Introduction

Human norovirus is an important cause of epidemic and sporadic gastroenteritis worldwide. This virus can induce severe in young children, the elderly individuals, and the immunocompromised persons1. Norovirus is a non-enveloped, positive-sense, single-stranded RNA virus belonging to the family Caliciviridae. It is transmitted via the fecal-oral route through contact with infected individuals and contaminated environments2. Due to their small size, norovirus particles presumably become aerosolized from vomit3 or toilet flushing4,5. Further susceptible people can be infected after they swallow the virus. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is an emerging coronavirus that causes coronavirus disease 19 (COVID-19). SARS-CoV-2 is an enveloped, positive-sense, single-stranded RNA virus of the genus Betacoronavirus, subgenus Sarbecovirus. The virus is transmitted via aerosols, droplets, and contaminated objects6. Previous studies have shown the presence of SARS-CoV-2 in indoor air samples collected from hospital settings7,8,−9. Human bocavirus (HBoV) is a parvovirus associated with acute respiratory infections in young children. HBoV is a small non-enveloped, single-stranded DNA virus, a member of the family Parvoviridae. HBoV1 is most likely transmitted via the respiratory route and it causes respiratory illness. Meanwhile HBoV2, HBoV3 and HBoV4 are mainly detected in human feces and are transmitted via the fecal-oral route10. Human bocavirus is detected in aerosols in hospitals11,12. Another study has examined the prevalence of HBoV, both as a single infection and as a co-infection with other viruses13.

Bioaerosol transmission is generally recognized as a transmission mode for respiratory viruses. However, the role of bioaerosols in viral spread is limited for gastrointestinal viruses. In controlled laboratory conditions, the norovirus RNA and SARS-CoV-2 RNA last in aerosols for up to 2–4 and 3 h, respectively14,15. A previous study investigated the presence of norovirus in aerosols for a specific time16. The concentration of SARS-CoV-2 RNA in aerosols was low in isolation wards and ventilated patient rooms but was relatively high in the toilet areas used by the patients17.

The viral load of aerosols is an important factor in determining the possible contribution of airborne transmission. However, the appropriate sampling and detecting methods of virus-laden aerosols are challenging to implement due to the low concentration levels of viruses in the air and the virus inactivation that may occur during sampling and processing. Thus far, numerous bioaerosol sampling methods for detecting norovirus and SARS-CoV-2 have been proposed. However, they have their own advantages and disadvantages16,18. Air sampling protocols have not been standardized. Therefore, to elucidate the presence and behavior of virus in aerosols, technological developments are still ongoing.

In Thailand, norovirus gastroenteritis, the emerging COVID-19 and acute respiratory disease caused by HBoV remain a public health issue19,20,−21. In addition, these enteric and respiratory viruses were detected in food and environmental samples. This finding indicates the importance of environmental contamination and potential vehicle transmission22,23,−24. However, there are no data on these viruses in air samples. This study aims to assess the presence of norovirus, SARS-CoV-2 and HBoV in air samples collected in water using the virus concentrating method developed in our laboratory14 and the molecular methods for the quantification and strain identification of the virus in aerosols.

Materials and methods

Aerosol samples

A cross-sectional study was performed between March 2023 and May 2023. In total, 60 air samples were collected from a tertiary care hospital in Bangkok, Thailand. The air samples were collected from 30 different areas of the indoor environments e.g. medical ward, internal medical ward, COVID-19 cohort ward, cohort ward patients under investigation (PUI), gynecology clinic, family planning clinic, ophthalmology clinic, and other areas including medical record department, and elevator hall (Supplementary Table S1). Most air samples were collected in front of the clinic, ward, department, and hall which had different size. Two air samples were collected from each area or sampling site in sterile distilled water with 5 and 20 mL collection vessels using the BioSampler (SKC Inc., Eighty-Four, PA, the USA) at a flow rate of 12.5 L/min for 30 min at a height of 1 m above the ground.

Virus concentration method

The water in the collection vessels was processed by speedVac centrifugation25. The volume of water was reduced using a vacuum centrifuge (UNIEQUIP Laborger ätebau- und-Vertriebs GmbH, Munich, Germany) to approximately 600 µL for 2 and 4 h with 5 mL and 20 mL of water collected, respectively. The air samples in the water concentrates were stored at −80 °C until used.

Nucleic acid extraction

Viral RNA and DNA were extracted from 200 µL each of the water concentrate from the air sample using the QIAamp Viral RNA Mini Kit and the QIAamp Viral DNA Mini Kit (Qiagen, Hilden, Germany), respectively, based on the manufacturer’s protocols.

