cond-mat0205559/a.tex
1: 
2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3: %                                                                     %
4: %  A LaTeX file with some definitions at the top.                     %
5: %                                                                     %
6: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
7: %
8: %                                                       MTU-PHY-HA-01/9
9: %                                                       Dec. 2,   2001
10: %                                               revised Mar.18,   2002 
11: 
12: %%%\documentstyle[aps,prb,preprint,floats]{revtex}
13: %\documentstyle[aps,prl,twocolumn,graphicx,floats]{revtex}
14: %options [aps,prl,pre,preprint,twocolumn,floats,graphicx]
15: \documentstyle[aps,prl,graphicx,floats]{revtex}
16: 
17: 
18: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
19: %    ONLY FOR ALVES PRINTER  %
20: \voffset=1.6cm 
21: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
22: 
23: 
24: \begin{document}
25: 
26: % \draft command makes pacs numbers print
27: \draft
28: % repeat the \author\address pair as needed
29: %{\wideabs{ 
30: \title{Helix Formation and Folding in an Artificial Peptide}   
31: \author{ Nelson A. Alves\footnote{E-mail: alves@quark.ffclrp.usp.br}}
32: \address{Departamento de F\'{\i}sica e Matem\'atica, FFCLRP
33:      Universidade de S\~ao Paulo. Av. Bandeirantes 3900. 
34:      CEP 014040-901 \, Ribeir\~ao Preto, SP, Brazil}
35: \author{Ulrich H.E. Hansmann \footnote{hansmann@mtu.edu, 
36:                         to whom all correspondence should be addressed}}
37: \address{Department of Physics, Michigan Technological University,
38:          Houghton, MI 49931-1291, USA}
39: \date{\today}
40: \maketitle
41: \begin{abstract}
42: We study the relation between $\alpha$-helix formation and 
43: folding  for a simple artificial peptide, Ala$_{10}$-Gly$_5$-Ala$_{10}$. 
44: Our data rely on multicanonical Monte Carlo simulations 
45: where the interactions among all atoms are taken into account. 
46: The free-energy landscape of the peptide is evaluated for various 
47: temperatures. Our data indicate that folding of this peptide is a
48: two-step process: in a first step two $\alpha$-helices are formed which
49: afterwards re-arrange themselves  into a U-like structure.
50: \end{abstract}
51: %\pacs{}
52: %}}
53: %\begin{multicols}{2}
54: 
55: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
56: \section{Introduction}
57: The mechanism by which a large class of proteins folds spontaneously
58: into  a unique  globular shape \cite{Anf} has remained elusive. 
59: Significant new insight was gained over the last few years from the 
60: studies of minimal protein models. For instance,  energy landscape theory
61: and funnel concept \cite{Bryngelson87,Onuchic97} proved to be
62: powerful tools for description of the general characteristics  of folding
63: not only in  minimalistic protein models  but also for real proteins 
64: \cite{HOO98b,HO01b}. However, many questions
65: on the details of the folding process remain to be solved. For instance,
66: folding of proteins involves one or more transitions  between  different 
67: thermodynamic states. The role of these transitions in the folding
68: process is an active area of research. An important example for
69: these transitions is  the formation of secondary 
70: structure elements.
71: %  such as $\alpha$-helices or $\beta$-sheets.  
72:  For the case of $\alpha$-helices it is long known that
73: there is a sharp transition towards a random coil state when the 
74: temperature is increased. The characteristics of this so-called 
75: helix-coil transition have been studied extensively \cite{Poland},
76: most recently in Refs.~\cite{Jeff,HO98c,AH99b,AH00b,MO2,PH01f}.
77: In this paper, we  research  the relation between helix-coil transition 
78: and folding. 
79: 
80: For this purpose, we have studied an artificial peptide, 
81: Ala$_{10}$-Gly$_5$-Ala$_{10}$, in a detailed representation 
82: where the interactions between all atoms are taken into account. 
83: Multicanonical simulations \cite{MU} with large statistics are 
84: used to evaluate the free energy landscape of our peptide at 
85: different temperatures.  The encountered transitions are further
86: investigated 
87: by partition zeros analysis which allows to characterize 
88: ``phase transitions'' in small systems \cite{BMH}.  Quantities such as 
89: the energy, specific heat, helicity and susceptibility were calculated 
90: as function of temperature. 
91: We have neglected in the simulations the interaction of our artifical
92: peptide with the surrounding solvent. While this is certainly
93: a crude approximation, it allows us not only to relate our results 
94: to our previous studies  on helix-coil transition
95: in poly-alanine that also relied on gas-phase simulations 
96: \cite{HO98c,AH99b,AH00b}, but also  
97:  to study the extend to that secondary structure formation and
98: folding are determined by intrinsic  properties of the peptide.
99: Our data suggest that the peptide  in gas-phase folds 
100: in a two-step process: in a first step two $\alpha$-helices are formed 
101: in what amounts to a first order transition. Afterwards these helices 
102: re-arrange themselves  into a U-like structure. 
103: The second step has the characteristics of a second order transition.
104: 
105: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
106: 
107: \section{Methods}
108: Our investigation of  Ala$_{10}$-Gly$_5$-Ala$_{10}$ is based on a detailed, 
109: all-atom representation of that peptide. The interaction between the 
110: atoms is described by a standard force field, ECEPP/2,\cite{EC}  
111: (as implemented in the  program package SMMP \cite{SMMP}) and is given by:
112: \begin{eqnarray}
113: E_{tot} & = & E_{C} + E_{LJ} + E_{HB} + E_{tor},\\
114: E_{C}  & = & \sum_{(i,j)} \frac{332q_i q_j}{\epsilon r_{ij}},\\
115: E_{LJ} & = & \sum_{(i,j)} \left( \frac{A_{ij}}{r^{12}_{ij}}
116:                                 - \frac{B_{ij}}{r^6_{ij}} \right),\\
117: E_{HB}  & = & \sum_{(i,j)} \left( \frac{C_{ij}}{r^{12}_{ij}}
118:                                 - \frac{D_{ij}}{r^{10}_{ij}} \right),\\
119: E_{tor}& = & \sum_l U_l \left( 1 \pm \cos (n_l \chi_l ) \right).
120: \end{eqnarray}
121: Here, $r_{ij}$ (in \AA) is the distance between the atoms $i$ and $j$,
122:  and $\chi_l$ is the $l$-th torsion angle. The peptide bond angles 
123: were set to their common value $\omega = 180^{\circ}$. We do not  
124: include explicitly  the interaction of the peptide with the solvent 
125: into our simulations and set the dielectric constant $\epsilon$ equal to 2.
