Coordinated nasal mucosa-mediated immunity accelerates recovery from COVID-19

doi:10.21203/rs.3.rs-2666607/v1

Download PDF

Article

Coordinated nasal mucosa-mediated immunity accelerates recovery from COVID-19

https://doi.org/10.21203/rs.3.rs-2666607/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 20 Feb, 2024

Read the published version in ERJ Open Research →

Version 1

posted

You are reading this latest preprint version

Viral infection due to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) induce a dynamic immune environment. Using nasal mucosal samples in 139 participants from the STOIC study (community-based randomised clinical trial for the use of budesonide in early onset SARS-CoV-2, NCT04416399), we applied predefined immune mediator nodes in relation to clinical outcomes and viral burden. Interferon- and chemokine-dominant nodes increased expression as compared to health, validating our modular approach. Next, we demonstrated that an increase in mucosal immunity-like node consisting of CCL13, CCL17, IL-33, among others was associated with a mean 3.7-day quicker recovery with no primary outcome events, irrespective of treatment arm. By day 14 the mucosal node divided into two daughter nodes linked to interferon molecules and was transcriptionally detectable in nasal cavity basal, hillock and ciliated cells (as per public single cell dataset EGAD00001007718). Our data suggest mucosal-associated mediators are key for early symptom resolution of SARS-CoV-2.

Biological sciences/Immunology/Infection

Health sciences/Diseases/Infectious diseases/Viral infection

Health sciences/Signs and symptoms/Respiratory signs and symptoms

The inflammatory environment generated by viral infections such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is multifaceted. Both the innate immune system, predominately triggered by pattern recognition receptors in epithelial and myeloid cells, as well as the adaptive immune system, produced by humoral and T cell-mediated processes, are vital to detect, control, and inhibit viral replication while simultaneously orchestrating repair and providing immunological memory¹. Importantly, cellular and molecular immunity do not function in isolation but are coordinated with multiple overlapping pathways to confer best protection against invading pathogens^2,3. By elucidating this immune component coordination unbiasedly in infection will provide insight into pathways critical to disease pathogenesis.

The global SARS-CoV-2 pandemic highlighted the importance of lung health and understanding airway immunopathology to prevent disease. The most common clinical symptoms of coronavirus disease 2019 (COVID-19) include fever, cough and dyspnoea alongside other upper respiratory infection-like symptoms; however, SARS-CoV-2 can result in complications leading to severe disease⁴. To prevent severe COVID-19 a total of 13.1 billion doses of vaccines⁵ using strategies including mRNA, chimpanzee adenoviral and inactivated virial platforms^6–9 have been used. In individuals who develop COVID-19, community-based studies have showed success in early use of inhaled budesonide to reduce time to clinical resolution^10,11. As new variants emerge which have the potential to be less susceptible to current vaccines and acquired immunity, continued effort to understand the immunopathology of SARS-CoV-2 to guide future therapeutics is warranted.

The Steroids in COVID-19 (STOIC) study (NCT04416399) examined clinical consequences of community acquired SARS-CoV-2 in two treatment arms (usual care or inhaled budesonide) and upper respiratory tract immune response¹⁰. This study provided detailed daily self-reported symptoms, self-reported recovery, nasopharyngeal viral burden, and upper respiratory tract profiling during the first two weeks of SARS-CoV-2 infection. Of note, recruitment in the STOIC study was halted early due to treatment efficacy and statistical futility with continuation. Previously, this dataset has proposed an attenuated interferon response with increased CCL24 can be used as a biomarker to predict severe COVID-19 disease¹². Notably, Baker et al. also demonstrated an early Th2-like (including IL-4 and CCL11) response comparable to nasal signatures seen in rhinovirus infection¹³ and defined inflammatory networks induced by SARS-CoV-2¹². These inflammatory networks offer an intriguing opportunity to elucidate key immune responses for disease intervention.

In this present study, we aimed to understand the contribution of inflammatory mediator expression networks on clinical outcomes in participants with high (viral^HIGH) or low (viral^LOW) SARS-CoV-2 burden. From an unbiased mathematical model that cluster mediators, we aimed to better capture immune complexity and avoid overinterpretation of singular pathways that may be overlooked by other approaches such as pathway analysis or gene ontology. Of four nodes present at enrolment (day 0), we showed mucosal-derived mediators including CCL13, CCL17, IL-33, IL-5 and IL-4 were associated quicker self-reported recovery in both usual care and budesonide treatment arms. Our work shows that immune networks offer insight into mechanisms of viral clearance and a framework for future studies of inflammatory diseases.

Viral burden is associated with differential immune node expression.

To begin, we stratified STOIC participants with SARS-CoV-2 infection and healthy volunteers based on four previously defined proteomic mediator nodes at day 0 enrolment¹². We labelled these nodes based on the predominant inflammatory pattern as, interferon (IFN: IFN-α2α and IFN-β), innate immunity-like (innate: TSLP, IFN-γ, VEGF, IL-1β, IL-10, IL-6, and CCL24), chemokine dominant (chemokine: CCL2, CCL3, CCL4, CCL11, CXCL8, CXCL10, CXCL11 and TNFα) and mucosal immunity-like (mucosal: CCL13, CCL17, CCL26, IL-2, IL-4, IL-5, IL-12p70, IL-33 and GM-CSF) (Fig. 1A). To account for the kinetics of viral acquisition and initiation of inflammatory processes across the cohort, we also stratified participants based on an arbitrary cycle threshold (Ct) value greater than or less than 30 (viral^LOW and viral^HIGH respectively) at time of recruitment (Sup. Figure 1A-B). Viral copy number in viral^HIGH participants equalled that of viral^LOW participants by day 7 (Fig. 1B). Increased viral copy number at day 0 was associated with a 0.6-day longer symptom onset in viral^HIGH participants. Other documented characteristics were equivalent between viral^HIGH and viral^LOW participants (Table 1).

