Synthesis, Optical Properties, and In Vivo Biodistribution Performance of Polymethine Cyanine Fluorophores

Near-infrared (NIR) cyanine dyes showed enhanced properties for biomedical imaging. A systematic modification within the cyanine skeleton has been made through a facile design and synthetic route for optimal bioimaging. Herein, we report the synthesis of 11 NIR cyanine fluorophores and an investigation of their physicochemical properties, optical characteristics, photostability, and in vivo performance. All synthesized fluorophores absorb and emit within 610–817 nm in various solvents. These dyes also showed high molar extinction coefficients ranging from 27,000 to 270,000 cm–1 M–1, quantum yields 0.01 to 0.33, and molecular brightness 208–79,664 cm–1 M–1 in the tested solvents. Photostability data demonstrate that all tested fluorophores 28, 18, 20, 19, 25, and 24 are more photostable than the FDA-approved indocyanine green. In the biodistribution study, most compounds showed tissue-specific targeting to selectively accumulate in the adrenal glands, lymph nodes, or gallbladder while excreted to the hepatobiliary clearance route. Among the tested, compound 23 showed the best targetability to the bone marrow and lymph nodes. Since the safety of cyanine fluorophores is well established, rationally designed cyanine fluorophores established in the current study will expand an inventory of contrast agents for NIR imaging of not only normal tissues but also cancerous regions originating from these organs/tissues.

I n the 21st century, with so many advancements in medical science, biomedical imaging modalities proved to be an effective tool for the detection of tumors, cellular structures, and functions. 1Despite the numerous research in this area, there are only six imaging modalities available to surgeons to identify cancerous cells, including X-ray, magnetic resonance imaging (MRI), ultrasound (US), computed tomography (CT), positron emission tomography (PET), and single-photon emission computed tomography (SPECT). 2 However, none of these imaging modalities can be used in real-time imaging.Moreover, radiation associated with some of the imaging modalities (MRI, PET, CT, X-ray) still tied clinicians to depend on their naked eyes during intricate surgery. 3,4Therefore, high demand is growing for the real-time imaging modality, which will not only guide surgeons during complex and arduous surgery but also work as a safer process for both patients and healthcare providers. 5In that context, optical imaging in the near-infrared (NIR) region (650−1700 nm) has gained immense interest from clinicians as an attractive imaging modality to track biological targets.
Exogenous contrast agent plays a vital role in optical imaging. 4,6Important features for any organic dyes to use as contrast agents in optical imaging are absorption/emission profile, molar extinction coefficient, quantum yield, Stokes shift, and photochemical stability.NIR cyanine fluorophores with their abilities of deeper tissue penetration, high imaging resolution, noninvasive performance, low-level attenuation, real-time imaging result, and ability to produce high-quality images are rigorously investigated to use in intraoperative imaging. 7,8Due to their highly tunable structures and improved physicochemical and optical properties, bright fluorophores provide detailed information about the cells under in vivo study with consistent imaging. 9,10The optical properties and fundamental structures of various classes of fluorophores that may be utilized as contrast agents for in vivo optical imaging are displayed in Figure 1.Generally, cyanine fluorophores have two terminal nitrogen-containing heterocycles connected by a polymethine bridge.These heterocycles create electron delocalization across the polymethine chain through electron donor and acceptor capabilities. 11Based on the length of the polymethine chain, cyanine dyes are classified as trimethine cyanine (Cy3), pentamethine (Cy5), and heptamethine (Cy7). 12n vivo biodistribution and clearance are strongly guided by the inherent properties of a fluorophore such as acid−base dissociation constant (pK a ), molecular weight (MW), hydrogen bond donors and acceptors, total surface area, polarizability, distribution or partition coefficients, and photo and thermal stability.Our prior studies showed that in vivo distribution of hydrophobic pentamethine cyanine fluorophores is significantly controlled by physicochemical properties, specially the log D value, 13 known as a distribution coefficient for any ionizable compounds, which measures the lipophilicity at certain pH.Generally, highly lipophilic molecules are less aqueous-soluble because of their poor solubility and low level of permeability.These characteristics increase the chance of off-center targeting and toxicity. 14,15However, several studies indicate that this range correlates with the increase of MW; log D > 1.7 is required for MW above 350 Da, log D > 3.1 is for MW above 400 Da, and log D > 4.5 is for compounds with MW above 500 Da. 7,16ppropriate biodistribution of any contrast agent is important to accumulate in the region or tissue of interest and reduce offcenter targeting.High tunable features of cyanine fluorophores allow for possessing suitable physicochemical properties, which can lead to unique biodistribution patterns.However, the lack of proper optical signals from the contrast agents could result in misleading the biodistribution and clearance patterns. 17Welldefined optical properties with a high extinction coefficient, large Stokes shift, high quantum yield, and high photostability are needed to influence the in vivo performance. 18Besides, several approaches usually are taken to target a specific tissue of interest,

ACS Pharmacology & Translational Science
by adding a targeting ligand to the cyanine molecule, without any targeting ligand, and activable targeting approach. 19Among those, one way is the "structure inherent targeting" approach. 20his concept uses the fluorophore without additional ligands (targeting and isolating) or additional structure complexity to reduce the overall complexity and toxicity during biodistribution.The underlying principle of the structure inherent targeting is that each fluorophore showed inherent pharmacophore properties owning to distinct chemical compositions that lead to a specific biodistribution pathway.