126: Since the charges at peptide termini are known to
127: reduce helix content \cite{SKY}, we removed them by taking a 
128: neutral NH$_2$-- group at the
129: N-terminus and a neutral --COOH group at the C-terminus.
130: 
131: Simulation of detailed protein models where the interaction between all
132: atoms are taken into account are extremely difficult. This is because 
133: the various competing interactions within the molecule lead to an energy 
134: landscape characterized by a multitude of local minima separated by
135: high energy barriers.  Hence, in the low-temperature region,
136: canonical Monte Carlo or molecular dynamics simulations will tend to get 
137: trapped in one of these minima and the simulation will not thermalize 
138: within the available CPU time. One example of the new and sophisticated
139: algorithms \cite{curr_op} that allow to overcome this difficulty are
140: {\it generalized-ensemble} techniques \cite{HO,OurReview98}, and 
141: it is one of these techniques, multicanonical 
142: sampling \cite{MU}, that we used for our investigations. 
143: 
144: In the multicanonical algorithm \cite{MU}
145: conformations with energy $E$ are assigned a weight
146: $  w_{mu} (E)\propto 1/n(E)$. Here, $n(E)$ is the density of states.
147: A  simulation with this weight
148: will  lead to a uniform distribution of energy:
149: \begin{equation}
150:   P_{mu}(E) \,  \propto \,  n(E)~w_{mu}(E) = {\rm const}~.
151: \label{eqmu}
152: \end{equation}
153: This is because the simulation generates a 1D random walk in the
154: energy space,
155: allowing itself to escape from any  local minimum.
156: Since a large range of energies are sampled, one can
157: use the reweighting techniques \cite{FS} to  calculate thermodynamic
158: quantities over a wide range of temperatures $T$ by
159: \begin{equation}
160: <{\cal{A}}>_T ~=~ \frac{{\int dx~{\cal{A}}(x)~w^{-1}(E(x))~
161:                  e^{-\beta E(x)}}}
162:               {{\int dx~w^{-1}(E(x))~e^{-\beta E(x)}}}~,
163: \label{eqrw}
164: \end{equation}
165: where $x$ stands for configurations.
166: 
167: Note that unlike in the case of canonical simulations the weights
168: \begin{equation}
169: w(E) = n^{-1}(E) = e^{-S(E)}
170: \label{eqweight}
171: \end{equation}
172: are not a priori known. Instead estimators for these weights have to
173: be determined  by  an iterative procedure \cite{Berg,PH01f}. In our
174: case we needed 500,000  sweeps for the  weight factor calculations. 
175: All thermodynamic quantities were  then estimated from one production 
176: run of $8,000,000$ Monte Carlo sweeps which followed 10,000 sweeps for
177: thermalization.  Our simulations were started from  completely random 
178: initial conformations (Hot Start) and  one Monte Carlo sweep updates 
179: every torsion angle of the peptide once.  At the end of every 10th 
180: sweep we stored the ECEPP/2 energies $E_{tot},E_C, E_{LJ}, E_{hb}$ 
181: and $E_{tor}$ of the conformation,  the corresponding number $n_H$ 
182: of helical residues and end-to-end distance $d_{e-e}$. 
183: Here, we follow  previous work \cite{OH95b} 
184: and consider a residue as helical if its backbone angle $(\phi,\psi)$ are 
185: within the range $(-70^{\circ}\pm 20^{\circ},-37^{\circ}\pm20^{\circ})$.
186: 
187: 
188: Using the results of our generalized-ensemble simulation, we explored for
189: various temperatures the free energies
190: \begin{equation}
191: G(n_H,d_{e-e}) = -k_B T \log P(n_H,d_{e-e})~.
192: \end{equation}
193: Here, $P(n_H,d_{e-e})$ is the probability to find a peptide conformation 
194: with values $\ell$, $d_{e-e}$ (at temperature $T$). We chose the 
195: normalization so that the lowest value of $G(n_H,d_{e-e})$ is set to 
196: zero for each temperature.
197: 
198: We finally used that the multicanonical algorithm allows us
199: to calculate estimates for the spectral density:
200: \begin{equation}
201:   n(E) = P_{mu} (E) w^{-1}_{mu} (E)~.
202: \end{equation}
203: We can therefore construct
204: the corresponding  partition function for our all-atom model of
205:  Ala$_{10}$-Gly$_5$-Ala$_{10}$ from these estimates by
206: \begin{equation}
207:      Z(\beta) = \sum_{E} n(E) e^{- \beta E} ,                   \label{eq:r1}
208: \end{equation}
209: with $\beta$ the inverse temperature, $\beta = 1/k_B T$.
210: The complex solutions of the partition function
211: determine the critical behavior of the model and were also studied by us.
212: 
213: 
214: 
215: 
216: 
217: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
218: 
219: \section{Results and Discussion}
220: Our peptide,  Ala$_{10}$-Gly$_5$-Ala$_{10}$, is build up out of two
221: chains of each 10 alanine residues connected by 5 glycine residues. 
222: In previous work \cite{OH95b,HO98c,AH99b} we could show that 
223: polyalanine has a pronounced transition between a disordered 
224: coil phase and an ordered state in which the polymer forms 
225: an $\alpha$-helix. For this reason, we  expect  formation 
226: of $\alpha$-helices in our peptide, and the average number of helical 
227: residues $<n_H>$ is therefore one of the quantities
228: that we have measured.  $<n_H>$ is displayed in Fig.~1 as a function of 
229: temperature, and we observe in this plot  two temperature
230: regions. At high temperature, few  residues are found with backbone
231: dihedral angles $(\phi,\psi)$ typical for an $\alpha$-helix. On the 
232: other hand, at low temperatures we observe helix-formation,
233: and almost all of the alanine residues are part of an $\alpha$-helix, 
234: i.e. have backbone dihedral angles  $(\phi,\psi)$ in the range 
235: ($-70^{\circ}\pm 20^{\circ},-37^{\circ}\pm 20^{\circ}$). 
236: The transition between the two temperature
237: regions is sharp indicating the existence of a helix-coil transition.
238: The  transition temperature $T_{hc}$ can be determined from the
239: corresponding peak in the susceptibility
240: \begin{equation}
241:  \chi(T) = <n^2_H(T)> - <n_H(T)>^2 \, ,
242: \end{equation}
243: which is ploted in the inset of Fig.~1, and we find the
244: transition temperature $T_{hc} = 485\pm 5$ K. 
