Quality assurance after process changes of the production of a therapeutic antibody.

Process development for the production of a therapeutic humanised antibody is a very complex operation. It involves recombinant genetics, verification of a strong expression system, gene amplification, characterisation of a stable host cell expression system, optimisation and design of the mammalian cell culture fermentation system and development of an efficient recovery process resulting in high yields and product quality. Rapid progress in the field and the wish of some pharmaceutical companies for outsourcing their production are the driving forces for process changes relatively late in the development phase. This literature survey is aimed at identifying the limits of acceptable process changes in up scaling of the fermentation and down stream processing of biopharmaceuticals and defining the demand in production validation to prove product equivalency and identity of the isolated, purified therapeutic antibody.

Compartmentalization of the periplasm at cell division sites in Escherichia coli as shown by fluorescence photobleaching experiments.

Morphological evidence has previously indicated that the periplasmic space of Escherichia coli is compartmentalized at sites corresponding to future sites of cell division. The borders of these morphological compartments are formed by localized zones of adhesion (periseptal annuli). In the present study, the technique of fluorescence recovery after photobleaching was used to determine whether these structures act as barriers to the free movement of proteins within the periplasm. The recovery of fluorescence in the ftsA filaments was found to be uniformly low over at potential sites of cell division and at the cell poles, indicating that these regions are biochemically sequestered from the remainder of the periplasmic space. Our results provide direct evidence for local compartments within the periplasm, primarily located at the sites of past or future cell divisions. The implications of this finding for cell division and other periplasmic processes are discussed.

Lateral diffusion of proteins in the periplasm of Escherichia coli.

We have introduced biologically active, fluorescently labeled maltose-binding protein into the periplasmic space of Escherichia coli and measured its lateral diffusion coefficient by the fluorescence photobleaching recovery method. Diffusion of this protein in the periplasm was found to be surprisingly low (lateral diffusion coefficient, 0.9 X 10(-10) cm2 s-1), about 1,000-fold lower than would be expected for diffusion in aqueous medium and almost 100-fold lower than for an equivalent-size protein in the cytoplasm. Galactose-binding protein, myoglobin, and cytochrome c were also introduced into the periplasm and had diffusion coefficients identical to that determined for the maltose-binding protein. For all proteins nearly 100% recovery of fluorescence was obtained after photobleaching, indicating that the periplasm is a single contiguous compartment surrounding the cell. These data have considerable implications for periplasmic structure and for the role of periplasmic proteins in transport and chemotaxis.

Ca2+-induced permeabilization of the Escherichia coli outer membrane: comparison of transformation and reconstitution of binding-protein-dependent transport.

Ca2+ treatment renders the outer membrane of Escherichia coli reversibly permeable for macromolecules. We investigated whether Ca2+-induced uptake of exogenous protein into the periplasm occurs by mechanisms similar to Ca2+-induced uptake of DNA into the cytoplasm during transformation. Protein import through the outer membrane was monitored by measuring reconstitution of maltose transport after the addition of shock fluid containing maltose-binding protein. DNA import through the outer and inner membrane was measured by determining the efficiency of transformation with plasmid DNA. Both processes were stimulated by increasing Ca2+ concentrations up to 400 mM. Plasmolysis was essential for a high efficiency; reconstitution and transformation could be stimulated 5- and 40-fold, respectively, by a high concentration of sucrose (400 mM) in cells incubated with a suboptimal Ca2+ concentration (50 mM). The same divalent cations that promote import of DNA (Ca2+, Ba2+, Sr2+, Mg2+, and Ni2+) also induced import of protein. Ca2+ alone was found to be inefficient in promoting reconstitution; successive treatment with phosphate and Ca2+ ions was essential. Transformation also was observed in the absence of phosphate, but could be stimulated by pretreatment with phosphate. The optimal phosphate concentrations were 100 mM and 1 to 10 mM for reconstitution and transformation, respectively. Heat shock, in which the cells are rapidly transferred from 0 to 42 degrees C, affected the two processes differently. Incubation of cells at 0 degrees C in Ca2+ alone allows rapid entry of protein, but not of DNA. Transformation was observed only when exogenous DNA was still present during the heat shock. Shock fluid containing maltose-binding protein inhibited transformation (with 6 microgram of DNA per ml, half-maximal inhibition occurred at around 300 microgram of shock fluid per ml). DNA inhibited reconstitution (with 5 microgram of shock fluid per ml, half-maximal inhibition occurred at around 3 mg of DNA per ml).

