Parkin promotes airway inflammatory response to interferon gamma

Background Increased type 2 interferon (i.e., IFN-γ) signaling has been shown to be involved in airway inflammation in a subset of asthma patients who often show high levels of airway neutrophilic inflammation and poor response to corticosteroid treatment. How IFN-γ mediates airway inflammation in a mitochondrial dysfunction setting (e.g., Parkin up-regulation) remains poorly understood. The goal of this study was to determine the role of Parkin, an E3 ubiquitin ligase, in IFN-γ-mediated airway inflammation and the regulation of Parkin by IFN-γ. Results Using a mouse model of IFN-γ treatment in wild-type and Parkin knockout mice, and cultured human primary airway epithelial cells with or without Parkin gene deficiency, we found that Parkin was necessary for the production of neutrophil chemokines (i.e., KC and IL-8) and airway neutrophilic inflammation. Mechanistically, Parkin was induced by IFN-γ treatment both in vivo and in vitro, which was associated with less expression of a Parkin transcriptional repressor Thap11. Overexpression of Thap11 inhibited Parkin expression in IFN-γ-stimulated airway epithelial cells. Conclusions Our data suggests a novel mechanism by which IFN-γ induces airway neutrophilic inflammation through the Thap11/Parkin axis. Inhibition of Parkin expression or activity may provide a new therapeutic target for the treatment of excessive neutrophilic inflammation in an IFN-γ high environment.


Background
Asthma is a heterogenous in ammatory airway disease characterized by various endotypes. The most common endotypes are type 2 in ammation high and type 2 in ammation low. Type 2 high asthma is featured by increased type 2 in ammation including higher levels of IL-4, IL-5, and IL-13 release and in ltration of eosinophils in the airways [1]. In contrast, airways of type 2 low asthma demonstrate neutrophilic and pauci-granulocytic airway in ltrates that are associated with cytokines such as interferon gamma (IFN-γ), IL-6 and IL-17 [2]. Currently, there are limited therapeutic options for type 2 low asthma compared to type 2 high asthma. In a study by Ricciardolo et al [3], high levels of IFN-γ expression were found in type 2 low asthma subjects. Additionally, increased IFN-γ in bronchoalveolar lavage uid was associated with neutrophilic in ammation in severe asthma patients [4]. This was further supported by a recent study demonstrating the association of increased IFN-γ expression in airway samples from a cohort of asthma patients with more severe disease [5]. The mechanisms of increased IFN-γ expression are still under investigation, although several major factors such as viral and bacterial infections have been proposed. Moreover, the modifying factors or regulators of the host in ammatory response to IFN-γ have not been fully understood.
Parkin, an E3 ubiquitin ligase, is encoded by PARK2 gene and expressed by various tissues (e.g., brain, kidney, lung) and different types of cells including neural cells and non-neural cells such as lung epithelial cells. Parkin protein consists of a N-terminal ubiquitin-like domain, the linker region and the C-terminal RING-box [6-8]. As an intracellular protein, the primary function of Parkin is related to its E3 ubiquitin ligase activity, a key component in 26S proteasome-mediated protein degradation. Recently, Parkin has been shown to exert non-classical/novel functions such as a pro-in ammatory role [9]. We have previously investigated the regulation of airway type 2 in ammation by Parkin [10]. Speci cally, Parkin promoted type 2 airway in ammation in a human airway epithelial cell air-liquid interface culture system, and in mouse models of IL-13 treatment and allergen challenges. A recent study [11] suggests that Parkin impairs host antiviral immunity by inhibiting the antiviral in ammation, but not affecting the type 1 interferon response. Whether Parkin regulates airway type 2 interferon (i.e., IFN-γ) response is unknown.
Regulation of Parkin expression or function in the airways has not been well understood. We have demonstrated that Parkin expression in airway epithelium of asthma patients was increased [10]. A previous study utilizing the genome-wide CRISPR screening approach identi ed transcription factor Thap11 as a major repressor of Parkin expression as Thap11 knockdown was shown to increase Parkin expression and activity [12]. It is unclear whether IFN-γ may regulate Thap11 expression, and subsequently affect Parkin expression. By using the human airway epithelial cell culture system and mouse models, we determined the interplay of IFN-γ, Thap11 and Parkin by testing whether IFN-γ upregulates Parkin by inhibiting Thap11, and whether Parkin in turn ampli es host pro-in ammatory response to IFN-γ.

