1: %\documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
2: \documentclass[preprint,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
3: \usepackage{graphicx}% Include figure files
4: \usepackage{dcolumn}% Align table columns on decimal point
5: \usepackage{bm}% bold math
6: \begin{document}
7: \title{Mass Fractal Dimension of the Ribosome and Implication of
8: its Dynamic Characteristics}
9: \author{Chang-Yong Lee}
10: \email{clee@kongju.ac.kr}
11: \affiliation{The Department of Industrial Information, Kongju National
12: University, Chungnam, 340-702 South Korea}
13: \date{\today}
14:
15: \begin{abstract}
16: Self-similar properties of the ribosome in terms of the mass
17: fractal dimension are investigated. We find that both the 30S subunit
18: and the 16S rRNA have fractal dimensions of 2.58 and 2.82,
19: respectively; while the 50S subunit as well as the 23S rRNA has the
20: mass fractal dimension close to 3, implying a compact three dimensional
21: macromolecule. This finding supports the dynamic and active
22: role of the 30S subunit in the protein synthesis, in contrast to the
23: pass role of the 50S subunit.
24: \end{abstract}
25:
26: \pacs{87.14.Gg, 87.15.-v, 61.43.Hv}
27:
28: \maketitle
29: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30: The structure of biomolecules is important because not only the
31: structure dictates its biological function, but it is the target of
32: antibiotics. In this sense, finding characteristics of the three-dimensional
33: shape of biomolecules is important for a better understanding of their
34: biological functions and associated applications in medicine. In the
35: case of the ribosome~\cite{ribosome}, as a large protein-RNA complex,
36: in contrast to most cellular machines, it has been known that the
37: ribosomal function heavily rely on the ribosomal RNA (rNRA), as a
38: ribozyme, rather than protein components~\cite{function}. In
39: particular, the protein synthesis is closely related to the dynamic
40: structure of the ribosome, which is too complicated for direct studies
41: by physical methods. However, careful study on the static structure from a
42: quantitative perspective may reveal an important aspect of its dynamic
43: properties.
44:
45: The quantitative investigation on the structure of the ribosome is
46: relatively less studied than that of protein, mainly due to the
47: difficulty of the highly resolved structure conformation. In contrast
48: to the ribosome, the geometric and self-similar properties of the
49: proteins have been studied extensively. It was reported that the
50: relation between the average radius and the mass of the protein chains
51: can be described by a fractal dimension~\cite{moret}. The mass fractal
52: dimension of the protein was shown to lie near 3, suggesting a compact
53: three-dimensional object~\cite{elber}. However, more recent and
54: extensive studies, including a statistical analysis in estimating the
55: dimension, argue smaller values that are consistent with the
56: result of the vibrational analysis of proteins~\cite{enright}. These
57: results suggest that the fractal dimension of less than 3
58: may be intrinsic and universal characteristics of the protein chain.
59:
60: Since the ribosome, as a tightly packed macromolecule, was considered
61: too large for a high-resolution structural analysis, quantitative
62: studies toward an understanding of the structure proved difficult
63: until recent progress in the high resolution crystallography has been
64: made. In fact, the ribosome and its subunits are the largest
65: asymmetric molecules that have been resolved at the atomic level so far
66: by the crystallography. The 2.4 $\mathring{A}$ high-resolution of the
67: 50S subunit from the {\it Haloarcula marismortuii}~\cite{ban} and the
68: 3.05 $\mathring{A}$ structure of the 30S subunit from the {\it
69: Thermus thermophilus}~\cite{wimberly} provided the first detailed
70: views of the structure of both ribosomal subunits; the intact 70S
71: ribosome from a {\it Escherichia coli} of 3.5 $\mathring{A}$
72: resolution~\cite{schuwirth} revealed the features of the inter-subunit
73: bridges. In addition to these, there are other X-ray crystal
74: structures available for the ribosome as well as its
75: subunits~\cite{harms,schluenzen,cate,yusupov}. With these considerable
76: advances in the ribosome structures at the atomic level, we are now
77: able to investigate quantitative characteristics of the ribosome from
78: the statistical physics perspective.
79:
80: In bacteria, the ribosome is a particle of size about 250
81: $\mathring{A}$ in diameter and consists mainly of two subunits: 30S and 50S
82: subunits, together forming the 70S. The unit ``S'' stands for Svedberg,
83: which is a measure of the sedimentation rate. The 30S subunit plays a
84: crucial role in decoding mRNA by monitoring base pairing between codon
85: and anticodon; whereas the 50S subunit catalyzes peptide bond
86: formation between the incoming amino acid and the nascent peptide
87: chain~\cite{ramakrishnan}. The 30S subunit, in turn, contains the 16S
88: rRNA molecule in addition to about 20 different proteins, and the 50S subunit
89: consists of the 5S and the 23S rRNAs besides about 30 different
90: proteins. The 16S and 23S rRNAs are composed of approximately 1500 and
91: 3000 nucleotides respectively, each of which is composed of one of four
92: different bases (denoted as A, C, G, and U), and sugar-phosphate backbones.
