Synergy of Key Properties Promotes Dendrimer ... - ACS Publications

Subscriber access provided by University of South Dakota

Article

A Synergy of Key Properties Promotes Dendrimer Conjugates as Prospective Ratiometric Bioresponsive MRI Probes Liam Connah, and Goran Angelovski Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/acs.biomac.8b01425 • Publication Date (Web): 09 Nov 2018 Downloaded from http://pubs.acs.org on November 10, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

A Synergy of Key Properties Promotes Dendrimer Conjugates as Prospective Ratiometric Bioresponsive MRI Probes

Liam Connah and Goran Angelovski*

MR Neuroimaging Agents, Max Planck Institute for Biological Cybernetics, D-72076 Tuebingen, Germany

ACS Paragon Plus Environment

1

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 28

ABSTRACT

Bioresponsive or smart contrast agents (SCAs) sensitive to Ca2+ are of extreme interest in the development of functional magnetic resonance imaging (MRI) techniques as they can aid in tracking neural activity in vivo. To this end, the design of macromolecular systems based on nanoscaffolds such as dendrimers functionalised with multiple MRI contrast agents have been used to conveniently increase the local concentration of paramagnetic MR reporters and slow the diffusion time of the probe, which are favourable in vivo characteristics. Moreover, previous studies with Ca-sensitive dendrimeric MRI probes revealed favourable properties crucial in the development of a ratiometric T2/T1-imaging method that provided a higher contrast-to-noise ratio compared to conventional T1- or T2-weighted imaging protocols. We therefore developed a series of novel dendrimeric MRI probes (DCAs) with differing structural properties and charge distributions. We thoroughly studied their features such as the relaxometric behaviour, size change and examined their electrostatic behaviours prior to and after the addition of Ca2+. The most active DCA displayed a common increase in r1 (3.11 mM-1 s-1 to 5.72 mM-1 s-1) and a remarkable increase in r2 (7.44 mM-1 s-1 to 34.57 mM-1 s-1), resulting in a r2/r1 ratio increase of the factor 2.52, which is greater than what was previously achieved. These changes in r1 and r2 were followed with a hydrodynamic diameter increase from 7.1 ± 1.2 to 8.5 ± 0.7 nm upon the addition of Ca2+, along with a decrease in the negative surface charge of the nanoparticle. Overall our findings indicate that highly responsive DCAs can be developed only through a combination of properties such as a change in hydration and size of the molecule, which come as a consequence of intramolecular structural and electrostatic changes in the particle. In turn, they provide a model for future preparations of responsive DCAs that can be utilized for both T1-


2

Page 3 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

weighted and ratiometric T2/T1-weighted imaging to visualize essential biological processes in a dynamic fashion.

KEYWORDS Calcium, Contrast agents, Dendrimers, Magnetic resonance imaging


3


Page 4 of 28

INTRODUCTION Functional imaging is an extremely attractive endeavour within magnetic resonance imaging (MRI) that aims to dynamically visualise biological processes in an unprecedented fashion. It is a spin-off of MRI, an important imaging technique which is routinely used for diagnostic purposes to acquire three-dimensional soft tissue images with excellent spatial resolution and without the negative effects of ionising radiation.1-3 Yet, as contrast agents (CAs) shorten the T1 and T2 relaxation times of the water protons in the vicinity of the agent thereby improving the specificity of the MRI experiment,4 bioresponsive or smart CAs (SCAs) enable the extraction of essential information such as organ or tissue function during functional MRI (fMRI) investigations. SCAs differentiate from conventional MRI CAs as they alter their behaviour upon a specific change in their microenvironment and therefore help to distinguish between different environmental states. To date various examples of SCAs have been developed and utilized, for instance those which examine changes in metal ion concentrations, enzyme activity, or pH changes.3, 5-7 Of the metal ion targets, calcium has received a lot of attention due to its critical involvement in cell signalling processes such as the secretion of neurotransmitters and hormones.8 Attempts to design SCAs responsive to calcium as ‘off-on’ sensors have been pursued over recent years with the development of several ‘small-molecule’ responsive probes.9 The coupling of these SCAs to different nanoscaffolds has resulted in multimeric systems characterised by a higher Gd3+ payload and a slower diffusion rate, which are ideal characteristics for in vivo application.9 Of the investigated nanoparticles, dendrimers were extensively studied in the pursuit of developing nano-sized calcium-responsive SCAs. They have a unique tree-like branched structure providing a precise core scaffold with a defined number of external functional groups depending on the


4

Page 5 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

generation (G) of the dendrimer (G1 = 4 functional groups, G2 = 8, G3 = 16 etc.).10-11 The overall size of the dendrimer can then be tuned through modifications of the surface and the environmental conditions such as solvent or pH.12-13 Previous investigations with a monomeric Ca-responsive SCA coupled to a polyamidoamine (PAMAM) dendrimer revealed a significant change in size of the dendrimeric SCA (DSCA) upon the addition of Ca2+ and a significant change in T2 compared to T1.14 These findings crucially led to the development of a ratiometric imaging method that exploits the changes in the T2/T1 ratio of the probe as a function of [Ca2+], allowing for rapid image acquisition with a high contrast-to-noise ratio. Moreover, they provided a significant perspective for the development of a dynamic fMRI methodology that allows functional studies with exceptionally high temporal resolution. Being intrigued with the behaviour of the previously developed DSCA, we embarked on a detailed investigation into the interactions of this class of nano-systems with the desired target (Ca2+) through studying major properties that affect the recorded T1 and T2 times, and subsequent assessment of their potential for performing dynamic fMRI studies. Namely, although the change in diameter was surprising, we anticipated that a possible explanation for this particular behaviour could be attributed to the previously mentioned size dependency of dendrimers on their environmental conditions and specifically in this case the fluctuating electrostatic interactions within the molecule upon Ca2+ coordination.12-13 Therefore we developed a novel series of dendrimeric CAs (DCAs) which differ through structural modifications of the macrocyclic MR reporter motif while all maintaining the same ethylene glycol tetraacetic acid (EGTA)-derived Ca2+ chelator. Varying the design of macrocyclic MR reporters allowed us to investigate which molecular characteristics (charge, type of macrocyclic unit or a linker distance) change upon the DCA interaction with Ca2+, thus


