See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/6792219

Culture media optimization for growth and phycoerythrin production from

Porphyridium purpureum

Article  in  Biotechnology and Bioengineering · February 2007

DOI: 10.1002/bit.21138 · Source: PubMed

CITATIONS READS
41 498

4 authors, including:

Ravi Sarada Ravishankar Gokare

Central Food Technological Research Institute Dayananda Sagar Institutions
97 PUBLICATIONS   2,977 CITATIONS    380 PUBLICATIONS   11,843 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Conservation of endangered plants by biotechnological applications View project

Production of bio-fuels from Malaysian micro-algal spices View project

All content following this page was uploaded by Ravi Sarada on 29 August 2017.

The user has requested enhancement of the downloaded file.

 ARTICLE

Culture Media Optimization for Growth

and Phycoerythrin Production from
Porphyridium purpureum
S. Kathiresan,1 R. Sarada,1 Sila Bhattacharya,2 G.A. Ravishankar1
1
Plant Cell Biotechnology Department, Central Food Technological Research Institute,
Mysore, India 570020; telephone: 91-821-2516501; fax: 91-821-2517233;
e-mail: sarada_ravi@yahoo.com
2
Grain Science Technology Department, Central Food Technological Research Institute,
Mysore, India 570020
Received 27 January 2006; accepted 13 July 2006
Published online 25 September 2006 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/bit.21138

algae), Cryptophyta (cryptomonad algae), and Rhodophyta

ABSTRACT: Porphyridium spp. is a red micro alga and is
gaining importance as a source of valuable products viz.,
phycobiliproteins (PB), sulfated exopolysaccharides, and
polyunsaturated fatty acids with potential applications in
the food and pharmaceutical industries. In the present study,
the effects of the major media constituents of Porphyridium
species were studied using response surface methodology
(RSM) on biomass yield, total PB and the production of
phycoerythrin (PE). A second order polynomial can be used
to predict the PB and PE production in terms of the
independent variables. The independent variables such as
the concentrations of sodium chloride, magnesium sulfate,
sodium nitrate, and dipotassium hydrogen phosphate influ-
enced the total PB and PE production. The optimum
conditions showed that total PB was 4.8% at the concen-
tration of sodium chloride 26.1 g/L, magnesium sulfate
5.23 g/L, sodium nitrate 1.56 g/L, and dipotassium hydrogen
phosphate 0.034 g/L. In case of optimum PE production
(3.3%), the corresponding values are 29.62, 6.11, 1.59, and
0.076 g/L, respectively. PE production depends greatly on
the concentrations of chloride, nitrate, and sulfate as well as
phosphate of which the former possess the maximum effect.
Biotechnol. Bioeng. 2007;96: 456–463.
ß 2006 Wiley Periodicals, Inc.
KEYWORDS: Porphyridium; phycobiliproteins (PB); phy-
coerythrin (PE); biomass; response surface methodology
(RSM); optimization
Introduction
Phycobiliproteins (PB) are the major accessory light-
harvesting pigments present in Cyanophyta (blue green
(red algae) (Borowitzka, 1994). The PB consists of
protein moiety and a linear tetrapyrrole prosthetic group
(phycobilins) attached to the polypeptide through thioether
bonds. Based on spectroscopic properties, these proteins are
broadly classified into three groups—allophycocyanin
(APC) (650–655 nm), C-phycocyanin (C-PC) (610–
620 nm), and R-phycoerythrin (R-PE) (540–570 nm). Each
of these PB is comprised of two subunits a and b to which
linear tetrapyrrole is covalently attached by a cysteine
thioether bond. The central portion of all the three groups is
an open tetrapyrrole containing skeletal system similar to
chlorophyll and haemoglobin. Four types of tetrapyrroles
are known—phycocyanobilin, phycoerythrobilin, phycour-
obilin, and phycoerythrocyanin. These chromophoretic
proteins along with some of the colorless linker peptides
form an organized structure called phycobilisome (Tandeau
and Houmard, 1988). These phycobilisomes are situated on
thylakoid membranes, and channels the absorbed light
energy to photosynthetic system II.
The Porphyridium spp. are unique in producing B-
phycoerythrin (B-PE), a red color pigment and R-
phycocyanin (R-PC), a blue color pigment which have
more applications as food and cosmetic colorant, and as
fluorescent marker in immunodiagnostics (Borowitzka,
1994). There are two types of PE namely R-PE and B-PE,
which were named after the taxa of the organism from
which they were isolated. R-PE was isolated from
Rhodophyta and B-PE from Bangiales order (red seaweed)
of Rhodophyta. However, later they were grouped based on
their spectral differences (for example, R-PE (565 nm) and
B-PE (545 nm) differed in their absorption maxima).

