Lower viral loads in rhinovirus-challenged allergic subjects despite reduced innate immunity

Abstract

Background:

Viral infections, especially those caused by rhinovirus, are the most common cause of asthma exacerbations. Previous studies have argued that impaired innate anti-viral immunity and, as a consequence, more severe infections contribute to these exacerbations.

Objective:

These studies explored the innate immune response in the upper airway of allergic rhinitis and asthmatic volunteers in comparison to healthy controls and interrogated how these differences corresponded to severity of infection.

Methods:

Allergic rhinitic, asthmatic, and healthy volunteers were inoculated with rhinovirus A16 and monitored for clinical symptoms. Tissue and nasal wash samples were evaluated for anti-viral signature and viral load.

Results:

Both allergic rhinitic and asthmatic subjects demonstrated more severe cold symptoms. Asthmatic subjects demonstrated worsened asthma control and increased bronchial hyperreactivity in the setting of higher exhaled breath FeNO and blood eosinophils. These studies confirmed reduced expression of interferons and virus-specific pattern recognition receptors in both atopic cohorts. However, despite this defect in innate immunity, allergic rhinitis/asthmatic volunteers demonstrated reduced rhinovirus concentrations in comparison to controls.

Conclusion:

These results confirm that the presence of an allergic inflammatory disorder of the airway is associated with reduced innate immune responsive to RV infection. Despite this, these allergic volunteers demonstrate reduced viral loads, arguing for the presence of a compensatory mechanism to clear the infection.

Introduction

Viral infections, especially those caused by rhinovirus (RV), are the most common cause of asthma exacerbations in children and can also cause exacerbations in adults.^1–5 The immune mechanisms underlying this pathogenic response to RV largely remain an enigma. One proposed explanation recognizes an impaired innate anti-viral immune response in asthmatics. Healthy airway epithelial cells (EpCs) effectively recognize and respond to RV through various pattern recognition receptors (PRRs) including the Toll-like receptors (TLRs), specifically TLR2, TLR3, TLR7, and TLR8, the RIG-I-like receptors (RLR) retinoic acid-inducible gene I (RIG-I) and melanoma differentiation-associated gene 5 (MDA5), and also through engagement of the inflammasome.^6–11 Inflammatory mediators released in response to these PRRs include the type I interferons (IFN) IFN-α/-ß, type III IFNs (IFN-λ), as well as numerous cytokines (including interleukin (IL)-1, IL-6, IL-18, and many others). Together these pathways orchestrate an anti-viral response within the EpC that is sufficient to eradicate the virus.^12,13 An impairment of IFN production by EpCs has been proposed as a mechanism that would lead to enhanced RV replication, with the consequent heightened inflammation driving the asthma exacerbation. This impairment of IFN production by EpCs has been established in numerous studies.^14–20 The current studies were designed to concomitantly explore the innate anti-viral immune response engaged by nasal EpCs and severity of subsequent viral infections. We focused on the nasal response as the nasopharynx is the primary sight of the RV infection and the practicality of having greater accessibility of tissue.²¹

Methods

Subjects.

To address the role of allergic (type 2^high) inflammation in modulating the anti-viral immune response in the airway we enrolled healthy controls (n=11) and allergic subjects, including those with allergic rhinitis (AR; n=17) and with concomitant asthma (n=17; not all studies were done on all subjects.) AR and asthma subjects reported upper respiratory symptoms with exposure-relevant positive prick skin tests (sensitization to aeroallergens to which they were currently being exposed) and were studied when symptomatic. Use of oral or nasal antihistamines, intranasal corticosteroids (CCS), or leukotriene modifiers was not permitted. Healthy controls consisted of non-asthmatic, non-allergic subjects with negative prick skin tests to aeroallergens. Subject characteristics are summarized in Table 1. All subjects gave written informed consent under protocols approved by the University of Virginia Institutional Review Board (#19512 and #19517).

Table 1.

