Characterization of MRI contrast agent-loaded polymeric nanocapsules as versatile vehicle for...

Full Paper

Received: 29 April 2009, Revised: 5 October 2009, Accepted: 5 December 2009, Published online in Wiley InterScience: 16 March 2010

(www.interscience.wiley.com) DOI:10.1002/cmmi.364

Characterization of MRI contrast agent-loadedpolymeric nanocapsules as versatile vehiclefor targeted imagingSonu Sharmaa, Umaporn Paiphansirib,c, Vinzenz Hombacha,Volker Mailanderb,d, Oliver Zimmermanna, Katharina Landfesterb,c

and Volker Raschea*

Various contrast agents (Magnevist1, Gadovist1

Contrast M

and Multihance1) loaded into polymeric nanocapsules weresynthesized by the inverse miniemulsion technique. The relaxivity of the resultant contrast agents was assessedat 1.5 T magnetic field strength. The ionic relaxivity of the contrast agents could at least be maintained after theirencapsulation in different polymer capsules. The chemical composition of the nanocapsules was characterized byFourier transform infrared spectroscopy. The distribution of the contrast agent in the nanocapsules could beidentified by energy filtered transmission electron microscopy and energy dispersive X-ray spectroscopy. The resultsindicate entrapment of the gadolinium complex into the inner shell of the polymeric nanocapsules. The payload ofcontrast agent per nanocapsule resulted in some 2.5T 106 Gd3R complexes yielding a particle-based relaxivity of10.75T 106mM

S1 sS1. Maintained or even slightly increased ionic relaxivity of the different contrast agents afterencapsulation in combination with high payloads and the possibility of functionalization of the capsules’ surfacefacilitate the application of the nanocapsules as promising targeted contrast agents for MRI. Copyright # 2010 JohnWiley & Sons, Ltd.

Keywords: polymer-encapsulated contrast agent; inverse miniemulsion; MRI relaxivity; nanocapsules

* Correspondence to: V. Rasche, Department of Internal Medicine II – Cardiology,University of Ulm, Albert-Einstein-Allee 23, 89081 Ulm, Germany.E-mail: [email protected]

a S. Sharma, V. Hombach, O. Zimmermann, V. Rasche

Department of Internal Medicine II – Cardiology, University of Ulm,

Albert-Einstein-Allee 23, 89081 Ulm, Germany

b U. Paiphansiri, V. Mailander, K. Landfester

Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz,

Germany

c U. Paiphansiri, K. Landfester

Department of Organic Chemistry III, University of Ulm, Albert-Einstein-Allee

11, 89081 Ulm, Germany

d V. Mailander

Institute of Clinical Transfusion Medicine/Institute of Transfusion Medicine,

University of Ulm, Helmholtzstr.11, 89081 Ulm, Germany 5

1. INTRODUCTION

Magnetic resonance imaging (MRI) has evolved as a prominentimaging technique in diagnostic clinical medicine andbiomedical research offering exquisite anatomical detail of softtissues based upon detection of protons in water and fat (1).The contrast in an MR image is the result of a complex interplayof numerous factors, including the relative T1 and T2 relaxationtimes, proton density of the imaged tissues and instrumentalparameters. The inherent contrast can be further enhanced byadministration of suitable MRI contrast agents (CA). These MRcontrast agents are chemical compounds which enhance therelaxation rate of protons in the tissue, thereby altering thesignal intensity in the image relative to areas not affected bythe contrast agent (2). Contrast agents that predominatelyreduce T1 are called positive contrast agents (e.g. paramagneticgadolinium complexes) whereas those that mainly affect T2 orT�2 are called negative contrast agents (e.g. SPIO [superparamagnetic iron oxide], USPIO [ultrasmall superparamagneticiron oxide]) (3,4). The precise effect of contrast agent dependson their concentration in the tissue concerned. At higherconcentrations, gadolinium reduces the signal intensity of thetissue generating negative contrast, whereas at low concen-tration of USPIOs, the T1 effect is prevalent in generatingpositive contrast (5,6). The efficiency of a contrast agent tochange the relaxation rate is represented quantitatively asrelaxivity, with r1 being the relaxivity corresponding to T1 orlongitudinal relaxation and r2 being the relaxivity correspond-

edia Mol. Imaging 2010, 5 59–69 Copyright #

ing to the T2 or transverse relaxation. Relaxivities are usuallyexpressed in mM

�1 s�1 (7,8).Although MRI provides good spatial resolution, it suffers from

its rather low sensitivity in detection of low concentrations oflocal contrast agent aggregations as required in the case oftargeted contrast agent imaging. Since, for this application, thelimiting factor is the number of local receptor sites, the relaxivityper binding site of a certain contrast agent must be optimized.This can either be achieved by optimization of the relaxivity of thecompound or by increasing the contrast agent load per bindingsite (9–12).

2010 John Wiley & Sons, Ltd.

9

S. SHARMA ET AL.

60

The relaxivity value of paramagnetic complexes depends onseveral parameters (3). Among these parameters the waterexchange between bulk water andwater bound to chelated Gd3þ

and rotational correlation time of the Gd3þ ions are of primeimportance (3). The rotational correlation time can be altered bye.g. conjugation of Gd3þ chelates to macromolecular objects likedendrimers (13,14), liposomes (15), micelles (16) and nano-particles (17).Several approaches for increasing the local contrast agent

load per potential binding site have been introduced based onGd3þ-labeled shell-cross-linked nanoparticles (16), perfluor-ocarbon particle (17), paramagnetic porous polymerosomes(18), paramagnetic silica Ru (bpy):GdIII/SiO2 nanoparticles (19),lipid encapsulated perfluorocarbon nanoparticles (20), encap-sulation of gadolinium complexes inside apoferritin spheres(21), gadolinium salt loaded in the nanoparticles (22) andencapsulation of gadolinium-based contrast agents in poly-meric nanocapsules (23).The enhancement of the relaxivity after encapsulation of

