Nanowell-mediated two-dimensional liquid chromatography enables deep proteome profiling of <1000 mammalian cells

Abstract

Multidimensional peptide separations can greatly increase the depth of coverage in proteome profiling. However, a major challenge for multidimensional separations is the requirement of large biological samples, often containing milligram amounts of protein. We have developed nanowell-mediated two-dimensional (2D) reversed-phase nanoflow liquid chromatography (LC) separations for in-depth proteome profiling of low-nanogram samples. Peptides are first separated using high-pH LC and the effluent is concatenated into 4 or 12 nanowells. The contents of each nanowell are reconstituted in LC buffer and collected for subsequent separation and analysis by low-pH nanoLC-MS/MS. The nanowell platform minimizes peptide losses to surfaces in offline 2D LC fractionation, enabling >5800 proteins to be confidently identified from just 50 ng of HeLa digest. Furthermore, in combination with a recently developed nanowell-based sample preparation workflow, we demonstrated deep proteome profiling of >6000 protein groups from small populations of cells, including ∼650 HeLa cells and 10 single human pancreatic islet thin sections (∼1000 cells) from a pre-symptomatic type 1 diabetic donor.

Fig. 1. Workflow for nanowell-mediated 2D LC. (a) Nanogram biological samples or small populations of cells are prepared using nanoPOTS. (b) Samples are separated using high-pH nanoflow RPLC. The eluent from the high-pH RPLC separation is concatenated into nanowells and allowed to dry. (c) Samples in each nanowell are reconstituted and collected into capillaries for subsequent low-pH nanoLC-MS/MS analysis.

Fig. 2. Sample recovery from nanowells following reconstitution of 2 ng (a, b) and 20 ng (c, d) HeLa digest. The log₂ ratio of intensities between direct-loaded samples (control, n = 3) and reconstituted samples (reconstituted, n = 3) (a, c) and their histograms (b, d).

Fig. 3. Proteome coverage. Cumulative identified unique peptides (a) and proteins (b) for 4 and 12 fractions using 50 ng of HeLa digest.

Fig. 4. Scatter plot of the proteins quantified from 12 and 4 fractions by using 50 ng HeLa digest. The Pearson coefficient for the scatter plot was calculated to be 0.96, indicating that the proteins quantified from 12 and 4 fractions had a high correlation and were very similar at the quantitative level.

Fig. 5. In-depth proteome analysis of ∼650 HeLa cells following concatenation into 12 fractions. (a) HeLa cells in a nanowell for sample preparation and analysis, (b–d) proteins ranked according to their (b) abundance levels (log₁₀ iBAQ values), (c) molecular weight, and (d) GRAVY. (e) Gene Ontology annotations for Cellular Component of the identified proteins.