Section 3.1.4.3. Baculoviruses and insect cells

Baculovirus expression systems are becoming increasingly important tools for the production of recombinant proteins for X-ray crystallography. The insect cell–virus expression systems are more experimentally demanding than bacteria or yeast, but they offer several advantages. Expression of some mammalian proteins has been achieved in baculovirus when simpler expression systems have failed. Because insects are higher eukaryotes, many of the difficulties associated with expression of proteins from higher eukaryotes in E. coli do not apply: there is no need for a Shine–Dalgarno sequence, no major problems with codon usage and fewer problems with a lack of appropriate chaperones. Although glycosylation is not the same in insect and mammalian cells, in some cases it is close enough to be acceptable. In addition, for many crystallography projects, minimizing glycosylation is helpful, so that it may be more appropriate to modify the gene or protein to avoid glycosylation (or minimize it) than to try to find ways to recapitulate the glycosylation pattern found in mammalian cells. As is the general case in biotechnology, the development of baculovirus expression systems is work in progress. Progress has been made towards making recombinant proteins in insect cells with glycosylation patterns that match those in mammalian cells (reviewed by Jarvis et al., 1998). Baculovirus systems allow expression of recombinant proteins at reasonable levels, typically ranging from 1–500 mg l⁻¹ of cell culture. Considerable work has gone into the development of convenient transfer vectors, and baculovirus expression kits are available from more than ten different commercial sources.

Baculoviruses usually infect insects; in terms of the expression of foreign proteins, the important baculoviruses are the Autographa californica nuclear polyhedrosis virus (AcNPV) and the Bombyx mori nuclear polyhedrosis virus (BmNPV). AcNPV has been used more widely than BmNPV in cell-culture systems; the BmNPV virus is used primarily to express recombinant proteins in insect larvae. The advantage of BmNPV is that it grows well in larger insect larvae, making the task of harvesting the haemolymph easier. Proteins expressed for crystallography have all been, to the best of our knowledge, expressed using the AcNPV virus system; we will not discuss the BmNPV virus expression system here. Anyone wishing to learn more about either AcNPV or BmNPV is urged to consult two useful monographs: Baculovirus Expression Protocols (Richardson, 1995) and Baculovirus Expression Vectors: A Laboratory Manual (O'Reilly et al., 1992). There are also shorter reviews that are quite helpful (Jones & Morikawa, 1996; Merrington et al., 1997; Possee, 1997).

In nature, in the late stage of replication in insect larvae, nuclear polyhedrosis viruses produce an occluded form, in which the virions are encased in a crystalline protein matrix, polyhedrin. After the virus is released from the insect larvae, this proteinaceous coat protects the virus from the environment and is necessary for the propagation of the virus in its natural state. However, replication of the virus in cell culture does not require the formation of occlusion bodies. In tissue culture, the production of occlusion bodies is dispensable, and the primary protein, polyhedrin, is not required for replication. Cultured cells infected with wild-type AcNPV produce large amounts of polyhedrin; cells infected with modified AcNPV vectors, with other genes inserted in place of the polyhedrin gene (or in place of another highly expressed gene, p10, that is dispensable in cultured cells), can express impressive amounts of the recombinant protein.

The AcNVP genome is 128 kb, which is too large for convenient direct manipulations. In most cases, novel genes are put into the AcNPV genome by homologous recombination using transfer vectors. Transfer vectors are small bacterial plasmids that contain AcNPV sequences that allow homologous recombination to direct the insertion of the transfer vector into the desired place in the AcNPV genome (often, but not always, the polyhedrin gene). Originally, the purified circular DNA from AcNPV and the appropriate transfer plasmids were simply cotransfected onto monolayers of insect cells. Plaques develop, and if the insertion is targeted to the polyhedrin gene, plaques that contain viruses that retain the ability to make polyhedrin (those that contain the wild-type virus) can be distinguished in the microscope from plaques that do not. This technique works, but has been largely replaced by systems that make it easier to obtain and/or find the recombinant plaques. The AcNPV genome is circular; if the DNA is linearized, it will not produce a replicating virus unless the break is repaired. The repair process is facilitated by the presence of homologous DNA flanking the break. Systems have been set up to exploit this property to increase the efficiency of the generation of vectors that carry the desired insert. Basically, the genome of the AcNPV vector is modified so that there is a unique restriction site at the site where the transfer vector would insert. Linear AcNPV DNA is cotransfected with a transfer vector. This can produce stocks in which greater than 90% of the virus is recombinant. Systems have also been developed in which a DNA insert can be ligated directly into a linearized AcNPV genome. This protocol also produces a high yield of recombinant virus (Lu & Miller, 1996).

