purification in a preparative scale and cannot be applied in parallel in
a single biological sample.
In contrast, the separation of native protein complexes according to their inherent biochemical properties, exemplified by size
separation via size-exclusion chromatography (SEC) or charge separation via ion exchange chromatography (IEXC), offers the possibility to isolate and characterize a large number of complexes in
parallel from the same sample, thereby opening the door for
system-wide studies of protein complexes. SEC–MS has been successfully applied for several years by now, for example, to analyze
proteome remodeling upon apoptosis [18], and biochemical fractionation established the first large-scale repository of human protein complexes [19].
Recently, we introduced the concept of complex-centric analysis for co-fractionation data [20]. In this method, protein complexes are statistically scored on the basis of protein co-elution
profiles using a target and decoy approach encrypted in the R
package CCprofiler (https://github.com/CCprofiler/CCprofiler).
Complex-centric analysis treats the parallel analysis of complexes as
a multiple hypothesis testing problem and thus greatly reduces the
main inherent problem of the approach, the fact that the number of
detected protein elution profiles is very large (e.g., 5000) in comparison to the number of SEC fractions analyzed per sample (e.g.,
70). Applying CCprofiler to HeLa cells in different mitotic states
[21] provided new insights into the mechanism of nuclear pore
complex assembly during cell division and identified numerous
additional differences in proteome organization between the states
tested.
CCprofiler together with novel LC solutions [22] for higher
throughput LC–MS/MS measurements allowed us to analyze the
protein contents of the consecutive fraction of an SEC separation in
approximately 1 day. This was achieved by using a 21-min injectionto-injection chromatographic method. The resulting gradient
compression is detrimental for data-dependent acquisition scheme,
however, the use of SWATH/DIA acquisition [23] largely mitigates this issue because multiple precursor ions are concurrently
fragmented. The resulting data set of convoluted fragment ion
spectra can be then queried using a peptide-centric search strategy
that depends on a spectral library as prior information. SWATH/
DIA acquisition has the proven capability of yielding high-quality
and reproducible quantification and identification of large number
of peptides and proteins across hundreds of samples or even laboratories [24] and is therefore ideally suited for the comparative analysis of the proteins in consecutive SEC fractions.
The combination of optimized sample preparation, fast LC–
MS acquisition, and a suite of software tools results in enhanced
throughput and robustness of the entire pipeline for the parallel
analysis of protein complexes from a broad range of biological
samples.
System-Wide Profiling of Protein Complexes Via Size Exclusion. . .
271
a single biological sample.
In contrast, the separation of native protein complexes according to their inherent biochemical properties, exemplified by size
separation via size-exclusion chromatography (SEC) or charge separation via ion exchange chromatography (IEXC), offers the possibility to isolate and characterize a large number of complexes in
parallel from the same sample, thereby opening the door for
system-wide studies of protein complexes. SEC–MS has been successfully applied for several years by now, for example, to analyze
proteome remodeling upon apoptosis [18], and biochemical fractionation established the first large-scale repository of human protein complexes [19].
Recently, we introduced the concept of complex-centric analysis for co-fractionation data [20]. In this method, protein complexes are statistically scored on the basis of protein co-elution
profiles using a target and decoy approach encrypted in the R
package CCprofiler (https://github.com/CCprofiler/CCprofiler).
Complex-centric analysis treats the parallel analysis of complexes as
a multiple hypothesis testing problem and thus greatly reduces the
main inherent problem of the approach, the fact that the number of
detected protein elution profiles is very large (e.g., 5000) in comparison to the number of SEC fractions analyzed per sample (e.g.,
70). Applying CCprofiler to HeLa cells in different mitotic states
[21] provided new insights into the mechanism of nuclear pore
complex assembly during cell division and identified numerous
additional differences in proteome organization between the states
tested.
CCprofiler together with novel LC solutions [22] for higher
throughput LC–MS/MS measurements allowed us to analyze the
protein contents of the consecutive fraction of an SEC separation in
approximately 1 day. This was achieved by using a 21-min injectionto-injection chromatographic method. The resulting gradient
compression is detrimental for data-dependent acquisition scheme,
however, the use of SWATH/DIA acquisition [23] largely mitigates this issue because multiple precursor ions are concurrently
fragmented. The resulting data set of convoluted fragment ion
spectra can be then queried using a peptide-centric search strategy
that depends on a spectral library as prior information. SWATH/
DIA acquisition has the proven capability of yielding high-quality
and reproducible quantification and identification of large number
of peptides and proteins across hundreds of samples or even laboratories [24] and is therefore ideally suited for the comparative analysis of the proteins in consecutive SEC fractions.
The combination of optimized sample preparation, fast LC–
MS acquisition, and a suite of software tools results in enhanced
throughput and robustness of the entire pipeline for the parallel
analysis of protein complexes from a broad range of biological
samples.
System-Wide Profiling of Protein Complexes Via Size Exclusion. . .
271