Maltose-binding protein does not modulate the activity of maltoporin as a general porin in Escherichia coli.

Maltoporin (lambda receptor) is part of the maltose transport system in Escherichia coli and is necessary for the facilitated diffusion of maltose and maltodextrins across the outer membrane. Maltoporin also allows the diffusion of nonmaltodextrin substrates, albeit with less efficiency. The preference of maltoporin for maltodextrins in vivo is thought to be the result of an interaction of maltoporin with the maltose-binding protein, the malE gene product. In a recent report Heuzenroeder and Reeves (J. Bacteriol. 144:431-435, 1980) suggested that this interaction establishes a gating mechanism which inhibits the diffusion of nonmaltodextrin substrates, such as lactose. To reinvestigate this important conclusion, we constructed ompR malTc strains carrying either the malE+ gene, the nonpolar malE444 deletion, or the malE254 allele, which specifies an interaction-deficient maltose-binding protein. Lactose uptake was measured at different concentrations below the Km of this transport system and under conditions where transport was limited by the diffusion through maltoporin. We found no difference in the kinetics of lactose uptake irrespective of the malE allele. We conclude that the maltose-binding protein does not modulate the activity of maltoporin as a general outer membrane porin.

Transposon Tn10-dependent expression of the lamB gene in Escherichia coli.

Among Tn10 insertions isolated in or near the malB region of Escherichia coli, one (zjb-729::Tn10) mapped between malK and lamB or late in malK and allowed MalT-independent expression of lamB. Tn10-dependent expression of a lamB-lacZ protein fusion was 25% of the expression of the fusion from the malK-lamB operon promoter in malTc constitutive strains. The maltoporin content of a strain carrying this Tn10 was about 20% that of a malTc malB+ strain. Transport of maltose at concentrations of below 10(-6) M was reduced about threefold. When maltoporin was present at about 50% of the level of malTc malB+ strains, maltose transport was largely restored. We conclude that maltoporin is not rate limiting for maltose transport in wild-type cells but becomes rate limiting when the ratio of maltoporin to other maltose transport components is reduced more than twofold.

Reconstitution of maltose chemotaxis in Escherichia coli by addition of maltose-binding protein to calcium-treated cells of maltose regulon mutants.

Maltose chemotaxis was reconstituted in delta malE cells lacking maltose-binding protein (MBP). Purified MBP was introduced into intact cells during incubation with 250 mM CaCl2 in Tris-hydrochloride buffer at 0 degrees C. After removal of extracellular CaCl2 and MBP, chemotaxis was measured with tethered bacteria in a flow chamber or with free-swimming cells in a capillary assay. About 20% of tethered cells responded to 10(-4) M maltose; the mean response times were about half those of CaCl2-treated wild-type cells (100 s as opposed to 190 s). In capillary tests, the maltose response of reconstituted cells was between 15 and 40% of the aspartate response, about the same percentage as in wild-type cells. The best reconstitution was seen with 0.5 to 1 mM MBP in the reconstitution mixture, which is similar to the periplasmic MBP concentration estimated for maltose-induced wild-type cells. Strains containing large deletions of the malB region and malT mutants lacking the positive regulator gene of the mal regulon also could be reconstituted for maltose chemotaxis, showing that no product of the mal regulon other than MBP is essential for maltose chemotaxis.

Reconstitution of maltose transport in Escherichia coli: conditions affecting import of maltose-binding protein into the periplasm of calcium-treated cells.

