Unique Associations Between Complement Split Products and Pollen Species in Adults with Respiratory Allergic Disease: A Pilot Study

Abstract

Background

Complement split product (CSP) anaphylatoxins, C3a and C5a, are important mediators of innate immune responses and have been implicated in the pathogenesis of asthma. Environmental pollen exposure is a trigger of asthma symptoms, leading to generation of CSP, which amplify allergic responses. The relationship between seasonal pollen levels, complement activation and clinical outcomes of asthma control remain unknown. This exploratory pilot study evaluated the association between concurrent tree, grass, and weed pollen exposure, CSP levels, and measures of asthma control in adults with asthma.

Methods

In this prospective study, which included 11 allergic adults, blood samples were drawn monthly for measurement of plasma C5a desArg and C3a desArg levels (ELISA). Concurrent local tree, grass, and weed pollen exposure, as well as clinical measures of asthma control (ACT) were determined. Spearman correlation matrices were computed.

Results

Tree pollen counts were positively associated with C3a desArg expression (r=0.562, p<0.0001). Weed pollen counts were negatively associated with C3a desArg expression (r=−0.419, p<0.0001). Increasing weed pollen counts decreased FeNO (β=0.956, p<0.001). Higher C5a desArg levels over time correlated with increased ACT scores (p=0.02), improved AQLQ and RQLQ (P=0.03, 0.03, respectively).

Conclusion

Seasonal pollen exposure is associated with distinct patterns of complement activation in asthma. Larger studies are required to elucidate the relation of these patterns of complement activation to immune mechanisms in asthma pathogenesis and clinical disease.

Introduction

Asthma is a chronic inflammatory airway disease characterized by airway hyperresponsiveness and recurring respiratory symptoms.¹ Environmental allergens, specifically seasonal pollens, are well‑recognized triggers of asthma.² Exposure to airborne pollens can induce both innate and adaptive immune responses, resulting in airway inflammation and impaired quality of life.³ While the role of allergen-specific IgE and type 2 inflammation has been studied, the contribution of innate immune pathways, including complement activation, remains undefined.⁴

The complement system is a critical component of innate immunity that participates in host defense and inflammatory regulation.⁵ The anaphylatoxins C3a and C5a, both complement split products, may be generated by allergens, including dust mite proteases and pollens⁶, tissue-based, phagocyte derived proteases,⁷ and human mast cell beta tryptase.⁸ In humans, mast cell-produced C5 and C3 have autacoid activity after tryptase cleavage which potentiates allergic inflammation.⁹ In a murine model of allergic asthma, Karp et al reported that a deficiency of C5 was linked to allergic inflammation.¹⁰ This suggests that a lack of C5 stimulation of macrophages skews responses from proinflammatory T helper (Th)1 toward Th2 allergic responses.¹¹ It is unknown how complement levels link systemic pro-allergic with non-allergic, pro-inflammatory responses in humans. Seasonal fluctuations in pollen exposure have been associated with changes in airway inflammation and clinical symptoms; however, little is known between ambient pollen levels and systemic complement responses. Understanding how pollen exposure affects complement activation may provide insight into mechanisms linking environmental triggers to asthma outcomes.

The aim of the current study was to prospectively evaluate the association between outdoor seasonal pollen exposure, circulating blood CSP and clinical measures of asthma control and quality of life in adults with asthma. We specifically examined the relationships between tree and weed pollen counts, circulating levels of C3a desArg and C5a desArg, asthma control test (ACT) scores and quality of life outcomes, to better understand the role of complement activation in allergen-driven asthma. We hypothesize that pollen counts associate with plasma complement levels, which then mediate clinical disease responses. The primary outcome was to explore whether CSP (C5a desArg and C3a desArg) associate with asthma control over time as determined by repeated measurements.