Table 1

**Subject characteristics in stratified viral**^HIGH **or viral**^LOW participants. *P = < 0.05 between health and both viral^HIGH or viral^LOW. *¹ P = < 0.05 between viral^HIGH or viral^LOW. Ordinary one-way ANOVA with Tukey’s multiple comparisons. Chi-squared test for proportional data. SD – standard deviation.
	High Viral Burden		Low Viral Burden		Healthy Control
	Mean	SD	Mean	SD	Mean	SD
n	60		79		22
Age	44.8	13.4	45.2	13.2	36.5*	11.5
Sex (% Male)	43.5		40.8		40.9
Ethnicity (% White Caucasian)	85.5		97.4		63.6
Body Mass Index	26.7	4.6	26.2	5.1
Smoking Status (% Never Smokers)	64.5		61.8		81.8
Smoking Pack Years	16.4	15.6	14.1	13.4
Clinical recovery or primary outcome (Days)	9.5	6.1	10.0	7.0
Primary outcome (n)	4.0		5.0
Symptom onset (Days)	3.6	1.7	3.0*¹	1.6
Day 0 CT Value	21.7	3.8	37.5*¹	3.1

Following immune and viral burden stratification, we examined the expression of each node at day 0 and day 14 independently to health (control participants were younger than SARS-CoV-2 infected individuals; Table 1). Virus^HIGH was associated with an elevated IFN node expression compared to viral^LOW and health at day 0 (Fig. 1C). Moreover, the chemokine node was increased independent of viral burden compared to health, with viral^HIGH participants expressing the greatest quantity of chemokine node at day 0. Mucosal node expression was increased at day 0 in viral^HIGH versus viral^LOW participants. At day 14 only participants with viral^LOW had increased IFN compared to health. We observed no difference in innate node expression between groups at either timepoint. As these nodes form part of a complex inflammatory environment and are not in observed in isolation, we performed hierarchical clustering which showed no clear grouping of participants based on node expression at say 0 or 14 (Fig. 1D; Sup. Figure 1C-D). These data support concepts of robust initial IFN and chemokine response to infection^14,15 and elevated mucosal node expression in viral^HIGH participants.

Budesonide and usual care impact node expression.

The primary aim of the STOIC study was to compare inhaled budesonide, a corticosteroid, versus usual care in early SARS CoV-2 infection and assess clinical outcome¹⁰. We thus investigated the impact of budesonide or usual care (anti-pyrectics) on viral burden and node expression in viral^LOW and viral^HIGH participants. Based on a random allocation to treatment, 54.4% of viral^LOW and 45.0% viral^HIGH participants were given budesonide. Viral copy number in both usual care and budesonide decreased over 14 days in viral^HIGH participants, while no change was observed in viral^LOW participants (Fig. 2A). Viral copy number was equivalent at each timepoint independent of treatment (Sup. Figures 2A-B). Budesonide demonstrated a 4.6-day accelerated recovery in viral^LOW participants and no recorded primary outcome events irrespective of viral burden (Table 2).

Table 2

**Budesonide was associated with improved recovery in viral**^LOW **participants.** Subject characteristics stratified based on viral burden and treatment arm. *P = < 0.05 usual care versus budesonide. *¹P = < 0.05 usual care; viral^HIGH versus viral^LOW. *²P = < 0.05 budesonide; viral^HIGH or viral^LOW. Unpaired t test with Welch’s correction SD – standard deviation.
	High Viral Burden				Low Viral Burden
	Usual Care		Budesonide		Usual Care		Budesonide
	Mean	SD	Mean	SD	Mean	SD	Mean	SD
n	33		27		36		43
Age	45.5	13.4	44.1	13.6	45.5	14.1	44.8	12.6
Sex (% Male)	48.5		40.7		25.6		46.5
Ethnicity (% White Caucasian)	90.9		85.2		79.1		95.3
Body Mass Index	26.7	4.8	26.6	4.5	25.8	4.3	26.6	5.6
Smoking Status (% Never Smokers)	63.6		70.4		61.1		58.1
Smoking Pack Years	13.0	15.9	21.5	14.6	14.7	12.6	13.7	14.4
Clinical recovery or primary outcome (Days)	10.3	7.5	8.4	3.5	12.5*	8.4	7.9*	4.8
Primary outcome (n)	4.0		0.0		5.0		0.0
Symptom onset (Days)	3.3	1.8	3.9	1.6	2.9	1.6	3.1	1.6
Day 0 CT Value	21.4*¹	3.7	22.0*²	4.0	38.2*¹	2.8	36.9*²	3.4

We next assessed modulation of SARS-CoV-2-induced node expression post treatment at day 14. No difference was observed in the IFN or innate node in virus^LOW participants in either treatment arm, while usual care only was sufficient to decreased chemokine (1.52-fold) and mucosal node (1.10-fold) expression (Fig. 2B). In viral^HIGH participants there was a reduction over time in both treatment arms in the IFN (2.31-fold usual care; 1.78-fold budesonide), chemokine (1.25-fold usual care; 1.29-fold budesonide) and mucosal nodes (3.10-fold usual care; 2.11-fold budesonide) (Fig. 2C). In contrast, budesonide-treated viral^HIGH participants had unaltered innate node expression. Overall, differences in node expression in viral^LOW participants was specific to usual care treatment. While in viral^HIGH participants all nodes decreased over 14 days regardless of treatment except for the innate node in budesonide-treated participants.

