All respiratory samples positive for coronavirus by real-time RT-PCR were screened for other respiratory viruses. Among the real-time RT-PCR CoV positive samples, 13% were also positive for Respiratory Syncytial Virus, Parainfluenza 2 and 3, and Rhinovirus (Data not shown). Among the patients harboring respiratory CoVs, detected by qRT-PCR and confirmed by RT-PCR, six (17%) were co-infected with a second respiratory virus.

The nucleic acid was subjected to multiplex amplification for all specimens using SureX 13 Respiratory Pathogen Multiplex Detection Kit (Cat. No. 1 060 144, Ningbo Health Gene Technology) on ABI GeneAmp PCR System 9700 (Thermo Fisher Scientific). The 13 respiratory pathogens were as following: influenza A virus, influenza A virus H1N1 (2009), seasonal H3N2 influenza virus, influenza B virus, adenovirus, boca virus, rhinovirus, parainfluenza virus, chlamydia, human metapneumovirus, Mycoplasma pneumoniae, coronavirus, and respiratory syncytial virus. The PCR product was subjected to capillary electrophoresis using GenomeLab™ GeXP Genetic Analysis System (Beckman Coulter) according to the instructions. Each pathogen, if detectable, produced a distinctive fragment size after PCR amplification. The results of fragment analysis were used to determine the outcomes of testing. In brief, if the peak height of a targeted fragment size is lower than the lower peak of the signal standard, the targeted pathogen is determined negative; if the peak height of a targeted fragment size is higher than the higher peak of the signal standard, the targeted pathogen is determined positive; if the peak height of a targeted fragment size is between the higher and the lower peaks of the signal standard, the targeted pathogen is determined uncertain and the test should be repeated.

A total of 420 oropharyngeal swabs were enrolled from 10 hospitals and 10 CDCs in Guangzhou from 2017 to 2018. Samples were collected from a wide range of ages, with the average age of 27.2 (Table 1). About 55% specimens were from male.

A pathogen‐positive result was determined when the pathogen‐specific fragment(s) was positive, as shown in Figure 1. A negative result was determined when none of the 13 pathogen‐specific fragment was positive, while the controls (huDNA, huRNA, and IC) were positive (Figure 2). In this study, the ResP detected positive results in 141 samples, accounting for 33.6%, while the comparator tests detected positive results in 127 samples, with positive rate 30.2%. Among the detected pathogens, rhinovirus was the most common, followed by adenovirus and influenza virus A pdmH1N1 (2009) (Table 2). Of the 420 specimens, the ResP yielded consistent positive results in 121 specimens (86.5%, 121/141), and consistent negative results in 273 specimens (97.8%, 273/279) comparing with pathogen‐specific PCRs, leading to an overall agreement of 93.8%.

No specimen was detected positive with coronavirus or Chlamydia. In six of the ten detected pathogens, the Cohen's kappa values were over 0.8 with P value <.01 (Table 2). The lowest kappa (0.70) was observed on human metapneumovirus.

When qRT-PCR was used to screen the samples for CoV, 79 out of 200 (39.5%) samples were positive. Thirty-five out of 200 (17.5%) were confirmed to be positive for coronavirus by RT-PCR, using primers for the ORF1b, N and/or S regions (Tables 2 and 3). Twenty-four samples were amplified using primers specific for the N or S regions of OC43, 5 for NL63, 3 for HKU1, 1 for 229E and 1 for feline-like (Table 3). Among those samples, 6 were amplified only after the filtered nasal aspirates were passed into cell culture; 5 of them were passed into Vero cells and 1 was passed into HRT-18 cells. All of them were amplified using primers to the S region of OC43.

The age distribution of coronavirus positive samples confirmed by RT-PCR is shown in Table 2. High incidence of respiratory CoV occurred in children up to 10 years of age (17.1%) and in adults above 40 years of age (11.4–25.7%). Gender distribution of coronavirus positive samples confirmed by RT-PCR showed a higher incidence of positive samples in females (13.0%) versus males (4.5%) (P<0.05).

Nasopharyngeal (NP) and oropharyngeal (OP) swabs for detecting viruses and blood cultures for detecting bacteria were taken from eligible patients on admission using operating procedures described by the WHO. Specimens were taken an average of 7 days after illness onset (range: 1–66 days).

