Radiolabeled somatostatin receptor antagonists are preferable to agonists for in vivo peptide receptor targeting of tumors

October 31, 2006
103 (44) 16436-16441


Targeting neuroendocrine tumors expressing somatostatin receptor subtypes (sst) with radiolabeled somatostatin agonists is an established diagnostic and therapeutic approach in oncology. While agonists readily internalize into tumor cells, permitting accumulation of radioactivity, radiolabeled antagonists do not, and they have not been considered for tumor targeting. The macrocyclic chelator 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) was coupled to two potent somatostatin receptor-selective peptide antagonists [NH2-CO-c(DCys-Phe-Tyr-DAgl8(Me,2-naphthoyl)-Lys-Thr-Phe-Cys)-OH (sst3-ODN-8) and a sst2-selective antagonist (sst2-ANT)], for labeling with 111/natIn. 111/natIn-DOTA-sst3-ODN-8 and 111/natIn-DOTA–[4-NO2-Phe-c(DCys-Tyr-DTrp-Lys-Thr-Cys)-DTyr-NH2] (111/natIn-DOTA-sst2-ANT) showed high sst3- and sst2-binding affinity, respectively. They did not trigger sst3 or sst2 internalization but prevented agonist-stimulated internalization. 111In-DOTA-sst3-ODN-8 and 111In-DOTA-sst2-ANT were injected intravenously into mice bearing sst3- and sst2-expressing tumors, and their biodistribution was monitored. In the sst3-expressing tumors, strong accumulation of 111In-DOTA-sst3-ODN-8 was observed, peaking at 1 h with 60% injected radioactivity per gram of tissue and remaining at a high level for >72 h. Excess of sst3-ODN-8 blocked uptake. As a control, the potent agonist 111In-DOTA–[1-Nal3]-octreotide, with strong sst3-binding and internalization properties showed a much lower and shorter-lasting uptake in sst3-expressing tumors. Similarly, 111In-DOTA-sst2-ANT was injected into mice bearing sst2-expressing tumors. Tumor uptake was considerably higher than with the highly potent sst2-selective agonist 111In-diethylenetriaminepentaacetic acid–[Tyr3,Thr8]-octreotide (111In-DTPA-TATE). Scatchard plots showed that antagonists labeled many more sites than agonists. Somatostatin antagonist radiotracers therefore are preferable over agonists for the in vivo targeting of sst3- or sst2-expressing tumors. Antagonist radioligands for other peptide receptors need to be evaluated in nuclear oncology as a result of this paradigm shift.
Peptide receptor targeting in vivo is a successful method to image and treat various types of cancers (1). The best example is somatostatin receptor targeting with 111In-, 90Y-, or 177Lu-labeled somatostatin radioligands that are injected into the patients intravenously and accumulate in their somatostatin receptor-expressing tumors. For this purpose, agonists have been selected. The rationale is that agonists, after high-affinity binding to the receptor, usually trigger internalization of the ligand–receptor complex (2). This process of internalization is the basis for an efficient accumulation of the radioligand in a cell over time (1, 3–5), and it has been considered a crucial step in the process of in vivo receptor targeting with radiolabeled peptides (46). Recently, a highly significant correlation between the rate of ligand internalization in vitro into AR42J cells expressing somatostatin receptor subtype 2 (sst2) and the in vivo uptake in the sst2-expressing rat tumor model has been reported (7). Therefore, when novel analogs are being designed for receptor targeting, their internalization properties are particularly thoroughly investigated (3).
Curiously, not much is known about the usefulness, for in vivo targeting of cancer, of high binding-affinity compounds lacking the ability to trigger receptor internalization. In this respect, little is known about antagonists, which, with a few exceptions (811), do not internalize (8, 12, 13), and one could therefore expect them not to be of particular interest as radioligands for receptor targeting. However, antagonists may have characteristics other than those related to internalization that may make their radiolabeled derivatives suitable tools for in vivo receptor targeting. Most relevant is the in vitro evidence that, in certain circumstances, antagonist radioligands may label a higher number of receptor-binding sites than agonist radioligands (14, 15).
The aim of the present study was to investigate to which extent somatostatin antagonist and agonist radioligands, with similar binding affinities for somatostatin receptors, differ in their in vivo tumor-targeting properties. The best clinically established system for in vivo tumor targeting with radiolabeled peptides (1) is based on the somatostatin receptor, and a particularly large number of excellent radioligands have been developed for that purpose, all derived from somatostatin agonists (16). The first part of the present study deals with somatostatin receptor subtype 3 (sst3). First, sst3 is characterized by very efficient internalization properties (17). Second, recently, sst3-selective antagonists with high binding affinity but without triggering receptor internalization have been described (18). Their radiolabeled derivatives may be used as antagonist radioligands in case the high affinity-binding and antagonistic properties are retained after conjugation with a chelator [e.g., 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA)] and 111In-complexation. Third, well characterized radiolabeled agonists, which can label sst3 receptors in vitro and in vivo, have recently been described (1921) and can be used as reference compounds in parallel experiments. Therefore, we have coupled the chelator DOTA to the sst3 antagonist NH2-CO-c(DCys-Phe-Tyr-DAgl8(Me,2-naphthoyl)-Lys-Thr-Phe-Cys)-OH (sst3-ODN-8), labeled the conjugate with nonradioactive natIn, and tested natIn-DOTA-sst3-ODN-8 for in vitro binding and signaling properties to establish whether it is suitable to be used for in vivo receptor targeting. We then compared the in vivo biodistribution of the 111In-labeled antagonist 111In-DOTA-sst3-ODN-8 with that of a similarly potent and well established agonist-radioligand 111In-DOTA-[1-Nal3]-octreotide (111In-DOTA-NOC) in an sst3 tumor-bearing nude mouse model. The properties of each compound to target normal and neoplastic tissue in vivo have been assessed quantitatively.
In a second part, we have performed comparable studies with the sst2 receptor system, to generalize the sst3-related observations. By using the same strategy, we have developed an natIn- or 111In-labeled sst2 antagonist 111/natIn-DOTA-[4-NO2-Phe-c(DCys-Tyr-DTrp-Lys-Thr-Cys)-DTyr-NH2] (111/natIn-DOTA-sst2-ANT), characterized it in in vitro experiments, and compared its in vivo sst2 tumor-targeting properties to that of the highly potent sst2 agonist 111In- diethylenetriaminepentaacetic acid –[Tyr3,Thr8]-octreotide (111In-DTPA-TATE) (ref. 22) in mice.


