This protocol provides researchers with a rapid, indirect method of measuring TLR-dependent NF-кB/AP-1 transcription factor activity in a murine macrophage cell line in response to a variety of polymeric surfaces and adsorbed protein layers that model the biomaterial implant microenvironment.
The persistent inflammatory host response to an implanted biomaterial, known as the foreign body reaction, is a significant challenge in the development and implementation of biomedical devices and tissue engineering constructs. Macrophages, an innate immune cell, are key players in the foreign body reaction because they remain at the implant site for the lifetime of the device, and are commonly studied to gain an understanding of this detrimental host response. Many biomaterials researchers have shown that adsorbed protein layers on implanted materials influence macrophage behavior, and subsequently impact the host response. The methods in this paper describe an in vitro model using adsorbed protein layers containing cellular damage molecules on polymer biomaterial surfaces to assess macrophage responses. An NF-кB/AP-1 reporter macrophage cell line and the associated colorimetric alkaline phosphatase assay were used as a rapid method to indirectly examine NF-кB/AP-1 transcription factor activity in response to complex adsorbed protein layers containing blood proteins and damage-associated molecular patterns, as a model of the complex adsorbed protein layers formed on biomaterial surfaces in vivo.
The foreign body reaction (FBR) is a chronic host response that can negatively impact the performance of an implanted material or device (e.g., drug delivery devices, biosensors), through the persistent release of inflammatory mediators and by impeding integration between the implanted material and the surrounding tissue1. This innate immune response is initiated by the implantation procedure and is characterized by the long-term presence of innate immune cells and fibrous capsule formation around the implant1. Within the context of material host responses, macrophage-material interactions have a significant impact on the progression of the host response and development of a FBR1. Macrophages are a diverse innate immune cell population, recruited to the implant site either from tissue-resident macrophage populations or from the blood as monocyte-derived macrophages. They begin to accumulate at the implant site shortly after implantation, and within days become the predominant cell population in the implant microenvironment. Material-adherent macrophages, along with foreign body giant cells (FBGC) formed through macrophage fusion, can persist at the material surface for the lifetime of the implant2,3. Consequently, macrophages are considered to be key players in the foreign body response due to their roles orchestrating the characteristic steps of the FBR: acute inflammatory response, tissue remodeling, and formation of fibrotic tissue1.
Toll-like receptors (TLRs) are a family of pattern recognition receptors that are expressed by many immune cells, including macrophages, and have been shown to play a significant role in inflammation and wound healing. In addition to pathogen-derived ligands, TLRs are able to bind endogenous molecules, known as damage-associated molecular patterns (DAMPs), which are released during cell necrosis and activate inflammatory signaling pathways resulting in the production of proinflammatory cytokines4. We and others have proposed that damage incurred during soft tissue biomaterial implantation procedures release DAMPs, which then adsorb to biomaterial surfaces in addition to blood proteins and modulate subsequent cell-material interactions5,6. When macrophages interact with the adsorbed protein layer on an implant, their surface TLRs may recognize adsorbed DAMPs and activate proinflammatory signaling cascades, leading to NF-κB and AP-1 transcription factor activation and production of proinflammatory cytokines. We have previously shown that murine macrophages have significantly increased NF-κB/AP-1 activity and tumor necrosis factor α (TNF-α, proinflammatory cytokine) secretion in response to DAMP-containing adsorbed protein layers on a variety of polymeric surfaces compared to surfaces with adsorbed serum or plasma only (i.e., no DAMPs present), and that this response is largely mediated by TLR2, while TLR4 plays a lesser role5.
The NF-κB/AP-1 reporter macrophage cell line (Table of Materials) used in this protocol is a convenient method to measure relative NF-κB and AP-1 activity in macrophages5,7,8. In combination with TLR pathway inhibitors, this cell line is a useful tool for investigating TLR activation and its role in inflammation in response to a variety of stimuli5,7,8. The reporter cells are a modified mouse macrophage-like cell line that can stably produce secreted embryonic alkaline phosphatase (SEAP) upon NF-κB and AP-1 transcription factor activation9. The colorimetric enzymatic alkaline phosphatase assay (Table of Materials) can then be used to quantify relative amounts of SEAP expression as an indirect measure of NF-κB/AP-1 activity. As NF-κB and AP-1 are downstream of many cell signaling pathways, neutralizing antibodies and inhibitors targeting specific TLRs (e.g., TLR2) or TLR adaptor molecules (e.g., MyD88) can be used to verify the role of a specific pathway. The methodology described in this article provides a simple and rapid approach for assessing the contribution of TLR signaling in murine macrophage responses to a variety of polymeric surfaces with adsorbed protein layers containing both blood proteins and DAMPs as an in vitro model of implanted biomaterials.
