Peptides as targeting probes against tumor vasculature for diagnosis and drug delivery

Tumor cells vs tumor vasculature as targets

Cancer cells express a large number of receptors on their surface. Some receptors are overexpressed and mediate important biological functions in tumor growth, migration, invasion and metastasis. Theoretically cancer cell is an excellent target for therapy and imaging. Conventional chemotherapy drugs are mainly designed to target cancer cells. However, it is difficult for these drugs to be absorbed by cancer cells because they can seldom accumulate into tumor mass due to poor blood perfusion, high interstitial pressure and abnormal vasculature inside the tumor mass[18,19]. In fact cancer cells are genetically unstable and often produce multidrug resistance to multiple chemotherapeutic drugs which is also considered as one of major reasons for the failure of cancer therapy [20-22]. Cultured cells which are usually utilized as target for peptide screening might lost tissue-specific characteristics or abnormally express some molecules that do not actually exist in the corresponding cells in vivo[23,24]. In contrast with tumor cells, endothelial cells of tumor display several advantages which promote tumor vasculature an attractive target for peptide discovery [5,18,25,26]. Tumor endothelial cells have the good genetic stability so that they rarely produce drug resistance. Tumor blood vessels are also highly accessible to any intravenously administered agents. A great quantity of data based on genomics and proteomics approaches have implicated that endothelial cells (from tumor mass or other organs) express distinct patterns of molecules regulated by their original organ tissues and microenvironment, which are the most important factors for selectivity and a prerequisite for phage peptide selection [27-29]. The unique zip codes of vasculature in different tissues exert vital roles in organ-specific physiological function or disease/tumor development[16].

In vivo phage display

Peptides recognizing specifically the tumor vasculature are promising agents to efficiently deliver drugs and imaging contrast to tumor sites. Tumor angiogenesis, the sprouting and growth of new blood vessels, is indispensible for tumor growth, development and progression [30,31]. Neovascularization of tumors not only provides nutrients and oxygen necessary to tumor growth through blood supply, but also carry cancer cells to adjacent/distant organs to induce metastasis [30,32]. It is likely that destruction of tumor vasculature will halt the tumor blood supply and further restrict tumor progression. In vivo phage peptide selection pioneered by Ruoshlahti’s laboratory since 1996[13] has been extensively exploiting the discovery of peptides targeting the vasculature in normal organs or angiogenesis-related diseased tissues. In brief, the procedure of in vivo phage library selection is closely similar to the classical in vitro screening based on the “binding-washing-eluting-amplifying” process against purified proteins. Only the target is now changed into live animal model. The phage peptide library is intravenously injected into animals and allowed to circulate for 5-15 minutes. It is understood that phage cannot leave the circulation in such a short period of time and the homing peptides displayed on the phage coat protein are able to bind to the target molecules expressed on the endothelial cells in the blood vessels. The subsequent phage enriching screening would allow the target phage to selectively bind to tumor blood vessels [11,27,33]. Using this method numerous peptides targeting the tumor vasculature have been identified. Some of the peptides are also found to recognize cancer cells besides tumor blood vessels owing to the fact that they may share the similar and related receptors in endothelial and cancer cells. Therefore these peptides may serve additional advantages as excellent candidates for drug delivery and cancer treatment [18,34-36].

Peptides targeting tumor vasculature

It has been shown that peptides obtained from the in vivo phage library selection against various tumor vasculatures are capable to efficiently deliver drug/imaging contrast to tumor sites. To this end some peptide conjugates are being tested in clinical studies and achieved promising results. Table ​Table11 gave a summary of tumor vasculature-homing peptides discovered in recent years through in vivo phage library screening.

Directed drug delivery by targeting-vasculature peptides

Arginine-glycine-aspartic acid (RGD) and asparagine-glycine-arginine (NGR) peptides are the two most famous peptides targeting tumor vasculature discovered earlier by phage display. RGD peptide can home to tumor vasculature selectively expressing α_vβ₃ and α_vβ₅ integrins, and NGR peptide (CNGRC) binds to CD13 specifically expressed in tumor vasculature [37]. Although intergrins are also expressed in normal tissue cells and blood vessels, endothelial cells in angiogenic vessels express a distinct set of integrins from its repertoire in quiescent endothelial cells [38]. α_vβ₃ and α_vβ₅ integrins are specifically upregulated in tumor endothelial cells undergoing angiogenesis [38,39]. The RGD peptide with two internal disulfide bonds recognizes selectively to α_vβ₃ and α_vβ₅ integrins, which promote RGD peptide or its conjugations to accumulate in tumor vasculature [40]. Similarly, CD13 is not exclusively expressed in tumor neovasculature. But NGR peptide specifically targets CD13 molecule expressed in tumor blood vessels rather than other CD13-rich tissues [36,41]. The exact mechanism so far remains unknown, but it is speculated that a unrevealed CD13 isoform is expressed in tumor blood vessels which might be involved in different glycosylation modification or conformational changes[36,42].

