Background: Toll-like receptors (TLRs) are key players in innate immunity and are causally related to arterial occlusive disease and arterial remodelling. The release of proinflammatory cytokines following TLR ligand binding is increased in patients with unstable angina.
Objective: To examine the effect of a percutaneous coronary intervention (PCI) on TLR2 and TLR4 response and expression.
Methods: In 70 PCI patients, blood samples were gathered after sheath insertion and 2 hours after the catheterisation. TLR2 and TLR4 expression on, and tumour necrosis factor α (TNFα) levels in, monocytes were measured with flow cytometry. Whole blood was stimulated overnight with the TLR2 ligand Pam3Cys and the TLR4 ligand lipopolysaccharide. TNFα was determined in the stimulated samples and considered to be a measure of the TLR response. Baseline TLR expression and response were studied in relation to angiographic luminal stenosis and fractional flow reserve (FFR) measurement.
Results: A significant relation was found between TLR response and the angiographic percentage diameter stenosis, number of diseased vessels and FFR outcome. Furthermore, 2 hours after PCI a significant decrease in TLR2 and TLR4 response (p<0.001) and TLR2 and TLR4 expression (p = 0.001 and p = 0.068, respectively) was seen.
Conclusion: TLR response is positively associated with percentage diameter stenosis, multivessel disease and FFR outcome. Systemic TLR2 and TLR4 response and expression decrease after PCI. These results suggests that the TLR signalling pathway encompasses a potential biomarker for myocardial ischaemia in stable coronary artery disease.
Statistics from Altmetric.com
Atherosclerosis is an important contributor to worldwide cardiovascular morbidity and mortality. Extensive evidence has shown that atherosclerotic plaque development is an inflammatory process, yet the exact aetiological mechanisms remain to be elucidated.1 Experimental research showed that toll-like receptors (TLRs) play a significant role not only in innate immune reactivity but also in arterial occlusive disease.2 3 Furthermore, TLRs contribute significantly to the development of atherosclerosis, arterial remodelling and neointima formation.4–6 The last of these is of particular importance with the emergence of percutaneous coronary intervention (PCI) and intracoronary stent deployment. The angioplasty procedure is inherently accompanied by (limited) vascular trauma which induces a local inflammatory response with subsequent cellular proliferation and migration. Although a TLR polymorphism has been correlated with late-term outcome following PCI,7 the immediate effect of the vascular trauma on the activity of TLRs has not been described.
Toll-like receptors (TLRs) are transmembrane receptors that recognise pathogen-associated molecular patterns which are highly conserved during evolution. Ligand binding of the extracellular domain triggers the production of proinflammatory cytokines like tumour necrosis factor α (TNFα) and interleukin 6.8 Different TLRs recognise different surfaces, intracellular components of micro-organisms and endogenous ligands released during (vascular) trauma.9 Cardiovascular research focusing on innate immunity often takes the expression and response of only TLR4 into account.10 11 Although TLR4 is an important contributor to inflammatory processes surrounding atherogenesis, recent observations also point to TLR2 activation as a potential trigger for intimal hyperplasia and acceleration of atherosclerotic plaque formation.12 13
Previously it has been shown that cytokine release after lipopolysaccharide (LPS) stimulation and TLR4 expression is significantly increased in patients with recurrent unstable angina compared with patients with stable angina and healthy controls.14 The determining factor for this increased cytokine release in patients with unstable angina remains unknown. A second stimulation of cells by TLR ligands results in an attenuated release of proinflammatory cytokines in comparison with the first exposure, a situation referred to as tolerance.15 Moreover, cross-tolerance of TLR2 for TLR4 ligands (and vice versa) has been reported, indicating a waning of the immune response upon ligand exposure. A state mimicking TLR ligand tolerance is observed upon the infliction of surgical trauma.16
Insight into the regulation of the TLR response in relation to vascular injury is relevant considering the important role of the innate immune system and TLRs in arterial occlusive disease. In this study we examined the release of the proinflammatory cytokine TNFα after ligand binding of both TLR2 and TLR4 before and after PCI. In addition, in a subgroup we studied the TLR responsive state in relation to baseline measures of myocardial ischaemia. A state of tolerance of TLRs following cerebral ischaemia has been reported in animal studies,17 but studies on the relation between myocardial ischaemia and TLR response are lacking. We report that TLR2 and TLR4 response differs significantly in relation to the baseline degree of angiographic stenosis and fractional flow reserve (FFR). In addition, arterial trauma significantly attenuates the full blood TLR2 and TLR4 response.
