Just because it’s natural doesn’t mean it’s safe

A major analytical challenge in natural products is the complexity of the samples.  Why does this matter?

Waters recently hosted a webinar, available on-demand, entitled “Authenticate Herbal Supplements with a Metabolomics Approach – Reporting and Analytics” featuring our own Dr Giorgis Isaac, whose current research is focused on novel analytical and informatics method development to solve the analytical challenges in natural product analyses. The webinar was based upon some collaborative work between Waters and the University of Mississippi, you can read the full paper here.

During the webinar Giorgis talked about the market in Natural Products, currently standing at $61.84 billion. There are many examples of reported adulteration. The take-home message was “Just because something is natural doesn’t mean that it is safe”. Herbal supplements are complex and so is their analysis. Giorgis talked about the problems of using a targeted approach. What compounds do you identify for your analysis? People are managing to meet required thresholds of active ingredients but what else is there? How can you be sure of the purity of your specimen?

A holistic analysis

Giorgis and his collaborators therefore wanted to explore an untargeted approach. Untargeted analysis is holistic so you can analyse ALL of the data in ALL of your samples and compare them, finding the differences without any prior knowledge. In short, you don’t need to know what you are looking for. Using Progenesis QI, you can produce a ‘metabolomic fingerprint’ for your samples which can then be compared against similar species which may be being used for adulteration. These metabolomics fingerprints can also be compared against a QC sample. Giorgis emphasized the importance of good experimental design in untargeted analysis. He emphasized the usefulness of using QC samples and how essential a standard workflow is. This is so you can minimise and measure technical variance in your analysis and focus on your biological variation.

QI: Quantify then Identify

Untargeted analysis has been made possible by the use of software that allows unbiased peak picking and retention time alignment of signal. Progenesis QI is renowned for its unique approach to alignment, co-detection and peak picking, best explained in this diagram:

Giorgis walked us through how Progenesis QI was used in the exploration of three botanicals: Hoodia, Terminalia, and chamomile. He pointed out the flexibility of Progenesis QI, being multi-vendor, able to handle profile and centroided data formats, as well as high and low energy formats, including MSE.

Useful statistics

He explained how he uses the PCA analysis to check the quality of the data and to spot any suspicious finding quickly. We have seen this done successfully time and time again, for example, if samples have got mixed up or there is an anomaly in the samples. Using the dendogram, you can also look at groups of compounds behaving in a similar way across the experiment. It is possible to tag such features of interest and drill further down into the data. Giorgis also described how Progenesis QI can export into EZinfo so you can run further multivariate statistical tests. You can then import the results back into Progenesis QI for further analysis.


So far so good: we have minimised our technical variation, we know what is changing significantly but… how do we identify these compounds of interest? One of the advantages of using Progenesis QI is its ability to combine results from multiple search methods and databases. Progenesis QI is able to score results from all the databases and search methods it supports, so you can compare search results obtained from different search methods. Progenesis QI currently supports these databases:

I don’t want to give the game away here, as the webinar and the paper are both worth viewing, but the results from the analyses of Hoodia, Terminalia, and chamomile were very clean and very encouraging, showing what can be done to identify marker compounds for each species in order to detect any adulteration. Progenesis QI had a large role to play in this neat piece of detective work. During the questions at the end of the webinar, Giorgis discussed how this untargeted approach compares to DNA methods and U-V NMR IR methods. He emphasised that LC-MS is more stable and easy-to-use these days, is very sensitive and, alongside Progenesis QI, can handle the analysis of complex mixtures. Once again questions led to the standardised workflow as it is so important to have maximum control of the variables. Why a QC group? Is that really necessary? Giorgis emphasized that a QC group gives you confidence: if your QC group lies in the centre of your PCA analysis, your variation is in the sample, not the methodology.

I came away from this webinar excited at the chance for Progenesis QI to play a pivotal role in the natural products market. As this market grows and the temptation to adulterate grows with it, it is reassuring to know that untargeted LC-MS analysis makes it virtually impossible to cheat the system.