RT-qPCR for the quantification of norovirus RNA

All 60 air samples were tested for the presence of norovirus GI and GII RNAs using quantitative reverse transcription-polymerase chain reaction (RT-qPCR). The quantitative detection of norovirus RNA was performed using primers, probes, and procedures as described by Rupprom et al.14. In total, 5 µL of undiluted or diluted RNA extract (1:4) in nuclease-free water was added into 15 µL of reaction mixture with LightCycler RNA Master Hybprobe (Roche Diagnostics, Mannheim, Germany) in separate tubes for norovirus GI and GII. Primers GITF and GITR, and probe GIT-TP for the GI or QNIF2 and COG2R, and probe QNIFs for GII were used. RT-PCR reactions were conducted in the LightCycler 96 Real-Time PCR System for GI or GII at 45 cycles of amplification. The norovirus genome copy numbers were quantified by comparing quantification cycle (Cq) values obtained from a standard curve using 10-fold serial dilutions of the GI or GII RNA transcript (103–107 RNA copies/mL). The norovirus GI or GII RNA concentration was calculated and presented as genome copies per mL and genome copies per m3 of air volume.

RT-qPCR for detecting SARS-CoV-2 RNA

All 60 air samples were tested for the presence of SARS-CoV-2 RNA using RT-qPCR. The qualitative detection of SARS-CoV-2 RNA was performed using primers and probe targeting a region of the N-gene coding for nucleocapsid protein26. In total, 5 µL of undiluted or diluted RNA extract (1:4) in nuclease-free water was added into 15 µL of reaction mixture comprising 1× LightCycler RNA Master Hybprobe, 3.25 mM Mn(OAc)2, 0.5 µM primer NIID_2019-nCOV_N_F2 (5′-AAATTTTGGGGACCAGGAAC-3′), 0.7 µM primer NIID_2019-nCOV_N_R2 (5′-TGGCAGCTGTGTAGGTCAAC-3′), 0.3 µM probe NIID_2019-nCOV_N_P2 (5′-FAM-ATGTCGCGCATTGGCATGGA-BHQ-TAMRA-3′), and PCR grade water. The reactions were conducted in the LightCycler 96 Real-Time PCR System. The PCR cycling conditions included 1 cycle of reverse transcription at 55 °C for 30 min; 1 cycle of initial denaturation at 95 °C for 5 min, then 45 cycles of denaturation at 95 °C for 15 s and annealing/ extension at 58 °C for 1 min. A Cq value of < 45 indicated positivity for SARS-CoV-2. A standard curve of the synthetic SARS-CoV-2 DNA plasmid was generated from 10-fold serial dilutions ranging from 10− 6 to 10− 10 ng/µL. The concentrations of SARS-CoV-2 DNA plasmid in the experiments were determined from the established standard curve.

RT-nested PCR for detecting norovirus

The RNA extract from 60 air samples was tested for the presence of noroviruses GI and GII using RT-nested PCR as described by Kittigul et al.22. A 2 µL of the RNA extract was added to the RT-PCR mixture (48 µL) using the SuperScript III One-Step RT-PCR System with Platinum Taq DNA polymerase (Invitrogen, Carlsbad, CA) in separate tubes for norovirus GI and GII. Primer GI (COG1F and G1-SKR) or GII (COG2F and G2-SKR) was used for RT-PCR. Then, primer GI (G1-SKF and G1-SKR) or GII (G2-SKF and G2-SKR) was used for nested PCR. The amplification reaction was performed on the BIO-RAD T100 Thermal cycler (Applied Biosystems, Foster City, CA, the USA) with established thermocycling profiles. By agarose gel electrophoresis, the amplicon sizes were found to be 330 bp, which is expected for norovirus GI, and 344 bp, which is expected for norovirus GII.

Nested PCR for detecting bocavirus DNA

The HBoV DNA in the nucleic acid extract from 60 air samples was tested using nested PCR, as described in a previous study24. A 5 µL DNA was mixed with 20 µL of reaction mixture (Invitrogen, Carlsbad, CA, the USA). Primers 2028F and 2029R were used for PCR. Then, primers 2030F and 2031R were used for nested PCR. Nucleic acid amplification was conducted on the BIO-RAD T100 Thermal cycler with PCR conditions. The amplicon size for HBoV was 382 bp.