Subject Characteristics at Baseline

Cohort	n	Age	Gender (%F)	Skin tests (# positive)	Absolute eosinophil count (cells/μL)	FEV1 (% predicted)	FEV1/FVC	FeNO (ppb)	Methacholine PC20
Healthy Control	11	22.8±0.4	54.5	--	90±10	99.2±4.5	.86±.01	12.9±2.0	>25
Allergic Rhinitis	17	21.5±0.7	58.8	4.8±0.5	150±20	98.9±2.2	.83±.01	24.1±3.3	>25
Asthma	17	21.7±0.5	52.9	4.3±1.1	250±140	88.3±5.1	.81±.03	53.8±18.0	1.89±0.7

Abbreviations: FeNO – fractional exhaled breath nitric oxide; FEV₁ – forced expiratory volume, 1 second; FVC – forced vital capacity; PC – provocative concentration

Experimental infections with RV-A16.

Each subject was inoculated with 300 TCID₅₀ of RV-A16 (IND# 15162). Nasal wash and nasal scraping samples were collected from each subject on days 0, 1, 2, 3, 4, and 7 post inoculation (dpi) as previously described.^22,23 Nasal biopsies were collected on day 4 from the inferior and middle turbinates and the nasopharynx.

Clinical response to RV infection.

Upper respiratory tract symptom scores were evaluated as previously described.²⁴ Briefly, this scoring system includes sneezing, nasal discharge, nasal obstruction, sore throat, headache, malaise, and chills with each scored on a severity scale of 0 to 3 (0 = symptoms not present; 1 = mild, but clearly present; 2 = moderate and uncomfortable; and 3 = severe, interfering with sleep or activity). Spirometry and fractional exhaled breath nitric oxide (FeNO) were measured at each visit. Methacholine reactivity and absolute eosinophil counts (AECs) were assessed on dpi 0 and 4.

Enzyme immunoassays.

EDN concentrations in nasal wash were quantified by enzyme immunoassay (EIA) according to manufacturer’s instructions (MBL, Nagoya, Japan). The lower limit of detection of the assay was 620 pg/ml.

Immunofluorescence staining.

Nasal biopsy samples were fixed in 4% paraformaldehyde, paraffin embedded and sectioned by the Histology Core Laboratory of the University of Virginia. For analyses, samples were deparaffinized in xylene and graded ethanol and then rehydrated in distilled water. Heat-induced antigen retrieval was performed by heating sections for 20 min in citrate buffer (Abcam; Cambridge, MA). Slides were blocked using 1% bovine serum albumin, 10% goat serum (Sigma, St. Louis, MO), 0.1% triton X-100 and 1 μg/ml Fc Block (BD Pharmingen; Sparks, MD) for 1.5 hrs, and then incubated with primary antibody for 16 hrs at 4°C (TLR3 (Novusbio, 40C1285.6, 1:100); RIG-I (Abcam, ab45428, 1:200); MDA5 (Abcam, ab79055, 1:200); human rhinovirus monoclonal antibody (QED Bioscience, 18758, 1:100); IFN-α (Proteintech, 18013–1-AP, 1:100); IFN-ß (Abcam, ab140211, 1:100); or appropriate isotype controls antibodies). Anti-major basic protein (MBP) (1:50) and anti-EDN (1:50) antibodies were gifts from Dr. Hirohito Kita. Sections were rinsed and then incubated with secondary allophycocyanin (APC) goat anti-rabbit/mouse IgG (1:200, Life Technologies) or Alexa Fluor 488 goat anti-rabbit IgG (1:200, Thermofisher) for 1 hour at room temperature. Nuclei were stained with 100 ng/ml DAPI (4’, 6-diamidino-2-phenylindole, Sigma) for 30 min at room temperature. Samples were washed and aqueous mounted with VectaMount AQ (Vector Laboratories; Burlington, CA).

Histological scoring.

The samples were analyzed using an EVOS FL Auto microscope (Life Technologies). EDN and MBP were quantified as positive area fraction (%) by ImageJ in multiple random high-power fields. IFN-α, IFN-ß, RIG-1, MDA5, TLR3 positive grade and RV positive cell numbers were scored in a blinded fashion based on multiple random high-power fields per slide. Grading was limited in the epithelial cells and scored as 0=no staining, 1=occasional positive staining/mild, 2=cluster of positive staining/moderate, 3=large area of positive staining/extensive; and 4=staining in all cells.