Innate node expression was weakly associated with accelerated recovery.

We explored if the persistently heightened innate node expression was associated with improvement with self-reported clinical outcomes in viral^HIGH participants considering the effectiveness of budesonide in recovery time^10,12. Low or high innate node expression was defined as an expression greater than 2 standard deviations above health. 36.4% usual care and 25.9% of budesonide-treated participants were classified as having high innate node expression at day 0. No difference was observed in innate node expression between treatment arms at either timepoint (Sup. Figure 3A) or the number of days of symptom onset before inclusion into the study (Table 3). Viral^HIGH budesonide-treated participants reported 1.7-day accelerated recovery compared to usual care arm participants (Table 3). Symptom severity defined by FLU-PRO questionnaire¹⁶ showed no significant difference between treatment arms regardless of innate node expression in viral^HIGH participants (Fig. 3B; Sup Fig. 3B); however, nasal symptoms started lower at day 0 in viral^LOW, usual care, high innate node participants (Sup Figs. 3C-D). Notably, budesonide treatment in viral^LOW participants was sufficient to accelerate days to recovery in low innate node participants to a comparable number of days to high innate node participants (Table 3). These data suggest that regardless of innate node expression, participants experienced similar symptom severity in either treatment, but high innate node expression was weakly associated with accelerated recovery.

Table 3

Innate node subject characteristics stratified based on treatment and viral burden. *P = < 0.05 low innate node; usual care versus budesonide. All stats within viral burden groups. Ordinary two-way ANOVA with Sidak’s multiple comparisons. Chi-squared test for proportional data. SD – standard deviation.
		Low Innate Node				High Innate Node
		Usual Care		Budesonide		Usual Care		Budesonide
		Mean	SD	Mean	SD	Mean	SD	Mean	SD
High Viral Burden	n	21		20		12		7
	Age	50.0	11.4	45.3	13.5	37.5	13.2	40.6	14.4
	Sex (% Male)	52.4		35.0		41.7		57.1
	Ethnicity (% White Caucasian)	90.5		95.0		91.7		57.1
	Body Mass Index	27.4	4.9	26.6	4.6	25.6	4.5	26.7	4.5
	Smoking Status (% Never Smokers)	47.6		65.0		91.7		85.7
	Smoking Pack Years	13.9	16.3	20.1	15.2	2.5	0.0	31.3	0.0
	Clinical recovery or primary outcome (Days)	11.0	7.8	8.4	3.8	9.3	7.2	8.5	2.6
	Primary outcome (n)	3.0		0.0		1.0		0.0
	Symptom onset (Days)	2.9	1.7	4.0	1.6	4.0	1.9	3.4	1.5
	Day 0 CT Value	21.3	3.5	22.4	4.1	21.6	4.2	21.1	3.9
Low Viral Burden	n	29		28		7		14
	Age	43.2	12.9	44.3	13.5	55.3	15.8	45.9	10.9
	Sex (% Male)	35.0		55.2		14.3		28.6
	Ethnicity (% White Caucasian)	96.6		96.6		85.7		92.9
	Body Mass Index	25.3	3.7	26.8	5.6	27.9	6.3	26.2	5.8
	Smoking Status (% Never Smokers)	62.1		55.2		57.1		64.3
	Smoking Pack Years	15.6	13.5	17.7	15.1	11.6	10.1	3.2	3.4
	Clinical recovery or primary outcome (Days)	13.5*	9.0	8.3*	4.0	8.7	3.5	7.1	6.1
	Primary outcome (n)	5.0		0.0		0.0		0.0
	Symptom onset (Days)	2.7	1.5	3.2	1.4	3.7	2.1	3.0	1.9
	Day 0 CT Value	38.2	2.8	36.2	3.5	38.5	2.8	38.1	2.9

To further explore the role of high innate node expression on symptom severity, we assessed high innate node irrespective of treatment arm. 31.7% of viral^HIGH and 26.9% of viral^LOW participants had high innate node expression. In viral^HIGH participants, high innate node was associated with accelerated symptom recovery post day 7 (Table 3; Sup. Figure 3E) but associated with increased nasal symptom severity at day 0 and 1 (Fig. 3B upper panel). A trend of fewer nasal, throat and body symptoms were observed in virus^LOW high innate note expressing participants (Fig. 3B lower panel). These data suggest that high innate node is associated weakly with accelerated recovery which is replicated by budesonide in low innate node expressing participants.

Mucosal node is associated with improved self-reported recovery and no primary outcome events.

The increase of the IFN and chemokine nodes at day 0 dependant on viral burden replicates the consensus immune response critical to virus defence^14,15. Notably, primary outcome events were observed in participants with both high and low IFN/chemokine node expression (Sup. Tables 1–2). Strikingly, high mucosal node expression was associated with no primary outcome events and accelerated recovery overall by 3.7-days (viral^LOW 2.5-days; viral^HIGH 4.8-days averaged between treatments respectively) (Table 4). Symptom onset before enrolment into the study was consistent across all groups Specifically, in viral^LOW usual care participants with high mucosal node had 7.9-day quicker symptom resolution. To note, high mucosal node expression was typically observed in younger participants in both viral burden arms (viral^LOW 12.5; viral^HIGH 16.3 years younger, respectively) (Table 4). To evaluate symptom severity, and mucosal node expression, we analysed our data combining usual care and budesonide as no difference in was observed between treatments (Fig. 4A).