5


Page 6 of 28

potentially maximising the T2/T1 ratio changes. The responses of the DCAs to Ca2+ additions were investigated by proton longitudinal and transverse relaxometric titrations, while the overall behaviour of the nanosystem in terms of size and surface charge were assessed with dynamic light scattering (DLS) and zeta potential studies respectively.

EXPERIMENTAL SECTION General remarks. Commercially available reagents and solvents were used without further purification. Purification of synthesized compounds was performed using silica gel 60 (0.03-0.2 mm) from Carl Roth (Germany). G4 Starburst‚ PAMAM dendrimer with cystamine core was purchased from Andrews ChemServices, USA. For the nomenclature of monomeric and dendrimeric intermediate products, we used the following abbreviations: L1-4 are reactive monomeric macrocycles functionalized with the isothiocyanate group and with carboxylate groups protected as t-butyl esters; D1a-5a are coupling products of commercial G4 PAMAM dendrimer with L1-4 (D5a is the product of D4a capping with the methyl isothiocyanate), which still possess carboxylate groups protected as t-butyl esters; D1b-5b are products of acid hydrolysis of t-butyl esters of D1a-5a; finally, DCAs1-5 are products of D1b-5b complexation with Gd3+. Dendrimers D1a-5a were purified using lipophilic Sephadex® LH-20 (bead size: 25-100 μm) from Sigma-Aldrich (Germany). DCAs1-5 were purified using hydrophilic Sephadex® G-15 (bead size: 40-120 μm) from GE Healthcare. Low resolution mass spectra were recorded on an ion trap SL 1100 system Agilent with an electrospray ionization source. High resolution mass spectra were recorded on a Bruker Daltonics APEX II (FT-ICR-MS) with an electrospray ionization source. MALDI-TOF-MS analysis was performed by The Scripps Center for Mass Spectrometry, La Jolla, CA. 1H and

13C

NMR spectra and relaxometric experiments were performed on a


6

Page 7 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Bruker Avance III 300 MHz spectrometer at 25 °C. Processing was performed using TopSpin 2.1 (Bruker GmbH) and ACD/SpecManager 9.0 (Advanced Chemistry Development, Inc.). The NMR spectra were obtained either in CDCl3 or D2O, using the deuterium lock frequency. The concentration of Gd3+ in analysed solutions was determined using the bulk magnetic susceptibility shift (BMS).15 DLS and zeta potential measurements were done on a MalvernNano-ZS (Zetasizer, software ver. 6.2) instrument. DCAs1-5 preparation. DCAs1-5 were each prepared following the same general procedure.16 G4 PAMAM dendrimer (1 equiv. with 64 amino surface groups, 10 % dendrimer solution in methanol) was dried under reduced pressure. The residue was dissolved in dimethylformamide, triethylamine (160 equiv.) was added and the mixture was stirred at 45 °C for 45 min. Macrocyclic isothiocyanate (L1-4, 1.5 equiv. per amino surface group of the dendrimer) was dissolved in dimethylformamide and added in 4 portions (one portion every 12 h) while maintaining the temperature of the mixture at 45 °C. After the final addition of isothiocyanate, the mixture was stirred for an additional 12 h at 45 °C. The solvent was then removed by bulbto-bulb vacuum distillation at 50 °C. The resulting residue was redissolved in methanol and purified by size exclusion chromatography using lipophilic Sephadex® with methanol as the eluent. The solvent was then evaporated under reduced pressure yielding the tert-butyl protected dendrimer D1a-4a. For the preparation of DCA5, the dendrimer D4a underwent an additional capping reaction with methyl isothiocyanate (62 equiv.) and triethylamine (103 equiv.) in dimethylformamide and was purified using the procedure described for the macrocyclic isothiocyanates L1-4, to result in the dendrimer D5a. The tert-butyl protected dendrimers D1a-5a were then dissolved in formic acid and the mixture stirred at 60 °C for 24 h. The solvent was removed under reduced pressure. The residue was


7


Page 8 of 28

redissolved in water and lyophilised to give the dendrimers D1b-5b. The resulting solids were then dissolved in water and the pH was adjusted to 7 by the addition of sodium hydroxide (0.1 M). A solution of gadolinium chloride hexahydrate (1.1 equiv. relative to the number of amino surface groups of the dendrimer) in water was then added to the mixture portionwise over 1 h while maintaining the pH at 7 with the addition of a solution of sodium hydroxide (0.05 M). The resulting solution was stirred for 24 h at room temperature. Ethylenediaminetetraacetic acid (EDTA, 1.5 equiv. relative to the number of amino surface groups of the dendrimer) was added to the solution over the period of 1 h while maintaining the pH at 7. The mixture was stirred for a further 2 h at room temperature and the solvent was evaporated. This was redissolved in a small volume of water and purified by size-exclusion chromatography with hydrophilic Sephadex® using water as the eluent. The fractions were collected and centrifuged using a 3 kDa centrifugal filter for 40 min at 2000 x g. This process was repeated until the filtrate showed an absence of GdEDTA and EDTA (checked by low resolution ESI-MS). The product was then lyophilised to obtain the final DCAs1-5 as off-white/yellow solids. Dendrimer characterisation. Dendrimeric conjugate products were characterised by 1H NMR and/or MALDI-TOF. Estimations of the number of monomeric units attached to the surface of the dendrimer unit were calculated using 1H NMR on dendrimeric conjugates before (D1a-4a) or after tBu ester hydrolysis (D1b-4b).16 MALDI-TOF was carried out on the dendrimers D1b-4b and the final DCA1-5. The MALDI-TOF spectra exhibited broad signals indicating the presence of species with high molecular masses, while preventing detailed quantitative analysis due to presence of multiple species with different m/z values. Relaxometric titrations. Proton longitudinal and transverse relaxometric titrations with Ca2+ were performed at 7.0 T, 25 °C and pH 7.4 (50 mM HEPES buffer) using inversion recovery (T1)


