High Throughput Metagenomic Analysis Exposes Mobile Phones as Potentially Hazardous Microbial Platforms Warranting Robust Public Health and Biosecurity Protocols

Advancements in technology and communication have revolutionised the 21 st century with the introduction of mobile phones and smartphones. These phones are known to be platforms harbouring microbes with recent research shedding light on the abundance and broad spectrum of organisms they harbour. Mobile phone use in the community and in professional sectors including health care settings is a potential source of microbial dissemination. Aim. To identify the diversity of microbial genetic signature present on mobile phones owned by hospital medical staff. Methods. Twenty-six mobile phones of health care staff were swabbed. DNA extraction for downstream next generation sequencing shotgun metagenomic microbial proling was performed. Survey questionnaires were handed to the staff to collect information on mobile phone usage and users’ behaviours. Results. A total of 11259 organisms derived from 26 phones were found with 2096 genes coding for antibiotic resistance and virulent factors. These organisms corresponded to 5717 bacteria, 675 fungi, 93 protists, 320 viruses, 4456 bacteriophages. The survey of medical staff showed that 46% (12/26) of the participants used their mobile phones in the bathroom.


Introduction
Mobile phones are ubiquitous and are used as primary communication devices. There are accounting for over 5 billion mobile phone users globally (over two-thirds of the world's population) with an increase of 100 million unique mobile phone users each year 1 . According to Statista in 2020, the number of mobile phone users accessing popular messaging apps to communicate was 2.77 billion 2 . There have been many risks identi ed linked to the use of mobile phones including addiction 3 , vision impairment in children 4 , dangerous driving 5 , distracted pedestrians 6 , psychological stress and general anxiety 7 . Mobile phones are used up to 3 hours and 37 minutes per person and touched with hands more than 2000 times a day 8 . A previously underestimated risk of mobile phones is associated with their role as fomite and several recent studies have con rmed the presence of viable microbes on their surface 9,10 . The United States Centre for Disease Control and Prevention (CDC) outlined that up to 80% of all infectious diseases was transmitted via hands 11 . Researchers have shown that mobile phones are reservoirs of microbes, users neglect and rarely decontaminate these devices, high rates of use and touch contact of mobile phone surfaces and individual' tendencies to touch their face regularly (up to 23 times an hour) 12 or/and other surrounding surfaces 13 . Olsen et al. (2020) stated that mobile phones act as 'Trojan horse' devices which: i) bypass gold standard hand hygiene practices; ii) are likely linked to pathogen movement via cross-contamination transmission pathways during epidemics and pandemics 14 ; and iii) contribute to global population infections and hospitalisations due to nosocomial infections. A recently published survey of 165 healthcare workers (HCW) demonstrated that 52% (86/165) of participants used their mobile phone in the bathroom/toilet and that 57% reported that they never wash their devices 15 . Mobile phones are platforms that host microbial vectors leading to the dissemination of infectious diseases. The use of phones by all professional sectors makes them ideal platform niches for micro-organism contamination 10 . Despite a massive increase in published articles describing the role of mobile phones as fomites there is still poor global awareness with continuing poor practices of standardised sanitisation. In 2020, a global scoping review of 56 studies identi ed that on average, 68% of mobile phones were contaminated with microbes with many harbouring antibiotic resistant bacteria 9 . While such scoping review was informative, microbial characterisation and identi cation from these studies most probably underestimated the overall spectrum and richness of microbes on mobile phones. These studies were based on traditional agar-based growths, biochemical testings or orthogonal polymerase chain reaction ampli cation of microbial genomic sequences 9 .
Improved methodology using a sequencing approach with 16S rRNA primers for metagenomic sequencing also highlighted the inadequacy of traditional culture-dependent identi cation techniques to capture the entire globality of microbes present on mobile phones 16 . A 2021 pilot next generation sequencing project was able to capture a wider population of micro-organisms with all mobile phones found to be contaminated with microbes . The ndings consisted of 235 bacteria, 8 fungi, 8 protists and 53 bacteriophages reported from only ve mobile phones derived swabs 14 . However, this study still could be considered as an underestimation of microbial nding as the samples were pre-cultured on agar plates prior to next generation sequencing metagenomic pro ling 14 . A similar study used a metagenomic shotgun sequencing-based approach of viable pre-cultured microbes collected from 30 mobile phones of HCW. These phones were swabbed across 4 different hospital wards and plated on ve different agar plates (Horse Blood agar, Nutrient agar, MacConkey agar, Bile Esculin agar, Mannitol Salt agar) before being subject to next generation sequencing 10 . The study identi ed a large range of microbial organisms with 399 operational taxonomic units (OTUs) bacteria, 155 bacteriophage OTUs and the identi cation of 133 antibiotic resistant genes (ARGs) and 347 virulence factor genes (VFGs).
To address the limitations identi ed in previous studies, this study collected swabs from mobile phones of health care staff working in a hospital. These swabs were subject to a direct shotgun next generation sequencing to identify the metagenomic presence of micro-organisms on these surfaces.