the contrast agent into polymeric shells or the attachmentto macromolecules in the above discussed cases is probablydue to a reduction in the rotational correlation time ofthe Gd3þ chelate. Besides the enhancement of the ion-based relaxivity, a further important advantage of themultimeric contrast agent encapsulation into polymericshells results from the potential high payload of thecontrast agent per vesicle and the potential for surfacefunctionalization. The resulting higher amount of contrastagent per binding site is supposed to significantly increasethe sensitivity of MRI, especially in molecular imagingapplications. For example, in cell marking (24), biologicalconstraints limit the amount of contrast material thatcan be delivered into a single cell without causing celldeath. Consequently, agents with many efficiently relaxingparamagnetic centres confined into a small space areadvantageous over large macromolecules with only fewGd3þ centers (9).Recently, we have introduced the concept of applying

the principle of inverse miniemulsion (25) to encapsulatedifferent clinically approved contrast agents like Magnevist1

(Schering, Germany), Gadovist1 (Schering, Germany) andMultihance1 (Bracco, Italy) in various types of polymeric shells(23), resulting in stable capsules with high concentrations ofcontrast media. The use of the miniemulsion techniqueproduces small stable droplets and subsequent polymerizationof these droplets leads to particles or capsules, which ideallykeep their size (26). In the presented work, contrast agen-t-loaded stable nanocapsules were synthesized using an inverseminemulsion technique. Various types of nanocapsules weresynthesized using different kinds of monomers and contrastagents. The relaxivity values of the resultant nanocapsules weremeasured by MRI. The chemical composition of the polymericnanocapsules was investigated by Fourier transform infraredspectroscopy (FTIR). Location of the contrast agents within thenanocapsules was studied by using various analytical techniqueslike transmission electron microscopy (TEM), energy filteredtransmission electron microscopy (EFTEM), scanning trans-mission electron microscopy (STEM) and energy dispersiveX-ray spectroscopy (EDX) spectroscopy. Further as a proof for thestability of these capsules, the effectiveness of the contrastagent loaded nanocapsules was assessed in human bloodplasma.

www.interscience.wiley.com/journal/cmmi Copyright # 2010 Jo

2. RESULTS AND DISCUSSION

2.1. Effect of variation of monomer ratios on relaxivity ofencapsulated Magnevist1 in polyurethane (PU)nanocapsules

For the assessment of the optimal nanocapsule design, shellproperties such as monomer ratios and the resulting stiffness andcapsule diameters of the shell were investigated. Themain impactof increasing the tolylene-2,4-diisocyanate (TDI) appears to beincreasing stiffness of the shell material. Increasing amount of TDIincreases the number of isocyanate groups (NCO), which can reactwith groups other than the OH of the diol. Reaction sites can eitherbe the meglumine- or N-methyl-D-glucamine-containing reactivegroups, for instance hydroxyl and amine groups, or waterresulting in the formation of urethane and urea, which cansignificantly increase in the stiffness of the formed polymer shell(27). The cross-linked structure formed will impact the chainflexibility, as can be demonstrated by the glass transitiontemperature (Tg) of the nanocapsules, which increases as afunction of the OH:TDI molar ratio of the monomers from 1:2 to1:10 (from �16.7, �15.9, �8.5 to �2.88C). No substantialinfluence on the relaxivity values with increase in amount of TDIwas observed. TEM images of the resultant capsules are found tobe similar without any obvious change in morphology. Forintravascular applications, the size of the nanocapsules should bein the range 150–280nm (28,29). Nanocapsules obtained withvarying diol:TDI ratios are all within this range and show almostsimilar relaxivity. For subsequent studies a molar ratio of diol:TDI(1:2) was used.

2.2. Effect of various biodegradable polymeric shells onrelaxivity of encapsulated Magnevist1

Efficient water exchange through the polymeric shell is one of thekey factors which affect the T1 relaxivity of encapsulated contrastagents. The encapsulation of Magnevist1 in nanocapsulesprepared from different biodegradable polymeric shells includ-ing polyurea, dextran and PU was done in order to investigatetheir relaxivity properties. The size of the resulting nanocapsulesas measured by dynamic light scattering (DLS) is shown inTable 1. The DLS measurements show higher diameters in thecase of dextran shell-encapsulated Magnevist1 nanocapsules ascompared with polyurea and PU shells-encapsulated Magne-vist1. Increased size in the case of dextran-encapsulated contrastagent nanocapsules could be due to the fact that dextran (whichis used as monomer for cross-linked dextran nanocapsules) is apolysaccharide having high molecular weight as compared with1,6-hexanediol (monomer used for PU shell nanocapsules) anddiethylenetriammine (monomer used for polyurea shell nano-capsules) and its structure comprises long chains as comparedwith short chain structures of 1,6-hexanediol and diethylene-triammine. Thus the high molecular weight of dextran, whichleads to a higher viscosity of the aqueous phase, is probablycausing the increased size of obtained aqueous core nanocap-sules.TEM images of these nanocapsules are shown in Fig. 1(a, b) for

polyurea and dextran shell nanocapsules, and in Fig. 2(d) for thePU shell. The T1 relaxivity values of these encapsulated contrastagents were 4.35� 0.04, 4.40� 0.08 and 4.60� 0.03mM

�1 s�1 forMagnevist1 encapsulated in PU, cross-linked dextran andpolyurea nanocapsules as shown in Table 1. The

hn Wiley & Sons, Ltd. Contrast Media Mol. Imaging 2010, 5 59–69

Table 1. Effect of different polymeric shells on diameter and relaxivity of encapsulated Magnevist1 nanocapsules in water phase

Polymeric shell Average size (nm)a PDI Average shell thickness (nm)b Relaxivity (mM�1 s�1) R2c

PU 250.45� 29.15 0.26 24 4.35� 0.04 0.999Dextran 301.25� 30.7 0.54 20 4.40� 0.08 0.999Polyurea 210.00� 30.7 0.29 23 4.60� 0.03 0.998

aMeasured by DLS;bmeasured by TEM.cR2 is measure of goodness of fit of linear regression.

