BIOTECHNOLOGICAL PRODUCTS AND PROCESS ENGINEERING

Enhancing itaconic acid production by Aspergillus terreus

Gregor Tevž & Mojca Benčina & Matic Legiša

Received: 12 February 2010 /Revised: 19 April 2010 /Accepted: 19 April 2010 /Published online: 12 May 2010# Springer-Verlag 2010

Abstract Aspergillus terreus is successfully used forindustrial production of itaconic acid. The acid is formedfrom cis-aconitate, an intermediate of the tricarboxylic(TCA) cycle, by catalytic action of cis-aconitate decarbox-ylase. It could be assumed that strong anaplerotic reactionsthat replenish the pool of the TCA cycle intermediateswould enhance the synthesis and excretion rate of itaconicacid. In the phylogenetic close relative Aspergillus niger,upregulated metabolic flux through glycolysis has beendescribed that acted as a strong anaplerotic reaction.Deregulated glycolytic flux was caused by posttranslationalmodification of 6-phosphofructo-1-kinase (PFK1) thatresulted in formation of a highly active, citrate inhibition-resistant shorter form of the enzyme. In order to avoidcomplex posttranslational modification, the native A. nigerpfkA gene has been modified to encode for an active shorterPFK1 fragment. By the insertion of the modified A. nigerpfkA genes into the A. terreus strain, increased specificproductivities of itaconic acid and final yields weredocumented by transformants in respect to the parentalstrain. On the other hand, growth rate of all transformantsremained suppressed which is due to the low initial pHvalue of the medium, one of the prerequisites for theaccumulation of itaconic acid by A. terreus mycelium.

Keywords Aspergillus terreus . Itaconic acid .

6-Phosphofructo-1-kinase . Posttranslational modification .

Primary metabolism . Anaplerosis

Introduction

Filamentous fungi of the genus Aspergillus are often usedfor the production of organic acids. Most abundant isaccumulation of citric acid by Aspergillus niger, whileAspergillus terreus is a well-known producer of itaconicacid (methylenesuccinic acid). Itaconic acid is a high-volume/low-price product, and its market price is about thesame as for citric acid (Okabe et al. 2009). Its primary useis as a copolymer with synthetic resins. The diesters ofitaconic acid are readily prepared; the dimetyl and dibutylesters are used as copolymers and are produced incommercial quantities (Mattey 1992). Recently, hexylita-conic acid has been reported as an inhibitor of p53-HDM2interaction, which could diminish the turnover of p53tumour suppressor in cancers (Tsukamoto et al. 2006).

The overflow of itaconic acid inAspergillus itaconicus wasfirst reported by Kinoshita (1932), while itaconic acid waslater shown to be excreted by A. terreus as well (Calam et al.1939). Sugars such as glucose and sucrose are primarily usedfor commercial production, and it is generally believed that,in aspergilli, glucose is metabolised to pyruvate mainly bythe glycolytic pathway (Flipphi et al. 2009). The mecha-nism of conversion of pyruvate to itaconate is stillunresolved although tricarboxylic acid cycle (TCA) inter-mediates were shown to be involved in the biosynthesisusing 14C- and 13C-labelled substrates (Bonnarme et al.1995). The fermentation conditions for itaconic acid produc-tion by A. terreus are similar to those for citric acidproduction by A. niger. It requires a high initial concentrationof easily metabolisable carbon source, high dissolved oxygentension, a limitation in metal ions and ammonium as anitrogen source (Miall 1978). Therefore, it could beanticipated that A. terreus strains that excrete itaconic acidmight have similar physiology to that of the citric acid

G. Tevž :M. Benčina :M. Legiša (*)National Institute of Chemistry,Hajdrihova 19,1000 Ljubljana, Sloveniae-mail: matic.legisa@ki.si

Appl Microbiol Biotechnol (2010) 87:1657–1664DOI 10.1007/s00253-010-2642-z

accumulating A. niger strains. However, A. terreus strains areknown to possess strong cis-aconitate decarboxylase activity.The cad gene from A. terreus has been recently characterised(Kanamasa et al. 2008).

