Inhibition of Collagenase with Triple Helical Peptide Mimics
IGERT Participant: John K. Whitehead
This project will: consist of the synthesis of triple-helical collagen substrate analogs and peptide followed by their purification, characterization, and verification of inhibition, and structure determination of collagenases with verified inhibitors.
Primary faculty co-advisors:
Robert P. Hammer (Peptide Synthesis and Phosphorus Chemistry)
Marcia Newcomer (Protein Crystallography)
Off campus participant: Gregg B. Fields, Florida Atlantic University (Enzymology)
Technical Proposal: Collagenase is a triple-helical enzyme belonging to the matrix metalloproteinase family. This family is mechanistically involved in a variety of diseases such as arthritis, atherosclerosis, and tumor cell metastasis. The goal of this project is to synthesize a triple-helical peptide mimic, which will successfully inhibit the enzyme. The success of inhibition will be determined though assays and the design of the mimic will be fine-tuned along the way.
Background:
Two classes of proteases, the aspartyl proteases and the metallo(zinc)-proteases use an acid catalyzed addition of water as one of the steps of the amide bond hydrolysis. The tetrahedral intermediate that results from water addition to the amide carbonyl has been the focus of many protease inhibitor designs and has given rise to two robust classes of inhibitors, namely the statines and phosphorus-based amide bond replacements such as phosphonates (Y = O), phosphinates (Y = CH2) and phosphonamides (Y = NH). Both the statine and phosphorus mimics are particularly attractive from an initial design point of view, because if you have a good substrate for your enzyme, one can almost invariably produces an inhibitor of the enzyme by the statine (mainly for aspartyl proteases) [Babine 1997] or phosphorus mimic (aspartyl and Zn proteases) approaches [Bartlett 1996]. Several laboratories have investigated phosphorus-based mimics for MMPs and found that both phosphonate [Gavuzzo 2000, D’Alessio 1999, Gallina 1999, Cirilli 1997, Mookhtiar 1987] and phosphinate [Cursio 2002, Shiot 2001, Buchardt 2000, Buchardt 1999, Caldwell 1996] peptide mimics are effective inhibitors of MMPs. Thiophosphonates and thiophosphinates [Nishino 1978, Grobelny 1985] have been studied less, but hold great promise as inhibitors due to the increased affinity of the sulfur atom for zinc in the enzyme active site and the increased hydrophobicity of the thio-derivatives.
Methods:
As both phosphinate and phosphonate mimics have been effective at inhibiting MMPs, we propose to make both phosphinate and phosphonate triple-helical collagen substrate analogs described in Table 1. Additionally we will make the thiophosphinate and thiophosphonate versions of peptide mimics which show promise as triple helical inhibitors. The MMP cleavage sight in each of sequences has been replaced by either a Gly-Leu phosphorus mimic or a Gly-Val phosphorus mimic (for the a1(V)436-450 sequence).
Figure 1 shows the typical sequence that will be prepared using stepwise solid-phase peptide synthesis (Fields ref) except that the Gly-Leu or Gly-Val phosphorus analog will be inserted as a phosphorus ester-protected, Fmoc-protected dipeptide 1 or 2.
Table 1. Phosphorus-based Tetrahedral Intermediate analogs of Collagen Cleavagea
Sequence Name |
Phosphinate or Phosphonate Analog of Cleavage Sequence |
|
a1(II)769-783 |
-Gly-Pro-Pro-Gly-Pro-Gln-Glyy[P(O)(XH)-Y]Leu-Ala-Gly-Gln-Arg-Gly-Ile-Val- |
|
a1(V)436-450 |
-Gly-Pro-Pro-Glyy[P(O)(XH)-Y]Val-Val-Gly-Glu-Gln-Gly-Glu-Gln-Gly-Pro-Pro- |
|
a1(IV)1366-1377 |
-Gly-Pro-Pro-Glyy[P(O)(XH)-Y]Leu-Lys-Gly-Leu-Gln-Gly-Leu-Pro- |
|
a1(IV)1426-1437 |
-Gly-Pro-Asp-Glyy[P(O)(XH)-Y]Leu-Pro-Gly-Ser-Met-Gly-Pro-Pro- |
a When X = O: Y = CH2, phosphinate; Y = O, phosphonate.