CHARACTERIZATION OF MRI CONTRAST AGENT-LOADED POLYMERIC NANOCAPSULES

non-encapsulated Magnevist1 solution showed a relaxivity of4.08� 0.01mM

�1 s�1. It was observed that all polymericnanocapsules are capable of at least maintaining the relaxivityof the contrast agent. It can be concluded that sufficient waterexchange between water-bound contrast agent and free watermolecules from continuous phase is well obtained through thesepolymeric shells.

2.3. Effect of the type of contrast agent encapsulatedwithin the PU nanocapsules on the relaxivity

Three different types of clinically approved gadolinium-basedcontrast agents [Magnevist1 (gadopenteate dimeglumine),Gadovist1 (gadobutrol), and Multihance1 (gadobenate dime-glumine)] were loaded in PU-shell nanocapsules, which wereprepared with a monomeric diol:TDI ratio of 1:2. These threecontrast agents were chosen because of the differences in theirchemical structures and functionalities. Magnevist1 andMultihance1 have a linear structure and are ionic in nature.In comparison to the gadopentetate chelate structure ofMagnevist1, the gadobenate chelate of Multihance1 ischaracterized by the presence of a (bulky) hydrophobicbenzyloxymethyl substituent. This imparts Multihance1 amarkedly higher T1 relaxivity in plasma among gadolinium-based contrast agents, due to the weak, transient interactionwith serum albumin (30,31). Gadovist1 possesses a cyclicstructure and is non-ionic in nature and is the only available

Figure 1. TEM image of (a) Polyurea encapsulated Magnevist1 nanocapsule

Contrast Media Mol. Imaging 2010, 5 59–69 Copyright # 2010 John

contrast agent with 1.0 M formulation, having a viscosity of 4.96CP at 378C (31–33), while Magnevist1 and Multihance1 areavailable in 0.5 M formulation having a viscosity of 2.9 and 5.3 CPat 378C, respectively. As shown in Table 2, the controlnanocapsules loaded with water and NaCl have an averagediameter of 105 nm. As compared with nanocapsules withoutcontrast agents, nanocapsules loaded with the differentcontrast agents show an increase in size (Table 2). An increasein the size of contrast agent-loaded nanocapsules is probablydue to water diffusion from the continuous aqueous phase intothe core as there is a higher concentration of the dissolvedmolecules in the core (NaCl þ gadolinium complex) ascompared with capsules without contrast agents (23). TEMimages show a dark rim on the inner border of the capsule shellin case of contrast agent-loaded nanocapsules, indicating thepresence of high density gadolinium on the inner surface of theshell. This rim can be observed in the case of air-dried samplesas well as freeze-dried samples (Fig. 2b–d and Fig. 7c). This rimcannot be appreciated in nanocapsules without contrast agent(Fig. 2a). The resultant relaxivity values of these PU encapsu-lated contrast agents are summarized in Table 2. For reference,we measured the relaxitivities of non-encapsulated or freecontrast agent solution under the same experimental con-ditions. The relaxivities values obtained for non-encapsulatedMagnevist1, Gadovist1 and Multihance1 were 4.08� 0.01,4.37� 0.03 and 5.63� 0.09 respectively. The relaxivity values ofthe encapsulated contrast agents show a trend of slight

s in water; (b) dextran encapsulated Magnevist1 nanocapsules in water.

Wiley & Sons, Ltd. www.interscience.wiley.com/journal/cmmi

61

Figure 2. TEM image of (a) PU encapsulated nanocapsules without contrast agents, (b) with contrast agent Gadovist1, (c) with contrast agent

Multihance1, and (d) with contrast agent Magnevist1. The insets show close ups of the respective capsules.

S. SHARMA ET AL.

62

increasing relaxivity values as compared with theirnon-encapsulated counterparts. For Magnevist1 and Multi-hance1 this may be explained by a synergetic effect of osmoticpressure gradient (aqueous core nanocapsules vs aqueouscontinuous phase) rendering improved water exchangethrough the nanocapsules shell and adsorption (or the

Table 2. Effect of PU encapsulated different contrast agents on d

PU nanocapsules Average size (nm)a PDI Avera

Without CAd 105.00� 10.01 0.125With Magnevist1 301.25� 48.45 0.281With Gadovist1 258.80� 11.10 0.308With Multihance1 217.25� 32.25 0.250

aMeasured by DLS;bmeasured by TEM.cR2 is measure of goodness of fit of linear regression;dCA, contrast agent.


entrapment) of the contrast agents to the inner surfaceof the polymeric shell. For Gadovist1, this might be a resultof the high concentration and high viscosity of the Gadovist1

(the concentration is 1 M and viscosity is 8.86 and 4.96 CP at20 and 378C, respectively), resulting in an enhanced waterdiffusion.

iameter and relaxivity in water phase

ge shell thickness(nm)b Relaxivity (mM�1 s�1) R2b

14 NA NA24 4.35� 0.04 0.99922 4.73� 0.04 0.99826 5.64� 0.18 0.998


Figure 3. Schematic view of the PU nanocapsule formation via polyaddition reaction in inverse miniemulsion between diol (from monomer-1),

meglumine and TDI (monomer 2) leading to a partial entrapment of gadopentetate within the polymeric shell (for more detailed description of the

process see Fig. 8).