In A. niger, strong anaplerotic reactions were oftenreported to accelerate acid excretion, amongst whichderegulated flux of metabolites through glycolysis seemedto be the most important (Peksel et al. 2002). Posttransla-tional modification of 6-phosphofructo-1-kinase (PFK1), akey allosteric regulatory enzyme of glycolysis, might beresponsible for that (Mesojednik and Legiša 2005). Detailedanalyses have shown that the native PFK1 protein wasmodified by a two step process in A. niger cells. First, theenzyme was cleaved by a specific protease which initiallyresulted in the formation of an inactive fragment. In thesecond stage, after the phosphorylation of the proteinmolecule, the enzyme regained activity. By measuring thekinetics of the shorter PFK1 fragment isolated from A. nigercells, it has been revealed that the modified enzyme isresistant to citrate and ATP inhibition while specific PFK1stimulators increased the activity of the shorter fragmentmore intensely than the activity of the native enzyme(Mlakar and Legiša 2006). There are strong indications thatspontaneous processing of the native PFK1 into a highlyactive, citrate inhibition-resistant shorter PFK1 fragmentrepresents a vital step in deregulation of A. niger glycolysiswhich results in enhanced metabolic flux that acts as a stronganaplerotic reaction (Legiša and Mattey 2007). In order toavoid complex posttranslational modification, modified pfkAgenes have been prepared (Capuder et al. 2009). Initially, atruncated gene of appropriate length was prepared thatenabled the synthesis of an initially inactive fragment whichmust be phosphorylated in the cells for activation. Addition-ally, specific mutations were introduced to replace theresponsible threonine residue with glutamic acid to avoidthe need for phosphorylation. By the construction of amodified pfkA gene, A. niger cells were forced to synthesisean active shorter PFK fragment after transformation. Signif-icantly increased production of citric acid was recorded bysome transformants in relation to the parental strain (Capuderet al. 2009). In the present paper, we show that itaconic acidproduction can be enhanced in A. terreus after the insertionof modified A. niger pfkA genes.

Materials and methods

Growth of A. terreus

A. terreus strain A 156 from the Culture Collection of theNational Institute of Chemistry, Ljubljana (MZKI) that isidentical to ATCC 20542 strain was used throughout allexperiments. Spores were harvested from 7-day-old agar

slants and suspended in sterile 0.1% Tween solution.Approximately 5.107 spores were used for the inoculationof 100 ml medium and incubated in 500-ml baffledErlenmeyer flasks on a rotary shaker (Kambič, Semič,Slovenia) at 100 rpm and 30°C. For the growth underminimal conditions, the medium with glucose as a solecarbon source has been taken (Mesojednik and Legiša2005) while for more rapid growth of mycelium, a complexmedium was used that consisted of, per litre, 2 g D-glucose,0.5 g yeast extract, 0.2 g casein acid hydrolysate (Amicase,Sigma), 6 g NaNO3, 1.5 g KH2PO4, 0.5 g MgSO4.7H2O,0.5 g KCl and 0.2 ml trace metal solution (Vishniac andSanter 1957) with the pH adjusted to 6.0.

Transformation of A. terreus protoplasts

A. terreus protoplasts were transformed with pRCR-PFK1expression plasmid, carrying the native A. niger pfkA gene,pRCR-pfk10 plasmid with a truncated A. niger (t-pfkA10)gene and pRCR-mt-pfkA10 plasmid carrying a mutatedtruncated (mt-pfkA10) genes that were constructed asreported previously (Capuder et al. 2009). For protoplastformation, the mycelium was obtained by growing A.terreus cells for 16–18 h in a liquid culture on completemedium. Cell wall lysis, release of protoplasts andsubsequent transformation of A. terreus protoplasts wereperformed essentially as described by Bagar et al. (2009)with the lytic enzyme Novozym234 replaced by Caylase-4(Cayla, Toulouse, France). Only dominant selectionmarkers were used for the co-transformation of A. terreuscells: a hygromycin resistance gene on pAN7-1 vector(Ventura and Ramon 1991) and a phleomycin resistancegene on pUT720 vector (Hendrickson et al. 1999). For co-transformation, 1 µg of individual marker vector and 15 µgof the co-transforming plasmid pRCR (carrying specific A.niger genes) were added to 1×108 protoplasts. Theselection factors used were added to the regenerationmedium in the following concentrations: phleomycin150 µg/ml and hygromycin 1 mg/ml. The transformantswere selected and purified by re-plating at low sporedensities on adequate selective medium.

Test of transformants for integrated genes

All transformants were tested for the presence of the insertedgenes, first by a PCR test using appropriate primers and finallyby Southern analysis that was essentially performed asdescribed previously (Capuder et al. 2009).

Test of transformants on a rotary shaker

Initially, all transformants were tested for itaconic acidaccumulation during the growth on a minimal medium in

1658 Appl Microbiol Biotechnol (2010) 87:1657–1664

500-ml Erlenmeyer flasks with buffles on a rotary shaker.For itaconic acid production, two-stage process was used,essentially as described previously (Riscaldati et al. 2000).In the medium A for vegetative seed culture development,100 ml of low-glucose (20 g/l) minimal medium wasinoculated by 107 spores and incubated on a rotary shakerat 100 rpm and 35°C for 24 h. Ten millilitres of culturefrom medium A was transferred into 90 ml of high-glucose(100 g/l) production medium B and incubated for severaldays on a rotary shaker at 100 rpm and 35°C. For eachtransformant, samples from three individual flasks havebeen withdrawn and tested for itaconic acid levels.

