| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online November 14, 2000 as doi:10.1096/fj.00-0422fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Laboratory of Experimental Oncology, National Institute for Cancer Research, 16132 Genoa, Italy;
Department of Experimental Medicine, Biochemistry Section, University of Genoa, Genoa, Italy;
Department of Experimental Medicine, Section of Human Anatomy, University of Genoa, Genoa, Italy;
§ Immunopharmacology Section, Advanced Biotechnology Center, Genoa, Italy; and
¶ Department of Oncology, Biology and Genetics, University of Genoa, 16132 Genoa, Italy
2Correspondence: Laboratory of Experimental Oncology, National Institute for Cancer Research, Largo R. Benzi 10, 16132 Genoa, Italy. E-mail: parodis{at}hp380.ist.unige.it
SPECIFIC AIMS
What we specifically tried to achieve here lies within a more general strategy for finding innovative antineoplastic drugs. The major points that characterize this new type of strategy are:
1. Choice of a reductionist target, like the protein of a major oncogene, as a primary screening object, rather than the global behavior of a rapidly proliferating cancer cell. The availability of a large body of knowledge about the selected target signaling protein and about the signaling proteins that make cross-talks with it appears useful for the reasons mentioned below (points 3 and 4).
2. To take full advantage of the recently offered opportunity of
internalizing a variety of molecules of several thousands daltons
carried inside the cell by a family of basic vector peptides, like the
III 
-helix of Antennapedia, a basic 11 amino acid-long motif from
the transcription-activating factor Tat of HIV, or a peptide made up of
a few clusters of up to nine arginine amino acids per cluster (as
recently reported at the 219th meeting of the American Chemical
Society).
3. Taking advantage of point 2, search for possible leads not among
small, very compact drug-like molecules of 
500 Da, but rather as an
inspiration and a lead among larger peptidic motifs of proteins
involved in cross-talks with our target signaling protein (or motifs of
the target protein itself).
4. Recent progress in computer modeling and reasoning around protein structure combined with technological progress in stepwise solid-phase peptide synthesis, with the consequent possibility of building and screening rapidly small peptidomimetic combinatorial libraries built around the initial inspiring peptidic motif, make the approach of point 3 potentially much more powerful than just 2–3 years ago (Chem. Biol., vol. 7, pp. 245–251, 2000; Nature (London), vol. 404, pp. 715–718, 2000).
5. Within the more general framework described above, the specific aim of this study was to build on our initial report of a Myc inhibitor peptide (Cancer Res., vol. 58, pp. 3654–3659, 1998). Here we tried to find a more potent and stable second-generation peptidomimetic molecule still capable of crossing mammalian cell membranes. We also wanted to show that the biological activity of this family of molecules was not restricted to a single type of cellular target.
PRINCIPAL FINDINGS
1. Antiproliferative and other biological activities toward a mammary cancer line (MCF-7 cells) originally reported for the L peptide Int-H1-S6A,F8A, containing an H1 of Myc motif and capable of internalization (Cancer Res., vol. 58, pp. 3654–3659, 1998), were confirmed and extended to a colon cancer line (HCT-116 cells) and to human peripheral blood lymphocytes (PBLs) stimulated with phytohemagglutinin (PHA).
2. We were able to move from a lead L peptide to a more interesting peptidomimetic molecule. We found that a corresponding retro-inverso (RI) peptide (RI-Int-H1-S6A,F8A) of the above H1 of Myc motif was also active; it was indeed 5- to 10-fold more active than the original L peptide and 30- to 35-fold more stable.
3. We confirmed that the similarity in terms of steric structure between a motif made of L-amino acids and its corresponding RI peptide can translate from the structure level to the level of molecular and biological activities.
4. According to our results, in all three cellular models the presence
of the third -helix of Antennapedia protein is necessary and
sufficient for internalization. This basic peptidic sequence was also
active in its RI variant, made of D-amino acids in the reverse order.
5. We synthesized iso-amphipathic-modified H1 of Myc motifs in both their L (Int-H1isoamph) and RI forms (RI-Int-H1isoamph). Among a set of 73 human helix-loop-helix (HLH) sequences, H1 motifs of the Myc family were still the closest to our two modified motifs (proximity=53%).