2.4. Distribution of the Magnevist1 within the PUnanocapsules

The slight trend of increasing relaxivity values of thenanocapsules containing Magnevist1 compared with itsnon-encapsulated counterpart could be explained by thebinding of meglumine into the polymeric shell (Fig. 3). Thechemical binding of the meglumine within the polymer chainwas proven by the formation of nanocapsules from meglumineand TDI. The meglumine was first adjusted to the pH 6.80in order to obtain both protonated and unprotonatedcompounds to mimic similar conditions as for Magnevist1.After performing the polyaddition reaction in the inverseminiemulsion, the obtained nanocapsules possess well-defined

Figure 4. (a) Morphology of aqueous core nanocapsules prepared by poly


core–shell morphology with about 280 nm in diameter (Fig. 4).FT-IR spectral (Fig. 5) analysis in the inverse miniemulsionsystem (with cyclohexane as continuous phase before transferto the water phase) supports this fact. Pure meglumine showscharacteristic peaks at 1389, 1078 and 1051 cm�1 correspond-ing to hydroxyl group from primary alcohol, C–N, C–O stretchfrom aliphatic amine and C–N, C–O stretch from primaryalcohol. After forming nanocapsules, the peak at 2230 cm�1 isdue to the isocyanate group (NCO) present in TDI (monomer 2),and is seen in the resultant polymer when 1,6-hexanediol isused as monomer 1 during synthesis (32). The isocyanate groupis very reactive and is also able to react with water and aminegroups of meglumine at the aqueous miniemulsion dropletinterface. Thus the formation of polyurethane–co-polyurea can

addition of meglumine and TDI. Chemical structure of meglumine (b).


63

Figure 5. IR Spectra of Magnevist1þ TDI, Diolþ TDI, Magnevist1þDiolþ TDI, pure meglumine and meglumineþ TDI.

S. SHARMA ET AL.

64

be observed. The ratios of urethane to urea units, which couldbe estimated from their carbonyl peak intensity at 1700 and1640 cm�1, were 1.11, 0.85, 0.85 and 0.87 corresponding tonanocapsules prepared by diol-TDI, Magnevist1-TDI, diol-Magnevist1-TDI and meglumine-TDI, respectively. Since themeglumine has a strong affinity to the gadopentetate complex,after reacting with TDI at the droplet interface formingthe polymeric shell, it is likely that the Gd3þ complex, whichis completely dissolved and solvated in the aqueous corenanocapsules, stays nearby and/or is entrapped withinthe polymeric shell consisting of meglumine subunits. Thepresence of the isocyanate peak in the resultant polymer shell

Figure 6. Chemical structure of Magnevist1 (gadopentetate dimeglumine)

and TDI (inset shows picture of enlarged capsule).


indicates that TDI still has some open sites for reaction, but acomplete disappearance of the 2230 cm�1 peak was observedwhen both 1,6-hexanediol and Magnevist1 (Gd-DTPA dime-glumine) were added during synthesis. This indicates acomplete consumption of binding sites of the reactiveisocyanate group of the TDI due to the presence of tworeactants. Additionally, the binding was further supported bysynthesis of nanocapsules using Magnevist1 alone (no diol wasused) where meglumine acts as ‘monomer 1’ and TDI asmonomer 2. The structure of Magnevist1 (34) along with TEMimages of the resultant nanocapsules are shown in Fig. 6. Theresulting nanocapsules showed a similar relaxivity value

along with TEM image of nanocapsules synthesized by using Magnevist1


Figure 7. EFTEM image of encapsulated Magnevist1 nanocapsules (a); elemental mapping of Gd3þ in the same capsule (b); TEM image of

encapsulated Magnevist1 nanocapsules after freeze-drying at higher magnification (c) and EDX spectrum (d).


6

(4.46� 0.04mM�1 s�1) to nanocapsules prepared by the

diol-Magnevist1-TDI.The distribution of the contrast agent was visualized by EFTEM

and EDX studies of PU nanocapsule encapsulating Magnevist1

(the nanocapsules were prepared from the diol-Magnevist1-TDI,with OH:TDI ratio from the monomers of 1:2). The EFTEM image(Fig. 7a) of the Magnevist1-loaded PU nanocapsules clearlydemonstrate that an element with a high atomic number (Gd3þ)forms a thin layer mostly on the inner surface of the shell. Theelemental map of Gd3þ (Fig. 7b) also proves the presence of Gd3þ.Further, to avoid diffusion of water during sample preparation thesamples were freeze-dried and EFTEM images along with EDXmeasurements were obtained (Fig. 7c, d). Analysis of the dark rimsection shows the presence of a gadolinium peak at 6.0 kV.

2.5. Stability studies

Since the solubility of the contrast agent complex in organicsolvent is very poor (23), gadolinium complexes are supposed tocompletely stay inside the capsules during the preparation in theinverse miniemulsion process with cyclohexane as continuousphase. However, during the transfer process to water ascontinuous phase, the complex may leak out of these capsuleseither by diffusion or by breakage of the capsules. To prove the


stability of these encapsulated nanocapsules two differentexperiments were performed:

1) d

Wile

ialysis experiment (in water medium) for PU-encapsulatedMagnevist1 loaded nanocapsules;

2) s
tability studies of PU and dextran shell encapsulated Multi-hance1 in the physiological fluid (blood plasma).
Dialysis studies were carried out for PU-encapsulatedMagnevist1 nanocapsules for 5 h and the Gd3þ content of thedialyzed sample was measured by ICP-OES. MRI studies done forthe dilution series (made from the dialysed samples) resulted in arelaxivity value of 4.33� 0.03mM

�1 s�1, indicating a stableencapsulation of the contrast agent.Furthermore, stability of the nanocapsules was studied in

physiological fluid (blood plasma). Multihance1 encapsulated intwo different polymeric shells, i.e. PU and dextran, was redispersedin water and plasma separately and the respective relaxivity valueswere measured. The r1 relaxivity values of non-encapsulatedMultihance1 in water and plasma were 5.63� 0.08 and7.18� 0.05mM

�1 s�1 respectively. The increase of the relaxivityvalues in plasma medium is due to the weak binding ofMultihance1 to the protein serum in human plasma (28,29).PU-encapsulated Multihance1 showed relaxitives of

y & Sons, Ltd. www.interscience.wiley.com/journal/cmmi

5

Figure 8. Schematic representation of the synthesis of contrast agent-loaded polymeric nanocapsules by inverse miniemulsion approach.

