1: \documentclass{kluwer}
2: %\documentclass[pdftex]{kluwer}
3: \usepackage{graphicx}
4: %
5: % math defs
6: \newcommand{\EQ}{\begin{equation}}
7: \newcommand{\EN}{\end{equation}}
8: \newcommand{\eq}[1]{(\ref{#1})}
9: \newcommand{\EEq}[1]{Equation~(\ref{#1})}
10: \newcommand{\Eq}[1]{Eq.~(\ref{#1})}
11: \newcommand{\Eqs}[2]{Eqs.~(\ref{#1}) and~(\ref{#2})}
12: \newcommand{\eqs}[2]{(\ref{#1}) and~(\ref{#2})}
13: \newcommand{\Eqss}[2]{Eqs.~(\ref{#1})--(\ref{#2})}
14: \newcommand{\eqss}[2]{(\ref{#1})--(\ref{#2})}
15: \newcommand{\Sec}[1]{Section~\ref{#1}}
16: \newcommand{\Fig}[1]{Figure~\ref{#1}}
17: \newcommand{\FFig}[1]{Figure~\ref{#1}}
18: \newcommand{\Tab}[1]{Table~\ref{#1}}
19: \newcommand{\TTab}[1]{Table~\ref{#1}}
20: \newcommand{\Figs}[2]{Figs~\ref{#1} and \ref{#2}}
21: \newcommand{\ea}{{\em et al., }}
22: \def\half{{\textstyle{1\over2}}}
23: \def\quarter{{\textstyle{1\over4}}}
24: \newcommand{\dd}{{\rm d} {}}
25: %
26: % journals
27: \newcommand{\ysci}[5]{: #1, #5, {\em Science }{\bf #2}, #3--#4.}
28: \newcommand{\ynat}[5]{: #1, #5, {\em Nature }{\bf #2}, #3--#4.}
29: \newcommand{\yprl}[5]{: #1, #5, {\em Phys.\ Rev.\ Lett. }{\bf #2}, #3--#4.}
30: \newcommand{\yphy}[5]{: #1, #5, {\em Physica }{\bf #2}, #3--#4.}
31: \newcommand{\yphl}[5]{: #1, #5, {\em Phys.\ Lett. }{\bf #2}, #3--#4.}
32: \newcommand{\yoleb}[5]{: #1, #5, {\em Orig.\ Life Evol.\ Biosph. }{\bf #2}, #3--#4.}
33: \newcommand{\yjour}[6]{: #1, #6, {\em #2} {\bf #3}, #4--#5.}
34: \newcommand{\ybook}[3]{: #1, {\em #2}, #3.}
35:
36: \begin{document}
37: \begin{article}
38: \begin{opening}
39: \title{Homochiral growth through enantiomeric cross-inhibition}
40:
41: \author{A. \surname{Brandenburg}$^*$}
42: \author{A.~C. \surname{Andersen}}
43: \author{S. \surname{H\"ofner}}
44: \author{M. \surname{Nilsson}}
45: \institute{Nordita, Blegdamsvej 17, DK-2100 Copenhagen \O, Denmark\\
46: ($*$ author for correspondence, e-mail: brandenb@nordita.dk,
47: phone +45 353 25228, fax: +45 353 89157)}
48:
49: \runningtitle{Homochiral growth through enantiomeric cross-inhibition}
50: \runningauthor{Brandenburg et al.}
51:
52: \begin{abstract}
53: The stability and conservation properties of a recently
54: proposed polymerization model are studied.
55: The achiral (racemic) solution is linearly unstable once the
56: relevant control parameter (here the fidelity of the catalyst)
57: exceeds a critical value.
58: The growth rate is calculated for different fidelity parameters and
59: cross-inhibition rates.
60: A chirality parameter is defined and shown to
61: be conserved by the nonlinear terms of the model.
62: Finally, a truncated version of the model
63: is used to derive a set of two ordinary
64: differential equations and it is argued that these equations are more
65: realistic than those used in earlier models of that form.
66: \end{abstract}
67:
68: \keywords{DNA polymerization, enantiomeric cross-inhibition,
69: origin of homochirality. $ $Revision: 1.63 $ $}
70:
71: \end{opening}
72:
73: \section{Introduction}
74:
75: The chirality of molecules in living organisms must have been fixed at
76: an early stage in the development of life.
77: All life that we know is based on RNA and DNA molecules with
78: dextrarotatory sugars.
79: There is growing evidence that the RNA world
80: (Woese, 1967; Crick, 1968; Orgel, 1968; see also Wattis \& Coveney 1999)
81: must have been preceded by a simpler pre-RNA world made up of
82: achiral constituents (Bada, 1995, Nelson \ea 2000).
83: An alternative carrier of genetic code are peptide nucleic acids or PNA
84: (Nielsen, 1993).
85: These can be rather simple and are currently discussed in connection
86: with the idea to build artificial life (Rasmussen \ea 2003).
87: Furthermore, although PNA can still be chiral (Tedeschi \ea 2002),
88: there are also forms of PNA that are achiral (Pooga \ea 2001),
89: suggesting that chirality may have
90: developed later when the first RNA molecules formed.
91:
92: In current proposals to build artificial life, chirality does not
93: seem to be crucial.
94: The PNA molecules is proposed to act primarily as charge carrier,
95: i.e.\ a very primitive functionality compared to the genetic code
96: in contemporary cells (Rasmussen \ea 2003).
97: At this stage, homochirality may have been introduced by chance.
98: This is also supported by the fact that
99: chiral polymers of the same chirality tend to have a more stable
100: structure (Pogodina \ea 2001) and would therefore be
101: genetically preferred.
102:
103: Since the introduction of chiral molecules is assumed to take place
104: at a stage when there is already growth and self-replication, it is
105: also plausible to assume that the existence of chiral molecules has
106: an autocatalytic effect in producing new chiral molecules of the
107: same chirality (Kondepudi \ea 1990).
108: This is the basis of the recently proposed polymerization model of
109: Sandars (2003); see also Wattis \& Coveney (2004).
110: The purpose of the present paper is to reconsider this model
111: (or a slightly modified version of it) and to analyze its
112: stability behavior and conservation properties.
113: We also discuss and illustrate some of the
114: salient features of the model in more detail.
115: The model is then compared with earlier models of homochirality where
116: the detailed polymerization process is ignored and the dynamics of single
117: variables representing left and right handed polymers are modeled instead
118: (Frank, 1953; Kondepudi and Nelson, 1984; Goldanskii and Kuzmin 1989;
119: Avetisov and Goldanskii, 1993; Saito and Hyuga, 2004).
120:
121: In order to appreciate the nature of the many terms in the model of Sandars we
122: begin by discussing first the basic principle of the model in connection
123: with homochiral polymer growth and then turn to the full set of reactions
124: that are included in the model.
125:
126: \section{Homochiral polymer growth}
127: \label{HomochiralGrowth}
128:
129: In this section we discuss the growth of polymers by adding monomers
130: of the same chirality, i.e.\ we ignore reactions with monomers of the
131: opposite chirality.
132: This is conceptually the simplest case, but its equilibrium solution
133: also corresponds to a solution of the full system discussed below.
134: We write down the equations for left-handed polymers, but the same
135: applies also to right-handed polymers.
