Y Huet et al 2013

ORIGINAL RESEARCH PAPER

Production and secretion of a heterologous protein by turniphairy roots with superiority over tobacco hairy roots

Yoann Huet • Jean-Pierre Ele Ekouna •

Aurore Caron • Katiba Mezreb •

Michele Boitel-Conti • Francois Guerineau

Received: 16 July 2013 / Accepted: 15 August 2013

� Springer Science+Business Media Dordrecht 2013

Abstract A fully contained and efficient heterolo-

gous protein production system was designed using

Brassica rapa rapa (turnip) hairy roots. Two expres-

sion cassettes containing a cauliflower mosaic virus

(CaMV) 35S promoter with a duplicated enhancer

region, an Arabidopsis thaliana sequence encoding a

signal peptide and the CaMV polyadenylation signal

were constructed. One cassette was used to express the

green fluorescent protein (GFP)-encoding gene in

hairy roots grown in flasks. A stable and fast-growing

hairy root line secreted GFP at [120 mg/l culture

medium. GFP represented 60 % of the total soluble

proteins in the culture medium. Turnip hairy roots

retained sustainable growth and stable GFP production

over 3 years. These results were superior to those

obtained using tobacco hairy roots.

Keywords Agrobacterium rhizogenes �Brassica � GFP � Green fluorescent protein �Hairy roots � Heterologous protein production �Protein � Secretion � Tobacco hairy roots

Introduction

Proteins have been developed as therapeutic agents

used to treat a wide range of viral, autoimmune, and

metabolic diseases and cancer (Brekke and Sandlie

2003; Stern and Herrmann 2005). In addition, amino

acid sequences derived from a number of pathogens

have been or could be developed as vaccines for either

animals or humans (Plotkin 2005; Pujol et al. 2007;

Patel and Heldens 2009). To date, most proteins for

therapeutic use have been produced in Escherichia

coli or in mammalian cells such as Chinese hamster

ovary cells. Mammalian cells require complex cultureYoann Huet and Jean-Pierre Ele Ekouna have equally

contributed to this work.

Y. Huet � J.-P. E. Ekouna � A. Caron �K. Mezreb � M. Boitel-Conti � F. Guerineau (&)

Biologie des Plantes et Innovation (BioPI), Universite de

Picardie Jules Verne, 33 rue St Leu, 80039 Amiens,

France

e-mail: [email protected]

Y. Huet


J.-P. E. Ekouna


A. Caron


K. Mezreb


M. Boitel-Conti


Present Address:

Y. Huet

Root Lines Technology, CRRBM, 33 rue St Leu,

80039 Amiens, France

123

Biotechnol Lett

DOI 10.1007/s10529-013-1335-y

media, often containing animal-derived compounds,

and the production of proteins can be plagued by

instability in some lines (Kim et al. 2011a). The

production costs are therefore high and the risk of

contamination by animal-derived pathogens cannot be

ruled out (Grillberger et al. 2009).

Plants are considered as valuable alternatives to

animal or microbial cells for the production of proteins

(Fischer et al. 2004; Ko and Koprowski 2005; Ma et al.

2005; Doran 2013). They are cheaper to grow and are

phylogenetically distant from humans, ruling out the

risk of virus or prion cross-transmission. A number of

strategies have been explored for the production of

proteins in plants. Many proteins have been produced in

leaves, which make up the most abundant biomass part

of annual plants. The yield has typically amounted to

1–2 % of total soluble proteins (Streatfield 2007).

Production in leaves is cheap and can easily be scaled

up by growing plants in open fields but such production

systems have raised a number of concerns. Among the

most important ones for the regulatory agencies is the

risk of leakage of transgenic pollen or seeds into the

environment. Moreover, plant biomass yield and quality

is highly dependent on weather conditions, phytopath-

ogen breakouts and pest attacks. Plants over-producing

active pharmaceuticals also represent a potential risk for

herbivorous fauna and possibly for humans, if plant

species of agro-industrial interest are involved. In such a

case, a tight control of the transgenic cultures along with

complex regulatory processes is mandatory to avoid any

cross-contamination with the human food supply chain.

A more technical limitation is the difficulty of extracting

proteins from plant cells, which are surrounded by a

rigid cell wall. Although using the protein from seed oil

bodies circumvented the need to break leaf cells and

provided faster protein purification, plants were still

required to grow to maturity (Boothe et al. 1997).

