Activation of platelet-adherent basophils in chronic rhinosinusitis with alcohol hypersensitivity

Abstract

Background:

Alcohol hypersensitivity (AH), an exacerbation of respiratory symptoms in response to alcohol consumption, is common in Aspirin Exacerbated Respiratory Disease (AERD) and other forms of chronic rhinosinusitis (CRS). We speculated that these reactions relate to the activation of innate immune cells including basophils and, in particular, by platelet-adherent basophils by polyphenolic compounds contained within eliciting alcoholic beverages.

Objectives:

We investigated the absolute numbers of these cells in patients with AH and the ability of relevant polyphenolic compounds to cause cellular activation.

Methods:

Data were collected from 412 consecutive adults presenting to a tertiary care sinonasal clinic in whom the presence of AH was elicited. CRS phenotype was determined and results from complete blood count and differential were analyzed. A subset of patients were invited to donate blood samples that were used to explore the ability of relevant compounds associated with alcohol consumption to activate platelet non-adherent and platelet-adherent basophils. Activation was quantified by flow cytometry as upregulated expression of CD63 and as secretion of lipid metabolites.

Results:

Of the 412 patients enrolled, 69 (16.7%) endorsed having AH. Significantly higher platelet counts were seen in patients reporting AH. Red wine extract and several polyphenolic compounds produced basophil activation and this was primarily observed amongst platelet-adherent basophils. Platelet activation was further established as the release of TxB2.

Conclusion.

The presence of AH is associated with significantly higher platelet levels and compounds present in alcoholic beverages can directly mediate both their activation as well as the activation of platelet-adherent basophils.

Introduction

Alcohol hypersensitivity (AH) comprises a syndrome characterized primarily by the development of upper respiratory vasomotor symptoms and flushing that occurs in response to alcohol consumption. AH is common in patients with asthma and chronic rhinosinusitis (CRS) with nasal polyps (NP) and has been reported to occur in anywhere from 12.9% to 43% of asthmatics and in a similar range of those having CRSwNPs.^1–4 These reactions are particularly common in aspirin-exacerbated respiratory disease (AERD), with a prevalence of up to 83%⁵ and Samter included alcohol sensitivity in his classic description of this syndrome.⁶ These reactions are not thought to be related to a sensitivity to inhibition of cycloxygenase-1 (cox-1), given its regular occurrence in aspirin tolerant patients and a more rapid onset of symptoms (typically <30 minutes) in contrast to the more delayed onset in AERD (typically 1-2 hrs).⁷ Furthermore, while adversely impacting quality of life (or at least the ability to imbibe alcoholic beverages), these reactions are typically less severe than those triggered by cox-1 inhibitors. Nor are these reactions related to the milder flushing reactions observed, especially in Asian populations, that are due to a polymorphism in acetaldehyde dehydrogenase (ADH).⁸ Another intriguing observation is that these respiratory reactions (in contrast to the more commonly occurring flushing reactions) do not appear related to alcohol itself. Specifically, these reactions are especially common in response to red wine and other beverages aged in wood barrels and are less common with (unconditioned) beer, white wine, or liquor.^5,9

We speculated that AH could reflect the ability of these beverages to trigger the activation of immune cells including mast cells, basophils, and eosinophils. Supportive of this concept is the observation that higher levels of eosinophil cation protein (ECP) are observed in the NPs of alcohol intolerant subjects.¹⁰ Our previous studies suggested the ability of alcoholic beverages to trigger activation of blood eosinophils and, especially, of basophils.¹¹ The production of cysteinyl leukotrienes (CysLTs), especially in AERD, is largely driven by the presence of platelet-adherent leukocytes.^12,13 Such activation involves, in part, the translocation of leukocyte-derived leukotriene (LT) A₄ to these adherent platelets, which then utilize their endogenous LTC₄ synthase to complete the production of CysLTs. It is also intriguing that alcoholic beverages, and in particular red wine, are capable of directly activating platelets ¹⁴. The current studies were therefore performed to further explore the ability of alcohol but, more importantly, components of alcoholic beverages, to drive the activation of basophils. And, in particular, we explored the role of platelets and platelet-adherent basophils in producing this syndrome.

Methods

Subjects.