Table 4

**Mucosal node was associated with improved recovery and no primary outcome events.** Subject characteristics stratified based on mucosal node and viral burden. *P = < 0.05 usual care; low versus high mucosal node. *¹ P = < 0.05 compared to usual care low mucosal node. All stats within viral burden groups. Ordinary two-way ANOVA with Sidak’s multiple comparisons. Chi-squared test for proportional data. SD – standard deviation.
		Low Mucosal Node				High Mucosal Node
		Usual Care		Budesonide		Usual Care		Budesonide
		Mean	SD	Mean	SD	Mean	SD	Mean	SD
High Viral Burden	n	24		21		9		6
	Age	49.8*	11.2	47.8	12.4	34.0*	12.5	31.0	9.5
	Sex (% Male)	45.8		33.3		55.6		66.7
	Ethnicity (% White Caucasian)	95.8		100.0		77.8		33.3
	Body Mass Index	27.3	4.2	27.1	4.6	25.2	6.1	24.8	3.9
	Smoking Status (% Never Smokers)	54.2		66.7		88.9		83.3
	Smoking Pack Years	11.6	16.0	18.4	12.5	27.6	0.0	43.6	0.0
	Clinical recovery or primary outcome (Days)	11.8	8.3	8.4	3.5	6.6	1.9	8.6	3.6
	Primary outcome (n)	4.0		0.0		0.0		0.0
	Symptom onset (Days)	3.0	1.9	4.0	1.6	4.1	1.4	3.3	1.4
	Day 0 CT Value	21.4	3.6	22.2	4.1	21.4	4.1	21.5	4.2
Low Viral Burden	n	30		33		6		10
	Age	48.5^*	12.7	46.5	12.7	30.8^*	12.0	39.3	11.1
	Sex (% Male)	23.3		51.5		66.7		30.0
	Ethnicity (% White Caucasian)	96.7		93.9		83.3		100.0
	Body Mass Index	25.6	3.8	26.8	5.7	26.6	6.5	25.9	5.6
	Smoking Status (% Never Smokers)	56.7		57.6		83.3		60.0
	Smoking Pack Years	15.8	12.4	16.8	14.9	1.0	0.0	3.0	4.0
	Clinical recovery or primary outcome (Days)	13.9^*1	8.5	8.3^*1	4.8	6.0^*1	3.0	6.7	4.9
	Primary outcome (n)	5.0		0.0		0.0		0.0
	Symptom onset (Days)	2.8	1.7	3.2	1.7	3.7	1.2	3.1	1.0
	Day 0 CT Value	38.1	2.7	37.4	2.8	38.7	3.2	34.9	4.3

20.2% of viral^LOW and 25.0% of viral^HIGH participants were classified as having high mucosal node expression. The number of viral^LOW participants reporting symptoms were comparable in low and high mucosal node groups (Fig. 4B). In viral^HIGH participants high mucosal node expression was associated with fewer reported symptoms from day 8 onwards. High mucosal node was associated with increased nasal and throat symptom severity at day 0 in viral^HIGH participants; however, no other severity scores were different (Fig. 4C). Overall, these data suggest while symptom severity is generally comparable in either low or high mucosal node participants, high node expression was also associated with accelerated recovery and no primary outcome.

The mucosal node forms two daughter networks by day 14.

Inflammatory pathways are dynamic with multiple feedback loops influencing mediator expression. We determined the local weighted degree of connectivity between mediators in the mucosal node at day 0 and 14. At day 0, CCL13 (1.14), CCL17 (1.12) and IL-33 (0.90) had the greatest weighted degree and thus the most locally connected mediators in the mucosal node (Fig. 5 middle panel; Sup. Table 3). By day 14 the mucosal node separated into two daughter nodes consisting of networks of CC17, IL-2, GM-CSF and IL-33 (Fig. 5 left panel; Sup. Table 3) or CCL26, CCL13 and IL4 (Fig. 5 right panels; Sup. Table 3.). Forming part of these day 14 daughter nodes other immune mediators became linked within this network (Sup. Table 3). Notably, daughter node 1 included IFN-β and IFN-γ in both budesonide and usual care subjects (Fig. 5 left panel; Sup. Table 3). Moreover, IFN-α was linked with daughter node 1 in budesonide while linked with daughter node 2 in usual care participants, proposing that IFN-α and other IFN molecules are closely linked to mucosal node mediators (Sup. Table 3). Together, these data propose the mucosal node divides into two daughter networks during the first 14 days of SARS-CoV-2 infection.

Nasal basal, hillock and ciliated cells are the predominant mucosal node expressors.