In vitro cultures in bioreactors are more expensive

to implement than field-grown plants but can be

developed under much more controlled conditions and

have none of the perceived environmental risks

associated with transgenic plants. Heterologous pro-

teins have successfully been produced in a number of

plant-based in vitro systems (Hellwig et al. 2004)

including cell suspensions, hairy roots, mosses and

microalgae (Xu et al. 2012; Ono and Tian 2011). The

secretion of proteins into the culture medium has also

been developed as an easier alternative to cell lysis for

protein recovery (Kim et al. 2011b).

Since the first report on the production of a murine

IgG1 by tobacco hairy roots (Wongsamuth and Doran

1997), several advantages of this system for the

production of recombinant protein have been high-

lighted (Doran 2013). In comparison with cultured

plant cells, hairy roots can maintain a much more

stable transgene expression (Sharp and Doran 2001;

Peebles et al. 2007) and can be cultured indefinitely on

hormone-free medium, the composition of which

remains extremely simple (minerals, vitamins and

sugar). They are also less sensitive to mechanical

damage and the biomass can be easily separated from

the culture medium. In contrast to mosses and

microalgae, growth of the non-photosynthetic hairy

roots does not require complex photo-bioreactors.

Their ability to secrete the protein of interest in the

culture medium (rhizosecretion) can greatly decrease

the complexity and associated costs of the downstream

purification steps. Moreover, rhizosecretion allows a

continuous mode of production in which the root

biomass is not destroyed after each production cycle.

As a model plant organism, tobacco has been largely

used for the production of recombinant proteins

through the generation of hairy root cultures, and

secretion of heterologous proteins has been achieved

(Medina-Bolivar and Cramer 2004; Nopo et al. 2012).

However, and although suitable for many academic

research purposes, the reported yield remained rather

low to meet the expectations of the industrial sector.

The work presented here aimed at developing a

hairy root platform for the production of heterologous

protein. In order to remain relevant in the context of

pharmaceutical bioproduction, this platform would

match the following criteria: Firstly, the use of a

Generally Recognized As Safe (GRAS) edible plant

species in order to reduce as much as possible the risk

of finding potentially hazardous secondary metabo-

lites in the system (e.g. alkaloids from the Solanaceae

plant family); secondly, the ability to secrete and

accumulate the recombinant protein into the culture

medium, while maintaining a low amount of endog-

enous secreted proteins; thirdly, a long-term (i.e.

several months) viability of the root cultures to allow

for the continuous production of the recombinant

protein. To achieve this, a number of food plant

species have been inoculated with Agrobacterium

rhizogenes transformed with a gfp construct. These

preliminary experiments singled out turnip (Brassica

rapa rapa) as a more efficient alternative to tobacco

Biotechnol Lett

123

for the production of a secreted heterologous protein

by hairy roots.

Materials and methods

Materials

Escherichia coli strain JM101 and A. rhizogenes strain

TR7 were used for cloning and plant transformation,

respectively. B. rapa L. subsp. rapa cv. ‘‘des vertus

marteau’’ was used in transformation experiments.

Restriction enzymes were purchased from New England

Biolabs, T4 DNA ligase from Promega, Taq DNA

polymerase from ATGC. Plant culture media compo-

nents and cefotaxime were obtained from Duchefa,

other antibiotics from Sigma-Aldrich. Oligonucleotides

were synthesized by Invitrogen. The trypsin gene was

synthesized by DNA2.0 Inc. DNA sequencing was

carried out by Eurofins-MWG. Bradford reagent, c-

globulin standard and Coomassie Blue G250 were

purchased from Bio-Rad and purified EGFP from

BioVision. Protein molecular weight markers were

Precision Plus from Bio-Rad or ColorBurst from Sigma.

Cloning experiments

Plasmids were extracted using a Wizard kit from

Promega. Bacterial transformations were performed

as in Nishimura et al. (1990) for E. coli and by

electroporation for A. rhizogenes. The signal peptide

(SP) coding sequence from the Arabidopsis

At1g69940 pme gene was amplified from the RIKEN

cDNA clone pda12853 (obtained from RIKEN) using

oligonucleotides 50-GAGAAGCTTAAAACAACAA

TGGGATACACAAATGTGTCC-30 and 50-GAGGT

CGACGACCATGGGATCGGCGAACACCATCGG

TG-30. The PCR product was digested by HindIII and

SalI and inserted into the same sites of the expression

cassette pJIT163 (Guerineau 1995), to give pPE41.

