De novo Sequencing of Proteins With Mass Spectrometry
69
with a minimum width requirement for peak consideration of 0.05 Da in the case
of quadrupole time of flight data and 0.5 Da for triple quadrupole data.
4
Results
4.1
Differential Scanning Based de Novo Sequencing of Proteins
with a Triple Quadrupole Mass Spectrometer
The advantage of 18 0 based de novo sequencing in comparison with the methylation method is that only one analytical run needs to be performed instead of two.
However, using a triple quadrupole mass spectrometer the fragments are generally not isotopically resolved to preserve the required sensitivity. This can lead to
misassignments of masses when 50 % 18 0 labelled peptides are investigated. The
fragment ion mass assignment can be off by up to 2 Da from its real mass when
its weight is over 1000 Da and its peak is of low abundance. Fragments with
masses over 1000 Da have a remarkable isotopic width due to the natural 13C and
15N contribution. If the peak contains about 20 ions in total, the quantity of the
different isotopes may not be correctly represented. In this situation 50 % 18 0
labeling makes a correct mass assignment to the first isotope considerably more
difficult than for unlabelled peptides. To circumvent this problem we label peptides only to 33 % when they are investigated on a triple quadrupole machine.
With 33 % labeling, the 18 0 isotope contributes to the fragment peak only when
sufficient ions can be accumulated and for fragment ions up to 1300 Da the first
isotope is still the most abundant one. Fig. 5.1 demonstrates that 33 % labeling
allows to assign the correct first isotopic mass to an abundant and a less abundant ion in the spectrum in comparison to a 50 % labelled peptide.
The purpose of isotopic labeling was to identify y ions throughout the spectrum. With the reduced labeling it is correspondingly more difficult to recognize
the y ions in the lower part of the spectrum. To facilitate the y ion recognition the
differential scanning technique is applied. Two tandem MS spectra are acquired,
one by transmitting the complete 16 0/ 18 0 isotopic distribution into the collision
zone and the second by selecting exclusively the 18 0 containing ions. This can be
done without compromising the sensitivity. The resolution of the precursor
selecting quadrupole does not need to be increased to transmit only a single isotope. When the first quadrupole is adjusted to high transmission, all masses
within a window of about 3 Da are transmitted into the collision zone. But, the
transmission curve has a sharply rising flank starting the transmission at the preset m/z value. Setting this value to a sufficiently higher m/z value it is possible to
exclude the smaller 16 0 containing pep tides but transmitting the 18 0 labelled.
This is done without changing other parameters of the quadrupole which could
affect the overall ion transmission. Therefore, it is possible to generate a spectrum from the complete 16 0/ 18 0 isotopic set of peptides and a second spectrum
exclusively from the 18 0 isotopes. When the two spectra are acquired they are
overlaid to visualize differences between them. Fragments which do not contain
the 18 0 isotope will be at the same place in both spectra, whereas C-terminal y
69
with a minimum width requirement for peak consideration of 0.05 Da in the case
of quadrupole time of flight data and 0.5 Da for triple quadrupole data.
4
Results
4.1
Differential Scanning Based de Novo Sequencing of Proteins
with a Triple Quadrupole Mass Spectrometer
The advantage of 18 0 based de novo sequencing in comparison with the methylation method is that only one analytical run needs to be performed instead of two.
However, using a triple quadrupole mass spectrometer the fragments are generally not isotopically resolved to preserve the required sensitivity. This can lead to
misassignments of masses when 50 % 18 0 labelled peptides are investigated. The
fragment ion mass assignment can be off by up to 2 Da from its real mass when
its weight is over 1000 Da and its peak is of low abundance. Fragments with
masses over 1000 Da have a remarkable isotopic width due to the natural 13C and
15N contribution. If the peak contains about 20 ions in total, the quantity of the
different isotopes may not be correctly represented. In this situation 50 % 18 0
labeling makes a correct mass assignment to the first isotope considerably more
difficult than for unlabelled peptides. To circumvent this problem we label peptides only to 33 % when they are investigated on a triple quadrupole machine.
With 33 % labeling, the 18 0 isotope contributes to the fragment peak only when
sufficient ions can be accumulated and for fragment ions up to 1300 Da the first
isotope is still the most abundant one. Fig. 5.1 demonstrates that 33 % labeling
allows to assign the correct first isotopic mass to an abundant and a less abundant ion in the spectrum in comparison to a 50 % labelled peptide.
The purpose of isotopic labeling was to identify y ions throughout the spectrum. With the reduced labeling it is correspondingly more difficult to recognize
the y ions in the lower part of the spectrum. To facilitate the y ion recognition the
differential scanning technique is applied. Two tandem MS spectra are acquired,
one by transmitting the complete 16 0/ 18 0 isotopic distribution into the collision
zone and the second by selecting exclusively the 18 0 containing ions. This can be
done without compromising the sensitivity. The resolution of the precursor
selecting quadrupole does not need to be increased to transmit only a single isotope. When the first quadrupole is adjusted to high transmission, all masses
within a window of about 3 Da are transmitted into the collision zone. But, the
transmission curve has a sharply rising flank starting the transmission at the preset m/z value. Setting this value to a sufficiently higher m/z value it is possible to
exclude the smaller 16 0 containing pep tides but transmitting the 18 0 labelled.
This is done without changing other parameters of the quadrupole which could
affect the overall ion transmission. Therefore, it is possible to generate a spectrum from the complete 16 0/ 18 0 isotopic set of peptides and a second spectrum
exclusively from the 18 0 isotopes. When the two spectra are acquired they are
overlaid to visualize differences between them. Fragments which do not contain
the 18 0 isotope will be at the same place in both spectra, whereas C-terminal y