Energy-Conformational Studies of ,B-Endorphins:
`
`
`
`Identification of Plausible Folded Conformers
`
`GILDA LOEW, JACK COLLINS, PHILIP PAYNE, AMRIT K. JUDD, and
`KEVIN H. WACKNOW
`
`SRI International. 333 Ravenswood Avenue,
`94025, USA
`Menlo Park, California
`
`Abstract
`
`{I-Endorphins are 31 amino acid endogenous opioid peptides with high receptor affinity and antinocicep­
`tive acvitity. Because of their importance as neurohormones and the significant experimental effort that has
`been made to understand their struc1ure activiiy profiles, we have begun to develop procedures that could
`be useful first to identify low-energy conformers of {I-endorphins and ul!imately their bioactive form. In
`the initial studies reported here, we have identified plausible initial structures of the full peptide by calcu­
`lating and comparing the conformational preference of all possible extended tetrapeptide fragments of
`/3-endorphin starting from each of the first 28 residues. Comparisons of fragment energies suggested two
`types of compact folded /3-endorphin conformers were plausible: a helix-rum-helix and an antiparallel
`,8-sheet conformer. These structures, as well as an extended a-helical and /3-slrand conformer, were assem­
`bled and total geometry optimization performed using the empirical-energy-based program AMBER. The
`resuhs yield an a-helical structure as the lowest energy form consistent with recently reported NMR stud­
`ies of ,B-endorphin. The two more compact folded structures obtained, however, are reasonable staning
`conformations for further planned molecular dynamics simulation studies and could yield competing low­
`energy structures as candidates for the bioactive form of these peptides.
`
`Introduction
`
`31 amino acid fragments
`of a larger prohormone, are potent en­
`{3-Endorphins,
`
`[l].
`
`
`dogenous opioid peptides with high receptor affinity and antinociceptive activity
`
`
`That the sequence of {3-endorphins is remarkably conserved across a variety of spe­
`
`
`cies is an indication that more than just the amino tenninal met-enkaphalin-like por­
`
`
`
`tion of the peptides is important for activity. Since its discovery, about one hundred
`
`
`
`different analogs of {3-endorphins have been synthesized in an attempt to detennine
`
`
`
`the importance of individual residues and regions to the affinity and activity of the
`
`
`
`
`
`peptide [1-3]. These extensive structure-activity studies include replacement, omis­
`bridges [l-3). In another
`
`
`
`sion, and addition of residues and incorporation of disulfide
`
`
`approach, variations in residues in the 13-31 regions were made based on the hy­
`
`
`
`is that this re­in {3-endorphins pothesis f 4] that all that is required for opioid activity
`
`
`
`gion form an amphiphilic helical structure. While the resulting analogs have shown
`
`
`binding affinity ranging from less than micromolar to nanomolar, and varying effica­
`
`
`
`cies as analgesics, the fundamental stereoelectronic properties that determine these
`
`variations have not yet been identified.
`
`International Journal of Quantum Chemistry, Quantum Biology Symposium, 15, 055-066 (1988)
`© 1988 by John Wiley & Sons, Inc.
`CCC 0360-8832188/010055-12$04.00
`
`1 of 12
`
`BI Exhibit 1076
`
`

