Schlick T. Molecular Modeling and Simulation: An Interdisciplinary Guide

Подождите немного. Документ загружается.

130 5. Nucleic Acids Structure Minitutorial

DNA plays a role in life rather like that played by the telephone direc-

tory in the social life of London: you can’t do anything much without

it, but, having it, you need a lot of other things — telephones, wires,

andsoon—aswell.

Review of The Double Helix in The Sunday Times, London (1968).

5.1 DNA, Life’s Blueprint

With the master molecule of heredity, deoxyribonucleic acid (or DNA), so fre-

quently mentioned in the media — in connections ranging from polymerase chain

reaction (PCR) applications (e.g., criminology, medical diagnoses) to cloning

and art [1112] — it is difﬁcult to imagine today that it was only a few decades

ago when Watson and Crick reported their description of the DNA double helix

[1354–1356]! Based on analysis of DNA ﬁber diffraction patterns and Chargaff’s

rules, they described a spiral image of an orderly helix — two intertwined polynu-

cleotide chains, with a sugar/phosphate backbone on the exterior and pairs of

hydrogen-bonded nitrogenous bases in the center. See [622] for a historical per-

spective of this discovery, including the contribution of all key players (a capsule

of which is given in Chapter 1), and anniversary issues of DNA, for example is-

sued in 2003 in many journals (e.g., Nature Vol. 421 and Science Vol. 300) at the

occasion of DNA’s golden anniversary.

Though the model was imperfect in several respects (e.g., sugar conforma-

tions and base positioning with respect to the helical axis were inexact), Francis

Crick, James Watson, and Maurice Wilkins received the 1962 Nobel Prize in

Chemistry for this monumental discovery. Indeed, soon after the description of the

DNA double helix came a burst of scientiﬁc excitement and discovery regarding

the mechanism of heredity. The ﬁeld of molecular biology gained extraordinary

momentum and, soon after, its tentacles extended to the disciplines of cellular bi-

ology, molecular genetics, genomics, and proteomics, all buttressed by impressive

advances in technology and computing.

5.1.1 The Kindled Field of Molecular Biology

The description of DNA’s three-dimensional (3D) structure provides a classic ex-

ample of the structure/function relationship. The two key functions of DNA —

replication and transcription — were suggested and later proven based on the or-

dered spatial arrangement of the DNA. Replication, or the accurate transmittal of

genetic information from parent to progeny, was explained by base pairing speci-

ﬁcity. Transcription, the transformation of DNA’s genetic information into a form

that directs protein synthesis, was later understood with the deciphering of the

5.1. DNA, Life’s Blueprint 131

genetic code relating triplet nucleotide codons (see Section 5.2 and Figure 5.2 for

a description of the basic building blocks, including the pyrimidine bases T and

C and purine bases A and G) to amino acids.

This genetic code (Table 5.1) reveals interesting patterns:

1. It is highly redundant: as 61 of the 64 triplet codons represent one of the

20 amino acids, and 3 signal ‘stop’ messages for protein synthesis (i.e.,

translation). In particular, Arg, Leu, and Ser are speciﬁed by 6 codons; Met

and Trp are each represented by only one; and the rest are speciﬁed by 2–4

triplets.

2. Pyrimidine-end commonalities exist: Codons B

TandB

C(where

and B

denote any of the 4 bases) always code for the same amino acid.

3. Purine-end commonalities exist: Codons B

AandB

G code for the

same amino acid except in the cases of AT (ATA = Ile, ATG = Met) and

TG (TGA = ‘

STOP’, TGG = Trp).

Though the genetic code was initially thought to be ‘universal’, DNA se-

quencing in the early 1980s revealed code variations for certain organisms.

Now it is clear that the genetic codes for mammalian and plant mitochondria

and certain protozoa, for instance, exhibit some variations from the code de-

scribed in Table 5.1. In organisms where nonstandard amino acids exist (e.g.,

certain Archaea and eubacteria), such a variation in the code may consist of a

STOP codon that encodes a nonstandard amino acid rather than instruct to halt

translation [68,511,1217].

Recent advances in experimental techniques that allow detailed study of the

components of the translation apparatus are also making possible the study of

how the modern code has evolved from a simpler form [662].