S. SHARMA ET AL.

66

5.64� 0.18mM�1 s�1 in water and 5.46� 0.04mM

�1 s�1 in plasma.Dextran-encapsulated Multihance1 showed relaxivities of5.42� 0.06mM

�1 s�1 in water and 5.58� 0.03mM�1 s�1 in plasma.

The results indicate a trend to a small increase of relaxivity of theencapsulated agent in plasma as comparedwithwater. Consideringthe relative large experimental error, this implies that the majorfraction of the encapsulated agent remains within the capsules.

2.6. Gadolinium content per PU nanocapsule containingMagnevist1

Gadolinium content per PU nanocapsules was obtained on thebasis of assumption that all Gd3þ will be encapsulated and allsolid content will be used for forming the polymer. Considering theaverage diameter of nanocapsules to be 200nm with an averageshell thickness of 20nm, the number of capsules obtained fromthe respective solid content was 7.6� 1013. The gadolinium massper capsule resulted as 0.67� 10�15g with the respective numberof gadolinium complexes per capsules of 2.5� 106. This resulted incapsule-based relaxivity of 10.75� 106mM

�1 s�1.

3. CONCLUSION

Polyurethane-, polyurea- and cross-linked dextran-encapsulatedcontrast agents (Magnevist1, Gadovist1 and Multihance1) weresynthesized using an inverse miniemulsion technique. Thediameter of the nanocapsules could be adjusted to the rangebetween 200 and 250 nm by proper adjustment of the processparameters. Maintained or even a trend in slight increase in therelaxivity of the encapsulated contrast agents in the differentbiodegradable polymeric shells indicates an efficient waterexchange through the shells and hence offers flexibility inchoosing the shell material for encapsulation. Various charac-terization techniques were used to locate the contrast agentswithin these nanocapsules. The EFTEM and EDX maps showed amore pronounced Gd3þ signal at the inner walls of the shell,indicating binding of the complex within the shell. FT-IR results


support the binding of the meglumine from the contrast agentcomplex to the polymer shell as the nanocapsules can beprepared through polyaddition reaction in inverse miniemulsionwithout addition of hexane 1,6-diol. Relaxivity values of thedextran- and PU-encapsulated Multihance1 nanocapsules in theplasma medium further support the gadolinium complex beingwell trapped within the nanocapsules. Relaxivity measurementsby MRI studies of these nanocapsules showed a slight trend inincreased relaxivity values for encapsulated contrast agents ascompared with non-encapsulated contrast agents which couldbe either due to the entrapment of contrast agents into thepolymeric shell or due to viscosity of the contrast agent resultingin a slight reduction of the tumbling rate of the gadoliniumcomplex. However, in direct comparison to previously reportedapproaches, the amplification of the ionic relaxivity is still limitedand further modification of the capsules by, for example, theusage of longer chain polymers might be needed for furtherimprovement. The high payload contrast agent per nanocapsulein the order of approximately 2.5� 106 complexes per capsuleresulted in a capsule relaxivity in the order of 10.75� 106mM

�1

s�1, enabling a strong local effect per available binding site.Thus, the simple production process (synthesis of these

nanocapsules is a single step process), the biocompatibility, efficientwater exchange through different polymeric shells, its stability inbiological medium like plasma, the high payload, trend of increasedrelaxivity seen in case of different systems, and the potential forfunctionalization (35,36) make these contrast agent loadednanocapsules a promising versatile vehicle for targeted imaging.

4. EXPERIMENTAL

4.1. Synthesis of nanocapsules

Magnetic resonance contrast agent-loaded nanocapsules wereprepared by an inverse miniemulsion technique (23) (Fig. 8). In atypical synthesis procedure, 0.05 g monomer 1 [1,6-hexanediol(Aldrich) for PU nanocapsules, diethylenetriamine (Aldrich) for



6

polyurea, and dextran (Mw � 70 000 g mol�1, Fluka) forcross-linked dextran nanocapsules] were dissolved in 0.65 gwater (for reference capsules without any contrast agents) orcontrast agent solution [Magnevist1(Gd-DTPA, Schering AGBerlin, Germany), Gadovist1 (Gadobuterol, Schering AG Berlin,Germany), Multihance1 (Gadobenate dimeglumine, Bracco,Italy)] for contrast agent loaded nanocapsules along with0.015 g NaCl (Fischer) in a 5ml glass vial. This mixture wasadded to a surfactant solution of P(B/E-b-EO) [consisting of apoly(butylene-co-ethylene) block (Mw¼ 3700 g mol�1) and apoly(ethylene oxide) block (Mw¼ 3600 g mol�1) synthesized byanionic polymerization (37)] in 3.75 g cyclohexane (VWR). Afterstirring for 1 h, the miniemulsion was homogenized in anice-cooled bath by ultrasonication for 3min at 70% amplitudewith a Branson sonifier W450 digital (6.4mm tip). Then a mixturecontaining 0.14 g monomer 2 [tolylene-2,4-diisocyanate (TDI,Fluka)], surfactant P(B/E-b-EO) and cyclohexane was addedcontinuously over a time period of 300 s with constant stirring. Inthe case of PU and cross-linked dextran capsules, the interfacialpolyaddition was done at 608C and for polyurea capsules at roomtemperature. Additionally, the reference sample prepared bypure meglumine (in the absence of hexanediol and gadopente-tate complex) and TDI via interfacial polyaddition were carriedout in order to confirm the possibility of the polymericnanocapsules formation. The dispersed phase consisted of0.127 g meglumine (equal to its amount present in theMagnevist1 solution), 0.015 g NaCl in 0.65 g distilled water.The pH of the dispersed phase was adjusted to 6.80, to matchthe pH of Magnevist1 solution. All further steps were asdescribed above.To redisperse the nanocapsules in an aqueous medium,