Total nucleic acid (TNA) was extracted by the automated KingFisher Flex Magnetic Particle Processor (Thermo Scientific, Waltham, MA, USA) using MagMAX Total Nucleic Acid Isolation Kit (Cat No. AM 1840, Applied Biosystems, Foster, CA, USA) according to the manufacturer’s instructions. The viral target was amplified using specific primers and probes produced by the CDC (Atlanta, GA, USA) and following standard protocol for reverse transcription polymerase chain reaction detection. From 2010 to 2012, testing for RSV, adenovirus, human parainfluenza viruses (hPIV) 1, 2 and 3, influenza (A and B) and human metapneumovirus was conducted at CUH laboratory and sent for confirmation by the Naval Medical Research Unit No.3 (NAMRU-3) laboratory. From 2013 to 2014, testing was conducted at CUH laboratory. For all samples, the human RNase P gene (RP) was tested as an internal positive control to ensure proper sample collection and nucleic acid extraction. Samples were considered positive to the viral target if the amplification curve crossed the threshold line before cycle 40. All clinical samples should be positive to RP with cutoff value ≤ 37, as prescribed previously. Blood samples were collected for detection of Mycoplasma pneumonia, Chlamydia pneumonia, and Legionella pneumophila, using RT-qPCR.

Prior to study initiation, the study protocol was reviewed and approved by Institutional Review Board at the NAMRU-3, as well as the ethical committee of CUH, in compliance with all applicable federal U.S. regulations governing the protection of human subjects. An informed written consent was obtained from the patients (in the case of adult patients) or patients’ parent/legal guardian (in the case of pediatric patients).

Nucleic acids were extracted from biological samples (nasopharyngeal swab) with Qiamp RNA Virus Kit (Qiagen) following the protocol provided by the manufacturer. Multiplex real-time PCR assays were performed with FTD Respiratory 21 PLUS pathogens panel (Fast-Track Diagnosis, Luxembourg). This kit allows the identification of major respiratory pathogens (2 atypical bacteria and 18 viruses) including: Influenza viruses A and B (FLUAV and FLUBV) and Influenza A virus subtype H1N1 2009 [FLUAV (H1N1/pdm09)], human coronaviruses: NL63 (HCoV-NL63), 229E (HCoV-229E), OC43 (HCoV-OC43) and HKU1 (HCoV-HKU1), human parainfluenza (HPIV-1, −2, −3, −4), human metapneumovirus A and B (HMPV A/B), human rhinovirus (HRV), respiratory syncytial virus A and B (RSV A/B), human adenovirus (HAdV), enterovirus (EV), parechovirus (PV), human bocavirus (HBoV), Mycoplasma pneumoniae (Mpneu), and Chlamydia pneumoniae (Cpneu). Internal control was added to each analysis. All PCR assays were performed with the AgPath-IDTM One-Step RT PCR kit (Ambion, cat#AM1005) as recommended by FTD company. For some analyses, the following pathogens were grouped: HCoV-OC43, HCoV-HKU1, HCoV-229E and HCoV-NL63 in human coronaviruses (HCoV group); FLUAV, FLUBV, (FLUAV/H1N1/pdm09) in Influenza virus (FLUV group); HPIV−2, HPIV−3 and HPIV−4 in human Parainfluenza virus (HPIV group).

Associations between clinical manifestations and etiological agents were analyzed with Fisher Exact test, considering a P-value <0.05 as significant value. Statistical analyses were calculated with EpiInfo software (version 3.5.1).

All environmental items were additionally tested for the presence of replicating virus using CRFK cell culture. No virus replication could be detected on CRFK cells as CPE was not observed at any time point for the tested samples. The supernatant from four samples tested weakly positive: The water samples collected on days 3 and 8 (Ct-values 39.3 and 38.3, respectively) and the samples collected from the two food bowls on day 3 (Ct-values 37.2 and 36.2, respectively). All other supernatants were FCV negative by RT-qPCR.

Clipped hair from either the right or left front leg of one cat (JJH3) collected on days 56, 71 and weekly thereafter until day 106 was analyzed by RT-qPCR and CRFK cell culture (Table 2). This cat had been shedding FCV until day 71. Only on day 85, the hair sample tested positive by direct RT-qPCR (Ct-value 38.7); all other samples were negative, and none of the samples tested in cell culture yielded CPE or RT-qPCR positive supernatant (Table 2). No samples from early infection were available.

