Enhanced tumor retention of NTSR1-targeted agents by employing a hydrophilic cysteine cathepsin inhibitor

Abstract

We explored the approach of using an analog of E-64, a well-known and hydrophilic cysteine cathepsin (CC) inhibitor, as a potent cysteine cathepsin-trapping agent (CCTA) to improve the tumor retention of low-molecular-weight, receptor-targeted radiopharmaceuticals. The synthesized hydrophilic CCTA-incorporated, NTSR1-targeted agents demonstrated a substantial increase in cellular retention upon uptake into the NTRS1-positive HT-29 human colon cancer cell line. Similarly, biodistribution studies using HT-29 xenograft mice revealed a significant and substantial increase in tumor retention for the CCTA-incorporated, NTSR1-targeted agent. The intracellular trapping mechanism of the CCTA-incorporated agents by macromolecular adduct formation was confirmed using multiple in vitro and in vivo techniques. Furthermore, utilization of the more hydrophilic CCTA greatly increased the hydrophilicity of the resulting NTSR1-targeted constructs leading to substantial decreases in most non-target tissues in contrast to our previously reported dipeptidyl acyloxymethyl ketone (AOMK) constructs. This work further confirms that the CCTA trapping approach can make significant improvements in the clinical potential of NTSR1- and other receptor-targeted radiopharmaceuticals.

Graphical Abstract

1. Introduction

The development of radiopharmaceuticals for oncology has largely centered on targeted delivery using molecular markers that are selectively overexpressed on the surface of cancer cells or in the tumor microenvironment [1–3]. Targeted radiopharmaceuticals utilizing low molecular weight constructs, such as small organic molecules and peptides, have been an attractive developmental option due to their inherently rapid tumor targeting and blood clearance properties, which are capable of yielding high tumor-to-non-target ratios relatively quickly post-administration. However, the rapid metabolic degradation and clearance of these low molecular weight carriers can substantially limit the retention and therefore efficacy of diagnostic and/or therapeutic radiopharmaceutical constructs [4, 5]. Due to this limitation, researchers have often sought approaches and constructs designs that would lengthen tumor residence time with, generally, limited success.

Cysteine cathepsins (CCs) are proteases that carry out a variety of biological functions but are primarily attributed to protein degradation and turnover in the endolysosomal compartment of cells [6]. As a consequence, CCs are known to be expressed in high concentrations (~1 mM) in these endolysosomal vesicles. A case in point, cysteine cathepsin B (CatB) and L (Cat L) were found to represent as much as 40% of the total protein content in endolysosomal compartments [7]. In addition, dysregulation, mostly up-regulation, and translocation of these CCs have been linked to several diseases, including cancer [8]. This upregulation of CCs in numerous cancer has driven the development of a variety of classes of reversible and irreversible CC inhibitor as chemotherapeutics. Furthermore, the role of CCs in cancer progression has prompted the development of diagnostic imaging probes [9–12].

Recently, we have begun exploring the use of irreversible inhibitors of CCs to prolong the tumor retention of receptor-targeted agents [13]. This approach incorporates irreversible CC inhibitors into the structure of receptor-targeted constructs. Conceptually, upon binding to the receptor, the agent is internalized and delivered to endolysosomal compartments of the cells. Once in these compartments, the attached inhibitor is able to form, through covalent bonds, high molecular weight adducts with CCs, thus prolonging the cellular retention of the receptor-targeted agent. Specifically, in the report eluded to above, we incorporated a dipeptidyl acyloxymethyl ketone (AOMK) AOMK1 (Fig. 1), a well-known class of irreversible inhibitor for CCs, into the structure of a peptide that targets the neurotensin receptor subtype 1 (NTSR1) [14, 15]. The NTSR1 has been shown to be overexpressed on numerous cancers and has attracted considerable interest as a diagnostic and therapeutic target. Incorporation of AOMK1 into the structure of NTSR1-targeted constructs (NA2f, see Fig. 1) resulted in a nearly two-fold improvement in in vivo tumor residualization over the analogous NA3b, which was conjugated with a trapping inactive control AOMK2 (unable to form CC adducts). While these proof of principle studies confirmed the potential of this approach, the relatively high lipophilic character of the AOMK1-incorporated peptides led to significant non-target uptake which we deemed detrimental to clinical translation [16].

Fig. 1.

Structures of the cysteine cathepsin inhibitors and peptidic conjugates studied in our previous and present work

Herein, we continue the examination of this CC trapping approach to enhance the tumor residualization of NTSR1-targeted peptides. Specifically, we explore the utilization of the more hydrophilic inhibitor 1a (Fig. 1), which is based on the well-known, irreversible E-64 inhibitor, as our CC trapping agent [17]. Our motivation for exploring this hydrophilic CC inhibitor is to improve the hydrophilicity of the resulting NTSR1-targeted conjugate thereby reducing uptake by non-target tissues. To that end, we synthesized inhibitor 1a and the matching, inactive control 1b. These two constructs were subsequently conjugated to the NTSR1-targeted peptides yielding NE1a and NE1b, correspondingly. With these conjugates in hand, we evaluated the biological performance of these analogs using in vitro and in vivo HT-29 cancer models. In particular, we examine how the utilization of a more hydrophilic inhibitor with the CC trapping approach can improve the clinical potential for NTSR1- and, likely by extension, other receptor-targeted agents.

2. Results and discussion

2.1. Synthesis and labeling

The first step in the development of the desired conjugates required the synthesis of the E-64 derivatives 1a (active CC inhibitor) and 1b (inactive control lacking the epoxysuccinyl moiety), outlined in Scheme 1. Firstly, Fmoc-Arg(Pbf)-OH moiety was conjugated to 2-azidoethanamine to produce 2. Using standard Fmoc peptide synthesis approaches, Fmoc-Leu-OH was subsequently conjugated to the amine to give 3, followed by coupling with (+/−)-trans-oxirane-2,3-dicarboxylic acid or succinate acid to yield intermediates 4a and 4b, respectively. Removal of the protecting groups furnished the desired CC inhibitor 1a and inactive control 1b. Note, the target inhibitor 1a will consist of two enantiomers of the epoxide given the use of the (+/−)-trans-oxirane-2,3-dicarboxylic acid(2S, 3S and 2R, 3R). However, while the 2S, 3S isomer has been shown to be a more (~7-fold) potent inhibitor than the 2R, 3R isomer for other cysteine proteases [18], the relatively similar inhibitor profiles and high endolysosomal CC concentration lead us to believe that the disparity between the trapping efficacies of the stereoisomers will likely to be limited with regard to our approach.

Scheme 1.

Synthetic Procedures for 1a and 1b^a

^aReagents and conditions: (a) (i) EDCI, NHS, DMF, 0°C, 2-3h (ii) 2-azidoethanamine, DIEA,DMF, r.t., overnight; (b) (i) 20% piperidine in DMF, r.t., 30min; (ii) COMU, Fmoc-Leu-OH, DIEA, DMF, r.t., 2h; (c) (i) 20% piperidine in DMF, r.t., 30min; (ii) COMU, (+/−)-trans-oxirane-2,3-dicarboxylic acid, DMF, DIEA, r.t., overnight; (c) (i) 20% piperidine in DMF, r.t., 30min; succinic anhydride, TEA, DMF, 50°C, overnight; (e) 90% TFA in DCM, r.t., 2h.

The synthetic schemes and details regarding the synthesis of these desired NTSR1-targeted peptides are outlined in Scheme 2. The synthesis of the starting peptides was carried out by standard SPPS techniques as described previously [13, 19]. Using the protected intermediates 4a and 4b, conjugation of the CC inhibitor or inactive control to the NTSR1-targeted peptide was carried out using copper (I)-catalyzed azide-alkyne cycloaddition (CuAAC) chemistry. After coupling of the chelator 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) to the N-terminus of the peptide, the peptide intermediates were deprotected to yield the desired conjugates NE1a and NE1b. The purification and characterization details of all the intermediates and final products are listed in Table S1.

Scheme 2.

Synthetic Procedures for NE1a and NE1b^a

^aReagents and conditions: (a) CuSO₄, ascorbic acid, TEA, H2O:n-Butanol:DMF (1:1:2), r.t., 1h; (b) (i) DOTA-NHS, DIEA, DMF, r.t., overnight; (ii) 90% TFA in DCM, r.t., 3h.

The NTSR1-targeted conjugates NE1a and NE1b were radiolabeled with lutetium-177 (¹⁷⁷Lu) trichloride by co-incubation in a sodium acetate buffer (pH = 5.3) for 45 min at 80 °C. The remaining unlabeled conjugate was complexed with copper (II) (^natCu) sulfate to aid in purification/separation of the radioconjugate during reverse phase-high performance liquid chromatography (RP-HPLC). Fig. 2A and 2B show the UV and radiometric RP-HPLC profiles indicating high ¹⁷⁷Lu-labeling efficiency of the conjugates. No significant degradation of the conjugates was detected under these radiolabeling conditions. The radiochemical yields for ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b, along with their chromatographic properties, are reported in Table S2.

Fig. 2.

The RP-HPLC profiles of radiolabeling NE1a (A) and NE1b (B) by ¹⁷⁷LuCl₃. The stability and purity of ¹⁷⁷Lu-NE1a (C) and ¹⁷⁷Lu-NE1a (D) in human serum at 37 °C at various time points.