Research subjects consisted of 478 patients consecutively evaluated in a tertiary care sinonasal clinic over a three year period. Data were collected under written protocols approved by the investigational review board of the University of Virginia (IRB# 20731 and #14457). The presence of AH was based upon history of having both recurrent exposures and consistent development of respiratory symptoms. Phenotypic characterization of CRS patients was determined based upon the absence or presence of NPs as determined on endoscopic exam. AERD was defined based upon either a compelling history of developing respiratory (upper and/or lower airway) symptoms after ingestion of a non-selective cyclooxygenase inhibitor or having a positive reaction to aspirin/NSAID challenge. No subject had been aspirin desensitized at the time of study and subjects were excluded who were on biologic therapies or who had received oral corticosteroids within 1 month of enrollment. Results from a complete blood count and differential performed within 1 year of initial presentation were obtained and absolute eosinophil, basophil, and platelet counts were recorded.

Alcohol hypersensitivity in relation to CRS phenotype and blood counts.

All subjects were inquired regarding their consumption of alcoholic beverages and if respiratory reactions subsequently occurred. We investigated the tendency of subjects to report AH (respiratory symptoms) as a function of CRS phenotype (subjects with non-respiratory complaints, e.g., flushing, were not considered to have AH for these analyses). Subsequently we examined the association of blood eosinophil, basophil, and platelet counts with the reported presence of AH and performed multivariate analyses to determine the ability of these cell counts and, in particular, platelets, to associate with AH independent of their impact on CRS phenotype.

Activation of platelet-adherent basophils by EtOH.

Several representative alcohol sensitive (n=12) along with age- and gender-matched healthy control subjects (n=10) were invited to participate in a subsequent study to ascertain the ability of alcohol and components of alcohol-containing beverages to activate platelet-adherent basophils. Healthy controls reported no sensitivity to alcohol-containing beverages and had no history of CRS or asthma. Heparinized peripheral blood samples (10 ml) were collected and red cells lysed using NH₄Cl. In previous preliminary studies,¹¹ we demonstrated the ability of red wine extract (RWE) and several biologically active polyphenolic compounds contained within red wine to activate basophils.^15,16 Basophil activation was determined by flow cytometry using a commercial basophil activation test (Bühlmann; Amherst, NH).^17,18 Blood samples were left in a resting state or activated with ethanol (0.1 and 1.0% v/v), red wine extract (13 and 65 μg/ml; Seppic, Paris, France), resveratrol (22 and 220 ng/ml), catechin (1 and 10 μg/ml; source), and epigallocatechin (1 and 10 μg/ml). Additional samples were stimulated with anti-IgE or formyl-methionyl-leucine-phenylalanine (fMLP) as positive controls per the manufacturer’s directions (Bühlmann). Cells were stimulated at 37°C for 25 minutes after which samples were centrifuged, supernatants collected, and cellular pellets analyzed by flow cytometry as described below. Ethanol concentrations reflected a common legal definition for inebriation (0.1%) as well as a much higher concentration (1%) to confirm the inability of alcohol to influence basophil activation. RWE concentrations were calculated based upon content present in one full glass of wine (200 ml) and the assumption that there would be 100% absorption, which would then be diluted across the intravascular volume of distribution (5 L). Similar assumptions were made for resveratrol, catechin, and epigallocatechin based upon their concentrations known to be present in one full glass of red wine.¹⁹ The absorption and bioavailability of polyphenolic compounds is incompletely understood and is likely variable depending on the respective class and conjugation of the specific polyphenol and this is further influenced by microflora through hydrolyzation or other modifications of the conjugated moieties. The catechins studied in this experiment are all known to be absorbed and reach a maximal plasma concentration at 1.5 hours.²⁰

Flow cytometry.

Resting and stimulated whole blood samples were labeled with a cocktail of antibodies including the basophil marker CCR3 (PE-conjugated anti-CCR3; Bühlmann). CCR3 is also expressed on eosinophils but basophils are distinguished by being side scatter^low (figure 1a). Platelets were identified via co-staining with PerCP-conjugated anti-CD61 and platelet-adherent basophils were recognized as CD61 positive expression within the basophil gate (figure 1b). Basophil activation was determined as the upregulation of fluorescein-conjugated anti-CD63 basophils (Bühlmann).¹⁸ CD63 is a lysosomal-associated membrane protein (LAMP) expressed exclusively within the granules of resting cells and which translocates to the surface when the granules fuse with the plasma membrane during granulation. Our analyses quantified activation of platelet-adherent as opposed to platelet non-adherent basophils (figure 1c). To visually confirm that this strategy specifically identified platelet-bound basophils, additional studies were performed with an Image Stream flow cytometer (Luminex ImageStream X Mark II).

