Rapid metagenomic sequencing for diagnosis and antimicrobial sensitivity prediction of canine bacterial infections
1The Roslin Institute, University of Edinburgh, UK
2Royal (Dick) School of Veterinary Studies, University of Edinburgh, UK
*Corresponding author and email address J. Ross Fitzgerald, ross.fitzgerald@roslin.ed.ac.ukAbstract
Antimicrobial resistance is one of the greatest current threats to human and animal health. There is an urgent need to ensure that antimicrobials are used appropriately to limit the emergence and impact of resistance. In the human and veterinary healthcare setting, traditional culture and antimicrobial sensitivity testing is typically conducted, requiring 48-72 h to identify appropriate antibiotics for treatment. In the meantime, broad-spectrum antimicrobials are often used, which may be ineffective or impact non-target commensal bacteria. Here, we present a rapid diagnostics pipeline, involving metagenomic Nanopore sequencing directly from clinical urine and skin samples of dogs. We have optimised this pipeline to be versatile and easily implementable in a clinical setting, with the potential for future adaptation to different sample types and animals. Using our approach, we can identify the bacterial pathogen present in a sample with 100% sensitivity within 5 hours. For urine samples, we can predict antibiotic sensitivity with up to 95% accuracy. However, skin swabs which exhibited lower bacterial abundance and higher host DNA, were less amenable and an additional host depletion step may be required prior to DNA extraction. In summary, our pipeline represents an important step towards the design of individually tailored veterinary treatment plans on the same day as presentation, facilitating effective use of antibiotics and promoting antimicrobial stewardship.
Impact statement
Antimicrobial resistance (AMR) is a major threat to veterinary and human healthcare. It is a one-health problem, as humans and dogs are in close contact, require similar antibiotics, and share bacterial pathogens and AMR genes. Limited treatments options due to AMR would have a catastrophic effect. The risk of infection would render much of modern healthcare (including critical care, orthopaedic and complex surgeries, implants and oncology) impossible. In addition, routine infections could become life threatening. It is therefore critical to preserve the efficacy of these drugs for the future. Inappropriate antimicrobial use is the single biggest factor driving AMR. Antimicrobial stewardship involves reducing antimicrobial use, using first-line narrow-spectrum drugs, and avoiding overly long treatment. Delays in culture-based diagnosis lead clinicians to speculatively use broad-spectrum antibiotics and prolong courses of treatment beyond clinical cure. Our rapid diagnostic approach will have a major impact in reducing, refining and replacing antibiotic use. This will advance antimicrobial stewardship in veterinary and human healthcare.
Data summary
All sequencing data mentioned in this work is available from NCBI, BioProject PRJNA925092, Biosamples SAMN32880396 to SAMN32880438, run accessions SRR23195371 to SRR23195413.
The authors confirm all supporting data, code and protocols have been provided within the article or through supplementary data files.
Article notes
Competing Interest Statement
The authors have declared no competing interest.
Introduction
Antimicrobial resistance (AMR) levels amongst human and veterinary bacterial pathogens are escalating globally, to the extent that the World Health Organization now classifies AMR as one of the biggest threats to global health, food security and development (1, 2). The current gold-standard diagnostic methods used in veterinary practice are culture-based or involve remote service providers for PCR, requiring several days to yield results. Broad-spectrum or inappropriate antibiotics are often started while waiting for the results; this inappropriate antimicrobial use is a major driver for AMR (3). Furthermore, animals with suspected highly contagious or potentially zoonotic infections like leptospirosis may require quarantining while waiting for a diagnosis leading to escalated costs, wasted resources, and unnecessary stress for the quarantined animal (4).
Novel methods that enable sensitive, same-day rapid diagnosis and prediction of antibiotic sensitivity across different infection types are required. One method which is of particular interest is metagenomic whole genome sequencing (WGS), involving the extraction of all genomic DNA (gDNA) present in a sample and subsequent identification of pathogens by unbiased DNA sequencing. One of the main sequencing platforms for this kind of rapid diagnosis is the MinION, produced by Oxford Nanopore Technologies (ONT), which has been tested with clinical sputum, endotracheal aspirate, blood, urine and other sample types (5-13). Importantly, the MinION’s small size and relative affordability could enable its usage in a wide variety of clinical settings. In addition, sequence data is accessible in real-time as it is produced, reducing the time required to identify pathogens and their AMR genotypes to as little as 10 minutes (14, 15). The long reads produced by nanopore sequencing readily facilitate whole genome assembly which potentially allows the linkage of AMR genes to specific bacterial strains (16).