Correspondence to: R. Sarada

Contract grant sponsor: CSIR
Contract grant sponsor: DBT

456 Biotechnology and Bioengineering, Vol. 96, No. 3, February 15, 2007 ß 2006 Wiley Periodicals, Inc.
Similarly, C-phycocyanin (Cyanophyta) and R-phycocyanin
(Rhodophyta) differed in their spectra. PE has also been
reported to possess beneficial effects as immunomodulators,
cancer preventive action (Dainippon Ink and Chemicals,
1983; Henrikson, 1989), as an elicitor for secondary
metabolite production in plant tissue culture (Rao et al.,
1996), and also for therapeutic purpose (Reddy et al.,
2000; Vadiraja and Madyastha, 2000). The blue colored
C-phycocyanin is prominent in blue green algae while PE is
in red algae. C-phycocyanin is commercially produced by
Spirulina while such a process is not available for the
manufacture of PE. Though the red microalga Porphyridium
is the potential organism for PE, it has been studied only
for extracellular polysaccharide production (Geresh et al.,
2002). With an increasing demand for natural colorants, the
microalgae gained focus. Since PE being a high-value
compound, the present study has relevance to develop
culture methods for high biomass and PE production in
P. purpureum. Growth of P. purpureum has been reported to
be slow in artificial seawater (ASW) medium (Tao et al.,
2004). Culture conditions such as light intensity and
residence time were reported to influence the content and
compositions of PB in red microalgae (Arad and Yaron,
1992; Fuentes et al., 2000). The conventional method of
medium optimization, one factor at a time, is time con-
suming and may lead to misinterpretation of results when
interactions between different components are present.
Statistical experimental designs can minimize the error
in determining the effect of parameters, and it allows
systematic and efficient variation of all parameters
(Ooijikass et al., 1999). These statistical experimental
designs can be adopted at various optimization processes,
such as for competitive objectives, for screening experi-
ments, or for finding the optimal conditions. Response
surface methodology (RSM) is commonly used tool for
design of experiments, analysis of results, and to obtain the
optimal conditions. This is an efficient statistical technique
for optimization of multiple variables with minimum
number of experiments (Vohra and Satyanarayana, 2002).
Therefore, the present study is aimed at understanding the
effects of media constituents and optimization employing
RSM for PB and PE production in P. purpureum.
Materials and Methods
Microorganism and Culture Conditions
Porphyridium purpureum (112.79) was obtained from
Sammlung von Algen Kulturen, Pflanzenphysiologisches
Institut, Universitat Gottingen, Germany. The stock culture
was maintained on ASW medium (Tao et al., 2004). ASW
medium consisted (g/L) of 27.0 g NaCl, 5.6 g MgCl2
6H2O,
1.5 g CaCl2
2H2O, 1.0 g KNO3, 0.07 g K2HPO4, 6.6 g
MgSO4
7H2O, 0.04 g NaHCO3, and 1 mL of trace metal
solution per liter and adjusted to pH 7.4.
The salt strength of the medium influences the growth
and pigment production in algae. As sodium chloride,
magnesium sulfate, sodium nitrate, and dipotassium
hydrogen phosphate are the major constituents of the
medium; these four constituents are taken for optimizing
their concentration using the RSM methodology.
A set of 250 mL conical flasks with 100 mL medium
containing different concentrations of the four variables
(sodium chloride, magnesium sulfate, sodium nitrate, and
dipotassium hydrogen phosphate) as mentioned in Table I
were taken and the flasks were inoculated and incubated on
Orbiteck shaker with 80 rpm speed maintained at 25  18C
under 1.5  0.2 klux (18.85 mmol photons/m2s or 2.2059 
107 erg/m2s) light intensity. Ten days old culture was used as
inoculum at 20% (v/v) for all the experiments. The cultures
were harvested at 14 and 28 days of growth. All the
experiments were carried out in triplicates.
Growth Measurement
The cultures were harvested by centrifugation (Remi C-24)
at 6,000 rpm. The biomass was washed with distilled water
and again centrifuged at the same speed. The wet biomass
after harvesting by centrifugation had moisture content of
90  2%. Biomass was estimated in terms of wet weight
(g/L). The PB and PE were expressed in terms of percent
(w/w) on dry weight basis.
Extraction and Estimation of Phycobiliproteins
The cells were harvested and the PB was extracted with
phosphate buffer by repeated freezing and thawing of cells.
Absorbance of the pooled extracts was read at 620, 650, and
565 nm in a Shimadzu UV spectrophotometer (UV-160).
The R-PC, APC, and B-PE contents were estimated using the
following equations (Tandeau and Houmard, 1988)
RPC ðmg=mLÞ ¼ OD620 nm  0:7 OD650 nm
7:38 (1)
APC ðmg=mLÞ ¼ OD650 nm 5:65
0:19 OD620 nm (2)
OD565 nm  2:8½R  PC  1:34½APC
BPE ðmg=mLÞ ¼ 12:7
(3)
Here, R-PC is the phycocyanin from red alga, APC is the
allophycocyanin, B-PE is the phycoerythrin from the red
algal order Bangiales and OD is the optical density of the
pigment at the particular wavelength. The total PB was
calculated by summing up of Equations (1)–(3).
Experimental Design and Analysis of Data
The design of the experiment was obtained by following the
method suggested by Akhnazarova and Kafarov (1982). The
Kathiresan et al.: Phycoerythrin Production 457
Biotechnology and Bioengineering. DOI 10.1002/bit
 458
Table I. Central composite experimental design in coded and actual level of variables and the response functions.