The reconstitution of active transport by the Ca2+ -induced import of exogenous binding protein was studied in detail in whole cells of a malE deletion mutant lacking the periplasmic maltose-binding protein. A linear increase in reconstitution efficiency was observed by increasing the Ca2+ - concentration in the reconstitution mixture up to 400 mM. A sharp pH optimum around pH 7.5 was measured for reconstitution. Reconstitution efficiency was highest at 0 degree C and decreased sharply with increasing temperature. The time necessary for optimal reconstitution at 0 degree C and 250 mM Ca2+ was about 1 min. The competence for reconstitution was highest in exponentially growing cultures with cell densities up to 1 X 10(9)/ml and declined when the cells entered the stationary-growth phase. The apparent Km for maltose uptake was the same as that of wild-type cells (1 to 2 microM). Vmax at saturating maltose-binding protein concentration was 125 pmol per min per 7.5 X 10(7) cells (30% of the wild-type activity). The concentration of maltose-binding protein required for half-maximal reconstitution was about 1 mM. The reconstitution procedure appears to be generally applicable. Thus, galactose transport in Escherichia coli could also be reconstituted by its respective binding protein. Maltose transport in E. coli was restored by maltose-binding protein isolated from Salmonella typhimurium. Finally, in S. typhimurium, histidine transport was reconstituted by the addition of shock fluid containing histidine-binding protein to a hisJ deletion mutant lacking histidine-binding protein. The method is fast and general enough to be used as a screening procedure to distinguish between transport mutants in which only the binding protein is affected and those in which additional transport components are affected.

Reconstitution of maltose transport in malB and malA mutants of Escherichia coli.

Pretreatment of lamB cells for 3 h with 10 mM Tris, pH 7.2, containing 25 mM Ca++ resulted in a Ca++-induced shift of the apparent Km of the maltose transport system from about 100 microM to about 15 microM. In contrast to maltose transport in untreated cells, that of Ca++-treated lamB cells was inhibited by anti-MBP (maltose-binding protein) antibodies. The calcium-induced permeability increase of the outer membrane allowed reconstitution of maltose transport in a malE mutant upon exposure to shock fluid or purified MBP. The efficiency of reconstitution, as judged by the Km of the maltose transport system in reconstituted cells, was rather high (Km = 5 microM). Vmax was around 20% of the wild-type. The rapid increase in maltose transport between 2' and 30' of incubation with shock fluid indicated that MBP readily entered the periplasm of Ca++-treated cells. Maltose transport continued for at least 1 h after washing the reconstituted cells. Surprisingly, Ca++ treatment also seemed to allow partial reconstitution of maltose transport in strain SF1701 malT::Tn10 after incubation with cell-free extracts of strain pop1740 malB,malTc.

Reconstitution of maltose transport in malB mutants of Escherichia coli through calcium-induced disruptions of the outer membrane.

The barrier function of the Escherichia coli outer membrane against low concentrations of maltose in strains missing the lambda receptor was partially overcome by treating the cells for 3 h with 25 mM Ca2+. Kinetic analysis of maltose-transport revealed a Ca2+-induced shift of the apparent Km of the system from about 100 microM in cells pretreated with Tris to about 15 microM in cells pretreated with Tris plus Ca2+. In contrast to maltose transport in untreated cells, that of Ca2+-treated lamB cells was inhibited by molecules with a high molecular weight, such as amylopectin (molecular weight, 20,000), and anti-maltose-binding protein antibodies. In addition, lysozyme was shown to attack Ca2+-treated cells in contrast to untreated cells. The Ca2+-induced permeability increase of the outer membrane allowed reconstitution of maltose transport in a mutant missing the maltose-binding protein with osmotic shock fluid containing the maltose-binding protein. Even though Ca2+-treatment allowed the entry of large molecules, the release of the periplasmic maltose-binding protein or alkaline phosphatase was negligible.