`

`56
`
`LOEW ET AL.
`
`This lack is not surprising since the /3-endorphins belong to a class of intermediate
`size bioactive peptides for which characterization of conformation profiles is most
`difficult. Such peptides pose discouraging difficulties for each of the three new disci­
`plines: x-ray crystal structure determination, NMR studies, and theoretical energy­
`conformational studies, that in principle could be most useful for such studies. They
`are, in general, difficult to crystallize, have conformational flexibilities at room tem­
`perature, and have many possible stable structures, i.e., many local conformational
`minima of varying relative energies.
`Recently, an NMR study of /3-endorphin has been published (5]. While conforma­
`tional flexibility in water did not allow an
`lysis of the spectra, dilution with methanol
`did. The extensive analysis made of the NMR results, in water-methanol solution, in­
`cluding NOE spectra, were consistent with a predominantly a-helical structure. The
`NMR data do not preclude, however, a possible tum near residues 10-15. However,
`these studies are only a first step in addressing the question of the bioactive conform­
`ers of ,8-endorphins, i.e., the form in which they bind to opioid receptors. There are
`undoubtedly a number of lower-energy candidate structures and the environment at
`the opioid receptor binding site is most likely devoid of bulk solvent.
`The two more formidable obstacles to energy conformation studies of peptides of
`the size of j3-endorphin, are the existence of large numbers of stable conformers, i.e.,
`the multiminimum problem, and the practical difficulties of constructing confonners
`which are good initial approximations to these minima. Human /3-endorphin, for ex­
`ample, has 184 nonhydrogen torsion angles which render impractical the search
`strategies routinely used for smaller peptides such as nested rotations and "buildup"
`from low energy conformers of single amino acids. Thus entirely different procedures
`must be developed to search conformational space of these peptides for low-energy
`conformers.
`The most common approach used to rel.ate amino acid sequence to secondary struc­
`'
`large peptides is based on statistical analysis of x-ray structure data of
`ture in
`proteins (6). By contrast, in the work reported here, a novel search strategy, based on
`comparisons of calculated optimized energies of sequential peptide fragments was de­
`veloped to help identify plausible secondary structure regions for j3-endorphin. These
`fragments were then used as guides to fold the peptide into a small number of qualita­
`tively different conformations, i.e., into a set of tertiary structures which were then
`subjected to complete geometry optimizations.
`The results obtained thus far indicate that the search strategy developed could be
`useful to construct initial conformers of bioactive peptides in general and to address
`aspects of the protein folding problem.
`
`.a
`
`Methods and Procedures
`
`As a guide to construction of plausible conformers of 13-endorphin, 28 overlapping
`elongated and derivatized tetrapeptide fragments were constructed of the form:
`CH3CONH-ala-(res;-res1+3)ala-CONHCH3, i == 1-28. Each of the 28 fragments
`were constructed in 6 idealized backbone conformations corresponding to an a-helix,
`a /3-strand, and four /3-turns; I, I', II, and II'. The tetramers were extended to hexam-
`
`2 of 12
`
`BI Exhibit 1076
`
`