Here, we set out to develop a rapid, culture-free metagenomic sequencing pipeline to identify pathogens and predict AMR in canine samples in a veterinary hospital setting. To initially develop this pipeline, we chose to focus on two common canine infections: urinary tract infections (UTIs) and skin infections (pyoderma). Antibiotic therapy is often the first line of treatment in these infections, but AMR, including multi-drug resistance, is frequently observed, and increasing in prevalence (17-21). Furthermore, recurrent infections are common, leading to frequent return visits to the clinic and further courses of antibiotics (22, 23). A rapid, sensitive method for diagnosing the bacterial pathogens and predicting their antimicrobial sensitivity could therefore prevent the use of inappropriate antibiotics, and limit the amount of clinical care required. Although we focussed on urine and skin swab samples here, we aimed to design a protocol which could be adapted to an array of other sample types (e.g. blood) from infections in other animals including humans. Moreover, the approach could be used in a variety of clinical settings, from small practices to large hospitals. By comparing and optimising a number of different kit-based gDNA extractions and sequencing techniques, combined with community-built DNA analysis tools, we have developed a pipeline which can identify the bacterial species present in clinical samples with around 100% sensitivity and specificity, in as little as 5 hours. It can also predict the antimicrobial resistance phenotype of those species with up to 95% accuracy, in around 8 hours.
Methods
Selection of DNA extraction kit
Three different kits recommended in the literature were tested: DNeasy Blood + Tissue (Qiagen, Hilden, Germany), DNeasy Powersoil (Qiagen) and MagAttract HMW DNA (Qiagen). To test each kit, overnight Escherichia coli CAN-50 growth in Luria-Bertani (LB, ThermoFisher, Massachusetts, USA) broth equivalent to 109 CFUs was spun down at 16,000 xg, and the cell pellet was resuspended in 1 ml healthy dog urine. The urine-cell suspension was then processed using each kit, according to the relevant manufacturer’s instructions for each, finishing with an elution into 50 µl nuclease-free water. The resulting extractions were used to compare the kits in terms of (i) gDNA yield in 50 µl nuclease-free water (quantified by Qubit dsDNA HS kit), (ii) lysis method for Gram-positive species (enzymatic vs bead-beating), (iii) speed, and (iv) cost.
Testing our extraction protocol on different species
The optimised MagAttract + Metapolyzyme extraction protocol was tested on the most commonly isolated species identified in the HfSA’s records: E. coli, S. pseudintermedius, Streptococcus canis, Enterococcus faecalis, Pseudomonas aeruginosa, Proteus mirabilis, Pasteurella canis, Klebsiella pneumoniae, Kocuria kristinae/Kocuria rosea, and Clostridium perfringens. For aerobic species, cells were grown on TSA plates at 37°C for 24h or 72h (Kocuria/Streptococcus), then in TSB medium at 37°C with shaking for 24h or 72h (Kocuria/Streptococcus). For two anaerobic or facultative anaerobic species (C. perfringens and P. canis), cells were grown on TSA plates at 37°C in a sealed box with Anaerogen sachets (Oxoid ThermoFisher) for 24h, then in TSB medium in growth flasks in a sealed box with Anaerogen sachets at 37°C with shaking for 24h. For all species, 3 ml of broth culture was pelleted by centrifuging at 16,000 xg for 3 min, the pellet was resuspended in 3 ml healthy dog urine, then pelleted again in the same way. The pellet was then processed as per the optimised protocol, starting with resuspension in 160 µl buffer P1 and 60 minutes lysis with 20 µl metapolyzyme. DNA was eluted into 50 µl nuclease-free water, and quantified using Qubit’s dsDNA HS kit.
Quality of the extracted DNA
The purity of the extracted DNA was assessed using a Nanodrop spectrophotometer (ThermoFisher) to measure the 260/280nm and 260/230nm absorbance ratios of 22 clinical samples (indicated in Supplementary Table S2). The samples were then cleaned up using the ProNex Size-Selective Purification System (Promega, Wisconsin, USA) according to manufacturer’s instructions, starting with 50 µl DNA and 200 µl ProNex beads, and eluting into 20 µl nuclease-free water at the end of the protocol. The purity of the cleaned-up DNA was then measured again by Nanodrop.