Enzyme assays

6-Phosphofructo-1-kinase activity

After 16 h of growth in the minimal medium with glucose,the mycelium was collected by suction filtration andextensively washed with icy cold 100 mM phosphatebuffer, pH=7.8, containing 1 mM dithiothreitol, 1 mMethylenediaminetetraacetic acid (EDTA) and 0.1 mM phe-nylmethylsulfonyl fluoride. Immediately after washing, themycelium was frozen under the liquid nitrogen. Homoge-nate preparation was performed and PFK1 activity wasmeasured as described previously (Mesojednik and Legiša2005).

When transformants carrying truncated t-pfkA10 geneswere tested for PFK1 activity, sodium azide was added tothe medium to a final concentration of 0.1 mM to inducephosphorylation. Fifteen minutes after the azide was added,the mycelium was collected by suction filtration andprocessed as described previously (Capuder et al. 2009).

Growth of A. terreus in a 5-l laboratory bioreactor

For testing A. terreus transformants in respect of itaconicacid production, the mycelium was grown in 5-l stirred tankbioreactor (IS-100, Infors, Bottmingen, Switzerland). Atwo-stage process was used essentially as describedpreviously (Riscaldati et al 2000). The vegetative phasewas conducted in five 500-ml Erlenmeyer flasks withbuffles. Hundred millilitres of vegetative medium in eachflask was inoculated by 107 spores and incubated on arotary shaker at 100 rpm and 35°C. After 24 h of growth,the fermentation broth was transferred into a laboratoryfermenter containing 4.5 l of production medium(Riscaldati et al. 2000). The growth temperature was setat 35°C, the medium was aerated with 5 l of air per minuteand the stirring rate adjusted to 500 rev min−1. Samples ofthe fermentation broth for analyses were collected in 10 mlpolypropylene tubes, sealed, frozen under the liquidnitrogen and stored at −80°C.

Analytical methods

The amount of dry biomass was determined gravimetricallyafter the washed mycelium was dried at 105°C to a constantweight. Total protein concentrations of the samples weredetermined by bicinchoninic acid protein assays (Smith etal 1985) performed with the Sigma kit (Sigma ChemicalCo., St. Louis, MO, USA). Ammonium ions in thefermentation broth were determined by ion exchangechromatography, essentially as reported previously (Šolaret al. 2008).

The amount of itaconic acid excreted was determined byhigh-pressure liquid chromatography using monolithicCIM® QA column (Bia Separation, Ljubljana, Slovenia)and 20 mM phosphate buffer, 100 mM NaCl, pH=8.0 as amobile phase. Itaconic acid was detected photometrically at214 nm using UV-M optical unit and monitor (Pharmacia).

Cyclic AMP levels in the mycelium were determined byHit Hunter EFC kit (Amersham Bioscience) after themycelium was collected by suction filtration, rapidlywashed by icy cold 50 mM Tris-HCl, 1 mM EDTA buffer,pH7.4 and frozen under the liquid nitrogen. The samplingprocedure was terminated in less than 30 s. After beingweighted, the samples were homogenised in a Mikro-Dismembrator (Sartorius, Germany), and cAMP wasextracted with perchloric acid and prepared for evaluationas described previously (Thevelein et al. 1987).

Results

Detecting inserted genes in transformants

After transformation of A. terreus protoplasts, severalcolonies were isolated that were able to grow in thepresence of hygromycin or phleomycin. By PCR analysis,the integration of the native or modified pfkA genes wasconfirmed in transformants. For each of the inserted genes,five individual transformants have been selected for furthertests. Different copy numbers of genes were inserted inthese transformants (Table 1) as estimated from Southernblot analysis.

Screening for the best itaconic acid producer

In the next step, transformants were tested for itaconic acidproduction during the growth on a rotary shaker. Eachtransformant was grown in three individual flasks, andresults are presented as means from triplicate determina-tions (Table 2). Itaconic acid yields were compared withthat of the parental strain. In a trial with transformantscarrying the native pfkA gene, no major differences in theamount of acid excreted or specific productivity could be

Appl Microbiol Biotechnol (2010) 87:1657–1664 1659

observed in relation to the parental strain. The majority oftransformants with the t-pfkA10 gene accumulated moreitaconic acid than the parental strain after 6 days offermentation. However, all tested, genetically modifiedstrains started to accumulate acid later than the parentalstrain with a lag phase lasting for approximately 40 h.Transformant A650, with 15 t-pfkA10 gene copies inserted,has shown the best results with approximately 50% moreacid excreted in comparison to the parental strain A156.With transformants carrying the mt-pfkA10 gene, signifi-cantly higher yields of itaconic acid were detected after1 week. Nearly double the amount of acid was detected inthe medium of the best transformant (A729), carrying 6 mt-pfkA10 gene copies, compared to the parental strain.