There are also a number of systems that allow either the selection or, more often, the ready identification of recombinant virus. The marker most commonly used for this purpose is β-galactosidase; a number of AcNPV vectors or transfer systems that make use of β-galactosidase are commercially available. Once a recombinant plaque is identified, it should be purified through multiple rounds of plaque purification to ensure that a homogeneous stock has been prepared. Several independent isolates should be prepared and each checked for expression of the desired protein.

There are several important things to consider when setting up the cell-culture system. Although most baculoviruses have a relatively restricted host range, and AcNPV was first isolated from alfalfa looper (Autographa californica), for the purpose of expressing foreign proteins, it is usually grown in cells of the fall armyworm (Spodoptera frugiperda) or the cabbage looper (Trichoplusia ni). The isolation and purification of the appropriate AcNPV vectors are usually done in monolayer cultures. In contrast, the production of large amounts of recombinant protein is usually done in suspension cultures. There is also the issue of whether or not to include fetal calf serum in the culture media. In theory, since the cells can be grown in serum-free media, which saves money and makes the subsequent purification of the recombinant protein simpler, serum-free culture is the appropriate choice. However, growing cells in serum-free media is a trickier proposition, and the cells are more sensitive to minor contaminants. As a general rule, high-level production of recombinant proteins using a baculovirus vector requires host cells that are growing rapidly; this is sometimes easier to achieve with serum-containing media. It is not always a simple matter to switch cells adapted to growth on plates to suspension culture, nor is it always easy to switch cells grown in the presence of serum to serum-free culture. Since the vector is a virus, it is usually more convenient to use cells adapted to different conditions than to try to adapt the cells. However, the relative yield of the recombinant protein will not necessarily be the same in different cells grown under different culture conditions.

Although baculoviruses, particularly AcNPV, are convenient vectors, the expression of the recombinant protein is carried out by the insect cell host. Baculovirus infection kills the host cell, so it is not possible to use baculoviruses to derive insect cell cultures that continuously express a recombinant protein. It is possible, however, to introduce DNA segments directly into insect cells and derive cell lines that stably express a recombinant protein; there are constitutive and inducible promoters that can be used in insect cell systems (McCarroll & King, 1997; Pfeifer, 1998). Basically, the protocols used to introduce DNA expression constructs into cultured insect cells are similar to those used in cultured mammalian cells (CaPO₄, electroporation, liposomes etc.), and similar selective protocols are used (G418, hygromycin, puromycin etc.).

Expression systems have been prepared based on baculovirus immediate early promoters and on cellular promoters, including the hsp70 promoter and metallothionein (McCarroll & King, 1997; Kwong et al., 1998; Pfeifer, 1998). Insect cells are, in general, easier (and cheaper) to grow in culture than mammalian cells, although many of the problems that exist in mammalian cell culture also exist in insect cell culture. Relative to the baculovirus system, the use of stable insect cell lines not only allows the continuous culture of cells that contain the desired expression system (provided the expressed protein is not too toxic), it also permits the use of Drosophila cell lines, which appear to have some advantages for the high-level production of recombinant proteins.

Compared to bacteria or yeast cells, cells from higher eukaryotes are quite delicate, and considerable care must be taken in cell culture. The cells are subject to shear stress, which can be a problem in stirred and/or shaken cultures; some researchers use airlift fermenters to help alleviate the problem. Compared to yeast and bacterial cells, cultured cells grow relatively slowly and require rich media that will support the rapid growth of a wide variety of unwanted organisms, so special care must be taken to avoid contaminating the cultures. Antibiotics are commonly used; however, antibiotics will not, in general, prevent contamination with yeasts or moulds, which often cause the greatest problems. If the baculovirus system is used, then the cells and viruses are kept separate, and the cells are relatively standard reagents. If there is contamination, the contaminated cultures can be discarded and replaced with fresh cells (and viruses). Stable transformed insect cells that express a recombinant protein must be kept free of all contaminants. As is always the case, both cells and viruses should be carefully stored. Any useful recombinant baculovirus can be easily stored as DNA.