Access to a main alphaherpesvirus receptor, located basolaterally in the respiratory epithelium, is masked by intercellular junctions

The respiratory epithelium of humans and animals is frequently exposed to alphaherpesviruses, originating from either external exposure or reactivation from latency. To date, the polarity of alphaherpesvirus infection in the respiratory epithelium and the role of respiratory epithelial integrity herein has not been studied. Equine herpesvirus type 1 (EHV1), a well-known member of the alphaherpesvirus family, was used to infect equine respiratory mucosal explants and primary equine respiratory epithelial cells (EREC), grown at the air-liquid interface. EHV1 binding to and infection of mucosal explants was greatly enhanced upon destruction of the respiratory epithelium integrity with EGTA or N-acetylcysteine. EHV1 preferentially bound to and entered EREC at basolateral cell surfaces. Restriction of infection via apical inoculation was overcome by disruption of intercellular junctions. Finally, basolateral but not apical EHV1 infection of EREC was dependent on cellular N-linked glycans. Overall, our findings demonstrate that integrity of the respiratory epithelium is crucial in the host’s innate defence against primary alphaherpesvirus infections. In addition, by targeting a basolaterally located receptor in the respiratory epithelium, alphaherpesviruses have generated a strategy to efficiently escape from host defence mechanisms during reactivation from latency.


Supplementary experimental procedures EHV1 purification and Dio-labelling
Culture fluids of EHV1-infected RK-13 cells were clarified by centrifugation at 60,000g for 2h at 4°C. The virus pellet was pooled onto a discontinuous OptiPrep™ gradient (Sigma-Aldrich, St. Louis, MO, USA) containing 10-30% (w/v) of iodixanol and centrifuged at 100,000g for 2.5h at 4°C. After centrifugation, purified opalescent virus bands were harvested at the interface of the 15% and 20% layers. To ensure efficient virus lipophilic labelling, the buffer was exchanged to HNE buffer (5 mM HEPES, 150 mM NaCl, 0.1 mM EDTA, pH 7.4) by the use of a 50K filter device (Millipore corporation, Bedford, MA, USA). While vortexing, 2 nM of 3,3'-Dioctadecyloxacarbocyanine perchlorate (Dio) dissolved in DMSO (Molecular probes, Oregon, USA) was added to the virus. Subsequently, unbound Dio was removed by centrifugation onto a MicroSpin™ G-50 fine column (GE Healthcare, Buckinghamshire, UK).
The degree of Dio-labelled virus purity (>90%) was evaluated by simultaneous immunofluorescent staining of EHV1 gB with mouse monoclonal antibody 3F6 (kindly provided by Prof. U. Balasuriya, University of Kentucky, USA) and quantitative analysis by confocal microscopy.

Respiratory mucosal explant isolation and cultivation
The respiratory mucosa was stripped from the underlying cartilage and washed in PBS to remove excess blood. Tissues were cut into small square pieces (25 mm 2 ), placed with the epithelial side facing upwards onto fine-meshed gauzes and cultured in a 37°C, 5% CO2, humidified incubator for 24h at air-liquid interface in serum-free medium containing DMEM/RPMI (Invitrogen, Paisley, UK), supplemented with 0.1 mg/mL gentamicin, 100 U/mL penicillin, 0.1 mg/mL streptomycin, and 0.25 µg/mL amphotericin B.

EREC isolation and cultivation
Tracheae were trimmed upon arrival in the lab and washed in PBS to remove excess blood.
After 5-7 days, the EREC attained a trans-epithelial electrical resistance (TEER) of ~500-700 Ω•cm -2 . TEER was measured using an epithelial voltohmmeter (Millipore). The net resistance was calculated by subtracting the background resistance and multiplying the resistance by the surface area of the membrane. 4

Disruption of intercellular bridges of respiratory mucosal explants
First, 24-well culture dishes were filled with 1 mL of a solution containing 50% sterile 3% agarose (low temperature gelling; Sigma-Aldrich) and 50% 2X MEM (Invitrogen). Explants were placed onto the solidified agarose with the epithelial surface facing upwards. Additional agarose was added until the lateral surfaces of the mucosa were fully occluded. Explants were then exposed for 1h at 37°C to different drugs (8 mM EGTA, 500 mM NAC, 20 mM DTT or 50 mM β-mercaptoethanol in PBS). PBS supplemented with calcium and magnesium was used as a control. Finally, explants were washed 3 times to remove excess drugs and were fixed in phosphate-buffered 3.5% formaldehyde solution, either immediately or after an additional 24h incubation. An automated system was used for paraffin embedding of the samples (Thermo Scientific™ STP 120 Spin Tissue Processor). Eight µm paraffin sections were first deparaffinised in xylene, then rehydrated in descending grades of alcohol, subsequently stained with haematoxylin-eosin, dehydrated in ascending grades of alcohol and xylene and finally mounted with DPX (Sigma-Aldrich). Ten pictures on five different sections per treated explant were taken with an Olympus IX50 light microscope fitted with 40X objective. The percentage of intercellular space in the epithelium was measured using ImageJ software (ImageJ, U.S. National Institutes of Health, Bethesda, Maryland, USA). The region of interest (ROI, i.e. the epithelium) was drawn manually for each picture in the "ROI manager tool". Next, the threshold value to distinguish blank spaces from cellular material was determined and the percentage of blank spaces between the cells (i.e. the intercellular space) was calculated.

