Protein Expression and Purification 63 (2009) 89–94

Contents lists available at ScienceDirect

Protein Expression and Purification

journal homepage: www.elsevier.com/ locate /yprep

Expression, purification and osteogenic bioactivity of recombinant human

BMP-4, -9, -10, -11 and -14

P.C. Bessa a,b,c,*, M.T. Cerqueira a,b, T. Rada a,b, M.E. Gomes a,b, N.M. Neves a,b, A. Nobre c, R.L. Reis a,b, M. Casal c

a 3B’s Research Group – Biomaterials, Biodegradables and Biomimetics, University of Minho, Headquarters of the European Institute of Excellence on Tissue Engineering and

Regenerative Medicine, AvePark, 4806-909 Taipas, Guimarães, Portugalb IBB – Institute for Biotechnology and Bioengineering, PT Associated Laboratory, Guimarães, Portugalc CBMA – Cen tro de Bi o lo gia Molec u lar e Am bi en tal, Depart ment of Biol ogy, Uni ver sity of Min ho, Cam pus de Gu al tar, 4710-057 Braga, Por tu gal

a r t i c l e i n f o a b s t r a c t

Article history:

Received 17 June 2008

and in revised form 13 September 2008

Available online 4 October 2008

Bone mor pho ge netic pro teins (BMPs) are cyto kines from the TGF-b super fam ily, with impor tant roles dur ing

embry onic devel op ment and in the induc tion of bone and car ti lage tis sue dif fer en ti a tion in the adult body.

In this con tri bu tion, we report the expres sion of recombinant human BMP-4, BMP-9, BMP-10, BMP-11 (or

growth dif fer en ti a tion fac tor-11, GDF-11) and BMP-14 (GDF-5), using Esch e richia coli pET-25b vec tor. BMPs

were over ex pres sed, puri fied by affin ity his-tag chro ma tog ra phy and shown to induce the expres sion of early

mark ers of bone dif fer en ti a tion (e.g. smad-1, smad-5, runx2/cbfa1, dlx5, os ter ix, oste o pon tin, bone sialo pro-

tein and alka line phos pha tase) in C2C12 cells and in human adi pose stem cells. The described approach is a

prom is ing method for pro duc ing large amounts of dif fer ent recombinant BMPs that show potential for novel

bio med i cal appli ca tions.

© 2008 Else vier Inc. All rights reserved.

Key words:

Bone mor pho ge netic pro tein-4

Bone mor pho ge netic pro tein-9

Bone mor pho ge netic pro tein-10

Growth dif fer en ti a tion fac tor-11

Growth dif fer en ti a tion fac tor-5

C2C12

Intro duc tion

Dis cov ered in 1965, bone mor pho ge netic pro teins (BMPs) are a

group of sev eral cyto kines belong ing to the trans form ing growth

fac tor-b (TGF-b)1 super fam ily. These pro teins have strong induc ing

activ ity in bone and car ti lage for ma tion, and also have impor tant

roles dur ing embry onic pat tern ing and early skel e tal devel op ment

[1]. Dur ing the last decades, BMPs have been used as pow er ful

os teo in duc tive com po nents in sev eral late-stage tis sue engi neer-

ing prod ucts for bone graft ing [2]. Since 2002, BMP-2 and BMP-7

received approval for spe cific clin i cal uses [3].

* Cor re spond ing author. Address: 3B’s Research Group Bio ma te ri als,

Bio de grad ables and Bi omi met ics, Dept. of Poly mer Engi neer ing, Uni ver sity of Min-

ho, Head quar ters of the Euro pean Insti tute of Excel lence on Tis sue Engi neer ing and

Regen er a tive Med i cine, Ave Park, Zona Indus trial da Gan dra, S. Cláu dio do Balco,

4806-909 Cal das das Tai pas, Gui marães, Por tu gal. Fax: +351 253 510 909.

E-mail address: paulo.bessa@dep.um in ho.pt (P. Bessa).

