MRI has recently been presented as a nondestructive in vivo readout to report perfusion capacity in biomaterials planted on the CAM in the living chick embryo in ovo. Perfusion capacity was assessed through changes in T1 relaxation pre- and post-injection of a paramagnetic contrast agent, Gd-DOTA (Dotarem®). Hence local contrast agent concentration was dependent on perfusion, vascular permeability, and extravascular compartment size. In the present study we, therefore, explore intravascular SPIO particles of the FeraSpin® series to deliver a more direct measure of vascularization in a 3D polymer DegraPol® scaffold. Furthermore, we present contrast enhancement upon SPIOs of different particle size, namely FeraSpin® series XS, M, XXL and Endorem® for comparison, and hence different efficiency on T1 and T2, and study respective dose-effects. No signal change was observed within the egg yolk, consistent with the SPIO remaining in the vasculature. Consequently, T1 positive signal enhancement (reduction in T1) and T2 negative contrast (reduction in T2) were observed only in the vasculature and hence were restricted mainly to the surface of the CAM at the interface to the biomaterial. Furthermore, the effect upon T2 appears stronger than in T1 with all SPIOs investigated and at blood concentrations between 0.46 mM to 4.65 mM. Comparison of different concentrations shows larger T1 enhancement at the highest dose, as expected. Vessel structures in and around the scaffold as seen in MRI were corroborated by histology. Different particle sizes show reduced T1 effect with larger particles, yet the effect on T2 was less apparent. In sum, SPIO-enhanced MRI provides measures for vascularization nondestructively in biomaterials connected to the CAM, based on intravascular contrast enhancement in T1 and T2, in ovo in the living chick embryo. Small SPIOs provide the best efficiency for that purpose, and contrast enhancement is most prominent in T2.

Tissue-engineered biomaterials in regenerative medicine provide a matrix for cells to attach and proliferate, stimulate angiogenesis and sustain long-term function and survival of the implant. The chorioallantoic membrane (CAM) of the chick embryo is a model for studying vascularization in vivo^[1]. As a consequence, effects of the biomaterial’s volume, pore size and pore interconnectivity on the vascularization capacity can be studied^[2].

Recently, an MRI method was presented as a nondestructive in vivo readout of perfusion capacity in biomaterials planted on the CAM in the living chick embryo in ovo^[3]. Perfusion capacity was assessed in various scaffold materials through changes in T1 relaxation pre- and post-injection of a paramagnetic contrast agent, Gd-DOTA (Dotarem®, Guerbet S.A.). Hence, local contrast agent concentration was dependent not only on perfusion but also on vascular permeability and extravascular compartment size, as Gd-DOTA diffuses into the interstitial space, particularly in “leaky” vessels. This is a different contrast compared to what is seen with blood pool agents staying in the vasculature.

In the present study, intravascular SPIO particles of the FeraSpin® series (Viscover, Miltenyi Biotec, Germany) are therefore explored to deliver a more direct measure of vascularization/ vessel architecture in 3D DegraPol® scaffolds (ab medica, Italy). We present contrast enhancement upon SPIOs of different size, namely FeraSpin® series XS, M, XXL and Endorem® (Guerbet S.A.) for comparison, and hence different efficiency on T1 and T2 and investigate respective dose effects.

To assess contrast enhancement in a 3D polymer (DegraPol®) scaffold upon injection of SPIOs as intravascular contrast agents in the CAM assay. SPIO particles of different size were used and dosages varied.

MR images were obtained from one sagittal slice positioned through the scaffold(s) planted on the CAM of the chicken embryos (Fig. 1A). A medetomidine anesthesia protocol offered proper sedation of the chick embryo throughout the MRI acquisition and the scaffold was clearly and reproducibly depicted.

No signal change was observed within the egg yolk, consistent with the SPIO remaining in the vasculature. Consequently, T1 positive signal enhancement (reduction in T1) and T2 negative contrast (reduction in T2) were observed, consistently in all samples, in the vasculature and hence were restricted mainly to the surface of the CAM at the interface to the scaffold material where vessel density is highest (Fig. 1B). The effect upon T2 appears stronger than in T1 with FeraSpin® M particles at 0.46 mM concentration in the Optimaix-3D™ scaffold.

Remarkably, no change in contrast was observed inside the scaffold, which might relate to partial volume effects between the small vascular compartment as compared to the scaffold structure, but also to our observation that the slightly viscous contrast agent did not distribute easily but rather slowly within the finer segments of the vasculature, like the vessels penetrating the scaffold. Consistent with this notion are the viscous quality of the FeraSpin® SPIO preparation and the observation that T1 and T2 contrast enhancement was more prominently seen at 140 min than 8 min after injection of the SPIO. Dose escalation from a second SPIO injection did not result in notable further contrast enhancement, possibly for the same reason.

Qualitative comparison of different FeraSpin® XS concentrations shows larger T1 enhancement at the highest dose, as expected (Fig. 1C). For comparison, doses typically suggested for use in mice are in a similar range, such as 0.66 mM (mouse example from manufacturer administering 0.1 mL of 10 mM FeraSpin®) and about 9 mM used for q-mapping^[4], administering monocrystalline iron oxide nanoparticles MION at 30 mg/kg body weight. Vessel structures as seen in MRI are corroborated by histology.

