WO1996028468A2

WO1996028468A2 - Amphiphilic peptide and analogs thereof

Info

Publication number: WO1996028468A2
Application number: PCT/EP1996/000844
Authority: WO
Inventors: Manmohan Bhakoo; Shail Patel; Peter Ian Stott
Original assignee: Unilever Plc; Unilever N.V.
Priority date: 1995-03-09
Filing date: 1996-02-27
Publication date: 1996-09-19
Also published as: AU720419B2; KR19980702853A; GB9504761D0; WO1996028468A3; EP0815128A2; BR9607242A; AU4831696A

Abstract

The invention relates to amphiphilic peptides and analogs thereof. The specification defines the molecules in three ways and discloses that antimicrobial peptides which comprise a sequence selected from the group comprising: +-++--++--+--++, +-++--++---+-+++, ++-+++-++-+++--, ++-+++-++-++++-++-, ++-+++-++--++--++-, --++--++-++--++-++, ++++-+++--+--++, ++-+++++++-++-++--, wherein + indicates a hydrophobic amino acid residue and - indicates a hydrophilic amino acid residue have improved antimicrobial properties. Specific sequences of active peptides are given. The specification also discloses two general equations which define the rule by whicc peptides falling within the scope of the invention are identified.

Description

AMPHIPHILIC PEPTIDE AND ANALOGS THEREOF

TECHNICAL FIELD

The present invention relates to amphiphilic peptides and to analogs thereof.

BACKGROUND TO THE INVENTION

Peptides are polymers of amino acids. There are around twenty or so naturally occurring amino acids and these vary somewhat in the nature of their chemistry. By arranging specific sequences of amino acids in peptide molecules it is possible to obtain peptides which express a very broad range of chemical functionality, ranging from activity as catalysts (e.g. enzymes) to chemical messengers (e.g.

peptide hormones such as insulin). It is known that small changes in the sequences of peptides can have significant consequences as regards the functionality of the molecules. It is known that the synthesis of particular peptides can be directed by nucleic acid molecules of related sequences.

Amphiphillic peptides are characterised by an amino acid structure which provides some regions of the peptide molecule with a hydrophobic character and other regions with a hydrophillic character. It is believed that these peptides are capable of interacting with the membranes of living cells to modify the life-process of the cells. A number of amphiphillic peptides occur in nature, others have been produced by synthetic methods. WO 92/01462 discloses amphiphilic peptides for inhibiting growth of a target cell. These peptides comprise both hydrophobic and hydrophilic amino acid residues. The hydrophobic and hydrophilic balance of the amino acids can be described by reference to the hydrophobicity values quoted by Eisenberg et al. [Nature 1982 299 p371].

Hydrophobic residues have a value greater than zero whereas hydrophillic residues have a value of less than zero.

For convenience some of these values are given in table 1 below, together with the codes used to represent the associated amino acid residues.

Lee at al. in Biochimica et Biophysica Acta 862 (1986) 211- 219 disclose amphiphilic antimicrobial peptides having the Structures: Ac-(LARL)_1-4-NHCH₃, Ac-(LLARL)_2-3-NHCH₃, Ac-(ARL)_3- ₄-NHCH₃ and Ac-(LAKL)_2-3-NHCH₃. A general review of the field is found in Biochimica et Biophysica Acta 1197 (1994) 109-131.

Despite the sequences of several amphiphilic peptides being known it has proved difficult to identify those peptides having particularly high antimicrobial activity. While each possible peptide could, in theory, be laboriously synthesised and tested against bacteria, fungi and other microbes and it must be appreciated that given the vast number of possible amino acid sequences of an appropriate length this would take an unacceptably long time.

Accordingly, there is a need to identify and define those peptides which might be expected to be effective by a definition other than one based only on an in-vitro

functional test of a molecule which is expensive to

synthesise.

