Article

Nature Chemical Biology 2, 381 - 389 (2006)
Published online: 11 June 2006 | doi:10.1038/nchembio798

Combinatorial chemistry identifies high-affinity peptidomimetics against alpha4bold beta1 integrin for in vivo tumor imaging

Li Peng1, Ruiwu Liu1, Jan Marik1, Xiaobing Wang1, Yoshikazu Takada2 & Kit S Lam1

Small peptide–based agents have attracted wide interest as cancer-targeting agents for diagnostic imaging and targeted therapy. There is a need to develop new high-affinity and high-specificity peptidomimetic or small-molecule ligands against cancer cell surface receptors. Here we report on the identification of a high-affinity peptidomimetic ligand (LLP2A; IC50 = 2 pM) against alpha4beta1 integrin using both diverse and highly focused one-bead-one-compound combinatorial peptidomimetic libraries in conjunction with high-stringency screening. We further demonstrate that LLP2A can be used to image alpha4beta1-expressing lymphomas with high sensitivity and specificity when conjugated to a near infrared fluorescent dye in a mouse xenograft model. Thus, LLP2A provides an important tool for noninvasive monitoring of alpha4beta1 expression and activity during tumor progression, and it shows great potential as an imaging and therapeutic agent for alpha4beta1-positive tumors.

Cell surface proteins are excellent targets for both diagnostic imaging and targeted therapy of cancers1. Several monoclonal antibodies against cancer cell surface proteins have been approved by the United States Food and Drug Administration for cancer therapy2, 3. Peptides represent another class of targeting agents and have several advantages over antibodies: favorable pharmacokinetics and better tumor penetration because of smaller size, less nonspecific binding to the reticuloendothelial system, easy derivatizing and manufacturing, and proteolytic stability if D and unnatural amino acids are used1, 4.

alpha4beta1 integrin has important roles in cancer metastasis and development5, 6, 7, 8, 9, 10, 11, 12. It is a noncovalent heterodimeric transmembrane receptor that recognizes the QIDS and ILDV motifs of two widely known ligands, the vascular cell adhesion molecule-1 (VCAM-1) and fibronectin, respectively13, 14. In adults, it is restricted to expression on hematopoietic cells regulating lymphocyte trafficking and homing; thus it has an important role in autoimmune diseases and inflammation15, 16. Recent studies have suggested that alpha4beta1 is also involved in regulating tumor growth, metastasis and angiogenesis. alpha4beta1 is expressed in leukemias and lymphomas, melanomas and sarcomas5. alpha4beta1 integrin promotes dissemination of tumor cells to distal organs by strengthening their adhesion to vasculature endothelium and by facilitating tumor cell extravasation5, 6. It prevents the apoptosis of malignant chronic lymphocytic leukemia (B-CLL) cells8, and it has important roles in the drug resistance of both multiple myeloma9 and acute myelogenous leukemia10. Antibodies against alpha4 integrin can inhibit multiple myeloma growth in murine models11. Furthermore, alpha4beta1 is expressed on proliferating endothelial cells (but not on quiescent ones) in angiogenesis during tumor development12. Therefore, alpha4beta1 integrin is an attractive imaging and therapeutic target for cancers, especially lymphoid malignancies. High-affinity and specific targeting peptides are required to noninvasively image alpha4beta1 integrin in vivo and to develop anti-alpha4beta1–directed tumor therapy.

Phage-display peptide libraries have been widely used to identify cell surface receptor-bound peptides by panning with immobilized, purified targets and intact cells or through in vivo selection17. However, these peptides are limited to L-amino acids, which require cyclization or modification with D-amino acids to be more resistant to proteolysis, as shown by somatostatin analogs for tumor imaging18. In contrast, synthetic peptidomimetic-library approaches can incorporate D-amino acids, unnatural amino acids and small-molecule moieties, making final ligands with high affinity and proteolytic stability.

We developed the one-bead-one-compound (OBOC) combinatorial library method over a decade ago19. OBOC libraries are generated such that each bead displays only one chemical entity. Millions of compound beads can be synthesized within a week and concurrently screened in a few days using on-bead binding or functional assays20, 21, 22. Peptide ligands against a variety of targets have been discovered with the OBOC method1, 23. Beads that interact with target proteins are isolated for structure determination using an automatic microsequencer with Edman chemistry19. Microsequencing requires that peptides consist of only alpha-amino acids and are N-terminally free, thereby limiting the flexibility of peptide-library design. To overcome this limitation, a series of methods have been developed using either MS or automatic sequencing to decode peptide, peptidomimetic or small-molecule OBOC libraries24, 25, 26. These include N-terminally blocked peptides, cyclic peptides, branched peptides, peptides with unsequenceable building blocks and peptides derivatized by a large number of small molecules. In our encoding methods, we have developed topologically segregated bilayer beads in which the library compounds are displayed on the bead surface and the encoding tags reside in the bead interior24. Therefore, the potential interference of the coding tag with target proteins during library screening is eliminated. With such advances, the diversity of OBOC combinatorial libraries can now be substantially increased, and there are many more options for library design. These developments enable us to fully use the power of OBOC libraries for the identification of cancer-targeting agents having both peptidic and nonpeptidic features. By screening random OBOC peptide libraries against live Jurkat T-lymphoid leukemia cells, we have identified peptide ligands that contain LDI, LDF and LDV motifs against alpha4beta1 integrin27. However, these ligands are from random peptide libraries and have relatively low affinity, and their binding on alpha4beta1-expressing cells is barely detectable using radiolabeled or fluorescence-labeled ligands. In this report, we describe both the strategy of combining diverse 'initial' and highly 'focused' OBOC combinatorial libraries and the ways this strategy generated a high-affinity and highly specific peptidomimetic ligand to alpha4beta1 integrin.

Top

Results

Design and synthesis of focused OBOC libraries

As alpha4beta1 integrin is an excellent therapeutic target for lymphoid malignancies, we used both diverse and focused OBOC peptidomimetic-library approaches to develop high-affinity imaging agents for this target. To achieve maximal structure and activity relationship (SAR) information, we first synthesized an initial OBOC peptidomimetic library having large permutations (up to 1010) of different lengths: hexamer, heptamer and octamer (Fig. 1a). We designed the library based on the LDV motif14, 27 and on the fact that a 2-(4-(3-o-toylureido)phenyl)acetyl N-terminal cap at LDV greatly enhances the interaction28. In this library, we diversified the 2-(4-(3-o-toylureido)phenyl)acetyl position with 420 combinations of 30 isocyanates (X1 position; Supplementary Table 1 online) and 14 analogs of 4-aminophenylacetic acid (X2 position; Supplementary Table 2 online). We diversified LDV positions with 6 L-leucine analogs and 20 L-lysine derivatives (X3), 3 L-aspartic acid analogs (X4) and 18 L-valine analogs (X5), respectively (Supplementary Tables 3 and 4 online). We randomized the C terminus with 45 D and unnatural amino acids in positions X6, X7 and X8 (Supplementary Table 5 online). We used alpha4beta1-expressing Jurkat T-lymphoid leukemia cells to screen this library with the standard whole cell–binding assay, as previously described27. We selected 23 positive beads covered by a monolayer of Jurkat cells from 2 ml of settled beads (about 1.5 times 106 compound beads, a fraction of the more than 1010 possible permutations). Sequence determination allowed the definition of consensus residues at several positions. On the amino side of LDV, 96% of ligand sequences at the X1 position contained 2-methylphenyl urea, and at X2 there was an 87% preference for 4-aminophenylacetic acid and a 13% preference for 2-(4-aminophenyl)propionic acid. The X3, X4, and X5 positions (LDV template) showed preference to the L-lysine derivatives, L-2-aminohexanedioic acid (Aad) and hydrophobic amino acids, respectively. There was no notable consensus in the carboxyl side of LDV (residues X6–X8). In fact, the beads isolated from hexamer, heptamer and octamer libraries had similar binding affinities (data not shown), indicating that the extension of the carboxyl end of the ligand beyond X5 did not contribute to additional interaction between ligands and receptors.

