L3

^^ Tentagel Resin Bead

^ Tetrapeptide Libraries with 15 AA/position

Fig. 3.1.2. Forceps libraries for the H-RAS receptors.

In the protein screen we employed fluorescein-5-maleimide (FM) to conjugate fluorescein to purified glutathione S-methyl transferase (FM-GST) and a GST-H-[12Val]RAS fusion (FM-RAS) protein. First, about 7000 library beads were incubated extensively with FM-GST and the green fluorescent beads were removed to eliminate false positive binders. The remaining unstained beads were then incubated with FM-RAS. Among the approximately 7000 library beads used in the screening we found 65 relatively strong green beads. All positive molecules have either L2 or L3 as core template, and no molecules based on the L1-scaffold were found (Figure 3.1.2). From the thirty sequences we selected four molecular forceps, "MF4" through "MF7", for further studies. We decided to use one molecular forceps from a bead which failed to bind to FM-GST and FM-RAS ("MF8") as a negative control.

In the epitope approach, we used the His-tagged octapeptide derived from the CaaX-box of H-RAS. In non-exhaustive screening of the library we identified 15 molecules binding to the CaaX-box octapeptide; molecules with four peptide arms L3 predominated as binders. We have re-synthesized three examples of peptide binders, "MF1" through "MF3". Each of the re-synthesized molecules that had been selected against the peptide binds with micromolar affinities selectively to the H-RAS protein in the presence of other proteins such as ovalbumin or BSA, supporting our assumption that we can screen libraries with isolated peptide epitopes to identify forceps binding to the parent full-length protein.

The re-synthesized molecular forceps (MF1 through MF8, except MF5 which was not soluble) were tested for inhibition of farnesylation by yeast FTase [25]. As expected, MF8, our negative selection, did not inhibit farnesylation of H-RAS protein. MF6 and MF7 from the protein screening also had no effect; neither did MF1 that we had obtained from peptide screening. MF4 had a weak impact on the far-

Peptide lenders

Peptide lenders

Fig. 3.1.3. Sequences and inhibitions.

MF1: v-f-E-e-L3 MF2: V-E-F-E-L3 MF3: f-G-F-E-L3 MF4: p-E-K-S-L3 MF5: Q-E-k-p-L3 MF6: S-k-K-E-L3 MF7: f-p-K-s-L3 MF8: F-k-G-F-L3

Fig. 3.1.3. Sequences and inhibitions.

nesylation, with an IC50 higher than 1 mM. The two molecular forceps MF2 and MF3 from the screening against the peptide epitope resulted in the best inhibition (Figure 3.1.3).

As discussed above, FTase processes many cellular proteins. We reasoned that if our forceps inhibited the farnesylation of RAS by binding to its carboxy terminus, they might not impede the processing of other proteins by FTase. Satisfy-ingly, molecular forceps MF3 had different effects on the farnesylation by FTase of four CaaX-containing peptides derived from Lamin B, K-RAS B, and H-RAS, and a chimera of H-RAS and K-RAS sequences (''H-RAS + K-RAS'') (Figure 3.1.4a). Whereas MF3 had no effect on the modification of the Lamin B peptide at concentrations up to 1 mM, it had a weak effect on the K-RAS B peptide and a stronger impact on the H-RAS/K-RAS B chimera, correlating with the binding strengths to these substrates and the sequence similarities to the H-RAS sequence.

To confirm further that inhibition of farnesylation of GFP-CaaX-proteins is because of the binding of the molecular forceps to their CaaX-sequence, and not because of inhibition of FTase, we performed in-vitro binding assays with FTase. Whereas MF3 clearly interacts with H-RAS at a concentration of 250 mM, two and a half times the observed IC50, we did not observe binding to Ftase (Figure 3.1.4b). Because the forceps bind to RAS and fail to bind FTase, we concluded that MF3 could not act as an enzyme inhibitor, but that its activity is caused by its interaction with the substrate [26].

0 0

Post a comment