Journal of General Microbiology (1993), 139, 2801-2810. 280 1 Printed in Great Britain Growth and product formation in chemostat and recycling cultures by Aspergillus niger N402 and a glucoamylase overproducing transformant, provided with multiple copies of the glaA gene JAAPM. SCHRICKX,'* AGNAS. KRAVE,'? JANC. V E R I ~ O E S ,CEES ~ A. M. J. J. VAN ADRIAAN H. STOUTHAMER' and HENKW. VAN VERSEVELD' DEN HONDEL,~ 'Department of Microbiology, Biological Laboratory, Vrije Universiteit, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands 'Department of Molecular Genetics and Gene Technology, TNO Medical Biological Laboratory, PO Box 5815, 2280 HV Rijswijk, The Netherlands (Received 5 March 1993; revised 24 June 1993; accepted 6 July 1993) Continuous and recycling cultures were carried out with Aspergillus niger N402 wild-type and a glucoamylase overproducing transformant to investigate growth and product formation characteristics. In shake flask cultures, the amount of glucoamylase produced by the transformant was about five times more than by the wild-type strain. In contrast with these results, a twofold overproduction was found in glucose-limited continuous cultures, while no overproduction was found under maltodextin-limitation. Two regions of specific growth rates could be distinguished, one at specific growth rates lower (domain I) and one at specific growth rates higher than 0.12 h-' (domain 11). In domain I changes in mycelium morphology and conidia formation were observed. It has been concluded that maintenance requirements are dependent on the specific growth rate over the whole range of measured growth rates. The deviation in linearity in the linear equation of substrate utilization, caused by this phenomenon, should be considered when continuous cultures with filamentous fungi are performed. In recycling cultures, xylose as limiting carbon source repressed glucoamylase production very strongly. Under maltodextrinlimitation a fivefold overproduction was found. After about 150 h, the total amount of glucoamylase produced was still increasing, while total amount of product, measured as carbon, remained constant. After this time no increase in the amount of biomass formed was observed. These results suggest autolysis and cryptic growth taking place in a recycling fermenter and cell death rate equalling growth rate. Introduction Filamentous fungi represent a physiologically diverse group of micro-organisms. They grow slowly, compared with bacteria. Their growth-form can be long, thin, branched threads of mycelium, but also compact mycelial pellets. These morphologically different growth-forms make them rather difficult to handle and introduce problems of heterogeneity. On a microscopic scale, for example, pellets are subject to internal substrate and oxygen limitations (Braun & ~ ~~ *Author for correspondence. Tel. +31 20 5482437; fax +31 20 6429202 ; e-mail verse@bio.vu.nl. t Present address: Biological Faculty, Universitas Kristen Satya Wacana, Jalan Diponegoro 52-60, Salatiga 5071 1, Indonesia. Vecht-Lifshitz, 1991). The filamentous growth-form, however, poses problems, especially in continuous culture fermentations, due to the tendency of the organism to accumulate on the walls and probes of the culture vessel. Furthermore, accumulation in the outflow system acts as a filter, allowing medium to flow through, while the mycelium partially remains (Pirt & Callow, 1960; Righelato & Pirt, 1967). The growth form of the mycelium can also influence the characteristics of product formation (Pirt & Callow, 1959; Whitaker, 1992). The economic and social importance of fungi as sources of food and biologically active metabolites has stimulated considerable interest in their growth and physiology (Berry, 1975; Righelato, 1975). The use of filamentous fungi and bacteria for commercial enzyme production has been developed in 0001-8199 0 1993 SGM Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17 2802 J. M . Schrickx and others this century. The filamentous fungi of the genus Aspergillus are organisms of considerable importance for a variety of biotechnological industries. In addition to their well documented use in starch processing industries (Norman, 1979), several Aspergillus species are used for the production of secondary metabolites and hydrolytic enzymes (Blain, 1975; Bhella & Altesaar, 1988). One of these enzymes is glucoamylase (GLA; a-1,4glucan glucohydrolase, EC 3.2.1 .3), which can be used for the conversion of starch, glycogen and other polyglucose dextrins to D-glucose and malto-oligosaccharides. Syrups containing high levels of glucose and maltose, industrially produced from starch by using glucoamylase, are widely used in the soft-candy, brewing, baking, soft drinks and canning industries (Norman, 1979; Manjunath et al., 1983). There is considerable interest in methods to improve the production of glucoamylase by Aspergillus. One way to achieve this is by increasing the number of copies of the glucoamylase gene. In general, an increase in the number of copies results in an increased production of the enzyme encoded by the multiplied gene. For the application of genetically engineered strains in industrial processes, it is important that the newly introduced gene copies should be expressed under various conditions used in production processes. Detailed studies on the physiology of protein overproduction by transformant fungi under different fermentation conditions have