Lastly, to address the cellular source of the mucosal node, we utilised CELLxGENE¹⁷ and publicly available nasal cavity single-cell RNA sequencing data by Yoshida and Worlock et al.¹⁸. Single cell mRNA expression from 37,513 nasal cavity cells from individuals infected with SARS-CoV-2 was projected as a uniform manifold approximation and projection (UMAP) and genes of interest were overlaid (CCL13, CCL17, IL-33, IL-5, IL-4, CCL26, IL-2, IL-12, and CSF2). The mucosal node was observed predominately across multiple epithelial cell types with minor lymphoid and myeloid cell expression (Fig. 6A-B; Sup Fig. 4A). IL-33 was the most abundant transcript and was predominately expressed in basal, secretory, duct and ciliated cell types in the nose (Fig. 6C). Hillock cells were associated with CCL26 while CCL17 and IL12A were most abundant in ciliated cells. Minimal IL5 was observed in this transcriptional dataset where it was only detected in ciliated cells. Notably, mucosal node encoding transcripts (IL-33, CSF2, IL12A, CCL13, CCL17 and CCL26) were observed in bronchial brushings from matched subjects (Sup. Figure 4B-C). Overall, IL-33, CCL26, CCL13 CCL17 and IL-5 were the predominant contributors to the mucosal node in the upper respiratory tract. These data also offer insight into the broad range of nasal epithelial cells that contribute to the mucosal node.

Our approach focusing on unbiased mediator network analysis demonstrating that elevated upper respiratory tract CCL13, CCL17, IL-33, IL-5, IL-4, CCL26, IL-2, IL-12, and GM-CSF was associated with a 3.7-day quicker recovery and no primary outcome events. While participants reported faster symptom resolution, symptom severity between those who had high or low mucosal node expression was similar except for initial nasal symptoms on days 0 and 1 in viral^HIGH participants. These data also showed time to symptom onset, symptom severities and clinical resolution was similar between participants with high or low viral burden.

The burden of SARS-CoV-2 and symptom presentation do not necessarily correlate and asymptomatic individuals capable of harbouring high viral loads^19,20. We observed that viral^LOW participants had equivalent days to recovery and demonstrated similar symptom severity to those viral^HIGH participants. Despite viral^LOW participants reporting statistically significant 0.6-day later symptom onset viral burden was equal to that of viral^HIGH from day 7 onwards. Moreover, viral^HIGH or viral^LOW participants stratified by node had equivalent days of symptom onset proposing that lower viral burden was not due to differential study enrolment. These viral kinetics for viral^HIGH and viral^LOW copy number and associated symptomology replicate previously studies^21,22. While viral burden did not directly correlate with symptoms, we observed increased IFN, chemokine and mucosal node expression at enrolment in viral^HIGH participants. Thus, we replicated previous findings of heterogenous viral burden and increased anti-viral responses^14,15.

A coordinated immune response composed of cell and mediator interactions is vital for host defence. In this cohort, SARS-CoV-2 infection defined four proteomic clusters which we named interferon, innate immunity-like, chemokine dominant and mucosal immunity-like. Increased mucosal node, which included CCL13, CCL17, IL-33, IL-5, IL-4 and CCL26, contributed to accelerated recovery and no primary outcome events. Previously these mediators have each been associated with viral infections in particular the alarmin, IL-33 (reviewed by Murdaca et al.²³). In COVID-19 specifically, IL-33 has been strongly associated with SARS-CoV-2 infection²⁴ and together with CCL26 has been found in plasma within 10 days of SARS-CoV-2 infection²⁵. Increased IL-4/5¹³, CCL13²⁶ and CCL17²⁷ has been observed in other respiratory infections in patients with atopy. The expression of these mediators supports the role of Th-2-like responses in the upper respiratory tract triggering cells including innate lymphoid cells, B cells and eosinophils to perform early viral control^13,14,28. Connectively analysis also highlights mucosal node’s role in infection with conserved links with IFN-α, -β and -γ by 14 days post enrolment. Cells expressing the transcripts encoding for these mediators encompassed mainly nasal epithelium, supporting a role of this node in barrier defence alongside interferon pathways in an effective antiviral response. Together the role of alarmin IL-33, pleiotropic eosinophil/B cell mediators IL-4/5 in parallel with chemokines CCL26, CCL13 and CCL17 need further investigation in viral infection and protection.

This study provides real world data assessing COVID-19 clinical outcomes through the assessment of inflammatory networks. Approximately 1-in-4 participants were determined to express high levels of the mucosal (viral^LOW − 20.2%; viral^HIGH − 25.0%) and innate (viral^LOW − 26.9%; viral^HIGH − 31.7%) nodes compared to health. These responses, in a predominately white Caucasian population, resulted in 3.7-day accelerated SARS-CoV-2 recovery. However, we observed limited differences in symptom severity between high and low node expressing participants, with differences only observed at days 0 and 1. As participants were enrolled 3 days into symptom onset the worst of symptoms may have passed²² and thus, equally, inflammatory resolution may have begun. Given that this is a community-based study with few participants reaching primary outcome these data may simply reflect a lower symptom severity in this cohort more reflective of common cold symptoms. The longitudinal clinical improvement observed in this cohort also decreased statistical power at later timepoints as participants stopped completing FLU-PRO questionnaires. We therefore restricted our analysis up to the day whereby 25% of subjects still reported symptoms and acknowledge that to observe symptom severity differences between immune node expression a larger cohort is required. Consequently, to understand nodal impact on symptom severity studies focusing on severe hospitalised COVID-19 patients is necessary.

Elderly individuals are at greater risk from SARS-CoV-2 infection due to epigenetic, immunosenescence, and higher basal inflammation processes²⁹, even following vaccination³⁰. Here, we demonstrated that a high mucosal node was typically observed in viral^LOW 12.5 and viral^HIGH 16.3 years younger adult participants. Given that COVID-19 caused more mortality in elderly patients³¹ this observation provides mechanistic insight into this phenomenon. The paradigm is that individuals with higher basal or inducible mucosal node mediators can clear virus quicker and thus avoid the worst symptoms and complications of disease. This basal or inducible response wains as individuals age subsequently predisposing elderly individuals to viral infection. We propose that a robust and coordinated upper respiratory tract mucosal node response is vital for a successful defence against SARS-CoV-2.

