Nasal polyp antibody secreting cells display proliferation signature in aspirin-exacerbated respiratory disease

Abstract

Background:

Chronic rhinosinusitis with nasal polyps (CRSwNP) causes nasal obstruction and olfactory dysfunction. Aspirin-exacerbated respiratory disease (AERD) is the triad of CRSwNP, asthma and respiratory reactions to cyclooxygenase-1 inhibitors. Patients with AERD have elevated nasal IL-5 levels and high numbers of antibody secreting cells (ASCs), plasma cells (PCs) and plasmablasts, in their polyp tissue, and nasal polyp IgE levels correlate with disease severity and recurrence of nasal polyposis.

Objective:

We sought to explore differences in the transcriptomic profile, activation markers, and IL-5Rα expression and function of nasal polyp ASCs from patients with AERD and CRSwNP.

Methods:

Nasal polyp tissue was collected from patients with AERD and CRSwNP and digested into single-cell suspensions. Nasal polyp cells were analyzed for protein expression by mass cytometry. For IL-5Rα functional studies, plasma cells were purified and cultured in vitro with/without IL-5 and analyzed by bulk RNA-sequencing.

Results:

AERD polyp tissue contained significantly more ASCs and had increased ASC expression of IL-5Rα compared to CRSwNP. AERD ASCs expressed higher protein levels of B cell activation and regulatory markers (CD40, CD19, CD32, and CD38) and the proliferation marker Ki-67. AERD ASCs also expressed more IL5RA, IGHE, and cell cycle and proliferation-related transcripts (CCND2, MKI67, CDC25A and CDC25B) compared to CRSwNP. Stimulation of AERD PCs with IL-5 induced key cell cycle genes (CCND2 and PTP4A3), whereas IL-5 stimulation of ASCs from CRSwNP induced few transcriptomic changes.

Conclusion:

Nasal polyp tissue ASCs from patients with AERD express higher levels of functional IL-5Rα and markers associated with cell cycling and proliferation compared to those from aspirin-tolerant CRSwNP patients.

Capsule Summary:

Nasal polyp tissue antibody secreting cells from patients with aspirin-exacerbated respiratory disease have increased expression of markers associated with cell cycle and growth compared to tissue antibody secreting cells from patients with aspirin-tolerant chronic rhinosinusitis with nasal polyps.

Introduction:

Chronic rhinosinusitis with nasal polyps (CRSwNP) is an inflammatory disorder marked by outgrowths of sinonasal mucosa leading to nasal blockage, olfactory dysfunction, and subsequent impairments in quality of life. Aspirin-exacerbated respiratory disease (AERD) is a severe phenotype of CRSwNP associated with co-morbid asthma and respiratory reactions to cyclooxygenase-1 inhibitors.¹ In patients with AERD and CRSwNP, antibody-secreting cells, including B cells, plasma cells (PCs) and plasmablasts, are present in nasal polyp tissue (NP) and generate antibodies locally.² Mechanisms by which NP antibodies have been hypothesized to contribute to CRSwNP disease severity include IgE-induced activation of NP mast cells or basophils,³ IgA-enhanced eosinophil survival,⁴ and IgG-directed local complement activation.⁵ In AERD specifically, NP IgE levels correlate with CRSwNP severity and disease recurrence,^{2, 6} and a role for IgE to staphylococcal enterotoxins has been proposed.⁷

Monoclonal antibodies targeting IL-5 or IL-5Rα are approved for the treatment of eosinophilic asthma and CRSwNP, and have efficacy in patients with AERD.⁸ While the efficacy of anti-IL-5 biologics for CRSwNP has generally been attributed to downstream impact on eosinophils, treatment of CRSwNP with dexpramipexole, which induced near-complete depletion of eosinophils from the peripheral blood and NP tissue in patients with severe CRSwNP, did not lead to significant improvement in CRSwNP symptoms or NP burden,⁹ suggesting that inhibition of eosinophils is unlikely to be the only mechanism by which anti-IL-5 drugs provide clinical efficacy.

We previously described a population of NP antibody secreting cells (ASC) that express both IL5RA and IGHE transcripts, and are more frequent in tissue from patients with AERD compared to aspirin-tolerant CRSwNP.² Additionally, we found that NP mast cells and ciliated epithelial cells express IL5RA transcript.¹⁰ Therefore, we suspect that the benefit of IL-5 inhibition for patients with CRSwNP may be due to the blockade of IL-5/IL-5Rα signaling on non-eosinophil upper airway cells including ASCs. In this study, we use Mass Cytometry (CyTOF) and RNA-sequencing of NP tissue from subjects with aspirin-tolerant CRSwNP and with AERD to further characterize the NP ASC compartment, with a specific focus on the IL-5Rα⁺ cell population. Our findings have implications for the potential mechanisms of ASC-mediated upper airway inflammation as well as mechanisms of benefit of drugs targeting the IL-5/IL-5Rα signaling pathway.