The two peptides have become a classic tool to be used to deliver various anti-cancer drugs including but not limited to chemotherapeutic drugs, cytokines, toxins, nucleic acids, radioactive isotopes and nanoparticles [37,43-46]. RGD and NGR peptides coupled with doxorubicin can specifically induce destruction of tumor vasculature in the breast cancer xenograft model in nude mice and further inhibit tumor growth but have no obvious cytotoxic effect on the vasculature of control organs. When conjugated with a pro-apoptotic peptide _D(KLAKLAK)₂ which by itself has no activity outside the cells, RGD/NGR peptides could deliver this pro-apoptotic peptide into the tumor vasculature through respective receptors. As a result, the conjugate selectively induced apoptosis of tumor endothelial cells and reduced the tumor growth as well as prolonged survival rate in animals [44]. NGR peptide (CNGRC) was also fused with human tumor necrosis factor alpha (TNF alpha) protein to constitute a new recombinant protein which could greatly enhance the activity of TNF alpha at very lower dose (0.3 μg) in tumor-bearing mice [47]. Further investigation disclosed that subnanogram (0.1 ng) of NGR-TNF alpha could synergistically enhance the tumor toxicity of doxorubicin and melphalan, cisplatin, paclitaxel, and gemcitabine in mouse tumor models, but did not increase the side effects [46]. Phase 1b study of NGR-TNF alpha was tested at low dose in combination with doxorubicin to patients with advanced solid tumors. Results showed that NGR-TNF alpha seldom induced a dose-limited systemic toxicity which previously limited TNF alpha usage for systemic treatment. Patients could well tolerate the side effects when TNF alpha and NGR combined [48]. Phase 1 clinical trial also showed similar conclusions [49]. Similarly, RGD peptide was also proven to be able to significantly enhance anti-tumor activity of TNF alpha at subnanogram level when combined with melphalan in tumor-bearing mice [50]. In addition to TNF alpha, other important cytokines such as truncated tissue factor (tTF), interferon gamma (IFN gamma) and interleukin-12(IL-12) which have great potential as anti-cancer agents, have been fused with NGR/RGD peptides or other peptides targeting tumor blood vessels to enhance their anti-cancer activity [51-55].

In recent years, other novel peptides exhibiting the homing ability to tumor blood vessels were also reported from different laboratories worldwide and tested in various systems [56-60]. Recently, we identified a vasculature-targeting peptide TCP-1 (CTPSPFSHC) using the in vivo phage library selection against an orthotopic colorectal cancer model. TCP-1 peptide can specifically recognize the blood vessels of orthotopic colorectal cancer in normal BABL/c mice induced by syngeneic colon cancer cells (colon 26) [19] and also orthotopic gastric cancer in nude mice induced by human gastric cancer cells (unpublished data). We also illustrated that TCP-1 peptide could efficiently deliver the fluorescein and the apoptosis-inducing peptide _D(KLAKLAK)₂ to the tumor site for imaging and targeted therapy. TCP-1 peptide appears to be a promising agent in molecular imaging and drug delivery for human gastrointestinal cancers since preliminary data shows that it could also recognize the blood vessels in colorectal cancer samples in humans.

