Skip to main content
NCBI home page
Protein Science : A Publication of the Protein Society logoLink to Protein Science : A Publication of the Protein Society
. 2001 Dec;10(12):2627–2632. doi: 10.1110/ps.ps.26601a

Preferred proline puckerings in cis andtrans peptide groups: Implications for collagen stability

Luigi Vitagliano 1, Rita Berisio 1, Antonio Mastrangelo 2, Lelio Mazzarella 1,2,3, Adriana Zagari 1,2,3
PMCID: PMC2374046  PMID: 11714932

Abstract

The interplay between side-chain and main-chain conformations is a distinctive characteristic of proline residues. Here we report the results of a statistical analysis of proline conformations using a large protein database. In particular, we found that proline residues with the preceding peptide bond in the cis state preferentially adopt a down puckering. Indeed, out of 178 cis proline residues, as many as 145 (81%) are down. By analyzing the 1–4 and 1–5 nonbonding distances between backbone atoms, we provide a structural explanation for the observed trend. The observed correlation between proline puckering and peptide bond conformation suggests a new mechanism to explain the reported shift of the cis-trans equilibrium in proline derivatives. The implications of these results for the current models of collagen stability are also discussed.

Keywords: Proline, cis peptide, hydroxyproline, statistical analysis, collagen

Proline is unique among the genetically encoded amino acids (Richardson and Richardson 1989). The covalent bond between the side-chain and the nitrogen backbone atom has relevant structural consequences on the properties of the residue. In particular, (1) the cyclic nature of proline imposes severe restrictions to the conformational freedom of this residue; (2) there is a strict dependence between main- and side-chain proline conformations; (3) the cis state of the peptide group preceding proline residues is significantly populated; and (4) the lack of a backbone hydrogen-bonding donor limits proline capability to form main-chain hydrogen bonds.

Because of these intrinsic properties, the presence of proline strongly influences the protein structure elements embodying this residue. Indeed, it is well known that proline residues cannot fit into the regular, internal parts of either α-helices or β-sheets (Richardson and Richardson 1989). In contrast, proline residues are easily accommodated in the collagen triple helix (Beck and Brodsky 1998). Finally, the insertion of proline in loop regions may induce the large conformational changes characteristic of three-dimensional domain swapping (Mazzarella et al. 1995; Bergdoll et al. 1997; Liu et al. 2001).

Proline residues have been the object of several statistical analyses using both protein and oligopeptide databases (Balasubramanian et al. 1971; Stewart et al. 1990; MacArthur and Thornton 1991; Milner-White et al. 1992; Weiss et al. 1998). In this context, the analysis of a limited number of protein structures has indicated that there is a correlation between proline puckering and the cis/trans peptide bond conformation (Milner-White et al. 1992). Here we report a more extensive investigation regarding this issue by using a large protein structure database. Furthermore, we propose a structural explanation for the observed correlation between proline side-chain conformation and the peptide group state. Finally, we discuss these findings in relation to the cis-trans equilibrium of proline derivatives (Eberhardt et al. 1996; Bretscher et al. 2001) and to the current collagen stabilization models (Bella et al. 1994; Holmgren et al. 1998; Vitagliano et al. 2001).

Results and Discussion

Statistical analyses on proline residues

Statistical analyses on proline conformations were performed on a database containing nonredundant protein chains sharing sequence homologies <25%. Four hundred seventy-eight protein chains were selected from the Protein Data Bank (Berman et al. 2000) by considering only structures refined to a resolution better than 2.0 Å and an R-factor <0.23. Because reliable estimates of proline side-chain conformations require accurate determinations of the local region embodying the residue, additional selecting criteria, based on B-factors and puckering amplitude of each proline, were also applied (see Materials and Methods). The final database contained 3424 prolines, which were classified according to their conformational properties. In particular, for 3246 and 178 prolines, the peptide bond preceding the residue was in trans (transPro) and cis (cisPro) state, respectively. The percentage of cisPro (5.2%) is in close agreement with some previous surveys performed on protein structure databases (MacArthur and Thornton 1991; Weiss et al. 1998), although slightly higher values have been reported in other earlier studies (Stewart et al. 1990; Milner-White et al. 1992). The proportion of cisPro found in the present study is, however, significantly different from that derived from analyses performed on peptide structure databases. Indeed, it has been reported that the percentage of cisPro may be as high as 35% in oligopeptides (Milner-White et al. 1992). To identify possible causes for this discrepancy, we searched the Cambridge Structural Database (Allen et al. 1979) for cis prolines in cyclic and acyclic peptides. The results of this analysis show that the proportion of cisPro is 57.4% and 5.6% in cyclic and acyclic peptides, respectively. As expected, this finding indicates that the high proportion of cisPro is characteristic of cyclic peptides, in which the conformational restrictions imposed by the ring closure enhance the occurrence of the cis state.

