Author: Alexandria Kilvington, Molecular Biology Technical Sales Specialist
Clinical diagnosis of microbial diseases has long been based on blood cultures and PCR, but can the sensitivity, time to result and the outcome be improved?
From uncovering more of the microbiome composition of multiple environments to clinical diagnostics, next-generation sequencing (NGS) has allowed further exploration of the microbial world. Shotgun metagenomic NGS (smNGS) is a high-throughput method allowing the simultaneous sequencing of multiple microorganisms within a single sample with high sensitivity. This enables the evaluation of bacterial diversity and abundance, which is usually difficult to analyse and increases the understanding of the potential links between complex microbial communities, with health and disease. This is compared to first-generation sequencing techniques such as Sanger sequencing, which are only capable of sequencing one sequence at a time, and other NGS technologies such as targeted metagenomic sequencing (tmNGS), which focus on specific target sequences. The use of mNGS has only been documented within the literature from the mid-2000s and so is relatively new compared to other more established methods. Although this method has revolutionised the field of microbial ecology (Taş et al., 2021), the implications of mNGS in a clinical setting remain controversial.
How has smNGS been used so far?
smNGS has been successful in identifying microbial presence in a range of sample types such as blood, cerebral spinal fluid, joint fluid, GI specimens, and urine as well as in upper and lower respiratory samples. The use of smNGS can identify pathogens that are not easily detected using the standard techniques of blood culture and broad-range polymerase chain reaction (PCR). It also allows an unbiased approach compared to targeted PCR, where particular microorganisms are chosen prior to analysis.
In comparison to culture, where a negative result may occur due to the presence of fastidious microorganisms or previous use of antibiotics, smNGS has the potential to identify the causative pathogen and so aid in the patient receiving the correct treatment. This was demonstrated in identifying the causative pathogen in a prosthetic joint infection where a novel pathogen was identified using smNGS (Thoendel et al., 2017) and in a study identifying and linking causative pathogens to cases of pneumonia (Bergin et al., 2022). This highlights the potential of smNGS in the clinical field, yet, the clinical utility of such NGS methods is poorly defined.
The UK currently use other NGS methods as a gold standard within the clinic with NGS panel testing being used for cancer diagnosis in combination with additional diagnostic tests. If NGS can be implemented in the oncogenic field as a gold standard, can this apply to other disciplines such as microbiology?
Can smNGS be improved by host cell depletion?
A problem faced using techniques such as smNGS is the fact that human DNA tends to be the dominant presence compared to microbial DNA in samples. This can make the detection of microbial DNA challenging by smNGS due to the additional sequencing of unwanted human DNA. This results in needing a higher read depth to detect the microbial DNA, equating to higher sequencing costs. The host DNA is then removed at the bioinformatics stage in analysis. One of VH Bio’s suppliers, Molzym, offers host-cell DNA depletion using their MolYsis technology which can combat this issue prior to downstream analysis.
Leo et al., (2017) used smNGS, to analyse the microbial versus human DNA reads in the contents of bronchoalveolar lavage (BAL) lower respiratory samples. Before host DNA depletion, Leo et al., identified only 0.05% prokaryote content with 0.22% of reads being unclassified, compared to 99.69% of reads being of human origin within 67 million reads. Another study using a different sample type; joint fluid, uncovered similar sequencing read proportions from raw samples (Thoendel et al., 2016). With the addition of a host depletion step, Leo et al. reduced the host DNA composition to 55.11% and increased the prokaryote reads from 0.05% to 23.55% within a lower read depth of 7 million reads (Leo et al., 2017). With the use of a host depletion step, Thoendel et al. also saw an increase in microbial content with a fold increase between 481 and 9580 compared to the control samples. In 2018, Thoendel et al., expanded on their study, producing the largest dataset of smNGS data on clinical samples at the time. They found that 43.9 % of culture-negative samples and 5.2% of culture-positive samples had new organisms detected by smNGS post-host depletion. In cases where prosthetic joints had failed for reasons other than infection, 3.6% of samples were shown to have new organisms detected by smNGS.
Host DNA depletion has been utilised in further patient samples. Gyarmati et al., (2016) depleted host DNA presence in blood samples from patients with acute leukaemia. This study discovered the presence of pathogens and associated antibiotic resistance genes in patients with a fever which correlated with low whole blood cell count. Gyarmati et al., (2016) concluded that smNGS can be implicated in personalised medicine for these patients. In another study, Büchler et al., (2022) were able to identify the causative microorganism in a patient resulting from an aortic valve replacement using smNGS which was missed by routine testing.
In summary, these studies identify that smNGS may have a part to play in the accurate diagnosis of infection in patients, even identifying microorganisms that are missed with standard practice (Thoendel et al., 2017; Büchler et al., 2022). Not only this, smNGS has the potential to identify antimicrobial resistance genes and virulence factors with high sensitivity which is not achieved through PCR methods such as 16S rDNA gene sequencing. But what challenges does this method have to face to reach the clinical setting?
