD14+/low or CD14++ expressing cells in Mo-mDC cultures. (D) CD14/CD209 dot plot of Wnt5a treated Mo cultures using GM-CSF only (and not IL-4).
sji12075-sup-0003-FigureS3.epsimage/eps6014KFigure S3. (A–E) Functional assays of cells in Mo-mDC cultures untreated or treated with rWnt3a or rWnt5a. (A–B) Pinocytosis by human Mo-mDC measured as FITC-dextran uptake. (C–D) Allogeneic mixed lymphocyte reaction (MLR) of primary human Mo-mDCs and T cells. (E) Mo-mDC secretion of IL-6 (left), IL-12 (middle), and IL-10 (right) measured by ELISA.
sji12075-sup-0004-FigureS4.epsimage/eps9178KFigure S4. (A) Representative CD14/CD209 dot plots and gates for Fig 3D. (B) Flow cytometric analyses of HLA-DR expression levels (MFI) on monocyte populations. (C) Wnt5a induced IL-6 production by MDA-MB-231 breast cancer cells. (D) Antibody control cultures for experiments in Fig. 5
sji12075-sup-0005-Figure-captions.docxWord document135K 

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.