Flow-Induced Shear Stress Primes NLRP3 Inflammasome Activation in Macrophages via Piezo1

Abstract

The NLRP3 inflammasome is a crucial component of the innate immune system, playing a pivotal role in initiating and regulating the body’s inflammatory response to various pathogens and cellular damage. Environmental stimuli, such as temperature, pH level, and nutrient availability, can influence the behavior and functions of innate immune cells, including immune cell activity, proliferation, and cytokine production. However, there is limited understanding regarding how mechanical forces, like shear stress, govern the intrinsic inflammatory reaction, particularly the activation of the NLRP3 inflammasome, and how shear stress impacts NLRP3 inflammasome activation through its capacity to induce alterations in gene expression and cytokine secretion. Here, we investigated how shear stress can act as a priming signal in NLRP3 inflammasome activation by exposing immortalized bone marrow-derived macrophages (iBMDMs) to numerous physiologically relevant magnitudes of shear stress before chemically inducing inflammasome activation. We demonstrated that shear stress of large magnitudes was able to prime iBMDMs more effectively for inflammasome activation compared to lower shear stress magnitudes, as quantified by the percentage of cells where ASC-CFP specks formed and IL-1β secretion, the hallmarks of inflammasome activation. Testing this in NLRP3 and caspase-1 knockout iBMDMs showed that the NLRP3 inflammasome was primarily primed for activation due to shear stress exposure. Quantitative polymerase chain reaction (qPCR) and a small-molecule inhibitor study mechanistically determined that shear stress regulates the NLRP3 inflammasome by upregulating Piezo1, IKKβ, and NLRP3. These findings offer insights into the mechanistic relationship among physiological shear stresses, inflammasome activation, and their impact on the progression of inflammatory diseases and their interconnected pathogenesis.

Graphical Abstract

INTRODUCTION

Chronic inflammatory diseases are the leading cause of death worldwide, and the World Health Organization cites them as the greatest threat to human health.^1,2 This class of diseases is regulated by a common protein complex called the NLRP3 inflammasome by responding to microbial invasion or stress and danger signals in innate immune cells.^3–6 Aberrant activation of the NLPR3 inflammasome can result in a plethora of diseases, including rheumatoid arthritis, chronic obstructive pulmonary disease, type II diabetes, Alzheimer’s disease, nonalcoholic steatohepatitis, inflammatory bowel disease, COVID-19, and cancers.^7–9 The current macroscale in vitro assays lack physiological relevance to how these diseases manifest in the body, and there is a need for in vitro assays that accurately recapitulate organ and tissue structure in drug development.¹⁰ There is also a need for a method that can more accurately bridge the gap between these 2D in vitro assays and animal models to clinical trials since approximately three-quarters of the cost pharmaceutical companies face in research and development is due to failure.¹¹ Failure often occurs due to clinical translation issues due to inaccurate computational, animal, or traditional cell culture models.^12,13 To fill this need in therapeutic development, microfluidic devices are utilized because they surpass traditional in vitro models by providing a platform to simulate in vivo 3D conditions through controlling flow rates, pressure, and concentration gradients, as well^10,as¹⁴ the ability to use human cell lines for in vitro models. One of the most useful advantages of microfluidic systems is their ability to offer a high-throughput screening method to validate the efficacy of therapeutics in a 3D environment.^15,16 Microfluidic devices have also been increasingly used to study inflammatory diseases due to their advantages for simulating in vivo environments in an in vitro setting.¹⁶ One of the physiological parameters that can be controlled with microfluidic devices in an in vitro setting is the flow rate and the resulting shear stress that can be applied to cells. The effect of shear stress on cells is widely studied, where shear stress can affect cell morphology, cytokine secretion, and gene expression; however, little is known about how shear stress plays a role in the inflammatory responses of immune cells.^17–19

Transcription factor NF-kB, which regulates NLRP3 inflammasome activation, has been shown in several cell types to be regulated via physiologically relevant shear stress magnitudes through receptor Piezo1, a mechanically sensitive ion channel that regulates protein synthesis, apoptosis, and protein secretion.^17,20–26 Canonical NLRP3 inflammasome activation with chemical stimulation occurs via a two-step signaling process. The priming signal (Signal 1) of NLRP3 inflammasome activation is chemically initiated when a toll-like receptor 4 (TLR4) senses well-conserved pathogen-associated molecular patterns, such as lipopolysaccharide (LPS) found on Gram-negative bacteria, which initiates downstream signaling to activate the transcription factor NF-kB.^6,7,22 NF-kB initiates the production of inactive proteins, which will then be necessary for inflammasome activation, including NOD-, LRR-, and pyrin-domain-containing protein 3 (NLRP3) as well as pro-IL-18 and pro-IL-1β. Pro-caspase-1, inactive gasdermin D, and ASC (apoptosis-associated speck-like protein-containing CARD) are also produced, but the production of these proteins is not regulated by NF-kB.^22,27 Signal 2 can be transduced through several methods, including mitochondrial reactive oxygen species (mROS) production, potassium efflux, calcium influx, lysosomal rupture, or damage-associated molecular patterns (DAMPs). The gold standard for Signal 2 is nigericin, a potassium ionophore which has been widely used to study NLRP3 inflammasome activation.^22,27 Signal 2 transduction results in the oligomerization of inactive proteins to form the NLRP3 inflammasome complex and the activation of the caspase-1 enzyme. Caspase-1 cleaves gasdermin D, pro-IL-18, and pro-IL-1β into their active forms upon Signal 2 for inflammasome activation being transduced. Active gasdermin D forms a pore on the membrane through which IL-18 and IL-1β are exported from the cell, resulting in pyroptotic cell death.^3,22,27 Despite numerous studies investigating how shear stress can regulate the activity of transcription factor NF-kB, no studies have mechanistically examined how physiologically relevant shear stress magnitudes in a microfluidic device system can prime NLRP3 inflammasome activation in macrophages.