245: 
246: In previous work \cite{OH95b,PH01f} we could show that in polyalanine 
247: the formation of $\alpha$-helices is related to a gain in potential 
248: energy. For this reason, we display in Fig.~2  the average total 
249: ECEPP/2 energy $<E_{tot}>$ and the thermodynamic averages of partial 
250: energies $<E_C>$, $<E_{LJ}>$, $<E_{hb}>$ and $<E_{tor}>$ of
251: Ala$_{10}$-Gly$_5$-Ala$_{10}$ as a function of temperature.
252: As expected, we observe around our transition temperature $T_{hc}$
253: a sharp decrease in the potential energy $<E_{tot}>$ that  is  due to 
254: a corresponding decrease   in the Coulomb energy $<E_C>$, 
255: Lennard-Jones energy $<E_{LJ}>$ and hydrogen-bond energy $<E_{hb}>$.
256: The change in  $<E_{tot}>$ with temperature can be described by the 
257: specific heat
258: \begin{equation}
259:   C(T)  = \beta^2 \ \frac{<E_{tot}^2> - <E_{tot}>^2}{25},
260: \end{equation}
261: which we display in Fig.~3.  A pronounced change in energy with 
262: temperature corresponds to a peak in the specific heat.  As one 
263: can see from Fig.~3, we observe indeed a pronounced peak in the 
264: specific heat at a temperature $T=480 \pm 10$ K that is consistent
265: with $T=485 \pm 5$ K,  the temperature where the peak in the 
266: susceptibility is located.  Combining both values we obtain as our 
267: final estimate for the  helix-coil transition temperature 
268: $T_{hc}=483\pm 8$. However, we find in Fig.~3 also a second, smaller 
269: peak in the specific heat at the lower temperature $T_f=265\pm 7$ K 
270: indicating yet another transition.
271: 
272: The sharp decrease in potential energy corresponding to
273: this second peak is clearly  visible in Fig.~2 for the total energy
274:  $<E_{tot}>$, however, not for all of the partial energies. Only the
275: Lennard-Jones term $<E_{LJ}>$ exhibits also a signal for the second
276: transition at $T_f$.  This energy term depends strongly on the
277: overall size of the molecule and the change in this quantity
278: indicates a transition between extended and compact structures.
279: Hence, we conjecture that the second peak in specific heat at the
280: lower temperature $T_f$ is related to a transition between extended
281: and compact structures.  A possible measure for such a change 
282: is the average end-to-end distance $<d_{e-e}>_T$.  We define here 
283:  $<d_{e-e}>$ as the distance between N of Ala$_1$ and O of Ala$_{25}$,
284: and plot this quantity in Fig.~4. We observe that this quantity decreases
285: with decreasing temperature. Below the helix-coil transition $T_{hc}$
286: the decrease slows down and the curve becomes almost flat at a value
287: of $<d_{e-e}> \approx 10$ \AA \ indicating that
288: there is little further change in the compactness of the molecule. 
289: However, at temperature $T_f$  the end-to-end distance decreases
290: again sharply towards a new value $<d_{e-e}> = 6.1$ \AA .  
291: Hence, $T_f$ marks the folding of the molecule into  a defined 
292: compact structure  with 
293: the two terminal ends of the peptide  close together.
294: 
295: The transition between extended and a more compact structure can also 
296: be seen when we display the free energy landscape of our peptide
297: as a function of helicity $n_H$ and end-to-end distance $d_{e-e}$.
298: At the temperature $T=480$ K (which is essentially the helix-coil 
299: transition temperature $T_{hc}= 483\pm 8$ K) the free energy 
300: landscape (displayed in Fig.~5a) is flat over a large range of values
301: of $n_H$ and $d_{e-e}$. The $3 k_BT$ contour line surrounds a region where
302: the helicity can take values between $0 \le n_H \lesssim  20$ and
303: the end-to-end distance values between $3 \lesssim d_{e-e} \lesssim 40$,
304: allowing the system to move freely between extended and compact
305: configurations, and between helical and coil configurations.
306: On the other hand, at the second and lower  temperature 
307:  $T=270$ K (which is essentially the  folding transition
308: temperature $T_f=265\pm 7$ K) the free energy grows rapidly with 
309: decreasing helicity $n_H$ favoring configuration in a small strip 
310: with $ 15 \lesssim n_H \lesssim 25$. Hence, the plot of the free
311: energy landscape in Fig.~5b is limited to  values $12 \lesssim 
312: n_H \lesssim 25$ of the helicity. Here,  two regions of minimal 
313: free energy can be seen (marked by the 3 $k_BT$ contour lines). 
314: The first minima is found at values of $d_{e-e}$ between 5 and 10 \AA \  
315: and $15 \le n_H \le 20$   characterising compact structures. 
316: A second region with slightly lower free energy (see the 1 $k_BT$ 
317: contour line) is found  at much larger values of $d_{e-e}$ 
318: between 35 and 40 \AA \ and $ 20 < n_H \le 25$ indicating a long 
319: stretched $\alpha$-helix. Both local free energy minima are 
320: separated by free energy barriers of  $\approx 8$ $k_BT$  that 
321: can be  overcome by thermal fluctuations. On the other hand,
322: configurations with helicity $n_H < 10$ are suppressed by free 
323: energy differences of more than 30 $k_BT$. 
324: 
325: 
326: 
327: Examples for the structures corresponding to the two free energy minima
328: are plotted in Fig.~6. The first one, displayed in Fig.~6a is the 
329: configuration with lowest energy ever found in our multicanonical
330: simulation of 8,000,000 sweeps and corresponds to the region in
331: the free-energy landscape at  values of $d_{e-e}$ between 5 and 10 \AA \
332: and $15 \le n_H \le 20$. This conformation (`A') consists  out
333: of two helixes (made up out of the alanine residues) connected by
334: a turn (build out of the flexible glycine residues) towards a 
335: U-turn-like structure that is consistent with the small value of
336: the end-to-end distance $d_{e-e}$ observed in Fig.~4 for 
337: temperatures below $T_f$.  For reference we show  in Fig.~6b 
338: also the configuration (`B') where all
339: 25 residues are part of an $\alpha$-helix and which corresponds to
340: the second local free-energy minimum in Fig.~5b at values of
341:  $d_{e-e}$ between 35 and 40 \AA \ and $ 20 < n_H \le 25$ . The  
342: dihedral angles of both configurations are listed in Table 1.  
343: Fig.~7 displays the frequency of both 
344: configurations as a function of temperature.  For $T > T_{hc}$ neither
345: configuration `A' nor `B' are observed. Below that temperature 
346: both structures appear with similar probability as long as the temperature
347: is higher than $T_f$. At $T=T_f$ the probability to find the maximal
348: helical structure `B' is with $\approx 30$ \% highest and decreases after that
349: with decreasing temperature. On the other hand, the frequency
350: for the U-turn structure `A' continues to grow with decreasing temperature. 