Species identification strategies
For each experiment detailed here, species were identified from sequencing reads using either ONT’s EPI2ME tool (v3.0.1-7052513, https://epi2me.nanoporetech.com/, ONT account required) or Kraken2 (v2.1.1, 25) with one of two custom databases, “pathogens_plus” or “bacteria_plus”. The bacteria_plus database was constructed from all bacterial representative genomes present in the NCBI RefSeq database in November 2022 (4,032 species) plus eight mammalian genomes (Canis lupus familiaris, Homo sapiens, Felis catus, Equus caballus, Oryctolagus cuniculus, Sus domesticus, Bos taurus and Ursus arctos). The “pathogens_plus” database contains 668 genomes of various bacterial, viral, protozoan and fungal pathogens, including the top 100 European human and animal pathogens identified in a 2014 study (26), plus selected other pathogens known to be important in veterinary samples, such as Leptospira and Mycobacterium spp (a full list of species included can be seen in Supplementary Table S3). The pathogens_plus database also contains the same eight mammalian genomes included in the bacteria_plus database. The tool and database used is noted for each step of the protocol development below.
Optimising the sequencing protocol
Two rapid library preparation kits were tested: rapid PCR barcoding (SQK-RPB004, Oxford Nanopore Technologies (ONT), Oxford, UK) and rapid barcoding (SQK-RBK004, ONT). The SQK-RPB004 PCR reaction was carried out on 22 clinical samples (Supplementary Table S2), according to the manufacturer’s instructions, and the concentration of each was measured by Qubit’s dsDNA HS kit before and after the reaction.
All 22 samples prepared with the rapid PCR barcoding kit were subsequently sequenced on MinION R9.4.1 flow cells (ONT) according to the SQK-RPB004 protocol, although ProNex beads were substituted into steps which required AMPure XP beads. Four of these samples, two urines (DTU09 and DTU16) and two skin swabs (SkSw08A and SkSw14), were re-sequenced with the rapid barcoding kit, to compare the results of sequencing with and without the PCR amplification. The DNA was prepared according to the manufacturer’s instructions for SQK-RBK004, except for the substitution of ProNex in place of AMPure XP again. The results of the two library preparation methods were compared in terms of (i) run yield, and (ii) bacterial reads vs eukaryotic host contamination (according to ONT’s online EPI2ME classification “WIMP” tool).
MinION vs Flongle flow cells
Two relatively highly concentrated gDNA samples were used to compare MinION flow cells with Flongle flow cells: a clinical skin swab sample (SkSw21, 55 ng µl-1) and a 1:1 mix of E. coli and P. aeruginosa isolate DNA, both extracted previously during the DNA extraction optimisation trials (80 ng µl-1). Each sample was prepared for sequencing on an R9.4.1 Flongle flow cell according to the Flongle-specific manufacturer’s instructions for SQK-RBK004 (using ProNex beads instead of AMPure XP), then each was sequenced separately for a full 24h on the Flongle device, and the total yield of the 24 h run was noted. More DNA from the same samples was then prepared for sequencing on an R9.4.1 MinION flow cell with SQK-RBK004 as above. Sequencing was started for the first (E. coli + P. aeruginosa) sample on a Mk1C MinION device, and the sequencing was monitored. The sequencing was stopped when the run yield matched that of the 24 h Flongle run, and the time taken to reach that yield was recorded. The MinION flow cell was then washed using the flow cell wash kit (EXP-WSH004, ONT) according to the manufacturer’s instructions, and the second sample (SkSw21) was loaded and sequenced. Again, the sequencing was monitored closely and stopped once the run yield matched that of the 24 h Flongle run.