1 Abbre vi a tions used: rhBMP, recombinant human bone mor pho ge netic pro tein;

TGF-b, trans form ing growth fac tor b; CHO, Chi nese Ham ster Ovary; PCR, poly mer-

ase chain reac tion; HSV, her pes sim plex virus; E. coli, Esch e richia coli; LB, Luria-

Ber tan i broth; SDS–PAGE, Sodium dode cyl sul phate poly acryl amide gel elec tro pho-

re sis; PBS, Phos phate buffer saline; DMEM, Dul becco’s mod i fied Eagle’s medium;

ALP, alka line phos pha tase; MTS, (3-(4,5)-dimeth yl thi a zol-2-yl)-5-(3-carboxy-

methoxy phenyl)-2-(4-sulf ophe nyl)-2H-tet ra zo lium); RT-PCR, reverse tran scrip tion

poly mer ase chain reac tion; BSP, bone sialo pro tein; Runx2/Cbfa1, runt-related tran-

scrip tion fac tor 2/ core bind ing fac tor alpha-1; Dlx5, dis tal-less homeo box 5.

1046-5928/$ - see front matter © 2008 Else vier Inc. All rights reserved.

doi:10.1016/j.pep.2008.09.014

Despite being exten sively inves ti gated, to date, most research

has been focused only in a few BMPs, par tic u larly in BMP-2 and

BMP-7, whilst little is known about the phys i o log i cal roles and the

bio med i cal rel e vance for the remain ing BMPs. Both, BMP-4 and

BMP-9 have been shown clearly bone-induc ing roles, as shown by

their abil ity to induce oste o genic dif fer en ti a tion in C2C12 cell line

[4] and ortho topic ossi fi ca tion in mice [5]. BMP-4 is also impli cated

in diverse roles dur ing embry onic devel op ment, and as a dif fer en-

ti a tion fac tor to he mat o po et ic and nerve cells [6–8]. BMP-14 (or

growth and dif fer en ti a tion fac tor-5) induces mainly ten don and

car ti lage for ma tion [9,10] and has been also stud ied in heal ing

of peri odon tal lig a ment [11]. BMP-11 and BMP-10 have not been

iden ti fied for any bone-induc ing potential. BMP-10 has expres sion

restricted solely to the heart tis sue and is essen tial to car diac mor-

pho gen e sis [12,13], while BMP-11 (or growth and dif fer en ti a tion

fac tor-11) has roles in the estab lish ment of axial pat tern ing [14].

In former reports, human BMP-4 and BMP-14 have been expressed

in E. coli and reported to induce alka line phos pha tase marker in

murine cells but these were not tested in human cells [15,16]. In

addi tion, the expres sion of human BMP-9, -10 and -11 in E. coli is

not yet reported. In this work, we described the pro duc tion, puri-

fi ca tion and bio ac tiv ity of recombinant human BMP-4, -9, -10, -11

and -14 by a novel method that was pre vi ously reported for the

over ex pres sion of human BMP-2 [17]. The dif fer ent BMPs were

ana lyzed by their abil ity to induce the mRNA expres sion of sev eral

mark ers of bone dif fer en ti a tion in C2C12 cells and in human adult

90 P. Bessa et al. / Protein Expression and Purification 63 (2009) 89–94

adjusted to 7.5 with 0.5 M HCl and applied to a pre-equil i brated

Hi sTrap chro ma tog ra phy col umn (Amersham). Briefly, the col umn

was equil i brated with five col umn vol umes of sodium phos phate

buffer con tain ing 30 mM imid az ole, washed exten sively with 100–

120 ml of sodium phos phate buffer con tain ing 60 mM imid az ole

and the sam ple eluted in 20 ml of phos phate buffer sup ple mented

with 400 mM imid az ole. A flow rate of 5 ml/min was kept dur ing

the whole pro ce dure by the use of a peri stal tic pump. Puri fied

pro tein was desalted with the use of a Hi Trap desalt ing col umn

(Amersham) and freeze–dried. BMPs were resus pended in ster-

ile phos phate buffer saline, prior to bio ac tiv ity tests. Pro tein con-

cen tra tion was mea sured by Brad ford method [20] using bovine

serum albu min stan dards of known con cen tra tion.

Pro tein purity was esti mated using SDS–PAGE image anal y sis

(Gel Pro Ana lyzer 4.0, Media Cyber net ics).

Western-blot detec tion of rhBMPs

Pro tein sam ples were mixed with SDS–PAGE load ing buffer and

heated at 95 °C for 5 min. Sam ples were sep a rated by using hand-

cast 10–12% reduced SDS–PAGE gels and coomassie blue-R was

employed for visu al i za tion and stain ing of gels. Sam ples were then

elec tro-trans ferred (wet Western-blot trans fer) to a nitro cel lu lose

mem brane at 100 V for 60 min. The mem brane was blocked with

2% (w/v) BSA in PBS-T buffer (PBS, 0.05% Tween 20) and incu bated

with per ox i dase con ju gated anti-His anti body (Sigma) diluted

1:2000. Image detec tion was per formed with Chem i Doc XRS and

Quan tity One soft ware (Bio Rad).

