Global monitoring of antimicrobial resistance based on metagenomics analyses of urban sewage

Antimicrobial resistance (AMR) is a serious threat to global public health, but obtaining representative data on AMR for healthy human populations is difficult. Here, we use metagenomic analysis of untreated sewage to characterize the bacterial resistome from 79 sites in 60 countries. We find systematic differences in abundance and diversity of AMR genes between Europe/North-America/Oceania and Africa/Asia/South-America. Antimicrobial use data and bacterial taxonomy only explains a minor part of the AMR variation that we observe. We find no evidence for cross-selection between antimicrobial classes, or for effect of air travel between sites. However, AMR gene abundance strongly correlates with socio-economic, health and environmental factors, which we use to predict AMR gene abundances in all countries in the world. Our findings suggest that global AMR gene diversity and abundance vary by region, and that improving sanitation and health could potentially limit the global burden of AMR. We propose metagenomic analysis of sewage as an ethically acceptable and economically feasible approach for continuous global surveillance and prediction of AMR.


Protocol; Global Sewage Surveillance Project
-global surveillance of infectious diseases and antimicrobial resistance from sewage HISTORY OF CHANGES; version 2 -Added that the shipment is not IATA restricted as per SP A197 (UN labels are not required) -Added references to Appendix 1, 2, 3, 4 and 5 -Added that an ethical approval is not required -Paragraph on sample storage and shipment adjusted -Editorial changes

EXECUTIVE SUMMARY
No single approach exists for adequately monitoring large populations and their environments for the emergence of novel pathogens. Recent developments in high-throughput sequencing offer the ability to rapidly identify nucleic acids from various organisms in clinical and environmental samples. Sewage systems are recognized as an important source of human pathogens, especially in crowded settings with poor infrastructure. A point-prevalence metagenomic analysis will be applied to sewage samples collected globally from the main sewage system of major cities prior to treatment plants inlet. The project will serve as proof-of-concept for applying metagenomic approaches, which could initiate a global surveillance of human infectious diseases including antimicrobial resistance from sewage collected in major cities around the world to detect, control, prevent and predict human infectious diseases.

Background
Human and animal populations are increasingly confronted with novel, emerging or re-emerging infectious, zoonotic, and communicable diseases including those that are multi-drug resistant. Many of these events can be attributed to increased globalization, urbanization, climate change, population growth, and intensive farming. According to the WHO, more than 25% of the total 58 million worldwide annual deaths are the direct result of infectious diseases.
Rapid detection and accurate identification of pathogens and antimicrobial resistance are essential in disease control and prevention strategies. Metagenomic analysis of genetic material through highthroughput sequencing offers the potential to greatly enhance our ability to rapidly detect emerging pathogens and related antimicrobial resistance genes. Human disease surveillance is often hampered due to ethical problems with sensitivity of data collected from individuals. Exposure to human waste is a well-established risk factor, why sewage has been suggested as an alternative to obtain population wide samples targeting various health hazards. This could be of considerable value in attempting to establish a global human disease surveillance program because it is virtually impossible to test an entire population and obtain samples from healthy humans. To date, there have only been limited applications of metagenomic analysis for monitoring large human populations. If monitoring of pathogens and antimicrobial resistance in sewage can provide timely information on pathogens of concern, this information can be used to assist risk managers with information on appropriate prevention and treatment strategies and potential needs for environmental remediation.
This joint study between the the World Health Organization (WHO) and National Food Institute, Technical University of Denmark (DTU Food) (WHO Collaborating Centre for Antimicrobial Resistance in Food borne Pathogens) will serve as proof-of-concept for applying these metagenomic approaches to initiate a global surveillance of human infectious diseases from sewage collected in major cities around the world to detect, control, prevent and predict human infectious disease.