136:
137: A left-handed polymer of length $n$ is assumed to react with
138: a left-handed monomer via the reaction
139: \begin{equation}
140: L_{n}+L_1\rightarrow L_{n+1}\quad\mbox{($n\ge1$)}.
141: \end{equation}
142: The reaction rate is $k_S$, but since $L_{n}$ can bind to $L_1$
143: on either side, the total reaction rate is $2k_S$ and proportional
144: to the product of the concentrations of the two constituents.
145: We denote the concentration of $L_n$ chains by $[L_n]$, so
146: in a volume $V$ the number of $L_n$ chains is $N_n\equiv[L_n]V$.
147: For $n\geq3$ the number of possible pairs of $L_{n-1}$ and $L_1$ is
148: $N_n\times N_1$.
149: A special situation arises for $n=2$, because then $L_1$ is interacting
150: with another $L_1$, and the number of possible pairs is only
151: $\half N_1(N_1-1)\approx\half N_1^2$.
152: [This problem is familiar from the physics of nuclear reactions; see, e.g.,
153: Kippenhahn and Weigert (1990).]
154: We therefore introduce the factor $\sigma_n^{(1/2)}$ defined by
155: \begin{equation}
156: \sigma_n^{(\alpha)}=\left\{
157: \begin{array}{cc}
158: \alpha & \quad\mbox{for}\quad n=2,\\
159: 1 & \quad\mbox{for}\quad n\ge3.
160: \end{array}
161: \right.
162: \label{sigman}
163: \end{equation}
164: (Later we shall use this factor also with $\alpha=0$ instead of $1/2$.)
165: The corresponding contribution to the evolution of the concentration of
166: $L_n$ is therefore
167: \begin{equation}
168: {\dd[L_{n+1}]\over\dd t}=...+2k_S\sigma_n^{(1/2)}[L_{n}][L_1],
169: \end{equation}
170: where the dots denote the presence of other terms that will be discussed
171: later.
172:
173: Obviously, the concentrations of $L_{n}$ and $L_1$ have to decrease
174: at the same rate by the same amount, so
175: \begin{equation}
176: {\dd[L_{n}]\over\dd t}=...-2k_S\sigma_n^{(1/2)}[L_{n}][L_1],
177: \label{Lna}
178: \end{equation}
179: \begin{equation}
180: {\dd[L_1]\over\dd t}=...-2k_S\sigma_n^{(1/2)}[L_{n-1}][L_1].
181: \label{L1a}
182: \end{equation}
183: In the following we regard
184: $n$ as a general index with $2\leq n\leq N$, the evolution
185: of $[L_n]$ is governed by the difference of two terms (gain from
186: $L_{n-1}$ chains and loss in favor of producing $L_{n+1}$ chains).
187: The production of each $L_n$ contributes to a loss of $L_1$ monomers,
188: so the right hand side of \Eq{L1a} becomes a sum over all $n$.
189: The full set of equations is then
190: \begin{equation}
191: {\dd[L_1]\over\dd t}=Q_L-\lambda_L[L_1],
192: \quad\mbox{where}\quad\lambda_L=2k_S\sum_{n=1}^{N-1}[L_{n}],
193: \label{L1b}
194: \end{equation}
195: \begin{equation}
196: {\dd[L_{n}]\over\dd t}=2k_S[L_1]\Big(\sigma_n^{(1/2)}[L_{n-1}]-[L_{n}]\Big),
197: \label{Lnb}
198: \end{equation}
199: where $Q_L$ denotes the production of new $L_1$ monomers (see below).
200: A corresponding set of equations applies also to right-handed polymers,
201: i.e.\ $R_1$ and $R_n$.
202: Note that \Eqs{L1b}{Lnb} obey the conservation law
203: \begin{equation}
204: {\dd E_L\over\dd t}=Q_L-2k_S[L_1][L_N],
205: \label{dELdt}
206: \end{equation}
207: where
208: \begin{equation}
209: E_L=\sum_{n=1}^N n[L_n]
210: \label{ELdef}
211: \end{equation}
212: is the total number of left-handed building blocks.
213: This number reaches an equilibrium if the supply of
214: new left-handed monomers, $Q_L$, balances the loss associated
215: with reactions involving the longest polymers possible for a
216: given value of $N$.
217:
218: \EEq{Lnb} shows that in the steady state we have $[L_n]=\half[L_1]$
219: for all $n\geq2$.
220: Using \Eq{L1b}, we find $\lambda_L=k_S N[L_1]$, and therefore
221: \begin{equation}
222: 2[L_n]=[L_1]=\sqrt{Q_L/k_SN}\quad\mbox{(steady state)}
223: \label{HomochiralEquilib}
224: \end{equation}
225: is a possible equilibrium solution.
226:
227: New left and right handed
228: monomers are assumed to be continuously reproduced from an
229: achiral (racemic) substrate.
230: The rates of regeneration, $Q_L$ and $Q_R$, depend on the concentration of the
231: substrate, $[S]$, and in some fashion on the relative concentrations
232: of right and left handed polymers.
233: So, in general, we write
234: \begin{equation}
235: Q_L=k_C[S]\Big\{\half(1+f)C_L+\half(1-f)C_R+C_{0L}\Big\},
236: \label{QLdef}
237: \end{equation}
238: \begin{equation}
239: Q_R=k_C[S]\Big\{\half(1+f)C_R+\half(1-f)C_L+C_{0R}\Big\},
240: \label{QRdef}
241: \end{equation}
242: where $C_L$ and $C_R$ are some measures of the catalytic effect of the
243: already existing right and left handed polymers, and the terms $C_{0L}$
244: and $C_{0R}$ allow for the possibility of non-catalytic production of
245: left and right handed monomers -- possibly at different rates.
246: (Unless noted otherwise, we keep $C_{0L}=C_{0R}=0$.)
247:
248: The concentration of the substrate is assumed to be maintained by
249: a source $Q$, so we have
250: \begin{equation}
251: {\dd[S]\over\dd t}=Q-\left(Q_L+Q_R\right),
252: \label{dSdt}
253: \end{equation}
254: where $Q_L+Q_R=k_C[S](C_L+C_R+C_{0L}+C_{0R})$; see \Eqs{QLdef}{QRdef}.
255: In general, we expect $C_L$ and $C_R$ to be some function of
256: $L_n$ and $R_n$, respectively.
257: Sandars (2003) assumed $C_L=[L_N]$ and $C_R=[R_N]$, i.e.\ the
258: catalytic effect depends on the concentrations of the longest possible
259: chains of left and right handed polymers.
260: This assumption imposes a dependence on the cutoff value $N$,
261: a dependence that should preferably be avoided in numerical or other
262: technical considerations.
263: The model should for example be stable and consistent in the
264: limit when $N$ is infinite.
265: Another option would be to assume $C_L=[L_M]$ and $C_R=[R_M]$, where $M<N$
266: is a fixed value that is independent of the maximum chain length.
267: Both alternatives have the disadvantage that $[L_1]$ and $[R_1]$ can never
268: grow unless $[L_M]$ or $[R_M]$ are initially also finite.
269: While it is plausible that long chains carry more catalytic weight than shorter
270: ones, the dependence of the results on the particular choice of $M$
271: seems artificial.