In addition to the radioconjugates, the compounds NE1a and NE1b were also labeled with europium (III) (^natEu) trichloride for utilization in fluorescent cell trafficking studies. Using similar labeling conditions as those used for ¹⁷⁷Lu, the conjugates were readily labeled with efficiencies > 95% using excess ^natEuCl₃ at room temperature (Table S2). The time-resolved fluorescence spectra for ^natEu-NE1a and ^natEu-NE1b were investigated and compared to free ^natEuCl₃. When Eu is complexed with a macrocyclic chelator, efficient energy transfer from excited europium to the chelator leads to blue-shift of the excitation spectra [20]. For ^natEu-NE1a and ^natEu-NE1b, an approximate 10 nm blue shift in excitation maximum from 390 to 380 nm was observed (Fig. S1) confirming the chelation of ^natEu.

2.2. Distribution coefficient and stability studies

To evaluate the hydrophilicity of the radioconjugates, the octanol-PBS distribution coefficients at pH 7.4 (LogD_7.4) were measured (Table S2). As expected, the LogD_7.4 of ¹⁷⁷Lu-NE1a (−3.78 ± 0.10) and ¹⁷⁷Lu-NE1b (−3.51 ± 0.15) were significantly lower than our previously reported analogs NA2f (−1.95 ± 0.05) and NA2f (−2.01 ± 0.06), indicating that the replacement of AOMK1 by the E-64 derivative markedly increased the hydrophilicity of the conjugates. Similarly, the LogD_7.4 values of the ^natEu-labeled conjugates were determined to be identical to their ¹⁷⁷Lu-labeled counterparts as measured by time-resolved fluorescence.

The stability of the radioconjugates in human serum was examined over a 24 h time period at 37 °C, see Fig. 2C and 2D. Both radioconjugates demonstrated modest and similar degradation patterns at 3 h. However, by 24 h, only 36.9 % of the ¹⁷⁷Lu-NE1a was shown to be intact while 49.5 % of ¹⁷⁷Lu-NE1b remained undegraded. This higher degradation of ¹⁷⁷Lu-NE1a is likely due to the reactivity of its electrophilic epoxysuccinyl ring over time. Given the rapid pharmacokinetic profile of most low molecular weight radiopharmaceuticals, the relatively slow degradation rates observed were not expected to substantially impact the biological performance of these NTSR1-targeted agents.

2.3. CC inhibition and in vitro competitive binding studies

To assess the CC-trapping potency, we investigated the CC inhibition affinity of the conjugates. CatB and CatL were chosen as our CC models due to their ubiquitous expression in mammalian cells and the high inhibition potency of E-64 for these proteases [21–23]. The Michaelis-Menten constants (K_m) of CatB and CatL were determined to be 189 ± 2 μM and 36 ± 3 μM respectively by monitoring the initial hydrolysis rates of their corresponding substrates (Z-Arg-Arg-AMC for CatB and Z-Phe-Arg-AMC for CatL) at different concentrations (Fig. S2). The K_m value for CatB was similar but lower than the reported value (390 ± 20 μM [24]); while the K_m for CatL was found to be modestly higher that the literature value (21.5 ± 3.2 μM [25]).

The pseudo-first-order rate constant (k_obs) for each CC was plotted against the florescence intensity hydrolyzed from the substrates by CatB and CatL in the presence of the conjugates at different concentrations (Fig. S3 and S4). The k_obs was plotted against the conjugate concentrations. The inhibition constants (K_i) and second order rate constant (k_i/K_i), where k_i represents the maximum inhibition rate, was determined using hyperbolic regression. Since CCTA 1a and 1b are new derivatives of E-64, we first evaluated their CC-inhibition potency before conjugating them to the peptide backbone. Although 1a was a mixture of diastereomers due to the racemic epoxysuccinyl moiety, the K_i and k_i/K_i for the cathepsins, as with many other E-64 derivatives, was determined to be in the nanomolar inhibition range (K_i = 64 ± 4 nM, k_i/K_i = 130000 ± 1900 M⁻¹s⁻¹ for CatB; K_i = 47 ± 6 nM, k_i/K_i = 16000 ± 1300 M⁻s⁻¹ for for CatL) (Fig. S3), thus demonstrating the high CC-trapping efficiency of the conjugate. No inhibition of 1b to CatB was detected in the test concentration ranges (Fig. S5), which is likely due to the absence of the epoxysuccinyl group in the structure of this compound. Compound NE1a also demonstrated potent nanomolar inhibition to CatB (K_i = 110 ± 7 nM, k_i/K_i = 67000 ± 600 M⁻¹s⁻¹) and CatL (K_i = 76 ± 8 nM, k_i/K_i = 24000 ± 2000 M⁻¹s⁻¹) (Fig. 3A and B). These results matched our expectations according to our previous studies with the conjugation of hydrophilic peptides to the AOMK based CCTA which resulted in minimal disturbances to the affinity towards the CCs. Similar to 1b, the NE1b did not demonstrate any inhibition to the CCs and showed identical observed Michaelis-Menten constant (K_obs) with the K_m over the concentrations investigated (Fig. S6).

Fig. 3.

The hyperbolic regression of pseudo-first-order rate constants (k_obs) (t = 0-60 min) by CatB (A) and CatL(B) versus the concentrations of NE1a ([C]). The substrates for CatB and CatL were Z-Arg-Arg-AMC and Z-Phe-Arg-AMC, respectively. The substrate concentration was from 1 to 500 nM. The K_i and k_i/K_i determined from k_obs = k_i[C]/(Ki + [C]). (C) A dose-response curve showing the inhibition of NTSR1 receptor binding of ¹⁷⁷Lu-N1 by NE1a and NE1b. Values are means ± SD (n = 3).

To discern if the NTSR1 binding was negatively impacted by the introduction of CCTAs into the NTSR1 targeting peptide, the affinity of the conjugates were investigated by in vitro competitive binding with ¹⁷⁷Lu-N1 (¹⁷⁷Lu-DOTA-β-Ala-[N-α-Me⁸, Dmt¹¹, Tle¹²]NT(6-13)), an NTSR1-targeted peptide we have reported elsewhere [19, 26]. The experiment was carried out using HT-29 human colon cancer cells, a well-known NTSR1-positive cell line. As shown in Fig. 3C, the 50% inhibitory concentrations (IC₅₀) of NE1a and NE1b were similar at approximately 60 nM, which is comparable to that of ¹⁷⁷Lu-N1 (31 ±5 nM). Overall, these results demonstrated that the incorporation of the hydrophilic CC inhibitor into the NTSR1-targeted peptide had a minimal impact on receptor affinity.

2.4. Cell uptake, efflux and confocal studies

The internalization rates and efflux studies of the radioconjugates NE1a and NE1b were carried out using HT-29 cells. At 15, 30, 60, 120, and 240 min time points, the cellular uptake, including both internalized and surface-bound radioactivity, of ¹⁷⁷Lu-NE1a (Fig. 4A) and ¹⁷⁷Lu-NE1b (Fig. 4B) were investigated. Both of the conjugates exhibited identical and efficient cellular uptake, approximately 6 % of the total radioactivity added, by the 240 min time point. The internalized activity was the primary component of the overall cellular uptake for these hydrophilic radioconjugates (LogD_7.4 < − 3.5) with very little surface-bound activity observed. This is in contrast with the more hydrophobic AOMK-incorporated NTSR1-targeted agents (e.g., ¹⁷⁷Lu-NA2f, LogD_7.4 = −1.95) reported previously, where approximately 2% was surface-bound at the same time point [13]. This discrepancy is likely due to higher non-specific binding associated with the more hydrophobic AOMK-incorporated analogs [27]. This demonstrates that the selection of the class of CC inhibitor and its impact on the overall hydrophobicity of the NTSR1-targeted analogs can have a significant effect on non-specific uptake.

Fig. 4.

The internalized and surface-bound ¹⁷⁷Lu-NE1a (A) and ¹⁷⁷Lu-NE1b (B) by HT-29 cells. (C) Efflux profiles of the internalized ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b over 24 h in HT-29 cells. Values are means ± SD (n = 3).

Efflux of the radioconjugates was examined at 1, 2, 4, 8, and 24 h time points (Fig. 4C). By 2 h, the effluxed fraction for ¹⁷⁷Lu-NE1a (27.3 ± 1.4 %), capable of CC adduct formation, was significantly lower than the control ¹⁷⁷Lu-NE1b (32.3 ± 1.4 %, p < 0.05). The contrast in efflux percentage between the two conjugates widened over the subsequent time points with ¹⁷⁷Lu-NE1a exhibiting about 12 % lower efflux (higher retention) compared to ¹⁷⁷Lu-NE1b by 24 h. Relative to the previous AOMK-incorporated analog ¹⁷⁷Lu-NA2f (39% effluxed by 24 h), the overall efflux percentage of ¹⁷⁷Lu-NE1a was modestly higher at 50%. This may be due to the AOMK inhibitor being a more efficient trapping agent relative to the E-64 based inhibitor. Although, we believe it likely that the increased non-specific binding of ¹⁷⁷Lu-NA2f, and metabolites, may account for this observed difference between the CCTA-incorporated, NTSR1-targeted agents.

