Ultrathin sP(EO-stat-PO) hydrogel coatings are biocompatibleand preserve functionality of surface bound growth factors in vivo

Carl Neuerburg • Stefan Recknagel • Jorg Fiedler • Jurgen Groll •

Martin Moeller • Kristina Bruellhoff • Heiko Reichel • Anita Ignatius •

Rolf E. Brenner

Received: 17 April 2013 / Accepted: 11 June 2013

� Springer Science+Business Media New York 2013

Abstract Hydrogel coatings prepared from reactive star

shaped polyethylene oxide based prepolymers (NCO-

sP(EO-stat-PO)) minimize unspecific protein adsorption

in vitro, while proteins immobilized on NCO-sP(EO-stat-

PO) coatings retain their structure and biological function.

The aim of the present study was to assess biocompatibility

and the effect on early osseointegrative properties of a

NCO-sP(EO-stat-PO) coating with additional RGD-pep-

tides and augmentation with bone morphogenetic protein-4

(BMP) used on a medical grade high-density polyethylene

(HDPE) base under in vivo circumstances. For testing of

biocompatibility dishes with large amounts of bulk NCO-

sP(EO-stat-PO) were implanted subcutaneously into 14

Wistar rats. In a second set-up functionalization of implants

with ultrathin surface layers by coating ammonia-plasma

treated HDPE with NCO-sP(EO-stat-PO), functionalization

with linear RGD-peptides, and augmentation with RGD

and BMP-4 was analyzed. Therefore, implants were placed

subcutaneously in the paravertebral tissue and transcortic-

ally in the distal femur of another 14 Wistar rats. Both tests

revealed no signs of enhanced inflammation of the sur-

rounding tissue analyzed by CD68, IL-1ß-/TNF-a-antibody

staining, nor systemic toxic reactions according to histo-

logical analysis of various organs. The mean thickness of

the fibrous tissue surrounding the femoral implants was

highest in native HDPE-implants and tended to be lower in

all NCO-sP(EO-stat-PO) modified implants. Micro-CT

analysis revealed a significant increase of peri-implant

bone volume in RGD/BMP-4 coated samples. These results

demonstrate that even very low amounts of surface bound

growth factors do have significant effects when immobi-

lized in an environment that retains their biological func-

tion. Hence, NCO-sP(EO-stat-PO)-coatings could offer an

attractive platform to improve integration of orthopedic

implants.

1 Introduction

Implant failure is often caused by poor osseointegration or

periprosthetic infection and might necessitate difficult

revision surgery. It has been shown, that this process can be

mediated by unspecific cell adhesion such as fibroblastic

cells or biofilm producing bacteria in the early phase after

implantation [1]. Considering the increasing demand of

joint replacement surgery, the necessity of revision surgery

is expected to rise similarly [2]. Therefore, great effort has

C. Neuerburg � H. Reichel

Department of Orthopaedics, University of Ulm, Ulm, Germany

Present Address:

C. Neuerburg

Experimental Surgery and Regenerative Medicine, Department

of Surgery, Ludwig-Maximilians-University, Munich, Germany

S. Recknagel � A. Ignatius

Institute of Orthopaedic Research and Biomechanics, University

of Ulm, Ulm, Germany

J. Fiedler � R. E. Brenner (&)

Division for Biochemistry of Joint and Connective Tissue

Diseases, Department of Orthopaedics, University of Ulm,

Oberer Eselsberg 45, 89081 Ulm, Germany

e-mail: rolf.brenner@uni-ulm.de

J. Groll � M. Moeller � K. Bruellhoff

DWI e.V. and Institute of Technical and Macromolecular

Chemistry, RWTH Aachen University, Aachen, Germany

Present Address:

J. Groll

Department of Functional Materials in Medicine and

Dentistry, University Hospital Wurzburg, Wurzburg, Germany

123

J Mater Sci: Mater Med

DOI 10.1007/s10856-013-4984-4

been taken in the past trying to reduce the risk of revision

surgery. Besides modification of the physicochemical

properties of orthopedic implants, organic and inorganic

surface modifications have been introduced such as

hydroxyapatite coatings [3]. Our group has recently per-

formed in vitro studies on a novel hydrogel coating based

on a star shaped (NCO-sP(EO-stat-PO); for better read-

ability of the article abbreviated as sPEOPO) that can be

applied on the implant surface using a spin-, dip- or spray

coating technique [1, 4–7]. The system consists of 80 %

ethylene oxide and 20 % propylene oxide with terminal

reactive isocyanate groups (NCO) that enable a covalent

binding to surfaces functionalized with isocyanate-reactive

groups. In vitro experiments have shown promising fea-

tures of star-shaped PEO layers. Ultrathin coatings of

cross-linked star-shaped polyethylene oxide with func-

tional isocyanate endings sPEOPO proved to be extremely

resistant to unspecific adsorption of proteins [4]. Thirty

nanometer thick sPEOPO coatings primarily prevented

mesenchymal cell adhesion but could be effectively func-

tionalized with specific adhesion peptides opening a unique

possibility to control and direct cell adhesion processes

under highly selective conditions [5, 8]. In addition, sPE-

OPO-coated glass slides showed up to 93 % germ-reduc-

tion after exposure to Staphylococcus aureus in comparison

to the native glass surface [1]. This opens the possibility to

guide eukaryotic cell adhesion and to reduce prokaryotic

adhesion by a common coating strategy to reduce the risk

of implant failure. In this context there is increasing

interest in the use of growth factors for functionalization of

orthopedic implants. Growth factors regulate cellular

events that are part of the processes of tissue repair and

regeneration [9]. Bone morphogenetic protein (BMP) can

be grouped into various proteins of which BMP-2, BMP-4

and BMP-7 belong to the most studied subtypes [10].

