Treatment of Peritoneal Carcinomatosis by Targeted Delivery of the Radio-Labeled Tumor Homing Peptide 213Bi-DTPA-[F3]2 into the Nucleus of Tumor Cells

Background α-particle emitting isotopes are effective novel tools in cancer therapy, but targeted delivery into tumors is a prerequisite of their application to avoid toxic side effects. Peritoneal carcinomatosis is a widespread dissemination of tumors throughout the peritoneal cavity. As peritoneal carcinomatosis is fatal in most cases, novel therapies are needed. F3 is a tumor homing peptide which is internalized into the nucleus of tumor cells upon binding to nucleolin on the cell surface. Therefore, F3 may be an appropriate carrier for α-particle emitting isotopes facilitating selective tumor therapies. Principal Findings A dimer of the vascular tumor homing peptide F3 was chemically coupled to the α-emitter 213Bi (213Bi-DTPA-[F3]2). We found 213Bi-DTPA-[F3]2 to accumulate in the nucleus of tumor cells in vitro and in intraperitoneally growing tumors in vivo. To study the anti-tumor activity of 213Bi-DTPA-[F3]2 we treated mice bearing intraperitoneally growing xenograft tumors with 213Bi-DTPA-[F3]2. In a tumor prevention study between the days 4–14 after inoculation of tumor cells 6×1.85 MBq (50 µCi) of 213Bi-DTPA-[F3]2 were injected. In a tumor reduction study between the days 16–26 after inoculation of tumor cells 6×1.85 MBq of 213Bi-DTPA-[F3]2 were injected. The survival time of the animals was increased from 51 to 93.5 days in the prevention study and from 57 days to 78 days in the tumor reduction study. No toxicity of the treatment was observed. In bio-distribution studies we found 213Bi-DTPA-[F3]2 to accumulate in tumors but only low activities were found in control organs except for the kidneys, where 213Bi-DTPA-[F3]2 is found due to renal excretion. Conclusions/Significance In conclusion we report that 213Bi-DTPA-[F3]2 is a novel tool for the targeted delivery of α-emitters into the nucleus of tumor cells that effectively controls peritoneal carcinomatosis in preclinical models and may also be useful in oncology.


Introduction
A single, high linear energy transfer a-particle can kill a tumor cell [1]. Therefore, a-particle emitting isotopes may emerge as effective novel tools in cancer therapy [2]. Due to its high linear energy transfer, short half life and uncomplicated use 213 Bi is promising for medical applications. To avoid toxic side effects, targeted delivery of such isotopes into tumors is a prerequisite of their application in oncology. a-particles kill cells due to induction of double strand brakes in DNA with a high relative biological effectiveness. As the range of the particles is only 28-100 mm in mammalian tissues, the development of carriers mediating specific uptake into the nucleus of tumor-or tumor endothelial cells is important to optimize therapeutic efficacy. Therefore, an ideal carrier for radionuclide therapy with a-emitters would mediate the specific uptake of the radionuclide selectively into the nuclei of tumor cells but not into cells in normal organs. Considering the short physical half life of 213 Bi (45.6 minutes) a quick internalization of the carrier is essential. Tumor-specific monoclonal antibodies coupled to a-emitters such as 225 Ac, 211 At, 149 Tb or to 213 Bi have already been successfully tested by several groups in a number of clinical and preclinical studies in different tumor entities including leukemia, glioma, melanoma, lymphoma, breast, ovarian, neuroblastoma as well as colon-and prostate cancer [1,3,4,5,6,7]. Peptides may also be used as conjugates for the targeted delivery of a-emitters into tumors and may be a promising option for cancer therapy [8]. Although it has been shown that in animal models 225 Ac and 213 Bi coupled to DOTATOC are effective in the treatment of neuroendocrine tumors, there is only limited experience with alpha particle emitting radio-peptide-conjugates in the therapy of cancer [9,10]. As a first step to develop radionuclide therapies using tumor homing peptides, it is important to test the binding affinity, stability and therapeutic efficacy of alpha-emitters coupled to such peptides in vitro and in vivo.
Peptide libraries displayed on phage can be screened in vivo for phage particles that home to a specific target [11]. A number of peptides capable of homing to tumor vasculature were isolated in this manner [12]. In preclinical models, targeted drug delivery into tumors has been achieved by conjugation of anti-cancer drugs with vascular tumor homing peptides [13]. F3 is a 32 amino acid vascular tumor homing peptide that binds specifically to nucleolin on the surface of tumor cells [14,15]. In resting cells, nucleolin is primarily expressed in the nucleus. In contrast, in proliferating tumor or tumor endothelial cells, nucleolin is cyclically transported from the nucleus to the cell surface and back by a specific shuttle mechanism [16]. Therefore, nucleolin is potentially an appropriate target for selective tumor therapies. Indeed, it has been reported that F3 is capable of transporting nano-particles or fluorescent dyes such as FITC into the nuclei of tumor cells and tumor endothelial cells in vitro and in vivo [17].
Peritoneal carcinomatosis is a widespread dissemination and implantation of tumors throughout the peritoneal cavity. It is a frequent and serious complication of intra-abdominal cancers of the gastrointestinal tract, including carcinomas of the colon, pancreas and stomach as well as ovarian cancer [18]. Peritoneal carcinomatosis also occurs in patients with primarily extraabdominal tumors such as breast cancer and melanoma. The majority of patients who are diagnosed with ovarian or pancreatic cancer already have peritoneal disease at the time of diagnosis. The vague signs and symptoms associated with peritoneal disease, such as bloating and early satiety, often delay the diagnosis and the initiation of treatment, leading to a poor prognosis [19]. Additionally, the lack of an effective treatment strategy leaves patients at significant risk of recurrence. Therefore, there is an urgent need to develop more effective tools for the treatment of peritoneal carcinomatosis. Recent evidence suggests that radionuclide therapy may be more beneficial than intra-peritoneal chemotherapy as treatment for peritoneal carcinomatosis [7].
We generated an F3 dimer coupled to DTPA which chelates 213 Bi and tested binding affinity, bio-distribution and anti-tumor activity of this construct ( 213 Bi-DTPA-[F3] 2 ) in vitro and in vivo in a pre-clinical model of peritoneal carcinomatosis. We found that 213 Bi-DTPA-[F3] 2 binds with high affinity to tumor cells, is internalized into the nuclei of tumor cells in vitro and is effective in treatment of peritoneal carcinomatosis.