1. Media and Reagent Preparation
2. Coating Cell Culture Surfaces with Poly(methyl methacrylate)
3. Coating Cell Culture Surfaces with Polydimethylsiloxane
4. Coating Cell Culture Surfaces with Fluorinated Poly(tetrafluoroethylene)
5. Making Lysate from 3T3 Cells
6. Assessing Effect of Adsorbed Protein Layers and Toll-like Receptors on NF-κB Activity of Macrophages
NOTE: For a schematic of the experimental workflow and plate layout, refer to Figure 1A and Supplemental Figure 1, respectively.
Cleaning methods for the polymer-coated surfaces were tested to ensure there was no disruption of the coating, which would be seen as a change in the water contact angle to an uncoated glass coverslip (Figure 2). Soaking PMMA-coated microscope slides in 70% ethanol for 1 h was found to remove the PMMA coating (Figure 2, left panel), likely due to the solubility of PMMA in 80 wt% ethanol13, therefore PMMA-coated surfaces were cleaned using 30 min of UV sterilization alone. The concentration of PMMA for coating was optimized previously5. A 1 h 70% ethanol soak was used to clean PDMS, and UV sterilization was neglected since UV light can cause chain scission and influence the surface wetting properties of PDMS14. Both 70% ethanol soak and UV sterilization did not influence the water contact angle of fPTFE-coated coverslips (Figure 2, right panel), therefore the two methods, in succession, were used to clean fPTFE coatings. The method of fPTFE coating was previously described by the Grainger group15.
A Western blot was performed on the 3T3 lysate to ensure that DAMP species were present in the complex molecular mixture. The results showed that both HMGB1 and HSP60, two well-documented DAMPs16,17, were present in the lysate (Figure 1B). The adsorption of TLR ligands from the lysate onto the polymer surfaces was confirmed by culturing reporter macrophages (untreated, TLR2 neutralized, or TLR4 inhibited) for 20 h on protein-adsorbed polymer surfaces (i.e., tissue culture-treated polystyrene [TCPS], PMMA, PDMS, fPTFE), and then indirectly assessing NF-кB/AP-1 activity based on SEAP production using an enzymatic assay (Figure 1C and Figure 3). Furthermore, the reporter macrophages had significantly increased NF-кB/AP-1 activity on adsorbed lysate compared to adsorbed FBS or plasma and no pre-adsorbed protein (media) (Figure 4). TLR ligands synthetic triacylated lipopeptide (Pam3CSK4, TLR2 ligand) and lipopolysaccharide (LPS, TLR4 ligand) were included as positive controls to confirm the antibody or inhibitor and the assay were working properly. TLR2 neutralization had a noticeably stronger reduction in NF-кB/AP-1 response of reporter macrophages to adsorbed lysate compared to TLR4 inhibition. As well, small amounts of lysate diluted in serum (based on total protein) induced significantly increased NF-кB/AP-1 response compared to serum alone, with the lowest effective dilution dependent on the polymer surface (Figure 5). These results demonstrate the potency of the adsorbed lysate-derived molecules on inducing TLR-dependent NF-кB/AP-1 activity in reporter macrophages on a variety of polymeric surfaces.
Figure 1: Methods and results for the alkaline phosphatase assay of NF-кB/AP-1 reporter macrophages on TCPS, PMMA, PDMS, and fPTFE. (A) Diagram of the workflow for the reporter macrophage alkaline phosphatase assay. (B) Western blot of lysate confirming the presence of DAMP species HMGB1 and HSP60, with β-actin as the loading control. (C) NF-кB/AP-1 activity (represented by absorbance) of reporter macrophages cultured on media (negative control), 10% FBS, lysate, and Pam3CSK4 (TLR2 ligand, positive control) for 20 h. Data shows the results of one experiment and is representative of results from at least 2 separate experiments, shown as mean ± standard deviation (SD). Each experiment used n = 3 separate wells per condition, and each well was plated in duplicate for the enzymatic assay. Analyzed using one-way ANOVA and Tukey post-hoc test. *** p < 0.001. This figure has been adapted with permission from McKiel and Fitzpatrick5. Copyright 2018 American Chemical Society. Please click here to view a larger version of this figure.