For further tests, one transformant was selected fromeach group of modified A. terreus strains. Selected trans-formants are marked with bold letters in the Table 2. First,specific PFK1 activities were measured in the selectedstrains. As shown in Fig. 1, relatively low specific activitieswere detected in the parental strain (A156), while in thetransformant (A571), carrying additional five copies of thenative A. niger pfkA genes, approximately three-fold higherspecific activities were recorded. Even higher activitieswere detected in homogenate of the transformants encodingshorter PFK1 fragments (A650, A729). Characteristically,in both transformants with active shorter PFK1 fragments,the increase in ATP concentration from 0.1 to 1 mM ATPcaused a decrease in PFK1 activities in the measuringsystem while after the addition of 4 μM of fructose-2,6-bisphosphate activities substantially increased. By previous

Table 2 Average productivities and yields of itaconic acid by A.terreus transformants carrying copies of the native or the modified A.niger pfkA gene in comparison to the A. terreus parental strain A156

Strain Productivity (g acid/h) Yield (g/l)

A156 0.079 21.35

A605 0.079 18.06

A606 0.062 17.78

A571 0.129 22.82

A607 0.103 18.97

A608 0.087 18.27

A650 0.282 37.1

A646 0.277 31.5

A649 0.272 21.0

A647 0.213 25.2

A655 0.243 28.0

A729 0.437 45.5

A730 0.301 33.25

A731 0.291 36.4

A732 0.403 39.9

A733 0.291 42.35

The average productivity of itaconic acid of each transformant wasdetermined between 48 and 96 h of growth on a rotary shaker. Thefinal yield of itaconic acid was measured after 168 h of growth. Theresults are means from triplicate determinations. The transformantsmarked with bold letters have been chosen for further testing in astirred tank laboratory bioreactor

Table 1 List of A. terreus transformants carrying different copynumbers of the native or the modified A. niger pfkA gene

A. terreus transformant Gene inserted Estimated gene copies

A605 n-pfkA 8

A606 n-pfkA 9

A571 n-pfkA 5

A607 n-pfkA 3

A608 n-pfkA 5

A650 t-pfkA10 15

A646 t-pfkA10 8

A649 t-pfkA10 7

A647 t-pfkA10 6

A655 t-pfkA10 4

A729 mt-pfkA10 6

A730 mt-pfkA10 10

A731 mt-pfkA10 10

A732 mt-pfkA10 3

A733 mt-pfkA10 3

All strains are deposited at the Culture collection (MZKI) of theNational Institute of Chemistry, Ljubljana. The estimated copy numberof genes inserted in each transformant, as determined by the Southernanalysis, is shown in the right column

0

200

400

600

800

1000

1200

A156 A571 A650 A729

PF

K1

spec

ific

act

ivit

y (m

U/m

g p

rot.

)

0.1 mM ATP 1 mM ATP 4 µM F-2,6-P

Fig. 1 Specific PFK1 activities in cell-free extracts of A. terreus wild-type strain (A156), transformant (A571) carrying copies of the nativeA. niger n-pfkA gene, transformant (A650) with inserted copies of thet-pfkA gene and transformant (A729) carrying copies of the mt-pfkAgene. F-2,6-P fructose-2,6-bisphosphate. All measurements wereconducted in the system with 6 mM of fructose-6-phosphate

1660 Appl Microbiol Biotechnol (2010) 87:1657–1664

studies of kinetic characteristics of the shorter PFK1fragments in A. niger, it has been revealed that they areinhibited by elevated levels of ATP, while fructose-2,6-bisphosphate prevented ATP inhibition and significantlyraised maximal velocity of the enzyme (Mesojednik andLegiša 2005).

In the second stage, all the selected transformants wereanalysed more in detail during the growth in a 5-l laboratory stirred tank fermenter. A two-stage processwas applied, combining the vegetative phase which wasperformed on a rotary shaker on the medium A, with that ina 5-l working volume bioreactor, which was conducted onthe medium B (Riscaldati et al 2000). All transformantswere tested in a bioreactor for at least three times. The mostrepresentative fermentations are shown in Figs. 2, 3, 4, 5and 6.

