Skip to Content

Cluster 1 Summary

Cluster 1 includes WP 2, 3, 4 and 5. The primary objective in this cluster is to identify appropriate molecular targets or markers and gene sequences that could be used to provide insights into the presence of pathogens in water, their viability and virulence or could be used for strain discrimination in microbial typing and source tracking. Because the issues facing microbiologists working with viruses, bacteria and protozoa can be quite different we have chosen to have separate work packages covering each of these categories of microorganism. In addition we have an independent WP on source tracking as this could include all three categories and also non microbiological markers. We have restricted our research to pathogens acquired orally. Targeted pathogens include: Norovirus, Rotavirus, Adenovirus and Hepatitis A and E viruses (WP2), A. butzleri, C. coli, C. jejuni, V. cholerae, P. aeruginosa, S. enterica, pathogenic E. coli and indicator bacteria (WP3), and Cryptosporidium, Giardia and Toxoplasma (WP4).


WP2: Active Months - February 2013 to January 2016


1.     To establish threshold maximum acceptance levels for genome copies of health-significant waterborne viruses that cannot be grown in cell cultures so that they represent an acceptable risk for the consumers.


2.     To indirectly ascertain the infectivity of viruses detected by molecular procedures. Several sample pre-treatments prior to molecular detection will be evaluated in order to provide a better estimation of infectious virus numbers. Additionally systems to evaluate specific virus nucleic acid or protein damage caused by some virucidal treatments will be developed.


3.     To determine the genetic diversity within the target viruses and how such diversity could be exploited in outbreak investigation to link human and environmental strains. Genetic characterization of virus isolates will determine the potential zoonotic origin of waterborne viruses.


Task 2.1 Establish robust quantitative molecular multiplex systems for detection of health-significant enteric viruses (UB, DTU-FOOD, IST SLU, UH)

A quadruplex Real-Time RT-qPCR assay fulfilling the requirements of the method developed by the European Committee on Standardization (CEN) was developed by the UB participants for the simultaneous quantitative detection of hepatitis A virus (HAV), norovirus (NoV) GI and GII, and mengovirus (used as process control for determination of the virus/nucleic acid extraction efficiency). The method was published in C. Fuentes et al., Food Microbiology, 2014, 40: 55).

While at IST several different combination RT-qPCR assays for different viral pairs were created (HAV+NoV; HAV+hepatitis E virus -HEV; NoV+rotavirus-RV), at SLU, VOCMA originally developed for detection on Luminex has been transferred to protocols suited for detection with RT-qPCR for NoV GI and II, sapovirus (SaV), HAV and HEV, creating one calici and one hepatitis panel respectively. Additionally, the use of internal quenchers has been investigated for NoV GII showing great potential for internal zen-quenchers and detection of a single target copy (manuscript submitted to Biotechnology).

At UH, duplex real-time RT-qPCR using NoV GI and GII primer-probe sets labelled with VIC and FAM were tested. The sensitivity of the duplex assay was compared to those of monoplex GI and GII assays with 10-fold dilutions of NoV GI.6 and GII.4 nucleic acid. In the conditions used, the sensitivity was comparable to that of monoplex, although the Cq-values for both genotypes obtained by duplex assay were about 2 cycles higher than Cq-values by monoplex assays. After some adjustments, multiplex assays may be suitable for testing viruses in water samples.


Task 2.2 Investigate pre-treatments of target viruses in order to provide a better estimation of infectivity through genome copy determination (UB, DTU-FOOD, IST, SLU, UH)

Methods involving the use of fluorescent dyes such as propidium monoazide (PMA) or ethidium monoazide (EMA), as well as mucin-coated beads have been used for the live-dead differentiation in molecular tests.

In the fulfillment of this task, PMA treatment before RT-qPCR amplification was optimized in all participant laboratories for the detection and quantification of viruses with intact cohesive undamaged capsids. As an example, low FC doses up to 2.5 mg/L only affected the infectivity assay with a signal reduction of 1.34±0.45 log. No effects were observed in any of the molecular tests used (RT-qPCR alone, PMA+RT-qPCR or PMA/Triton+RT-qPCR). With higher doses up to 5 and 10 mg/L of free chlorine, infectivity was reduced over 4.5 logs, and in most cases viral titers decreased below the detection limit. Log reductions obtained by molecular tests were higher when PMA was included, but TritonX100 did not result in a significant increase in log reduction.

