Optimizing the Performance of 68Ga Labeled FSHR Ligand in Prostate Cancer Model by Means of Aprotinin

Purpose: Radiolabeled FSH1 peptides are potential specic probes for FSHR imaging. However, moderate uptakes and fast washout from the tumors may limit its widespread use. In this study, 68 Ga labeled modied FSH1 analogs was prepared and the imaging properties were determined in the prostate cancer model with or without aprotinin. Methods: NOTA-MAL-FSH4 was synthesized and labeled with 68 Ga. The pharmacokinetic prole of the peptide after co-administration with aprotinin was determined through metabolism analyses and microPET imaging. Results: 68 Ga-NOTA-MAL-FSH4 was successfully prepared. The IC50 value of displacement 68 Ga-NOTA-MAL-FSH4 with FSH1 was 139.4±1.16 nM. The PC-3 prostate tumor was visible after administration of the 68 Ga labeled tracer. In vitro RP-HPLC analysis revealed that the average percentage of intact peptide in the plasma, liver and tumor was 8.30, 9.57 and 7.06 % respectively. In presence of aprotinin, the amounts of intact peptide increased to 34.32%, 20.63 % and 15.39 % in the counterparts respectively. MicroPET imaging showed that the uptakes of PC-3 tumors at 60mins after co-administration of 100μg, 200μg or 400μg enzyme inhibitors were 2.91±0.21%ID/g, 3.89±0.16%ID/g and 9.21±0.22%ID/g respectively. Conclusion: With the aid of a serine protease inhibitor, the performance of the 68 Ga labeled peptide was optimized, which may benet further clinical application.


Introduction
Follicle stimulating hormone receptor (FSHR) is a glycosylated transmembrane protein and plays an important role in reproductive processes through binding to the glycoprotein, FSH. [1]It was mainly located in the testicular sertoli cells and ovarian granulose cells with low levels in mammals. [2] Recently, FSHR was observed to be overexpressed in numerous tumors such as prostate cancer, ovarian cancer, head and neck squamous cancer, etc. [3,4]On the contrary, the receptor was absent in normal and non-malignant tissues. It was found that FSHR participated in promoting the growth of tumors. [5,6]Animal studies showed that a DNA vaccine towards FSHR could e ciently delay the progression of FSHR positive tumors and increasing the immunity against the tumor. [7]Also, FSHR may play a speci c role in neoangiogenesis since the percentage of FSHR positive vessels was signi cantly higher in the patients with kidney cancer responded to the treatment with sunitinib than those in the non-responsive group.
[8]These features imply that FSHR might be an appealing target for tumor diagnosis and therapy. [9,10]Development of speci c ligands to the receptors may provide effective theranostic strategies against cancer for personalized treatment.
It was proved that the 33-53 region of FSHβ chain, FSH33-53 peptide(denoted as FSH1), owns a strong binding a nity to FSHR and becomes an attractive vector for drug delivery to enhance the antitumor effect. [11,12] For example, FSH1 peptide-conjugated gro-α siRNA-loaded polyethylene glycolpolyethylenimine nanoparticles orpaclitaxel nanoparticles can e ciently suppress the growth of ovarian cancers. [13] Besides, radiolabeled FSH1 analog might be a potential speci c probe for FSHR imaging.
Previous studies revealed that 18 FAl labeled FSH1 peptides were speci c FSHR probes and the tumors were visualized via a microPET scan. [14]To further improve the pharmacokinetic pro le, a hydrophilic linker (GGGRDN) was introduced at the N-terminals of FSH1. [15]It showed that the modi cation could decrease the abdomen radioactivity background. However, moderate uptakes and fast washout of the tracer (~ 3 %ID/g at 0.5 h post-injection and ~ 1%ID/g at 2 hours postinjection) from the tumors may partially hamper their application.
Degradation of peptides by various proteases present in the plasma and digestive system may be a possible reason. The proteases decrease the metabolic stability of the peptides and affect their performance in molecular imaging. Enzyme inhibitors were proofed to increase metabolic stability and prolong tumor localization of the peptides.
[18]Aprotinin (also named Trasylol) is a competitive serine protease inhibitor consisted of a single peptide chain with three disul de bonds. It forms stable complexes with the serine protease including trypsin, chymotrypsin, plasmin as well as other related proteolytic enzymes, and blocks the active sites of enzymes. [19]Thus, we speculated that the tumor uptake of the radiolabeled FSH1 peptides might be enhanced via coadministration of aprotinin. 68 Ga is an attractive nuclide for PET imaging due to its simple availability by eluting the 68 Ge/ 68 Ga generator every 4 hours in the same day without an onsite cyclotron. Besides, 68 Ga is easily labeled to peptides by conjugating with macrocyclic chelators, and the labeling yields were higher than those of 18 F labeled counterparts. [20][21][22][23] In this study, a FSH1 analog, FSH4, was obtained by introducing a hydrophilic linker (GGGRDN) into FSH1 at C-terminals and conjugated with malemide-NOTA. The resulting peptides, NOTA-MAL-FSH4 was labeled with 68 Ga. (Fig. 1). The properties for FSHR PET imaging with 68 Ga-NOTA-MAL-FSH4 were rstly investigated in prostate cancer models. Also, the effect of aprotinin on the tumor imaging performance of the radiolabeled peptide was further evaluated.