BRIEF DESCRIPTION OF THE INVENTION We have determined that improved amphiphilic antimicrobial peptides can be prepared which comprise sequences selected from the group comprising:

wherein + indicates a hydrophobic amino acid residue and - indicates a hydrophillic amino acid residue.

In particular, we have found utility in peptides having the sequences

serine, E = glutamate, Q = glutamine, T= threonine and R = arginine.

We have also determined that effective antimicrobial peptides are discriminated from ineffective peptides by a vectorial analysis on dimensions corresponding to charge, hydrophile/lipophile balance, hydrophobic moment, and amphiphilicity of the peptides whereby effective peptides fall into the region which is anti-clockwise of the amphiphilicity dimension and clockwise of the charge dimension.

We have further determined that effective peptides are discriminated from ineffective peptides by means of an equation relating certain physical properties of the peptides to their biological activity against specific microorganisms.

As discussed above, it will be appreciated that a very large number of possible peptides can be envisaged with a chain length of 10-30 amino acids. We have determined what is believed to be a general rule as regards the peptides which have effective antimicrobial properties and, as will be discussed in further detail below, have demonstrated that a strong correlation is shown between those peptides which satisfy this rule and those which have effective antimicrobial properties. Conversely, it appears that the majority of the peptides which exhibit properties outside of the scope of the rule do not show effective

antimicrobial properties.

Accordingly, the present invention extends to antimicrobial peptides having a length of 10-30 amino acid residues wherein the predicted log kill to S. aureus ATCC 6538, (L_s _aureus) is greater than 5, L_{s aureus} being given by the

wherein : z₁= sum of the hydrophobic and hydrophilic values z₂= Y component of charge dipole

z₃= sum of negative charge

z₄= closeness of fit to sine wave

z₅= X component of charge dipole

z_€= closeness of fit to square wave

The present invention also extends to antimicrobial peptides having a length of 10-30 amino acid residues wherein the predicted log kill to E. coli ATCC 11229, (L_E _coli) is greater than 4, L_{E coli} being given by the equation L_E _coli = y[27] * (4.41875) + (3.615)

z₀ is the L_{S aureus,}

z₁ is the molecular weight,

z₂ is the modulus of the hydrophobicity dipole moment in

Z-plane,

z₃ is the hydrophobic solvent accessible surface area, z₄ is the number of atoms which are greater than 8 Angstrom Units from the axis of the alpha helix,

z₅ is the 2D radius of gyration (R_xy)

While the above-mentioned equations seem complex it is a simple computational task to determine whether any

particular peptide falls within the scope of the rule or not. The equation is that of a so-called neural net and the general type is familiar to those engaged in this art.

As will be discussed in further detail and demonstrated by way of example below, it is possible, given the rule, to search for sequences which obey this rule and consequently would be expected to exhibit significant antimicrobial properties. The synthesis of the peptides which have been thus identified as potentially antimicrobial is therefore a significantly simpler task than the synthesis of peptides and the selection of desirable molecules on the basis of their measured antimicrobial properties.

DETAILED DESCRIPTION OF THE INVENTION: In the current specification, the following terms are used:

The molecular weight, is the molecular weight of the peptide in Daltons (grams per mole). The hydrophobic solvent accessible surface area (in square Angstroms) is determined using the SYBYL [TM] molecular modelling package (ex. TRIPOS [TM]). The hydrophilic solvent accessible surface area (in square Angstroms) is determined using the SYBYL [TM] molecular modelling package (ex. TRIPOS [TM]).

The X- and Y- components of the charge dipole are obtained from SYBYL after aligning the molecule such that the z-axis is along the long axis of the molecule, i.e. along the axis of the alpha-helix and the y axis is aligned with the projection into a plane orthogonal with the z-axis of the vector P wherein:

P = sum over hydrophobic residues of H_iC_i, wherein, C_i= centre of mass of the alpha-carbon of each

residue, and,

H_i= the hydrophobicity of the residue according to Eisenberg The sum of negative charge is the sum of the negative charges on the residues at pH 7.0

The hydrophobicity dipole moment in XY-plane is the sum of vectors H⁺-H wherein:

H+ = sum over hydrophobic residues of H_iC_i ^xy, and, H- = sum over hydrophilic residues of H_iC_i ^xy

The sum of the hydrophobic and hydrophilic values is the sum of the hydrophobicity and hydrophilicity values according to Eisenberg (cit. ultra).