Figure 1:

Representative flow cytometry. The cytometer was gated on basophils (side scatter^lowCCR3^high). 1a. Gating strategy for activated (CD63^+ve) basophils (CCR3^highside scatter^high cells comprise eosinophils). 1b. Basophil-adherent platelets were demonstrated as the co-expression of CD61 among the gated eosinophils. 1c. Activation of platelet-non-adherent and adherent basophils in response to cell activation.

Enzyme immunoassays (EIAs).

Supernatants collected after cell stimulation were analyzed for the presence of thromboxane B₂ (TxB₂) using a commercial EIA (Abcam, Cambridge, UK). The sensitivity of this EIAs is 10.54 pg/ml.

Statistical analysis.

Differences in AH sensitivity among the various cohorts was evaluated by Chi² analysis. Differences in basophil, eosinophil and platelet counts among the different CRS phenotypes were determined via ANOVA. Next, the impact of blood counts on having AH was determined via t-tests. To determine which factors in these data were independent predictors of AH, a multivariate model was constructed. A raw model was created using the predictors of platelet count, absolute eosinophil count, age, sex, race, and polyp status. Then model-building continued in a stepwise backward approach, removing the least significant predictors until only statistically significant predictors remained. Basophil activation data were analyzed between stimulated and resting samples and between platelet-adherent and non-adherent basophils by paired t-tests after correcting for multiple comparison errors.

Results

Alcohol hypersensitivity in relation to CRS phenotype.

Of 478 consecutive patients with CRS, 412 were able to report on whether alcohol consumption resulted in respiratory symptoms. The remainder had either not imbibed or could not remember if any symptoms ever occurred in the past when they had. Only these 412 were included in the current study (Table 1). Of these, 69 (16.7%) endorsed having AH in the form of respiratory reactions. A significant difference in eosinophil counts (p<0.0001) was seen among the different CRS phenotypes, along with trends for basophil counts (p=0.06), but no differences were observed for platelets (p=0.33). Significantly higher platelet and eosinophil counts were seen in patients reporting AH (277±64 vs. 249±74 (x 10³) platelets; p<0.02 and 353±286 vs. 252±239 eosinophils/μL, p=0.001)(Table 2). No differences were seen in basophil counts (p=0.58). On the multivariate analysis, platelet (but not eosinophil) count and phenotype of sinusitis remained independent predictors of having AH. The presence of nasal polyposis (AERD or CRSwNPs) was associated with a 4.02-fold increase in the odds of having AH and for each additional 1000/μL increase in platelet number, there was a 0.5% increase in the odds of having AH (Table 3).

Table 1.

Alcohol hypersensitivity and blood counts in relation to CRS phenotype

	All (n=412)	CRSsNPs (n=274)	CRSwNPs (n=116)	AERD (n=22)	p-value*
Alcohol hypersensitivity (n, %)	69 (17%)	23 (8%)	31 (27%)	15 (68%)	<0.0001
Platelet count (x 10−3) (mean±SD)	254±73	251±73	256±75	276±64	0.34
Eosinophil count (cells/μL) (mean±SD)	271±251	191±152	405±317	498±337	<0.0001
Basophil count (cells/μL) (mean±SD)	49±48	44±50	56±46	62±38	.06

^*AH significance determined via Chi² and blood counts analyzed by ANOVA.

AERD – aspirin-exacerbated respiratory disease, AH – alcohol hypersensitivity, CRS – chronic rhinosinusitis, NP – nasal polyps, NS – not significant.

Table 2.

Blood counts (x10⁻³) in alcohol tolerant and hypersensitive CRS patients

	Tolerant	Sensitive	p-value*
Platelet count (x10−3) (mean±SD)	249±74	277±64	<0.02
Absolute eosinophil count (cells/μL; mean±SD)	252±239	353±286	0.001
Absolute basophil count (cells/μL; mean±SD)	48.4±33.3	50.6±33.2	NS

^*Comparisons assessed via paired t-tests;

NS – not significant.

Table 3.

