WO2014149440A1 - Autophagy-inducing peptide analogs - Google Patents

Abstract

An autophagy-inducing compound comprises an autophagy-inducing peptide comprising Beclin 1 peptides immediately N- and C-terminally flanked by moieties R₁ and R₂, respectively, wherein up to six of said peptide residues may be substituted, R₁ and R₂ do not naturally flank the Belclin 1 residues, and F270 and F274 are optionally substituted and optionally linked. The compounds may be used to induce autophagy.

Description

Autophagy-Inducing Peptide Analogs

Inventors: Beth C. Levine and Sanae Shoji-Kawata, of Dallas, TX

Nick V. Grishin and Lisa N. Kinch, of Dallas, TX

Olivier Lichtarge and Angela D. Wilkins, of Houston, TX

Applicants/: University of Texas, Board of Regents

Assignees Baylor College of Medicine

[001] This invention was made with government support under Grant Number

U54AI057156 awarded by the National Institutes of Health (NIH). The government has certain rights in the invention.

Applicants claim priority to US Ser No. 61/803,095, filed: Mar 18, 2013.

INTRODUCTION

[002] Beclin 1 encodes a 450 amino acid protein with a central coiled coil domain. Within its N' terminus, it contains a BH3-only domain, which mediates binding to anti-apoptotic molecules such as Bcl-2 and Bcl-xL. The most highly conserved region, referred to as the evolutionarily conserved domain (ECD), spans from amino acids 244-337, which is important for its interaction with Vps34. Overexpression of Beclin 1 is sufficient to induce autophagy. Furuya, et al., Autophagy 1, 46-52, 2005; Pattingre, S. et al. Cell 122, 927-39, 2005.

[003] We disclosed an autophagy-inducing beclin 1 peptide in PCT/US 13/22350.

SUMMARY OF THE INVENTION

[004] The invention provides methods and compositions for inducing autophagy.

[005] In one aspect the invention is an autophagy-inducing compound comprising an autophagy-inducing peptide comprising Beclin 1 residues 270-278 (FNATFHIWH; SEQ ID NO:01 , residues 2-10), or the D-retro-inverso sequence thereof, immediately N- and C- terminally flanked by moieties Ri and R₂, respectively, wherein up to six of said residues may be substituted, Ri and R₂ do not naturally flank the Beclin 1 residues, and F270 and F274 are optionally substituted and optionally linked.

[006] In particular embodiments the invention is a subject autophagy-inducing compound, [007] - having a peptide sequence disclosed herein, unsubstituted, or wherein up to six of said residues may be substituted, and the two F residues are Fi and F₂ and are optionally substituted and optionally linked, or the D-retro-inverso sequence of the peptide;

[008] - wherein Ri is a heterologous moiety that promotes therapeutic stability or delivery of the compound;

[009] - wherein Rl comprises a transduction domain, a homing peptide, or a serum stabilizing agent;

[010] - wherein Rl is a tat protein transduction domain linked to the peptide through a diglycine linker, particularly a diglycine-T-N linker;

[011] - wherein R₂ is carboxyl;

[012] - wherein R₂ comprises an affinity tag or detectable label, particularly a fluorescent label;

[013] - wherein F270 and F274 are substituted and linked;

[014] - wherein F270 and F274 are substituted with crosslinkable moieties and/or linked, and each optionally comprises an additional a-carbon substitution selected from substituted, optionally hetero- lower alkyl, particularly optionally substituted, optionally hetero- methyl, ethyl, propyl and butyl;

[015] - wherein F270 and F274 are substituted with homocysteines connected through a disulfide bridge to generate a ring and tail cyclic peptide;

[016] - wherein the side chains of F270 and F274 are replaced by a linker:

[017] -(CH_2)nONHCOX(CH_2)m-, wherein X is CH₂, NH or O, and m and n are integers 1 -4, forming a lactam peptide;

[018] -CH₂OCH₂CHCHCH₂OCH₂-, forming an ether peptide; or

[019] -(CH_2)nCHCH(CH_2)m-, forming a stapled peptide;

[020] - wherein 1 to 6 residues are alanine substituted;

[021] - wherein the peptide comprises at least one of substitutions: H275E and S279D.