The modulus of the hydrophobicity dipole moment in the Z-plane. This hydrophobicity dipole moment H_z is given by:

H_z = H_z- - H_z- where H_z ⁺ is the sum of (H_i * z_i) over all hydrophobic amino acids, and, H_z- is the sum of (-H_i * z_i) over all

hydrophilic amino acids, H_i is the Eisenberg (cit ultra) hydrophobicity of amino acid i, and z. is the z coordinate of the C-alpha carbon atom of amino acid i, in units of Angstroms, using the same co-ordinate system as described above, i.e. aligning the molecule such that the Z-axis is along the long axis of the molecule The number of atoms which are greater than 8 Angstrom Units from the axis of the alpha helix is obtained from SYBIL and is a simple number. The closeness of fit to a sine wave, s_sin, is given by the 'least squares' equation:

where N is the number of peptides, s_sin is the minimum value which can be obtained by varying the offset θ, and where H_i is in this instance normalised such that the largest positive and largest negative value of hydrophobicity for the residues are set at 1 and -1 respectively.

The closeness of fit to a square wave is s_sqr, is given by the 'least squares' equation:

where N is the number of amino acid residues, and H_i is the hydrophobicity, i.e. s_sqr is the minimum value which can be obtained by varying the offset θ. The radius of gyration (R_zyz) is a known measure of the sphericalness of a molecule. The 2D radius of gyration (R_xy) is related to how circular the peptide appears, when looking down the alpha helix axis.

Where x_i and y_i are the coordinates of atom i. N is the number of atoms in the peptide. It is believed that this value is related to the ease of diffusion of the molecule through the outer membrane of the E. coli bacteria. The software used was SYBYL version 6.2 (R4000 version running on a Silicon Graphics [TM] Power Challenge [TM]). The peptides were built using the biopolymer module, building an individual peptide with an alpha helix

conformation using the command BIOPOLYMER BUILD M1

<SEQUENCE> ALPHA_HELIX. Hydrogen atoms were then added using the command BIOPOLYMER ADDH M1 ( * ) ESSENTIAL_ONLY. The peptide was then orientated as specified earlier, then the protein structure was refined using the following commands:

BIOPOLYMER ADD_SIDECHAINS Ml ( * )

BIOPOLYMER FIX_END_GROUPS Ml NEUTRAL

BIOPOLYMER ADDH Ml ( * ) ALL Charges were calculated using the Gasteiger-Huckel method. All the calculations mentioned above can be automated using computational techniques.

The forcefield used was the "Tripos 5.2 force field, as reported by M. Clark, R.D. Cramer III & N. van Opdenbosch; J. Comp. Chem, vol 10, p 982-1012 (1989), and amended for the SYBYL 6.0 release, as described in Appendix 2 of the "Sybyl Theory Manual." From the above it can be seen that the parameters used in defining the peptides according to the present invention can be simply determined given the structure of the peptide and no measurement of the actual physical parameters of the peptides need be made. This process can be performed by a computer. Given that there are a finite number of peptide molecules having 10-30 amino acids it is a computationally simple task to determine the sequences and parameters of each of these peptides and then determine whether the peptide has a predicted log kill against E . coli or S .

aureus above the required value. The peptides mentioned above have several additional features which make them particularly suitable as

amphiphilic antimicrobial peptides. The peptides are generally in the L-configuration, although it is envisaged that helices of D-amino acids would be effective.