351: This different behavior is due to the energy differences between
352: both structures. The minimal
353: energy conformation `A' has with $E_{tot} =-34.7$ Kcal/mol a $10.8 $ Kcal/mol
354: lower potential energy than the extended helix conformation `B'
355: ($E_{tot}=-23.9 $ Kcal/mol). This difference is mainly due to the
356: Lennard-Jones terms: $E_{LJ}=-132.5$ Kcal/mol for `A' 
357: vs. $E_{LJ}=-118.9$ Kcal/mol
358: for `B'. The gain in $E_{LJ}$ is  in part compensated by the
359: hydrogen-bonding terms: $E_{hb}=-30.2$ Kcal/mol  for `A' vs. $E_{hb}=-34.7$
360: Kcal/mol for `B'. Coulomb and torsion energies differed little between
361: the two configurations: $E_C=126.7$ Kcal/mol and $E_{tor}=1.3$ Kcal/mol for
362: conformation `A' vs. $E_C=126.4$ Kcal/mol and $E_{tor}=3.3$ Kcal/mol for
363: conformation `B'. 
364: 
365: It is an interesting question whether our two observed transitions
366: (occuring in a finite and small system) can be
367: related to phase transitions which in a strict sense are defined only
368: for macroscopic (that is very large) systems. In order to study this
369: question we have  calculated the 
370: complex zeros $\beta \rightarrow {\rm Re}(\beta) + i \tau$
371: of the partition function $Z(\beta)$ of our molecule.  
372: In the case of a temperature driven phase transition, 
373: we expect that the complex zeros
374: $\beta_j,~(j=1,2, ...)$
375: (or at least the ones close to the real axis) 
376: condense for large enough system size  on a single line.
377: As the system size  increases, those zeros will move towards
378: the positive real $\beta$-axis  and the corresponding value is for large
379: system size  the inverse of the physical critical temperature $T_c$.
380: Crucial information on phase transitions can be obtained from
381: the way in which the first zero approaches the real $\beta$-axis.
382: However, such an analysis depends on the extrapolation
383: towards the infinite large system and does not allow  characterization
384: of the situation in small systems such as Ala$_{10}$-Gly$_5$-Ala$_{10}$.
385: One possible extension of the above ideas to ``phase transitions'' in
386: biological molecules and other small systems is the classification 
387: scheme by  Borrmann {\it et al.} \cite{BMH}. 
388:  In this approach one computes the discrete line density of zeros as
389: an average of the inverse distances between neighboring zeros,
390: \begin{equation}
391: \phi(\tau_k) = \frac{1}{2} \left(\frac{1}{|\beta_k - \beta_{k-1}|}
392:        + \frac{1}{|\beta_{k+1} - \beta_{k}|}\right)\,,     \label{eq:phi}
393: \end{equation}
394: and  approximates $\phi(\tau)$ by a simple power law
395: $\phi(\tau) \sim \tau^{\alpha}$. Taking the first four complex zeros,
396: one obtains
397: \begin{equation}
398:  \alpha = \frac{ {\rm ln}\,\phi(\tau_3) - {\rm ln}\,\phi(\tau_2)}
399:                { {\rm ln}\,\tau_3 - {\rm ln}\,\tau_2} \, .    \label{alpha}
400: \end{equation}
401: With a second parameter $\gamma$, related to the crossing angle of this
402: line with the real axis,  
403: \begin{equation}
404:   \gamma = 
405:   [{\rm Re}(\beta_2) -{\rm Re}(\beta_1)]/(\tau_2 - \tau_1)\,, \label{gamma}
406: \end{equation}
407: and following the classification scheme by Grossmann
408: and Rosenhauer \cite{GR1,GR2},
409: phase transitions can now be classified according to the values of these
410: two parameters:
411: for $\alpha \leq 0$ and $\gamma=0$ one has a phase transition of
412: first order, it is of second order if $0 < \alpha < 1$ and
413: %$\gamma=0$ (or $\gamma \neq 0$)
414: arbitrary $\gamma$, and for $\alpha > 1$
415: and arbitrary $\gamma$ one has a higher order transition. We have
416: evaluated the usefulness of this approach both for spin systems and
417: polyalanine chains \cite{AFH01d,AHP01e}. Preliminary results for
418:  Ala$_{10}$-Gly$_5$-Ala$_{10}$ are also listed in Ref.~\cite{AHP01e}.
419: 
420: For Ala$_{10}$-Gly$_5$-Ala$_{10}$, we  find  two lines of complex zeros.
421: The corresponding first four zeros for each characteristic line are 
422: listed in Table 2. Our error estimate is based on the jackknife 
423: method \cite{jack} with 16 bins. These lines lead to  two critical 
424: temperatures $T_{hc}=480$ K  and $T_f=271$ K (estimated from the real
425: part of $\beta$) that agree  with 
426: the corresponding values $T_{hc}=483 \pm 8$ K and $T_f=265 \pm 7$ K
427: found  by us above with different methods.
428: 
429: 
430: Using these zeros  we  have calculated
431: the parameters $\alpha$ and $\gamma$ that characterize
432: in the Borrmann {\it et al.} approach phase transitions in small systems.
433: For the first transition, at $T=480$ K,
434: we find $\alpha= 1.1(1.5)$ and $\gamma=-0.4(2)$. The  errors 
435: reflect  large fluctuations in the values of the two parameters
436: $\alpha$ and $\gamma$ that do not allow us  to determine whether
437: the helix-coil transition  is a weak first  order or a strong
438: second order phase transition.
439: This problem was also observed in our earlier work on 
440: polyalanine \cite{AH99b,AFH01d} where we were also 
441: not able to establish clearly the order of the helix-coil transition.
442: However, our results illustrate the strength of this transition
443: that also leads to the pronounced peak in the specific heat observed 
444: in Fig.~3, and suggest a nucleation mechanism for $\alpha$-helix 
445: formation.  Our data are more decisive in the
446: case of the second transition, at $T=265$ K, which marks the
447: collapse and folding of the peptide. Here we find $\alpha=0.32(8)$
448: and $\gamma=0.36(2)$. These values indicate  a second-order transition
449: which is consistent with what one would expect for a transition between
450: extended and compact structures and imply that collapse and folding
451: of the Ala$_{10}$-Gly$_5$-Ala$_{10}$ is connected with long range 
452: correlations between the residues. 