6. In all three biological systems studied (following paragraphs a, b, c), Int-H1isoamph or RI-Int-H1isoamph were at least three- to fivefold less active than their corresponding reference peptides, whereas lack of the H1 motif implied substantial lack of biological activity.
a) After treatments on days 1, 4, and 7 in MCF-7 cells, during the last 3 days of growth the number of cells counted for the L-iso-amphipathic-modified peptide was always three- to fivefold higher than in cells treated with Int-H1-S6A,F8A. This difference was qualitatively relevant and statistically significant (P < 0.002, one-tailed). As expected and reported previously (Cancer Res. vol. 58, pp. 3654–3659, 1998), the Int sequence from Antennapedia was practically inactive.
Things were even more clear-cut when MCF-7 cells were tested with our RI peptides. For instance, whereas RI-Int-H1isoamph 5 µM would increase the doubling time of only 23%, the same concentration of RI-Int-H1-S6A,F8A increased the doubling time of 231% (10-fold more!). This difference was statistically extremely significant (P < 5 x 10-5, one-tailed).
The results obtained in MCF-7 cells using both L and RI peptides are
reported in Fig. 1
.
|
b) The relevance of the presence of a Myc H1 motif capable of internalization for inhibiting growth of cancer cells hyper-expressing c-Myc was completely confirmed in the colon cancer line HCT-116. Int alone (lacking the Myc H1 motif) and H1-S6A,F8A (lack of the internalization sequence) were completely inactive. Int-H1isoamph was about 10-fold less active than Int-H1wt or Int-H1-S6A,F8A. Even these differences were extremely significant (P < 10-4, one-tailed).
c) We investigated the activity of our peptides in a system of normal human cells that is well known to display a peak of c-Myc activity after PHA stimulation and is sensitive to c-myc mRNA inhibition by antisense phosphorothioates, followed by downstream inhibition of 3H-thymidine incorporation. Inhibition of 3H-thymidine incorporation by the two L peptides capable of internalization and containing the c-Myc H1 motif (Int-H1-S6A,F8A and Int-H1wt) was always at least threefold stronger than inhibition by Int-H1isoamph (difference significant at P < 10-3, one-tailed). Int (no H1 motif) and H1-S6A,F8A (no internalization) were practically inactive. In this model of normal human cells, the L peptides Int-H1-S6A,F8A and Int-H1wt were equally active. This is similar to what we observed for the colon cancer line HCT-116, whereas Int-H1wt displayed a much weaker activity than Int-H1-S6A,F8A in the mammary cancer line MCF-7. We do not understand the reason for this discrepancy.
Working with RI peptides, potency on a molar basis was at least threefold higher than for L peptides, and again RI-Int-H1-S6A,F8A was about threefold more active than RI-Int-H1isoamph (difference significant at P < 10-3, one-tailed).
The results obtained in human PBLs using both L and RI peptides are
reported in Fig. 2
.
|
7. A semi-quantitative reverse transcription-polymerase chain reaction was performed on RNA extracted from PHA stimulated PBLs treated with different peptides. Two bona fide c-Myc target genes—ODC and p53—were assayed in comparison with the housekeeper gene GAPDH, which is not insensitive to stimulation of cell division. GAPDH is involved in fragmentation of exoses to trioses and downstream production of energy. Both ODC and p53 responded mainly to our active L peptides Int-H1-S6A,F8A or Int-H1wt, more weakly to Int-H1isoamph, and just slightly to the control L peptides Int (lacking any kind of H1-related sequence) or H1-S6A,F8A (lacking an internalization sequence).
As expected considering that we had to use a lower concentration, RI
peptides appeared more potent than the corresponding L peptides. Again,
RI-Int-H1-S6A,F8A was more active than its corresponding
iso-amphipathic-modified peptide. The target gene p53 tended to be more
sensitive to inhibition than ODC. It was reassuring to notice that
3H-thymidine incorporation and mRNA levels were
rank correlated reasonably well (Spearman’s rho test), especially for
ODC (P 0.002) but also for p53 (P 0.021),
whereas GAPDH was essentially unaffected, even if this last gene is
probably useful for burning the exoses and supplying energy to dividing
cells.