Several statistical and energetic studies have shown that proline side-chains adopt two different conformational states usually denoted as up (upPro) and down (downPro; Némethy et al. 1992). UpPro and downPro can be identified by considering the distribution of proline side-chain dihedral angles. In particular, upPro is characterized by negative values of χ1 and χ3 and positive values of χ2 and χ4. The distribution of upPro and downPro in our database shows that the two states are equally populated. Indeed, out of 3424 prolines, 1738 (51%) are downPro, whereas 1686 are upPro. In the Cremer and Pople (1975) definition, upPro and downPro adopt phase angles clustered around 282° and 93°, respectively. These values are similar to those adopted by a regular five-membered cycle in the Cγ-exo envelope (288°) and twisted Cγ-endo–Cβ-exo (90°) conformations. In agreement with previous reports (Némethy et al. 1992; Vitagliano et al. 2001), upPro and downPro also show significantly different main-chain dihedral angles (Fig. 1A,B ▶). Indeed, the average values of ϕ are −58.7° and −69.8° for upPro and downPro, respectively.

Fig. 1.

Fig. 1.

Fig. 1.

Open in a new tab

ϕ versus χ1 (A) and ψ versus χ1 (B) in transPro (open circles) and cisPro (solid squares). It should be noted that only ψ angles >100° have been plotted.

Subsequently, we analyzed the distribution of cisPro and transPro among upPro and downPro. As many as 145 cisPro adopt a down conformation, whereas only 33 cisPro are upPro (Fig. 2 ▶). In contrast, the populations of down and up conformers are comparable for transPro. Similar preferences were obtained in previous studies performed on a more limited database constituted of 68 protein structures (Milner-White et al. 1992). It is worth noting that cisPro adopt ϕ angles characterized by more negative values (Figs. 1A, 2 ▶ ▶). Interestingly, conformations with ϕ > −60° are widely populated among transPro and essentially forbidden for cisPro. Because ϕ angles of downPro are significantly more negative than ϕ angles of upPro (Fig. 1A ▶), it is not surprising that cisPro preferentially adopts the down state.

Fig. 2.

Fig. 2.

Open in a new tab

ϕ versus ω in downPro (open triangles) and upPro (solid triangles).

Why should more negative values of the ϕ angle be required in cis prolines? As expected on the bases of pure geometrical considerations, several 1–4 and 1–5 distances between backbone atoms strongly depend on the value of the ϕ angle (Fig. 3A, B ▶). In particular, we focused our attention on the distances in the Ci−1–Ci, Cαi−1–Ci, and Oi−1–Ci pairs (Figs. 3, 4 ▶ ▶). As shown in Figure 4 ▶, all of these distances increase as ϕ assumes more negative values in both cis and trans conformations. Therefore, because downPro adopts more negative ϕ angles, these nonbonding distances are usually larger in downPro than in upPro (Fig. 4 ▶). The larger distances found in downPro may attenuate the severe steric hindrance occurring particularly in the cis state between the Cα atom (and its substituents) of the residue that precedes the proline and the carbonyl oxygen of the proline (Fig. 4D ▶). On the other hand, the interaction between Oi−1 and Ci occurring in transPro has a potential stabilizing effect (Bretscher et al. 2001). This allows rather short distances and thus large values of ϕ (Fig. 4C ▶). Consequently, both up and down states are compatible with trans conformations.

Fig. 3.

Fig. 3.

Open in a new tab

Schematic representation of 1–4 and 1–5 nonbonding backbone distances occurring in transPro (A) and cisPro (B).

