Clickable, hydrophilic ligand for fac-[M(I)(CO)3](+) (M = Re/(99m)Tc) applied in an S-functionalized α-MSH peptide.

The copper(I)-catalyzed azide-alkyne cycloaddition (CuAAC) click reaction was used to incorporate alkyne-functionalized dipicolylamine (DPA) ligands (1 and 3) for fac-[M(I)(CO)3](+) (M = Re/(99m)Tc) complexation into an α-melanocyte stimulating hormone (α-MSH) peptide analogue. A novel DPA ligand with carboxylate substitutions on the pyridyl rings (3) was designed to increase the hydrophilicity and to decrease in vivo hepatobiliary retention of fac-[(99m)Tc(I)(CO)3](+) complexes used in single photon emission computed tomography (SPECT) imaging studies with targeting biomolecules. The fac-[Re(I)(CO)3(3)] complex (4) was used for chemical characterization and X-ray crystal analysis prior to radiolabeling studies between 3 and fac-[(99m)Tc(I)(OH2)3(CO)3](+). The corresponding (99m)Tc complex (4a) was obtained in high radiochemical yields, was stable in vitro for 24 h during amino acid challenge and serum stability assays, and showed increased hydrophilicity by log P analysis compared to an analogous complex with nonfunctionalized pyridine rings (2a). An α-MSH peptide functionalized with an azide was labeled with fac-[M(I)(CO)3](+) using both click, then chelate (CuAAC reaction with 1 or 3 followed by metal complexation) and chelate, then click (metal complexation of 1 and 3 followed by CuAAC with the peptide) strategies to assess the effects of CuAAC conditions on fac-[M(I)(CO)3](+) complexation within a peptide framework. The peptides from the click, then chelate strategy had different HPLC tR's and in vitro stabilities compared to those from the chelate, then click strategy, suggesting nonspecific coordination of fac-[M(I)(CO)3](+) using this synthetic route. The fac-[M(I)(CO)3](+)-complexed peptides from the chelate, then click strategy showed >90% stability during in vitro challenge conditions for 6 h, demonstrated high affinity and specificity for the melanocortin 1 receptor (MC1R) in IC50 analyses, and led to moderately high uptake in B16F10 melanoma cells. Log P analysis of the (99m)Tc-labeled peptides confirmed the enhanced hydrophilicity of the peptide bearing the novel, carboxylate-functionalized DPA chelate (10a') compared to the peptide with the unmodified DPA chelate (9a'). In vivo biodistribution analysis of 9a' and 10a' showed moderate tumor uptake in a B16F10 melanoma xenograft mouse model with enhanced renal uptake and surprising intestinal uptake for 10a' compared to predominantly hepatic accumulation for 9a'. These results, coupled with the versatility of CuAAC, suggests this novel, hydrophilic chelate can be incorporated into numerous biomolecules containing azides for generating targeted fac-[M(I)(CO)3](+) complexes in future studies.

Introduction

The copper(I)-catalyzed n class="Chemical">azide–alkyne cycloaddition (CuAAC) reaction has become the icon of click chemistry since its introduction by the Sharpless and Meldal laboratories.[1,2] Numerous scientific disciplines (e.g., biology, chemical synthesis, materials science, etc.) have utilized the rapid kinetics, mild conditions, and orthogonality of CuAAC reactions for applications ranging from bioconjugation to polymer synthesis.[3−5] Methods of introducing the azide or alkyne moiety into polypeptides range from standard solid and solution phase techniques to biological incorporation using engineered codons during recombinant expression of proteins.[6−8] Due to these benefits, the number and types of biological applications using CuAAC have grown tremendously in recent years. In the field of radiopharmaceuticals, CuAAC has been harnessed for labeling a variety of biomolecules (e.g., small molecules, peptides, antibodies, etc.) with short-lived radionuclides during targeted nuclear imaging studies.[9,10] The diagnostic radionuclide 99mTc [t1/2 = 6.02 h, 140 keV γ (89%)] has nearly ideal nuclear properties for single photon computed emission tomography (SPECT). The fac-[99mTcI(OH2)3(CO)3]+ core developed by Alberto has gained recognition for SPECT applications due to its small molecular volume, versatile chelation chemistry, straightforward preparation from an aqueous-based kit, and amenability for readily labeling a wide variety of biomolecules.[11−14]

CuAAC has been used to n class="Chemical">incorporate ligands for the fac-[99mTcI(CO)3]+ core into biomolecules by several strategies.[15] In the click, then chelate strategy, a biomolecule functionalized with an azide or an alkyne undergoes cycloaddition with a bifunctional chelator containing the orthogonal coupling partner prior to complexation with fac-[99mTcI(CO)3]+. The click to chelate strategy developed by Schibli and Mindt is a variant of this method where the 1,2,3-triazole generated by the click reaction acts as an integral ligand donor for the fac-[99mTcI(CO)3]+ core.[16,17] In an alternative chelate, then click strategy, a bifunctional chelator is first complexed with 99mTc and then joined to a biomolecule via CuAAC to generate the final radiopharmaceutical. As the click reaction can be accomplished under physiological temperature and pH, this radiolabeling strategy enables the pairing of sensitive biomolecules with ligands that require harsh labeling conditions (e.g., high temperature or pH) to drive complexation kinetics with the fac-[99mTcI(CO)3]+ core.

Tridentate ligands containing aromatic nitrogen donors, such as histidine or 2,2′-dipicolylamine (DPA), have been shown to form kinetically and thermodynamically inert fac-[99mTcI(CO)3]+ complexes due to the chelate effect (tridentate as opposed to mono- or bidentate ligands) and the low-spin d6 metal center.[12,14,18−21] Despite the high in vitro and in vivo stabilities of these complexes, their formation often requires high temperatures (>90 °C) to achieve quantitative radiolabeling yields in acceptable time frames (30–60 min). Additionally, the lipophilic fac-[99mTcI(CO)3(DPA)]+ complexes can lead to undesirable uptake and retention of the labeled radiopharmaceutical in nontarget tissues such as the liver and gastrointestinal tract.[22,23] In an effort to generate a versatile fac-[99mTcI(CO)3]+ complex with increased hydrophilicity and high stability, we installed carboxylate groups on the aromatic rings of an alkyne-functionalized DPA ligand. Using CuAAC would allow the ligand or radiolabeled complex to be easily incorporated into a wide spectrum of targeting molecules containing an azide group. An azide-functionalized α-melanocyte stimulating hormone (α-MSH) analogue was chosen as a model targeting biomolecule to test this new chelate system with fac-[99mTcI(CO)3]+.

Malignant melanoma n class="Chemical">remains the deadliest form of skin cancer due to its high metastatic rate and to challenges in detecting and treating these metastases at early stages.[24−27] Development of targeting peptides labeled with SPECT isotopes to identify metastatic melanoma at early stages continues to be an active area of medical research.[28−35] The majority of these peptides are α-MSH analogues which bind to the melanocortin-1 receptor (MC1R), a biomarker that is overexpressed on primary and metastatic melanoma cells (∼400 to 22 000 receptors/cell) compared to normal cells.[26,36−39] Several α-MSH analogues labeled with fac-[99mTcI(CO)3]+ have shown promising results for detecting models of melanoma in vivo during preclinical studies.[22,32,33,40−42] Utilizing various chelates and attachment strategies with α-MSH peptides has been shown to modulate the in vivo properties of the ensuing fac-[99mTcI(CO)3]-labeled peptides. While many conjugates have demonstrated undesired uptake in the liver and gastrointestinal tract,[22,32,40] introducing carboxylates on a pyrazolyl-based ligand has recently been shown to markedly reduce the hepatobiliary retention of α-MSH peptide analogues after labeling with the fac-[99mTcI(CO)3]+ core.[33]

We hypothesized that the addn class="Chemical">ition of carboxylates on a DPA chelate would increase the hydrophilicity and decrease the nonspecific hepatobiliary uptake of the resulting fac-[99mTcI(CO)3]+ complex compared to the original DPA ligand while maintaining complex stability. To test this hypothesis, we employed a previously characterized DPA-alkyne ligand, 1, and the new, carboxylate-functionalized DPA-alkyne ligand, 3, in parallel studies to incorporate the fac-[MI(CO)3]+ (M = Re/99mTc) core into an α-MSH peptide by convergent pathways (click, then chelate and chelate, then click) as shown in Scheme 1. The in vitro stabilities of the complexed ligands and peptides were determined, and the melanoma targeting potentials of the labeled peptides were evaluated through (1) in vitro MC1R affinity, (2) in vitro B16F10 cell uptake assays, and (3) in vivo biodistribution analyses in B16F10 xenograft models.