not been performed to date. This paper reports on the growth and production characteristics of wild-type Aspergillus niger and a glucoamylase overproducing A . niger transformant, grown in continuous culture and in the recycling fermenter with 100 % biomass feedback. Methods Strains and growth conditions. Wild-type Aspergillus niger N402 (cspAl derivative of ATCC 9029; Bos et al., 1988) and a glucoamylase overproducing transformant, A . niger N402bAB6- 1O]B1 (Verdoes et al., 1993), supplied by and constructed at TNO Medical Biological Laboratory, Rijswijk, The Netherlands, were used. Transformation was based on the use of the amdS selectionmarker and a cosmid vector, containing four copies of the glucoamylase gene (glaA) (Verdoes et al., 1993). Both organisms were maintained on potato-dextrose agar slants. The medium composition was as described by Anderson & Smith (1971) with slight modifications. It contained, per litre of distilled water: (NH4),S04, 1.98 g; KH,P04, 1 g; MgS0,.7H,O, 0.25 g; CuSO,. 5H,O, 0.234 mg; FeSO,. 7H20, 6.3 mg; ZnS04.7H,0, 1.1 mg; and MnC1,. 4H20, 3.5 mg. The medium was sterilized in the absence of carbon source by autoclaving. The carbon source was sterilized for 30 min at 100 "C and added aseptically. In batch and continuous cultures carbon and energy sources were either 10 mM-glucose or 1-64g maltodextrin 1-' (dextrin 10, Fluka). The composition of dextrin 10 is not known exactly, but it contains about 10% of reducing matter, which means that the mean chain length of this maltodextrin is 10 glucose units. Therefore the mean M , is 163.8. Thus 1-64g 1-' corresponds to 10 mM of C, units. In recycling experiments carbon- and energy-sources were either 20 mM-glucose or 24 mM-xylose. Initial characterization was carried out in 500 ml Erlenmeyer flasks containing 100 ml of medium. They were grown in a New Brunswick Scientific shaker incubator at 250 r.p.m. at 30 "C. The cultures were harvested 48 h after inoculation and the dry weight and glucoamylase activity were determined. Aerobic carbon-limited chemostat experiments were carried out in a fermenter apparatus designed and manufactured by the electronics and mechanics workshops of the Faculty of Biology, Vrije Universiteit, Amsterdam, The Netherlands (Hanegraaf et al., 1991). A 2 litre fermenter vessel with a working volume of 1.5 litre was used. The culture volume was controlled by means of a stainless steel standpipe (overflow) inserted through the baseplate; the exhaust gas was led through this pipe also, to avoid accumulation of the fungus in the overflow system. Agitation was obtained by flat blade, propeller-type impellers operating at about 700 r.p.m. When necessary wall growth was removed daily with a Teflon covered ring bar magnet, which could be moved over the wall with an external horseshoe magnet. Culture pH was controlled at 4-5kO.1 by addition of 1 M-KOH, and the temperature was controlled at 30 "C. Steady states were assumed at the moment the rates of CO, production and 0, consumption remained constant. Samples were taken for measurements after at least six volume changes had taken place. For every steady state a new culture was started by inoculation with spore suspension, thus eliminating the possibility of carry-over of material from previous steady states. Aerobic carbon limited recycling experiments were carried out in a 2 litre fermenter vessel with a working volume of 1 litre. The culture volume was controlled by means of a liquid-level indicator and a recycling unit, both inserted through the top plate of the fermenter. Fig. 1 is a photograph of the internal recycling unit. The Teflon filters used had a pore-size of 0.22 pm (type GV, Millipore). The filters were sealed and placed over a socket of sintered steel. The liquid-level indicator regulated the speed of a peristalticpump that kept the volume constant by withdrawing filtrate from the culture through the internal recycling unit, thus letting the biomass remain behind in the culture vessel. The whole recycling system has been designed and built by the electronics and mechanics workshops of the Faculty of Biology, Vrije Universiteit, Amsterdam, The Netherlands. Fig. 2 is a diagram of the recycling system. Agitation was obtained by flat blade, propeller-type impellers operating at about 700 r.p.m. Culture pH was controlled at 4.5 'I0.1 by addition of 1 M-KOH,and the temperature was controlled at 30 "C. In order to get good filamentous growth, for both chemostat and recycling experiments, the fermenter vessel was supplemented with CaCl,, added aseptically after sterilization of the medium to avoid precipitation (fhal concentration 46-7mg 1-'), inoculated with a spore suspension (lo5 spores ml-', final concentration), and a batch phase started in which the pH was initially kept at 2.0, air was led over with a flow of 10 1 h-' and the agitation speed was kept at 250 r.p.m. to get oxygen limitation. The pH was shifted to a value of 4.5,the airflow to 30 litres h-I, led in through the bottom of the vessel and the agitation speed to 700r.p.m. after batch growth was nearly completed, and growth in continuous or recycling culture started. Analytical procedures. Dry weight was determined by filtration over Millipore prefilters of constant weight (type AP 40), washing twice with equal volumes of water and drying at 105 "C. Oxygen uptake and carbon dioxide production were measured with a mass spectrometer (MM8-80F, VG Gas Analysis