Figure 1: Chemical structures and screening of the OBOC libraries for alpha4bold beta1-targeting ligands.

Figure 1 : Chemical structures and screening of the OBOC libraries for |[alpha]|4|[beta]|1-targeting ligands.

(a) The initial library (permutations: 30 times 14 times 26 times 3 times 18 times 45 times 45 times 45 = 5.4 times 1010). (b) The focused library (permutations: 2 times 10 times 3 times 26 = 1,560). (c,d) Competitive OBOC cell-based screening of the focused library using Jurkat cells. Without competitors (the alpha4beta1 antagonist BIO-1211) in the screening medium (c), most beads were covered entirely with Jurkat cells. In the presence of a soluble binding competitor (BIO-1211) (d), only a few beads displaying high-affinity ligands were coated with Jurkat cells (one positive bead is seen here). Scale bar, 50 mum.

Full size image (62 KB)

Based on the SAR results from the initial library, we designed and synthesized a highly focused pentamer library by fixing or limiting the number of building blocks to eliminate the nonligands and low-binding-affinity ligands (Fig. 1b). Therefore, the number of permutations was reduced from billions to 1,560 with the following selected building blocks in each position: 2-methylphenyl urea at the X1 position, 4-aminophenylacetic acid and 2-(4-aminophenyl)propionic acid at the X2 position, 7 L-leucine analogs and 3 L-lysine derivatives at X3, 3 L-aspartic acid analogs at X4 and 26 hydrophobic amino acids at X5 (Supplementary Table 6 online). Because both the initial and focused libraries were N-terminally blocked, they could not be sequenced directly with Edman chemistry. We used a bilayer bead-partition approach and peptide encoding method to construct the libraries24. The outer layer of the beads contained the library compounds, and the inner core contained the coding tags, which could be decoded by microsequencing, as previously described24.

Identification of high-affinity ligands with stringent screening

To screen this highly focused library, we developed a competitive cell-based screening method with high stringency so that only the beads carrying high-affinity ligands to alpha4beta1 integrin were selected. To achieve this, we incorporated an increasing amount of the known alpha4beta1 antagonist BIO-1211 into the screening solution to compete with the interaction between alpha4beta1 integrin and the immobilized library compounds on beads. BIO-1211 is one of the best alpha4beta1 antagonists reported in the literature28. Before the screening, we determined that 10 muM BIO-1211 is sufficient to completely eliminate the binding of Jurkat cells (3 times 105 ml-1) to the control beads, which contained immobilized BIO-1211. To identify high-affinity targeting agents for alpha4beta1, we increased the amount of BIO-1211 up to 500 muM under the same conditions as during screening (Fig. 1c,d). Under these highly stringent conditions, we screened approximately 75,000 beads, which amounted to a 48-fold excess of the library permutation. We isolated 20 positive beads bound Jurkat cells and treated them with guanidine HCl (8 M) to strip the beads of all cells and proteins. Then we retested the beads with freshly prepared normal peripheral-blood lymphocytes. We retrieved and sequenced 12 beads that bound preferentially to Jurkat cells but not to normal lymphocytes.

Characterization of alpha4beta1-targeting peptidomimetics

Upon structurally decoding the 12 beads, we obtained only two sequences, named LLP1A (1Compound 11) and LLP2A (2Compound 22), as two sets of six beads each shared an identical sequence (Fig. 2a). This repetition of identified sequences was not surprising, because we screened a large excess of beads in the small, highly focused library to cover all permutations. In both ligands, the small-molecule moieties (a 2-methylphenyl urea group at the X1 position and a 4-aminophenylacetic acid at the X2 position) turned out to be the same as those in BIO-1211 (ref. 28), whereas the residues at the X3 position (epsilon-6-(2E)-1-oxo-3-(3-pyridinyl-2-propenyl)-L-lysine), the X4 position (Aad) and the X5 position (1-amino-1-cyclohexane carboxylic acid in LLP2A and D-phenylalanine in LLP1A) were quite different from the LDV residues of BIO-1211.

Figure 2: The stability and specificity of LLP2A.

Figure 2 : The stability and specificity of LLP2A.

(a) The chemical structure of LLP1A (1Compound 11), LLP2A (2Compound 22), and LLP2A-biotin (3Compound 33). (b) The in vitro stability of LLP2A in human plasma. We quantified LLP2A (plusminus s.d., triplicate samples) by reverse-phase HPLC and found it to be stable over 18 d in 80% human plasma at 37 °C. (c) Flow cytometry analysis of CHO cells expressing alpha4beta1, alpha2beta1, alpha6beta1, alpha9beta1, alphaLbeta2 or alphaMbeta2 integrins, using antibodies raised against the alpha chain (anti-alpha mAb, left) or LLP2A (right) as the probe. Only those CHO cells transfected with alpha4 were shown to bind LLP2A.

Full size image (47 KB)

Because LLP1A and LLP2A were identified from competition screening under high concentration of soluble BIO-1211, we envisioned that they would have higher binding affinities than BIO-1211. We performed a cell-based alpha4beta1-mediated adhesion assay using Jurkat cells with the immobilized CS-1 peptide (a 25-amino-acid linear peptide of fibronectin that is responsible for its interaction with alpha4beta1) to determine the binding affinities of the ligands. We found the half-maximal inhibitory concentration (IC50) of BIO-1211 to be 0.3 plusminus 0.1 nM, which is similar to that previously reported28. In contrast, we found the IC50 of LLP2A and LLP1A to be 2.0 plusminus 1.4 pM and 22 plusminus 18 pM, respectively—an improvement of up to 150-fold. We selected LLP2A for further characterization because its affinity was higher than that of LLP1A. We showed LLP2A to be protease resistant and highly stable in plasma. Incubation of LLP2A with human plasma (80%) did not result in any degradation over 18 d (Fig. 2). We further determined the structure of LLP2A with NMR spectroscopy. We found that LLP2A adopted a well-ordered structure in solution with a turn in the unnatural amino acid region (Supplementary Fig. 1 online). We also evaluated the effect of LLP2A on cell proliferation and survival, and we saw no effect of LLP2A on either alpha4beta1-positive or alpha4beta1-negative cells (data not shown).

Specificity of LLP2A to alpha4beta1 integrin

We evaluated the specificity of LLP2A to alpha4beta1 integrin with cell-binding assays against different integrins using fluorescent microscopy and flow cytometry. We conjugated LLP2A to biotin (LLP2A-biotin, 3Compound 33; Fig. 2a) for histochemical staining and in vivo imaging. LLP2A-biotin (detected by Qdots605-streptavidin (SA) conjugate; Fig. 3) showed strong binding to alpha4beta1-expressing cancer cells such as Molt-4 (acute lymphoblastic leukemia) cells (Fig. 3a,e), but we observed no staining on alpha4beta1-negative A549 (non–small cell lung cancer) cells, which express alphavbeta1, alphavbeta3 and alpha3beta1 integrins (Fig. 3d,h)29. The staining of Molt-4 cells by LLP2A was completely abolished using either inhibitory monoclonal antibody (HP2/1) to alpha4 (Fig. 3b,f) or an excess of unlabeled LLP2A (Fig. 3c,g). Specific binding of LLP2A to alpha4beta1 was further confirmed on alpha4-transfected Chinese hamster ovary (CHO) and K562 (chronic myelogenous leukemia) cells, whereas no staining was observed on either of the parent cell lines, which express alpha5beta1 but not alpha4beta1 integrins (Supplementary Fig. 1). To investigate LLP2A's binding to integrins from each subfamily, we also evaluated the binding of LLP2A to CHO cells separately expressing the integrins alpha2beta1, alpha6beta1, alpha9beta1 (closely related to alpha4beta1), alphaLbeta2 and alphaMbeta2 and found no detectable binding (Fig. 2c). These data strongly indicate that LLP2A is highly specific for alpha4beta1 integrin.

