A Comparison of Evans Blue and 4-(p-Iodophenyl)butyryl Albumin Binding Moieties on an Integrin αvβ6 Binding Peptide

Serum albumin binding moieties (ABMs) such as the Evans blue (EB) dye fragment and the 4-(p-iodophenyl)butyryl (IP) have been used to improve the pharmacokinetic profile of many radiopharmaceuticals. The goal of this work was to directly compare these two ABMs when conjugated to an integrin αvβ6 binding peptide (αvβ6-BP); a peptide that is currently being used for positron emission tomography (PET) imaging in patients with metastatic cancer. The ABM-modified αvβ6-BP peptides were synthesized with a 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetracetic acid (DOTA) chelator for radiolabeling with copper-64 to yield [64Cu]Cu DOTA-EB-αvβ6-BP ([64Cu]1) and [64Cu]Cu DOTA-IP-αvβ6-BP ([64Cu]2). Both peptides were evaluated in vitro for serum albumin binding, serum stability, and cell binding and internalization in the paired engineered melanoma cells DX3puroβ6 (αvβ6 +) and DX3puro (αvβ6 −), and pancreatic BxPC-3 (αvβ6 +) cells and in vivo in a BxPC-3 xenograft mouse model. Serum albumin binding for [64Cu]1 and [64Cu]2 was 53–63% and 42–44%, respectively, with good human serum stability (24 h: [64Cu]1 76%, [64Cu]2 90%). Selective αvβ6 cell binding was observed for both [64Cu]1 and [64Cu]2 (αvβ6 (+) cells: 30.3–55.8% and 48.5–60.2%, respectively, vs. αvβ6 (−) cells <3.1% for both). In vivo BxPC-3 tumor uptake for both peptides at 4 h was 5.29 ± 0.59 and 7.60 ± 0.43% ID/g ([64Cu]1 and [64Cu]2, respectively), and remained at 3.32 ± 0.46 and 4.91 ± 1.19% ID/g, respectively, at 72 h, representing a >3-fold improvement over the non-ABM parent peptide and thereby providing improved PET images. Comparing [64Cu]1 and [64Cu]2, the IP-ABM-αvβ6-BP [64Cu]2 displayed higher serum stability, higher tumor accumulation, and lower kidney and liver accumulation, resulting in better tumor-to-organ ratios for high contrast visualization of the αvβ6 (+) tumor by PET imaging.


Introduction
The use of biologically active molecules such as peptides and antibodies continues to increase for both diagnosis and therapy [1][2][3]. Peptides are attractive platforms for diagnostics due to their ability to achieve high target binding affinity and in part due to their small size which results in short biological half-life and rapid clearance from nontarget tissues, producing good target-to-non-target contrast, low toxicity, and generally low or absent immunogenicity [1]. Synthetic advantages of peptides include simple preparation and easy, flexible functionalization or chemical modification to further improve affinity, stability, selectivity, and overall pharmacokinetic properties [1,4]. However, some of the properties that are desirable for a diagnostic agent can hamper the translation to a therapeutic, which relies on a prolonged circulation for high and persistent uptake in the targeted tissue. Too rapid clearance can render the therapeutic ineffective, and poor clearance from non-target tissue can lead to off-target toxicity. Thus, peptides typically require fine-tuning for therapeutic applications to balance circulation time and provide high target accumulation with sufficient clearance from non-target tissues [5][6][7].