Table 1 summarizes the binding data of the sst3 antagonist (sst3-ODN-8) and its DOTA analog with or without natIn complexation at all five sst. For comparison, the values of the natural somatostatin-28 (SS-28) as well as that of a potent sst3 agonist, natIn-DOTA-NOC, are shown as references. sst3-ODN-8 and its derivatives show high selectivity and binding affinity for sst3. The reference agonist natIn-DOTA-NOC has comparable sst3-binding affinity, whereas the sst2-selective analog (natIn-DTPA- TATE), used in its 111In-labeled form as a negative control for sst3-expressing tissues in the biodistribution assays, shows high sst2 but no sst3 affinity (Table 1).
Table 1.
In vitro binding, signaling, and internalization properties of somatostatin analogs
 Binding affinity*SignalingInternalization
    sst3-ODN-8>1,000>1,0008.6 ± 1.87>1,000>1,000Antagonist (sst3)No internalization (sst3)
    DOTA-sst3-ODN-8>1,000>1,0005.2 ± 1.3>1,000>1,000Antagonist (sst3)No internalization (sst3)
    natIn-DOTA-sst3-ODN-8>1,000>1,00015 ± 5.2>1,000>1,000Antagonist (sst3)No internalization (sst3)
    sst2-ANT>1,0003.6 ± 0.4>1,000349 ± 30276 ± 119Antagonist (sst2)No internalization (sst2)
    DOTA-sst2-ANT>1,0001.5 ± 0.4>1,000287 ± 27>1,000Antagonist (sst2)No internalization (sst2)
    natIn-DOTA-sst2-ANT>1,0009.4 ± 0.4>1,000380 ± 57>1,000Antagonist (sst2)No internalization (sst2)
    SS-283.2 ± 0.22.3 ± 0.13.7 ± 0.32.6 ± 0.12.4 ± 0.2Agonist (all sst)Internalization (sst2,3,5)
    natIn-DOTA-NOC>1,0002.9 ± 0.311 ± 3.2503 ± 2229.4 ± 3.7Agonist (sst2,3)Internalization (sst2,3)
    natIn-DTPA-TATE>1,0001.3 ± 0.2>1,000>1,000>1,000Agonist (sst2)Internalization (sst2)
*Values represent IC50 in nM; mean ± SEM n ≥ 3.
Tested with cAMP assay in sst-transfected cells.
Tested as DTPA-TATE.
The compounds were evaluated for their effect on forskolin-stimulated cAMP accumulation in CCL39 cells stably expressing sst3 (Table 1). SS-28 and natIn-DOTA-NOC, used as controls, act as agonists; they potently inhibit forskolin-stimulated cAMP accumulation by >77% and 58%, respectively, at a peptide concentration of 100 nM. sst3-ODN-8 and its two derivatives given alone do not inhibit forskolin-stimulated cAMP accumulation up to 10 μM. However, the agonistic effect of SS-28 can be competitively antagonized with a fixed concentration of 1 μM each of the sst3-ODN-8 derivatives applied individually. Fig. 1B illustrates the antagonistic properties of the sst3-ODN-8 derivatives.
Fig. 1.
In vitro characteristics of somatostatin analogs. (A) Structure of the sst3 antagonist DOTA-sst3-ODN-8. (B) Effects of somatostatin analogs on forskolin-stimulated cAMP accumulation in CCL-sst3 cells. Concentration-response curves with increasing concentrations of SS-28 (○), sst3-ODN-8 (▴), DOTA-sst3-ODN-8 (▾), or natIn-DOTA-sst3-ODN-8 (♦), and of increasing concentrations of SS-28 in the presence of 10−6 M DOTA-sst3-ODN-8 (□) or 10−6 M natIn-DOTA-sst3-ODN-8 (▵). Data are expressed as a percentage of the 10 μM forskolin response. SS-28 inhibits forskolin-stimulated cAMP formation in CCL-sst3 cells, whereas the sst3-ODN-8 derivatives alone have no effect; however, they reverse the SS-28-induced effect on cAMP. (C) Effect of soma tostatin analogs on sst3 internalization detected by immunofluorescence in HEK-sst3 cells. Control experiment showing membrane-bound sst3 with no peptide (a); 100 nM of the agonists SS-28 (b), or natIn-DOTA-NOC (c) trigger sst3internalization. The antagonist DOTA-sst3-ODN-8 (d) or its natIn-derivative (e) at 10 μM are not able to induce internalization. Internalization triggered by 100 nM of SS-28 is abolished by 10 μM natIn-DOTA-sst3-ODN-8 (f). (D) Scatchard plots from saturation-binding experiments on HEK-sst3 cells show a higher Bmax for 111In-DOTA-sst3-ODN-8 than for 111In-DOTA-NOC.
The antagonistic property of DOTA-sst3-ODN-8 and its derivatives was also confirmed in an immunofluorescence internalization assay (3) with HEK293 cells stably expressing sst3 (Table 1). Fig. 1C illustrates that, although the control agonists SS-28 and natIn-DOTA-NOC can induce sst3 internalization, the DOTA-sst3-ODN-8 analogs have no effect when given alone, even at a concentration of 10 μM. Moreover, they prevent sst3 internalization induced by SS-28 (Fig. 1C) or by the agonist natIn-DOTA-NOC.
Furthermore, Fig. 1D shows a Scatchard analysis in HEK-sst3 cells comparing the Bmax for the agonist 111In-DOTA-NOC (68 ± 9 pM) with that of the antagonist 111In-DOTA-sst3-ODN-8 (5,180 ± 70 pM). The antagonist labels 76 times more sst3 sites in cultured HEK-sst3 cells than the agonist.