Fig. 4.

Fig. 4.

Fig. 4.

Fig. 4.

Fig. 4.

Open in a new tab

Dependence of 1–4 and 1–5 nonbonding distances in downPro (open triangles ) and upPro (solid triangles ) on ϕ: Ci−1–Ci distances in transPro (A), Ci−1–Ci distances in cisPro (B), Oi−1–Ci distances in transPro (C), and Cαi−1–Ci distances in cisPro (D).

These observations do not necessarily imply that the presence of bulky residues preceding proline disfavors cis conformations, as specific interactions involving side-chains may occur. In fact, in our database ∼20% of residues preceding cisPro are aromatic, in accordance with previous reports (Reimer et al. 1998; Jabs et al. 1999). The high frequency of aromatic residues preceding cisPro may be attributed to the CH…π interactions between the proline and the aromatic rings, as reported by Jabs et al. (1999).

A new mechanism to explain the cis-trans equilibrium observed in proline derivatives

The cis-trans equilibrium of substituted pyrrolidine rings has been the object of several recent investigations for its implications on collagen stability. In a series of interesting studies, Raines and coworkers have shown that the presence of electron-withdrawing atoms on proline ring may significantly affect the cis-trans equilibrium of the preceding peptide group (Eberhardt et al. 1996; Bretscher et al. 2001). Indeed, they have shown that this equilibrium is more shifted toward the trans state in 4R-hydroxyproline (4R-Hyp) and 4R-fluoroproline (4R-Flp) than in proline (Eberhardt et al. 1996). Interestingly, in 4S-hydroxyproline (4S-Hyp) and 4S-fluoroproline (4S-Flp) the equilibrium is shifted toward the cis state (Bretscher et al. 2001). These experimental data have been interpreted considering the interactions of the proline carbonyl group with the carbonyl group of the residue preceding the proline and with the ring substituent (Bretscher et al. 2001).

We propose an additional mechanism that explains the experimental data on the bases of the above described correlation between proline puckering and peptide bond conformation. Several experimental reports and statistical analyses have indicated that 4R-Hyp and 4R-Flp preferentially adopt the up state (Holmgren et al. 1999; Vitagliano et al. 2001). Because we have shown that up conformers strongly adverse the cis state, the shift observed for 4R-Hyp and 4R-Flp toward the trans state (Eberhardt et al. 1996) is straightforwardly explained. On the other hand, there are indications that 4S-Hyp and 4S-Flp preferentially adopt the down state (Gerig and McLeod 1973; Shamala et al. 1976; Holmgren et al. 1999). Therefore, they are more prone to assume the cis state.

Although other effects such as interactions of backbone atoms and pyrrolidine substituents may play a role in the cis-trans equilibrium, the above described mechanism represents a reliable explanation for the observed experimental data (Eberhardt et al. 1996; Bretscher et al. 2001). Furthermore, the mechanism highlights the interplay between the conformational properties of backbone and side-chain in imino acids. Because the conformational preferences of the substituted imino acid side-chains are likely driven by stereo-electronic effects (Holmgren et al. 1999; Bretscher et al. 2001), our mechanism emphasizes the importance of inductive effects in directing the conformational properties of proteins.

Implications for collagen stability

The structural role of hydroxyproline residues on collagen stability is a long-standing issue that recently has received much attention. Earlier hypotheses on the role of the hydroxyproline were essentially based on its interaction with the solvent (Ramachandran et al. 1973; Privalov 1982). Although high-resolution X-ray structures of collagen-like polypeptides have indeed shown that hydroxyproline residues strongly interact with the solvent (Bella et al. 1994, 1995; Berisio et al. 2001), recent findings have shifted the attention toward other effects. Indeed, Raines and coworkers (Holmgren et al. 1998,1999) have indicated that collagen stabilization induced by hydroxyproline relies on the shift of the cis-trans equilibrium toward the trans state, which is the one required in collagen triple helix. This stabilization model does not explain the inability of polypeptides with repeating sequence Hyp-Pro-Gly to fold in triple helix (Inouye et al. 1982), even when the three chains are covalently linked (Henkel et al. 1999). Based on the observation that upPro and downPro ϕ dihedral angles suited the X and Y positions, respectively, in the repetitive X-Y-Gly sequence of the collagen triple helix and that 4R-Hyp preferentially adopts an up state, we have recently proposed a new stabilization model (Vitagliano et al. 2001). This model is able to explain the effects of the presence of 4R-Hyp in both the X and Y positions on the stability of imino acid–rich collagen-like polypeptides. The suitability of ψ angles of downPro and upPro to the X and Y positions, respectively, is confirmed in the present statistical analysis. Furthermore, we show that ψ angles typical of imino acids in the X position (∼160°) are allowed mainly to downPro (Fig. 1B ▶).