Mice were euthanized and bronchoalveolar lavage (BAL) fluid and lung cells obtained as in Price et al., 2009. Briefly, for BAL fluid, lungs were flushed with 1 ml phosphate-buffered saline (PBS). Lung cells were isolated by gradient centrifugation of minced and collagenase-digested lung tissue.

Splenocytes were depleted of erythrocytes by treatment with ACK lysis buffer. Sera from blood collected from the abdominal vena cava were isolated using BD Microtainers (Franklin Lakes,NJ), and decomplemented by heat-treating at 56°C for 30 minutes.

Comparisons between the groups were performed with the non-parametric Mann-Whitney U test. Spearman rank correlation was used to evaluate the correlation between variables. Results are expressed as median and 25th to 75th percentiles. The statistical analysis was performed using Statistica (StatSoft, USA); p values < 0.05 were accepted as statistically significant.

The study was approved by the local Bioethics Committee, and all study subjects provided their informed consent.

Viral DNA or RNA was extracted from sputum using a QIAamp® MinElute® Virus Spin Kit (Qiagen), according to the manufacturer’s instructions. To synthesise cDNA, reverse transcription was performed using a RevertAid H Minus First-Strand cDNA Synthesis Kit (Thermo Fisher Scientific Inc., Waltham, USA). Each cDNA was subjected to multiplex PCR using a SEEPLEX RV15 ACE Detection Kit (Seegene), according to the manufacturer’s instructions. The multiplex PCR detected rhinoviruses (HRV) and enteroviruses (HEV), parainfluenza viruses 1 (PIV1), 2 (PIV2), 3 (PIV3), and 4 (PIV4), influenza viruses A (FluA) and B (FluB), respiratory syncytial viruses A (RSVA) and B (RSVB), coronaviruses (229E/NL63 and OC43), adenovirus (AdV), human metapneumovirus (MPV), and human bocaviruses 1/2/3/4 (HBoV). The multiplex PCR products were visualised by electrophoresis on 2% agarose gel.

At all sites, 78 out of 123 (63.4%) workers invited to participate agreed, and their nasal wash samples were collected. Further, 55 pig fecal, 49 pig oral or water, and 45 bioaerosol samples were collected across these sites (S2 Table). Of these, 21 (38.2%; average CT 35.09) pig fecal, 43 (87.8%; average CT 32.82) pig oral or water, 3 (14.2%; CT 32.05, 34.20, and 36.19) bioaerosol, and 4 (5.1%; CT 30.60, 36.00, 36.32, and 37.30) worker nasal wash samples were positive for PCV2 by rPCR (Table 1). PCV2 was detected in at least one sample at all 11 farms, as well as in a bioaerosol sample at one abattoir and one worker nasal wash at a market (Fig 1). Porcine RVC was detected in one pig fecal sample (1.8%; CT 28.31). EMCV and porcine RVA were not detected in any of the sample types tested (Table 1). ADV was detected in three (3.8%; CT 30.45, 32.20, and 35.03), CoV was detected in two (2.6%; CT 25.50 and 37.84), and IBV (CT 27.66) and IDV (CT 16.18) were each detected in one (1.3%) worker nasal wash sample. IAV was detected in two (4.1%; CT 35.19 and 38.48) pig oral secretions, and EV was detected in two (15.4%; CT 36.60 and 36.71) bioaerosol samples. ICV was not detected in any of the sample types tested (Table 2). Of the 16 sites, 13 had at least one virus detected in any sample type, and 8 had positive results for more than one virus (S3 Table).

While virus-positive pig oral secretions were not associated with the potential risk factors of pig breed and production type, there were associations with PCV2 detection in feces. The local pig breed, babi kampung, exhibited a higher prevalence of PCV2 results compared to standard domestic breeds (OR = 10.55, 95% C.I. 1.65, 110.87). Additionally, there was a nonstatistically significant trend towards an association of PCV2 in the feces for sows or gilts (breeding stock) as compared to other pig types (OR = 6.25, 95% C.I. 0.37, 353.48) (Table 3).

For infection and cytotoxicity assays Madin-Darby Canine Kidney Epithelial Cells (MDCK II, NBL-2, CCL-34) were used (ATCC). Cells were cultivated under standard cell culture conditions with DMEM (supplemented with 10% fetal calf serum (FCS), 2 mM L-glutamine) in humid atmosphere at 37°C and 5% CO2. Hemagglutination inhibition assays (HAI) were performed using either human erythrocytes (α-2,6’-sialosugars) for Aichi H3N2, or turkey erythrocytes (α-2,3’-sialosugars) for Rostock H7N1.