Originally BMPs were known to induce the formation of

bone and cartilage, whereas in recent years their impor-

tance in the prevention of cancers has been additionally

stressed [11]. Since biotechnology offers fast and efficient

production of biologically active growth factors such as

BMP-2 [12], their potential to improve implant integration

is actively investigated. Implant surface modification using

growth factures such as BMP has been shown to induce

bone regeneration [9, 13, 14]. Unsolved questions are the

optimal strategy of local growth factor delivery and the

high concentrations of BMPs needed so far to achieve a

therapeutic effect. In this context biodegradable hydrogel

coatings that have been shown to enable a continuous

release of bioactive proteins may accessorily support bone

regeneration [15].

There are only few studies investigating hydrogel

coatings under in vivo conditions in general [15, 16], and

none has been performed with the NCO-sP(EO-stat-PO)-

coating so far. Therefore, the purpose of the present study

was to investigate the biocompatibility of sPEOPO in vivo

and the functionality of incorporated growth factors with

respect to bone tissue integration of a biomaterial covered

with NCO-sP(EO-stat-PO)-RGD-peptide functionalized

hydrogel coatings, supplemented with rhBMP-4.

2 Materials and methods

2.1 Preparation of sPEOPO samples

for biocompatibility testing

sPEOPO was synthesized as previously reported [17].

Hydrogels were formed by dissolution of sPEOPO at

20 wt% in sterile water under a flow bench, initial stirring

for homogeneous distribution of the polymer and sub-

sequent sterile storage under humid atmosphere to prevent

drying [18]. After storage for 24 h to make sure that all

isocyanate groups are hydrolyzed and the cross-linking

reaction is completed, discs with a diameter of 10 mm and

a thickness of 4 mm were pinned out of these samples

under sterile conditions.

2.2 Preparation of HDPE implants for testing

of functionality and osseointegration

Medical grade high-density polyethylene (HDPE) in disc

(diameter 10 mm, thickness 5 mm) and cylinder shape

(diameter 2 mm, height 3 mm) was used as implant

material. HDPE was kindly provided by Aesculap (Ae-

sculap, Germany). For easier handling during the coating

procedure, the cylindrical substrates were spiked onto

0.4 9 12 mm needles. HDPE samples were cleaned in

70 % ethanol and subsequently left for drying under the

flow-bench for 24 h to ensure complete evaporation of

ethanol. Ammonia plasma activation was carried out using

low-pressure ammonia plasma (Microwave Discharge AK

330 Plasma Apparatus from Roth and Rau Oberflachen-

technik, Germany).

XPS analysis and protein adsorption were carried out to

ensure homogeneous activation of the samples from all

sides. Protein adsorption experiments were performed by

incubation of the samples in a 100 mg/mL solution of

bovine serum albumin (BSA) tetramethylrhodamine conju-

gate in PBS-buffer (pH 7.4) for 20 min, followed by five

times washing with PBS buffer, a final thorough washing

step with deionized water and subsequent drying in a stream

of nitrogen. Analysis of protein adsorption was performed

via fluorescence microscopy using a Zeiss Axiovert 100A

microscope equipped with halogen lamp, fluorescence filters

and a digital camera. Integration time for all images was

identical to allow comparability of the results.

J Mater Sci: Mater Med

123

Linear RGD-functional coatings were produced with

identical coating procedure using a RGD-supplemented

solution by dissolving 50 mg sPEOPO in 0.5 mL THF

followed by addition of 3.5 mL sterile water and 1 mL of

an aqueous solution of GRGDS (Bachem, Switzerland)

with a concentration of 1 mg/mL.

In addition to RGD, growth factors were immobilized on

the surface of some disc and cylindrical HDPE samples.

Therefore, the freshly sPEOPO coated and still NCO-

functional substrates were incubated in a 0.01 lg/mL

aqueous solution of BMP-4 for 10 min in an Eppendorf

vial followed by four times rinsing with sterile water and

subsequent drying in the Eppendorf vials under the clean

bench overnight. The samples were then stored in sterile

Eppendorf vials until further use within 10 days. A com-

plete list of substrates prepared for the animal experiments

is displayed in Table 1.

2.3 Animal model

For testing the biocompatibility of bulk sPEOPO, two

hydrogel disks were implanted in the left and right sub-

cutaneous dorsal tissue of 14 Wistar rats (weight

300–400 g).

For testing the preservation of functionality of biomol-

ecules that are covalently bound to a sPEOPO-coated

surface, 14 adult Wistar rats (weight 400–500 g) were

randomly chosen for bilateral implantation of a cylindrical

implant in the distal femur. Additional implantation of two

discoidal polyethylene implants was performed in the left

and right paravertebral subcutaneous tissue.

All animal studies were performed according to the

principles of the Guide for the Care and Use of Laboratory

Animals and approved by the local regulatory agency

(registered under Reg.-Nr.: 850).

2.4 Surgical procedure

Before surgery the rats received subcutaneous injection

of atropine sulphate (Braun, Germany) at 0.05 mg/kg to

stabilize respiration and counter low-pulse-induced

anesthesia. Ten minutes thereafter inhalation anesthesia

was performed in an induction chamber at 5 % Isofluran

and 800 mL/min oxygen. During surgery anesthesia was

obtained using a facemask at 2 % of Isofluran. All animals

received preoperative Clindamycin-2-dihydrogenpho-

sphate, (Pfizer, Karlsruhe, Germany) as antibiotic prophy-

laxis (45 mg/kg) and Tramadol (Grunenthal, Germany,

20 mg/kg) for pain relief. Skin incision was performed at

the medial aspect of the knee joint, the fascia latae was cut

and a transmuscular approach performed to expose the

insertion of the medial collateral ligament at the distal left

and right femur. While cooling with saline, a monocortical

drill hole was placed at the distal femur using a conven-

tional reaming drill with a diameter of 2.4 mm. Then a

cylindrical implant was placed into the hole and additional

saline lavage applied. After muscular suture and fascior-

rhaphy skin closure was performed.

