338
TIBTECH - DECEMBER 1990 [Vol. 8]
Interestingly, some biological polymers bind copper in the Cu ÷ form, some in the Cu 2÷ form, and some fail to bind copper ions. Most of the evidence for microbial involvement in metal corrosion is circumstantial, but Geesey believes that recent understanding of microbial biofilm processes should make it easier to relate specific biochemical activities of the organisms to the electrochemical reactions leading to corrosion. Bob Tatnell (DuPont de Nemours & Co., DE, USA) provided a series of case histories of biocorrosion in various metals in different industrial situations and under different environmental conditions. He noted that, although precise figures are not available, biological factors probably play a role in about half of the metal-corrosion events reported worldwide. Allan Hamilton (University of Aberdeen, UK) noted that the sulphate-reducing bacteria (SRB) are widely implicated in biocorrosion in natural and industrial environments.
[]
[]
[]
These anaerobic bacteria function efficiently at the biofilm-substratum interface where oxygen availability is limited both by slow diffusion through the biofilm matrix and by bacterial utilization closer to the biofilm-water interface. In the absence of oxygen, protons may serve as electron acceptors at the cathode leading to the production of molecular hydrogen, which the SRB subsequently oxidize during sulphate reduction, giving rise to sulphide and the potential for metal sulphide corrosion products. The reactions are very complex, as SRBinduced corrosion is stimulated by oxygen and involves consortium relationships between the SRB and other microorganisms; points emphasized by both Allan Hamilton and Bill Costerton (University of Calgary, Alberta, Canada). Costerton described a variety of newly developed techniques for the detection, monitoring and control of biocorrosion. Some of the techniques have found ready acceptance (e.g. the
[]
[]
[]
[]
[]
[]
Separations for biotechnology Robust and efficient separation processes are essential for the manufacture of biochemical products. Over the past decade, the demand for improved processes to purify proteins has increased dramatically following the successful molecular cloning and expression of recombinant proteins. In response to this demand, research activities in downstream processing have increased, focusing on purification of high molecular-weight macromolecules (primarily proteins), developing a better understanding of individual unit operations; integration of process strategies for purification, and devel* The Second International Conference on Separations for Biotechnology was held at the University of Reading, UK, 10-13 September 1990. The proceedings of this conference have been published in Separations for Biotechnology, edited by D. L. Pyle, Elsevier Applied Science Press, London, 1990. (~) 1990, Elsevier Science Publishers Ltd (UK)
oping new and innovative methods for purification. The papers and posters presented at the recent conference on Separations for Biotechnology* clearly reflect these current trends. The plenary speaker was Prof. A. S. Michaels (Chestnut Hill, MA, USA) who placed the state and direction of research in downstream processing in perspective. Achieving reproducible high standards of product purity and quality must remain the goal for those in the business of biochemical process manufacturing. Today's goal is to make high purity therapeutic compounds. In the future, a new challenge will emerge: to extend our capability to make a wider range of materials for 'speciality' as well as 'commodity' markets. The technical challenge is to maintain high product standards while adapting new techniques to largerscale operation; competitiveness will depend increasingly on efficient implementation of process technology.
0167 - 9430/90/$2.00
Robbins device for monitoring biofilm and biocorrosion development), whereas others remain to be more thoroughly tested (e.g. a hydrogenase test for the presence of SRB). The fact that this workshop was held in Germany indicates an increasing awareness in Europe, in recent years, of the importance of biofouling and biocorrosion. Workshops similar in scope have been held during the last five years by Water Micro Associates in the USA. Problems resulting from biofilm formation in flowing water systems are of worldwide occurrence, and the costs resulting from biofouling and biocorrosion are enormous. It will be interesting to note developments over the next few years as more engineers and microbiologists realize the importance to industry of these microbial processes. KEVIN M A R S H A L L
School of Microbiology, The University of N e w South Wales, Kensington, Sydney, Australia.