`

`/HNDORPHlNS
`
`57
`
`ers by adding an alanine on each side to allow even-handed comparisons of the ener­
`gies of these various secondary structures since a minimum of 6-, 5-, and
`4-contiguous residues are required for favorable H-bonding in a-helical, {3-strand and
`/3-tum fragments respectively. The N-terminal and C-terminal ends were appropri­
`ately derivatized by N-acetyl groups and carboxy N-methyl to more realistically
`mimic the conformational behavior of the segment as part of the larger peptide chain.
`Facile construction of these fragments was possible using the capabilities of an inter­
`active structure generating program called MOLECULE, described elsewhere (7],
`which has a library of single amino acid structures and the ability to automatically
`generate peptides of a chosen backbone conformation with extended side chain tor­
`sion angles. For fragments 10, 11, 12, and 13, which include the proline residue 13,
`two optimized proline ring geometries called PROu and PROd were used. In frag­
`ment 11, in which Prol3 is the second residue in the P-tum, only tum types 1 and ll'
`are possible. For fragment 12, in which Pro 13 is the first residue in the turn, only
`turn types II and II are possible. All of the initial structures generated for the
`28 fragments were optimized in two steps, side chain angles only and then full tor­
`sion angle optimization using a quasi-Newton-Raphson energy minimization pro­
`gram caJled PEP that was developed in our laboratory. The five-term empirical
`energy expression in the program called ECCEP (8) formed the basis of this opti­
`mization. It contains contributions from electrostatic, hydrogen-bonded, dispersion,
`repulsion, and torsion angle potentials, and is described in detail elsewhere [8, 9).
`In a buildup procedure similar to that used by Scheraga and co-workers [10), plau­
`sible folded structures of {3-endorphin were constructed by linking energy optimized
`fragments corresponding to different types of secondary structures. This process was
`not automatic, but involved extensive use of graphics capabilities, and distance opti­
`mization to obtain interfragment side-chain conformers which eliminated major steric
`repulsions.
`A set of 5 initial conformers generated for P-endorphin were energy optimized us­
`ing the empirical energy expression contained in the program AMBER [l l) described
`in detail elsewhere. This program allows total geometry optimization and contains a
`7-term energy expression including bond angle and bond length variations in addition
`to the torsion angle variation and four other types of terms similar to those in the
`ECCEP potential. In these calculations, all atoms were explicitly included; with a
`nonbonding atom distance cutoff of IO A. A distance-dependent dielectric, e = r,
`was used. In this program, the polar side chain residues are assumed to be ionized.
`To achieve charge neutrality, salt bridges were formed between oppositely charged
`nearby amino acid side chains and the remainder of the charged residues (Lys) were
`neutralized by addition of negative counterions with Van Der Waal's radii of 3 A.
`This procedure resulted in salt bridge formation in the antiparallel {3-sheet between
`Lys29-Glu3 l and Lysl9-Glu8 and counterions at Lys9, 24, and 28. In the {3-strand,
`salt bridges were formed between Glu8-Lys9; Lys29-Glu3 l, and counterions were
`placed at Lys19, 24, and 28. In the a-helical structure, salt bridges were formed be­
`tween Glu8-Ly9, Lys24-Glu31, and counterions placed at Lysl9, 28, and 25. Fi­
`nally, in the turn-helix-tum-helix-turn structure in the left-handed helix, salt
`bridges were formed between Gly8-Ly9, Lys28-Glu31, and counterions were placed
`
`3 of 12
`
`BI Exhibit 1076
`
`

`

`58
`
`LOEW ET AL.
`
`at Lysl9, 24, and 29; while in the right-handed helix salt bridges were fanned be­
`tween Glu8-Lys9, Lys29-Glu31, and counterions placed at Lysl9, 24, and 28.
`
`Results and Discussion
`
`The optimized energies of the 28 extended tetrameric fragments of 13-endorphin in
`the different backbone confonnations are summarized in Table I relative to that of the
`alpha helical form. As shown in this table, a /3-tum conformation i� preferred for the
`enkephalin portion of the peptide. These results are consistent with our own [121 and
`other previously reported energy-conformation studies of both tetra and pentapeptides
`and NMR studies of metenkephalin [13, 14] in which evidence for both a gly-gly and
`a gly-phe /3-bends have been reported. For fragments 10, 11, 12, and 13, which in­
`clude the proline 13 residue, one proline ring geometry, Pro-D, definitely favored an
`optimized alpha helical structure with some deviations from ideal backbone angles.
`No turns were possible with this proline ring geometry. For the other proline ring
`geometry however, /31-type turns, again deviating from ideal, were favored for
`fragments 10, 11, and 12, identifying a crucial turn region in the middle of the
`13-endorphin sequence which could allow a highly folded structure. A third turn re­
`gion involving C-terminal residues 28-31 was also suggested by these results, though
`an a-helix is somewhat favored.
`In addition to identification of possible tum regions, the results suggest that the re­
`maining contiguous region of the peptide, i.e., residues 5-10, and residues 15-27 are
`in modified alpha helical rather than /3-strand conformations. Thus the most plausible
`folded conformer of ,8-endorphins, is predicted to be of the helix-tum-helix pattern
`with possible additional turns at both the N-terminal and C-terminal end and an inter­
`nal tum beginning at residue 11 or 12. Two highly folded structures of this type were
`constructed, one with a right-handed and the other a left-handed helix, to explore the
`effect of the sense of the helical portions of the conformation and energy. A totally
`helical structure was also constructed as a possible variation of these compact folded
`structures.
`While /3-strands, with only a few exceptions, were not energetically favorable sec­
`ondary structures for fragments, nevertheless, the possibility existed that a fully ex­
`tended /3-strand could be a low-energy conformer of the full peptide since /j-strands
`are likely to have less hindered interfragment interactions than a-helices. It is also
`possible that a highly folded /3-sheet structure, which can be formed from antiparallel
`/3-strands by interstrand H-bonding could be energetically favorable if the energy
`gained from interstrand interactions outweighs the favorable a-helical versus /3-strand
`contiguous domain energies. These possibilities were explored by including among
`the initial conformers to be optimized an extended ,8-strand and an antiparallel
`/3-sheet conformer with turns at residues l-4, 11-14, and 23-26.
`The results of total geometry optimization of the five types of conformers of
`/3-endorphins using the AMBER program are summarized in Table II. These opti­
`mized conformations are shown in Figure 1; their backbone conformations in Fig­
`ure 2, and the corresponding backbone torsion angles are listed in Tables III-Vl. As
`seen in Tables Ill and lV, fairly regular a-helical [Fig. l(a)), and /3-strand [Fig. l(b)]
`structures are retained after optimization. This is not true for the more folded struc-
`
`4 of 12
`
`BI Exhibit 1076
`
`

