Letter pubs.acs.org/NanoLett
Frequency-Encoded MRI-CEST Agents Based on Paramagnetic Liposomes/RBC Aggregates Giuseppe Ferrauto,†,§ Enza Di Gregorio,†,§ Simona Baroni,† and Silvio Aime*,†,‡ †
Molecular Imaging Center, Department of Molecular Biotechnologies & Health Sciences, University of Torino, Torino, Italy Institute for Advanced Study, Technische Universitat Munchen, Munich, Germany
‡
S Supporting Information *
ABSTRACT: Paramagnetic liposomes containing DyHPDO3A in their inner water compartment and carrying a residual positive charge on their outer surface have been electrostatically bound to the membrane of red blood cells (RBCs). These aggregates yield two chemical exchange saturation transfer (CEST) pools represented by liposomal water protons (LipoCEST) and cytoplasmatic water protons (ErythroCEST), respectively. The absorption frequencies of the two pools fall at the negative and positive side of the solvent water resonance as expected from the dipolar (LipoCEST) and BMS (bulk magnetic susceptibility) (ErythroCEST) origin of the paramagnetic induced shift of their water protons resonances, respectively. In vivo magnetic resonance imaging (MRI) shows that the liposomes/RBC aggregates report about the vascular volume whereas the residual LipoCEST effect informs about the presence of released liposomes in the region of interest (ROI). Besides being an innovative blood cell labeling for MRI, the LipoCEST/RBC aggregates provide a route to improve the circulation lifetime of the liposomes and the CEST procedure allows assessing the deassembly of the aggregates and accumulation of the liposomes in the ROI. KEYWORDS: Red blood cells, paramagnetic liposomes, MRI, CEST contrast agents, drug delivery, bulk magnetic susceptibility
R
The SR induces a dipolar shift to the liposomal water resonance that makes it distinguishable from the “bulk” water signal. The proper choice of saturated/unsaturated phospholipids in the liposome membrane composition allows the control of the water exchange rate across the liposomal membrane, in order to optimize the saturation transfer (ST) in the CEST experiment. The extension of this concept leads us to use the even larger cytoplasmatic water pool as source of the exchangeable proton pool by loading RBCs with paramagnetic molecules.19 Actually, the discoidal shape of RBCs allows the efficient exploitation of the BMS contribution to induce marked shifts of the intracellular water resonance.16−18 It has been reported that this system acts efficiently as CEST probe with a sensitivity that is the highest shown by CEST agents in the fM−pM range. The availability of erythroCEST agents opens interesting possibilities to visualize blood circuit by the MRICEST modality. Of course, some concern may be raised about the long-term stability of the paramagnetic complexes entrapped in the RBC that may affect the toxicity of these agents. This prompted us to rethink the erythroCEST agents by inducing the BMS effect from the external side of the RBCs. Herein it is shown that this task can be well addressed by anchoring cationic liposomes loaded with paramagnetic
ecently, chemical exchange saturation transfer (CEST) magnetic resonance imaging (MRI) agents have been under intense scrutiny for applications in molecular imaging and in MRI guided drug delivery.1−4 CEST agents generate a frequency-encoded contrast in MRI by irradiating with a selective radiofrequency field, the exchangeable protons of interest. Various CEST contrast agents (CAs) have been developed and are now under scrutiny for in vivo applications, including both diamagnetic (diaCEST) and paramagnetic (paraCEST) molecules.5−7 This type of CAs have received much attention by virtue of their capability, unique for 1H-MRI scanners, to make possible the multiplex visualization of more than one probe present in the same region, simply by changing the irradiation offset.8−10 Moreover, CEST molecules responsive to microenvironmental parameters (pH, temperature, enzymes, metal ions, etc.) have also been developed and applied in preclinical studies.11−14 In the past few years, CEST CAs with a large number of equivalent exchangeable protons have been intensively investigated to increase the sensitivity (typically in the millimolar range for exchanging protons), making it possible to overcome the intrinsic drawback of the MRI-CEST approach. Liposomes that contain 106−109 water molecules have been proposed in 2007 as CEST probes, provided that a paramagnetic shift reagent (SR) is selectively entrapped in their inner aqueous cavity (LipoCEST).15−18 © XXXX American Chemical Society
Received: July 14, 2014 Revised: October 19, 2014
A
dx.doi.org/10.1021/nl5026612 | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
Letter
Figure 1. (Top) Formation of LipoCEST/RBCs aggregates. (Bottom) Chemical structure of the shift reagent Ln-HPDO3A (Ln = Eu- or Dy-).