Flexible tunable features of cyanine fluorophores by adding or removing certain functional groups are advantageous for achieving target-selective imaging.Our group previously reported the correlation between structure and biodistribution patterns, which led to targeting various organs such as thyroid and parathyroid, cartilage, and bone-specific imaging. 13,21ecently, we reported that one of the heptamethine cyanine fluorophore 18, which has a log D value of 8.64 at pH 7.4 showed ubiquitous tumor targetability in breast and lung cancer tumors with a high tumor-to-background ratio (TBR). 22Herein, we reported the synthesis of 11 pentamethine and heptamethine fluorophores, quantum computation, optical characteristics, photostability, and in vivo biodistribution performances for these fluorophores.The main objective of this research is to investigate the physicochemical, optical, and photostability properties of the selected cyanine fluorophores and show their effects on the biodistribution character.To the best of our knowledge, no scientific report has been made yet to correlate the effect of these properties combined on their biodistribution pattern.

■ RESULTS AND DISCUSSION
Synthesis.As presented in Scheme 1, the synthesis started with the formation of indole rings 5−7 through Fischer indole synthesis.A starting material 4-substituted phenyl hydrazine derivative 1, 2, or 3 was refluxed under an acidic condition with 3-methylbutan-2-one 4 for cyclization.Each cyclization reaction proceeded through the formation of imine derivatives and refluxed for 48−72 h at 110 °C to complete the reaction.After cooling the reaction mixture to room temperature, substituted indole rings 5−7 were achieved as a brown oil by extracting the reaction mixture in DCM and NaHCO 3 (aq).In the next step, heterocyclic 3H-indolium salts 8−13 were synthesized according to a previously reported procedure by our group, 10 where Nalkylation to the cyclic indole rings was obtained by refluxing with various alkyl halides (Iodomethane, 1-Iodobutane, 1bromo-3-phenyl propane) in boiling acetonitrile.The heterocyclic salts 8−13 were purified by performing several recrystallizations in DCM:ether, acetone:ether, EtOAc/ether, and MeOH/ether.After purification, each of these individual salts was allowed to be condensed with various linkers 14, 15, 16, or 17 separately under basic conditions to form the final desired products.For example, heptamethine cyanine derivatives containing cyclohexenyl rings 18−21 were achieved by the condensation reaction between Vilsmeier−Haack reagent 23 14 and individual heterocyclic salt 8, 9, 10, 12, or 13.Another version of heptamethine cyanine fluorophores with an open chain polymethine bridge as seen in dyes 22−25 were accomplished by the condensation reaction between commercially available linker 15 and individual salt 8, 9, 11, or 10.These condensation reactions were performed under basic conditions between individual salt with linker 14 or 15 at a 2:1 ratio in boiling acetic anhydride for 4−6 h.For the pentamethine cyanine fluorophores 26−28, the same condensation reaction conditions were maintained.For these products, individual heterocyclic salt 9, 8 or 11 was allowed to condense with polymethine chain 16 or 17. 24 Immediately after the reaction mixture was quenched, it was cooled down to room temperature.Finally, the crude products were purified by flush column chromatography in 5% methanol in DCM and dried under vacuum to obtain green crystals for the heptamethine cyanine fluorophores and blue crystals for the pentamethine cyanine fluorophores.