Bio ac tiv ity tests

C2C12 cells were seeded at 2.5 £ 104 cells/ml per well in a 24-well

plate, attached in Dul becco’s mod i fied Eagle’s medium (DMEM) with

1% (v/v) fetal bovine serum, 100 U/ml pen i cil lin and 100 lg/ml strep-

to my cin, at 37 °C with 5% CO2 in a humid i fied envi ron ment. Cells

were incu bated for 5 days with BMP-4, -9, -10, -11 and -14, puri fied

by the method described above (500 ng/ml) and a con trol with no

cyto kine. MTS cell via bil ity assay was per formed in C2C12 cells after

3 days of cul ture [21]. Primary cul tures of human adult stem cells

iso lated from adi pose tis sue [22] were cul tured under sim i lar con-

di tions and incu bated with BMP-4, -9 and -14 (500 ng/ml) for 3 days

of cell cul ture. For RT-PCR, the mRNA of cells was extracted with Tri-

Zol Reagent (Invit ro gen, USA) from trip li cate 24-well assays fol low-

ing the pro ce dure pro vide by the sup plier. Sam ples were quan ti fied

using a Nano Drop ND-1000 Spec tro pho tom e ter (Nano Drop Tech nol-

o gies, USA) and cDNA syn the sis per formed with iScript cDNA syn-

the sis Kit (Bio Rad, USA). PCR was per formed with spe cific prim ers

for oste o genic mark ers. Aga rose gels were visu al ized using Eagle Eye

imag ing sys tem (Strat a gene) and gene expres sion was ana lyzed with

Gel-Pro Ana lyzer (Media Cyber net ics), using poly mer ase II expres-

sion as a house keep ing marker.

stem cells from adi pose tis sue. Our results sug gest a prom is ing and

straight for ward way for the pro duc tion of dif fer ent BMPs, all capa-

ble of induc ing cell dif fer en ti a tion in the afore men tioned mod els.

Mate ri als and meth ods

Clon ing and expres sion of rhBMP-4, -9, -10, -11 and -14

The sequences cod ing the mature (bio ac tive) domains of human

BMP-4, BMP-9, BMP-10, BMP-11/GDF-11 and BMP-14/GDF-5 were

obtained from dif fer ent bac te rial clones (San ger Insti tute, UK),

which con tained the full genes encod ing the human BMPs: clone

ref er ence RP11-808F04 for BMP-4, RP11-463P17 for BMP-9, RP11-

88N13 for BMP-10, RP11-55N12 for BMP-11, and RP11-173P06 for

BMP-14. Accord ing to the data from the Uni ProtKB/Swiss-Prot

data base (http://www.exp asy.org) and from the Human Genome

Pro ject (http://www.en sem bl.org/homo_sapi ens), the sequences

cod ing for the bio ac tive domains of these BMPs, are located in one

sole exon, thus allow ing direct PCR clon ing of these BMPs from the

above men tioned clones, as was pre vi ously described [17].

DNA cod ing for the mature/bio ac tive pro tein domains was

ampli fied by PCR using the prim ers indi cated in Table 1. PCR-prod-

ucts were cloned in a pET-25b vec tor (Nova gen, USA) via the two

restric tion sites, Bam HI and XhoI, that were used to trans form

E. coli BL21(DE3) strain (Invit ro gen, UK). The restric tion sites for

these enzymes were spe cific.

DNA clon ing and manip u la tion were per formed accord ing

to stan dard pro to cols [18]. The integ rity of the cloned PCR prod-

ucts was ver i fied by DNA sequenc ing [19] using a ABI PRISM310

Genetic Ana lyzer. Trans formed E. coli BL21(DE3) strain with pET-

25b/rhBMP were inoc u lated in 50 ml Luria-Ber tan i (LB) medium

with 50 lg ampi cil lin/ml and incu bated at 37 °C, until an OD600 of

0.8 was obtained. To induce recombinant pro tein expres sion IPTG

1 mM was then added to the cul ture media, and the tem per a ture

low ered to 25 °C dur ing 24 h. Bio mass was col lected by cen tri fu-

ga tion (4000g, 20 min, 4 °C), washed once with PBS and stored at

¡20 °C.