[]
[]
[]
The conference organization followed closely the sequence of unit operations in biochemical processes.
Cell harvesting Since many biochemical products of commercial interest are intracellular, the initial papers dealt with methods of cell harvesting and cell disruption. The two common approaches to cell harvesting are centrifugation and filtration. P. N. Ward (ICI Biological Products, Billingham, UK) described work clone in collaboration with M. Hoare (University College London, UK) on elucidating the dependence of dewatering performance in a scroll decanter centrifuge on rheological properties of the suspension. An understanding of the chemistry, as well as the physics of the problem is essential for rational optimization of performance. Viscoelasticity and network property characterization were shown to provide a basis for the design of a de-watering operation. The viscoelasticity of borax-flocculated cells was characterized as both shear moduli and loss angle as a
TIBTECH
-
DECEMBER 1990 [Vol. 8]
function of oscillation frequency in a pulse shearometer. With this information, as well as the network modulus measured as a function of cell volume fraction, the authors were able to design operating conditions to de-water borax-flocculated cells to 24% solids in a decanter centrifuge. The significance of this lies in the ability to use laboratory characterization of rheology to design practical operating conditions. An alternative to centrifugation is membrane filtration for cell harvesting. M. Pritchard, with colleagues J. A. Scott and J. A. Howell (University of Bath, UK), presented a paper on concentrating yeast by cross-flow filtration with ultra- and microfiltration membranes. While these devices have been available for many years, membrane filtration is used infrequently for cell harvesting. Part of the problem is the inability to predict performance from first principles. As described by Pritchard, the filtration flux during yeast concentration generally decreases with increasing concentration. However, a surprising result was observed when comparing tubular and flatsheet systems; in the flat-sheet membrane module, when the wall shear stress exceeded a threshold value, the filtration flux began to increase, rather than decrease, with concentration. As a consequence, it was possible to operate the module at its maximum pressure drop before the flux decreased to zero with increasing suspension viscosity. In tubular modules, this shear-induced effect did not occur before there was a transition from turbulent to laminar flow followed by a drop in flux to near zero. These results offer further encouragement to the use of membrane filters for cell harvesting.
Cell disruption Alternative approaches to cell disruption were covered in a series of papers and reviewed by J. A. Asenjo (University of Reading, UK). Although high-pressure homogenization and bead mills are routinely used, it became apparent at this conference that directed protein secretion was becoming the method of choice for separating cells from product. This avoids brute force disruption which could create problems downstream. T. J. Naglak and
339
H: Y. Wang (University of Michigan, USA) described the use of selective chemical permeabilization for recovery of proteins from Pichia pastoris and E. coll. Guanidine-HC1 with Triton X-100 were shown to act synergistically to release ~-lactamase from E. coil Both yeast and bacteria remain intact, thus facilitating separation of cells from broth after protein release. As much as 40-fold purification could be achieved in a protein-release step that often does not provide any purification. While the practitioner usually tries to avoid the addition of chemicals during purification, the results described here may provide sufficient justification to pursue this approach as part of an integrated strategy for efficient recovery.
Product concentration Once a product is separated from the cells, the next step is to reduce the volume by concentration. Several approaches were taken to address this problem. In current practice, filtration and adsorption are most commonly used for this purpose with large molecular-weight molecules, and extraction is used for low molecular-weight materials. C. L. Cooney (Massachusetts Institute of Technology, Cambridge, MA, USA) described the enhanced performance of vortex-flow ultrafiltration for protein and cell concentration. By creating a secondary flow (Taylor vortices) with a high-speed rotating membrane, it is possible to maintain high flux rates even when protein and cell concentrations are high. The important operating and design parameters were identified and high performance and predictable operation will lead to broader use of this technique in bioprocessing. Several approaches to the use of aqueous biphasic extraction of high proteins were described by J. G. Huddlestone (University of Birmingham, UK), and V. Riveros-Moreno (Wellcome Biotech, UK). This technique continues to be interesting and works well for specific products, however, its use is still restricted to only a few materials. Presumably this is a consequence of the difficulty in predicting performance (and thus designing from first principles), and in recovering the used solvent. Both of these aspects of the technology require further study.