Platelet and CRS phenotype as predictors of alcohol hypersensitivity

Alcohol Hypersensitivity	Odds Ratio (± standard error)	z-score	p-value	[95% confidence interval]
Platelet	1.005 ± 0.002	2.18	0.029	[1.0005, 1.0089]
CRS phenotype*	4.023 ± 0.964	5.81	<0.001	[2.515, 6.436]

^*Presence of nasal polyposis (CRSwNPs/AERD) in comparison to not having NPs (CRSsNPs).

AERD – aspirin-exacerbated respiratory disease, CRS – chronic rhinosinusitis, NP – nasal polyps.

Activation of platelet-adherent basophils by polyphenolic compounds.

Control subjects demonstrated basophil activation after exposure to anti-IgE or fMLP and no differences between platelet-adherent and non-adherent cells was observed. No significant activation was observed with ethanol, red wine extract, or any of the polyphenolic compounds (figure 2a). In contrast, a dose-dependent stimulation of basophil degranulation was observed with red wine extract (p<0.05 for RWE 65 μg/ml) and this was significantly greater (~2-fold) for platelet-adherent in comparison to non-adherent basophils (figure 2b). Of note, basophils from subjects reporting AH displayed significantly more platelet-adherence (43.1±1.8%) than those not reporting AH (33.4±1.5%; p<0.001). As we have previously shown¹¹ and consistent with the ability of these patients to tolerate non-conditioned alcoholic drinks, no activation was observed with ethanol. When we addressed specific bioactive polyphenolic compounds known to be present in conditioned beverages, no activation was observed with resveratrol, however significant activation was observed with both catechin and epigallocatechin and, as with RWE, these were all significantly greater for the platelet-adherent basophils. Indeed, no significant induction of non-adherent basophils was observed with catechin or with the lower concentration of epigallocatechin (or with any other stimulant).

Figure 2.

Activation of platelet-non-adherent and adherent basophils. 2a. Control subjecyd. 2b. Alcohol hypersensitive subjects. EtOH – ethanol; ResV – resveratrol, Cat – catechin, EGC – epigallocatechin, RWE – red wine extract. *p<0.05 compared to platelet non-adherent; ^†p<0.05 compared to background.

Visualization of activated platelet-adherent basophils.

To unambiguously confirm that activation was largely limited to platelet-adherent basophils, we performed image stream flow cytometry. Representative activated (CD63⁺ degranulated) basophils are displayed in figure 3, clearly displaying the presence of adhered basophils.

Figure 3.

Image stream visualization of red wine extract-activated platelet-adherent basophils. From left to right: dark field appearance, staining for adherent platelets (CD61; green), basophils (CCR3, yellow), activated (degranulated) basophils (CD63; red), and their merged co-expression.

Arachidonate mediator expression.

To further explore the activation of platelet-adherent basophils, we quantified secretion of the platelet-derived mediator, thromboxane (TxB₂). No TxB₂ release was observed with anti-IgE reflecting the absence of IgE receptors on these cells and implying that basophil activation did not cause the activation of adherent platelets. In contrast, significant platelet activation was observed with RWE but only among those subjects reporting alcohol hypersensitivity (figure 4).

Figure 4:

Thromboxane production in response to red wine extract. Whole blood samples were stimulated and supernatants collected. Thromboxane B₂ secretion was quantified by EIA. *p<0.05 compared to controls.

Discussion

Sensitivity to alcoholic beverages is common in patients with asthma, CRS, and nasal polyposis and is particularly common in patients with AERD.^1–4 These reactions are not related to the flushing reactions that occur in individuals with reduced expression of ADH.⁸ In addition to occurring with normal ADH expression, these episodes are contrasted by being less likely to be related to the alcohol itself. This is suggested by the recognition that these patients are particularly sensitive to oak- or other wood-conditioned beverages, such as red wine and conditioned beers but typically tolerate white wine and unconditioned highly distilled beverages such as vodka.⁵ Nor are these reactions related to the hypersensitivity responses driven by inhibition of cyclooxygenase (cox)-1 that occur in AERD patients, insofar as these reactions are common in asthmatics and CRS patients (and many healthy individuals) without sensitivity to these agents. In further contrast to AERD, these alcohol hypersensitivity reactions are seldom serious and never life-threatening.