Chemical modifications of peptides offer a route to improving these pharmacokinetic properties; this includes incorporation of polyethylene glycol (PEG; PEGylation), glycosylation, or the formation of protein conjugates (e.g., with serum albumin) [4,[8][9][10][11]. PEGylation is a convenient approach as PEGs are commercially available in a variety of molecular sizes, including mono-disperse PEGs with various functional groups for synthetic orthogonality [1,9]. PEGylation increases hydrophilicity (reducing kidney, lung, and liver accumulation) [12,13], provides increased stability (by protection from proteases), and reduces immunogenicity (by masking the peptide) [9,13]. The size and placement of the PEG on the peptides can significantly affect the pharmacokinetics and tumor accumulation [12][13][14]. Stability, circulation time, and tumor uptake of peptides can also be increased by chemical ligation ex vivo to serum albumin (taking advantage of albumin's size, long circulation time, and renal recycling) [8,[15][16][17]. Alternatively, the same benefit can be achieved by direct attachment of a small albumin binding moiety (ABM) onto the peptide without substantially increasing the size. The ABM binds reversibly to albumin in the blood, thereby increasing circulation time and facilitating renal recycling, which, in turn, increases target tissue accumulation [8,15,16,18]. Several ABMs have been employed to modify pharmaceuticals currently used in the clinic, with some being used on their own, primarily for measuring plasma volume [16]; among the first ABMs used to modify pharmaceuticals were long-chain fatty acids, such as myristic and palmitic acid [5], and later other lipophilic molecules including benoxaprofen, phenytoin, ibuprofen, and naproxen [16].
More recently, two ABMs in particular, a fragment of Evans blue (EB) dye and the 4-(p-iodophenyl)butytryl (IP) group, have also been used to modify the pharmacokinetic profile of radiopharmaceuticals, in particular small molecules (folic acid and prostate specific membrane antigen (PSMA) agents) and peptides (octreotide, exendin-4, and cRGDfK) [10,11,16,[18][19][20]. The EB-ABM was derived from Evans blue dye, a dye which has been used clinically for over 90 years to measure plasma volume and determine blood-brain barrier integrity [16,17,21]. The EB-ABM fragment was first used in 2004 as an MRI contrast agent for imaging blood vessels [22] and has since been used for a variety of applications, including determining blood volume, vascular permeability, and as a conjugate to enhance receptor targeting agents (small molecules and peptides) for both cancer imaging and therapy [9,17,[23][24][25][26]. The IP-ABM has also been studied extensively to enhance radiopharmaceuticals (small molecules and peptides), where the group at the para-position of the aromatic ring of the IP-ABM can be tuned to adjust serum albumin affinity [15,27,28], and a neighboring aspartate residue (D) has been shown to provide a more sustained tumor retention [29]. Numerous preclinical studies have evaluated both ABMs and noted prolonged blood circulation, with an increase in tumor uptake that can also lead to a reduction of kidney accumulation [7,11,18].
In the present study, we describe a head-to-head comparison of the α v β 6 -BP modified with either EB-ABM or IP-ABM, with the goal to examine if fine tuning of the ABM could further increase tumor accumulation.

Synthesis of EB-ABM 8
The synthesis of the Evans blue fragment (EB-ABM 8) was based on previously described methods [7,17,45] (Scheme 1). In brief, o-tolidine 5 (531 mg, 2.5 mmol; TCI America, Inc., OR, USA) was dissolved in anhydrous pyridine (1 mL) followed by the addition of succinic anhydride (300 mg, 3.0 mmol) in DMF (1 mL) and allowed to react overnight at room temperature. The crude reaction mixture was concentrated under vacuum and purified by silica-gel column chromatography using a four solvent gradient system beginning with EtOAc/n-hexanes (1/1, v/v) to remove unreacted o-tolidine (5, yellow band). The solvent was then changed to 100% EtOAc before switching to MeOH/DCM (1/9, v/v) and gradually ramping to 3/7 (v/v) to obtain 6 (648 mg, rt = 0.13, 1:1 hexanes:EtOAc) as a white solid in 83% yield. Compound 6 was analyzed by analytical RP-HPLC and ESI mass spectrometry ( Figure S17). Compound 6 (300 mg, 0.96 mmol) was added to a 25 mL round bottom flask with stir bar containing MeOH (7 mL) and water (5 mL). The contents were cooled to 0 • C (ice/brine solution) and allowed to stir for 15 min prior to addition of concentrated hydrochloric acid 240 µL (HCl, 12.1 N; EMD). The diazonium formation of 7 was most successful when the addition of sodium nitrite was done in two portions; the first portion of sodium nitrite (NaNO 2 , 70 mg, 1.01 mmol; Sigma-Aldrich) was allowed to react for 5 min before the addition of the second portion (NaNO 2 , 70 mg, 1.01 mmol), after which the reaction was stirred an additional 30 min to generate 7 in situ, which was produced in better yields using the methanol co-solvent than water alone [46]. During in situ formation of 7, sodium bicarbonate (350 mg, 4.17 mmol; EMD) was dissolved in water (4 mL) with 1-amino-8napthol-2,4-disulfonic acid (377 mg, 1.18 mmol; TCI America, Inc.) in a separate 25 mL round bottom flask and the contents cooled in an ice/brine solution (~20 min). Next, the diazonium 7 reaction mixture (yellow) was cannulated into the 1-amino-8-napthol-2,4-disulfonic acid (brown-purple) solution by drop-wise addition over 20 min while maintaining both solutions at 0 • C. Upon complete addition of 7, the reaction contents were allowed to stir for 3 h at 0 • C, and the crude reaction mixture was lyophilized and purified by semi-preparative RP-HPLC, and the collected material lyophilized. The EB-ABM 8 was afforded as a fluffy purple solid (480 mg, 78%) and was analyzed by analytical RP-HPLC, ESI mass spectrometry, and NMR (Figures S18 and S19).

