It is well established in the published literature that compromise of epithelial barrier function is intricately involved in the etiology of a wide range of diseases (Guttman and Finlay, 2009; Torres-Flores and Arias, 2015; DiGuilio et al., 2022). From cancer to chronic inflammation to a diverse spectrum of infectious diseases, even dust mite rhinitis, a barrier leak at the level of the tight junctional (TJ) complex can play a fundamental role in the disease process (Soler et al., 1999; Wan et al., 1999; Guttman and Finlay, 2009; Noureddine et al., 2022). A vibrant and rapidly growing body of literature also exists now, documenting those micronutrients—beneficial compounds and minerals that occur naturally in the diet at very low concentrations—that can improve barrier function and/or protect barrier function from the action of disease modifiers (Amasheh et al., 2009; DiGuilio et al., 2022). Both actions typically ensue at micronutrient concentrations that are above recommended dietary allowance (RDA) levels or, in the case of cell culture models, levels above those found in the culture medium with 10% serum (DiGuilio et al., 2022).

Flavonoids are polyphenolic phytochemical compounds found in plants and are some of the most beneficial micronutrients in our diet. In addition to their well-known role as antioxidants, flavonoids have also become recognized as being very adept at tight junction regulation and epithelial barrier improvement, actions that cannot be attributed simply to their antioxidant capability (Suzuki and Hara, 2011). Flavonoids are also well-described inhibitors of pro-inflammatory cytokine synthesis and secretion (Leyva-López et al., 2016; Martínez et al., 2019; Al-Khayri et al., 2022), actions that would certainly augment, if not account for, their improvement of epithelial barrier function.

Quercetin is one of the most widely abundant flavonoids in fruits and vegetables and, consequently, one of the most abundant in our diet (Formica and Regelson, 1995). It is also one of the most investigated flavonoids with respect to tight junctions and epithelial barrier function. Quercetin has been shown to support epithelial barrier function in a variety of epithelial cell culture models: rat intestinal IEC-6 cell layers (Fan et al., 2021; Fan et al., 2022), human retinal RPE cell layers (Li et al., 2021), human intestinal Caco-2 cell layers (Carrasco-Pozo et al., 2013; Valenzano et al., 2015; Sharma et al., 2020), human intestinal HT29/B6 cell layers (Amasheh et al., 2012), human gingival Gie-3B11 cell layers (Rybakovsky et al., 2017), human umbilical vein ECV304 cell layers (Chuenkitiyanon et al., 2010), and porcine renal LLC-PK1 cell layers (Mercado et al., 2013). In addition to epithelial cell culture models, quercetin also exhibited barrier improvement in epithelial tissue models: hen intestinal tissue (Amevor et al., 2022), pig intestine (Xu et al., 2021), rat lung (Tripathi et al., 2021), and rat ileum and colon (Amasheh et al., 2012). However, it is worth noting that quercetin has been reported to decrease the transepithelial electrical resistance (TER) of dog kidney MDCK II cell layers (Nakashima et al., 2020), although the opposite was reported by Gamero-Estevez et al. (2019), who showed increased transepithelial electrical resistance (TER) along with increased claudins-3 and -4 but decreased claudin-2. A quercetin-induced decrease in TER and an increase in mannitol leak were also reported in a human airway (16HBE) model (Callaghan et al., 2020).

When asserting that quercetin supports epithelial barrier function, it is essential to distinguish between two aspects of that action. The first is that quercetin improves basal-state barrier function, i.e., quercetin acts to increase TER and/or decrease leak of paracellular probe molecules in epithelial cell layers without being exposed to any other agent (Mercado et al., 2013; Valenzano et al., 2015; Rybakovsky et al., 2017; Gamero-Estevez et al., 2019). The second is that quercetin can protect an epithelial cell layer from the compromising effects of substances known to make these cell layers leaky. Quercetin has been shown to successfully protect against leaks induced by acrylamide (Fan et al., 2022), aging (Amevor et al., 2022), microcystin (Zhou et al., 2021), hypoxia (Tripathi et al., 2021), peroxide (Chuenkitiyanon et al., 2010; Li et al., 2021), indomethacin (Carrasco-Pozo et al., 2013; Fan et al., 2021), hyperglycemia (Sharma et al., 2020), and tumor necrosis factor (Amasheh et al., 2012).

This study sought to investigate, for the first time, the epithelial barrier improvement and protection of the Calu-3 airway epithelial model by quercetin. Barrier improvement was recorded by both TER increase and ¹⁴C-D-mannitol leak decrease, as well as a remodeling effect on TJ proteins. Quercetin’s inhibition of DNA synthesis, leading to a decrease in cell replication in the Calu-3 cell layer, may also be a mechanism of barrier improvement through the resulting decrease in linear junctional density. Protection by quercetin from the barrier compromise induced by the cytokine TNF-α is also shown, a phenomenon that appears to be associated with decreased ERK phosphorylation and a dramatic decrease in the TJ protein claudin-5.