High-resolution purification The most challenging problems of downstream processing are in highresolution separations. H. Chase (University of Cambridge, Cambridge, UK) reviewed the use of adsorptive and extractive processes for purification. This area of downstream processing has drawn more creativity and innovation than any other. As pointed out by Chase, one can use specific affinity techniques to adsorb, extract, precipitate or neutralize target molecules. Alternatively, a genetic approach to solving process problems can be taken. Modifying the product gene to incorporate a target in the protein product that can be recognized by a specific affinity ligand places a strategic 'fish hook' on the molecule of interest. However, these procedures are class separations and, alone, do not lead to sufficient purity for therapeutic use. One of the problems that emerges in manufacturing of proteins is the formation of isomeric or modified forms of the desired molecule. Affinity techniques pick out molecules with a specific epitope or recognition site b u t may not distinguish between variations in other regions of the molecule. In this, as in other conferences on separations, too little attention is given to the problem of purifying the desired product from other species. This difficult problem requires further investigation. The problem of structural variation for 7-interferon was nicely described by E. Curling (University of Kent, UK). The variability of posttranslational modification of 7-interferon produced by recombinant CHO cells occurring in response to process conditions suggests that care is needed in design and operation of the bioreactor. Obviously, upstream operation can affect downstream product purification, and the entire process needs to be examined as an integrated unit and not as a sequence of individual steps.
Process integration The problems of process integration and design were considered in a series of papers and were reviewed by K. Mosbach (University of Lund, Sweden) and by H. Y. Wang (University of Michigan, USA). Some dealt with the integration of multiple tasks into a single-unit operation
340
TIBTECH - D E C E M B E R 1990 [Vol. 8]
while others examined methodologies to design fully integrated processes. The consensus opinion was that integration of processes is the direction to take, though there were diverse views as to h o w best to achieve this goal. Combining multiple functions into a single operation has the benefit of reducing the number of steps and increasing the overall yield, but does not necessarily lead to process robustness. When two steps are used in sequence, for example, if the first is variable in performance, then the second can be designed to remove impurities or contaminants that flow through the first. If a single-unit operation is used then this possibility is lost. Application of computer-aided design techniques works well for cost minimalization but can not predict trace impurities or contaminants that determine product quality because analytical descriptions are, as yet, too imprecise. []
[]
[]
Product finishing An important area of downstream processing is product finishing. This topic is neglected in most conferences but was nicely addressed on this occasion. C. R. Hill (Celltech Ltd, UK) reviewed the problems associated with product finishing and pointed out that impurities and contaminants can be introduced into the process from raw materials, hostcell synthesis, and in-process reactive conditions. Attention to detail in operation and design, development of sensitive analytical methods, and establishing rigorous quality controls are essential to the success of any manufacturing process for therapeutic materials. These points were amplified by the other speakers in an entire session devoted to this important topic.
1990, Elsevier Science Publishers Ltd (UK)
C H A R L E S L. C O O N E Y
The search for purity This conference provided an excellent opportunity for researchers []
[]
[]
[]
[]
[]
Altered gene expression in plants due to trans interactions between homologous genes Two recent publications 1,2 on the genetic engineering of flower color have reported the intriguing observation that ectopic transgenes can participate in a homology-based interaction with endogenous plant genes, resulting in suppression of the expression of both the ectopic gene and the endogenous gene. Better understanding of this phenomenon may eventually shed some light on the dynamic behavior of genes in the nucleus. Furthermore, the phenomenon is likely to have significant practical applications in the manipulation of plant phenotypes for use in agriculture and the plant sciences.
and practitioners to come together and discuss work on separations for biotechnology. While advancements were cited on all fronts, it is disappointing that little attention was given by the individual researchers to the problems of removing lowlevel impurities and contaminants such as viruses, nucleic acids and isomeric or modified forms of proteins. The need to reach levels of 99.999% purity, or better, is the most difficult challenge today in protein purification for therapeutic use; for some reason researchers seem unwilling or unable to accept this challenge. Perhaps we can look forward to advances i n this area, too, in the next of this series of conferences on Separations for Biotechnology.