Determining the optimal MinION flow cell usage strategy
Two flow cells and a variety of DNA samples were used to test how many times a MinION flow cell can be washed and re-used. For the first flow cell, the DNA sample was a relatively high concentration previously extracted E. coli isolate (110 ng µl-1) which had both been stored at 4°C since extraction. An initial sample of 7.5 µl of DNA was prepared for sequencing with SQK-RBK004 as previously described, using barcode01. Sequencing commenced on a fresh flow cell and was stopped when a target of 200 Mbp was reached. The time taken to reach 200 Mbp was recorded, and the flow cell was washed using EXP-WSH004 as previously described. A second 7.5 µl was processed in the same way, with barcode02, followed by another wash with EXP-WSH004. This protocol continued until the time taken to reach 200 Mbp was longer than 2 h, using the next subsequent barcode (barcode03, barcode04, etc.). For each sequencing step, the starting number of available sequencing pores was recorded at the beginning of the run. After each run, the sequenced reads were quality controlled by EPI2ME and NanoStat (v1.6.0, 27), and the percentage of DNA assigned to the wrong barcode was noted.
For the second flow cell, the same protocol was followed, but freshly extracted DNA from clinical samples, ranging from 0.76 to >120 ng µl-1, was sequenced. Each sample was sequenced for up to 3 h, and some samples were sequenced simultaneously (with different barcodes). Full details of the samples sequenced are given in Supplementary Table S4.
Testing adaptive sampling on a GridION to reduce contamination with host DNA
Two flow cells and two DNA samples were used to test the efficiency of adaptive sampling for reducing levels of host DNA in clinical samples. The two DNA samples, one an E. coli isolate and the other a sample of previously sequenced canine DNA, were 75 ng µl-1 each. The samples were mixed in various ratios (90:10, 75:25, 50:50, 25:75 and 10:90). 15 µl of each mix, plus 15 µl of a previously sequenced clinical sample known to be contaminated with host DNA, were prepared for sequencing with SQK-RBK004 (volumes doubled), each with a different barcode, and using nuclease-free water with barcode01 as a negative control. Half of the pooled library was then sequenced on a fresh R9.4.1 flow cell on a GridION, with real-time super accuracy basecalling and no adaptive sampling. The other half was sequenced at the same time on a second fresh R9.4.1 flow cell on the GridION, with real-time super accuracy basecalling and adaptive sampling selected to deplete DNA which mapped to a canine reference genome provided to the software (GCA_014441545.1 ROS_Cfam_1.0).
The resulting datasets for each barcode were analysed using Kraken2 (v2.1.1, 25) with the bacteria_plus database described above. The percentage of reads from each sample assigned to Escherichia (or Shigella) and Canis lupus familiaris were recorded, and the differences between the percentages of each with and without adaptive sampling were tested for significance using a paired Wilcoxon signed-rank test in R (v4.1.2).
Testing clinical samples
Sample selection, DNA extraction and sequencing, and flow cell usage
During the development of this protocol, DNA from 45 clinical urine and skin swab samples was extracted and sequenced, using either the rapid barcoding or rapid PCR barcoding library preparation kit (Supplementary Table S3). These samples were not usually processed on the same day they were received; instead, they were processed and sequenced in batches, and the raw sample and/or extracted DNA was stored at 4°C in the meantime. After these development steps, a further nine urine samples were processed in batches of one to four. These were collected from the HfSA and processed immediately using the finalised extraction protocol (https://www.protocols.io/view/magattract-metapolyzyme-metagenomic-gdna-extractio-chnrt5d6).
The extracted DNA was prepared for sequencing using the rapid barcoding kit (SQK-RBK004) as described above, using barcode 01 for a negative control (nuclease-free water, which was processed using the same extraction protocol as the real samples) and the remaining sequential barcodes for the real samples. Samples were sequenced for 2h, or until 100 Mbp had been sequenced, whichever was sooner. In between sequencing runs, the flow cell was stopped, washed using the flow cell wash kit (EXP-WSH004) according to manufacturer’s instructions and stored with storage buffer at 4°C until the next use.
Species identification and AMR prediction
During sequencing, the data produced underwent preliminary analysis using EPI2ME’s Fastq Antimicrobial Resistance workflow. After sequencing, the data were further analysed, including species identification with Kraken2 (pathogens_plus database), genome assembly with Flye (v2.9-b1774, 28) using the --meta flag and no --genome-size flag, genome annotation using Prokka (v1.14.5, 29) using the appropriate --species and --genus flags along with --compliant, and --usegenus, and AMR prediction using Abricate (v1.0.1, 30) with the NCBI database downloaded on 18 January 2022 and AMRFinderPlus (v3.10.45, 31) with the database version 2022-10-11.2.