Figure 3: In vitro cell binding assays with LLP2A-biotin.

Figure 3 : In vitro cell binding assays with LLP2A-biotin.

(a–h) Photomicrographs of cultured cells incubated with LLP2A-biotin followed by Qdots605-SA. (e–g) We observed strong fluorescence in Molt-4 cells (e), whereas the signal was abolished by monoclonal antibody (HP2/1) to alpha4 (f) or excess unlabeled LLP2A (g). (h) We detected no signal on A549 cells. Scale bar, 10 mum. (i) Flow cytometry analysis: both Jurkat cells and peripheral blood lymphocytes expressed alpha4beta1; LLP2A bound to normal lymphocytes only after activation of alpha4beta1 with Mn2+; LLP2A showed strong binding to Jurkat cells even without activation with Mn2+; and beta1-associated integrins were activated in Jurkat cells as shown by monoclonal antibody to beta1 LIBS (anti-beta1 LIBS mAb). Anti-alpha4 mAb, monoclonal antibody to alpha4.

Full size image (58 KB)

In addition, we showed that LLP2A binds preferentially to leukemia and lymphoma cells over normal lymphocytes. During screening, we selected only the beads that bound to Jurkat cells but not to normal peripheral blood lymphocytes. As expected, LLP2A showed strong binding to Jurkat cells but weak binding to normal lymphocytes, even though normal lymphocytes also express alpha4beta1 integrin (Fig. 3i). Using monoclonal antibody raised against integrin beta1 ligand-induced binding site (LIBS), which recognizes activated integrins, we showed that on normal peripheral blood lymphocytes alpha4beta1 integrin is in the resting form, which has low affinity to ligands (Fig. 3i). Treatment of normal lymphocytes with Mn2+, which activates alpha4beta1 integrin, did increase the binding of these cells to LLP2A (Fig. 3i). A variety of alpha4beta1-expressing leukemia and lymphoma cell lines, including Molt-4, Reh and HL-60 cells, also bound to LLP2A at levels comparable to those of Jurkat cells (Supplementary Fig. 1). We further demonstrated that LLP2A also binds to fresh leukemia cells obtained from patients with acute lymphocytic leukemia or chronic lymphocytic leukemia (Supplementary Fig. 1).

Molecular interaction between LLP2A and alpha4beta1 integrin

We characterized the molecular interaction between LLP2A and alpha4beta1 integrin using alanine-scanning mutagenesis of alpha4. We expressed human alpha4 as a heterodimer with endogenous hamster beta1 in CHO cells. We tested LLP2A for binding to CHO cells that had been stably transfected with wild-type or different mutant alpha4 cDNAs. Mutations of Tyr187, Trp188 and Gly190 in alpha4 inhibit alpha4beta1-mediated cell adhesion to both VCAM-1 and CS-1 (ref. 30). These three critical residues are located in the ligand-binding site (residues 108–268) and clustered in the predicted beta-turn structure (residues 181–190) of the third N-terminal repeat of the alpha4 subunit. As LLP2A blocked alpha4beta1-mediated Jurkat cell adhesion to VCAM-1 and CS-1 (Supplementary Fig. 1), we hypothesized that the binding site of LLP2A on alpha4beta1 is close to or overlapping with that of the natural ligands VCAM-1 and fibronectin. Therefore, we selected 16 mutations around these three critical amino acids of the alpha4 chain (residues 184–203) for the binding assay with LLP2A. We allowed LLP2A-biotin to bind to CHO cells expressing wild-type or mutant alpha4, and then we stained them with SA-PE and detected them with flow cytometry. We determined the binding of LLP2A to each mutant after normalizing with the expression level of alpha4beta1 evaluated with monoclonal antibody B5G10 to alpha4. Most CHO cells expressing mutants bound to LLP2A at levels comparable to those of cells expressing wild-type alpha4. W188A and G190A alpha4 mutants yielded substantially lower binding to LLP2A than wild type and other mutants (Fig. 4). To a lesser degree, the Y187A mutation also reduced the binding to LLP2A (Fig. 4). This binding result is consistent with the fact that Tyr187, Trp188 and Gly190 are crucial for the binding of VCAM-1 and CS-1 (ref. 30), suggesting that the binding site for LLP2A on alpha4beta1 is very close to the binding sites for VCAM-1 and fibronectin.

Figure 4: Effect of single amino acid mutation of alpha4 on its binding to LLP2A.

Figure 4 : Effect of single amino acid mutation of |[alpha]|4 on its binding to LLP2A.

We incubated CHO cells expressing wild-type or mutant alpha4 with LLP2A-biotin in TBS (1 mM Mn2+) followed by SA-PE. We used monoclonal antibody (B5G10) raised against alpha4 to monitor the expression level of alpha4. Parent CHO cells neither express alpha4 nor bind LLP2A (data not shown). All data are expressed as means of (staining by LLP2A) / (alpha4 B5G10 expression) plusminus s.d. (triplicate samples). Thus, the relative binding of LLP2A to these mutants was normalized with the expression level of alpha4.

Full size image (28 KB)

Accumulation of LLP2A in alpha4beta1-expressing tumors

We used in vivo optical-imaging studies of mouse xenografts to investigate the targeting efficiency of LLP2A in live mice. Optical imaging is becoming an increasingly important tool for visualizing molecular processes in vivo31. Near-infrared fluorescent (NIRF) dyes allow the imaging of deeper tissues because of their high penetration and low tissue absorption and scattering. NIRF dye–labeled peptide is a sensitive tracer for imaging somatostatin receptors in tumors32. We coupled LLP2A to the NIRF dye Alexa680 by using the strong interaction between SA and biotin. We prepared the targeting complex (LLP2A-SA-Alexa680) by incubating LLP2A-biotin with the commercially available Alexa680-conjugated SA (SA-Alexa680). We first assessed the binding activity of the NIRF targeting complex in vitro using cell-binding assays. Immunocytochemical data confirmed strong and specific binding to alpha4beta1-positive cells (data not shown). We injected various amounts of LLP2A-SA-Alexa680 into nude mice bearing human subcutaneous xenografts of Molt-4 tumors (which are alpha4beta1-positive) to establish the optimal dose. We imaged the mice (n = 3) 0, 1, 3, 6, 12, 16, 24 and 44 h after administration with the established dose of 60 nmol kg-1. We observed substantial contrasts between Molt-4 tumors and normal tissue from 6 h until 2 d after injection (Supplementary Fig. 1). We derived semiquantitative information from NIRF images by integrating fluorescence intensities from equal areas within tumor and normal tissue regions. The time course of fluorescence intensity shows that LLP2A uptake by tumors was much higher than that of normal tissue starting 3 h (and reaching the greatest difference 24 h) after injection (Supplementary Fig. 1).

