Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/223733740

Biotechnologicalvalorisationofrawglyceroldischargedafterbio-diesel(fattyacidmethylesters)manufacturingprocess...

ArticleinBiomassandBioenergy·January2008

DOI:10.1016/j.biombioe.2007.06.007

CITATIONS

254

READS

559

8authors,including:

Someoftheauthorsofthispublicationarealsoworkingontheserelatedprojects:

bioactivecompoundsViewproject

CharacterizationofphosphatidicacidphosphataseactivityintheoleaginousyeastYarrowia

lipolyticaViewproject

SeraphimPapanikolaou

AgriculturalUniversityofAthens

128PUBLICATIONS4,894CITATIONS

SEEPROFILE

IsabelleChevalot

FrenchNationalCentreforScientificResearch

60PUBLICATIONS1,508CITATIONS

SEEPROFILE

IvanMarc

FrenchNationalCentreforScientificResearch

78PUBLICATIONS1,673CITATIONS

SEEPROFILE

GeorgeAggelis

UniversityofPatras

121PUBLICATIONS4,989CITATIONS

SEEPROFILE

AllcontentfollowingthispagewasuploadedbyGeorgeAggelison21January2015.

Theuserhasrequestedenhancementofthedownloadedfile.Allin-textreferencesunderlinedinblueareaddedtotheoriginaldocumentandarelinkedtopublicationsonResearchGate,lettingyouaccessandreadthemimmediately.

This article was published in an Elsevier journal. The attached copyis furnished to the author for non-commercial research and

education use, including for instruction at the author’s institution,sharing with colleagues and providing to institution administration.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Available at www.sciencedirect.com

http://www.elsevier.com/locate/biombioe

Biotechnological valorisation of raw glycerol dischargedafter bio-diesel (fatty acid methyl esters) manufacturingprocess: Production of 1,3-propanediol, citric acid andsingle cell oil

Seraphim Papanikolaoua,b,�, Stylianos Fakasb,c, Michel Ficka, Isabelle Chevalota,Maria Galiotou-Panayotoub, Michael Komaitisb, Ivan Marca, George Aggelisc

aLaboratoire des Sciences du Genie Chimique, CNRS, ENSIC/ENSAIA, UPR 6811, 2 Avenue de la Foret de Haye,

54505 Vandœuvre-les-Nancy, FrancebDepartment of Food Science & Technology, Agricultural University of Athens, 75 Iera Odos, 11855 Athens, GreececDivision of Genetics, Cell and Development Biology, Department of Biology, University of Patras, 26504 Patras, Greece

a r t i c l e i n f o

Article history:

Received 18 December 2006

Received in revised form

12 June 2007

Accepted 25 June 2007

Available online 21 August 2007

Keywords:

Clostridium butyricum

Yarrowia lipolytica

Mortierella isabellina

Bio-diesel

Citric acid

Fatty acid methyl esters

Single-cell oil

1,3-Propanediol

Raw glycerol

a b s t r a c t

Raw glycerol, byproduct from bio-diesel production process, is used as carbon substrate in

several biotechnological applications. Using Clostridium butyricum F2b, 47.1 g L�1 of

1,3-propanediol was produced in batch anaerobic cultures while substrate uptake rate

(rS, expressed in g L�1 h�1) increased with increase in glycerol concentration in the medium.

In continuous cultures, microbial behaviour was studied in transitory states after addition

of 1,3-propanediol in the chemostat vessel. Microbial growth was not affected by the high

1,3-propanediol (which was added in the chemostat vessel) concentration, while butyric

and acetic acids concentrations were increased. In a two-stage continuous culture,

43.5 g L�1 of 1,3-propanediol was produced with a total volumetric productivity of

1.33 g L�1 h�1.

Yarrowia lipolytica ACA-DC 50109 was grown in nitrogen-limited aerobic cultures on raw

glycerol and it exhibited remarkable biomass production even at high glycerol concentra-

tion media, while rS decreased with increase in glycerol concentration. Citric acid was

produced after nitrogen depletion in the medium, with the highest quantity of 62.5 g L�1,

and yield on glycerol consumed was 0.56 g g�1. Fatty acid analysis of total cellular lipids

showed that glycerol concentration increase in the growth medium somehow increased

the cellular unsaturated fatty acids content of lipids.

Mortierella isabellina ATHUM 2935 exhibited satisfactory growth in nitrogen-limited

aerobic cultures with raw glycerol used as sole substrate. When high initial glycerol

quantities were employed (e.g. 100 g L�1), 4.4 g L�1 of lipid were accumulated corresponding

to around 51% (wt/wt) of lipid in dry weight. rS constantly decreased with increase in

glycerol concentration in the medium, and in all cases notable glycerol quantities remained

unconsumed in the medium.

& 2007 Elsevier Ltd. All rights reserved.

ARTICLE IN PRESS

0961-9534/$ - see front matter & 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.biombioe.2007.06.007

�Corresponding author. Laboratory of Food Microbiology & Biotechnology, Agricultural University of Athens, Athens, Greece.Tel./fax: +30 210 5294700.

E-mail address: spapanik@aua.gr (S. Papanikolaou).

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 1

Author's personal copy

1. Introduction

Bio-diesel fuels defined as fatty acid methyl esters derived

from various renewable lipid resources (e.g. rapeseed oil,

soybean oil, palm oil, etc.) constitute an alternative type of

fuel for various types of diesel engines and heating systems

[1]. With the obligatory application of bio-fuels in a large

commercial scale in Europe, enormous quantities of glycerol

deposit in the market are likely to be generated in the near

future [1], while glycerol-containing residue is also produced

in significant amounts from fat saponification and alcoholic

beverage fabrication units [2,3]. Conversion, thus, of glycerol

to various high-value added products attracts much interest.

The main technique of biotechnological valorisation of

glycerol is related to its biotransformation into 1,3-propane-

diol, a substance of notable importance for the textile and

chemical industry [4,5]. This conversion is carried out by a

number of prokaryotic cells belonging to the family Enter-

obacteriaceae and to the genus Clostridium [4,6]. Glycerol has

also been alternatively utilised as sole substrate by various

types of microorganisms for production of other metabolites

such as ethanol and hydrogen [7] or biomass and a-amylase

[8]. Organic acids have also been produced by strains using

glycerol [5,9], but citric acid in spite of being one of the most

important metabolites produced via biotechnological meth-

ods in industrial scale [10], has been rarely produced using

glycerol as starting material [11]. Microbial lipids (also called

single-cell oils, SCO) present a potential industrial interest

due to capacity of various microorganisms to store lipids with

unusual composition or structure [12,13]. In spite of the

significant number of studies dealing with the production of

SCO by various microorganisms grown on a plethora of

carbon sources and culture configurations, data dealing with

growth of oleaginous microorganisms on (raw) glycerol are

quite limited [14,15]. Through utilisation of raw glycerol

discharged from bio-diesel units for the production of lipid,

we could recycle the by-product, decreasing significantly the

production cost of the whole process.

Although utilisation of raw glycerol in the fermentation

medium without prior purification offers a remarkable

advantage against the traditional use of pure glycerol as

substrate, only few reports have appeared in the literature on

the use of this substrate as sole carbon source [1–3,11,16–18].

ARTICLE IN PRESS

Fig. 1 – Pathways involved in glycerol breakdown by Clostridium butyricum, Yarrowia lipolytica and Mortierella isabellina.

TAGs: triacylglycerols; 3-HPA: 3-hydroxypropionaldehyde; (a)–(c) systems transporting pyruvic acid from cytosol to

mitochondrion and inversely; (d) system transporting citric and malic acid from cytosol to mitochondrion and inversely; ACL:

ATP-citrate lyase; FAS: fatty acid synthetase; ICDH: iso-citrate dehydrogenase; MDc: malate dehydrogenase (cytoplasmic);

MDm: malate dehydrogenase (mitochondrial); ME: NADPH+-malic enzyme; PD: pyruvate dehydrogenase; CS: citrate synthase;

ICL: iso-citrate lyase; GK: glycerol kinase; GDHt: glycerol dehydratase; GDH: glycerol dehydrogenase; 3-P-GDH: 3-P-glycerol

dehydrogenase; DHAk: di-hydroxyacetone kinase; PDOR: 1,3-propanediol oxidoreductase; FD: ferredoxine oxido-reductase.