It is envisaged that peptides simultaneously recognizing the tumor vasculature and cancer cells are more effective for cancer therapy than those only targeting tumor blood vessels or cancer cells. Such peptides like NGR and GRD display a dual targeting ability. TCP-1 peptide was also found to be able to bind to some colorectal cancer cells (unpublished data). This kind of peptide conjugate therefore not only can exert targeted anti-cancer function but also provide selective anti-angiogenesis therapy. However, in order to achieve both actions the dose of the new conjugate should be large so as to be able to fully infiltrate into the tumor mass which has a high interstitial pressure and abnormal vasculature structure. The conjugate may only penetrate three to five cell diameters close to blood vessels [18,61]. Recently, the discovery of tumor-penetrating peptides might hold good promise for increasing overall tumor accumulation and penetration abilities of drugs to a great extent. The peptides containing the consensus motif R/KXXR/K (R: arginine, K: lysine, X: any amino acids) [24] can bind to neourophilin-1 (NRP-1) protein expressed on the cell surface. The arginine or lysine in the C-terminal is indispensible for the binding activity. The peptide internally containing the motif can be cleaved by protease to expose the C-terminal so that the binding activity to NRP-1 is switched on (termed C-end rule, CendR). NRP-1 is a cell-surface receptor expressed in various cells and is also found to be overexpressed in tumor tissues [62]. Normally NRP-1 is involved in angiogenesis, regulation of vascular permeability, and development of the nervous system. It is suggested that CendR peptide can increase the vascular permeability, penetrate tissues in vivo through binding to NRP-1 of endothelial cells after intravenous injection [24]. The peptide CRGDKGPDC (termed iRGD) obtained through the in vivo phage library screening against experimental metastasis mouse model of human prostate cancer contains both RGD and CendR (RGDK) motifs [63]. Further studies showed that i.v. injected iRGD was first recruited to tumor endothelial cells through RGD motif interacting with integrins, and subsequently the CendR motif exposed by the proteolytic cleavage finished the interaction with NRP-1 which drove the peptide to cross the vascular wall and penetrate into the tumor parenchyma. iRGD peptide chemically conjugated with abraxane (a 130 nm albumin-based nanoparticle embedded paclitaxel) could increase more than 4-fold drug accumulation in tumor tissues when compared to conventional RGD conjugates [63]. Moreover, co-administration of the iRGD peptide with various anti-cancer drugs but not chemical conjugation also highly enhanced the drug penetration and accumulation and improved their therapeutic index [64]. Actually, several early-discovered peptides may have the tumor-penetrating activity similar to iRGD because they contain a potential CendR motif, such as F3 (a 31 amino acid peptide homing to cancer cells and tumor endothelial cells) [65], LyP-1(CGNKRTRGC), homing to cancer cells, tumor macrophages and lymphatic vessels [66], CREKA and CSRPRRSEC (two peptides homing to tumor blood vessels) [58,67].

Directed molecular imaging by targeting-vasculature peptides

The tumor-homing peptides have been investigated in various imaging detection system for tumor diagnosis including MRI, positron emission tomography (PET), single photon emission computed tomography (SPECT) and fluorescence confocal microendoscope when they are coupled with different dyes or imaging agents [68]. RGD and NGR are extensively used to deliver different imaging agents for molecular imaging studies in various tumor types or other angiogenesis diseases. [18F]-labelling RGD conjugate has been developed for PET diagnosis and administered to patients with squamous cell carcinoma of the head and neck (SCCHN) [69]. Results showed that [18F]-labelling RGD can successfully produce specific imaging in α_vβ3-expression SCCHN patients and the conjugate may be useful to assess angiogenesis and response of α_vβ3-targeted therapies for patients with SCCHN [69]. Other phase I trial of PET based on 18F-AH111585/RGD conjugate has been performed in breast cancer patients [70]. The conjugate is retained in the tumor tissues and can detect the breast cancer by PET. Many other agents have also been coupled with RGD peptide for imaging detection such as (64)Cu, (68)Ga [71], near infrared fluorescent [72] and 99mTc [73]. Recently, RGD or NGR peptide coupled with nanoparticles (quantum dots) was tested in mice tumor model for molecular imaging [74-76]. Intravenous injection of NGR peptide-labelled paramagnetic quantum dots (NGR-pQDs) into tumor-bearing mice was used to evaluate the angiogenesis activity of tumors by the MRI system [76]. NGR-pQDs were found to be capable of specifically detecting the tumor region with the highest angiogenic activity [76].

TCP-1 peptide identified by our laboratory was labelled by FITC and administered to mice bearing orthotopic colorectal tumor for ex vivo detection under the blue light. The simple conjugate was found to accumulate within the tumor mass with a high specificity, and even it could produce obvious signal in tumor mass as small as 2 millimetres in the ex vivo level. If this situation can be reproduced in clinical setting, TCP-1 peptide conjugate combined with PET, MRI or endoscope may improve the diagnosis of patients with colorectal cancer[19].

For the newly discovered peptide iRGD, the iRGD peptide-linked superparamagnetic iron oxide nanoworms can produce hypo-intense vascular signals and low intensity regions in the tumor mass detected by MRI after intravenous injection into tumor-bearing mice[63]. Its effect is obviously better than the conventional RGD peptide for tumor visualization. Histological staining confirms that iRGD nanoworms have the ability to more deeply penetrate into the tumor tissues than the RGD conjugate, suggesting the great potential of iRGD as a diagnostic agent in clinical practice[63].