To further characterize the in vivo targeting potential of LLP2A, we divided nude mice bearing Molt-4 tumors into four groups for imaging. We injected group 1 (n = 3) intravenously with the LLP2A-SA-Alexa680 conjugate. As controls, group 2 (n = 3) received SA-Alexa680 (without LLP2A), group 3 (n = 2) received the conjugate of SA-Alexa680 and scrambled LLP2A-biotin (4Compound 44; Supplementary Fig. 1), and group 4 (n = 2) received monoclonal antibody (HP2/1) to alpha4 1 h before administration of the targeting agent (LLP2A-SA-Alexa680). We performed imaging 24 h after injection (Fig. 5) and found the autofluorescence of the Molt-4 tumor to be negligible (Fig. 5a). SA-Alexa680 alone without LLP2A showed no notable accumulation in Molt-4 tumors (Fig. 5b). We observed substantial fluorescent uptake in Molt-4 tumors after injection with LLP2A-SA-Alexa680 (Fig. 5e). Controls with the scrambled LLP2A-SA-Alexa680 conjugate also revealed negligible uptake into the tumor (Fig. 5c). Pretreatment of the Molt-4 tumor–bearing mouse with 10 mg kg-1 of monoclonal antibody HP2/1 raised against alpha4 substantially reduced fluorescent uptake into the tumor (Fig. 5d). We observed no fluorescence localization in the corresponding region of healthy mice (mice having no xenografts) after injection with the LLP2A-SA-Alexa680 conjugate (Fig. 5f).

Figure 5: In vivo NIRF imaging of Molt-4 tumor–bearing mice.

Figure 5 : In vivo NIRF imaging of Molt-4 tumor|[ndash]|bearing mice.

We obtained images 24 h after injection. (a) We observed a negligible autofluorescent signal in this tumor-bearing mouse before injection of any fluorescent probe. (b) Mouse received SA-Alexa680 alone. (c) Mouse received scrambled LLP2A-SA-Alexa680 conjugate. (d) Mouse was pretreated with monoclonal antibody HP2/1 raised against alpha4 before administration of LLP2A-SA-Alexa680 conjugate, resulting in substantial decrease in fluorescence uptake in tumor. (e) Mouse received LLP2A-SA-Alexa680 conjugate, resulting in high fluorescence uptake in the tumor. (f) Healthy mouse (without xenografts) received LLP2A-SA-Alexa680 conjugate. In b–f, we observed fluorescence uptake in the kidneys, but it was underestimated owing to the limited optical penetration. Fluorescence intensity is shown in arbitrary units.

Full size image (35 KB)

Ex vivo imaging of excised tumors and organs further confirmed the targeting specificity of LLP2A. In whole-animal imaging, fluorescent signals of deep organs are often underestimated because of optical impedance by soft tissues. We therefore also performed ex vivo imaging immediately after obtaining the 24 h postinjection in vivo imaging data. The LLP2A-SA-Alexa680 conjugate accumulated primarily in the Molt-4 tumor and kidney, whereas uptake in other normal organs was low (Fig. 6a). Experiments using SA-Alexa680 alone without LLP2A yielded notable localization only in the kidney (Fig. 6b). It is not surprising to see renal uptake in both experiments because SA is known to bind to kidneys, as shown in pretargeted radioimmunotherapy of cancers using radiolabeled SA33. We also performed microscopic analysis of frozen tumor tissues from the control (Fig. 5b) and experimental (Fig. 5e) mice to confirm the targeting of LLP2A at the cellular level. Cellular fluorescence intensity was very high for the tumor from the mice injected with LLP2A-SA-Alexa680 conjugate (Fig. 6c–e). In contrast, the tumor from the control mouse injected with SA-Alexa680 alone only yielded a small amount of fluorescent staining (Fig. 6f,g). Notably, some blood vessels within the Molt-4 tumors were also stained by the LLP2A-SA-Alexa680 conjugate (Fig. 6c,e), an observation that is consistent with the finding that alpha4beta1 integrin is expressed on proliferating but not quiescent endothelial cells in neovascularization during tumor angiogenesis12. Large blood vessels in tumors tended to have higher uptake than small capillaries (Supplementary Fig. 1).

Figure 6: Ex vivo NIRF images and microscopic analysis of tumors and organs that were excised from mice 24 h after receiving (a,c–e) LLP2A-SA-Alexa680 or (b,f,g) SA-Alexa680 alone.

Figure 6 : Ex vivo NIRF images and microscopic analysis of tumors and organs that were excised from mice 24 h after receiving (a,c|[ndash]|e) LLP2A-SA-Alexa680 or (b,f,g) SA-Alexa680 alone.

(a,b) Ex vivo imaging. Fluorescence (in arbitrary units) uptake into Molt-4 tumors was high in the mice that received LLP2A-SA-Alexa680 conjugate. (c–g) Histological analysis of tumor cryosections. Fluorescence signals from LLP2A probes (red) were detected only in tumor cells and some tumor vessels from mice that received LLP2A-SA-Alexa680 (c–e), whereas low signals were observed in the tumors from mice injected with SA-Alexa680 (f–g). We visualized tumor and vascular endothelial cells by anti-CD3 (orange) and anti-CD31 staining (green), respectively. Scale bar, 25 mum.

Full size image (49 KB)

Furthermore, we imaged a group of mice (n = 4) bearing tumors of alpha4-transfected K562 (alpha4beta1-positive) and K562 (alpha4beta1-negative) parent myeloid leukemia cells with the LLP2A-SA-Alexa680 conjugate to investigate whether alpha4beta1-negative tumors can be visualized by targeting alpha4beta1-expressing blood vessels or tumor-associated inflammatory cells. The uptake of LLP2A-conjugate in alpha4-transfected K562 tumors was substantially higher than that of K562 tumors as demonstrated by both in vivo and ex vivo imaging and microscopic analysis of the tumor cryosections (Fig. 7). Although some vascular endothelial cells of K562 tumors were also stained by LLP2A (Fig. 7), the fluorescent signal of the entire K562 tumor in whole-animal imaging was low. More sensitive imaging modalities, such as positron emission tomography, may be required to image tumors in vivo if only the tumor vascular endothelium is targeted. Together, these results demonstrate that LLP2A specifically targets alpha4beta1-expressing tumors in vivo in a xenograft model.

Figure 7: Specific accumulation of LLP2A probes in alpha4bold beta1-expressing tumors.

Figure 7 : Specific accumulation of LLP2A probes in |[alpha]|4|[beta]|1-expressing tumors.

We injected mice bearing bilateral alpha4-trasfected K562 and normal K562 tumors with LLP2A-SA-Alexa680. (a) NIRF in vivo image. (b,c) White light (b) and NIRF (c) images of excised tumors. Fluorescence intensity is color coded as in Figures 5 and 6. Fluorescence uptake into K562alpha4 tumors was substantially higher than that of K562 tumors. Histological examination of cryosections of (d–f) K562alpha4 and (g–i) K562 tumors showed high fluorescence uptake (red) in (d) K562alpha4 tumor cells and some uptake in (d,g) some but not all tumor vessels. We visualized tumor and vascular endothelial cells by anti-CD33 (orange) and anti-CD31 staining (green), respectively. Scale bar, 25 mum.

Full size image (80 KB)

Top

Discussion

Through screening a diverse initial OBOC peptidomimetic library and then screening a focused peptidomimetic library (containing 1,560 discrete compounds) with high stringency, we have succeeded in identifying LLP2A, a high-affinity (IC50 = 2 pM) ligand for alpha4beta1 integrin. Replacement of the LDV motif with the unnatural amino acids epsilon-6-[(2E)-1-oxo-3-(3-pyridinyl-2-propenyl)-L-lysine), Aad and 1-amino-1-cyclohexane carboxylic acid has greatly contributed to the high binding affinity of LLP2A. LLP2A is much more potent than ligands discovered by substituting the LDV region with small molecules through in silico screening, which uses a model based on the X-ray conformation of the binding region of VCAM-1 (ref. 34). The OBOC combinatorial library approach is still applicable for identifying new ligands with high affinity even in the absence of structural information pertaining to natural protein ligands and target proteins.