Pathways described in [4,13].

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 1 61

Author's personal copy

Recent investigations showed that promising results can be

achieved with Clostridium butyricum F2b, which produced

considerable quantities of 1,3-propanediol on raw glycerol in

continuous mode [19]. Furthermore, Yarrowia lipolytica ACA-

DC 50109 [LGAM S(7)1] presented accumulation of citric acid

in the medium when grown on raw glycerol in nitrogen-

limited flask cultures [20]. Additionally, the mould Mortierella

isabellina ATHUM 2935 produced huge quantities of lipid when

cultivated on high-glucose nitrogen-limited media [21]. The

aim of the present study is to elucidate further the possibi-

lities of raw glycerol valorisation through different metabolic

pathways using the above-mentioned microorganisms as

model systems. These microorganisms were chosen as in

the case of C. butyricum, glycerol undergoes assimilation that

involves different pathways compared with the one of the

eukaryotic cells (DHA regulon against traditional phosphor-

ylation of glycerol to 3-P-glycerol), while the production of

SCO or citric acid is a non-growth-coupled process occurring

after nitrogen exhaustion from the medium in contrast to

biosynthesis of 1,3-propanediol [4,10,13]. Pathways involving

glycerol breakdown and biosynthesis of metabolites by

C. butyricum, Y. lipolytica and M. isabellina are shown in Fig. 1.

Bio-kinetic considerations are considered and discussed.

2. Materials and methods

2.1. Microorganisms, growth media and cultivationconditions

C. butyricum F2b, Y. lipolytica ACA-DC 50109 and M. isabellina

ATHUM 2935 were used. Strains were maintained as pre-

viously described [19–21]. C. butyricum preculturing and

bioreactor culturing were performed in the medium described

by Homann et al. [22]. Batch and single-stage continuous

cultures were carried out in a 2-L bioreactor (Biolafitte) as

described in Papanikolaou et al. [19]. Anaerobic conditions

were maintained by self-generated anaerobiosis. Two-stage

culture was performed with the aid of two 2-L bioreactors

(Biolafitte, France), one filled with 0.5 of medium and the

other with 1.3 L (agitation 200 rpm, incubation temperature

T ¼ 33 1C, pH ¼ 7.070.1, and flow rate is 5575 mL h�1). In

continuous cultures, steady state was obtained after contin-

uous flow of at least 5 working volumes of the medium

through the vessel.

Salt composition of the medium used for Y. lipolytica

cultures is described in Papanikolaou et al. [20], while that

for M. isabellina cultures is as in Fakas et al. [23]. Ammonium

sulphate and yeast extract, 0.5 g L�1 each, were used as

nitrogen sources. Experiments were performed in 250-mL

conical flasks, containing 5071 mL of growth medium

inoculated with 1 mL of exponential preculture (in Y. lipolytica

1 mL contains 1–3�108 cells, in M. isabellina 1 mL contains

1–3�105 spores; in both cases initial biomass inoculums

concentration was 0.2070.05 g L�1) as described in [20] in

T ¼ 28 1C. In yeast cultures, in order to prevent excessive pH

drop, highly buffered media were employed (utilisation of

KH2PO4 and Na2HPO4 at concentrations 12 g L�1 of each) while

additionally a volume (500–600mL) of KOH at 5 M was

periodically and aseptically added in order to maintain a

medium pH value greater than 5.0. All yeast or mould cultures

were carried out at initial pH ¼ 6.170.2. The carbon source

used was raw glycerol issued from bio-diesel (fatty acid

methyl esters) production unit [Agro-chemical industry

EPILOR-R&D, Compiegne, France—glycerol content 65%

(wt/wt)]. Impurities in the industrial glycerol solution were

potassium and sodium salts [4–5% (wt/wt)], methanol

[3% (wt/wt)], heavy metals and lignin [1% (wt/wt)] other

organic materials [0.5% (wt/wt)] and water [26% (wt/wt)]. For

all strains, control experiments were done using pure glycerol

[95% (wt/wt)].

2.2. Quantitative determinations and chemical analyses

Biomass, glycerol and products concentration were estimated

as previously described [19,20]. Carbon recovery (Rc) was

calculated in C. butyricum cultures as indicated elsewhere

[4,24]. Total cellular lipid was extracted with a mixture of

chloroform and methanol 2:1 (v/v), converted to methyl esters

and analysed by GLC as described earlier in [23]. Methods for

determining ammonium concentration, pH, dissolved oxygen

(DO) concentration and specific rate of oxygen uptake

(qO2Fmg mg�1 h�1) in the shaker cultures were as previously

described in [21].

2.3. Notation

Ac: acetic acid (g L�1), CAS 64-19-7; But: butyric acid (g L�1),

CAS 107-92-6; Cit: citric acid (g L�1), CAS 77-92-7; GLA: g-linolenic acid, CAS 506-26-3; L: microbial total lipid (g L�1); PD:

1,3-propanediol (g L�1), CAS 504-63-2; S: glycerol (g L�1), CAS

56-81-5; X: biomass (g L�1); YX/S: global biomass yield (g of

biomass per g of glycerol consumed); Yp/S: product yield (g of

product—p per g of glycerol consumed, where p was Ac, But,

PD, Cit, L); Pp: volumetric productivity (g of product—p per L

per h); Subscripts 0, r and max indicate the initial, remaining

and maximum quantity, respectively, of the components, in

the kinetics performed.

3. Results

3.1. Cultures of C. butyricum F2b

Anaerobic batch cultures of C. butyricum were carried out with

raw glycerol utilised as the sole substrate, at initial concen-

trations 39 and 90 g L�1, and satisfactory biomass production

was observed (Xmax ¼ 1.2–2.6 g L�1). mmax evaluated at the

early exponential phase was 0.3670.02 h�1 with lower values

obtained, at high initial glycerol concentration culture

(S0 ¼ 90 g L�1). Substrate uptake rate (rS), expressed as rS ¼

�DS=Dt and estimated during exponential growth phase,

showed some increase when initial substrate quantity in

the growth medium increased (at S0 ¼ 39 g L�1 rS was

2.3 g L�1 h�1 while at S0 ¼ 90 g L�1 rS was 3.4 g L�1 h�1), while

at the end of cultures insignificant S quantities remained

unconsumed. The fermentation kinetics at S0 ¼ 90 g L�1 is

shown in Fig. 2. YX/S presented a slightly lower value for

the cultivation of S0 ¼ 90 g L�1 compared with S0 ¼ 39 g L�1

(0.026 against 0.030 g g�1). PD was the principal metabolic

ARTICLE IN PRESS

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 162

Author's personal copy

product with a maximum concentration of 47.1 g L�1. Yields

YPD/S, YAc/S and YBut/S presented almost equivalent values

regardless of S0 concentration (values 0.53, 0.028 and

0.09 g g�1, respectively).

In order to investigate the effect of PD addition upon cell

growth and secretion of metabolites, a continuous culture

was carried out (dilution rate D imposed 0.0470.005 h�1,

S0 ¼ 80 g L�1), and when steady state was attained (at steady

state PD ¼ 41.2, X ¼ 1.5, Sr ¼ 5.5, Ac ¼ 1.0 and But ¼ 9.9 g L�1), a

43 g pulse of PD was carried out in the fermentation medium.

