Proteomic analysis of Chinese hamster ovary cells

Abstract

To complement the recent genomic sequencing of Chinese hamster ovary (CHO) cells, proteomic analysis was performed on CHO cells including the cellular proteome, secretome, and glycoproteome using tandem mass spectrometry (MS/MS) of multiple fractions obtained from gel electrophoresis, multidimensional liquid chromatography, and solid phase extraction of glycopeptides (SPEG). From the 120 different mass spectrometry analyses generating 682,097 MS/MS spectra, 93,548 unique peptide sequences were identified with at most 0.02 false discovery rate (FDR). A total of 6164 grouped proteins were identified from both glycoproteome and proteome analysis, representing an 8-fold increase in the number of proteins currently identified in the CHO proteome. Furthermore, this is the first proteomic study done using the CHO genome exclusively, which provides for more accurate identification of proteins. From this analysis, the CHO codon frequency was determined and found to be distinct from humans, which will facilitate expression of human proteins in CHO cells. Analysis of the combined proteomic and mRNA data sets indicated the enrichment of a number of pathways including protein processing and apoptosis but depletion of proteins involved in steroid hormone and glycosphingolipid metabolism. Five-hundred four of the detected proteins included N-acetylation modifications, and 1292 different proteins were observed to be N-glycosylated. This first large-scale proteomic analysis will enhance the knowledge base about CHO capabilities for recombinant expression and provide information useful in cell engineering efforts aimed at modifying CHO cellular functions.

Figure 1

Flow diagram of the CHO proteome sample preparation and analysis. Spent media and cells were separated followed by cell lysis. Proteins were extracted from both sources and subjected to gel separation and in gel proteolysis (lysates) or in-solution digestion and bRPLC separation (lysates and secretome). These fractions were then subjected to LC/MS/MS and bioinformatics analysis. The glycosylated peptides were also analyzed by using SPEG coupled LC/MS/MS analysis.

Figure 2

Number of peptides identified for each protein.

Figure 3

Percentage of identified CHO protein with greatest genetic sequence homology to Mouse, Rat, Human, None (do not have any homology) and other Mammalian such as Sheep, Orangutan or Bovine.

Figure 4

Codon frequency for CHO cells. The ratio of codon frequency in CHO cells to humans is indicated by the color of the codon boxes using the heat map scale.

Figure 5

Significantly enriched functional groups of a) proteome, and b) glycoproteome. Identified proteins both in proteome and glycoproteome were functionally annotated with GO-slim terms and statistically enriched GO functions were categorized into 17 main groups.

Figure 6

A) mRNA and protein intensities for each identified protein. The normalized mRNA values, FPKM values versus normalized protein values NSAF values are plotted to figure out the significantly unstable and stable proteins.

Figure 7

Detailed mapping of A) Protein processing in endoplasmic reticulum B) Apoptosis pathways (Yellow: Genome, Red: mRNA and Protein, Orange : mRNA, Green: Protein).