Participant recruitment and sample collection
Informed consent was obtained from all subjects of this study with a total of 26 health care workers from the Paediatric Intensive Care Unit (PICU) and the General Paediatric Department (GPD) of the Gold Coast University Hospital, Australia. An information sheet was provided to all participants, detailing the nature of the research, with no personal identifying information collected. Informed consent was provided verbally and agreeing to participate on the day of sampling. Samples were collected each of the 26 mobile phones using culture swab EZ II swabs (Becton Dickson) pre-moistened with sterile saline. During the sample collection phase, gloves were worn and changed between participants to prevent cross-contamination. The mobile phones were swabbed on both the front and back of the devices with swabs then placed in portable containers and transported immediately to the laboratory for processing.

Survey Questionnaire
The complete survey data set has been published previously 15 , however, for this paper some results have been extracted to enable comparison with the microorganisms discovered. The 26 questionnaire survey responses were included in this paper.

Swab and DNA extraction
The preliminary step of the DNA extraction process involved the use of bead beating with 0.1 mm diameter glass beads (BioSpec Products, Bartlesville, OK USA) on a Powerlyser 24 homogenizer (Mo-Bio, Carlsbad, CA USA) at the Australian Centre for Ecogenomics (ACE), Brisbane, Australia. Brie y, samples were transferred to a bead tube and 800µl of bead solution (Qiagen, Germantown, MD USA) was added and bead-beat for ve minutes at 2,000 rpm, then centrifuged at 10,000 g for one minute. Following the addition of 60µl of cell lysis buffer, tubes were vortexed and then heated at 65°C for 10 minutes (while mixing at 1,000 rpm), then vortexed again for 30 seconds and stored at -20°C pending DNA extraction. Prior to DNA extraction, samples were thawed at room temperature; vortexed and centrifuged for one minute at 10,000 g. The resulting lysate was transferred to a new collection tube and DNA extraction carried out using DNeasy Powersoil Kit (Qiagen), as per manufacturer protocol with a nal elution volume of 50 µl using sterile, EDTA-free elution buffer.

Metagenomic sequencing and Bioinformatic Analysis
Libraries were prepared according to the manufacturer's protocol using Nextera DNA Flex Library Preparation Kit (Illumina San Diego, CA USA). Preparation and bead clean-up were run on the Mantis Liquid Handler (Formulatrix) and Epmotion (Eppendorf) automated platform. On completion of the library prep protocol, each library was quanti ed, and quality control (QC) was performed using the Quant-iT™ dsDNA HS Assay Kit (Invitrogen, Carlsbad, CA USA) and Agilent D1000 HS tapes on the TapeStation 4200 (Agilent Technologies, Santa Clara, CA USA) as per manufacturer's protocol. Library Pooling, QC and Loading Nextera DNA Flex libraries were pooled at equimolar amounts of 2nM per library to create a sequencing pool. The library pool was quanti ed in triplicates using the Qubit™ dsDNA HS Assay Kit (Invitrogen). Sequencing was carried out on the NextSeq500 (Illumina) using NextSeq 500/550 High Output v2 2 x 150bp paired end chemistry according to manufacturer's protocol. 12 The post-sequencing derived raw data were retained and transferred into Illumina base space platform (https://basespace.illumina.com). Following the sequencing runs, data as demultiplexed FASTQ les were uploaded into CosmosID platform (https://www.cosmosid.com/). Raw datasets Fastq les were analysed using the CosmosID software to identify bacteria, fungi, virulence factor genes and antibiotic resistance genes. The datasets were then analysed with proper mining bio-informatic analytic tools using high performance datamining k-mer algorithm and highly dynamic comparator databases (GenBook ® ). Through this process, the raw data of millions of short reads can be distinctively aligned against sequences of microbial genomes and genes (CosmosID Metagenomics Cloud).
Microbial 'Richness' corresponds to the cumulative amount of all distinct microbes detected across all phones whereas the number of occurrences across all phones for each of these distinct microbes are represented by Hits.