The cell trafficking of the conjugates was studied by using ^natEu-labeled NE1a and NE1b. Fluorescent lanthanides, such as europium, have been introduced as nonradioactive labels in many biomedical applications [28]. For examples, fluoroimmunoassays using ^natEu-chelates have been consequently developed for the labeling of proteins [29, 30], prostate-specific antigen [31], and other biomolecules [32–35] due to its high sensitivity due to time-resolved fluorescence. Using confocal microscopy, cell trafficking studies of the ^natEu-NE1a and ^natEu-NE1b demonstrated that both conjugates were efficiently internalized by the HT-29 cells within 2 h, imparting a strong fluorescence intensity (red) in compartments within the cytoplasm (Fig. S7A and Fig. S8A). Lysotracker® Green was utilized to identify the endolysosomal compartments. There was strong colocalization of the conjugate signal with the endolysosomal vesicles suggesting, as expected, that the ^natEu-labeled conjugates were being trafficked to the endolysosomal system (Fig. S7B and Fig. S8B) upon internalization. Using 20 μM of the NTSR1-targeted conjugate N1 to block receptor-mediated uptake, the cellular uptake of ^natEu-NE1a and ^natEu-NE1b were inhibited by 63.7% and 65.0%, respectively, at the 2h time point. Although, a weak signal that did not overlap with the Lysotracker® was observed on the cell surface and was presumably due to non-specific binding. These data supports the conclusion that the conjugates were internalized by the HT-29 cells via the receptor mediated endocytosis and trafficked to the endolysosomal compartments of the cell.

The cellular retention of ^natEu-NE1a and ^natEu-NE1b was also examined by confocal microscopy, see Fig. 5A. Cells were incubated with ^natEu labeled NE1a or NE1b for 2 h, washed with fresh medium and the efflux evaluated over a 24 time period. At 2 and 4 h, the mean fluorescence of ^natEu-NE1a was higher than the control ^natEu-NE1b, but the differences were not quite statistically significant (p = 0.058 for 4h). However, by the 24 h time point, the mean fluorescence of ^natEu-NE1a was approximately 2-fold (p < 0.001) higher than the control, demonstrating that the active conjugate resulted in a substantial increase in cellular retention (Fig. 5B). Additionally, ^natEu-NE1a exhibited a substantially higher percent colocalization, between the conjugate and Lysotracker®, (Fig. 5C) with the endolysosomal compartments of the cell relative to ^natEu-NE1b over the 24 h time period. Colocalization efficiency values for the active agent remained largely constant at about 60% throughout the study. In contrast, the colocalization efficiency of the inactive control ^natEu-NE1b decreased by approximately 20%. The residualization of florescence intensity and co-localization efficiency for ^natEu-NE1a demonstrates that the E-64 based CC inhibitor is able to increase the cellular retention of the NTSR1-targeted agents in the endolysosomal compartments of cells upon receptor-mediated internalization. This increased residualization is attributed to adduct formation, investigated below, in the CC-rich, endolysosomal vesicles of the cell [36].

Fig. 5.

(A) Representative confocal microscopy images of the efflux of ^natEu labeled NE1a and NE1b (red) form HT-29 cells. Cell endolysosomal compartments were stained with Lysotracker (green). Scale bar = 50 μm. (B) Time-dependent fluorescence intensity of ^natEu per cell as quantified from the confocal images. (C) Co-localization efficiencies of ^natEu (red) overlapping with Lysotracker® (green). All the analysis was performed in 6 random images and were presented as mean ± SD. **p < 0.01, ***p < 0.001, NS = not significant.

2.5. Intracellular trapping of the radioconjugates

In the endolysosomal compartments of cells, CatB and CatL are highly expressed and the active proteases exist in two different forms: single and two chain, with the two chain consisting of a heavy (~25 kDa) and a light chain (~5 kDa) [37–39]. To confirm adduct formation, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed to examine the ability of the radioconjugates to irreversibly bind to CatB and CatL. As the autoradiographic SDS-PAGE shows in Fig. 6A, ¹⁷⁷Lu-NE1a, which contains the active CC inhibitor, demonstrated efficient binding with both human CatB and CatL, providing bands at about 23 - 25 kDa. As expected, no adduct bands were observed for the control ¹⁷⁷Lu-NE1b which contains an inactive inhibitor that is unable to irreversibly bind to CCs. In addition, co-incubation of ¹⁷⁷Lu-NE1a with E-64 prevented adduct formation with both proteases, demonstrating that adduct formation of the radioconjugate occurs through the same mechanism as the E-64 inhibitor.

Fig. 6.

The autoradiography of the SDS-PAGE showing the CatB and CatL binding with the ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b in the presence or absence of the cysteine proteases inhibitor E-64 (A) and the NTSR1 ligand N1 (B). (C). Autoradiographic image of an SDS-PAGE gel examining the time-dependent retention of CC-conjugate adducts in HT-29 cells after pre-incubation with ¹⁷⁷Lu-NE1a for 4 h.

After incubation of ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b with HT-29 cells, the cells were washed to remove extracellular radioactivity and lysed. The cell lysate underwent SDS-PAGE analysis and revealed multiple intracellular, macromolecule adducts with ¹⁷⁷Lu-NE1a. The two most intense bands are about 30 kDa and 25 kDa, which corresponds to the single chain form and heavy chain component (two-chain form) of both mature CatB and CatL [40, 41]. Note that while these adducts are consistent with the molecular weights of these proteases, we did not unambiguously identify the CCs involved in intracellular adduct formation. As expected, the cell lysate of the inactive control ¹⁷⁷Lu-NE1b demonstrated no adduct formation. Interestingly, the level of intracellular adducts formation for ¹⁷⁷Lu-NE1a was unaffected when coincubated with the E-64 inhibitor. This suggests two possibilities: 1) E-64, due to its hydrophilicity, isn’t able to effectively passively diffuse to the endolysosomal compartments and prevent the radioconjugate from forming adducts and/or 2) the CC concentration in these endolysosomal compartments is too high for the limited amount of internalized E-64 to be effective.

Further studies were conducted to examine the impact of an excess (20 μM) of NTSR1-targeted peptide N1 on adduct formation with both CatB, CatL and in HT-29 cells. In Fig. 6B, the presence of N1 has no impact on the ability of ¹⁷⁷Lu-NE1a to form adducts with either commercial purified CatB or CatL. However, coincubation of N1 and ¹⁷⁷Lu-NE1a with HT-29 cells eliminated the adduct signal, strongly arguing that NTSR1-mediated internalization is a requirement for adduct formation. Overall, these results are consistent with our original proposal regarding the trapping mechanism of NTSR1-targeted agents incorporating CC trapping agents [13].

Lastly, using HT-29 cells, the intracellular adduct signal of ¹⁷⁷Lu-NE1a was monitored (Fig. 6C) and quantified at several time points over a 24 h period. The radioactive signal from the SDS-PAGE bands revealed that 55 % of the adduct signals still remained in the cells at 24 h in comparison to the initial intensity at 0 h. While our data suggests the high molecular weight ¹⁷⁷Lu-NE1a-CC adducts are efficiently and rapidly formed in vitro, these adducts are eventually degraded and/or slowly excreted from the cell.

2.6. Biodistribution studies

Biodistribution studies were carried out using an HT-29 tumor xenograft mouse model (Table 1). The radiolabeled conjugates were administrated to mice via tail vein injection. Both of the conjugates showed rapid blood and muscle clearance with distribution values around the detection limit at 4 h post-injection (p.i.). With exception of the intestine, spleen and kidneys, ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b cleared rapidly from all other non-target tissues with values < 1%ID/g at 4 h post-injection. Overall, ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b exhibited dramatically lower radioactive levels in non-target tissues in contrast to the previous AOMK-incorporated agents (¹⁷⁷Lu-NA2f and ¹⁷⁷Lu-NA3b, biodistribution data are presented in Table S3). Indeed, the retention of ¹⁷⁷Lu-NE1a in the liver was about 50, 35 and 17 times lower than ¹⁷⁷Lu-NA2f at 4, 24 and 72 h post injection. In our previous report, we postulated that the high uptake of ¹⁷⁷Lu-NA2f in the liver and other non-target tissues was likely due to hepatic clearance and non-specific internalization. The lower retention of the more hydrophilic ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b in the liver and other non-target tissue suggest that this supposition was correct.

Table 1.