Phylogenetic analysis

The RT-nested PCR products obtained from the norovirus-positive air samples were subjected to DNA sequencing. The nucleotide sequences of norovirus were initially genotyped by the norovirus genotyping tool (https://www.rivm.nl/mpf/typingtool/norovirus/). The nucleotide sequences were compared with those of the norovirus reference strains available in the GenBank database using the BLAST server. Phylogenetic analysis of norovirus was performed using MEGA 11.0. Two nucleotide sequences of norovirus coded A09-20 and A28-5 were submitted to the GenBank database with the accession numbers PP843640 and PP843641, respectively.

RT-PCR inhibition test

RT-PCR inhibition for norovirus GI, GII and SARS-CoV-2 was tested using water concentrates from the collected air samples with undiluted and 1:4 diluted RNA extract. In total, 1 µL aliquot of external control (EC) (norovirus GI RNA transcript [104 RNA copies/µL], GII RNA transcript [104 RNA copies/µL], or SARS-CoV-2 DNA [10− 7 ng/µL]) was added to 5 µL of the undiluted or 1:4 diluted RNA extract (sample RNA + EC RNA). Then, 1 µL of EC was also added to 5 µL of PCR-grade water (water + EC RNA). RT-qPCR was performed for norovirus GI, GII and SARS-CoV-2 in separate tubes. RT-PCR inhibition was calculated using the following equation:

RT-PCR inhibition = (1−10ΔCq/m) × 100%,

where ΔCq = Cq value (sample RNA + EC RNA) − Cq value (water + EC RNA) and m = slope of the standard curve. The acceptable level of RT-PCR inhibition was ≤75% based on the result analysis27.

Results

Detection of noroviruses and SARS-CoV-2 in air samples

During air sampling periods in a hospital, the temperature and relative humidity were 24.8 °C ± 2.7 °C (range 20.9–32.3 °C) and 66.3% ± 6.3% (53–79%), respectively. In total, 60 aerosol samples were collected from different areas. Next, the presence of noroviruses was simultaneously assessed using RT-qPCR and the RT-nested PCR approaches. In separate experiments, all 60 samples were examined for SARS-CoV-2 by RT-qPCR and HBoV by RT-nested PCR. Based on RT-qPCR, norovirus RNA and SARS-CoV-2 RNA were detected in 13/60 (21.7%) and 3/60 (5.0%) samples, respectively, with the 5 and 20 mL collection vessels. One air sample presented with a weak positivity for norovirus and SARS-CoV-2 RNAs. Some norovirus-positive samples could be quantified as genome copies. However, the three air samples had a weak fluorescence signal for SARS-CoV-2, and their Cq value could not be obtained. By RT-nested PCR, norovirus RNA was detected in two (3.3%) samples with the 5 and 20 mL collection vessels. All air samples in 5 and 20 mL collection vessels from the same collection sites tested negative for HBoV DNA (Table 1).

Table 1 Test of respiratory and gastrointestinal viruses in air samples.
Full size table

Table 2 shows the norovirus genogroups detected by either RT-qPCR or RT-nested PCR. Using RT-qPCR, norovirus GII (8/60, 13.3%) had a higher detection rate than norovirus GI (4/60, 6.7%). One air sample (1/60, 1.7%) tested positive for the GI and GII noroviruses. Based on RT-nested PCR, GII was detected in two (3.3%) samples. Meanwhile, all samples tested negative for GI.

Table 2 Detection of noroviruses in air samples using RT-qPCR and RT-nested PCR.
Full size table

Quantification of noroviruses in the air samples

Of 13 norovirus-positive samples, the viral load in the air samples for one GI and five GII could be determined in air samples. The norovirus GI concentration was 1.1 × 103 genome copies/mL or 6.0 × 102 genome copies/m3 with the 20 mL collection vessel. The norovirus GII concentrations ranged from 6.6 × 101 to 1.0 × 104 genome copies/mL or from 3.4 × 101 to 5.0 × 103 genome copies/m3 with the 5 mL (three samples) and 20 mL (two samples) collection vessels. Four air sampling sites tested positive for norovirus (either GI and/or GII) based on RT-qPCR and/or RT-nested PCR for the 5 and 20 mL collection vessels from the same areas. The aerosol samples No. A10-5, A10-20 and A11-5, A11-20 were collected from gynecology clinic and family planning clinic, respectively at the same day (31 Mar 23). The aerosol samples No. A24-5 and A24-20 were collected from ophthalmology clinic (28 Apr 23). Notably, one aerosol sample No. A28-20 collected in the 20 mL collection vessel from cohort ward PUI 3 (29 Apr 23) tested positive for GI and GII using RT-qPCR. Meanwhile, GII collected in the 5 mL collection vessel (A28-5) at the same air sampling site was also identified by RT-nested PCR (Table 3 and Supplementary Table S2).