Ethics
Ethical approval was obtained from Bond University Human Research Ethics Committee (16004) and the GCUH Human Research Ethics committee with Site Speci c approval (GC HREA 46569). All methods were performed in accordance with the relevant guidelines and regulations.

Funding Support
Funding for the DNA sequencing was made available thanks to a consultation research-based account owned by LT and administered at Bond University.

Participant features and Questionnaire ndings
In total, there were 26 health care workers who participated in this study: 16 nurses, 8 doctors, 1 outpatient clinical staff and 1 unspeci ed participant. 16 staff members were from the General Paediatric Department and 10 were from the Paediatric Intensive Care Unit. Majority of the participants (77%; N=20/26) were completing their shift and whilst 23% (n/N=6/26) commencing their shift. 77% (20/26) reported using their mobile phones at work with 88% (23/26) believing their mobile phones were essential tools for their job. 96% (25/26) of participants believed their mobile phones would harbour potentially pathogenic microorganisms. Figure 1 illustrates the hygiene habits associated with mobile phone use in the professional setting, 46% (12/26) of the participants had recently used their mobile phones in the bathroom (Figure 1a). Of the medical staff using mobile phones in the bathroom, 58% (7/12) reported using their devices for social media access, 25% (3/12) did not specify the purpose of use and 16% (2/12) reported using their phone for work-related purposes. Over half of the participants (54%; n?N=14/26) of participants had never cleaned their mobile phone ( Figure 1b). Of the 46% (12/26) of participants who had cleaned their mobile phones at some point, 25% (3/12) did so within the past year, 33% (4/12) did so within the past month, 16% (2/12) did so within the past week and 25% (3/12) did so within the past day. Of those who reported cleaning their phones, 41% (5/12) used an alcohol-based wipe and 33% (4/12) used a disinfectant spray (Figure 1c).

Reads
The average amount of sequencing reads per mobile phone was approximately 53 million reads Sample 26 (NS313-110) contained the lowest (33 million) and sample 12 (NS250-72) the highest number of reads (156 million) respectively.

Sequencing reads and metagenomic overview
A total of 11259 microorganisms and 2096 genes coding for antibiotic and virulent factors were identi ed in this metagenomic shotgun next generation sequencing study. In total, there were 5717 bacteria, 675 fungi, 93 protists, 320 viruses, 4456 bacteriophages, 560 antibiotic resistant genes and 1 536 virulence factor genes identi ed across the 26 mobile phones from GPD and PICU (Table 1).
On average, mobile phones from the GPD contained higher amounts of pathogens and genes, compared to the phones sampled from PICU. Additionally, mobile phones of nurses contained in average a slightly higher number of pathogens compared to doctors with 460.2 and 403.6 respectively. Across all 26 mobile phones, the average number of pathogens was calculated to be 433 with an average of 477.7 on the GPD phones and 361.6 in the PICU phones. Pathogen numbers ranged from 138 to 669 per phone and genes (ARG and VRG) ranged from 7 to 144 per phone. (Table 1). Bacteria and bacteriophages represented the largest proportion of the microorganism distribution ( Figure 2 3).
Mobile phones microbial composition varied with a subset of microbes uniquely present in either department: 170 and 317 bacteria in PICU and GPD respectively. These unique ward bacterial signatures showed different bacterial phylum pro les with the bacterial Actinobacteria phylum demonstrating the larger signature subset of PICU derived mobile phones while Bacteroidetes, Firmicutes, and Proteobacteria phylum were predominant in GPD derived devices (Figure 4).