In addition to its high affinity, LLP2A is also highly specific to alpha4beta1 integrin and binds preferentially to leukemia and lymphoma cells over normal lymphocytes. The affinity modulation of alpha4beta1 may have a critical role in the preference of LLP2A to malignant lymphoid cells. Conformational change in integrins (known as affinity regulation) is a very important process of regulating ligand binding35, 36. Many integrins are expressed in an inactive state (low affinity to ligands) on cells, especially circulating blood cells. They can be rapidly activated through inside-out and outside-in signaling to achieve high affinity (activated) states. We have proven that LLP2A does bind to normal lymphocytes after alpha4beta1 integrin has been activated with Mn2+. However, it still could be possible that these malignant lymphoid cells have mutated alpha4beta1 integrins that favor binding to LLP2A. Although we showed that LLP2A does not bind to a series of integrins from each subfamily, we cannot eliminate the possibility that LLP2A may still interact with alpha4beta7 integrin, which is expressed in a subpopulation of memory T cells that are capable of homing to intestinal sites and that have an important role in chronic inflammatory bowel disease37.

The binding site of LLP2A on alpha4beta1 integrin is close to or overlaps with the binding sites of VCAM-1 and fibronectin, based on their similar binding profiles to alpha4 mutants. Both Trp188 and Gly190 (and to a lesser degree Tyr187) were shown to be important for LLP2A binding, findings that are consistent with their known crucial roles for the binding of VCAM-1 and CS-1 (ref. 30). The binding profile of LLP2A to various alpha4beta1 mutants not only facilitates the design of more potent alpha4beta1-targeting ligands but will also enable us to develop LLP2A analogs that bind only to alpha4beta1 mutants and not to wild-type alpha4beta1.

In vivo NIRF imaging demonstrates that LLP2A specifically targeted to alpha4beta1-positive tumors in nude-mouse xenografts. There was a substantial accumulation of fluorescence from the targeting conjugate (LLP2A-SA-Alexa680) in Molt-4 tumors. We confirmed unambiguously the specific labeling of alpha4beta1-expressing tumors by ex vivo and microscopic analysis of tumors and organs that were excised immediately after in vivo imaging. Histological analysis of tumor cryosections showed that some tumor neovasculature was also targeted by LLP2A. However, the application of LLP2A to target tumor vascular endothelial cells will require more characterization. PEG and dendrimers have been used as scaffolds for the development of imaging probes to achieve high-avidity multimers and to reduce tumor washing-off rate38, 39, 40. Here, we have demonstrated that SA can also be exploited as a tetravalent and bifunctional scaffold carrying both targeting peptides and NIRF dyes or radionuclides. In pretargeted imaging and radioimmunotherapy of cancers, SA has been used as a carrier of radionuclides to greatly increase tumor uptake and to accelerate blood clearance, because of its small size compared with radiolabeled antibodies41, 42. The SA-biotin ligand system is not ideal for clinical application because SA is immunogenic in humans. Nevertheless, this system serves as a very useful and simple preclinical imaging platform for evaluation of potential cancer cell surface–targeting agents. Using optical-imaging techniques, we have demonstrated the targeting capability of LLP2A. However, more quantitative data on tumor uptake and pharmacokinetics will require radiolabeled LLP2A and nuclear medicine imaging modalities such as positron emission tomography and single-photon emission computed tomography.

To the best of our knowledge, this is the first report on the successful use of peptide or peptidomimetic ligands for in vivo imaging of alpha4beta1 integrin. alpha4beta1 is transiently expressed during embryonic development and is crucial in placental and cardiac development43. Expression of alpha4 in human adults is normally restricted to hematopoietic cells, but during the tumorigenic process expression of alpha4beta1 occurs in a number of metastatic melanomas, sarcomas and tumor vasculature. alpha4beta1 has an important role in cancer metastasis5, 6, 7, drug resistance9, 10 and tumor angiogenesis12. The specific targeting potential of LLP2A as well as its resistance to proteolysis makes it an excellent candidate for the development of imaging and therapeutic agents for alpha4beta1-positive cancers, especially lymphoma and lymphocytic leukemia.

Although monoclonal antibodies to CD20 (for example, rituximab (Rituxan), ibritumomab tiuxetan (Zevalin) and tositumomab (Bexxar)) have been approved by the Food and Drug Administration for the treatment of low-grade B-cell lymphoma, they do have limitations. First, they do not bind to T cells and therefore they are ineffective in the treatment of T-cell lymphoma. Second, they bind to the reticuloendothelial system, as demonstrated by the recent report on the expected high liver and spleen uptake of 111In-labeled ibritumomab tiuxetan (a monoclonal antibody to CD20) in B-lymphoma patients44. Third, anti-CD20 monoclonal antibodies also bind to normal B lymphocytes. In contrast, LLP2A ligand (i) binds to both B- and T-cell lymphoid cancer cell lines, (ii) has low uptake in liver and spleen, as demonstrated in NIRF imaging, and (iii) has low affinity to normal T or B lymphocytes. We believe LLP2A is an ideal vehicle for the delivery of radionuclides (such as 90Y or 131I), cytotoxic agents, toxins, cytokines and nanoparticles to the lymphoid cancer cells, and therefore it has great potential for becoming a new, effective therapeutic agent for patients with lymphoid cancer.

Owing to the versatility of the synthetic and encoding methods used in the OBOC combinatorial libraries, many different structural designs can be incorporated. Now we are able to substantially increase the diversity of OBOC libraries and synthesize not only peptide but also peptidomimetic and small-molecule libraries. These advances enable us to fully use the power of OBOC libraries for identifying new ligands that are 'drug like' and protease resistant. Here we have validated the strategy of first screening an OBOC library with great diversity to provide comprehensive SAR information, and then designing highly focused peptidomimetic or even small-molecule libraries and screening them under high-stringency conditions to identify compounds of high potency and specificity. This library design strategy and screening approach is general and can be readily applied to the discovery and optimization of drug leads for diverse biological targets.

In summary, we have described the identification of a alpha4beta1-targeting peptidomimetic ligand using an OBOC combinatorial library approach in conjunction with a high-stringency, competitive, cell-based screening method. The identified ligand LLP2A possesses high affinity (IC50 = 2 pM) and specificity to the alpha4beta1 receptor. Using NIRF imaging, we have demonstrated the specific targeting of LLP2A to alpha4beta1-expressing lymphoma xenografts in nude mice in vivo, a result that is consistent with ex vivo analysis of tumors and organs. Thus, LLP2A is an attractive targeting agent for noninvasively imaging alpha4beta1 integrin during tumor progression and for developing anti-alpha4beta1–directed therapy.

Top

Methods

Synthesis of the initial and focused OBOC libraries.

We synthesized the OBOC libraries on TentaGel S NH2 resin (Rapp Polymere Gmbh) using a "split-mix synthesis" method and a newly developed encoding strategy involving the bilayer bead approach24 (Supplementary Methods online). Briefly, we used a biphasic-solvent approach to create bilayer beads so that only the library compounds were displayed on the outer layer of the beads; the coding tags resided in the inner core. We synthesized the peptide moiety of the peptidomimetic by standard solid-phase peptide synthesis techniques using 9-fluorenylmethoxycarbonyl (Fmoc) chemistry and N-hydroxybenzotriazole (HOBt)/N,N'-diisopropylcarbodiimide (DIC) coupling45. We confirmed the completion of coupling with a ninhydrin test46.