Biodistribution data of the radioconjugates in an HT-29 xenograft mouse model^a

Tissue (ID %/g)	4 h		24 h				72 h
	NE1a	NE1b	NE1a	NE1a + Colchicineb	NE1a + D-lysinec	NE1b	NE1a	NE1b
Blood	0.02 ± 0.01	0.01 ± 0.02	0.01 ± 0.02	0.06 ± 0.04	0.02 ± 0.01	n.d.d	n.d.d	n.d.d
Heart	0.14 ± 0.07	0.13 ± 0.10	0.08 ± 0.06	0.20 ± 0.10	0.14 ± 0.17	0.01 ± 0.08	0.04 ± 0.13	n.d.d
Lung	0.53 ± 0.07	0.37 ± 0.06	0.45 ± 0.13	3.2 ± 0.9	0.41 ± 0.08	0.25 ± 0.07	0.3 ± 0.3	0.24 ± 0.10
Liver	0.78 ± 0.15	0.42 ± 0.05	0.73 ± 0.08	1.4 ± 0.3	0.65 ± 0.09	0.40 ± 0.07	0.67 ± 0.07	0.26 ± 0.05
Pancreas	0.19 ± 0.08	0.12 ± 0.02	0.18 ± 0.14	0.7 ± 1.2	0.14 ± 0.08	0.11 ± 0.06	0.15 ± 0.07	0.08 ± 0.08
Stomach	0.34 ± 0.12	0.27 ± 0.13	0.33 ± 0.18	0.29 ± 0.08	0.21 ± 0.05	0.15 ± 0.07	0.21 ± 0.07	0.08 ± 0.03
Spleen	1.4 ± 0.6	1.0 ± 0.8	1.7 ± 0.6	4 ± 2	1.1 ± 0.9	1.6 ± 0.8	1.9 ± 0.7	0.8 ± 0.3
Small int.	3.5 ± 0.5	2.9 ± 0.6	3.2 ± 0.7	2.2 ± 0.3	1.47 ± 0.09	1.9 ± 0.2	2.3 ± 0.4	1.07 ± 0.06
Large int.	1.2 ± 0.2	1.0 ± 0.3	1.1 ± 0.3	0.67 ± 0.12	0.8 ± 0.3	0.73 ± 0.19	0.73 ± 0.08	0.36 ± 0.07
Kidney	67 ± 12	57 ± 13	80 ± 9	26 ± 12	34 ± 6	39 ± 6	46 ± 4	15 ± 4
Muscle	0.09 ± 0.07	0.04 ± 0.03	0.04 ± 0.05	0.06 ± 0.07	0.07 ± 0.07	0.05 ± 0.03	0.03 ± 0.03	0.01 ± 0.05
Bone	0.8 ± 0.3	0.18 ± 0.07	1.1 ± 0.5	0.36 ± 0.11	0.6 ± 0.2	0.15 ± 0.07	0.6 ± 0.3	0.12 ± 0.03
Brain	0.01 ± 0.02	0.02 ± 0.01	0.02 ± 0. 01	0.07 ± 0. 08	0.02 ± 0. 01	0.01 ± 0.01	0.02 ± 0.01	0.01 ± 0.02
Tumor	6.4 ± 0.6	5.6 ± 1.2	6.2 ± 0.9	1.7 ± 0.6	6.0 ± 0.8	4.4 ± 1.1	3.9 ± 0.5	2.6 ± 0.4
Excretione	70 ± 4	75 ± 4	69 ± 4	85 ± 2	83 ± 3	84.7 ± 1.1	80.8 ± 0.9	91.6 ± 0.9

^aThe radioconjugates were injected to the mice at a dose of 10 μCi (0.37 MBq) /mouse. Data are represented as mean ± SD (n = 5).

^bColchicine was administrated by i.p. injection with a dose of 1mg/kg at 1 h before the i.v. injection of ¹⁷⁷Lu-NE1a.

^cD-lysine was co-injected with ¹⁷⁷Lu-NE1a at a dose of 40mg/kg.

^dThe radioactive signal was non-detectable in the samples.

^eData are presented as ID%.

One exception to the overall favorable clearance of ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b was renal uptake. ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b showed high and statistically identical renal uptake values, 67 ± 12 and 57 ± 13 %ID/g, respectively, at the 4 h time point. By 24 h, the renal uptake of ¹⁷⁷Lu-NE1a increased to 80 ± 9 %ID/g while ¹⁷⁷Lu-NE1b decreased to 26 ± 12 %ID/g. At 72 h, the renal retention for ¹⁷⁷Lu-NE1a (46 ± 4 %ID/g) was 3-fold higher than the control ¹⁷⁷Lu-NE1b (15 ± 4 %ID/g). The substantially increased renal retention of ¹⁷⁷Lu-NE1a, the active CC trapping conjugate, is most likely due to the non-specific cellular uptake mechanisms of the proximal tubule cells within the kidney [19, 26, 42]. Once taken up in the proximal tubule cells, ¹⁷⁷Lu-NE1a would be expected to form intracellular adducts in a manner analogous to that proposed for NTSR1-mediated internalization. This increased renal retention is consistent with our previously reported AOMK-incorporated NTSR1-targeted peptide [13].

In the HT-29 tumor xenograft mouse model, there are two tissues that can be used to evaluate in vivo NTSR1 targeting; the HT-29 tumors (human NTSR1) and the intestines (mouse NTSR1), particularly the small intestines [43]. HT-29 tumor uptake of ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b was statistically identical with values of 6.4 ± 0.6 and 5.6 ±1.2 ID%/g, correspondingly, at 4 hours p.i. By 24 hours post-administration, ¹⁷⁷Lu-NE1a maintained a retention of 6.2 ± 0.9 ID%/g in the tumor, while the tumor retention of ¹⁷⁷Lu-NE1b decreased by 21 % giving a significantly lower uptake at 4.4 ± 1.1 ID%/g (p < 0.05). At the 72 time point, ¹⁷⁷Lu-NE1a exhibited a tumor retention that was 52% higher (3.9 ± 0.5 ID%/g) compared to the inactive control ¹⁷⁷Lu-NE1b (2.6 ± 0.4 ID%/g). The retention profile in the NTSR1-positive small intestines was analogous to the tumor profile. Comparison of the tumor retention profiles of ¹⁷⁷Lu-NE1a with that of other reported NTSR1-targeted agents (retention of 38% - 78% of initial uptake by 24 h p.i.) revealed that ¹⁷⁷Lu-NE1a exhibited superior tumor retention [44–46]. These results demonstrate that the hydrophilic CCTA 1a is able to enhance the tumor retention of ¹⁷⁷Lu-NE1a by the intracellular CC trapping mechanism we discussed above. In comparison to the relatively hydrophobic AOMK-incorporated analogs ¹⁷⁷Lu-NA2f and ¹⁷⁷Lu-NA3b, the hydrophilic radioconjugates reported in this paper had lower (approximately 38%) initial tumor uptake values, but similar retention profiles. Nevertheless, due to the substantially lower uptake in most non-target tissues,¹⁷⁷Lu-NE1a demonstrated superior tumor to non-targeted (T/NT) ratios (Table S4) relative to ¹⁷⁷Lu-NA2f at all-time points investigated.

De Jong and colleagues demonstrated that colchicine, a known inhibitor of microtubule-dependent receptor-mediated endocytosis [47], was able to substantially reduce [¹¹¹In-DTPA⁰]octreotide renal uptake in a non-tumor bearing rat model [48]. Following the principle in vivo biodistribution studies of ¹⁷⁷Lu-NE1a above, the effects of colchicine (1 mg/kg), administered by intraperitoneal (i.p.) injection 1 h prior to the radioconjugate, on the renal uptake and in vivo NTSR1-targeting was investigated. At the 24 h time point, colchicine lowered the renal uptake (Table 1) of ¹⁷⁷Lu-NE1a by 67%, which is nearly identical to that reported (63%) in the rat model described above. Uptake of ¹⁷⁷Lu-NE1a in NTSR1-positive small intestines and HT-29 tumors were also significantly inhibited with the administration of colchicine. The small intestines demonstrated a 32% reduction in uptake at 24 h post-administration (p < 0.05). Microtubule inhibition by colchicine had an even greater effect on the HT-29 tumors which yielded a 73% reduction in ¹⁷⁷Lu-NE1a uptake at the same time point (p < 0.05). This result is particularly significant since it confirms that inhibition of internalization leads to a substantial reduction in ¹⁷⁷Lu-NE1a uptake. This would indicate that ¹⁷⁷Lu-NE1a is not predominantly targeting and forming in vivo adducts with extracellular CCs, but instead suggests that targeting/retention of this radioconjugate relies upon NTSR1-mediated internalization. The finding that colchicine led to observable inhibition of uptake in the NTSR1-positive small intestines is dissimilar to that of [¹¹¹In-DTPA⁰]octreotide in somatostatin-positive organs in which colchicine had no observable effect on receptor-mediated uptake.⁴⁴ Many variables (e.g., animal model, PK profile of the agent, receptors) are different between their study and ours. To date, it is not clear to us which factor(s) may be leading to these differing results.

Co-injection of positive amino acids is well-known to mitigate renal uptake and is currently utilized as a renal protectant in the clinic [49–51]. In order to examine the impact of this approach to reduce the renal uptake of ¹⁷⁷Lu-NE1a, D-Lys (40 mg/kg) was co-administered with the radioconjugate. At 24 h post-injection, the renal uptake of ¹⁷⁷Lu-NE1a with D-Lys dropped by 58 % in comparison to the data without the D-Lys co-injection (p < 0.001). Importantly, the HT-29 tumor uptake mediated by the NTSR1 receptor was unaffected. This result confirms the feasibility of using D-lysine as an efficient renal blocking agent for NTSR1-targeted agents synergistically designed with CC trapping agents.

2.7. Evaluation of tissue adduct formation of the radioconjugates

To elucidate the mechanism of the extended in vivo retention of ¹⁷⁷Lu-NE1a, SDS-PAGE analysis was performed on tumor as well as liver and kidney samples after the injection of ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b (Fig. 7A). Both radioconjugates (800 μCi (29.6 MBq)) were administrated to the HT-29 xenograft-bearing mice and sacrificed at 72 h p.i. to collect the desired tissues samples. In the tumor samples, similar adduct profiles as compared to the in vitro studies for ¹⁷⁷Lu-NE1a were observed suggesting the in vivo macromolecular adducts may be the same as the in vitro studies. Conversely, the control radioconjugate ¹⁷⁷Lu-NE1b demonstrated no signs of adduct formation. Not surprisingly, ¹⁷⁷Lu-NE1a formed a significant amount of macromolecular adducts in the kidney at 20-30 kDa which was similar to profiles of tumor tissues, suggesting that the increased renal retention was also due to the CC trapping effect. Co-injection of D-lysine yielded no inhibition of tumor adduct formation, but a remarkable drop (47 % lower than the non-blocked sample) in adduct formation was observed in the kidney samples. Again, this confirms that D-Lys is able to inhibit adduct formation through reduction of the endocytic uptake of the radioconjugate. Adduct formation for ¹⁷⁷Lu-NE1a in the liver samples could not be detected. This was in contrast with ¹⁷⁷Lu-NA3f in which significant amount of adducts were observed due to the more hydrophobic nature of the radioconjugate resulting in non-specific uptake.