The evolution of biomass, 1,3-propanediol, substrate and

organic acids, followed until 25 h after the injection and

illustrated in Figs. 3(a and b), shows that despite the

significant quantity of PD found in the chemostat vessel

(84.2 g L�1), biomass concentration remained almost constant

(X ¼ 1.570.1 g L�1), obviously without inhibition exerted by

1,3-propanediol. Likewise, increment in PD in the vessel

resulted in a small decrease in glycerol consumption, and

acetic acid and butyric acid concentrations increased. When

the excess of PD was washed out from the bioreactor, S, Ac

and But concentrations tended to reach the values of steady-

state conditions (Fig. 3b). Furthermore, it appears that when

PD concentration increased, YX/S showed the tendency to

decrease despite constancy of biomass concentration in the

transitory stage, while when PD excess was discharged from

ARTICLE IN PRESS

0

2

4

6

8

0

0

X(g L-1)

X, A

c, B

ut (g

L-1

)

12

10

Time (h)

5040302010S

, P

D (

g L

-1)

100

80

60

40

20

S (g L-1)

PD (g L-1)But (g L-1)

Ac (g L-1)

Fig. 2 – Kinetics of biomass (X, g L�1), 1,3-propanediol (PD,

g L�1), acetic acid (Ac, g L�1) and butyric acid (But, g L�1)

production and glycerol (S, g L�1) consumption of Clostridium

butyricum during growth in batch bioreactor experiment.

Culture conditions: S0 ¼ 90.2 g L�1; pH ¼ 7.070.1; incubation

temperature T ¼ 33 1C. Kinetics were conducted in duplicate

by using different inocula.

0

1

2

3

0

0 5

0

5

0

1

2

0

X p

roduced (

g L

-1)

2.5

1.5

0.5

Time after transition (h)

3025201510-5

0 5

Time after transition (h)

3025201510-5

0 5

Time after transition (h)

3025201510-5

100

PD

(g L

-1)

80

60

40

20

Pulse of 1,3-propanediol

PD (g L-1) X (g L-1)

S r

em

ain

ing, B

ut pro

duced (

g L

-1) 20

15

10

Ac p

roduced (

g L

-1)1.5

0.5

Ac (g L-1)But (g L-1)

S (g L-1)

YX

/S, Y

Ac/S

(g g

-1)

0.05

0.04

0.03

0.02

0.01

0

YB

ut/S (

g g

-1)

0.2

0.15

0.1

0.05

YBut/S (g g-1) YX/S (g g-1)

YAc/S (g g-1)

Fig. 3 – Addition of 1,3-propanediol at steady-state

conditions in single-stage anaerobic continuous culture of

raw glycerol by Clostridium butyricum. Evolution of

1,3-propanediol and biomass concentrations (a), glycerol,

acetic acid and butyric acid concentrations (b) and biomass,

acetic acid and butyric acid yields on glycerol consumed (c).

Culture conditions: S0 ¼ 80 g L�1; pH ¼ 7.070.1;

D ¼ 0.0470.005 h�1; incubation temperature T ¼ 33 1C; pulse

of 43 g of PD at t ¼ 0.

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 1 63

Author's personal copy

the chemostat vessel, YX/S values tended to be the ones of the

steady state. Likewise, yields YAc/S and YBut/S clearly increased

with PD quantity increase, decreasing afterwards when PD

concentration gradually decreased (Fig. 3c). It should be noted

that similar biochemical behaviour has been observed when

lower 1,3-propanediol concentrations were added at steady-

state conditions into the chemostat vessel (added PD quan-

tities of 10, 20 or 39 g—data not presented).

A two-stage culture was also carried out (S0 ¼ 90 g L�1, flow

rate at 5575 mL h�1). The first stage of the culture (active

volume 0.5 L) presented a higher dilution rate (D ¼ 0.11 h�1),

are hence somewhat increased 1,3-propanediol volumetric

productivity could have been achieved. The second stage

(active volume 1.3 L), presenting a lower dilution rate

(D ¼ 0.04 h�1), was mainly used to further increase the

product concentration. It was also desirable to study the

biochemical behaviour of the microorganism when continu-

ously significant quantities of metabolites (mainly PD) were

injected into the second-stage vessel. The obtained result for

both stages is shown in Table 1. Indeed, for the first stage of

the culture satisfactory PD quantity and productivity were

obtained (PD ¼ 32.5 g L�1, PPD ¼ 3.58 g L�1 h�1), whereas

around 28% (wt/wt) of the substrate remained. In the second

stage, almost all of the remaining glycerol was assimilated

and a significant final 1,3-propanediol concentration

(43.5 g L�1) was achieved, but X quantity was significantly

lower as compared with the first one. Furthermore, in the

second stage of the culture a remarkably lower carbon

recovery was observed and a lower YPD/S was obtained.

Moreover, acetic acid was produced in higher amounts and

butyric acid in lower ones compared with the first stage.

3.2. Cultures of Y. lipolytica ACA-DC 50109

Aerobic nitrogen-limited flask cultures of Y. lipolytica were

carried out with raw glycerol utilised as sole substrate.

Literature suggests that enhancement of Cit production by

yeast strains occurs only in nitrogen-limited media in which

pH is maintained greater than 4.5 (see for instance Rymowicz

et al. [11]), while the strain under investigation produced low

quantities of organic acids in media that did not present a

satisfactory buffer capacity and medium pH dropped rapidly

to levels below 3.5 [20]. Therefore, in order to quantify the

effect of glycerol concentration upon growth and citric acid

production by the microorganism, cultures in a controlled

mode concerning pH evolution in the medium were done

(use of highly buffered media and daily correction of the

medium pH so that pH value was maintained within 5.0 and

6.2—see Section 2). Nitrogen concentration in the medium

remained constant [initial (NH4)2SO4 and yeast extract at

0.5 g L�1] while that of initial glycerol varied (S0 ranged

between 20 and 164 g L�1). The obtained result of cultures is

summarised in Table 2. In all cases and despite differentia-

tions in S0 value (which in some cases presented notably high

values, e.g., 90 and 164 g L�1), no lag phase was observed,

while mmax, as evaluated at the early exponential growth

phase, obtained the value of 0.1870.03 h�1, with the lower

values observed for the high S0 concentration cultures

presumably due to inhibition exerted by the high substrate

concentration. Up to 50710 h after inoculation, the micro-

organism consumed almost all of the quantity of the available

extracellular nitrogen of the growth medium, and at that

period, glycerol consumption led mainly to the creation of

biomass, since insignificant quantities of organic acids

(e.g. citric acid up to 0.5–1.0 g L�1) were produced. Then, when

nitrogen became limiting, X presented a slight further

increase before reaching the kinetics plateau, which was in

the same magnitude regardless of S0 quantity (Table 2). It is

noted that S0 concentration increment decreased rS, while at

high initial substrate amount media, significant glycerol

quantities remained unconsumed at the culture medium

even after long incubation periods (when S0 ¼ 164 g L�1, Sr

was 52.5 g L�1 at t ¼ 600 h—see Table 2). From mmax and rS

values, it may be assumed that at high initial glycerol

concentration media (specifically at S0 ¼ 164 g L�1) slight

inhibition of the microbial growth was observed, although

Xmax concentration was unaffected by the significant S0

concentration in the medium (concentration within the range

of 7.0–7.9 g L�1 in all trials—see Table 2). However, remarkable

citric acid quantities were accumulated in the growth

medium mainly at high S0 concentration media (when

S0 ¼ 164 g L�1, Citmax ¼ 62.5 g L�1), while yield YCit/S signifi-

cantly increased with glycerol concentration increase in the

culture medium (highest YCit/S value 0.56 g g�1) and maximum

ARTICLE IN PRESS

Table 1 – Growth parameters, conversion yields and carbon recoveries in the first and second stage of the culture at steadystate, during an anaerobic two-stage continuous fermentation of raw glycerol by Clostridium butyricum F2b

D(h�1)

X(g L�1)

S(g L�1)

PD(g L�1)

Ac(g L�1)

But(g L�1)

Rc (% wt/wt)

YPD/S

(g g�1)YBut/S

(g g�1)YAc/S

(g g�1)PPD

(g L�1 h�1)

First stage of the culture

0.11 2.2 25.4 32.5 2.0 7.8 95 0.50 0.12 0.031 3.58

Second stage of the culture

0.04 1.4 0.5 43.5 3.4 9.6 76 0.44 0.07 0.056 1.74

Global fermentation

1.4 0.5 43.5 3.4 9.6 88 0.49 0.11 0.038 1.33

Culture conditions: S0 ¼ 90 g L�1; flow rate 5575 mL h�1; pH ¼ 7.070.1; incubation temperature T ¼ 33 1C.