In Tables 2 and 3, the in vivo biodistribution of the antagonist111In-DOTA-sst3-ODN-8 is reported in nude mice bearing the sst3-expressing tumor (Table 2) and compared with that of the agonist 111In-DOTA-NOC (Table 3). Of note is the accumulation of the radiolabeled antagonist in the tumor, which peaks at 1 h [>60% injected activity per gram of tissue (IA/g) uptake] and remains very high at 4 h (50%IA/g), at 24 h (>30%IA/g) and even at 72 h (>10%IA/g). The pituitary, an organ known to express various sst, including sst3 (17), is also labeled with the antagonist 111In-DOTA-sst3-ODN-8 (Table 2). A blocking experiment performed in separate mice by adding 1,000 times excess of sst3-ODN-8 together with the radioligand documents that the majority of the labeling in the sst3-expressing tumor and pituitary represents binding to specific somatostatin receptors. The blocking agent did not affect radioactivity uptake in the kidneys and blood, indicating that this uptake is not receptor-mediated (Table 2). For comparison, the accumulation of radioactivity in the sst3-expressing tumors by using the agonist 111In-DOTA-NOC is considerably less than by using the antagonist 111In-DOTA-sst3-ODN-8 and amounts to ≈7%IA/g at 4 h and decreases at 24 h (Table 3). The calculated tumor/tissue ratios, representing important parameters to evaluate the quality of a targeting agent (16), are considerably higher with 111In-DOTA-sst3-ODN-8 than with 111In-DOTA-NOC, especially for the tumor/blood and tumor/muscle ratios at 4 and 24 h. Established sst2-expressing organs, such as the stomach, adrenals, or pancreas, show a blockable accumulation of 111In-DOTA-NOC (Table 3), whereas they show no significant accumulation of 111In-DOTA-sst3-ODN-8 (Table 2). As a negative control, labeling of the sst3-expressing tumor with a specific sst2 agonist radioligand, 111In-DTPA-TATE (22), is found to be negligible (Table 3). The high uptake of 111In-DOTA-NOC and 111In-DTPA-TATE in the pituitary reflects the high expression of sst2 in this organ. Fig. 2 shows scans taken at 30 min and 4 h of animals bearing sst3-expressing tumors in one flank and, as control, sst2-expressing tumors in the other flank. Mice injected with the sst3 antagonist 111In-DOTA-sst3-ODN-8 showed a massive uptake in the sst3- but not the sst2-expressing tumor, whereas those injected with the sst2/sst3 agonist 111In-DOTA-NOC showed a weak uptake in both tumors.
Fig. 2.
In vivo scans taken 30 min and 4 h after injection of 111In-DOTA-sst3-ODN-8 or 111In-DOTA-NOC. Each mouse was bearing two tumors, an sst3-expressing tumor and, as control, an sst2-expressing tumor. Mice were placed directly on one head of a two-headed gamma-camera system (PRISM 2000, Philips, Eindhoven, The Netherlands) equipped with medium-energy collimators (acquisition time, 10 min per time point). Strong uptake is seen with the antagonist radioligand in the sst3-expressing tumor exclusively, whereas weak uptake was found in both tumors with the sst2/sst3 agonist.
Table 2.
Biodistribution in HEK-sst3 tumor-bearing nude mice at 0.25, 0.5, 1, 4, 24, or 72 h after injection of 111In-DOTA-sst3-ODN-8
Organ0.25 h0.25 h, blocked*0.5 h1 h4 h24 h72 h
Blood10.4 ± 1.39.2 ± 0.55.3 ± 0.21.8 ± 0.80.1 ± 0.000.03 ± 0.010.01 ± 0.00
Stomach2.7 ± 0.13.1 ± 0.42.0 ± 0.61.1 ± 0.60.2 ± 0.000.23 ± 0.130.12 ± 0.00
Kidney20.3 ± 1.121.7 ± 4.917.8 ± 1.815.8 ± 3.114.1 ± 2.96.8 ± 1.23.5 ± 0.3
Bowel2.4 ± 0.092.2 ± 0.21.4 ± 0.10.7 ± 0.30.17 ± 0.030.08 ± 0.010.07 ± 0.00
Pancreas1.8 ± 0.11.3 ± 0.051.1 ± 0.10.5 ± 0.30.16 ± 0.000.08 ± 0.010.06 ± 0.00
Spleen2.3 ± 0.082.4 ± 0.21.3 ± 0.080.8 ± 0.30.34 ± 0.040.17 ± 0.030.15 ± 0.01
Liver3.8 ± 0.83.8 ± 0.042.1 ± 0.31.3 ± 0.40.63 ± 0.020.41 ± 0.210.18 ± 0.00
Heart1.3 ± 0.081.4 ± 0.040.9 ± 0.050.6 ± 0.10.22 ± 0.050.15 ± 0.010.08 ± 0.00
sst3 tumor22.1 ± 3.59.3 ± 0.234.8 ± 1.461.3 ± 10.149.7 ± 11.830.8 ± 5.010.9 ± 1.9
Muscle3.6 ± 0.43.6 ± 0.12.0 ± 0.10.9 ± 0.60.2 ± 0.040.09 ± 0.0040.03 ± 0.00
Adrenal4.8 ± 0.63.9 ± 0.23.6 ± 0.21.7 ± 1.01.0 ± 0.20.68 ± 0.120.61 ± 0.08
Bone2.1 ± 0.12.2 ± 0.11.2 ± 0.10.6 ± 0.20.2 ± 0.010.2 ± 0.040.22 ± 0.05
Pituitary24.4 ± 0.77.4 ± 1.814.1 ± 1.05.3 ± 0.63.6 ± 0.21.7 ± 0.31.11 ± 0.2
Tumor/tissue ratios
    Tumor/blood2.1 ± 0.33 6.5 ± 0.2634.0 ± 5.6497.0 ± 981026 ± 1661090 ± 190
    Tumor/kidney1.08 ± 0.1 1.95 ± 0.073.9 ± 0.63.5 ± 0.84.5 ± 0.73.1 ± 0.5
    Tumor/muscle6.1 ± 0.9 17.4 ± 0.768.1 ± 11248 ± 59342 ± 55363 ± 63
The results are expressed as the percentage of the %IA/g, mean ± SEM, n ≥ 3. Bold text indicates the tumor as the most important of the listed tissues.
*Blocked with excess sst3-ODN-8 coinjected with the radioligand.
P < 0.001.
Table 3.