[022] - wherein the peptide comprises one or more D-amino acids, Ε-β-homo amino acids, Ο-β-homo amino acids, or N-methylated amino acids;

[023] -comprising the D-retro-inverso sequence, particularly, RRQRRKKKRGY-GG- DHWIEFTANFV (SEQ ID NO:08);

[024] - wherein the peptide is acetylated, acylated, formylated, amidated, phosphorylated, sulfated or glycosylated;

[025] -comprising an N-terminal acetyl, formyl, myristoyl, palmitoyl, carboxyl or 2-furosyl group, and/or a C-terminal hydroxyl, amide, ester or thioester group; and/or [026] - wherein the peptide is cyclized.

[027] The invention also provides pharmaceutical compositions comprising a subject compound in unit dosage, administrable form, and methods of inducing autophagy, comprising administering to a person in need thereof an effective amount of a subject compound or composition.

[028] The invention includes all combinations of the recited particular embodiments as if each combination had been laboriously separately recited.

DESCRIPTION OF PARTICULAR EMBODIMENTS OF THE INVENTION

[029] In one aspect the invention is an autophagy-inducing compound comprising an autophagy-inducing peptide comprising Beclin 1 residues 270-278 (FNATFHrWH; SEQ ID NO:01 , residues 2-10), or the D-retro-inverso sequence thereof, immediately N- and C- terminally flanked by moieties Ri and R₂, respectively, wherein up to six of said residues may be substituted, Ri and R₂ do not naturally flank the Beclin 1 residues, and F270 and F274 are optionally substituted and optionally linked.

[030] In embodiments the peptide comprises an unsubstituted mutant sequence of Table 1 , below, or the D-retro-inverso sequence thereof, immediately N- and C-terminally flanked by moieties Ri and R₂, respectively, wherein Ri and R₂ do not naturally flank the Beclin 1 residues.

[031] Peptide and compound activity are tolerant to a variety of additional moieties, flanking residues, and substitutions within the defined boundaries. Peptide and compound activity are also tolerant to backbone modification and replacement, side-chain modifications, and N- and C-terminal modifications, all conventional in the art of peptide chemistry.

[032] Chemical modifications of the peptide bonds may be used to provide increased metabolic stability against enzyme-mediated hydrolysis; for example, peptide bond replacements (peptide surrogates), such as trifluoroethylamines, can provide metabolically more stable and biologically active peptidomimetics.

[033] Modifications to constrain the peptide backbone include, for example, cyclic peptides/peptidomimetics which can exhibit enhanced metabolic stability against

exopeptidases due to protected C- and N-terminal ends. Suitable techniques for cyclization include Cys-Cys disulfide bridges, peptide macrolactam, peptide thioether, parallel and anti- parallel cyclic dimers, etc. ; see, e.g. PMID 22230563 (stapled peptides), PMID 23064223 (use of click variants for peptide cyclization ), PMID 23133740 (optimizing PK properties of cyclic peptides: effects of side chain substitutions), PMID: 22737969 (identification of key backbone motifs for intestinal permeability, PMID 12646037 (cyclization by coupling 2- amino-d,l-dodecanoic acid (Laa) to the N terminus (LaaMII), and by replacing Asn with this lipoamino acid).

[034] In particular embodiments F270 and F274 are substituted and linked, such as wherein the side chains of F270 and F274 replaced by a linker. For example, these residues may be substituted with homocysteines connected through a disulfide bridge to generate a ring and tail cyclic peptide. In addition, the side chains of these residues can be substituted and cross- linked to form a linker, such as -CH2)_nONHCOX(CH2)_m-, wherein X is C¾, NH or O, and m and n are integers 1-4, forming a lactam peptide; -CH₂OCH₂CHCHCH₂OCH₂-, forming an ether peptide; -(CH₂)_nCHCH(CH₂)_m-, forming a stapled peptide. The linkers may incorporate additional atoms, heteroatoms, or other functionalities, and are typically generated from reactive side chain at F270 and F274. The crosslinkable moieties may include additional a- carbon substititions, such as optionally substituted, optionally hetero- lower alkyl, particularly optionally substituted, optionally hetero- methyl, ethyl, propyl and butyl. Exemplary crosslinkable moieties include:

[035]

formylation, amidation, phosphorylation (on Ser, Thr and/or Tyr), etc. which can be used to improve peptide bioavailability and/or activity, glycosylation, sulfonation, incorporation of chelators (e.g. DOTA, DPT A), etc. PEGylation can be used to increase peptide solubility, bioavailability, in vivo stability and/or decrease immunogenicity, and includes a variety of different PEGs: HiPEG, branched and forked PEGs, releasable PEGs; heterobifunctional PEG (with endgroup N-Hydroxysuccinimide (NHS) esters, maleimide, vinyl sulfone, pyridyl disulfide, amines, and carboxylic acids), etc.

