Stability of recombinant protein production by Penicillium chrysogenum in prolonged chemostat culture

Stability of recombinant protein production by Penicillium chrysogenum in prolonged chemostat culture

FEMS Microbiology Letters 133 (1995) 245-2.51 Stability of recombinant protein production by Penicillium chrysogenum in prolonged chemostat culture ...

659KB Sizes 0 Downloads 34 Views

FEMS Microbiology

Letters 133 (1995) 245-2.51

Stability of recombinant protein production by Penicillium chrysogenum in prolonged chemostat culture Julie M. Withers ‘, Marilyn G. Wiebe a, Geoffrey D. Robson a, David Osborne b, Geoffrey Turner b, Anthony P.J. Trinci a3* d School

qf Biological

h Deprrrtment

Sciences, 1.800 Stopford

qf Molecular

Biology

Building,

Unic:ersi~

and Biotechnology,

ofManchester.Manchester,

Unil,ersit~

Received 22 June 1995; revised 6 September

of Shefield.

1995; accepted

Shefield.

I8 September

MI3

YPT. UK

SIO 2TN. UK

1995

Abstract A strain (WKW2) of fi-galactosidase

glucose-limited

Penicillium

chryysogenum

transformed

with heterologous

fungal

acetamidase

(am&)

and bacterial

(lucZ) was grown at a dilution rate of 0.17 h-’ (doubling time of approx. 4.1 h) for 1600 h in a culture. By the end of the experiment the original strain had been almost completely replaced by

spontaneous, morphological mutants, but the acetamidase and P-galactosidase activities of the culture were essentially unaltered. Furthermore, when WKW2 and the non-transformed parental strain (NRRL1951) were grown together in glucoseor NH:-limited chemostat cultures, neither strain had a selective advantage over the other. Thus, heterologous gene expression does not result in NRRL1951 having a selective advantage over WKW2. These results suggest that continuous flow culture systems could be used for efficient (and cost effective) production of recombinant proteins. Keywords:

Colonial

lium chryso~enum;

mutant: Continuous Transformant

flow culture; Culture stability;

1. Introduction When microorganisms are grown under constant conditions in a chemostat, they evolve and become better adapted to their environment [I]. According to the Monod model of microbial growth kinetics [2], a selective advantage may be gained (depending upon dilution rate) by a mutant either from an increase in maximum specific growth rate ( pm,,), or from a decrease in the saturation constant (K,) for the limit-

* Corresponding author. Tel: +44 (161) 275 3893; Fax: +44 (0161) 275 5656: E-mail: [email protected]. Federation of European Microbiological SSDI 0378- IO97(95)00362-2

Societies

Heterologous

protein production;

Morphological

mutant:

Penicil-

ing nutrient. However, other factors conferring a selective advantage have been identified, including mutations conferring increased I’stickiness ” or cell clumping [3,4] which result in the preferential retention of mutant cells within the fermenter vessel. Although a continuous flow fermentation system is the most efficient way of producing microbial biomass or a biomass-associated product [5,6], it demands greater strain stability than that required for either a batch or fed-batch fermentation. Strain instability has been observed in industrial and laboratory continuous cultures of fungi and streptomycetes [791. For example, highly branched, so-called “colonial” mutants replace the more sparsely branched

246

J.M. Withers et al. / FEMS Microbiology

parental strain during prolonged cultivation of Fusarium graminearum [IO] and non-sporulating mutants accumulate in chemostat populations of sporulating strains of Bacillus subtilis [ 1 I] and Streptomyces tendue [ 121. Stable transformants of bacteria and yeast usually contain one or more autonomously replicating episoma1 (i.e. non-integrated) plasmids, derived from a native plasmid, and the successful exploitation of recombinant microorganisms in large scale fermentations requires the hybrid plasmids to be stable and maintained at a sufficient copy number. Loss of recombinant plasmids has been observed in chemostat cultures [13], with the kinetics of segregation of episomal plasmids being dependent on dilution rate, the nature of the growth-limiting substrate and temperature [ 14,121. The rate of accumulation of plasmid-free cells will, of course, be higher if they have a selective advantage over plasmid-containing cells than if no such advantage is present. Differences in /-lmax and K, values have been reported between plasmid-carrying and plasmid-free strains and these differences have been used to explain the decrease in the frequency of plasmid-carrying organisms in chemostat culture [ 15,161. The growth rate advantage of plasmid-free cells over their transformed counterparts can be explained in terms of the maintenance costs of reproducing plasmid DNA, synthesizing RNA from plasmid-encoded genes and translating the messengers produced into protein. However, the significance of this metabolic load in imparting a selective disadvantage to plasmid-containing cells has been disputed by some investigators [ 151 and alternative explanations for the selective disadvantage have been proposed. In transformants of filamentous fungi, stably maintained recombinant plasmids are integrated into host chromosomes, by homologous or non-homologous recombination, in single or multiple copies. Thus, foreign DNA can only be lost by excision of the plasmid, and therefore, fungal transformants should show greater stability than bacteria or yeast transformants. However, with time, deletions, insertions and substitutions will occur in both native and foreign DNA sequences in fungi and the changes conferring maximum advantage may be those which alter the expression of heterologous genes (and hence reduce the energetic cost of carrying the foreign

