Integrin β6 microparticles in nasal lavage fluids; potential new biomarkers for basal cell activation in chronic rhinosinusitis

To the editor

Microparticles (MP) are extracellular vesicles that are released by shedding from the plasma membrane of activated cells. Different from exosome, another type of extracellular vesicles which are released through exocytosis of multivesicular bodies, typically MP share the same membrane markers and intracellular proteins with their parent cells. We previously reported the successful use of MP analysis in nasal lavage fluids (NLFs) as a biomarker to evaluate activation of structural and immune cells in subjects with chronic rhinosinusitis (CRS) and aspirin-exacerbated respiratory disease (AERD)¹. Abnormal epithelial repair characterized by basal cell hyperplasia was recently reported², with evidence of epithelial-mesenchymal transition (EMT) in hyperplastic basal cells in nasal polyps³, suggesting that epithelial repair fails due to abnormal basal cell function or differentiation, accompanied by aberrant EMT in severe phenotypes of CRS with polyps (CRSwNP) with high risk of polyp recurrence. We report here that elevated levels of the putative biomarkers of activated basal cells undergoing EMT, MP expressing both the basal cell marker, integrin β6 (ITGB6), and canonical EMT/mesenchymal markers, Snail (SNAI 1) or alpha-smooth muscle actin (ACTA2- αSMA) in NLF are associated with severe polyp phenotype which had previously necessitated revision surgery.

We enrolled 196 CRSwNP, 70 CRS without polyp (CRSsNP) and 31 AERD cases who had endoscopic sinus surgery (ESS) at Northwestern Memorial Hospital between January 2014 and December 2017 (Table 1). We also enrolled 47 control cases who underwent surgery other than ESS without sinus disease during the same period. NLFs were collected by spraying 8 mL of phosphate buffered saline toward the middle meatus, or polyps arising from the middle meatus and their surrounding mucosa, under direct visual guidance at the time of surgery, and MP were measured using a BD FACS LSRII flow cytometer (BD Biosciences, Erembodegem, Belgium) as previously described (Fig.E1A,B)^1,4. Three types of epithelial MP were defined according to the expression of epithelial specific antigens: EpCAM(+)MP, E-cadherin(+)MP and ITGB6(+)MP; we also tested the EMT markers, Snail, Slug (SNAI 2), and αSMA. After staining the MP surface with an epithelial antibody, permeabilization of MP was performed to allow intravesicular staining. We compared MP assay results with and without permeabilization to see localization of EMT/mesenchymal markers on the MPs (surface or inside of MP) (Fig.E1C), and decided to use the MP assay with permeabilization to detect αSMA but not Snail or Slug on epithelial MP.

Table 1.

Demographics of the cases

Characteristics	Control (n=47)	CRSsNP (n=70)	CRSwNP (n=196)	AERD (n=31)	p value
Gender, male/female	31/16	38/32	129/67	15/16	.13
Ethnicity, White/ non-White	39/8	54/16	150/46	20/11	.33
Age, mean± Standard deviation	44.8±15.6	46.4±15.1	47.5±13.1	47.5±13.4	.63
Smoking status, never/prior /current smokers	34/12/1	51/16/3	133/53/10	19/12/0	.43
Comorbidity
Asthma, yes (%)	-	17 (24%)	83 (42%)*	31 (100%)*#	<.0001
Atopy, yes (%)	-1)	40 (57%)	131 (67%)	22 (71%)	.27
Steroid use, yes (%)
Oral	-	2 (3%)	6 (3%)	6 (19%)*#	<.006
Intranasal	-	21 (30%)	52 (27%)	10 (32%)	.73
Inhaled	-	4 (6%)	40 (20%)*	13 (42%)*#	<.0001
Preoperative oral steroid use	-	1 (1%)	29 (15%)*	11 (35%)*#	<.0001

Cases with autoimmune diseases, immunodeficiency, systemic coagulation disorders, cancer, metabolic diseases, cystic fibrosis, an antrochoanal polyp, or unilateral nasal polyp were excluded from the study. Control subjects had no comorbid asthma and no steroid therapy.

¹⁾Thirteen out of 46 control cases had history of atopy, but none of them was on medication for atopy at the time of enrollment. The chi-squared test or Kruskal-Wallis test (Age) was used to compare demographics among the 4 groups.

^*p< 0.05, vs. CRSsNP,

^#p< 0.05, vs. CRSwNP

The increase of epithelial MP subtypes varied according to phenotype of CRS (Fig.1A). Levels of E-cadherin(+)MP and EpCAM(+)MP were higher in CRSsNP and AERD than in CRSwNP, whereas ITGB6(+)MP were higher in CRSwNP and AERD than in CRSsNP. EpCAM and E-cadherin are constitutively expressed on epithelial cells and down-regulated during EMT³. The release of EpCAM(+)MP and E-cadherin(+)MP may thus reflect epithelial injury more than repair and EMT¹. In contrast, ITGB6 is rarely expressed on normal epithelial cells and is up-regulated on basal cells during epithelial repair^5,6. Thus, ITGB6(+)MP release may better reflect basal cell activation during repair. Our findings suggest that basal cells are more highly activated during epithelial repair in the nasal mucosa of patients with polypoid CRS than in patients with non-polypoid CRS.

Figure 1. Comparisons of MP levels.

(A) Comparison of epithelial MP among control, CRSsNP, CRSwNP and AERD.