Methods

Subjects

This is a prospective study of adults (ages 18–75 years) receiving ongoing outpatient allergy care at the University Hospital at Downstate (UHD) of SUNY Downstate Health Sciences University (Brooklyn, NY) between December 2006 and June 2008. Written informed consent was obtained from all participants prior to their enrollment in the study. Approval was obtained for this study from the SUNY Downstate Health Sciences University Institutional Review Board (IRB# 569596; initial approval 8/7/2006). The study was conducted in accordance with the ethical principles of the Declaration of Helsinki and all applicable institutional regulations. Subjects had been diagnosed with allergic asthma and/or rhinoconjunctivitis (EPR-3).¹²

Eligible subjects underwent previous allergy evaluation with either epicutaneous skin prick testing (DermaPik device; Greer Laboratories, Lenoir, NC, USA) and/or with allergen-specific in vitro testing. Skin testing was done with following extracts: dust mite (Dermatophagoides pteronyssinus and/or Dermatophagoides farinae; Hollister-Stier, Spokane, WA); American cockroach; ragweed (short, tall), 5-grass mix (Timothy, Orchard, June, Red Top, Sweet Vernal); English plantain, 10-Tree Mix A (Ash, Beech, Birch, Elm, Hickory, Maple, Oak, Poplar, Sycamore, Alder); Cladosporium sphaerospermum; Alternaria; dog epithelium; and cat pelt (Alk-Abello, Round Rock, TX). Respiratory allergen-specific IgE levels were measured in the Diagnostic Immunology Laboratory at SUNY using a fluoroenzyme immunoassay (Pharmacia Upjohn, Unicap System). Study measurements were obtained monthly (median 8 months) from 11 adults; with continual enrollment; subjects began the study throughout the year. Rolling enrollment was used to accommodate subjects’ availability to ensure adherence to visit schedule. Data points were analyzed independently due to the small sample size.

Blood Complement Split Product Measurement

Blood (5 mL) was obtained monthly. Samples were drawn for plasma C5a-desArg and C3a-desArg levels into tubes with disodium EDTA (Quest Diagnostics, USA). The samples were centrifuged within 1 hour for 15 minutes at 1000 x g. Aliquots of plasma were stored at −70° C until assayed. CSP levels were determined using the OptEIA™ ELISA test (Pharmingen, San Diego, CA), according to manufacturer’s recommendation, and KCjunior software (Bio-Tek Instruments, Inc). The normal ranges for C3a desArg were approximately 1080 ± 189 ng/mL and for C5a is approximately 3.9 ± 1.5ng/mL in healthy controls.^8–10

Total Serum IgE

A baseline total serum IgE level was determined at the first visit at the Diagnostic Immunology Laboratory at SUNY Downstate Health Science University (Brooklyn, NY) using the fluoroenzyme immunoassay. Normal IgE range for a healthy adult is <100 IU/mL.

Clinical Disease Measurements

Clinical Questionnaires

Concurrent scores were determined for asthma control (ACT)¹³ and measures of QOL (AQLQ,¹⁴ RQLQ).¹⁵ The subjects’ symptoms were assessed by using both the Asthma Quality of Life Questionnaire (AQLQ) and Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ). The Asthma Quality of Life Questionnaire (AQLQ) is a validated, asthma-specific document. Patients rate the degree of impairment caused by asthma during the preceding 14 days and respond to each of 32 items using a 7‑point scale from 1 (maximal impairment) to 7 (no impairment). The responses were grouped and assessed for each domain. Total scores were used in statistical analysis. The RQLQ has a similar design, except that higher scores are associated with increased symptoms.

Exhaled Nitric Oxide (FeNO)

Exhaled nitric oxide (FeNO) levels were measured (NIOX Mino; NIOX Group, Oxford, United Kingdom) according to guidelines.¹⁶

Pollen Counts

Concurrent local grass, tree, and weed pollen counts were obtained with permission from the American Academy of Allergy, Asthma and Immunology (AAAAI) National Allergy Bureau (NAB) and its participating station at Long Island College Hospital (LICH) (Brooklyn, NY).

Statistical Analysis

Data was analyzed using Spearman correlation matrices and log-linear (Poisson) regression.