Physicochemical Properties.In silico calculations of physicochemical properties, which included well-defined rotatable bonds, hydrogen bond donor−acceptor groups, hydrophobicity, and polar surface areas are crucial for rationally designing drug molecules.These properties can correlate the The data calculated (at pH 7.4) include log D, polarizability, number of rotatable bonds (nrotb), molecular volume (MV, Å 3 ), topological polar surface area (TPSA, Å 2 ), and molecular weight (MW, Da).
relationship between an optimized structure and its biological activities, such as permeability, retention, and clearance.Modifying the structures of any small molecules can alter these properties.For example, for an increased lipophilicity of a certain fluorophore at a certain pH, log D (distribution coefficient) predicts that molecules will partition faster into lipid cell membranes.The topological polar surface area (TPSA) predicts intestinal absorption. 25Higher TPSA means low permeability because of more retention in the vascular system, such as a larger molecular volume greater than 40 kDa can increase the residence time in the peripheral compartments. 26owever, the low MW can increase the permeability.Therefore, to predict the localization of the fluorophores in a specific tissue, permeability, cellular uptake, retention, and clearance the physicochemical properties were examined by ChemAxon (JChem plugin). 27The results obtained from in silico calculations are presented in Table 1.Most molecular volumes of the synthesized fluorophores lied between 531 and 699 Å 3 , while their TPSA remained the same at 6.25 Å 2 .The log D values at pH 7.4 for the fluorophores 18−28 are fallen between 7.44 and 10.91, and the polarizability value changes from 64 to 88.Among the log D values presented in Table 1, fluorophore 21 showed the lowest hydrophobicity value, whereas fluorophore 19 showed the highest.
Our preliminary observation is that a large alkyl chain increases hydrophobicity.In addition to large alkyl chains (phenyl propyl), fluorophore 19 also has three Cl atoms in the structure and a cyclohexenyl ring in the middle, which are the reason behind the highest log D at pH 7.4.Although TPSA remains the same for all of the fluorophores, polarizability and molecular volume change with the structures.We observed a linear relationship between molecular volumes and polarizabilities.Polarizability increased with the increment of molecular volume.Fluorophore 19 is the most polarizable among the fluorophores with its largest molecular volume (698.88Å 3 ), while fluorophore 27 is the least polarizable with its lowest molecular volume (531.45Å 3 ).

Optical Characterization and Explanation of Absorption and Emission Profiles of the Selected Fluorophores by DFT Calculations.
In vivo success of any contrast agent highly depends on the absorption and emission profiles, molar extinction coefficient, quantum yield, molecular brightness, physicochemical properties, and photochemical stability. 24The synthesized fluorophores can be classified into two main groups, which are distinguishable based on their polymethine chain lengths: heptamethine cyanine fluorophores, ranging from 18 to 25, and pentamethine cyanine fluorophores, specifically 26 to 28.Different types of polymethine bridges have a substantial impact on the structural geometry of the fluorophores and the absorption and emission profiles.Initially, we tried to understand the substituent effect on optical properties in four different media.We measured optical properties in polar protic EtOH, polar aprotic DMSO, and two buffer solutions, phosphatebuffered saline (PBS, pH ∼ 7.4) and 4-(2-hydroxyethyl)-1piperazine-ethanesulfonic acid (HEPES, pH ∼ 7.4).Organic solvents such as ethanol (EtOH) and dimethyl sulfoxide (DMSO) are often used in spectroscopic experiments to observe the absorption and emission maximum, whereas buffer solvents PBS and HEPES were used to mimic the biological environment, as these synthesized fluorophores are intended to be used in bioimaging.The absorbance profiles of these fluorophores were recorded at various concentrations (0.8, 1.6, 2.4, 3.2, 4.0 μM) and the absorbance values obtained were plotted against each of the concentrations.The measured media showed a linear correlation between absorbance and concentration, which is observed in Figures S1−S11, in accordance with the Beer−Lambert law.Measured absorption and emission profiles are summarized in Table 2, where heptamethine cyanine dyes 18−25 absorb around 649−815 nm and pentamethine cyanine dyes 26−28 absorb within 610−659 nm.The emission maxima for heptamethine cyanine fluorophores were also within around 668−817 nm, and pentamethine cyanine fluorophores were recorded in between 667 and 675 nm.