The field of laboratory-acquired viral infections has progressed since Sulkin and Pike’s seminal report in 1949, but the problem has yet to be completely eliminated. The paucity of experimental data is an important reason for this. Our study, presented herein, provides the first evidence of laboratory procedures that generate viral contamination that can potentially infect laboratory personnel.

In fact, all of the studies performed to date by others to determine whether laboratory studies of pathogens, including viruses, bacteria and fungi, could generate infective aerosols have been based on statistical analysis of actual infections. The influenza viruses present a particular challenge to monitoring in the lab by traditional methods, since their small size and low concentration make them difficult to collect and detect from the air. Most of the bioaerosol samplers currently on the market are not suitable for collection of viruses. Furthermore, effective detectors need to be able to conveniently collect air samples at various time points over the course of an entire experiment since virus-containing aerosols may be generated by any number of steps in the process and then fall from the air prior to an end-of-procedure sampling time point. The currently available samplers typically run for short periods of time (minutes), making it difficult to capture large volumes or integrate sample collection over time. We designed a collector terminal with six air pollution sampling probes that are able to continuously sample the air for ≥10 h; in addition, this device is portable and can be easily moved around the laboratory to locales where different steps of the experiment are performed. The feature of remote control allows for simultaneous or sequential collection in different locales from two or more of the probes. In the study described herein, to test the capability of such a detector, the aerosols collected during the experiment reflected the virus produced in most operations; this study was not designed to clearly delineate a complete picture of all potential applications of this device.

Measuring infectious virus from air samples is logistically difficult. Few researchers have reported airborne influenza virus from the laboratory, and even fewer have detected the infectious capacity of influenza viruses sampled from air. Traditional methods of determining the presence of virus in an aerosol include directly applying the concentrated aerosol to infect a host cell, or as template in Rt-PCR to detect virus-specific genes. One of the most commonly used methods is Rt-PCR because of its remarkable sensitivity and specificity; however, Rt-PCR is unable to determine whether the detected virus is infectious. In our study, we combined experimental approaches that would determine both the concentration of virus in aerosols and the activity of those viruses. By exploiting the functional character of the hemagglutinin of influenza virus, we were able to estimate the concentration of influenza virus in aerosols according to the amount of adsorption that occurred with red blood cells (RBCs). By also using the aerosols to directly inoculate chick embryos, we were able to determine the ability of the aerosolized virus to proliferate in vivo, indicating the infectivity of the contaminating pathogens. Future studies will aim to determine the feasibility of this approach for quantifying virus in the aerosol.

This study, to our knowledge, represents the first successful attempt to directly detect influenza virus-containing aerosols generated by routine laboratory procedures. Our results provide evidence that many of the laboratory techniques used to process influenza virus for experimental analysis produce aerosols and, thereby, represent significant risks of infection to laboratory personnel and potential spread beyond the laboratory. Working in the laboratory is inevitably dangerous, but nearly all risks can be sufficiently minimized by GLP and careful monitoring of risk factors, such as presence of pathogen-containing aerosols. According to the results of this study, we plan to extend our investigations to emergency operating procedures that follow laboratory accidents with virus samples and to generate more effective strategies to prevent laboratory-generated aerosols and laboratory-acquired infections or spread.

A total of 517 serum samples from wild-caught cynomolgus macaques were collected from 2010 to 2017 in Singapore. An initial screening was performed using a PRV3M whole virus-based ELISA. By using three standard deviations of the geometric mean as a cut-off, 67 serum samples were identified as reactive to PRV3M (Figure 2(a), Table 1). To further confirm the positivity of the samples and increase specificity, a serum neutralization test against PRV3M was performed on all ELISA positive serum samples. A total of 34 serum samples were able to neutralize PRV3M at a dilution of at least 1:10 (Table 1). To further confirm the specificity, Western blot analysis was conducted using purified virions. Shown in Figure 2(b) is a representative Western blot conducted with sample #6301, a known positive sample with a high neutralizing antibody titre, and a negative control sample, both used at dilution of 1:500. The Western blot analysis not only confirmed findings from ELISA and VNT, but also revealed that the major outer capsid protein (mu1) was the most reactive viral protein (Figure 2(b)).

Our experiment was carried out in a negative pressure isolation unit with mechanical arm (Figure 3).

