Ultraviolet-A light increases mitochondrial anti-viral signaling protein in confluent human tracheal cells even at a distance from the light source

Mitochondrial antiviral signaling (MAVS) protein mediates innate antiviral responses, including responses to certain coronaviruses such as severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2). We have previously shown that ultraviolet-A (UVA) therapy can prevent virus-induced cell death in human ciliated tracheal epithelial cells (HTEpC) infected with coronavirus-229E, and that UVA treatment results in an increase in intracellular levels of MAVS. In this study, we set out to determine the mechanisms by which UVA light can activate MAVS, and whether local UVA light application can activate MAVS at locations distant from the light source (such as via cell-to-cell communication). MAVS levels were compared in HTEpC exposed to 2 mW/cm2 narrow band (NB)-UVA for 20 minutes and in unexposed controls, at 30-40% and at 100% confluency. MAVS levels were also compared in unexposed HTEpC treated with supernatants or lysates from UVA-exposed cells or from unexposed controls. Also, MAVS was assessed in different sections of confluent monolayer plates where only one section was exposed to NB-UVA. The results show that UVA increases the expression of MAVS protein. Cells in a confluent monolayer exposed to UVA were able to confer an elevation in MAVS in cells adjacent to the exposed section, and even cells in the most distant sections not exposed to UVA. In this study, human ciliated tracheal epithelial cells exposed to UVA demonstrate increased MAVS protein, and also appear to transmit this influence to distant confluent cells not exposed to light.


22
The human body has various defense mechanisms against infections, the most well-23 known of which involve innate immune responses where immune cells are recruited to sites of 24 infection via cytokine signaling [1,2]. Host intracellular responses to infection are also 25 important, particularly in the defense against viruses. In the past decade, it has been discovered 26 that mitochondria can mediate innate and adaptive immune responses via several mechanisms 27 [3], including the production of mitochondrial anti-viral signaling (MAVS) protein [4].

29
The MAVS protein is primarily localized to the outer membrane of the mitochondria, and  In this study, we explore the effects of narrow band (NB)-UVA light on MAVS 52 expression in uninfected human ciliated tracheal epithelial cells in vitro. We also explore 53 whether the effects of UVA light were limited to cells directly exposed to UVA, or were also 54 seen in cells not directly exposed to UVA.  Once the cells reached 10 5 cells per plate (30-40% confluency), HTEpC were washed 3 66 times with sterile 1x PBS pH 7.4 (cat. 10010072, ThermoFisher), and fresh media was added to 67 each plate. Cells were exposed to 2 mW/cm 2 of NB-UVA for 20 minutes based on previously 68 validated ideal UVA irradiation levels [6]. Unexposed cells were used as controls. After 24 hours 69 the supernatants were collected, and cell were washed 3 times with sterile 1x PBS, pH 7.4.

70
Following the removal of any remaining PBS, cells were lysed in the plate using 1 mL of RTL HTEpC that were exposed to NB-UVA were transferred to 30-40% confluent naïve HTEpC.

84
-To determine if cell-to-cell signaling was involved, areas of 100% confluent HTEpC exposed 85 or not exposed to NB-UVA were analyzed. Supernatants collected from UVA-exposed and control HTEpC from the previous 90 experiment were transferred to a new 60x15mm tissue culture dish containing 10 5 naïve HTEpC 91 (i.e. cells that were never exposed to UVA). Before receiving the supernatant from UVA-exposed or control cells, the naïve HTEpC were washed 3 times with sterile 1x PBS, pH 7.4. The

93
PBS was completely removed, and 4 mL of the supernatant collected from UVA-exposed or 94 control HTEpC were added to the naïve cells. After 24 hours of incubation, the cells were 95 washed 3 times, and were then lysed in the plate using 1 mL of RTL buffer from an AllPrep 96 DNA/RNA/Protein isolation kit (Qiagen). Experiments were performed in triplicate. HTEpC were cultured at 37ºC (5% CO 2 ) in 60x15mm standard tissue culture dishes (cat.  Once the cells reached 10 5 cells per plate (30-40% confluency), HTEpC were washed 3 106 times with sterile 1x PBS pH 7.4 (cat. 10010072, ThermoFisher), and fresh media was added to 107 each plate. Cells were exposed to 2 mW/cm 2 of NB-UVA for 20 minutes. Unexposed cells were 108 used as controls. After 24 hours, the cells were washed 3 times with sterile 1x PBS, pH 7.4, 109 scraped from the culture dishes, and transferred to a 15mL sterile tube. Cells were pelleted, and 110 new fresh Airway Epithelial Cell Growth Medium was added. A single sterile 5 mm stainless 111 steel bead (Qiagen) was added to each tube, and cells were lysed by vortexing the tube for 5 112 minutes. Lysates from UVA-exposed and control HTEpC were transferred to a new 60x15mm 113 tissue culture dish containing 10 5 naïve HTEpC (i.e. HTEpC that had never been exposed to 114 UVA). Before receiving the lysate from UVA-exposed or control cells, naïve HTEpC were 115 washed 3 times with sterile 1x PBS, pH 7.4. The PBS was completely removed, and 4 mL of the 116 lysate from either UVA-exposed or control HTEpC were added to the naïve cells. After 24 hours 117 of incubation, the cells were washed 3 times with sterile 1x PBS and were then lysed in the plate 118 using 1 mL of RTL buffer from an AllPrep DNA/RNA/Protein isolation kit (Qiagen).