Cells.

We obtained all cells from American Type Culture Collection, except as otherwise described. alpha4-transfected K562 cells and alpha2, alpha4/alpha4 mutant, alpha6, alpha9, alphaLbeta2 and alphaMbeta2-transfected CHO cells have been previously described30, 47, 48, 49, 50. We prepared normal peripheral lymphocytes and patient leukemia cells using the Ficoll-Paque gradient method from peripheral blood of healthy donors and leukemia and lymphoma patients, respectively.

High-stringency, competitive, cell-based screening method.

We washed beads extensively with double-distilled water and phosphate-buffered saline (PBS) before screening. We performed the cell-bead binding by incubating beads with Jurkat cells (3 times 105 ml-1) in complete RPMI 1640 medium in Petri dishes in a humidified CO2 incubator with shaking (60 rpm). Beads bound by cells appeared as rosettes with a central bead covered by a monolayer of cells under a microscope. To determine the minimum amount of competitors to include in the screening medium, we titrated serial dilutions of the competitor BIO-1211 to inhibit the binding of Jurkat cells on BIO-1211–displayed beads (control beads) as described above. We found that 10 muM of BIO-1211 blocks the binding completely. Therefore, we included increasing amounts of BIO-1211 (from 10 muM up to 500 muM) in the screening buffer to compete with ligands displayed on beads until the top candidate beads (about 20 beads) among approximately 75,000 beads remained covered by lymphoma cells. We selected beads bound by cells and washed them with guanidine HCl (8 M) to strip all cells and proteins. Then we retested the beads with normal peripheral blood lymphocytes. We selected and sequenced beads that bound preferentially to Jurkat cells but not to normal lymphocytes.

Synthesis of LLP2A, LLP2A-biotin and scrambled LLP2A-biotin.

The synthetic chemistry of LLP2A, LLP2A-biotin and scrambled LLP2A-biotin for biological testing is similar to that of the library using HOBt/DIC coupling (Supplementary Methods). We used Rink amide resin as solid support to prepare compounds with carboxyl amide.

Cell adhesion assays.

We studied the binding affinities (IC50s) of the ligands in a Jurkat-cell adhesion assay by inhibiting the alpha4beta1-mediated cell adhesion to CS-1 peptides, which contain the binding motif of fibronectin to alpha4beta1. We coated 96-well plates with 1 mug ml-1 neutravidin, followed by biotin-conjugated CS-1 peptides (synthesis described in Supplementary Methods) after washing. We blocked the wells with 1% bovine serum albumin in PBS. We added Jurkat cells with serial dilutions of tested ligands in 100 mul binding buffer (Tris-buffered saline or TBS, 1 mM Mn2+) and allowed them to bind for 30 min; we removed unbound cells by gentle washing. We fixed bound cells with 3.7% formaldehyde and stained them with 0.1% crystal violet. We dissolved the dye in 1% SDS and recorded on a 96-well plate reader (Safire, Tecan) at 570 nm. We calculated IC50 data from the inhibition curves resulting from the concentration-dependent inhibition.

In vitro stability assay of LLP2A in human plasma.

We prepared heparin-, citrate- and EDTA-treated plasma from the peripheral blood of healthy donors. We added LLP2A to the plasma (80%) and incubated it at 37 °C. At each of the indicated time points, we withdrew and terminated samples by precipitating plasma proteins with trifluoroacetic acid (10%). We then analyzed the supernatant, which contained the peptides, by reverse-phase HPLC and measured the area of the peak corresponding to the intact samples.

Fluorescence microscopy.

We incubated cells with the LLP2A-biotin conjugate (10 nM) for 1 h at 4 °C in TBS buffer (1 mM Mg2+). For blocking experiments, we incubated cells with inhibitory monoclonal antibody to alpha4 (HP2/1) (Serotec) or an excess of LLP2A (100 muM) for 1 h before adding LLP2A-biotin. Then we washed the cells three times with TBS and then incubated them with Qdots605-SA (10 nM) (Quantum Dot Corp.). We washed and examined the cells using an inverted Olympus fluorescence microscope (IX70).

Flow cytometry.

For the binding of LLP2A, we incubated 0.5 times 106 cells with LLP2A-biotin (100 nM) for 1 h at 4 °C in TBS (1 mM Mg2+/Mn2+) and then with SA-PE (Molecular Probes) after washing. For the expression of alpha4beta1, we incubated cells with mouse monoclonal antibody P1H4 (Chemicon International), raised against human alpha4 integrin, in PBS and then with goat anti–mouse IgG–FITC (Serotec). We analyzed the activation state of alpha4beta1 with mouse monoclonal antibody to human beta1 integrin LIBS (Chemicon International). We measured the fluorescence of 10,000 cells using a Coulter Epics XL-MCL flow cytometer (Beckman Coulter, Inc.).

In vivo and ex vivo mouse imaging.

We prepared the LLP2A-Alexa Fluor680 complexes by incubating LLP2A-biotin conjugates with SA-Alexa Fluor680 (Molecular Probes) at molar ratios of 5:1 for 1 h at 4 °C. We confirmed the fluorescence labeling by in vitro cell-binding assays as described in the fluorescence microscopy section above.

We performed in vivo NIRF imaging experiments on nude (nu/nu) mice bearing either Molt-4 tumors or both alpha4-transfected K562 and normal K562 tumors. We injected Molt-4 cells (about 5 times 106 cells) or K562/alpha4-K562 cells (about 2 times 106 cells) subcutaneously into one side of the shoulder of mice. Tumors measured about 0.5–1.0 cm in diameter at the time of imaging. We administered various amounts of the imaging complex intravenously (0.6, 6, 60, 300 nmol kg-1) to establish the optimal dosage. We anesthetized mice using intraperitoneal injection of pentobarbital (60 mg kg-1), and we performed imaging using a Kodak multimodal-imaging system IS2000MM (Kodak) equipped with an excitation bandpass filter at 625 nm and an emission at 700 nm. Exposure time was 30 s per image. We analyzed images using the imaging station IS2000MM software (Kodak 1D Image Analysis Software; Kodak).

After in vivo imaging, we killed the mice by CO2 overdose and excised and imaged tumors, organs and muscle tissue with the Kodak imaging system as described above.

We performed all experiments that included animals in compliance with institutional guidelines and according to protocol no. 04-11527, approved by the Animal Use and Care Administrative Advisory Committee of University of California Davis.

Microscopic analysis of mouse xenografts after conjugate application.

We froze the tumors, cryosectioned them into Sections 7 mum thick and fixed them with acetone for histological analysis. To detect the expression of CD31 (a marker for the vascular endothelium) and CD3 (a marker for T-cell lymphoblastic leukemia) or CD33 (a marker for myeloid leukemia) in tumors, we rehydrated slides in PBS for 5 min and then blocked them in 5% BSA in PBS for 1 h at room temperature (23 °C). We incubated slides with primary antibodies at 2–10 mug ml-1 for 1 h at room temperature, and then we washed the slides three times in PBS and incubated them in secondary antibodies at 5 mug ml-1 for 1 h at room temperature. We washed slides three times in PBS and then mounted them with coverslips. We examined the slides with a confocal microscope (Zeiss LSM 510). We used the following primary antibodies: rat monoclonal antibody to mouse CD31 (Chemicon International), rabbit polyclonal antibody to CD3 (Santa Cruz Biotechnology, Inc) and rabbit polyclonal antibody to human CD33 (Santa Cruz Biotechnology, Inc). The secondary antibodies used were donkey anti–rat IgG–FITC (Chemicon International) and goat anti–rabbit IgG–AlexaFluor555 (Invitrogen).