272: (The allowance of finite values of $C_{0L}$ and $C_{0R}$ would remove
273: this problem, although in practice both of these values should still be quite small.)
274:
275: On may expect that the catalytic properties of the existing left
276: and right handed polymers depend on the length of the
277: polymer. The exact functional expression for this dependence
278: is not known. It is therefore important that a model that explains
279: homochirality is not sensitive to the details of the catalytic
280: properties and hence the functional form of $C _L$ and $C _R$.
281: It turns out that the qualitative
282: behavior of the model of Sandars is indeed robust in this respect,
283: e.g.\ a pitchfork bifurcation exists in both Sandars' original
284: and in our model.
285: To avoid artificial dependence on the maximal chain length
286: $N$, we chose to let the catalytic functions have the following
287: form
288: \begin{equation}
289: C_L=E_L,\quad
290: C_R=E_R,
291: \label{Cdef}
292: \end{equation}
293: where $E_L$ is given by \Eq{ELdef}, and $E_R$ is defined analogously.
294: This is similar to the choice of Wattis \& Coveney (2004) who assumed,
295: independently of us, $C_L=E_L-[L_1]$ and $C_R=E_R-[R_1]$.
296:
297: We now comment on another aspect of the model of Sandars.
298: He assumed that in the evolution of $[L_N]$ the loss is not
299: $2k_S[L_1][L_N]$, as it would be if \Eq{Lnb} were applied to $n=N$,
300: but he introduced an explicit linear damping term instead.
301: This implies that the model behaves discontinuously at the end
302: of the chain.
303: We feel that an ``extrapolating'' (continuous) behavior
304: is more reasonable, so we choose to apply \Eq{Lnb} also at $n=N$.
305:
306: \begin{figure}[t!]\begin{center}
307: \includegraphics[width=\textwidth]{fig1}
308: \end{center}\caption[]{
309: Wave-like propagation of a finite amplitude perturbation.
310: The initial profile is a gaussian.
311: Note the undisturbed propagation of the wave out of the chain at $n=N$.
312: The time difference between the different curves is $20/(k_SQ)^{1/2}$.
313: We have shown the first and last times as solid and dashed lines,
314: and all other times as dotted lines.
315: The parameters are $N=50$ and $k_C/k_S=1$.
316: }\label{pwave}\end{figure}
317:
318: It is interesting to note that in the continuous limit, \Eq{Lnb}
319: becomes
320: \begin{equation}
321: \left({\partial\over\partial t}+2k_S[L_1]
322: {\partial\over\partial n}\right)[L_n]=0,
323: \end{equation}
324: which describes waves
325: traveling toward larger $n$.
326: This is shown in \Fig{pwave}, where we have perturbed
327: the equilibrium solution \eq{HomochiralEquilib} by a gaussian
328: and have solved \Eqss{L1b}{Lnb} numerically.
329: The wave is damped and has a speed that is proportional to $(k_SQ)^{1/2}$,
330: because for the steady state background solution $[L_1]\sim (k_S/Q)^{1/2}$.
331: Note that the extrapolating boundary condition at $n=N$ allows
332: the wave to escape freely.
333:
334: In this paper we do not adopt the nondimensionalization of Sandars.
335: Instead, we note that there are only two physical dimensions in this
336: problem: time and volume.
337: Characteristic quantities with the dimensions of time and volume are
338: $(k_SQ)^{-1/2}$ and $(k_S/Q)^{1/2}$, respectively.
339: We therefore present all results by explicitly quoting these dimensions.
340: In practice this means that from now on we use $Q=k_S=1$ as numerical
341: values, but we keep the symbols in the equations for clarity.
342: Throughout this paper we also assume $k_C/k_S=1$; calculations with
343: different values do not seem to affect our results in any important way.
344:
345: The fact that the equilibrium solution is constant for all $n\geq2$
346: implies that this value will decrease for longer choices of $N$.
347: In that sense the solution is never converged.
348: This situation changes when we allow the ends of the left-handed polymers
349: to be spoiled by right-handed monomers, as done by Sandars (2003).
350: This will be discussed in the next section.
351:
352: \section{Enantiomeric cross-inhibition}
353:
354: Already 20 years ago, Joyce \ea (1984) showed in an important paper
355: describing experiments with template-directed polymerization that,
356: once a monomer of the opposite chirality is bound to one end of the
357: chain, the polymerization terminates on that end of the chain.
358: Sandars (2003) incorporated this effect in his model and showed that
359: this can lead to a bifurcation into two possible solutions of opposite
360: chirality and hence to homochirality.
361:
362: The full set of reactions included in his model is (for $n\ge2$)
363: \begin{eqnarray}
364: L_{n}+L_1&\stackrel{2k_S~}{\longrightarrow}&L_{n+1},
365: \label{react1}\\
366: L_{n}+R_1&\stackrel{2k_I~}{\longrightarrow}&L_nR_1,
367: \label{react2}\\
368: L_1+L_{n}R_1&\stackrel{k_S~}{\longrightarrow}&L_{n+1}R_1,
369: \label{react3}\\
370: R_1+L_{n}R_1&\stackrel{k_I~}{\longrightarrow}&R_1L_nR_1,
371: \label{react4}
372: \end{eqnarray}
373: and for all four equations we have the complementary reactions
374: obtained by exchanging $L\rightleftarrows R$.
375: Following Sandars (2003), we have introduced the new parameter $k_I$,
376: which quantifies that rate of enantiomeric cross-inhibition.
377: The special case $k_I=0$ corresponds to the case discussed in the
378: previous section.
379:
380: The most important effect of enantiomeric cross-inhibition is that
381: a certain fraction of chains becomes spoiled by producing $L_nR_1$
382: and $R_nL_1$ polymers.
383: \EEq{Lnb} and its complementary equation for right-handed polymers
384: suffer therefore a loss proportional to $2k_I$, so we have instead
385: \begin{equation}
386: {\dd[L_{n}]\over\dd t}=2k_S[L_1]\Big(\sigma_n^{(1/2)}[L_{n-1}]-[L_{n}]\Big)
387: -2k_I[L_{n}][R_1],
388: \label{Ln_new2}
389: \end{equation}
390: \begin{equation}
391: {\dd[R_{n}]\over\dd t}=2k_S[R_1]\Big(\sigma_n^{(1/2)}[R_{n-1}]-[R_{n}]\Big)
392: -2k_I[R_{n}][L_1].
393: \label{Rn_new2}
394: \end{equation}
395: These equations allow us to see what happens in the racemic case
396: with $[R_n]=[L_n]$.
397: In a steady state we have (for $n\ge2$)
398: \begin{equation}
399: [L_n]=\half a^{-(n-1)}[L_1]
400: \quad\mbox{(racemic solution)},
401: \end{equation}
402: where we have defined $a=1+k_I/k_S$.
403: In particular, if $k_I=k_S$, then $[L_n]=2^{-n}[L_1]$,
404: i.e.\ $[L_n]$ drops by a factor of 2 from one $n$ to the next,
405: except for $n=1$ to 2, where it drops by a factor of 4.
406: We should note, however, that this solution can be unstable
407: (see \Sec{SStability}).