Co-suppression of anthocyanin biosynthetic enzymes in petunia In many plants, flower color is determined by the production of anthocyanin pigments. In an attempt to overproduce flower pigments in petunias, a chimeric gene encoding chalcone synthase (CHS), a key enzyme in the anthocyanin pigment biosynthetic pathway (Fig. 1), was introduced; the results, however, were surprising in several ways. First, the introduction of a CHS transgene to ectopic sites in the petunia genome was found to suppress the expression of both alleles of the endogenous CHS gene in trans, causing the production of pure
0167 - 9430/90/$2.00
Massachusetts Institute of Technology, Cambridge, MA 02139, USA. []
[]
[]
white or patterned flowers 1 (see Box 1). Similar results were obtained independently with CHS, and then also with another petunia anthocyanin gene encoding the enzyme dihydroflavonol-4-reductase (DFR) (Fig. 1) 2. It was demonstrated that in both cases only homologous gene expression was suppressed, not that of other genes in the pathway 1.2. However, the really surprising result was that suppression of the endogenous gene was not simply the result of transgene expression. This could be shown because the suppression phenomenon was somatically reversible in both CHS and DFR transformants (i.e. white flowers sometimes reverted to wildtype purple flowers) (see Box 1). In purple revertant flowers the expression levels of both the ectopic transgene and the endogenous homologous gene were found to increase coordinately and substantially (50fold) restoring expression of the endogenous gene to normal levels 1,2. This indicated that both genes must be coordinately suppressed in the white flowers, suggesting the provisional term 'co-suppression' for this phenomenon 1.
TIBTECH - DECEMBER 1990 [Vol. 8]
Interestingly, some biological polymers bind copper in the Cu ÷ form, some in the Cu 2÷ form, and some fail to bind copper ions. Most of the evidence for microbial involvement in metal corrosion is circumstantial, but Geesey believes that recent understanding of microbial biofilm processes should make it easier to relate specific biochemical activities of the organisms to the electrochemical reactions leading to corrosion. Bob Tatnell (DuPont de Nemours & Co., DE, USA) provided a series of case histories of biocorrosion in various metals in different industrial situations and under different environmental conditions. He noted that, although precise figures are not available, biological factors probably play a role in about half of the metal-corrosion events reported worldwide. Allan Hamilton (University of Aberdeen, UK) noted that the sulphate-reducing bacteria (SRB) are widely implicated in biocorrosion in natural and industrial environments.