Materials And Methods
General FSH4 peptide (YTRDLVYKDPARPKIQKTCTFNDRGGG) and FSH1 peptide were obtained from Apeptide Co., Ltd. (Shanghai, China). Malemide-NOTA (denoted as MAL-NOTA) was purchased from CheMatech (Dijon, France). Aprotinin was bought from Sigma-Aldrich. [ 68 Ga]GaCl 3 was obtained by eluting a 68 Ge/ 68 Ga-generator (ITG, Germany) with 0.05 M hydrochloric acid. All reagents were analytical grade and employed without further puri cation. HPLC systems for analysis and puri cation peptides were the same as the literature report. [14,15,24] The animal experiments were performed according to the national regulations and approved by the animal welfare committee.
NOTA Conjugation of Peptides FSH4 (3mg, 0.96μmol) was added to a vial and followed by 3ml Malemide-NOTA (0.5mg, 1,1μmol) in 2M ammonium acetate solutions. Then the mixture was stirred at 40 °C overnight. After puri cation with HPLC, the desired product was collected and lyophilized as a white powder. The mass spectrum was determined using a high resolution LC-MS system(Waters, Milford, USA).

Preparation of 68 Ga-NOTA-MAL-FSH4
Labeling the peptide with 68 Ga was carried out in an Eppendorf vial containing NOTA-MAL-FSH4 (20μg, 6.7 nmol). Fresh [ 68 Ga] GaCl 3 (185 MBq) eluate was added to the vial followed by 0.25M sodium acetate solutions. After incubating at 100 °C for 10 min, the complex was diluted with deionized water and then loaded into a Varian BOND ELUT C18 column. The labeled peptide was obtained after eluting the column with 200μL 10 mM HCl in ethanol. The product was reconstituted in saline and sterilized using a 0.22 μm Millipore lter. Radiochemical purity was analyzed with HPLC.
Cell lines and animal models PC-3 human prostate cancer cells were purchased from Cell Bank of Shanghai Institutes for Biological Sciences. The cells were cultured in DMEM (Gibco, USA) and supplemented with 10% (v/v) fetal bovine serum and kept in a humidi ed atmosphere containing 5% CO 2 at 37℃.
Male Balb/c nude mice (4 weeks old) were obtained from CAVENS Laboratory animal co.ltd, China. A suspension of 5×10 6 PC-3 cells in sterilized saline was subcutaneously injected into the right front ank of the mice. When the tumors developed proper sizes (100-300 mm 3 ), the following animal experiments were performed.
Cell Binding Assay PC-3 human prostate cancer cells (2×10 5 cells) were seeded in 6 wells plates and cultured at 37℃ overnight. After washing with binding buffer (RPMI, 0.5 % bovine serum albumin), 68 Ga-NOTA-MAL-FSH4 ( 37 KBq) and FSH1 peptide ranged from 0 to 5,000 nM were added to each well. The medium was removed after incubation for 2 hours at 37 °C. After washing 3 times with PBS, the cells were lysed using 1M NaOH and the radioactivity was determined in a γ-counter (PerkinElmer). Inhibitory concentration of 50 %(IC50) values were calculated by GraphPadPrism software. Experiment was performed in triplicate.