Systems). Extracellular products were measured with a Total Organic Carbon Analyzer (type 915A, Beckman). The glucoamylase activity was measured as described by Yamasaki et al. (1977) with slight modifications as follows. A reaction mixture containing 0 2 ml 1 % Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17 Glucoamylase overproduction by Aspergillus niger I II II Nutrient Fig. 1 Culture vessel r----- 2803 I Filtrate Fig. 2 Fig. 1. Photograph of the internal recycling unit. In the middle, the socket of sintered steel is shown. It is covered by a sealed filter with a pore size of 0.22 pm, shown in the lower part of the picture. The upper part of the picture shows the assembled recycling unit, ready to be placed in the fermenter vessel through the top plate. Fig. 2. Schematic diagram of the recycling system. Nutrient is added at a constant speed to the culture vessel via pump A. The volume of the culture fluid in the culture vessel is kept constant by the liquid-level sensor B, which regulates the speed of pump D via level control unit C. Pump D removes the culture fluid through the internal recycling unit E, which is covered by a 0.22 pm Teflon filter, resulting in 100 YObiomass retention in the culture vessel. (w/v) maltose, 0.5 ml 1 M-acetate buffer (PH 5.3) and 0.3 ml enzyme solution (culture supernatant sample), was incubated at 40 "C for 10min. After incubation the reaction was stopped by heating in a boiling water bath. The glucose formed was measured by the UV method (Boehringer Mannheim kit). One unit of glucoamylase is defined as the amount of enzyme which forms 1 pmol glucose min-' under the conditions described. Results Glucoamylase production in shake flask cultures Aspergillus niger N402 wild-type and transformant B 1 (A. niger N402[pAB6-10]B1) were initially characterized in Erlenmeyer shake flasks. From Table 1 it can be seen that transformant B1 (containing 20 copies of the glaA Table 1. Copy numbers of glaA and glucoamylase production of A . niger N402 wild-type and glucoamylase overproducing transformant BI after 48 h incubation in shake flasks with glucose and maltodextrin as sole carbon sources Copy numbers are as determined by Verdoes et al. (1993). Strain gZuA COPY number N402 (wild-type) N402[pAB6- 1O]B1 20 1 Glucoamylase activity [units (g dry wt)-'] on: Glucose Maltodextrin 127.7 579 268.7 1236 gene; see Verdoes et al., 1993) produced 4.6 times more units glucoamylase per gram dry wt than the wild-type strain on both glucose and maltodextrin as sole carbon sources. On maltodextrin both strains produced about twice as much glucoamylase as on glucose. To investigate the dependency of glucoamylase production on the specific growth rate @) of wild-type A . niger and overproducing transformant B1, both strains were grown in continuous culture. In this way it is also possible to determine the maximum growth yield ( Yxsm) and the maintenance coefficient (ms),which are important parameters to investigate the energetic costs of glucoamylase production. When steady states are reached, it is possible to observe differences in culture characteristics and mycelium morphology at different specific growth rates. Growth and glucoamylase production in continuous culture The growth parameters of continuous culture experiments in which A . niger N402 wild-type and transformant Bl were grown on glucose as limiting carbon and energy source are given in Table 2. The maximum growth yields, Yxm and Yx,, and the maintenance coefficients, m, and m were calculated using the linear equation of substrate u8ization (Pirt, 1965) as shown in equation 1 Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17 2804 J. M . Schrickx and others Table 2. Growth parameters of A . niger N402 and transformant Bl grown aerobically in glucose limited continuous culture qs,go, and qco are, respectively, specific rates of glucose and and oxygen consumptionand CO, , and ,Y are, respectively, maximum growth yields production [mhols (g dry wt)-' h-'I. Y on glucose and oxygen (g dry wt mol- ). M,, mo2 and mco2 are, respectively, maintenance utilization of substrate and oxygen and production of CO, [mmols (g dry wt)-' h-'1. The values of maintenance requirements and maximum yield were calculated by means of linear regression consideringall data obtained from dilution rates within the range of 0.025 h-' < D < 0225 h-' ; 95 % confidence intervals are given in parentheses. A. niger N402 (wild-type) Parameter + + (Is (Ico, 40, YX, YXO, ms mco, mO, Carbon balance Energy balance 0.061 10.00 p 0.245+ 19.488 p 0.369 18.287 p 100 (88-14-115.6) 54.68 (46-08-67-25) 0.061 (-0.1364258) 0.245 (- 0.2904779) 0.369 (-0.1304868) 94.4 % 105.4 Yo + + + 0,029 10.385 p 0-371 19.336 p 0.426 18-47p 96.29 (85.94109.5) 54.14 (42.59-74.28) 0.029 (-0*108-O*110) 0.371 (-0*350-1*093) 0.426 (-0*175-1*027) 97.1 Yo 104.9 Yo Table 3. SpeciJic rate of glucoamylase production (qGLA) [units (g dry wt)-' h-l] of A . niger N402 wild-type and transformant B l , grown aerobically in glucose- and maltodextrin-limited continuous culture 200 160 Data were fitted within the range of 0.025 h-* < y < 0.15 h-' in the case of maltodextrin-limitationand within the range of 0-025 h-' < y < 0-225 h-' in the case of glucose-limitation. 120 - A . niger N402[pAB6-10]B1 80 h CI L . . 