93:
94: With this structural information of the ribosome at the atomic level,
95: we investigate the self-similar property of the ribosome structure and
96: its biological implication. We especially focus on the scale
97: invariance, by estimating the mass fractal dimension, for structures of
98: {\it Thermus thermophilus} 30S subunit including the 16S rRNA, and
99: {\it Haloarcula marismortui} 50S subunit including the 23S
100: rRNA~\cite{data}. It has been known that the protein synthesis occurs
101: in the context of the intact ribosome and the moving part of the
102: ribosome enables the dynamic process of the translation. In this
103: sense, the ribosome function is closely related to its spatial
104: conformation in the physiological medium. Thus, the mass fractal
105: dimension analysis may help to reveal any characteristics of the ribosome,
106: especially the dynamics of the ribosome.
107:
108: The mass fractal dimension, which can be used as a measure of the compactness,
109: is defined as the number of monomers (atoms in our case), $N$,
110: enclosed in a sphere of radius $R$. When a molecule has a fractal
111: structure, it is expected that
112: \begin{equation}
113: N \propto R^{D_{M}}~,
114: \end{equation}
115: where $D_{M}$ is the mass fractal dimension. It can be estimated by
116: plotting the number of all atoms contained inside concentric spheres
117: of varying radius $R$ on a log-log scale.
118:
119: To test the sensitivity of the result to the choice of the origin, we
120: set the origin at the geometric center of each molecule,
121: and vary the origin by $\pm 2~\mathring{A}$ in the XYZ directions
122: (thus 27 different origins) with respect to the coordinate system
123: adopted from the PDB data. We ignore hydrogen atoms since the X-ray crystal
124: structures do not contain the geometric information of hydrogen atoms,
125: which cannot be seen but in very high resolution. Incidentally, this
126: is also true for the protein case. Thus, most of descriptions of the
127: ribosome focus on the position of the heavy atoms, such as C, N, O,
128: and P. Note also that the PDB for the 30S
129: subunit does not contain water molecules; where as that for the 50S
130: does. For a fair comparison, we exclude water molecules from the mass
131: fractal calculation.
132: Due to the finite-size effect, there are both upper and lower size
133: limits beyond which a macromolecule is no longer fractal.
134:
135: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
136: \begin{table}
137: \begin{ruledtabular}
138: \caption{The number of atoms, the average, and its standard deviation
139: of the mass fractal dimension for each molecule. For the 16S rRNA
140: and the 30S subunit, the scaling property emerges for all 27 measurements;
141: while for the 23S rRNA and the 50S subunit, respectively 17 and 19
142: out of 27 measurements show scaling properties. The parentheses are
143: the standard deviation of the corresponding number of estimations.}
144: \setlength{\extrarowheight}{4pt}
145: \begin{tabular}{c c c}
146: molecule & number of atoms & $D_{M}$ \\ \hline
147: 16S rRNA & 32514 & 2.58 (0.06) \\
148: 30S subunit & 51742 & 2.82 (0.07) \\
149: 23S rRNA & 59017 & 3.11 (0.07) \\
150: 50S subunit & 62673 & 3.07 (0.08) \\
151: \end{tabular}
152: \end{ruledtabular}
153: \end{table}
154: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
155:
156: We estimate the mass fractal dimension for the 16S rRNA, the 30S subunit,
157: the 23S rRNA, and the 50S subunit with 27 different origins. The
158: result is summarized in Table~1, and Fig.~1 and 2 present a typical
159: log-log plot of the enclosed ``mass'' $N$ as a function of radius $R$.
160: Note that the geometric origin and the center of mass for all molecules
161: are almost identical, and evaluating the corresponding physical
162: mass rather than the number of atoms does not affect the result.
163: For the 23S rRNA and the 30S subunit, we are able to estimate the
164: fractal dimensions for all 27 cases; while for the 23S rRNA and the
165: 50S subunit, respectively 17 and 19 cases out of 27 shows the scaling
166: behavior.