During development, large volumes of data were sequenced for some samples. For the analysis described above, these large datasets were randomly down-sampled to 100 Mbp using Rasusa (v0.6.1, 32). The results from all data analyses were collated to predict which species had been present in the original sample, and the AMR phenotypes expected from them. In addition, the length of time taken for each sample was recorded. The results from this pipeline were then compared to the results from the Veterinary Pathology Unit (VPU) at the HfSA, which were produced using the current gold-standard techniques of culture followed by species identification and antibiotic susceptibility testing (AST) using a VITEK® 2 (bioMérieux) instrument.
Results and Discussion
Comparison of DNA extraction methods
A wide variety of DNA extraction methods are available, and many of them have been used in previous studies aiming to develop rapid diagnostic protocols or to extract metagenomic DNA from urine samples (7, 9, 11, 14, 16, 33-37). We selected three commonly used kits to test here (Table 1), based on their previous successful application and their potential flexibility for application to other sample types in the future. A metagenomic approach using enzymatic lysis rather than mechanical cell disruption should result in high molecular weight (HMW) fragments, facilitating better genome assembly and more efficient species and AMR gene identification. Nonetheless, we tested one kit using mechanical disruption, to compare the efficiency of mechanical versus enzymatic lysis.
Of the three kits we trialled, the one with mechanical lysis (the DNeasy Powersoil) was the most expensive per-sample, and produced lowest yield of gDNA when tested (Table 1). The two kits that included optional enzymatic Gram-positive lysis steps were the DNeasy blood + tissue kit and the MagAttract HMW kit, each producing similar yields of DNA in a similar time-frame when tested with E. coli. Although the spin column-based DNeasy kit had a lower cost-per-sample, we ultimately selected the MagAttract in our protocol due to its optimisation towards extracting HMW DNA, and its extensive washing steps that should limit the presence of inhibitors that affect library preparation.
Our optimised protocol can efficiently extract DNA from the bacterial pathogens most commonly identified in canine urinary tract and skin infections
We examined records from the HfSA to identify the array of bacterial species most commonly associated with clinical urine or skin swab samples from dogs (Figure 1, Supplementary Figure S1). Overall, 90% of culture-positive cases were associated with ten different genera, with a total of 53 different genera in total, including 48% Gram-positive and 52% Gram-negative (Supplementary Table S1). Thus, we required a lysis method that would allow the extraction of DNA from a wide variety of different species. Metapolyzyme contains six enzymes optimised for the lysis of bacterial and/or fungal cell walls. Of these six, lysostaphin, mutanolysin and lysozyme should lyse all of the Gram-positive species most commonly seen in the 2018-2019 HfSA data. We determined that 3.3 mg ml-1 of metapolyzyme resulted in better yield of extracted DNA than either lysozyme or lysostaphin for the most common Gram-positive species in skin swabs (Staphylococcus pseudintermedius, Table 2). The optimised MagAttract + Metapolyzyme extraction protocol was next tested on the top ten species identified in urine and/or skin swab samples (Figure 1, Table 3). Each species was grown in appropriate media and spiked into 3 ml healthy dog urine to simulate a clinical sample, and our protocol was successful in extracting micrograms of DNA from all ten species, including anaerobes, Gram-positives, and slow-growing bacteria (e.g. Kocuria). We therefore concluded that the protocol would be efficacious in extracting DNA from the vast majority, if not all, of the species encountered in clinical canine urine or skin swab samples.
Enhancement of DNA purity during extraction
As clinical samples may contain contaminants or inhibitors that impact library preparation or sequencing, we tested if a further clean-up step would be enhance the protocol. Nanodrop 260/280nm and 260/230nm absorbance ratios were used to assess the purity of 22 clinical samples (14 urine, 8 skin swabs, Supplementary Figure S2) before and after a clean-up protocol using the ProNex Size-Selective Purification System. This kit was selected because it should maintain the high molecular weight of the DNA extracted by the MagAttract protocol, and for its affordability relative to the AMPure XP system. After the clean-up, the mean value across the 22 samples was close to ideal at 1.82, although the difference between pre- and post-clean-up ratios was not assessed to be significant according to the Wilcoxon signed rank test for paired samples (p=0.7, n=22). In contrast, the 260/230nm ratios were significantly improved post-clean-up (p=0.004, n=22); the mean value of 1.36 was not optimal but still enabled efficient library preparation. Our findings support the inclusion of the additional clean-up step between DNA extraction and library preparation, despite adding extra time to the final length of the rapid diagnostics protocol. An additional benefit of the clean-up protocol is that by eluting into a final volume lower than the starting volume (20 µl vs 50 µl) the DNA is further concentrated.