The three groups of procedures with simulated accidents (broken glass containers with influenza virus suspension, syringe-ejected influenza virus suspension, and centrifuge tube rupture) all produced aerosols. The controls for each were all negative. All of the aerosols that were produced contained H5N1 that was detectable by Rt-PCR and HA text (Table 1, Figure 4).

Semiquantitative histopathological assessment of influenza virus-associated inflammation in the lung was performed as reported earlier (23), with the following modifications. For the extent of alveolitis, we scored the total number of alveoli affected in the lung slide as follows: 0, 0%; 1, 1 to 25%; 2, 25 to 50%; and 3, >50%. For the severity of alveolitis, bronchiolitis, and bronchitis, we scored the slides as follows: 0, no inflammatory cells; 1, few inflammatory cells; 2, moderate numbers of inflammatory cells; and 3, many inflammatory cells in all alveoli, bronchioles, and bronchi of the lung slide. For the presence of alveolar edema and the presence of alveolar epithelial necrosis, we reported scores as follows: 0, not present; and 1, present in the lung slide. For the overall score for pathological changes in the alveoli, the total score (scores for extent of alveolitis plus severity of alveolitis plus presence of alveolar edema plus presence of alveolar epithelial necrosis) was used. Slides were examined without knowledge of the identities of animals.

Quantitative PCR (qPCR) was performed using QuantiTect Probe RT–PCR Kit (Qiagen) reagents and with the CFX96 Real-Time System (Bio-Rad). Each 25 µL PCR reaction contained 12.5 µL 2X QuantiTect PCR master mix, 1 µL of each 10 mM primer, 0.5 μL 0.2 mM probe, 0.5 μL reverse transcriptase, 3 µL RNA template and 6.5 µL H2O. Every PCR was performed as follows: reverse transcription at 50°C for 10 min, initial PCR activation at 95°C for 5 min and 45 amplification cycles consisting of a 95°C denaturation for 10 sec and a 60°C annealing/extension for 30 s. Sequences of primers and probes are as follows; PRV3M-S4-F: 5′-CAT TGT CAC TCC GAT TAT GG -3′, PRV3M-S4-R; 5′- TGG GAG GGT GCA GAG CAT AG -3′, PRV3M-S4-probe5′- /56-FAM/ GTA GGC ATG CCG CTC GTG GAA TCC A /3BHQ_1/ -3′. Each PCR was performed in duplicate to obtain an average Ct for each sample. Amplicons were quantified by plotting the Ct values against standard curves made using 10-fold dilutions of cDNA produced from in vitro transcription RNA samples. Data were expressed as molecules of RNA.

Univariate logistic analysis was performed with all variables in Tables 1, 2, and 3 for M. pneumoniae CAP, and the following variables with P < 0.2 were selected for multivariate logistic regression analysis: age, coexisting disease, smoking, fever, max temperature, neutrophil and C-reactive protein (Table 4). Through the multivariate logistic analysis, two independent factors with P < 0.1 associated with M. pneumoniae CAP were identified, namely age (OR 0.94, P < 0.05) and coexisting disease (OR 0.33, P < 0.1) (Table 5). Subsequently, discriminatory analysis and ROC curves were used to establish and assess a predictive diagnostic model of CAP with these two independent factors. The predictive diagnostic model of M. pneumoniae CAP included the characteristics of being younger than 45 years of age and not having a coexisting disease, with an AUC of 0.61 (95% CI: 0.53 to 0.69). The sensitivity and specificity of this model were 54.9% (95% CI: 51.2% to 64.8%) and 58.0% (95% CI: 43.4% to 66.5%), respectively (P < 0.05).