In this study, we investigated how priming immortalized bone marrow-derived macrophages (iBMDMs), a widely studied murine macrophage cell line, with five physiologically relevant shear stress magnitudes in a microfluidic device affects NLRP3 inflammasome activation.^28–30 In addition to the magnitude of shear stress exposure, we varied the exposure time and the seeding density of cells in the microfluidic device system to discern how shear stress exposure time and cell-to-cell interactions play a role in shear-stress-primed inflammasome activation. When varying all three parameters, we investigated how ASC–CFP speck formation was affected by shear-stress-priming with differing magnitudes. We confirmed NLRP3 inflammasome activation by measuring the proinflammatory cytokine IL-1β concentration. We also investigated ASC–CFP speck formation in NLRP3 and caspase-1 knockout (KO) iBMDMs to confirm that shear stress regulates only the NLRP3 inflammasome. To discern the mechanism through which shear stress affects the activation of the NLRP3 inflammasome, we performed quantitative polymerase chain reaction (qPCR) and an inhibitor study with Dooku1, a well-characterized Piezo1 inhibitor. We assessed the relative amounts of NLRP3, caspase-1, p50/p105 NF-kB, p52/p100 NF-kB, p65 NF-kB, I kappa B kinase (IKKβ), Piezo1, and IL-1β compared to the β-actin control in iBMDMs after shear stress exposure. To determine how Dooku1 inhibited shear stress, we pretreated iBMDMs with varying concentrations of Dooku1 before treating cells with a high shear stress. We then assessed the ASC–CFP speck formation as well as the gene expression of NLRP3 and Piezo1. Studying how shear stress can induce NLRP3 inflammasome activation in a physiologically relevant system can provide clarity on how the progression of one inflammatory disease can affect the pathogenesis of others.

RESULTS

Shear-Stress-Primed iBMDMs Form ASC–CFP Specks after Nigericin Exposure.

To determine the effect of shear stress exposure on NLRP3 inflammasome activation, iBMDMs were exposed in vitro to five different magnitudes of shear stress ranging from 1.0 to 50 dyn/cm² by varying the flow rate in a syringe pump. Shear stresses around this magnitude have been investigated widely for their role in modulating the activity of transcription factors and the production of cytokines in endothelial cells and T cells.^20,21 Additionally, shear stresses of this magnitude are physiologically relevant compared to those found in the body in the central veins of the liver, capillaries, and arteries under healthy conditions.¹⁷ Shear stresses found in the body span 4 orders of magnitude ranging from 0.1 dyn/cm² in hepatic sinusoid to 1000 dyn/cm² under turbulent flow conditions in the heart, where shear stresses on the upper end of this spectrum have been shown to be detrimental to cell structure.^17,31 While shear stresses on the order of 10³ dyn/cm² can be physiologically relevant in the body, we wanted to focus on laminar flow and more physiologically similar to those observed more widely.¹⁷ In addition to shear stress magnitude, we varied the shear stress exposure time and seeding density of cells in the microfluidic chip to elucidate how shear stress exposure time and cell-to-cell interaction play roles in shear-stress-primed activation of the NLRP3 inflammasome. Previous studies have demonstrated the ability of the shear stress to increase the activation of NF-kB via the IKK enzyme, an upstream element involved in the priming (Signal 1) of NLRP3 inflammasome activation.^17,20–25 However, these studies utilized 2D in vitro methods and lacked a physiologically relevant 3D in vitro method to analyze the effect of shear stress on NF-kB. This study utilized the idenTx 3 microfluidic device from AIM Biotech to study the effect of shear stress on NLRP3 inflammasome activation to mimic the extracellular matrix and recapitulate how inflammatory diseases manifest in the body.

To apply shear stress in this system, we uniformly suspended iBMDMs in a 2.5 mg/mL collagen gel to mimic macrophages accumulated at the site of inflammation or in tissues in the body. Collagen type I was used due to its high presence in the vertebrate extracellular matrix, high force propagation efficiency, and ability to generate higher forces more effectively in comparison to other gels such as fibronectin, laminin, or collagen type IV.³² The cell suspension was injected into the central channel of a microfluidic device and hardened at 37 °C for 30–35 min (Figure 1A,B). A syringe pump applied shear stress by flowing media flanking both sides of the cells where the effluent media was collected in a vessel. Correlations provided by the device’s manufacturer (Figure 1C) were used to equate the flow rates to applied shear stresses in the microfluidic device system. This microfluidic system was designed to mimic the shear stress applied by the blood flow to tissue-resident macrophages in the body under various physiological conditions, as demonstrated by the shear stress magnitudes (1.0–50 dyn/cm²) and exposure time selected for this study (15 and 30 min). We primed iBMDMs with five different magnitudes of shear stress for 15 or 30 min, followed by a 10 μM nigericin treatment for 1 h which initiates the oligomerization of an inactive ASC adaptor protein into an active ASC adaptor protein, often called an ASC speck (Figure 2A,B). Following nigericin treatment, the cells were stained and imaged via confocal microscopy (Figure 2B,C). The iBMDMs used in this study have an ASC adaptor protein engineered to express a cyan fluorescent protein (CFP) tag, so that the formation of ASC–CFP specks could be quantified via confocal microscopy since the ASC speck formation is a well-studied indicator of inflammasome activation.^22,25 In addition to applying shear stress to iBMDMs, we performed static controls to assess the percentage of ASC–CFP speck formation and cell death in iBMDMs primed with 100 ng/mL of LPS, followed by 10 μM nigericin treatment for 1 h (Figure S1A–C). This was important to assess the relative amount of inflammasome activation when LPS priming was performed compared with shear-stress-priming. Untreated, LPS-primed, and nigericin-treated iBMDMs were also analyzed via confocal microscopy as controls to compare with the ASC–CFP speck formation and cell death observed in shear-stress-primed iBMDMs (Figure S1D,E). When primed with shear stress for 15 min, we observed with confocal microscopy analysis that shear stresses higher than 25 dyn/cm²exhibited a significantly higher percentage of ASC–CFP speck formation (Figure 2D), indicating that priming for speck formation is more effective with 25 and 50 dyn/cm² compared to shear stress magnitudes less than 25 dyn/cm². The percentage of cell death due to 15 min shear stress exposure was significantly lower than the nigericin control, regardless of the shear stress magnitude, indicating that 15 min was sufficient for priming iBMDMs for speck formation, while it was not sufficient to initiate cell death in iBMDMs (Figure S2A). When iBMDMs were primed with shear stress for 30 min, confocal microscopy analysis determined that 50 dyn/cm² shear stress caused an eightfold higher percentage of ASC–CFP speck formation compared to the untreated control and threefold higher when compared to the lower magnitudes of shear stress exposure and 1 h nigericin control (Figure 2D,E). Magnitudes of shear-stress-priming from 1.0 to 10 dyn/cm² did not demonstrate ASC–CFP speck formation significantly higher than the 1 h nigericin control for this time of shear stress exposure. These results indicate that compared to the other four shear stress magnitudes, 50 dyn/cm² provided exceptional priming to initiate the formation of ASC–CFP specks. This can be especially seen with the magnified images of shear stress exposure at 1.0 and 50 dyn/cm² (Figure 2F,G). The percentage of cell death, marked by propidium iodide (PI), as a result of ASC–CFP speck formation at the highest shear stress magnitude for 30 min was not significantly different than the 1 h nigericin control (Figure S2B). The difference that we observed between these two treatment groups was that in the case of nigericin treatment, cell death was a byproduct of nigericin toxicity, while shear-stress-primed iBMDMs underwent pyroptosis as a result of inflammasome activation.^4,5,33 However, we observed that iBMDMs primed with 1.0 dyn/cm² shear stress exhibited a significantly lower percentage of cell death than all other groups (Figure S2B). This indicates that shear stress mitigates the effects of nigericin toxicity at this shear stress rate, suggesting that 1.0 dyn/cm² before nigericin treatment can provide a protective effect. This trend has been observed in bioprinting, where shear stress and cell viability have an exponential relationship with shear stress, where high shear stress results in a decrease in cell viability and a decrease in shear stress results in an increase in cell viability.^34,35 Shearstress-priming iBMDMs with 50 dyn/cm² for 30 min demonstrated a significantly higher degree of ASC–CFP speck formation compared to relevant controls and lower shear stresses, yet we still need to investigate the mechanism that shear-stress-priming affects NLRP3 inflammasome activation in macrophages.

