Schlick T. Molecular Modeling and Simulation: An Interdisciplinary Guide

Подождите немного. Документ загружается.

210 7. Topics in Nucleic Acids Structure: Noncanonical Helices and RNA Structure

is shifted in the bp plane relative to its partner due to steric misalignment of the

bases. Experimental evidence for bp wobbling has come from tRNA structural

analysis.

Thus, various non-WC base-pairing schemes can occur when bases are chemi-

cally modiﬁed, found in the rare enol and imino tautomeric forms, or mismatched,

for example. The alternative arrangements for mispaired bases, in particular, can

in turn lead to mutations upon DNA replication if not corrected by the elaborate

repair machinery of the cell.

Other Patterns

For further details regarding these and many other base-pairing schemes in nu-

cleic acids, see [1191]and[737] for an early encyclopedic compendium of

the “bewildering” variety of observed canonical and non-canonical nucleic acid

arrangements, as compiled from RNA base-pairing interactions. This variety

led to many further works and a proposal for new nomenclature for RNAs

[738] which stimulated a large international initiative termed the RNA Ontol-

ogy Consortium (ROC) [735]. ROC aims to establish a standard vocabulary for

studying RNA by creating a common system to describe RNA sequences, 2D, and

3D structures and by developingsoftware tools that merge sequence and structural

databases for RNA for use by the general scientiﬁc community. See ROC website

(http://roc.bgsu.edu/) for details.

In general, non WC bps are generally lower in stability because they typically

incur a greater distortion in the sugar/phosphate backbone (assuming that the rest

of the DNA is in a canonical A, B, Z-type conformation). Though they are more

difﬁcult to accommodate in general, many local alterations in sequence compo-

sition, environmental factors, and conformational patterns favor these alternative

base-pairing arrangements.

7.2.2 Hybrid Helical/Nonhelical Forms

Alternative Helical Geometries

Numerous variations are noted in the helical structures of polynucleotides (see

[139, 893, 1080], for example). For example, C-DNA, D-DNA, and T-DNA

helices have been deﬁned with 9.3, 8.5, and 8.0 bps/turn, respectively. The

C-DNA motif forms in ﬁbers of low humidity (57–66%), D-DNA is observed

in poly(dA)·poly(dT) strands, and T-DNA has been noted in bacteriophages that

contain modiﬁed C residues and glucose derivatives for the sugar attachments.

Within polynucleotide structures, local distortions can also produce hairpins

(DNA and RNA strands that fold back on themselves), cruciforms (intrastrand

hydrogen bonds between complementary bases, often leaving a few unpaired

bases at the hairpin tip), inlcuding bulges and loops (unpaired extrusions)

(Figure 7.4). Such motifs are especially important in stabilizing the folding of

single-stranded RNA.

7.2. Variations on a Theme 211

DNA Triplexes and Quadruplexes

Though Linus Pauling was ridiculed for his early suggestion of a triple helix

structure for DNA [975], we now recognize the existence or the possibility for for-

mation of various DNA triplexes (involving WC and Hoogsteen bps) [421,1008].

Other existing or possible forms are quadruplexes (stabilized by hydrogen bond-

ing and cations as in guanine-quartets; see Figure 7.1)[961,968,998,1160,1183,

1319,1320,1337]; parallel DNA (requiring reverse WC base pairing); DNA ana-

logues; and hybrids of RNA, DNA, and other polymers such as oligonucleotide

mimics (see below) [712].

Triplex forming oligonucleotides are promising pharmaceutical targets since

they modulate gene activity in vivo through recognition of duplex DNA [905].

Triplexes and other unusual oligonucleotides are good subjects for simulation

techniques since the factors that govern sequence-dependent properties can be

explored [784].

Triple-helical DNA gained wide attention in the mid-1980s when various

groups demonstrated that homopyrimidine (polynucleotide chains containing

pyrimidines, such as poly(dT) or poly(dCT)) and some purine-rich oligomers

(such as poly(dA) or poly(dAG)) can form stable and sequence-speciﬁc triple-

stranded complexes with corresponding sites on duplex DNA; the third strand

associates with the duplex via Hoogsteen pairing (see Figure 7.1)[421].

Homopurine/homopyrimidine stretches in supercoiled DNA were also found to

adopt an unusual structure which includes a triplex called H-DNA.