The highest absorbance maximum was observed at 815 nm for open chain heptamethine cyanine 22 in PBS; however, heptamethine cyanine with cyclohexenyl ring 19 exhibited the highest emission wavelength at 817 nm in DMSO.We observed that the absorbance maxima red-shifted by 5−20 nm in DMSO but blue-shifted (20−130 nm) in buffer solutions.As reported, the emission spectra are highly influenced by the nature of the media, and large spectral shifts are observed in buffer solutions. 28Absorbance maxima (λ max ) of heptamethine cyanine fluorophores 18−21 with cyclohexenyl rings were observed in the range of 661−805 nm, whereas open chain heptamethine fluorophores 22−25 were observed between 649−815 nm.However, peak broadening was observed for these fluorophores as they aggregate in buffer solutions.Among the tested fluorophores, most of the fluorophores 18, 23, 24, 26, and 28 showed H-aggregates in both buffer solutions.In contrast, heptamethine cyanine dye 22 exhibited distinct aggregations when placed in different buffer solutions.Specifically, in PBS, Jaggregates were observed, causing a red shift in the absorption maximum to 815 nm.Meanwhile, in HEPES, H-aggregates were observed, which resulted in a blue shift of the absorption maximum to 726 nm.Different substituted groups, such as chlorine and bromine, have a negligible effect on the dye absorption and emission spectra.However, an extension of the pi system led to a red-shifted absorbance and fluorescence when comparing fluorophore 18 to fluorophore 26 (Table 2).Upon excitation at the absorbance maximum, 661 nm, the emission wavelength of fluorophore 18 in HEPES buffers was seen at 802 nm, which gave the largest Stokes shift, 140 nm, observed among these fluorophores.
To understand the contribution of the geometry and the pi system on the optical properties, quantum chemical calculations were performed on the selected fluorophores by frontier molecular orbitals (FMOs) in optimized structures based on the density functional theory (DFT) method with a 6-311G basis set.In comparison to pentamethine, the presence of one extra double bond conjugation in heptamethine fluorophores  introduced more vibrational energy levels and decreased the energy difference between the ground states and the excited states.Moreover, the presence of a cyclohexenyl ring in the polymethine bridge enhances rigidity, made the overall structure flat, and decreased the energy band gap between FMOs. 29igure 2 shows the energy difference of optimized molecular structures between FMOs of fluorophores 19, 23, and 26.The energy differences between the highest occupied molecular orbitals (HOMOs) and lowest unoccupied molecular orbitals (LUMOs) increased from 1.00 to 2.07 eV.These fluorophores have the same terminal indolium heterocycles but different polymethine chains such as fluorophore 19 with a cyclohexenyl ring in the polymethine bridge, fluorophore 23 with an open polymethine chain, and fluorophore 26 with one double bond shorter conjugation length polymethine chain.As reported, energy is inversely proportional to its wavelength.The wavelength for absorption and emission maxima decreased as following fluorophores 19λ abs /λ em > 23λ abs /λ em > 26λ abs /λ em with the increase of the energy difference from left to right as shown in Figure 2.
In addition, we performed DFT calculations on fluorophores 18, 20, and 21 presented in Figure 3 to understand the substituent effect on the absorption and emission wavelengths.We observed that the substituents, such as the halogen atoms (Cl/Br), do not affect the energy gap, whereas the presence of a hydrogen atom changes the HOMO−LUMO gap slightly.Therefore, fluorophore 18, having a substituent Cl atom in the heterocyclic rings, and fluorophore 20, containing a Br atom in the heterocyclic rings, showed almost similar absorption and emission wavelengths.But fluorophore 21, without any halogen substituent showed slightly less absorption and emission wavelengths.This outcome can be explained by the resonance and induction effect of the halogen atom.Halogen atoms in the heterocyclic rings possibly increase the electron density by pushing electrons to the conjugation system.On the other hand, it stabilized the conjugation of the polymethine bridge by withdrawing electrons through an inductive effect.This phenomenon pushes the overall HOMO and LUMO gap narrower.For that reason, halogen atoms substituents in the heterocyclic rings and in the polymethine bridge provide a little more stable structures with higher absorption and emission maxima than their substituent hydrogen atoms.