same size and membrane composition whose aqueous phase is deprived of the paramagnetic shift reagent have been prepared and anchored on the outer membrane of RBCs. No CEST property is present in control specimens (b-Ctrl-lipo, Figure 2 black). Conversely, in the case of RBCs/Dy-liposomes aggregates, two peaks are clearly present at +4.8 ppm and −4.2 ppm, respectively. The CEST result has been reported in Supporting Information Figure.S12. For the positive peak, a CEST effect of ca. 50% is present. The negative peak is assigned to the water inside the liposome (shift determined by the dipolar contribution generated by Dy-HPDO3A). The magnitude of the shift (−4.2 ppm) is consistent with what expected for a liposomes containing ca. 250 mM of Dy-HPDO3A in the inner cavity (the shift given by Dy-HPDO3A is ca. −19 ppm/M). The positive chemical shift centered at +4.8 ppm from bulk water signal is ascribable to intracellular water shifted by the changes in magnetic susceptibility due to the paramagnetic Dy-liposomes anchored on the RBCs’ outer surface. The two peaks show comparable ST% effects. The proton transfer efficiency (PTE) for compartmental exchange systems (such as LipoCESTs or ErytroCESTs) is the result of a complex interplay of different factors7 among which the volume of the compartment and the Kex are the most important ones. Despite the much larger volume of RBC, the ST% shown by the two compartments appears similar. The observed behavior can be accounted in terms of a very high water permeability of the liposomal membrane due to the presence of the unsaturated DOTAP component in the bilayer composition. Kex has been estimated to be ca. 700 Hz for liposomes whereas it is equal to ca. 25 Hz for RBC. Moreover, it is likely that the presence of liposomes on the surface of RBC may further hamper the water exchange in the latter systems. To get more insights into the origin of the two CEST peaks shown by the RBCs/Dy-loaded liposomes the following experiments have been carried out: (i) Dy-HPDO3A-loaded RBCs have been used in place of native RBCs. The procedures for the preparation and assessment of the CEST properties of erythroCEST systems has been recently reported.19,20 Herein, Dy-HPDO3A-loaded RBCs display the CEST peak at +4.5 ppm (Figure 3A,B blue lines) as expected for an intracellular concentration of the
complexes to the negatively charged molecules present on the outer side of the RBCs’ membrane (Figure 1). Positively charged liposomes have been obtained by hydrating a thin layer mixture of phospholipids (DPPC/ DOTAP/cholesterol, 75/15/10 molar ratio) with an aqueous solution of the neutral, stable, and well-tolerated complex LnHPDO3A (300 mM, 300mOsm/L; Ln = Dy or Eu).21 Liposomes have been extruded and the resulting unilamellar vesicles have been purified by dialysis. The intraliposomal concentration of Ln-HPDO3A complexes corresponds to ca. 250 mM as evaluated by high resolution NMR spectra of the liposomes (Supporting Information Figure.S10). The extensive characterization of these liposomes is reported in Supporting Information (Figure.S1−S9, Table S1). RBCs were incubated for 10 min at 4 °C in the presence of Dy-HPDO3A-loaded liposomes, then washed three times with fresh saline phosphate buffer (PBS) solution. This procedure allows the binding of Dy-liposomes on the surface of the RBC. Upon binding, the RBC cell suspension has been diluited 1:6 in PBS (v/v) to acquire CEST images. Z-spectra have been recorded at 7.1 T, 21 °C, pH 7 by using a Bruker 300 MHz spectrometer equipped with a microimaging probe. A rapid acquisition with relaxation enhancement (RARE) sequence was preceded by the presaturation scheme that consists of a continuous, rectangular block pulse (B2 = 3 μT, irradiation time 2 s). The Z-spectra of RBCs/Dy-liposomes aggregates are reported in Figure 2 (red). As the control, liposomes of the
Figure 2. Z-spectra of RBCs with Dy-loaded liposomes (red) or control liposomes (black) anchored on their surface. B
dx.doi.org/10.1021/nl5026612 | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
Letter
Figure 3. (A) Z-spectra and (B) LD% of Dy-RBCs (blue), RBCs with bH-Dy-lipo (red), and Dy-RBCs with bH-Dy-lipo (green). (C) Z-spectra of a suspension of Dy-TTHA-loaded liposomes. (D) Z-spectra of a RBCs with (red line) and without (black line) Dy-TTHA-loaded liposomes anchored on cell membrane. (E) Z-spectra of RBCs with bL-Dy-lipo (light blue), sH-Dy-lipo (violet) and bH-Dy-lipo (red) anchored on the surface. (F) Intracellular water chemical shift versus number of Dy-liposomes anchored on the cell surface. The arrows indicate the data points which Z-spectra are reported in E.