PATIENTS AND METHODS
Patient inclusion and clinical data
The study was approved by the local medical ethical board. All participating patients signed an informed consent before inclusion. Patients scheduled for a PCI and more than 18 years old were eligible for the study. Exclusion criteria were chronic total occlusion, malignant neoplastic disease, participation in another non-observational study, admittance for ST elevation myocardial infarction within 30 days of the procedure, admittance for non-ST elevation myocardial infarction within 30 days of the procedure and intravenous administration of corticosteroids before, during or after the interventional procedure. Cardiovascular risk factors, sex, previous medical history and medication use were gathered from questionnaires and patient medical records. All PCI procedures and FFR measurements were performed routinely according to hospital protocol. FFR measurement was performed by introducing a pressure wire into the target vessel (PressureWire5, 0.014″, 175 cm, Radi Medical Systems, Sweden). The pressure wire was equalised to aortic pressure and advanced over the coronary lesion. Under continuous intravenous adenosine infusion (8.4 mg/kg/h) FFR values were obtained and recorded.
A blood sample of 10 ml was drawn immediately after sheath insertion. Two hours after removal of all catheters a second blood sample of 10 ml was obtained.
Blood samples were stored in both lithium-heparin (LH) anticoagulated and ethylenediaminetetraacetic acid anticoagulated tubes. To prevent premature leucocyte activation all tubes were kept on ice until further processing. Before initiation of this study, this protocol had been validated and control samples did not show altered TLR expression patterns and response over time (data not shown).
Quantitative coronary analysis
The severity of the coronary lesions was measured by two trained independent technicians in a blinded fashion, using quantitative coronary analysis software (QCAPlus, SDS, Palo Alto, CA, USA).
White blood cell count
To correct for the number of leucocytes present in the stimulation assay, in the LH samples a total white blood cell count was performed using a haematology analyser (Celldyn 1700, Abbott Diagnostics, IL, USA).
Measurement of TLR2 and TLR4 expression
TLR2 and TLR4 expression on circulating CD14-positive monocytes was examined by flow cytometry (Cytomics FC500, Beckman Coulter, Fullerton, CA, USA) of EDTA anticoagulated blood. The samples obtained before and after the catheterisation were processed simultaneously. Whole blood was stained for TLR2 (fluorescein isothiocyanate (FITC)), TLR4 (phycoerythrin (PE)) and CD14 (PE-Cy5) (all Serotec, Oxford, UK). In separate samples intracellular TNFα levels (FITC, Serotec) were assessed in CD14-positive monocytes (PE; Dako, CA, USA). TLR2, TLR4 and intracellular TNFα expression levels are referred to as mean fluorescence intensity.
Stimulation of whole blood samples
LH blood (100 μl) of the samples drawn before and after the catheterisation was stimulated with 100 μl of the synthetic TLR2 ligand Pam3Cys (Novabiochem, Cambridge, MA, USA) at 5, 50 and 500 ng/ml and 100 μl of the TLR4 ligand LPS (Sigma, St Louis, MO, USA) at 1, 10 and 100 ng/ml. Samples were incubated overnight at 37°C and 5% CO2, centrifuged at 1000 rpm for 5 minutes and the resulting supernatant was transferred to a separate 96-well plate and stored at −20°C until further analysis.
As a surrogate of TLR2 and TLR4 response, TNFα was measured in the supernatants of the blood samples using an enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's protocol (PeliKine-compact, Sanquin, Amsterdam, The Netherlands). All TNFα concentrations were corrected for the white blood cells present in the sample.
The Wilcoxon signed ranks test was used to compare differences between TLR response and TLR expression before and after the catheterisation. Differences between two independent variables were calculated with the Mann–Whitney U test. A one-way analysis of variance was performed to compare differences between categorical data consisting of more than two variables. A p value <0.05 was considered significant. All statistical analysis was performed with SPSS, version 12 (SPSS Inc, Chicago, IL, USA).
A total of 90 patients with stable coronary artery disease were included in this study. In five patients a blood sample after catheterisation could not be obtained for logistical reasons. Two patients received intravenous corticosteroids during the procedure owing to contrast allergy, resulting in a complete absence of TLR response. Thirteen patients did not undergo a balloon dilatation, either because of a non-significant FFR measurement (n = 6), the impossibility of passing the coronary lesion (n = 5) or the impossibility of visualising a coronary lesion despite an indication on the diagnostic angiogram (n = 2). In the remaining 70 cases a balloon dilatation and stent placement was performed (either with or without pre-dilatation). Table 1 summarises the clinical and angiographic characteristics of these 70 patients.
TLR response in relation to percentage stenosis and number of diseased vessels
A total of 65 angiograms were available for offline quantitative coronary angiography analysis performed by two individual observers, reaching a mean interobserver variation of 3.65%. Both minimal luminal diameter and percentage diameter stenosis were measured with quantitative coronary angiography. However, to account for variations in vessel size the percentage diameter stenosis was used in all comparisons. The highest percentage diameter stenosis measured in a patient was assigned to one of the categories <70%, 71–90%, 90–99% stenosis and correlated with the TLR response measured in the sample drawn immediately after sheath insertion.
