Purification trials of the human UDP-Xyl/UDP-GlcNAc transporter

Many NSTs have been characterized from different species but the mechanism of recognition and transport, as well as the structure of the protein is not well investigated. This is due to the difficulties associated with expression and purification of membrane proteins in general. In comparison with soluble proteins, limited numbers of structures for transporters have been solved. NSTs are membrane proteins of 300-400 amino acids with ten transmembrane domains. The prediction of transmembrane helices demonstrated that the majority of the protein is very hydrophobic and is in the membrane. In contrast to other membrane proteins the inter-helical loops of transporters are relatively small. To purify the NST from other membrane proteins, solubilization of the protein is required.

NSTs can be produce in E. coli, however the protein is in inclusion bodies. Expression of the transporter as a membrane protein in E. coli CD43 (DE3) failed. The mammalian system is not suitable for high level of NSTs expression. Therefore the yeast expression system was chosen as production platform for protein production.

To express the 5xHis tagged SLC35B4 the yeast strain BY4741 was transformed with SLC35B4 cloned in the pYES2cup vector utilizing the BamHI/XbaI sites (pYES2cup-SLC35B4). The vector leads to translation, under control of the CuP1 promoter, of fusion protein with an N-terminal 5xHis tag. The tag facilitates the detection of the protein as well as purification via Ni-agarose affinity chromatography. The amount of protein and distribution in different fractions were controlled by SDS-PAGE, Western blot and immunostaining with anti 5xHis antibody.

To concentrate the recombinant protein Golgi and ER rich fractions were collected as described in material and methods. Furthermore, both fractions were diluted in phosphate or Tris buffer pH 8.0 to a protein concentration of 2 mg/ml. Aliquots of 100 µl were mixed with solubilization buffer containing different detergents. Several detergents Tween-20, Triton-X100,

NP-40, n-Dodecyl- β -D-maltoside and n-Octyl-β-D-glucopyranoside, were tested. The complete tests were done in both buffers in two variants with final protein concentration of 1 and 0.1 mg/ml. Most tested conditions resulted in precipitation of SLC35B4. Good results were obtained using 0.5% n-Octyl-β-D-glucopyranoside and phosphate buffer. Hundred percent soluble proteins was obtained with a final concentration of 0.1 mg/ml. These conditions were applied for ER and Golgi membranes isolated from 1 l culture and the obtained soluble fraction was used for Ni-agarose affinity chromatography. Nickel agarose chromatography is based on noncovalent binding of histidine residues to the immobilized Ni. The method was applied for purification of huge number of proteins. The obtained soluble fraction was passed via a 1 ml HisTrap column.

Disappointingly, the 5xHis tagged SLC35B4 precipitated in the column and could only be eluted by stripping buffer containing 500 mM Imidazole, 1% SDS and 8 M Urea. Further study demonstrated that precipitation occurred due to concentration of the protein. No soluble protein was detected in the trials when the final protein concentration was above 1 mg/ml.

Due to the first good results obtain with OGP, other detergent from the same group but with longer hydrophobic chain was tested. N-dodecyl-β-D-glucopyranoside (DDG) has twelve carbon aliphatic chain and glucose as a hydrophilic part. The solubility of DDG in water is about 0.008% when the critical mycell concentration is 0.007%. Therefore only one concentration of DDG was tested using two different buffers (phosphate and TRIS) and two final concentrations of the protein. No protein in the soluble fractions was detected. Furthermore, several mixtures of DDG with other detergents were applied for solubilization of SLC35B4. As is shown in figure 21 the DDG in combination with Triton-X100 in phosphate buffer (20 mM sodium phosphate pH 8.0, 10 % Glycerol, 50 mM NaCl) was able to reduce the protein in the pellet.

Although, no soluble protein was detected, this was the first indication of the capability of DDG/Triton-X100 mixtures to solubilize the protein. Better results were obtained by increasing the concentration of Triton-X100. Hundred percent solubility was detected when the concentration of Triton-X100 was increased to 3%. This detergent mixture was further applied for solubilization of large scale ER and Golgi membranes and further purification via a Ni-agarose column.

Results

Figure 21: Solubilization of SLC35B4. Golgi membranes isolated from yeast transformed with N-terminally his tagged SLC35B4 were divided in equal aliquots. The solubilization capabilities of a DDG alone and as combinations with a Triton-X100 and CHAP in phosphate buffer (20 mM sodium phosphate pH 8.0, 10 % Glycerol, 50 mM NaCl) were tested. The soluble protein (s) and the pellet (p) obtained after centrifugation at 15 000xg were analyzed by a SDS-PAGE, western blot and immunostaing with a 5xHis Ab. Arrows indicate reduced amount of the SLC35B4 in the pellet. Soluble SLC35B4 could not be detected.

To facilitate efficient binding, overnight circulation via the column was performed. The bound proteins were washed and further elute as is described in Materials and Methods.

Fractions from different steps were collected and analyzed by SDS-PAGE, Western blot and immunostaining with anti 5xHis antibody. Purity of the protein was estimated after SDS-PAGE and silver staining. As is shown in figure 22 the binding to Ni-agarose of 5xHis tagged SLC35B4 is very efficient and no detectible protein was washed of the column. The His-tag signal appeared in the second elution fraction and is at the level of the dimer SLC35B4. Maximum amounts of the SLC35B4 were detected in the third elution fraction. Additional positive bands were observed at the level of 45-60 kDa. The silver staining clearly illustrated co purification of the SLC35B4 with other proteins. Triton-X100 is able to form very big micelles and often is spotted as a reason for co-purification of different proteins. Therefore, several optimizations of the Ni-agarose affinity chromatography including exchange of the Triton-X100 in the washing and elution buffers with other detergents were tested but that did not improve the results

significantly. Moreover in some of the conditions, precipitations of the protein in the column were detected. The purification trials did not lead to a pure product. As the idea was to separate the recombinantly expressed transporter from endogenous transporters to have a zero background assay system, reconstitution experiments into proteoliposomes were not pursued.

Figure 22: Purification trials of 5xHis tagged SLC35B4 by Ni2+-affinity chromatography. Protein fractions obtained through purification were analyzed by silver staining (lower panel) and a western blot immunostained with the anti 5xHis Ab. Fractions are indicated as: flow trough (FT), washings (W1-5), elutions (E1-9) and striping of the colomn (S1-6). Arrows point at bands that correspond to the expected monomer and dimmer of the SLC35B4 protein.

Results

3.2.3 In vitro expression, reconstitution in artificial proteoliposomes and transport assay of