Dipotassium hydrogen
Experiment No Sodium chloride Magnesium sulfate Sodium nitrate phosphate

Coded level Actual levela Coded level Actual levela Coded level Actual levela Coded level Actual levela
(x1) (X1) (x2) (X2) (x3) (X3) (x4) (X4) Bio mass yield (g/L) Y1 PB content (%) Y2 PE content (%) Y3
1 1.00 25.608 1.00 8.536 1.00 1.707 1.00 0.171 3.44 4.35 2.89
2 1.00 4.392 1.00 1.464 1.00 1.707 1.00 0.171 5.00 2.91 1.45
3 1.00 25.608 1.00 1.464 1.00 0.293 1.00 0.171 6.44 1.79 1.67
4 1.00 4.392 1.00 8.536 1.00 0.293 1.00 0.171 5.08 2.43 1.63
5 1.00 25.608 1.00 1.464 1.00 1.707 1.00 0.029 6.00 4.07 2.63
6 1.00 4.392 1.00 8.536 1.00 1.707 1.00 0.029 4.04 1.80 1.76
7 1.00 25.608 1.00 8.536 1.00 0.293 1.00 0.029 4.84 3.81 1.75
8 1.00 4.392 1.00 1.464 1.00 0.293 1.00 0.029 4.96 1.37 1.14
9 1.00 25.608 1.00 1.464 1.00 1.707 1.00 0.171 4.23 3.42 2.49
10 1.00 4.392 1.00 8.536 1.00 1.707 1.00 0.171 5.48 1.68 1.07
11 1.00 25.608 1.00 8.536 1.00 0.293 1.00 0.171 5.20 3.04 2.24

Biotechnology and Bioengineering, Vol. 96, No. 3, February 15, 2007

12 1.00 4.392 1.00 1.464 1.00 0.293 1.00 0.171 5.28 2.48 1.51
13 1.00 25.608 1.00 8.536 1.00 1.707 1.00 0.029 4.68 4.78 3.32
14 1.00 4.392 1.00 1.464 1.00 1.707 1.00 0.029 7.36 2.38 1.96
15 1.00 25.608 1.00 1.464 1.00 0.293 1.00 0.029 6.64 2.20 1.30
16 1.00 4.392 1.00 8.536 1.00 0.293 1.00 0.029 5.00 1.65 1.08
17 0.00 15.00 0.00 5.00 0.00 1.00 0.00 0.10 8.28 3.63 2.31
18 1.414 30.00 0.00 5.00 0.00 1.00 0.00 0.10 7.52 4.25 2.92
19 1.414 0.00 0.00 5.00 0.00 1.00 0.00 0.10 7.96 2.05 1.53
20 0.00 15.00 1.414 10.00 0.00 1.00 0.00 0.10 6.80 2.87 2.01
21 0.00 15.00 1.414 0.00 0.00 1.00 0.00 0.10 8.44 1.99 1.52
22 0.00 15.00 0.00 5.00 1.414 2.00 0.00 0.10 5.68 3.07 2.13
23 0.00 15.00 0.00 5.00 1.414 0.00 0.00 0.10 6.56 1.87 1.35
24 0.00 15.00 0.00 5.00 0.00 1.00 1.414 0.20 8.32 3.51 2.47
25 0.00 15.00 0.00 5.00 0.00 1.00 1.414 0.00 7.32 3.05 2.48
26 0.00 15.00 0.00 5.00 0.00 1.00 0.00 0.10 8.56 3.35 2.28
27 0.00 15.00 0.00 5.00 0.00 1.00 0.00 0.10 9.36 3.11 2.28
28 0.00 15.00 0.00 5.00 0.00 1.00 0.00 0.10 9.40 3.10 2.27
29 0.00 15.00 0.00 5.00 0.00 1.00 0.00 0.10 8.68 3.05 2.23
30 0.00 15.00 0.00 5.00 0.00 1.00 0.00 0.10 9.12 3.09 2.20
a
Actual level in (g/L).