For the discoidal implants (bulk material or coated

HDPE), a skin incision was done at the thoracic back and

two pockets build in the left- and right paravertebral sub-

cutaneous tissue. One implant was placed into each pocket

and subcutaneous sutures applied to prevent dislocation of

the implants, followed by closure of the skin.

Postoperative analgesia was obtained for three days by

using Tramadol at 20 mg/kg diluted in the animals drink-

ing water. To prevent infection the same antibiotic used

preoperatively was also administered subcutaneously

3 days postoperatively at 45 mg/kg.

2.5 Blood flow analysis

For blood flow analysis of the subcutaneously located

polyethylene-based implants a laser-Doppler-measure-

ment-unit was used connected with a tissue-spectrometer

(O2C, LEA Medizintechnik, Germany). The system is

based on a continuous wave white light Laser device class

3 B, protective class 1 with a power \30 mW, a detection

range of 450–850 nm and a resolution of 1 nm. Blood flow

analysis was performed in anesthesia by placing the laser

probe over the adjacent tissue of the left- and right para-

vertebral tissue in the area of operation. Measurement was

done before and at day 3, 7, 14 and 21 after surgery. The

parameters for flow and velocity were recorded at a mea-

surement interval of 1 s and a recording time of 10 s.

2.6 Specimen retrieval and processing

Half of the bulk sPEOPO treated animals were killed after

3 weeks, whereas the other half of the group was killed

after 12 weeks. Animals treated with a HDPE based

implant were all sacrificed at day 21 after surgery

according to the principles of the Guide for the Care and

Table 1 Complete list of HDPE samples prepared for the animal

experiments

Sample Number of

samples

Uncoated HDPE 14

NCO-sP(EO-stat-PO) coated HDPE 14

NCO-sP(EO-stat-PO) coated HDPE ? RGD 14

NCO-sP(EO-stat-PO) coated HDPE ? RGD ? BMP-4 14

J Mater Sci: Mater Med

123

Use of Laboratory Animals. In these animals femurs were

dissected and stored in formalin for histological processing

and l-CT analysis. The discoidal implant surrounding tis-

sue and a selection of internal organs (liver, kidney and

spleen) were harvested and stored in formalin.

2.7 l-CT scanning and image segmentation

All femoral samples were scanned using a Skyscan1172

l-CT scanner (Skyscan, Belgium). Resolution was set at

8.68 lm operating at a source voltage of 49 kV and

200 mA. After three-dimensional reconstruction of the

images the sagittal sections were isolated using the soft-

ware Data Viewer (Version 1.4.1.0, Skyscan, Belgium).

The bone volume (BV), including bone volume/tissue

volume ratio (BV/TV), bone surface (BS), trabecular

thickness (TbTh), trabecular number (TbN) and trabecular

separation (TbSp) of the implant surrounding bone were

measured using CT-Analyser software (Version 1.9.3.2,

Skyscan, Belgium). Therefore, the region of interest (ROI)

was set from the implant surface to a distance of 0.2 mm

around the implant to assess the implant/bone interface. By

combining the ROIs of 51 slices a volume of interest was

generated for each sample for densitometric measurement.

A global threshold of 25 % of the maximum gray value

was chosen to distinguish mineralized tissue from unmin-

eralized and poorly mineralized tissue as previously

described by Morgan et al. [19].

2.8 Histomorphometry

After fixation in formalin the femurs were treated with

graded concentrations of alcohol and embedded in methyl

methacrylate (Technovit, Heraeus, Germany). Undecalci-

fied slices were cut with the use of a diamond saw (Exact,

Grunewald Inc., Germany) and subsequently manually

guided grinding down to a thickness of 100 lm (Exact,

Germany). Then the sections were stained with trichrome

Masson-Goldner to show new bone formation as previ-

ously described [20].

Histomorphometric analysis was performed with an

Axioskop 2mot plus (Zeiss, Germany) connected to a

camera (Axio Cam MRc, Zeiss, Germany) and a computer.

Thickness of the fibrous tissue surrounding the implant was

measured at a tenfold magnification in 45� clockwise

rotation around the implants surface by using the software

Axio Vision 4.8.00 (Zeiss, Germany).

2.9 Histology

As an indicator of systemic toxic effects hematoxylin-eosin

staining of the liver, kidney and spleen was performed for

spot test analysis.

2.10 Immunohistochemistry

For immunohistochemistry, deparaffinised sections of tis-

sue samples surrounding the implants were pretreated with

a primary rat specific monoclonal CD68 antibody (ACRIS,

Germany) for macrophage detection and primary rat-spe-

cific IL-1ß-/TNF-a-antibodies (both from ACRIS, Ger-

many) for cytokine detection. Staining without primary

antibody served as negative control.

2.11 Statistical analysis

Analysis was performed using SPSS software, version 19.0

(SPSS, Inc., Chicago, IL). Data are presented as

mean ± standard deviation (SD) or median and confidence

interval (Fig. 5). The data were analysed using Levene’s

test for homogeneity of variance and student’s t test, sig-

nificance was set at (P \ 0.05).

3 Results

The experimental setup was divided in two parts. First, we

were interested to test the biocompatibility of NCO-sP(EO-

stat-PO) in vivo and screen for signs of inflammation and

toxicity. Second, we wanted to check the in vivo-applica-

bility of thin NCO-sP(EO-stat-PO) coatings on a model

implant with respect to osseointegration. In this context the

work also focussed on effects of RGD-peptide integration

and the question whether NCO-sP(EO-stat-PO) coating

maintains the functionality of additionally integrated

growth factors.