The molar extinction coefficient (Figures S12−S15) is an intrinsic property of any fluorophore and is important for high image resolution.Generally, fluorophores with high molar absorptivity and quantum yield showed high molecular brightness, which means that less dosing is required for in vivo applications.These properties vary with the structure change and are difficult to predict beforehand.The optical properties of the synthesized fluorophores were measured in various media.All of the data are summarized in Table 3. Synthesized fluorophores showed higher molar extinction coefficients and molecular brightness in organic solvents but lower extinction coefficients and brightness in buffer solutions.Due to their hydrophobic nature, all of the synthesized fluorophores tend to aggregate when in aqueous environments.This aggregation causes a reduction in their molar extinction coefficient and molecular brightness, ultimately affecting their ability to absorb and emit light.Furthermore, the aggregated state of the dye molecules can result in fluorescence quenching, which further diminishes their brightness. 30We observed molar extinction coefficients of fluorophore 18 in the decreasing order 270,000, 203,000, 44,000, and 41,000 cm −1 M −1 in EtOH, DMSO, HEPES, and PBS, respectively.We also found that fluorophore 18 showed significantly lower quantum yield and molecular brightness in buffer solutions.We observed that the molar extinction coefficient of fluorophore 18 decreased in the order of 270,000, 203,000, 44,000, and 41,000 cm −1 M −1 in EtOH, DMSO, HEPES, and PBS, respectively.Fluorophore 18 also exhibited significantly lower quantum yield and molecular brightness in buffer solutions.The highest molar extinction coefficient (270,000 cm −1 M −1 ) was observed for fluorophore 18, while the highest molecular brightness (79,664 cm −1 M −1 ) was observed for fluorophore 21, and the highest quantum yield (0.33) was observed for fluorophores 21 and 24 in EtOH.In contrast, the lowest molar extinction coefficient (27,000 cm −1 M −1 ) was observed for fluorophore 24, the lowest molecular brightness (208 M −1 cm −1 ) was observed for fluorophore 22, and the lowest quantum yield (0.01) was observed for fluorophores 20 and 21 in PBS.Furthermore, it was observed that the heptamethine cyanine dyes (18−22) had higher extinction coefficients in organic solvents than pentamethine cyanine dyes (26−28).
Photostability Study.In consideration of theranostic (therapeutics and diagnostic) applications, NIR cyanine fluorophores are synthesized for fluorescence imaging, photodynamic therapy (PDT), 31 photothermal therapy (PTT), 32 optoacoustic imaging, 33 and cancer immunotherapy. 22One of the approaches before using these fluorophores in those applications is to study their biodistribution and clearance pattern in vivo.Generally, after intravenous injection of these contrast agents, 4−8 h is required to localize in the targeted area and clear from other background tissues and organs. 29,34rganic fluorophores used as contrast agents in those processes are light-sensitive, and prolonged exposure to light can induce photodegradation.Photobleaching is an outcome of the photodegradation process, which may cause unwanted toxicity and harmful effects.Also, it can limit the overall optical applications of the fluorophores during the in vivo process.Photobleaching is typically observed in the long-wavelength cyanine fluorophores because of their long polymethine bridge when they are in solution. 35Besides, it is important to know about the handling and storage protocols of these fluorophores.Therefore, we performed photostability studies on selected fluorophores 18, 19, 20, 24, 25, and 28 versus the FDAapproved indocyanine green (ICG) by continuous irradiation with a xenon lamp at 150 W for 2 h.The power density of 150 W is much higher than the required power for fluorescence imaging, considering that the time it takes for a fluorophore to clear from the background and accumulate in sufficient quantities in the targeted tissues is unknown.However, using high-power sources can be justified if the fluorophore remains stable and demonstrates better photostability than ICG under these conditions.This is because a fluorophore, which showed better photostability by enduring a high light source would exhibit less photobleaching if there is a low energy source throughout the process.By considering this, an aliquot of the stock solution (1 mM in DMSO) was diluted in ethanol, and the fluorescence intensity was measured at the excitation wavelength same as their absorption maxima (nm).Photodegradation rates were measured every 20 min intervals for the selected fluorophores based on the % reduced fluorescence intensity starting from 100%.We observed that in dark conditions at room temperature, the selected fluorophores showed no photobleaching for over 24 h, while in the presence of light, all of the fluorophores showed photodegradation.Obtained results are presented in Table S16 and Figure 4.All of the tested fluorophores showed better stability than ICG possibly due to their rigid structure with fewer rotatable bonds, which reduces their susceptibility to chemical and photo-degradation.In contrast, ICG has a more flexible structure with more rotatable bonds (14 nrotb), making it more prone to degradation over time under exposure to light. 