Rapid barcoding enables sequencing and identification of species from extracted concentrations of DNA as low as 0.04 ng µl-1
Our experience extracting DNA from the first few clinical samples (Supplementary Table S2) indicated that DNA concentrations may be low in some samples (the mean concentration of the first five skin swab samples was just 4.86 ng µl-1, while the mean from the first five urine samples was 21.98 ng µl-1). Some previous protocols developed for rapid bacterial identification by whole genome nanopore sequencing have used the rapid PCR barcoding library preparation kit, SQK-RPB004 (7, 9, 38, 39). The major benefit of this kit is the amplification of DNA during the PCR step, which may be important for low abundance samples, though with the disadvantage of the additional time required. However, we found that this step resulted in unpredictable yields of DNA and a tendency to amplify host DNA (Supplementary figures S3 and S4). We therefore trialled the rapid barcoding kit without the PCR step (SQK-RBK004) instead, aiming to determine the lowest concentration which could be sequenced and still produce enough usable data to identify selected bacterial species from our samples. We conducted three serial dilutions of DNA samples extracted from cultured E. coli, S. pseudintermedius and S. canis, and found that the relevant bacterial species was identifiable at concentrations much lower than our means of 4.86 (skin swabs) and 21.98 (urine) ng µl-1. For E. coli (Figure 2B) and S. canis (Figure 2D), the original species could be detected above background contamination even at the lowest concentrations tested (0.07 and 0.04 ng µl-1. For S. pseudintermedius (Figure 2C), the original species could be detected above background contamination at 0.16 ng µl-1. We therefore concluded that we could use the rapid barcoding kit to sequence our clinical samples without the need for the PCR amplification step.
Flongle flow cell use and re-use of MinION flow cells
The Flongle is an adapter which fits into MinION or GridION sequencers and allows the use of Flongle flow cells, which are a single-use, lower yield alternative to MinION flow cells. These characteristics of Flongle flow cells are desirable for clinical use: single-use means no potential cross-contamination between different samples, and the lower yield means the flow cells are correspondingly less expensive than MinION flow cells. Clinical applications for which Flongle flow cells are already being used include rapid sequencing of viruses such as SARS-CoV-2 and monkeypox (40, 41), HLA-typing (42, 43), and 16S metagenomics (44, 45). A small number of previous studies have also investigated the use of Flongle flow cells for the rapid identification and typing of bacterial infections (5, 46, 47).
We sought to evaluate the utility of Flongle flow cells in our protocol compared to the classical MinION using two clinical samples (Supplementary Table S4). The volume of DNA sequence data produced by Flongle in a 24h period was matched by MinION within 1.5h for one sample, and 3h for the other. However, although MinION offers greater sequencing speed, the price of each MinION flow cell precludes the use of one flow cell per sample. Barcoding allows multiplexing of up to 12 samples, but in a clinical setting, there may not be sufficient samples to load a full flow cell and still produce results within the desired timeframe. We therefore examined the possibility of reusing MinION flow cells. This would involve sequencing a sample for as long as necessary to produce the sequence data required, stopping the run and performing a DNase wash of the flow cell, then storing the flow cell until another sample was received. In this way, the same flow cell could theoretically be used many times, thus reducing the cost-per-sample without the need for simultaneous sequencing of multiple samples. Accordingly, we aimed to establish i) how many times could we re-use a flow cell to produce sufficient sequence data in a timely manner, and ii) how much residual DNA from previous samples would remain in the flow cell after washing. Employing a cultured E. coli sample, we examined the capacity to produce 200 Mbp of DNA within 2h, and what percentage of the reads had the correct barcode attached. We were able to use the same flow cell eight times before it was exhausted, and the percentage of reads with the correct barcode never fell below 98.8% (Figure 3, Supplementary Table S5). Further, using different barcodes for each subsequent sample reduces the risk of cross-contamination between runs to negligible. We also later decided that, for most samples, only 100 Mbp of sequence will be sufficient for species identification and AMR prediction, hence flow cells could potentially be used even more times.