453: 
454: Our above analysis of the thermodynamics of our peptide suggests that
455: Ala$_{10}$-Gly$_5$-Ala$_{10}$ folds in a 2 step process. The first step
456: is the formation of $\alpha$-helices and can be characterized by a  
457: helix-coil transition temperature $T_{hc} = 483\pm 8$ K. The formation
458: of $\alpha$-helices then restricts the possible configuration space. 
459: Energetically most favorable is the folding of two $\alpha$-helices  
460: (made out of the alanine residues)  into a hairpin.   This second 
461: step can be characterized by a lower folding temperature
462:  $T_f = 265\pm 7$ K. Note that this folding temperature is in the  
463: biological relevant temperature regime while helix-formation can 
464: also happen at much higher temperatures. The above described two step 
465: folding of our artificial peptide is reminiscent of the well known 
466: framework \cite{Ptitsyn,KB} and collision-diffusion model \cite{KW} 
467: of folding which also propose that local elements of native local 
468: secondary structure form independently of tertiary structure. These 
469: elements diffuse until they collide and coalesce to give a tertiary 
470: structure. In our case, the temperature region of $265 - 480$ K is the 
471: one where the thermal energy of the molecule does not allow 
472: coalescing of the helix-fragments that therefore form and decay. 
473: Some stabilization happens when these fragments try to form one 
474: extended helix, however, the inherent flexibility of the glycine 
475: residues, connecting the two alanine chains,  and the gain in 
476: Lennard-Jones energy lead instead at temperatures below $T_f$
477: to  a U-turn-like bundle of two (antiparallel) $\alpha$-helices 
478: connected by a turn of glycine residues as the most stable structure. 
479: Note that this picture is consistent with energy landscape theory 
480: and funnel concept \cite{Bryngelson87,Onuchic97}.  Fig.~5b depicts 
481: the appearance of a folding funnel at $T=270$ K towards our
482: ``native structure'' `A'. The competing structure `B', that at
483: this temperature has a slightly higher  ($\approx 1 k_BT$) free 
484: energy ( see the 1 $k_BT$ contour line for structure `A' that is 
485: missing for conformer `B'), acts as a local trap. However, the free 
486: energy barriers of $\approx  6~k_BT$ can be  overcome at this
487: temperature by thermal fluctuations. Below that temperature the 
488: relative weight of structure `B' decreases (see Fig.7) and its
489: free energy difference to `A' increases: the energy landscape 
490: becomes even more funnel like (data not shown). The energy landscape 
491: of Fig.~5b allows for a multitude of folding pathways that all, however, 
492:  follow the above described two-step process.
493: 
494: An interesting question is how general the above obtained results  are.
495: A direct comparison with experimental data is difficult since
496: solvent effects were neglected in the simulation of 
497:  Ala$_{10}$-Gly$_5$-Ala$_{10}$ and most experiments study
498: solvated peptides. An exception are the techniques developed by 
499: Jarrold and collaborators for examination of gas-phase conformations 
500: of proteins and peptides \cite{jarrold}.  An experimental study
501: of  Ala$_{10}$-Gly$_5$-Ala$_{10}$ using these techniques is now planed.
502: The authors are not aware of experimental results for the solvated 
503: peptide. In order to compare our  work with experiments of 
504: other short helical peptides \cite{Brooks,Oas}, we therefore started 
505: now simulations of Ala$_{10}$-Gly$_5$-Ala$_{10}$ where the
506: solvation effects are approximated  by a solvent accessible surface
507: term \cite{inplan}. This will allows us also to test the dependency
508: of our results on the solvation model. Simulating another, slightly 
509: more complicated, artifical peptide, 
510: Ala$_{10}$-Gly$_5$-Ala$_{10}$-Gly$_5$-Ala$_{10}$  
511: that presumably will fold in a three-helix bundle, will  allow 
512: in addition a direct comparison with recent experimental work by 
513: Myers and Oas \cite{Oas}
514: where the relation between helix-formation and folding was studied 
515: for the 58-residue B domain of protein A. 
516: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
517:  
518: 
519: \section{conclusion}
520: In summary, we have performed multicanonical simulations with high
521: statistics of a simple artificial peptide, the 25 residue
522:  Ala$_{10}$-Gly$_5$-Ala$_{10}$. We found that this peptide 
523: folds into a specific  structure that is determined solely by
524: the intrinsic properties of the molecule (since solvent interactions
525: are absent in our simulations. In gas-phase, the peptide exhibits
526: two characteristic transitions. At $T_{hc}=483\pm 8$ K 
527: we observed a  helix-coil transition that is either a weak first 
528: order transition or a strong second order transition. Our results 
529: indicate that there is a second transition at $T_f=265\pm 7$ K, 
530: the folding transition, that is second order-like. These results 
531: suggest that folding of this peptide in gas-phase is a two-step 
532: process. In a first step, the alanine residues form independently 
533: helical segments which then afterwards in a second step assemble to 
534: a U-turn like structure of two antiparallel $\alpha$-helices
535: connected by a turn.  By using an implicit solvent model in
536: our simulations we started now to investigate whether the 
537: final structure or this two-step process  changes in the present of
538: water. 
539: 
540: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
541:  
542: \noindent
543: {\bf Acknowledgements}: \\
544:  U. Hansmann gratefully acknowledges support by  a research grant
545: from the National Science Foundation (CHE-9981874), and
546: N.A. Alves support by CNPq (Brazil).
547: 
548: %\end{multicols}
549: 
550: 
551: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
552:  
553: %\begin{thebibliography}{99}
554: \begin{references}
555: \bibitem{Anf} C.B. Anfinsen, {\it Science} {\bf 181}, 223 (1973).
556: \bibitem{Bryngelson87} J.D.~Bryngelson and P.G.~Wolynes, 
557:                        {\it Proc.~Natl.~Acad.~Sci.~(USA)} {\bf  84},
558:                        7524 (1987).
559: \bibitem{Onuchic97} J.N.~Onuchic, Z.~Luhey-Schulten and  P.G.~Wolynes, 
560:                     {\it Annu. Rev. Phys. Chem.} {\bf 48},  545 (1997).
561: \bibitem{HOO98b} U.H.E.~Hansmann, Y.~Okamoto and J.N.~Onuchic,
562:                 {\it Proteins} {\bf 34}, 472 (1999).
563: \bibitem{HO01b} U.H.E.~Hansmann and J.N.~Onuchic, {\it J. Chem. Phy.}
564:                 {\bf 115}, 1601 (2001).
565: \bibitem{Poland} D.~Poland and H.A.~Scheraga, {\it Theory of Helix-Coil
566:          Transitions in Biopolymers} (Academic Press, New York, 1970).
567: \bibitem{Jeff} J.P.~Kemp and Z.Y. Chen,  {\it Phys.~Rev.~Lett.}
568:               {\bf 81}, 3880 (1998).
569: \bibitem{HO98c} U.H.E.~Hansmann and Y.~Okamoto, {\it J. Chem.~Phys.} {\bf 110},
570:                 1267 (1999); {\bf 111} (1999) 1339(E).
571: \bibitem{AH99b} N.A. Alves  and U.H.E. Hansmann,\
572:                 {\it Phys. Rev. Lett.}  {\bf 84}, 1836 (2000).
573: \bibitem{AH00b} N.A. Alves and U.H.E.~Hansmann, {\it Physica A}
574:                 {\bf 292}, 509 (2001).
575: \bibitem{MO2}  A. Mitsutake and Y. Okamoto, {\it J. Chem. Phys.} {\bf 112},
576:                10638 (2000).
577: \bibitem{PH01f} Y.~Peng and U.H.E.~Hansmann, {\it Solvation model dependency
578:                of helix-coil transition in polyalanine}, {\it Biophysical
579:                J.} (2002), in press.
580: \bibitem{MU} B.A. Berg  and T. Neuhaus,
581:              {\it Phys. Lett.} B {\bf 267}, 249 (1991).
582: \bibitem{BMH} P. Borrmann, O. M\"ulken and J. Harting,
583:               Phys. Rev. Lett. 84 (2000)3511;
584:               O. M\"ulken, P. Borrmann, J. Harting and H. Stamerjohanns,
585:               Phys. Rev. A64 (2001) 013611;
586:               O. M\"ulken and P. Borrmann, Phys. Rev. C63 (2001) 024306.
587: \bibitem{SKY} K.R.~Shoemaker, P.S.~Kim, E.J.~York, J.M.~Stewart, 
588:               R.L.~Baldwin, {\it Nature} {\bf 326}, 563 (1987).
589: \bibitem{EC} M.J. Sippl, G. N{\'e}methy, and H.A. Scheraga,
590:              {\it J. Phys. Chem.} {\bf 88}, 6231~(1984), 
591:              and references therein.
592: \bibitem{SMMP} F.~Eisenmenger, U.H.E.~Hansmann, Sh.~Hayryan, C.-K.~Hu,
593:                {\it Comp.~Phys.~Comm.} {\bf 138}, 192 (2001).
594: \bibitem{curr_op}  U.H.E.~Hansmann and Y.~Okamoto,
595:         {\it Curr.~Opin.~Struc.~Biol.} {\bf 9}, 177 (1999).
596: \bibitem{HO} U.H.E. Hansmann and Y. Okamoto,  J.~Comp.~Chem.
597:   {\bf 14}, 1333 (1993).
598: \bibitem{OurReview98} U.H.E.~Hansmann and Y.~Okamoto, 
599:          in: Stauffer, D. (ed.) ``{\it Annual Reviews in Computational 
600:          Physics VI}''
601:          (Singapore: World Scientific), p.129.  (1998).
602: \bibitem{FS} A.M. Ferrenberg and R.H. Swendsen, {\it Phys.\ Rev.\ Lett.}
603:              {\bf  61}, 2635 (1988); {\it Phys. Rev. Lett.} {\bf 63},
604:              1658(E) (1989), and
605:               references given in the erratum.
606: \bibitem{Berg}  B.A. Berg, {\it J. Stat. Phys.} {\bf 82}, 323 (1996).
607: \bibitem{OH95b} Y.~Okamoto and U.H.E.~Hansmann,\  J.~Phys.~Chem.
608:                 {\bf 99}, 11276 (1995).
609: \bibitem{GR1} S. Grossmann and W. Rosenhauer, Z. Physik 207 (1967)138.
610: \bibitem{GR2} S. Grossmann and W. Rosenhauer, Z. Physik 218 (1969)437;
611:               S. Grossmann and W. Rosenhauer, Z. Physik 218 (1969)449.
612: \bibitem{AFH01d} N.A. Alves, J.P.N. Ferrite and U.H.E. Hansmann,
613:                 {\it Phys. Rev. E} {\bf 65}, 036110 (2002).
614: \bibitem{AHP01e}  N.A. Alves, U.H.E. Hansmann and Y. Peng,
615:                 {\it Int. J. Mol. Sci.} {\bf 3}, 17 (2002).
616: \bibitem{jack} R.G. Miller, {\it Biometrika} {\bf 61}, 1 (1974).
617: \bibitem{Ptitsyn} O.B. Ptitsyn, {\it Protein Eng.} {\bf 7}, 593 (1994).
618: \bibitem{KB} P.D. Kim and R.L. Baldwin, {\it Annu. Rev. Biochem.} {\bf 59},
619:              631 (1990).
620: \bibitem{KW} M. Karplus and D.L. Weaver, {\it Protein Sci.} {\bf 3},
621:              650 (1994). 
622: \bibitem{jarrold} R.R. Hudgins, M.A. Ratner and M.F. Jarrold,
623:                   {\it J. Am. Chem. Soc.} {\bf 120} 12974 (1998).
624: \bibitem{Brooks} E.M.~Boczko and C.L.~Brooks  III,
625:                  {\it Science} {\bf 269},  393 (1995).
626: \bibitem{Oas} J.K. Myers and T.G. Oas, 
627:               {\it Nat. Str. Bio.} {\bf 8} 552 (2001).
628: \bibitem{inplan} U.H.E. Hansmann, in preparation.