408:
409: So far, we have not yet considered the evolution equations for the
410: concentrations of the mixed terms, $L_{n}R_1$ and $R_{n}L_1$.
411: Following Sandars (2003), we abbreviate the corresponding
412: concentrations by $[L_{n}R]$ and $[R_{n}L]$, respectively, i.e.\ without
413: the subscript 1 on the terminating end of the chain.
414: The effect of generating these terms was already manifested in
415: \Eqs{Ln_new2}{Rn_new2} through the appearance of the last term proportional
416: to $2k_I$.
417: Nevertheless, we do need to solve for $[L_{n}R]$ and $[R_{n}L]$
418: explicitly, because the reactions \eqs{react3}{react4} consume
419: $L_1$ and $R_1$ monomers, respectively.
420: The evolution equations for $[L_1]$ and $[R_1]$ are therefore given by
421: \begin{equation}
422: {\dd[L_1]\over\dd t}=Q_L-\lambda_L[L_1],\quad
423: {\dd[R_1]\over\dd t}=Q_R-\lambda_R[R_1],
424: \label{L1R1}
425: \end{equation}
426: where
427: \begin{equation}
428: \lambda_L=
429: 2k_S\sum_{n=1}^{N-1}[L_n]
430: +2k_I\sum_{n=1}^{N}[R_n]
431: +k_S\sum_{n=2}^{N-1}[L_nR]
432: +k_I\sum_{n=2}^{N}[R_nL],
433: \label{lambdaL}
434: \end{equation}
435: \begin{equation}
436: \lambda_R=
437: 2k_S\sum_{n=1}^{N-1}[R_n]
438: +2k_I\sum_{n=1}^{N}[L_n]
439: +k_S\sum_{n=2}^{N-1}[R_nL]
440: +k_I\sum_{n=2}^{N}[L_nR],
441: \label{lambdaR}
442: \end{equation}
443: are the decay rates that quantify
444: the losses associated with the reactions
445: \eqss{react1}{react4}, respectively.
446:
447: In \Eqs{lambdaL}{lambdaR} the concentrations $[L_{n}R]$ and $[R_{n}L]$
448: enter, so we have to solve their corresponding
449: evolution equations (for $n\geq2$)
450: \begin{equation}
451: {\dd[L_{n}R]\over\dd t}=
452: k_S[L_1]\Big(\sigma_n^{(0)}[L_{n-1}R]-[L_{n}R]\Big)
453: +k_I[R_1]\Big(2[L_{n}]-[L_nR]\Big),
454: \label{LRevol}
455: \end{equation}
456: \begin{equation}
457: {\dd[R_{n}L]\over\dd t}=
458: k_S[R_1]\Big(\sigma_n^{(0)}[R_{n-1}L]-[R_{n}L]\Big)
459: +k_I[L_1]\Big(2[R_{n}]-[R_nL]\Big),
460: \label{RLevol}
461: \end{equation}
462: where the $\sigma_n^{(0)}$ factor turns off the first term for $n=2$;
463: see \Eq{sigman}.
464: In \Eqs{LRevol}{RLevol}
465: the first two terms proportional to $k_S$ correspond to
466: the homochiral growth on the unspoiled end, i.e.\ reaction \eq{react3}.
467: The third term comes from reaction \eq{react2} and the fourth
468: term comes from reaction \eq{react4} and enters here and also in
469: \Eqs{lambdaL}{lambdaR} as a loss term.
470: For completeness, we note that the corresponding gain enters in
471: the evolution equations
472: \begin{equation}
473: {\dd[RL_{n}R]\over\dd t}=k_I[R_1][L_nR],\quad
474: {\dd[LR_{n}L]\over\dd t}=k_I[L_1][R_nL],
475: \label{RLnR+LRnL}
476: \end{equation}
477: which are not explicitly required for constructing a solution,
478: because these polymers no longer react with the monomers.
479: Nevertheless, solving \Eq{RLnR+LRnL} simultaneously with
480: \Eqs{Ln_new2}{Rn_new2} and \Eqss{L1R1}{RLevol} can be useful for monitoring
481: the evolution of the net chirality; see \Sec{Conservation}.
482:
483: Note that, in contrast to the equations given by Sandars (2003), the
484: truncation levels for the terms $[L_n]$, $[L_nR]$, and $[RL_nR]$ are
485: here the same,
486: i.e.\ $n\leq N$ for both terms, whereas in the model of Sandars the longest
487: $L_nR_1$ chain has $n=N-1$, and the longest $R_1L_nR_1$ chain has only
488: $n=N-2$.
489: The reason we need to keep the same truncation levels for all three types
490: of polymers is that we want to ensure that the behavior near the end of
491: the chain ($n=N$) does not deviate from the behavior elsewhere ($n<N$);
492: see the discussion in \Sec{HomochiralGrowth}.
493: For example, to ensure continuous behavior of $[L_n]$ at $n=N$ we need
494: to keep the term $-2k_I[L_N][R_1]$ in \Eq{Ln_new2}.
495: This term, however, is the loss resulting from the gain of $[L_NR]$, so
496: we have to keep the evolution equation for this term as well.
497: Furthermore, the evolution equation for this term involves, in turn,
498: the term $-k_I[R_1][L_NR]$, which is the loss corresponding to the gain
499: of $[RL_NR]$.
500: If one regards the truncation level $N$ as an unrealistic feature of the
501: model, as we do, then all three polymer types should be truncated
502: at the same level.
503:
504: \begin{figure}[t!]\begin{center}
505: \includegraphics[width=\textwidth]{fig2}
506: \end{center}\caption[]{
507: $[L_n]$ (left) and $[L_nR]$ (right) of equilibrium solutions for different
508: values of $f$.
509: For $f=1$ we have $[L_nR]=0$, which cannot be seen in the logarithmic
510: representation.
511: }\label{pspectra}\end{figure}
512:
513: In \Fig{pspectra} we show $[L_n]$ and $[L_nR]$ for a number of
514: equilibrium solutions for different values of $f$ and $k_I/k_S=1$.
515: The corresponding values of $[R_n]$ and $[R_nL]$ are small and
516: not shown, except when $f=0$ in which case the solution is fully
517: racemic with $[R_n]=[L_n]$ and $[R_nL]=[L_nR]$ and is simply
518: \begin{equation}
519: [L_nR]=(n-1)[L_n]=(n-1)2^{-n}[L_1].
520: \label{RacemicSolution}
521: \end{equation}
522: For $f=1$ the solution is given by \Eq{HomochiralEquilib}.
523:
524: For $k_I/k_S=0.1$ the results are similar to those for $k_I/k_S=1$
525: provided $f>0.8$.
526: For $f<0.7$, however, the solution is fully racemic and therefore
527: the curves are independent of $f$.
528: This racemic solution is similar to the case
529: $k_I/k_S=1$ and $f=0.8$ that is shown in \Fig{pspectra}.
530:
531: \section{Stability of the racemic equilibrium}
532: \label{SStability}
533:
534: A realistic model of homochirality must also have an achiral (racemic)
535: equilibrium solution.
536: It is generally anticipated that this racemic solution can be destabilized
537: in the presence of catalytic reactions (Frank, 1953;
538: Avetisov and Goldanskii, 1993).