Four-stranded DNA structures called G-quadruplexes are believed to play im-

portant functional roles in genome maintenance through their stabilization of

chromosome ends or telomeres. Likened to plastic caps ﬁrming shoe-lace ends

(by Stu Borman, C&EN writer), telomeres are protein/DNA assemblies that con-

ﬁne chromosome ends. Such protection from the possible fusing of chromosomes

into one another helps maintain the integrity of the genome in living cells.

Telomere structure is of great biomedical importance since aging and cancer

can be associated with malfunctioning telomeres. Thus, drugs targeted to

telomeres can affect telomere function in DNA synthesis and transcription.

Telomeric DNA contains short guanine-rich repeats. At the very end of telom-

eres, G-quadruplexes are found, elaborately folded stacks of G-quartets (also

known as GGGG tetrads) — the planar arrays of hydrogen-bonded guanines as

shown in Figure 7.1.

Direct evidence that quadruplexes exist in human cells was provided by Hurley

and coworkers [1183], who reported a chair-shaped quadruplex in a purine-rich

strand of DNA in a promoter region of the oncogene c-Myc activated in cancer

cells. Their work also highlights the potential role of quadruplex-targeted com-

pounds as agents to combat disease, for example through control of oncogene

expression to induce tumor shrinkage [601].

The architecture of G-quartets in G-quadruplexes remains an area of intense

study. The planar G-quartets, as shown in Figure 7.1, can stack vertically with

monovalent ions (Na

or K

) sandwiched in between the layers. Wang and

212 7. Topics in Nucleic Acids Structure: Noncanonical Helices and RNA Structure

POO

Base

POS

Base

POO

Base

BaseBase

DNA

Phosphate changed

(phosphorothioate)

Sugar changed

(homo-DNA)

Linkage changed

(3'-thioformacetal)

Backbone changed

(PNA)

Base

O Base

Base

Figure 7.3. Various oligonucleotide analogues that involve modiﬁcations of the DNA

phosphate unit, sugar, entire phosphodiester link, or the entire sugar/phosphate backbone.

Patel [1337] revealed in 1993 an NMR structure of a G-quadruplex in sodium

solution. The sequence studied, with three repeats of TTAGGG, has a topol-

ogy of G-quartets connected by loops (the TTA sequences that join each GGG

series to the next) at the top and bottom of the quadruplex rather than the

sides (see original article for ﬁgures). An exciting 2002 crystal structure by the

Neidle group [961] of the same single-stranded sequence motif (speciﬁcally,

TTAGGG

and TTAGGG

) in the presence of potassium ions reveals a radi-

cally different arrangement, with propeller-like connections through the sides of

the stack (double-chain reversal topology [968]) rather than top and bottom (or

one diagonal and two edgewise loops [968]).

Since mammalian cells are likely to have quadruplexes coordinated by potas-

sium ions (since K

concentration exceeds that of Na

by an order of magnitude),

this propeller-form quadruplex structure is of great importance and has clinical

implications as well [961]. The topologies of quadruplexes and their interactions

with other DNA, RNA, and proteins continue to deﬁne an active research fron-

tier with signiﬁcant potential for biomedical applications, for example in cancer

therapy. See [1319,1320] for recent computational studies.

DNA Mimics

A growing number of DNA analogues (or ‘DNA mimics’) is also appearing (see

Figure 7.3)[905]. Analogues can be sought in several ways (see Figure 7.3 and

Box 7.1):

• by modifying the phosphate backbone segment (e.g., one P–O link changed

to P–S or P–CH

);

• by modifying the sugar (e.g., 6-membered instead of 5-membered ring);

• by modifying the entire phosphodiester linkage of 4 atoms, O3



through



(to increase DNA hybrid stability through charge neutrality); or

• by replacing the entire sugar/phosphate backbone, retaining just the bases

as in PNA, peptide (or polyamide) nucleic acids.

7.2. Variations on a Theme 213

Such close analogues of oligonucleotides are designed to bind selectively to DNA

bps through triple strands or duplexes. Various duplex DNA/RNA hybrids are

key elements in transcription and replication and thus are potential therapeutic

agents in anti-sense techniques. Triplex structures are also potential agents for

sequence-speciﬁc cutting of DNA and drugs (by attacking various disease targets

at the genetic level). Non-complementary or ‘anti-sense’ binding to a sequence of

complementary messenger RNA or single-stranded DNA can also be exploited in

antisense technology to suppress gene expression.