8

Page 9 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

and Car-Purcell-Meiboom-Gill (T2) pulse sequences. A CaCl2 solution of known concentration was added stepwise to the DCA1-5 solution (starting concentration 2.5 mM Gd3+) and measurements of T1 and T2 were performed after each addition of the analyte. The longitudinal and transverse relaxivities, r1 and r2, were calculated from Eq. 1, where Ti,obs is the measured T1 (i = 1) or T2 (i = 2), Tid is the diamagnetic contribution of the solvent and [Gd] is the actual Gd3+ concentration at each point of the titration. All reported relaxivity values were the average of three independent measurements. 1/Ti,obs= Ti,d +ri × [Gd]

(1)

DLS and zeta potential measurements. DLS and zeta potential measurements were carried out with DCAs1-5 (0.3 mM Gd3+, 25 mM HEPES, pH 7.4) with and without the addition of 2 equivalents of Ca2+. Sample solutions were filtered before measuring. After the addition of Ca2+, samples were left to equilibrate for 5 min before measurement. Each DLS measurement included 5 repetitions of 15 scans (1 scan = 12 sec, refractive index 1.345, absorption 1 %), without delays in between the scans, and with an equilibration time of 30 sec prior to recording. The reported size was obtained from the values provided by a distribution analysis. Each zeta potential measurement included 5 repetitions (number of scans was determined by the device depending on the consistency of the data) after an initial calibration time (10 min).

RESULTS AND DISCUSSION Design of nano-sized DCAs. The design of the new DCA series was based upon a previously reported nano-sized SCA from our group,14,

17

particularly a 1,4,7,10-tetraazacyclododecane-

1,4,7-tricarboxylic acid (DO3A)-based MR reporter and an EGTA-derived Ca2+ chelator were conjugated to the nanoparticle of interest, a G4 PAMAM dendrimer.14 In this study, we


9


Page 10 of 28

expanded the number of different DCAs that could potentially provide diverse responses to Ca2+, which resulted in a series of five newly developed DCAs. These DCAs can be broadly split into two classes based on the choice of MR reporter used. The first group is analogous to the previously investigated responsive probe with a propyl linker between the DO3A and EGTAderived units. Here expanding the distance between these moieties resulted in two DCAs with butyl and pentyl linkers (DCA1 and DCA2, respectively). The second group of DCAs consisted of MR reporters that form 8 coordination bonds with the paramagnetic Gd3+, namely possessing DO3A-monoamide (DCA3) or 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetate (DOTA)type (DCA4-5) units. Among these, the nano-sized probe DCA5 is a close analogue of DCA4, with the exception that the remaining unfunctionalised dendrimeric amino surface groups were additionally capped with methyl-thiourea units to result in the system DCA5 (Figure 1).

R

COO-

H N

N

OOC

R = -

OOC

N N Gd3+ N N

O

OOC

COO-

N N Gd3+ N N

-

OOC

S

O

O

-

COO-

H N

N

O

-

COO-

N H

-

OOC

N N Gd3+ N N

-

OOC

DCA

DCA

COO-

-

OOC

O N H

-

OOC

N N Gd3+ N N

COO-

COO-

DCA4

DCA3

2

1

N H

NH S -

OOC

-

OOC

N N Gd3+ N N

NH

COOCOO-

H N COO-

COO-

N O

H N

N O

O

O

S N H

N H

DCA5

Figure 1. Chemical structures of the synthesized DO3A-, DO3A-monoamide and DOTA-based DCAs.


10

Page 11 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

The rationale for using different DO3A and DOTA-type macrocycles or capping of the amino groups on the dendrimer surface was to achieve different charge distributions across the investigated DCAs. Namely, dendrimers DCA1-2 possess Gd-DO3A MR reporting units with zero net charge, while also containing EGTA-derived units that likely bear a net negative charge,18 and a positively charged dendrimeric core. Similar to them, the net charge of the macrocyclic MR reporting units of DCA3 is also zero; however this unit is not known to have an effective communication with the EGTA-derived part upon interaction with Ca2+ to result in hydration number change.3 Finally, the systems DCA4 and DCA5 with DOTA-type macrocyclic MR reporters are also not expected to change their hydration number upon interaction with Ca2+, nevertheless their distribution of charges is different than in the cases of DCA1-3. Both DCA4-5 possess negatively charged MR reporting units, while the positive surface charge of the dendrimer is suppressed in DCA5 through the capping of the remaining amino groups and their conversion to the thioureas. Having the above mentioned differences amongst these five systems, we were particularly interested in studying their behaviour upon the addition of Ca2+. It was anticipated that the charge of the EGTA-derived component would change from negative to neutral through its coordination to Ca2+, thus altering the overall charge distribution across the molecule (Figure 2).