`

`,B·ENDORPHINS
`
`59
`
`TABJ..El. Optimized energies' of .Bh·endor·
`phin fragments CH,CONH-ala[Res,-
`Res<1+lJ)-ala-CONHCH3•
`
`Fragment
`
`AE11s
`
`6..Br
`- 4(1I')b
`3
`7
`- l J(I)
`17(11)
`21
`20
`19(1)
`17
`I 1(1)
`14
`8(1')
`16
`5(1)
`- 4(1)
`17
`23
`25(Il)
`-I
`- 9(1)
`HIND1
`2
`-6
`- 2(1)
`I
`HIND
`- 1(1)
`0
`2 H1ND
`7
`4(1)
`7
`HIND
`21
`23(I)b
`l 5(ll)
`14
`9(1)
`12
`13(1)
`15
`14
`9(11)
`16
`9(1)
`15
`17(1)
`16
`14(!)
`16(11)
`16
`14
`12(1)
`14
`11(11)
`14
`9(11)
`15
`7(1)
`15
`9(1)
`3
`2(ll)
`
`II
`
`12
`
`13
`
`Tyr
`I
`2
`Gly
`Gly
`3
`Phe
`4
`5 Met
`6
`Thr
`7
`Ser
`8
`Glu
`Lys
`9
`10
`Seru•
`SerD
`GlnU<·•
`GlnD
`ThrUC,C
`ThrD
`Proue
`ProD
`14
`Leu
`Val
`15
`16
`Thr
`17
`Leu
`18
`Phe
`19
`Lys
`Asn
`20
`2 1 Ala
`Ile
`22
`Ile
`23
`24
`Lys
`Asn
`25
`Ala
`26
`27
`Tyr
`Lys-8
`28
`
`'tJ.E in kcal/mo! relative to energy of a-
`helix.
`b( ) = optimized tum with lowest energy
`of I, I', II, II'.
`•Two proline ring geometries called U
`and D were used in these fragments.
`"only /3-tum types I and II' possible with
`Prol3.
`•only ,8-tum types I and II possible with
`Prol3.
`'Tums sterically hindered.
`'The C-term.inal fragment is Lys-Lys-
`Gly-Gln.
`
`5 of 12
`
`BI Exhibit 1076
`
`

