Unraveling the Chemistry of meso-Cl Tricarbocyanine Dyes in Conjugation Reactions for the Creation of Peptide Bonds

Tricarbocyanine dyes have become popular tools in life sciences and medicine. Their near-infrared (NIR) fluorescence makes them ideal agents for imaging of thick specimens or in vivo imaging, e.g., in fluorescence-guided surgery. Among other types of cyanine dyes, meso-Cl tricarbocyanine dyes have received a surge of interest, as it emerged that their high reactivity makes them inherently tumor-targeting. As such, significant research efforts have focused on conjugating these to functional moieties. However, the syntheses generally suffer from low yields. Hereby, we report on the reaction of meso-Cl dyes with a small selection of coupling reagents to give the corresponding keto-polymethines, potentially explaining low yields and the prevalence of monofunctionalized cyanine conjugates in the current state of the art of functional near-infrared dyes. We present the synthesis and isolation of the first keto-polymethine-based conjugate and present preliminary investigation in the prostate cancer cell lines PC3 and DU145 by confocal microscopy and discuss changes to optical properties in biological media.

T he development of functional near-infrared (NIR)emitting fluorescent probes for image-guided surgery as well as multimodal imaging and personalized medicine is of great interest for biomedical applications and life sciences research. Some of the essential design requirements for these imaging probes for in vivo applications have recently been reviewed. 1−3 In the NIR region of the electromagnetic spectrum, scattering coefficients, absorption, and autofluorescence of biological tissues reach local minima, resulting in overall low attenuation and allowing for larger penetration depths of the emitted light. 4−6 Most prominent commercially accessible dyes for imaging in this region are those based on pentamethine and heptamethine cyanine frameworks, as well as the closely related tricarbocyanine dyes. 1,2,7−10 The latter offer an additional handle for functionalization if the mesoposition features a halide. 8,11 Such dyes are highly reactive, with substitution reactions in the meso-position believed to proceed via an S RN 1-type mechanism. 12 In this context, derivatization of this Cy7 site can drastically alter the photophysical properties, 11 due to substantial changes of the charge distribution along the polymethine backbone. 13 In addition, this high reactivity gives the meso-Cl dyes inherent cancer-targeting properties, as reactions with thio nucleophiles, like cysteine sidechains in proteins and peptides, occur under physiological conditions. 14,15 It has been reported that the reaction of meso-substituted NIR dyes with human serum albumin (HSA) may be responsible for the tumor-targeting properties observed in vivo. Such bioconjugates would readily form in vivo. 16,17 In this context, it has been of interest to investigate the role of subsets of endothelial cells, such as the recently described nanoparticle-transporting endothelial cells (NTECs). 18 Additionally, meso-substitution reactions have been applied to prepare targeted NIR-emitting probes, 8,19−21 improve photophysical properties, or introduce functionalities for the design of activatable fluorescent probes. 22, 23 Mellanby et al. demonstrated the potential of meso-N-triazole-functionalized dyes for long-term imaging, 24 and Wang et al. demonstrated the emergence of targeted, multimodal imaging probes using the N-triazole scaffold. 25 We aimed to synthesize bioconjugates of meso-Cl dyes with a selection of standard peptide coupling agents and to incorporate tumor-homing targeting groups (Scheme 1). We focus hereby on addressing the synthetic challenges which were raised by the laboratory-scale synthesis and functionalization of dye 1, a sulfonated analogue of the popular MHI-148 (Figure 1), and the isolation and characterization of new bioconjugates derived from this Cy7 framework. Synthetic avenues targeted at a scalable peptide coupling chemistry were explored hereby with an aim to access a new toolkit for the assembly of modular probes for fluorescence imaging. Taken   20,26,27 together, these observations prompted our curiosity-driven investigation aiming to describe the reasons behind the low conjugation yields at the peptide bond formation and analytical purification under typical reverse-phase chromatography conditions (including the use traces of trifluoroacetic acid (TFA) in the reverse-phase high-performance liquid chromatography (HPLC)) as well as the prevalence of monofunctionalized cyanine conjugates in the current state of the art of functional NIR dyes. Literature accounts of amide bond formation from meso-Cl dyes report low yields, and mostly monofunctionalization of dyes. 32 Laboratory-scale syntheses of water-soluble dyes were complicated by difficult separations in aqueous environment and low yields which appeared to be caused by side reactions and decompositions during purification and HPLC analysis. To obtain the meso-Cl tricarbocyanine 1, an early synthetic procedure (reported in 2004) was followed. 