11


Page 12 of 28

Figure 2. The anticipated Ca2+ response mechanism for DCAs1-5 with the associated changes in electrostatic charge across the system. Synthesis. The synthesis of the DO3A-based monomeric SCAs (L1,2, Scheme 1) commenced from tBu-DO3A. Alkylation of the free secondary amine with alkyl bromides 1 or 2 (prepared from 4-aminobutan-1-ol and 5-aminopentan-1-ol respectively), gave 3 and 4.19-21 Removal of the Cbz protecting group of 3 or 4 was achieved by catalytic hydrogenation with 10 wt% Pd(OH)2/C in ethanol yielding amines 5 and 6. Carbodiimide activated coupling with N,N’dicyclohexycarbodiimide (DCC) and bromoacetic acid in dichloromethane afforded macrocyclic bromides 7 and 8. EGTA derivative 11 was prepared via monoalkylation of 9 with 10 in acetonitrile with potassium carbonate as the base. Alkylation of 11 with 7 or 8 formed nitro compounds 12 and 13 respectively. Reduction of the nitro group was then carried out by hydrogenation with Pd/C in ethanol giving amines 14 and 15, which were converted further to isothiocyanates using thiophosgene in dichloromethane to yield the final monomeric SCAs L1 and L2.


12

Page 13 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

tBuOOC i)

tBu-DO3A Br

NHCbz

( )n

tBuOOC

1, n=1 2, n=2

tBuOOC

N H

N

N

N

N

( )n

tBuOOC

NHCbz

N

ii)

COOtBu

N

tBuOOC

3, n=1 4, n=2

O

O

COOtBu

Br

+

N

tBuOOC

NH2

COOtBu

11

N

tBuOOC

N H

v)

N

N

N

N

( )n

tBuOOC

O

N H

COOtBu

H N

O

tBuOOC

N

H N

Br

COOtBu

N O

O

NO2

O N

N H

11

COOtBu

N O

( )n

7, n=1 8, n=2

tBuOOC iv)

N

O

NO2

tBuOOC +

N

iii)

10

9

7 or 8

( )n

5, n=1 6, n=2

O

H N

N

H N

O

R

COOtBu

12, n=1, R=NO2 14, n=1, R=NH2 L1, n=1, R=NCS

vi) vii)

13, n=2, R=NO2 15, n=2, R=NH2 L2, n=2, R=NCS

vi) vii)

Scheme 1. Synthesis of DO3A-based SCA monomers (L1, L2). Reagents and conditions: (i) 1 or 2, K2CO3, CH3CN, 70 °C. 3 = 94% , 4 = 84%. (ii) H2, Pd(OH)2/C, NH3/MeOH, EtOH, RT. 5 = 97%, 6 = 99%. (iii) BrCH2COOH, DCC, DCM, RT. 7 = 53%, 8 = 69%. (iv) 9, 10, K2CO3, CH3CN, RT, 34%. (v) 7 or 8, 11, K2CO3, CH3CN, 70 °C. 12 = 58%, 13 = 44%. (vi) H2, Pd/C, EtOH, RT. 14 = 99%, 15 = 93%. (vii) CSCl2, NaHCO3 (sat), DCM, RT. L1 = 41%, L2 = 42%.

The DOTA-based monomeric CAs L3,4 were synthesized following procedures similar to those described for L1,2 (Scheme 2). Aliphatic bromide 21 was prepared following a previously published procedure, in which H-Lys(Z)-OH underwent diazotization with sodium nitrite before bromination using potassium bromide and finally conversion of the carboxylic acid to a tertbutyl ester through treatment with 2,2,2-trichloroacetimidate.22 Linkers 10 and 21 were then used to alkylate the fourth N-position of DO3A to form the DO3A-monoamide and DOTA-like scaffolds, giving compounds 16 and 22 respectively. Hydrogenation with Pd(OH)2/C or Pd/C respectively afforded amines 17 and 23. Then, by following the steps previously outlined in Scheme 1, monomers L3 and L4 were synthesized.


13


a)

tBuOOC i)

tBu-DO3A

N

COOtBu

N

tBuOOC

N

N

tBuOOC

tBuOOC

+

18

N

N

COOtBu

tBuOOC i)

N

tBu-DO3A Br

NHR

tBuOOC

N

N

N

N

O

N H

tBuOOC 11

+

24

iv) tBuOOC

H N

N O

COOtBu

O

O

R

v) vi)

tBuOOC iii) NHR

22, R=Cbz 23, R=H

21

Br O

N H

COOtBu

N

tBuOOC

COOtBu

COOtBu

N

COOtBu

H N

19, R=NO2 20, R=NH2 L3, R=NCS

b)

H N

18

O tBuOOC

COOtBu

N

N

ii)

N

iv)

N

O

N H

16, R=NO2 17, R=NH2

11

tBuOOC iii)

R

O

10

Page 14 of 28

N

N

N

N

COOtBu H N

Br O

COOtBu

24

ii)

N

N

N

N

COOtBu COOtBu

H N

N O

COOtBu

COOtBu

25, R=NO2 26, R=NH2 L4, R=NCS

N O

O

O

H N R

v) vi)

Scheme 2. a). Synthesis of the DO3A-monoamide based SCA monomer (L3). Reagents and conditions: (i) 10, K2CO3, DMF, RT, 67%. (ii) H2, Pd/C, EtOH, RT, 90%. (iii) BrCH2COOH, DCC, DCM, RT, 51%. (iv) 11, 18, K2CO3, CH3CN, 70 °C, 59%. (v) H2, Pd/C, EtOH, RT, 76%. (vi) CSCl2, NaHCO3 (sat), DCM, RT, 55%. b). Synthesis of the DOTA-based SCA monomer (L4). Reagents and conditions: (i) 21, K2CO3, DMF, 45 °C, 79%. (ii) H2, Pd(OH)2/C, EtOH, RT, 99%. (iii) BrCH2COOH, DCC, DCM, RT, 51%. (iv) 11, 24, K2CO3, CH3CN, 70 °C, 76%. (v) H2, Pd/C, EtOH, RT, 100%. (vi) CSCl2, NaHCO3 (sat), DCM, RT, 42%.