36 4), 25 nmol of each fluorophore was injected intravenously.As shown in Figure 5, heptamethine fluorophores 18 and 20, containing butyl chains on nitrogen atoms of heterocyclic backbone exhibited high   22 On the other hand, as shown in Figure 6, fluorophore 22 has a similar structure to fluorophore 18, without a cyclohexenyl ring and a mesochlorine, and showed reduced signals in the bone marrow, lymph node, adrenal gland, liver, and gallbladder compared to fluorophore 18.Interestingly, lipophilic heptamethine fluoro- Abbreviations used are AG, adrenal gland; BM, bone marrow; Du, duodenum; Ga, gallbladder; He, heart; In, intestine; Ki, kidney; Li, liver; Lu, lung; LN, lymph node; Pa, pancreas; and Sp, spleen.Arrows and arrowheads indicate the targeted organs.The SBR of each organ/tissue relative to the muscle was quantified and labeled as −, 1 to 2; +, 2 to 3; ++, 3 to 5; and +++, > 5. phores 23, 24, and 25 without a central cyclohexenyl ring and a meso-chlorine atom had high signals in the sternum bone marrow, gallbladder, and spleen, with improved targetability to the adrenal glands.Among these, fluorophore 23 with a terminal chlorine substituent and N-phenyl propyl showed the highest signal intensity in the bone marrow, spleen, and lymph nodes compared to other fluorophores, including those with bromine and N-butyl chain substitution.Interestingly, the pentamethinebased structures 26, 27, and 28, such as halogen substituents and the N-butyl or N-phenyl propyl chain, provided high uptake in the liver and rapid hepatobiliary clearance with decreased signals in the bone marrow, adrenal gland, and lymph nodes due to the increased refractivity compared with heptamethine cyanines (Figure 7).
Table 4 shows the summary of targeting properties and biodistribution of heptamethine (18−25) and pentamethine (27−28) fluorophores.Overall, the meso-chlorinated heptamethine fluorophore 21 showed higher background tissue uptake with circulation in the bloodstream for a longer period of time compared to other substitutions due to the rapid binding to serum proteins. 22Pentamethine cyanines (26−28) are smaller than heptamethines and more susceptible to MW and log D values, thus excreting faster than larger lipophilic heptamethines (23−25). 13

■ CONCLUSIONS
In a search of suitable contrast agents for fluorescence imaging in the NIR window I region, eight heptamethine cyanine fluorophores 18−25 and three pentamethine cyanine fluorophores 26−28 were synthesized successfully with relatively high yields.We examined their physicochemical properties including well-defined rotatable bonds, hydrogen bond donor−acceptor groups, hydrophobicity, and polar surface areas.The calculated log D values at pH 7.4 for the fluorophores 18−28 are greater than 7.0, their polarizability values change from 64 to 88, and their molecular volumes are between 531 and 699 Å 3 , while their TPSA remained the same at 6.25 Å 2 .All presented fluorophores showed a high molar extinction coefficient and high quantum yield, and heptamethine cyanine fluorophores with a cyclohexenyl ring showed significantly higher extinction coefficients than other synthesized fluorophores.DFT calculations showed that the cyclohexenyl ring in the polymethine bridge enhances rigidity, made the overall structure flat, and decreased the energy band gap between frontier molecular orbitals (FMOs).We also observed that selected fluorophores 18, 19, 20, 24, 25, and 28 showed way better photostability than FDA-approved heptamethine cyanine dye ICG.Due to the relatively high log D values (>7.0 at pH 7.4), most of the synthesized compounds showed strong signals in the bone marrow, gallbladder, adrenal gland, and lymph node with hepatobiliary clearance.Among the tested, hydrophobic fluorophore 23 targets the bone marrow and lymph nodes within 4 h post-injection upon intravenous injection, showing that changes in physicochemical properties significantly affect the biodistribution of these fluorophores via cellular uptake and clearance.Interestingly, pentamethine fluorophores 26−28 provided rapid hepatobiliary clearance with decreased signals in the secondary lymphoid tissues, despite the substitution of a halogen and an N-butyl or N-phenyl propyl chain due to the increased refractivity compared with heptamethine cyanines.