Ethics Statement
All experiments were performed according to the German law for the protection of animals. The animal experiments were approved and supervised by the Government of Upper Bavaria (Regierung von Oberbayern) (record number 55.2-1-54-2531-34-02 and 55.2-1-54-2531-52-07).

Materials
All materials not further specified were purchased from Sigma, Deisenhofen, Germany.

Cell Culture
The tumor cell lines MDA-MB-435 (human breast cancer), MIAPACA (human pancreas carcinoma), OVCAR 3 (ovarian cancer) and CMT 93 (murine colon cancer) were grown in RPMI To test stability the constructs were diluted 1:10 with murine serum, cell culture media or ascites from mice with peritoneal carcinomatosis and incubated at 37uC for the indicated times. 1-2 ml of the solution were then submitted to ITLC. The amount of activity bound to the peptide was then determined using a c-counter.

Internalization of 213 Bi-DTPA-[F3] 2 and cell fractionation in vitro
MDA-MB-435 cells were grown to a density of 1610 7 cells/dish in 10 mm culture dishes. For fractionation of cultured MDA-MB-435 cells after incubation with 1.85 MBq (33 ng/ml) 213 Bi-DTPA-[F3] 2 for 5, 10, 20, 45, and 90 minutes, cells were washed three times with ice cold PBS. Culture dishes were placed on ice and 4 ml of ice cold Nuclei EZ lysis buffer (Nuclei Isolation Kit from Sigma, Deisenhofen, Germany) were added. Cells were harvested using a bladed cell scraper. The entire cell lysate from each plate was transferred into a 15 ml centrifuge tube (Falcon), vortexed and set on ice for five minutes. The nuclei were then pelleted by centrifugation at 5006g for five minutes, washed by three times with EZ lysis buffer and resuspended in nuclei EZ storage buffer. The supernatant contains the cytoplasm and membrane fraction. The integrity of the nuclei was checked by microscopy. To determine the 213 Bi-radioactivity concentration within the nuclei of tumor cells we determined the volume and the number of nuclei using a haemocytometer. The radioactivity present was determined by a using a c-counter and used for calculation of the radioactivity concentration (MBq/ml). Soft agar assay for colony formation 1610 5 MDA-MB-435, MIAPACA, OvCAR3 and CMT 93 cells were cultivated in 6 well plates for 24 h. Cells were treated with different activities of 213 Bi-DTPA-[F3] 2 in cell culture media (RPMI 1640). After 12 h cells were detached with trypsin, washed 3 times and resuspended in RPMI 1640. 200 ml cell culture medium with 4610 4 tumor cells were filled into a 15 ml Falcon tube and mixed with 0.7% low melting agarose and 3 ml 26 RPMI 1640. 1.5 ml of this suspension were added per well of 6well plates covered with base agar. Cell culture dishes were then incubated at 37uC for one week and stained with 0.5 ml 0.005% Cresyl-violet-solution for the minimum of 1 hour. Colonies were counted using an inverted microscope (Zeiss, Oberkochen, Germany).

