Comparing Strategies in the Design of Responsive Contrast Agents for

Jun 14, 2017 - be exploited to design responsive MRI contrast agents. The majority ..... Society and a New Talent: Americas by the Royal Society of Ch...
0 downloads 0 Views 6MB Size
Article Cite This: Acc. Chem. Res. 2018, 51, 342−351

pubs.acs.org/accounts

Comparing Strategies in the Design of Responsive Contrast Agents for Magnetic Resonance Imaging: A Case Study with Copper and Zinc Valérie C. Pierre,* Sarah M. Harris, and Sylvie L. Pailloux Department of Chemistry, University of Minnesota, 207 Pleasant Street SE, Minneapolis, Minnesota 55455, United States CONSPECTUS: Magnetic resonance imaging (MRI) has emerged over the years as one of the preferred modalities for medical diagnostic and biomedical research. It has the advantage over other imaging modalities such as positron emission tomography and Xray of affording high resolution three-dimensional images of the body without using harmful radiation. The use of contrast agents has further expanded this technique by increasing the contrast between regions where they accumulate and background tissues. As MRI most often measures the relaxation rate of water throughout the body, contrast agents function by modulating the intensity of the water signal either via improved relaxation or via saturation transfer to selected exchangeable proton. Among the growing class of MRI contrast agents, a subset of them called “smart” contrast agents function as responsive probes. Their ability to increase or decrease their signal intensity is modulated by the presence of an analyte. These probes offer the unique ability to image the distribution of an analyte in vivo, thereby opening new possibilities for diagnostics and for elucidating the role of specific analytes in various pathologies or biological processes. A number of different strategies can be exploited to design responsive MRI contrast agents. The majority of contrast agents are based on GdIII complexes. These complexes can be rendered responsive in either of two ways: either by modulating the number of inner-sphere water molecules, q, or via modulating the rotational correlation time, τR, of the contrast agent upon substrate binding. The longitudinal relaxivity increases with the number of inner-sphere water molecules. GdIII complexes can be rendered responsive if they contain a recognition moiety that can bind to both the open coordination site of GdIII and to the analyte. When the recognition moiety leaves the lanthanide ion to bind to the analyte, q increases and therefore so does the relaxivity. The dependence of relaxivity on rotational correlation time is more complex and more pronounced at lower magnetic fields. In general, slower tumbling macromolecules have longer rotational correlation times and higher relaxivities. Analyte-triggered formation of macromolecules thus also increases relaxivity. Such macromolecules can either be analyte-templated supramolecular assemblies, or analyteenhanced protein-contrast agent complexes. Chemical Exchange Saturation Transfer (CEST) agents are a newer class of contrast agents that offer the possibility of multifrequency and thus ratiometric imaging, which in turn enables quantitative mapping of the concentration of an analyte in vivo under conditions where the concentration of the contrast agent is not known. Such agents can be rendered responsive if the analyte changes the number of exchangeable proton(s), its exchange rate, or its chemical shift. All of these approaches have been successfully employed for detecting and imaging both copper and zinc, including in vivo. Magnetic Iron Oxide Nanoparticles (MIONs) are powerful MRI transverse relaxation agents. They can also be rendered responsive to an analyte if the latter can control the aggregation of the nanoparticles. For metal ions, this can be achieved via chemical functionalities that only react to form conjugates in the presence of the metal ion analyte.



IMAGING COPPER AND ZINC Zinc and copper are essential to several biological pathways. Disruption of their homeostasis is implicated in the pathology of Alzheimer’s, Parkinson’s, and Wilson’s diseases, Menkes syndrome and certain tumors. The large difference in concentration between healthy and diseased tissues render them good targets for the development of in vivo imaging agents. Before designing imaging probes, though, one should also consider the sensitivity of the technique. MRI is poorly sensitive, and the limits-of-detection of clinical contrast agents is ∼0.1 mM. One must therefore also keep in mind how the limit of detection of the probe compares to the anticipated concentration of the analyte. In certain cases, the pool of labile copper and zinc reaches concentrations that are within the range of sensitivity of MRI contrast agents. This is the case for copper and zinc which © 2018 American Chemical Society

are present in 10−100-fold higher concentrations in Aβ plaques than in the extracellular environment of a healthy brain, at concentrations of up to 0.4 mM for Cu2+ and 0.2−1 mM for Zn2+. It is because the concentrations of zinc can reach >0.1 mM in the pancreas that it could be mapped by magnetic resonance imaging (MRI).



GADOLINIUM-BASED RESPONSIVE CONTRAST AGENTS Gadolinium has seven unpaired electrons and a long electronic relaxation time, making it an ideal candidate as a relaxation agent. The toxicity of gadolinium, however, imposes that the lanthanide Received: June 14, 2017 Published: January 22, 2018 342

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351

Article

Accounts of Chemical Research ion be tightly chelated in order to prevent leaching into the body. Advantageously, the substantial effect that the ligand has on the properties of the GdIII complex can be readily exploited to design responsive contrast agents. The image-enhancing capability of a gadolinium contrast agent is directly proportional to its longitudinal relaxivity, r1, which is the relaxation rate increase of bulk water-molecules by the paramagnetic ion per unit concentration of the Gd complex. This relaxivity includes both an inner-sphere and an outer-sphere component. The former arises from water molecules directly coordinated to the GdIII ion which exchanges with those of the bulk. The latter is the result of water molecules that interact with the GdIII center according to a dipolar intermolecular mechanism. The longitudinal proton relaxation rate is described by the Solomon−Bloembergen− Morgan (SBM) theory. A complete description of this theory and a review of Gd-based responsive contrast agents can be found elsewhere.1,2 A responsive GdIII contrast agent is one whose relaxivity varies as a function of the presence of an analyte. From the SBM equations, there are only two parameters that can be exploited, sometimes simultaneously, to render GdIII complexes responsive: q, the number of inner-sphere water molecules, and τR, the rotational correlation time (Figure 1). For macromolecular

Figure 2. Design of q-based responsive contrast agents.