Quantitative real-time polymerase chain reaction (qPCR).

qPCR was performed on RNA extracted from nasal scrapes. Total RNA was extracted using TRI^® reagent (Sigma, St. Louis, MO). Conversion of mRNA to cDNA was performed using a Taqman Reverse Transcription kit (Roche, Branchburg, NJ). Total RNA (200 ng) was added to each reaction along with oligo dT primers, 5.5 mM MgCl₂, 2 mM dNTPs, RNase inhibitor, and reverse transcriptase. Reactions went through 10 min at 25° C, 30 min at 48° C and 5 min at 95° C in a Bio-Rad iCycler thermocycler (Bio-Rad). The PCR mix consisted of Sensimix (Bioline; Taunton, MA), cDNA and 400 nM of each primer. qPCR was performed for IFN-α, IFN-ß, IFN-λ, RIG-I, MDA5, and TLR3 using commercial primer pairs (Integrated DNA Technologies, Inc., Coralville, IA). Data were analyzed as the difference in C_T (ΔCT) of each transcript from the C_T ß-actin (ΔC_T).

Proteomics.

EIAs performed on nasal lavage fluid for IFN-α, -ß, and -λ were all below the threshold of sensitivity of these assays, reflecting the extensive dilution. To address expression of other cytokines we utilized a commercial proximity extension assay (PEA)²⁵ which included ability to assess cytokines associated with inflammasome activation (IL-1 and IL-18). The PEA is an affinity-based assay which semi-quantitatively characterizes protein abundance levels. For each measured protein, a pair of oligonucleotide-labeled antibody probes target the protein and if both probes are in close proximity, a PCR target sequence is formed by a proximity-dependent DNA polymerization event. The resulting sequence is then detected and quantified using standard real-time PCR.

RV viral titer determination.

RNA was extracted from nasal wash fluid using RNeasy RNA isolation kits (Qiagen, Crawley, United Kingdom) and cDNA was generated as previously described.²⁶ The cDNA was amplified and detected via qPCR utilizing primers specific for conserved regions of RV and detected via qPCR (RV forward 5’- CCTCCGGCCCCTGAAT-3’; RV reverse 5’- AAACACGGACACCCAAAGTAGT-3’; Integrated DNA Technologies). The qPCR mix consisted of 2X SYBR green Supermix (BioRad), cDNA, and 400 nM of each primer. Nasal wash samples from the experimental RV-A16 challenge were evaluated in duplicate and the values compared to a standard curve to determine the copies of viral RNA/μL of nasal wash supernatant. The neat concentration of the RV-A16 pool used to generate the standard curve in this analysis contained 5×10⁷ copies of viral RNA/μL.

Statistical analyses.

For comparisons between the cohorts and with baseline (dpi0) samples, data were compared using ANOVA or t-tests with Bonferroni correction for multiple comparisons. A p-value of <0.05 was considered statistically significant. Statistical analyses were performed using GraphPad Prism 6 (La Jolla, CA).

Results

Clinical response to RV-A16 inoculation.

RV-A16 inoculation produced symptomatic infections in all volunteers. Allergic inflammation (AR and allergic asthma) was associated with both more severe and persistent symptoms (figure 1A; p<0.05 for days 2 and 3). Asthmatics demonstrated significant worsening of FEV₁ (87% predicted at dpi0 to 76% at dpi4) and of FEV₁/FVC ratio (.81 to .68; p<0.05; figure 1B). In addition, asthmatics had an approximately 2 log₂ increase in bronchial hyperreactivity (PC₂₀ 1.89 to .40 mg/ml).

Figure 1:

Clinical response to RV infection. 1A. Upper respiratory symptoms in control (C), AR, and asthmatics (A) post-viral inoculation. 1B. FEV1/FVC ratio. 1C. Circulating absolute eosinophil count. 1D. FeNO. *p<0.05 compared to control subjects; ^§p<0.01 compared to day 0; ^†p<0.05 compared to day 0.

Tissue eosinophilic inflammation post-infection.