Description of Innovation
By applying the proposed innovative approach of using a combination of bench-top whole community sequencing (WCS) with metagenomic analysis on sewage samples collected from main sewage systems of major cities prior to treatment plants inlet, it may be possible to detect and monitor all known microbial agents and associated epidemiological markers, such a virulence and antimicrobial resistance, in a large healthy human population in relative real-time and at a low cost. This will enable the implementation of control measures that could potentially save lives and prevent further local or global spread and even predict future events.
The primary objective of the study is to evaluate the possibility of using WCS technology directly on sewage for surveillance of infectious, zoonotic, and communicable diseases in a global context. This will be achieved through the collection of two random consecutive and representative sewage samples from each city representing the entire population with connection to the main sewage outlet. Each sample will consist of approximately 1L of sewage collected at a designated sampling point (e.g. main sewage pipe prior to waste water treatment plants (WWTP inlet or disposal directly into rivers or similar, following a detailed sampling protocol (see below)). Sampling locations will be georeferenced by GPS and an image of the sampling site will be captured. The samples will be stored at -80°C until shipping to the National Food Institute, Technical University of Denmark (DTU Food) for WCS and further metagenomic analysis. The shipment is not IATA 1 restricted as per SP A197 (see Appendix 2 and 3), this means that the parcel can be sent without a UN-label.
At DTU Food, extraction of DNA and sequencing will be performed using Illumina Hiseq technology. An expected 50-100 million reads will be sequenced from each sample. A subset of collected sewage samples will be spiked with a constructed pool of known pathogens to evaluate sensitivity of the extraction and sequencing methodology.

____________________________________________________________________________________________________
Global Sewage Surveillance Protocol -version 2 (January-2016) -Page 3 of 7 Different available and in-house programs hosted by DTU's Center of Genomic Epidemiology will be used to filter data and map sequence reads against databases of reference genomes (ca. 2,000), selected genes, and human microbiome data (MetaHIT gene catalogue and HMP genomes). The remaining part of the sequence reads, that do not map to anything, will be compared with Genbank (NT) using Blastn. For all abundant microbial taxonomic groups, the matching sequence reads will be counted and used to define an abundance profile. The data will be analyzed and presented per country and subsequently be related to available data such as antimicrobial usage. At a later stage, additional hazards such as the content of virus and parasites as well as bacterial virulence markers will be examined. All sequence data will be submitted and deposited in the public domain at NCBI /ENA.

Impact
The most important outcome will be a proof-of-concept of "real-time" large-scale population surveillance combining state-of-the-art technology and analytic facilities that provide better and faster detection and control of health risks. In addition, the project is also expected to provide a proof-of-principle regarding the evaluation of WCS performed directly on sewage samples. The impact from this project could establish the foundation for the first surveillance of a large, healthy human population and possibly animal populations. Thus, it could reduce morbidity and mortality through rapid disease detection, reduce the development of antimicrobial resistance through proper drug adherence and enable earlier clinical treatment (interventions?), and ultimately improve treatment outcome and minimize disease spread. The outcome of this study could lead to a complete paradigm shift in the way infectious disease surveillance of nationwide or disease hot spots are conducted.

Key risks
It is impossible to accurately predict whether infectious diseases will be present in the sampling site or period. To-date, WCS studies have used clinical samples such as urine, blood and feces with success; however, there is limited knowledge on the sensitivity and specificity of WCS from sewage samples. In addition, there may be a limitation in the detection of all microbial agents since this depends on the availability of reference data. However, the knowledge gained even if the expected outcomes are not achieved will be valuable since it will identify the limitation in the proposed approach and technology enabling researchers to focus on those areas for future applications.

Ethical Issues
All analysis will be conducted in accordance with the Danish Act on scientific ethical treatment of health research administrated and confirmed by the Research Ethics Committees of the Capital Region of Denmark (www.regionh.dk), Journal no.: H-14013582. Thus, it will not be possible to trace back samples or data to any individuals, i.e. an ethical approval is not required.