Table 3 Norovirus-positive air samples detected using RT-qPCR or RT-nested PCR.
Full size table

Molecular characterization of norovirus GII

Two air samples had strong intensity bands of norovirus GII at 344 bp under gel electrophoresis. One air sample No. A28-5 was collected in the 5 mL collection vessel from cohort ward PUI 3 (29 Apr 23) and the other No. A09-20 was collected in the 20 mL collection vessel from pharmacy of the medical ward (30 Mar 23). The nucleotide sequences of these norovirus GII-positive samples were obtained via DNA sequencing of the amplicons. They were genotyped as GII.17 by the norovirus genotyping tool. Regarding the sequence alignment using the BLAST program, similar to the ASE_06/KOR/ 2019 and 38/CHN/2019 strains, these sequences had nucleotide identity of 98.5% and 100.0%, respectively. By phylogenetic analysis, the two sequences clustered in the reference GII.17 strain (Kawasaki308/JPN/2015; LC037415) and belonged to the GII.17 norovirus. These strains were similar to the norovirus GII.17 strains found in Korea and China in 2019. Moreover, they were closely related to the GII.17 strains detected in patients with norovirus gastroenteritis in Thailand from 2021 to 2023 (unpublished data) (Fig. 1).

Fig. 1
figure a

Phylogenetic analysis of the 298 bp norovirus nucleotide sequence compared to reference strains. The norovirus GII strains detected in this study (AIR28-5/THA/2023 and AIR09-20/THA/2023) and the reference sequences retrieved from the GenBank were constructed using the neighbour-joining method based on the Tamura-Nei model with bootstrap analysis of 1,000 replicates. Numbers at each node (> 80) are shown. The nucleotide sequences obtained in this study are indicated as bold. The scale bars indicate the nucleotide distance.

Full size image

RT-PCR inhibition in the air samples

The use of undiluted RNA extract from the water concentrates of air samples revealed that the RT-PCR inhibition was higher than the acceptable level (≤ 75%) for both norovirus and SARS-CoV-2. Thus, the RNA extract was diluted 1:4 and tested for the presence of viruses as well as RT-PCR inhibition. All air samples collected in the 5 and 20 mL collection vessels had RT-PCR inhibition within the acceptable level for norovirus GI. For norovirus GII, RT-PCR inhibition in the 5 and 20 mL collection vessels were within the acceptable level (73.3% and 100%, respectively). Nevertheless, the frequency of samples within the acceptable level of RT-PCR inhibition was lower in the SARS-CoV-2 assay (56.7% in 5 mL and 76.7% in 20 mL collection vessels) (Table 4).

Table 4 RT-PCR inhibition of norovirus and SARS-CoV-2 from air sampling in water using RT- qPCR.
Full size table

Discussion

The investigations of viruses in aerosols are improving, thereby providing a better understanding on environmental contamination caused by respiratory and gastrointestinal viruses and the sources of transmission routes of aerosols. To use a developed concentrating method and molecular techniques for detecting and quantifying of viruses in air samples14, the current study was performed in a hospital, which is a closed environment. The viruses in the air were collected from the water using the BioSampler, processed for virus concentration by speedVac centrifugation, and tested for the presence of viral RNA and DNA by RT-qPCR and/ or RT-nested PCR. Numerous factors affect the results of virus detection from indoor aerosols. These factors include air sampling, virus lasting and stability in air, viral concentration, ventilation, and personal behaviors affecting the spread of virus etc. Potential air sampling and sensitive detection methods are significant factors that should be considered in virus aerosol studies28. This study used the effective air sampling and concentrating method as reported in a previous study14. The highly sensitive molecular methods were also used to quantify genome copy and characterize norovirus strain in aerosol samples since the methods could successfully determine for concentration and identification of the virus in environmental samples22,25.