negligible dipolar shift of the water resonance of its aqueous solution, and when entrapped in the aqueous compartment of spherical liposomes it does not generate a LipoCEST agent as shown by its Z-Spectrum (Figure 3C). However, in the asymmetric environment of the Dy-TTHA/RBC aggregates, these paramagnetic liposomes are able to induce a marked magnetic susceptibility effect on the cytoplasmatic water resonance of RBCs as shown by the CEST peak at 3.5 ppm (Figure 3D). (iii) Eu-HPDO3A-loaded liposomes (of the same size, composition and concentration of the paramagnetic agent as Dy-HPDO3A-loaded systems) have been used to form the aggregates with RBCs. The smaller μeff value of the Eu(III) ions with respect to Dy(III)ions (3.5 vs 10.6) does not generate a sufficient magnetic susceptibility change to induce a detectable shift in the water resonance of RBCs (Supporting Information Figure.S14). (iv) To assess the relationship between the paramagnetic loading of the liposomes and the shift of the CEST peak due to the RBC water resonance two types of Dy-HPDO3A-loaded liposomes (bH-Dy-lipo and sH-Dy-lipo) have been compared.
paramagnetic agent of ca. 3 mM. Upon anchoring DyHPDO3A loaded liposomes on the surface of the DyHPDO3A-loaded RBCs, the Z-spectrum shows two CEST peaks at −4.2 ppm (due to the LipoCEST) and at +9pmm (due to the water inside RBCs, Figure 3A,B green lines). The latter shift appears then as the sum of two BMS contributions, one arising from the paramagnetic agents inside the RBCs and one arising from the paramagnetic agents in the liposomes anchored on RBC membranes. The control made by empty RBCs with Dy-liposomes anchored on the surface shows two CEST peaks at −4.2 ppm (due to the LipoCEST) and at +4.8 ppm (due to the water inside RBCs, Figure 3A,B red lines). The latter shift is given by the BMS contributions arising from the paramagnetic agents in the liposomes anchored on RBC membranes. (ii) Dy-TTHA loaded liposomes have been used (chemical structure of Dy-TTHA is reported in Supporting Information Figure.S13) to form the aggregates with the RBCs. Lanthanide (III) complexes of TTHA are known as only outer sphere complexes as the wrapping of the ligand around the metal ion excludes any coordination of water molecules in the inner coordination sphere.22 It follows that Dy-TTHA displays a C
dx.doi.org/10.1021/nl5026612 | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
Letter
Figure 4. (A) Z-spectrum of in vivo liposome/RBCs aggregates, at t = 0 (left) and t = 1 h (right), reporting positive (3.2 ppm) and negative (−4.2 ppm) signals belonging to RBCs and LipoCESTs, respectively. CEST% map of tumor region with irradiation RF offset at 3.2 ppm at t = 0 (B) and t = 1 h (C). CEST% map with irradiation RF offset at −4.2 ppm signal at t = 0 (D) and t = 1 h (E). The ROI has been circled with a white line.
Supporting Information). A third kind of Dy-HPDO3A-loaded liposomes (bL-Dy-lipo) has been considered for the binding with RBCs surface. These liposomes are large-sized (hydrodynamic diameter = 180 nm) and poorly loaded (ca. 100 mM) with Dy-HPDO3A complexes. Upon interaction with these liposomes, no sign of shift of intracellular water signal is detected (Figure 3E light blue line). As a proof of concept on the potential “theranostic” role of the LipoCEST/RBC aggregates developed in this work, it has been deemed of interest to visualize the release of the anchored liposomes in the tumor region. RBCs with bH-Dy-liposomes anchored on the surface (ca. 300 bH-Dy-liposomes/RBC) have been administered to a TSA breast cancer bearing mouse by iv injection in the tail vein. CEST maps have been acquired at t = 0 and t = 1 h after the LipoCEST/RBCs administration. After administration of RBCs/bH-Dy-lipo, at t = 0 a clear CEST% effect is present when the irradiation frequency is set at 3.2 ppm (mean CEST% in the entire tumor is ca. 9.5% Figure 4B). A similar CEST% effect has been obtained for the entire tumor