Puri fi ca tion of rhBMPs

Fro zen bac te ria were resus pended in lysis buffer (20 mM

sodium phos phate buffer, 0.5 M NaCl, 1 mg/ml lyso zyme), with a

pro te ase inhib i tors (Com plete Mini EDTA-free, Roche). Bac te ria

cells were ultr aso ni cat ed 6 times for 30 s with inter vals of 5 min

on ice and super na tant and pellet frac tions col lected by cen tri fu-

ga tion (4000g, 30 min, 4 °C). The pellet was then incu bated over-

night in sol u bi li za tion buffer (phos phate saline buffer, 30 mM

imid az ole, 0.7 M l-argi nine, pH 10.0) at 18 °C with gen tle stir ring

and the super na tant con tain ing sol u ble rhBMP col lected after

cen tri fu ga tion (4000g, 20 min, 4 °C). The pH of super na tant was

Sta tis ti cal anal y sis

Exper i ments were per formed at least in trip li cate and expressed

as means ± stan dard devi a tions. Stu dent’s t test was used for sta-

tis ti cal anal y sis using a two-tailed paired test and ana lyzed by a

paired anal y sis of var i ance. Sta tis ti cal sig nifi cance was defined as

p < 0.05 for a 95% con fi dence.

Results

Clon ing, expres sion and puri fi ca tion of rhBMPs

The PCR orig i nated sin gle prod ucts for the cod ing regions of

human BMP-4, -9, -10, -11 and -14 (data not shown). DNA sequenc ing

Table 1

Oli go nu cleo tide prim ers for clon ing of human BMP-4, -9, -10, -11 and -14. The

nucle o tides for restric tion by Bam HI and XhoI are under lined in the for ward and

reverse prim ers, respec tively. The codons rep re sent ing the begin ning and the end

of BMP sequences are rep re sented in bold.

Gene Primer sequence: for ward/reverse

BMP-4 59 CGG GAT CCA AGC CCT AAG CAT CAC TCA C 39

59 CCC TCG AGG CGG CAC CCA CAT CC 39BMP-9 59 CGG GAT CCA AGC GCC GGG GCT GGC AG 39

59 CCC TCG AGC CTG CAC CCA CAC TCT GCC AC 39BMP-10 59 CGG GAT CCA AAC GCC AAA GGA AAC TAC 39

59 CCC TCG AGT CTA CAG CCA CAT TCG GAG 39BMP-11 59 CGG GAT CCA AAC CTG GGT CTG GAC TG 39

59 CCC TCG AGA GAG CAG CCA CAG CGA TC 39BMP-14 59 CGG GAT CCA GCC CCA CTG GCC ACT C 39

59 CCC TCG AGC CTG CAG CCA CAC GAC TC 39

P. Bessa et al. / Protein Expression and Purification 63 (2009) 89–94 91

revealed a total cor re spon dence to the encoded nucle o tides and cor-

re spond ing amino acids. Esch e richia coli BL21(DE3) strain trans for-

mants were then cul ti vated and col lected as described in Mate rial

and meth ods sec tion. After cell lysis, rhBMPs were found mainly in

the pellet (frac tion cor re spond ing to cell debris) (data not shown).

Sol u ble pro tein was recov ered using phos phate buffer con tain ing

0.7 M l-argi nine, pH 10.0, as indi cated in Mate ri als and meth ods. After

puri fi ca tion by affin ity chro ma tog ra phy, the rhBMPs were vis i ble as

mono mer, dimer and poly mer, with Western-blot, using a anti body

against the six-his ti dine tag (Fig. 1). Esti mat ing from coomassie blue-

stained SDS–PAGE, the BMPs were pure up to 95% (Table 2).

Cyto tox ic ity of rhBMPs

Bio as says were per formed in murine myo blast C2C12 cell line,

by the admin is tra tion of puri fied rhBMP-2. The dif fer ent pro teins

did not show cyto tox ic ity in C2C12 cells after three days of cell

cul ture at 50 and 500 ng/ml (Fig. 2). MTS is a via bil ity/pro lif er a tion

test and an inverse rela tion ship of tox ic ity to cells can be assumed.

BMP-4 (50 ng/ml), BMP-9 (50 ng/ml) and BMP-10 (500 ng/ml) sig-

nifi cantly stim u lated cell pro lif er a tion by approx i mately 10–25%

(p < 0.05).