Biodistribution in HEK-sst3 tumor-bearing nude mice at 0.5, 4, or 24 h after injection of 111In-DOTA-NOC and at 4 h after 111In-DTPA-TATE
0.5 h4 h4 h, blocked*24 h4 h
Blood2.6 ± 0.30.3 ± 0.020.5 ± 0.030.1 ± 0.000.1 ± 0.01
Stomach7.4 ± 2.03.4 ± 0.40.46 ± 0.011.7 ± 0.45.2 ± 0.2
Kidney11.7 ± 1.314.2 ± 0.817.9 ± 1.19.0 ± 1.217.7 ± 1.5
Bowel1.9 ± 0.10.9 ± 0.10.3 ± 0.050.6 ± 0.11.1 ± 0.1
Pancreas13.4 ± 3.42.7 ± 0.20.2 ± 0.021.5 ± 0.14.8 ± 0.4
Spleen1.7 ± 0.070.5 ± 0.032.2 ± 0.80.5 ± 0.090.5 ± 0.04
Liver1.8 ± 0.10.9 ± 0.071.5 ± 0.70.6 ± 0.050.2 ± 0.04
Heart1.5 ± 0.20.3 ± 0.030.3 ± 0.010.1 ± 0.020.2 ± 0.00
sst3 tumor17.5 ± 4.36.5 ± 0.74.08 ± 0.223.5 ± 0.20.3 ± 0.02
Muscle1.4 ± 0.20.2 ± 0.010.2 ± 0.010.1 ± 0.010.2 ± 0.01
Adrenal6.5 ± 1.34.7 ± 0.60.8 ± 0.052.6 ± 0.24.6 ± 0.2
Bone2.6 ± 0.40.7 ± 0.080.4 ± 0.070.8 ± 0.10.7 ± 0.05
Pituitary6.1 ± 0.58.3 ± 3.33.3 ± 0.85.5 ± 0.418.7 ± 1.2
Tumor/tissue ratios
    Tumor/blood6.7 ± 1.621.6 ± 2.3 35 ± 2 
    Tumor/kidney1.5 ± 0.30.45 ± 0.04 0.4 ± 0.01 
    Tumor/muscle12.5 ± 2.832.5 ± 3.5 35 ± 2 
The results are expressed as percentage of the %IA/g, mean ± SEM, n ≥ 3. Bold text indicates the tumor as the most important of the listed tissues.
*Blocked with excess natIn-DOTA-NOC coinjected with the radioligand.
P < 0.001.
Similarly, we investigated the sst2 receptor system with sst2-transfected HEK293 and CCL39 cells in vitro or HEK293 cells transplanted in animals, to expand the above sst3 study. We chose to conjugate the established sst2-ANT (23) with DOTA, namely DOTA-[4-NO2-Phe-c(DCys-Tyr-DTrp-Lys-Thr-Cys)-DTyr-NH2] (DOTA-sst2-ANT). sst2-ANT and unlabeled or In-labeled DOTA-sst2-ANT have high binding affinity and selectivity for sst2 (Table 1). Both the In-labeled and unlabeled DOTA-sst2-ANT are sst2-antagonists in the cAMP assay; although inactive at 10 μM concentration in the absence of SS-28, they reverse completely the 100 nM SS-28-induced inhibition of cAMP formation triggered by 30 μM forskolin in CCL-sst2 cells. Further, they also act as antagonists in the immunofluorescence internalization assay, because they do not induce internalization of the sst2 receptor up to 10,000 nM but inhibit completely at 1,000 nM the internalization induced by potent somatostatin agonists such as DTPA-TATE (Fig. 3). The tissue biodistribution after 111In-DOTA-sst2-ANT injection in mice reveals a high but also long-lasting tumor uptake in sst2 tumor-bearing animals (Table 4). It is impressive to see that 4- and 24-h values of tumor uptake are twice the ones obtained with the highly potent sst2 agonist 111In-DTPA-TATE given under the same conditions. As proof of specificity, uptake in the tumor is massively blocked by excess DOTA-sst2-ANT (Table 4). Moreover, Scatchard analysis in HEK-sst2 cells identifies more sites labeled with the radiolabeled antagonist (Bmax = 354 ± 14 pM) than with the agonist (Bmax = 23 ± 1.0 pM).
Fig. 3.
Effect of somatostatin analogs on sst2 internalization detected by immunofluorescence with R2–88 in HEK-sst2 cells. Compared with control with no peptide (a), 100 nM DTPA-TATE (b) triggers strong sst2 internalization. natIn-DOTA-sst2-ANT (10 μM) (c) does not induce sst2 internalization but abolishes internalization induced by 100 nM DTPA-TATE (d).
Table 4.
Biodistribution in HEK-sst2 tumor bearing nude mice after injection of 111In-DOTA-sst2-ANT or 111In-DTPA-TATE
0.5 h4 h4 h, blocked*24 h0.5 h4 h24 h
Blood2.76 ± 0.190.14 ± 0.030.13 ± 0.010.05 ± 0.010.99 ± 0.250.13 ± 0.10.06 ± 0.01
Stomach7.82 ± 2.030.61 ± 0.180.19 ± 0.070.25 ± 0.0613.93 ± 8.167.04 ± 2.024.86 ± 1.68
Kidney22.92 ± 2.6210.5 ± 1.09.67 ± 1.387.38 ± 0.0929.25 ± 7.711.44 ± 0.867.08 ± 1.21
Bowel1.72 ± 0.250.16 ± 0.030.15 ± 0.030.08 ± 0.031.7 ± 0.530.97 ± 0.310.62 ± 0.16
Pancreas24.16 ± 6.580.71 ± 0.210.09 ± 0.020.13 ± 0.0218.18 ± 12.596.06 ± 3.262.26 ± 0.09
Spleen1.67 ± 0.230.23 ± 0.040.21 ± 0.020.15 ± 0.021.13 ± 0.320.39 ± 0.050.21 ± 0.05
Liver1.74 ± 0.180.43 ± 0.070.49 ± 0.030.32 ± 0.020.46 ± 0.070.16 ± 0.040.17 ± 0.03
Heart1.23 ± 0.050.11 ± 0.030.08 ± 0.010.04 ± 0.00.53 ± 0.210.18 ± 0.070.1 ± 0.01
sst2tumor22.33 ± 3.2729.12 ± 3.93.62 ± 0.2622.84 ± 0.418.36 ± 4.3715.83 ± 3.9412.3 ± 1.32
Muscle0.97 ± 0.360.11 ± 0.020.09 ± 0.020.06 ± 0.030.51 ± 0.150.09 ± 0.030.05 ± 0.01
Adrenal4.74 ± 3.00.49 ± 0.120.24 ± 0.040.46 ± 0.264.68 ± 1.461.95 ± 0.262.28 ± 0.67
Bone1.84 ± 0.381.29 ± 0.750.58 ± 0.220.48 ± 0.140.8 ± 0.270.87 ± 0.650.74 ± 0.37
Pituitary27.7 ± 6.4820.23 ± 6.383.14 ± 0.912.08 ± 1.7221.99 ± 5.3612.1 ± 5.016.99 ± 4.61
The results are expressed in percent of the %IA/g, mean ± SEM, n ≥ 3. Bold text indicates the tumor as the most important of the listed tissues.
*Blocked with excess DOTA-sst2-ANT coinjected with the radioligand.