bH-Dy-lipo are large-sized (hydrodynamic diameter = 200 nm) systems whereas sH-Dy-lipo are small-sized ones (hydrodynamic diameter = 120 nm). Both these systems contain a high concentration of Dy-HPDO3A (ca. 250 mM). Figure 3E,F reports the chemical shift of the RBCs water resonance as a function of the number of liposomes anchored on their surface (red lines for bH-Dy-lipo and violet line for sHDy-lipo). Clearly, the higher the paramagnetic loading the higher is the shift of the CEST peak. In the case of sH-Dy-lipo/ RBC aggregates, the water shift is lower than the one of bH-Dylipo/RBC aggregates. In fact, bH-Dy-lipo can induce a high chemical shift close to 7 ppm in the case of ca. 300 liposome vesicles anchored on RBCs surface; the smaller liposomes (sHDy-lipo) induce a shift that reaches only 4 ppm by anchoring ca. 550 liposomes vesicles on the RBCs surface. These results indicate that higher is the number of paramagnetic liposomes attached to cell membrane the higher is the induced chemical shift. Analogously, bigger, highly loaded paramagnetic liposomes are able to induce a larger susceptibility effect (see D
dx.doi.org/10.1021/nl5026612 | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
Letter
region when the irradiation offset is shifted to −4.2 ppm, that is, the position of the intraliposomal water resonance. These findings indicate that at t = 0 the LipoCEST agents are bound to the RBC outer membrane and the images report about the extent of vascularization in the ROI. The chemical shift of the intracellular shifted water is at 3.2 ppm and not, as expected, at 6−7 ppm. This finding can be accounted for in terms of the redistribution of liposomes from the injected labeled-RBCs to the endogenous unlabeled RBCs (by considering that the administrated liposomes-anchored RBCs are ca. 10% of the total number of endogenous RBCs). The CEST% effect due to ErythroCEST in the entire tumor region rapidly decreases and at 1 h the mean CEST% at 3.2 ppm is ca. 2%. Conversely, the LipoCEST water signal at −4.2 ppm does not change so much and at t = 1 h its CEST% effect is still good (ca. 6%). This observation suggests that at t = 1 h more than half of the LipoCEST have been detached from RBCs and likely are extravasated in the tumor extracellular matrix. It has to be taken in account that in principle the CEST results may be affected by NOE-relayed effects.23 To get rid of the endogenous CEST interferences from RBCs in place of using the absolute ST values we considered the difference between post- and pre-iv injection of Liposome/RBC aggregates. CEST response arising from liposomes/RBCs aggregates are much more intense than endogenous NOE effects. The endogenous signals remain almost unaltered before and after the administration of liposomes/RBCs aggregates. Conversely, the LipoCEST and ErythroCEST signals are present only after the administration of the contrast agent. To our knowledge, the idea of anchoring liposomes on the outer surface of RBCs is new and it may open new avenues for drug delivery systems (the electrostatic binding may be replaced by cleavable spacer responsive to endogenous or external stimuli). Moreover, our results show that when LipoCESTs are used, the MRI visualization provides key information on the biodistribution of the aggregates and on the availability of the liposomes at the site of interest.
■
farmaci e la valutazione contestuale della risposta mediante imaging funzionale) and by the EU COST Action TD1004. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS Support from Bracco Imaging S.p.A. and CIRCMSB (Consorzio Interuniversitario di Ricerca sulla Chimica dei Metalli nei Sistemi Biologici) is also gratefully acknowledged. We are grateful to Dr. Marta Tapparo and Dr. Cristina Grange for blood supply.
■
ABBREVIATIONS MRI, magnetic resonance imaging; CEST, chemical exchange saturation transfer; BMS, bulk magnetic susceptibility; RBCs, red blood cells; CA, contrast agents; μeff, effective magnetic moment; ROI, region of interest
■
ASSOCIATED CONTENT
S Supporting Information *
Methods (chemicals; liposomes preparation; isolation of RBCs and liposomes binding; binding of fluorescent liposomes to RBCs; ICP-MS measurements of metal content; MRI experiments and data analysis; in vivo experiments), supporting results (liposomes characterization; LD analysis; Dy-TTHAloaded liposomes; Z-spectra and LD% of RBCs+ bH-Eu-lipo; different number of bH-Dy-lipo anchored on RBCs surface; binding of fluorescent liposomes to RBCs; stability of DyHPDO3A-loaded liposomes; biodistribution of Gd-liposomes/ RBC aggregates), and supporting references. This material is available free of charge via the Internet at http://pubs.acs.org.