Thus the model for the DNA double helix led the way to establishment of

the ﬁrst iteration of the central dogma of biology: DNA is self-replicating, DNA

makes messenger ribonucleic acid (mRNA) through transcription,

and mRNA

makes protein through translation (DNA → RNA →protein).

Although other ar-

rows in the dogma have been suggested or veriﬁed (e.g., RNA → RNA in certain

viruses, RNA → DNA in retroviruses), the main components appear known.

In transcription, an RNA polymerase glides along one strand of the DNA double helix and

builds an RNA complement by coupling ribonucleotides through dehydration synthesis. The faithful

replicate of one DNA strand then functions as the messenger RNA.

In translation, the genetic code of the messenger RNA is read by transfer RNA molecules on

cellular structures called ribosomes. Every transfer RNA molecule carries a speciﬁc sequence of three

nucleotides on one end of an L-shaped structure and the corresponding amino acid on the other. The

transfer RNA’s main task is to deposit its amino acids on the ribosomes in proper sequence. In this

process of matching each messenger-RNA codon with the complementary transfer RNA molecule, the

polypeptide chain is assembled. As the amino acids link to one another on the ribosomes, polypeptide

folding is thought to begin.

132 5. Nucleic Acids Structure Minitutorial

Table 5.1. The genetic code in terms of the parental DNA; the mRNA transcript has uracil

(U) instead of thymine (T).

Amino Acid Encoding Triplets

(or Instruction) in Parental DNA

Arginine (Arg) CG(*), AG(A,G)

Leucine (Leu) CT(*), TT(A,G),

Serine (Ser) TC(*), AG(T,C)

Alanine (Ala) GC(*)

Glycine (Gly) GG(*)

Proline (Pro) CC(*)

Threonine (Thr) AC(*)

Valine (Val) GT(*)

Isoleucine (Ile) AT(T,C,A)

Asparagine (Asp) GA(T,C)

Aspartic acid (Asn) AA(T,C)

Cysteine (Cys) TG(T,C)

Histidine (His) CA(T,C)

Phenylalanine (Phe) TT(T,C)

Tyrosine (Tyr) TA(T,C)

Glutamine (Gln) CA(A,G)

Glutamic acid (Glu) GA(A,G)

Lysine (Lys) AA(A,G)

Methionine (Met) ATG

Tryptophan (Trp) TGG

STOP TA(A,G), TGA

Short-hand notation is used for the third base of the codon when the ﬁrst and second bases are

the same: GA(T,C) speciﬁes both GAT and GAC, while GC(*) denotes all four possibilities, namely

GCT, GCC, GCA, and GCG.

The codons ATG and (less frequently) GTG form part of the initiation signal in addition to their

coding for Met and Val, respectively.

Much attention has now expanded to new areas of molecular biology, such as

structural molecular biology, computational molecular biology, and offspring of

genomics, such as structural genomics and functional genomics. The functional

relationships among genes and the interaction of genes within the context of the

complete organism are also of great interest.

5.1.2 Fundamental DNA Processes

As the genetic material, DNA carries structural information in the primary se-

quence that not only controls faithful duplication but also regulates expression

5.1. DNA, Life’s Blueprint 133

of the hereditary information [1021]. Genetic variability can result from errors

(or mutations), such as insertions, deletions, or substitutions in the daughters com-

pared to the parental DNA, during the template-copying process; these changes

can lead to altered triplet codes and hence different polypeptide composition.

Since many basic biological processes rely on protein/nucleic-acidinteractions,

the base sequence of polynucleotides also affects profoundly the characteristic 3D

structure of DNA and RNA and hence the nature of fundamental biological pro-

cesses. On a higher level of structure (hundreds and thousands of base pairs),

compact forms of long DNA — such as supercoils, knots, and chromosomes —

are central to fundamental mechanisms for replication, transcription, and recom-

bination [195,1384]. DNA supercoiling, namely the coiling in space of the double

helix axis itself, can condense the DNA by several orders of magnitude and read-

ily store energy for various activities. This supercoiled state is essential for the

storage of DNA in eukaryotic cells, where DNA is wrapped as a left-handed su-

percoil around cores of proteins to form the chromatin ﬁber. The wide range of

characteristic timescales associated with the conﬁgurational rearrangements and

hydrodynamicproperties of supercoiled (or superhelical) DNA represents another

area of intense study.