To be suitable, DNA analogues must also be biologically stable (i.e., resistant

to nuclease digestion) and have favorable cellular absorption properties. All issues

of geometry, ﬂexibility, hydrophobicity, and triplex stability must be weighed in

designing DNA mimics with practical utility.

Box 7.1: PNA (Peptide Nucleic Acid)

PNA is an especially interesting oligonucleotide analogue that forms very stable com-

plexes with double-stranded DNA [228, 350, 351, 905, 907, 1161]. In studies of evolution,

PNA has been proposed as a candidate for the backbone of the ﬁrst genetic material

that predates the “RNA world”. While RNA is unstable and difﬁcult to synthesize, the

polymeric constituent of PNA, N-(2-aminoethyl)glycine (or AEG), may spontaneously

polymerize and is accessible in prebiotic syntheses. No ﬁrm evidence, however, exists

regarding this hypothesis [662].

The sugar/phosphate backbone of oligonucleotides is replaced in PNA by a pseudopep-

tide chain of AEG units with the N-acetic acids of the bases (N9 for purines and N1 for

pyrimidines) linked via amide bonds (Figure 7.3). This substitution of the DNA backbone

makes the PNA backbone electrically neutral rather than negatively charged; though the

same number of chemical bonds as found in DNA and RNA exist in PNA (albeit different

types), the replacement of the sugar ring by a linear bond sequence introduces additional

torsional ﬂexibility.

Stable hybrids between PNA and complementary DNA or RNA oligonucleotides form

in a sequence-selective manner. Binding of PNA to double-stranded DNA leads not only

to triple helices but also to P-loops strand-displacement complexes [905]. Model building,

molecular mechanics and dynamics calculations, as well as experimental studies have

shown how the replacement of DNA by PNA disrupts the canonical DNA triple helix, by

displacing the hydrogen-bonded bases away from the global helical axis with respect to

their positions in B-DNA helices, and changes the groove structure substantially [1216].

This makes PNA/DNA hybrids (involving both single and double helical DNA) A-like

rather than an intermediate between A and B-DNA forms, as is typical for canonical DNA

triplexes. Helix stabilization in PNA/DNA hybrids stems from favorable base pairing,

stacking, and (reduced) electrostatic interactions.

PNA molecules have already found practical applications as probes — an alternative

to conventional DNA probes — for the detection of bacteria like Salmonella enterica or

Staphylococcus aureus; chemical advantages stem from PNA’s strong resistance to degra-

dation by proteases and nucleases in the cell and its salt-independent hybridization kinetics.

214 7. Topics in Nucleic Acids Structure: Noncanonical Helices and RNA Structure

Triplexes and DNA analogues like PNA also have potential as drugs in gene therapy [905].

In particular, PNA is quite attractive for drug development, given its high triplex stability.

However, given the A-like form of PNA/DNA hybrids, the design of stable B-like hybrids

requires a different chemical approach than that used to construct PNA.

7.2.3 Unusual Forms: Overstretched and Understretched DNA

Single-Molecule Manipulations

Recently, a new experimental achievement of single-molecule observation and

manipulation experiments using laser tweezers or single glass ﬁbers [187, 1196]

has been applied to DNA [682,1049], re-invigorating interest in DNA’s structural

versatility [120,182,188,250,990,1203,1233]. Such experiments can also be used

to determine DNA force constants, such as torsional rigidity [182]. It has been

found that DNA subjected to previously unattainable forces of 10–150 picoNew-

tons (pN) displays a highly cooperative, sharp, and reversible transition to a new

DNA structure at around 70 pN. This new overstretched helical structure, termed

the S-DNA ladder, has been suggested to have a rise per residue of around 5.8

(compare to values in Table 5.3 of Chapter 5) and a notable inclination (possibly

as large as 70

)[671], but structural details remain controversial. The transition

under a force of extension is also affected by changes in the ionic strength of the

medium, sequence, and addition of intercalators. Interestingly, DNA can sustain

extensions to roughly twice its original length without severe distortions to the

base pairing.

Modeling studies soon followed these experiments to investigate global con-

formational features of overstretched DNA duplexes [671, 711] and the local

morphological features associated with the deformation process [678](see

Box 7.2).