3.1 Testing the biocompatibility of NCO-sP(EO-stat-

PO) in vivo

In case of biocompatibility testing all animals tolerated

the bulk material insertion well. After three (n = 7) and

12 weeks (n = 7) the subcutaneous inserted bulk disks

were surrounded by a thin fibrous capsule with mild signs

of an inflammatory reaction as shown in Fig. 1a, b. The

disks could be easily removed from the capsule because of

lacking adhesion or integration into the surrounding tis-

sue. Macroscopically the capsule presented no prominent

vascularization and contained only few CD68 positive

cells and a faint staining for IL-1ß and TNF-a in histo-

logical sections of paraffin embedded tissue (Fig. 1c–e).

The analyzed internal organs showed no signs of systemic

toxic effects (Fig. 1f–h). Overall these results indicated

that the insertion of huge amounts of NCO-sP(EO-stat-

PO) neither showed toxicity in liver, spleen, or kidney nor

induced severe inflammatory reactions in the surrounding

tissue.

J Mater Sci: Mater Med

123

3.2 Testing the applicability of NCO-sP(EO-stat-PO)

for osseointegration

3.2.1 Coating of implants

The coating technology could be successfully adapted to

the implant geometry and medical grade polyethylene as

implant material resulting in hydrogel films of \100 nm.

Using ammonia plasma activation at 4.0 9 10-1 mbar

with 400 W at a gas inlet flow rate of 30 sccm ammonia for

300 s only minimal residual protein adsorption tested with

BSA as a model substance was detectable (Fig. 2).

3.2.2 In vivo experiments with coated HDPE

One animal had to be sacrificed postoperatively due to a

femoral fracture. No wound healing disorders occurred. For

testing the biofunctionality of growth factors integrated in

the NCO-sP(EO-stat-PO)-surface 13 animals could be

evaluated with at least 6 femoral and subcutaneous samples

Fig. 1 In vivo testing of NCO-

sP(EO-stat-PO) in Wistar rats.

a After 12 weeks of

subcutaneous integration the

bulk NCO-sP(EO-stat-PO)-

discs were surrounded by a

fibrous capsule. b The bulk disc

could easily be removed after

cutting the capsule into pieces.

The thickness of the tissue was

\1 mm and only a few blood

vessels were detectable. Only a

mild inflammatory reaction

within the surrounding tissue

was detectable by c CD68

staining. d TNF-alpha staining.

e IL-1beta staining. HE staining

from harvested tissues of

f spleen, g kidney, h liver

showed a normal appearance

(magnification 9100)

J Mater Sci: Mater Med

123

of each coating. Spot test histologic analysis of internal

organs such as the kidney, liver and spleen again didn’t

reveal signs of systemic toxic reaction at the time of sac-

rifice after 21 days (data not shown).

3.2.3 Subcutaneous implants

For blood flow analysis the preoperative blood supply of

the operating area was set at 100 % and blood flow changes

were measured with the Laser-Doppler-measurement-unit

located in the adjacent tissue of the implant. There was a

slight decrease in blood flow in the peri-implant tissue at

day 3 after surgery seen in all samples. At day 7 postop-

eratively, blood flow percentage returned to the preopera-

tive value in all animals disregarding the implant surface.

Fourteen days after surgery blood flow percentage showed

a drop in the sPEOPO-RGD coated group (77 %) while an

increased blood flow of RGD/BMP-4 augmented samples

was measurable (120 %), yet there was no significance

compared to uncoated HDPE implants. At the end of the

experimental time blood flow of the adjacent tissue

returned almost completely to the index perfusion level.

Thus a mean flow of 108 % in the sPEOPO-RGD and

105 % in the solely sPEOPO modified implants respec-

tively could be measured. Adjacent tissues of RGD/BMP-4

augmented implant surfaces showed a small reduction in

blood flow percentage 85 % compared to their corre-

sponding uncoated HDPE implants. Though there was no

significance for the differences observed and all data resi-

due within the range of technical reproducibility (Fig. 3).

3.2.4 Femoral implants

Histomorphometric analysis revealed an increased mean

thickness of the fibrous tissue surrounding femoral

implants of uncoated HDPE samples (149.1 lm), com-

pared to sPEOPO and sPEOPO-RGD (67.3 lm/108.8 lm)

or BMP-4 augmented implants (61.44 lm). The differ-

ences did not reach statistical significance either in relation

to uncoated HDPE, nor sPEOPO-RGD modified samples

(Fig. 4; Table 2).

Bone density was calculated in the femoral sections

using l-CT analysis, thus the peri-implant bone structure

was assessed at a given ROI of 0.2 mm around the implant.

Analysis of the samples treated with sPEOPO-RGD coated

implants revealed a significant increase in trabecular

thickness (TbTh) and trabecular separation (TbSp) (0.13/

0.69 mm) compared to the corresponding uncoated HDPE

surfaces (0.09/0.49 mm). Furthermore bone volume (BV)

of BMP-4 augmented coatings (0.60 mm3) was signifi-

cantly increased compared to uncoated HDPE samples

(0.33 mm3; P B 0.05). Comparison of the growth factor

augmented surfaces with solely sPEOPO-RGD coated

implants revealed a significant increase in BV (P B 0.05)

and trabecular number (TbN) (P B 0.05) of the BMP-4

augmented implants (0.42 mm3/0.89/mm compared to

0.60 mm3/1.43/mm), as shown in Table 2 and Fig. 5.