It is worth mentioning that very recent comparative studies on (Pro-Thr-Gly)10 and (4R-Hyp-Thr-Gly)10 have shown that 4R-Hyp in the X position may have stabilizing effects on the triple helix (Bann and Bächinger 2000). Based on this observation, it can be surmised that intrinsic conformational properties of imino acids may play a crucial role in the stabilization of imino acid–rich collagen polypeptides, in which even small differences in main-chain dihedral angles may not be tolerated. In contrast, main-chain distortions may be more easily accommodated in less rigid collagen-like polypeptides as (Pro-Thr-Gly)10. Therefore, it is likely that for these polypeptides, other effects such as the shifting of the cis-trans equilibrium and networking through water molecules may explain the stabilizing role of 4R-Hyp. Because real collagen molecules consist of both regions with low and high content of imino acids, this observation leads to the suggestive hypothesis that stabilization effects of 4R-Hyp may be different in the distinct regions of the collagen molecule. This idea is supported by the finding that the triple helical parameters are different in regions with high and low content of imino acids (Kramer et al. 1999) and that Gly-Pro-Hyp and Gly-X-Hyp triplets contribute differently to the triple helix stability (Burjanadze 2000).

Materials and methods

The May 1999 release of the Protein Data Bank (Berman et al. 2000) was searched for proline residues. In particular, 478 nonredundant protein chains having sequence homology <25% and belonging to structures refined to a resolution better than 2.0 Å and to an R-factor <0.23 were considered. These chains contained 5168 proline residues, which were further selected according to their B-factors and puckering amplitudes. Proline residues showing average B-factors >1.2 times the average B-factor of the protein structure they belonged to were rejected. In addition, the Cremer and Pople (1975) algorithm was used to exclude proline residues with an unusual puckering (Hooft et al. 1996). Indeed, only proline residues with puckering amplitudes in the range of 0.20–0.45 were included in the final database, which contained 3424 proline residues.

Up or down puckerings were assigned depending on the combination of pyrrolidine ring dihedral angles. Namely, proline rings with (χ1 + χ3 − χ2 − χ4) > 40° were regarded as down, whereas those with (χ1 + χ3 − χ2 − χ4) < −40° were regarded as up.

Acknowledgments

We are indebted to Mr. Luca De Luca for technical assistance. This work was financially supported by the Agenzia Spaziale Italiana and by the CNR Agenzia2000.

The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact.

Abbreviations

  • cisPro and transPro, proline residues with the peptide bond of the preceding residue in the cis and trans conformations, respectively

  • upPro and downPro, proline residues with up and down side-chain puckerings, respectively

  • 4R-Hyp, 4R-hydroxyproline

  • 4S-Hyp, 4S-hydroxyproline

  • 4R-Flp, 4R-fluoroproline

  • 4S-Flp, 4S-fluoroproline

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1101/ps.26601.