Expres sion of oste o genic mark ers induced by rhBMPs

Semi-quan ti ta tive RT-PCR revealed an increase in expres sion of

sev eral early osteo blast dif fer en ti a tion mark ers in C2C12 cell line

after five days of cell cul ture with the dif fer ent BMPs (Fig. 3). The

increased mRNA lev els were mean val ues of inde pen dent assays,

and nor mal ized against the expres sion of poly mer ase II. Expres-

sion of early osteo chon dral tran scrip tion fac tor runx2/cbfa1 was

induced by 2- to 3-fold with BMP-4 and BMP-11, 4-fold with

BMP-10 and BMP-14, and 6-fold with BMP-9 (p < 0.05 in all cases).

Expres sion of smad-1 and smad-5, two sig nal ing mol e cules for

BMPs, were also induced but in a lower extent. The expres sion

of smad-5 with induced 2-fold with BMP-4 and BMP-9 (p < 0.05).

Expres sion of smad-1 was induced with BMP-9 (2-fold, p < 0.05).

With BMP-10, BMP-11 and BMP-14, there was no sig nifi cant

increase in the expres sion of smad-1 or -5. Os ter ix, a tran scrip tion

fac tor down stream of runx2/cbfa-1 and a marker of com mit ment

to osteo blas tic line age, was induced both with BMP-4 (2-fold) and

BMP-9 (2-fold, p < 0.05) and its expres sion main tained with the

other BMPs. Os ter ix expres sion was slightly decreased by BMP-11

(not sta tis ti cally sig nifi cant). The expres sion of tran scrip tion fac-

tor dlx5 was nearly the same with all BMPs, with a slight increase

with BMP-4 (1.5-fold) which is not sig nifi cant. Alka line phos pha-

tase (ALP), bone sialo pro tein and oste o pon tin, three mark ers of

bone for ma tion, were induced with some of these BMPs. ALP was

induced with BMP-4, BMP-9 (2-fold, p < 0.05) and BMP-10 and -14

(1.5-fold), and reduced by half (0.5-fold) with BMP-11 (p < 0.05).

Bone sialo pro tein expres sion was up-reg u lated with BMP-14

(3-fold) and with BMP-4 (5-fold, p < 0.05). Oste o pon tin expres sion

was up-reg u lated mainly with BMP-9, BMP-14 and less with BMP-4

and BMP-10 (p < 0.05), com pared with basal expres sion in the con-

trol and with BMP-11. Oste o cal cin, a marker spe cific of late-stages

of bone min er al i za tion, was not detected in any assay. The obser-

va tions on cell mor phol ogy cor re lated well with the induc tion of

the sev eral oste o genic mark ers (Fig. 4). BMP-9, and to a less degree

BMP-4 and BMP-14, induced an osteo blast-like mor phol ogy.

In the primary cul tures of fat-derived human stem cells, BMP-4,

BMP-9 and BMP-14 were also able to induce an increase in the

expres sion of the dif fer ent early mark ers of oste o genic dif fer en ti a-

tion (Fig. 5). BMP-4 induced the high est lev els of oste o genic mark-

ers, par tic u larly runx2 (7-fold, p < 0.05), bone sialo pro tein (7-fold,

p < 0.05) and oste o pon tin (8-fold, p < 0.05). BMP-9 induced runx2

(2-fold) and alka line phos pha tase (2-fold), while BMP-14 induced

the expres sion of alka line phos pha tase (5-fold, p < 0.05) and bone

sialo pro tein (3-fold, p < 0.05). Bac te ria trans formed with blank vec-

tor (e.g. no BMP) did not induce any expres sion of oste o genic genes

(data not shown).

Dis cus sion

In this work we report a straight for ward meth od ol ogy for the

expres sion of recombinant human BMP-4, BMP-9, BMP-10, BMP-

11/GDF-11 and BMP-14/GDF-5 in E. coli, that showed bio ac tiv ity in

both C2C12 cell line and human adult stem cells.

Although the con cept of over express ing BMPs in E. coli is not

novel, we have shown, for the first time, the over ex pres sion of

func tional BMP-9, -10 and -11 using E. coli and char ac ter ized the

bio ac tiv ity of BMP-4, BMP-14 and BMP-9 using primary cul tures of

human adult stem cells, which has not been shown before. Using

a sin gle-step method, which was pre vi ously reported for the sol-

u bi li za tion of rhBMP-2 [17], the recov ery with 0.7 M l-argi nine

buffer resulted in mono mer, dimer and poly mer, in a sim i lar way

to former reports [23,24], with up to 50 mg puri fied BMPs per liter

of cul ture broth.