DOI 10.1002/bit
analysis of experiment was done by self-developed software.
The basis of the functioning of this software is the response
surface method (RSM) as suggested by Myers (1971) in
which matrix operation was the main feature to find the
coefficients of the regression equation. Optimization was
done by using canonical method suggested by Myers (1971)
and Khuri and Cornell (1989).
The experimental design employed for the present study
was a four-variables (five levels of each variable), second
order central composite design with five replications at the
center points (0,0,0,0) in coded levels of variables (1.414,
1, 0, 1, and 1.414) (Akhnazarova and Kafarov, 1982). The
four independent variables were X1 (concentration of
sodium chloride), X2 (concentration of magnesium sulfate),
X3 (concentration of sodium nitrate), and X4 (concentration
of dipotassium hydrogen phosphate).
The experimental design in the actual (X) and coded (x)
levels of variables is shown in Table I. The response
functions (Yijk), that is, yield of biomass, PB content and PE
production in the culture was approximated by a second-
degree polynomial (Eq. 4) with linear, quadratic, and
interaction effects (in coded level of variables) using the
method of least squares (Little and Hills, 1978).
X
n X
n X
n
yijk ¼ b0 þ b i xi þ bij xi xj þ 2ijk
i¼1 i¼1; ij j¼1; ij
The number of variables is denoted by n and j while k and i
are integers. The coefficients of the polynomials are
represented by b0, bi and bij, and 2ijk is the random error;
when i < j, bij represents the interaction effects of the
variables xi and xj. The response surface graphs were
obtained from the regression equations in actual level of
variables, keeping the response function on the Z axis with X
and Y axes representing the two independent variables while
keeping the other variables (third and fourth) constant at
their center (corresponding to 0 level in coded level) points.
The detailed analysis of variance (ANOVA) was conducted
in coded level of variables to know the effects of individual
variables. Stepwise deletion of individual non-significant
( P < 0.10) terms was conducted followed by recalculation of
the coefficients of the regression equation, to arrive at the
final regression equation in coded level which is later
converted to actual level of variables.
Optimization
Optimization was conducted by employing canonical
analysis (Khuri and Cornell, 1989; Myers, 1971) wherein
the levels of the variables (x1, x2, x3, x4) (within the
experimental range) were determined to obtain the
maximum yield of biomass, PB and PE production in-
dividually. Optimization of the response functions consists
of the translation of the response function (yk) from the
origin to the stationary points (Myers, 1971). Then the
response function was expressed in terms of the new
variables, the axes of which correspond to the principal axes
of the contour system. Further the roots (l1, l2, l3, l4) of
the auxiliary equation (l2l þ 1 ¼ 0) were calculated
initially to know the nature of optimum. The response
function is maximum if all the roots have negative values
and minimum if all the roots have positive values. If some of
the roots have positive values and some of them have
negative values, then it is the situation of a saddle point
(Myers, 1971; Bhattacharya and Prakash, 1994; Sarada et al.,
2002). Finally, the optimum levels of the variables were
obtained by superimposing the contours.
Results and Discussion
The present results on PB composition indicated PE as a
major component (70%) followed by R-PC (20%) and APC
(10%). PE is the major component among all the PB in
P. purpureum (Gnatt and Lipschultz, 1972, 1974). Fuentes
et al. (2000) studied the influence of dilution rates and
conditions of solar irradiation on biomass nutrient profiles
of the microalga P. cruentum using photobioreactor system,
and reported a maximum value of 3.6% PB and 2.8% PE
with the mean values of 2.5% and 2.0%, respectively. In the
(4) present study, optimization of the medium constituents for
increased production of PB and PE in batch mode was
achieved by employing RSM.
The experimental results on the effect of the four
independent variables (concentrations of sodium chloride,
magnesium sulfate, sodium nitrate, and dipotassium
hydrogen phosphate) on the response functions or targeted
parameters (yield of biomass, total PB and PE production)
are shown in Table I. The results on the ANOVA (in coded
level of variables) are shown in Table II for all the three
response functions. The response surfaces (Figs. 1 and 2) are
presented to aid in visualizing the effect of the four variables
on the response functions.
Yield of Biomass
The yield of biomass (Y1) varies between 3.4 and 9.4 g/L
(Table I). A high multiple correlation coefficient (r ¼ 0.97,
P  0.01) indicates the suitability of the second order
polynomial to predict the yield of the biomass in terms of
independent variables (Table II). Among the variables, the
total quadratic effect (significant at P  0.01) dominates
over the total linear effect ( P  0.05), but the effect of
interaction terms ( P  0.10) is marginal. Of the individual
variables, magnesium sulfate has the maximum linear
negative effect ( P  0.01) on yield of biomass whereas other
linear effects are negligible (such as sodium chloride). The
quadratic effects of all the variables are also significant
( P  0.01). Sodium nitrate possess the maximum negative
effect ( P  0.01) (Table II) on Y1. Among the various
interactions, negative effect of sodium chloride  sodium
Kathiresan et al.: Phycoerythrin Production 459
Biotechnology and Bioengineering. DOI 10.1002/bit
Table II. Analysis of variance (ANOVA) for the yield of biomass (Y1), phycobiliprotein content (Y2), and phycoerythrin production (Y3) in coded level of
variables.