Fig. 7.

(A) The autoradiography of SDS-PAGE of the tumor, liver, and kidney samples at 72 h p.i. of ¹⁷⁷Lu-NE1a and ¹⁷⁷Lu-NE1b in mice. The radioconjugates were injected to the mice at a dose of 800 μCi (29.6 MBq) /mouse. (B) Percentage of the macromolecules associated radioactivity (Mw > 10 kDa) in tumor, liver, and kidney samples at 72 h after administration of the radioconjugates (n=3). *p < 0.05, **p < 0.01, ***p < 0.001, NS = not significant.

Lastly, the CC trapping ability of ¹⁷⁷Lu-NE1a was further examined using centrifugal filtration (10 kDa MWCO) to separate radioactive macromolecular adducts from the low-molecular-weight radioactive agents/metabolites in the tissue lysates (Fig. 7B). In the tumor at 72 h, the results show that the macromolecular adducts were responsible for about 80 % of the radioactivity resident in the tumor, liver and kidney (with or without D-lysine blocking). Overall, this study suggests that ¹⁷⁷Lu-NE1a forms in vivo macromolecule adducts with CCs and are responsible for the longer retention profile. By using this strategy in the development of low molecular weight targeted agents, significant improvement in the tumor retention may be achievable resulting in improved diagnostic and/or therapeutic potential.

3. Conclusion

Small molecules, peptides and other low-molecular-weight receptor-targeted platforms offer a promising route for the development of clinically viable diagnostic and therapeutic agents for cancer. However, the intrinsically high diffusion and/or rate of metabolic degradation of these agents often diminishes their potential by reducing tumor retention. In this study, we continue the exploration of CC inhibitors to increase the tumor retention of receptor-targeted agents. Specifically, we examined the ability of a novel hydrophilic inhibitor 1a, derived from E-64, to act as an efficient CCTA to increase the tumor retention of an NTSR1-targeted peptide. In vitro studies revealed that the NTSR1-targeted analog incorporating the CCTA, ¹⁷⁷Lu-NE1a, was able to efficiently form macromolecular adducts, presumably with CCs, upon internalization by NTSR1-positive cells. In vivo biodistribution and SDS-PAGE autoradiography studies demonstrated that ¹⁷⁷Lu-NE1a had a substantially higher (2-fold) increase in tumor retention relative to the control through the proficient formation of macromolecules adducts within NTSR1-positive tumors. Furthermore, this study also revealed the profound impact of the hydrophilicity of the CCTA on the biological performance of the incorporated NTSR1-targeted peptide. Relative to our previous work using a peptidyl (acyloxy)methyl ketone CC inhibitor, the more hydrophilic ¹⁷⁷Lu-NE1a demonstrated a substantially lower amount of non-target accumulation with exception to renal retention. The high renal retention could be partially ameliorated through the use of D-Lys or colchicine, both inhibitors of non-specific binding/cellular internalization in the renal proximal tubule cells. Overall, the results obtained herein demonstrate the potential of the CCTA approach to increase the efficacy of receptor-targeted agents through increased retention at the target site. This approach has considerable potential for increasing the effectiveness of many other receptor-targeted agent constructs. Though, radiotherapeutic applications utilizing this methodology will likely be hampered due to renal uptake and retention. Current efforts in our laboratory are exploring structure-activity relationships and alternative design approaches to reduce the renal retention of CCTA-incorporated, receptor-targeted agents.

4. Experimental section

4.1. General

4.1.1. Materials

N,N-dimethylformamide (DMF), dichloromethane (DCM), methanol, ethyl acetate, acetonitrile, formic acid, acetone, diethyl ether, trifluoroacetic acid (TFA), pyridine, piperidine and N-methylpyrrolidone (NMP) were purchased from Fisher Scientific. Fluorenylmethyloxycarbonyl (Fmoc)-protected natural amino acids, Fmoc-Tle-OH, Fmoc-Gly(Propargyl)-OH and N,N-diisopropylethylamine (DIEA) were purchased from Chem-Impex International. (+/−)-Trans-oxirane-2,3-dicarboxylic acid, Europium(III) chloride (^natEuCl₃), copper(II) sulfate (CuSO₄), 1-butanol, ascorbic acid, triethylamine (TEA), ethylenediaminetetraacetic acid (EDTA), N-hydroxysuccinimide (NHS), Brij®35, and E-64 were obtained from Sigma-Aldrich. The 2-azidoethanamine was prepared according to the reported method [52]. Fmoc-D-Ser(t-Bu)-OH was purchased from NovaBiochem. (1-Cyano-2-ethoxy-2-oxoethylidenaminooxy)dimethylamino-morpholino-carbenium hexafluorophosphate (COMU) was purchased from AK Scientific. Fmoc-Leu-SASRIN™ resin (200-400 mesh), Z-Arg-Arg-AMC, Z-Phe-Arg-AMC, N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride (EDCI), and sodium dodecyl sulfate (SDS) were obtained from Bachem. Fmoc-N-Me-Arg(Pbf)-OH was produced by ChemPep. Fmoc-2,6-dimethyl-L-tyrosine (Dmt) was purchased from Key Organics. DOTA-NHS ester was produced by Macrocyclics. Lutetium-177 chloride (¹⁷⁷LuCl₃) was obtained from the U.S. Department of Energy’s National Isotope Development Center. McCoy’s 5A medium (1×; Iwakata & Grace Modification) with L-glutamine were obtained from Mediatech, Inc. Human serum was obtained from MP Biomedicals. TrypLE Express was obtained from Invitrogen. Penicillin-streptomycin solution and 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) were procured from HyClone Laboratories, Inc. Fetal Bovine Serum (FBS) was purchased from Gibco by Life Technologies Corporation. BD Cytofix Fixation buffer was obtained from BD Biosciences. Novex™ Tris-Glycine SDS sample buffer, Pierce™ RIPA buffer, PageRuler™ Prestained protein ladder, Halt™ Protease inhibitor cocktail, LysoTracker Green DND-26, NuPAGE® sample reducing reagent (10×), and Tween™ 20 were purchased from Thermo Fisher Scientific. Cathepsin B (CatB, human liver), Cathepsin L (Cat L, human liver) and Amicon Ultra-4 centrifugal filters (MWCO 10 kDa) were purchased from EMD Millipore. Five-week-old female SCID mice were purchased from The Jackson Laboratory. The human colon cancer cell line HT-29 was obtained from the American Type Culture Collection and cultured under vendor-recommended conditions. All procedures utilizing animals conform to the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and were approved by the Institutional Animal Care and Use Committee at the University of Nebraska Medical Center.

4.1.2. Instrumentation

Peptides were synthesized by solid phase peptide synthesis (SPPS) on a Liberty microwave peptide synthesizer from CEM. A Waters e2695 system equipped with a Waters 2489 absorption detector and a Waters Qtof Micro electrospray ionization mass spectrometer was used to perform high-performance liquid chromatography(HPLC)/mass spectrometry (MS) analyses. ¹H-NMR and ¹³C-NMR spectra were recorded on a Bruker Avance-III HD 600 MHz instrument using CDCl3 and DMSO-d₆ as the solvent. A Phenomenex Jupiter C12 Proteo 250 × 10 mm semi-prep column was used for the purification of bulk amounts of peptides. Evaluation and purification of radiolabeled conjugates were performed on a Waters 1515 binary pump equipped with a Waters 2489 absorption detector and a Bioscan Flow Count radiometric detector system. The radioactivity of the cell samples and tissue homogenates was quantified by Multi-Wiper™ multi-well wipe test counter. Gamma decay detection of ¹⁷⁷Lu-labeled conjugates for biodistribution studies was accomplished using a NaI (Tl) well detector constructed by AlphaSpectra Inc. Fluorescence intensities were measured by a SpectraMax® M5 multimode plate reader. Lab-Tek chambered #1.0 borosilicate coverglass disks (4 well) were used for confocal cell imaging. Confocal microscopy images were taken on a Leica LSM510 META Microscope equipped with an argon laser. Autoradiography was recorded via BAS storage phosphor screens and scanned by GE Lifesciences Typhoon FLA 9500 variable mode imager.

4.2. Chemistry

4.2.1. Synthesis of CCTA (1a) and complimentary inactive control (1b)

Corresponding ¹H-NMR, ¹³C-NMR and LRMS-ESI spectra of intermediates and final products are presented in the Supplementary Material.

(9H-fluoren-9-yl)methyl (S)-(1-((2-azidoethyl)amino)-1-oxo-5-(3-((2,2,4,6,7-pentamethyl-2,3-dihydrobenzofuran-5-yl)sulfonyl)guanidino)pentan-2-yl)carbamate (2).