Fig. 1. Western-blot of puri fied rhBMP-4, -9, -10, -11 and -14 and a neg a tive con trol

(respec tively, 1–6) using an anti 6-His tag anti body. The BMPs are vis i ble in mono-

mer, dimer and poly mer.

Table 2

Puri fi ca tion of recombinant human BMP-4, BMP-9, BMP-10, BMP-11 (GDF-11) and BMP-14 (GDF-5).

Steps Crude lysateal-Argi nine sol u bi li za tion Ni-NTA chro ma tog ra phy

Pro tein Total pro tein (mg)b Purity (%)c Total pro tein (mg) Purity (%) Total pro tein (mg) Purity (%)

hBMP-4 497.5 Yield 100% 24.4 203.6 Yield 70.5% 42.3 53.6 Yield 40.0% 91.0

hBMP-9 531.5 Yield 100% 17.2 220.9 Yield 66.5% 27.5 45.9 Yield 44.5% 88.7

hBMP-10 533.3 Yield 100% 14.9 261.3 Yield 77.4% 23.7 22.7 Yield 25.5% 92.4

hBMP-11 506.4 Yield 100% 13.5 278.5 Yield 80.1% 19.7 23.4 Yield 30.1% 87.7

hBMP-14 569.1 Yield 100% 19.7 252.5 Yield 73.4% 32.5 35.0 Yield 29.5% 95.1

a The recombinant hBMPs were puri fied from 1 L E. coli cul ture. b Total pro tein quan ti fi ca tion was deter mined using Brad ford method. c Esti mated from SDS–PAGE image anal y sis.

92 P. Bessa et al. / Protein Expression and Purification 63 (2009) 89–94

Fig. 2. Tet ra zo lium salt (MTS) test per formed in C2C12 cells as a func tion of the con cen tra tion of the dif fer ent rhBMPs, after three days of cell cul ture. Cell via bil ity is

expressed in per cent age (mean ± S.D., n = 3, *p < 0.05).

Fig. 3. mRNA expres sion of sig nal ing mol e cules and tran scrip tion fac tors (A) and oste o genic-spe cific mark ers (B) induced by BMPs, mea sured by semi-quan ti ta tive RT-PCR

after five days of cell cul ture in C2C12 cell line. Bars rep re sent mean val ues ± SD of gene expres sion fold var i a tion of at least three inde pen dent deter mi na tions (*p < 0.05).

Smad-1 and Smad-5 were induced mostly for BMP-4, -9 and -14, Os ter ix for BMP-4 and -9, Runx2 was induced for all BMPs but higher for BMP-9. Bone sialo pro tein was

induced mostly for BMP-4 and -11; ALP was induced 1.5- to 2-fold for all BMPs except BMP-11; oste o pon tin was induced mainly for BMP-9, BMP-10 and BMP-14.

P. Bessa et al. / Protein Expression and Purification 63 (2009) 89–94 93

Fig. 4. Cell mor phol ogy of C2C12 cells after five days of cell cul ture with (A) no growth fac tor (con trol), (B) 500 ng/ml rhBMP-4, (C) 500 ng/ml rhBMP-9, (D) 500 ng/ml rhBMP-

10, (E) 500 ng/ml rhBMP-11, and (F) 500 ng/ml rhBMP-14. Puri fied BMP was used £400. Osteo blast-like mor phol ogy (o) was observed in C2C12 cells stim u lated with BMP-4,

BMP-9 and BMP-14 and myo blast-like mor phol ogy (m) in con trol and cells stim u lated with BMP-10.

Fig. 5. mRNA expres sion of early mark ers of oste o genic dif fer en ti a tion induced by BMPs, mea sured by semi-quan ti ta tive RT-PCR after three days of human adult stem cell

cul tures. Bars rep re sent mean val ues ± SD of gene expres sion fold var i a tion of at least three inde pen dent deter mi na tions (*p < 0.05). The expres sion of oste o genic mark ers was

induced mainly with BMP-4. BMP-9 induced runx2 and ALP (2-fold) and BMP-14 induced solely ALP (5-fold) and bone sialo pro tein (3-fold).

94 P. Bessa et al. / Protein Expression and Purification 63 (2009) 89–94

All tested BMPs have shown evi dence of bio ac tiv ity in C2C12

cells. C2C12 cell line is a well-known model for study ing the oste-

o genic dif fer en ti a tion induced by BMPs as has been reported by

sev eral inves ti ga tors [17,24,25]. The dif fer ent BMPs induced sev eral

early mark ers of BMP sig nal ing (smad-1/-5, runx2/cbfa-1 or os ter ix)

and of bone min er al i za tion (bone sialo pro tein, oste o pon tin or ALP).

