The same instrument shows a limit of detection of 10
À12 to
10
À9
mole injected when analyzing purified lipid
standards [12].
14. Each M. tuberculosis lipid family is composed of not one, but
more often of several dozen acyl-forms that share a common
polar head. Hence, the signal corresponding to a given lipid
family is characterized by a series of molecular ions detected
with m/z values that differ by 14.0156 a.m.u (CH 2 group),
which elute within a narrow retention time window on diol
chromatography. When performing the semi-quantitative
comparison of the lipid profile of different samples, it is possible to perform the comparison using one representative acylform by lipid family, if the acyl-form profile of the lipid families
is conserved between samples.
15. While MS/MS databases are not available for M. tuberculosis
lipids, fragmentation profiles of several M. tuberculosis lipids
families have been previously published [12].
16. When the goal is the semi-quantitative comparison of the lipid
profiles of different strains (for example, WT, mutant, and
complemented strains) or of the same strain grown in different
conditions, it is important to ensure that the injected quantity
of lipid extract is within the linear response range of the lipids
of interest. Indeed, the area under the chromatogram of a
given molecular lipid species increases proportionally with its
abundance until the signal saturates or even gradually extinguishes. Because each lipid family has its own response factor
under given mass spectrometry settings, it is important that the
concentration of the studied lipid extract (i.e., the injected
quantity) is suitable to assess changes in abundance for as
many, if not all, lipid families as possible.
17. These databases represent a unique resource of information,
but it is important to keep in mind that they consist of lists of
theoretical calculated m/z for the acyl-forms of the different
mycobacterial lipid families described in the literature. The
MycoMass was built to provide a tool to assist MS signal
assignments or at least to propose possible formulas. As culture/infection conditions modulate the acyl-form profile of
lipids, the listed acyl-forms are not limited to the published
data but are sometimes extended in terms of acyl chain length
and combination of fatty acids (Table 2). In addition, these
databases do not contain retention time criteria, which are
method- and therefore laboratory-dependent. Hence, a
match between a detected m/z and an entry in these databases
does not constitute, by itself, a definitive assignment. The use
of these databases for the annotation of large HPLC-MS datasets currently requires an extensive manual curating to remove
annotations that are incoherent with observed RTs. When
574
Emilie Layre
À12 to
10
À9
mole injected when analyzing purified lipid
standards [12].
14. Each M. tuberculosis lipid family is composed of not one, but
more often of several dozen acyl-forms that share a common
polar head. Hence, the signal corresponding to a given lipid
family is characterized by a series of molecular ions detected
with m/z values that differ by 14.0156 a.m.u (CH 2 group),
which elute within a narrow retention time window on diol
chromatography. When performing the semi-quantitative
comparison of the lipid profile of different samples, it is possible to perform the comparison using one representative acylform by lipid family, if the acyl-form profile of the lipid families
is conserved between samples.
15. While MS/MS databases are not available for M. tuberculosis
lipids, fragmentation profiles of several M. tuberculosis lipids
families have been previously published [12].
16. When the goal is the semi-quantitative comparison of the lipid
profiles of different strains (for example, WT, mutant, and
complemented strains) or of the same strain grown in different
conditions, it is important to ensure that the injected quantity
of lipid extract is within the linear response range of the lipids
of interest. Indeed, the area under the chromatogram of a
given molecular lipid species increases proportionally with its
abundance until the signal saturates or even gradually extinguishes. Because each lipid family has its own response factor
under given mass spectrometry settings, it is important that the
concentration of the studied lipid extract (i.e., the injected
quantity) is suitable to assess changes in abundance for as
many, if not all, lipid families as possible.
17. These databases represent a unique resource of information,
but it is important to keep in mind that they consist of lists of
theoretical calculated m/z for the acyl-forms of the different
mycobacterial lipid families described in the literature. The
MycoMass was built to provide a tool to assist MS signal
assignments or at least to propose possible formulas. As culture/infection conditions modulate the acyl-form profile of
lipids, the listed acyl-forms are not limited to the published
data but are sometimes extended in terms of acyl chain length
and combination of fatty acids (Table 2). In addition, these
databases do not contain retention time criteria, which are
method- and therefore laboratory-dependent. Hence, a
match between a detected m/z and an entry in these databases
does not constitute, by itself, a definitive assignment. The use
of these databases for the annotation of large HPLC-MS datasets currently requires an extensive manual curating to remove
annotations that are incoherent with observed RTs. When
574
Emilie Layre