The peptides according to the invention are predominantly alpha-helical and laterally amphipathic in their secondary-structure, in that they have two distinguishable faces one being predominantly hydrophilic the other being

predominantly hydrophobic. The peptides are relatively short having 12-26, preferably 13-20, most preferably 15-18 amino acid residues. It is believed that these relatively short chain lengths reduce the risk of haematotoxicity. Preferably, the peptides are rich in basic amino acids, particularly lysine and arginine. In preferred embodiments 15-50% by number of the amino acid residues present are lysine or arginine. Preferably, the peptides are poor in the so-called 'beta-turn' formers, particularly lso-leucine, tyrosine and valine. In preferred embodiments of the invention less than 20% by number of the amino acid residues present are iso-leucine, tyrosine or valine.

It is preferable that the numeric ratio of the numbers of lysine or arginine residues to the numbers of iso-leucine, tyrosine or valine residues is one or more. Preferably, the peptides are rich in alpha-helix formers, particularly lysine, alanine and glutamine. It is preferable that at least 50% by number of the amino acid residues are lysine, alanine or glutamine.

Preferably, the overall charge is net positive, such that the molecules are cationic at pH 7.00.

Preferably, specific amino acids, such as cystine, proline and histidine are not prevalent. In embodiments of the invention these amino acid residues are typically absent.

Preferably, the peptides have an amphiphilicity between 0.15 and -0.34. The amphiphilicity for a peptide with N residues is given as the sum from 1-N of the H_n (hydrophobic face) minus the sum from 1-N of H_n(hydrophilic face): where H_n is the numerical hydrophobicity of the nth residue, as given by Eisenberg.

Preferably, the peptides have a hydrophobic moment between 0.25 and 0.5. The hydrophobic moment is a scalar quantity derived from the vector sum of the hydrophobic moments of the individual residues, divided by the total number of residues. The hydrophobic moment for a peptide with N residues is given by:

wherein N is the number of residues, H_i is the

hydrophobicity of the ith residue and δ_i is the angle between the unit vectors of successive residues, i.e. 100 degrees of arc.

Preferably, the peptides are have both L_s>5 and L_e>4, a length of 13-20 amino acids, and consist of the residues of at least three of leucine, alanine, phenylalanine, valine, isoleucine, glycine, lysine, serine, glutamine, glutamate and arginine. More preferably other amino acids than those listed above are absent. The most preferable amino acids are those which show high activity against both S. aureus ( L_s greater than about 5) and E coli (L_e greater than about 4) these have sequences which comprise one of:

Preferably the peptides are not homopolymers of oligomers consisting of 3-5 amino acid residues.

The present invention also relates to compositions of matter comprising one or more of the peptides of the invention and the use thereof as antimicrobial (preferably antibacterial or antifungal) compositions either alone or in combination with other antimicrobial materials and treatments (including the application of heat and or pressure). The afore-mentioned compositions find utility in disinfection, spoilage prevention, preservation and other hygiene processes. It will be apparent to the skilled worker that the peptides will find applications in many situations in which bacterial or fungal kill is required. According a further aspect of the present invention

subsists in a method for the disinfection of surfaces which comprises the step of treating a surface with a composition comprising a peptide according to any one of claims 1-4 as appended hereto. While the peptides disclosed herein have been synthesised in vitro by the stepwise method of Merifield, it is

envisaged that the peptides can be produced by synthesis, in vivo using a nucleic acid molecule capable of directing the synthesis of the peptides. It is within the scope of the art to produce a DNA molecule which codes for the peptides disclosed herein and introduce the said DNA molecule into a cell such that the cell produces the peptides or a precursor thereof as a metabolic product. In order that the present invention may be further

understood it will be described hereinafter with reference to the accompanying figure wherein:

Figure 1: Is a SAS/RGSYS (TM) vectorial analysis method (internal unfolding) on dimensions corresponding to Total charge, Hydrophile/Lipophile balance (HLB) , hydrophobic moment (as described above), and amphiphilicity (as

described above) of the peptides.