Intranasal administration of recombinant IFN-γ protein in mice
Wild-type (WT) C57BL/6 and Parkin knockout (PKO) mice on C57BL/6 background were purchased from the Jackson Laboratory (Bar Harbor, Maine, USA). All mouse procedures were approved by the Institutional Animal Care and Use Committee (protocol #AS2792-03-23) at National Jewish Health.
To induce lung in ammation in mice, WT and PKO mice were intranasally inoculated with recombinant mouse IFN-γ (PeproTech, Cranbury, NJ) at 25 ng/mouse prepared in 0.1% bovine serum albumin (BSA) or 0.1% BSA (control) once daily for three consecutive days. We used this dose based on our pilot experiment comparing lung neutrophil levels between 10 ng/mouse and 25 ng/mouse, and we found that 25 ng/mouse induced about 15% of neutrophils in bronchoalveolar lavage uid (BALF), while 10 ng/ml IFN-γ resulted in minimal lung neutrophilic in ammation. Raundhal et al. reported this similar percentage of neutrophils in BALF from human severe asthma subjects with more IFN-γ + T cells in the airways [4].
Mice were sacri ced after 24 hours of the last IFN-γ treatment. BAL was performed with 1 mL of sterile saline. BAL cells were used for the leukocyte count and BALF was analyzed for pro-in ammatory mediators.

Ifn-γ Stimulation In Cultured Primary Human Airway Epithelial Cells
Human tracheobronchial epithelial (HTBE) cells from donors (n = 6) without lung disease or smoking history were used for cell culture. The human study was approved by the National Jewish Health Institutional Review Board (Protocol # HS-3209). HTBE cells were cultured in 12-well cell culture plates (10 5 cells/well) under the submerged condition in the presence or absence of IFN-γ (5 ng/ml) for 24 hours to measure Thap11 and Parkin protein expression.

Thap11 Overexpression In Primary Human Airway Epithelial Cells
To determine the role of Thap11 in Parkin expression in the absence or presence of IFN-γ, Thap11 overexpression was performed in HTBE cells from a donor without lung disease and smoking history. Lentivirus system was utilized to overexpress Thap11 using reagents from Genecopoeia (Rockville, MD).
Thap11 open reading frame (ORF) or scrambled control sequence was cloned into a lentiviral vector (pReceiver-Lv205) and subsequently packaged in 293FT cells using the Lenti-Pac HIV Expression Packaging Kit. After 48 hours, the supernatant containing the viral particles was harvested. HTBE cells grown in 24-well plates under submerged condition were transduced with 70% viral particle supernatant plus 30% culture medium or 100% viral particle supernatant for 72 hours before treatment with IFN-γ (10 ng/ml) [13]. After 24 hours, epithelial supernatant was collected for ELISA while cell lysates were harvested in RIPA for western blot.

Parkin Knockout (Pko) In Primary Htbe Cells
Control and PKO HTBE cells from a donor without lung disease and smoking history were generated using CRISPR-Cas9 as detailed in our previous publication [10]. Brie y, a single guide RNA (sgRNA, 5' AGTCTAAGCAAATCACGTGG 3') was designed to target exon 7 of human Parkin in HTBE cells. For the control CRISPR, a scrambled sgRNA was used. The transduced HTBE cells were expanded and cultured at air-liquid interface (ALI) for 21 days to induce mucociliary differentiation as previously described [10]. At day 21, cells were stimulated with recombinant human IFN-γ (5 ng/ml) or 0.1% BSA (control solution for preparing IFN-γ) for three days. Basolateral supernatants were collected for ELISA while cell lysates were harvested in RIPA for western blot analysis.
Densitometry was performed to quantify the protein expression levels using the National Institutes of Health's ImageJ software. ELISA Human IL-8 and mouse LIX/CXCL5 were measured using Duoset ELISA kits from R&D systems (Minneapolis, MN).