Annealed oligonucleotides 50-CATGCATCATCAT-

CATCACCACCC-30 and 50-CATGGGGTGGTGAT-

GATGATGATG-30 were inserted into the NcoI site of

pPE41, giving pPE45. The integrity of the cloning

junctions of the two expression cassettes was verified

by nucleotide sequencing. The egfp (enhanced gfp)

coding sequence was extracted from pEGFP (Clon-

tech) using NcoI and EcoRI and inserted into pPE45,

giving pPE47. The 35S promoter-SP-His tag-egfp-

CaMV polyA fusion was recovered from pPE47 as an

Asp718-BglII fragment and inserted into the Asp718

and BamHI sites of the binary vector pRD400 (Datla

et al. 1992), to give pRP49.

Plant transformation and hairy root cultivation

Agrobacterium rhizogenes was grown in MGL med-

ium (Baranski et al. 2006) devoid of biotin. Turnip (B.

rapa L. var. rapa cv. des vertus marteau) seeds were

sown in 180 ml plastic pots containing approx. 20 ml

0.5xMS agar (0.8 %) medium supplemented with 1 %

(w/v) sucrose and 2 mM MES at pH 5.7. After 10 days

growth under dim light at 25 �C, the elongated stems

were punctured at three positions using a 26G needle

dipped in an A. rhizogenes suspension. Only one root

was excised from each infection site and placed on

0.5xMS medium supplemented with 3 % (w/v)

sucrose, 2 mM MES, 300 lg cefotaxime/ml and

50 lg kanamycin/ml. Root lines were screened for

egfp expression by observation under a fluorescence

microscope (Nikon) fitted with the Semrock GFP-

3035B filter (Exciter 472/30; Emitter 520/35). Roots

from fast-growing lines were then grown in liquid

Gamborg B5 medium supplemented with 3 % (w/v)

sucrose and 300 lg cefotaxime/ml. For protein pro-

duction, roots were grown in 250 ml Erlenmeyer flasks

containing 100 ml Gamborg B5 medium supple-

mented with 3 % (w/v) sucrose, with shaking at

120 rpm and at 23 �C. Tobacco transformation was

performed as follows. Nicotiana tabacum cv. SR1

seeds were surface-sterilized and germinated on half

strength (0.5x) MS solid medium supplemented with

1 % (w/v) sucrose. Leaves of 4-week old plants were

cut off and transferred on 0.5xMS solid medium at pH

5.7 supplemented with 3 % (w/v) sucrose. The main

central veins were longitudinally sectioned and an

Agrobacterium suspension was applied onto the

injured tissue. After 4 days, leaves were transferred

onto solid 0.5xMS medium supplemented with 3 %

(w/v) sucrose and 300 lg cefotaxime/ml. Hairy roots

started to emerge from the middle veins after 4 weeks.

They were then grown according to the procedure

previously stated for turnip roots.

Protein analysis

Total proteins were assayed by the Bradford method

using the Bio-Rad Bradford reagent and c-globulin as a

Biotechnol Lett

123

standard. Unless stated otherwise, proteins were sepa-

rated by electrophoresis on 12 % (v/v) PAGE, stained

with Coomassie Brilliant Blue G250 as in Kang et al.

(2002). Proteins in culture media were concentrated on

Millipore Microcon YM-10 columns. GFP was quan-

tified on a Bio-Rad VersaFluor fluorimeter using

EX490/10 and EM510/10 filters at the medium gain

setting. The instrument was calibrated using a 10 mg

EGFP/l standard. This concentration was within the

linear range of a standard curve. For the quantification,

the culture media was diluted 10 or 20 times in 50 mM

Tris/HCl pH 7.5 and the fluorescence of the solution was

measured. For western blotting analysis, root extracts

were obtained by grinding 200 mg roots in 600 ll

50 mM sodium phosphate buffer, pH 6.8, at 4 �C. After

centrifugation at 4 �C for 10 min at 10,000 9 g, 1.5 lg

proteins from the supernatant was separated by SDS-

PAGE in an mini-Protean TGX precast gel (Bio-Rad).

The proteins were transferred to a nitrocellulose mem-

brane using a Trans-Blot Turbo Transfer instrument and

pack (Bio-Rad). The membrane was blocked and

incubated overnight at 4 �C with a monoclonal anti-

polyhistidine antibody-peroxidase conjugate (Sigma)

diluted at 1/1,000. The signal was revealed in Tris

60 mM pH 7.5, 0.7 mg/ml 3,30-diaminobenzidine,

1.6 mg H2O2/ml. After western blotting, the proteins

on the membrane were stained for 5 min in a 0.1 % (w/

v) Ponceau S in 5 % (v/v) acetic acid solution. For

Edman protein sequencing, 500 ll culture medium was

concentrated on a Millipore Microcon YM10 column.