40 5 5 a Fig. 3. Specific glucoamylase production (qGLA)versus the specific growth rate (p) of glucose- (a) and maltodextrin- (b) limited chemostat cultures of wild type A. niger N402 ( 0 )and transformant A. niger N402[pAB6-10]Bl (0). in which q is the specific rate of substrate consumption, Y is the molar growth yield and m is the maintenance coefficient. As can be seen from Table 2, there was no significant difference between the values of the measured Carbon source A. niger N402 (wild-type) A . niger N402[pAB6-10]B1 Glucose Maltodextrin 11-6+271-6p 6.8 948.21 p 3046+663.75 p 13.94 + 1 106.9p + parameters of the wild-type and transformant B 1. Carbon recovery was somewhat lower than loo%, because the carbon balances were calculated without considering product formation. The energy balance was calculated as described by Metwally et al. (1991), using the balance method described by Roels (1980). Fig. 3(a) shows the specific rate of glucoamylase production, qGLA(units per gram dry wt per h) versus the specific growth rate of wild-type A . niger N402 and transformant B 1, grown in glucose-limited continuous culture. In Table 3 the calculated relation between the specific rate of glucoamylase production and the specific growth rate is given for both strains, grown on glucose. The specific rate of glucoamylase production of transformant B1 was about twice that of the wild-type strain. The growth parameters of continuous culture experiments in which A . niger N402 wild-type and transformant B1 were grown on maltodextrin as limiting carbon and energy source are given in Table 4. The same Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17 Glucoamylase overproduction by Aspergillus niger 2805 Table 4. Growth parameters (on a c 6 base) of A . niger N402 and transformant BI grown aerobically in maltodextrin-limited continuous culture See Table 2 legend for parameter definitions. Parameter 4s A . niger N402[pAB6-1O]Bl 0.061 + 11.14p 0298 19.646p 0.292 19.082p 89.60 (79.47-1 03.12) 52.41 (44-90-62.93) 0.057 (-0*146-0*259) 0.298 (-0.133-0.729) 0.292 ( -0.1554.739) 0.0003+ 11.357p 0.293 18.157 p 0,253 18.39p 88.05 (80.80-96.72) 54.37 (48.07-62.58) 0.0003 (-0*149-0*150) 0.293 (- 0.0134599) 0.253 (-0.10143608) 871 1 Yo 106.7Yo + + 4co* 40, Yxsm Yxom 4 mco* mo2 A . niger N402 (wild-type) Carbon balance Energy balance 85.5 Yo 106% + + Fig. 4. Mycelium of A . niger growing in submerged culture. (a) Mycelium growing at a dilution rate of about 0.05 h-' ; (b) mycelium growing at a dilution rate of about 022 h-'. Bars, 20 pm. calculations have been carried out as with the glucoselimited cultures. For a good comparison between the glucose and maltodextrin data, the specific rate of maltodextrin consumption, the maximum growth yield on maltodextrin and the maintenance requirements for maltodextrin are expressed on a c6 basis. Here again the determined parameters of transformant B1 were not significantly different from those of the wild-type. Fig. 3(b) shows the specific rate of glucoamylase production, qGLA (units per gram dry wt per h) versus the specific growth rate of wild-type A . niger N402 and transformant B 1, grown in maltodextrin-limited continuous culture. In Table 3 the calculated relation between the specific rate of glucoamylase production and the specific growth rate in the first linear part of Fig. 3 (b) is given for both strains, grown on maltodextrin. At specific growth rates higher than 0.15 h-', the specific rate of glucoamylase production by the transformant did not increase with increasing p. Therefore the relation between qGLAand p in the case of maltodextrin-limitation as mentioned in Table 3 only applies to specific growth rates lower than 0.15. There was no apparent difference in specific rate of glucoamylase production between the wild-type strain and the transformant. Moreover, there was no difference in specific rate of glucoamylase production between transformant B 1, grown on glucose and the two strains, grown on maltodextrin in the range p = 0 h-' to 0.15 h-l, to which linear regression was applied. Mycelium morphology at diflerent speciJic growth rates At specific growth rates lower than 0.12 h-' we observed a change in mycelium morphology: the hyphae were more branched and conidiation took place. At specific growth rates higher than 0.12 h-' the hyphae were less branched and no conidiophores were formed. These morphological changes at different specific growth rates were most obvious in chemostat cultures with the wildtype strain. Photographs of culture samples taken at a low growth rate @ about 0.05 h-l) and a high growth rate (p about 0.22 h-') are shown in Fig. 4. Because of these observations we analysed the chemostat experiments in two domains, although, when linear Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17 2806 J. M . Schrickx and others 0.05 0.1 0-15 0.2 0.25 0.05 0.1 0.15 P (h-9 P @-9 Fig. 5 Fig. 6 0.2 0-25 Fig. 5. Specific rate of glucose consumption (qdco$ (a) and oxygen consumption (q0J (b) versus the specific growth rate 0) in glucoselimited continuous culture of wild-type A. niger N402. Linear regression has been carried out in two domains I and 11. The regression lines for domain I1 are shown in the figure. Domain I : Y,,, = 101-55g (mol glucose)-' ; m, = 0.0998 mmols glucose g-' h-' ; Y,,, = 112.3 g (mol oxygen)-' ; mo,= 1.045 mmols oxygen g-' h-l, Domain I1 : Y,, = 78-25g (mol glucose)-' ; rn, = -0.447 mmols glucose g-' h-'; Y,,, = 40.26 g (mol oxygen)-'; mO2= -0.79 mmols oxygen g-' h-'. Fig. 6. Specific rate of glucose consumption (qgl,,$ (a) and oxygen consumption (44)(b) versus the specific growth rate (p)in glucoselimited