Integrin α v β 6 Affinity ELISA
Affinity for the integrin α v β 6 was determined by competitive binding ELISA of [ Nat Cu]1 and [ Nat Cu]2 against biotinylated-LAP (G&P Biosciences, Santa Clara, CA, USA) as previously described to determine the half-maximum inhibitory concentration (IC 50 ) [44]. Briefly, in a 96 well Nunc Immuno maxisorp plate, capturing anti-α v antibody (P2W7, 5 µg/mL, Abcam, MA, USA) was plated (50 µL/well) at 37 • C for 1 h, washed with PBS (3×), and blocked overnight with blocking buffer (300 µL/well, 0.5% non-fat dry milk powder (w/v), 1% Tween 20, in PBS). It was then washed with wash buffer that consisted of 2 mmol/L of Tris buffer (pH = 7.6), 150 mmol/L sodium chloride, 1 mmol/L manganese chloride, and 0.1% Tween 20 (v/v) in deionized water (3×). Purified integrin α v β 6 (R&D Systems, Minneapolis, MN, USA) in conjugate buffer (50 µL/well, 20 mM Tris, 1 mM MnCl 2 , 150 mM NaCl, 0.1% Tween, 0.1% milk powder in water) was then added to each well, incubated at 37 • C for 1 h, followed by washing using wash buffer 3×). Serial dilutions of each peptide stock of 2 mmol/L in 10% DMSO (v/v) into PBS and biotinylated natural ligand LAP were premixed in equal volumes and placed onto the plate in triplicate for each peptide concentration (50 µL/well) and allowed to incubate at 37 • C for 1 h then washed with wash buffer (3×). A 1:1000 dilution of ExtrAvidin Horseradish Peroxidase (HRP; Fisher, NH, USA) was added to each well (50 µL/well), incubated at 37 • C for 1 h, and then washed with wash buffer (3×). The ExtrAvidin HRP was detected with TMB One solution (50 µL/well; Promega Corp., Madison, WI, USA) for 10-15 min at room temperature. The reaction was stopped by adding 1N sulfuric acid (H 2 SO 4 , 50 µL/well; EMD, MA, USA) and the absorbance was measured in a Multiscan Ascent plate reader (Thermo Fisher, Waltham, MA, USA) at 450 nm. Half-maximal inhibitory concentration (IC 50 ) of peptides was determined by fitting to sigmoidal dose-response model in GraphPad Prism 8.0 (GraphPad, CA, USA). For the positive control no peptide was added and for the negative controls either no biotinylated-LAP or no integrin α v β 6 was added.