Semi-quan ti ta tive RT-PCR anal y sis of mRNA expres sion was selected

since it is well estab lished that upon bind ing of the BMPs to their cell

recep tors, the expres sion of these genes is highly induced, trig ger-

ing oste o genic dif fer en ti a tion [1]. In C2C12 cells, BMP-4, -9 and -14

appeared to have had the most sig nifi cant effects in the mark ers of

bone for ma tion. BMP-11 solely induced sig nifi cantly the BMP-acti-

vated tran scrip tion fac tor runx2, show ing that prob a bly this BMP

does not have any impor tant role in bone for ma tion and heal ing,

besides those described in the lit er a ture dur ing embry onic devel op-

ment [14]. In human stem cells, BMP-4, -9 and -14 also dem on strated

bio ac tiv ity, despite the fact that the oste o genic potential was higher

for BMP-4 than BMP-9 or -14, which could point for the stron ger

bone-induc ing potential of this cyto kine.

Despite the fact that these recombinant BMPs show bio ac tiv ity,

fur ther stud ies will be per formed to assess the detailed response

of cells to BMPs, includ ing the induc tion of the dif fer ent mark ers

by RT-PCR, and immu no stain ing anal y sis.

In con clu sion, the approach we detailed herein shows a

prom is ing route for pro duc ing large amounts of sev eral dif fer ent

bio ac tive recombinant BMPs. The results in human adult stem cell

cul tures sug gest that some of these BMPs could be used with suc-

cess for future clin i cal appli ca tions.

Acknowl edg ments

The author wishes to acknowl edge the San ger Insti tute for kindly

offer ing the bac te rial clones for clon ing of human BMP-9 to -14. This

work was sup ported by Fun dação para a Ci ên cia e Tecn o lo gia (PhD

grant SFRH/BD/17049/2004, pro ject ElastM POCI/CTM/57177/2004

funded by FEDER and the Fun dação para a Ci ên cia e Tecn o lo gia; and

Euro pean STREP Pro ject HIP PO CRATES (NMP3-CT-2003-505758).

This work was car ried out under the scope of the Euro pean NoE

EX PERT IS SUES (NMP3-CT-2004-500283).

Ref er ences

[1] P.C. Bessa, M. Ca sal, R.L. Reis, Bone mor pho ge netic pro teins in tis sue engi neer-ing: the road from lab o ra tory to the clinic, part I (basic con cepts), J. Tis sue Eng. Regen. Med. 2 (2008) 1–13.

[2] R.J. West er huis, R.L. van Be zooi jen, P. Kloen, Use of bone mor pho ge netic pro-teins in trau ma tol ogy, Injury 36 (2005) 1405–1412.

[3] P.C. Bessa, M. Ca sal, R.L. Reis, Bone mor pho ge netic pro teins in tis sue engi neer-ing: the road from lab o ra tory to the clinic, part II (BMP deliv ery), J. Tis sue Eng. Regen. Med. 2 (2008) 81–96.

[4] H. Cheng, W. Ji ang, F.M. Phil lips, et al., Oste o genic activ ity of the four teen types of human bone mor pho ge netic pro teins (BMPs), J. Bone Joint Surg. Am. 85-A (2003) 1544–1552.

[5] Q. Kang, M.H. Sun, H. Cheng, et al., Char ac ter iza tion of the dis tinct ortho topic bone-form ing activ ity of 14 BMPs using recombinant ade no vi rus-med i ated gene deliv ery, Gene Ther. 11 (2004) 1312–1320.

[6] F. Li, S. Lu, l. Vida, et al., Bone mor pho ge netic pro tein 4 induces effi cient hema-to poi etic dif fer en ti a tion of rhe sus mon key embry onic stem cells in vitro, Blood 98 (2001) 335–342.

[7] D. Gia co mini, M. Páez-Pe re da, M. The od or o pou lou, et al., Bone mor pho ge netic pro tein-4 con trol of pitu i tary path o phys i ol ogy, Front. Horm. Res. 35 (2006) 22–31.

[8] E. Re iss mann, U. Erns ber ger, P.H. Fran cis-West, et al., Involve ment of bone mor pho ge netic pro tein-4 and bone mor pho ge netic pro tein-7 in the dif fer en ti-a tion of the adren er gic phe no type in devel op ing sym pa thetic neu rons, Devel-op ment 122 (1996) 2079–2088.