Figure 2: Optimization of cleaning methods for PMMA- and fPTFE-coated surfaces, assessed using water contact angle (WCA). Measurements were taken on 2 separate spots of at least 3 coverslips. Data is shown as mean ± SD. Analyzed using one-way ANOVA and Tukey post-hoc test. * p < 0.05. This figure has been adapted with permission from McKiel and Fitzpatrick5. Copyright 2018 American Chemical Society. Please click here to view a larger version of this figure.
Figure 3: TLR-mediated NF-кB/AP-1 activity (represented by absorbance) of reporter macrophages cultured on 10% FBS (control), lysate, and positive control for 20 h. (A) Influence of TLR2 neutralization on reporter macrophage responses to adsorbed lysate. Positive control is Pam (Pam3CSK4, TLR2 ligand). (B) Influence of TLR4 inhibition on reporter macrophage response to adsorbed lysate. Positive control is LPS (TLR4 ligand). Data shows the results of one experiment and is representative of results from at least 2 separate experiments, shown as mean ± SD. Each experiment used n = 3 separate wells per condition, and each well was plated in duplicate for the enzymatic assay. Analyzed using one-way ANOVA and Tukey post-hoc test. ** p < 0.01, *** p < 0.001. This figure has been adapted with permission from McKiel and Fitzpatrick5. Copyright 2018 American Chemical Society. Please click here to view a larger version of this figure.
Figure 4: NF-кB/AP-1 activity (represented by absorbance) of reporter macrophages cultured on media (negative control), 30 min and 24 h adsorbed protein layers, and Pam3CSK4 (positive control) on TCPS for 20 h. Data is combined from 3 separate experiments and shown as mean ± SD. Each experiment used n = 3 separate wells per condition, and each well was plated in duplicate for the enzymatic assay (i.e., n = 9 non-independent cell culture wells and n = 18 non-independent enzymatic assay wells). Analyzed using one-way ANOVA and Tukey post-hoc test. *** p < 0.001. Please click here to view a larger version of this figure.
Figure 5: Reporter macrophage NF-кB/AP-1 activity (represented by absorbance) after 20 h in response to dilutions of lysate in FBS (total protein = 280 µg/well) adsorbed to polymer surfaces for 30 min. (A) TCPS. (B) PMMA. (C) PDMS. (D) fPTFE. Data shows the results of one experiment and is representative of results from at least 2 separate experiments, shown as mean ± SD. Each experiment used n = 3 separate wells per condition, and each well was plated in duplicate for the enzymatic assay. Analyzed using one-way ANOVA and Tukey post-hoc test. * p < 0.05, ** p < 0.01, *** p < 0.001. This figure has been adapted with permission from McKiel and Fitzpatrick5. Copyright 2018 American Chemical Society. Please click here to view a larger version of this figure.
Supplemental Figure 1: Example layouts used for NF-кB/AP-1 reporter macrophage cell culture assay in 8-chamber and 48-well plate formats. Please click here to view a larger version of this figure.
A primary focus of our lab is the host response to solid biomaterial soft tissue implants, and in particular how the cellular damage incurred during the implantation procedure impacts the host response. The work presented here describes preliminary experiments using a reporter macrophage cell line and in vitro-generated DAMP-containing cellular lysate, to investigate the influence of molecules released during cellular damage (i.e., from the implant surgery) on macrophage responses to biomaterials. Fibroblast cell lysate was used to model the cellular damage and release of DAMPs due to biomaterial placement. Fibroblasts were chosen to create the lysate because of the prevalence of fibroblasts in soft tissue, as well as their ability to secrete a variety of extracellular matrix (ECM) proteins, including fibronectin18. Freeze-thaw cycling was chosen as the method of lysis to produce both intracellular and ECM-derived DAMPs, similar to what would be present in the implant environment. Protease inhibitors were not used to make this lysate. While uncontrolled cell lysis like freeze-thaw cycling can result in the release of proteases that may degrade DAMPs, these enzymes would also likely be present in the biomaterial implant environment when cells are damaged during the implantation procedure. The presence of DAMPs in the complex molecular mixture of the lysate was confirmed by Western blot (Figure 1B; HMGB1 and HSP60) and SEAP reporter assay (Figure 1C; NF-кB/AP-1 activity in response to adsorbed lysate). We have also performed assays where lysate was diluted in FBS based on total protein concentration and adsorbed onto cell culture surfaces (Figure 5) to better reflect the complexity of the implant environment, since it will contain an abundance of blood proteins as well as DAMPs6. Reporter macrophage NF-кB/AP-1 activity remained significantly increased on adsorbed layers from lysate diluted in FBS, and the lowest dilution to achieve significant activation was surface dependent, ranging from 0.1% (TCPS) to 10% (PDMS and fPTFE).
