Comparing metagenomics and metatranscriptomics

To study the microbiome, metagenomics is frequently employed to characterize the microbes present within a specified community. Metagenomics relies on extracting and utilizing the genomic content of microbes to identify them. In addition, metagenomics can also be used to give an approximation of the abundance of each of the different microbes within a community.  

On the other hand, a more recent approach being used to characterize the microbiome is metatranscriptomics—a technique which instead assesses which genes are active within microbial communities. Like metagenomics, metatranscriptomics is also quantifiable, but may provide scientists with a better idea of the functional activity of the microbiome.  

The value of metatranscriptome analysis 

Metatranscriptomics represents a deeper layer of analysis, complementary to metagenomics. The value of metatranscriptomic profiling was discussed in a series of questions posed to Dr. Brice Le François, Research and Development Scientist – Microbiome at DNA Genotek. Dr. Le François provided answers in the context of a “Microbiome Connect Interview”, one of a series of microbiome-related interviews conducted by Kisaco Research. Specifically, he highlighted the value of metatranscriptomic data and its differences from metagenomics sequencing. 

1. What are the key insights that metatranscriptomic profiling provides over metagenomics profiling?

Over the past decade, a wealth of data has been generated on the human microbiome. This has mostly been achieved through metagenomic sequencing (16S and/or whole genome sequencing). Metagenomic sequencing can help determine the composition of the samples (i.e., which bacteria are present), as well as help predict the biological functions that are present in the sample, based on genomic DNA.

However, it doesn’t reflect bacterial activity (species or taxa that are metabolically active) or provide any information on the genes/pathways that are actively expressed in the sample. High relative abundance of a species, based on DNA profiles, does not necessarily suggest that this particular species is highly active and vice versa, a low relative abundance species may be highly active. In addition, given the difference in stability of DNA and RNA molecules within sample matrix, RNA signal is more likely to represent bacteria that are alive in the sample.  

[Metatranscriptomics] can help researchers determine which species are active at the time of sample collection, but more importantly also give us a sense of what these species are doing.

One of the main advantages of metatranscriptomic profiling is that it directly identifies bacterial activity in microbiome samples by analyzing the RNA these microorganisms are producing as part of their biological activities (independent of relative abundance of DNA). Metatranscriptomics opens up a new avenue for confidently targeting specific species, segmenting cohorts for more personalized treatments, or modulating the microbiome based on its metabolic activity.

2. How can metatranscriptomics complement metagenomics in understanding microbiome-related diseases?

That’s a great question. A lot of recent data has highlighted the critical role of the microbiome in health and disease. Most of the research has focused on differences in community structure through metagenomic sequencing, which only gives a snapshot of the species that are present.  

Interestingly enough, there are many examples of diseases where the microbial community composition vary very little (if at all) based on whole genome sequencing data. Acne is a great example. I think, it’s now widely accepted that the disease is driven by Cutibacterium acnes, a species that can promote inflammation of the skin. However, based on DNA profiles, the overall relative abundance of C. acnes is generally the same in individuals suffering from acne and healthy controls. Recent data suggest that specific strains of C. acnes species are driving the inflammation process as opposed to overall abundance of the species alone.  

I really believe that there is a lot of untapped data/insights lying in metatranscriptomic profiles, and I think generating these datasets will help us better understand the interplay between the microbiome and its host in health and disease. It will also help us gain actionable insights on how to better treat microbiome-associated diseases. The field is at a turning point and I’m excited to see what will come out of the data that will be generated in the next few years. 

3. Why do we see more researchers incorporating metatranscriptomics into microbiome studies?

I think it’s multi-factorial. The cost of sequencing has come down significantly in recent years and a number of recently developed tools/technologies have helped make metatranscriptomics of complex samples a reality. Overall, metatranscriptomics feels more accessible than ever to the average researcher.  

I also believe that people are starting to realize that metagenomic sequencing has limitations and cannot answer all the outstanding questions in the microbiome field. In recent years, there has been an increased number of studies looking at metabolomics which can be a useful readout of metabolic activity in microbiome samples. However, as with any other approach, metabolomics has its own limitations and is limited to the identification of metabolic by-products (host and microbial). As such, it cannot always identify the organism or the pathway that led to their production. The metabolomics databases, particularly for microbial derived metabolites are also still being improved, with researchers still focusing on specific metabolite assessment, such as short chain fatty acids or bile acids, rather than metabolome-wide untargeted analysis.  

Metatranscriptomics on the other hand has the advantage of looking at gene expression directly and by doing so it can help tie taxonomy with metabolic functions/pathways and to provide an extra layer than neither metagenomic sequencing nor metabolomics can provide. Many researchers have been (or are looking at biobanking samples for the future) and I expect that we will see more and more of them going back to these samples to generate additional insights with metatranscriptomics.  