Data processing and statistics.

All the data are given as mean plusminus s.d. of n independent measurements. For determination of tumor contrast, we calculated mean fluorescence intensities of the tumor area and of the normal tissue area by means of the region-of-interest function using Kodak 1D Image Analysis Software (Kodak).



Top

Acknowledgments

This work was supported by the National Institutes of Health (R33CA-86364, R33CA-99136 and U19CA113298) and the National Science Foundation (CHE-0302122). The 500-MHz NMR spectrometer was purchased in part with the grant NSF 9724412. We thank R. Wisdom, A. Lehman, S.M. Dixon and A. Enstrom for editorial assistance.

Author Contributions

L.P., study concept and design, acquisition and analysis of the majority of the data of the manuscript and drafting of the manuscript; R.L., design and synthesis of OBOC libraries and compounds, sequencing and decoding of positive beads and drafting of the manuscript; J.M., obtaining the NMR data of LLP2A and synthesis of CS-1 peptides; X.W., analysis of the NMR data of LLP1A, LLP1A-biotin and scrambled LLP2A-biotin; Y.T., providing all the CHO cell clones transfected with alpha4 mutants and critical revision of the manuscript; K.L., principal investigator of the project, study concept and design, critical revision of the manuscript, obtained funding and study supervision.

Note: Supplementary information is available on the Nature Chemical Biology website.

Competing interests statement

The authors declare no competing financial interests.

Received 28 February 2006; Accepted 9 May 2006; Published online 11 June 2006.