The polymers PMMA, PDMS, and PTFE were chosen for this work because they are nondegradable and have been used extensively in the literature to assess protein adsorption and macrophage response to biomaterials19,20,21,22,23,24,25. TCPS was also used for comparison since it is a common substrate used for in vitro macrophage and TLR signaling work21,26,27,28. The materials used in our work are representative examples of non-degradable, solid biomaterials. However, many other materials could be used with this model, provided the material can be coated onto cell culture plates or microscope slides and properly decontaminated. The NF-κB/AP-1 reporter macrophage cell line was selected for this in vitro model because it enables rapid, indirect measurement of NF-κB/AP-1 activity through the NF-κB/AP-1 inducible expression of SEAP. NF-κB/AP-1 reporter macrophages require the use of phleomycin D1 in the culture media as a selective antibiotic to ensure that only cells with the NF-кB/AP-1 inducible SEAP gene are present29. For the alkaline phosphatase assay, it is critical to use HI-FBS in the cell culture media to avoid potential false positive results generated by alkaline phosphatases present in serum. Our research to date suggests that FBS-adsorbed surfaces do not generate a detectable false positive result, likely because the serum molecules are strongly adsorbed to the culture surface and are not released into the supernatant. The culture time point for the reporter macrophages (20 h), assay incubation timepoint (2.5 h), and absorbance reading wavelength (635 nm) for the alkaline phosphatase assay were optimized with this system to ensure robust and reproducible measurements for all conditions.
An initial protein adsorption timepoint of 30 min was chosen for this work due to its common use in protein adsorption literature (Figure 1C)30,31,32,33,34. However, we have also explored longer adsorption times (i.e., 60 min and 24 h, Figure 4) to better represent the adsorbed protein layer that macrophages would interact with in vivo, which is likely to occur 4−24 h following implantation1. It has been postulated that the majority of protein adsorption and exchange occurs in the first 60 min of exposure to a surface26,35,36, therefore a 60 min adsorption time may be a more relevant timepoint. We have also moved from using FBS as a negative control for the presence of DAMPs in the adsorbed protein layer to commercial mouse plasma. The rationale for using plasma instead of serum is that plasma proteins are known to play significant roles in protein adsorption and macrophage response1, and that plasma provides a better representation of the proteins in the wound environment. Plasma used in protein adsorption experiments is commonly prepared as a 1−10% dilution26,36,37, which motivated our use of 10% plasma. Human plasma is commonly used26,36, as it is easier to obtain in large quantities and more clinically relevant, compared to mouse plasma. However, we chose to use commercial mouse plasma for in this model to keep the species of the protein solutions consistent with that of the reporter cells.