We observed a significant relation between the degree of stenosis and the TLR response after stimulation with all concentrations of Pam3Cys (p = 0.003, p = 0.004 and p = 0.016) and 1 ng/ml LPS (p = 0.002) (figs 1A and B). Furthermore, the TLR response after Pam3Cys stimulation in patients with two- or three-vessel disease was significantly higher in comparison with that in patients with one-vessel disease (fig 2A, p = 0.007, p = 0.003 and p = 0.017). A similar relation was also present following LPS stimulation, but did not reach statistical significance (fig 2B). The TNFα production after LPS and Pam3Cys stimulation were all dose dependent.
TLR response in relation to FFR
In 13 patients an FFR was measured. Seven patients had a significant lesion (FFR <0.75) and six lesions were not significant (FFR >0.80). Although the six patients with a non-significant FFR were not part of the main study group as described above, stimulation experiments with Pam3Cys and LPS were performed. In conjunction with the finding that percentage diameter stenosis is related to TLR response, we observed a striking difference in TLR response at baseline in relation to FFR outcome. The TLR response after Pam3Cys stimulation in patients with an FFR <0.75 was significantly higher than in those with an FFR >0.80 (p = 0.008, p = 0.014 and p = 0.014). A similar relation was also present after LPS stimulation, but did not reach statistical significance (figs 3A and B).
Comparison of TLR response and expression before and after PCI
Both, the Pam3Cys- (fig 4A) and LPS- (fig 4B) induced TNFα release were significantly lower in the samples gathered 2 hours after PCI than in the samples gathered before PCI (p = 0.002 for 100 ng/ml LPS and p<0.001 for all other concentrations). In conjunction with the decrease in TLR response, the TLR2 and TLR4 expression on monocytes as measured with FACS significantly decreased after PCI (p = 0.001 and p = 0.068, respectively). The intracellular TNFα level in monocytes did not change after PCI.
In this study we describe a relation between the TLR2 and TLR4 response and the presenting percentage diameter stenosis before balloon dilatation and the number of diseased vessels. The relation between TLR response and severity of coronary stenosis is further supported by the finding, that the production of TNFα after TLR ligand incubation is significantly higher in patients with an FFR measurement <0.75 than in patients with an FFR measurement of >0.80.
Luminal stenosis and TLR expression
Previous studies by Liuzzo et al14 and Methe et al18 showed an increased response to LPS stimulation and an increased expression of TLR4, in patients with unstable angina pectoris. As a consequence of plaque rupture, unstable angina pectoris is thought to arise owing to intermittent and increasing obstruction of the coronary vessel by thrombus formation. This study provides a hypothesis for the observation of increased TLR response in patients with unstable angina. We examined patients who mostly had stable coronary artery disease. Our study might provide a possible explanation for the previously described increase in TLR response in patients with unstable angina pectoris as we demonstrated a relationship between the highest degree of stenosis and the response of the whole blood samples. A low degree of chronic stenosis results in a lower responsive state, whereas a high degree of stenosis leads to a higher TLR response. Therefore, not just in unstable angina, but also in stable coronary syndromes, TLR2 and TLR4 responses may vary substantially in relation to the degree of coronary ischaemia. FFR has been widely described as an accurate measure of the severity of coronary lesions and the necessity to perform an intervention on them. In view of our observation that full blood TNFα release after TLR2 and TLR4 stimulation is more than twofold higher in patients with an FFR <0.75 than in patients with an FFR >0.80 (fig 3), we speculate that TLR responsiveness could serve as a surrogate marker for myocardial ischaemia.
The described coronary obstructions eventually lead to collateral vessel formation (arteriogenesis), which is an important mechanism by which the myocardium increases blood flow to ischaemic regions.19 20 Important contributors to this process are inflammatory cells and cytokines.21–23 We hypothesise that the relation between severity of coronary lesions and the size of the TLR response reflects a stimulatory trigger for arteriogenesis. This hypothesis is further supported by the finding that FFR outcome is related to TLR response before catheterisation. Future studies looking specifically into the relation between the innate immune response and myocardial ischaemia should provide more insight into the mechanism of this process. This is relevant for clinical practice since the strong temporal changes in the TLR response may be a double-edged sword. On the one hand, an increase in TLR2 and TLR4 response may be a trigger for arteriogenesis, and on the other, it may elicit acute coronary events by plaque destabilisation. Therefore, future experimental studies are needed to increase our understanding of the role of the TLR system in coronary artery disease.