continuous culture of wild-type A. niger N402. Non-linear regression has been carried out using all data points. Linear regression lines, supposing growth rate independent maintenance requirements to be zero are shown as dotted lines. No maintenance requirements are supposed at ,u = 0.12 h-'. regression considering all data was carried out, good fits for linearity were found (see the confidence limits in Tables 2 and 4). The values of maintenance requirements and maximum yield shown in Tables 2 and 4 were calculated by means of linear regression considering all data points. The results for the glucose-limited continuous culture with wild-type A . niger N402 are shown in Fig. 5. In domain I @ c 0.12 h-l) the Yxsmand rn, values were not significantly different from the values determined considering all data points (compare with Table 2). The Yxom and rnO2 values in domain I were both significantly higher, indicating an increased energy need when compared with a fit with all data or domain I1 @ > 0-12h-l). It is shown that when linear regression in domain I1 is used, both maximum yields and maintenance requirements on glucose (a) and oxygen (b) are lowered, compared to results of linear regression when all data points are considered. These low, even negative, maintenance requirements indicate decreasing costs for maintenance requirements at increasing specific growth rates up to the shift from domain I to 11, and increasing costs again at increasing specific growth rates in domain 11. Considering maintenance requirements to be growth rate dependent, we applied non-linear regression to the data points of Fig. 5. The results are shown Fig. 6. The dotted lines in Fig. 6 are linear regression lines, supposing growth rate independent maintenance requirements to be zero. For clarity no maintenance requirements are supposed at ,u = 0.12 h-', the p value at which a shift from domain I to domain I1 occurs. This presentation clearly shows that maintenance requirements decrease with increasing growth rates up to p = 0-12h-' and again increase with increasing growth rates at higher ,u values. Chemostat experiments as described above are convenient for relatively high values of specific growth rate. In chemostat cultures with Bacillus Zicheniforrnis it Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17 Glucoamy lase overproduction by Asperg illus n iger 800 L 2807 i 600 400 200 5 4 h eO 3 2 3 2 Q.7 1 50 100 150 Time (h) 200 250 50 100 150 Time (h) Fig. 7. Glucoamylase production (a, b) and product formation (Pt) (c, d) versus time in maltodextrin- (a, c) and xylose- (byd)limited recycling cultures of wild-type A . niger N402 ( 0 )and transformant A. niger N402[pAB6-10]Bl (0). was found that exocellular protease production was negatively correlated with the specific growth rate (Frankena et al., 1986). To investigate glucoamylase production at low growth rates by A . niger wild-type and overproducing transformant B1, both strains were grown in the recycling fermenter with 100YObiomass retention. This fermentation technique resembles industrial processes, which are usually carried out in fed-batch type fermentations. Growth and glucoamylase production in the recycling fermenter A . niger N402 wild-type and transformant A . niger N402[pAB6-10]Bl were grown in xylose- and maltodextrin-limited recycling fermenters with 100% biomass feedback (van Verseveld et al., 1986; Frankena et al., 1988) with a rate of substrate addition corresponding to 1 mmol of C, units h-'. Production of biomass, total product (measured as total organic carbon) and glucoamylase were determined. Fig. 7 ( a ,b) shows the total production of glucoamylase by both strains versus time for the maltodextrin(a) and xylose- (b) limited recycling cultures. It is clear from Fig. 7(a,b) that much more glucoamylase (about 30 times) was produced on maltodextrin as limiting carbon source than on xylose. This was expected, since maltodextrin highly induces the expression of the glaA gene, while xylose strongly represses it (Fowler et al., 1990). Overproduction of glucoamylase by transformant Bl could be clearly observed when maltodextrin was used as sole carbon source. During the whole experiment transformant Bl produced about 4.5 times more glucoamylase than the wildtype strain, as was found in shake flask cultures. The total amount of product, measured as total organic carbon, produced by both wild-type and transformant, was about the same during the course of recycling cultivation as shown in Fig. 7(c, d). Glucoamylase production on xylose was negligible. The model used for the description of growth and product formation in the recycling fermenter has been treated extensively by van Verseveld et al. (1986), Stouthamer & van Verseveld (1987) and Ykema et al. (1989). It is based on carbon- and energy-balances (Roels, 1980), the linear equation for substrate utilization, rs = m,x, (Pirt, 1965)and a linear equation for product formation, rp = ax,+br,. Equation 2 gives the amount of biomass formed with time +&+& in which rs is the rate of substrate consumption (mols h-l), m,* = m,+% is the maintenance coefficient Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17 J. M . Schrickx and others 2808 7 0 0 50 100 150 Time (h) 200 250 Fig. 8. Total amount of biomass formed versus time in maltodextrinlimited recycling cultures of wild-type A . niger N402 ( 0 ) and transformant A . niger N402[pAB6-10]Bl (0). -h 0.02 n -e Discussion 0.015 2M 