When X = S: Y = CH2, thiophosphinate; Y = O, thiophosphonate.
Figure 1. Typical phosphinate collagen substrate analog to be prepared by solid-phase peptide synthesis.
The synthesis of phosphinate mimic is shown in Scheme 1. The procedure follows the recently developed synthesis devised by Yiotakis and coworkers [Georgiadis 2001a]. First the phosphinate analog of glycine 4 is prepared by the method of Baylis [1984] from the imine cyclotrimer 3 and hyphophosporus acid followed by hydrolysis to cleave off the benzhydryl group. Next it is protected as the Fmoc derivative to give 5. Mild silylation of the phosphinate with TMS-Cl and DIEA gives the trivalent phosphonite which reacts readily in a Arbusov like reaction with allyl acrylate 6 (R = iPr [Huntington 2000] or R = iBu [Mori 1970]) to give the phosphinic acid 7. The acrylate esters are readily available from the nucleophilic esterification of the parent acids with allylbromide under phase-transfer conditions [Friedrich 1989]. Finally the phosphinate is protected as the adamantyl ester by reaction with silver oxide and adamantylbromide (Ad-Br) and the terminal allyl ester removed by Pd(0) catalysis to give the protected phosphinate dipeptide mimic 1. The thiophosphinate mimic is made by first activating the phosphinate to the acid chloride with thionyl chloride and then coupling with adamantanethiol which is then again deblocked on the C-terminus with Pd(0) catalysis. Several investigations have shown the adamantyl ester to be the best choice as it doesn’t prematurely hydrolyze but it is readily cleaved under the TFA-based final deblock conditions used in Fmoc/tBu solid-phase strategies [Georgiadis 2001b, Yiotakis 1996]. The scheme shown below has given overall yields of 70-80% from compound 5 to the desired product 1 for very similar substrates.
Scheme 1. Synthesis of phosphinate dipeptide mimic.
Scheme 2 shows the preparation of the phosphonate dipeptide analog. Starting from the Fmoc-protected glycine phosphinate 5, the a-hydroxyacid 8 is coupled to it using water-soluble carbodiimide (EDC) and a catalytic amount of dimethylaminopyridine (DMAP) to make the H-phosphinate ester 9 [Karanewsky 1986]. The a-hydroxyacids are readily available from the corresponding amino acids by diazotization in water [Campbell 1992]. These can then be converted to the corresponding allyl esters by reaction with allylbromide under phase-transfer conditions [Friedrich 1989]. This carbodiimide-mediated coupling reaction for sterically unhindered phosphinates like 5 is very high yielding. Next, using methods developed in the Hammer laboratory [Rushing 2001, Fernandez 1995], the H-phosphinate 9 can be non-oxidatively activated to the phosphonochloridite 10 with dichlorotriphenylphosphorane and then coupled with adamantyl alcohol and oxidized to give the fully protected phosphonate derivative, which upon Pd-catalyzed removal of the allyl ester gives the desired phosphonodipeptide 2. Alternatively the H-phosphinate ester 9 could be oxidized by periodate [Karanewsky 1986] and then esterified by the Ag2O/Ad-Br method described above [Georgiadis 2001a] to give fully blocked phosphono-dipeptide 11. This is again readily converted to desired free acid 2 by Pd-catalyzed removal of the allyl ester. The thiophosphonate can be prepared by sulfurizing compound 9 with sulfur under basic conditions and the resultant thiophosphinate can be alkylated Ag2O/Ad-Br as before. Deprotection of the allyl ester then gives the thiophosphinate protected derivative 2 suitable for solid-phase peptide synthesis.
Scheme 2. Synthesis of protected phosphonate dipeptide mimic.