Importantly, we also trialled a second flow cell, using real clinical urine samples that had been extracted on the same day they were sequenced (Table 4). This second flow cell was able to produce data for nine samples before being used to exhaustion, with samples being sequenced in batches of one to four per run, and over a period of roughly one month.
Ultimately, we estimate that the average MinION flow cell can be used and re-used around eight times to sequence 100-200 Mbp per sample. This slight increase in cost is balanced against producing the data at least eight times faster than it would be on a single-use Flongle flow cell, as well as producing less plastic waste (especially as MinION flow cells can be returned to ONT for recycling, whilst Flongle flow cells cannot).
Adaptive sampling on a GridION significantly reduced host DNA contamination, but not enough to benefit sample analysis
The pipeline developed here is primarily aimed at processing clinical samples and it is therefore likely that many of the DNA samples extracted will contain host DNA alongside the bacterial DNA. In some cases, such as purulent skin swabs or urine samples containing high numbers of white blood cells, the vast majority of the DNA extracted may be derived from the host, which could impact on the accuracy of the diagnostic and AMR prediction from bacterial DNA. Lab-based methods to reduce the numbers of host cells in the sample prior to DNA extraction exist, but can be costly and/or slow, so were not included in this pipeline. As an alternative, a sequencing-based sampling method was trialled.
Adaptive sampling is technique unique to nanopore-based sequencers, in which DNA reads are compared to a reference genome whilst the strand of DNA is still being sequenced. The technique can be used either to enhance or to deplete sequences which map to the reference genome. Due to the computer power needed, MinION sequencers are unlikely to be capable of adaptive sampling, unless connected to a GPU-powered computer. The GridION sequencer, however, is equipped with GPU processors, and adaptive sampling is therefore possible.
Although adaptive sampling did significantly reduce the proportion of eukaryotic reads (p=0.03135, n=6) and increase the proportion of bacterial reads (p=0.03135, n=6) in our paired samples, the absolute differences in percentages were small: 1% fewer eukaryotic reads, and 0.7% more bacterial reads (Supplementary Table S6). These differences would realistically have little effect on our ability to identify bacterial species or detect AMR genes. Where a GridION is available, we would therefore recommend its use in our protocol, but where only a MinION is available, the ability to produce accurate results will not be affected.
Our optimised protocol produces up to 100% sensitivity and specificity for species prediction, within five hours
During the development of this protocol, DNA was extracted from 45 urine and skin swab samples (20 urines and 25 skin swabs). Subsequently, we processed a further nine urine samples using the final optimised protocol, for a total of 54 clinical samples tested (Supplementary Tables S2, S7, and S8, and Table 4).
A wide variety of DNA concentrations were retrieved from these samples, ranging from 0 to >120 ng µl-1 for the urine samples and 0 to 116 ng µl-1 for the skin swabs (Figure 4). As mentioned previously, the lower limit of detection for sequencing was determined to be < 0.1 ng µl-1 of cleaned-up DNA. In total, only five urine samples and five skin swabs produced DNA with lower concentrations than our lower limit and colony forming units (CFU) per ml of sample indicated that those samples had either no growth of any kind, or contained fewer than 1×105 cells. We therefore estimate our lower limit for sufficient DNA from clinical samples to be between 1×105 and 1×106 CFU/ml. The exact lower limit is likely to depend on other sample characteristics, such as the presence of host cells, sediment or inhibitors.