Cell Binding and Internalization Assay
Binding of [ 64 Cu]1-4 and internalization to DX3puro, DX3puroβ6, and BxPC-3 cells were determined as previously described [44]. Prior to the experiment, the cells were analyzed by flow cytometry to confirm levels of integrin α v β 6 expression. Non-fat dry milk powder (0.5% w/v in PBS) was used to pretreat the assay tubes to prevent non-specific binding. Aliquots of [ 64 Cu]1-4 (≤1 µL, 7.4-18.5 KBq) in 50 µL serum free medium (pH 7.2) were added to a cell suspension (3.75 × 10 6 cells in 50 µL serum free medium) and incubated for 1 h at room temperature in closed microfuge tubes (n = 3/cell line/compound) and gently agitated every 3 min to ensure mixing. The cells were pelleted by centrifugation at 200 (RCF) for 3 min and the supernatant collected. The cell pellet was washed with 0.5 mL serum free medium and the wash medium combined with the original supernatant. The cells were resuspended in 0.6 mL serum free medium for γ-counting. The fraction of bound radioactivity was determined with a γ-counter (by measuring cell pellet and combined supernatants). To determine the fraction of internalized radioactivity, the cells were repelleted, and re-suspended in acidic wash buffer (0.2 mol/L sodium acetate, 0.5 mol/L sodium chloride, pH 2.5, 300 µL, 4 • C, 5 min) to release surface-bound activity, followed by a wash with PBS (300 µL). The internalized fraction was determined with a γ-counter (cell pellet vs. radioactivity released into supernatant).

Human and Mouse Serum Binding Assay and Stability Assay
Serum protein binding of [ 64 Cu]1 and [ 64 Cu]2 was assessed following the previously reported method [42]. Peptides [ 64 Cu]1 and [ 64 Cu]2 were evaluated by ultrafiltration using Centrifree Ultrafiltration devices (EMD) according to the manufacturer's recommendations. Experiments were carried out in triplicate. The Centrifree Ultrafiltration devices were pretreated with PBS containing Tween 20 (5% v/v), followed by triplicate rinses with PBS. An aliquot of each peptide [ 64 Cu]1 or [ 64 Cu]2 in PBS (≤25 µL, 20-60 KBq) was thoroughly mixed with 0.5 mL of serum at 37 • C in a microfuge tube. The mixture was incubated at 37 • C for 5 min, and an aliquot (50 µL) was transferred to a tube for γ-counting. The remaining sample was transferred to a Centrifree Ultrafiltration device and centrifuged for 40 min at 1500 (RCF) at ambient temperature (20-24 • C). An aliquot (50 µL) of the filtrate was transferred to a tube for γ-counting. For each radiolabeled peptide, a blank was run using 0.5 mL PBS/Tween 20 (5% v/v) instead of serum (n = 3) to determine nonspecific binding. Following γ-counting, the protein-bound radioactivity was calculated by subtracting the counts measured in the filtrate aliquot (i.e., not protein-bound) from the counts in the corresponding serum aliquot. The data are expressed as mean ± standard deviation of fraction of radioactivity bound to protein after subtraction of non-specific binding determined in the blank.

Biodistribution
All animal procedures conformed to the Animal Welfare Act and were approved by the University of California, Davis Institutional Animal Care and Use Committee. Female athymic nu/nu-nude mice (6-8 weeks old) were purchased from Charles River Laboratories (Wilmington, MA, USA) and provided food and water on an ad libitum basis. BxPC-3 xenografts were implanted according to previous methods [42,44]. Briefly, BxPC-

Statistical Analysis
Quantitative data are reported as mean ± standard deviation (SD). Statistical significance was determined by a paired two-tailed Student's t-test from the two independent sample means to give a significance value (p-value) at 95% confidence interval (CI). A p-value of <0.05 was considered statistically significant.

PET Imaging
Overall, the BxPC-3 tumors were clearly visualized by PET imaging with b tides at all time points ( Figure 6); the PET imaging also showed that [ 64 Cu]2 prov clearest images based on its superior tumor-to-background ratios. Most notably, ously discussed for the biodistribution data, the PET images for [ 64 Cu]1 had muc kidney accumulation and higher levels of radiation in the liver, indicative of po vivo instability of [ 64 Cu]1, which had shown substantially higher degradation i serum compared to [ 64 Cu]2.

PET Imaging
Overall, the BxPC-3 tumors were clearly visualized by PET imaging with both tides at all time points ( Figure 6); the PET imaging also showed that [ 64 Cu]2 provide clearest images based on its superior tumor-to-background ratios. Most notably, as p ously discussed for the biodistribution data, the PET images for [ 64 Cu]1 had much h kidney accumulation and higher levels of radiation in the liver, indicative of possib vivo instability of [ 64 Cu]1, which had shown substantially higher degradation in m serum compared to [ 64 Cu]2. ).