119
Experiments were performed four times.     The remaining UVA-exposed HTEpC from areas 1, 2 and 3 (still attached to the plate) 154 were washed 3 times with sterile 1x PBS, pH 7.4. 10 mL of sterile 1x PBS, pH 7.4, was added to 155 the plate and cells from area 3 were carefully scraped and lysed as described above. The same 156 process was used to harvest the cells from areas 2 and 1 (in this order). AllPrep DNA/RNA/Protein Mini Kits (Qiagen) were used to extract total proteins from 160 UVA-exposed and non-exposed HTEpC from all experiments, according to the manufacturer's 161 protocol. Total proteins were quantitated using Qubit Protein Assays (ThermoFisher) and equal 162 loads of total protein were separated on a NuPAGE 4-12% Bis-Tris mini gel (NP0336BOX,  confluency which were exposed to 2 mW/cm 2 NB-UVA for 20 minutes and in unexposed 190 controls. Normalized MAVS levels, as detected by western blot, were increased in NB-UVA 191 exposed cells when compared to unexposed controls (P=0.0193, Fig 2).    205 When naïve 30-40% confluent HTEpC were treated with supernatants from NB-UVA 206 exposed 30-40% confluent HTEpC, no changes in MAVS levels were observed (P=0.4022, Fig   207   4). However, when naïve 30-40% confluent HTEpC were incubated with cell lysates from NB-208 UVA exposed 30-40% confluent HTEpC, normalized levels of MAVS tended to increase (Fig 5,   209 P=0.1256).   Next, levels of MAVS were analyzed in different areas of culture plates containing 100% 226 confluent monolayers of HTEpC, after only one part of the plate (area 1) was exposed to 2 227 mW/cm 2 NB-UVA for 20 min (Fig 1). Normalized MAVS levels gradually increased from area 228 4 (farthest unexposed area) through area 1 (exposed to NB-UVA) (ANOVA P=0.08, Fig 6A,B), 229 and there was a statistically significant increase in MAVS levels in area 1 (exposed to NB-UVA) 230 when compared to unexposed area 4 (P=0.0382, Fig 6A,B). Importantly, levels of MAVS were 231 also significantly increased in unexposed areas 2 and 3 when compared to controls from 232 unexposed plates (P=0.0289 and P=0.0402 respectively, Fig 6A). Normalized MAVS levels in 233 area 4 (farthest unexposed area) also appeared to be higher than in controls, but did not reach 234 statistical significance (P=0.1262, Fig 6A). NB-UVA for 20 minutes. Area 1 was directly exposed to NB-UVA, but areas 2, 3 and 4 were not 238 exposed to NB-UVA. B -Western blot prepared from cell lysates of 100% confluent HTEpC 239 from three experiments, exposed to NB-UVA (area 1 -lanes 1, 5 and 9) and from lysates of 240 confluent HTEpC not exposed to NB-UVA from the same culture plate (area 2 -lanes 2, 6 and 241 10; area 3 -lanes 3, 7 and 11; area 4 -lanes 4, 8, and 12).    improvements, despite the fact that only small portions of the trachea were exposed to UVA 306 light, suggested the possibility that the antiviral effects of UVA light might not be confined to 307 cells directly exposed to UVA, but might also be transmitted to neighboring cells.

309
To explore the potential mechanisms underlying this transmission, we first took 310 supernatants from UVA-exposed cells and added them to fresh plates of cells that were never 311 exposed to UVA light. No increase in MAVS protein levels were seen in these cells, indicating 312 that a secreted extracellular mediator was not involved. Next, to explore whether a cytosolic 313 mediator was involved, we lysed UVA-exposed cells and non-exposed controls and added the 314 lysates to fresh plates of cells that were never exposed to UVA light. There was a trend towards 315 an increase in MAVS protein levels in naïve HTEpC incubated with lysates from UVA-exposed 316 cells, but this did not reach significance. In contrast, when we compared MAVS levels in 317 confluent monolayers of HTEpC directly exposed to UVA light and in adjacent areas from the 318 same plate that were blocked from UVA light, we found that MAVS was not only increased in 319 cells in area 1 (directly exposed to UVA light), but was also increased in cells in the adjacent 320 areas 2, 3, and 4 which were blocked from direct UVA light, in a gradient that decreased with 321 increasing distance from UVA-exposed cells. These findings confirm that an increase in MAVS 322 in response to UVA light can be transmitted from directly exposed cells to neighboring 323 unexposed cells, and suggest that cell-to-cell signaling is involved, although further work is 324 required to determine the mechanisms involved.  In conclusion, this study begins to unravel the possible mechanisms by which UVA light 341 could influence innate intracellular immunity. It appears that NB-UVA increases MAVS protein 342 levels in human ciliated tracheal epithelial cells. This increase in MAVS protein appears to be 343 transmissible to adjacent cells not directly exposed to UVA light. Further, our results suggest that  The authors thank Chandrima Chatterjee for her assistance in preparing the figures. We 353 would also like to thank Ewan Seo for applying our ideas to paper in the early CAD drawings for 354 the human UVA intra-tracheal device. In addition, the authors thank Frank Lee for his support of