Bacteriophage identi cation
In total there were 512 different bacteriophage viruses accounting for 4453 hits. Figure 5 Figure 5).
A signi cant difference in the number of bacteriophages was observed between the two wards (GPD and PICU) (P-value: 0.0022) (Wilcoxon Rank Sum Test) (Figure 6).

Viral identi cation
Eighty-seven different viruses accounting for 320 richness hits was found on the mobile phones. Seven different human herpes viruses (HHV) were identi ed and corresponded to a total richness of 29 hits. 15 phones had at least one HHV and in one phone  Figure 7 highlights the range of protozoa identi ed with several amoebae of the protozoal group Sarcodina with Acanthamoeba polyphaga, Acanthamoeba palestinensis, Naegleria fowleri, Entamoeba dispar, Entamoeba histolytica ( Figure 7).

Discussion
This study performed metagenomic pro ling of swabs derived from 26 mobile phones of HCW, predominantly doctors and nurses, were also more common on mobile phones from the GPD versus PICU with 194 and 135 respectively. This ward microbial burden difference was observed in both nursing and medical staff. The reduction of mobile phone microbial burden in PICU might be associated with higher frequency of hand hygiene practices or more stringent infection control measures. Interestingly, the average number of microbes irrespective of the ward was always higher in mobile phones owned by nurses than doctors with the exception of fungi and protists that were found in higher number on doctor phones from the GPD. Additionally, d mobile phones of the doctors from the GPD had a higher number of antibiotic resistant and virulent factor genes than those of nurses. However, in PICU, nurses' mobile phones had a higher number of antibiotic resistant and virulent factor genes compared to doctors within that department. Overall, the microbial load on phones from both departments was at levels that should be considered problematic. This study has identi ed a serious general hospital infection control concern that may escalate future public health threats.
The study also identi ed other microbial presence on mobile phones that raises concerns. Clinically relevant pathogens such as Bordetella pertussis, responsible for whooping cough was present on 69% of all phones studied, Streptococcus pneumoniae and the emergent nosocomial bacteria Stenotrophomonas maltophilia were each present on 81% of all phones studied (21/26).
Food borne bacteria (Bacillus cereus) was identi ed on HCW mobile phones. While this study was done in a hospital setting, it con rms that other industries such as the food industry are also at risk of microbial cross contamination from mobile phones. Other concerning organisms including Clostridioides di cile, Moraxella catarrhalis, Proteus mirabilis, Elizabethkingia meningoseptica and the sexually transmitted infectious bacteria Neisseria gonorrhoeae were identi ed on phones in this study Clostridioides di cile infections has been shown to spread from contaminated surfaces with the risk of infection higher when using bathrooms preceded by infected individuals 17 . Finding HCW mobile phones to be microbial laden fomites possibly confers appropriate conditions to disseminate infections to susceptible hosts and immune-compromised patients and is a real public health risk. One example is nding Elizabethkingia meningoseptica, a nosocomial bacterium that has disastrous consequences on premature babies with known past outbreaks linked to phone receivers 18 .
Human behaviours and the constant contact with mobile phones in toilets provide cumulative evidence that such devices are exposed to unsanitary conditions leading to the presence of a range of viable microbes on these platforms. Based on this study and others, it appears mobile phones are rarely or ever cleaned and even when cleaned this may occur in an ineffective manner.
Mobile phones act as fomites turning these devices into ideal platforms for disease transmission either by means of selfinoculation when touching your own mobile phone and face orby simple microbial dissemination in the environment, public places, or professional sectors.
Bacteriophages were also found in association with bacteria with 512 different phages found and accounting for 4453 hits across 26 HCW mobile phones. Additionally, 87 different viruses including animal and human viruses were detected. These consisted of seven different human herpes viruses with 15 phones found with at least one HHV and one phone harboured up to 5 HHVs. In hospitals, it is now commonplace for mobile phones to be used by the majority of HCW, they may however be counteracting World Health Organisation hand hygiene campaigns. The efforts to limit exposure of microbes to patients may be nulli ed if mobile phones are not decontaminated regularly 19 . The number of microbes identi ed on phones does suggest that new measures of infection control in these vulnerable areas should be implemented. This should include mobile phone sanitisation as a corollary to the Five Moments of Hand Hygiene 20 . Mobile phones should now be considered as the 'third hand' from their users and subject to frequent decontaminations in hospitals (both health care staff and patients/visitors). An infographic shows the dissemination route of microbes derived from healthcare staff users and users of the community (Figure 11). Figure 11 illustrates the transmission dynamics of microbes derived from mobile phones and the possible inter-related dissemination of germs in and out healthcare and community settings.
A. Mobile phones exposed to all sorts of community environments will harbour diverse microbes from the user's hands. These