Respiratory mucosal explants
Sixteen µm thick cryosections were cut using a cryostat at -20°C and loaded onto 3aminopropyltriethoxysilane-coated (Sigma-Aldrich) glass slides. Slides were then fixed in 4% paraformaldehyde for 15min and subsequently permeabilized in 0.1% Triton-X 100 diluted in 5 PBS. Non-specific binding sites were blocked by 15min incubation with avidin and biotin (Invitrogen) at 37°C. To label late viral glycoproteins, a polyclonal biotinylated horse anti-EHV1 was used for 1h at 37°C 1 , followed by incubation with streptavidin-FITC ® (Invitrogen) for 1h at 37°C. The basement membrane of the tissues was stained with monoclonal mouse anti-collagen VII antibodies (Sigma-Aldrich), followed by secondary Texas Red ® labelled goat anti-mouse antibodies (Invitrogen). Nuclei were detected by staining with Hoechst 33342 (Invitrogen). Slides were mounted with glycerol-DABCO and analysed using a Leica (TCS SPE) confocal microscope. The total number of plaques was counted on 50 cryosections and plaque latitude was measured using the Leica confocal software package. Five cryosections per explant were completely photographed and the percentage of infection in the epithelium (i.e. ROI) was determined using Image J software. The ROI (i.e. the epithelium) was drawn manually for each picture in the "ROI manager tool". Next, the threshold value to distinguish the FITC positive signal from the background signal was determined and the percentage of

EREC
Antibodies were incubated directly in the transwells for 1h at 37°C. Cells were first incubated with a 1:1,000 dilution of a polyclonal rabbit anti-IEP antibody, kindly provided by Dr. D.
O'Callaghan, Louisiana State University, USA. The diluent used was PBS containing 10% negative goat serum. This was followed by incubation with a goat anti-rabbit IgG FITC ® conjugated antibody (Invitrogen). Nuclei were counterstained with Hoechst 33342 for 10min 6 at 37°C. Transwell membranes were excised from the culture inserts and mounted on glass slides using glycerol-DABCO. Slides were examined using a Leica confocal microscope. The total number of plaques was counted on 5 random fields of approximately 3•10 4 cells per insert.
Plaque latitude was measured on 10 individual plaques using the Leica confocal software package.

Enzymatic removal of cell surface N-linked glycans and sialic acids prior to EHV1
inoculation PNGase F (New England Biolabs, Ipswich, UK) removes complex, hybrid and oligomannose N-glycosylations and was applied onto apical or basolateral EREC surfaces for 12h at a concentration of 25,000 U/mL, diluted in EREC medium, supplemented with 10% glycobuffer (New England Biolabs; Ipswich; UK). Neuraminidase from Vibrio cholera (Sigma-Aldrich) has a broad substrate spectrum for sialic acids and was used for 1h at 50 mU/mL in PBS. The Correct cleavage of the respective sialic acids was corroborated by immunofluorescent staining of EREC with biotinylated Maackia Amurensis Lectin II (Vector laboratories). The complex was subsequently stained with Streptavidin-FITC ® (Invitrogen). Ten z-stack confocal pictures were taken to distinguish apical from basolateral treatment. The means of the fluorescent apical 7 or basolateral signals were compared with Image J software and dropped significantly after treatment with neuraminidase, compared to control. Following enzymatic treatment, cells were washed 3 times with DMEM/F12 and the inoculum was delivered on top of the respective surfaces for 1h at 37°C. Concurrently, and as a positive control for enzymatic treatment, MDCK cells were treated with neuraminidase before inoculation with EIV. After 1h inoculation, unbound virus particles were removed by washing and cells were incubated for 10 hours before fixation in methanol, as described above. Figure S1. Cell viability in respiratory mucosal explants TUNEL-staining data of tracheal ME (left) and nasal ME (right) after different treatments.

Supplementary figures
Three independent experiments were performed and data are represented as means + SD. The lower case letters indicate significant (P<0.05) differences in the epithelium, while the upper case letters indicate significant differences in the lamina propria.