The first peptide to be prepared will be the a1(II)769-783 sequence. Initially, just the central 769-783 sequence will be prepared both for purposes of methodology development and as a benchmark inhibitor for a single stranded phosphonopeptide mimic of this sequence. The coupling of the phosphinate mimic 1a will be monitored by determination of Fmoc group upon coupling. Once it is established that the phosphinate mimic can be incorporated and purified intact, then the full a1(II)769-783 sequence will be prepared with 5 Gly-Pro-Hyp repeats on both the N- and C-termini and a hexanoic acid coupled to the N-terminus. This sequence will be purified by HPLC and then checked for triple helicity using our standard protocol with denaturing CD (GREGG REF). Adjustments will be made as need be with longer or shorter repeats on the termni as needed to produce stable triple-helical tris-phosphinate mimics. Another factor that could be varied to provide triple helicity is the length of the alkyl chain on the N-terminus. As the phosphorus mimics will be more hydrophilic than their corresponding substrates, some additional hydrophobicity in the tail may be tolerated before aggregation and precipitation make purification intractable. For all the sequences shown in Table 1, first the short sequences (no Gly-Pro-Hyp repeats) will be prepared and then the “full-length” sequences with appropriate Gly-Pro-Hyp repeats will be made. Both the shortened and full-length phosphinates will be assayed with various MMPs for inhibition. Once phosphinates are made and assayed, the phosphonates will be prepared using the phosphonate mimic 2a and 2b. Lastly, thiophosphinate and thiophosphonate inhibitors will be made of the derivatives 1c, 1d, 2c and 2d (as appropriate) that show the most selective inhibition.
Once we have identified triple helical sequences that are good inhibitors (<100 nM), we will assess the role that residues in the P3, P2, P2' and P3' play in both affinity and selectivity of MMP 1 and MMP-13 inhibition by doing scans of Lys, Tyr, Leu, and Gln of each position. We have chosen these particular residues as examples of classes of either charged (Lys), aromatic (Tyr), hydrophobic (Leu), and polar (Gln) residues. The fifteen-sixteen analogs of each sequence (Table 2) will be prepared with the most promising mimics (phosphinate, phosphonate, thiophosphinate, thiophosphonate) using the Multiple Peptide Synthesis (MPS) accessory on the Pioneer peptide synthesizer in the LSU Peptide Facility. The MPS accessory can do up to sixteen sequences at once at a scale of 0.1 mole. Each derivative will be assayed for triple helicity, and then assayed for enzyme inhibitory activity as appropriate. Where increases in inhibition and particularly selectivity of inhibition are found, further more detailed scans (up to all 20 residues at a particular position) will be performed at these positions and perhaps at P4 and P4'. Finally, we may also look at varying the side-chain at the P1' sight by preparing additional phosphorus peptide mimics according to Scheme 1 or 2 as appropriate. Enzyme assays will be done in collaboration with the Fields laboratory with an expectation of bringing assay techniques
Additionally, MMP-1 and MMP-13 clones will be obtained from the Fields laboratory and these proteins will be expressed in E. coli in the Newcomer laboratory according to standard protocols. A variety of crystallization conditions of MMP-1 and MMP-13 with the verified inhibitors will be attempted using parallel crystallization conditions. S
Number of IGERT apprentices to be recruited and probable home departments: A minimum of one from the chemistry or biological sciences department as well as one at Florida Atlantic University with other positions possible.
Consistency with the Macromolecular Education, Research & Training theme: This project requires a complete understanding of polymers in the form of peptides and proteins and advanced instrumental methods for their isolation, purification, and characterization.
How does the project form a vector cross-product of existing research themes by the participants? Hammer’s expertise of peptide synthesis, Fields’ knowledge of MMP inhibition, and Newcomer’s protein crystallography skills will be combined in a new research direction which will ultimately yield a variety of inhibitors of collagenase and insight into the structural determinants of collagenase substrate recognition.
How do students benefit from the team-oriented research, beyond what would be available to them from either advisor separately? In addition to learning organic synthesis, this project will enable the student to study from experts in hot areas such as solid-phase peptide synthesis, enzymology, protein expression, and protein crystallography.
Commitment of faculty & off-campus participants to work side-by-side with apprentices: All three of the faculty members are open to working with the student personally to complete their respective leg of the project.
Briefly describe the support level available to each individual faculty or off-campus participant (i.e., without IGERT)
Table 2. Scanning mutagenesis of the P3, P2, P2', and P3' sights of collagen inhibitors.