(B-D) Comparison of ITGB6(+)MP subtypes (B) among control, CRSsNP, CRSwNP and AERD, (C) CRSwNP with history of prior surgery (cases undergoing surgery to treat polyp recurrence) vs. without prior surgery (primary surgical cases), (D) CRSwNP with asthma vs. without asthma (including those with history of prior surgery and those without prior surgery, as indicated). The Steel-Dwass test was used to analyze MP levels between each two groups.

*; p<.05, **; p<.01, ***; p<.001, ****; p<.0001, ##-fold; significantly higher by ##-fold vs control.

Epithelial repair has been reported to occur actively in parallel with inflammation starting at an early stage of polyp formation⁷. Among CRSwNP cases, high ITGB6(+)MP levels were associated with those with the most severe phenotype, i.e. revision surgery, independent of asthma, atopy and steroid use (Fig.1C, Table E2). Conversely, ITGB6(+)MP were not increased compared with controls in CRSwNP cases with no history of prior surgery despite the need for surgery (Fig.1C). We speculated that ITGB6(+)MP may be released from the nasal epithelium on the polyp perimeter (where acanthosis is more frequently observed)⁷ rather than from the epithelium spanning the mature polyp itself (which is often either acanthotic or consists of only a simple layer of basal cells)^3,7. Our findings suggest that basal cells may be more highly activated in nasal epithelium of patients with recurrent CRSwNP.

We examined commonly used EMT markers to test level of expression on ITGB6(+)MP to evaluate the phenotype of basal cells undergoing EMT and/or epithelial repair in CRSwNP. All three MP subtypes expressing ITGB6 along with EMT markers were significantly higher in CRSwNP cases with history of prior surgery than those without; especially Snail(+)ITGB6(+)MP and αSMA(+)ITGB6(+)MP (Fig.1C). In vitro studies confirmed that these ITGB6(+)MP subtypes were all consistently released from A549 cells along with up-regulation of ITGB6 during TGF-β-induced EMT (Fig.E2). Because myoepithelial cells have not been identified in nasal mucosa of CRS, these results may indicate aberrant or excessive EMT and basal cell activation during repair in this subset of subjects with particularly severe disease. While snail, slug, and/or αSMA positive ITGB6(+)MPs did overlap significantly, they were not fully equivalent, implying that they may not control or be derived from identical signaling pathways.

We classified CRSwNP cases into 4 groups according to asthma comorbidity and history of prior endoscopic sinus surgery (Table E3) and compared levels of ITGB6(+)MP subtypes. CRSwNP subjects with comorbid asthma have a higher recurrence rate after surgery than those without asthma and are regarded as a severe phenotype with high type 2 inflammation⁸. However, dual Snail(+)ITGB6(+)MP and αSMA(+)ITGB6(+)MP were elevated to the highest levels in patients with AERD (Fig.1B) and in patients with CRSwNP with history of prior surgery regardless of whether there was comorbid asthma or not (Fig.1D), and correlations between levels of the eosinophil cation protein (EPC) and levels of Snail(+)ITGB6(+)MP (r_s=.20, p<.06) or αSMA(+)ITGB6(+)MP (r_s=.27, p<.02) were weak. Assessment of epithelial repair using ITGB6(+)MP subtypes therefore appears not to represent a simple surrogate of type 2 inflammation. Biopsy tissue-based studies reported that higher counts of eosinophils, mast cells and basophils, and high levels of type 2 cytokines including EPC, were detected in mature polyps, rather than in tissues around polyps such as uncinate/ethmoid tissues^8,9. As mentioned above, the perimeter of polyps appears to have the highest level of acanthosis⁷; whether ITGB6(+)MP originate from nasal epithelium on the perimeter of nasal polyps is worthy of further study. Because the MP assay in NLFs is representative of the condition in the entire nasal cavity, it would be preferable to collect MPs released from nasal tissues selected for biopsy using a sponge-based approach in order to directly compare histological features in specific regions of the sinonasal cavity to the MP profiles.

There are limitations to this study. NLFs were only collected from surgical patients in the operating room before surgery and we were not able to follow their levels with subsequent measurements after surgery. The degree of basal cell activation and EMT might be elevated in response to prolonged chronic inflammation in severe CRSwNP endotypes with prior surgery. There was no significant difference in levels of ITGB6(+)MP subtypes between CRSsNP with and without prior surgery (Fig.E3). In addition, among 196 CRSwNP cases enrolled in the study, with the exception of one subject who had had a revision surgery 1.7 year after the prior surgery, no CRSwNP subjects had undergone surgery previously between January 2014 and December 2017. Therefore, the surgical procedure itself used during prior surgery may not directly influence ITGB6(+)MP levels in cases undergoing revision surgery.

In conclusion and based on the results of our MP assays, basal cells are significantly activated with up-regulation of ITGB6 and markers of EMT in patients with AERD or a severe phenotype of CRSwNP which had previously necessitated revision surgery. Biopsy of nasal tissues is invasive and not generally practical in the clinic. Basal cells undergoing EMT are unevenly distributed in tissue and are difficult to quantify. While it is generally accepted that CRSwNP is mostly characterized by type 2 inflammation, inflammatory profiles are highly variable and polyp recurrence necessitating revision surgery occurs in diverse inflammatory endotypes, with or without comorbid asthma. We believe that ITGB6(+)MP subtypes in NLFs may help to clarify roles of basal cells in pathophysiology, provide new insights for CRS endotyping and play valuable roles in testing the link between status of basal cells and the natural history, comorbidities and long term outcomes of CRS.