The use of the reporter macrophage cell line introduced some limitations within the study. First, using a murine leukemic macrophage cell line has inherent limitations, as the phenotype and behavior may vary from primary macrophage cultures. While this limitation will be addressed in future work using primary macrophages, the parental macrophage cell line was shown to closely mimic mouse bone marrow-derived macrophages in terms of their cell surface receptors and response to microbial ligands for TLRs 2, 3 and 438. Furthermore, the NF-κB/AP-1 reporter macrophages yielded similar results in response to HMGB1 and LPS stimulation when compared to peritoneal primary murine macrophages39. It should be noted that the NF-κB/AP-1 reporter macrophages, and their parental strain, do not express TLR540. Researchers have demonstrated that HMGB1 was able to activate NF-κB transcription factors via the TLR5 signaling pathways in HEK-293 cells stably transfected with human TLR541. Therefore, the contribution of HMGB1-TLR5 signaling to overall NF-κB activity on lysate-coated surfaces was neglected in this model. Additionally, the reporter macrophages and their parental strain do not express the ASC adaptor protein, and consequently do not form most types of inflammasomes and cannot process inactive IL-1β or inactive IL-18 to their mature forms42. Therefore, the model we have used does not account for the contribution of ASC-dependent inflammasome activity and subsequent autocrine IL-1β and IL-18 signaling in macrophage responses to lysate-adsorbed surfaces. Consequently, this assay is intended as a preliminary examination of TLR-dependent NF-κB activation, and subsequent research using primary macrophages is recommended to provide a more complete and representative understanding of macrophage activation and phenotype on material surfaces of interest.
The alkaline phosphatase assay indirectly measures the NF-кB/AP-1 activity of the reporter macrophages. However, there are many signaling pathways other than TLRs that involve NF-кB/AP-1 (e.g., interleukin-1 receptor [IL-1R]43 and tumor necrosis factor receptor [TNFR]44). Therefore, it was necessary to assess the contribution of TLR2 and TLR4 signaling in the increased NF-кB/AP-1 response to lysate-adsorbed surfaces using inhibition assays (Figure 3). The rationale for selecting these two surface TLRs was that at least 23 DAMPs that have been shown to signal through TLR2 and TLR445, including the well-characterized HMGB1, and both receptors are expressed on the cell surface and can interact directly with the biomaterial surface6. The TLR2 and TLR4 inhibition assays demonstrated that when TLR2 or TLR4 signaling were blocked, the NF-кB/AP-1 response of the reporter macrophages to adsorbed lysate was reduced, indicating that both pathways are involved. However, there was a noticeably larger reduction in NF-кB/AP-1 activity when TLR2 signaling was neutralized, suggesting that TLR2 may play a primary role in the response of reporter macrophages to adsorbed lysate. We recognize that there may be some off-target inhibition with the TLR signaling pathway neutralizing antibodies and inhibitors. A neutralizing antibody was used to inhibit the TLR2 pathway since there were not commercially available TLR2 inhibitor molecules at the time of this work.
The methods presented here use lysate, as a complex source of DAMPs, and NF-κB/AP-1 reporter macrophages as an in vitro model for macrophage responses to DAMPs and other proteins adsorbed to polymeric biomaterials (Figure 1). We anticipate our protocol can be used to quickly analyze NF-κB/AP-1 responses and upstream TLR signaling of reporter macrophages to a variety of materials (including degradable materials, porous scaffolds or hydrogels) and adsorbed protein layers (Figure 3). However, the use of porous materials and hydrogels will introduce complexity within the system, as it may be challenging to distinguish between adsorbed molecules and entrained molecules. We also anticipate that this protocol can be easily adapted to investigate the contribution of other signaling pathway upstream of NF-кB/AP-1 (e.g., C-type lectin receptors46 and nucleotide-binding oligomerization domain (NOD)-like receptors47) with the appropriate inhibitors. Furthermore, the NF-κB/AP-1 response of reporter macrophages could be compared between different materials, provided responses are normalized to baseline cell activity (i.e., cells in media on each surface with no pre-adsorbed protein) and all materials have undetectable endotoxin levels.
The authors have nothing to disclose.
The authors gratefully acknowledge operational funding from Canadian Institutes of Health Research Project (PTJ 162251), Queen's University Senate Advisory Research Committee and infrastructure support from the Canadian Foundation for Innovation John Evan's Leadership Fund (Project 34137) and the Ministry of Research and Innovation Ontario Research Fund (Project 34137). L.A.M. was supported by a Queen's University R. Samuel McLaughlin Fellowship, a Natural Sciences and Engineering Research Council of Canada Canadian Graduate Scholarship Master's Award and an Ontario Graduate Scholarship. The authors would like to thank Dr. Myron Szewczuk for his generous gift of the NF-κB/AP-1 reporter macrophage cell line and Drs. Michael Blennerhassett and Sandra Lourenssen for the use of their gel imaging system and plate reader.