[9] P. Aspen berg, C. Forsl und, Enhanced ten don heal ing with GDF 5 and 6, Acta Ort hop. Scand. 70 (1999) 51–54.

[10] J.S. Dines, L. Weber, P. Razz ano, et al., The effect of growth dif fer en ti a tion fac-tor-5-coated sutures on ten don repair in a rat model, J. Shoul der Elbow Surg. 16 (2007) S215–S221.

[11] T. Na kam ura, M. Ya mam ot o, M. Tam ura, Y. Iz umi, Effects of growth/dif fer en-ti a tion fac tor-5 on human peri odon tal lig a ment cells, J. Peri odon tal Res. 38 (2003) 597–605.

[12] H. Neu haus, V. Rosen, R.S. Thies, Heart spe cific expres sion of mouse BMP-10 a novel mem ber of the TGF-beta super fam ily, Mech. Dev. 80 (1999) 181–184.

[13] H. Chen, S. Shi, L. Ac o sta, BMP10 is essen tial for main tain ing car diac growth dur ing murine car di o gen e sis, Devel op ment 131 (2004) 2219–2231.

[14] O. An ders son, E. Re iss mann, C.F. Ibáñez, Growth dif fer en ti a tion fac tor-11 sig-nals through the trans form ing growth fac tor-beta recep tor ALK5 to region al ize the ante rior-pos ter ior axis, EMBO Rep. 7 (2006) 831–837.

[15] N.R. Kübler, M. Mo ser, K. Berr, Bio log i cal activ ity of E. coli expressed BMP-4, Mund Kie fer Ge sic hts chir 2 (1) (1998) S149–S152.

[16] C. Sie ber, F. Plö ger, R. Schwapp ach er, Mono meric and dimeric GDF-5 show equal type I recep tor bind ing and olig o mer i za tion capa bil ity and have the same bio log i cal activ ity, Biol. Chem. 387 (2006) 451–460.

[17] P.C. Bessa, A.J. Pedro, B. Klösch, et al., Osteo in duc tion in human fat-derived stem cells by recombinant human bone mor pho ge netic pro tein-2 pro duced in E. coli, Bio tech nol. Lett. 30 (2008) 15–21.

[18] J. Sam bro ock, E.F. Frit sch, T. Ma nia tis, Molec u lar Clon ing: A Lab o ra tory Man-ual, sec ond ed., Cold Spring Har bor Lab o ra tory, New York, 1989.

[19] F. San ger, S. Nick len, A.R. Coul son, DNA sequenc ing with chain-ter mi nat ing inhib i tors, Proc. Natl. Acad. Sci. USA 74 (1977) 5463–5467.

[20] M.M. Brad ford, A rapid and sen si tive method for the quan ti ta tion of micro-gram quan ti ties of pro tein uti liz ing the prin ci ple of pro tein–dye bind ing, Anal. Bio chem. 72 (1976) 248–254.

[21] A.J. Sal gad o, M.E. Go mes, A. Chou, et al., Preliminary study on the adhe sion and pro lif er a tion of human oste o blasts on starch-based scaf folds, Mater. Sci. Eng. 20 (2002) 27–33.

[22] P.P. Ma la faya, A.J. Pedro, A. Pet er bauer, et al., Chito san par ti cles agglom er ated scaf folds for car ti lage and osteo chon dral tis sue engi neer ing approaches with adi pose tis sue derived stem cells, J. Mater. Sci. Mater. Med. 16 (2005) 1077–1085.

[23] B. Klösch, W. Fürst, R. Knei ding er, et al., Expres sion and puri fi ca tion of bio log i-

cally active rat bone mor pho ge netic pro tein-4 pro duced as inclu sion bodies in recombinant E. coli, Bio tech nol. Lett. 27 (2005) 1559–1564.

[24] S. Long, L. Tru ong, K. Ben nett, Expres sion, puri fi ca tion, and rena tur ation of bone mor pho ge netic pro tein-2 from E. coli, Pro tein Expr. Purif. 46 (2006) 374–378.

[25] L.F. Val lejo, M. Brokel mann, S. Mar ten, et al., Rena tur ation and puri fi ca tion of bone mor pho ge netic pro tein-2 pro duced as inclu sion bodies in high-cell-den-sity cul tures of recombinant E. coli, J. Bio tech nol. 94 (2002) 185–194.