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 164

Author's personal copy

PCit value obtained was 0.13 g L�1 h�1 (see Table 2). The

fermentation kinetics of one characteristic culture of Y.

lipolytica on raw glycerol (e.g. when S0 ¼ 164 g L�1) is shown

in Fig. 4.

In all the trials and despite high S0 concentration employed,

Y. lipolytica grew in full aerobic conditions, since the DO level

corresponded to oxygen saturation higher than 30% (v/v) for

all growth steps. In the first growth step (0–50 h), qO2

presented high values (0.3570.05 mg mg�1 h�1) significantly

decreasing afterwards (0.0670.02 mg mg�1 h�1). The impact of

initial glycerol concentration upon the composition of fatty

acids of total cellular lipids of Y. lipolytica was also studied.

Total lipid was extracted and in all growth steps its quantity

did not exceed the 6–14% (wt/wt) of dry matter. Slight

modifications in the fatty acid composition of the cellular

lipid were observed with time or glycerol concentration

increase in the growth medium. However, cultivation on

increased S0 concentration media was accompanied by

slightly higher quantity of total cellular unsaturated fatty

acid (D9C16:1, D9C18:1, and D9,12C18:2) quantity synthesised

when compared with that on lower S0 cultures (Table 3).

3.3. Cultures of M. isabellina ATHUM 2935

Kinetic studies were done in nitrogen-limited media at

various initial glycerol (S0 ¼ 50, 100 and 135 g L�1) and

constant nitrogen [initial (NH4)2SO4 and yeast extract at

0.5 g L�1] concentration media. The obtained result of cultures

is summarised in Table 4. Up to 4575 h after inoculation, the

fungus consumed all of the available ammonium nitrogen,

while low organic acid amounts were produced throughout

culture, since pH value of the growth medium presented a

little drop in all growth steps (final pH ¼ 5.370.1). The

principal organic acid produced was citric acid and its

maximum concentration achieved was around 2.070.5 g L�1.

M. isabellina presented satisfactory cell growth in all initial

glycerol concentration media (Table 4). The highest biomass

concentration achieved (8.5 g L�1) was obtained at

S0 ¼ 100.5 g L�1, while, at higher initial glycerol concentration

media (e.g. S0 ¼ 134 g L�1), Xmax quantity decreased. mmax

values, as evaluated at the early exponential growth phase,

were around 0.07 h�1 with slightly lower values obtained at

high S0 concentration media. Moreover, in spite of the

remarkably long incubation period employed, significant

quantities of substrate remained unconsumed at the med-

ium, principally when high initial S0 concentrations were

used; at S0 ¼ 134.5 g L�1, around 75% (wt/wt) of glycerol

remained untouched while rS (in g L�1 h�1) significantly

ARTICLE IN PRESS

Table 2 – Quantitative data of Yarrowia lipolytica originated from kinetics in media with various initial glycerol (S0)concentrations

Fermentation time(h)

S0

(g L�1)Sr

(g L�1)rS

a

(g L�1 h�1)mmax

b

(h�1)Xmax

(g L�1)Citmax

(g L�1)PCit

(g L�1 h�1)YCit/S

(g g�1)

92 20.5 0.0 0.39 0.21 7.1 6.0 0.07 0.29

200 45.9 0.9 0.38 0.21 7.9 17.8 0.09 0.40

240 60.5 2.8 0.35 0.19 7.1 29.0 0.12 0.50

280 90.3 16.9 0.26 0.17 7.0 36.5 0.13 0.50

600 164.0 52.5 0.19 0.16 7.4 62.5 0.10 0.56

Representation of maximum concentration biomass (Xmax, g L�1), remaining glycerol quantity (Sr, g L�1) substrate uptake rate (rS, g L�1 h�1),

maximum specific growth rate (mmax, h�1), maximum concentration of citric acid (Citmax, g L�1), citric acid volumetric productivity

(PCit, g L�1 h�1) and total conversion yield of citric acid produced per glycerol consumed (YCit/S, g g�1). Sr, YCit/S and PCit values and fermentation

time are presented when the maximum concentration of citric acid had been achieved.

Culture conditions: growth on flasks at 18075 rpm, initial pH ¼ 6.170.2, initial (NH4)2SO4 0.5 g L�1, pH ranging between 5.0 and 6.2, incubation

temperature T ¼ 28 1C, oxygen saturation higher than 30% (v/v). Kinetics carried out at least in duplicate experiments by using different

inocula.a rS was estimated as rs ¼ �DS/Dt within the period in which remarkable quantity of glycerol was available.b mmax was estimated by fitting the equation ln(X/X0) ¼ f(t) within the early exponential growth phase.

0 0

0

X, C

it p

roduced (

g L

-1)

70

60

50

40

30

20

10

Time (h)

700600500400300200100

S r

em

ain

ing (

g L

-1)

200

150

100

50

X (g L-1)

Cit (g L-1)

S (g L-1)

Fig. 4 – Kinetics of biomass (X, g L�1) and citric acid (Cit, g L�1)

production and glycerol (S, g L�1) consumption of Yarrowia

lipolytica during growth in batch flask experiment. Culture

conditions: S0 ¼ 164.0 g L�1; pH ranging from 5.0 to 6.2;

incubation temperature T ¼ 28 1C. Kinetics were conducted

in duplicate by using different inocula.

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 1 65

Author's personal copy

decreased with glycerol concentration increase in the med-

ium (Table 4). From the decrease of mmax, rS and principally

Xmax values when S0 concentration increased, it may be

assumed that noticeable substrate inhibition was exerted

mainly at significantly high S0 quantity media (this was

obvious at S0 ¼ 134 g L�1). In contrast, in the culture with

S0 ¼ 100.5 g L�1, despite slight decrease of mmax value, rS was

comparable with the trial with S0 ¼ 50.5 g L�1, while biomass

production and lipogenesis were significant; Xmax was

8.5 g L�1 with 51.7% of lipids (wt/wt in dry matter) accumu-

lated inside the fungal mycelia (quantity corresponding to

4.4 g L�1), while lipid yield (YL/S) obtained the value of

0.0770.01 g g�1 in all trials (Table 4). Finally, total biomass

yield (YX/S) presented the value of 0.1570.01 g g�1, with

the lower values obtained at high S0 concentration media

(data not presented).

In all the trials, M. isabellina grew in full aerobic conditions,

since the DO level corresponded to oxygen saturation higher

than 70% (v/v) for all growth steps. As previously (Y. lipolytica

fermentation), in the first growth step (0–45 h), in which

ammonium nitrogen was found in the medium, qO2presented

elevated values (around 0.4070.05 mg mg�1 h�1) and then the

respiratory activity of the strain significantly decreased

(qO2values around 0.0870.02 mg mg�1 h�1). The fermentation

kinetics of M. isabellina when S0 ¼ 100.5 g L�1 is shown in

Fig. 5.

In M. isabellina, the principal cellular fatty acids were oleic

(D9C18:1), linoleic (D9,12C18:2) and palmitic (C16:0) acid and

concentrations of these remained almost constant during

growth. Cellular fatty acids produced in lower quantities were

stearic acid (C18:0), palimitoleic acid (D9C16:1) and g-linolenic

acid (D6,9,12C18:3—GLA) (Table 5). In the first growth step, GLA

ARTICLE IN PRESS

Table 3 – Fatty acid composition of lipid produced by Yarrowia lipolytica (%, wt/wt), when this microorganism was grown onraw glycerol

Fermentation time C16:0 D9C16:1 C18:0 D9C18:1 D9,12C18:2 D9C16:1/C16:0

D9C18:1/C18:0

D9,12C18:2/D9C18:1

Growth on raw glycerol, S0 ¼ 45.9 g L�1

Late exponential

(50–70 h)

15.7 7.1 12.5 40.1 22.2 0.45 3.21 0.55

Mid stationary

(120–140 h)

13.1 8.1 9.9 45.1 19.2 0.62 4.55 0.42

Late stationary

(220–250 h)

14.0 8.0 11.0 44.9 20.2 0.57 4.08 0.45

Growth on raw glycerol, S0 ¼ 164.0 g L�1

Late exponential

(50–70 h)

15.9 8.9 7.9 46.9 20.0 0.55 5.93 0.42

Mid stationary

(120–140 h)

12.1 13.9 5.9 44.1 21.8 1.14 7.47 0.49

Late stationary

(220–250 h)

16.9 15.5 5.2 42.9 19.4 0.92 8.25 0.45

Culture conditions as in Table 2.