0.6 g of the nanocapsules in the cyclohexane phase weremixed with an aqueous solution of 30mg sodium dodecylsulfate dissolved in 18 g of water. The mixture was stirredfor 1 h at 408C. Then the redispersion was kept in anultrasonication bath for 5min to avoid the aggregation ofthe nanocapsules before further vigorous stirring at 408Cfor 12 h.For the evaluation of the impact of monomer 1 vs monomer 2

ratios on the relaxivity of encapsulated contrast agents, themolar ratio of monomer 1(diol) vs monomer 2 (TDI) was varied.Different nanocapsules were prepared using four differentdiol:TDI ratios (1:2, 1:4, 1:8, 1:10), keeping the amount of diolconstant.

4.2. Preparation of dilution series for reference experimentof gadolinium-based contrast agent

For reference, dilution series of Magnevist1, Gadovist1 andMultihance1 were prepared in water. A stock solutionwith a Gd3þ concentration of 5mM was prepared by adding100ml of Magnevist1 (concentration of commerciallyavailable Magnevist1 is 0.5 M solution) in 10ml of water,50ml of Gadovist1 (concentration of commercial availableGadovist1 is 1 M) solution in 10ml of water and 100ml ofMultihance1 (the concentration of commercially availableMultihance1 is 0.5 M) solution in 10ml of water. This stocksolution was further diluted with a requisite amount of waterto get dilution series ranging from 1 to 0.0625mM of Gd3þ

concentration and the T1 values were measured using a 1.5 TMRI instrument.


4.3. Quantification of gadolinium content

The total concentrations of gadolinium (Gd3þ) in the stock solutionsof encapsulated nanocapsules in aqueous phase were determinedby Inductively Coupled Plasma–Optical Emission Spectroscopy(Varian Vista-Pro, CCD-simultaneous ICP-OES). The sample prep-aration for analysis included a microwave digestion for 1 h at 1508Cafter adding HNO3–aq. H2O2. Then the resultant mixture was usedfor the analysis of Gd3þcontent by ICP-OES. The concentration ofGd3þ-containing solution includes the concentration error of� 4%.

4.4. Characterization by electron microscopy

Transmission electron microscopy experiments were performedon a Philips EM400 microscope operating at 80 kV. A drop ofdiluted aqueous nanocapsules suspension was mounted on400-mesh carbon coated copper grids and dried at roomtemperature. The size distribution of the nanocapsules wasmeasured by DLS using a zeta nanosizer (Malvern Instruments,UK) at a single scattering angle of 1738 and a temperature of258C. The shell thickness was determined by averaging the shellthicknesses of several capsules in TEM images.Energy filtered TEM was investigated using a CM 20 TEM at

200 kV in conventional bright field (BF) mode. Samples wereprepared from concentrated colloidal solutions. A soapy bubblewas inflated from the above solution and a carbon-coated copperTEM grid was placed on it, so that a thin film of the solutioncovered the whole grid. The grid was then dried in air at roomtemperature for 30 h before analysis.TEM analysis was also performed after freeze-drying the

sample. For freeze-drying, samples were adsorbed on a carbon-coated 3mm electron microscopical copper grid and were fastfrozen in liquid propane super-cooled by liquid nitrogen. Laterthe grids were transferred to a freeze etching device BAF 300(Blazers) and freeze-dried at �908C in a vacuum of about 10�6

mbar. The sample was then coated with 6 nm of carbon byelectron beam evaporation. Then these samples were warmed upto room temperature and imaged in a 200 kV Philips TEM at anaccelerating voltage of 10 kV.

4.5. Gadolinium content per PU nanocapsule containingMagnevist1

The number of Gd3þ ions per capsule was calculated on the basisof assumption that all Gd3þ will be encapsulated and all solidcontent will be used for forming the polymer. On the basis of solidmass an overall volume of the polymer is calculated by Pv ¼ PM=rwith r being the density of the polymeric nanocapsule (1.21 gcm3 for PU) and PM¼ 0.19 g being the overall mass of thepolymer. The volume per capsule was calculated from the derivedgeometric properties of the capsule according to

PCv ¼ 4

3p r3c � rc � STð Þ3h i

;

with rc being the radius of the capsule and ST the shell thickness.The number of capsules (nc) was then computed using theformula

nc ¼PvPCv

:

The gadolinium load per capsule (GdCM) was obtainedaccording to GdCM ¼ GdM=nc; with GdM being the overall massof the Gd.