28 However, in our hands, material of sufficient purity was not forthcoming following this protocol. As such, we developed a modified synthetic purification procedure (reliant on the selective precipitation of 1 from dilute hydrochloric acid, see the Supporting Information) which enabled the facile separation from starting materials and by-products. The optical properties of compound 1 were subsequently determined (see the Experimental Section and the Supporting Information): overall, all spectroscopic measurements are in agreement with recent literature of similar dyes in aqueous systems and broadly comparable to the recently developed QuatCy and the established MHI-148 dye ( Figure 1). 29 In vitro investigations of 1 in living cells (specifically, in the common PC3 cell line, derived from the bone metastasis of a stage 4 prostate cancer patient) indicated no significant inhibition of cell metabolism or growth by the compound used in concentrations of up to 250 μM, similar to other results highlighting the low cytotoxicity of sulfonated cyanines. 30,31 We then aimed to synthesize a library of imaging agents incorporating the cyanine framework 1 as a highly modular building block for fluorescence imaging of various cancer cells, using coupling reactions of this meso-Cl dye with glucosamine, a simple [7,13]bombesin peptide fragment, or the urea-based derivative EuK (or LysCOGlu).
As mentioned above, literature accounts of meso-Cl tricarbocyanine dye conjugates obtained through amide bond formation with demanding reaction partners typically report low yields 32 and the state of the art also predominantly features monofunctionalized derivatives (Scheme 1). The same is not true, however, for cyanine dyes that do not feature meso-Cl substituents, where coupling chemistry was reported to proceed more successfuly. 33,34 The reasons for this feature or attempts at explaining the nature of the low yields have not been fully explored, and the outcomes of coupling chemistry involved therefore confined this otherwise promising synthetic building block′s functionalization chemistry to the realm of analytical, rather than synthetic scaled-up approaches. Hereby, upon attempting the standard peptide coupling conjugation reactions to deliver various bioconjugates, we noticed that the color of the solutions changed drastically during the reaction (see Figure 2 and the Supporting Information). For the case study of coupling one or two EuK tags (denoted LysCOGlu) to the −COOH groups of 1 using the reagent PyBOP, the reaction mixture was monitored by ultraviolet−visible (UV− vis) spectroscopy in the region of 350−900 nm. The absorption spectra were recorded (a) before and (b) after the peptide coupling reaction was carried out, and the corresponding spectra are depicted in Figure 2. This shows the absorption assignable to 1 and that of the resulting postcoupling crude reaction mixture. The intense band at ca. 560 nm suggested the formation of the corresponding ketopolymethine as a side product. 13,35 Earlier work on pH sensors considered the conversion of its cyanine ketone derivatives to hydroxy cyanines upon protonation and established that the inherent visible light absorption of a simple keto-polymethine type MHI-148-O (with λ max = 535 nm) shifts significantly to the NIR region (with λ max = 709 nm) due to pH variations. 36 Attempts to isolate the desired meso-Cl dye conjugate 1-EuK were challenging and typically gave indications of the formation of the desired compound in very low yields (<10%) or gave only some evidence in spectroscopic measurements of the formation thereof. This is in line with the observations by other authors for similar approaches on peptides. 29,32 The peptide coupling reactions endeavored hereby to obtain meso-Cl dye bioconjugates on a laboratory scale, which simultaneously retain the −Cl site and incorporate

ACS Bio & Med Chem Au
pubs.acs.org/biomedchemau Article at least one conjugated peptide bond-linked functionality, only resulted in the isolation of traces of the desired conjugates on an analytical scale. Evidence for the formation of such meso-Cl bioconjugates was gathered by mass spectrometry (see the Supporting Information). It is well-known that the reaction of meso-Cl tricarbocyanine dyes with N-hydroxysuccinimide (NHS) gives rise to keto-polymethines. 36 Analogous functionalization reactions upon treatment with other coupling reagents have not been described. Upon testing the stability of 1 in DMF, in the presence of N-hydroxysuccinimide (NHS), 1-hydroxybenzotriazole (HOBt), or K-Oxyma Pure, we determined that all three reagents lead to the formation of the keto-polymethine 2, as shown in Figure 3. Apart from the main product 2, other side products seem to be produced for coupling chemistry reactions of the type shown in Figure 2, and which have not been unambiguously identified despite extensive mass spectrometry investigations of the crude reaction mixtures (Supporting Information). The presence of side products was most apparent from reactions with HOBt, and reactions with both NHS and K-Oxyma Pure yielded unassignable side products. A possible reason may be that intermediates formed with HOBt may be more reactive or that water (from the addition of HOBt·H 2 O) may facilitate their formation.