5.1.3 Challenges in Nucleic Acid Structure

As described above, there are several levels of nucleic acid structure, from the

base-pair level to the cellular level of organization in the chromosomes (see end

of next chapter and [1119] for example). This study of DNA’s rich conﬁgurational

levels is particularly challenging to modelers.

Much focus has been placed on protein structure (including protein/DNA com-

plexes) and the protein folding problem — predicting the 3D architecture of a

system from the primary sequence. Yet an analogous folding problem is also rele-

vant to DNA and RNA polymers. In fact, the nucleic-acid analogue of the ‘protein

folding problem’ might be viewed as more challenging than protein structure pre-

diction in the sense that it extends over much larger spatial scales (thousands of

Angstroms) as well as temporal scales (picoseconds to minutes and longer) than

typically associated with proteins. Biologically, elucidating the folding of DNA

in the cell — supercoiling and chromosome condensation, for example — is im-

portant for understanding the regulatory role of DNA in fundamental biological

processes.

For small segments of nucleic acids, X-ray crystallography and NMR data have

been invaluable for providing detailed, atomic-resolution data on single, double,

triple and other forms of nucleic acid oligonucleotides, as well as their complexes

with proteins, other biomolecules, and ligands [126]. The Nucleic Acid Data-

base (NDB, ndbserver.rutgers.edu/)[127] has been beautifully cataloging these

structures and offers many utilities for their analysis.

All-atom molecular dynamics (MD) simulations of nucleic acids have also

shed important insights on DNA sequence/structure relationships, the nature of

the hydration geometries surrounding nucleic acids, nucleic acid ﬂexibility, and

134 5. Nucleic Acids Structure Minitutorial

nucleic-acid protein structures [126, 217, 808, 953, 1026, 1417, 1419, 1421,for

example]. Indeed, simulation improvements and the availability of ultra-high

resolution nucleic-acid crystal structures [234,641,1182] have made possible the

study of fully solvated nucleic acid MD trajectories, with representative ionic

atmospheres [219,239,283,1412, 1416], and even millisecond simulations [987].

The theoretical advances resulted from improvements in long-range electro-

static modeling, force ﬁeld parameters, representation of the ionic atmosphere,

and novel conformational sampling approaches [217,953,988,1116,1117,1319],

as well as increases in computer memory and speed. The experimental advances

reﬂect improved methods for crystallization and phase determination, the in-

creased availability of very strong X-ray sources, improvements in algorithms

for model reﬁnement, and innovative approaches such as single-force microscopy

which allows studies of DNA and RNA energetics and dynamics of folding and

unfolding (e.g., [187, 759]) and improved ultra-structural visualization tools for

DNA’s higher levels of structural organization (e.g., chromatin, see next chapter).

Signiﬁcantly, such modeling and experimental advances have made it possible to

simulate solvated RNA at atomic resolution [86, 502, 525, 526, 1446]. Advances

on both the computational and experimental fronts are ongoing.

Computer scientists also ﬁnd interest in DNA with the emerging possibility of

using the strands of DNA in vitro for practical applications, such as to produce

electronic devices like nanowires, or as parallel computers to solve very difﬁcult,

combinatorial optimization problems that have non-polynomial complexity [40,

117,162, 825, 1156]. Very recently, a ‘DNA Sudoku’ approach for parallel DNA

sequencing by combinatorial pooling strategies reminiscent of solving sudoku

puzzles has also been reported for analyzing short genome segments associated

with disease markers [366].

Undoubtedly, progress is expected in the bridging between all-atom and macro-

scopic representations of nucleic acids and between experiment and theory. This

unity will enhance our understanding of DNA structure and DNA/protein interac-

tions and, in turn, will likely have important biomedical applications, for example,

in the design of new anti-viral drugs, antibiotics, and anti-cancer agents, some of

which are being designed with DNA or RNA agents such as DNA plasmids and

silencing RNAs.

5.1.4 Chapter Overview

In this chapter, the basic elements of nucleic acids and DNA structure on the base-

pair and helical level will be introduced: the fundamental building blocks and how

they are linked to form polynucleotides, aspects of nucleic acid conformational

ﬂexibility (sugar pseudorotation, torsion angle preferences, and global and local

base-pair parameters), and the three canonical DNA helices (A, B, and Z-DNA).