Itaconic acid production by the parental A. terreus strain(A156)

Initially, the parental strain (A156) was tested underitaconic acid production conditions. The strain showedmoderate itaconic acid overflow rate reaching approximate-ly 13.5 g of acid accumulated per litre of the medium after150 h of fermentation. Itaconic acid started to accumulate inthe medium only after 35 h of growth when ammoniumions have been depleted from the medium. Mycelium drybiomass increased rapidly up to 70 h of fermentation when10 g of dry biomass was recorded per litre. At the end offermentation, the amount of dry biomass reached the valueof 12.5 g/l (Fig. 2).

Itaconic acid production by an A. terreus transformantcarrying copies of the native A. niger pfkA gene (A571)

As shown in Fig. 3, no increased productivity of itaconicacid has been observed with the transformant (A571)carrying the higher number of native pfkA genes (five

copies) compared to the parental strain (A156). Very similaramounts of itaconic acid and dry biomass were detected atthe end of fermentation as with the parental strain (Fig. 1).The values of the two strains varied for less than 15%. Inboth strains, a higher specific growth rate lasted forapproximately 50 h. At the end of trophophase, itaconicacid started to excrete. Finally, about 13 g of acid weredetected in the medium after approximately 150 h ofgrowth.

Itaconic acid production by an A. terreus transformantcarrying copies of the t-pfkA10 gene (A650)

The transformant (A650) carrying 15 copies of truncated t-pfkA10 gene performed a different type of fermentation(Fig. 4). Although the increase in dry biomass showed asimilar pattern as in the previous two strains, itaconic acidstarted to accumulate with higher specific rate at about 32 hof growth. Initially, productivity reached the value of 0.53 gof acid formed per hour, while after 50 h, excretion ratedeclined to the level of 0.2 g of acid per hour. However,

0

5

10

15

20

25

30

35

0 25 50 75 100 125 150 175

Time (hours)

Dry

bio

mas

s (g

/l); N

H4 (

mM

);ita

coni

c ac

id (g

/l)

0

1

2

3

4p

H v

alu

e

Fig. 4 Itaconic acid production by A. terreus transformant (A650)carrying copies of the t-pfkA10 gene encoding a shorter PFK1fragment that is activated by phosphorylation. Symbols as in Fig. 1

0

5

10

15

20

25

30

35

0 25 50 75 100 125 150 175

Time (hours)

Dry

bio

mas

s (g

/l);

NH

4 (m

M);

itac

on

ic a

cid

(g

/l)

0

1

2

3

4

pH

val

ue

Fig. 3 Itaconic acid production by A. terreus transformant (A571)carrying copies of the A. niger n-pfkA gene encoding the wild-typePFK1 enzyme. Symbols as in Fig. 1

0

5

10

15

20

25

30

35

0 25 50 75 100 125 150 175

Time (hours)

Dry

wei

gh

t (g

/l);

NH

4 (m

M);

itaco

nic

acid

(g/l)

0

1

2

3

4

pH

val

ue

Fig. 2 Itaconic acid production by the parental A. terreus strain(A156) during the growth in a 5-l bioreactor. Itaconic acid (triangle),dry biomass (diamond), pH value (square), ammonium ions (circle)

Appl Microbiol Biotechnol (2010) 87:1657–1664 1661

significantly higher yield of acid was detected at the end offermentation with the values of 23 g of acid per litre, whichwas nearly twice as much as with the parental strain.Moreover, the fermentation lasted for 120 h only. It is worthnoting that the t-pfkA10 gene encoded a shorter PFK1fragment that was initially inactive and must have beenphosphorylated to become active. By measuring freeammonium levels in the medium, rapid depletion wasdetected simultaneously with the initiation of acid accumu-lation. At the same time, a distinct, transient peak of cAMPconcentration was recorded in the cells (Fig. 5).

Itaconic acid production by an A. terreus transformantcarrying copies of the mt-pfkA10 gene (A729)

The best results with respect to itaconic acid overflow wereobserved with transformants constitutively encoding activeshorter PFK1 fragments (Fig. 6). With site-specific muta-tions introduced into the t-pfkA10 gene, the need forphosphorylation of a specific threonine residue was eluded,

and an active enzyme was encoded by the modified gene.The transformant (A729) tested for increased itaconic acidproduction was assessed by Southern blot analysis tocontain six copies of inserted mt-pfkA10 gene. The overallyield after 100 h of fermentation reached the value of 31 gof acid per litre. Maximum productivity was similar to thatof the transformants carrying copies of the t-pfkA10 gene,reaching the value of 0.48 g of acid formed per hour;however, such a high production rate persisted more or lessunchanged until the end of fermentation, which terminatedat approximately 100 h.