Letters

133 (1995) 245-251

DNA). The spontaneous appearance of highly branched morphological mutants during prolonged continuous flow culture of Streptomyces sp., has been correlated with large deletions of DNA (up to 40% of the total DNA; Prof. S.G. Oliver, personal communication), and major chromosomal changes have been observed in prolonged chemostat cultures of Succharomyces cereuisiae [ 171. This paper considers the stability of a transformed strain of PenicilZium chrysogenum in chemostat culture. The study was made to assess the potential of continuous flow systems for the production of recombinant proteins.

2. Materials and methods

2.1. Strains P. chrysogenum NRRL195 1 [ 181 is a wild-type strain, and WKW2 was derived from it by cotransformation [19] with plasmids p3SR2 (containing the acetamidase gene, amdS of A. nidulans) and pKW100 (containing the E. coli P-galactosidase coding sequence 1acZ under the oliC promoter of A. niduluns [20]). The copy number and sites of integration of plasmids in the WKW2 genome are unknown. Stock cultures were maintained as conidia at - 80°C in 20% (v/v) glycerol. Inocula were prepared by harvesting conidia (using 10 ml volumes of sterile distilled water) from plates of 5-8 day old cultures grown on agar-solidified medium at 30°C. 2.2. Media Modified Vogel’s medium [21] containing 3 g, 6 g or 10 g glucose 1-l as the carbon source instead of sucrose was used for glucose-limited chemostat cultures, NH:-limited chemostat cultures and batch cultures, respectively, and 1.65, 3.3 and 3.3 g (NH,),SO, 1-i was substituted for NH,NO, as the nitrogen source for NH:-limited cultures, glucoselimited cultures and batch cultures, respectively. The Vogel’s mineral salts solution was prepared at X 100 final concentration and sterilised by membrane (0.2 pm pore size) filtration before addition to sterile glucose solution. Glucose solutions for chemostat cultures were prepared in 10 1 volumes. For some experiments, medium was solidified with agar (15 g 1-l medium; Taiyo powdered agar, Davis Gelatine).

J.M. Withers et al. / FEMS Microbiology

Selective media were used to identify colonies expressing heterologous genes from non-expressing colonies or non-transformed colonies. To detect acetamidase-producing colonies, 2 mM acetamide (final concentration) was used in place of (NH,),SO, as the nitrogen source in the medium. Colonies producing acetamidase grew vigorously, while colonies unable to utilise acetamide grew very poorly and failed to sporulate. To detect @galactosidase producing colonies, modified Vogel’s medium was adjusted to pH 7.2, before adding 40 mg 1-l X-gal (5-bromo-Cchloro-3indolyl P-D-galactopyranosidase), which had been first dissolved in a small volume of dimethylformamide. The underside of a P-galactosidase-producing colony grown on this medium was bright blue in colour as compared to the grey or yellow colour of non-producing colonies. The simultaneous production of both P-galactosidase and acetamidase by colonies was detected by using 2 mM acetamide modified Vogel’s medium containing 40 mg X-gal I - ’ . Colonies producing both enzymes were large and dense, with blue undersides. Colonies expressing neither were sparse and grey. Colonies producing one or the other enzyme had an intermediate appearance. 2.3. Chemostat Chemostat cultures were grown in a Braun Biostat M (2 1) fermenter operated at full working volume [IO]. The fermenter was inoculated with 10 ml of a suspension containing approx. IO6 conidia ml-’ in sterile distilled water. All chemostat cultures were grown at a dilution rate of 0.17 hh’ (doubling time approx. 4.1 h), pH 5.8 -t 0.1 and 30 + 0. 1°C. Samples (IO ml volumes) were taken at approx. 24 h intervals from the overflow and from the culture vessel for biomass measurements, and a further sample (5 ml volume) was taken from the vessel and used for total and viable counts. Morphological mutants can be detected at concentrations above approx. 0.25 to 0.5% of the total population (200 to 500 colonies were screened from fragments harvested in the daily chemostat sample). 2.4. Total counts Culture samples were diluted ( X lo- ’ ) in 0.16% (w/v) Junlon 100 (an anionic polymer which serves