We next investigated several parameters of eosinophilic inflammation as a correlate to the clinical impact observed in response to infection. Asthmatic (but not AR) subjects demonstrated a significant increase in AEC at day 4 (250 to 400/μL); figure 1C; p<0.01). Indirect evidence of increased type 2 inflammation in the lower airways of asthmatics was supported by significant increases in FeNO (figure 1D). Intact eosinophils were only rarely observed in nasal wash fluid or tissue samples. The presence of an eosinophilic inflammatory response was instead quantified in the day 4 biopsy samples by performing IF staining for EDN and MBP. Because of limitations in tissue sample availability, these tissue and proteomic analyses (below) were performed as a comparison between healthy controls and subjects with a type 2^high upper airway inflammatory state (pooled samples from AR and allergic asthmatics). Figure 2A displays representative data from a healthy control and allergic subject. Negligible staining was observed with isotype control antibodies (on line supplement, eFigure 1). Blinded semiquantitative scoring for IF was performed as described. Little or no staining for either peptide was observed in the control specimens whereas elevated staining for both was observed in the allergic cohort (p<0.01 for EDN; figure 2B–C). Under the study protocol, we were not permitted to obtain baseline IF data regarding expression of EDN or MBP prior to inoculation, so it cannot be categorically argued that the IF data reflected de novo RV-induced expression of these peptides. However, induction of eosinophilic inflammation by RV was demonstrated in the allergic volunteers as demonstrated by the higher EDN expression in the allergic (AR and asthma) cohort (p<0.01) with significantly higher release of EDN into the nasal wash fluid observed at day 4 (p<0.05; figure 2D).

Figure 2:

Immunofluorescence staining of eosinophil-derived EDN and MBP at dpi 4. 2A. Representative staining. 2B: Summary of expression of EDN. **p<0.01. 2C: Summary of expression of MBP. 2D: Nasal wash EDN concentrations as determined by EIA. **p<0.01 atopic compared to controls. ^#p<0.05 atopic subjects day 4 compared to day 0.

mRNA transcript expression of RV-detecting and responding molecules.

To explore the innate immune anti-viral signature in healthy controls and subjects with allergic inflammation of their upper airways, initially, we investigated the innate immune response of nasal epithelium at varying time points after inoculation. We examined innate molecules responsible for RV recognition (toll-like receptors for viral RNA and the RIG-I-like receptors RIG-I and MDA5) as well as expression of type I and III interferons (IFN-α, -ß, and -λ). RNA was extracted from nasal scrapings of infected individuals prior to inoculation (dpi 0) and post-infection on dpi 1–4 and 7 and gene expression determined by qPCR. Over the course of the infection expression of both type I (IFN-α/IFN-ß) and type III (IFN-λ) interferons was significantly higher in the control compared to the atopic subjects (p<0.01 for type I and <0.001 for type III IFNs). Expression of IFN-α and IFN-ß were similar at baseline whereas IFN-λ was significantly decreased in the allergic subjects (figure 3). After inoculation there was no change in expression of any of these IFN transcripts in the type 2^high subjects. In contrast, statistically significant increases in transcript expression for IFN-α and IFN-ß were observed at dpi 2 and dpi 1, respectively in the healthy controls (p<0.05). Transcript expression for the TLRs and RLRs were similar in all 3 cohorts at baseline and no significant change in expression was observed with infection (data for TLR3, RIG-I and MDA5 are displayed in figure 3).

Figure 3.

qPCR for RV sensing and responding molecules: 3A – IFN-α, 3B – IFN-ß, 3C – IFN-λ, 3D – RIG-I, 3E – TLR3, 3F – MDA5. Vertical bars on the right of each graph indicate significant difference between the two groups. and the * above, indicate the specific day these differences were significant. *p<0.05, **p<0.01, ***p<0.001.

Protein expression of RV-detecting and -responding molecules.

Enzyme immunoassays for IFN-α, -ß, and -λ were performed on nasal lavage fluid but were below the detection limit in all samples collected, presumably reflecting excessive sample dilution. We performed ultrasensitive proximity extension assays (PEA) to detect cytokines released into the nasal lavage fluid in response to inflammasome activation, specifically IL-1 and IL-18. While these proteins were both detectable via this methodology, no differences were observed between the control and allergic subjects and only a slight insignificant (1–2 log₂ (~2–4-fold)) increase in IL-18 was observed in response to the infection (online supplement, eFigure 2).

Immunofluorescence.

Given the inadequacies of lavage fluid assays, protein expression was subsequently analyzed via immunofluorescence (IF) of nasal biopsies obtained at day 4. We performed IF staining to quantify expression of IFN-α, IFN-ß, TLR3, RIG-I, and MDA5. Representative IF data from a healthy control and allergic subject are displayed in figure 4A (with isotype control antibodies displayed in eFigure 1). Compared to controls, significantly reduced expression of IFN-α and IFN-ß (p<0.05) was observed in the type 2^high (allergic) cohort (figure 4B) with smaller insignificant differences observed for TLR3, RIG-I, and MDA5 (p=0.09 for TLR3).