■ EXPERIMENTAL SECTION Materials.All chemicals and solvents used in the synthesis and purification were of American Chemical Society (ACS) grade or HPLC grade.These chemicals were purchased from Fisher Scientific (Pittsburgh, PA), Sigma-Aldrich (Saint Louis, MO), Combi-Blocks (San Diego, CA), and Acros Organics (Pittsburgh, PA).TLC plates have silica gel 60 F 254 (Merck EMD Millipore, Darmstadt, Germany) used to guide the completion of the reaction.A flash column chromatographic technique was used to purify the final fluorophores by using 60− 200u, 60A classic column silica gel (Dynamic Adsorbents, Norcross, GA).Both 1 H NMR and 13 C NMR spectra were obtained using high-quality Kontes NMR tubes (Kimble Chase, Vineland, NJ) rated to 500 MHz and were recorded on a Bruker Avance (400 MHz) spectrometer using chloroform-d or DMSO-d 6 containing tetramethylsilane (TMS) as an internal calibration standard set to 0.0 ppm.All chemical shifts were recorded in parts per million (ppm).Signals are labeled as follows: s = singlet, d = doublet, t = triplet, m = multiplet, dd = doublet doublets, br = broad, and coupling constants (J) are measured in hertz (Hz) throughout the experimental section.UV−vis/NIR absorption spectra were recorded on a Varian Cary 50 spectrophotometer.The melting points were determined with a Mel-temp melting point apparatus and were corrected.
Optical Study, Photostability, Physicochemical Properties, and DFT Calculation. 1 mM stock solution in DMSO of the synthesized fluorophores was prepared before starting any spectral measurement.Optical properties for the synthesized fluorophores were measured in four different solvents (EtOH, DMSO, HEPES, and PBS).Absorbance spectra were measured in a Varian Cary 50 absorbance spectrophotometer (190−1100 nm).Fluorescence emission and photostability were measured and emission spectra were recorded on a Shimadzu RF-5301PC spectrofluorometer.Physicochemical properties (Log D, molecular mass, TPSA, H bond D/A, polarizability, molecular volume, and rotatable bonds) were predicted by ChemAxon (JChem plugin).Theoretical calculations of frontier molecular orbitals (FMOs) for the optimized structures of selected fluorophores were calculated based on DFT calculations at the B3LYP/6-311G(d,p) level.
In Vivo Biodistribution Study of NIR Fluorophores (Compounds 18−28).Under the supervision of the MGH IACUC, animals were housed in an AAALAC-certified facility and studied according to institutional protocol #2016N000136.In preparation for injection, CD-1 mice (25−30 g, 6 weeks, Charles River Laboratories, Wilmington, MA) were anesthetized with isoflurane and oxygen.A 10 mM stock solution of NIR fluorophores was prepared in DMSO, and 25 nmol of each molecule was diluted in BSA-containing saline (5 wt/v%) for intraoperative imaging.After 4 h post-injection, mice were imaged.For each experiment, camera exposure time and image normalization were held constant.
Quantitative Analysis.The fluorescence and background intensity of a region of interest (ROI) over each tissue was quantified using ImageJ software (NIH, Bethesda, MD) version 1.45q.The signal-to-background ratio (SBR) was calculated as SBR = fluorescence/background, where background is the signal intensity of neighboring tissues such as muscle.All NIR fluorescence images for a particular fluorophore were normalized identically for all conditions of an experiment.