Fig. 2 Adsorption of BSA

tetramethylrhodamine conjugate

on a cleaned disc shaped HDPE

control substrate (a) and on

sPEOPO coated disc shaped

HDPE substrates after ammonia

plasma activation at

4.0 9 10-1 mbar with 400 W at

a gas inlet flow rate of 30 sccm

ammonia for 180 s (b), 240 s

(c) and 300 s (d). Only after

300 s of treatment, protein

adsorption on the substrate was

minimized

J Mater Sci: Mater Med

123

4 Discussion

In order to improve osseous integration of implants many

organic and inorganic surface modifications have been

introduced. Although there have been encouraging reports

on the use of surface functionalization in in vivo experi-

ments on dental implants [21, 22], the clinical use of bio-

mimetic implants in orthopedic and trauma surgery

remains rare.

In the present study we therefore investigated the bio-

compatibility and early osseointegrative properties of a

hydrogel coating based on star-shaped poly ethylene oxide-

stat-propylene oxide (sPEOPO) augmented with RGD-

peptides and growth factors like bone morphogenetic pro-

teins (BMP-4) in an animal model.

Using in vitro studies we previously could show that

coating of a variety of substrates with sPEOPO leads to

minimization of protein adsorption and thus prevention of

cell adhesion. Most interestingly, cell adhesion mediating

peptide sequences may be introduced in a very simple one-

step layer preparation which leads to cell adhesion that is

exclusively induced via the immobilized peptides [5, 23,

24]. The non-interacting background in which the peptides

are embedded retains the specific biological function of the

peptides [8, 24]. This coating has mostly been performed

using spin coating, a technique useful for preparation of

films on flat model substrates like glass or titanium [23].

Here, we transferred the coating to 3D substrates by simple

incubation of aminofunctionalized HDPE substrates in

sPEOPO solutions. As reported before, introduction of

ammonia-groups on the substrates is necessary to ensure

covalent binding and long-term stability of the sPEOPO

coating [23]. We have used our previously reported

microwave induced ammonia plasma treatment method for

silicone [8] as starting point and optimized the process

parameters for HDPE. In contrast to silicone, the surface of

HDPE does not recover after plasma treatment, so that

shorter treatment times were tested. However, protein

adsorption tests on sPEOPO coated substrates showed that

only after plasma for 300 s, the adsorption of BSA as

model protein was minimized. Hence, the HDPE substrates

for the in vivo study were plasma treated accordingly.

Coating of these substrates with sPEOPO by incubation

could be achieved with solutions of identical concentration

as for spin coating.

Fig. 3 Blood flow percentage in the adjacent soft tissue of the ectopic

implants. Before surgery and at 3, 7, 14 and 21 days post surgery

Fig. 4 Representative images of implant positioning in bone tissue of

sPEOPO-RGD coated HDPE samples with the implant surrounding

fibrous tissue illustrated at higher magnification (a) and a cross

section of the distal femur after trichrome Masson-Goldner staining

(b). Cross sectional images obtained via Micro-CT (c)

J Mater Sci: Mater Med

123

Till today no information is available on the use of

sPEOPO in vivo and how it will interact with the sur-

rounding tissue over a longer time period. So before using

sPEOPO for implant coatings, we tested the overall toxicity

and biocompatibility of a large amount of sPEOPO material

through formation of sPEOPO hydrogel discs containing

20 wt% polymer. Our results go in line with data presented

by Wang et al. [25] about the non toxicity of similar 3 armed

polymers in short time experiments in mice. They also

didn’t observe any toxic effects and described a well toler-

ated usage of small amounts for drug delivery.

For evaluation of the inflammatory response to the novel

hydrogel coating we investigated the ectopic host foreign

body response. Bridges et al. [26] have shown that the

subcutaneous implantation of surface coated implants in rats

remains a feasible model to investigate the biocompatibility

of microgel coatings in vivo. For analysis of the ectopic

biocompatibility during the experimental setup we investi-

gated the blood supply of the implant surrounding subcu-

taneous tissue, since blood flow analysis has been shown to

be a reliable proof of an inflammatory response [27]. In our

series a decrease in blood flow was detectable in all implant

surrounding soft tissues in the early post-operative phase,

which might be due to the primary surgical procedure. One

week after surgery the blood flow was back to the index

perfusion-level again. At 14 days after surgery there was a

mild increase in blood flow of BMP-4 augmented samples.

With regards to implant integration an improved vasculari-

zation induced via growth factor augmented surfaces could

lead to accelerated bone regeneration in the peri-implant

interface which could offer an advantage in osseous tissues.

Similarly, it has also been shown by Bouet et al. [27] in a

similar experimental setup, that an inflammatory response is

associated by hyperemia and thus one has to take into

account that the observed effect could be due to an

inflammation. However, at the end of the experimental setup

blood flow decreased again, although inflammation of the

peri-implant tissue around the ectopic implants could have

occurred temporarily. To further test the inflammatory

response, sections of the peri-implant tissue were stained for

CD68 antigen a known immunohistochemical marker for

multinucleated giant cells/macrophages [28]. Likewise, it

has been shown that an inflammatory response mediated by

cytokines such as IL-1ß, TNF-a and IL-6 can cause

impaired bone healing in an experimental model of fracture

healing [29]. Thus the inflammatory response in the ectopic

tissue has been further investigated by IL-1ß and TNF-astaining as an indicator of monocyte and macrophage

activity, since osseointegration with the peri-implant surface

might be similarly affected.