The proteins were separated by SDS-PAGE and the gel

was stained with Coomassie Brilliant Blue G250. The

bands containing the GFP were cut out and the protein

was extracted and blotted onto a polyvinylidene difluo-

ride membrane with the ProSorb system (Applied

Biosystems). The N-terminal sequence of the protein

on the blot was determined by Edman degradation using

a Procise P494 automated protein sequencer (Applied

Biosystems). The identification of EGFP by LC–MS–

MS was performed as follows. The 27 kDa band cut off

from a Coomassie-stained SDS-PAGE gel was reduced

at 50 �C for 1 h in 10 mM DTT and alkylated for 1 h in

55 mM iodoacetamide, in the dark. Fragments were

washed several times with water and (NH4)2CO3,

dehydrated with acetonitrile and dried. Trypsin diges-

tion was performed overnight using a MultiPROBE II

system (Perkin–Elmer). Peptides were extracted by

soaking the gel piece for 2 9 15 min in an acetonitrile/

water solution, 15 min in a 1 % (v/v) formic acid and in

100 % acetonitrile. Peptide extracts were dried and

dissolved in chromatography starting buffer made of

3 % (v/v) acetonitrile and 0.1 % formic acid. Peptides

were analysed using a nano-LC1200 system coupled to a

6340 ion trap mass spectrometer equipped with an

HPLC-chip cube interface (Agilent Technologies). The

tandem mass spectrometry peak list was extracted using

the DataAnalysis program (v 3.4, Bruker Daltonic) and

referred to the protein database using the Mascot

Daemon (v 2.1.3) search engine. The search was

performed with a maximum of one missed cleavage,

with no fixed modification and with variable modifica-

tions for carbamidomethyl and oxidation of methio-

nines. The identification from the tandem mass

spectrometry spectrum was performed with a mass

tolerance of 1.6 Da for precursor ions and 0.8 for MS/

MS fragments.

Results

Expression cassettes for the secretion of proteins

by plant cells

The choice of the signal peptide sequence of the

Arabidopsis At1g69940 pectin methylesterase has been

made on predictions using the SignalP 3.0 analysis tool

(Bendtsen et al. 2004). The probability of the presence

of a signal peptide in this protein was 0.998 and the

probability of its cleavage between amino acids 23 and

24 was 0.997 using the HMM prediction method. The

sequence encoding this signal peptide has been inserted

into the expression cassette pJIT163 (Guerineau 1995),

which contained a duplicated cauliflower mosaic virus

(CaMV) 35S promoter followed by a polylinker and the

CaMV polyadenylation signal, giving pPE41 (Fig. 1).

A His tag-coding sequence has been added upstream of

the polylinker of pPE41, giving pPE45 (Fig. 1). The

egfp coding sequence has been inserted into the pPE45

polylinker, in frame with the His tag-coding sequence

and the SP sequence. The protein was expected to be

synthesized as a fusion protein containing the signal

peptide and the His tag at its N-terminus.

Secretion of GFP by turnip and tobacco hairy roots

The 35S-SP-His tag-egfp-polyA fusion has been

inserted between the T-DNA borders of a binary

vector, giving pRP49. Turnip and tobacco plants have

Biotechnol Lett

123

been infected with pRP49-containing A. rhizogenes.

During a first screen, 38 tobacco root lines and 20

turnip root lines were selected. GFP was assayed in the

culture medium of those roots after 12 days cultiva-

tion. The results obtained with the top 10 producers

from each species are shown on Fig. 2. The fast

growing turnip line 2M1 was chosen for further

investigations. A strong fluorescence could be seen in

the roots under the microscope and some GFP was

present in intercellular spaces (Fig. 3). No fluores-

cence could be observed in wild-type roots. Roots

from the 2M1 line were grown in Erlenmeyer flasks in

order to monitor their growth and the concentration of

GFP in the medium at various times (Fig. 4). After the

A

B

Fig. 1 Expression cassettes for the secretion of proteins in plant

cells. a Plasmid maps. SP signal peptide, CaMV cauliflower

mosaic virus. The polyadenylation signal is located within the

EcoRI-BglII fragment (Guerineau et al. 1991). b Nucleotide

sequence encoding the Arabidopsis At1g69940 signal peptide

and the same sequence in plasmids pPE41 and pPE45. The

numbered amino acid sequences are given below the nucleotide

sequences ?1 indicates the predicted transcription start

nucleotide in the two plasmids. The NcoI sites are indicated in

bold characters and the SalI sites are underlined. Black triangles

indicate the predicted SP cleavage site

Biotechnol Lett

123

growth phase, that lasted for approx. 10 days, GFP

kept accumulating in the medium for two additional

weeks, up to 123 ± 4.7 lg/ml (n = 5) (Fig. 4).