Compound 1 has limited solubility in common organic solvents: it is possible that certain cyanine/solvent combinations could limit the formation of the keto-polymethine or indeed lead to other products or product mixtures. The underlining process for its reactivity in coupling chemistry is expected to involve the addition of N-hydroxy-based coupling agents to the meso-Cl position, followed by the N−O bond homolysis to proceed to the ketone product, although the exact mechanism has not been established to date.
These coupling reactions were performed at room temperature and under conditions suitable for coupling chemistry for the formation of bioconjugates: as such further optimization and work with the control of the kinetic rather than thermodynamic factors involved may be necessary to fully assign the transformation pathways involved.
Interestingly, compound 1 reacted with glutathione in a straightforward manner, giving rise to a discrete, new mesosubstituted compound (3, which was isolated and fully characterized spectroscopically, as shown in the Experimental Section), whereas compound 2 did not react with glutathione under the same conditions, indicating stability toward endogenous N-and S-based nucleophiles.
Further investigations on the isolated keto-polymethine 2 indicated that the acidic pH of the initially used eluent mixtures containing 0.1% trifluoroacetic acid (TFA) led to protonation and subsequently to decomposition of dye 2 and its derivatives, which we postulate as the reason deemed to complicate purification by semiprep HPLC. We observed that the use of TFA, at concentrations commonly used in standard HPLC methods (0.1 vol %), leads to protonation, evident from a significant bathochromic shift in the absorption spectroscopy, and subsequent decomposition, evidenced by absorption spectroscopy and mass spectrometry. We also noted that the use of traces of TFA in the standard HPLC purification leads to decomposition, whereas mild acid conditions confirm the expected conversion of the cyanine ketone derivatives to hydroxy cyanines (Supporting Information). However, this well-documented pH dependence did not form the topic of this investigation into coupling chemistries involving the peptide bond formation at the exocyclic substituents. We discarded photobleaching as the primary reason of decomposition for the protonated compound, as samples dissolved in water and acidified with 0.1 vol % TFA decomposed at a similar rate, irrespective of whether they were exposed to light or kept in the dark (Figures 3 and S12−S14). In organic solvents, their stability was markedly increased, suggesting an involvement of water in the decomposition. Based on these observations, a slightly modified work-up procedure was subsequently used, as follows: The dye was loaded onto the cartridge again and washed with a 3 M solution of NaCl in water, before being eluted with an unmodified mobile phase (deionized water/acetonitrile gradient, 0−50% CH 3 CN). This procedure allowed the isolation of the new dye, compound 2 on a milligram scale, which was then characterized by 1 H NMR spectroscopy, including DOSY, high-resolution mass spectrometry (ESI-HRMS, negative mode), UV−vis, and fluorescence spectroscopies, and deemed to be of purity >95% by integration in HPLC. Optimized geometries for simplified analogues of 1 and 2, featuring N-methyl substituents, instead of N-hexanoic acid substituents, were investigated by computational approaches (TDDFT, with details given in the Supporting Information).