Author's Recommendation
This direct swab to metagenomic analysis study has revealed that hospital derived mobile phones used by health care workers, are accommodating niches for large amount of diverse pathogenic germs that are equipped with an arsenal of virulence genes and large spectrum of antibiotic resistance.
While this study took place in a hospital, the research highlights the need for the scienti c community and public health authorities to further investigate the role mobile phones play as fomites. The potential for them to be vehicles for transmission and propagation of infectious microbes across health care settings needs to be addressed. Additionally, mobile phones harbouring a plethora of viable microbes are in circulation, with billions currently owned globally, and may be the means to establish, maintain or spread epidemics and pandemics. As an example, SARS-CoV-2 was detected on mobile phones and shown to survive on such platforms up to 28 days. 21 Undetected introductions of biothreats and invasive pathological organisms might be due to the billions of passengers travelling around the globe with 'uncleaned' mobile phones. Presence of SARS-CoV-2 omicron or delta variants on mobile phones need to be investigated.
Additionally, this research emphasises that the density of microbes found on mobile phones may be the ideal platforms for horizontal genetic transfers to occur among different species of micro-organisms such as transformation, conjugation, and transduction. Mobile phones may act as platforms for microbial multiplication and as a dynamic training 'school' for superbugs to evolve (and disseminate).
Mobile phones are dynamically contaminated with all sorts of microbes touched by the hands of their users thousands of times a day, even while in bathrooms. Mobile phones therefore have become our third hand. They are 'dirty' as are infrequently cleaned/sanitised and are completely negating rst the worldwide gold standard hygienic hand washing practices and secondly the cost-effective public health and biosecurity prophylactic measures. Mitigation resides in sanitising mobile phones as frequently as we wash our hands with the adoption of new technology driven solution a like safety-certi ed enclosed ultraviolet-C emitting mobile phone sanitisers to clean phones in 10-20 seconds. This fast and e cient technology driven sanitisation of phones is practical as could be performed while health care workers practise hand hygiene. Presence in healthcare facilities of stations that can decontaminate both hands and mobile phones will prevent the risks of cross contamination and should be implemented in the ve moments of hand washing.
It also sends a strong message to the general community to prevent further global microbial dissemination. These metagenomics analysis ndings revealed a real biosecurity concern with possible economically important disease repercussions that authorities must take seriously. Not only were some microbes on mobile phones highly resistant to multiple antibiotics, but cancer related viruses such as herpes viruses, polyomaviruses and human papillomaviruses are also of high concern for public health if mobile   Tables   Table 1:   Distribution of different types of microorganisms across the 26 mobile phone samples. Distribution of bacteriophages identi ed across the 26 mobile phones.

Figure 7
Distribution of protists identi ed across 26 mobile phones. Heatmap representation of antibiotic resistant genes clustered by occupation. Heatmap representation by healthcare occupation of the 419 distinct virulence factor genes identi ed by metagenomic analysis.
Page 19/20 Figure 11 Contaminated mobile phones potential vectors of dissemination of germs in and out healthcare and community settings Figure 12 Mobile phone contaminated with microbes pose national and global biosecurity threats.