CONCLUSIONS AND SIGNIFICANCE
It would be difficult not to seriously consider the possibility
that the effects we have observed and reported here, together with our
previous results, suggest that the antiproliferative activities
registered are indeed related to the presence in our active molecules
of a specific Myc motif. The molecular interference probably is taking
place at the interior of the boundary of the set of related molecules
we can refer to as the Myc/Max family, and we suspect that this happens
at the level of the common domain present in these homo/hetero-dimers,
the four -helix bundle, because H1 of Myc belongs to this
domain. We could be seeing a direct disturbance in the formation of the
bundle, but also interference at the level of the enhanceosome at the
interface between the bundle and surrounding proteins of the higher
order structure. If these protein–protein interactions with the
outside of the bundle were less strong than the cooperative forces
keeping the hetero-dimer together, they could be even more sensitive to
an excess of competing presence of our peptides.
We propose a general strategy of pharmacological attack of
intracellular protein functions not through a small ( 500 Da) and
compact drug-like molecule, but through a larger structure (on the
order of 3500 Da in this case, but we have started to move in the
direction of doubling the size of our peptidomimetic molecules). In
fact, we have just synthesized and started to study peptidomimetic
molecules containing the [H1,Loop,H2] motif of c-Myc.
If it becomes possible to use a retro-inverso conformation of Int (or
the like) to introduce interacting motifs at least 5- to 10-fold more
extended in terms of interaction surface space than the usual size of a
typical ‘drug-like molecule’, we can hope to achieve a ‘broader
vision’ of the target. In addition, structural motifs used by natural
signaling proteins for their cross-talks can become an useful
inspiration, especially if RI--helices resistant to the peptidases
(and other peptidase resistant peptidomimetic motifs) can be used as a
starting point for finding further improvements in terms of affinity
and selectivity. It might be useful to have a molecule able to ‘see’
at least (25–50 Å)2 of the surface of a target
protein instead of the usual (5–10 Å)2 ‘seen’
by a typical more compact and smaller ‘drug-like molecule’.
Perhaps it could be a limitation/disadvantage for several
varieties of combinatorial chemistry approaches to generate libraries
of compact ‘drug-like molecules’ endowed with an enormous diversity
of multiple inter-actors, both in terms of combinations, space
distribution, and orientation, but in too limited/confined a space. We
could advance the consideration that the typical informational content
of signaling proteins is extremely high, but often more diluted in
space than the informational content of drug-like molecules. We are
curious about the potential of relatively small peptidomimetic
combinatorial libraries not starting from scratch, but implanted over
the scaffold of a lead peptidic motif suggested by nature (Chem.
Biol., vol. 7, pp. 245–251, 2000), guided by intelligent
reasoning and adequate computer modeling. A synthesis of the conceptual
background of our line of investigation is given in Fig. 2
.
FOOTNOTES
1 To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.00-0422fje To cite this
article, use (November 14, 2000) FASEB J. 10.1096/fj.00-0422fje 
This article has been cited by other articles:
|
|
 |
|
  C. Avignolo, L. Bagnasco, B. Biasotti, A. Melchiori, V. Tomati, I. Bauer, A. Salis, L. Chiossone, M. C. Mingari, P. Orecchia, et al. Internalization via Antennapedia protein transduction domain of an scFv antibody toward c-Myc protein FASEB J, April 1, 2008; 22(4): 1237 - 1245. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Bagnasco, L. Tortolina, B. Biasotti, N. Castagnino, R. Ponassi, V. Tomati, E. Nieddu, G. Stier, D. Malacarne, and S. Parodi Inhibition of a protein-protein interaction between INI1 and c-Myc by small peptidomimetic molecules inspired by Helix-1 of c-Myc: identification of a new target of potential antineoplastic interest FASEB J, April 1, 2007; 21(4): 1256 - 1263. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. R. Eberhardy and P. J. Farnham Myc Recruits P-TEFb to Mediate the Final Step in the Transcriptional Activation of the cad Promoter J. Biol. Chem., October 11, 2002; 277(42): 40156 - 40162. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