Tumor model of peritoneal carcinomatosis
To generate intraperitoneal xenograft tumors, 6-8 week old SCID-mice (Charles River, Sulzfeld, Germany) were injected i.p. with 1610 7 MDA-MB435 tumor cells. Mice were weighed weekly and were sacrificed when the mean normal body weigth decreased ,25% or when mice developped tumor side effects like ascites or reduced general condititon. All experiments were performed according to the German law for the protection of animals.

Biodistribution of 213 Bi-DTPA-[F3] 2
213 Bi-DTPA-[F3] 2 (3.7 MBq) was injected i.p. into SCID-mice with intra-peritoneal growing tumors in an advanced stage 20 days after inoculation of tumor cells. After 45 minutes, mice were sacrificed, the tumors and the organs were surgically removed and the radioactivity present in each organ (percent of the injected activity/g = % ID/g), the tumors and the blood was measured with a c-counter.

Autoradiography and histology
Tumor bearing mice were injected i. p. with 3.7 MBq 213 Bi-DTPA-[F3] 2 . The tumors and control organs were resected, snap frozen with liquid nitrogen and cut in 10 mm sections using a microtome (HM 500 O, Microm, Germany). Sections were then exposed to a micro-imager (mImager TM , Biospace, Germany). For H&E staining 7.5 mm sections were cut, fixed with 37% formalin and stained with haematoxiline and eosine.

Macroscopic reduction of cell clones in vivo
Two SCID-mice with intraperitoneal MDA-MB-435 xenografts were treated with 661.85 MBq 213 Bi-DTPA-[F3] 2 between the days 4 and 14 after inoculation of tumor cells as described above, two others with 66100 ml PBS. 50 days after injection of xenografts, mice were sacrificed and the small intestine and the colon were preparated for macroscopic examination of tumor spread.

Statistical Analysis
Statistical analysis of cell-survival-studies was performed using the student t-test for paired values and with t-test according to Lord and Moore for unpaired values. Kaplan-Meier curves were analyzed by a Log Rank Test.

Binding and internalization of 213 Bi-DTPA-[F3] 2
To couple F3 to the a-particle emitting isotope 213 Bi, DTPA-[F3] 2 was synthesized which is a F3-dimer linked to DTPA. The chelator DTPA effectively binds metal ions such as 213 Bi. It was reported recently by Porkka et al. reported [14] that F3 is internalized into the nucleus of tumor cells and tumor endothelial cells in vitro and in vivo. We therefore asked whether also 213