Scheme 1

Reaction Schemes and Structures of Compounds

Results and Discussion

In an effort to generate a versatile ligand for the fac-[MI(CO)3]+ core with enhanced hydrophilicity, we synthesized a DPA chelate with carboxylate substitutions on the pyridal rings and a terminal alkyne on the central amine. Reductive amination between propargyl amine and methyl 6-formylnicotinate using sodium triacetoxyborohydride followed by basic hydrolysis of the methyl ester groups provided the desired ligand, 3, in moderate overall yield (55%) following isolation and purification. Standard characterization techniques were used to confirm the structure of the isolated material as the desired compound. The IR spectrum contained a band at 3244 cm–1 which suggested a terminal alkyne group (CH stretching), although the corresponding expected absorption in the 2000–2200 cm–1 region (C≡C stretching) was not present. However, 1H NMR analysis indicated the presence of a propargyl-type system due to the resonances at 3.21 and 4.06 ppm for the terminal and methylene protons, respectively. Reaction of 3 with fac-[ReI(OH2)3(CO)3](SO3CF3) at 70 °C in mildly acidic (pH 6–6.5) aqueous conditions produced the desired complex in moderate yield (66%) following purification by preparatory RP-HPLC. As anticipated, 1H NMR analysis of the isolated product, 4, showed the methylene protons of the DPA moiety as an AB quartet (5.03/4.96 ppm) with a downfield shift from their original appearance as a singlet in the spectrum of 3 (4.62 ppm). ESI-MS analysis confirmed a ligand:Re(CO)3 ratio of 1:1 in complex 4 as expected for tridentate coordination at the DPA chelate site. IR analysis indicated a facial arrangement of the carbonyls on the metal center following complexation as has been observed with other fac-[ReI(CO)3(DPA)]+ complexes.[47,49,50]

The absolute coordn class="Chemical">ination environment of 4 was confirmed through single-crystal X-ray analysis (Figure 1). Crystallographic details, bond lengths, and bond angles are given in the Supporting Information (Tables S5–S7). Crystals of 4 packed in the P21/c space group with four molecules in the unit cell. The bond lengths and angles between the carbonyl ligands and the metal center in 4 (1.918–1.939 Å and 89.18–91.16°, respectively) were as expected for distorted octahedral fac-[ReI(CO)3(DPA)]+ complexes.[47,49] The bond lengths between rhenium and the central nitrogen (2.227 Å) or the pyridal nitrogens (2.163 and 2.176 Å) in 4 were also similar to those in complex 2 (2.223 and 2.171 Å, respectively) as previously characterized.[47] This suggests that including the carboxylates did not significantly perturb complexation with the fac-[ReI(CO)3]+ core. The alkyne was pointed away from the metal center as would be necessary for participation in CuAAC reactions with azides.

Figure 1

X-ray crystal structure of 4 with thermal ellipsoids shown at the 50% probability level. Hydrogen atoms were omitted for clarity. Atoms are colored as follows: C, gray; N, light blue; O, red; Re, dark blue.

Having established that 3 coordinates fac-[ReI(CO)3]+ as desired, we assessed its labeling efficiency with fac-[99mTcI(CO)3]+ to determine its potential for use in SPECT imaging applications. Radiolabeling 3 with fac-[99mTcI(OH2)3(CO)3]+ at 70 °C yielded a single peak with a similar RP-HPLC tR (19.2 min) as complex 4 (18.9 min), indicating successful formation of the desired product, 4a (Figure 2). After determining that complexation occurred as expected, decreasing concentrations of 3 were used to examine the effects of the carboxylate substitutions on the radiolabeling efficiency compared to the original DPA-alkyne chelate, 1. Quantitative labeling was achieved in 30 min at 70 °C when 3 was employed at 1 × 10–4 M with decreasing yields at lower concentrations (Table S1). While extending the reaction time to 1 h increased yields of 4a, labeling efficiency with 3 was lower than with 1 to produce 2a when identical experimental conditions were used.[47] High decay corrected yields of 4a (>91%) could be achieved following RP-HPLC purification. A major objective in preparing the new ligand was to increase the hydrophilicity of the resulting metal complex and to decrease hepatobiliary retention of radiopharmaceuticals with this complex compared to the regular DPA complex. Log P analysis (octanol/PBS partition coefficient) of complexes 2a and 4a yielded values of −0.23 ± 0.01 and −1.85 ± 0.02, respectively, confirming that the carboxylate-functionalized DPA complex was more hydrophilic than the regular DPA complex at physiological pH.

Figure 2

Normalized and offset RP-HPLC chromatograms of 4 (UV absorbance, 254 nm; lower blue trace) at tR 18.9 min and 4a (radiodetector, counts per minute (cpm); higher red trace) at tR 19.2 min.

Targeted radiopharmaceuticals should maintain high biological stability with the radionuclide to avoid nonspecific uptake or retention of radioactive species in nontarget tissues during in vivo applications. Serum stability analyses and amino acid challenge assays were performed with purified solutions of 4a to assess its resistance to transchelation or oxidation under biologically relevant conditions (Table S2). As expected from previous studies employing DPA complexes with fac-[99mTcI(CO)3]+,[20,21,23,51]4a was >98% stable in serum and in the presence of 1 mM histidine or cysteine up to 24 h. These in vitro results suggested that the chelate would be suitable for further exploration within a radiopharmaceutical construct for in vivo applications.

The encouraging radiolabeling results with the novel chelate prompted its evaluation in a targeting peptide framework. A NAPamide α-MSH peptide containing a nonessential cysteine residue (Cys-NAPamide) was chosen as an initial model system to incorporate the ligand or fac-[MI(CO)3]+ complex in a site-specific manner. NAPamide was chosen because of its high MC1R affinity and in vivo stability as demonstrated in previous studies;[39,52,53] we have recently employed a similar peptide with fac-[MI(CO)3]+ using an alternative chelate.[41] The thiol group of cysteine presents an attractive moiety for functionalization due to its lower pKa compared to other nucleophiles in biomolecules (e.g., amines, alcohols). Native cysteine residues are absent from many targeting biomolecules (e.g., peptides, affibodies) or are not accessible due to disulfide bridge formation (e.g., proteins, antibodies).[54,55] Therefore, synthetic or recombinant techniques can be employed to install a nonessential cysteine residue into these scaffolds for selective functionalization at locations that would not impair affinity for the biological target. Cysteine was utilized to append an azide functional group for attaching the alkyne-functionalized chelates to the peptide via CuAAC. The 2-bromo-N-(3-azidopropyl)-acetamide azide linker was chosen because it is stable to hydrolysis in aqueous solution, and it has previously been used for selective S-alkylation of proteins to incorporate an azide functionality for subsequent CuAAC reactions.[56]

The Cys-NAPamide peptide, 5, was pn class="Chemical">repared using solid phase peptide synthesis with Fmoc-protected amino acids in high purity as indicated by RP-HPLC analysis (Figure S1). Following TCEP reduction of the peptide, alkylation was performed overnight at room temperature with a slight excess of 2-bromo-N-(3-azidopropyl)-acetamide in a PBS/MeCN mixture. These conditions ensured maximum alkylation of the starting peptide with minimal side product (e.g., dialkylation) formation. The alkylated peptide, 6, was isolated in moderately high yields (>70%) and high purity (>99% by RP-HPLC, Figure S2) following RP-HPLC purification. ESI-MS analysis confirmed the identity of the product as the desired monoalkylated species (1341.1 m/z, Table 1). IR analysis showed a prominent absorption at 2103 cm–1 as anticipated for the presence of an azide group. Evaluation of 6 with Ellman’s reagent confirmed that that the azide linker had been incorporated at the cysteine residue as no free thiols (<1%) were detected during analysis.

Table 1

Characterization Results for Non-Radioactive Peptides

		calcd MW (m/z)	obsvd MW (m/z)	RP-HPLC tR (min)	IC50 (nM)
	5	[M+2H]2+: 601.8	602.1	19.0a	2.19 ± 0.54
	6	[M+2H]2+: 671.8	672.0	18.7a	9.60 ± 1.74
	7	[M+2H]2+: 790.4,	790.7,	25.2b	--
		[M+2H+Cu]2+: 821.9	821.2
	8	[M+2H]2+: 834.4,	834.7,	25.7b	--
		[M+2H+Cu]2+: 865.8	866.2
Click, then chelate	9	[M+H]2+: 925.4	925.7	14.6c	--
	10	[M+H]2+: 968.9	968.7	14.4c	--
Chelate, then click	9′	[M+H]2+: 925.4	925.5	14.9c	3.57 ± 1.15
	10′	[M–H]−: 1936.7	1936.3	15.3c	3.13 ± 0.29

HPLC method 3.

HPLC method 4.

HPLC method 6.