Letters 133 Cl Y951 245-251

241

to disaggregate hyphal fragments and thereby facilitate their counting; Honeywell and Stein Ltd., Wallington, Surrey), and counts (fragments plus spores) were made under X 100 magnification using an improved Neubauer counting chamber. 2.5. Viable counts Samples (containing predominantly mycelial fragments) from glucose-limited chemostat cultures were diluted serially with sterile distilled water to yield a suspension containing approx. 4 X lo2 cfu ml-‘, and 0.1 ml volumes of these suspensions were spread evenly over the surface of 9 cm Petri dishes containing 20 ml agar-solidified media. The plates (10 replicates per culture sample) were incubated for six days at 30°C. 2.6. Measurement

of biomass

Fungal biomass was harvested using dried Whatman No. 1 filter papers, washed with 2 X 20 ml distilled water and dried to a constant weight at 60°C. Although retention of biomass in the vessel occurred from time to time (as a result of biomass accumulation on internal surfaces) the biomass concentrations in the vessel and overflow remained relatively constant throughout the fermentation. 2.7. Mycelial morphology Measurements of hyphal growth unit length [22] were made on 22 h old mycelia grown in submerged batch culture using a MeasureMouse graphics system (Analytical Measuring Systems, Pamisford, Cambridge, UK) and an Amstrad PC 1512 connected to a Nikon microscope [lo]. Final magnifications of the myceha on the computer monitor were X 140 or X 350; only mycelia with five or more hyphal tips were measured and, for each sample, the hyphal growth unit lengths of 16 to 25 mycelia were measured. 2.8. Colony radial growth rate measurements For measurements of colony radial growth rates, 9 cm Petri dishes containing 20 ml agar medium were inoculated centrally with a drop of spore suspension and incubated at 30°C. Measurements of colony

248

J.M. Withers et al. / FEMS Microbiology

diameters were made with rule at a magnification of X 10 using a Shadowmaster (Baty and Co., Burgess Hill, Sussex) at approx. 24 h intervals, for a total of 96 h. For each strain, 3 colonies were measured, two diameters (perpendicular to each other) being measured for each colony. 2.9. Analyses Samples for acetamidase and /3-galactosidase determinations were prepared by filtering culture samples through 1 pm pore diam. filters (Whatman No. 1) and immediately harvesting the supematant for storage at - 20°C. In addition, for assays of intracellular enzyme activity, culture samples (10 ml> were vacuum filtered through muslin (four layers), and the biomass was washed with distilled water (20 ml> and suspended in a similar volume of Tris buffer (pH 7.2). An ultrasonic probe was inserted (W-225R Sonicator cell disruptor, Heat Systems Ultrasonics Inc.) into the suspension and ultrasound applied (4 X 20 s, maximum power); with intervals (1 min) between successive ultrasonics applications for the suspension to cool to room temperature. The disrupted mycelium was filtered (0.2 pm pore diam., Whatman No. 1) and the filtrate frozen at - 20°C. Acetamidase activity was measured by incubating the sample with acetamide and measuring the ammonia liberated by a phenol-hypochlorite reaction [23]. /?-Galactosidase activity was measured by incubating the sample with lactose using a glucose oxidase/peroxidase reagent to measure the glucose liberated [241).

