Clara Esteve1, Alfonsina D’Amato, Pier Giorgio Righetti
Department of Chemistry, Materials and Chemical Engineering “Giulio Natta”, Proteomics Laboratory, Politecnico di Milano, 20131 Milan, Italy
1Department of Analytical Chemistry, Faculty of Chemistry, University of Alcalá, Spain
It all started, perhaps, in the year 2001 when Hidalgo et al. [1] claimed finding proteins in olive oils, ranging in concentration from 10 to 50 micro-grams per 100 grams of oil. Curiously, though, there was no identification, via the classical tool of mass spectrometry (MS) of any of these presumptive proteins. In addition, all the proteins they claimed having seen amounted to just a single polypeptide chain, of apparent Mr of 4600 Da (but of undisclosed nature). The fact that such an important finding, surely of great interest in the field of olive oil industry, had no follow up in the scientific literature for more that 11 years should have rung a bell.
Notwithstanding that, considering that at least since two years our lab has started an extensive investigation on the proteome content of alcoholic and non-alcoholic beverages, in order to find if they could be classified according to the set of proteins they contained, and if any adulteration could be detected [2-6], we recently decided to visit again this obscure aspect of olive oils, with the following aims:
- in order to see if enough proteins could be detected so as to classify the various oils according to different cultivars;
- in order to see if high-quality oils could be distinguished form industrial oils available in supermarket.
As
a pre-requisite to that, we first started investigating the proteome
content of the olive fruit, on which not much was known up to the
present. To that aim, both the proteinaceous content of the seed as
well as of the pulp were analyzed, since both compartments could
contribute to the presence, if any, of trace proteins in the final
olive oil processed. Indeed our work resulted in a very extensive
exploration of both proteomes, amounting to the identification of a
grand total of 61 proteins in the seed and as many as 231 species in
the pulp [7].
Having established these two proteomes, we started
investigating the trace proteins in commercial olive oils, as found
on the shelves of supermarkets. On the few major olive oils available
(Monini, Dante, Carapelli) we could not detect any
proteinaceous material. We thus resorted to the analysis of
high-quality oils, that had not undergone any industrial refinement
process, to ascertain the possible presence of proteins in traces.
For those analyses, we were kindly provided with three types of
high-quality oils by the Tenuta Zimarino Masseria Don Vincenzo
(Vasto, Chieti, courtesy of Mr. Tieri) namely:
Per Liliana “Ascolana Tenera” (monocultivar);
Costa dei Trabocchi “Gentile di Chieti” (monocultivar);
Don Vincenzo “Colline Teatine”, Vastese DOP.
A
representative label of these three oils is given in Fig. 1
We
had to test different extraction protocols in order to obtain
reasonable results, but finally we succeed in seeing bands of protein
zones in a sodium dodecyl sulphate polyacrylamide gel slab
(SDS-PAGE), see Fig. 2. This gel is interesting, since it shows a
spectrum of fine bands spanning the Mr region from ca. 12 to 40 kDa,
suggesting indeed the presence of quite a few proteins extracted from
these types of oils (this gel is representative of all extractions
performed in all oils from "Masseria don Vincenzo").
It also gives us
an important clue: considering that the proteins were extracted from
400 grams of oils, and that detection has been performed with silver
staining (the most sensitive stain available), likely the trace
proteins present should be of the order of tens of micro-grams per
litre, i.e. two orders of magnitude lower than that reported in [1].
For
the analyses of these presumptive proteins, the gel was segmented
into 21 zones (marked by numbers and brackets in Fig. 2), the
proteins digested with trypsin, the resulting peptides captured,
purified and injected in two mass spectrometer: an Orbitrap (at the
Faculty of Pharmacy, Prof. Aldini) and a Velos (at the Fondazione
Filarete).
Identifications were not easy at all, since the signals
was very feeble and there were plenty of traces of any possible
contaminant. Finally, after rejecting all spurious spectra and
submitting the few surviving ones to stringent statistical tests, we
enucleated just three proteins that had survived the threshold test.
It should be noted that one of them, histone H4, is highly
homologous to a histone that we found in the seed of the olive fruit,
which suggest that this protein was transferred to the oil form the
proteome of the seed. The other two proteins are homologous to
components of the pulp.
Finally,
Fig. 3 gives the fragmentation spectrum of one peptide of the
histone, giving the determination of the correct sequence of this
peptide.
Conclusions
It
would appear that only traces of proteins can be found in olive oils,
of which at the moment only three have been correctly identified.
Surely quite a few more should be present, but ascertaining their IDs
might be a difficult task, since such proteins could be very
hydrophobic and contain too few Lys and Arg residues, where the
trypsin attack occurs. We will thus have to test different enzymes
for protein digestion, in the hope of obtaining the correct cuts. The
second important information is that it appears that only high
quality oils, which have not been subjected to any industrial
treatment, contain traces of proteins. Commercial oils commonly found
in the supermarkets did not appear to contain any trace proteins
(which presumably are removed in the industrial fining process). Thus
the present data seem to suggest that the presence of trace proteins
might enable consumers and control agencies to distinguish
high-quality oils (those that have been prepared by cold pressing in
the absence of any subsequent industrial treatment) vs. the
industrial ones that have undergone refining processes.
References
[1] Hidalgo FJ, Alaiz M,
Zamora R, Determination of peptides and proteins in fats and oils.
Anal Chem. (2001) 73, 698-702.
[2] D'Amato
A, Fasoli E, Righetti PG. Harry
Belafonte and the secret proteome of coconut milk.
J Proteomics (2012) 75, 914-920.
[3] Fasoli E, D'Amato A,
Citterio A, Righetti PG. Ginger
Rogers? No, Ginger Ale and its invisible proteome.
J
Proteomics (2012)
75, 1960-1965.
[4] Di
Girolamo F, D'Amato A,
Righetti
PG.
Horam
nonam exclamavit: sitio.
The trace proteome of your daily vinegar.
J Proteomics
(2011) 75, 718-24.
[5] Fasoli E, D'Amato A,
Kravchuk AV, Citterio A, Righetti PG. In-depth
proteomic analysis of non-alcoholic beverages with peptide ligand
libraries. I: Almond milk and orgeat syrup. J
Proteomics (2011) 74, 1080-1090.
[6] D'Amato
A, Fasoli E, Kravchuk AV, Righetti PG. Going
Nuts for Nuts? The Trace Proteome of a Cola Drink, as Detected via
Combinatorial Peptide Ligand Libraries.
J Proteome Res. (2011) 10, 2684-2686.
[7] Clara Esteve C,
D’Amato A, Marina ML, García MC, Citterio A, Righetti
PG. Identification of olive (Olea
europaea) seed and pulp proteins by
nLC-MS/MS via combinatorial peptide ligand libraries. J Proteomics
(2012) 75, 2396-403.
Figure 1 – Label of the
monocultivar Ascolana Tenera Oil
Figure 2 – SDS-PAGE analysis of
proteins extracted from olive oil “Masseria Don Vincenzo” from
Tenuta Zimarino. Mr: molecular mass ladder. The track “Olive oil”
has been segmented into 21 zones, whose content has been digested
with trypsin and subjected to MS analysis on a Orbitrap.
Figure 3 - MS spectrum of the
second peptide of histone R4 (ISGLIYEETR).