Following n class="Chemical">incorporation of the azide linker, two pathways were examined to label the peptide with fac-[MI(CO)3]+: click, then chelate and chelate, then click. Both pathways have previously been utilized with DPA ligands to form metal complexes with biologically relevant small molecules through CuAAC reactions.[47,51] The click, then chelate route is analogous to the prevailing strategy in the literature where a targeting species is functionalized with a bifunctional chelator prior to complexation with the metal. The chelate, then click approach was used as an alternative strategy to generate the metal-complexed peptides to determine if the presence of the fac-[MI(CO)3]+ core prior to the CuAAC reaction affected formation of the desired products. CuAAC reactions were performed to append ligands 1 or 3 to peptide 6 in order to yield constructs 7 and 8 containing the standard DPA chelate and the new DPA chelate bearing carboxylates, respectively. Due to the likelihood of copper coordination by the DPA chelates,[51,57] excess Cu(OAc)2 was used in the CuAAC reactions to ensure that enough copper(I) would be available to catalyze the desired cycloaddition reactions. After RP-HPLC analysis indicated completion of the reactions, EDTA or sodium sulfide was added to remove copper and the crude peptides were purified to a single peak by RP-HPLC (Table 1, Figures S3 and S4). ESI-MS analysis indicated that the peptides were produced as desired due to the resonances at 790.7 m/z and 834.7 m/z for the [M+2H]2+ ions of 7 and 8, respectively (Table 1), although copper complexes of these peptides were also apparent as minor species in the MS spectra (821.2 m/z and 866.2 m/z for [M+2H+Cu]2+ ions of 7 and 8, respectively, Table 1). This result is similar to what has been observed previously with the DPA-alkyne ligand, 1, following CuAAC reaction with a functionalized azide.[51] Using excess sodium sulfide to remove residual copper from the peptides by forming CuS was only partially successful; employing precoordinated copper complexes (e.g., tris-(benzyl-triazolylmethyl)amine) during the CuAAC reactions also failed to yield 7 and 8 without residual copper present (data not shown). Despite the presence of copper in the peptides, subsequent reactions of 7 and 8 with fac-[ReI(OH2)3(CO)3](SO3CF3) yielded the rhenium-complexed peptides 9 and 10, respectively, following RP-HPLC purification (Figures S5 and S6). ESI-MS analysis of 9 and 10, which showed the anticipated [M+H]2+ ions at 925.7 m/z and 968.7 m/z, respectively, confirmed coordination of the fac-[ReI(CO)3]+ core to the peptides (Table 1), and no copper was evident during analysis following fac-[ReI(CO)3]+ complexation. Overall yields of 9 and 10 from the starting azide-functionalized peptide 6 via the click, then chelate strategy were 50% and 39%, respectively.

The chelate, then click route was then exn class="Chemical">amined to compare overall reaction yields from 6 and to determine if the products obtained corresponded to those via the click, then chelate route. A particular advantage of the chelate, then click synthetic strategy is that the ligand is complexed with the fac-[MI(CO)3]+ center prior to cycloaddition with the azide on the peptide, assuring that the final metal-complexed product is obtained as desired without the possibility of mixed coordination species. Surprisingly, RP-HPLC analysis of the chelate, then click reactions between peptide 6 and complexes 2 and 4 indicated that species with slightly different tR’s were formed compared to those from the click, then chelate pathway (Table 1). ESI-MS analysis of these new species, designated 9′ for the analogue with the regular DPA complex and 10′ for the analogue with the carboxylate-functionalized DPA complex, following RP-HPLC purification of the reactions confirmed that the fac-[ReI(CO)3]+-labeled peptides were successfully obtained without residual copper via this synthetic pathway (925.5 m/z for [M+H]2+ ions of 9′ and 1936.3 m/z for [M–H]− ions of 10′, Table 1). Yields for 9′ and 10′ obtained by the chelate, then click strategy from peptide 6 were 47% and 70%, respectively. The different tR’s of 9/9′ and 10/10′ suggest that the presence of copper prior to rhenium complexation negatively affected the desired coordination of the fac-[ReI(CO)3]+ core in the click, then chelate strategy. In subsequent reactions, the peptides obtained by the chelate, then click strategy (9′ and 10′) were hypothesized to possess the desired N,N,N coordination mode exclusively at the chelate position because metal complexation with the DPA ligands, which possess well-defined coordination and high stability with the fac-[ReI(CO)3]+ core,[47,49,58] was performed prior to attachment to the peptide. The rhenium-complexed peptides produced by the click, then chelate pathway (9 and 10) were likely coordination isomers of the desired products and were not pursued in the biological assays below.

Peptides 5, 6, 9′, and 10′ wen class="Chemical">re used in B16F10 melanoma cell competitive binding assays with a standard radiolabeled MC1R binding peptide, 125I-(Tyr2)-NDP, to confirm that the modifications made to the peptides through linker attachment and metal complexation did not abrogate their affinity for the MC1R. These assays yielded IC50 values in the low nanomolar range for the peptides (Table 1), although the IC50 values were higher than that of the previously reported NAPamide sequence when tested with B16F1 melanoma cells (0.27 nM).[59] The high affinity of the metalated analogues supported further exploration of the peptides with 99mTc for targeting melanoma in SPECT applications.

Both synthetic pathways used wn class="Chemical">ith the rhenium analogues were explored for producing the 99mTc-complexed peptides. In the click, then chelate pathway, peptides 7 and 8 containing the DPA chelate and the carboxylate-functionalized DPA chelate, respectively, were directly complexed with fac-[99mTcI(OH2)3(CO)3]+ at 70 °C for 1 h in PBS. RP-HPLC analysis showed similar tR’s for these 99mTc-labeled peptides, designated 9a and 10a (Table 2, Figures S5 and S6), as was observed for the respective rhenium analogues, 9 and 10 (Table 1), produced by the click, then chelate strategy. This suggested that similar coordination modes were obtained with both fac-[MI(CO)3]+ cores using these peptides. Radiolabeling yields of 9a and 10a after 1 h at 70 °C were greater than 93% (conversion from fac-[99mTcI(OH2)3(CO)3]+) when the peptides were used at ∼5 × 10–5 M with declining yields at shorter times and lower concentrations as was observed with the ligands (1 and 3) before attachment to the peptides. Purification of the products by RP-HPLC yielded 9a and 10a in moderate decay-corrected yields (Table 2).

Table 2

RP-HPLC tR’s and Isolated Radiolabeling Yields of Peptides Labeled with 99mTc Following RP-HPLC Purification

	Click, then chelate		Chelate, then click
	9a	10a	9a′	10a′
RP-HPLC tR (min)	14.6a	14.4a	15.1a	15.5a
isolated, decay-corrected yields	84%	47%	42%	71%

HPLC method 6.

The chelate, then click pathway was next explored to produce the 99mTc-complexed peptides 9a′ and 10a′. Performing the CuAAC reactions with ∼5 × 10–5 M 6 and crude reaction mixtures of 2a or 4a at 50 °C in PBS yielded 9a′ and 10a′ quantitatively (>98% conversion from 2a or 4a) in as little as 30 min, although decay-corrected yields of 9a′ and 10a′ were moderate following RP-HPLC isolation (Table 2). Comparison of the RP-HPLC tR’s of the Re and 99mTc analogues produced by the chelate, then click route validated the identities of 9a′ and 10a′ (Tables 1 and 2, Figures 3 and 4). As was observed with the rhenium-complexed peptides, different RP-HPLC tR’s of 9a/9a′ and 10a/10a′ were observed for the products of the chelate, then click route compared to the click, then chelate route (Table 2).

Figure 3

Normalized and offset RP-HPLC chromatograms of 9′ (UV absorbance, 220 nm; lower blue trace) at tR 14.9 min and 9a′ (radiodetector, counts per minute (cpm); higher red trace) at tR 15.1 min produced by the chelate, then click route.

Figure 4

Normalized and offset RP-HPLC chromatograms of 10′ (UV absorbance, 238 nm; earlier blue trace) at tR 15.3 min and 10a′ (radiodetector, counts per minute (cpm); later red trace) at tR 15.5 min produced by the chelate, then click route.