`

`60
`
`LOEW ET AL.
`
`TABLE ll. Optimized energies for four types /34-endorphin structures.
`
`Helix-tum-helix
`
`Energy'
`
`a.t·helix
`
`aR
`
`aL
`
`J3-sheet
`
`J3-strand
`
`A£ total
`bond
`non bond
`1-4 nb
`angle
`eel
`1-4 eel
`dihed
`hbond
`
`0.0
`0.0
`0.0
`0.0
`0.0
`0.0
`0.0
`0.0
`0.0
`
`41.7
`0.4
`-16.8
`1.2
`9.4
`4.9
`24.9
`17.l
`0.4
`
`125.5
`2.0
`- 1.0
`- 5.5
`35.7
`16.9
`27.9
`53.5
`- 3.9
`
`89.3
`0.4
`4.6
`- 2.7
`5.4
`59.8
`0.5
`20.8
`0.4
`
`247.4
`- 1.3
`62.4
`5.l
`4.1
`65.J
`11.9
`- 0.7
`0.7
`
`'All energies relative to a-helix.
`
`(a)
`
`(b)
`
`(d)
`
`(e)
`
`Figure I. Five optimized conformers of /3-endorphin showing both backbone and side
`chains: (a) /3-strand, (b) a-helix, (c) antiparallel /3-shect, (d) /3r-arfJr-arJ3r. (d) f3r-aL -
`f3r-a,_ -/Jr.
`
`tures [Fig. l(c)-(e)] in which extensive interstrand side chain interaction causes large
`deviations from ideal secondary structure. As seen in Table V [Fig. l(c)], in the
`antiparallel ,B-sheet structure residues 1-4 form a distorted ,Bll' tum; 11-14, a dis-
`
`6 of 12
`
`BI Exhibit 1076
`
`

`

`,8-ENDORPHINS
`
`61
`
`(a)
`
`(b)
`
`(c)
`
`(e)
`
`(d) f;
`
`Figure 2. Five optimized backbone conformers of ,6-endorphin: (a) ,6-strand, (b) a-helix,
`(c) antiparallel /3-shect, (d) f3r-ar/3r-ap-/3r, (d) /3r-aL-/3r-0tL -.Br·
`
`torted ,8lII tum, and residues 23-26 a distorted {31 tum. The remainder of the struc­
`ture consists of antiparallel /3-strand distorted from ideal values to minimize steric
`hindrance. Similar distortions from ideal secondary structure can be seen in the other
`highly folded structures.
`As seen in Table V, the left-handed helical structure, /3r-acf3r-aL.-f3r, has a ,8II' -
`type turn in residues 1-4 and 11-14, with the remainder of the structure in distorted
`al helical form. The right-handed helical compact structure <f3r-arf3r-aR-/3r) has a
`/3ll' tum at the N-terminal and C-terminal segments, a very broad tum around Prol3,
`with the remaining segments distorted aR-helices.
`As seen in Table II, the alpha helical structure is the most stable. Of the more
`folded structures, our results thus far favor the f3r-aR-,8r-arf3r structure. This
`structure is significantly stabilized over the left-handed helical folded structure and
`the antiparallel ,B-sheet. While the interstrand interactions in the ,8-sheet has con­
`siderably lower energy relative to the {3-strand, the energy gained by interstrand
`H-bonding was not sufficient to make it the more favored compact structure.
`The finding that the two lowest energy conformers are predominantly a-helical is
`in agreement with the recently published NMR studies of ,8-endorphin which also in­
`fers large helical regions [5]. Other features of the structures which are consistent
`with those observed NMR spectra are a turn at the C-terminal end and a relatively
`close distance of the Phe 18 side chain with the delta methyl group of isoleucine 23
`
`7 of 12
`
`BI Exhibit 1076
`
`