Top

References

  1. Aina, O.H., Sroka, T.C., Chen, M.L. & Lam, K.S. Therapeutic cancer targeting peptides. Biopolymers 66, 184–199 (2002). | Article | PubMed | ChemPort |
  2. Coiffier, B. et al. Rituximab (anti-CD20 monoclonal antibody) for the treatment of patients with relapsing or refractory aggressive lymphoma: a multicenter phase II study. Blood 92, 1927–1932 (1998). | PubMed | ISI | ChemPort |
  3. Wiseman, G.A. et al. Ibritumomab tiuxetan radioimmunotherapy for patients with relapsed or refractory non-Hodgkin lymphoma and mild thrombocytopenia: a phase II multicenter trial. Blood 99, 4336–4342 (2002). | Article | PubMed | ISI | ChemPort |
  4. Okarvi, S.M. Peptide-based radiopharmaceuticals: future tools for diagnostic imaging of cancers and other diseases. Med. Res. Rev. 24, 357–397 (2004). | Article | PubMed | ChemPort |
  5. Holzmann, B., Gosslar, U. & Bittner, M. alpha4 integrins and tumor metastasis. Curr. Top. Microbiol. Immunol. 231, 125–141 (1998). | PubMed | ISI | ChemPort |
  6. Vincent, A.M., Cawley, J.C. & Burthem, J. Integrin function in chronic lymphocytic leukemia. Blood 87, 4780–4788 (1996). | PubMed | ChemPort |
  7. Marco, R.A., Diaz-Montero, C.M., Wygant, J.N., Kleinerman, E.S. & McIntyre, B.W. Alpha 4 integrin increases anoikis of human osteosarcoma cells. J. Cell. Biochem. 88, 1038–1047 (2003). | Article | PubMed | ISI | ChemPort |
  8. de la Fuente, M.T., Casanova, B., Garcia-Gila, M., Silva, A. & Garcia-Pardo, A. Fibronectin interaction with alpha4beta1 integrin prevents apoptosis in B cell chronic lymphocytic leukemia: correlation with Bcl-2 and Bax. Leukemia 13, 266–274 (1999). | Article | PubMed | ISI | ChemPort |
  9. Damiano, J.S. & Dalton, W.S. Integrin-mediated drug resistance in multiple myeloma. Leuk. Lymphoma 38, 71–81 (2000). | PubMed | ISI | ChemPort |
  10. Matsunaga, T. et al. Interaction between leukemic-cell VLA-4 and stromal fibronectin is a decisive factor for minimal residual disease of acute myelogenous leukemia. Nat. Med. 9, 1158–1165 (2003). | Article | PubMed | ISI | ChemPort |
  11. Olson, D.L., Burkly, L.C., Leone, D.R., Dolinski, B.M. & Lobb, R.R. Anti-alpha4 integrin monoclonal antibody inhibits multiple myeloma growth in a murine model. Mol. Cancer Ther. 4, 91–99 (2005). | PubMed | ChemPort |
  12. Garmy-Susini, B. et al. Integrin alpha4beta1-VCAM-1-mediated adhesion between endothelial and mural cells is required for blood vessel maturation. J. Clin. Invest. 115, 1542–1551 (2005). | Article | PubMed | ChemPort |
  13. Osborn, L. et al. Arrangement of domains, and amino acid residues required for binding of vascular cell adhesion molecule-1 to its counter-receptor VLA-4 (alpha 4 beta 1). J. Cell Biol. 124, 601–608 (1994). | Article | PubMed | ChemPort |
  14. Komoriya, A. et al. The minimal essential sequence for a major cell type-specific adhesion site (CS1) within the alternatively spliced type III connecting segment domain of fibronectin is leucine-aspartic acid-valine. J. Biol. Chem. 266, 15075–15079 (1991). | PubMed | ISI | ChemPort |
  15. Lin, K.C. & Castro, A.C. Very late antigen 4 (VLA4) antagonists as anti-inflammatory agents. Curr. Opin. Chem. Biol. 2, 453–457 (1998). | Article | PubMed | ISI | ChemPort |
  16. Yusuf-Makagiansar, H., Anderson, M.E., Yakovleva, T.V., Murray, J.S. & Siahaan, T.J. Inhibition of LFA-1/ICAM-1 and VLA-4/VCAM-1 as a therapeutic approach to inflammation and autoimmune diseases. Med. Res. Rev. 22, 146–167 (2002). | Article | PubMed | ISI | ChemPort |
  17. Lowman, H.B. Bacteriophage display and discovery of peptide leads for drug development. Annu. Rev. Biophys. Biomol. Struct. 26, 401–424 (1997). | Article | PubMed | ISI | ChemPort |
  18. Weiner, R.E. & Thakur, M.L. Radiolabeled peptides in the diagnosis and therapy of oncological diseases. Appl. Radiat. Isot. 57, 749–763 (2002). | Article | PubMed | ISI | ChemPort |
  19. Lam, K.S. et al. A new type of synthetic peptide library for identifying ligand-binding activity. Nature 354, 82–84 (1991). | Article | PubMed | ISI | ChemPort |
  20. Lam, K.S., Lebl, M. & Krchnak, V. The "one-bead-one-compound" combinatorial library method. Chem. Rev. 97, 411–448 (1997). | Article | PubMed | ISI | ChemPort |
  21. Meldal, M., Svendsen, I., Breddam, K. & Auzanneau, F.I. Portion-mixing peptide libraries of quenched fluorogenic substrates for complete subsite mapping of endoprotease specificity. Proc. Natl. Acad. Sci. USA 91, 3314–3318 (1994). | Article | PubMed | ChemPort |
  22. Spring, D.R., Krishnan, S., Blackwell, H.E. & Schreiber, S.L. Diversity-oriented synthesis of biaryl-containing medium rings using a one bead/one stock solution platform. J. Am. Chem. Soc. 124, 1354–1363 (2002). | Article | PubMed | ChemPort |
  23. Copeland, G.T. & Miller, S.J. Selection of enantioselective acyl transfer catalysts from a pooled peptide library through a fluorescence-based activity assay: an approach to kinetic resolution of secondary alcohols of broad structural scope. J. Am. Chem. Soc. 123, 6496–6502 (2001). | Article | PubMed | ISI | ChemPort |
  24. Liu, R., Marik, J. & Lam, K.S. A novel peptide-based encoding system for "one-bead one-compound" peptidomimetic and small molecule combinatorial libraries. J. Am. Chem. Soc. 124, 7678–7680 (2002). | Article | PubMed | ChemPort |
  25. Song, A.M., Zhang, J.H., Lebrilla, C.B. & Lam, K.S. A novel and rapid encoding method based on mass spectrometry for "One-Bead One-Compound" small molecule combinatorial libraries. J. Am. Chem. Soc. 125, 6180–6188 (2003). | Article | PubMed | ChemPort |
  26. Affleck, R.L. Solutions for library encoding to create collections of discrete compounds. Curr. Opin. Chem. Biol. 5, 257–263 (2001). | Article | PubMed | ChemPort |
  27. Park, S.I. et al. The use of one-bead one-compound combinatorial library method to identify peptide ligands for alpha-4-beta-1 integrin receptor in non-Hodgkin's lymphoma. Lett. Pept. Sci. 8, 171–178 (2002). | Article
  28. Lin, K. et al. Selective, tight-binding inhibitors of integrin alpha4beta1 that inhibit allergic airway responses. J. Med. Chem. 42, 920–934 (1999). | Article | PubMed | ISI | ChemPort |
  29. Falcioni, R. et al. Expression of beta 1, beta 3, beta 4, and beta 5 integrins by human lung carcinoma cells of different histotypes. Exp. Cell Res. 210, 113–122 (1994). | Article | PubMed | ISI | ChemPort |
  30. Irie, A., Kamata, T., Puzon-McLaughlin, W. & Takada, Y. Critical amino acid residues for ligand binding are clustered in a predicted beta-turn of the third N-terminal repeat in the integrin alpha 4 and alpha 5 subunits. EMBO J. 14, 5550–5556 (1995). | PubMed | ISI | ChemPort |
  31. Weissleder, R. & Ntziachristos, V. Shedding light onto live molecular targets. Nat. Med. 9, 123–128 (2003). | Article | PubMed | ChemPort |
  32. Becker, A. et al. Receptor-targeted optical imaging of tumors with near-infrared fluorescent ligands. Nat. Biotechnol. 19, 327–331 (2001). | Article | PubMed | ISI | ChemPort |
  33. Wilbur, D.S., Hamlin, D.K., Sanderson, J. & Lin, Y. Streptavidin in antibody pretargeting. 4. Site-directed mutation provides evidence that both arginine and lysine residues are involved in kidney localization. Bioconjug. Chem. 15, 1454–1463 (2004). | Article | PubMed | ChemPort |
  34. Singh, J. et al. Identification of potent and novel alpha4beta1 antagonists using in silico screening. J. Med. Chem. 45, 2988–2993 (2002). | Article | PubMed | ISI | ChemPort |
  35. Hynes, R.O. Integrins: bidirectional, allosteric signaling machines. Cell 110, 673–687 (2002). | Article | PubMed | ISI | ChemPort |
  36. Shimaoka, M., Takagi, J. & Springer, T.A. Conformational regulation of integrin structure and function. Annu. Rev. Biophys. Biomol. Struct. 31, 485–516 (2002). | Article | PubMed | ISI | ChemPort |
  37. Rivera-Nieves, J. et al. L-selectin, alpha 4 beta 1, and alpha 4 beta 7 integrins participate in CD4+ T cell recruitment to chronically inflamed small intestine. J. Immunol. 174, 2343–2352 (2005). | PubMed | ChemPort |
  38. Chen, X. et al. Micro-PET imaging of alphavbeta3-integrin expression with 18F-labeled dimeric RGD peptide. Mol. Imaging 3, 96–104 (2004). | Article | PubMed | ChemPort |
  39. Venditto, V.J., Regino, C.A. & Brechbiel, M.W. PAMAM dendrimer based macromolecules as improved contrast agents. Mol. Pharm. 2, 302–311 (2005). | Article | PubMed | ChemPort |
  40. McIntyre, J.O. et al. Development of a novel fluorogenic proteolytic beacon for in vivo detection and imaging of tumour-associated matrix metalloproteinase-7 activity. Biochem. J. 377, 617–628 (2004). | Article | PubMed | ISI | ChemPort |
  41. Boerman, O.C., van Schaijk, F.G., Oyen, W.J. & Corstens, F.H. Pretargeted radioimmunotherapy of cancer: progress step by step. J. Nucl. Med. 44, 400–411 (2003). | PubMed | ISI |
  42. Rosebrough, S.F. Two-step immunological approaches for imaging and therapy. Q. J. Nucl. Med. 40, 234–251 (1996). | PubMed | ChemPort |
  43. Yang, J.T., Rayburn, H. & Hynes, R.O. Cell adhesion events mediated by alpha 4 integrins are essential in placental and cardiac development. Development 121, 549–560 (1995). | PubMed | ISI | ChemPort |
  44. Spies, S.M. Imaging and dosing in radioimmunotherapy with Yttrium 90 ibritumomab tiuxetan (Zevalin). Semin. Nucl. Med. 34, 10–13 (2004). | Article | PubMed
  45. Fields, G.B. & Noble, R.L. Solid phase peptide synthesis utilizing 9-fluorenylmethoxycarbonyl amino acids. Int. J. Pept. Protein Res. 35, 161–214 (1990). | PubMed | ISI | ChemPort |
  46. Kaiser, E., Colescott, R.L., Bossinger, C.D. & Cook, P.I. Color test for detection of free terminal amino groups in the solid-phase synthesis of peptides. Anal. Biochem. 34, 595–598 (1970). | Article | PubMed | ChemPort |
  47. Kamata, T., Puzon, W. & Takada, Y. Identification of putative ligand binding sites within I domain of integrin alpha 2 beta 1 (VLA-2, CD49b/CD29). J. Biol. Chem. 269, 9659–9663 (1994). | PubMed | ISI | ChemPort |
  48. Kamata, T., Wright, R. & Takada, Y. Critical threonine and aspartic acid residues within the I domains of beta 2 integrins for interactions with intercellular adhesion molecule 1 (ICAM-1) and C3bi. J. Biol. Chem. 270, 12531–12535 (1995). | Article | PubMed | ISI | ChemPort |
  49. Eto, K. et al. Functional classification of ADAMs based on a conserved motif for binding to integrin alpha 9beta 1: implications for sperm-egg binding and other cell interactions. J. Biol. Chem. 277, 17804–17810 (2002). | Article | PubMed | ISI | ChemPort |
  50. Tarui, T., Miles, L.A. & Takada, Y. Specific interaction of angiostatin with integrin alpha(v)beta(3) in endothelial cells. J. Biol. Chem. 276, 39562–39568 (2001). | Article | PubMed | ISI | ChemPort |
  1. Division of Hematology & Oncology, Department of Internal Medicine, UC Davis Cancer Center, University of California, Davis, 4501 X Street, Sacramento, California 95817, USA.
  2. Department of Dermatology, University of California, Davis, 4501 X Street, Sacramento, California 95817, USA.

Correspondence to: Kit S Lam1 e-mail: Kit.Lam@ucdmc.ucdavis.edu

MORE ARTICLES LIKE THIS

These links to content published by NPG are automatically generated.

NEWS AND VIEWS

Significance of one-bead-one-compound combinational chemistry

Nature Chemical Biology News and Views (01 Jul 2006)

Extra navigation

Subscribe to Nature Chemical Biology

Subscribe

See also

Search PubMed for

Open Innovation Challenges

naturejobs

More science jobs
Post a job for free