Immune interactions include multiple feedback loops to stimulate the immune response and then restore tissue homeostasis^2,3. Consequently, immune networks are dynamic and the ability to control these networks critical for determining if the host will only experience mild, moderate or severe inflammation³². These dynamic processes result in mediators swapping node allegiances based on currently unknown factors. Previously, we have shown that overall these upper respiratory tract immune nodes share 72% homology in usual care but only 27% in budesonide at day 0 and 14, respectively¹². Here, we expand on this analysis and determined that the mucosal node defines two daughter networks independent of treatment at day 14. Moreover, these subsequent networks are linked with interferon pathways supporting an important role in viral defence³³. Based on a shared inflammatory responses to other viruses^13,26,27 the mucosal node and its subsequent networks may be common to multiple viral infections. Frequent sampling is necessary to identify the kinetic switch to these daughter networks during the first 14 days of infection. The dynamics of mediator nodes will offer insight into viral immunopathology and highlight potential immunological mechanisms to hasten viral symptom recovery.

In conclusion, we demonstrated that an upper respiratory tract mucosal response which includes CCL13, CCL17, IL-33, IL-5, IL-4, CCL26, IL-2, IL-12, and GM-CSF was associated with 3.7-day quicker symptom resolution in SARS-CoV-2 infected participants. Elevated mucosal node was also associated with younger adults and no primary outcome events. Together these data offer insight into potential coordinating inflammatory mediators key to the control and swift symptom resolution of SARS-CoV-2. Studies aimed at inducing mucosal node mediators may provide clinical benefit in the accelerated resolution of viral infection.

Acknowledgments

This study was funded by the Oxford NIHR BRC and AstraZeneca (Gothenburg, Sweden). We would like to specifically thank Karolina Krassowska, Beverly Langford, Helen Jeffers, and members of the Oxfordshire primary care network for helping enrol and collect respiratory samples.

Author Contributions

SPC and MB designed the study and wrote the manuscript. JRB, MM, CM and SR were responsible for sample collection and processing. SPC, DVN, RTMN analysed the data. SPC, DVN, JRB, CM, PJB, LED, RTMN, REKR, and MB were responsible for data interpretation and critical revision of the work. PJB, LED, REKR and MB conceived and supervised the study. All authors read and approved the contents of the manuscript.

Competing Interests

SR, LED and MB reports grants from AstraZeneca during the conduct of this study paid to the institute.

Participants and ethics.

This study included 139 participants from the STOIC open-label, parallel-group, phase 2, randomised controlled trial (Fulham London Research Ethics Committee and the National Health Research Authority- 20/HRA/2531, NCT04416399)¹⁰ and 22 healthy controls recruited at the University of Oxford (18/SC/0361). These participants represent a subset of enrolled STOIC participants and inclusion in this study was based on symptom questionnaire and nasal mediator assessment completion.

Study Design.

Healthy controls recruited at the University of Oxford were adults aged 18 years or older without any known history of respiratory disease or COVID-19 symptoms. Healthy controls provided a negative COVID-19 antigen test on the day of nasal mucosa sampling. Nasal sampling was performed using a nasosorption FX.I device (Hunt Developments UK, London, UK). Briefly, the synthetic absorptive matrix (SAM) strip was gently pressed against the inferior turbinate for 1 min. SAM strips were processed in an elution buffer containing 1.2% Triton-X, 1.0% bovine serum albumin (BSA) in phosphate-buffered saline (PBS).

In STOIC, participants aged 18 years or older experiencing COVID-19 symptoms as defined by onset of cough, fever, anosmia, or a combination of these symptoms for less than 7 days were recruited by local primary care networks, local COVID-19 testing sites, and via multichannel advertising. Recruitment was undertaken from July 2020 until January 2021. Participants were randomised into two treatment groups, budesonide, or usual care on day 0. Usual care consisted of supportive therapy and encouragement to take anti-pyrectics for symptoms of fever. Participants randomised to receive budesonide were provided with a dry powder inhaler dose of 400 µg per actuation (two puffs to be taken twice per day: total dose 1600 µg, Pulmicort Turbuhaler, AstraZeneca, Gothenburg, Sweden). Participants allocated to budesonide were asked to halt budesonide treatment when they felt they had recovered (clinical recovery). Primary outcome was defined by an urgent care, emergency department visit or hospitalisation for COVID-19.

Participants recruited as part of STOIC were seen at their homes at day 0, day 7, and day 14 by a trained respiratory research nurse. Self-performed nasopharyngeal swabs for SARS-CoV-2 quantification and daily Influenza Patient-Reported Outcome (FLU-PRO)¹⁶ symptoms scores were collected. At day 0 (pre-randomisation) and day 14 self-performed nasosorption FX.I device (Hunt Developments UK, London, UK) were collected.

SARS-CoV-2 detection and quantification.