To a solution of Fmoc-Arg(Pbf)-OH (500 mg, 0.77 mmol) in DMF (5 mL), EDCI (162 mg, 0.85 mmol) and NHS (116 mg, 1.00 mmol) in DMF (1 mL) were added dropwise at 0 °C. The solution was stirred at room temperature for 2 h prior to the addition of a 3 mL DMF solution of 2-azidoethanamine (132 mg, 1.54 mmol) and DIEA (405 μL, 2.31 mmol). The mixture was stirred overnight and partitioned in ethyl acetate (30 mL × 3) and brine (20 mL). The combined organic phases were washed with water and concentrated by rotary evaporation to yield a white foam (508 mg, 92.1%). ¹H-NMR (500 MHz, CDCl₃): δ 7.73 (d, J = 7.5 Hz, 2H), 7.55 (d, J = 7.0 Hz, 2H), 7.45 (br, 1H), 7.37 (d, J = 7.5 Hz, 2H), 7.24-7.22 (m, 2H), 6.27 (s, 2H), 6.05 (d, J = 7.0 Hz, 1H), 4.33-4.31 (m, 3H), 4.14 (m, 1H), 3.43-3.27 (m, 6H), 2.90 (s, 2H), 2.58 (s, 3H), 2.50 (s, 3H), 2.07 (s, 3H), 1.86-1.55 (m, 5H), 1.42 (s, 6H). LRMS-ESI (m/z): [M+H]⁺ calcd. for C₃₆H₄₄N₈O₆S H⁺ 717.3, found 717.3.

(9H-fluoren-9-yl)methyl ((S)-1-(((S)-1-((2-azidoethyl)amino)-1-oxo-5-(3-((2,2,4,6,7-pentamethyl-2,3-dihydrobenzofuran-5-yl)sulfonyl)guanidino)pentan-2-yl)amino)-4-methyl-1-oxopentan-2-yl)carbamate (3).

Conjugate 2 (500 mg, 0.70 mmol) was dissolved in 20% piperidine in DMF (10 mL) and stirred for 30 min. The solution was extracted with ethyl acetate (30 mL × 3) and the organic phases combined and concentrated by rotary evaporation. The residue was re-dissolved in DMF (5 mL) and added to a solution of COMU (342 mg, 0.80 mmol) and Fmoc-Leu-OH (247mg, 0.70 mmol) in DMF (10 mL). The solution was stirred for 5 min before DIEA (250 μL, 1.4 mmol) was added. This mixture was allowed to stir for an additional 2 h. The product was extracted with brine (100 ml) and ethyl acetate (30 mL × 3). The combined organic phases were washed twice with aq. HCl (100 mL, 0.5N), saturated NaHCO₃ (100 mL) and DI water (100 mL) before finally drying over Na₂SO₄. The organic layer was filtered, evaporated to dryness and purified by flash column chromatography (silica gel, DCM/methanol = 15:1) to afford a pale foam (424 mg, 73.0%). ¹H-NMR (500 MHz, CDCl₃): δ 7.75 (d, J = 7.5 Hz, 2H), 7.57 (d, J = 7.5 Hz, 2H), 7.46 (br, 1H), 7.38 (t, J = 7.5 Hz, 2H), 7.8 (m, 2H), 6.08 (s, 2H), 5.47 (br, 1H), 4.72-4.33 (m, 5H), 4.14 (m, 1H), 3.46-3.21 (m, 6H), 2.92 (s, 2H), 2.57 (s, 3H), 2.50 (s, 3H), 2.01 (s, 3H), 1.65 (m, 2H), 1.43 (s, 6H), 1.26 (m, 2H), 1.11 (s, 2H), 0.96-0.92 (m, 8H). LRMS-ESI (m/z): [M+H]⁺ calcd. for C₄₂H₅₅N₉O₇S H⁺ 830.4, found 830.2.

3-(((S)-1-(((S)-1-((2-azidoethyl)amino)-1-oxo-5-(3-((2,2,4,6,7-pentamethyl-2,3-dihydrobenzofuran-5-yl)sulfonyl)guanidino)pentan-2-yl)amino)-4-methyl-1-oxopentan-2-yl)carbamoyl)oxirane-2-carboxylic acid (4a).

Compound 3 (200 mg, 0.24 mmol) was dissolved in 20% piperidine in DMF (5 mL) and stirred for 30 min. The solution was extracted with ethyl acetate (20 mL ×3) and the organic phase was concentrated by rotary evaporation. The resulting residue was dissolved in DMF (5 mL) and was added to a 10 mL DMF solution containing a mixture of COMU (171 mg, 0.40 mmol) and (+/−)-trans-oxirane-2,3-dicarboxylic acid (158 mg, 1.20 mmol). This mixture was stirred for 5 min before DIEA (500 μL, mmol) was added and was allowed to stir overnight. The product was isolated by extraction using brine (100 ml) and ethyl acetate (30 mL × 3). The organic phases were combined, washed twice with aq. HCl (100 mL, 0.5 N) and water (100 mL × 3), and dried over Na₂SO₄. The organic phase was filtered, evaporated to dryness and the residue purified by a semi-preparative Proteo C12 HPLC column with a 15 min gradient and a flow rate of 5.0 mL/min to yield compound 4a as a yellow powder (99 mg, 57.2%). See Table S1 for chromatography details. ¹H-NMR (500 MHz, (DMSO-d₆): δ 8.59 (dd, J = 29.5, 8.0 Hz, 2H), 8.13 (m, 1H), 8.08 (m, 1H), 6.64 (br, 1H), 6.38 (br, 1H), 4.39 (m, 1H), 4.19 (m, 1H), 3.66 (s, 1H), 3.48 (d, J = 9.0 Hz, 1H), 3.26-3.16 (m, 4H), 3.02 (s, 2H), 2.96 (s, 2H), 2.54 (s, 1H), 2.47 (s, 3H), 2.42 (s, 3H), 2.01 (s, 3H), 1.59-1.23 (m, 13H), 0.88-0.84 (m, 6H). ¹³C-NMR (125 MHz; (DMSO-d₆): δ 172.0, 169.2, 163.5, 157.9, 156.5, 131.9, 124.8, 86.8, 53.1, 52.8, 51.9, 51.8, 51.7, 50.4, 42.9, 41.3, 28.8, 24.7, 23.5, 23.4, 22.0, 19.4, 18.1, 12.7. LRMS-ESI (m/z): [M+H]⁺ calcd. for C₃₁H₄₇N₉O₉S H⁺ 722.3, found 722.2.

4-(((S)-1-(((S)-1-((2-azidoethyl)amino)-1-oxo-5-(3-((2,2,4,6,7-pentamethyl-2,3-dihydrobenzofuran-5-yl)sulfonyl)guanidino)pentan-2-yl)amino)-4-methyl-1-oxopentan-2-yl)amino)-4-oxobutanoic acid (4b).

Compound 3 (200 mg, 0.24 mmol) was dissolved in 20% piperidine in DMF (5 mL) and stirred for 30 min. The solution was extracted with ethyl acetate (20 mL ×3) and the organic phase was concentrated by rotary evaporation. The resulting residue was dissolved in DMF (5 mL). To this solution was added succinic anhydride (75 mg, 0.75 mmol) and TEA (140 μL, 1 mmol) followed by stirring and heating to 50 °C for overnight. Once cooled, the produce was extracted using brine (100 ml) and ethyl acetate (30 mL × 3). The combined organic phases were washed twice with aq. HCl (100 mL, 0.5 N), water (100 mL × 3), and dried over Na₂SO₄. The organic layer was filtered, evaporated to dryness and purified by a semi-preparative Proteo C12 HPLC column to give a white powder (87 mg, 51.3%). See Table S1 for chromatography details. ¹H-NMR (500 MHz, (DMSO-d₆): δ 8.55-7.98 (m, 4H), 7.40 (br, 3H), 4.43 (dq, J = 51.5, 6.0 Hz, 1H), 4.24 (m, 1H), 3.51 (s, 1H), 3.26-3.17 (m, 4H), 3.10-3.06 (m, 3H), 1.66 (m, 1H), 1.54-1.44 (m, 6H), 0.88-0.83 (m, 6H). ¹³C-NMR (125 MHz; (DMSO-d₆): δ 171.8, 171.7, 171.1, 170.7, 167.6, 167.2, 164.2, 157.4, 157.3, 54.3, 54.2, 52.7, 52.4, 52.2, 52.1, 51.4, 51.1, 50.1, 48.8, 40.9, 38.3, 29.2, 29.1, 25.1, 25.0, 24.4, 24.3, 23.2, 21.9, 21.7. LRMS-ESI (m/z): [M+H]⁺ calcd. for C₃₁H₄₉N₉O₈S H⁺ 708.4, found 708.2.

3-(((S)-1-(((S)-1-((2-azidoethyl)amino)-5-guanidino-1-oxopentan-2-yl)amino)-4-methyl-1-oxopentan-2-yl)carbamoyl)oxirane-2-carboxylic acid (1a).

To compound 4a (20 mg, 28 μmol) was added a solution of 90% TFA in DCM (1 mL). The solution was stirred at room temperature for 2 h before the solvent was removed by nitrogen flow. The product was purified by a semi-preparative Proteo C12 HPLC column with a 15 min gradient and a flow rate of 5.0 mL/min to give compound 1a as a white film (5.6 mg, 42.6%). See Table S1 for chromatography details. ¹H-NMR (500 MHz, (DMSO-d₆): δ 8.05 (d, J = 7.5 Hz, 1H), 7.99 (s, 1H), 7.82 (d, J = 8.0 Hz, 1H), 6.65 (br, 1H), 6.38 (br, 1H), 4.25 (q, J = 7.5 Hz, 1H), 4.17 (q, J = 5.5 Hz, 1H), 3.35-3.17 (m, 4H), 3.03-3.01 (m, 2H), 2.96 (s, 2H), 2.47 (s, 3H), 2.39 (s, 3H), 2.38-2.33 (m, 4H), 2.01 (s, 3H), 1.62-1.45 (m, 13H), 0.87 (dd, J = 14.5, 6.0 Hz, 6H). ¹³C-NMR (125 MHz; (DMSO-d₆): δ 173.9, 172.0, 171.5, 157.4, 156.0, 131.4, 124.3, 86.3. 52.1, 51.3, 49.9, 42.5, 40.6, 38.1, 29.9, 29.2, 28.3, 24.1, 23.1, 21.5, 18.9, 17.6, 12.3. LRMS-ESI (m/z): [M+H]⁺ calcd. for C₁₈H₃₁N₉O₆ H⁺ 470.2, found 470.1.