Figure 4:

Immunofluorescent staining of RV-sensing and anti-viral responding molecules. 4A. Representative immunofluorescent staining. Images are shown at 200x with digital enlargement of key areas (dotted lines) to enhance visibility of details (inset, solid lines). 4B. Expression of RV-sensing and anti-viral responding molecules via IF. Statistical analyses involved unpaired t-tests. *p<0.05.

RV titer post inoculation.

Having confirmed previous studies demonstrating defects in components of the innate immune system in atopic subjects,^14–20 we queried whether this would impact RV titer. RNA was extracted from nasal wash collected during the infection and viral titer determined via qPCR. In addition, we quantified RV-infected cells by IF in the nasal biopsy samples. In the day 4 biopsy samples we were no longer able to identify infected epithelial cells (presumably reflecting the rapid clearance of infected EpCs via apoptosis and engulfment).^27,28 However, in contrast to epithelium, positive IF was observed in the submucosal tissue, especially in the healthy controls (figure 5A). Positive IF for RV was significantly reduced in the atopic subjects (p<0.05; figure 5B). This lower expression of RV in the atopic subjects despite their reduced expression of type I and type III IFNs was confirmed when we quantified virions in the nasal lavage fluid (p<0.01; figure 5C).

Figure 5:

Rhinovirus expression post-inoculation. 5A. Representative RV immunofluorescence staining showing viral protein in submucosal (but not epithelial) tissue. 5B. Summary of IF of RV positive cells at dpi 4. *p<0.05. 5C. Viral load in nasal wash fluid. NWF – nasal wash fluid. **p<0.01.

Discussion

These studies further establish the experimental RV inoculation model as a robust methodology to explore the cellular and molecular mechanisms underlying the capability of this virus to induce asthma exacerbations. In contrast to previous studies dependent on a less potent RV (e.g., RV-A39),^24,29 the current studies with RV-A16 demonstrated significant worsening of asthma as manifested as more severe respiratory symptoms, decreased lung function, and increases in FeNO and bronchial hyperreactivity (figure 1). In addition, in comparison to our more recent published studies using RV-A16,²³ the current challenges were also more likely to have further induced worsening asthma as a result of the removal of volunteers from an aeroallergen-free environment and allowing volunteers to be exposed to relevant bystander allergens. This outcome is consistent with the concept that RV-induced exacerbations are mediated in large part by the tendency of the virus to potentiate an ongoing immune response to bystander allergens.^3,23,29–33

In asthmatics, RV infection is associated with rapid expression of numerous cytokines/chemokines associated directly or indirectly with eosinophil activation and recruitment.^34–37 It is well recognized that activation products of eosinophils damage airway epithelium, promote bronchospasm, and enhance bronchial hyperreactivity (reviewed in³⁸). We therefore investigated the recruitment of eosinophils into the nasal airspace in a timeframe consistent with the clinical response to RV inoculation. Intact eosinophils were only rarely observed in nasal wash fluid or tissue samples. However, these studies demonstrated early expression of the eosinophil-derived peptide EDN in nasal lavage fluid (figure 2D) and biopsies obtained at day 4 post-inoculation confirm diffuse staining for both MBP and EDN in the AR and allergic asthmatic volunteers (figure 2B and 2C). It is intriguing to note that both this influx of eosinophilic mediators as well as the induction of blood eosinophilia (in asthmatics only; figure 1C), while within a timeframe consistent with the worsening of asthma symptoms, is well prior to the time required for an adaptive immune response to develop. Effector memory T cells will not be engaged until dpi 5–7 as suggested by our group’s previous publications.^39,40 Mechanisms for epithelial recruitment of eosinophils include epithelial cell generation of innate lymphoid cell 2-activating cytokines and eosinophil-targeting chemokines.^41,42 These data invite further investigations regarding the specific pathways initiated by RV-infected epithelium that drive the rapid induction of eosinophilic inflammation. This is of particular importance given the likely central relevance of these processes in driving the adverse clinical consequences.