Measurements of the optical properties of 2 in PBS suggested a molecular brightness of ∼12,800 at low μM concentrations: a comparison with Compound 1 is given in Figure 3. For a wider comparison, the molecular brightness of the clinically used PpIX gives values below 10,000 in phosphate buffered saline (PBS) and fetal bovine serum (FBS). 37 The presence of the oxo-substituted framework does not lead to NIR-emitting fluorophores; however, they may be of interest as microscopy tools in a wide range of life sciences applications, e.g., biochemical research tools. Given the success of fluorescein and protoporphyrin IX in fluorescence-guided surgery, the lack of NIR fluorescence may not be prohibitive of in vivo use. 38, 39 The related meso-amine dyes, which also show blue-shifted absorption and emission profiles, tend to give lower molecular brightness but are of interest to biomedical applications. 11,40,41 MTT assays with dye 2 indicated that this does not cause a reduction in metabolic activity in PC3 cells ( Figure S42) and laser confocal imaging indicated a rather low cellular uptake overall ( Figure 4 and the Supporting Infomation).
Interestingly, upon treating the well-studied, commercial dye MHI-148 with either HOBt or K-Oxyma in DMF, the reactions with these standard coupling agents appeared to proceed much faster than for the case of 1, and typically reached completion within 30 min, giving rise to the analogous species to 2 on an analytical scale (Supporting Information and Figure S1). The scale-up reaction protocols for the resulting keto-polymethine derivative of MHI-148 (denoted MHI-148-O) proved difficult in our hands. The exclusive use of carbodiimide coupling agents failed to give sufficient conversions and instead appeared to lead to the formation of N-acylisoureas as side products (Figures S2−S5). Attempts at forming cyanine conjugates using alternative coupling agents, such as propylphosphonic anhydride (T3P), failed. As expected, the solubility of the resulting product emerging from the oxidation of MHI-148 in PBS was greatly reduced with respect to that of compound 2. Also, unlike the case of 2, for the product resulting from the oxidation of MHI-148, no fluorescence emission was observed in PBS ( Figures 5 and  S51): we postulate that a significant level of aggregation occurs in aqueous solutions. This was also suggested by the fact that the same substance in methanol did show a fluorescence response, in agreement with a keto-polymethine structure. Confocal fluorescence images were obtained from living cancer Fluorescence microscopy experiments showed only limited emission from cells incubated with 2. However, at higher concentrations (100 μM) and after longer incubation times (2 h) at 37°C, a visible cellular uptake was observed for 2 ( Figure  S49). The low uptake of highly polar keto-polymethines into human cell lines was previously observed with structurally related dyes, and our results further confirm this for prostate cancer cells. 35 The oxo form of the ubiquitous MHI-148 (the nonsulfonated analogous of compound 2, MHI-148-O) appears to be readily taken up by cancer cells even at 10 μM concentrations. Other authors have previously noted the influence of structure and charge on the pharmacokinetics of cyanine dyes and cyanine−antibody conjugates. 42,43 Further cellular imaging micrographs are given in the Supporting Information (Figures S44−S50). Also, the excitation−emission maps (EEMs) of the merocyanine 2 and the product of the MHI-148 oxidation, MHI-148-O, showed contrasting emission−excitation profiles when measured in PBS or 10% serum medium, as shown in Figure 5. The results suggest that in the presence of proteins, supramolecular interactions occur which lead to the emergence of a distinct, hypsochromically shifted absorption and emission profile (also see the Supporting Information, Figure S51). Pascal et al. previously hypothesized that this emission stems from aprotic environments around the dye and may be used to probe intracellular environments. 35 Specifically, our experiments suggest that the aggregation to hydrophobic surface regions of proteins may be the central reason for the observed spectroscopic behavior in aqueous environments.
Keto-polymethines may well be used as dyes in multiphoton imaging applications in their own right, as indicated in the state of the art. 35 Therefore, an in-depth incursion into the formation and isolation of new, water-soluble functionalized keto-polymethine dyes as alternatives to their meso-Cl counterparts was deemed worthwhile. A new representative of this class, compound 2, already showed the advantage that it could be isolated on a milligram scale, in HPLC purity of above 95%. Following the optimization of work-up procedures, its functionalization protocols were developed (see the Experimental Section) and the new, symmetric, glucosaminefunctionalized derivative, conjugate 4, was also synthesized and purified ( Figure 6, Experimental Section, and Supporting Information). Its characterization was performed using HPLC, ESI-HRMS, and NMR in line with the standard approaches for other Cy7-based dyes, commercial or noncommercial, and the

ACS Bio & Med Chem Au pubs.acs.org/biomedchemau
Article purity was deemed above 95% according to HPLC quantification at 560 nm detection. The cellular uptake of compound 4 was also investigated under similar conditions to those applied for its precursor, compound 2. Following cellular uptake experiments in living PC3 cells, statistical analysis of these in vitro revealed very limited uptake of compound 4 at low concentration (10 μM) and 30 min incubation time, and this was enhanced at 100 μM concentration and 2 h incubation time at 37°C (Figure 4). The fact that compound 4 was not internalized more efficiently and rapidly by living cells at the low concentrations tested may be related to the high hydrophilicity of the large cyanine backbone. Previously, it has been reported that water-soluble keto-polymethines were not efficiently taken up by T24 cells and that the uptake of a Cy5.5 glucosamine conjugate was not glucose transporter (GLUT) specific in certain cell lines. 35,44,45 Additionally, the synthesis of a new bioconjugate from the peptide coupling reaction of compound 2 and the [7,13]bombesin fragment led to isolation of compound 5 on the analytical scale, and its characterization only by ESI-HRMS and HPLC. (Supporting Information, Figures S38 and S39). The scale-up synthesis for this compound and the assessment of its ability to target prostate cancer cell lines expressing the gastrin-releasing peptide receptor (GRPR) will be the focus of future work. As highlighted above, the development of smallmolecule conjugates of cyanine dyes has been an ongoing endeavor for several decades; however, only very few meso-Cl dye conjugates have been described in the literature. 5 Our results seem to suggest that the reason for this, apart from the generally challenging nature of indocyanine chemistry, resides in the high reactivity of the meso-Cl center of these dyes. Future work in our group will focus on alternative methods of synthesizing conjugates of meso-Cl dyes: our trials toward the synthesis of peptide bioconjugates of 1 without the use of Nhydroxy-based agents also gave mixtures of multiple species. Mass spectroscopy analysis of isolated fractions suggested the presence of N-acylurea side products following peptide coupling chemistry protocols (see the Supporting Information): the formation of these may well be suppressed by solvents with lower dielectric constant. Using more recently developed methods to form amide bonds, such as the method developed by Mishra et al., 46 may be a way to address the difficult syntheses of conjugates of meso-Cl tricarbocyanines. We postulate that the nonpolar solvents necessarily used in many of these reactions would likely be incompatible with the nature of sulfonated cyanine dyes.
In conclusion, we recognized that meso-Cl-substituted NIR dyes react with a variety of coupling reagents to give the corresponding keto-polymethines as the major product, as well as other decomposition species, which could not be fully identified in the course of this work. The resulting ketopolymethines are unstable in acidic environments, which may be the reason why these, and their conjugates, have so far been elusive, and the reports on these species in the current literature remain surprisingly scarce, despite the widespread need for dyes of relevance to clinical imaging applications. As stated above, a recent review by Mieog et al. discussed the requirements for developing agents for fluorescence-guided surgery, with a particular focus on biomolecules. 3 To address this limitation of the state of the art, we developed and reported hereby on our new approach and underlining synthetic method, e.g., to isolate and characterize the ketopolymethines, and their rationally designed, synthesized, and characterized new conjugates, which were discussed hereby. We further investigated the optical properties of some ketopolymethines in biological media, revealing increased fluorescence as well as supramolecular chemistry, which may involve aggregation to proteins. The findings described here are of relevance in the efforts to improve the development of robust and scalable synthetic methodologies for meso-Cl cyanine functionalization. This will lead to discrete, wellunderstood bioconjugated and functional probes that open opportunities for new synthetic chemistry for multimodality imaging applications.

