Regulatory issues in the use of insect-cell culture.

CONCLUSION: The insect cell as host for protein production is relative new. Therefore few data are available. This creates a vicious circle because it makes the choice of insect cells as basis for a pharmaceutical process less attractive. There are three main issues when comparing insect-cells to "traditional" systems as mammalian and bacterial cells. First, since the expression vector is not incorporated in the cells, a virus stock similar to the cell bank system has to be laid down and tested. This will cost time and money. Secondly the vector is subject to mutation and therefore the decrease in infectivity has to be characterized and validated. Third, the post-translational modification of the protein may differ. None of the mentioned issues, however, forms an obstacle that can not be overcome.

Continuous beta-galactosidase production in insect cells with a p10 gene based baculovirus vector in a two-stage bioreactor system.

Continuous production of polyhedra or baculovirus-expressed proteins in insect cell cultures is limited to about 4 weeks. The decrease in production has been ascribed to the interference of defective deletion mutants with wild-type baculoviruses. The deleted genome sequences include the polyhedrin gene (or the heterologous gene of interest); in the remaining part, the major late p10 gene is always maintained. In the present study, the productivity of a recombinant baculovirus with the lacZ gene from Escherichia coli cloned downstream of the p10 promoter at the p10 locus was investigated. It was hypothesized that this p10 promoter driven gene is preserved over a longer period of time in a continuously operated two-stage bioreactor system than foreign genes behind the polyhedrin promoter at the polyhedrin locus. In two separate runs, beta-galactosidase production with the p10-lacZ recombinant reached quasi-steady-state levels of 30 and 60 units/cm3. Polyhedron production was about 3 x 10(6) and 6 x 10(6) polyhedra/cm3, respectively. However, both polyhedron and beta-galactosidase production decreased after about 30 days of relatively constant production. In the infection reactor, deletion mutants of the virus, which contained both the polyhedrin and lacZ gene, were predominant. Therefore, the presence of the polyhedrin or p10 gene alone in deletion mutants is not sufficient for prolonged expression; other genes involved in major late gene expression and not present in the deleted virus are probably necessary.

A structured dynamic model for the baculovirus infection process in insect-cell reactor configurations.

A mathematical description of the infection of insect cells with baculovirus in a continuously operated reactor configuration is presented. The reactor configuration consists of one bioreactor in which insect cells (Spodoptera frugiperda) are grown followed by one or two bioreactors in which cells are infected by a baculovirus (Autographa californica nuclear polyhedrosis virus). It was demonstrated that the so-called passage effect is responsible for the observed difference in run time between a configuration with one or with two infection vessels. Furthermore, a model is presented based on the hypothesis that the limited run time of series of continuously operated bioreactors is associated with the occurrence of a virus particle (so-called virion) that is defective and has interfering properties. With the assumption that not all nonoccluded virions are capable of establishing a correct infection leading to new virus production, infection levels in continuously operated reactor configurations could be described well with the model.

Continuous beta-galactosidase production with a recombinant baculovirus insect-cell system in bioreactors.

Insect cells were exploited to produce bacterial beta-galactosidase by infecting them with a recombinant nuclear polyhedrosis virus (baculovirus) of Autographa californica. The insect cells were cultured in a continuous stirred tank reactor (CSTR) and led to a second CSTR where they were infected with a recombinant virus in which the lacZ gene from Escherichia coli was inserted. In the effluent of the production reactor, maximum activities of 15 units beta-galactosidase per 10(6) cells were measured. For about 25 d beta-galactosidase production remained constant, but then rapidly declined. This drop was due to a decrease in production of active beta-galactosidase rather than to inactivation of this enzyme. It was concluded that the reduced production was due to reduced polyhedrin promoter-driven synthesis.

Detection and analysis of Autographa californica nuclear polyhedrosis virus mutants with defective interfering properties.

Defective interfering particles (DIPs) were generated upon continuous production of Autographa californica nuclear polyhedrosis virus (AcNPV) in bioreactors. This configuration mimicked the serial undiluted passaging of virus, which is known to result in plaque-morphology mutants. Restriction enzyme analysis of DIP-containing preparations of extracellular virus showed the presence of many DNA fragments in less than equimolar amounts. These fragments were colinear on the physical map of AcNPV and extended from map position 1.7 to 45. These DIPs thus lacked 43% of the genetic information of the standard virus, including the polyhedrin and DNA polymerase genes. The existence of DIPs was confirmed by electron microscopy, where virions were observed with reduced length. Among the less than equimolar fragments in DIP-containing preparations, fragments were observed linking sequences from map positions 1.7 and 45 via a TGTT linker of unknown origin. The DIPs could not be plaque-purified and needed standard (helper) virus to replicate; DIP-containing preparations interfered with standard virus replication in an interference assay, which explained the reduction in productivity of an AcNPV expression vector-insect cell system in continuous bioreactor operations. The origin of these DIPs and their possible generation mechanism are discussed.

Continuous production of baculovirus in a cascade of insect-cell reactors.

Insect cells (Spodoptera frugiperda) were cultured in a continuous stirred-tank reactor. The effluent was led to a cascade of another two reactors, each containing half the volume of the cell-growth reactor, where the cells were infected with Autographa californica nuclear polyhedrosis virus. For about 10 days production of 10(7) polyhedra (virus particles embedded in a protein capsule) per cm3 was achieved. This short production time compared to previous experiments involving an analogous system with a single infection vessel of equal volume to the cell-growth vessel is ascribed to the accelerated occurrence of the so-called passage effect (a decrease of infectious virus with time). From the results of a computer model it was concluded that this passage effect was accelerated by the change in residence time distribution as compared to the earlier experiments.