7

S. SHARMA ET AL.

68

4.6. Stability studies

To prove the stability of the nanocapsules two separateexperiments were done. In the first experiment the redispersed(in water) contrast agent-loaded nanocapsules were dialyzedwith an Amicon membrane (exclusion Mw 30 000) against waterfor 5 h at 2000 rpm in a centrifuge. The concentration of Gd3þ inthe resultant solution containing the capsules (supernatantsolution) was quantified using ICP-OES and on the basis of theconcentration so obtained a series of concentrations of Gd3þwasmade. The relaxivity value of the resultant series was thenmeasured by MRI.The second experiment was done in a physiological fluid,

human blood plasma, which was collected from the blood bank.The aim was to prove that nanocapsules were stable inphysiological fluid and that the Gd3þ complexes were wellinside the nanocapsules. For this purpose, two different types ofpolymeric shell (PU and dextran)-encapsulated, Multihance1-loaded nanocapsules were used. These nanocapsules (whichwere in the water phase) were then added in requisite amountto plasma medium. The concentration of Gd3þ in the resultantplasma medium containing the capsules was quantified usingICP-OES. On the basis of the concentration so obtained a series ofconcentrations of Gd3þ was made. The relaxivity value of theresultant series was then measured by MRI.

4.7. MRI relaxivity measurements and data analysis

Relaxivity values of different encapsulated and non-encapsulatedcontrast agents were determined in water and plasma at roomtemperature. All imaging experiments were performed at 1.5 Tona clinical scanner (Intera, Philips Medical Systems) using astandard head coil. SSFP-Look Locker (38–40) technique (1.5/12000/168 echo time, ms/repetition time, ms/flip angle, deg),spatial resolution of 1mm3 and a temporal resolution of 30mswere used to measure longitudinal relaxation rate (1/T1, in s�1).Longitudinal relaxivities (r1) were retrieved from a dilution seriescontaining varying gadolinium (Gd3þ) concentrations in waterand in human plasma. T1 values of different concentrations werecalculated from the MR magnitude data of a region of interest(ROI) in the center of the sample by filling of the data to athree-parameter exponential function (1), applying a Leven-berg–Marquardt optimization algorithm (Mat Lab, Math WorksInc.). The longitudinal relaxation rate (1/T1, in s�1) was plottedagainst Gd3þ concentration (mM) of the sample and the slope wascalculated according to equation (2), yielding the value of r1relaxivity:

SðtÞ ¼ a� b � e�t=T1

�� (1)

1

T1obs¼ 1

T1dþ r1 C½ � (2)

where 1/Tobs1 is the relaxation rate of the sample containing Gd3þ

and 1/Td1 is the relaxation rate of the reference sample withoutGd3þ (i.e. pure water/plasma), C is the concentration of Gd3þ (inmM) and r1 is the slope of the linear correlation (equal to relaxivityin mM

�1 s�1).

Acknowledgements

The authors would like to thank Professor P. Walther (CentralElectron Microscopy Unit, University of Ulm) for freeze-drying


TEM, and Professor U. Kaiser and Dr A. Chuvilin (Electron Micro-scopy Group of Materials Science, University of Ulm) for EFTEM,STEM and EDX analyses. The research fellowship from the Alex-ander von Humboldt Foundation to Umaporn Paiphansiri isgratefully acknowledged.

References

1. Woods M, Woessner DE, Sherry AD. Paramagnetic lanthanidecomplexes as PARACEST agents for medical imaging. Chem SocRev 2006; 35: 500–511.

2. Aime S, Crich SG, Gianolio E, Giovenzana GB, Tei L, Terreno E. Highsensitivity lanthanide (III) based probes for MR-medical imaging.Coord Chem Rev 2006; 250: 1562–1579.

3. Caravan P. Strategies for increasing the sensitivity of gadoliniumbased MRI contrast agents. Chem Soc Rev 2006; 35: 512–523.

4. Bottrill M, Kwok L, Long NJ. Lanthanides in magnetic resonanceimaging. Chem Soc Rev 2006; 35: 557–571.

5. McRobbie DW, Moore EA, Graves MJ, Prince MR. MRI from Pictureto Proton. Cambridge University Press: Cambridge, 2006.

6. Modo MMJ, Bulte JWM. Molecular and Cellular MR Imaging. CRCPress: Danvers, MA, 2007.

7. Werner EJ, Datta A, Jocher CJ, Raymond KN. High-relaxivity MRIcontrast agents: where coordination chemistry meets medicalimaging. Angew Chem Int Ed 2008; 47: 8568–8580.

8. Lauffer RB. Paramagnetic metal complexes as water protonrelaxation agents for NMR imaging: theory and design. ChemRev 1987; 87: 901–927.

9. Livramento JB, Toth E, Sour A, Borel A, Merbach AE, Ruloff R. Highrelaxivity confined to a small molecular space: a metallostar-based, potential MRI contrast agent. Angew Chem Int Ed 2005;44: 1480–1484.

10. Bianchi A, Calabi L, Corana F, Fontana S, Losi P, Maiocchi A, PaleariL, Valtancoli B. Thermodynamic and structural properties of Gd(III) complexes with polyamino-polycarboxylic ligands: basiccompounds for the development of MRI contrast agents. CoordChem Rev 2000; 204: 309–393.

11. MulderWJM, Strijkers GJ, Van Tilborg GAF, Griffioen AW, Nicolay K.Lipid-based nanoparticles for contrast-enhanced MRI and mol-ecular imaging. NMR Biomed 2006; 19: 142–164.

12. Toth E, Helm L, Merbach AE. Relaxivity of MRI contrast agents. TopCurr Chem 2002; 221(Contrast Agents I): 61–101.

13. Wiener E, Brechbiel MW, Brothers H, Magin RL, Gansow OA,Tomalia DA, Lauterbur PC. Dendrimer-based metal chelates: Anew class of magnetic resonance imaging contrast agents. MagnReson Med 1994; 31: 1–8.

14. Bryant LH, Brechbiel MW, Wu C, Bulte JWM, Herynek V, Frank JA.Synthesis and relaxometry of high-generation (G¼ 5, 7, 9 and 10)PAMAM dendrimer–DOTA–gadolinium chelates. J Magn ResonImag 1999; 9: 348–352.