Agonists have been used exclusively as radioligands in the past decade for the development and implementation of peptide receptor targeting of tumors in vivo, because such radioligands are readily internalized together with the receptor, permitting an active accumulation of radioactivity in the tumor cells (1, 3–5). The present study may build the foundation for a change of paradigm in this respect. Indeed, it shows, unexpectedly, that adequately labeled sst2 and sst3 antagonists, even though they do not internalize, may be useful radioligands to target tumors in vivo. More importantly, it also shows that antagonists may be even better candidates to target tumors than agonists with comparable binding characteristics.
In the in vivo model of an sst3-expressing tumor, we have compared the biodistribution of the sst3 antagonist, 111In-DOTA-sst3-ODN-8, with that of an established agonist, 111In-DOTA-NOC (1921). natIn-DOTA-NOC has high binding affinity to sst3 receptors in vitro (19) and efficiently triggers the sst3 internalization into HEK-sst3 cells. The DOTA-linked sst3-ODN-8 antagonist, with or without indium, also shows a comparably high sst3-binding affinity but, differently from natIn-DOTA-NOC, it cannot trigger sst3 internalization into HEK-sst3 cells. Therefore, although one can expect to see a specific in vivo uptake of 111In-DOTA-NOC in sst3-expressing tumors, it is unexpected, based on current knowledge (1, 3–5), to see such a high uptake with 111In-DOTA-sst3-ODN-8. The uptake in the tumors and in the sst3-expressing pituitary is specific in both cases, because it can be specifically blocked by the corresponding cold peptide, indicating a somatostatin receptor-mediated process.
One of the most impressive findings is that the amount of uptake of the antagonist radioligand is particularly high in these tumors: 60%IA/g uptake has indeed never been achieved by any somatostatin receptor agonist ligand, not even by those developed most recently (19, 21). Not only is the uptake at the peak time point very high, but also the long-lasting accumulation of the antagonist radioligand up to 72 h after injection is a remarkable result and represents a considerable advantage over labeling with established agonists. Of crucial importance for potential clinical use are the high tumor/tissue ratios obtained with the radiolabeled antagonist.
The same observation of a much better labeling is obtained in HEK-sst2 tumors with the sst2 antagonist 111In-DOTA-sst2-ANT. Knowing of the outstanding targeting abilities of 111In-DTPA-TATE (22), it is striking to see that the in vivo labeling at 4 and 24 h for the sst2 antagonist is twice as high, despite the fact that the antagonist is not internalized into the tumor cells and that its sst2-binding affinity is lower than for the agonist.
Explanations for these excellent in vivo targeting properties of antagonists may be found, at least in part, in previous in vitro studies dealing with other G protein-coupled receptors. Indeed, a higher number of 5-HT2A (15) or corticotropin releasing factor (CRF) receptors (14) were reported to be labeled in vitro with radiolabeled antagonists than with agonists, probably reflecting a difference in the receptor interaction with the G proteins (14). Similar conclusions can be drawn for sst2 and sst3 receptors labeled with antagonists, as shown in our Scatchard data. It appears, therefore, that, in an in vivo situation, an agonist that triggers a strong internalization but binds to a limited number of high-affinity receptors is a less-efficient targeting agent than an antagonist lacking internalization capabilities but binding to a larger variety of receptor conformations.
The long-lasting in vivo labeling of antagonists, in particular 111In-DOTA-sst3-ODN-8, may be brought in association to earlier findings by Wynn et al. (13) with gonadotropin-releasing hormone (GnRH) analogs. Although in vitro quantitative autoradiography did not detect any significant internalization at the electron-microscopic level of a GnRH antagonist during in vitro incubation for up to 120 min, in vivo cellular accumulation and binding of this antagonist to the pituitary GnRH receptor increased slowly during the 10 h after i.v. injection (13). These data were interpreted as reflecting the slow dissociation of the bound antagonist, with a persistence of specific binding up to 8 days after injection (13) but possibly also a delayed and slow internalization of the antagonist occurring 4–36 h after injection (13). Not to be ignored in this regard is that antagonists are also more chemically stable and more hydrophobic than agonists, resulting in longer duration of action and possible stabilization in the lipid-rich environment of the receptors.
What are the consequences of the present data for in vivo targeting of human cancers? First, after safety and efficacy evaluation in humans, radiolabeled DOTA-sst3-ODN-8 should be evaluated in patients with sst3-expressing tumors, such as inactive pituitary adenomas or pheochromocytomas (24, 25). Radiolabeled DOTA-sst2-ANT should be studied in patients with sst2-expressing tumors, consisting of a majority of neuroendocrine tumors. Then, it will be necessary to know whether the proposed change of paradigm can be generalized to other peptide receptor systems that are currently involved in peptide receptor targeting, including cholecystokinin, gastrin-releasing peptide, vasoactive intestinal peptide, neurotensin, and neuropeptide Y receptors (1). The answer to this question represents a long-term project for the nuclear oncology/molecular imaging community, requiring adequate tumor models, in vitro testing methods, and most importantly, adequate and potent antagonists as radioligands for the respective receptors. For the moment, peptidic antagonists are not available for every peptide receptor of interest and therefore need to be developed. The present results with 111In-DOTA-sst3-ODN-8 and 111In-DOTA-sst2-ANT, two radiolabeled peptide antagonists ready for nuclear medicine investigations, indicate the importance of developing such antagonist peptides. If the present observation can be confirmed for other receptors, the use of potent radiolabeled antagonists for in vivo tumor targeting may considerably improve the sensitivity of diagnostic procedures and the efficacy of receptor-mediated radiotherapy.