PET Imaging
Overall, the BxPC-3 tumors were clearly visualized by PET imaging with both peptides at all time points ( Figure 6); the PET imaging also showed that [ 64 Cu]2 provided the clearest images based on its superior tumor-to-background ratios. Most notably, as previously discussed for the biodistribution data, the PET images for [ 64 Cu]1 had much higher kidney accumulation and higher levels of radiation in the liver, indicative of possible in vivo instability of [ 64 Cu]1, which had shown substantially higher degradation in mouse serum compared to [ 64 Cu]2.

Discussion
Cancer remains a leading cause of death globally [48,49]. Many cancers exhibit high expression of the cell surface receptor integrin α v β 6 , and expression levels correlate with poor prognosis and reduced progression-free and overall survival [31,32,38]. Therefore, integrin α v β 6 has been identified as an important target both for imaging and treatment [50,51]. Receptor targeted delivery of radiopharmaceuticals is an important part of new approaches for improved cancer detection and therapy [48]. Peptides are attractive radiopharmaceuticals for both detection and treatment, because they are readily synthesized and can be chemically modified to optimize pharmacokinetics and metabolic stability. The addition of albumin binding moieties (ABMs) to numerous radiopharmaceuticals has demonstrated increased circulation time, reduced kidney uptake, and substantially increased tumor accumulation [18,52,53]. However, differences in the chemical structures of the ABM have been found at times to significantly affect the biodistribution, which ultimately determines target uptake, therapeutic efficacy, and off-target toxicity [52,[54][55][56]. Thus, evaluation of different ABMs is important for optimal radiopharmaceutical performance towards the development of an α v β 6 -targeted radiotherapeutic agent. Our laboratory continues to develop integrin α v β 6 -targeting radiopharmaceuticals, including optimization of the core peptide structure [30] via PEGylation [14], and most recently the addition of an 4-(piodophenyl)butyryl (IP) ABM, which has demonstrated improved accumulation in tumors for both the [ 18 F]AlF NOTA and [ 64 Cu]Cu DOTA radiolabeled IP-ABM-α v β 6 -BP compared to the parent non-ABM α v β 6 -BP [42,44]. To further evaluate the choice of preferred ABM for α v β 6 -BP, the comparison of the IP-ABM with another prominent ABM, the Evans blue fragment (EB-ABM), was explored. The synthesis of both α v β 6 -BP peptides containing different ABMs, [ 64 Cu]1 or [ 64 Cu]2 (Scheme 1), was done efficiently using a solid-phase approach, which allowed installation of the respective ABM-peptide from the same batch of peptidyl-resin by first coupling an orthogonally protected lysine allowing for the attachment of the DOTA-chelator at the N-terminus and either the EB-ABM 8 or the IP-ABM at the sidechain. The IP-ABM included an aspartate (D) residue as it is reported to result in better tumor retention [28]. After removal from the resin and purification, both DOTA-ABM-α v β 6 -BP peptides (1 and 2) were efficiently radiolabeled with copper-64 to yield [ 64 Cu]1 and [ 64 Cu]2 in high radiochemical purity >97%.