Publication and IP
The samples will be stored temporarily at DTU prior to disposal. The samples will not be transferred to third parties and will solely be used for this international sewage project. All metagenomic raw data for each individual sample will be made available for the individual sample provider as soon as the results are available and subsequently deposited in the public domain at the time of publication. All metagenomic data in combination with minimum meta-data (location and date) for all samples will be made available for the COMPARE consortium (www.compare-europe.eu).
Please contact Dr. Rene S. Hendriksen rshe@food.dtu.dk from DTU Food if there is a need for signing a contract, agreement or material transfer agreement (MTA).
It is expected that the study will result in a number of scientific publications. It is expected that all partners, either country mediator or sample providers, will be co-authors on the first publication (e.g. one per country). Persons also actively participating in the analysis of the samples and/or data analysis may become co-authors on multiple publications. All co-authors will have the opportunity to comment on the manuscript prior to submission.

Sampling site
The intention is to collect two samples representative of the entire population with connection to the main sewage outlet. Note that all WWTPs will most likely have an individual infrastructure and the project design allows for some individual variations as to the exact spot from where to collect the samples.
Not all sites have special equipment for continuous sampling concentrated over an entire day, and we will have to rely on crude point sampling in these cases.
From each location, two representative sewage samples (1L each, in total 2L) are collected from the main sewage flow on consecutive days from the city's main sewage pipelines prior to WWTP inlets or from the main outlet to rivers or similar. Samples can be obtained following the first filtering step, but it is important that there has been no processing of the sewage.
It is preferred to collect concentrated flow proportion sampling over 24 hours, however, should this not be possible due to lack of equipment, three crude point samples should be collected in a short time interval, i.e. at least 5 minutes between each individual sample, to ensure as much randomness as possible.
An image of the sampling site should be captured indicating the sampling environment. Please send the image to Mrs. Susanne Carlsson (suca@food.dtu.dk) indicating in the file name country and city. Please also share epidemiological data for each sample, including GPS coordinates of sampling location; WGS84 geodetic datum (e.g. N43⁰38'19.39'' / W116⁰14'28.86'') also used by Google Earth, temperature of the sample at the time of sampling, storage temperature of sample, etc.
A survey has been set up to collect these data, as well as other general information. Please submit a.s.a.p. after sampling via the following SurveyMonkey link: https://www.surveymonkey.com/r/SewageSampleDetails (see also Appendix 5)

Sampling procedure
A checklist relevant for the sampling is listed in Appendix 1.
Please follow the below basic instructions for each sample: 1) Using a permanent marker, label two clean 1000 mL plastic containers (no soap or disinfectant residue) marked with country, city, name of the collector, sample number (001 or 002) and date. Seal the label with tape to avoid the text being smeared.
2) Fill 1L of sewage into each of the containers -one per day. Leave a bit of space to allow the sewage to expand during freezing.
a) Collect each flow proportion sewage sample over 24 hours if possible from the mid-stream of the sewage inlet to the WWTP.
b) In the case where it is not possible to collect samples over a 24-hour period e.g. the sewage is not treated and runs directly into rivers, then collect the sample in short time intervals, i.e. with at least 5 minutes between each of three individual sample of approximately 300 ml, and subsequently pool the three samples to ensure as much randomness as possible.
c) Record the temperature of the sewage flow the sample is taken from.
3) Close the containers with the corresponding cap and tighten to avoid leakage. The sample must be kept as cool as possible and transported to the local laboratory within 8 hours of collection. Wipe the surface of the containers clean with alcohol. Pack each container in a plastic bag individually storing at -80°C as soon as possible.

Sample storage and shipment
Store the containers at -80°C for at least 48 hours (preferred) and prepare shipping the samples to the DTU Food in Denmark. The samples should be packed according to the description in Appendix 2. See also Appendix 4 that presents an example of the packing procedure. The samples should not be shipped using dry ice as this will complicate shipping and increase costs, for the same reasons, the shipment should be sent without temperature restrictions (no cool-chain necessary during shipment). Please send the samples frozen, packed directly from the freezer.
The international courier services of DHL or FedEx must be used. Please use the shipping account numbers provided below:

Timeline
The overview of the project will be as indicated in the table below.
The sampling window in which all participants should arrange for the sampling to take place will be the last week of January and first week of February.