Again as an example, after heat treatment at 70ºC, infectivity was reduced 2.48±1.30 logs. Treatment at 85ºC and 99ºC resulted in a loss of 3.58±0.32 and 4.50±0.58 logs of infectious viruses, respectively. Despite this high effect on infectivity, RT-qPCR alone only caused a reduction lower than 1 log in all cases, confirming that the main target for inactivation using high temperature is the viral capsid. When assayed by PMA/Triton+RT-qPCR, log reductions obtained at 85ºC and 99ºC were 2.81±0.38 and 3.63±0.48, respectively, which only differ in less than 1 log with infectivity log reductions.

However, quantification by PMA-RT-qPCR showed a limited and varying distinction between heat treated non-intact and non-heat treated intact viral particles with the best effect for samples initially containing moderate levels of viruses (3.64 ×102 – 2.71×105 GC or RT-qPCR units). Thus a mean reduction of 1.53 ± 0.68 Log (spanning from 1.18 Log for SaV to 2.05 log for MNV) could be demonstrated for samples containing viruses in levels ≥ 3.64 ×102 GC or RT-qPCR units. Interpretation of data for samples with an initial viral load ≤ tLOQ (approximately 2×102 genome units) was difficult due to uncertainty of the RT-qPCR assays. In conclusion, these data show that the application of PMA-RT-qPCR provides a limited and insufficient distinction between viral RNA from heat inactivated and infectious virions of NoV, SaV and MNV. The efficiency of the method depends on the type and level of virus in the sample.

Additionally, DTU-Food have tested PGM-coated beads (mucin-conjugated beads) for the efficiency to capture viruses with intact capsids present in different levels without treatment and after being heat treated at 80ºC for 10 minutes. The application of PGM-RT-qPCR on heat treated samples resulted in overall reductions in detectable viral genomes from 58.51 to 3.97% for NoV GI and from 68.69 to 4.80% of NoV GII. For MNV, a reduction from 12.27 to 1.00% PCR detectable genomes could be observed in the heat treated suspensions. The loss of intact viruses during PGM capture was determined by quantifying the remaining part of intact viruses present in the supernatant subjected to RNAse treatment prior to nucleic acid extraction. This indicated an escape from PGM capture of a smaller fraction of viruses with intact capsids which might be due to destroyed surface proteins.    

The data confirms that capturing of viruses using PGM coated beads, may indeed facilitate selection of viruses with intact surface proteins.


Task 2.3 Investigate specific virus damage through some specific virucidal treatments (UB)

Work at the UB has been focused at elucidating the effect of different concentrations of free chlorine (FC) on the genome of human norovirus (NoV) GII.4 (New Orleans1805/2009/USA).

Primers were designed in order to cover the full-length NoV genome in 15 fragments. The experimental design was to apply to the NoV suspension increasing concentration of FC (from 0 to more than 9 mg/l) for 30 min at room temperature in the dark. Our data point is a differential degradation of the NoV genome segments. Some segments are more resistant than others to the effect of FC. These findings open the possibility to employ specific genome fragments for different purposes; e.g., a resistant fragment to trace the presence of contaminant NoV in a sample NoV-tracking), or a sensitive fragment to ascertain the inactivation of NoV after FC treatment.

At DTU-Food, inactivation studies of MNV in suspensions and on surfaces using electrolyzed oxidizing water (EOW) produced in hand hold spraying bottles had been conducted to test the efficiency to inactivate MNV, NoV GI nd NoV GII. Results from suspension tests applying 0.5 and 1 minute contact time have shown 1 and 2 log reduction in MNV plaque forming units (PFU) using 100 or 200 ppm AFC, respectively. Results from surface tests on MNV dried on stain less steel disks applying 0.5 minute contact time with liquid EOW, showed 1 and 2 log reductions in MNV PFU using 200 or 400 ppm AFC, respectively, and >4 log10 reductions using 1, 5 and 10 minutes contact times combined with either 200 and 400 ppm AFC. Applying sprayed EOW on stainless steel surfaces with dried MNV required 10-20 minutes contact time to reduce MNV by 3-4 log. The determination of the reductions in MNV, NoV GI and NoV GII genomes using traditional RT-qPCR has been delayed to autumn 2015. And so has the determination of potential genomic hot spots sensible to EOW using e.g. the newly developed PCR assays which amplify the whole norovirus genome in 15 segments (UB lab, unpublished).