Nasopharyngeal swabs were collected from and stored immediately in RNAse solution. RNA was extracted from clinical samples using the QIAamp Viral RNA Mini Kit (Catalogue: 52906, Qiagen) following manufacturer’s instructions. 10 µl of eluted RNA was assayed using the Taqman fast virus 1-step master mix (ThermoFisher Scientific, Loughborough, UK), utilising oligonucleotide primers (600 nM forward and 800 nM reverse per reaction) and fluorescent conjugated probes (two probes 100 nM each) (Eurofins Genomics, Wolverhampton, UK) for the detection of SARS-CoV-2 RNase P gene (RdRP) gene region using the ABI-7500 SDS Instruments (Applied Biosystems). A standard curve was determined using a plasmid expressing a known concentration of the RdRP gene. Sequences for primer and probes for quantitative real-time RT-PCR (RT-qPCR):

Dual Labelled probe P2: CAG GTG GAA CCT CAT CAG GAG ATG C

Dual labelled probe P1: CCA GGT GGW ACR TCA TCM GGT GAT GC

PCR primer RdRP F: GTG ARA TGG TCA TGT GTG GCG G

PCR primer RdRP R: CAR ATG TTA AAS ACA CTA TTA GCA TA

Nasal Immune Mediator Assessment.

Chemokine (C-C motif) ligand (CCL)2, CCL3, CCL4, CCL11, CCL13, CCL17, CCL24, CCL26, chemokine (C-X-C motif) ligand (CXCL)8, CXCL10, CXCL11, granulocyte-macrophage colony-stimulating factor (GM-CSF), interferon (IFN)-α2α, IFN-β, IFN-γ, interleukin (IL)-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12p70, IL-33, thymic stromal lymphopoietin (TSLP), tumour necrosis factor alpha (TNFα) and vascular endothelial growth factor (VEGF) were quantified using MSD Quickplex SQ 120 1300 (MesoScale Diagnostics, Rockville, MD, USA). All values at or below the lower limit of detection or above the upper limit of detection were replaced by the appropriate standard value.

FLU-PRO questionnaire and symptom indexing.

To understand symptoms from a respiratory virus infection the validated FLU-PRO respiratory questionnaire was used for SARS-CoV-2¹⁶. The questionnaire comprised of 32 questions which covered symptom severity in the nose, throat, eyes, chest/respiratory tract, gastrointestinal tract, and body/systemic. Participants scored symptoms from not at all (0) to very much (4).

Single-cell transcriptomic overlay analysis.

Transcripts encoding for mucosal mediators were assessed using publicly available data via the Lung Cell Atlas consortium and CELLxGENE¹⁷. Briefly, Yoshida and Worlock et al.¹⁸ nasal cavity (37,513 cells) or bronchial brushing (11,672 cells) data was used to visualise CCL13, CCL17, IL-33, IL-5, IL-4, CCL26, IL-2, IL-12, and CSF2 (encoding for GM-CSF) transcripts in adults (18 + years) with SARS-CoV-2 infection.

Statistical Analysis.

Networks analysis was performed by first eliminating from the correlations matrix the contribution corresponding to ‘noise’ eigenvalues, i.e., those present in the Marcenko-Pastur distribution. The remaining, corrected matrix can be taxonomized into functional clusters; here, we used the Louvain algorithm, which finds – in these conditions – anti- or un-correlated clusters. All networks considered here showed a moderate degree of modularity, with four functional clusters in each case. All inflammatory mediator data was Log₂ transformed. Transformed data was stratified based on network analysis determined by Baker et al.¹² and averaged per node, per subject (inflammatory score). High node expression was determined by an increased expression greater than 2 standard deviations above health. Kolmogorov-Smirnov test was performed to test for normality. Heatmaps were produced using mean inflammatory score in R using heatmap.2 within gplots. All packages are available at CRAN: https://cran.r-project.org/. We also performed hierarchical clustering with Manhattan distance with Pearson correlation coefficient using R, to detect possible clustering patients based on treatment arm. Qualitative FLU-PRO scores were averaged daily or stratified on questions targeted physiology daily and averaged. Symptom data is presented as a daily mean plus or minus standard error of the mean (SEM) up to a maximum of day 14, or to the day when greater than a quarter of participants still reported symptoms. Statistical analysis was performed with GraphPad Prism v9.1 (GraphPad Software, Inc). A P value of < 0.05 was considered statistically significant.