In vitro metabolism analyses
Mice bearing PC-3 tumors were sacri ced and the main organs (blood, liver, kidney and tumor) were collected. After centrifugation at 4℃, plasma was obtained from the blood. Other organs were homogenized with ice PBS at 4℃. Each plasma or tissue homogenate was added 37KBq 68 Ga-NOTA-MAL-FSH4 with or without 100μg aprotinin and incubated 30 min at 37℃. Plasma and tissue homogenates were denatured with an equivalent volume of acetonitrile in centrifuge tubes. After vortexing, the mixture was centrifuged and the supernatant was separated and analyzed by HPLC.
After anesthetized with iso urane, the mice were performed PET imaging using a microPET scanner (Inveon, Siemens). At 30min, 60 min and 120min postinjection, static PET images were acquired. To investigate the effects of enzyme inhibition by aprotinin, 100 μL 3.7 MBq 68 Ga-NOTA-MAL-FSH4 together with 100μL aprotinin (100μg, 200μg or 400μg respectively) in saline were injected into mice via a tail vein. Static PET images were acquired for 10 min at 30min, 60 min and 120min after injection. The quanti cation analysis of PET images was carried out according to the previous method. [14,15,24] Ex vivo biodistribution experiments Mice bearing PC-3 tumors were injected with about 740KBq radiolabeled peptides in the absence or presence of 400μg aprotinin and sacri ced at selected time points. Blood, tumor and major organs were collected and weighed. The radioactivity was measured by a γ-counter. Data were determined by the percentage injected dose per gram of tissue (%ID/g). Blocking studies were performed through coadministration of excessive unlabeled FSH1 with the tracer.
Statistical analysis.
One-way analysis of variance (ANOVA) and Student'st test were used for statistical analysis of data. P values of < 0.05 were considered to be statistically signi cant.

Results:
Chemistry NOTA-MAL-FSH4 was achieved with yields of nearly 50%. Analytical HPLC showed that the chemical purity was greater than 90%. High resolution LC-MS system measured m/z 3524. Radiochemistry detected at about 14 min by radio-HPLC and the radiochemical purity was greater than 95% ( Figure 2). The speci c activity was calculated to be at least 30GBq/μmol.

Cell Binding Assay
The binding of 68 Ga-NOTA-MAL-FSH4 to FSHR can be displaced by increasing amounts of unlabeled FSH1. (Figure 3). The IC50 value of displacement 68 Ga-NOTA-MAL-FSH4 with FSH1 was determined to be 139.4±1.16 nM.
In vitro metabolism analysis RP-HPLC analysis showed that the percentages of intact peptides remained in plasma, liver and tumor homogenates were 8.30 %, 9.57% and 7.06% respectively. In the presence of aprotinin, the corresponding values were increased to 34.32%, 20.63% and 15.35% respectively. (Figure 4) No intact peptide was found in the kidney homogenates.
Consistent with the results of in vitro metabolism analysis results, visualization of PC-3 tumors was substantially improved by co-injection of aprotinin, as shown by the increased signal intensity in the tumors. After coadministration of 100μg, 200μg and 400μg aprotinin, the uptakes of PC-3 tumors were 2.91±0.21%ID/g, 3.89±0.16%ID/g and 9.21± 0.22%ID/g respectively at 60 min p.i. At the same time, tumor to muscle uptake ratios were 6.33±0.58, 7.98±1.23 and 13.15±1.16 respectively.

Ex vivo Biodistribution Experiments
The data of the biodistribution experiments are listed in Table 1. Radioactivity accumulated in PC-3 tumors were 3.13±0.27 %ID/g, 2.21±0.37%ID/g and 1.78±0.16 %ID/g at 30, 60 and 120 min postinjection respectively. The tumor to blood and tumor to muscle uptake ratios were 1.46±0.13, 2.59±0.32, 3.76±0.31, and 4.05±0.37, 6.29±0.67 and 11.92±1.10 at the same time points respectively. Except for kidney, uptakes of the tracer in normal organs were below 2%ID/g at 60min after administration of 68 Ga labeled peptides. The levels of radioactivity accumulated in tumors signi cantly reduced to 0.51±0.13 %ID/g at 1h postinjection in the presence of excessive FSH1 peptide.
Treatment with aprotinin resulted in a profound increase in 68 Ga-NOTA-MAL-FSH4 uptake in PC-3 tumors. In accordance with PET imaging, the tumor uptakes signi cantly increased to 7.81±0.62 %ID/g at 60mins p.i. after treatment with 400μg aprotinin. Also, tumor to blood and tumor to muscle uptake ratios were raised to 8.13±0.76 and 12.67±1.18 respectively.