[]
[]
[]
These anaerobic bacteria function efficiently at the biofilm-substratum interface where oxygen availability is limited both by slow diffusion through the biofilm matrix and by bacterial utilization closer to the biofilm-water interface. In the absence of oxygen, protons may serve as electron acceptors at the cathode leading to the production of molecular hydrogen, which the SRB subsequently oxidize during sulphate reduction, giving rise to sulphide and the potential for metal sulphide corrosion products. The reactions are very complex, as SRBinduced corrosion is stimulated by oxygen and involves consortium relationships between the SRB and other microorganisms; points emphasized by both Allan Hamilton and Bill Costerton (University of Calgary, Alberta, Canada). Costerton described a variety of newly developed techniques for the detection, monitoring and control of biocorrosion. Some of the techniques have found ready acceptance (e.g. the
[]
[]
[]
[]
[]
[]
Separations for biotechnology Robust and efficient separation processes are essential for the manufacture of biochemical products. Over the past decade, the demand for improved processes to purify proteins has increased dramatically following the successful molecular cloning and expression of recombinant proteins. In response to this demand, research activities in downstream processing have increased, focusing on purification of high molecular-weight macromolecules (primarily proteins), developing a better understanding of individual unit operations; integration of process strategies for purification, and devel* The Second International Conference on Separations for Biotechnology was held at the University of Reading, UK, 10-13 September 1990. The proceedings of this conference have been published in Separations for Biotechnology, edited by D. L. Pyle, Elsevier Applied Science Press, London, 1990. (~) 1990, Elsevier Science Publishers Ltd (UK)
oping new and innovative methods for purification. The papers and posters presented at the recent conference on Separations for Biotechnology* clearly reflect these current trends. The plenary speaker was Prof. A. S. Michaels (Chestnut Hill, MA, USA) who placed the state and direction of research in downstream processing in perspective. Achieving reproducible high standards of product purity and quality must remain the goal for those in the business of biochemical process manufacturing. Today's goal is to make high purity therapeutic compounds. In the future, a new challenge will emerge: to extend our capability to make a wider range of materials for 'speciality' as well as 'commodity' markets. The technical challenge is to maintain high product standards while adapting new techniques to largerscale operation; competitiveness will depend increasingly on efficient implementation of process technology.
0167 - 9430/90/$2.00
Robbins device for monitoring biofilm and biocorrosion development), whereas others remain to be more thoroughly tested (e.g. a hydrogenase test for the presence of SRB). The fact that this workshop was held in Germany indicates an increasing awareness in Europe, in recent years, of the importance of biofouling and biocorrosion. Workshops similar in scope have been held during the last five years by Water Micro Associates in the USA. Problems resulting from biofilm formation in flowing water systems are of worldwide occurrence, and the costs resulting from biofouling and biocorrosion are enormous. It will be interesting to note developments over the next few years as more engineers and microbiologists realize the importance to industry of these microbial processes. KEVIN M A R S H A L L
School of Microbiology, The University of N e w South Wales, Kensington, Sydney, Australia.
[]
[]
[]
The conference organization followed closely the sequence of unit operations in biochemical processes.
Cell harvesting Since many biochemical products of commercial interest are intracellular, the initial papers dealt with methods of cell harvesting and cell disruption. The two common approaches to cell harvesting are centrifugation and filtration. P. N. Ward (ICI Biological Products, Billingham, UK) described work clone in collaboration with M. Hoare (University College London, UK) on elucidating the dependence of dewatering performance in a scroll decanter centrifuge on rheological properties of the suspension. An understanding of the chemistry, as well as the physics of the problem is essential for rational optimization of performance. Viscoelasticity and network property characterization were shown to provide a basis for the design of a de-watering operation. The viscoelasticity of borax-flocculated cells was characterized as both shear moduli and loss angle as a
TIBTECH
-
DECEMBER 1990 [Vol. 8]
function of oscillation frequency in a pulse shearometer. With this information, as well as the network modulus measured as a function of cell volume fraction, the authors were able to design operating conditions to de-water borax-flocculated cells to 24% solids in a decanter centrifuge. The significance of this lies in the ability to use laboratory characterization of rheology to design practical operating conditions. An alternative to centrifugation is membrane filtration for cell harvesting. M. Pritchard, with colleagues J. A. Scott and J. A. Howell (University of Bath, UK), presented a paper on concentrating yeast by cross-flow filtration with ultra- and microfiltration membranes. While these devices have been available for many years, membrane filtration is used infrequently for cell harvesting. Part of the problem is the inability to predict performance from first principles. As described by Pritchard, the filtration flux during yeast concentration generally decreases with increasing concentration. However, a surprising result was observed when comparing tubular and flatsheet systems; in the flat-sheet membrane module, when the wall shear stress exceeded a threshold value, the filtration flux began to increase, rather than decrease, with concentration. As a consequence, it was possible to operate the module at its maximum pressure drop before the flux decreased to zero with increasing suspension viscosity. In tubular modules, this shear-induced effect did not occur before there was a transition from turbulent to laminar flow followed by a drop in flux to near zero. These results offer further encouragement to the use of membrane filters for cell harvesting.