EXAMPLES

The peptides listed in Table 2 below were prepared by

Merifield Solid Phase Synthesis (R. B. Merrifield J.Am.

Chem. Soc. [1963] volume 85 page 2149) .

The actual bactericidal activity of these peptides was determined by a total viable counts technique using both a Gram negative bacteria (Escherichia coli ATCC 11229) and a Gram positive bacteria (Staphylococcus aureus ATCC 6538) which were cultured overnight, in nutrient broth (Oxoid (TM) Nr. 2). Cells were harvested and suspended in Ringers solution to obtain initial cell numbers of 10⁶-10⁸. Cells were incubated with the peptides aseptically in 96-well microtitre plates and relevant controls undertaken. The concentration of the peptides was 0.1-1 mg/ml. In both cases (E coli and S aureus) a log kill was determined. The measured log kill values are given in Table 2 below:

The results were analysed by a SAS/RGSYS (TM) vectorial analysis method (internal unfolding) on dimensions

corresponding to Total charge, Hydrophile/Lipophile balance (HLB), hydrophobic moment (as described above), and

amphiphilicity (as described above) of the peptides. This analysis showed that these parameters could be used to predict which of a large number of peptides would be effective antibacterials. Results are presented in figure 1. In figure 1, AMP is amphiphilicity, EC is log kill against E. coli, STPH is log kill against S. aureus, HM is hydrophobic moment, CH is total positive charge and HLB is hydrophile/ lipophile balance.

Table 2 below shows correlations between the measured log kill (as 'act') and the predicted log kills (as 'pre') against E. coli and S. aureus for the peptides listed.

Predicted values were determined by means of the equation which is provided by the present invention. It can be seen that there is very good agreement between the predicted values and the observed values. The peptides which show a predicted log kill of 5 or more against S . aureus or a predicted log kill 4 or more against E. coli are examples of the present invention, the other examples are comparative and show that relatively minor changes in structure can have significant consequences as regards the antimicrobial activity of the molecules. The samples identified by the codes MB_nn (SEQ. ID. ) were of peptides which were identified as possibly antimicrobial without reference to the predictive equation provided by the present invention. The samples coded MC_nn (SEQ. ID. ) were peptides whose sequence showed high predicted activity. These molecules were synthesised using the

Merifield Solid Phase Synthesis method and the actual activity of the molecules against S . aureus was determined. It can be seen that there is a good correlation between the predicted activity and the observed activity of these molecules. A computer-assisted method was used to arrive at the sequences of the CM-n (SEQ. ID. ) series of

examples. Of these it can be seen that there is again a high correlation between the predicted and expected

antimicrobial activity.

Claims

1. Amphiphilic antimicrobial peptides which comprise a sequence selected from the group comprising:

2. Peptides according to claim 1 having a sequence

selected from the group comprising

3. Antimicrobial peptides having a length of 10-30 amino acid residues wherein the predicted log kill to S. aureus ATCC 6538, (L_{S aureus}) is greater than 5, L_{S aureus} being given by the equation L_{S aureus} = y[25] *

(4.8362068) + (3.5316215) wherein:

4. Antimicrobial peptides having a length of 10-30 amino acid residues wherein the predicted log kill to E. coli ATCC 11229, (L_{E coli}) is greater than 4, L_{E coli} being given by the equation L_{E coli} = y[27] * (4.41875) + (3.615); wherein,

z₀ is the L_{S aureus,}

z₁ is the molecular weight,

z₂ is the modulus of the hydrophobicity dipole moment in Z-plane,

z₃ is the hydrophobic solvent accessible surface area, z₄ is the number of atoms which are' greater than 8 Angstrom Units from the axis of the alpha helix,

z₅ is the 2D radius of gyration (R_xy)

5. A method for the disinfection of surfaces which comprises the step of treating a surface with a composition comprising a peptide according to any one of claims 1-4.