The next chapter presents further topics regarding the structural diversity of DNA

and RNA, and DNA folding on a higher level, namely supercoiling and chromatin

organization.

5.2. The Basic Building Blocks of Nucleic Acids 135

While RNA is introduced in the next chapter, after the variety of base pairing

arrangements is presented, this in no way means RNA is less fundamentally im-

portant than DNA. In fact, recent discoveries on RNA’s prevalent regulatory roles

in the cell place RNA in the forefront of biology and chemistry.

For basic books on DNA structure, see [100,139,195,269,893,895,1080,1191,

1305]. Lively, less technical introductions can be enjoyed from [272, 420, 1353].

The reader is also well advised to examine some of the rich information on

nucleic acid structure available in the public domain through the NDB [127],

ndbserver.rutgers.edu/.

Throughout this chapter and the next, we abbreviate a base pair as bp and

base pairs as bps. Note that though we mention the three canonical helices be-

fore introducing them formally (using all notation we systematically develop),

novices should be able to follow the material. Students are encouraged to re-read

the chapter after they are more familiar with all the presented material.

5.2 The Basic Building Blocks of Nucleic Acids

In the classic DNA double helix described by Watson and Crick, a ﬂexible ladder-

like structure is formed with the polymer wrapped around an imaginary central

axis (Figure 5.1). The two rails of the ladder consist of alternating sugar (deoxyri-

bose) and phosphate units; the rungs of the ladder consist of nitrogenous bps held

together by hydrogen bonds (see Box 3.2 of Chapter 3 for a deﬁnition).

5.2.1 Nitrogenous Bases

Four nitrogenous bases can be found in DNA: the pyrimidines cytosine (C) and

thymine (T) which are 6-membered rings, and the purines guanine (G) and ade-

nine (A), each of which is a fused system with a 5 and 6-membered ring. In

the single-stranded RNA polynucleotide, ribose replaces deoxyribose and uracil

(U) replaces thymine (Figure 5.2). Some DNAs contain bases that are chemically

modiﬁed (e.g., substitutions of a hydrogen by a methyl group).

Alternative but

rare tautomeric forms due to proton shifts in aromatic molecules (as introduced in

Chapter 1, subsection on the discovery of DNA structure),

are possible but may

not be biologically signiﬁcant.

For example, N6-methyl-dA is a modiﬁed adenine base with the N6H

(attachment to the ring

carbon C6) moiety replaced by N6HCH

; 5-methyl-dC is a modiﬁed cytosine where the C5H becomes

C5CH

Two classic tautomerization reactions are keto/enol and amino/imino; the keto and amino forms

are typically favored and are shown in Figure 5.2. Keto/enol tautomerization involves alteration of the

carbonyl group (–C=O) to a hydroxyl group (=C–O–H), shifting the double bond from the carbonyl

group to the nitrogen-carbon bond in the ring (e.g., C6=O of G becomes C6–OH, accompanied by the

change of H–N1–C6 to N1=C6). Similarly, an amino/imino tautomerization involves a change in an

amino nitrogen –NH

to an imino form, =NH (e.g., C6–NH

of A becomes C6=NH, accompanied

by the change of N1=C6 to H–N1–C6).

136 5. Nucleic Acids Structure Minitutorial

Figure 5.1. The two strands of the classic DNA double helix: alternating subunits of phos-

phates and deoxyriboses held together by hydrogen-bonded bps, normally with adenine

(A) pairing with thymine (T) and guanine (G) pairing with cytosine (C). The two chains of

the DNA double helix run in opposite directions. The chain ending in the lower right ter-

minates with a phosphate group linked to the sugar carbon denoted as C5



, while the chain

ending in the lower left terminates with a free hydroxyl group linked to the sugar carbon

denoted as C3



. Individual bases are numbered 1–16 (or 1 to n where n is the number of

bps) for the sequence strand (Strand I) and 17–32 (n +1to 2n in general) for Strand II,

both in the 5



→ 3



direction; the n base pairs 1–16 are numbered so that they coincide

with the bases of Strand I. The pitch of the helix P

is the length along the helix axis for

one complete turn, and n

is the number of bps per turn (around 10.5 for DNA in solution).

The unit twist is deﬁned as Ω = 360

, and the helical rise is h = P

. Mgr and mgr

denote the major and minor grooves, respectively.