Letters 133 lIYY5)

245-251

3. Results 3. I. Culture morphology When P. chrysogenum WKW2 was grown at 30°C and pH5.8 in glucose-limited culture at a dilution rate of 0.17 hP ’ , only WKW2 colonies were observed in chemostat samples taken from the first 400 h of cultivation, but thereafter the concentration of WKW2 colonies decreased progressively until by the end of the experiment, after 1600 h cultivation, they formed less than 5% of the population (Fig. 1). Fig. 1 shows that WKW2 was supplanted by a pale green mutant, a white mutant and a white colonial mutant; the morphology and colony radial growth rates of the four strains are shown in Table 1. When P. chrysogenum NRRL195 1 was grown under the same conditions as WKW2, morphological mutants began to accumulate in the culture approx. 500 h after the onset of continuous flow and included both white mutant and pale green mutants similar to those observed in the WKW2 chemostat culture (results not shown) and by Christensen et al. [25]. 3.2. Acetamidase

and P-galactosidase

production

NRRL195 1 grew poorly on medium containing acetamide as the sole nitrogen source and, compared to WKW2, did not produce substantial amounts of acetamidase or P-galactosidase activity (Table 2). WKW2 produced cell-bound acetamidase and /?galactosidase.

Table 1 Colony morphology, colony radial growth rate and hyphal growth unit length of f. chrysogenum WKW2 and three spontaneous morphological mutants isolated from a glucose-limited chemostat culture grown at a dilution rate of 0.17 hh’ at 30°C and pH 5.8 on modified Vogel’s medium Strain

Colony morphology 5 days incubation

Parental strain WKW2 Mutant WKWZ/P Mutant WKW2/W

Large colonies producing many green spores Large pale green colonies Large white colonies producing relatively few white spores Small white colonies producing relatively few white spores

Mutant WKWC/WC

after

Colony radial growth rate * (K,, pmh-‘)

Hyphal growth unit length + (G, pm)

74kI” 77+1b 74*1”

111+4” 112+3 110*2

48+2’

91+2b

a B

Colonies of the isolates were grown at 30°C on agar-solidified modified Vogel’s medium containing 10 g glucose I- ’ + Mean k S.E. of 25 replicates. * Mean f S.E. of 6 replicates. Within a column, values with the same. superscript letter are not significantly different (P > 0.5; Scheffi’s multiple range test).

249

J. M. Withers et al. / FEMS Microbiology Letters 133 ( 19951245-251

0

400 Time

(h) since

1200

600 onset

of

continuous

1600 flow

Fig. I. Concentrations (expressed as a % of the total population) of f. ch~.~oRenum WKW2 (0) and spontaneous morphological mutants (as observed on modified Vogel’s medium containing X-gal) grown in a glucose-limited chemostat culture at a dilution rate of 0.17 h- ’ at 30°C and pH 5.8. The bars give the SE.; white mutant (m ); pale green mutant (0); white colonial mutant (0).

3.3. Stability of heterologous

gene expression

No significant loss of heterologous gene expression was observed during prolonged cultivation of P. chrysogenum WKW2 in glucose-limited chemostat cultures. This conclusion was inferred from the appearance of colonies produced on the three selective media, and from measurements made of the activities of the enzymes in culture samples (Table 2). During

Time

ante

onset

of

continuous

Acetamidase activity (pmolNH: g- ’ biomass h- ’ )

2

107+35 122+46’

56

a

a

(h)

Fig. 2. Concentrations (expressed as a % of the total population) of propagules of P. chrysogenum WKW2 displaying various characteristics of genetic transformation ( P-galactosidase and/or acetamidase activity) in a glucose-limited chemostat of a mixed culture of WKW2 (carrying the plasmids pKWlO0 and p3SR2 incorporating the E. co/i /ucZ and the A. niduluns ctmdS genes respectively) and the parental strain NRRLl951 grown at 30°C and at a dilution rate of 0. I7 hh’

the experiment an occasional colony was observed which showed little or no blue coloration on X-gal Vogel’s medium, or which grew poorly on acetamide medium, but the appearance of such colonies was sporadic and rare.

Table 2 Acetamidase and P-galactosidase activities in biomass of P. chryvsogenum WKW2 grown at a dilution rate of 0.17 hchemostat culture at 30°C and pH 5.8 Time (days) after onset of continuous flow

flow

PGalactosidase activity b ( pg glucose mg biomass-

’ in a glucose-limited

’ h- ’ )

78 + 28 s 64 f 30 b

Within a column, values with the same superscript letter are not significantly different (P > 0.05; Scheffe’s multiple range test). a NRRL1951 grew poorly on acetamide as the sole nitrogen source. and very little acetamidase activity (18 + 4 kmol NH: g- ’ biomass h- ’ ) was detected in extracts of the biomass. h NRRLl951 grown on glucose had a low fi-galactosidase activity (21 + 5 Fg glucose mg biomass-’ h- ‘) in extracts of the biomass.