Specific Safety Requirements and Responsibilities
This protocol describes how to collect sewage samples containing human and potentially animal faeces. Blood-and airborne pathogen protection (relevant personal protection: gloves, lab coat, mask, etc.) must be utilized when handling human and animal clinical samples.
Biosafety level-3 (BSL-3 / RG-3) practices and procedures must be followed when handling sewage or clinical samples suspected to contain pathogenic organisms. It is the responsibility of the collector and local laboratory to comply with the biosafety rules. b) In the case where it is not possible to collect samples over a 24-hour period e.g. the sewage is not treated and runs directly into rivers, then collect the sample in short time intervals, i.e. with at least 5 minutes between each of three individual sample of approximately 300 ml, and subsequently pool the three samples to ensure as much randomness as possible.
c) Record the temperature of the sewage flow the sample is taken from.
3) Close the containers with the corresponding cap and tighten to avoid leakage. The sample must be kept as cool as possible and transported to the local laboratory within 8 hours of collection. Wipe the surface of the containers clean with alcohol. Pack each container in a plastic bag individually storing at -80°C as soon as possible.

Appendix 2 Packing the sewage samples
The shipping of the sewage samples to Denmark is not IATA 1 restricted as per SP A197.
(Note: the contents of the parcel fall under the UN-category UN3082 environmentally hazardous substance, liquid, n.o.s. -BUT since the amount shipped is less than 5L, the consequence is that it is not IATA restricted as per SP A197) This means that when shipping the parcel, there should be no UN-diamond label on the outside of the parcel.
At departure, the sample material must be frozen (at -80⁰C), and the parcel must be packed to keep the sample material frozen for as long as possible, must protect the contents from leaking, and it must ensure that the liquid is collected inside the parcel, should one of the sample containers leak. This is described as follows by IATA in A197 and references (5.0.2.4.1, 5.0.2.6.1.1 and 5.0.2.8).
IATA A197 and references indicate that the packaging used must be good quality packagings which must be strong enough to withstand the shocks and loadings normally encountered in transport, including removal from a pallet, unit load device or overpack for subsequent manual or mechanical handling. Packages must be constructed and closed as to prevent any loss of contents when prepared for transport which might be caused under normal conditions of transport, by vibration or by changes in temperature, humidity or pressure (resulting from altitude, for example). Packages (including inner packagings and receptacles) must be closed in accordance with the information provided by the manufacturer. No dangerous residue must adhere to the outside of packages during transport. These provisions apply, as appropriate, to new, reconditioned or remanufactured packagings.
Parts of packagings which are in direct contact with the sample material: a) Must not be affected or significantly weakened by the sample material; b) Must not cause a dangerous effect, e.g. catalyzing a reaction or reacting with the sampling material; and c) Must not allow permeation of the sample material that could constitute a danger under normal conditions of transport When filling packagings for liquids, sufficient ullage (outage) must be left to ensure that neither leakage nor permanent distortion of the packaging will occur as a result of an expansion of the ________________________________________________________________________________________________ Global Sewage Surveillance Protocol, Appendix 2 -version 2 (January-2016) -Page 2 of 3 liquid caused by temperatures likely to prevail during transport. Liquids must not completely fill a packaging at a temperature of 55⁰C.

In summary:
When shipping, the sewage samples must be frozen at -80⁰C.
The contents of one parcel must not exceed 5L of waste water.
Packagings must be strong and of good quality.
The primary receptacles should be packed in secondary packagings in such a way that, under normal conditions of transport, they cannot break, be punctured or leak their contents into the secondary packaging. Secondary packagings should be secured in outer packagings with suitable cushioning material. Any leakage of the contents must not compromise the integrity of the cushioning material or of the outer packaging.
Any instructions from packaging manufacturers or distributors on filling and closing the packages must be followed to enable the package to be correctly prepared for transport.