Results and Discussion:

There were no significant differences in age, sex, absolute eosinophil count, and lifetime number of sinus surgeries between subjects with CRSwNP (n=14) and AERD (n=12). A greater proportion of patients with AERD had co-morbid asthma compared to those with CRSwNP (Table E1). All patients with AERD had physician-diagnosed asthma, and their AERD diagnosis had been confirmed with a physician-observed oral aspirin challenge.

Since our initial description of IL-5Rα⁺ ASCs in patients with AERD², we have further characterized this activated population of ASCs with bulk RNA-sequencing of NP ASCs, CyTOF, and in vitro IL-5 stimulation experiments. In contrast to murine B cells, information on the expression of IL-5Rα in human B cells is limited, but re-analysis of existing transcriptomic data shows induction of IL5RA expression during the differentiation of B cells into PCs (Fig E1).^{11, 12}

Consistent with our prior findings with flow cytometric analysis,² in this study we used CyTOF in a distinct cohort of patients and found that there are more ASCs in the NP tissue from participants with AERD compared to aspirin-tolerant CRSwNP (15.3±3.3% vs 8.975±5.9% of CD45⁺ cells, P < 0.011, Fig 1A) and IL-5Rα protein expression is elevated on the NP ASCs from patients with AERD compared to those from aspirin-tolerant CRSwNP (P < 0.001, Fig 1B). Using a panel of 42 antibodies designed to assess canonical cell type markers and B cell development and activation (Table E2), we identified multiple immune cell populations (Fig E2 A) including ASCs. ASCs were identified as CD56⁻/CD20⁻/CD38^+/CD19⁺/CD27⁺ cells (Fig E2 B). Cells from all donors were cryopreserved for batched CyTOF, which led to a significant reduction of expression levels of the PC marker, CD138, consistent with previously reported findings.¹³ Thus CD138 was not included for PC identification in our study. When we used CyTOF to evaluate the expression of activation markers on ASCs from donors with AERD and CRSwNP, AERD ASCs displayed significantly greater levels of CD19, CD40, CD38, and CD27, which have previously been described as markers of activated IgE-producing ASCs by Horst et al. (Fig 1C).¹⁴ In another recent study, expression of Ki-67, a marker of cellular proliferation, was reported at higher levels in IgE⁺ antibody-secreting cells compared to IgD⁺ naïve B cells.¹⁵ We also observed that the NP ASCs from both AERD and CRSwNP express Ki-67, but in tissue from patients with AERD we observed a significantly higher proportion of Ki-67⁺ cells (P < 0.001, Fig 1D). Additionally, we found that EBI2 (GPR183), a marker of B cell activation within NPs that is important for B cell migration and the induction of extrafollicular ASCs within secondary lymphoid tissue,¹⁶ is higher in AERD ASCs compared to aspirin-tolerant CRSwNP (Fig 1C). When IL-5Rα+ ASCs were analyzed separately as a subpopulation to better profile their phenotype, it was found that IgE protein levels were greater in IL-5Rα+ PCs from patients with AERD compared to the same ASC subpopulation in CRSwNP (Fig 1E).

Figure 1:

Protein markers quantification by CyTOF. Number of antibody secreting cells (ASCs) relative to total CD45⁺ cells in nasal polyps of AERD and CRSwNP (A). Violin plots show expression of IL-5Rα protein (B) on nasal polyp ASCs, with each dot presenting one plasma cell. Differential expression of proteins from CyTOF panel between AERD ASCs and CRSwNP ASCs (C). Violin plots show Ki-67 protein expression (D). Differential expression analysis was performed on IL-5Rα⁺ ASCs from patients with AERD and CRSwNP (E).

Bulk RNA-sequencing (RNA-seq) of purified NP ASCs from a cohort of patients independent from those included in the CyTOF analyses was performed to characterize the AERD and CRSwNP ASCs. A principal component analysis showed tendencies toward global differences in the transcriptome of ASCs populations from patients with AERD compared to aspirin-tolerant CRSwNP, but with no clear separation (Fig 2A). Differential expression analysis between AERD and CRSwNP returned a total of 258 significantly differentially expressed genes (DEGs) (p-adjusted < 0.1, Fig 2B). Among these DEGs, 84 genes were more highly expressed in AERD, whereas 174 genes were more highly expressed in CRSwNP. In ASCs from donors with AERD, IL5RA was one of the top differentially expressed genes along with CCND2 and IGHE, which are transcripts that were previously found to be differentially expressed in NP ASCs.^2,6 Gene enrichment analysis performed on DEGs shows several cell cycle and cell growth-related pathways were enriched in the purified ASCs from patients with AERD (Fig 2C). Top enriched pathways were MYC targets and cell cycle progression pathways (E2F targets, G2M_checkpoint, and mitotic spindle assembly),¹⁷ suggesting a strong proliferation transcriptomic signature in ASCs from participants with AERD. The Ki-67 protein transcript MKI67 was differentially expressed by AERD ASCs compared to CRSwNP ASCs, confirming our CyTOF findings and demonstrating the increased proliferation in AERD ASCs(Fig 2D and 1C). Interestingly, a recent transcriptional analysis study of human IgE-expressing PCs also suggests that they have a high proliferation and cell cycling profile.¹⁸ CCND2 expression activates cell cycle progression and enhances cell proliferation.¹⁹ Moreover, cell division cycle-25 family genes (CDC25A and CDC25B), which are key regulator genes of cell cycle progression and cell survival, were also expressed at significantly higher levels in AERD ASCs (Table E3).²⁰ In addition to increased cell cycle and oxidation-related transcripts, there were also PC homeostasis and regulatory genes noted to be upregulated in AERD ASCs, including BACH2, CXCR4, and LGALS1 (Table E3).^{15, 21, 22}