15. Unger E, Shen DK, Wu G, Fritz T. Liposomes as MR contrast agents:pros and cons. Magn Reson Med 1991; 22: 304–308.

16. Turner JL, Pan D, Plummer R, Chen Z, Whittaker AK, Wooley KL.Synthesis of gadolinium-labeled shell- crosslinked nanoparticlesfor magnetic resonance imaging applications. Adv Func Mater2005; 15: 1248–1254.

17. Morawski AM, Winter PM, Crowder KC, Caruthers SD, FuhrhopRW, Scott MJ, Robertson JD, Abendschein DR, Lanza GM, WicklineSA. Targeted nanoparticles for quantitative imaging of sparsemolecular epitopes with MRI. Magn Reson Med 2004; 51:480–486.

18. Cheng Z, Tsourkas A. Paramagnetic porous polymersomes. Lang-muir 2008; 24: 8169–8173.

19. Santra S, Bagwe RP, Dutta D, Stanley JT, Walter GA, TanW,MoudgilBM, Mericle RA. Synthesis and characterization of fluorescent,radio-opaque, and paramagnetic silica nanoparticles for multi-modal bio imaging applications. Adv Mater 2005; 17: 2165–2169.

20. Winter PM, Caruthers SD, Yu X, Song S-K, Chen J, Miller B, BulteJWM, Robertson JD, Gaffney PJ, Wickline SA, Lanza GM. Improvedmolecular imaging contrast agent for detection of human throm-bus. Magn Reson Med 2003; 50: 411–416.

21. Aime S, Frullano L, Crich SG. Compartmentalization of a gado-linium complex in the apoferritin cavity: A route to obtain high



relaxivity contrast agents for magnetic resonance imaging.Angew Chem 2002; 114: 1059–1061.

22. Reynolds CH, Annan N, Beshah K, Huber JH, Shaber SH, LenkinskiRE, Wortman JA. Gadolinium-loaded nanoparticles: new contrastagents for magnetic resonance imaging. J Am Chem Soc 2000;122: 8940–8945.

23. Jagielski N, Sharma S, Hombach V, Mailander V, Rasche V, Land-fester K. Nanocapsules synthesized by miniemulsion techniquefor application as new contrast agent material. Macromol ChemPhys 2007; 208: 2229–2241.

24. Chen HH, Visage CL, Qiu B, Du X, Ouwerkerk R, Leong KM, Yang X.MR imaging of biodegradable polymeric microparticles: a poten-tial method of monitoring local drug delivery. Magn Reson Med2005; 53: 614–620.

25. Crespy D, Stark M, Richter CH, Ziener U, Landfester K. Polymericnanoreactors for hydrophilic reagents synthesized by interfacialpolycondensation on miniemulsion droplets. Macromolecules2007; 40: 3122–3135.

26. Landfester K, Willert M, Antonietti M. Preparation of polymerparticles in nonaqueous direct and inverse miniemulsions.Macromolecules 2000; 33: 2370–2376.

27. Lamba NMK, Woodhouse KA, Cooper SL, Lelah MD. Poly-urethanes in Biomedical Applications. CRC Press: Danvers, MA,1997.

28. Torchilin VP. Polymeric contrast agents for medical imaging. CurrPharm Biotechnol 2000; 1: 183–215.

29. Garnett MC, Kallinteri P. Nanomedicines and nanotoxicology:some physiological principles. Occup Med 2006; 56: 307–311.

30. Schick F, Pintaske J, Martirosian P, Graf H, Erb G, Lodemann KP,Claussen CD. Relaxivity of gadopentetate dimeglumine, gadobu-


trol and gadobenate dimeglumine in human blood plasma at 0.2,1.5 and 3T. Contrast Med Mol Imag 2006; 1: 52–92.

31. Shellock FG, Parker JR, Venetianer C, Pirovano G, Spinazzi A.Safety of gadobenate dimeglumine (MultiHance). Summary offindings from clinical studies and postmarketing surveillance.Invest Radiol 2006; 41: 500–509.

32. Laurent S, Elst LV, Muller RN. Comparative study of the physico-chemical properties of six clinical low molecular weight gado-linium contrast agents. Contrast Med Mol Imag 2006; 1: 128–137.

33. Karabulut N, Elmas N. Contrast agents used in MR imaging of theliver. Diagn Interv Radiol 2006; 12: 22–30.

34. Frullano L, Rohovec J, Peters AJ, Geraldes CFGC. Structure of MRIcontrast agents in solution. Top Curr Chem 2002; 221: 25–60.

35. Zimmerman SC, Zeng F, Reichert DEC, Kolotuchin SV. Self-assembling dendrimers. Science 1996; 271: 1095–1098.

36. Paiphansiri U, Dausend J, Musyanovych A, Mailander V, Land-fester K. Fluorescent polyurethane nanocapsules prepared viainverse miniemulsion: surface functionalisation for use as bio-carriers. Macromol Biosci 2009; 9: 575–584.

37. Schlaad H, Kukula H, Rudloff J, Below I. Synthesis ofa,v-heterobifunctional poly(ethylene glycol)s by metal-free anio-nic ring-opening polymerization. Macromolecules 2001; 34:4302–4304.

38. Look DC, Locker DR. Time saving in measurement of NMR andEPR relaxation times. Rev Sci Instrum 1970; 29: 1237–1246.

39. Crawle AP, Henkelman RM. A comparison of one-shot andrecovery methods in T1 imaging. Magn Reson Med 1998; 7:23–34.

40. Scheffler K, Hennig J. T1 quantification with inversion recoveryTrueFISP. Magn Reson Med 2001; 45: 720–723.


69

Characterization of MRI contrast agent-loaded polymeric nanocapsules as versatile vehicle for...

Documents

Transcript of Characterization of MRI contrast agent-loaded polymeric nanocapsules as versatile vehicle for...