(SPSS software v25.0; Chicago, IL). A P value of <0.05 was determined statistically significant.

Results

Demographics

Demographics and clinical factors of study subjects are shown in Table 1. Measurements were obtained monthly (median 8 months) from 11 adults. The study subjects had a mean age of 50.7 yrs. ±9.3; subjects were female 8/11 (73%) and male 3/11 (27%). Diagnoses included allergic rhinoconjunctivitis with asthma (N=10) and allergic rhinoconjunctivitis alone (N=1). Clinical factors included mean total serum IgE levels (557.4 IU/mL ±529.54), tree, grass, and weed pollen counts, C3a desArg and C5a desArg levels, FeNO, ACT, AQLQ and RQLQ scores (Table 1). Further baseline characteristics of the study subjects include the following: Allergen sensitizations were as follows: tree mix –7/11, grass mix- 6/11, ragweed –7/11, dust mite, 11/11, cockroach – 7/11, cat – 6/11, English plantain 1/11. All subjects had allergic rhinoconjunctivitis, Asthma severity was as follows: 9/11 had moderate persistent asthma treated with inhaled corticosteroids, 1/11 with mild persistent asthma treated with inhaled corticosteroids, 1/11 with allergic rhinoconjunctivitis treated with intranasal corticosteroids. Three of 11 patients were current smokers.

Table 1.

Patient Demographics and Clinical Factors

Patient	Age (yrs)	Sex	Tree Pollen Count (Grains/hpf)	Grass Pollen Count (Grains/hpf)	Weed Pollen Count (Grains/hpf)	C3a Des Arg (nl: 1080 ±189 ng/mL)	C5a Des Arg (nl: 3.9 ± 1.5 ng/mL)	FeNO (ppb) (<25)	ACT Score (0–25)	AQLQ Score (0–224)	RQLQ Score (0–196)
1	45	F	0.5±0.8	0.8±0.8	1.5±2.8	1139.1±529.1	5.1±3.1	9.0±1.7	24.2±1.1	201.2±16.9	77.0±22.6
2	53	F	0.4±0.9	0.6±0.9	1.8±3.0	1055.1±137.3	3.3±1.4	24.8±17.0	11.4±6.2	96.1±25.8	88.6±11.9
3	38	F	0	2.0±1.7	1±1.7	1298.5±336.6	4.3±1.7	8.7 5±1.7	19.8±3.9	156.3±25.9	51.2±15.
4	70	M	2.5±5.0	2.5±1.9	2.3±2.6	1040.4±67.5	5.0±2.0	20.7±2.8	14.4±3.3	165.2±19.2	14.6±12.3
5	45	M	1±1.4	2.5±2.1	1.0±1.4	1083.6±230.3	1.4±1.1	22.3±2.5	N/A	N/A	N/A
6	53	F	24.5±42.6	1.3±1.0	2.0±4.0	1118.7±234.3	4.1±2.6	17.8±10.1	10.0±2.6	80.7±11.5	81.6±7.5
7	44	F	19.0±36.7	5.5±8.5	0.3±0.5	1049.3±188.5	3.5±2.5	54.5±11.8	13.4±2.1	113.6±6.6	62.6±18.5
8	45	F	14.0±29.7	3.5±7.2	0.8±1.2	994.1±244.4	5.9±2.8	58.6±28.4	10.9±1.6	118.0±15.0	23.9±13.7
9	61	F	0.5±1.0	0.3±0.5	0.8±1.0	1391.1±728.6	2.6±1.0	54±22.3	13.2±2.3	133.0±17.4	105.2±3.6
10	58	F	12.6±30.1	3.3±7.2	1.2±1.8	1079.2±174.2	6.7±0.9	70.1±14.4	17.1±4.0	181.1±17.2	51.6±18.4
11	46	M	2.5±5.0	1.8±2.4	1.3±1.3	1246.3±384.3	6.9±2.6	8.4±6.5	17.8±2.6	160.1±32.3	50.6±8.7

Notes: Patients 1–10 had asthma and allergic rhinoconjunctivitis. Patient 11 had allergic rhinoconjunctivitis only (F: female, M: male). Pollen counts are presented as mean± standard deviations (grains/hpf), summarized across all measurements for each participant. Total serum IgE levels were 557.4 IU.mL + 529.54.