On the external surface of the outer packaging
One label is necessary; i.e. 'This way up' indicated by two arrows (if it is not already printed on the external surface of the outer packaging). This example of the looks of the label may be printed and taped to the parcel.
No diamond-shaped labels are required for the shipment of this parcel (because the content is not IATA restricted as per SP A197).
Indicate the name and address of the shipper and of the consignee. ________________________________________________________________________________________________ Global Sewage Surveillance Protocol, Appendix 2 -version 2 (January-2016) -Page 3 of 3

Documentation
In the AWB for the 'full desription of contents', indicate Watersamples, not IATA restricted as per SP A197.
In Appendix 3 you find a letter from Professor Frank Aarestrup. This indicates the contents of the shipment, and also states that the shipment does not contain dry-ice. Print the letter, and make sure to add it with the AWB when sending the parcel.
(the letter is intended for those that handle the parcel that when they see on the x-ray that something in the parcel is very cold, they are informed that it is not a sign of dry-ice, and to confirm to them that the parcel is not IATA restricted as per SP A197) An import permit is not required.
A Shipper's Declaration for Dangerous Goods is not required.

Appendix 4
Packing the sewage samples -example This appendix presents an example for packing the sewage samples. When shipping the sewage samples to Denmark, the samples must be packed according to Appendix 2 in the Global Sewage Surveillance protocol.

(packing example)
Sewage sample is filled into sample containers Make sure to leave empty space in the sample containers

(packing example)
After filling sewage sample into the sample container, close them firmly

(packing example)
Place the sample containers in the -80 ⁰C freezer at least 48 hours until completely frozen

(packing example)
Cover the frozen sewage samples with vermiculite

(packing example)
Make sure the box is full, meaning that -The samples must not move inside the box when 'shaking' the package -The box must not be too full, but must be easily closed with the lid ________________________________________________________________________________________________ Global Sewage Surveillance Protocol, Appendix 4 -version 2 (January-2016) -Page 3 of 5

(packing example)
Place the lid on the box Use tape (could be fortified tape) for fixing the lid thoroughly to the box, i.e. all the way around the box where the lid is joined to the box

(packing example)
Make sure the tape fixes to the box

(packing example)
Fold the plastic bag and close it using tape

(packing example)
Make sure the tape is fixed well down the sides of the box ________________________________________________________________________________________________ Global Sewage Surveillance Protocol, Appendix 4 -version 2 (January-2016) -Page 4 of 5

(packing example)
With the palm of your hand, press gently on the tape allowing it to fix well to the plastic bag

(packing example)
Place the closed bag containing the polystyrene box inside the cardboard box Make sure the polystyrene box does not move inside the cardboard box when 'shaking' the package Make sure to tape all way around the parcel three times, i.e. once in the middle fixing the flaps, and twice all the way around the sides of the parcel as indicated on the photo Make sure the ends of the tape fix each other (see photo)

(packing example)
Make sure the tape correctly fixes the flaps to the cardboard box

(packing example)
Make sure that all corners and ends of the fortified tape is fixed to the cardboard box NO! ________________________________________________________________________________________________ Global Sewage Surveillance Protocol, Appendix 4 -version 2 (January-2016) -Page 5 of 5 The package is now correctly packed and ready for sending, using the description in Appendix 2 of the Global Sewage Surveillance protocol.
Should you need clarification to any of the steps mentioned above, please send an email directly to suska@food.dtu.dk (Susanne Karlsmose Pedersen). ---------

(packing example)
This Purpose A point-prevalence metagenomic analysis will be applied to sewage samples collected globally from the main sewage system of major cities prior to treatment plants inlet. The project will serve as proof-of-concept for applying metagenomic approaches, which could initiate a global surveillance of human infectious diseases including antimicrobial resistance from sewage collected in major cities around the world to detect, control, prevent and predict human infectious diseases.

Procedure
From each location, two representative sewage samples (1L each, in total 2L) are collected from the main sewage flow on consecutive days from the city's main sewage pipelines prior to waste water treatment plant inlets or from the main outlet to rivers or similar. Samples can be obtained following the first filtering step, but it is important that there has been no processing of the sewage. 14. For SAMPLE NUMBER 1 -How would you characterize the viscosity of the sample? (Please indicate 1, 2, 3, 4 or 5, following the graduation where 'like water' is '5' and 'solid' is '1')