Single-force spectroscopy has also been applied to study the mechanical prop-

erties of nucleosome arrays of the chromatin ﬁber (see Section 6.5), to suggest

chromatin organization and to examine the effect of linker histones on ﬁber

compaction [682].

Similar experiments have also been applied to RNA [759, 771, 1452]and

proteins [1453].

Biological Relevance and Other Applications

The extreme deformations of DNA are of great interest because ﬂexible DNA

molecules undergo a variety of distortions in biological environments. Notable

examples occur during DNA recombination and cell division. It has been sug-

gested, for example, that a helix to ladder transition in DNA near the chromosomal

centromere region may occur during cell division and thus play a regulatory

role [671]. Intriguingly, the overstretched DNA in recombination ﬁlaments of

5.1

A per bp is close to that associated with the phase transition observed in

single-molecular micro-manipulation experiments.

7.2. Variations on a Theme 215

It has also been suggested that longitudinal deformations of DNA might be

associated with many DNA/protein binding events, such as DNA binding to the

TATA-box binding protein TBP (where DNA is locally compressed, strongly bent,

and unwound) and DNA binding to nucleosomes [1050], where variable lengths

of DNA associated with nucleosome wrapping might accommodate sequence and

ionic variations [678].

More generally, single-molecule biochemistry experiments help investigate the

energetics and dynamics of folding and unfolding, the stability of various 2D and

3D intermediates, conformational landscapes, and folding pathways.

Box 7.2: Modeling Overstretched DNA

Results of modeling studies of overstretched DNA — by energy minimization or molec-

ular dynamics — are protocol dependent. For example, results and interpretations depend

on which end of the DNA is pulled, what minimization method is used, how minimization

is implemented (e.g., how ﬁne the helical rise increments are), what force ﬁeld is em-

ployed, and how solvent is treated.

The minimization work of Lavery and coworkers [711] suggested that the stretched con-

formation may be a ﬂat, unwound duplex or a narrow ﬁber with substantial bp inclination

[711]. The molecular dynamics simulations of Konrad and Bolonick [671] reproduced the

helix to ladder transition and analyzed the geometric and energetic properties stabilizing

the S-ladder. The constrained minimization studies of Olson, Zhurkin and coworkers [678]

produced a wealth of structural analyses for both compressed and stretched DNA duplexes

(from 2 to 7

A per bp) of poly(dA)·poly(dT) and poly(dG)·poly(dC) homopolymers under

high and low salt conditions. It was found that DNA can stretch to about double, and

compress to half, its length before the internal energy rises sharply. Energy proﬁles span-

ning four families of right-handed structures revealed that DNA extension/compression

deformations can be related to concerted changes in rise, twist, roll, and slide parameters.

The lowest energy conﬁgurations correspond to canonical A and B-DNA. These models

may be relevant to nucleoprotein ﬁlaments between bacterial Rec-A-like proteins and

overstretched, undertwisted DNA [349].

Such fascinating single-molecule biochemistry manipulations have also shown that un-

der much smaller forces (<3 pN), a new DNA phase is achieved, with about 2.6 bps/turn

and thus about 75% more extended than B-DNA [21]. This new DNA conformation,

termed P-DNA (for Pauling, see below), occurs at moderate positive supercoiling for

molecules that cannot relieve the torsional stress via writhing (nonplanar bending). (For

negative supercoiling, DNA denatures in this force range). Intriguingly, some modeling

suggests that such a structure is ‘inside-out’, with bases on the helical exterior and the

sugar/phosphate backbone at the center, as Linus Pauling once suggested for the structure

of the double helix based on a model with un-ionized bases [975]. Such a conformational

transition arrives from major rotation of torsion angles (notably α and γ,towardthetrans

conformation) and is compatible with both C2



-endo and C3



-endo sugar puckers. The

intrastrand P–P distances in P-DNA (around 7.5

A) are larger than the corresponding

values in canonical A and B-DNA (5.8 and 6.6

A, respectively). Evidence shows that this

P-DNA conformation is found in the packed DNA inside some virus complexes, where

216 7. Topics in Nucleic Acids Structure: Noncanonical Helices and RNA Structure

the DNA is constrained by the helical coat protein [289, 775]. Still, the interpretation of

positively-supercoiled, overstretched DNA as an ‘inside-out’ model remains controversial.