Figure 1.

idenTx 3 AIM Biotech microfluidic device schematic and shear stress correlations. (A) Schematic of the microfluidic device is shown where the device contains a central channel as well as two flanking channels. A further magnified image shows that the central channel will contain iBMDMs suspended in a collagen gel and flanking channels where media is flowed to apply shear stress. Created with Biorender.com (B) Dimensions of the AIM Biotech idenTx 3 microfluidic chip provided from the manufacturer’s Web site. (C) Correlations from the microfluidic device’s manufacturer to equate shear stresses to flow rates applied.

Figure 2.

Shear-stress-primed iBMDMs show ASC–CFP speck formation after nigericin exposure. (A) Schematic shows shear stress initiates a conformational change in Piezo1 mechanosensitive ion channel, which activates NF-kB and the production of NLRP3, and proinflammatory cytokines. Shear stress is applied to iBMDMs for 30 min by flowing media on both sides of a microfluidic device. Cells are treated with nigericin for 1 h to act as Signal 2 which initiates the formation of the ASC–CFP speck complex where the inactive ASC protein oligomerizes to form a speck-like-complex. The cells are imaged with ASC–CFP specks indicated with arrows. (B, C) Representative imaging is shown for the microfluidic system where ASC–CFP-expressed iBMDMs were primed with shear stress of varying magnitudes for 15 min (B) and 30 min (C), followed by 1 h treatment of nigericin. Nigericin controls were also done where iBMDMs were treated with nigericin for 1 h. NucRed stains the nucleus of living cells in blue, ASC–CFP specks which indicate inflammasome activation are shown in cyan, and propidium iodide (PI) showing the nucleus of dead cells is shown in red. Scale bar: 100 μm. (D, E) Quantification of ASC–CFP specks after 15 min (D) or 30 min (E) shear stress exposure and 1 h nigericin treatment was normalized by the total number of living cells characterized by the NucRed signal. Data shown are ± SEM (n = 3). Statistical analysis was performed using one-way ANOVA and then a Tukey post-test. **p < 0.01, ***p < 0.001, and ****p < 0.0001. (F, G) Further enhanced representative images of 1.0 dyn/cm² (F) and 50 dyn/cm² (G) shear-stress-primed iBMDMs subsequently treated with nigericin for 1 h. NucRed stains the nucleus of living cells in blue; ASC–CFP specks which indicate inflammasome activation are shown in cyan, and propidium iodide (PI) showing the nucleus of dead cells is shown in red. Scale bar: 100 μm.

