FOOD-Lab Interview with Dr. Hans
Mol, Wageningen Food Safety Research,
part of Wageningen University
Detecting the unexpected
Targeted vs. Non-Targeted approach on pesticides
FOOD-Lab: Dr. Mol, what is the biggest
threat in the area of food safety?
Hans MoL: Pesticides are not the
biggest risk but the perception of
consumers is a different thing since
pesticides are highly ranked even if
EFSA or national governments state
that there is little risk actually.
FL: You have worked with several
methodologies to detect residues.
What is the actual status?
HM: The number of pesticides exceeds
1000. In the EC more than
450 are approved, globally it is much
more. A default maximum residue
limit of 0,01 mg/kg applies. Even if
not explicitly listed, all pesticides are
regulated. So basically, you have to
look for everything. From the historical
point of view methods were always targeted, i.e. the
classical methods based on gas chromatography (GC) with
element selective detectors and liquid chromatography (LC)
with UV or fluorescence detection started in the 70ies and
80ies. The selectivity of these instrumental measurements
was limited. This had to be compensated for by selective
extraction and cleanup procedures, dedicated to specific
groups of pesticides. However, a highly selective sample
preparation also means exclusion of other pesticides. In
time, the analysis scopes gradually increased, starting in the
late 80ies when mass spectrometry was coupled to GC. An
even bigger impact had the coupling of tandem MS (MS/
MS) to LC, from 2000s onwards. The very high selectivity
of chromatography combined with MS, and the continuous
increase in instrument sensitivity, eliminated the need for
extensive cleanup. Nowadays you can use generic sample
preparation techniques like the QuEChERS-method. That
changed the whole system because we could simultaneously
measure several hundreds of compounds. Currently,
LC-MS/MS is the dominant technique because it tends to
be more robust than GC. Although there is a large overlap,
many newer pesticides can only be analysed with LC-MS,
while there are others that can only be analysed by GC-MS/
MS. In the end, you still need both. Combined, you can
cover 100s of pesticides, but with these targeted methods,
you'll only find the ones you tell the instrument to look for.
FL: In a non-target-approach you
would not go for only known pesticides
but use the tool more for
screening purposes, wouldn’t you?
HM: Well, known pesticides as you
refer to here are the usual suspects,
the 200 or so that are most frequently
present in food. As I mentioned, there
a many more which are not often
found, but if they are there we would
like to know it. For this purpose we
are using non-target 'full scan high
resolution MS' based screening approaches.
With a full-scan method
you basically measure everything in a
sample. As long as a substance ends
up in the sample extract, comes out
of the chromatographic column, and
gets ionized, it can be detected. Data
handling after the measurement is
mostly done by extracting signals, diagnostic for pesticides,
out of the raw data. Depending on the instrument, signals
can be MS-spectra or exact masses of certain ions. So actually,
we do a non-target measurement but in terms of data
handling it is still targeted because detection is based on
pesticides included in a database or library. The difference
with target measurement is that the number of compounds
you can detect is in principle unlimited, and that you can
retrospectively search for additional pesticides or other contaminants
in old raw data.
FL: The confidence that is found may be only like 95%.
What does that mean then?
HM: Screening methods are relying on automated data
processing to find and report pesticides, and are never
perfect. What we are aiming at in screening is a confidence
of detection of 95%. The challenge here is that on
one hand you want to achieve the high detection confidence
rate, while at the same time you do not want
to get a lot of detects that upon further investigation
turn out to be false i.e. something else than a pesticide.
The result of a screening typically is that the pesticide
has been tentatively identified. This triggers a follow up
measurement by LC-MS/MS together with measurement
of the analytical reference standard to confirm the identity
and to quantify it.