References

  1. Allen, F.H., Bellard, S., Brice, M.D., Cartwright, B.A., Doubleday, A., Higgs, H., Hummelink, T., Hummelink-Peters, B.G., Kennard, O., Motherwell, W.D.S., Rodgers, J.R., and Watson, D.G. 1979. The Cambridge Crystallographic Data Centre: Computer-based search, retrieval, analysis and display of information. Acta Crystallogr. B 35 2331–2339. [Google Scholar]
  2. Balasubramanian, R., Lakshminarayanan, A.V., Sabesan, M.N., Tegoni, G., Venkatesan, K., and Ramachandran, G.N. 1971. Studies on the conformation of amino acids, VI: Conformation of the proline ring as observed in crystal structures of amino acids and peptides. Int. J. Protein Res. 3 25–33. [PubMed] [Google Scholar]
  3. Bann, J.G. and Bächinger, H.P. 2000. Glycosylation/hydroxylation-induced stabilization of the collagen triple helix: 4-trans-hydroxyproline in the Xaa position can stabilize the triple helix. J. Biol. Chem. 275 24466–24469. [DOI] [PubMed] [Google Scholar]
  4. Beck, K. and Brodsky, B. 1998. Supercoiled protein motifs: The collagen triple-helix and the α-helical coiled coil. J. Struct. Biol. 122 17–29. [DOI] [PubMed] [Google Scholar]
  5. Bella, J., Eaton, M., Brodsky, B., and Berman, H.M. 1994. Crystal and molecular structure of a collagen-like peptide at 1.9 Å resolution. Science 266 75–81. [DOI] [PubMed] [Google Scholar]
  6. Bella, J., Brodsky, B., and Berman, H.M. 1995. Hydration structure of a collagen peptide. Structure 3 893–906. [DOI] [PubMed] [Google Scholar]
  7. Bergdoll, M., Remy, M.H., Cagnon, C., Masson, J.M., and Dumas, P. 1997. Proline-dependent oligomerization with arm exchange. Structure 5 391–401. [DOI] [PubMed] [Google Scholar]
  8. Berisio, R., Vitagliano, L., Mazzarella, L., and Zagari, A. 2001. Crystal structure determination of the collagen-like polypeptide with repeating sequence Pro-Hyp-Gly: Implications for hydration. Biopolymers 56 8–13. [DOI] [PubMed] [Google Scholar]
  9. Berman, H.M., Bhat, T.N., Bourne, P.E., Feng, Z., Gilliland, G., Weissig, H., and Westbrook, J. 2000. The Protein Data Bank and the challenge of structural genomics. Nat. Struct. Biol. 7 957–959. [DOI] [PubMed] [Google Scholar]
  10. Bretscher, L.E., Jenkins, C.L., Taylor, K.M., DeRider, M.L., and Raines, R.T. 2001. Conformational stability of collagen relies on a stereoelectronic effect. J. Am. Chem. Soc. 123 777–778. [DOI] [PubMed] [Google Scholar]
  11. Burjanadze, T.V. 2000. New analysis of the phylogenetic change of collagen thermostability. Biopolymers 53 523–528. [DOI] [PubMed] [Google Scholar]
  12. Cremer, D. and Pople, J.A. 1975. General definition of ring puckering coordinates. J. Amer. Chem. Soc. 97 1354–1358. [Google Scholar]
  13. Eberhardt, E.S., Panisik, N.J., and Raines, R.T. 1996. Inductive effects on the energetics of prolyl peptide bond isomerization: Implications for collagen folding and stability. J. Am. Chem. Soc. 118 12261–12266. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gerig, J.T. and McLeod, R.S. 1973. Conformations of cis- and trans-4-fluoro-l-proline in aqueous solution. J. Am. Chem. Soc. 95 5725–5729. [DOI] [PubMed] [Google Scholar]
  15. Henkel, W., Vogl, T., Echner, H., Voelter, W., Urbanke, C., Schleuder, D., and Rauterberg, J. 1999. Synthesis and folding of native collagen III model peptides. Biochemistry 38 13610–13622. [DOI] [PubMed] [Google Scholar]
  16. Holmgren, S.K., Bretscher, L.E., Taylor, K.M., and Raines, R.T. 1999. A hyperstable collagen mimic. Chem. Biol. 6 63–70. [DOI] [PubMed] [Google Scholar]
  17. Holmgren, S.K., Taylor, K.M., Bretscher, L.E., and Raines, R.T. 1998. Code for collagen's stability deciphered. Nature 392 666–667. [DOI] [PubMed] [Google Scholar]