Fig. 5 Boxplot analysis illustrating the median bone volume (BV)

and 95 % confidence interval of the femoral samples at a given ROI

of 0.2 mm around the implant. Coatings are indicated on the x-axis

(*P B 0.05)

Table 2 Mean thickness of the fibrous tissue surrounding the implant according to histomorphometric analysis (lm) and bone structure

parameters assessed via l-CT analysis in the femoral samples at a given ROI of 2.4 mm

Uncoated HDPE sPEOPO sPEOPO-RGD sPEOPO-RGD ? BMP-4

CpTh 149.10 ± 89.18 (70.91–304.68) 67.27 ± 37.17 (40.99–93.55) 108.82 ± 91.47 (28.25–223.11) 61.44 ± 11.16 (49.27–78.09)

BV 0.33 ± 0.21 (0.07–0.66) 0.52 ± 0.30 (0.21–0.81) 0.42 ± 0.11 (0.28–0.51) 0.60 ± 0.08 */� (0.48–0.69)

BV/TV 10.67 ± 617 (3.55–20.19) 13.34 ± 7.66 (5.52–20.83) 10.87 ± 2.97 (7.07–13.40) 15.74 ± 2.49 (13.26–18.36)

BS 16.81 ± 7.09 (9.26–29.86) 29.26 ± 16.99 (16.20–48.46) 14.69 ± 5.35 (9.80–21.72) 23.10 ± 7.26 (18.15–40.12)

TbTh. 0.09 ± 0.04 (0.04–0.13) 0.09 ± 0.04 (0.05–0.13) 0.13 ± 0.01 * (0.11–0.14) 0.11 ± 0.02 (0.07–0.14)

TbNo. 1.33 ± 1.01 (0.49–3.55) 1.60 ± 0.94 (1.05–2.69) 0.89 ± 0.32 (0.51–1.17) 1.43 ± 0.21� (1.23–1.89)

TbSp. 0.49 ± 0.20 (0.28–0.73) 0.59 ± 0.12 (0.50–0.72) 0.69 ± 0.09 * (0.59–0.80) 0.61 ± 0.07 (0.51–0.72)

Data are given as mean ± standard deviation and range put in parentheses. Significant values in comparison to uncoated HDPE implants are

indicated * and significance in comparison to sPEOPO-RGD coated implants indicated � respectively

CpTh mean thickness of the fibrous tissue surrounding the implant according to histomorphometry (lm), BV bone volume (mm3), BV/TV bone

volume/tissue volume (mm3), BS bone surface (mm2), TbTh. trabecular thickness (mm), TbNo. trabecular number (1/mm), TbSp. trabecular

separation (mm)

J Mater Sci: Mater Med

123

Both experimental setups showed that there was no

increased inflammatory response in the vicinity of bulk

sPEOPO discs or sPEOPO modified HDPE implants under

ectopic conditions after 12 and 3 weeks respectively. Nor

did an additional surface-modification with RGD-peptides

and BMP-4 cause any inflammatory response. This is in

accordance with other in vivo experiments that evaluated

the inflammatory response of similar hydrogel surfaces.

Kim et al. tested another biodegradable PEG/sebacic acid-

based hydrogel (PEGSDA) in a similar experimental setup

to our model. They found almost the same acute and

chronic inflammatory responses between the control

material and the PEGSDA-based hydrogels while they

reported a normal wound healing response and neither a

massive inflammatory response nor a thick fibrous capsule

formation around PEGSDA hydrogel implants [30]. Thus

the local subcutaneous and systemic effect of sPEOPO

appears to be insignificant.

For evaluation of the sPEOPO surface modification in

bone contact, we inserted a cylindrical implant in the medial

aspect of the femoral cortical bone. Biocompatibility and

functionality was assessed using histological and l-CT

analysis. A complex cascade of events, which involves

protein adsorption, inflammatory response and mesenchy-

mal cell recruitment mediates the host foreign body response

to synthetic implants, resulting in fibrous encapsulation of

the implant [26]. Therefore, histomorphometric analysis of

the implants was performed, which revealed that the fibrous

tissue surrounding sPEOPO modified implant surfaces ten-

ded to be thinner compared to native HDPE implants. This

effect was most distinct in solely sPEOPO modified sur-

faces, though there was no significance.

Micro-CT analysis was performed in order to investigate

the peri-implant bone structure. The results indicated a

significantly higher BV in the peri-implant interface of

RGD/BMP-4 augmented hydrogel coatings compared to

untreated HDPE implants. This observation is in accor-

dance with other studies reporting improved osteoinduc-

tivity of surfaces modified by the addition of BMP [14, 31].

In particular BMP-4 has been shown to enhance early

implant stability as reported by Lai et al. [32] in femurs of

ovariectomized rabbits. Furthermore, it is important to

highlight the release of growth factors. Thus, Liu et al. [14]

used 1–3 lm thick calcium phosphate coatings containing

at least 10 lg BMP-2 per implant. Kang et al. used a

system similar to our approach, with surface initiated

polymerization PEG-methacrylate and subsequent immo-

bilization of BMP, although the experiments were per-

formed under in vitro circumstances [31]. In our approach,

the growth factors are covalently immobilized onto an

ultrathin hydrogel film during the cross-linking reaction of

the hydrogel. Hence, the proteins cannot penetrate into the

already partially formed network so that only a molecular

monolayer of proteins is immobilized. Taking into account

the average size of the growth factors resulting in an

approximate surface coverage of 50 nm2 per molecule, we

can estimate the growth factor density on the surface as

2 9 1012 molecules/cm2. With the sample geometry of

cylinders (2 mm diameter and 3 mm height) and 30 kDa as

molecular weight, the cylindrical samples were augmented

with a maximal total amount of 25 ng growth factor (BMP-

4). Our study thus demonstrated in vivo that covalently

immobilized growth factors even in very low amounts

(about 1 ng/mm2) can cause significant effects when pre-

sented in an environment that preserves their functional

conformation [4].