0.01 W La 0.005 20 40 60 80 100 120 n d 0.04 G 2 0.03 t I 0 bD La biomass was practically zero after about 100 h in the wild-type culture and after about 150 h in the culture with the transformant strain. Fig. 9 shows the total amount of product (measured as organic carbon) and the rate of production (in carbon units per h) versus time in maltodextrin-limited recycling cultures of wild-type A . niger N402 and transformant A . niger N402[pAB6-10]Bl. At the time the amount of mycelial biomass in the fermenter vessel stopped increasing, the carbon production rate strongly decreased and consequently the total amount of organic carbon units produced almost remained stable. This phenomenon was most obvious in the culture with transformant B1, because this culture lasted long enough to reach this stage in the recycling process. The last data point from the recycling culture with the wild-type strain (pointed out with an arrow in Fig. 9a) suggests that this phenomenon was about to happen in this culture as well. 0.02 0.01 50 100 150 Time (h) 200 250 Fig. 9. Product formation (Pt) ( 0 )and rate of product formation (rp) ( 0 )versus time in maltodextrin-limitedrecycling cultures of wild-type A . niger N402 (a) and transformant A . niger N402j-pAB6-101Bl (b). [mols substrate (g biomass)-' h-I], not corrected for product formation and Ymm = [&+&]-' is the maximum yield of biomass on substrate (gmol-'), not corrected for product formation. The full derivation of this equation is given by van Verseveld et al. (1986, 1991). According to equation 2 the increase in the amount of biomass, i.e., the growth rate, will decrease with time. Fig. 8 shows that this happened in the maltodextrin-limited recycling cultures. The increase in In continuous culture we did not observe overproduction of glucoamylase by the transformant strain when maltodextrin was the growth-limiting substrate. Under glucose-limitation there was obvious glucoamylase overproduction as can be seen from Fig. 3 and from Table 3. These results are in agreement with Metwally et al. (199 l), who found overproduction under glucose-limitation and not under maltose-limitation. As expected, the specific rate of glucoamylase production was higher under maltodextrin-limitationthan under glucose-limitation, because glucose either induces the gZaA gene at a lower level than maltodextrin or acts as a carbon catabolite repressor (Fowler et al., 1990). The difference in production rates, when growing on the two growthlimiting substrates in the case of the transformant, was nevertheless rather small. There may be a maximum limit for glucoamylase production in the chemostat by the fungus due to a shortage of specific regulatory proteins in the presence of a large number of gZaA gene copies, as suggested by Verdoes et al. (1993). This also explains why transformant B1, grown in shake flask culture, only overproduces glucoamylase 4.6 times, while provided with 20 copies of the gZaA gene. Brown & Zainudeen (1977) and Metwally et al. (199 1) found two distinct domains of specific growth rates for the specific rate of substrate and oxygen consumption in continuous cultures of Trichoderma viride and A . niger respectively. Our data show a comparable deviation of the linear equation of substrate utilization, showing two domains of linearity at specific growth rates lower than 0-12h-' and higher than 0.12 h-'. The shifting point 0.12 h-' is taken somewhat arbitrarily, but it fits best in the graphs for both qglucose and qo, data. The shift takes Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17 Glucoamylase overproduction by Aspergillus niger place somewhere between p = 0-11 h-' and p = 0.14 h-I, because at specific growth rates of 0.1 1 h-I and lower we observed changes in the mycelium morphology and differentiation (conidia formation) as shown in Fig. 4. The deviation mentioned above was probably caused by these morphological changes, demanding higher maintenance requirements at specific growth rates of 0.1 1 h-' and lower. Pitt & Bull (1982) observed the same phenomenon in cultures of the filamentous fungus Trichoderma aureoviride. Since the respiratory quotient (RQ) equalled 1 at all specific growth rates, this deviation cannot be explained by changes in the pathways of glucose catabolism, as found by Carter & Bull (1969) in chemostat cultures with Aspergillus nidulans. When non-linear regression is applied to these data, as shown in Fig. 5, it is obvious that maintenance requirements are dependent upon the specific growth rate. At a specific growth rate of 0.12 h-l, the shift from domain I to domain 11, maintenance requirements could be virtually zero or at the most very low, as has been found before for other filamentous fungi (Pitt & Bull, 1982). At increasing specific growth rates, when ,u > 0.12 h-', maintenance requirements will again increase with increasing specific growth rates. This may be caused by the increasing protein and RNA contents of the fungal biomass, resulting in higher energy demands for protein and RNA turnover at increasing specific growth rates, as found before in chemostat experiments with the yeast Saccharomyces cerevisiae (Furukawa et al., 1983; Verduyn et al., 1990) and the filamentous fungus T. aureoviride (Pitt & Bull, 1982). These phenomena should always be considered when the linear equation of substrate utilization is applied to cultures of filamentous fungi. It is difficult to reach a steady state situation when growing