In order to test relevance to a clinical scenario, the nine post-optimisation urine samples were processed in real-time. The average time taken to process the samples through to the commencement of sequencing was just under 5 hours (Table 4 and Supplementary Table S8). Using the online EPI2ME tool, species identification was carried out in real-time as soon as the first sequencing reads were produced. The speed of EPI2ME can be affected by the number of concurrent online users, but we were able to produce preliminary species identification calls within approximately 20 min of sequencing. In each case, we continued to sequence until 100 Mbp of reads had been produced and, to reduce the likelihood of false positive calls, we also performed species identification using Kraken2 after sequencing was complete. Of note, the initial rapid EPI2ME species calls differed from the Kraken2 calls for only one sample in which EPI2ME detected both P. mirabilis and E. coli, whilst Kraken2 only identified P. mirabilis at an abundance greater than 1% of the reads. The culture-based diagnostic process also only identified P. mirabilis. Across all the clinical samples processed during the study, EPI2ME and Kraken2 agreed on the species present in all samples except this one, and one other (where EPI2ME detected Finegoldia magna in addition to E. coli and E. faecalis, but Kraken2 did not). Likewise, the culture-based identification for the samples were in almost 100% agreement with both EPI2ME and Kraken2 (Table 4 and Supplementary Table S8). For 2 samples, SkSw8A and SkSw10, our sequence-based protocol revealed additional species according to EPI2ME (Kraken2 only identified an additional species for SkSw10). Although these may be false-positives, it is also highly feasible that the sequence-based approach is be more sensitive than the current gold-standard culture-based techniques. It is worth noting that while culture-based species identification and ASTs are “gold standard” they are themselves not 100% accurate, with problems including small sample sizes compared to the infecting population, results skewed to easily cultured and fast growing organisms, and laboratory error (e.g. single doubling dilution steps between sensitive, intermediate and resistant). A 2007 evaluation estimated VITEK® 2 accuracy for species identification to be 98.3%, while AST accuracy was estimated to be 97.7% (48).
Supporting information
Acknowledgements
Many thanks to Jennifer Harris, Easter Bush Pathology, R(D)SVS, University of Edinburgh, for sharing clinical samples and the results thereof.
The authors are grateful for the fantastic bioinformatics resource, CLIMB-BIG-DATA (developed by the MRC, grant number MR/T030062/1), without which much of the data analysis undertaken here would not have been possible.
For the purpose of open access, the authors have applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising from this submission.
Concluding comments and future considerations
We have developed and validated a protocol for the rapid, culture-free, agnostic identification of pathogenic species from clinical canine samples, by cost-effective metagenomic sequencing. We have shown that this protocol is capable of detecting a wide array of species, representing over 90% of the urine and skin infections seen in the R(D)SVS HfSA. Although we did not test the remaining 10% of species due to their large number and relative rarity, our metagenomic extraction should effectively extract DNA from any species present in a sample, and we have shown up to 100% sensitivity and specificity in the identification of species from sequencing data alone. We intentionally developed a protocol that can also be adapted to a variety of other sample types, and the MagAttract protocol can be easily adapted for tissue, blood and other bodily fluids. In this way, one simple protocol can be deployed in a clinical setting to detect pathogens in a wide variety of infections, in different animals, in as little as 5 hours, compared to the 48 hours plus commonly seen in the current gold-standard diagnostics techniques.
Although the large amounts of host cell contamination we saw in some samples was problematic with regards to the prediction of AMR, a number of depletion techniques exist which could be incorporated into our protocol, at the cost of time and likely money, but with the benefit of allowing accurate AMR prediction from a greater range of samples. Outwith the samples with the highest levels of host DNA contamination, our ability to predict AMR was relatively accurate, approaching 95% in all but the most complicated co-infections. However, according to phenotypic AST a number of samples displayed intermediate levels of resistance to certain antimicrobials, mainly chloramphenicol, which were not predicted from the sequencing data. This suggests that as-yet unknown mechanisms may be responsible for intermediate resistance, or that certain mechanisms are currently missing from the AMR databases used (NCBI, ResFinder and CARD). Our current AMR prediction pipeline includes combining the results of three different tools (EPI2ME, Abricate and AMRFinderPlus), which function in different ways, in order to capture all potential information from our sequencing data. Future developments may include combining these tools into one easy-to-run workflow, as well as testing alternative tools and databases, such as the newly released, ISO-certified, abritAMR (55).
Lastly, we note that the flow cells (R9.4.1) and kits (SQK-RBK004) used here throughout development are now being replaced with R10.4.1 and SQK-RBK114. We expect these new flow cells and kits to incorporate seamlessly into our existing protocol and, indeed, they will likely improve the accuracy of the sequencing data produced, which could in turn improve the accuracy of our prediction of SNP-based AMR.