The ABM containing peptides [ 64 Cu]1 and [ 64 Cu]2 both demonstrated high tumor uptake at 4 h p.i., over 5% and 7.5% ID/g, respectively; representing a greater than 3to-4.5-fold increase, respectively, from the non-ABM bearing [ 64 Cu]Cu DOTA-α v β 6 -BP (1.61 ± 0.70% ID/g) [44]. The improvement in tumor accumulation was greater for the IP-ABM peptide [ 64 Cu]2 than for the EB-ABM peptide [ 64 Cu]1, and was in concordance with the cell binding to both DX3puroβ6 and BxPC-3 cells (Figure 2). Furthermore, the prolonged tumor uptake and retention ( Figure 4A) were maintained for 72 h, and, in conjunction with rapid renal clearance, provided a high tumor-to-background ratio ( Figure 5) and high contrast PET-images ( Figure 6). Since the only difference between [ 64 Cu]1 and [ 64 Cu]2 is the ABM, and [ 64 Cu]2 showed significantly higher stability in serum compared to [ 64 Cu]1 (Figure 3), the observed differences in the tumor-to-background ratios could be attributed to the improved stability. This study adds to the growing number of literature reports describing improved tumor uptake following the incorporation of ABMs [4,7,11]. For example, the small molecule PSMA-617, a radiopharmaceutical targeting the prostate specific membrane antigen (PSMA), exhibited approximately a fivefold increase in tumor accumulation with the addition of an EB-ABM at 4 h and a twofold increase for the IP-ABM modified PSMA-617, compared to the unmodified (non-ABM bearing) PSMA-617; furthermore, the EB-ABM PSMA-617 maintained tumor accumulation over time (65.6-77.3% ID/g from 4 h to 48 h) [55]. In another study with PSMA-617, the addition of the IP-ABM also resulted in twofold higher accumulation in tumor tissue as compared to the non-ABM containing PSMA-617 agent (non-ABM PSMA-617: 38% ID/g vs. IP-ABM PSMA-617: 75.7% ID/g at 24 h) [28,57]. Other small molecule PSMA agents modified with ABMs have also shown improvements in tumor accumulation, with the EB-ABM MCG PSMA agent having around a fourfold increase in tumor accumulation (MCG non-ABM: 10.9% ID/g vs. MCG-ABM: 40.4% ID/g at 24 h) [53] and an IP-ABM PSMA agent CTT1403 exhibiting >18-fold improvement in tumor accumulation (CTT1401 non-ABM: 2.2% ID/g vs. CTT1403-ABM: 40% ID/g at 24 h [54]. The addition of ABMs to other small molecule radiopharmaceuticals has also been shown to improve tumor accumulation with the small molecule radioligand folic acid modified with the IP-ABM having a threefold increase in tumor accumulation (ABM: 19.5% ID/g vs. non-ABM: 7% ID/g at 24 h, p.i.) with a considerably lower kidney accumulation (ABM: 28% ID/g vs. non-ABM: 70% ID/g at 4 h) [52,58].
Aside from small molecule radiopharmaceuticals, substantial benefits from the addition of ABMs to peptide radiopharmaceuticals have been shown; for example, the large peptide exendin-4 (39 amino acids), which targets the glucagon-like peptide 1 (GLP-1) receptor, when modified with the IP-ABM, demonstrated an improved stability and a twofold increase in tumor accumulation at 4 h, along with reduced kidney retention by more than half [7,59]. The small five amino acid integrin α v β 3 targeting cyclic peptide (cRGDfK) modified with EB-ABM and radiolabeled as [ 64 Cu]Cu NOTA-EB-cRGDfK displayed a >16-fold improvement (vs. [ 64 Cu]Cu NOTA-cRGDfK) in tumor accumulation in a U87MG glioblastoma tumor model (with ABM: 16.6% ID/g vs. non-ABM: <1.1% ID/g), but only had about a fivefold improvement in MDA-MB-435 melanoma and HT29 colorectal adenocarcinoma models [18]. The somatostatin receptor targeting peptide octreotide (TATE), which is eight amino acids in size, has seen some of the greatest improvements in tumor accumulation upon modification with an ABM. For example, the EB-ABM modified [ 177 Lu]Lu DOTA-EB-TATE provided a greater than eightfold increase in the tumor accumulation at 24 h (with ABM: 78.8% ID/g vs. non-ABM: 9.3% ID/g, respectively) [60] and the [ 86 Y]Y DOTA-EB-TATE showed a larger enhancement with a between 30-and 60-fold increase in tumor accumulation compared to [ 86 Y]Y DOTA-TATE, depending on the tumor model [6]. These studies paved the way for clinical trials where [ 177 Lu]Lu DOTA-EB-TATE showed an extended circulation which led to a 7.9-fold increase in tumor dose delivery [61]. Overall, these studies illustrate the potential benefits of including an ABM on targeted peptide receptor radionuclide therapy (PRRT).