5.2.2 Hydrogen Bonds

In the Watson-Crick base pairing scheme (termed WC herewith), cytosine pairs

with guanine by forming three hydrogen bonds, and thymine pairs with adenine

by forming two hydrogen bonds (Figure 5.3). This arrangement produces CG and

TA bps whose widths are nearly identical. The approximately uniform width is

signiﬁcant because any pyrimidine/purine or purine/pyrimidine sequence can be

accumulated on one strand, with a pyrimidine opposing a purine, without much

alteration in overall structure. The discovery of this orderly 3D structure of DNA

5.2. The Basic Building Blocks of Nucleic Acids 137

O3'H

2'3'

β−D-2-Deoxyribose

Ribose

Thymine

Cytosine

Uracil

Adenine

Guanine

Sugars Pyrimidine Bases

Purine Bases

C5M

C5'

O5'H

O2'H

O3'H

2'3'

C5'

O5'H

Figure 5.2. Chemical structures and atomic labels for nucleic acid sugars (deoxyribose in

DNA and ribose in RNA) and nitrogenous bases: T, C, A, G in DNA and U, C, A, G in

RNA. The broken-line pattern in each compound indicates the bond where a link to another

building block is made (see Figure 5.4).

helices explained for the ﬁrst time how a regular polymer made of repeating

phosphate/sugar/base units (nucleotides) could replicate and encode hereditary

information.

5.2.3 Nucleotides

The basic building block of nucleic acid polymers, the nucleotide, consists of a

5-membered sugar ring — deoxyribose in DNA and ribose in RNA — a phos-

phate, and a purine or pyrimidine base (see Figure 5.4). The unit containing

just the sugar and base is called the nucleoside. Nucleotides are linked together

through the phosphate group to form the polynucleotide chain.

5.2.4 Polynucleotides

When nucleotides are polymerized into nucleic acid chains by the chemi-

cal removal of water molecules, a sugar/phosphate backbone is formed. The

138 5. Nucleic Acids Structure Minitutorial

Sugar

Watson - Crick Base Pairs

Sugar

. . .

Figure 5.3. In the classic Watson-Crick base pairing scheme of DNA double helices,

thymine (T) pairs with adenine (A) by forming two hydrogen bonds, and cytosine (C)

pairs with guanine (G) by forming three hydrogen bonds. The hydrogen bonds are often

represented by dots (···). The structural ﬁt (e.g., as measured by C1



–C1



distances)

is such that the difference in width between the two base pairs is less than 3%, allowing

formation of a double helix with a nearly constant diameter.



–hydroxyl group of the nth nucleotide sugar is joined to the C5



–hydroxyl

group of the (n+1)th nucleotide by a phosphodiester bridge.

This negative phos-

phate group, located on the exterior of the helix, is readily available for physical

and chemical interactions with solvent water molecules and metal ions present

in the cell (see next chapter). Hydration effects are of great importance for DNA

molecules in solution, since water plays a central role in stabilizing a particular

helical conformation.

An ester is an –OR group where R represents an organic chemical group.

5.2. The Basic Building Blocks of Nucleic Acids 139

O2P

O1P

Base

O2P

Base

Nucleotide Unit

C4⬘

O4⬘

HO−C5⬘ − H

5⬘− end

3⬘− end

O1P

O4⬘

O2P

O1P

Base

N9/N1

C3⬘

C5⬘

C4⬘

C3⬘

C2⬘

C1⬘

O3⬘

O5⬘

C4⬘

C3⬘

C2⬘

C1⬘

O3⬘

O3⭈

C5⬘

O5⬘

O4⬘

O3⬘

C2⬘

C1⬘

Figure 5.4. The polynucleotide chain, with standard atom labeling shown, runs from the



end (of the sugar atom C5



)tothe3



end (of the sugar atom C3



). Nucleotide linkage

is via the 3



to 5



phosphodiester bonds (i.e., C3



–hydroxyl group of one nucleotide sugar

to C5



–hydroxyl group of another). Nucleic acid torsion angles are labeled as α, β, γ, δ, 

and ζ along the polynucleotide chain, τ

through τ

about the endocyclic bonds of the

sugar, and χ about the glycosyl bond, connecting the sugar and base. See Table 5.2 for full

quadruplet sequences of these torsion angles.