Importantly, the relaxivity of a GdIII contrast agents includes both an inner-sphere and an outer-sphere component.1 A change in q, the number of inner-sphere water molecules, only affects the inner-sphere component. Therefore, even in the “off” state with q = 0, the probe can have a substantial outer-sphere relaxivity which is, for the most part, unaffected by the presence or absence of the analyte. This class of probes should therefore not be expected to behave as “relaxometric light-switches”. Moreover, the requirements with respect to the safety of the probe for in vivo use governs that the GdIII complex remains stable in the “on” state.3 This stability is due in large part to the number of coordination sites on the metal occupied by the ligand. Therefore, q cannot be increased at will. In practice, complexes with up to 2 inner-sphere water molecules are stable enough for in vivo use. As such, q-based responsive contrast agents are limited to a Δq of 2. The combination of these two factors indicates that the maximum response of q-based responsive contrast agents remains modest. This small increase in r1 is compounded by the difficulty in mapping independently the distribution of the contrast agent in vivo. GdIII contrast agents can only be detected via one parameter: the T1 of the protons of water. For such probes, this parameter is a function of both the concentration of the analyte and that of the probe. When the concentration of the agent in a tissue can vary as much as the response to the analyte, one must be careful with any conclusion drawn from in vivo imaging. This class of responsive contrast agents includes the first responsive contrast agent for zinc, Gd-1 (Figure 3).4 Gd-1 is a DTPA derivative bearing two (2-pyridylmethyl)amine moieties which are extensively used for fluorescent ZnII probes.5 Addition of 1 equiv of zinc initially decreases r1, whereas a second one increases r1 back to its original value. Accordingly, the initial complex has a q = 1, the first equivalent of ZnII blocks solvent access to the GdIII center, and the second equivalent of ZnII reopens it. Gd-2 was designed in order to resolve this “on”−“off”−“on” issue. It contains only two pyridyl moieties, the other two having been replaced by carboxylates.6 This probe functions essentially via the same mechanism as Gd-1, however, since only two pyridyl are present, the probe can only bind one equivalent of ZnII resulting in an “on”−“off” signal, albeit with a small decrease in r1 (30%). The first true “off”−“on” q-based GdIII-responsive contrast agent for ZnII was obtained with Gd-3.7 Gd-3 contains a recognition moiety with two carboxylates that can coordinate weakly to GdIII, thereby replacing its inner-sphere water molecule. In binding zinc, the carboxylates release the GdIII, which opens up one site on the lanthanide ion for water coordination. As a result, r1 doubles from 2.33 to 5.07 mM−1 s−1. Further investigation demonstrated the requirements for efficient ZnII recognition and response. Evaluations of the analogues Gd-4, Gd-5, and Gd-6 indicated that at least one appended aminoacetate must be present on the recognition motif to bind GdIII and inhibit water coordination in the “off”

Figure 1. Parameters influencing the inner-sphere relaxivity of GdIII contrast agents.

agents where τR is long, τm, the water residence time, can also be used to modulate r1. Although it is not necessary, most responsive GdIII contrast agents, including those for copper and zinc, employ a polyaminocarboxylate ligand. These types of ligands, which are used in all clinical GdIII contrast agents, have the advantages of forming stable complexes which are sufficiently water-soluble for most applications. The response of q-based responsive contrast agents is modulated by the number of inner-sphere water molecules (Figure 2). This class of contrast agent simply requires the addition of a reporting functionality on the ligand which must be able to coordinate both GdIII and the analyte with, importantly, higher affinity for the analyte. The principle governing these probes is simple. In the absence of analyte, the functional group coordinates GdIII, thereby displacing any coordinated water molecule. In the presence of analyte, the functional group releases GdIII in favor of the analyte, consequently opening up a coordination site on the lanthanide ion which immediately fills with water. 343

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351

Article

Accounts of Chemical Research

Figure 3. q-Based GdIII-responsive contrast agents for zinc.