Table 4 – Quantitative data of Mortierella isabellina originated from kinetics in media with various initial glycerol (S0)concentrations

Fermentation time(h)

S0

(g L�1)Sr

(g L�1)rS

a

(g L�1 h�1)mmax

b

(h�1)Xmax

(g L�1)Lmax

(g L�1)Lipid (%, wt/

wt)YL/S

(g g�1)

310 50.5 9.8 0.13 0.08 6.5 2.4 36.9 0.06

420 100.5 45.1 0.13 0.07 8.5 4.4 51.7 0.08

380 135.1 99.5 0.09 0.07 5.0 2.5 50.0 0.07

Representation of maximum concentration biomass (Xmax, g L�1), remaining glycerol quantity (Sr, g L�1) substrate uptake rate (rS, g L�1 h�1),

maximum specific growth rate (mmax, h�1), maximum concentration of cellular lipid (Lmax, g L�1), corresponding lipid in dry weight (%, wt/wt)

and total conversion yield of lipid produced per glycerol consumed (YL/S, g g�1).

Sr and YL/S values and fermentation time are presented when the maximum concentration of lipid had been achieved. Culture conditions:

growth on flasks at 18075 rpm, initial pH ¼ 6.170.2, initial (NH4)2SO4 0.5 g L�1, pH ranging between 5.4 and 6.0, incubation temperature

T ¼ 28 1C, oxygen saturation higher than 70% (v/v). Kinetics carried out at least in duplicate experiments by using different inocula.a rS was estimated as rs ¼ �DS/Dt throughout culture.b mmax was estimated by fitting the equation ln(X/X0) ¼ f(t) within the early exponential growth phase.

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 166

Author's personal copy

content was elevated, while later, during lipid accumulation,

GLA concentration somewhat decreased, and then remained

constant. Similar total cellular fatty acid profiles were

observed for all fermentations of M. isabellina, regardless of

the concentration of glycerol employed in the culture

medium. GLAmax concentration was achieved at S0 ¼

100.5 g L�1 and corresponded to 120 mg L�1.

3.4. Comparison between growth on pure and rawglycerol

In order to investigate if the impurities of the industrial

feedstock could affect microbial growth and production of

metabolites, control experiments for all microorganisms

(batch cultures) were realised using pure glycerol as sole

carbon source and they were compared with the results

obtained when raw glycerol was used as substrate. The

cultures showed similar growth and production of metabo-

lites for all microorganisms used (C. butyricum, Y. lipolytica, M.

isabellina, data not presented). Therefore, it may be consid-

ered that the impurities of the industrial feedstock (salts,

methanol, etc.) do not have any impact on the metabolism of

the studied strains.

4. Discussion

Batch cultures of C. butyricum F2b were accompanied by PDmax

production of 47.1 g L�1, a quantity comparable with literature

values for batch and fed-batch fermentations of Clostridium

sp. or Enterobacteria cultured on glycerol used as sole

substrate [2,3,16–18,22,24–26]. Though, in some cases utilisa-

tion of C. butyricum mutants in fed-batch fermentations

resulted in significant PD quantities, of more than 70 g L�1

[27], while, the highest final PD concentrations reported in the

literature were 80.1–87.7 g L�1, obtained in fed-batch cultures

of a newly isolated Clostridium sp. strain on pure or raw

glycerol [1]. In the present study, YPD/S remained constant

despite differentiations in the S0 concentration in the

medium. Generally, the fermentation of glycerol by Clostri-

dium sp. strains presents a higher YPD/S than that conducted

by the Klebsiella sp. since in the later case the metabolism is

directed towards the synthesis of more end-fermentation

products compared with the former one [6]. With Klebsiella

sp., YPD/S can be substantially increased when differentiations

in the culture conditions are employed such as utilisation of

glycerol in significant excess or culture carried out under

micro-aerobic conditions or differentiations employed in

the extracellular oxido-reduction potential of culture

[25,26,28,29]. The spectrum of end-fermentation products in

K. pneumoniae cultures can also be critically influenced by the

‘‘purity’’ of the substrate used (use of ‘‘raw’’ or ‘‘pure’’ glycerol)

or the addition of organic compounds (e.g. fumarate) into the

medium [18,30]. In contrast, C. butyricum cultures are carried

out only in strict anaerobiosis, with distribution between final

concentrations of PD, Ac and But being strain dependent

[6,16].

In the present study, in continuous culture and steady-state

conditions, PD pulses were carried out. It was shown that at

increased PD concentrations at transitory stage, although X

concentration remained practically constant, yield YX/S de-

creased. Simultaneously, slight decrease of S uptake, and

increment of Ac and But concentrations and yields were

observed. The above findings suggest that even at very high

PD concentrations (e.g. 84.2 g L�1 into the chemostat vessel),

growth was not inhibited, while carbon flow was mainly

channelled towards the synthesis of organic acids instead of

biomass, resulting, thus, in carbon losses through the

reaction of decarboxylation of pyruvic acid (see reactions in

Fig. 1). Similarly, continuous cultures at constant D and

ARTICLE IN PRESS

0

2

4

6

8

0

0

L (g L-1)100

X, L p

roduced (

g L

-1)

Time (h)

500400300200100

10

12

S r

em

ain

ing (

g L

-1)

120

80

60

40

20

S (g L-1)X (g L-1)

Fig. 5 – Kinetics of biomass (X, g L�1) and lipid (L, g L�1)

production and glycerol (S, g L�1) consumption of Mortierella

isabellina during growth in batch flask experiment. Culture

conditions: S0 ¼ 100.5 g L�1; pH ranging from 6.1 to 5.2;

incubation temperature T ¼ 28 1C. Kinetics were conducted

in duplicate by using different inocula.

Table 5 – Fatty acid composition of lipid produced by Mortierella isabellina (%, wt/wt), when this microorganism was grownon raw glycerol

Fermentation time C16:0 D9C16:1 C18:0 D9C18:1 D9,12C18:2 D6,9,12C18:3

Late exponential (40–50 h) 20.1 1.3 5.8 46.6 18.0 8.3

Mid stationary (130–150 h) 21.2 2.8 5.2 52.0 13.5 3.8

Late stationary (220–250 h) 22.0 4.2 4.3 50.0 16.7 3.7

Culture conditions as in Table 4.

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 1 67

Author's personal copy

various inlet concentrations of the employed strain showed

that when S0 concentration increased, decrease in YX/S and Rc

values was observed. It was assumed that at high S0

concentrations (when the 1,3-propanediol pathway was

somehow saturated), carbon flow towards biomass and PD

synthesis decreased and the percentage of NAD+ regenerated

via the pathway of butyric acid biosynthesis increased,

resulting, hence, in carbon losses [19]. In contrast, in C.

butyricum DSM 5431 and its product-tolerant mutants, with

increase of inlet glycerol and thus 1,3-propanediol produced

concentration, although the yield YPD/S and the ratio of GDH

versus GDHt remained constant in the range of substrate

concentrations tested, a net increase of the NADH2/NAD+

ratio was observed, suggesting insufficient electron flow

through the reductive branch of the pathway of glycerol

catabolism, while the pathway of Ac biosynthesis was

favoured instead of that of But due to high intracellular

concentrations of acetyl-CoA detected, and that suggest

bottlenecks in the enzymes (e.g. aceto-acetyl-CoA thiolase,

butyrate kinase) leading to the formation of butyric acid [31].