We next wanted to investigate if the mechanism involved in shear-stress-priming was affected by varying the seeding density of cells in this microfluidic system. This was done to ensure that the assay was sensitive enough in the various shear stress scenarios and to investigate the role and importance of cell-to-cell signaling in NLRP3 inflammasome activation. A more widely studied example of this is cell-to-cell interactions between circulating cells and resident cells, as is the case in tumor extravasation, which is more widely studied when investigating the effect of shear stress on the ability of tumor cells to extravasate and metastasize to other organs.^17,36–38 However, the cell-to-cell signaling cascade in the context of NLRP3 inflammasome activation due to shear stress has not been widely studied and will provide important insights into how these mechanisms are affected by the NLRP3 inflammasome activation pathway. One study investigated the effect of pannexins, a group of gap-junction proteins that can form intercellular channels, on cell-to-cell communication and determined that pannexin1 can be open due to mechanical stress, such as shear stress, applied to cells and are active at physiological extracellular concentrations of calcium ions creating channels that are conducive to cell-to-cell communication.^39–42 Pannexin 1 facilitates the transport of extracellular calcium ions and mediates the release of ATP, indicating that it has a clear involvement in NLRP3 inflammasome activation and the subsequent secretion of IL-1β.^39,41–45 This indicates that shear stress has the potential to be conducive to increased cell-to-cell communication under conditions of inflammasome activation. To test this hypothesis, we seeded microfluidic devices as previously described with 1 million, 2 million, or 4 million cells/mL to investigate how cell-to-cell interactions play a role in shear-stress-primed inflammasome activation with confocal microscopy (Figure S3A,B). At a seeding density of 1 million cells/mL, we observed no significant changes in ASC–CFP speck formation when cells were exposed to shear stress, regardless of the magnitude (Figure S3C), but we observed a significant increase in cell death at 50 dyn/cm² (Figure S3D). However, at 4 million cells/mL, we observed an increase in the magnitude of shear stress exposed, resulting in a larger percentage of ASC–CFP speck formation (Figure S3E). This was coupled with an increase in PI-positive cells as shear stress increased (Figure S3F). While the cell seeding densities that we selected showed no significant changes in the formation of ASC–CFP specks, we did observe a trend in the percentage of ASC–CFP speck formation at 50 dyn/cm². At 1 million cells/mL, there was a decrease in ASC–CFP speck formation when compared to the 2 million and 4 million cells/mL densities (Figure S3G). This indicates that a change in the density of cells does not impact speck formation over the range of densities selected, but there may be changes in speck formation observed over a wider range of densities. We also observed that as the seeding density increased, the percentage of PI-positive cells increased, indicating that increased cell-tocell proximity under shear stress conditions can result in higher cell death, regardless of the shear stress magnitude (Figure S3H). These studies suggest that at 2 million cells/mL, the shear-stress-primed NLRP3 inflammasome activation assay is sensitive enough to ASC–CFP speck formation, while increasing the seeding density to 4 million cells/mL does not impact the fraction of ASC–CFP speck formation over the range of seeding densities tested. These studies show that the optimum shear stress magnitude, shear stress exposure time, and cell seeding density to induce NLRP3 inflammasome activation were 2 million cells/mL exposed to 50 dyn/cm² over 30 min.

Shear-Stress-Primed ASC–CFP Speck Formation is Regulated by the NLRP3 Inflammasome.

We anticipated that shear stress would only prime the NLRP3 inflammasome for activation and would not affect the priming of the NLRP1 and AIM2 inflammasomes since previous studies suggested that shear stress would not prime these inflammasomes.^24,46 While NF-kB is involved in the priming of the NLRC4 inflammasome pathway, Gram-negative bacteria are required to initiate activation; so, even if shear stress could prime this pathway for activation, the NLRC4 inflammasome could not become activated.^47–49 To confirm this, we separately primed NLRP3 KO and caspase-1 KO iBMDMs with 3 magnitudes of shear stress for 30 min, followed by 1 h nigericin treatment before imaging via confocal microscopy (Figure 3A,B). Similar to the WT iBMDMs used in this study, the KO iBMDMs were engineered to express a CFP tag on the ASC adaptor protein. Untreated, 4 h LPS-primed, 1 h nigericin-treated, and 4 h LPS-primed, followed by 1 h nigericin-treated controls were performed for KO iBMDMs to assess the ASC–CFP speck formation and percentage of PI-positive cells (Figure S4A,B). Again, this modification was made to facilitate the quantification of inflammasome activation via confocal microscopy and to determine how shear-stress-priming in KO iBMDMs affects the formation of ASC–CFP specks. We expected these results to indicate that regardless of the magnitude of shear-stress-priming the KO iBMDMs, there should be a significant reduction in the formation of ASC–CFP specks when compared to the WT iBMDMs exposed to the same shear stress. We observed no significant differences in the percentage of ASC–CFP speck formation when NLRP3 KO or caspase-1 KO cells primed with shear stress were compared to the untreated controls (Figure 3C,D). Additionally, we observed that at all shear stress magnitudes, the percentage of cells that formed ASC–CFP specks was significantly lower for both NLRP3 KO and caspase-1 KO iBMDMs compared with the WT iBMDMs (Figure 3E). Under shear stress conditions, the percentage of PI-positive cells was significantly lower than the nigericin control regardless of the shear stress magnitude for NLRP3 and caspase-1 KO iBMDMs, with the exception of 10 dyn/cm² for caspase-1 KO iBMDMs which was not significantly different (Figure S4C,D). This indicates that the absence of NLRP3 or caspase-1 does not result in an increase in the percentage of PI-positive cells. Additionally, under shear stress conditions, there were no cell-line-specific trends with regard to the percentage of PI-positive cells (Figure 3F), indicating that knocking out NLRP3 or capase-1 did not selectively increase or decrease cell death. We also wanted to measure the concentration of IL-1β with ELISA in the supernatants of cells exposed to shear stress, followed by nigericin treatments as well as other control groups (Figure 3G). We determined that priming cells with 50 dyn/cm² before nigericin treatment resulted in significant IL-1β secretion compared to that of untreated and LPS-primed iBMDMs. These experiments demonstrated that both NLRP3 and caspase-1 are essential for inflammasome activation upon priming iBMDMs with shear stress due to the percentage of ASC–CFP specks being much higher for WT iBMDMs primed with high shear stresses. Neither NLRP3 nor caspase-1 has been evaluated previously as the direct mechanism through which shear stress primes cells, utilizing KO iBMDMs provides insight into the involvement of the NLRP3 inflammasome in this process, demonstrating that shear-stress-priming directly affects the activation of only the NLRP3 inflammasome in iBMDMs under flow conditions.^17,20–25

Figure 3.