Approx. 10 mg secreted GFP and 1.1 g of dry root

biomass were produced in each flask. The productivity

was 348 ± 49 lg GFP/gDW.day. A fast-growing root

line obtained from transformation with wild-type A.

rhizogenes TR7 was used as a control. The growth rate

of the control line was similar to that of the GFP-

producing line and no fluorescence could be detected

in the medium of wild-type root cultures (not shown).

A single band of approx. 27 kDa could be seen on

PAGE gels in the transgenic root culture medium but

not in the wild type hairy root culture medium

(Fig. 5a). The identity of GFP was confirmed by

LC–MS–MS: eleven peptides, identified following a

trypsin digest of the band extracted from the PAGE

gel, matched the Aequorea victoria GFP sequence,

ensuring 50 % coverage (Fig. 5b). Protein concentra-

tions from wild-type and GFP-producing line media

were 88 ± 4.5 lg/ml and 210 ± 6 lg/ml (n = 5)

respectively, indicating that GFP represented up to

60 % of the proteins found in the culture medium. The

selected GFP-expressing line has been maintained by

sub-culturing for over three years without any change

in both growth rate and GFP secretion.

To find out whether the 2M1 turnip hairy root GFP-

producing line could sustain production over culture

medium changes, the culture time in Erlenmeyer flasks

was extended to 100 days. The medium was replaced

Fig. 2 GFP secretion by ten

turnip (Br) and ten tobacco

(Nt) hairy root lines

expressing a 35S-SP-GFP

construct. The GFP in the

medium was assayed after

12 days. The lines were

selected from 20 and 38

independent transformants

respectively

Fig. 3 Light micrographs

of wild-type (1, 2) or GFP-

expressing (3, 4) turnip hairy

roots. 1, 3, visible light; 2, 4,

UV light. The arrowed inset

shows the presence of GFP

in the intercellular space

Biotechnol Lett

123

after 33 and 67 days. A 4.2 tobacco hairy root line

transformed with pRP49 was grown in parallel. GFP

concentration in the medium was also monitored. The

amount found in the culture medium of the tobacco

line reached 0.77 ± 0.11 lg/ml (n = 5) after 31 days

cultivation (Fig. 6). GFP production resumed in the

turnip and tobacco lines after the first medium change

but the production from the tobacco line collapsed

after 55 days due to root death. In contrast, the turnip

line kept on producing GFP after the second medium

change and until the end of the experiment (100 days).

At the end of the experiment, turnip roots had kept

their typical whitish color, indicating that they

remained healthy after 100 days whereas tobacco

roots were dark brown.

A western blotting experiment has been carried out

to assess whether the His tag was present on the GFP

protein. A conjugated antibody recognizing the His tag

has been used to probe proteins extracted from wild-

type roots, GFP-producing roots and proteins recov-

ered from the culture media of these roots. A protein

matching the expected size of the His-GFP (27.8 kDa)

has been detected in the GFP-producing roots but not

in the medium (Fig. 7a). Ponceau S staining of the

proteins on the membrane revealed that GFP in the

culture medium was smaller than the protein in the

root extract which had been stained by the anti-His

antibody (Fig. 7b). Edman sequencing revealed that

the amino acid sequence VSKGE was present at the N-

terminus of the GFP recovered from the culture

medium. This sequence corresponds to the five amino

acids following the first methionine of GFP.

Discussion

Agrobacterium rhizogenes-induced hairy roots are

fast-growing organs which are easy to grow in vitro

and do not require any light or an exogenous supply of

hormones. These features have led to much interest in

the use of hairy roots for academic or industrial

research, such as for the production of secondary

metabolites or heterologous proteins (Ono and Tian

2011). For example, hepatitis B surface antigen has

been produced in potato hairy roots at five times higher

than in the original transgenic plants (Sunil Kumar

et al. 2006). Hairy roots used for the production of

proteins originated mainly from solanaceous plants

(Gaume et al. 2003; Woods et al. 2008; De Guzman

et al. 2011). The production yields of the E. coli

B-subunit heat-labile toxin antigen in hairy roots were

approx. seven times higher in Petunia and tobacco

than in tomato (De Guzman et al. 2011), indicating a

high variability in the efficiency of various species for

the production of heterologous proteins.