167:
168: The mass fractal dimension exhibits the molecule's space-filling
169: ability: the larger is $D_{M}$, the more atoms are in the sphere. When
170: fractal dimension is less than 3, the structure has ``empty'' or
171: ``void'' space. From the above result, we see that the mass fractal
172: dimensions for both the 23S rRNA and the 50S subunit are close to 3,
173: implying that these are compact three-dimensional collapsed
174: objects. On the other hand, the average $D_{M}$ for the 16S rRNA and
175: the 30S subunit are found to be 2.58 and 2.82, respectively, smaller
176: than that of a completely compact three-dimensional collapsed polymer.
177: This indicates that the mass inside a radius $R$ does not increase with
178: the Euclidean dimension as an exponent but with some lesser
179: power. Thus, we find that both the 23S rRNA and 50S subunit are more
180: compact than either the 16S rRNA or the 30S subunit.
181:
182: The fact that the 16S rRNA and the 30S subunit have the mass fractal
183: dimension less than 3 leaves a room for the 16S rRNA to
184: make any movement during the protein synthesis, and can be related to
185: the rigid body motion of domains within the subunit.
186: It has been found from many studies that it is the 30S subunit that
187: makes a motion in translocation. It was reported that the
188: 30S subunit makes a ratchet-like rotations relative to the large
189: 50S subunit~\cite{schuwirth,frank}, particularly, the
190: rotational rigid body motions of the ``head'' domain~\cite{domain}
191: within the 30S subunit~\cite{schuwirth}. This reveals a high degree of
192: flexibility between the head and the rest of the 30S subunit.
193: Furthermore, not only domains for the entrance channel such as the
194: ``shoulder'' and the head domains in the 30S subunit are dynamic
195: during decoding, but the exit channel formed between the
196: ``platform'' and the head domains are known to be
197: variable~\cite{schuwirth,frank,spirin,serdyuk}. Thus, it is the 30S
198: subunit that is either dynamic or variable, playing an active role,
199: which is possible due to the fractal characteristics of the 30S subunit.
200:
201: The 50S subunit, on the other hand, cannot make movements due
202: to a highly compact structure except peripheral regions. As an
203: example, the L1 stalk, located in a peripheral region of the 50S
204: subunit, makes a bifurcation~\cite{frank} and moves toward the
205: inter-subunit space playing a pivotal role in the transcription
206: process~\cite{valle}.
207:
208: In this paper, we investigated a symmetry embedded in the ribosome structure
209: under a change of the length scale via the mass fractal dimension. We
210: found the 30S and the 50S subunit (also the 16S and the 23S rRNAs)
211: differs in their fractal dimensions: the 30S subunit and the 16S rRNA
212: have fractal dimensions, while the 50S subunit and the 23S rRNA can be
213: regarded as three-dimensional compact molecules. The fractality of
214: both the 16S rRNA and the 30S subunit supports the dynamic nature of
215: the ribosome in the protein synthesis.
216:
217: Although the power of the self-similarity approach to the ribosome
218: structure is in its simplicity and generality, it also true that their
219: detail dynamic properties and realization are not obvious because
220: detailed properties of the ribosome which determine their function
221: are averaged out. Nevertheless, the fractal property of the 30S
222: subunit (also the 16S rRNA) provides a partial, if not whole, evidence of
223: its movement during the transcription.
224:
225: This work was supported by the Korea Research Foundation Grant funded
226: by the Korean Government (MOEHRD) (KRF-2005-041-H00052).
227: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
228: \begin{thebibliography}{99}
229: \bibitem{ribosome} For a general reference of the ribosome and its
230: function, see, for example, {\it Protein Synthesis and Ribosome Structure:
231: Translating the Genome}, edited by Knud H. Nierhaus and Daniel
232: N. Wilson (Wiley-VCH, Weinheim, 2004); {\it The Ribosome: Structure,
233: Function, Antibiotics, and Cellular Interactions}, edited by R. Garrett,
234: S. Douthwaite, A. Liljas, A. Matheson, P. Moore, and H. Noller (ASM
235: Press, Washington, DC, 2000).
236: \bibitem{function} H. Noller, Science {\bf 309}, 1508 (2005);
237: P. Nissen, J. Hansen, N. Ban, P. B. Moore, T. A. Steitz, Science {\bf
238: 289}, 920 (2000).
239: \bibitem{moret} M. A. Moret, J. G. V. Miranda, E. Nogueira, Jr., M. C. Santana,
240: and G. F. Zebende, Phys. Rev. E {\bf 71}, 012901 (2005).
241: \bibitem{elber} R. Elber, in {\it Fractal analysis of protein in The Fractal
242: Approach to Heterogeneous Chemistry}, edited by D. Avnir (John Wiley $\&$
243: Sons, New York, 1989), p. 407.