539: If the probabilities of adding left and right handed monomers to a
540: homochiral polymer are equal, i.e.\ if $k_I=k_S$, the racemic solution
541: given by \Eq{RacemicSolution} is also a possible solution for other
542: values of the fidelity than $f=0$, but it may of course be unstable.
543:
544: We have carried out a numerical stability analysis by adding a small
545: ($10^{-5}$) relative perturbation to the value of $[L_1]$ of the racemic
546: solution.
547: It turns out that for certain values of the fidelity $f$ the departure
548: of $[L_1]$ from the racemic equilibrium solution, $\delta[L_1]$, growth
549: exponentially in time like $e^{\lambda t}$.
550: In \Figs{bifurc}{bifurc01} we plot $\lambda$ obtained from the slope
551: of the graph of $\ln\delta[L_1](t)$ during the exponential growth phase,
552: i.e.\ before a new nonlinear equilibrium is attained.
553: In \Figs{bifurc}{bifurc01} we also plot the corresponding
554: chiral polarization parameter, $\eta$, as a function of $f$.
555: Here we have chosen to define $\eta$ as
556: \begin{equation}
557: \eta=\left(E_R-E_L\right)/\left(E_R+E_L\right).
558: \label{eta}
559: \end{equation}
560: It turns out that for $k_I/k_S=1$ the racemic solution is unstable when
561: $f>0.39$, and for $k_I/k_S=0.1$ it is unstable when $f>0.735$.
562: The transition from an achiral to a chiral solution is a typical
563: example of a pitchfork bifurcation; see \Figs{bifurc}{bifurc01}.
564: This result is in qualitative agreement with Sandars (2003) who
565: found that for $k_I/k_S=1$ the critical value of $f$ is around 0.21.
566: The differences in the numerical values are explained by differences
567: in the model (e.g., the coupling to the substrate and the length
568: of the maximum polymer length).
569:
570: \begin{figure}[t!]\begin{center}
571: \includegraphics[width=\textwidth]{fig3}
572: \end{center}\caption[]{
573: Growth rate (left) and bifurcation diagram
574: showing a classical pitchfork bifurcation (right)
575: as a function of fidelity for $k_I/k_S=1$, and $N=50$.
576: The dashed line indicates an unstable solution.
577: }\label{bifurc}\end{figure}
578:
579: \begin{figure}[t!]\begin{center}
580: \includegraphics[width=\textwidth]{fig4}
581: \end{center}\caption[]{
582: Same as \Fig{bifurc}, but for $k_I/k_S=0.1$.
583: }\label{bifurc01}\end{figure}
584:
585: The growth rate of the instability is important for determining the time
586: it takes for an almost racemic solution to become homochiral
587: (or at least non-racemic for $f\neq1$).
588: When $k_I/k_S=1$, the growth rate $\lambda$ is around 0.5, but it becomes
589: significantly smaller when the value of $k_I$ is reduced.
590: This shows explicitly that homochirality emerges as being due to
591: enantiomeric cross-inhibition.
592:
593: \section{Conservation of chirality}
594: \label{Conservation}
595:
596: For homochiral growth the relevant conservation law is given by
597: \Eq{dELdt} for $E_L$, and similarly for $E_R$.
598: In general, however, because of the interaction with left and
599: right handed monomers, there are no longer separate conservation
600: laws for $E_L$ and $E_R$.
601: Instead, the complete set of equations, \Eqs{Ln_new2}{Rn_new2}
602: together with \Eqss{L1R1}{lambdaR}, satisfies
603: \begin{equation}
604: {\dd\over\dd t}\Delta\tilde{E}=\Delta Q-\Delta\Lambda,
605: \label{H_cons}
606: \end{equation}
607: where $\Delta Q=Q_R-Q_L$ and $\Delta\Lambda=\Lambda_R-\Lambda_L$
608: are the net input and output rates of chirality, respectively, and
609: $\Delta\tilde{E}=\tilde{E}_R-\tilde{E}_L$ is the total chirality, where
610: \begin{equation}
611: \tilde{E}_R=\sum_{n=1}^N n[R_n]+\sum_{n=2}^N(n-1)[R_nL]
612: +\sum_{n=3}^N(n-2)[LR_nL],
613: \label{ERtilde}
614: \end{equation}
615: \begin{equation}
616: \tilde{E}_L=\sum_{n=1}^N n[L_n]+\sum_{n=2}^N(n-1)[L_nR]
617: +\sum_{n=3}^N(n-2)[RL_nR],
618: \label{ELtilde}
619: \end{equation}
620: denote the total numbers of right and left handed building blocks
621: (or enantiomers),
622: where opposite enantiomers are counted such that they annihilate
623: enantiomers of the opposite chirality.
624: The loss terms resulting from the finite truncation level, $N$, are
625: denoted by
626: \begin{equation}
627: \Lambda_R=2k_S N[R_1][R_N]+k_S(N-1)[R_1][R_NL],
628: \end{equation}
629: \begin{equation}
630: \Lambda_L=2k_S N[L_1][L_N]+k_S(N-1)[L_1][L_NR].
631: \end{equation}
632: In order to evaluate the quantities $\tilde{E}_R$ and
633: $\tilde{E}_L$ we have to integrate the evolution equations \eq{RLnR+LRnL}
634: for the production of terminally spoiled polymers -- even though they
635: undergo no further evolution.
636: In a sense the integration of the terminally spoiled polymers acts
637: only as counters that keeps track of the number
638: of polymers that are lost during the polymerization process.
639:
640: The expressions for $\tilde{E}_R$ and $\tilde{E}_L$ involve sums over
641: $[LR_nL]$ and $[RL_nR]$, but since these quantities do not occur on the
642: right hand sides of the governing evolution equations, their values are
643: not constrained by the dynamics and depend on the initial conditions
644: and continue to evolve in time even though the system may have reached
645: an equilibrium.
646: The so defined net chirality can therefore not be used to characterize
647: a particular solution, and we have to restrict ourselves either to
648: $E_R$ and $E_L$, or to $\hat{E}_R$ and $\hat{E}_L$, which are defined
649: by taking only the first two sums in \Eqs{ERtilde}{ELtilde}, i.e.\
650: \begin{equation}
651: \hat{E}_R=\sum_{n=1}^N n[R_n]+\sum_{n=2}^N(n-1)[R_nL],
652: \label{ERtilde}
653: \end{equation}
654: \begin{equation}
655: \hat{E}_L=\sum_{n=1}^N n[L_n]+\sum_{n=2}^N(n-1)[L_nR].
656: \label{ELtilde}
657: \end{equation}
658: In \Tab{Tres} we list the resulting values of $\eta$,
659: defined in \Eq{eta}, an analogously defined
660: $\hat\eta=(\hat{E}_R-\hat{E}_L)/(\hat{E}_R+\hat{E}_L)$,
661: $\Delta E=E_R-E_L$, and $\Delta\hat{E}=\hat{E}_R-\hat{E}_L$.
662: We also give the mean polymer lengths, $N_R=\sum n[R_n]/\sum[R_n]$
663: and $N_L=\sum n[L_n]/\sum[L_n]$, of right and left handed polymers.