Task 2.4 Generate typing tools for health significant human and animal viral pathogens in clinical and environmental samples (UB, DTU-FOOD, IST, SLU, UH)

Several genotyping tools have been developed in this Task, enabling the generation of genotyping tools applicable to samples, usually water, environmental or food samples, with a low quantity of viruses (Sabrià et al., J.  Clin. Microbiol. 2014; D'Andrea et al., Int. J. Mol. Sci. 2015; Kantala et al., Foodborne Pathogens and Disease 2015; Jalava et al., PLoS One. 2014; Oristo et al., Food Environ Virol.2016; Kauppinen et al., in Press 2016).

One study of the epidemiology of food and waterborne outbreaks of norovirus gastroenteritis occurring in Catalonia during 2010-2012 compared clinical features and levels of viral shedding of the most prevalent GII.4 2012 variant with its predecessor. In another study focused on the effect of a universal hepatitis A vaccination program among preadolescents implemented in Catalonia, Spain, during the period of 1999–2013, revealed the emergence of genotype IC during a foodborne outbreak, the short-lived circulation of vaccine-escape variants isolated during an outbreak among the men-having-sex-with-men group, and the association of genotype IIIA associated with the increase of symptomatic cases among the very young.

At DTU-Food, work on the development on NGS- pre-processing has been carried out on a NoV GII.1 positive stool sample adding mengovirus as an index. The use of endonuclease was very important to remove extracellular DNA/RNA, increasing the fraction of reads mapping to viruses from ~2 % to ~40 %. It was found necessary to amplify the extracted RNA/DNA quite a lot (40 PCR cycles) to reach concentrations suitable for NGS. This method has currently been applied on African sewage samples, where preliminary data shows the presence of more than 450 different virus species, including NoV, enteroviruses (EV) and RV. However the most numerous viruses that could be identified are plant viruses associated with human feces and bacteriophages. Since these viruses constitute the biggest part of the sewage viral community, deep sequencing is needed (~1,000,000 reads per sample) to identify human and animal pathogens present in a sewage sample. Highlight significant results of your work package for the 2nd period:


  • More reliable singleplex and multiplex molecular assays for detection of waterborne viral agents.
  • Methods for better discrimination of infectivity through determination of genome copies.
  • Procedures to estimate the inactivation of non-cultivable viruses.
  • Molecular typing methodologies for characterization of waterborne viral outbreaks.




WP3: Active Months - February 2013 to January 2016

Molecular tools were developed for the identification and quantification of the major waterborne bacterial pathogens and successfully tested with water samples representing the drinking water supply chain.

Deliverable D3.5 
Analytical modules for targeted pathogens ready for integration into platforms:  Submission date: January 30, 2016
Authors: Manfred Höfle, HZI, Germany; Carla Pruzzo, UNIGE, Italy; Maria Jose Figueras, URV, Spain; *Antonio Martinez-Murcia, GPSTM, Spain; Jörg Peplies, Ribocon, Germany. 
*Lead beneficiary: Genetic PCR SolutionsTM, Alicante, Spain.


Deliverable 3.5 - Analytical modules for targeted pathogens ready for integration into platforms

1. Executive Summary 

The overall aim of WP 3 was to develop analytical modules for the identification and quantification the targeted bacterial pathogens which can be adapted in the platforms to enable commercialisation. These modules consist of sets of Standard Operational Protocols (SOPs) for the targeted pathogens to identify and quantify them at the genus, species or strain level. At a higher Technology Readiness Level (TRL) molecular test kits, containing all chemicals, enzymes and molecular standards for these tasks, were developed and validated in the relevant environments. 