Abbreviations: C3a des Arg, complement component 3a des-arginine; C5a des Arg, complement component 5a des-arginine; FeNO, fractional exhaled nitric oxide; ACT, Asthma Control Test; AQLQ, Asthma Quality of Life Questionnaire; RQLQ, Rhinoconjunctivitis Quality of Life Questionnaire.

Association Between Seasonal Pollen Counts and CSP

Tree pollen counts were moderately positively correlated with C3a desArg expression (r=0.562, p<0.0001, Figure 1); weed pollen counts were strongly and negatively associated with C3a desArg expression (r=−0.419, p<0.0001, Figure 2). No association was found with grass pollen, and pollen counts did not correlate with C5a desArg or FeNO (p>0.05).   Increasing tree (β=1.001) and grass pollen (β=1.005) counts were associated with increased C3a desArg expression (p<0.0001 and p<0.001, respectively); increasing weed pollen counts decreased C3a desArg expression (β=0.791, p<0.0001). No significant associations were observed between pollen counts and C5a desArg or CSP correlation coefficient (r) levels over time.

Figure 1.

Association between tree pollen counts and plasma C3a des Arg levels. N=11. Each point represents an individual participant’s repeated measurements over time.

Figure 2.

Association between weed pollen counts and plasma C3a des Arg levels. N=11. Each point represents an individual participant’s repeated measurements over time.

Association Between Seasonal Pollen Counts and CSP

Tree and grass pollen counts did not predict FeNO levels. Weed pollen counts decreased FeNO levels (β=0.956, p<0.001) (Figure 3).

Figure 3.

Association between weed pollen counts and exhaled nitric oxide levels.N=11. Each point represents an individual participant’s repeated measurements over time.

Association Between CSP Measurements and Clinical Disease Activity

While C3a desArg levels were associated with pollen counts, plasma C5a desArg levels correlated positively over time with improved asthma control (ACT) (P=0.023); and asthma quality of life AQLQ) (p=0.032) and rhinoconjunctivitis QOL (RQLQ) (p=0.033) (Table 2). C5a desArg levels were not significantly correlated with FeNO levels.

Table 2.

Correlation of Plasma C5adesArg Levels with Clinical Parameters in Adults with Respiratory Disease

Parameter	R	P value
Asthma control (ACT)	0.260	0.023
Asthma QOL (AQLQ)	0.240	0.032
Rhinoconjunctivitis QOL (RQLQ)	0.239	0.033

Note: Spearmans’ rank correlation analyses.

Regression Associations

Exploratory log-linear Poisson regression was attempted; however, estimates were unstable with wide confidence intervals because of the small sample size (data not shown).

Discussion

In the current pilot study, blood levels of C3a desArg positively correlated with concurrent tree and grass pollen counts but negatively correlated with weed counts. Greater weed pollen counts correlated with lower FeNO levels. Higher blood levels of C5a desArg over time correlated with improved asthma and rhinoconjunctivitis scores. These data suggest that pollen exposure in adults with respiratory allergic disease may be associated through localized inflammatory responses in the bronchi, with altered blood complement levels, which may, in turn, affect clinical disease activity. These findings demonstrate that environmental pollen exposure is associated with differential complement activation and that CSP may have distinct relationships with asthma control and quality of life. The underlying mechanisms between weed pollen count associations with both lower C3a desArg and FeNO levels are not well-established.¹⁷