664:
665: \begin{table}[t!]\caption{
666: Numerical results for $\eta$, $\hat\eta$, $\Delta E$, and $\Delta\hat{E}$
667: for different values of $f$.
668: The $\pm$ indicates that these values can have either sign.
669: The last column gives the typical value of $N$ necessary for obtaining
670: a converged result representing the limit $N\to\infty$.
671: For $f=1$ the results for $\Delta E$ ($=\Delta\hat{E}$) and $N_R$
672: do not converge
673: and we give the analytic expression for arbitrary $N$ instead.
674: }\vspace{12pt}{\begin{tabular}{cccccccc}
675: $f$ & $\pm\eta$ & $\pm\hat\eta$ & $\pm\Delta E$ & $\pm\Delta\hat{E}$
676: & $N_R$ & $N_L$ & $N$ \\
677: \hline
678: 1 & $1$ & $1$ & \multicolumn{2}{c}{$\quarter[N(N+1)+2]/N^{1/2}$}
679: & $\half N+1/(N+1)$& -- & $N$\\
680: 0.9 & $0.999$ & $0.9999$& $30.61$ & $143.73$ &19.0 & 1.0 & 500 \\
681: 0.8 & $0.995$ & $0.9986$& $10.28$ & $45.06$ & 9.0 & 1.1 & 200 \\
682: 0.7 & $0.978$ & $0.993$ & $5.170$ & $20.99$ & 5.6 & 1.1 & 100 \\
683: 0.6 & $0.933$ & $0.975$ & $2.949$ & $11.01$ & 3.9 & 1.2 & 100 \\
684: 0.5 & $0.813$ & $0.907$ & $1.648$ & $5.597$ & 2.8 & 1.2 & 50 \\
685: 0.4 & $0.368$ & $0.482$ & $0.491$ & $1.500$ & 1.9 & 1.5 & 50 \\
686: 0.38& $0$ & $0$ & $0$ & $0$ & 1.7 & 1.7 & 20 \\
687: \label{Tres}\end{tabular}}\end{table}
688:
689: \section{Comparison with other models}
690:
691: The polymerization model of Sandars (2003) is significantly different
692: from all the previously proposed models of homochirality that ignore
693: the detailed polymerization process by only describing
694: some scalar quantities, say $x$ and $y$, that are representative
695: of the number of left and right handed polymers.
696: In the papers by Saito and Hyuga (2004) it was shown that neither linear
697: nor nonlinear autocatalytic behavior suffice to produce homochirality,
698: and that a backreaction term is needed.
699: Their model equations are
700: \begin{eqnarray}
701: \begin{array}{l}
702: \dot{x}=x^2(1-r)-\epsilon x\cr
703: \dot{y}=y^2(1-r)-\epsilon y
704: \end{array}
705: \quad\mbox{(SH model)}
706: \label{SaitoHyuga}
707: \end{eqnarray}
708: where $r=x+y$ and $\epsilon$ is the feedback parameter.
709: For $\epsilon=0$ there is a continuous range of solutions along the line
710: $r\equiv x+y=1$, i.e.\ homochirality does not emerge unless the initial
711: condition is already homochiral.
712: For finite (but small) values of $\epsilon$ there are two nontrivial
713: stable fixed points.
714: (The trivial solution, $x=y=0$, is always a stable fixed point in
715: this model.)
716:
717: The model of Saito and Hyuga (2004), hereafter the SH model,
718: does capture the expected behavior,
719: but it remains unsatisfactory in that its functional form has been introduced
720: ad hoc.
721: It is therefore desirable to derive simple model equations based
722: on the polymerization equations of Sandars (2003).
723: It turns out that, without changing the basic properties of the model,
724: a minimal version is still meaningful for $N=2$,
725: and that the equations for the semi-spoiled polymers, $[L_2R]$ and
726: $[R_2L]$, can be ignored (as already done by Sandars).
727: Thus, we only solve \Eqs{Ln_new2}{Rn_new2} together with \Eqss{L1R1}{lambdaR}.
728: Following Sandars (2003), we also assume that $C_L=[L_2]$ and $C_R=[R_2]$
729: (instead of $C_L=E_L$ and $C_R=E_R$, which would yield more complicated
730: expressions).
731: A further simplification can be made by regarding $[L_2]$ as a rapidly
732: adjusting variable that is enslaved to $[L_1]$ (and similarly for $[R_2]$).
733: This technique is also known as the adiabatic elimination of rapidly
734: adjusting variables (e.g., Haken, 1983).
735: \EEq{Ln_new2} becomes
736: \begin{equation}
737: 0=k_S[L_1]^2-2[L_2]\Big(k_S[L_1]+k_I[R_1]\Big),
738: \end{equation}
739: which is solved for $[L_2]$ (and similarly for $[R_2]$), which in turn
740: couples back to the equations for $[L_1]$ and $[R_1]$ via $Q_L$ and $Q_R$.
741: Finally, we also treat the substrate $[S]$ as a rapidly adjusting variable,
742: i.e.\ we have $k_C[S]=Q/([L_2]+[R_2])$.
743: We emphasize that the adiabatic elimination does not affect the
744: accuracy of steady solutions.
745: It is convenient to introduce new dimensionless variables,
746: \begin{equation}
747: x=[R_1](2k_S/Q)^{1/2},\quad
748: y=[L_1](2k_S/Q)^{1/2},\quad
749: \tau=t(Qk_S/2)^{1/2}.
750: \end{equation}
751: In order to compare first with the SH model we restrict ourselves
752: to the special case $k_I/k_S=f=1$, which leads to the revised model
753: equations
754: \begin{eqnarray}
755: \begin{array}{l}
756: \dot{x}=x^2/\tilde{r}^2-rx,\cr
757: \dot{y}=y^2/\tilde{r}^2-ry,
758: \end{array}
759: \label{modeleqn}
760: \end{eqnarray}
761: where dots denote derivatives with respect to $\tau$ and
762: $r=x+y$ and $\tilde{r}^2=x^2+y^2$ has been introduced for brevity.
763: Equations \eq{modeleqn} resemble the equations of the SH model
764: in that both have a quadratic term proportional to $x^2$ (or $y^2$),
765: which is quenched either by a $1-r$ factor (in the SH model) or by
766: a $1/\tilde{r}^2$ factor in our model.
767: Furthermore, both models have a backreaction term proportional to $-x$
768: (or $-y$), but the coefficient in front of this term
769: ($\epsilon$ in the SH model) is not constant but equal to $r$.
770:
771: \begin{figure}[t!]\begin{center}
772: \includegraphics[width=\textwidth]{fig5}
773: \end{center}\caption[]{
774: Phase diagram showing the trajectories of solutions
775: of \Eq{modeleqnf} for two different values of $f$.
776: The starting points of each trajectory are marked by small dots
777: and stable fixed points are marked by big dots.