■ EXPERIMENTAL SECTION General Experimental Methods
All reactions were performed under ambient atmosphere unless otherwise mentioned. Solvents were peptide (N,N-dimethylformamide, DMF, and dimethyl sulfoxide, DMSO) or HPLC (acetonitrile, methanol, ethanol) grade. Reagents were obtained commercially and used without further purification. Degassed solvents were obtained using three consecutive freeze−pump−thaw cycles. Deoxygenated solvents were obtained by bubbling argon gas through the respective solvent for 10 min. Solid-phase peptide synthesis (SPPS) was performed using a Biotage Alstra Initiator +. For synthetic operations, 30 mL reactor vials were used according to manufacturer specifications. Fmoc-protected L-amino acids were dissolved in DMF, for a final concentration of 0.6 mol/L. Syntheses of known dyes, compound 1 and MHI-148, are based on adapted protocols based on the literature methods, 47−51 and these are detailed in the Supporting Information.
Chromatography was performed using a Biotage Isolera system equipped with a reverse-phase C 18 -silica cartridge (Sfar Bio C18 − Duo 300 Å 20 mm, 30 g). A gradient of acetonitrile and water or methanol and water was used to purify and elute compounds. For details on buffers used, the reader is referred to the relevant experimental procedure. To remove excess ammonium formate from purified meso-oxo dyes, they were eluted once more using an unbuffered mobile-phase. High-performance liquid chromatography (HPLC) was performed using a Dionex UltiMate 3000 preparative system, equipped with a 50 μL loop for analytical and a 2 mL loop for semipreparative work and an eight-channel UV−vis detector (UltiMate 3000 Diode Array Detector). For analytical work, a C 18silica column by Hamilton (PRP1, internal diameter 4.1 mm, length 150 mm, particle size 10 μm, pore size 100 Å) was utilized. Integration of chromatograms was performed using commercial Chromeleon software. Methods used were as follows: method A: water/methanol ( The final sulfonated dye compounds isolated on a laboratory scale display a purity of >95% (NMR, ESI-HRMS, and HPLC analysis, taken together). The specific batches of the as-synthesized compound 1 and of its new derivatives 2, 3, and 4 which were used for biochemical investigations hereby were of analytical purity by integration in HPLC using the Chromeleon software (>95%). All spectroscopic data including NMR, HPLC, and corresponding ESI-HRMS with matched isotopic patterns are given below and in the Supporting Information.
NMR spectra were recorded on a Bruker Neo (400 MHz) spectrometer with SampleCase sample changer. Spectra were measured at 400.13, 376.49, and 100.61 MHz for the acquisition of 1 H, 19 F, and 13 C, respectively. Chemical shifts are reported in ppm, with the solvent residual peak used as an internal standard: DMSO-d 6 , δ = 2.50 for 1 H NMR spectra and δ = 39.52 for 13 C; CD 3 -OD, δ = 3.31 for 1 H NMR spectra and δ = 49.00 for 13 C. Data is reported as ACS Bio & Med Chem Au pubs.acs.org/biomedchemau Article follows: s = singlet, d = doublet, t =triplet, m = multiplet, and br = broad. Coupling constants are given in Hz.
High-resolution mass spectra of cyanine dyes were recorded on a Bruker MAXIS HD ESI-QTOF. Parameters are given in the Supporting Information. For mass spectrometry of starting materials, intermediates, and peptides, an automated Agilent QTOF was used. For peptides, an HPLC/MS method was chosen, eluting the injected sample over a reverse-phase C8-HPLC column, with a water/ acetonitrile mobile phase (0.1% formic acid). Analysis of mass spectrometry data was performed using commercial Bruker and Agilent software. Deviations between measured m/z values and predicted m/z values are given as absolute values in ppm. Fouriertransform infrared (FTIR) spectra were recorded on a PerkinElmer Spectrum 100 IR spectrometer with an attenuated total reflection (ATR) module. The signals listed are those unambiguously assigned to a functional group.