Shear-stress-priming only affects NLRP3 inflammasome activation in iBMDMs. (A, B) Representative imaging is shown for the microfluidic system where ASC–CFP-expressed NLRP3 KO iBMDMs (A) and caspase-1 KO iBMDMs (B) were primed with shear stress of varying magnitudes for 30 min, followed by 1 h treatment of nigericin. NucRed stains the nucleus of living cells in blue, ASC–CFP specks which indicate inflammasome activation are shown in cyan, and propidium iodide (PI) showing the nucleus of dead cells is shown in red. Scale bar: 100 μm. (C, D) Quantification of ASC–CFP specks in NLPR3 KO (C) and caspase-1 KO (D) iBMDMs after 30 min shear stress exposure and 1 h nigericin treatment was normalized by the total number of living cells characterized by the NucRed signal. Data shown are ± SEM (n = 3). Statistical analysis was performed using one-way ANOVA and then a Tukey post-test. ns: not significant. (E, F) Quantification of ASC–CFP specks (E) and PI-positive cells (F) after 30 min shear stress exposure and 1 h nigericin treatment was normalized by the total number of living cells characterized by the NucRed signal. Data shown are ± SEM (n = 3). Statistical analysis was performed using a two-way ANOVA, followed by a Tukey post-test. (G) Supernatants were taken from microfluidic device experiments to perform IL-1β ELISA to measure the protein concentration relative to untreated iBMDMs. Data shown are ± SEM (n = 3). Statistical analysis was performed using a one-way ANOVA, followed by a Tukey post-test. *p < 0.05, ***p < 0.001, ****p < 0.0001.

Shear Stress-Primed iBMDMs Show Increases in NLRP3, Piezo1, IKKβ, p65 NF-kB, and p52/p100 NF-kB mRNA.

We next elucidated the mechanism through which shear stress primes the NLRP3 inflammasome for activation. Recent studies have shown that shear stress magnitudes consistent with those used in this study activate NF-kB through an enzyme IKK, which would prime the inflammasome for activation.^17,20–25 To determine the mechanism through which shear stress affects the activation of the NLRP3 inflammasome, qPCR was performed on iBMDMs in the microfluidic chip after 50 dyn/cm² shear stress exposure. This method has been widely used to quantitatively determine the relative amounts of RNA in NLRP3 inflammasome signaling.^50–54 Based on the relative amounts of RNA present in the cells primed with 50 dyn/cm², we should be able to discern the mechanism through which shear stress primes the NLRP3 inflammasome for activation (Figure 4A). We evaluated the relative changes in RNA for NLRP3, caspase-1, p50/p105 NF-kB, p65 NF-kB, p52/p100 NF-kB, IKKβ, IL-1β, and Piezo1 (Figure 4B–I) relative to β-actin in untreated, LPS-primed, and shear-stress-primed iBMDMs. Results from qPCR showed significant increases in NLRP3 (Figure 4B), Piezo1 (Figure 4E), IKKβ (Figure 4G), p52/p100 NF-kB (Figure 4H), and p65 NF-kB (Figure 4I) RNA when exposed to shear stress but no significant changes in IL-1β (Figure 4C), p50/p105 NF-kB (Figure 4D), and caspase-1 (Figure 4F) expression. However, we did observe significant changes in the relative expression of NLRP3, IL-1β, p50/p105 NF-kB, p52/p100 NF-kB, and p65 NF-kB when cells were primed with LPS. A recent study investigated which subunits of NF-kB were affected by shear stress and determined that the p50/p105 subunit was not affected by shear stress while both the p65 and p52/p100 subunits showed increased expression after treatment with shear stress.^38,55 Additionally, NF-kB signaling upregulates the production of NLRP3, IL-18, and IL-1β, so caspase-1 expression should not be upregulated in the presence of shear stress.^22,56,57 Shear stress has also been shown in the literature to initiate the second signal in NLRP3 inflammasome activation through calcium influx.^58,59 This explains why a decrease in caspase-1, a protein not regulated by NF-kB, is observed; shear stress converts the caspase-1 RNA into the caspase-1 protein. Binding to the pro-IL-1β gene is regulated by the p50/p65 heterodimer, so in the absence of the p50/p105 subunit of NF-kB, IL-1β RNA will be greatly reduced in comparison to LPS-treated cells.^60,61 However, with LPS stimulation, the p50/p65 heterodimer of NF-kB forms and initiates the upregulation of IL-1β. We treated iBMDMs with 10 or 50 μM Dooku1, a well-documented Piezo1 inhibitor, before exposing cells to 50 dyn/cm² shear stress for 30 min, followed by 1 h nigericin treatment.⁶² To determine the inhibition we see with Dooku1 treatment, we quantified the ASC–CFP speck formation (Figure 5A) and cell death with PI staining (Figure 5B). With both concentrations of Dooku1, we observed a threefold reduction in ASC–CFP speck formation compared to the group that was not treated with Dooku1. We also saw no significant differences between the 50 μM Dooku1 group and the untreated cells, indicating that Dooku1 completely inhibited shear-stress-induced NLRP3 inflammasome activation. Dooku1 treatments also greatly reduced the cell death we observed with PI staining, indicating that inhibiting Piezo1 and therefore reducing ASC–CFP speck formation increased cell viability to the levels statistically similar to untreated cells. Next, we performed qPCR to assess the relative gene expression of NLRP3 (Figure 5C) and Piezo1 (Figure 5D) when cells were pretreated with 50 μM Dooku1. We observed significant reductions in both NLRP3 and Piezo1, indicating that Dooku1 completely inhibited Piezo1 which resulted in no significant changes in NLRP3 expression. These experiments provided conclusive results about the mechanism through which shear stress primes the NLRP3 inflammasome for activation. Upregulation of Piezo1, IKKβ, p52/p100 NF-kB, p65 NF-kB, and NLRP3 RNA under shear stress conditions and the inhibition of ASC–CFP specks and NLRP3 and Piezo1 gene expression prove that shear stress interacts through the Piezo1 receptor to upregulate the IKKβ enzyme, which activates the NF-kB transcription factor for NLRP3 inflammasome priming (Figure 6).

Figure 4.

NLRP3 inflammasome shear-stress-priming acts through Piezo1. (A) Concept figure of qPCR is shown. The collagen in the gel channel of the microfluidic chip is digested using 1 mg/mL collagenase D in basal DMEM for 1 h at 37 °C. The cells are then removed from the chip, and RNA extraction is done with TRIzol. The total RNA is synthesized into cDNA before being amplified and quantified by using a qPCR machine. (B–I) Quantification of the relative expression levels of NLRP3 (B), IL-1β C), p50/p105 NF-kB (D), Piezo1 (E), caspase-1 (F), IKKβ (G), p52/p100 NF-kB (H), and p65 NF-kB (I) is normalized to the changes in expression of β-actin mRNA for untreated iBMDMs, iBMDMs primed with 50 dyn/cm² for 30 min, and LPS-primed iBMDMs for 4 h. Data shown are ± SEM (n = 3). Statistical analysis was performed using ***one-way ANOVA, followed by the Tukey post-test. ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 5.