In a two-stage culture carried out, large PD quantities

(43.5 g L�1) comparable with the highest values reported in the

literature for single- or two-stage 1,3-propanediol production

processes [28,32,33] were produced. The second stage of the

culture showed lower YPD/S, Rc and X values compared with

those of the first stage. This is probably due to incomplete

reduction of 3-HPA (the only metabolic intermediate of PD

biosynthesis pathway) to 1,3-propanediol and accumulation

of this toxic intermediate in the growth medium. This

intermediate product was not measured, and, possibly, for

this reason in the second stage low Rc values were observed.

Furthermore, in the second stage a metabolic shift favouring

the synthesis of acetic acid against that of butyric acid was

observed (see Table 1). It may be assumed, hence, that in the

above conditions, in the second stage of the culture a

bottleneck for the carbon flux through the butyrate biosynth-

esis occurred, resulting, thus, in shifting of the cellular

metabolism towards the biosynthesis of the acetic acid.

Y. lipolytica ACA-DC 50109 presented significant cell growth

with slight inhibition when raw glycerol was utilised as the

sole carbon source, although in some cases extremely high

initial glycerol quantities (S0 ¼ 164 g L�1) were used. Incre-

ment of glycerol concentration in the medium resulted in

significant decrease of rS, while mainly at high-substrate

concentration media, notable quantities of glycerol remained

unconsumed even though the fermentation time was

remarkably high. In M. isabellina ATHUM 2935 cultures,

biomass production declined at relatively high S0 concentra-

tions in the medium (e.g. at 134 g L�1) due to substrate

inhibition, while in all cultures significant quantities of

glycerol remained unconsumed in spite of the fact that the

fermentations were carried out for long incubation periods

(e.g. 300–400 h). In the literature, data dealing with the

cultivation of Y. lipolytica or lipid-accumulating strains on

glycerol used as sole substrate are rather equivocal. The

recombinant Y. lipolytica CX161-1B strain was batch cultured

on various S0 concentration media and significant growth

with similar mmax values was observed, regardless of S0

concentration, while no substrate inhibition was observed

even when S0 quantity was 150 g L�1 and all of the available

glycerol quantity was consumed [8]. The lipid-accumulating

C. curvata ATCC 20509 strain showed almost a 2-fold decrease

of Xmax value, when glycerol concentration increased from 32

to 128 g L�1 in flask nitrogen-limited cultures, while further S0

quantity increase, almost ceased microbial growth due to

inhibition exerted in high initial glycerol concentrations [14].

However, in fed-batch stirred tank bioreactor experiments,

the above microorganism consumed 300 g L�1 of glycerol

within 55 h of culture (rS ¼ 5.5 g L�1 h�1), producing around

120 g L�1 of biomass that contained 25% (wt/wt) of lipid [14],

while with a specified pumped external loop bioreactor the

respective values were 100 g L�1 and 32% (wt/wt), within the

same fermentation time [34]. In contrast, the oleaginous

Cunninghamella echinulata CCRC 31840 presented decreased

microbial growth and accumulation of lipid when glycerol

was used as substrate [35], and similar results were obtained

for other oleaginous Mucorales [36]. Additionally, fed-batch

nitrogen-limited cultures of the thiamine auxotroph

Y. lipolytica VKM-374/4 strain growing on glycerol, showed

increased substrate uptake rate despite S0 amounts used [9],

while citric-acid-producing Y. lipolytica mutants cultivated on

raw glycerol at extremely high S0 quantities (i.e. 200 g L�1)

presented efficient cell growth (X ¼ 16.5–26.5 g L�1) and com-

plete substrate consumption within 80–140 h without obvious

inhibition observed (rS from 1.4 to 2.5 g L�1 h�1—values

significantly higher compared to the ones obtained in the

present study) [11].

In the lipid- or citric acid-producing microorganisms that

are cultivated on glycerol used as substrate, glycerol, which

enters inside the microbial cell by facilitated diffusion, is

firstly converted to 3-P-glycerol and then to 3-P-dihydroxya-

cetone, reactions catalysed by GK and 3-P-GDH, thus entering

the second branch of the glucolytic pathway to yield pyruvate.

Subsequent metabolic steps to lipid or citric acid biosynthesis

are the same as for glucose (see Fig. 1). However, a portion of

glycerol must be channelled to hexose synthesis and then to

pentose synthesis, which are necessary for nucleic acid

biosynthesis and NADPH2 generation. Hexose synthesis is

carried out through gluconeogenesis that in the case of

glycerol involves the reversal of carbon flow from

3-P-glyceraldehyde. In the present study, both M. isabellina

and Y. lipolytica consumed with reduced rates glycerol mainly

at high S0 concentration media, since significant quantities of

this substrate remained in the medium despite high fermen-

tation duration. Considering for the case of M. isabellina that

huge quantities of biomass and fat (36 and 18 g L�1, respec-

tively) were produced at high initial glucose media

(e.g. 100 g L�1) with glucose almost completely consumed

[21], the reason for the slow assimilation of glycerol and the

lower production of biomass and lipid compared with the

glucose fermentation can be poor regulation of the enzymes

involved in the primary metabolic steps of glycerol assimila-

tion (GK, 3-P-GDH), or decreased activity of gluconeogenic

enzymes, or, finally slow activity of the NADP+-malic enzyme

(ME) that can severely curtail substrate uptake and accumula-

tion of fat [13]. In contrast, in glucose fermentation of

Y. lipolytica, despite notable growth and citric acid produc-

tion at high initial substrate concentration media

(i.e. 150 g L�1), as in the case of glycerol fermentation, glucose

remained unconsumed in significant quantities at the end of

ARTICLE IN PRESS

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 168

Author's personal copy

growth [37]. The fact, hence, that non-negligible glycerol

amounts remain unconsumed in the medium after long

incubation periods, seems not related with the regulatory

mechanisms of glycerol assimilation or gluconeogenesis

process but with the fact of being a strain-dependant event.

Other Y. lipolytica strains cultured on various initial glucose

carbon-excess batch cultures (glucose variations from 50 to

150 g L�1), consumed almost all available glucose quantity

[38,39], while in our laboratory, growth of newly isolated or

genetically engineered Y. lipolytica strains on glucose or raw

glycerol at various concentrations and carbon-excess condi-

tions, showed indeed strain-dependent discrepancies on

substrate uptake rate and complete exhaustion of the carbon

source from the medium (Papanikolaou, unpublished data).

In the present investigation, in Y. lipolytica cultures citric

acid was synthesised in significant quantities (Citmax ¼

62.5 g L�1), comparable with values reported in the literature

(values within 30 and 80 g L�1) [38–43]. It should be noted that

very high concentrations of citric acid (Citmax ¼ 120–135 g L�1)

have been achieved by mutant Y. lipolytica strains cultivated

on ethanol [44], crude lipids [45] or raw glycerol [11] in batch,

fed-batch or continuous operations. Likewise, YCit/S signifi-

cantly increased with glycerol concentration increment in the

medium due to ‘‘metabolic overflow’’ at carbon-excess con-

ditions [10,40]. Additionally, the highest yield YCit/S presented

the value of 0.56 g g�1 which was satisfactory and comparable

with the values reported in the literature for various Candida

sp. strains cultured on sugar-based media (yields from 0.50 to

0.77 g g�1 [38–40,43]). YCit/S is significantly dependent on the

fermented carbon source utilised, since substances with

higher carbon and reductance degree compared with glycerol,

can lead to higher conversion yields; for instance, growth of Y.

lipolytica strains on ethanol was accompanied by a citric acid

yield of 0.87 g g�1 [44], while growth on hydrophobic sub-

strates can be accompanied by conversion yields, which may

be close to or even higher than 1.0 g g�1 [41,45].

Although Y. lipolytica ACA-DC 50109 was reported capable to

produce huge quantities of lipid during growth on raw

glycerol in continuous bioreactor nitrogen-limited cultures

[15], in flask experiments the microorganism did not accu-

mulate notable lipid quantities. Cellular lipids showed slightly

higher quantity of unsaturated fatty acids with high initial

glycerol media. The higher ‘‘fluidity’’ on the membrane,

hence, could reflect the increased osmotic pressure of the

medium due to high glycerol concentration.