Materials and Methods


Peptides were synthesized as described (18, 19). Radioactive or nonradioactive metals were chelated to the DOTA-coupled sst3-ODN-8, DOTA-NOC, DOTA-sst2-ANT, and DTPA-TATE, as described (21). The structure of DOTA-sst3-ODN-8 is shown in Fig. 1A.

Cell Culture.

The HEK293 cell lines (HEK-sst2, HEK-sst3) and the CCL39 cell lines (CCL-sst2, CCL- sst3) stably expressing the human sst2 or sst3 were grown as described (3, 26).

sst1, sst2, sst3, sst4, and sst5 Receptor Binding in Vitro.

The sst1-, sst2-, sst3-, sst4-, and sst5-binding affinities of various compounds listed in Table 1 were measured by using in vitro receptor autoradiography, as described (26).
Saturation-binding experiments for 111/natIn-DOTA-sst2-ANT and 111/natDTPA-TATE or 111/natIn-DOTA-sst3-ODN-8 and 111/natIn-DOTA-NOC were performed on HEK-sst2 or -sst3 cells, respectively, at 4°C by using increasing concentrations of the 111/natIn-labeled peptide ranging from 0.1 to 1,000 nM, as described (27). One micromolar cold peptide was used to quantify nonspecific binding. Bmax was calculated for each radioligand from Scatchard plotting of the obtained data by using Origin 5.0 software (Microcal Software, Northampton, MA).

Adenylate Cyclase Activity.

Forskolin-stimulated cAMP accumulation was determined in CCL-sst2 and -sst3 cells by using a commercially available cAMP scintillation proximity assay, as described (18).

sst2 and sst3 Receptor Internalization.