T1 and T2 efficiency for contrast enhancement is dependent on relaxivity of the SPIO particles and hence on particle size, as well as on clearance of the contrast agent from the blood compartment (blood circulation time itself dependent on particle size). T1 and T2 contrast enhancement with FeraSpin XS® and Endorem® in Degrapol® scaffolds is depicted in figure 1D for comparison. At the same contrast SPIO concentration (4.65 mM) contrast enhancement appears more prominent in T2 than in T1. Moreover, larger vessels present strong T2 effect (signal reduction) outweighing T1 enhancement (signal increase) and hence appear dark in T1w images due to these competing effects (“T2-shine-through”).

T1 and T2 efficiency of different particle sizes of the FeraSpin® SPIO particle family are shown in figure 1E. Feraspin® particles with XS 10–20 nm, M 30–40 nm, and XXL 60–70 nm hydrodynamic diameter, respectively, show reduced T1 effect with larger particle size, as expected; effect on T2 (larger effect/signal reduction with increasing particle size) is not apparent at 0.46 mM concentration.

Quantitative analysis of contrast enhancement within the whole scaffold revealed the same trends as observed qualitatively in the MR images (Fig. 1F).

Our data demonstrate that SPIO-enhanced MRI is feasible in the chick embryo in the CAM assay. SPIO-enhanced MRI provides measures for vascularization non-destructively in biomaterials connected to the CAM, based on intravascular contrast enhancement in T1 and T2, in ovo in the living chick embryo. Small SPIOs provide the best efficiency for that purpose, and contrast enhancement is most prominent in T2.

We compared MR images qualitatively for contrast enhancement and to the corresponding histological sections. In our MR assessments, the intravascular contrast was most apparent in large vessels directly at the interface between the CAM and the biomaterial planted on it. We are now working on a quantitative assessment of this region in order to quantify the effect of different SPIO particle sizes and concentrations of SPIOs applied. A thorough detailed quantitative analysis, however, requires:

-more samples per condition (SPIO particle type, concentration) and,

-to tackle dependence on ROI selection, as it varies greatly with positioning within the scaffold (especially when touching the more vascularized region towards the interface to the CAM), the size of the selected ROI (due to partial voluming between vessels and scaffold material), and the effect of single big vessels. Alternatively, a well-defined control region within the yolk may be impeded by the motion of the chick embryo between pre/post-injection MRI sessions.

Yet, a quantitative assessment of the scaffold region with the purpose to quantify the effect of different SPIO particle sizes and concentrations of SPIOs was conducted using a histogram analysis of color-coded T1 and T2 maps, which allowed to assess the more subtle signal changes within the scaffold region. This analysis confirmed contrast enhancement (reflected in the increase in the blue distribution, which in turn corresponds to decrease in T1 and T2 relaxation times within the scaffold, upon contrast agent administration) and reproduced the trends observed qualitatively, as described above, for the interface region between CAM and scaffold (Fig. 1F): Larger effects were observed at higher contrast agent concentrations, and contrast enhancement was more prominent in T2 than T1. Furthermore, increased effect on T2 was expected, and observed, with increased particle size (in the FeraSpin® series, not Endorem, however) in the scaffold region, while effect on T1 was expected to be reduced but was not obvious in the scaffold region, arguably due to underlying T2-effects (T2 shine-through).

Full thorough quantitative analysis of contrast enhancement, in particular within the scaffold, requires more samples per condition and was replaced by a semi-quantitative histogram-based analysis on our preliminary data, which is supposedly more sensitive towards smaller signal changes within the scaffold. Results from this analysis reproduced the qualitative observations/trends seen in the (more obvious) interface region between the CAM and scaffold.

Scaffold material and preparation

DegraPol® is a biocompatible and biodegradable material based on poly-hydroxy-butyrate as a crystalline segment and $\varepsilon$-caprolactone as a soft segment^[5]. Originally, its synthesis aimed at generating a suitable scaffold material for bone tissue engineering^[6], but has recently also been shown to be a beneficial scaffold material for cartilage^[7], nerves^[8] and as a growth factor delivery device for tendon regeneration.

DegraPol® foams were kindly provided by ab medica, Italy. An Optimaix-3D™ scaffold (Matricel GmbH, Germany) was used in one egg in a preparatory experiment.

After soaking in cyclohexane (Fluka, puriss.) and freezing at -20°C overnight, foams were cut into equal pieces of about 8×4×3 mm3, dried and sterilized with ethylen oxide before placing onto the CAM.

CAM assay

Fertilized Lowman white LSL chick eggs (Animalco AG Geflügelzucht, Staufen, Switzerland) were incubated at 37°C and 65% relative humidity. On incubation day (ID) 3.5 a circular window was excised into the eggshell after removing 2 mL albumen so that the developing CAM detached from the eggshell. On ID 7, Optimaix-3D™ (in one egg) and DegraPol® foam scaffolds (in 11 eggs) were planted on top of the CAM, one or two scaffolds into each egg, in the middle of 1 cm diameter plastic rings to flatten the CAM surface. Eggs were then incubated until ID 14.