Treatment of Calu-3 cell layers with 400 µM of quercetin for 72 h resulted in a statistically significant increase of over 40% in TER and a simultaneous significant decrease of over 25% in ¹⁴C-D-mannitol transepithelial diffusion (Figure 1). The simultaneous occurrence of both events strongly suggests that quercetin was causing a decrease, specifically in the paracellular permeability of Calu-3 cell layers. This quercetin-induced improvement of barrier function was concentration-dependent, with activity observed at concentrations as low as 25–50 µM (Figure 2).

The time course of quercetin’s (400 µM) action on Calu-3 barrier function was unusual. The dramatic improvement in barrier function, as evidenced by an over 3-fold increase in TER and a 70% decrease in ¹⁴C-D-mannitol leak, was seen as early as 5 h post-treatment. By 24 h, this situation reversed completely, with a 30% highly significant decrease in TER and an over 150% increase in ¹⁴C-D-mannitol leak compared to time-matched controls. At 48 h post-exposure, the situation spontaneously reversed itself again, with a 30% highly significant increase in TER and a 35% decrease in ¹⁴C-D-mannitol leak. The barrier then tightened further by 72 h on both metrics, all this occurring with only one treatment with quercetin (Figure 3).

The tightening of the barrier did not apply to only small solutes. Increased TER would be indicative of decreased paracellular conductance of Na⁺ and Cl⁻ ions, whose molecular weights are 23 and 35, respectively. The decreased permeability to D-mannitol is for a 182 molecular weight solute. However, Figure 4 shows that a 72-h treatment of Calu-3 cell layers with 400 µM quercetin also decreased the (paracellular) leak of ¹⁴C-inulin—a 3,500 mw solute. This signifies that the quercetin-treated barrier is likely tightening to mono- and di-saccharides, amino acids, peptides, and small proteins, as well as salts and water.

An analysis of TJ proteins by PAGE and Western immunoblot of Calu-3 cell layers treated with 400 µM qQuercetin for 72 h showed statistically significant changes in several TJ proteins in particulate (detergent-soluble) fractions (Figure 5). Specifically, we observed a statistically significant 25% decrease in claudin-2, a 20% decrease in claudin-3, and a borderline significant (p < 0.05, one-tailed Student’s t-test) 35% decrease in claudin-5 because of quercetin treatment. Claudin-1, -4, and -7 and occludin abundance were not significantly changed.

Calu-3 cultures treated with quercetin were observed to have increased cell size (Figure 6). Of distinct note for a study on barrier function, the increased cell size of the quercetin-treated cell layers translated to a statistically significant decrease in linear junctional density (total cell border length) as compared to matched controls (Figure 6E). This results in less junctional density per unit area of cell layer, and along with the altered tight junctional composition (Figure 5), both these factors can be reasons for the decreased paracellular permeability to probes such as 14C-D-mannitol as well as the increased TER. Proceeding from the microscopy observations, the potential effect of 400 µM quercetin on Calu-3 DNA synthesis was tested in cell layers cultured on permeable filters. Subconfluent cultures treated with quercetin exhibited a statistically significant decrease in DNA synthesis, as evidenced by a reduction in total DNA and a reduction in the incorporation of 14C-thymidine into DNA (Figure 7).

Quercetin was effective not only in improving the barrier function of normal, control Calu-3 cell layers but also in reducing the barrier compromise caused by the pro-inflammatory cytokine TNF-α. 150 ng/ml of TNF-α reduced TER by approximately 50% while simultaneously increasing 14C-D-mannitol leak by approximately 40% (Figure 8). 400 μM quercetin significantly reduced these actions of TNF-α when TNF-α and quercetin were simultaneously presented. The transepithelial leak produced by TNF-α may have been transduced by dramatically increased ERK 1/2 phosphorylation; this increased phosphorylation was totally inhibited by 400 µM quercetin (Figure 9). The barrier compromise caused by 72-h exposure to TNF-α coincided with a dramatic (over 200%) increase in the tight junction protein, claudin-5. This increase was significantly reduced by simultaneous exposure to 400 µM quercetin (Figure 10).