`

`62
`
`LOEW ET AL.
`
`TABLE Ill. Backbone torsion angles for
`a-Helical' 13.·endorphin structure.
`
`<t>.
`
`0
`-49
`-54
`-64
`-60
`-67
`-56
`-55
`-64
`-73
`-50
`-56
`-64
`-54
`-61
`-60
`-59
`-55
`-53
`-59
`-60
`-57
`-54
`-60
`-64
`-62
`-51
`-59
`-60
`-87
`
`tli;
`
`-42
`-43
`-49
`-49
`-47
`-36
`-52
`-52
`-31
`-65
`-56
`-62
`-43
`-48
`-45
`-49
`-49
`-so
`-51
`-44
`-47
`-48
`-51
`-40
`-41
`-54
`-36
`-27
`-29
`-57
`
`W;
`
`176
`175
`179
`-177
`178
`172
`178
`-180
`171
`-171
`-171
`175
`-175
`177
`179
`179
`-177
`176
`178
`-179
`176
`173
`179
`179
`176
`173
`180
`174
`172
`-176
`
`Tryl
`Gly2
`Gly3
`Phe4
`Met5
`Thr6
`Se r7
`Glu8
`Lys9
`Ser IO
`Glall
`Thrl2
`Prol3
`Leul4
`Vall5
`Thrl6
`Levl7
`Phe18
`Lysl9
`A sn20
`Alu21
`lll2 2
`11123
`Lys24
`Asn25
`Ala26
`Try22
`Lys28
`Lys29
`Gly30
`
`'Ideal torsion angle values for a-helix
`are q,, = -72°, t/11 = -54°. Fairly regular
`structure retained after optimization.
`
`(5 A) which is consistent with the upfield shift of this methyl group due to ring cur­
`
`
`
`rent perturbations. These results also support the growing experimental evidence that
`
`small peptides can fold with regular secondary structure patterns most often favoring
`
`
`those with predominantly a-helical regions [15]. While it is premature to select these
`
`
`
`
`conformers as the bioactive form in which �-endorphins bind to opioid receptors, the
`
`
`
`result thus far also supports the hypothesis that structures with alpha helical segments
`16-27 could be involved (4).
`in the regions of residues
`In conclusion, the results obtained thus far indicate that {3-endorphin can fold into
`
`
`
`
`
`conformers with clearly de.fined structural motifs that can be inferred from energy op-
`
`8 of 12
`
`BI Exhibit 1076
`
`

`

`,8-EN DORPHINS
`
`63
`
`TABLE IV. Backbone torsion angles for op-
`timized ,8-strand' confonner /3.-endorphin.
`
`<P1
`
`0
`-173
`-178
`-164
`-157
`-155
`-174
`-157
`-153
`-152
`-169
`-156
`- 77
`- 67
`-131
`-153
`-166
`-152
`-158
`-147
`-149
`-160
`-151
`-142
`-129
`-109
`-169
`-163
`-160
`-170
`
`l/J;
`
`174
`173
`171
`159
`154
`175
`156
`158
`139
`159
`155
`141
`65
`126
`154
`172
`153
`152
`143
`145
`148
`153
`150
`129
`126
`169
`152
`153
`158
`169
`
`W;
`
`177
`180
`178
`175
`168
`177
`172
`-170
`179
`-175
`176
`-176
`173
`172
`172
`-179
`173
`175
`171
`176
`177
`176
`-179
`175
`170
`176
`180
`179
`173
`-179
`
`Try!
`Gly2
`Gly3
`Phe4
`Met.5
`Thr6
`Ser?
`Glu8
`Lys9
`SertO
`Glal l
`Thrl2
`Prol3
`Leul4
`Vall5
`Thrl6
`Levi?
`Phe18
`Lys19
`Asn20
`Alu21
`Ill22
`lll23
`Lys24
`Asn25
`Ala26
`Try22
`Lys28
`Lys29
`Gly30
`
`'Ideal values for ,8-strand: <P1 "" -140°,
`tji1 = 135°. Fairly regular s tructure main-
`tained except for a "kink" near Pro 13.
`
`timized fragments and are not random conformations. While an a-helix appears to be
`
`
`
`
`favored thus far, it is possible that additional side chain variations and some changes
`in backbone structure will lead to lower energy forms of the more compact folded
`
`
`
`structures. We are continuing to explore the conformational behavior of this large
`peptide, by molecular dynamic studies using each of the optimized folded conformers
`
`
`
`obtained here as starting configurations. The results should lead to additional varia­
`
`
`
`tions and refinements of the lowest energy conformers among which the bioactive
`
`form will ultimately be identified.
`
`9 of 12
`
`BI Exhibit 1076
`
`