UV−vis spectra were recorded on a PerkinElmer Lambda 650 spectrometer using quartz cuvettes with a path length of 1 cm. Solutions were prepared by dissolving weighed samples of the purified compounds in the appropriate solvent. Dilutions were prepared using standard Eppendorf pipettors. Extinction coefficients were determined with at least four different concentrations using a linear regression according to the Beer−Lambert law. To assess stability toward endogenous species, dilutions were prepared using deionized water either 10 mM glutathione or 1 mM L-ascorbic acid, and the samples were kept in the dark. At various time points (t = 0, 2, 4, 8, 24, 72 h), 60 μL aliquots from these solutions were taken, diluted to 1.5 mL, and their absorption was measured.
Fluorescence spectra were recorded on a PerkinElmer LS55 fluorescence spectrometer using quartz cuvettes with a path of 1 cm × 1 cm. Fluorescence maps (EEMs) were recorded with 10 nm excitation increments at a scan speed of 100 nm/min. Relative fluorescence quantum yields were determined with both the PerkinElmer LB650 and the PerkinElmer LS55 spectrometers.
Cell lines used in MTT assays and live cell imaging were prostate cancer cells (PC3 and DU145), obtained from the American Type Cell Culture (ATCC). Cell culturing experimental details are given in the Supporting Information. Before microscopy experiments, cells were seeded onto sterile glass dishes and incubated for 48 h before addition of fluorescent compounds to allow them to adhere to the surface. Confocal microscopy was performed using a Nikon Eclipse Ti2-E inverted confocal microscope with an LU-N3 laser unit (405, 488, and 561 nm). Compound 1.
Compound 1 was isolated on a 1.35 g scale overall, 18% yield using an adapted multistep protocol, as described in the Supporting Information. 47 In a microwave vial, compound 2, synthesized as above (12 mg, 14 μmol), was dissolved in 2.5 mL of DMF. To this, HBTU (30 mg, 79 μmol) was added, followed by DIPEA (23 μL, ca. 140 μmol) and glucosamine hydrochloride (20 mg, 92 μmol). The mixture was heated to 75°C under microwave irradiation for 90 min. After cooling to room temperature, the mixture was poured into 20 mL of deionized water. The solution was loaded on a reverse-phase flash chromatography column and eluted with a water/MeOH gradient (10 mM ammonium formate buffer in mobile phases). The obtained product fraction was concentrated under reduced pressure to remove excess methanol. The obtained aqueous solution was loaded onto a reverse-phase flash chromatography cartridge and washed with a 3 M solution of sodium chloride in water before the product was eluted with unbuffered methanol. The obtained product was concentrated under reduced pressure and finally lyophilized to yield compound 4 as a red solid (11 mg, 66%). 1  In a microwave vial, compound 1 (6 mg, 6.8 μmol) was dissolved in 1.5 mL of DMF. To this, HBTU (3 mg, 7.9 μmol) was added, followed by DIPEA (3 μL, 17 μmol) and [7,13]bombesin (5 mg, 5.4 μmol). The mixture was heated to 50°C under microwave irradiation for 90 min. After cooling to room temperature, the mixture was poured into 20 mL of deionized water and filtered over a pad of C 18silica, which was washed with a 1:1 mixture of acetonitrile and water to elute trapped product. The solution was loaded on a reverse-phase flash chromatography column and eluted with a water/MeOH gradient (10 mM ammonium formate buffer in mobile phases). The obtained product fraction was concentrated under reduced pressure to remove excess methanol. The obtained aqueous solution was loaded onto a reverse-phase flash chromatography cartridge and washed with a 3 M solution of sodium chloride in water before the product was eluted with a gradient of deionized water and methanol (0% MeOH to 75% MeOH). The obtained product was concentrated under reduced pressure and finally lyophilized to yield compound 5 as a red solid, isolated on the analytical scale only (3.7 mg). Integration in the analytical HPLC was approximate due to significant peak broadening assigned to protonation, in line with our previously reported conjugates of [7,13]

■ ASSOCIATED CONTENT Data Availability Statement
The available data is included in the Supplementary Information published with the article.
File denoted the ESI in manuscript contains detailed description of experimental methods and characterization of new compounds, including HPLCs, 1 NMR, UV−vis, fluorescence, and ESI-HRMS spectrometry data; additional cellular assays; imaging micrographs, and density functional theory (DFT) calculation details (PDF)