  18. Hooft, R.W., Vriend, G., Sander, C., and Abola, E.E. 1996. Errors in protein structures. Nature 381 272. [DOI] [PubMed] [Google Scholar]
  19. Inouye, K., Kobayashi, Y., Kyogoku, Y., Kishida, Y., Sakakibara, S., and Prockop, D. J. 1982. Synthesis and physical properties of (hydroxyproline-proline-glycine)10: Hydroxyproline in the X-position decreases the melting temperature of the collagen triple helix. Arch. Biochem. Biophys. 219 198–203. [DOI] [PubMed] [Google Scholar]
  20. Jabs, A., Weiss, M.S., and Hilgenfeld, R. 1999. Non-proline cis peptide bonds in proteins. J. Mol. Biol. 286 291–304. [DOI] [PubMed] [Google Scholar]
  21. Kramer, R.Z., Bella, J., Mayville, P., Brodsky, B., and Berman, H.M. 1999. Sequence dependent conformational variations of collagen triple-helical structure. Nat. Struct. Biol. 6 454–457. [DOI] [PubMed] [Google Scholar]
  22. Liu, Y., Gotte, G., Libonati, M., and Eisenberg, D. 2001. A domain-swapped RNase A dimer with implications for amyloid formation. Nat. Struct. Biol. 8 211–214. [DOI] [PubMed] [Google Scholar]
  23. MacArthur, M.W. and Thornton, J.M. 1991. Influence of proline residues on protein conformation. J. Mol. Biol. 218 397–412. [DOI] [PubMed] [Google Scholar]
  24. Mazzarella, L., Vitagliano, L., and Zagari, A. 1995. Swapping structural determinants of ribonucleases: An energetic analysis of the hinge peptide 16–22. Proc. Natl. Acad. Sci. 92 3799–3803. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Milner-White, E.J., Bell, L.H., and Maccallum, P.H. 1992. Pyrrolidine ring puckering in cis and trans-proline residues in proteins and polypeptides: Different puckers are favoured in certain situations. J. Mol. Biol. 228 725–734. [DOI] [PubMed] [Google Scholar]
  26. Némethy, G., Gibson, K.D., Palmer, K.A., Yoon, C.N., Paterlini, G., Zagari, A., Rumsey, S., and Scheraga, H.A. 1992. Energy parameters in polypeptides. 10. Improved geometrical parameters and nonbonded interactions for use in ECEPP/3 algorithm, with application to proline-containing peptides. J. Phys. Chem. 96 6472–6484. [Google Scholar]
  27. Privalov, P.L. 1982. Stability of proteins: Proteins which do not present a single cooperative system. Adv. Protein Chem. 35 1–104. [PubMed] [Google Scholar]
  28. Ramachandran, G.N., Bansal, M., and Bhatnagar, R.S. 1973. A hypothesis on the role of hydroxyproline in stabilizing collagen structure. Biochim. Biophys. Acta 322 166–171. [DOI] [PubMed] [Google Scholar]
  29. Reimer, U., Scherer, G., Drewello, M., Kruber, S., Schutkowski, M., and Fischer, G. 1998. Side-chain effects on peptidyl-prolyl cis/trans isomerisation. J. Mol. Biol. 279 449–60. [DOI] [PubMed] [Google Scholar]
  30. Richardson, J.S. and Richardson, D.C. 1989. Principles and patterns of protein conformation. In Prediction of protein structure and the principles of protein conformation (ed. G.D. Fasman), pp. 1–98. Plenum Press, New York.
  31. Shamala, N., Guru Row, T.N., and Venkatasen, K. 1976. The crystal and molecular structure of allo-4-hydroxy-l-proline dihydrate. Acta Crystallogr. B 32 3267–3270. [Google Scholar]
  32. Stewart, D.E., Sarkar, A., and Wampler, J.E. 1990. Occurrence and role of cis peptide bonds in protein structures. J. Mol. Biol. 214 253–260. [DOI] [PubMed] [Google Scholar]
  33. Vitagliano, L., Berisio, R., Mazzarella, L., and Zagari, A. 2001. Structural bases of collagen stabilization induced by proline hydroxylation. Biopolymers 36 459–464. [DOI] [PubMed] [Google Scholar]
  34. Weiss, M.S., Jabs, A., and Hilgenfeld, R. 1998. Peptide bonds revisited. Nat. Struct. Biol. 5 676. [DOI] [PubMed] [Google Scholar]

Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society

RESOURCES