Analysis of complex coffee bean contaminants with high-resolution mass spectrometry

[Susanne] My name is Susanne Ekroth and I have been working at the Swedish Food Agency since 1992, particularly focusing on pesticide analysis, food packaging materials, and hi-res technology.

I'd like to present Julio Cesar España, a chemist doing his Ph.D. focusing on analysis of residues in coffee at Chemistry Department at Swedish Food Agency.

[Julio] Currently, I am working with high-resolution mass spectrometry to address both field and processing coffee contaminants, particularly ochratoxin A and acrylamide.

[Susanne] The main goal of this project was to focus on which contaminants were relevant for this specific matrix. By doing so, we needed to merge a multi-residue method with a single-residue method, and to succeed with that, a great deal of method development was performed. The contamination of acrylamide or roasted coffee is a single-residue method (SRM), not only because of its particular structure but due to the highly complex nature of its challenging matrix.

The goal of the project was to combine relevant analytes in a simplified way to include this SRM into an ethyl acetate-based method to a multi-analyte workflow.

The possibility to develop a method that merges other different contaminants such as mycotoxins in a single approach to look for relevant assets is very appealing. The lab can take advantage of a screening method for monitoring purposes to comply regulations and to help improve consumer health.

[Julio] We have been relying on the Orbitrap technology, a platform that has been widely applied in several laboratories just like triple quads, but also for full scan, trigger tandem modes that leads to high-quality spectral information.

I am particularly focusing on the ochratoxin and acrylamide, but the plans are to expand the scope with all the relevant analytes as more mycotoxins, along with pesticides and contaminants of concern, that can be applied to other products using a simple and environmental-green alternative.

Full scan has become quite popular because it's simple. It has been used in untargeted scopes based on the exact mass of a molecule. High-res enables us to get an idea of the identity of the components of an extract with particular extraction methods.

The other aspect of our hybrid platform is that it not only enables to have a high resolution for isotope pattern matching, but also to obtain MS/MS spectrum which is used to fulfill the classification system to confidently assign identity.

Compound Discoverer, or CD, is a software that automates several processes such as alignment, merged features, and adaptation that help us to figure out identity. But what I find particularly important is to highlight the flexibility and simplify a way to set up experiments for small datasets or large batches.

The workload can be customized with different parameters which can be easily added or removed to search against local databases and huge online spectral libraries, such as mzCloud, to automate the process in the same layout. And very nicely, CD is ideal for tailoring the workflow for a particular comparison in a very straightforward way and visualizing statistical differences very quickly.

In practice, the compounds of interest can be differentiated from artifacts or contaminants derived from sample prep. It is possible to compare procedures, different samples such as the types of coffee samples, or other variations that helps to obtain a comprehensive picture beyond targeted approaches and without missing priceless information.

So, since I'm trying to find the effect of changes during method development, the good thing about CD layouts is that they are always directly linked to the raw data. The results can be further reduced by filtering sections on volcano plots, PCA, PLS, that are interactive and updated in real-time.

Having in mind that sometimes the bottleneck of nontargeted screening is the competent identification of structures of unknown features of no significance, in the latest version of CD, additional specificity can be achieved by isotope labeling, an orthogonal approach that can be used to assign identity to fewer picks meant as biologically relevant.

[Susanne] As in all, the main goal across all food safety application is to develop new methods to widen the current scope of the laboratories for relevant compounds and advance our understanding towards specific concerns on the food chain that may have an impact on the consumer health.

There are basically two main challenges that we can address now and extrapolate to further scenarios. The first one is that the base amount of data obtained from the different acquisition modes on the hi-res platform is now automatically processed.

And secondly, that is crucial to collect, organize, and compare data to make sense for the LC/MS spectral information that may help to reach conclusions for a particular contaminant.

Today, as we're finding untargeted approaches very challenging, the use of hi-res coupled to a comprehensive automatic software, such as Compound Discoverer, researchers have the potential to improve their identification workflow to achieve reliable solutions.