Merad, M., Blish, C. A., Sallusto, F. & Iwasaki, A. The immunology and immunopathology of COVID-19. Science (80-. ). 375, 1122–1127 (2022).
Rieckmann, J. C. et al. Social network architecture of human immune cells unveiled by quantitative proteomics. Nat. Immunol. 18, 583–593 (2017).
Nish, S. & Medzhitov, R. Host Defense Pathways: Role of Redundancy and Compensation in Infectious Disease Phenotypes. Immunity vol. 34 629–636 (2011).
Wiersinga, W. J., Rhodes, A., Cheng, A. C., Peacock, S. J. & Prescott, H. C. Pathophysiology, Transmission, Diagnosis, and Treatment of Coronavirus Disease 2019 (COVID-19): A Review. JAMA - Journal of the American Medical Association vol. 324 782–793 (2020).
World Health Organization (WHO). WHO Coronavirus (COVID-19) Dashboard. https://covid19.who.int/ (2023).
Baden, L. R. et al. Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine. N. Engl. J. Med. 384, 403–416 (2021).
Polack, F. P. et al. Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine. N. Engl. J. Med. 383, 2603–2615 (2020).
Voysey, M. et al. Single-dose administration and the influence of the timing of the booster dose on immunogenicity and efficacy of ChAdOx1 nCoV-19 (AZD1222) vaccine: a pooled analysis of four randomised trials. Lancet 397, 881–891 (2021).
Kim, J. H., Marks, F. & Clemens, J. D. Looking beyond COVID-19 vaccine phase 3 trials. Nature Medicine vol. 27 205–211 (2021).
Ramakrishnan, S. et al. Inhaled budesonide in the treatment of early COVID-19 (STOIC): a phase 2, open-label, randomised controlled trial. Lancet Respir. Med. 9, 763–772 (2021).
Yu, L. M. et al. Inhaled budesonide for COVID-19 in people at high risk of complications in the community in the UK (PRINCIPLE): a randomised, controlled, open-label, adaptive platform trial. Lancet 398, 843–855 (2021).
Baker, J. R. et al. Early Th2 inflammation in the upper respiratory mucosa as a predictor of severe COVID-19 and modulation by early treatment with inhaled corticosteroids: a mechanistic analysis. Lancet Respir. Med. 10, 545–556 (2022).
Hansel, T. T. et al. A Comprehensive Evaluation of Nasal and Bronchial Cytokines and Chemokines Following Experimental Rhinovirus Infection in Allergic Asthma: Increased Interferons (IFN-γ and IFN-λ) and Type 2 Inflammation (IL-5 and IL-13). EBioMedicine 19, 128–138 (2017).
Schwarze, J. & Mackenzie, K. J. Novel insights into immune and inflammatory responses to respiratory viruses. Thorax 68, 108–110 (2013).
Melchjorsen, J., Sørensen, L. N. & Paludan, S. R. Expression and function of chemokines during viral infections: from molecular mechanisms to in vivo function. J. Leukoc. Biol. 74, 331–343 (2003).
Powers, J. H. et al. Performance of the inFLUenza Patient-Reported Outcome (FLU-PRO) diary in patients with influenza-like illness (ILI). PLoS One 13, 1–15 (2018).
CZ CELLxGENE Discover. Chan Zuckerberg Initative www.cellxgene.cziscience.com/gene-expression.
Yoshida, M. et al. Local and systemic responses to SARS-CoV-2 infection in children and adults. Nature 602, 321–327 (2022).
Argyropoulos, K. V. et al. Association of Initial Viral Load in Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Patients with Outcome and Symptoms. Am. J. Pathol. 190, 1881–1887 (2020).
Ra, S. H. et al. Upper respiratory viral load in asymptomatic individuals and mildly symptomatic patients with SARS-CoV-2 infection. Thorax 76, 61–63 (2021).
Néant, N. et al. Modeling SARS-CoV-2 viral kinetics and association with mortality in hospitalized patients from the French COVID cohort. Proc. Natl. Acad. Sci. 118, 1–11 (2021).
He, X. et al. Temporal dynamics in viral shedding and transmissibility of COVID-19. Nat. Med. 26, 672–675 (2020).
Murdaca, G. et al. Involvement of Il-33 in the Pathogenesis and Prognosis of Major Respiratory Viral Infections: Future Perspectives for Personalized Therapy. Biomedicines vol. 10 (2022).
Stanczak, M. A. et al. IL-33 expression in response to SARS-CoV-2 correlates with seropositivity in COVID-19 convalescent individuals. Nat. Commun. 12, (2021).
Lucas, C. et al. Longitudinal analyses reveal immunological misfiring in severe COVID-19. Nature 584, 463–469 (2020).
Jha, A. et al. Increased nasal mucosal interferon and CCL13 response to a TLR7/8 agonist in asthma and allergic rhinitis. J. Allergy Clin. Immunol. 147, 694-703.e12 (2021).
Williams, T. C. et al. Rhinovirus-induced CCL17 and CCL22 in asthma exacerbations and differential regulation by STAT6. Am. J. Respir. Cell Mol. Biol. 64, 344–356 (2021).
Chang, Y. J. et al. Innate lymphoid cells mediate influenza-induced airway hyper-reactivity independently of adaptive immunity. Nat. Immunol. 12, 631–638 (2011).
Mueller, A. L., Mcnamara, M. S. & Sinclair, D. A. Why does COVID-19 disproportionately affect older people? Aging vol. 12 9959–9981 (2020).
Antonelli, M. et al. Risk factors and disease profile of post-vaccination SARS-CoV-2 infection in UK users of the COVID Symptom Study app: a prospective, community-based, nested, case-control study. Lancet Infect. Dis. 22, 43–55 (2022).
Gray, W. K., Navaratnam, A. V., Day, J., Wendon, J. & Briggs, T. W. R. COVID-19 hospital activity and in-hospital mortality during the first and second waves of the pandemic in England: an observational study. Thorax 77, 1113–1120 (2022).
Fadai, N. T. et al. Infection, inflammation and intervention: Mechanistic modelling of epithelial cells in COVID-19. J. R. Soc. Interface 18, (2021).
Katze, M. G., He, Y. & Gale, M. Viruses and interferon: A fight for supremacy. Nature Reviews Immunology vol. 2 675–687 (2002).

Yes there is potential Competing Interest. SR, LED and MB reports grants from AstraZeneca during the conduct of this study paid to the institute.

SupplementaryFiguresandTables.docx

Download PDF

Journal Publication

published 20 Feb, 2024

Read the published version in ERJ Open Research →

Version 1

posted

You are reading this latest preprint version

Coordinated nasal mucosa-mediated immunity accelerates recovery from COVID-19

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Discussion

Declarations

Methods

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1