Challenges that smNGS may face
With any method comes its challenges and smNGS comes with pitfalls and questions. These include:
- Are the microorganisms detected a result of sample contamination or are they true pathogens?
- Are these microorganisms implicated in the patient’s disease?
- How would the results differ if the patient was already receiving antibiotics prior to testing?
These questions highlight areas which need attention to be able to successfully identify the pathogenic cause of a disease in a patient. Many of these questions can be answered with the implementation of appropriate clinical protocols and controls. The use of clinical databases containing known pathogenic microorganisms may also aid in detecting the pathogenic microorganisms rather than microorganisms which may be either harmless or transient. The difficulty may arise in detecting whether a potential pathogenic microorganism is dead, ruling them out of being the pathogenic cause. These factors may highlight a need for cultures alongside the method.
Other aspects to address in the implication of smNGS in the clinical setting include that:
- Human DNA in samples tends to be the dominant presence.
- smNGS is complex and costly.
- The chosen bioinformatics pipeline and taxonomic databases used downstream are important.
The first point has been previously addressed above with help from the products from Molzym, the presence of human DNA can be reduced at the sample processing stage\ rather than at the bioinformatics stage, reducing sequencing costs. The cost of sequencing has reduced since its development and as the technology around sequencing becomes more advanced, costs may decrease further. The full workflow of smNGS can however be complex, requiring the need for qualified personnel; laboratory staff as well as bioinformaticians.
The final point may be harder to address. With an increase in data produced from methods such as smNGS, comes the necessity of well-developed and well-controlled downstream analysis methods to make sense of the large and complex datasets to produce a clinically relevant answer. As expected, the choice of bioinformatics used downstream has been shown to play a large role in the outcome of an smNGS study (Yap et al., 2021) and with any taxonomic database use, there is the possibility of introducing biases, raising the questions; how do you select the correct bioinformatic process and database? Are the taxonomic databases up to date enough for clinical application? Can database usage be standardised across laboratories?
Currently, there are many questions around the implication of NGS in a clinical setting, yet smNGS methods have the potential to add valuable insight and improved turnaround times to the microbiology field. Is it just a matter of time before it revolutionises the clinical diagnosis of microbial infection? Watch this space!
Get in touch to discuss your questions about smNGS or to find out how VH Bio can help with your lab’s requirements.
References:
Bergin, S., Chemaly, R., Dadwal, S., Hill, J., Joo Lee, Y., Haidar, G., Luk, A., Drelick, A., Chin-Hong, P., Benamu, E., Khawaja, F., Nanayakkara, D., Papanicolaou, G., Butkus Small, C., Fung, M., Barron, M., Davis, T., McClain, M., Maziarz, E., Madut, D., Bedoya, A., Gilstrap, D., Todd, J., Barkauskas, C., Bigelow, R., Leimberger, J., Tsalik, E., Wolf, O., Mughar, M., Hollemon, D., Duttagupta, R., Lupu, D., Bercovici, A., Perkins, B., T, Blauwkamp, T., Fowler, V., Holland, T. (2023) Plasma Microbial Cell-Free DNA Sequencing in Immunocompromised Patients with Pneumonia: A Prospective Observational Study. Clinical Infectious Diseases. DOI: 10.1093/cid/ciad599
Büchler, A., Lazarevic, V., Gaïa, N., Girard, M., Eckstein, F., Egil, A., Tschudin Sutter, S., Schrenzel, J. (2022) Mycobacterium chelonae Infection Identified by Metagenomic Next-Generation Sequencing as the Probable Cause of Acute Contained Rupture of a Biological Composite Graft – A Case Report. International Journal of Molecular Sciences. 23(1) 381
Gyarmati, P., Kjellander, C., Aust, C., Song, Y., Öhrmalm, L., Giske, C. (2016) Metagenomic analysis of bloodstream infections in patients with acute leukaemia and therapy-induced neutropenia. Scientific Reports. 6(23532)
Taş, N., de Jong, A., Li, Y., Trubl, G., Xue, Y., Dove, N. (2021) Metagenomic tools in microbial ecology research. Current Opinion in Biotechnology. 67 184-191
Thoendel, M., Jeraldo, P., Greenwood-Quaintance, K., Chia, N., Abdel, M., Steckelberg, J., Osmon, D., Patel, R. (2017) A Novel Prosthetic Joint Infection Pathogen, Mycoplasma salivarium, Identified by Metagenomic Shotgun Sequencing. Clinical Infectious Diseases. 65(2) 332-335.
Thoendel, M., Jeraldo, P., Greenwood-Quaintance, K., Yao, J., Chia, N., Hanssen, A., Abdel, M., Patel, R. (2016) Comparison of microbial DNA enrichment tools for metagenomic whole genome sequencing. Journal of Microbiological Methods. 127 141-145
Thoendel, M., Jeraldo, P., Greenwood-Quaintance, K., Yao, J., Chia, N., Hanssen, A., Abdel, M., Patel, R. (2018) Identification of prosthetic joint infection pathogens using a shotgun metagenomics approach. Clinical Infectious Diseases. 67(9) 1333-1338