Statistical Analyses
Data were analyzed using Graph Pad Prism software. For parametric data, a paired Student's t-test was performed for two-group comparisons or two-way ANOVA followed by the Tukey's multiple comparison test. For non-parametric data, comparisons were done using the Mann-Whitney test for two group comparisons or using the Kruskal-Wallis test for multiple group comparisons. A p value of < 0.05 was considered to be statistically signi cant.

Results
Parkin was essential to mouse lung neutrophilic in ammation induced by IFN-γ To determine the in vivo function of Parkin in lung neutrophilic in ammation following IFN-γ treatment, wild-type (WT) and Parkin knockout (PKO) mice were treated with IFN-γ. WT mice treated with IFN-γ signi cantly increased the number and percentage of neutrophils ( Fig. 1A and B) as well as neutrophilic chemoattractant LIX in BALF (Fig. 1C). However, PKO mice was not able to signi cantly increase LIX and neutrophil levels following IFN-γ treatment. These ndings suggest that Parkin enhances neutrophilic in ammation after IFN-γ challenge in vivo. As expected, Parkin was de cient in the lungs of PKO mice (Fig. 1D). Notably, IFN-γ induced Parkin expression in WT mouse lung tissues. IFN-γ directly increased Parkin, while inhibited Thap11 expression in primary human airway epithelial cells Next, we sought to determine whether IFN-γ increases Parkin expression in human airway epithelial cells, which is associated with reduction of Thap11, a transcriptional repressor of Parkin [12]. In submerged culture of primary human airway epithelial cells, IFN-γ signi cantly increased Parkin expression ( Fig. 2A), which further supported our mouse model data. Importantly, increased Parkin expression by IFN-γ was accompanied by decreased Thap11 expression (Fig. 2B), suggesting a negative relationship between Thap11 and Parkin under IFN-γ stimulation.
Overexpression of Thap11 decreased Parkin expression in primary human airway epithelial cells treated with IFN-γ Knockdown of Thap11 in the HEK293 cell line was shown to increase Parkin mRNA and protein expression in a previous study [12]. Whether Thap11 regulates Parkin expression in human primary airway epithelial cells has not been investigated. To directly test if Thap11 regulates Parkin, we overexpressed Thap11 in HTBE cells with two different doses of lentiviral particles containing the Thap11 ORF plasmid. Cells transduced with 100% lentivirus particles expressed Thap11 protein that was about 5 times higher than the control cells (Fig. 3A). Although cells transduced with 70% viral particles showed higher Thap11 expression, it was not signi cantly higher than the control cells. We measured IL-8 to determine if the lentivirus particles induced a pro-in ammatory response. Notably, IL-8 levels were similar between cells transduced with 70% or 100% viral particles and control cells (Fig. 3B), suggesting that the lentiviral particles did not induce the proin ammatory response. We proceeded to stimulate the cells receiving 100% viral particles with IFN-γ. Enhancing Thap11 expression in HTBE cells stimulated with IFNγ decreased Parkin expression as well as IL-8 levels compared to the scrambled control cells. This suggests that restoring Thap11 expression in IFN-γ-stimulated cells decreased Parkin expression and the pro-in ammatory response.
Parkin is essential to maintain human airway epithelial proin ammatory response to IFN-γ Primary HTBE cells with Parkin knockout using the CRISPR-Cas9 approach were utilized to test if Parkin induction by IFN-γ contributes to the production of pro-neutrophilic molecules. Similar to the mouse lung data, IFN-γ increased Parkin protein expression in scrambled control CRISPR cells (Fig. 4A). We con rmed the loss of Parkin protein expression in the absence and particularly in the presence of IFN-γ stimulation.
The pro-neutrophilic chemokine IL-8 was increased in IFN-γ stimulated scrambled control cells compared to untreated cells (Fig. 4B). Importantly, Parkin de ciency decreased IL-8 production after IFN-γ stimulation, suggesting that Parkin may maintain or promote the pro-neutrophilic response to IFN-γ.