778: }\label{pruns051}\end{figure}
779:
780: In the general
781: case with $k_I/k_S\neq1$, $f\neq1$, as well as finite values of
782: $C_{0x}=C_{0R}(2k_S/Q)^{1/2}$, and $C_{0y}=C_{0L}(2k_S/Q)^{1/2}$,
783: the equations read
784: \begin{eqnarray}
785: \begin{array}{l}
786: \dot{x}=(p\tilde{x}^2+q\tilde{y}^2+C_{0x})/\tilde{r}^2-r_x x,\cr
787: \dot{y}=(p\tilde{y}^2+q\tilde{x}^2+C_{0y})/\tilde{r}^2-r_y y,
788: \end{array}
789: \label{modeleqnf}
790: \end{eqnarray}
791: where we have introduced the abbreviations
792: $r_x=x+yk_I/k_S$, $r_y=y+xk_I/k_S$,
793: $\tilde{x}^2=x^2/2r_x$, $\tilde{y}^2=y^2/2r_y$,
794: $\tilde{r}^2=\tilde{x}^2+\tilde{y}^2+C_{0x}+C_{0y}$,
795: $p=(1+f)/2$, and $q=(1-f)/2$.
796:
797: In \Fig{pruns051} we show trajectories of solutions of \Eq{modeleqnf}
798: for two different values of $f$ in an $(x,y)$ phase diagram.
799: Note that all equilibrium solutions lie on the line $r=1$.
800: This property allows us to calculate equilibrium solutions for
801: general values of $f$.
802: Inserting $y=1-x$ yields a cubic equation of which one solution
803: is always $x=1/2$.
804: This reduces the problem to a quadratic equation with the solution
805: \begin{eqnarray}
806: x=\half\left\{
807: \begin{array}{ll}
808: 1\pm\sqrt{2f-1}\quad&\mbox{for $f\geq1/2$},\\
809: 1 & \mbox{otherwise}.
810: \end{array}
811: \right.
812: \end{eqnarray}
813: Linearizing the equations around the racemic solution, $x=y=1/2$, yields
814: the growth rate
815: \begin{equation}
816: \lambda=2f-1.
817: \end{equation}
818: In agreement with our numerical results for large values of $N$,
819: this equation gives a linear dependence of the growth rate on
820: the fidelity.
821: This result also shows that for $f<1/2$ perturbations decay exponentially.
822:
823: \begin{figure}[t!]\begin{center}
824: \includegraphics[width=.8\textwidth]{fig6}
825: \end{center}\caption[]{
826: Imperfect bifurcation obtained by solving \Eq{modeleqnf} for
827: $C_{0x}=0.001$ and $C_{0y}=0$ using the Newton-Raphson method.
828: }\label{pbranches}\end{figure}
829:
830: In the presence of a biased, non-catalytic generation of monomers
831: (finite $C_{0x}$ or $C_{0y}$ with $C_{0x}\neq C_{0y}$) there is no
832: longer a perfectly racemic equilibrium solution.
833: The sign of $\eta$ for the solution for $f=0$ depends on the
834: sign of $C_{0x}-C_{0y}$.
835: Along this solution branch $\eta$ goes further away from zero in
836: a continuous fashion until $f=1$.
837: At some value of $f$ a pair of new solutions emerges, one is stable and
838: the other one unstable, but both have the opposite sign of $\eta$;
839: see \Fig{pbranches}.
840: Among these new branches, the stable one can only be reached via a
841: finite amplitude perturbation.
842: This behavior is called an imperfect bifurcation and has long been
843: anticipated in this context (Kondepudi and Nelson, 1983;
844: Kondepudi \ea 1986; Goldanskii and Kuzmin, 1989).
845:
846: The steady solutions shown in \Fig{pbranches}
847: have been obtained by solving \Eq{modeleqnf} using the
848: Newton-Raphson method.
849: This method allows us to find both stable and unstable solutions.
850: Near the bifurcation point the diagram is extremely sensitive to the
851: addition of a bias parameter.
852: It is remarkable that for a value as small as $C_{0x}=10^{-3}$ a relatively
853: large gap has been produced in the bifurcation diagram.
854:
855: Finite values of $C_{0x}$ and $C_{0y}$ could result from physical
856: influences, for example polarized synchrotron radiation from neutron stars
857: (but see Bonner, 1999), UV radiation in star-forming regions (Bailey, 2001),
858: or the parity violation of the electroweak force (e.g., Hegstrom, 1984).
859: In all these cases the expected effect is however very small (Bada, 1995).
860: We emphasize, however, that the main reason for homochirality is the
861: instability of the racemic (or nearly racemic) solution, which is
862: hardly modified by a finiteness of $C_{0x}$ or $C_{0y}$.
863:
864: \section{Conclusions}
865:
866: The origin of homochirality has long been thought to be the result
867: of a bifurcation process that can vastly amplify a very small random
868: enantiomeric excess which can then prevail forever.
869: Generic model equations reproducing the expected bifurcation behavior
870: have so far mostly been proposed on an ad hoc basis.
871: It was therefore difficult to establish a connection between model
872: and reality.
873: According to the work of Saito and Hyuga (2004) one expects two
874: effects to be important: nonlinearity and backreaction.
875: However, the functional form of these terms remained open.
876: Furthermore, the meaning of non-perfect catalytic fidelity
877: and enantiomeric cross-inhibition within the framework of the
878: model were not clear.
879: In the present paper we have established a direct connection between
880: the more detailed polymerization model of Sandars (2003) and
881: the simpler model equation approach with only two ordinary differential
882: equations.
883: In particular, the present work has confirmed that the relevant
884: nonlinearity is indeed quadratic (as in the SH model), but it is not
885: quenched like $1-r$, but rather like $1/\tilde{r}^2$, where $r$ and
886: $\tilde{r}$ are measures of the total concentrations of monomers
887: (both right and left handed).
888: Furthermore, the feedback coefficient is not a small constant, as in
889: the SH model, but it is itself proportional to $r$.
890: More importantly, imperfect fidelity and enantiomeric cross-inhibition,
891: as well as the effects of a weakly biased non-catalytic production of
892: new monomers, have a quantitative meaning within the framework of
893: the reduced model.
894:
895: For a more quantitative comparisons of the polymerization process with
896: experiments the full set of equations of Sandars (with the revisions
897: discussed above) is to be preferred.
898: A number of features that can only be captured by the full model.
899: An example is the wave-like propagation in the distribution of homochiral
900: polymers.
901: An experimental confirmation would help to quantify the growth coefficient
902: $k_S$ characterizing the probability that a polymer grows by a monomer
903: of the same chirality.
904: On the other hand, the growth coefficient for enantiomeric
905: cross-inhibition, $k_I$, determines primarily the minimum fidelity
906: parameter, $f$, above which bifurcation and hence homochiral growth is
907: at all possible.
908: It is indeed quite remarkable, that the main reason homochiral
909: growth occurs is that binding with a wrong enantiomer spoils further
910: polymerization on the corresponding end of the chain.
911: This leads to competition which is always a key feature of natural
912: selection processes such as these.
913:
914: Homochirality in living organisms is a singular phenomenon.
915: Non-living chemical systems do in general not have a preferred
916: chirality.
917: In the models presented in this paper this is reflected in
918: \Figs{bifurc}{pbranches}.
919: The region of the phase diagram displaying homochirality is
920: characterized by high fidelity, i.e.\ high auto-catalytic accuracy.
921: The fidelity is expected to be significantly higher in living
922: systems.
923: When an organism dies the auto-catalytic polymerization
924: stops and as a consequence the fidelity is sharply decreased.