14

Page 15 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

DCAs1-4 were synthesized by conjugating monomers L1-4 to the surface amines of G4 PAMAM dendrimers through step wise additions of the monomer in dimethylformamide to give the protected dendrimeric compounds D1a-4a (Scheme 3). An estimation of the number of attached monomeric units was performed on purified D1a-4a, and reconfirmed also after the preparation of D1b-4b (see below). Out of 64 surface amines available for functionalisation on a G4 dendrimer, the estimations indicated average monomeric loadings of between 52 – 77 % for all dendrimeric systems, which are consistent with loadings previously achieved.14,

16

Deprotection of the tert-butyl esters was carried out using formic acid to provide D1b-4b and complexation with Gd3+ was achieved at pH 7 to give the final dendrimeric contrast agents DCA1-4. Finally, DCA5 was prepared through an additional capping step on D4a with methyl isothiocyanate to give D5a. Deprotection with formic acid yielded D5b, which underwent complexation with Gd3+ to obtain DCA5.


15


+

SCN

Page 16 of 28

R

R

O

O

N

R

N

N

R

N

N

N

R

L1-4

i) G4 PAMAM dendrimer with the cystamine core

R N

R

N

R

R

D4a R

N

N

R

R

O

O

N

N

R D1a-4a, R = tBu

N N

N

R N

N

N

ii)

D1b-4b, R = H

R

O

O

n

n

iv)

-

O

iii)

O-

N

N

N

N

-

-

O

O

O-

O-

O

O

N

O

N

N

O-

N

N

N

O-

O

N N

O-

DCA1-4

O n

n

R

N

N

R

R R

R

N

N

R

N

N N

O

O

R

N

N

R

N

N

N

R

D5a, R = tBu

R

D5b, R = H

ii)

O n

O

n

iv)

OO-

N

N

N

N

O-

OO-

O-

O

O

N

N N

O

O-

N

N

O-

N

N

N

O-

O-

DCA5

O n

n

Scheme 3. Synthesis of DCA1-5. Reagents and conditions: (i) L1-4, Et3N, DMF, 45 °C. (ii) HCO2H, 60 °C. (iii) MeNCS, Et3N, DMF, 45 °C. (iv) GdCl3.6H2O, H2O, pH 7, RT, followed by EDTA.2Na.2H2O, H2O, pH 7, RT.


16

Page 17 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Relaxometric characterisation. Proton longitudinal and transverse relaxometric titrations were carried out for DCAs1-5. All measurements were conducted at 7 T and 25 °C with identical starting Gd3+ concentrations (2.5 mM). T1 and T2 relaxation times were determined as a function of [Ca2+] and the r1 and r2 relaxivities were calculated after every addition of Ca2+. Relaxometric titration curves revealed two distinct behaviours depending on the type of DCA examined (Figures 3 and 4). The first class of DCAs, namely DCA1 and DCA2, exhibited remarkable increases in relaxivity as a function of [Ca2+], whereas the second class (DCA3-5) showed profiles with markedly lower changes in relaxivity (Figure 4).

Figure 3. Relaxometric titration curves for DCA1 and DCA2. [Gd3+] = 2.5 mM, pH 7.4 (50 mM HEPES), 25 °C, 7 T.


17


Page 18 of 28

The relaxometric titrations show that the butyl based derivative DCA1 displays a greater increase in both r1 and r2 (84 % and 365 % vs. 76 % and 280 %, respectively) upon Ca2+ addition when compared to the pentyl derivative DCA2 (Figure 3). Specifically, the r1 values increased from 3.11 mM-1s-1 to 5.72 mM-1s-1, or from 3.30 mM-1s-1 to 5.80 mM-1s-1 for DCA1 and DCA2, respectively. At the same time, the r2 values increased from 7.44 mM-1s-1 to 34.57 mM-1s-1, and from 7.39 mM-1s-1 to 28.05 mM-1s-1 for DCA1 and DCA2, respectively. When the r2/r1 ratios were calculated for both DCAs and normalized to the initial values, we obtained the factors 2.52 and 2.14 for DCA1 and DCA2, respectively (Figure 5 and Table 1). Compared to both DCA2 and to the previously reported ‘smart’ dendrimeric contrast agent,17 DCA1 exhibited a remarkable increase in r2 and especially in the r2/r1 ratio, making this derivative a candidate for further possible use in vivo. Obviously, by changing the linker from propyl (previously reported dendrimeric probe) to butyl and thereby extending the length and distance between the EGTAderived and MR reporting units, a slight increase in the overall effect (r2/r1 ratio) was achieved. However, by increasing the distance between the carboxylic flipping group of EGTA and the macrocycle even further (DCA2 bearing the pentyl linker), it was observed that changes in r1 and r2 are less pronounced compared to their two structural analogues. As an alternate approach to maximising the r2/r1 ratios, we investigated the second class of DCAs, the three dendrimeric agents DCA3-5. This approach aimed to restrict a change in the hydration number of the complex upon the addition of Ca2+. In both the DO3A-monoamide and the DOTA-type systems (DCA3 and DCA4-5, respectively), the fourth pendant group of the macrocycle remains coordinated to Gd3+ instead of the carboxylate from the EGTA-derived moiety. Therefore upon Ca2+ binding to the Ca-chelator, no change in the hydration number takes place since the Gd3+ metal ion is already saturated. Consequently, the relaxometric


18

Page 19 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

titrations with the DCA3-5 revealed remarkably different behaviours to that of the DO3A systems. In all cases the r1 remained constantly between 5.50 and 6.50 mM-1s-1, the values which are typical for DOTA-type dendrimeric contrast agents.23 On the other hand, the overall r2 increase did not exceed 25 %, with values ranging from 17.50 to 22.50 mM-1s-1 (Figure 4). Interestingly, while the control of r1 was achieved, the enhancement in r2 as a function of [Ca2+] was considerably less than desired, so supressing the change in r1 could not boost the r2/r1 ratio due to moderate r2 changes (Figure 4, Table 1). In turn the results obtained with the DCAs3-5 proved that they are unsuitable for the desired application, while additionally indicating that a change in hydration is of paramount importance to achieve the astonishing increase in r2 observed within the DO3A-type systems.

Figure 4. Relaxometric titration curves for DCAs3-5. [Gd3+] = 2.5 mM, pH 7.4 (50 mM HEPES), 25 °C, 7 T.


19


Page 20 of 28

Figure 5. Relaxometric titration curves showing the r2/r1 ratio change of DCAs1-2 (top) and DCAs3-5 (bottom).

Table 1. r2/r1 ratios of DCAs1-5 at Ca2+ saturation (< 2 equiv.). DCA

1

2

3

4

5

r2/r1

2.52

2.14

1.32

1.26

1.25

Dynamic Light Scattering. Given the incredible differences in relaxometric behaviour between the two classes of investigated systems, further insights into the overall structural behaviour of these DCAs upon Ca2+ binding were investigated. Previously, a DO3A-type DCA


20

Page 21 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

analogue with a propyl linker was reported to undergo a substantial increase in size upon transition from the off- to on-state through the addition of Ca2+, which was unexpected for a G4 PAMAM dendrimer conjugate.14,

24

Indeed, similar behaviour for the DO3A-systems

investigated in this study was observed.

Table 2. Particle hydrodynamic diameter determined using the DLS method before and after the addition of Ca2+ (2 equiv.). Conditions: 0.3 mM [Gd3+], pH 7.4 (25 mM HEPES). DCA

1

2

3

4 a)

5 a)

No Ca2+

7.1 ± 1.2

5.6 ± 0.9

7.4 ± 0.4

–

8.9 ± 1.4

+ Ca2+ (2 equiv.)

8.5 ± 0.7

10.2 ± 0.9

8.9 ± 0.6

–

–

a) Values for hydrodynamic diameters not provided for polydisperse samples.

All of the examined DCAs exhibited changes in particle diameter distribution upon the addition of Ca2+. Of the responsive DO3A derivatives DCAs1-2, DCA2 underwent the most significant change by almost doubling in size upon binding Ca2+ (Table 2). This expansion is greater than that observed for DCA1 and the previously reported propyl derivative.14 Interestingly, DCAs3-5 also displayed observable changes in diameter through the addition of Ca2+. Nonetheless, the values for the hydrodynamic diameters of DCA4 and Ca-DCA5 have not been reported here due to the polydisperse nature of the samples, indicating possible aggregate formation of the dendrimeric particles, especially in presence of Ca2+. The diameter change that was observed with all DCAs is likely to be influenced by parameters such as the increased rigidity of the EGTA-derived chelator and the electrostatic interactions that occur across the molecule due to changes in charge of the EGTA-derived unit upon the binding of Ca2+.


21


Page 22 of 28

Additionally, these observations suggest that size change alone cannot explain the remarkably different relaxometric behaviours previously discussed. The diameter change obviously affects the r2 of all the investigated DCAs; however, the magnitude of change differs in the two different classes, specifically resulting in significant r2 changes for DCAs1-2 and moderate for DCAs3-5. These findings suggest that only through a combination of both hydration change and an increase in diameter can these incredible changes be realised. The ‘non-responsive’ DCAs3-5 gave various starting diameters with different magnitudes of change when exposed to Ca2+ (Table 2). DCA3 possesses a neutrally charged macrocycle due to the amide in the fourth N-position whereas DCAs4-5 contained an extra carboxylic acid which gives an overall net charge of the macrocyclic unit of -1. The impact of the variance in charge in addition to the structural differences between the two linkers resulted in different particle diameters. Interestingly, when comparing DCA4 and DCA5 the structural differences lie on the dendrimer surface. The capping of the remaining free primary amines on the surface of DCA4 with methyl isothiocyanate was carried out with the intention to neutralise the surface charge of the dendrimer unit. Even though the absolute distributions for DCA4 and Ca-DCA5 complexes have not been reported here, the capping procedure appeared to prevent some of the intermolecular interactions or aggregation that was observed for DCA4 in the DLS measurements. These observations clearly establish that creating variations in the charge distribution across the system can have significant effects on the overall behaviour of the nanoparticle in terms of size, further highlighting the sensitivity of DCAs to their environment.12 Additional manipulations of the environmental conditions through the use of acidic and basic buffered media (pH 5.5 and pH 8.5) emphasised these sensitivities, resulting in a set of completely different size changing behaviours and polydispersities prior to and after Ca2+


22

Page 23 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

addition (data not shown). Most notably, the effect of acidic and basic environments seemed to reduce the expansive capabilities of the responsive DO3A-type dendrimers DCA1-2, limiting them to negligible/no change in diameter upon Ca2+ addition.

Zeta Potential studies. The zeta potential of a system is an important characteristic that is routinely investigated in nanosystems such as dendrimers, liposomes or colloidal gold.25 This term describes the electrical potential at the interface of the electric double layer of the nanoparticle and the bulk medium, which is related to the surface charge of the nanosystem.25-27 To this end, we investigated the zeta potential properties for each dendrimeric system in order to learn more on their organization in solution and surface charge. Consequently, the values were measured both before and after the addition of Ca2+. The initial negative charge of each system decreased upon Ca2+ binding, due to the discharging of the EGTA-derived unit (Table 3). These results are indeed logical, indicating that the binding of Ca2+ takes place with all DCAs, and as a consequence the negative surface charge of the nanoparticles drops due to the addition and complexation of positively charged Ca2+ ions. Moreover, the initial values indicate exposure of the negatively charged EGTA-derived moieties along with the neutral or negatively charged MR reporters on the nanoparticle surface. Also, there were no major differences observed between the absolute zeta potential values across the DCAs, neither before nor after Ca2+ addition. Finally, although discharging of the Ca-DCA complexes converges towards net neutral charge, none of the performed relaxometric and DLS studies indicated poor solubility or suspension formation for any of the investigated systems, except of the aggregate formation for DCA4 and Ca-DCA5, as observed in DLS experiments (see above).


23


Page 24 of 28

Table 3. The zeta potential change of each DCA upon the addition of Ca2+. Conditions: 0.3 mM [Gd3+], pH 7.4 (25 mM HEPES). DCA

1

2

3

4

5

No Ca2+

-20.4 ± 1.3

-18.8 ± 1.0

-20.2 ± 2.0

-15.2 ± 1.4

-13.2 ± 1.8

+ Ca2+ (2 equiv.)

-5.2 ± 0.4

-3.8 ± 1.3

-0.6 ± 1.0

-3.9 ± 0.3

-6.0 ± 1.2

CONCLUSIONS In this work we reported the development of a novel series of DCAs and characterised their relaxometric and size changing behaviours. The findings of this study have established that the macrocyclic structure of any attached monomeric unit and the environmental conditions impact the overall size changing capabilities of the DCA. It could be concluded that the requirements to maximise an increase in the r2/r1 ratio change are a combination of changes in size, hydration number, and rigidity of the system. Neither of these properties alone gives rise to a significant enhancement of r2 values, suggesting that only their synergistic combination leads to the most pronounced changes. Overall the results of this systematic study have aided in determining specific characteristics that are important when aiming to design a highly responsive dendrimeric system which can be utilised for ratiometric MR imaging, thus providing a blueprint for any future preparations of these types of nanoparticles. Moreover, with the development of the particular system bearing the butyl linker between the Ca-chelating and MR reporting moieties (DCA1), we have improved the r2/r1 ratio as well as the overall r1 response compared to that which was previously published.14 These results enable flexibility in the use of this nanoprobe for both conventional T1-weighted and rapid r2/r1 ratiometric imaging for dynamic fMRI studies.


24

Page 25 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Combined with the expected slow diffusion in tissue, the remarkable response of DCA1 to Ca2+ also makes it an ideal candidate for future investigations in vivo.

SUPPORTING INFORMATION The synthetic procedures for obtaining ligands L1-4. This material is available free of charge via the Internet at http://pubs.acs.org.

ACKNOWLEDGMENT The authors thank German Research Foundation (DFG, grant AN 716/7-1) for the financial support and Prof. Thilo Stehle for access to the DLS instrument.

AUTHOR INFORMATION Corresponding Author [email protected] Notes The author declares no competing financial interest.

REFERENCES 1. Caravan, P., Strategies for increasing the sensitivity of gadolinium based MRI contrast agents. Chem. Soc. Rev. 2006, 35 (6), 512-523. 2. Geraldes, C. F. G. C.; Laurent, S., Classification and basic properties of contrast agents for magnetic resonance imaging. Contrast Media Mol. Imaging 2009, 4 (1), 1-23. 3. Davies, G. L.; Kramberger, I.; Davis, J. J., Environmentally responsive MRI contrast agents. Chem. Commun. 2013, 49 (84), 9704-9721.


25


Page 26 of 28

4. Merbach, A. E.; Helm, L.; Tóth, É., The chemistry of contrast agents in medical magnetic resonance imaging. 2nd ed.; Wiley: Chichester, 2013. 5. Heffern, M. C.; Matosziuk, L. M.; Meade, T. J., Lanthanide Probes for Bioresponsive Imaging. Chem. Rev. 2014, 114 (8), 4496-4539. 6. Hingorani, D. V.; Bernstein, A. S.; Pagel, M. D., A review of responsive MRI contrast agents: 2005–2014. Contrast Media Mol. Imaging 2015, 10 (4), 245-265. 7. Carril, M., Activatable probes for diagnosis and biomarker detection by MRI. J. Mater. Chem. B 2017, 5 (23), 4332-4347. 8. Zamponi, G. W., A Crash Course in Calcium Channels. ACS Chem. Neurosci. 2017, 8 (12), 2583-2585. 9. Angelovski, G., Heading toward Macromolecular and Nanosized Bioresponsive MRI Probes for Successful Functional Imaging. Acc. Chem. Res. 2017, 50 (9), 2215-2224. 10. Martinelli, J.; Thangavel, K.; Tei, L.; Botta, M., Dendrimeric β-Cyclodextrin/GdIII Chelate Supramolecular Host-Guest Adducts as High-Relaxivity MRI Probes. Chem. Eur. J. 2014, 20 (35), 10944-10952. 11. Longmire, M. R.; Ogawa, M.; Choyke, P. L.; Kobayashi, H., Dendrimers as high relaxivity MR contrast agents. WIREs Nanomed. Nanobiotechnol. 2014, 6 (2), 155-162. 12. Margerum, L. D.; Campion, B. K.; Koo, M.; Shargill, N.; Lai, J. J.; Marumoto, A.; Sontum, P. C., Gadolinium(III) DO3A macrocycles and polyethylene glycol coupled to dendrimers - Effect of molecular weight on physical and biological properties of macromolecular magnetic resonance imaging contrast agents. J. Alloys Compd. 1997, 249 (1-2), 185-190. 13. Young, J. K.; Baker, G. R.; Newkome, G. R.; Morris, K. F.; Johnson, C. S., "Smart" Cascade Polymers - Modular Syntheses of 4-Directional Dendritic Macromolecules with Acidic, Neutral, or Basic Terminal Groups and the Effect of pH Changes on Their Hydrodynamic Radii. Macromolecules 1994, 27 (13), 3464-3471. 14. Gündüz, S.; Savić, T.; Pohmann, R.; Logothetis, N. K.; Scheffler, K.; Angelovski, G., Ratiometric Method for Rapid Monitoring of Biological Processes Using Bioresponsive MRI Contrast Agents. ACS Sens. 2016, 1 (5), 483-487. 15. Corsi, D. M.; Platas-Iglesias, C.; van Bekkum, H.; Peters, J. A., Determination of paramagnetic lanthanide(III) concentrations from bulk magnetic susceptibility shifts in NMR spectra. Magn. Reson. Chem. 2001, 39 (11), 723-726. 16. Gündüz, S.; Savić, T.; Toljić, D.; Angelovski, G., Preparation and In Vitro Characterization of Dendrimer-based Contrast Agents for Magnetic Resonance Imaging. J. Vis. Exp. 2016, (118), e54776. 17. Gündüz, S.; Nitta, N.; Vibhute, S.; Shibata, S.; Maier, M. E.; Logothetis, N. K.; Aoki, I.; Angelovski, G., Dendrimeric calcium-responsive MRI contrast agents with slow in vivo diffusion. Chem. Commun. 2015, 51 (14), 2782-2785. 18. Cakić, N.; Verbić, T. Z.; Jelić, R. M.; Platas-Iglesias, C.; Angelovski, G., Synthesis and characterisation of bismacrocyclic DO3A-amide derivatives - an approach towards metalresponsive PARACEST agents. Dalton Trans. 2016, 45 (15), 6555-6565. 19. Wei, W. H.; Tomohiro, T.; Kodaka, M.; Okuno, H., Selective synthesis and kinetic measurement of 1 : 1 and 2 : 2 cyclic compounds containing 1,4,7,10-tetraazacyclododecane and azobenzene units. J. Org. Chem. 2000, 65 (26), 8979-8987. 20. Chatterjee, M. N.; Kay, E. R.; Leigh, D. A., Beyond switches: Ratcheting a particle energetically uphill with a compartmentalized molecular machine. J. Am. Chem. Soc. 2006, 128 (12), 4058-4073.


26

Page 27 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

21. Boseggia, E.; Gatos, M.; Lucatello, L.; Mancin, F.; Moro, S.; Palumbo, M.; Sissi, C.; Tecilla, P.; Tonellato, U.; Zagotto, G., Toward efficient Zn(II)-based artificial nucleases. J. Am. Chem. Soc. 2004, 126 (14), 4543-4549. 22. Vibhute, S. M.; Engelmann, J.; Verbić, T.; Maier, M. E.; Logothetis, N. K.; Angelovski, G., Synthesis and characterization of pH-sensitive, biotinylated MRI contrast agents and their conjugates with avidin. Org. Biomol. Chem. 2013, 11 (8), 1294-1305. 23. Gündüz, S.; Power, A.; Maier, M. E.; Logothetis, N. K.; Angelovski, G., Synthesis and Characterization of a Biotinylated Multivalent Targeted Contrast Agent. ChemPlusChem 2015, 80 (3), 612-622. 24. Nwe, K.; Bryant, L. H.; Brechbiel, M. W., Poly(amidoamine) Dendrimer Based MRI Contrast Agents Exhibiting Enhanced Relaxivities Derived via Metal Preligation Techniques. Bioconjugate Chem. 2010, 21 (6), 1014-1017. 25. Clogston, J. D.; Patri, A. K., Zeta Potential Measurement. In Characterization of Nanoparticles Intended for Drug Delivery, McNeil, S. E., Ed. Humana Press: Totowa, NJ, 2011; pp 63-70. 26. Isailović, T. M.; Todosijević, M. N.; Đorđević, S. M.; Savić, S. D., Natural SurfactantsBased Micro/Nanoemulsion Systems for NSAIDs–Practical Formulation Approach, Physicochemical and Biopharmaceutical Characteristics/Performances. In Microsized and Nanosized Carriers for Nonsteroidal Anti-Inflammatory Drugs, Čalija, B., Ed. Academic Press: Boston, 2017; pp 179-217. 27. Lu, G. W.; Gao, P., Emulsions and Microemulsions for Topical and Transdermal Drug Delivery. In Handbook of Non-Invasive Drug Delivery Systems, Kulkarni, V. S., Ed. William Andrew Publishing: Boston, 2010; pp 59-94.


27


Page 28 of 28

Table of Contents (TOC) graphic


28

Synergy of Key Properties Promotes Dendrimer ... - ACS Publications

Recommend Documents