However, the experimental setup implies some con-

straints of our study. First, the short observation period of

3 weeks does not allow definite conclusions on osseointe-

gration. Since there have been reports on a pronounced

hydrogel biodegradation after a period of 4 weeks [30], a

longer investigation with different endpoints of the experi-

mental time would be of great interest. A promising future

concept would be to integrate adhesive peptides that support

primary attachment of osteogenic cells which may optimize

the long-term anchorage after degradation of the ultrathin

sPEOPO-coating. In this context, a second limitation of our

study is the use of HDPE as implant material. It allows lCT-

analysis of the periimplant tissue and assessment of effects

on fibrous capsule formation but not on osseointegration

after degradation of the ultrathin coating.

5 Conclusions

In conclusion, we found no signs of inflammation due to

plain or biofunctionalized sPEOPO coatings under ectopic

and orthotopic insertion. Furthermore, we observed a

reduced fibrous encapsulation of sPEOPO-RGD and RGD-

growth factor (BMP-4) augmented coatings compared to

uncoated HDPE implants. Also, periimplant bone density

was improved according to l-CT analysis with ng-immo-

bilized growth factor. Therefore, surface modification with

sPEOPO could represent an attractive platform for opti-

mization of implant integration and prevention of implant

loosening by infection. Furthermore, the biodegradable

properties of sPEOPO hydrogel coatings offer the chance

of a subsequent release of bioactive substances to influence

primary processes during implant anchorage. However,

further investigations are necessary to examine the medium

and long-term biological and mechanical effects of this

coating strategy with selective functionalization for cell

adhesion and growth factor delivery on osseointegration.

Acknowledgments The authors wish to thank Aesculap (Tuttlin-

gen, Germany) for providing the high density polyethylene carrier

J Mater Sci: Mater Med

123

implants. Also we would like to thank Nadine Jansen, Sarah Kraut-

hausen and Giovanni Ravalli for excellent technical assistance.

References

1. Groll J, Fiedler J, Bruellhoff K, Moeller M, Brenner RE. Novel

surface coatings modulating eukaryotic cell adhesion and pre-

venting implant infection. Int J Artif Organs. 2009;32:655–62.

2. Archibeck MJ, White RE Jr. What’s new in adult reconstructive

knee surgery. J Bone Joint Surg Am. 2006;88(7):1677–86. doi:

10.2106/JBJS.F.00450.

3. de Jonge LT, Leeuwenburgh SC, Wolke JG, Jansen JA. Organic–

inorganic surface modifications for titanium implant surfaces. Pharm

Res. 2008;25(10):2357–69. doi:10.1007/s11095-008-9617-0.

4. Groll J, Amirgoulova EV, Ameringer T, Heyes CD, Rocker C,

Nienhaus GU, et al. Biofunctionalized, ultrathin coatings of

cross-linked star-shaped poly(ethylene oxide) allow reversible

folding of immobilized proteins. J Am Chem Soc. 2004;126(13):

4234–9. doi:10.1021/ja0318028.

5. Groll J, Fiedler J, Engelhard E, Ameringer T, Tugulu S, Klok HA,

et al. A novel star PEG-derived surface coating for specific cell

adhesion. J Biomed Mater Res A. 2005;74(4):607–17. doi:10.1002/

jbm.a.30335.

6. Groll J, Haubensak W, Ameringer T, Moeller M. Ultrathin

coatings from isocyanate terminated star PEG prepolymers: pat-

terning of proteins on the layers. Langmuir. 2005;21(7):3076–83.

7. Heuts J, Salber J, Goldyn AM, Janser R, Moller M, Klee D. Bio-

functionalized star PEG-coated PVDF surfaces for cytocompat-

ibility-improved implant components. J Biomed Mater Res A.

2010;92(4):1538–51. doi:10.1002/jbm.a.32478.

8. Fiedler J, Groll J, Engelhardt E, Gasteier P, Dahmen C, Kessler

H, et al. NCO-sP(EO-stat-PO) surface coatings preserve bio-

chemical properties of RGD peptides. Int J Mol Med.

2011;27(1):139–45. doi:10.3892/ijmm.2010.553.

9. Kaigler D, Avila G, Wisner-Lynch L, Nevins ML, Nevins M,

Rasperini G, et al. Platelet-derived growth factor applications in

periodontal and peri-implant bone regeneration. Expert Opin Biol

Ther. 2011;11(3):375–85. doi:10.1517/14712598.2011.554814.

10. Lissenberg-Thunnissen SN, de Gorter DJ, Sier CF, Schipper IB.

Use and efficacy of bone morphogenetic proteins in fracture

healing. Int Orthop. 2011;35(9):1271–80. doi:10.1007/s00264-

011-1301-z.

11. Bleuming SA, He XC, Kodach LL, Hardwick JC, Koopman FA,

Ten Kate FJ, et al. Bone morphogenetic protein signaling sup-

presses tumorigenesis at gastric epithelial transition zones in

mice. Cancer Res. 2007;67(17):8149–55. doi:10.1158/0008-

5472.CAN-06-4659.

12. von Einem S, Schwarz E, Rudolph R. A novel TWO-STEP

renaturation procedure for efficient production of recombinant

BMP-2. Protein Expr Purif. 2010;73(1):65–9. doi:10.1016/j.pep.

2010.03.009.

13. Hagi TT, Wu G, Liu Y, Hunziker EB. Cell-mediated BMP-2

liberation promotes bone formation in a mechanically unstable

implant environment. Bone. 2010;46(5):1322–7. doi:10.1016/

j.bone.2010.02.010.

14. Liu Y, Enggist L, Kuffer AF, Buser D, Hunziker EB. The

influence of BMP-2 and its mode of delivery on the osteocon-

ductivity of implant surfaces during the early phase of osseoin-

tegration. Biomaterials. 2007;28(16):2677–86. doi:10.1016/

j.biomaterials.2007.02.003.