[037] Suitable terminal modifications include N-terminal acetyl, formyl, myristoyl, palmitoyl, carboxyl and 2-furosyl, and C-terminal hydroxyl, amide, ester and thioester groups, which can make the peptide more closely mimic the charge state in the native protein, and/or make it more stable to degradation from exopeptidases. [038] The peptides may also contain atypical or unnatural amino acids, including D-amino acids, L- -homo amino acids, Ο-β-homo amino acids, N-methylated amino acids, etc.

[039] A wide variety of flanking moieties Ri and/or R₂ may be employed, such as affinity tags, transduction domains, homing or targeting moieties, labels, or other functional groups, such as to improve bioavailability and/or activity, and/or provide additional properties.

[040] One useful class of such moieties include transduction domains which facilitate cellular penetrance or uptake, such as protein-derived ( e.g. tat, smac, pen, pVEC, bPrPp, PIsl , VP22, M918, pep-3); chimeric (e.g. TP, TP10, ΜΡΟΔ) or synthetic (e.g. MAP, Pep-1, Oligo Arg) cell-penetrating peptides; see, e.g. "Peptides as Drugs: Discovery and

Development", Ed. Bernd Groner, 2009 WILEY- VCH Verlag GmbH & Co, KGaA,

Weinheim, esp. Chap 7: "The Internalization Mechanisms and Bioactivity of the Cell- Penetrating Peptides", Mats Hansen, Elo Eriste, and Ulo Langel, pp. 125-144.

[041] Another class are homing biomolecules, such as RGD-4C (CCDCRGDCFC; SEQ ID NO:04), NGR (CCNGRC; SEQ ID NO:05) , CREKA, LyP-1 (CGNKRTRGC; SEQ ID NO:06), F3, SMS (SMSIARL; SEQ ID NO:07), IF7, and H2009.1 (Li et al. Bioorg Med Chem. 2011 Sep 15;19(18):5480-9), particularly cancer cell homing or targeting

biomolecules, wherein suitable examples are known in the art, e.g. e.g. Homing peptides as targeted delivery vehicles, Pirjo Laakkonen and Kirsi Vuorinen, Integr. Biol., 2010, 2, 326- 337; Mapping of Vascular ZIP Codes by Phage Display, Teesalu T, Sugahara KN, Ruoslahti E., Methods Enzymol. 2012;503:35-56.

[042] Other useful classes of such moieties include stabilizing agents, such as PEG, oligo- N-methoxyethylglycine (NMEG), albumin, an albumin-binding protein, or an

immunoglobulin Fc domain; affinity tags, such as immuno-tags, biotin, lectins, chelators, etc.; labels, such as optical tags (e.g. Au particles, nanodots), chelated lanthanides, fluorescent dyes (e.g. FITC, FAM, rhodamines), FRET acceptor/donors, etc.

[043] The moieties, tags and functional groups may be coupled to the peptide through linkers or spacers known in the art, such as polyglycine, ε-aminocaproic, etc.

[044] The compound and/or peptide can also be presented as latent or activatable forms, such as a prodrug, wherein the active peptide is metabolically liberated; for example, release of the linear peptide from cyclic prodrugs prepared with an acyloxyalkoxy promoiety

(prodrug 1) or a 3-(2'-hydroxy-4',6'-dimethylphenyl)-3,3-dimethyl propionic acid promoiety (prodrug 2). of the compound).

[045] Particular embodiments include all combinations of particular embodiments, as though each had be separately set forth; for example, wherein the peptide is N-terminally flanked with T-N and C-terminally flanked by T, the first moiety is a tat protein transduction domain linked to the peptide through a diglycine linker; and wherein the peptide is N- terminally flanked with T-N and C-terminally flanked by T, the first moiety is a tetrameric integrin a(v) (6)-binding peptide known as H2009.1 , linked to the peptide through a maleimide - PEG(3) linker.