244: \bibitem{enright} Matthew B. Enright and David M. Leitner,
245: Phys. Rev. E {\bf 71}, 011912 (2005); X. Yu and D. M. Leitner,
246: J. Chem. Phys. {\bf 119}, 12673 (2003).
247: \bibitem{ban} N. Ban, P. Nissen, J. Hansen, P. Moore, T. Steitz,
248: Science {\bf 289}, 905 (2000).
249: \bibitem{wimberly} B. Wimberly, D. Brodersen, W. Claemons,
250: R. Morgan-Warren, A. Carter, C. Vonhein, T. Hartsch, and
251: V. Ramakrishnan, Nature {\bf 407}, 327 (2000).
252: \bibitem{schuwirth} B. Schuwirth, M. Borovinskaya, C. Hau, W. Zhang,
253: A. Vila-Sanjurjo, J. Holton, J. Cate, Science {\bf 310}, 827 (2005).
254: \bibitem{harms} J. Harms, F. Schluenzen, R. Zarivach, A. Bashan,
255: S. Gat, I. Agmon, H. Bartels, F. Franceschi, and A. Yonath, Cell
256: {\bf 107}, 679 (2001).
257: \bibitem{schluenzen} F. Schluenzen, A. Tocilj, R. Zarivach, J. Harms,
258: M. Gluehmann, D. Janell, A. Bashan, H. Bartels, I. Agmon,
259: F. Franceschi, and A. Yonath, Cell {\bf 102}, 615 (2000).
260: \bibitem{cate} J. H. Cate, M. M. Yusupov, G. Z. Yusupova,
261: T. N. Earnest, H. F. Noller, Science {\bf 285}, 2095 (1999).
262: \bibitem{yusupov} M. Yusupov {\it et al}, Science {\bf 292}, 883 (2001).
263: \bibitem{ramakrishnan} V. Ramakrishnan, Cell {\bf 108}, 557 (2002).
264: \bibitem{data} These are the highest resolution results for each
265: subunit. Anything worse than about 3.5 $\mathring{A}$ would normally
266: not be possible to construct an accurate model of a
267: macromolecule. We also exclude the 5S subunit from the analysis
268: because it is too small to perform any statistical analysis. The
269: structure information for the subunits can be found in the Protein
270: Data Bank (PDB) at \url{http://www.rcsb.org/pdb/}. The access
271: numbers are 1J5E for {\it Thermus thermophilus} 30S subunit
272: including the 16S rRNA, and 1JJ2 for {\it Haloarcula marismortui}
273: 50S including the 23S rRNA.
274: \bibitem{frank} J. Frank and R. Agrawal, Nature {\bf 406}, 318 (2000).
275: \bibitem{domain} The structure of the 16S rRNA is commonly organized
276: into four domains of a few hundred nucleotides each: the head, the body, and
277: the platform, and $3^{\prime}$ minor domains.
278: \bibitem{spirin} A. Spirin, V. Baranov, G. Polubesov, I. Serdyuk, and
279: R. May, J. Mol. Biol. {\bf 194}, 119 (1987).
280: \bibitem{serdyuk} I. Serdyuk, V. Baranov, T. Tsalkova, D. Gulyamova,
281: M. Pavlov, A. Spirin, and R. May, Biochimie {\bf 74}, 299-306 (1992).
282: \bibitem{valle} M. Valle, A. Zavialov, J. Sengupta, U. Rawat,
283: M. Ehrenberg, and J. Frank, Cell {\bf 114}, 123 (2003).
284: \end{thebibliography}
285:
286: \newpage
287: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
288: \begin{figure}
289: \centerline{
290: \includegraphics{fig1}
291: }
292: \caption{A typical log-log plot of the number of atoms versus the
293: radius for the 16S rRNA ($\blacksquare$) and the 30S subunit
294: ($\Box$). The least-square fits on the slope of the 16S rRNA (solid
295: line) and the 30S subunit (dotted line) yield 2.62 and 2.85,
296: respectively. The plot for the 30S subunit is shifted to the
297: right by the factor of 10 for the display purpose. }
298: \end{figure}
299: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
300: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
301: \begin{figure}
302: \centerline{
303: \includegraphics{fig2}
304: }
305: \caption{A typical log-log plot of the number of atoms versus the
306: radius for the 23S rRNA ($\blacksquare$) and the 50S subunit
307: ($\Box$). The least-square fits on the slope of the 23S rRNA (solid
308: line) and the 50S subunit (dotted line) yield 3.01 and 3.06,
309: respectively. The plot for the 50S subunit is shifted to the
310: right by the factor of 10 for the display purpose.}
311: \end{figure}
312: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
313: \end{document}
314: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
315: