2~7~5 The invention is a metal chelating peptide (MCP) which can be unidirectionally and selectively coupled to a targeting molecule such as an antibody. The peptide of the invention has a plurality of metal binding sites attached to a branched portion thereof.
Metal ions including radionuclides may be useful in diagnosis and treatment of disease. For example, ~ or B
emitting radionuclides are useful for treatment of some cancers, and diagnostic tumor or thrombus imaging may be accomplished using paramagnetic metals in conjunction with magnetic resonance imaging or by using y-ray emitting isotopes for visualization with y-ray cameras.
Lanthanide metal fluorescence is also useful in diagnostic testing, and the ability to craate an electron dense targeting agent for electron microscopy has applications in basic research.
Delivering metal ions to a desired site, either in vivo or in vitro, and maintaining a sufficient metal ion concentration at the desired site for a sufficient time to accomplish the diagnostic or therapeutic objective, has been a goal of researchers for many years. The invention addresses this need by providing a metal chelating peptide that may take any of several structures depending on the intended application. The peptide has a reactive group at one end to unidirectionally, i.e.
irreversibly, couple the peptide to the targetinq molecule without affecting the inherent bioselectivity of the targeting molecule. The peptide is branched to provide a plurality of site~ for attachment of metal chelating groups such as ethylenedia~ine tetraacetic acid (EDTA), and can be constructed using d-amino acids in order to resist degradation. Importantly, the peptides of the invention are readily synthesized using standard techniques and may be customized in accordance with the needs of a particular application.
2~9~78~
Typically, the metal chelating peptides of the invention will be coupled to antibo~ies, but may also be coupled to drugs, oligonucleotides, or peptides or other targeting molecules. In constructing antibody conjugates, the peptides may be coupled to the Fc domain of the antibody, thereby not affect~ng the immunoreactivity of the antibody with antigen.
Accordingly, the invention provides a metal chelating peptide which can be coupled with a targeting molecule. The peptide has a reactive group at one end to allow unidirectional and selective coupling to the target molecule, and is branched at the oth~r end to provide a plurality of metal binding sites. A preferred reactive group for coupling the peptide is provided by a lysine residue having a free ~-NH2. Likewise, the branching of the peptide is preferably provided by a polylysine construct having free ~ and ~ NH2 groups to which metal chelating groups can be attached.
Brief ~escription of ~he Drawinq~
Figure 1 is a schematic representation of a general structure for an MCP of the invention.
Figure 2 is a schematic representation of a particular example of an MCP of the invention.
Figure 3 is a schematic representation of one possible pathway for coupling an MCP to an antibody.
Figures 4A-F show six possible derivatives of an MCP
for use in various coupling procedures.
While the invention has broad application, it will be described with reference to a preferred e~bodiment used for conjugation with antibodies such as monoclonal 2 0 9 ~ 7 8 ~
antibodies. Indeed, the driving force for the creation of metal chelating peptides arose from the desire to produce new generations of radioimmunoconjugates of monoclonal antibodies specific for tumor associated antigens.
With the advent of hybridoma biotechnology in the mid-1~70's, there has been a worldwide research effort to develop monoclonal antibodies to tumor associated antigens. In most cases, these antigens either represant altered forms of antigens present on normal cells or are the same antigens but are over expressed on cancer cells.
The therapeutic and diagnostic potential of monoclonal antibodies relates to their ability to bind tumor associated antigens. Antibodies themselves may have antitumor properties by virtue of immunological mechanisms such as antibody dependent cell mediated cytotoxicity, complement ackivation, and anti-idiotype activity. However, rapid antigenic modulation and the lack of measurable clinical benefits from the use of antibody alone as a therapeutic agent have caused researchers to turn to an examination of immunoconjugates.
To date, three major families of immunoconjugates have been developed: drug-antibody, toxin-antibody and radionuclide-antibody conjugates. Of these, only the radioimmunoconjugates do not require endocvtosis or binding to all tumor cells to deliver their toxic action.
Additionally, radioimmunoconjugates are usèful diagnostic tools, for example, in locating tumors.
Radioimmunoconjugates have been prepared for both therapeutic and diagnostic purposes; however, the success of such conjugates has been minimal owing to a number of factors which the present invention addresses. One problem arises from the use of monoclonal antibodies 2~9~178t~j which recognize poorly defined epitopes often located on the surface of both normal and malignant cells. Thus, other targeting agents may prove more suitable for conjugate formation. For example, peptide hormones, growth factors and lymphokines bind to defined cell surface receptors and may prove to be superior candidate~
to antibodies for conjugate construction in many instances.
Another problem associated with the use of radioimmunoconjugates is the need for the targeting agent to retain its binding capacity to a tumor cell marker after having been modified to carry an effective therapeutic or diagnostic probe. Past experience has shown that antibodies have often been poorly labelled with radioisotopes, thereby providing radioconjugates which are too weakly radioactive to be useful, or in which the metal is not adequately complexed leading to release of radioactive elements in vivo, or which are denatured during the conjugation process so that binding to the target antigen is impaired.
The metal chelating peptides of the invention are readily synthesized using existing or modified procedures of classical solid phase peptide synthesis. In addition to being rapid and simple, solid phase peptide synthesis permits the removal of unreacted reagents and by-products at each step of the synthesis. Also, each synthetic step may be repeated until satisfactory yields are obtained~
The peptides of the invention are branched to provide clusters of metal binding sites. These peptides may be customized for various diagnostic and therapeutic applications in association with a wide range of targeting agents.
As shown in Fig. 1, the first step in the :
~ ~, ",.. .
209~7~
construction of a metal chelating peptide of the invention is to couple at least one amino acid, and preferably for most purposes several amino acids, to a solid support for the purpose of providing a molecular arm Xl which separates the metal chelating portion of the peptide from the targe~ing agent coupling portion of the peptide. This C-terminal end of the peptide Xl has a reactive moiety permitting the unidirectional coupling of the peptide to a targeting agent. Thus, the molecular arm Xl may have a free carboxyl or amide end and may have a hydrazide, a~ide, activated ester, thiol, aldehyde, amino, maleimido, isothiocyanate, haloacetate, photo-reactive or other reactive group capable of effecting the desired unidirectional coupling to the targeting agent.
The second part of the peptide of the invention is a branched structure formed from diamino carboxylic acids such as lysine, ornithine, or 1,3-diaminopropionic acid.
Referring to Fig. 1, the branched structure is connected to the C-terminus Xl via the spacer X2 which may be omitted or may be up to a few amino acids in length depending on the spacer requirements for separating the branched portion of the peptide from the targeting agent coupling site at the C-terminus Xl.
In Fig. 1, X3 represents a branch forming amino aaid, preferably a diamino acid such as lysine, ornithine or 1,3-diaminopropionic acid. The group X4 in Fig. 1 provides the capability for the attachment of a plurality of chelation sites X5. The preferred branch forming amino acid is L-lysine; however, the use of D-lysine and D-amino acids for construction of the peptide will reduce the rate of degradation of the peptide in vivo, and the use of shorter diaminocarboxylic acids for branching produce a more compact structure.
When the desired number of amino groups have been .
-6- 2 ~ 8 ~
assembled onto the peptide, a metal chelator X5 is attached to each amino group. A preferred metal chelator is ethylenediamine tetraacetic acid ~EDTA), which may be coupled to the amino end of an amino acid by using the derivative I, the simple Il - 0~
O ~ O
~ 11 / 11 .
t - c~ rc o~
N N
> ~ 11 ot H0 - C o O
synthesis of which is reported in Arya, R. and Gari~py, J. (1991) Bioconjugate Chemistry 2, 323-326. The advantage of using an EDTA derivative of the type I is that the full EDTA structure is available for metal chelation as a hexadentate ligand. The skilled person will appreciate that a similar chelation strategy can be employed using other chelators such as diethylene-triamine pentaacetic acid.
The following is a typical synthetic scheme for preparation of a peptide of the invention. The peptide, termed MCP-8-EDTA-NH2, is shown in Fig. 2.
Preparation of Metal Chelating Peptide, MCP-8-EDTA-NH2 .
The peptide shown in Fig. 2 was prepared by solid phase peptide synthesis on an automated Applied Biosystems model 430 Peptide Synthesizer using t-Boc protected amino acids and PAM resin supports. Unless ':, .
.
- ~
2~7~
indicated, all couplings were carried out using symmetric anhydride derivatives of protected amino acids. All synthesis protocols employed were those established ~y the manufacturer (Applied Biosystems, Foster City, ~A).
Each coupling step was monitored by the quantitative determination of free amino yroups present on the resin (quantitative ninhydrin test). Typically, the efflciency of each coupling step was close to 99%O As illustrated in Figure 2, the first re~ e coupled to the PAM resin was B-alanine (B-aminopropionic acid) and the substitution on the resin support was 0.l mmole/gram resin. The initial low substitution value on the resin insures that crowding of the resin with peptide chains will not occur as a result of three branching steps li~e.
maximal substitution of 23 x 0.1 mmole = O.8 mmole/gram resin). The B-alanine serves as an internal standard.
The second ~esidue is a lysine (N~-Boc, N~-FMOC-Lysine) having its ~-amino group protected with a 9-fluoromethoxycarbonyl amino group (FMOC). The base-sensitive FMOC ~roup provides an option to introduce a reactive site at this position on the resin prior to chain cleavage from the resin. A more classical lysine derivative with an acid labile side chain protecting group (cleavable with H~ 2 ClZ derivative for example) has also been used. A protected cysteine residue (for example Cys (Acm)) could have been introduced at this position if a thiol group would have been preferred later for conjugation purposes (for labelling mAb fragments for example). The third and fouxth residues coupled were alanine and tyrosine respectively. These residues act as molecular spacers between the paptide branches and the C-terminus reactive group. They also permit one to monitor the quality of the synthesis when samples taken at each step during the construction of the polymer are analyzed by amino aaid analysis. In addition, the aromatic side chain of tyrosine allows for the determination spectroscopically the concentration of the peptide stock .
' . .
.
-8- 2~9~78~
solutions and can be radioiodinated providing an approach to determine the number of peptides coupled to anti~odies. The first through fourth residues here correspond to the - ~ - ~ chain of Fig. 1. The fi~th residue was N~,N~-Boc-lysine(Boc), an amino acid having its amino groups at the C~ and CE positions protected with acid labile ~oc protecting groups. After deprotecting these sites with TFA, branchinqLls initiated by coupling two equivalents of N~,N~-Boc-lysine(Boc) to the two available amino positions. After another round of acid deprotection, the branchin~ step was repeated with this time four amino sites available for coupling.
N~,N~-Boc-lysine(Boc) and glycine(Boc) were successively coupled again. Upon removal of the Boc groups, eight new amino sites were exposed. The t-butyl protected EDTA
carboxy derivatives I were then coupled to these amino sites (2 molar excess over the number of moles of free amino groups) in the presence of an equivalent molar amount each of dicyclohexylcarbodiimide and N-hydroxybenzotriazole in 50% DMF/DCM. The re~ction was performed at room temperature for 2 hours. This coupling step was repeated and the completion of this reaction~a~
monitored by quantitative ninhydrin proved to be greater than 99%. Any remaining unreacted amino groups were acylated with acetyl anhydride in the presence of diisopropylethylamine. The t-butyl groups were cleavad in 50% TFA/DCM for 2 hours and the protecting groups removed in the presence of trifluoroacetic acid (TFA~.
The peptide-resin was then treated with 20~
piperidine/DMF for 20 minutes to deblock the FMOC group present on the ~-amino group of the C-terminus lysine.
Finally, the MCP-8-EDTA-NH2 (Fig. 2) was cleaved from the support with 10% anisole/anhydrous hydrogen fluoride at 0 C for 45 minutes. The resin was extracted with several ether washes to remove the anisole and the cleaved protecting groups. The branched peptide was then recovered by extracting the resin with 100~ TFA and : ' ' . ~ ' . ~ , ' . ' :. .' ', ', : ' 2~78~
rapidly evaporating the acid. The peptide was then suspended in 50 mM ammonium bicarbonate and lyophilized.
The skilled person will appreciate that the number of amino groups available for the introduction of chelator sites can be increased readily to 16 or more, with the upper limit depending on the ability to resolve problems of low coupling rates associated with the crowding of the resin support and ~he accessibility of free amino groups. The rapidity of the synthesis strategy (two days) and its potential for automation represent major assets in developing the technology.
With an expected molecular weight of 4307 daltons, MCP-8-EDTA-NH2 is desalted initially on a Sephadex G-10 column to remove any low molecular weight impurities resulting from the cleavage step. The recovered polymer has 8 (+/- 2) EDTA groups and the proper amino acid composition. The chelating properties of MCP-8-EDTA-NH2 has also been confirmed by the binding of radionuclides such as 99mTc.
Cou~lina of MCP-8-EDTA to the ca~bohydr~te domain of monoclonal antibQdies Aldahyde groups can be generated on monoclonal antibodies by the periodate oxidation of carbohydrate moieties located in the Fc domain. These sites have been recently exploited to selectively label antibadies and glycoproteins with hydrazide derivatives resulting in a minimal loss of activity.
Introducing a maleimide functionality on MCP~8-EDTA-NH2 A number of strategies can be devised to couple peptides of the invention to antibodies. This flexibility is one of the powerful feat~res of the invention. Fig. 4 illustrates 6 examples of how simple ~0~78~
modifications of the peptide can lead to alternate approaches for coupling it to a targeting agent. In the selected example outline in Fig. 3, MCP-8-EDTA-NH2 is first reacted with the bifunctional crosslinking agent, m-maleimido-benzoyl-N-hydroxysulfosuccinimide ester (sulfo-MBS; Fig. 3, pathway a). Typically 3 mg of sulfo-MBS dissolved in PBS is mixed with 1 mg of the peptide dissolved in PBS. The reaction is left ~o process at room temperature for 1 hour with occasional stirring.
The polymer-MBS is desalted from unreact~d sulfo~MBS by gel filtration using Sephadex G-25. Using the reactivity of the resultant conjugate in a quantitative DTNB/B-mercaptoethanol assay, it has been confirmed that one mole of MBS is coupled to one mole of the peptide under these conditions. This compound, abbreviated as MCP-8-EDTA-NH-MB, is stabl~ for days in aqueous buffers (PBS) at 4 C.
Construction of MCP-antibody_coniuqateS
The maleimide derivative of the peptide (MCP-8-EDTA-NH-MB) selectively and rapidly reacts with any free thiols present on the surface of targeting agents. As an example, thiol groups can be created on an antibody by partially reducing with DTT one or more of the disulfide bridges present at the hinge region of IgG molecules.
Alternatively, the Fc domain of antibodies can be targeted by oxidi2ing the carbohydrate moieties located on each heavy chain to create free aldehyde groups (Fig.
3). Indeed, in the pr~sence of sodium periodate, aldehyde groups are generated which can then be readily reacted with a thiol-~ontaining hydra2ide compound such as 2-acetamido-4-mercaptobutyric acid hydrazine (Fig. 3, pathway b). The resulting thiol-containing antibody preparation can then be reacted with MCP-8~EDTA-NH-MB
(50-~old excess in relation to the number o~ moles o~
antibody) overnight at room temperature in an aqueous . .
':
. , :' ~ "
2~9~78~
buffered solution adjusted to a pH between 4 and 8 ~Fig.
3, pathway c). The complex is then desalted on Protein PakI 125 (Wat:ers; Millipore) HPLC gel filtration column from the unreacted ~CP-8-EDTA-NH-MB.
Construction of MCP-antibody fragment conjuqates Amino groups on F(ab')2 fragments can be linked through a bifunctional crosslinker to the MCP-B-EDTA
carrying a cysteine moiety (Fig. 4A as an example), while the single reactive thiol group of F(ab') fragments will be coupled to the peptide carrying a lysine group ~Fig.
4B as an example). A variety of commercially available heterobifunctional crosslinkers containing a maleimido group at one end (reacts with thiol group) and a succidimidyl ester group at the other (reacts with amino groups) can be used to link together MCPs and mA~
fragments. In the case of F(ab')2-MCP conjugates, the antibody fragment will first be reacted with MBS and the resulting construct purified by gel filtration (Sephadex G-50). Excess MCPs (50-fold molar excess over amino groups on fragment) containing a free thiol group is then added to the desalted fragment-MBS conjugate and the reaction allowed to proceed for 2 hours at room temperature (with occasional stirring). In all cases [F(ab') and F(ab')2 conjugates], the final conjugates can be purigied from the excess polymer by gel filtration (Sephadex G-50).
VariatiQns of the_,ba~ic concept_lead to the desiqn of a multitude of MCPs Fig. 4 illustrates six approaches to link MCPs to targeting agents. In Fig. 4, only a few variations of the basic polymer are shown that would permit an MCP to be coupled unidirectionally to amino, thiol, or carbohydrate groups present on a targeting agent. Non-2~9~78~
specific labelling strategies such as the use of photoreactive groups (aryl azides or benzoyl benzoate derivatives, for example, Fig. 4F) are also illustrated.
From the foregoing description and examples, the skilled person will appreciate that the invention is of broader scope than the specific peptides illustrated. A
significant strength of the invention is its versatility and wide applicability to a variety of metal chelating applications.
.