[046] In another aspect the invention provides a method of inducing autophagy, comprising administering to a person in need thereof an effective amount of a subject compound or composition. Applications broadly encompass persons in need of enhanced autophagy, and include diseases and pathologies where the upregulation of autophagy is therapeutically beneficial, including infection with intracellular pathogens, neurodegenerative diseases, cancers, cardiomyopathy, and aging.

[047] Autophagy can be detected directly, indirectly or inferentially by conventional assays, such as disclosed and/or exemplified herein, including biochemically (by assessing the generation of Atg8-PE or LC3-II or the degradation of p62) or microscopically (e.g. by observing the localization pattern of fluorescently tagged Atg8 or LC3).

[048] It is understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof will be suggested to persons skilled in the art and are to be included within the spirit and purview of this application and scope of the appended claims. All publications, patents, and patent applications cited herein, including citations therein, are hereby incorporated by reference in their entirety for all purposes.

EXAMPLES

[049] We used the beclinl ECD structure as a basis for rational design and optimizing the beclin 1 peptide as a therapeutic.

[050] A sequence alignment of Atg6/beclin 1 orthologues revealed two conserved aromatic residues (F270 and F274) within the Tat-beclinl peptide that are required for its autophagy- inducing activity [1]. When mapped to the beclin 1 ECD structure, these aromatic side chains interact in a T-shaped geometry and help to position the intervening residues in a surface- exposed loop. We inferred that the aromatic residues play an analogous role in the Tat-beclinl peptide, thus contributing to its structural stability, and that replacing this aromatic interaction with a more permanent chemical linkage should further improve conformational stability [2]. For example, the two aromatics could be replaced with homocysteines that exhibit a similar length as the native sidechains and can be cyclized to form a stabilized ring-and-tail peptide structure. Similar biologically active cyclic peptide structures exist in nature, such as the 6- amino acid ring and 3 amino acid tail structure of the peptide hormone vasopressin [3] or the 8-amino acid ring and 13 amino acid tail structure of the RNA polymerase inhibiting peptide microcin [4].

[051] To further assess the components of the beclinl peptide required for autophagy- inducing activity, a series of shorter peptides were developed. Activity was measured by biochemical autophagy assays including western blot detection of p62 (a protein degraded by autophagy) and western blot detection of the conversion of LC3-I into LC3-II (the lipidated autophagosome-associated form of LC3). 10 micomolar concentration of Tat-GG- VFNATFEIWHD (SEQ ID NO:09) results in increased p62 degradation and increased LC3-II conversion as compared to 10 micomolar concentrations of the Tat-Beclin 1 compound; see, Shoji-Kawata et al. (2013), below.

[052] The Tat-beclinl peptides could tolerate losing the two N-terminal beclinl amino acids, but lost potency with the removal of the third N-terminal residue. The two required structural residues (F270 and F274) make up the N-terminal half of the beclinl peptide, forming a surface loop that may dictate activity. TAT-fusions corresponding to this N- terminal half were tested alone and with the C-terminal half added in trans. Neither of these peptides, nor their combination, was able to induce autophagy. This lack of activity suggests a requirement for some component of the C-terminus in addition to the N-terminal aromatic stabilized loop. In the ECD structure, the C-terminal residues following the D (corresponding to S279 mutated for solubility) are largely buried and unlikely to contribute directly to an interaction surface. Accordingly, omission of these five C-terminal residues from the beclinl peptide yielded the most potent inducer of autophagy thus tested. The surface components of the minimal peptide include 1) the N-terminal V 2) the aromatic intervening NAT and 2) the C-terminal WH. These residues are inferred to contribute to a binding surface in the peptide that mimics the ECD and is required for induction of autophagy by the peptide.

[053] We used targeted mutagenesis of amino acids that are predicted to be important based on their substitution patterns during evolution [PMID 8609628; 15037084]. This approach has been successfully applied to design point mutants that modify the activity of functional sites [PMID 20385837] and that selectively block function [PMID 16280323]. We similarly use a series of three or four different mutations at each position to probe both conservative to non- conservative substitutions based on standard transition matrices [PMID 1438297], using mutations to residues predicted to be less important as negative controls to map out efficiently the key binding residues in the minimal peptide. To enhance potency and identify activity determinants we can increasingly perturb residues through increasingly less conservative substitutions, for example, V269 to I, L, M then A, or 1276 to V, L, F, etc. We also use scanning mutagenesis to systematically change each residue one at a time.