Toxicity of 213 Bi-DTPA-[F3] 2 -treatment
Our biodistribution study showed that 213 Bi-DTPA-[F3] 2 after i. p. injection accumulates in tumors and is excreted via the kidneys. Therefore, primarily renal toxic side effects are to be expected. The serum creatinine level is a highly specific marker of renal function. The serum concentration of creatinine was determined in all tumor bearing mice when they were sacrificed after therapy. We did not observe any significant elevation of the serum creatinine level after treatment with 213 Bi-DTPA-[F3] 2 (data not shown) at time of sacrifice. Discussion a-particle emitting isotopes such as 213 Bi are promising tools for cancer therapy. Due to the high toxicity of a-particles targeted delivery into tumor cells is a prerequisite of their application in oncology. Therefore, carriers that accumulate in tumor tissue but not in other organs are required. Ideally, the unbound molecules should be rapidly excreted and the carrier should mediate an internalization of the nuclide into the nucleus of tumor cells to enhance the DNA-damage and to reduce the cytotoxic effects on neighbouring cells within normal tissue. We generated conjugates of the vascular tumor homing peptide F3 and the metal chelator DTPA which effectively chelate 213 Bi. 213 Bi-DTPA-[F3] 2 is internalized into the cytoplasm and the nucleus of tumor cells, as already reported for FITC-labelled F3-peptide [10]. The peak activity of 213 Bi-DTPA-[F3] 2 within the nucleus was found approximately 5 minutes after application but significantly elevated levels were found for at least 90 minutes. As the halflife of 213 Bi is approximately 46 minutes, a considerable amount of the isotope will decay in the nucleus, damaging the DNA of cancer cells but not the surrounding tissue. Therefore, 213 Bi-DTPA-[F3] 2 induces cell death in tumor cells and effectively reduces the number of malignant tumor cell clones in vitro. This was demonstrated for of a variety of tumor cell lines from different types of carcinoma including pancreatic-, colon-and ovary carcinoma as well as breast cancer. Within the tumors 213 Bi-DTPA-[F3] 2 is primarily located in the tumor periphery and in the peri-vascular region. In contrast, only small amounts of radioactivity were detected in non-target organs except for the kidneys were 213 Bi-DTPA-[F3] 2 is present presumably due to secretion of the peptide. In previous studies by Porkka et al. [14] FITC-F3 was also found in single cells in the intestine and in the skin, whereas in our study no relevant activities were found in these tissues. This  difference is most likely due to the blood pool present in these organs which masks the quantitatively low yet potentially specific uptake in a small group of cells. Application of 213 Bi-DTPA-[F3] 2 by i.p. injection into animals with peritoneal carcinomatosis of MDA-MB-435 cells blocks tumor growth and increases the survival of the animals. Moreover, 213 Bi-DTPA-[F3] 2 not only blocks the growth of small tumor nodules but also reduces the growth of large intraperitoneal tumors.
The internalization of 213 Bi-DTPA-[F3] 2 into the nucleus seems to be necessary for its anti tumor activity as 213 Bi-DTPA had no relevant effect in vivo and was less active in vitro. As F3 is a fragment of the HMGN2 protein which is known to be associated with chromatin in a cell cycle dependent manner F3 transports 213 Bi potentially into close proximity of DNA enhancing the effect of the a-particles [23]. Taken together, these findings indicate that 213 Bi-DTPA-[F3] 2 is a promising new tool in the treatment of peritoneal carcinomatosis which may be helpful in many patients. As for example a high number of patients with ovarian cancer presents with peritoneal carcinomatosis at the time of diagnosis, 213 Bi-DTPA-[F3] 2 may be useful in treating small tumor nodules which are not visible by clinical inspection and improve the diagnosis of these patients by preventing tumor growth. 213 Bi-DTPA-[F3] 2 may also be useful for palliative treatment in patients with advanced peritoneal carcinomatosis spread throughout the whole abdomen as we found a trend of longer survival of animals treated with 213 Bi-DTPA-[F3] 2 . As advanced peritoneal carcinomatosis is poorly responsive to chemotherapy and frequently leads to complications such as ascites 213 Bi-DTPA-[F3] 2 may also help to alleviate such complications. We found that 213 Bi-DTPA-[F3] 2 blocks the clone formation of tumor cells from many tumor types causing peritoneal carcinomatosis, it may be helpful to reduce tumor growth and to reduce complications such as ascites. It has been shown that F3 binds to many tumor entities including breast and prostate cancer suggesting that F3-directed therapy may be widely applicable.
Though a-particles are highly cytotoxic, we did not observe severe side effects in animals i. p. treated with 213 Bi-DTPA-[F3] 2 over a three month observation period. We can not exclude the possibility that pathological changes might become apparent at later time points. As biodistribution studies revealed that 213 Bi-DTPA-[F3] 2 is excreted via the kidneys but does not accumulate in other organs, primarily renal side effects are to be expected. Assessment of renal function in mice treated with 213 Bi-DTPA-[F3] 2 did not show any pathological changes in creatinine levels. As it has been shown that the internalization of F3 into the nucleus is correlated to cell proliferation, in kidney cells nuclear internalization of F3 may be low due to a low proliferation rate resulting in a low nephro-toxicity of 213 Bi-DTPA-[F3] 2 [14,15].
It has been shown that a-emitting isotopes such as 225 Ac and 213 Bi coupled to cancer-specific antibodies are effective in the treatment of solid tumors as well as peritoneal carcinomatosis in pre-clinical models [24,25,26,27,28]. Here we report for the first time on a tumor-homing-peptide as a carrier for the a-particle emitting isotope 213 Bi. Compared to antibodies peptides may have some advantages For example, an advantage of peptides is the rapid renal excretion within hours. In contrast antibodies show circulation times of up to a week. Therefore radioactively labeled antibodies not bound to their target molecules on tumor cells will lead to higher radiation exposure of the reminder body compared to radioactive labeled peptides, when labeled to a-emitters with long half lifes such as 225 Ac. As we reported recently that the somatostatin receptor binding peptide DOTATOC coupled to 225 Ac is more effective in the therapy of neuroendocrine tumors than DOTATOC coupled to ß-particle emitting isotopes [9]. Therefore, we propose that specific tumor-homing-peptides such as F3 will emerge as carriers for a-particle emitting isotopes in the future.