Dooku1 inhibits shear-stress-primed NLRP3 inflammasome activation. (A, B) Quantification of ASC–CFP specks (A) and PI-positive cells (B) after pretreatment of 10 or 50 μM Dooku1 treatment for 2 h, followed by 30 min shear stress exposure and 1 h nigericin treatment was normalized by the total number of living cells characterized by the NucRed signal. Data shown are ± SEM (n = 3). (C, D) Quantification of gene expression levels for NLRP3 (C) and Piezo1 (D) was normalized relative to the changes in the expression of β-actin mRNA for untreated iBMDMs. All statistical analyses were performed using one-way ANOVA, followed by the Tukey post-test. ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 6.

Schematic showing shear-stress-primed NLRP3 inflammasome activation pathway. Activation of the NLRP3 inflammasome requires two signals. Signal 1 is transduced when shear stress causes a conformational change in the Piezo1 mechanosensitive ion channel, which activates IκB kinase (IKK), the enzyme responsible for the activation of transcription factor NF-kB. The increased activity of IKK, and subsequently NF-kB, regulates the production of pro-IL-18, pro-IL-1β, and NLRP3. Signal 2 is transduced through nigericin, a well-studied toxin that induces a potassium efflux and an increase in the number of mitochondrial reactive oxygen species (mROS). In response to Signal 2, inactive NLRP3, procaspase 1, and inactive ASC oligomerize to form the NLRP3 inflammasome complex. Active caspase 1 cleaves pro-IL-18, pro-IL-1β, and Gasdermin D into their active forms. Inflammatory cytokines IL-18 and IL-1β are exported from the cell via Gasdermin D pores, which results in pyroptosis. Created with Biorender.com.

DISCUSSION

Shear stress is one of the many stimuli that macrophages are exposed to; yet, its effect on NLRP3 inflammasome activation is poorly understood. We aimed to gain a deeper understanding of the relationship shear stress can have with the progression of inflammatory diseases in the context of the NLRP3 inflammasome. In this study, we identified a mechanism through which physiologically relevant shear stresses prime the NLRP3 inflammasome for activation. We assessed the degree of inflammasome activation in a microfluidic device by priming macrophages with several physiologically relevant shear stresses by investigating how ASC–CFP speck formation is affected by the shear stress magnitude, shear stress exposure time, and cell seeding density. These assays led us to conclude that 50 dyn/cm²shear stress resulted in significant ASC–CFP speck formation for both 15- and 30 min exposure. Cell seeding density across the range tested only exhibited differences in the formation of ASC–CFP specks at 50 dyn/cm² shear stress for 30 min. We also supported these data by utilizing NLRP3 and Caspase-1 KO iBMDMs to demonstrate that the ASC–CFP speck formation due to shear stress exposure correlates only with NLRP3 inflammasome activation. Additionally, we provided qPCR results for numerous genes involved in NLRP3 inflammasome activation and pretreated cells with a small-molecule inhibitor to confirm that Piezo1 plays a critical role in shear-stress-priming of the NLRP3 inflammasome. From this, we provided a mechanistic pathway of the players involved in shear-stress-primed NLRP3 inflammasome activation in a microfluidic device system.

These results demonstrated that shear stress regulates the NLRP3 inflammasome through Piezo1 and IKKβ, which other studies indicated that shear stress upregulates. These results also proved that the ASC–CFP speck formation observed in iBMDMs exposed to shear stress is regulated by only the NLRP3 inflammasome. For example, the lack of ASC–CFP speck formation in NLRP3 and caspase-1 KO iBMDMs at the same shear stresses indicated that this phenomenon is only observed in NLRP3 inflammasomes, where NLRP3, caspase-1, and NF-kB play extensive roles in its activation. There were no significant changes in the formation of ASC–CFP specks for both NLRP3 and caspase-1 KO iBMDMs, regardless of the magnitude of shear stress they were exposed to, while we noted significant changes in ASC–CFP speck formation for WT iBMDMs at shear stress magnitudes of 25 and 50 dyn/cm². This observation suggests that both NLRP3 and caspase-1 are involved in the process of shear-stress-priming and subsequent inflammasome activation. To determine the direct mechanism of this activation, qPCR results confirmed that shear-stress-primed NLRP3 inflammasome activation is regulated upstream by Piezo1 and IKKβ, as has been previously reported in the literature.^17,19,20,26 The relative expression of NLRP3 when compared to β-actin was upregulated threefold in cells treated with shear stress when compared to the untreated group, indicating that the presence of NLRP3 is integral to this activation pathway via shear stress, which we also confirmed with NLRP3 KO via confocal microscopy. In addition to NLRP3, the relative expression of Piezo1 and IKKβ was increased 1.5-fold and 2-fold, respectively, in the presence of shear stress, indicating that shear stress primes NLRP3 inflammasome activation through Piezo1 and IKKβ. These results supported previous literature and microscopy data that indicated a positive correlation between the shear stress magnitude and ASC–CFP speck formation in iBMDMs. We also demonstrated that when the Piezo1 receptor was inhibited with Dooku1, there were no significant changes in ASC–CFP speck formation, cell death, relative NLRP3 expression, or relative Piezo1 expression, indicating that Piezo1 plays an integral role in transducing NLRP3 inflammasome activation through shear stress exposure. The increased expression of some players was not expected due to NF-kB not regulating their transcription. For example, the relative expression of caspase-1 was affected by LPS, but shear stress showed a decrease in the level of caspase-1 expression. The ability of shear stress to decrease caspase-1 RNA levels is consistent with the findings of other studies that shear stress can induce a calcium influx in the cell membrane, acting as Signal 2 for NLRP3 inflammasome activation.^58,59 This could also account for the decrease in IL-1β expression in shear-stress-treated cells, since IL-1β ELISA data demonstrated an upregulation of IL-1β in shear-stress-primed cells. Since we demonstrated the upregulation of other markers of NLRP3 inflmmasome activation, it is logical that IL-1β would also be upregulated, as we demonstrated. Additionally, one in vivo study indicated that when the p50/p105 subunit of NF-kB was knocked out, IL-1β expression was greatly increased compared to the control group.⁶¹ Since shear stress does not upregulate the expression of the p50/p105 subunit of NF-kB, it is logical that IL-1β would be upregulated under shear stress conditions.⁶⁰

Data from qPCR also indicated that shear stress could selectively upregulate the transcription of different subunits of the transcription factor NF-kB in comparison to those upregulated from the chemical stimulation of NLRP3 inflammasome priming. LPS-treated cells upregulated p50/p105 and p65 NF-kB subunits since a p50/p65 heterodimer is formed when NF-kB is translocated to the nucleus due to LPS treatment. Additionally, LPS-treated groups indicated an increase in the p52/p100 NF-kB subunit, which can also be produced but happens at a later stage compared to the p50/p65 heterodimer formation and translocation.^63,64 Under shear stress conditions, the expression of the p50/p105 subunit was not significantly different, while the expression of both p52/p100 and p65 subunits was increased in shear-stress-treated cells by 3.5-fold and 4-fold, respectively. Previous literature supports this finding with the p52/p100 and p65 subunits of NF-kB being increased in the presence of shear stress, while the p50/p105 subunit was unaffected by shear stress.^38,55 The upregulation of the p65 subunit is also supported by the upregulation of IKKβ expression, which cleaves p65 into its active form.⁶⁵ This indicates that shear-stress-priming and LPS priming affect NF-kB activation differently, as the subunits interacting with DNA to initiate transcription differ. Under shear stress conditions, the p50/p105 subunit is not upregulated, while the expression of the p52/p100 and p65 subunits is upregulated. Based on the qPCR data indicating the lack of involvement of the p50/p105 subunit, the dimer that binds to DNA to initiate transcription in shear-stress-induced NLRP3 inflammasome priming is likely the p52/p65 heterodimer since p52 homodimers have low affinities for DNA binding and require coactivators to initiate transcription.⁶⁶ Homodimers of the p65 subunit are also unlikely to form since the affinity for homodimerization is low unless the IκBβ protein is present, which increases the binding affinity of this homodimer.^66,67 Since IKKβ is upregulated in the presence of shear stress, which phosphorylates IκBβ resulting in its proteasomal degradation, the affinity for p65 homodimerization should be unfavorable under shear stress conditions.^67,68 Therefore, the noncanonical p52/p65 heterodimer of NF-kB subunits is translocated to the nucleus to initiate the transcription of inactive proteins for inflammasome activation under shear stress conditions. To further emphasize the difference between shear-stress-priming and LPS priming of the NLRP3 inflammasome, several studies have shown that the p50/p105 subunit can homodimerize to aid in the recovery of inflammatory diseases, and its absence has been associated with disease progression.^69,70 This provides evidence that shear stress, which does not induce p50/p105 transcription, can result in a more sustained inflammatory disease state due to the body’s inability to attenuate the inflammatory response with a p50 homodimer. Additionally, higher shear stresses could hinder the body’s ability to repress an inflammatory environment in other organs of the body, increasing the probability of the occurrence of additional inflammatory diseases. These experiments provided a better understanding of how shear stress primes the NLRP3 inflammasome for activation in iBMDMs and insights into the mechanism through which this occurs. The upregulation of Piezo1, NLRP3, p52 NF-kB, p65 NF-kB, and IKKβ proved their involvement in shear-stress-induced NLRP3 inflammasome activation, in addition to suggesting that the transcription factor NF-kB behaves differently under shear stress exposure than traditional chemical stimulation.

CONCLUSIONS

In summary, we have provided a mechanistic study on how exposure to shear stress affects the activation of the NLRP3 inflammasome. We demonstrated through KO iBMDMs that NLRP3 and caspase-1 are involved in forming ASC–CFP specks and IL-1β secretion after shear-stress-priming and that shear-stress-priming only regulates the NLRP3 inflammasome. We also determined shear-stress-induced NLRP3 inflammasome activation through the Piezo1 and IKKβ pathways with qPCR and utilizing a small-molecule inhibitor of Piezo1. We also determined that the production of different NF-kB subunits can be affected by either LPS-priming or shear-stress-priming that is performed. These results indicate that shear stresses can modulate the progression of inflammatory diseases without the presence of chemical activators. This work can provide a foundation for more physiologically accurate organ- and disease-on-a-chip models that utilize physiological shear stress and cell lines specific to the disease model investigated. Designing disease-specific models to incorporate these physiologically relevant shear stresses can provide an avenue for significant advancement in the field of pharmaceutical therapeutic development by increasing the quality and specificity of treatments given to patients who need them.

MATERIALS AND METHODS

Materials.

All of the reagents used in these experiments were purchased from commercial suppliers. Ultrapure lipopolysaccharide (LPS) and nigericin were purchased from Invivogen and Tocris, respectively. Cell culture materials, such as Dulbecco’s modified Eagle’s medium (DMEM), penicillin/streptomycin, trypsin-EDTA, and heat-inactivated fetal bovine serum (FBS), were purchased from Gibco, Life Technologies. The microfluidic devices and luer-lock connectors used in shear stress experiments were purchased from AIM BIOTECH. Sterile syringes were acquired from Fisher Scientific. Materials for the shear stress experiments, such as tubing and luer-lock to barbed end adaptors, were purchased from Cole Parmer. All other staining compounds, such as NucRed Live 647 ReadyProbes Reagent and propidium iodide, were acquired from ThermoFisher Scientific. Selective Piezo1 inhibitor Dooku1 was acquired from MedChemExpress.

Methods.

Cell Culture.

All in vitro experiments were conducted using immortalized bone marrow-derived macrophages (iBMDMs) engineered to express a CFP residue tagged on the ASC. These cells and the NLRP3 KO and caspase-1 KO cells were a gift from Dr. Kate Fitzgerald from the University of Massachusetts Chan Medical School. All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% v/v fetal bovine serum (FBS) and 1% v/v antibiotic mixture containing penicillin (50 μg mL⁻¹) and streptomycin (50 μg mL⁻¹). Cells were split every 3 days or at 90% confluency by detachment with 0.25% trypsin-EDTA in 1X PBS and seeded in a T-25 cell culture flask at a 1/6 ratio.

Device Design and Fabrication.

Unless otherwise specified, the microfluidic device used for in vitro experiments was the ‘3D culture chip’ from AIM BIOTECH. The device fabrication was conducted at AIM BIOTECH using a cyclic olefin polymer (COP). The device design consists of a single-layer slide format (75 × 25 mm) device with three ‘chips’ containing a central gel channel (1.3 mm) surrounded by two media channels (0.5 mm) on opposing sides. The height of all channels in each chip is 0.25 mm.

Collagen Gel Synthesis.

Type I rat tail collagen (Corning) at a concentration of 2.8 mg/mL was combined with sterile-filtered 10× PBS, sterile-filtered 1.0 M NaOH, and sterile-filtered Milli-Q water. The collagen gel (2.5 mg/mL) was pH-tested using pH strips to ensure it is between 7.0 and 7.5.

Seeding Cells in Microfluidic Device.

Luer-lock adaptors were inserted into the media ports of the microfluidic device before seeding cells. Cells were suspended in the 2.5 mg/mL collagen gel at a seeding density of 2 million cells/mL. For this, 10 μL of the suspension was injected into the central channel of the gel. After seeding cells, the gel inlets were covered with covers provided with the luer-lock adaptors from AIM BIOTECH. The chip was then incubated at 37 °C for 30–35 min until the gel hardened. Media was injected into the outer channels after the incubation period.

Shear Stress Exposure.

The luer-lock adaptors were filled with 105 μL of media, and the luer-lock to barbed end adaptors were attached to the microfluidic device. Sterile luer-lock syringes were filled with media, and biocompatible tubing was attached to the syringes. A Harvard Apparatus PhD Ultra Syringe Pump was used to apply flow in the shear stress experiments. The syringes were filled with supplemented DMEM and loaded onto the syringe pump. The tubing was primed with media before the tubing was attached to the adaptors on the microfluidic device. A collection vessel was used to collect the effluent media from the microfluidic device. Shear stress experiments were conducted at 37 °C and 5% CO₂.

Light Microscopy for Static ASC Speck Imaging.

As previously dictated, iBMDMs were seeded in the central channel of a microfluidic device at 2 million cells/mL. The cells were treated with 100 ng mL⁻¹ LPS for 4 h, followed by 10 μM nigericin for 1 h. After nigericin treatment, the cells were stained with NucRed (2 drops per mL) and 2 μg mL⁻¹ propidium iodide incubated at 37 °C for 30 min. LPS and nigericin treatment controls were performed for their respective time points and stained as previously described. Cells were imaged at 20× on an A1R-TIRF confocal microscope.

Light Microscopy for Dynamic ASC Speck Imaging.

To investigate shear stress as Signal 1, iBMDMs were seeded in a microfluidic device at 2 million cells/mL, as previously described. The cells were pretreated with shear stress for 15 or 30 min before being treated with 10 μM nigericin for 1 h. After nigericin treatment, the cells were stained with NucRed (2 drops per mL) and 2 μg mL⁻¹ propidium iodide incubated at 37 °C for 30 min. Nigericin and shear stress controls were used for the time points as previously described. Cells were imaged at 20× on an A1R-TIRF confocal microscope.

IL-1β ELISA Treatment.

Media samples from microfluidic devices were extracted and treated with 1 % v/v of HALT Protease Inhibitor Cocktail (ThermoFisher) according to the manufacturer’s protocol. IL-1β ELISA (Invitrogen) was used to quantify protein concentration in supernatants from microfluidic experiments according to manufacturer’s protocols.

RNA Extraction.

The collagen in the gel channel of microfluidic devices was digested using 1 mg/mL collagenase D (Roche) in basal DMEM for 1 h at 37 °C and 5% CO₂. RNA was extracted from the cells using TRIzol Reagent (Invitrogen) based on manufacturer’s protocol, and the purity was assessed using a Nanodrop. Total RNA was reverse-transcribed using the High-Capacity cDNA Reverse Transcriptase Kit (Applied Biosystems) according to the manufacturer’s protocol.

qPCR Analysis.

qPCR was performed with TaqMan Fast Advanced Master Mix and TaqMan Gene Expression Assay (FAM) following the manufacturer’s protocols. Relative transcription was normalized for that of β-actin mRNA. Information about the RT-PCR primers is provided in Table S1.

Dooku1 Inhibitor Study.

For the microscopy of the Dooku1 inhibitor study, iBMDMs were pretreated with 10 or 50 μM Dooku1 for 2 h before undergoing shear stress exposure at 50 dyn/cm² for 30 min, 1 h 10 μM nigericin treatment, and staining, as previously described. For the qPCR experiments, iBMDMs were pretreated with 50 μM Dooku1 for 2 h and treated with shear stress 50 dyn/cm² for 30 min before performing RNA extraction and running qPCR as previously described.

Statistics.

GraphPad Prism 8 was used to plot all the graphs and analyze their respective statistics. Comparative analysis between two groups was executed using unpaired t test (two-tailed). For multiple-group comparative analysis, ordinary one-way and two-way ANOVA, followed by Tukey’s post-test were used. All the data are displayed as mean ± SEM (standard error of the mean) or mean ± SD (standard deviation). p-value <0.05 was considered as significant.

Supplementary Material

Supporting Information

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsami.3c18645.

Static NLPR3 inflammasome activation on a microfluidic device; cell death studies for WT iBMDMs in dynamic inflammasome activation; cell seeding density does not affect ASC–CFP speck formation in shearstress-primed iBMDMs; cell death studies for KO iBMDMs in dynamic inflammasome activation; and list of qPCR primers (PDF).

Data Availability Statement

The authors declare that all data supporting the findings of this study are available within the paper and its Supporting Information. Additional data related to this paper may be requested from the authors.