Microfluidics-based electrospray ionization enhances the intrasource separation of lipid classes and extends identification of individual molecular species through multi-dimensional mass spectrometry: development of an automated high-throughput platform for shotgun lipidomics

Abstract

Herein, we exploit the use of microfluidics and optimized Taylor cones for improved intrasource separation/selective ionization of lipid classes during electrospray ionization. Increased differential ionization of multiple phospholipid classes was achieved through microfluidics with chip-based ionization resulting in substantial enhancement of intrasource separation/selective ionization of phospholipid classes in comparison to the conventional ion source. For example, using myocardial lipid extracts, 3-fold improvements in intrasource separation/selective ionization of myocardial phospholipid classes were routinely realized in the negative-ion mode in the absence of LiOH or other basic modifiers in the infused sample solutions. Importantly, the relative ratios of ions corresponding to individual molecular species in each lipid class to a selected internal standard from myocardial extracts were nearly identical between the chip-based interface and the syringe-pump-driven capillary interface. Therefore, quantitation of individual lipid molecular species directly from biological extracts through comparisons with internal standards in each lipid class was readily accomplished with an accuracy and dynamic range nearly identical to those documented using the well-established direct syringe-pump-driven capillary interface. Collectively, the use of microfluidics and robotic sample handling substantially enhances intrasource separation of lipids in comparison to routine capillary interfaces and greatly facilitates the use of multi-dimensional mass spectrometry using shotgun lipidomics, thereby providing an automated and high-throughput platform for global analyses of cellular lipidomes.

Figure 1

Comparisons of ESI mass spectra of mouse myocardial lipids acquired from an ion source coupled to a syringe-pump-driven interface to that obtained using a nanomate interface. Mass spectrum in (A) was acquired in the negative-ion mode directly from a diluted lipid extract (<50 pmol of total lipids/μL in 1:1 CHCl₃/MeOH) infused with a syringe-pump-driven interface. The mass spectrum in (B) was acquired in the negative-ion mode directly from a diluted mouse myocardial lipid extract (<50 pmol of total lipids/μL in chloroform/methanol/isopropanol (1:2:4, v/v/v)) infused with a nanomate interface. ‘IS’ denotes internal standard. CL, PtdCho, PtdEtn, PtdGro, PtdIns, and PtdSer stand for cardiolipin, choline glycerophospholipid, ethanolamine glycerophospholipid, phosphatidylglycerol, phosphatidylinositol, and phosphatidylserine, respectively.

Figure 2

Mass spectrometric analyses of mouse myocardial lipid extracts infused by either a syringe-pump-driven interface or a nanomate interface. The lipid extracts were appropriately diluted to less than 50 pmol of total lipids/μL with either 1:1 CHCl₃/MeOH or chloroform/methanol/isopropanol (1:2:4, v/v/v) in the presence of approximately 25 pmol of LiOH/μL prior to infusion with a syringe-pump-driven interface (A, C) or a nanomate interface (B, D), respectively, in the positive (A, B) or negative (C, D) mode. ‘IS’ denotes internal standard. CL, PtdCho, PtdEtn, and TAG stand for cardiolipin, choline glycerophospholipid, ethanolamine glycerophospholipid, and triacylglycerol, respectively.

Figure 3

Comparison of the relative contents of individual molecular species of choline and ethanolamine glycerophospholipids present in the lipid extracts of mouse myocardium using either a syringe-pump-driven interface or a nanomate device. The levels of individual PtdCho (A) and PtdEtn (B) molecular species in the lipid extracts of mouse myocardium were determined by using either a nanomate device (circles) or a syringe-pump-driven interface (squares). The levels of the relatively abundant molecular species of both lipid classes (solid symbols) were determined from the survey scans in the positive- and negative-ion modes, respectively, in the presence of a small amount of LiOH (i.e., the first step of quantitation) by comparison to their respective internal standards. The levels of the low-abundance molecular species of both lipid classes (open symbols) were determined using ratiometric comparisons in the second step by employing endogenous molecular species whose contents were determined in the first step as standards (solid symbols). Note that some error bars are within the symbols. *p <0.05 indicating the significant differences of the determined levels between the interfaces. The numbers on the x-axis correspond to those used in Tables 1 and 2, respectively.

Figure 4

ESI mass spectrometric total ion chromatograms of diluted extracts of mouse myocardial lipids in both positive-and negative-ion modes. Total ion counts of the ionized lipid solution in both positive (A) and negative (B) ion modes were acquired in the mass range of m/z 400–1000 at a scan rate of 1 s/scan for a total of 50 min.