Source of Coefficient of Coefficient of Coefficient of

variation polynomial for Y1 F-value polynomial for Y2 F-value polynomial for Y3 F-value
Constant 9.019 — 3.149 — 2.255 —
x1 0.068 0.229 (NS) 0.694 167.242c 0.433 367.975c
x2 0.523 13.722c 0.208 15.074c 0.114 25.597c
x3 0.223 2.484 (NS) 0.416 60.125c 0.318 198.237c
x4 0.098 0.479 (NS) 0.034 0.414 (NS) 0.0002 0.00008 (NS)
x21 0.729 12.401c 0.055 0.485 (NS) 0.011 0.101 (NS)
x22 0.789 14.5287c 0.305 15.126c 0.241 53.034c
x23 1.539 55.315c 0.285 13.209c 0.253 58.690c
x24 0.689 11.077c 0.120 2.324 (NS) 0.114 12.020c
x1x2 0.134 0.723 (NS) 0.380 40.164c 0.164 42.465c
x1x3 0.396 6.270b 0.309 26.514c 0.218 74.777c
x1x4 0.146 0.849 (NS) 0.285 22.592c 0.036 1.995 (NS)
x2x3 0.109 0.479 (NS) 0.204 11.547c 0.036 1.995 (NS)
x2x4 0.291 3.384a 0.070 1.363 (NS) 0.011 0.177 (NS)
x3x4 0.281 3.155a 0.086 2.069 (NS) 0.222 77.371c
TLE — 4.231b — 60.748c — 147.647c
TQE — 23.353c — 7.791c — 30.882c
TIE — 2.479a — 17.374c — 33.039c
Error — — — — — —
R 0.965c — 0.981c — 0.992c —
a
Significant at P  0.10.
b
Significant at P  0.05.
c
Significant at P  0.01.
NS ¼ Non-significant at P  0.10.
Variables: x1: Concentration of sodium chloride (g/L), x2: Concentration of magnesium sulfate (g/L), x3: Concentration of sodium nitrate (g/L), and x4:
Concentration of dipotassium hydrogen phosphate (g/L).
TLE: Total linear effect.
TQE: Total quadratic effect.
TIE: Total interaction effect.

Figure 1. Response surface for PB content as a function of sodium chloride and Figure 2. Pigment production as a function of sodium chloride and magnesium
magnesium sulfate when sodium nitrate and dipotassium hydrogen phosphate con- sulfate when sodium nitrate and dipotassium hydrogen phosphate concentration was
centration was kept constant at 1 and 0.1 g/L, respectively. kept constant at 1 and 0.1 g/L, respectively.