state.8 Moreover, two binding moieties, carboxylates or pyridines, are required for efficient ZnII binding. The length of the linker between the recognition element and the GdIII chelate also affects the response of the probe.9 Increased effectiveness requires efficient coordination of the GdIII in the absence of the analyte, which in turn requires linkers of intermediate flexibility. The linker, however, affects neither the affinity nor the selectivity of the recognition moiety for its analyte. These conclusions likely also hold true in the design of other q-based responsive GdIII contrast agents. Similar responsive contrast agents for ZnII have been developed. Gd-7 turns-on 30−40% in the presence of both CaII and ZnII, demonstrating the difficulty in achieving selectivity for ZnII only with carboxylate donors.10 The response of this probe is affected by the presence of carbonate anions, which can also coordinate GdIII and replace the inner-sphere water molecule in the “on” state, thereby making it less bright. This drawback is common to all neutral or positively charged q-based GdIII-responsive contrast agents. The most recent example, Gd8, incorporates the best of the previous probes, with a linker of medium length and a rigid macrocycle for ZnII binding.11 It

shows the highest turn-on relaxivity for this class: 150% increase upon addition of 1 equiv of ZnII at high field while maintaining high selectivity over both CaII and MgII. This approach can also be applied to transition-metal based T1 agents, as exemplified by the MnIII-porphyrin complex, Mn-9.12 All probes have excellent selectivity for Zn2+ over Ca2+, a necessity given the higher extracellular concentrations of Ca2+ (1−2 mM). q-Based responsive contrast agents for CuI and CuII were developed at the same time as their ZnII-targeted counterparts. Differentiation between CuI and CuII is possible via the recognition motif. CuI, a soft metal, favors softer ligands; CuII, which is harder, favors harder ligands. Therefore, all CuIresponsive Gd-based contrast agents, Gd-10−Gd-15, incorporate thioether ligands (Figure 4).14 Replacement of one thioether by a harder carboxylate is detrimental to selectivity for CuI; r1 of Gd-17 increases more in the presence of CuII than CuI.14 The design of this class of probes was improved in two ways: first by increasing the change in the number of inner-sphere water molecules from 1 to 2, all the while maintaining the stability of the GdIII complex.14 The substantialincrease in r1 of the CuI probes Gd-10, Gd-11, and Gd-12 (∼3-fold) is 3−6 times 344

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351

Article

Accounts of Chemical Research

Figure 4. q-Based GdIII-responsive contrast agents for copper.

The second significant improvement was the addition of a pendant carboxylate group which increases the negative charge of the GdIII complex.15 In serum, the open coordination site of GdIII complexes can be readily filled by oxoanions, particularly bicarbonate.16 The affinity of endogenous hard anions for the hard GdIII is one of the Achilles’ heel of this class of probes. When the open coordination site is partially filled by an anion as opposed to water, the relaxivity turn-on is lowered. One way to mitigate this problem is to add negative charges to the complex

greater than other q-based responsive contrast agents (including those for ZnII) is primarily due to a structure that enables two water molecules to coordinate GdIII in the presence of copper. This strategy is only viable if inert macrocyclic GdIII complexes are used. Opening up further the coordination cage of the GdIII ion to water molecules would result in complexes that would be unstable in vivo. Note that no q = 2 contrast agents are approved clinically. 345

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351

Article

Accounts of Chemical Research

The first ZnII contrast agents that function via the formation of a dimer is Gd-24 (Figure 6).25 In the presence of 0.5 equiv of ZnII, Gd-24 forms a dimer whose slower rotational relaxation time results in a 55% increase in r1. Unfortunately, further addition of ZnII breaks the dimer to form two Gd-24·ZnII monomers with a concomitant decrease in r1. These agents do not have a straightforward “off”−“on” response. The similar analog Gd-25 answers likewise, although its binding to the protein albumin leads to a greater increase in r1.26 Similar results were observed with Gd-26,27 Gd-27, Gd-28 and Gd-29.28,29 Gd24, Gd-28 and Gd-29 respond both by relaxivity and by luminescence, enabling detection of ZnII by two complementary modalities. The second approach for τR-based responsive contrast agents is to design a complex that has low affinity for a protein in the absence of the targeted metal, but high affinity for the protein when it is complexed to the metal. For extracellular agents, one protein of choice is human serum albumin (HSA), a large protein present at 0.4 mM in the blood and which binds to a wide range of substrates. Since the protein−probe adducts are substantially bigger than probe dimers, the former have longer τR. As such, the response of protein-targeted probes, if designed with appropriate linkers that minimize rotational freedom, is typically higher. A further advantage of this approach is that it is not plagued by the “off”−“on”−“off” response observed with complexes that rely on the formation of dimers. The first example of such a probe is Gd-30 (Figure 7).30 In the absence of ZnII, Gd-30 has low affinity for HSA and tumbles rapidly in blood. The complex has high affinity for ZnII and high affinity for HSA only when bound to the zinc. This results in a near 2-fold increase in r1 upon ZnII binding in blood. A secondgeneration analogue, Gd-31, with tighter binding to HSA, has a 114% turn-on.31 One caveat with this approach is that the turnon is highly dependent on the magnetic field. At the higher magnetic field increasingly used in biomedical research, the effect of τR on r1 is mitigated; particularly if the water residence time, τm, is not optimum. At high field, the water residence time of GdDOTA is too slow to fully take advantage of the increase in τR upon protein binding. With this in mind, the turn-on of this class of probe can be increased by decreasing τm.32 This can be achieved by increasing the length of the polycarboxylate arms, as in Gd-32 to Gd-36. Two complexes in particular, Gd-34 and Gd35, showed increase turn-on relaxivity at low field (0.47 T) due to their optimized τm. Unfortunately, even with optimized τm, the benefit of τR is diminished above 3 T, and lower responses were observed at higher magnetic field. This is a disadvantage of τRbased agents compared to q-based ones whose responses are independent of magnetic field strength. Importantly, regardless of its class, as discussed above, the signal of a responsive contrast agents is dependent not only on the concentration of its substrate, but also on temperature, magnetic field strength, and media; hence the necessity to calibrate each probe before any in vivo experiments.