M. isabellina ATHUM 2935 cultivated on raw glycerol utilised

as the sole substrate in nitrogen-limited experiments pro-

duced relatively satisfactory quantities of microbial lipid

[up to 4.4 g L�1, around 50% (wt/wt) of lipid in dry weight].

This is an interesting result, given that glycerol is not

considered as an adequate substrate for lipid production for

various oleaginous Zygomycetes [12,35,36]. In contrast, sub-

stances favouring production of SCO from oleaginous Zygo-

mycetes in higher quantities compared with the present

study (e.g. quantities higher than 8 g L�1) are various sugars

and polysaccharides [12,35,36,46] or fatty materials [47–49]. In

general, a scarce number of investigations have been carried out

dealing with the production of lipid using glycerol as the sole

substrate, with YL/S values 0.08–0.15 g g�1 (values comparable

with the present study) [14,15,34]. Additionally, the fatty acid

profile of the microbial oil produced was largely unaffected by

initial concentration of glycerol, while GLA concentration in the

produced lipid decreased as a function of the fermentation time

(in accordance with Chen and Chang [35] and Chen and Liu [46]).

It is noted that in oleaginous Zygomycetes, GLA resides mostly

in the TAG fraction that comprises the edible portion of the lipid

[23]. GLAmax concentration of 120 mg L�1 was obtained, which is

2.5–10-fold lower compared with that obtained in various

Mucorales strains. The same strain (ATHUM 2935) produced

around 800 mg L�1 of GLA during growth on high-glucose media

[21]. GLAmax quantity of Mucor spp. strains was 370mg L�1 of

medium during growth on mixtures of fats and glucose [47].

A Mucor racemosus strain cultivated on olive oil produced huge

quantities of lipids (13.4 g L�1) with GLA produced at 725 mg L�1

[49]. C. echinulata CCRC 31840 produced 964 mg L�1 of GLA after

optimisation of the growth conditions [35] while additional

optimisation of the inoculation led to a production of

1350 mg L�1, after 5 days of culture [46].

In conclusion, raw glycerol, by-product discharged after

bio-diesel production process, was an adequate substrate for

the growth of C. butyricum F2b, Y. lipolytica ACA-DC 50109 and

M. isabellina ATHUM 2935. High value-added metabolic

products (1,3-propanediol, citric acid and SCO) in remarkable

quantities and satisfactory conversion yields were produced.

As for the economic significance of the results obtained from

the present study, it should be stressed that the utilisation of

bio-diesel (fatty acid methyl esters) in continuously increas-

ing quantities has dramatically decreased the price of

glycerol, resulting in the necessity of discovery of various

integrated bioprocesses of valorisation of this residue

(‘‘biorefinery approach’’). By 2007, it is expected that there

will be an over-capacity of more than 600,000 metric tonnes of

glycerol residue produced, based on the bio-diesel industry

growth in Europe. Currently, in various countries of Western

Europe (e.g. Germany), crude glycerine water derived from

various bio-diesel plants is treated as a typical ‘‘industrial

waste water’’ (with a cost of 0$ per kg—it is, hence, a waste

material) being used directly for biogas production. The

current cost of 1,3-propanediol is estimated to be 1.5–2.0$

per kg while that of citric acid is 0.8–1.2$ per kg. As for plant

lipids that contain g-linolenic acid and are currently com-

mercialised [in general, it is the case of the oil deriving from

the plant Oenothera biennis called also Evening primrose oil,

that contains GLA in concentrations 8–10% (wt/wt)], their cost

is estimated to be 45–50$ per kg. Therefore, although our

experiments have been carried out in small-scale operations

and scale-up studies are envisaged in future investigations of

our research teams, the significantly low (or even zero) cost or

raw glycerol and the (relatively) high price of the metabolites

produced, can potentially result in the viability of the

proposed approaches dealing with the biotransformation of

raw glycerol into 1,3-propanediol, citric acid and SCO contain-

ing g-linolenic acid.

Acknowledgments

Financial support was provided by: (a) the project entitled

‘‘Bio-diesel production from agro-industrial by-products’’

funded by the Greek Fuel Company DRACOIL SA; (b) the

ARTICLE IN PRESS

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 1 69

Author's personal copy

bilateral project between Greece and Slovak Republic entitled

‘‘Biotechnological production of bioactive lipids from

agro-industrial by-products’’.

R E F E R E N C E S

[1] Hirschmann S, Baganz K, Koschik I, Vorlop KD. Developmentof an integrated bioconversion process for the production of1,3-propanediol from raw glycerol waters. LandbauforschungVolkenrode 2005;55:261–7.

[2] Barbirato F, Himmi EH, Conte T, Bories A. 1,3-Propanediolproduction by fermentation: an interesting way to valorizeglycerin from the ester and ethanol industries. IndustrialCrops and Products 1998;7:281–9.

[3] Himmi EH, Bories A, Barbirato F. Nutrient requirements forglycerol conversion to 1,3-propanediol by Clostridium butyr-icum. Bioresource Technology 1999;67:123–8.

[4] Zeng AP, Biebl H. Bulk chemicals from biotechnology: thecase of 1,3-propanediol production and the new trends.Advances in Biochemical Engineering and Biotechnology2002;74:239–59.

[5] Lee SY, Hong SH, Lee SH, Park SJ. Fermentative production ofchemicals that can be used for polymer synthesis. Macro-molecular Bioscience 2004;4:157–64.

[6] Biebl H, Menzel K, Zeng AP, Deckwer WD. Microbial produc-tion of 1,3-propanediol. Applied Microbiology and Biotech-nology 1999;52:289–97.

[7] Ito T, Nakashimada Y, Senba K, Matsui T, Nishio N. Hydrogenand ethanol production from glycerol-containing wastesdischarged after biodiesel manufacturing process. Journal ofBioscience and Bioengineering 2005;100:260–5.

[8] Kim JW, Park TJ, Ryu DDY, Kim JY. High cell density culture ofYarrowia lipolytica using a one-step feeding process. Biotech-nology Progress 2000;16:657–60.

[9] Morgunov I, Kamzolova S, Perevoznikova O, Shishkanova N,Finogenova T. Pyruvic acid production by a thiamineauxotroph of Yarrowia lipolytica. Process Biochemistry2004;39:1469–74.

[10] Soccol CR, Vandenberghe LPS, Rondriques C, Pandey A. Newperspectives for citric acid production and application. FoodTechnology and Biotechnology 2006;44:141–9.

[11] Rymowicz W, Rywinska A, Zarowska B, Juszczyk P. Citric acidproduction from raw glycerol by acetate mutants of Yarrowialipolytica. Chemical Papers 2006;60:391–4.

[12] Certık M. Fermentation physiology and regulation of micro-bial polyunsaturated fatty acid biosynthesis. Research inAdvanced Bioscience and Bioengineering 2000;1:45–64.

[13] Ratledge C. Regulation of lipid accumulation in oleaginousmicro-organisms. Biochemical Society Transactions2002;30:1047–50.

[14] Meesters PAEP, Huijberts GNM, Eggink G. High-cell-densitycultivation of the lipid accumulating yeast Cryptococcuscurvatus using glycerol as a carbon source. Applied Micro-biology and Biotechnology 1996;45:575–9.

[15] Papanikolaou S, Aggelis G. Lipid production by Yarrowialipolytica growing on industrial glycerol in a single-stagecontinuous culture. Bioresource Technology 2002;82:43–9.

[16] Petitdemange E, Durr C, Abbad-Andaloussi S, Raval G. Fermen-tation of raw glycerol to 1,3-propanediol by new strains ofClostridium butyricum. Journal of Industrial Microbiology 2005;15.

[17] Gonzalez-Pajuelo M, Andrade JC, Vasconcelos I. Production of1,3-propanediol by Clostridium butyricum VPI 3266 using asynthetic medium and raw glycerol. Journal of IndustrialMicrobiology and Biotechnology 2004;31:442–6.