As a preliminary work towards finding a GRAS

plant suitable for protein secretion by hairy roots, a

number of food plant species have been challenged

with A. rhizogenes harboring a 35S-egfp construct in

which the egfp coding sequence has been fused to the

signal peptide coding sequence from an Arabidopsis

pme gene (Fig. 1). This study highlighted turnip (B. r.

rapa) as being susceptible to A. rhizogenes, producing

fast-growing and persistent hairy roots and secreting

GFP into the culture medium. In particular, the

difference between turnip and tobacco transgenic

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 5 10 15 20 25 30 35

days

GFP(mg/l)

0

20

40

60

80

100

120

140

160

180

200

DW(g)

DW GFP

Fig. 4 Growth and GFP

production by turnip hairy

roots grown in flasks. DW

root dry weight per flask.

Means and confidence

intervals were calculated

from five independent

experiments

Biotechnol Lett

123

lines was obvious at an early stage (Fig. 2). Trans-

genic turnip roots were strongly fluorescent and GFP

could be seen in the intercellular spaces indicating that

the signal peptide was efficient at directing the

passenger protein to the apoplastic space (Fig. 3).

The GFP concentration in the medium of Brassica

roots reached over 120 mg/l after 4 weeks (Fig. 4).

The amount of secreted GFP was compared to the one

obtained from tobacco hairy roots. After 30 days, GFP

from turnip root cultures was present in the medium at

levels over two orders of magnitude higher than from

tobacco roots (Fig. 6). GFP production from tobacco

roots was close to that previously indicated in another

study using a similar construct (Medina-Bolivar and

Cramer 2004), which measured secreted GFP concen-

tration at 800 lg/l in bioreactors. In a later study, in

which a mas-derived promoter has been used to drive

the transcription of the egfp gene in tobacco, a

concentration of 27 lg/l has been obtained in the

medium of flask grown hairy roots (Nopo et al. 2012).

The differences in the yields of recovered GFP from

Nicotiana and Brassica hairy roots may originate from

various parameters such as the level of expression,

secretion efficiency, extent of diffusion of the protein

into the medium or the level of extra-cellular protease

activity. The amount of protease secreted by plant

cells is highly species-dependent (Plasson et al. 2009).

This can result in large variations in the amount of

heterologous protein secreted into the medium. For

example, the amount of recombinant human granulo-

cyte–macrophage colony stimulating factor recovered

from the medium of rice cells was 1,000 times higher

than the amount recovered from the medium of

tobacco cells (Shin et al. 2003). Another noticeable

difference with tobacco is the ability of turnip hairy

roots to undergo medium replacement and continuous

production of GFP. Whereas tobacco root cultures

turned brown and collapsed after 55 days turnip roots

remained in a healthy and productive state for at least

100 days, having undergone two changes of medium

(Fig. 6). This behaviour would make them suitable for

protein production using fed-batch type cultures

(Mairet et al. 2010).

The gfp coding sequence has been fused to a 6xHis

coding sequence and a SP-encoding sequence from an

A. pme gene. Based on software prediction (SignalP,

Bendtsen et al. 2004), the SP was expected to be

cleaved between the 23rd and 24th amino acid

(Fig. 1). To support this prediction, an antibody raised

against the His tag has been used to detect the fusion

protein extracted from roots or recovered from the

medium. The His tag was linked to a 27 kDa protein in

GFP-expressing roots (Fig. 7a). This indicated that the

SP cleavage occurred upstream of the His tag, as

predicted (Fig. 1). However, the His tag was no longer

attached to GFP in the medium, suggesting that some

Fig. 5 Identification of GFP in hairy root culture medium.

a PAGE analysis of culture media of wild-type hairy roots (1) or

hairy roots expressing the egfp gene (2). A concentrate of 50 ll

of culture medium was loaded into each lane. M molecular

weight marker. b Amino acid sequence of the Aequorea victoria

GFP. The peptide contigs identified by LC–MS–MS in the

27 kDa band extracted from the PAGE gel are indicated in bold

Biotechnol Lett

123

trimming occurred during or after secretion. This has

been confirmed by Edman sequencing of the GFP

protein purified from the medium, which showed that

the N-terminus of the protein was VSKGE, the valine

being the second amino acid of GFP. From this, it can

be inferred that the DPMHHHHHHPM sequence had

been deleted from the fusion protein found in the

medium. This might have been the result of some

uncharacterized protease activity in the cell wall or in

the medium. The presence of a smaller band stained by

Ponceau S in the root protein extract suggests that

some degradation occurred within the roots (Fig. 7b).

The fact that only one band was detected by the

antibody in the root extract indicated that no signif-

icant amount of SP–His–GFP protein (expected size

30.5 kDa) accumulated in the cytosol. It can therefore

be inferred that most, if not all, was transferred to the

endoplasmic reticulum for secretion.