Polymer statistical-mechanics calculations suggest that the experimental data involved

in these force-versus-extension measurements of DNA can be ﬁt to the elastic theory of

the entropic force required to extend a worm-like chain [190]. However, such studies

of entropic force are only relevant to small ﬂuctuations about the B-DNA equilibrium

conformation.

7.3 RNA Structure and Function

7.3.1 DNA’s Cousin Shines

While proteins are household words and DNA is an icon, in science as well as

art (for the latter, see [1112] for an overview), their biomolecular cousin, RNA,

was largely left behind until recently. Indeed, RNA’s starring role in the cell has

emerged with new discoveries concerning RNA’s vital regulatory roles, as in-

troduced in Chapter 1. Namely, our appreciation for RNA has heightened with

discoveries that RNA molecules are integral components of the cellular machin-

ery for protein synthesis and transport, RNA editing, chromosome replication and

regulation, catalysis, and many other functions (see Table 7.1 for some of RNA’s

diverse roles).

At a 2006 symposium organized by NCI at Cold Spring Harbor, Susan

Gottesman suggested on a presentation slide: “Anything that DNA can do, RNA

candobetter”. At the same symposium, Gary Ruvkun announced 23 challenges

facing scientists concerning regulatory RNAs. James Watson simply summed the

atmosphere: “This is a revolution”. Indeed, we are now discovering that bio-

logical and synthetic RNAs prepared in the laboratory perform a broad range

of functions and have numerous applications. This comes with the discovery not

only of regulatory RNAs but also of many synthetic RNAs developed from in vitro

selection experiments that have signiﬁcantly expanded our knowledge of RNA’s

repertoire.

7.3.2 RNA Chains Fold Upon Themselves

Many of the base pairing variations described for DNA in the prior sections

are common for RNA, the single-stranded polynucleotide chain, which can fold

upon itself to form double-stranded segments interspersed with loops. Such pat-

terns are governed by the primary sequence and can accommodate a variety of

hydrogen bonding patterns. The double stranded regions (stems) can be imper-

fect with bulges, mismatched pairs, and unusual hydrogen bond schemes, as

shown in Figure 7.4. The stem/loop structures themselves can fold further, seeking

7.3. RNA Structure and Function 217

Table 7.1. Some classes of non-coding RNA (ncRNA).

RNA Function

transfer RNA (tRNA) protein synthesis

ribosomal RNA (rRNA) protein synthesis

Signal recognition particle (SRP) protein recognition

small nucleolar RNA (snoRNA) rRNA modiﬁcation

micro RNA (miRNA) translation regulation

transfer-messenger RNA (tmRNA) protein stability in ribosome

telomerase RNA replication

guide RNA (gRNA) mRNA editing

spliced leader RNA (SL RNA) mRNA trans-splicing

small nuclear RNA (snRNA) RNA splicing

hairpin, hammerhead, and HDV ribozymes self-cleavage

Group I intron self-splicing

Group II intron self-splicing

RNase P pre-tRNA processing

23S rRNA peptide bond formation

G, A, glmS, TPP and other riboswitches gene regulation

favorable stacking, hydrogen bonding, and other interactions between distant part-

ners. See [1113] for an introduction to RNA structure and associated biological

problems.

The multitude of RNA secondary structures and tertiary interactions are sta-

bilized by various double-stranded regions, interior and terminal loops, bulges,

K-turns, U-turns, S-turns, A-platforms, tetraloops, and other motifs (e.g., [102,

545,654,736,739,740,875]).

RNA pseudoknot motifs, which can be considered supersecondary structural

elements, are produced from a special intertwining of secondary structural ele-

ments by Watson-Crick base pairing (see Figure 7.5) Namely, pseudoknots form

when a consecutive single-stranded domain with segments a,a



, b,b



, c,c



and

d (a



,andc



are connectors) folds to form two hydrogen bonds: a with c,and

b with d. This is a pseudoknot rather than a knot since the strands do not actu-

ally pass through one another [1367]. See Figure 7.6 for examples of RNAs with

pseudoknots.