We show that systemic TLR2 and TLR4 response decreases significantly in patients undergoing PCI. In conjunction with this, TLR2 and TLR4 expression on individual monocytes decreases within 2 hours after the interventional procedure.
It is known that mononuclear cells stimulated with LPS are less responsive to a subsequent stimulation, a situation known as LPS or endotoxin tolerance.15 Major trauma mimic this endotoxin-tolerant state as was shown in previous research.24 In another study eight patients who underwent laparoscopic surgery showed a decreased postoperative response to LPS despite the lack of previous exposure to LPS.16 Our study demonstrates a comparable effect after minimal vascular trauma. Although excluded from the main population, the samples of the patients who did not undergo balloon dilatation did undergo TLR stimulation and FACS analysis of TLR2 and TLR4. These patients showed a less pronounced downtoning of the TLR response or expression after 2 hours (data not shown). This supports the idea that not only coronary catheterisation but also local arterial dilation is a primary trigger for the attenuation of TLR response and expression.
The mechanisms for these findings are not fully understood. One might suggest a release of TLR ligand from the plaque during the PCI, which might well elicit a tolerance-like state as described above. A possible candidate for this TLR ligand is oxidised LDL (oxLDL). It is abundantly present in atherosclerotic plaque, and oxLDL has been described as a TLR4 ligand.25 Paradoxically, the release of oxLDL might stimulate TLR response in vivo upon first ligation, but most likely elicit an inhibitory effect on subsequent (ex vivo) TLR stimulation as described by Nomura et al.15
Attenuation of the proinflammatory response may serve as a protection against an excess of cytokine release after an ischaemic episode or injury that can be provoked during balloon angioplasty. An abundance of proinflammatory cytokines can be harmful to the host environment. Thus, a decreased cytokine release after trauma could be functional in that it prevents additional damage inflicted by the immunological response. Our observations might be comparable to the role of TLR inhibition and cytokine signalling after cerebral ischaemia as described by Karikó et al.17 An ischaemic episode leads to the production of proinflammatory cytokines, which increase vascular permeability leading to secondary ischaemia and accumulation of immune cells in the affected region. To prevent an excess of proinflammatory cytokines being released in a subsequent ischaemic period, a negative regulation of the TLR response occurs. This so-called preconditioning might also take place after acute balloon-induced myocardial ischaemia. Local inflammation is a trigger in excessive vascular repair responses following arterial injury that may eventually result in restenosis. The strong fluctuations in innate immune responses early after arterial injury may influence the downstream release of cytokines, chemokines and growth factors, key players in the local response to injury. Whether the strong attenuation of the TLR response after PCI is functional—that is, that the inflammatory responses and hence intimal hyperplasia is reduced, remains to be investigated.
It is surprising that a small-vessel injury elicits this evident systemic decrease in TLR response, as expressed in TNFα release. This indicates that the innate immune system is highly responsive to minimal vascular trauma or short ischaemic events, and that the innate immune response is tightly regulated. Next to local ischaemia, local release of endogenous ligands might also explain the systemic response.
Alternatively, it has been suggested that the autonomic nervous system has a role in immune modulation.26–29 The systemic decreased TLR response and expression might be mediated through the so called nicotinic anti-inflammatory pathway.30 Vagal nerve activation can lead to a decrease in proinflammatory cytokine excretion by inflammatory cells and macrophages in particular. Furthermore, the vagal nerve serves as a bidirectional connection between the brain, and the immune system. Sensory fibres can be activated by the immune system, relaying a negative signal to itself through the brain. Thus a reflex similar to a knee-jerk reflex arises, the so-called inflammatory reflex. The changes in TLR response seen in this study could theoretically be accounted for by this inflammatory reflex.
Interestingly, only the excretion of TNFα is described as affected by this neural effect. The transcription of proinflammatory genes has been shown to continue intracellularly, thus resulting in an intracellular accumulation of cytokines. Indeed, in accordance with this inflammatory reflex model, our flow cytometry measurements show a modest increase in intracellular TNFα.
We conclude that baseline TLR2 and TLR4 response is associated with percentage diameter stenosis and FFR outcome. In addition, TLR response is attenuated immediately after PCI. This decreased response can be partially explained by decreased TLR2 and TLR4 expression on monocytes after PCI The observation that innate immunity response is strongly influenced by luminal stenosis as well as vascular injury is relevant for an understanding of the pathogenesis of atherosclerotic lesion initiation and progression. Future studies will show whether TLR response could serve as a surrogate inflammatory marker for stable ischaemic coronary artery disease.
The Bokalis foundation had no involvement in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.
Funding: This publication was supported by the Bokalis foundation.
Competing interests: None.
Ethics approval: Approved by the local medical ethical board.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.