Factors affecting enveloped or non-enveloped viruses play an important role in virus resistance in the environment29. Norovirus in the air samples had a higher detection rate than SARS-CoV-2. Therefore, non-enveloped viruses are more stable because of the more resistance of the viral capsid30. Some norovirus-positive samples and SARS-CoV-2-positive samples had a weak fluorescence signal on RT-qPCR, thereby indicating the presence of RNAs at the lower limit of detection. However, a non-enveloped bocavirus could not be detected in the collected air samples in the water in the 5 and 20 mL collection vessels using the highly sensitive RT-nested PCR. This could be caused by an extremely lower detection rate in aerosols compared with that in other respiratory viruses12. The prevalence of HBoV1 which is transmitted via aerosols might be low when compared to that of the fecally transmitted HBoV2, HBoV3 and HBoV4. Our previous study was able to detect these three genotypes in recycled water and sewage sludge thereby exhibiting the high viral load of HBoVs in fecally-contaminated samples24.

Under experimental conditions, norovirus GII in an aerosol chamber could be detected within 2 h in the 5 mL collection vessel and 4 h in the 20 mL collection vessel after aerosol generation14. Naturally occurring noroviruses GI and GII RNAs were quantified in the air samples collected in the two collection vessels, thereby representing a potential concentrating procedure coupled with the molecular method for assessing airborne transmission. RT-qPCR is widely used for detecting norovirus31 and SARS-CoV-212,28 in aerosols due to quantitation and high sensitivity, specificity, rapidity, and reduction in cross-contamination. One norovirus GI-positive sample contained RNA at the amount of 6.0 × 102 genome copies/m3 and one-half of the norovirus GII-positive samples could be quantified with the RNA viral load ranging from 3.4 × 101 to 5.0 × 103 genome copies/m3. This result is in accordance with that of previous studies, which reported norovirus concentrations at a range of 1.35 × 101–2.35 × 103 genome copies/m332 and 5.0–2.15 × 102 genome copies/m333 during outbreaks in healthcare settings.

Two air samples tested positive for norovirus GII using RT-nested PCR, and they were identified as GII.17 strains. The norovirus genotype in aerosols is not commonly registered due to a low viral load. In a previous study, norovirus GII.4 Sydney was genotyped33. From 2014 to 2015, the outbreaks of GII.17 norovirus gastroenteritis have been increasing in Asian countries such as China34 and Japan35. Thus far, the GII.17 norovirus strains have been identified in patients with acute gastroenteritis at a global scale36,37,−38. Moreover, the GII.17 strains were found to be associated with foodborne and waterborne outbreaks39,40. In Thailand, the GII.17 norovirus strains were found in sporadic cases41, bivalve shellfish42 and environmental waters25. Hence, the circulation of this genotype might be responsible for norovirus outbreaks caused by the consumption of contaminated food and water. The current study first characterized the norovirus GII.17 in air samples and provided additional information on the possible route of norovirus transmission via aerosols. Nevertheless, further studies may evaluate the viability of norovirus to elucidate the potential role of norovirus-laden aerosols.

The present study found norovirus in aerosol samples collected from different areas in the hospital whether these areas related to fecal contamination. It is likely that norovirus from infected patients might spread throughout collection areas in the form of droplet nuclei and then fall to the floor. The limitation of this study is no information available regarding norovirus episode from infected patients or norovirus contamination in the detected area as well as the possibility of norovirus dispersion from the floor surface and movement in air with dust. A previous study of feline calicivirus as a norovirus surrogate suggested that dust might be the main route of airborne transmission for norovirus via droplet nuclei43.

Each of the aerosol samples collected in this study was tested for norovirus GI and GII RNAs using RT-qPCR to quantify genome copy and RT-nested PCR to characterize norovirus strain. The discordance of RT-qPCR and RT-nested PCR results might be the difference in primers, probe, genetic sequence amplification of target and PCR condition. This finding is consistent with our previous study of norovirus in environmental samples25. Nevertheless, one air sample (No. A28) revealed positive results by both methods thereby emphasizing the presence of norovirus GII in aerosols. The negative results on the evaluated viruses in aerosols might be partly attributed to the presence of RT-PCR inhibitors including biological materials, and organic and inorganic substances44. A previous study has found that a 1:10 dilution of RNA extracts can reduce RT-PCR inhibitors in aerosols effectively14. However, the sensitivity of norovirus detection might also be reduced by a corresponding dilutional factor. Norovirus RNA could be detected in air samples with acceptable levels of RT-PCR inhibition. Thus, the 1:4 dilution of RNA extract in this study was probably appropriate.

In conclusion, aerosols can be a source of norovirus and SARS-CoV-2 transmission. The detection rate of RT-qPCR for norovirus testing in air samples is higher than that of RT-nested PCR. The advantage of the former is quantification. Meanwhile, the advantage of the latter is genotyping. The combination of aerosol sampling in water and molecular assays can be used for the surveillance of viruses in hospital facilities.