`

`64
`
`LOEW ET AL.
`
`TABLEV. Backbone torsion angles from op-
`timized anti parallel /3-sheet' ,8. -endorphin
`eon former.
`
`c/>1
`
`0
`70
`-47
`-134
`-145
`-134
`-179
`- 75
`- 71
`- 49
`-165
`- 44
`55
`-171
`54
`-124
`-147
`-131
`- 96
`- 58
`-138
`-126
`- 87
`- 50
`-135
`-162
`-122
`-14·6
`- 57
`- 81
`
`t/11
`
`36
`-73
`-54
`148
`164
`146
`53
`67
`86
`129
`96
`-46
`-39
`107
`135
`157
`114
`75
`67
`145
`JIO
`55
`149
`-28
`40
`107
`109
`126
`98
`45
`
`w,
`
`179
`177
`156
`174
`173
`179
`172
`178
`158
`-165
`-165
`176
`-179
`172
`173
`-161
`166
`179
`-168
`172
`168
`-179
`-175
`173
`168
`-170
`162
`156
`-155
`-177
`
`Tryl
`01)'2
`Oly3
`Phe4
`Met5
`Thr6
`Ser7
`Glu8
`Lys9
`Serio
`Glall
`Thrl2
`Pro13
`Leu14
`Va115
`Thrl6
`Levi?
`Phcl8
`Lysl9
`Asn20
`Alu21
`Ill22
`lll23
`Lys24
`Asn25
`Ala26
`Try22
`Lys28
`Lys29
`Oly30
`
`'Residue 1-4 fonn a distort.ed /3fi' tum;
`11-14 a distorted /3fil tum and 23-26 a dis-
`toned /31 tum. The remainder of the struc-
`tures are antiparallel {3-strands distorted
`from ideal values to minimize steric hin-
`drance of side chains.
`
`10 of 12
`
`BI Exhibit 1076
`
`

`

`P-ENDORPHINS
`
`65
`
`TABLE VI. Backbone torsion angles for optimi1.ed Pr-a-fir-a-Pr conformers of PA·
`endorphin.
`
`aR·helix
`
`A
`
`166.84
`- 91.29
`52.37
`-168.99
`- 48.49
`
`-170.91
`172.59
`177.69
`-168.72
`-177.08
`
`0.00
`80.01
`- 57.32
`-126.26
`66.72
`
`94.56
`- 46.79
`173.00
`- 47.78
`178.21 -135.73
`169.51
`78.87
`- 27.78
`40.97
`- 58.67 -172.73
`151.34
`- 38.86
`177.91
`
`63.74
`-160.56
`- 39.69
`121.01
`-156.72
`- 51.65
`- 66.80 -175.21 - 48.18
`-172.93 -114.39
`125.96
`179.48 -116.06
`- 49.41
`
`Tyrl
`Gly2
`Gly3
`Phe4
`Met5
`Thr6
`Ser7
`Glu8
`Lys9
`SerlO
`G lnl l
`Thrl2
`Prol3
`Leul4
`Va115
`
`0.00
`62.12
`-139.84
`-138.65
`- 39.92
`
`- 67.04
`- 57.56
`- 55.36
`- 80.80
`- 46.79
`- 53.57
`- 51.19
`- 48.67
`-IOl.47
`- 45.49
`
`ai -helix
`
`B
`
`14.55
`- 72.17
`- 46.40
`137.07
`107.31
`
`175.07
`177.48
`170.44
`-175.35
`-159.85
`
`146.36
`- 16.61
`-140.54
`121.72
`-156.89
`138.77
`41.57
`-165.67
`178.05 -170.12
`
`63.04
`173.64
`165.57
`- 53.29
`- 42.25
`166.03
`166.78
`147.14
`130.43 -138.00
`
`147.67
`51.94
`99.69
`84.27
`96.92
`
`-156.63
`-159.97
`-173.12
`-145.84
`-153.17
`
`-147.21
`104.92
`-167.45
`115.34
`87.58 -166.72
`-175.08
`134.17
`60. 47
`-147.68
`92.84 -136.60
`-162.93
`116.96
`I 19.22
`-176.12
`169.79
`- 71.70
`-166.89
`60.75
`
`69.16
`53.11
`156.06
`70.58
`85.46
`
`74.52
`62.85
`64.95
`81.09
`59.48
`
`179.36
`- 60.55
`- 41.66
`- 69.75 - 37.15
`169.09
`- 56.53
`- 53.14
`-176.53
`- 47.29
`177.01
`- 56.55
`- 61.43
`- 52.82 -171.13
`
`- 41.16
`- 37.08
`- 44.58
`- 53.07
`- 44.32
`
`177.68
`174.04
`172.39
`172.41
`176.36
`
`Thrl6
`Leul7
`Phe18
`Lys l 9
`Asn20
`Ala2l
`- 64.62
`- 63.07
`lle23
`- 66.89
`Ile23
`- 51.91
`Lys24
`Asn25 - 49.34
`- 59.75 - 45.72
`119.24
`273.47
`Ala26
`61.47
`- 44.27 -174.93
`Try27
`- 58.20
`74.52
`-170.96
`Lys28
`- 69.53 - 39.51
`67.84
`Lys29
`-137.88 - 21.99
`172.64
`- 43.50
`171.61 - 79.45
`Gly30
`- 63.34
`
`11 of 12
`
`BI Exhibit 1076
`
`