Our preliminary findings on the association of pollen with blood complement responses broaden extant knowledge of how inhaled allergens relate to blood complement. It is well known that C3a contributes to the production of nitric oxide (NO).¹⁷ We posit that weed pollen may potentiate Th2-driven airway inflammation, leading to a decrease in C3a desArg, which may then be subsequently associated with lower FeNO levels.¹⁸ In addition, it is known that dust mite proteases cleave complement.^6,19 The relation of plasma CSP levels to indoor dust mite (Der F1 and Der p1) bedding allergen levels has previously been studied.¹⁹ While plasma CSP levels were not found to correlate with measured bedding dust mite allergen levels, CSP (both C5a desArg and C3a desArg) correlated with Der f1- and Der p1‑specific IgE levels.¹⁹ The correlation between CSP levels and the degree of dust mite-specific IgE suggests that ongoing, high-level Th2 sensitization is functionally linked to chronic complement consumption and activation. This CSP–dust mite specific IgE link suggests that in sensitized individuals, complement is potentially chronically activated by the allergen, making the resulting CSP a direct measure of current systemic allergic inflammatory burden and therefore a useful potential biomarker of disease activity or control. The complement activating capacity of dust mite allergen may also potentially increase CSP levels in those who inhale it, but are not dust mite allergic, but who, nonetheless, have generated CSP drive allergic responses. Furthermore, the capacity to monitor reductions in C3a desArg or C5a desArg could serve as a valuable, potential pharmacodynamic endpoint to assess the efficacy of immunomodulatory treatments notably subcutaneous immunotherapy (SCIT), sublingual immunotherapy,²⁰ and monoclonal antibody (mAb) therapy in their ability to suppress the allergen-induced complement cascade.

The association of increasing weed pollen with decreased FeNO suggests a functional link between complement activation and T helper (Th)2-mediated airway inflammation, as FeNO is well-known to be a surrogate marker for IL-13-driven Th2 inflammation. Although higher pollen exposure would generally be expected to increase airway inflammation, the inverse relationship observed in our cohort may reflect seasonal variation in inflammatory phenotypes, differences in allergen sensitization, environmental confounders, or the effects of ongoing controller therapy. These findings underscore the complexity of interactions between environmental exposures and indicators of airway inflammation.

Another finding that we observed was that increased C5a desArg levels over time were associated with better asthma control (higher ACT scores), and asthma QOL measures. Charlot et al reported an inverse relationship of plasma C5a desArg levels to asthma severity in allergic children.²¹ C5a/C5aR signaling can regulate dendritic cell function and modulate type 2 immune responses in allergic airway disease, suggesting a context-dependent immunoregulatory role for C5a in asthma.²²

Higher blood levels of C5a desArg over time correlated with improved clinical responses. When detected in tissue, C5a and C3a have been associated with symptoms of organ-specific clinical allergic disease.^23–26 This dual functionality of C5a and C3a is in line with previous context-associated roles of these CSP²² C5a/C5a desArg is present in induced sputum from asthmatics and patients with chronic obstructive pulmonary disease (COPD).^23–25 Levels of CSP are increased in bronchoalveolar lavage (BAL) fluid after segmental allergen provocation.²⁵ Fregonese et al demonstrated increased expression of receptors C5a and C3a in lung tissue of patients with fatal asthma.²⁴ Overexpression of these receptors is also present in the nasal mucosa of patients with allergic rhinitis.²⁶ In murine models, ingestion of peanuts has been shown to contribute to anaphylactic shock by directly activating the complement system, leading to C3a generation in the absence of adaptive T-cell responses.²⁷ The differing relationships observed for C3a and C5a further support the concept that individual complement pathways may contribute uniquely to asthma pathophysiology.