which decrease the affinity of anions for GdIII due to electrostatic repulsion. Indeed, the response of the negatively charged Gd-14 to CuI, is far less affected by bicarbonate levels than its neutral analog, Gd-10.15 Note that polyaminocarboxylate-based GdIII complexes are extracellular agents with poor cell penetration.17 Responsive agents are thus intended to image analytes that are also extracellular. To image changes in intracellular concentrations of substrate, the complex must be modified to enable cell penetration. This can be achieved by adding a cell penetrating peptide such as a polylysine to the probe, as in Gd-15.18 Unlike CuI, CuII prefers harder ligands such as carboxylates and pyridyls. The first responsive contrast agent for CuII, Gd-16, includes a carboxylate moiety that coordinate to GdIII in the absence of copper, resulting in q = 0 in the “off” state. r1 of Gd-16 increases by 35% as the carboxylate releases GdIII for CuII, thereby opening a coordination sphere on GdIII for water. The structure of Gd-16 is similar in structure to zinc-responsive agent Gd-3, highlighting the difficulty to design probes that can distinguish between Cu2+ and Zn2+.13,14 Similar q-based responsive contrast agents for CuII have been developed. Gd-18, employs a quinolone as a recognition element.19 Gd-19, uses instead a tryptophan,20 Gd-22, uses (2pyridylmethyl)amine most commonly used in ZnII sensing,21 and Gd-23 uses a dicarboxylate.22 The bimetallic Gd-21, also functions as a q-based contrast agent.23 The increase in signal intensity of these probes is less than 2-fold. Gd-19, GD-20, Gd22, and Gd-23 also display a change in fluorescence intensity upon binding CuII, highlighting the ability to render GdIIIresponsive contrast agents multimodal.24 The second class of GdIII responsive contrast agents functions by modulating the rotational correlation time, τR (Figure 5).



Figure 5. Design of τR-based responsive contrast agents.

CEST-BASED RESPONSIVE CONTRAST AGENTS Chemical Exchange Saturation Transfer (CEST) agents require an exchangeable proton belonging either to coordinated water molecule(s) or to NH or OH groups that resonate at a frequency different from that of bulk water.33 Because the protons are exchangeable, their direct saturation by an appropriate radiofrequency also transfers to bulk water. Therefore, the signal of the exchangeable proton disappears whereas that of the bulk water attenuates (Figure 8). The first advantage of CEST agents is that

There are two possibilities to do this; the first involves using the targeted metal ion to template the formation of dimers or trimers, the second is to use the metal ion to enhance binding to a larger macromolecule such as a protein. An advantage of this class of probe over the previous one is that since the response is not dependent on q, it is little affected by anion binding such as bicarbonate. 346

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351

Article

Accounts of Chemical Research

Figure 6. τR-Based responsive contrast agents for zinc that function via dimer formation.

Figure 7. τR-Based responsive contrast agents for zinc that function via protein binding.

unambiguously image the distribution of analyte when the concentration of the agent in a tissue is not uniform. Their disadvantage is that they have higher limit-of-detections than GdIII complexes (mM range). None have been approved for clinical use.

their MR signal can be switched on and off at will. The second is that the saturation pulse is frequency-encoded. Hence, probes with different exchangeable protons positioned at different chemical shifts can be used for multifrequency imaging. If these protons respond differently to a analyte, the corresponding CEST probe is ratiometric. This property is necessary to 347

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351

Article

Accounts of Chemical Research

The first CEST agent responsive to zinc, Eu-37 (Figure 10), functions by modulating the rate of exchange of the CEST proton.34 In the absence of ZnII, Eu-37 exhibits only one CEST signal, that of the slowly exchanging inner-sphere water molecules. Addition of ZnII leads to the formation of a complex of presumably similar structure as Gd-1 in which the ZnII is coordinated by the four pyridines. The macrocycle in such a structure adopts a different geometry which hastens the exchange rate of the water molecule to the point where the necessary condition Δω > kex is no longer maintained. As a result, the CEST signal decreases by ca. 50%. The two Tm-based paraCEST agents for copper and zinc, Tm38 and Tm-39, also function by modulating the rate of exchange of the labile proton, although in these cases the labile protons belong to amides and not to inner sphere water molecules.36 The addition of pendant carboxylates, which are expected to make the complex less toxic by decreasing its charge, also substantially decrease its sensitivity. In both cases, zinc, or copper coordination increases notably the exchange rate of the amide protons. Consequently, the CEST signal completely disappears. These are the only true “MRI signal-switch” responsive contrast agents: they are completely silent in the “off” state. Both probes form a 2:1 complex with CuII and ZnII and a 1:1 complex with CuI with affinity slightly higher than Eu-37. Interestingly, unlike the Gd-analogues, Gd-10 and Gd-14, the number of inner-sphere water molecule on the Ln ions does not change upon binding copper or zinc. The probe 40 is a diamagnetic CEST agent.35 It is less sensitive than the three paraCEST probes. Nonetheless, this probe can distinguish between different metal ions via frequency encoding, since each lead to a different Δω.

Figure 8. Principles of CEST contrast agents.

The CEST process is best explained by a system of two exchanging pools of protons which transfer magnetization saturation. The first one corresponds to bulk water, the second to the exchangeable protons of the CEST agent. For saturation transfer to occur, coalescence between the signals of those two proton pools must be avoided; thus the shift difference between the two proton pools must be higher than their exchange rate (Δω > kex). This highlights the advantage of CEST agents incorporating TmIII or EuIII, since the lanthanide induces a larger Δω, they enable the detection of protons with faster kex which in turn leads to higher saturation transfer and improved sensitivity. The sensitivity of a CEST agent is dependent on a number of parameters, of which three can be exploited to design responsive CEST agents (Figure 9): the number of exchangeable protons on the CEST agent, n, their exchange rate with the bulk water, kex, and their chemical shift compared to the bulk solvent, Δω.



IRON OXIDE NANOPARTICLE-BASED RESPONSIVE CONTRAST AGENTS Unlike GdIII-based contrast agents, magnetic iron oxide nanoparticles (MIONs) function entirely as outer-sphere relaxation agents. They are most commonly used as T2 agents but can also be used as T1 contrast agents depending on the size of the nanoparticle and the pulse sequence used. As T2 agents they darken the region where they accumulate in T2-weighted scans. As T1 agents they can brighten the image if fast cycling pulse sequences such as SWIFT that reduce the effect of T2 are used.37 The relaxivity of the nanoparticles is a function of their anisotropy energy.37,39 The anisotropy of the iron oxide core is dependent on (1) its bulk magnetocrystalline anisotropy field, which is a function of its chemical composition and crystal structure, (2) its demagnetizing field, which is a function of the shape of the crystal, (3) its surface anisotropy field, which is a function of the size of the core and its surface functionalization, in particular the anchoring group, and (4) its mutual anisotropy from dipolar coupling, which is a function of the clustering or agglomeration of the nanoparticles.37,38 The nature and thickness of the coating also affect both r1 and r2.39 A detailed description of the different parameters influencing relaxivity can be found elsewhere.40 Of these parameters, only one can be easily manipulated to yield responsive contrast agents: the clustering of the nanoparticles. Responsive MION contrast agents thus contain moieties on their peripheries that bind to or react with the analyte in such a way that they form clusters (Figure 11). Importantly, the level of magnetic interactions between iron oxide nanoparticles depends substantially on the distance separating them. This interaction is more substantial when

Figure 9. Design of responsive CEST contrast agents. 348

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351

Article

Accounts of Chemical Research

Figure 10. Responsive CEST contrast agents for copper and zinc.

alkynes (Figure 12).42 In the absence of CuI, the two sets of nanoparticles do not react and cluster, resulting in intermediate

Figure 11. Design of responsive MION contrast agents.

intercore distance is below 7 nm.41 The thickness of the coating is thus a crucial parameter influencing the response of MIONs. Thick silica coating, typically >20 nm thick, will not enable iron oxide core to come close enough to interact significantly, thus resulting in small responses. More porous and thinner polymer coatings are more appropriate for this application. Part of the difficulty in sensing ions with nanoparticles is that the ions must cause a change in clustering; simple coordination of a metal ion by terminal groups does not suffice. MIONs can nonetheless be rendered responsive via the use of two nanoparticles, each conjugated with complementary reactive groups that react only in the presence of the metal ion. The two responsive MION contrast agents for copper exploit the ability of CuI to catalyze Huisgen cycloaddition between alkyne and azides. The first one, MION-41, consists of two components, MION-41a is functionalized with PEGs terminated with azides, whereas those of MION-41b are terminated with

Figure 12. Design MION contrast agents for copper.

values of r1 and r2. Addition of CuI catalyzes the click chemistry between the azides and alkynes, resulting in the formation of clusters. The formation of clusters decreases the longitudinal relaxivity by half. This decrease is due to the formation of two pools of water.38 The smaller pool of water inside the cluster which relaxes rapidly due to its vicinity to multiple MIONs exchanges 349

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351

Article

Accounts of Chemical Research slowly with the bulk water outside the cluster which relaxes much slower. On the other hand, the formation of cluster results in an initial increase followed by a near complete decrease in the transverse relaxivity. This observation is in agreement with the theory of relaxivity. Initially, for small clusters, the cluster behaves as a large magnetized sphere whose total magnetic moment, and hence r2, increases with cluster size according to Langevin’s law. However, when the cluster becomes larger than 250 nm, r2 is governed by the static dephasing regime.38 At this point r2 decreases substantially with increasing cluster size. This construct can be rendered multimodal readily. Magnetoplasmonic agents behave both as MRI contrast agents and as plasmonic probes.43 The former functionality can be achieved with iron oxide nanoparticles, the latter with gold or silver nanoparticles or shells. Advantageously, both types of nanoparticles respond to analytes via the formation of clusters. MION-42 is a derivative of MION-41 that yields both a relaxometric and a plasmonic response by replacing one of the iron oxide nanoparticles with gold (MION-42b).44 As for the previous probe, addition of CuI leads to the formation of clusters and a decrease of r1 concomitant with an initial increase followed by a decrease of r2. Clustering also leads to a decrease in the intensity of the plasmonic band of the gold nanoparticles that can be visualized both by UV−visible spectroscopy and dark-field microscopy.