The addition of either EB-ABM or the IP-ABM on the α v β 6 -BP did significantly increase tumor accumulation (three-to-fivefold from the non-ABM-α v β 6 -BP) and the overall clearance properties of the ABM-modified α v β 6 -BP peptides [ 64 Cu]1 and [ 64 Cu]2 were similar with predominantly renal excretion. The organ with the highest accumulation was the kidneys, with the initial kidney uptake of the EB-ABM peptide [ 64 Cu]1 having more than double that of the IP-ABM peptide [ 64 Cu]2 (4 h: 75.5 ± 7.3% ID/g vs. 33.6 ± 5.4% ID/g, p = 0.0013), with both dropping to approximately one third of their initial value at 72 h p.i. (20.0 ± 6.9% ID/g and 11.4 ± 1.0% ID/g, respectively, p = 0.10, Figure 4). The introduction of the IP-ABM to the α v β 6 -BP significantly reduced kidney accumulation, which we hypothesize is due to the higher stability of the IP-ABM [ 64 Cu]2 over the EB-ABM [ 64 Cu]1. These data are promising and indicate that renal toxicity would be less of a concern for PRRT of α v β 6 -BP agents using the IP-ABM. The observed effects of the different ABMs on kidney uptake and retention are comparable to other radiopharmaceutical ABMadducts, for example, the ABM modified peptide [ 177 Lu]Lu DOTA-TATE showed that the IP-ABM-analogue also provided lower kidney accumulation that was more rapidly cleared (dropping from~20% ID/g at 4 h to~5% ID/g at 72 h) compared to the EB-ABM-analogue (~30% ID/g at 4 h to~15% ID/g at 72 h) [29,60]. This similar kidney accumulation and retention trend was also observed with the small molecule PSMA-617 agent, where the EB-PSMA-617 had considerably higher kidney accumulation and retention compared to the IP-PSMA-617, which had rapid kidney clearance (EB-PSMA-617: >20% ID/g at 4 h, which remained at 48 h vs. IP-ABM-PSMA-617:~10% ID/g at 4 h dropping to <5% ID/g at 48 h) [55]. Both [ 64 Cu]1 and [ 64 Cu]2 also displayed some secondary clearance through the gastrointestinal (GI) tract and excretion of radioactivity in the feces (Figures S20 and S21). The IP-ABM modified peptide [ 64 Cu]2 had higher GI accumulation, with the highest uptake in the stomach of 18.1 ± 2.9% ID/g at 4 h, though, gratifyingly, both peptide's GI accumulation dropped down to less than one-fifth of their respective original value (≤3.2% ID/g at 72 h, Figure 4).
The non-α v β 6 -targeting ABM controls [ 64 Cu]3 and [ 64 Cu]4 were used to evaluate non-specific uptake and demonstrate that the enhanced tumor accumulation of [ 64 Cu]1 or [ 64 Cu]2 resulted from integrin α v β 6 receptor mediated uptake, as opposed to the enhanced permeability and retention (EPR) effect. As expected, [ 64 Cu]3 and [ 64 Cu]4 largely remained in the blood, thus mostly acting as blood pool imaging agents with high blood accumulation of 39.0% ID/g and 9.5% ID/g, respectively, at 4 h ( Figure S25) and mirrored other similar non-targeted ABMs, such as the EB-ABM compound [ 64 Cu]Cu NOTA-EB (NEB,~15% ID/g at 4 h, dropping to~10% ID/g at 1 d) [16,23] Figure 5A) but was lower in the kidneys (though the EB compound was still higher than the IP compound with 18.6% ID/g and 4.3% ID/g, respectively, at 4 h; Figure 5A), highlighting the effect of both the properties of the ABM as well as the targeting peptide moiety on kidney uptake.  