In conclusion: our results support B. rapa hairy

roots as a promising GRAS platform and a valuable

alternative to tobacco for the production of heterolo-

gous proteins. This fully contained system, which

features none of the risks associated with transgenic

plants in an open environment, would make a safe

alternative for the production of proteins for medical

or industrial use. The presence of the protein of

interest in the culture medium would greatly facilitate

the downstream purification process.

Acknowledgments We thank Dr. Laurent Coquet, from the

‘‘Plateforme de Proteomique Pissaro’’, University of Rouen, for

the GFP N-terminus determination and LC–MS-MS analysis,

RIKEN for the pme cDNA clone, Katarina Bradfield, Carol

Robins and Jean-Luc Henrioul for critical reading of the

manuscript.

References

Baranski R, Klocke E, Schumann G (2006) Green fluorescent

protein as an efficient selection marker for Agrobacterium

rhizogenes mediated carrot transformation. Plant Cell Rep

25:190–197

Bendtsen JD, Nielsen H, von Heijne G, Brunak S (2004)

Improved prediction of signal peptides: SignalP 3.0. J Mol

Biol 340:783–795

Boothe JG, Saponja JA, Parmentier DL (1997) Molecular

farming in plants: oilseeds as vehicles for the production of

pharmaceutical proteins. Drug Dev Res 42:172–181

Brekke OH, Sandlie I (2003) Therapeutic antibodies for human

diseases at the dawn of the twenty-first century. Nature Rev

Drug Discov 2:52–62

Chilton MD, Tepfer DA, Petit A, David D, Casse-Delbart F,

Tempe J (1982) Agrobacterium rhizogenes inserts T-DNA

0

20

40

60

80

100

120

140

0 20 40 60 80 100

days

BrGFP(mg/l)

0

0.2

0.4

0.6

0.8

1

NtGFP(mg/l)

Br NtFig. 6 Accumulation of

GFP in the culture medium

of tobacco (Nt) and turnip

(Br) hairy root lines. Arrows

indicate medium changes.

Means and confidence

intervals were calculated

from five independent

experiments

Fig. 7 His tag immuno-detection by western blotting and

Ponceau S staining of the proteins in turnip hairy roots and in the

culture medium. a Western blotting. b Ponceau S staining on the

membrane after western blotting. M molecular weight marker;

1 proteins concentrated from wild-type root culture medium;

2 proteins concentrated from GFP-producing root culture

medium; 3 proteins extracted from wild-type roots; 4 proteins

extracted from GFP-producing roots

Biotechnol Lett

123

into the genomes of the host plant root cells. Nature

295:432–434

Datla RSS, Hammerlindl JK, Panchuk B, Pelcher LE, Keller W

(1992) Modified binary plant transformation vectors with

the wild-type gene encoding NPTII. Gene 211:383–384

De Guzman G, Walmsley AM, Webster DE, Hamill JD (2011)

Hairy roots cultures from different Solanaceous species

have varying capacities to produce E coli B-subunit heat-

labile toxin antigen. Biotechnol Lett 33:2495–2502

Doran PM (2013) Therapeutically important proteins from

in vitro plant tissue culture systems. Curr Med Chem

20:1047–1055

Fischer R, Stoger E, Schillberg S, Christou P, Twyman RM

(2004) Plant-based production of biopharmaceuticals. Curr

Opin Plant Biol 7:152–158

Gaume A, Komarnytsky S, Borisjuk N, Raskin I (2003) Rhiz-

osecretion of recombinant proteins from plant hairy roots.

Plant Cell Rep 21:1188–1193

Grillberger L, Kreil TR, Nasr S, Reiter M (2009) Emerging

trends in plasma-free manufacturing of recombinant pro-

tein therapeutics expressed in mammalian cells. Biotechnol

J 4:186–201

Guerineau F (1995) Tools for expressing foreign genes in plants.