460 Biotechnology and Bioengineering, Vol. 96, No. 3, February 15, 2007

DOI 10.1002/bit
nitrate concentration has the predominant effect on yield of
biomass ( P  0.05). These significant interaction effects
mean that the effect of chloride on yield of biomass is
dependent on the level of nitrate used. The effect of
phosphate on yield of biomass is dependent on the level of
sulfate and nitrate to a low extent.
The linear effect of chloride, nitrate, and phosphate on
biomass yield is negligible but their quadratic effects are
significant ( P  0.01). This indicates that a low level of their
concentration has marginal effect but at a high level, they
tend to decrease the yield of biomass markedly. Therefore,
the level of all these three variables are required to be kept
at minimum levels in order to increase the yield of biomass
for the present situation, and in the experimental ranges
used. The optimum medium conditions of sodium chloride,
magnesium sulfate, sodium nitrate, and dipotassium
hydrogen phosphate are found to be 15.17, 4.87, 0.97,
and 0.09 g/L, respectively, for maximization of biomass yield
to 9.12 g/L at 14 days of growth, whereas in control (ASW)
medium, the biomass is found to be 4.8 g/L. The major
difference between optimized medium constituents for high
biomass yield and ASW medium is in the higher con-
centration of sodium chloride in ASW medium. The ASW
medium may lead to low biomass yield of Porphyridium due
to high sodium chloride concentration as reported for
Chlorella emersonii (Tel-Or, 1980).
Total Phycobiliprotein Content
The total PB content (Y2) shows a wide variation (1.64–
4.78%) due to the variations in experimental conditions
(Table I). Total PB content in the culture shows a high
multiple correlation coefficient (r ¼ 0.98, P  0.01), which
indicates the suitability of the second order polynomial to
predict the Y2 values in terms of these four independent
variables (Table II). Among these variables, total linear effect
dominates over the interaction and quadratic effects
( P  0.01); the linear positive effect of sodium chloride
(Fig. 1) is maximum followed by the linear positive effect of
sodium nitrate and magnesium sulfate. Low effects of
individual quadratic terms are observed for the chloride and
phosphate concentrations whereas for sulfate and nitrate
high negative effects ( P  0.01) are observed. It shows that
high concentrations of sulfate and nitrate suppress the PB
production.
Phycoerythrin Production
The PE production (Y3) shows about threefold variation
(1.07–3.32%) when the variables are changed (Table I). The
second order polynomial is suitable (r ¼ 0.99, P  0.01) to
predict the Y3 values in terms of the four variables (Table II).
The total linear effect ( P  0.01) dominates over the total
interaction effect and the total quadratic effects. The linear
positive effect of sodium chloride (Fig. 2) is maximum
followed by the linear positive effect of sodium nitrate and
magnesium sulfate. Low effect of individual quadratic term
is observed for the chloride concentrations whereas for
sulfate and nitrate high negative effects ( P  0.01) are ob-
served. The linear and quadratic effects of nitrate are positive
and negative, respectively, such that it shows an overall
curvilinear effect. PE production was not affected by sodium
chloride concentrations whereas sodium nitrate and mag-
nesium sulfate at higher concentrations decrease the PE
production. Although there are several reports on Porphyr-
idium, they are limited to exopolysaccharide production
(Adda et al., 1986), PE and exopolysaccharide under
mixotrophic conditions (Fabregas et al., 1999), light quality
on exopolysaccharide (Friedman et al., 1991; Tao et al.,
2004) but none on the effect of media constituents on PE.
The total PB as well as PE obtained at optimal conditions is
higher than the reported value of 1.66% (w/w PE) (Bermejo
et al., 2002). However, the yields are not comparable with
the reported values as they differ in the mode of cultivation
such as batch versus semi-continuous with different renewal
rates (Fabregas et al., 1998), outdoor tubular photobior-
eactors (Fuentes et al., 2000; Bermejo et al., 2002) versus
glass tubes (Fabregas et al., 1999), and autotrophic versus
mixotrophic medium (Fabregas et al., 1999).
Optimization
The process of optimization (maximization) of the yield of
biomass (Y1), total PB content (Y2), and PE production (Y3)
has been conducted separately. The roots (l1, l2, l3, l4) of
the auxiliary equation have been determined and are shown
in Table III. The root of the auxiliary equation for biomass
yield indicates maximization of response function, whereas
for PB and PE production, these are the cases of saddle
point. Hence, the canonical test has been employed, and the
optimum conditions (within the range of the present
experimental variables) in coded and actual level of variables
are obtained (Table III). Accordingly, the maximum value
for the yield of biomass is 9.12 g/L, maximum PB content is
4.78% and the maximum PE production is 3.30%. The
predicted optimized levels for all the four variables as shown
in Table III for production of biomass, PB, and PE were
confirmed experimentally.
Figure 3 shows the superimposed contour plot for
optimization (maximization) of Y2 and Y3. The effect of x1
terms on Y2 and Y3 is positive. Optimum situation is in
between 26 and 29.6 g/L, that is, at an average of 27.8 g/L for
x1. The effect of x3 on Y2 and Y3 is complex as there are both
positive and negative effects, and hence, x3 concentration is
averaged at 1.573 g/L. Contour plots against magnesium
sulfate and dipotassium hydrogen phosphate have been
separately constructed and later superimposed (Fig. 3) to
know the levels of these two variables. In the superimposed
graph, we found the optimum conditions.
Here, we have chosen a value of PB content to be more
than 5.1% and PE production to be more than 3.4% (from
Fig. 3). The optimum condition (marked as double hatched
Kathiresan et al.: Phycoerythrin Production 461
Biotechnology and Bioengineering. DOI 10.1002/bit
Table III. Results of the optimization study.