4-(((S)-1-(((S)-1-((2-azidoethyl)amino)-5-guanidino-1-oxopentan-2-yl)amino)-4-methyl-1-oxopentan-2-yl)amino)-4-oxobutanoic acid (1b).

This compound was synthesized by the method described for compound 1a. The compound was purified by a semi-preparative Proteo C12 HPLC column to provide a white film (4.9 mg, 38.4%). See Table S1 for chromatography details. ¹H-NMR (500 MHz, (DMSO-d₆): δ 8.38 (br, 1H), 8.25 (d, J = 6.5, 1H), 8.16 (m, 1H), 7.89 (s, 1H), 7.24 (br, 3H), 4.15-4.09 (m, 2H), 3.24-3.16 (m, 3H), 3.09-3.01 (m, 3H), 2.47-2.25 (m, 4H), 1.68-1.61 (m, 4H), 1.49-1.46 (m, 2H), 0.87 (dd, J = 26.5, 6.5 Hz, 6H). ¹³C-NMR (125 MHz; (DMSO-d₆): δ 175.6, 173.1, 172.6, 172.1, 157.2, 52.9, 52.0, 50.0, 40.7, 38.3, 30.9, 28.6, 24.6, 24.7, 23.3, 21.4. LRMS-ESI (m/z): [M+H]⁺ calcd. for C₁₈H₃₃N₉O₅ H⁺ 456.3, found 456.3.

4.2.2 Synthesis of NE1a and NE1b.

All of the corresponding LRMS-ESI and HRMS-ESI spectra are presented in the Supplementary Material. The chromatography details and yields are listed in Table S1.

Synthesis of NTSR1 peptide backbone 5.

The peptides were produced by SPPS. Briefly, Fmoc-Leu-SASRIN™ resin (150 mg, 0.1 mmol) was deprotected by 20% piperidine in DMF (7 mL) to expose the primary amine. Fmoc-L-Tle-OH (177 mg, 0.5 mmol) was coupled to the resin in the presence of COMU (214 mg, 0.5 mmol) and DIEA (90 μl, 1 mmol) in DMF (5 mL). This process of deprotection and conjugation was repeated until the desired peptide was synthesized. Cleavage of the protected peptide (orthogonal protecting groups intact) from the resin was achieved by shaking the resin with 1% TFA in dry DCM (5 × 3 mL) for 2 min. The filtrates were immediately neutralized with 5% pyridine in methanol (1 mL) and evaporated to dryness. This residue was dissolved in methanol (1 mL) and the crude peptides were precipitated in cold water (30 mL). The peptides were purified by a semi-preparative Proteo C12 HPLC column with a 15 min gradient and a flow rate of 5.0 mL/min to give the desired protected peptide.

General procedure for the synthesis of conjugates 6a and 6b.

To a solution of peptide 5 (2 μmol) and compound 4a or 4b (5 μmol) in water/n-butanol/DMF (200 uL, v/v/v=1:1:2) was added CuSO₄ (200 μg, 1.25 μmol) in water (50 μL). After stirring for 5 min, a solution of ascorbic acid (1 mg, 6 μmol) and TEA (5 μL) in water (50 μL) was added to the mixture. The reaction mixture was stirred for 1 h at room temperature under nitrogen. The product was obtained by purification via a semi-preparative Proteo C12 HPLC column with a 15 min gradient and a flow rate of 5.0 mL/min to give the target compound.

General procedure for the synthesis of conjugates NE1a and NE1b.

Conjugates 6a or 6b (1 μmol) and DOTA-NHS ester (2.3 mg, 3 μmol) were dissolved in DMF (3 mL). The solution was basified with DIEA (0.081 mL, 0.47 mmol) and stirred at room temperature for overnight. The completion of the conjugation reaction was confirmed by HPLC before the removal of the solvent under nitrogen flow. To this residue was added a 90% TFA in DCM (300 μL). This solution was stirred at room temperature for 3 h under nitrogen. The solvent was removed by nitrogen flow and the residue redissolved in DMF (300 μL). This solution was purified via a semi-preparative Proteo C12 HPLC column with a 15 min gradient and a flow rate of 5.0 mL/min to give the target compound.

4.3. General procedure for the conjugate labeling

An aliquot of the conjugate (50 μg) in 0.1 M sodium acetate buffer (pH 5.5, 100 μL) was mixed with a predetermined amount of ¹⁷⁷LuCl₃ (37 MBq (1 mCi)) and incubated at 85 °C for 20 min. Subsequently, CuSO₄ (3 mg, 38.5 μmol) was added and incubated for 1 min at 85 °C in order to complex to the unlabeled conjugate and enhance separation. The mixture was purified by the radio-RP-HPLC system and the radiolabeling efficiency (RE) was calculated based on the analysis of the chromatograms. For the ^natEu-labeling, the conjugate (150 μg) was dissolved in 100 μL of sodium acetate buffer and incubated with ^natEuCl₃ (3 mg) at room temperature for 10 min. The ^natEu-labeled conjugates were purified by RP-HPLC system. To remove organic eluent, the labeled conjugate was loaded onto an Empore C18 high-performance extraction cartridge followed by washing with water (3×3 mL) and elution by ethanol/saline solution (v/v = 6:4, 200 μL. See Table S2 for chromatography details and labeling yields.

4.4. Distribution coefficient (Log D_7.4) of the conjugates

For the ¹⁷⁷Lu-labeled conjugates: In a 1.5 mL centrifuge tube, 0.5 mL of 1-octanol was added to 0.5 mL of PBS (pH 7.4) containing the radiolabeled peptide (500, 000 cpm). The solution was vigorously vortexed for 2 min at room temperature and subsequently centrifuged to yield two immiscible layers. The radioactivity of the aliquots (100 μL) taken from each layer was quantified by a gamma counter. From this data, the LogD_7.4 for each radioconjugate was calculated. For the ^natEu-labeled conjugates: In a 1.5 mL centrifuge tube, 0.2 mL of 1-octanol was added to 0.2 mL of PBS (pH 7.4) containing the ^natEu-labeled peptide (300 μg). The aliquots (100 μL) from each layer were dried under vacuum and recovered in D.I. water (100 μL). The time-resolved fluorescence of the samples from each layer was measured at 590 nm using excitation at 380 nm.

4.5. Stability of the radioconjugates in human serum

The radioconjugates (1.5 mCi, 55.5 MBq,) was added to 300 μL of human serum and incubated at 37 °C for 24 h. At predetermined time points (1, 3, and 24 h), acetonitrile (50 μL) was added and the mixture was centrifuged at 12,000×g for 5 min. The supernatant was collected and dried with a nitrogen flow. The sample was reconstituted in water (100 μL) and analyzed by radio-HPLC using the gradient described in the radiolabeling procedure.

4.6. Determination of CCs inhibitory constants of the conjugates

4.6.1. Determination of the K_m and V_max of CatB and CatL to their fluorogenic substrates

Two types of different assay buffer were prepared before the study. Buffer 1: phosphate buffer (0.1 M, pH = 5.8, containing EDTA (1 mM), DTT (2.7 mM), and Brij®35 (0.03%)). Buffer 2: acetate buffer (pH = 5.5, containing sodium acetate (340 mM), acetic acid (60 mM), EDTA (1 mM) and dithiothreitol (3 mM)). The solution of CatB (50 μL, 2 nM) in buffer 1 and CatL (50 μL, 2 nM) in buffer 2 were mixed with Z-Arg-Arg-AMC (50 μL solution in assay buffer 1 at concentrations of 12.5 μM, 25 μM, 50 μM, 100 μM, 150 μM, 250 μM, 500 μM and 1mM) and Z-Phe-Arg-AMC (50 μL solution in assay buffer 2 at concentrations of 2.5 μM, 5 μM, 10 μM, 25 μM, 50 μM, 100 μM, 200 μM, and 300μM), respectively. Each mixture was incubated at 37 °C and the fluorescence of the liberated aminomethylcoumarin at 460 nM using 355 nM excitation was measured at predetermined time points (0, 2, 4, 6, 8, and 10 min). The fluorescence intensity was plotted versus time and the reaction rate (ν₀) for each substrate concentration was calculated as the slope of the trend line obtained by linear regression. The K_m and maximum reaction rate (V_max-obs) for the CatB and CatL were separately determined from the equation ν₀ = V_max[S]/(K_m + [S]) where [S] stands for the substrate concentration. The equation was solved by nonlinear regression using GraphPad Prism 5. All measurements were performed in biological triplicate.