[054] F270 and F274 function to restrict the conformational flexibility of the intervening residues and reduce the entropy cost of binding. We prepared a number of more chemically- stable linkers via standard solid-phase peptide synthesis. For instance, replacement of F270 with a lysine and F274 with an aspartic or glutamic acid residue allows the synthesis of macrocyclic lactam peptides of incrementally-larger ring size. Even more chemically-stable linkages can be achieved through ring-closing metathesis strategies, where terminal alkene handles derived from commercial non-natural amino acids offer a chemical handle for cyclization. This latter carbon-carbon based cyclization modality is often called peptide "stapling", and has been reviewed extensively (PMID 22230563). An advantage of backbone "staples" is the reduced susceptibility of carbon-carbon linkages to proteolytic cleavage. "Click" variants derived from [3+2] cycloaddition reactions may also be used for peptide cyclization (PMID 23064223).

[055] These various cyclization strategies enforce the bioactive conformation of the beclin peptide, and can be used to increase the cellular permeability relative to the uncyclized peptide. Peptide conformation can significantly influence the ability of macromolecules to penetrate into cells (PMID: 23133740). In particular, backbone modifications can have dramatic impact on cellular permeability, potentially creating opportunities for oral administration of cyclic peptides (PMID 22737969 and PMID 12646037). Optionally combined with retro-inverso peptide strategies to further limit amide hydrolysis (PMID 23382963), these compositions enable use the disclosed compounds as peptide therapeutics.

[056] Table 1. Targeted mutagensis

Mutant 7 VFNATFDIWHD SEQ ID NO: 16 ~wt

Mutant 8 VFNATFELWHD SEQ ID NO: 17 >wt

Mutant 9 VFNATFE IFHD SEQ ID NO: 18 ~wt

Mutant 10 VFNATFE IWYD SEQ ID NO: 19 ~wt

Mutant 11 VFNATFE IWHE SEQ ID NO:20 ~wt

Mutant 12 VWNATFELWHD SEQ ID NO:21 ~wt

Mutant 13 VYNATFE IWHD SEQ ID NO:22 <wt

Mutant 14 VFNATFE VWHD SEQ ID NO:23 ~wt

Mutant 15 VLNATFE IWHD SEQ ID NO:24 ~wt

Mutant 16 VFNATFEMWHD SEQ ID NO:25 ~wt

Mutant 17 VWNATFHIWHD SEQ ID NO:26 »wt

Mutant 18 VFNATFEFWHD SEQ ID NO:27 ~wt

Mutant 19 VFNATFEYWHD SEQ ID NO:28 ~wt

Mutant 20 VFNATFERWHD SEQ ID NO:29 ~wt

Mutant 21 FNATFEIWHD SEQ ID NO: 30 >wt

Mutant 25 VFNATFE IWH SEQ ID NO:31 >wt

Mutant 26 FNATFEIWH SEQ ID NO: 32 »wt

Mutant 27 WNATFHIWH SEQ ID NO:33 »wt

Mutant 28 VWNATFHIWH SEQ ID NO: 34 »wt

Mutant 29 WNATFHIWHD SEQ ID NO:35 »wt

Mutant 30 DHWIHFTANWV SEQ ID NO:36 »wt

[057] * The 11 amino acids PTD from HIV TAT and the diglycine linker are not listed; this is the same for all peptides. Mutant 17 and its derivatives are the most active, and activity is retained with the corresponding D-retro-inverso sequences (e.g. Mutant 30). Activity was measured as autophagy inducing activity, as described supra.

[058] References

[1] S. Shoji-Kawata, R. Sumpter, M. Leveno, G.R. Campbell, Z. Zou, L. Kinch, A.D. Wilkins, Q. Sun, K. Pallauf, D. MacDuff, C. Huerta, H.W. Virgin, J.B. Helms, R. Eerland, S.A. Tooze, R. Xavier, D.J. Lenschow, A. Yamamoto, D. King, O. Lichtarge, N.V. Grishin, S.A. Spector, D.V. Kaloyanova, B. Levine, Identification of a candidate therapeutic autophagy-inducing peptide, Nature, 494 201-206, 2013.

[2] A. Bryan, L. Joseph, J.A. Bennett, H.I. Jacobson, T.T. Andersen, Design and synthesis of biologically active peptides: a 'tail' of amino acids can modulate activity of synthetic cyclic peptides, Peptides, 32 2504-2510.