Figure 5A). However, the non-targeted [ 64 Cu]3 and [ 64 Cu]4 showed minimal binding (<4.3%) in cell binding studies to both the α v β 6 -expressing and α v β 6 -null cells ( Figure S24), thus their higher tumor accumulation compared to [ 64 Cu]1 and [ 64 Cu]2 was attributed to the EPR effect (which, together with the long circulation, resulted in the expectedly low tumor-to-blood ratios of <0.9/1 ( Figure 5B, Figure S26). By comparison, [ 64 Cu]1 and [ 64 Cu]2 showed high and α v β 6 -dependent cell binding (>30-60% binding;~20:1 for DX3puroβ6 (+)/DX3puro (−) cells), and in vivo tumor uptake was efficiently blocked by the pre-administration of metal free 1 and 2, respectively, supporting integrin α v β 6 -dependent tumor accumulation ( Figure S23). Taken together, the tumor uptake observed for the integrin α v β 6 -binding peptides [ 64 Cu]1 and [ 64 Cu]2 was attributed to specific targeting of the integrin α v β 6 receptor. Both ABM modified α v β 6 -BP peptides had improved pharmacokinetic profiles from the parent peptide and overall [ 64 Cu]2 demonstrated a more favorable biodistribution. Tumor retention of [ 64 Cu]1 and [ 64 Cu]2 was good over the three day study period, with each retaining about two-thirds of the original (4 h) uptake at 72 h p.i. The PET image quality improved, most notably for [ 64 Cu]2 over time after the initial uptake period (i.e., after 24 h p.i.) as a result of faster washout from non-target tissues ( Figure 6). The high absolute tumor uptake of [ 64 Cu]2, its efficient binding and internalization to α v β 6 -expressing cells (Figure 2), and its better serum stability (Figure 3) demonstrate the potential of using the [ 64 Cu]2 as an integrin α v β 6 -targeted peptide receptor radionuclide therapy (PRRT) agent where the copper-64 is replaced by a therapeutic radioisotope such as lutetium-177.

Conclusions
The effect of Evans blue (EB) and 4-(p-iodophenyl)butyryl (IP)-based albumin binding moieties (ABMs) on the pharmacokinetics of α v β 6 -BP, a peptide targeting the cancerassociated cell surface receptor integrin α v β 6 was investigated. The albumin binding moieties on α v β 6 -BP did not interfere with integrin α v β 6 affinity or selectivity in vitro. In vivo in a BxPC-3 pancreatic tumor xenograft mouse model, the IP-ABM-modified α v β 6 -BP [ 64 Cu]2 had a considerably more favorable pharmacokinetic profile compared to the EB-ABM-modified α v β 6 -BP [ 64 Cu]1, with higher tumor uptake, reduced kidney and liver uptake, and improved tumor-to-background ratios that led to a clearer tumor visualization by PET imaging. Furthermore, the IP-ABM-modified α v β 6 -BP [ 64 Cu]2 had superior serum stability, making it a lead candidate for future integrin α v β 6 -targeted imaging and therapy studies.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/ 10.3390/pharmaceutics14040745/s1, S1-S26. Table S1-S3: Table of Contents, Table S3: RP-HPLC methods, Figure S4: Schematic for solid phase reaction of EB-ABM 8 to peptidyl resin of DOTA-K(NH 2 -α v β 6 -BP to produce DOTA-EB-α v β 6 -BP 1 after cleavage and pictorial of the reaction of 8 with peptidyl resin of DOTA-K(NH 2 )-α v β 6 -BP, Figure S5: RP-HPLC and MALDI-TOF of DOTA-EB-α v β 6 -BP 1, Figure S6: Radio-RP-HPLC of [ 64 Cu]1 and co-injection radio-RP-HPLC of [ Nat Cu]1 and [ 64 Cu]1, Figure S7: MALDI-TOF of [ Nat Cu]1, Figure S8: RP-HPLC and MALDI-TOF of DOTA-IP-α v β 6 -BP 2, Figure S9  Institutional Review Board Statement: Radioactive work was conducted under radioactive use authorization 9098 managed by University of California, Davis Radiation Safety Services. All animal and biological research were conducted under biological use authorization R1580 and all animal work was conducted in accordance with procedures pre-approved by the Institutional of Animal Care and Use Committee (IACUC) at the University of California, Davis which is regulated by several independent resources. Accreditation and oversight has been approved since 1966 by AAALAC #000029 and by the Office of Laboratory Animal Welfare (OLAW) #D16-00272 (A3433-01).

Informed Consent Statement: Not applicable.
Data Availability Statement: Additional data supporting the reported results can be found in the Supplementary Materials (S1-S26).