4.6.2. Determination of the inhibition constant of the conjugates to CatB and CatL

For CatB, the enzyme in assay buffer 1 (30 μL) was mixed with the conjugate (30 μL) in 96-well plate. After the solution was mixed, the Z-Arg-Arg-AMC in assay buffer (30 μL) was added to the well, yielding a final cathepsin B concentration of 2 nM, substrate concentration ([S]) of 800 μM, and conjugate concentration ([C]) of 1 nM, 5 nM, 12.5 nM, 25 nM, 50 nM, 100 nM, 133.3 nM, and 500 nM. The mixture was incubated at 37 °C and the liberated fluorescence at 460 nM using 355 nM excitation was recorded at every 5 min for 1 h. The fluorescence intensity (Y) was plotted versus time (t) by nonlinear regression (Y=Y_max*(1−exp(−k_obs*t))) to determine the pseudo-first order rate constant (k_obs). The apparent inactivation rate (K_i) and second order rate constant (ki/Ki) were determined from the equation k_obs = k_i[C]/(K_i + [C]) which solved by nonlinear regression using GraphPad Prism 5. For Cat L, the Ki was determined by the same methodology described for CatB except buffer 2, Z-Phe-Arg-AMC ([S] =100 μM), different conjugate concentrations (1 nM, 5 nM, 12.5 nM, 25 nM, 50 nM, 100 nM, 133.3 nM, and 500 nM) were applied. All measurements were in biological triplicate.

4.7. In vitro NTSR1 receptor competitive binding

The IC₅₀ values for the conjugates binding to the NTSR1 were determined using the HT-29 human colon cancer cell line. In these studies, ¹⁷⁷Lu-N1 synthesized according to our previous publication,[13] served as the competitive radioligand for comparing the relative binding affinities of the conjugates. HT-29 cells (~1 × 10⁶) suspended in 100 μL of McCoy’s 5A medium (pH 7.4, 4.8 mg/mL HEPES, and 2 mg/mL BSA) were incubated with ¹⁷⁷Lu-N1 (100,000 cpm, 100 μL) at 37 °C for 45 min in the presence of the conjugates with predetermined concentrations (5 nM - 1 μM) in 100 μL of medium. At the end of the incubation, the cells were centrifuged, aspirated, and washed with fresh medium (5 × 500 pL). The cell-associated radioactivity was measured using a gamma counter and the IC₅₀ values determined by nonlinear regression using GraphPad Prism 5. All measurements were performed in biological triplicate.

4.8. Cell internalization

HT-29 cells (~1 × 10⁶ cells in 1 mL of culture medium) were incubated with the ¹⁷⁷Lu-labeled conjugate (100,000 cpm) in a 6-well plate at 37 °C for 4 h. At 15, 30, 60, 120 and 240 min time points, the culture medium was removed and the cells were washed with fresh medium (2 × 1 mL). The surface-bound fraction of radioactivity was removed by washing the cells twice with an acidic buffer (1 mL, 50 mM glycine-HCl/0.1 M NaCl buffer, pH 2.8). The cells were then lysed by 1% SDS solution and the amount of radioactivity remaining in the lysate was assigned as the internalized fraction. The radioactivity for each fraction was measured by a gamma counter. The cellular uptake of the radioconjugates was presented as a percentage of the surface-bound and internalized radioactivity relative to the total activity added.

4.9. Cell efflux

The radioconjugate (100,000 cpm/100 μL) was added to a 6-well plate containing HT-29 cells (~1 × 10⁶) with 1 mL of cell culture medium and incubated for 2 h at 37 °C. After the removal of the culture medium, the cells were washed with fresh medium (2 × 1 mL) followed by the addition of 1 mL of fresh medium for the efflux assay. At 0, 1, 2, 4, 8, and 24 h time interval, fresh medium (1 mL) was added to the tube to replace the old medium which was harvested for quantitative analysis of the effluxed radioactivity using a gamma counter. The cells were lysed with a 10% aqueous SDS solution at 24 h to quantify the remaining internalized radioactivity. The effluxed fraction is expressed as a percentage of the total internalized radioactivity which was the sum of the effluxed and internalized fractions obtained from the study.

4.10. Confocal studies

4.10.1. The uptake and cell trafficking studies

HT-29 cells (1.25 × 10⁵ / well in 500 μL medium) in a Lab-Tek chambered #1.0 borosilicate coverglass disk (four-well) were incubated with the ^natEu labeled conjugate (5 μM) in the presence (blocking) or absence of the NTSR1 peptide N1 (20 μM) at 37 °C for 2 h. Lysotracker-green (100 nM) was added in the media for 1 h to stain the cell endosomes/lysosomes. The cells were washed with PBS (400 μL) and fixed with formaldehyde (400 μL) prior to imaging. The images were obtained using an excitation wavelength of 405 nm (blue excitation) and 488 nm (green excitation). ImageJ software was used for quantifying the fluorescence of the ^natEu-labeled conjugates. Mean pixel intensities in each image were normalized to the total cell number by counting the number of cells.

4.10.2. The intracellular retention of ^natEu labeled conjugates

HT-29 cells (1.25 × 10⁵ / well) in Lab-Tek chambered #1.0 borosilicate coverglass disk (four-well) were incubated with the ^natEu-labeled conjugate (5 μM) in 500 μL of medium for 4 h. The cells were washed with fresh medium and cultured for up to 24 h. At 0, 2 and 22 h time points, Lysotracker-green (100 nM) was added to the cells and incubated for 2 h. The cells were washed with PBS (400 μL) and fixed with formaldehyde (400 μL) prior to imaging. The images were obtained using an excitation wavelength of 405 nm (blue excitation) and 488 nm (green excitation). ImageJ software was used to quantifying the fluorescence and co-localization efficiency. Mean pixel intensities in each image were normalized to the total cell number.

4.11. SDS-PAGE analysis of the in vitro samples

For preparation of CatB and CatL samples, the enzyme (3 nM, 10 μL) in buffer (50mM sodium acetate and 1mM EDTA, pH 5.0) was incubated with the radioconjugates (500,000 cpm, in 30 μL of binding assay buffer (5 mM Tris, 5 mM MgCl₂, and 2 mM DTT, pH = 5.5)) in the absence and presence of cysteine proteases inhibitor E-64 (10 μM, 10 μL) or the NTSR1 ligand N1 (20 μM, 10 μL) for 30 min. For preparation of cell samples, the radioconjugate (20 μCi, 0.74 MBq) in culture medium (1 mL) was incubated with HT-29 cells (1 × 10⁶ / well) seeded in 6-well plates in the absence and presence of the cysteine proteases inhibitor E-64 (10 μM, 10 μL) or the NTSR1 ligand N1 (20 μM, 10 μL). The cells were incubated at 37 °C for 4 h. The cells were washed with PBS (3 × 2 mL) prior to the fresh medium being added. At the desired time points, the cells (~3 × 10⁶ - from 3 wells) were trypsinized, combined and microcentrifuged. RIPA buffer (100 μL) containing the Halt™ protease inhibitor (100 ×, 1 μL) was added to the cell pellet and vigorously vortexed for 1 min. The suspension was incubated on ice for 15 min and centrifuged to remove the pellet. All the prepared samples (40 μL) were mixed with Novex Tris-Glycine SDS sample buffer (2×) (40 μL) and heated to 85 °C for 2 min. The mixture from each sample (20 μL) was loaded onto a Novex 16% tris-glycine gel and analyzed by SDS-PAGE at 110 V for 90 min. After shaking overnight in a shrinking buffer (50 mL, 65% methanol, and 0.5% glycerol in water) at 4 °C, the gel was dried for 6 h at room temperature and the ladders were painted with small amount of radioactivity. The gel was exposed to a phosphor plate for 72 h which was subsequently scanned by a Typhoon FLA 9500 imaging system at a 25 μm resolution to achieve the autoradiograph.

4.12. Biodistribution studies

Female SCID mice (5 weeks of age) received subcutaneous injections of HT-29 cells (5 × 10⁶ in 100 μL) suspended in Matrigel® into the flanks. When the tumor size reached 80 mm³ (two weeks after injection), the mice were randomized into groups and intravenously injected with 10 μCi (0.37 MBq) of the purified ¹⁷⁷Lu labeled conjugates via the tail vein. For the renal blocking study, Colchicine (1 mg/kg) was administrated by intraperitoneal injection 4 h prior to the injection of ¹⁷⁷Lu labeled conjugates while the D-lysine (40 mg/kg) was mixed with The mice were sacrificed and the tissues were excised at 4, 24, and 72 h post-injection time points. The blood, tumor, and other excised tissues were weighed and the radioactivity for each sample was measured using a gamma counter. The percentage injected dose per gram (ID%/g) and the radioactivity ratios between tumor and non-targeted tissues were calculated.

4.13. SDS-PAGE analysis of the in vivo samples

The radioconjugates (800 μCi (29.6 MBq) /mouse) were intravenously injected into the HT-29 tumor-bearing mice. The mice were sacrificed and the tumor, liver, and kidneys were excised at the 72 h post-injection time point. The tumor and organs were homogenized in RIPA buffer (50 mg / 100 μL) containing Halt™ protease inhibitor (100 ×, 50 mg / 1 μL) on ice and centrifuged to remove the pellet. An aliquot (20 μL) of the supernatant was mixed with Novex Tris-Glycine SDS sample buffer (2×, 20 μL) and heated to 85 °C for 2 min. The SDS-PAGE and the corresponding autoradiograph were performed according to the same procedure described above. Additionally, aliquots (100 μL) of supernatants of all the tissue samples were centrifuged with Pierce™ protein concentrators (MWCO = 10kDa) to separate the low molecular weight radioactivity. The radioactivity in each fraction was quantified using a gamma counter to calculate the percentage of the macromolecule-associated radioconjugate from the sum of the total counts.

Highlights.

• An E-64 analogue was incorporated to the NTSR1-targeting peptide as trapping agent.

• The trapping agent significantly enhanced tumor retention both in vitro and in vivo.

• The radioconjugate could substantially reduce the non-targeted tissues uptake.