Discussion
Radiolabeled peptides are useful approaches for cancer diagnosis and therapy. [25,26]A close relationship between FSHR and cancers suggested that it is a potential target of the disease. Although FSH1 analogs labeled with PET nuclides can speci cally bind to FSHR, low tumor uptakes may decrease the diagnostic sensitivity. Previous showed that co-injection of the enzyme inhibitors could prolong the accelerated radioactivity of 99m Tc or 111 In labeled peptides in tumors. [27] [28]To evaluate the feasibility of improving the performance of radiolabeled FSH1 peptides through the strategy, a 68 Ga labeled modi ed FSH1 peptide was synthesized and the imaging properties were determined in prostate cancer models with or without aprotinin. 68 Ga-NOTA-MAL-FSH4, was successfully prepared in 20 min through solid-phase extraction with satisfactory radiochemical purity. The yields were identical to other 68 Ga labeled FSH1 peptide and signi cantly higher than those of 18 FAl labeled counterparts (~ 50%.) respectively. [14] In vitro competing assay revealed that the IC50 value of 68 Ga-NOTA-MAL-FSH4 towards FSHR was similar to those of 68 Ga-NOTA-MAL-FSH1 (139.4 ± 1.16nM vs 123.7 ± 1.21 nM). [24]It implied that the a nity of FSH4 to FSHR was not affected after introducing the hydrophilic linker to FSH1 peptides.
Consistent with the behaviors of other FSHR tracers, radioactivity accumulated higher in FSHR positive PC-3 xenografts than normal organs such as heart, brain, muscle and livers etc except kidneys after administration of 68 Ga-NOTA-MAL-FSH4. It implies that 68 Ga-NOTA-MAL-FSH4 is a candidate probe targeting FSHR. Preclinical studies revealed that tumor uptakes of the 68 Ga labeled modi ed FSH1 peptide was signi cantly higher than those of unmodi ed counterpart (2.07 ± 0.11 %ID/g vs 1.26 ± 0.06%ID/g, p < 0.05) at 60min postinjection respectively. [24] It means that hydrophilic linkers may bene t for enhancing the retention values of the peptide in tumors.
The stability of 68 Ga-NOTA-MAL-FSH4 in main metabolism organs (blood, liver,kidney) and the tumor were rstly analyzed using in vitro assay. Many radioactive metabolites were observed in the tissue homogenates after incubation of the peptide. It implies that 68 Ga labeled peptide may be easily cleaved by enzymes. HPLC analysis con rmed that polar radioactive metabolites with a retention time of about 3 minutes were disappeared in presence of aprotinin. At the same time, the intact peptides in blood, tumor and liver were signi cantly increased nearly 4, 2 and 2 times respectively. It primarily suggested that serine proteases might be responsible for the in vivo degradation of FSH1 peptides. Aprotinin e ciently inhibited enzyme degradation and improved the stabilization of the peptide in tissues.
In agreement with in vitro stability data, aprotinin coinjection resulted in favorable improvement on the pharmacokinetic pro le in vivo PET imaging. The tumor uptakes increased about 1.5, 2 and 4 times than control after coinjection of 100µg, 200µg and 400µg aprotinin at 60min p.i.respectively. Ex vivo biodistribution experiments revealed that radioactivities accelerated in normal organs such as liver, spleen and stomach etc after co-administration of 400µg enzyme inhibitor was signi cantly higher than those of control. Although the detailed mechanism was not cleared, the uptake ratios between tumor and non-target organ were similar due to stronger uptakes in tumors. For example, tumor to liver and tumor to stomach were about 2 and 3 respectively in the absence or presence of aprotinin at 60 min p.i. It was also noted that tumor to blood and tumor to muscles ratios could signi cantly increase nearly 3 and 2 times through preventing degradation of 68 Ga labeled peptide using aprotinin. Thus, coadministration of 400µg aprotinin with 68 Ga labeled FSH1 analog could gain satisfactory performance of FSHR target PET imaging.

Conclusion:
In summary, a 68 Ga labeled FSH1 analogs, 68 Ga-NOTA-MAL-FSH4, was successfully prepared. Preclinical data indicate that the tracer is potential for non-invasive visualization of FSHR expression.
Coadministration of aprotinin signi cantly enhanced the stability of 68 Ga labeled peptide, which promotes the pro le of tumor targeting.  Figure 1 Schematic representation for preparing 68Ga-NOTA-MAL-FSH4.