15. Ichinohe N, Kuboki Y, Tabata Y. Bone regeneration using tita-

nium nonwoven fabrics combined with fgf-2 release from gelatin

hydrogel microspheres in rabbit skull defects. Tissue Eng Part A.

2008;14(10):1663–71. doi:10.1089/ten.tea.2006.0350.

16. Rao L, Zhou H, Li T, Li C, Duan YY. Polyethylene glycol-

containing polyurethane hydrogel coatings for improving the

biocompatibility of neural electrodes. Acta Biomater. 2012;8(6):

2233–42. doi:10.1016/j.actbio.2012.03.001.

17. Gotz HBU, Bartelink CF, Grunbauer HJM, Moller M. Preparation

of isophorone diisocyanate terminated star polyethers. Macromol

Mater Eng. 2002;287(4):223–30.

18. Dalton PD, Hostert C, Albrecht K, Moeller M, Groll J. Structure

and properties of urea-crosslinked star poly[(ethylene oxide)-ran-

(propylene oxide)] hydrogels. Macromol Biosci. 2008;8(10):

923–31. doi:10.1002/mabi.200800080.

19. Morgan EF, Mason ZD, Chien KB, Pfeiffer AJ, Barnes GL,

Einhorn TA, et al. Micro-computed tomography assessment of

fracture healing: relationships among callus structure, composi-

tion, and mechanical function. Bone. 2009;44(2):335–44. doi:

10.1016/j.bone.2008.10.039.

20. Koepp HE, Schorlemmer S, Kessler S, Brenner RE, Claes L,

Gunther KP, et al. Biocompatibility and osseointegration of beta-

TCP: histomorphological and biomechanical studies in a weight-

bearing sheep model. J Biomed Mater Res B Appl Biomater.

2004;70(2):209–17. doi:10.1002/jbm.b.30034.

21. Schliephake H, Aref A, Scharnweber D, Bierbaum S, Sewing A.

Effect of modifications of dual acid-etched implant surfaces on

peri-implant bone formation. Part I: organic coatings. Clin Oral

Implants Res. 2009;20(1):31–7. doi:10.1111/j.1600-0501.2008.

01603.x.

22. Schliephake H, Scharnweber D, Dard M, Sewing A, Aref A,

Roessler S. Functionalization of dental implant surfaces using

adhesion molecules. J Biomed Mater Res B Appl Biomater.

2005;73(1):88–96. doi:10.1002/jbm.b.30183.

23. Gasteier P, Reska A, Schulte P, Salber J, Offenhausser A, Moeller

M, et al. Surface grafting of PEO-based star-shaped molecules for

bioanalytical and biomedical applications. Macromol Biosci.

2007;7(8):1010–23. doi:10.1002/mabi.200700064.

24. Salber J, Grater S, Harwardt M, Hofmann M, Klee D, Dujic J,

et al. Influence of different ECM mimetic peptide sequences

embedded in a nonfouling environment on the specific adhesion

of human-skin keratinocytes and fibroblasts on deformable sub-

strates. Small. 2007;3(6):1023–31. doi:10.1002/smll.200600596.

25. Wang X, Kan B, Wang Y, Dong P, Shi S, Guo G, et al. Safety

evaluation of amphiphilic three-armed star-shaped copolymer

micelles. J Pharm Sci. 2010;99(6):2830–8. doi:10.1002/jps.22042.

26. Bridges AW, Singh N, Burns KL, Babensee JE, Lyon Andrew L,

Garcia AJ. Reduced acute inflammatory responses to microgel

conformal coatings. Biomaterials. 2008;29(35):4605–15. doi:

10.1016/j.biomaterials.2008.08.015.

27. Bouet T, Schmitt M, Desuzinges C, Eloy R. Quantitative in vivo

studies of hyperemia in the course of the tissue response to bio-

material implantation. J Biomed Mater Res. 1990;24(11):

1439–61. doi:10.1002/jbm.820241104.

28. Kunisch E, Fuhrmann R, Roth A, Winter R, Lungershausen W,

Kinne RW. Macrophage specificity of three anti-CD68 mono-

clonal antibodies (KP1, EBM11, and PGM1) widely used for

immunohistochemistry and flow cytometry. Ann Rheum Dis.

2004;63(7):774–84. doi:10.1136/ard.2003.013029.

29. Recknagel S, Bindl R, Kurz J, Wehner T, Ehrnthaller C, Knoferl

MW, et al. Experimental blunt chest trauma impairs fracture

healing in rats. J Orthop Res. 2011;29(5):734–9. doi:10.1002/

jor.21299.

30. Kim J, Dadsetan M, Ameenuddin S, Windebank AJ, Yaszemski

MJ, Lu L. In vivo biodegradation and biocompatibility of PEG/

sebacic acid-based hydrogels using a cage implant system.

J Biomed Mater Res A. 2010;95(1):191–7. doi:10.1002/jbm.a.

32810.

31. Kang SM, Kong B, Oh E, Choi JS, Choi IS. Osteoconductive

conjugation of bone morphogenetic protein-2 onto titanium/

J Mater Sci: Mater Med

123

titanium oxide surfaces coated with non-biofouling

poly(poly(ethylene glycol) methacrylate). Colloids Surf B Bio-

interfaces. 2010;75(1):385–9. doi:10.1016/j.colsurfb.2009.08.

039.

32. Lai YL, Kuo NC, Hsiao WK, Yew TL, Lee SY, Chen HL. In-

tramarrow bone morphogenetic protein 4 gene delivery enhances

early implant stability in femurs of ovariectomized rabbits. J Pe-

riodontol. 2011;82(7):1043–50. doi:10.1902/jop.2011.100404.

J Mater Sci: Mater Med

123