SPIO-enhanced MRI to assess vascularization of the 3D DegraPol® scaffold

On ID 14, vascularization of the scaffold by capillaries of the chick embryo’s chorioallantoic membrane was studied in situ on the CAM (“in ovo”) of the living chicken embryo (“in vivo”) using Magnetic Resonance Imaging (MRI). For the MRI examination, the eggs were placed onto a custom-built sliding bed and enveloped by warm water tubing (37°C) to maintain the temperature of the chick embryo in a physiological range. Chick embryos were sedated with 0.3 mg/kg medetomidine (diluted 1:100, volume 0.3 mL) dripped onto the CAM in 3 doses 30 min prior start of MRI examinations and immediately before pre- and post-contrast-enhanced MRI.

MRI was performed on a 4.7 T/16 cm Bruker PharmaScan small animal scanner (Bruker BioSpin, Ettlingen, Germany), equipped with an actively decoupled two-coil system, consisting of a 72 mm bird cage resonator for excitation and a 20 mm single loop surface coil for reception. The surface coil was fixed onto a Petri dish cover plate that covered the eggshell window directly above the scaffold for optimal sensitivity.

Anatomical reference images were acquired in coronal, transversal and sagittal slice orientations with a routine FLASH sequence. T1- and T2-weighted MR images were then obtained from one sagittal slice placed through the scaffold with a RARE sequence of variable TR and TE for quantitative T1 and T2 mapping with TR 200/400/800/1500/3000/4500 ms, TE 9.3/27.9/46.5/65.1/83.7 ms, field of view 55×22 mm, image matrix 275×100, spatial resolution 200×200 um2, slice thickness 1 mm, RARE-factor 2, 2 averages, total scan time 13 min.

T1 and T2 maps were collected before and after injection of SPIOs. Intravenous injection was performed under a surgical microscope with a 300 uL insulin syringe and 30G needle into medium size vein. Quantitative T1 and T2 maps were computed using ParaVision® 5.1 software package (Bruker BioSpin, Ettlingen, Germany).

SPIO protocols

Different SPIO preparations and doses of the FeraSpin® series (Viscover, Miltenyi Biotec, Germany) were investigated in the present study, and Endorem® (Guerbet SA, France) was used in 2 eggs for comparison. Different SPIO particle sizes affect contrast efficiency with respect to T1 and T2 relaxation as well as blood circulation time and liver uptake.

In a preparatory experiment tolerance of high-dose iron administration into the chick embryo was tested in one egg planted with a Optimaix-3D™ scaffold and was studied prior and 8 and 140 min post-injection of SPIOs of the FeraSpin® series M (particle size 30–40 nm, 10 mM) a 0.46 mM dose in blood upon injection. In a ‘dose escalation’ test it additionally received a second 0.46 mM Fe dose 3 hrs after the first.

Throughout the paper blood concentrations upon injection are given at the corresponding places. Doses were varied by adaption of injection volume between 50 and 100 uL, and iron concentration in contrast agent preparation by dilution of contrast agent in saline. The following SPIOs were explored in the present study: FeraSpin® XXL (particle size 60–70 nm, 10 mM), FeraSpin® M (particle size 30–40 nm, 10 mM), FeraSpin® XS (particle size 10–20 nm, prepared to a custom-tailored 200 mM Fe molarity by the manufacturer). For comparison, Endorem® (particle size 120–180 nm, 200 mM) was used in 2 eggs at concentrations of 0.46 mM and 4.65 mM.

Comparative histology

After completion of the MRI measurements the scaffold-CAM complex was fixed overnight using 4% formalin solution in PBS, then excised, embedded in paraffin, cross-sectioned into 5 µm slices and stained with H&E and Haemalaun Sudan Red.

Analysis of contrast enhancement

Histograms of color-coded quantitative T1 and T2 maps obtained pre- and post-injection of the contrast agent were used to compute contrast enhancement within the scaffold. We analyzed, using a freely downloadable software package (http://arohatgi.info/WebPlotDigitizer/citation.html ) in the blue channel, decrease in T1 and T2 values, respectively, which translates into increased distribution in the blue values with contrast enhancement. Results were then explored with respect to SPIO particle size and SPIO concentration of the contrast agent.

Departmental fundings and Matching Funds 2011 (University of Zürich).

We would like to thank Ms. Pia Fuchs for H&E staining. Prof. Dr. Jan Klohs is acknowledged for providing the FeraSpin XS contrast agent. The Matching Funds of University Hospital Zürich are highly acknowledged for financial support. We are thankful to ab medica, Italy, for providing the DegraPol® foams.

The authors declare no conflicts of interest.

Not applicable.

No fraudulence is committed in performing these experiments or during processing of the data. We understand that in the case of fraudulence, the study can be retracted by ScienceMatters.