We speculated that AH reactions could reflect the pharmacologic activity of polyphenolic compounds contained within conditioned alcoholic beverages and, specifically, we reasoned that these compounds could directly activate innate immune cells. We focused on the direct activation of immune cells central to the pathogenesis of these conditions, specifically mast cells, basophils, and eosinophils. While AH was associated with eosinophilia, by multivariate analysis we found no influence of eosinophils independent of their association with presence of CRS, NPs, or AERD (Tables 1–2). We were intrigued, however, that the most compelling association of blood cell distribution with AH was with platelet number (Table 3).

Previous studies have explored the role of platelets in these conditions and, in particular, the role of platelet-adherent leukocytes. This work includes the demonstration that activated leukocyte-adherent platelets via transcellular translocation of lipid intermediates contribute to the production of CysLTs.¹² In addition, previous studies have also demonstrated that alcoholic beverages can directly induce platelet activation.¹⁴ We therefore explored the ability of polyphenolic compounds to induce innate cell activation, focusing for the present, on basophil activation. (And the current results should suggest the possibility of similar activation of platelet-adherent eosinophils, neutrophils, and monocytes). Basophils, in addition to contributing to the presence and severity of NP disease²¹ readily lend themselves to investigating cell-specific activation. Specifically, we took advantage of flow cytometric methodologies to both identify platelet-adherent basophils and to identify and quantify their activation as the induced expression of surface CD63. The current studies confirm that it is not alcohol that is responsible for cellular activation (figure 2) but that compounds present with red wine extract are instead biologically active in regards to basophil activation. Intriguingly, no activation was observed in subjects not reporting AH (figure 2a) arguing for specific induction of expression of biochemical pathways being engaged in subjects who develop this condition. Because of its extensive pharmacologic behaviors,^22–25 we explored the role of resveratrol as a relevant component within red wine, but demonstrated no role in driving basophil activation. Interestingly, the biological activities of resveratrol include inhibition of cyclooxygenase, again, supporting the concept that AH is not related to the more serious reactions that occur in AERD. In contrast, the biologically active polyphenolic compounds catechin and epigallocatechin²⁴ at physiologically relevant concentrations did produce basophil activation in the AH cohort (figure 2b). The inability of these compounds to fully reproduce the activity of RWE argues for a role for additional uncharacterized mediators.

In support of the concept that platelets could mediate a pivotal role in driving basophil activation we interrogated the relative activation of platelet-non-adherent and platelet-adherent basophils (figure 2b). While some activation of non-adherent basophils was observed, approximately 2-fold greater activation was observed by platelet-adherent basophils. Direct evidence for platelet activation was confirmed as the secretion of the platelet-specific mediator TxB₂ in response to RWE (figure 4). Importantly, these studies argue for an influence of platelets promoting basophil activation and not the reverse insofar as neither fMLP nor anti-IgE were capable of producing TxB₂ secretion. It has previously been established that, at least in the AERD population, aspirin desensitization and continuing administration restores tolerance to alcoholic beverages.²⁶ Given the ability of aspirin to inactivate platelets, these observations invite exploration as to whether platelet activation by itself suffices to drive AH without need for the involvement of associated leukocytes. Platelet-derived mediators including serotonin and thromboxane) would suffice to explain the flushing responses characteristic of AH, although the full spectrum of respiratory symptoms that these patients develop requires the additional contribution of CysLTs and other vasoactive mediators (figure 5).

Figure 5.

Model of polyphenolic activation of alcohol hypersensitivity reactions. See text for details.

In summary, these studies confirm the high prevalence of AH in individuals with CRS and, especially, AERD. Individuals with AH demonstrated higher numbers of circulating platelets and, at least in regards to the ability of polyphenolic compounds contained within red wine, it was the platelet-adhered basophils that were largely being activated. These studies invite exploration as to the specific pathways being engaged by polyphenolic compounds in AH subjects as well as determining whether it is platelet activation alone that suffices to explain these vasomotor/flushing reactions.

Abbreviations:

ADH

acetaldehyde dehydrogenase

AERD

aspirin-exacerbated respiratory disease

AH

alcohol hypersensitivity

Cat

catechin

COX

cyclooxygenase

CRS

chronic rhinosinusitis

CysLT

cysteinyl leukotrienes

EGC

epigallocatechin

EtOH

ethanol

LT

leukotriene

NP

nasal polyps

PLT

platelet

ResV

resveratrol

RWE

red wine extract

TxB₂

thromboxane B₂