[18] Mu Y, Zhang D, Teng H, Wang W, Xiu ZL. Microbial productionof 1,3-propanediol by Klebsiella pneumoniae using crude

glycerol from bio-diesel preparation. Biotechnology Letters2006;28:1755–9.

[19] Papanikolaou S, Fick M, Aggelis G. The effect of raw glycerolconcentration on the production of 1,3-propanediol byClostridium butyricum. Journal of Chemical Technology andBiotechnology 2004;79:1189–96.

[20] Papanikolaou S, Muniglia L, Chevalot I, Aggelis G, Marc I.Yarrowia lipolytica as a potential producer of citric acid fromraw glycerol. Journal of Applied Microbiology 2002;92:737–44.

[21] Papanikolaou S, Komaitis M, Aggelis G. Single cell oil (SCO)production by Mortierella isabellina grown on high-sugarcontent media. Bioresource Technology 2004;95:287–91.

[22] Homann T, Tag C, Biebl H, Decker WD, Schink B. Fermenta-tion of glycerol to 1-3 propanediol by Klebsiella and Citrobacterstrains. Applied Microbiology and Biotechnology1990;33:121–6.

[23] Fakas S, Papanikolaou S, Galiotou-Panayotou M, Komaitis M,Aggelis G. Lipids of Cunninghamella echinulata with emphasisto g-linolenic acid distribution among lipid classes. AppliedMicrobiology and Biotechnology 2006;73:676–83.

[24] Chen Z, Xiu ZL, Wang JF, Zhang D, Xu P. Stoichiometricanalysis and experimental investigation of glycerol biocon-version to 1,3-propanediol by Klebsiella pneumoniae undermicroaerobic conditions. Enzyme and Microbial Technology2003;33:386–94.

[25] Chen X, Zhang DJ, Qi WT, Gao SJ, Xiu ZL, Xu P. Microbial fed-batch production of 1,3-propanediol by Klebsiella pneumoniaeunder micro-aerobic conditions. Applied Microbiology andBiotechnology 2003;63:143–6.

[26] Du C, Yan H, Zhang Y, Li Y, Cao Z. Use of oxidoreductionpotential as an indicator to regulate 1,3-propanediol fer-mentation by Klebsiella pneumoniae. Applied Microbiology andBiotechnology 2006;69:554–63.

[27] Reimann A, Biebl H. Production of 1,3-propanediol byClostridium butyricum DSM 5431 and product tolerant mu-tants in fed-batch cultures: feeding strategy for glycerol andammonium. Biotechnology Letters 1996;18:827–32.

[28] Menzel K, Zeng AP, Deckwer WD. High concentration andproductivity of 1,3-propanediol from continuous fermenta-tion of glycerol by Klebsiella pneumoniae. Enzyme and Micro-bial Technology 1997;20:82–6.

[29] Menzel K, Zeng AP, Deckwer WD. Enzymatic evidence for aninvolvement of pyruvate dehydrogenase in the anaerobicglycerol metabolism of Klebsiella pneumoniae. Journal ofBiotechnology 1997;56:135–42.

[30] Lin R, Liu H, Hao J, Cheng K, Liu D. Enhancement of 1,3-propanediol production by Klebsiella pneumoniae with fuma-rate addition. Biotechnology Letters 2005;27:1755–9.

[31] Reimann A, Abbad-Andaloussi S, Biebl H, Petitdemange H.1,3-Propanediol formation with product tolerant mutants ofClostridium butyricum DSM 5431 in continuous culture:productivity, carbon and electron flow. Journal of AppliedMicrobiology 1998;84:1125–30.

[32] Boenigk R, Bowien S, Gottschalk G. Fermentation of glycerolin continuous cultures of Citrobacter freundii. Applied Micro-biology and Biotechnology 1993;38:453–7.

[33] Xiu ZL, Song BH, Wang ZT, Sun LH, Feng EM, Zeng AP.Optimization of dissimilation of glycerol to 1,3-propanediolby Klebsiella pneumoniae in one- and two-stage anaerobiccultures. Biochemical Engineering Journal 2004;19:189–97.

[34] Meesters PAEP, van der Wal H, Weusthuis R, Eggink G.Cultivation of the oleaginous yeast Cryptococcus curvatus ina new reactor with improved mixing and mass transfercharacteristics (Super-s). Biotechnology Techniques1996;10:277–82.

[35] Chen HC, Chang CC. Production of g-linolenic acid by thefungus Cunninghamella echinulata CCRC 31840. BiotechnologyProgress 1996;12:338–41.

ARTICLE IN PRESS

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 170

Author's personal copy

[36] Sajbidor J, Certık M, Dobronova S. Influence of differentcarbon sources on growth, lipid content and fatty acidcomposition in four strains belonging to Mucorales. Bio-technology Letters 1988;10:347–50.

[37] Papanikolaou S, Galiotou-Panayotou M, Chevalot I, KomaitisM, Marc I, Aggelis G. Influence of glucose and saturated free-fatty acid mixtures on citric acid and lipid production byYarrowia lipolytica. Current Microbiology 2006;52:134–42.

[38] Rane KD, Sims KA. Production of citric acid by Candidalipolytica Y 1095: effect of glucose concentration on yield andproductivity. Enzyme and Microbial Technology1993;15:646–51.

[39] Moresi C. Effect of glucose concentration on citric acidproduction by Yarrowia lipolytica—kinetics of the tropho-phase, citrate lag phase and idiophase. Journal of ChemicalTechnology and Biotechnology 1994;60:387–95.

[40] Anastassiadis S, Aivasidis A, Wandrey C. Citric acid produc-tion by Candida strains under intracellular nitrogen limita-tion. Applied Microbiology and Biotechnology 2002;60:81–7.

[41] Crolla A, Kennedy KJ. In-line mixing for production of citricacid by Candida lipolytica grown on n-paraffins. Journal ofChemical Technology and Biotechnology 2004;79:720–8.

[42] Anastassiadis S, Rehm HJ. Continuous citric acid secretion bya high specific pH dependent active transport system in yeastCandida oleophila ATCC 20177. Electronic Journal of Biotech-nology 2005;8:146–61.

[43] Rymowicz W, Cibis E. Optimization of citric acid productionfrom glucose syrup by Yarrowia lipolytica using responsesurface methodology. Electronic Journal of Polish Agricul-tural Universities 2006;9:#20.

[44] Kamzolova S, Shishkanova N, Morgunov I, Finogenova T.Oxygen requirements for growth and citric acid productionof Yarrowia lipolytica. FEMS Yeast Research 2003;3:217–22.

[45] Kamzolova S, Morgunov I, Aurich A, Perevoznikova S,Shiskanova N, Stottmeister U, et al. Lipase secretion andcitric acid production in Yarrowia lipolytica yeast grown onanimal and vegetable fat. Food Technology and Biotechnol-ogy 2005;43:113–22.

[46] Chen HC, Liu TM. Inoculum effects on the production of g-linolenic acid by the shake culture of Cunninghamella echinulataCCRC 31840. Enzyme and Microbial Technology 1997;21:137–42.

[47] Certık M, Balteszova L, Sajbidor J. Lipid formation and g-linolenic acid production by Mucorales fungi grown onsunflower oil. Letters in Applied Microbiology 1997;25:101–5.

[48] Mantzouridou F, Tsimidou M, Roukas T. Performance ofcrude olive pomace oil and soybean oil during carotenoidproduction by Blakeslea trispora in submerged fermentation.Journal of Agricultural and Food Chemistry 2006;54:2575–81.

[49] Szczesna-Antczak M, Antczak T, Piotrowicz-Wasiak M,Rzyska M, Binkowska N, Bielecki S. Relationships betweenlipases and lipids in mycelia of two Mucor strains. Enzymeand Microbial Technology 2006;39:1214–22.

ARTICLE IN PRESS

B I O M A S S A N D B I O E N E R G Y 3 2 ( 2 0 0 8 ) 6 0 – 7 1 71