Liying Low1,2, Pablo Fuentes-Utrilla3, James Hodson4, John D O'Neil5, Amanda E Rossiter6, Ghazala Begum5,7, Kusy Suleiman1, Philip I Murray1,2, Graham R Wallace1,2, Nicholas J Loman3, Saaeha Rauz1,2. 1. Academic Unit of Ophthalmology, Institute of Inflammation and Ageing, University of Birmingham, Birmingham, West Midlands, UK. 2. Birmingham and Midland Eye Centre, Sandwell and West Birmingham Hospitals National Health Service (NHS) Trust, Birmingham, West Midlands, UK. 3. MicrobesNG/School of Biosciences, University of Birmingham, Birmingham, West Midlands, UK. 4. Queen Elizabeth Hospital, University Hospitals Birmingham NHS Foundation Trust, Birmingham, West Midlands, UK. 5. Institute of Inflammation and Ageing, University of Birmingham, Birmingham, West Midlands, UK. 6. Institute of Microbiology and Infection, University of Birmingham, Birmingham, West Midlands, UK. 7. National Institute for Health Research Surgical Reconstruction and Microbiology Research Centre, Birmingham, UK.
Abstract
BACKGROUND: Microbial keratitis is a leading cause of preventable blindness worldwide. Conventional sampling and culture techniques are time-consuming, with over 40% of cases being culture-negative. Nanopore sequencing technology is portable and capable of generating long sequencing reads in real-time. The aim of this study is to evaluate the potential of nanopore sequencing directly from clinical samples for the diagnosis of bacterial microbial keratitis. METHODS: Using full-length 16S rRNA amplicon sequences from a defined mock microbial community, we evaluated and benchmarked our bioinformatics analysis pipeline for taxonomic assignment on three different 16S rRNA databases (NCBI 16S RefSeq, RDP and SILVA) with clustering at 97%, 99% and 100% similarities. Next, we optimised the sample collection using an ex vivo porcine model of microbial keratitis to compare DNA recovery rates of 12 different collection methods: 21-gauge needle, PTFE membrane (4 mm and 6 mm), Isohelix™ SK-2S, Sugi® Eyespear, Cotton, Rayon, Dryswab™, Hydraflock®, Albumin-coated, Purflock®, Purfoam and Polyester swabs. As a proof-of-concept study, we then used the sampling technique that provided the highest DNA recovery, along with the optimised bioinformatics pipeline, to prospectively collected samples from patients with suspected microbial keratitis. The resulting nanopore sequencing results were then compared to standard microbiology culture methods. RESULTS: We found that applying alignment filtering to nanopore sequencing reads and aligning to the NCBI 16S RefSeq database at 100% similarity provided the most accurate bacterial taxa assignment. DNA concentration recovery rates differed significantly between the collection methods (p < 0.001), with the Sugi® Eyespear swab providing the highest mean rank of DNA concentration. Then, applying the optimised collection method and bioinformatics pipeline directly to samples from two patients with suspected microbial keratitis, sequencing results from Patient A were in agreement with culture results, whilst Patient B, with negative culture results and previous antibiotic use, showed agreement between nanopore and Illumina Miseq sequencing results. CONCLUSION: We have optimised collection methods and demonstrated a novel workflow for identification of bacterial microbial keratitis using full-length 16S nanopore sequencing.
BACKGROUND: Microbial keratitis is a leading cause of preventable blindness worldwide. Conventional sampling and culture techniques are time-consuming, with over 40% of cases being culture-negative. Nanopore sequencing technology is portable and capable of generating long sequencing reads in real-time. The aim of this study is to evaluate the potential of nanopore sequencing directly from clinical samples for the diagnosis of bacterial microbial keratitis. METHODS: Using full-length 16S rRNA amplicon sequences from a defined mock microbial community, we evaluated and benchmarked our bioinformatics analysis pipeline for taxonomic assignment on three different 16S rRNA databases (NCBI 16S RefSeq, RDP and SILVA) with clustering at 97%, 99% and 100% similarities. Next, we optimised the sample collection using an ex vivo porcine model of microbial keratitis to compare DNA recovery rates of 12 different collection methods: 21-gauge needle, PTFE membrane (4 mm and 6 mm), Isohelix™ SK-2S, Sugi® Eyespear, Cotton, Rayon, Dryswab™, Hydraflock®, Albumin-coated, Purflock®, Purfoam and Polyester swabs. As a proof-of-concept study, we then used the sampling technique that provided the highest DNA recovery, along with the optimised bioinformatics pipeline, to prospectively collected samples from patients with suspected microbial keratitis. The resulting nanopore sequencing results were then compared to standard microbiology culture methods. RESULTS: We found that applying alignment filtering to nanopore sequencing reads and aligning to the NCBI 16S RefSeq database at 100% similarity provided the most accurate bacterial taxa assignment. DNA concentration recovery rates differed significantly between the collection methods (p < 0.001), with the Sugi® Eyespear swab providing the highest mean rank of DNA concentration. Then, applying the optimised collection method and bioinformatics pipeline directly to samples from two patients with suspected microbial keratitis, sequencing results from Patient A were in agreement with culture results, whilst Patient B, with negative culture results and previous antibiotic use, showed agreement between nanopore and Illumina Miseq sequencing results. CONCLUSION: We have optimised collection methods and demonstrated a novel workflow for identification of bacterial microbial keratitis using full-length 16S nanopore sequencing.