Cell culture reagents | |||
anti-mouse/human CD282 (TLR2) | Biolegend | 121802 | |
CLI-095 (TLR4 inhibitor) | Invivogen | TLRL-CLI95 | |
C57 complement plasma K2 EDTA 10ml, innovative grade US origin | InnovativeResearch | IGMSC57-K2 EDTA-Compl-10ml | Mouse plasma |
Dulbecco's modified eagle medium (DMEM) | Sigma Aldrich | D6429-500ML | |
Dulbecco's phosphate buffered saline (DPBS) | Fisher Scientific | 14190250 | No calcium, no magnesium |
Fetal bovine serum (FBS), research grade | Wisent | 98150 | |
LPS-EK | Invivogen | TLRL-EKLPS | Lipopolysaccharide from Escherichia coli K12 |
NIH/3T3 fibroblasts | ATCC | CRL-1658 | |
Pam3CSK4 | Invivogen | tlrl-pms | Synthetic triacylated lipopeptide – TLR1/2 ligand |
Penicillin/streptomycin | Sigma Aldrich | P4333-100ML | |
Plasmocin | Invivogen | ANT-MPP | Mycoplasma elimination reagent |
RAW-Blue cells | Invivogen | raw-sp | NF-κB/AP-1 reporter macrophage cell line |
Trypan blue solution, 0.4% | Fisher Scientific | 15250061 | |
TrypLE express enzyme (1X) | Fisher Scientific | 12604021 | animal origin-free recombinant cell dissociation enzyme |
Zeocin | Invivogen | ANT-ZN-1 | |
Kits and assays | |||
ELISA precoated plates, mouse IL-6 | Biolegend | B213022 | |
ELISA precoated plates, mouse TNF-α | Biolegend | B220233 | |
Endotoxin (Escherichia coli) – Control standard endotoxin (CSE) | Associates of Cape Cope Inc. | E0005-5 | Endotoxin for standard curve in chromogenic endotoxin assay |
LAL water, 100 mL | Associates of Cape Cope Inc. | WP1001 | Used with chromogenic endotoxin assay |
Micro BCA protein assay | Fisher Scientific | PI23235 | |
Limulus amebocyte lysate (LAL) Pyrochrome endotoxin test kit | Associates of Cape Cope Inc. | C1500-5 | Chromogenic endotoxin assay reagent |
QUANTI-Blue alkaline phosphatase detection medium | Invivogen | rep-qb2 | Alkaline phosphatase assay to indirectly measure NF-κB/AP-1 activity |
Polymeric coating reagents | |||
Chloroform, anhydrous | Sigma Aldrich | 288306-1L | |
Ethyl alcohol anhydrous | Commercial Alcohols | P006EAAN | Sigma: Reagent alcohol, anhydrous, 676829-1L |
Straight tapered fine tip forceps | Fisher Scientific | 16-100-113 | |
Fluorinert FC-40 solvent | Sigma Aldrich | F9755-100ML | Fluorinated solvent for fPTFE |
Cell culture grade water (endotoxin-free) | Fisher Scientific | SH30529LS | |
Poly(methyl methacrylate) (PMMA) | Sigma Aldrich | 182230-25G | |
Sylgard 184 elastomer kit | Fisher Scientific | 50822180 | |
Teflon-AF (fPTFE) | Sigma Aldrich | 469610-1G | Poly[4,5-difluoro-2,2-bis(trifluoromethyl)-1,3-dioxole-co-tetrafluoroethylene] |
Consumables | |||
Adhesive plate seals | Fisher Scientific | AB-0580 | |
Axygen microtubes, 1.5 mL | Fisher Scientific | 14-222-155 | |
Borosilicate glass scintillation vials, with white polypropylene caps | Fisher Scientific | 03-337-4 | |
Clear PS 48-well plate | Fisher Scientific | 08-772-52 | |
Clear TCPS 96-well plate | Fisher Scientific | 08-772-2C | |
Clear TCPS 48-well plate | Fisher Scientific | 08-772-1C | |
Cover glasses, circles | Fisher Scientific | 12-545-81 | |
Falcon tissue culture treated flasks, T25 | Fisher Scientific | 10-126-10 | |
sticky-Slide 8 Well | Ibidi | 80828 | |
Superfrost microscope slides | Fisher Scientific | 12-550-15 | |
Tissue culture treated flasks, T150 | Fisher Scientific | 08-772-48 |