Cell disruption Alternative approaches to cell disruption were covered in a series of papers and reviewed by J. A. Asenjo (University of Reading, UK). Although high-pressure homogenization and bead mills are routinely used, it became apparent at this conference that directed protein secretion was becoming the method of choice for separating cells from product. This avoids brute force disruption which could create problems downstream. T. J. Naglak and
339
H: Y. Wang (University of Michigan, USA) described the use of selective chemical permeabilization for recovery of proteins from Pichia pastoris and E. coll. Guanidine-HC1 with Triton X-100 were shown to act synergistically to release ~-lactamase from E. coil Both yeast and bacteria remain intact, thus facilitating separation of cells from broth after protein release. As much as 40-fold purification could be achieved in a protein-release step that often does not provide any purification. While the practitioner usually tries to avoid the addition of chemicals during purification, the results described here may provide sufficient justification to pursue this approach as part of an integrated strategy for efficient recovery.
Product concentration Once a product is separated from the cells, the next step is to reduce the volume by concentration. Several approaches were taken to address this problem. In current practice, filtration and adsorption are most commonly used for this purpose with large molecular-weight molecules, and extraction is used for low molecular-weight materials. C. L. Cooney (Massachusetts Institute of Technology, Cambridge, MA, USA) described the enhanced performance of vortex-flow ultrafiltration for protein and cell concentration. By creating a secondary flow (Taylor vortices) with a high-speed rotating membrane, it is possible to maintain high flux rates even when protein and cell concentrations are high. The important operating and design parameters were identified and high performance and predictable operation will lead to broader use of this technique in bioprocessing. Several approaches to the use of aqueous biphasic extraction of high proteins were described by J. G. Huddlestone (University of Birmingham, UK), and V. Riveros-Moreno (Wellcome Biotech, UK). This technique continues to be interesting and works well for specific products, however, its use is still restricted to only a few materials. Presumably this is a consequence of the difficulty in predicting performance (and thus designing from first principles), and in recovering the used solvent. Both of these aspects of the technology require further study.
High-resolution purification The most challenging problems of downstream processing are in highresolution separations. H. Chase (University of Cambridge, Cambridge, UK) reviewed the use of adsorptive and extractive processes for purification. This area of downstream processing has drawn more creativity and innovation than any other. As pointed out by Chase, one can use specific affinity techniques to adsorb, extract, precipitate or neutralize target molecules. Alternatively, a genetic approach to solving process problems can be taken. Modifying the product gene to incorporate a target in the protein product that can be recognized by a specific affinity ligand places a strategic 'fish hook' on the molecule of interest. However, these procedures are class separations and, alone, do not lead to sufficient purity for therapeutic use. One of the problems that emerges in manufacturing of proteins is the formation of isomeric or modified forms of the desired molecule. Affinity techniques pick out molecules with a specific epitope or recognition site b u t may not distinguish between variations in other regions of the molecule. In this, as in other conferences on separations, too little attention is given to the problem of purifying the desired product from other species. This difficult problem requires further investigation. The problem of structural variation for 7-interferon was nicely described by E. Curling (University of Kent, UK). The variability of posttranslational modification of 7-interferon produced by recombinant CHO cells occurring in response to process conditions suggests that care is needed in design and operation of the bioreactor. Obviously, upstream operation can affect downstream product purification, and the entire process needs to be examined as an integrated unit and not as a sequence of individual steps.