Growth behaviour of strains

It is important to realise that all strains grew similarly in theproduction medium (Figs. 2, 3, 4 and 6) reaching the valueof about 9±0.6 g of dry biomass per litre at 60 to 70 h,when the trophophase terminated. They continued to growwith reduced growth rates 0.032±0.008 (g dry biomass/l)till the end of fermentation. In this regard, higher specificproductivities were detected in transformants carryingmodified pfkA genes. In the wild-type strain (A156) andin the transformant with inserted native pfkA genes (A571),maximum specific activities reached the value of 0.02 g ofacid per grams dry biomass of mycelium per hour. Thetransformant (A650) with integrated copies of the t-pfkA10genes reached the highest rates of 0.04 g of acid formed pergram of dry biomass per hour, while in the transformant(A729) with the inserted copies of the mt-pfkA10 gene,nearly 0.06 g of acid were excreted per gram of dry weightin 1 h.

Discussion

Up till now, relatively moderate attention was devoted tothe improvements of the primary metabolism with an aim toincrease productivity and yields in microbial biotechnology.Yet, it is the primary metabolism that is responsible for thesynthesis of precursors of both primary and secondarymetabolites. Since the majority of precursors originate fromthe intermediates of the TCA cycle, strong anapleroticreactions seem to be vital. The expression ‘anapleroticsequence’ is a term used in biochemistry to describe a seriesof enzymatic reactions or pathways that replenish the poolof metabolic intermediates of the TCA cycle (Owen et al.2002). Normally, regulatory properties of allostericenzymes sustain the metabolic fluxes under tight control,which is even more profound in eukaryotic organisms. Themost complex control over the glycolytic flux as a part ofprimary metabolism is attributed to PFK1 enzyme (EC2.7.1.11) that catalyzes the phosphorylation of fructose-6-phosphate to fructose-1,6-bisphosphate, using MgATP as a

0

5

10

15

20

25

30

35

0 25 50 75 100 125 150 175

Time (hours)

Dry

bio

mas

s (g

/l);

NH

4 (m

M);

itaco

nic

acid

(g/l)

0

1

2

3

4

pH v

alue

Fig. 6 Itaconic acid production by A. terreus transformant (A729)carrying copies of the mt-pfkA10 gene encoding an active shorterPFK1 fragment. Symbols as in Fig. 1

0

2

4

6

8

10

12

14

16

18

0 10 20 30 40 50

Time (hours)

NH

4 (m

M)

0

1

2

3

4

5

6

7

8

9

cAM

P (n

M)

Fig. 5 Ammonium ion depletion (square) and intracellular cAMPvalues (diamond) during fermentation with an A. terreus transformant(A650) carrying copies of the t-pfkA10 gene

1662 Appl Microbiol Biotechnol (2010) 87:1657–1664

phosphoryl donor. Effectors like AMP, ammonium ions andfructose-2,6-bisphosphate increase, while inhibitors likeATP and citrate decrease substrate binding affinity andconcomitantly determine overall PFK1 activity (Dunaway1983). Eukaryotic PFK1 enzymes developed during evolu-tion by duplication and tandem fusion of two prokaryoticgenes where more allosteric effector binding sites divergedfrom prokaryotic ancestor. Sequencing analyses of eukary-otic PFK1s revealed that the active site residues werelocated in the N-terminal half of the enzyme while citratebinding sites are scattered both on the N-terminal and C-terminal parts of the protein (Poorman et al. 1984).Posttranslational modification of the A. niger PFK1 enzymewhich resulted in a formation of a 49-kDa fragment from a85-kDa native enzyme represented a kind of a reversedevolutionary process where more complex enzyme wasreduced to a smaller, less regulated enzyme. However,proteolytic cleavage of a part of the eukaryotic moleculemight leave the shorter fragment not only resistant to citrateinhibition but also more susceptible to the positive effectsof some specific effectors (Legiša and Mattey 2007). Such amodified PFK1 enzyme causes in fact upregulation of theoverall glycolytic flux in the cells that started to act as astrong anaplerotic reaction.

By inserting a modified A. niger pfkA gene, significantincrease in itaconic acid production by A. terreus trans-formants was recorded, which was not the case with thegenes encoding the native PFK1 enzymes, where produc-tivity similar to that of the parental strain was observed. Infact, the native PFK1 enzymes remained sensitive to feedback inhibition, albeit the increased amount of the enzymein the cells. Moderate effects of an increased level of nativeglycolytic enzymes on primary metabolism were describedalso in other microorganisms. Overexpression of the keyglycolytic enzymes in Escherichia coli did not increase theglucose consumption rate (Emmerling et al. 1999), andsimilar results were obtained with the yeast Saccharomycescerevisiae after overexpression of seven glycolytic enzymes(Hauf et al. 2000). An increased copy number of the nativepfkA gene neither affected primary metabolism in A. niger.No improved citric acid accumulation has been observed bytransformants, while the detailed analysis of intracellularmetabolites revealed a decrease of fructose-2,6-phosphate, apotent activator of eukaryotic PFK1 enzymes under suchconditions (Ruijter et al 1997).