Four different types of molecular markers/targets were explored for the identification and quantification of bacteria to the genus until the strain level: i) 16S ribosomal RNA genes, ii) highly conserved but variable functional genes (gyrB and rpoB), iii) specific virulence genes and iv) tandem repeats. The applicability of these four types of markers for molecular quantification of bacterial pathogens was assessed at three levels of validation: i) in-silico meaning comparison with International sequence data bases, ii) in-vitro meaning validation in the laboratory using reference strains, and iii) in-situ meaning validation with real water samples. The 16S rRNA gene, currently the gold standard for bacterial taxonomy, was successfully applied to identify the targeted bacteria at the genus and species level.  In combination with Next Generation Sequencing (NGS) technologies, genus-specific identification and quantification was demonstrated for all relevant species of the genus Pseudomonas, including the targeted species P. aeruginosa. In addition, pan-bacterial primers targeting the 16S RNA gene were applied in combination with NGS technologies to obtain a complete overview about all bacteria present in a given water sample to achieve the “detection of the unknown”. For the functional genes, the gyrase B (gyrB) gene was most promising because it could be demonstrated in-silico and in-vitro that it is possible to differentiate the major serovars of Salmonella enterica and identify clearly Escherichia coli and Shigella flexneri. In-situ application in a set of contaminated water samples confirmed this high resolution if specific primers were applied in a patho-printing approach using Illumina technology. New primers for the high resolution identification of Arcobacter butzleri and Campylobacter jejuni targeting the RNA polymerase B (rpoB) gene were developed and validated in-vitro using reference strains. Virulence genes for all targeted pathogenic species were assessed and validated for quantification by qPCR. A novel phylogenetic marker for a virulence related gene was discovered for Vibrio cholerae based on which a highly specific qPCR assay was developed and validated. The value of tandem repeats was explored in a MLVA (Multiple-Locus Variable number tandem repeat (VNTR) Analysis) approach for Pseudomonas aeruginosa and Legionella pneumophila because of their commercial relevance. Both MLVA assays could be validated in-vitro with a comprehensive set of reference strains using multiplex PCR with fluorescent primers and capillary electrophoresis (CE) for detection and separation of the amplicons. 

Molecular quantification kits and genome standards for all targeted species were validated and are now commercially available from SMEs. In conclusion, this deliverable report demonstrates that WP 3 achieved the final Milestone MS 26 “Molecular tools for high resolution identification” of bacterial targeted species & genus-specific bacterial fingerprints available for integration into platforms developed in Cluster 2 and for field trails in Cluster 3. 

Implications of the results of Deliverable Report D3.5

Implications of the results for the Work Package (WP 3) 
This deliverable report provides a large variety of Standard Operational Protocols (SOPs) to identify and quantify the targeted bacterial pathogens at the genus, species or strain level. Molecular test kits, containing all chemicals, enzymes and molecular standards for these tasks, were developed and validated. In essence, it is the final report of the major results of WP3.

Implications of the results for the Cluster 1, Molecular detection of bacterial pathogens 

Overall, D3.5 comprises the results of work at a higher Technology Readiness Level (TRL) that fulfil the aim of WP 3 within Cluster 1, related to the bacteria selected in the project. This deliverable report delivers a set of analytical modules developed for the identification and quantification of the targeted bacterial pathogens which can be adapted in the platforms developed of Cluster 2 to enable commercialisation.

Implications of the results for the whole project 

This deliverable report provides the analytical modules developed in WP3 at two different Technology Readiness Levels (TRL): i) molecular identification and quantification technologies validated in relevant environments and SOPs developed, and ii) commercially available molecular test kits and molecular standards validated in relevant environment.  In conclusion, this deliverable report demonstrates that WP 3 achieved the final Milestone MS 26 “Molecular tools for high resolution identification” of bacterial targeted species and genus-specific bacterial fingerprints available for integration into platforms developed in Cluster 2 and for field trails in Cluster 3.

Indicate key external stakeholders interested in the results of Deliverable Report 3.5 
SOPs are of interest to partners of Cluster 2 and 3, and consequently to external stakeholders (large and small water suppliers, food industry, and regulators). Molecular quantification kits and genome standards for all targeted bacterial species were validated and are now commercially available from SMEs which should be specifically advertised by the coordinator.