a filamentous fungus in continuous culture, mainly because of wall growth and (partial) blocking of the outlet pipe. For this reason one should always take into account that non-homogeneity can influence the culture results. This can lead to pseudo steady states, partial recycling of the culture and inaccurate dry weight determinations. In our experiments we confirmed that a steady state was reached at the moment the rates of CO, production and 0, consumption remained constant. We found no significant effect of glucoamylase production on yield parameters, as shown in Tables 2 and 4, because glucoamylase overproduction was low in glucose-limited continuous culture and absent in maltodextrin-limited continuous culture. In the recycling fermenter the overproduction of glucoamylase by the transformant strain was obvious and equal to the results found in shake flask cultures, when maltodextrin was the limiting carbon source. In xylose-limited recycling culture the glucoamylase pro- 2809 duction was almost completely repressed as can be seen in Fig. 7. For this reason it can be concluded that the introduced extra copies of the glaA gene in the overproducing transformant were subject to the same regulatory mechanisms as the parental glaA gene. In the yeast S. cerevisiae secretion of exocellular proteins happens at the bud where vesicles, transporting these proteins from the Golgi apparatus, fuse with the plasma membrane to secrete their contents (Novick et al., 1981). In filamentous fungi the hyphal tip is the chief site of secretion of exocellular enzymes into the medium (Chang & Trevithick, 1974). The apical vesicles, produced by the Golgi apparatus and accumulating in the hyphal tips of a growing fungus, have two essential functions in hyphal tip extension: they carry and secrete enzymes and/or materials for the growth of the plasmalemma and the cell wall and they extend the plasmalemma surface (Girbardt, 1969; Bartnicki-Garcia, 1973). We assume therefore that without growth of the fungus, there will be a strongly decreased secretion of exocellular enzymes. As shown in Figs 7 and 8, the rate of production of glucoamylase was still increasing when the amount of biomass in the culture vessel stayed constant. On the other hand, the total product formation, measured as organic carbon, produced by the transformant strain did not increase after about 150 h. Consequently, from this point of time the total production rate (rp) decreased rapidly to almost zero (Fig. 9(b). Most likely this phenomenon also happened in the wild-type culture, indicated by a decreasing rp after 100 h. As opposed to the glucoamylase production data, no overproduction of total products was found. After about 150 h, dry weight did not increase, while glucoamylase was still produced, suggesting hyphal tip extension was still taking place. Because of the above observations we assume that after about 150 h, death of the mycelium took place at the same rate as growth at the hyphal tips, resulting in a constant dry weight. Mycelial death could be caused by autolysis due to starvation, and by mechanical damage to the mycelium by the impeller. The contents of the broken mycelial threads would be released into the medium and act as an extra carbon source for the growing parts of mycelium. To confirm these conclusions further investigations (e.g. at a microscopic level) are required. It has been reported that during glucose starvation of the mycelial fungus Penicillium chrysogenum, autolysis and cryptic growth on cell contents took place (Trinci & Righelato, 1970). Release of contents of hyphae probably took place before 150 h as well. The increase in the amount of total products, measured as carbon, produced in the first 150 h suggests that in this case there was no carbon-limitation, but only energy-limitation.After this time growth became carbon- Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17 2810 J . M . Schrickx and others and energy-limited again, when an equilibrium was established between carbon production and consumption. References ANDERSON, J. G. & SMITH,J. E. (1971). Synchronous initiation and maturation of Aspergillus niger conidiophores in culture. Transactions of the British Mycological Society 56, 9-29. BARTNICKI-GARCIA, S. (1973). Fundamental aspects of hyphal morphogenesis. Symposia of the Society of General Microbiology 23, 245-267. BERRY, D. R. (1975). The environmental control of the physiology of filamentous fungi. In The Filamentous Fungi, vol. 1, Industrial Mycology, pp. 16-32. Edited by J. E. Smith & D. R. Berry. London: Edward Arnold. I. (1988). Role of CAMPin the mediation BHELLA, R. S. & ALTESAAR, of glucose catabolite repression of glucoamylase synthesis in Aspergillus awamori. Current Genetics 14, 247-252. BLAIN,J. A. (1975). Industrial enzyme production. In The Filamentous Fungi, vol. 1, Industrial Mycology, pp. 193-211. Edited by J. E. Smith & D. R. Berry. London: Edward Arnold. A. J. M., SWART,K., HUYBERS, A., KOBUS,G. & Bos, C. J., DEBETS, SLAKHORST, S. M. (1988). Genetic analysis and the construction of master strains for assignment of genes to six linkage groups in Aspergillus niger. Current Genetics 14, 437443. BRAUN, S. & VECHT-LIFSHITZ, S. E. (1991). Mycelial morphology and metabolite production. Trends in Biotechnology 9, 63-68. BROWN,D.E. & ZAINUDEEN, M.A. (1977). Growth kinetics and cellulase biosynthesis in the continuous culture of Trichoderma viriak. Biotechnology and