A. terreus transformants with inserted copies of thetruncated t-pfkA10 genes showed substantially enhanceditaconic acid production and yields. Compared to theparental strain, accumulation started much more vigorouslyafter approximately 25 h of fermentation; however, a steadydecline in specific productivity was soon observed. Thetruncated t-pfkA10 gene encoded for an inactive shorterPFK1 fragment that was obviously activated after phos-

phorylation. Covalent modification of the target moleculemight be induced by a transient increase in cyclic AMPconcentration that occurred approximately at the same timeas ammonium depletion from the medium. In all fermen-tation trials, ammonium ions, as a sole nitrogen source,have been taken from the medium very rapidly. Virtually noions have been detected in samples after 35 h offermentation, although growth of the fungus proceededexponentially till about 70 h. Uptake of ammonium fromthe medium corresponded to the drop of the extracellularpH value. In all fermentations, it decreased from an initialvalue of 3.5 to a value below 2.0. A similar situation wasobserved in A. niger citric acid fermentation, whereammonium ions were reported to be converted intoglucosamine that accumulated in the medium (Papagianniet al. 2005). Glucosamine-6-phosphate deaminase withspecific kinetic characteristics was suggested to conductsuch a rapid reaction in which fructose-6-phosphate, anintermediate of glycolytic flux, was used as a carbon donor(Šolar et al. 2008). Rapid consumption of ammonium ionsmight result in a slight intracellular acidification that couldtrigger cAMP synthesis by adenylate cyclase. Due to theshort phylogenetic distance between A. niger and A. terreus(Flipphi et al. 2009), similar events could be expected tooccur in the later fungus; however, for the final confirma-tion, more experiments should be conducted. Since atransient increase in cAMP level induced phosphorylationonly once, inactive shorter PFK1 fragments encoded by thet-pfkA10 genes were activated only at the early stage. Later,in the newly formed cells, the enzyme might have beensynthesised but most probably remained inactive. Thiscould explain the gradual decline in specific productivity ofitaconic acid by a transformant with inserted t-pfkA10 genecopies. On the other hand, transformant A729 carryingcopies of the mt-pfkA10 gene that was able to synthesise anactive shorter PFK1 fragment throughout the entire fer-mentation maintained a high specific rate of acid produc-tion till the exhaustion of glucose from the medium.

All different strains showed very similar specific growthrates, in spite of different types of PFK1 enzymes that theysynthesised. Under the normal growth conditions in acomplete medium, one would expect that overall anabolicreactions in the cells would be enhanced after thederegulation of metabolic flux through the glycolysis.However, the minimal medium used for the production ofitaconic acid contained a high C-to-N ratio and had a highdissolved oxygen tension, a limitation of metal ions andammonium as a nitrogen source (Miall 1978). Besides, alow starting pH of the medium restrained the growth of themycelium. Under such conditions, the balance betweenanaplerotic and cataplerotic reactions could be maintainedonly by an overflow of itaconic acid. Moreover, thepresented data show that itaconic acid accumulation can

Appl Microbiol Biotechnol (2010) 87:1657–1664 1663

be additionally enhanced by deregulating glycolytic fluxthat is achieved by introducing a gene encoding for a highlyactive, citrate inhibition resistant modified PFK1 enzyme.

Acknowledgements This work was supported by grants of the EUCommission (project QLK3-CT-2002-02038) and project No. L4-4323 financed by the Slovenian Research Agency.

References

Bagar T, Altenbach K, Read ND, Benčina M (2009) Live-cell imagingand measurement of intracellular pH in filamentous fungi using agenetically encoded ratiometric probe. Eukaryot Cell 8:703–712

Bonnarme P, Gillet B, Sepulchre AM, Role C, Beloeil JC, Ducrocq C(1995) Itaconate biosynthesis in Aspergillus terreus. J Bacteriol177:3573–3578

Calam CT, Oxford AE, Raistrick H (1939) Studies in the biochemistryof micro-organisms: itaconic acid, a metabolic product of a strainof Aspergillus terreus Thom. Biochem J 33:1488–1495

Capuder M, Šolar T, Benčina M, Legiša M (2009) Highly active,citrate inhibition resistant form of Aspergillus niger 6-phosphofructo-1-kinase encoded by a modified pfkA gene. JBiotechnol 144:51–57

Dunaway GA (1983) A review of animal phosphofructokinaseisozymes with an emphasis on their physiological role. Mol CellBiochem 52:75–91

Emmerling M, Bailey JE, Sauer U (1999) Glucose catabolism ofEscherichia coli strains with increased activity and alteredregulation of key glycolytic enzymes. Metab Eng 1:117–127