Previous studies have provided evidence for decreased innate anti-viral immunity in the airways of asthmatics, suggesting an enhanced capacity for viral replication. This model argues that this would be associated with an increase in viral load after infection, thereby driving the asthma exacerbation. Despite this, our published studies quantifying RV in the nose showed that asthmatics displayed lower viral load in their nasal passages in comparison to either controls presenting to the ED with cold symptoms or after experimental viral inoculation.^22,23 While RV does infect the lungs of asthmatics and infected bronchial epithelial cells can be identified,⁴³ the primary target of RV infection is the nose.⁴⁴ The reduction in innate immunity is thought to develop as a consequence of the type 2 inflammatory state present in asthma and – in the upper airway – also in allergic rhinitis.^45,46 We therefore initially investigated whether the defect in innate anti-viral immunity extended to the nose, reflecting this “unified airway.” Our initial investigations confirmed a defect in the expression of type I and type III IFN immune responses in the upper airway of atopic volunteers (figures 3–4).^14–20 However, despite this deficiency, we confirmed our earlier studies that the viral load observed in the nose of inoculated allergic volunteers was lower than in non-allergic healthy controls (figure 5).

These seemingly discordant findings signify that compensatory pathways are being engaged in the allergic subjects that must be responsible for eliminating the virus. One possibility is that it is the potentiated eosinophilic inflammation that provides this protection. Numerous studies demonstrate the capacity of eosinophils to eliminate respiratory RNA viruses including respiratory syncytial virus (RSV) and influenza. This reflects, at least in part, the potent RNase functionality of their cationic proteins,^47–54 such as eosinophil-derived neurotoxin (EDN), that degrade the viral RNA.^48,54 More recently, a favorable outcome to SARS-CoV2 infection was observed in asthmatics, but specifically only in those asthmatics with the highest eosinophil counts.⁵⁵ An argument for the concept that eosinophils may target RV is supported by a study in which induction of eosinophilic inflammation led to less severe infections.⁵⁶ And, more impressively, was the demonstration that depletion of eosinophils by administration of an anti-IL-5 biologic was associated with ~10-fold increased viral loads.^57,58 The capacity of eosinophils to target rhinovirus has not been extensively explored. It should be noted, however, that in one conflicting study involving resting or fMLP-stimulated eosinophils, at least ex vivo, there was no reduction of viral load in an RV-infected epithelial cell line.⁵⁹ And it is certainly plausible that a more robust response by, for example, NK cells or tissue memory cytotoxic lymphocytes, if present in the allergic volunteers, could provide alternative explanations for the decreased viral loads that were observed.

Together these findings suggest a specific pathogenic mechanism for RV-mediated asthma exacerbations (this model is displayed in figure 6). While induction of an eosinophil-mediated anti-viral response could compensate for reduced IFN-mediated anti-viral immunity, this compensation will occur at the cost of inducing the asthma exacerbation. This mechanism is supported by the paradox that whereas, eosinophil eradication via treatment with an anti-IL-5 monoclonal antibody, as noted, promoted viral replication, the concomitant eradication of the eosinophils prevented this from impacting symptoms.⁵⁸ Indeed, the absence of worsened asthma despite much higher viral loads, points to the overarching significance of exacerbation of a bystander type 2 inflammatory state in driving RV-induced asthma exacerbations as opposed to any direct adverse impact from the RV infection itself. This concept is certainly consistent with the recognition that targeting type 2 inflammation with, in addition to anti-IL-5, either with anti-IL-4Rα or anti-IgE typically suffices to mitigate asthma exacerbations, many or most of which can be presumed to have been associated with natural RV infections.

Figure 6.

Model for the development of asthma exacerbations in response to RV infections. See text for details.

Abbreviations

APC

allophycocyanin

AR

allergic rhinitis

CCS

corticosteroids

DAPI

4’, 6-diamidino-2-phenylindole

dpi

days post inoculation

EDN

eosinophil-derived neurotoxin

EpCs

epithelial cells

IF

immunofluorescence

IFN

interferon

IL

interleukin

MBP

major basic protein

MDA5

melanoma differentiation-associated receptors

PEA

primer extension assay

PRR

pattern recognition receptors

RIG-I

retinoic acid-inducible gene I

RLR

RIG-I-like receptor

qPCR

quantitative real-time polymerase chain reaction

RSV

respiratory syncytial virus

RV

rhinovirus

TCID

tissue culture infective dose

TLR

toll-like receptors