■
REFERENCES
(1) Ward, K. M.; Aletras, A. H.; Balaban, R. S. J. Magn. Reson. 2000, 143, 79−87. (2) Sherry, A. D.; Woods, M. Annu. Rev. Biomed. Eng. 2008, 10, 391− 411. (3) Langereis, S.; Keupp, J.; van Velthoven, J. L.; de Roos, I. H.; Burdinski, D.; Pikkemaat, J. A.; Grüll, H. J. Am. Chem. Soc. 2009, 131, 1380−1. (4) Burdinski, D.; Pikkemaat, J. A.; Emrullahoglu, M.; Costantini, F.; Verboom, W.; Langereis, S.; Grüll, H.; Huskens, J. Angew. Chem., Int. Ed. 2010, 49, 2227−9. (5) Terreno, E.; Delli Castelli, D.; Aime, S. Contrast Media Mol. Imaging 2010, 5, 78−98. (6) Zhang, S.; Merritt, M.; Woessner, D. E.; Lenkinski, R. E.; Sherry, D. Acc. Chem. Res. 2003, 36, 783−790. (7) van Zijl, P. C.; Yadav, N. N. Magn. Reson. Med. 2011, 65, 927−48. (8) Ferrauto, G.; Delli Castelli, D.; Terreno, E.; Aime, S. Magn. Reson. Med. 2013, 69, 1703−11. (9) Aime, S.; Carrera, C.; Delli Castelli, D.; Geninatti Crich, S.; Terreno, E. Angew.Chem., Int. Ed. 2005, 44, 1813−1815. (10) McMahon, M. T.; Gilad, A. A.; DeLiso, M. A.; Cromer Berman, S. M.; Bulte, J. W. M.; van Zijl, P. C. M. Magn. Reson. Med. 2008, 60, 803−812. (11) De Leon-Rodriguez, L. M.; Lubag, A. J.; Malloy, C. R.; Martinez, G. V.; Gillies, R. J.; Sherry, A. D. Acc. Chem. Res. 2009, 42, 948−57. (12) Delli Castelli, D.; Ferrauto, G.; Cutrin, J. C.; Terreno, E.; Aime, S. Magn. Reson. Med. 2014, 71, 326−32. (13) Yoo, B.; Sheth, V. R.; Howison, C. M.; Douglas, M. J.; Pineda, C. T.; Maine, E. A.; Baker, A. F.; Pagel, M. D. Magn. Reson. Med. 2013, DOI: 10.1002/mrm.24763. (14) Chen, L. Q.; Howison, C. M.; Jeffery, J. J.; Robey, I. F.; Kuo, P. H.; Pagel, M. D. Magn. Reson. Med. 2013, DOI: 10.1002/mrm.25053. (15) Terreno, E.; Delli Castelli, D.; Violante, E.; Sanders, H. M.; Sommerdijk, N. A.; Aime, S. Chemistry 2009, 15, 1440−8. (16) Aime, S.; Delli Castelli, D.; Terreno, E. Methods Enzymol. 2009, 464, 193−210. (17) Terreno, E.; Cabella, C.; Carrera, C.; Delli Castelli, D.; Mazzon, R.; Rollet, S.; Stancanello, J.; Visigalli, M.; Aime, S. Angew. Chem., Int. Ed. 2007, 46, 966−8. (18) Peters, J. A.; Huskens, J.; Raber, D. J. Prog. Nucl. Magn. Reson. Spectrosc. 1996, 28, 283. (19) Ferrauto, G.; Delli Castelli, D.; Di Gregorio, E.; Langereis, S.; Burdinski, D.; Grüll, H.; Terreno, E.; Aime, S. J. Am. Chem. Soc. 2014, 136, 638−41. (20) Di Gregorio, E.; Ferrauto, G.; Gianolio, E.; Aime, S. Contrast Media Mol. Imaging 2013, 8, 475−86. (21) Di Gregorio, E.; Gianolio, E.; Stefania, R.; Barutello, G.; Digilio, G.; Aime, S. Anal. Chem. 2013, 85, 5627−31.
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Tel:+390116706451. Author Contributions §
G.F. and E.D.G. contributed equally.
Funding
This research was founded by the AIRC Investigator Grant IG2013, by the University of Genova (Progetto San Paolo; Title: Validazione di molecole per il rilascio tumore specifico di E
dx.doi.org/10.1021/nl5026612 | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
Letter
(22) Caravan, P.; Ellison, J. J.; McMurry, T. J.; Lauffer, R. B. Chem. Rev. 1999, 99, 2293−2352. (23) Vinogradov, E.; Sherry, A. D.; Lenkiski, R. E. J. Magn. Reson. 2013, 229, 155−172.
F
dx.doi.org/10.1021/nl5026612 | Nano Lett. XXXX, XXX, XXX−XXX