Flipphi M, Sun J, Robellet X, Karaffa L, Fekete E, Zeng A, KubicekCP (2009) Biodiversity and evolution of primary carbonmetabolism in Aspergillus nidulans and other Aspergillus spp.Fungal Genet Biol 46(Suppl 1):S19–S44

Hauf J, Zimmermann FK, Müller S (2000) Simultaneous genomicoverexpression of seven glycolytic enzymes in the yeastSaccharomyces cerevisiae. Enzyme Microb Technol 26:688–698

Hendrickson L, Davis CR, Roach C, Nguyen DK, Aldrich T, McAda PC,Reeves CD (1999) Lovastatin biosynthesis in Aspergillus terreus:characterisation of blocked mutants, enzyme activities and amultifunctional polyketide synthase gene. Chem Biol 6(7):429–439

Kanamasa S, Dwiarti L, Okabe M, Park EY (2008) Cloning andfunctional characterization of the cis-aconitic acid decarboxylase(CAD) gene from Aspergillus terreus. Appl Microbiol Biotechnol80:223–229

Kinoshita K (1932) Über die Production von Itaconsäure und Mannitdurch einem neuen Schimmelpilz Aspergillus itaconicus. ActaPhytochim 5:271–287

Legiša M, Mattey M (2007) Changes in primary metabolism leading tocitric acid overflow in Aspergillus niger. Biotechnol Lett 29:181–190

Mattey M (1992) The production of organic acids. Crit RevBiotechnol 12(1–2):87–132

Mesojednik S, Legiša M (2005) Posttranslational modification of 6-phosphofructo-1-kinase in Aspergillus niger. Appl EnvironMicrobiol 71:1425–1432

Miall LM (1978) Itaconic acid. In: Rose AH (ed) Economicmicrobiology, vol. 2. Primary products of metabolism. Academic,New York, pp 47–119

Mlakar T, Legiša M (2006) Citrate inhibition-resistant form of 6-phosphofructo-1-kinase from Aspergillus niger. Appl EnvironMicrobiol 72:4515–4521

Okabe M, Lies D, Kanamasa S, Park EY (2009) Biotechnologicalproduction of itaconic acid and its biosynthesis in Aspergillusterreus. Appl Microbiol Biotechnol 84:597–606

Owen OE, Kalhan SC, Hanson RW (2002) The key role of anaplerosisand cataplerosis for citric acid cycle function. J Biol Chem277:30409–30412

Papagianni M, Wayman F, Mattey M (2005) Fate and role ofammonium ions during fermentation of citric acid by Aspergillusniger. Appl Environ Microbiol 71:7178–7186

Peksel A, Torres NV, Liu J, Juneau G, Kubicek CP (2002) 13C-NMR analysis of glucose metabolism during citric acidproduction by Aspergillus niger. Appl Microbiol Biotechnol58:157–163

Poorman RA, Randolph A, Kemp RG, Heinrikson RL (1984)Evolution of phosphofructokinase-gene duplication and creationof new effector sites. Nature 309:467–469

Riscaldati E, Moresi M, Federici F, Petruccioli M (2000) Effect of pHand stirring rate on itaconate production by Aspergillus terreus. JBiotechnol 83:219–230

Ruijter GJ, Panneman H, Visser J (1997) Overexpression ofphosphofructokinase and pyruvate kinase in citric acid-producing Aspergillus niger. Biochim Biophys Acta 1334:317–326

Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH,Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC(1985) Measurement of protein using bicinchoninic acid. AnalBiochem 150:76–85

Šolar T, Turšič J, Legiša M (2008) The role of glucosamine-6-phosphate deaminase at the early stages of Aspergillus nigergrowth in a high-citric-acid-yielding medium. Appl MicrobiolBiotechnol 78:613–619

Thevelein JM, Beullens M, Honshoven F, Hoebeeck G, Detremerie K,den Hollander JA, Jans AW (1987) Regulation of the cAMP levelin the yeast Saccharomyces cerevisiae: intracellular pH and theeffect of membrane depolarizing compounds. J Gen Microbiol133:2191–2196

Tsukamoto S, Yoshida T, Hosono H, Ohta T, Yokosawa H (2006)Hexylitaconic acid: a new inhibitor of p53-HDM2 interactionisolated from a marine-derived fungus, Arthrinium sp. BioorgMed Chem Lett 16:69–71

Ventura L, Ramon D (1991) Transformation of Aspergillus terreuswith the hygromycin B resistance marker from Escherichia coli.FEMS Microbiol Lett 82:189–194

Vishniac W, Santer M (1957) The thiobacilli. Bacteriol Rev 21:185–213

1664 Appl Microbiol Biotechnol (2010) 87:1657–1664