Methods Mol Biol 49:1–32

Guerineau F, Brooks L, Mullineaux P (1991) Effects of dele-

tions in the cauliflower mosaic virus polyadenylation

sequence on the choice of the polyadenylation sites in

tobacco protoplasts. Mol Gen Genet 226:141–144

Hellwig S, Drossard J, Twyman RM, Fischer R (2004) Plant cell

cultures for the production of recombinant proteins. Nature

Biotechnol 22:1415–1422

Kang D, Gho YS, Suh M, Kang C (2002) Highly sensitive and

fast protein detection with Coomassie brilliant blue in

sodium dodecyl sulfate-polyacrylamide gel electrophore-

sis. Bull Korean Chem Soc 23:1511–1512

Kim M, O’Callaghan PM, Droms KA, James DC (2011a) A

mechanistic understanding of production instability in

CHO cell lines expressing recombinant monoclonal anti-

bodies. Biotechnol Bioeng 108:2434–2446

Kim NS, Yu HY, Chung ND, Shin YJ, Kwon TH, Yang MS

(2011b) Production of functional recombinant bovine

trypsin in transgenic rice cell suspension cultures. Protein

Express Purif 76:121–126

Ko K, Koprowski H (2005) Plant biopharming of monoclonal

antibodies. Virus Res 111:93–100

Ma JKC, Chikwamba R, Sparrow P, Fischer R, Moloney R,

Twyman RM (2005) Plant-derived pharmaceuticals—the

road forward. Trends Plant Sci 10:581–585

Mairet F, Villon P, Boitel-Conti M, Shakourzadeh K (2010)

Modeling and optimization of hairy root growth in fed-

batch process. Biotechnol Prog 26:847–856

Medina-Bolivar F, Cramer C (2004) Production of recombinant

proteins by hairy roots cultured in plastic sleeve bioreac-

tors. Methods Mol Biol 267:351–363

Nishimura A, Morita M, Nishimura Y, Sugino Y (1990) A rapid

and highly efficient method for the preparation of compe-

tent Escherichia coli cells. Nucleic Acids Res 18:6169

Nopo L, Woffenden BJ, Reed DG, Buswell S, Zhang C, Medina-

Bolivar F (2012) Super-promoter: TEV, a powerful gene

expression system for tobacco hairy roots. Methods Mol

Biol 824:501–526

Ono NN, Tian L (2011) The multiplicity of hairy root cultures:

prolific possibilities. Plant Sci 180:439–446

Patel JR, Heldens JGM (2009) Immunoprophylaxis against

important diseases of horses, farm animals and birds.

Vaccines 27:1797–1810

Peebles CA, Gibson SI, Shanks JV, San KY (2007) Long-term

maintenance of a transgenic Catharanthus roseus hairy

root line. Biotechnol Prog 23:1517–1518

Peters J, Stoger E (2011) Transgenic crops for the production of

recombinant vaccines and anti-microbial antibodies. Hum

Vaccin 7:367–374

Plasson C, Michel R, Lienard D, Saint-Jore-Dupas C,

Sourrouille C, Grenier de March G, Gomord V (2009)

Production of recombinant proteins in suspension-cultured

plant cells. Methods Mol Biol 483:145–161

Plotkin SA (2005) Vaccines: past, present and future. Nature

Med 11:S5–S11

Pujol M, Gavilondo J, Ayala M, Rodriguez M, Gonzales EM,

Perez L (2007) Fighting cancer with plant-expressed

pharmaceuticals. Trends Biotechnol 25:455–459

Sharp JM, Doran PM (2001) Strategies for enhancing mono-

clonal antibody accumulation in plant cell and organ cul-

tures. Biotechnol Prog 17:979–992

Shih SMH, Doran PM (2009) Foreign protein production using

plant cell and organ cultures: advantages and limitations.

Biotechnol Adv 27:1036–1042

Shin YJ, Hong SY, Kwon TH, Jang YS, Yang MS (2003) High

level of expression of recombinant human granulocyte-

macrophage colony stimulating factor in transgenic rice

cell suspension culture. Biotechnol Bioeng 82:778–783

Stern M, Herrmann R (2005) Overview of monoclonal anti-

bodies in cancer therapy: present and promise. Crit Rev

Oncol Hematol 54:11–29

Streatfield SJ (2007) Approaches to achieve high-level heter-

ologous protein production in plants. Plant Biotechnol J

5:2–15

Sunil Kumar GB, Ganapathi TR, Srinivas L, Revathi CJ, Bapat

VA (2006) Expression of hepatitis B surface antigen in

potato hairy roots. Plant Sci 170:918–925

Twining SS (1984) Fluorescein isothiocyanate-labeled casein

assay for proteolytic enzymes. Anal Biochem 143:30–34

Wongsamuth R, Doran PM (1997) Production of monoclonal

antibodies by tobacco hairy roots. Biotechnol Bioeng

54:401–415

Woods RR, Geyer BC, Mor TS (2008) Hairy-root organ cultures

for the production of human acetylcholinesterase. BMC

Biotechnol 8:95

Xu J, Dolan MC, Medrano G, Cramer CL, Weathers PJ (2012)

Green factory: plants as bioproduction platforms forrecombinant proteins. Biotechnol Adv 30:1171–1184

Biotechnol Lett

123

Y Huet et al 2013

Documents