Structure-Activity Relationships of Bioengineered Heparin/Heparan Sulfates Produced in Different Bioreactors

Abstract

Heparin and heparan sulfate are structurally-related carbohydrates with therapeutic applications in anticoagulation, drug delivery, and regenerative medicine. This study explored the effect of different bioreactor conditions on the production of heparin/heparan sulfate chains via the recombinant expression of serglycin in mammalian cells. Tissue culture flasks and continuously-stirred tank reactors promoted the production of serglycin decorated with heparin/heparan sulfate, as well as chondroitin sulfate, while the serglycin secreted by cells in the tissue culture flasks produced more highly-sulfated heparin/heparan sulfate chains. The serglycin produced in tissue culture flasks was effective in binding and signaling fibroblast growth factor 2, indicating the utility of this molecule in drug delivery and regenerative medicine applications in addition to its well-known anticoagulant activity.

Keywords: bioreactor; heparan sulfate; heparin; proteoglycan; recombinant expression; serglycin.

Figure 1

(A) Schematic of different bioreactors used to culture the HEK-293 cells expressing serglycin including (i) tissue culture flasks, (ii) continuously stirred tank reactors (CSTR), and (iii) shaker flasks; and (B) phase contrast images of cells after three days of culture in the different bioreactor conditions. The scale bar represents 50 μm.

Figure 2

The relative number of cells measured over three days in the different bioreactors, including tissue culture flasks, CSTR, and shaker flasks. Data are presented as means ± standard deviation (n = 3). * indicates significant differences (p < 0.05) compared to tissue culture flasks at day 3 analyzed by one-way ANOVA.

Figure 3

Yield of (A) protein; (B) glycosaminoglycan (GAG); and (C) the ratio of GAG to protein from HEK-293 cells expressing serglycin cultured in different bioreactors over three days and purified by anion exchange chromatography. Protein concentration was measured by Coomassie protein assay and GAG concentration was measured by Dimethylmethylene Blue (DMMB) assay.

Figure 4

The effect of bioreactors on the production of serglycin, heparin/heparan sulfate and chondroitin sulfate. The schematic indicates the structure of serglycin with eight glycosaminoglycan attachment sites that can be decorated with either chondroitin/dermatan sulfate or heparin/heparan sulfate chains. The effect of glycosaminoglycan lyase digestion on the glycosaminoglycan chains are indicated in panels (B,C) with HepIII removing heparin/heparan sulfate chains to reveal a single stub structure and chondroitinase ABC (C’ase ABC) removing chondroitin/dermatan sulfate chains to reveal a stub structure. ELISA for the presence of (A) serglycin core protein; (B) heparin/heparan sulfate stubs detected using anti-heparan sulfate/heparin-stub antibody clone 3G10 following HepIII digestion, and (C) chondroitin sulfate stubs detected using anti-4-sulfated chondroitin sulfate stub antibody clone 2B6 and anti-6-sulfated chondroitin sulfate stub antibody clone 3B3 following C’ase ABC digestion. Data are presented as means ± standard deviation (n = 3). * indicates significant differences (p < 0.05) compared to tissue culture flasks analyzed by one-way ANOVA.

Figure 5

Effect of bioreactors on heparan and chondroitin sulfate structure. ELISA for the presence of (A) heparan sulfate chains detected using anti-heparan sulfate chain antibody clone 10E4 and (B) chondroitin sulfate chains detected using anti-chondroitin sulfate chain antibody clone CS-56. Data are presented as means ± standard deviation (n = 3). * indicates significant differences (p < 0.05) compared to batch cultures analyzed by one-way ANOVA.

Figure 6

The effects of altering glucose concentrations in media for the production of serglycin, heparan sulfate/heparin and chondroitin sulfate. ELISA for the presence of (A) serglycin was detected using a polyclonal anti-serglycin antibody; (B) heparan sulfate/heparin stubs were detected using anti-heparan sulfate stub antibody clone 3G10 following HepIII digestion; (C) heparan sulfate chains were detected using anti-heparan sulfate antibody clone 10E4; and (D) chondroitin sulfate chains were detected using anti- chondroitin sulfate chain antibody clone CS-56. Data are presented as means ± standard deviation (n = 3). * indicates significant differences (p < 0.05) compared to 25 mM glucose analyzed by one-way ANOVA.

Figure 7

Activity of serglycin with heparin/heparan sulfate and chondroitin sulfate chains determined by the signaling of FGF receptor type 1c expressing BaF32 cells in the presence of FGF-2 as mesured by the MTS assay. Cells in the presence of FGF-2 and heparin were used as a control for the formation of active ternary complexes. Negative controls were cells in the presence of no additives, heparin, or FGF-2. Selected serglycin preparations were digested with either chondroitinase ABC, hepIII or both glycosaminoglycan lysases prior to the assay. Cell proliferation was measured after 72 h. * indicated significant differences (p < 0.05) within treatments for cells and FGF-2 compared to cells only as determined by a one-way ANOVA; ** indicated significant differences (p < 0.05) as determined by two-way ANOVA.