The establishment of stable recombinant protein-producing mammalian cell lines is an expensive, time-consuming, tedious procedure. In some cases, expressed recombinant proteins have adverse effects on host cell function, including cell adhesion. Based on the adhesive properties of SCARA5, a scavenger receptor (SR) of the class A SR family, we developed a method for selection of stable recombinant protein-producing cell clones that relies on an internal ribosome entry site (IRES) vector where the protein of interest is expressed in the same messenger RNA as SCARA5, resulting in improved adhesion and increased cell viability of recombinant protein-producing cells in serum-free media. This method does not depend on antibiotics, complicated selective cell culture media or equipment, and thus offers the advantages of being inexpensive, environmentally friendly, and simple.
Various approaches have been developed for large-scale recombinant protein production for protein crystallization and biotechnological purposes. Bacterial expression systems have proven to be reliable, fast, inexpensive, and easy to apply in high-throughput studies. In structural biology, approximately 90% of proteins have been produced using Escherichia coli (1, 2). Cell-free, yeast, baculoviral, and mammalian cell-based expression systems are also available and are often necessary for solving structures of mammalian proteins, since many have proven to be difficult to express in bacterial systems (1). In the biotechnology industry, more than 50% of approved therapeutic proteins are manufactured in mammalian expression systems (3, 4), the only choice when correct glycosylation of a recombinant protein is needed. Indeed, to ensure stability and function and to prevent immunological reactions, recombinant proteins used as biopharmaceuticals must closely resemble their native variants (4, 5).
For protein crystallization, most mammalian proteins are produced in transiently transfected human embryonic kidney (HEK) 293T, 293/EBNA, or 293S GnTI-cells (6, 7). Transient transfection systems are relatively inexpensive and diminish time-consuming selection steps for establishing stable clones (1, 8, 9). An additional advantage of transient systems is that they do not require selection antibiotics, which are expensive and have to be destroyed after a production run. However, transient expression systems provide relatively low recombinant protein yields and require repeated transfections.
For protein crystallography and biotechnology, there is still a need for fast, high-throughput, low-cost selection methods for establishing stable, high-producing mammalian cell lines. In the biotechnology industry, clonality of protein-producing cells is often required. To date, many improvements have been made in protein yields by generating specific host cell lines (e.g., dihydrofolate reductase-deficient CHO cell line) and expression constructs, formulating various cell culture media, and improving bioprocess conditions (2, 3, 10). However, current methods for establishing stable cell lines for large-scale protein production are still very time-consuming, laborious, relatively low-throughput, and dependent on specialized equipment or selection media with antibiotics. Costs and time needed for protein production runs may also increase, since there is often no guarantee that clones with the highest expression of a protein of interest are selected (11).
Scavenger receptors of the class A (SCARA) family are trimeric type II membrane proteins, which are able to bind diverse polyanionic ligands including modified low-density lipoprotein (LDL), polynucleotides, bacterial products, and modified extracellular matrix proteins (12-14). SCARA5 has a similar domain structure to the first identified member in the class, scavenger receptor A (SR-A, SCARA1), which has been shown to promote cell adhesion to serum-coated tissue culture dishes in a cation-independent manner (15, 16). In this article, we show how coexpression of SCARA5 can be advantageous for production of a recombinant protein with adverse effects on such an important host-cell function as its adhesive capacity. We also utilize the adhesion capacity of SCARA5 for developing a selection method for stable protein-producing cell clones. The selection method is simple, fast, and inexpensive, since there is no need for specialized equipment or expensive antibiotics. The method can be applied to HEK-293 and CHO cells. The method also provides selection pressure for clones with high recombinant protein expression levels, which is highly desirable from both protein crystallographic and biotechnologic points of view.Materials and methods LN3G-producing construct
A 1.7-kb fragment encoding the G subdomains 1–3 (LN3G) of the mouse laminin α3 chain was PCR-amplified from a mouse small intestine cDNA library (BD Biosciences, Stockholm, Sweden) using the primers 5′-CGGCCTGCAGGGCCAGAGATGCTGCAAACAAGG-3′ and 5′-GGGGTACCTTAATCTTGCAACAGCTGATTGACATTG-3′ and cloned in frame after a cassette composed of the human TIMP-2 secretion signal sequence and a polyhistidine-tag in the pcDNA3.1/Zeo(-) vector (Invitrogen, Stockholm, Sweden). The plasmid was sequenced to confirm its identity.Generation of LN3G-producing HEK-293/EBNA cell clones
HEK-293/EBNA cells were transfected using the calcium-phosphate method with 40 µg LN3G-expressing construct. Transfected cells were selected with 500 µg/mL Zeocin. Zeocin-resistant clones were tested for recombinant protein production using Western blot analysis with Penta-His antibodies (Qiagen, Sollentuna, Sweden).