Immunofluorescence microscopy-based internalization assay for sst2 and sst3 was performed as described (3). For sst3 immunocytochemistry, the commercially available sst3-specific antibody SS-850 (Gramsch Laboratories, Schwabenhausen, Germany) was used (3). For sst2 immunocytochemistry, the sst2-specific antibody SS-800 (Gramsch Laboratories) or the R2–88 antibody (A. Schonbrunn, Universtiy of Texas Health Science Center, Houston, TX) were used, with equivalent results, as reported (3, 28). HEK-sst2 and -sst3 cells were treated with the sst2 or sst3 agonists, respectively, at a concentration of 100 nM, or with agonists at a concentration of 100 nM in the presence of an excess of antagonists (100 times the concentration of the agonist) or with antagonists alone at a concentration of 10 μM and processed for immunofluorescence microscopy (3).

HEK-sst2 and -sst3 Cell Implantation in Nude Mice.

Animals were kept, treated, and cared for in compliance with the guidelines of the Swiss regulations (approval 789). Athymic female nude mice were implanted s.c. with 10–12 million HEK-sst2 and -sst3 cells, respectively, freshly suspended in sterile PBS. Ten to fourteen days after inoculation, the mice showed solid palpable tumor masses (tumor weights, 60–150 mg) and were used for the in vivo biodistribution experiments.
Confirmation that the transfected tumors were indeed expressing solely sst2 or sst3, respectively, was obtained in resected tumor samples tested in vitro with somatostatin receptor autoradiography by using subtype selective ligands (24).

In Vivo Biodistribution of 111In-Labeled Antagonists and Agonists.

Mice were injected into a tail vein with 10 pmol 111In-radiolabeled peptide (≈0.15–0.2 MBq) in 0.1 ml of NaCl solution (0.9%, with 0.1% BSA). To determine the nonspecific uptake of the radiolabeled peptides, mice were injected with 49 nmol sst3-ODN-8 or natIn-DOTA-NOC (sst3 study) or 20 nmol DOTA-sst2-ANT (sst2 study) in 0.05 ml of NaCl solution (0.9%) as a coinjection with the radioligand.
To study the biodistribution of 111In-DOTA-sst3-ODN-8, mice were killed at 0.25 h, 0.5 h, 1 h, 4 h, 24 h, or 72 h postinjection. For the biodistribution study of 111In-DOTA-NOC or 111In-DOTA-sst2-ANT, mice were killed at 0.5 h, 4 h, or 24 h postinjection. The biodistribution of 111In-DTPA-TATE was studied at 0.5 h or 4 h after injection. The organs of interest were collected, blotted dried, and weighed; their radioactivity was measured, and the %IA/g was calculated.


somatostatin receptor subtype
1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid
diethylenetriaminepentaacetic acid
injected activity per gram of tissue
CCL39 cells stably expressing sst3
HEK293 cells stably expressing sst3


We thank V. Eltschinger, R. Kaiser, W. Low, and C. Miller for technical assistance and D. Doan and M. Kuonen for manuscript preparation. This work was supported by National Institutes of Health Grant R01 DK59953, National Science Foundation Grant 3100A0-100390, and the European Molecular Imaging Laboratories.


JC Reubi Endocr Rev 24, 389–427 (2003).
JA Koenig, JM Edwardson Trends Pharmacol Sci 18, 276–287 (1997).
R Cescato, S Schulz, B Waser, V Eltschinger, J Rivier, HJ Wester, M Culler, M Ginj, Q Liu, A Schonbrunn, et al. J Nucl Med 47, 502–511 (2006).
L Bodei, G Paganelli, G Mariani J Nucl Med 47, 375–377 (2006).
LJ Hofland, SW Lamberts Endocr Rev 24, 28–47 (2003).
G Weckbecker, I Lewis, R Albert, HA Schmid, D Hoyer, C Bruns Nat Rev Drug Discov 2, 999–1017 (2003).
D Storch, M Behe, MA Walter, J Chen, P Powell, R Mikolajczak, HR Macke J Nucl Med 46, 1561–1569 (2005).
JA Gray, BL Roth Brain Res Bull 56, 441–451 (2001).
BF Roettger, D Ghanekar, R Rao, C Toledo, J Yingling, D Pinon, LJ Miller Mol Pharmacol 51, 357–362 (1997).
N Bhowmick, P Narayan, D Puett Endocrinology 139, 3185–3192 (1998).
LH Pheng, Y Dumont, A Fournier, JG Chabot, A Beaudet, R Quirion Br J Pharmacol 139, 695–704 (2003).
Q Liu, R Cescato, DA Dewi, J Rivier, JC Reubi, A Schonbrunn Mol Pharmacol 68, 90–101 (2005).
PC Wynn, CA Suarez-Quian, GV Childs, KJ Catt Endocrinology 119, 1852–1863 (1986).
MH Perrin, SW Sutton, LA Cervini, JE Rivier, WW Vale J Pharmacol Exp Ther 288, 729–734 (1999).
AJ Sleight, NJ Stam, V Mutel, PM Vanderheyden Biochem Pharmacol 51, 71–76 (1996).
A Heppeler, S Froidevaux, AN Eberle, HR Maecke Curr Med Chem 7, 971–994 (2000).
Y Patel Front Neuroendocrinol 20, 157–198 (1999).
JC Reubi, JC Schaer, S Wenger, C Hoeger, J Erchegyi, B Waser, J Rivier Proc Natl Acad Sci USA 97, 13973–13978 (2000).
D Wild, JS Schmitt, M Ginj, HR Macke, BF Bernard, E Krenning, M De Jong, S Wenger, JC Reubi Eur J Nucl Med Mol Imaging 30, 1338–1347 (2003).
D Wild, HR Macke, B Waser, JC Reubi, M Ginj, H Rasch, J Muller-Brand, M Hofmann Eur J Nucl Med Mol Imaging 32, 724–724 (2005).
M Ginj, J Chen, MA Walter, V Eltschinger, JC Reubi, HR Maecke Clin Cancer Res 11, 1136–1145 (2005).
M de Jong, WAP Breeman, WH Bakker, PPM Kooij, BF Bernard, LJ Hofland, TJ Visser, A Srinivasan, MA Schmidt, JL Erion, et al. Cancer Res 58, 437–441 (1998).
RT Bass, BL Buckwalter, BP Patel, MH Pausch, LA Price, J Strnad, JR Hadcock Mol Pharmacol 50, 709–715 (1996).
JC Reubi, B Waser, JC Schaer, JA Laissue Eur J Nucl Med 28, 836–846 (2001).
J Mundschenk, N Unger, S Schulz, V Hollt, R Steinke, H Lehnert J Clin Endocrinol Metab 88, 5150–5157 (2003).
JC Reubi, JC Schaer, B Waser, S Wenger, A Heppeler, J Schmitt, HR Mäcke Eur J Nucl Med 27, 273–282 (2000).
M Ginj, K Hinni, S Tschumi, S Schulz, HR Maecke J Nucl Med 46, 2097–2103 (2005).
M Korner, V Eltschinger, B Waser, A Schonbrunn, JC Reubi Am J Surg Pathol 29, 1642–1651 (2005).