250

J.M. Withers et al. / FEMS Microbiology

3.4. Growth of transformed (WKW2) and non-transformed (NRRL1951) strains of P. chrysogenum in mixed culture in glucoseand NH,+-limited chemostats Mixed cultures of non-transformed and transformed strains of P. chrysogenum were grown under two different selective conditions, viz. glucose- and NH:-limited conditions. Culture samples were taken from the chemostat at intervals and plated onto three types of selective media which enabled transformed and non-transformed colonies to be distinguished. Fig. 2 shows that the proportion of transformed fragments present in the mixed culture remained constant for the duration (280 h) of the glucoselimited chemostat experiment and an identical result was obtained when the experiment was repeated using a NH:-limited chemostat (results not shown).

4. Discussion Dunn-Coleman et al. [26] reported an approximate reduction of 70% in the yield of heterologous chymosin during 191 generations of continuous cultivation of a strain of A. niger; this reduction in yield was correlated with the accumulation of a non-conidiating, low producing mutant in the population. However, no loss of heterologous gene expression ( P-galactosidase or acetamidase) was detected during 56 days (approx. 330 generations) of glucoselimited chemostat culture of P. chrysogenum WKW2 (Table 2), although, the accumulation of morphological mutants in the culture showed that the culture had evolved during this period (Fig. 1). Similar white and pale green mutants have been isolated from prolonged glucose-limited chemostat cultures of P. chrysogenum Novo Nordisk P8 by Christensen et al. [25] who showed that the white mutant had lost the three genes involved in penicillin biosynthesis and that the pale green mutant had extra copies of the last two genes in the pathway. Although the activity of heterologous enzymes in individual morphological variants of WKW2 was not determined here, measurements of enzyme activity in the fermenter population at the start and end of the fermentation were not significantly different (Table 2), suggesting that the morphological mutants which sup-

Letters 133 (1995) 245-251

planted the parental strain were not significantly altered in their ability to produce heterologous protein. In A. niger, the spontaneous loss of heterologous DNA was found to occur at a frequency of 10m4 per generation 1271. If plasmid loss occurred at a broadly similar rate in a fermenter population of approx. lo9 propagules, approximately 104- 1O5 propagules would be expected to lose their plasmids per generation. Assuming that plasmid loss did not confer any selective advantage on plasmid-cured propagules), at this rate of plasmid loss, plasmid-free propagules would represent approx. 3.3% of the population by the end of the experiment (330 generations). That no plasmid-cured mutants of WKW2 were detected indicates that such a mutation is selectively neutral under the conditions employed and this is supported by the data in Fig. 2 (if an advantageous mutant arose which happened to be plasmid-free, it would accumulate rapidly in the population, a phenomenon related to the ‘jackpot’ effect observed in batch activities). The results obtained in this study indicate that at least some transformed strains of filamentous fungi are sufficiently stable to enable continuous flow fermentations to be used for recombinant protein production.

Acknowledgements The authors are grateful to the BBSRC for the research studentship held by J.M. Withers and to Smith Kline Beecham for supporting D. Osborne.

References [l] Novick, A. and Szilard, L. (1950) Description of the chemostat. Science 112, 715-716. [2] Monod, J. (1942) Recherches sur la croissance des cultures bacteriennes. Hermann et Cie, Paris, France. [3] Francis, J.C. and Hansche, P.E. (1972) Directed evolution of metabolic pathways in microbial populations. I. Modification of the acid phosphatase pH optimum in Succcharomyces cerecisiae. Genetics 70, 59-73. [4] Francis, J.C. and Hansche, P.E. (1973) Directed evolution of metabolic pathways in microbial populations. II. A repeatable adaptation in Succharomyces cereuisiae. Genetics 74, 259-265.

