Hyperosmolality in CHO cell culture: effects on the proteome

Abstract

Chinese hamster ovary (CHO) cells are the most commonly used host cell lines for therapeutic protein production. Exposure of these cells to highly concentrated feed solution during fed-batch cultivation can lead to a non-physiological increase in osmolality (> 300 mOsm/kg) that affects cell physiology, morphology, and proteome. As addressed in previous studies (and indeed, as recently addressed in our research), hyperosmolalities of up to 545 mOsm/kg force cells to abort proliferation and gradually increase their volume-almost tripling it. At the same time, CHO cells also show a significant hyperosmolality-dependent increase in mitochondrial activity. To gain deeper insight into the molecular mechanisms that are involved in these processes, as detailed in this paper, we performed a comparative quantitative label-free proteome study of hyperosmolality-exposed CHO cells compared with control cells. Our analysis revealed differentially expressed key proteins that mediate mitochondrial activation, oxidative stress amelioration, and cell cycle progression. Our studies also demonstrate a previously unknown effect: the strong regulation of proteins can alter both cell membrane stiffness and permeability. For example, we observed that three types of septins (filamentous proteins that form diffusion barriers in the cell) became strongly up-regulated in response to hyperosmolality in the experimental setup. Overall, these new observations correlate well with recent CHO-based fluxome and transcriptome studies, and reveal additional unknown proteins involved in the response to hyperosmotic pressure by over-concentrated feed in mammalian cells.Key points• First-time comparative proteome analysis of CHO cells exposed to over-concentrated feed.• Discovery of membrane barrier-forming proteins up-regulation under hyperosmolality.• Description of mitochondrial and protein chaperones activation in treated cells.

Keywords: CHO; Cell size; Fed-batch; Hyperosmolality; LFQ proteomics.

Fig. 1

Workflow of LFQ proteomics used in this work. Created with BioRender.com (https://biorender.com/)

Fig. 2

Western blot analysis of oversupplemented feed-exposed (F) and control (K) whole protein lysates of the fed-batch cultivation of CHO-DP12 sampled on days 2, 6, and 8. a Antibody against septin 7, detecting a band at 50 kDa next to product. b Antibody against Tinagl1, detecting a specific band at about 52 kDa, at expected Tinagl1 size (glycosylated) and an additional band at 44 kDa (verified by HEK-cell line lysate), probably unglycosylated

Fig. 3

Proteome analysis data across three cultivation time points for the fed-batch cultivation of CHO DP-12 cells exposed to high osmolality (“feed,” F) or without osmotic change (“control,” C). a Numbers of quantified and significantly regulated (permutation-based FDR < 0.05) proteins between F and C found in CHO proteomes; b profile plots of the log2-ratios F vs. C of the two clusters based on heat map c): cluster 1, proteins with significantly decreased expression on day six and eight in the “feed” condition; 2, proteins with a significantly increased expression on day six and eight in the “feed” condition. c Hierarchical clustering of significantly regulated proteins across three cultivation time points for the fed-batch cultivation of CHO DP-12 cells. High and low expression is shown in red and green, respectively (“T” is an abbreviation for “day” (Tag), “C” indicates “control,” and “F” indicates “feed” condition. Number after the letter indicates biological replicate). d Volcano plots of fold change (LFQ-intensity) in four biological replicates for significantly regulated proteins on day six and day eight in F and C. The plot is represented as a function of statistical significance (t test p ≤ 0.01) between “control” and “feed” condition isolates. The Y-axis indicates p value (− log 10). The X-axis shows the protein ratio (log2 change) in C vs. F conditions. Proteins significantly up-regulated in the feed are highlighted with pink ovals, proteins significantly down-regulated with light blue ones. The top ten proteins (up-regulated in feed) are marked with red dots; the top ten proteins down-regulated in “feed”—blue dots. Proteins with no statistically significant expression differences between the two conditions are shown in gray under the significance cut-off curve. The S0 parameter was set to 0

Fig. 4

STRING visualization of the significantly regulated proteins with enriched annotations resulting from Fischer exact test BH-FDR < 0.02 listed in Table 1. The nodes (spheres) are the proteins and the connecting lines represent STRING interaction (according to STRING: red line, fusion evidence; green line, neighborhood evidence; blue line, co-occurrence evidence; purple line, experimental evidence; yellow line, textmining evidence; light blue line, database evidence; black line, co-expression evidence). Proteins annotated as extracellular cluster are circled in light blue. The table below represents confidence scores (the approximate probability that a predicted link exists in the same KEGG metabolic map) of the interaction nodes. The table below includes interactions with confidence scores ≥ 0.7 (“high confidence”) and ≥ 0.9 (“very high confidence”)

Fig. 5

The major differentially expressed proteins on day 6 and 8 of the fed-batch cultivation between hyperosmolality-exposed CHO-DP12 cells (“feed”) and culture under physiological conditions (“control”). The proteins are arranged according to cellular structures they belong to. Green: proteins, significantly down-regulated in “feed” (Log2 fold change day 8 “feed” vs. day 8 “control” ≥ -0.7); red: proteins, significantly up-regulated in “feed” (Log2 fold change day 8 “feed” vs. day 8 “control” ≥  + 0.7). For full names of the proteins, please refer to Tables 1 and 2. Created with BioRender.com