Authors: Mónika Szabó; Tibor Nagy; Tímea Wilk; Tibor Farkas; Anna Hegyi; Ferenc Olasz; János Kiss Journal: Antimicrob Agents Chemother Date: 2016-10-21 Impact factor: 5.191
Authors: K Schmidt; S Mwaigwisya; L C Crossman; M Doumith; D Munroe; C Pires; A M Khan; N Woodford; N J Saunders; J Wain; J O'Grady; D M Livermore Journal: J Antimicrob Chemother Date: 2016-09-25 Impact factor: 5.790
Authors: N R Faria; J Quick; I M Claro; J Thézé; J G de Jesus; M Giovanetti; M U G Kraemer; S C Hill; A Black; A C da Costa; L C Franco; S P Silva; C-H Wu; J Raghwani; S Cauchemez; L du Plessis; M P Verotti; W K de Oliveira; E H Carmo; G E Coelho; A C F S Santelli; L C Vinhal; C M Henriques; J T Simpson; M Loose; K G Andersen; N D Grubaugh; S Somasekar; C Y Chiu; J E Muñoz-Medina; C R Gonzalez-Bonilla; C F Arias; L L Lewis-Ximenez; S A Baylis; A O Chieppe; S F Aguiar; C A Fernandes; P S Lemos; B L S Nascimento; H A O Monteiro; I C Siqueira; M G de Queiroz; T R de Souza; J F Bezerra; M R Lemos; G F Pereira; D Loudal; L C Moura; R Dhalia; R F França; T Magalhães; E T Marques; T Jaenisch; G L Wallau; M C de Lima; V Nascimento; E M de Cerqueira; M M de Lima; D L Mascarenhas; J P Moura Neto; A S Levin; T R Tozetto-Mendoza; S N Fonseca; M C Mendes-Correa; F P Milagres; A Segurado; E C Holmes; A Rambaut; T Bedford; M R T Nunes; E C Sabino; L C J Alcantara; N J Loman; O G Pybus Journal: Nature Date: 2017-05-24 Impact factor: 49.962
Authors: Christian Quast; Elmar Pruesse; Pelin Yilmaz; Jan Gerken; Timmy Schweer; Pablo Yarza; Jörg Peplies; Frank Oliver Glöckner Journal: Nucleic Acids Res Date: 2012-11-28 Impact factor: 16.971
Authors: James R Cole; Qiong Wang; Jordan A Fish; Benli Chai; Donna M McGarrell; Yanni Sun; C Titus Brown; Andrea Porras-Alfaro; Cheryl R Kuske; James M Tiedje Journal: Nucleic Acids Res Date: 2013-11-27 Impact factor: 16.971
Authors: Paulo J M Bispo; Samaneh Davoudi; Matthew L Sahm; Ai Ren; John Miller; John Romano; Lucia Sobrin; Michael S Gilmore Journal: Invest Ophthalmol Vis Sci Date: 2018-01-01 Impact factor: 4.799
Authors: Clarisse A Marotz; Jon G Sanders; Cristal Zuniga; Livia S Zaramela; Rob Knight; Karsten Zengler Journal: Microbiome Date: 2018-02-27 Impact factor: 14.650
Authors: Raf Winand; Bert Bogaerts; Stefan Hoffman; Loïc Lefevre; Maud Delvoye; Julien Van Braekel; Qiang Fu; Nancy Hc Roosens; Sigrid Cj De Keersmaecker; Kevin Vanneste Journal: Int J Mol Sci Date: 2019-12-31 Impact factor: 5.923