[3] J.P. Rose, C.K. Wu, CD. Hsiao, E. Breslow, B.C. Wang, Crystal structure of the neurophysin-oxytocin complex, Nat Struct Biol, 3 (1996) 163-169.

[4] K.J. Rosengren, R.J. Clark, N.L. Daly, U. Goransson, A. Jones, D.J. Craik, Microcin J25 has a threaded sidechain-to-backbone ring structure and not a head-to-tail cyclized backbone, J Am Chem Soc, 125 (2003) 12464-12474.

Claims

WHAT IS CLAIMED IS:

1. An autophagy-inducing compound comprising an autophagy-inducing peptide comprising Beclin 1 residues 270-278 (VFNATFHIWHS; SEQ ID NO:01, residues 2-10), or the D-retro- inverso sequence thereof, immediately N- and C-terminally flanked by moieties Ri and R₂, respectively, wherein up to six of said residues may be substituted, Ri and R₂ do not naturally flank the Beclin 1 residues, and F270 and F274 are optionally substituted and optionally linked.

2. The compound of claim 1 wherein the peptide comprises a sequence, unsubstituted, selected from:

Mutant 18 VFNATFEFWHD SEQ ID NO:27;

Mutant 19 VFNATFEYWHD SEQ ID NO:28;

Mutant 20 VFNATFERWHD SEQ ID NO:29;

Mutant 21 FNATFEIWHD SEQ ID NO: 30;

Mutant 25 VFNATFEIWH SEQ ID NO:31 ;

Mutant 26 FNATFEIWH SEQ ID NO: 32;

Mutant 27 WNATFHIWH SEQ ID NO:33;

Mutant 28 VWNATFHIWH SEQ ID NO: 34; and

Mutant 29 WNATFHIWHD SEQ ID NO:35,

or the D-retro-inverso sequence of the peptide.

3. The compound of claim 1, wherein Ri comprises a transduction domain, a homing peptide, or a serum stabilizing agent.

4. The compound of claim 1 wherein Ri is a tat protein transduction domain linked to the peptide through a diglycine linker, particularly a diglycine-T-N linker.

5. The compound of claim 1 wherein R₂ is carboxyl or R₂ comprises an affinity tag or detectable label, particularly a fluorescent label.

6. The compound of claim 1 wherein:

a) F270 and F274 are substituted with crosslinkable moieties and/or linked, and each optionally comprises an additional a-carbon substitution selected from substituted, optionally hetero- lower alkyl, particularly optionally substituted, optionally hetero- methyl, ethyl, propyl and butyl; or

b) F270 and F274 are substituted with homocysteines connected through a disulfide bridge to generate a ring and tail cyclic peptide.

7. The compound of claim 1 wherein the side chains of F270 and F274 are replaced by a linker:

-(CH_2)nONHCOX(CH_2)m-, wherein X is CH₂, NH or O, and m and n are integers 1-4, forming a lactam peptide; -CH₂OCH₂CHCHCH₂OCH₂-, forming an ether peptide; or

-(CH_2)nCHCH(CH_2)m-, forming a stapled peptide.

8. The compound of claim 1 wherein:

1 to 6 residues are alanine substituted; or

the peptide comprises at least one of substitutions: H275E and S279D; or the peptide comprises one or more D-amino acids, Ε-β-homo amino acids, Ο-β-homo amino acids, or N- methylated amino acids; or

the peptide comprises the D-retro-inverso sequence.

9. The compound of claim 1 comprising the D-retro-inverso sequence:

RRQRRKKKRGY-GG-DHWIEFTANFV (SEQ ID NO:08).

10. The compound of claim 1 wherein the peptide is acetylated, acylated, formylated, amidated, phosphorylated, sulfated or glycosylated.

11. The compound of claim 1 comprising an N-terminal acetyl, formyl, myristoyl, palmitoyl, carboxyl or 2-furosyl group, and/or a C-terminal hydroxyl, amide, ester or thioester group.

12. The compound of claim 1 wherein the peptide is cyclized.

13. The compound of claim 1 wherein the peptide comprises the sequence VWNATFHIWHD (SEQ ID NO:26) or the D-retro-inverso- sequence thereof, DHWIHFTANWV (SEQ ID NO:30).

14. A pharmaceutical composition comprising a compound of claim 1 in unit dosage, administrable form.

15. A method of inducing autophagy, comprising administering to a person in need thereof an effective amount of a compound of claim 1.