General Procedure for the Preparation of Fluorophores 18−28.Our research group previously reported the synthetic procedure of the precursors of the cyanine fluorophores. 13,19,22,37Individual heterocycle indolenium salts 8−13 (2 mol equiv) were taken in a clean dried round bottom flask (100 mL).These salts were then dissolved in acetic anhydride (5 mL) with Vilsmeier−Haack reagent 14 (1 mol equiv), commercially available glutaconaldehydedianil hydrochloride 15 (1 mol equiv), N-((1E,2Z)-2-chloro-3-(phenylamino)allylidene)benzenaminium 16 (1 mol equiv), or N-((1E,2Z)-2-bromo-3-(phenylamino) allylidene) benzenaminium 17 (1 mol equiv) in the presence of acetic anhydride (5 mL) and sodium acetate (3.5 mol equiv) to yield the cyanine fluorphores 18−28, respectively.The reaction mixture was vigorously stirred for 2−8 h at 65 °C.During this period, the solution gradually turns dark green for the heptamethine cyanine fluorophores and dark blue for the pentamethine cyanine fluorophores.The reaction mixtures were monitored by vis/NIR spectrophotometry via analyzing the change of the relative ratios between the expected absorption band (> 750 nm for fluorophores 18−21, 22−25, >650 nm for fluorophores 26− 28) and the starting material absorption peaks (<500 nm) in methanol and the synthesis, followed by using thin layer chromatography (TLC) in dichloromethane (DCM) and 5% methanol as the eluting solvent.The individual reaction mixture was allowed to cool down to room temperature, and then the solid of each dye is collected and purified via flash column chromatography using 5−10% methanol in DCM.The solution of the pure fractions was collected and condensed under reduced pressure to produce dark green solids, which are then dried under a vacuum.All of the synthesized fluorophores were characterized and confirmed by 1 H NMR and 13 C NMR.The final fluorophores were determined to be >95% pure with fair to very good yields (41−72%).

Figure 1 .
Figure 1.Core structures of various classes of fluorophores with their optical characteristics and absorption/emission profile considered as potential contrast agents for in vivo optical imaging.Scheme 1. Synthetic Route for the Preparation of Cyanine Fluorophores 18−28

Figure 2 .
Figure 2. Theoretical calculations of frontier molecular orbitals for the optimized structures of cyanine 19, 23, and 26 based on DFT calculations at the B3LYP/6-311G(d,p) level.

Figure 3 .
Figure 3. Theoretical calculations of frontier molecular orbitals for the optimized structures of fluorophores 18, 20, and 21 based on DFT calculations at the B3LYP/6-311G(d,p) level.
Fluorophores 18 and 28 degraded 1−4%, fluorophores 20 and 19 degraded 7− 12%, and fluorophores 24 and 25 degraded 16−22%, while ICG degraded 41% at the end of 2 h.The photostability of the fluorophores decreases in the order 28 > 18 > 20 > 19 > 25 > 24 > ICG, which demonstrates that these dyes with short polymethine bridge fluorophore 28 and the presence of cyclohexenyl rings in fluorophores 18, 19, and 20 in the polymethine bridge showed increased photostability.Among the tested fluorophores having terminal heterocyclic indolium rings with a phenyl propyl group and polymethine bridge without the cyclohexenyl ring in the middle 24 showed significant photobleaching than its counterparts.Biodistribution Study on Synthesized Compounds 18−28.Newly synthesized fluorophores 18−28 were injected into CD-1 male mice to study their biodistribution and tissue/ organ-targeting characteristics.4 h prior to sequential intraoperative imaging (Figures 5−7, Table

Figure 4 .
Figure 4. Photostability data of the selected fluorophores 18, 19, 20, 24, 25, and 28 compared against commercially available FDA-approved heptamethine cyanine dye ICG.All of the data were collected in every 20 min intervals under continuous irradiation with a xenon lamp at 150 W for 2 h and presented based on the % reduced fluorescence intensity starting from 100%.

Figure 5 .
Figure 5. Targeting and biodistribution of heptamethine NIR fluorophores 18−21.25 nmol of each fluorophore was injected intravenously into CD-1 mice 4 h prior to imaging and resection.Abbreviations used are Du, duodenum; He, heart; In, intestine; Ki, kidney; Li, liver; Lu, lung; Mu, muscle; Pa, pancreas; and Sp, spleen.Arrows and arrowheads indicate the targeted organs.

Figure 6 .
Figure 6.Targeting and biodistribution of heptamethine NIR fluorophores 22−25.25 nmol of each fluorophore was injected intravenously into CD-1 mice 4 h prior to imaging and resection.Abbreviations used are Du, duodenum; He, heart; In, intestine; Ki, kidney; Li, liver; Lu, lung; Mu, muscle; Pa, pancreas; and Sp, spleen.Arrows and arrowheads indicate the targeted organs.

Table 1 .
Physicochemical Properties of Fluorophores 18−28 Were Calculated by Using Marvin and JChem Calculator Plug-Ins (ChemAxon) a

Table 3 .
Summarized Optical Properties of Fluorophores 18−28 a All measurements were performed in 2 different organic solvents (EtOH and DMSO) and 2 different buffer (HEPES and PBS), and ICG was used as a reference (Φ ICG = 0.043). a