The role of flavonoid micronutrients in improving epithelial barrier function and ameliorating the barrier compromise/leak caused by pro-inflammatory cytokines is becoming increasingly established in the published literature (Amasheh et al., 2008; Suzuki and Hara, 2009; Suzuki and Hara, 2011; Carrasco-Pozo et al., 2013; Valenzano et al., 2015; Zhang et al., 2022). The flavonoid, quercetin, and airway epithelial barriers, such as Calu-3, are mere specific examples of an overall flavonoid trend regarding this class of polyphenols improving barrier function. In the series of experiments presented in this study, we add to that body of evidence of barrier improvement and protection by flavonoids (Figure 1). This activity is shown to be concentration-dependent (Figure 2) and exhibits a curious, biphasic temporal response (Figure 3) that has also been previously reported in MDCK renal epithelia (Gamero-Estevez et al., 2019). It encompasses improvement of normal (control) epithelial barrier function and protection from the effects of a leak inducer (Figure 8).

The clinical compromise of airway barrier function—and resultant lung water accumulation—was evident in dramatic fashion very recently in the COVID-19 pandemic and the significant need that it generated for critical care medicine and mechanical ventilators (Rasch et al., 2021; Virot et al., 2021). However, those needs are equally relevant in acute respiratory distress syndrome (Kreuzfelder et al., 1988). Both instances become compelling reasons to search out and research therapeutics for barrier compromise. The fact that naturally occurring—even dietary—compounds are very adept in that regard offers a safety and economic attribution that is very appealing. Although not a flavonoid or polyphenol, vitamin D offers yet another example of a dietary compound that proved clinically useful in COVID-19 and exhibited the ability to modify and improve the barrier function of airway epithelial models (Bae et al., 2022; Rastogi et al., 2022; Gao et al., 2023; Rybakovsky et al., 2023).

Among researchers who study barrier remodeling, it is always tempting to think that induced TJ remodeling is the reason for the ability of a particular agent, such as quercetin, to improve barrier function. The TJ protein changes observed in this study and reported in Figure 5 (such as decreased claudins-2, -3, and -5) may indeed be evidence of such activity. However, simpler explanations may also be available, and the observations that quercetin inhibits cell replication (Figure 7) and decreases linear junctional density (Figure 6) may be equally important. Rabito (1986) made the original observation in LLC-PK₁ renal epithelial cell layers that decreased linear junctional density would translate to improved barrier function. Our observation that quercetin can affect ERK 1/2 phosphorylation (Figure 9) represents a more fundamental molecular mechanism consistent with observations by others that inhibiting ERK phosphorylation decreases TJ leaks and cell cycle activity (Crews et al., 1992; Kevil et al., 2000; Petecchia et al., 2012). Therefore, TJ remodeling and the decreased linear junctional density resulting from increased cell size (both mediated by changes in ERK 1/2 phosphorylation) become viable potential mechanisms to explain the improved Calu-3 barrier function seen here.

In addition, regarding the molecular mechanism, it is worth noting that flavonoids (e.g., genistein but also quercetin), as a class, exhibit protein kinase C-inhibitory activity (Suzuki and Hara, 2011). Given that protein kinase C activation—including protein kinase C delta—is a highly effective way to increase TJ leak (Mullin et al., 1998a), inhibition of protein kinase C by quercetin could be another means by which quercetin improves Calu-3 barrier function.

Turning to the biomedical implications of these and similar findings, we would point out that a TJ (barrier) leak is intrinsic to a wide variety of diseases and disease processes and that an agent that opposes such a leak would likely have therapeutic potential, a subject covered in a recent extensive review (DiGuilio et al., 2022). Diarrhea, liver insufficiency, renal (tubular) insufficiency, and accumulation of lung water (think multi-organ failure here) are clinical manifestations of the type of barrier leak that we observe in this study (e.g., the TNF-α-induced leak) as revealed by decreased TER (i.e., increased passive transepithelial leak (conductance) of salt and water) and increased ¹⁴C-D-mannitol permeability (indicative of leak of sugars, amino acids, and other small molecules). The ability of quercetin to partially reduce these aspects of leak suggests its utility in treating those manifestations. However, the quercetin-induced decrease in leak of ¹⁴C-inulin (Figure 4) indicates that it is not simply physiological osmolytes/substrates and water whose leak is being reduced by flavonoids but also very likely the pathophysiological transepithelial leak of peptides and small proteins that could then result in abnormal regulatory effects in these tissues (Mullin and McGinn, 1987; Mullin et al., 1998b). Therefore, the therapeutic action of naturally occurring leak reducers, such as quercetin, could be biomedically substantial.

In summary, our results demonstrated that quercetin can improve Calu-3 epithelial barrier function and reduce TNF-α-induced barrier compromise, mediated in part by changes in the tight junctional complex. In this study, we add to the knowledge of the regulation of barrier function in a widely used and applicable model of human airway physiology. We also lay out for discussion whether flavonoids such as quercetin could have adjuvant therapeutic utility in airway diseases that involve barrier compromise and fluid accumulation, such as but not exclusive to the recent and ongoing COVID-19 outbreaks.