Influenza strain A/Aichi/2/68 H3N2 X31 (Aichi H3N2), reassorted with A/PuertoRico/8/1934 H1N1, and low pathogenic A/Mute Swan/Rostock/R901/2006 H7N1 K3141 (Rostock H7N1) were harvested from allantoic fluid of hen eggs. Virus isolates were clarified upon low speed centrifugation (300 x g, 10 min) and concentrated by ultracentrifugation (100 000 x g, 1 h). For safety reasons, viruses prepared for SPR experiments were inactivated by 5 min irradiation with UV-light on ice. Infectivity of inactivated virus was precluded using MDCK II based cell-assays, remaining binding ability of HA was proven using standard hemagglutination assay (HA) with human red blood cells. For SPR measurements, also monovalent split vaccine influenza strains A/NewYork/55/2004 H3N2, NIH accession No. ABO37541 (New York H3N2), A/Victoria/210/2009 H3N2, NIH accession No. AFM71802 (Victoria H3N2) and the split vaccine Pandemrix (GlaxoSmithKline) of influenza strain A/California/7/2009 H1N1, NIH accession No. ACP44189 (California H1N1) were used. Protein concentrations were determined using standard BCA assay (Thermo Fisher Scientific). Additionally viruses were titrated for HA units. Here, 50 μl virus concentrate was serially diluted twofold in PBS using 96-well microtiter plates. Then, 50 μl of 1% human red blood cells (German Red Cross, Berlin, AB+) were added to each well, followed by an incubation of 60 min at room temperature. The last well showing hemagglutination provided the HA units per 50 μl virus solution. For the Rostock H7N1 strain turkey erythrocytes have been used. Additional material used was HA from Aichi H3N2 (Sino Biological, 11707-V08H).

Lungs were prepared and immersion fixed for 24 h in 4% buffered formaldehyde solution (pH 7.4), dehydrated in a graded ethanol series, and embedded in paraffin. Sections (0.5 μm) were cut from three evenly distributed levels of paraffin blocks and stained with hematoxylin and eosin. For immunohistochemical studies, sections were stained overnight at 4°C with a primary antibody against influenza A virus nucleoprotein (clone hb65; ATCC, Wessel, Germany). Subsequently, tissue sections were incubated for 30 min with horseradish peroxidase (HRP)-labeled goat anti-mouse IgG2a (Biozol) and counterstained with hematoxylin (22). Semiquantitative assessment of influenza virus antigen expression in the lungs was performed as reported earlier (23), with the following minor modifications. For the alveoli, 25 arbitrarily chosen 20× objective fields of lung parenchyma per lung slide were examined by light microscopy for the presence of influenza virus nucleoprotein, without the knowledge of the identities of the animals. The score for each animal was presented as the percentage of positive fields. For the bronchi and bronchioles, the percentage of positively staining epithelium was estimated for every slide to provide the score per animal, as follows: 0, 0% staining; 1, 1 to 25% staining; 2, 25 to 50% staining; and 3, >50% staining. For immunofluorescence analyses, 12-μm-thick cryo-sections were air dried, fixed in acetone at −20°C, and rehydrated in PBS. Slides were blocked with anti-FCR (1:500; purified rat anti-mouse CD16/CD32). Primary antibodies used were AF350-conjugated rabbit anti-TMPRSS4 polyclonal antibody (Bioss) and Cy5-conjugated rabbit anti-TMPRSS2 antibody (Antikoerper-Online). Both antibodies were checked on appropriate knockout slides to exclude potential cross-reactivity. After staining, the slides were washed with PBS, dried, and mounted with Neo-Mount (Merck, Darmstadt, Germany). Analyses were performed using a Zeiss LSM510 laser scanning microscope with a 40× oil immersion objective. For in situ hybridization (ISH) studies of type I alveolar epithelial cells (AECI) and type II alveolar epithelial cells (AECII), 5-μm-thick formalin-fixed, paraffin-embedded lung tissue sections were used with a QuantiGene ViewRNA ISH tissue assay kit (Affymetrix, Cleveland, OH) following the manufacturer's instructions. Type 6 (Sftpc; for AECII) and type 1 (Aqp5; for AECI) QuantiGene ViewRNA probes were generated based on Affymetrix probe sets. The GenBank accession numbers for the probe sequences are as follows: for type 6 probes (Sftpc), NM_011359 (region 2–784); and for type 1 probes (Aqp5), NM_009701 (region 239–1533). Pretreatment was performed for 10 min at 90 to 95°C, and protease digestion was performed for 20 min at 40°C. Slides were stained with Fast red for type 1 probes. Type 6 probes were detected with Fast blue. Slides were then counterstained with Gill's hematoxylin. Negative controls (antisense) were run in parallel and showed no hybridization signals (not shown) (type 6 [Sftpc-neg] negative control [region 23–803], covered by probe set NM_011359-N; and type 1 [Aqp5-neg] negative control [region 172–1122], covered by probe set NM_009701-N). Bright-field images were recorded using Aperio Scanscope and analyzed with Aperio Imagescope software. For detection of fluorescence images, a Zeiss LSM710 laser scanning microscope with a 20× objective was used.