Bioengineering 19, 941-958. CARTER,B. L. A. & BULL,A. T. (1969). Studies of fungal growth and intermediary carbon metabolism under steady and non-steady state conditions. Biotechnology and Bioengineering 11,785-804. CHANG,P. L. Y. & TREVITHICK, J. R. (1974). How important is secretion of exoenzymes through apical cell walls of fungi? Archives of Microbiology 101,281-293. FOWLER, T., BERKA, R. M. &WARD,M. (1990). Regulation of theglaA gene of Aspergillus niger. Current Genetics 18,537-545. FRANKENA, J., KONINGSTEIN, G. M., VAN VERSEVELD, H. W. & STOUTHAMER, A. H. (1986). Effect of different limitations in chemostat cultures on growth and production of exocellular protease by Bacillus licheniformis. Applied Microbiology and Biotechnology 24, 106-1 12. FRANKENA, J., VAN VERSEVELD, H. W. & STOUTHAMER, A. H. (1988). Substrate and energy costs of the production of exocellular enzymes by Bacillus licheniformis. Biotechnology and Bioengineering 32, 803-812. FURUKAWA, K.,HEINZLE, E. & DUNN,I. J. (1983). Influence of oxygen on the growth of Saccharomyces cerevisiae in continuous culture. Biotechnology and Bioengineering 25, 2293-23 17. GIRBARDT,M. (1969). Die Ultrastruktw der Apikalregion von Pilzhyphen. Protoplasma 67, 413-441. HANEGRAAF, P. P. F., PUNT, P. J., VAN DEN HONDEL,C. A. M. J. J., DEKKER, J., YAP, W.,VAN VERSEVELD, H. W.& STOUTHAMER, A. H. (1991). Construction and physiological characterization of glyceraldehyde-3-phosphatedehydrogenase overproducing transformants of Aspergillus nidulans. Applied Microbiology and Biotechnology 34, 765-77 1. MANJUNATH, P., SHENOY, B. C.& RAGHAWNDRA, RAO,M. R. (1983). Fungal glucoamylases. Journal of Applied Biochemistry 5, 235-260. METWALLY, M., EL SAYED,M., OSMAN,H., HANEGRAAF, P. P. F., STOUTHAMER, A. H. & VAN VERSEVELD, H. W. (1991). Bioenergetic consequences of glucoamylaseproduction in carbon-limited chemostat cultures of Aspergillus niger. Antonie van Leeuwenhoek 59,3543. NORMAN, B. E. (1979). The application of polysaccharide degrading enzymes in the starch industry. In Microbial Polysaccharides and Polysaccharases, pp. 339-376. Edited by R. C. W. Berkely, G. W. Gooday & D. C. Ellwood. New York: Academic Press. R. (1981). Order of events in the NOVICK, P., FERRO,S. & SCHEKMAN, yeast secretory pathway. Cell 25,46 1469. PIRT, S. J. (1965). The maintenance energy of bacteria in growing cultures. Proceedings of the Royal Society 163,224-23 1. PIRT, S. J. & CALLOW, D. S. (1959). Continuous flow culture of the filamentous mould Penicillium chrysogenum and the control of its morphology. Nature, London 184,307-3 10. PIRT,S. J. & CALLOW, D. S. (1960). Studieson the growth of Penicillium chrysogenum in continuous-flow culture with reference to penicillin production. Journal of Applied Bacteriology 23, 87-98. PITT,D.E. & BULL,A. T. (1982). Influence of culture conditions on the physiology and composition of Trichoderma aureoviride. Journal of General Microbiology 128, 15 17-1 527. RIGHELATO, R. C. (1975). Growth kinetics of mycelial fungi. In The Filamentous Fungi, vol. 1, Industrial Mycology, pp. 79-103. Edited by J. E. Smith & D. R. Berry. London: Edward Arnold. RIGHELATO, R. C. & PIRT,S. J. (1967). Improved control of organism concentration in continuous cultures of filamentous micro-organisms. Journal of Applied Bacteriology 30, 246-250. ROELS, J. A. (1980). Application of macroscopic principles to microbial metabolism. Biotechnology and Bioengineering 22,2457-25 14. STOUTHAMER, A. H. & VAN VERSEVELD, H. W. (1987). Microbial energetics should be considered in manipulating metabolism for biotechnological purposes. Trends in Biotechnology 5, 149-155. TRINCI,A. P. J. & RIGHELATO, R. C. (1970). Changes in constituents and ultrastructure of hyphal compartments during autolysis of glucose-starved Penicillium chrysogenum. Journal of General Microbiology 60,239-249. VAN VERSEVELD, H. w ., DE HOLLANDER, J. A., FRANKENA, J., BRASTER, M., LEEUWERIK, F. J. & STOUTHAMER, A. H. (1986). Modelling of microbial substrate conversion, growth and product formation in a recycling fermentor. Antonie van Leeuwenhoek 52, 325-342. VAN VERSEVELD, H. W., METWALLY, M., EL SAYED,M., OSMAN,M., SCHRICKX, J. M. & STOUTHAMER, A. H. (1991). Determination of the maximum product yield from glucoamylase-producingAspergillus niger grown in the recycling fermentor. Antonie van Leeuwenhoek 60, 313-323. VERDOES, J. C., PUNT,P. J., SCHRICM,J. M., VAN VERSEVELD, H. W., STOUTHAMER, A. H. & VAN DEN HONDEL,C. A. M. J. J. (1993). Glucoamylase overexpressionin Aspergillus niger : molecular genetic analysis of strains containing multiple copies of the glaA gene. Transgenic Research 2, 84-92. VERDUYN, C.,POSTMA, E., SCHEFFERS, W. A. & VAN DIJKEN,J. P. (1990). Energetics of Saccharomyces cerevisiae in anaerobic glucoselimited chemostat cultures. Journal of General Microbiology 136, 405-412. WHITAKER, A. (1 992). Actinomycetes in submerged culture. Applied Biochemistry and Biotechnology 32, 23-35. YAMASAKI,Y., SUZUKI,Y. & OZAWA,J. (1977). Purification and properties of two forms of glucoamylase from Penicillium oxalicum. Agricultural and Biological Chemistry 41,755-762. YKEMA, A., BAKELS, R. H. A,, VERWOERT, I. I. G. S., SMIT,H. & VAN VERSEVELD, H. W. (1989). Growth yield, maintenance requirements, and lipid formation in the oleaginous yeast Apiotrichum curvatum. Biotechnology and Bioengineering 34, 1268-1276. Downloaded from www.microbiologyresearch.org by IP: 54.160.99.180 On: Sun, 18 Sep 2016 01:29:17