7.3.3 RNA’s Diversity

The wonderful capacity of RNA to form complex, stable tertiary structures has

been exploited by evolution. RNA molecules are integral components of the cel-

lular machinery for protein synthesis and transport, RNA editing, chromosome

218 7. Topics in Nucleic Acids Structure: Noncanonical Helices and RNA Structure

Stem I

Stem II

Stem III

Hammerhead

ribozyme

AUG

Stem I

Stem II

Stem III

’

Hairpin:

SL3 element of

Ψ site of

human immuno−

deficiency virus

type−1

’

Cruciform:

4−way branch

junction

AGC

’

Bulge loops

& Hairpin:

hairpin ribozyme

of tobacco ringspot

satellite virus

Bulge

loops

Hairpin

’

UUU UC

GAA

’

Loops, Bulge Loops

& Hairpins:

P4−P6 domain of

the group I intron of

Tetrahymena

thermophila

Loops

Tetraloop

Receptor

Tetraloop

P6b

P6a

P5a

P5c

P5b

Figure 7.4. Various nucleotide-chain folding motifs from solved RNA systems: loops,

bulge loops, hairpins, hairpin loops, cruciforms (four-way junctions, found in recombi-

nation intermediates and synthetic designs), and the complex tertiary contacts found in

RNA (illustrated on the hammerhead ribozyme), which compact the secondary-structure

elements. Broken connections are used for non-WC bps, including mismatches. The PDB

ﬁles for the structures (clockwise from left) are as follows: tetrahymena intron, 1GID; to-

bacco ringspot ribozyme, 1B36; HIV-1 Ψ-stem loop 3, 1A1T; and hammerhead ribozyme,

1MME. The tertiary interactions shown for the hammerhead drawing (ellipses or connect-

ing line segments) are based on [1154]. The nucleotide C17 shown in space-ﬁlling form is

the strand cleavage site of the ribozyme.

7.3. RNA Structure and Function 219

Figure 7.5. RNA pseudoknots have an intertwined form of base pairing, which can be

evident from a circular representation of base pairing.

replication and regulation, catalysis and many other functions

. Indeed, some

scientists hypothesize that life was based on RNA (‘RNA World’) before the

evolution of the modern nucleoprotein universe [454, 906]. RNA/DNA hybrid

double helices also occur in transcription of RNA onto DNA templates and in

the initiation of DNA replication by short RNA segments.

As mentioned in Chapter 1, a highlight of RNA’s functional diversity is its

catalytic capability, as discovered in the 1980s and recognized in the 1989 No-

bel Prize in Chemistry to Thomas Cech and Sidney Altman. Such catalysis is

performed by RNA molecules termed ribozymes; hundreds of ribozyme types

are now known in a diverse range of organisms (e.g., [1142, 1208, 1251, 1275],

and some have been designed in the laboratory by in vitro selection (e.g., [1228,

1245]), as spare in composition as two base building-block units (rather than four)

[1042].

Many ribozymes make or break phosphodiester bonds in nucleic acid

backbones, but other biological and chemical functions are continuously being

discovered.

RNA’s conformational ﬂexibility, modularity, and versatility [493] are impor-

tant for recognition by, and interactions with, ligands (e.g., as in the RNA of the

human immunodeﬁciency virus, HIV) and other molecules. These features are

of practical importance for the design of therapeutic agents that exploit RNA’s

functional sites as potential drug targets (e.g., [76,211,539,546,725,981,1268]).

The construction of ligands that bind RNA and interfere with protein synthesis,

transcription, or viral replication have potential as new therapeutic agents, such

as antibiotic/antiviral drugs [984]. These pharmaceuticals are urgent given the

resurgence of previously known viral and bacterial diseases and the emergence of

resistant mutants of common antibiotic and antiviral drugs.

For example, tRNA molecules carry amino acids and deposit them in correct order, mRNAs

translate hereditary information from DNA into protein, rRNA are involved in protein biosynthesis

(within a complex of ribosomal RNA and numerous proteins), gRNAs edit RNA messages (by ‘guide’

sequences), cRNAs play roles in catalysis and autocatalysis, and snRNAs are splicing agents (by

‘small nuclear’ components of the mRNA splicing machinery known as the splicosome).

The 83-nucleotide ribozyme composed only of two different building blocks — uracil and 2,6-

diaminopurine — was shown to catalyze the ligation of two RNA molecules with a rate 36,000 times

faster than the uncatalyzed reaction [1042]. The fact that RNA’s genetic code may be simpler than

today’s four bases lends further support to the “RNA world” hypothesis.