Parameters Biomass Y1 Phycobiliprotein Y2 Phycoerythrin Y3

Roots of the auxiliary equation l1 0.5329 0.2947 0.1467
l2 0.7061 0.0122 0.0450
l3 0.8933 0.1941 0.2365
l4 1.6125 0.5048 0.3403
Optimum conditions in coded level x1 0.0157 1.0425 1.3778
x2 0.3544 0.0647 0.3127
x3 0.0394 0.7934 0.8275
x4 0.1358 0.9378 0.3403
Optimum conditions in actual level X1 15.17 g/L 26.06 g/L 29.62 g/L
X2 3.75 g/L 5.23 g/L 6.11 g/L
X3 0.97 g/L 1.56 g/L 1.59 g/L
X4 0.09 g/L 0.03 g/L 0.076 g/L

X1, X2, X3, and X4 are same as mentioned in Table II.

zone in Fig. 3) is, therefore, sodium chloride concentration
is 27.8 g/L and that of sodium nitrate to be 1.573 g/L.
For the commercialization of micro algal products, two
major aspects are considered as important. One is the
effect of environmental conditions such as pH, aeration,
illumination, agitation etc, while another is the selection of
a suitable nutrient medium (Sarra et al., 1993). It is well
known that culture medium affects not only the cell
productivity but also the product composition and yield
(Gong and Chen, 1997). In the present study the aspect of
environmental conditions as mentioned above have been
kept unchanged and only optimization of the media cons-
tituents has been achieved through RSM. Though low
concentrations of sodium chloride and magnesium sulfate
are required for biomass growth, relatively high concentra-
tion of these salts increases both PB and PE. When optimum
conditions for growth and metabolite production are found
to be different, maximum yield of both can be achieved
either by two-stage cultivation (Sarada et al., 2002) or
providing optimum condition for growth first followed by
supplementation of required nutrients in the later stage.
Modifications of cultural conditions for growth and
metabolite production are reported for algae. Harker
et al., 1996 reported growth of Haematococcus culture in
a closed photobioreactor and astaxanthin accumulation (a
ketocarotenoid) in thin columns with addition of sodium
chloride under high light intensity. Improved growth under
low salinity has also been reported for Isochrysis and
Tetraselmis (Laing and Utting, 1980). Since PB production is
not growth linked and it acts as a nitrogen sink (nitrogen

Figure 3. Superimposed contour plots for optimization of phycobiliprotein content (—) and phycoerythrin production (- - -).