The in vn class="Chemical">itro stabilities of the radiolabeled peptides were examined prior to testing their in vivo properties. It was likely that the stabilities of 9a and 10a would differ from 9a′ and 10a′ due to the potentially different fac-[99mTcI(CO)3]+ coordination modes resulting from the click, then chelate and chelate, then click strategies. Peptide 9a′ was >99% stable in the presence of either histidine or cysteine at 37 °C for 4 h with slightly lower stability (84%) in cysteine at 18 h (Table S3). Peptide 10a′ exhibited >99% stability under these conditions for 18 h (Table S3). Both 9a′ and 10a′ were >90% stable for 6 h at 37 °C in serum (Table S4) which indicated they would be candidates for further biological analysis. RP-HPLC analysis of 9a and 10a produced via the click, then chelate route, however, showed multiple species at times beyond 2 h in the above challenge conditions and when incubated in PBS at room temperature following purification (data not shown). This suggests that residual copper which was apparent in peptides 7 and 8 prior to complexation with the fac-[99mTcI(CO)3]+ core led to less stable coordination modes or may have negatively impacted the integrity of the radiolabeled peptides (e.g., through metal-mediated peptide hydrolysis[60,61]). Employing the precomplexed ligands in the chelate, then click strategy did not suffer these drawbacks, thus highlighting the benefit of this synthetic pathway for generating the desired fac-[99mTcI(CO)3]+-complexed peptides, 9a′ and 10a′. Having determined which species were stable in solution, the effects of the two chelates (regular DPA vs carboxylate-functionalized DPA) on the overall hydrophilicity of the peptides was examined. Under the acidic conditions (0.1% TFA) used for RP-HPLC analysis, the analogues bearing the novel chelate (8, 10′, 10a′) had slightly longer tR’s compared to their counterparts with the unmodified DPA chelate (7, 9′, 9a′). This observation could be explained by protonation of one of the carboxylates on the modified DPA chelate at low pH, leading to a zwitterionic form of the chelate in 8 and an overall neutral charge for the metal complexes in 10′ and 10a′. The DPA ligand in 7, on the other hand, was protonated at low pH; complexation with the metal in 9′ and 9a′ retained the positive charge from the fac-[MI(CO)3]+ core regardless of pH. Thus, the tR’s of the compounds under the acidic RP-HPLC conditions did not adequately correlate to the anticipated hydrophilicity of the compounds at physiological pH. Log P analysis at neutral pH confirmed the increased hydrophilicity of 10a′ (−2.03 ± 0.01) compared to 9a′ (−0.68 ± 0.04) as would be desired for generating a targeted peptide with less hepatobiliary retention compared to the more hydrophobic analogue.

Following the radiolabeling and characterization studies above, the in vitro B16F10 melanoma cell uptake and MC1R specificity of 9a′ and 10a′ were explored at 0.5, 1, and 2 h (Figure 5). 9a′ exhibited moderate accumulation in B16F10 melanoma cells at 37 °C, reaching 2.71 ± 0.18%, 3.30 ± 0.44%, and 5.86 ± 1.07% of applied activity at 0.5, 1, and 2 h, respectively. Interestingly, increased cell uptake values were observed for 10a′, reaching 2.75 ± 0.72%, 4.01 ± 0.47%, and 9.40 ± 1.08% at 0.5, 1, and 2 h, respectively. The significantly higher uptake of 10a′ compared to 9a′ at 2 h (P < 0.05) in the melanoma cells was surprising considering that the rhenium-complexed analogues exhibited similar IC50 values in the MC1R binding assays (3.57 ± 1.15 nM for 9′ and 3.13 ± 0.29 for 10′). To confirm whether the binding of 9a′ and 10a′ with B16F10 cells was due to specific interactions with MC1R, the cells were coincubated with solutions containing 9a′ or 10a′ and excess amounts of unlabeled NDP for 0.5, 1, and 2 h at 37 °C. As shown in Figure 5, these blocking conditions caused a decrease in the uptake of the radiolabeled peptides at 2 h from 5.86 ± 1.07% to 1.42 ± 0.08% (P < 0.05) for 9a′ and from 9.40 ± 1.08% to 1.20 ± 0.22% (P < 0.05) for 10a′. Thus, unlabeled NDP significantly reduced the binding of 9a′ and 10a′ to B16F10 cells, indicating that the uptake of the radiolabeled peptides was mediated by the MC1R.

Figure 5

In vitro B16F10 cell uptake assay of 9a′ and 10a′. B16F10 cells were incubated with 1 μCi of 9a′ or 10a′ for 0.5, 1, and 2 h at 37 °C with or without NDP (3 μM) as a blocking agent. Data represent the amount of activity associated with the cells as a percentage of the total activity applied ± standard deviation and are the average of two individual experiments performed in quadruplicate. 9a′ only (black squares with solid black line), 9a′ with NDP (black circles with dotted black line), 10a′ only (red triangles with solid red line), and 10a′ with NDP (red triangles with dotted red line).

In vn class="Chemical">itro B16F10 cell uptake assay of 9a′ and 10a′. B16F10 cells were incubated with 1 μCi of 9a′ or 10a′ for 0.5, 1, and 2 h at 37 °C with or without NDP (3 μM) as a blocking agent. Data represent the amount of activity associated with the cells as a percentage of the total activity applied ± standard deviation and are the average of two individual experiments performed in quadruplicate. 9a′ only (black squares with solid black line), 9a′ with NDP (black circles with dotted black line), 10a′ only (red triangles with solid red line), and 10a′ with NDP (red triangles with dotted red line).

The above in vn class="Chemical">itro stability and cell uptake results encouraged the investigation of 9a′ and 10a′ as potential SPECT peptides for targeting melanoma in vivo. The biodistribution patterns of 9a′ and 10a′ were examined in B16F10 melanoma mouse xenograft models at 1 and 2 h to quantify the uptake of these peptides following intravenous administration. The results are presented in Figure 6. Peptide 9a′ accumulated to the greatest extent in the liver and kidneys (5.92 ± 0.87% and 2.28 ± 0.18% at 2 h p.i., respectively), although retention in the spleen and lung was also observed at 2 h p.i. (1.24 ± 0.35% and 1.00 ± 0.53%, respectively). Other peptides and small molecule targeting ligands with fac-[99mTc(CO)3(DPA)]+ complexes have shown similar biodistribution patterns with high uptake in the hepatobiliary system.[23,62−64] Using the carboxylate-substituted DPA complex in peptide 10a′ was anticipated to decrease the hepatobiliary retention of this analogue relative to 9a′ due to the increased hydrophilicity of the novel chelate. Compared to 9a′, peptide 10a′ exhibited greater renal (3.69 ± 0.40%) than hepatic (2.52 ± 0.61%) accumulation at 2 h p.i.; retention of 10a′ in the blood (1.91 ± 0.33%), lung (1.50 ± 0.48%), pancreas (0.91 ± 0.25%), and intestine (2.87 ± 0.63%) was also higher compared to 9a′ at 2 h p.i. (0.61 ± 0.10%, 1.00 ± 0.53%, 0.07 ± 0.03%, and 0.49 ± 0.15% for blood, lung, pancreas, and intestine, respectively). The different accumulations of 9a′ and 10a′ observed in the liver and kidneys confirmed that incorporation of the carboxylates on the DPA complex in 10a′ altered the properties of the resulting radiopharmaceutical in vivo compared to the original DPA complex in 9a′. However, the amount of 10a′ observed in the intestines suggests the desired goal of decreasing hepatobiliary retention by using the hydrophilic complex was not achieved. The lower level of 9a′ in the intestines at 2 h (0.49 ± 0.15%) compared to 10a′ (2.87 ± 0.63%) suggests that the carboxylate substitutions may have increased the rate of clearance through the liver without significantly modifying total hepatobiliary uptake. Although the greater levels of 10a′ in the kidneys and in the blood and could be due to its increased hydrophilicity (compared to 9a′), prolonged blood retention of the radionuclide using this analogue would adversely affect signal-to-noise ratios of MC1R-expressing tissues during SPECT imaging.

Figure 6

Biodistribution (A) and tumor-to-normal tissue ratios (B) of 9a′ and 10a′ in B16F10 tumor-bearing C57BL/6 mice 1 and 2 h after injection of approximately 100 μCi of 9a′ and 10a′ via the tail vein. Data are expressed as % ID/g ± standard deviation (n = 4 per group).

The two diffen class="Chemical">rent chelates did not appear to affect MC1R affinity in vivo as tumor uptake was comparable for both peptides, reaching 0.79 ± 0.19% for 9a′ and 1.06 ± 0.16% for 10a′ at 1 h p.i. and declining to 0.73 ± 0.05% for 9a′ and 0.71 ± 0.14% for 10a′ at 2 h p.i. The tumor-to-blood ratio (Figure 6) slightly increased for 9a′ at 2 h p.i. (1.2 ± 0.13) compared to 1 h p.i. (0.7 ± 0.17) as would be desired for obtaining SPECT images of target tissues with low background levels at later time points. Although the amount of 10a′ in the blood also decreased over time, its higher overall blood retention led to lower tumor-to-blood ratios (∼0.4 at both 1 and 2 h p.i.) compared to 9a′. Many tissues (stomach, brain, muscle, bone, skin) exhibited lower uptake and retention of 9a′ and 10a′ compared to the melanoma tumors (Figure 6) as would be desired for diagnostic imaging of malignant melanoma. However, due to the rates of melanoma metastases in the hepatobiliary and gastrointestinal systems,[65] additional modifications to 9a′ and 10a′ would be required to minimize the nonspecific retention of these peptides in soft tissues prior to further development in SPECT imaging applications.