925: The characteristic behavior of the polymerization changes from the chiral
926: to the racemic region of the phase diagram. The relaxation
927: of the system from the homochiral to the racemic state is
928: often very slow. It was in fact suggested by Hare and Mitterer
929: (1967) and later by Bada \ea (1970)
930: that racemization of amino acids in fossil material could
931: be used as a dating method.
932: Unfortunately it has turned out that the rate of racemization
933: is strongly temperature dependent, which tends to make this
934: dating method unreliable.
935:
936: \begin{thebibliography}{}
937:
938: \bibitem[]{}
939: Avetisov, V. A. and Goldanskii, V.\yphl{1993}{A 172}{407}
940: {410}{Chirality and the equation of `biological big bang'}
941:
942: \bibitem[]{}
943: Bada, J. L., Luyendyk, B. P., and Maynard, J. B.\ysci{1970}{170}{730}
944: {732}{Marine Sediments: Dating by the Racemization of Amino Acids}
945:
946: \bibitem[]{}
947: Bada, J. L.\ynat{1995}{374}{594}
948: {595}{Origins of homochirality}
949:
950: \bibitem[]{}
951: Bailey, J.\yoleb{2001}{31}{167}
952: {183}{Astronomical sources of circularly polarized light and the origin
953: of homochirality}
954:
955: \bibitem[]{}
956: Bonner, W. A.\yoleb{1999}{29}{615}
957: {624}{Chirality amplification -- the accumulation principle revisited}
958:
959: \bibitem[]{}
960: Crick, F. H. C.\yjour{1968}{J.\ Mol.\ Biol.}{38}{367}
961: {379}{The origin of the genetic code}
962:
963: \bibitem[]{}
964: Frank, F.: 1953, On Spontaneous Asymmetric Synthesis,
965: Biochim. Biophys. Acta 11, 459-464.
966:
967: \bibitem[]{}
968: Goldanskii, V. I. and Kuzmin, V. V.\yjour{1989}{Sov.\ Phys.\ Uspekhi}{32}{1}
969: {29}{Spontaneous breaking of mirror symmetry in nature and origin of life}
970:
971: \bibitem[]{}
972: Haken, H.\ybook{1983}{Synergetics -- An Introduction}{Springer: Berlin}
973:
974: \bibitem[]{}
975: Hare, P. E. and Mitterer, R. M.\yjour{1967}
976: {Yearbook Carnegie Institution of Washington}{65}{362}
977: {364}{Nonprotein amino acids in fossil shells}
978:
979: \bibitem[]{}
980: Hegstrom, R. A.\yjour{1984}{Orig.\ Life}{14}{405}
981: {414}{Parity nonconservation and the origin of biological chirality --
982: theoretical calculations}
983:
984: \bibitem[]{}
985: Joyce, G. F., Visser, G. M., van Boeckel, C. A. A., van Boom, J. H.,
986: Orgel, L. E., and Westrenen, J.\ynat{1984}{310}{602}
987: {603}{Chiral selection in poly(C)-directed synthesis of oligo(G)}
988:
989: \bibitem[]{}
990: Kippenhahn, R. and Weigert, A.\ybook{1990}{Stellar structure and evolution}
991: {Springer: Berlin}
992:
993: \bibitem[]{}
994: Kondepudi, D. K. and Nelson, G. W.\yprl{1983}{50}{1023}
995: {1026}{Chiral symmetry breaking in nonequilibrium chemical systems}
996:
997: \bibitem[]{}
998: Kondepudi, D. K. and Nelson, G. W.\yphl{1984}{106A}{203}
999: {206}{Chiral symmetry breaking in nonequilibrium chemical systems:
1000: time scales for chiral selection}
1001:
1002: \bibitem[]{}
1003: Kondepudi, D. K., Moss, F., and McClintock, P. V. E.\yphy{1986}{21D}{296}
1004: {306}{Observation of symmetry breaking, state selection and sensitivity in a
1005: noisy electron system}
1006:
1007: \bibitem[]{}
1008: Kondepudi, D. K., Kaufman, R. J., and Singh, N.\ysci{1990}{250}{975}
1009: {976}{Chiral symmetry breaking in sodium chlorate crystallization}
1010:
1011: \bibitem[]{}
1012: Nelson, K. E., Levy, M., and Miller, S. L.\yjour{2000}
1013: {Proc.\ Nat.\ Acad.\ Sci.\ U.S.A.}{97}{3868}
1014: {3871}{Peptide Nucleic Acids rather than RNA may have been
1015: the first genetic molecule}
1016:
1017: \bibitem[]{}
1018: Nielsen, P. E.\yoleb{1993}{23}{323}
1019: {327}{Peptide nucleic acid (PNA): A model structure for the primordial
1020: genetic material}
1021:
1022: \bibitem[]{}
1023: Orgel, L. E.\yjour{1968}{J.\ Mol.\ Biol.}{38}{381}
1024: {393}{Evolution of the genetic apparatus}
1025:
1026: \bibitem[]{}
1027: Pogodina, N. V., Lavrenko, V. P., Srinivas, S., and Winter, H. H.\yjour{2001}
1028: {Polymer}{42}{9031}
1029: {9043}{Rheology and structure of isotactic polypropylene near the gel
1030: point: quiescent and shear-induced crystallization}
1031:
1032: \bibitem[]{}
1033: Pooga, M., Land, T., Bartfai, T., Langel, \"U.\yjour{2001}{Biomol.\ Eng.}{17}{183}
1034: {192}{PNA oligomers as tools for specific modulation of gene expression}
1035:
1036: \bibitem[]{}
1037: Rasmussen, S., Chen, L., Nilsson, M., and Abe, S.\yjour{2003}{Artif.\ Life}{9}{269}
1038: {316}{Bridging nonliving and living matter}
1039:
1040: \bibitem[]{}
1041: Sandars, P. G. H.\yoleb{2003}{33}{575}
1042: {587}{A toy model for the generation of homochirality during polymerization}
1043:
1044: \bibitem[]{}
1045: Saito, Y. and Hyuga, H.\yjour{2004}{J.\ Phys.\ Soc.\ Jap.}{73}{33}
1046: {35}{Complete homochirality induced by the nonlinear autocatalysis
1047: and recycling}
1048:
1049: \bibitem[]{}
1050: Tedeschi, T., Corradini, R., Marchelli, R., Pushl, A., and
1051: Nielsen, P. E.\yjour{2002}{Tetrahedron: Asymmetry}{13}{1629}
1052: {1636}{Racemization of chiral PNAs during solid-phase synthesis: effect
1053: of the coupling conditions on enantiomeric purity}
1054:
1055: \bibitem[]{}
1056: Wattis, J. A. D. and Coveney, P. V.\yjour{1999}{J.\ Phys.\ Chem.\ B}{103}{4231}
1057: {4250}{The origin of the RNA world: a kinetic model}
1058:
1059: \bibitem[]{}
1060: Wattis, J. A. D. and Coveney, P. V.: 2004, Symmetry-breaking in
1061: chiral polymerisation, arXiv:physics/0402091.
1062:
1063: \bibitem[]{}
1064: Woese, C.\ybook{1967}{The Genetic Code}{New York: Harper and Row}
1065:
1066: \end{thebibliography}
1067: \end{article}
1068: \end{document}
1069: