Bioengineered Chinese hamster ovary cells with Golgi-targeted 3-O-sulfotransferase-1 biosynthesize heparan sulfate with an antithrombin-binding site

Abstract

HS3st1 (heparan sulfate 3-O-sulfotransferase isoform-1) is a critical enzyme involved in the biosynthesis of the antithrombin III (AT)-binding site in the biopharmaceutical drug heparin. Heparin is a highly sulfated glycosaminoglycan that shares a common biosynthetic pathway with heparan sulfate (HS). Although only granulated cells, such as mast cells, biosynthesize heparin, all animal cells are capable of biosynthesizing HS. As part of an effort to bioengineer CHO cells to produce heparin, we previously showed that the introduction of both HS3st1 and NDST2 (N-deacetylase/N-sulfotransferase isoform-2) afforded HS with a very low level of anticoagulant activity. This study demonstrated that untargeted HS3st1 is broadly distributed throughout CHO cells and forms no detectable AT-binding sites, whereas Golgi-targeted HS3st1 localizes in the Golgi and results in the formation of a single type of AT-binding site and high anti-factor Xa activity (137 ± 36 units/mg). Moreover, stable overexpression of HS3st1 also results in up-regulation of 2-O-, 6-O-, and N-sulfo group-containing disaccharides, further emphasizing a previously unknown concerted interplay between the HS biosynthetic enzymes and suggesting the need to control the expression level of all of the biosynthetic enzymes to produce heparin in CHO cells.

Keywords: Antithrombin; Golgi; Heparan Sulfate; Heparin; Metabolic Engineering; Sulfotransferase.

FIGURE 1.

Anticoagulant heparin. a, mode of action of heparin. During the blood coagulation cascade, serine proteases are activated, resulting in activation of prothrombin to form thrombin and causing blood clot formation. Heparin catalyzes the binding of AT and thrombin, resulting in increased efficacy of the AT-thrombin interaction and decreased clot formation. b, biosynthesis of anticoagulant heparin. The critical steps in heparin biosynthesis involve the concerted action of N-sulfonation (NDST2), epimerization (GLCE), 6-O-sulfonation (6-sulfotransferases, HS6sts), and, finally, GAG chain modification by 3-O-sulfonation (HS3st1). Anticoagulant heparin is characterized by AT pentasaccharide-binding sites. IdoA, IdoUA; GlcA, GlcUA; C5epi, C5-epimerase.

FIGURE 2.

CHO-S dual clones demonstrate highly N-sulfated heparin/HS. a, metabolic engineering strategy showing that CHO-S cells were first bioengineered to stably overexpress NDST2 before being stably transfected with HS3st1 to prepare 40 stable dual clones. b, HS disaccharide composition analysis performed on GAGs isolated from CHO-S, CHO-ndst2, and dual clones. Heparin/HS disaccharide composition analysis was performed on GAGs isolated from the cell pellet. All of the dual clones showed highly N-sulfated HS. Expression of NDST2 and HS3st1 varied among the selected clones (Dual-3, Dual-10, Dual-20, Dual-22, and Dual-29). c, metabolic engineering strategy showing CHO-S cells stably transfected with Golgi-targeted HS3st1 plasmid. d, HS disaccharide composition analysis performed on GAGs isolated from parental CHO-S and CHO-gt31 cell pellets. CHO-gt31 cells demonstrated an increase in TriS. HS disaccharide analysis was performed by AMAC/LC-MS. Heparin was used as a positive control.

FIGURE 3.

Expression and localization of HS3st1 and NDST2 in CHO-S, Dual-29, Dual-10, and CHO-gt31 cells. Overexpression of HS3st1 in Dual-29 cells resulted in a broad distribution of HS3st1. Overexpression of HS3st1 in Dual-10 cells localized HS3st1 in the Golgi. Overexpression of Golgi-targeted HS3st1 in CHO-gt31 localized HS3st1 in the Golgi. NDST2 was localized in the Golgi. DAPI was used to visualize nuclei, and anti-GM130 Golgi marker antibodies were used to stain the Golgi compartment. Colocalization experiments were performed using confocal microscopy. Dual-29 cells were used as the positive control. For imaging, signal intensities were adjusted for Dual-29 cells, and other cells were imaged at the same frequencies and configurations.

FIGURE 4.

CS/DS structural glycomics of CHO-S, CHO-ndst2, Dual-10, Dual-29, and CHO-gt31 cells. a, ratio of HS and CS/DS from total GAGs isolated from cell pellets. Data represents the mean ± S.D. (n = 2). b, CS/DS disaccharide composition analysis.

FIGURE 5.

Proteglycanomics (GAGs from the cell pellet) of wild-type CHO-S cells and CHO-S cells overexpressing EXT1, EXT2 and EXT1/2 biosynthetic enzymes. a, HS GAGs were isolated from the cell pellet, and HS disaccharide composition was analyzed by AMAC/LC-MS. An increase in TriS HS disaccharides, suggestive of elevated action of N-, 2-O-, and 6-O-sulfotransferases, was observed in CHO-S cells overexpressing EXT2 (CHO-ext2) and EXT1 and EXT2 (CHO-ext1/2). b, CS/DS GAGs disaccharide analysis. An increase in 0S CS disaccharides was observed in CHO-S cells overexpressing EXT1 and EXT2. c, percentage of HS and CS/DS GAGs. An increase in CS/DS GAGs was seen in CHO-S cells overexpressing EXT1 and EXT2. d, normalized binding activity of AT and FGF-2 in wild-type, EXT1-overexpressing, EXT2-overexpressing, and EXT1/2-overexpressing CHO-S cells by flow cytometry (n = 10000 events). wrt, with respect to.

FIGURE 6.

Fractionation of GAGs by strong anion exchange chromatography demonstrates the existence of different populations of sulfated HS in wild-type and recombinant CHO cells. The CS/DS GAGs were removed from the preparation before heparin/HS GAG quantification. a, experimental design showing that total GAGs were isolated from cell samples and subjected to a salt gradient. Fraction 1 (eluted at 0.5 m NaCl) contained mainly less sulfated HS GAGs, and Fraction 2 (eluted at 1.6 m NaCl) contained mainly highly sulfated heparin/HS GAGs. b, quantification of Fractions 1 and 2. c, disaccharide profile of fractionated GAGs from Fraction 1. d, disaccharide analysis of fractionated GAGs from Fraction 2. The composition of pharmaceutical heparin is shown in both c and d for comparison.

FIGURE 7.

Oligosaccharide mapping of bovine lung heparin and HS GAGS from CHO-gt31, Dual-29, Dual-10, and wild-type CHO-S cells by AMAC/LC-MS analysis. a, tetrasaccharide structures in CHO-S, Dual-29, Dual-10, and CH-GT31 cell pellets are compared with bovine lung heparin. b, structures were assigned by MS analysis. The degree of sulfation of tetrasaccharide structures increased in successive tetrasaccharides, i.e. T1 < T2 < T3 < T4 < T5.

FIGURE 8.

Anti-factor Xa assay. Bioengineered HS from CHO-gt31 cells was fractionated, and anti-factor Xa activity was assessed with heparin as the control. Values are the mean ± S.E. (n = 3).