A variety of chelates for the fac-[MI(CO)3]+ core have been designed with modifications to increase the hydrophilicity and renal clearance of the resulting complexes.[21,23,33,64,66] Several have employed the single amino acid chelate (SAAC) approach which is useful for inserting ligands at nearly any desired location in peptides made by solid phase synthesis.[21,23,58] Pyrazolyl chelates modified with carboxylate groups have also been employed in α-MSH peptides for targeting melanoma.[33] Both strategies have successfully produced peptides with lower soft tissue retention and greater excretion through renal rather than hepatobiliary pathways compared to unmodified chelate structures. The DPA chelates used in the present study were intended to build on the above concepts by employing a versatile linker strategy that could be utilized with peptides or even larger macromolecules produced by both synthetic and recombinant techniques. Such techniques could be used to insert azide-bearing residues (or orthogonal reactive groups for selectively incorporating azides) for chelate attachment via CuAAC at potentially at any location in the sequence of the biomolecule, thus promoting the flexibility of the chelation strategy used here compared to the other strategies mentioned above.

While the chelates and CuAAC strategies used here are expected to be generally applicable for labeling biomolecules with the fac-[MI(CO)3]+ core, several limitations are worth noting. First, it is possible that incorporating modified residues within highly optimized targeting ligands (e.g., small molecules, short peptides) could negatively affect their binding affinities or pharmacokinetic properties compared to the original targeting moiety. Second, the unstable nature of the metal-complexed α-MSH peptides generated through the click, then chelate route in this study suggests that residual Cu from the CuAAC reaction can potentially lead to undefined fac-[MI(CO)3]+ complexes during applications with complex biomolecules. This drawback, along with other limitations associated with CuAAC click reactions (e.g., purification requirements to avoid Cu toxicity, transchelation, etc.) must be considered during future applications with the click, then chelate strategy in particular. Additionally, using the carboxylate-modified DPA complex with the α-MSH peptide in the present study did not lead to the desired decrease in hepatobiliary retention compared to the unmodified DPA complex. Despite these observations, the hydrophilic ligand for the fac-[MI(CO)3]+ core in the chelate, then click strategy confirmed that its incorporation into a targeting peptide did not significantly impair peptide function when compared with a known DPA complex containing the fac-[MI(CO)3]+ core. The chelate, then click strategy prevents the possibility of undefined or unstable coordination isomers of the metal core at undesired sites in targeting biomolecules and can allow even temperature-sensitive biomolecules to be rapidly labeled with the fac-[MI(CO)3]+ core under mild conditions.[47] Therefore, utilizing this strategy with either of the metal complexes from the present study would be suitable for generating a wide variety of targeting radiopharmaceuticals in future applications.

Conclusions

In summn class="Chemical">ary, a novel chelate based on DPA with enhanced hydrophilicity has been developed and utilized with the fac-[MI(CO)3]+ core. Structural assessment using Re and radiolabeling studies with 99mTc demonstrated that this chelate’s performance was comparable to a well characterized DPA ligand. An α-MSH analogue with an S-functionalized azide linker was used as a model biomolecule to incorporate the alkyne-functionalized chelates and fac-[MI(CO)3]+ complexes into the peptides via CuAAC reactions. Metal complexation through the click, then chelate route produced peptides with inferior in vitro properties compared to those generated by the chelate, then click strategy. The fac-[MI(CO)3]+-complexed analogues generated by the chelate, then click strategy were stable to transchelation, possessed high affinity for the MC1R, and demonstrated encouraging uptake values in B16F10 melanoma cells in vitro. Biodistribution analysis of the peptides bearing the two different DPA chelates demonstrated moderately low uptake in melanoma xenografts in vivo. While the α-MSH peptide with the new carboxylate-functionalized complex showed slightly increased renal accumulation and less hepatic retention than the alternative hydrophobic analogue, the desired goal of decreasing hepatobiliary retention by increasing the hydrophilicity of this analogue was not obtained. However, the versatility of the chelate, then click strategy with either of the chelates used here maintains the potential to incorporate the fac-[MI(CO)3]+ core into a wide variety of biomolecules bearing azide moieties for generating targeted radiopharmaceuticals via CuAAC.

Experimental Procedures

All reagents and solvents wen class="Chemical">re of reagent grade or higher from commercial suppliers (Aldrich, Fluka, Acros, Fisher) and used as received unless noted otherwise. fac-[ReI(OH2)3(CO)3](SO3CF3),[43] methyl 6-formylnicotinate,[44] 2-bromo-N-(3-azidopropyl)-acetamide,[45]N,N-bis(pyridine-2-ylmethyl)prop-2-yn-1-amine[46] (ligand 1), and fac-[ReI(CO)3(1)](SO3CF3)[47] (complex 2) were prepared from literature procedures. [99mTcO4]− was obtained from Cardinal Health (Spokane, WA) or from Stanford Nuclear Medicine Clinic and was used to prepare fac-[99mTcI(OH2)3(CO)3]+ via commercially available Isolink kits (Tyco, Inc.) as previously described.[41]125I-(Tyr2)-[Nle4,d-Phe7]-α-MSH [125I-(Tyr2)-NDP] was purchased from Perkin-Elmer (Waltham, MA). Rink Amide LS resin, hydroxybenzotriazole (HOBt), and 9-fluorenylmethoxycarbonyl (Fmoc) protected amino acids were purchased from Advanced ChemTech (Louisville, KY). Elemental analysis was performed by Intertek Pharmaceutical Services, Inc. (Whitehouse, NJ). Nuclear magnetic resonance (NMR) spectra were recorded at 293 K on 300 or 400 MHz Varian Mercury Vx spectrometers using 5 mm NMR tubes. 1H and 13C NMR spectra peak positions were referenced using residual solvent signals, and spectra were processed using Varian VNWR 6.1 software. Mass spectra were obtained on a Thermo-Finnigan LCQ Advantage instrument for electrospray ionization mass spectrometry (ESI-MS). Infrared (IR) spectra were recorded on a Thermo Nicolett 6700 FTIR with an ATR cell and analyzed with OMNIC 7.1a software. UV/vis spectra were recorded on a Varian Cary 50 Bio spectrophotometer and analyzed with Cary WinUV 3.00 software. Solutions of the compounds were prepared in UV/vis quality methanol and measured in 1 cm path length quartz cuvettes. Analytical separation and identification of small molecules and peptides by reversed-phase high performance liquid chromatography (RP-HPLC) were conducted using a Varian Pursuit XRs column (C18, 5 μm, 4.6 × 250 mm) with a Phenomenex security guard cartridge (C18, 4.0 × 3.0 mm). A Perkin-Elmer Series 200 analytical chromatography system equipped with a Perkin-Elmer Radiomatic 610TR detector and a Hitachi D-7000 series analytical chromatography system (L-7100 pump, L-7400 UV detector) equipped with a Berthold FlowStar LB 513 radiodetector were used for the above compounds. Unless noted otherwise, elutions performed on these systems used a gradient profile (HPLC method 1, previously described as “Gradient 1”)[48] with 0.1% TFA in water (solvent A) and MeOH (solvent B) as solvents. Chromatograms were plotted using OriginPro 8.5.1 (OriginLab Corporation, MA). Preparatory RP-HPLC separations and purifications of small molecules used a Phenomenex Gemini-NX column (C18, 5 μm, 21.2 × 250 mm) with a Phenomenex security guard cartridge (C18, 21.2 × 15 mm) on a Hitachi D-7000 series semipreparative chromatography system (L-7150 pump, L-7400 UV detector). Chromatographic analysis and purification of the starting peptide (5; Scheme 1) was performed on a Dionex Ultimate 300 HPLC system (Dionex Corp.) using both semipreparative (Vydac; 218TP510-C18, 5 μm, 10 × 250 mm) and analytic (Vydac; 214TP54-C18, 5 μm, 4.6 × 250 mm) peptide RP-HPLC columns. The mobile phase gradients used 0.1% TFA in water (solvent A) and 0.1% TFA in acetonitrile (MeCN, solvent C). Radio-HPLC purification and analysis with the radiolabeled peptides used for biological analysis (9a′ and 10a′; Scheme 1) was conducted with a semipreparative RP-HPLC Vydac peptide column on a Dionex 680 chromatography system with a UVD 170U absorbance detector and model 105S single-channel radiation detector (Carroll & Ramsey Associates, Berkley, CA) using a gradient method (HPLC method 2: 1 mL/min flow rate; 0–3 min 95% A/5% C, 3–8 min linear gradient to 72% A/28% C, 8–33 min linear gradient to 67% A/33% C, 33–36 min linear gradient to 5% A/95% C, 36–39 min hold 5% A/95% C, return to 95% A/5% C after 39 min and equilibrate). The recorded data were processed with use of Chromeleon v 6.50 software (Sunnyvale, CA) and were plotted using Origin 8.0 (MicroCal). The B16F10 murine melanoma cell line was obtained from American Type Culture Collection (Manassas, VA). Female C57BL/6 mice were purchased from Charles River Laboratory (Wilmington, MA).