Information & Authors


Published in

Go to Proceedings of the National Academy of Sciences
Go to Proceedings of the National Academy of Sciences
Proceedings of the National Academy of Sciences
Vol. 103 | No. 44
October 31, 2006
PubMed: 17056720


Submission history

Received: May 6, 2006
Published online: October 31, 2006
Published in issue: October 31, 2006


  1. antagonist radioligands
  2. tumor targeting
  3. peptide hormones
  4. neuropeptides
  5. receptor internalization


We thank V. Eltschinger, R. Kaiser, W. Low, and C. Miller for technical assistance and D. Doan and M. Kuonen for manuscript preparation. This work was supported by National Institutes of Health Grant R01 DK59953, National Science Foundation Grant 3100A0-100390, and the European Molecular Imaging Laboratories.



Mihaela Ginj
Division of Radiological Chemistry, Institute of Nuclear Medicine, Department of Radiology, University Hospital Basel, Petersgraben 4, CH-4031 Basel, Switzerland;
Hanwen Zhang
Division of Radiological Chemistry, Institute of Nuclear Medicine, Department of Radiology, University Hospital Basel, Petersgraben 4, CH-4031 Basel, Switzerland;
Beatrice Waser
Division of Cell Biology and Experimental Cancer Research, Institute of Pathology, University of Berne, Murtenstrasse 31, CH-3010 Berne, Switzerland; and
Renzo Cescato
Division of Cell Biology and Experimental Cancer Research, Institute of Pathology, University of Berne, Murtenstrasse 31, CH-3010 Berne, Switzerland; and
Damian Wild
Division of Radiological Chemistry, Institute of Nuclear Medicine, Department of Radiology, University Hospital Basel, Petersgraben 4, CH-4031 Basel, Switzerland;
Xuejuan Wang
Division of Radiological Chemistry, Institute of Nuclear Medicine, Department of Radiology, University Hospital Basel, Petersgraben 4, CH-4031 Basel, Switzerland;
Judit Erchegyi
Clayton Foundation Laboratories for Peptide Biology, The Salk Institute, 10010 North Torrey Pines Road, La Jolla, CA 92037
Jean Rivier
Clayton Foundation Laboratories for Peptide Biology, The Salk Institute, 10010 North Torrey Pines Road, La Jolla, CA 92037
Helmut R. Mäcke
Division of Radiological Chemistry, Institute of Nuclear Medicine, Department of Radiology, University Hospital Basel, Petersgraben 4, CH-4031 Basel, Switzerland;
Jean Claude Reubi§ [email protected]
Division of Cell Biology and Experimental Cancer Research, Institute of Pathology, University of Berne, Murtenstrasse 31, CH-3010 Berne, Switzerland; and


To whom correspondence should be addressed. E-mail: [email protected]
Communicated by Roger Guillemin, The Salk Institute for Biological Studies, La Jolla, CA, September 5, 2006
Author contributions: M.G. and H.Z. contributed equally to this work; H.R.M., and J.C.R. designed research; M.G., H.Z., B.W., R.C., D.W., and X.W. performed research; J.E., J.R., and H.R.M. contributed new reagents/analytic tools; M.G., H.Z., R.C., J.R., H.R.M., and J.C.R. analyzed data; and J.C.R. wrote the paper.

Competing Interests

The authors declare no conflict of interest.

Metrics & Citations


Note: The article usage is presented with a three- to four-day delay and will update daily once available. Due to ths delay, usage data will not appear immediately following publication. Citation information is sourced from Crossref Cited-by service.

Citation statements



If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

Cited by


    View Options

    View options

    PDF format

    Download this article as a PDF file


    Get Access

    Login options

    Check if you have access through your login credentials or your institution to get full access on this article.

    Personal login Institutional Login

    Recommend to a librarian

    Recommend PNAS to a Librarian

    Purchase options

    Purchase this article to get full access to it.

    Single Article Purchase

    Radiolabeled somatostatin receptor antagonists are preferable to agonists for in vivo peptide receptor targeting of tumors
    Proceedings of the National Academy of Sciences
    • Vol. 103
    • No. 44
    • pp. 16059-16615







    Share article link

    Share on social media