Discussion
Whether Parkin regulates type 2-low (e.g., IFN-γ high) lung in ammatory responses has not been reported. Here, for the very rst time, we demonstrated that Parkin promotes IFN-γ-mediated lung neutrophilic in ammation. Moreover, we discovered that IFN-γ serves as a positive feedback loop mechanism to increase Parkin expression.
We have reported increased Parkin expression in asthmatic airway epithelium [10]. However, the role of Parkin in lung in ammation has not been well understood. An early mouse model study suggested that Parkin regulates lipopolysaccharide (LPS)-induced proin ammatory responses as Parkin-de cient mice demonstrated less severe lung in ammation (e.g., neutrophils) as compared to the wild-type mice [14]. Our data generated in Parkin-de cient mice treated with IFN-γ further suggests a role of Parkin in promoting lung neutrophilic in ammation. How Parkin promotes or maintains the pro-in ammatory response to IFN-γ remains to be explored. IFN-γ-mediated effects have been shown to be associated with activation of STAT1 and NF-κB [15,16]. A previous study showed that a RBR E3 ubiquitin ligase Natural Killer Lytic-Associated Molecule (NKLAM) may interact with STAT1 to maintain STAT1-mediated transcriptional activity in macrophages [17]. Parkin has been functionally considered as a RBR E3 ligase, but recent studies suggest that it may also act as RING-between-RING and HECT E3 ligases [18]. It remains to be determined if Parkin promotes IFN-γ-mediated in ammatory response through increasing STAT1-induced gene expression. Additionally, Parkin was shown to up-regulate NF-κB activation [19], which may be another potential mechanism underlying the pro-in ammatory effect of Parkin.
We not only found that Parkin regulates IFN-γ-mediated in ammatory response, but also discovered that Parkin was up-regulated by IFN-γ stimulation. Parkin expression can be regulated at multiple (e.g., genetic, epigenetic, transcriptional) levels. While transcription factor activating transcription factor 4 (ATF4) was shown to up-regulate Parkin under ER stress [20], Thap11 may serve as a transcriptional repressor of Parkin [12]. Our data in airway epithelial culture demonstrated that IFN-γ reduced Thap11, which was coupled with increased Parkin expression. In IFN-γ-stimulated airway epithelial cells, Thap11 overexpression reduced Parkin expression. Together, our data suggests that IFN-γ increases Parkin expression in part through reducing the expression of Thap11. At the moment, it is unclear whether IFN-γ may also regulate E3 ligase activity of Parkin, but this warrants further investigation in order to fully understand the mutual regulation of IFN-γ and Parkin.
We realize several limitations to this study. First, we utilized an acute mouse model of IFN-γ treatment to study the role of Parkin in the in ammatory process. Future studies are necessary to determine the role of Parkin in chronic type 2-low in ammation as asthma and other lung diseases are of chronic nature. Second, how Parkin exactly regulates IFN-γ signaling was not addressed in this study, but the role of Parkin in regulating STAT1 and NF-κB pathways will be considered in our future studies. Third, a previous study has shown the ability of IFN-γ to enhance ATF4 mRNA expression under ER stress in pancreatic beta cells [21]. Whether IFN-γ utilizes other transcription factors such as ATF4 to up-regulate Parkin in human airway epithelial cells could be further studied. Lastly, we established Thap11 overexpressing and Parkin knockout airway epithelial cell lines only from one human donor. While these cell lines provided clear evidence about the role of Thap11 and Parkin in IFN-γ-mediated in ammatory response, our results may need to be replicated in cells from additional donors.

Conclusions
Our current study has deepened our understanding about the role of Parkin in type 2-low in ammation. As Parkin dysregulation may involve multiple biological processes such as mitochondrial dysfunction, it is conceivable that unraveling its regulation and functions under physiological and pathological conditions may provide insights into mechanisms of a variety of diseases with elevated IFN-γ or type 1 or type 3 interferons as seen in respiratory viral infections.