Sarah M. Harris received her B. A. in 2012 at Occidental College and her Ph.D. with Valerie Pierre at the UMN in 2017. Sylvie L. Pailloux received her B.Sc. from the Universities of Poitiers and Lille, France and her Ph.D. from the University of Pharmaceutical and Biological Science of Lille in 2003 under the supervision of Pascal Berthelot. She then joined the groups of Robert Paine at the University of New Mexico and Kenneth Raymond at the University of California Berkeley and Lumiphore Inc. before joining Valerie Pierre at University of Minnesota.



(1) Toth, E.; Helm, L.; Merbach, A. Relaxivity of Gadolinium(III) Complexes: Theory and Mechanism. In The Chemistry of Contrast Agents in Medical Magnetic Resonance Imaging; Merbach, A., Helm, L., Toth, E., Eds.; John Wiley & Sons, Ltd.: West Sussex, United Kingdom, 2013; pp 25−82. (2) Que, E. L.; Chang, C. J. Responsive magnetic resonance imaging contrast agents as chemical sensors for metals in biology and medicine. Chem. Soc. Rev. 2010, 39, 51−60. (3) Brucher, E.; Tircso, G.; Baranyai, Z.; Kovacs, Z.; Sherry, A. D. Stability and Toxicity of Contrast Agents. In The Chemistry of Contrast Agents in Medical Magnetic Resonance Imaging; Merbach, A., Helm, L., Toth, E., Eds.; John Wiley & Sons, Ltd.: West Sussex, United Kingdom, 2013; pp 157−208. (4) Hanaoka, K.; Kikuchi, K.; Urano, Y.; Nagano, T. Selective sensing of zinc ions with a novel magnetic resonance imaging contrast agent. Perkin Trans. 2001, 2, 1840−1843. (5) Huang, Z.; Zhang, X. A.; Bosch, M.; Smith, S. J.; Lippard, S. J. Tris(2-pyridylmethyl)amine (TPA) as a membrane-permeable chelator for interception of biological mobile zinc. Metallomics 2013, 5, 648−655. (6) Hanaoka, K.; Kikuchi, K.; Urano, Y.; Narazaki, M.; Yokawa, T.; Sakamoto, S.; Yamaguchi, K.; Nagano, T. Design and synthesis of a novel magnetic resonance imaging contrast agent for selective sensing of zinc ion. Chem. Biol. 2002, 9, 1027−1032. (7) Major, J. L.; Parigi, G.; Luchinat, C.; Meade, T. J. The synthesis and in vitro testing of a zinc-activated MRI contrast agent. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 13881−13886. (8) Major, J. L.; Boiteau, R. M.; Meade, T. J. Mechanisms of ZnIIactivated magnetic resonance imaging agents. Inorg. Chem. 2008, 47, 10788−10795. (9) Matosziuk, L. M.; Leibowitz, J. H.; Heffern, M. C.; MacRenaris, K. W.; Ratner, M. A.; Meade, T. J. Structural Optimization of Zn(II)Activated Magnetic Resonance Imaging Probes. Inorg. Chem. 2013, 52, 12250−12261. (10) Mishra, A.; Logothetis, N. K.; Parker, D. Critical In Vitro Evaluation of Responsive MRI Contrast Agents for Calcium and Zinc. Chem. - Eur. J. 2011, 17, 1529−1537. (11) Regueiro-Figueroa, M.; Gunduz, S.; Patinec, V.; Logothetis, N. K.; Esteban-Gomez, D.; Tripier, R.; Angelovski, G.; Platas-Iglesias, C. Gd3+Based Magnetic Resonance Imaging Contrast Agent Responsive to Zn2+. Inorg. Chem. 2015, 54, 10342−10350. (12) Zhang, X. A.; Lovejoy, K. S.; Jasanoff, A.; Lippard, S. J. Watersoluble porphyrins as dual-function imaging platform for MRI and fluorescence zinc sensing. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 10780−10785. (13) Que, E. L.; Chang, C. J. A Smart Magnetic Resonance Contrast Agent for Selective Copper Sensing. J. Am. Chem. Soc. 2006, 128, 15942−15943. (14) Que, E. L.; Gianolio, E.; Baker, S. L.; Wong, A. P.; Aime, S.; Chang, C. J. Copper-Responsive Magnetic Resonance Imaging Contrast Agents. J. Am. Chem. Soc. 2009, 131, 8527−8536. (15) Que, E. L.; Gianolio, E.; Baker, S. L.; Aime, S.; Chang, C. J. A copper-activated magnetic resonance imaging contrast agent with improved turn-on relaxivity response and anion compatibility. Dalton Trans. 2010, 39, 469−476. (16) Dickins, R. S.; Aime, S.; Batsanov, A. S.; Beeby, A.; Botta, M.; Bruce, J. I.; Howard, J. A. K.; Love, C. S.; Parker, D.; Peacock, R. D.;



CONCLUSIONS Responsive contrast agents for zinc and copper have been among the most successful. Although not ideal, they have enabled in vivo imaging of the extracellular distribution of zinc in the pancreas,29,31,32,45 brain, and cancer46,47 in small animals. The design approaches described above for zinc and copper can be applied to other substrates, including ions, biomolecules, enzymes, genes, and receptors. In the future, they are expected to yield novel diagnostics tools and enlighten our understanding of biological processes.



REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Phone: 612-625-0921. E-mail: [email protected]. ORCID

Valérie C. Pierre: 0000-0002-0907-8395 Sarah M. Harris: 0000-0002-6515-3548 Sylvie L. Pailloux: 0000-0001-7318-7089 Funding

This work was supported by the National Science Foundation Grant No. CAREER 1151665. Notes

The authors declare no competing financial interest. Biographies Valerie C. Pierre received her Engineering Degree from Lyon in 2001 and her Ph.D. with Kenneth Raymond in 2005. Following a postdoctoral scholarship with Jacqueline Barton at the California Institute of Technology, she joined the University of Minnesota in 2007. She is currently an Associate Professor of Chemistry and Medicinal Chemistry and is the recipient of a CAREER award from the NSF and the Edward Stiefel award in bioinorganic chemistry, and named an Emerging Investigator in Bioinorganic Chemistry by the American Chemical Society and a New Talent: Americas by the Royal Society of Chemistry. 350

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351

Article

Accounts of Chemical Research Puschmann, H. Structural, Luminescence, and NMR Studies of the Reversible Binding of Acetate, Lactate, Citrate, and Selected Amino Acids to Chiral Diaqua Ytterbium, Gadolinium, and Europium Complexes. J. Am. Chem. Soc. 2002, 124, 12697−12705. (17) Peterson, K. L.; Dang, J. V.; Weitz, E. A.; Lewandowski, C.; Pierre, V. C. Effect of Lanthanide Complex Structure on Cell Viability and Association. Inorg. Chem. 2014, 53, 6013−6021. (18) Que, E. L.; New, E. J.; Chang, C. J. A cell-permeable gadolinium contrast agent for magnetic resonance imaging of copper in a Menkes disease model. Chem. Sci. 2012, 3, 1829−1834. (19) Li, W. S.; Luo, J. A.; Chen, Z. N. A gadolinium(III) complex with 8-amidequinoline based ligand as copper(II) ion responsive contrast agent. Dalton Trans. 2011, 40, 484−488. (20) Kasala, D.; Lin, T. S.; Chen, C. Y.; Liu, G. C.; Kao, C. L.; Cheng, T. L.; Wang, Y. M. Gd(Try-TTDA)(H2O)(2‑): A new MRI contrast agent for copper ion sensing. Dalton Trans. 2011, 40, 5018−5025. (21) Zhang, X. L.; Jing, X.; Liu, T.; Han, G.; Li, H. Q.; Duan, C. Y. DualFunctional Gadolinium-Based Copper(II) Probe for Selective Magnetic Resonance Imaging and Fluorescence Sensing. Inorg. Chem. 2012, 51, 2325−2331. (22) Jang, J. H.; Bhuniya, S.; Kang, J.; Yeom, A.; Hong, K. S.; Kim, J. S. Cu2+-Responsive Bimodal (Optical/MRI) Contrast Agent for Cellular Imaging. Org. Lett. 2013, 15, 4702−4705. (23) Xiao, Y. M.; Zhao, G. Y.; Fang, X. X.; Zhao, Y. X.; Wang, G. H.; Yang, W.; Xu, J. W. A smart copper(II)-responsive binuclear gadolinium(III) complex-based magnetic resonance imaging contrast agent. RSC Adv. 2014, 4, 34421−34427. (24) Andrews, M.; Amoroso, A. J.; Harding, L. P.; Pope, S. J. A. Responsive, di-metallic lanthanide complexes of a piperazine-bridged bis-macrocyclic ligand: modulation of visible luminescence and proton relaxivity. Dalton Trans. 2010, 39, 3407−3411. (25) Luo, J.; Li, W. S.; Xu, P.; Zhang, L. Y.; Chen, Z. N. Zn2+ Responsive Bimodal Magnetic Resonance Imaging and Fluorescent Imaging Probe Based on a Gadolinium(III) Complex. Inorg. Chem. 2012, 51, 9508−9516. (26) Xu, Y. Q.; Luo, J.; Chen, Z. N. Zn2+-Responsive Bimodal Magnetic Resonance Imaging and Fluorescence Imaging Agents and Their Interaction with Human Serum Albumin. Eur. J. Inorg. Chem. 2014, 2014, 3208−3215. (27) Bonnet, C. S.; Caille, F.; Pallier, A.; Morfin, J. F.; Petoud, S.; Suzenet, F.; Toth, E. Mechanistic Studies of Gd3+-Based MRI Contrast Agents for Zn2+ Detection: Towards Rational Design. Chem. - Eur. J. 2014, 20, 10959−10969. (28) Rivas, C.; Stasiuk, G. J.; Sae-Heng, M.; Long, N. J. Towards understanding the design of dual-modal MR/fluorescent probes to sense zinc ions. Dalton Trans. 2015, 44, 4976−4985. (29) Stasiuk, G. J.; Minuzzi, F.; Sae-Heng, M.; Rivas, C.; Juretschke, H.P.; Piemonti, L.; Allegrini, P. R.; Laurent, D.; Duckworth, A. R.; Beeby, A.; Rutter, G. A.; Long, N. J. Dual-Modal Magnetic Resonance/ Fluorescent Zinc Probes for Pancreatic β-Cell Mass Imaging. Chem. Eur. J. 2015, 21, 5023−5033. (30) Esqueda, A. C.; López, J. A.; Andreu-de-Riquer, G.; AlvaradoMonzón, J. C.; Ratnakar, J.; Lubag, A. J. M.; Sherry, A. D.; De LeónRodríguez, L. M. A New Gadolinium-based MRI Zinc Sensor. J. Am. Chem. Soc. 2009, 131, 11387−11391. (31) De Leon-Rodriguez, L. M.; Lubag, A. J. M.; Lopez, J. A.; Andreude-Riquer, G.; Alvarado-Monzon, J. C.; Sherry, A. D. A second generation MRI contrast agent for imaging zinc ions in vivo. MedChemComm 2012, 3, 480−483. (32) Yu, J.; Martins, A. F.; Preihs, C.; Jordan, V. C.; Chirayil, S.; Zhao, P. Y.; Wu, Y. K.; Nasr, K.; Kiefer, G. E.; Sherry, A. D. Amplifying the Sensitivity of Zinc(II) Responsive MRI Contrast Agents by Altering Water Exchange Rates. J. Am. Chem. Soc. 2015, 137, 14173−14179. (33) van Zijl, P. C.; Yadav, N. N. Chemical exchange saturation transfer: what is in a name and what isn’t? Magn. Reson. Med. 2011, 65, 927−948. (34) Trokowski, R.; Ren, J.; Kálmán, F. K.; Sherry, A. D. Selective Sensing of Zinc Ions with a PARACEST Contrast Agent. Angew. Chem., Int. Ed. 2005, 44, 6920−6923.