Process integration The problems of process integration and design were considered in a series of papers and were reviewed by K. Mosbach (University of Lund, Sweden) and by H. Y. Wang (University of Michigan, USA). Some dealt with the integration of multiple tasks into a single-unit operation
340
TIBTECH - D E C E M B E R 1990 [Vol. 8]
while others examined methodologies to design fully integrated processes. The consensus opinion was that integration of processes is the direction to take, though there were diverse views as to h o w best to achieve this goal. Combining multiple functions into a single operation has the benefit of reducing the number of steps and increasing the overall yield, but does not necessarily lead to process robustness. When two steps are used in sequence, for example, if the first is variable in performance, then the second can be designed to remove impurities or contaminants that flow through the first. If a single-unit operation is used then this possibility is lost. Application of computer-aided design techniques works well for cost minimalization but can not predict trace impurities or contaminants that determine product quality because analytical descriptions are, as yet, too imprecise. []
[]
[]
Product finishing An important area of downstream processing is product finishing. This topic is neglected in most conferences but was nicely addressed on this occasion. C. R. Hill (Celltech Ltd, UK) reviewed the problems associated with product finishing and pointed out that impurities and contaminants can be introduced into the process from raw materials, hostcell synthesis, and in-process reactive conditions. Attention to detail in operation and design, development of sensitive analytical methods, and establishing rigorous quality controls are essential to the success of any manufacturing process for therapeutic materials. These points were amplified by the other speakers in an entire session devoted to this important topic.
1990, Elsevier Science Publishers Ltd (UK)
C H A R L E S L. C O O N E Y
The search for purity This conference provided an excellent opportunity for researchers []
[]
[]
[]
[]
[]
Altered gene expression in plants due to trans interactions between homologous genes Two recent publications 1,2 on the genetic engineering of flower color have reported the intriguing observation that ectopic transgenes can participate in a homology-based interaction with endogenous plant genes, resulting in suppression of the expression of both the ectopic gene and the endogenous gene. Better understanding of this phenomenon may eventually shed some light on the dynamic behavior of genes in the nucleus. Furthermore, the phenomenon is likely to have significant practical applications in the manipulation of plant phenotypes for use in agriculture and the plant sciences.
and practitioners to come together and discuss work on separations for biotechnology. While advancements were cited on all fronts, it is disappointing that little attention was given by the individual researchers to the problems of removing lowlevel impurities and contaminants such as viruses, nucleic acids and isomeric or modified forms of proteins. The need to reach levels of 99.999% purity, or better, is the most difficult challenge today in protein purification for therapeutic use; for some reason researchers seem unwilling or unable to accept this challenge. Perhaps we can look forward to advances i n this area, too, in the next of this series of conferences on Separations for Biotechnology.
Co-suppression of anthocyanin biosynthetic enzymes in petunia In many plants, flower color is determined by the production of anthocyanin pigments. In an attempt to overproduce flower pigments in petunias, a chimeric gene encoding chalcone synthase (CHS), a key enzyme in the anthocyanin pigment biosynthetic pathway (Fig. 1), was introduced; the results, however, were surprising in several ways. First, the introduction of a CHS transgene to ectopic sites in the petunia genome was found to suppress the expression of both alleles of the endogenous CHS gene in trans, causing the production of pure
0167 - 9430/90/$2.00
Massachusetts Institute of Technology, Cambridge, MA 02139, USA. []
[]
[]
white or patterned flowers 1 (see Box 1). Similar results were obtained independently with CHS, and then also with another petunia anthocyanin gene encoding the enzyme dihydroflavonol-4-reductase (DFR) (Fig. 1) 2. It was demonstrated that in both cases only homologous gene expression was suppressed, not that of other genes in the pathway 1.2. However, the really surprising result was that suppression of the endogenous gene was not simply the result of transgene expression. This could be shown because the suppression phenomenon was somatically reversible in both CHS and DFR transformants (i.e. white flowers sometimes reverted to wildtype purple flowers) (see Box 1). In purple revertant flowers the expression levels of both the ectopic transgene and the endogenous homologous gene were found to increase coordinately and substantially (50fold) restoring expression of the endogenous gene to normal levels 1,2. This indicated that both genes must be coordinately suppressed in the white flowers, suggesting the provisional term 'co-suppression' for this phenomenon 1.