Small Molecule Synthesis and Complexation Reactions

6,6′-((Prop-2-yn-1-ylazanediyl)bis(methylene))dinicotinic Acid, 3

Methyl 6-formylnicotinate (0.250 g, 1.514 × 10–3 mol) was dissolved in 1,2-dichloroethane (10 mL) while stirring under N2. Propargyl amine (0.042 mL, 6.58 × 10–4 mol) was added followed by sodium triacetoxyborohydride (0.349 g, 1.645 × 10–3 mol) and the resulting mixture was stirred for 3 h. Water (10 mL) was added to quench the reaction and the solution was stirred for an additional 5 min. The solution was diluted with 20 mL CH2Cl2 and 20 mL 1 M NaOH. The organic layer was collected and the aqueous layer was washed with CH2Cl2 (2 × 20 mL). The combined organic layers were dried with MgSO4, filtered, and concentrated to dryness. The resulting yellow oil was dissolved in MeOH and a 1 M solution of NaOH (0.724 mL, 7.24 × 10–4 mol) was added. The reaction was heated to reflux for 2 h and then concentrated to dryness under vacuum. The crude residue was redissolved in H2O and brought to pH 2–3 with 1 M HCl. The precipitate that formed during acidification was collected by filtration and rinsed with ice-cold H2O to yield 41.3 mg of the desired product. The remaining filtrate was purified by preparatory RP-HPLC using a gradient method (prep-HPLC method 1) with the following conditions: 10 mL/min flow rate; 0–3 min 100% A, 3–9 min 75% A/25% B, 9–25 min linear gradient to 100% B, 25–32 min 100% B, 32–33 min linear gradient to 100% A, 33–38 min equilibrate in 100% A; UV at 220 nm. Product peak tR: 20.6 min. Purified portions were combined and concentrated to dryness under vacuum to yield an additional 104 mg of the desired product as a brown powder. Combined yield, 55%. Anal. Calcd for C17H15N3O4·2.25 HCl: C, 50.12; H, 4.27; N, 10.32. Found: C, 50.05; H, 3.61; N, 10.06. 1H NMR [δ (ppm), 400 MHz, CD3OD]: 9.20 (d, CPyH, 2 H), 8.50 (dd, CPyH, 2 H), 7.72 (d, CPyH, 2 H), 4.62 (s, Py-CH-N, 4 H), 4.06 (d, N-CH-C, 2 H), 3.21 (dd, CCH, 1 H). 13C NMR [δ (ppm), 100 MHz, D2O]: 164.9 (COO), 155.4 (Ar-C), 146.2 (d, 32.8 Hz, Ar-C), 143.7 (d, 28.8 Hz, Ar-C), 129.4 (Ar-C), 127.0 (t, 36.2 Hz, Ar-C), 76.8 (d, 73.2 Hz, N-CH2), 75.6 (d, 14.8 Hz, N-CH2), 55.0 (t, 25.7 Hz, CH2-C≡CH), 42.6 (t, 14.1 Hz, C≡CH). MS (+ESI): 326.1 m/z; calcd for [M+H]+: 326.1 m/z. IR (solid, cm–1): 3244, 2716, 1738, 1607, 1194, 1121. UV/vis εmax (223 nm): 27 500 M–1 cm–1.

fac-[ReI(CO)3(N,N,N-3)], 4

Ligand 3 (4.9 × 10–5 mol) was suspended n class="Chemical">in 2 mL H2O and dissolved by adding 1 M NaHCO3 until pH ∼7.5 was achieved. The pH was adjusted to 6.5–7 with 0.1 M HCl, and a 0.1 M solution of fac-[ReI(OH2)3(CO)3](SO3CF3) (0.5 mL, 5 × 10–5 mol) was added. The solution was stirred at 70 °C for 6.5 h while maintaining pH 6.5–7 by addition of 1 M NaHCO3 or 0.1 M HCl as needed. The reaction was then cooled, filtered, and purified by preparatory RP-HPLC using a gradient method (prep-HPLC method 2) with the following conditions: 10 mL/min; 0–3 min 100% A, 3–10 min 75% A/25% B, 10–32 min linear gradient to 100% B, 32–40 min 100% B, 40–45 min equilibrate in 100% A; UV at 220 nm. Product peak tR: 26.6 min. Purified portions were combined and concentrated to dryness under vacuum to yield 19 mg (66%) of the desired product as a grayish-brown powder. X-ray quality crystals were obtained by vapor diffusion of ethyl acetate into a solution of 4 dissolved in EtOH. Anal. Calcd for C20H14N3O7Re·H2O: C, 39.22; H, 2.63; N, 6.86. Found: C, 39.77; H, 2.40; N, 6.47. 1H NMR [δ (ppm), 300 MHz, CD3OD]: 9.33 (d, CPyH, 2 H), 8.48 (dd, CPyH, 2 H), 7.71 (d, CPyH, 2 H), 5.03 (d, J = 18.0 Hz, Py-CH-N, 2 H), 4.96 (d, J = 17.7 Hz, Py-CH-N, 2 H), 4.64 (d, N-CH-C, 2 H), 3.41 (dd, CCH, 1 H). 13C NMR [δ (ppm), 75 MHz, CD3OD]: 165.3 (COO), 165.1 (Ar-C), 154.1 (Ar-C), 142.2 (Ar-C), 130.7 (Ar-C), 124.7 (Ar-C), 80.8 (N-CH2), 77.3 (N-CH2), 69.1 (CH2-C≡CH), 58.9 (C≡CH). MS (+ESI): 596.1 m/z; calcd for [M]+: 596.1 m/z. IR (solid, cm–1): 2033, 1937, 1678, 1183, 1124. UV/vis εmax (213 nm): 19 600 M–1 cm–1.

Cys-NAPamide Analogue Synthesis, Functionalization, and Complexation Reactions

Ac-Cys-Nle-Asp-His-D-Phe-Arg-Trp-Gly-Lys-NH2, 5

The startn class="Chemical">ing peptide 5 (also abbreviated as Cys-NAPamide) was synthesized on an automated CS Bio CS336 peptide synthesizer (CS Bio Company, Inc., Menlo Park, CA) using standard solid-phase techniques with Fmoc-protected amino acids. Briefly, Rink Amide LS resin (220 mg, 0.2 mmol, 0.44 mmol/g loading capacity) was presuspended and swollen in N,N-dimethylformamide (DMF) for 30 min. Fmoc groups on the resin beads were deprotected with 20% piperidine in DMF. For every step, each Fmoc-protected amino acid (1 mmol) was activated in a solution containing 1 mmol HOBt and 1 M diisopropylcarbodiimide in DMF. After deprotection of the Fmoc protection group of the last amino acid residue, the N-terminal residue was acetylated by acetic anhydride with HOBt/diisopropylethylamine. The resultant peptide was cleaved and deprotected by a 3 h incubation in a mixture of TFA/triisopropylsilane/1,2-ethanedithiol/H2O (94:2:2:2). The crude peptide was filtered, and then precipitated with ice-cold anhydrous diethyl ether. The resulting peptide pellet was washed four times with ice-cold anhydrous diethyl ether, dried under a flow of N2, and dissolved in a 1 mM dithiothreitol solution. The crude peptide was purified by semipreparative RP-HPLC with solvents A (0.1% TFA in H2O) and C (0.1% TFA in MeCN) using a linear gradient from 95% A/5% C to 35% A/65% C over 45 min at a flow rate of 4 mL/min. Purified fractions were collected and lyophilized to yield the desired peptide, 5, in 83% yield and >95% purity by analytical RP-HPLC using HPLC method 3 (gradient system with solvents A and C; 1 mL/min flow rate; 0–3 min 95% A/5% C, 3–33 min linear gradient to 35% A/65% C, 33–36 min linear gradient to 15% A/85% C, hold 15% A/85% C from 36 to 39 min, return to 95% A/5% C at 39 min and equilibrate in 95% A/5% C; UV at 238 nm). MS (+ESI): 602.1 m/z, 1202.5 m/z; calcd for [M+2H]2+: 601.8 m/z; calcd for [M+H]+: 1202.6 m/z.

Ac-S-(3-Azidopropyl)-acetamido-Cys-Nle-Asp-His-d-Phe-Arg-Trp-Gly-Lys-NH2, 6

Peptide 5 (2.1 mg, 1.36 × 10–6 mol) was dn class="Chemical">issolved in 1.2 mL 0.1 M pH 7.4 PBS and mixed with 1.2 mL TCEP gel (tris-2-carboxyethyl phosphine, Pierce) in 0.1 M pH 7.4 PBS at room temperature for 1.75 h. The gel was removed by filtration, and the filtrate was immediately frozen and lyophilized overnight. N2-sparged water (1 mL), N2-sparged MeCN (0.35 mL), and 2-bromo-N-(3-azidopropyl)-acetamide (0.45 mg, 2.0 × 10–6 mol) were added to the reduced peptide, and the solution was stirred at room temperature for 12 h. The reaction mixture was subsequently concentrated by rotary evaporation and purified by RP-HPLC using HPLC method 3. MeCN was removed from the purified product fractions under vacuum, and the resulting purified solution was frozen and lyophilized overnight to yield 1.6 mg (70%) of 6. MS (+ESI): 672.0 m/z, 1343.1 m/z; calcd for [M+2H]2+: 671.8 m/z; calcd for [M+H]+: 1342.7 m/z. IR (lyophilized solid, cm–1): 3271, 3206, 2103, 1659, 1642, 1536, 1427, 1181, 1130.

Ellman’s Test for Thiol Content

Peptide 6 was analyzed by a 5,5′-dn class="Chemical">ithio-bis-(2-nitrobenzoic acid) (DTNB, Ellman’s reagent) assay according to instructions from Pierce, with slight modifications. Briefly, triplicate solutions of 6 at 3.5 × 10–4 to 4 × 10–4 M in 50 μL H2O were combined with 500 μL Ellman’s buffer (0.1 M pH 8 sodium phosphate with 1 mM EDTA) and 10 μL DTNB dissolved at 4 mg/mL in Ellman’s buffer. The solutions were vortexed and allowed to sit at room temperature for 15 min, and the UV absorbance of the solutions at 412 nm was determined. The amount of free thiol remaining in the peptide solutions was quantified by comparing the absorbance values at 412 nm to a standard curve generated by performing the assay with freshly prepared solutions of l-cysteine at known concentrations.

CuAAC Conjugate of Peptide 6 and Ligand 1, 7

Peptide 6 (300 μg, 1.78 × 10–7 mol) n class="Chemical">in 70 μL H2O, 1 (2.29 × 10–7 mol) in 10.9 μL MeOH, sodium ascorbate (6.4 × 10–7 mol), copper(II) acetate (Cu(OAc)2, 3.2 × 10–7 mol), and 100 μL of 0.1 M pH 7.4 PBS were combined, and the reaction was stirred under argon at room temperature for 2 h. A solution of 0.1 M EDTA (100 μL, 1 × 10–5 mol) was added and the solution was stirred at room temperature overnight. The reaction mixture was then acidified with 10% TFA in H2O and purified by RP-HPLC using a gradient method (HPLC method 4: 1 mL/min flow rate; 0–3 min 95% A/5% B, 3–33 min linear gradient to 35% A/65% B, 33–36 min linear gradient to 1% A/99% B, hold 1% A/99% B from 36 to 39 min, return to 95% A/5% B at 39 min and equilibrate in 95% A/5% B; UV at 238 nm). MeOH was removed from the purified product fractions under vacuum, and the resulting purified solution was frozen and lyophilized overnight to yield 200 μg (51%) of 7. MS (+ESI): 790.7 m/z, 527.6 m/z, 821.2 m/z; calcd for [M+2H]2+: 790.4 m/z; calcd for [M+3H]3+: 527.3 m/z; calcd for [M+2H+Cu]2+: 821.9 m/z. This reaction was repeated several times to obtain additional product for reactions below.

CuAAC Conjugate of Peptide 6 and Ligand 3, 8

Peptide 6 (420 μg, 2.49 × 10–7 mol) n class="Chemical">in 70 μL 0.1 M pH 7.4 PBS, 3 (3 × 10–7 mol) in 32.5 μL H2O, sodium ascorbate (9 × 10–7 mol), Cu(OAc)2 (4.5 × 10–7 mol), and 20 μL of 0.1 M pH 7.4 PBS were combined, and the reaction was stirred under argon at room temperature for 3 h. Sodium sulfide (3 mg, 1.25 × 10–5 mol) was added and the solution was stirred at room temperature for 2.75 h. The reaction mixture was then acidified with 10% TFA in H2O, centrifuged, and the supernatant was purified by RP-HPLC with HPLC method 4. Methanol was removed from the purified product fractions under vacuum, and the resulting purified solution was frozen and lyophilized overnight to yield 350 μg (61%) of 8. MS (+ESI): 834.7 m/z, 557.0 m/z, 866.2 m/z, 1667.6 m/z; calcd for [M+2H]2+: 834.4 m/z; calcd for [M+3H]3+: 556.6 m/z; calcd for [M+2H+Cu]2+: 865.8 m/z; calcd for [M+H]+: 1667.8 m/z. This reaction was repeated several times to obtain additional product for reactions below.

Click, then Chelate Route for Generating Conjugate of Peptide 7 and fac-[ReI(CO)3]+, 9

Peptide 7 (200 μg, 9.5 × 10–8 mol) n class="Chemical">in 300 μL 0.1 M pH 7.4 PBS and 1 μL 0.1 M fac-[ReI(OH2)3(CO)3](SO3CF3) (1 × 10–7 mol) were combined in a vial and stirred under argon at 60 °C for 1.75 h. Following the reaction, the solution was acidified with 0.1% TFA and purified by RP-HPLC with solvents A and B using a gradient method (HPLC method 5: 1 mL/min flow rate; 0–15 min 62% A/38% B, 15–40 min linear gradient to 100% B, 40–45 min hold 100% B, return to 62% A/38% B at 34 min and equilibrate; UV at 220 nm). Methanol was removed from the purified product fractions under vacuum, and the resulting purified solution was frozen and lyophilized overnight to yield 200 μg (9.3 × 10–8 mol, 98%) of 9. MS (+ESI): 617.7 m/z, 925.7 m/z; calcd for [M+2H]3+: 617.2 m/z; calcd for [M+H]2+: 925.4 m/z.

Chelate, then Click Route for Generating CuAAC Conjugate of Peptide 6 and Complex 2, 9′

Peptide 6 (300 μg, 1.78 × 10–7 mol) n class="Chemical">in 25 μL H2O and 2 (1.81 × 10–7 mol) in MeOH were combined, and the MeOH was removed by rotary evaporation. Sodium ascorbate (3.8 × 10–7 mol) in 3.55 μL 0.1 M pH 6 PBS and Cu(OAc)2 (1.9 × 10–7 mol) in 3.33 μL 0.1 M pH 6 PBS were added, and the reaction vial was purged with N2 for 2 min. The reaction was stirred at room temperature overnight and then purified by RP-HPLC with HPLC method 5. Methanol was removed from the purified product fractions under vacuum, and the resulting purified solution was frozen and lyophilized overnight to yield 200 μg (8.3 × 10–8 mol, 47%) of 9′. MS (+ESI): 617.7 m/z, 925.5 m/z; calcd for [M+2H]3+: 617.2 m/z; calcd for [M+H]2+: 925.4 m/z.

Click, then Chelate Route for Generating Conjugate of Peptide 8 and fac-[ReI(CO)3]+, 10

Peptide 8 (300 μg, 1.37 × 10–7 mol) n class="Chemical">in 200 μL H2O, 100 μL 0.1 M pH 7.4 PBS, and 1.5 μL 0.1 M fac-[ReI(OH2)3(CO)3](SO3CF3) (1.5 × 10–7 mol) were combined in a vial and stirred under argon at 60 °C for 2.25 h. Following the reaction, the solution was acidified with 0.1% TFA and purified by RP-HPLC with HPLC method 5. Methanol was removed from the purified product fractions under vacuum, and the resulting purified solution was frozen and lyophilized overnight to yield 200 μg (8.8 × 10–8 mol, 64%) of 10. MS (+ESI): 968.7 m/z; calcd for [M+H]2+: 968.9 m/z.

Chelate, then Click Route for Generating CuAAC Conjugate of Peptide 6 and Complex 4, 10′

Peptide 6 (200 μg, 1.19 × 10–7 mol), 4 (1.6 × 10–7 mol), n class="Chemical">sodium ascorbate (2.55 × 10–7 mol), and Cu(OAc)2 (1.28 × 10–7 mol) were dissolved in 100 μL of 0.1 M pH 6 PBS; the reaction vial was purged with N2 for 2 min and then stirred at room temperature for a total of 48 h with the addition of more sodium ascorbate (2.55 × 10–7 mol) and Cu(OAc)2 (5.8 × 10–8 M) after the first 24 h. After a total reaction time of 48 h, EDTA (2 × 10–6 mol) was added and the reaction was stirred for 1 h. The reaction was then acidified with TFA (2.5 μL), and the solution was centrifuged to pellet the resulting precipitate. The supernatant was collected and purified by RP-HPLC with HPLC method 5. Methanol was removed from the purified product fractions under vacuum, and the resulting purified solution was frozen and lyophilized overnight to yield 200 μg (8.8 × 10–8 mol, 70%) of 10′. MS (-ESI): 1936.3 m/z; calcd for [M-H]−: 1936.7 m/z.