(35) Bar-Shir, A.; Yadav, N. N.; Gilad, A. A.; van Zijl, P. C. M.; McMahon, M. T.; Bulte, J. W. M. Single 19F Probe for Simultaneous Detection of Multiple Metal Ions Using miCEST MRI. J. Am. Chem. Soc. 2015, 137, 78−81. (36) Srivastava, K.; Ferrauto, G.; Harris, S. M.; Botta, M.; Aime, S.; Pierre, V. C. Complete On/Off Responsive ParaCEST Contrast Agents for Copper and Zinc. Submitted. (37) Smolensky, E. D.; Park, H. Y. E.; Zhou, Y.; Rolla, G. A.; Marjanska, M.; Botta, M.; Pierre, V. C. Scaling laws at the nanosize: the effect of particle size and shape on the magnetism and relaxivity of iron oxide nanoparticle contrast agents. J. Mater. Chem. B 2013, 1, 2818−2828. (38) Roch, A.; Gossuin, Y.; Muller, R. N.; Gillis, P. Superparamagnetic colloid suspensions: Water magnetic relaxation and clustering. J. Magn. Magn. Mater. 2005, 293, 532−539. (39) Smolensky, E. D.; Park, H. Y. E.; Berquo, T. S.; Pierre, V. C. Surface functionalization of magnetic iron oxide nanoparticles for MRI applications - effect of anchoring group and ligand exchange protocol. Contrast Media Mol. Imaging 2011, 6, 189−199. (40) Roch, A.; Muller, R. N.; Gillis, P. Theory of proton relaxation induced by superparamagnetic particles. J. Chem. Phys. 1999, 110, 5403−5411. (41) Frankamp, B. L.; Boal, A. K.; Tuominen, M. T.; Rotello, V. M. Direct Control of the Magnetic Interaction between Iron Oxide Nanoparticles through Dendrimer-Mediated Self-Assembly. J. Am. Chem. Soc. 2005, 127, 9731−9735. (42) Smolensky, E. D.; Marjanska, M.; Pierre, V. C. A responsive particulate MRI contrast agent for copper(I): a cautionary tale. Dalton Trans. 2012, 41, 8039−8046. (43) Smolensky, E. D.; Neary, M. C.; Zhou, Y.; Berquo, T. S.; Pierre, V. C. Fe3O4@organic@Au: core-shell nanocomposites with high saturation magnetisation as magnetoplasmonic MRI contrast agents. Chem. Commun. 2011, 47, 2149−2151. (44) Weitz, E. A.; Lewandowski, C.; Smolensky, E. D.; Marjanska, M.; Pierre, V. C. A Magnetoplasmonic Imaging Agent for Copper(I) with Dual Response by MRI and Dark Field Microscopy. ACS Nano 2013, 7, 5842−5849. (45) Lubag, A. J. M.; De Leon-Rodriguez, L. M.; Burgess, S. C.; Sherry, A. D. Noninvasive MRI of beta-cell function using a Zn2+-responsive contrast agent. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 18400−18405. (46) Clavijo Jordan, M. V.; Lo, S. T.; Chen, S. W.; Preihs, C.; Chirayil, S.; Zhang, S. R.; Kapur, P.; Li, W. H.; De Leon-Rodriguez, L. M.; Lubag, A. J. M.; Rofsky, N. M.; Sherry, A. D. Zinc-sensitive MRI contrast agent detects differential release of Zn(II) ions from the healthy vs. malignant mouse prostate. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, E5464−E5471. (47) Ghosh, S. K.; Kim, P.; Zhang, X. A.; Yun, S. H.; Moore, A.; Lippard, S. J.; Medarova, Z. A novel imaging approach for early detection of prostate cancer based on endogenous zinc sensing. Cancer Res. 2010, 70, 6119−6127.

351

DOI: 10.1021/acs.accounts.7b00301 Acc. Chem. Res. 2018, 51, 342−351