Which internal partners should your deliverable be sent to?   
All the 39 internal partners in Aquavalens should receive a copy, it will be particularly important to those active in Clusters 2 and 3. This deliverable report will help with planning the integration of molecular assays for the detection of bacteria into the platforms developed in Cluster 2 and help perform the field trials in Cluster 3.     



International Publications of WP 3 until January 31, 2016

  • Fisher, J.C., Levican, A., Figueras, M.J., McLellan., S.L. 2014. Population dynamics and ecology of Arcobacter in sewage. Frontiers in Microbiology 5: 525; doi:10.3389/fmicb.2014.00525
  • Figueras, M.J., Beaz-Hidalgo, R., Hossain, M.J., Liles, M.R. 2014.Taxonomic affiliation of new genomes should be verified using average nucleotide identity and multilocus phylogenetic analysis. Genome Announcements 2: e00927-14; doi:10.1128/genomeA.00927-14 
  • Figueras, M.J., Levican, A., Pujol, I., Ballester, F., Rabada, Quilez, M.Jm, Gomez-Bertomeu, F. 2014. A severe case of persistent diarrhoea associated with Arcobacter cryaerophilus but attributed to Campylobacter sp. and a review of the clinicalinc idence of Arcobacter spp.. New Microbes New Infect. 2:31-37; doi: 10.1002/2052-2975.35. 
  • Hossain, M.J., Beaz-Hidalgo, R., Figueras, M.J., Liles, M.R. 2014. Draft genome sequences of two novel Aeromonas species recovered in association with cyanobacterial blooms.  Genome Announcements 2: e01181-14; doi: 10.1128/genomeA.01181-14
  • Vezzulli L., E. Pezzati, I. Brettar, M. G. Höfle, Pruzzo, C. 2015. Effects of global warming on Vibrio ecology. Microbiol. Spectrum 3: VE-0004-2014 doi:10.1128/microbiolspec.VE-0004-2014
  • Beaz-Hidalgo, R., Hossain, M.J., Liles, M.R., Figueras, M. J. 2015. Strategies to avoid wrongly labelled genomes using as example the detected wrong taxonomic affiliation for Aeromonas genomes in the GenBank database.  PLoS ONE 10:e0115813; doi:10.1371/journal.pone.0115813
  • Levican A, Rubio-Arcos S, Martinez-Murcia A, Collado L, Figueras M. J. 2015. Arcobacter ebronensis sp. nov. and Arcobacter aquimarinus sp. nov., two new species isolated from marine environment. Systematic and Applied Microbiology    38:30-35; doi:10.1016/j.syapm.2014.10.011.
  • Martinez-Murcia, A., Lamy, B. 2015. Molecular Diagnostics by Genetic Methods.     Aeromonas, Caister Academic Press, Chapter 8, p. 180-195, ISBN: 978-1-908230-56-0
  • Vezzulli L, Stauder M, Grande C, Pezzati E, Verheye HM, Owens NJ, Pruzzo C. 2015. gbpA as a Novel qPCR Target for the Species-Specific Detection of Vibrio cholerae O1, O139, non-O1/non-O139 in environmental, stool, and historical continuous plankton recorder samples. PLoS ONE 10: e0123983; doi:10.1371/journal.pone.0123983
  • Lesnik R., Brettar I., Höfle MG. 2015. Legionella species diversity and dynamics from surface reservoir to tap water: from cold adaptation to thermophily. ISME Journal; DOI: 10.1038/ismej.2015.199    
  • Martínez-Murcia A., Beaz-Hidalgo R., Navarro A., Carvalho M.J., Aravena-Román M., Correia A., Figueras M.J., Saavedra M. J. 2016. Aeromonas lusitana sp. nov., isolated from untreated waters and vegetables. Current Microbiology, 72:795–803

WP4: Active Months - February 2013 to January 2016

This work package has validated and applied molecular targets for detection and human infectivity potential of the protozoan parasites; Cryptosporidium, Giardia and Toxoplasma, and investigated targets that might be associated with virulence. 

Deliverable D4.4 (Version 2, 02/12/15)

Molecular tools for parasite lineage to support source tracking in outbreak investigations: Authors: Rachel Chalmers, Sophie May, Guy Robinson, Gregorio Perez, Stephen Hadfield (PHL), Frank Katzer, Alison Burrells, Beth Wells, Emily Hotchkiss, Elisabeth Innes (MRI), Kevin Tyler, Johanna Nader, Maha Bouzid (UEA).