99mTc Reactions with Small Molecules and Peptides

General Procedures for 99mTc Reactions with Small Molecule Ligands

Solutions of 1 or 3 n class="Chemical">in 10 mM pH 7.2 PBS (400–450 μL) were sealed in 5 mL labeling vials and the vials were purged with N2 for 5–10 min. Solutions of fac-[99mTcI(OH2)3(CO)3]+ (50–100 μL, 0.98–1.37 mCi) were added to bring the final ligand concentrations to 1 × 10–4 to 1 × 10–6 M in final volumes of 500 μL, and the vials were heated at 70 °C for 30–60 min. The vials were cooled prior to analysis and purification by radio-HPLC using HPLC method 1.

General Procedures for 99mTc Reactions with Peptides

Click, then chelate route: Solutions of peptides 7 or 8 (2.5 × 10–8 mol) dissolved in 400 μL 10 mM pH 7.4 PBS were sealed in 5 mL labeling vials and purged with N2 for 5–10 min. 100 uL of fac-[99mTc(OH2)3(CO)3]+ (0.76–0.81 mCi) was added to yield final peptide concentrations of 5 × 10–5 M. The reactions were heated at 70 °C for a total of 1 h with monitoring by RP-HPLC using HPLC method 1 or HPLC method 6 (solvents A and B; 1 mL/min flow rate; 0–8 min 62% A/38% B, 8–26 min linear gradient to 100% B, 26–30 min hold 100% B, return to 62% A/38% B after 30 min and equilibrate; UV at 220, 238, or 254 nm). Peptides were purified by HPLC method 1 or HPLC method 6. Chelate, then click route: Crude reaction mixtures of either 2a or 4a (200 μL, 0.34–0.60 mCi) containing ligands 1 or 3 at 5.25 × 10–5 M were added to a solution containing peptide 6 (2.6 × 10–8 mol) and sodium ascorbate (2 × 10–6 mol) in 10 mM pH 7.4 PBS (275 μL) in 5 mL labeling vials. After the vials were sealed and purged with N2 for 5 min, Cu(OAc)2 (1 × 10–6 mol) was added to give a final volume of 500 μL with peptide 6 at 5.2 × 10–5 M, ligands 1 or 3 at 2.1 × 10–5 M, sodium ascorbate at 4 × 10–3 M, and Cu(OAc)2 at 2 × 10–3 M. The vials were then heated with stirring at 50 °C for a total of 1 h with periodic monitoring by radio-HPLC using HPLC method 6. Peptides were purified by either HPLC method 1 or HPLC method 6.

Stability Analysis

Solutions of HPLC-purified 4a, 9a, 9a′, 10a, or 10a′ (200 μL, MeOH removed under a stream of N2 following isolation) were combined with solutions containing l-histidine or l-cysteine at 2 × 10–3 M in 10 mM pH 7.4 PBS (200 μL) in 5 mL labeling vials. Vials were sealed, purged with N2 for 5 min, and incubated at 37 °C for a total of 18 h (9a, 9a′, 10a, and 10a′) or 24 h (4a) with periodic analysis by radio-HPLC using HPLC methods 1 or 6. Serum stability of 4a was determined by incubating solutions of HPLC-purified 4a (200 μL) in mouse serum (200 μL, clarified by centrifugation) in 5 mL labeling vials. Vials were sealed, purged with N2 for 5 min, and incubated at 37 °C for a total of 24 h. At periodic time points, aliquots were removed, counted, and added to an equal volume of ice-cold ethanol to precipitate proteins from the solution. This mixture was centrifuged and the supernatants were counted and analyzed by radio-HPLC using HPLC method 1. Serum stability of the peptides was analyzed by incubating solutions of HPLC-purified 9a, 9a′, 10a, or 10a′ (100 μCi) in 50 μL of 10 mM pH 7.4 PBS with mouse serum (500 μL) at 37 °C for a total of 6 h. At periodic time points, the solutions were filtered through a centrifugal filter (10 K; Millipore Corp.) and the filtrate was analyzed by radio-HPLC using HPLC method 2.

Log P analysis

Equal volumes of 10 mM pH 7.4 PBS and n class="Chemical">1-octanol (500 μL each) were vortexed (30 s) in centrifuge tubes and allowed to separate (2 min). HPLC-purified solutions of 2a, 4a, 9a′, or 10a′ (0.1 μCi in 2–15 μL, MeOH removed under a stream of N2 following isolation) were added to triplicate sets of tubes after an equal volume of the aqueous phase was removed, and the tubes were vortexed (1 min) and centrifuged (2000 × g, 5 min). Aliquots (50 μL) from each layer were removed and counted to determine the ratio of radioactivity present in the octanol layer compared to the aqueous layer.

In Vitro and In Vivo Biological Evaluations

Cell Culture and Animal Model

Cell growth and tumor implantation procedures were performed as previously described.[41] Briefly, B16F10 murine melanoma cells were cultured in Dulbecco’s modified Eagle’s high-glucose medium (GIBCO, Carlsbad, CA) and supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin in a humidified incubator containing 5% CO2 at 37 °C. A 70–80% confluent monolayer was detached with 0.25% trypsin–EDTA and dissociated into a single-cell suspension for further cell culture and assays. All animal studies were carried out in compliance with federal and local institutional rules for the conduct of animal experimentation. Approximately 1 × 106 cultured B16F10 cells were suspended in 100 μL of PBS and subcutaneously implanted in the right shoulders of C57BL/6 mice. Tumors were grown to a size of 0.5–1 cm in diameter (1–2 weeks) prior to biodistribution studies.

In Vitro IC50 Cell Binding Analysis

The MC1R bn class="Chemical">inding affinity of 5, 6, 9′, and 10′ in B16F10 cells was performed as previously described.[41] Briefly, B16F10 cells (5 × 105) suspended in Dulbecco’s Modified Eagle’s Medium (DMEM) containing 25 mM N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid), 0.2% bovine serum albumin, and 0.3 mM 1,10-phenanthroline were seeded at a density of 0.3 million per well in 96-well plates and allowed to attach overnight. The cells were then incubated at 37 °C for 2 h with 125I-(Tyr2)-NDP (20 000 counts per minute) and 5, 6, 9′, or 10′ (peptide concentration varying from 10–11 to 10–5 M). Cells were washed three times with ice-cold 10 mM pH 7.4 PBS and lysed in 1 mL of 1.0 M NaOH containing 0.1% sodium dodecyl sulfate. The radioactivity of the lysed cells was measured by a γ-counter (Perkin-Elmer model 1470). The experiment was performed in quadruplicate wells. The IC50 values (the concentration of competitor required to inhibit 50% of the radioligand binding) of the peptides were calculated by using Origin 8.0 (MicroCal).

In Vitro B16F10 Melanoma Cell Uptake Assays

Cell uptake studies of 9a′ and 10a′ wen class="Chemical">re performed as previously described.[41] Briefly, B16F10 cells (3 × 105) were seeded in 12-well tissue culture plates and incubated at 37 °C overnight. The cells were washed with 10 mM pH 7.4 PBS and then incubated with 9a′ or 10a′ (1 μCi per well, in DMEM) with or without NDP (3 μM/well) at 37 °C for 0.5, 1, and 2 h. The cells were then washed 3 times with 10 mM pH 7.4 PBS and lysed in 1 mL of 1.0 M NaOH containing 0.1% sodium dodecyl sulfate and transferred to γ-counter tubes. Radioactivity was measured by a γ-counter (Perkin-Elmer model 1470). Cell uptake was expressed as the percentage of added radioactivity. Experiments were performed twice with quadruplicate wells.

In Vivo B16F10 Melanoma Xenograft Mouse Biodistribution Analysis

For biodistribution studies, female C57BL/6 mice bearing B16F10 xenografts (n = 4 per group) were injected via the tail vein with approximately 100 μCi of 9a′ or 10a′ and were euthanized at 1 and 2 h post injection (p.i.). Tumor and normal tissues of interest were removed and weighed, and their radioactivity was measured using a γ-counter. Radioactivity uptake was expressed as a percentage of the injected radioactive dose per gram of tissue (% ID/g).

Statistical Methods

Statistn class="Chemical">ical analysis was performed using the Student’s t test for unpaired data. A 95% confidence level was chosen to determine the significance between groups, with P < 0.05 being designated as significantly different.