J.M. Withers et al. / FEMS Microbiology [51 Pitt, S.J. (1975) Principles of Microbe and Cell Cultivation. Blackwell, Oxford, UK. a twenty-year [61Trinci, A.P.J. (19921 Myco-protein overnight success story. Mycol. Res. 96, l-13. I71 Righelato, R.C. (1976) Selection of strains of Penicillium chtyogenum with reduced penicillin yields in continuous cultures. J. Appl. Chem. Biotechnol. 26, 153-159. [Sl Trinci, A.P.J., Robson G.D., Wiebe, M.G., Cunliffe, B. and Naylor, T.W. (1990) Growth and morphology of Fusarium graminearum and other fungi in batch and continuous culture. In Microbial Growth Dynamics, Eds. R.K. Poole, M.J. Bazin and C.W. Keevil, pp. 17-38, IRL Press, Oxford. [91 Gravius B., Bezmalinovic, T., Hranueli, D. and Cullum, J. t 19931 Genetic instability and strain degeneration in Streptomyces rimosus. Appl. Environ. Microbial. 59, 2220-2228. [lOI Wiebe, M.G. and Trinci, A.P.J. (19911 Dilution rate as a determinant of mycelial morphology in continuous culture. Biotechnol. Bioeng. 38, 75-8 I, [l II Dawes, I.W. and Thomley, J.H.M. (1970) Sporulation in Batiks subtiks. Theoretical and experimental studies in continuous culture systems, J. Gen. Microbial. 62, 49-66. 1121 Roth, M. and Noack, D. (19821 Genetic stability of differentiated functions in Streptomwes hygroscopicus in relation to conditions of continuous culture. J. Gen. Microbial. 128, 107-I 14. iI31 Dwivedi, C.P., Imanaka, T. and Aiba, S. (1982) Instability of plasmid-harbouring strain of Escherichia coli in continuous culture. Biotechnol. Bioeng. 24, 1465-1468. [I41 Roth, M., Noack, D. and Geuter, R. (1985) Maintenance of the recombinant plasmid pIJ2 in chemostat cultures of Streptomyces lividans 66 (pIJ21. J. Basic Microbial. 25, 265-27 I. [I51 Helling, R.B., Kinney, T. and Adams, J. (1981) The maintenance of plasmid-containing organisms in populations of Escherichia co/i. J. Gen. Microbial. 123, 129- 141. [I61 Numan, 2.. W.A. Venables and Wimpenny, J.W.T. (1991) Competition between strains of Escherichia co/i with and without plasmid RP4 during chemostat growth. Can. J. Microbiol. 37, 509-5 12.

Letters 133 (1995) 245-251

251

S., Simlar, J. and Wilke, C.M. [I71 Adams, J., Puskas-Roza, (1992) Adaptation and major chromosomal changes in populations of Sacchuromyces cerecisiae. Curr. Genet. 22, I3- 19. [I81 Raper, K.B., Alexander, D.F., Coghill, R.D. (19441 Natural variation and penicillin production in Penicillium notatum and allied species. J. Bacterial. 48, 639-659. of Pencil[191 Beri, R.I. and Turner, G. (19871 Transformation lium chvsogenum using the Aspergillus nidulans amdS gene as a dominant selective marker. Curr. Genet. 1I, 639-64 I [201 De Lucas, J.R., Gregory, S. and Turner, G. (1994) Analysis of the regulation of the Aspergillus nidulans acuD gene, encoding isocitrate ylyase, by construction of a hybrid promter. MGG 243, 654-659. [211 Vogel, H.J. (19561 A convenient growth medium for Nelcrospora (Medium Nl. Microbial Genet. Bull. 13. 42-44. [221 Trinci, A.P.J. (1974) A study of the kinetics of hyphal extension and branch initiation of fungal mycelia. J. Gen. Microbial. 8 1, 225-236. reaction for [231 Muftic, M.K. (19641 A new phenol-hypochlorite ammonia. Nature 203, 622-623. In: Methods in enzy[241 Dallqvist, A. (19841 P-galactosidase. matic analysis IV, 3rd. ed. ted. H.V. Bergmeyer), Verlag Chemie. Weinheim FDR, pp. 227-230. b51 Christensen, L.H., Henriksen, C.M., Nielsen, J. and Villadsen, J. (19951 Continuous cultivation of P. chtysogenum. Growth on glucose and penicillin production. J. Biotechnol. In Press. I261 Dunn-Coleman, N.S.. Bodie, E., Carter, G.L. and Armstrong, G.L. (19931 Stability of recombinant strains under fermentation conditions. In: Applied Molecular Genetics of Filamentous Fungi. Eds. J.R. Kinghom and G. Turner, pp. 152-174, Blackie and Son Ltd., Scotland. b71 Ward, M., Wilson L.J. and Kodama, K.H. (19931 Use of Aspergillus overproducing mutants, cured for integrated plasmid, to overproduce heterologous protein. Appl. Microbial. Biotechnol. 39, 738-743.