D4.4 Executive Summary

Waterborne outbreaks of human disease have occurred caused by the protozoan parasites Cryptosporidium species, Giardia duodenalis and Toxoplasma gondii in drinking water.  To investigate sources of contamination, transmission routes and outbreaks of disease, high resolution genotyping is needed. To underpin the development of standardised, multi-locus schemes, this work focussed mainly on the major human pathogens Cryptosporidium hominis and Cryptosporidium parvum, especially the zoonotic C. parvum. The initial focus was on multiple-locus variable-numbers of tandem repeats analysis (MLVA). Potentially-useful loci were identified and examined in silico for suitability for inclusion in inter-laboratory MLVA schemes. An economical capillary electrophoresis (CE) platform (QIAxcel, Qiagen) was optimised for performance with single round PCR amplicons, and epidemiological utility was demonstrated by the investigation of a panel of sporadic and outbreak-related clinical samples. The loci were tested in vitro in different laboratories using different CE machines through the distribution of diverse C. parvum DNA. Although many loci were more complex than previously thought, the use of sequenced reference standards enabled consistent allele calling based on adjusted fragment sizes. Additionally, the Cryptosporidium whole genomes generated in Aquavalens D4.1 were investigated for new subtyping markers, and suitable candidates were identified for potential inclusion in MLVA.

For T. gondii, we developed and tested a multiplex-nested-PCR-RFLP assay for genotyping based on five markers that originate from four distinct gene loci within the parasite genome. Although sensitive, being able to detect single parasites within a PCR reaction, the assay was less sensitive than the 529 bp qPCR used for T. gondii enumeration. The assay did not provide enough discriminatory power to distinguish between Type II strains of the parasite which are the most commonly found strains within Europe.

WP5: Active Months - April 2013 to January 2016

A data matrix based on the first sampling results has been created. This was used to develop MST predictive models and select MST markers for further testing in 2015.

Deliverable D5.4

MST predictive models validated at European level.

Authors: Leena Maunula1, Anicet R. Blanch2, Lluis A. Belanche3,

1 University of Helsinki, Finland, 2 University of Barcelona, Spain

3  Polytechnic University of Catalonia, Spain


Abbreviations:  Mitochondrial DNA markers: Human-specific (HMMit), Porcine (PGMit), Bacteroidetes DNA: Human-specific (HF183Taqman), Ruminant (BacR), Porcine (Pig2Bac), Bifidobacterium DNA: Human-specific (HMBif), Bovine (CWBif), Poultry (PLBif), Total (TLBif), Norovirus RNA (NoV)

Results and Discussion

The selected predictive models were evaluated for prediction and their corresponding accuracy was estimated by repeated rounds of k-CV, as explained above. Furthermore, four validated MST predictive models at the five participant countries (Austria, Germany, Finland, Portugal and Spain) were selected based on molecular markers only, in order to be considered for future use on technological platforms.

These selected models determine a subset of nine molecular parameters (see below and in abbreviations) to be considered as preliminary candidates for the final list of molecular targets on MST due to the end of the WP5 (Month 32). However, the final list of selected MST markers to include in the technological platforms will be confirmed through the second sampling campaign.

The validated MST molecular-based models (linear or non-linear) and the corresponding scenarios providing solutions are:

Linear Discriminating Analysis

Human vs Non-human  /  point source

HMBif, PLBif, TLBif, Pig2Bac, HF183TaqMan  (accuracy 100% by LOOCV)

Four sources  /  point source

HMBif, BacR, Pig2Bac  (accuracy 100% by LOOCV)

Non-linear Analysis (Random Forest)

Human vs Non-human

NoV, PGMit (accuracy 88.5% by 10x10 CV)

Four sources

NoV, PGMit, CWBif, HMBif, TLBif/CWBif (accuracy by 79.1% by 10x10 CV)

Although adding some additional molecular markers to the subsets selected could increase accuracy, it could not be worthy when considering costs. If other combinations of molecular markers are requested with the purpose of implementation on technological platforms, it could be possible to select other predictive models out of the already performed calculations at close accuracy values.