Plasma CSP measurements may help in assessing allergic disease activity, including tissue responses, symptom control, and measures of quality of life (QoL). C3a and C5a levels in blood are typically measured using their stable cleavage products, C3a-desArg and C5a-desArg (CSP), using an enzyme‑linked immunosorbent assay (ELISA).²⁸ CSP levels were found to be elevated at presentation in adults requiring emergency room treatment for asthma exacerbations (total serum IgE 500IU/mL ± 561); blood C3a des Arg levels were also increased five-fold and C5a‑desArg levels were increased two-fold.²⁹ These levels did not change at the completion of emergency room treatment of acute asthma.²⁹ Elevated C3a levels in hospitalized patients with severe acute asthma were reported; these levels subsequently decreased after inpatient treatment.³⁰ Bowser et al reported changes in blood CSP levels measured up to 3 hours after receiving routine SCIT (for respiratory allergic disease) correlated with change in mean induced skin wheal diameter at the site of the injection.¹⁰ In that study, the majority of subjects received SCIT regimens which included dust mite, which is known to activate complement.¹⁰ Allergen introduction, either via inhalation (ie, allergic acute asthmatics),²⁹ or by injection, as after SCIT,³⁰ is associated with the generation of CSP which may acutely potentiate allergic responses.

Tissue-based, locally generated CSP anaphylatoxins can drive mast cell-mediated allergic responses.³¹ C3a and C5a are chemoattractants for proinflammatory leukocytes, including dendritic cells and neutrophils.^32,33 Inhalation of tree and grass pollen likely contains proteolytic enzymes that directly cleave complement components (eg, C3), thus rapidly generating C3a. Given that C3a is a strong chemoattractant and pro-inflammatory mediator, its systemic increase following exposure may trigger a broader inflammatory cascade.

This systemic inflammatory response may therefore be relevant to clinical disease assessment and should be interpreted in conjunction with C3d complement workup. C3d assessment, along with IgG, is also a component of the direct antiglobulin testing (DAT) immunohematology workup of various hemolytic diseases where complement is suspected of binding to red blood cells (RBC).³⁴ It is unclear whether atopic patients may exhibit modulations in their C3/C3d levels that may affect the result of a DAT in the process of hemolysis work-up.³⁵ Interestingly, certain patients with severe refractory asthma, sinusitis, atopic dermatitis (AD), or urticaria may receive intravenous immune globulin (IVIG) as adjunctive therapy,³⁶ impacting total IgG levels, that have been reported to adversely affect DAT results.³⁷ Further studies examining the relationship between C3, downstream complement proteases,³⁸ and other select components of the complement system in the setting of atopy and comprehensive immunohematologic assessment are warranted.

This exploratory pilot study has notable limitations, most important being its small sample size and the use of established manufacturer control CSP levels. However, prospective, repeated assessments were helpful in identifying longitudinal trends. These findings may help elucidate the relationship between blood C3a and C5a in allergic disease.

Seasonal pollen exposure was associated with distinct patterns of complement activation in adults with asthma. Tree pollen exposure was associated with increased C3a expression, whereas weed pollen exposure was associated with reduced C3a levels and lower FeNO. Higher circulating C5a levels were associated with improved asthma control and quality of life, suggesting a complex role for complement activation in asthma. Increased C5a desArg may be a marker of immune regulation rather than a direct factor of asthma improvement. These findings support further investigation of complement pathways as mediators of environmental allergen responses and potential biomarkers of asthma disease activity.

This pilot study is exploratory, and the results raise questions that merit further investigation. Future studies should account for potential confounding factors, including seasonality, asthma treatment, adherence to asthma therapy, respiratory infections, and residence distance from environmental monitoring station, to further elucidate these findings. As this study was performed in a large urban center, the potential impact of ambient air pollution on complement activation should be addressed. Validation and expansion of these findings require larger prospective cohort studies. Future studies should also consider evaluating more complex causal pathways using mediation analyses, in which age may serve as a confounder and complement biomarkers as potential mediators.

Funding Statement

Study was supported by SUNY Downstate Health Sciences University divisional funds for Allergy/Immunology. No external or extramural funding (grant or otherwise) was obtained for this study.

Data Sharing Statement

Data is available upon request from Drs Nagarajan and Joks.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas, took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agreed to be accountable for all aspects of the work.

Disclosure

Dr. Nagarajan receives employment income from Merck & Co. Inc, Rahway, NJ, USA. Dr Rauno Joks reports patents from US Patent No. 7,407,764 B2, outside the submitted work. The other authors have no relevant conflicts of interest to disclose.