629: \end{references}
630: %\end{thebibliography}
631: \vfil
632: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
633: \newpage
634: %{\huge Tables:}
635: \begin{table}[!h]
636: \renewcommand{\tablename}{Table}
637: \caption{Dihedral angles for the two configurations shown in Fig.~5}
638: \begin{center}
639: {\small
640: \begin{tabular}{|c|ccc|ccc|} 
641: State &\multicolumn{3}{c}{Conformer {\it A}} &
642:         \multicolumn{3}{c}{Conformer {\it B}}\\ \hline
643: $E_{tot}$ (Kcal/mol) & \multicolumn{3}{c}{-34.7} & \multicolumn{3}{c}{-23.9}\\
644: \hline
645: Residue & $\phi$ & $\psi$ & $\chi$ & $\phi$ & $\psi$ & $\chi$ \\
646: \hline
647:  Ala-1& -101.7 &  154.4 & -178.1 &  -55.2 &   -54.9 &    61.5\\
648:  Ala-2&  -65.4 &  -35.5 & -173.8 &  -70.6 &   -37.8 &    49.2\\
649:  Ala-3&  -68.2 &  -36.1 &  -61.7 &  -60.4 &   -30.8 &  -178.3\\
650:  Ala-4&  -70.6 &  -39.0 &  -69.7 &  -78.6 &   -37.6 &   -43.4\\
651:  Ala-5&  -64.3 &  -40.6 &  -54.1 &  -66.5 &   -28.8 &   176.5\\
652:  Ala-6&  -66.3 &  -38.3 &  175.1 &  -70.6 &   -50.4 &   -55.8\\
653:  Ala-7&  -69.5 &  -31.9 & -172.6 &  -64.2 &   -32.6 &  -178.6\\
654:  Ala-8&  -75.5 &  -31.6 &   57.2 &  -68.6 &   -43.4 &   -58.3\\
655:  Ala-9&  -62.6 &  -45.2 & -172.9 &  -67.8 &   -41.0 &  -164.8\\
656:  Ala-10&  -73.4 &  -54.4 &  -59.0 &  -60.8 &   -43.3 &    53.4\\
657:  Gly-11&  -91.9 &   65.0 &        &  -64.3 &   -40.5 &        \\
658:  Gly-12&  156.2 &  -80.6 &        &  -64.9 &   -44.9 &        \\
659:  Gly-13&  151.5 & -176.6 &        &  -64.6 &   -47.0 &        \\
660:  Gly-14&  -60.1 &  -35.5 &        &  -63.0 &   -33.0 &        \\
661:  Gly-15&  -63.5 &  -38.0 &        &  -72.6 &   -39.9 &        \\
662:  Ala-16&  -72.6 &  -34.6 &  -48.7 &  -70.3 &   -31.1 &    83.1\\
663:  Ala-17&  -69.4 &  -32.9 &  -52.2 &  -69.2 &   -39.2 &    63.3\\
664:  Ala-18&  -70.2 &  -38.4 &  -53.6 &  -68.4 &   -39.9 &    63.8\\
665:  Ala-19&  -70.5 &  -35.1 &  174.4 &  -66.1 &   -38.7 &   -58.5\\
666:  Ala-20&  -66.6 &  -40.4 &  179.5 &  -67.1 &   -41.9 &  -170.9\\
667:  Ala-21&  -68.7 &  -40.3 &  173.3 &  -66.0 &   -33.3 &    66.1\\
668:  Ala-22&  -61.2 &  -36.0 &   61.7 &  -73.9 &   -35.5 &    61.9\\
669:  Ala-23&  -71.1 &  -52.6 &  -51.3 &  -65.2 &   -37.8 &  -176.5\\
670:  Ala-24& -153.3 &  111.8 &   58.6 &  -73.0 &   -39.3 &   -39.5\\
671:  Ala-25&  -62.8 &  -62.6 & -173.5 &  -69.5 &    -8.8 &    71.0\\
672: \end{tabular}
673: }
674: \end{center}
675: \end{table}
676: 
677: \begin{table}[!h]
678: \renewcommand{\tablename}{Table}
679: \caption{\baselineskip=0.8cm Partition function zeros for
680: the two transitions observed for Ala$_{10}$-Gly$_5$-Ala$_{10}$.}
681: \begin{center}
682: \begin{tabular}{cccccccc}\\
683: %%%\hline
684: \\[-0.3cm]
685: Re$(\beta_1)$ &$\tau_1$  &Re$(\beta_2)$ & $\tau_2$ &
686: Re$(\beta_3)$ &$\tau_3$  &Re$(\beta_4)$ & $\tau_4$        \\
687: \\[-0.35cm]
688: \hline
689: \\[-0.3cm]
690:  1.0463(46)   & 0.1307(53) & 1.0144(99) & 0.2112(55) &
691:  1.051(44)    & 0.310(36)  & 1.055(46)  & 0.347(41) \\
692:  1.855(21)    & 0.263(12)  & 1.991(25)  & 0.637(41) &
693:  2.057(53)    & 0.923(43)  & 2.070(43)  & 1.229(55) \\
694: %%%\hline\\
695: \end{tabular}
696: \end{center}
697: \label{table 1}
698: \end{table}
699: 
700: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
701: \clearpage 
702: \newpage
703: {\huge Figure Captions:} \\
704: \begin{description}
705: \item[Fig.~1] The average helicity $<n_H>(T)$ of Ala$_{10}$-Gly$_5$-Ala$_{10}$ 
706:               as a function of temperature $T$. The corresponding values
707:               of the susceptibility $\chi (T)$ are ploted in the inset.
708:               Our data rely on a single
709:               multicanonical simulation of 8,000,000 Monte Carlo sweeps.
710: \item[Fig.~2] The average ECEPP/2 energy $<E_{tot}>$ and the corresponding
711:               partial energy terms, Coulomb energy $<E_C>$, Lennard-Jones
712:               term $<E_{LJ}>$, hydrogen-bond energy $<E_{hb}>$ and
713:               torsion energy $<E_{tor}>$, as a function of temperature
714:               $T$ for Ala$_{10}$-Gly$_5$-Ala$_{10}$.
715: \item[Fig.~3]  The specific heat $C(T)$ of  Ala$_{10}$-Gly$_5$-Ala$_{10}$ 
716:               as a function of temperature $T$.
717: \item[Fig.~4] The average end-to-end distance $<d_{e-e}>(T)$ as a function
718:               of temperature $T$ for  Ala$_{10}$-Gly$_5$-Ala$_{10}$.
719: \item[Fig.~5] Free energy landscape of Ala$_{10}$-Gly$_5$-Ala$_{10}$  as
720:               a function of helicity $n_H$ and end-to-end distance
721:               $d_{e-e}$ at temperature (a) $T_{hc}=480$ K  and 
722:               (b) $T_f =270$ K. The contour lines are drawn in
723:               multiples of $k_BT$ indicated in the key.
724: \item[Fig.~6] Lowest-energy conformation `A' of Ala$_{10}$-Gly$_5$-Ala$_{10}$
725:               (a) and the  conformation `B' (which has the largest helicity) 
726:               (b) as found in our multicanonical simulation of 8,000,000 
727:               Monte Carlo sweeps.
728: \item[Fig.~7] Relative weight  of the lowest-energy conformation `A' and
729:               conformation `B' (which has maximal helicity) as a function
730:               of temperature.
731: \end{description}
732: 
733: 
734: 
735: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
736: 
737: %\end{document}
738: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
739: 
740: 
741: %FIGURE 1
742: \begin{figure}[b]
743: %\begin{figure}[!ht]
744: \begin{center}
745: \begin{minipage}[t]{0.95\textwidth}
746: \centering
747: \includegraphics[angle=90,width=0.72\textwidth]{Fig1.eps}
748: \renewcommand{\figurename}{(Fig.1)}
749: \caption{The average helicity $<n_H>(T)$ of Ala$_{10}$-Gly$_5$-Ala$_{10}$
750:               as a function of temperature $T$. The corresponding values
751:               of the susceptibility $\chi (T)$ are ploted in the inset.