Figure 2:

Bulk RNA-sequencing of purified nasal polyp antibody secreting cells (ASCs). (A) Principal component analysis show global differences between the AERD and CRSwNP ASCs transcriptome, (B) Volcano plot of differentially expressed genes in nasal polyp ASCs in AERD compared aspirin-tolerant CRSwNP, (C) Hallmark Geneset enrichment analysis performed on pre-ranked genes from differential expression analysis, (D) Heatmap showing expression of cell cycle and proliferation related genes (expression scaled on rows).

In CRSwNP ASCs, estrogen response and tumor necrosis factor-alpha (TNF-α) signaling via NF-κB pathways were enriched compared to ASCs from AERD NP tissue, likely due to a higher expression of several NF-κB target genes (EGR1, EGR2, EGR3, FOS, FOSB, JUN, JUNB, and NFKBIA). NF-κB plays a critical role in balancing cell survival and cell death.²³ Interestingly, both pro-survival (BCL2, BNIP2) and pro-apoptotic genes (BCL10, JUN) were differentially expressed in ASCs from CRSwNP. This was confirmed at the protein level by CyTOF with higher expression of BCL2 protein noted on CRSwNP ASCs compared to AERD ASCs, which was consistent with previous findings by Zhang and colleagues.²⁴

Next, we conducted differential analyses of the transcriptome in ASCs from AERD (re-analysis of our previously reported data),² and CRSwNP (new cohort) after IL-5 stimulation (Fig 3 A & B). P-adjusted values and log-fold change differences upon IL-5 stimulation show that IL-5 induced changes in a greater number of transcripts in AERD ASCs than it did in CRSwNP ASCs (63 vs 21) suggesting that the IL-5Rα in ASCs from patients with AERD is either more highly expressed and/or more functionally active. In particular, IL-5 stimulation of ASCs from AERD NP tissue induced a number of cell cycle and growth-related genes. Specifically, CCND2, CD44 and PTP4A3 were upregulated in vitro by IL-5 stimulation (Fig 3 A) and were also expressed in vivo at significantly higher levels in AERD ASCs compared to CRSwNP (Fig 2B), further confirming that ASCs within the NP of AERD patients have functional IL-5Rα and respond to IL-5 signaling. However, IL-5 stimulation led to fewer DEGs in the PCs from CRSwNP patients (Fig 3 B), which we hypothesize may be due to the fact that there are fewer IL-5Rα⁺ ASCs in patients with aspirin-tolerant CRSwNP compared to AERD, or possibly that ASCs from patients with aspirin-tolerant CRSwNP have a less functionally active IL-5Rα.

Figure 3:

Bulk RNA-sequencing of antibody secreting cells treated with IL-5. (A) Differential expression between IL-5 stimulated ASC from tissue from patients with AERD (partially published data²). (B) Differential expression between IL-5 stimulated ASC from tissue from aspirin-tolerant patients with CRSwNP.

Our study demonstrates the importance of IL-5 as a stimulus driving ASC activation and cycling in a type 2 inflammatory environment using tissue from patients with a severe phenotype of upper airway inflammation. We have found that IL5RA-expressing ASCs in patients with AERD are functionally more active and produce higher levels of IgE. We have shown previously that higher levels of tissue IgE in nasal polyps lead to faster growth and a greater likelihood of recurrence after surgery.² Our transcriptomic data also indicates that ASCs from patients with AERD contain programming suggestive of high rates of cell proliferation including induction of Myc targets, cell cycle progression pathways, and cell survival pathways, possibly induced by higher levels of IL-5 protein in the NP microenvironment.²⁵ IL-5/IL-5Rα signaling induces cell cycle and proliferation transcripts in ASCs, as established by in vitro stimulation experiments. We conclude that, in addition to promoting eosinophil differentiation and survival, IL-5 may play a role in propagating ASC-mediated inflammation^3-5 in the diseased upper airway tissue of AERD.

Key Messages:

• Nasal polyp tissue antibody secreting cells from patients with AERD have increased expression of a functional IL-5Rα compared to patients with aspirin-tolerant CRSwNP.

• The nasal polyp antibody secreting cells from patients with AERD have increased expression of proteins and transcripts associated with cell cycle and growth-related pathways.

Abbreviations used:

AERD

aspirin-exacerbated respiratory disease

ASC

antibody secreting cells

CRSwNP

Chronic rhinosinusitis with nasal polyps

CyTOF

Mass cytometry

DEGs

Differentially expressed genes

IL

Interleukin

PCs

Plasma cells