462 Biotechnology and Bioengineering, Vol. 96, No. 3, February 15, 2007

DOI 10.1002/bit
 reserve), and whenever there is a limitation of nitrogen, PB
will be degraded and vice versa (Andria et al., 1999). The
present results also show PB production to be affected by
nitrate to a greater extent while growth is affected in
presence of high sodium chloride concentration. Therefore,
Porphyridium can be grown in low salt medium (sodium
chloride 15.17 g/L and magnesium sulfate 4.87 g/L) for
maximum growth, and high salt medium for enhancement
of PB and PE production (sodium chloride 29.62 g/L and
magnesium sulfate 6.11 g/L).
It is concluded that PE production depends on concen-
tration of sodium chloride, sodium nitrate, and magnesium
sulfate as well as dipotassium hydrogen phosphate. Among
these variables, sodium chloride possesses the maximum
effect. The second order polynomials can be used to predict
yield of biomass, PB, and PE production in terms of these
four variables. The superimposition of response curves is
used to predict the values of these variables when the PB and
PE production can be more than 5.1% and 3.4% (w/w),
respectively. This information has more relevance to culti-
vation of algae in scaled up systems. Production of PB may
be improved by adopting the method mentioned in the
present research.
The authors thank Dr. V. Prakash, Director, CFTRI for his support to
carry out this work and acknowledge the CSIR and DBT for their
financial support.
References
Adda M, Merchuk CJ, Arad SM. 1986. Effect of nitrate on growth and
production of cell wall polysaccharides by the unicellular red alga
Porphyridium. Biomass 10:131–140.
Akhnazarova S, Kafarov V. 1982. Experiment optimization in chemistry and
chemical engineering. Moscow: MIR Publishers. pp 151–240.
Andria JR, Vergara JJ, Perez-Liarens JL. 1999. Biochemical response and
photosynthetic performance of Gracilaria sp. (Rhodophyta) from
Cadiz, Spain, cultured under different inorganic carbon and nitrogen
levels. Eur J Phycol 34:497–504.
Arad SM, Yaron A. 1992. Natural pigments from red microalga for use in
foods and cosmetics. Trends Food Sci Technol 3:92–97.
Bermejo R, Alarez-Pez JM, Acien Fernandez FG, Molina E. 2002. Recovery
of B-phycoerythrin from the microalga Porphyridium cruentum.
J Biotechnol 93:73–85.
Bhattacharya S, Prakash M. 1994. Extrusion cooking of blends of rice and
chickpea flour: A response surface analysis. J Food Engg 21:315–330.
Borowitzka MA. 1994. Products from algae. In: Phang SM, Lee YK,
Borowitzka MA, Whitton BA, editors. Algal biotechnology in the
Asia-Pacific region. Kuala Lumpur: Institute of Advanced studies.
pp 5–15.
Dainippon Ink and Chemicals, 1983. Antitumoral agents containing phy-
cobilin-also used to treat ulcers and hemorrhoidal bleeding. Tokyo,
Kenkyukai patent JP 58065216.
Fabregas J, Garcia D, Morales E, Dominguez A, Othero A. 1998. Renewal
rate of semicontinuous cultures of the microalga Porphyridium cruen-
tum modifies phycoerythrin, exopolysaccharide and fatty acid produc-
tivity. J Fermen Bioengg 86:477–481.
View publication stats
Fabregas J, Garcia D, Morales E, Lamela T, Othero A. 1999. Mixotrophic
production of phycoerythrin and exopolysaccharide by the microalga
Porphyridium cruentum. Cryptogamie Algol 20:89–94.
Friedman O, Dubinsky A, Arad SM. 1991. Effect of light intensity on growth
and polysaccharide production in red and blue-green Rhodophyta
unicells. Biores Technol 38:105–110.
Fuentes RM, Fernandez AG, Prerez SJ, Guerrero GJ. 2000. Biomass, nutrient
profiles of the microalga Porphyridium purpureum. Food Chem 70:345–
353.
Geresh S, Adin I, Yarmolinsky E, Karapasas M. 2002. Characterization of the
extracellular polysaccharide of Porphyridium sp: Molecular weight
determination and rheological properties. Carbohydrate polymers
50:183–189.
Gnatt E, Lipschultz CA. 1972. Phycobilisomes of Porphyridium cruentum I.
Isolation. J Cell Biol 54:313–324.
Gnatt E, Lipschultz CA. 1974. Phycobilisomes of Porphyridium cruentum:
Pigment analysis. Biochemistry 13:2960–2966.
Gong XD, Chen F. 1997. Optimization of culture medium for growth of
Haematococcus pluvialis. J Appl Phycol 9:437–444.
Harker M, Tsavalos AJ, Young AJ. 1996. Autotrophic growth and carote-
noid production of Haematococcus pluvialis in a 30 liter air-lift photo-
bioreactor. J Ferment Bioeng 82:113–118.
Henrikson R. 1989. Earth food Spirulina. New clinical research reveals
health benefits. Laguna Beach, California: Ronore Enterprises. p 69.
Khuri AI, Cornell JA. 1989. Response surfaces: Designs and analyses.
New York: Marcel Dekker. pp 19–69.
Laing I, Utting SD. 1980. The influence of salinity on the production of two
commercially important unicellular marine algae. Aquaculture 21:
79–86.
Little TM, Hills FJ. 1978. Agricultural experimentation: Design and analysis.
New York: John Wiley & Sons. pp 247–266.
Myers RH. 1971. Response Surface Methodology. Boston: Allyn and Bacon.
pp 61–106.
Ooijikass LP, Wilinson EC, Tramper J, Buitelaar RM. 1999. Medium
optimization for spore production of Coniothyrium minitans using
statistically based experimental designs. World J Microbiol Biotechnol
14:185–190.
Rao RS, Sarada R, Ravishankar GA. 1996. Phycocyanin, a new elicitor for
capsaicin and anthocyanin accumulation in plant cell cultures. Appl
Microbiol Biotechnol 46:619–621.
Reddy CM, Vadiraja BB, Kiranmai G, Reddy MN, Reddanna P, Madyastha
KM. 2000. Selective inhibition of cyclooxygenase-2 by C-phycocyanin,
a biliprotein from Spirulina platensis. Biochem Biophy Res Comm 277:
599–603.
Sarada R, Bhattacharya S, Ravishankar GA. 2002. Optimization of culture
conditions for growth of the green alga Haematococcus pluvialis. World
J Microbiol Biotechnol 18:517–521.
Sarra M, Redin I, Ochin F, Godia F, Casas C. 1993. Application of factorial
design to the optimization of media composition in batch cultures of
Streptomyces lividans TK21 producing a hybrid antibiotic. Biotechnol
Lett 15:559–564.
Tandeau N, Houmard J. 1988. Complementary chromatic adaptation:
Physiological conditions and action spectra. Methods Enzymol 167:
318–328.
Tao Y, Barnett SM. 2004. Effect of light quality on production of
extracellular polysaccharides and growth rate of Porphyridium cruen-
tum. Biochem Eng J 19:251–258.
Tel-Or E. 1980. Response of N2-fixing cyanobacteria to salt. Appl Environ
Microbiol 40:689–693.
Vadiraja BB, Madyastha KM. 2000. C-Phycocyanin: A potent peroxyl
radical scavenger in vivo and in vitro. Biochem Biophy Res Comm
275:20–25.
Vohra A, Satyanarayana T. 2002. Statistical optimization of medium
components by response surface methodology to enhance phytase
production by Picha anomala. Proc Biochem 37: 999–1004.
Kathiresan et al.: Phycoerythrin Production 463
Biotechnology and Bioengineering. DOI 10.1002/bit