`

`66
`
`LOEW ET AL.
`
`Acknowledgment
`
`Support for this work from NIDA grant DA02622 is gratefully acknowledged. We
`are also grateful the use of the San Diego Supercomputer Center Class VI computers
`and the helpful guidance and support of the SDSC staff.
`
`Bibliography
`
`[I] C.H. Li, in Hormonal Proteins and Peptides. C.H. Li, Ed. (Academic, New York. 1980). p. 10.
`12) P. Nicholas, R. G. Hammonds, Jr., and C.H. Li, Proc. Natl. Acad. Sci. U.S.A. 81, 3074 (1984).
`(3) P. Nicholas, R. G. Hammonds, Jr., and C.H. Li, Proc. Natl. Acad. Sci. U.S.A. 79, 2191 (1982).
`[4] J. W. Taylor and E.T. Kaiser, Phannacol. Rev. 38, 291 (1986).
`(SJ 0. Lichtarge, 0. Jardetsky, and C.H. Li, Biochemistry 26, 5916 (1987).
`[6] P. Y. Chou and G.D. Fasman, J. Mol. Biol. 115, 135 (1977).
`(7] J. T. Egan, J. Hart, S. K. Burt, and R. D. MacElroy, Comput. Graphics 6, 1977 (1982).
`[8) F. A. Momany, R. F. McGuire, A. W. Burgess, and H. A. Scheraga, J. Phys. Chem. 79, 7381
`(1975).
`[9} G. Nemethy, M. S. Pottle, and H. A. Scheraga, J. Phys. Chem. 87, 1883 (1983).
`[!OJ M. Vasquez and H. A. Scheraga, Biopolymers 24, 1437 (1985).
`[ 11] J. S. Weiner, P.A. Kollman, D. A. Case, U. C. Singh, C. Ghio, G. Alagona, S. Profeta, and P. J.
`Weiner, J. Am. Chem. Soc. 106, 765 (1984).
`[12] G. Loew, H. Hashimoto, L. Williamson, S. Burt, and W. Ande.rson, Molec. Phann. 22, 467 (1982).
`[13) G. Gupta, FEBS Letters 198, 245 (1986).
`[ 14] Z. Li and H. A. Scheraga, Proc. Natl. Acad. Sci. U.S.A., 84, 6611 (1987).
`[151 S. Marqusee and R. L. Baldwin. "The Prot.ein Folding Problem," Abstracts, AAAS Meeting, Boston
`MA, February 1988.
`
`Received May 3, 1988.
`
`12 of 12
`
`BI Exhibit 1076
`
`