4. What are the main challenges associated with metatranscriptomics characterizing the microbial function?

There are many. However, the great news is that a lot of issues that existed a few years ago have been solved by new technologies/products. Optimal sample collection and processing is a major consideration for anyone interested in doing metatranscriptomic profiling. RNA is much more unstable and prone to degradation than DNA. Coming up with a stabilization chemistry able to preserve stool RNA was a significant undertaking and took many, many iterations. 

 Microbiome samples also tend to contain high levels of host and microbial derived nucleases, especially RNAses, that can be pretty difficult to keep in check. This is something we’ve learned the hard way during the development of the OMNIgene™•GUT DNA and RNA (OMR-205) device.  

Additionally, bacteria are, by design, able to alter their gene expression profiles within seconds to respond to environmental cues or stressors. In order to generate profiles that reflect the true in vivo state of the samples, it’s important to ensure that the collection process has as little impact as possible on gene expression profiles and fully interacts with the entire sample rapidly. For example, impact of collection can be mitigated by the stabilization chemistry (lytic with faster diffusion rate) or by efficiently homogenizing the sample, a key feature of our OMNIgene•GUT devices. Choosing a high-performance extraction kit is also important in order to recover the highest quality and purity of RNA possible. Some commercially available extraction kits aren’t able to mitigate the impact of nucleases, which results in lower quality nucleic acids regardless of the quality of the input material.  

Another key consideration for metatranscriptomics is around sample library prep and analysis of the resulting profiles. A staggering high proportion (>98%) of the total RNA in a microbial samples consists of ribosomal RNA (rRNA) which provide very little information/insights into microbial activity. Additionally, unlike their mammalian counterparts, bacterial mRNAs are not polyadenylated and cannot be selectively recovered using oligo(dT)s. In metatranscriptomics workflows, bacterial rRNA has to be depleted during library preparation. rRNA depletion can be achieved using probes that remove rRNA by hybridization capture (i.e., biotin-streptavidin) or enzymatic digestion (RNAse H). Designing probes targeting 23S and 16S from specific bacterial species is fairly easy, however microbiome samples are complex, highly variable between donors, and can contain hundreds of bacterial species making probe design and efficient rRNA depletion a significant hurdle. Our early microbiome metatranscriptomics work has shown that a significant number of rRNA depletion kits perform well with low taxonomic complexity samples but fail with higher complexity samples, especially stool. Lastly, analysis of resulting metatranscriptomics profiles is also a significant challenge and require optimized pipelines with proper sequencing depth, filtering and databases for optimal outcomes.

5. What do you think is the next frontier in the advancement of microbiome analysis?

Metatranscriptomics is definitely an important step forward for microbiome research. It will enable the field to look at microbial activity and expressed functions and provide the microbiome field with novel insights on the interplay between the microbiome and its host as well as how it can lead to disease.  

Metagenomic sequencing has generated a wealth of data on the composition of the human microbiome across sites as well as compositional changes associated with certain disease states. However, to this day, we lack actionable targets to treat microbiome-linked diseases. In some cases, this may have been due to incorrect cohorts being selected for clinical testing; for example, wrongly assuming that patients with highly similar metagenomic sequencing profiles will have shared microbial functional and metabolic activities.  

Metatranscriptomics could help fill some of the remaining gaps in our understanding by identifying the metabolic pathways/specific genes that are driving the imbalance in the system and finding consistent drivers within a given cohort. Ultimately, the next frontier in microbiome research will be to look at the microbiome from multiple angles in multiomics-type of studies. Looking at several analytes at once (DNA, RNA, and metabolites for example), will help further our understanding of the super-organism humans and their microbiome constitute. Each analyte can provide a unique perspective on the microbiome and help unveil the intrinsic ties it shares with its host, from presence or relative abundance to metabolic activity and its by-products. We still have lots to learn. 

Incorporating microbial DNA and RNA analysis seamlessly into your workflow 

Imagine being able to collect a single sample from which you could analyze both microbial DNA and RNA in your research. On July 14, 2022, the OMNIgene•GUT DNA and RNA (OMR-205) device was made available to gut microbiome researchers to do exactly that—enable the study of both DNA and RNA.  

Developed by DNA Genotek, a subsidiary company of OraSure Technologies, Inc., the OMNIgene•GUT DNA and RNA device allows for the self-collection, stabilization, storage, and transportation of both microbial DNA and RNA at ambient temperature for gut microbiome profiling. According to Kathleen Weber, President of Molecular Solutions for OraSure Technologies, Inc.:  

“This product gives researchers in the pharmaceutical, biotechnical, academic, and consumer products fields a valuable tool to gain insight into the gene expression of gut microbes.”