752:               Our data rely on a single
753:               multicanonical simulation of 8,000,000 Monte Carlo sweeps.}
754: \label{Fig. 1}
755: \end{minipage}
756: \end{center}
757: \end{figure}
758: 
759: 
760: 
761: \newpage
762: \cleardoublepage
763: 
764: 
765: 
766: %FIGURE 2
767: \begin{figure}[b]
768: %\begin{figure}[!ht]
769: \begin{center}
770: \begin{minipage}[t]{0.95\textwidth}
771: \centering
772: \includegraphics[angle=-90,width=0.72\textwidth]{Fig2.eps}
773: \renewcommand{\figurename}{(Fig.2)}
774: \caption{ The average ECEPP/2 energy $<E_{tot}>$ and the corresponding
775:               partial energy terms, Coulomb energy $<E_C>$, Lennard-Jones
776:               term $<E_{LJ}>$, hydrogen-bond energy $<E_{HB}>$ and
777:               torsion energy $<E_{tor}>$, as a function of temperature
778:               $T$ for Ala$_{10}$-Gly$_5$-Ala$_{10}$. }
779: \label{Fig. 2}
780: \end{minipage}
781: \end{center}
782: \end{figure}
783: 
784: 
785: %FIGURE 3
786: \begin{figure}[b]
787: %\begin{figure}[!ht]
788: \begin{center}
789: \begin{minipage}[t]{0.95\textwidth}
790: \centering
791: \includegraphics[angle=-90,width=0.72\textwidth]{Fig3.eps}
792: \renewcommand{\figurename}{(Fig.3)}
793: \caption{ The specific heat $C(T)$ of  Ala$_{10}$-Gly$_5$-Ala$_{10}$ 
794:               as a function of temperature $T$. }
795: \label{Fig. 3}
796: \end{minipage}
797: \end{center}
798: \end{figure}
799: 
800: 
801: \newpage
802: \cleardoublepage
803: 
804: 
805: %FIGURE 4
806: \begin{figure}[b]
807: %\begin{figure}[!ht]
808: \begin{center}
809: \begin{minipage}[t]{0.95\textwidth}
810: \centering
811: \includegraphics[angle=-90,width=0.72\textwidth]{Fig4.eps}
812: \renewcommand{\figurename}{(Fig.4)}
813: \caption{The average end-to-end distance $<d_{e-e}>(T)$ as a function
814:               of temperature $T$ for  Ala$_{10}$-Gly$_5$-Ala$_{10}$.}
815: \label{Fig. 4}
816: \end{minipage}
817: \end{center}
818: \end{figure}
819: 
820: 
821: \newpage
822: \cleardoublepage
823: 
824: 
825: %FIGURE 5a
826: \begin{figure}[b]
827: %\begin{figure}[!ht]
828: \begin{center}
829: \begin{minipage}[t]{0.95\textwidth}
830: \centering
831: \includegraphics[angle=-90,width=0.72\textwidth]{Fig5aa.eps}
832: \renewcommand{\figurename}{(Fig.5a)}
833: \caption{ Free energy landscape of Ala$_{10}$-Gly$_5$-Ala$_{10}$  as
834:               a function of helicity $n_H$ and end-to-end distance
835:               $d_{e-e}$ at temperature (a) $T_{hc}=480$ K  and 
836:               (b) $T_f =270$ K.}
837: \label{Fig. 5a}
838: \end{minipage}
839: \end{center}
840: \end{figure}
841: 
842: 
843: %FIGURE 5b
844: \begin{figure}[b]
845: %\begin{figure}[!ht]
846: \begin{center}
847: \begin{minipage}[t]{0.95\textwidth}
848: \centering
849: \includegraphics[angle=-90,width=0.72\textwidth]{Fig5bb.eps}
850: \renewcommand{\figurename}{(Fig.5b)}
851: \caption{ }
852: \label{Fig. 5b}
853: \end{minipage}
854: \end{center}
855: \end{figure}
856: 
857: 
858: 
859: \newpage
860: \cleardoublepage
861: 
862: 
863: %FIGURE 6a
864: \begin{figure}[b]
865: %\begin{figure}[!ht]
866: \begin{center}
867: \begin{minipage}[t]{0.95\textwidth}
868: \centering
869: \includegraphics[angle=-90,width=0.72\textwidth]{Fig6a.eps}
870: \renewcommand{\figurename}{(Fig.6a)}
871: \caption{ Lowest-energy conformation `A' of Ala$_{10}$-Gly$_5$-Ala$_{10}$
872:               (a) and the  conformation `B' (which has the largest helicity) 
873:               (b) as found in our multicanonical simulation of 8,000,000 
874:               Monte Carlo sweeps.}
875: \label{Fig. 6a}
876: \end{minipage}
877: \end{center}
878: \end{figure}
879: 
880: 
881: \newpage
882: \cleardoublepage
883: 
884: 
885: 
886: %FIGURE 6b
887: \begin{figure}[b]
888: %\begin{figure}[!ht]
889: \begin{center}
890: \begin{minipage}[t]{0.95\textwidth}
891: \centering
892: \includegraphics[angle=-90,width=0.72\textwidth]{Fig6b.eps}
893: \renewcommand{\figurename}{(Fig.6b)}
894: \caption{ }
895: \label{Fig. 6b}
896: \end{minipage}
897: \end{center}
898: \end{figure}
899: 
900: 
901: \newpage
902: \cleardoublepage
903: 
904: 
905: %FIGURE 7
906: \begin{figure}[b]
907: %\begin{figure}[!ht]
908: \begin{center}
909: \begin{minipage}[t]{0.95\textwidth}
910: \centering
911: \includegraphics[angle=-90,width=0.72\textwidth]{Fig7.eps}
912: \renewcommand{\figurename}{(Fig.7)}
913: \caption{ Relative weight  of the lowest-energy conformation `A' and
914:               conformation `B' (which has maximal helicity) as a function
915:               of temperature. }
916: \label{Fig. 7}
917: \end{minipage}
918: \end{center}
919: \end{figure}
920: 
921: %%%%%%%%%%%%%%%%%%%%%%%%%%%
922: \end{document}
923: %%%%%%%%%%%%%%%%%%%%%%%%%%%%
924: 
925: 
926: 
927: 
928: 
929: 
930: 
931: 
932: 
933: